From b87a03cc0838e2b92ceccef3ac7f3a77287e4254 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 1 Feb 2024 13:52:17 -0600 Subject: [PATCH 001/590] Track additional metrics with W&B in `megatron/training.py` --- megatron/training.py | 25 +++++++++++++++++++++++-- 1 file changed, 23 insertions(+), 2 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index c7ac2574fb7..29d2181d463 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1039,7 +1039,8 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, tokens_per_gpu_per_second = tokens_per_sec / args.world_size tokens_per_gpu_per_second_per_replica = tokens_per_gpu_per_second / args.data_parallel_size if wandb is not None and getattr(wandb, 'run', None) is not None: - tput = { + assert wandb.run is not None + wandb_metrics = { 'throughput/iteration-time': elapsed_time_per_iteration, # 1000 ms / s 'throughput/samples_per_sec': samples_per_sec, 'throughput/samples_per_sec_per_replica': samples_per_sec_per_replica, @@ -1050,8 +1051,13 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, 'throughput/tflops': tflops, 'throughput/approx_params_in_billions': approx_parameters_in_billions, 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, + 'throughput/iteration': iteration, } - wandb.run.log(tput) + if loss_dict is not None: + wandb_metrics |= { + f'loss/{k}': v for k, v in loss_dict.items() + } + wandb_metrics |= {'loss/iteration': iteration} if writer: if args.log_timers_to_tensorboard: writer.add_scalar('iteration-time/iteration-time', @@ -1060,6 +1066,21 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration, args.consumed_train_samples) writer.add_scalar('iteration-time/iteration-time vs tokens', elapsed_time_per_iteration, args.consumed_train_tokens) + if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb_metrics |= { + 'iteration': iteration, + 'iteration_time': elapsed_time_per_iteration, + 'iteration_time_vs_tokens': ( + (elapsed_time_per_iteration + / args.consumed_train_tokens) + ), + 'iteration_time_vs_samples': ( + (elapsed_time_per_iteration + / args.consumed_train_samples), + ), + } + if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb.log(wandb_metrics) log_string = ' iteration {:8d}/{:8d} |'.format( iteration, args.train_iters) log_string += ' consumed samples: {:12d} |'.format( From 1302c61795e60c14de4ec0a5ab3ea1c3214cf45f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 21 Feb 2024 08:27:08 -0600 Subject: [PATCH 002/590] Update `megatron/training.py` --- megatron/training.py | 32 +++++++++++++++++--------------- 1 file changed, 17 insertions(+), 15 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 29d2181d463..ef32cd38566 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1066,21 +1066,6 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration, args.consumed_train_samples) writer.add_scalar('iteration-time/iteration-time vs tokens', elapsed_time_per_iteration, args.consumed_train_tokens) - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb_metrics |= { - 'iteration': iteration, - 'iteration_time': elapsed_time_per_iteration, - 'iteration_time_vs_tokens': ( - (elapsed_time_per_iteration - / args.consumed_train_tokens) - ), - 'iteration_time_vs_samples': ( - (elapsed_time_per_iteration - / args.consumed_train_samples), - ), - } - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb.log(wandb_metrics) log_string = ' iteration {:8d}/{:8d} |'.format( iteration, args.train_iters) log_string += ' consumed samples: {:12d} |'.format( @@ -1091,6 +1076,21 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration * 1000.0) log_string += ' learning rate: {:.3E} |'.format(learning_rate) log_string += ' global batch size: {:5d} |'.format(batch_size) + if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb_metrics |= { + 'training/iteration': iteration, + 'training/iteration_time': elapsed_time_per_iteration, + 'training/iteration_time_vs_tokens': ( + (elapsed_time_per_iteration + / args.consumed_train_tokens) + ), + 'training/iteration_time_vs_samples': ( + (elapsed_time_per_iteration + / args.consumed_train_samples), + ), + 'training/consumed_samples': args.consumed_train_samples, + 'training/consumed_tokens': args.consumed_train_tokens, + } for key in total_loss_dict: if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: @@ -1099,6 +1099,8 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, if avg > 0.0: log_string += ' {}: {:.6E} |'.format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) + if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb.log(wandb_metrics) if loss_scale is not None: log_string += ' loss scale: {:.1f} |'.format(loss_scale) if grad_norm is not None: From de60d86cf2928f43b894c3825d8ee90872cdd549 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 26 Feb 2024 08:22:46 -0600 Subject: [PATCH 003/590] Remove `assert num_datasets < 255` in `megatron/data/blendable_dataset.py` --- megatron/data/blendable_dataset.py | 54 ++++++++++++++++++++++-------- 1 file changed, 40 insertions(+), 14 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 2516e584153..590375a0f28 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -13,16 +13,39 @@ from megatron import print_rank_0 from megatron.core import mpu -class BlendableDataset(torch.utils.data.Dataset): - - def __init__(self, datasets, weights, size, *, - data_cache_path=None): +class BlendableDataset(torch.utils.data.Dataset): + def __init__( + self, + datasets, + weights, + size, + *, + data_cache_path=None + ): self.datasets = datasets num_datasets = len(datasets) + ndsets = np.array([num_datasets], dtype=np.int64).item() + len_weights = np.array([len(weights)], dtype=np.int64).item() + print_rank_0(f'{len(datasets)=}') + print_rank_0(f'{len(weights)}') + print_rank_0(f'{ndsets=}') + print_rank_0(f'{len_weights=}') + # if int(num_datasets) != int(len(weights)): + # if f'{num_datasets}' == f'{len(weights)}': + if ndsets != len_weights: + warr = np.array(weights) + print_rank_0('\n'.join([ + f'{num_datasets=}', + f'{len(weights)=}', + f'{warr.shape=}', + f'{warr.sum()=}', + # f'{num_datasets=} != {len(warr)=}', + ])) assert num_datasets == len(weights) - + # else: + # raise IndexError(f'{num_datasets=} != {len(weights)=}') self.size = size # Normalize weights. @@ -34,8 +57,8 @@ def __init__(self, datasets, weights, size, *, # Build indicies. def _build_indices(): start_time = time.time() - assert num_datasets < 255 - dataset_index = np.zeros(self.size, dtype=np.uint8) + # assert num_datasets < 255 + dataset_index = np.zeros(self.size, dtype=np.int64) dataset_sample_index = np.zeros(self.size, dtype=np.int64) from megatron.data import helpers @@ -92,16 +115,21 @@ def _build_indices(): # Load on all ranks. print_rank_0(f'> loading blendable dataset index: {index_path}') - self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') + self.dataset_index = np.load( + index_path, + allow_pickle=True, + mmap_mode='r' + ) assert self.dataset_index.size == self.size - print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') + print_rank_0( + f'> loading blendable dataset sample index: ' + f'{sample_index_path}' + ) self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size else: self.dataset_index, self.dataset_sample_index = _build_indices() - - # Check size _ = self.__getitem__(self.size - 1) try: @@ -112,15 +140,13 @@ def _build_indices(): print_rank_0('> size of blendable dataset: ' '{} samples'.format(self.size)) - def __len__(self): return self.size - def __getitem__(self, idx): dataset_idx = self.dataset_index[idx] sample_idx = self.dataset_sample_index[idx] return { - "dataset_idx" : dataset_idx, + "dataset_idx": dataset_idx, **self.datasets[dataset_idx][sample_idx], } From c11589da93bc109aa98be19434ee9d984783d424 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 26 Feb 2024 08:25:50 -0600 Subject: [PATCH 004/590] [format] `megatron/data/indexed_dataset.py` --- megatron/data/indexed_dataset.py | 98 ++++++++++++++++++++++++-------- 1 file changed, 73 insertions(+), 25 deletions(-) diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index 219ffe8031c..efd3310af0e 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -47,21 +47,32 @@ def infer_dataset_impl(path): return None else: print(f"Dataset does not exist: {path}") - print("Path should be a basename that both .idx and .bin can be appended to get full filenames.") + print( + "Path should be a basename that both .idx and .bin can be " + "appended to get full filenames." + ) return None def make_builder(out_file, impl, vocab_size=None): if impl == 'mmap': - return MMapIndexedDatasetBuilder(out_file, dtype=__best_fitting_dtype(vocab_size)) + return MMapIndexedDatasetBuilder( + out_file, + dtype=__best_fitting_dtype(vocab_size) + ) else: return IndexedDatasetBuilder(out_file) def make_dataset(path, impl, skip_warmup=False): if not IndexedDataset.exists(path): - print(f"Dataset does not exist: {path}") - print("Path should be a basename that both .idx and .bin can be appended to get full filenames.") + print( + f"Dataset does not exist: {path}" + ) + print( + "Path should be a basename that both .idx and .bin " + "can be appended to get full filenames." + ) return None if impl == 'infer': impl = infer_dataset_impl(path) @@ -167,13 +178,15 @@ def __del__(self): self.data_file.close() # @lru_cache(maxsize=8) - def __getitem__(self, idx): + def __getitem__(self, idx) -> np.ndarray: if not self.data_file: self.read_data(self.path) if isinstance(idx, int): i = idx self.check_index(i) - tensor_size = self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + tensor_size = ( + self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + ) a = np.empty(tensor_size, dtype=self.dtype) self.data_file.seek(self.data_offsets[i] * self.element_size) self.data_file.readinto(a) @@ -181,15 +194,16 @@ def __getitem__(self, idx): elif isinstance(idx, slice): start, stop, step = idx.indices(len(self)) if step != 1: - raise ValueError("Slices into indexed_dataset must be contiguous") + raise ValueError( + "Slices into indexed_dataset must be contiguous" + ) sizes = self.sizes[self.dim_offsets[start]:self.dim_offsets[stop]] size = sum(sizes) a = np.empty(size, dtype=self.dtype) self.data_file.seek(self.data_offsets[start] * self.element_size) self.data_file.readinto(a) offsets = list(accumulate(sizes)) - sents = np.split(a, offsets[:-1]) - return sents + return np.split(a, offsets[:-1]) def __len__(self): return self._len @@ -203,7 +217,8 @@ def size(self, index): @staticmethod def exists(path): return ( - os.path.exists(index_file_path(path)) and os.path.exists(data_file_path(path)) + os.path.exists(index_file_path(path)) + and os.path.exists(data_file_path(path)) ) @property @@ -251,7 +266,9 @@ def __getitem__(self, idx): if isinstance(idx, int): i = idx self.check_index(i) - tensor_size = self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + tensor_size = ( + self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + ) a = np.empty(tensor_size, dtype=self.dtype) ptx = self.cache_index[i] np.copyto(a, self.cache[ptx: ptx + a.size]) @@ -286,7 +303,9 @@ def __init__(self, out_file, dtype=np.int32): def add_item(self, tensor): bytes = self.out_file.write(np.array(tensor.numpy(), dtype=self.dtype)) - self.data_offsets.append(self.data_offsets[-1] + bytes / self.element_size) + self.data_offsets.append( + self.data_offsets[-1] + bytes / self.element_size + ) for s in tensor.size(): self.sizes.append(s) self.dim_offsets.append(self.dim_offsets[-1] + len(tensor.size())) @@ -325,7 +344,9 @@ def finalize(self, index_file): index.write(b'TNTIDX\x00\x00') index.write(struct.pack(' Date: Mon, 26 Feb 2024 08:26:33 -0600 Subject: [PATCH 005/590] Add debug logic (+ formatting fixes) in `megatron/data/gpt_dataset.py` --- megatron/data/gpt_dataset.py | 204 +++++++++++++++++++++++++---------- 1 file changed, 149 insertions(+), 55 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 1d9b7e1c1d9..f9cf5ee6f75 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -32,16 +32,23 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, # Single dataset. if len(data_prefix) == 1: - return _build_train_valid_test_datasets(data_prefix[0], - data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + return _build_train_valid_test_datasets( + data_prefix[0], + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path + ) # Blending dataset. # Parse the values. - output = get_datasets_weights_and_num_samples(data_prefix, - train_valid_test_num_samples) + output = get_datasets_weights_and_num_samples( + data_prefix, + train_valid_test_num_samples + ) prefixes, weights, datasets_train_valid_test_num_samples = output train_num_samples, valid_num_samples, test_num_samples = map( sum, @@ -69,55 +76,92 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, # Blend. blending_train_dataset = None if train_datasets: - blending_train_dataset = BlendableDataset(train_datasets, weights, train_num_samples, - data_cache_path=data_cache_path) + blending_train_dataset = BlendableDataset( + train_datasets, + weights, + train_num_samples, + data_cache_path=data_cache_path + ) blending_valid_dataset = None if valid_datasets: - blending_valid_dataset = BlendableDataset(valid_datasets, weights, valid_num_samples, - data_cache_path=data_cache_path) + blending_valid_dataset = BlendableDataset( + valid_datasets, + weights, + valid_num_samples, + data_cache_path=data_cache_path + ) blending_test_dataset = None if test_datasets: - blending_test_dataset = BlendableDataset(test_datasets, weights, test_num_samples, - data_cache_path=data_cache_path) + blending_test_dataset = BlendableDataset( + test_datasets, + weights, + test_num_samples, + data_cache_path=data_cache_path + ) return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) else: - print_rank_0("Separate data paths provided for train, valid & test. Split string will be ignored.") + print_rank_0( + "Separate data paths provided for train, valid & test. " + "Split string will be ignored." + ) train_dataset, valid_dataset, test_dataset = None, None, None # Single dataset. if train_data_prefix is not None: - train_dataset = build_dataset("train", train_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[0], - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + train_dataset = build_dataset( + "train", + train_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[0], + seq_length, seed, skip_warmup, + data_cache_path=data_cache_path + ) if valid_data_prefix is not None: - valid_dataset = build_dataset("valid", valid_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[1], - seq_length, seed, False, - data_cache_path=data_cache_path) - + valid_dataset = build_dataset( + "valid", + valid_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[1], + seq_length, + seed, + False, + data_cache_path=data_cache_path + ) if test_data_prefix is not None: - test_dataset = build_dataset("test", test_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[2], - seq_length, seed, False, - data_cache_path=data_cache_path) + test_dataset = build_dataset( + "test", + test_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[2], + seq_length, + seed, + False, + data_cache_path=data_cache_path + ) return (train_dataset, valid_dataset, test_dataset) -def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - return_doc_ids=False, *, - data_cache_path=None): +def _build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + return_doc_ids=False, + *, + data_cache_path=None +): """Build train, valid, and test datasets.""" # Indexed dataset. @@ -195,10 +239,18 @@ def build_dataset(dataset_name, data_prefix, data_impl, return dataset -def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - *, - data_cache_path=None): +def _build_dataset( + dataset_name, + data_prefix, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + *, + data_cache_path=None +): """ Build dataset. This method is called when individual train, valid, test datasets are provided @@ -211,18 +263,34 @@ def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, total_num_of_documents = indexed_dataset.sizes.shape[0] - print_rank_0(' {}:'.format(dataset_name)) - print_rank_0(' document indices in [0, {}) total of {} ' - 'documents'.format(total_num_of_documents, total_num_of_documents)) - - documents = np.arange(start=0, stop=total_num_of_documents, - step=1, dtype=np.int32) - - dataset = GPTDataset(dataset_name, data_prefix, documents, indexed_dataset, - splits_string, num_samples, seq_length, seed, - data_cache_path=data_cache_path) - - return dataset + print_rank_0(f' {dataset_name}:') + print_rank_0( + f' ' + f'document indices in [0, {total_num_of_documents}) ' + f'total of {total_num_of_documents} documents' + ) + # 'documents'.format( + # total_num_of_documents, + # total_num_of_documents + # )) + + documents = np.arange( + start=0, + stop=total_num_of_documents, + step=1, + # dtype=np.int32 + ) + return GPTDataset( + dataset_name, + data_prefix, + documents, + indexed_dataset, + splits_string, + num_samples, + seq_length, + seed, + data_cache_path=data_cache_path + ) def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): @@ -273,7 +341,31 @@ def __getitem__(self, idx): args = get_args() orig_idx = idx # Get the shuffled index. - idx = self.shuffle_idx[idx] + try: + idx = self.shuffle_idx[idx] + except IndexError as exc: + if is_rank_0(): + import json + from rich import print_json + print(exc) + print( + '\n'.join( + ['-------------------------------------------------', + f'Trying to access {idx=} from self.shuffle_idx,', + f'but {len(self.shuffle_idx)=}', + '-------------------------------------------------'] + ) + ) + print_json( + json.dumps( + { + 'doc_idx': len(self.doc_idx), + 'sample_idx': len(self.sample_idx), + 'shuffle_idx': len(self.shuffle_idx), + }, + indent=4, + ) + ) # Start and end documents and offsets. doc_index_f = self.sample_idx[idx][0] doc_index_l = self.sample_idx[idx + 1][0] @@ -283,9 +375,11 @@ def __getitem__(self, idx): doc_ids = [] if doc_index_f == doc_index_l: doc_ids.append(self.doc_idx[doc_index_f]) - sample = self.indexed_dataset.get(self.doc_idx[doc_index_f], - offset=offset_f, - length=offset_l - offset_f + 1) + sample = self.indexed_dataset.get( + self.doc_idx[doc_index_f], + offset=offset_f, + length=offset_l - offset_f + 1 + ) else: # Otherwise, get the rest of the initial document. doc_ids.append(self.doc_idx[doc_index_f]) From 1906d31689b81e59a3000362f0bbb0dd8e7d0bb1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 26 Feb 2024 08:27:12 -0600 Subject: [PATCH 006/590] Update `.gitignore` --- .gitignore | 26 +++++++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/.gitignore b/.gitignore index 64270f07524..74f73f1cde3 100644 --- a/.gitignore +++ b/.gitignore @@ -1,3 +1,27 @@ +*.bak +**index-cache** +**pbslogs** +ezpz +*.o17* +*.e17* +*hostfile* +.deepspeed_env +*.DS_Store +old/* +**venv** +*.json +*.o1 +*.e1 +outputs/ +venvs/ +wandb/ +llama-logs/ +checkpoints/ +*.gz +*.txt +*.idx +*.bin +*.log __pycache__ # Distribution / packaging @@ -20,4 +44,4 @@ slurm* logs # Data folder -bookcorpus_data/ \ No newline at end of file +bookcorpus_data/ From ee7ce6fbc82f5020c82df15033fa6dcea710cad4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 26 Feb 2024 13:08:35 -0600 Subject: [PATCH 007/590] Update `.gitignore` --- .gitignore | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/.gitignore b/.gitignore index 74f73f1cde3..c972bfbd5bd 100644 --- a/.gitignore +++ b/.gitignore @@ -1,21 +1,10 @@ -*.bak -**index-cache** -**pbslogs** -ezpz -*.o17* -*.e17* -*hostfile* .deepspeed_env *.DS_Store -old/* **venv** *.json -*.o1 -*.e1 outputs/ venvs/ wandb/ -llama-logs/ checkpoints/ *.gz *.txt From 99d12f4c62b507b5e5f0d05cadcf70c82333b28d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 12 Mar 2024 00:54:28 -0500 Subject: [PATCH 008/590] added support for multiprocessing_context --- megatron/arguments.py | 1 + megatron/data/data_samplers.py | 2 +- train_llama_alcf_polaris_hzheng.sh | 22 +++++++++------------- 3 files changed, 11 insertions(+), 14 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index faeac611bc9..a83c369018d 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1307,6 +1307,7 @@ def _add_data_args(parser): help='Force to use certain index file.') group.add_argument('--repeated-dataloader', action='store_true', help='Once all the data has been loaded, reuse the DataLoader.') + group.add_argument('--multiprocessing-context', type=str, default='fork') return parser diff --git a/megatron/data/data_samplers.py b/megatron/data/data_samplers.py index 8eb2f2a6687..0aae13abce0 100644 --- a/megatron/data/data_samplers.py +++ b/megatron/data/data_samplers.py @@ -46,7 +46,7 @@ def build_pretraining_data_loader(dataset, consumed_samples): batch_sampler=batch_sampler, num_workers=args.num_workers, pin_memory=True, - # multiprocessing_context='spawn' + multiprocessing_context=args.multiprocessing_context ) if args.repeated_dataloader: loader=RepeatingLoader(loader) diff --git a/train_llama_alcf_polaris_hzheng.sh b/train_llama_alcf_polaris_hzheng.sh index 0ca7cb78bbf..1fb5345eda7 100755 --- a/train_llama_alcf_polaris_hzheng.sh +++ b/train_llama_alcf_polaris_hzheng.sh @@ -4,25 +4,21 @@ #PBS -q debug-scaling #PBS -l select=2 #PBS -l filesystems=eagle:grand:home -export PPN=4 -export MD=/home/hzheng/ALCF-Megatron-DeepSpeed -module load conda/2023-10-04 -#conda activate /soft/datascience/megatron-deepspeed/2023-10-04 -conda activate $HOME/PolarisAT/pyenvs/megatron/2023-10-04 cd ${PBS_O_WORKDIR} +export PPN=4 +export MD=/eagle/argonne_tpc/soft/Megatron-DeepSpeed +source /eagle/argonne_tpc/soft/conda.sh export PBS_JOBSIZE=$(cat $PBS_NODEFILE | uniq | wc -l) export TP=1 export PP=1 export MBS=1 export BS=$((MBS*PBS_JOBSIZE*PPN/PP/TP)) export SP=$((PBS_JOBSIZE*PPN/PP/TP)) -#export DATA_PATH="/eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/" - -export export DATE_TAG=$(date +"%Y-%m-%d-%H-%M-%S") -export DATA_PATH="/eagle/datasets//dolma/data_Llama2Tokenizer/wiki-en-simple/" -#export DATA_FILE_LIST="/eagle/datasets//dolma//data_file_list_select.txt" -DATA_FILE_LIST=$PWD/test.txt -echo "BS: $BS\n PP:$PP \n TP: $TP, PBS_JOBSIZE: $PBS_JOBSIZE" +export DATE_TAG=$(date +"%Y-%m-%d-%H-%M-%S") +export DATA_FILE_LIST="/eagle/datasets//dolma/data_file_list_reweighted.txt" +echo "BS: $BS - PP:$PP - TP: $TP, PBS_JOBSIZE: $PBS_JOBSIZE" +# First time running, it will compile the fused kernels, which will take about 10 mins +# >>> done with compiling and loading fused kernels. Compilation time: 545.468 seconds HIDDEN_SIZE=4096 NUM_LAYERS=32 @@ -32,7 +28,7 @@ TRAIN_ITERS=10 ZERO_STAGE=2 MODEL=LLAMA_7B OUTPUT_PREFIX=${MODEL}_z${ZERO_STAGE}_seqlen_mp${MP}_pp${PP}_sp${SP}_nl${NUM_LAYERS}_hs${HIDDEN_SIZE}_gb${BS}_mb${MBS} -MASTER_ADDR=localhost MASTER_PORT=6543 mpiexec -n $((PBS_JOBSIZE*PPN)) -ppn $PPN --cpu-bind depth -d 16 --hostfile $PBS_NODEFILE python3 ./pretrain_gpt_alcf.py \ +MASTER_ADDR=localhost MASTER_PORT=6543 mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN)) -ppn $PPN --cpu-bind depth -d 16 --hostfile $PBS_NODEFILE python3 ./pretrain_gpt_alcf.py \ --tensor-model-parallel-size ${TP} \ --pipeline-model-parallel-size ${PP} \ --num-layers ${NUM_LAYERS} \ From d91b237ba1460556f96c04e22d766b741d190e7d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 12 Mar 2024 01:12:14 -0500 Subject: [PATCH 009/590] changed script to common environement --- ALCF/test_blend_full.sh | 14 +++++++------- ALCF/test_blendable_dataset.py | 26 +++++++++++++++----------- 2 files changed, 22 insertions(+), 18 deletions(-) diff --git a/ALCF/test_blend_full.sh b/ALCF/test_blend_full.sh index 4245304456f..0218383980b 100755 --- a/ALCF/test_blend_full.sh +++ b/ALCF/test_blend_full.sh @@ -6,16 +6,16 @@ #PBS -l filesystems=eagle:grand:home cd ${PBS_O_WORKDIR} export PPN=4 -export MD=/home/hzheng/ALCF-Megatron-DeepSpeed -module load conda/2023-10-04 -#conda activate /soft/datascience/megatron-deepspeed/2023-10-04 -conda activate $HOME/PolarisAT/pyenvs/megatron/2023-10-04 +export MD=/eagle/argonne_tpc/soft/Megatron-DeepSpeed +source /eagle/argonne_tpc/soft/conda.sh +export TRITON_CACHE_DIR=/tmp/.cache/ + export TP=1 export PP=1 export SP=128 export MBS=1 export BS=$((MBS*SP)) -export export DATE_TAG=$(date +"%Y-%m-%d-%H-%M-%S") +export DATE_TAG=$(date +"%Y-%m-%d-%H-%M-%S") export DATA_FILE_LIST="/eagle/datasets//dolma/data_file_list_reweighted.txt" HIDDEN_SIZE=4096 @@ -25,8 +25,9 @@ EMBEDDINGS=2048 TRAIN_ITERS=80797 ZERO_STAGE=2 MODEL=LLAMA_7B +export PBS_JOBSIZE=$(cat $PBS_NODEFILE | uniq | wc -l) OUTPUT_PREFIX=${MODEL}_z${ZERO_STAGE}_seqlen_mp${MP}_pp${PP}_sp${SP}_nl${NUM_LAYERS}_hs${HIDDEN_SIZE}_gb${BS}_mb${MBS} -python3 ALCF_utils/test_blendable_dataset.py \ +mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN)) --ppn $PPN --cpu-bind depth -d 16 python3 ALCF/test_blendable_dataset.py \ --tensor-model-parallel-size ${TP} \ --pipeline-model-parallel-size ${PP} \ --num-layers ${NUM_LAYERS} \ @@ -51,7 +52,6 @@ python3 ALCF_utils/test_blendable_dataset.py \ --lr-warmup-iters 2 \ --optimizer adam \ --adam-beta1 0.9 \ - --mmap_warmup False \ --adam-beta2 0.95 \ --log-interval 1 \ --cpu-optimizer \ diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index a0cccbb6cb3..503f499fab8 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -5,12 +5,14 @@ from megatron.arguments import parse_args from megatron.initialize import initialize_megatron from megatron.data.data_samplers import build_pretraining_data_loader - +from mpi4py import MPI +comm = MPI.COMM_WORLD initialize_megatron(allow_no_cuda=True) args = get_args() data_file_list = args.data_file_list -print(f"Reading data from {args.data_file_list}") +if comm.rank==0: + print(f"Reading data from {args.data_file_list}") files = [] weights = [] flist = [] @@ -28,10 +30,10 @@ num_samples = args.global_batch_size*args.train_iters num_datasets = len(weights) - -print(f"Number of datasets: {num_datasets}") -print(f"Global batch size: {args.global_batch_size}") -print(f"Training iterations: {args.train_iters}") +if comm.rank==0: + print(f"Number of datasets: {num_datasets}") + print(f"Global batch size: {args.global_batch_size}") + print(f"Training iterations: {args.train_iters}") train_valid_test_num_samples = [num_samples, 0, 0] seed=args.seed data_impl = args.data_impl @@ -48,21 +50,23 @@ ratio_select=np.zeros(num_datasets) #for i in range(num_datasets): # ratio_select[i] = np.sum([i==d for d in dataset_idx])/num_samples - -print(f"Total number of samples: {len(train_ds)}") -print(f"Weights set: {weights[:min(8, num_datasets)]}") +if comm.rank ==0: + print(f"Total number of samples: {len(train_ds)}") + print(f"Weights set: {weights[:min(8, num_datasets)]}") #print(f"Weights across training: {ratio_select[:min(8, num_datasets)]}") for e in range(min(100, args.train_iters)): ratio_select=np.zeros(num_datasets) for i in range(num_datasets): ratio_select[i] = np.sum([i==d for d in dataset_idx[e*args.global_batch_size:(e+1)*args.global_batch_size]])/args.global_batch_size - print(f"iter-{e}: {ratio_select[:min(8, num_datasets)]}") + if comm.rank==0: + print(f"iter-{e}: {ratio_select[:min(8, num_datasets)]}") print("First 10 samples") for i in range(10): - print(f"Sample: {i} \t dataset_idx: {train_ds.dataset_index[i]}, sample_idx: {train_ds.dataset_sample_index[i]}") + if comm.rank==0: + print(f"Sample: {i} \t dataset_idx: {train_ds.dataset_index[i]}, sample_idx: {train_ds.dataset_sample_index[i]}") #### Build data loaders From 15d422e306d97c6aab6cd0e4e42383a2039604b5 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 14 Mar 2024 19:20:00 -0500 Subject: [PATCH 010/590] fixing some NCCL issue and updated the script with the common environment --- ALCF/test_blend_full.sh | 4 ++-- ALCF/test_blendable_dataset.py | 11 +++++++---- megatron/data/gpt_dataset.py | 6 ++++-- pretrain_gpt_alcf.py | 3 ++- train_llama_alcf_polaris_hzheng.sh | 7 +++++-- 5 files changed, 20 insertions(+), 11 deletions(-) diff --git a/ALCF/test_blend_full.sh b/ALCF/test_blend_full.sh index 0218383980b..459652a2ee8 100755 --- a/ALCF/test_blend_full.sh +++ b/ALCF/test_blend_full.sh @@ -27,7 +27,7 @@ ZERO_STAGE=2 MODEL=LLAMA_7B export PBS_JOBSIZE=$(cat $PBS_NODEFILE | uniq | wc -l) OUTPUT_PREFIX=${MODEL}_z${ZERO_STAGE}_seqlen_mp${MP}_pp${PP}_sp${SP}_nl${NUM_LAYERS}_hs${HIDDEN_SIZE}_gb${BS}_mb${MBS} -mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN)) --ppn $PPN --cpu-bind depth -d 16 python3 ALCF/test_blendable_dataset.py \ +APRUN_PMI=pmix aprun -n $((PBS_JOBSIZE*PPN)) -N $PPN --cc depth -d 16 ${MD}/local_rank.sh python3 ALCF/test_blendable_dataset.py \ --tensor-model-parallel-size ${TP} \ --pipeline-model-parallel-size ${PP} \ --num-layers ${NUM_LAYERS} \ @@ -70,4 +70,4 @@ mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN)) --ppn $PPN --cpu-bind depth -d 16 py --data-path ${DATA_PATH} \ --data-cache-path /tmp/hzheng-megatron-deepspeed-cache/ \ --vocab-file ${MD}/dataset/gpt2-vocab.json --merge-file ${MD}/dataset/gpt2-merges.txt \ - --zero-stage=${ZERO_STAGE} --deepspeed_config=${MD}/ds_config-gpt.json --deepspeed + --zero-stage=${ZERO_STAGE} --deepspeed_config=${MD}/ds_config-gpt.json --deepspeed diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 503f499fab8..96811982510 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -6,6 +6,7 @@ from megatron.initialize import initialize_megatron from megatron.data.data_samplers import build_pretraining_data_loader from mpi4py import MPI +from megatron.core import mpu comm = MPI.COMM_WORLD initialize_megatron(allow_no_cuda=True) args = get_args() @@ -68,10 +69,12 @@ if comm.rank==0: print(f"Sample: {i} \t dataset_idx: {train_ds.dataset_index[i]}, sample_idx: {train_ds.dataset_sample_index[i]}") - #### Build data loaders -train_dataloader = build_pretraining_data_loader( +rank_in_parallel_group = mpu.get_sequence_parallel_rank() +print(rank_in_parallel_group) +if rank_in_parallel_group == 0: + train_dataloader = build_pretraining_data_loader( train_ds, args.consumed_train_samples) -valid_dataloader = build_pretraining_data_loader( + valid_dataloader = build_pretraining_data_loader( valid_ds, args.consumed_valid_samples) -test_dataloader = build_pretraining_data_loader(test_ds, 0) + test_dataloader = build_pretraining_data_loader(test_ds, 0) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 210a92c85e8..84ccff9a8b5 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -504,8 +504,10 @@ def _build_index_mappings(name, data_prefix, documents, sizes, data_cache_success = False counts = get_accelerator().LongTensor([data_cache_success]) - torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + if mpu.get_data_parallel_world_size() > 1: + torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) + if mpu.get_pipeline_model_parallel_world_size() > 1: + torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) if counts[0].item() != ( torch.distributed.get_world_size() // torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 4fefef795fe..fae3309d5cf 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -466,6 +466,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): 'for GPT ...') files = [] if args.data_file_list is not None: + print_rank_0(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: for f in flist.readlines(): w, fname = f.split() @@ -479,7 +480,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): files.append(path + f.split(".bin")[0]) else: files = args.data_path - print_rank_0(f"file list {files}") + train_ds, valid_ds, test_ds = build_train_valid_test_datasets( data_prefix=files, data_impl=args.data_impl, diff --git a/train_llama_alcf_polaris_hzheng.sh b/train_llama_alcf_polaris_hzheng.sh index 1fb5345eda7..83d8a2c5a7e 100755 --- a/train_llama_alcf_polaris_hzheng.sh +++ b/train_llama_alcf_polaris_hzheng.sh @@ -8,6 +8,7 @@ cd ${PBS_O_WORKDIR} export PPN=4 export MD=/eagle/argonne_tpc/soft/Megatron-DeepSpeed source /eagle/argonne_tpc/soft/conda.sh + export PBS_JOBSIZE=$(cat $PBS_NODEFILE | uniq | wc -l) export TP=1 export PP=1 @@ -27,8 +28,9 @@ EMBEDDINGS=2048 TRAIN_ITERS=10 ZERO_STAGE=2 MODEL=LLAMA_7B +#LAUNCHER="//eagle/argonne_tpc/soft/Megatron-DeepSpeed/..//conda/2024-03-11/lib/python3.10/site-packages/deepspeed/launcher/launcher_helper.py --launcher mpich " OUTPUT_PREFIX=${MODEL}_z${ZERO_STAGE}_seqlen_mp${MP}_pp${PP}_sp${SP}_nl${NUM_LAYERS}_hs${HIDDEN_SIZE}_gb${BS}_mb${MBS} -MASTER_ADDR=localhost MASTER_PORT=6543 mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN)) -ppn $PPN --cpu-bind depth -d 16 --hostfile $PBS_NODEFILE python3 ./pretrain_gpt_alcf.py \ +APRUN_PMI=pmix aprun -n $((PBS_JOBSIZE*PPN)) -N $PPN --cc depth -d 16 /eagle/argonne_tpc/soft/Megatron-DeepSpeed/local_rank.sh python3 $LAUNCHER ./pretrain_gpt_alcf.py \ --tensor-model-parallel-size ${TP} \ --pipeline-model-parallel-size ${PP} \ --num-layers ${NUM_LAYERS} \ @@ -70,4 +72,5 @@ MASTER_ADDR=localhost MASTER_PORT=6543 mpiexec --pmi=pmix -n $((PBS_JOBSIZE*PPN) --data-file-list ${DATA_FILE_LIST} \ --data-path ${DATA_PATH} \ --vocab-file ${MD}/dataset/gpt2-vocab.json --merge-file ${MD}/dataset/gpt2-merges.txt \ - --zero-stage=${ZERO_STAGE} --deepspeed_config=${MD}/ds_config-gpt.json --deepspeed + --zero-stage=${ZERO_STAGE} --deepspeed_config=${MD}/ds_config-gpt.json --deepspeed \ + --data-cache-path ./data_cache_path/ From 73dc82a532d4d5b55cdcaf151887f7e867aeab96 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 14 Mar 2024 21:35:53 -0500 Subject: [PATCH 011/590] further allreduce check --- megatron/data/blendable_dataset.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index f3276c68230..0220fa22e78 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -80,8 +80,10 @@ def _build_indices(): counts = get_accelerator().LongTensor([cache_success]) - torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + if mpu.get_data_parallel_world_size() > 1: + torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) + if mpu.get_pipeline_model_parallel_world_size() > 1: + torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) if counts[0].item() != ( torch.distributed.get_world_size() // torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // From 4fe87d0f71708404a437684593442607558e4bb4 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 14 Mar 2024 23:41:38 -0500 Subject: [PATCH 012/590] Update gpt_dataset.py --- megatron/data/gpt_dataset.py | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 84ccff9a8b5..210a92c85e8 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -504,10 +504,8 @@ def _build_index_mappings(name, data_prefix, documents, sizes, data_cache_success = False counts = get_accelerator().LongTensor([data_cache_success]) - if mpu.get_data_parallel_world_size() > 1: - torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - if mpu.get_pipeline_model_parallel_world_size() > 1: - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) + torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) if counts[0].item() != ( torch.distributed.get_world_size() // torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // From 39be39565aceb16dbb8b2e04d41b45d183cfee71 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 14 Mar 2024 23:47:58 -0500 Subject: [PATCH 013/590] Update blendable_dataset.py --- megatron/data/blendable_dataset.py | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 0220fa22e78..f3276c68230 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -80,10 +80,8 @@ def _build_indices(): counts = get_accelerator().LongTensor([cache_success]) - if mpu.get_data_parallel_world_size() > 1: - torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - if mpu.get_pipeline_model_parallel_world_size() > 1: - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) + torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) if counts[0].item() != ( torch.distributed.get_world_size() // torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // From 2d46039a7d58188bb66816030c60f2f9ef389fc6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 22 Mar 2024 10:47:41 -0500 Subject: [PATCH 014/590] Add `--train-iters-to-skip` option for skipping backprop on certain train iters --- megatron/arguments.py | 8 +++++--- megatron/core/pipeline_parallel/schedules.py | 11 +++++++++++ 2 files changed, 16 insertions(+), 3 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index d83fe99856a..3b83d16299f 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -959,6 +959,11 @@ def _add_training_args(parser): group.add_argument('--use-dataset-only', type=bool, required=False, default=False, help='If set to True, only use the megatron dataset for external trainer ') group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') + group.add_argument('--train-iters-to-skip', action="extend", nargs="+", type=str, + help=( + "Specific train iterations to skip when training. " + "Load the data and just perform a noop." + )) return parser @@ -1510,7 +1515,6 @@ def _add_activation_checkpoint_args(parser): def _add_distillation_args(parser): group = parser.add_argument_group('Knowledge distillation', 'Distillation Configurations') - group.add_argument('--num-layers-teacher', type=int, default=None, help='Number of the teacher transformer layers.') group.add_argument('--num-experts-teacher', type=int, nargs='+', default=[1,], @@ -1519,7 +1523,6 @@ def _add_distillation_args(parser): help='Tansformer teacher hidden size.') group.add_argument('--num-attention-heads-teacher', type=int, default=None, help='Number of teacher transformer attention heads.') - group.add_argument('--mos', action='store_true', help='Enable Mixture-of-Students via knolwedge distillation.') group.add_argument('--kd', action='store_true', @@ -1529,7 +1532,6 @@ def _add_distillation_args(parser): group.add_argument('--kd-temp', default=1.0, type=float) group.add_argument('--reset-iteration', action='store_true', help='Reset the iteration count.') - group.add_argument('--load-teacher', type=str, default=None, help='Directory containing a teacher model checkpoint.') diff --git a/megatron/core/pipeline_parallel/schedules.py b/megatron/core/pipeline_parallel/schedules.py index 407bb16d569..bb876481acf 100644 --- a/megatron/core/pipeline_parallel/schedules.py +++ b/megatron/core/pipeline_parallel/schedules.py @@ -241,6 +241,17 @@ def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, c # needs to be modified slightly to support arbitrary numbers of skip # connections. args = get_args() + assert args is not None + if config.timers is not None: + config.timers('backward-compute', log_level=2).start() + # if (to_skip := args.train_iters_to_skip) is not None and len(to_skip) > 0: + to_skip = getattr(args, 'train_iters_to_skip', None) + if to_skip is not None: + if config.timers is not None: + config.timers('backward-compute').stop() + if args.iteration in [int(i) for i in to_skip]: + print(f'Caught {args.iteration=} in `iters_to_skip`! Skipping!') + return [None] if args.deepspeed: assert model is not None From d96a58535327ac0ef3bdead29ffd833a0661a631 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 22 Mar 2024 12:23:08 -0500 Subject: [PATCH 015/590] [fix] `megatron/training.py` Fix bug when passing `--train-iters-to-skip` when running **without** setting `CPU_OPTIMIZER=1` --- megatron/training.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 7e6c7dc6bb2..e987eb158df 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -728,8 +728,11 @@ def train_step(forward_step_func, data_iterator, increment = get_num_microbatches() * \ args.micro_batch_size * \ args.data_parallel_size - model[0].step(lr_kwargs={'increment': increment}) - update_successful = model[0].was_step_applied() + try: + model[0].step(lr_kwargs={'increment': increment}) + update_successful = model[0].was_step_applied() + except Exception: + update_successful = False else: update_successful, grad_norm, num_zeros_in_grad = optimizer.step(args, timers) timers('optimizer').stop() From 038109af825d62029f00c9717a2948b2adafa362 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 25 Mar 2024 20:33:52 -0500 Subject: [PATCH 016/590] Add `--optimizer adamw` for `torch.optim.AdamW` --- megatron/arguments.py | 4 +-- megatron/optimizer/__init__.py | 50 ++++++++++++++++++++++------------ 2 files changed, 34 insertions(+), 20 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 3b83d16299f..cd618b285e5 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -800,7 +800,7 @@ def _add_training_args(parser): ' ' ' ' 'For example:' - ' --rampup-batch-size 16 8 300000 \ ' + ' --rampup-batch-size 16 8 300000 \\ ' ' --global-batch-size 1024' 'will start with global batch size 16 and over ' ' (1024 - 16) / 8 = 126 intervals will increase' @@ -914,7 +914,7 @@ def _add_training_args(parser): help='Disable bias in the linear layers', dest='add_bias_linear') group.add_argument('--optimizer', type=str, default='adam', - choices=['adam', 'sgd'], + choices=['adam', 'adamw', 'sgd'], help='Optimizer function') group.add_argument('--dataloader-type', type=str, default=None, choices=['single', 'cyclic'], diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 12a458375da..e7400c39b52 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -1,12 +1,13 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == 'cuda': - from apex.optimizers import FusedAdam as Adam - from apex.optimizers import FusedSGD as SGD -else: - from torch.optim import Adam - from torch.optim import SGD +# if get_accelerator().device_name() == 'cuda': +# from apex.optimizers import FusedAdam as Adam +# from apex.optimizers import FusedSGD as SGD +# else: +# from torch.optim import Adam +# from torch.optim import SGD +import torch from megatron import get_args @@ -93,24 +94,37 @@ def get_megatron_optimizer(model, betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps) else: - if args.optimizer == 'adam': + if str(args.optimizer).lower() == 'adamw': + optimizer = torch.optim.AdamW( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) + elif args.optimizer == 'adam': if args.ds_fused_adam: global Adam from deepspeed.ops.adam import FusedAdam Adam = FusedAdam - optimizer = Adam(param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps) + else: + Adam = torch.optim.Adam + optimizer = Adam( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) elif args.optimizer == 'sgd': - optimizer = SGD(param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - momentum=args.sgd_momentum) + optimizer = torch.optim.SGD( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + momentum=args.sgd_momentum + ) else: - raise Exception('{} optimizer is not supported.'.format( - args.optimizer)) + raise Exception(f'{args.optimizer} optimizer is not supported.') if args.deepspeed: return optimizer From 66401b8e197dbcc649f779119e5cb8f2303ea929 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 08:42:02 -0500 Subject: [PATCH 017/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 4fefef795fe..e296a637be0 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -153,12 +153,14 @@ def model_provider(pre_process=True, post_process=True): see_memory_usage("After Building Model", force=True) if wandb.run is not None: wandb.run.config.update({'num_params': num_params}) - # wandb.run.watch( - # model, - # log='all', - # log_graph=True, - # ) - # wandb.run.config.update({'num_params': num_params}) + try: + wandb.run.watch( + model, + log='all', + log_graph=True, + ) + except Exception: + pass return model From c87ea6aafe82924e4b9b44c066ebcd238572d722 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:22:50 -0500 Subject: [PATCH 018/590] Remove redundant `generate_config_cpu_optimizer.sh` --- generate_config_cpu_optimizer.sh | 151 ------------------------------- 1 file changed, 151 deletions(-) delete mode 100644 generate_config_cpu_optimizer.sh diff --git a/generate_config_cpu_optimizer.sh b/generate_config_cpu_optimizer.sh deleted file mode 100644 index 99dec979581..00000000000 --- a/generate_config_cpu_optimizer.sh +++ /dev/null @@ -1,151 +0,0 @@ -#!/bin/bash --login - -for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ - "$PP" "$DTYPE" -do - if [ -z $v ]; then - echo "Please export required envs before execute $0" - exit 1 - fi -done - -if [ $# -ne 1 ]; then - echo "Usage: $0 config_file" - exit 1 -fi - -extra="" -common="\ - \"train_batch_size\": $GLOBAL_BATCH, - \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, - \"steps_per_print\": 1, - \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, - \"optimizer\": { - \"type\": \"AdamW\", - \"params\": { - \"lr\": ${LR}, - \"beta1\": 0.9, - \"beta2\": 0.95, - \"eps\": 1e-5, - \"weight_decay\": 1e-1 - } - }, - \"scheduler\": { - \"type\": \"WarmupLR\", - \"params\": { - \"warmup_min_lr\": 0.00003, - \"warmup_max_lr\": 0.0003, - \"warmup_num_steps\": 5000 - } - }, - \"zero_allow_untested_optimizer\": true, - \"gradient_clipping\": 1.0, - \"activation_checkpointing\": { - \"partition_activations\": true, - \"contiguous_memory_optimization\": false - }, - \"wall_clock_breakdown\": false," - -flops_profiler="\ - \"flops_profiler\": { - \"enabled\": false, - \"profile_step\": 45, - \"module_depth\": -1, - \"top_modules\": 1, - \"detailed\": true, - \"output_file\": null - }" - -if [[ $DTYPE == "bf16" ]]; then -dtype="\ - \"communication_data_type\": \"bfp16\", - \"fp16\": { - \"enabled\": false, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": true, - \"loss_scale\": 1.0 - }," -else -dtype="\ - \"communication_data_type\": \"fp16\", - \"fp16\": { - \"enabled\": true, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": false, - \"loss_scale\": 1.0 - }," -fi - -if [ $ZERO_STAGE == 3 ]; then -zero="\ - \"zero_optimization\": { - \"stage\": 3, - \"reduce_scatter\": false, - \"stage3_max_live_parameters\": 3e9, - \"stage3_max_reuse_distance\": 3e9, - \"stage3_param_persistence_threshold\": 1e5, - \"stage3_prefetch_bucket_size\": 5e7, - \"contiguous_gradients\": true, - \"overlap_comm\": true, - \"reduce_bucket_size\": 90000000, - \"sub_group_size\": 1e9, - \"offload_optimizer\": { - \"device\": \"none\", - \"buffer_count\": 4, - \"pipeline_read\": false, - \"pipeline_write\": false, - \"pin_memory\": true - } - }," -elif [ $ZERO_STAGE == 2 ] || [ $ZERO_STAGE == 1 ]; then -zero="\ - \"zero_optimization\": { - \"stage\": $ZERO_STAGE, - \"offload_optimizer\": { - \"device\": \"cpu\", - \"buffer_count\": 4, - \"pipeline_read\": false, - \"pipeline_write\": false, - \"pin_memory\": true - } - }," - if [ $ZERO_STAGE == 1 ]; then - if [ $PP > 1 ]; then - extra="\ - \"data_types\": { - \"grad_accum_dtype\": \"fp32\" - }, - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," - else - echo 'please add the config for zero_stage 1 without pipeline-parallelism' - fi - fi -else - echo 'Please add the correct config set!!!' -fi - -# flops_profiler must at the end because no ',' is allowed at the end -cat < $1 -{ -$common -$zero -$dtype -$extra -$flops_profiler -} -EOT From 37f6d3a757e7393751cdbaa48bea6e4754024a9e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:23:08 -0500 Subject: [PATCH 019/590] Update `train_llama_alcf_sunspot.sh` --- train_llama_alcf_sunspot.sh | 35 +++++++++-------------------------- 1 file changed, 9 insertions(+), 26 deletions(-) diff --git a/train_llama_alcf_sunspot.sh b/train_llama_alcf_sunspot.sh index d5e83c57a0e..f000506bb92 100644 --- a/train_llama_alcf_sunspot.sh +++ b/train_llama_alcf_sunspot.sh @@ -41,15 +41,6 @@ module () { return $__lmod_my_status } -# -# eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" -# conda activate q4-drop - -if [[ $(hostname) == x1* || $(hostname) == x4* ]] ; then - echo "!!!! Caught Intel XPU, using CPU_OPTIMIZER !!!!" - export CPU_OPTIMIZER=1; -fi - # +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ # ---- 0. Navigate into `$PBS_O_WORKDIR` ------------------------------------- @@ -67,7 +58,7 @@ saveDSenv || exit # 2. save env vars to `.deepspeed_ ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` -buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ +buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} setArgs || exit # 8. specify additional `deepspeed` arguments setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset @@ -81,21 +72,15 @@ custom_args=" $@" # Assert `./hostfile_deepspeed` exists export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit -# hf="${HOSTFILE:-${PBS_NODEFILE}}" -# nh=$(wc -l "${hf}") -# if [[ "${nh}" -gt 1 ]]; then -# launch_cmd="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" -# else -# launch_cmd="python3 ${EXEC}" -# fi -# -# echo "launch_cmd: ${launch_cmd}" - # --use-flash-attn-v2 \ - # python3 ${EXEC} \ + # --use-flash-attn \ + # --$DTYPE \ + # --optimizer adamw \ + # --adam-beta1 0.9 \ + # --adam-beta2 0.95 \ run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - --$DTYPE \ + --optimizer ${OPT} \ --num-workers 0 \ --split 100,0,0 \ --log-interval 1 \ @@ -108,9 +93,9 @@ run_cmd=" --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ --lr ${LR} \ - --seq-length $SEQ \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ + --seq-length ${SEQ} \ --num-layers ${NLAYERS} \ --hidden-size ${HIDDEN} \ --train-iters ${TRAIN_ITER} \ @@ -155,12 +140,10 @@ run_cmd=" # |& tee $OUTPUT_DIR/output.log # ${EXTRA_ARGS} \ -echo "All DeepSpeed(s): $(which -a deepspeed)" -echo "Using $(which deepspeed)" +echo "! Using $(which deepspeed)" ds_report echo "${run_cmd}" - printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" # echo "${OUTPUT_LOG}" From 7befd20801d4f62e9e217636a9e1ed3e73c13f71 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:24:03 -0500 Subject: [PATCH 020/590] Update `megatron/core/pipeline_parallel/schedules.py` --- megatron/core/pipeline_parallel/schedules.py | 32 +++++++++----------- 1 file changed, 14 insertions(+), 18 deletions(-) diff --git a/megatron/core/pipeline_parallel/schedules.py b/megatron/core/pipeline_parallel/schedules.py index bb876481acf..453019f8c9d 100644 --- a/megatron/core/pipeline_parallel/schedules.py +++ b/megatron/core/pipeline_parallel/schedules.py @@ -14,7 +14,7 @@ from megatron.core.enums import ModelType from megatron.core.utils import get_attr_wrapped_model, get_model_type, get_model_config -from megatron.utils import unwrap_model +from megatron.utils import print_rank_0, unwrap_model from megatron.model import DistributedDataParallel as LocalDDP from megatron.model import Float16Module @@ -228,7 +228,14 @@ def forward_step(forward_step_func, return [output_tensor] -def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config, model=None): +def backward_step( + input_tensor, + output_tensor, + output_tensor_grad, + model_type, + config, + model=None +): """Backward step through passed-in output tensor. If last stage, output_tensor_grad is None, otherwise gradient of loss @@ -244,20 +251,16 @@ def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, c assert args is not None if config.timers is not None: config.timers('backward-compute', log_level=2).start() - # if (to_skip := args.train_iters_to_skip) is not None and len(to_skip) > 0: - to_skip = getattr(args, 'train_iters_to_skip', None) - if to_skip is not None: + if (to_skip := getattr(args, 'train_iters_to_skip', None)) is not None: if config.timers is not None: config.timers('backward-compute').stop() - if args.iteration in [int(i) for i in to_skip]: - print(f'Caught {args.iteration=} in `iters_to_skip`! Skipping!') + if len(to_skip) > 0 and args.iteration in [int(i) for i in to_skip]: + print_rank_0( + f'Caught {args.iteration=} in `iters_to_skip`! Skipping!' + ) return [None] if args.deepspeed: assert model is not None - - if config.timers is not None: - config.timers('backward-compute', log_level=2).start() - # Retain the grad on the input_tensor. unwrap_input_tensor_grad = False if not isinstance(input_tensor, list): @@ -266,24 +269,20 @@ def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, c for x in input_tensor: if x is not None: x.retain_grad() - if not isinstance(output_tensor, list): output_tensor = [output_tensor] if not isinstance(output_tensor_grad, list): output_tensor_grad = [output_tensor_grad] - # Backward pass. if args.deepspeed: model.backward(output_tensor[0]) else: if output_tensor_grad[0] is None and config.grad_scale_func is not None: output_tensor[0] = config.grad_scale_func(output_tensor[0]) - if config.deallocate_pipeline_outputs: custom_backward(output_tensor[0], output_tensor_grad[0]) else: torch.autograd.backward(output_tensor[0], grad_tensors=output_tensor_grad[0]) - # Collect the grad of the input_tensor. input_tensor_grad = [None] if input_tensor is not None: @@ -293,7 +292,6 @@ def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, c input_tensor_grad.append(None) else: input_tensor_grad.append(x.grad) - # Handle single skip connection if it exists (encoder_hidden_state in # model with encoder and decoder). if parallel_state.get_pipeline_model_parallel_world_size() > 1 and \ @@ -303,10 +301,8 @@ def backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, c input_tensor_grad[-1].add_(output_tensor_grad[1]) if unwrap_input_tensor_grad: input_tensor_grad = input_tensor_grad[0] - if config.timers is not None: config.timers('backward-compute').stop() - return input_tensor_grad From 25353fda499d23bd3d6b42986cbf0f5353537eab Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:24:30 -0500 Subject: [PATCH 021/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 31 +++++++++++++++++++++---------- 1 file changed, 21 insertions(+), 10 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index bc2adb26fa6..30d9add2867 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -40,8 +40,11 @@ setParams() { export BE="${CCL}" # BE = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 - echo "!!!! Using CPU_OPTIMIZER on Intel XPU by Default !!!!" - export CPU_OPTIMIZER=${CPU_OPTIMIZER:-1} # CPU OPTIMIZER ON INTEL XPU + # if [[ -z "${CPU_OPTIMIZER}" ]]; then + # CPU_OPTIMIZER=1 + # fi + # echo "!!!! Using CPU_OPTIMIZER on Intel XPU by Default !!!!" + # export CPU_OPTIMIZER=${CPU_OPTIMIZER:-1} # CPU OPTIMIZER ON INTEL XPU # -------- [Polaris] ----------------------------------- elif [[ $(hostname) == x3* ]]; then TP=${TP:-2} # TP = 2 @@ -53,8 +56,10 @@ setParams() { MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 fi # ------------------------------------------------------------------------ + # export OFFLOAD_DEVICE="${OFFLOAD_DEVICE:-none}" export PP="${PP}" export TP="${TP}" + export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} # ---- Llama2 7B Config ------------------------------ @@ -87,9 +92,13 @@ setParams() { export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" # if [[ "${CPU_OPTIMIZER:-0}" ]]; then - if [[ -n "${CPU_OPTIMIZER}" ]]; then + # if [[ -n "${CPU_OPTIMIZER}" ]]; then + if [[ "${CPU_OPTIMIZER}" == 1 ]]; then + export OFFLOAD_DEVICE="cpu" echo "\n!!! Appending \`--cpu-optimizer\` to LLAMA_ARGS..." export LLAMA_ARGS="${LLAMA_ARGS} --cpu-optimizer" + else + export OFFLOAD_DEVICE="none" fi # ---------------------------------------------------- } @@ -171,13 +180,15 @@ buildDSconfig() { # ---- Build DeepSpeed Config --------------------------------- export DS_CONFIG="ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" echo "DS_CONFIG: ${DS_CONFIG}" - printf "ZS: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" ${ZERO_STAGE} ${MICRO_BATCH} ${GLOBAL_BATCH} ${PP} ${DTYPE} - if [[ -z "${CPU_OPTIMIZER}" ]]; then - bash "${PBS_O_WORKDIR}/generate_config.sh" "${DS_CONFIG}" #|| exit 1 - else - echo "!!! Using CPU Optimizer !!!" - bash "${PBS_O_WORKDIR}/generate_config_cpu_optimizer.sh" "${DS_CONFIG}" - fi + printf "ZS: %s, CPU_OPTIMIZER: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" + bash "${PBS_O_WORKDIR}/generate_config.sh" "${DS_CONFIG}" + # if [[ -z "${CPU_OPTIMIZER}" ]]; then + # echo "!!! Using GPU Optimizer !!!" + # bash "${PBS_O_WORKDIR}/generate_config.sh" "${DS_CONFIG}" #|| exit 1 + # else + # echo "!!! Using CPU Optimizer !!!" + # bash "${PBS_O_WORKDIR}/generate_config_cpu_optimizer.sh" "${DS_CONFIG}" + # fi # ------------------------------------------------------------- } From 29f1e307095188a750da7ba3f97d1697055f2809 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:24:46 -0500 Subject: [PATCH 022/590] Update `generate_config.sh` --- generate_config.sh | 75 +++++++++++++++++++++++++++++----------------- 1 file changed, 47 insertions(+), 28 deletions(-) diff --git a/generate_config.sh b/generate_config.sh index 6bea420a2a1..64f01610876 100644 --- a/generate_config.sh +++ b/generate_config.sh @@ -14,22 +14,23 @@ if [ $# -ne 1 ]; then exit 1 fi +# \"optimizer\": { +# \"type\": \"AdamW\", +# \"params\": { +# \"lr\": ${LR}, +# \"beta1\": 0.9, +# \"beta2\": 0.95, +# \"eps\": 1e-5, +# \"weight_decay\": 1e-1 +# } +# }, + extra="" common="\ \"train_batch_size\": $GLOBAL_BATCH, \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, - \"optimizer\": { - \"type\": \"AdamW\", - \"params\": { - \"lr\": ${LR}, - \"beta1\": 0.9, - \"beta2\": 0.95, - \"eps\": 1e-5, - \"weight_decay\": 1e-1 - } - }, \"scheduler\": { \"type\": \"WarmupLR\", \"params\": { @@ -70,7 +71,7 @@ dtype="\ \"enabled\": true, \"loss_scale\": 1.0 }," -else +elif [[ $DTYPE == "fp16" ]]; then dtype="\ \"communication_data_type\": \"fp16\", \"fp16\": { @@ -84,6 +85,8 @@ dtype="\ \"enabled\": false, \"loss_scale\": 1.0 }," +else + dtype="\"communication_data_type\": \"fp32\"," fi if [ $ZERO_STAGE == 3 ]; then @@ -107,27 +110,43 @@ zero="\ \"pin_memory\": true } }," -elif [ $ZERO_STAGE == 2 ] || [ $ZERO_STAGE == 1 ]; then + +# elif [[ $ZERO_STAGE == 2 ]]; then +elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then + +if [[ -n "${CPU_OPTIMIZER}" ]]; then +echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" + +zero="\ + \"zero_optimization\": { + \"stage\": $ZERO_STAGE, + \"offload_optimizer\": { + \"device\": \"cpu\" + } + }," + +else zero="\ \"zero_optimization\": { \"stage\": $ZERO_STAGE }," - if [ $ZERO_STAGE == 1 ]; then - if [ $PP > 1 ]; then - extra="\ - \"data_types\": { - \"grad_accum_dtype\": \"fp32\" - }, - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," - else - echo 'please add the config for zero_stage 1 without pipeline-parallelism' - fi - fi +fi + +# elif [[ $ZERO_STAGE == 1 ]]; then +if [[ $PP > 1 ]]; then + extra="\ + \"data_types\": { + \"grad_accum_dtype\": \"fp32\" + }, + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," +else + echo 'please add the config for zero_stage 1 without pipeline-parallelism' +fi else echo 'Please add the correct config set!!!' fi From 336332508baeb4e51dc7dfcd480749a13f01b38f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 09:25:37 -0500 Subject: [PATCH 023/590] `assert args is not None` in `megatron/training.py` --- megatron/training.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/megatron/training.py b/megatron/training.py index e987eb158df..f511efc4ffa 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -143,6 +143,8 @@ def pretrain(train_valid_test_dataset_provider, args = get_args() timers = get_timers() + assert args is not None + assert timers is not None if args.deepspeed: args.deepspeed_config_dict = _create_ds_config_dict() @@ -319,6 +321,7 @@ def setup_teacher_model(args, model_provider): def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap_with_ddp=True): """Build the model.""" args = get_args() + assert args is not None args.model_type = model_type # Build model. From 0dcba4ff4c8262a0fe30412509a22792896d552f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Mar 2024 11:08:51 -0500 Subject: [PATCH 024/590] Remove `"scheduler": {...}` from `generate_config.sh` --- generate_config.sh | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/generate_config.sh b/generate_config.sh index 64f01610876..e140b7a274b 100644 --- a/generate_config.sh +++ b/generate_config.sh @@ -24,6 +24,14 @@ fi # \"weight_decay\": 1e-1 # } # }, +# \"scheduler\": { +# \"type\": \"WarmupLR\", +# \"params\": { +# \"warmup_min_lr\": 0.00003, +# \"warmup_max_lr\": 0.0003, +# \"warmup_num_steps\": 5000 +# } +# }, extra="" common="\ @@ -31,14 +39,6 @@ common="\ \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, - \"scheduler\": { - \"type\": \"WarmupLR\", - \"params\": { - \"warmup_min_lr\": 0.00003, - \"warmup_max_lr\": 0.0003, - \"warmup_num_steps\": 5000 - } - }, \"zero_allow_untested_optimizer\": true, \"gradient_clipping\": 1.0, \"activation_checkpointing\": { From 367e4ae534090dd6f016c123bab9f04db1b32b67 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 09:33:57 -0500 Subject: [PATCH 025/590] Track optimizer states --- train_llama_alcf_polaris.sh | 2 ++ 1 file changed, 2 insertions(+) diff --git a/train_llama_alcf_polaris.sh b/train_llama_alcf_polaris.sh index 2e1a23010c5..7cc1454f603 100644 --- a/train_llama_alcf_polaris.sh +++ b/train_llama_alcf_polaris.sh @@ -51,6 +51,7 @@ export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit # --launcher_args='--pmi=pmix' # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ # ${launch_cmd} \ + # --optimizer adam \ run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ --use-flash-attn-v2 \ @@ -66,6 +67,7 @@ run_cmd=" --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ + --log-optimizer-states-to-tensorboard \ --lr ${LR} \ --seq-length $SEQ \ --save ${CKPT_DIR} \ From 31644f068b962cf03479d1c8840254921ef240f1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 09:34:20 -0500 Subject: [PATCH 026/590] Track optimizer states with W&B in `megatron/training.py` --- megatron/training.py | 147 +++++++++++++++++++++++++++---------------- 1 file changed, 93 insertions(+), 54 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index f511efc4ffa..bf87e7b52c2 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -755,7 +755,6 @@ def train_step(forward_step_func, data_iterator, skipped_iter = 0 grad_norm = None num_zeros_in_grad = None - loss_reduced = {} for key in losses_reduced[0]: losses_reduced_for_key = [x[key] for x in losses_reduced] @@ -793,7 +792,7 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, args = get_args() timers = get_timers() writer = get_tensorboard_writer() - + wandb_metrics = {} # Advanced, skipped, and Nan iterations. advanced_iters_key = 'advanced iterations' skipped_iters_key = 'skipped iterations' @@ -852,11 +851,15 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, 'optimizer'] # Calculate batch size. - batch_size = args.micro_batch_size * args.data_parallel_size * \ - get_num_microbatches() - - total_iterations = total_loss_dict[advanced_iters_key] + \ - total_loss_dict[skipped_iters_key] + batch_size = ( + args.micro_batch_size + * args.data_parallel_size + * get_num_microbatches() + ) + total_iterations = ( + total_loss_dict[advanced_iters_key] + + total_loss_dict[skipped_iters_key] + ) # Tensorboard values. # Timer requires all the ranks to call. @@ -870,6 +873,10 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, writer.add_scalar('steps-vs-tokens/y=steps,x=tokens', iteration, args.consumed_train_tokens) writer.add_scalar('steps-vs-tokens/y=tokens,x=steps', args.consumed_train_tokens, iteration) if args.log_learning_rate_to_tensorboard: + wandb_metrics |= { + 'learning-rate/iteration': iteration, + 'learning-rate/learning-rate': learning_rate, + } writer.add_scalar('learning-rate/learning-rate', learning_rate, iteration) writer.add_scalar('learning-rate/learning-rate vs samples', learning_rate, args.consumed_train_samples) @@ -881,7 +888,12 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, args.consumed_train_samples) writer.add_scalar('batch-size/batch-size vs tokens', batch_size, args.consumed_train_tokens) + wandb_metrics |= { + 'lm-loss-training/iteration': iteration, + 'lm-loss-training/consumed_train_tokens': args.consumed_train_tokens, + } for key in loss_dict: + wandb_metrics |= {f'lm-loss-training/{key}': loss_dict[key]} writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) writer.add_scalar(f"lm-loss-training/{key}" + ' vs samples', loss_dict[key], args.consumed_train_samples) @@ -900,18 +912,21 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, writer.add_scalar('world-size/world-size vs tokens', args.world_size, args.consumed_train_tokens) if grad_norm is not None: + wandb_metrics |= {'training/grad-norm': grad_norm} writer.add_scalar('grad-norm/grad-norm', grad_norm, iteration) writer.add_scalar('grad-norm/grad-norm vs samples', grad_norm, args.consumed_train_samples) writer.add_scalar('grad-norm/grad-norm vs tokens', grad_norm, args.consumed_train_tokens) if num_zeros_in_grad is not None: + wandb_metrics |= {'training/num-zeros': num_zeros_in_grad} writer.add_scalar('num-zeros/num-zeros', num_zeros_in_grad, iteration) writer.add_scalar('num-zeros/num-zeros vs samples', num_zeros_in_grad, args.consumed_train_samples) writer.add_scalar('num-zeros/num-zeros vs tokens', num_zeros_in_grad, args.consumed_train_tokens) if params_norm is not None: + wandb_metrics |= {'training/params-norm': params_norm} writer.add_scalar('params-norm/params-norm', params_norm, iteration) writer.add_scalar('params-norm/params-norm vs samples', params_norm, args.consumed_train_samples) @@ -955,7 +970,6 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, mem_stats["allocation.all.current"], iteration, ) - if iteration % args.tensorboard_log_interval == 0: # This logging write various optimizer states to tensorboard. This # feature may consume extra GPU memory thus is set at false by default. @@ -979,26 +993,49 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if args.zero_stage > 0: # ZeRO partiions optimizer states + # opt_stats = opt_stats.clone().detach() + # opt_stats = get_accelerator().FloatTensor opt_stats = get_accelerator().FloatTensor(opt_stats) torch.distributed.all_reduce(opt_stats, group=mpu.get_sequence_data_parallel_group()) + # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_sequence_data_parallel_group()) if args.tensor_model_parallel_size > 1: + # opt_stats = opt_stats.clone().detach() opt_stats = get_accelerator().FloatTensor(opt_stats) torch.distributed.all_reduce(opt_stats, group=mpu.get_tensor_model_parallel_group()) + # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_tensor_model_parallel_group()) if args.pipeline_model_parallel_size > 1: + # opt_stats = opt_stats.clone().detach() opt_stats = get_accelerator().FloatTensor(opt_stats) torch.distributed.all_reduce(opt_stats, group=mpu.get_pipeline_model_parallel_group()) + # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_pipeline_model_parallel_group()) - + wandb_metrics |= { + 'optimizer/iteration': args.iteration, + 'optimizer/consumed_train_tokens': args.consumed_train_tokens, + 'optimizer/variance_l2': opt_stats[0]**0.5, + 'optimizer/variance_sqrt_l2': opt_stats[1]**0.5, + 'optimizer/momentum_l2': opt_stats[2]**0.5, + 'optimizer/weight_l2': opt_stats[3]**0.5, + 'optimizer/variance_l1': opt_stats[4], + 'optimizer/variance_sqrt_l1': opt_stats[5], + 'optimizer/momentum_l1': opt_stats[6], + 'optimizer/weight_l1': opt_stats[7], + 'optimizer/variance_abs_max': opt_stats_2[0], + 'optimizer/variance_sqrt_abs_max': opt_stats_2[1], + 'optimizer/momentum_abs_max': opt_stats_2[2], + 'optimizer/weight_abs_max': opt_stats_2[3], + } # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if writer and is_last_rank(): writer.add_scalar('optimizer/variance_l2 vs tokens', opt_stats[0]**0.5, args.consumed_train_tokens) @@ -1045,27 +1082,24 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, tokens_per_sec_per_replica = tokens_per_sec / args.data_parallel_size tokens_per_gpu_per_second = tokens_per_sec / args.world_size tokens_per_gpu_per_second_per_replica = tokens_per_gpu_per_second / args.data_parallel_size - wandb_metrics = {} - if wandb is not None and getattr(wandb, 'run', None) is not None: - assert wandb.run is not None - wandb_metrics = { - 'throughput/iteration-time': elapsed_time_per_iteration, # 1000 ms / s - 'throughput/samples_per_sec': samples_per_sec, - 'throughput/samples_per_sec_per_replica': samples_per_sec_per_replica, - 'throughput/tokens_per_sec': tokens_per_sec, - 'throughput/tokens_per_sec_per_replica': tokens_per_sec_per_replica, - 'throughput/tokens_per_gpu_per_sec': tokens_per_gpu_per_second, - 'throughput/tokens_per_gpu_per_sec_per_replica': tokens_per_gpu_per_second_per_replica, - 'throughput/tflops': tflops, - 'throughput/approx_params_in_billions': approx_parameters_in_billions, - 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, - 'throughput/iteration': iteration, + wandb_metrics |= { + 'throughput/iteration-time': elapsed_time_per_iteration, # 1000 ms / s + 'throughput/samples_per_sec': samples_per_sec, + 'throughput/samples_per_sec_per_replica': samples_per_sec_per_replica, + 'throughput/tokens_per_sec': tokens_per_sec, + 'throughput/tokens_per_sec_per_replica': tokens_per_sec_per_replica, + 'throughput/tokens_per_gpu_per_sec': tokens_per_gpu_per_second, + 'throughput/tokens_per_gpu_per_sec_per_replica': tokens_per_gpu_per_second_per_replica, + 'throughput/tflops': tflops, + 'throughput/approx_params_in_billions': approx_parameters_in_billions, + 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, + 'throughput/iteration': iteration, + } + if loss_dict is not None: + wandb_metrics |= { + 'loss/iteration': iteration, + **{f'loss/{k}': v for k, v in loss_dict.items()} } - if loss_dict is not None: - wandb_metrics |= { - 'loss/iteration': iteration, - **{f'loss/{k}': v for k, v in loss_dict.items()} - } if writer and args.log_timers_to_tensorboard: writer.add_scalar('iteration-time/iteration-time', elapsed_time_per_iteration, iteration) @@ -1073,31 +1107,36 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration, args.consumed_train_samples) writer.add_scalar('iteration-time/iteration-time vs tokens', elapsed_time_per_iteration, args.consumed_train_tokens) - log_string = ' iteration {:8d}/{:8d} |'.format( - iteration, args.train_iters) - log_string += ' consumed samples: {:12d} |'.format( - args.consumed_train_samples) - log_string += ' consumed tokens: {:12d} |'.format( - args.consumed_train_tokens) - log_string += ' elapsed time per iteration (ms): {:.1f} |'.format( - elapsed_time_per_iteration * 1000.0) - log_string += ' learning rate: {:.3E} |'.format(learning_rate) - log_string += ' global batch size: {:5d} |'.format(batch_size) - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb_metrics |= { - 'training/iteration': iteration, - 'training/iteration_time': elapsed_time_per_iteration, - 'training/iteration_time_vs_tokens': ( - (elapsed_time_per_iteration - / args.consumed_train_tokens) - ), - 'training/iteration_time_vs_samples': ( - (elapsed_time_per_iteration - / args.consumed_train_samples), - ), - 'training/consumed_samples': args.consumed_train_samples, - 'training/consumed_tokens': args.consumed_train_tokens, - } + log_string = f' iteration {iteration:8d}/{args.train_iters:8d} |' + # .format( iteration, args.train_iters) + log_string += ( + f' consumed samples: {args.consumed_train_samples:12d} |' + # .format(args.consumed_train_samples) + ) + log_string += f' consumed tokens: {args.consumed_train_tokens:12d} |' + # .format( args.consumed_train_tokens) + log_string += ( + ' elapsed time per iteration (ms): ' + f'{elapsed_time_per_iteration * 1000.0:.1f} |' + # .format( elapsed_time_per_iteration * 1000.0) + ) + log_string += f' learning rate: {learning_rate:.3E} |' + log_string += f' global batch size: {batch_size:5d} |' + # if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb_metrics |= { + 'training/iteration': iteration, + 'training/iteration_time': elapsed_time_per_iteration, + 'training/iteration_time_vs_tokens': ( + (elapsed_time_per_iteration + / args.consumed_train_tokens) + ), + 'training/iteration_time_vs_samples': ( + (elapsed_time_per_iteration + / args.consumed_train_samples), + ), + 'training/consumed_samples': args.consumed_train_samples, + 'training/consumed_tokens': args.consumed_train_tokens, + } for key in total_loss_dict: if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: From 5273630333e0b4024f50b01f5741dae869182773 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 13:18:34 -0500 Subject: [PATCH 027/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 37 ++++++++++++++++++++++++++++++------- 1 file changed, 30 insertions(+), 7 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index e296a637be0..af88c83a0fe 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -25,6 +25,7 @@ checkpoint_throughput_calculator ) from pathlib import Path +from enrich import get_logger import deepspeed from deepspeed.runtime.utils import see_memory_usage @@ -43,6 +44,7 @@ # backend='deepspeed', # port='5432', # ) +log = get_logger(__name__) RANK = get_rank() WORLD_SIZE = get_world_size() LEVEL = "DEBUG" if RANK == 0 else "CRITICAL" @@ -51,6 +53,10 @@ DISABLE_WANDB = ( WANDB_MODE is not None and str(WANDB_MODE).lower() == 'disabled' ) +if RANK == 0: + log.setLevel("INFO") +else: + log.setLevel("CRITICAL") if RANK == 0 and not DISABLE_WANDB: project_name = ( @@ -74,9 +80,18 @@ def model_provider(pre_process=True, post_process=True): see_memory_usage("Before Building Model", force=True) args = get_args() config = core_transformer_config_from_args(args) - if wandb.run is not None: - print(f"Updating WandB run: [{wandb.run.name}]({wandb.run.url})") - wandb.run.config.update({"args": vars(args)}) + # if wandb.run is not None and RANK == 0: + # print(f"Updating WandB run: [{wandb.run.name}]({wandb.run.url})") + # try: + # wandb.run.config.update({"args": vars(args)}) + # except Exception: + # log.error( + # 'Unable to `wandb.run.config.update({"args": vars(args)})`' + # ) + # if wandb is not None and wandb.run is not None: + # assert wandb is not None and wandb.run is not None + # print(f'Updating {wandb.run.name=} at {wandb.run.url=}') + # wandb.run.config.update({'args': vars(args)}) if RANK == 0: git_ds_info() if hasattr(mpu, 'get_sequence_parallel_group'): @@ -85,10 +100,6 @@ def model_provider(pre_process=True, post_process=True): dpg = mpu.get_data_parallel_group() else: dpg = None - if wandb is not None and wandb.run is not None: - assert wandb is not None and wandb.run is not None - print(f'Updating {wandb.run.name=} at {wandb.run.url=}') - wandb.run.config.update({'args': vars(args)}) with deepspeed.zero.Init( data_parallel_group=dpg, remote_device=( @@ -153,6 +164,18 @@ def model_provider(pre_process=True, post_process=True): see_memory_usage("After Building Model", force=True) if wandb.run is not None: wandb.run.config.update({'num_params': num_params}) + if "args" not in wandb.run.config: + log.info( + f"Updating WandB run.config: [{wandb.run.name}]({wandb.run.get_url()})" + ) + try: + wandb.run.config.update( + {"args": dict(sorted(vars(args).items()))} + ) + except Exception: + log.error( + 'Unable to `wandb.run.config.update({"args": vars(args)})`' + ) try: wandb.run.watch( model, From b4a310ad19aeba318bd2238fc09796a93549ac57 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 13:18:51 -0500 Subject: [PATCH 028/590] Update `train_llama_alcf_sunspot.sh` --- train_llama_alcf_sunspot.sh | 2 ++ 1 file changed, 2 insertions(+) diff --git a/train_llama_alcf_sunspot.sh b/train_llama_alcf_sunspot.sh index f000506bb92..ea4115aacd9 100644 --- a/train_llama_alcf_sunspot.sh +++ b/train_llama_alcf_sunspot.sh @@ -92,6 +92,8 @@ run_cmd=" --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ --lr ${LR} \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ From 4f7ee536f14601f3febbd18f6f6da9f3c1945ed7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 13:19:53 -0500 Subject: [PATCH 029/590] Update `megatron/training.py` --- megatron/training.py | 46 +++++++++++++++++++++++++++++++++++++++----- 1 file changed, 41 insertions(+), 5 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index bf87e7b52c2..a0dd659b7c5 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -63,26 +63,55 @@ def print_datetime(string): time_str = datetime.now().strftime('%Y-%m-%d %H:%M:%S') print_rank_0('[' + string + '] datetime: {} '.format(time_str)) + +def num_floating_point_operations(args, batch_size): + # Group Query Attention. + # if not args.group_query_attention: + if not args.num_key_value_heads: + args.num_key_value_heads = args.num_attention_heads + # args.num_query_groups = args.num_attention_heads + # MoE. + # num_experts_routed_to = 1 if args.num_experts is None else args.moe_router_topk + num_experts_routed_to = 1 if args.num_experts is None else args.topk + gated_linear_multiplier = 3 / 2 if args.swiglu else 1 + return ( + 12 + * batch_size + * args.seq_length + * args.num_layers + * args.hidden_size + * args.hidden_size + * ( + 1 + + ( + (args.ffn_hidden_size / args.hidden_size) + * num_experts_routed_to + * gated_linear_multiplier + ) + + (args.num_key_value_heads / args.num_attention_heads) + + (args.seq_length / args.hidden_size) + + (args.padded_vocab_size / (2 * args.num_layers * args.hidden_size)) + ) + ) + ''' Since v0.9.0, deepspeed.initialize() has forbidden simultaneous setting of args.deepspeed_config (Path) and ds_config dict. So, we use ds_config dict which is the more flexible option. ''' def _create_ds_config_dict(): args = get_args() + assert args is not None if isinstance(args.deepspeed_config, dict) : ds_config_dict = args.deepspeed_config else: with open(args.deepspeed_config, 'r', encoding='utf-8') as config_file: ds_config_dict = json.load(config_file) - if args.universal_checkpoint: ds_config_dict["checkpoint"] = {"load_universal": True} - # Clear config path - args.deepspeed_config = None - + args.deepspeed_config = None return ds_config_dict - + def pretrain(train_valid_test_dataset_provider, model_provider, @@ -1021,6 +1050,7 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_pipeline_model_parallel_group()) wandb_metrics |= { + 'optimizer/learning_rate': learning_rate, 'optimizer/iteration': args.iteration, 'optimizer/consumed_train_tokens': args.consumed_train_tokens, 'optimizer/variance_l2': opt_stats[0]**0.5, @@ -1077,6 +1107,10 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time, total_iterations ) + num_flops = num_floating_point_operations(args, batch_size) + # throughput = ( + # num_floating_point_operations_so_far - arg + # ) samples_per_sec_per_replica = samples_per_sec / args.data_parallel_size tokens_per_sec = samples_per_sec * seq_len tokens_per_sec_per_replica = tokens_per_sec / args.data_parallel_size @@ -1091,6 +1125,8 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, 'throughput/tokens_per_gpu_per_sec': tokens_per_gpu_per_second, 'throughput/tokens_per_gpu_per_sec_per_replica': tokens_per_gpu_per_second_per_replica, 'throughput/tflops': tflops, + 'throughput/flops': num_flops, + 'throughput/tflops-new': num_flops / elapsed_time_per_iteration, 'throughput/approx_params_in_billions': approx_parameters_in_billions, 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, 'throughput/iteration': iteration, From 2f9cf05ce56149d3d492914e1c3fe1fd11a5a209 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Mar 2024 13:33:11 -0500 Subject: [PATCH 030/590] Update `train_llama_alcf_{polaris,sunspot}.sh` --- train_llama_alcf_polaris.sh | 13 +++++++++---- train_llama_alcf_sunspot.sh | 5 +++++ 2 files changed, 14 insertions(+), 4 deletions(-) diff --git a/train_llama_alcf_polaris.sh b/train_llama_alcf_polaris.sh index 7cc1454f603..aae931fdba6 100644 --- a/train_llama_alcf_polaris.sh +++ b/train_llama_alcf_polaris.sh @@ -45,6 +45,8 @@ custom_args=" $@" # Assert `./hostfile_deepspeed` exists export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit +TBDIR="${CKPT_DIR}/tensorboard" +mkdir -p "${TBDIR}" # source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit # echo "Using $(which python3)" @@ -52,10 +54,11 @@ export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ # ${launch_cmd} \ # --optimizer adam \ + # --use-flash-attn-v2 \ run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - --use-flash-attn-v2 \ --$DTYPE \ + --optimizer ${OPT} \ --num-workers 0 \ --split 100,0,0 \ --log-interval 1 \ @@ -67,11 +70,13 @@ run_cmd=" --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ + --tensorboard-dir ${TBDIR} \ + --log-timers-to-tensorboard \ --log-optimizer-states-to-tensorboard \ --lr ${LR} \ - --seq-length $SEQ \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ + --seq-length ${SEQ} \ --num-layers ${NLAYERS} \ --hidden-size ${HIDDEN} \ --train-iters ${TRAIN_ITER} \ @@ -98,8 +103,8 @@ run_cmd=" " -echo "All DeepSpeed(s): $(which -a deepspeed)" -echo "Using $(which deepspeed)" +# echo "All DeepSpeed(s): $(which -a deepspeed)" +echo "! Using $(which deepspeed)" ds_report echo "${run_cmd}" diff --git a/train_llama_alcf_sunspot.sh b/train_llama_alcf_sunspot.sh index ea4115aacd9..700f5dbfdbf 100644 --- a/train_llama_alcf_sunspot.sh +++ b/train_llama_alcf_sunspot.sh @@ -72,6 +72,9 @@ custom_args=" $@" # Assert `./hostfile_deepspeed` exists export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit +TBDIR="${CKPT_DIR}/tensorboard" +mkdir -p "${TBDIR}" + # --use-flash-attn-v2 \ # --use-flash-attn \ # --$DTYPE \ @@ -80,6 +83,7 @@ export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit # --adam-beta2 0.95 \ run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ + --${DTYPE} \ --optimizer ${OPT} \ --num-workers 0 \ --split 100,0,0 \ @@ -92,6 +96,7 @@ run_cmd=" --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ + --tensorboard-dir ${TBDIR} \ --log-timers-to-tensorboard \ --log-optimizer-states-to-tensorboard \ --lr ${LR} \ From 5b9ad9a91d71c9b0573e6ba678ff1d2355a587f1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 1 Apr 2024 23:16:47 -0500 Subject: [PATCH 031/590] Add support for `--optimizer={apex.adam,apex.sgd,adamw,adam,sgd}` --- megatron/optimizer/__init__.py | 31 +++++++++++++++++++++---------- 1 file changed, 21 insertions(+), 10 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index e7400c39b52..cdcc3445412 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -1,12 +1,6 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. from deepspeed.accelerator import get_accelerator -# if get_accelerator().device_name() == 'cuda': -# from apex.optimizers import FusedAdam as Adam -# from apex.optimizers import FusedSGD as SGD -# else: -# from torch.optim import Adam -# from torch.optim import SGD import torch from megatron import get_args @@ -94,7 +88,25 @@ def get_megatron_optimizer(model, betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps) else: - if str(args.optimizer).lower() == 'adamw': + if str(args.optimizer).lower() == 'apex.adam': + assert get_accelerator().device_name() == 'cuda' + from apex.optimizers import FusedAdam as Adam + optimizer = Adam( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) + elif str(args.optimizer).lower() == 'apex.sgd': + from apex.optimizers import FusedSGD as SGD + optimizer = SGD( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + momentum=args.sgd_momentum + ) + elif str(args.optimizer).lower() == 'adamw': optimizer = torch.optim.AdamW( param_groups, lr=args.lr, @@ -104,7 +116,7 @@ def get_megatron_optimizer(model, ) elif args.optimizer == 'adam': if args.ds_fused_adam: - global Adam + # global Adam from deepspeed.ops.adam import FusedAdam Adam = FusedAdam else: @@ -124,8 +136,7 @@ def get_megatron_optimizer(model, momentum=args.sgd_momentum ) else: - raise Exception(f'{args.optimizer} optimizer is not supported.') - + raise TypeError(f'{args.optimizer} optimizer is not supported.') if args.deepspeed: return optimizer From 294d81f2678f99a747f2bb30238a574e9e6fd5e1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 1 Apr 2024 23:17:17 -0500 Subject: [PATCH 032/590] Add support for \`--optimizer={apex.adam,apex.sgd,adamw,adam,sgd}\` - Updates: - `megatron/arguments.py` --- megatron/arguments.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index cd618b285e5..c26bac6fe99 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -914,7 +914,7 @@ def _add_training_args(parser): help='Disable bias in the linear layers', dest='add_bias_linear') group.add_argument('--optimizer', type=str, default='adam', - choices=['adam', 'adamw', 'sgd'], + choices=['adam', 'adamw', 'sgd', 'apex.adam', 'apex.sgd'], help='Optimizer function') group.add_argument('--dataloader-type', type=str, default=None, choices=['single', 'cyclic'], From 888a63ad8e06d0e84b0d5904a310fa4218f23d9c Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Tue, 2 Apr 2024 23:13:47 +0800 Subject: [PATCH 033/590] fix an issue for DP on Megatron-DeepSpeed (#368) --- pretrain_gpt.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pretrain_gpt.py b/pretrain_gpt.py index 785a1291560..52681e5f8fc 100644 --- a/pretrain_gpt.py +++ b/pretrain_gpt.py @@ -36,8 +36,8 @@ def model_provider(pre_process=True, post_process=True): args = get_args() config = core_transformer_config_from_args(args) - if hasattr(mpu, 'get_sequence_parallel_group'): - dpg = mpu.get_sequence_parallel_group() + if hasattr(mpu, 'get_sequence_data_parallel_group'): + dpg = mpu.get_sequence_data_parallel_group() elif hasattr(mpu, 'get_data_parallel_group'): dpg = mpu.get_data_parallel_group() else: From 3c5f47563f697702c1e305fa01b7563f54b747fc Mon Sep 17 00:00:00 2001 From: Yizhou Wang Date: Wed, 3 Apr 2024 01:50:47 +0800 Subject: [PATCH 034/590] remove contiguous copy for flash-attn opbuilder (#372) * remove unnecessary codes for latest flash-attn opbuilder * add use-flash-attn-builder to make flash_attn usage clear and compatible * use hasattr --- megatron/arguments.py | 4 ++- megatron/model/transformer.py | 63 +++++++++++++++++++++++------------ 2 files changed, 45 insertions(+), 22 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 461a2441c39..631d4b12e87 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -421,7 +421,7 @@ def validate_args(args, defaults={}): args.compression_training = False # FlashAttention - args.use_flash_attn = args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2 + args.use_flash_attn = args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2 or args.use_flash_attn_builder # AML if args.aml_data_download_path is not None: @@ -910,6 +910,8 @@ def _add_training_args(parser): 'https://arxiv.org/abs/2307.08691') group.add_argument('--use-flash-attn-triton', action='store_true', help='use FlashAttention implementation of attention using Triton.') + group.add_argument('--use-flash-attn-builder', action='store_true', + help='use FlashAttention op builder.') group.add_argument('--disable-bias-linear', action='store_false', help='Disable bias in the linear layers', dest='add_bias_linear') diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index 924f46c6888..3449cf26f3b 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -381,7 +381,19 @@ def __init__(self, causal=False, softmax_scale=None, attention_dropout=0.0, # Use FlashAttention-2 when args.use_flash_attn_v2 is True args = get_args() - self.flash_attn_func = flash_attn_varlen_func if args.use_flash_attn_v2 else flash_attn_unpadded_func + self.use_flash_attn_builder_v1 = False + self.use_flash_attn_builder_v2 = False + self.use_flash_attn = False + if args.use_flash_attn_builder: + if hasattr(flash_attn_builder, 'flash_attn_func'): + self.flash_attn_func = flash_attn_builder.flash_attn_func + self.use_flash_attn_builder_v1 = True + else: + self.flash_attn_func = flash_attn_builder.flash_attn_func_v2 + self.use_flash_attn_builder_v2 = True + else: + self.flash_attn_func = flash_attn_varlen_func if args.use_flash_attn_v2 else flash_attn_unpadded_func + self.use_flash_attn = True def forward(self, q, k, v): """Implements the multihead softmax attention. @@ -392,22 +404,19 @@ def forward(self, q, k, v): assert all((i.dtype in [torch.float16, torch.bfloat16] for i in (q,k,v))) assert all((get_accelerator().on_accelerator(i) for i in (q, k, v))) - # if get_accelerator().device_name() == 'cuda': - # assert all((i.is_cuda for i in (q,k,v))) - # else: - # assert all((i.is_xpu for i in (q,k,v))) batch_size, seqlen_q = q.shape[0], q.shape[1] seqlen_k = k.shape[1] - if get_accelerator().device_name() == 'cuda': - # goes for cuda device + if self.use_flash_attn: q, k, v = [rearrange(x, 'b s ... -> (b s) ...') for x in [q, k, v]] cu_seqlens_q = torch.arange(0, (batch_size + 1) * seqlen_q, step=seqlen_q, dtype=torch.int32, device=q.device) - else: - # goes for other device + elif self.use_flash_attn_builder_v1: q, k, v = [rearrange(x, 'b s h d -> b h s d').contiguous() for x in [q, k, v]] + else: + # use_flash_attn_builder_v2 + q, k, v = [rearrange(x, 'b s h d -> b h s d') for x in [q, k, v]] if self.training: # during training q,k,v always have same seqlen @@ -424,16 +433,26 @@ def forward(self, q, k, v): device=q.device) if get_accelerator().device_name() == 'cuda' else None dropout_p = 0 - output = self.flash_attn_func( - q, k, v, cu_seqlens_q, cu_seqlens_k, seqlen_q, seqlen_k, - dropout_p, - softmax_scale=self.softmax_scale, causal=is_causal - ) if get_accelerator().device_name() == 'cuda' else flash_attn_builder.flash_attn_func( - q, k, v, self.dropout_p, self.softmax_scale, is_causal - ) + if self.use_flash_attn: + output = self.flash_attn_func( + q, k, v, cu_seqlens_q, cu_seqlens_k, seqlen_q, seqlen_k, + dropout_p, + softmax_scale=self.softmax_scale, causal=is_causal + ) + else: + # use_flash_attn_builder + output = self.flash_attn_func( + q, k, v, self.dropout_p, self.softmax_scale, is_causal + ) + + if self.use_flash_attn: + output = rearrange(output, '(b s) ... -> b s ...', b=batch_size) + elif self.use_flash_attn_builder_v1: + output = rearrange(output, 'b h s d -> b s h d').contiguous() + else: + # use_flash_attn_builder_v2: + output = rearrange(output, 'b h s d -> b s h d') - output = rearrange(output, '(b s) ... -> b s ...', b=batch_size) if get_accelerator().device_name() == 'cuda' else rearrange( - output, 'b h s d -> b s h d').contiguous() return output class FlashSelfAttentionTriton(torch.nn.Module): @@ -492,7 +511,8 @@ def __init__(self, config, layer_number, self.num_key_value_heads = config.num_key_value_heads self.use_gqa = (self.num_attention_heads != self.num_key_value_heads) - self.use_flash_attn = (args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2) \ + self.use_flash_attn = (args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2 or \ + args.use_flash_attn_builder) \ and attention_type == AttnType.self_attn \ and self.attn_mask_type == AttnMaskType.causal self.use_flash_attn_triton = args.use_flash_attn_triton @@ -504,12 +524,13 @@ def __init__(self, config, layer_number, flash_attn_builder = None if args.use_flash_attn_v1: - assert flash_attn_unpadded_func != None or flash_attn_builder != None, ("Cannot import FlashAttention v1 " - "and Cannot find FlashAttention Builder") + assert flash_attn_unpadded_func != None, "Cannot import FlashAttention v1 " if args.use_flash_attn_v2: assert flash_attn_varlen_func != None, "Cannot import FlashAttention v2 " if args.use_flash_attn_triton: assert flash_attn_func != None, "Cannot import FlashAttention triton " + if args.use_flash_attn_builder: + assert flash_attn_builder != None, "Cannot find FlashAttention op builder " assert attention_type == AttnType.self_attn, ('FlashAttention code path only supports ' 'self-attention for now') From bed55a033e79bd98e9ffe917dda78926a43c33d1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:13:21 -0500 Subject: [PATCH 035/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 23 +++++++++-------------- 1 file changed, 9 insertions(+), 14 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 30d9add2867..27e2a65d457 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -80,26 +80,21 @@ setParams() { export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" export SAVE_INTERVAL=${SAVE_INTERVAL:-200} export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} - # export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-0} - # export GLOBAL_BATCH=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" tm="${PBS_O_WORKDIR}/ALCF/tokenizer.model" - # tm_a=/home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/tokenizer.model - # tm_p="/eagle/datasets/dolma/utils/tokenizer.model" - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm_p:-${tm_a}}}" export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - # if [[ "${CPU_OPTIMIZER:-0}" ]]; then - # if [[ -n "${CPU_OPTIMIZER}" ]]; then - if [[ "${CPU_OPTIMIZER}" == 1 ]]; then - export OFFLOAD_DEVICE="cpu" - echo "\n!!! Appending \`--cpu-optimizer\` to LLAMA_ARGS..." - export LLAMA_ARGS="${LLAMA_ARGS} --cpu-optimizer" - else - export OFFLOAD_DEVICE="none" - fi + # # if [[ "${CPU_OPTIMIZER:-0}" ]]; then + # # if [[ -n "${CPU_OPTIMIZER}" ]]; then + # if [[ "${CPU_OPTIMIZER}" == 1 ]]; then + # export OFFLOAD_DEVICE="cpu" + # echo "\n!!! Appending \`--cpu-optimizer\` to LLAMA_ARGS..." + # export LLAMA_ARGS="${LLAMA_ARGS} --cpu-optimizer" + # else + # export OFFLOAD_DEVICE="none" + # fi # ---------------------------------------------------- } From fceb3738351b08671053d2d87cba02a256e7c748 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:37:21 -0500 Subject: [PATCH 036/590] Update `megatron/data/data_samplers.py` --- megatron/data/data_samplers.py | 89 +++++++++++++++++++++++----------- 1 file changed, 61 insertions(+), 28 deletions(-) diff --git a/megatron/data/data_samplers.py b/megatron/data/data_samplers.py index 0aae13abce0..b242101b3a6 100644 --- a/megatron/data/data_samplers.py +++ b/megatron/data/data_samplers.py @@ -2,7 +2,6 @@ """Dataloaders.""" - import random import torch import numpy as np @@ -46,7 +45,10 @@ def build_pretraining_data_loader(dataset, consumed_samples): batch_sampler=batch_sampler, num_workers=args.num_workers, pin_memory=True, - multiprocessing_context=args.multiprocessing_context + multiprocessing_context=( + args.multiprocessing_context if args.num_workers > 0 + else None + ) ) if args.repeated_dataloader: loader=RepeatingLoader(loader) @@ -54,28 +56,39 @@ def build_pretraining_data_loader(dataset, consumed_samples): class MegatronPretrainingSampler: - def __init__(self, total_samples, consumed_samples, micro_batch_size, - data_parallel_rank, data_parallel_size, drop_last=True): + def __init__( + self, + total_samples, + consumed_samples, + micro_batch_size, + data_parallel_rank, + data_parallel_size, + drop_last=True + ): # Keep a copy of input params for later use. self.total_samples = total_samples self.consumed_samples = consumed_samples self.micro_batch_size = micro_batch_size self.data_parallel_rank = data_parallel_rank - self.micro_batch_times_data_parallel_size = \ + self.micro_batch_times_data_parallel_size = ( self.micro_batch_size * data_parallel_size + ) self.drop_last = drop_last # Sanity checks. - assert self.total_samples > 0, \ - 'no sample to consume: {}'.format(self.total_samples) - assert self.consumed_samples < self.total_samples, \ - 'no samples left to consume: {}, {}'.format(self.consumed_samples, - self.total_samples) + assert self.total_samples > 0, ( + f'no sample to consume: {self.total_samples}' + ) + assert self.consumed_samples < self.total_samples, ( + 'no samples left to consume: ' + f'{self.consumed_samples}, {self.total_samples}' + ) assert self.micro_batch_size > 0 assert data_parallel_size > 0 - assert self.data_parallel_rank < data_parallel_size, \ - 'data_parallel_rank should be smaller than data size: {}, ' \ - '{}'.format(self.data_parallel_rank, data_parallel_size) + assert self.data_parallel_rank < data_parallel_size, ( + f'data_parallel_rank should be smaller than data size: ' + f'{self.data_parallel_rank}, {data_parallel_size}' + ) def __len__(self): return self.total_samples @@ -125,8 +138,16 @@ def __getitem__(self, idx): class MegatronPretrainingRandomSampler: - def __init__(self, dataset, total_samples, consumed_samples, micro_batch_size, - data_parallel_rank, data_parallel_size, data_sharding): + def __init__( + self, + dataset, + total_samples, + consumed_samples, + micro_batch_size, + data_parallel_rank, + data_parallel_size, + data_sharding + ): # Keep a copy of input params for later use. self.dataset = dataset self.total_samples = total_samples @@ -135,19 +156,23 @@ def __init__(self, dataset, total_samples, consumed_samples, micro_batch_size, self.data_parallel_rank = data_parallel_rank self.data_parallel_size = data_parallel_size self.data_sharding = data_sharding - self.micro_batch_times_data_parallel_size = \ + self.micro_batch_times_data_parallel_size = ( self.micro_batch_size * data_parallel_size - self.last_batch_size = \ + ) + self.last_batch_size = ( self.total_samples % self.micro_batch_times_data_parallel_size + ) # Sanity checks. - assert self.total_samples > 0, \ - 'no sample to consume: {}'.format(self.total_samples) + assert self.total_samples > 0, ( + f'no sample to consume: {self.total_samples}' + ) assert self.micro_batch_size > 0 assert data_parallel_size > 0 - assert self.data_parallel_rank < data_parallel_size, \ - 'data_parallel_rank should be smaller than data size: {}, ' \ - '{}'.format(self.data_parallel_rank, data_parallel_size) + assert self.data_parallel_rank < data_parallel_size, ( + f'data_parallel_rank should be smaller than data size: ' + f'{self.data_parallel_rank}, {data_parallel_size}' + ) def __len__(self): return self.total_samples @@ -163,23 +188,31 @@ def __iter__(self): # data sharding and random sampling if self.data_sharding: - bucket_size = (self.total_samples // self.micro_batch_times_data_parallel_size) \ - * self.micro_batch_size + bucket_size = ( + self.micro_batch_size * ( + self.total_samples + // self.micro_batch_times_data_parallel_size + ) + ) bucket_offset = current_epoch_samples // self.data_parallel_size start_idx = self.data_parallel_rank * bucket_size - g = torch.Generator() g.manual_seed(self.epoch) random_idx = torch.randperm(bucket_size, generator=g).tolist() idx_range = [start_idx + x for x in random_idx[bucket_offset:]] else: - full_bucket_size = (self.total_samples // self.micro_batch_size) \ - * self.micro_batch_size + full_bucket_size = ( + self.micro_batch_size * ( + self.total_samples + // self.micro_batch_size + ) + ) full_bucket_offset = current_epoch_samples g = torch.Generator() g.manual_seed(self.epoch) - idx_range_total = \ + idx_range_total = ( torch.randperm(full_bucket_size, generator=g).tolist() + ) idx_range_active = idx_range_total[full_bucket_offset:] idx_range = idx_range_active[self.data_parallel_rank::self.data_parallel_size] From 07bb7bf87a7710a114645840de8ffe60dd854467 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:40:39 -0500 Subject: [PATCH 037/590] Add `train_llama_alcf.sh` --- train_llama_alcf.sh | 112 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 112 insertions(+) create mode 100644 train_llama_alcf.sh diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh new file mode 100644 index 00000000000..0a2e7e5643b --- /dev/null +++ b/train_llama_alcf.sh @@ -0,0 +1,112 @@ +#!/bin/bash --login +#PBS -l walltime=06:00:00 +#PBS -A argonne_tpc +#PBS -q prod +#PBS -l select=48 +#PBS -l filesystems=eagle:home + +function sourceFile() { + fp="$1" + echo "source-ing ${fp}" + if [[ -f "${fp}" ]]; then + # shellcheck source="${fp}" + source "${fp}" + else + echo "ERROR: UNABLE TO SOURCE ${fp}" + fi +} + +# ----[0. Navigate into `$PBS_O_WORKDIR`]------------------------------------- +cd "${PBS_O_WORKDIR}" || exit +HERE=$(python3 -c 'import os; print(os.getcwd())') +export HERE +# ----[1. Assert `./pretrain_gpt_alcf.py` exists:]----------------------------- +export EXEC="${HERE}/pretrain_gpt_alcf.py" +[ -f "${EXEC}" ] || exit +# ----[2. `source ./ALCF/helpers_alcf.sh`:]------------------------------------ +sourceFile "${HERE}/ALCF/helpers.sh" || exit +# ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ +setEnv || exit # 1. load `conda` environment +saveDSenv || exit # 2. save env vars to `.deepspeed_env` +ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars +makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` +buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ +setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} +setArgs || exit # 8. specify additional `deepspeed` arguments +setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset +setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` +printJobInfo || exit # 11. print job info +# ----------------------------------------------------------------------------- + +# Take custom args +custom_args=" $@" + +# Assert `./hostfile_deepspeed` exists +export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit +TBDIR="${CKPT_DIR}/tensorboard" +mkdir -p "${TBDIR}" + +# source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit +# echo "Using $(which python3)" +# --launcher_args='--pmi=pmix' + # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ + # ${launch_cmd} \ + # --optimizer adam \ + # --use-flash-attn-v2 \ + # --num-workers 0 \ +run_cmd=" + deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ + --$DTYPE \ + --optimizer ${OPT} \ + --split 100,0,0 \ + --log-interval 1 \ + --no-bias-gelu-fusion \ + --lr-decay-style cosine \ + --no-bias-dropout-fusion \ + --no-masked-softmax-fusion \ + --tokenizer-type Llama2Tokenizer \ + --no-gradient-accumulation-fusion \ + --accumulate-allreduce-grads-in-fp32 \ + --use-checkpoint-opt_param-scheduler \ + --tensorboard-dir ${TBDIR} \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ + --lr ${LR} \ + --save ${CKPT_DIR} \ + --load ${CKPT_DIR} \ + --seq-length ${SEQ} \ + --num-layers ${NLAYERS} \ + --hidden-size ${HIDDEN} \ + --train-iters ${TRAIN_ITER} \ + --eval-iters ${EVAL_ITERS} \ + --distributed-backend ${BE} \ + --num-attention-heads ${HEADS} \ + --save-interval ${SAVE_INTERVAL} \ + --eval-interval ${EVAL_INTERVAL} \ + --max-position-embeddings ${SEQ} \ + --micro-batch-size ${MICRO_BATCH} \ + --data-file-list ${DATA_FILE_LIST} \ + --tensor-model-parallel-size ${TP} \ + --global-batch-size ${GLOBAL_BATCH} \ + --pipeline-model-parallel-size ${PP} \ + --num-key-value-heads ${NUM_KV_HEAD} \ + --data-cache-path ${DATA_CACHE_PATH} \ + --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ + --tokenizer-model ${TOKENIZER_MODEL} \ + ${LLAMA_ARGS} \ + $ds_args \ + ${gpt_args[*]} \ + $custom_args \ + |& tee ${OUTPUT_LOG} + " + +echo "! Using $(which deepspeed)" +ds_report + +echo "${run_cmd}" + +printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" +printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" +eval "${run_cmd}" +set +x From 3c1cdb4c31bdde9503f0fe9f0a7869704b52bc2b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:41:18 -0500 Subject: [PATCH 038/590] Update `megatron/global_vars.py` --- megatron/global_vars.py | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/megatron/global_vars.py b/megatron/global_vars.py index cb284b3c34f..9f833fbd19b 100644 --- a/megatron/global_vars.py +++ b/megatron/global_vars.py @@ -96,7 +96,7 @@ def set_global_variables(args): if args.exit_signal_handler: _set_signal_handler() - + def set_args(args): global _GLOBAL_ARGS @@ -137,11 +137,10 @@ def _set_tensorboard_writer(args): global _GLOBAL_TENSORBOARD_WRITER _ensure_var_is_not_initialized(_GLOBAL_TENSORBOARD_WRITER, 'tensorboard writer') - if hasattr(args, 'tensorboard_dir') and \ args.tensorboard_dir and args.rank == (args.world_size - 1): try: - from torch.utils.tensorboard import SummaryWriter + from torch.utils.tensorboard.writer import SummaryWriter print('> setting tensorboard ...') _GLOBAL_TENSORBOARD_WRITER = SummaryWriter( log_dir=args.tensorboard_dir, @@ -179,9 +178,9 @@ def _set_timers(args): def _ensure_var_is_initialized(var, name): """Make sure the input variable is not None.""" - assert var is not None, '{} is not initialized.'.format(name) + assert var is not None, f'{name} is not initialized.' def _ensure_var_is_not_initialized(var, name): """Make sure the input variable is not None.""" - assert var is None, '{} is already initialized.'.format(name) + assert var is None, f'{name} is already initialized.' From 5fe64ac06fa31a735d146b1154e82665850f92f5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:45:55 -0500 Subject: [PATCH 039/590] Turn on flops profiler in `generate_config.sh` --- generate_config.sh | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/generate_config.sh b/generate_config.sh index e140b7a274b..d5b12642191 100644 --- a/generate_config.sh +++ b/generate_config.sh @@ -49,7 +49,7 @@ common="\ flops_profiler="\ \"flops_profiler\": { - \"enabled\": false, + \"enabled\": true, \"profile_step\": 45, \"module_depth\": -1, \"top_modules\": 1, @@ -94,6 +94,8 @@ zero="\ \"zero_optimization\": { \"stage\": 3, \"reduce_scatter\": false, + \"mics_shard_size\": 4, + \"mics_hierarchical_params_gather\": true, \"stage3_max_live_parameters\": 3e9, \"stage3_max_reuse_distance\": 3e9, \"stage3_param_persistence_threshold\": 1e5, @@ -145,7 +147,14 @@ if [[ $PP > 1 ]]; then \"debug\": false }," else - echo 'please add the config for zero_stage 1 without pipeline-parallelism' + # echo 'please add the config for zero_stage 1 without pipeline-parallelism' + extra="\ + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," fi else echo 'Please add the correct config set!!!' From 316fd938f076ef7e75ba692c2ac4ca0fa99bc857 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:46:31 -0500 Subject: [PATCH 040/590] Update `megatron/model/language_model.py` --- megatron/model/language_model.py | 132 +++++++++++++++++++++---------- 1 file changed, 89 insertions(+), 43 deletions(-) diff --git a/megatron/model/language_model.py b/megatron/model/language_model.py index ec2ae1877ad..ceef5be7257 100644 --- a/megatron/model/language_model.py +++ b/megatron/model/language_model.py @@ -389,10 +389,16 @@ def __init__(self, post_process=True, num_experts=[1]): args = get_args() - # TODO: passing share_embeddings_and_output_weights=False will not work correctly for T5 and embeddings will not be synced. Fix later for T5. - if args.untie_embeddings_and_output_weights: assert not add_decoder - super(TransformerLanguageModel, self).__init__(share_embeddings_and_output_weights=not args.untie_embeddings_and_output_weights) - + # TODO: passing `share_embeddings_and_output_weights=False` + # will not work correctly for T5 and embeddings will not be synced. + # Fix later for T5. + if args.untie_embeddings_and_output_weights: + assert not add_decoder + super(TransformerLanguageModel, self).__init__( + share_embeddings_and_output_weights=( + not args.untie_embeddings_and_output_weights + ) + ) self.pre_process = pre_process self.post_process = post_process self.hidden_size = config.hidden_size @@ -405,27 +411,35 @@ def __init__(self, self.add_pooler = add_pooler self.encoder_hidden_state = None self.add_retriever = args.retro_add_retriever - self.untie_embeddings_and_output_weights = args.untie_embeddings_and_output_weights + self.untie_embeddings_and_output_weights = ( + args.untie_embeddings_and_output_weights + ) self.num_experts = num_experts # Embeddings. if self.pre_process: - self.embedding = Embedding(self.hidden_size, - args.padded_vocab_size, - args.max_position_embeddings, - args.hidden_dropout, - config, - self.num_tokentypes, - args.embedding_weights_in_fp32) + self.embedding = Embedding( + self.hidden_size, + args.padded_vocab_size, + args.max_position_embeddings, + args.hidden_dropout, + config, + self.num_tokentypes, + args.embedding_weights_in_fp32 + ) self._embedding_key = 'embedding' # Rotary positional embeddings - self.use_rotary_position_embeddings = \ - args.use_rotary_position_embeddings + self.use_rotary_position_embeddings = ( + args.use_rotary_position_embeddings + ) if args.use_rotary_position_embeddings: self.seq_length = args.seq_length - rotary_dim = args.hidden_size // args.num_attention_heads \ - if args.kv_channels is None else args.kv_channels + rotary_dim = ( + args.hidden_size // args.num_attention_heads + if args.kv_channels is None + else args.kv_channels + ) if args.rotary_percent < 1.0: rotary_dim = int(rotary_dim * args.rotary_percent) @@ -433,15 +447,22 @@ def __init__(self, # partial rotary embeddings, which is better than full rotary # Wang and Komatsuzaki et al # https://github.com/kingoflolz/mesh-transformer-jax/ - self.rotary_pos_emb = RotaryEmbedding(rotary_dim, theta=args.rope_theta) + self.rotary_pos_emb = RotaryEmbedding( + rotary_dim, + theta=args.rope_theta + ) # Encoder (usually set to True, False if part of an encoder-decoder # architecture and in encoder-only stage). if self.add_encoder: self.encoder = ParallelTransformer( config, - model_type=args.model_type if not args.retro_add_retriever \ - else ModelType.retro_decoder, + # args.model_type if not args.retro_add_retriever + # else ModelType.retro_decoder + model_type=( + ModelType.retro_decoder if args.retro_add_retriever + else args.model_type + ), self_attn_mask_type=self.encoder_attn_mask_type, pre_process=self.pre_process, post_process=self.post_process, @@ -461,7 +482,8 @@ def __init__(self, self_attn_mask_type=self.decoder_attn_mask_type, pre_process=self.pre_process, post_process=self.post_process, - num_experts=self.num_experts) + num_experts=self.num_experts + ) self._decoder_key = 'decoder' else: self.decoder = None @@ -478,24 +500,30 @@ def __init__(self, args.padded_vocab_size, config=config, init_method=self.init_method, - bias=False) # Setting bias to False always to keep it consistent with embedding tying that also does not have a bias. + # Setting bias to False always to keep it consistent with + # embedding tying that also does not have a bias. + bias=False + ) self._output_layer_key = 'output_layer' def set_input_tensor(self, input_tensor): """ See megatron.model.transformer.set_input_tensor()""" - # This is usually handled in schedules.py but some inference code still # gives us non-lists or None if not isinstance(input_tensor, list): input_tensor = [input_tensor] if self.add_encoder and self.add_decoder: - assert len(input_tensor) == 1, \ - 'input_tensor should only be length 1 for stage with both encoder and decoder' + assert len(input_tensor) == 1, ( + 'input_tensor should only be length 1 ' + 'for stage with both encoder and decoder' + ) self.encoder.set_input_tensor(input_tensor[0]) elif self.add_encoder: - assert len(input_tensor) == 1, \ - 'input_tensor should only be length 1 for stage with only encoder' + assert len(input_tensor) == 1, ( + 'input_tensor should only be length 1 ' + 'for stage with only encoder' + ) self.encoder.set_input_tensor(input_tensor[0]) elif self.add_decoder: if len(input_tensor) == 2: @@ -505,32 +533,50 @@ def set_input_tensor(self, input_tensor): self.decoder.set_input_tensor(None) self.encoder_hidden_state = input_tensor[0] else: - raise Exception('input_tensor must have either length 1 or 2') + raise Exception( + 'input_tensor must have either length 1 or 2' + ) else: - raise Exception('Stage must have at least either encoder or decoder') - - def forward(self, enc_input_ids, enc_position_ids, enc_attn_mask, - dec_input_ids=None, dec_position_ids=None, dec_attn_mask=None, - retriever_input_ids=None, - retriever_position_ids=None, - retriever_attn_mask=None, - enc_dec_attn_mask=None, tokentype_ids=None, - inference_params=None, - pooling_sequence_index=0, - enc_hidden_states=None, output_enc_hidden=False): + raise Exception( + 'Stage must have at least either encoder or decoder' + ) + + def forward( + self, + enc_input_ids, + enc_position_ids, + enc_attn_mask, + dec_input_ids=None, + dec_position_ids=None, + dec_attn_mask=None, + retriever_input_ids=None, + retriever_position_ids=None, + retriever_attn_mask=None, + enc_dec_attn_mask=None, + tokentype_ids=None, + inference_params=None, + pooling_sequence_index=0, + enc_hidden_states=None, + output_enc_hidden=False + ): args = get_args() # Encoder embedding. if self.pre_process: - encoder_input = self.embedding(enc_input_ids, enc_position_ids, - tokentype_ids=tokentype_ids) + encoder_input = self.embedding( + enc_input_ids, + enc_position_ids, + tokentype_ids=tokentype_ids + ) else: encoder_input = None # Retriever embedding. if self.add_retriever and self.pre_process: - retriever_input = self.embedding(retriever_input_ids, - retriever_position_ids, - tokentype_ids=tokentype_ids) + retriever_input = self.embedding( + retriever_input_ids, + retriever_position_ids, + tokentype_ids=tokentype_ids + ) else: retriever_input = None From 69bb53e70f6ce27233b24e39633adde9f94c6e89 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:47:03 -0500 Subject: [PATCH 041/590] Remove `--num-workers 0` in `train_llama_alcf_polaris.sh` --- train_llama_alcf_polaris.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train_llama_alcf_polaris.sh b/train_llama_alcf_polaris.sh index aae931fdba6..06f268f10e0 100644 --- a/train_llama_alcf_polaris.sh +++ b/train_llama_alcf_polaris.sh @@ -55,11 +55,11 @@ mkdir -p "${TBDIR}" # ${launch_cmd} \ # --optimizer adam \ # --use-flash-attn-v2 \ + # --num-workers 0 \ run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ --$DTYPE \ --optimizer ${OPT} \ - --num-workers 0 \ --split 100,0,0 \ --log-interval 1 \ --no-bias-gelu-fusion \ From 2a36f142296bd4a3da3cccdb50cdc22dc6786663 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 08:54:18 -0500 Subject: [PATCH 042/590] Update `megatron/timers.py` --- megatron/timers.py | 28 +++++++++++++++++++++++++--- 1 file changed, 25 insertions(+), 3 deletions(-) diff --git a/megatron/timers.py b/megatron/timers.py index 384c7c37a3c..870ba8996fb 100644 --- a/megatron/timers.py +++ b/megatron/timers.py @@ -8,8 +8,14 @@ import torch from deepspeed.accelerator import get_accelerator +from tensorboard.summary import Writer from packaging import version +try: + import wandb +except Exception: + wandb = None + class TimerBase(ABC): @@ -292,8 +298,15 @@ def log(self, names, rank=None, normalizer=1.0, reset=True, barrier=False): print(output_string, flush=True) - def write(self, names, writer, iteration, normalizer=1.0, - reset=False, barrier=False): + def write( + self, + names: list[str], + writer: Writer, + iteration: int, + normalizer: float = 1.0, + reset: bool = False, + barrier: bool = False + ): """Write timers to a tensorboard writer Note that we only report maximum time across ranks to tensorboard. """ @@ -303,7 +316,16 @@ def write(self, names, writer, iteration, normalizer=1.0, assert normalizer > 0.0 name_to_min_max_time = self._get_global_min_max_time( names, reset, barrier, normalizer) + timer_data = { + 'timers/iteration': iteration, + **{ + f'timers/{k}-time': name_to_min_max_time[k][1] + for k in name_to_min_max_time + } + } + if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb.log(timer_data, commit=False) if writer is not None: for name in name_to_min_max_time: _, max_time = name_to_min_max_time[name] - writer.add_scalar(name + '-time', max_time, iteration) + writer.add_scalar(f'{name}-time', max_time, iteration) From 58b1696fe3e349b9c29a223886e103bf23f5c196 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 07:07:26 -0700 Subject: [PATCH 043/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 71 ++++++++++++++++++++++++------------------------- 1 file changed, 35 insertions(+), 36 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 27e2a65d457..d5d70420890 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -12,6 +12,19 @@ printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } + +function setupSrun() { + if [[ $(hostname) == login* || $(hostname) == nid* ]]; then + export NHOSTS="${SLURM_NNODES:-1}" + export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" + export NGPUS="$(( NHOSTS * NGPU_PER_HOST ))" + export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" + else + echo "Skipping setupSrun() on $(hostname)" + fi +} + + function setDSlauncher() { # launcher setting outdir=$1 @@ -35,11 +48,11 @@ setParams() { # -------- [Aurora] ---- || ----- [SunSpot] ------------ if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 - PP=${PP:-1} # PP = 1 export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # BE = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + export WORKING_DIR="${PBS_O_WORKDIR}" # if [[ -z "${CPU_OPTIMIZER}" ]]; then # CPU_OPTIMIZER=1 # fi @@ -48,17 +61,24 @@ setParams() { # -------- [Polaris] ----------------------------------- elif [[ $(hostname) == x3* ]]; then TP=${TP:-2} # TP = 2 - PP=${PP:-1} # PP = 1 export NCCL=${NCCL:-nccl} # NCCL export BE="${NCCL}" # BE = NCCL # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 + export WORKING_DIR="${PBS_O_WORKDIR}" + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + TP="${TP:-2}" + export NCCL="${NCCL:-nccl}" + export BE="${CCL}" + export DTYPE="${DTYPE:-bf16}" + MICRO_BATCH="${MICRO_BATCH:-8}" + export WORKING_DIR="${SLURM_SUBMIT_DIR}" fi # ------------------------------------------------------------------------ - # export OFFLOAD_DEVICE="${OFFLOAD_DEVICE:-none}" - export PP="${PP}" export TP="${TP}" + export PP="${PP:-1}" + export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} @@ -82,19 +102,10 @@ setParams() { export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" - tm="${PBS_O_WORKDIR}/ALCF/tokenizer.model" + tm="${WORKING_DIR}/ALCF/tokenizer.model" export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - # # if [[ "${CPU_OPTIMIZER:-0}" ]]; then - # # if [[ -n "${CPU_OPTIMIZER}" ]]; then - # if [[ "${CPU_OPTIMIZER}" == 1 ]]; then - # export OFFLOAD_DEVICE="cpu" - # echo "\n!!! Appending \`--cpu-optimizer\` to LLAMA_ARGS..." - # export LLAMA_ARGS="${LLAMA_ARGS} --cpu-optimizer" - # else - # export OFFLOAD_DEVICE="none" - # fi # ---------------------------------------------------- } @@ -176,14 +187,8 @@ buildDSconfig() { export DS_CONFIG="ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" echo "DS_CONFIG: ${DS_CONFIG}" printf "ZS: %s, CPU_OPTIMIZER: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" - bash "${PBS_O_WORKDIR}/generate_config.sh" "${DS_CONFIG}" - # if [[ -z "${CPU_OPTIMIZER}" ]]; then - # echo "!!! Using GPU Optimizer !!!" - # bash "${PBS_O_WORKDIR}/generate_config.sh" "${DS_CONFIG}" #|| exit 1 - # else - # echo "!!! Using CPU Optimizer !!!" - # bash "${PBS_O_WORKDIR}/generate_config_cpu_optimizer.sh" "${DS_CONFIG}" - # fi + working_dir="${PBS_O_WORKDIR:-${SLURM_SUBMIT_DIR:-$(pwd)}}" + bash "${working_dir}/generate_config.sh" "${DS_CONFIG}" # ------------------------------------------------------------- } @@ -224,29 +229,21 @@ setEnv() { echo "Running on Polaris !!" # ---- [load conda] --------------------- module load conda/2023-10-04; conda activate cu118-pt221 ; unset PYTHONUSERBASE - # module load conda/2023-10-04 ; conda activate /lus/eagle/projects/datascience/foremans/miniconda3/envs/polaris/py311-cu118 - # ; conda activate /lus/eagle/projects/datascience/foremans/miniconda3/envs/polaris/2024-03-06 - # export PYTHONUSERBASE="${HOME}/.local/polaris/conda/py311-cu118" - # mkdir -p "${PYTHONUSERBASE}" - # if [[ "${VIRTUAL_ENV}" ]]; then - # echo "Caught VIRTUAL_ENV = ${VIRTUAL_ENV} from environment!!" - # else - # echo "Not using VIRTUAL_ENV" - # # sourceFile "${HERE}/venvs/polaris/2023-10-04/bin/activate" || exit - # fi + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + echo "Running on Perlmutter !!" + module load pytorch + source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" else # ------------------------------------- [Unknown] ------------------- echo "Unknown hostname $(hostname)" exit 1 fi + echo "[python] Using: $(which python3)" } makeHostfiles() { - # GPUS_PER_NODE=$(python3 -Wignore -c 'import ezpz; print(ezpz.get_gpus_per_node())') - # source $(python3 -c 'import ezpz; print(ezpz.SAVEJOBENV.as_posix())') || exit - # source $(python3 -c 'import ezpz; print(ezpz.GETJOBENV.as_posix())') || exit source ezpz/src/ezpz/bin/savejobenv || exit #> /tmp/savejobenv.log 2>&1 & source ezpz/src/ezpz/bin/getjobenv || exit - export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST}}" + export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" export hostfile_mpich=hostfile_mpich @@ -264,6 +261,8 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- dfl_fallback="/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_file_list_reweighted.txt" elif [[ $(hostname) == x3* ]]; then dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" else echo "Unknown hostname. Must manually specify DATA_FILE_LIST." fi From 9ac01590afc89016a3cfb699fc5c5d15e42f15a0 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 07:31:56 -0700 Subject: [PATCH 044/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d5d70420890..c0976bd6dec 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -53,11 +53,6 @@ setParams() { export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 export WORKING_DIR="${PBS_O_WORKDIR}" - # if [[ -z "${CPU_OPTIMIZER}" ]]; then - # CPU_OPTIMIZER=1 - # fi - # echo "!!!! Using CPU_OPTIMIZER on Intel XPU by Default !!!!" - # export CPU_OPTIMIZER=${CPU_OPTIMIZER:-1} # CPU OPTIMIZER ON INTEL XPU # -------- [Polaris] ----------------------------------- elif [[ $(hostname) == x3* ]]; then TP=${TP:-2} # TP = 2 @@ -67,10 +62,11 @@ setParams() { export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 export WORKING_DIR="${PBS_O_WORKDIR}" + # -------- [Perlmutter] --------------------------------- elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then TP="${TP:-2}" export NCCL="${NCCL:-nccl}" - export BE="${CCL}" + export BE="${NCCL}" export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-8}" export WORKING_DIR="${SLURM_SUBMIT_DIR}" From e54063b8267a00c5328dd4a4ef966b13537e6227 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 07:33:04 -0700 Subject: [PATCH 045/590] Add `train_llama_nersc_perlmutter.sh` --- train_llama_nersc_perlmutter.sh | 141 ++++++++++++++++++++++++++++++++ 1 file changed, 141 insertions(+) create mode 100644 train_llama_nersc_perlmutter.sh diff --git a/train_llama_nersc_perlmutter.sh b/train_llama_nersc_perlmutter.sh new file mode 100644 index 00000000000..81315798098 --- /dev/null +++ b/train_llama_nersc_perlmutter.sh @@ -0,0 +1,141 @@ +#!/bin/bash --login +#SBATCH -A m4388_g +#SBATCH -C 'gpu&hbm80g' +#SBATCH -q regular +#SBATCH -t 00:30:00 +#SBATCH --nodes 128 +#SBATCH --gpus 512 +# + +function sourceFile() { + fp="$1" + echo "source-ing ${fp}" + if [[ -f "${fp}" ]]; then + # shellcheck source="${fp}" + source "${fp}" + else + echo "ERROR: UNABLE TO SOURCE ${fp}" + fi +} + +# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ +# ---- 0. Navigate into `$PBS_O_WORKDIR` ------------------------------------- +# cd "${PBS_O_WORKDIR}" || exit +cd "${SLURM_SUBMIT_DIR}" || exit +HERE=$(python3 -c 'import os; print(os.getcwd())') +export HERE +# dflfb="${HERE}/genslm-subsample.txt" +# ---- 1. Assert `./pretrain_gpt_alcf.py` exists: ----------------------------- +export EXEC="${HERE}/pretrain_gpt_alcf.py" +[ -f "${EXEC}" ] || exit +# ---- 2. `source ./ALCF/helpers_alcf.sh`: ------------------------------------ +sourceFile "${HERE}/ALCF/helpers.sh" || exit +# ---- 3. Call fns from `./ALCF/helpers_alcf.sh` ------------------------------ +setEnv || exit # 1. load `conda` environment +saveDSenv || exit # 2. save env vars to `.deepspeed_env` +ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars +makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` +buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ +setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} +setArgs || exit # 8. specify additional `deepspeed` arguments +setData "${DATA_FILE_LIST:-${dflfb}}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset +setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` +printJobInfo || exit # 11. print job info +# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ + +# Take custom args +custom_args=" $@" + +# Assert `./hostfile_deepspeed` exists +export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit +TBDIR="${CKPT_DIR}/tensorboard" +mkdir -p "${TBDIR}" + +# source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit +# echo "Using $(which python3)" +# --launcher_args='--pmi=pmix' + # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ + # ${launch_cmd} \ + # --optimizer adam \ + # --use-flash-attn-v2 \ + # deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ +# source ezpz/src/ezpz/bin/getjobenv || exit +# if [[ -z "${DIST_LAUNCH}" ]]; then +# setupSrun || exit +# echo "Using SRUN_EXEC: ${SRUN_EXEC}" +# else +# SRUN_EXEC="${DIST_LAUNCH}" +# fi +# echo "Using SRUN_EXEC: ${SRUN_EXEC}" +# +export NHOSTS="${SLURM_NNODES:-1}" +export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" +export NGPUS="$(( NHOSTS * NGPU_PER_HOST ))" +export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" + + # srun --gpus ${NGPUS} \ + # --gpus-per-node ${NGPU_PER_HOST} \ + # -N ${NHOSTS} \ + # -n ${NGPUS} \ + # -l -u --verbose python3 ${EXEC} \ +run_cmd=" + ${SRUN_EXEC} python3 ${EXEC} \ + --$DTYPE \ + --optimizer ${OPT} \ + --num-workers 0 \ + --split 100,0,0 \ + --log-interval 1 \ + --no-bias-gelu-fusion \ + --lr-decay-style cosine \ + --no-bias-dropout-fusion \ + --no-masked-softmax-fusion \ + --tokenizer-type Llama2Tokenizer \ + --no-gradient-accumulation-fusion \ + --accumulate-allreduce-grads-in-fp32 \ + --use-checkpoint-opt_param-scheduler \ + --tensorboard-dir ${TBDIR} \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ + --lr ${LR} \ + --save ${CKPT_DIR} \ + --load ${CKPT_DIR} \ + --seq-length ${SEQ} \ + --num-layers ${NLAYERS} \ + --hidden-size ${HIDDEN} \ + --train-iters ${TRAIN_ITER} \ + --eval-iters ${EVAL_ITERS} \ + --distributed-backend ${BE} \ + --num-attention-heads ${HEADS} \ + --save-interval ${SAVE_INTERVAL} \ + --eval-interval ${EVAL_INTERVAL} \ + --max-position-embeddings ${SEQ} \ + --micro-batch-size ${MICRO_BATCH} \ + --data-file-list ${DATA_FILE_LIST} \ + --tensor-model-parallel-size ${TP} \ + --global-batch-size ${GLOBAL_BATCH} \ + --pipeline-model-parallel-size ${PP} \ + --num-key-value-heads ${NUM_KV_HEAD} \ + --data-cache-path ${DATA_CACHE_PATH} \ + --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ + --tokenizer-model ${TOKENIZER_MODEL} \ + ${LLAMA_ARGS} \ + $ds_args \ + ${gpt_args[*]} \ + $custom_args \ + |& tee ${OUTPUT_LOG} + " + +run_cmd=$(echo "${run_cmd}" | sed -e 's/ */ /g') + +# echo "All DeepSpeed(s): $(which -a deepspeed)" +echo "! Using $(which deepspeed)" +ds_report + +echo "${run_cmd}" + +printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" +printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" +# echo "${OUTPUT_LOG}" +eval "${run_cmd}" +set +x From 8ac8bdc651f0fe31c5960294b9f1a21d57b006e8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 10:01:38 -0500 Subject: [PATCH 046/590] Update `{train_llama_alcf.sh,ALCF/helpers.sh}` --- ALCF/helpers.sh | 3 +++ train_llama_alcf.sh | 10 +++++----- 2 files changed, 8 insertions(+), 5 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index c0976bd6dec..ee97709ef47 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -102,6 +102,9 @@ setParams() { export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" + if [[ -z "${NO_FLASH_ATTN}" ]]; then + export LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" + fi # ---------------------------------------------------- } diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 0a2e7e5643b..f0c939b9f70 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -50,11 +50,11 @@ mkdir -p "${TBDIR}" # source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit # echo "Using $(which python3)" # --launcher_args='--pmi=pmix' - # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ - # ${launch_cmd} \ - # --optimizer adam \ - # --use-flash-attn-v2 \ - # --num-workers 0 \ +# deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ +# ${launch_cmd} \ +# --use-flash-attn-v2 \ +# --num-workers 0 \ + run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ --$DTYPE \ From 8c6c91f5ecc2eccf87a192cdfe4a86b620165013 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 10:02:01 -0500 Subject: [PATCH 047/590] Update `megatron/training.py` --- megatron/training.py | 23 ++++++++++++++--------- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index a0dd659b7c5..9b701ea6871 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -113,15 +113,17 @@ def _create_ds_config_dict(): return ds_config_dict -def pretrain(train_valid_test_dataset_provider, - model_provider, - model_type, - forward_step_func, - process_non_loss_data_func=None, - extra_args_provider=None, - args_defaults={}, - data_post_process=None, - external_args={}): +def pretrain( + train_valid_test_dataset_provider, + model_provider, + model_type, + forward_step_func, + process_non_loss_data_func=None, + extra_args_provider=None, + args_defaults={}, + data_post_process=None, + external_args={}, +) -> torch.nn.Module: """Main training program. This function will run the followings in the order provided: @@ -149,6 +151,9 @@ def pretrain(train_valid_test_dataset_provider, to it. It is used for programs to add their own arguments. args_defaults: a dictionary from argument-name to argument-value. It to set already parse arguments. + + Returns: + model (torch.nn.Module) """ # Initalize and get arguments, timers, and Tensorboard writer. From 7794fc07fe74fcdecc53fa1061e69d2eb8084e6b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 10:02:20 -0500 Subject: [PATCH 048/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index a91217afa6e..f5061bf5b26 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -571,18 +571,21 @@ def main(): model_provider, ModelType.encoder_or_decoder, forward_step, - args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, + # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, data_post_process=data_post_process ) - - prof.export_chrome_trace(f"{args.tensorboard_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}.json") + args = get_args() + prof.export_chrome_trace( + f"{args.tensorboard_dir}" + "/torch-trace-{RANK}-of-{WORLD_SIZE}.json" + ) else: model = pretrain( train_valid_test_datasets_provider, model_provider, ModelType.encoder_or_decoder, forward_step, - args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, + # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, data_post_process=data_post_process ) return model From 590630e79e2895452a908b73b3f37678484ef773 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 4 Apr 2024 11:13:02 -0500 Subject: [PATCH 049/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index ee97709ef47..bb612def4ae 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -44,6 +44,7 @@ function setDSlauncher() { } setParams() { + LLAMA_ARGS="" # ---- [Parallelism Settings] -------------------------------------------- # -------- [Aurora] ---- || ----- [SunSpot] ------------ if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then @@ -53,6 +54,9 @@ setParams() { export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 export WORKING_DIR="${PBS_O_WORKDIR}" + if [[ -z "${NO_FLASH_ATTN}" ]]; then + LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn" + fi # -------- [Polaris] ----------------------------------- elif [[ $(hostname) == x3* ]]; then TP=${TP:-2} # TP = 2 @@ -62,6 +66,9 @@ setParams() { export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 export WORKING_DIR="${PBS_O_WORKDIR}" + if [[ -z "${NO_FLASH_ATTN}" ]]; then + LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" + fi # -------- [Perlmutter] --------------------------------- elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then TP="${TP:-2}" @@ -70,6 +77,9 @@ setParams() { export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-8}" export WORKING_DIR="${SLURM_SUBMIT_DIR}" + if [[ -z "${NO_FLASH_ATTN}" ]]; then + LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" + fi fi # ------------------------------------------------------------------------ export TP="${TP}" @@ -101,10 +111,7 @@ setParams() { tm="${WORKING_DIR}/ALCF/tokenizer.model" export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" - export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - if [[ -z "${NO_FLASH_ATTN}" ]]; then - export LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" - fi + export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" # ---------------------------------------------------- } From d03aac040e93045fe9b68c516e9528676a2c39e3 Mon Sep 17 00:00:00 2001 From: Varuni Sastry <88804132+vksastry@users.noreply.github.com> Date: Mon, 8 Apr 2024 11:19:47 -0500 Subject: [PATCH 050/590] Update README.md with data preprocessing info and links. --- ALCF/README.md | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/ALCF/README.md b/ALCF/README.md index 1a8612ed8a9..d95c70125d1 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -196,3 +196,14 @@ modules and launch # [...] ``` + +### Data Preprocessing + +AuroraGPT is trained on the Dolma dataset (initially v0), now in the process of moving to v6. For more details on the dataset, refer to https://huggingface.co/datasets/allenai/dolma. The dolma dataset downloaded is already preprocessing to remove the duplicates (dedup) and filtering the data (mixing). For more details refer to https://github.com/allenai/dolma/tree/main/docs and https://github.com/vksastry/dolma_alcf/blob/main/ALCF/Readme.md. + +The data preprocessing of Dolma dataset before training consists of tokenization of the data using a specific tokenizer (LlamaTokenizer is what we are currently using), Use the below script to tokenize the entire dataset. Example shown for Polaris. + +``` bash +cd /eagle/datasets/dolma/utils +./tokenization.sh +``` From bcedecd1ff788d4d363f3365fd396053a08d65be Mon Sep 17 00:00:00 2001 From: Moshe Island Date: Tue, 9 Apr 2024 11:22:45 +0300 Subject: [PATCH 051/590] Support MoE for GPTModelPipe (#373) * MOE: Support MoE layers creation for GPTModelPipe Signed-off-by: Moshe Island * MOE: Support MoE aux loss for GPTModelPipe Propagate aux loss along GPTModelPipe layers by forwarding the aggregated loss from each transformer layer to the next transformer layer. In addition, add a layer to GPTModelPipe, after the last transformer layer, to catch the final aggregated aux loss and cache it for use in the loss function. Signed-off-by: Moshe Island * MOE: Support display of MoE loss for GPTModelPipe Signed-off-by: Moshe Island * MOE: Verify MoE with no pipe/grad partitioned Currently PipelineEngine supports only a single tensor partitioning with grad. MoE model requires to forward with grad both the activations and the aux_loss. Therefore, until PilelineEngine limitation is removed, verify no partitioning when using MoE. Signed-off-by: Moshe Island --------- Signed-off-by: Moshe Island Co-authored-by: Moshe Island --- megatron/model/gpt_model.py | 61 ++++++++++++++++++++++++--- megatron/model/transformer.py | 79 +++++++++++++++++++++++++---------- megatron/training.py | 8 +++- 3 files changed, 121 insertions(+), 27 deletions(-) diff --git a/megatron/model/gpt_model.py b/megatron/model/gpt_model.py index 0527765f16e..8968c96655d 100644 --- a/megatron/model/gpt_model.py +++ b/megatron/model/gpt_model.py @@ -3,6 +3,7 @@ """GPT-2 model.""" import torch +from collections import OrderedDict from megatron import get_args from megatron.core import mpu, tensor_parallel, sequence_parallel @@ -16,7 +17,7 @@ from megatron.model import LayerNorm, RMSNorm from .language_model import EmbeddingPipe -from .transformer import ParallelTransformerLayerPipe, LMHeadPipe +from .transformer import ParallelTransformerLayerPipe, LMHeadPipe, get_num_experts_per_layer from deepspeed.pipe import PipelineModule, LayerSpec, TiedLayerSpec @@ -360,12 +361,33 @@ def _to_float16(inputs): embedding_weights_in_fp32=args.embedding_weights_in_fp32, tied_weight_attr='word_embeddings_weight')) + experts_per_layer = get_num_experts_per_layer(args.num_experts, args.num_layers, args.expert_interval) + self.is_moe_model = any(n_experts > 1 for n_experts in experts_per_layer) + + # Currently PipelineEngine does not support more than 1 pipe and/or grad partitioned tensors that + # require grads. + # When using MoE, we have 2 tensors that are passed along pipeline stages and both require grads. + # Therefore, verify that both pipe_partitioned / grad_partitioned are not enabled + if self.is_moe_model and args.pipeline_model_parallel_size > 1 and args.tensor_model_parallel_size > 1: + pipe_partitioned_enabled = args.deepspeed_config_dict.get('pipeline', {}).get('pipe_partitioned', False) + grad_partitioned_enabled = args.deepspeed_config_dict.get('pipeline', {}).get('grad_partitioned', False) + assert not pipe_partitioned_enabled and not grad_partitioned_enabled, \ + 'Pipe and/or Grad partitioning are not supported for MoE model' + for layer_idx in range(args.num_layers): self.specs.append( LayerSpec(ParallelTransformerLayerPipe, - config, - layer_number=layer_idx, - self_attn_mask_type=AttnMaskType.causal)) + config, + layer_number=layer_idx, + self_attn_mask_type=AttnMaskType.causal, + num_experts=experts_per_layer[layer_idx], + input_aggregated_moe_loss=(self.is_moe_model and layer_idx > 0), + return_aggregated_moe_loss=self.is_moe_model)) + + # if model has experts, add a layer to get and cache the aggregated moe loss from the + # last transformer layer + if self.is_moe_model: + self.specs.append(self._calculate_moe_loss) # Final layernorm after transformer layers if args.normalization == 'layernorm': @@ -404,6 +426,11 @@ def _logits_helper(embedding, lm_output): if args.fp16 or args.bf16: self.specs.append(float16_to_fp32) + # Cache losses + self.moe_loss = None + self.last_lm_loss = None # detached, for display only + self.last_moe_loss = None # detached, for display only + if args.checkpoint_activations: interval = args.checkpoint_num_layers elif args.recompute_granularity == "full" and args.recompute_method == 'uniform': @@ -418,10 +445,34 @@ def _logits_helper(embedding, lm_output): num_dp=mpu.get_data_parallel_world_size()) super().__init__(layers=self.specs, - loss_fn=CrossEntropy, + loss_fn=self.loss_func, topology=topo, activation_checkpoint_interval=interval, partition_method='type:transformer') + def _calculate_moe_loss(self, inputs): + """ Calculate MoE auxiliary loss """ + assert isinstance(inputs, tuple) and len(inputs) == 2 + hidden, aggregated_moe_loss = inputs[0], inputs[1] + args = get_args() + self.moe_loss = aggregated_moe_loss * args.moe_loss_coeff + return hidden + + def loss_func(self, output, labels): + loss = CrossEntropy(output, labels) + self.last_lm_loss = loss.clone().detach() + if self.moe_loss is not None: + loss += self.moe_loss + self.last_moe_loss = self.moe_loss.clone().detach() + return loss + def universal_checkpoint_info(self): return UniversalCheckpointInfo(using_model_pipe=True).get() + + def get_additional_losses(self): + if not self.is_moe_model: + return None + return OrderedDict({ + 'lm loss': self.last_lm_loss, + 'moe loss': self.last_moe_loss + }) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index 3449cf26f3b..e75f13a24f4 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -1229,7 +1229,8 @@ def forward(self, hidden_states, attention_mask=None, retriever_output=None, retriever_attn_mask=None, inference_params=None, - rotary_pos_emb=None): + rotary_pos_emb=None, + aggregated_moe_loss=None): # hidden_states: [s, b, h] # Layer norm at the beginning of the transformer layer. @@ -1321,6 +1322,10 @@ def forward(self, hidden_states, attention_mask=None, else: mlp_output, moe_loss, _ = self.mlp(layernorm_output) + # when aggregated_moe_loss received, returned moe_loss is the aggregated moe loss + if aggregated_moe_loss is not None: + moe_loss += aggregated_moe_loss + # Second residual connection. if self.apply_residual_connection_post_layernorm: residual = layernorm_output @@ -1381,23 +1386,51 @@ class ParallelTransformerLayerPipe(ParallelTransformerLayer): If no mask is provided, the module will query `self._args.attn_mask` for the mask and only return `super().forward(...)` """ + def __init__(self, config, + layer_number, layer_type=LayerType.encoder, + self_attn_mask_type=AttnMaskType.padding, + drop_path_rate=0., num_experts=1, + input_aggregated_moe_loss=False, return_aggregated_moe_loss=False): + self.input_aggregated_moe_loss = input_aggregated_moe_loss + self.return_aggregated_moe_loss = return_aggregated_moe_loss + super().__init__(config, layer_number, layer_type, self_attn_mask_type, drop_path_rate, num_experts) + def forward(self, inputs, **kwargs): assert torch.is_tensor(inputs) or isinstance(inputs, tuple) if not hasattr(self, '_args'): self._args = get_args() rotary_pos_emb = self._args.rotary_pos_emb if self._args.use_rotary_position_embeddings else None if torch.is_tensor(inputs) or len(inputs) == 1: + assert not self.input_aggregated_moe_loss, f'Expecting an input tuple of size >= 2' # No attention mask forwarded, search for args.attn_mask hidden_states, attention_mask = inputs, self._args.attn_mask - # HACK: currently MoE model does not support pipeline parallel, so - # here we just ignore the moe_loss returned by forward() - return super().forward(hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb)[0] - elif len(inputs) == 2: - # Attention mask is an activation. - hidden_states, attention_mask = inputs[0], inputs[1] - # HACK: currently MoE model does not support pipeline parallel, so - # here we just ignore the moe_loss returned by forward() - return super().forward(*inputs, **kwargs, rotary_pos_emb=rotary_pos_emb)[0], attention_mask + output, moe_loss = super().forward(hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb) + return (output, moe_loss) if self.return_aggregated_moe_loss else output + elif len(inputs) in (2, 3): + # Attention mask and aggregated_moe can both be activations. + return_attention_mask = False + if len(inputs) == 2: + if self.input_aggregated_moe_loss: + hidden_states, aggregated_moe_loss = inputs[0], inputs[1] + attention_mask = self._args.attn_mask + else: + hidden_states, attention_mask = inputs[0], inputs[1] + return_attention_mask = True + else: + hidden_states, attention_mask, aggregated_moe_loss = inputs[0], inputs[1], inputs[2] + + # Forward aggregated_moe_loss to ParallelTransformerLayer for further accumulation + if self.input_aggregated_moe_loss: + kwargs.update({'aggregated_moe_loss': aggregated_moe_loss}) + + output, moe_loss = super().forward(hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb) + + ret = (output, ) + if return_attention_mask: + ret += (attention_mask, ) + if self.return_aggregated_moe_loss: + ret += (moe_loss, ) + return ret else: raise RuntimeError('Received more inputs than understood.') @@ -1499,6 +1532,19 @@ def _get_layer_type(model_type, default_layer_type, retro_layer_numbers, return default_layer_type +def get_num_experts_per_layer(num_experts: list, num_layers: int, expert_interval: int, offset: int = 0) -> list: + assert len(num_experts) == 1 or len(num_experts) == num_layers // expert_interval, \ + 'num_experts must be either a single value or a list of the same length as the number of MoE layers' + if len(num_experts) == 1: + num_experts = num_experts * (num_layers // expert_interval) + experts_per_layer = [] + for i in range(num_layers): + layer_num = i + 1 + offset + n_e = num_experts[(layer_num-1) // expert_interval] if layer_num % expert_interval == 0 else 1 + experts_per_layer.append(n_e) + return experts_per_layer + + class ParallelTransformer(MegatronModule): """Transformer class.""" @@ -1682,21 +1728,12 @@ def build_layer(layer_number, n_e): self.num_layers = 1 self.layers = torch.nn.ModuleList([ NoopTransformerLayer(1) ]) else: - assert len(num_experts) == 1 or len(num_experts) == args.num_layers // args.expert_interval, \ - 'num_experts must be either a single value or a list of the same length as the number of MoE layers' - - # Create the list of MoE experts - if len(num_experts) == 1: - num_experts = num_experts * (args.num_layers // args.expert_interval) - # Build the layers self.layers = [] + experts_per_layer = get_num_experts_per_layer(num_experts, self.num_layers, args.expert_interval, offset) for i in range(self.num_layers): layer_num = i + 1 + offset - if layer_num % args.expert_interval == 0: - n_e = num_experts[(layer_num-1) // args.expert_interval] - else: - n_e = 1 + n_e = experts_per_layer[i] self.layers.append(build_layer(layer_num, n_e)) self.layers = torch.nn.ModuleList(self.layers) diff --git a/megatron/training.py b/megatron/training.py index ef32cd38566..7b6acffedec 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -10,6 +10,7 @@ # The earliest we can measure the start time. _TRAIN_START_TIME = time.time() import torch +from collections import OrderedDict from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP from megatron import get_args @@ -667,8 +668,13 @@ def train_step(forward_step_func, data_iterator, num_zeros_in_grad = 0 assert isinstance(model[0], deepspeed.PipelineEngine) loss = model[0].train_batch(data_iter=data_iterator) + additional_losses = model[0].get_additional_losses() + loss_key = 'lm loss' if additional_losses is None else 'loss' # use "lm loss" for backward compatibility + loss_dict = OrderedDict({loss_key: loss}) + if additional_losses is not None: + loss_dict.update(additional_losses) grad_norm = model[0].get_global_grad_norm() - return {'lm loss' : loss}, skipped_iter, grad_norm, num_zeros_in_grad + return loss_dict, skipped_iter, grad_norm, num_zeros_in_grad # Set grad to zero. if not args.deepspeed: From c72914f65ad8542eebc612da19a37dc2ad42651f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 11:59:36 -0500 Subject: [PATCH 052/590] Update `megatron/core/tensor_parallel/cross_entropy.py` --- megatron/core/tensor_parallel/cross_entropy.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/core/tensor_parallel/cross_entropy.py b/megatron/core/tensor_parallel/cross_entropy.py index 9dcdc0459f1..d0453d25eae 100644 --- a/megatron/core/tensor_parallel/cross_entropy.py +++ b/megatron/core/tensor_parallel/cross_entropy.py @@ -69,14 +69,14 @@ def forward(ctx, vocab_parallel_logits, target, label_smoothing=0.0): vocab_size = exp_logits.size(-1) if label_smoothing > 0: - """ + r""" We'd like to assign 1 / (K - 1) probability mass to every index that is not the ground truth. = (1 - alpha) * y_gt + alpha * mean(y_{i for i != gt}) = (1 - alpha) * y_gt + (alpha / (K - 1)) * \sum_{i != gt} y_i = ((K - 1) * (1 - alpha) / (K - 1)) * y_gt + (alpha / (K - 1)) * \sum_{i != gt} y_i = (K * (1 - alpha) - 1) / (K - 1)) * y_gt + (alpha / (K - 1)) * \sum_{i} y_i = (1 - (alpha * K) / (K - 1)) * y_gt + ( (alpha * K) / (K - 1) ) * \sum_{i} y_i / K - From: https://github.com/NVIDIA/NeMo/blob/main/nemo/collections/common/losses/smoothed_cross_entropy.py + From: """ assert 1.0 > label_smoothing > 0.0 smoothing = label_smoothing * vocab_size / (vocab_size - 1) From 7848cd49c70592432c46910a0efb5d01ace2b430 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 12:00:47 -0500 Subject: [PATCH 053/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index f5061bf5b26..46036efbad7 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -94,8 +94,8 @@ def model_provider(pre_process=True, post_process=True): # wandb.run.config.update({'args': vars(args)}) if RANK == 0: git_ds_info() - if hasattr(mpu, 'get_sequence_parallel_group'): - dpg = mpu.get_sequence_parallel_group() + if hasattr(mpu, 'get_sequence_data_parallel_group'): + dpg = mpu.get_sequence_data_parallel_group() elif hasattr(mpu, 'get_data_parallel_group'): dpg = mpu.get_data_parallel_group() else: @@ -154,6 +154,7 @@ def model_provider(pre_process=True, post_process=True): pre_process=pre_process, post_process=post_process ) + num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) # print_rank_0('\n ------------------------ ') # print_rank_0(f'num of parameters {num_params}') @@ -588,6 +589,18 @@ def main(): # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, data_post_process=data_post_process ) + try: + from megatron.text_generation import generate_and_post_process + import ezpz as ez + with torch.autocast(device_type=ez.get_torch_device(), dtype=torch.float16): + response, _, _, _ = generate_and_post_process(model, prompts=["Hello world", "Nature is", "Turing test comprises", "Explain solar eclipse"], tokens_to_generate=32) + if RANK == 0: + log.info(f'generation completed..\n response:{response}') + except ValueError as ve: + log.critical(f'ValueError: {ve}') + pass + # dist.barrier() + model.train() return model From f2b82b964beba5eac36be836a2a185a3515baf0a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 12:01:16 -0500 Subject: [PATCH 054/590] Removes old `train_sbatch_pp64.sh` --- train_sbatch_pp64.sh | 34 ---------------------------------- 1 file changed, 34 deletions(-) delete mode 100755 train_sbatch_pp64.sh diff --git a/train_sbatch_pp64.sh b/train_sbatch_pp64.sh deleted file mode 100755 index b7baf2539e2..00000000000 --- a/train_sbatch_pp64.sh +++ /dev/null @@ -1,34 +0,0 @@ -#!/bin/bash --login -#SBATCH -A m3957_g -#SBATCH -C 'gpu&hbm80g' -#SBATCH -q regular -#SBATCH -t 00:30:00 -#SBATCH --nodes 128 -#SBATCH --gpus 512 - - -# TODO:: -# - Add logic for catching / killing hung process at end of run to ensure -# second run starts up (otherwise, it will wait for the hung process, which -# will run until the job is killed) -# - This wll let us try running multiple experiments in a single slurm job -# allocation. -# - Existing (similar implementation) from my `~/bin/kill-match`: -# ```bash -# #!/bin/bash --login -# TO_KILL=$1 -# kill $(ps aux | grep -E "$USER.+($TO_KILL)" | grep -v grep | awk '{print $2}') - - -PPSIZE=64 \ - MODEL_SIZE_KEY="GPT1T_$(( 2 * PPSIZE ))L" \ - SEQ_LEN=2048 \ - MICRO_BATCH=2 \ - GAS=$(( 8 * PPSIZE )) \ - SP_TYPE=megatron \ - ZERO_STAGE=1 \ - USE_SEQUENCE_PARALLEL=0 \ - MPSIZE=8 \ - SPSIZE=1 \ - USE_ACTIVATION_CHECKPOINTING=1 \ - ./ALCF/train-gpt3.sh From 5c3b5b7f98a2e09d75160d03a61a5e67ecba9ed1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 15:36:24 -0500 Subject: [PATCH 055/590] Update `generate_config.sh` --- generate_config.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/generate_config.sh b/generate_config.sh index d5b12642191..b164b5e6107 100644 --- a/generate_config.sh +++ b/generate_config.sh @@ -50,7 +50,7 @@ common="\ flops_profiler="\ \"flops_profiler\": { \"enabled\": true, - \"profile_step\": 45, + \"profile_step\": 4, \"module_depth\": -1, \"top_modules\": 1, \"detailed\": true, From 4e5c38321ca8d099b8bdad6fb0d47c54febd483e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 15:37:56 -0500 Subject: [PATCH 056/590] Add support for `schedulefree.{AdamWScheduleFree,SGDScheduleFree}` --- megatron/arguments.py | 10 +++- megatron/optimizer/__init__.py | 94 +++++++++++++++++++--------------- megatron/training.py | 43 +++++++++++----- 3 files changed, 94 insertions(+), 53 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 994a727ec32..5f9b6f41448 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -916,7 +916,15 @@ def _add_training_args(parser): help='Disable bias in the linear layers', dest='add_bias_linear') group.add_argument('--optimizer', type=str, default='adam', - choices=['adam', 'adamw', 'sgd', 'apex.adam', 'apex.sgd'], + choices=[ + 'adam', + 'adamw', + 'sgd', + 'apex.adam', + 'apex.sgd', + 'adamwschedulefree', + 'sgdschedulefree' + ], help='Optimizer function') group.add_argument('--dataloader-type', type=str, default=None, choices=['single', 'cyclic'], diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index cdcc3445412..fdaa1f7dc72 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -87,8 +87,22 @@ def get_megatron_optimizer(model, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps) - else: - if str(args.optimizer).lower() == 'apex.adam': + elif str(args.optimizer).lower() == 'adamwschedulefree': + import schedulefree + optimizer = schedulefree.AdamWScheduleFree( + param_groups, + lr=args.lr, + warmup_steps=args.lr_warmup_iters, + ) + elif str(args.optimizer).lower() == 'sgdschedulefree': + import schedulefree + optimizer = schedulefree.SGDScheduleFree( + param_groups, + lr=args.lr, + warmup_steps=args.lr_warmup_iters, + ) + # else: + elif str(args.optimizer).lower() == 'apex.adam': assert get_accelerator().device_name() == 'cuda' from apex.optimizers import FusedAdam as Adam optimizer = Adam( @@ -98,45 +112,45 @@ def get_megatron_optimizer(model, betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps ) - elif str(args.optimizer).lower() == 'apex.sgd': - from apex.optimizers import FusedSGD as SGD - optimizer = SGD( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - momentum=args.sgd_momentum - ) - elif str(args.optimizer).lower() == 'adamw': - optimizer = torch.optim.AdamW( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps - ) - elif args.optimizer == 'adam': - if args.ds_fused_adam: - # global Adam - from deepspeed.ops.adam import FusedAdam - Adam = FusedAdam - else: - Adam = torch.optim.Adam - optimizer = Adam( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps - ) - elif args.optimizer == 'sgd': - optimizer = torch.optim.SGD( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - momentum=args.sgd_momentum - ) + elif str(args.optimizer).lower() == 'apex.sgd': + from apex.optimizers import FusedSGD as SGD + optimizer = SGD( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + momentum=args.sgd_momentum + ) + elif str(args.optimizer).lower() == 'adamw': + optimizer = torch.optim.AdamW( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) + elif args.optimizer == 'adam': + if args.ds_fused_adam: + # global Adam + from deepspeed.ops.adam import FusedAdam + Adam = FusedAdam else: - raise TypeError(f'{args.optimizer} optimizer is not supported.') + Adam = torch.optim.Adam + optimizer = Adam( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) + elif args.optimizer == 'sgd': + optimizer = torch.optim.SGD( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + momentum=args.sgd_momentum + ) + else: + raise TypeError(f'{args.optimizer} optimizer is not supported.') if args.deepspeed: return optimizer diff --git a/megatron/training.py b/megatron/training.py index e4b25117679..634713eafb6 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1018,18 +1018,37 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, opt_stats_2 = [0.0] * 4 for _, group in enumerate(optimizer.param_groups): for _, param in enumerate(group['params']): - opt_stats[0] += (torch.norm(optimizer.state[param]['exp_avg_sq']).item())**2 - opt_stats[1] += (torch.norm(optimizer.state[param]['exp_avg_sq'].sqrt()).item())**2 - opt_stats[2] += (torch.norm(optimizer.state[param]['exp_avg']).item())**2 - opt_stats[3] += (torch.norm(param).item())**2 - opt_stats[4] += torch.norm(optimizer.state[param]['exp_avg_sq'],p=1).item() - opt_stats[5] += torch.norm(optimizer.state[param]['exp_avg_sq'].sqrt(),p=1).item() - opt_stats[6] += torch.norm(optimizer.state[param]['exp_avg'],p=1).item() - opt_stats[7] += torch.norm(param,p=1).item() - opt_stats_2[0] = max(opt_stats_2[0], abs(optimizer.state[param]['exp_avg_sq'].max().item()), abs(optimizer.state[param]['exp_avg_sq'].min().item())) - opt_stats_2[1] = max(opt_stats_2[1], optimizer.state[param]['exp_avg_sq'].sqrt().abs_().max().item()) - opt_stats_2[2] = max(opt_stats_2[2], abs(optimizer.state[param]['exp_avg'].max().item()), abs(optimizer.state[param]['exp_avg'].min().item())) - opt_stats_2[3] = max(opt_stats_2[3], abs(param.max().item()), abs(param.min().item())) + state_param = getattr(optimizer, 'state', None) + if state_param is not None: + exp_avg_sq = state_param.get('exp_avg_sq', torch.tensor(0.)) + exp_avg = state_param.get('exp_avg', torch.tensor(0.)) + opt_stats[0] += (torch.norm(exp_avg_sq).item()) ** 2 + opt_stats[1] += (torch.norm(exp_avg_sq.sqrt()).item()) ** 2 + opt_stats[2] += (torch.norm(exp_avg).item()) ** 2 + opt_stats[3] += (torch.norm(param).item()) ** 2 + opt_stats[4] += torch.norm(exp_avg_sq, p=1).item() + opt_stats[5] += torch.norm(exp_avg_sq.sqrt(), p=1).item() + opt_stats[6] += torch.norm(exp_avg, p=1).item() + opt_stats[7] += torch.norm(param, p=1).item() + opt_stats_2[0] = max( + opt_stats_2[0], + abs(exp_avg_sq.max().item()), + abs(exp_avg_sq.min().item()) + ) + opt_stats_2[1] = max( + opt_stats_2[1], + exp_avg_sq.sqrt().abs_().max().item() + ) + opt_stats_2[2] = max( + opt_stats_2[2], + abs(exp_avg.max().item()), + abs(exp_avg.min().item()) + ) + opt_stats_2[3] = max( + opt_stats_2[3], + abs(param.max().item()), + abs(param.min().item()) + ) # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if args.zero_stage > 0: # ZeRO partiions optimizer states From a68ed8ebaeef43b2d5bcb63d8e8da905f40b7284 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 15:38:25 -0500 Subject: [PATCH 057/590] update `train_llama_alcf.sh` --- train_llama_alcf.sh | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index f0c939b9f70..ce188428504 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -47,6 +47,15 @@ export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" +# TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') +# printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" +# +# if [[ "${TORCH_DEVICE}" == "cuda" ]]; then +# printf %s "Setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True" +# PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True +# fi + + # source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit # echo "Using $(which python3)" # --launcher_args='--pmi=pmix' @@ -55,6 +64,9 @@ mkdir -p "${TBDIR}" # --use-flash-attn-v2 \ # --num-workers 0 \ + # aprun -n "${NGPUS}" -N "${NGPU_PER_HOST}" --pmi=pmix ${PBS_O_WORKDIR}/local_rank.sh + # ${DIST_LAUNCH} $(which python3) ${EXEC} \ +# yeet="${DIST_LAUNCH} ./local_rank.sh" run_cmd=" deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ --$DTYPE \ From a70aa6e95b8c9e24a86c59f6f9401828f7c49aaf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Apr 2024 15:38:49 -0500 Subject: [PATCH 058/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 28 +++++++++++++--------------- 1 file changed, 13 insertions(+), 15 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 46036efbad7..0139330277b 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -36,34 +36,33 @@ import time from torch import nn import torch.nn.functional as F +import ezpz as ez -# from ezpz import get_logger -from ezpz.dist import get_world_size, setup_wandb, get_rank -# RANK = setup_torch( +# ---- SETUP DISTRIBUTED COMMS ---- +# RANK = ez.setup_torch( # backend='deepspeed', # port='5432', # ) +RANK = ez.get_rank() +WORLD_SIZE = ez.get_world_size() +DEVICE = ez.get_torch_device() + +# --- TURN OFF LOGGER ON ALL RANK != 0 ---- log = get_logger(__name__) -RANK = get_rank() -WORLD_SIZE = get_world_size() -LEVEL = "DEBUG" if RANK == 0 else "CRITICAL" +log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") +# ---- SETUP WANDB FROM RANK 0 ---------------- WANDB_MODE = os.environ.get('WANDB_MODE', None) DISABLE_WANDB = ( WANDB_MODE is not None and str(WANDB_MODE).lower() == 'disabled' ) -if RANK == 0: - log.setLevel("INFO") -else: - log.setLevel("CRITICAL") - if RANK == 0 and not DISABLE_WANDB: project_name = ( os.environ.get( - 'WB_PROJECT', + 'WB_PROJECT', # look for WB_PROJECT in env os.environ.get( - 'WANDB_PROJECT', + 'WANDB_PROJECT', # look for WANDB_PROJECT in env 'AuroraGPT' ), ) @@ -71,7 +70,7 @@ print('--------------------------------------------------') print(f"Setting up W&B from: {RANK} with {project_name}") print('--------------------------------------------------') - setup_wandb(project_name=project_name) + ez.setup_wandb(project_name=project_name) def model_provider(pre_process=True, post_process=True): @@ -591,7 +590,6 @@ def main(): ) try: from megatron.text_generation import generate_and_post_process - import ezpz as ez with torch.autocast(device_type=ez.get_torch_device(), dtype=torch.float16): response, _, _, _ = generate_and_post_process(model, prompts=["Hello world", "Nature is", "Turing test comprises", "Explain solar eclipse"], tokens_to_generate=32) if RANK == 0: From 1de3c662eb2302dc99290cfd80cf80055469c811 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 17 Apr 2024 08:05:53 -0500 Subject: [PATCH 059/590] Fix checkpointing with `schedulefree.*` optimizers --- ALCF/helpers.sh | 5 ++- megatron/checkpointing.py | 88 +++++++++++++++++++++++++++++++++++++-- megatron/training.py | 64 ++++++++++++++++------------ 3 files changed, 126 insertions(+), 31 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index bb612def4ae..cd04f890b99 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -178,7 +178,7 @@ saveDSenv() { setOutput() { # ---- Specify output location -------------------------------- - export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}" + export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" @@ -235,6 +235,9 @@ setEnv() { echo "Running on Polaris !!" # ---- [load conda] --------------------- module load conda/2023-10-04; conda activate cu118-pt221 ; unset PYTHONUSERBASE + if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then + source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" module load pytorch diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index 3e4d20035e3..04201fcc06d 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -8,6 +8,12 @@ import numpy as np from deepspeed.accelerator import get_accelerator import torch +import ezpz as ez +import logging +import torch.distributed as tdist + +import yaml +from pathlib import Path from megatron import update_num_microbatches, get_tokenizer from megatron.core import mpu, tensor_parallel @@ -24,6 +30,12 @@ UNIVERSAL_CHECKPOINT_VERSION_VALUE, ) +RANK = ez.get_rank() +WORLD_SIZE = ez.get_world_size() +DEVICE = ez.get_torch_device() +log = logging.getLogger(__name__) +log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") + _CHECKPOINT_VERSION = None @@ -225,6 +237,10 @@ def get_rng_state(): def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): """Save a model checkpoint.""" args = get_args() + assert args is not None + iteration = args.iteration + + save_lr_state_dict() # Only rank zero of the data parallel writes to the disk. if not args.deepspeed: @@ -529,15 +545,81 @@ def _set_arg(arg_name, old_arg_name=None, force=False): return args, checkpoint_args -def load_checkpoint(model, optimizer, opt_param_scheduler, load_arg='load', strict=True, load_only_weights=False): +def load_lr_state_dict(strict: bool = False) -> dict: + """Load {iteration, lr} from .yaml file when restoring from checkpoint.""" + args = get_args() + assert args is not None + lr_state_dict_fp = Path(args.load).joinpath( + f"lr_state_dict_{RANK}_of_{WORLD_SIZE}.yaml" + ) + lr_state_dict = {} + if lr_state_dict_fp.is_file(): + with lr_state_dict_fp.open('r') as f: + lr_state_dict = yaml.safe_load(f) + args.lr = lr_state_dict['lr'] + else: + if strict: + raise FileNotFoundError( + f"{lr_state_dict_fp=}.is_file() is False" + ) + log.info( + f"Unable to load lr_state_dict from {lr_state_dict_fp=}, " + f"but strict=False. Returning empty dictionary: {lr_state_dict=}" + ) + return lr_state_dict + + +def save_lr_state_dict() -> None: + """Save {iteration, lr} to .yaml file for safe-keeping. + + Make sure we're only saving from RANK == 0. + """ + if RANK != 0: + return None + args = get_args() + assert args is not None + outdir = getattr(args, 'save', None) + assert outdir is not None + lr_state_dict_fp = Path(args.save).joinpath( + "lr_state_dict.yaml" + ) + log.info(f"Saving lr_state_dict to {lr_state_dict_fp.as_posix()}") + with lr_state_dict_fp.open('w') as f: + yaml.dump( + {'iteration': args.iteration, 'lr': args.lr}, + f + ) + + +def load_checkpoint( + model, + optimizer, + opt_param_scheduler, + load_arg: str = 'load', + strict: bool = True, + load_only_weights: bool = False, + strict_lr_state_dict: bool = False +): """Load a model checkpoint and return the iteration. strict (bool): whether to strictly enforce that the keys in :attr:`state_dict` of the checkpoint match the names of parameters and buffers in model. """ args = get_args() + assert args is not None load_dir = getattr(args, load_arg) - + if RANK == 0: + lr_state_dict = load_lr_state_dict(strict=strict_lr_state_dict) + lr_tensor = torch.tensor( + lr_state_dict['lr'], + requires_grad=False, + device=DEVICE + ) + else: + lr_state_dict = {} + lr_tensor = torch.tensor(0., requires_grad=False, device=DEVICE) + tdist.broadcast(lr_tensor, 0) + args.lr = lr_tensor.item() if args.deepspeed: if args.finetune: loaded_dir, state_dict = model[0].load_checkpoint(load_dir, @@ -553,7 +635,7 @@ def load_checkpoint(model, optimizer, opt_param_scheduler, load_arg='load', stri print_rank_0(' will not load any checkpoints and will start from ' 'random') return 0 - release = False + release = False else: model = unwrap_model(model) diff --git a/megatron/training.py b/megatron/training.py index 634713eafb6..c02ca8d0efd 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -10,6 +10,7 @@ # The earliest we can measure the start time. _TRAIN_START_TIME = time.time() import torch +import torch.distributed as tdist from collections import OrderedDict from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP @@ -49,9 +50,16 @@ from deepspeed.compression.compress import init_compression, redundancy_clean from deepspeed.runtime.data_pipeline.data_routing.helper import convert_to_random_ltd from megatron.model.transformer import ParallelTransformerLayer +import ezpz as ez +import logging from deepspeed import comm as dist +RANK = ez.get_rank() +WORLD_SIZE = ez.get_world_size() +log = logging.getLogger(__name__) +log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") + try: import wandb except (ImportError, ModuleNotFoundError): @@ -60,9 +68,9 @@ def print_datetime(string): """Note that this call will sync across all ranks.""" - torch.distributed.barrier() + tdist.barrier() time_str = datetime.now().strftime('%Y-%m-%d %H:%M:%S') - print_rank_0('[' + string + '] datetime: {} '.format(time_str)) + log.info('[' + string + '] datetime: {} '.format(time_str)) def num_floating_point_operations(args, batch_size): @@ -169,10 +177,11 @@ def pretrain( # image ... launches. global _TRAIN_START_TIME start_time_tensor = get_accelerator().DoubleTensor([_TRAIN_START_TIME]) - torch.distributed.all_reduce(start_time_tensor, - op=torch.distributed.ReduceOp.MIN) + tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) + # torch.distributed.all_reduce(start_time_tensor, + # op=torch.distributed.ReduceOp.MIN) _TRAIN_START_TIME = start_time_tensor.item() - print_rank_0('time to initialize megatron (seconds): {:.3f}'.format( + log.info('time to initialize megatron (seconds): {:.3f}'.format( time.time() - _TRAIN_START_TIME)) print_datetime('after megatron is initialized') @@ -247,16 +256,16 @@ def pretrain( args.teacher_model = setup_teacher_model(args, model_provider) # Print setup timing. - print_rank_0('done with setup ...') + log.info('done with setup ...') timers.log(['model-and-optimizer-setup', 'train/valid/test-data-iterators-setup'], barrier=True) if not args.skip_train: - print_rank_0('training ...') + log.info('training ...') if args.dataloader_type == 'cyclic' and args.retro_add_retriever: args.train_iters = args.retro_cyclic_train_iters - print_rank_0("retro cyclic train iters : %d" % args.train_iters) + log.info("retro cyclic train iters : %d" % args.train_iters) iteration = 0 if args.do_train and args.train_iters > 0: @@ -273,7 +282,7 @@ def pretrain( if args.save and iteration != 0: save_checkpoint(iteration, model, optimizer, opt_param_scheduler) else: - print_rank_0('skipping training (--skip-train is on) ...') + log.info('skipping training (--skip-train is on) ...') iteration = args.iteration @@ -321,12 +330,12 @@ def update_train_iters(args): args.global_batch_size args.train_iters = iterations - print_rank_0('setting training iterations to {}'.format(args.train_iters)) + log.info('setting training iterations to {}'.format(args.train_iters)) def setup_teacher_model(args, model_provider): - print_rank_0('***>>>>> Student model checkpoint iteration:{}'.format(args.iteration)) + log.info('***>>>>> Student model checkpoint iteration:{}'.format(args.iteration)) iteration_stuent = args.iteration num_layers_student = args.num_layers num_experts_student = args.num_experts @@ -334,7 +343,7 @@ def setup_teacher_model(args, model_provider): num_attention_heads_student = args.num_attention_heads load_student = args.load - print_rank_0('***>>>>> Setting up the teacher model') + log.info('***>>>>> Setting up the teacher model') args.num_layers = args.num_layers_teacher args.num_experts = args.num_experts_teacher @@ -342,7 +351,7 @@ def setup_teacher_model(args, model_provider): args.num_attention_heads = args.num_attention_heads_teacher args.load = args.load_teacher teacher_model, _, _ = load_model_weights_only(model_provider) - print_rank_0('***>>>>> Teacher model:{}'.format(teacher_model)) + log.info('***>>>>> Teacher model:{}'.format(teacher_model)) args.num_layers = num_layers_student args.num_experts = num_experts_student @@ -518,7 +527,7 @@ def get_optimizer_param_scheduler(optimizer): def load_model_weights_only(model_provider_func): """Setup model and optimizer.""" args = get_args() - print_rank_0('***>>>>> Args:{}'.format(args)) + log.info('***>>>>> Args:{}'.format(args)) model = get_model(model_provider_func) @@ -577,7 +586,7 @@ def setup_model_and_optimizer(model_provider_func, else: args.iteration = 0 student_global_steps = model[0].global_steps - print_rank_0('***>>>>> Student model, global step:{}'.format(student_global_steps)) + log.info('***>>>>> Student model, global step:{}'.format(student_global_steps)) if args.compression_training: model, _, _, _ = deepspeed.initialize( @@ -605,7 +614,7 @@ def setup_model_and_optimizer(model_provider_func, opt_param_scheduler = get_optimizer_param_scheduler(optimizer) if args.deepspeed: - print_rank_0("DeepSpeed is enabled.") + log.info("DeepSpeed is enabled.") pp = mpu.get_pipeline_model_parallel_world_size() if args.data_efficiency_curriculum_learning and build_train_valid_test_datasets_provider is not None: train_ds = None @@ -681,7 +690,7 @@ def setup_model_and_optimizer(model_provider_func, # get model without FP16 and/or TorchDDP wrappers if args.iteration == 0 and len(unwrapped_model) == 1 \ and hasattr(unwrapped_model[0], 'init_state_dict_from_bert'): - print_rank_0("Initializing ICT from pretrained BERT model") + log.info("Initializing ICT from pretrained BERT model") unwrapped_model[0].init_state_dict_from_bert() if args.fp16: optimizer.reload_model_params() @@ -1238,7 +1247,8 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 - print_rank_last(log_string) + # print_rank_last(log_string) + log.info(log_string) if report_memory_flag and learning_rate > 0.: # Report memory after optimizer state has been initialized. report_memory('(after {} iterations)'.format(iteration)) @@ -1465,7 +1475,7 @@ def evaluate(forward_step_func, while iteration < args.eval_iters: iteration += 1 if verbose and iteration % args.log_interval == 0: - print_rank_0('Evaluating iter {}/{}'.format(iteration, + log.info('Evaluating iter {}/{}'.format(iteration, args.eval_iters)) forward_backward_func = get_forward_backward_func() @@ -1576,9 +1586,9 @@ def evaluate_and_print_results(prefix, forward_step_func, process_non_loss_data_func(collected_non_loss_data, iteration, writer) length = len(string) + 1 - print_rank_last('-' * length) - print_rank_last(string) - print_rank_last('-' * length) + log.info('-' * length) + log.info(string) + log.info('-' * length) def cyclic_iter(iter): @@ -1603,10 +1613,10 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): train_val_test_num_samples = [train_samples, eval_iters * args.global_batch_size, test_iters * args.global_batch_size] - print_rank_0(' > datasets target sizes (minimum size):') - print_rank_0(' train: {}'.format(train_val_test_num_samples[0])) - print_rank_0(' validation: {}'.format(train_val_test_num_samples[1])) - print_rank_0(' test: {}'.format(train_val_test_num_samples[2])) + log.info(' > datasets target sizes (minimum size):') + log.info(' train: {}'.format(train_val_test_num_samples[0])) + log.info(' validation: {}'.format(train_val_test_num_samples[1])) + log.info(' test: {}'.format(train_val_test_num_samples[2])) # Build the datasets. return build_train_valid_test_datasets_provider(train_val_test_num_samples) @@ -1620,7 +1630,7 @@ def build_train_valid_test_data_loaders( (train_dataloader, valid_dataloader, test_dataloader) = (None, None, None) - print_rank_0('> building train, validation, and test datasets ...') + log.info('> building train, validation, and test datasets ...') # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: From aa2cd591cec5a355b59abd5ec248aed1f462f79f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 17 Apr 2024 08:24:58 -0500 Subject: [PATCH 060/590] Fix checkpointing with `schedulefree.*` optimizers --- megatron/checkpointing.py | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index 04201fcc06d..d585baf7177 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -608,16 +608,16 @@ def load_checkpoint( args = get_args() assert args is not None load_dir = getattr(args, load_arg) + lr_state_dict = {} + lr_tensor = torch.tensor(args.lr, requires_grad=False, device=DEVICE) if RANK == 0: lr_state_dict = load_lr_state_dict(strict=strict_lr_state_dict) - lr_tensor = torch.tensor( - lr_state_dict['lr'], - requires_grad=False, - device=DEVICE - ) - else: - lr_state_dict = {} - lr_tensor = torch.tensor(0., requires_grad=False, device=DEVICE) + if len(lr_state_dict.keys()) > 0 and 'lr' in lr_state_dict: + lr_tensor = torch.tensor( + lr_state_dict['lr'], + requires_grad=False, + device=DEVICE, + ) tdist.broadcast(lr_tensor, 0) args.lr = lr_tensor.item() if args.deepspeed: From a365a182a3f17edf90d59021088ef28855ead611 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 17 Apr 2024 09:03:41 -0500 Subject: [PATCH 061/590] Add `--schedulefree-foreach` flag From https://github.com/facebookresearch/schedule_free/blob/6db4953ccde631baca5f5fd2d9e5cab21caf9ed4/schedulefree/adamw_schedulefree.py#L39 --- megatron/arguments.py | 35 +++++++++---- megatron/optimizer/__init__.py | 95 +++++++++++++++++++--------------- 2 files changed, 78 insertions(+), 52 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 5f9b6f41448..b982337f517 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -915,17 +915,30 @@ def _add_training_args(parser): group.add_argument('--disable-bias-linear', action='store_false', help='Disable bias in the linear layers', dest='add_bias_linear') - group.add_argument('--optimizer', type=str, default='adam', - choices=[ - 'adam', - 'adamw', - 'sgd', - 'apex.adam', - 'apex.sgd', - 'adamwschedulefree', - 'sgdschedulefree' - ], - help='Optimizer function') + group.add_argument( + '--optimizer', + type=str, + default='adam', + choices=[ + 'adam', + 'adamw', + 'sgd', + 'apex.adam', + 'apex.sgd', + 'adamwschedulefree', + 'sgdschedulefree' + ], + help='Optimizer function' + ) + group.add_argument( + "--schedulefree-for-each", + action="store_true", + help=""" + Use a foreach-backed implementation of the schedulefree optimizers. + Should be significantly faster, + but will have a higher peak memory usage. + """, + ) group.add_argument('--dataloader-type', type=str, default=None, choices=['single', 'cyclic'], help='Single pass vs multiple pass data loader') diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index fdaa1f7dc72..6dfd0ea40f7 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -60,21 +60,29 @@ def get_param_groups(modules, return param_groups -def get_megatron_optimizer(model, - no_weight_decay_cond=None, - scale_lr_cond=None, - lr_mult=1.0): + +def get_megatron_optimizer( + model, + no_weight_decay_cond=None, + scale_lr_cond=None, + lr_mult=1.0 +): args = get_args() # Base optimizer. - param_groups = get_param_groups(model, - no_weight_decay_cond, - scale_lr_cond, - lr_mult) + param_groups = get_param_groups( + model, + no_weight_decay_cond, + scale_lr_cond, + lr_mult + ) if args.create_moe_param_group: - from deepspeed.moe.utils import split_params_into_different_moe_groups_for_optimizer - param_groups = split_params_into_different_moe_groups_for_optimizer(param_groups) - + from deepspeed.moe.utils import ( + split_params_into_different_moe_groups_for_optimizer + ) + param_groups = split_params_into_different_moe_groups_for_optimizer( + param_groups + ) if args.cpu_optimizer: assert args.optimizer == 'adam', 'CPU offloading is for Adam' if args.cpu_torch_adam: @@ -82,36 +90,44 @@ def get_megatron_optimizer(model, else: from deepspeed.ops.adam import DeepSpeedCPUAdam cpu_adam_optimizer = DeepSpeedCPUAdam - optimizer = cpu_adam_optimizer(param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps) + optimizer = cpu_adam_optimizer( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, + ) elif str(args.optimizer).lower() == 'adamwschedulefree': import schedulefree optimizer = schedulefree.AdamWScheduleFree( param_groups, lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, warmup_steps=args.lr_warmup_iters, + foreach=args.schedulefree_for_each, ) elif str(args.optimizer).lower() == 'sgdschedulefree': import schedulefree optimizer = schedulefree.SGDScheduleFree( param_groups, lr=args.lr, + momentum=args.sgd_momentum, + weight_decay=args.weight_decay, warmup_steps=args.lr_warmup_iters, + foreach=args.schedulefree_for_each, ) - # else: elif str(args.optimizer).lower() == 'apex.adam': - assert get_accelerator().device_name() == 'cuda' - from apex.optimizers import FusedAdam as Adam - optimizer = Adam( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps - ) + assert get_accelerator().device_name() == 'cuda' + from apex.optimizers import FusedAdam as Adam + optimizer = Adam( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) elif str(args.optimizer).lower() == 'apex.sgd': from apex.optimizers import FusedSGD as SGD optimizer = SGD( @@ -153,18 +169,15 @@ def get_megatron_optimizer(model, raise TypeError(f'{args.optimizer} optimizer is not supported.') if args.deepspeed: return optimizer - # Determine whether the params have main-grad field. params_have_main_grad = False if args.use_contiguous_buffers_in_local_ddp: params_have_main_grad = True - # Mixed precision optimizer. # - Note: both the Float16Optimizer and the DistributedOptimizer inherit # from the MixedPrecisionOptimizer, which manages any optimizer where # the model params and main params are distinct. if args.fp16 or args.bf16 or args.use_distributed_optimizer: - # Grad scaler: # if loss-scale is provided, instantiate the constant scaler. # if we are using fp16 and loss-scale is not present, use a @@ -172,11 +185,9 @@ def get_megatron_optimizer(model, # otherwise we are running in bf16 with no loss-scale so # leave it as None. grad_scaler = None - # Constant loss scale. if args.loss_scale: grad_scaler = ConstantGradScaler(args.loss_scale) - # Dynamic loss scale. else: if args.fp16: @@ -187,11 +198,11 @@ def get_megatron_optimizer(model, backoff_factor=0.5, growth_interval=args.loss_scale_window, hysteresis=args.hysteresis) - # Megatron optimizer. - opt_ty = DistributedOptimizer \ - if args.use_distributed_optimizer else \ - Float16OptimizerWithFloat16Params + opt_ty = ( + DistributedOptimizer if args.use_distributed_optimizer + else Float16OptimizerWithFloat16Params + ) return opt_ty(optimizer, args.clip_grad, args.log_num_zeros_in_grad, @@ -202,10 +213,12 @@ def get_megatron_optimizer(model, args.params_dtype, grad_scaler, model) - # FP32. - return FP32Optimizer(optimizer, args.clip_grad, - args.log_num_zeros_in_grad, - params_have_main_grad, - args.use_contiguous_buffers_in_local_ddp, - model) + return FP32Optimizer( + optimizer, + args.clip_grad, + args.log_num_zeros_in_grad, + params_have_main_grad, + args.use_contiguous_buffers_in_local_ddp, + model + ) From eecf70d55ea4c8b46cb714bc127087aa645299aa Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 17 Apr 2024 09:05:16 -0500 Subject: [PATCH 062/590] Replace `{:.6E}` with `{:.6f}` in `log_string` formatting --- megatron/training.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training.py b/megatron/training.py index c02ca8d0efd..43fd359ab0d 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1218,7 +1218,7 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, avg = total_loss_dict[key].item() / \ float(max(1, total_loss_dict[advanced_iters_key])) if avg > 0.0: - log_string += ' {}: {:.6E} |'.format(key, avg) + log_string += ' {}: {:.6f} |'.format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) if loss_scale is not None: log_string += ' loss scale: {:.1f} |'.format(loss_scale) From 6969dc2d09bc845e3982aa0fb416760830cca9c3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 17 Apr 2024 11:06:11 -0500 Subject: [PATCH 063/590] Update `megatron/training.py` --- megatron/training.py | 45 ++++++++++++++++++++++---------------------- 1 file changed, 23 insertions(+), 22 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 43fd359ab0d..67cd47bda15 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -70,7 +70,7 @@ def print_datetime(string): """Note that this call will sync across all ranks.""" tdist.barrier() time_str = datetime.now().strftime('%Y-%m-%d %H:%M:%S') - log.info('[' + string + '] datetime: {} '.format(time_str)) + log.info('[' + string + '] datetime={} '.format(time_str)) def num_floating_point_operations(args, batch_size): @@ -181,7 +181,7 @@ def pretrain( # torch.distributed.all_reduce(start_time_tensor, # op=torch.distributed.ReduceOp.MIN) _TRAIN_START_TIME = start_time_tensor.item() - log.info('time to initialize megatron (seconds): {:.3f}'.format( + log.info('time to initialize megatron (seconds)={:.3f}'.format( time.time() - _TRAIN_START_TIME)) print_datetime('after megatron is initialized') @@ -435,7 +435,7 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap # Print number of parameters. if mpu.get_data_parallel_rank() == 0: print(' > number of parameters on (tensor, pipeline) ' - 'model parallel rank ({}, {}): {}'.format( + 'model parallel rank ({}, {})={}'.format( mpu.get_tensor_model_parallel_rank(), mpu.get_pipeline_model_parallel_rank(), sum([sum([p.ds_numel if hasattr(p,'ds_id') else p.nelement() for p in model_module.parameters()]) @@ -1185,18 +1185,18 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, log_string = f' iteration {iteration:8d}/{args.train_iters:8d} |' # .format( iteration, args.train_iters) log_string += ( - f' consumed samples: {args.consumed_train_samples:12d} |' + f' consumed samples={args.consumed_train_samples:12d} |' # .format(args.consumed_train_samples) ) - log_string += f' consumed tokens: {args.consumed_train_tokens:12d} |' + log_string += f' consumed tokens={args.consumed_train_tokens:12d} |' # .format( args.consumed_train_tokens) log_string += ( ' elapsed time per iteration (ms): ' f'{elapsed_time_per_iteration * 1000.0:.1f} |' # .format( elapsed_time_per_iteration * 1000.0) ) - log_string += f' learning rate: {learning_rate:.3E} |' - log_string += f' global batch size: {batch_size:5d} |' + log_string += f' learning rate={learning_rate:.3f} |' + log_string += f' global batch size={batch_size:5d} |' # if wandb is not None and getattr(wandb, 'run', None) is not None: wandb_metrics |= { 'training/iteration': iteration, @@ -1218,32 +1218,32 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, avg = total_loss_dict[key].item() / \ float(max(1, total_loss_dict[advanced_iters_key])) if avg > 0.0: - log_string += ' {}: {:.6f} |'.format(key, avg) + log_string += ' {}={:.6f} |'.format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) if loss_scale is not None: - log_string += ' loss scale: {:.1f} |'.format(loss_scale) + log_string += ' loss scale={:.1f} |'.format(loss_scale) wandb_metrics |= {'loss/loss_scale': loss_scale} if grad_norm is not None: - log_string += ' grad norm: {:.3f} |'.format(grad_norm) + log_string += ' grad norm={:.3f} |'.format(grad_norm) wandb_metrics |= {'loss/grad_norm': grad_norm} if num_zeros_in_grad is not None: - log_string += ' num zeros: {:.1f} |'.format(num_zeros_in_grad) + log_string += ' num zeros={:.1f} |'.format(num_zeros_in_grad) wandb_metrics |= {'loss/num_zeros_in_grad': num_zeros_in_grad} if params_norm is not None: - log_string += ' params norm: {:.3f} |'.format(params_norm) + log_string += ' params norm={:.3f} |'.format(params_norm) wandb_metrics |= {'loss/params_norm': params_norm} if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - log_string += ' curriculum seqlen: {:5d} |'.format(args.curriculum_seqlen) + log_string += ' curriculum seqlen={:5d} |'.format(args.curriculum_seqlen) if args.random_ltd: - log_string += ' random ltd reserved length: {:5d} |'.format(args.random_ltd_reserved_length) - log_string += ' actual seqlen: {:5d} |'.format(seq_len) - log_string += ' number of skipped iterations: {:3d} |'.format( + log_string += ' random ltd reserved length={:5d} |'.format(args.random_ltd_reserved_length) + log_string += ' actual seqlen={:5d} |'.format(seq_len) + log_string += ' number of skipped iterations={:3d} |'.format( total_loss_dict[skipped_iters_key]) - log_string += ' number of nan iterations: {:3d} |'.format( + log_string += ' number of nan iterations={:3d} |'.format( total_loss_dict[nan_iters_key]) - log_string += ' samples per second: {:.3f} |'.format(samples_per_sec) - log_string += ' tokens per gpu per second (tgs): {:.3f} |'.format(tokens_per_gpu_per_second) - log_string += ' TFLOPs: {:.2f} |'.format(tflops) + log_string += ' samples per second={:.3f} |'.format(samples_per_sec) + log_string += ' tokens per gpu per second (tgs)={:.3f} |'.format(tokens_per_gpu_per_second) + log_string += ' TFLOPs={:.2f} |'.format(tflops) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 @@ -1560,9 +1560,10 @@ def evaluate_and_print_results(prefix, forward_step_func, process_non_loss_data_func, config, verbose) string = ' validation loss at {} | '.format(prefix) for key in total_loss_dict: - string += '{} value: {:.6E} | '.format(key, total_loss_dict[key].item()) + string += f"{key} value={total_loss_dict[key].item():.6f}" ppl = math.exp(min(20, total_loss_dict[key].item())) - string += '{} PPL: {:.6E} | '.format(key, ppl) + string += f"{key} PPL={ppl:.6f}" + # string += '{} PPL={:.6f} | '.format(key, ppl) if writer and is_last_rank(): data_type = 'test' if test else 'validation' writer.add_scalar(f'lm-loss-validation/{key} {data_type}', From 1d30d41dd33de0e26d9253e581120c10aa8c99c7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 08:39:03 -0500 Subject: [PATCH 064/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 63 ++++++++----------------------------------------- 1 file changed, 10 insertions(+), 53 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index cd04f890b99..151c3e857a3 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -28,9 +28,6 @@ function setupSrun() { function setDSlauncher() { # launcher setting outdir=$1 - # hfds=$1 - # hfmpi=$2 - # here=$(python3 -c 'import os; print(os.getcwd())') export hfds="$outdir/hostfile_deepspeed" export hfmpi="$outdir/hostfile_mpich" [ -f "$hfds" ] || exit @@ -90,8 +87,8 @@ setParams() { export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} # ---- Llama2 7B Config ------------------------------ export MODEL_KEY="Llama-7B" - export HEADS=${HEADS:-32} - export NLAYERS=${NLAYERS:-32} + export HEADS=${HEADS:-${NHEADS:-32}} + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} export HIDDEN=${HIDDEN:-4096} export NUM_KV_HEAD=${NUM_KV_HEAD:-8} export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} @@ -146,6 +143,10 @@ setArgs() { export gpt_args } +# +---------------------------------------+ +# | 1. Git clone `ezpz` (if not found) | +# | 2. Install `ezpz` (if not installed) | +# +---------------------------------------+ ezpz() { if [[ ! -d ezpz ]]; then git clone https://github.com/saforem2/ezpz @@ -164,6 +165,10 @@ ezpz() { # source ezpz/src/ezpz/bin/getjobenv || exit } +# +------------------------------------------------------------------------+ +# | Save important environment variables to .deepspeed_env, which will be | +# | forwarded to ALL ranks with DeepSpeed | +# +------------------------------------------------------------------------+ saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { @@ -202,8 +207,6 @@ buildDSconfig() { sumWeights() { local file_list=$1 weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") - # weights=$(echo "$weights" | tr ",]" "]") - # echo "weights: $weights" python3 -c "import numpy as np; print(np.sum(${weights}))" } @@ -298,52 +301,6 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- echo "--------------------" } -# buildCLIargs() { # ---- [BROKEN] ------------------------------------------- -# custom_args=" $@" -# export CLI_ARGS=" -# --$DTYPE \ -# --num-workers 0 \ -# --split 100,0,0 \ -# --log-interval 1 \ -# --use-flash-attn-v2 \ -# --no-bias-gelu-fusion \ -# --lr-decay-style cosine \ -# --no-bias-dropout-fusion \ -# --no-masked-softmax-fusion \ -# --tokenizer-type Llama2Tokenizer \ -# --no-gradient-accumulation-fusion \ -# --accumulate-allreduce-grads-in-fp32 \ -# --use-checkpoint-opt_param-scheduler \ -# --lr ${LR} \ -# --save ${CKPT_DIR} \ -# --load ${CKPT_DIR} \ -# --seq-length ${SEQ} \ -# --num-layers ${NLAYERS} \ -# --hidden-size ${HIDDEN} \ -# --train-iters ${TRAIN_ITER} \ -# --eval-iters ${EVAL_ITERS} \ -# --distributed-backend ${BE} \ -# --num-attention-heads ${HEADS} \ -# --save-interval ${SAVE_INTERVAL} \ -# --eval-interval ${EVAL_INTERVAL} \ -# --max-position-embeddings ${SEQ} \ -# --micro-batch-size ${MICRO_BATCH} \ -# --data-file-list ${DATA_FILE_LIST} \ -# --tensor-model-parallel-size ${TP} \ -# --global-batch-size ${GLOBAL_BATCH} \ -# --pipeline-model-parallel-size ${PP} \ -# --num-key-value-heads ${NUM_KV_HEAD} \ -# --data-cache-path ${DATA_CACHE_PATH} \ -# --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ -# --tokenizer-model ${TOKENIZER_MODEL} \ -# $ds_args \ -# ${LLAMA_ARGS} \ -# ${gpt_args[*]} \ -# ${custom_args} \ -# " -# } - - printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } From 981b7d9896e4c416a842e3b3af09b3f36d46f889 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 08:39:35 -0500 Subject: [PATCH 065/590] Update logging in `megatron/training.py` --- megatron/training.py | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 67cd47bda15..2bdf61f9084 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1182,21 +1182,21 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration, args.consumed_train_samples) writer.add_scalar('iteration-time/iteration-time vs tokens', elapsed_time_per_iteration, args.consumed_train_tokens) - log_string = f' iteration {iteration:8d}/{args.train_iters:8d} |' + log_string = f' iteration={iteration:8d}/{args.train_iters:8d} |' # .format( iteration, args.train_iters) log_string += ( - f' consumed samples={args.consumed_train_samples:12d} |' + f' consumed_samples={args.consumed_train_samples:12d} |' # .format(args.consumed_train_samples) ) - log_string += f' consumed tokens={args.consumed_train_tokens:12d} |' + log_string += f' consumed_tokens={args.consumed_train_tokens:12d} |' # .format( args.consumed_train_tokens) log_string += ( - ' elapsed time per iteration (ms): ' + ' elapsed_time_per_iteration_ms=' f'{elapsed_time_per_iteration * 1000.0:.1f} |' # .format( elapsed_time_per_iteration * 1000.0) ) - log_string += f' learning rate={learning_rate:.3f} |' - log_string += f' global batch size={batch_size:5d} |' + log_string += f' learning_rate={learning_rate:.6f} |' + log_string += f' global_batch_size={batch_size:5d} |' # if wandb is not None and getattr(wandb, 'run', None) is not None: wandb_metrics |= { 'training/iteration': iteration, @@ -1221,28 +1221,28 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, log_string += ' {}={:.6f} |'.format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) if loss_scale is not None: - log_string += ' loss scale={:.1f} |'.format(loss_scale) + log_string += ' loss_scale={:.1f} |'.format(loss_scale) wandb_metrics |= {'loss/loss_scale': loss_scale} if grad_norm is not None: - log_string += ' grad norm={:.3f} |'.format(grad_norm) + log_string += ' grad_norm={:.3f} |'.format(grad_norm) wandb_metrics |= {'loss/grad_norm': grad_norm} if num_zeros_in_grad is not None: - log_string += ' num zeros={:.1f} |'.format(num_zeros_in_grad) + log_string += ' num_zeros={:.1f} |'.format(num_zeros_in_grad) wandb_metrics |= {'loss/num_zeros_in_grad': num_zeros_in_grad} if params_norm is not None: - log_string += ' params norm={:.3f} |'.format(params_norm) + log_string += ' params_norm={:.3f} |'.format(params_norm) wandb_metrics |= {'loss/params_norm': params_norm} if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - log_string += ' curriculum seqlen={:5d} |'.format(args.curriculum_seqlen) + log_string += ' curriculum_seqlen={:5d} |'.format(args.curriculum_seqlen) if args.random_ltd: - log_string += ' random ltd reserved length={:5d} |'.format(args.random_ltd_reserved_length) - log_string += ' actual seqlen={:5d} |'.format(seq_len) - log_string += ' number of skipped iterations={:3d} |'.format( + log_string += ' random_ltd reserved_length={:5d} |'.format(args.random_ltd_reserved_length) + log_string += ' actual_seqlen={:5d} |'.format(seq_len) + log_string += ' number_of_skipped_iterations={:3d} |'.format( total_loss_dict[skipped_iters_key]) - log_string += ' number of nan iterations={:3d} |'.format( + log_string += ' number_of_nan_iterations={:3d} |'.format( total_loss_dict[nan_iters_key]) - log_string += ' samples per second={:.3f} |'.format(samples_per_sec) - log_string += ' tokens per gpu per second (tgs)={:.3f} |'.format(tokens_per_gpu_per_second) + log_string += ' samples_per_second={:.3f} |'.format(samples_per_sec) + log_string += ' tokens_per_gpu_per_second_tgs={:.3f} |'.format(tokens_per_gpu_per_second) log_string += ' TFLOPs={:.2f} |'.format(tflops) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 From 7e6a3a42a8a8707fee77c844ce83015c6ce9e4d5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 08:40:44 -0500 Subject: [PATCH 066/590] Remove redundant `train_llama_alcf_*.sh` --- train_llama_alcf_aurora.sh | 234 ------------------------------------ train_llama_alcf_polaris.sh | 116 ------------------ train_llama_alcf_sunspot.sh | 158 ------------------------ 3 files changed, 508 deletions(-) delete mode 100644 train_llama_alcf_aurora.sh delete mode 100644 train_llama_alcf_polaris.sh delete mode 100644 train_llama_alcf_sunspot.sh diff --git a/train_llama_alcf_aurora.sh b/train_llama_alcf_aurora.sh deleted file mode 100644 index 48651dbeb1c..00000000000 --- a/train_llama_alcf_aurora.sh +++ /dev/null @@ -1,234 +0,0 @@ -#!/bin/bash --login -#PBS -l walltime=06:00:00 -#PBS -A argonne_tpc -#PBS -q prod -#PBS -l select=48 -#PBS -l filesystems=eagle:home -# - -function sourceFile() { - fp="$1" - if [[ -f "${fp}" ]]; then - echo "Found ${fp}, \`source\`-ing" - # shellcheck source="${fp}" - source "${fp}" - else - echo "ERROR: UNABLE TO SOURCE ${fp}" - fi -} - -# +++++++++++++++ SCRIPT START ++++++++++++++++++++++ -# ---- source ./helpers_alcf.sh --------------------- -cd "${PBS_O_WORKDIR}" || exit -HERE=$(python3 -c 'import os; print(os.getcwd())') -sourceFile "${HERE}/ALCF_utils/helpers_alcf.sh" || exit -# cd ~/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed || exit -# eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate anl_release_q4v2 -ezpz || exit -setEnv || exit -saveDSenv || exit -makeHostfiles || exit -setupData "${DATA_FILE_LIST:-${HERE}/data_file_list_reweighted.txt}" || exit -# dfl_fallback="${HERE}/data_file_list_shuf_debug.txt" - -# # ---- DATA SETUP ------------------------------------ -# dfl_debug="./data_file_list_shuf_debug.txt" -# DATA_FILE_LIST="${DATA_FILE_LIST:-${dfl_debug}}" && export DATA_FILE_LIST="${DATA_FILE_LIST}" -# NUM_DOCS=$(wc -l < "${DATA_FILE_LIST}") && export NUM_DOCS="${NUM_DOCS}" -# WEIGHT_SUM="$(sumWeights "${DATA_FILE_LIST}")" && export WEIGHT_SUM="${WEIGHT_SUM}" -# DFL_STEM=$(echo "$DATA_FILE_LIST" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") && export DFL_STEM="${DFL_STEM}" -# dcp="${HERE}/.cache/${DFL_STEM}-index-cache" -# DATA_CACHE_PATH="${DATA_CACHE_PATH:-${dcp}}" && export DATA_CACHE_PATH="${DATA_CACHE_PATH}" -# mkdir -p "${DATA_CACHE_PATH}" -# if [[ -n "${DOLMA_CHUNK_IDX}" ]]; then -# echo "Using DOLMA CHUNK ${DOLMA_CHUNK_IDX} from ${DATA_FILE_LIST} with ${NUM_DOCS} documents..." -# else -# echo "Using NUM_DOCS=${NUM_DOCS} documents from DATA_FILE_LIST=${DATA_FILE_LIST}" -# fi - - -# ---- Parallelism Settings -------------------------- -PP=${PP:-1} -TP=${TP:-1} -export PP="${PP}" -export TP="${TP}" -export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" -export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} -# export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${PBS_NODEFILE}")} -# ---------------------------------------------------- - -# ---- Llama2 7B Config ----------------------- -export HEADS=${HEADS:-32} -export NLAYERS=${NLAYERS:-32} -export HIDDEN=${HIDDEN:-4096} -export NUM_KV_HEAD=${NUM_KV_HEAD:-8} -export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" -# --------------------------------------------- - -# ---- Run Settings --------------------------- -export LR=${LR:-0.0003} -export SEQ=${SEQ:-4096} -export DTYPE=${DTYPE:-bf16} -export ZERO_STAGE=${ZERO_STAGE:-2} -export MICRO_BATCH=${MICRO_BATCH:-4} -export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} -export TRAIN_ITER=${TRAIN_ITER:-317892} -export SAVE_INTERVAL=${SAVE_INTERVAL:-200} -export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} -export GLOBAL_BATCH=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) -export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-0} -export TOKENIZER_MODEL="/lus/gecko/projects/Aurora_deployment/AuroraGPT/datasets/dolma/utils/tokenizer.model" -# export EXTRA_ARGS="" -export LLAMA_ARGS="--no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" -# --------------------------------------------- - -# ---- Build DeepSpeed Config --------------------------------- -export DS_CONFIG="ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" -bash "${HERE}/generate_config.sh" "${DS_CONFIG}" || exit -# ------------------------------------------------------------- - - -# ---- Specify output location -------------------------------- -export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}" -# OUTPUT_DIR=logs/ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_`date +%m%d%H%M%S`_${HOSTNAME} -OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" -export OUTPUT_DIR="${OUTPUT_DIR}" -export OUTPUT_LOG="${OUTPUT_DIR}/output.log" -export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" -echo "${OUTPUT_LOG}" >> "logs/latest" -mkdir -p "${OUTPUT_DIR}" -echo "!!!Please see logs at ${OUTPUT_DIR}" - - -gpt_args=() -ds_args=" " -ds_args=" --deepspeed ${ds_args}" -if [ "$PP" == 1 ]; then - ds_args=" --no-pipeline-parallel ${ds_args}" -fi -ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" -ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" - -# BUG: [???] ---------------------------------------------------------------- -# I dont know where this came from... -# > we are now using activation checkpoint provided by megatron, see below. -# --------------------------------------------------------------------------- -# -# NOTE: [???] --------------------------------------------------------------- -# In `train_llama_alcf_polaris.sh` we also pass -# `"--checkpoint-num-layers 1"` -# ---------------------------------------------------------------------------- -if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then - echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - ds_args=" --deepspeed-activation-checkpointing ${ds_args}" - gpt_args+=( - "--checkpoint-activations" - ) - # "--checkpoint-num-layers 1" - # --checkpoint-activations \ - # --deepspeed-activation-checkpointing -fi - -# take custom args -custom_args=" $@" - -# Ensure `./hostfile_deepspeed` and `./hostfile_mpich` exist in $(pwd) -hfds="${HERE}/hostfile_deepspeed" -hfmpi="${HERE}/hostfile_mpich" -[ -f "$hfds" ] || exit -[ -f "$hfmpi" ] || exit - -# launcher setting -LAUNCHER=${LAUNCHER:-MPICH} -if [[ $LAUNCHER == "deepspeed" ]]; then - launcher="" -else - launcher="--force_multi --hostfile ${hfds} --launcher=${LAUNCHER} --launcher_args='-hostfile ${hfmpi}'" -fi - - -if [[ $(hostname) == x4* ]]; then - CCL=${CCL:-ccl} - BE="${CCL}" -elif [[ $(hostname) == x3* ]]; then - NCCL=${NCCL:-nccl} - BE="${NCCL}" -fi -# NCCL=${NCCL:-nccl} -EXEC=pretrain_gpt_alcf.py - -# MODEL=LLAMA_7B -# OUTPUT_PREFIX=${MODEL}_z${ZERO_STAGE}_seqlen_tp${TP}_pp${PP}_sp${SP}_nl${NUM_LAYERS}_hs${HIDDEN_SIZE}_gb${BS}_mb${MBS} -echo "++++++++++++++++++++++++++++++++++++++++++++++++++" -echo "- WORLD_SIZE:${WORLD_SIZE}" -echo "- BACKEND: ${BE}" -echo "- MODEL_TYPE: ${MODEL_TYPE}" -echo "- DOCUMENT WEIGHT_SUM: ${WEIGHT_SUM}" -echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST}" -echo "- Using NUM_DOCS=${NUM_DOCS} documents from DATA_FILE_LIST=${DATA_FILE_LIST}" -echo "++++++++++++++++++++++++++++++++++++++++++++++++++" - -run_cmd=" - deepspeed $launcher ${EXEC} \ - --use-flash-attn \ - --num-key-value-heads ${NUM_KV_HEAD} \ - --tensor-model-parallel-size $TP \ - --pipeline-model-parallel-size $PP \ - --num-layers $NLAYERS \ - --hidden-size $HIDDEN \ - --num-attention-heads $HEADS \ - --seq-length $SEQ \ - --max-position-embeddings $SEQ \ - --micro-batch-size $MICRO_BATCH \ - --global-batch-size $GLOBAL_BATCH \ - --train-iters $TRAIN_ITER \ - --lr ${LR} \ - --lr-decay-style cosine \ - --log-interval 1 \ - --save-interval ${SAVE_INTERVAL} \ - --split 100,0,0 \ - --$DTYPE \ - --no-masked-softmax-fusion \ - --no-bias-gelu-fusion \ - --no-bias-dropout-fusion \ - --no-gradient-accumulation-fusion \ - --distributed-backend ${BE} \ - --tokenizer-type Llama2Tokenizer \ - --save checkpoints/${OUTPUT_PREFIX} \ - --load checkpoints/${OUTPUT_PREFIX} \ - --use-checkpoint-opt_param-scheduler \ - --tokenizer-model ${TOKENIZER_MODEL} \ - --data-file-list ${DATA_FILE_LIST} \ - --data-cache-path ${DATA_CACHE_PATH} \ - $ds_args \ - ${LLAMA_ARGS} \ - ${gpt_args[*]} \ - $custom_args \ - |& tee ${OUTPUT_LOG} - " - # >> ${OUTPUT_LOG} 2>&1 & - # |& tee $OUTPUT_DIR/output.log - -# --ffn-hidden-size 11008 \ -# --vocab-file $VOCAB_FILE \ -# --merge-file $MERGE_FILE \ -# --lr-decay-iters 320000 \ -# --num-workers 0 \ -# --eval-iters ${EVAL_ITERS} \ -# --eval-interval ${EVAL_INTERVAL} \ -# --lr-warmup-iters 5000 \ -# --lr-decay-iters 10000 \ -# --accumulate-allreduce-grads-in-fp32 \ -# --data-impl mmap \ - -echo "All DeepSpeed(s): $(which -a deepspeed)" -echo "Using $(which deepspeed)" -ds_report - -echo "${run_cmd}" - -printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" -printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" - -eval "${run_cmd}" -set +x diff --git a/train_llama_alcf_polaris.sh b/train_llama_alcf_polaris.sh deleted file mode 100644 index 06f268f10e0..00000000000 --- a/train_llama_alcf_polaris.sh +++ /dev/null @@ -1,116 +0,0 @@ -#!/bin/bash --login -#PBS -l walltime=06:00:00 -#PBS -A argonne_tpc -#PBS -q prod -#PBS -l select=48 -#PBS -l filesystems=eagle:home - -function sourceFile() { - fp="$1" - echo "source-ing ${fp}" - if [[ -f "${fp}" ]]; then - # shellcheck source="${fp}" - source "${fp}" - else - echo "ERROR: UNABLE TO SOURCE ${fp}" - fi -} - -# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ -# ---- 0. Navigate into `$PBS_O_WORKDIR` ------------------------------------- -cd "${PBS_O_WORKDIR}" || exit -HERE=$(python3 -c 'import os; print(os.getcwd())') -export HERE -# ---- 1. Assert `./pretrain_gpt_alcf.py` exists: ----------------------------- -export EXEC="${HERE}/pretrain_gpt_alcf.py" -[ -f "${EXEC}" ] || exit -# ---- 2. `source ./ALCF/helpers_alcf.sh`: ------------------------------------ -sourceFile "${HERE}/ALCF/helpers.sh" || exit -# ---- 3. Call fns from `./ALCF/helpers_alcf.sh` ------------------------------ -setEnv || exit # 1. load `conda` environment -saveDSenv || exit # 2. save env vars to `.deepspeed_env` -ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars -makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` -buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ -setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} -setArgs || exit # 8. specify additional `deepspeed` arguments -setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset -setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` -printJobInfo || exit # 11. print job info -# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - -# Take custom args -custom_args=" $@" - -# Assert `./hostfile_deepspeed` exists -export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit -TBDIR="${CKPT_DIR}/tensorboard" -mkdir -p "${TBDIR}" - -# source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit -# echo "Using $(which python3)" -# --launcher_args='--pmi=pmix' - # deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ - # ${launch_cmd} \ - # --optimizer adam \ - # --use-flash-attn-v2 \ - # --num-workers 0 \ -run_cmd=" - deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - --$DTYPE \ - --optimizer ${OPT} \ - --split 100,0,0 \ - --log-interval 1 \ - --no-bias-gelu-fusion \ - --lr-decay-style cosine \ - --no-bias-dropout-fusion \ - --no-masked-softmax-fusion \ - --tokenizer-type Llama2Tokenizer \ - --no-gradient-accumulation-fusion \ - --accumulate-allreduce-grads-in-fp32 \ - --use-checkpoint-opt_param-scheduler \ - --tensorboard-dir ${TBDIR} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ - --lr ${LR} \ - --save ${CKPT_DIR} \ - --load ${CKPT_DIR} \ - --seq-length ${SEQ} \ - --num-layers ${NLAYERS} \ - --hidden-size ${HIDDEN} \ - --train-iters ${TRAIN_ITER} \ - --eval-iters ${EVAL_ITERS} \ - --distributed-backend ${BE} \ - --num-attention-heads ${HEADS} \ - --save-interval ${SAVE_INTERVAL} \ - --eval-interval ${EVAL_INTERVAL} \ - --max-position-embeddings ${SEQ} \ - --micro-batch-size ${MICRO_BATCH} \ - --data-file-list ${DATA_FILE_LIST} \ - --tensor-model-parallel-size ${TP} \ - --global-batch-size ${GLOBAL_BATCH} \ - --pipeline-model-parallel-size ${PP} \ - --num-key-value-heads ${NUM_KV_HEAD} \ - --data-cache-path ${DATA_CACHE_PATH} \ - --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - --tokenizer-model ${TOKENIZER_MODEL} \ - ${LLAMA_ARGS} \ - $ds_args \ - ${gpt_args[*]} \ - $custom_args \ - |& tee ${OUTPUT_LOG} - " - - -# echo "All DeepSpeed(s): $(which -a deepspeed)" -echo "! Using $(which deepspeed)" -ds_report - -echo "${run_cmd}" - -printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" -printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" -# echo "${OUTPUT_LOG}" -eval "${run_cmd}" -set +x diff --git a/train_llama_alcf_sunspot.sh b/train_llama_alcf_sunspot.sh deleted file mode 100644 index 700f5dbfdbf..00000000000 --- a/train_llama_alcf_sunspot.sh +++ /dev/null @@ -1,158 +0,0 @@ -#!/bin/bash --login -#PBS -l walltime=06:00:00 -#PBS -A argonne_tpc -#PBS -q prod -#PBS -l select=48 -#PBS -l filesystems=eagle:home - -function sourceFile() { - fp="$1" - echo "source-ing ${fp}" - if [[ -f "${fp}" ]]; then - # shellcheck source="${fp}" - source "${fp}" - else - echo "ERROR: UNABLE TO SOURCE ${fp}" - fi -} - -module () { - if [ -z "${LMOD_SH_DBG_ON+x}" ] - then - case "$-" in - (*v*x*) __lmod_sh_dbg='vx' ;; - (*v*) __lmod_sh_dbg='v' ;; - (*x*) __lmod_sh_dbg='x' ;; - esac - fi - if [ -n "${__lmod_sh_dbg:-}" ] - then - set +$__lmod_sh_dbg - echo "Shell debugging temporarily silenced: export LMOD_SH_DBG_ON=1 for Lmod's output" >&2 - fi - eval "$($LMOD_CMD $LMOD_SHELL_PRGM "$@")" && eval "$(${LMOD_SETTARG_CMD:-:} -s sh)" - __lmod_my_status=$? - if [ -n "${__lmod_sh_dbg:-}" ] - then - echo "Shell debugging restarted" >&2 - set -$__lmod_sh_dbg - fi - unset __lmod_sh_dbg - return $__lmod_my_status -} - - -# +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ -# ---- 0. Navigate into `$PBS_O_WORKDIR` ------------------------------------- -cd "${PBS_O_WORKDIR}" || exit -HERE=$(python3 -c 'import os; print(os.getcwd())') -export HERE -# ---- 1. Assert `./pretrain_gpt_alcf.py` exists: ----------------------------- -export EXEC="${HERE}/pretrain_gpt_alcf.py" -[ -f "${EXEC}" ] || exit -# ---- 2. `source ./ALCF/helpers_alcf.sh`: ------------------------------------ -sourceFile "${HERE}/ALCF/helpers.sh" || exit -# ---- 3. Call fns from `./ALCF/helpers_alcf.sh` ------------------------------------------------------------------ -setEnv || exit # 1. load `conda` environment -saveDSenv || exit # 2. save env vars to `.deepspeed_env` -ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars -makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` -buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ -setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} -setArgs || exit # 8. specify additional `deepspeed` arguments -setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset -setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` -printJobInfo || exit # 11. print job info -# ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ - -# Take custom args -custom_args=" $@" - -# Assert `./hostfile_deepspeed` exists -export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit - -TBDIR="${CKPT_DIR}/tensorboard" -mkdir -p "${TBDIR}" - - # --use-flash-attn-v2 \ - # --use-flash-attn \ - # --$DTYPE \ - # --optimizer adamw \ - # --adam-beta1 0.9 \ - # --adam-beta2 0.95 \ -run_cmd=" - deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - --${DTYPE} \ - --optimizer ${OPT} \ - --num-workers 0 \ - --split 100,0,0 \ - --log-interval 1 \ - --no-bias-gelu-fusion \ - --lr-decay-style cosine \ - --no-bias-dropout-fusion \ - --no-masked-softmax-fusion \ - --tokenizer-type Llama2Tokenizer \ - --no-gradient-accumulation-fusion \ - --accumulate-allreduce-grads-in-fp32 \ - --use-checkpoint-opt_param-scheduler \ - --tensorboard-dir ${TBDIR} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ - --lr ${LR} \ - --save ${CKPT_DIR} \ - --load ${CKPT_DIR} \ - --seq-length ${SEQ} \ - --num-layers ${NLAYERS} \ - --hidden-size ${HIDDEN} \ - --train-iters ${TRAIN_ITER} \ - --eval-iters ${EVAL_ITERS} \ - --distributed-backend ${BE} \ - --num-attention-heads ${HEADS} \ - --save-interval ${SAVE_INTERVAL} \ - --eval-interval ${EVAL_INTERVAL} \ - --max-position-embeddings ${SEQ} \ - --micro-batch-size ${MICRO_BATCH} \ - --data-file-list ${DATA_FILE_LIST} \ - --tensor-model-parallel-size ${TP} \ - --global-batch-size ${GLOBAL_BATCH} \ - --pipeline-model-parallel-size ${PP} \ - --num-key-value-heads ${NUM_KV_HEAD} \ - --data-cache-path ${DATA_CACHE_PATH} \ - --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - --tokenizer-model ${TOKENIZER_MODEL} \ - ${LLAMA_ARGS} \ - $ds_args \ - ${gpt_args[*]} \ - $custom_args \ - |& tee ${OUTPUT_LOG} - " - - # --------------------------------------------------- - # --vocab-file $VOCAB_FILE \ - # --merge-file $MERGE_FILE \ - # --lr-decay-iters 320000 \ - # --lr-warmup-iters 5000 \ - # --lr-decay-iters 10000 \ - # --num-workers 4 \ - # launch python3 ${EXEC} \ - # --data-impl mmap \ - # source ./ezpz/src/ezpz/bin/getjobenv || exit - # --------------------------------------------------- - # ${DIST_LAUNCH} ./local_rank.sh python3 ${EXEC} \ - # ${DIST_LAUNCH} python3 ${EXEC} \ - # deepspeed $launcher ${EXEC} \ - # >> ${OUTPUT_LOG} 2>&1 & - # >> ${OUTPUT_LOG} 2>&1 & - # |& tee $OUTPUT_DIR/output.log - # ${EXTRA_ARGS} \ - -echo "! Using $(which deepspeed)" -ds_report - -echo "${run_cmd}" -printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" -printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" -# echo "${OUTPUT_LOG}" -eval "${run_cmd}" -set +x From d243489c1c6b0e41e39e3e81d78b42a04538a849 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 11:27:20 -0500 Subject: [PATCH 067/590] Update README.md --- ALCF/README.md | 463 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 461 insertions(+), 2 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index d95c70125d1..5f0fd1b6bd5 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,6 +1,9 @@ # Megatron-DeepSpeed @ ALCF -## Polaris +## ✅ TODOs + +
+TODOs: - [ ] Ensure / double check that optimizer settings from `ds_config.json` aren't being overwritten by some defaults in `megatron/arguments.py` - [ ] specifically, `momentum, beta{1, 2}, etc` @@ -69,6 +72,456 @@
+ + +## 🏃‍♂️ Running + +> [!NOTE] +> [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching +> distributed training on {Polaris, Aurora, Sunspot} @ ALCF. + +To launch: + +```bash +$ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I +$ cd /path/to/Megatron-DeepSpeed/ +# load your favorite {conda, venv} environment, requires: {pytorch, deepspeed} +# e.g. on Polaris: +$ module load conda/2023-10-04 # ; conda activate cu118-pt221 ; unset PYTHONUSERBASE +$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./convergence_debug_small.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh +``` + +
[output]: + +
+ +
[Sunspot]: + +```bash +# [09:07:32 AM] [foremans@x1921c0s0b0n0] ~/q/llm.devkit/Megatron-DeepSpeed  main !1 ?27 q4-drop 26s ✘ INT +$ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./convergence_debug_small.txt bash train_llama_alcf.sh +source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh +Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... + UMD: agama-ci-devel-736.9 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-736.9 +Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" + +Please check the spelling or version number. Also try "module spider ..." +It is also possible your cache file is out-of-date; it may help to try: + $ module --ignore_cache load "gcc/12.1.0" + +Also make sure that all modulefiles written in TCL start with the string #%Module + +Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. + +Running on SunSpot !! +[python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 +Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env +Found ezpz! +/lus/gila/projects/Aurora_deployment/foremans/locations/sunspot/projects/saforem2/ezpz/src/ezpz/__init__.py +Has ezpz installed. Nothing to do. +Done with ezpz. +┌─────────────────────────────────────────────────────────────────── +│ Writing PBS vars to /home/foremans/.pbsenv +│ HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 +│ NHOSTS: 2 +│ NGPU_PER_HOST: 12 GPUs per host +│ NGPUS: 24 GPUs total +└─────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [Hosts]: +│ • [host:0] - x1921c0s0b0n0.hostmgmt2000.cm.americas.sgi.com +│ • [host:1] - x1921c0s1b0n0.hostmgmt2000.cm.americas.sgi.com +└────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [DIST INFO]: +│ • Loading job env from: /home/foremans/.pbsenv +│ • HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 +│ • NHOSTS: 2 +│ • NGPU_PER_HOST: 12 +│ • NGPUS (NHOSTS x NGPU_PER_HOST): 24 +│ • WORLD_SIZE: 24 +│ • DIST_LAUNCH: mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001 +└────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [Launch]: +│ • Use: 'launch' (=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001) +│ to launch job +└────────────────────────────────────────────────────────────────── +DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json +ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 +!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! +!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! +Calling: `setData()` with ./convergence_debug_small.txt +-------------------- +Updated environment: +DATA_FILE_LIST: ./convergence_debug_small.txt +NUM_DOCS: 15 + WEIGHT_SUM: 15.0 +DFL_STEM: convergence_debug_small +DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache +-------------------- +++++++++++++++++++++++++++++++++++++++++++++++++++ +- MPICH_DIR= +- Using /home/foremans/miniconda3/envs/q4-drop/bin/python3 +- WORLD_SIZE:24 +- NCCL: nccl +- MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden +- Using DATA_FILE_LIST: ./convergence_debug_small.txt +++++++++++++++++++++++++++++++++++++++++++++++++++ +! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed +/home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html + __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') +/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you don't plan on using image function +ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch +vision` from source? + warn( +[2024-04-04 09:07:45,585] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) +[2024-04-04 09:07:45,818] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) +-------------------------------------------------- +DeepSpeed C++/CUDA extension op report +-------------------------------------------------- +NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. +-------------------------------------------------- +JIT compiled ops requires ninja +ninja .................. [OKAY] +-------------------------------------------------- +op name ................ installed .. compatible +-------------------------------------------------- +async_io ............... [NO] ....... [OKAY] +cpu_adagrad ............ [NO] ....... [OKAY] +cpu_adam ............... [NO] ....... [OKAY] +flash_attn ............. [NO] ....... [OKAY] +fused_adam ............. [NO] ....... [OKAY] +quantizer .............. [NO] ....... [OKAY] +transformer ............ [NO] ....... [OKAY] +transformer_inference .. [NO] ....... [OKAY] +utils .................. [NO] ....... [OKAY] +-------------------------------------------------- +DeepSpeed general environment info: +torch install path ............... ['/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch'] +torch version .................... 2.1.0a0+cxx11.abi +deepspeed install path ........... ['/lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/DeepSpeed/deepspeed'] +deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD +deepspeed wheel compiled w. ...... torch 2.1 +shared memory (/dev/shm) size .... 503.18 GB + + deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ +foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay +-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 + --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/tensorboard --log-timers-to-tensorboard --log-optimizer +-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 + --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend ccl --num-attention-heads 32 --save-interval 20 +0 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 4 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 1 --global-bat +ch-size 96 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ +.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/tokenizer.model --no-query- +key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z +ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 +_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + +[!! NOTE] View output at: +logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + +# ... + +/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.010017 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_doc_idx.npy + > loading sample-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_sample_idx.npy + > loading shuffle-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_shuffle_idx.npy + loaded indexed file in 0.056 seconds + total number of samples: 2318461 + total number of epochs: 8 +> loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy +> loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy +/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. + +[after dataloaders are built] datetime: 2024-04-04 09:09:27 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (64818.18, 64858.22) + train/valid/test-data-iterators-setup ..........: (1968.10, 2288.56) +training ... +[before the start of training step] datetime: 2024-04-04 09:09:27 +[2024-04-04 09:09:27,718] [INFO] [checkpointing.py:540:forward] Activation Checkpointing Information +[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:541:forward] ----Partition Activations False, CPU CHECKPOINTING False +[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:542:forward] ----contiguous Memory Checkpointing False with 32 total layers +[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:544:forward] ----Synchronization False +[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:545:forward] ----Profiling time in checkpointing False +[2024-04-04 09:09:33][INFO][utils:145] - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. +[2024-04-04 09:09:33][INFO][utils:148] - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. +[2024-04-04 09:09:33][INFO][utils:160] - NumExpr defaulting to 8 threads. +^[c[2024-04-04 09:09:53,311] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 884.11 | optimizer_gradients: 6.43 | optimizer_step: 23.44 +[2024-04-04 09:09:53,312] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] +[2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 6567.68 | bwd_microstep: 17950.36 | bwd_inner_microstep: 17711.20 | bwd_allreduce_microstep: 239.11 | step_microstep: 1139.27 +[2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 6567.66 | bwd: 17950.35 | bwd_inner: 17711.19 | bwd_allreduce: 239.11 | step: 1139.29 +[Rank 0] (after 1 iterations) memory (MB) | allocated: 18244.640625 | max allocated: 41299.50146484375 | reserved: 46764.0 | max reserved: 46764.0 + iteration 1/ 317892 | consumed samples: 96 | consumed tokens: 393216 | elapsed time per iteration (ms): 25849.1 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 1.117136E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.714 | tokens per gpu per second(tgs): 633.832 | TFLOPs: 38.61 | +[2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 327.85 | optimizer_gradients: 6.26 | optimizer_step: 23.60 +[2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] +[2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 4022.74 | bwd_microstep: 15738.67 | bwd_inner_microstep: 15556.80 | bwd_allreduce_microstep: 181.82 | step_microstep: 371.01 +[2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 + iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | +``` + +
+ +
[Polaris]: + +```bash +[09:31:35 AM] [foremans@x3112c0s13b0n0] ~/pol/p/a/Megatron-DeepSpeed  main !4 ?24 cu118-pt221 ✘ INT +$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./convergence_debug_small.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh +source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh +Running on Polaris !! + +[python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 +Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env +Found ezpz! +/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/ezpz/src/ezpz/__init__.py +Has ezpz installed. Nothing to do. +Done with ezpz. +┌─────────────────────────────────────────────────────────────────── +│ Writing PBS vars to /home/foremans/.pbsenv +│ HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov +│ NHOSTS: 2 +│ NGPU_PER_HOST: 4 GPUs per host +│ NGPUS: 8 GPUs total +└─────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [Hosts]: +│ • [host:0] - x3112c0s13b0n0.hsn.cm.polaris.alcf.anl.gov +│ • [host:1] - x3112c0s13b1n0.hsn.cm.polaris.alcf.anl.gov +└────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [DIST INFO]: +│ • Loading job env from: /home/foremans/.pbsenv +│ • HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov +│ • NHOSTS: 2 +│ • NGPU_PER_HOST: 4 +│ • NGPUS (NHOSTS x NGPU_PER_HOST): 8 +│ • WORLD_SIZE: 8 +│ • DIST_LAUNCH: mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov +└────────────────────────────────────────────────────────────────── +┌────────────────────────────────────────────────────────────────── +│ [Launch]: +│ • Use: 'launch' (=mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov) +│ to launch job +└────────────────────────────────────────────────────────────────── +DS_CONFIG: ds_stage2_mb8_gb32_pp1_bf16.json +ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 +!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! +!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! +Calling: setData() with ./convergence_debug_small.txt +-------------------- +Updated environment: +DATA_FILE_LIST: ./convergence_debug_small.txt +NUM_DOCS: 15 + WEIGHT_SUM: 15.0 +DFL_STEM: convergence_debug_small +DATA_CACHE_PATH: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache +-------------------- +++++++++++++++++++++++++++++++++++++++++++++++++++ +- MPICH_DIR=/opt/cray/pe/mpich/8.1.25/ofi/gnu/9.1 +- Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 +- WORLD_SIZE:8 +- NCCL: nccl +- MODEL_TYPE: llama-seq4096-pp1-tp2-32layers-32heads-4096hidden +- Using DATA_FILE_LIST: ./convergence_debug_small.txt +++++++++++++++++++++++++++++++++++++++++++++++++++ +! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed +[2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda (auto detect) +-------------------------------------------------- +DeepSpeed C++/CUDA extension op report +-------------------------------------------------- +NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. +-------------------------------------------------- +JIT compiled ops requires ninja +ninja .................. [OKAY] +-------------------------------------------------- +op name ................ installed .. compatible +-------------------------------------------------- +async_io ............... [NO] ....... [OKAY] +fused_adam ............. [NO] ....... [OKAY] +cpu_adam ............... [NO] ....... [OKAY] +cpu_adagrad ............ [NO] ....... [OKAY] +cpu_lion ............... [NO] ....... [OKAY] + [WARNING] Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH +evoformer_attn ......... [NO] ....... [NO] +fused_lamb ............. [NO] ....... [OKAY] +fused_lion ............. [NO] ....... [OKAY] +inference_core_ops ..... [NO] ....... [OKAY] +cutlass_ops ............ [NO] ....... [OKAY] +transformer_inference .. [NO] ....... [OKAY] +quantizer .............. [NO] ....... [OKAY] +ragged_device_ops ...... [NO] ....... [OKAY] +ragged_ops ............. [NO] ....... [OKAY] +random_ltd ............. [NO] ....... [OKAY] + [WARNING] sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.2 + [WARNING] using untested triton version (2.2.0), only 1.0.0 is known to be compatible +sparse_attn ............ [NO] ....... [NO] +spatial_inference ...... [NO] ....... [OKAY] +transformer ............ [NO] ....... [OKAY] +stochastic_transformer . [NO] ....... [OKAY] +-------------------------------------------------- +DeepSpeed general environment info: +torch install path ............... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/torch'] +torch version .................... 2.2.1 +deepspeed install path ........... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/deepspeed'] +deepspeed info ................... 0.14.0, unknown, unknown +torch cuda version ............... 11.8 +torch hip version ................ None +nvcc version ..................... 11.8 +deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 +shared memory (/dev/shm) size .... 251.61 GB + + deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc +e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion + --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- +grads-in-fp32 --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/tensorboard --log-timers-to-tensorboard - +-log-optimizer-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_ +pp1_tp2_bf16 --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend nccl --num-attention-heads 32 --s +ave-interval 200 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 8 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 2 + --global-batch-size 32 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-l +cf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokeniz +er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 + --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- +num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + +[!! NOTE] View output at: +logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + +# ... + +/eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.001280 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_doc_idx.npy + > loading sample-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_sample_idx.npy + > loading shuffle-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_shuffle_idx.npy + loaded indexed file in 0.004 seconds + total number of samples: 869423 + total number of epochs: 3 +> loading blendable dataset index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_index.npy +> loading blendable dataset sample index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_sample_index.npy +> size of blendable dataset: 10223415 samples +> finished creating GPT datasets ... +[after dataloaders are built] datetime: 2024-04-04 09:36:07 +done with setup ... +(min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (4794.78, 4795.23) + train/valid/test-data-iterators-setup ..........: (589.69, 721.20) +training ... +[before the start of training step] datetime: 2024-04-04 09:36:07 +[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:539:forward] Activation Checkpointing Information +[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:540:forward] ----Partition Activations False, CPU CHECKPOINTING False +[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:541:forward] ----contiguous Memory Checkpointing False with 32 total layers +[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:543:forward] ----Synchronization False +[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:544:forward] ----Profiling time in checkpointing False +[2024-04-04 09:36:28,429] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1626.54 | optimizer_gradients: 19.29 | optimizer_step: 419.48 +[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] +[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 11336.34 | bwd_microstep: 7134.73 | bwd_inner_microstep: 7090.02 | bwd_allreduce_microstep: 44.65 | step_microstep: 2564.02 +[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 11336.33 | bwd: 7134.75 | bwd_inner: 7090.01 | bwd_allreduce: 44.66 | step: 2564.02 + iteration 1/ 317892 | consumed samples: 32 | consumed tokens: 131072 | elapsed time per iteration (ms): 21133.8 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.119983E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 1.514 | tokens per gpu per second(tgs): 775.250 | TFLOPs: 47.23 | +[Rank 1] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 35824.0 +[Rank 0] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 32994.0 +[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1605.55 | optimizer_gradients: 11.56 | optimizer_step: 50.92 +[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] +[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 +[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 + iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | + +# ... +``` + +
+ +
+ +## 📦 Install + +
Install Instructions + +1. Clone [`argonne-lcf/Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed) + + ```bash + $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed + $ cd Megatron-DeepSpeed + ``` + + > [!NOTE] + > In the `conda create` command below, + > you can replace `--name "${DAY}"` with + > `--prefix /path/to/your/conda/envs`, if you prefer: + +2. Create `conda` env: + + ```bash + $ module load conda/2023-10-04 + $ export MPICC="cc -shared -taret-accel=nvidia80" + $ export DAY=$(date "+%Y-%m-%d") + $ export PYTHONUSERBASE="${HOME}/.local/polaris/conda/${DAY}" + $ conda create --solver libmamba -c pytorch -c nvidia --name "${DAY}" "python==3.12" + ``` + +3. Install dependencies: + + ```bash + $ conda activate "${DAY}" # e.g. 2024-03-07 + $ conda install -c pytorch -c nvidia --solver libmamba mpi4py ninja transformers xformers triton pytorch torchvision torchaudio pytorch-cuda=11.8 + $ conda install --solver libmamba mpi4py -c conda-forge -c pytorch -c nvidia + $ python3 -m pip install --upgrade pip pybind11 toolong appdirs wandb sentencepiece ipython setuptools wheel ninja + $ python3 -m pip install --upgrade deepspeed wandb + ``` + + - [`ezpz`](https://github.com/saforem2/ezpz): + +
install: + + ```bash + $ git clone https://github.com/saforem2/ezpz + $ python3 -m pip install -e "ezpz[dev]" + ``` + +
+ + - [**OPTIONAL**] [`NVIDIA/apex`](https://github.com/NVIDIA/apex): + +
install: + + ```bash + $ git clone https://github.com/NVIDIA/apex + $ cd apex + # NOTE: need GCC < 11 for APEX ¯\_(ツ)_/¯ ?? + $ module swap gcc gcc/10.3.0 + $ python3 -m pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./ + ``` + +
+ +
+ + +## 📝 Data Preprocessing + +
Data Pre-Processing: + AuroraGPT is trained on the Dolma dataset (initially v0), now in the process of moving to v6. For more details on the dataset, refer to https://huggingface.co/datasets/allenai/dolma. The dolma dataset downloaded is already preprocessing to remove the duplicates (dedup) and filtering the data (mixing). For more details refer to https://github.com/allenai/dolma/tree/main/docs and https://github.com/vksastry/dolma_alcf/blob/main/ALCF/Readme.md. The data preprocessing of Dolma dataset before training consists of tokenization of the data using a specific tokenizer (LlamaTokenizer is what we are currently using), Use the below script to tokenize the entire dataset. Example shown for Polaris. @@ -207,3 +664,5 @@ The data preprocessing of Dolma dataset before training consists of tokenization cd /eagle/datasets/dolma/utils ./tokenization.sh ``` + +
From 78f378539eb4932a82dced1cfb8262ee608eca28 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 11:45:49 -0500 Subject: [PATCH 068/590] Update README.md --- ALCF/README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 5f0fd1b6bd5..d405986fdf7 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -152,7 +152,7 @@ DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! -Calling: `setData()` with ./convergence_debug_small.txt +Calling: setData() with ./convergence_debug_small.txt -------------------- Updated environment: DATA_FILE_LIST: ./convergence_debug_small.txt @@ -172,7 +172,7 @@ DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/l ! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed /home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') -/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you don't plan on using image function +/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you dont plan on using image function ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch vision` from source? warn( From 2dc5aebb1eb4345c1eab32d8247a813c7c6bba98 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 18 Apr 2024 12:05:35 -0500 Subject: [PATCH 069/590] Add `ALCF/data-lists/polaris/*.txt` --- ALCF/data-lists/polaris/data_file_list_c4.txt | 86 + ALCF/data-lists/polaris/data_file_list_cc.txt | 2878 +++++++++++ .../data_file_list_gutenberg-books.txt | 3 + .../polaris/data_file_list_peS2o.txt | 42 + .../polaris/data_file_list_stack-code.txt | 4435 +++++++++++++++++ .../polaris/data_file_list_wiki-en-simple.txt | 2 + 6 files changed, 7446 insertions(+) create mode 100644 ALCF/data-lists/polaris/data_file_list_c4.txt create mode 100644 ALCF/data-lists/polaris/data_file_list_cc.txt create mode 100644 ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt create mode 100644 ALCF/data-lists/polaris/data_file_list_peS2o.txt create mode 100644 ALCF/data-lists/polaris/data_file_list_stack-code.txt create mode 100644 ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt diff --git a/ALCF/data-lists/polaris/data_file_list_c4.txt b/ALCF/data-lists/polaris/data_file_list_c4.txt new file mode 100644 index 00000000000..9ff6f90ff92 --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_c4.txt @@ -0,0 +1,86 @@ +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0012_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0001_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0073_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0045_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0084_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0065_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0032_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0085_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0064_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0025_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0042_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0055_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0023_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0028_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0036_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0037_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0016_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0000_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0019_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0046_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0059_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0017_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0072_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0033_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0006_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0061_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0071_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0057_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0011_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0047_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0004_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0009_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0070_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0018_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0054_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0049_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0003_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0021_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0051_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0074_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0027_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0050_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0079_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0022_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0030_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0034_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0020_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0035_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0015_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0066_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0044_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0010_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0002_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0041_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0067_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0048_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0013_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0083_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0053_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0008_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0014_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0069_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0056_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0062_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0031_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0007_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0077_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0058_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0076_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0078_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0005_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0081_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0040_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0068_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0075_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0063_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0029_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0039_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0026_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0052_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0024_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0043_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0060_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0082_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0080_text_document +0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0038_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_cc.txt b/ALCF/data-lists/polaris/data_file_list_cc.txt new file mode 100644 index 00000000000..4a20a99b89c --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_cc.txt @@ -0,0 +1,2878 @@ +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0553_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0299_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0366_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0753_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0429_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0372_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0124_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0437_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0053_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0615_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0182_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0713_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0688_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0166_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0768_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0692_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0041_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0416_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0630_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0639_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0225_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0035_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0365_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0368_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0196_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0328_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0624_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0081_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0488_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0189_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0118_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0150_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0314_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0209_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0229_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0265_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0532_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0478_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0140_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0256_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0047_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0607_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0023_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0111_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0613_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0748_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0000_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0127_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0106_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0563_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0577_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0502_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0705_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0538_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0088_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0263_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0460_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0571_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0653_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0172_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0524_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0652_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0322_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0447_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0387_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0612_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0290_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0339_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0487_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0396_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0178_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0091_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0193_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0408_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0496_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0755_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0773_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0547_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0384_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0574_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0533_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0464_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0489_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0050_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0060_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0114_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0033_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0561_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0208_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0233_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0744_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0326_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0313_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0482_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0436_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0588_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0080_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0660_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0038_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0282_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0745_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0406_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0116_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0059_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0503_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0357_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0171_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0770_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0286_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0544_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0698_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0155_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0341_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0463_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0676_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0595_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0174_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0198_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0480_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0687_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0145_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0004_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0583_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0449_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0204_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0715_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0521_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0320_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0568_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0151_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0197_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0709_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0499_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0006_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0269_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0525_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0413_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0656_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0646_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0246_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0535_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0333_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0238_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0241_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0469_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0689_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0403_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0404_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0360_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0191_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0236_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0032_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0445_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0614_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0490_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0651_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0703_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0702_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0623_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0719_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0728_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0031_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0253_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0551_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0327_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0027_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0491_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0395_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0473_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0662_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0312_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0605_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0455_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0580_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0005_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0311_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0305_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0260_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0566_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0670_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0129_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0742_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0549_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0058_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0501_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0071_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0450_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0375_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0131_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0697_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0415_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0560_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0643_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0699_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0515_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0739_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0092_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0046_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0083_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0443_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0746_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0655_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0427_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0603_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0367_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0318_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0520_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0749_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0771_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0369_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0434_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0602_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0349_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0763_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0731_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0338_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0462_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0347_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0649_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0194_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0134_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0734_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0632_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0280_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0184_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0089_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0095_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0555_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0016_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0168_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0665_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0767_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0666_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0737_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0037_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0648_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0064_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0764_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0323_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0009_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0545_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0212_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0015_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0267_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0727_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0661_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0211_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0220_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0278_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0721_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0718_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0207_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0619_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0400_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0754_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0610_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0358_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0758_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0298_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0756_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0729_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0468_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0397_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0247_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0149_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0119_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0010_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0093_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0386_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0045_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0066_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0393_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0600_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0440_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0350_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0214_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0714_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0161_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0775_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0203_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0077_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0332_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0700_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0123_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0024_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0013_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0587_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0148_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0513_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0674_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0188_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0599_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0158_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0425_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0003_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0534_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0254_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0121_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0099_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0373_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0479_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0379_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0344_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0684_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0720_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0391_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0575_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0319_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0336_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0531_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0474_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0432_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0766_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0342_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0476_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0237_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0061_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0250_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0752_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0329_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0376_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0640_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0634_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0682_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0181_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0076_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0244_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0690_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0303_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0228_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0477_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0224_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0199_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0343_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0399_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0707_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0760_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0774_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0270_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0144_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0451_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0180_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0025_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0363_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0516_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0647_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0581_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0679_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0635_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0201_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0133_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0351_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0325_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0183_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0287_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0683_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0316_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0275_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0424_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0461_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0576_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0390_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0052_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0086_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0492_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0216_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0772_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0439_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0249_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0493_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0593_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0442_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0218_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0484_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0346_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0157_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0352_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0441_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0486_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0537_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0485_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0164_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0022_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0458_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0497_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0170_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0154_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0751_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0048_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0428_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0418_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0112_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0757_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0421_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0471_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0510_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0466_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0641_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0601_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0740_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0594_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0276_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0383_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0232_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0717_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0644_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0518_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0743_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0673_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0044_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0667_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0308_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0675_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0572_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0579_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0723_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0381_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0759_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0504_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0708_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0049_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0642_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0074_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0039_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0401_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0409_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0014_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0098_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0146_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0616_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0101_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0446_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0565_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0295_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0730_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0498_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0638_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0301_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0139_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0192_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0001_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0268_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0527_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0359_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0315_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0251_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0546_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0262_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0659_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0567_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0190_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0078_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0175_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0054_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0008_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0452_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0187_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0011_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0138_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0087_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0206_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0611_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0509_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0205_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0620_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0677_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0132_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0296_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0495_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0444_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0598_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0691_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0761_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0417_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0317_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0578_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0374_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0055_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0481_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0307_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0736_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0136_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0550_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0084_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0511_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0380_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0356_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0310_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0110_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0668_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0306_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0115_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0324_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0202_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0294_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0704_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0629_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0608_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0627_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0725_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0472_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0230_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0407_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0556_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0505_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0040_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0606_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0096_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0281_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0179_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0557_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0288_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0769_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0370_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0017_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0694_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0385_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0130_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0562_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0506_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0036_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0217_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0289_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0712_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0724_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0564_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0105_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0120_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0141_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0431_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0142_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0570_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0512_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0227_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0411_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0389_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0735_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0585_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0122_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0042_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0309_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0765_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0636_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0539_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0467_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0586_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0750_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0200_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0671_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0530_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0012_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0082_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0160_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0438_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0195_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0185_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0215_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0173_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0710_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0348_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0590_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0073_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0030_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0079_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0072_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0019_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0239_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0410_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0453_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0543_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0007_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0733_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0125_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0569_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0331_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0062_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0043_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0433_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0235_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0448_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0696_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0559_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0392_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0664_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0483_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0591_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0272_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0271_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0340_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0159_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0153_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0541_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0028_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0067_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0222_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0165_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0117_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0669_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0103_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0258_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0097_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0419_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0459_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0609_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0104_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0430_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0582_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0457_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0529_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0029_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0426_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0279_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0596_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0631_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0517_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0507_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0252_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0626_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0056_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0335_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0542_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0672_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0284_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0257_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0654_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0678_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0018_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0528_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0422_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0068_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0245_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0617_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0255_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0176_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0732_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0221_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0371_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0137_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0177_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0382_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0291_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0597_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0321_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0292_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0273_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0716_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0094_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0026_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0021_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0242_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0226_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0100_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0519_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0552_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0057_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0776_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0109_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0219_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0706_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0508_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0722_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0456_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0633_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0618_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0354_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0514_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0475_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0034_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0300_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0454_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0693_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0658_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0304_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0090_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0548_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0063_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0465_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0414_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0169_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0762_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0701_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0135_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0143_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0167_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0523_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0540_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0334_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0297_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0738_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0500_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0554_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0695_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0747_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0108_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0113_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0686_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0302_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0362_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0355_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0645_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0102_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0681_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0536_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0156_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0163_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0277_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0573_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0293_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0741_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0377_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0423_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0592_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0657_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0584_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0628_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0240_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0361_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0680_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0388_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0435_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0345_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0637_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0420_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0378_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0398_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0234_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0266_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0210_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0711_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0162_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0264_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0070_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0223_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0470_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0002_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0069_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0625_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0285_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0065_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0147_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0085_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0394_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0330_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0621_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0248_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0274_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0526_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0685_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0589_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0128_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0405_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0126_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0075_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0494_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0283_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0558_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0402_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0261_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0107_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0522_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0604_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0231_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0663_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0726_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0152_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0353_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0243_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0364_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0213_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0412_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0622_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0650_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0354_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0200_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0391_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0559_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0473_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0528_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0073_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0170_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0180_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0582_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0067_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0036_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0023_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0099_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0342_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0353_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0044_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0575_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0107_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0566_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0101_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0542_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0217_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0437_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0583_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0172_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0060_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0072_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0393_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0414_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0355_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0082_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0281_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0051_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0133_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0471_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0346_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0034_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0300_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0282_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0309_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0125_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0434_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0460_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0186_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0504_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0487_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0132_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0225_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0401_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0477_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0607_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0562_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0497_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0174_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0569_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0591_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0110_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0455_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0166_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0453_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0581_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0341_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0544_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0606_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0291_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0086_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0578_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0083_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0357_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0188_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0411_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0218_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0003_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0001_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0543_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0307_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0169_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0149_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0561_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0310_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0222_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0547_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0183_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0214_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0111_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0127_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0054_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0007_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0572_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0103_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0334_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0114_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0513_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0237_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0397_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0306_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0120_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0129_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0262_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0459_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0085_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0206_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0271_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0610_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0031_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0390_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0043_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0012_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0409_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0574_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0596_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0588_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0532_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0236_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0501_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0269_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0540_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0175_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0290_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0233_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0440_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0071_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0037_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0420_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0404_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0141_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0592_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0164_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0162_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0388_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0159_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0372_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0476_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0323_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0008_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0151_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0533_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0344_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0481_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0204_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0179_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0496_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0469_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0055_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0367_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0277_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0603_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0512_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0340_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0143_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0140_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0285_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0124_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0531_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0375_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0013_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0522_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0066_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0332_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0228_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0445_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0430_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0018_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0392_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0505_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0485_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0130_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0026_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0489_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0006_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0157_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0467_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0454_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0600_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0163_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0248_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0339_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0534_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0038_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0597_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0303_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0425_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0000_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0352_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0226_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0461_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0545_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0102_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0587_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0048_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0336_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0502_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0427_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0090_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0369_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0216_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0292_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0243_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0326_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0602_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0611_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0499_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0032_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0599_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0097_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0182_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0378_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0509_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0139_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0456_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0322_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0221_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0153_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0076_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0057_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0284_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0075_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0422_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0288_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0077_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0305_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0273_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0242_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0050_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0525_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0232_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0173_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0294_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0016_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0365_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0604_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0138_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0178_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0377_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0042_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0065_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0112_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0142_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0184_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0495_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0276_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0301_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0210_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0494_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0447_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0333_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0424_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0224_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0105_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0081_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0579_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0536_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0168_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0293_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0021_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0230_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0260_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0450_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0465_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0394_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0319_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0028_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0608_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0538_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0155_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0517_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0240_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0515_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0158_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0209_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0321_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0296_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0576_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0080_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0091_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0259_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0121_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0580_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0595_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0062_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0436_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0337_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0059_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0115_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0412_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0462_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0106_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0423_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0366_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0215_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0263_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0295_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0443_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0557_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0010_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0136_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0376_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0235_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0135_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0144_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0548_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0537_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0128_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0287_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0194_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0069_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0324_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0364_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0187_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0096_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0558_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0063_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0551_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0286_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0449_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0255_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0358_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0383_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0283_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0470_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0403_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0349_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0524_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0426_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0486_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0519_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0380_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0387_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0122_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0518_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0554_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0027_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0418_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0104_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0039_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0268_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0201_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0094_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0347_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0416_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0514_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0231_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0330_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0523_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0570_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0421_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0009_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0478_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0417_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0482_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0213_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0568_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0312_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0550_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0219_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0093_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0530_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0035_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0356_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0406_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0498_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0297_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0266_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0428_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0074_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0244_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0431_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0410_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0134_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0246_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0408_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0563_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0360_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0468_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0402_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0370_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0148_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0092_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0089_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0251_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0061_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0609_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0539_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0475_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0261_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0304_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0203_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0084_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0395_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0131_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0197_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0087_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0327_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0279_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0196_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0555_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0041_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0317_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0193_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0441_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0202_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0511_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0220_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0474_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0577_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0014_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0171_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0361_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0264_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0024_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0432_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0451_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0552_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0584_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0239_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0458_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0407_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0045_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0541_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0371_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0480_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0590_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0015_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0108_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0320_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0145_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0483_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0521_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0419_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0150_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0526_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0589_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0119_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0315_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0546_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0510_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0373_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0413_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0249_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0484_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0493_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0491_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0385_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0005_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0374_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0185_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0345_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0571_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0167_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0234_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0318_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0520_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0256_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0116_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0088_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0556_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0302_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0238_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0205_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0019_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0191_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0199_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0078_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0594_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0195_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0030_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0439_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0448_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0350_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0267_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0275_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0348_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0560_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0181_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0329_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0516_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0363_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0258_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0359_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0299_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0457_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0379_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0049_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0368_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0265_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0046_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0311_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0177_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0058_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0040_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0549_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0605_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0160_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0472_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0020_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0553_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0211_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0052_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0466_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0382_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0351_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0433_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0270_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0593_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0529_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0095_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0308_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0152_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0064_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0189_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0527_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0070_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0400_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0257_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0229_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0154_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0362_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0573_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0161_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0252_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0386_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0280_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0585_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0464_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0338_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0278_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0506_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0033_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0137_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0444_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0247_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0109_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0004_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0011_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0508_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0126_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0017_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0254_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0567_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0100_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0398_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0117_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0147_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0176_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0156_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0490_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0022_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0190_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0047_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0207_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0446_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0227_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0435_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0289_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0146_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0598_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0503_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0165_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0208_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0025_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0442_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0325_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0429_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0056_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0212_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0002_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0245_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0068_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0381_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0452_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0500_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0396_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0328_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0384_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0389_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0586_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0488_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0298_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0463_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0118_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0479_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0274_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0272_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0492_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0113_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0415_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0405_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0198_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0313_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0331_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0314_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0053_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0507_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0438_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0029_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0399_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0564_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0601_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0241_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0343_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0223_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0316_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0123_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0535_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0250_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0335_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0253_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0079_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0565_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0192_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0098_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0247_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1166_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1192_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0818_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0166_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0529_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1356_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0858_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0823_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1339_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0627_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1370_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0341_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0185_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1259_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0981_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0515_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0545_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1167_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0077_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0665_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1195_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0085_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0566_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0230_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0311_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0055_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0810_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1270_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0966_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0517_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0843_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0348_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0797_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0736_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0943_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1054_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1105_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0556_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0849_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0492_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1121_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0817_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0874_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0579_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1250_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0146_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0589_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0169_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1084_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1041_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0526_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0551_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0193_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1402_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1106_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0782_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0659_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0588_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0990_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0833_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0845_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1098_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0402_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0878_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0930_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0046_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0440_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1293_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0393_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0049_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0305_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0868_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1126_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0531_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1087_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1442_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0997_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0366_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0165_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1078_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0957_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1002_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0269_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0460_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1397_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0250_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0951_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1246_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0876_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0302_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0564_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0584_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0622_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0694_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0335_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1189_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0215_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1390_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0204_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1038_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0713_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0567_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1130_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0221_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0538_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1232_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1265_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0628_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0090_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0968_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0248_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0885_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0977_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0749_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0527_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0985_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0934_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0993_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0088_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0674_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0171_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1160_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0640_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1419_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0488_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0704_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0887_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1055_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1258_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0924_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0390_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0612_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1276_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0744_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0399_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0053_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0025_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0371_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0161_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1444_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0051_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0367_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0036_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1398_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0434_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0965_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1226_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0093_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0702_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0949_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1343_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1480_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0503_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0979_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1039_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1034_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0725_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1185_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1288_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0742_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0242_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1071_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0975_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1211_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0435_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0841_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1303_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1380_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0558_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0522_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0181_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0098_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0318_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1101_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1183_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0054_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0962_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0072_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1283_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1269_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0379_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0109_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1257_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0683_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0932_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0724_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0259_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0752_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0748_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1456_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0038_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1311_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0224_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1181_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0292_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0006_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1077_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1376_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0082_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1107_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1305_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1263_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1484_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0470_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0232_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0312_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1478_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0176_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0825_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0587_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0726_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1069_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0478_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1470_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0751_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0034_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0557_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0087_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0554_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0680_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0095_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1193_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1290_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0956_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1433_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1409_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0546_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1137_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1042_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1361_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0983_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1099_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0675_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0000_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0540_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0593_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1182_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0984_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0463_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0406_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0963_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1112_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0572_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0894_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0901_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0539_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1462_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0504_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1196_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0916_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0840_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1059_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0895_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1439_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0521_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0227_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0157_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0134_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0091_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0080_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1405_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0856_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0355_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0904_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1256_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0886_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0703_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0870_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1307_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1463_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1278_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1176_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0999_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0352_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1216_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0506_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0942_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0164_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0071_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0362_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0285_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0819_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1447_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0794_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1111_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0452_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1008_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1064_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0786_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0275_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1004_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1020_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1198_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0258_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0179_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0473_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0706_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0829_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1348_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0489_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0721_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1418_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0657_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0182_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1268_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0677_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0211_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1199_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1161_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1159_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1060_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0203_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0634_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0214_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1292_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0168_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1342_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0105_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1346_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0145_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0174_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1393_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1295_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0888_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0673_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1412_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1337_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1089_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1031_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0542_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0135_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0604_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0905_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1028_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0760_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0052_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0361_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0408_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0514_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1144_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0086_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0016_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0363_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0026_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0013_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0048_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0343_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0991_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1151_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0325_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0209_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1325_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0042_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0162_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0902_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0625_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0482_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0502_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0225_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1190_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0498_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0198_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1146_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1197_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0889_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0873_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1013_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1212_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0107_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0005_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0528_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0101_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0792_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0638_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1253_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1449_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0761_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1330_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0268_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0372_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0757_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0031_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0815_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0074_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0771_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0936_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1355_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0764_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0516_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1460_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0394_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0178_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1005_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1103_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0899_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1204_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1061_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0327_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0233_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0057_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0922_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0996_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1457_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0513_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1323_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0426_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1428_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0879_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0297_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1395_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0333_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0202_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1171_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0746_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0111_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1459_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0806_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1332_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0003_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1082_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0745_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1187_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0803_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0326_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1131_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0678_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1458_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1465_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0662_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0715_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0043_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1029_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0108_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0450_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0465_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0125_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1052_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0988_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0487_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0872_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0096_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0865_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0643_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0507_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0032_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1228_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1076_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1229_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0219_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1341_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0384_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0583_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0236_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1267_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0682_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1324_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0672_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1237_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0256_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0689_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0801_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1392_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0630_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0041_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1027_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0173_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1476_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1455_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0172_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0228_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0812_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0980_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0781_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1280_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0141_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1046_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0656_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0137_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0234_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0909_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0410_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0836_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1313_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1326_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1113_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1040_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1345_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1492_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0831_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0293_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1063_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0309_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0353_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0597_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0011_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0189_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1174_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0474_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1079_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0066_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0697_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0734_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0756_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1488_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0238_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0235_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0274_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0417_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0316_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1451_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0864_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0127_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0900_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0244_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0291_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1205_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1344_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1351_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0441_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0213_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1143_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1320_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0437_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0927_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0324_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1125_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1421_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0735_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0573_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0002_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1416_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0776_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1430_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1322_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0180_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0272_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0896_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0543_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0570_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0859_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0255_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0263_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1299_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0835_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0039_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0611_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0369_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0732_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1485_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0045_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1138_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0067_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0642_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0961_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0249_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1221_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0496_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1312_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0599_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0497_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1413_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1382_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0661_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1251_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0560_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0350_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0018_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0354_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0007_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0594_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1003_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1375_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0298_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0199_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0555_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0419_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0400_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0608_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0789_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0618_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0212_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0493_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0633_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0920_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0621_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0129_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1033_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0982_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0300_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1000_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1284_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0241_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0163_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0971_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0455_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1124_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1464_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0332_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1225_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0160_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0132_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0953_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1396_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0319_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0014_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0867_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0765_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0731_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0121_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1340_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1414_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1149_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0897_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1140_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1406_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1491_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0811_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0047_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0907_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0184_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1202_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0595_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1333_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0149_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1234_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1289_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0331_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1415_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0769_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0935_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0216_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0940_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0762_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0445_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0378_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0280_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0376_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1047_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1045_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1056_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0356_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0210_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1386_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1110_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0476_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1163_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0598_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0511_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0279_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1254_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0115_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0365_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1298_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0839_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1227_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1282_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0030_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0254_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0658_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0978_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0851_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0130_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0357_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0152_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0952_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0834_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1436_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1302_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1210_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1445_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1328_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0188_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1152_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0340_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0534_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0986_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0892_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0062_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1173_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0009_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0537_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0058_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0286_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1350_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0837_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0068_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1291_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1108_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0158_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0425_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1007_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0717_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1314_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1097_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0131_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1425_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0050_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1432_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0257_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1487_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0793_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0655_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0339_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1109_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0151_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0830_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0912_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0700_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0959_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1156_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0798_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1371_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0023_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1483_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1357_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0122_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0462_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1021_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0562_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0505_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0787_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1420_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1399_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1024_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1296_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0454_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0844_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0133_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1377_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1037_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0705_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0623_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0950_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0547_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0240_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1011_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0893_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0117_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1178_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1120_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0346_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1354_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0880_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1019_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0477_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1231_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0660_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0471_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1168_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1214_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1378_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1142_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1102_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1015_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1438_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1372_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0945_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1248_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1119_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0064_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0973_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0509_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1135_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0019_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0663_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1025_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0261_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1431_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1403_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0004_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1374_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1164_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0740_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1217_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0387_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0911_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1203_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0995_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0535_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1319_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0795_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1184_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0915_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0576_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0329_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0929_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0142_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0413_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0520_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1080_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1334_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1207_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0869_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1012_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0475_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0167_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0494_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0264_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1358_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0641_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1318_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1194_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0059_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0578_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1095_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0692_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1422_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0195_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0431_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0928_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0388_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0690_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0467_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0415_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0119_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0187_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0424_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0716_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0854_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0252_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0635_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0755_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0294_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0719_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0785_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0914_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1364_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0021_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1242_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0220_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0139_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1220_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1383_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0102_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1424_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0921_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1215_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0722_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0908_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0805_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0414_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0190_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0089_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1150_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1360_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0453_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0827_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1273_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0670_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0891_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1329_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1261_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1147_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1262_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0359_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0328_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1240_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1391_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0156_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0989_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1219_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0923_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0654_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0698_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0065_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1287_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0383_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0532_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0260_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1349_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1086_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0788_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0113_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1363_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0687_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1083_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0918_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0585_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0392_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0253_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1043_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0449_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0568_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0421_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0747_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1175_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1373_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1304_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1474_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1482_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1385_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0317_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0603_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0519_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1090_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0020_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1489_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0871_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0510_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0104_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1018_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0245_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1440_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0758_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1247_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1010_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1179_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0552_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0276_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0679_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0571_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1017_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0301_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1477_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1450_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0712_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0676_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0577_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0684_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0881_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0644_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0076_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1446_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0154_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0284_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0824_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0231_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0701_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0790_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1224_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1454_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0405_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0177_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0267_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0944_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1368_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0389_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0610_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0128_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1085_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1057_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0459_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0027_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0759_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0458_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0472_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0816_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0022_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0832_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0097_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0407_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0820_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0777_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0941_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0821_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0796_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1022_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0967_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0548_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1117_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0197_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0175_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0218_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0397_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0850_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0710_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0592_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0447_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0607_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1249_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0436_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0784_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0728_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0646_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1104_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0443_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1154_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0194_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1206_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1233_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0170_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0299_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0605_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0033_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0037_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0631_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0877_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0565_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0626_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0723_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0650_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0590_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0938_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1132_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1479_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0313_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0651_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0926_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1274_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0828_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0602_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0637_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0423_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0681_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0411_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0223_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1186_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0591_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1441_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0533_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1366_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1466_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1423_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1075_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0939_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1327_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0647_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0420_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0330_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1016_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0600_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0606_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1096_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0246_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0530_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0344_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1471_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0774_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0733_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0283_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0561_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0466_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1032_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1188_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1285_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0029_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1081_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0063_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0739_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0954_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0448_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1331_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1218_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0688_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1169_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0799_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0866_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0853_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0955_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0910_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0490_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0861_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0948_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1336_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0349_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0808_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1277_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0601_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0044_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0315_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1230_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1255_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0271_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1051_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0481_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0970_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0412_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0040_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1068_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1091_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1394_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0862_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1461_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1453_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0070_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0852_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0012_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0484_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0140_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0919_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0288_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1490_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0863_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1129_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0737_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0791_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0884_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0708_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0667_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1452_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1048_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0609_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1049_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0336_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0800_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0196_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0931_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1072_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0207_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0144_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0395_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0648_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1389_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0582_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0433_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0553_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1088_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0391_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0304_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0110_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1310_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0398_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0446_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0574_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0772_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0686_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1335_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1417_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1388_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0855_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0457_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1093_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0903_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0669_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0860_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0754_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0289_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0946_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1275_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0024_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0310_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1153_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0958_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0385_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1115_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0550_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0653_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1243_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1030_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0775_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0382_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1369_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1429_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1400_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1213_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0709_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0842_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0691_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1066_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0766_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1139_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0499_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0699_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0200_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0239_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0439_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1155_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0525_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0523_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0079_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0416_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1481_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0727_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0307_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0138_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1427_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1475_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0429_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0206_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0375_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1448_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1352_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0270_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0103_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0778_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1408_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1223_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1472_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0813_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1384_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0629_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0337_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0848_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0124_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0753_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1410_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0243_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0714_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1001_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1317_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0112_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1260_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0251_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1272_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0624_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1158_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0444_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0925_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1050_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1134_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0695_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0508_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0615_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0265_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1316_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1148_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1315_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0226_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0320_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0237_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1353_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0287_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0262_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1067_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0430_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0001_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0360_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0017_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1468_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0501_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1141_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0998_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1145_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1473_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0396_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0913_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0380_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0544_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1411_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0106_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0442_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0418_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0491_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0368_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0296_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0483_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0536_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1306_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1347_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1467_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0969_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1036_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0010_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1044_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0619_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1437_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0282_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1162_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0750_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0126_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0192_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1362_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1435_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1359_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0685_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1469_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1338_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0273_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0060_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0229_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0358_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0743_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1294_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1136_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0155_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1122_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0707_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1301_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0814_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0083_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0303_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1208_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1116_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0512_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0205_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0847_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0464_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1365_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1266_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0278_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0075_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0201_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1308_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1235_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0056_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0061_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0917_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0401_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0617_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1236_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0974_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0693_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1245_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0524_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0541_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0351_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0099_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0614_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0403_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0321_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0370_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1379_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0084_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0666_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0783_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0028_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0208_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1244_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1300_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0883_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1239_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0308_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0035_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0486_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1271_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0078_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1127_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0377_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0652_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1443_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0711_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0838_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0322_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1222_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0290_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0857_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1062_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0186_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1114_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1073_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0802_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1123_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0281_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0770_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0620_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0581_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0456_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0094_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0092_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0645_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1426_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0334_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0809_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0422_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1170_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0153_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0147_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0100_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0636_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1026_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0639_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0081_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0720_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1094_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0381_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0342_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1387_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0347_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0191_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0409_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1118_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0159_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0485_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1009_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1165_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0116_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0586_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1381_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0668_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0992_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0073_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0664_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1264_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0183_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0822_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0338_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1200_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0479_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0972_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0826_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0780_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0432_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0373_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0767_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0469_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0779_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0217_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1177_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0364_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1014_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0461_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1209_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0976_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0804_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0549_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1180_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0120_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1401_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0559_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0933_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1157_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0114_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0480_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0964_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1023_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0580_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1053_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0266_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0438_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0875_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1286_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0123_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0729_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0937_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0295_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0277_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0069_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1006_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0994_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1100_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0960_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1407_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0741_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0374_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1065_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0730_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0306_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1074_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0222_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0008_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0763_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0563_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1172_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0632_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1092_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0906_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0613_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0890_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1070_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0136_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0947_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1238_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0882_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0495_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1128_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1279_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0575_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1201_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0404_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0773_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0898_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1321_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0846_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0518_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1367_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0345_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0015_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0451_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1297_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1241_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0987_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0150_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0696_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0807_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1035_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1434_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0649_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1404_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1281_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0768_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0738_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1309_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1133_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1058_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0427_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0468_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0500_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0148_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0386_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0616_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0428_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0118_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0569_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0323_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0718_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1252_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1191_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0143_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0314_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0671_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1486_text_document +0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0596_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt b/ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt new file mode 100644 index 00000000000..18109f946d5 --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt @@ -0,0 +1,3 @@ +0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0001_text_document +0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0000_text_document +0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0002_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_peS2o.txt b/ALCF/data-lists/polaris/data_file_list_peS2o.txt new file mode 100644 index 00000000000..3f2ddfb2992 --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_peS2o.txt @@ -0,0 +1,42 @@ +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0039_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0014_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0034_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0007_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0020_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0026_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0036_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0030_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0015_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0018_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0033_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0027_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0023_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0024_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0009_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0025_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0010_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0032_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0029_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0021_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0040_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0000_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0013_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0005_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0022_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0011_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0038_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0003_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0019_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0031_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0012_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0041_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0004_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0001_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0037_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0006_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0016_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0002_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0017_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0028_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0035_text_document +0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0008_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_stack-code.txt b/ALCF/data-lists/polaris/data_file_list_stack-code.txt new file mode 100644 index 00000000000..f5049cd9e4f --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_stack-code.txt @@ -0,0 +1,4435 @@ +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rhtml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl6/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gap/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/alloy/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/squirrel/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rouge/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/textile/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/labview/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mask/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/openscad/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/openscad/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/maxscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/modelica/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/modelica/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unrealscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lookml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opal/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/module-management-system/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opa/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ren'py/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ren'py/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/maple/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/numpy/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/realbasic/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pure-data/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pure-data/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ceylon/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0102_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0104_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0105_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0106_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0103_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0107_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0108_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0101_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/augeas/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rdoc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logtalk/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c2hs-haskell/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pov-ray-sdl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ioke/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/verilog/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clips/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/chuck/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pod/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pod/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xojo/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/component-pascal/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/component-pascal/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lex/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lex/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/grace/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sas/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sas/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/netlogo/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autoit/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autoit/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zephir/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/robotframework/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/propeller-spin/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xquery/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/txl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nu/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ampl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tea/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/csound/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/brightscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slim/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slim/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/red/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/thrift/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pony/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-coffeescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ats/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot-internal-representation/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lolcode/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/digital-command-language/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/abap/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/abap/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lsl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hlsl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hlsl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scaml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ags-script/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xs/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mtml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rmarkdown/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kit/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mako/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/igor-pro/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sourcepawn/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sourcepawn/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/apl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nginx/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/piglatin/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/awk/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vcl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/applescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webidl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glyph/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/papyrus/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/boo/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hy/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/d/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/aspectj/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/isabelle/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/isabelle/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sqf/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sqf/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/volt/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/monkey/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lfe/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clarion/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/oxygene/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/metal/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nsis/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/muf/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dylan/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xbase/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autohotkey/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autohotkey/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clean/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ston/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/creole/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ecl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eiffel/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eiffel/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/darcs-patch/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dm/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot-assembly/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cobol/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/io/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/saltstack/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ox/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/matlab/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/renderscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purescript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dogescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/omgrofl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/supercollider/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/flux/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-haskell/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcsh/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+eex/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/golo/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pawn/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/oz/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/idl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-agda/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/click/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/krl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/inform-7/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stan/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/livescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion-cfc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion-cfc/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gosu/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lilypond/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/moonscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/idris/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fish/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/myghty/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/latte/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turing/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bison/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emberscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xpages/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/linker-script/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/linker-script/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cycript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scilab/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-eclass/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy-server-pages/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy-server-pages/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dns-zone/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/netlinx/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/irc-log/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mirah/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purebasic/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cartocss/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/j/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jflex/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qmake/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/api-blueprint/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pan/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/brainfuck/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ecere-projects/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fantom/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/grammatical-framework/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/befunge/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objdump/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/urweb/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/blitzmax/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/liquid/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/liquid/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/genshi/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ninja/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ninja/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gams/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lasso/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/desktop/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/agda/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/agda/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m4/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m4/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coq/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-ebuild/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-ebuild/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/factor/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/uno/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/apacheconf/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pogoscript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jasmin/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bluespec/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nit/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ec/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/raml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rebol/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-j/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bro/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sparql/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/chapel/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pike/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/harbour/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unified-parallel-c/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xtend/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ooc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shen/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/self/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/x10/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cirru/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/redcode/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mupad/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/inno-setup/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/wisp/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/http/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/http/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/forth/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ragel-in-ruby-host/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nesc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nesc/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slash/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zimpl/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sage/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/blitzbasic/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/octave/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fancy/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/antlr/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bitbake/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bitbake/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vala/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cap'n-proto/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/prolog/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shellsession/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python-traceback/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0087_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0025_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0019_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0015_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0009_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0064_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0050_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0058_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0093_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0059_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0051_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0021_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0071_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0008_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0044_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0072_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0097_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0068_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0080_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0057_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0047_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0052_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0067_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0014_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0023_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0034_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0032_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0007_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0060_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0086_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0094_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0091_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0017_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0090_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0066_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0046_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0096_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0100_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0020_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0076_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0092_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0048_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0049_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0075_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0026_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0012_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0095_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0074_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0083_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0054_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0070_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0082_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0055_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0024_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0037_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0088_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0078_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0065_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0062_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0089_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0099_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0061_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0041_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0079_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0035_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0002_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0028_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0045_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0040_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0004_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0098_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0038_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0056_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0063_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0011_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0003_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0077_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0042_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0013_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0029_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0039_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0022_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0084_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0085_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0069_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0081_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0006_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0073_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0027_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0010_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0043_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0005_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0016_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0033_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0018_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0036_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0030_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0001_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0053_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0031_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xproc/v3-0000_text_document +0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/racket/v3-0000_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt b/ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt new file mode 100644 index 00000000000..134c1473b11 --- /dev/null +++ b/ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt @@ -0,0 +1,2 @@ +0.0045 /eagle/datasets//dolma/data_Llama2Tokenizer/wiki-en-simple//en_simple_wiki-0000_text_document +0.0045 /eagle/datasets//dolma/data_Llama2Tokenizer/wiki-en-simple//en_simple_wiki-0001_text_document From 27f66fd87053c0f3764e4dedc9afe8f3fd65c963 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Apr 2024 11:35:52 -0500 Subject: [PATCH 070/590] Add support for DeepSpeed `FusedLamb` optimizer --- megatron/arguments.py | 7 +- megatron/optimizer/__init__.py | 146 +++++++++++++++++++++++++++++++-- 2 files changed, 144 insertions(+), 9 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index b982337f517..70dfa88ef6e 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -923,10 +923,15 @@ def _add_training_args(parser): 'adam', 'adamw', 'sgd', + 'ds.fusedlamb', 'apex.adam', 'apex.sgd', 'adamwschedulefree', - 'sgdschedulefree' + 'sgdschedulefree', + 'galoreadamw', + 'adam8bit', + 'galoreadamw8bit', + 'galoreadamw8bitperlayer' ], help='Optimizer function' ) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 6dfd0ea40f7..80adeb85210 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -3,6 +3,7 @@ from deepspeed.accelerator import get_accelerator import torch +from typing import Callable, Any from megatron import get_args from .distrib_optimizer import DistributedOptimizer @@ -10,19 +11,60 @@ from .optimizer import Float16OptimizerWithFloat16Params, FP32Optimizer -def get_param_groups(modules, - no_weight_decay_cond, - scale_lr_cond, - lr_mult): - """creates param groups based on weight decay condition (regularized vs non regularized) - and learning rate scale condition (args.lr vs lr_mult * args.lr) - scale_lr_cond is used during finetuning where head of the network requires a scaled - version of the base learning rate. +import ezpz as ez +RANK = ez.get_rank() + + +def get_param_groups( + modules: torch.nn.Module | iter[torch.nn.Module], + no_weight_decay_cond: Callable[[str, torch.Tensor], bool], + scale_lr_cond: Callable[[str, torch.Tensor], bool], + lr_mult: Any, + use_galore: bool = False, +): + """ + Creates param groups (regularized vs non) based on: + + - weight decay condition. + - learning rate scale condition (args.lr vs lr_mult * args.lr) + - scale_lr_cond is used during finetuning, where head of the network + requires a scaled version of the base learning rate. + # if 'galore' in args.optimizer.lower(): + # # make parameters with "rank" to a single group, if param_name has "mlp" or "attn" + # galore_params = [] + # target_modules_list = ["attn", "mlp"] + # # for module_name, module in param_groups: + # for group_id, group in enumerate(param_groups): + # for param, p in enumerate(group['params']): + # if not isinstance(module, torch.nn.Linear): + # continue + # if not any(target_key in module_name for target_key in target_modules_list): + # continue + # print('enable GaLore for weights in module: ', module_name) + # galore_params.append(module.weight) + # id_galore_params = [id(p) for p in galore_params] + # # make parameters without "rank" to another group + # regular_params = [p for p in param_groups if id(p) not in id_galore_params] + # # then call galore_adamw + # param_groups = [ + # { + # 'params': regular_params + # }, + # { + # 'params': galore_params, + # 'rank': RANK, + # 'update_proj_gap': args.update_proj_gap, + # 'scale': args.galore_scale, + # 'proj_type': args.proj_type + # } + # ] """ wd_no_scale_lr = [] wd_scale_lr = [] no_wd_no_scale_lr = [] no_wd_scale_lr = [] + galore_params = [] + target_modules_list = ["attn", "mlp"] for module in modules: for name, param in module.named_parameters(): if not param.requires_grad: @@ -83,6 +125,7 @@ def get_megatron_optimizer( param_groups = split_params_into_different_moe_groups_for_optimizer( param_groups ) + if args.cpu_optimizer: assert args.optimizer == 'adam', 'CPU offloading is for Adam' if args.cpu_torch_adam: @@ -97,6 +140,93 @@ def get_megatron_optimizer( betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps, ) + + elif args.optimizer.lower() == "galore_adamw": + from galore_torch import GaLoreAdamW, GaLoreAdamW8bit + # redefine way to call galore_adamw + optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) + elif args.optimizer.lower() == "galore_adamw": + # redefine way to call galore_adamw + optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) + # implement adafactor + elif args.optimizer.lower() == "adafactor": + import transformers + args.beta1 = None if args.beta1 == 0.0 else args.beta1 + optimizer = transformers.optimization.Adafactor( + param_groups, + lr=args.lr, + eps=(1e-30, 1e-3), + clip_threshold=1.0, + decay_rate=-0.8, + beta1=args.beta1, + weight_decay=args.weight_decay, + relative_step=False, + scale_parameter=False, + warmup_init=False, + ) + # low-rank adafactor + elif args.optimizer.lower() == "galore_adafactor": + args.beta1 = None if args.beta1 == 0.0 else args.beta1 + optimizer = GaLoreAdafactor( + param_groups, + lr=args.lr, + eps=(1e-30, 1e-3), + clip_threshold=1.0, + decay_rate=-0.8, + beta1=args.beta1, + weight_decay=args.weight_decay, + relative_step=False, + scale_parameter=False, + warmup_init=False, + ) + # 8-bit Adam + elif args.optimizer.lower() == "adam8bit": + import bitsandbytes as bnb + optimizer = bnb.optim.Adam8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + elif args.optimizer.lower() == "galore_adamw8bit": + optimizer = GaLoreAdamW8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + elif args.optimizer.lower() == 'galore_adamw8bit_per_layer': + # TODO: seems scheduler call twice in one update step, need to check, for now double the num_training_steps, warmup_steps and update_proj_gap + optimizer_dict = {} + for p in model.parameters(): + if p.requires_grad: + if id(p) in id_galore_params: + optimizer_dict[p] = GaLoreAdamW8bit([{'params': [p], 'rank': args.rank, 'update_proj_gap': args.update_proj_gap * 2, 'scale': args.galore_scale, 'proj_type': args.proj_type}], lr=args.lr, weight_decay=args.weight_decay) + else: + optimizer_dict[p] = bnb.optim.Adam8bit([p], lr=args.lr, weight_decay=args.weight_decay) + # get scheduler dict + scheduler_dict = {} + from galore_torch.peft_pretraining import training_utils + for p in model.parameters(): + if p.requires_grad: + scheduler_dict[p] = training_utils.get_scheculer( + optimizer=optimizer_dict[p], + scheduler_type=args.scheduler, + num_training_steps=args.num_training_steps * 2, + warmup_steps=args.warmup_steps * 2, + min_lr_ratio=args.min_lr_ratio, + ) + + def optimizer_hook(p): + if p.grad is None: + return + optimizer_dict[p].step() + optimizer_dict[p].zero_grad() + scheduler_dict[p].step() + # Register the hook onto every parameter + for p in model.parameters(): + if p.requires_grad: + p.register_post_accumulate_grad_hook(optimizer_hook) + layer_wise_flag = True + elif str(args.optimizer).lower() == 'ds.fusedlamb': + from deepspeed.ops.lamb import FusedLamb + optimizer = FusedLamb( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, + ) elif str(args.optimizer).lower() == 'adamwschedulefree': import schedulefree optimizer = schedulefree.AdamWScheduleFree( From fc1b347ca8e6c246309e43fa1a8600ee81b4de68 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Apr 2024 12:20:58 -0500 Subject: [PATCH 071/590] Add `ALCF/data-lists/sunspot/*.txt` --- ...erg-books.txt => data_file_list_books.txt} | 0 ...tack-code.txt => data_file_list_stack.txt} | 0 ...-en-simple.txt => data_file_list_wiki.txt} | 0 .../sunspot/data_file_list_books.txt | 3 + ALCF/data-lists/sunspot/data_file_list_c4.txt | 86 + ALCF/data-lists/sunspot/data_file_list_cc.txt | 2880 +++++++++++++++++ .../sunspot/data_file_list_peS2o.txt | 26 + .../sunspot/data_file_list_reddit.txt | 78 + .../sunspot/data_file_list_stack.txt | 149 + .../sunspot/data_file_list_wiki.txt | 2 + 10 files changed, 3224 insertions(+) rename ALCF/data-lists/polaris/{data_file_list_gutenberg-books.txt => data_file_list_books.txt} (100%) rename ALCF/data-lists/polaris/{data_file_list_stack-code.txt => data_file_list_stack.txt} (100%) rename ALCF/data-lists/polaris/{data_file_list_wiki-en-simple.txt => data_file_list_wiki.txt} (100%) create mode 100644 ALCF/data-lists/sunspot/data_file_list_books.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_c4.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_cc.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_peS2o.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_reddit.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_stack.txt create mode 100644 ALCF/data-lists/sunspot/data_file_list_wiki.txt diff --git a/ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt b/ALCF/data-lists/polaris/data_file_list_books.txt similarity index 100% rename from ALCF/data-lists/polaris/data_file_list_gutenberg-books.txt rename to ALCF/data-lists/polaris/data_file_list_books.txt diff --git a/ALCF/data-lists/polaris/data_file_list_stack-code.txt b/ALCF/data-lists/polaris/data_file_list_stack.txt similarity index 100% rename from ALCF/data-lists/polaris/data_file_list_stack-code.txt rename to ALCF/data-lists/polaris/data_file_list_stack.txt diff --git a/ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt b/ALCF/data-lists/polaris/data_file_list_wiki.txt similarity index 100% rename from ALCF/data-lists/polaris/data_file_list_wiki-en-simple.txt rename to ALCF/data-lists/polaris/data_file_list_wiki.txt diff --git a/ALCF/data-lists/sunspot/data_file_list_books.txt b/ALCF/data-lists/sunspot/data_file_list_books.txt new file mode 100644 index 00000000000..9187565a5ed --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_books.txt @@ -0,0 +1,3 @@ +0.0031007020167215667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0000_text_document +0.003100207465277759 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0001_text_document +0.000999090518000674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0002_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_c4.txt b/ALCF/data-lists/sunspot/data_file_list_c4.txt new file mode 100644 index 00000000000..ca7df1839e9 --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_c4.txt @@ -0,0 +1,86 @@ +0.0011545953050729803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0000_text_document +0.0011570295715413383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0001_text_document +0.001156438391210766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0002_text_document +0.0011556820995190797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0003_text_document +0.001156780334924253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0004_text_document +0.0011563528368937514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0005_text_document +0.0011574632716369762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0006_text_document +0.0011577445131424494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0007_text_document +0.0011599182963630329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0008_text_document +0.0011550792360663698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0009_text_document +0.001154948574643344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0010_text_document +0.0011560157369398198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0011_text_document +0.0011551344387810997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0012_text_document +0.0011586914190552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0013_text_document +0.00115559584811127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0014_text_document +0.0011562917764239204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0015_text_document +0.0011582019252872318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0016_text_document +0.0011585605528399534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0017_text_document +0.0011567600261132287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0018_text_document +0.0011561323235067436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0019_text_document +0.0011568948157687324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0020_text_document +0.0011562184926986983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0021_text_document +0.001155171968076667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0022_text_document +0.001156245876059478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0023_text_document +0.0011591826911770261 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0024_text_document +0.0011564400126070828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0025_text_document +0.0011571005158517765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0026_text_document +0.0011560050453907214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0027_text_document +0.0011559074476966407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0028_text_document +0.0011567638698290205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0029_text_document +0.0011558972055942165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0030_text_document +0.001157532269673901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0031_text_document +0.0011559883017581377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0032_text_document +0.001155556362078353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0033_text_document +0.0011544735837522018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0034_text_document +0.0011547315955415466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0035_text_document +0.0011570980852521353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0036_text_document +0.0011562552591307868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0037_text_document +0.001156640315842092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0038_text_document +0.0011587257748187634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0039_text_document +0.0011563083526351268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0040_text_document +0.0011554464046007336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0041_text_document +0.001155442922136426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0042_text_document +0.0011557081619451221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0043_text_document +0.001156421357082161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0044_text_document +0.0011562730825316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0045_text_document +0.001157525507046117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0046_text_document +0.0011552936629887162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0047_text_document +0.0011578959437852875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0048_text_document +0.0011568910557636293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0049_text_document +0.0011578444955946039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0050_text_document +0.001157076096248001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0051_text_document +0.0011568459536403974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0052_text_document +0.0011555352450605598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0053_text_document +0.0011557650508322967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0054_text_document +0.0011567625802857914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0055_text_document +0.0011568533734967437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0056_text_document +0.0011562185375437102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0057_text_document +0.0011558740426473278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0058_text_document +0.0011549825990520978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0059_text_document +0.0011572314079774744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0060_text_document +0.0011576031815962752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0061_text_document +0.0011567937670018521 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0062_text_document +0.001154956951193276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0063_text_document +0.001157226898064118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0064_text_document +0.001156096958730414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0065_text_document +0.001155844223704128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0066_text_document +0.0011571187084765205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0067_text_document +0.0011573954893981501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0068_text_document +0.0011566700251641518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0069_text_document +0.0011550051959552815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0070_text_document +0.0011559629359246125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0071_text_document +0.001157971629210032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0072_text_document +0.0011561725903411443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0073_text_document +0.001157160385935682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0074_text_document +0.0011568864860569239 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0075_text_document +0.0011576433208715313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0076_text_document +0.0011571382379808948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0077_text_document +0.0011590178523739284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0078_text_document +0.001156347684201892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0079_text_document +0.0011552550374817486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0080_text_document +0.0011570794132840427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0081_text_document +0.0011570932061148482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0082_text_document +0.0011561938025300182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0083_text_document +0.0011560757016965283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0084_text_document +0.00019284851714729888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0085_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_cc.txt b/ALCF/data-lists/sunspot/data_file_list_cc.txt new file mode 100644 index 00000000000..4b9a797878b --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_cc.txt @@ -0,0 +1,2880 @@ +0.0002329030984435853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0000_text_document +0.00023018699207949078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0001_text_document +0.00024373839803694205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0002_text_document +0.00023608269234913788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0003_text_document +0.00024813091225197464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0004_text_document +0.00023520818074126314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0005_text_document +0.0002374607329273171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0006_text_document +0.00023738412849923294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0007_text_document +0.0002443634316582533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0008_text_document +0.00023847622533166118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0009_text_document +0.00023199871587697545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0010_text_document +0.0002385337709567312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0011_text_document +0.0002432839071745339 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0012_text_document +0.00023508523674007346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0013_text_document +0.00032603226617680567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0014_text_document +0.00023789141182395846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0015_text_document +0.0002461407443245122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0016_text_document +0.00023499257215518966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0017_text_document +0.00024846537508068473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0018_text_document +0.0002386611981191132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0019_text_document +0.0002476214516386151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0020_text_document +0.00023922963334203518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0021_text_document +0.0002566637890877035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0022_text_document +0.0002480836116312675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002500957846859012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0024_text_document +0.00023232303192858133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0025_text_document +0.0002402109920207785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0026_text_document +0.00032458741378655037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0027_text_document +0.00023711130623699136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0028_text_document +0.0002473092752915358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024517111812673547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0030_text_document +0.00024145261714879915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0031_text_document +0.0002441832095655324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0032_text_document +0.00024533720808111173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0033_text_document +0.00024615543201451354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0034_text_document +0.00029788578618284437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0035_text_document +0.00026821245945822444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0036_text_document +0.0002451138188102186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0037_text_document +0.00023812823651070536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0038_text_document +0.00023799603175215714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0039_text_document +0.00024128396884325748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0040_text_document +0.00024158008848876737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0041_text_document +0.00024722330373436316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0042_text_document +0.00023308404070500205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002554252556503107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0044_text_document +0.0003132025339147037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0045_text_document +0.00024278622445373792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0046_text_document +0.0003214585004572529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0047_text_document +0.0003329131703028111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0048_text_document +0.0002361664236831262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0049_text_document +0.0002643368247294079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0050_text_document +0.00024766538637149724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0051_text_document +0.0002627167479901225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0052_text_document +0.00025033496855447236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0053_text_document +0.00024160037266449382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0054_text_document +0.00022926708072112655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0055_text_document +0.00023577632399723273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0056_text_document +0.00024916378421745264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0057_text_document +0.00024065956580145883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0058_text_document +0.00032914757231594763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0059_text_document +0.000382735213415281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0060_text_document +0.00019876415914729903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0061_text_document +0.0002455041228482986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0062_text_document +0.0002360975192355561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0063_text_document +0.00035687225557611647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0064_text_document +0.00034010734287544296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0065_text_document +0.00024289772720050695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0066_text_document +0.0002298464162081398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0067_text_document +0.00032731880189343956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0068_text_document +0.00024593154050122983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0069_text_document +0.00024184757636917526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0070_text_document +0.0002619883069796127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0071_text_document +0.00023707630401459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0072_text_document +0.0003648802259322563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0073_text_document +0.00034821518419266554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0074_text_document +0.00025687739808269634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0075_text_document +0.00025210376457187776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0076_text_document +0.00025341417049958763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0077_text_document +0.00026096750660126574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002557323323244081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0079_text_document +0.0003306928457892949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0080_text_document +0.00034038835131844906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0081_text_document +0.00025944099107910257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0082_text_document +0.00011523229485833962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002577986281049885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0084_text_document +0.00011763411767853355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0085_text_document +0.00025348268598695737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0086_text_document +0.00032333206004171266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0087_text_document +0.00030755087408648437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0088_text_document +0.00023006508933660387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0089_text_document +0.00023529378653763827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0090_text_document +0.0002316006671871909 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0091_text_document +0.0002467080329046101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002812385280195195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0093_text_document +0.0002999655363830447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0094_text_document +0.00030366253916544147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0095_text_document +0.00034483134052353947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0096_text_document +0.0002264669007084511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0097_text_document +0.0002601377797129039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0098_text_document +0.000243683175313779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0099_text_document +0.0002458323373867855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0100_text_document +0.00023061358738763293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0101_text_document +0.0002383240957413279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0102_text_document +0.00024652411741760106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0103_text_document +0.00024356064371899462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0104_text_document +0.00023826916720633669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0105_text_document +0.00023583636824734604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0106_text_document +0.00023310828235332517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0107_text_document +0.00024133699058477928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0108_text_document +0.00023757818755491814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0109_text_document +0.00024650642737935284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0110_text_document +0.00023587507176169633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0111_text_document +0.0002394516652010616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0112_text_document +0.00026115753562452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0113_text_document +0.00023919185015293048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0114_text_document +0.0002328737948830104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0115_text_document +0.0002449581587150213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0116_text_document +0.00023488566807302266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0117_text_document +0.0002461692650286432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0118_text_document +0.00023193321359714746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0119_text_document +0.00024814319189332457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0120_text_document +0.0002502054369100928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0121_text_document +0.0002294119999864264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0122_text_document +0.00023986985689573848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0123_text_document +0.00023333209217509475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0124_text_document +0.0002268247786450586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0125_text_document +0.0002289098412617007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0126_text_document +0.00023635954118858026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0127_text_document +0.00024647215050850076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0128_text_document +0.00024326708810109974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0129_text_document +0.0002931046025004214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0130_text_document +0.00022529330733557138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0131_text_document +0.00024288319647667783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0132_text_document +0.0003170441859608398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0133_text_document +0.00032183678547706126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0134_text_document +0.00020557308761968548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0135_text_document +0.00020890924417592562 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0136_text_document +0.00021111297420597103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0137_text_document +0.00021993650550023244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0138_text_document +0.0002123163519100286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0139_text_document +0.0002103629651549111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0140_text_document +0.00021370932994199264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0141_text_document +0.00020399994203827728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0142_text_document +0.00021563034464531022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0143_text_document +0.0002119386189866467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0144_text_document +0.00020333697838057754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0145_text_document +0.00020812225502998168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0146_text_document +0.0002192034455873437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0147_text_document +0.0002146433860256116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0148_text_document +0.00022498320338620924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0149_text_document +0.00020605974297327904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0150_text_document +0.00020911517614300505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0151_text_document +0.00022086517759478398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0152_text_document +0.00021332423639106333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0153_text_document +0.00020576019154376813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0154_text_document +0.00020504347709097317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0155_text_document +0.00020777754226086552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0156_text_document +0.00021294564928541406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0157_text_document +0.00020775275197134613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0158_text_document +0.00021002644029417448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0159_text_document +0.00021013797882725636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0160_text_document +0.00019076903434985646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0161_text_document +0.00019137766426576477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0162_text_document +0.0001841037351078922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0163_text_document +0.0001952863228508793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0164_text_document +0.00018602295481911772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0165_text_document +0.0001931370361427833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0166_text_document +0.0001801085437374987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0167_text_document +0.000188289716886196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0168_text_document +0.0001852865203803285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0169_text_document +0.00018892492640726607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0170_text_document +0.0001867706345514145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0171_text_document +0.00018688900901065678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0172_text_document +0.00018978617486719294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0173_text_document +0.00019074400515584856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0174_text_document +0.00018895644551080948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0175_text_document +0.0002014139475504348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0176_text_document +0.00019178652165604014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0177_text_document +0.00019538713758341256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0178_text_document +0.00019221603071045457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0179_text_document +0.00018559646736351844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0180_text_document +0.00018839424919962872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0181_text_document +0.00031369302654824313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0182_text_document +0.00029283955302533026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003003216050130351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0184_text_document +0.00030560088357585723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0185_text_document +0.00030852297965873606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0186_text_document +0.00030137151200383515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0187_text_document +0.000287675564141583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002865118305148982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002942394807592494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0190_text_document +0.0002892999122858095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0191_text_document +0.00029726222843843534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0192_text_document +0.0002865106197035132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0193_text_document +0.0002831383377282607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0194_text_document +0.00029911101649033976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0195_text_document +0.0002879193266837814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0196_text_document +0.000293888834619463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0197_text_document +0.00028471984768159116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0198_text_document +0.0002880090219919074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0199_text_document +0.0002916398711835823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0200_text_document +0.00029790830243728387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0201_text_document +0.00028328873748227157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0202_text_document +0.000295084201372288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0203_text_document +0.0002870500420988019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0204_text_document +0.00028061238206088403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0205_text_document +0.00028268741759946835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0206_text_document +0.0002832900433124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0207_text_document +0.0002821269671667503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0208_text_document +0.00028388007298379026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0209_text_document +0.0002811354392519064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0210_text_document +0.0002811576793347316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0211_text_document +0.000291266961761568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0212_text_document +0.0002930917058536775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0213_text_document +0.00029247722771384336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0214_text_document +0.00030253733431717943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0215_text_document +0.0002988938219536017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0216_text_document +0.0003002888817617649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0217_text_document +0.00028686614758997625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0218_text_document +0.00032046548753382687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0219_text_document +0.00027752519729998216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0220_text_document +0.00026529350985605245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0221_text_document +0.0002654493836819182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0222_text_document +0.00026232091015406547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0223_text_document +0.0002599081762104853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0224_text_document +0.0002835817651903514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0225_text_document +0.00026294839748242733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0226_text_document +0.0002610835823452124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0227_text_document +0.000260110886669002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0228_text_document +0.000253371820236557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0229_text_document +0.0002581811396117453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0230_text_document +0.0002514852630632709 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0231_text_document +0.00025726705673313424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0232_text_document +0.00025592912496079053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0233_text_document +0.00025012268192543976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0234_text_document +0.00024391340520007348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0235_text_document +0.0002384383639062725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0236_text_document +0.00023975576001149118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0237_text_document +0.0002338016280970284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0238_text_document +0.0002439200883556984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0239_text_document +0.00024142268942556778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0240_text_document +0.0002427966777591219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0241_text_document +0.00024280144153436732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0242_text_document +0.00024065658615901044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0243_text_document +0.00024455143739741974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0244_text_document +0.00023239795390635735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0245_text_document +0.0002582911684560293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0246_text_document +0.00024625861259252923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0247_text_document +0.0002391576312805854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0248_text_document +0.000238078180343909 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0249_text_document +0.00023486425304981024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0250_text_document +0.0002355893518655022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0251_text_document +0.0002366129403678232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0252_text_document +0.00023595832035066449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0253_text_document +0.00023327574008525872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0254_text_document +0.00024148789011315923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0255_text_document +0.0002373778500991465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0256_text_document +0.00023955987733466374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0257_text_document +0.000230949882722363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0258_text_document +0.00023691636140836262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0259_text_document +0.0002296963977634624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0260_text_document +0.0002332661069034444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0261_text_document +0.00023843042502126992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0262_text_document +0.00023511746712743498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0263_text_document +0.0002347369877896436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0264_text_document +0.0002323753243697275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0265_text_document +0.00026669348300156857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0266_text_document +0.00025799845912273273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0267_text_document +0.00027628560903016796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0268_text_document +0.00026519284616643963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0269_text_document +0.00026441815097637077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0270_text_document +0.0002662131391195505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0271_text_document +0.00027728803868991606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0272_text_document +0.0002769764618252775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0273_text_document +0.00027646939593325287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0274_text_document +0.0002624622460988396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0275_text_document +0.0002597094641937235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0276_text_document +0.00026414993058715923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0277_text_document +0.00027056496256926013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0278_text_document +0.0002594411680362496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0279_text_document +0.00026263805833060905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0280_text_document +0.0002560343870682032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0281_text_document +0.0002624349038750109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0282_text_document +0.00025919416325410714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0283_text_document +0.0002611522977423299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0284_text_document +0.00023679129688303509 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0285_text_document +0.0002424050866477902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0286_text_document +0.00022701047777126036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0287_text_document +0.00023885339653333248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0288_text_document +0.00024106734540671208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0289_text_document +0.0002258801520250309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0290_text_document +0.0003279882524990489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0291_text_document +0.00033565261995537515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0292_text_document +0.0003289323356607256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0293_text_document +0.0003074095430777535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0294_text_document +0.0003207680812935341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0295_text_document +0.00031455349141131964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0296_text_document +0.0003292847953027658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0297_text_document +0.0003336588045388259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0298_text_document +0.00031509118791912046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0299_text_document +0.0003142598967986839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0300_text_document +0.00030783273695855995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0301_text_document +0.0003180584048660508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0302_text_document +0.0003132932087805931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0303_text_document +0.00031883257979717144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0304_text_document +0.00030944547256766847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0305_text_document +0.00030308947812968015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0306_text_document +0.00027546560713402303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0307_text_document +0.0002849896883269672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0308_text_document +0.00028854314233644503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0309_text_document +0.00028915140229591915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0310_text_document +0.00028785031389006415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0311_text_document +0.00029386612956137296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0312_text_document +0.00027190973100817075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0313_text_document +0.00028482862326451903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0314_text_document +0.00028103519882799385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0315_text_document +0.00027510038584601916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0316_text_document +0.00028413351954904745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0317_text_document +0.0002766838847779375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0318_text_document +0.00026734717208098886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0319_text_document +0.0002798212098651715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0320_text_document +0.0002747771651023886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0321_text_document +0.0002653649112010507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0322_text_document +0.0002631895073950362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0323_text_document +0.00027233897055462913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0324_text_document +0.00026295942114759743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0325_text_document +0.00030523368071333024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0326_text_document +0.00022951852300606208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0327_text_document +0.00022441558532523096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0328_text_document +0.00022508048810748277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0329_text_document +0.00021854625167048365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0330_text_document +0.00032578339433634126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0331_text_document +0.0003234065091465547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0332_text_document +0.00031578848940780525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0333_text_document +0.0003211733834987297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0334_text_document +0.00030598592011548813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0335_text_document +0.00030636342203205056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0336_text_document +0.0003057832116313887 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0337_text_document +0.000314036788141844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0338_text_document +0.00030966829419359915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0339_text_document +0.00030590256959722885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0340_text_document +0.0003098044211320355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0341_text_document +0.00031610551467687426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0342_text_document +0.0003181946275637243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0343_text_document +0.00030594263323826957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0344_text_document +0.0003126680759448145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0345_text_document +0.0002992280964722656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0346_text_document +0.00029925238994904177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0347_text_document +0.0003002679127100512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0348_text_document +0.00029525568123898354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0349_text_document +0.0003024653097967333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0350_text_document +0.0002953978348393056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0351_text_document +0.0003002611325611784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0352_text_document +0.0002957202302765213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0353_text_document +0.00029316969879070013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0354_text_document +0.00029927093466316167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0355_text_document +0.00029673566591636904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0356_text_document +0.0002937689672539696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0357_text_document +0.0002973606684406085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0358_text_document +0.0002964111065178358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0359_text_document +0.0003023024169175062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0360_text_document +0.0003023653161749783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0361_text_document +0.0003041586406248139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0362_text_document +0.00029561553630767535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0363_text_document +0.00024185982713467274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0364_text_document +0.00023843085692504566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0365_text_document +0.00024640440430345615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0366_text_document +0.0002514283272863322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0367_text_document +0.0002428429062712565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0368_text_document +0.00023806417358106035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0369_text_document +0.000241345504518809 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0370_text_document +0.00023475737093303525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0371_text_document +0.00024315922889458298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0372_text_document +0.0002509834540572025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0373_text_document +0.00025303820591366467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0374_text_document +0.00023678822937901864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0375_text_document +0.00023171129872234371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0376_text_document +0.00024461347186013167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0377_text_document +0.00023799008209254456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0378_text_document +0.00023090419051131675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0379_text_document +0.0002236725770641727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0380_text_document +0.00023567214707890686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0381_text_document +0.0002262722125540663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0382_text_document +0.00034312492202384507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0383_text_document +0.00021814471912144287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0384_text_document +0.00023259303719099642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0385_text_document +0.00031953022508126173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0386_text_document +0.00023554778297810253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0387_text_document +0.0002460294175476815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0388_text_document +0.0002407153820838108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0389_text_document +0.0002374237316074476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0390_text_document +0.00023980889380119253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0391_text_document +0.0002511495625217406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0392_text_document +0.0002455758117178104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0393_text_document +0.00024203242698955926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0394_text_document +0.00024139601603558614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0395_text_document +0.00024286894291167163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0396_text_document +0.00023208951019510916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0397_text_document +0.0002357404012027918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0398_text_document +0.00023446333528494393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0399_text_document +0.0002366761658977476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0400_text_document +0.0002382598783135322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0401_text_document +0.00023065268726624828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0402_text_document +0.00022821836479753894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0403_text_document +0.00023184541693801962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0404_text_document +0.00023323789396160382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0405_text_document +0.00022765013762883577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0406_text_document +0.00023714308028716352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0407_text_document +0.00028689301916209046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0408_text_document +0.0003409253474017267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0409_text_document +0.0003375051344730567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0410_text_document +0.0003292176313040109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0411_text_document +0.00032955022485317955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0412_text_document +0.0003279397699428092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0413_text_document +0.0003197789907967984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0414_text_document +0.00031901270687106177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0415_text_document +0.000321273794216131 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0416_text_document +0.0003220857325921838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0417_text_document +0.00031002969769902754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0418_text_document +0.00031282247512778876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0419_text_document +0.0003087408247659614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0420_text_document +0.0003000588357430778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0421_text_document +0.0003050525128747414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0422_text_document +0.0003038755807622741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0423_text_document +0.00029692774685276133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0424_text_document +0.0003116160903862434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0425_text_document +0.00031084101832927995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0426_text_document +0.00030708350656830715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0427_text_document +0.00031743538194191725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0428_text_document +0.00031694261996253895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0429_text_document +0.0003146446823405206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0430_text_document +0.00030156651655858596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0431_text_document +0.000303240651608455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0432_text_document +0.00032558453868072364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0433_text_document +0.0002973680179620588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0434_text_document +0.0002971760577119216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0435_text_document +0.0002973002298006474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0436_text_document +0.0002878620791957177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0437_text_document +0.00029632190555443135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0438_text_document +0.0002946733596926658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0439_text_document +0.00029877307004917556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0440_text_document +0.00029551091884749816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0441_text_document +0.0002976670701108049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0442_text_document +0.0002888352867396029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0443_text_document +0.0002866799361024954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0444_text_document +0.0002859222006630905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0445_text_document +0.00028581831052887173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0446_text_document +0.00028506927387831265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0447_text_document +0.0002803249093757669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0448_text_document +0.0002809203104492272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0449_text_document +0.00028454145587367076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0450_text_document +0.00028584177277598123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0451_text_document +0.00028086934160805217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0452_text_document +0.000270936293938279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0453_text_document +0.00028304258342716634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0454_text_document +0.00028276074943094315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0455_text_document +0.0002602100764561298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0456_text_document +0.00028012504824815937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0457_text_document +0.0002608944608134916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0458_text_document +0.0002845289889094832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0459_text_document +0.0002717532367216808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0460_text_document +0.0002643974553814476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0461_text_document +0.0002758213344366294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0462_text_document +0.0002753861114186629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0463_text_document +0.00031845649723981725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0464_text_document +0.00032153756772406746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0465_text_document +0.0003223378422301534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0466_text_document +0.0002996787108131847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0467_text_document +0.00030486709979224023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0468_text_document +0.00031053773722556385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0469_text_document +0.0003002771838331003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0470_text_document +0.00029794449770130684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0471_text_document +0.0003033670930430196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0472_text_document +0.0002965031647098184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0473_text_document +0.0002837085032811094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0474_text_document +0.0002828420727162801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0475_text_document +0.00028941167269403106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0476_text_document +0.00029157564190928313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0477_text_document +0.00029812762761704826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0478_text_document +0.0002961388642406645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0479_text_document +0.0002838466433847451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0480_text_document +0.0002788779144959817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0481_text_document +0.0003402152386086791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0482_text_document +0.00037332501068667467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0483_text_document +0.0002413675200116708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0484_text_document +0.0003704235275199961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0485_text_document +0.0002379466982220781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0486_text_document +0.00035089333509974934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0487_text_document +0.00023630817154070126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0488_text_document +0.00023857309295728839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0489_text_document +0.0002435822475458576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0490_text_document +0.00023387703405383536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0491_text_document +0.00034319854187343774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0492_text_document +0.0003622737409420836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0493_text_document +0.00023570573166970698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0494_text_document +0.00022641527241191097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0495_text_document +0.00034243292431352653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0496_text_document +0.00024045245535407698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0497_text_document +0.00023676532885361976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0498_text_document +0.00022335363118071338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0499_text_document +0.00023448598925498735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0500_text_document +0.00033737048365832474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0501_text_document +0.0003357751601882351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0502_text_document +0.0003383236392673138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0503_text_document +0.0003397838415177592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0504_text_document +0.00033705937300296186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0505_text_document +0.00033448155827902774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0506_text_document +0.00034576892094196856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0507_text_document +0.00033674871522955814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0508_text_document +0.0003328110361659434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0509_text_document +0.00032432631363958473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0510_text_document +0.00032731656932112217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0511_text_document +0.00032024116066153716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0512_text_document +0.0003040305172335454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0513_text_document +0.00031659687802842567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0514_text_document +0.000303687860573204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0515_text_document +0.0003155611705529593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0516_text_document +0.00030697272991348575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0517_text_document +0.00032874805540012775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0518_text_document +0.0003195460475675836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0519_text_document +0.00029999019685462926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0520_text_document +0.0003031992730055188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0521_text_document +0.0003004957313392662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0522_text_document +0.00029242533089655584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0523_text_document +0.0002940539652538529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0524_text_document +0.0003042748602544184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0525_text_document +0.00029329988520120374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0526_text_document +0.00028533980088048884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0527_text_document +0.0002995523399640371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0528_text_document +0.00024445982369612285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0529_text_document +0.0002341949821161716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0530_text_document +0.0002448827406649086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0531_text_document +0.0002464661023748273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0532_text_document +0.0002458273043503861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0533_text_document +0.000234131092194839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0534_text_document +0.00023502842288340058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0535_text_document +0.00023472409854696446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0536_text_document +0.0002353934437680525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0537_text_document +0.00023298716740292522 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0538_text_document +0.00023724345571185632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0539_text_document +0.0002463911915031484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0540_text_document +0.00023298903026561056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0541_text_document +0.00022884149754863258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0542_text_document +0.00023103945956545342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0543_text_document +0.0002444088792883614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0544_text_document +0.00022581722858094737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0545_text_document +0.0002370810502668904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0546_text_document +0.00022632319324174496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0547_text_document +0.00023710168144645038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0548_text_document +0.00022964923090952467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0549_text_document +0.00023748320722538985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0550_text_document +0.0003222624777361089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0551_text_document +0.0002939065142920207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0552_text_document +0.0003163669341858318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0553_text_document +0.0002875568128154461 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0554_text_document +0.0002891174847690085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0555_text_document +0.0002845830978145091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0556_text_document +0.0002834617830618547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0557_text_document +0.0002825955578364204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0558_text_document +0.0002770681818983043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0559_text_document +0.00027398693963975244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0560_text_document +0.00026761486776881346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0561_text_document +0.0002709662939745425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0562_text_document +0.0002715205476986883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0563_text_document +0.0002694875173937183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0564_text_document +0.0002691404382855153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0565_text_document +0.0002555696578650904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0566_text_document +0.00025938400199289785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0567_text_document +0.00025330279781755557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0568_text_document +0.00025455190919542185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0569_text_document +0.0002596474980952091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0570_text_document +0.0002593765878092823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0571_text_document +0.00026530976177812846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0572_text_document +0.00026521586959931293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0573_text_document +0.00027156192778243744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0574_text_document +0.00026542489893346987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0575_text_document +0.0002637742757379441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0576_text_document +0.0002660391549513622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0577_text_document +0.0002622961692249776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0578_text_document +0.0002668259130904866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0579_text_document +0.00026393281403990296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0580_text_document +0.0002573087912247817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0581_text_document +0.0002689284845925933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0582_text_document +0.0002587878565641303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0583_text_document +0.0002591277179432351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0584_text_document +0.00025645748667058553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0585_text_document +0.0002576834953920859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0586_text_document +0.0002574007659976351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0587_text_document +0.00026215195926907863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0588_text_document +0.0002550452573299244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0589_text_document +0.0002580549425113166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0590_text_document +0.0002580184320809385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0591_text_document +0.00026135902243793944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0592_text_document +0.0002499110939933153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0593_text_document +0.00023602977130289638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0594_text_document +0.0002179537404034863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0595_text_document +0.000217790844069029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0596_text_document +0.00021511798361299487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0597_text_document +0.00025422459968044684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0598_text_document +0.00026310640293852807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0599_text_document +0.0003408740036680742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0600_text_document +0.00025777786217145044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0601_text_document +0.00025244460970438263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0602_text_document +0.00025351648924446906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0603_text_document +0.0003423231978018855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0604_text_document +0.0003423953052478566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0605_text_document +0.0003318569148201118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0606_text_document +0.00032767638761629247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0607_text_document +0.00033215390937927025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0608_text_document +0.00032618622802635676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0609_text_document +0.00032507622347617733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0610_text_document +0.00031030763419557833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0611_text_document +0.00024643590119480534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0000_text_document +0.0002095902169870633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0001_text_document +0.00021403593088797332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0002_text_document +0.0002227102409085263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00020197706221244385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00022874875522106917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00023280550472601052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0006_text_document +0.00022749578163623905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0007_text_document +0.00023802912323224644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0008_text_document +0.00023176496190267302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0009_text_document +0.0002278986856648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0010_text_document +0.00021833909531790053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0011_text_document +0.0003080057114591217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0012_text_document +0.00021694016663911526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0013_text_document +0.0002668830492707773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0014_text_document +0.00024523658363304193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0015_text_document +0.0002894756615830288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00020347856162111349 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002134325832786435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00021673235231198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0019_text_document +0.0002654127125833355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0020_text_document +0.0002158672209137081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00023947604851382316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0022_text_document +0.00026152140024106367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0023_text_document +0.00021518621527788343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0002439782139658387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0025_text_document +0.0002905141391659118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0026_text_document +0.00021642682185339982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00019960430947798375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00026322267340937706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022334429465509248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0030_text_document +0.00022855119280875728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00028578658731994404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0032_text_document +0.0002584277862839571 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0033_text_document +0.00021861958226794765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00026614391185475836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00028970533715167736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0036_text_document +0.0002235814952215254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0037_text_document +0.00022032188312044515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0038_text_document +0.00022884461811511293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0002551680347396578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0040_text_document +0.00022883355545520197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0002232938120141678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0002691617763064546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00023572139842386745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0002552819803341825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0045_text_document +0.00027155660031106415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0046_text_document +0.00021551548292117663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0047_text_document +0.00020620735756494168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0002166820604491231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0049_text_document +0.00018501398539579828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0050_text_document +0.00027225222848112053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00023371832644559636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00023566702124489628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0053_text_document +0.00023686334707090557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0054_text_document +0.00022423975285568458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0055_text_document +0.0002528257228301147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0002561855163693918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00022810786925037496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0058_text_document +0.0002762405538154904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0059_text_document +0.00022261162863844723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00022540915157909426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0061_text_document +0.00022299985657677767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0062_text_document +0.00022755525774778565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00024165856540482104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0064_text_document +0.00025687628451136137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0065_text_document +0.0002231870244226192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0066_text_document +0.00026580529164370396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0067_text_document +0.00028870521089646587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00021540624754582923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00025778332069476944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0070_text_document +0.00021926796929661694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00026029886649394187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00022285796310592967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0073_text_document +0.00023080628286139754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0074_text_document +0.00025245808263416443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00022457772027503216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0076_text_document +0.00024435224362284627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0077_text_document +0.00022526086938759533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0002673487094116284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00023263089713557213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0080_text_document +0.00021778225362633044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0081_text_document +0.00021409630017652816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0082_text_document +0.00022267016739539933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00020585884947224638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00022993683686780696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0085_text_document +0.00024242353683668374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0002092411836993767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0087_text_document +0.0002197488902020793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00021875038642425168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0002494827261520774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0090_text_document +0.00023601123399284122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0091_text_document +0.00021826172481591926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0092_text_document +0.000236632672200321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00025074570040713444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00022642420961164095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00023812142057551977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0096_text_document +0.0002428821562055837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00022488741946885592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0098_text_document +0.00020317409833506262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0099_text_document +0.00021856439903312987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0100_text_document +0.0002106925714107645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0101_text_document +0.00021119826681040816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0002592340274790045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00023255611509461946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0104_text_document +0.00020894883617804318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00022615604129768463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0106_text_document +0.000203728797783905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0107_text_document +0.0001989690950208705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0108_text_document +0.00021734356057002846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0109_text_document +0.0002433390106922548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00022031295850762523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00022344289507866802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0112_text_document +0.00022230083290263739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0113_text_document +0.00021439002065826426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0002041951415667326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0115_text_document +0.00022877491032651992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00021999090587860643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0117_text_document +0.00025682432698074305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0118_text_document +0.00024400030399295212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0119_text_document +0.00022789294060424558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0120_text_document +0.00021497724986548528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0121_text_document +0.00023813142494777905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0122_text_document +0.00021895635220322673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00023328497887722523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0124_text_document +0.00022164528342855325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0125_text_document +0.0002484042811809953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00021121568758750245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00020558498767931708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00024543621326022564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00019902438240619879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00023691721805865155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0131_text_document +0.00021791494779355714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0132_text_document +0.0002240264291639859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0133_text_document +0.0002473539109425455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0002071473371471445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00021022258828332134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0136_text_document +0.00022311670653909265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0137_text_document +0.00022930107525031038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0002214421423002716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00021570132519262982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002197681200389886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0141_text_document +0.0002800029152388595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0142_text_document +0.00026843440765131945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002849765317975514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0144_text_document +0.00027096319463304773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027086227426919104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002526247335698449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0147_text_document +0.00027363240217034764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0002623467059155748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0149_text_document +0.00027346078063921375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0150_text_document +0.00025920642956814055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0151_text_document +0.00025705335691494745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0152_text_document +0.00025922805782841715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0153_text_document +0.0002788336705199961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0154_text_document +0.00024845909125095083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00028656519284339746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0156_text_document +0.00025647131598268287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0157_text_document +0.0002784068234736532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0158_text_document +0.0002528120161786896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0159_text_document +0.0002488190053053583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0160_text_document +0.0002704389893183884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0161_text_document +0.00025616941425622545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0162_text_document +0.00026029019534693783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00025685556571703545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0164_text_document +0.00019723833812640722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0165_text_document +0.0001895418580073486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00019011078486016846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0167_text_document +0.00018779376696334834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0168_text_document +0.00018563641007150188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00018754827458482748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0170_text_document +0.00019755194962803275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00028610572842390993 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0172_text_document +0.00019902354772130188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0173_text_document +0.00020283251106846995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00018722834815639619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0175_text_document +0.00018348325202476222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002739432916909774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00014534657139819037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00015282753276716084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0001549244865585569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0180_text_document +0.0001465220076427807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0181_text_document +0.00015309131688759006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0001462273984264752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0183_text_document +0.00014903597785697923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0184_text_document +0.0001547302246314982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0001486478323505694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00014887945296702178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00014582128695700495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00015040846513981096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0001492663985213415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0001491503509128408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0191_text_document +0.00014485595166153977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0192_text_document +0.00014471245274265675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0193_text_document +0.0001539836098505113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0194_text_document +0.00014269340600113259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0001366015589763494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0196_text_document +0.00014275967558886846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0197_text_document +0.00012216291308335102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0198_text_document +9.860253447438225e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0199_text_document +0.00013395002197992724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0200_text_document +0.00013095775634161855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0201_text_document +0.00013244501748701574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0202_text_document +0.00013344638268905827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00013599432127141194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0204_text_document +0.0001319495730149868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0001286425479982177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0001288175023456875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0207_text_document +0.00014061678080985136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0208_text_document +0.000128553766351679 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00013865417327932483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00012918889813006947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0211_text_document +0.00013369372633056305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0212_text_document +0.00012818148109232114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00013087168186794624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0214_text_document +0.00012209941459024034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0001170049632015973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0216_text_document +0.00013033065279061172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00012782387759971287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0218_text_document +0.00012594444140907917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0219_text_document +0.00012747350244869554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0220_text_document +0.00011189052700824495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0221_text_document +0.000118474284791765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0222_text_document +0.00012947220948400783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0223_text_document +0.00011563584378100779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0224_text_document +0.00012898102925965738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0225_text_document +0.000122859118523654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0226_text_document +0.00013841949453733798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00012735223374055142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00013005120882648248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0229_text_document +0.000133953509788018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00012898361006981912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00012385687424414202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0232_text_document +0.00012495169231715962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0001334287109141697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0234_text_document +0.0001251557347669207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00012458204389205325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0236_text_document +0.00013142493999218836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0001234876747521603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0238_text_document +0.00011414056156548952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0239_text_document +0.00023536944102421793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0240_text_document +0.00020899836320101376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00020694945512603853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0001985515975806629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0243_text_document +0.00020332234597425947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00019901014809176087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0245_text_document +0.00019730742496077176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0246_text_document +0.0002086531104287768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00019880240459684486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0248_text_document +0.0001934729054969894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0249_text_document +0.00020006177554040137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0001941325758266985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00020329878081065027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0252_text_document +0.00020327608562464652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00019798005487177493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0001954984594242001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0001990223203741723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0256_text_document +0.00019108660381768295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00019716779886134537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0258_text_document +0.0001928475026596504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0259_text_document +0.00019634937526499807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0260_text_document +0.00019298574642019224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0261_text_document +0.00018884134414178089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0262_text_document +0.00018997833083144106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0263_text_document +0.0001905325885044214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0264_text_document +0.00020263821458910917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0002079379871094917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0266_text_document +0.00019785431238092052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0267_text_document +0.00018722610077594935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00019937636744768995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00018558334637361332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0270_text_document +0.00019000469868035166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0001853064471865308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000189466635918149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0273_text_document +0.00019109828052136198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0274_text_document +0.00018290456266579745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0275_text_document +0.00017877060456109023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0276_text_document +0.00018344271945962216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0277_text_document +0.0001937669621232641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0278_text_document +0.00019434311583686195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0279_text_document +0.0001805150932807986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0001914582846585569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00020025771498172507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0282_text_document +0.00019924956568197525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0283_text_document +0.000189496868442045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0001929642820365483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0285_text_document +0.0001903124937955297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0286_text_document +0.00019497565890742164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00018960064504727124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00018568951646616373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0289_text_document +0.00018239686989629257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00018605553146990633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0291_text_document +0.0001844096767388669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00017898307999377337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0001739406120499752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0294_text_document +0.0001911537409150027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00017663348174413226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00017913373123918278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00017455805527093036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0298_text_document +0.00017536417503931625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00017329247651270448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0300_text_document +0.00017912565587258707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0301_text_document +0.00017228776664782256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0302_text_document +0.0001825947205735245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0001696263054898423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00017175867341643253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0305_text_document +0.0001668734295531042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0306_text_document +0.00016312507834781404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0001687262224636195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00017236097186979052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0309_text_document +0.0002586993024691808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00026219934972577114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0311_text_document +0.0002566784476550503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0312_text_document +0.0002530671575343629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00025526495987018773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0314_text_document +0.0002510505062545801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00024743741398453804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00024882602559273036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00024230881628338428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00025005854915078414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0319_text_document +0.00024477471955617643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0320_text_document +0.0002480463985551468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0321_text_document +0.00024335328103980772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00024464696562773777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0323_text_document +0.00023820565587951385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0324_text_document +0.00024537554558786237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0325_text_document +0.00024052017934692743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00023660347377746528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00023823292504990384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00023564543049854766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002370415962271789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00023453319757168757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0331_text_document +0.000236480621339876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002391149628895737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0333_text_document +0.00023165934662137285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00023331169915961683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0335_text_document +0.0002348226454144718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0336_text_document +0.00023564045570745751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0337_text_document +0.00016411316830860297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002007359738791159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0339_text_document +0.00019930606930833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0340_text_document +0.00019598670739211644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00019115600211637036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0342_text_document +0.00018957338451495675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0343_text_document +0.0001997256344570198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0344_text_document +0.0001924339501051294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0345_text_document +0.0001929492409258573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00019129356692417672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0347_text_document +0.0001927097658307402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0348_text_document +0.00018744016832935095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00018898826127054628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00019337725386559253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0351_text_document +0.00018434878571055096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0352_text_document +0.00018454731188528818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0353_text_document +0.00018197801455061398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00018615322144032256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0355_text_document +0.00017981075274777777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00018028813451030057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0357_text_document +0.0001760055343765487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0358_text_document +0.00018306121836089844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00018110213343756692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00017839531596627688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0361_text_document +0.00017668405792307465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00018382867977972885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0363_text_document +0.00017812146256462094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0364_text_document +0.00017866992260811773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00017457542446637375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00017144357690622488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0367_text_document +0.00017669299438239817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0368_text_document +0.00017721730286035934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0002573630336497748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00025158500395961657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00025871208953576674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0372_text_document +0.0002522219361597465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0373_text_document +0.00025035546177162626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0374_text_document +0.00024714234522261514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0375_text_document +0.00024296206951019436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0376_text_document +0.00023797488747091152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0377_text_document +0.0002417964809184933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0378_text_document +0.0002317015633644362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00023529081059722227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0380_text_document +0.00022865050303533797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0381_text_document +0.00022350627510674308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0382_text_document +0.00022416076407195612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0383_text_document +0.0002237152481700081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0384_text_document +0.00022673308251184112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0385_text_document +0.00021988509315558021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0386_text_document +0.00021791186375379613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0387_text_document +0.00021902394687174658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0388_text_document +0.00022390913224296354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0389_text_document +0.0002159569838456253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0390_text_document +0.000193074631476835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0391_text_document +0.00019772710141722856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0392_text_document +0.0001918863050023569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0393_text_document +0.0001968641761834432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0394_text_document +0.00019269495646727515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0395_text_document +0.0001986463032193898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0396_text_document +0.0001855871337995234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0397_text_document +0.00019041152711008963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0398_text_document +0.00018277849340888642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0399_text_document +0.00018810546599505484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0400_text_document +0.00018711834399232793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0401_text_document +0.000180125082690484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0402_text_document +0.00023744084906469025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0403_text_document +0.00023803845013258319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0404_text_document +0.00023586547263857976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0405_text_document +0.00023222402329423718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0406_text_document +0.00023270999204422837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0407_text_document +0.00023378783679246331 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0408_text_document +0.00017304047941651873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0409_text_document +0.00017585076104150352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0410_text_document +0.00017101296884180275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0411_text_document +0.00017561096140154923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0412_text_document +0.0001713420333669203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0413_text_document +0.00022603582939637927 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0414_text_document +0.0001703733924033566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0415_text_document +0.0002396801442728503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0416_text_document +0.0001676400523382032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0417_text_document +0.00017193438273170229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0418_text_document +0.00017529021040710947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0419_text_document +0.0001630741415909194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0420_text_document +0.00024179471702347313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0421_text_document +0.00016581358754145113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0422_text_document +0.0002456894490564403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0423_text_document +0.0002456073517995372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0424_text_document +0.00024937580109172706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0425_text_document +0.0002457208726475487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0426_text_document +0.00024399607429757567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0427_text_document +0.00023977003702270238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0428_text_document +0.0002453131498067917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0429_text_document +0.0001621090466807557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0430_text_document +0.00024557101413066944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0431_text_document +0.00024662307150866836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0432_text_document +0.00015758980646827074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0433_text_document +0.00024391288666874046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0434_text_document +0.00023509503922816786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0435_text_document +0.00023489539281843744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0436_text_document +0.00023286637378893443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0437_text_document +0.00023379369093964089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0438_text_document +0.00023205784424428202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0439_text_document +0.00023009948269807432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0440_text_document +0.00023187584394201576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0441_text_document +0.00023202252759594008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0442_text_document +0.00022728777233539934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0443_text_document +0.00022582666382743133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0444_text_document +0.00022616733175598707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0445_text_document +0.00022768677294110565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0446_text_document +0.00022367789565066836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0447_text_document +0.00022752055218158585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0448_text_document +0.00021819243338256605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0449_text_document +0.0002241455531613807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0450_text_document +0.00022437797440403226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0451_text_document +0.00022445007197791702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0452_text_document +0.00022150502971124016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0453_text_document +0.0002225145672731263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0454_text_document +0.00022368982014371355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0455_text_document +0.00022402755606263736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0456_text_document +0.00023016090138940315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0457_text_document +0.0002260342841680707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0458_text_document +0.00022458279279977673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0459_text_document +0.00021839974448010203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0460_text_document +0.0002264409368746725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0461_text_document +0.000223550215762877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0462_text_document +0.00021610601829010048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0463_text_document +0.00022408120517524368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0464_text_document +0.00021671066876802013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0465_text_document +0.00016072298972169641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0466_text_document +0.0001722371396276357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0467_text_document +0.00017303760343097654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0468_text_document +0.00016820735177759604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0469_text_document +0.0001782239553050235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0470_text_document +0.0001749477598265696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0471_text_document +0.0001700037698924768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0472_text_document +0.0001721297434219665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0473_text_document +0.00017082606704868714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0474_text_document +0.00017400024710211123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0475_text_document +0.00017016210162102983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0476_text_document +0.00016745166973214216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0477_text_document +0.0001684428163376526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0478_text_document +0.0001648685852885396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0479_text_document +0.00017387645508870812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0480_text_document +0.00016594906144137858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0481_text_document +0.00016042654972698604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0482_text_document +0.00014860104507835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0483_text_document +0.00016227281398002708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0484_text_document +0.00016502091577582913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0485_text_document +0.00016106235650927743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0486_text_document +0.00015987309712264371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0487_text_document +0.0001642815421701454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0488_text_document +0.00016531915249024665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0489_text_document +0.00015833872193897052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0490_text_document +0.00015639158495488916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0491_text_document +0.00015342548972376501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0492_text_document +0.00015518921543764528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0493_text_document +0.0001621958240469728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0494_text_document +0.00015155749799598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0495_text_document +0.00014939896262383117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0496_text_document +0.00015490092509698006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0497_text_document +0.00017977881778259884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0498_text_document +0.00018596378104021417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0499_text_document +0.00017898738743182946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0500_text_document +0.00018286541046512472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0501_text_document +0.00018092409134830376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0502_text_document +0.00017788220095337013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0503_text_document +0.00017903547090898037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0504_text_document +0.0001797342122414524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0505_text_document +0.00018405110997743763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0506_text_document +0.00016587458814992502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0507_text_document +0.00018323507493237133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0508_text_document +0.00017881236669457928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0509_text_document +0.00017083385044833047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0510_text_document +0.0001730201559992492 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0511_text_document +0.00016901593018907565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0512_text_document +0.00017121838351155997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0513_text_document +0.0001762157419442059 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0514_text_document +0.00017000047903250774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0515_text_document +0.00017628842147757824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0516_text_document +0.0001760014416563697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0517_text_document +0.00017080626611158523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0518_text_document +0.00017077485831581488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0519_text_document +0.0001740210774510124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0520_text_document +0.00017310752988628116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0521_text_document +0.00016563538206915967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0522_text_document +0.0001698038028867437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0523_text_document +0.00022989652913943246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0524_text_document +0.00023802118237282655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0525_text_document +0.00023209291976691602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0526_text_document +0.00023478978296678473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0527_text_document +0.00023185674392304132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0528_text_document +0.0002223151271899996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0529_text_document +0.0002212980337800594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0530_text_document +0.0002177142043482363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0531_text_document +0.00022071160791386127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0532_text_document +0.0002155092901614389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0533_text_document +0.00021709946336410436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0534_text_document +0.0002091085371649664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0535_text_document +0.00021301299764538067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0536_text_document +0.00020514046046681228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0537_text_document +0.00020554350961511138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0538_text_document +0.0002032929572669402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0539_text_document +0.00020017696773262392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0540_text_document +0.0002041760983122544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0541_text_document +0.00019610775249750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0542_text_document +0.0001972797535028649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0543_text_document +0.00019987201182946655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0544_text_document +0.00023221090921479249 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0545_text_document +0.00022866265656078542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0546_text_document +0.00022846213721182363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0547_text_document +0.00022028779604045222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0548_text_document +0.00023019534411130514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0549_text_document +0.00021499063838892918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0550_text_document +0.0002238747556640398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0551_text_document +0.000219139079337847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0552_text_document +0.00022466810662919942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0553_text_document +0.00021354111452743537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0554_text_document +0.0002116352569318229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0555_text_document +0.00021742490236552721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0556_text_document +0.00020976053145397075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0557_text_document +0.0002121893598598504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0558_text_document +0.00020611700008662688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0559_text_document +0.00020771394257887023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0560_text_document +0.00020861778045311834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0561_text_document +0.00020549717473124685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0562_text_document +0.00021168253336591858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0563_text_document +0.00020292362079976103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0564_text_document +0.0002053579978117472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0565_text_document +0.0002025742316233632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0566_text_document +0.00019721191770863706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0567_text_document +0.00020263891920926902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0568_text_document +0.0002047513235561355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0569_text_document +0.0002058192920224309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0570_text_document +0.00020762611235464895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0571_text_document +0.00020536767369033477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0572_text_document +0.000208726602681654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0573_text_document +0.00020670689006790867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0574_text_document +0.0001987029852837105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0575_text_document +0.00019743671572624558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0576_text_document +0.00020347237873346202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0577_text_document +0.00019483561225711876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0578_text_document +0.00019876706376189147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0579_text_document +0.00019418407035646924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0580_text_document +0.00019094739234588127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0581_text_document +0.00018896169178427298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0582_text_document +0.00019336957140803166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0583_text_document +0.00019246034436187084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0584_text_document +0.00019234601030075014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0585_text_document +0.00018937638801999214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0586_text_document +0.00019243149393005724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0587_text_document +0.00018564518487541217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0588_text_document +0.00018349694905090308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0589_text_document +0.00018632405912780405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0590_text_document +0.0001859374743982387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0591_text_document +0.00018735943662878573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0592_text_document +0.00018429223346416512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0593_text_document +0.00018743951405683122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0594_text_document +0.0002231790070545305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0595_text_document +0.00023691491440731282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0596_text_document +0.00022732583835977663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0597_text_document +0.00023280690754947414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0598_text_document +0.00023098339919576762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0599_text_document +0.00022742109041848038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0600_text_document +0.00023387941495424947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0601_text_document +0.00022226509841824269 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0602_text_document +0.00022342786655488707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0603_text_document +0.00022237713376406775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0604_text_document +0.00021379459835981835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0605_text_document +0.00021934823034546768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0606_text_document +0.00022299117012803982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0607_text_document +0.0002249652818475372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0608_text_document +0.00021549803647665793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0609_text_document +0.00021082391557018925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0610_text_document +0.0002063290532408184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0611_text_document +0.0002098859538424268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0612_text_document +0.00020927123951292785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0613_text_document +0.00020988114416198002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0614_text_document +0.00020708947339409333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0615_text_document +0.00020681735599881374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0616_text_document +0.00020862989695824213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0617_text_document +0.00017921617624032585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0618_text_document +0.0001869630178204498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0619_text_document +0.0001837421970952879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0620_text_document +0.0002703540624747488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0621_text_document +0.0002773524903329593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0622_text_document +0.00026751943505093036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0623_text_document +0.00026849089128670544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0624_text_document +0.00017768273890485142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0625_text_document +0.00026694167218726514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0626_text_document +0.00026851367038544275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0627_text_document +0.00017178448275206052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0628_text_document +0.00026146356857229295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0629_text_document +0.0002631494175929668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0630_text_document +0.00026756049947472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0631_text_document +0.0002600735435281443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0632_text_document +0.00026162102069795645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0633_text_document +0.0002546230805208093 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0634_text_document +0.00025384118907342997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0635_text_document +0.00024898898905737453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0636_text_document +0.0002560021645785107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0637_text_document +0.00025001876340897294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0638_text_document +0.00024817567624010623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0639_text_document +0.00025419118513633326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0640_text_document +0.00025520008446783997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0641_text_document +0.00024355226527934937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0642_text_document +0.00024233948860872504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0643_text_document +0.00024413553528635867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0644_text_document +0.00024287456234999737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0645_text_document +0.0002471744870080021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0646_text_document +0.00024318841473052868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0647_text_document +0.00024268080340573577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0648_text_document +0.000242363177173413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0649_text_document +0.00025310826613573865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0650_text_document +0.0002450433802404371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0651_text_document +0.0002429196089265994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0652_text_document +0.00023818874203405117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0653_text_document +0.00023814010078402416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0654_text_document +0.0002258262625271231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0655_text_document +0.0002359106231188901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0656_text_document +0.00023984369117779496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0657_text_document +0.00022677878582898447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0658_text_document +0.00023019334994987196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0659_text_document +0.0002326106169086802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0660_text_document +0.00023296218608853588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0661_text_document +0.00021930251468821644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0662_text_document +0.00022685746290158792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0663_text_document +0.00022204375118840136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0664_text_document +0.00022312982876300855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0665_text_document +0.00022347955655196657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0666_text_document +0.00021968416238742178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0667_text_document +0.00022148339454050315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0668_text_document +0.00022133417129237745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0669_text_document +0.00021840518548046784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0670_text_document +0.00021501258675160414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0671_text_document +0.00016302293581967305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0672_text_document +0.00018778970953587786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0673_text_document +0.00018910837173616491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0674_text_document +0.00019201324078164315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0675_text_document +0.00018432054093123207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0676_text_document +0.0001907749590824511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0677_text_document +0.00017971831966331778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0678_text_document +0.00018425384289495064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0679_text_document +0.00018353447605936826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0680_text_document +0.00018776194922919426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0681_text_document +0.000181858547251418 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0682_text_document +0.00017663862855632625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0683_text_document +0.00017879513620194847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0684_text_document +0.00017779569087388971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0685_text_document +0.00017180299093946108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0686_text_document +0.00018535528192944938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0687_text_document +0.0001710147287879884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0688_text_document +0.000173577199328182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0689_text_document +0.0001768576763304655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0690_text_document +0.0001796376911260544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0691_text_document +0.00016782824293218567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0692_text_document +0.00016074324428116396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0693_text_document +0.0001687513348299545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0694_text_document +0.00016077518171436444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0695_text_document +0.00017132313128327624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0696_text_document +0.00016190186959679132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0697_text_document +0.00016290842504820753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0698_text_document +0.00016156811558387776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0699_text_document +0.00023555665280084346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0700_text_document +0.0002284718177796522 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0701_text_document +0.00022374123273516798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0702_text_document +0.00021994541999416394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0703_text_document +0.00022338573100973358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0704_text_document +0.00022049767881647008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0705_text_document +0.00022416439897413284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0706_text_document +0.00021895546198784436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0707_text_document +0.0002142388294097341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0708_text_document +0.00022017166748084383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0709_text_document +0.00021104350754308596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0710_text_document +0.00020985391201191366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0711_text_document +0.00021778183924550787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0712_text_document +0.00021271266854227129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0713_text_document +0.0002086433619903549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0714_text_document +0.00021568150697937684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0715_text_document +0.00020764802098217656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0716_text_document +0.0002151205404833473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0717_text_document +0.00020430590834946775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0718_text_document +0.00020866543326050432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0719_text_document +0.00020818319961436583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0720_text_document +0.00020070798626764516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0721_text_document +0.00019693995826673832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0722_text_document +0.00020030234076064843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0723_text_document +0.00019788654054706263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0724_text_document +0.0001993257554824347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0725_text_document +0.00021386012383904914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0726_text_document +0.00021978412787373083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0727_text_document +0.0002175599344895926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0728_text_document +0.00021091594587352813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0729_text_document +0.0002034137316303627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0730_text_document +0.00021253423082914959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0731_text_document +0.00020816999471172712 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0732_text_document +0.00021853522405647908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0733_text_document +0.0002051944662085363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0734_text_document +0.00020978726975291983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0735_text_document +0.00020468921406556763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0736_text_document +0.0002007846124143192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0737_text_document +0.00020366090300396152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0738_text_document +0.0001993156168498017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0739_text_document +0.00020150340666889603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0740_text_document +0.00020188286325854645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0741_text_document +0.00020072267667247027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0742_text_document +0.00019591912629771525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0743_text_document +0.00020056463740447396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0744_text_document +0.0001962511050627094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0745_text_document +0.00018969020412060633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0746_text_document +0.00018711981666080213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0747_text_document +0.00019064086480658448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0748_text_document +0.0001893430509717561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0749_text_document +0.00018823938035214858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0750_text_document +0.000191049243153872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0751_text_document +0.00015215085209234548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0752_text_document +0.00013881666461144156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0753_text_document +0.0001511979467407442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0754_text_document +0.00015091819106548992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0755_text_document +0.00013896830454629422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0756_text_document +0.00014286084497610213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0757_text_document +0.00013829013170563417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0758_text_document +0.00014842506748913496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0759_text_document +0.0001621698681108632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0760_text_document +0.00025658329333000087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0761_text_document +0.0002625776226522738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0762_text_document +0.00018893904126945972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0763_text_document +0.00019173419836462428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0764_text_document +0.00024972708669590365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0765_text_document +0.0002565621075859928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0766_text_document +0.0002548091984702725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0767_text_document +0.00024781120449025493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0768_text_document +0.00024190263274768403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0769_text_document +0.00024935480538538027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0770_text_document +0.00024565807926820224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0771_text_document +0.00024335665926774057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0772_text_document +0.0002407471035651234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0773_text_document +0.00024409063432302957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0774_text_document +0.00025048184051844287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0775_text_document +2.7431736503196682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0776_text_document +0.0001542652540558753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0000_text_document +0.0001414689533672357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00014218991553196462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0002_text_document +0.00014380616486339045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00014537826992690233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00015240156803853129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0005_text_document +0.0001508299161037807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0001645724380011881 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0007_text_document +0.0001636434127327491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0008_text_document +0.0001425695379726649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0009_text_document +0.00015038309042278246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0010_text_document +0.00015551331010771582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0011_text_document +0.00014395190746068794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0012_text_document +0.00014572155617954775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00014985257363654754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00016517178815597176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00015368391453534256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0016_text_document +0.00013802907993189142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0001438832947332681 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0018_text_document +0.0001453654604013201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00015126685069470999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00014666492015973732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0021_text_document +0.00015372684675786069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0022_text_document +0.0001466694423156705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00014645983052842166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00014464707855314855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0025_text_document +0.00014224079429035223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00015150561574001976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0027_text_document +0.00014869251464718684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00014975351070572874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00014693519813853656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0030_text_document +0.00015177096878176463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0031_text_document +0.0001541385774188545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0032_text_document +0.00014686140972307025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0033_text_document +0.00014836061485888312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00015908940031748178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00014335960523511807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00014014336145596836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00014804788542816872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00014447262570766296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0039_text_document +0.0001490836674378867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0040_text_document +0.00015491171627451768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00014704465686983656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0042_text_document +0.00015578029994136968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00014442509556094932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0044_text_document +0.00016401352835433973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0045_text_document +0.0001426617272165932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00014952006301290383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00014858509055287383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0048_text_document +0.0001452147802800582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00014648995026373163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0050_text_document +0.000150292569067835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00015359505638013499 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0052_text_document +0.00014342220561517732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0053_text_document +0.00015037020981817882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0054_text_document +0.0001442503228598675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00015512168691210362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0056_text_document +0.000141978855262853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0001433238477981227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0058_text_document +0.0001522852753554881 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00015750021259583146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0060_text_document +0.0001620583984355833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0061_text_document +0.00014425968431250636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00015502607180742606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0063_text_document +0.00014808719854384823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00014037741406088144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0065_text_document +0.00014415351915599912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00014669998038063754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00014168851942590583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00016615444649487683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00017314227247280456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0070_text_document +0.00014511886160872687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0071_text_document +0.0001589885117911034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0072_text_document +0.0001468857466370262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00014409172483178647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00017524066610798787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0001423201779575328 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00014813204150867026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0001426495065609589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00015198519700337085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0079_text_document +0.0001407239353962083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0080_text_document +0.00015564799275992607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00014044706039573722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00014271692599994692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0083_text_document +0.000145622079855115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0001420329587382314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00014388823447845187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0086_text_document +0.0001386395317413269 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00014615498445222442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00014100731560794867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0089_text_document +0.0001412468938663676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0001448361986040564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00015041376595655126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00015393889374199827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0093_text_document +0.0001424230223910099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00013832238850082653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00014573052620396468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00014871061906625763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0097_text_document +0.0001474653563212365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00014332440162216428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0099_text_document +0.00013995360169386805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0100_text_document +0.0001396957447740551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00014451429874557317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00014667057760559536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00014311302174425863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0104_text_document +0.0001486303888676766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00014984904337848564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00014471364010783683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00014422564733335141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00014833706425660122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001547519654335586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0110_text_document +0.00016861028196725518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0111_text_document +0.00014655906054430117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00016230445673145143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0113_text_document +0.0001608744287595928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00014838797263124772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00013772432541929463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00014338391080519997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00013969596121954725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0118_text_document +0.00014433977111903384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00013940910504621967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00013841015875212353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0121_text_document +0.00015084460181936482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00015609034169658813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00014131566380676185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0124_text_document +0.0001489310284479002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0125_text_document +0.000142611271970708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00013893968956373896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00014729183194546773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0128_text_document +0.00013844600256987405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00014038359448051134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00014148398954188355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00014453817241187933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0132_text_document +0.00014513564218102443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00013715869534969562 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0134_text_document +0.00013954027841855143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0135_text_document +0.0001412204761634212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0136_text_document +0.00014007212448120704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00013479800901952131 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00014295404043242684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0139_text_document +0.00013573518591642275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00013489486257742554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0141_text_document +0.00014869208126259815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0142_text_document +0.0001475290332523071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00013460777613768496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00013367855194670696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00014765624643721848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0146_text_document +0.000134270744838921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0147_text_document +0.0001482262234332188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0148_text_document +0.00013864007544648005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00014100224826604942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0150_text_document +0.00013048509121512907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0151_text_document +0.00012998584056022605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0152_text_document +0.00013999889926826433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0153_text_document +0.00015896524130927037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00013250422088217822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00013171712561130657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0156_text_document +0.0001353745598377907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0157_text_document +0.00014385989862913682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0158_text_document +0.0001337953809308385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00013398910556960017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0160_text_document +0.00013559702582181232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0161_text_document +0.0001479395819777683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00013591800338063272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0163_text_document +0.0001378006151746279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0164_text_document +0.0001348466006623147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0165_text_document +0.0001333909410523815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0166_text_document +0.0001395413623736275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0001340932768114764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0168_text_document +0.0001401874805500622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0169_text_document +0.00013331732046238236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00013697107190707125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00014902856836260464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0172_text_document +0.0001360841851454116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0001430140955418351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00013769551793105646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0175_text_document +0.00013655894858384475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0176_text_document +0.00013428329448183135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00013659792851661152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0001390035871359384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00013373343260207954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0180_text_document +0.0001377384027675603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00013688101750180593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00013942483868376482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0183_text_document +0.00013069676947684327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0184_text_document +0.00013248181223347942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00013404120081582244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0186_text_document +0.0001361765930560515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00012895481023244784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0001269948854413741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00012923062571125647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00013146444734116587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00012866221788337398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00012734919091675074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00012491017673902597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00012532940253474304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00012942822588429847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0196_text_document +0.00012954876208363892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00012757889363363662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00012925483823692497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0199_text_document +0.00012887114306702046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00014305235456549959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00012924991650829868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00013939886744592149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00013473816912159447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00012443870588817695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00012352413384768962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00012363992848397884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00012876521187895858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0208_text_document +0.00012998676310669105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0209_text_document +0.00013573707197851088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00012914628304832383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00012135846145074816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0212_text_document +0.0001272491158502837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0213_text_document +0.00014048669089899133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00012821863542952837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00012843614908145614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0216_text_document +0.00012566972592748682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0217_text_document +0.00012623965035462757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00012745682281848042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00012684031670531754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00013734922167929733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0221_text_document +0.00012364311692105248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00012150014908859676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0223_text_document +0.00013255947544281956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00013080450775030287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00012642072366799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00012748944981690917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0227_text_document +0.0001272640012288133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0228_text_document +0.00012676436334132504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00011463874381385243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0230_text_document +0.0001259763726722257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00013265355691888996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00012800075083395775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00012600035320386608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00012796669337022373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0001281363666451258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00013103924202277517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00013710099201804686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00012390494315996567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00012375130141281296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00012654460329615904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00013347917998097572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00012957465780002206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00012365300899833007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00012759104863989702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0245_text_document +0.00012669826503428652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00019585621938937627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0247_text_document +0.00017199211271798405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00017151003159557893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0249_text_document +0.0001704765251017538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00016915887010107177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0251_text_document +0.000164507074917777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0252_text_document +0.0001707345009802067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00016235282921392888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0254_text_document +0.0001603312806389334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00016063927887228715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0256_text_document +0.0001682293216120587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00016945118701893779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00016510575549830714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00015878514261762818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00016058925849180358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0261_text_document +0.00016806270202025228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0262_text_document +0.0001601743221175851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0263_text_document +0.00016728028661189246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00016271828570438892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0265_text_document +0.0001663197659329172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00016211838369998094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00016174818095722866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00016660916885770873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0269_text_document +0.00016749279166083448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00015990162967327836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0271_text_document +0.00016050019425679443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00015826664805809287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00015906002765230277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0274_text_document +0.00016496336225309003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00015969348413764765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0276_text_document +0.00015888249989873604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001588217905168081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0278_text_document +0.0001579176192128451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0279_text_document +0.0001599592014593771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0280_text_document +0.00015860202306757735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00015475539919197688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0282_text_document +0.0001606154789998261 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0283_text_document +0.00015967691482799697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00015467004809542842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0285_text_document +0.00015681467419158087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0001622263618651377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00016071879902106084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00015926245724996415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00015865169965265541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0001558589009989086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00015834413702510978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00015984235618630313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0293_text_document +0.00015906347325722462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0001540401129832678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00015709268423517463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0296_text_document +0.00016150611616707217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0297_text_document +0.0001575761279522917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0298_text_document +0.00015145845456253164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00015531545597525365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0300_text_document +0.00015290580088858923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0301_text_document +0.00015077381822016696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00016026706987479596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00015143811781794564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00015335594803302406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00015760769888428818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00016811053178478525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00021456946285616728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00021300214303968855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00020349194545531642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0310_text_document +0.00021281325399560017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00020973400589848146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00020126033912157333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00020674507357011296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0314_text_document +0.00021222543863325493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002050723383820817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0316_text_document +0.00021804813803312056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002008803314227051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002150047024098784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0319_text_document +0.00020318723314588857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0320_text_document +0.00020021945595806058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00020351797666608406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00020832621085218548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0323_text_document +0.0002035206051090622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00020272338181805027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00020460676190716195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00020717814792849565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0001955827435950214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0328_text_document +0.00020417807396352577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0329_text_document +0.0002011029760914888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00020292765823625672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00020035339845060027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002019662525247444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0333_text_document +0.000206838061219021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0001941713531348939 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00010283055875342613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00010052128921034293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00021410992316202177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0338_text_document +0.0002019050315219438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0339_text_document +0.0001986035353671086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0340_text_document +0.00019334420113344198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0341_text_document +0.0002040134561840194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0342_text_document +0.00019786749210973914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00021292248961774976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0344_text_document +0.000198975254462317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00019270601369753864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0346_text_document +0.0001938662101557011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0347_text_document +0.0002029610545170106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002024962737322469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002075197885043544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00019871985248356538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0351_text_document +0.0001949694696029141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0352_text_document +0.00020180408203543252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0353_text_document +0.00019545199817763088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0354_text_document +0.00019734611243298183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0355_text_document +0.00021047242956266074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0356_text_document +0.0001968562822164333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0357_text_document +0.00019972266049102106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00020126122390730825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0359_text_document +0.00019799642896091898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002021712802087185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0001941903201275054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0362_text_document +0.00019307283352311706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00016598318480128866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0364_text_document +0.00016504803365649659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00016630327313193533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00016601923469884076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0367_text_document +0.0001681694501234557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00016859564709291555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00016845101707974437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001643037792913447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00016186624765418046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00016697344045101027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00016669715111205908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00016364850623567704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0375_text_document +0.0001634811496926281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00016825687707295152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0377_text_document +0.0001627585946667742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00016582351614544805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0379_text_document +0.0001630893218980273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0001568416151151013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0381_text_document +0.00017273044852059518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0382_text_document +0.00016016530273273665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00015777742226002822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00016385370668116144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0385_text_document +0.00016954547679602915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0001676626705219338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00016250610371947111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00016004510983519738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0389_text_document +0.000161815220320894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0390_text_document +0.00016744693680716642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0391_text_document +0.00015604191096880147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0001636895622681933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0393_text_document +0.000158886517344257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0001558972054341701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0001591533045533395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0001657955386528658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00016060726764524156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00016167923208527019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00015484394662326808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00016052047349647775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0401_text_document +0.0001582576585363055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0402_text_document +0.0001545777833300399 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00016303475566860345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0404_text_document +0.0001627904173369714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0405_text_document +0.0001567550665344843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001587287727580368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0407_text_document +0.0001606889088117574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0408_text_document +0.00016206324217472778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00015712668987045555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0410_text_document +0.0001607143430081059 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00015230600229428526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00016067822548676263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00015993580979768466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00016379843410396262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0415_text_document +0.0001533135627240871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00016861285265852845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0417_text_document +0.0001632799417656467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00015962871905586431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00015014915949133304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0420_text_document +0.00015059096546324844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0421_text_document +0.00015841934874861877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0422_text_document +0.000152377097357806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0423_text_document +0.00014942797865989248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0424_text_document +0.00015640838403734855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0001557305888039896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0426_text_document +0.00014992907934376868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00015847297170019638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0428_text_document +0.0001563057066889321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00015425884830587555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00015294599138593887 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00015307387809393826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00016021533866175615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0433_text_document +0.00015819924688246454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0434_text_document +0.00014854336050366525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00015428039783626384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00015380539006369472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0437_text_document +0.00015543551510602353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00015792640857808265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00015591945366146652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0440_text_document +0.00014809559672766608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00015190843215388426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0442_text_document +0.00014890757113683386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0443_text_document +0.0001610286090290533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0444_text_document +0.00015061787553649923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0445_text_document +0.00014811603935037767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0446_text_document +0.00015254163073097444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0447_text_document +0.00015300211863900935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0448_text_document +0.00015063192030688013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0449_text_document +0.00015300622789493292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0450_text_document +0.00015096280425750327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0451_text_document +0.00015205454100558358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0452_text_document +0.00015121161958027361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0453_text_document +0.0001493611157597698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0454_text_document +0.00015838957873196607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0455_text_document +0.0001497669779590609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0456_text_document +0.00015173657097785533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0457_text_document +0.0001542516903028995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0458_text_document +0.000149139532833868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0459_text_document +0.00014644441551246194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0460_text_document +0.00015166787754612994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0461_text_document +0.00014923555170687534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0462_text_document +0.00015589324574035403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0463_text_document +0.00015022803227804745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0464_text_document +0.00015127324533861265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0465_text_document +0.00014783676790095657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0466_text_document +0.00014927753645591052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0467_text_document +0.00014753911610765252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0468_text_document +0.00014886425094132403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0469_text_document +0.00014432622711023067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0470_text_document +0.00015087353567030766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0471_text_document +0.00015318739523991737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0472_text_document +0.00014716603935377532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0473_text_document +0.00015032310787320853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0474_text_document +0.00014425315738264723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0475_text_document +0.0001507311940067415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0476_text_document +0.0001735562949386336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0477_text_document +0.0001664225151007615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0478_text_document +0.00017016223341338198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0479_text_document +0.0001686337558140661 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0480_text_document +0.00018737654520115072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0481_text_document +0.00016696818282464752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0482_text_document +0.00017542891864931188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0483_text_document +0.000168925038877694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0484_text_document +0.0001769097096293462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0485_text_document +0.00017465563985682533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0486_text_document +0.0001704723163845607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0487_text_document +0.00017113194080906855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0488_text_document +0.00017056770492485763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0489_text_document +0.0001736825492971628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0490_text_document +0.00017060994856935613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0491_text_document +0.00017539355807018588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0492_text_document +0.00017512560274649157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0493_text_document +0.00017536288179601056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0494_text_document +0.00017214679473623093 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0495_text_document +0.00017372473469635212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0496_text_document +0.00016968876198424372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0497_text_document +0.00017328658337078598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0498_text_document +0.00016545006523949998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0499_text_document +0.0001712623636560391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0500_text_document +0.00017259544872761246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0501_text_document +0.00016731532955664165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0502_text_document +0.00017234554920296389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0503_text_document +0.00016824263782247044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0504_text_document +0.00017046154865322805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0505_text_document +0.00016701775451880861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0506_text_document +0.0001640723558698162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0507_text_document +0.00016912021224512063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0508_text_document +0.00016148128416798815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0509_text_document +0.00017033021559990035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0510_text_document +0.00016742449903581303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0511_text_document +0.00016604941440707502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0512_text_document +0.00016168218680070063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0513_text_document +0.00016545734985198287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0514_text_document +0.00016617264790719555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0515_text_document +0.00016903898126379064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0516_text_document +0.00016251470403425602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0517_text_document +0.00016741321573477316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0518_text_document +0.00016314387702135404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0519_text_document +0.00016261766224352778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0520_text_document +0.00016043765927930694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0521_text_document +0.0001581188444159775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0522_text_document +0.0001675593630876091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0523_text_document +0.00016225811098829194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0524_text_document +0.00016027854790273813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0525_text_document +0.00015477514040295668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0526_text_document +0.00016132027735084922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0527_text_document +0.00016144543812901825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0528_text_document +0.00016356924967160763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0529_text_document +0.00016721507926064277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0530_text_document +0.0001623283758093546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0531_text_document +0.00016540060361910116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0532_text_document +0.00016618517731232895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0533_text_document +0.0001661140965633334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0534_text_document +0.00016521134906101744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0535_text_document +0.0001605250452596446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0536_text_document +0.00016158626615495202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0537_text_document +0.00016348402666537893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0538_text_document +0.00015887094758334445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0539_text_document +0.00016216761850919694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0540_text_document +0.00016125922688833952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0541_text_document +0.00015719662175540762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0542_text_document +0.00016177908132776304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0543_text_document +0.0001616654955707841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0544_text_document +0.0001575744247706023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0545_text_document +0.00016594502227726776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0546_text_document +0.00016680360478028852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0547_text_document +0.00016969508752354227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0548_text_document +0.00018702211879271686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0549_text_document +0.00019358085009705273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0550_text_document +0.0001871367897387826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0551_text_document +0.00018452058370522755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0552_text_document +0.0001850164319455863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0553_text_document +0.00018589455402222413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0554_text_document +0.00018848818876445855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0555_text_document +0.00018677441309244695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0556_text_document +0.00018806266359047162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0557_text_document +0.00018742615490284408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0558_text_document +0.00018308658912909244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0559_text_document +0.00017917024956722993 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0560_text_document +0.0001796815083811096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0561_text_document +0.00018830762534435366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0562_text_document +0.0001850705756497164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0563_text_document +0.00018620607609678367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0564_text_document +0.00018735293561315315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0565_text_document +0.00018406055855123805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0566_text_document +0.00018296049025592247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0567_text_document +0.00018407127494772196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0568_text_document +0.0001809459590066732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0569_text_document +0.00018206921683271417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0570_text_document +0.0001823423927624476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0571_text_document +0.00017843504198196598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0572_text_document +0.0001849074668186014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0573_text_document +0.0001812163144813499 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0574_text_document +0.00018309068999374263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0575_text_document +0.00018500613289155086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0576_text_document +0.00017930403632760822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0577_text_document +0.0001846380543749688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0578_text_document +0.0001805411790348431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0579_text_document +0.00017815258406988848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0580_text_document +0.00017771149209661494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0581_text_document +0.000179212119800064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0582_text_document +0.0001770710081666354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0583_text_document +0.00018076802304233783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0584_text_document +0.00018266780486243524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0585_text_document +0.00017952537023013302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0586_text_document +0.00017482592939671484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0587_text_document +0.00017479307237867526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0588_text_document +0.00017947982239834899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0589_text_document +0.00017800230944457152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0590_text_document +0.0001768045667273756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0591_text_document +0.00018432659029891628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0592_text_document +0.00017860310980883306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0593_text_document +0.00017352563618741148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0594_text_document +0.000177967402241009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0595_text_document +0.0001761394507080597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0596_text_document +0.0001727461411889822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0597_text_document +0.00017520765607261058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0598_text_document +0.00017389963918978602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0599_text_document +0.00017297383567671195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0600_text_document +0.00017186248654837811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0601_text_document +0.00018016764298215066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0602_text_document +0.00017252933018279703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0603_text_document +0.0001720498259217191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0604_text_document +0.00017208910794032673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0605_text_document +0.0001638288329725128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0606_text_document +0.00015774370365565657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0607_text_document +0.00015428183891406193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0608_text_document +0.0001579263490987627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0609_text_document +0.00015679781661701012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0610_text_document +0.00015686067490532405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0611_text_document +0.00015476043642401294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0612_text_document +0.0001538144005636655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0613_text_document +0.00015471809257783847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0614_text_document +0.00014950254548936378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0615_text_document +0.00015189343275275787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0616_text_document +0.00016808135779534307 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0617_text_document +0.00015331380459020154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0618_text_document +0.00015025506525877266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0619_text_document +0.00015705079524537657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0620_text_document +0.00014843144411648014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0621_text_document +0.0001536670204340525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0622_text_document +0.00014701650982417206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0623_text_document +0.0001470830903826265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0624_text_document +0.00014669457615379322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0625_text_document +0.00015327731341039172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0626_text_document +0.00016421071093813112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0627_text_document +0.00014320086554259857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0628_text_document +0.00014733292080267092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0629_text_document +0.00014574339323444963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0630_text_document +0.00014508510524362508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0631_text_document +0.0001510667294376284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0632_text_document +0.00014448955337404646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0633_text_document +0.00015189242851477872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0634_text_document +0.0001408976680729981 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0635_text_document +0.00014495438771487836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0636_text_document +0.00014607129482780071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0637_text_document +0.0001425703250247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0638_text_document +0.00014772556798043487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0639_text_document +0.0001454755294743558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0640_text_document +0.00014604759342940054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0641_text_document +0.000144987966876031 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0642_text_document +0.00014159362399631978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0643_text_document +0.00015166107543186514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0644_text_document +0.00013872638536941069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0645_text_document +0.00014392691133816916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0646_text_document +0.00014527538230304764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0647_text_document +0.0001445241296159157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0648_text_document +0.00014566980102669863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0649_text_document +0.00014105957349679274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0650_text_document +0.00014407711883329926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0651_text_document +0.00014304333666146412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0652_text_document +0.00014480474786471068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0653_text_document +0.00014513562095603888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0654_text_document +0.00014216954843071324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0655_text_document +0.0001472056417215835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0656_text_document +0.0001411732545194045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0657_text_document +0.00014472737242668624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0658_text_document +0.0001412212585262607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0659_text_document +0.00020834639482623596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0660_text_document +0.00019484913874296875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0661_text_document +0.00019400182473285833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0662_text_document +0.000192581173021768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0663_text_document +0.0001958163408499538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0664_text_document +0.00019017201894348343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0665_text_document +0.00018748712836308062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0666_text_document +0.00019398325978096153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0667_text_document +0.00018740362852951608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0668_text_document +0.00018769931256921782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0669_text_document +0.00018841740417805205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0670_text_document +0.0001897879160564146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0671_text_document +0.00018663113185306689 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0672_text_document +0.00018894652949372258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0673_text_document +0.0001929378648272062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0674_text_document +0.00019134942047365448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0675_text_document +0.00018699153383533985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0676_text_document +0.00018610331853766602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0677_text_document +0.0001863160274451902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0678_text_document +0.00018636405144302115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0679_text_document +0.00018489348621678148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0680_text_document +0.0001860176372198307 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0681_text_document +0.00018315031813541827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0682_text_document +0.00019049993633217256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0683_text_document +0.00018374255446481207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0684_text_document +0.00017918235151102646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0685_text_document +0.00018078078222027994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0686_text_document +0.00018377134048126254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0687_text_document +0.00018119048712916442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0688_text_document +0.00018226290667237163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0689_text_document +0.00018539016766122422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0690_text_document +0.00018304864675259609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0691_text_document +0.00018006283819913595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0692_text_document +0.00017853375396011673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0693_text_document +0.0001806080666151815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0694_text_document +0.00018287085590792935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0695_text_document +0.00018102703894508278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0696_text_document +0.00017985249563069855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0697_text_document +0.00018055111208127884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0698_text_document +0.00017436715651687287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0699_text_document +0.0001750410902836745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0700_text_document +0.0001755658852086883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0701_text_document +0.00017704710809249836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0702_text_document +0.00017563712144304312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0703_text_document +0.00017646118668991032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0704_text_document +0.0001738273848965312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0705_text_document +0.00017355052248297015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0706_text_document +0.00017182494917422235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0707_text_document +0.0001796801127149085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0708_text_document +0.0001535678074475219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0709_text_document +0.00016509131806569352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0710_text_document +0.0001660762988129014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0711_text_document +0.00017181117317139103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0712_text_document +0.00016385189811495075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0713_text_document +0.00016321938466065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0714_text_document +0.0001627668114510062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0715_text_document +0.0001667874841569603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0716_text_document +0.0001647336272051215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0717_text_document +0.00015927038206724374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0718_text_document +0.000163069807004626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0719_text_document +0.00016643362662749963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0720_text_document +0.0001598347201275479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0721_text_document +0.00016414824852047793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0722_text_document +0.00016387374849716915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0723_text_document +0.00016218986007283508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0724_text_document +0.00016170100645242406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0725_text_document +0.00016794279442600715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0726_text_document +0.00016410407241508566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0727_text_document +0.00016663924614304762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0728_text_document +0.0001610334643678992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0729_text_document +0.00016082817926927476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0730_text_document +0.00016483710320531984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0731_text_document +0.00015950564573034403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0732_text_document +0.00016176598872010603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0733_text_document +0.00016374799045777884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0734_text_document +0.00016207070843359862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0735_text_document +0.000161310121195263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0736_text_document +0.0001590930806312555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0737_text_document +0.00015872700071854542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0738_text_document +0.0001601426608559989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0739_text_document +0.0001592737504230903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0740_text_document +0.0001599609389465664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0741_text_document +0.0001573951015313392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0742_text_document +0.00015918138446881715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0743_text_document +0.00016063409035052854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0744_text_document +0.00015479247307168076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0745_text_document +0.0001590206266750552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0746_text_document +0.00016413616409963463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0747_text_document +0.00015909403254717725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0748_text_document +0.00015912638065916792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0749_text_document +0.00015509170291798033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0750_text_document +0.00015668221053756931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0751_text_document +0.00015993661313870757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0752_text_document +0.00015986553041529475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0753_text_document +0.0001551253906720823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0754_text_document +0.0001569044427999477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0755_text_document +0.00015512319487328638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0756_text_document +0.00016021882869106635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0757_text_document +0.00015415106017838012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0758_text_document +0.00015711650631982987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0759_text_document +0.00015512670736159294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0760_text_document +0.00016200410442893923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0761_text_document +0.00015949285619573655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0762_text_document +0.0001625616727060612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0763_text_document +0.00016316486655764686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0764_text_document +0.0001571167311565954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0765_text_document +0.00016128213234978153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0766_text_document +0.00015535324730882956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0767_text_document +0.0001579934311592013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0768_text_document +0.00015195311864613838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0769_text_document +0.0001615190125670139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0770_text_document +0.00015867133202388371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0771_text_document +0.00015932910049616658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0772_text_document +0.00015735730575532447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0773_text_document +0.00016192787415292593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0774_text_document +0.00015443514945271916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0775_text_document +0.00015290872574095856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0776_text_document +0.0001586657525675075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0777_text_document +0.0001561292345081933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0778_text_document +0.0001584146414910674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0779_text_document +0.00015282231142071527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0780_text_document +0.0001561252202711004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0781_text_document +0.00015508367049496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0782_text_document +0.00015211947613405347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0783_text_document +0.00014976529550875275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0784_text_document +0.00015418186133444713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0785_text_document +0.00015777360151582686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0786_text_document +0.000152640262498424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0787_text_document +0.00015418142572863903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0788_text_document +0.00015502601134089746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0789_text_document +0.00015405733434421877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0790_text_document +0.00015484459497253604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0791_text_document +0.0001541867208689297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0792_text_document +0.00015014404352940876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0793_text_document +0.00015357544967633106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0794_text_document +0.00015037823631794736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0795_text_document +0.00015025795679285078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0796_text_document +0.00014876992710553488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0797_text_document +0.00015032669711698612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0798_text_document +0.00015596697517010466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0799_text_document +0.00015498394440674378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0800_text_document +0.00014757314272111684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0801_text_document +0.00014919071614611802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0802_text_document +0.00014686280514246915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0803_text_document +0.00015882771228777683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0804_text_document +0.00014763597756322578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0805_text_document +0.00014785441795725526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0806_text_document +0.00015313024795352964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0807_text_document +0.0001497627986113246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0808_text_document +0.00014499607432690722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0809_text_document +0.0001461719027401259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0810_text_document +0.00014839933441537366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0811_text_document +0.0001475840995029022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0812_text_document +0.00015065512711375653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0813_text_document +0.00015285087358760883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0814_text_document +0.00014861957547794477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0815_text_document +0.00014996949492468605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0816_text_document +0.0001472998668365096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0817_text_document +0.0001464012147691964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0818_text_document +0.00015227635617231567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0819_text_document +0.0001491494017117428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0820_text_document +0.00014464475787246092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0821_text_document +0.00014410767861685618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0822_text_document +0.000144919516791233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0823_text_document +0.00014507990635617585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0824_text_document +0.0001468797342896656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0825_text_document +0.0001422000420712919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0826_text_document +0.00014228987139298954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0827_text_document +0.00014481016912090385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0828_text_document +0.000142802473797815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0829_text_document +0.00014812295450003065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0830_text_document +0.00014697991622146685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0831_text_document +0.000143946325289488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0832_text_document +0.0001418544716646782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0833_text_document +0.00014706985092768576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0834_text_document +0.0001411487598988699 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0835_text_document +0.0001583983550166893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0836_text_document +0.00015370277071378533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0837_text_document +0.0001574284524004961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0838_text_document +0.00016033599900258183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0839_text_document +0.00016159470012508268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0840_text_document +0.00015624921021983388 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0841_text_document +0.0001603288323615303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0842_text_document +0.00016421653645625842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0843_text_document +0.00016136751182857813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0844_text_document +0.0001644008542307843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0845_text_document +0.00016320230298972016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0846_text_document +0.00016176830866038722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0847_text_document +0.00015883945834286212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0848_text_document +0.00015854734059433728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0849_text_document +0.00015424048326372636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0850_text_document +0.00015913631543321879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0851_text_document +0.00016242367155204024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0852_text_document +0.00016352898883564303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0853_text_document +0.00016283852574114027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0854_text_document +0.0001597064012689706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0855_text_document +0.00015723207463854053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0856_text_document +0.00016082454091186785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0857_text_document +0.00015148430437371348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0858_text_document +0.00015699196205345046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0859_text_document +0.00016323993834433252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0860_text_document +0.00015419189482936103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0861_text_document +0.00014984592429281824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0862_text_document +0.0001540327550705441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0863_text_document +0.00015559458082419316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0864_text_document +0.00015809601003355687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0865_text_document +0.00015561437781246056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0866_text_document +0.00015650965510707114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0867_text_document +0.00015654223175785975 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0868_text_document +0.00015966194232830576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0869_text_document +0.0001542791440813034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0870_text_document +0.00016358133853488976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0871_text_document +0.0001610108148402946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0872_text_document +0.0001567861463301872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0873_text_document +0.00015916579076809533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0874_text_document +0.00015834187212170972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0875_text_document +0.00015492852942470005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0876_text_document +0.0001565761307746086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0877_text_document +0.00016111787860345758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0878_text_document +0.00015262185821473176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0879_text_document +0.00015609313599061615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0880_text_document +0.00015265109415151545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0881_text_document +0.00015596676711588585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0882_text_document +0.00015602244000618423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0883_text_document +0.00015533087814847594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0884_text_document +0.000148761688602713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0885_text_document +0.00015124065708812265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0886_text_document +0.00015177148904071277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0887_text_document +0.00015551510213818192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0888_text_document +0.00015328016792414618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0889_text_document +0.00014826652573194586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0890_text_document +0.00015618973632950672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0891_text_document +0.00016465597460827412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0892_text_document +0.00017729797829003265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0893_text_document +0.00017645710877786075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0894_text_document +0.000173993320599559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0895_text_document +0.0001752697954262395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0896_text_document +0.00017545831920313468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0897_text_document +0.00017512052874093406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0898_text_document +0.00017596295211949001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0899_text_document +0.0001763343681416489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0900_text_document +0.00016737628055788186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0901_text_document +0.00017659674006013248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0902_text_document +0.00017521085067973818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0903_text_document +0.00018110203496350606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0904_text_document +0.00016887408015540739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0905_text_document +0.0001730418383091983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0906_text_document +0.00017084812178309202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0907_text_document +0.00016928946570955264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0908_text_document +0.00017272373105947043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0909_text_document +0.00016793546933797045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0910_text_document +0.00016510473373737477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0911_text_document +0.0001656625036518595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0912_text_document +0.00016849674877913583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0913_text_document +0.00017492155042464418 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0914_text_document +0.00017092357710033054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0915_text_document +0.00016970730743877006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0916_text_document +0.00016573665091766286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0917_text_document +0.00016358480536479716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0918_text_document +0.0001653802811890403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0919_text_document +0.00017231807148475074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0920_text_document +0.00017361608596973323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0921_text_document +0.00017404933358323055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0922_text_document +0.00016371945617952907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0923_text_document +0.00017000836658266155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0924_text_document +0.00017142976487027857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0925_text_document +0.00017006281434704977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0926_text_document +0.0001751965302313473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0927_text_document +0.00016954848753554936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0928_text_document +0.0001683555446267139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0929_text_document +0.00016921278107076727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0930_text_document +0.00016808682594394623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0931_text_document +0.00017711704047105475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0932_text_document +0.0001675247295876393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0933_text_document +0.00017061773073498863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0934_text_document +0.0001644856648306077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0935_text_document +0.00016530682645009105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0936_text_document +0.00016993430076157017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0937_text_document +0.00016716870217360928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0938_text_document +0.0001672477045314564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0939_text_document +0.00016150529456268964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0940_text_document +0.0001642955368396883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0941_text_document +0.0001650135010986092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0942_text_document +0.0001719916971031507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0943_text_document +0.0001663860254017646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0944_text_document +0.00016810785027934324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0945_text_document +0.00016663511368772123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0946_text_document +0.00017120237493641126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0947_text_document +0.0001651698100366788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0948_text_document +0.00016069571413445028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0949_text_document +0.0001631772602215936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0950_text_document +0.00016994484266892867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0951_text_document +0.00016821930169126347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0952_text_document +0.0001680542144940534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0953_text_document +0.00015807234911071054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0954_text_document +0.00016287290799651364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0955_text_document +0.00016674360421415713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0956_text_document +0.0001663549971877126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0957_text_document +0.0001699417467826641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0958_text_document +0.0001661066433849769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0959_text_document +0.00016736976350010906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0960_text_document +0.00016160049405253383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0961_text_document +0.0001625500850979611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0962_text_document +0.00016172349111618741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0963_text_document +0.00016041582790085466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0964_text_document +0.00016369413378455798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0965_text_document +0.00016245798272839223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0966_text_document +0.00016458727969573578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0967_text_document +0.0001618972714257936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0968_text_document +0.00016149423535800886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0969_text_document +0.00015886933917368354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0970_text_document +0.00015721961433801126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0971_text_document +0.00015609496997744904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0972_text_document +0.0001608435755282705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0973_text_document +0.00015730100598754584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0974_text_document +0.00015955845719642757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0975_text_document +0.00015469663090901824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0976_text_document +0.00015812452037199733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0977_text_document +0.00015443940925795885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0978_text_document +0.00015678701926941855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0979_text_document +0.00015787925332384637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0980_text_document +0.00015669644312439214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0981_text_document +0.00015342587917756964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0982_text_document +0.00015642024238741553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0983_text_document +0.0001540823378708023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0984_text_document +0.00015238224416999995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0985_text_document +0.0001522695061784323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0986_text_document +0.00020085620305657233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0987_text_document +0.00014698197479826313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0988_text_document +0.00014796924883111914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0989_text_document +0.0001483800966807953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0990_text_document +0.00014550940307048242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0991_text_document +0.00015052597307667803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0992_text_document +0.00014866583878918362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0993_text_document +0.00014440801314961302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0994_text_document +0.00014295564464645108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0995_text_document +0.00014903049761507035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0996_text_document +0.00014820091066353183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0997_text_document +0.0001429454882440627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0998_text_document +0.00015048550764172483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0999_text_document +0.0001430543312039796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1000_text_document +0.00014661342883839465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1001_text_document +0.00014721354013103223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1002_text_document +0.00014780017824708586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1003_text_document +0.0001463184859455721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1004_text_document +0.00014654870719379106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1005_text_document +0.00020943212095457075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1006_text_document +0.00021205821955900777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1007_text_document +0.00014176730212983274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1008_text_document +0.00014026276433980122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1009_text_document +0.00013570196535880505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1010_text_document +0.00014776685378575983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1011_text_document +0.00014138218982193943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1012_text_document +0.0001412602382122253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1013_text_document +0.00013944232659104602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1014_text_document +0.00014570617769030735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1015_text_document +0.00014233071172042007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1016_text_document +0.00014016762901851798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1017_text_document +0.0001434413757259645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1018_text_document +0.00014003324697133565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1019_text_document +0.00014567282904236987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1020_text_document +0.00013992559507863123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1021_text_document +0.00021096883039305026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1022_text_document +0.00014274603730164107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1023_text_document +0.00013914595792215918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1024_text_document +0.00013666688380542608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1025_text_document +0.00014001152690065646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1026_text_document +0.00021392615254787925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1027_text_document +0.00014251166508793392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1028_text_document +0.00013886942449587415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1029_text_document +0.0002078004025575127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1030_text_document +0.00020928673622040174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1031_text_document +0.00020558733131260538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1032_text_document +0.0002036663760886078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1033_text_document +0.00014592860566679667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1034_text_document +0.00014346325128200297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1035_text_document +0.00014068142446497316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1036_text_document +0.000142996292961803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1037_text_document +0.00020633185839414136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1038_text_document +0.00013684538988274547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1039_text_document +0.0002033768324865864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1040_text_document +0.000200593087523188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1041_text_document +0.0002297294147093001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1042_text_document +0.00022971372080690233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1043_text_document +0.00023092966691083417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1044_text_document +0.00015159247973379415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1045_text_document +0.00015257723761865372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1046_text_document +0.00015750287090187065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1047_text_document +0.00015557071949799488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1048_text_document +0.00015138603787345713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1049_text_document +0.00014966823068820163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1050_text_document +0.00015481393029806212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1051_text_document +0.0001521335747073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1052_text_document +0.00015447866363472483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1053_text_document +0.0001564823000495303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1054_text_document +0.00015484698673224505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1055_text_document +0.00022305811126444646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1056_text_document +0.00015308102523761935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1057_text_document +0.00022494528198789627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1058_text_document +0.0002206911435725598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1059_text_document +0.00021440132246946592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1060_text_document +0.00014934935094772055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1061_text_document +0.00015275047150828305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1062_text_document +0.00021692931968428998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1063_text_document +0.00023057843831795596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1064_text_document +0.00022061661869945533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1065_text_document +0.0001475889972917192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1066_text_document +0.00014965255899799802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1067_text_document +0.000146325773766483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1068_text_document +0.00021849119850040293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1069_text_document +0.00021649545481859658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1070_text_document +0.00014463616989778393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1071_text_document +0.00014301572221485565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1072_text_document +0.00014804643324427358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1073_text_document +0.0002143783669071859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1074_text_document +0.0001479303814401362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1075_text_document +0.00015068744684349907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1076_text_document +0.00021658806091136903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1077_text_document +0.00021333945668012075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1078_text_document +0.000142221472149436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1079_text_document +0.0002158096794842747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1080_text_document +0.00021541031163695796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1081_text_document +0.0002160301031804424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1082_text_document +0.00014484879119054217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1083_text_document +0.00014717950537309672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1084_text_document +0.00021016132927298846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1085_text_document +0.00021433713539833563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1086_text_document +0.0001438233936284062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1087_text_document +0.0001447086593934949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1088_text_document +0.00021440017582664183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1089_text_document +0.00020841624205804798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1090_text_document +0.000213227136771408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1091_text_document +0.00020931414236598925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1092_text_document +0.0002134545412666026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1093_text_document +0.0002126803251195216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1094_text_document +0.00014114550507201583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1095_text_document +0.00016444080384922814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1096_text_document +0.0001542515002652382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1097_text_document +0.0001608177523717217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1098_text_document +0.0001577693965006662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1099_text_document +0.0001615213258436368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1100_text_document +0.00014975169893108998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1101_text_document +0.00015902857074290308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1102_text_document +0.00015523901418979132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1103_text_document +0.00015842052994374488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1104_text_document +0.0001543439686424067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1105_text_document +0.0001559141331005536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1106_text_document +0.0001558557495821586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1107_text_document +0.00016108187362389814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1108_text_document +0.0001605357063724452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1109_text_document +0.0001588416921491903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1110_text_document +0.00015452564563384654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1111_text_document +0.0001575925464658241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1112_text_document +0.000155416389913229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1113_text_document +0.00015834897089216795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1114_text_document +0.00015376802717866433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1115_text_document +0.00015257616131444455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1116_text_document +0.00015333466381495513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1117_text_document +0.00015356006723825613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1118_text_document +0.00015392513748333956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1119_text_document +0.00015808193589371923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1120_text_document +0.00015572715307115401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1121_text_document +0.00015677288071421776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1123_text_document +0.00015564703516755468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1124_text_document +0.00015473730933423342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1125_text_document +0.00015227152970932222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1126_text_document +0.00015062363935408713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1127_text_document +0.0001608838990519831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1128_text_document +0.00016058746991656767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1129_text_document +0.00015232158785053588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1130_text_document +0.00015216796930278597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1131_text_document +0.00015531087359959403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1132_text_document +0.00017455174602057423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1133_text_document +0.00015220395996782025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1134_text_document +0.00022536045257736233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1135_text_document +0.00023391977994072452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1136_text_document +0.00022316737354122904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1137_text_document +0.00023097409031198833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1138_text_document +0.0001536444602488289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1139_text_document +0.00015290857223001657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1140_text_document +0.00015053717764782956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1141_text_document +0.0001487906308449292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1142_text_document +0.00022796481136694752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1143_text_document +0.00022388054021300896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1144_text_document +0.00015633876287631285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1146_text_document +0.00015683128496399404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1147_text_document +0.0001498588984354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1148_text_document +0.00015466674094651695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1149_text_document +0.00015104328866230663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1150_text_document +0.0001510288850415886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1151_text_document +0.00015453329995596143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1152_text_document +0.0001717890160140908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1153_text_document +0.00016303689223488152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1154_text_document +0.00017438742884609578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1155_text_document +0.00017195307231868866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1156_text_document +0.00016630614911747752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1157_text_document +0.0001738954845222655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1158_text_document +0.00016759158755171884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1159_text_document +0.00017061259922452842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1160_text_document +0.00017196072417278202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1161_text_document +0.00016824585118656202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1162_text_document +0.00016301309236242047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1163_text_document +0.0001718575393991296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1164_text_document +0.00017003663826341565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1165_text_document +0.00017018328983305946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1166_text_document +0.00017218141861091656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1167_text_document +0.00016559619112054818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1168_text_document +0.00016284882257395627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1169_text_document +0.0001617104078870124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1170_text_document +0.00016849349395228177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1171_text_document +0.00016378319727916067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1172_text_document +0.00017114019486042634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1173_text_document +0.0001726823065513329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1174_text_document +0.00016244897469644304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1175_text_document +0.0001613681046473606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1176_text_document +0.00018118661924575096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1177_text_document +0.00016563345750593493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1178_text_document +0.00016790014898759615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1179_text_document +0.0001629142142864177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1180_text_document +0.00016191717527939525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1181_text_document +0.0001671004065869619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1182_text_document +0.0001675370141650324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1183_text_document +0.00016799445480682778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1184_text_document +0.0001719736620354862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1185_text_document +0.00016261057260474936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1186_text_document +0.00015865991174764644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1187_text_document +0.00015739800441831657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1188_text_document +0.00016171134746282626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1189_text_document +0.00016720238820009615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1190_text_document +0.00016497201020069133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1191_text_document +0.00016081080933342493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1192_text_document +0.0001598451415954535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1193_text_document +0.00016189725587725768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1194_text_document +0.00015376149407875128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1195_text_document +0.00015923032632387212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1196_text_document +0.000161420662154024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1197_text_document +0.00015926844960634996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1198_text_document +0.000156372807999939 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1199_text_document +0.00016050285429044874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1200_text_document +0.00015617925982472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1201_text_document +0.00016514079794945202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1202_text_document +0.00016522274070820443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1203_text_document +0.0001597381170738336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1204_text_document +0.0001616744058690789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1205_text_document +0.00016029435854255644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1206_text_document +0.0001600416279503584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1207_text_document +0.0001607379715998696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1208_text_document +0.0001593514911283079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1209_text_document +0.00015864317782095664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1210_text_document +0.00015911735436385907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1211_text_document +0.0001556275795066712 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1212_text_document +0.0001656764173702947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1213_text_document +0.00015679155524627255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1214_text_document +0.00016376988600479205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1215_text_document +0.0001581538165285075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1216_text_document +0.0001610240227045592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1217_text_document +0.00015776131940645536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1218_text_document +0.00015818231748846595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1219_text_document +0.0001625550897521123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1220_text_document +0.0001547371099180901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1221_text_document +0.00015414283944531357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1222_text_document +0.00016266088273096592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1223_text_document +0.00016083169545961368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1224_text_document +0.0001573027086756309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1225_text_document +0.00015728313997935927 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1226_text_document +0.00016781226249248295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1227_text_document +0.00014976228995207784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1228_text_document +0.00015444629923379175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1229_text_document +0.00015203154472094758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1230_text_document +0.00015416974359531256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1231_text_document +0.00015545110214308707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1232_text_document +0.0001510309557116906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1233_text_document +0.000150151986610048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1234_text_document +0.00014833490597173326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1235_text_document +0.00014730918386476007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1236_text_document +0.00014903663558472915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1237_text_document +0.00014834903218682616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1238_text_document +0.00015322537809196756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1239_text_document +0.0001511230642513134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1240_text_document +0.00015357591909403477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1241_text_document +0.00015295542934724653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1242_text_document +0.00015013958035919124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1243_text_document +0.00015023610122778707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1244_text_document +0.00014784318253583398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1245_text_document +0.00015065966876706016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1246_text_document +0.0001481405433493943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1247_text_document +0.00014721741369089534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1248_text_document +0.00014730057861393202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1249_text_document +0.00015235999841072513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1250_text_document +0.00014541040677624616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1251_text_document +0.00014639042630648248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1252_text_document +0.00015068532335773535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1253_text_document +0.00015516053357170532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1254_text_document +0.00014515004876336832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1255_text_document +0.0001488593805475465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1256_text_document +0.0001506759742452044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1257_text_document +0.0001429840653957083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1258_text_document +0.00014437998012654534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1259_text_document +0.0001428860592717282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1260_text_document +0.0001475220383855572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1261_text_document +0.00014640582972274082 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1262_text_document +0.0001505350968588391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1263_text_document +0.00014784485165882563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1264_text_document +0.00014770697193146622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1265_text_document +0.0001433464625266231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1266_text_document +0.00014139730694769496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1267_text_document +0.00014139435371307747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1268_text_document +0.00014164383589527758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1269_text_document +0.0001429075740030123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1270_text_document +0.00014605872692153072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1271_text_document +0.0001424796215298057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1272_text_document +0.00014112515203848743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1273_text_document +0.00014039188160335826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1274_text_document +0.00014502736267043328 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1275_text_document +0.00014184146815260007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1276_text_document +0.0001453216584479987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1277_text_document +0.00014226985746562565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1278_text_document +0.00013903471234323833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1279_text_document +0.00014633669945119654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1280_text_document +0.00015567823959834718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1281_text_document +0.00016711998145328748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1282_text_document +0.00016716820782888765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1283_text_document +0.00016788189624042867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1284_text_document +0.00016762149528397544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1285_text_document +0.00016394982452183396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1286_text_document +0.00017499487929449305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1287_text_document +0.00017285598246362648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1288_text_document +0.0001813127546456402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1289_text_document +0.00016923644001919636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1290_text_document +0.00016671545149204298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1291_text_document +0.0001691584149978932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1292_text_document +0.00016279240063910965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1293_text_document +0.00016581675179191334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1294_text_document +0.00016709742151486606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1295_text_document +0.00016462921631835026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1296_text_document +0.0001635773235573904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1297_text_document +0.0001629499633321397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1298_text_document +0.00016244603775076793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1299_text_document +0.00016565874682941692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1300_text_document +0.00016704769334813707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1301_text_document +0.00016527793060668047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1302_text_document +0.0001614670182628741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1303_text_document +0.00016090321773766912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1304_text_document +0.00016205158644923216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1305_text_document +0.00016115649647745916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1306_text_document +0.00016750884342636079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1307_text_document +0.0001593023982303325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1308_text_document +0.00015894512446540672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1309_text_document +0.00016391499925658774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1310_text_document +0.0001615310219600013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1311_text_document +0.00016109142610140696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1312_text_document +0.0001622135071747606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1313_text_document +0.00016686311075489617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1314_text_document +0.00016322992039795453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1315_text_document +0.00015923727775344227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1316_text_document +0.00016528070219491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1317_text_document +0.00016089805290891765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1318_text_document +0.00016142731643379644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1319_text_document +0.00016164621217780662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1320_text_document +0.00015738061325748116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1321_text_document +0.0001591233926254462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1322_text_document +0.00016649327648776514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1323_text_document +0.00016299925243783037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1324_text_document +0.00016490993699004063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1325_text_document +0.0001589061309585213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1326_text_document +0.00015701373074415468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1327_text_document +0.00015755460137450403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1328_text_document +0.00016368403834230255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1329_text_document +0.0001619141257919363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1330_text_document +0.0002274793692927606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1331_text_document +0.0001567633247814788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1332_text_document +0.00022905033511751312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1333_text_document +0.0001548301064518758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1334_text_document +0.000226605319945327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1335_text_document +0.00022667037674726058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1336_text_document +0.00022923961805784498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1337_text_document +0.00014906828549341607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1338_text_document +0.00015829222539969273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1339_text_document +0.0001509036911919305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1340_text_document +0.00022536653378252486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1341_text_document +0.00015104016760222197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1342_text_document +0.00015099364342110257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1343_text_document +0.00022777331115603203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1344_text_document +0.00021580582739619934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1345_text_document +0.0001492017484493636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1346_text_document +0.0002232038326367584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1347_text_document +0.0002173110715340058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1348_text_document +0.0002106853410947563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1349_text_document +0.00021523392953900664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1350_text_document +0.00021996424976477582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1351_text_document +0.00021735745725911482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1352_text_document +0.00014743618479981591 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1353_text_document +0.00021587099328468655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1354_text_document +0.00021669175360386172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1355_text_document +0.00021667379282364665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1356_text_document +0.0002192120523189847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1357_text_document +0.00021547193097844086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1358_text_document +0.00021621049112421326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1359_text_document +0.00021196265801039842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1360_text_document +0.00021115416894129982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1361_text_document +0.00021548122875612305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1362_text_document +0.0002167839127379268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1363_text_document +0.00021388435981092266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1364_text_document +0.00021247309275187394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1365_text_document +0.00020865156988970925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1366_text_document +0.00021232420243985875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1367_text_document +0.00020288941772275403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1368_text_document +0.00020534370920083462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1369_text_document +0.00014906807620518648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1370_text_document +0.0002110153701227056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1371_text_document +0.00020709542453451886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1372_text_document +0.00020465988557797482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1373_text_document +0.000195974694790701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1374_text_document +0.0002006410964660873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1375_text_document +0.00020083864604468702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1376_text_document +0.00020640909562295756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1377_text_document +0.0002009390668809768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1378_text_document +0.00019660322090934407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1379_text_document +0.0002031382964736789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1380_text_document +0.00019629671755665872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1381_text_document +0.00019754174238439996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1382_text_document +0.0002056909946356413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1383_text_document +0.0001979138566098626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1384_text_document +0.0001932131948461709 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1385_text_document +0.00020416546879013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1386_text_document +0.00020460391232945065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1387_text_document +0.00019389888059130955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1388_text_document +0.00019783854863351214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1389_text_document +0.000200961415063147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1390_text_document +0.0001956818423121531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1391_text_document +0.00020637040765714317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1392_text_document +0.00020119793791085526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1393_text_document +0.0002019159752232148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1394_text_document +0.00020709690510066213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1395_text_document +0.00019733093804912572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1396_text_document +0.0001880608678579731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1397_text_document +0.0002016375431479316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1398_text_document +0.00019179791527764437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1399_text_document +0.00018506553224762644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1400_text_document +0.00019958850500821938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1401_text_document +0.0001963985599733761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1402_text_document +0.00019686962952391687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1403_text_document +0.00019466431453041557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1404_text_document +0.00019423474723069192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1405_text_document +0.00018645004940802463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1406_text_document +0.0001957563417646353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1407_text_document +0.00019567310057973193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1408_text_document +0.00019820964060443815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1409_text_document +0.0001922448994056278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1410_text_document +0.00018809380854194413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1411_text_document +0.00019183325882742152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1412_text_document +0.00018979529371331087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1413_text_document +0.00018194205843788177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1414_text_document +0.000185326810832552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1415_text_document +0.00018768967790659056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1416_text_document +8.238038512980449e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1417_text_document +7.037628876350043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1418_text_document +7.031761895460266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1419_text_document +6.852561440270574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1420_text_document +7.163053214543125e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1421_text_document +6.965337217248569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1422_text_document +7.217926984135532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1423_text_document +6.887448282655111e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1424_text_document +7.065036798913058e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1425_text_document +7.00112034634854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1426_text_document +7.22545398101735e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1427_text_document +7.06581990215903e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1428_text_document +8.704526082745054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1429_text_document +7.647895905010174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1430_text_document +6.667599117230014e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1431_text_document +9.175439580281598e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1432_text_document +6.936196694178977e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1433_text_document +6.994579003243415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1434_text_document +6.85501978720171e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1435_text_document +6.733846418731063e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1436_text_document +6.760126406073544e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1437_text_document +6.979877393600358e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1438_text_document +6.866399513844505e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1439_text_document +6.599386727589954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1440_text_document +7.022110351565428e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1441_text_document +6.889110495186351e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1442_text_document +7.249533430962498e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1443_text_document +7.061312850517899e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1444_text_document +7.026495137417699e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1445_text_document +7.053710208774785e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1446_text_document +7.079302654666706e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1447_text_document +7.142821385554296e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1448_text_document +6.884074447800683e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1449_text_document +6.775299728680366e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1450_text_document +6.935640081273007e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1451_text_document +7.071164131398859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1452_text_document +7.251697614402021e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1453_text_document +0.00012391766284956256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1454_text_document +6.876051279861284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1455_text_document +7.000563116437178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1456_text_document +7.021430732464126e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1457_text_document +7.320305084935923e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1458_text_document +6.93854906426365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1459_text_document +7.268065730933861e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1460_text_document +0.00015861223006440801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1461_text_document +0.00018316051674097559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1462_text_document +0.00018895109829526356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1463_text_document +0.0001814266629730391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1464_text_document +0.00019157095403912478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1465_text_document +0.0001816102282477865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1466_text_document +0.00018024098024327291 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1467_text_document +0.00018881439877582162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1468_text_document +0.00018219752655961166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1469_text_document +0.00018211134990984607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1470_text_document +0.00018153994220173833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1471_text_document +0.00018033319169939537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1472_text_document +0.00017832883141386175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1473_text_document +0.00018126942359083546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1474_text_document +0.0001780495979719729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1475_text_document +0.00017938729404825616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1476_text_document +0.00017959022144586935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1477_text_document +0.00017639465376427234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1478_text_document +0.00017486395149030808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1479_text_document +0.0001808240528085561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1480_text_document +0.00017274905690967175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1481_text_document +0.00017486036693577152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1482_text_document +0.00017698269026135074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1483_text_document +0.00016841434450034874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1484_text_document +0.00016863721573351308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1485_text_document +0.00017483420640067329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1486_text_document +0.0001732828833128397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1487_text_document +0.00017295290907496933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1488_text_document +0.00017133232046800912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1489_text_document +0.00016977827391836668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1490_text_document +0.0001740935435005184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1491_text_document +3.6341265268857285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1492_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_peS2o.txt b/ALCF/data-lists/sunspot/data_file_list_peS2o.txt new file mode 100644 index 00000000000..dff1249cf5e --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_peS2o.txt @@ -0,0 +1,26 @@ +0.001258392312111664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0000_text_document +0.0012624388132337304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0001_text_document +0.0012626279540316713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0002_text_document +0.0012611745647392154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0003_text_document +0.001261744080471196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0004_text_document +0.0012540395730196387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0005_text_document +0.0012674794155994474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0006_text_document +0.0015181894311854882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0007_text_document +0.003353431842116585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0008_text_document +0.0033457523561418873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0009_text_document +0.0035360737173355393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0010_text_document +0.0035328021064248917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0011_text_document +0.003537853575841124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0012_text_document +0.0033495442948704096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0013_text_document +0.0033515559656802623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0014_text_document +0.0035292441286648877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0015_text_document +0.0033472466636064995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0016_text_document +0.003347244907254542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0017_text_document +0.003361109976122766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0018_text_document +0.003527949940706846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0019_text_document +0.0033629959027952918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0020_text_document +0.003534363177394335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0021_text_document +0.0033534091101340303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0022_text_document +0.003362863367631581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0023_text_document +0.0035295619700253587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0024_text_document +0.00039091085286111746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0025_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_reddit.txt b/ALCF/data-lists/sunspot/data_file_list_reddit.txt new file mode 100644 index 00000000000..644d7170213 --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_reddit.txt @@ -0,0 +1,78 @@ +0.0011541728836721287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0000_text_document +0.001194214065746794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0001_text_document +0.0012074645870644872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0002_text_document +0.0011669676257397446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0003_text_document +0.0011730429598479002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0004_text_document +0.0011829645036126979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0005_text_document +0.0011882713613863669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0006_text_document +0.0011409601969657492 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0007_text_document +0.0011370779956530767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0008_text_document +0.0011566277261230336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0009_text_document +0.0011377530435595722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0010_text_document +0.0011422212106036002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0011_text_document +0.0011359648236479313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0012_text_document +0.0011422819300771266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0013_text_document +0.0011613422088431185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0014_text_document +0.001149222546698594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0015_text_document +0.0011520428345756523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0016_text_document +0.0011408015787470732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0017_text_document +0.001145413257179254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0018_text_document +0.0011543340882314167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0019_text_document +0.0011397083750923865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0020_text_document +0.001163788652940794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0021_text_document +0.0011441686420414542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0022_text_document +0.0011429505546541332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0023_text_document +0.00117471168582067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0024_text_document +0.0011456585273133617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0025_text_document +0.0011738665177335344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0026_text_document +0.0011646176186295262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0027_text_document +0.0011629386473461694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0028_text_document +0.0011421097688385183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0029_text_document +0.0011459477142114253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0030_text_document +0.0011756431096178663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0031_text_document +0.0011482680809577622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0032_text_document +0.0011445710176100962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0033_text_document +0.001142534803152167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0034_text_document +0.0011422043218494292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0035_text_document +0.0011678344410475695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0036_text_document +0.0011562147470581413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0037_text_document +0.0011468122833549663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0038_text_document +0.0011532706690152916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0039_text_document +0.0011292882378850658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0040_text_document +0.0011300177059999066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0041_text_document +0.0011287171558685828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0042_text_document +0.0011295841562723513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0043_text_document +0.0011279954847952854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0044_text_document +0.0011283817109930107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0045_text_document +0.001128286479630481 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0046_text_document +0.0011276081740353844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0047_text_document +0.0011268985652144736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0048_text_document +0.0011261863340342809 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0049_text_document +0.0011248860240274238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0050_text_document +0.0011253794147731645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0051_text_document +0.0011242857628861397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0052_text_document +0.0011228472942657042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0053_text_document +0.00112269047698053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0054_text_document +0.0011234938283922757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0055_text_document +0.0011230927745087202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0056_text_document +0.0011247141749506225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0057_text_document +0.0011241207913742775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0058_text_document +0.0011220187728072355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0059_text_document +0.0011227320045060405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0060_text_document +0.0011217839100677303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0061_text_document +0.0011210875921360617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0062_text_document +0.0011221651716921029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0063_text_document +0.0011248396609954611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0064_text_document +0.0012275703827670792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0065_text_document +0.0011056036331311184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0066_text_document +0.001107902944963784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0067_text_document +0.0010968114497626087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0068_text_document +0.0011027306309299484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0069_text_document +0.0010853624892717291 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0070_text_document +0.0011051858405711837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0071_text_document +0.0010808015771539223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0072_text_document +0.0010855928806935572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0073_text_document +0.0010442141182932184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0074_text_document +0.0011804749731815143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0075_text_document +0.0011670805522744465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0076_text_document +0.0008366052616529944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0077_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_stack.txt b/ALCF/data-lists/sunspot/data_file_list_stack.txt new file mode 100644 index 00000000000..cbaf3cedde1 --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_stack.txt @@ -0,0 +1,149 @@ +0.0010659025986423038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0000_text_document +0.001089820700651703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0001_text_document +0.0010894690468995446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0002_text_document +0.0010893103153582777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0003_text_document +0.001092968830569157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0004_text_document +0.0010927822953669655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0005_text_document +0.0010948538530423937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0006_text_document +0.0010914947459084862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0007_text_document +0.0011531345061061805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0008_text_document +0.0009273732822541429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0009_text_document +0.0009298094568342398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0010_text_document +0.0009269985376241653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0011_text_document +0.0009299414467502114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0012_text_document +0.0009281292496915194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0013_text_document +0.0009300797305068478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0014_text_document +0.0009575658299825903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0015_text_document +0.001124706364232967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0016_text_document +0.0011201757618238954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0017_text_document +0.001126433347327465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0018_text_document +0.0011299837668245817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0019_text_document +0.001127851225271931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0020_text_document +0.0011265589698280143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0021_text_document +0.0011227970380980016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0022_text_document +0.001131300918127052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0023_text_document +0.00112588381546472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0024_text_document +0.0011692456277892793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0025_text_document +0.0011330744556493294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0026_text_document +0.001041946972706877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0027_text_document +0.0010493121881969634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0028_text_document +0.0009912570469629923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0029_text_document +0.0012717963903526445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0030_text_document +0.0014051955824199262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0031_text_document +0.0011248653480876683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0032_text_document +0.0015096975127629315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0033_text_document +0.001056885183600456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0034_text_document +0.0010523010671513575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0035_text_document +0.001055691055690255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0036_text_document +0.0012434898779499373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0037_text_document +0.0009615620261395163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0038_text_document +0.0011689290747945063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0039_text_document +0.0012610288149681123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0040_text_document +0.0012183045747008489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0041_text_document +0.0012232394891956877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0042_text_document +0.0012316862572191265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0043_text_document +0.001171858466558184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0044_text_document +0.0009288715082322405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0045_text_document +0.0009096255640660796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0046_text_document +0.0009098493089021282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0047_text_document +0.000908428701094243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0048_text_document +0.0009115948236386599 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0049_text_document +0.0009109761446993803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0050_text_document +0.0009097199236925156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0051_text_document +0.0009103946801923116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0052_text_document +0.0009109038594994949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0053_text_document +0.0009098133932243314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0054_text_document +0.0009111744494635876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0055_text_document +0.0008961257268851344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0056_text_document +0.0008499219991848833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0057_text_document +0.000848817192629684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0058_text_document +0.0008469931268429987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0059_text_document +0.0008487804660301039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0060_text_document +0.0008535293627452302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0061_text_document +0.0008508082359285502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0062_text_document +0.000847764423021283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0063_text_document +0.0008661814491784624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0064_text_document +0.0012598427266996145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0065_text_document +0.0015411645064455006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0066_text_document +0.0015500690406153115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0067_text_document +0.0010431702414192465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0068_text_document +0.0010103298065465376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0069_text_document +0.0009173697763272889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0070_text_document +0.0009149081716719212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0071_text_document +0.0009223001515794829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0072_text_document +0.0009231205497115238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0073_text_document +0.0009205400022638854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0074_text_document +0.000921891356231865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0075_text_document +0.0009206550523916788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0076_text_document +0.000919101114727538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0077_text_document +0.0009189314293443922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0078_text_document +0.0009187845413397615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0079_text_document +0.0009212488966514148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0080_text_document +0.0009193937503280587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0081_text_document +0.0013803871878583557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0082_text_document +0.0009950213666737198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0083_text_document +0.000927893134699511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0084_text_document +0.0009256115426841411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0085_text_document +0.0009245248815034989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0086_text_document +0.0009239324963431647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0087_text_document +0.00093017264782812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0088_text_document +0.0009246774971430524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0089_text_document +0.0009246651817682976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0090_text_document +0.0009220962135479767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0091_text_document +0.0009218191222144196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0092_text_document +0.0009271314108370893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0093_text_document +0.0011393174361636815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0094_text_document +0.0010056046636817732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0095_text_document +0.000985188940051775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0096_text_document +0.0009834908338499898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0097_text_document +0.0009841221104671695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0098_text_document +0.0009846688252964021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0099_text_document +0.0009846837273836892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0100_text_document +0.000983200779763785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0101_text_document +0.000983626091844726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0102_text_document +0.0009227550215195058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0103_text_document +0.0008517634745985513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0104_text_document +0.0009820984183696825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0105_text_document +0.001062956613371643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0106_text_document +0.0009446580160861343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0107_text_document +0.000849273787178016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0108_text_document +0.0010838798124933814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0109_text_document +0.0016259767652594482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0110_text_document +0.0009261166233974987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0111_text_document +0.0013044836937627727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0112_text_document +0.0017111272224419217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0113_text_document +0.0017274616815008634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0114_text_document +0.0017204942871235126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0115_text_document +0.0017119592701771347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0116_text_document +0.0016979912192342588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0117_text_document +0.001701886248500233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0118_text_document +0.0017227272126357288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0119_text_document +0.0017014517255794117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0120_text_document +0.0016995002579026628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0121_text_document +0.0016958447424626011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0122_text_document +0.0017111887981161064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0123_text_document +0.0017172926007805738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0124_text_document +0.0016938659465618113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0125_text_document +0.0016877576226485259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0126_text_document +0.0017144361080061983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0127_text_document +0.0017173753931755767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0128_text_document +0.001713308056226134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0129_text_document +0.0017126769067653286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0130_text_document +0.0017129095633438736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0131_text_document +0.001704961253905759 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0132_text_document +0.0009282082505873367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0133_text_document +0.0007973220067601047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0134_text_document +0.0008407445714413182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0135_text_document +0.0008403726198530843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0136_text_document +0.0008371632157580058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0137_text_document +0.0013060325919558903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0138_text_document +0.0014100060700040244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0139_text_document +0.0008750222172256031 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0140_text_document +0.0016918433420911735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0141_text_document +0.001838605753011377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0142_text_document +0.0016004536814984726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0143_text_document +0.0011738110086663097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0144_text_document +0.0011269892510041232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0145_text_document +0.0011251329530758676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0146_text_document +0.0011788404279377853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0147_text_document +0.0007876495048700586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0148_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_wiki.txt b/ALCF/data-lists/sunspot/data_file_list_wiki.txt new file mode 100644 index 00000000000..65169566eb0 --- /dev/null +++ b/ALCF/data-lists/sunspot/data_file_list_wiki.txt @@ -0,0 +1,2 @@ +0.0035577638528123345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/en_simple_wiki_v0-0000_text_document +0.0018422361471876658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/en_simple_wiki_v0-0001_text_document From 80dc91c416e08cd72ec3db74b49976e2767adead Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Apr 2024 13:17:25 -0500 Subject: [PATCH 072/590] Add support for `--optimizer ipex.{fused}lamb` --- megatron/optimizer/__init__.py | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 80adeb85210..48f2737a062 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -3,7 +3,7 @@ from deepspeed.accelerator import get_accelerator import torch -from typing import Callable, Any +from typing import Callable, Any, Iterable, Union from megatron import get_args from .distrib_optimizer import DistributedOptimizer @@ -16,7 +16,7 @@ def get_param_groups( - modules: torch.nn.Module | iter[torch.nn.Module], + modules: Union[torch.nn.Module, Iterable[torch.nn.Module]], no_weight_decay_cond: Callable[[str, torch.Tensor], bool], scale_lr_cond: Callable[[str, torch.Tensor], bool], lr_mult: Any, @@ -110,6 +110,7 @@ def get_megatron_optimizer( lr_mult=1.0 ): args = get_args() + assert args is not None # Base optimizer. param_groups = get_param_groups( @@ -218,6 +219,25 @@ def optimizer_hook(p): if p.requires_grad: p.register_post_accumulate_grad_hook(optimizer_hook) layer_wise_flag = True + elif str(args.optimizer) == 'ipex.lamb': + from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, + ) + elif str(args.optimizer) == 'ipex.fusedlamb': + from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, + fused=True, + ) elif str(args.optimizer).lower() == 'ds.fusedlamb': from deepspeed.ops.lamb import FusedLamb optimizer = FusedLamb( From 36af9fc9cf90c0f41cbb60d920302f10ca9f525e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Apr 2024 13:17:36 -0500 Subject: [PATCH 073/590] Add support for `--optimizer ipex.{fused}lamb` --- megatron/arguments.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/megatron/arguments.py b/megatron/arguments.py index 70dfa88ef6e..702a4bad093 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -924,6 +924,8 @@ def _add_training_args(parser): 'adamw', 'sgd', 'ds.fusedlamb', + 'ipex.lamb', + 'ipex.fusedlamb', 'apex.adam', 'apex.sgd', 'adamwschedulefree', From 997c39fb3a80fe684e878fbdd40ab1429f324886 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Apr 2024 16:49:13 -0500 Subject: [PATCH 074/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 151c3e857a3..d124ecb242a 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -144,12 +144,13 @@ setArgs() { } # +---------------------------------------+ -# | 1. Git clone `ezpz` (if not found) | -# | 2. Install `ezpz` (if not installed) | +# | 1. Git clone ezpz (if not found) | +# | 2. Install ezpz (if not installed) | # +---------------------------------------+ ezpz() { - if [[ ! -d ezpz ]]; then - git clone https://github.com/saforem2/ezpz + if [[ ! -d "${PBS_O_WORKDIR}/deps/ezpz" ]]; then + mkdir -p "${PBS_O_WORKDIR}/deps" + git clone https://github.com/saforem2/ezpz "${PBS_O_WORKDIR}/deps" else echo "Found ezpz!" fi @@ -157,12 +158,12 @@ ezpz() { echo "Has ezpz installed. Nothing to do." else echo "Does not have ezpz installed. Installing..." - echo "Using $(which python3) to install \`ezpz\`:" - python3 -m pip install -e ezpz > ezpz-install.log 2>&1 + echo "Using $(which python3) to install ezpz:" + python3 -m pip install -e "${PBS_O_WORKDIR}/edps/ezpz" # > ezpz-install.log 2>&1 fi echo "Done with ezpz." - # source ezpz/src/ezpz/bin/savejobenv || exit # > /tmp/savejobenv.log 2>&1 || exit - # source ezpz/src/ezpz/bin/getjobenv || exit + source ezpz/src/ezpz/bin/savejobenv > /tmp/savejobenv.log 2>&1 || exit + source ezpz/src/ezpz/bin/getjobenv || exit } # +------------------------------------------------------------------------+ @@ -270,9 +271,11 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- if [[ $(hostname) == x4* ]]; then # ---- [AURORA] ---- dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then - dfl_fallback="/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_file_list_reweighted.txt" + dfl_fallback="../ALCF/data-lists/sunspot/data_file_list_books.txt" + # dfl_fallback="/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_file_list_reweighted.txt" elif [[ $(hostname) == x3* ]]; then - dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" + # dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" + dfl_fallback="../ALCF/data-lists/polaris/data_file_list_books.txt" elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" else @@ -280,7 +283,7 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- fi dfl="${1:-${dfl_fallback}}" # dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" - printf "Calling: \`setData()\` with %s\n" "${dfl}" + printf "Calling: setData() with %s\n" "${dfl}" ndocs=$(wc -l < "${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") From e179fe05af7c19e5a9b590be57b0ffad8da38723 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 22 Apr 2024 21:56:57 -0500 Subject: [PATCH 075/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 43 +++++++++++++++++++++++++++++-------------- 1 file changed, 29 insertions(+), 14 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d124ecb242a..216798876b1 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,5 +1,16 @@ #!/bin/bash --login +if [[ -n "${PBS_O_WORKDIR}" ]]; then + WORKING_DIR="${PBS_O_WORKDIR}" +elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then + WORKING_DIR="${SLURM_SUBMIT_DIR}" +else + WORKING_DIR="$(realpath $(pwd))" +fi + +export WORKING_DIR="${WORKING_DIR}" +printf "Using WORKING_DIR: %s" ${WORKING_DIR} + printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" @@ -50,7 +61,7 @@ setParams() { export BE="${CCL}" # BE = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 - export WORKING_DIR="${PBS_O_WORKDIR}" + # export WORKING_DIR="${PBS_O_WORKDIR}" if [[ -z "${NO_FLASH_ATTN}" ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn" fi @@ -62,7 +73,7 @@ setParams() { # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 - export WORKING_DIR="${PBS_O_WORKDIR}" + # export WORKING_DIR="${PBS_O_WORKDIR}" if [[ -z "${NO_FLASH_ATTN}" ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi @@ -73,7 +84,7 @@ setParams() { export BE="${NCCL}" export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-8}" - export WORKING_DIR="${SLURM_SUBMIT_DIR}" + # export WORKING_DIR="${SLURM_SUBMIT_DIR}" if [[ -z "${NO_FLASH_ATTN}" ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi @@ -162,8 +173,8 @@ ezpz() { python3 -m pip install -e "${PBS_O_WORKDIR}/edps/ezpz" # > ezpz-install.log 2>&1 fi echo "Done with ezpz." - source ezpz/src/ezpz/bin/savejobenv > /tmp/savejobenv.log 2>&1 || exit - source ezpz/src/ezpz/bin/getjobenv || exit + source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" > /tmp/savejobenv.log 2>&1 || exit + source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit } # +------------------------------------------------------------------------+ @@ -236,11 +247,15 @@ setEnv() { fi # ----- [Polaris] --------------------------------------- elif [[ $(hostname) == x3* ]]; then - echo "Running on Polaris !!" - # ---- [load conda] --------------------- - module load conda/2023-10-04; conda activate cu118-pt221 ; unset PYTHONUSERBASE - if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then - source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + if [[ "${PBS_O_HOST}" == sirius* ]]; then + export MACHINE="Running on Sirius !!" + else + echo "Running on Polaris !!" + # ---- [load conda] --------------------- + # module load conda/2023-10-04; conda activate cu118-pt221 ; unset PYTHONUSERBASE + if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then + source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + fi fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" @@ -254,8 +269,8 @@ setEnv() { } makeHostfiles() { - source ezpz/src/ezpz/bin/savejobenv || exit #> /tmp/savejobenv.log 2>&1 & - source ezpz/src/ezpz/bin/getjobenv || exit + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" || exit #> /tmp/savejobenv.log 2>&1 & + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" @@ -271,11 +286,11 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- if [[ $(hostname) == x4* ]]; then # ---- [AURORA] ---- dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then - dfl_fallback="../ALCF/data-lists/sunspot/data_file_list_books.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/data_file_list_books.txt" # dfl_fallback="/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_file_list_reweighted.txt" elif [[ $(hostname) == x3* ]]; then # dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" - dfl_fallback="../ALCF/data-lists/polaris/data_file_list_books.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/data_file_list_books.txt" elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" else From 133f24484f302bb8fc439b6bba772eb0dec9653a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Apr 2024 16:07:10 -0500 Subject: [PATCH 076/590] Remove `apex` deps from `megatron/*` --- megatron/model/__init__.py | 12 +++++++++--- megatron/optimizer/clip_grads.py | 10 +++++++--- megatron/optimizer/distrib_optimizer.py | 7 ++++--- megatron/utils.py | 13 ++++++++----- 4 files changed, 28 insertions(+), 14 deletions(-) diff --git a/megatron/model/__init__.py b/megatron/model/__init__.py index 2306749fcb4..141c901ffab 100644 --- a/megatron/model/__init__.py +++ b/megatron/model/__init__.py @@ -1,12 +1,18 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. -from deepspeed.accelerator.real_accelerator import get_accelerator -if get_accelerator().device_name() == 'cuda': +# from deepspeed.accelerator.real_accelerator import get_accelerator +# if get_accelerator().device_name() == 'cuda': +try: from .fused_layer_norm import MixedFusedLayerNorm as LayerNorm from apex.normalization import MixedFusedRMSNorm as RMSNorm -else: + HAS_APEX = True +except Exception: + HAS_APEX = False from .rmsnorm import RMSNorm from torch.nn import LayerNorm +# else: +# from .rmsnorm import RMSNorm +# from torch.nn import LayerNorm from .distributed import DistributedDataParallel from .bert_model import BertModel diff --git a/megatron/optimizer/clip_grads.py b/megatron/optimizer/clip_grads.py index afec8f220c4..b5141d00590 100644 --- a/megatron/optimizer/clip_grads.py +++ b/megatron/optimizer/clip_grads.py @@ -8,10 +8,14 @@ except ModuleNotFoundError: from torch import inf as inf -from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == 'cuda': +# from deepspeed.accelerator import get_accelerator +# if get_accelerator().device_name() == 'cuda': +try: from apex.multi_tensor_apply import multi_tensor_applier import amp_C + HAS_APEX = True +except Exception: + HAS_APEX = False from megatron.model.module import param_is_not_shared from megatron.core.tensor_parallel import param_is_not_tensor_parallel_duplicate @@ -71,7 +75,7 @@ def clip_grad_norm_fp32(parameters, grads_for_norm, else: if norm_type == 2.0: - if get_accelerator().device_name() == 'cuda': + if get_accelerator().device_name() == 'cuda' and HAS_APEX: dummy_overflow_buf = torch.cuda.IntTensor([0]) # Use apex's multi-tensor applier for efficiency reasons. # Multi-tensor applier takes a function and a list of list diff --git a/megatron/optimizer/distrib_optimizer.py b/megatron/optimizer/distrib_optimizer.py index 1aeeac34449..10331607d99 100644 --- a/megatron/optimizer/distrib_optimizer.py +++ b/megatron/optimizer/distrib_optimizer.py @@ -2,10 +2,11 @@ """Megatron distributed optimizer.""" -from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == 'cuda': +# from deepspeed.accelerator import get_accelerator +# if get_accelerator().device_name() == 'cuda': +try: from apex.optimizers import FusedAdam as Adam -else: +except Exception: from torch.optim import Adam import math diff --git a/megatron/utils.py b/megatron/utils.py index 97294070af7..f6a293281ac 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -10,8 +10,12 @@ from deepspeed.accelerator import get_accelerator if get_accelerator().device_name() == 'cuda': - from apex.multi_tensor_apply import multi_tensor_applier - import amp_C + try: + from apex.multi_tensor_apply import multi_tensor_applier + import amp_C + HAS_APEX = True + except Exception: + HAS_APEX = False from megatron import ( get_args, @@ -74,15 +78,14 @@ def calc_params_l2_norm(model): # Calculate norm dummy_overflow_buf = get_accelerator().IntTensor([0]) - if get_accelerator().device_name() == 'cuda': - + if get_accelerator().device_name() == 'cuda' and HAS_APEX: norm, _ = multi_tensor_applier( amp_C.multi_tensor_l2norm, dummy_overflow_buf, [params_data], False # no per-parameter norm ) - else : + else: norm = torch.norm(params_data,p=2.0) norm_2 = norm * norm # Sum across all model-parallel GPUs. From 42a27fb40cd5758b3aa8e03bcf94d0e896e45fce Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Apr 2024 16:09:54 -0500 Subject: [PATCH 077/590] Move `generate_config.sh` logic into `ALCF/helpers.sh` --- ALCF/helpers.sh | 179 +++++++++++++++++++++++++++++++++++++++++++-- generate_config.sh | 172 ------------------------------------------- 2 files changed, 173 insertions(+), 178 deletions(-) delete mode 100644 generate_config.sh diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 216798876b1..18fe965f829 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -95,7 +95,11 @@ setParams() { export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" - export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} + NHOSTS=$(wc -l < "${HOSTFILE}") + NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') + export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" + # export WORLD_SIZE="${WORLD_SIZE:-${NGPUS:-$(( ))}}" + # export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} # ---- Llama2 7B Config ------------------------------ export MODEL_KEY="Llama-7B" export HEADS=${HEADS:-${NHEADS:-32}} @@ -207,11 +211,13 @@ setOutput() { buildDSconfig() { # ---- Build DeepSpeed Config --------------------------------- - export DS_CONFIG="ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" + export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" + mkdir -p $(dirname "${DS_CONFIG}") echo "DS_CONFIG: ${DS_CONFIG}" printf "ZS: %s, CPU_OPTIMIZER: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" - working_dir="${PBS_O_WORKDIR:-${SLURM_SUBMIT_DIR:-$(pwd)}}" - bash "${working_dir}/generate_config.sh" "${DS_CONFIG}" + # working_dir="${PBS_O_WORKDIR:-${SLURM_SUBMIT_DIR:-$(pwd)}}" + generateDSconfig "${DS_CONFIG}" + # bash "${WORKING_DIR}/ALCF/generate_ds_config.sh" "${DS_CONFIG}" # ------------------------------------------------------------- } @@ -302,8 +308,8 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- ndocs=$(wc -l < "${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") - dcp="${HERE}/.cache/${dfl_stem}/index-cache" - mkdir -p dcp + dcp=".cache/${dfl_stem}/index-cache" + # mkdir -p dcp export DATA_FILE_LIST="${dfl}" export NUM_DOCS="${ndocs}" export WEIGHT_SUM="${ws}" @@ -319,6 +325,167 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- echo "--------------------" } +generateDSconfig() { + for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ + "$PP" "$DTYPE" + do + if [ -z $v ]; then + echo "Please export required envs before execute $0" + exit 1 + fi + done + if [ $# -ne 1 ]; then + echo "Usage: $0 config_file" + exit 1 + fi + # \"optimizer\": { + # \"type\": \"AdamW\", + # \"params\": { + # \"lr\": ${LR}, + # \"beta1\": 0.9, + # \"beta2\": 0.95, + # \"eps\": 1e-5, + # \"weight_decay\": 1e-1 + # } + # }, + # \"scheduler\": { + # \"type\": \"WarmupLR\", + # \"params\": { + # \"warmup_min_lr\": 0.00003, + # \"warmup_max_lr\": 0.0003, + # \"warmup_num_steps\": 5000 + # } + # }, + extra="" + common="\ + \"train_batch_size\": $GLOBAL_BATCH, + \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, + \"steps_per_print\": 1, + \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, + \"zero_allow_untested_optimizer\": true, + \"gradient_clipping\": 1.0, + \"activation_checkpointing\": { + \"partition_activations\": true, + \"contiguous_memory_optimization\": false + }, + \"wall_clock_breakdown\": false," + flops_profiler="\ + \"flops_profiler\": { + \"enabled\": true, + \"profile_step\": 4, + \"module_depth\": -1, + \"top_modules\": 1, + \"detailed\": true, + \"output_file\": null + }" + if [[ $DTYPE == "bf16" ]]; then + dtype="\ + \"communication_data_type\": \"bf16\", + \"fp16\": { + \"enabled\": false, + \"loss_scale\": 0, + \"loss_scale_window\": 1000, + \"hysteresis\": 2, + \"min_loss_scale\": 1 + }, + \"bfloat16\": { + \"enabled\": true, + \"loss_scale\": 1.0 + }," + elif [[ $DTYPE == "fp16" ]]; then + dtype="\ + \"communication_data_type\": \"fp16\", + \"fp16\": { + \"enabled\": true, + \"loss_scale\": 0, + \"loss_scale_window\": 1000, + \"hysteresis\": 2, + \"min_loss_scale\": 1 + }, + \"bfloat16\": { + \"enabled\": false, + \"loss_scale\": 1.0 + }," + else + dtype="\"communication_data_type\": \"fp32\"," + fi + if [ $ZERO_STAGE == 3 ]; then + zero="\ + \"zero_optimization\": { + \"stage\": 3, + \"reduce_scatter\": false, + \"mics_shard_size\": 4, + \"mics_hierarchical_params_gather\": true, + \"stage3_max_live_parameters\": 3e9, + \"stage3_max_reuse_distance\": 3e9, + \"stage3_param_persistence_threshold\": 1e5, + \"stage3_prefetch_bucket_size\": 5e7, + \"contiguous_gradients\": true, + \"overlap_comm\": true, + \"reduce_bucket_size\": 90000000, + \"sub_group_size\": 1e9, + \"offload_optimizer\": { + \"device\": \"none\", + \"buffer_count\": 4, + \"pipeline_read\": false, + \"pipeline_write\": false, + \"pin_memory\": true + } + }," + # elif [[ $ZERO_STAGE == 2 ]]; then + elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then + if [[ -n "${CPU_OPTIMIZER}" ]]; then + echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" + zero="\ + \"zero_optimization\": { + \"stage\": $ZERO_STAGE, + \"offload_optimizer\": { + \"device\": \"cpu\" + } + }," + else + zero="\ + \"zero_optimization\": { + \"stage\": $ZERO_STAGE + }," + fi + # elif [[ $ZERO_STAGE == 1 ]]; then + if [[ $PP > 1 ]]; then + extra="\ + \"data_types\": { + \"grad_accum_dtype\": \"fp32\" + }, + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," + else + # echo 'please add the config for zero_stage 1 without pipeline-parallelism' + extra="\ + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," + fi + else + echo 'Please add the correct config set!!!' + fi +# flops_profiler must at the end because no ',' is allowed at the end +cat < $1 +{ +$common +$zero +$dtype +$extra +$flops_profiler +} +EOT +} + printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } diff --git a/generate_config.sh b/generate_config.sh deleted file mode 100644 index b164b5e6107..00000000000 --- a/generate_config.sh +++ /dev/null @@ -1,172 +0,0 @@ -#!/bin/bash --login - -for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ - "$PP" "$DTYPE" -do - if [ -z $v ]; then - echo "Please export required envs before execute $0" - exit 1 - fi -done - -if [ $# -ne 1 ]; then - echo "Usage: $0 config_file" - exit 1 -fi - -# \"optimizer\": { -# \"type\": \"AdamW\", -# \"params\": { -# \"lr\": ${LR}, -# \"beta1\": 0.9, -# \"beta2\": 0.95, -# \"eps\": 1e-5, -# \"weight_decay\": 1e-1 -# } -# }, -# \"scheduler\": { -# \"type\": \"WarmupLR\", -# \"params\": { -# \"warmup_min_lr\": 0.00003, -# \"warmup_max_lr\": 0.0003, -# \"warmup_num_steps\": 5000 -# } -# }, - -extra="" -common="\ - \"train_batch_size\": $GLOBAL_BATCH, - \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, - \"steps_per_print\": 1, - \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, - \"zero_allow_untested_optimizer\": true, - \"gradient_clipping\": 1.0, - \"activation_checkpointing\": { - \"partition_activations\": true, - \"contiguous_memory_optimization\": false - }, - \"wall_clock_breakdown\": false," - -flops_profiler="\ - \"flops_profiler\": { - \"enabled\": true, - \"profile_step\": 4, - \"module_depth\": -1, - \"top_modules\": 1, - \"detailed\": true, - \"output_file\": null - }" - -if [[ $DTYPE == "bf16" ]]; then -dtype="\ - \"communication_data_type\": \"bfp16\", - \"fp16\": { - \"enabled\": false, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": true, - \"loss_scale\": 1.0 - }," -elif [[ $DTYPE == "fp16" ]]; then -dtype="\ - \"communication_data_type\": \"fp16\", - \"fp16\": { - \"enabled\": true, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": false, - \"loss_scale\": 1.0 - }," -else - dtype="\"communication_data_type\": \"fp32\"," -fi - -if [ $ZERO_STAGE == 3 ]; then -zero="\ - \"zero_optimization\": { - \"stage\": 3, - \"reduce_scatter\": false, - \"mics_shard_size\": 4, - \"mics_hierarchical_params_gather\": true, - \"stage3_max_live_parameters\": 3e9, - \"stage3_max_reuse_distance\": 3e9, - \"stage3_param_persistence_threshold\": 1e5, - \"stage3_prefetch_bucket_size\": 5e7, - \"contiguous_gradients\": true, - \"overlap_comm\": true, - \"reduce_bucket_size\": 90000000, - \"sub_group_size\": 1e9, - \"offload_optimizer\": { - \"device\": \"none\", - \"buffer_count\": 4, - \"pipeline_read\": false, - \"pipeline_write\": false, - \"pin_memory\": true - } - }," - -# elif [[ $ZERO_STAGE == 2 ]]; then -elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then - -if [[ -n "${CPU_OPTIMIZER}" ]]; then -echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" - -zero="\ - \"zero_optimization\": { - \"stage\": $ZERO_STAGE, - \"offload_optimizer\": { - \"device\": \"cpu\" - } - }," - -else -zero="\ - \"zero_optimization\": { - \"stage\": $ZERO_STAGE - }," -fi - -# elif [[ $ZERO_STAGE == 1 ]]; then -if [[ $PP > 1 ]]; then - extra="\ - \"data_types\": { - \"grad_accum_dtype\": \"fp32\" - }, - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," -else - # echo 'please add the config for zero_stage 1 without pipeline-parallelism' - extra="\ - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," -fi -else - echo 'Please add the correct config set!!!' -fi - -# flops_profiler must at the end because no ',' is allowed at the end -cat < $1 -{ -$common -$zero -$dtype -$extra -$flops_profiler -} -EOT From 3be7efc120849627404da4607f20103089532064 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Apr 2024 16:25:46 -0500 Subject: [PATCH 078/590] Add option to launch with `mpiexec` --- ALCF/helpers.sh | 2 +- pretrain_gpt_alcf.py | 27 ++++++++++++++++----------- train_llama_alcf.sh | 32 ++++++++++++++++++++++++++++---- 3 files changed, 45 insertions(+), 16 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 18fe965f829..be04c082016 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -9,7 +9,7 @@ else fi export WORKING_DIR="${WORKING_DIR}" -printf "Using WORKING_DIR: %s" ${WORKING_DIR} +printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} printJobInfo() { diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 0139330277b..6530340c19b 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -39,20 +39,19 @@ import ezpz as ez -# ---- SETUP DISTRIBUTED COMMS ---- -# RANK = ez.setup_torch( -# backend='deepspeed', -# port='5432', -# ) -RANK = ez.get_rank() +# ---- [SETUP COMMS] ------------------------ +RANK = ez.setup_torch(backend="deepspeed") +# RANK = ez.get_rank() WORLD_SIZE = ez.get_world_size() +LOCAL_RANK = ez.get_local_rank() DEVICE = ez.get_torch_device() - -# --- TURN OFF LOGGER ON ALL RANK != 0 ---- +if torch.cuda.is_available(): + torch.cuda.set_device(LOCAL_RANK) +# ------------------------------------------- +# --- [TURN OFF LOGGER ON ALL RANK != 0] ---- log = get_logger(__name__) log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") - -# ---- SETUP WANDB FROM RANK 0 ---------------- +# ---- [SETUP WANDB FROM RANK 0] -------------- WANDB_MODE = os.environ.get('WANDB_MODE', None) DISABLE_WANDB = ( WANDB_MODE is not None and str(WANDB_MODE).lower() == 'disabled' @@ -70,7 +69,7 @@ print('--------------------------------------------------') print(f"Setting up W&B from: {RANK} with {project_name}") print('--------------------------------------------------') - ez.setup_wandb(project_name=project_name) + _ = ez.setup_wandb(project_name=project_name) def model_provider(pre_process=True, post_process=True): @@ -163,6 +162,12 @@ def model_provider(pre_process=True, post_process=True): print_rank_0(80 * '-') see_memory_usage("After Building Model", force=True) if wandb.run is not None: + tbdir = args.tensorboard_dir + # tbdir = args.getattr('tensorboard_dir', None) + if tbdir is not None: + log.info(f'Patching tensorboard from {tbdir}') + wandb.tensorboard.patch(root_logdir=tbdir) + wandb.run.config.update({'num_params': num_params}) if "args" not in wandb.run.config: log.info( diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index ce188428504..09b175b1cd2 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -29,7 +29,11 @@ sourceFile "${HERE}/ALCF/helpers.sh" || exit setEnv || exit # 1. load `conda` environment saveDSenv || exit # 2. save env vars to `.deepspeed_env` ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars -makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +if [[ -z "${HOSTFILE}" ]]; then + makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +else + echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" +fi setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} @@ -65,10 +69,30 @@ mkdir -p "${TBDIR}" # --num-workers 0 \ # aprun -n "${NGPUS}" -N "${NGPU_PER_HOST}" --pmi=pmix ${PBS_O_WORKDIR}/local_rank.sh - # ${DIST_LAUNCH} $(which python3) ${EXEC} \ # yeet="${DIST_LAUNCH} ./local_rank.sh" + # deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ + # +# export MPICH_GPU_SUPPORT_ENABLED=1 +# export CUDA_DEVICE_MAX_CONNECTIONS=1 +# export NCCL_DEBUG=INFO + # deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ + # +data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" +mkdir -p "${data_cache_path}" + +if [[ -n "${DIST_LAUNCH}" ]]; then + LAUNCHER="${DIST_LAUNCH} python3 ${EXEC}" +else + LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" +fi + +rstr=$(printRed "Launching with:") +gstr=$(printGreen " ${LAUNCHER}") +printf "%s %s\n" "${rstr}" "${gstr}" + + # ${DIST_LAUNCH} python3 ${EXEC} \ run_cmd=" - deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ + ${LAUNCHER} \ --$DTYPE \ --optimizer ${OPT} \ --split 100,0,0 \ @@ -103,7 +127,7 @@ run_cmd=" --global-batch-size ${GLOBAL_BATCH} \ --pipeline-model-parallel-size ${PP} \ --num-key-value-heads ${NUM_KV_HEAD} \ - --data-cache-path ${DATA_CACHE_PATH} \ + --data-cache-path ${data_cache_path} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ --tokenizer-model ${TOKENIZER_MODEL} \ ${LLAMA_ARGS} \ From a8a9a591d72f1177a78ca3639c1eebca9515eff8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Apr 2024 23:01:56 -0500 Subject: [PATCH 079/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 09b175b1cd2..2c5eff24101 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -20,11 +20,14 @@ function sourceFile() { cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') export HERE + # ----[1. Assert `./pretrain_gpt_alcf.py` exists:]----------------------------- export EXEC="${HERE}/pretrain_gpt_alcf.py" [ -f "${EXEC}" ] || exit + # ----[2. `source ./ALCF/helpers_alcf.sh`:]------------------------------------ sourceFile "${HERE}/ALCF/helpers.sh" || exit + # ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ setEnv || exit # 1. load `conda` environment saveDSenv || exit # 2. save env vars to `.deepspeed_env` @@ -81,16 +84,20 @@ data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" mkdir -p "${data_cache_path}" if [[ -n "${DIST_LAUNCH}" ]]; then - LAUNCHER="${DIST_LAUNCH} python3 ${EXEC}" + LAUNCHER="${DIST_LAUNCH} python3 -Wignore ${EXEC}" else LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" fi -rstr=$(printRed "Launching with:") -gstr=$(printGreen " ${LAUNCHER}") -printf "%s %s\n" "${rstr}" "${gstr}" +rstr=$(printRed "Launching with:\n") +mstr=$(printBlue " ${LAUNCHER}") +printf "%s" "${rstr}'" +printf " %s" "${mstr}" +# printf "%s %s\n" "${rstr}" "${mstr}" # ${DIST_LAUNCH} python3 ${EXEC} \ + # --log-num-zeros-in-grad \ + # --log-memory-to-tensorboard \ run_cmd=" ${LAUNCHER} \ --$DTYPE \ @@ -130,6 +137,9 @@ run_cmd=" --data-cache-path ${data_cache_path} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ --tokenizer-model ${TOKENIZER_MODEL} \ + --timing-log-level ${TIMING_LOG_LEVEL} \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ ${LLAMA_ARGS} \ $ds_args \ ${gpt_args[*]} \ From 42140d7bd0af08ff718db69c10b34c8af8ba039c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Apr 2024 23:02:24 -0500 Subject: [PATCH 080/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 1 + 1 file changed, 1 insertion(+) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index be04c082016..06fadf74f39 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -117,6 +117,7 @@ setParams() { export TRAIN_ITER=${TRAIN_ITER:-317892} export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" export SAVE_INTERVAL=${SAVE_INTERVAL:-200} + export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" From fa0c5a66272a2dbb3295a0540c7c544e9e334640 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 13:06:48 -0500 Subject: [PATCH 081/590] Update `ALCF/helpers.sh`, `train_llama_alcf.sh` --- ALCF/helpers.sh | 57 +++++++++++++++++++++++++++++------------ train_llama_alcf.sh | 62 +++++++++++++++------------------------------ 2 files changed, 61 insertions(+), 58 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 06fadf74f39..8e9e508087f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -36,6 +36,19 @@ function setupSrun() { } +function setupLauncher() { + # outdir=$1 + if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then + export LAUNCH_CMD="${DIST_LAUNCH} python3 -Wignore ${EXEC}" + else + # Assert `./hostfile_deepspeed` exists + export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit + export LAUNCH_CMD="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" + fi + printf "%s" "$(printRed 'Launching with:')" + printf " %s" "$(printMagenta ${LAUNCH_CMD})" +} + function setDSlauncher() { # launcher setting outdir=$1 @@ -159,14 +172,27 @@ setArgs() { export gpt_args } + +function make_ds_hostfile() { + export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" + # ---- Make MPICH hostfile ---------------- + hf="${HOSTFILE:-${PBS_NODEFILE}}" + export hostfile_mpich=hostfile_mpich + cat "${hf}" > "${hostfile_mpich}" + # ---- Make DeepSpeed hostfile ------------------- + export hostfile_deepspeed=hostfile_deepspeed + cat "${hf}" > "${hostfile_deepspeed}" + sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" +} + # +---------------------------------------+ # | 1. Git clone ezpz (if not found) | # | 2. Install ezpz (if not installed) | # +---------------------------------------+ ezpz() { - if [[ ! -d "${PBS_O_WORKDIR}/deps/ezpz" ]]; then - mkdir -p "${PBS_O_WORKDIR}/deps" - git clone https://github.com/saforem2/ezpz "${PBS_O_WORKDIR}/deps" + if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then + mkdir -p "${WORKING_DIR}/deps" + git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps" else echo "Found ezpz!" fi @@ -175,11 +201,12 @@ ezpz() { else echo "Does not have ezpz installed. Installing..." echo "Using $(which python3) to install ezpz:" - python3 -m pip install -e "${PBS_O_WORKDIR}/edps/ezpz" # > ezpz-install.log 2>&1 + python3 -m pip install -e "${WORKING_DIR}/edps/ezpz" # > ezpz-install.log 2>&1 fi echo "Done with ezpz." source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" > /tmp/savejobenv.log 2>&1 || exit source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit + make_ds_hostfile || exit } # +------------------------------------------------------------------------+ @@ -201,7 +228,8 @@ saveDSenv() { setOutput() { # ---- Specify output location -------------------------------- export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" - OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" + # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" + OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" @@ -275,18 +303,15 @@ setEnv() { echo "[python] Using: $(which python3)" } + makeHostfiles() { - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" || exit #> /tmp/savejobenv.log 2>&1 & - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit - export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" - # ---- Make MPICH hostfile ---------------- - hf="${HOSTFILE:-${PBS_NODEFILE}}" - export hostfile_mpich=hostfile_mpich - cat "${hf}" > "${hostfile_mpich}" - # ---- Make DeepSpeed hostfile ------------------- - export hostfile_deepspeed=hostfile_deepspeed - cat "${hf}" > "${hostfile_deepspeed}" - sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" + if [[ -n "${HOSTFILE}" ]]; then + printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" + else + make_ds_hostfile + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" || exit #> /tmp/savejobenv.log 2>&1 & + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit + fi } setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 2c5eff24101..af689ab235b 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -30,29 +30,24 @@ sourceFile "${HERE}/ALCF/helpers.sh" || exit # ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ setEnv || exit # 1. load `conda` environment -saveDSenv || exit # 2. save env vars to `.deepspeed_env` +# saveDSenv || exit # 2. save env vars to `.deepspeed_env` ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars -if [[ -z "${HOSTFILE}" ]]; then - makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -else - echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" -fi + +# if [[ -z "${HOSTFILE}" ]]; then +# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +# else +# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" +# fi setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} setArgs || exit # 8. specify additional `deepspeed` arguments setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset -setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` +# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` printJobInfo || exit # 11. print job info +setupLauncher || exit # ----------------------------------------------------------------------------- -# Take custom args -custom_args=" $@" - -# Assert `./hostfile_deepspeed` exists -export hfds="${HERE}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit -TBDIR="${CKPT_DIR}/tensorboard" -mkdir -p "${TBDIR}" # TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') # printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" @@ -63,44 +58,27 @@ mkdir -p "${TBDIR}" # fi -# source "${HERE}/venvs/polaris/2024-03-14/bin/activate" || exit -# echo "Using $(which python3)" -# --launcher_args='--pmi=pmix' -# deepspeed --hostfile $hfds --launcher ${LAUNCHER} ${EXEC} \ -# ${launch_cmd} \ -# --use-flash-attn-v2 \ -# --num-workers 0 \ - - # aprun -n "${NGPUS}" -N "${NGPU_PER_HOST}" --pmi=pmix ${PBS_O_WORKDIR}/local_rank.sh -# yeet="${DIST_LAUNCH} ./local_rank.sh" - # deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - # # export MPICH_GPU_SUPPORT_ENABLED=1 # export CUDA_DEVICE_MAX_CONNECTIONS=1 # export NCCL_DEBUG=INFO - # deepspeed --hostfile $hfds --launcher MPICH ${EXEC} \ - # +# +# +# Assert TBDIR exists inside our $CKPT_DIR +TBDIR="${CKPT_DIR}/tensorboard" +mkdir -p "${TBDIR}" + data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" mkdir -p "${data_cache_path}" +module list -if [[ -n "${DIST_LAUNCH}" ]]; then - LAUNCHER="${DIST_LAUNCH} python3 -Wignore ${EXEC}" -else - LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" -fi - -rstr=$(printRed "Launching with:\n") -mstr=$(printBlue " ${LAUNCHER}") -printf "%s" "${rstr}'" -printf " %s" "${mstr}" -# printf "%s %s\n" "${rstr}" "${mstr}" +# Take custom args +custom_args=" $@" - # ${DIST_LAUNCH} python3 ${EXEC} \ # --log-num-zeros-in-grad \ # --log-memory-to-tensorboard \ run_cmd=" - ${LAUNCHER} \ - --$DTYPE \ + ${LAUNCH_CMD} \ + --${DTYPE} \ --optimizer ${OPT} \ --split 100,0,0 \ --log-interval 1 \ From 4b9c2f2c966664e0f67a4840b3d88ab6691c76d1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 13:11:22 -0500 Subject: [PATCH 082/590] Add `ALCF/sunspot-env.sh` --- ALCF/sunspot-env.sh | 27 +++++++++++++++++++++++++++ 1 file changed, 27 insertions(+) create mode 100644 ALCF/sunspot-env.sh diff --git a/ALCF/sunspot-env.sh b/ALCF/sunspot-env.sh new file mode 100644 index 00000000000..e97545aafec --- /dev/null +++ b/ALCF/sunspot-env.sh @@ -0,0 +1,27 @@ +#!/bin/bash --login + +export CCL_OP_SYNC=1 # Required by current oneCCL (HPCS-8067) +export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich +export FI_PROVIDER=cxi # Required by Aurora mpich +export PALS_PMI=pmix # Required by Aurora mpich +export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich +export FI_MR_CACHE_MONITOR=disabled # Required by Aurora mpich (HPCS-6501) +export CCL_SKIP_SCHEDULER=1 # Required by current oneCCL, will remove when set by default +export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1 +export CCL_TOPO_COLOR="card:{0,1},{2,3},{4,5},{6,7},{8,9},{10,11};plane:{0,3,4,6,8,11},{1,2,5,7,9,10}" +export UR_L0_IN_ORDER_BARRIER_BY_SIGNAL=0 # Required by current oneCCL + + +export LLM_DK_DIR=/home/$(whoami)/q4-drop_sunspot/llm.devkit + +module load oneapi/release/2023.12.15.001 +unset MPIR_CVAR_CH4_COLL_SELECTION_TUNING_JSON_FILE +unset MPIR_CVAR_COLL_SELECTION_TUNING_JSON_FILE +unset MPIR_CVAR_CH4_POSIX_COLL_SELECTION_TUNING_JSON_FILE + + +module use /home/ftartagl/graphics-compute-runtime/modulefiles +module load graphics-compute-runtime/agama-ci-devel-736.9 +source /home/$(whoami)/q4-drop_sunspot/llm.devkit/torch-ccl/third_party/oneCCL/build/_install/env/vars.sh +module load gcc/12.1.0 +module unload intel_compute_runtime/release/agama-devel-647 From c2e9147546e03cfdb0afe85e6ed8317eecd7ea66 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 13:34:10 -0500 Subject: [PATCH 083/590] Update `train_llama_alcf.sh`, `ALCF/helpers.sh` --- ALCF/helpers.sh | 53 ++++++++++++++++++++++++++++----------------- train_llama_alcf.sh | 5 ++++- 2 files changed, 37 insertions(+), 21 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8e9e508087f..2e34e6c425f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,11 +1,15 @@ #!/bin/bash --login +# +# set -euxo pipefail if [[ -n "${PBS_O_WORKDIR}" ]]; then WORKING_DIR="${PBS_O_WORKDIR}" elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then WORKING_DIR="${SLURM_SUBMIT_DIR}" else - WORKING_DIR="$(realpath $(pwd))" + echo "Unable to detect PBS or SLURM working directory info..." + WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') + echo "Using ${WORKING_DIR} as working directory..." fi export WORKING_DIR="${WORKING_DIR}" @@ -14,7 +18,7 @@ printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" - echo "- MPICH_DIR=$MPICH_DIR" + echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" echo "- Using $(which python3)" echo "- WORLD_SIZE:${WORLD_SIZE}" echo "- NCCL: ${NCCL:-nccl}" @@ -68,6 +72,7 @@ setParams() { LLAMA_ARGS="" # ---- [Parallelism Settings] -------------------------------------------- # -------- [Aurora] ---- || ----- [SunSpot] ------------ + NO_FLASH_ATTN="${NO_FLASH_ATTN:-0}" if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 export CCL=${CCL:-ccl} # CCL @@ -75,7 +80,7 @@ setParams() { export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 # export WORKING_DIR="${PBS_O_WORKDIR}" - if [[ -z "${NO_FLASH_ATTN}" ]]; then + if [[ "${NO_FLASH_ATTN}" != 0 ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn" fi # -------- [Polaris] ----------------------------------- @@ -87,7 +92,8 @@ setParams() { export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 # export WORKING_DIR="${PBS_O_WORKDIR}" - if [[ -z "${NO_FLASH_ATTN}" ]]; then + # if [[ -z "${NO_FLASH_ATTN}" ]]; then + if [[ "${NO_FLASH_ATTN}" != 0 ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi # -------- [Perlmutter] --------------------------------- @@ -98,7 +104,8 @@ setParams() { export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-8}" # export WORKING_DIR="${SLURM_SUBMIT_DIR}" - if [[ -z "${NO_FLASH_ATTN}" ]]; then + # if [[ -z "${NO_FLASH_ATTN}" ]]; then + if [[ "${NO_FLASH_ATTN}" != 0 ]]; then LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi fi @@ -192,20 +199,21 @@ function make_ds_hostfile() { ezpz() { if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" - git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps" + git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" else echo "Found ezpz!" fi + echo "Done with clone. Now, checking if ezpz is installed..." if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then echo "Has ezpz installed. Nothing to do." else echo "Does not have ezpz installed. Installing..." echo "Using $(which python3) to install ezpz:" - python3 -m pip install -e "${WORKING_DIR}/edps/ezpz" # > ezpz-install.log 2>&1 + python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" # > ezpz-install.log 2>&1 fi echo "Done with ezpz." - source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" > /tmp/savejobenv.log 2>&1 || exit - source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit + source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv # > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit + source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv || exit make_ds_hostfile || exit } @@ -240,10 +248,11 @@ setOutput() { buildDSconfig() { # ---- Build DeepSpeed Config --------------------------------- + export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" mkdir -p $(dirname "${DS_CONFIG}") echo "DS_CONFIG: ${DS_CONFIG}" - printf "ZS: %s, CPU_OPTIMIZER: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" + printf "ZS: %s, , MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" # working_dir="${PBS_O_WORKDIR:-${SLURM_SUBMIT_DIR:-$(pwd)}}" generateDSconfig "${DS_CONFIG}" # bash "${WORKING_DIR}/ALCF/generate_ds_config.sh" "${DS_CONFIG}" @@ -269,16 +278,19 @@ sumFiles() { setEnv() { # ---- [SunSpot] ------- || ---- [Aurora] -------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then - PBS_PARENT=$(dirname ${PBS_O_WORKDIR}) - echo "Sourcing ${PBS_PARENT}/setenv.sh..." - source "${PBS_PARENT}/setenv.sh" || exit + # PBS_PARENT=$(dirname ${PBS_O_WORKDIR}) + # echo "Sourcing ${PBS_PARENT}/setenv.sh..." + # source "${PBS_PARENT}/setenv.sh" || exit + source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit # ----- [Aurora] ----------------------------------- - if [[ $(hostname) == x4* ]]; then - eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate anl_release_q4v2 - # ----- [SunSpot] ---------------------------------- - elif [[ $(hostname) == x1* ]]; then - echo "Running on SunSpot !!" - eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + if [[ $(hostname) == x4* ]]; then + eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 + # ----- [SunSpot] ---------------------------------- + elif [[ $(hostname) == x1* ]]; then + echo "Running on SunSpot !!" + eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + fi fi # ----- [Polaris] --------------------------------------- elif [[ $(hostname) == x3* ]]; then @@ -460,7 +472,8 @@ generateDSconfig() { }," # elif [[ $ZERO_STAGE == 2 ]]; then elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then - if [[ -n "${CPU_OPTIMIZER}" ]]; then + # if [[ -n "${CPU_OPTIMIZER}" ]]; then + if [[ "${CPU_OPTIMIZER}" != 0 ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" zero="\ \"zero_optimization\": { diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index af689ab235b..7731df52dd7 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -5,6 +5,8 @@ #PBS -l select=48 #PBS -l filesystems=eagle:home +set -euxo pipefail + function sourceFile() { fp="$1" echo "source-ing ${fp}" @@ -125,7 +127,8 @@ run_cmd=" |& tee ${OUTPUT_LOG} " -echo "! Using $(which deepspeed)" +# ds_exec +# echo "! Using $(which deepspeed)" ds_report echo "${run_cmd}" From 41a3f3523a29a18462d926bbec5cd86c278bfedf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 14:10:04 -0500 Subject: [PATCH 084/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 2e34e6c425f..d0716739af3 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -43,7 +43,7 @@ function setupSrun() { function setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then - export LAUNCH_CMD="${DIST_LAUNCH} python3 -Wignore ${EXEC}" + export LAUNCH_CMD="${DIST_LAUNCH} --cpu-bind depth -d 16 python3 -Wignore ${EXEC}" else # Assert `./hostfile_deepspeed` exists export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit From 71c725ec9aab266d25963fa0d8a37d15c582d7f2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 14:17:00 -0500 Subject: [PATCH 085/590] Much faster check if `ezpz` installed Replace: ```bash if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then ``` with ```bash if python3 -c "import sys; any(['ezpz' in s for s in sys.path])" 2> '/dev/null'; then ``` in `ezpz()` from `ALCF/helpers.sh` --- ALCF/helpers.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d0716739af3..a097d72efcd 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -204,7 +204,8 @@ ezpz() { echo "Found ezpz!" fi echo "Done with clone. Now, checking if ezpz is installed..." - if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then + # if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then + if python3 -c "import sys; any(['ezpz' in s for s in sys.path])" 2> '/dev/null'; then echo "Has ezpz installed. Nothing to do." else echo "Does not have ezpz installed. Installing..." From ae0b4d8b40aec3d43c1842442166829463563d15 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 15:30:50 -0500 Subject: [PATCH 086/590] Add option to run in `DEBUG` mode (i.e. `set -euxo pipefail`) --- ALCF/helpers.sh | 84 ++++++++++++++++++++++++--------------------- train_llama_alcf.sh | 5 ++- 2 files changed, 49 insertions(+), 40 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a097d72efcd..fdd42b56b82 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -28,7 +28,7 @@ printJobInfo() { } -function setupSrun() { +setupSrun() { if [[ $(hostname) == login* || $(hostname) == nid* ]]; then export NHOSTS="${SLURM_NNODES:-1}" export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" @@ -40,7 +40,7 @@ function setupSrun() { } -function setupLauncher() { +setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then export LAUNCH_CMD="${DIST_LAUNCH} --cpu-bind depth -d 16 python3 -Wignore ${EXEC}" @@ -53,7 +53,7 @@ function setupLauncher() { printf " %s" "$(printMagenta ${LAUNCH_CMD})" } -function setDSlauncher() { +setDSlauncher() { # launcher setting outdir=$1 export hfds="$outdir/hostfile_deepspeed" @@ -70,20 +70,23 @@ function setDSlauncher() { setParams() { LLAMA_ARGS="" - # ---- [Parallelism Settings] -------------------------------------------- - # -------- [Aurora] ---- || ----- [SunSpot] ------------ - NO_FLASH_ATTN="${NO_FLASH_ATTN:-0}" + # +----[Parallelism Settings] -------------------------------------------+ + # +------[Aurora]--------||-------[SunSpot]-------------+ if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # BE = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 - # export WORKING_DIR="${PBS_O_WORKDIR}" - if [[ "${NO_FLASH_ATTN}" != 0 ]]; then - LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn" + ####################################################### + # if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" fi - # -------- [Polaris] ----------------------------------- + ####################################################### + # +--------[Polaris]-----------------------------------+ elif [[ $(hostname) == x3* ]]; then TP=${TP:-2} # TP = 2 export NCCL=${NCCL:-nccl} # NCCL @@ -91,61 +94,64 @@ setParams() { # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 - # export WORKING_DIR="${PBS_O_WORKDIR}" - # if [[ -z "${NO_FLASH_ATTN}" ]]; then - if [[ "${NO_FLASH_ATTN}" != 0 ]]; then + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi - # -------- [Perlmutter] --------------------------------- + # +--------[Perlmutter]---------------------------------+ elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then TP="${TP:-2}" export NCCL="${NCCL:-nccl}" export BE="${NCCL}" export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-8}" - # export WORKING_DIR="${SLURM_SUBMIT_DIR}" - # if [[ -z "${NO_FLASH_ATTN}" ]]; then - if [[ "${NO_FLASH_ATTN}" != 0 ]]; then + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi fi - # ------------------------------------------------------------------------ + # +----------------------------------------------------------------------+ export TP="${TP}" export PP="${PP:-1}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") - NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') + if [[ -z "${NGPU_PER_HOST-}" ]]; then + NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') + fi export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" # export WORLD_SIZE="${WORLD_SIZE:-${NGPUS:-$(( ))}}" # export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} - # ---- Llama2 7B Config ------------------------------ + # +---[Llama2 7B Config]-----------------------------+ export MODEL_KEY="Llama-7B" export HEADS=${HEADS:-${NHEADS:-32}} export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} export HIDDEN=${HIDDEN:-4096} export NUM_KV_HEAD=${NUM_KV_HEAD:-8} export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} - # ---- Run Settings ---------------------------------- - export LR=${LR:-0.0003} + # +---[Run Settings]------------------------------------------------------+ + export LR=${LR:-0.0003} # LEARNING_RATE export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export ZERO_STAGE=${ZERO_STAGE:-2} - export MICRO_BATCH=${MICRO_BATCH:-8} - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} - export EVAL_ITERS="${EVAL_ITERS:-10}" - export TRAIN_ITER=${TRAIN_ITER:-317892} - export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" - export SAVE_INTERVAL=${SAVE_INTERVAL:-200} - export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} - export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) - export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" - tm="${WORKING_DIR}/ALCF/tokenizer.model" - export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" - export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" + export ZERO_STAGE=${ZERO_STAGE:-2} # ZERO OFFLOADING STAGE + export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS + export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN + export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS + export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL + export SAVE_INTERVAL=${SAVE_INTERVAL:-200} # HOW FREQUENTLY TO SAVE CKPTS + export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? + export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) # MAX GLOBAL BATCH SIZE + export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT + tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL + # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - # ---------------------------------------------------- + # +----------------------------------------------------------------------+ } @@ -180,7 +186,7 @@ setArgs() { } -function make_ds_hostfile() { +make_ds_hostfile() { export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" @@ -213,7 +219,7 @@ ezpz() { python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" # > ezpz-install.log 2>&1 fi echo "Done with ezpz." - source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv # > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit + source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv || exit make_ds_hostfile || exit } diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 7731df52dd7..4aac1153c7d 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -5,7 +5,10 @@ #PBS -l select=48 #PBS -l filesystems=eagle:home -set -euxo pipefail +if [[ -n "${DEBUG-}" ]]; then + printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" + set -euxo pipefail +fi function sourceFile() { fp="$1" From 2d6608a9260cf36ac844d87efd7a4e73b08398de Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 17:00:23 -0500 Subject: [PATCH 087/590] Update `ALCF/data-lists/sunspot/*.txt` --- ALCF/data-lists/sunspot/algebraic.txt | 16 + ALCF/data-lists/sunspot/arxiv.txt | 100 + ALCF/data-lists/sunspot/books.txt | 3 + ALCF/data-lists/sunspot/c4.txt | 171 + ALCF/data-lists/sunspot/cc.txt | 1108 +++++++ .../sunspot/data_file_list_books.txt | 3 - ALCF/data-lists/sunspot/data_file_list_c4.txt | 86 - ALCF/data-lists/sunspot/data_file_list_cc.txt | 2880 ----------------- .../sunspot/data_file_list_peS2o.txt | 26 - .../sunspot/data_file_list_reddit.txt | 78 - .../sunspot/data_file_list_stack.txt | 149 - .../sunspot/data_file_list_wiki.txt | 2 - ALCF/data-lists/sunspot/falcon.txt | 501 +++ ALCF/data-lists/sunspot/megawiki.txt | 262 ++ .../sunspot/open-web-math-train.txt | 13 + ALCF/data-lists/sunspot/pes2o.txt | 26 + ALCF/data-lists/sunspot/reddit.txt | 78 + ALCF/data-lists/sunspot/stack.txt | 26 + ALCF/data-lists/sunspot/starcoder.txt | 50 + ALCF/data-lists/sunspot/tulu.txt | 66 + ALCF/data-lists/sunspot/wiki.txt | 2 + 21 files changed, 2422 insertions(+), 3224 deletions(-) create mode 100644 ALCF/data-lists/sunspot/algebraic.txt create mode 100644 ALCF/data-lists/sunspot/arxiv.txt create mode 100644 ALCF/data-lists/sunspot/books.txt create mode 100644 ALCF/data-lists/sunspot/c4.txt create mode 100644 ALCF/data-lists/sunspot/cc.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_books.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_c4.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_cc.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_peS2o.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_reddit.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_stack.txt delete mode 100644 ALCF/data-lists/sunspot/data_file_list_wiki.txt create mode 100644 ALCF/data-lists/sunspot/falcon.txt create mode 100644 ALCF/data-lists/sunspot/megawiki.txt create mode 100644 ALCF/data-lists/sunspot/open-web-math-train.txt create mode 100644 ALCF/data-lists/sunspot/pes2o.txt create mode 100644 ALCF/data-lists/sunspot/reddit.txt create mode 100644 ALCF/data-lists/sunspot/stack.txt create mode 100644 ALCF/data-lists/sunspot/starcoder.txt create mode 100644 ALCF/data-lists/sunspot/tulu.txt create mode 100644 ALCF/data-lists/sunspot/wiki.txt diff --git a/ALCF/data-lists/sunspot/algebraic.txt b/ALCF/data-lists/sunspot/algebraic.txt new file mode 100644 index 00000000000..0f25e30498d --- /dev/null +++ b/ALCF/data-lists/sunspot/algebraic.txt @@ -0,0 +1,16 @@ +0.0018520780893211373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document diff --git a/ALCF/data-lists/sunspot/arxiv.txt b/ALCF/data-lists/sunspot/arxiv.txt new file mode 100644 index 00000000000..c50df905037 --- /dev/null +++ b/ALCF/data-lists/sunspot/arxiv.txt @@ -0,0 +1,100 @@ +0.0002583902668716813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document diff --git a/ALCF/data-lists/sunspot/books.txt b/ALCF/data-lists/sunspot/books.txt new file mode 100644 index 00000000000..7aa37a00d2e --- /dev/null +++ b/ALCF/data-lists/sunspot/books.txt @@ -0,0 +1,3 @@ +0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document diff --git a/ALCF/data-lists/sunspot/c4.txt b/ALCF/data-lists/sunspot/c4.txt new file mode 100644 index 00000000000..9504bcbfe65 --- /dev/null +++ b/ALCF/data-lists/sunspot/c4.txt @@ -0,0 +1,171 @@ +0.0002406272620255565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document diff --git a/ALCF/data-lists/sunspot/cc.txt b/ALCF/data-lists/sunspot/cc.txt new file mode 100644 index 00000000000..0a2a0ce35bb --- /dev/null +++ b/ALCF/data-lists/sunspot/cc.txt @@ -0,0 +1,1108 @@ +0.0003742481815405742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document + diff --git a/ALCF/data-lists/sunspot/data_file_list_books.txt b/ALCF/data-lists/sunspot/data_file_list_books.txt deleted file mode 100644 index 9187565a5ed..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_books.txt +++ /dev/null @@ -1,3 +0,0 @@ -0.0031007020167215667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0000_text_document -0.003100207465277759 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0001_text_document -0.000999090518000674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/books-0002_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_c4.txt b/ALCF/data-lists/sunspot/data_file_list_c4.txt deleted file mode 100644 index ca7df1839e9..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_c4.txt +++ /dev/null @@ -1,86 +0,0 @@ -0.0011545953050729803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0000_text_document -0.0011570295715413383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0001_text_document -0.001156438391210766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0002_text_document -0.0011556820995190797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0003_text_document -0.001156780334924253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0004_text_document -0.0011563528368937514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0005_text_document -0.0011574632716369762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0006_text_document -0.0011577445131424494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0007_text_document -0.0011599182963630329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0008_text_document -0.0011550792360663698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0009_text_document -0.001154948574643344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0010_text_document -0.0011560157369398198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0011_text_document -0.0011551344387810997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0012_text_document -0.0011586914190552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0013_text_document -0.00115559584811127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0014_text_document -0.0011562917764239204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0015_text_document -0.0011582019252872318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0016_text_document -0.0011585605528399534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0017_text_document -0.0011567600261132287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0018_text_document -0.0011561323235067436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0019_text_document -0.0011568948157687324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0020_text_document -0.0011562184926986983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0021_text_document -0.001155171968076667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0022_text_document -0.001156245876059478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0023_text_document -0.0011591826911770261 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0024_text_document -0.0011564400126070828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0025_text_document -0.0011571005158517765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0026_text_document -0.0011560050453907214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0027_text_document -0.0011559074476966407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0028_text_document -0.0011567638698290205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0029_text_document -0.0011558972055942165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0030_text_document -0.001157532269673901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0031_text_document -0.0011559883017581377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0032_text_document -0.001155556362078353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0033_text_document -0.0011544735837522018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0034_text_document -0.0011547315955415466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0035_text_document -0.0011570980852521353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0036_text_document -0.0011562552591307868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0037_text_document -0.001156640315842092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0038_text_document -0.0011587257748187634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0039_text_document -0.0011563083526351268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0040_text_document -0.0011554464046007336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0041_text_document -0.001155442922136426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0042_text_document -0.0011557081619451221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0043_text_document -0.001156421357082161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0044_text_document -0.0011562730825316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0045_text_document -0.001157525507046117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0046_text_document -0.0011552936629887162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0047_text_document -0.0011578959437852875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0048_text_document -0.0011568910557636293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0049_text_document -0.0011578444955946039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0050_text_document -0.001157076096248001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0051_text_document -0.0011568459536403974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0052_text_document -0.0011555352450605598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0053_text_document -0.0011557650508322967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0054_text_document -0.0011567625802857914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0055_text_document -0.0011568533734967437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0056_text_document -0.0011562185375437102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0057_text_document -0.0011558740426473278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0058_text_document -0.0011549825990520978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0059_text_document -0.0011572314079774744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0060_text_document -0.0011576031815962752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0061_text_document -0.0011567937670018521 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0062_text_document -0.001154956951193276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0063_text_document -0.001157226898064118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0064_text_document -0.001156096958730414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0065_text_document -0.001155844223704128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0066_text_document -0.0011571187084765205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0067_text_document -0.0011573954893981501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0068_text_document -0.0011566700251641518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0069_text_document -0.0011550051959552815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0070_text_document -0.0011559629359246125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0071_text_document -0.001157971629210032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0072_text_document -0.0011561725903411443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0073_text_document -0.001157160385935682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0074_text_document -0.0011568864860569239 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0075_text_document -0.0011576433208715313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0076_text_document -0.0011571382379808948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0077_text_document -0.0011590178523739284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0078_text_document -0.001156347684201892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0079_text_document -0.0011552550374817486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0080_text_document -0.0011570794132840427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0081_text_document -0.0011570932061148482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0082_text_document -0.0011561938025300182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0083_text_document -0.0011560757016965283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0084_text_document -0.00019284851714729888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/c4-0085_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_cc.txt b/ALCF/data-lists/sunspot/data_file_list_cc.txt deleted file mode 100644 index 4b9a797878b..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_cc.txt +++ /dev/null @@ -1,2880 +0,0 @@ -0.0002329030984435853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0000_text_document -0.00023018699207949078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0001_text_document -0.00024373839803694205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0002_text_document -0.00023608269234913788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0003_text_document -0.00024813091225197464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0004_text_document -0.00023520818074126314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0005_text_document -0.0002374607329273171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0006_text_document -0.00023738412849923294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0007_text_document -0.0002443634316582533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0008_text_document -0.00023847622533166118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0009_text_document -0.00023199871587697545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0010_text_document -0.0002385337709567312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0011_text_document -0.0002432839071745339 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0012_text_document -0.00023508523674007346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0013_text_document -0.00032603226617680567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0014_text_document -0.00023789141182395846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0015_text_document -0.0002461407443245122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0016_text_document -0.00023499257215518966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0017_text_document -0.00024846537508068473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0018_text_document -0.0002386611981191132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0019_text_document -0.0002476214516386151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0020_text_document -0.00023922963334203518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0021_text_document -0.0002566637890877035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0022_text_document -0.0002480836116312675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0023_text_document -0.0002500957846859012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0024_text_document -0.00023232303192858133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0025_text_document -0.0002402109920207785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0026_text_document -0.00032458741378655037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0027_text_document -0.00023711130623699136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0028_text_document -0.0002473092752915358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0029_text_document -0.00024517111812673547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0030_text_document -0.00024145261714879915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0031_text_document -0.0002441832095655324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0032_text_document -0.00024533720808111173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0033_text_document -0.00024615543201451354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0034_text_document -0.00029788578618284437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0035_text_document -0.00026821245945822444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0036_text_document -0.0002451138188102186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0037_text_document -0.00023812823651070536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0038_text_document -0.00023799603175215714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0039_text_document -0.00024128396884325748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0040_text_document -0.00024158008848876737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0041_text_document -0.00024722330373436316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0042_text_document -0.00023308404070500205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0043_text_document -0.0002554252556503107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0044_text_document -0.0003132025339147037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0045_text_document -0.00024278622445373792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0046_text_document -0.0003214585004572529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0047_text_document -0.0003329131703028111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0048_text_document -0.0002361664236831262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0049_text_document -0.0002643368247294079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0050_text_document -0.00024766538637149724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0051_text_document -0.0002627167479901225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0052_text_document -0.00025033496855447236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0053_text_document -0.00024160037266449382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0054_text_document -0.00022926708072112655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0055_text_document -0.00023577632399723273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0056_text_document -0.00024916378421745264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0057_text_document -0.00024065956580145883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0058_text_document -0.00032914757231594763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0059_text_document -0.000382735213415281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0060_text_document -0.00019876415914729903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0061_text_document -0.0002455041228482986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0062_text_document -0.0002360975192355561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0063_text_document -0.00035687225557611647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0064_text_document -0.00034010734287544296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0065_text_document -0.00024289772720050695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0066_text_document -0.0002298464162081398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0067_text_document -0.00032731880189343956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0068_text_document -0.00024593154050122983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0069_text_document -0.00024184757636917526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0070_text_document -0.0002619883069796127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0071_text_document -0.00023707630401459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0072_text_document -0.0003648802259322563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0073_text_document -0.00034821518419266554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0074_text_document -0.00025687739808269634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0075_text_document -0.00025210376457187776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0076_text_document -0.00025341417049958763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0077_text_document -0.00026096750660126574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0078_text_document -0.0002557323323244081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0079_text_document -0.0003306928457892949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0080_text_document -0.00034038835131844906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0081_text_document -0.00025944099107910257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0082_text_document -0.00011523229485833962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0083_text_document -0.0002577986281049885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0084_text_document -0.00011763411767853355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0085_text_document -0.00025348268598695737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0086_text_document -0.00032333206004171266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0087_text_document -0.00030755087408648437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0088_text_document -0.00023006508933660387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0089_text_document -0.00023529378653763827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0090_text_document -0.0002316006671871909 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0091_text_document -0.0002467080329046101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0092_text_document -0.0002812385280195195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0093_text_document -0.0002999655363830447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0094_text_document -0.00030366253916544147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0095_text_document -0.00034483134052353947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0096_text_document -0.0002264669007084511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0097_text_document -0.0002601377797129039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0098_text_document -0.000243683175313779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0099_text_document -0.0002458323373867855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0100_text_document -0.00023061358738763293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0101_text_document -0.0002383240957413279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0102_text_document -0.00024652411741760106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0103_text_document -0.00024356064371899462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0104_text_document -0.00023826916720633669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0105_text_document -0.00023583636824734604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0106_text_document -0.00023310828235332517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0107_text_document -0.00024133699058477928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0108_text_document -0.00023757818755491814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0109_text_document -0.00024650642737935284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0110_text_document -0.00023587507176169633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0111_text_document -0.0002394516652010616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0112_text_document -0.00026115753562452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0113_text_document -0.00023919185015293048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0114_text_document -0.0002328737948830104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0115_text_document -0.0002449581587150213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0116_text_document -0.00023488566807302266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0117_text_document -0.0002461692650286432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0118_text_document -0.00023193321359714746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0119_text_document -0.00024814319189332457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0120_text_document -0.0002502054369100928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0121_text_document -0.0002294119999864264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0122_text_document -0.00023986985689573848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0123_text_document -0.00023333209217509475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0124_text_document -0.0002268247786450586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0125_text_document -0.0002289098412617007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0126_text_document -0.00023635954118858026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0127_text_document -0.00024647215050850076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0128_text_document -0.00024326708810109974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0129_text_document -0.0002931046025004214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0130_text_document -0.00022529330733557138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0131_text_document -0.00024288319647667783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0132_text_document -0.0003170441859608398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0133_text_document -0.00032183678547706126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0134_text_document -0.00020557308761968548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0135_text_document -0.00020890924417592562 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0136_text_document -0.00021111297420597103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0137_text_document -0.00021993650550023244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0138_text_document -0.0002123163519100286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0139_text_document -0.0002103629651549111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0140_text_document -0.00021370932994199264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0141_text_document -0.00020399994203827728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0142_text_document -0.00021563034464531022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0143_text_document -0.0002119386189866467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0144_text_document -0.00020333697838057754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0145_text_document -0.00020812225502998168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0146_text_document -0.0002192034455873437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0147_text_document -0.0002146433860256116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0148_text_document -0.00022498320338620924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0149_text_document -0.00020605974297327904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0150_text_document -0.00020911517614300505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0151_text_document -0.00022086517759478398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0152_text_document -0.00021332423639106333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0153_text_document -0.00020576019154376813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0154_text_document -0.00020504347709097317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0155_text_document -0.00020777754226086552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0156_text_document -0.00021294564928541406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0157_text_document -0.00020775275197134613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0158_text_document -0.00021002644029417448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0159_text_document -0.00021013797882725636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0160_text_document -0.00019076903434985646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0161_text_document -0.00019137766426576477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0162_text_document -0.0001841037351078922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0163_text_document -0.0001952863228508793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0164_text_document -0.00018602295481911772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0165_text_document -0.0001931370361427833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0166_text_document -0.0001801085437374987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0167_text_document -0.000188289716886196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0168_text_document -0.0001852865203803285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0169_text_document -0.00018892492640726607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0170_text_document -0.0001867706345514145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0171_text_document -0.00018688900901065678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0172_text_document -0.00018978617486719294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0173_text_document -0.00019074400515584856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0174_text_document -0.00018895644551080948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0175_text_document -0.0002014139475504348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0176_text_document -0.00019178652165604014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0177_text_document -0.00019538713758341256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0178_text_document -0.00019221603071045457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0179_text_document -0.00018559646736351844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0180_text_document -0.00018839424919962872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0181_text_document -0.00031369302654824313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0182_text_document -0.00029283955302533026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0183_text_document -0.0003003216050130351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0184_text_document -0.00030560088357585723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0185_text_document -0.00030852297965873606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0186_text_document -0.00030137151200383515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0187_text_document -0.000287675564141583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0188_text_document -0.0002865118305148982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0189_text_document -0.0002942394807592494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0190_text_document -0.0002892999122858095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0191_text_document -0.00029726222843843534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0192_text_document -0.0002865106197035132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0193_text_document -0.0002831383377282607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0194_text_document -0.00029911101649033976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0195_text_document -0.0002879193266837814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0196_text_document -0.000293888834619463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0197_text_document -0.00028471984768159116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0198_text_document -0.0002880090219919074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0199_text_document -0.0002916398711835823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0200_text_document -0.00029790830243728387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0201_text_document -0.00028328873748227157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0202_text_document -0.000295084201372288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0203_text_document -0.0002870500420988019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0204_text_document -0.00028061238206088403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0205_text_document -0.00028268741759946835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0206_text_document -0.0002832900433124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0207_text_document -0.0002821269671667503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0208_text_document -0.00028388007298379026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0209_text_document -0.0002811354392519064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0210_text_document -0.0002811576793347316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0211_text_document -0.000291266961761568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0212_text_document -0.0002930917058536775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0213_text_document -0.00029247722771384336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0214_text_document -0.00030253733431717943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0215_text_document -0.0002988938219536017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0216_text_document -0.0003002888817617649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0217_text_document -0.00028686614758997625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0218_text_document -0.00032046548753382687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0219_text_document -0.00027752519729998216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0220_text_document -0.00026529350985605245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0221_text_document -0.0002654493836819182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0222_text_document -0.00026232091015406547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0223_text_document -0.0002599081762104853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0224_text_document -0.0002835817651903514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0225_text_document -0.00026294839748242733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0226_text_document -0.0002610835823452124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0227_text_document -0.000260110886669002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0228_text_document -0.000253371820236557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0229_text_document -0.0002581811396117453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0230_text_document -0.0002514852630632709 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0231_text_document -0.00025726705673313424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0232_text_document -0.00025592912496079053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0233_text_document -0.00025012268192543976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0234_text_document -0.00024391340520007348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0235_text_document -0.0002384383639062725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0236_text_document -0.00023975576001149118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0237_text_document -0.0002338016280970284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0238_text_document -0.0002439200883556984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0239_text_document -0.00024142268942556778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0240_text_document -0.0002427966777591219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0241_text_document -0.00024280144153436732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0242_text_document -0.00024065658615901044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0243_text_document -0.00024455143739741974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0244_text_document -0.00023239795390635735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0245_text_document -0.0002582911684560293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0246_text_document -0.00024625861259252923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0247_text_document -0.0002391576312805854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0248_text_document -0.000238078180343909 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0249_text_document -0.00023486425304981024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0250_text_document -0.0002355893518655022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0251_text_document -0.0002366129403678232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0252_text_document -0.00023595832035066449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0253_text_document -0.00023327574008525872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0254_text_document -0.00024148789011315923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0255_text_document -0.0002373778500991465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0256_text_document -0.00023955987733466374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0257_text_document -0.000230949882722363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0258_text_document -0.00023691636140836262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0259_text_document -0.0002296963977634624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0260_text_document -0.0002332661069034444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0261_text_document -0.00023843042502126992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0262_text_document -0.00023511746712743498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0263_text_document -0.0002347369877896436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0264_text_document -0.0002323753243697275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0265_text_document -0.00026669348300156857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0266_text_document -0.00025799845912273273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0267_text_document -0.00027628560903016796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0268_text_document -0.00026519284616643963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0269_text_document -0.00026441815097637077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0270_text_document -0.0002662131391195505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0271_text_document -0.00027728803868991606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0272_text_document -0.0002769764618252775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0273_text_document -0.00027646939593325287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0274_text_document -0.0002624622460988396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0275_text_document -0.0002597094641937235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0276_text_document -0.00026414993058715923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0277_text_document -0.00027056496256926013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0278_text_document -0.0002594411680362496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0279_text_document -0.00026263805833060905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0280_text_document -0.0002560343870682032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0281_text_document -0.0002624349038750109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0282_text_document -0.00025919416325410714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0283_text_document -0.0002611522977423299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0284_text_document -0.00023679129688303509 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0285_text_document -0.0002424050866477902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0286_text_document -0.00022701047777126036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0287_text_document -0.00023885339653333248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0288_text_document -0.00024106734540671208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0289_text_document -0.0002258801520250309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0290_text_document -0.0003279882524990489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0291_text_document -0.00033565261995537515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0292_text_document -0.0003289323356607256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0293_text_document -0.0003074095430777535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0294_text_document -0.0003207680812935341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0295_text_document -0.00031455349141131964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0296_text_document -0.0003292847953027658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0297_text_document -0.0003336588045388259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0298_text_document -0.00031509118791912046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0299_text_document -0.0003142598967986839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0300_text_document -0.00030783273695855995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0301_text_document -0.0003180584048660508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0302_text_document -0.0003132932087805931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0303_text_document -0.00031883257979717144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0304_text_document -0.00030944547256766847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0305_text_document -0.00030308947812968015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0306_text_document -0.00027546560713402303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0307_text_document -0.0002849896883269672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0308_text_document -0.00028854314233644503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0309_text_document -0.00028915140229591915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0310_text_document -0.00028785031389006415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0311_text_document -0.00029386612956137296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0312_text_document -0.00027190973100817075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0313_text_document -0.00028482862326451903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0314_text_document -0.00028103519882799385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0315_text_document -0.00027510038584601916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0316_text_document -0.00028413351954904745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0317_text_document -0.0002766838847779375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0318_text_document -0.00026734717208098886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0319_text_document -0.0002798212098651715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0320_text_document -0.0002747771651023886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0321_text_document -0.0002653649112010507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0322_text_document -0.0002631895073950362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0323_text_document -0.00027233897055462913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0324_text_document -0.00026295942114759743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0325_text_document -0.00030523368071333024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0326_text_document -0.00022951852300606208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0327_text_document -0.00022441558532523096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0328_text_document -0.00022508048810748277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0329_text_document -0.00021854625167048365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0330_text_document -0.00032578339433634126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0331_text_document -0.0003234065091465547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0332_text_document -0.00031578848940780525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0333_text_document -0.0003211733834987297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0334_text_document -0.00030598592011548813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0335_text_document -0.00030636342203205056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0336_text_document -0.0003057832116313887 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0337_text_document -0.000314036788141844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0338_text_document -0.00030966829419359915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0339_text_document -0.00030590256959722885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0340_text_document -0.0003098044211320355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0341_text_document -0.00031610551467687426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0342_text_document -0.0003181946275637243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0343_text_document -0.00030594263323826957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0344_text_document -0.0003126680759448145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0345_text_document -0.0002992280964722656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0346_text_document -0.00029925238994904177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0347_text_document -0.0003002679127100512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0348_text_document -0.00029525568123898354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0349_text_document -0.0003024653097967333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0350_text_document -0.0002953978348393056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0351_text_document -0.0003002611325611784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0352_text_document -0.0002957202302765213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0353_text_document -0.00029316969879070013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0354_text_document -0.00029927093466316167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0355_text_document -0.00029673566591636904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0356_text_document -0.0002937689672539696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0357_text_document -0.0002973606684406085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0358_text_document -0.0002964111065178358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0359_text_document -0.0003023024169175062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0360_text_document -0.0003023653161749783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0361_text_document -0.0003041586406248139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0362_text_document -0.00029561553630767535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0363_text_document -0.00024185982713467274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0364_text_document -0.00023843085692504566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0365_text_document -0.00024640440430345615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0366_text_document -0.0002514283272863322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0367_text_document -0.0002428429062712565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0368_text_document -0.00023806417358106035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0369_text_document -0.000241345504518809 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0370_text_document -0.00023475737093303525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0371_text_document -0.00024315922889458298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0372_text_document -0.0002509834540572025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0373_text_document -0.00025303820591366467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0374_text_document -0.00023678822937901864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0375_text_document -0.00023171129872234371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0376_text_document -0.00024461347186013167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0377_text_document -0.00023799008209254456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0378_text_document -0.00023090419051131675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0379_text_document -0.0002236725770641727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0380_text_document -0.00023567214707890686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0381_text_document -0.0002262722125540663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0382_text_document -0.00034312492202384507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0383_text_document -0.00021814471912144287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0384_text_document -0.00023259303719099642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0385_text_document -0.00031953022508126173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0386_text_document -0.00023554778297810253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0387_text_document -0.0002460294175476815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0388_text_document -0.0002407153820838108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0389_text_document -0.0002374237316074476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0390_text_document -0.00023980889380119253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0391_text_document -0.0002511495625217406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0392_text_document -0.0002455758117178104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0393_text_document -0.00024203242698955926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0394_text_document -0.00024139601603558614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0395_text_document -0.00024286894291167163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0396_text_document -0.00023208951019510916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0397_text_document -0.0002357404012027918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0398_text_document -0.00023446333528494393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0399_text_document -0.0002366761658977476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0400_text_document -0.0002382598783135322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0401_text_document -0.00023065268726624828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0402_text_document -0.00022821836479753894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0403_text_document -0.00023184541693801962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0404_text_document -0.00023323789396160382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0405_text_document -0.00022765013762883577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0406_text_document -0.00023714308028716352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0407_text_document -0.00028689301916209046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0408_text_document -0.0003409253474017267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0409_text_document -0.0003375051344730567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0410_text_document -0.0003292176313040109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0411_text_document -0.00032955022485317955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0412_text_document -0.0003279397699428092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0413_text_document -0.0003197789907967984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0414_text_document -0.00031901270687106177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0415_text_document -0.000321273794216131 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0416_text_document -0.0003220857325921838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0417_text_document -0.00031002969769902754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0418_text_document -0.00031282247512778876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0419_text_document -0.0003087408247659614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0420_text_document -0.0003000588357430778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0421_text_document -0.0003050525128747414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0422_text_document -0.0003038755807622741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0423_text_document -0.00029692774685276133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0424_text_document -0.0003116160903862434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0425_text_document -0.00031084101832927995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0426_text_document -0.00030708350656830715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0427_text_document -0.00031743538194191725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0428_text_document -0.00031694261996253895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0429_text_document -0.0003146446823405206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0430_text_document -0.00030156651655858596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0431_text_document -0.000303240651608455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0432_text_document -0.00032558453868072364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0433_text_document -0.0002973680179620588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0434_text_document -0.0002971760577119216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0435_text_document -0.0002973002298006474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0436_text_document -0.0002878620791957177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0437_text_document -0.00029632190555443135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0438_text_document -0.0002946733596926658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0439_text_document -0.00029877307004917556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0440_text_document -0.00029551091884749816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0441_text_document -0.0002976670701108049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0442_text_document -0.0002888352867396029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0443_text_document -0.0002866799361024954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0444_text_document -0.0002859222006630905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0445_text_document -0.00028581831052887173 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0446_text_document -0.00028506927387831265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0447_text_document -0.0002803249093757669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0448_text_document -0.0002809203104492272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0449_text_document -0.00028454145587367076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0450_text_document -0.00028584177277598123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0451_text_document -0.00028086934160805217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0452_text_document -0.000270936293938279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0453_text_document -0.00028304258342716634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0454_text_document -0.00028276074943094315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0455_text_document -0.0002602100764561298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0456_text_document -0.00028012504824815937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0457_text_document -0.0002608944608134916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0458_text_document -0.0002845289889094832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0459_text_document -0.0002717532367216808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0460_text_document -0.0002643974553814476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0461_text_document -0.0002758213344366294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0462_text_document -0.0002753861114186629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0463_text_document -0.00031845649723981725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0464_text_document -0.00032153756772406746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0465_text_document -0.0003223378422301534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0466_text_document -0.0002996787108131847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0467_text_document -0.00030486709979224023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0468_text_document -0.00031053773722556385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0469_text_document -0.0003002771838331003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0470_text_document -0.00029794449770130684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0471_text_document -0.0003033670930430196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0472_text_document -0.0002965031647098184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0473_text_document -0.0002837085032811094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0474_text_document -0.0002828420727162801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0475_text_document -0.00028941167269403106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0476_text_document -0.00029157564190928313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0477_text_document -0.00029812762761704826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0478_text_document -0.0002961388642406645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0479_text_document -0.0002838466433847451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0480_text_document -0.0002788779144959817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0481_text_document -0.0003402152386086791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0482_text_document -0.00037332501068667467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0483_text_document -0.0002413675200116708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0484_text_document -0.0003704235275199961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0485_text_document -0.0002379466982220781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0486_text_document -0.00035089333509974934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0487_text_document -0.00023630817154070126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0488_text_document -0.00023857309295728839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0489_text_document -0.0002435822475458576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0490_text_document -0.00023387703405383536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0491_text_document -0.00034319854187343774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0492_text_document -0.0003622737409420836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0493_text_document -0.00023570573166970698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0494_text_document -0.00022641527241191097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0495_text_document -0.00034243292431352653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0496_text_document -0.00024045245535407698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0497_text_document -0.00023676532885361976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0498_text_document -0.00022335363118071338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0499_text_document -0.00023448598925498735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0500_text_document -0.00033737048365832474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0501_text_document -0.0003357751601882351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0502_text_document -0.0003383236392673138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0503_text_document -0.0003397838415177592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0504_text_document -0.00033705937300296186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0505_text_document -0.00033448155827902774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0506_text_document -0.00034576892094196856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0507_text_document -0.00033674871522955814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0508_text_document -0.0003328110361659434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0509_text_document -0.00032432631363958473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0510_text_document -0.00032731656932112217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0511_text_document -0.00032024116066153716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0512_text_document -0.0003040305172335454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0513_text_document -0.00031659687802842567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0514_text_document -0.000303687860573204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0515_text_document -0.0003155611705529593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0516_text_document -0.00030697272991348575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0517_text_document -0.00032874805540012775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0518_text_document -0.0003195460475675836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0519_text_document -0.00029999019685462926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0520_text_document -0.0003031992730055188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0521_text_document -0.0003004957313392662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0522_text_document -0.00029242533089655584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0523_text_document -0.0002940539652538529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0524_text_document -0.0003042748602544184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0525_text_document -0.00029329988520120374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0526_text_document -0.00028533980088048884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0527_text_document -0.0002995523399640371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0528_text_document -0.00024445982369612285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0529_text_document -0.0002341949821161716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0530_text_document -0.0002448827406649086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0531_text_document -0.0002464661023748273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0532_text_document -0.0002458273043503861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0533_text_document -0.000234131092194839 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0534_text_document -0.00023502842288340058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0535_text_document -0.00023472409854696446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0536_text_document -0.0002353934437680525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0537_text_document -0.00023298716740292522 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0538_text_document -0.00023724345571185632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0539_text_document -0.0002463911915031484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0540_text_document -0.00023298903026561056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0541_text_document -0.00022884149754863258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0542_text_document -0.00023103945956545342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0543_text_document -0.0002444088792883614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0544_text_document -0.00022581722858094737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0545_text_document -0.0002370810502668904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0546_text_document -0.00022632319324174496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0547_text_document -0.00023710168144645038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0548_text_document -0.00022964923090952467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0549_text_document -0.00023748320722538985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0550_text_document -0.0003222624777361089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0551_text_document -0.0002939065142920207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0552_text_document -0.0003163669341858318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0553_text_document -0.0002875568128154461 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0554_text_document -0.0002891174847690085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0555_text_document -0.0002845830978145091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0556_text_document -0.0002834617830618547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0557_text_document -0.0002825955578364204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0558_text_document -0.0002770681818983043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0559_text_document -0.00027398693963975244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0560_text_document -0.00026761486776881346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0561_text_document -0.0002709662939745425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0562_text_document -0.0002715205476986883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0563_text_document -0.0002694875173937183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0564_text_document -0.0002691404382855153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0565_text_document -0.0002555696578650904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0566_text_document -0.00025938400199289785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0567_text_document -0.00025330279781755557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0568_text_document -0.00025455190919542185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0569_text_document -0.0002596474980952091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0570_text_document -0.0002593765878092823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0571_text_document -0.00026530976177812846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0572_text_document -0.00026521586959931293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0573_text_document -0.00027156192778243744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0574_text_document -0.00026542489893346987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0575_text_document -0.0002637742757379441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0576_text_document -0.0002660391549513622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0577_text_document -0.0002622961692249776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0578_text_document -0.0002668259130904866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0579_text_document -0.00026393281403990296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0580_text_document -0.0002573087912247817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0581_text_document -0.0002689284845925933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0582_text_document -0.0002587878565641303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0583_text_document -0.0002591277179432351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0584_text_document -0.00025645748667058553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0585_text_document -0.0002576834953920859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0586_text_document -0.0002574007659976351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0587_text_document -0.00026215195926907863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0588_text_document -0.0002550452573299244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0589_text_document -0.0002580549425113166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0590_text_document -0.0002580184320809385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0591_text_document -0.00026135902243793944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0592_text_document -0.0002499110939933153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0593_text_document -0.00023602977130289638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0594_text_document -0.0002179537404034863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0595_text_document -0.000217790844069029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0596_text_document -0.00021511798361299487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0597_text_document -0.00025422459968044684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0598_text_document -0.00026310640293852807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0599_text_document -0.0003408740036680742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0600_text_document -0.00025777786217145044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0601_text_document -0.00025244460970438263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0602_text_document -0.00025351648924446906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0603_text_document -0.0003423231978018855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0604_text_document -0.0003423953052478566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0605_text_document -0.0003318569148201118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0606_text_document -0.00032767638761629247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0607_text_document -0.00033215390937927025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0608_text_document -0.00032618622802635676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0609_text_document -0.00032507622347617733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0610_text_document -0.00031030763419557833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_head-0611_text_document -0.00024643590119480534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0000_text_document -0.0002095902169870633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0001_text_document -0.00021403593088797332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0002_text_document -0.0002227102409085263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0003_text_document -0.00020197706221244385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0004_text_document -0.00022874875522106917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0005_text_document -0.00023280550472601052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0006_text_document -0.00022749578163623905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0007_text_document -0.00023802912323224644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0008_text_document -0.00023176496190267302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0009_text_document -0.0002278986856648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0010_text_document -0.00021833909531790053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0011_text_document -0.0003080057114591217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0012_text_document -0.00021694016663911526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0013_text_document -0.0002668830492707773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0014_text_document -0.00024523658363304193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0015_text_document -0.0002894756615830288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0016_text_document -0.00020347856162111349 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0017_text_document -0.0002134325832786435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0018_text_document -0.00021673235231198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0019_text_document -0.0002654127125833355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0020_text_document -0.0002158672209137081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0021_text_document -0.00023947604851382316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0022_text_document -0.00026152140024106367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0023_text_document -0.00021518621527788343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0024_text_document -0.0002439782139658387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0025_text_document -0.0002905141391659118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0026_text_document -0.00021642682185339982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0027_text_document -0.00019960430947798375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0028_text_document -0.00026322267340937706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0029_text_document -0.00022334429465509248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0030_text_document -0.00022855119280875728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0031_text_document -0.00028578658731994404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0032_text_document -0.0002584277862839571 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0033_text_document -0.00021861958226794765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0034_text_document -0.00026614391185475836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0035_text_document -0.00028970533715167736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0036_text_document -0.0002235814952215254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0037_text_document -0.00022032188312044515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0038_text_document -0.00022884461811511293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0039_text_document -0.0002551680347396578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0040_text_document -0.00022883355545520197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0041_text_document -0.0002232938120141678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0042_text_document -0.0002691617763064546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0043_text_document -0.00023572139842386745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0044_text_document -0.0002552819803341825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0045_text_document -0.00027155660031106415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0046_text_document -0.00021551548292117663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0047_text_document -0.00020620735756494168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0048_text_document -0.0002166820604491231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0049_text_document -0.00018501398539579828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0050_text_document -0.00027225222848112053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0051_text_document -0.00023371832644559636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0052_text_document -0.00023566702124489628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0053_text_document -0.00023686334707090557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0054_text_document -0.00022423975285568458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0055_text_document -0.0002528257228301147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0056_text_document -0.0002561855163693918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0057_text_document -0.00022810786925037496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0058_text_document -0.0002762405538154904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0059_text_document -0.00022261162863844723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0060_text_document -0.00022540915157909426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0061_text_document -0.00022299985657677767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0062_text_document -0.00022755525774778565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0063_text_document -0.00024165856540482104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0064_text_document -0.00025687628451136137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0065_text_document -0.0002231870244226192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0066_text_document -0.00026580529164370396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0067_text_document -0.00028870521089646587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0068_text_document -0.00021540624754582923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0069_text_document -0.00025778332069476944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0070_text_document -0.00021926796929661694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0071_text_document -0.00026029886649394187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0072_text_document -0.00022285796310592967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0073_text_document -0.00023080628286139754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0074_text_document -0.00025245808263416443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0075_text_document -0.00022457772027503216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0076_text_document -0.00024435224362284627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0077_text_document -0.00022526086938759533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0078_text_document -0.0002673487094116284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0079_text_document -0.00023263089713557213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0080_text_document -0.00021778225362633044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0081_text_document -0.00021409630017652816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0082_text_document -0.00022267016739539933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0083_text_document -0.00020585884947224638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0084_text_document -0.00022993683686780696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0085_text_document -0.00024242353683668374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0086_text_document -0.0002092411836993767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0087_text_document -0.0002197488902020793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0088_text_document -0.00021875038642425168 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0089_text_document -0.0002494827261520774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0090_text_document -0.00023601123399284122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0091_text_document -0.00021826172481591926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0092_text_document -0.000236632672200321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0093_text_document -0.00025074570040713444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0094_text_document -0.00022642420961164095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0095_text_document -0.00023812142057551977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0096_text_document -0.0002428821562055837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0097_text_document -0.00022488741946885592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0098_text_document -0.00020317409833506262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0099_text_document -0.00021856439903312987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0100_text_document -0.0002106925714107645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0101_text_document -0.00021119826681040816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0102_text_document -0.0002592340274790045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0103_text_document -0.00023255611509461946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0104_text_document -0.00020894883617804318 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0105_text_document -0.00022615604129768463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0106_text_document -0.000203728797783905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0107_text_document -0.0001989690950208705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0108_text_document -0.00021734356057002846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0109_text_document -0.0002433390106922548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0110_text_document -0.00022031295850762523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0111_text_document -0.00022344289507866802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0112_text_document -0.00022230083290263739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0113_text_document -0.00021439002065826426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0114_text_document -0.0002041951415667326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0115_text_document -0.00022877491032651992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0116_text_document -0.00021999090587860643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0117_text_document -0.00025682432698074305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0118_text_document -0.00024400030399295212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0119_text_document -0.00022789294060424558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0120_text_document -0.00021497724986548528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0121_text_document -0.00023813142494777905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0122_text_document -0.00021895635220322673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0123_text_document -0.00023328497887722523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0124_text_document -0.00022164528342855325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0125_text_document -0.0002484042811809953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0126_text_document -0.00021121568758750245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0127_text_document -0.00020558498767931708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0128_text_document -0.00024543621326022564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0129_text_document -0.00019902438240619879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0130_text_document -0.00023691721805865155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0131_text_document -0.00021791494779355714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0132_text_document -0.0002240264291639859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0133_text_document -0.0002473539109425455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0134_text_document -0.0002071473371471445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0135_text_document -0.00021022258828332134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0136_text_document -0.00022311670653909265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0137_text_document -0.00022930107525031038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0138_text_document -0.0002214421423002716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0139_text_document -0.00021570132519262982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0140_text_document -0.0002197681200389886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0141_text_document -0.0002800029152388595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0142_text_document -0.00026843440765131945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0143_text_document -0.0002849765317975514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0144_text_document -0.00027096319463304773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0145_text_document -0.00027086227426919104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0146_text_document -0.0002526247335698449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0147_text_document -0.00027363240217034764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0148_text_document -0.0002623467059155748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0149_text_document -0.00027346078063921375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0150_text_document -0.00025920642956814055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0151_text_document -0.00025705335691494745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0152_text_document -0.00025922805782841715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0153_text_document -0.0002788336705199961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0154_text_document -0.00024845909125095083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0155_text_document -0.00028656519284339746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0156_text_document -0.00025647131598268287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0157_text_document -0.0002784068234736532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0158_text_document -0.0002528120161786896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0159_text_document -0.0002488190053053583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0160_text_document -0.0002704389893183884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0161_text_document -0.00025616941425622545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0162_text_document -0.00026029019534693783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0163_text_document -0.00025685556571703545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0164_text_document -0.00019723833812640722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0165_text_document -0.0001895418580073486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0166_text_document -0.00019011078486016846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0167_text_document -0.00018779376696334834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0168_text_document -0.00018563641007150188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0169_text_document -0.00018754827458482748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0170_text_document -0.00019755194962803275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0171_text_document -0.00028610572842390993 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0172_text_document -0.00019902354772130188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0173_text_document -0.00020283251106846995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0174_text_document -0.00018722834815639619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0175_text_document -0.00018348325202476222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0176_text_document -0.0002739432916909774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0177_text_document -0.00014534657139819037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0178_text_document -0.00015282753276716084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0179_text_document -0.0001549244865585569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0180_text_document -0.0001465220076427807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0181_text_document -0.00015309131688759006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0182_text_document -0.0001462273984264752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0183_text_document -0.00014903597785697923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0184_text_document -0.0001547302246314982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0185_text_document -0.0001486478323505694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0186_text_document -0.00014887945296702178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0187_text_document -0.00014582128695700495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0188_text_document -0.00015040846513981096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0189_text_document -0.0001492663985213415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0190_text_document -0.0001491503509128408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0191_text_document -0.00014485595166153977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0192_text_document -0.00014471245274265675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0193_text_document -0.0001539836098505113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0194_text_document -0.00014269340600113259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0195_text_document -0.0001366015589763494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0196_text_document -0.00014275967558886846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0197_text_document -0.00012216291308335102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0198_text_document -9.860253447438225e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0199_text_document -0.00013395002197992724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0200_text_document -0.00013095775634161855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0201_text_document -0.00013244501748701574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0202_text_document -0.00013344638268905827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0203_text_document -0.00013599432127141194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0204_text_document -0.0001319495730149868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0205_text_document -0.0001286425479982177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0206_text_document -0.0001288175023456875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0207_text_document -0.00014061678080985136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0208_text_document -0.000128553766351679 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0209_text_document -0.00013865417327932483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0210_text_document -0.00012918889813006947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0211_text_document -0.00013369372633056305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0212_text_document -0.00012818148109232114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0213_text_document -0.00013087168186794624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0214_text_document -0.00012209941459024034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0215_text_document -0.0001170049632015973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0216_text_document -0.00013033065279061172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0217_text_document -0.00012782387759971287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0218_text_document -0.00012594444140907917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0219_text_document -0.00012747350244869554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0220_text_document -0.00011189052700824495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0221_text_document -0.000118474284791765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0222_text_document -0.00012947220948400783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0223_text_document -0.00011563584378100779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0224_text_document -0.00012898102925965738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0225_text_document -0.000122859118523654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0226_text_document -0.00013841949453733798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0227_text_document -0.00012735223374055142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0228_text_document -0.00013005120882648248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0229_text_document -0.000133953509788018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0230_text_document -0.00012898361006981912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0231_text_document -0.00012385687424414202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0232_text_document -0.00012495169231715962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0233_text_document -0.0001334287109141697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0234_text_document -0.0001251557347669207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0235_text_document -0.00012458204389205325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0236_text_document -0.00013142493999218836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0237_text_document -0.0001234876747521603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0238_text_document -0.00011414056156548952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0239_text_document -0.00023536944102421793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0240_text_document -0.00020899836320101376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0241_text_document -0.00020694945512603853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0242_text_document -0.0001985515975806629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0243_text_document -0.00020332234597425947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0244_text_document -0.00019901014809176087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0245_text_document -0.00019730742496077176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0246_text_document -0.0002086531104287768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0247_text_document -0.00019880240459684486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0248_text_document -0.0001934729054969894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0249_text_document -0.00020006177554040137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0250_text_document -0.0001941325758266985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0251_text_document -0.00020329878081065027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0252_text_document -0.00020327608562464652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0253_text_document -0.00019798005487177493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0254_text_document -0.0001954984594242001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0255_text_document -0.0001990223203741723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0256_text_document -0.00019108660381768295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0257_text_document -0.00019716779886134537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0258_text_document -0.0001928475026596504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0259_text_document -0.00019634937526499807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0260_text_document -0.00019298574642019224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0261_text_document -0.00018884134414178089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0262_text_document -0.00018997833083144106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0263_text_document -0.0001905325885044214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0264_text_document -0.00020263821458910917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0265_text_document -0.0002079379871094917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0266_text_document -0.00019785431238092052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0267_text_document -0.00018722610077594935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0268_text_document -0.00019937636744768995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0269_text_document -0.00018558334637361332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0270_text_document -0.00019000469868035166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0271_text_document -0.0001853064471865308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0272_text_document -0.000189466635918149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0273_text_document -0.00019109828052136198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0274_text_document -0.00018290456266579745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0275_text_document -0.00017877060456109023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0276_text_document -0.00018344271945962216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0277_text_document -0.0001937669621232641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0278_text_document -0.00019434311583686195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0279_text_document -0.0001805150932807986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0280_text_document -0.0001914582846585569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0281_text_document -0.00020025771498172507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0282_text_document -0.00019924956568197525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0283_text_document -0.000189496868442045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0284_text_document -0.0001929642820365483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0285_text_document -0.0001903124937955297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0286_text_document -0.00019497565890742164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0287_text_document -0.00018960064504727124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0288_text_document -0.00018568951646616373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0289_text_document -0.00018239686989629257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0290_text_document -0.00018605553146990633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0291_text_document -0.0001844096767388669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0292_text_document -0.00017898307999377337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0293_text_document -0.0001739406120499752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0294_text_document -0.0001911537409150027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0295_text_document -0.00017663348174413226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0296_text_document -0.00017913373123918278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0297_text_document -0.00017455805527093036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0298_text_document -0.00017536417503931625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0299_text_document -0.00017329247651270448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0300_text_document -0.00017912565587258707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0301_text_document -0.00017228776664782256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0302_text_document -0.0001825947205735245 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0303_text_document -0.0001696263054898423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0304_text_document -0.00017175867341643253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0305_text_document -0.0001668734295531042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0306_text_document -0.00016312507834781404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0307_text_document -0.0001687262224636195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0308_text_document -0.00017236097186979052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0309_text_document -0.0002586993024691808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0310_text_document -0.00026219934972577114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0311_text_document -0.0002566784476550503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0312_text_document -0.0002530671575343629 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0313_text_document -0.00025526495987018773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0314_text_document -0.0002510505062545801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0315_text_document -0.00024743741398453804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0316_text_document -0.00024882602559273036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0317_text_document -0.00024230881628338428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0318_text_document -0.00025005854915078414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0319_text_document -0.00024477471955617643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0320_text_document -0.0002480463985551468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0321_text_document -0.00024335328103980772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0322_text_document -0.00024464696562773777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0323_text_document -0.00023820565587951385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0324_text_document -0.00024537554558786237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0325_text_document -0.00024052017934692743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0326_text_document -0.00023660347377746528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0327_text_document -0.00023823292504990384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0328_text_document -0.00023564543049854766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0329_text_document -0.0002370415962271789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0330_text_document -0.00023453319757168757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0331_text_document -0.000236480621339876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0332_text_document -0.0002391149628895737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0333_text_document -0.00023165934662137285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0334_text_document -0.00023331169915961683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0335_text_document -0.0002348226454144718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0336_text_document -0.00023564045570745751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0337_text_document -0.00016411316830860297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0338_text_document -0.0002007359738791159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0339_text_document -0.00019930606930833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0340_text_document -0.00019598670739211644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0341_text_document -0.00019115600211637036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0342_text_document -0.00018957338451495675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0343_text_document -0.0001997256344570198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0344_text_document -0.0001924339501051294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0345_text_document -0.0001929492409258573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0346_text_document -0.00019129356692417672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0347_text_document -0.0001927097658307402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0348_text_document -0.00018744016832935095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0349_text_document -0.00018898826127054628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0350_text_document -0.00019337725386559253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0351_text_document -0.00018434878571055096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0352_text_document -0.00018454731188528818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0353_text_document -0.00018197801455061398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0354_text_document -0.00018615322144032256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0355_text_document -0.00017981075274777777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0356_text_document -0.00018028813451030057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0357_text_document -0.0001760055343765487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0358_text_document -0.00018306121836089844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0359_text_document -0.00018110213343756692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0360_text_document -0.00017839531596627688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0361_text_document -0.00017668405792307465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0362_text_document -0.00018382867977972885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0363_text_document -0.00017812146256462094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0364_text_document -0.00017866992260811773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0365_text_document -0.00017457542446637375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0366_text_document -0.00017144357690622488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0367_text_document -0.00017669299438239817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0368_text_document -0.00017721730286035934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0369_text_document -0.0002573630336497748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0370_text_document -0.00025158500395961657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0371_text_document -0.00025871208953576674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0372_text_document -0.0002522219361597465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0373_text_document -0.00025035546177162626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0374_text_document -0.00024714234522261514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0375_text_document -0.00024296206951019436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0376_text_document -0.00023797488747091152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0377_text_document -0.0002417964809184933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0378_text_document -0.0002317015633644362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0379_text_document -0.00023529081059722227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0380_text_document -0.00022865050303533797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0381_text_document -0.00022350627510674308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0382_text_document -0.00022416076407195612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0383_text_document -0.0002237152481700081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0384_text_document -0.00022673308251184112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0385_text_document -0.00021988509315558021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0386_text_document -0.00021791186375379613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0387_text_document -0.00021902394687174658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0388_text_document -0.00022390913224296354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0389_text_document -0.0002159569838456253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0390_text_document -0.000193074631476835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0391_text_document -0.00019772710141722856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0392_text_document -0.0001918863050023569 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0393_text_document -0.0001968641761834432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0394_text_document -0.00019269495646727515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0395_text_document -0.0001986463032193898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0396_text_document -0.0001855871337995234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0397_text_document -0.00019041152711008963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0398_text_document -0.00018277849340888642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0399_text_document -0.00018810546599505484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0400_text_document -0.00018711834399232793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0401_text_document -0.000180125082690484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0402_text_document -0.00023744084906469025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0403_text_document -0.00023803845013258319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0404_text_document -0.00023586547263857976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0405_text_document -0.00023222402329423718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0406_text_document -0.00023270999204422837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0407_text_document -0.00023378783679246331 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0408_text_document -0.00017304047941651873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0409_text_document -0.00017585076104150352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0410_text_document -0.00017101296884180275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0411_text_document -0.00017561096140154923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0412_text_document -0.0001713420333669203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0413_text_document -0.00022603582939637927 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0414_text_document -0.0001703733924033566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0415_text_document -0.0002396801442728503 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0416_text_document -0.0001676400523382032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0417_text_document -0.00017193438273170229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0418_text_document -0.00017529021040710947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0419_text_document -0.0001630741415909194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0420_text_document -0.00024179471702347313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0421_text_document -0.00016581358754145113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0422_text_document -0.0002456894490564403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0423_text_document -0.0002456073517995372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0424_text_document -0.00024937580109172706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0425_text_document -0.0002457208726475487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0426_text_document -0.00024399607429757567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0427_text_document -0.00023977003702270238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0428_text_document -0.0002453131498067917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0429_text_document -0.0001621090466807557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0430_text_document -0.00024557101413066944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0431_text_document -0.00024662307150866836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0432_text_document -0.00015758980646827074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0433_text_document -0.00024391288666874046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0434_text_document -0.00023509503922816786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0435_text_document -0.00023489539281843744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0436_text_document -0.00023286637378893443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0437_text_document -0.00023379369093964089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0438_text_document -0.00023205784424428202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0439_text_document -0.00023009948269807432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0440_text_document -0.00023187584394201576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0441_text_document -0.00023202252759594008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0442_text_document -0.00022728777233539934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0443_text_document -0.00022582666382743133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0444_text_document -0.00022616733175598707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0445_text_document -0.00022768677294110565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0446_text_document -0.00022367789565066836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0447_text_document -0.00022752055218158585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0448_text_document -0.00021819243338256605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0449_text_document -0.0002241455531613807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0450_text_document -0.00022437797440403226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0451_text_document -0.00022445007197791702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0452_text_document -0.00022150502971124016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0453_text_document -0.0002225145672731263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0454_text_document -0.00022368982014371355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0455_text_document -0.00022402755606263736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0456_text_document -0.00023016090138940315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0457_text_document -0.0002260342841680707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0458_text_document -0.00022458279279977673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0459_text_document -0.00021839974448010203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0460_text_document -0.0002264409368746725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0461_text_document -0.000223550215762877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0462_text_document -0.00021610601829010048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0463_text_document -0.00022408120517524368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0464_text_document -0.00021671066876802013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0465_text_document -0.00016072298972169641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0466_text_document -0.0001722371396276357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0467_text_document -0.00017303760343097654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0468_text_document -0.00016820735177759604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0469_text_document -0.0001782239553050235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0470_text_document -0.0001749477598265696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0471_text_document -0.0001700037698924768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0472_text_document -0.0001721297434219665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0473_text_document -0.00017082606704868714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0474_text_document -0.00017400024710211123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0475_text_document -0.00017016210162102983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0476_text_document -0.00016745166973214216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0477_text_document -0.0001684428163376526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0478_text_document -0.0001648685852885396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0479_text_document -0.00017387645508870812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0480_text_document -0.00016594906144137858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0481_text_document -0.00016042654972698604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0482_text_document -0.00014860104507835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0483_text_document -0.00016227281398002708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0484_text_document -0.00016502091577582913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0485_text_document -0.00016106235650927743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0486_text_document -0.00015987309712264371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0487_text_document -0.0001642815421701454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0488_text_document -0.00016531915249024665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0489_text_document -0.00015833872193897052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0490_text_document -0.00015639158495488916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0491_text_document -0.00015342548972376501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0492_text_document -0.00015518921543764528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0493_text_document -0.0001621958240469728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0494_text_document -0.00015155749799598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0495_text_document -0.00014939896262383117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0496_text_document -0.00015490092509698006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0497_text_document -0.00017977881778259884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0498_text_document -0.00018596378104021417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0499_text_document -0.00017898738743182946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0500_text_document -0.00018286541046512472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0501_text_document -0.00018092409134830376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0502_text_document -0.00017788220095337013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0503_text_document -0.00017903547090898037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0504_text_document -0.0001797342122414524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0505_text_document -0.00018405110997743763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0506_text_document -0.00016587458814992502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0507_text_document -0.00018323507493237133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0508_text_document -0.00017881236669457928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0509_text_document -0.00017083385044833047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0510_text_document -0.0001730201559992492 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0511_text_document -0.00016901593018907565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0512_text_document -0.00017121838351155997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0513_text_document -0.0001762157419442059 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0514_text_document -0.00017000047903250774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0515_text_document -0.00017628842147757824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0516_text_document -0.0001760014416563697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0517_text_document -0.00017080626611158523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0518_text_document -0.00017077485831581488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0519_text_document -0.0001740210774510124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0520_text_document -0.00017310752988628116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0521_text_document -0.00016563538206915967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0522_text_document -0.0001698038028867437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0523_text_document -0.00022989652913943246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0524_text_document -0.00023802118237282655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0525_text_document -0.00023209291976691602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0526_text_document -0.00023478978296678473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0527_text_document -0.00023185674392304132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0528_text_document -0.0002223151271899996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0529_text_document -0.0002212980337800594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0530_text_document -0.0002177142043482363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0531_text_document -0.00022071160791386127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0532_text_document -0.0002155092901614389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0533_text_document -0.00021709946336410436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0534_text_document -0.0002091085371649664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0535_text_document -0.00021301299764538067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0536_text_document -0.00020514046046681228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0537_text_document -0.00020554350961511138 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0538_text_document -0.0002032929572669402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0539_text_document -0.00020017696773262392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0540_text_document -0.0002041760983122544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0541_text_document -0.00019610775249750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0542_text_document -0.0001972797535028649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0543_text_document -0.00019987201182946655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0544_text_document -0.00023221090921479249 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0545_text_document -0.00022866265656078542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0546_text_document -0.00022846213721182363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0547_text_document -0.00022028779604045222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0548_text_document -0.00023019534411130514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0549_text_document -0.00021499063838892918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0550_text_document -0.0002238747556640398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0551_text_document -0.000219139079337847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0552_text_document -0.00022466810662919942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0553_text_document -0.00021354111452743537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0554_text_document -0.0002116352569318229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0555_text_document -0.00021742490236552721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0556_text_document -0.00020976053145397075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0557_text_document -0.0002121893598598504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0558_text_document -0.00020611700008662688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0559_text_document -0.00020771394257887023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0560_text_document -0.00020861778045311834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0561_text_document -0.00020549717473124685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0562_text_document -0.00021168253336591858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0563_text_document -0.00020292362079976103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0564_text_document -0.0002053579978117472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0565_text_document -0.0002025742316233632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0566_text_document -0.00019721191770863706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0567_text_document -0.00020263891920926902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0568_text_document -0.0002047513235561355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0569_text_document -0.0002058192920224309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0570_text_document -0.00020762611235464895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0571_text_document -0.00020536767369033477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0572_text_document -0.000208726602681654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0573_text_document -0.00020670689006790867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0574_text_document -0.0001987029852837105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0575_text_document -0.00019743671572624558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0576_text_document -0.00020347237873346202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0577_text_document -0.00019483561225711876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0578_text_document -0.00019876706376189147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0579_text_document -0.00019418407035646924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0580_text_document -0.00019094739234588127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0581_text_document -0.00018896169178427298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0582_text_document -0.00019336957140803166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0583_text_document -0.00019246034436187084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0584_text_document -0.00019234601030075014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0585_text_document -0.00018937638801999214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0586_text_document -0.00019243149393005724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0587_text_document -0.00018564518487541217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0588_text_document -0.00018349694905090308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0589_text_document -0.00018632405912780405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0590_text_document -0.0001859374743982387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0591_text_document -0.00018735943662878573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0592_text_document -0.00018429223346416512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0593_text_document -0.00018743951405683122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0594_text_document -0.0002231790070545305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0595_text_document -0.00023691491440731282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0596_text_document -0.00022732583835977663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0597_text_document -0.00023280690754947414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0598_text_document -0.00023098339919576762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0599_text_document -0.00022742109041848038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0600_text_document -0.00023387941495424947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0601_text_document -0.00022226509841824269 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0602_text_document -0.00022342786655488707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0603_text_document -0.00022237713376406775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0604_text_document -0.00021379459835981835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0605_text_document -0.00021934823034546768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0606_text_document -0.00022299117012803982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0607_text_document -0.0002249652818475372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0608_text_document -0.00021549803647665793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0609_text_document -0.00021082391557018925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0610_text_document -0.0002063290532408184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0611_text_document -0.0002098859538424268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0612_text_document -0.00020927123951292785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0613_text_document -0.00020988114416198002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0614_text_document -0.00020708947339409333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0615_text_document -0.00020681735599881374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0616_text_document -0.00020862989695824213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0617_text_document -0.00017921617624032585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0618_text_document -0.0001869630178204498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0619_text_document -0.0001837421970952879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0620_text_document -0.0002703540624747488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0621_text_document -0.0002773524903329593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0622_text_document -0.00026751943505093036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0623_text_document -0.00026849089128670544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0624_text_document -0.00017768273890485142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0625_text_document -0.00026694167218726514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0626_text_document -0.00026851367038544275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0627_text_document -0.00017178448275206052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0628_text_document -0.00026146356857229295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0629_text_document -0.0002631494175929668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0630_text_document -0.00026756049947472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0631_text_document -0.0002600735435281443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0632_text_document -0.00026162102069795645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0633_text_document -0.0002546230805208093 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0634_text_document -0.00025384118907342997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0635_text_document -0.00024898898905737453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0636_text_document -0.0002560021645785107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0637_text_document -0.00025001876340897294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0638_text_document -0.00024817567624010623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0639_text_document -0.00025419118513633326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0640_text_document -0.00025520008446783997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0641_text_document -0.00024355226527934937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0642_text_document -0.00024233948860872504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0643_text_document -0.00024413553528635867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0644_text_document -0.00024287456234999737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0645_text_document -0.0002471744870080021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0646_text_document -0.00024318841473052868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0647_text_document -0.00024268080340573577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0648_text_document -0.000242363177173413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0649_text_document -0.00025310826613573865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0650_text_document -0.0002450433802404371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0651_text_document -0.0002429196089265994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0652_text_document -0.00023818874203405117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0653_text_document -0.00023814010078402416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0654_text_document -0.0002258262625271231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0655_text_document -0.0002359106231188901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0656_text_document -0.00023984369117779496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0657_text_document -0.00022677878582898447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0658_text_document -0.00023019334994987196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0659_text_document -0.0002326106169086802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0660_text_document -0.00023296218608853588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0661_text_document -0.00021930251468821644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0662_text_document -0.00022685746290158792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0663_text_document -0.00022204375118840136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0664_text_document -0.00022312982876300855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0665_text_document -0.00022347955655196657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0666_text_document -0.00021968416238742178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0667_text_document -0.00022148339454050315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0668_text_document -0.00022133417129237745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0669_text_document -0.00021840518548046784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0670_text_document -0.00021501258675160414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0671_text_document -0.00016302293581967305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0672_text_document -0.00018778970953587786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0673_text_document -0.00018910837173616491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0674_text_document -0.00019201324078164315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0675_text_document -0.00018432054093123207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0676_text_document -0.0001907749590824511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0677_text_document -0.00017971831966331778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0678_text_document -0.00018425384289495064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0679_text_document -0.00018353447605936826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0680_text_document -0.00018776194922919426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0681_text_document -0.000181858547251418 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0682_text_document -0.00017663862855632625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0683_text_document -0.00017879513620194847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0684_text_document -0.00017779569087388971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0685_text_document -0.00017180299093946108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0686_text_document -0.00018535528192944938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0687_text_document -0.0001710147287879884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0688_text_document -0.000173577199328182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0689_text_document -0.0001768576763304655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0690_text_document -0.0001796376911260544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0691_text_document -0.00016782824293218567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0692_text_document -0.00016074324428116396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0693_text_document -0.0001687513348299545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0694_text_document -0.00016077518171436444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0695_text_document -0.00017132313128327624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0696_text_document -0.00016190186959679132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0697_text_document -0.00016290842504820753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0698_text_document -0.00016156811558387776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0699_text_document -0.00023555665280084346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0700_text_document -0.0002284718177796522 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0701_text_document -0.00022374123273516798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0702_text_document -0.00021994541999416394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0703_text_document -0.00022338573100973358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0704_text_document -0.00022049767881647008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0705_text_document -0.00022416439897413284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0706_text_document -0.00021895546198784436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0707_text_document -0.0002142388294097341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0708_text_document -0.00022017166748084383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0709_text_document -0.00021104350754308596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0710_text_document -0.00020985391201191366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0711_text_document -0.00021778183924550787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0712_text_document -0.00021271266854227129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0713_text_document -0.0002086433619903549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0714_text_document -0.00021568150697937684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0715_text_document -0.00020764802098217656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0716_text_document -0.0002151205404833473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0717_text_document -0.00020430590834946775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0718_text_document -0.00020866543326050432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0719_text_document -0.00020818319961436583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0720_text_document -0.00020070798626764516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0721_text_document -0.00019693995826673832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0722_text_document -0.00020030234076064843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0723_text_document -0.00019788654054706263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0724_text_document -0.0001993257554824347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0725_text_document -0.00021386012383904914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0726_text_document -0.00021978412787373083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0727_text_document -0.0002175599344895926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0728_text_document -0.00021091594587352813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0729_text_document -0.0002034137316303627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0730_text_document -0.00021253423082914959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0731_text_document -0.00020816999471172712 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0732_text_document -0.00021853522405647908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0733_text_document -0.0002051944662085363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0734_text_document -0.00020978726975291983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0735_text_document -0.00020468921406556763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0736_text_document -0.0002007846124143192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0737_text_document -0.00020366090300396152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0738_text_document -0.0001993156168498017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0739_text_document -0.00020150340666889603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0740_text_document -0.00020188286325854645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0741_text_document -0.00020072267667247027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0742_text_document -0.00019591912629771525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0743_text_document -0.00020056463740447396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0744_text_document -0.0001962511050627094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0745_text_document -0.00018969020412060633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0746_text_document -0.00018711981666080213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0747_text_document -0.00019064086480658448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0748_text_document -0.0001893430509717561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0749_text_document -0.00018823938035214858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0750_text_document -0.000191049243153872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0751_text_document -0.00015215085209234548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0752_text_document -0.00013881666461144156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0753_text_document -0.0001511979467407442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0754_text_document -0.00015091819106548992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0755_text_document -0.00013896830454629422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0756_text_document -0.00014286084497610213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0757_text_document -0.00013829013170563417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0758_text_document -0.00014842506748913496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0759_text_document -0.0001621698681108632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0760_text_document -0.00025658329333000087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0761_text_document -0.0002625776226522738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0762_text_document -0.00018893904126945972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0763_text_document -0.00019173419836462428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0764_text_document -0.00024972708669590365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0765_text_document -0.0002565621075859928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0766_text_document -0.0002548091984702725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0767_text_document -0.00024781120449025493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0768_text_document -0.00024190263274768403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0769_text_document -0.00024935480538538027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0770_text_document -0.00024565807926820224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0771_text_document -0.00024335665926774057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0772_text_document -0.0002407471035651234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0773_text_document -0.00024409063432302957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0774_text_document -0.00025048184051844287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0775_text_document -2.7431736503196682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_middle-0776_text_document -0.0001542652540558753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0000_text_document -0.0001414689533672357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0001_text_document -0.00014218991553196462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0002_text_document -0.00014380616486339045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0003_text_document -0.00014537826992690233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0004_text_document -0.00015240156803853129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0005_text_document -0.0001508299161037807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0006_text_document -0.0001645724380011881 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0007_text_document -0.0001636434127327491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0008_text_document -0.0001425695379726649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0009_text_document -0.00015038309042278246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0010_text_document -0.00015551331010771582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0011_text_document -0.00014395190746068794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0012_text_document -0.00014572155617954775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0013_text_document -0.00014985257363654754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0014_text_document -0.00016517178815597176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0015_text_document -0.00015368391453534256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0016_text_document -0.00013802907993189142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0017_text_document -0.0001438832947332681 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0018_text_document -0.0001453654604013201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0019_text_document -0.00015126685069470999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0020_text_document -0.00014666492015973732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0021_text_document -0.00015372684675786069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0022_text_document -0.0001466694423156705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0023_text_document -0.00014645983052842166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0024_text_document -0.00014464707855314855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0025_text_document -0.00014224079429035223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0026_text_document -0.00015150561574001976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0027_text_document -0.00014869251464718684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0028_text_document -0.00014975351070572874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0029_text_document -0.00014693519813853656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0030_text_document -0.00015177096878176463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0031_text_document -0.0001541385774188545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0032_text_document -0.00014686140972307025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0033_text_document -0.00014836061485888312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0034_text_document -0.00015908940031748178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0035_text_document -0.00014335960523511807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0036_text_document -0.00014014336145596836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0037_text_document -0.00014804788542816872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0038_text_document -0.00014447262570766296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0039_text_document -0.0001490836674378867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0040_text_document -0.00015491171627451768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0041_text_document -0.00014704465686983656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0042_text_document -0.00015578029994136968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0043_text_document -0.00014442509556094932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0044_text_document -0.00016401352835433973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0045_text_document -0.0001426617272165932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0046_text_document -0.00014952006301290383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0047_text_document -0.00014858509055287383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0048_text_document -0.0001452147802800582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0049_text_document -0.00014648995026373163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0050_text_document -0.000150292569067835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0051_text_document -0.00015359505638013499 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0052_text_document -0.00014342220561517732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0053_text_document -0.00015037020981817882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0054_text_document -0.0001442503228598675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0055_text_document -0.00015512168691210362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0056_text_document -0.000141978855262853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0057_text_document -0.0001433238477981227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0058_text_document -0.0001522852753554881 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0059_text_document -0.00015750021259583146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0060_text_document -0.0001620583984355833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0061_text_document -0.00014425968431250636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0062_text_document -0.00015502607180742606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0063_text_document -0.00014808719854384823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0064_text_document -0.00014037741406088144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0065_text_document -0.00014415351915599912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0066_text_document -0.00014669998038063754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0067_text_document -0.00014168851942590583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0068_text_document -0.00016615444649487683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0069_text_document -0.00017314227247280456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0070_text_document -0.00014511886160872687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0071_text_document -0.0001589885117911034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0072_text_document -0.0001468857466370262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0073_text_document -0.00014409172483178647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0074_text_document -0.00017524066610798787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0075_text_document -0.0001423201779575328 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0076_text_document -0.00014813204150867026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0077_text_document -0.0001426495065609589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0078_text_document -0.00015198519700337085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0079_text_document -0.0001407239353962083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0080_text_document -0.00015564799275992607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0081_text_document -0.00014044706039573722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0082_text_document -0.00014271692599994692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0083_text_document -0.000145622079855115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0084_text_document -0.0001420329587382314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0085_text_document -0.00014388823447845187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0086_text_document -0.0001386395317413269 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0087_text_document -0.00014615498445222442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0088_text_document -0.00014100731560794867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0089_text_document -0.0001412468938663676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0090_text_document -0.0001448361986040564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0091_text_document -0.00015041376595655126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0092_text_document -0.00015393889374199827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0093_text_document -0.0001424230223910099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0094_text_document -0.00013832238850082653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0095_text_document -0.00014573052620396468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0096_text_document -0.00014871061906625763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0097_text_document -0.0001474653563212365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0098_text_document -0.00014332440162216428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0099_text_document -0.00013995360169386805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0100_text_document -0.0001396957447740551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0101_text_document -0.00014451429874557317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0102_text_document -0.00014667057760559536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0103_text_document -0.00014311302174425863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0104_text_document -0.0001486303888676766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0105_text_document -0.00014984904337848564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0106_text_document -0.00014471364010783683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0107_text_document -0.00014422564733335141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0108_text_document -0.00014833706425660122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0109_text_document -0.0001547519654335586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0110_text_document -0.00016861028196725518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0111_text_document -0.00014655906054430117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0112_text_document -0.00016230445673145143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0113_text_document -0.0001608744287595928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0114_text_document -0.00014838797263124772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0115_text_document -0.00013772432541929463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0116_text_document -0.00014338391080519997 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0117_text_document -0.00013969596121954725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0118_text_document -0.00014433977111903384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0119_text_document -0.00013940910504621967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0120_text_document -0.00013841015875212353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0121_text_document -0.00015084460181936482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0122_text_document -0.00015609034169658813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0123_text_document -0.00014131566380676185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0124_text_document -0.0001489310284479002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0125_text_document -0.000142611271970708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0126_text_document -0.00013893968956373896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0127_text_document -0.00014729183194546773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0128_text_document -0.00013844600256987405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0129_text_document -0.00014038359448051134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0130_text_document -0.00014148398954188355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0131_text_document -0.00014453817241187933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0132_text_document -0.00014513564218102443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0133_text_document -0.00013715869534969562 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0134_text_document -0.00013954027841855143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0135_text_document -0.0001412204761634212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0136_text_document -0.00014007212448120704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0137_text_document -0.00013479800901952131 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0138_text_document -0.00014295404043242684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0139_text_document -0.00013573518591642275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0140_text_document -0.00013489486257742554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0141_text_document -0.00014869208126259815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0142_text_document -0.0001475290332523071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0143_text_document -0.00013460777613768496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0144_text_document -0.00013367855194670696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0145_text_document -0.00014765624643721848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0146_text_document -0.000134270744838921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0147_text_document -0.0001482262234332188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0148_text_document -0.00013864007544648005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0149_text_document -0.00014100224826604942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0150_text_document -0.00013048509121512907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0151_text_document -0.00012998584056022605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0152_text_document -0.00013999889926826433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0153_text_document -0.00015896524130927037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0154_text_document -0.00013250422088217822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0155_text_document -0.00013171712561130657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0156_text_document -0.0001353745598377907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0157_text_document -0.00014385989862913682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0158_text_document -0.0001337953809308385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0159_text_document -0.00013398910556960017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0160_text_document -0.00013559702582181232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0161_text_document -0.0001479395819777683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0162_text_document -0.00013591800338063272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0163_text_document -0.0001378006151746279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0164_text_document -0.0001348466006623147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0165_text_document -0.0001333909410523815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0166_text_document -0.0001395413623736275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0167_text_document -0.0001340932768114764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0168_text_document -0.0001401874805500622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0169_text_document -0.00013331732046238236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0170_text_document -0.00013697107190707125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0171_text_document -0.00014902856836260464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0172_text_document -0.0001360841851454116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0173_text_document -0.0001430140955418351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0174_text_document -0.00013769551793105646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0175_text_document -0.00013655894858384475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0176_text_document -0.00013428329448183135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0177_text_document -0.00013659792851661152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0178_text_document -0.0001390035871359384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0179_text_document -0.00013373343260207954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0180_text_document -0.0001377384027675603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0181_text_document -0.00013688101750180593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0182_text_document -0.00013942483868376482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0183_text_document -0.00013069676947684327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0184_text_document -0.00013248181223347942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0185_text_document -0.00013404120081582244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0186_text_document -0.0001361765930560515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0187_text_document -0.00012895481023244784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0188_text_document -0.0001269948854413741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0189_text_document -0.00012923062571125647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0190_text_document -0.00013146444734116587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0191_text_document -0.00012866221788337398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0192_text_document -0.00012734919091675074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0193_text_document -0.00012491017673902597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0194_text_document -0.00012532940253474304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0195_text_document -0.00012942822588429847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0196_text_document -0.00012954876208363892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0197_text_document -0.00012757889363363662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0198_text_document -0.00012925483823692497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0199_text_document -0.00012887114306702046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0200_text_document -0.00014305235456549959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0201_text_document -0.00012924991650829868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0202_text_document -0.00013939886744592149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0203_text_document -0.00013473816912159447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0204_text_document -0.00012443870588817695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0205_text_document -0.00012352413384768962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0206_text_document -0.00012363992848397884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0207_text_document -0.00012876521187895858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0208_text_document -0.00012998676310669105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0209_text_document -0.00013573707197851088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0210_text_document -0.00012914628304832383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0211_text_document -0.00012135846145074816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0212_text_document -0.0001272491158502837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0213_text_document -0.00014048669089899133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0214_text_document -0.00012821863542952837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0215_text_document -0.00012843614908145614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0216_text_document -0.00012566972592748682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0217_text_document -0.00012623965035462757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0218_text_document -0.00012745682281848042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0219_text_document -0.00012684031670531754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0220_text_document -0.00013734922167929733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0221_text_document -0.00012364311692105248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0222_text_document -0.00012150014908859676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0223_text_document -0.00013255947544281956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0224_text_document -0.00013080450775030287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0225_text_document -0.00012642072366799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0226_text_document -0.00012748944981690917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0227_text_document -0.0001272640012288133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0228_text_document -0.00012676436334132504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0229_text_document -0.00011463874381385243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0230_text_document -0.0001259763726722257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0231_text_document -0.00013265355691888996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0232_text_document -0.00012800075083395775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0233_text_document -0.00012600035320386608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0234_text_document -0.00012796669337022373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0235_text_document -0.0001281363666451258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0236_text_document -0.00013103924202277517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0237_text_document -0.00013710099201804686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0238_text_document -0.00012390494315996567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0239_text_document -0.00012375130141281296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0240_text_document -0.00012654460329615904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0241_text_document -0.00013347917998097572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0242_text_document -0.00012957465780002206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0243_text_document -0.00012365300899833007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0244_text_document -0.00012759104863989702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0245_text_document -0.00012669826503428652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0246_text_document -0.00019585621938937627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0247_text_document -0.00017199211271798405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0248_text_document -0.00017151003159557893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0249_text_document -0.0001704765251017538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0250_text_document -0.00016915887010107177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0251_text_document -0.000164507074917777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0252_text_document -0.0001707345009802067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0253_text_document -0.00016235282921392888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0254_text_document -0.0001603312806389334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0255_text_document -0.00016063927887228715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0256_text_document -0.0001682293216120587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0257_text_document -0.00016945118701893779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0258_text_document -0.00016510575549830714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0259_text_document -0.00015878514261762818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0260_text_document -0.00016058925849180358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0261_text_document -0.00016806270202025228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0262_text_document -0.0001601743221175851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0263_text_document -0.00016728028661189246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0264_text_document -0.00016271828570438892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0265_text_document -0.0001663197659329172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0266_text_document -0.00016211838369998094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0267_text_document -0.00016174818095722866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0268_text_document -0.00016660916885770873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0269_text_document -0.00016749279166083448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0270_text_document -0.00015990162967327836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0271_text_document -0.00016050019425679443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0272_text_document -0.00015826664805809287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0273_text_document -0.00015906002765230277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0274_text_document -0.00016496336225309003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0275_text_document -0.00015969348413764765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0276_text_document -0.00015888249989873604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0277_text_document -0.0001588217905168081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0278_text_document -0.0001579176192128451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0279_text_document -0.0001599592014593771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0280_text_document -0.00015860202306757735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0281_text_document -0.00015475539919197688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0282_text_document -0.0001606154789998261 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0283_text_document -0.00015967691482799697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0284_text_document -0.00015467004809542842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0285_text_document -0.00015681467419158087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0286_text_document -0.0001622263618651377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0287_text_document -0.00016071879902106084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0288_text_document -0.00015926245724996415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0289_text_document -0.00015865169965265541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0290_text_document -0.0001558589009989086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0291_text_document -0.00015834413702510978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0292_text_document -0.00015984235618630313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0293_text_document -0.00015906347325722462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0294_text_document -0.0001540401129832678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0295_text_document -0.00015709268423517463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0296_text_document -0.00016150611616707217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0297_text_document -0.0001575761279522917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0298_text_document -0.00015145845456253164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0299_text_document -0.00015531545597525365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0300_text_document -0.00015290580088858923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0301_text_document -0.00015077381822016696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0302_text_document -0.00016026706987479596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0303_text_document -0.00015143811781794564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0304_text_document -0.00015335594803302406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0305_text_document -0.00015760769888428818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0306_text_document -0.00016811053178478525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0307_text_document -0.00021456946285616728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0308_text_document -0.00021300214303968855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0309_text_document -0.00020349194545531642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0310_text_document -0.00021281325399560017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0311_text_document -0.00020973400589848146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0312_text_document -0.00020126033912157333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0313_text_document -0.00020674507357011296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0314_text_document -0.00021222543863325493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0315_text_document -0.0002050723383820817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0316_text_document -0.00021804813803312056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0317_text_document -0.0002008803314227051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0318_text_document -0.0002150047024098784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0319_text_document -0.00020318723314588857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0320_text_document -0.00020021945595806058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0321_text_document -0.00020351797666608406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0322_text_document -0.00020832621085218548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0323_text_document -0.0002035206051090622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0324_text_document -0.00020272338181805027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0325_text_document -0.00020460676190716195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0326_text_document -0.00020717814792849565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0327_text_document -0.0001955827435950214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0328_text_document -0.00020417807396352577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0329_text_document -0.0002011029760914888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0330_text_document -0.00020292765823625672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0331_text_document -0.00020035339845060027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0332_text_document -0.0002019662525247444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0333_text_document -0.000206838061219021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0334_text_document -0.0001941713531348939 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0335_text_document -0.00010283055875342613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0336_text_document -0.00010052128921034293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0337_text_document -0.00021410992316202177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0338_text_document -0.0002019050315219438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0339_text_document -0.0001986035353671086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0340_text_document -0.00019334420113344198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0341_text_document -0.0002040134561840194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0342_text_document -0.00019786749210973914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0343_text_document -0.00021292248961774976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0344_text_document -0.000198975254462317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0345_text_document -0.00019270601369753864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0346_text_document -0.0001938662101557011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0347_text_document -0.0002029610545170106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0348_text_document -0.0002024962737322469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0349_text_document -0.0002075197885043544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0350_text_document -0.00019871985248356538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0351_text_document -0.0001949694696029141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0352_text_document -0.00020180408203543252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0353_text_document -0.00019545199817763088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0354_text_document -0.00019734611243298183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0355_text_document -0.00021047242956266074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0356_text_document -0.0001968562822164333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0357_text_document -0.00019972266049102106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0358_text_document -0.00020126122390730825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0359_text_document -0.00019799642896091898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0360_text_document -0.0002021712802087185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0361_text_document -0.0001941903201275054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0362_text_document -0.00019307283352311706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0363_text_document -0.00016598318480128866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0364_text_document -0.00016504803365649659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0365_text_document -0.00016630327313193533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0366_text_document -0.00016601923469884076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0367_text_document -0.0001681694501234557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0368_text_document -0.00016859564709291555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0369_text_document -0.00016845101707974437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0370_text_document -0.0001643037792913447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0371_text_document -0.00016186624765418046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0372_text_document -0.00016697344045101027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0373_text_document -0.00016669715111205908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0374_text_document -0.00016364850623567704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0375_text_document -0.0001634811496926281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0376_text_document -0.00016825687707295152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0377_text_document -0.0001627585946667742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0378_text_document -0.00016582351614544805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0379_text_document -0.0001630893218980273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0380_text_document -0.0001568416151151013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0381_text_document -0.00017273044852059518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0382_text_document -0.00016016530273273665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0383_text_document -0.00015777742226002822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0384_text_document -0.00016385370668116144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0385_text_document -0.00016954547679602915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0386_text_document -0.0001676626705219338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0387_text_document -0.00016250610371947111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0388_text_document -0.00016004510983519738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0389_text_document -0.000161815220320894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0390_text_document -0.00016744693680716642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0391_text_document -0.00015604191096880147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0392_text_document -0.0001636895622681933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0393_text_document -0.000158886517344257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0394_text_document -0.0001558972054341701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0395_text_document -0.0001591533045533395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0396_text_document -0.0001657955386528658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0397_text_document -0.00016060726764524156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0398_text_document -0.00016167923208527019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0399_text_document -0.00015484394662326808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0400_text_document -0.00016052047349647775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0401_text_document -0.0001582576585363055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0402_text_document -0.0001545777833300399 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0403_text_document -0.00016303475566860345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0404_text_document -0.0001627904173369714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0405_text_document -0.0001567550665344843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0406_text_document -0.0001587287727580368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0407_text_document -0.0001606889088117574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0408_text_document -0.00016206324217472778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0409_text_document -0.00015712668987045555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0410_text_document -0.0001607143430081059 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0411_text_document -0.00015230600229428526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0412_text_document -0.00016067822548676263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0413_text_document -0.00015993580979768466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0414_text_document -0.00016379843410396262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0415_text_document -0.0001533135627240871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0416_text_document -0.00016861285265852845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0417_text_document -0.0001632799417656467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0418_text_document -0.00015962871905586431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0419_text_document -0.00015014915949133304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0420_text_document -0.00015059096546324844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0421_text_document -0.00015841934874861877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0422_text_document -0.000152377097357806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0423_text_document -0.00014942797865989248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0424_text_document -0.00015640838403734855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0425_text_document -0.0001557305888039896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0426_text_document -0.00014992907934376868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0427_text_document -0.00015847297170019638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0428_text_document -0.0001563057066889321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0429_text_document -0.00015425884830587555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0430_text_document -0.00015294599138593887 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0431_text_document -0.00015307387809393826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0432_text_document -0.00016021533866175615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0433_text_document -0.00015819924688246454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0434_text_document -0.00014854336050366525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0435_text_document -0.00015428039783626384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0436_text_document -0.00015380539006369472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0437_text_document -0.00015543551510602353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0438_text_document -0.00015792640857808265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0439_text_document -0.00015591945366146652 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0440_text_document -0.00014809559672766608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0441_text_document -0.00015190843215388426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0442_text_document -0.00014890757113683386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0443_text_document -0.0001610286090290533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0444_text_document -0.00015061787553649923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0445_text_document -0.00014811603935037767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0446_text_document -0.00015254163073097444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0447_text_document -0.00015300211863900935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0448_text_document -0.00015063192030688013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0449_text_document -0.00015300622789493292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0450_text_document -0.00015096280425750327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0451_text_document -0.00015205454100558358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0452_text_document -0.00015121161958027361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0453_text_document -0.0001493611157597698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0454_text_document -0.00015838957873196607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0455_text_document -0.0001497669779590609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0456_text_document -0.00015173657097785533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0457_text_document -0.0001542516903028995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0458_text_document -0.000149139532833868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0459_text_document -0.00014644441551246194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0460_text_document -0.00015166787754612994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0461_text_document -0.00014923555170687534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0462_text_document -0.00015589324574035403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0463_text_document -0.00015022803227804745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0464_text_document -0.00015127324533861265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0465_text_document -0.00014783676790095657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0466_text_document -0.00014927753645591052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0467_text_document -0.00014753911610765252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0468_text_document -0.00014886425094132403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0469_text_document -0.00014432622711023067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0470_text_document -0.00015087353567030766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0471_text_document -0.00015318739523991737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0472_text_document -0.00014716603935377532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0473_text_document -0.00015032310787320853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0474_text_document -0.00014425315738264723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0475_text_document -0.0001507311940067415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0476_text_document -0.0001735562949386336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0477_text_document -0.0001664225151007615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0478_text_document -0.00017016223341338198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0479_text_document -0.0001686337558140661 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0480_text_document -0.00018737654520115072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0481_text_document -0.00016696818282464752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0482_text_document -0.00017542891864931188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0483_text_document -0.000168925038877694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0484_text_document -0.0001769097096293462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0485_text_document -0.00017465563985682533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0486_text_document -0.0001704723163845607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0487_text_document -0.00017113194080906855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0488_text_document -0.00017056770492485763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0489_text_document -0.0001736825492971628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0490_text_document -0.00017060994856935613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0491_text_document -0.00017539355807018588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0492_text_document -0.00017512560274649157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0493_text_document -0.00017536288179601056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0494_text_document -0.00017214679473623093 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0495_text_document -0.00017372473469635212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0496_text_document -0.00016968876198424372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0497_text_document -0.00017328658337078598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0498_text_document -0.00016545006523949998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0499_text_document -0.0001712623636560391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0500_text_document -0.00017259544872761246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0501_text_document -0.00016731532955664165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0502_text_document -0.00017234554920296389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0503_text_document -0.00016824263782247044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0504_text_document -0.00017046154865322805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0505_text_document -0.00016701775451880861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0506_text_document -0.0001640723558698162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0507_text_document -0.00016912021224512063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0508_text_document -0.00016148128416798815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0509_text_document -0.00017033021559990035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0510_text_document -0.00016742449903581303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0511_text_document -0.00016604941440707502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0512_text_document -0.00016168218680070063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0513_text_document -0.00016545734985198287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0514_text_document -0.00016617264790719555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0515_text_document -0.00016903898126379064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0516_text_document -0.00016251470403425602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0517_text_document -0.00016741321573477316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0518_text_document -0.00016314387702135404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0519_text_document -0.00016261766224352778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0520_text_document -0.00016043765927930694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0521_text_document -0.0001581188444159775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0522_text_document -0.0001675593630876091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0523_text_document -0.00016225811098829194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0524_text_document -0.00016027854790273813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0525_text_document -0.00015477514040295668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0526_text_document -0.00016132027735084922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0527_text_document -0.00016144543812901825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0528_text_document -0.00016356924967160763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0529_text_document -0.00016721507926064277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0530_text_document -0.0001623283758093546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0531_text_document -0.00016540060361910116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0532_text_document -0.00016618517731232895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0533_text_document -0.0001661140965633334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0534_text_document -0.00016521134906101744 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0535_text_document -0.0001605250452596446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0536_text_document -0.00016158626615495202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0537_text_document -0.00016348402666537893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0538_text_document -0.00015887094758334445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0539_text_document -0.00016216761850919694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0540_text_document -0.00016125922688833952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0541_text_document -0.00015719662175540762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0542_text_document -0.00016177908132776304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0543_text_document -0.0001616654955707841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0544_text_document -0.0001575744247706023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0545_text_document -0.00016594502227726776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0546_text_document -0.00016680360478028852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0547_text_document -0.00016969508752354227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0548_text_document -0.00018702211879271686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0549_text_document -0.00019358085009705273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0550_text_document -0.0001871367897387826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0551_text_document -0.00018452058370522755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0552_text_document -0.0001850164319455863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0553_text_document -0.00018589455402222413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0554_text_document -0.00018848818876445855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0555_text_document -0.00018677441309244695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0556_text_document -0.00018806266359047162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0557_text_document -0.00018742615490284408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0558_text_document -0.00018308658912909244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0559_text_document -0.00017917024956722993 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0560_text_document -0.0001796815083811096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0561_text_document -0.00018830762534435366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0562_text_document -0.0001850705756497164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0563_text_document -0.00018620607609678367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0564_text_document -0.00018735293561315315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0565_text_document -0.00018406055855123805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0566_text_document -0.00018296049025592247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0567_text_document -0.00018407127494772196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0568_text_document -0.0001809459590066732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0569_text_document -0.00018206921683271417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0570_text_document -0.0001823423927624476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0571_text_document -0.00017843504198196598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0572_text_document -0.0001849074668186014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0573_text_document -0.0001812163144813499 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0574_text_document -0.00018309068999374263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0575_text_document -0.00018500613289155086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0576_text_document -0.00017930403632760822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0577_text_document -0.0001846380543749688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0578_text_document -0.0001805411790348431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0579_text_document -0.00017815258406988848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0580_text_document -0.00017771149209661494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0581_text_document -0.000179212119800064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0582_text_document -0.0001770710081666354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0583_text_document -0.00018076802304233783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0584_text_document -0.00018266780486243524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0585_text_document -0.00017952537023013302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0586_text_document -0.00017482592939671484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0587_text_document -0.00017479307237867526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0588_text_document -0.00017947982239834899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0589_text_document -0.00017800230944457152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0590_text_document -0.0001768045667273756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0591_text_document -0.00018432659029891628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0592_text_document -0.00017860310980883306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0593_text_document -0.00017352563618741148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0594_text_document -0.000177967402241009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0595_text_document -0.0001761394507080597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0596_text_document -0.0001727461411889822 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0597_text_document -0.00017520765607261058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0598_text_document -0.00017389963918978602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0599_text_document -0.00017297383567671195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0600_text_document -0.00017186248654837811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0601_text_document -0.00018016764298215066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0602_text_document -0.00017252933018279703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0603_text_document -0.0001720498259217191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0604_text_document -0.00017208910794032673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0605_text_document -0.0001638288329725128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0606_text_document -0.00015774370365565657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0607_text_document -0.00015428183891406193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0608_text_document -0.0001579263490987627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0609_text_document -0.00015679781661701012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0610_text_document -0.00015686067490532405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0611_text_document -0.00015476043642401294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0612_text_document -0.0001538144005636655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0613_text_document -0.00015471809257783847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0614_text_document -0.00014950254548936378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0615_text_document -0.00015189343275275787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0616_text_document -0.00016808135779534307 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0617_text_document -0.00015331380459020154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0618_text_document -0.00015025506525877266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0619_text_document -0.00015705079524537657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0620_text_document -0.00014843144411648014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0621_text_document -0.0001536670204340525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0622_text_document -0.00014701650982417206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0623_text_document -0.0001470830903826265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0624_text_document -0.00014669457615379322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0625_text_document -0.00015327731341039172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0626_text_document -0.00016421071093813112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0627_text_document -0.00014320086554259857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0628_text_document -0.00014733292080267092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0629_text_document -0.00014574339323444963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0630_text_document -0.00014508510524362508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0631_text_document -0.0001510667294376284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0632_text_document -0.00014448955337404646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0633_text_document -0.00015189242851477872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0634_text_document -0.0001408976680729981 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0635_text_document -0.00014495438771487836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0636_text_document -0.00014607129482780071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0637_text_document -0.0001425703250247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0638_text_document -0.00014772556798043487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0639_text_document -0.0001454755294743558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0640_text_document -0.00014604759342940054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0641_text_document -0.000144987966876031 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0642_text_document -0.00014159362399631978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0643_text_document -0.00015166107543186514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0644_text_document -0.00013872638536941069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0645_text_document -0.00014392691133816916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0646_text_document -0.00014527538230304764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0647_text_document -0.0001445241296159157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0648_text_document -0.00014566980102669863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0649_text_document -0.00014105957349679274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0650_text_document -0.00014407711883329926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0651_text_document -0.00014304333666146412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0652_text_document -0.00014480474786471068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0653_text_document -0.00014513562095603888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0654_text_document -0.00014216954843071324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0655_text_document -0.0001472056417215835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0656_text_document -0.0001411732545194045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0657_text_document -0.00014472737242668624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0658_text_document -0.0001412212585262607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0659_text_document -0.00020834639482623596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0660_text_document -0.00019484913874296875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0661_text_document -0.00019400182473285833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0662_text_document -0.000192581173021768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0663_text_document -0.0001958163408499538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0664_text_document -0.00019017201894348343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0665_text_document -0.00018748712836308062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0666_text_document -0.00019398325978096153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0667_text_document -0.00018740362852951608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0668_text_document -0.00018769931256921782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0669_text_document -0.00018841740417805205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0670_text_document -0.0001897879160564146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0671_text_document -0.00018663113185306689 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0672_text_document -0.00018894652949372258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0673_text_document -0.0001929378648272062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0674_text_document -0.00019134942047365448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0675_text_document -0.00018699153383533985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0676_text_document -0.00018610331853766602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0677_text_document -0.0001863160274451902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0678_text_document -0.00018636405144302115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0679_text_document -0.00018489348621678148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0680_text_document -0.0001860176372198307 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0681_text_document -0.00018315031813541827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0682_text_document -0.00019049993633217256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0683_text_document -0.00018374255446481207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0684_text_document -0.00017918235151102646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0685_text_document -0.00018078078222027994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0686_text_document -0.00018377134048126254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0687_text_document -0.00018119048712916442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0688_text_document -0.00018226290667237163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0689_text_document -0.00018539016766122422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0690_text_document -0.00018304864675259609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0691_text_document -0.00018006283819913595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0692_text_document -0.00017853375396011673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0693_text_document -0.0001806080666151815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0694_text_document -0.00018287085590792935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0695_text_document -0.00018102703894508278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0696_text_document -0.00017985249563069855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0697_text_document -0.00018055111208127884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0698_text_document -0.00017436715651687287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0699_text_document -0.0001750410902836745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0700_text_document -0.0001755658852086883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0701_text_document -0.00017704710809249836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0702_text_document -0.00017563712144304312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0703_text_document -0.00017646118668991032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0704_text_document -0.0001738273848965312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0705_text_document -0.00017355052248297015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0706_text_document -0.00017182494917422235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0707_text_document -0.0001796801127149085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0708_text_document -0.0001535678074475219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0709_text_document -0.00016509131806569352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0710_text_document -0.0001660762988129014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0711_text_document -0.00017181117317139103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0712_text_document -0.00016385189811495075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0713_text_document -0.00016321938466065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0714_text_document -0.0001627668114510062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0715_text_document -0.0001667874841569603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0716_text_document -0.0001647336272051215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0717_text_document -0.00015927038206724374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0718_text_document -0.000163069807004626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0719_text_document -0.00016643362662749963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0720_text_document -0.0001598347201275479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0721_text_document -0.00016414824852047793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0722_text_document -0.00016387374849716915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0723_text_document -0.00016218986007283508 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0724_text_document -0.00016170100645242406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0725_text_document -0.00016794279442600715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0726_text_document -0.00016410407241508566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0727_text_document -0.00016663924614304762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0728_text_document -0.0001610334643678992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0729_text_document -0.00016082817926927476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0730_text_document -0.00016483710320531984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0731_text_document -0.00015950564573034403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0732_text_document -0.00016176598872010603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0733_text_document -0.00016374799045777884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0734_text_document -0.00016207070843359862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0735_text_document -0.000161310121195263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0736_text_document -0.0001590930806312555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0737_text_document -0.00015872700071854542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0738_text_document -0.0001601426608559989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0739_text_document -0.0001592737504230903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0740_text_document -0.0001599609389465664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0741_text_document -0.0001573951015313392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0742_text_document -0.00015918138446881715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0743_text_document -0.00016063409035052854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0744_text_document -0.00015479247307168076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0745_text_document -0.0001590206266750552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0746_text_document -0.00016413616409963463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0747_text_document -0.00015909403254717725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0748_text_document -0.00015912638065916792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0749_text_document -0.00015509170291798033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0750_text_document -0.00015668221053756931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0751_text_document -0.00015993661313870757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0752_text_document -0.00015986553041529475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0753_text_document -0.0001551253906720823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0754_text_document -0.0001569044427999477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0755_text_document -0.00015512319487328638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0756_text_document -0.00016021882869106635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0757_text_document -0.00015415106017838012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0758_text_document -0.00015711650631982987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0759_text_document -0.00015512670736159294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0760_text_document -0.00016200410442893923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0761_text_document -0.00015949285619573655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0762_text_document -0.0001625616727060612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0763_text_document -0.00016316486655764686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0764_text_document -0.0001571167311565954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0765_text_document -0.00016128213234978153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0766_text_document -0.00015535324730882956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0767_text_document -0.0001579934311592013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0768_text_document -0.00015195311864613838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0769_text_document -0.0001615190125670139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0770_text_document -0.00015867133202388371 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0771_text_document -0.00015932910049616658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0772_text_document -0.00015735730575532447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0773_text_document -0.00016192787415292593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0774_text_document -0.00015443514945271916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0775_text_document -0.00015290872574095856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0776_text_document -0.0001586657525675075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0777_text_document -0.0001561292345081933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0778_text_document -0.0001584146414910674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0779_text_document -0.00015282231142071527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0780_text_document -0.0001561252202711004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0781_text_document -0.00015508367049496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0782_text_document -0.00015211947613405347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0783_text_document -0.00014976529550875275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0784_text_document -0.00015418186133444713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0785_text_document -0.00015777360151582686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0786_text_document -0.000152640262498424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0787_text_document -0.00015418142572863903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0788_text_document -0.00015502601134089746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0789_text_document -0.00015405733434421877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0790_text_document -0.00015484459497253604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0791_text_document -0.0001541867208689297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0792_text_document -0.00015014404352940876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0793_text_document -0.00015357544967633106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0794_text_document -0.00015037823631794736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0795_text_document -0.00015025795679285078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0796_text_document -0.00014876992710553488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0797_text_document -0.00015032669711698612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0798_text_document -0.00015596697517010466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0799_text_document -0.00015498394440674378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0800_text_document -0.00014757314272111684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0801_text_document -0.00014919071614611802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0802_text_document -0.00014686280514246915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0803_text_document -0.00015882771228777683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0804_text_document -0.00014763597756322578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0805_text_document -0.00014785441795725526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0806_text_document -0.00015313024795352964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0807_text_document -0.0001497627986113246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0808_text_document -0.00014499607432690722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0809_text_document -0.0001461719027401259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0810_text_document -0.00014839933441537366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0811_text_document -0.0001475840995029022 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0812_text_document -0.00015065512711375653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0813_text_document -0.00015285087358760883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0814_text_document -0.00014861957547794477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0815_text_document -0.00014996949492468605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0816_text_document -0.0001472998668365096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0817_text_document -0.0001464012147691964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0818_text_document -0.00015227635617231567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0819_text_document -0.0001491494017117428 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0820_text_document -0.00014464475787246092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0821_text_document -0.00014410767861685618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0822_text_document -0.000144919516791233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0823_text_document -0.00014507990635617585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0824_text_document -0.0001468797342896656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0825_text_document -0.0001422000420712919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0826_text_document -0.00014228987139298954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0827_text_document -0.00014481016912090385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0828_text_document -0.000142802473797815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0829_text_document -0.00014812295450003065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0830_text_document -0.00014697991622146685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0831_text_document -0.000143946325289488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0832_text_document -0.0001418544716646782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0833_text_document -0.00014706985092768576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0834_text_document -0.0001411487598988699 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0835_text_document -0.0001583983550166893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0836_text_document -0.00015370277071378533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0837_text_document -0.0001574284524004961 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0838_text_document -0.00016033599900258183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0839_text_document -0.00016159470012508268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0840_text_document -0.00015624921021983388 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0841_text_document -0.0001603288323615303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0842_text_document -0.00016421653645625842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0843_text_document -0.00016136751182857813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0844_text_document -0.0001644008542307843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0845_text_document -0.00016320230298972016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0846_text_document -0.00016176830866038722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0847_text_document -0.00015883945834286212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0848_text_document -0.00015854734059433728 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0849_text_document -0.00015424048326372636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0850_text_document -0.00015913631543321879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0851_text_document -0.00016242367155204024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0852_text_document -0.00016352898883564303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0853_text_document -0.00016283852574114027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0854_text_document -0.0001597064012689706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0855_text_document -0.00015723207463854053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0856_text_document -0.00016082454091186785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0857_text_document -0.00015148430437371348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0858_text_document -0.00015699196205345046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0859_text_document -0.00016323993834433252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0860_text_document -0.00015419189482936103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0861_text_document -0.00014984592429281824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0862_text_document -0.0001540327550705441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0863_text_document -0.00015559458082419316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0864_text_document -0.00015809601003355687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0865_text_document -0.00015561437781246056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0866_text_document -0.00015650965510707114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0867_text_document -0.00015654223175785975 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0868_text_document -0.00015966194232830576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0869_text_document -0.0001542791440813034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0870_text_document -0.00016358133853488976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0871_text_document -0.0001610108148402946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0872_text_document -0.0001567861463301872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0873_text_document -0.00015916579076809533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0874_text_document -0.00015834187212170972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0875_text_document -0.00015492852942470005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0876_text_document -0.0001565761307746086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0877_text_document -0.00016111787860345758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0878_text_document -0.00015262185821473176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0879_text_document -0.00015609313599061615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0880_text_document -0.00015265109415151545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0881_text_document -0.00015596676711588585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0882_text_document -0.00015602244000618423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0883_text_document -0.00015533087814847594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0884_text_document -0.000148761688602713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0885_text_document -0.00015124065708812265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0886_text_document -0.00015177148904071277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0887_text_document -0.00015551510213818192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0888_text_document -0.00015328016792414618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0889_text_document -0.00014826652573194586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0890_text_document -0.00015618973632950672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0891_text_document -0.00016465597460827412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0892_text_document -0.00017729797829003265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0893_text_document -0.00017645710877786075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0894_text_document -0.000173993320599559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0895_text_document -0.0001752697954262395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0896_text_document -0.00017545831920313468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0897_text_document -0.00017512052874093406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0898_text_document -0.00017596295211949001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0899_text_document -0.0001763343681416489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0900_text_document -0.00016737628055788186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0901_text_document -0.00017659674006013248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0902_text_document -0.00017521085067973818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0903_text_document -0.00018110203496350606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0904_text_document -0.00016887408015540739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0905_text_document -0.0001730418383091983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0906_text_document -0.00017084812178309202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0907_text_document -0.00016928946570955264 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0908_text_document -0.00017272373105947043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0909_text_document -0.00016793546933797045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0910_text_document -0.00016510473373737477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0911_text_document -0.0001656625036518595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0912_text_document -0.00016849674877913583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0913_text_document -0.00017492155042464418 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0914_text_document -0.00017092357710033054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0915_text_document -0.00016970730743877006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0916_text_document -0.00016573665091766286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0917_text_document -0.00016358480536479716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0918_text_document -0.0001653802811890403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0919_text_document -0.00017231807148475074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0920_text_document -0.00017361608596973323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0921_text_document -0.00017404933358323055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0922_text_document -0.00016371945617952907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0923_text_document -0.00017000836658266155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0924_text_document -0.00017142976487027857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0925_text_document -0.00017006281434704977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0926_text_document -0.0001751965302313473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0927_text_document -0.00016954848753554936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0928_text_document -0.0001683555446267139 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0929_text_document -0.00016921278107076727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0930_text_document -0.00016808682594394623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0931_text_document -0.00017711704047105475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0932_text_document -0.0001675247295876393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0933_text_document -0.00017061773073498863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0934_text_document -0.0001644856648306077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0935_text_document -0.00016530682645009105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0936_text_document -0.00016993430076157017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0937_text_document -0.00016716870217360928 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0938_text_document -0.0001672477045314564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0939_text_document -0.00016150529456268964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0940_text_document -0.0001642955368396883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0941_text_document -0.0001650135010986092 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0942_text_document -0.0001719916971031507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0943_text_document -0.0001663860254017646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0944_text_document -0.00016810785027934324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0945_text_document -0.00016663511368772123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0946_text_document -0.00017120237493641126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0947_text_document -0.0001651698100366788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0948_text_document -0.00016069571413445028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0949_text_document -0.0001631772602215936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0950_text_document -0.00016994484266892867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0951_text_document -0.00016821930169126347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0952_text_document -0.0001680542144940534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0953_text_document -0.00015807234911071054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0954_text_document -0.00016287290799651364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0955_text_document -0.00016674360421415713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0956_text_document -0.0001663549971877126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0957_text_document -0.0001699417467826641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0958_text_document -0.0001661066433849769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0959_text_document -0.00016736976350010906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0960_text_document -0.00016160049405253383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0961_text_document -0.0001625500850979611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0962_text_document -0.00016172349111618741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0963_text_document -0.00016041582790085466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0964_text_document -0.00016369413378455798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0965_text_document -0.00016245798272839223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0966_text_document -0.00016458727969573578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0967_text_document -0.0001618972714257936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0968_text_document -0.00016149423535800886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0969_text_document -0.00015886933917368354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0970_text_document -0.00015721961433801126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0971_text_document -0.00015609496997744904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0972_text_document -0.0001608435755282705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0973_text_document -0.00015730100598754584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0974_text_document -0.00015955845719642757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0975_text_document -0.00015469663090901824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0976_text_document -0.00015812452037199733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0977_text_document -0.00015443940925795885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0978_text_document -0.00015678701926941855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0979_text_document -0.00015787925332384637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0980_text_document -0.00015669644312439214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0981_text_document -0.00015342587917756964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0982_text_document -0.00015642024238741553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0983_text_document -0.0001540823378708023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0984_text_document -0.00015238224416999995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0985_text_document -0.0001522695061784323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0986_text_document -0.00020085620305657233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0987_text_document -0.00014698197479826313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0988_text_document -0.00014796924883111914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0989_text_document -0.0001483800966807953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0990_text_document -0.00014550940307048242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0991_text_document -0.00015052597307667803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0992_text_document -0.00014866583878918362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0993_text_document -0.00014440801314961302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0994_text_document -0.00014295564464645108 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0995_text_document -0.00014903049761507035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0996_text_document -0.00014820091066353183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0997_text_document -0.0001429454882440627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0998_text_document -0.00015048550764172483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-0999_text_document -0.0001430543312039796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1000_text_document -0.00014661342883839465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1001_text_document -0.00014721354013103223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1002_text_document -0.00014780017824708586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1003_text_document -0.0001463184859455721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1004_text_document -0.00014654870719379106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1005_text_document -0.00020943212095457075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1006_text_document -0.00021205821955900777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1007_text_document -0.00014176730212983274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1008_text_document -0.00014026276433980122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1009_text_document -0.00013570196535880505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1010_text_document -0.00014776685378575983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1011_text_document -0.00014138218982193943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1012_text_document -0.0001412602382122253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1013_text_document -0.00013944232659104602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1014_text_document -0.00014570617769030735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1015_text_document -0.00014233071172042007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1016_text_document -0.00014016762901851798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1017_text_document -0.0001434413757259645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1018_text_document -0.00014003324697133565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1019_text_document -0.00014567282904236987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1020_text_document -0.00013992559507863123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1021_text_document -0.00021096883039305026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1022_text_document -0.00014274603730164107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1023_text_document -0.00013914595792215918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1024_text_document -0.00013666688380542608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1025_text_document -0.00014001152690065646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1026_text_document -0.00021392615254787925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1027_text_document -0.00014251166508793392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1028_text_document -0.00013886942449587415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1029_text_document -0.0002078004025575127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1030_text_document -0.00020928673622040174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1031_text_document -0.00020558733131260538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1032_text_document -0.0002036663760886078 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1033_text_document -0.00014592860566679667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1034_text_document -0.00014346325128200297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1035_text_document -0.00014068142446497316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1036_text_document -0.000142996292961803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1037_text_document -0.00020633185839414136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1038_text_document -0.00013684538988274547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1039_text_document -0.0002033768324865864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1040_text_document -0.000200593087523188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1041_text_document -0.0002297294147093001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1042_text_document -0.00022971372080690233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1043_text_document -0.00023092966691083417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1044_text_document -0.00015159247973379415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1045_text_document -0.00015257723761865372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1046_text_document -0.00015750287090187065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1047_text_document -0.00015557071949799488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1048_text_document -0.00015138603787345713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1049_text_document -0.00014966823068820163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1050_text_document -0.00015481393029806212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1051_text_document -0.0001521335747073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1052_text_document -0.00015447866363472483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1053_text_document -0.0001564823000495303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1054_text_document -0.00015484698673224505 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1055_text_document -0.00022305811126444646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1056_text_document -0.00015308102523761935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1057_text_document -0.00022494528198789627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1058_text_document -0.0002206911435725598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1059_text_document -0.00021440132246946592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1060_text_document -0.00014934935094772055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1061_text_document -0.00015275047150828305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1062_text_document -0.00021692931968428998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1063_text_document -0.00023057843831795596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1064_text_document -0.00022061661869945533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1065_text_document -0.0001475889972917192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1066_text_document -0.00014965255899799802 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1067_text_document -0.000146325773766483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1068_text_document -0.00021849119850040293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1069_text_document -0.00021649545481859658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1070_text_document -0.00014463616989778393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1071_text_document -0.00014301572221485565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1072_text_document -0.00014804643324427358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1073_text_document -0.0002143783669071859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1074_text_document -0.0001479303814401362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1075_text_document -0.00015068744684349907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1076_text_document -0.00021658806091136903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1077_text_document -0.00021333945668012075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1078_text_document -0.000142221472149436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1079_text_document -0.0002158096794842747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1080_text_document -0.00021541031163695796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1081_text_document -0.0002160301031804424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1082_text_document -0.00014484879119054217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1083_text_document -0.00014717950537309672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1084_text_document -0.00021016132927298846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1085_text_document -0.00021433713539833563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1086_text_document -0.0001438233936284062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1087_text_document -0.0001447086593934949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1088_text_document -0.00021440017582664183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1089_text_document -0.00020841624205804798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1090_text_document -0.000213227136771408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1091_text_document -0.00020931414236598925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1092_text_document -0.0002134545412666026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1093_text_document -0.0002126803251195216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1094_text_document -0.00014114550507201583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1095_text_document -0.00016444080384922814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1096_text_document -0.0001542515002652382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1097_text_document -0.0001608177523717217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1098_text_document -0.0001577693965006662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1099_text_document -0.0001615213258436368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1100_text_document -0.00014975169893108998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1101_text_document -0.00015902857074290308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1102_text_document -0.00015523901418979132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1103_text_document -0.00015842052994374488 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1104_text_document -0.0001543439686424067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1105_text_document -0.0001559141331005536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1106_text_document -0.0001558557495821586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1107_text_document -0.00016108187362389814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1108_text_document -0.0001605357063724452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1109_text_document -0.0001588416921491903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1110_text_document -0.00015452564563384654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1111_text_document -0.0001575925464658241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1112_text_document -0.000155416389913229 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1113_text_document -0.00015834897089216795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1114_text_document -0.00015376802717866433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1115_text_document -0.00015257616131444455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1116_text_document -0.00015333466381495513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1117_text_document -0.00015356006723825613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1118_text_document -0.00015392513748333956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1119_text_document -0.00015808193589371923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1120_text_document -0.00015572715307115401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1121_text_document -0.00015677288071421776 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1123_text_document -0.00015564703516755468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1124_text_document -0.00015473730933423342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1125_text_document -0.00015227152970932222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1126_text_document -0.00015062363935408713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1127_text_document -0.0001608838990519831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1128_text_document -0.00016058746991656767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1129_text_document -0.00015232158785053588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1130_text_document -0.00015216796930278597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1131_text_document -0.00015531087359959403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1132_text_document -0.00017455174602057423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1133_text_document -0.00015220395996782025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1134_text_document -0.00022536045257736233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1135_text_document -0.00023391977994072452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1136_text_document -0.00022316737354122904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1137_text_document -0.00023097409031198833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1138_text_document -0.0001536444602488289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1139_text_document -0.00015290857223001657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1140_text_document -0.00015053717764782956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1141_text_document -0.0001487906308449292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1142_text_document -0.00022796481136694752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1143_text_document -0.00022388054021300896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1144_text_document -0.00015633876287631285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1146_text_document -0.00015683128496399404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1147_text_document -0.0001498588984354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1148_text_document -0.00015466674094651695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1149_text_document -0.00015104328866230663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1150_text_document -0.0001510288850415886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1151_text_document -0.00015453329995596143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1152_text_document -0.0001717890160140908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1153_text_document -0.00016303689223488152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1154_text_document -0.00017438742884609578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1155_text_document -0.00017195307231868866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1156_text_document -0.00016630614911747752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1157_text_document -0.0001738954845222655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1158_text_document -0.00016759158755171884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1159_text_document -0.00017061259922452842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1160_text_document -0.00017196072417278202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1161_text_document -0.00016824585118656202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1162_text_document -0.00016301309236242047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1163_text_document -0.0001718575393991296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1164_text_document -0.00017003663826341565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1165_text_document -0.00017018328983305946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1166_text_document -0.00017218141861091656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1167_text_document -0.00016559619112054818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1168_text_document -0.00016284882257395627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1169_text_document -0.0001617104078870124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1170_text_document -0.00016849349395228177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1171_text_document -0.00016378319727916067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1172_text_document -0.00017114019486042634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1173_text_document -0.0001726823065513329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1174_text_document -0.00016244897469644304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1175_text_document -0.0001613681046473606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1176_text_document -0.00018118661924575096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1177_text_document -0.00016563345750593493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1178_text_document -0.00016790014898759615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1179_text_document -0.0001629142142864177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1180_text_document -0.00016191717527939525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1181_text_document -0.0001671004065869619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1182_text_document -0.0001675370141650324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1183_text_document -0.00016799445480682778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1184_text_document -0.0001719736620354862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1185_text_document -0.00016261057260474936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1186_text_document -0.00015865991174764644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1187_text_document -0.00015739800441831657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1188_text_document -0.00016171134746282626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1189_text_document -0.00016720238820009615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1190_text_document -0.00016497201020069133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1191_text_document -0.00016081080933342493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1192_text_document -0.0001598451415954535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1193_text_document -0.00016189725587725768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1194_text_document -0.00015376149407875128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1195_text_document -0.00015923032632387212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1196_text_document -0.000161420662154024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1197_text_document -0.00015926844960634996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1198_text_document -0.000156372807999939 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1199_text_document -0.00016050285429044874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1200_text_document -0.00015617925982472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1201_text_document -0.00016514079794945202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1202_text_document -0.00016522274070820443 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1203_text_document -0.0001597381170738336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1204_text_document -0.0001616744058690789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1205_text_document -0.00016029435854255644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1206_text_document -0.0001600416279503584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1207_text_document -0.0001607379715998696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1208_text_document -0.0001593514911283079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1209_text_document -0.00015864317782095664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1210_text_document -0.00015911735436385907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1211_text_document -0.0001556275795066712 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1212_text_document -0.0001656764173702947 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1213_text_document -0.00015679155524627255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1214_text_document -0.00016376988600479205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1215_text_document -0.0001581538165285075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1216_text_document -0.0001610240227045592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1217_text_document -0.00015776131940645536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1218_text_document -0.00015818231748846595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1219_text_document -0.0001625550897521123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1220_text_document -0.0001547371099180901 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1221_text_document -0.00015414283944531357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1222_text_document -0.00016266088273096592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1223_text_document -0.00016083169545961368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1224_text_document -0.0001573027086756309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1225_text_document -0.00015728313997935927 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1226_text_document -0.00016781226249248295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1227_text_document -0.00014976228995207784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1228_text_document -0.00015444629923379175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1229_text_document -0.00015203154472094758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1230_text_document -0.00015416974359531256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1231_text_document -0.00015545110214308707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1232_text_document -0.0001510309557116906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1233_text_document -0.000150151986610048 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1234_text_document -0.00014833490597173326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1235_text_document -0.00014730918386476007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1236_text_document -0.00014903663558472915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1237_text_document -0.00014834903218682616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1238_text_document -0.00015322537809196756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1239_text_document -0.0001511230642513134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1240_text_document -0.00015357591909403477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1241_text_document -0.00015295542934724653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1242_text_document -0.00015013958035919124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1243_text_document -0.00015023610122778707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1244_text_document -0.00014784318253583398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1245_text_document -0.00015065966876706016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1246_text_document -0.0001481405433493943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1247_text_document -0.00014721741369089534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1248_text_document -0.00014730057861393202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1249_text_document -0.00015235999841072513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1250_text_document -0.00014541040677624616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1251_text_document -0.00014639042630648248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1252_text_document -0.00015068532335773535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1253_text_document -0.00015516053357170532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1254_text_document -0.00014515004876336832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1255_text_document -0.0001488593805475465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1256_text_document -0.0001506759742452044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1257_text_document -0.0001429840653957083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1258_text_document -0.00014437998012654534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1259_text_document -0.0001428860592717282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1260_text_document -0.0001475220383855572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1261_text_document -0.00014640582972274082 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1262_text_document -0.0001505350968588391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1263_text_document -0.00014784485165882563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1264_text_document -0.00014770697193146622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1265_text_document -0.0001433464625266231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1266_text_document -0.00014139730694769496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1267_text_document -0.00014139435371307747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1268_text_document -0.00014164383589527758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1269_text_document -0.0001429075740030123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1270_text_document -0.00014605872692153072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1271_text_document -0.0001424796215298057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1272_text_document -0.00014112515203848743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1273_text_document -0.00014039188160335826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1274_text_document -0.00014502736267043328 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1275_text_document -0.00014184146815260007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1276_text_document -0.0001453216584479987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1277_text_document -0.00014226985746562565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1278_text_document -0.00013903471234323833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1279_text_document -0.00014633669945119654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1280_text_document -0.00015567823959834718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1281_text_document -0.00016711998145328748 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1282_text_document -0.00016716820782888765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1283_text_document -0.00016788189624042867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1284_text_document -0.00016762149528397544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1285_text_document -0.00016394982452183396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1286_text_document -0.00017499487929449305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1287_text_document -0.00017285598246362648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1288_text_document -0.0001813127546456402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1289_text_document -0.00016923644001919636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1290_text_document -0.00016671545149204298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1291_text_document -0.0001691584149978932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1292_text_document -0.00016279240063910965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1293_text_document -0.00016581675179191334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1294_text_document -0.00016709742151486606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1295_text_document -0.00016462921631835026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1296_text_document -0.0001635773235573904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1297_text_document -0.0001629499633321397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1298_text_document -0.00016244603775076793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1299_text_document -0.00016565874682941692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1300_text_document -0.00016704769334813707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1301_text_document -0.00016527793060668047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1302_text_document -0.0001614670182628741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1303_text_document -0.00016090321773766912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1304_text_document -0.00016205158644923216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1305_text_document -0.00016115649647745916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1306_text_document -0.00016750884342636079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1307_text_document -0.0001593023982303325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1308_text_document -0.00015894512446540672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1309_text_document -0.00016391499925658774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1310_text_document -0.0001615310219600013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1311_text_document -0.00016109142610140696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1312_text_document -0.0001622135071747606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1313_text_document -0.00016686311075489617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1314_text_document -0.00016322992039795453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1315_text_document -0.00015923727775344227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1316_text_document -0.00016528070219491 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1317_text_document -0.00016089805290891765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1318_text_document -0.00016142731643379644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1319_text_document -0.00016164621217780662 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1320_text_document -0.00015738061325748116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1321_text_document -0.0001591233926254462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1322_text_document -0.00016649327648776514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1323_text_document -0.00016299925243783037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1324_text_document -0.00016490993699004063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1325_text_document -0.0001589061309585213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1326_text_document -0.00015701373074415468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1327_text_document -0.00015755460137450403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1328_text_document -0.00016368403834230255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1329_text_document -0.0001619141257919363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1330_text_document -0.0002274793692927606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1331_text_document -0.0001567633247814788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1332_text_document -0.00022905033511751312 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1333_text_document -0.0001548301064518758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1334_text_document -0.000226605319945327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1335_text_document -0.00022667037674726058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1336_text_document -0.00022923961805784498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1337_text_document -0.00014906828549341607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1338_text_document -0.00015829222539969273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1339_text_document -0.0001509036911919305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1340_text_document -0.00022536653378252486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1341_text_document -0.00015104016760222197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1342_text_document -0.00015099364342110257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1343_text_document -0.00022777331115603203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1344_text_document -0.00021580582739619934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1345_text_document -0.0001492017484493636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1346_text_document -0.0002232038326367584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1347_text_document -0.0002173110715340058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1348_text_document -0.0002106853410947563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1349_text_document -0.00021523392953900664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1350_text_document -0.00021996424976477582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1351_text_document -0.00021735745725911482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1352_text_document -0.00014743618479981591 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1353_text_document -0.00021587099328468655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1354_text_document -0.00021669175360386172 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1355_text_document -0.00021667379282364665 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1356_text_document -0.0002192120523189847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1357_text_document -0.00021547193097844086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1358_text_document -0.00021621049112421326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1359_text_document -0.00021196265801039842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1360_text_document -0.00021115416894129982 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1361_text_document -0.00021548122875612305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1362_text_document -0.0002167839127379268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1363_text_document -0.00021388435981092266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1364_text_document -0.00021247309275187394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1365_text_document -0.00020865156988970925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1366_text_document -0.00021232420243985875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1367_text_document -0.00020288941772275403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1368_text_document -0.00020534370920083462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1369_text_document -0.00014906807620518648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1370_text_document -0.0002110153701227056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1371_text_document -0.00020709542453451886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1372_text_document -0.00020465988557797482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1373_text_document -0.000195974694790701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1374_text_document -0.0002006410964660873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1375_text_document -0.00020083864604468702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1376_text_document -0.00020640909562295756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1377_text_document -0.0002009390668809768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1378_text_document -0.00019660322090934407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1379_text_document -0.0002031382964736789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1380_text_document -0.00019629671755665872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1381_text_document -0.00019754174238439996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1382_text_document -0.0002056909946356413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1383_text_document -0.0001979138566098626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1384_text_document -0.0001932131948461709 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1385_text_document -0.00020416546879013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1386_text_document -0.00020460391232945065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1387_text_document -0.00019389888059130955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1388_text_document -0.00019783854863351214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1389_text_document -0.000200961415063147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1390_text_document -0.0001956818423121531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1391_text_document -0.00020637040765714317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1392_text_document -0.00020119793791085526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1393_text_document -0.0002019159752232148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1394_text_document -0.00020709690510066213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1395_text_document -0.00019733093804912572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1396_text_document -0.0001880608678579731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1397_text_document -0.0002016375431479316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1398_text_document -0.00019179791527764437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1399_text_document -0.00018506553224762644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1400_text_document -0.00019958850500821938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1401_text_document -0.0001963985599733761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1402_text_document -0.00019686962952391687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1403_text_document -0.00019466431453041557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1404_text_document -0.00019423474723069192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1405_text_document -0.00018645004940802463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1406_text_document -0.0001957563417646353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1407_text_document -0.00019567310057973193 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1408_text_document -0.00019820964060443815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1409_text_document -0.0001922448994056278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1410_text_document -0.00018809380854194413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1411_text_document -0.00019183325882742152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1412_text_document -0.00018979529371331087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1413_text_document -0.00018194205843788177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1414_text_document -0.000185326810832552 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1415_text_document -0.00018768967790659056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1416_text_document -8.238038512980449e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1417_text_document -7.037628876350043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1418_text_document -7.031761895460266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1419_text_document -6.852561440270574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1420_text_document -7.163053214543125e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1421_text_document -6.965337217248569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1422_text_document -7.217926984135532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1423_text_document -6.887448282655111e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1424_text_document -7.065036798913058e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1425_text_document -7.00112034634854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1426_text_document -7.22545398101735e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1427_text_document -7.06581990215903e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1428_text_document -8.704526082745054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1429_text_document -7.647895905010174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1430_text_document -6.667599117230014e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1431_text_document -9.175439580281598e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1432_text_document -6.936196694178977e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1433_text_document -6.994579003243415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1434_text_document -6.85501978720171e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1435_text_document -6.733846418731063e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1436_text_document -6.760126406073544e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1437_text_document -6.979877393600358e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1438_text_document -6.866399513844505e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1439_text_document -6.599386727589954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1440_text_document -7.022110351565428e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1441_text_document -6.889110495186351e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1442_text_document -7.249533430962498e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1443_text_document -7.061312850517899e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1444_text_document -7.026495137417699e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1445_text_document -7.053710208774785e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1446_text_document -7.079302654666706e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1447_text_document -7.142821385554296e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1448_text_document -6.884074447800683e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1449_text_document -6.775299728680366e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1450_text_document -6.935640081273007e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1451_text_document -7.071164131398859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1452_text_document -7.251697614402021e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1453_text_document -0.00012391766284956256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1454_text_document -6.876051279861284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1455_text_document -7.000563116437178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1456_text_document -7.021430732464126e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1457_text_document -7.320305084935923e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1458_text_document -6.93854906426365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1459_text_document -7.268065730933861e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1460_text_document -0.00015861223006440801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1461_text_document -0.00018316051674097559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1462_text_document -0.00018895109829526356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1463_text_document -0.0001814266629730391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1464_text_document -0.00019157095403912478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1465_text_document -0.0001816102282477865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1466_text_document -0.00018024098024327291 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1467_text_document -0.00018881439877582162 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1468_text_document -0.00018219752655961166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1469_text_document -0.00018211134990984607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1470_text_document -0.00018153994220173833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1471_text_document -0.00018033319169939537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1472_text_document -0.00017832883141386175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1473_text_document -0.00018126942359083546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1474_text_document -0.0001780495979719729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1475_text_document -0.00017938729404825616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1476_text_document -0.00017959022144586935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1477_text_document -0.00017639465376427234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1478_text_document -0.00017486395149030808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1479_text_document -0.0001808240528085561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1480_text_document -0.00017274905690967175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1481_text_document -0.00017486036693577152 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1482_text_document -0.00017698269026135074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1483_text_document -0.00016841434450034874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1484_text_document -0.00016863721573351308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1485_text_document -0.00017483420640067329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1486_text_document -0.0001732828833128397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1487_text_document -0.00017295290907496933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1488_text_document -0.00017133232046800912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1489_text_document -0.00016977827391836668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1490_text_document -0.0001740935435005184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1491_text_document -3.6341265268857285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/cc_en_tail-1492_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_peS2o.txt b/ALCF/data-lists/sunspot/data_file_list_peS2o.txt deleted file mode 100644 index dff1249cf5e..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_peS2o.txt +++ /dev/null @@ -1,26 +0,0 @@ -0.001258392312111664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0000_text_document -0.0012624388132337304 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0001_text_document -0.0012626279540316713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0002_text_document -0.0012611745647392154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0003_text_document -0.001261744080471196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0004_text_document -0.0012540395730196387 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0005_text_document -0.0012674794155994474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0006_text_document -0.0015181894311854882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0007_text_document -0.003353431842116585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0008_text_document -0.0033457523561418873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0009_text_document -0.0035360737173355393 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0010_text_document -0.0035328021064248917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0011_text_document -0.003537853575841124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0012_text_document -0.0033495442948704096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0013_text_document -0.0033515559656802623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0014_text_document -0.0035292441286648877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0015_text_document -0.0033472466636064995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0016_text_document -0.003347244907254542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0017_text_document -0.003361109976122766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0018_text_document -0.003527949940706846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0019_text_document -0.0033629959027952918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0020_text_document -0.003534363177394335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0021_text_document -0.0033534091101340303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0022_text_document -0.003362863367631581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0023_text_document -0.0035295619700253587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0024_text_document -0.00039091085286111746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/pes2o_v2-0025_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_reddit.txt b/ALCF/data-lists/sunspot/data_file_list_reddit.txt deleted file mode 100644 index 644d7170213..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_reddit.txt +++ /dev/null @@ -1,78 +0,0 @@ -0.0011541728836721287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0000_text_document -0.001194214065746794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0001_text_document -0.0012074645870644872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0002_text_document -0.0011669676257397446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0003_text_document -0.0011730429598479002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0004_text_document -0.0011829645036126979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0005_text_document -0.0011882713613863669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0006_text_document -0.0011409601969657492 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0007_text_document -0.0011370779956530767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0008_text_document -0.0011566277261230336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0009_text_document -0.0011377530435595722 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0010_text_document -0.0011422212106036002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0011_text_document -0.0011359648236479313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0012_text_document -0.0011422819300771266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0013_text_document -0.0011613422088431185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0014_text_document -0.001149222546698594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0015_text_document -0.0011520428345756523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0016_text_document -0.0011408015787470732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0017_text_document -0.001145413257179254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0018_text_document -0.0011543340882314167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0019_text_document -0.0011397083750923865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0020_text_document -0.001163788652940794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0021_text_document -0.0011441686420414542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0022_text_document -0.0011429505546541332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0023_text_document -0.00117471168582067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0024_text_document -0.0011456585273133617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0025_text_document -0.0011738665177335344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0026_text_document -0.0011646176186295262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0027_text_document -0.0011629386473461694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0028_text_document -0.0011421097688385183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0029_text_document -0.0011459477142114253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0030_text_document -0.0011756431096178663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0031_text_document -0.0011482680809577622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0032_text_document -0.0011445710176100962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0033_text_document -0.001142534803152167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0034_text_document -0.0011422043218494292 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0035_text_document -0.0011678344410475695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0036_text_document -0.0011562147470581413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0037_text_document -0.0011468122833549663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0038_text_document -0.0011532706690152916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0039_text_document -0.0011292882378850658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0040_text_document -0.0011300177059999066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0041_text_document -0.0011287171558685828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0042_text_document -0.0011295841562723513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0043_text_document -0.0011279954847952854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0044_text_document -0.0011283817109930107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0045_text_document -0.001128286479630481 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0046_text_document -0.0011276081740353844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0047_text_document -0.0011268985652144736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0048_text_document -0.0011261863340342809 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0049_text_document -0.0011248860240274238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0050_text_document -0.0011253794147731645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0051_text_document -0.0011242857628861397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0052_text_document -0.0011228472942657042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0053_text_document -0.00112269047698053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0054_text_document -0.0011234938283922757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0055_text_document -0.0011230927745087202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0056_text_document -0.0011247141749506225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0057_text_document -0.0011241207913742775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0058_text_document -0.0011220187728072355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0059_text_document -0.0011227320045060405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0060_text_document -0.0011217839100677303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0061_text_document -0.0011210875921360617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0062_text_document -0.0011221651716921029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0063_text_document -0.0011248396609954611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0064_text_document -0.0012275703827670792 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0065_text_document -0.0011056036331311184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0066_text_document -0.001107902944963784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0067_text_document -0.0010968114497626087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0068_text_document -0.0011027306309299484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0069_text_document -0.0010853624892717291 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0070_text_document -0.0011051858405711837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0071_text_document -0.0010808015771539223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0072_text_document -0.0010855928806935572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0073_text_document -0.0010442141182932184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0074_text_document -0.0011804749731815143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0075_text_document -0.0011670805522744465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0076_text_document -0.0008366052616529944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/reddit-v5-dedupe-pii-nsfw-toxic-0077_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_stack.txt b/ALCF/data-lists/sunspot/data_file_list_stack.txt deleted file mode 100644 index cbaf3cedde1..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_stack.txt +++ /dev/null @@ -1,149 +0,0 @@ -0.0010659025986423038 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0000_text_document -0.001089820700651703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0001_text_document -0.0010894690468995446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0002_text_document -0.0010893103153582777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0003_text_document -0.001092968830569157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0004_text_document -0.0010927822953669655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0005_text_document -0.0010948538530423937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0006_text_document -0.0010914947459084862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0007_text_document -0.0011531345061061805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0008_text_document -0.0009273732822541429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0009_text_document -0.0009298094568342398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0010_text_document -0.0009269985376241653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0011_text_document -0.0009299414467502114 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0012_text_document -0.0009281292496915194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0013_text_document -0.0009300797305068478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0014_text_document -0.0009575658299825903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0015_text_document -0.001124706364232967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0016_text_document -0.0011201757618238954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0017_text_document -0.001126433347327465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0018_text_document -0.0011299837668245817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0019_text_document -0.001127851225271931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0020_text_document -0.0011265589698280143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0021_text_document -0.0011227970380980016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0022_text_document -0.001131300918127052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0023_text_document -0.00112588381546472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0024_text_document -0.0011692456277892793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0025_text_document -0.0011330744556493294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0026_text_document -0.001041946972706877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0027_text_document -0.0010493121881969634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0028_text_document -0.0009912570469629923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0029_text_document -0.0012717963903526445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0030_text_document -0.0014051955824199262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0031_text_document -0.0011248653480876683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0032_text_document -0.0015096975127629315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0033_text_document -0.001056885183600456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0034_text_document -0.0010523010671513575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0035_text_document -0.001055691055690255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0036_text_document -0.0012434898779499373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0037_text_document -0.0009615620261395163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0038_text_document -0.0011689290747945063 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0039_text_document -0.0012610288149681123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0040_text_document -0.0012183045747008489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0041_text_document -0.0012232394891956877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0042_text_document -0.0012316862572191265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0043_text_document -0.001171858466558184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0044_text_document -0.0009288715082322405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0045_text_document -0.0009096255640660796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0046_text_document -0.0009098493089021282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0047_text_document -0.000908428701094243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0048_text_document -0.0009115948236386599 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0049_text_document -0.0009109761446993803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0050_text_document -0.0009097199236925156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0051_text_document -0.0009103946801923116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0052_text_document -0.0009109038594994949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0053_text_document -0.0009098133932243314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0054_text_document -0.0009111744494635876 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0055_text_document -0.0008961257268851344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0056_text_document -0.0008499219991848833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0057_text_document -0.000848817192629684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0058_text_document -0.0008469931268429987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0059_text_document -0.0008487804660301039 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0060_text_document -0.0008535293627452302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0061_text_document -0.0008508082359285502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0062_text_document -0.000847764423021283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0063_text_document -0.0008661814491784624 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0064_text_document -0.0012598427266996145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0065_text_document -0.0015411645064455006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0066_text_document -0.0015500690406153115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0067_text_document -0.0010431702414192465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0068_text_document -0.0010103298065465376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0069_text_document -0.0009173697763272889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0070_text_document -0.0009149081716719212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0071_text_document -0.0009223001515794829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0072_text_document -0.0009231205497115238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0073_text_document -0.0009205400022638854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0074_text_document -0.000921891356231865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0075_text_document -0.0009206550523916788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0076_text_document -0.000919101114727538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0077_text_document -0.0009189314293443922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0078_text_document -0.0009187845413397615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0079_text_document -0.0009212488966514148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0080_text_document -0.0009193937503280587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0081_text_document -0.0013803871878583557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0082_text_document -0.0009950213666737198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0083_text_document -0.000927893134699511 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0084_text_document -0.0009256115426841411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0085_text_document -0.0009245248815034989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0086_text_document -0.0009239324963431647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0087_text_document -0.00093017264782812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0088_text_document -0.0009246774971430524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0089_text_document -0.0009246651817682976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0090_text_document -0.0009220962135479767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0091_text_document -0.0009218191222144196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0092_text_document -0.0009271314108370893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0093_text_document -0.0011393174361636815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0094_text_document -0.0010056046636817732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0095_text_document -0.000985188940051775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0096_text_document -0.0009834908338499898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0097_text_document -0.0009841221104671695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0098_text_document -0.0009846688252964021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0099_text_document -0.0009846837273836892 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0100_text_document -0.000983200779763785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0101_text_document -0.000983626091844726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0102_text_document -0.0009227550215195058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0103_text_document -0.0008517634745985513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0104_text_document -0.0009820984183696825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0105_text_document -0.001062956613371643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0106_text_document -0.0009446580160861343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0107_text_document -0.000849273787178016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0108_text_document -0.0010838798124933814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0109_text_document -0.0016259767652594482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0110_text_document -0.0009261166233974987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0111_text_document -0.0013044836937627727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0112_text_document -0.0017111272224419217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0113_text_document -0.0017274616815008634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0114_text_document -0.0017204942871235126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0115_text_document -0.0017119592701771347 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0116_text_document -0.0016979912192342588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0117_text_document -0.001701886248500233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0118_text_document -0.0017227272126357288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0119_text_document -0.0017014517255794117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0120_text_document -0.0016995002579026628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0121_text_document -0.0016958447424626011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0122_text_document -0.0017111887981161064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0123_text_document -0.0017172926007805738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0124_text_document -0.0016938659465618113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0125_text_document -0.0016877576226485259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0126_text_document -0.0017144361080061983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0127_text_document -0.0017173753931755767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0128_text_document -0.001713308056226134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0129_text_document -0.0017126769067653286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0130_text_document -0.0017129095633438736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0131_text_document -0.001704961253905759 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0132_text_document -0.0009282082505873367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0133_text_document -0.0007973220067601047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0134_text_document -0.0008407445714413182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0135_text_document -0.0008403726198530843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0136_text_document -0.0008371632157580058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0137_text_document -0.0013060325919558903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0138_text_document -0.0014100060700040244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0139_text_document -0.0008750222172256031 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0140_text_document -0.0016918433420911735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0141_text_document -0.001838605753011377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0142_text_document -0.0016004536814984726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0143_text_document -0.0011738110086663097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0144_text_document -0.0011269892510041232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0145_text_document -0.0011251329530758676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0146_text_document -0.0011788404279377853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0147_text_document -0.0007876495048700586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/stack-v4-train-0148_text_document diff --git a/ALCF/data-lists/sunspot/data_file_list_wiki.txt b/ALCF/data-lists/sunspot/data_file_list_wiki.txt deleted file mode 100644 index 65169566eb0..00000000000 --- a/ALCF/data-lists/sunspot/data_file_list_wiki.txt +++ /dev/null @@ -1,2 +0,0 @@ -0.0035577638528123345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/en_simple_wiki_v0-0000_text_document -0.0018422361471876658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/en_simple_wiki_v0-0001_text_document diff --git a/ALCF/data-lists/sunspot/falcon.txt b/ALCF/data-lists/sunspot/falcon.txt new file mode 100644 index 00000000000..0b2fd6d43f7 --- /dev/null +++ b/ALCF/data-lists/sunspot/falcon.txt @@ -0,0 +1,501 @@ +0.0003547982093445404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document + diff --git a/ALCF/data-lists/sunspot/megawiki.txt b/ALCF/data-lists/sunspot/megawiki.txt new file mode 100644 index 00000000000..9fc9ca5dab8 --- /dev/null +++ b/ALCF/data-lists/sunspot/megawiki.txt @@ -0,0 +1,262 @@ +6.322825248625475e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document diff --git a/ALCF/data-lists/sunspot/open-web-math-train.txt b/ALCF/data-lists/sunspot/open-web-math-train.txt new file mode 100644 index 00000000000..b36e9977c05 --- /dev/null +++ b/ALCF/data-lists/sunspot/open-web-math-train.txt @@ -0,0 +1,13 @@ +0.001451215788905126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document diff --git a/ALCF/data-lists/sunspot/pes2o.txt b/ALCF/data-lists/sunspot/pes2o.txt new file mode 100644 index 00000000000..63f805c06d5 --- /dev/null +++ b/ALCF/data-lists/sunspot/pes2o.txt @@ -0,0 +1,26 @@ +0.0012499632072059553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document diff --git a/ALCF/data-lists/sunspot/reddit.txt b/ALCF/data-lists/sunspot/reddit.txt new file mode 100644 index 00000000000..59eafce1ee0 --- /dev/null +++ b/ALCF/data-lists/sunspot/reddit.txt @@ -0,0 +1,78 @@ +0.0005759963691850877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document diff --git a/ALCF/data-lists/sunspot/stack.txt b/ALCF/data-lists/sunspot/stack.txt new file mode 100644 index 00000000000..297783ac228 --- /dev/null +++ b/ALCF/data-lists/sunspot/stack.txt @@ -0,0 +1,26 @@ +0.0009994361338078242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document diff --git a/ALCF/data-lists/sunspot/starcoder.txt b/ALCF/data-lists/sunspot/starcoder.txt new file mode 100644 index 00000000000..37e6333de51 --- /dev/null +++ b/ALCF/data-lists/sunspot/starcoder.txt @@ -0,0 +1,50 @@ +0.004474659408857016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document + diff --git a/ALCF/data-lists/sunspot/tulu.txt b/ALCF/data-lists/sunspot/tulu.txt new file mode 100644 index 00000000000..2b758025015 --- /dev/null +++ b/ALCF/data-lists/sunspot/tulu.txt @@ -0,0 +1,66 @@ +0.00032927705604725614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document diff --git a/ALCF/data-lists/sunspot/wiki.txt b/ALCF/data-lists/sunspot/wiki.txt new file mode 100644 index 00000000000..52af00d57b5 --- /dev/null +++ b/ALCF/data-lists/sunspot/wiki.txt @@ -0,0 +1,2 @@ +0.003548077173506675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document From 3648af572d926278429f35045abfd72eb1d76a76 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 17:51:23 -0500 Subject: [PATCH 088/590] Add `ALCF/test_sunspot.sh` --- ALCF/test_sunspot.sh | 48 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+) create mode 100755 ALCF/test_sunspot.sh diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh new file mode 100755 index 00000000000..a8a4a21f32e --- /dev/null +++ b/ALCF/test_sunspot.sh @@ -0,0 +1,48 @@ +#!/bin/bash --login +# +# Run complete test of +# https://github.com/argonne-lcf/Megatron-DeepSpeed +# on Sunspot @ ALCF + +# EXIT ON ERROR(s) +set -euxo pipefail + +######################################################## +# Setup / activate conda environment, +# mine is called q4-drop +######################################################## +setup_conda() { + if [[ "${SHELL}" = "/bin/zsh" ]]; then + eval "$(~/miniconda3/bin/conda shell.zsh hook)" + else + eval "$(~/miniconda3/bin/conda shell.bash hook)" + fi + conda activate q4-drop +} + + +######################################## +# Make sure ./tmp/Megatron-DeepSpeed +# does not already exist +######################################## +setup_megatron_deepspeed() { + mkdir tmp && cd tmp + if [[ -d "Megatron-DeepSpeed" ]]; then + # rm -rfv Megatron-DeepSpeed/ + echo "Found existing Megatron-DeepSpeed. + Remove existing directory to run test." + exit + fi + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + git checkout remove-apex-deps +} + + +main() { + setup_conda + setup_megatron_deepspeed + # NOTE: to use OPT=adamwschedulefree, you will need to pip install schedulefree + DEBUG=1 PBS_O_WORKDIR="$(pwd)" DATA_FILE_LIST=./ALCF/data-lists/sunspot/books.txt LR=0.0008 GRAD_ACC_STEPS=8 ZERO_STAGE=1 NUM_LAYERS=10 MICRO_BATCH=8 OPT=adamwschedulefree TIMING_LOG_LEVEL=1 bash train_llama_alcf.sh +} + +main From 9796eacf3c7392e20ff03aa2ea594ef3eb46ecfd Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 19:09:33 -0500 Subject: [PATCH 089/590] Add `ALCF/data-lists/sirius/books.txt` --- ALCF/data-lists/sirius/books.txt | 3 +++ 1 file changed, 3 insertions(+) create mode 100644 ALCF/data-lists/sirius/books.txt diff --git a/ALCF/data-lists/sirius/books.txt b/ALCF/data-lists/sirius/books.txt new file mode 100644 index 00000000000..7567ba52279 --- /dev/null +++ b/ALCF/data-lists/sirius/books.txt @@ -0,0 +1,3 @@ +0.006 /lus/tegu/projects/PolarisAT/foremans/projects/argonne-lcf/Megatron-DeepSpeed/data/books-0000_text_document +0.006 /lus/tegu/projects/PolarisAT/foremans/projects/argonne-lcf/Megatron-DeepSpeed/data/books-0001_text_document +0.006 /lus/tegu/projects/PolarisAT/foremans/projects/argonne-lcf/Megatron-DeepSpeed/data/books-0002_text_document From 7b2ab6d984ddf4aaa380b8421ae39d1c9dec5d2a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 19:10:01 -0500 Subject: [PATCH 090/590] Add `ALCF/test_sirius.sh` --- ALCF/test_sirius.sh | 66 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 66 insertions(+) create mode 100755 ALCF/test_sirius.sh diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh new file mode 100755 index 00000000000..273d511afa2 --- /dev/null +++ b/ALCF/test_sirius.sh @@ -0,0 +1,66 @@ +#!/bin/bash --login +# +# Run complete test of +# https://github.com/argonne-lcf/Megatron-DeepSpeed +# on Sunspot @ ALCF + +# EXIT ON ERROR(s) +set -euxo pipefail + +NOW="$(date "+%Y-%m-%d-%H%M%S")" + +######################################################## +# Setup / activate conda environment, +# mine is called q4-drop +######################################################## +setup_conda() { + export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" + micromamba activate 2024-04-23 + # if [[ "${SHELL}" = "/bin/zsh" ]]; then + # eval "$(~/miniconda3/bin/conda shell.zsh hook)" + # else + # eval "$(~/miniconda3/bin/conda shell.bash hook)" + # fi + # conda activate q4-drop +} + + +######################################## +# Make sure ./tmp/Megatron-DeepSpeed +# does not already exist +######################################## +setup_megatron_deepspeed() { + OUTDIR="test-sirius-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" + echo "Running test in: ${OUTDIR}" + echo "WORKING DIRECTORY: $(realpath $(pwd .))" + if [[ -d "Megatron-DeepSpeed" ]]; then + # rm -rfv Megatron-DeepSpeed/ + echo "Found existing Megatron-DeepSpeed. + Remove existing directory to run test." + exit + fi + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + git checkout remove-apex-deps +} + + +main() { + setup_conda + setup_megatron_deepspeed + # NOTE: to use OPT=adamwschedulefree, you will need to pip install schedulefree + DEBUG=1 + PBS_O_WORKDIR="$(pwd)" + DATA_FILE_LIST=./ALCF/data-lists/sirius/books.txt + # LR=0.0008 + # GRAD_ACC_STEPS=8 + ZERO_STAGE=1 + NUM_LAYERS=10 + MICRO_BATCH=8 + # OPT=adamwschedulefree + TIMING_LOG_LEVEL=1 + bash train_llama_alcf.sh |& tee test-sirius.log +} + +main From 58cdccaa2438ac1da2a310652df3dc410e7a1138 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 19:17:30 -0500 Subject: [PATCH 091/590] Update `ALCF/test_sirius.sh` --- ALCF/test_sirius.sh | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-) diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh index 273d511afa2..6199c5c1571 100755 --- a/ALCF/test_sirius.sh +++ b/ALCF/test_sirius.sh @@ -2,7 +2,14 @@ # # Run complete test of # https://github.com/argonne-lcf/Megatron-DeepSpeed -# on Sunspot @ ALCF +# on Sirius @ ALCF +# to launch (inside an interactive `qsub -I` job) on Sirius: +# +# ```bash` +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed/ALCF +# $ bash test_sirius.sh +# ```` # EXIT ON ERROR(s) set -euxo pipefail @@ -18,12 +25,6 @@ setup_conda() { shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" micromamba activate 2024-04-23 - # if [[ "${SHELL}" = "/bin/zsh" ]]; then - # eval "$(~/miniconda3/bin/conda shell.zsh hook)" - # else - # eval "$(~/miniconda3/bin/conda shell.bash hook)" - # fi - # conda activate q4-drop } @@ -32,7 +33,7 @@ setup_conda() { # does not already exist ######################################## setup_megatron_deepspeed() { - OUTDIR="test-sirius-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" + OUTDIR="OUTPUTS/test-sirius-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" echo "Running test in: ${OUTDIR}" echo "WORKING DIRECTORY: $(realpath $(pwd .))" if [[ -d "Megatron-DeepSpeed" ]]; then @@ -49,18 +50,17 @@ setup_megatron_deepspeed() { main() { setup_conda setup_megatron_deepspeed - # NOTE: to use OPT=adamwschedulefree, you will need to pip install schedulefree - DEBUG=1 - PBS_O_WORKDIR="$(pwd)" - DATA_FILE_LIST=./ALCF/data-lists/sirius/books.txt + export DEBUG=1 + export PBS_O_WORKDIR="$(pwd)" + export DATA_FILE_LIST=./ALCF/data-lists/sirius/books.txt # LR=0.0008 # GRAD_ACC_STEPS=8 - ZERO_STAGE=1 - NUM_LAYERS=10 - MICRO_BATCH=8 - # OPT=adamwschedulefree - TIMING_LOG_LEVEL=1 - bash train_llama_alcf.sh |& tee test-sirius.log + export ZERO_STAGE=1 + export NUM_LAYERS=10 + export MICRO_BATCH=8 + export TRAIN_ITERS=20 + export TIMING_LOG_LEVEL=1 + bash train_llama_alcf.sh |& tee "test-sirius-${NOW}".log } main From 02a955c410429e8d1a39356b89ad4d1c191e88e5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 19:46:53 -0500 Subject: [PATCH 092/590] Create `alcf-tests` branch --- ALCF/test_sunspot.sh | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-) diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh index a8a4a21f32e..29ec46fc038 100755 --- a/ALCF/test_sunspot.sh +++ b/ALCF/test_sunspot.sh @@ -3,10 +3,19 @@ # Run complete test of # https://github.com/argonne-lcf/Megatron-DeepSpeed # on Sunspot @ ALCF +# to launch (inside an interactive `qsub -I` job) on Sirius: +# +# ```bash` +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed/ALCF +# $ bash test_sunspot.sh +# ```` # EXIT ON ERROR(s) set -euxo pipefail +NOW="$(date "+%Y-%m-%d-%H%M%S")" + ######################################################## # Setup / activate conda environment, # mine is called q4-drop @@ -26,9 +35,10 @@ setup_conda() { # does not already exist ######################################## setup_megatron_deepspeed() { - mkdir tmp && cd tmp + OUTDIR="OUTPUTS/test-sunspot-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" + echo "Running test in: ${OUTDIR}" + echo "WORKING DIRECTORY: $(realpath $(pwd .))" if [[ -d "Megatron-DeepSpeed" ]]; then - # rm -rfv Megatron-DeepSpeed/ echo "Found existing Megatron-DeepSpeed. Remove existing directory to run test." exit @@ -41,8 +51,15 @@ setup_megatron_deepspeed() { main() { setup_conda setup_megatron_deepspeed - # NOTE: to use OPT=adamwschedulefree, you will need to pip install schedulefree - DEBUG=1 PBS_O_WORKDIR="$(pwd)" DATA_FILE_LIST=./ALCF/data-lists/sunspot/books.txt LR=0.0008 GRAD_ACC_STEPS=8 ZERO_STAGE=1 NUM_LAYERS=10 MICRO_BATCH=8 OPT=adamwschedulefree TIMING_LOG_LEVEL=1 bash train_llama_alcf.sh + export DEBUG=1 + export PBS_O_WORKDIR="$(pwd)" + export DATA_FILE_LIST=./ALCF/data-lists/sunspot/books.txt + export ZERO_STAGE=1 + export NUM_LAYERS=10 + export MICRO_BATCH=8 + export TRAIN_ITER=20 + export TIMING_LOG_LEVEL=1 + bash train_llama_alcf.sh |& tee "test-suntpot-${NOW}".log } main From 23c953124d6025e9aef4eb7516c2d9dbe01169e2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 19:49:41 -0500 Subject: [PATCH 093/590] Update `ALCF/{test_sirius.sh,test_sunspot.sh}` --- ALCF/test_sirius.sh | 4 ++-- ALCF/test_sunspot.sh | 6 +++--- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh index 6199c5c1571..19cb9405dc3 100755 --- a/ALCF/test_sirius.sh +++ b/ALCF/test_sirius.sh @@ -43,7 +43,7 @@ setup_megatron_deepspeed() { exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed - git checkout remove-apex-deps + git checkout alcf-tests } @@ -58,7 +58,7 @@ main() { export ZERO_STAGE=1 export NUM_LAYERS=10 export MICRO_BATCH=8 - export TRAIN_ITERS=20 + export TRAIN_ITER=20 export TIMING_LOG_LEVEL=1 bash train_llama_alcf.sh |& tee "test-sirius-${NOW}".log } diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh index 29ec46fc038..ca3a59cc870 100755 --- a/ALCF/test_sunspot.sh +++ b/ALCF/test_sunspot.sh @@ -5,7 +5,7 @@ # on Sunspot @ ALCF # to launch (inside an interactive `qsub -I` job) on Sirius: # -# ```bash` +# ```bash # $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed # $ cd Megatron-DeepSpeed/ALCF # $ bash test_sunspot.sh @@ -44,7 +44,7 @@ setup_megatron_deepspeed() { exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed - git checkout remove-apex-deps + git checkout alcf-tests } @@ -59,7 +59,7 @@ main() { export MICRO_BATCH=8 export TRAIN_ITER=20 export TIMING_LOG_LEVEL=1 - bash train_llama_alcf.sh |& tee "test-suntpot-${NOW}".log + bash train_llama_alcf.sh |& tee "test-sunspot-${NOW}.log" } main From 5fff0af7bdaafebb75f713ad088edac3be3d2d60 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 20:08:57 -0500 Subject: [PATCH 094/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 65 ++++++++++++++++++++++++-------------------- 1 file changed, 36 insertions(+), 29 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 6530340c19b..fe903e04f7d 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -8,7 +8,7 @@ import math from functools import partial from megatron import get_args -from megatron import print_rank_0 +from megatron import log.info from megatron import get_timers from megatron import get_tokenizer from megatron.core import mpu, tensor_parallel @@ -74,7 +74,7 @@ def model_provider(pre_process=True, post_process=True): """Build the model.""" - print_rank_0('building GPT model ...') + log.info('building GPT model ...') see_memory_usage("Before Building Model", force=True) args = get_args() config = core_transformer_config_from_args(args) @@ -118,7 +118,7 @@ def model_provider(pre_process=True, post_process=True): # We need to call model.set_batch_fn after deepspeed.initialize model._megatron_batch_fn = get_batch_pipe - # Predompute the attention mask and store it in args. + # Precompute the attention mask and store it in args. # This avoids having to pipeline it # as an activation during training. # The mask is constant, and thus we can reuse it. @@ -154,12 +154,9 @@ def model_provider(pre_process=True, post_process=True): ) num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) - # print_rank_0('\n ------------------------ ') - # print_rank_0(f'num of parameters {num_params}') - # print_rank_0('------------------------\n ') - print_rank_0(80 * '-') - print_rank_0(f"Number of parameters in model: {num_params}") - print_rank_0(80 * '-') + log.info(80 * '-') + log.info(f"Number of parameters in model: {num_params}") + log.info(80 * '-') see_memory_usage("After Building Model", force=True) if wandb.run is not None: tbdir = args.tensorboard_dir @@ -342,7 +339,7 @@ def loss_func(loss_mask, moe_loss, mos_loss, output_tensor): 'moe loss': moe_loss, 'kd loss': mos_loss } - print_rank_0( + log.info( f'>>> total loss: {loss}, ' f'lm loss {averaged_loss[0]}, ' f'kd loss {mos_loss}' @@ -419,7 +416,8 @@ def forward_step(data_iterator, model): # Get the batch. timers('batch-generator', log_level=2).start() tokens, labels, loss_mask, attention_mask, position_ids = get_batch( - data_iterator) + data_iterator + ) timers('batch-generator').stop() if args.data_efficiency_curriculum_learning: @@ -492,11 +490,12 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): """Build train, valid, and test datasets.""" args = get_args() - print_rank_0('> building train, validation, and test datasets ' - 'for GPT ...') + log.info( + '> building train, validation, and test datasets for GPT ...' + ) files = [] if args.data_file_list is not None: - print_rank_0(f"Reading datasets from {args.data_file_list}") + log.info(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: for f in flist.readlines(): w, fname = f.split() @@ -523,8 +522,9 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): train_data_prefix=args.train_data_path, valid_data_prefix=args.valid_data_path, test_data_prefix=args.test_data_path, - data_cache_path=args.data_cache_path) - print_rank_0("> finished creating GPT datasets ...") + data_cache_path=args.data_cache_path, + ) + log.info("> finished creating GPT datasets ...") return train_ds, valid_ds, test_ds @@ -566,8 +566,6 @@ def git_ds_info(): def main(): - # if RANK == 0: - # setup_wandb() if os.getenv('TORCH_PROFILER_ENABLED') == '1': from torch.profiler import profile, record_function, ProfilerActivity with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA]) as prof: @@ -593,17 +591,26 @@ def main(): # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, data_post_process=data_post_process ) - try: - from megatron.text_generation import generate_and_post_process - with torch.autocast(device_type=ez.get_torch_device(), dtype=torch.float16): - response, _, _, _ = generate_and_post_process(model, prompts=["Hello world", "Nature is", "Turing test comprises", "Explain solar eclipse"], tokens_to_generate=32) - if RANK == 0: - log.info(f'generation completed..\n response:{response}') - except ValueError as ve: - log.critical(f'ValueError: {ve}') - pass + # try: + # from megatron.text_generation import generate_and_post_process + # with torch.autocast(device_type=DEVICE, dtype=args.dtype): + # response, _, _, _ = generate_and_post_process( + # model, + # prompts=[ + # "Hello world", + # "Nature is", + # "Turing test comprises", + # "Explain solar eclipse" + # ], + # tokens_to_generate=32 + # ) + # if RANK == 0: + # log.info(f'generation completed..\n response:{response}') + # except ValueError as ve: + # log.critical(f'ValueError: {ve}') + # pass # dist.barrier() - model.train() + # model.train() return model @@ -623,4 +630,4 @@ def main(): print(f"wandb.run.name: {wandb.run.name}") print(f"wandb.run.url: {wandb.run.url}") wandb.finish() - sys.exit() + sys.exit(0) From 005272b586b08e88619b4c7e341bf7951ae563d3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 20:10:25 -0500 Subject: [PATCH 095/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index fe903e04f7d..7ed38614a71 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -8,7 +8,7 @@ import math from functools import partial from megatron import get_args -from megatron import log.info +# from megatron import print_rank_0 from megatron import get_timers from megatron import get_tokenizer from megatron.core import mpu, tensor_parallel @@ -19,12 +19,12 @@ from megatron.utils import get_ltor_masks_and_position_ids from megatron.utils import average_losses_across_data_parallel_group, update_rotary_pos_emb from megatron.arguments import core_transformer_config_from_args -from megatron.utils import ( - report_memory, - throughput_calculator, - checkpoint_throughput_calculator -) -from pathlib import Path +# from megatron.utils import ( +# # report_memory, +# # throughput_calculator, +# # checkpoint_throughput_calculator +# ) +# from pathlib import Path from enrich import get_logger import deepspeed @@ -33,7 +33,7 @@ import subprocess import wandb -import time +# import time from torch import nn import torch.nn.functional as F import ezpz as ez From fdb17075f0a84f0ca61bb76e6bb4d3d37087b649 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 20:58:38 -0500 Subject: [PATCH 096/590] Remove `ds_report` from `train_llama_alcf.sh` --- train_llama_alcf.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 4aac1153c7d..cd2d8213dd9 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -132,7 +132,7 @@ run_cmd=" # ds_exec # echo "! Using $(which deepspeed)" -ds_report +# ds_report echo "${run_cmd}" From 936c423355d36102b51c9f885079952e15a856b8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 21:17:41 -0500 Subject: [PATCH 097/590] Update `.gitignore` --- .gitignore | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) diff --git a/.gitignore b/.gitignore index 3e46cef4c56..edbde3c246e 100644 --- a/.gitignore +++ b/.gitignore @@ -1,3 +1,33 @@ +# User Added +*tmp* +*core.* +*old* +*.bak +**index-cache** +**pbslogs** +ezpz +*.o17* +*.e17* +*hostfile* +.deepspeed_env +*.DS_Store +old/* +**venv** +*.json +*.o1 +*.e1 +outputs/ +venvs/ +wandb/ +llama-logs/ +checkpoints/ +*.gz +*.txt +*.idx +*.bin +*.log +__pycache__ + .deepspeed_env *.bak .cache/* From a59a532b332b788d26919d94c6470a8593314775 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Apr 2024 21:39:34 -0500 Subject: [PATCH 098/590] Update `ALCF/test_{sunspot,sirius}.sh` --- ALCF/test_sirius.sh | 8 ++------ ALCF/test_sunspot.sh | 5 ++--- 2 files changed, 4 insertions(+), 9 deletions(-) diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh index 19cb9405dc3..108649612c4 100755 --- a/ALCF/test_sirius.sh +++ b/ALCF/test_sirius.sh @@ -37,13 +37,11 @@ setup_megatron_deepspeed() { echo "Running test in: ${OUTDIR}" echo "WORKING DIRECTORY: $(realpath $(pwd .))" if [[ -d "Megatron-DeepSpeed" ]]; then - # rm -rfv Megatron-DeepSpeed/ - echo "Found existing Megatron-DeepSpeed. - Remove existing directory to run test." + echo "Found existing Megatron-DeepSpeed in ${OUTDIR}" + echo "Remove Megatron-DeepSpeed from ${OUTDIR} to run test." exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed - git checkout alcf-tests } @@ -53,8 +51,6 @@ main() { export DEBUG=1 export PBS_O_WORKDIR="$(pwd)" export DATA_FILE_LIST=./ALCF/data-lists/sirius/books.txt - # LR=0.0008 - # GRAD_ACC_STEPS=8 export ZERO_STAGE=1 export NUM_LAYERS=10 export MICRO_BATCH=8 diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh index ca3a59cc870..67f6868d435 100755 --- a/ALCF/test_sunspot.sh +++ b/ALCF/test_sunspot.sh @@ -39,12 +39,11 @@ setup_megatron_deepspeed() { echo "Running test in: ${OUTDIR}" echo "WORKING DIRECTORY: $(realpath $(pwd .))" if [[ -d "Megatron-DeepSpeed" ]]; then - echo "Found existing Megatron-DeepSpeed. - Remove existing directory to run test." + echo "Found existing Megatron-DeepSpeed in ${OUTDIR}" + echo "Remove Megatron-DeepSpeed from ${OUTDIR} to run test." exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed - git checkout alcf-tests } From c9c87d9ff40d013e93dd623ed68cada4086cce9f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 10:53:29 -0500 Subject: [PATCH 099/590] Update `ALCF/data-lists/polaris/*.txt` --- ALCF/data-lists/polaris/algebraic.txt | 16 + ALCF/data-lists/polaris/arxiv.txt | 100 + ALCF/data-lists/polaris/books.txt | 3 + ALCF/data-lists/polaris/c4.txt | 171 + ALCF/data-lists/polaris/cc.txt | 1108 ++++ .../polaris/data_file_list_books.txt | 3 - ALCF/data-lists/polaris/data_file_list_c4.txt | 86 - ALCF/data-lists/polaris/data_file_list_cc.txt | 2878 ----------- .../polaris/data_file_list_peS2o.txt | 42 - .../polaris/data_file_list_stack.txt | 4435 ----------------- .../polaris/data_file_list_wiki.txt | 2 - .../polaris/dolma_v1_7_file_list.txt | 2419 +++++++++ ALCF/data-lists/polaris/falcon.txt | 501 ++ ALCF/data-lists/polaris/megawiki.txt | 262 + .../polaris/open-web-math-train.txt | 13 + ALCF/data-lists/polaris/pes2o.txt | 26 + ALCF/data-lists/polaris/reddit.txt | 78 + ALCF/data-lists/polaris/stack.txt | 26 + ALCF/data-lists/polaris/starcoder.txt | 50 + ALCF/data-lists/polaris/tulu.txt | 66 + ALCF/data-lists/polaris/wiki.txt | 2 + 21 files changed, 4841 insertions(+), 7446 deletions(-) create mode 100644 ALCF/data-lists/polaris/algebraic.txt create mode 100644 ALCF/data-lists/polaris/arxiv.txt create mode 100644 ALCF/data-lists/polaris/books.txt create mode 100644 ALCF/data-lists/polaris/c4.txt create mode 100644 ALCF/data-lists/polaris/cc.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_books.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_c4.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_cc.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_peS2o.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_stack.txt delete mode 100644 ALCF/data-lists/polaris/data_file_list_wiki.txt create mode 100644 ALCF/data-lists/polaris/dolma_v1_7_file_list.txt create mode 100644 ALCF/data-lists/polaris/falcon.txt create mode 100644 ALCF/data-lists/polaris/megawiki.txt create mode 100644 ALCF/data-lists/polaris/open-web-math-train.txt create mode 100644 ALCF/data-lists/polaris/pes2o.txt create mode 100644 ALCF/data-lists/polaris/reddit.txt create mode 100644 ALCF/data-lists/polaris/stack.txt create mode 100644 ALCF/data-lists/polaris/starcoder.txt create mode 100644 ALCF/data-lists/polaris/tulu.txt create mode 100644 ALCF/data-lists/polaris/wiki.txt diff --git a/ALCF/data-lists/polaris/algebraic.txt b/ALCF/data-lists/polaris/algebraic.txt new file mode 100644 index 00000000000..505276d3bfb --- /dev/null +++ b/ALCF/data-lists/polaris/algebraic.txt @@ -0,0 +1,16 @@ +0.0018520780893211373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document diff --git a/ALCF/data-lists/polaris/arxiv.txt b/ALCF/data-lists/polaris/arxiv.txt new file mode 100644 index 00000000000..cae6e2da693 --- /dev/null +++ b/ALCF/data-lists/polaris/arxiv.txt @@ -0,0 +1,100 @@ +0.0002583902668716813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document diff --git a/ALCF/data-lists/polaris/books.txt b/ALCF/data-lists/polaris/books.txt new file mode 100644 index 00000000000..195aca5339f --- /dev/null +++ b/ALCF/data-lists/polaris/books.txt @@ -0,0 +1,3 @@ +0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document diff --git a/ALCF/data-lists/polaris/c4.txt b/ALCF/data-lists/polaris/c4.txt new file mode 100644 index 00000000000..833b0958829 --- /dev/null +++ b/ALCF/data-lists/polaris/c4.txt @@ -0,0 +1,171 @@ +0.0002406272620255565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document diff --git a/ALCF/data-lists/polaris/cc.txt b/ALCF/data-lists/polaris/cc.txt new file mode 100644 index 00000000000..edf6aab8c1c --- /dev/null +++ b/ALCF/data-lists/polaris/cc.txt @@ -0,0 +1,1108 @@ +0.0003742481815405742 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document + diff --git a/ALCF/data-lists/polaris/data_file_list_books.txt b/ALCF/data-lists/polaris/data_file_list_books.txt deleted file mode 100644 index 18109f946d5..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_books.txt +++ /dev/null @@ -1,3 +0,0 @@ -0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0001_text_document -0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0000_text_document -0.006 /eagle/datasets//dolma/data_Llama2Tokenizer/gutenberg-books//books-0002_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_c4.txt b/ALCF/data-lists/polaris/data_file_list_c4.txt deleted file mode 100644 index 9ff6f90ff92..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_c4.txt +++ /dev/null @@ -1,86 +0,0 @@ -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0012_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0001_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0073_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0045_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0084_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0065_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0032_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0085_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0064_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0025_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0042_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0055_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0023_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0028_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0036_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0037_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0016_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0000_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0019_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0046_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0059_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0017_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0072_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0033_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0006_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0061_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0071_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0057_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0011_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0047_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0004_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0009_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0070_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0018_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0054_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0049_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0003_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0021_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0051_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0074_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0027_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0050_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0079_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0022_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0030_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0034_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0020_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0035_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0015_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0066_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0044_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0010_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0002_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0041_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0067_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0048_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0013_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0083_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0053_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0008_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0014_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0069_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0056_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0062_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0031_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0007_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0077_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0058_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0076_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0078_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0005_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0081_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0040_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0068_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0075_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0063_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0029_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0039_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0026_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0052_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0024_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0043_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0060_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0082_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0080_text_document -0.0875 /eagle/datasets//dolma/data_Llama2Tokenizer/c4//c4-0038_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_cc.txt b/ALCF/data-lists/polaris/data_file_list_cc.txt deleted file mode 100644 index 4a20a99b89c..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_cc.txt +++ /dev/null @@ -1,2878 +0,0 @@ -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0553_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0299_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0366_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0753_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0429_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0372_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0124_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0437_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0053_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0615_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0182_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0713_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0688_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0166_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0768_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0692_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0041_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0416_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0630_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0639_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0225_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0035_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0365_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0368_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0196_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0328_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0624_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0081_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0488_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0189_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0118_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0150_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0314_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0209_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0229_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0265_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0532_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0478_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0140_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0256_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0047_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0607_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0023_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0111_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0613_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0748_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0000_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0127_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0106_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0563_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0577_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0502_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0705_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0538_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0088_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0263_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0460_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0571_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0653_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0172_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0524_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0652_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0322_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0447_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0387_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0612_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0290_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0339_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0487_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0396_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0178_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0091_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0193_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0408_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0496_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0755_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0773_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0547_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0384_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0574_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0533_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0464_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0489_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0050_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0060_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0114_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0033_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0561_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0208_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0233_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0744_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0326_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0313_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0482_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0436_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0588_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0080_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0660_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0038_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0282_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0745_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0406_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0116_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0059_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0503_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0357_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0171_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0770_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0286_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0544_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0698_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0155_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0341_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0463_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0676_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0595_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0174_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0198_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0480_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0687_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0145_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0004_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0583_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0449_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0204_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0715_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0521_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0320_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0568_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0151_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0197_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0709_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0499_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0006_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0269_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0525_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0413_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0656_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0646_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0246_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0535_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0333_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0238_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0241_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0469_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0689_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0403_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0404_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0360_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0191_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0236_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0032_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0445_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0614_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0490_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0651_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0703_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0702_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0623_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0719_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0728_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0031_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0253_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0551_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0327_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0027_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0491_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0395_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0473_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0662_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0312_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0605_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0455_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0580_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0005_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0311_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0305_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0260_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0566_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0670_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0129_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0742_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0549_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0058_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0501_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0071_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0450_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0375_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0131_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0697_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0415_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0560_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0643_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0699_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0515_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0739_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0092_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0046_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0083_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0443_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0746_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0655_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0427_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0603_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0367_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0318_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0520_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0749_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0771_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0369_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0434_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0602_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0349_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0763_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0731_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0338_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0462_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0347_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0649_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0194_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0134_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0734_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0632_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0280_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0184_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0089_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0095_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0555_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0016_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0168_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0665_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0767_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0666_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0737_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0037_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0648_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0064_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0764_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0323_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0009_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0545_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0212_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0015_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0267_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0727_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0661_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0211_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0220_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0278_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0721_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0718_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0207_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0619_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0400_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0754_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0610_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0358_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0758_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0298_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0756_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0729_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0468_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0397_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0247_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0149_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0119_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0010_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0093_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0386_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0045_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0066_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0393_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0600_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0440_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0350_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0214_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0714_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0161_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0775_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0203_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0077_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0332_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0700_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0123_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0024_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0013_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0587_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0148_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0513_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0674_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0188_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0599_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0158_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0425_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0003_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0534_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0254_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0121_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0099_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0373_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0479_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0379_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0344_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0684_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0720_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0391_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0575_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0319_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0336_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0531_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0474_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0432_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0766_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0342_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0476_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0237_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0061_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0250_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0752_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0329_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0376_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0640_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0634_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0682_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0181_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0076_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0244_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0690_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0303_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0228_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0477_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0224_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0199_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0343_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0399_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0707_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0760_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0774_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0270_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0144_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0451_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0180_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0025_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0363_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0516_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0647_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0581_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0679_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0635_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0201_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0133_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0351_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0325_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0183_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0287_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0683_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0316_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0275_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0424_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0461_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0576_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0390_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0052_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0086_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0492_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0216_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0772_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0439_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0249_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0493_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0593_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0442_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0218_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0484_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0346_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0157_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0352_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0441_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0486_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0537_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0485_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0164_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0022_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0458_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0497_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0170_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0154_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0751_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0048_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0428_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0418_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0112_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0757_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0421_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0471_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0510_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0466_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0641_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0601_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0740_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0594_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0276_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0383_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0232_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0717_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0644_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0518_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0743_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0673_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0044_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0667_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0308_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0675_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0572_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0579_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0723_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0381_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0759_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0504_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0708_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0049_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0642_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0074_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0039_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0401_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0409_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0014_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0098_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0146_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0616_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0101_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0446_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0565_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0295_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0730_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0498_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0638_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0301_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0139_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0192_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0001_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0268_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0527_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0359_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0315_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0251_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0546_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0262_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0659_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0567_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0190_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0078_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0175_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0054_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0008_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0452_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0187_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0011_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0138_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0087_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0206_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0611_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0509_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0205_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0620_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0677_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0132_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0296_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0495_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0444_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0598_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0691_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0761_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0417_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0317_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0578_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0374_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0055_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0481_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0307_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0736_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0136_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0550_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0084_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0511_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0380_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0356_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0310_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0110_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0668_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0306_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0115_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0324_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0202_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0294_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0704_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0629_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0608_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0627_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0725_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0472_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0230_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0407_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0556_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0505_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0040_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0606_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0096_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0281_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0179_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0557_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0288_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0769_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0370_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0017_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0694_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0385_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0130_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0562_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0506_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0036_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0217_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0289_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0712_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0724_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0564_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0105_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0120_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0141_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0431_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0142_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0570_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0512_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0227_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0411_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0389_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0735_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0585_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0122_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0042_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0309_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0765_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0636_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0539_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0467_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0586_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0750_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0200_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0671_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0530_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0012_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0082_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0160_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0438_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0195_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0185_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0215_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0173_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0710_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0348_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0590_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0073_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0030_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0079_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0072_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0019_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0239_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0410_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0453_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0543_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0007_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0733_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0125_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0569_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0331_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0062_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0043_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0433_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0235_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0448_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0696_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0559_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0392_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0664_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0483_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0591_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0272_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0271_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0340_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0159_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0153_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0541_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0028_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0067_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0222_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0165_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0117_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0669_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0103_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0258_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0097_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0419_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0459_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0609_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0104_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0430_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0582_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0457_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0529_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0029_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0426_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0279_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0596_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0631_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0517_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0507_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0252_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0626_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0056_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0335_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0542_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0672_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0284_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0257_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0654_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0678_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0018_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0528_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0422_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0068_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0245_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0617_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0255_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0176_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0732_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0221_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0371_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0137_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0177_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0382_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0291_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0597_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0321_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0292_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0273_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0716_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0094_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0026_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0021_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0242_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0226_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0100_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0519_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0552_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0057_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0776_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0109_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0219_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0706_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0508_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0722_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0456_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0633_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0618_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0354_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0514_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0475_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0034_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0300_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0454_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0693_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0658_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0304_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0090_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0548_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0063_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0465_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0414_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0169_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0762_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0701_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0135_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0143_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0167_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0523_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0540_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0334_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0297_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0738_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0500_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0554_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0695_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0747_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0108_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0113_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0686_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0302_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0362_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0355_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0645_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0102_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0681_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0536_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0156_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0163_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0277_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0573_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0293_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0741_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0377_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0423_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0592_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0657_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0584_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0628_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0240_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0361_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0680_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0388_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0435_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0345_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0637_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0420_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0378_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0398_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0234_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0266_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0210_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0711_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0162_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0264_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0070_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0223_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0470_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0002_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0069_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0625_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0285_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0065_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0147_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0085_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0394_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0330_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0621_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0248_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0274_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0526_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0685_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0589_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0128_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0405_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0126_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0075_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0494_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0283_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0558_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0402_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0261_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0107_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0522_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0604_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0231_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0663_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0726_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0152_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0353_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0243_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0364_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0213_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0412_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0622_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0650_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0354_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0200_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0391_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0559_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0473_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0528_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0073_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0170_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0180_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0582_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0067_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0036_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0023_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0099_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0342_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0353_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0044_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0575_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0107_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0566_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0101_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0542_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0217_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0437_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0583_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0172_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0060_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0072_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0393_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0414_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0355_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0082_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0281_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0051_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0133_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0471_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0346_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0034_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0300_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0282_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0309_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0125_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0434_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0460_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0186_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0504_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0487_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0132_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0225_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0401_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0477_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0607_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0562_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0497_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0174_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0569_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0591_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0110_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0455_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0166_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0453_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0581_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0341_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0544_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0606_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0291_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0086_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0578_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0083_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0357_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0188_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0411_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0218_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0003_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0001_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0543_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0307_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0169_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0149_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0561_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0310_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0222_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0547_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0183_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0214_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0111_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0127_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0054_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0007_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0572_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0103_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0334_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0114_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0513_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0237_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0397_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0306_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0120_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0129_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0262_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0459_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0085_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0206_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0271_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0610_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0031_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0390_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0043_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0012_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0409_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0574_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0596_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0588_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0532_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0236_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0501_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0269_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0540_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0175_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0290_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0233_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0440_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0071_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0037_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0420_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0404_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0141_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0592_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0164_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0162_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0388_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0159_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0372_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0476_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0323_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0008_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0151_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0533_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0344_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0481_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0204_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0179_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0496_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0469_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0055_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0367_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0277_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0603_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0512_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0340_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0143_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0140_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0285_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0124_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0531_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0375_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0013_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0522_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0066_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0332_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0228_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0445_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0430_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0018_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0392_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0505_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0485_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0130_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0026_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0489_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0006_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0157_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0467_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0454_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0600_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0163_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0248_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0339_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0534_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0038_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0597_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0303_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0425_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0000_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0352_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0226_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0461_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0545_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0102_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0587_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0048_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0336_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0502_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0427_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0090_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0369_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0216_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0292_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0243_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0326_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0602_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0611_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0499_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0032_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0599_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0097_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0182_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0378_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0509_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0139_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0456_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0322_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0221_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0153_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0076_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0057_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0284_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0075_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0422_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0288_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0077_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0305_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0273_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0242_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0050_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0525_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0232_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0173_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0294_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0016_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0365_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0604_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0138_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0178_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0377_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0042_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0065_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0112_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0142_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0184_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0495_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0276_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0301_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0210_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0494_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0447_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0333_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0424_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0224_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0105_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0081_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0579_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0536_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0168_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0293_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0021_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0230_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0260_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0450_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0465_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0394_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0319_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0028_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0608_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0538_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0155_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0517_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0240_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0515_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0158_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0209_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0321_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0296_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0576_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0080_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0091_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0259_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0121_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0580_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0595_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0062_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0436_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0337_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0059_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0115_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0412_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0462_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0106_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0423_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0366_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0215_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0263_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0295_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0443_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0557_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0010_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0136_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0376_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0235_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0135_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0144_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0548_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0537_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0128_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0287_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0194_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0069_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0324_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0364_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0187_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0096_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0558_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0063_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0551_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0286_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0449_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0255_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0358_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0383_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0283_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0470_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0403_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0349_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0524_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0426_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0486_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0519_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0380_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0387_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0122_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0518_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0554_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0027_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0418_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0104_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0039_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0268_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0201_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0094_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0347_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0416_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0514_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0231_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0330_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0523_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0570_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0421_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0009_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0478_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0417_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0482_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0213_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0568_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0312_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0550_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0219_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0093_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0530_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0035_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0356_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0406_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0498_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0297_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0266_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0428_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0074_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0244_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0431_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0410_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0134_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0246_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0408_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0563_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0360_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0468_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0402_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0370_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0148_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0092_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0089_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0251_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0061_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0609_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0539_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0475_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0261_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0304_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0203_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0084_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0395_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0131_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0197_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0087_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0327_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0279_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0196_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0555_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0041_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0317_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0193_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0441_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0202_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0511_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0220_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0474_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0577_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0014_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0171_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0361_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0264_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0024_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0432_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0451_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0552_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0584_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0239_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0458_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0407_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0045_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0541_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0371_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0480_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0590_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0015_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0108_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0320_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0145_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0483_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0521_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0419_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0150_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0526_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0589_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0119_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0315_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0546_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0510_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0373_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0413_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0249_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0484_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0493_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0491_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0385_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0005_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0374_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0185_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0345_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0571_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0167_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0234_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0318_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0520_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0256_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0116_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0088_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0556_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0302_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0238_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0205_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0019_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0191_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0199_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0078_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0594_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0195_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0030_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0439_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0448_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0350_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0267_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0275_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0348_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0560_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0181_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0329_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0516_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0363_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0258_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0359_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0299_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0457_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0379_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0049_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0368_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0265_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0046_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0311_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0177_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0058_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0040_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0549_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0605_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0160_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0472_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0020_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0553_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0211_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0052_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0466_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0382_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0351_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0433_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0270_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0593_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0529_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0095_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0308_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0152_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0064_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0189_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0527_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0070_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0400_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0257_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0229_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0154_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0362_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0573_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0161_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0252_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0386_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0280_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0585_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0464_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0338_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0278_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0506_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0033_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0137_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0444_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0247_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0109_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0004_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0011_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0508_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0126_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0017_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0254_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0567_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0100_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0398_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0117_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0147_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0176_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0156_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0490_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0022_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0190_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0047_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0207_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0446_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0227_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0435_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0289_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0146_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0598_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0503_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0165_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0208_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0025_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0442_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0325_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0429_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0056_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0212_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0002_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0245_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0068_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0381_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0452_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0500_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0396_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0328_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0384_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0389_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0586_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0488_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0298_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0463_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0118_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0479_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0274_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0272_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0492_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0113_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0415_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0405_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0198_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0313_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0331_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0314_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0053_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0507_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0438_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0029_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0399_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0564_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0601_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0241_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0343_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0223_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0316_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0123_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0535_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0250_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0335_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0253_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0079_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0565_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0192_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_head/cc_en_head-0098_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0247_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1166_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1192_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0818_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0166_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0529_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1356_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0858_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0823_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1339_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0627_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1370_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0341_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0185_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1259_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0981_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0515_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0545_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1167_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0077_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0665_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1195_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0085_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0566_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0230_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0311_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0055_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0810_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1270_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0966_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0517_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0843_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0348_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0797_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0736_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0943_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1054_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1105_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0556_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0849_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0492_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1121_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0817_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0874_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0579_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1250_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0146_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0589_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0169_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1084_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1041_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0526_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0551_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0193_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1402_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1106_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0782_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0659_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0588_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0990_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0833_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0845_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1098_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0402_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0878_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0930_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0046_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0440_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1293_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0393_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0049_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0305_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0868_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1126_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0531_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1087_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1442_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0997_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0366_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0165_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1078_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0957_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1002_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0269_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0460_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1397_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0250_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0951_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1246_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0876_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0302_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0564_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0584_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0622_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0694_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0335_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1189_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0215_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1390_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0204_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1038_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0713_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0567_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1130_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0221_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0538_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1232_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1265_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0628_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0090_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0968_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0248_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0885_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0977_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0749_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0527_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0985_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0934_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0993_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0088_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0674_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0171_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1160_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0640_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1419_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0488_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0704_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0887_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1055_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1258_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0924_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0390_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0612_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1276_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0744_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0399_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0053_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0025_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0371_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0161_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1444_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0051_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0367_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0036_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1398_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0434_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0965_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1226_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0093_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0702_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0949_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1343_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1480_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0503_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0979_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1039_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1034_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0725_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1185_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1288_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0742_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0242_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1071_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0975_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1211_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0435_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0841_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1303_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1380_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0558_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0522_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0181_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0098_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0318_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1101_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1183_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0054_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0962_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0072_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1283_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1269_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0379_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0109_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1257_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0683_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0932_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0724_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0259_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0752_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0748_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1456_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0038_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1311_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0224_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1181_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0292_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0006_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1077_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1376_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0082_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1107_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1305_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1263_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1484_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0470_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0232_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0312_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1478_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0176_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0825_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0587_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0726_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1069_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0478_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1470_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0751_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0034_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0557_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0087_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0554_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0680_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0095_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1193_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1290_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0956_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1433_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1409_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0546_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1137_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1042_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1361_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0983_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1099_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0675_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0000_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0540_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0593_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1182_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0984_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0463_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0406_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0963_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1112_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0572_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0894_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0901_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0539_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1462_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0504_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1196_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0916_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0840_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1059_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0895_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1439_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0521_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0227_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0157_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0134_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0091_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0080_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1405_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0856_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0355_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0904_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1256_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0886_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0703_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0870_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1307_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1463_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1278_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1176_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0999_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0352_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1216_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0506_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0942_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0164_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0071_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0362_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0285_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0819_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1447_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0794_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1111_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0452_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1008_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1064_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0786_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0275_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1004_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1020_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1198_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0258_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0179_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0473_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0706_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0829_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1348_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0489_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0721_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1418_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0657_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0182_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1268_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0677_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0211_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1199_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1161_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1159_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1060_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0203_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0634_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0214_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1292_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0168_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1342_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0105_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1346_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0145_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0174_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1393_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1295_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0888_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0673_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1412_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1337_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1089_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1031_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0542_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0135_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0604_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0905_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1028_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0760_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0052_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0361_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0408_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0514_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1144_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0086_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0016_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0363_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0026_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0013_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0048_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0343_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0991_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1151_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0325_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0209_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1325_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0042_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0162_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0902_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0625_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0482_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0502_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0225_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1190_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0498_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0198_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1146_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1197_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0889_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0873_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1013_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1212_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0107_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0005_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0528_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0101_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0792_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0638_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1253_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1449_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0761_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1330_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0268_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0372_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0757_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0031_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0815_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0074_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0771_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0936_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1355_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0764_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0516_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1460_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0394_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0178_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1005_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1103_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0899_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1204_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1061_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0327_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0233_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0057_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0922_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0996_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1457_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0513_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1323_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0426_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1428_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0879_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0297_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1395_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0333_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0202_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1171_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0746_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0111_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1459_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0806_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1332_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0003_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1082_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0745_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1187_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0803_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0326_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1131_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0678_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1458_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1465_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0662_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0715_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0043_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1029_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0108_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0450_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0465_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0125_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1052_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0988_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0487_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0872_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0096_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0865_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0643_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0507_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0032_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1228_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1076_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1229_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0219_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1341_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0384_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0583_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0236_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1267_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0682_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1324_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0672_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1237_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0256_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0689_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0801_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1392_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0630_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0041_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1027_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0173_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1476_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1455_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0172_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0228_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0812_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0980_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0781_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1280_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0141_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1046_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0656_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0137_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0234_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0909_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0410_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0836_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1313_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1326_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1113_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1040_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1345_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1492_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0831_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0293_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1063_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0309_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0353_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0597_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0011_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0189_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1174_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0474_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1079_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0066_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0697_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0734_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0756_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1488_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0238_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0235_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0274_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0417_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0316_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1451_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0864_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0127_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0900_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0244_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0291_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1205_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1344_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1351_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0441_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0213_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1143_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1320_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0437_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0927_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0324_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1125_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1421_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0735_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0573_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0002_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1416_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0776_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1430_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1322_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0180_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0272_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0896_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0543_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0570_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0859_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0255_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0263_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1299_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0835_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0039_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0611_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0369_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0732_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1485_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0045_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1138_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0067_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0642_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0961_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0249_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1221_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0496_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1312_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0599_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0497_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1413_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1382_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0661_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1251_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0560_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0350_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0018_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0354_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0007_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0594_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1003_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1375_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0298_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0199_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0555_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0419_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0400_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0608_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0789_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0618_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0212_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0493_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0633_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0920_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0621_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0129_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1033_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0982_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0300_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1000_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1284_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0241_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0163_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0971_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0455_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1124_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1464_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0332_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1225_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0160_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0132_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0953_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1396_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0319_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0014_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0867_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0765_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0731_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0121_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1340_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1414_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1149_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0897_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1140_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1406_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1491_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0811_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0047_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0907_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0184_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1202_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0595_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1333_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0149_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1234_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1289_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0331_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1415_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0769_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0935_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0216_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0940_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0762_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0445_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0378_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0280_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0376_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1047_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1045_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1056_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0356_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0210_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1386_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1110_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0476_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1163_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0598_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0511_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0279_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1254_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0115_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0365_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1298_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0839_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1227_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1282_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0030_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0254_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0658_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0978_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0851_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0130_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0357_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0152_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0952_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0834_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1436_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1302_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1210_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1445_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1328_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0188_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1152_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0340_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0534_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0986_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0892_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0062_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1173_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0009_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0537_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0058_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0286_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1350_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0837_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0068_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1291_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1108_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0158_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0425_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1007_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0717_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1314_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1097_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0131_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1425_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0050_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1432_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0257_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1487_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0793_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0655_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0339_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1109_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0151_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0830_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0912_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0700_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0959_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1156_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0798_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1371_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0023_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1483_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1357_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0122_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0462_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1021_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0562_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0505_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0787_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1420_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1399_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1024_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1296_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0454_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0844_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0133_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1377_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1037_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0705_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0623_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0950_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0547_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0240_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1011_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0893_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0117_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1178_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1120_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0346_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1354_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0880_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1019_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0477_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1231_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0660_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0471_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1168_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1214_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1378_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1142_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1102_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1015_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1438_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1372_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0945_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1248_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1119_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0064_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0973_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0509_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1135_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0019_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0663_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1025_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0261_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1431_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1403_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0004_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1374_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1164_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0740_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1217_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0387_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0911_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1203_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0995_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0535_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1319_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0795_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1184_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0915_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0576_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0329_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0929_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0142_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0413_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0520_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1080_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1334_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1207_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0869_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1012_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0475_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0167_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0494_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0264_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1358_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0641_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1318_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1194_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0059_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0578_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1095_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0692_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1422_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0195_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0431_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0928_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0388_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0690_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0467_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0415_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0119_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0187_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0424_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0716_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0854_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0252_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0635_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0755_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0294_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0719_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0785_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0914_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1364_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0021_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1242_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0220_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0139_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1220_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1383_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0102_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1424_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0921_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1215_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0722_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0908_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0805_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0414_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0190_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0089_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1150_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1360_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0453_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0827_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1273_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0670_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0891_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1329_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1261_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1147_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1262_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0359_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0328_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1240_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1391_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0156_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0989_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1219_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0923_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0654_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0698_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0065_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1287_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0383_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0532_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0260_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1349_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1086_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0788_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0113_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1363_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0687_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1083_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0918_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0585_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0392_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0253_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1043_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0449_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0568_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0421_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0747_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1175_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1373_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1304_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1474_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1482_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1385_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0317_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0603_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0519_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1090_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0020_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1489_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0871_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0510_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0104_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1018_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0245_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1440_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0758_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1247_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1010_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1179_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0552_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0276_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0679_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0571_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1017_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0301_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1477_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1450_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0712_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0676_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0577_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0684_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0881_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0644_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0076_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1446_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0154_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0284_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0824_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0231_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0701_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0790_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1224_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1454_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0405_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0177_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0267_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0944_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1368_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0389_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0610_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0128_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1085_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1057_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0459_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0027_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0759_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0458_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0472_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0816_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0022_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0832_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0097_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0407_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0820_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0777_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0941_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0821_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0796_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1022_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0967_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0548_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1117_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0197_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0175_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0218_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0397_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0850_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0710_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0592_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0447_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0607_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1249_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0436_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0784_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0728_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0646_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1104_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0443_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1154_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0194_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1206_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1233_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0170_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0299_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0605_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0033_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0037_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0631_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0877_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0565_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0626_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0723_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0650_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0590_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0938_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1132_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1479_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0313_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0651_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0926_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1274_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0828_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0602_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0637_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0423_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0681_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0411_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0223_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1186_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0591_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1441_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0533_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1366_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1466_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1423_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1075_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0939_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1327_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0647_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0420_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0330_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1016_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0600_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0606_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1096_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0246_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0530_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0344_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1471_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0774_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0733_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0283_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0561_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0466_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1032_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1188_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1285_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0029_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1081_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0063_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0739_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0954_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0448_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1331_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1218_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0688_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1169_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0799_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0866_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0853_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0955_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0910_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0490_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0861_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0948_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1336_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0349_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0808_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1277_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0601_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0044_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0315_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1230_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1255_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0271_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1051_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0481_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0970_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0412_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0040_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1068_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1091_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1394_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0862_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1461_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1453_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0070_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0852_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0012_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0484_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0140_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0919_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0288_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1490_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0863_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1129_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0737_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0791_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0884_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0708_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0667_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1452_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1048_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0609_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1049_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0336_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0800_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0196_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0931_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1072_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0207_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0144_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0395_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0648_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1389_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0582_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0433_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0553_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1088_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0391_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0304_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0110_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1310_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0398_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0446_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0574_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0772_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0686_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1335_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1417_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1388_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0855_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0457_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1093_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0903_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0669_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0860_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0754_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0289_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0946_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1275_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0024_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0310_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1153_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0958_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0385_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1115_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0550_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0653_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1243_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1030_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0775_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0382_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1369_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1429_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1400_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1213_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0709_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0842_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0691_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1066_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0766_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1139_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0499_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0699_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0200_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0239_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0439_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1155_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0525_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0523_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0079_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0416_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1481_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0727_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0307_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0138_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1427_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1475_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0429_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0206_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0375_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1448_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1352_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0270_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0103_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0778_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1408_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1223_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1472_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0813_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1384_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0629_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0337_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0848_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0124_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0753_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1410_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0243_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0714_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1001_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1317_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0112_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1260_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0251_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1272_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0624_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1158_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0444_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0925_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1050_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1134_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0695_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0508_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0615_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0265_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1316_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1148_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1315_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0226_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0320_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0237_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1353_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0287_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0262_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1067_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0430_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0001_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0360_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0017_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1468_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0501_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1141_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0998_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1145_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1473_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0396_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0913_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0380_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0544_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1411_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0106_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0442_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0418_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0491_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0368_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0296_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0483_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0536_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1306_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1347_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1467_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0969_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1036_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0010_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1044_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0619_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1437_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0282_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1162_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0750_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0126_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0192_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1362_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1435_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1359_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0685_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1469_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1338_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0273_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0060_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0229_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0358_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0743_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1294_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1136_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0155_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1122_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0707_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1301_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0814_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0083_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0303_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1208_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1116_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0512_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0205_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0847_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0464_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1365_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1266_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0278_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0075_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0201_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1308_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1235_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0056_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0061_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0917_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0401_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0617_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1236_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0974_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0693_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1245_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0524_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0541_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0351_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0099_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0614_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0403_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0321_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0370_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1379_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0084_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0666_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0783_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0028_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0208_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1244_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1300_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0883_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1239_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0308_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0035_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0486_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1271_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0078_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1127_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0377_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0652_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1443_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0711_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0838_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0322_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1222_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0290_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0857_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1062_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0186_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1114_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1073_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0802_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1123_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0281_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0770_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0620_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0581_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0456_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0094_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0092_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0645_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1426_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0334_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0809_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0422_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1170_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0153_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0147_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0100_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0636_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1026_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0639_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0081_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0720_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1094_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0381_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0342_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1387_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0347_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0191_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0409_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1118_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0159_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0485_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1009_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1165_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0116_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0586_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1381_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0668_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0992_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0073_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0664_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1264_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0183_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0822_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0338_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1200_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0479_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0972_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0826_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0780_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0432_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0373_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0767_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0469_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0779_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0217_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1177_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0364_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1014_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0461_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1209_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0976_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0804_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0549_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1180_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0120_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1401_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0559_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0933_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1157_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0114_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0480_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0964_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1023_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0580_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1053_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0266_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0438_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0875_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1286_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0123_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0729_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0937_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0295_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0277_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0069_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1006_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0994_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1100_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0960_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1407_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0741_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0374_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1065_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0730_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0306_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1074_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0222_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0008_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0763_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0563_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1172_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0632_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1092_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0906_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0613_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0890_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1070_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0136_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0947_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1238_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0882_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0495_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1128_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1279_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0575_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1201_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0404_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0773_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0898_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1321_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0846_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0518_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1367_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0345_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0015_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0451_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1297_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1241_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0987_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0150_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0696_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0807_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1035_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1434_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0649_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1404_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1281_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0768_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0738_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1309_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1133_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1058_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0427_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0468_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0500_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0148_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0386_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0616_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0428_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0118_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0569_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0323_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0718_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1252_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1191_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0143_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0314_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0671_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-1486_text_document -0.673 /eagle/datasets//dolma/data_Llama2Tokenizer/common-crawl/cc_en_tail/cc_en_tail-0596_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_peS2o.txt b/ALCF/data-lists/polaris/data_file_list_peS2o.txt deleted file mode 100644 index 3f2ddfb2992..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_peS2o.txt +++ /dev/null @@ -1,42 +0,0 @@ -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0039_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0014_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0034_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0007_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0020_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0026_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0036_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0030_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0015_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0018_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0033_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0027_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0023_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0024_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0009_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0025_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0010_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0032_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0029_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0021_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0040_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0000_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0013_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0005_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0022_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0011_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0038_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0003_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0019_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0031_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0012_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0041_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0004_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0001_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0037_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0006_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0016_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0002_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0017_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0028_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0035_text_document -0.057 /eagle/datasets//dolma/data_Llama2Tokenizer/peS2o//s2_v3-0008_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_stack.txt b/ALCF/data-lists/polaris/data_file_list_stack.txt deleted file mode 100644 index f5049cd9e4f..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_stack.txt +++ /dev/null @@ -1,4435 +0,0 @@ -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scss/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+django/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/text/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elixir/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rhtml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl6/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gap/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/alloy/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/squirrel/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webassembly/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/common-lisp/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rouge/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/textile/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/labview/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mask/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/batchfile/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/openscad/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/openscad/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/maxscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/modelica/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/modelica/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unrealscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pascal/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lookml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opal/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lua/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/module-management-system/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opa/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ren'py/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ren'py/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smalltalk/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/maple/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/numpy/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/realbasic/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pure-data/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pure-data/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ceylon/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sql/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0102_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0104_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0105_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0106_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0103_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0107_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0108_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0101_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/augeas/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/swift/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rdoc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logtalk/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c2hs-haskell/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pov-ray-sdl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ioke/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/verilog/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clips/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/chuck/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stylus/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pod/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pod/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xojo/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/makefile/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/f-sharp/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/component-pascal/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/component-pascal/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/powershell/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emacs-lisp/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lex/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lex/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/grace/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/processing/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sas/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sas/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/netlogo/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c++/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autoit/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autoit/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zephir/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/robotframework/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/propeller-spin/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xquery/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/txl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nu/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/toml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ampl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tea/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/csound/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/brightscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/perl/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slim/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slim/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/red/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/thrift/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pony/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stata/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-coffeescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ats/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot-internal-representation/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lolcode/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/digital-command-language/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/abap/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/abap/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lsl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nix/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jade/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hlsl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hlsl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scaml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ags-script/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tex/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yaml/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clojure/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xs/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mtml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rmarkdown/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kit/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mako/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/igor-pro/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sourcepawn/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sourcepawn/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/apl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/protocol-buffer/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nginx/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scala/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ocaml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/piglatin/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shell/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/solidity/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/awk/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vcl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gdscript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/applescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/webidl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/max/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glyph/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/papyrus/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/boo/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hy/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/d/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/aspectj/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/isabelle/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/isabelle/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sqf/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sqf/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/volt/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/monkey/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lfe/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mathematica/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clarion/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/oxygene/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/metal/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nsis/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zig/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/muf/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dylan/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xbase/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smt/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autohotkey/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/autohotkey/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/erlang/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arduino/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/clean/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ston/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/creole/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ecl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/elm/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eiffel/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eiffel/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turtle/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/darcs-patch/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+php/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cython/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dm/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cmake/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/opencl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot-assembly/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asciidoc/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cobol/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/io/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/saltstack/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ox/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/matlab/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/renderscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/less/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purescript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dogescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/omgrofl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/supercollider/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/typescript/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/flux/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/viml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-haskell/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcsh/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fortran/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+eex/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/golo/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pawn/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kicad/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/oz/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/idl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/literate-agda/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/click/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/freemarker/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/krl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/inform-7/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smarty/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/stan/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/livescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dockerfile/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion-cfc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion-cfc/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gosu/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/kotlin/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/parrot/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lilypond/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/moonscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jupyter-notebook/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/idris/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mediawiki/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/logos/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fish/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/julia/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coffeescript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/asp/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/systemverilog/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/org/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/css/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/actionscript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/myghty/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/arc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphviz-(dot)/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/latte/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/turing/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/java-server-pages/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bison/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/emberscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xpages/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/linker-script/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/linker-script/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coldfusion/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/eagle/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/graphql/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cycript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scilab/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-eclass/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/go/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy-server-pages/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy-server-pages/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dns-zone/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/scheme/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/hcl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/netlinx/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/irc-log/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mirah/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rust/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xml/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haskell/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/visual-basic/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/purebasic/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/handlebars/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cartocss/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/j/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jflex/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unity3d-asset/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/qmake/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/api-blueprint/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gettext-catalog/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pan/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/brainfuck/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/edn/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ecere-projects/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fantom/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/tcl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/grammatical-framework/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/befunge/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sass/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/restructuredtext/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objdump/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ada/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vhdl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/twig/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/urweb/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ruby/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/blitzmax/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/liquid/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/liquid/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/genshi/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/g-code/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ninja/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ninja/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gams/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lasso/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/desktop/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/postscript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/agda/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/agda/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-c++/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m4/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m4/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/coq/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/php/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-ebuild/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gentoo-ebuild/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/factor/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/uno/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/apacheconf/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pogoscript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nimrod/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jasmin/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/javascript/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/html+erb/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bluespec/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nit/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ec/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/raml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/rebol/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/diff/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/objective-j/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bro/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sparql/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/chapel/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/pike/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ini/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/harbour/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/llvm/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/crystal/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/lean/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/unified-parallel-c/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/c-sharp/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xtend/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ooc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shen/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/self/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/m/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/x10/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cirru/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/redcode/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/mupad/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/markdown/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/glsl/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/inno-setup/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/smali/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groff/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/wisp/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/haxe/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cucumber/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/http/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/http/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yacc/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/forth/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/yang/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/r/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gnuplot/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xslt/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/ragel-in-ruby-host/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nesc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/nesc/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/standard-ml/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/slash/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/dart/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/zimpl/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/sage/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/blitzbasic/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/octave/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/fancy/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/antlr/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bitbake/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/bitbake/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/gas/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/web-ontology-language/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/jsx/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vala/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cap'n-proto/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/prolog/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/cuda/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/shellsession/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/python-traceback/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/groovy/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0087_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0025_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0019_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0015_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0009_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0064_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0050_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0058_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0093_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0059_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0051_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0021_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0071_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0008_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0044_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0072_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0097_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0068_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0080_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0057_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0047_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0052_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0067_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0014_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0023_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0034_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0032_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0007_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0060_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0086_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0094_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0091_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0017_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0090_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0066_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0046_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0096_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0100_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0020_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0076_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0092_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0048_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0049_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0075_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0026_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0012_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0095_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0074_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0083_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0054_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0070_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0082_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0055_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0024_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0037_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0088_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0078_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0065_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0062_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0089_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0099_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0061_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0041_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0079_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0035_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0002_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0028_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0045_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0040_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0004_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0098_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0038_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0056_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0063_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0011_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0003_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0077_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0042_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0013_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0029_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0039_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0022_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0084_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0085_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0069_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0081_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0006_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0073_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0027_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0010_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0043_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0005_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0016_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0033_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0018_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0036_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0030_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0001_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0053_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/vue/v3-0031_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/xproc/v3-0000_text_document -0.172 /eagle/datasets//dolma/data_Llama2Tokenizer/stack-code/racket/v3-0000_text_document diff --git a/ALCF/data-lists/polaris/data_file_list_wiki.txt b/ALCF/data-lists/polaris/data_file_list_wiki.txt deleted file mode 100644 index 134c1473b11..00000000000 --- a/ALCF/data-lists/polaris/data_file_list_wiki.txt +++ /dev/null @@ -1,2 +0,0 @@ -0.0045 /eagle/datasets//dolma/data_Llama2Tokenizer/wiki-en-simple//en_simple_wiki-0000_text_document -0.0045 /eagle/datasets//dolma/data_Llama2Tokenizer/wiki-en-simple//en_simple_wiki-0001_text_document diff --git a/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt b/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt new file mode 100644 index 00000000000..6f34558ec33 --- /dev/null +++ b/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0002583902668716813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0031025147279277244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0002406272620255565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0003742481815405742 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003547982093445404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +6.322825248625475e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +0.001451215788905126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.0012499632072059553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0005759963691850877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0009994361338078242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.004474659408857016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.00032927705604725614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.003548077173506675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document diff --git a/ALCF/data-lists/polaris/falcon.txt b/ALCF/data-lists/polaris/falcon.txt new file mode 100644 index 00000000000..68aeb2f27bb --- /dev/null +++ b/ALCF/data-lists/polaris/falcon.txt @@ -0,0 +1,501 @@ +0.0003547982093445404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document + diff --git a/ALCF/data-lists/polaris/megawiki.txt b/ALCF/data-lists/polaris/megawiki.txt new file mode 100644 index 00000000000..4c4f47df5ff --- /dev/null +++ b/ALCF/data-lists/polaris/megawiki.txt @@ -0,0 +1,262 @@ +6.322825248625475e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document diff --git a/ALCF/data-lists/polaris/open-web-math-train.txt b/ALCF/data-lists/polaris/open-web-math-train.txt new file mode 100644 index 00000000000..caab74fb9f5 --- /dev/null +++ b/ALCF/data-lists/polaris/open-web-math-train.txt @@ -0,0 +1,13 @@ +0.001451215788905126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document diff --git a/ALCF/data-lists/polaris/pes2o.txt b/ALCF/data-lists/polaris/pes2o.txt new file mode 100644 index 00000000000..7bb62d46b62 --- /dev/null +++ b/ALCF/data-lists/polaris/pes2o.txt @@ -0,0 +1,26 @@ +0.0012499632072059553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document diff --git a/ALCF/data-lists/polaris/reddit.txt b/ALCF/data-lists/polaris/reddit.txt new file mode 100644 index 00000000000..4f46ee0e648 --- /dev/null +++ b/ALCF/data-lists/polaris/reddit.txt @@ -0,0 +1,78 @@ +0.0005759963691850877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document diff --git a/ALCF/data-lists/polaris/stack.txt b/ALCF/data-lists/polaris/stack.txt new file mode 100644 index 00000000000..971329159b0 --- /dev/null +++ b/ALCF/data-lists/polaris/stack.txt @@ -0,0 +1,26 @@ +0.0009994361338078242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document diff --git a/ALCF/data-lists/polaris/starcoder.txt b/ALCF/data-lists/polaris/starcoder.txt new file mode 100644 index 00000000000..c675b0b84df --- /dev/null +++ b/ALCF/data-lists/polaris/starcoder.txt @@ -0,0 +1,50 @@ +0.004474659408857016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document + diff --git a/ALCF/data-lists/polaris/tulu.txt b/ALCF/data-lists/polaris/tulu.txt new file mode 100644 index 00000000000..a65ae2b93a3 --- /dev/null +++ b/ALCF/data-lists/polaris/tulu.txt @@ -0,0 +1,66 @@ +0.00032927705604725614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document diff --git a/ALCF/data-lists/polaris/wiki.txt b/ALCF/data-lists/polaris/wiki.txt new file mode 100644 index 00000000000..7759120d876 --- /dev/null +++ b/ALCF/data-lists/polaris/wiki.txt @@ -0,0 +1,2 @@ +0.003548077173506675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document From caa1a4bfdd159594de2615fc36a580a86a735386 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 11:02:07 -0500 Subject: [PATCH 100/590] Add `ALCF/test_polaris.sh` --- ALCF/test_polaris.sh | 84 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 84 insertions(+) create mode 100644 ALCF/test_polaris.sh diff --git a/ALCF/test_polaris.sh b/ALCF/test_polaris.sh new file mode 100644 index 00000000000..cbe2b41c5ec --- /dev/null +++ b/ALCF/test_polaris.sh @@ -0,0 +1,84 @@ +#!/bin/bash --login +# +# Run complete test of +# https://github.com/argonne-lcf/Megatron-DeepSpeed +# on Polaris @ ALCF +# to launch (inside an interactive `qsub -I` job) on Polaris: +# +# ```bash` +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed/ALCF +# $ bash test_polaris.sh +# ```` + +# EXIT ON ERROR(s) +set -euxo pipefail + +NOW="$(date "+%Y-%m-%d-%H%M%S")" + +######################################################## +# Setup / activate conda environment, +# mine is called q4-drop +######################################################## +setup_conda() { + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" + micromamba activate 2024-04-25 + else + echo "Found existing python at: $(which python3)" + fi +} + + +######################################## +# Make sure ./tmp/Megatron-DeepSpeed +# does not already exist +######################################## +setup_megatron_deepspeed() { + OUTDIR="OUTPUTS/test-polaris-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" + echo "Running test in: ${OUTDIR}" + echo "WORKING DIRECTORY: $(realpath $(pwd .))" + if [[ -d "Megatron-DeepSpeed" ]]; then + echo "Found existing Megatron-DeepSpeed in ${OUTDIR}" + echo "Remove Megatron-DeepSpeed from ${OUTDIR} to run test." + exit + fi + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + if [[ -z "${GIT_BRANCH-}" ]]; then + git checkout "${GIT_BRANCH}" + fi +} + + +main() { + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + setup_conda + else + echo "Unable to setup python. Exiting" + exit 1 + fi + setup_megatron_deepspeed + export DEBUG=1 + export PBS_O_WORKDIR="$(pwd)" + export DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt + export ZERO_STAGE=1 + export NUM_LAYERS=10 + export MICRO_BATCH=8 + export TRAIN_ITER=20 + export TIMING_LOG_LEVEL=1 + bash train_llama_alcf.sh |& tee "test-polaris-${NOW}".log +} + +main From b534e09d000a343af0d65b0b6feb652303b74aa1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 11:02:26 -0500 Subject: [PATCH 101/590] Fix duplicate loggers in `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 7ed38614a71..f83ef05fc8d 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -25,7 +25,7 @@ # # checkpoint_throughput_calculator # ) # from pathlib import Path -from enrich import get_logger +import logging import deepspeed from deepspeed.runtime.utils import see_memory_usage @@ -49,7 +49,7 @@ torch.cuda.set_device(LOCAL_RANK) # ------------------------------------------- # --- [TURN OFF LOGGER ON ALL RANK != 0] ---- -log = get_logger(__name__) +log = logging.getLogger(__name__) log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") # ---- [SETUP WANDB FROM RANK 0] -------------- WANDB_MODE = os.environ.get('WANDB_MODE', None) From 2c4d7728fa15e762ad9ef896c2bf15c440d70f5e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 11:03:11 -0500 Subject: [PATCH 102/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 88 ++++++++++++++++++++++++++++++++----------------- 1 file changed, 57 insertions(+), 31 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index fdd42b56b82..9b057e0c03c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -16,7 +16,19 @@ export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} -printJobInfo() { +function setupSrun() { + if [[ $(hostname) == login* || $(hostname) == nid* ]]; then + export NHOSTS="${SLURM_NNODES:-1}" + export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" + export NGPUS="$(( NHOSTS * NGPU_PER_HOST ))" + export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" + else + echo "Skipping setupSrun() on $(hostname)" + fi +} + + +function printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" echo "- Using $(which python3)" @@ -27,20 +39,33 @@ printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } +function setupVenv() { + VENV_DIR="$1" + if [[ -d "${VENV_DIR}" ]]; then + echo "Found venv at: ${VENV_DIR}" + source "${VENV_DIR}/bin/activate" + else + echo "Skipping setupVenv() on $(hostname)" + fi +} -setupSrun() { - if [[ $(hostname) == login* || $(hostname) == nid* ]]; then - export NHOSTS="${SLURM_NNODES:-1}" - export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" - export NGPUS="$(( NHOSTS * NGPU_PER_HOST ))" - export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" +function loadCondaEnv() { + if [[ "${CONDA_EXE}" ]]; then + echo "Already inside ${CONDA_EXE}, exiting!" else - echo "Skipping setupSrun() on $(hostname)" + MODULE_STR="$1" + module load "conda/${MODULE_STR}" + nargs="$#" + if [[ "${nargs}" -ge 2 ]]; then + conda activate "$2" + else + conda activate base + fi fi } -setupLauncher() { +function setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then export LAUNCH_CMD="${DIST_LAUNCH} --cpu-bind depth -d 16 python3 -Wignore ${EXEC}" @@ -53,7 +78,7 @@ setupLauncher() { printf " %s" "$(printMagenta ${LAUNCH_CMD})" } -setDSlauncher() { +function setDSlauncher() { # launcher setting outdir=$1 export hfds="$outdir/hostfile_deepspeed" @@ -68,7 +93,7 @@ setDSlauncher() { fi } -setParams() { +function setParams() { LLAMA_ARGS="" # +----[Parallelism Settings] -------------------------------------------+ # +------[Aurora]--------||-------[SunSpot]-------------+ @@ -155,7 +180,7 @@ setParams() { } -setArgs() { +function setArgs() { # ---- Set DeepSpeed arguments -------------------------------- ds_args=" " ds_args=" --deepspeed ${ds_args}" @@ -186,7 +211,7 @@ setArgs() { } -make_ds_hostfile() { +function make_ds_hostfile() { export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" @@ -202,7 +227,7 @@ make_ds_hostfile() { # | 1. Git clone ezpz (if not found) | # | 2. Install ezpz (if not installed) | # +---------------------------------------+ -ezpz() { +function ezpz() { if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" @@ -228,7 +253,7 @@ ezpz() { # | Save important environment variables to .deepspeed_env, which will be | # | forwarded to ALL ranks with DeepSpeed | # +------------------------------------------------------------------------+ -saveDSenv() { +function saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { echo "PATH=${PATH}" ; @@ -240,7 +265,7 @@ saveDSenv() { } > .deepspeed_env } -setOutput() { +function setOutput() { # ---- Specify output location -------------------------------- export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" @@ -253,7 +278,7 @@ setOutput() { echo "!!!Please see logs at ${OUTPUT_DIR}" } -buildDSconfig() { +function buildDSconfig() { # ---- Build DeepSpeed Config --------------------------------- export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" @@ -267,13 +292,13 @@ buildDSconfig() { } -sumWeights() { +function sumWeights() { local file_list=$1 weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") python3 -c "import numpy as np; print(np.sum(${weights}))" } -sumFiles() { +function sumFiles() { local rd=$1 for f in $("${rd}/*.txt"); do ws=$(sumWeights "${rd}/${f}") @@ -282,7 +307,7 @@ sumFiles() { } -setEnv() { +function setEnv() { # ---- [SunSpot] ------- || ---- [Aurora] -------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then # PBS_PARENT=$(dirname ${PBS_O_WORKDIR}) @@ -323,7 +348,7 @@ setEnv() { } -makeHostfiles() { +function makeHostfiles() { if [[ -n "${HOSTFILE}" ]]; then printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" else @@ -333,7 +358,7 @@ makeHostfiles() { fi } -setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- +function setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- if [[ $(hostname) == x4* ]]; then # ---- [AURORA] ---- dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then @@ -370,7 +395,7 @@ setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- echo "--------------------" } -generateDSconfig() { +function generateDSconfig() { for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ "$PP" "$DTYPE" do @@ -532,33 +557,34 @@ $flops_profiler EOT } -printBlack() { +function printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } -printRed() { +function printRed() { printf "\e[1;31m%s\e[0m\n" "$@" } -printGreen() { +function printGreen() { printf "\e[1;32m%s\e[0m\n" "$@" } -printYellow() { +function printYellow() { printf "\e[1;33m%s\e[0m\n" "$@" } -printBlue() { +function printBlue() { printf "\e[1;34m%s\e[0m\n" "$@" } -printMagenta() { +function printMagenta() { printf "\e[1;35m%s\e[0m\n" "$@" } -printCyan() { +function printCyan() { printf "\e[1;36m%s\e[0m\n" "$@" } -printWhite() { + +function printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } From cfa6b52e82c29018ebe92642e0379f393fe26eb5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 11:14:02 -0500 Subject: [PATCH 103/590] Update `ALCF/test_{polaris,sirius,sunspot}.sh` --- ALCF/test_polaris.sh | 10 +++++++--- ALCF/test_sirius.sh | 38 ++++++++++++++++++++++++++++++++------ ALCF/test_sunspot.sh | 35 +++++++++++++++++++++++++++++------ 3 files changed, 68 insertions(+), 15 deletions(-) diff --git a/ALCF/test_polaris.sh b/ALCF/test_polaris.sh index cbe2b41c5ec..3a6734f61eb 100644 --- a/ALCF/test_polaris.sh +++ b/ALCF/test_polaris.sh @@ -22,7 +22,7 @@ NOW="$(date "+%Y-%m-%d-%H%M%S")" ######################################################## setup_conda() { if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then - export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba + export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" micromamba activate 2024-04-25 @@ -46,7 +46,7 @@ setup_megatron_deepspeed() { exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed - if [[ -z "${GIT_BRANCH-}" ]]; then + if [[ -n "${GIT_BRANCH-}" ]]; then git checkout "${GIT_BRANCH}" fi } @@ -72,7 +72,11 @@ main() { setup_megatron_deepspeed export DEBUG=1 export PBS_O_WORKDIR="$(pwd)" - export DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt + export DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/books.txt" + if [[ ! -f "${DATA_FILE_LIST}" ]]; then + echo "Unable to find / use ${DATA_FILE_LIST}. Exiting." + exit 1 + fi export ZERO_STAGE=1 export NUM_LAYERS=10 export MICRO_BATCH=8 diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh index 108649612c4..39170fa47cf 100755 --- a/ALCF/test_sirius.sh +++ b/ALCF/test_sirius.sh @@ -21,10 +21,14 @@ NOW="$(date "+%Y-%m-%d-%H%M%S")" # mine is called q4-drop ######################################################## setup_conda() { - export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba - shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" - micromamba activate 2024-04-23 + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" + micromamba activate 2024-04-23 + else + echo "Found existing python at: $(which python3)" + fi } @@ -42,15 +46,37 @@ setup_megatron_deepspeed() { exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + if [[ -n "${GIT_BRANCH-}" ]]; then + git checkout "${GIT_BRANCH}" + fi } main() { - setup_conda + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + setup_conda + else + echo "Unable to setup python. Exiting" + exit 1 + fi setup_megatron_deepspeed export DEBUG=1 export PBS_O_WORKDIR="$(pwd)" - export DATA_FILE_LIST=./ALCF/data-lists/sirius/books.txt + export DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/sirius/books.txt" + if [[ ! -f "${DATA_FILE_LIST}" ]]; then + echo "Unable to find / use ${DATA_FILE_LIST}. Exiting." + exit 1 + fi export ZERO_STAGE=1 export NUM_LAYERS=10 export MICRO_BATCH=8 diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh index 67f6868d435..b39bff7736a 100755 --- a/ALCF/test_sunspot.sh +++ b/ALCF/test_sunspot.sh @@ -21,12 +21,13 @@ NOW="$(date "+%Y-%m-%d-%H%M%S")" # mine is called q4-drop ######################################################## setup_conda() { - if [[ "${SHELL}" = "/bin/zsh" ]]; then - eval "$(~/miniconda3/bin/conda shell.zsh hook)" + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$(~/miniconda3/bin/conda shell hook -s posix)" + conda activate q4-drop else - eval "$(~/miniconda3/bin/conda shell.bash hook)" + echo "Found existing python at: $(which python3)" fi - conda activate q4-drop } @@ -44,15 +45,37 @@ setup_megatron_deepspeed() { exit fi git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + if [[ -n "${GIT_BRANCH-}" ]]; then + git checkout "${GIT_BRANCH}" + fi } main() { - setup_conda + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + setup_conda + else + echo "Unable to setup python. Exiting" + exit 1 + fi setup_megatron_deepspeed export DEBUG=1 export PBS_O_WORKDIR="$(pwd)" - export DATA_FILE_LIST=./ALCF/data-lists/sunspot/books.txt + export DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/sunspot/books.txt" + if [[ ! -f "${DATA_FILE_LIST}" ]]; then + echo "Unable to find / use ${DATA_FILE_LIST}. Exiting." + exit 1 + fi export ZERO_STAGE=1 export NUM_LAYERS=10 export MICRO_BATCH=8 From a3114bf898c035b4c4db366af4e6745c9bc82276 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 11:15:36 -0500 Subject: [PATCH 104/590] Add `ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt` --- .../sunspot/dolma_v1_7_file_list.txt | 2419 +++++++++++++++++ 1 file changed, 2419 insertions(+) create mode 100644 ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt diff --git a/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt b/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt new file mode 100644 index 00000000000..5d142522a76 --- /dev/null +++ b/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0002583902668716813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0002406272620255565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0003742481815405742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003547982093445404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +6.322825248625475e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +0.001451215788905126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.0012499632072059553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0005759963691850877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0009994361338078242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.004474659408857016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.00032927705604725614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.003548077173506675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document From f63aad14f230c60a0ad523019c09c34075326445 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:02:46 -0500 Subject: [PATCH 105/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 68 +++++++++++++++++++++++++++++++++++++------------ 1 file changed, 52 insertions(+), 16 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 9b057e0c03c..a0d191bb8ca 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -306,13 +306,46 @@ function sumFiles() { done } +######################################################## +# Setup / activate conda environment, +# mine is called q4-drop +######################################################## +setup_conda_sunspot() { + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$(~/miniconda3/bin/conda shell hook -s posix)" + conda activate q4-drop + else + echo "Found existing python at: $(which python3)" + fi +} + +setup_conda_sirius() { + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" + micromamba activate 2024-04-23 + else + echo "Found existing python at: $(which python3)" + fi +} + +setup_conda_polaris() { + if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" + micromamba activate 2024-04-25 + else + echo "Found existing python at: $(which python3)" + fi +} + function setEnv() { # ---- [SunSpot] ------- || ---- [Aurora] -------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then - # PBS_PARENT=$(dirname ${PBS_O_WORKDIR}) - # echo "Sourcing ${PBS_PARENT}/setenv.sh..." - # source "${PBS_PARENT}/setenv.sh" || exit source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit # ----- [Aurora] ----------------------------------- if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then @@ -321,20 +354,22 @@ function setEnv() { # ----- [SunSpot] ---------------------------------- elif [[ $(hostname) == x1* ]]; then echo "Running on SunSpot !!" - eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + setup_conda_sunspot + # eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop fi fi # ----- [Polaris] --------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then export MACHINE="Running on Sirius !!" + setup_conda_sirius else echo "Running on Polaris !!" # ---- [load conda] --------------------- - # module load conda/2023-10-04; conda activate cu118-pt221 ; unset PYTHONUSERBASE - if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then - source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" - fi + setup_conda_polaris + # if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then + # source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + # fi fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" @@ -353,8 +388,6 @@ function makeHostfiles() { printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" else make_ds_hostfile - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" || exit #> /tmp/savejobenv.log 2>&1 & - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" || exit fi } @@ -362,24 +395,27 @@ function setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------ if [[ $(hostname) == x4* ]]; then # ---- [AURORA] ---- dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/data_file_list_books.txt" - # dfl_fallback="/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_file_list_reweighted.txt" + # shellcheck: source ./data-lists/sunspot/books.txt + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/books.txt" elif [[ $(hostname) == x3* ]]; then - # dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/data_file_list_books.txt" + if [[ "${PBS_O_HOST}" == sirius* ]]; then + # shellcheck: source ./data-lists/sirius/books.txt + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/books.txt" + elif [[ "${PBS_O_HOST}" == polaris* ]]; then + # shellcheck: source ./data-lists/polaris/books.txt + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/books.txt" + fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" else echo "Unknown hostname. Must manually specify DATA_FILE_LIST." fi dfl="${1:-${dfl_fallback}}" - # dfl_fallback="/eagle/datasets/dolma/data_file_list_reweighted.txt" printf "Calling: setData() with %s\n" "${dfl}" ndocs=$(wc -l < "${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") dcp=".cache/${dfl_stem}/index-cache" - # mkdir -p dcp export DATA_FILE_LIST="${dfl}" export NUM_DOCS="${ndocs}" export WEIGHT_SUM="${ws}" From d329801a8e48a6da10b0a4d2da59961d293691dd Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:02:54 -0500 Subject: [PATCH 106/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a0d191bb8ca..868f0e194bd 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -361,7 +361,7 @@ function setEnv() { # ----- [Polaris] --------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then - export MACHINE="Running on Sirius !!" + echo "Running on Sirius !!" setup_conda_sirius else echo "Running on Polaris !!" From 585c15e3557157df2f31e1d5d3ccbf343ca6088c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:03:08 -0500 Subject: [PATCH 107/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 32 +++++++++++++------------------- 1 file changed, 13 insertions(+), 19 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index cd2d8213dd9..462da0c4e12 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -35,39 +35,38 @@ sourceFile "${HERE}/ALCF/helpers.sh" || exit # ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ setEnv || exit # 1. load `conda` environment -# saveDSenv || exit # 2. save env vars to `.deepspeed_env` +# saveDSenv || exit # 2. save env vars to `.deepspeed_env` ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars - -# if [[ -z "${HOSTFILE}" ]]; then -# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -# else -# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" -# fi setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} setArgs || exit # 8. specify additional `deepspeed` arguments setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset -# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` printJobInfo || exit # 11. print job info setupLauncher || exit # ----------------------------------------------------------------------------- - +#### [DEPRECATED] ########################################################### +# if [[ -z "${HOSTFILE}" ]]; then +# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +# else +# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" +# fi +# ---------------------------------------------------------------------------- +# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` +# ---------------------------------------------------------------------------- # TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') # printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" -# # if [[ "${TORCH_DEVICE}" == "cuda" ]]; then # printf %s "Setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True" # PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True # fi - - +# ---------------------------------------------------------------------------- # export MPICH_GPU_SUPPORT_ENABLED=1 # export CUDA_DEVICE_MAX_CONNECTIONS=1 # export NCCL_DEBUG=INFO -# -# +############################################################################# + # Assert TBDIR exists inside our $CKPT_DIR TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" @@ -130,12 +129,7 @@ run_cmd=" |& tee ${OUTPUT_LOG} " -# ds_exec -# echo "! Using $(which deepspeed)" -# ds_report - echo "${run_cmd}" - printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" eval "${run_cmd}" From 3444b99dd8639659a4c5ed987952df093d09d46a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:03:25 -0500 Subject: [PATCH 108/590] Update `.gitignore` --- .gitignore | 3 +++ 1 file changed, 3 insertions(+) diff --git a/.gitignore b/.gitignore index edbde3c246e..3b18d4956a0 100644 --- a/.gitignore +++ b/.gitignore @@ -1,4 +1,7 @@ # User Added +deps/* +OUTPUTS/* +ALCF/OUTPUTS/* *tmp* *core.* *old* From 505aef062892ef1da3e1124e7bab8c383dc54fe9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:15:03 -0500 Subject: [PATCH 109/590] Update defaults in `ALCF/helpers.sh` --- ALCF/helpers.sh | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 868f0e194bd..7fa8e748c1f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -113,12 +113,13 @@ function setParams() { ####################################################### # +--------[Polaris]-----------------------------------+ elif [[ $(hostname) == x3* ]]; then - TP=${TP:-2} # TP = 2 + TP=${TP:-1} # TP = 2 export NCCL=${NCCL:-nccl} # NCCL export BE="${NCCL}" # BE = NCCL # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - MICRO_BATCH=${MICRO_BATCH:-8} # MICRO_BATCH = 8 + export GRAD_ACC_STEPS=4 # GRADIENT_ACCUMULATION_STEPS + MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else @@ -160,7 +161,7 @@ function setParams() { # +---[Run Settings]------------------------------------------------------+ export LR=${LR:-0.0003} # LEARNING_RATE export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export ZERO_STAGE=${ZERO_STAGE:-2} # ZERO OFFLOADING STAGE + export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN From e31bb23b9c7e865b873c9d888439ba81018ba387 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 12:15:24 -0500 Subject: [PATCH 110/590] Add `train_agpt.sh` --- train_agpt.sh | 11 +++++++++++ 1 file changed, 11 insertions(+) create mode 100644 train_agpt.sh diff --git a/train_agpt.sh b/train_agpt.sh new file mode 100644 index 00000000000..4129f553e28 --- /dev/null +++ b/train_agpt.sh @@ -0,0 +1,11 @@ +#!/bin/bash --login + +NOW="$(date "+%Y-%m-%d-%H%M%S")" +cd "${PBS_O_WORKDIR}" || exit + +OUTDIR="${PBS_O_WORKDIR}/pbslogs" +mkdir -p "${OUTDIR}" +OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" +echo "${OUTFILE}" >> "${OUTDIR}/latest" +echo "Logging job output to: ${OUTFILE}" +bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From a73e9af36ef12ae8435ed72bc15bd78e58bdc670 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 13:58:31 -0500 Subject: [PATCH 111/590] Update `ALCF/test_{polaris,sirius,sunspot}.sh` --- ALCF/test_polaris.sh | 4 ++-- ALCF/test_sirius.sh | 2 +- ALCF/test_sunspot.sh | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/test_polaris.sh b/ALCF/test_polaris.sh index 3a6734f61eb..a18c87fad74 100644 --- a/ALCF/test_polaris.sh +++ b/ALCF/test_polaris.sh @@ -21,7 +21,7 @@ NOW="$(date "+%Y-%m-%d-%H%M%S")" # mine is called q4-drop ######################################################## setup_conda() { - if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" @@ -54,7 +54,7 @@ setup_megatron_deepspeed() { main() { local virtual_env="${VIRTUAL_ENV-}" - local conda_prefix="${CONDA_PREFIX}" + local conda_prefix="${CONDA_PREFIX-}" if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "Using conda from: ${conda_prefix}" elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then diff --git a/ALCF/test_sirius.sh b/ALCF/test_sirius.sh index 39170fa47cf..0a528a95198 100755 --- a/ALCF/test_sirius.sh +++ b/ALCF/test_sirius.sh @@ -54,7 +54,7 @@ setup_megatron_deepspeed() { main() { local virtual_env="${VIRTUAL_ENV-}" - local conda_prefix="${CONDA_PREFIX}" + local conda_prefix="${CONDA_PREFIX-}" if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "Using conda from: ${conda_prefix}" elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then diff --git a/ALCF/test_sunspot.sh b/ALCF/test_sunspot.sh index b39bff7736a..b3b22c78b41 100755 --- a/ALCF/test_sunspot.sh +++ b/ALCF/test_sunspot.sh @@ -53,7 +53,7 @@ setup_megatron_deepspeed() { main() { local virtual_env="${VIRTUAL_ENV-}" - local conda_prefix="${CONDA_PREFIX}" + local conda_prefix="${CONDA_PREFIX-}" if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "Using conda from: ${conda_prefix}" elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then From 57f1c964052074e8b75d04ec1a0d4242e3ef1678 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 13:58:47 -0500 Subject: [PATCH 112/590] Add `ALCF/test_alcf.sh` --- ALCF/test_alcf.sh | 166 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 166 insertions(+) create mode 100644 ALCF/test_alcf.sh diff --git a/ALCF/test_alcf.sh b/ALCF/test_alcf.sh new file mode 100644 index 00000000000..367b40d17c2 --- /dev/null +++ b/ALCF/test_alcf.sh @@ -0,0 +1,166 @@ +#!/bin/bash --login +# +# Run complete test of +# https://github.com/argonne-lcf/Megatron-DeepSpeed +# on {Polaris, Sunspot, Sirius} @ ALCF +# to launch (inside an interactive `qsub -I` job) on Polaris: +# +# ```bash` +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed/ALCF +# $ bash test_alcf.sh +# ```` + +# EXIT ON ERROR(s) +set -euxo pipefail + +NOW="$(date "+%Y-%m-%d-%H%M%S")" + +setup_conda_sunspot() { + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$(~/miniconda3/bin/conda shell hook -s posix)" + conda activate q4-drop + else + echo "Found existing python at: $(which python3)" + fi +} + +setup_conda_sirius() { + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then + export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" + micromamba activate 2024-04-23 + else + echo "Found existing python at: $(which python3)" + fi +} + +setup_conda_polaris() { + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then + # export CUDA_HOME=/soft/compilers/cudatoolkit/cuda-12.2.2 + # && export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba && eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" ; mm activate 2024-04-25 + export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba + shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" + micromamba activate 2024-04-25 + else + echo "Found existing python at: $(which python3)" + fi +} + + +function setEnv() { + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX-}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + # setup_conda + # ---- [SunSpot] ------- || ---- [Aurora] -------------- + if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then + source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit + # ----- [Aurora] ----------------------------------- + if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + if [[ $(hostname) == x4* ]]; then + eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 + # ----- [SunSpot] ---------------------------------- + elif [[ $(hostname) == x1* ]]; then + echo "Running on SunSpot !!" + setup_conda_sunspot + # eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + fi + fi + # ----- [Polaris] --------------------------------------- + elif [[ $(hostname) == x3* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + echo "Running on Sirius !!" + setup_conda_sirius + else + echo "Running on Polaris !!" + # ---- [load conda] --------------------- + setup_conda_polaris + # if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then + # source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + # fi + fi + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + echo "Running on Perlmutter !!" + module load pytorch + source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" + else # ------------------------------------- [Unknown] ------------------- + echo "Unknown hostname $(hostname)" + exit 1 + fi + else + echo "Unable to setup python environment. Exiting" + exit 1 + fi + echo "[python] Using: $(which python3)" +} + + + +######################################## +# Make sure ./tmp/Megatron-DeepSpeed +# does not already exist +######################################## +setup_megatron_deepspeed() { + OUTDIR="OUTPUTS/test-polaris-${NOW}" && mkdir -p "${OUTDIR}" && cd "${OUTDIR}" + echo "Running test in: ${OUTDIR}" + echo "WORKING DIRECTORY: $(realpath $(pwd .))" + if [[ -d "Megatron-DeepSpeed" ]]; then + echo "Found existing Megatron-DeepSpeed in ${OUTDIR}" + echo "Remove Megatron-DeepSpeed from ${OUTDIR} to run test." + exit + fi + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + if [[ -n "${GIT_BRANCH-}" ]]; then + git checkout "${GIT_BRANCH}" + fi +} + + +main() { + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX-}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + setup_conda + else + echo "Unable to setup python. Exiting" + exit 1 + fi + setup_megatron_deepspeed + export DEBUG=1 + export PBS_O_WORKDIR="$(pwd)" + SUBMITTED_FROM=$(echo $PBS_O_HOST | tr '-' ' ' | awk '{print $1}') + export DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/${SUBMITTED_FROM}/books.txt" + if [[ ! -f "${DATA_FILE_LIST}" ]]; then + echo "Unable to find / use ${DATA_FILE_LIST}. Exiting." + exit 1 + fi + export ZERO_STAGE=1 + export NUM_LAYERS=10 + export MICRO_BATCH=8 + export TRAIN_ITER=20 + export TIMING_LOG_LEVEL=1 + bash train_llama_alcf.sh |& tee "test-${SUBMITTED_FROM}-${NOW}".log +} + +main + From 455126caa6a4bf21c2ab87eaee53b1ac7404acde Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 13:59:04 -0500 Subject: [PATCH 113/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 162 ++++++++++++++++++++++++++++-------------------- 1 file changed, 95 insertions(+), 67 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 7fa8e748c1f..516bc59f4cd 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -16,6 +16,17 @@ export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} +function check_and_kill_if_running() { + # kill $(ps aux | grep -E "$USER.+(mpi|main.py)" | grep -v grep | awk '{print $2}') + RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') + if [[ -n "${RUNNING_PIDS}" ]]; + then echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}"; + else + echo "Not currently running. Continuing!" + fi +} + + function setupSrun() { if [[ $(hostname) == login* || $(hostname) == nid* ]]; then export NHOSTS="${SLURM_NNODES:-1}" @@ -100,26 +111,27 @@ function setParams() { if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 export CCL=${CCL:-ccl} # CCL - export BE="${CCL}" # BE = CCL + export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 - ####################################################### - # if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! + ############################################################## + # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" fi - ####################################################### + ############################################################## # +--------[Polaris]-----------------------------------+ elif [[ $(hostname) == x3* ]]; then - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL - # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? - export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - export GRAD_ACC_STEPS=4 # GRADIENT_ACCUMULATION_STEPS - MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL + # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? + export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-16} # GRADIENT_ACC_STEPS + # NOTE: MICRO_BATCH is exported below + MICRO_BATCH=${MICRO_BATCH:-1} # MICRO_BATCH = 8 if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else @@ -142,42 +154,40 @@ function setParams() { export TP="${TP}" export PP="${PP:-1}" export DTYPE="${DTYPE:-bf16}" - export OPT="${OPT:-adamw}" + export OPT="${OPT:-adamwschedulefree}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") if [[ -z "${NGPU_PER_HOST-}" ]]; then NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') fi export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" - # export WORLD_SIZE="${WORLD_SIZE:-${NGPUS:-$(( ))}}" - # export WORLD_SIZE=${WORLD_SIZE:-$(wc -l < "${HOSTFILE}")} - # +---[Llama2 7B Config]-----------------------------+ + # +---[Llama2 7B Config]--------------------------------------------------+ export MODEL_KEY="Llama-7B" - export HEADS=${HEADS:-${NHEADS:-32}} - export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} - export HIDDEN=${HIDDEN:-4096} - export NUM_KV_HEAD=${NUM_KV_HEAD:-8} - export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} + export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export LR=${LR:-0.0003} # LEARNING_RATE - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS - export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN - export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS - export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL - export SAVE_INTERVAL=${SAVE_INTERVAL:-200} # HOW FREQUENTLY TO SAVE CKPTS - export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS + export LR=${LR:-0.0003} # LEARNING_RATE + export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE + export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS + export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN + export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS + export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL + export SAVE_INTERVAL=${SAVE_INTERVAL:-200} # HOW FREQUENTLY TO SAVE CKPTS + export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - # +----------------------------------------------------------------------+ } @@ -312,7 +322,7 @@ function sumFiles() { # mine is called q4-drop ######################################################## setup_conda_sunspot() { - if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$(~/miniconda3/bin/conda shell hook -s posix)" conda activate q4-drop @@ -322,7 +332,7 @@ setup_conda_sunspot() { } setup_conda_sirius() { - if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" @@ -333,7 +343,9 @@ setup_conda_sirius() { } setup_conda_polaris() { - if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then + if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then + # export CUDA_HOME=/soft/compilers/cudatoolkit/cuda-12.2.2 + # && export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba && eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" ; mm activate 2024-04-25 export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" @@ -345,39 +357,55 @@ setup_conda_polaris() { function setEnv() { - # ---- [SunSpot] ------- || ---- [Aurora] -------------- - if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then - source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit - # ----- [Aurora] ----------------------------------- - if [[ -z "${CONDA_PREFIX}" && -z "${VIRTUAL_ENV}" ]]; then - if [[ $(hostname) == x4* ]]; then - eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 - # ----- [SunSpot] ---------------------------------- - elif [[ $(hostname) == x1* ]]; then - echo "Running on SunSpot !!" - setup_conda_sunspot - # eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + local virtual_env="${VIRTUAL_ENV-}" + local conda_prefix="${CONDA_PREFIX-}" + if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "Using conda from: ${conda_prefix}" + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix or virtual_env found in environment..." + echo "Setting up conda" + # setup_conda + # ---- [SunSpot] ------- || ---- [Aurora] -------------- + if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then + source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit + # ----- [Aurora] ----------------------------------- + if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + if [[ $(hostname) == x4* ]]; then + eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 + # ----- [SunSpot] ---------------------------------- + elif [[ $(hostname) == x1* ]]; then + echo "Running on SunSpot !!" + setup_conda_sunspot + # eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop + fi fi + # ----- [Polaris] --------------------------------------- + elif [[ $(hostname) == x3* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + echo "Running on Sirius !!" + setup_conda_sirius + else + echo "Running on Polaris !!" + # ---- [load conda] --------------------- + setup_conda_polaris + # if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then + # source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" + # fi + fi + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + echo "Running on Perlmutter !!" + module load pytorch + source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" + else # ------------------------------------- [Unknown] ------------------- + echo "Unknown hostname $(hostname)" + exit 1 fi - # ----- [Polaris] --------------------------------------- - elif [[ $(hostname) == x3* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then - echo "Running on Sirius !!" - setup_conda_sirius - else - echo "Running on Polaris !!" - # ---- [load conda] --------------------- - setup_conda_polaris - # if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then - # source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" - # fi - fi - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then - echo "Running on Perlmutter !!" - module load pytorch - source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" - else # ------------------------------------- [Unknown] ------------------- - echo "Unknown hostname $(hostname)" + else + echo "Unable to setup python environment. Exiting" exit 1 fi echo "[python] Using: $(which python3)" @@ -473,7 +501,7 @@ function generateDSconfig() { \"gradient_clipping\": 1.0, \"activation_checkpointing\": { \"partition_activations\": true, - \"contiguous_memory_optimization\": false + \"contiguous_memory_optimization\": true }, \"wall_clock_breakdown\": false," flops_profiler="\ From 2a49f6dcb55771fba1b8737060848a4f04ecdd2f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 13:59:49 -0500 Subject: [PATCH 114/590] Update `train_agpt.sh` --- train_agpt.sh | 1 + 1 file changed, 1 insertion(+) diff --git a/train_agpt.sh b/train_agpt.sh index 4129f553e28..a21b9880734 100644 --- a/train_agpt.sh +++ b/train_agpt.sh @@ -8,4 +8,5 @@ mkdir -p "${OUTDIR}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" echo "${OUTFILE}" >> "${OUTDIR}/latest" echo "Logging job output to: ${OUTFILE}" +# export DEBUG=1 bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From 482dffd1b27fa85f8f8b2365e65151624c078e95 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 26 Apr 2024 13:59:57 -0500 Subject: [PATCH 115/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 462da0c4e12..5d728068a5d 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -41,7 +41,7 @@ setParams || exit # 5. set command line arguments to pass to ` buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} setArgs || exit # 8. specify additional `deepspeed` arguments -setData "${DATA_FILE_LIST}"|| exit # 9. specify `DATA_FILE_LIST` for dolma dataset +setData "${DATA_FILE_LIST-}" || exit # 9. specify `DATA_FILE_LIST` for dolma dataset printJobInfo || exit # 11. print job info setupLauncher || exit # ----------------------------------------------------------------------------- @@ -75,6 +75,17 @@ data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" mkdir -p "${data_cache_path}" module list +if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then + TIMING_STR="\ + --timing-log-level ${TIMING_LOG_LEVEL} \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ + " +else + TIMING_STR="" +fi + + # Take custom args custom_args=" $@" @@ -83,7 +94,6 @@ custom_args=" $@" run_cmd=" ${LAUNCH_CMD} \ --${DTYPE} \ - --optimizer ${OPT} \ --split 100,0,0 \ --log-interval 1 \ --no-bias-gelu-fusion \ @@ -94,16 +104,17 @@ run_cmd=" --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ - --tensorboard-dir ${TBDIR} \ --log-timers-to-tensorboard \ --log-optimizer-states-to-tensorboard \ --lr ${LR} \ + --optimizer ${OPT} \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ --seq-length ${SEQ} \ --num-layers ${NLAYERS} \ --hidden-size ${HIDDEN} \ --train-iters ${TRAIN_ITER} \ + --tensorboard-dir ${TBDIR} \ --eval-iters ${EVAL_ITERS} \ --distributed-backend ${BE} \ --num-attention-heads ${HEADS} \ @@ -119,16 +130,16 @@ run_cmd=" --data-cache-path ${data_cache_path} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ --tokenizer-model ${TOKENIZER_MODEL} \ - --timing-log-level ${TIMING_LOG_LEVEL} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ + --lr-warmup-fraction ${LR_WARMUP_FRAC} \ ${LLAMA_ARGS} \ + ${TIMING_STR} \ $ds_args \ ${gpt_args[*]} \ $custom_args \ |& tee ${OUTPUT_LOG} " +check_and_kill_if_running || exit echo "${run_cmd}" printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" From c04c42da256eeb809a63b758a1f782514c243958 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 27 Apr 2024 00:22:39 -0500 Subject: [PATCH 116/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 516bc59f4cd..8976247b09f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -129,9 +129,9 @@ function setParams() { export BE="${NCCL}" # BE = NCCL # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-16} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS # NOTE: MICRO_BATCH is exported below - MICRO_BATCH=${MICRO_BATCH:-1} # MICRO_BATCH = 8 + MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else @@ -180,6 +180,7 @@ function setParams() { export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL export SAVE_INTERVAL=${SAVE_INTERVAL:-200} # HOW FREQUENTLY TO SAVE CKPTS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS + export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT @@ -215,7 +216,7 @@ function setArgs() { echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" gpt_args+=( "--checkpoint-activations" - "--checkpoint-num-layers 1" + "--checkpoint-num-layers ${ACT_CKPT_NUM_LAYERS}" ) fi export gpt_args @@ -432,7 +433,7 @@ function setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------ dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/books.txt" elif [[ "${PBS_O_HOST}" == polaris* ]]; then # shellcheck: source ./data-lists/polaris/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/books.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" From 36fa52093fa09602bbce890169623c2ac50def6a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 1 May 2024 12:56:41 -0500 Subject: [PATCH 117/590] Fix for `conda/2024-04-29` on Polaris --- ALCF/helpers.sh | 17 ++++++++++++----- pretrain_gpt_alcf.py | 5 ++++- 2 files changed, 16 insertions(+), 6 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8976247b09f..0f7b1f3559e 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -79,7 +79,7 @@ function loadCondaEnv() { function setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then - export LAUNCH_CMD="${DIST_LAUNCH} --cpu-bind depth -d 16 python3 -Wignore ${EXEC}" + export LAUNCH_CMD="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" else # Assert `./hostfile_deepspeed` exists export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit @@ -124,6 +124,7 @@ function setParams() { ############################################################## # +--------[Polaris]-----------------------------------+ elif [[ $(hostname) == x3* ]]; then + # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" TP=${TP:-1} # TP = 2 export NCCL=${NCCL:-nccl} # NCCL export BE="${NCCL}" # BE = NCCL @@ -347,10 +348,16 @@ setup_conda_polaris() { if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then # export CUDA_HOME=/soft/compilers/cudatoolkit/cuda-12.2.2 # && export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba && eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" ; mm activate 2024-04-25 - export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba - shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" - micromamba activate 2024-04-25 + # export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba + # shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') + # eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" + # micromamba activate 2024-04-25 + module use /soft/modulefiles + module load conda/2024-04-29 ; conda activate base + # unset MPICH_GPU_SUPPORT_ENABLED + # if [[ -d "${WORKING_DIR}/venvs/polaris/2024-04-29" ]]; then + # source "${WORKING_DIR}/venvs/polaris/2024-04-29/bin/activate" + # fi else echo "Found existing python at: $(which python3)" fi diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index f83ef05fc8d..2c521a400af 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -2,6 +2,7 @@ """Pretrain GPT""" +from mpi4py import MPI import os from rich import print import torch @@ -40,8 +41,10 @@ # ---- [SETUP COMMS] ------------------------ +# if str(os.environ.get('LAUNCH_CMD', 'mpich')).lower() == 'mpich': RANK = ez.setup_torch(backend="deepspeed") -# RANK = ez.get_rank() +# else: +# RANK = ez.get_rank() WORLD_SIZE = ez.get_world_size() LOCAL_RANK = ez.get_local_rank() DEVICE = ez.get_torch_device() From 3b83b36b1dac258b3ae40d6ba3a3fcfe41165b06 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 1 May 2024 12:57:54 -0500 Subject: [PATCH 118/590] Add `train_agpt_polaris_7B_production.sh` --- train_agpt_polaris_7B_production.sh | 13 +++++++++++++ 1 file changed, 13 insertions(+) create mode 100644 train_agpt_polaris_7B_production.sh diff --git a/train_agpt_polaris_7B_production.sh b/train_agpt_polaris_7B_production.sh new file mode 100644 index 00000000000..8b45ddc30ae --- /dev/null +++ b/train_agpt_polaris_7B_production.sh @@ -0,0 +1,13 @@ +#!/bin/bash --login + +NOW="$(date "+%Y-%m-%d-%H%M%S")" +cd "${PBS_O_WORKDIR}" || exit + +OUTDIR="${PBS_O_WORKDIR}/pbslogs" +mkdir -p "${OUTDIR}" +OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" +echo "${OUTFILE}" >> "${OUTDIR}/latest" +echo "Logging job output to: ${OUTFILE}" +export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 +# export DEBUG=1 +MICRO_BATCH=2 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From a916a8dd724e232c2db46b5ceb90143469681774 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 1 May 2024 13:50:58 -0500 Subject: [PATCH 119/590] Update `ALCF/helpers.sh` on Sunspot --- ALCF/helpers.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 0f7b1f3559e..d42888b56ec 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -113,6 +113,7 @@ function setParams() { export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 ############################################################## # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! @@ -326,7 +327,7 @@ function sumFiles() { setup_conda_sunspot() { if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$(~/miniconda3/bin/conda shell hook -s posix)" + eval "$(~/miniconda3/bin/conda shell.${shell_name} hook)" conda activate q4-drop else echo "Found existing python at: $(which python3)" From 5257721b69e40c635df1e42f08d04bd6f21cdf73 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 8 May 2024 08:27:09 -0500 Subject: [PATCH 120/590] Update `ALCF/test_alcf.sh` --- ALCF/test_alcf.sh | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ALCF/test_alcf.sh b/ALCF/test_alcf.sh index 367b40d17c2..853addc59d1 100644 --- a/ALCF/test_alcf.sh +++ b/ALCF/test_alcf.sh @@ -154,9 +154,9 @@ main() { echo "Unable to find / use ${DATA_FILE_LIST}. Exiting." exit 1 fi - export ZERO_STAGE=1 - export NUM_LAYERS=10 - export MICRO_BATCH=8 + # export ZERO_STAGE=1 + # export NUM_LAYERS=10 + # export MICRO_BATCH=8 export TRAIN_ITER=20 export TIMING_LOG_LEVEL=1 bash train_llama_alcf.sh |& tee "test-${SUBMITTED_FROM}-${NOW}".log From 1ad039c4d76bd661ddd4fddf59d14eb1fffce0b1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 9 May 2024 12:45:49 -0500 Subject: [PATCH 121/590] Add `train_agpt_polaris_7B_production_NCCL_OFI.sh` --- train_agpt_polaris_7B_production_NCCL_OFI.sh | 23 ++++++++++++++++++++ 1 file changed, 23 insertions(+) create mode 100644 train_agpt_polaris_7B_production_NCCL_OFI.sh diff --git a/train_agpt_polaris_7B_production_NCCL_OFI.sh b/train_agpt_polaris_7B_production_NCCL_OFI.sh new file mode 100644 index 00000000000..fd787dc80fe --- /dev/null +++ b/train_agpt_polaris_7B_production_NCCL_OFI.sh @@ -0,0 +1,23 @@ +#!/bin/bash --login + +NOW="$(date "+%Y-%m-%d-%H%M%S")" +cd "${PBS_O_WORKDIR}" || exit + +OUTDIR="${PBS_O_WORKDIR}/pbslogs" +mkdir -p "${OUTDIR}" +OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" +echo "${OUTFILE}" >> "${OUTDIR}/latest" +echo "Logging job output to: ${OUTFILE}" + +# export NCCL_NET_GDR_LEVEL=PHB +export NCCL_CROSS_NIC=1 +export NCCL_COLLNET_ENABLE=1 +export NCCL_NET="AWS Libfabric" +export LD_LIBRARY_PATH=/soft/libraries/aws-ofi-nccl/v1.9.1-aws/lib:$LD_LIBRARY_PATH +export LD_LIBRARY_PATH=/soft/libraries/hwloc/lib/:$LD_LIBRARY_PATH +export FI_CXI_DISABLE_HOST_REGISTER=1 +export FI_MR_CACHE_MONITOR=userfaultfd +export FI_CXI_DEFAULT_CQ_SIZE=131072 +# export DEBUG=1 +MICRO_BATCH=4 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" +# MICRO_BATCH=1 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/books.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From 7eb36a11b3a9c48ed07b93692ccf22bfb5577f7e Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Mon, 13 May 2024 15:25:56 +0530 Subject: [PATCH 122/590] get distributed backend name via accelerator and check loss_scale before writing to tb (#374) * check loss_scale before writing to tb * get distributed backend name via accelerator * add hccl distributed backend support --- megatron/arguments.py | 2 +- megatron/initialize.py | 2 +- megatron/training.py | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 631d4b12e87..dad993be04d 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1138,7 +1138,7 @@ def _add_distributed_args(parser): help='overlap pipeline parallel communication with forward and backward chunks', dest='overlap_p2p_comm') group.add_argument('--distributed-backend', default='nccl', - choices=['nccl', 'gloo', 'ccl'], + choices=['nccl', 'gloo', 'ccl', 'hccl'], help='Which backend to use for distributed training.') group.add_argument('--distributed-timeout-minutes', type=int, default=10, help='Timeout minutes for torch.distributed.') diff --git a/megatron/initialize.py b/megatron/initialize.py index 31f26c5086e..85c37a57675 100644 --- a/megatron/initialize.py +++ b/megatron/initialize.py @@ -222,7 +222,7 @@ def _initialize_distributed(): else: if not torch.distributed.is_initialized(): torch.distributed.init_process_group( - backend=args.distributed_backend, + backend=get_accelerator().communication_backend_name(), world_size=args.world_size, rank=args.rank, timeout=timedelta(minutes=args.distributed_timeout_minutes)) diff --git a/megatron/training.py b/megatron/training.py index 7b6acffedec..19b8a6c71f6 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -887,7 +887,7 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, args.consumed_train_samples) writer.add_scalar(f"lm-loss-training/{key}" + ' vs tokens', loss_dict[key], args.consumed_train_tokens) - if args.fp16 and args.log_loss_scale_to_tensorboard: + if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: writer.add_scalar('loss-scale/loss-scale', loss_scale, iteration) writer.add_scalar('loss-scale/loss-scale vs samples', loss_scale, args.consumed_train_samples) From 328dfda088ac98090627ac95e9abd9932f1d57a3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 May 2024 20:43:19 -0500 Subject: [PATCH 123/590] Create `flash-attn-fix` branch --- ALCF/aws_ofi_nccl_plugin.sh | 20 ++++ ALCF/helpers.sh | 103 ++++++++++++------- pretrain_gpt_alcf.py | 2 +- train_agpt_polaris_7B_production.sh | 28 +++-- train_agpt_polaris_7B_production_NCCL_OFI.sh | 23 ----- 5 files changed, 111 insertions(+), 65 deletions(-) create mode 100644 ALCF/aws_ofi_nccl_plugin.sh delete mode 100644 train_agpt_polaris_7B_production_NCCL_OFI.sh diff --git a/ALCF/aws_ofi_nccl_plugin.sh b/ALCF/aws_ofi_nccl_plugin.sh new file mode 100644 index 00000000000..ffd1471cd3b --- /dev/null +++ b/ALCF/aws_ofi_nccl_plugin.sh @@ -0,0 +1,20 @@ +#!/bin/bash --login + +# AWS NCCL OFI Plugin settings below +export NCCL_CROSS_NIC=1 +export NCCL_COLLNET_ENABLE=1 +export NCCL_NET="AWS Libfabric" +export LD_LIBRARY_PATH=/soft/libraries/aws-ofi-nccl/v1.9.1-aws/lib:$LD_LIBRARY_PATH +export LD_LIBRARY_PATH=/soft/libraries/hwloc/lib/:$LD_LIBRARY_PATH +export FI_CXI_DISABLE_HOST_REGISTER=1 +export FI_MR_CACHE_MONITOR=userfaultfd +export FI_CXI_DEFAULT_CQ_SIZE=131072 +######################################################### +# WARNING: !!! +# - Currently, `export NCCL_NET_GDR_LEVEL=PHB` +# causes a hang on Polaris. +# so, we don't set it for the time being [2024-05-14]. +# - Seems to work on Perlmutter ??? +# +# export NCCL_NET_GDR_LEVEL=PHB +######################################################### diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d42888b56ec..a257c3fb3f5 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -156,7 +156,7 @@ function setParams() { export TP="${TP}" export PP="${PP:-1}" export DTYPE="${DTYPE:-bf16}" - export OPT="${OPT:-adamwschedulefree}" + export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") if [[ -z "${NGPU_PER_HOST-}" ]]; then @@ -322,9 +322,8 @@ function sumFiles() { ######################################################## # Setup / activate conda environment, -# mine is called q4-drop ######################################################## -setup_conda_sunspot() { +setup_conda_sunspot() { # mine is called `q4-drop` on Sunspot if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') eval "$(~/miniconda3/bin/conda shell.${shell_name} hook)" @@ -334,6 +333,9 @@ setup_conda_sunspot() { fi } +######################## +# Setup conda on Sirius +######################## setup_conda_sirius() { if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba @@ -345,20 +347,27 @@ setup_conda_sirius() { fi } +######################## +# Setup conda on Polaris +######################## setup_conda_polaris() { - if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then - # export CUDA_HOME=/soft/compilers/cudatoolkit/cuda-12.2.2 - # && export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba && eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" ; mm activate 2024-04-25 - # export MAMBA_ROOT_PREFIX=/eagle/argonne_tpc/micromamba - # shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - # eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook -s posix)" - # micromamba activate 2024-04-25 + # unset MPICH_GPU_SUPPORT_ENABLED + ###### check if CONDA_PREFIX non-empty ################ + if [[ -z "${CONDA_PREFIX-}" ]]; then + # if so, load the default conda/2024-04-29 + # module and activate base environment module use /soft/modulefiles module load conda/2024-04-29 ; conda activate base - # unset MPICH_GPU_SUPPORT_ENABLED - # if [[ -d "${WORKING_DIR}/venvs/polaris/2024-04-29" ]]; then - # source "${WORKING_DIR}/venvs/polaris/2024-04-29/bin/activate" - # fi + else + echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + fi + ###### check if VIRTUAL_ENV non-empty ################# + if [[ -z "${VIRTUAL_ENV:-}" ]]; then + DEFAULT_VENV_PATH=${WORKING_DIR}/venvs/polaris/2024-04-29 + if [[ -d "${DEFAULT_VENV_PATH}" ]]; then + echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" + source "${WORKING_DIR}/venvs/polaris/2024-04-29/bin/activate" + fi else echo "Found existing python at: $(which python3)" fi @@ -369,42 +378,42 @@ function setEnv() { local virtual_env="${VIRTUAL_ENV-}" local conda_prefix="${CONDA_PREFIX-}" if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No virtual environment found." echo "Using conda from: ${conda_prefix}" elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "No conda found." echo "Using virtual_env from: ${virtual_env}" elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then - echo "Using virtual_env: ${virtual_env} on top of CONDA: ${conda_prefix}" + echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "No conda_prefix or virtual_env found in environment..." - echo "Setting up conda" - # setup_conda + echo "Setting up conda..." + ######################## setup_conda ############################ # ---- [SunSpot] ------- || ---- [Aurora] -------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit - # ----- [Aurora] ----------------------------------- + # ----- [Aurora] -------------------------------------------- if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then if [[ $(hostname) == x4* ]]; then eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 - # ----- [SunSpot] ---------------------------------- + # ----- [SunSpot] --------------------------------------- elif [[ $(hostname) == x1* ]]; then echo "Running on SunSpot !!" setup_conda_sunspot - # eval "$(/home/foremans/miniconda3/bin/conda shell.zsh hook)" && conda activate q4-drop fi fi - # ----- [Polaris] --------------------------------------- + # ----- [Polaris] ----------------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then echo "Running on Sirius !!" setup_conda_sirius else echo "Running on Polaris !!" - # ---- [load conda] --------------------- + # ---- [load conda] ------------------------------------- setup_conda_polaris - # if [[ -d "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221" ]]; then - # source "${PBS_O_WORKDIR}/venvs/polaris/cu118-pt221/bin/activate" - # fi fi + echo "Setting up AWS NCCL OFI Plugin on Polaris..." + source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" module load pytorch @@ -417,10 +426,19 @@ function setEnv() { echo "Unable to setup python environment. Exiting" exit 1 fi - echo "[python] Using: $(which python3)" + printf "\n" + pystr="Using: $(which python3)" + printf "[python] %s" "$(printMagenta ${pystr})" + printf "\n" } +###################################################################### +# `makeHostiles`: +# Detect if `HOSTFILE` set in active environment. +# - If so, use this. +# - Otherwise, make default HOSTFILEs from "${PBS_NODEFILE}" +###################################################################### function makeHostfiles() { if [[ -n "${HOSTFILE}" ]]; then printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" @@ -429,25 +447,40 @@ function makeHostfiles() { fi } -function setData() { # ---- [dfl: abbrv. for DATA_FILE_LIST] ------------------------- - if [[ $(hostname) == x4* ]]; then # ---- [AURORA] ---- +############################################### +# `setData`: +# Ensure `DATA_FILE_LIST` is set, +# fallback to default values if necessary. +############################################### +function setData() { # ----------------------[dfl: abbrv. for DATA_FILE_LIST] + # dfldir="${WORKING_DIR}/ALCF/data-lists" + # =====[Set DATA_FILE_LIST_FALLBACK based on current machine]============== + if [[ $(hostname) == x4* ]]; then # ---------------------------[AURORA] dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" - elif [[ $(hostname) == x1* ]]; then + + elif [[ $(hostname) == x1* ]]; then # --------------------------[SUNSPOT] # shellcheck: source ./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/books.txt" - elif [[ $(hostname) == x3* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then + + elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] + if [[ "${PBS_O_HOST}" == sirius* ]]; then # -------------------[SIRIUS] # shellcheck: source ./data-lists/sirius/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/books.txt" - elif [[ "${PBS_O_HOST}" == polaris* ]]; then + + elif [[ "${PBS_O_HOST}" == polaris* ]]; then # ---------------[POLARIS] # shellcheck: source ./data-lists/polaris/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" fi - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + + elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" - else + + else # -----------------------------------------------------------[UNKNOWN] echo "Unknown hostname. Must manually specify DATA_FILE_LIST." fi + # ========================================================================== + # set `dfl` to `dfl_fallback` if not passed as an argument, + # use this data file list to call `setData` dfl="${1:-${dfl_fallback}}" printf "Calling: setData() with %s\n" "${dfl}" ndocs=$(wc -l < "${dfl}") @@ -516,7 +549,7 @@ function generateDSconfig() { flops_profiler="\ \"flops_profiler\": { \"enabled\": true, - \"profile_step\": 4, + \"profile_step\": 2, \"module_depth\": -1, \"top_modules\": 1, \"detailed\": true, diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 2c521a400af..750bd21e48d 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -42,7 +42,7 @@ # ---- [SETUP COMMS] ------------------------ # if str(os.environ.get('LAUNCH_CMD', 'mpich')).lower() == 'mpich': -RANK = ez.setup_torch(backend="deepspeed") +RANK = ez.setup_torch(backend="deepspeed", timeout=7200) # else: # RANK = ez.get_rank() WORLD_SIZE = ez.get_world_size() diff --git a/train_agpt_polaris_7B_production.sh b/train_agpt_polaris_7B_production.sh index 8b45ddc30ae..f83b6ebc29b 100644 --- a/train_agpt_polaris_7B_production.sh +++ b/train_agpt_polaris_7B_production.sh @@ -1,13 +1,29 @@ #!/bin/bash --login +# +# This script can be submitted with `qsub` via: +# +# ```bash +# $ git clone https://github.com/argonee-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed +# $ qsub train_agpt_polaris_7B_production.sh +# ``` -NOW="$(date "+%Y-%m-%d-%H%M%S")" cd "${PBS_O_WORKDIR}" || exit -OUTDIR="${PBS_O_WORKDIR}/pbslogs" -mkdir -p "${OUTDIR}" +TODAY="$(date "+%Y-%m-%d")" +NOW="$(date "+%Y-%m-%d-%H%M%S")" +OUTDIR="${PBS_O_WORKDIR}/pbslogs/${TODAY}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" -echo "${OUTFILE}" >> "${OUTDIR}/latest" +mkdir -p $(dirname "${OUTFILE}") + +echo "${OUTFILE}" >> "$(dirname ${OUTDIR})/latest" echo "Logging job output to: ${OUTFILE}" -export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 + # export DEBUG=1 -MICRO_BATCH=2 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" +# export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 + +# Path to the data file list: +DFL="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" + +# Launch: +MICRO_BATCH=2 DATA_FILE_LIST="${DFL}" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" diff --git a/train_agpt_polaris_7B_production_NCCL_OFI.sh b/train_agpt_polaris_7B_production_NCCL_OFI.sh deleted file mode 100644 index fd787dc80fe..00000000000 --- a/train_agpt_polaris_7B_production_NCCL_OFI.sh +++ /dev/null @@ -1,23 +0,0 @@ -#!/bin/bash --login - -NOW="$(date "+%Y-%m-%d-%H%M%S")" -cd "${PBS_O_WORKDIR}" || exit - -OUTDIR="${PBS_O_WORKDIR}/pbslogs" -mkdir -p "${OUTDIR}" -OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" -echo "${OUTFILE}" >> "${OUTDIR}/latest" -echo "Logging job output to: ${OUTFILE}" - -# export NCCL_NET_GDR_LEVEL=PHB -export NCCL_CROSS_NIC=1 -export NCCL_COLLNET_ENABLE=1 -export NCCL_NET="AWS Libfabric" -export LD_LIBRARY_PATH=/soft/libraries/aws-ofi-nccl/v1.9.1-aws/lib:$LD_LIBRARY_PATH -export LD_LIBRARY_PATH=/soft/libraries/hwloc/lib/:$LD_LIBRARY_PATH -export FI_CXI_DISABLE_HOST_REGISTER=1 -export FI_MR_CACHE_MONITOR=userfaultfd -export FI_CXI_DEFAULT_CQ_SIZE=131072 -# export DEBUG=1 -MICRO_BATCH=4 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" -# MICRO_BATCH=1 DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/polaris/books.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From bd8cb0917cf0ca4700f68a0b33ffc26fc20e4718 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 May 2024 20:44:13 -0500 Subject: [PATCH 124/590] Add + pass default `LR_DECAY_ITERS` --- ALCF/helpers.sh | 9 +++++++-- train_llama_alcf.sh | 1 + 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a257c3fb3f5..e66d5ff500c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -173,6 +173,7 @@ function setParams() { # +---[Run Settings]------------------------------------------------------+ export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE @@ -281,7 +282,10 @@ function saveDSenv() { function setOutput() { # ---- Specify output location -------------------------------- - export OUTPUT_PREFIX="ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + export OUTPUT_PREFIX="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + if [[ -z "${NO_FLASH_ATTN:-}" ]]; then + export OUTPUT_PREFIX="${OUTPUT_PREFIX}_flash" + fi # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" @@ -289,7 +293,8 @@ function setOutput() { export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" echo "${OUTPUT_LOG}" >> "logs/latest" mkdir -p "${OUTPUT_DIR}" - echo "!!!Please see logs at ${OUTPUT_DIR}" + printf "Please see logs at: %s\n" $(printGreen "${OUTPUT_DIR}") + printf "Checkpoints will be saved to: %s\n" $(printYellow "${CKPT_DIR}") } function buildDSconfig() { diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 5d728068a5d..3e0f9390cb1 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -124,6 +124,7 @@ run_cmd=" --micro-batch-size ${MICRO_BATCH} \ --data-file-list ${DATA_FILE_LIST} \ --tensor-model-parallel-size ${TP} \ + --lr-decay-iters ${LR_DECAY_ITERS} \ --global-batch-size ${GLOBAL_BATCH} \ --pipeline-model-parallel-size ${PP} \ --num-key-value-heads ${NUM_KV_HEAD} \ From 59f60522b486738e443177296af43c5807f8a57b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 May 2024 20:45:16 -0500 Subject: [PATCH 125/590] Add `aGPT_7B.sh` --- aGPT_7B.sh | 34 ++++++++++++++++++++++++++++++++++ 1 file changed, 34 insertions(+) create mode 100644 aGPT_7B.sh diff --git a/aGPT_7B.sh b/aGPT_7B.sh new file mode 100644 index 00000000000..0c192b1b4f6 --- /dev/null +++ b/aGPT_7B.sh @@ -0,0 +1,34 @@ +#!/bin/bash --login +# +HOSTNAME=$(hostname) +if [[ "${HOSTNAME}" == x3* ]]; then + MACHINE="polaris" +elif [[ "${HOSTNAME}" == x1* ]]; then + MACHINE="sunspot" +elif [[ "${HOSTNAME}" == x4* ]]; then + MACHINE="aurora" +fi + +NOW="$(date "+%Y-%m-%d-%H%M%S")" +cd "${PBS_O_WORKDIR}" || exit +nhosts=$(wc -l < "${HOSTFILE}") + +if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then + MBS=1 +elif [[ "${nhosts}" -ge 2 ]]; then + MBS=2 +elif [[ "${nhosts}" -ge 8 ]]; then + MBS=4 +fi + +printf "Detected %s hosts. Running with micro_batch:\n" ${nhosts} ${MBS} + +OUTDIR="${PBS_O_WORKDIR}/pbslogs" +mkdir -p "${OUTDIR}" +OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" +echo "Running on: ${MACHINE}" +echo "${OUTFILE}" >> "${OUTDIR}/latest" +echo "Logging job output to: ${OUTFILE}" +export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 +# export DEBUG=1 +MICRO_BATCH="${MBS}" DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From 9f98c0912437f2abbd05c663e54c5329210c7d91 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 May 2024 20:46:21 -0500 Subject: [PATCH 126/590] Update `.gitignore` --- .gitignore | 2 ++ 1 file changed, 2 insertions(+) diff --git a/.gitignore b/.gitignore index 3b18d4956a0..0a1387f56af 100644 --- a/.gitignore +++ b/.gitignore @@ -1,4 +1,6 @@ # User Added +**.e1** +**.o1** deps/* OUTPUTS/* ALCF/OUTPUTS/* From 2cc2965c4831cda9621ce4781b3b24f478faa68c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 May 2024 20:47:54 -0500 Subject: [PATCH 127/590] Rename `aGPT_7B.sh` -> `train_aGPT_7B.sh` --- aGPT_7B.sh => train_aGPT_7B.sh | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename aGPT_7B.sh => train_aGPT_7B.sh (100%) diff --git a/aGPT_7B.sh b/train_aGPT_7B.sh similarity index 100% rename from aGPT_7B.sh rename to train_aGPT_7B.sh From 8b26db3b1369ea802ec53ce6e8cdb6ec1141ff40 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 15 May 2024 04:08:11 +0000 Subject: [PATCH 128/590] added pfw trace decorators --- megatron/arguments.py | 2 ++ megatron/checkpointing.py | 23 ++++++++++++----------- megatron/data/blendable_dataset.py | 7 +++++-- megatron/data/gpt_dataset.py | 22 +++++++++++++--------- megatron/data/indexed_dataset.py | 25 +++++++++++++++---------- megatron/training.py | 27 ++++++++++++++++++--------- megatron/utils.py | 25 +++++++++++++++++++++++++ pretrain_gpt_alcf.py | 4 +++- 8 files changed, 93 insertions(+), 42 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 702a4bad093..b7b71572e12 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -874,6 +874,8 @@ def _add_training_args(parser): 'training if SIGTERM is received') group.add_argument('--tensorboard-dir', type=str, default=None, help='Write TensorBoard logs to this directory.') + group.add_argument('--trace-dir', type=str, default="./trace/", + help='Write trace logs to this directory.') group.add_argument('--no-masked-softmax-fusion', action='store_false', help='Disable fusion of query_key_value scaling, ' diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index d585baf7177..b7f4b30bde8 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -21,6 +21,7 @@ from .utils import (unwrap_model, print_rank_0, is_rank_0) +from .utils import PerfTrace, Profile from deepspeed.checkpoint import ( ORIGINAL_VOCAB_SIZE, @@ -38,7 +39,7 @@ _CHECKPOINT_VERSION = None - +dlp = Profile("CHECKPOINT") def set_checkpoint_version(value): global _CHECKPOINT_VERSION if _CHECKPOINT_VERSION is not None: @@ -167,7 +168,7 @@ def get_checkpoint_tracker_filename(checkpoints_path): training to restart from.""" return os.path.join(checkpoints_path, 'latest_checkpointed_iteration.txt') - +@dlp.log def read_metadata(tracker_filename): # Read the tracker file and either set the iteration or # mark it as a release checkpoint. @@ -207,7 +208,7 @@ def read_metadata(tracker_filename): max_iter = iteration return max_iter, release - +@dlp.log def get_rng_state(): """ collect rng state across data parallel ranks """ args = get_args() @@ -233,7 +234,7 @@ def get_rng_state(): return rng_state_list - +@dlp.log def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): """Save a model checkpoint.""" args = get_args() @@ -338,7 +339,7 @@ def state_dict_for_save_checkpoint_deepspeed(destination=None, prefix='', keep_v if torch.distributed.is_initialized(): torch.distributed.barrier() - +@dlp.log def _transpose_first_dim(t, num_splits, num_splits_first, model): input_shape = t.size() # We use a self_attention module but the values extracted aren't @@ -408,7 +409,7 @@ def fix_query_key_value_ordering(model, checkpoint_version): print_rank_0(" succesfully fixed query-key-values ordering for" " checkpoint version {}".format(checkpoint_version)) - +@dlp.log def _load_base_checkpoint(load_dir, rank0=False): """ Load the base state_dict from the given directory @@ -463,7 +464,7 @@ def _load_base_checkpoint(load_dir, rank0=False): return state_dict, release - +@dlp.log def load_args_from_checkpoint(args, load_arg='load'): """Set required arguments from the checkpoint specified in the arguments. @@ -544,7 +545,7 @@ def _set_arg(arg_name, old_arg_name=None, force=False): _set_arg('num_layers_per_virtual_pipeline_stage') return args, checkpoint_args - +@dlp.log def load_lr_state_dict(strict: bool = False) -> dict: """Load {iteration, lr} from .yaml file when restoring from checkpoint.""" args = get_args() @@ -568,7 +569,7 @@ def load_lr_state_dict(strict: bool = False) -> dict: ) return lr_state_dict - +@dlp.log def save_lr_state_dict() -> None: """Save {iteration, lr} to .yaml file for safe-keeping. @@ -590,7 +591,7 @@ def save_lr_state_dict() -> None: f ) - +@dlp.log def load_checkpoint( model, optimizer, @@ -811,7 +812,7 @@ def load_checkpoint( return iteration - +@dlp.log def load_biencoder_checkpoint(model, only_query_model=False, only_context_model=False, custom_load_path=None): """ diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index f3276c68230..3830b5dfd4f 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -12,10 +12,12 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0 from megatron.core import mpu +from megatron.utils import Profile, PerfTrace +dlp = Profile("DATASET") class BlendableDataset(torch.utils.data.Dataset): - + @dlp.log def __init__(self, datasets, weights, size, *, data_cache_path=None): @@ -32,6 +34,7 @@ def __init__(self, datasets, weights, size, *, weights /= sum_weights # Build indicies. + @dlp.log def _build_indices(): start_time = time.time() dataset_index = np.zeros(self.size, dtype=np.int64) @@ -115,7 +118,7 @@ def _build_indices(): def __len__(self): return self.size - + @dlp.log def __getitem__(self, idx): dataset_idx = self.dataset_index[idx] sample_idx = self.dataset_sample_index[idx] diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 210a92c85e8..eb7de86fa3a 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -16,7 +16,11 @@ from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset +from megatron.utils import PerfTrace, Profile +dlp = Profile("DATASET") + +@dlp.log def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_valid_test_num_samples, seq_length, seed, skip_warmup, @@ -112,7 +116,7 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, return (train_dataset, valid_dataset, test_dataset) - +@dlp.log def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_valid_test_num_samples, seq_length, seed, skip_warmup, @@ -159,7 +163,7 @@ def build_dataset(index, name): return (train_dataset, valid_dataset, test_dataset) - +@dlp.log def build_dataset(dataset_name, data_prefix, data_impl, splits_string, num_samples, seq_length, seed, skip_warmup, @@ -194,7 +198,7 @@ def build_dataset(dataset_name, data_prefix, data_impl, return dataset - +@dlp.log def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, num_samples, seq_length, seed, skip_warmup, *, @@ -242,7 +246,7 @@ def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): class GPTDataset(torch.utils.data.Dataset): - + @dlp.log def __init__(self, name, data_prefix, documents, indexed_dataset, splits_string, num_samples, seq_length, seed, return_doc_ids=False, *, @@ -268,7 +272,7 @@ def __len__(self): # -1 is due to data structure used to retieve the index: # sample i --> [sample_idx[i], sample_idx[i+1]) return self.sample_idx.shape[0] - 1 - + @dlp.log def __getitem__(self, idx): args = get_args() orig_idx = idx @@ -341,7 +345,7 @@ def __getitem__(self, idx): return sample_dict - +@dlp.log def _build_index_mappings(name, data_prefix, documents, sizes, splits_string, num_samples, seq_length, seed, *, @@ -552,7 +556,7 @@ def _num_epochs(tokens_per_epoch, seq_length, num_samples): if ((total_tokens - 1) // seq_length) >= num_samples: return num_epochs - +@dlp.log def _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch): """Build an array with length = number-of-epochs * number-of-dcuments. Each index is mapped to a corresponding document.""" @@ -568,7 +572,7 @@ def _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch): doc_idx_last = _build_doc_idx(documents, 1, np_rng, False) return np.concatenate((doc_idx_first, doc_idx_last)) - +@dlp.log def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): """Sample index mapping is a 2D array with sizes @@ -617,7 +621,7 @@ def _build_sample_idx(sizes, doc_idx, seq_length, return sample_idx - +@dlp.log def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" print(' > building shuffle index with split [0, {}) and [{}, {}) ' diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index 62ebdc9813a..7810a6496bb 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -22,7 +22,9 @@ import numpy as np import torch from megatron import print_rank_0 +from megatron.utils import Profile +dlp = Profile("DATASET") def __best_fitting_dtype(vocab_size=None): if vocab_size is not None and vocab_size < 65500: @@ -136,7 +138,7 @@ def __init__(self, path): self.path = path self.data_file = None self.read_index(path) - + @dlp.log def read_index(self, path): with open(index_file_path(path), 'rb') as f: magic = f.read(8) @@ -154,7 +156,7 @@ def read_index(self, path): self.data_offsets = read_longs(f, self._len + 1) self.sizes = read_longs(f, self.s) self.doc_idx = read_longs(f, self.doc_count) - + @dlp.log def read_data(self, path): self.data_file = open(data_file_path(path), 'rb', buffering=0) @@ -167,6 +169,7 @@ def __del__(self): self.data_file.close() # @lru_cache(maxsize=8) + @dlp.log def __getitem__(self, idx): if not self.data_file: self.read_data(self.path) @@ -221,7 +224,7 @@ def __init__(self, path): @property def supports_prefetch(self): return True - + @dlp.log def prefetch(self, indices): if all(i in self.cache_index for i in indices): return @@ -247,6 +250,7 @@ def prefetch(self, indices): self.data_file = None # @lru_cache(maxsize=8) + @dlp.log def __getitem__(self, idx): if isinstance(idx, int): i = idx @@ -283,7 +287,7 @@ def __init__(self, out_file, dtype=np.int32): self.sizes = [] self.element_size = self.element_sizes[self.dtype] self.doc_idx = [0] - + @dlp.log def add_item(self, tensor): bytes = self.out_file.write(np.array(tensor.numpy(), dtype=self.dtype)) self.data_offsets.append(self.data_offsets[-1] + bytes / self.element_size) @@ -293,7 +297,7 @@ def add_item(self, tensor): def end_document(self): self.doc_idx.append(len(self.sizes)) - + @dlp.log def merge_file_(self, another_file): index = IndexedDataset(another_file) assert index.dtype == self.dtype @@ -333,7 +337,7 @@ def finalize(self, index_file): write_longs(index, self.doc_idx) index.close() - +@dlp.log def _warmup_mmap_file(path): with open(path, 'rb') as stream: while stream.read(100 * 1024 * 1024): @@ -399,7 +403,7 @@ def _get_pointers(sizes, npdtype): # compute element sizes in bytes pointers, _ = get_pointers_with_total(sizes, dtype().itemsize, npdtype) return pointers - + @dlp.log def write(self, sizes, doc_idx): self._file.write(struct.pack(' Date: Thu, 16 May 2024 10:43:50 -0500 Subject: [PATCH 129/590] Add `ALCF/aws_ofi_nccl_plugin.sh` for Polaris --- ALCF/helpers.sh | 44 +++++++++++++++++++++++++++++++++----------- train_aGPT_7B.sh | 25 ++++++++++++++++--------- train_llama_alcf.sh | 21 +++++++++++---------- 3 files changed, 60 insertions(+), 30 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index e66d5ff500c..1bd0febf248 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -16,6 +16,27 @@ export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} +function get_machine() { + if [[ $(hostname) == x4* ]]; then + machine="aurora" + elif [[ $(hostname) == x1* ]]; then + machine="sunspot" + elif [[ $(hostname) == x3* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + machine="sirius" + else + machine="polaris" + fi + elif [[ $(hostname) == nid* ]]; then + machine="perlmutter" + else + echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." + fi + export MACHINE="${machine}" + printf "Running on: %s\n" "$(printBlue ${MACHINE})" +} + + function check_and_kill_if_running() { # kill $(ps aux | grep -E "$USER.+(mpi|main.py)" | grep -v grep | awk '{print $2}') RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') @@ -139,6 +160,8 @@ function setParams() { else LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" fi + echo "Setting up AWS NCCL OFI Plugin on Polaris..." + source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # +--------[Perlmutter]---------------------------------+ elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then TP="${TP:-2}" @@ -293,7 +316,7 @@ function setOutput() { export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" echo "${OUTPUT_LOG}" >> "logs/latest" mkdir -p "${OUTPUT_DIR}" - printf "Please see logs at: %s\n" $(printGreen "${OUTPUT_DIR}") + printf "\n Please see logs at: %s\n" $(printGreen "${OUTPUT_DIR}") printf "Checkpoints will be saved to: %s\n" $(printYellow "${CKPT_DIR}") } @@ -394,12 +417,13 @@ function setEnv() { echo "No conda_prefix or virtual_env found in environment..." echo "Setting up conda..." ######################## setup_conda ############################ - # ---- [SunSpot] ------- || ---- [Aurora] -------------- + # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit # ----- [Aurora] -------------------------------------------- if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then if [[ $(hostname) == x4* ]]; then + # TODO: Update once Aurora back online eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 # ----- [SunSpot] --------------------------------------- elif [[ $(hostname) == x1* ]]; then @@ -407,7 +431,7 @@ function setEnv() { setup_conda_sunspot fi fi - # ----- [Polaris] ----------------------------------------------- + # ----- [Polaris @ ALCF] -------------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then echo "Running on Sirius !!" @@ -417,8 +441,7 @@ function setEnv() { # ---- [load conda] ------------------------------------- setup_conda_polaris fi - echo "Setting up AWS NCCL OFI Plugin on Polaris..." - source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit + # ----- [Perlmutter @ NERSC] ---------------------------------------- elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" module load pytorch @@ -431,10 +454,9 @@ function setEnv() { echo "Unable to setup python environment. Exiting" exit 1 fi - printf "\n" + ##################################################################### pystr="Using: $(which python3)" - printf "[python] %s" "$(printMagenta ${pystr})" - printf "\n" + printf "\n[python] %s\n" "$(printMagenta ${pystr})" } @@ -457,13 +479,13 @@ function makeHostfiles() { # Ensure `DATA_FILE_LIST` is set, # fallback to default values if necessary. ############################################### -function setData() { # ----------------------[dfl: abbrv. for DATA_FILE_LIST] +function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # dfldir="${WORKING_DIR}/ALCF/data-lists" # =====[Set DATA_FILE_LIST_FALLBACK based on current machine]============== - if [[ $(hostname) == x4* ]]; then # ---------------------------[AURORA] + if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" - elif [[ $(hostname) == x1* ]]; then # --------------------------[SUNSPOT] + elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] # shellcheck: source ./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/books.txt" diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 0c192b1b4f6..6bc8f1f9dcd 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -11,24 +11,31 @@ fi NOW="$(date "+%Y-%m-%d-%H%M%S")" cd "${PBS_O_WORKDIR}" || exit -nhosts=$(wc -l < "${HOSTFILE}") +export nhosts=$(wc -l < "${PBS_NODEFILE}") -if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then - MBS=1 +if [[ "${nhosts}" == 1 ]]; then + export MBS=1 +elif [[ "${nhosts}" == 2 ]]; then + export MBS=1 elif [[ "${nhosts}" -ge 2 ]]; then - MBS=2 + export MBS=2 elif [[ "${nhosts}" -ge 8 ]]; then - MBS=4 + export MBS=4 fi -printf "Detected %s hosts. Running with micro_batch:\n" ${nhosts} ${MBS} +# printf "Detected %s hosts. Running with micro_batch:\n" ${nhosts} ${MBS} OUTDIR="${PBS_O_WORKDIR}/pbslogs" mkdir -p "${OUTDIR}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" -echo "Running on: ${MACHINE}" -echo "${OUTFILE}" >> "${OUTDIR}/latest" -echo "Logging job output to: ${OUTFILE}" + +echo "+---------------------------------------------------------+" +echo "| Running on: ${MACHINE}" +echo "| Detected ${nhosts} hosts. Running with micro batch: ${MBS}" +echo "| Logging job output to: ${OUTFILE}" +echo "+---------------------------------------------------------+" + export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 +echo "${OUTFILE}" >> "${OUTDIR}/latest" # export DEBUG=1 MICRO_BATCH="${MBS}" DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 3e0f9390cb1..237bcf7c159 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -34,16 +34,17 @@ export EXEC="${HERE}/pretrain_gpt_alcf.py" sourceFile "${HERE}/ALCF/helpers.sh" || exit # ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ -setEnv || exit # 1. load `conda` environment -# saveDSenv || exit # 2. save env vars to `.deepspeed_env` -ezpz || exit # 3. determine WORLD_SIZE, etc. from `PBS_*` vars -setParams || exit # 5. set command line arguments to pass to `"${EXEC}"` -buildDSconfig || exit # 6. create `deepspeed_config.json` from runtime params from ^ -setOutput || exit # 7. specify output directory for {logs, checkpoints, etc.} -setArgs || exit # 8. specify additional `deepspeed` arguments -setData "${DATA_FILE_LIST-}" || exit # 9. specify `DATA_FILE_LIST` for dolma dataset -printJobInfo || exit # 11. print job info -setupLauncher || exit +get_machine || exit # 01. Identify machine we're on +setEnv || exit # 02. Load `conda` environment +# saveDSenv || exit # 03. Save env vars to `.deepspeed_env` +ezpz || exit # 04. Determine WORLD_SIZE, etc. from `PBS_*` vars +setParams || exit # 05. Set command line arguments to pass to `"${EXEC}"` +buildDSconfig || exit # 06. Create `deepspeed_config.json` from runtime params from ^ +setOutput || exit # 07. Specify output directory for {logs, checkpoints, etc.} +setArgs || exit # 08. Specify additional `deepspeed` arguments +setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset +printJobInfo || exit # 11. Print job info +setupLauncher || exit # 12. set launcher to one of `MPICH` (default), or `deepspeed` # ----------------------------------------------------------------------------- #### [DEPRECATED] ########################################################### From 6b4ea4c7085c0588214dca07b28a9fa76b0f6657 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 16 May 2024 10:49:20 -0500 Subject: [PATCH 130/590] Update `ALCF/{helpers.sh,train_llama_alcf.sh}` --- ALCF/helpers.sh | 52 +++++++++++++++++++++++++++++++++++---------- train_llama_alcf.sh | 13 ++++++------ 2 files changed, 48 insertions(+), 17 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 1bd0febf248..2422ecbe7a9 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -125,6 +125,18 @@ function setDSlauncher() { fi } +function set_lr_args() { + LR_ARGS="--lr ${LR} --lr-decay-style cosine" + if [[ -n "${LR_DECAY_ITERS:-}" ]]; then + LR_ARGS="${LR_ARGS} --lr-decay-iters ${LR_DECAY_ITERS}" + fi + if [[ -n "${LR_WARMUP_FRAC}" ]]; then + LR_ARGS="${LR_ARGS} --lr-warmup-fraction ${LR_WARMUP_FRAC}" + fi + echo "LR_ARGS: ${LR_ARGS}" + export LR_ARGS="${LR_ARGS}" +} + function setParams() { LLAMA_ARGS="" # +----[Parallelism Settings] -------------------------------------------+ @@ -194,9 +206,6 @@ function setParams() { export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export LR=${LR:-0.0003} # LEARNING_RATE - export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP - export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE @@ -215,6 +224,11 @@ function setParams() { export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" + export LR=${LR:-0.0003} # LEARNING_RATE + export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS + export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS + set_lr_args } @@ -305,10 +319,15 @@ function saveDSenv() { function setOutput() { # ---- Specify output location -------------------------------- - export OUTPUT_PREFIX="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}_hs${HIDDEN}_mb${MICRO_BATCH}_seq${SEQ}_gb${GLOBAL_BATCH}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + OUTPUT_PREFIX="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_hs${HIDDEN}_mb${MICRO_BATCH}" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_seq${SEQ}_gb${GLOBAL_BATCH}" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_lr${LR}_lwf${LR_WARMUP_FRAC}_ldi${LR_DECAY_ITERS}" if [[ -z "${NO_FLASH_ATTN:-}" ]]; then - export OUTPUT_PREFIX="${OUTPUT_PREFIX}_flash" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_flash" fi + export OUTPUT_PREFIX="${OUTPUT_PREFIX}" # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" @@ -351,11 +370,20 @@ function sumFiles() { ######################################################## # Setup / activate conda environment, ######################################################## -setup_conda_sunspot() { # mine is called `q4-drop` on Sunspot - if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then - shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$(~/miniconda3/bin/conda shell.${shell_name} hook)" - conda activate q4-drop +setup_conda_sunspot() { + ###### check if CONDA_PREFIX non-empty ################ + if [[ -z "${CONDA_PREFIX-}" ]]; then + module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 + else + echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + fi + ###### check if VIRTUAL_ENV non-empty ################# + if [[ -z "${VIRTUAL_ENV:-}" ]]; then + DEFAULT_VENV_PATH=${WORKING_DIR}/venvs/frameworks_2024_5_v2 + if [[ -d "${DEFAULT_VENV_PATH}" ]]; then + echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" + source "${WORKING_DIR}/${DEFAULT_VENV_PATH}/bin/activate" + fi else echo "Found existing python at: $(which python3)" fi @@ -456,7 +484,9 @@ function setEnv() { fi ##################################################################### pystr="Using: $(which python3)" - printf "\n[python] %s\n" "$(printMagenta ${pystr})" + printf "[python] %s" "$(printMagenta ${pystr})" + printf "\n" + export "PYTHON_EXEC=$(which python3)" } diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 237bcf7c159..d877c19a06e 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -75,6 +75,7 @@ mkdir -p "${TBDIR}" data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" mkdir -p "${data_cache_path}" module list +printenv > "${CKPT_DIR}/.env" if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then TIMING_STR="\ @@ -87,6 +88,8 @@ else fi + + # Take custom args custom_args=" $@" @@ -98,7 +101,6 @@ run_cmd=" --split 100,0,0 \ --log-interval 1 \ --no-bias-gelu-fusion \ - --lr-decay-style cosine \ --no-bias-dropout-fusion \ --no-masked-softmax-fusion \ --tokenizer-type Llama2Tokenizer \ @@ -107,7 +109,6 @@ run_cmd=" --use-checkpoint-opt_param-scheduler \ --log-timers-to-tensorboard \ --log-optimizer-states-to-tensorboard \ - --lr ${LR} \ --optimizer ${OPT} \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ @@ -125,14 +126,13 @@ run_cmd=" --micro-batch-size ${MICRO_BATCH} \ --data-file-list ${DATA_FILE_LIST} \ --tensor-model-parallel-size ${TP} \ - --lr-decay-iters ${LR_DECAY_ITERS} \ --global-batch-size ${GLOBAL_BATCH} \ --pipeline-model-parallel-size ${PP} \ --num-key-value-heads ${NUM_KV_HEAD} \ --data-cache-path ${data_cache_path} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ --tokenizer-model ${TOKENIZER_MODEL} \ - --lr-warmup-fraction ${LR_WARMUP_FRAC} \ + ${LR_ARGS} \ ${LLAMA_ARGS} \ ${TIMING_STR} \ $ds_args \ @@ -143,7 +143,8 @@ run_cmd=" check_and_kill_if_running || exit echo "${run_cmd}" -printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" -printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" +printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow ${OUTPUT_LOG})" +# printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" +# printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" eval "${run_cmd}" set +x From 14970b9a01a1055c143b1969dff89d6aed99bfb1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 16 May 2024 11:54:21 -0500 Subject: [PATCH 131/590] Update `ALCF/helpers.sh` on Sunspot --- ALCF/helpers.sh | 34 +++++++++++++++++++++++----------- 1 file changed, 23 insertions(+), 11 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 2422ecbe7a9..9fe9934dab2 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -372,18 +372,30 @@ function sumFiles() { ######################################################## setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ - if [[ -z "${CONDA_PREFIX-}" ]]; then - module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 - else - echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + if [[ -z "${CONDA_PREFIX:-}" ]]; then + eval "$(~/miniconda3/bin/conda shell.zsh hook)" + conda activate anl_24_q2_release fi - ###### check if VIRTUAL_ENV non-empty ################# + # ------------------------------------------------------------------------ + # XXX: Jerome's `frameworks_2024_5_v2` seems broken ?? + # - seems to be missing `python3 -c 'from mpi4py import MPI'` ??? + # - consequently, we leave the setup below commented out (for the time + # being): + # if [[ -z "${CONDA_PREFIX-}" ]]; then + # module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 + # else + # echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + # fi + # + # ------------------------------------------------------------------------ + ###### check if VIRTUAL_ENV non-empty #################################### if [[ -z "${VIRTUAL_ENV:-}" ]]; then - DEFAULT_VENV_PATH=${WORKING_DIR}/venvs/frameworks_2024_5_v2 - if [[ -d "${DEFAULT_VENV_PATH}" ]]; then - echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" - source "${WORKING_DIR}/${DEFAULT_VENV_PATH}/bin/activate" - fi + DEFAULT_VENV_PATH="${WORKING_DIR}/venvs/anl_24_q2_release" + # venvs/anl_24_q2_release/bin/activate + # if [[ -d "${DEFAULT_VENV_PATH}" ]]; then + echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" + source "${DEFAULT_VENV_PATH}/bin/activate" || exit + # fi else echo "Found existing python at: $(which python3)" fi @@ -447,7 +459,6 @@ function setEnv() { ######################## setup_conda ############################ # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then - source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit # ----- [Aurora] -------------------------------------------- if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then if [[ $(hostname) == x4* ]]; then @@ -459,6 +470,7 @@ function setEnv() { setup_conda_sunspot fi fi + source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit # ----- [Polaris @ ALCF] -------------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then From d1aec5d7afe4ca68196e325f32bd185584caaf9a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 16 May 2024 11:54:55 -0500 Subject: [PATCH 132/590] Update `ALCF/sunspot-env.sh` with new modules for `anl_24_q2_release` --- ALCF/sunspot-env.sh | 31 ++++++------------------------- 1 file changed, 6 insertions(+), 25 deletions(-) diff --git a/ALCF/sunspot-env.sh b/ALCF/sunspot-env.sh index e97545aafec..8b02542b20e 100644 --- a/ALCF/sunspot-env.sh +++ b/ALCF/sunspot-env.sh @@ -1,27 +1,8 @@ #!/bin/bash --login - -export CCL_OP_SYNC=1 # Required by current oneCCL (HPCS-8067) -export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich -export FI_PROVIDER=cxi # Required by Aurora mpich -export PALS_PMI=pmix # Required by Aurora mpich -export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich -export FI_MR_CACHE_MONITOR=disabled # Required by Aurora mpich (HPCS-6501) -export CCL_SKIP_SCHEDULER=1 # Required by current oneCCL, will remove when set by default -export SYCL_PI_LEVEL_ZERO_USE_IMMEDIATE_COMMANDLISTS=1 -export CCL_TOPO_COLOR="card:{0,1},{2,3},{4,5},{6,7},{8,9},{10,11};plane:{0,3,4,6,8,11},{1,2,5,7,9,10}" -export UR_L0_IN_ORDER_BARRIER_BY_SIGNAL=0 # Required by current oneCCL - - -export LLM_DK_DIR=/home/$(whoami)/q4-drop_sunspot/llm.devkit - -module load oneapi/release/2023.12.15.001 -unset MPIR_CVAR_CH4_COLL_SELECTION_TUNING_JSON_FILE -unset MPIR_CVAR_COLL_SELECTION_TUNING_JSON_FILE -unset MPIR_CVAR_CH4_POSIX_COLL_SELECTION_TUNING_JSON_FILE - - +# module use /home/ftartagl/graphics-compute-runtime/modulefiles -module load graphics-compute-runtime/agama-ci-devel-736.9 -source /home/$(whoami)/q4-drop_sunspot/llm.devkit/torch-ccl/third_party/oneCCL/build/_install/env/vars.sh -module load gcc/12.1.0 -module unload intel_compute_runtime/release/agama-devel-647 +module load graphics-compute-runtime/agama-ci-devel-803.29 +module load spack-pe-gcc/0.6.1-23.275.2 +module load gcc/12.2.0 +module use /soft/preview-modulefiles/24.086.0 +module load oneapi/release/2024.04.15.001 From 13700cfc05ef71f6052cf876f8431d5d2eba58bc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 17 May 2024 07:19:09 -0500 Subject: [PATCH 133/590] Update `ALCF/helpers.sh` with new release on Sunspot --- ALCF/helpers.sh | 88 +++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 70 insertions(+), 18 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 9fe9934dab2..a805a34e289 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -229,6 +229,15 @@ function setParams() { # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args + if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then + TIMING_STR="\ + --timing-log-level ${TIMING_LOG_LEVEL} \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard \ + " + else + TIMING_STR="" + fi } @@ -323,7 +332,10 @@ function setOutput() { OUTPUT_PREFIX="${OUTPUT_PREFIX}_hs${HIDDEN}_mb${MICRO_BATCH}" OUTPUT_PREFIX="${OUTPUT_PREFIX}_seq${SEQ}_gb${GLOBAL_BATCH}" OUTPUT_PREFIX="${OUTPUT_PREFIX}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" - OUTPUT_PREFIX="${OUTPUT_PREFIX}_lr${LR}_lwf${LR_WARMUP_FRAC}_ldi${LR_DECAY_ITERS}" + OUTPUT_PREFIX="${OUTPUT_PREFIX}_lr${LR}_lwf${LR_WARMUP_FRAC}" + if [[ -n "${LR_DECAY_ITERS}" ]]; then + OUTPUT_PREFIX="${OUTPUT_PREFIX}_ldi${LR_DECAY_ITERS}" + fi if [[ -z "${NO_FLASH_ATTN:-}" ]]; then OUTPUT_PREFIX="${OUTPUT_PREFIX}_flash" fi @@ -373,29 +385,39 @@ function sumFiles() { setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ if [[ -z "${CONDA_PREFIX:-}" ]]; then - eval "$(~/miniconda3/bin/conda shell.zsh hook)" - conda activate anl_24_q2_release + # eval "$(~/miniconda3/bin/conda shell.zsh hook)" + # conda activate anl_24_q2_release + module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua fi - # ------------------------------------------------------------------------ - # XXX: Jerome's `frameworks_2024_5_v2` seems broken ?? + # XXX: ------------------------------------------------------------------- + # Jerome's `frameworks_2024_5_v2` seems broken ?? # - seems to be missing `python3 -c 'from mpi4py import MPI'` ??? # - consequently, we leave the setup below commented out (for the time # being): - # if [[ -z "${CONDA_PREFIX-}" ]]; then - # module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 - # else - # echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" - # fi - # + # if [[ -z "${CONDA_PREFIX-}" ]]; then + # module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 + # else + # echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + # fi # ------------------------------------------------------------------------ + ###### check if VIRTUAL_ENV non-empty #################################### + # venvs/anl_24_q2_release/bin/activate + # if [[ -d "${DEFAULT_VENV_PATH}" ]]; then if [[ -z "${VIRTUAL_ENV:-}" ]]; then - DEFAULT_VENV_PATH="${WORKING_DIR}/venvs/anl_24_q2_release" - # venvs/anl_24_q2_release/bin/activate - # if [[ -d "${DEFAULT_VENV_PATH}" ]]; then - echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" - source "${DEFAULT_VENV_PATH}/bin/activate" || exit - # fi + if [[ -n "${CONDA_PREFIX}" ]]; then + VENV_DIR="${WORKING_DIR}/venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')" + else + VENV_DIR="${WORKING_DIR}/venvs/anl_24_q2_release" + fi + echo "Caught virtual env at ${VENV_DIR}!" + # source "${VENV_DIR}/bin/activate" || + if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then + printf "[!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" + # echo "[!ERROR]: Unable to locate ${VENV_DIR}/bin/activate !!" + else + source "${VENV_DIR}/bin/activate" + fi else echo "Found existing python at: $(which python3)" fi @@ -415,6 +437,10 @@ setup_conda_sirius() { fi } +setup_venv_from_conda() { + source "venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')/bin/activate" +} + ######################## # Setup conda on Polaris ######################## @@ -470,7 +496,12 @@ function setEnv() { setup_conda_sunspot fi fi - source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit + # MPICH_MODULES=$(echo $LOADEDMODULES | tr ':' '\n' | grep mpich) + # if [[ -z "${MPICH_MODULES" ]]; then + # source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit + # else + # echo "Caught MPICH_MODULES: ${MPICH_MODULES}" + # fi # ----- [Polaris @ ALCF] -------------------------------------------- elif [[ $(hostname) == x3* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then @@ -764,3 +795,24 @@ function printCyan() { function printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } + +#### [DEPRECATED] ########################################################### +# if [[ -z "${HOSTFILE}" ]]; then +# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` +# else +# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" +# fi +# ---------------------------------------------------------------------------- +# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` +# ---------------------------------------------------------------------------- +# TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') +# printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" +# if [[ "${TORCH_DEVICE}" == "cuda" ]]; then +# printf %s "Setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True" +# PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True +# fi +# ---------------------------------------------------------------------------- +# export MPICH_GPU_SUPPORT_ENABLED=1 +# export CUDA_DEVICE_MAX_CONNECTIONS=1 +# export NCCL_DEBUG=INFO +############################################################################# From 530c7c80cee7ff10a01b432fed01e5e642fc0cd2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 17 May 2024 07:19:30 -0500 Subject: [PATCH 134/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 65 +++++++++++++++++---------------------------- 1 file changed, 24 insertions(+), 41 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index d877c19a06e..b1d4ed9825f 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -5,12 +5,25 @@ #PBS -l select=48 #PBS -l filesystems=eagle:home -if [[ -n "${DEBUG-}" ]]; then +#### Make it easy to track experiments by date ################### +YEAR="$(date "+%Y")" +MONTH="$(date "+%m")" +DAY="$(date "+%Y-%m-%d")" +TODAY="$(date "+%Y-%m-%d")" # kept for backwards compatibility +STARTED_AT="$(date "+%Y-%m-%d-%H%M%S")" +################################################################## + +if [[ -n "${DEBUG-}" ]]; then # to use: `DEBUG=1 bash train_llama_alcf.sh` printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" set -euxo pipefail fi -function sourceFile() { +if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` + echo "Run NOOP mode" + set -o noexec # same as set -n +fi + +sourceFile() { fp="$1" echo "source-ing ${fp}" if [[ -f "${fp}" ]]; then @@ -21,7 +34,7 @@ function sourceFile() { fi } -# ----[0. Navigate into `$PBS_O_WORKDIR`]------------------------------------- +# ----[0. Navigate into `$PBS_O_WORKDIR`]-------------------------------------- cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') export HERE @@ -47,48 +60,18 @@ printJobInfo || exit # 11. Print job info setupLauncher || exit # 12. set launcher to one of `MPICH` (default), or `deepspeed` # ----------------------------------------------------------------------------- -#### [DEPRECATED] ########################################################### -# if [[ -z "${HOSTFILE}" ]]; then -# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -# else -# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" -# fi -# ---------------------------------------------------------------------------- -# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` -# ---------------------------------------------------------------------------- -# TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') -# printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" -# if [[ "${TORCH_DEVICE}" == "cuda" ]]; then -# printf %s "Setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True" -# PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True -# fi -# ---------------------------------------------------------------------------- -# export MPICH_GPU_SUPPORT_ENABLED=1 -# export CUDA_DEVICE_MAX_CONNECTIONS=1 -# export NCCL_DEBUG=INFO -############################################################################# - -# Assert TBDIR exists inside our $CKPT_DIR +################################################ +# Assert `$TBDIR` exists inside our `$CKPT_DIR` +# to ensure metrics are tied to checkpoint +################################################ TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" -data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" -mkdir -p "${data_cache_path}" -module list -printenv > "${CKPT_DIR}/.env" - -if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then - TIMING_STR="\ - --timing-log-level ${TIMING_LOG_LEVEL} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ - " -else - TIMING_STR="" -fi - - +data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" +# Print info about loaded modules and runtime environment +module list +printenv |& tee "${CKPT_DIR}/.env" # Take custom args custom_args=" $@" From b8cb2e8bfd3459d2f08a6c55b9b07f9f20447ac6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 17 May 2024 07:51:15 -0500 Subject: [PATCH 135/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 6bc8f1f9dcd..d2719b088eb 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -38,4 +38,6 @@ echo "+---------------------------------------------------------+" export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 echo "${OUTFILE}" >> "${OUTDIR}/latest" # export DEBUG=1 -MICRO_BATCH="${MBS}" DATA_FILE_LIST="${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" +MICRO_BATCH="${MICRO_BATCH:-${MBS}}" +DATA_FILE_LIST="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt}" +bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From 0f180319cc9e48b4c52b32400a26297a91937c40 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 09:43:45 -0500 Subject: [PATCH 136/590] Add `setup_venv_from_conda` fn to `ALCF/helpers.sh` --- ALCF/helpers.sh | 33 ++++++++++++++++++++++++++++++++- 1 file changed, 32 insertions(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a805a34e289..9e7e2b582bb 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -438,7 +438,38 @@ setup_conda_sirius() { } setup_venv_from_conda() { - source "venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')/bin/activate" + if [[ -z "${CONDA_PREFIX}" ]]; then + echo "No ${CONDA_PREFIX} found." # Exiting." + # exit 1 + else + if [[ -n "${VIRTUAL_ENV}" ]]; then + echo "Already inside virtual env at ${VENV_DIR}!" + elif [[ -z "${VIRTUAL_ENV}" ]]; then + echo "No VIRTUAL_ENV found in environment!" + echo " - Trying to setup from ${CONDA_PREFIX}" + CONDA_NAME=$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') + VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" + echo " - Using VENV_DIR=${VENV_DIR}" + # VENV_DIR="venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}')" + # VENV_DIR="${WORKING_DIR}/venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')" + # VENV_DIR="${WORKING_DIR}/venvs/anl_24_q2_release" + # if [[ -f "${VENV_DIR}/bin/activate" ]]; then + if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then + printf "\n - Creating a new virtual env on top of %s in %s" "$(printBlue "${CONDA_NAME}")" "$(printGreen "${VENV_DIR}")" + mkdir -p "${VENV_DIR}" + python3 -m venv "${VENV_DIR}" --system-site-packages + source "${VENV_DIR}/bin/activate" || exit + elif [[ -f "${VENV_DIR}/bin/activate" ]]; then + echo " - Found existing venv, activating from $(printBlue "${VENV_DIR}")" + source "${VENV_DIR}/bin/activate" + else + printf "\n [!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" + fi + fi + # else + # printf "[!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" + fi + } ######################## From 061e2cc9e0f4a6c98e857b850f0113dc310f1b09 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 09:44:17 -0500 Subject: [PATCH 137/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index d2719b088eb..0907f8163c6 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -1,30 +1,34 @@ #!/bin/bash --login # + +NOW="$(date "+%Y-%m-%d-%H%M%S")" +cd "${PBS_O_WORKDIR}" || exit + HOSTNAME=$(hostname) if [[ "${HOSTNAME}" == x3* ]]; then MACHINE="polaris" + # XXX: ¯\_(ツ)_/¯ + # - On Polaris, we see that: + # - on 1 or 2 nodes, only MICRO_BATCH=1 will fit in memory + # - on 8 nodes, MICRO_BATCH=2 will fit in memory + # - on 48 nodes, MICRO_BATCH=4 will fit in memory + # + export nhosts=$(wc -l < "${PBS_NODEFILE}") + if [[ "${nhosts}" == 1 ]]; then + export MBS=1 + elif [[ "${nhosts}" == 2 ]]; then + export MBS=1 + elif [[ "${nhosts}" -ge 3 ]]; then + export MBS=2 + elif [[ "${nhosts}" -ge 8 ]]; then + export MBS=4 + fi elif [[ "${HOSTNAME}" == x1* ]]; then MACHINE="sunspot" elif [[ "${HOSTNAME}" == x4* ]]; then MACHINE="aurora" fi -NOW="$(date "+%Y-%m-%d-%H%M%S")" -cd "${PBS_O_WORKDIR}" || exit -export nhosts=$(wc -l < "${PBS_NODEFILE}") - -if [[ "${nhosts}" == 1 ]]; then - export MBS=1 -elif [[ "${nhosts}" == 2 ]]; then - export MBS=1 -elif [[ "${nhosts}" -ge 2 ]]; then - export MBS=2 -elif [[ "${nhosts}" -ge 8 ]]; then - export MBS=4 -fi - -# printf "Detected %s hosts. Running with micro_batch:\n" ${nhosts} ${MBS} - OUTDIR="${PBS_O_WORKDIR}/pbslogs" mkdir -p "${OUTDIR}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" @@ -38,6 +42,6 @@ echo "+---------------------------------------------------------+" export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 echo "${OUTFILE}" >> "${OUTDIR}/latest" # export DEBUG=1 -MICRO_BATCH="${MICRO_BATCH:-${MBS}}" -DATA_FILE_LIST="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt}" +export MICRO_BATCH="${MICRO_BATCH:-${MBS}}" +export DATA_FILE_LIST="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt}" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From 47bf9b50088069f960a9ee52d4d535530427d8ae Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 09:44:34 -0500 Subject: [PATCH 138/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index b1d4ed9825f..fee2e392dc3 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -71,7 +71,9 @@ data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path} # Print info about loaded modules and runtime environment module list -printenv |& tee "${CKPT_DIR}/.env" +dotenv_file="${CKPT_DIR}/.env" +echo "Saving environment to ${dotenv_file}" +printenv | grep -v "LS_COLORS" > "${dotenv_file}" # Take custom args custom_args=" $@" From e68d270faf4780018ae954f38e83b002adadb55f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 09:47:57 -0500 Subject: [PATCH 139/590] Update `ALCF/README.md` --- ALCF/README.md | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index d405986fdf7..8c097c699a0 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,4 +1,4 @@ -# Megatron-DeepSpeed @ ALCF +- [>] # Megatron-DeepSpeed @ ALCF ## ✅ TODOs @@ -82,13 +82,16 @@ To launch: +### Polaris + ```bash $ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I $ cd /path/to/Megatron-DeepSpeed/ -# load your favorite {conda, venv} environment, requires: {pytorch, deepspeed} -# e.g. on Polaris: $ module load conda/2023-10-04 # ; conda activate cu118-pt221 ; unset PYTHONUSERBASE -$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./convergence_debug_small.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh +$ module use /soft/modulefiles +$ module load conda/2024-04-29 +$ conda activate base +$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh ```
[output]: From ac414a0479b69f4968405c682cd49700eb3faebb Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 09:57:28 -0500 Subject: [PATCH 140/590] Update README.md --- ALCF/README.md | 60 ++++++++++++++++++++++++++++++++------------------ 1 file changed, 39 insertions(+), 21 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 8c097c699a0..0b436792d99 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,4 +1,4 @@ -- [>] # Megatron-DeepSpeed @ ALCF +# Megatron-DeepSpeed @ ALCF ## ✅ TODOs @@ -80,19 +80,35 @@ > [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching > distributed training on {Polaris, Aurora, Sunspot} @ ALCF. -To launch: - -### Polaris - -```bash -$ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I -$ cd /path/to/Megatron-DeepSpeed/ -$ module load conda/2023-10-04 # ; conda activate cu118-pt221 ; unset PYTHONUSERBASE -$ module use /soft/modulefiles -$ module load conda/2024-04-29 -$ conda activate base -$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh -``` +To launch on Polaris @ ALCF: + +> [!IMPORTANT] +> **Launch Instructions** on Polaris @ ALCF +> +> 1. Request an interactive job with `qsub -I`: +> +> ```bash +> $ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I +> ``` +> +> 2. Clone repo + navigate into it: +> ```sh +> $ cd /path/to/Megatron-DeepSpeed/ # or git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +> ``` +> +> 3. Load required modules: +> +> ```sh +> $ module use /soft/modulefiles +> $ module load conda/2024-04-29 +> $ conda activate base +> ``` +> +> 4. Launch: +> +> ```sh +> $ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh +> ```
[output]: @@ -461,6 +477,8 @@ training ...
+ + + + ## 📝 Data Preprocessing
Data Pre-Processing: From 9aa7fabfc70ead7470183c31e9ea5c7449d6da08 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 20 May 2024 13:26:09 -0500 Subject: [PATCH 141/590] Fix path in `prof.export_chrome_trace()` from `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 750bd21e48d..94186fea83b 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -2,6 +2,7 @@ """Pretrain GPT""" +from pathlib import Path from mpi4py import MPI import os from rich import print @@ -581,9 +582,13 @@ def main(): data_post_process=data_post_process ) args = get_args() - prof.export_chrome_trace( - f"{args.tensorboard_dir}" - "/torch-trace-{RANK}-of-{WORLD_SIZE}.json" + assert args is not None + trace_output = Path(f"{args.tensorboard_dir}").joinpath( + f"torch-trace-{RANK}-of-{WORLD_SIZE}.json" + ) + prof.export_chrome_trace(trace_output.as_posix()) + log.info( + f'Saved trace output to: {trace_output.as_posix()}' ) else: model = pretrain( From 0508cf6194835482c02705360b8b0b28508b1f25 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 20 May 2024 20:31:43 +0000 Subject: [PATCH 142/590] changed environment variable --- pretrain_gpt_alcf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 2a15ed52abd..2334a585a0e 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -571,7 +571,7 @@ def git_ds_info(): def main(): - if os.getenv('TORCH_PROFILER_ENABLED') == '1': + if os.getenv('TORCH_PROFILER_ENABLE') == '1': from torch.profiler import profile, record_function, ProfilerActivity with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA]) as prof: model = pretrain( From c4250a110c1891901b15082c66931f609372362c Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 20 May 2024 22:05:14 +0000 Subject: [PATCH 143/590] added torch profiler per step output support --- megatron/core/pipeline_parallel/schedules.py | 27 ++++++++------- megatron/training.py | 36 ++++++++++++++------ 2 files changed, 41 insertions(+), 22 deletions(-) diff --git a/megatron/core/pipeline_parallel/schedules.py b/megatron/core/pipeline_parallel/schedules.py index 453019f8c9d..c24959c64f1 100644 --- a/megatron/core/pipeline_parallel/schedules.py +++ b/megatron/core/pipeline_parallel/schedules.py @@ -17,7 +17,9 @@ from megatron.utils import print_rank_0, unwrap_model from megatron.model import DistributedDataParallel as LocalDDP from megatron.model import Float16Module +from megatron.utils import Profile +dlp = Profile("CORE") # Types Shape = Union[List[int], torch.Size] @@ -124,6 +126,7 @@ def deallocate_output_tensor(out, deallocate_pipeline_outputs=False): dtype = out.dtype, ) +@dlp.log def custom_backward(output, grad_output): '''Directly call C++ autograd engine. @@ -162,7 +165,7 @@ def custom_backward(output, grad_output): - +@dlp.log def forward_step(forward_step_func, data_iterator, model, @@ -227,7 +230,7 @@ def forward_step(forward_step_func, return output_tensor return [output_tensor] - +@dlp.log def backward_step( input_tensor, output_tensor, @@ -305,7 +308,7 @@ def backward_step( config.timers('backward-compute').stop() return input_tensor_grad - +@dlp.log def forward_backward_no_pipelining(*, forward_step_func, data_iterator: Union[Iterator, List[Iterator]], @@ -352,7 +355,7 @@ def forward_backward_no_pipelining(*, forward_data_store = [] input_tensor, output_tensor_grad = None, None with no_sync_func(): - for i in range(num_microbatches - 1): + for i in dlp.iter(range(num_microbatches - 1)): output_tensor = forward_step(forward_step_func, data_iterator, model, num_microbatches, input_tensor, forward_data_store, config, collect_non_loss_data) if not forward_only: @@ -370,7 +373,7 @@ def forward_backward_no_pipelining(*, return forward_data_store - +@dlp.log def forward_backward_pipelining_with_interleaving(*, forward_step_func, data_iterator: Union[Iterator, List[Iterator]], @@ -923,7 +926,7 @@ def get_tensor_shapes(*, return tensor_shapes - +@dlp.log def recv_forward(tensor_shapes, config): input_tensors = [] for tensor_shape in tensor_shapes: @@ -933,7 +936,7 @@ def recv_forward(tensor_shapes, config): input_tensors.append(p2p_communication.recv_forward(tensor_shape, config)) return input_tensors - +@dlp.log def recv_backward(tensor_shapes, config): output_tensor_grads = [] for tensor_shape in tensor_shapes: @@ -943,7 +946,7 @@ def recv_backward(tensor_shapes, config): output_tensor_grads.append(p2p_communication.recv_backward(tensor_shape, config)) return output_tensor_grads - +@dlp.log def send_forward(output_tensors, tensor_shapes, config): if not isinstance(output_tensors, list): output_tensors = [output_tensors] @@ -952,7 +955,7 @@ def send_forward(output_tensors, tensor_shapes, config): continue p2p_communication.send_forward(output_tensor, config) - +@dlp.log def send_backward(input_tensor_grads, tensor_shapes, config): if not isinstance(input_tensor_grads, list): input_tensor_grads = [input_tensor_grads] @@ -961,7 +964,7 @@ def send_backward(input_tensor_grads, tensor_shapes, config): continue p2p_communication.send_backward(input_tensor_grad, config) - +@dlp.log def send_forward_recv_backward(output_tensors, tensor_shapes, config): if not isinstance(output_tensors, list): output_tensors = [output_tensors] @@ -975,7 +978,7 @@ def send_forward_recv_backward(output_tensors, tensor_shapes, config): output_tensor_grads.append(output_tensor_grad) return output_tensor_grads - +@dlp.log def send_backward_recv_forward(input_tensor_grads, tensor_shapes, config): if not isinstance(input_tensor_grads, list): input_tensor_grads = [input_tensor_grads] @@ -989,7 +992,7 @@ def send_backward_recv_forward(input_tensor_grads, tensor_shapes, config): input_tensors.append(input_tensor) return input_tensors - +@dlp.log def forward_backward_pipelining_without_interleaving(*, forward_step_func, data_iterator: Union[Iterator, List[Iterator]], diff --git a/megatron/training.py b/megatron/training.py index c830468f4cc..254e5b0fdd1 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -279,9 +279,9 @@ def pretrain( iteration = 0 if args.do_train and args.train_iters > 0: iteration = train(forward_step_func, - model, optimizer, opt_param_scheduler, - train_data_iterator, valid_data_iterator, - process_non_loss_data_func) + model, optimizer, opt_param_scheduler, + train_data_iterator, valid_data_iterator, + process_non_loss_data_func) print_datetime('after training is done') # Clean the model @@ -1340,13 +1340,29 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, update_rotary_pos_emb(curriculum_seqlen) args.curriculum_seqlen = curriculum_seqlen args.curr_iteration = iteration - loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ - train_step(forward_step_func, - train_data_iterator, - model, - optimizer, - opt_param_scheduler, - config) + if os.getenv("TORCH_PROFILER_ENABLE")=='2': + from torch.profiler import profile, record_function, ProfilerActivity + with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA]) as prof: + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ + train_step(forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config) + prof.export_chrome_trace( + f"{args.trace_dir}" + "/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" + ) + else: + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ + train_step(forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config) + iteration += 1 args.iteration = iteration new_samples = mpu.get_data_parallel_world_size() * \ From fa04d11e656ae010bb372b96a4c5b816b279adf8 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 22 May 2024 22:29:00 +0000 Subject: [PATCH 144/590] local changes --- megatron/training.py | 3 +-- pretrain_gpt_alcf.py | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 254e5b0fdd1..12dcc25829c 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1351,8 +1351,7 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, opt_param_scheduler, config) prof.export_chrome_trace( - f"{args.trace_dir}" - "/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" + f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" ) else: loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 2334a585a0e..c82dfc8ebe3 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -584,8 +584,7 @@ def main(): ) args = get_args() prof.export_chrome_trace( - f"{args.trace_dir}" - "/torch-trace-{RANK}-of-{WORLD_SIZE}.json" + f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}.json" ) else: model = pretrain( From 6cca87f4db36b4e43149376e7a47a5bc678d0d38 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 22 May 2024 23:37:25 +0000 Subject: [PATCH 145/590] distributed loading --- megatron/__init__.py | 1 + megatron/data/gpt_dataset.py | 67 +++++++++++++++++++----------------- megatron/utils.py | 3 ++ 3 files changed, 40 insertions(+), 31 deletions(-) diff --git a/megatron/__init__.py b/megatron/__init__.py index d92a279ec67..ce44561e2d0 100644 --- a/megatron/__init__.py +++ b/megatron/__init__.py @@ -14,6 +14,7 @@ from .initialize import initialize_megatron from .utils import (print_rank_0, + print_flush, is_last_rank, print_rank_last, is_rank_0, diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 210a92c85e8..ea350bcd288 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -9,13 +9,14 @@ import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0, is_rank_0, get_args +from megatron import print_rank_0, is_rank_0, get_args, print_flush from megatron.core import mpu from megatron.data.blendable_dataset import BlendableDataset from megatron.data.dataset_utils import get_datasets_weights_and_num_samples from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset - +from mpi4py import MPI +comm = MPI.COMM_WORLD def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_valid_test_num_samples, @@ -52,7 +53,7 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_datasets = [] valid_datasets = [] test_datasets = [] - for i in range(len(prefixes)): + for i in range(torch.distributed.get_rank(), len(prefixes), torch.distributed.get_world_size()): train_ds, valid_ds, test_ds = _build_train_valid_test_datasets( prefixes[i], data_impl, splits_string, datasets_train_valid_test_num_samples[i], @@ -65,7 +66,11 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, valid_datasets.append(valid_ds) if test_ds: test_datasets.append(test_ds) - + def flatten_list(xss): + return [x for xs in xss for x in xs] + train_datasets = flatten_list(comm.allgather(train_datasets)) + valid_datasets = flatten_list(comm.allgather(valid_datasets)) + test_datasets = flatten_list(comm.allgather(test_datasets)) # Blend. blending_train_dataset = None if train_datasets: @@ -129,11 +134,11 @@ def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + print_flush(' > dataset split:') def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' + print_flush(' {}:'.format(name)) + print_flush(' document indices in [{}, {}) total of {} ' 'documents'.format(splits[index], splits[index + 1], splits[index + 1] - splits[index])) print_split_stats('train', 0) @@ -211,8 +216,8 @@ def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, total_num_of_documents = indexed_dataset.sizes.shape[0] - print_rank_0(' {}:'.format(dataset_name)) - print_rank_0(' document indices in [0, {}) total of {} ' + print_flush(' {}:'.format(dataset_name)) + print_flush(' document indices in [0, {}) total of {} ' 'documents'.format(total_num_of_documents, total_num_of_documents)) documents = np.arange(start=0, stop=total_num_of_documents, @@ -233,9 +238,9 @@ def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) - print_rank_0(' > finished creating indexed dataset in {:4f} ' + print_flush(' > finished creating indexed dataset in {:4f} ' 'seconds'.format(time.time() - start_time)) - print_rank_0(' number of documents: {}'.format( + print_flush(' number of documents: {}'.format( indexed_dataset.sizes.shape[0])) return indexed_dataset @@ -413,9 +418,9 @@ def _build_index_mappings(name, data_prefix, documents, sizes, data_cache_success = True # Build the indexed mapping if not exist. - if build_indices and is_rank_0(): - print_rank_0(' > WARNING: could not find index map files, building ' - 'the indices on rank 0 ...') + if build_indices: + print_flush(f" > WARNING: could not find index map files, building " + "the indices on rank {torch.distributed.get_rank()} ...") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -480,7 +485,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch) np.save(idx_path['sample'], sample_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save sample-idx mapping ' + print_flush(' > elasped time to build and save sample-idx mapping ' '(seconds): {:4f}'.format(time.time() - start_time)) # shuffle-idx. start_time = time.time() @@ -493,7 +498,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, shuffle_idx = _build_shuffle_idx(num_samples_, sample_idx.shape[0] - 1, np_rng) np.save(idx_path['shuffle'], shuffle_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save shuffle-idx mapping' + print_flush(' > elasped time to build and save shuffle-idx mapping' ' (seconds): {:4f}'.format(time.time() - start_time)) except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_dir})') @@ -503,32 +508,32 @@ def _build_index_mappings(name, data_prefix, documents, sizes, print('write access to.') data_cache_success = False - counts = get_accelerator().LongTensor([data_cache_success]) - torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) - if counts[0].item() != ( - torch.distributed.get_world_size() // - torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // - torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): - print_rank_0("Data index creation unsuccessful, exiting.") - exit() + #counts = get_accelerator().LongTensor([data_cache_success]) + #torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) + #torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + #if counts[0].item() != ( + # torch.distributed.get_world_size() // + # torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // + # torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): + # print_rank_0("Data index creation unsuccessful, exiting.") + # exit() # Load mappings. start_time = time.time() - print_rank_0(f" > loading doc-idx mapping from {idx_path['doc']}") + print_flush(f" > loading doc-idx mapping from {idx_path['doc']}") doc_idx = np.load(idx_path['doc'], allow_pickle=True, mmap_mode='r') - print_rank_0(f" > loading sample-idx mapping from {idx_path['sample']}") + print_flush(f" > loading sample-idx mapping from {idx_path['sample']}") sample_idx = np.load(idx_path['sample'], allow_pickle=True, mmap_mode='r') - print_rank_0(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") + print_flush(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") shuffle_idx = np.load(idx_path['shuffle'], allow_pickle=True, mmap_mode='r') - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( + print_flush(' loaded indexed file in {:3.3f} seconds'.format( time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( + print_flush(' total number of samples: {}'.format( sample_idx.shape[0])) - print_rank_0(' total number of epochs: {}'.format(num_epochs)) + print_flush(' total number of epochs: {}'.format(num_epochs)) return doc_idx, sample_idx, shuffle_idx, desc, desc_hash diff --git a/megatron/utils.py b/megatron/utils.py index f6a293281ac..06b0ee693db 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -234,6 +234,9 @@ def print_rank_0(message): else: print(message, flush=True) +def print_flush(message): + print(f"[{torch.distributed.get_rank()}] {message}", flush=True) + def is_last_rank(): return torch.distributed.get_rank() == ( torch.distributed.get_world_size() - 1) From 62f8f56d12d8629c518624fceb724660535cfecf Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 23 May 2024 03:27:42 +0000 Subject: [PATCH 146/590] fixed print issue --- ALCF/test_blendable_dataset.py | 13 +++++++------ megatron/data/gpt_dataset.py | 7 +++++-- 2 files changed, 12 insertions(+), 8 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 96811982510..24570b4d768 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -1,11 +1,12 @@ #!/usr/bin/env python +from mpi4py import MPI from megatron.data.gpt_dataset import build_train_valid_test_datasets import numpy as np from megatron.global_vars import set_args, set_global_variables, get_args from megatron.arguments import parse_args from megatron.initialize import initialize_megatron from megatron.data.data_samplers import build_pretraining_data_loader -from mpi4py import MPI + from megatron.core import mpu comm = MPI.COMM_WORLD initialize_megatron(allow_no_cuda=True) @@ -72,9 +73,9 @@ #### Build data loaders rank_in_parallel_group = mpu.get_sequence_parallel_rank() print(rank_in_parallel_group) -if rank_in_parallel_group == 0: - train_dataloader = build_pretraining_data_loader( - train_ds, args.consumed_train_samples) - valid_dataloader = build_pretraining_data_loader( +#if rank_in_parallel_group == 0: +train_dataloader = build_pretraining_data_loader( + train_ds, args.consumed_train_samples) +valid_dataloader = build_pretraining_data_loader( valid_ds, args.consumed_valid_samples) - test_dataloader = build_pretraining_data_loader(test_ds, 0) +test_dataloader = build_pretraining_data_loader(test_ds, 0) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index ea350bcd288..1db14bccdb2 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -53,6 +53,8 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_datasets = [] valid_datasets = [] test_datasets = [] + print_rank_0(" >>> Building datasets in distributed way ****") + t0 = time.time() for i in range(torch.distributed.get_rank(), len(prefixes), torch.distributed.get_world_size()): train_ds, valid_ds, test_ds = _build_train_valid_test_datasets( prefixes[i], data_impl, splits_string, @@ -71,6 +73,7 @@ def flatten_list(xss): train_datasets = flatten_list(comm.allgather(train_datasets)) valid_datasets = flatten_list(comm.allgather(valid_datasets)) test_datasets = flatten_list(comm.allgather(test_datasets)) + print_rank_0(f" >>> Done with building datasets in {time.time() - t0} seconds") # Blend. blending_train_dataset = None if train_datasets: @@ -232,7 +235,7 @@ def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): """Build indexed dataset.""" - print_rank_0(' > building dataset index ...') + print_flush(' > building dataset index ...') start_time = time.time() indexed_dataset = make_indexed_dataset(data_prefix, @@ -420,7 +423,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Build the indexed mapping if not exist. if build_indices: print_flush(f" > WARNING: could not find index map files, building " - "the indices on rank {torch.distributed.get_rank()} ...") + f"the indices on rank {torch.distributed.get_rank()} ...") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. From 2f0154394bbdf3c64b4669f9d944645e2cdb8f2b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 23 May 2024 17:49:32 -0500 Subject: [PATCH 147/590] Update README.md --- ALCF/README.md | 251 +++++++++++++++++++++++++++++++------------------ 1 file changed, 162 insertions(+), 89 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 0b436792d99..82577ab372e 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,114 +1,84 @@ # Megatron-DeepSpeed @ ALCF -## ✅ TODOs -
-TODOs: +## 🆘 Getting Started -- [ ] Ensure / double check that optimizer settings from `ds_config.json` aren't being overwritten by some defaults in `megatron/arguments.py` - - [ ] specifically, `momentum, beta{1, 2}, etc` - -
✅ Completed +> [!NOTE] +> [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching +> distributed training on {Polaris, Aurora, Sunspot} @ ALCF. -- Continue runs on Polaris @ - - [x] 48 Nodes - - [x] 32 Nodes - - [x] 16 Nodes - - [x] 8 Nodes - - [x] 4 Nodes -- [x] Then, try re-creating ( / fixing) conda with `cuda==12.1` - - 😔, failed. - -- ~~‼️ Unable to save checkpoints with `torch==2.1` + `cuda==11.8`~~: - - Fixed in [a57a21f](https://github.com/argonne-lcf/Megatron-DeepSpeed/commit/a57a21f6b2a8abf847f5ef599e1b1edcb5a5e1b5) + -
🐛 Bug - - - Training progresses OK: +## 🏃‍♂️ Running - ```bash - [2024-03-07 15:27:02,646] [INFO] [timer.py:260:stop] epoch=0/micro_step=199/global_step=199, RunningAvgSamplesPerSec=58.730622229657506, CurrSamplesPerSec=61.35304005128382, MemAllocated=6.01GB, MaxMemAllocated=19.52GB - iteration 199/ 317892 | consumed samples: 152832 | consumed tokens: 625999872 | elapsed time per iteration (ms): 14287.5 | learning rate: 2.407E-04 | global batch size: 768 | lm loss: 5.905366E+00 | loss scale: 8192.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 53.753 | tokens per gpu per second (tgs): 1146.733 | TFLOPs: 69.85 | - [2024-03-07 15:27:15,063] [INFO] [logging.py:96:log_dist] [Rank 0] step=200, skipped=4, lr=[0.000240653265864008, 0.000240653265864008], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-03-07 15:27:17,188] [INFO] [timer.py:260:stop] epoch=0/micro_step=200/global_step=200, RunningAvgSamplesPerSec=58.730745476291396, CurrSamplesPerSec=58.75503515561452, MemAllocated=6.01GB, MaxMemAllocated=19.52GB - iteration 200/ 317892 | consumed samples: 153600 | consumed tokens: 629145600 | elapsed time per iteration (ms): 14541.4 | learning rate: 2.407E-04 | global batch size: 768 | lm loss: 5.897035E+00 | loss scale: 8192.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 52.815 | tokens per gpu per second (tgs): 1126.713 | TFLOPs: 68.63 | - saving checkpoint at iteration 200 to checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb768_pp1_tp2_fp16 - # ... - ``` +To launch on Polaris @ ALCF: - - Then crashes with: - ```python - Traceback (most recent call last): - Traceback (most recent call last): - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/pretrain_gpt_alcf.py", line 575, in - model = main() - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/pretrain_gpt_alcf.py", line 554, in main - model = pretrain( - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/training.py", line 226, in pretrain - iteration = train(forward_step_func, - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/training.py", line 1290, in train - save_checkpoint_and_time(iteration, model, optimizer, - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/training.py", line 1151, in save_checkpoint_and_time - save_checkpoint(iteration, model, optimizer, opt_param_scheduler) - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/checkpointing.py", line 259, in save_checkpoint - state_dict[UNIVERSAL_CHECKPOINT_INFO] = _universal_checkpoint_info(model) - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/checkpointing.py", line 783, in _universal_checkpoint_info - info.update(model[0].universal_checkpoint_info()) - File "/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/megatron/model/gpt_model.py", line 203, in universal_checkpoint_info - info[TP_REPLICATED_PARAMETER_PATTERNS] = self._get_tp_replicated_param_patterns() - File "/lus/eagle/projects/datascience/foremans/miniconda3/envs/polaris/2024-03-06/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1695, in __getattr__ - raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'") - AttributeError: 'GPTModel' object has no attribute '_get_tp_replicated_param_patterns' - ``` - 🤔 +
⏳ Request an interactive job with qsub -I: + +```bash +qsub -A -
+
⬇️ Clone repo + navigate into it: -
+```bash +git clone "https://github.com/argonne-lcf/Megatron-DeepSpeed" +cd Megatron-DeepSpeed +```
+
🐍 Setup Python: + +```bash +module use /soft/modulefiles ; module load conda ; conda activate base +PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_venv_from_conda +``` + +- 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): + + ```bash + mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz + python3 -m pip install -e deps/ezpz --require-virtualenv + ``` +
-## 🏃‍♂️ Running -> [!NOTE] -> [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching -> distributed training on {Polaris, Aurora, Sunspot} @ ALCF. +
🚀 Launch: -To launch on Polaris @ ALCF: +In this case, train a ~ 2B Model (with 10 layers), +for 1000 iterations using the data file list in: -> [!IMPORTANT] -> **Launch Instructions** on Polaris @ ALCF -> -> 1. Request an interactive job with `qsub -I`: -> -> ```bash -> $ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I -> ``` -> -> 2. Clone repo + navigate into it: -> ```sh -> $ cd /path/to/Megatron-DeepSpeed/ # or git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -> ``` -> -> 3. Load required modules: -> -> ```sh -> $ module use /soft/modulefiles -> $ module load conda/2024-04-29 -> $ conda activate base -> ``` -> -> 4. Launch: -> -> ```sh -> $ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh -> ``` +[`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) + +with a micro-batch-size of 2, with the `torch.optim.AdamW` optimizer. + +**Note** that _any_ of the options in the + +[`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) + +function from + +[`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) + +can be overridden dynamically at runtime using this technique. + +```bash +PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh +```
[output]: @@ -477,6 +447,34 @@ training ...
+
+ + + + + + +
🚀 Launch: @@ -66,14 +100,8 @@ for 1000 iterations using the data file list in: with a micro-batch-size of 2, with the `torch.optim.AdamW` optimizer. -**Note** that _any_ of the options in the - -[`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) - -function from - -[`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) - +**Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) +function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) can be overridden dynamically at runtime using this technique. ```bash From 06ac06512893f36f27c35064627cab872025d898 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 23 May 2024 22:13:59 -0500 Subject: [PATCH 149/590] Added function for on-the-fly building the dataset --- ALCF/test_blendable_dataset.py | 8 +- megatron/data/blendable_dataset.py | 16 ++- megatron/data/gpt_dataset.py | 163 ++++++++++++++++++++--------- 3 files changed, 136 insertions(+), 51 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 24570b4d768..05086a5edb1 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -55,7 +55,6 @@ if comm.rank ==0: print(f"Total number of samples: {len(train_ds)}") print(f"Weights set: {weights[:min(8, num_datasets)]}") -#print(f"Weights across training: {ratio_select[:min(8, num_datasets)]}") for e in range(min(100, args.train_iters)): ratio_select=np.zeros(num_datasets) @@ -72,10 +71,13 @@ #### Build data loaders rank_in_parallel_group = mpu.get_sequence_parallel_rank() -print(rank_in_parallel_group) -#if rank_in_parallel_group == 0: train_dataloader = build_pretraining_data_loader( train_ds, args.consumed_train_samples) valid_dataloader = build_pretraining_data_loader( valid_ds, args.consumed_valid_samples) test_dataloader = build_pretraining_data_loader(test_ds, 0) + + +# Run through all the batches in data loader +for i in train_dataloader: + print(i.shape) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index f3276c68230..c1148c469e9 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -10,7 +10,7 @@ import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0 +from megatron import print_rank_0, print_flush from megatron.core import mpu class BlendableDataset(torch.utils.data.Dataset): @@ -123,3 +123,17 @@ def __getitem__(self, idx): "dataset_idx" : dataset_idx, **self.datasets[dataset_idx][sample_idx], } + +class DistributedBlendableDataset(BlendableDataset): + def __getitem__(self, idx): + dataset_idx = self.dataset_index[idx] + sample_idx = self.dataset_sample_index[idx] + if (self.datasets[dataset_idx].build): + dataset = self.datasets[dataset_idx].dataset + else: + dataset = self.datasets[dataset_idx].Build() + return { + "dataset_idx" : dataset_idx, + **dataset[sample_idx], + } + \ No newline at end of file diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 1db14bccdb2..5ff31983b7e 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -11,12 +11,10 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0, is_rank_0, get_args, print_flush from megatron.core import mpu -from megatron.data.blendable_dataset import BlendableDataset +from megatron.data.blendable_dataset import BlendableDataset, DistributedBlendableDataset from megatron.data.dataset_utils import get_datasets_weights_and_num_samples from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset -from mpi4py import MPI -comm = MPI.COMM_WORLD def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, train_valid_test_num_samples, @@ -49,43 +47,74 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, zip(*datasets_train_valid_test_num_samples) ) - # Build individual datasets. + class DatasetBuilder: + ''' + This is for building individual dataset from each dataset file + ''' + def __init__(self, prefix, data_impl, splits_string, + num_samples, seq_length, seed, skip_warmup, + return_doc_ids, + data_cache_path=data_cache_path, name='train'): + self.prefix = prefix + self.data_impl = data_impl + self.splits_string = splits_string + self.num_samples = num_samples + self.seq_length = seq_length + self.seed = seed + self.skip_warmup = skip_warmup + self.return_doc_ids = return_doc_ids + self.data_cache_path = data_cache_path + self.dataset = None + self.name = name + self.desc = prefix + f"{num_samples}" + f"{seq_length}" + f"{seed}" + self.build = False + def Build(self): + self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, + self.num_samples, self.seq_length, self.seed, self.skip_warmup, self.return_doc_ids, name, + data_cache_path=self.data_cache_path) + self.build = True + return self.dataset + + # Predetermine whether need to build the specific dataset or not. + + print_rank_0(" >>> Started building datasets in distributed way ... ") + + a, b, c = [int(d) for d in splits_string.split(",")] train_datasets = [] valid_datasets = [] test_datasets = [] - print_rank_0(" >>> Building datasets in distributed way ****") - t0 = time.time() - for i in range(torch.distributed.get_rank(), len(prefixes), torch.distributed.get_world_size()): - train_ds, valid_ds, test_ds = _build_train_valid_test_datasets( - prefixes[i], data_impl, splits_string, + # Build individual datasets. + if a > 0: + train_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, datasets_train_valid_test_num_samples[i], seq_length, seed, skip_warmup, - return_doc_ids, - data_cache_path=data_cache_path) - if train_ds: - train_datasets.append(train_ds) - if valid_ds: - valid_datasets.append(valid_ds) - if test_ds: - test_datasets.append(test_ds) - def flatten_list(xss): - return [x for xs in xss for x in xs] - train_datasets = flatten_list(comm.allgather(train_datasets)) - valid_datasets = flatten_list(comm.allgather(valid_datasets)) - test_datasets = flatten_list(comm.allgather(test_datasets)) - print_rank_0(f" >>> Done with building datasets in {time.time() - t0} seconds") + return_doc_ids,data_cache_path, 'train') for i in range(len(weights))] + for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): + train_datasets[i].Build() + if b > 0: + valid_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, seed, skip_warmup, + return_doc_ids,data_cache_path, 'valid') for i in range(len(weights))] + if c > 0: + test_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, seed, skip_warmup, + return_doc_ids,data_cache_path, 'test') for i in range(len(weights))] + print_rank_0(" >>> Finished building datasets in distributed way ... ") + # Blend. blending_train_dataset = None if train_datasets: - blending_train_dataset = BlendableDataset(train_datasets, weights, train_num_samples, + blending_train_dataset = DistributedBlendableDataset(train_datasets, weights, train_num_samples, data_cache_path=data_cache_path) blending_valid_dataset = None if valid_datasets: - blending_valid_dataset = BlendableDataset(valid_datasets, weights, valid_num_samples, + blending_valid_dataset = DistributedBlendableDataset(valid_datasets, weights, valid_num_samples, data_cache_path=data_cache_path) blending_test_dataset = None if test_datasets: - blending_test_dataset = BlendableDataset(test_datasets, weights, test_num_samples, + blending_test_dataset = DistributedcBlendableDataset(test_datasets, weights, test_num_samples, data_cache_path=data_cache_path) return (blending_train_dataset, blending_valid_dataset, @@ -137,11 +166,11 @@ def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_flush(' > dataset split:') + print_rank_0(' > dataset split:') def print_split_stats(name, index): - print_flush(' {}:'.format(name)) - print_flush(' document indices in [{}, {}) total of {} ' + print_rank_0(' {}:'.format(name)) + print_rank_0(' document indices in [{}, {}) total of {} ' 'documents'.format(splits[index], splits[index + 1], splits[index + 1] - splits[index])) print_split_stats('train', 0) @@ -167,6 +196,54 @@ def build_dataset(index, name): return (train_dataset, valid_dataset, test_dataset) +def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_string, + train_valid_test_num_samples, + seq_length, seed, skip_warmup, name + return_doc_ids=False, *, + data_cache_path=None): + """Build train, valid, and test datasets.""" + + # Each rank print out information + print_flush(f" >> building dataset for {dataset_prefix}") + # Indexed dataset. + indexed_dataset = get_indexed_dataset_(data_prefix, + data_impl, + skip_warmup) + + total_num_of_documents = indexed_dataset.sizes.shape[0] + splits = get_train_valid_test_split_(splits_string, total_num_of_documents) + + # Print stats about the splits. + print_rank_0(' > dataset split:') + + def print_split_stats(name, index): + print_rank_0(' {}:'.format(name)) + print_rank_0(' document indices in [{}, {}) total of {} ' + 'documents'.format(splits[index], splits[index + 1], + splits[index + 1] - splits[index])) + print_split_stats('train', 0) + print_split_stats('validation', 1) + print_split_stats('test', 2) + + def build_dataset(index, name): + dataset = None + if splits[index + 1] > splits[index]: + documents = np.arange(start=splits[index], stop=splits[index + 1], + step=1, dtype=np.int32) + dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, seed, + return_doc_ids, + data_cache_path=data_cache_path) + return dataset + if name.find("train")!=-1: + return build_dataset(0, 'train') + if name.find("valid")!=-1: + return build_dataset(1, 'valid') + if name.find("test")!=-1: + return build_dataset(2, 'test') + def build_dataset(dataset_name, data_prefix, data_impl, splits_string, num_samples, @@ -219,8 +296,8 @@ def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, total_num_of_documents = indexed_dataset.sizes.shape[0] - print_flush(' {}:'.format(dataset_name)) - print_flush(' document indices in [0, {}) total of {} ' + print_rank_0(' {}:'.format(dataset_name)) + print_rank_0(' document indices in [0, {}) total of {} ' 'documents'.format(total_num_of_documents, total_num_of_documents)) documents = np.arange(start=0, stop=total_num_of_documents, @@ -235,15 +312,15 @@ def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): """Build indexed dataset.""" - print_flush(' > building dataset index ...') + print_rank_0(' > building dataset index ...') start_time = time.time() indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) - print_flush(' > finished creating indexed dataset in {:4f} ' + print_rank_0(' > finished creating indexed dataset in {:4f} ' 'seconds'.format(time.time() - start_time)) - print_flush(' number of documents: {}'.format( + print_rank_0(' number of documents: {}'.format( indexed_dataset.sizes.shape[0])) return indexed_dataset @@ -422,8 +499,10 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Build the indexed mapping if not exist. if build_indices: - print_flush(f" > WARNING: could not find index map files, building " - f"the indices on rank {torch.distributed.get_rank()} ...") + # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the + # ranks will first create the document files, and then read it. + # There will not be contension effects going on either + print_flush(f" > WARNING: could not find index map files, building ") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -476,7 +555,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, doc_idx = _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch) np.save(idx_path['doc'], doc_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save doc-idx mapping ' + print_flush(' > elasped time to build and save doc-idx mapping ' '(seconds): {:4f}'.format(time.time() - start_time)) # sample-idx. start_time = time.time() @@ -511,16 +590,6 @@ def _build_index_mappings(name, data_prefix, documents, sizes, print('write access to.') data_cache_success = False - #counts = get_accelerator().LongTensor([data_cache_success]) - #torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - #torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) - #if counts[0].item() != ( - # torch.distributed.get_world_size() // - # torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // - # torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): - # print_rank_0("Data index creation unsuccessful, exiting.") - # exit() - # Load mappings. start_time = time.time() print_flush(f" > loading doc-idx mapping from {idx_path['doc']}") From 120a2b52b380fa6539ecafb506da5d1284349681 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 23 May 2024 22:18:59 -0500 Subject: [PATCH 150/590] fixed minor issue in _build_train_valid_test_datasets_single --- megatron/data/gpt_dataset.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 5ff31983b7e..9d457f09742 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -198,7 +198,7 @@ def build_dataset(index, name): def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_string, train_valid_test_num_samples, - seq_length, seed, skip_warmup, name + seq_length, seed, skip_warmup, name, return_doc_ids=False, *, data_cache_path=None): """Build train, valid, and test datasets.""" From 6fdbfd3a1f253044aab367d673a2633659c1e67a Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 23 May 2024 22:22:23 -0500 Subject: [PATCH 151/590] fixed variable order in Builder --- megatron/data/gpt_dataset.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 9d457f09742..bb381567a8c 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -70,7 +70,7 @@ def __init__(self, prefix, data_impl, splits_string, self.build = False def Build(self): self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, - self.num_samples, self.seq_length, self.seed, self.skip_warmup, self.return_doc_ids, name, + self.num_samples, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, data_cache_path=self.data_cache_path) self.build = True return self.dataset @@ -204,7 +204,7 @@ def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_strin """Build train, valid, and test datasets.""" # Each rank print out information - print_flush(f" >> building dataset for {dataset_prefix}") + print_flush(f" >> building dataset for {data_prefix}") # Indexed dataset. indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, From 3e2aa234e46aa4ef65f30167e0a1a3efc5904725 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 23 May 2024 22:40:00 -0500 Subject: [PATCH 152/590] fixed minor issue --- ALCF/test_blendable_dataset.py | 2 +- megatron/data/blendable_dataset.py | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 05086a5edb1..2c2ad84c22d 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -80,4 +80,4 @@ # Run through all the batches in data loader for i in train_dataloader: - print(i.shape) + print(i) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index c1148c469e9..3f4ea32c637 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -131,6 +131,7 @@ def __getitem__(self, idx): if (self.datasets[dataset_idx].build): dataset = self.datasets[dataset_idx].dataset else: + print_flush(f" First time reading samples from {self.datasets[dataset_idx].prefix}. Building the dataset now.") dataset = self.datasets[dataset_idx].Build() return { "dataset_idx" : dataset_idx, From b37174223577f1b6dcc3085f58df77424e3bc222 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 06:23:09 -0500 Subject: [PATCH 153/590] Add `setup_tokenizer_and_data()` function to `ALCF/helpers.sh` --- ALCF/helpers.sh | 145 +++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 138 insertions(+), 7 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 9e7e2b582bb..6a647952edb 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -16,6 +16,22 @@ export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} +save_dotenv() { + if [[ "$#" -ne 1 ]]; then + estr="[error]" + # echo "Expected exactly one argument, specifying outputdir. Received $#" + printf "%s Expected one argument (outdir). Received: %s" "$(printRed ${estr})" "$#" + else + outdir="$1" + module list + dotenv_file="${outdir}/.env" + echo "Saving environment to ${dotenv_file}" + printenv | grep -v "LS_COLORS" > "${dotenv_file}" + export DOTENV_FILE="${dotenv_file}" + fi +} + + function get_machine() { if [[ $(hostname) == x4* ]]; then machine="aurora" @@ -100,14 +116,14 @@ function loadCondaEnv() { function setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then - export LAUNCH_CMD="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" else # Assert `./hostfile_deepspeed` exists export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit - export LAUNCH_CMD="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" + export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" fi - printf "%s" "$(printRed 'Launching with:')" - printf " %s" "$(printMagenta ${LAUNCH_CMD})" + printf "Launching with: %s\n" "$(printRed "${LAUNCH_CMD}")" + printf " %s" "$(printMagenta ${LAUNCHER})" } function setDSlauncher() { @@ -219,8 +235,8 @@ function setParams() { export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" @@ -498,6 +514,23 @@ setup_conda_polaris() { fi } +########################################################## +# Check that we can find the `.py` file we wish to launch +########################################################## +check_executable() { + fp=$1 + if [[ -f "${fp}" ]]; then + export EXEC="${EXEC}" + # ----[1.5 Keep track of stem from file path]------------------------- + exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") + export EXEC_STEM="${exec_stem}" + else + estr="Unable to locate executable ${fp}" + printf "[ALCF.helpers:check_executable] %s" "$(printRed ${estr})" + fi +} + + function setEnv() { local virtual_env="${VIRTUAL_ENV-}" @@ -578,17 +611,109 @@ function makeHostfiles() { fi } +################################################## +# Setup tokenizer as either Llama2 or GPT2 style +################################################## +setup_tokenizer_and_data() { + # if [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then + # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + # export TOKENIZER_TYPE="Llama2" + # setData + # elif [[ ${tok} == gpt* || ${tok} == GPT* ]]; then + # export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + # export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" + # export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" + # export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" + # export TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file $MERGE_FILE" + # # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" + # export TOKENIZER_TYPE="GPT2" + # else + # echo "Unknown tokenizer ${tok} passed" + # fi + if [[ "$#" == 1 ]]; then + tok="$1" + dfl="${DATA_FILE_LIST:-}" + # echo "Setting up tokenizer with ${tok}" + # elif [[ "$#" -ne 2 ]]; then + elif [[ "$#" == 2 ]]; then + tok="$1" + dfl="$2" + # tok="${TOKENIZER_TYPE:-Llama2}" + else + echo "Incorrect number of arguments passed. Received: $#, expected 2" + fi + echo "Setting up tokenizer with ${tok}" + echo "Using data_file_list: ${dfl}" + if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then + export TOKENIZER_TYPE="GPT2" + export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" + export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" + export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" + export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" + # TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file ${MERGE_FILE}" + export DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" + # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" + # else [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then + else + export DATA_FLAGS="" + export TOKENIZER_TYPE="Llama2" + tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + export TOKENIZER_FLAGS="${TOKENIZER_FLAGS} --tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" + if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then + echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" + setData "${dfl}" || exit + # setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset + fi + # --tokenizer-model ${TOKENIZER_MODEL} \ + fi + # export DATA_FLAGS="${DATA_FLAGS}" + # export TOKENIZER_TYPE="${TOKENIZER_TYPE}" + # export TOKENIZER_FLAGS="${TOKENIZER_FLAGS}" + printf "[setData] DATA_FLAGS: %s\n" "$(printGreen ${DATA_FLAGS})" + printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" + # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then + # echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" + # setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset + # fi +} + + ############################################### # `setData`: # Ensure `DATA_FILE_LIST` is set, # fallback to default values if necessary. ############################################### function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] + # if [[ "$#" -ne 1 ]]; then + # tok="${TOKENIZER_TYPE:-Llama2}" + # else + # tok="$1" + # fi + # echo "Setting up tokenizer with ${tok}" + # setup_tokenizer "${tok}" + # tok="${TOKENIZER_TYPE:-}" + # if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then + # export TOKENIZER_TYPE="GPT2" + # export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + # export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" + # export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" + # export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" + # # TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file ${MERGE_FILE}" + # DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" + # # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" + # # else [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then + # else + # export TOKENIZER_TYPE="Llama2" + # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + # TOKENIZER_FLAGS="${TOKENIZER_FLAGS} --tokenizer-type Llama2Tokenizer" # dfldir="${WORKING_DIR}/ALCF/data-lists" # =====[Set DATA_FILE_LIST_FALLBACK based on current machine]============== if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" - elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] # shellcheck: source ./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/books.txt" @@ -623,6 +748,7 @@ function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export WEIGHT_SUM="${ws}" export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" + export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST} --data-cache-path ${DATA_CACHE_PATH}" echo "--------------------" echo "Updated environment:" printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" @@ -631,6 +757,11 @@ function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] printf "DFL_STEM: %s\n" "${DFL_STEM}" printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" echo "--------------------" + # fi + # export DATA_FLAGS="${DATA_FLAGS}" + # export TOKENIZER_FLAGS="${TOKENIZER_FLAGS}" + # printf "[setData] DATA_FLAGS: %s\n" "$(printGreen ${DATA_FLAGS})" + # printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" } function generateDSconfig() { From d93fb7f41eeda59a3c90f90984979da4db4212bf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 06:23:22 -0500 Subject: [PATCH 154/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 83 +++++++++++++++++++++++++++++---------------- 1 file changed, 54 insertions(+), 29 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index fee2e392dc3..9a6d39923c0 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -5,14 +5,26 @@ #PBS -l select=48 #PBS -l filesystems=eagle:home + #### Make it easy to track experiments by date ################### -YEAR="$(date "+%Y")" -MONTH="$(date "+%m")" -DAY="$(date "+%Y-%m-%d")" -TODAY="$(date "+%Y-%m-%d")" # kept for backwards compatibility -STARTED_AT="$(date "+%Y-%m-%d-%H%M%S")" +year="$(date "+%Y")" +month="$(date "+%m")" +day="$(date "+%Y-%m-%d")" +today="$(date "+%Y-%m-%d")" # kept for backwards compatibility +started_at="$(date "+%Y-%m-%d-%H%M%S")" +export YEAR="${year}" +export MONTH="${month}" +export DAY="${day}" +export TODAY="${today}" +export STARTED_AT="${started_at}" ################################################################## + +############################################################################# +# Check if running in `DEBUG=1` mode. +# - If so, this will print each command before it is ran and exit if any of +# them return a nonzero exit status. +############################################################################# if [[ -n "${DEBUG-}" ]]; then # to use: `DEBUG=1 bash train_llama_alcf.sh` printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" set -euxo pipefail @@ -23,6 +35,9 @@ if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` set -o noexec # same as set -n fi +################################################## +# Helper function for `source`-ing another file +################################################## sourceFile() { fp="$1" echo "source-ing ${fp}" @@ -34,6 +49,8 @@ sourceFile() { fi } +############################################################################## +###################### MAIN LOGIC ############################################ # ----[0. Navigate into `$PBS_O_WORKDIR`]-------------------------------------- cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') @@ -43,6 +60,11 @@ export HERE export EXEC="${HERE}/pretrain_gpt_alcf.py" [ -f "${EXEC}" ] || exit +# ----[1.5 Keep track of ] +exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") +export EXEC_STEM="${exec_stem}" + + # ----[2. `source ./ALCF/helpers_alcf.sh`:]------------------------------------ sourceFile "${HERE}/ALCF/helpers.sh" || exit @@ -55,10 +77,15 @@ setParams || exit # 05. Set command line arguments to pass t buildDSconfig || exit # 06. Create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 07. Specify output directory for {logs, checkpoints, etc.} setArgs || exit # 08. Specify additional `deepspeed` arguments -setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset -printJobInfo || exit # 11. Print job info -setupLauncher || exit # 12. set launcher to one of `MPICH` (default), or `deepspeed` +dfl="${DATA_FILE_LIST:-}" # 09. Setup data + tokenizer +tok="${TOKENIZER_TYPE:-Llama2}" # via `DATA_FILE_LIST` and `TOKENIZER_TYPE` +setup_tokenizer_and_data "${tok}" "${dfl}" || exit +printJobInfo || exit # 10. Print job info +setupLauncher || exit # 11. set launcher to one of `MPICH` (default), or `deepspeed` +save_dotenv "${CKPT_DIR}" || exit # 12. Print info about loaded modules and runtime environment +check_and_kill_if_running || exit # 13. Check that were not already running, if so, exit. # ----------------------------------------------------------------------------- +############################################################################## ################################################ # Assert `$TBDIR` exists inside our `$CKPT_DIR` @@ -69,31 +96,33 @@ mkdir -p "${TBDIR}" data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" -# Print info about loaded modules and runtime environment -module list -dotenv_file="${CKPT_DIR}/.env" -echo "Saving environment to ${dotenv_file}" -printenv | grep -v "LS_COLORS" > "${dotenv_file}" - -# Take custom args -custom_args=" $@" - # --log-num-zeros-in-grad \ - # --log-memory-to-tensorboard \ -run_cmd=" - ${LAUNCH_CMD} \ - --${DTYPE} \ +export DEFAULTS="\ --split 100,0,0 \ --log-interval 1 \ --no-bias-gelu-fusion \ --no-bias-dropout-fusion \ --no-masked-softmax-fusion \ - --tokenizer-type Llama2Tokenizer \ --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ --use-checkpoint-opt_param-scheduler \ --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ + --log-optimizer-states-to-tensorboard" + +# Take custom args +custom_args=" $@" + + # --log-num-zeros-in-grad \ + # --log-memory-to-tensorboard \ + # --data-file-list ${DATA_FILE_LIST} \ + # --data-file-list ${DATA_FILE_LIST} \ + # --data-cache-path ${data_cache_path} \ + # --tokenizer-type Llama2Tokenizer \ + # --tokenizer-model ${TOKENIZER_MODEL} \ +run_cmd=" + ${LAUNCHER} \ + --${DTYPE} \ + ${DEFAULTS} \ --optimizer ${OPT} \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ @@ -109,27 +138,23 @@ run_cmd=" --eval-interval ${EVAL_INTERVAL} \ --max-position-embeddings ${SEQ} \ --micro-batch-size ${MICRO_BATCH} \ - --data-file-list ${DATA_FILE_LIST} \ --tensor-model-parallel-size ${TP} \ --global-batch-size ${GLOBAL_BATCH} \ --pipeline-model-parallel-size ${PP} \ --num-key-value-heads ${NUM_KV_HEAD} \ - --data-cache-path ${data_cache_path} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - --tokenizer-model ${TOKENIZER_MODEL} \ ${LR_ARGS} \ ${LLAMA_ARGS} \ ${TIMING_STR} \ + ${DATA_FLAGS} \ + ${TOKENIZER_FLAGS} \ $ds_args \ ${gpt_args[*]} \ $custom_args \ |& tee ${OUTPUT_LOG} " -check_and_kill_if_running || exit echo "${run_cmd}" printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow ${OUTPUT_LOG})" -# printf "[!! \e[1;31m%s\e[0m] View output at:\n" "NOTE" -# printf "\e[1;34m%s\e[0m\n" "${OUTPUT_LOG}" eval "${run_cmd}" set +x From 05d82c3911f2643d55ef3da5f497c9bfba70313b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 06:23:56 -0500 Subject: [PATCH 155/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 0907f8163c6..e5035ae27b2 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -7,13 +7,12 @@ cd "${PBS_O_WORKDIR}" || exit HOSTNAME=$(hostname) if [[ "${HOSTNAME}" == x3* ]]; then MACHINE="polaris" - # XXX: ¯\_(ツ)_/¯ + # XXX: # - On Polaris, we see that: # - on 1 or 2 nodes, only MICRO_BATCH=1 will fit in memory # - on 8 nodes, MICRO_BATCH=2 will fit in memory # - on 48 nodes, MICRO_BATCH=4 will fit in memory - # - export nhosts=$(wc -l < "${PBS_NODEFILE}") + nhosts=$(wc -l < "${PBS_NODEFILE}") if [[ "${nhosts}" == 1 ]]; then export MBS=1 elif [[ "${nhosts}" == 2 ]]; then @@ -29,6 +28,7 @@ elif [[ "${HOSTNAME}" == x4* ]]; then MACHINE="aurora" fi +export nhosts OUTDIR="${PBS_O_WORKDIR}/pbslogs" mkdir -p "${OUTDIR}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" From 6de8496e435fdf9a2c32ac5b627afecc62f1aa32 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 06:24:21 -0500 Subject: [PATCH 156/590] Update `ALCF/README.md` --- ALCF/README.md | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 29169932d24..907cbd36cca 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -19,10 +19,10 @@ ## 🏃‍♂️ Running -To launch on Polaris @ [ALCF](https://alcf.anl.gov): +To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov):
⏳ Request an interactive job with qsub -I: - + ```bash qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I ``` @@ -42,14 +42,22 @@ cd Megatron-DeepSpeed 1. 📂 Load `conda` module and activate base environment: - ```bash - module use /soft/modulefiles ; module load conda ; conda activate base - ``` + - **Polaris**: + + ```bash + module use /soft/modulefiles ; module load conda ; conda activate base + ``` + + - **Sunspot**: + + ```bash + source ALCF/sunspot-env-2024-04-15-002.sh + ``` 3. 👻 Create virtual environment _on top of the base `conda`_[^venv]: ```bash - PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_venv_from_conda + export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_venv_from_conda ``` From 03aa7c11c0d5afd5c516b52b341b078c04c594fe Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 08:22:23 -0500 Subject: [PATCH 157/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 265 ++++++++++++++++++++++++------------------------ 1 file changed, 130 insertions(+), 135 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 6a647952edb..a9af22416dd 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -23,6 +23,7 @@ save_dotenv() { printf "%s Expected one argument (outdir). Received: %s" "$(printRed ${estr})" "$#" else outdir="$1" + mkdir -p "${outdir}" module list dotenv_file="${outdir}/.env" echo "Saving environment to ${dotenv_file}" @@ -32,7 +33,26 @@ save_dotenv() { } -function get_machine() { +where_am_i() { + if [[ $(hostname) == x4* ]]; then + machine="aurora" + elif [[ $(hostname) == x1* ]]; then + machine="sunspot" + elif [[ $(hostname) == x3* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + machine="sirius" + else + machine="polaris" + fi + elif [[ $(hostname) == nid* ]]; then + machine="perlmutter" + else + machine=$(hostname) + fi + echo "${machine}" +} + +get_machine() { if [[ $(hostname) == x4* ]]; then machine="aurora" elif [[ $(hostname) == x1* ]]; then @@ -53,7 +73,7 @@ function get_machine() { } -function check_and_kill_if_running() { +check_and_kill_if_running() { # kill $(ps aux | grep -E "$USER.+(mpi|main.py)" | grep -v grep | awk '{print $2}') RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') if [[ -n "${RUNNING_PIDS}" ]]; @@ -64,7 +84,7 @@ function check_and_kill_if_running() { } -function setupSrun() { +setupSrun() { if [[ $(hostname) == login* || $(hostname) == nid* ]]; then export NHOSTS="${SLURM_NNODES:-1}" export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" @@ -76,7 +96,7 @@ function setupSrun() { } -function printJobInfo() { +printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" echo "- Using $(which python3)" @@ -87,7 +107,7 @@ function printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } -function setupVenv() { +setupVenv() { VENV_DIR="$1" if [[ -d "${VENV_DIR}" ]]; then echo "Found venv at: ${VENV_DIR}" @@ -97,7 +117,7 @@ function setupVenv() { fi } -function loadCondaEnv() { +loadCondaEnv() { if [[ "${CONDA_EXE}" ]]; then echo "Already inside ${CONDA_EXE}, exiting!" else @@ -113,7 +133,7 @@ function loadCondaEnv() { } -function setupLauncher() { +setupLauncher() { # outdir=$1 if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" @@ -126,7 +146,7 @@ function setupLauncher() { printf " %s" "$(printMagenta ${LAUNCHER})" } -function setDSlauncher() { +setDSlauncher() { # launcher setting outdir=$1 export hfds="$outdir/hostfile_deepspeed" @@ -141,7 +161,7 @@ function setDSlauncher() { fi } -function set_lr_args() { +set_lr_args() { LR_ARGS="--lr ${LR} --lr-decay-style cosine" if [[ -n "${LR_DECAY_ITERS:-}" ]]; then LR_ARGS="${LR_ARGS} --lr-decay-iters ${LR_DECAY_ITERS}" @@ -153,7 +173,22 @@ function set_lr_args() { export LR_ARGS="${LR_ARGS}" } -function setParams() { + +get_batch_size_on_polaris() { + if [[ $(hostname) == x3* ]]; then + local nhosts=$(wc -l < "${PBS_NODEFILE}") + if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then + mbs=1 + elif [[ "${nhosts}" -ge 3 ]]; then + mbs=2 + elif [[ "${nhosts}" -ge 8 ]]; then + mbs=4 + fi + fi + echo "${mbs}" +} + +setParams() { LLAMA_ARGS="" # +----[Parallelism Settings] -------------------------------------------+ # +------[Aurora]--------||-------[SunSpot]-------------+ @@ -182,7 +217,8 @@ function setParams() { export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS # NOTE: MICRO_BATCH is exported below - MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 + # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 + export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else @@ -257,7 +293,7 @@ function setParams() { } -function setArgs() { +setArgs() { # ---- Set DeepSpeed arguments -------------------------------- ds_args=" " ds_args=" --deepspeed ${ds_args}" @@ -288,7 +324,7 @@ function setArgs() { } -function make_ds_hostfile() { +make_ds_hostfile() { export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" @@ -304,33 +340,45 @@ function make_ds_hostfile() { # | 1. Git clone ezpz (if not found) | # | 2. Install ezpz (if not installed) | # +---------------------------------------+ -function ezpz() { +ezpz() { if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" else echo "Found ezpz!" fi - echo "Done with clone. Now, checking if ezpz is installed..." - # if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then - if python3 -c "import sys; any(['ezpz' in s for s in sys.path])" 2> '/dev/null'; then - echo "Has ezpz installed. Nothing to do." + ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') + if [[ -n "${ezloc}" ]]; then + echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" + source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 + source "${ezloc}/src/ezpz/bin/getjobenv" || exit + make_ds_hostfile || exit else - echo "Does not have ezpz installed. Installing..." - echo "Using $(which python3) to install ezpz:" - python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" # > ezpz-install.log 2>&1 + echo "No ezpz detected. Attempting to install with $(which python3)" + python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --require-virtualenv fi echo "Done with ezpz." - source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit - source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv || exit - make_ds_hostfile || exit + # echo "Done with clone. Now, checking if ezpz is installed..." + # if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then + # if [[ $(python3 -c "import sys; any(['ezpz' in s for s in sys.path])") 2> '/dev/null' ]]; then + # echo "Has ezpz installed. Nothing to do." + # else + # echo "Does not have ezpz installed. Installing..." + # echo "Using $(which python3) to install ezpz:" + # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --verbose --require-virtualenv # > ezpz-install.log 2>&1 + # fi + # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --verbose --require-virtualenv + # # echo "Done with ezpz." + # source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit + # source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv || exit + # make_ds_hostfile || exit } # +------------------------------------------------------------------------+ # | Save important environment variables to .deepspeed_env, which will be | # | forwarded to ALL ranks with DeepSpeed | # +------------------------------------------------------------------------+ -function saveDSenv() { +saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { echo "PATH=${PATH}" ; @@ -342,7 +390,7 @@ function saveDSenv() { } > .deepspeed_env } -function setOutput() { +setOutput() { # ---- Specify output location -------------------------------- OUTPUT_PREFIX="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" OUTPUT_PREFIX="${OUTPUT_PREFIX}_hs${HIDDEN}_mb${MICRO_BATCH}" @@ -367,7 +415,7 @@ function setOutput() { printf "Checkpoints will be saved to: %s\n" $(printYellow "${CKPT_DIR}") } -function buildDSconfig() { +buildDSconfig() { # ---- Build DeepSpeed Config --------------------------------- export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" @@ -381,13 +429,13 @@ function buildDSconfig() { } -function sumWeights() { +sumWeights() { local file_list=$1 weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") python3 -c "import numpy as np; print(np.sum(${weights}))" } -function sumFiles() { +sumFiles() { local rd=$1 for f in $("${rd}/*.txt"); do ws=$(sumWeights "${rd}/${f}") @@ -397,46 +445,23 @@ function sumFiles() { ######################################################## # Setup / activate conda environment, +# NOTE: +# +# Jerome's `frameworks_2024_5_v2` seems broken ?? +# - seems to be missing `python3 -c 'from mpi4py import MPI'` ??? +# - consequently, we leave the setup below commented out (for the time +# being): +# if [[ -z "${CONDA_PREFIX-}" ]]; then +# module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 +# else +# echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" +# fi ######################################################## setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ if [[ -z "${CONDA_PREFIX:-}" ]]; then - # eval "$(~/miniconda3/bin/conda shell.zsh hook)" - # conda activate anl_24_q2_release module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua fi - # XXX: ------------------------------------------------------------------- - # Jerome's `frameworks_2024_5_v2` seems broken ?? - # - seems to be missing `python3 -c 'from mpi4py import MPI'` ??? - # - consequently, we leave the setup below commented out (for the time - # being): - # if [[ -z "${CONDA_PREFIX-}" ]]; then - # module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 - # else - # echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" - # fi - # ------------------------------------------------------------------------ - - ###### check if VIRTUAL_ENV non-empty #################################### - # venvs/anl_24_q2_release/bin/activate - # if [[ -d "${DEFAULT_VENV_PATH}" ]]; then - if [[ -z "${VIRTUAL_ENV:-}" ]]; then - if [[ -n "${CONDA_PREFIX}" ]]; then - VENV_DIR="${WORKING_DIR}/venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')" - else - VENV_DIR="${WORKING_DIR}/venvs/anl_24_q2_release" - fi - echo "Caught virtual env at ${VENV_DIR}!" - # source "${VENV_DIR}/bin/activate" || - if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then - printf "[!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" - # echo "[!ERROR]: Unable to locate ${VENV_DIR}/bin/activate !!" - else - source "${VENV_DIR}/bin/activate" - fi - else - echo "Found existing python at: $(which python3)" - fi } ######################## @@ -453,6 +478,21 @@ setup_conda_sirius() { fi } +######################## +# Setup conda on Polaris +######################## +setup_conda_polaris() { + # unset MPICH_GPU_SUPPORT_ENABLED + ###### check if CONDA_PREFIX non-empty ################ + if [[ -z "${CONDA_PREFIX-}" ]]; then + # if so, load the default conda/2024-04-29 + # module and activate base environment + module use /soft/modulefiles ; module load conda ; conda activate base + else + echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" + fi +} + setup_venv_from_conda() { if [[ -z "${CONDA_PREFIX}" ]]; then echo "No ${CONDA_PREFIX} found." # Exiting." @@ -488,32 +528,6 @@ setup_venv_from_conda() { } -######################## -# Setup conda on Polaris -######################## -setup_conda_polaris() { - # unset MPICH_GPU_SUPPORT_ENABLED - ###### check if CONDA_PREFIX non-empty ################ - if [[ -z "${CONDA_PREFIX-}" ]]; then - # if so, load the default conda/2024-04-29 - # module and activate base environment - module use /soft/modulefiles - module load conda/2024-04-29 ; conda activate base - else - echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" - fi - ###### check if VIRTUAL_ENV non-empty ################# - if [[ -z "${VIRTUAL_ENV:-}" ]]; then - DEFAULT_VENV_PATH=${WORKING_DIR}/venvs/polaris/2024-04-29 - if [[ -d "${DEFAULT_VENV_PATH}" ]]; then - echo "Caught virtual env at ${DEFAULT_VENV_PATH}!" - source "${WORKING_DIR}/venvs/polaris/2024-04-29/bin/activate" - fi - else - echo "Found existing python at: $(which python3)" - fi -} - ########################################################## # Check that we can find the `.py` file we wish to launch ########################################################## @@ -532,12 +546,14 @@ check_executable() { -function setEnv() { - local virtual_env="${VIRTUAL_ENV-}" - local conda_prefix="${CONDA_PREFIX-}" +setEnv() { + local virtual_env="${VIRTUAL_ENV:-}" + local conda_prefix="${CONDA_PREFIX:-}" if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "No virtual environment found." echo "Using conda from: ${conda_prefix}" + echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" + setup_venv_from_conda elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then echo "No conda found." echo "Using virtual_env from: ${virtual_env}" @@ -589,6 +605,9 @@ function setEnv() { echo "Unable to setup python environment. Exiting" exit 1 fi + if [[ -z "${virtual_env}" ]]; then + setup_venv_from_conda + fi ##################################################################### pystr="Using: $(which python3)" printf "[python] %s" "$(printMagenta ${pystr})" @@ -603,7 +622,7 @@ function setEnv() { # - If so, use this. # - Otherwise, make default HOSTFILEs from "${PBS_NODEFILE}" ###################################################################### -function makeHostfiles() { +makeHostfiles() { if [[ -n "${HOSTFILE}" ]]; then printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" else @@ -615,31 +634,12 @@ function makeHostfiles() { # Setup tokenizer as either Llama2 or GPT2 style ################################################## setup_tokenizer_and_data() { - # if [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then - # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - # export TOKENIZER_TYPE="Llama2" - # setData - # elif [[ ${tok} == gpt* || ${tok} == GPT* ]]; then - # export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" - # export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" - # export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" - # export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - # export TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file $MERGE_FILE" - # # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" - # export TOKENIZER_TYPE="GPT2" - # else - # echo "Unknown tokenizer ${tok} passed" - # fi if [[ "$#" == 1 ]]; then tok="$1" dfl="${DATA_FILE_LIST:-}" - # echo "Setting up tokenizer with ${tok}" - # elif [[ "$#" -ne 2 ]]; then elif [[ "$#" == 2 ]]; then tok="$1" dfl="$2" - # tok="${TOKENIZER_TYPE:-Llama2}" else echo "Incorrect number of arguments passed. Received: $#, expected 2" fi @@ -648,14 +648,18 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" - export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + local machine=$(where_am_i) + if [[ ${machine} == "polaris" ]]; then + export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" + elif [[ ${machine} == "sunspot" ]]; then + export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + else + export DATA_PARENT="${DATA_PARENT:-${WORKING_DIR}/dataset}" + fi export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - # TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file ${MERGE_FILE}" export DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" - # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" - # else [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then else export DATA_FLAGS="" export TOKENIZER_TYPE="Llama2" @@ -665,19 +669,10 @@ setup_tokenizer_and_data() { if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" setData "${dfl}" || exit - # setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset fi - # --tokenizer-model ${TOKENIZER_MODEL} \ fi - # export DATA_FLAGS="${DATA_FLAGS}" - # export TOKENIZER_TYPE="${TOKENIZER_TYPE}" - # export TOKENIZER_FLAGS="${TOKENIZER_FLAGS}" printf "[setData] DATA_FLAGS: %s\n" "$(printGreen ${DATA_FLAGS})" printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" - # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then - # echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" - # setData "${DATA_FILE_LIST-}" || exit # 09. Specify `DATA_FILE_LIST` for dolma dataset - # fi } @@ -686,7 +681,7 @@ setup_tokenizer_and_data() { # Ensure `DATA_FILE_LIST` is set, # fallback to default values if necessary. ############################################### -function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] +setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # if [[ "$#" -ne 1 ]]; then # tok="${TOKENIZER_TYPE:-Llama2}" # else @@ -764,7 +759,7 @@ function setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" } -function generateDSconfig() { +generateDSconfig() { for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ "$PP" "$DTYPE" do @@ -926,35 +921,35 @@ $flops_profiler EOT } -function printBlack() { +printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } -function printRed() { +printRed() { printf "\e[1;31m%s\e[0m\n" "$@" } -function printGreen() { +printGreen() { printf "\e[1;32m%s\e[0m\n" "$@" } -function printYellow() { +printYellow() { printf "\e[1;33m%s\e[0m\n" "$@" } -function printBlue() { +printBlue() { printf "\e[1;34m%s\e[0m\n" "$@" } -function printMagenta() { +printMagenta() { printf "\e[1;35m%s\e[0m\n" "$@" } -function printCyan() { +printCyan() { printf "\e[1;36m%s\e[0m\n" "$@" } -function printWhite() { +printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } From 3cd3f1a9a695341643dd29c87a47969de90a4812 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 08:22:48 -0500 Subject: [PATCH 158/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 21 ++------------------- 1 file changed, 2 insertions(+), 19 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index e5035ae27b2..72322d655c9 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -1,5 +1,5 @@ #!/bin/bash --login -# + NOW="$(date "+%Y-%m-%d-%H%M%S")" cd "${PBS_O_WORKDIR}" || exit @@ -7,30 +7,13 @@ cd "${PBS_O_WORKDIR}" || exit HOSTNAME=$(hostname) if [[ "${HOSTNAME}" == x3* ]]; then MACHINE="polaris" - # XXX: - # - On Polaris, we see that: - # - on 1 or 2 nodes, only MICRO_BATCH=1 will fit in memory - # - on 8 nodes, MICRO_BATCH=2 will fit in memory - # - on 48 nodes, MICRO_BATCH=4 will fit in memory - nhosts=$(wc -l < "${PBS_NODEFILE}") - if [[ "${nhosts}" == 1 ]]; then - export MBS=1 - elif [[ "${nhosts}" == 2 ]]; then - export MBS=1 - elif [[ "${nhosts}" -ge 3 ]]; then - export MBS=2 - elif [[ "${nhosts}" -ge 8 ]]; then - export MBS=4 - fi elif [[ "${HOSTNAME}" == x1* ]]; then MACHINE="sunspot" elif [[ "${HOSTNAME}" == x4* ]]; then MACHINE="aurora" fi -export nhosts -OUTDIR="${PBS_O_WORKDIR}/pbslogs" -mkdir -p "${OUTDIR}" +OUTDIR="${PBS_O_WORKDIR}/pbslogs" && mkdir -p "${OUTDIR}" OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" echo "+---------------------------------------------------------+" From bc1dbfdd7943b4513c9a167d934ae0e5e1d5e171 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 24 May 2024 13:59:39 -0500 Subject: [PATCH 159/590] Fix `--data-cache-path` in `ALCF/helpers.sh, train_llama_alcf.sh` --- ALCF/helpers.sh | 37 +++++++++++++++++++++++-------------- train_llama_alcf.sh | 5 +++-- 2 files changed, 26 insertions(+), 16 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a9af22416dd..c0927766f27 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -135,12 +135,13 @@ loadCondaEnv() { setupLauncher() { # outdir=$1 - if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then - export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" - else + if [[ "${LAUNCH_CMD:-"MPICH"}" == "deepspeed" ]]; then # Assert `./hostfile_deepspeed` exists export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" + # if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then + else + export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" fi printf "Launching with: %s\n" "$(printRed "${LAUNCH_CMD}")" printf " %s" "$(printMagenta ${LAUNCHER})" @@ -390,21 +391,27 @@ saveDSenv() { } > .deepspeed_env } -setOutput() { + +get_output_prefix() { # ---- Specify output location -------------------------------- - OUTPUT_PREFIX="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" - OUTPUT_PREFIX="${OUTPUT_PREFIX}_hs${HIDDEN}_mb${MICRO_BATCH}" - OUTPUT_PREFIX="${OUTPUT_PREFIX}_seq${SEQ}_gb${GLOBAL_BATCH}" - OUTPUT_PREFIX="${OUTPUT_PREFIX}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" - OUTPUT_PREFIX="${OUTPUT_PREFIX}_lr${LR}_lwf${LR_WARMUP_FRAC}" + pre="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" + pre="${pre}_hs${HIDDEN}_mb${MICRO_BATCH}" + pre="${pre}_seq${SEQ}_gb${GLOBAL_BATCH}" + pre="${pre}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" if [[ -n "${LR_DECAY_ITERS}" ]]; then - OUTPUT_PREFIX="${OUTPUT_PREFIX}_ldi${LR_DECAY_ITERS}" + pre="${pre}_ldi${LR_DECAY_ITERS}" fi if [[ -z "${NO_FLASH_ATTN:-}" ]]; then - OUTPUT_PREFIX="${OUTPUT_PREFIX}_flash" + pre="${pre}_flash" fi - export OUTPUT_PREFIX="${OUTPUT_PREFIX}" + export OUTPUT_PREFIX="${pre}" + echo "${pre}" +} + +setOutput() { # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" + OUTPUT_PREFIX=$(get_output_prefix) OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" @@ -665,7 +672,7 @@ setup_tokenizer_and_data() { export TOKENIZER_TYPE="Llama2" tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - export TOKENIZER_FLAGS="${TOKENIZER_FLAGS} --tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" + export TOKENIZER_FLAGS="--tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" setData "${dfl}" || exit @@ -737,13 +744,14 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ndocs=$(wc -l < "${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") + # dcp="${OUTPUT_PREFIX:-$(get_output_prefix)}/.cache/${dfl_stem}/index-cache" dcp=".cache/${dfl_stem}/index-cache" export DATA_FILE_LIST="${dfl}" export NUM_DOCS="${ndocs}" export WEIGHT_SUM="${ws}" export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" - export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST} --data-cache-path ${DATA_CACHE_PATH}" + export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" echo "--------------------" echo "Updated environment:" printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" @@ -751,6 +759,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" printf "DFL_STEM: %s\n" "${DFL_STEM}" printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" + printf "DATA_FLAGS: %s\n" "${DATA_FLAGS}" echo "--------------------" # fi # export DATA_FLAGS="${DATA_FLAGS}" diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 9a6d39923c0..bf346e2144d 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -95,6 +95,7 @@ TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" +echo "Using data_cache_path: ${data_cache_path}" export DEFAULTS="\ @@ -116,7 +117,6 @@ custom_args=" $@" # --log-memory-to-tensorboard \ # --data-file-list ${DATA_FILE_LIST} \ # --data-file-list ${DATA_FILE_LIST} \ - # --data-cache-path ${data_cache_path} \ # --tokenizer-type Llama2Tokenizer \ # --tokenizer-model ${TOKENIZER_MODEL} \ run_cmd=" @@ -143,10 +143,11 @@ run_cmd=" --pipeline-model-parallel-size ${PP} \ --num-key-value-heads ${NUM_KV_HEAD} \ --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ + --data-cache-path ${data_cache_path} \ + ${DATA_FLAGS} \ ${LR_ARGS} \ ${LLAMA_ARGS} \ ${TIMING_STR} \ - ${DATA_FLAGS} \ ${TOKENIZER_FLAGS} \ $ds_args \ ${gpt_args[*]} \ From c3a4451f646b3ef17a1a8f2d371ae159974c4ca3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 25 May 2024 11:21:00 -0500 Subject: [PATCH 160/590] Add `ALCF/sunspot-env-2024-04-15-002.sh` --- ALCF/sunspot-env-2024-04-15-002.sh | 4 ++++ 1 file changed, 4 insertions(+) create mode 100644 ALCF/sunspot-env-2024-04-15-002.sh diff --git a/ALCF/sunspot-env-2024-04-15-002.sh b/ALCF/sunspot-env-2024-04-15-002.sh new file mode 100644 index 00000000000..3b7155675df --- /dev/null +++ b/ALCF/sunspot-env-2024-04-15-002.sh @@ -0,0 +1,4 @@ +#!/bin/bash --login + +module use /soft/preview-modulefiles/24.086.0 +module load frameworks/2024.04.15.002.lua From 0fc3919d1587adc3b13b4afe2f9e1b90cbd6874f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 25 May 2024 11:22:03 -0500 Subject: [PATCH 161/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 72322d655c9..9dc0f1d9469 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -25,6 +25,6 @@ echo "+---------------------------------------------------------+" export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 echo "${OUTFILE}" >> "${OUTDIR}/latest" # export DEBUG=1 -export MICRO_BATCH="${MICRO_BATCH:-${MBS}}" +# export MICRO_BATCH="${MICRO_BATCH:-${MBS}}" export DATA_FILE_LIST="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt}" bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From efb2a3a00918c40d0b176e7b063f606104fa7962 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 28 May 2024 10:07:10 -0500 Subject: [PATCH 162/590] added a barrier to make sure all the datasets are built before other ranks touch it --- megatron/data/gpt_dataset.py | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index bb381567a8c..171f55d3af0 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -95,12 +95,21 @@ def Build(self): valid_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, datasets_train_valid_test_num_samples[i], seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, 'valid') for i in range(len(weights))] + return_doc_ids,data_cache_path, 'valid') for i in range(len(weights))] + for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): + valid_datasets[i].Build() if c > 0: test_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, datasets_train_valid_test_num_samples[i], seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, 'test') for i in range(len(weights))] + return_doc_ids,data_cache_path, 'test') for i in range(len(weights))] + for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): + test_datasets[i].Build() + + # This barrier is critical to make sure that all the datasets are built once + # and the metadata were written to the cache folder before other ranks touch them + torch.distributed.barrier() + print_rank_0(" >>> Finished building datasets in distributed way ... ") # Blend. From ccf8835e6c67d8ca14745e18c43d1650f04b77e7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 29 May 2024 09:14:58 -0500 Subject: [PATCH 163/590] Update `{train_llama_alcf.sh,ALCF/helpers.sh}` --- ALCF/helpers.sh | 196 +++++++++++++++++++++++++++++++++++++++----- train_llama_alcf.sh | 133 +++--------------------------- 2 files changed, 188 insertions(+), 141 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index c0927766f27..6f77c565197 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -15,6 +15,120 @@ fi export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} +############################################################################## +# `setup`: All-in-one helper function. +# +# - Explicitly, this will: +# 01. Identify the machine we're on +# 02. Setup `python` +# 1. Load `conda` +# 2. Setup `venv` on top of `conda` +# 03. Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) +# 1. Additionally, call `source deps/ezpz/src/ezpz/bin/savejobenv`, +# which will automatically build a `alias launch=mpiexec ...` +# according to the specifics of our active job. +# 04. Set runtime options +# 05. Build `deepspeed_config.json` +# 06. Build {logs, checkpoints, etc} dirs, named according to specifics of +# current run +# 07. Specify additional `deepspeed` arguments +# 08. Ensure executable exists at expected path +# 09. Setup data + tokenizer via `TOKENIZER_TYPE` +# 10. Print job info +# 11. Save `.env` to `CKPT_DIR` for safe keeping +# 12. Check that we're not already running, and if so, exit. +# 13. Setup run command to be executed. +############################################################################## +setup() { + get_machine || exit # 01. Identify machine we're on + setEnv || exit # 02. Load `conda` environment + # saveDSenv || exit + ezpz || exit # 03. Determine WORLD_SIZE, etc. from `PBS_*` vars + setParams || exit # 04. Set command line arguments to pass to `"${EXEC}"` + buildDSconfig || exit # 05. Create `deepspeed_config.json` from runtime params from ^ + setOutput || exit # 06. Specify output directory for {logs, checkpoints, etc.} + setArgs || exit # 07. Specify additional `deepspeed` arguments + export EXEC="${EXEC:-${HERE}/pretrain_gpt_alcf.py}" + check_executable "${EXEC}" # 08. Ensure executable exists in expected path + dfl="${DATA_FILE_LIST:-}" # 09. Setup data + tokenizer + tok="${TOKENIZER_TYPE:-Llama2}" # via `DATA_FILE_LIST` and `TOKENIZER_TYPE` + setup_tokenizer_and_data "${tok}" "${dfl}" || exit + printJobInfo || exit # 10. Print job info + save_dotenv "${CKPT_DIR}" || exit # 11. Print info about loaded modules and runtime environment + check_and_kill_if_running || exit # 12. Check that were not already running, if so, exit. + setup_run_cmd || exit # 13. Setup run command to be executed +} + +##################################################### +# `setup_run_cmd`: Build run command to be executed. +##################################################### +setup_run_cmd() { + #### Make it easy to track experiments by date ################### + year="$(date "+%Y")" + month="$(date "+%m")" + day="$(date "+%Y-%m-%d")" + today="$(date "+%Y-%m-%d")" # kept for backwards compatibility + started_at="$(date "+%Y-%m-%d-%H%M%S")" + export YEAR="${year}" + export MONTH="${month}" + export DAY="${day}" + export TODAY="${today}" + export STARTED_AT="${started_at}" + ################################################################## + setupLauncher || exit + TBDIR="${CKPT_DIR}/tensorboard" + mkdir -p "${TBDIR}" + export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" + printf "\n" + echo "Using data_cache_path: ${data_cache_path}" + export DEFAULTS="\ + --split 100,0,0 \ + --log-interval 1 \ + --no-bias-gelu-fusion \ + --no-bias-dropout-fusion \ + --no-masked-softmax-fusion \ + --no-gradient-accumulation-fusion \ + --accumulate-allreduce-grads-in-fp32 \ + --use-checkpoint-opt_param-scheduler \ + --log-timers-to-tensorboard \ + --log-optimizer-states-to-tensorboard" + if [[ "${SP}" -ge 2 ]]; then + export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" + fi + run_cmd=" + ${LAUNCHER} \ + --${DTYPE} \ + ${DEFAULTS} \ + --optimizer ${OPT} \ + --save ${CKPT_DIR} \ + --load ${CKPT_DIR} \ + --seq-length ${SEQ} \ + --num-layers ${NLAYERS} \ + --hidden-size ${HIDDEN} \ + --train-iters ${TRAIN_ITER} \ + --tensorboard-dir ${TBDIR} \ + --eval-iters ${EVAL_ITERS} \ + --distributed-backend ${BE} \ + --num-attention-heads ${HEADS} \ + --save-interval ${SAVE_INTERVAL} \ + --eval-interval ${EVAL_INTERVAL} \ + --max-position-embeddings ${SEQ} \ + --micro-batch-size ${MICRO_BATCH} \ + --tensor-model-parallel-size ${TP} \ + --global-batch-size ${GLOBAL_BATCH} \ + --pipeline-model-parallel-size ${PP} \ + --num-key-value-heads ${NUM_KV_HEAD} \ + --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ + --data-cache-path ${data_cache_path} \ + ${DATA_FLAGS} \ + ${LR_ARGS} \ + ${LLAMA_ARGS} \ + ${TIMING_STR} \ + ${TOKENIZER_FLAGS} \ + $ds_args \ + ${gpt_args[*]} + " +} save_dotenv() { if [[ "$#" -ne 1 ]]; then @@ -133,6 +247,19 @@ loadCondaEnv() { } +############################################################################# +# `setupLauncher`: Launch with one of `{mpiexec, deepspeed}`. +# +# Explicitly, look for `LAUNCH_CMD` in environment and launch accordingly. +# Will use `mpiexec` by default. +# To launch with `deepspeed` instead, specify `LAUNCH_CMD=deepspeed`, e.g. +# +# ```bash +# PBS_O_WORKDIR=$(pwd) LAUNCH_CMD=deepspeed bash train_llama_alcf.sh +# ``` +# +# will launch with `deepspeed` instead of `mpiexec`. +############################################################################# setupLauncher() { # outdir=$1 if [[ "${LAUNCH_CMD:-"MPICH"}" == "deepspeed" ]]; then @@ -175,6 +302,17 @@ set_lr_args() { } +######################################################################### +# `get_batch_size_on_polaris`: Identify MICRO_BATCH to use on Polaris. +# +# - In particular, it seems that different node counts allow for different +# `MICRO_BATCH` sizes. +# Explicitly: +# - [NHOSTS <= 3]: `MICRO_BATCH=1` +# - [NHOSTS >= 4]: `MICRO_BATCH=2` +# - [NHOSTS >= 8]: `MICRO_BATCH=4` +# are the largest batch sizes that fit in memory at various node counts. +######################################################################### get_batch_size_on_polaris() { if [[ $(hostname) == x3* ]]; then local nhosts=$(wc -l < "${PBS_NODEFILE}") @@ -189,6 +327,16 @@ get_batch_size_on_polaris() { echo "${mbs}" } + +############################################################################## +# `setParams`: Set / configure run options by parsing environment. +# +# - any of the declared options below can be overridden +# dynamically at runtime, e.g. to run with a `MICRO_BATCH` size of 2: +# ```bash +# $ PBS_O_WORKDIR=$(pwd) MICRO_BATCH=2 bash train_llama_alcf.sh +# ``` +############################################################################## setParams() { LLAMA_ARGS="" # +----[Parallelism Settings] -------------------------------------------+ @@ -243,6 +391,7 @@ setParams() { # +----------------------------------------------------------------------+ export TP="${TP}" export PP="${PP:-1}" + export SP="${SP:-1}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" @@ -252,7 +401,7 @@ setParams() { fi export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" # +---[Llama2 7B Config]--------------------------------------------------+ - export MODEL_KEY="Llama-7B" + # export MODEL_KEY="Llama-7B" export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE @@ -270,13 +419,18 @@ setParams() { export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? - export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP )) # MAX GLOBAL BATCH SIZE + export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP / $SP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ - export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" + if [[ "${SP}" == 1 ]]; then + export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" + else + export LLAMA_ARGS="" + echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" + fi export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS @@ -397,8 +551,11 @@ get_output_prefix() { pre="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" pre="${pre}_hs${HIDDEN}_mb${MICRO_BATCH}" pre="${pre}_seq${SEQ}_gb${GLOBAL_BATCH}" - pre="${pre}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" + pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" + if [[ -n "${TOKENIZER_TYPE:-}" ]]; then + pre="${pre}_tok${TOKENIZER_TYPE}" + fi if [[ -n "${LR_DECAY_ITERS}" ]]; then pre="${pre}_ldi${LR_DECAY_ITERS}" fi @@ -413,11 +570,10 @@ setOutput() { # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" OUTPUT_PREFIX=$(get_output_prefix) OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" - export OUTPUT_DIR="${OUTPUT_DIR}" + export OUTPUT_DIR="${OUTPUT_DIR}" && mkdir -p "${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" echo "${OUTPUT_LOG}" >> "logs/latest" - mkdir -p "${OUTPUT_DIR}" printf "\n Please see logs at: %s\n" $(printGreen "${OUTPUT_DIR}") printf "Checkpoints will be saved to: %s\n" $(printYellow "${CKPT_DIR}") } @@ -467,7 +623,8 @@ sumFiles() { setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ if [[ -z "${CONDA_PREFIX:-}" ]]; then - module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua + # module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua + source "${WORKING_DIR}/ALCF/sunspot-env-2024-q2.sh" fi } @@ -573,16 +730,16 @@ setEnv() { # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then # ----- [Aurora] -------------------------------------------- - if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then - if [[ $(hostname) == x4* ]]; then - # TODO: Update once Aurora back online - eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 - # ----- [SunSpot] --------------------------------------- - elif [[ $(hostname) == x1* ]]; then - echo "Running on SunSpot !!" - setup_conda_sunspot - fi + # if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + if [[ $(hostname) == x4* ]]; then + # TODO: Update once Aurora back online + eval "$(conda shell.zsh hook)" && conda activate anl_release_q4v2 + # ----- [SunSpot] --------------------------------------- + elif [[ $(hostname) == x1* ]]; then + echo "Running on SunSpot !!" + setup_conda_sunspot fi + # fi # MPICH_MODULES=$(echo $LOADEDMODULES | tr ':' '\n' | grep mpich) # if [[ -z "${MPICH_MODULES" ]]; then # source "${WORKING_DIR}/ALCF/sunspot-env.sh" || exit @@ -718,18 +875,15 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] # shellcheck: source ./data-lists/sunspot/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/books.txt" - + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt" elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] if [[ "${PBS_O_HOST}" == sirius* ]]; then # -------------------[SIRIUS] # shellcheck: source ./data-lists/sirius/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/books.txt" - + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma_v1_7_file_list.txt" elif [[ "${PBS_O_HOST}" == polaris* ]]; then # ---------------[POLARIS] # shellcheck: source ./data-lists/polaris/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" fi - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index bf346e2144d..98ef99abfd0 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -5,21 +5,6 @@ #PBS -l select=48 #PBS -l filesystems=eagle:home - -#### Make it easy to track experiments by date ################### -year="$(date "+%Y")" -month="$(date "+%m")" -day="$(date "+%Y-%m-%d")" -today="$(date "+%Y-%m-%d")" # kept for backwards compatibility -started_at="$(date "+%Y-%m-%d-%H%M%S")" -export YEAR="${year}" -export MONTH="${month}" -export DAY="${day}" -export TODAY="${today}" -export STARTED_AT="${started_at}" -################################################################## - - ############################################################################# # Check if running in `DEBUG=1` mode. # - If so, this will print each command before it is ran and exit if any of @@ -42,120 +27,28 @@ sourceFile() { fp="$1" echo "source-ing ${fp}" if [[ -f "${fp}" ]]; then - # shellcheck source="${fp}" source "${fp}" else echo "ERROR: UNABLE TO SOURCE ${fp}" fi } -############################################################################## -###################### MAIN LOGIC ############################################ -# ----[0. Navigate into `$PBS_O_WORKDIR`]-------------------------------------- +##################### +# MAIN PROGRAM LOGIC +##################### +# ----[1. Navigate into `$PBS_O_WORKDIR`]-------------------------------------- cd "${PBS_O_WORKDIR}" || exit -HERE=$(python3 -c 'import os; print(os.getcwd())') -export HERE - -# ----[1. Assert `./pretrain_gpt_alcf.py` exists:]----------------------------- -export EXEC="${HERE}/pretrain_gpt_alcf.py" -[ -f "${EXEC}" ] || exit - -# ----[1.5 Keep track of ] -exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") -export EXEC_STEM="${exec_stem}" - - -# ----[2. `source ./ALCF/helpers_alcf.sh`:]------------------------------------ +HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE sourceFile "${HERE}/ALCF/helpers.sh" || exit - -# ----[3. Call fns from `./ALCF/helpers_alcf.sh`]------------------------------ -get_machine || exit # 01. Identify machine we're on -setEnv || exit # 02. Load `conda` environment -# saveDSenv || exit # 03. Save env vars to `.deepspeed_env` -ezpz || exit # 04. Determine WORLD_SIZE, etc. from `PBS_*` vars -setParams || exit # 05. Set command line arguments to pass to `"${EXEC}"` -buildDSconfig || exit # 06. Create `deepspeed_config.json` from runtime params from ^ -setOutput || exit # 07. Specify output directory for {logs, checkpoints, etc.} -setArgs || exit # 08. Specify additional `deepspeed` arguments -dfl="${DATA_FILE_LIST:-}" # 09. Setup data + tokenizer -tok="${TOKENIZER_TYPE:-Llama2}" # via `DATA_FILE_LIST` and `TOKENIZER_TYPE` -setup_tokenizer_and_data "${tok}" "${dfl}" || exit -printJobInfo || exit # 10. Print job info -setupLauncher || exit # 11. set launcher to one of `MPICH` (default), or `deepspeed` -save_dotenv "${CKPT_DIR}" || exit # 12. Print info about loaded modules and runtime environment -check_and_kill_if_running || exit # 13. Check that were not already running, if so, exit. -# ----------------------------------------------------------------------------- -############################################################################## - -################################################ -# Assert `$TBDIR` exists inside our `$CKPT_DIR` -# to ensure metrics are tied to checkpoint -################################################ -TBDIR="${CKPT_DIR}/tensorboard" -mkdir -p "${TBDIR}" - -data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" -echo "Using data_cache_path: ${data_cache_path}" - - -export DEFAULTS="\ - --split 100,0,0 \ - --log-interval 1 \ - --no-bias-gelu-fusion \ - --no-bias-dropout-fusion \ - --no-masked-softmax-fusion \ - --no-gradient-accumulation-fusion \ - --accumulate-allreduce-grads-in-fp32 \ - --use-checkpoint-opt_param-scheduler \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard" +setup || exit +############################################################################### # Take custom args -custom_args=" $@" - - # --log-num-zeros-in-grad \ - # --log-memory-to-tensorboard \ - # --data-file-list ${DATA_FILE_LIST} \ - # --data-file-list ${DATA_FILE_LIST} \ - # --tokenizer-type Llama2Tokenizer \ - # --tokenizer-model ${TOKENIZER_MODEL} \ -run_cmd=" - ${LAUNCHER} \ - --${DTYPE} \ - ${DEFAULTS} \ - --optimizer ${OPT} \ - --save ${CKPT_DIR} \ - --load ${CKPT_DIR} \ - --seq-length ${SEQ} \ - --num-layers ${NLAYERS} \ - --hidden-size ${HIDDEN} \ - --train-iters ${TRAIN_ITER} \ - --tensorboard-dir ${TBDIR} \ - --eval-iters ${EVAL_ITERS} \ - --distributed-backend ${BE} \ - --num-attention-heads ${HEADS} \ - --save-interval ${SAVE_INTERVAL} \ - --eval-interval ${EVAL_INTERVAL} \ - --max-position-embeddings ${SEQ} \ - --micro-batch-size ${MICRO_BATCH} \ - --tensor-model-parallel-size ${TP} \ - --global-batch-size ${GLOBAL_BATCH} \ - --pipeline-model-parallel-size ${PP} \ - --num-key-value-heads ${NUM_KV_HEAD} \ - --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - --data-cache-path ${data_cache_path} \ - ${DATA_FLAGS} \ - ${LR_ARGS} \ - ${LLAMA_ARGS} \ - ${TIMING_STR} \ - ${TOKENIZER_FLAGS} \ - $ds_args \ - ${gpt_args[*]} \ - $custom_args \ - |& tee ${OUTPUT_LOG} - " +export custom_args=" $@" +export run_cmd="${run_cmd} ${custom_args}" -echo "${run_cmd}" -printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow ${OUTPUT_LOG})" -eval "${run_cmd}" +echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" +# eval "${run_cmd}" >> "${OUTPUT_LOG}" 2>&1 & +eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" set +x From 358139fb8b5699c8cd84a505b524aaa6128cad74 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 30 May 2024 17:02:45 -0500 Subject: [PATCH 164/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 58 +++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 46 insertions(+), 12 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 6f77c565197..24e048c06d5 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -40,6 +40,10 @@ printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} # 13. Setup run command to be executed. ############################################################################## setup() { + python3 -m ezpz.jobs && source "./.jobenv" + # echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" + # source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 + # source "${ezloc}/src/ezpz/bin/getjobenv" || exit get_machine || exit # 01. Identify machine we're on setEnv || exit # 02. Load `conda` environment # saveDSenv || exit @@ -57,6 +61,7 @@ setup() { save_dotenv "${CKPT_DIR}" || exit # 11. Print info about loaded modules and runtime environment check_and_kill_if_running || exit # 12. Check that were not already running, if so, exit. setup_run_cmd || exit # 13. Setup run command to be executed + # make_ds_hostfile || exit } ##################################################### @@ -75,7 +80,10 @@ setup_run_cmd() { export TODAY="${today}" export STARTED_AT="${started_at}" ################################################################## - setupLauncher || exit + # to launch with DeepSpeed instead of mpiexec, run with: + # $ LAUNCH_WITH=deepspeed bash train_llama_alcf.sh + ################################################################## + setupLauncher "${LAUNCH_WITH:-MPICH}" || exit TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" @@ -262,15 +270,27 @@ loadCondaEnv() { ############################################################################# setupLauncher() { # outdir=$1 - if [[ "${LAUNCH_CMD:-"MPICH"}" == "deepspeed" ]]; then + if [[ "$#" == 1 ]]; then + local dist_launcher="$1" + else + local dist_launcher="${LAUNCH_WITH:-${LAUNCH_CMD:-"MPICH"}}" + fi + if [[ "${dist_launcher}" == "deepspeed" ]]; then # Assert `./hostfile_deepspeed` exists export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" # if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then else - export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + if [[ -n "${DIST_LAUNCH}" ]]; then + export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + else + echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" + printenv | tee ".env-${PBS_JOBID}" + echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" + fi fi - printf "Launching with: %s\n" "$(printRed "${LAUNCH_CMD}")" + + printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" printf " %s" "$(printMagenta ${LAUNCHER})" } @@ -491,6 +511,16 @@ make_ds_hostfile() { sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } +ezpz_savejobenv() { + local outfile="${WORKING_DIR}/savejobenv" + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" +} + +ezpz_getjobenv() { + local outfile="${WORKING_DIR}/getjobenv" + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" +} + # +---------------------------------------+ # | 1. Git clone ezpz (if not found) | # | 2. Install ezpz (if not installed) | @@ -504,9 +534,11 @@ ezpz() { fi ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -n "${ezloc}" ]]; then - echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" - source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 - source "${ezloc}/src/ezpz/bin/getjobenv" || exit + ezpz_savejobenv + python3 -m ezpz.jobs && source "./.jobenv" + # echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" + # source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 + # source "${ezloc}/src/ezpz/bin/getjobenv" || exit make_ds_hostfile || exit else echo "No ezpz detected. Attempting to install with $(which python3)" @@ -662,13 +694,15 @@ setup_venv_from_conda() { echo "No ${CONDA_PREFIX} found." # Exiting." # exit 1 else - if [[ -n "${VIRTUAL_ENV}" ]]; then - echo "Already inside virtual env at ${VENV_DIR}!" - elif [[ -z "${VIRTUAL_ENV}" ]]; then + echo "Found conda at: ${CONDA_PREFIX}" + CONDA_NAME=$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') + export CONDA_NAME + # if [[ -n "${VIRTUAL_ENV}" ]]; then + # echo "Already inside virtual env at ${VENV_DIR}!" + if [[ -z "${VIRTUAL_ENV}" ]]; then echo "No VIRTUAL_ENV found in environment!" echo " - Trying to setup from ${CONDA_PREFIX}" - CONDA_NAME=$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') - VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" + export VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" echo " - Using VENV_DIR=${VENV_DIR}" # VENV_DIR="venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}')" # VENV_DIR="${WORKING_DIR}/venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')" From 47585ed00d1c4781b1424510cd3b8dd7259f3308 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 30 May 2024 21:02:34 -0500 Subject: [PATCH 165/590] Concat datasets that belongs the same corpus --- megatron/data/blendable_dataset.py | 17 +----- megatron/data/dataset_utils.py | 12 ++++ megatron/data/gpt_dataset.py | 90 ++++++++++++++++++++---------- pretrain_gpt_alcf.py | 8 ++- 4 files changed, 79 insertions(+), 48 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 3f4ea32c637..32b2005516b 100644 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -122,19 +122,4 @@ def __getitem__(self, idx): return { "dataset_idx" : dataset_idx, **self.datasets[dataset_idx][sample_idx], - } - -class DistributedBlendableDataset(BlendableDataset): - def __getitem__(self, idx): - dataset_idx = self.dataset_index[idx] - sample_idx = self.dataset_sample_index[idx] - if (self.datasets[dataset_idx].build): - dataset = self.datasets[dataset_idx].dataset - else: - print_flush(f" First time reading samples from {self.datasets[dataset_idx].prefix}. Building the dataset now.") - dataset = self.datasets[dataset_idx].Build() - return { - "dataset_idx" : dataset_idx, - **dataset[sample_idx], - } - \ No newline at end of file + } diff --git a/megatron/data/dataset_utils.py b/megatron/data/dataset_utils.py index cadca053cf5..b73a6cf85fd 100644 --- a/megatron/data/dataset_utils.py +++ b/megatron/data/dataset_utils.py @@ -40,6 +40,18 @@ DSET_TYPES = [DSET_TYPE_BERT, DSET_TYPE_ICT, DSET_TYPE_T5] +def get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_num_samples): + assert len(data_prefix) % 3 == 0 + num_datasets = len(data_prefix) // 3 + data_new_prefix = [] + corpuses = [] + for i in range(num_datasets): + data_new_prefix += [data_prefix[3*i], data_prefix[3*i+1]] + corpuses.append(data_prefix[3*i+2]) + prefixes, weights, datasets_train_valid_test_num_samples = \ + get_datasets_weights_and_num_samples(data_prefix, + train_valid_test_num_samples) + return prefixes, corpuses, weights, datasets_train_valid_test_num_samples def get_datasets_weights_and_num_samples(data_prefix, train_valid_test_num_samples): diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 171f55d3af0..98089dfdcb7 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -11,8 +11,8 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0, is_rank_0, get_args, print_flush from megatron.core import mpu -from megatron.data.blendable_dataset import BlendableDataset, DistributedBlendableDataset -from megatron.data.dataset_utils import get_datasets_weights_and_num_samples +from megatron.data.blendable_dataset import BlendableDataset +from megatron.data.dataset_utils import get_datasets_weights_and_num_samples, get_datasets_corpuses_weights_and_num_samples from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset @@ -39,9 +39,10 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, # Blending dataset. # Parse the values. - output = get_datasets_weights_and_num_samples(data_prefix, + output = get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_num_samples) - prefixes, weights, datasets_train_valid_test_num_samples = output + prefixes, corpuses, weights, datasets_train_valid_test_num_samples = output + corpus_list = set(corpuses) train_num_samples, valid_num_samples, test_num_samples = map( sum, zip(*datasets_train_valid_test_num_samples) @@ -51,10 +52,10 @@ class DatasetBuilder: ''' This is for building individual dataset from each dataset file ''' - def __init__(self, prefix, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - return_doc_ids, - data_cache_path=data_cache_path, name='train'): + def __init__(self, prefix, corpus, data_impl, splits_string, + num_samples, seq_length, seed, skip_warmup, + return_doc_ids, + data_cache_path=data_cache_path, name='train'): self.prefix = prefix self.data_impl = data_impl self.splits_string = splits_string @@ -68,6 +69,7 @@ def __init__(self, prefix, data_impl, splits_string, self.name = name self.desc = prefix + f"{num_samples}" + f"{seq_length}" + f"{seed}" self.build = False + self.corpus = corpus def Build(self): self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, self.num_samples, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, @@ -75,36 +77,66 @@ def Build(self): self.build = True return self.dataset + def BuildConcatDatasets(Dataset): + def __init__(self, dataset_builders): + self.dataset_builders = dataset_builders + self.num_datasets = len(dataset_builders) + self.num_samples = np.sum([d.num_samples for d in dataset_builders]) + self.indices=[] + self.desc="ConcatDatasets" + for i in range(self.num_datasets): + self.desc += dataset_builders[i].prefix + dataset_builders[i].corpus + for j in range(dataset_builders[i].num_samples): + self.indices.append((i, j)) + self.desc += f"{self.num_samples}" + f"{seq_length}" + f"{seed}" + def __len__(self): + return self.num_samples + def __getitem__(self, idx): + i, j = self.indices[idx] + if self.dataset_builders[i].build: + return self.dataset_builders.dataset[j] + else: + return self.dataset_builders[i].Build()[j] + + # Predetermine whether need to build the specific dataset or not. print_rank_0(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] + train_datasets = [] valid_datasets = [] test_datasets = [] # Build individual datasets. - if a > 0: - train_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, 'train') for i in range(len(weights))] + def build_corpus_datasets(dataset_type='train'): + datasets = [] + corpus_builders = {} + corpus_weights = {} + for c in corpus_list: + corpus_builders[c] = [] + corpus_weights[c] = 0.0 + dataset_builders = [DatasetBuilder(prefixes[i], corpuses[i], data_impl, splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, seed, skip_warmup, + return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): - train_datasets[i].Build() + dataset_builders[i].Build() + for i, d in zip(range(len(weights)), train_datasets_builders): + corpus_builders[d.corpus].append(d) + corpus_weights[d.corpus] += weights[i] + for c in corpus_list: + datasets.append(BuildConcatDatasets(train_corpus_builders[c])) + + return datasets, [corpus_weights[c] for c in corpus_list] + + if a > 0: + train_datasets, train_weights = build_corpus_datasets('train') + if b > 0: - valid_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, 'valid') for i in range(len(weights))] - for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): - valid_datasets[i].Build() + valid_datasets, valid_weights = build_corpus_datasets('valid') if c > 0: - test_datasets = [DatasetBuilder(prefixes[i], data_impl, splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, 'test') for i in range(len(weights))] - for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): - test_datasets[i].Build() + test_datasets, test_weights = build_corpus_datasets('test') # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them @@ -115,15 +147,15 @@ def Build(self): # Blend. blending_train_dataset = None if train_datasets: - blending_train_dataset = DistributedBlendableDataset(train_datasets, weights, train_num_samples, + blending_train_dataset = BlendableDataset(train_datasets, train_weights, train_num_samples, data_cache_path=data_cache_path) blending_valid_dataset = None if valid_datasets: - blending_valid_dataset = DistributedBlendableDataset(valid_datasets, weights, valid_num_samples, + blending_valid_dataset = BlendableDataset(valid_datasets, valid_weights, valid_num_samples, data_cache_path=data_cache_path) blending_test_dataset = None if test_datasets: - blending_test_dataset = DistributedcBlendableDataset(test_datasets, weights, test_num_samples, + blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, data_cache_path=data_cache_path) return (blending_train_dataset, blending_valid_dataset, diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 94186fea83b..56a8d9919c8 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -498,13 +498,16 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): '> building train, validation, and test datasets for GPT ...' ) files = [] + # making it the default input method + assert(agrs.data_file_list is not None) if args.data_file_list is not None: log.info(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: for f in flist.readlines(): - w, fname = f.split() + w, fname, c = f.split() files.append(float(w)) files.append(fname) + files.append(c) elif len(args.data_path) == 1 and os.path.isdir(args.data_path[0]): path = args.data_path[0] + "/" for f in os.listdir(path): @@ -521,8 +524,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): train_valid_test_num_samples=train_val_test_num_samples, seq_length=args.seq_length, seed=args.seed, - skip_warmup=True, - # skip_warmup=(not args.mmap_warmup), + skip_warmup=(not args.mmap_warmup), train_data_prefix=args.train_data_path, valid_data_prefix=args.valid_data_path, test_data_prefix=args.test_data_path, From 2b5b41f7acb8bb685e7e7f794af12ef6af317674 Mon Sep 17 00:00:00 2001 From: Varuni Sastry <88804132+vksastry@users.noreply.github.com> Date: Thu, 30 May 2024 22:29:53 -0500 Subject: [PATCH 166/590] convert MDS checkpoint to Hf Llama model --- ALCF/mds_to_hf.py | 91 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 91 insertions(+) create mode 100644 ALCF/mds_to_hf.py diff --git a/ALCF/mds_to_hf.py b/ALCF/mds_to_hf.py new file mode 100644 index 00000000000..a3367882740 --- /dev/null +++ b/ALCF/mds_to_hf.py @@ -0,0 +1,91 @@ +# Usage : python mds_to_hf.py --mds_checkpoint --output_dir +# Tips : Do not run on login node. +# This script currently only takes care of tp=1. Takes a AuroraGPT Llama model trained with Megatron-DeepSpeed and converts to LLamaCausalForLM architecture from HuggingFace. + +import argparse +import torch +import pdb +import os +from transformers import LlamaConfig, LlamaForCausalLM, LlamaTokenizer + +def repeat_kv_wt(x,np): + return torch.repeat_interleave(x, dim=0, repeats=np) + +def Update_llama_config(Llama_config, mds_args): + if mds_args['swiglu']: + Llama_config.hidden_act = "silu" + Llama_config.hidden_size = mds_args['hidden_size'] + Llama_config.intermediate_size = mds_args['ffn_hidden_size'] + Llama_config.max_position_embeddings = mds_args['max_position_embeddings'] + Llama_config.num_attention_heads = mds_args['num_attention_heads'] + Llama_config.num_hidden_layers = mds_args['num_layers'] + Llama_config.num_key_value_heads = mds_args['num_key_value_heads'] + Llama_config.rms_norm_eps = mds_args['layernorm_epsilon'] + Llama_config.rope_theta = mds_args['rope_theta'] + Llama_config.vocab_size = mds_args['padded_vocab_size'] + return Llama_config + + +if __name__ == "__main__": + parser = argparse.ArgumentParser() + parser.add_argument('--mds_checkpoint', required=True) + parser.add_argument('--output_dir', required=True) + args = parser.parse_args() + + # make output_dir if it does not exits. + if not os.path.exists(args.output_dir): + os.makedirs(args.output_dir) + + filename = str(args.mds_checkpoint) + if not filename.split("/")[-1].startswith('mp_rank') and not filename.split("/")[-1].endswith('.pt'): + assert ("Provide the right file path, The file should be of format mp_rank_*.pt") + print(f"loading mds checkpoint {filename}") + + mds_model = torch.load(args.mds_checkpoint,map_location=torch.device('cpu')) + Llama_model = LlamaForCausalLM.from_pretrained("meta-llama/Llama-2-7b-hf",cache_dir='/eagle/datascience/vsastry/huggingface') + + Llama_config = Llama_model.config + Updated_Llama_config = Update_llama_config(Llama_config, mds_model['args'].__dict__) + # save the updated config.json file + Updated_Llama_config.to_json_file(os.path.join(args.output_dir,'config.json')) + + state_dict = {} + dim = mds_model['args'].__dict__['kv_channels'] + inv_freq = 1.0 / (mds_model['args'].__dict__['rope_theta'] ** (torch.arange(0,dim, 2).float() / dim)) + hidden_size = mds_model['args'].__dict__['hidden_size'] + kv_dim = mds_model['args'].__dict__['kv_channels'] * mds_model['args'].__dict__['num_key_value_heads'] + kv_groups = mds_model['args'].__dict__['num_attention_heads'] // mds_model['args'].__dict__['num_key_value_heads'] + for layer_i in range(Updated_Llama_config.__dict__['num_hidden_layers']): + # SELF ATTENTION layers. + # get the q, k, v weights separately. Keeping k and v at the GQA head dim, since the transformers/models/llama/modelling_utils will take care of it. + fused_qkv = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.query_key_value.weight"] + state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = fused_qkv[0:hidden_size] + state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = fused_qkv[hidden_size:hidden_size+kv_dim] + #state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size:hidden_size+kv_dim], kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim] + #state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim],kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.dense.weight"] + + # MLP Layers + fused_mlp = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_h_to_4h.weight"] + chunked_mlp = torch.chunk(fused_mlp,2,dim=0) + state_dict[f"model.layers.{layer_i}.mlp.gate_proj.weight"] = chunked_mlp[0] + state_dict[f"model.layers.{layer_i}.mlp.up_proj.weight"] = chunked_mlp[1] + state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_4h_to_h.weight"] + + #LayerNorm weights and RoPe + state_dict[f"model.layers.{layer_i}.input_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.input_layernorm.weight"] + state_dict[f"model.layers.{layer_i}.post_attention_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.post_attention_layernorm.weight"] + + state_dict[f"model.layers.{layer_i}.self_attn.rotary_emb.inv_freq"] = inv_freq + + # Get the non-encoder layer weights. + state_dict["model.embed_tokens.weight"] = mds_model['module']['language_model']['embedding']['word_embeddings']['weight'] + state_dict["model.norm.weight"] = mds_model['module']['language_model']['encoder']['final_layernorm.weight'] + state_dict["lm_head.weight"] = mds_model['module']['language_model']['output_layer']['weight'] + + # Save the model in the hf output path. + torch.save(state_dict, os.path.join(args.output_dir,"pytorch_model.bin")) + + + From 10a34ea26ca7a29cce713c4177031514fcc13ce2 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 31 May 2024 05:06:31 +0000 Subject: [PATCH 167/590] fixed bugs --- ALCF/test_blendable_dataset.py | 79 ++++++++++++++++++------------ megatron/data/blendable_dataset.py | 2 +- megatron/data/gpt_dataset.py | 43 +++++++++------- 3 files changed, 75 insertions(+), 49 deletions(-) mode change 100644 => 100755 megatron/data/blendable_dataset.py mode change 100644 => 100755 megatron/data/gpt_dataset.py diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 2c2ad84c22d..86857c819a1 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -6,25 +6,32 @@ from megatron.arguments import parse_args from megatron.initialize import initialize_megatron from megatron.data.data_samplers import build_pretraining_data_loader - +import time from megatron.core import mpu comm = MPI.COMM_WORLD + +import datetime +def print_rank_0(msg): + if comm.rank==0: + print(f" [INFO][{datetime.datetime.now()}] {msg}", flush=True) + + initialize_megatron(allow_no_cuda=True) args = get_args() data_file_list = args.data_file_list -if comm.rank==0: - print(f"Reading data from {args.data_file_list}") +print_rank_0(f"Reading data from {args.data_file_list}") files = [] weights = [] flist = [] with open(data_file_list, 'r') as fin: for f in fin.readlines(): - w, fname = f.split() + w, fname, c = f.split() weights.append(float(w)) flist.append(fname) files.append(float(w)) files.append(fname) + files.append(c) splits_string="100,0,0" weights = np.array(weights) @@ -32,10 +39,9 @@ num_samples = args.global_batch_size*args.train_iters num_datasets = len(weights) -if comm.rank==0: - print(f"Number of datasets: {num_datasets}") - print(f"Global batch size: {args.global_batch_size}") - print(f"Training iterations: {args.train_iters}") +print_rank_0(f"Number of datasets: {num_datasets}") +print_rank_0(f"Global batch size: {args.global_batch_size}") +print_rank_0(f"Training iterations: {args.train_iters}") train_valid_test_num_samples = [num_samples, 0, 0] seed=args.seed data_impl = args.data_impl @@ -44,40 +50,51 @@ splits_string = "1,0,0" # Build datasets +start_build_dataset = time.time() +print_rank_0(f"Starting to build the blendable dataset") train_ds, valid_ds, test_ds = build_train_valid_test_datasets(files, data_impl, splits_string, train_valid_test_num_samples, seq_length, seed, skip_warmup, data_cache_path=args.data_cache_path) -dataset_idx = [train_ds.dataset_index[i] for i in range(num_samples)] -ratio_select=np.zeros(num_datasets) -#for i in range(num_datasets): -# ratio_select[i] = np.sum([i==d for d in dataset_idx])/num_samples -if comm.rank ==0: - print(f"Total number of samples: {len(train_ds)}") - print(f"Weights set: {weights[:min(8, num_datasets)]}") - -for e in range(min(100, args.train_iters)): - ratio_select=np.zeros(num_datasets) - for i in range(num_datasets): - ratio_select[i] = np.sum([i==d for d in dataset_idx[e*args.global_batch_size:(e+1)*args.global_batch_size]])/args.global_batch_size - if comm.rank==0: - print(f"iter-{e}: {ratio_select[:min(8, num_datasets)]}") - -print("First 10 samples") -for i in range(10): - if comm.rank==0: - print(f"Sample: {i} \t dataset_idx: {train_ds.dataset_index[i]}, sample_idx: {train_ds.dataset_sample_index[i]}") +end_build_dataset = time.time() +print_rank_0(f"Finished building the blendable dataset in {end_build_dataset - start_build_dataset} second") +print_rank_0(f"Total number of samples: {len(train_ds)}") +print_rank_0(f"Weights set: {weights[:min(8, num_datasets)]}") -#### Build data loaders +start_build_dataloader = time.time() +print_rank_0(f"Starting to build the data loader") rank_in_parallel_group = mpu.get_sequence_parallel_rank() train_dataloader = build_pretraining_data_loader( train_ds, args.consumed_train_samples) valid_dataloader = build_pretraining_data_loader( valid_ds, args.consumed_valid_samples) test_dataloader = build_pretraining_data_loader(test_ds, 0) +end_build_dataloader = time.time() +print_rank_0(f"Finished building the data loader in {end_build_dataloader - start_build_dataloader} second") - -# Run through all the batches in data loader +print_rank_0(f"Starting loading the data") +start_loading_time = time.time() +NUM_ITEMS=100 +n=0 for i in train_dataloader: - print(i) + print(f"[{comm.rank}] DATA {i}") + n+=1 + if (n%NUM_ITEMS==0): + print_rank_0(f"Loaded {n} batches in {time.time() - start_loading_time}") + if n>=5000: + break +end_loading_time = time.time() +print_rank_0(f"Finished loading the data ({n} batches) in {end_loading_time - start_loading_time}") + + + +dataset_idx = [train_ds.dataset_index[i] for i in range(num_samples)] +ratio_select=np.zeros(num_datasets) + +for e in range(min(100, args.train_iters)): + ratio_select=np.zeros(num_datasets) + for i in range(num_datasets): + ratio_select[i] = np.sum([i==d for d in dataset_idx[e*args.global_batch_size:(e+1)*args.global_batch_size]])/args.global_batch_size + print_rank_0(f"iter-{e}: {ratio_select[:min(8, num_datasets)]}") + diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py old mode 100644 new mode 100755 index 32b2005516b..8478ef2e038 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -78,7 +78,7 @@ def _build_indices(): print('write access to.') cache_success = False - + counts = get_accelerator().LongTensor([cache_success]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py old mode 100644 new mode 100755 index 98089dfdcb7..6324dda755c --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -42,7 +42,7 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, output = get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_num_samples) prefixes, corpuses, weights, datasets_train_valid_test_num_samples = output - corpus_list = set(corpuses) + corpus_list = sorted(set(corpuses)) train_num_samples, valid_num_samples, test_num_samples = map( sum, zip(*datasets_train_valid_test_num_samples) @@ -59,7 +59,13 @@ def __init__(self, prefix, corpus, data_impl, splits_string, self.prefix = prefix self.data_impl = data_impl self.splits_string = splits_string - self.num_samples = num_samples + if name == 'train': + self.num_samples = num_samples[0] + elif name == 'valid': + self.num_samples = num_samples[1] + else: + self.num_samples = num_samples[2] + self.num_samples_train_valid_test = num_samples self.seq_length = seq_length self.seed = seed self.skip_warmup = skip_warmup @@ -67,40 +73,40 @@ def __init__(self, prefix, corpus, data_impl, splits_string, self.data_cache_path = data_cache_path self.dataset = None self.name = name - self.desc = prefix + f"{num_samples}" + f"{seq_length}" + f"{seed}" + self.desc = prefix + f"{self.num_samples}" + f"{seq_length}" + f"{seed}" self.build = False self.corpus = corpus def Build(self): self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, - self.num_samples, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, + self.num_samples_train_valid_test, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, data_cache_path=self.data_cache_path) self.build = True return self.dataset - def BuildConcatDatasets(Dataset): + class BuildConcatDataset(torch.utils.data.Dataset): def __init__(self, dataset_builders): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) self.indices=[] - self.desc="ConcatDatasets" + self.desc="ConcatDataset:" for i in range(self.num_datasets): - self.desc += dataset_builders[i].prefix + dataset_builders[i].corpus + self.desc += dataset_builders[i].prefix + "," for j in range(dataset_builders[i].num_samples): self.indices.append((i, j)) - self.desc += f"{self.num_samples}" + f"{seq_length}" + f"{seed}" + self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): return self.num_samples def __getitem__(self, idx): i, j = self.indices[idx] if self.dataset_builders[i].build: - return self.dataset_builders.dataset[j] + return self.dataset_builders[i].dataset[j] else: return self.dataset_builders[i].Build()[j] # Predetermine whether need to build the specific dataset or not. - + start_time = time.time() print_rank_0(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] @@ -110,6 +116,8 @@ def __getitem__(self, idx): test_datasets = [] # Build individual datasets. def build_corpus_datasets(dataset_type='train'): + start_time = time.time() + print_rank_0(" >>> Building {dataset_type} corpus datasets ...") datasets = [] corpus_builders = {} corpus_weights = {} @@ -122,12 +130,12 @@ def build_corpus_datasets(dataset_type='train'): return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): dataset_builders[i].Build() - for i, d in zip(range(len(weights)), train_datasets_builders): + for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] for c in corpus_list: - datasets.append(BuildConcatDatasets(train_corpus_builders[c])) - + datasets.append(BuildConcatDataset(corpus_builders[c])) + print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") return datasets, [corpus_weights[c] for c in corpus_list] if a > 0: @@ -135,14 +143,14 @@ def build_corpus_datasets(dataset_type='train'): if b > 0: valid_datasets, valid_weights = build_corpus_datasets('valid') + if c > 0: test_datasets, test_weights = build_corpus_datasets('test') # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them torch.distributed.barrier() - - print_rank_0(" >>> Finished building datasets in distributed way ... ") + print_rank_0(f" >>> Finished building datasets in distributed way in {time.time() - start_time} seconds") # Blend. blending_train_dataset = None @@ -157,7 +165,8 @@ def build_corpus_datasets(dataset_type='train'): if test_datasets: blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, data_cache_path=data_cache_path) - + torch.distributed.barrier() + print_rank_0(f" >>> Finished building BlendableDataset") return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) @@ -543,7 +552,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the # ranks will first create the document files, and then read it. # There will not be contension effects going on either - print_flush(f" > WARNING: could not find index map files, building ") + print_flush(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. From 7c80c2cee064831afebd9a3fc8e060d601361512 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 31 May 2024 00:10:55 -0500 Subject: [PATCH 168/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 67 ++++++++++++++++++++++++++----------------------- 1 file changed, 35 insertions(+), 32 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 24e048c06d5..8dc519d501d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -2,6 +2,12 @@ # # set -euxo pipefail +#################################################################### +# `WORKING_DIR`: +# 1. if `${PBS_O_WORKDIR}` has nonzero value, use this +# 2. else, if `${SLURM_SUBMIT_DIR}` has nonzero value, use that +# 3. else, use `$(pwd)` +#################################################################### if [[ -n "${PBS_O_WORKDIR}" ]]; then WORKING_DIR="${PBS_O_WORKDIR}" elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then @@ -13,7 +19,7 @@ else fi export WORKING_DIR="${WORKING_DIR}" -printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} +printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" ############################################################################## # `setup`: All-in-one helper function. @@ -40,14 +46,13 @@ printf "Using WORKING_DIR: %s\n" ${WORKING_DIR} # 13. Setup run command to be executed. ############################################################################## setup() { - python3 -m ezpz.jobs && source "./.jobenv" # echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" # source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 # source "${ezloc}/src/ezpz/bin/getjobenv" || exit get_machine || exit # 01. Identify machine we're on setEnv || exit # 02. Load `conda` environment # saveDSenv || exit - ezpz || exit # 03. Determine WORLD_SIZE, etc. from `PBS_*` vars + setup_ezpz || exit # 03. Determine WORLD_SIZE, etc. from `PBS_*` vars setParams || exit # 04. Set command line arguments to pass to `"${EXEC}"` buildDSconfig || exit # 05. Create `deepspeed_config.json` from runtime params from ^ setOutput || exit # 06. Specify output directory for {logs, checkpoints, etc.} @@ -103,7 +108,7 @@ setup_run_cmd() { if [[ "${SP}" -ge 2 ]]; then export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" fi - run_cmd=" + export run_cmd=" ${LAUNCHER} \ --${DTYPE} \ ${DEFAULTS} \ @@ -335,7 +340,7 @@ set_lr_args() { ######################################################################### get_batch_size_on_polaris() { if [[ $(hostname) == x3* ]]; then - local nhosts=$(wc -l < "${PBS_NODEFILE}") + nhosts=$(wc -l < "${PBS_NODEFILE}") if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then mbs=1 elif [[ "${nhosts}" -ge 3 ]]; then @@ -513,19 +518,21 @@ make_ds_hostfile() { ezpz_savejobenv() { local outfile="${WORKING_DIR}/savejobenv" - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/jobs-cleanup/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" } ezpz_getjobenv() { local outfile="${WORKING_DIR}/getjobenv" - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/jobs-cleanup/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" } # +---------------------------------------+ # | 1. Git clone ezpz (if not found) | # | 2. Install ezpz (if not installed) | # +---------------------------------------+ -ezpz() { +setup_ezpz() { + [ -n "${PBS_NODEFILE}" ] && ezpz_savejobenv || ezpz_getjobenv + python3 -m ezpz.jobs && source "./.jobenv" if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" @@ -536,29 +543,25 @@ ezpz() { if [[ -n "${ezloc}" ]]; then ezpz_savejobenv python3 -m ezpz.jobs && source "./.jobenv" - # echo "ezpz detected. Sourcing ${ezloc}/bin/savejobenv" - # source "${ezloc}/src/ezpz/bin/savejobenv" > /dev/null 2>&1 - # source "${ezloc}/src/ezpz/bin/getjobenv" || exit make_ds_hostfile || exit else echo "No ezpz detected. Attempting to install with $(which python3)" python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --require-virtualenv fi echo "Done with ezpz." - # echo "Done with clone. Now, checking if ezpz is installed..." - # if python3 -c 'import ezpz; print(ezpz.__file__)' 2> '/dev/null'; then - # if [[ $(python3 -c "import sys; any(['ezpz' in s for s in sys.path])") 2> '/dev/null' ]]; then - # echo "Has ezpz installed. Nothing to do." - # else - # echo "Does not have ezpz installed. Installing..." - # echo "Using $(which python3) to install ezpz:" - # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --verbose --require-virtualenv # > ezpz-install.log 2>&1 - # fi - # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --verbose --require-virtualenv - # # echo "Done with ezpz." - # source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv > /dev/null 2>&1 #> /tmp/savejobenv.log 2>&1 || exit - # source ${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv || exit - # make_ds_hostfile || exit +} + +####################################################################### +# `ezpz_test`: Run simple test to make sure all nodes in working order +####################################################################### +ezpz_test() { + printf "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" + # [ -n "${PBS_O_WORKIR}" ] && ezpz_savejobenv || ezpz_getjobenv + # python3 -Wignore -m ezpz.jobs && source "${PBS_O_WORKDIR}/.jobenv" + printf "[$(printBlue 'ezpz:test_dist')] Running test: ${eztest}\n" + eztest="TRAIN_ITERS=50 ${LAUNCH_CMD} python3 -Wignore -m ezpz.test_dist" + eval "${eztest}" + printf "[$(printBlue 'ezpz:test_dist')] Done with test!\n" } # +------------------------------------------------------------------------+ @@ -908,14 +911,14 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] - # shellcheck: source ./data-lists/sunspot/books.txt + # shellcheck source=./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt" elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] if [[ "${PBS_O_HOST}" == sirius* ]]; then # -------------------[SIRIUS] - # shellcheck: source ./data-lists/sirius/books.txt + # shellcheck source=./data-lists/sirius/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma_v1_7_file_list.txt" elif [[ "${PBS_O_HOST}" == polaris* ]]; then # ---------------[POLARIS] - # shellcheck: source ./data-lists/polaris/books.txt + # shellcheck source=./data-lists/polaris/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] @@ -960,7 +963,7 @@ generateDSconfig() { for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ "$PP" "$DTYPE" do - if [ -z $v ]; then + if [ -z "$v" ]; then echo "Please export required envs before execute $0" exit 1 fi @@ -1040,7 +1043,7 @@ generateDSconfig() { else dtype="\"communication_data_type\": \"fp32\"," fi - if [ $ZERO_STAGE == 3 ]; then + if [ "$ZERO_STAGE" == 3 ]; then zero="\ \"zero_optimization\": { \"stage\": 3, @@ -1082,7 +1085,7 @@ generateDSconfig() { }," fi # elif [[ $ZERO_STAGE == 1 ]]; then - if [[ $PP > 1 ]]; then + if [[ "${PP}" -gt 1 ]]; then extra="\ \"data_types\": { \"grad_accum_dtype\": \"fp32\" @@ -1107,7 +1110,7 @@ generateDSconfig() { echo 'Please add the correct config set!!!' fi # flops_profiler must at the end because no ',' is allowed at the end -cat < $1 +cat < "$1" { $common $zero From 93db2a966f4077a3d8df867ec4b66bcea2f6291d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 31 May 2024 01:05:36 -0500 Subject: [PATCH 169/590] optimized loading blendable dataset meta data, by loading and broadcasting --- megatron/data/blendable_dataset.py | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 8478ef2e038..504f76c813e 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -52,7 +52,8 @@ def _build_indices(): desc += f"Weights: {weights}\n" desc += f"Size: {size}\n" self.desc = desc - + self.dataset_index = [] + self.dataset_sample_index = [] if data_cache_path: desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() desc_path = os.path.join(data_cache_path, desc_hash + ".dsc") @@ -77,8 +78,8 @@ def _build_indices(): print('ensure you have write access to this directory or specify one that you do have') print('write access to.') cache_success = False - - + self.dataset_index = dataset_index + self.dataset_sample_index = dataset_sample_index counts = get_accelerator().LongTensor([cache_success]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) @@ -89,14 +90,16 @@ def _build_indices(): print_rank_0("Data index creation unsuccessful, exiting.") exit() - # Load on all ranks. - print_rank_0(f'> loading blendable dataset index: {index_path}') - self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') - assert self.dataset_index.size == self.size + if cache_hit: + print_rank_0(f'> loading blendable dataset index: {index_path}') + self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_index.size == self.size - print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') - self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') - assert self.dataset_sample_index.size == self.size + print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') + self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_sample_index.size == self.size + torch.distributed.broadcast_object_list(self.dataset_sample_index, src=0) + torch.distributed.broadcast_object_list(self.dataset_index, src=0) else: self.dataset_index, self.dataset_sample_index = _build_indices() From 89f2a95ff50de1b8caf9dc3e4b03988c45e7bbdb Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 31 May 2024 01:09:33 -0500 Subject: [PATCH 170/590] added broadcast --- megatron/data/blendable_dataset.py | 1 + 1 file changed, 1 insertion(+) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 504f76c813e..532059edb4c 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -91,6 +91,7 @@ def _build_indices(): exit() if cache_hit: + print_rank_0(f"> index map files exists already, rank 0 will read them and broadcast to all the rank") print_rank_0(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_index.size == self.size From 96cb1e5bb65166c708088c53f7ffec9b92a8a834 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 31 May 2024 09:12:55 +0000 Subject: [PATCH 171/590] fixed overflow issue --- megatron/data/blendable_dataset.py | 29 +++++++++++++++-------------- megatron/data/gpt_dataset.py | 18 ++++++++++++++++-- 2 files changed, 31 insertions(+), 16 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 532059edb4c..f8b3c0429c1 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -12,7 +12,7 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0, print_flush from megatron.core import mpu - +from mpi4py import MPI class BlendableDataset(torch.utils.data.Dataset): @@ -52,8 +52,8 @@ def _build_indices(): desc += f"Weights: {weights}\n" desc += f"Size: {size}\n" self.desc = desc - self.dataset_index = [] - self.dataset_sample_index = [] + self.dataset_index = np.zeros(self.size, dtype=np.int64) + self.dataset_sample_index = np.zeros(self.size, dtype=np.int64) if data_cache_path: desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() desc_path = os.path.join(data_cache_path, desc_hash + ".dsc") @@ -66,12 +66,15 @@ def _build_indices(): ' dataset, building indices on rank 0 ...', flush=True) dataset_index, dataset_sample_index = _build_indices() try: + print_rank_0(" > saving index map files") + start_time = time.time() os.makedirs(os.path.dirname(index_path), exist_ok=True) with open(desc_path, 'wt') as fd: fd.write(desc) np.save(index_path, dataset_index, allow_pickle=True) np.save(sample_index_path, dataset_sample_index, allow_pickle=True) + print_rank_0(f" > finished saving index map files in {time.time() - start_time} seconds") except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_path})') print('or a file in it. This is set with the --data-cache-path argument. Please') @@ -90,17 +93,15 @@ def _build_indices(): print_rank_0("Data index creation unsuccessful, exiting.") exit() - if cache_hit: - print_rank_0(f"> index map files exists already, rank 0 will read them and broadcast to all the rank") - print_rank_0(f'> loading blendable dataset index: {index_path}') - self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') - assert self.dataset_index.size == self.size - - print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') - self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') - assert self.dataset_sample_index.size == self.size - torch.distributed.broadcast_object_list(self.dataset_sample_index, src=0) - torch.distributed.broadcast_object_list(self.dataset_index, src=0) + start_time = time.time() + print_rank_0(f'> loading blendable dataset index: {index_path}') + self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_index.size == self.size + print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') + self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_sample_index.size == self.size + torch.distributed.barrier() + print_rank_0(f'> finished loading in {time.time() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 6324dda755c..b65fe1cf8b1 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -94,11 +94,17 @@ def __init__(self, dataset_builders): self.desc += dataset_builders[i].prefix + "," for j in range(dataset_builders[i].num_samples): self.indices.append((i, j)) + assert(len(self.indices)==self.num_samples) self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): return self.num_samples def __getitem__(self, idx): - i, j = self.indices[idx] + if idx >= self.num_samples: + print_flush(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") + id = np.random.randint(self.num_samples) + else: + id = idx + i, j = self.indices[id] if self.dataset_builders[i].build: return self.dataset_builders[i].dataset[j] else: @@ -133,10 +139,18 @@ def build_corpus_datasets(dataset_type='train'): for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] + total = 0 + print_rank_0(" > number of samples for each corpus ") + corpus_weights_achieved={} for c in corpus_list: datasets.append(BuildConcatDataset(corpus_builders[c])) + total += datasets[-1].num_samples + corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples + print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") + + print_rank_0(f" > total number of samples: {total}") print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") - return datasets, [corpus_weights[c] for c in corpus_list] + return datasets, [corpus_weights_achieved[c] for c in corpus_list] if a > 0: train_datasets, train_weights = build_corpus_datasets('train') From cb2f1dc731d74680f9a7ff62a91727bea7607ec3 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 31 May 2024 09:15:30 +0000 Subject: [PATCH 172/590] removed unnecessary mpi4py --- megatron/data/blendable_dataset.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index f8b3c0429c1..bcc13b021c3 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -12,7 +12,7 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0, print_flush from megatron.core import mpu -from mpi4py import MPI + class BlendableDataset(torch.utils.data.Dataset): From 0dea6aaa715cd1b391cc6f12fdb6d57f2cb341c7 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 3 Jun 2024 22:35:16 -0500 Subject: [PATCH 173/590] Update dataset_utils.py fixed a minor bug --- megatron/data/dataset_utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/data/dataset_utils.py b/megatron/data/dataset_utils.py index b73a6cf85fd..61dd3909e4d 100644 --- a/megatron/data/dataset_utils.py +++ b/megatron/data/dataset_utils.py @@ -49,7 +49,7 @@ def get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_ data_new_prefix += [data_prefix[3*i], data_prefix[3*i+1]] corpuses.append(data_prefix[3*i+2]) prefixes, weights, datasets_train_valid_test_num_samples = \ - get_datasets_weights_and_num_samples(data_prefix, + get_datasets_weights_and_num_samples(data_new_prefix, train_valid_test_num_samples) return prefixes, corpuses, weights, datasets_train_valid_test_num_samples From 5d26dfe2fc193e8940dcd4a46a42082ad11f1ad9 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 4 Jun 2024 22:49:51 -0500 Subject: [PATCH 174/590] fixed a minor bug --- megatron/utils.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 383be237583..cd32f9a946f 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -38,7 +38,7 @@ else: from functools import wraps class Profile: - def __init__(type="PROFILER"): + def __init__(self, type="PROFILER"): self.type = type def log(self, func): @wraps(func) @@ -49,7 +49,9 @@ class dlio_logger: def __init__(self,): self.type = None def initialize_log(self, logfile=None, data_dir=None, process_id=-1): - return + return + def iter(self, a): + return a PerfTrace = dlio_logger() From 3dc424f6dda0c0ec87b0bc9b0d61b3a48f1090f3 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 4 Jun 2024 23:10:43 -0500 Subject: [PATCH 175/590] remove unnecessary barrier --- megatron/data/gpt_dataset.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 8ecc115399b..b1914db5147 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -175,7 +175,8 @@ def build_corpus_datasets(dataset_type='train'): # and the metadata were written to the cache folder before other ranks touch them torch.distributed.barrier() print_rank_0(f" >>> Finished building datasets in distributed way in {time.time() - start_time} seconds") - + print_rank_0(f" >>> Starting to build BlendableDataset") + start_time = time.time() # Blend. blending_train_dataset = None if train_datasets: @@ -189,8 +190,8 @@ def build_corpus_datasets(dataset_type='train'): if test_datasets: blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, data_cache_path=data_cache_path) - torch.distributed.barrier() - print_rank_0(f" >>> Finished building BlendableDataset") + end_time = time.time() + print_rank_0(f" >>> Finished building BlendableDataset in {end_time - start_time} seconds") return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) From 60fc482f6d5013918faae1f2033a41b20d630328 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 4 Jun 2024 23:31:17 -0500 Subject: [PATCH 176/590] added pfw tracing for test_blendable_dataset --- ALCF/test_blendable_dataset.py | 12 +++++++++++- megatron/training.py | 5 +++-- 2 files changed, 14 insertions(+), 3 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 86857c819a1..2b189dd8458 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -1,5 +1,6 @@ #!/usr/bin/env python from mpi4py import MPI +import os from megatron.data.gpt_dataset import build_train_valid_test_datasets import numpy as np from megatron.global_vars import set_args, set_global_variables, get_args @@ -9,7 +10,9 @@ import time from megatron.core import mpu comm = MPI.COMM_WORLD +from megatron.utils import PerfTrace, Profile +dlp = Profile("TEST_BLENDABLEDATASET") import datetime def print_rank_0(msg): if comm.rank==0: @@ -18,6 +21,13 @@ def print_rank_0(msg): initialize_megatron(allow_no_cuda=True) args = get_args() +if os.getenv('DLIO_PROFILER_DATASET_DIR') is not None: + extra_trace_path = os.environ['DLIO_PROFILER_DATASET_DIR'] +else: + extra_trace_path='' + +os.makedirs(args.trace_dir, exist_ok=True) +PerfTrace.initialize_log(f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=comm.rank) data_file_list = args.data_file_list print_rank_0(f"Reading data from {args.data_file_list}") @@ -77,7 +87,7 @@ def print_rank_0(msg): start_loading_time = time.time() NUM_ITEMS=100 n=0 -for i in train_dataloader: +for i in dlp.iter(train_dataloader): print(f"[{comm.rank}] DATA {i}") n+=1 if (n%NUM_ITEMS==0): diff --git a/megatron/training.py b/megatron/training.py index a89486d6b12..9968ad1a34a 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -190,10 +190,11 @@ def pretrain( print_datetime('after megatron is initialized') args = get_args() - if os.getenv('DLIO_PROFILER_DATA_DIR') is not None: - extra_trace_path = os.environ['DLIO_PROFILER_DATA_DIR'] + if os.getenv('DLIO_PROFILER_DATASET_DIR') is not None: + extra_trace_path = os.environ['DLIO_PROFILER_DATASET_DIR'] else: extra_trace_path='' + makedirs(args.trace_dir, exist_ok=True) PerfTrace.initialize_log(f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=ez.get_rank()) timers = get_timers() assert args is not None From b1f17d598d9bd6541879c13c2e016c4b6afe2341 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 4 Jun 2024 23:33:49 -0500 Subject: [PATCH 177/590] fixed bug --- ALCF/test_blendable_dataset.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 2b189dd8458..13168a00b89 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -27,7 +27,7 @@ def print_rank_0(msg): extra_trace_path='' os.makedirs(args.trace_dir, exist_ok=True) -PerfTrace.initialize_log(f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=comm.rank) +PerfTrace.initialize_log(f"{args.trace_dir}/trace-{comm.rank}-of-{comm.size}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=comm.rank) data_file_list = args.data_file_list print_rank_0(f"Reading data from {args.data_file_list}") From 10a3737dc4800df51cca3b358d4a5753cd882904 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 4 Jun 2024 23:55:18 -0500 Subject: [PATCH 178/590] added more loging --- megatron/data/dataset_utils.py | 2 +- megatron/data/gpt_dataset.py | 10 ++++++---- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/megatron/data/dataset_utils.py b/megatron/data/dataset_utils.py index b73a6cf85fd..61dd3909e4d 100644 --- a/megatron/data/dataset_utils.py +++ b/megatron/data/dataset_utils.py @@ -49,7 +49,7 @@ def get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_ data_new_prefix += [data_prefix[3*i], data_prefix[3*i+1]] corpuses.append(data_prefix[3*i+2]) prefixes, weights, datasets_train_valid_test_num_samples = \ - get_datasets_weights_and_num_samples(data_prefix, + get_datasets_weights_and_num_samples(data_new_prefix, train_valid_test_num_samples) return prefixes, corpuses, weights, datasets_train_valid_test_num_samples diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index b65fe1cf8b1..26c6683b71c 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -163,10 +163,12 @@ def build_corpus_datasets(dataset_type='train'): # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them + print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") torch.distributed.barrier() - print_rank_0(f" >>> Finished building datasets in distributed way in {time.time() - start_time} seconds") - + print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") + print_rank_0(f" >>> Starting to build BlendableDataset") # Blend. + start_time = time.time() blending_train_dataset = None if train_datasets: blending_train_dataset = BlendableDataset(train_datasets, train_weights, train_num_samples, @@ -179,8 +181,8 @@ def build_corpus_datasets(dataset_type='train'): if test_datasets: blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, data_cache_path=data_cache_path) - torch.distributed.barrier() - print_rank_0(f" >>> Finished building BlendableDataset") + end_time = time.time() + print_rank_0(f" >>> Finished building BlendableDataset in {end_time - start_time} seconds") return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) From bc28f843bf74e1e33be4b8e840232e0c9d2124fe Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 00:49:08 -0500 Subject: [PATCH 179/590] removed allreduce calls that are not needed --- megatron/data/blendable_dataset.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index bcc13b021c3..524b72737b1 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -83,6 +83,7 @@ def _build_indices(): cache_success = False self.dataset_index = dataset_index self.dataset_sample_index = dataset_sample_index + ''' I don't think the following piece of code is needed any more. counts = get_accelerator().LongTensor([cache_success]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) @@ -92,7 +93,8 @@ def _build_indices(): torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): print_rank_0("Data index creation unsuccessful, exiting.") exit() - + ''' + torch.distributed.barrier() start_time = time.time() print_rank_0(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') @@ -100,7 +102,6 @@ def _build_indices(): print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size - torch.distributed.barrier() print_rank_0(f'> finished loading in {time.time() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() From 6eb21b73f2401cca98c7548acfc8421989745994 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 00:49:53 -0500 Subject: [PATCH 180/590] removed allreduce call that are not needed any more --- ALCF/test_blendable_dataset.py | 6 ++++-- megatron/data/blendable_dataset.py | 14 ++++++++------ megatron/data/gpt_dataset.py | 11 +++++++---- 3 files changed, 19 insertions(+), 12 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 13168a00b89..5303f3fc5b1 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -12,7 +12,7 @@ comm = MPI.COMM_WORLD from megatron.utils import PerfTrace, Profile -dlp = Profile("TEST_BLENDABLEDATASET") + import datetime def print_rank_0(msg): if comm.rank==0: @@ -25,9 +25,11 @@ def print_rank_0(msg): extra_trace_path = os.environ['DLIO_PROFILER_DATASET_DIR'] else: extra_trace_path='' +PerfTrace.initialize_log(f"{args.trace_dir}/trace-{comm.rank}-of-{comm.size}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=comm.rank) +dlp = Profile("TEST_BLENDABLEDATASET") os.makedirs(args.trace_dir, exist_ok=True) -PerfTrace.initialize_log(f"{args.trace_dir}/trace-{comm.rank}-of-{comm.size}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=comm.rank) + data_file_list = args.data_file_list print_rank_0(f"Reading data from {args.data_file_list}") diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 40194692f75..a9fbe7b6206 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -76,7 +76,7 @@ def _build_indices(): np.save(index_path, dataset_index, allow_pickle=True) np.save(sample_index_path, dataset_sample_index, allow_pickle=True) - print_rank_0(f" > finished saving index map files in {time.time() - start_time} seconds") + print_rank_0(f" > finished saving index map files in {time.time() - start_time} seconds") except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_path})') print('or a file in it. This is set with the --data-cache-path argument. Please') @@ -85,16 +85,18 @@ def _build_indices(): cache_success = False self.dataset_index = dataset_index self.dataset_sample_index = dataset_sample_index + ''' I don't think the following piece of code is necessary any more; I commented them out now counts = get_accelerator().LongTensor([cache_success]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) if counts[0].item() != ( - torch.distributed.get_world_size() // - torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // - torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): + torch.distributed.get_world_size() // + torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // + torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): print_rank_0("Data index creation unsuccessful, exiting.") exit() - + ''' + torch.distributed.barrier() start_time = time.time() print_rank_0(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') @@ -102,7 +104,7 @@ def _build_indices(): print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size - torch.distributed.barrier() + print_rank_0(f'> finished loading in {time.time() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index b1914db5147..28c78ab3fd3 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -94,13 +94,15 @@ def __init__(self, dataset_builders): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) - self.indices=[] + self.indices=np.zeros((self.num_samples, 2), dtype=np.uint64) self.desc="ConcatDataset:" + m = 0 for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," for j in range(dataset_builders[i].num_samples): - self.indices.append((i, j)) - assert(len(self.indices)==self.num_samples) + self.indices[m] = [i, j] + m+=1 + assert(m==self.num_samples) self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): return self.num_samples @@ -173,8 +175,9 @@ def build_corpus_datasets(dataset_type='train'): # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them + print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") torch.distributed.barrier() - print_rank_0(f" >>> Finished building datasets in distributed way in {time.time() - start_time} seconds") + print_rank_0(f" >>> Finished (all ranks) building datasets in distributed way in {time.time() - start_time} seconds") print_rank_0(f" >>> Starting to build BlendableDataset") start_time = time.time() # Blend. From 20a243082f38c68b347602da342cb4802e9e5568 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 09:26:35 -0500 Subject: [PATCH 181/590] fixed a bug --- ALCF/test_blendable_dataset.py | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 86857c819a1..94adb686a1b 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -1,4 +1,6 @@ #!/usr/bin/env python +import time +start_time = time.time() from mpi4py import MPI from megatron.data.gpt_dataset import build_train_valid_test_datasets import numpy as np @@ -6,17 +8,22 @@ from megatron.arguments import parse_args from megatron.initialize import initialize_megatron from megatron.data.data_samplers import build_pretraining_data_loader -import time + +import torch from megatron.core import mpu + + comm = MPI.COMM_WORLD import datetime def print_rank_0(msg): if comm.rank==0: print(f" [INFO][{datetime.datetime.now()}] {msg}", flush=True) - - +end_time = time.time() +print_rank_0(f"Loaded python modules in {end_time - start_time} seconds") initialize_megatron(allow_no_cuda=True) +torch.distributed.barrier() +print_rank_0(f"Barrier synchonization time: {time.time() - end_time} seconds") args = get_args() data_file_list = args.data_file_list @@ -51,6 +58,7 @@ def print_rank_0(msg): # Build datasets start_build_dataset = time.time() + print_rank_0(f"Starting to build the blendable dataset") train_ds, valid_ds, test_ds = build_train_valid_test_datasets(files, data_impl, splits_string, train_valid_test_num_samples, From b4c832eb086f98ff14c6e3526946cc6cca032a4d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 12:27:43 -0500 Subject: [PATCH 182/590] added more logging for index_dataset --- megatron/data/indexed_dataset.py | 5 ++++- megatron/training.py | 4 ++-- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index 7810a6496bb..1eb9b7842bb 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -278,7 +278,7 @@ class IndexedDatasetBuilder(object): np.float32: 4, np.float64: 8, } - + @dlp.log def __init__(self, out_file, dtype=np.int32): self.out_file = open(out_file, 'wb') self.dtype = dtype @@ -425,6 +425,7 @@ def __exit__(self, exc_type, exc_val, exc_tb): return _Writer() + @dlp.log def __init__(self, path, skip_warmup=False): with open(path, 'rb') as stream: magic_test = stream.read(9) @@ -500,6 +501,7 @@ def __getstate__(self): def __setstate__(self, state): self._do_init(state, skip_warmup=True) + @dlp.log def _do_init(self, path, skip_warmup): self._path = path self._index = self.Index(index_file_path(self._path), skip_warmup) @@ -544,6 +546,7 @@ def __getitem__(self, idx): else: raise TypeError("Unexpected type received for idx: {}".format(type(idx))) + @dlp.log def get(self, idx, offset=0, length=None): """ Retrieves a single item from the dataset with the option to only return a portion of the item. diff --git a/megatron/training.py b/megatron/training.py index 9968ad1a34a..b89cc559324 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -4,7 +4,7 @@ from datetime import datetime import math -import sys +import sys, os import time import json # The earliest we can measure the start time. @@ -194,7 +194,7 @@ def pretrain( extra_trace_path = os.environ['DLIO_PROFILER_DATASET_DIR'] else: extra_trace_path='' - makedirs(args.trace_dir, exist_ok=True) + os.makedirs(args.trace_dir, exist_ok=True) PerfTrace.initialize_log(f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=ez.get_rank()) timers = get_timers() assert args is not None From 1719b0e1014fd7b258728eeab8b3afb885e1cc00 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 17:13:42 -0500 Subject: [PATCH 183/590] added new log --- ALCF/test_blendable_dataset.py | 25 ++++++++++--------------- megatron/data/gpt_dataset.py | 12 +++++++----- 2 files changed, 17 insertions(+), 20 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 227bf4ebaa9..7f52169a3a5 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -95,26 +95,21 @@ def print_rank_0(msg): print_rank_0(f"Starting loading the data") start_loading_time = time.time() -NUM_ITEMS=100 +NUM_ITEMS=1 +SLEEP_TIME=10.0 +@dlp.log +def compute(ct): + time.sleep(ct) n=0 +start_time = time.time() for i in dlp.iter(train_dataloader): + compute(SLEEP_TIME) print(f"[{comm.rank}] DATA {i}") n+=1 if (n%NUM_ITEMS==0): - print_rank_0(f"Loaded {n} batches in {time.time() - start_loading_time}") - if n>=5000: + print_rank_0(f"Proccessed {n}th-batch in {time.time() - start_time}") + if n>=100: break + start_time = time.time() end_loading_time = time.time() print_rank_0(f"Finished loading the data ({n} batches) in {end_loading_time - start_loading_time}") - - - -dataset_idx = [train_ds.dataset_index[i] for i in range(num_samples)] -ratio_select=np.zeros(num_datasets) - -for e in range(min(100, args.train_iters)): - ratio_select=np.zeros(num_datasets) - for i in range(num_datasets): - ratio_select[i] = np.sum([i==d for d in dataset_idx[e*args.global_batch_size:(e+1)*args.global_batch_size]])/args.global_batch_size - print_rank_0(f"iter-{e}: {ratio_select[:min(8, num_datasets)]}") - diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index f93c5180ffa..e758e6cb213 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -101,9 +101,8 @@ def __init__(self, dataset_builders): m = 0 for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," - for j in range(dataset_builders[i].num_samples): - self.indices[m] = [i, j] - m+=1 + self.indices[m:dataset_builders[i].num_samples+m] = [[i, j] for j in range(dataset_builders[i].num_samples)] + m+=dataset_builders[i].num_samples assert(m==self.num_samples) self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): @@ -150,6 +149,8 @@ def build_corpus_datasets(dataset_type='train'): return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): dataset_builders[i].Build() + print_rank_0(f" >>> Finished building individual datasets in time.time() - start_time() seconds") + start_concating_time = time.time() for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] @@ -161,8 +162,9 @@ def build_corpus_datasets(dataset_type='train'): total += datasets[-1].num_samples corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") - - print_rank_0(f" > total number of samples: {total}") + + print_rank_0(f" > total number of samples: {total}") + print_rank_0(f" >>> Finished concating datasets in {time.time() - start_concating_time} seconds") print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") return datasets, [corpus_weights_achieved[c] for c in corpus_list] From 053b42ddbcdf296976e1039cf6ad55e43f6ef641 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 5 Jun 2024 18:11:15 -0500 Subject: [PATCH 184/590] changed things into helper --- megatron/data/gpt_dataset.py | 28 +++++++++++++++++++++------- megatron/data/helpers.cpp | 19 ++++++++++++++++++- 2 files changed, 39 insertions(+), 8 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index e758e6cb213..bc5691fdd79 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -11,6 +11,7 @@ from deepspeed.accelerator import get_accelerator from megatron import print_rank_0, is_rank_0, get_args, print_flush from megatron.core import mpu +from megatron.data import helpers from megatron.data.blendable_dataset import BlendableDataset from megatron.data.dataset_utils import get_datasets_weights_and_num_samples, get_datasets_corpuses_weights_and_num_samples from megatron.data.dataset_utils import get_train_valid_test_split_ @@ -99,11 +100,24 @@ def __init__(self, dataset_builders): self.indices=np.zeros((self.num_samples, 2), dtype=np.uint64) self.desc="ConcatDataset:" m = 0 + num_samples_list = np.array([d.num_samples for d in dataset_builders]) + self.num_samples = np.sum(num_samples_list) + def _build_indices(): + start_time = time.time() + dataset_index = np.zeros(self.num_samples, dtype=np.int64) + dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) + helpers.build_concat_indices(dataset_index, dataset_sample_index, + num_samples_list, + self.num_datasets, + torch.distributed.get_rank()==0) + print_rank_0('> elapsed time for building concat dataset indices: ' + '{:.2f} (sec)'.format(time.time() - start_time)) + return dataset_index, dataset_sample_index + + self.dataset_index, self.dataset_sample_index = _build_indices() for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," - self.indices[m:dataset_builders[i].num_samples+m] = [[i, j] for j in range(dataset_builders[i].num_samples)] - m+=dataset_builders[i].num_samples - assert(m==self.num_samples) + self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): return self.num_samples @@ -115,7 +129,8 @@ def __getitem__(self, idx): id = np.random.randint(self.num_samples) else: id = idx - i, j = self.indices[id] + i = self.dataset_index[idx] + j = self.dataset_sample_index[idx] if self.dataset_builders[i].build: return self.dataset_builders[i].dataset[j] else: @@ -149,7 +164,7 @@ def build_corpus_datasets(dataset_type='train'): return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): dataset_builders[i].Build() - print_rank_0(f" >>> Finished building individual datasets in time.time() - start_time() seconds") + print_rank_0(f" >>> Finished building individual datasets in {time.time() - start_time} seconds") start_concating_time = time.time() for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) @@ -164,7 +179,7 @@ def build_corpus_datasets(dataset_type='train'): print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") print_rank_0(f" > total number of samples: {total}") - print_rank_0(f" >>> Finished concating datasets in {time.time() - start_concating_time} seconds") + print_rank_0(f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds") print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") return datasets, [corpus_weights_achieved[c] for c in corpus_list] @@ -179,7 +194,6 @@ def build_corpus_datasets(dataset_type='train'): # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them - start_time = time.time() print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") torch.distributed.barrier() print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") diff --git a/megatron/data/helpers.cpp b/megatron/data/helpers.cpp index 142f159dd37..8dfed20f86d 100644 --- a/megatron/data/helpers.cpp +++ b/megatron/data/helpers.cpp @@ -15,7 +15,23 @@ namespace py = pybind11; using namespace std; const int32_t LONG_SENTENCE_LEN = 512; - +void build_concat_indices(py::array_t& dataset_index, py::array_t& dataset_sample_index, + const py::array_t &num_samples, + const int64_t num_datasets, const bool verbose) { + if (verbose) { + std::cout << "> building indices for corpus datasets ..." << std::endl; + } + auto dataset_index_ptr = dataset_index.mutable_unchecked<1>(); + auto num_samples_ptr = num_samples.unchecked<1>(); + auto dataset_sample_index_ptr = dataset_sample_index.mutable_unchecked<1>(); + int64_t m = 0; + for(uint64_t i=0; i& dataset_index, py::array_t& dataset_sample_index, @@ -698,4 +714,5 @@ PYBIND11_MODULE(helpers, m) { m.def("build_blocks_mapping", &build_blocks_mapping); m.def("build_sample_idx", &build_sample_idx); m.def("build_blending_indices", &build_blending_indices); + m.def("build_concat_indices", &build_concat_indices); } From 52b2cca8b7bdfe640590dbc20bdb83559f50d86c Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 6 Jun 2024 21:51:37 +0000 Subject: [PATCH 185/590] fixed issue with dlioprofiler --- megatron/utils.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index cd32f9a946f..ee9ef5bec4a 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -41,10 +41,7 @@ class Profile: def __init__(self, type="PROFILER"): self.type = type def log(self, func): - @wraps(func) - def wrapper(*args, **kwargs): - x = func(*args, **kwargs) - return x + return func class dlio_logger: def __init__(self,): self.type = None From cbc78308ba4e8c6b157c33c30e11dbc5f772af37 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 6 Jun 2024 17:03:06 -0500 Subject: [PATCH 186/590] fixed some bugs --- ALCF/test_blendable_dataset.py | 5 ++--- megatron/data/blendable_dataset.py | 4 ++-- megatron/data/gpt_dataset.py | 3 ++- pretrain_gpt_alcf.py | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index 7f52169a3a5..a3cabddd293 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -25,7 +25,7 @@ def print_rank_0(msg): end_time = time.time() print_rank_0(f"Loaded python modules in {end_time - start_time} seconds") initialize_megatron(allow_no_cuda=True) -torch.distributed.barrier() +comm.Barrier() print_rank_0(f"Barrier synchonization time: {time.time() - end_time} seconds") args = get_args() if os.getenv('DLIO_PROFILER_DATASET_DIR') is not None: @@ -103,12 +103,11 @@ def compute(ct): n=0 start_time = time.time() for i in dlp.iter(train_dataloader): - compute(SLEEP_TIME) print(f"[{comm.rank}] DATA {i}") n+=1 if (n%NUM_ITEMS==0): print_rank_0(f"Proccessed {n}th-batch in {time.time() - start_time}") - if n>=100: + if n>=1000: break start_time = time.time() end_loading_time = time.time() diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 005231738f3..1f6619dc39c 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -13,7 +13,7 @@ from megatron import print_rank_0, print_flush from megatron.core import mpu from megatron.utils import Profile, PerfTrace - +from mpi4py import MPI dlp = Profile("DATASET") class BlendableDataset(torch.utils.data.Dataset): @dlp.log @@ -96,7 +96,7 @@ def _build_indices(): print_rank_0("Data index creation unsuccessful, exiting.") exit() ''' - torch.distributed.barrier() + MPI.COMM_WORLD.Barrier() start_time = time.time() print_rank_0(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index bc5691fdd79..1b85f7fce8b 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -18,6 +18,7 @@ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset from megatron.utils import PerfTrace, Profile +from mpi4py import MPI dlp = Profile("DATASET") @@ -195,7 +196,7 @@ def build_corpus_datasets(dataset_type='train'): # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") - torch.distributed.barrier() + MPI.COMM_WORLD.Barrier() print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") print_rank_0(f" >>> Starting to build BlendableDataset") # Blend. diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 23bc9f2e56b..6abcd03a423 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -501,7 +501,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): ) files = [] # making it the default input method - assert(agrs.data_file_list is not None) + assert(args.data_file_list is not None) if args.data_file_list is not None: log.info(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: From 36a267111c94e89fca8fb0197888c23f14d6a4e6 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 6 Jun 2024 22:08:57 +0000 Subject: [PATCH 187/590] fixed profiler issue --- megatron/utils.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/megatron/utils.py b/megatron/utils.py index ee9ef5bec4a..bdbb3210e88 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -42,6 +42,8 @@ def __init__(self, type="PROFILER"): self.type = type def log(self, func): return func + def iter(self, a): + return a class dlio_logger: def __init__(self,): self.type = None From 5c8d376124065f1d6ec05bbde5423908e29e352d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 6 Jun 2024 18:31:15 -0500 Subject: [PATCH 188/590] reduced printing --- megatron/__init__.py | 1 - megatron/data/blendable_dataset.py | 2 +- megatron/data/gpt_dataset.py | 26 +++++++++++++------------- megatron/utils.py | 3 --- 4 files changed, 14 insertions(+), 18 deletions(-) diff --git a/megatron/__init__.py b/megatron/__init__.py index ce44561e2d0..d92a279ec67 100644 --- a/megatron/__init__.py +++ b/megatron/__init__.py @@ -14,7 +14,6 @@ from .initialize import initialize_megatron from .utils import (print_rank_0, - print_flush, is_last_rank, print_rank_last, is_rank_0, diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 1f6619dc39c..a4872613f96 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -10,7 +10,7 @@ import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0, print_flush +from megatron import print_rank_0 from megatron.core import mpu from megatron.utils import Profile, PerfTrace from mpi4py import MPI diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 1b85f7fce8b..b7a9634fffe 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -9,7 +9,7 @@ import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0, is_rank_0, get_args, print_flush +from megatron import print_rank_0, is_rank_0, get_args from megatron.core import mpu from megatron.data import helpers from megatron.data.blendable_dataset import BlendableDataset @@ -126,7 +126,7 @@ def __len__(self): @dlp.log def __getitem__(self, idx): if idx >= self.num_samples: - print_flush(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") + print_rank_0(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") id = np.random.randint(self.num_samples) else: id = idx @@ -303,7 +303,7 @@ def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_strin """Build train, valid, and test datasets.""" # Each rank print out information - print_flush(f" >> building dataset for {data_prefix}") + print_rank_0(f" >> building dataset for {data_prefix}") # Indexed dataset. indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, @@ -601,7 +601,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the # ranks will first create the document files, and then read it. # There will not be contension effects going on either - print_flush(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") + print_rank_0(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -654,7 +654,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, doc_idx = _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch) np.save(idx_path['doc'], doc_idx, allow_pickle=True) - print_flush(' > elasped time to build and save doc-idx mapping ' + print_rank_0(' > elasped time to build and save doc-idx mapping ' '(seconds): {:4f}'.format(time.time() - start_time)) # sample-idx. start_time = time.time() @@ -666,7 +666,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch) np.save(idx_path['sample'], sample_idx, allow_pickle=True) - print_flush(' > elasped time to build and save sample-idx mapping ' + print_rank_0(' > elasped time to build and save sample-idx mapping ' '(seconds): {:4f}'.format(time.time() - start_time)) # shuffle-idx. start_time = time.time() @@ -679,7 +679,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, shuffle_idx = _build_shuffle_idx(num_samples_, sample_idx.shape[0] - 1, np_rng) np.save(idx_path['shuffle'], shuffle_idx, allow_pickle=True) - print_flush(' > elasped time to build and save shuffle-idx mapping' + print_rank_0(' > elasped time to build and save shuffle-idx mapping' ' (seconds): {:4f}'.format(time.time() - start_time)) except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_dir})') @@ -691,20 +691,20 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Load mappings. start_time = time.time() - print_flush(f" > loading doc-idx mapping from {idx_path['doc']}") + print_rank_0(f" > loading doc-idx mapping from {idx_path['doc']}") doc_idx = np.load(idx_path['doc'], allow_pickle=True, mmap_mode='r') - print_flush(f" > loading sample-idx mapping from {idx_path['sample']}") + print_rank_0(f" > loading sample-idx mapping from {idx_path['sample']}") sample_idx = np.load(idx_path['sample'], allow_pickle=True, mmap_mode='r') - print_flush(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") + print_rank_0(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") shuffle_idx = np.load(idx_path['shuffle'], allow_pickle=True, mmap_mode='r') - print_flush(' loaded indexed file in {:3.3f} seconds'.format( + print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( time.time() - start_time)) - print_flush(' total number of samples: {}'.format( + print_rank_0(' total number of samples: {}'.format( sample_idx.shape[0])) - print_flush(' total number of epochs: {}'.format(num_epochs)) + print_rank_0(' total number of epochs: {}'.format(num_epochs)) return doc_idx, sample_idx, shuffle_idx, desc, desc_hash diff --git a/megatron/utils.py b/megatron/utils.py index bdbb3210e88..22e9a926512 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -260,9 +260,6 @@ def print_rank_0(message): else: print(message, flush=True) -def print_flush(message): - print(f"[{torch.distributed.get_rank()}] {message}", flush=True) - def is_last_rank(): return torch.distributed.get_rank() == ( torch.distributed.get_world_size() - 1) From d9085b6d685771969b381469a8cb8b7ddca9b939 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 10 Jun 2024 10:26:08 -0500 Subject: [PATCH 189/590] added more timing info --- megatron/training.py | 3 +++ pretrain_gpt_alcf.py | 11 ++++++++--- 2 files changed, 11 insertions(+), 3 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index b89cc559324..3eec8ba23cc 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -179,11 +179,14 @@ def pretrain( # Adjust the startup time so it reflects the largest value. # This will be closer to what scheduler will see (outside of # image ... launches. + before_allreduce = time.time() global _TRAIN_START_TIME + log.info(f"time to finish initialize_megatron: {time.time() - _TRAIN_START_TIME} seconds") start_time_tensor = get_accelerator().DoubleTensor([_TRAIN_START_TIME]) tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) # torch.distributed.all_reduce(start_time_tensor, # op=torch.distributed.ReduceOp.MIN) + log.info(f"allreduce call time: {time.time()} seconds") _TRAIN_START_TIME = start_time_tensor.item() log.info('time to initialize megatron (seconds)={:.3f}'.format( time.time() - _TRAIN_START_TIME)) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 6abcd03a423..9d64c74609e 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -1,7 +1,8 @@ # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain GPT""" - +import time +python_start_time = time.time() from pathlib import Path from mpi4py import MPI import os @@ -41,11 +42,12 @@ from torch import nn import torch.nn.functional as F import ezpz as ez - - +import_time = time.time() - python_start_time +start_setup_time = time.time() # ---- [SETUP COMMS] ------------------------ # if str(os.environ.get('LAUNCH_CMD', 'mpich')).lower() == 'mpich': RANK = ez.setup_torch(backend="deepspeed", timeout=7200) +setup_time = time.time() - start_setup_time # else: # RANK = ez.get_rank() WORLD_SIZE = ez.get_world_size() @@ -57,6 +59,9 @@ # --- [TURN OFF LOGGER ON ALL RANK != 0] ---- log = logging.getLogger(__name__) log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") +log.info(f"Import python modules in {import_time} seconds") +log.info(f"ez.setup_torch time: {setup_time} seconds") + # ---- [SETUP WANDB FROM RANK 0] -------------- WANDB_MODE = os.environ.get('WANDB_MODE', None) DISABLE_WANDB = ( From 0ef6bfd2b0b0dbecf9e8b5bc2c2e6ee76b175df6 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 10 Jun 2024 10:33:29 -0500 Subject: [PATCH 190/590] fixed timing issue for all reduce --- megatron/training.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training.py b/megatron/training.py index 3eec8ba23cc..498099c2a44 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -186,7 +186,7 @@ def pretrain( tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) # torch.distributed.all_reduce(start_time_tensor, # op=torch.distributed.ReduceOp.MIN) - log.info(f"allreduce call time: {time.time()} seconds") + log.info(f"allreduce call time: {time.time()-before_allreduce} seconds") _TRAIN_START_TIME = start_time_tensor.item() log.info('time to initialize megatron (seconds)={:.3f}'.format( time.time() - _TRAIN_START_TIME)) From f6363fbf9ab60b87cffac48d2d2b57dc5499925d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 12 Jun 2024 14:33:20 +0000 Subject: [PATCH 191/590] changed init --- pretrain_gpt_alcf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 9d64c74609e..f9e7a4d0a1d 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -109,7 +109,7 @@ def model_provider(pre_process=True, post_process=True): dpg = mpu.get_data_parallel_group() else: dpg = None - with deepspeed.zero.Init( + with deepspeed.zero.MiCS_Init( data_parallel_group=dpg, remote_device=( None if args.remote_device == 'none' else args.remote_device From a55df512752d5a18ae49bfe298c4ee63c2479bb6 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 12 Jun 2024 22:51:29 +0000 Subject: [PATCH 192/590] reducing printing from non-root ranks --- megatron/data/gpt_dataset.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index b7a9634fffe..47beadc27ad 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -610,8 +610,8 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # not mean anything. if num_epochs == 1: separate_last_epoch = False - print(' > only one epoch required, setting ' - 'separate_last_epoch to False', flush=True) + print_rank_0(' > only one epoch required, setting ' + 'separate_last_epoch to False') else: # Get the number of samples for the last epoch @@ -638,8 +638,8 @@ def _build_index_mappings(name, data_prefix, documents, sizes, string = ' > last epoch number of samples ({}) is larger '\ 'than 80% of number of samples per epoch ({}), '\ 'setting separate_last_epoch to False' - print(string.format(last_epoch_num_samples, - num_samples_per_epoch), flush=True) + print_rank_0(string.format(last_epoch_num_samples, + num_samples_per_epoch)) try: From a24f01be0f7138bc71a781725f62c815d8050983 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 12 Jun 2024 22:53:17 +0000 Subject: [PATCH 193/590] reduce printing --- megatron/data/gpt_dataset.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index b7a9634fffe..47beadc27ad 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -610,8 +610,8 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # not mean anything. if num_epochs == 1: separate_last_epoch = False - print(' > only one epoch required, setting ' - 'separate_last_epoch to False', flush=True) + print_rank_0(' > only one epoch required, setting ' + 'separate_last_epoch to False') else: # Get the number of samples for the last epoch @@ -638,8 +638,8 @@ def _build_index_mappings(name, data_prefix, documents, sizes, string = ' > last epoch number of samples ({}) is larger '\ 'than 80% of number of samples per epoch ({}), '\ 'setting separate_last_epoch to False' - print(string.format(last_epoch_num_samples, - num_samples_per_epoch), flush=True) + print_rank_0(string.format(last_epoch_num_samples, + num_samples_per_epoch)) try: From 5a5414959702d4b2775e66fd99d2ffbc9212420d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 12 Jun 2024 23:28:43 +0000 Subject: [PATCH 194/590] reducing printing --- megatron/data/gpt_dataset.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 47beadc27ad..734bdbfe43b 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -796,8 +796,8 @@ def _build_sample_idx(sizes, doc_idx, seq_length, @dlp.log def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" - print(' > building shuffle index with split [0, {}) and [{}, {}) ' - '...'.format(num_samples, num_samples, total_size), flush=True) + print_rank_0(' > building shuffle index with split [0, {}) and [{}, {}) ' + '...'.format(num_samples, num_samples, total_size)) dtype_ = np.uint32 if total_size >= (np.iinfo(np.uint32).max - 1): From 3acdda7e1e895375aead8b2da0a4a7d418d3cbfe Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 13 Jun 2024 22:40:12 +0000 Subject: [PATCH 195/590] added MiCS as an option --- megatron/arguments.py | 2 ++ megatron/data/gpt_dataset.py | 2 +- megatron/data/helpers.cpp | 23 ++++++++++++----------- pretrain_gpt_alcf.py | 11 +++++++++-- 4 files changed, 24 insertions(+), 14 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 3143e449c41..9a5e4b8da7e 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1504,6 +1504,8 @@ def _add_zero_args(parser): help='Remote device for ZeRO-3 initialized parameters.') group.add_argument('--use-pin-memory', action='store_true', help='Use pinned CPU memory for ZeRO-3 initialized model parameters.') + group.add_argument('--use-mics', action='store_true', + help='Use MiCS') return parser def _add_memoryopt_args(parser): diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 734bdbfe43b..686ccecd70d 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -664,7 +664,7 @@ def _build_index_mappings(name, data_prefix, documents, sizes, assert doc_idx.dtype == np.int32 assert sizes.dtype == np.int32 sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, - num_epochs, tokens_per_epoch) + num_epochs, tokens_per_epoch, torch.distributed.get_rank()==0) np.save(idx_path['sample'], sample_idx, allow_pickle=True) print_rank_0(' > elasped time to build and save sample-idx mapping ' '(seconds): {:4f}'.format(time.time() - start_time)) diff --git a/megatron/data/helpers.cpp b/megatron/data/helpers.cpp index 8dfed20f86d..9dee0589b6d 100644 --- a/megatron/data/helpers.cpp +++ b/megatron/data/helpers.cpp @@ -100,7 +100,7 @@ py::array build_sample_idx(const py::array_t& sizes_, const py::array_t& doc_idx_, const int32_t seq_length, const int32_t num_epochs, - const int64_t tokens_per_epoch) { + const int64_t tokens_per_epoch, const bool verbose=false) { /* Sample index (sample_idx) is used for gpt2 like dataset for which the documents are flattened and the samples are built based on this 1-D flatten array. It is a 2D array with sizes [number-of-samples + 1, 2] @@ -119,16 +119,17 @@ py::array build_sample_idx(const py::array_t& sizes_, // Mapping and it's length (1D). int64_t num_samples = (num_epochs * tokens_per_epoch - 1) / seq_length; int64_t* sample_idx = new int64_t[2*(num_samples+1)]; - - cout << " using:" << endl << std::flush; - cout << " number of documents: " << - doc_idx_.shape(0) / num_epochs << endl << std::flush; - cout << " number of epochs: " << num_epochs << - endl << std::flush; - cout << " sequence length: " << seq_length << - endl << std::flush; - cout << " total number of samples: " << num_samples << - endl << std::flush; + if (verbose) { + cout << " using:" << endl << std::flush; + cout << " number of documents: " << + doc_idx_.shape(0) / num_epochs << endl << std::flush; + cout << " number of epochs: " << num_epochs << + endl << std::flush; + cout << " sequence length: " << seq_length << + endl << std::flush; + cout << " total number of samples: " << num_samples << + endl << std::flush; + } // Index into sample_idx. int64_t sample_index = 0; diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index f9e7a4d0a1d..de48dc6dcb5 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -2,9 +2,12 @@ """Pretrain GPT""" import time +from mpi4py import MPI +comm = MPI.COMM_WORLD +comm.Barrier() python_start_time = time.time() from pathlib import Path -from mpi4py import MPI + import os from rich import print import torch @@ -109,7 +112,11 @@ def model_provider(pre_process=True, post_process=True): dpg = mpu.get_data_parallel_group() else: dpg = None - with deepspeed.zero.MiCS_Init( + + deepspeed_zero_init = deepspeed.zero.Init + if args.use_mics: + deepspeed_zero_init = deepspeed.zero.MiCS_Init + with deepspeed_zero_init( data_parallel_group=dpg, remote_device=( None if args.remote_device == 'none' else args.remote_device From 712d08d861d9e80926af79a5fec002889bf63e90 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 13 Jun 2024 21:11:24 -0500 Subject: [PATCH 196/590] Update `dropout` in `ALCF/helpers.sh` Set `--attention-dropout` and `--hidden-dropout` to `0.0` in `LLAMA_ARGS` in `ALCF/helpers.sh` for Llama style models --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8dc519d501d..c182d9cbafd 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -363,7 +363,7 @@ get_batch_size_on_polaris() { # ``` ############################################################################## setParams() { - LLAMA_ARGS="" + LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" # +----[Parallelism Settings] -------------------------------------------+ # +------[Aurora]--------||-------[SunSpot]-------------+ if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then From 482c2354f9068b7d58b331ea52cc3c6996ace7c4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 13 Jun 2024 23:37:46 -0500 Subject: [PATCH 197/590] Update {`ALCF/helpers.sh`, `train_llama_alcf.sh`} --- ALCF/helpers.sh | 44 +++++++++++++++++++++++++++------------- train_llama_alcf.sh | 49 ++++++++++++++++++--------------------------- 2 files changed, 50 insertions(+), 43 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index c182d9cbafd..02d1faa12c2 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,5 +1,17 @@ #!/bin/bash --login # +# Contains helper functions for launching `../train_llama_alcf.sh` +# +# Example for interactive use: +# +# ```bash +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed +# $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh +# $ setEnv # will setup conda + virtual environment +# ``` + +# for debug mode, uncomment below: # set -euxo pipefail #################################################################### @@ -8,18 +20,20 @@ # 2. else, if `${SLURM_SUBMIT_DIR}` has nonzero value, use that # 3. else, use `$(pwd)` #################################################################### -if [[ -n "${PBS_O_WORKDIR}" ]]; then - WORKING_DIR="${PBS_O_WORKDIR}" -elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then - WORKING_DIR="${SLURM_SUBMIT_DIR}" -else - echo "Unable to detect PBS or SLURM working directory info..." - WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') - echo "Using ${WORKING_DIR} as working directory..." -fi - -export WORKING_DIR="${WORKING_DIR}" -printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" +main() { + if [[ -n "${PBS_O_WORKDIR}" ]]; then + WORKING_DIR="${PBS_O_WORKDIR}" + elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then + WORKING_DIR="${SLURM_SUBMIT_DIR}" + else + echo "Unable to detect PBS or SLURM working directory info..." + WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') + echo "Using ${WORKING_DIR} as working directory..." + fi + + export WORKING_DIR="${WORKING_DIR}" + printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" +} ############################################################################## # `setup`: All-in-one helper function. @@ -541,8 +555,8 @@ setup_ezpz() { fi ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -n "${ezloc}" ]]; then - ezpz_savejobenv - python3 -m ezpz.jobs && source "./.jobenv" + # ezpz_savejobenv + # python3 -m ezpz.jobs && source "./.jobenv" make_ds_hostfile || exit else echo "No ezpz detected. Attempting to install with $(which python3)" @@ -1153,6 +1167,8 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } +main + #### [DEPRECATED] ########################################################### # if [[ -z "${HOSTFILE}" ]]; then # makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 98ef99abfd0..778c4e82fe3 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -20,35 +20,26 @@ if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` set -o noexec # same as set -n fi -################################################## -# Helper function for `source`-ing another file -################################################## -sourceFile() { - fp="$1" - echo "source-ing ${fp}" - if [[ -f "${fp}" ]]; then - source "${fp}" - else - echo "ERROR: UNABLE TO SOURCE ${fp}" - fi -} +main() { + ##################### + # MAIN PROGRAM LOGIC + ##################### + #### 1. Navigate into `$PBS_O_WORKDIR` + cd "${PBS_O_WORKDIR}" || exit + #### 2. source `ALCF/helpers.sh` + HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE + source "${HERE}/ALCF/helpers.sh" || exit + #### 3. call `setup` from `./ALCF/helpers.sh` + setup || exit -##################### -# MAIN PROGRAM LOGIC -##################### -# ----[1. Navigate into `$PBS_O_WORKDIR`]-------------------------------------- -cd "${PBS_O_WORKDIR}" || exit -HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE -sourceFile "${HERE}/ALCF/helpers.sh" || exit -setup || exit -############################################################################### + # Take custom args + export custom_args=" $@" + export run_cmd="${run_cmd} ${custom_args}" -# Take custom args -export custom_args=" $@" -export run_cmd="${run_cmd} ${custom_args}" + echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" + printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" + set +x +} -echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" -printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" -# eval "${run_cmd}" >> "${OUTPUT_LOG}" 2>&1 & -eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" -set +x +main From f4c2c16940be1fd04c9d94e50c96014361298ab7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 11:58:58 -0500 Subject: [PATCH 198/590] Add `ALCF/data-lists/aurora/*.txt` --- ALCF/data-lists/aurora/algebraic.txt | 16 + ALCF/data-lists/aurora/arxiv.txt | 100 + ALCF/data-lists/aurora/books.txt | 3 + ALCF/data-lists/aurora/c4.txt | 171 ++ ALCF/data-lists/aurora/cc.txt | 1108 ++++++++ .../aurora/dolma_v1_7_file_list.txt | 2419 +++++++++++++++++ ALCF/data-lists/aurora/falcon.txt | 501 ++++ ALCF/data-lists/aurora/megawiki.txt | 262 ++ .../data-lists/aurora/open-web-math-train.txt | 13 + ALCF/data-lists/aurora/pes2o.txt | 26 + ALCF/data-lists/aurora/reddit.txt | 78 + ALCF/data-lists/aurora/stack.txt | 26 + ALCF/data-lists/aurora/starcoder.txt | 50 + ALCF/data-lists/aurora/tulu.txt | 66 + ALCF/data-lists/aurora/wiki.txt | 2 + 15 files changed, 4841 insertions(+) create mode 100644 ALCF/data-lists/aurora/algebraic.txt create mode 100644 ALCF/data-lists/aurora/arxiv.txt create mode 100644 ALCF/data-lists/aurora/books.txt create mode 100644 ALCF/data-lists/aurora/c4.txt create mode 100644 ALCF/data-lists/aurora/cc.txt create mode 100644 ALCF/data-lists/aurora/dolma_v1_7_file_list.txt create mode 100644 ALCF/data-lists/aurora/falcon.txt create mode 100644 ALCF/data-lists/aurora/megawiki.txt create mode 100644 ALCF/data-lists/aurora/open-web-math-train.txt create mode 100644 ALCF/data-lists/aurora/pes2o.txt create mode 100644 ALCF/data-lists/aurora/reddit.txt create mode 100644 ALCF/data-lists/aurora/stack.txt create mode 100644 ALCF/data-lists/aurora/starcoder.txt create mode 100644 ALCF/data-lists/aurora/tulu.txt create mode 100644 ALCF/data-lists/aurora/wiki.txt diff --git a/ALCF/data-lists/aurora/algebraic.txt b/ALCF/data-lists/aurora/algebraic.txt new file mode 100644 index 00000000000..febdf441fb4 --- /dev/null +++ b/ALCF/data-lists/aurora/algebraic.txt @@ -0,0 +1,16 @@ +0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document diff --git a/ALCF/data-lists/aurora/arxiv.txt b/ALCF/data-lists/aurora/arxiv.txt new file mode 100644 index 00000000000..3a822f5f44c --- /dev/null +++ b/ALCF/data-lists/aurora/arxiv.txt @@ -0,0 +1,100 @@ +0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document diff --git a/ALCF/data-lists/aurora/books.txt b/ALCF/data-lists/aurora/books.txt new file mode 100644 index 00000000000..946013b617b --- /dev/null +++ b/ALCF/data-lists/aurora/books.txt @@ -0,0 +1,3 @@ +0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document diff --git a/ALCF/data-lists/aurora/c4.txt b/ALCF/data-lists/aurora/c4.txt new file mode 100644 index 00000000000..4bc8e358205 --- /dev/null +++ b/ALCF/data-lists/aurora/c4.txt @@ -0,0 +1,171 @@ +0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document diff --git a/ALCF/data-lists/aurora/cc.txt b/ALCF/data-lists/aurora/cc.txt new file mode 100644 index 00000000000..5d4fce7cff2 --- /dev/null +++ b/ALCF/data-lists/aurora/cc.txt @@ -0,0 +1,1108 @@ +0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document + diff --git a/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt b/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt new file mode 100644 index 00000000000..2bb2ea805e7 --- /dev/null +++ b/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document diff --git a/ALCF/data-lists/aurora/falcon.txt b/ALCF/data-lists/aurora/falcon.txt new file mode 100644 index 00000000000..3aeb4a37513 --- /dev/null +++ b/ALCF/data-lists/aurora/falcon.txt @@ -0,0 +1,501 @@ +0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document + diff --git a/ALCF/data-lists/aurora/megawiki.txt b/ALCF/data-lists/aurora/megawiki.txt new file mode 100644 index 00000000000..4983cd797c8 --- /dev/null +++ b/ALCF/data-lists/aurora/megawiki.txt @@ -0,0 +1,262 @@ +6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document diff --git a/ALCF/data-lists/aurora/open-web-math-train.txt b/ALCF/data-lists/aurora/open-web-math-train.txt new file mode 100644 index 00000000000..a1a80aa1677 --- /dev/null +++ b/ALCF/data-lists/aurora/open-web-math-train.txt @@ -0,0 +1,13 @@ +0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document diff --git a/ALCF/data-lists/aurora/pes2o.txt b/ALCF/data-lists/aurora/pes2o.txt new file mode 100644 index 00000000000..751cce20a02 --- /dev/null +++ b/ALCF/data-lists/aurora/pes2o.txt @@ -0,0 +1,26 @@ +0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document diff --git a/ALCF/data-lists/aurora/reddit.txt b/ALCF/data-lists/aurora/reddit.txt new file mode 100644 index 00000000000..ea235a34685 --- /dev/null +++ b/ALCF/data-lists/aurora/reddit.txt @@ -0,0 +1,78 @@ +0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document diff --git a/ALCF/data-lists/aurora/stack.txt b/ALCF/data-lists/aurora/stack.txt new file mode 100644 index 00000000000..bc7f3efbc75 --- /dev/null +++ b/ALCF/data-lists/aurora/stack.txt @@ -0,0 +1,26 @@ +0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document diff --git a/ALCF/data-lists/aurora/starcoder.txt b/ALCF/data-lists/aurora/starcoder.txt new file mode 100644 index 00000000000..401f053d944 --- /dev/null +++ b/ALCF/data-lists/aurora/starcoder.txt @@ -0,0 +1,50 @@ +0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document + diff --git a/ALCF/data-lists/aurora/tulu.txt b/ALCF/data-lists/aurora/tulu.txt new file mode 100644 index 00000000000..ac59c679df1 --- /dev/null +++ b/ALCF/data-lists/aurora/tulu.txt @@ -0,0 +1,66 @@ +0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document diff --git a/ALCF/data-lists/aurora/wiki.txt b/ALCF/data-lists/aurora/wiki.txt new file mode 100644 index 00000000000..f0ab87cb679 --- /dev/null +++ b/ALCF/data-lists/aurora/wiki.txt @@ -0,0 +1,2 @@ +0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document From 231d2b59393d2c06093a11fe90e71165357ccceb Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 11:59:16 -0500 Subject: [PATCH 199/590] Add `setup_conda_aurora` to `ALCF/helpers.sh` --- ALCF/helpers.sh | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 02d1faa12c2..a0b9793c11c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -677,6 +677,12 @@ setup_conda_sunspot() { fi } +setup_conda_aurora() { + if [[ -z "${CONDA_PREFIX:-}" ]]; then + module use -a /soft/modulefiles ; module load frameworks/2024.1 + fi +} + ######################## # Setup conda on Sirius ######################## From aaf61526b275c7237c45b8205ead45fb8cdca81b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 12:26:27 -0500 Subject: [PATCH 200/590] Fix `ezpz_{save,get}jobenv` in `ALCF/helpers.sh` --- ALCF/helpers.sh | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a0b9793c11c..50efae5afc6 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -531,13 +531,13 @@ make_ds_hostfile() { } ezpz_savejobenv() { - local outfile="${WORKING_DIR}/savejobenv" - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/jobs-cleanup/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" + local outfile="${WORKING_DIR}/bin/savejobenv" ; mkdir -p $(dirname "${outfile}") + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" } ezpz_getjobenv() { - local outfile="${WORKING_DIR}/getjobenv" - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/jobs-cleanup/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" + local outfile="${WORKING_DIR}/bin/getjobenv" ; mkdir -p $(dirname "${outfile}") + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" } # +---------------------------------------+ From b905e5306d92add62417ffc2967275b002b610a3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 12:42:11 -0500 Subject: [PATCH 201/590] Correctly set `dfl_fallback` on Aurora if no `DATA_FILE_LIST` specified --- ALCF/helpers.sh | 26 +------------------------- 1 file changed, 1 insertion(+), 25 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 50efae5afc6..39b6ce6c8aa 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -903,33 +903,9 @@ setup_tokenizer_and_data() { # fallback to default values if necessary. ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] - # if [[ "$#" -ne 1 ]]; then - # tok="${TOKENIZER_TYPE:-Llama2}" - # else - # tok="$1" - # fi - # echo "Setting up tokenizer with ${tok}" - # setup_tokenizer "${tok}" - # tok="${TOKENIZER_TYPE:-}" - # if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then - # export TOKENIZER_TYPE="GPT2" - # export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" - # export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" - # export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" - # export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - # # TOKENIZER_FLAGS="--data-path $DATA_PATH--vocab-file $VOCAB_FILE --merge-file ${MERGE_FILE}" - # DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" - # # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-GPT2}" - # # else [[ ${tok} == Llama* || ${tok} == llama* || ${tok} == LLAMA* ]]; then - # else - # export TOKENIZER_TYPE="Llama2" - # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - # TOKENIZER_FLAGS="${TOKENIZER_FLAGS} --tokenizer-type Llama2Tokenizer" - # dfldir="${WORKING_DIR}/ALCF/data-lists" # =====[Set DATA_FILE_LIST_FALLBACK based on current machine]============== if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] - dfl_fallback="/home/foremans/anl_24_release_q4/llm.devkit/Megatron-DeepSpeed/data_file_list_reweighted.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt" elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] # shellcheck source=./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt" From ba5f871a78e473e22b2eab3eb4faadabc3166ca2 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 14 Jun 2024 13:45:21 -0500 Subject: [PATCH 202/590] added warning if the file list is not provided correctly --- megatron/training.py | 6 +++--- pretrain_gpt_alcf.py | 13 +++++++++++-- 2 files changed, 14 insertions(+), 5 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 498099c2a44..703c1a15127 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -345,7 +345,7 @@ def update_train_iters(args): log.info('setting training iterations to {}'.format(args.train_iters)) - +@dlp.log def setup_teacher_model(args, model_provider): log.info('***>>>>> Student model checkpoint iteration:{}'.format(args.iteration)) @@ -488,7 +488,7 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap return model - +@dlp.log def get_optimizer_param_scheduler(optimizer): """Build the learning rate scheduler.""" args = get_args() @@ -570,7 +570,7 @@ def load_model_weights_only(model_provider_func): return model, optimizer, lr_scheduler - +@dlp.log def setup_model_and_optimizer(model_provider_func, model_type, no_wd_decay_cond=None, diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index de48dc6dcb5..b5ec065c008 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -517,8 +517,17 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): if args.data_file_list is not None: log.info(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: - for f in flist.readlines(): - w, fname, c = f.split() + lines = flist.readlines() + if len(lines[0].split()) < 3: + log.warning(f"> Corpus type is not provided in {args.data_file_list}; will assume the datasets are from different corpuses; suggesting to add corpus type to improve the performance.") + for i, f in enumerate(lines): + try: + w, fname, c = f.split() + except: + w, fname = f.split() + c = str(i) + if fname.find(".bin") != -1: + fname = fname.split(".bin")[0] files.append(float(w)) files.append(fname) files.append(c) From c6902029505ebf3969b27245be140e481ac68a35 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 14 Jun 2024 13:46:18 -0500 Subject: [PATCH 203/590] make it still compatible to previous --- pretrain_gpt_alcf.py | 2 -- 1 file changed, 2 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index b5ec065c008..823c84b17fe 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -512,8 +512,6 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): '> building train, validation, and test datasets for GPT ...' ) files = [] - # making it the default input method - assert(args.data_file_list is not None) if args.data_file_list is not None: log.info(f"Reading datasets from {args.data_file_list}") with open(args.data_file_list, 'r') as flist: From a96bceaf1123f8b94c092d711a4c8b53d269c43f Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 14 Jun 2024 13:58:49 -0500 Subject: [PATCH 204/590] added support for XPU --- megatron/training.py | 6 +++++- pretrain_gpt_alcf.py | 6 +++++- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 703c1a15127..de7d809890b 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1346,7 +1346,11 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, args.curr_iteration = iteration if os.getenv("TORCH_PROFILER_ENABLE")=='2': from torch.profiler import profile, record_function, ProfilerActivity - with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA]) as prof: + try: + activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA, ProfilerActivity.XPU] + except: + activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA] + with profile(activities=activities) as prof: loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ train_step(forward_step_func, train_data_iterator, diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 823c84b17fe..846e340f643 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -595,7 +595,11 @@ def git_ds_info(): def main(): if os.getenv('TORCH_PROFILER_ENABLE') == '1': from torch.profiler import profile, record_function, ProfilerActivity - with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA]) as prof: + try: + activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA, ProfilerActivity.XPU] + except: + activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA] + with profile(activities=activities) as prof: model = pretrain( train_valid_test_datasets_provider, model_provider, From 30fe47955c0d1a99c452ffc9aae581c341037b92 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 17:45:14 -0500 Subject: [PATCH 205/590] Update README.md --- ALCF/README.md | 874 +++++++++++++++++++++++++------------------------ 1 file changed, 453 insertions(+), 421 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 907cbd36cca..00a70dbd42f 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -3,7 +3,7 @@ ## 🆘 Getting Started -> [!NOTE] +> [!IMPORTANT] > [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching > distributed training on {Polaris, Aurora, Sunspot} @ ALCF. @@ -21,59 +21,91 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): -
⏳ Request an interactive job with qsub -I: +1.
⏳ Request an interactive job with qsub -I: -```bash -qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I -``` + ```bash + qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I + ```
-
⬇️ Clone repo + navigate into it: +2.
⬇️ Clone repo + navigate into it: -```bash -git clone "https://github.com/argonne-lcf/Megatron-DeepSpeed" -cd Megatron-DeepSpeed -``` + ```bash + git clone "https://github.com/argonne-lcf/Megatron-DeepSpeed" + cd Megatron-DeepSpeed + ```
-
🐍 Setup Python: +3.
🐍 Setup Python: -1. 📂 Load `conda` module and activate base environment: +
+ + > **NOTE**: The following commands should be ran from [`Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed), following the `cd` command from 2. - - **Polaris**: + 1. Load `conda` module and activate base environment: + + ```bash + PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + ``` + -
[output] Sunspot @ ALCF: + + ```bash + # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed][main *1 31 ?32] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + Running on SunSpot !! + + Due to MODULEPATH changes, the following have been reloaded: + 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 + 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 + 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 + 4) mpfr/4.2.0-w7v7yjv + + The following have been reloaded with a version change: + 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 + 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 + UMD: agama-ci-devel-803.29 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-803.29 + + The following have been reloaded with a version change: + 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 + + Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + No VIRTUAL_ENV found in environment! + - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + - Using VENV_DIR=/gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 + ``` + +
+ + + + 2. 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): + ```bash - source ALCF/sunspot-env-2024-04-15-002.sh + mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz + python3 -m pip install -e deps/ezpz --require-virtualenv ``` -3. 👻 Create virtual environment _on top of the base `conda`_[^venv]: - - ```bash - export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_venv_from_conda - ``` - - -4. 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): - - ```bash - mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz - python3 -m pip install -e deps/ezpz --require-virtualenv - ``` - -[^venv]: Its generally a good practice to keep separate virtual Python environments different projects. - We provide a helper function, [`setup_venv_from_conda()`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/2f0154394bbdf3c64b4669f9d944645e2cdb8f2b/ALCF/helpers.sh#L440), - that helps take care of this for you. -
- This will: activate (or build, if necessary) a `venv` in your working dir, - _automatically_ matching the name of your active `conda` environment (e.g. `2024-04-29`, on Polaris_. + [^venv]: Its generally a good practice to keep separate virtual Python environments different projects. + We provide a helper function, [`setup_venv_from_conda()`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/2f0154394bbdf3c64b4669f9d944645e2cdb8f2b/ALCF/helpers.sh#L440), + that helps take care of this for you. +
+ This will: activate (or build, if necessary) a `venv` in your working dir, + _automatically_ matching the name of your active `conda` environment (e.g. `2024-04-29`, on Polaris_.
@@ -99,389 +131,389 @@ setup_venv_from_conda # 3. will (1.) --> -
🚀 Launch: - -In this case, train a ~ 2B Model (with 10 layers), -for 1000 iterations using the data file list in: - -[`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) - -with a micro-batch-size of 2, with the `torch.optim.AdamW` optimizer. - -**Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) -function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) -can be overridden dynamically at runtime using this technique. - -```bash -PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh -``` - -
[output]: - -
- -
[Sunspot]: - -```bash -# [09:07:32 AM] [foremans@x1921c0s0b0n0] ~/q/llm.devkit/Megatron-DeepSpeed  main !1 ?27 q4-drop 26s ✘ INT -$ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./convergence_debug_small.txt bash train_llama_alcf.sh -source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh -Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... - UMD: agama-ci-devel-736.9 successfully loaded: - UMD: graphics-compute-runtime/agama-ci-devel-736.9 -Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" - -Please check the spelling or version number. Also try "module spider ..." -It is also possible your cache file is out-of-date; it may help to try: - $ module --ignore_cache load "gcc/12.1.0" - -Also make sure that all modulefiles written in TCL start with the string #%Module - -Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. - -Running on SunSpot !! -[python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 -Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env -Found ezpz! -/lus/gila/projects/Aurora_deployment/foremans/locations/sunspot/projects/saforem2/ezpz/src/ezpz/__init__.py -Has ezpz installed. Nothing to do. -Done with ezpz. -┌─────────────────────────────────────────────────────────────────── -│ Writing PBS vars to /home/foremans/.pbsenv -│ HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 -│ NHOSTS: 2 -│ NGPU_PER_HOST: 12 GPUs per host -│ NGPUS: 24 GPUs total -└─────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [Hosts]: -│ • [host:0] - x1921c0s0b0n0.hostmgmt2000.cm.americas.sgi.com -│ • [host:1] - x1921c0s1b0n0.hostmgmt2000.cm.americas.sgi.com -└────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [DIST INFO]: -│ • Loading job env from: /home/foremans/.pbsenv -│ • HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 -│ • NHOSTS: 2 -│ • NGPU_PER_HOST: 12 -│ • NGPUS (NHOSTS x NGPU_PER_HOST): 24 -│ • WORLD_SIZE: 24 -│ • DIST_LAUNCH: mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001 -└────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [Launch]: -│ • Use: 'launch' (=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001) -│ to launch job -└────────────────────────────────────────────────────────────────── -DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json -ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 -!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! -!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! -Calling: setData() with ./convergence_debug_small.txt --------------------- -Updated environment: -DATA_FILE_LIST: ./convergence_debug_small.txt -NUM_DOCS: 15 - WEIGHT_SUM: 15.0 -DFL_STEM: convergence_debug_small -DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --------------------- -++++++++++++++++++++++++++++++++++++++++++++++++++ -- MPICH_DIR= -- Using /home/foremans/miniconda3/envs/q4-drop/bin/python3 -- WORLD_SIZE:24 -- NCCL: nccl -- MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden -- Using DATA_FILE_LIST: ./convergence_debug_small.txt -++++++++++++++++++++++++++++++++++++++++++++++++++ -! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed -/home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html - __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') -/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you dont plan on using image function -ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch -vision` from source? - warn( -[2024-04-04 09:07:45,585] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) -[2024-04-04 09:07:45,818] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) --------------------------------------------------- -DeepSpeed C++/CUDA extension op report --------------------------------------------------- -NOTE: Ops not installed will be just-in-time (JIT) compiled at - runtime if needed. Op compatibility means that your system - meet the required dependencies to JIT install the op. --------------------------------------------------- -JIT compiled ops requires ninja -ninja .................. [OKAY] --------------------------------------------------- -op name ................ installed .. compatible --------------------------------------------------- -async_io ............... [NO] ....... [OKAY] -cpu_adagrad ............ [NO] ....... [OKAY] -cpu_adam ............... [NO] ....... [OKAY] -flash_attn ............. [NO] ....... [OKAY] -fused_adam ............. [NO] ....... [OKAY] -quantizer .............. [NO] ....... [OKAY] -transformer ............ [NO] ....... [OKAY] -transformer_inference .. [NO] ....... [OKAY] -utils .................. [NO] ....... [OKAY] --------------------------------------------------- -DeepSpeed general environment info: -torch install path ............... ['/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch'] -torch version .................... 2.1.0a0+cxx11.abi -deepspeed install path ........... ['/lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/DeepSpeed/deepspeed'] -deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD -deepspeed wheel compiled w. ...... torch 2.1 -shared memory (/dev/shm) size .... 503.18 GB - - deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ -foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay --style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 - --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/tensorboard --log-timers-to-tensorboard --log-optimizer --states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 - --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend ccl --num-attention-heads 32 --save-interval 20 -0 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 4 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 1 --global-bat -ch-size 96 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ -.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/tokenizer.model --no-query- -key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z -ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 -_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - -[!! NOTE] View output at: -logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - -# ... - -/gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin - creating memory view of numpy buffer... - > finished creating indexed dataset in 0.010017 seconds - number of documents: 1498927 - > dataset split: - train: - document indices in [0, 1498927) total of 1498927 documents - validation: - document indices in [1498927, 1498927) total of 0 documents - test: - document indices in [1498927, 1498927) total of 0 documents - > loading doc-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_doc_idx.npy - > loading sample-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_sample_idx.npy - > loading shuffle-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_shuffle_idx.npy - loaded indexed file in 0.056 seconds - total number of samples: 2318461 - total number of epochs: 8 -> loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy -> loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy -/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. - -[after dataloaders are built] datetime: 2024-04-04 09:09:27 -done with setup ... -(min, max) time across ranks (ms): - model-and-optimizer-setup ......................: (64818.18, 64858.22) - train/valid/test-data-iterators-setup ..........: (1968.10, 2288.56) -training ... -[before the start of training step] datetime: 2024-04-04 09:09:27 -[2024-04-04 09:09:27,718] [INFO] [checkpointing.py:540:forward] Activation Checkpointing Information -[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:541:forward] ----Partition Activations False, CPU CHECKPOINTING False -[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:542:forward] ----contiguous Memory Checkpointing False with 32 total layers -[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:544:forward] ----Synchronization False -[2024-04-04 09:09:27,719] [INFO] [checkpointing.py:545:forward] ----Profiling time in checkpointing False -[2024-04-04 09:09:33][INFO][utils:145] - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. -[2024-04-04 09:09:33][INFO][utils:148] - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. -[2024-04-04 09:09:33][INFO][utils:160] - NumExpr defaulting to 8 threads. -^[c[2024-04-04 09:09:53,311] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 884.11 | optimizer_gradients: 6.43 | optimizer_step: 23.44 -[2024-04-04 09:09:53,312] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] -[2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 6567.68 | bwd_microstep: 17950.36 | bwd_inner_microstep: 17711.20 | bwd_allreduce_microstep: 239.11 | step_microstep: 1139.27 -[2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 6567.66 | bwd: 17950.35 | bwd_inner: 17711.19 | bwd_allreduce: 239.11 | step: 1139.29 -[Rank 0] (after 1 iterations) memory (MB) | allocated: 18244.640625 | max allocated: 41299.50146484375 | reserved: 46764.0 | max reserved: 46764.0 - iteration 1/ 317892 | consumed samples: 96 | consumed tokens: 393216 | elapsed time per iteration (ms): 25849.1 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 1.117136E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.714 | tokens per gpu per second(tgs): 633.832 | TFLOPs: 38.61 | -[2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 327.85 | optimizer_gradients: 6.26 | optimizer_step: 23.60 -[2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] -[2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 4022.74 | bwd_microstep: 15738.67 | bwd_inner_microstep: 15556.80 | bwd_allreduce_microstep: 181.82 | step_microstep: 371.01 -[2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 - iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | -``` - -
- -
[Polaris]: - -```bash -[09:31:35 AM] [foremans@x3112c0s13b0n0] ~/pol/p/a/Megatron-DeepSpeed  main !4 ?24 cu118-pt221 ✘ INT -$ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./convergence_debug_small.txt DTYPE=bf16 OPT=adamw bash train_llama_alcf.sh -source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh -Running on Polaris !! - -[python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 -Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env -Found ezpz! -/lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/ezpz/src/ezpz/__init__.py -Has ezpz installed. Nothing to do. -Done with ezpz. -┌─────────────────────────────────────────────────────────────────── -│ Writing PBS vars to /home/foremans/.pbsenv -│ HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov -│ NHOSTS: 2 -│ NGPU_PER_HOST: 4 GPUs per host -│ NGPUS: 8 GPUs total -└─────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [Hosts]: -│ • [host:0] - x3112c0s13b0n0.hsn.cm.polaris.alcf.anl.gov -│ • [host:1] - x3112c0s13b1n0.hsn.cm.polaris.alcf.anl.gov -└────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [DIST INFO]: -│ • Loading job env from: /home/foremans/.pbsenv -│ • HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov -│ • NHOSTS: 2 -│ • NGPU_PER_HOST: 4 -│ • NGPUS (NHOSTS x NGPU_PER_HOST): 8 -│ • WORLD_SIZE: 8 -│ • DIST_LAUNCH: mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov -└────────────────────────────────────────────────────────────────── -┌────────────────────────────────────────────────────────────────── -│ [Launch]: -│ • Use: 'launch' (=mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov) -│ to launch job -└────────────────────────────────────────────────────────────────── -DS_CONFIG: ds_stage2_mb8_gb32_pp1_bf16.json -ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 -!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! -!! Caught USE_ACTIVATION_CHECKPOINTING=1 !! -Calling: setData() with ./convergence_debug_small.txt --------------------- -Updated environment: -DATA_FILE_LIST: ./convergence_debug_small.txt -NUM_DOCS: 15 - WEIGHT_SUM: 15.0 -DFL_STEM: convergence_debug_small -DATA_CACHE_PATH: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --------------------- -++++++++++++++++++++++++++++++++++++++++++++++++++ -- MPICH_DIR=/opt/cray/pe/mpich/8.1.25/ofi/gnu/9.1 -- Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 -- WORLD_SIZE:8 -- NCCL: nccl -- MODEL_TYPE: llama-seq4096-pp1-tp2-32layers-32heads-4096hidden -- Using DATA_FILE_LIST: ./convergence_debug_small.txt -++++++++++++++++++++++++++++++++++++++++++++++++++ -! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed -[2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda (auto detect) --------------------------------------------------- -DeepSpeed C++/CUDA extension op report --------------------------------------------------- -NOTE: Ops not installed will be just-in-time (JIT) compiled at - runtime if needed. Op compatibility means that your system - meet the required dependencies to JIT install the op. --------------------------------------------------- -JIT compiled ops requires ninja -ninja .................. [OKAY] --------------------------------------------------- -op name ................ installed .. compatible --------------------------------------------------- -async_io ............... [NO] ....... [OKAY] -fused_adam ............. [NO] ....... [OKAY] -cpu_adam ............... [NO] ....... [OKAY] -cpu_adagrad ............ [NO] ....... [OKAY] -cpu_lion ............... [NO] ....... [OKAY] - [WARNING] Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH -evoformer_attn ......... [NO] ....... [NO] -fused_lamb ............. [NO] ....... [OKAY] -fused_lion ............. [NO] ....... [OKAY] -inference_core_ops ..... [NO] ....... [OKAY] -cutlass_ops ............ [NO] ....... [OKAY] -transformer_inference .. [NO] ....... [OKAY] -quantizer .............. [NO] ....... [OKAY] -ragged_device_ops ...... [NO] ....... [OKAY] -ragged_ops ............. [NO] ....... [OKAY] -random_ltd ............. [NO] ....... [OKAY] - [WARNING] sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.2 - [WARNING] using untested triton version (2.2.0), only 1.0.0 is known to be compatible -sparse_attn ............ [NO] ....... [NO] -spatial_inference ...... [NO] ....... [OKAY] -transformer ............ [NO] ....... [OKAY] -stochastic_transformer . [NO] ....... [OKAY] --------------------------------------------------- -DeepSpeed general environment info: -torch install path ............... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/torch'] -torch version .................... 2.2.1 -deepspeed install path ........... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/deepspeed'] -deepspeed info ................... 0.14.0, unknown, unknown -torch cuda version ............... 11.8 -torch hip version ................ None -nvcc version ..................... 11.8 -deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 -shared memory (/dev/shm) size .... 251.61 GB - - deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc -e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion - --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- -grads-in-fp32 --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/tensorboard --log-timers-to-tensorboard - --log-optimizer-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_ -pp1_tp2_bf16 --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend nccl --num-attention-heads 32 --s -ave-interval 200 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 8 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 2 - --global-batch-size 32 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-l -cf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokeniz -er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 - --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- -num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - -[!! NOTE] View output at: -logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - -# ... - -/eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin - creating memory view of numpy buffer... - > finished creating indexed dataset in 0.001280 seconds - number of documents: 1498927 - > dataset split: - train: - document indices in [0, 1498927) total of 1498927 documents - validation: - document indices in [1498927, 1498927) total of 0 documents - test: - document indices in [1498927, 1498927) total of 0 documents - > loading doc-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_doc_idx.npy - > loading sample-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_sample_idx.npy - > loading shuffle-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_shuffle_idx.npy - loaded indexed file in 0.004 seconds - total number of samples: 869423 - total number of epochs: 3 -> loading blendable dataset index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_index.npy -> loading blendable dataset sample index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_sample_index.npy -> size of blendable dataset: 10223415 samples -> finished creating GPT datasets ... -[after dataloaders are built] datetime: 2024-04-04 09:36:07 -done with setup ... -(min, max) time across ranks (ms): - model-and-optimizer-setup ......................: (4794.78, 4795.23) - train/valid/test-data-iterators-setup ..........: (589.69, 721.20) -training ... -[before the start of training step] datetime: 2024-04-04 09:36:07 -[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:539:forward] Activation Checkpointing Information -[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:540:forward] ----Partition Activations False, CPU CHECKPOINTING False -[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:541:forward] ----contiguous Memory Checkpointing False with 32 total layers -[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:543:forward] ----Synchronization False -[2024-04-04 09:36:07,407] [INFO] [checkpointing.py:544:forward] ----Profiling time in checkpointing False -[2024-04-04 09:36:28,429] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1626.54 | optimizer_gradients: 19.29 | optimizer_step: 419.48 -[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] -[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 11336.34 | bwd_microstep: 7134.73 | bwd_inner_microstep: 7090.02 | bwd_allreduce_microstep: 44.65 | step_microstep: 2564.02 -[2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 11336.33 | bwd: 7134.75 | bwd_inner: 7090.01 | bwd_allreduce: 44.66 | step: 2564.02 - iteration 1/ 317892 | consumed samples: 32 | consumed tokens: 131072 | elapsed time per iteration (ms): 21133.8 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.119983E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 1.514 | tokens per gpu per second(tgs): 775.250 | TFLOPs: 47.23 | -[Rank 1] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 35824.0 -[Rank 0] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 32994.0 -[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1605.55 | optimizer_gradients: 11.56 | optimizer_step: 50.92 -[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] -[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 -[2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 - iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | - -# ... -``` - -
- -
+4.
🚀 Launch: + + In this case, train a ~ 2B Model (with 10 layers), + for 1000 iterations using the data file list in: + + [`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) + + with a micro-batch-size of 2 (`MICRO_BATCH=2`), with the `torch.optim.AdamW` optimizer (`OPT=adamw`). + + **Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) + function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) + can be overridden dynamically at runtime using this technique. + + ```bash + PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh + ``` + +
[output]: + +
+ +
[Sunspot]: + + ```bash + # [09:07:32 AM] [foremans@x1921c0s0b0n0] ~/q/llm.devkit/Megatron-DeepSpeed  main !1 ?27 q4-drop 26s ✘ INT + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_llama_alcf.sh + source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh + Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... + UMD: agama-ci-devel-736.9 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-736.9 + Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" + + Please check the spelling or version number. Also try "module spider ..." + It is also possible your cache file is out-of-date; it may help to try: + $ module --ignore_cache load "gcc/12.1.0" + + Also make sure that all modulefiles written in TCL start with the string #%Module + + Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. + + Running on SunSpot !! + [python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 + Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env + Found ezpz! + /lus/gila/projects/Aurora_deployment/foremans/locations/sunspot/projects/saforem2/ezpz/src/ezpz/__init__.py + Has ezpz installed. Nothing to do. + Done with ezpz. + ┌─────────────────────────────────────────────────────────────────── + │ Writing PBS vars to /home/foremans/.pbsenv + │ HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 + │ NHOSTS: 2 + │ NGPU_PER_HOST: 12 GPUs per host + │ NGPUS: 24 GPUs total + └─────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Hosts]: + │ • [host:0] - x1921c0s0b0n0.hostmgmt2000.cm.americas.sgi.com + │ • [host:1] - x1921c0s1b0n0.hostmgmt2000.cm.americas.sgi.com + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [DIST INFO]: + │ • Loading job env from: /home/foremans/.pbsenv + │ • HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 + │ • NHOSTS: 2 + │ • NGPU_PER_HOST: 12 + │ • NGPUS (NHOSTS x NGPU_PER_HOST): 24 + │ • WORLD_SIZE: 24 + │ • DIST_LAUNCH: mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001 + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Launch]: + │ • Use: 'launch' (=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001) + │ to launch job + └────────────────────────────────────────────────────────────────── + DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json + ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Calling: setData() with ./convergence_debug_small.txt + -------------------- + Updated environment: + DATA_FILE_LIST: ./convergence_debug_small.txt + NUM_DOCS: 15 + WEIGHT_SUM: 15.0 + DFL_STEM: convergence_debug_small + DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache + -------------------- + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR= + - Using /home/foremans/miniconda3/envs/q4-drop/bin/python3 + - WORLD_SIZE:24 + - NCCL: nccl + - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./convergence_debug_small.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + ! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed + /home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html + __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') + /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you dont plan on using image function + ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch + vision` from source? + warn( + [2024-04-04 09:07:45,585] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) + [2024-04-04 09:07:45,818] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) + -------------------------------------------------- + DeepSpeed C++/CUDA extension op report + -------------------------------------------------- + NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. + -------------------------------------------------- + JIT compiled ops requires ninja + ninja .................. [OKAY] + -------------------------------------------------- + op name ................ installed .. compatible + -------------------------------------------------- + async_io ............... [NO] ....... [OKAY] + cpu_adagrad ............ [NO] ....... [OKAY] + cpu_adam ............... [NO] ....... [OKAY] + flash_attn ............. [NO] ....... [OKAY] + fused_adam ............. [NO] ....... [OKAY] + quantizer .............. [NO] ....... [OKAY] + transformer ............ [NO] ....... [OKAY] + transformer_inference .. [NO] ....... [OKAY] + utils .................. [NO] ....... [OKAY] + -------------------------------------------------- + DeepSpeed general environment info: + torch install path ............... ['/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch'] + torch version .................... 2.1.0a0+cxx11.abi + deepspeed install path ........... ['/lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/DeepSpeed/deepspeed'] + deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD + deepspeed wheel compiled w. ...... torch 2.1 + shared memory (/dev/shm) size .... 503.18 GB + + deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ + foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay + -style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 + --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/tensorboard --log-timers-to-tensorboard --log-optimizer + -states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 + --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend ccl --num-attention-heads 32 --save-interval 20 + 0 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 4 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 1 --global-bat + ch-size 96 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ + .cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/tokenizer.model --no-query- + key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z + ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 + _nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + + [!! NOTE] View output at: + logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + + # ... + + /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.010017 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_doc_idx.npy + > loading sample-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_sample_idx.npy + > loading shuffle-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_shuffle_idx.npy + loaded indexed file in 0.056 seconds + total number of samples: 2318461 + total number of epochs: 8 + > loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy + > loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy + /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. + + [after dataloaders are built] datetime: 2024-04-04 09:09:27 + done with setup ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (64818.18, 64858.22) + train/valid/test-data-iterators-setup ..........: (1968.10, 2288.56) + training ... + [before the start of training step] datetime: 2024-04-04 09:09:27 + [2024-04-04 09:09:27,718] [INFO] [checkpointing.py:540:forward] Activation Checkpointing Information + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:541:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:542:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:544:forward] ----Synchronization False + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:545:forward] ----Profiling time in checkpointing False + [2024-04-04 09:09:33][INFO][utils:145] - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. + [2024-04-04 09:09:33][INFO][utils:148] - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. + [2024-04-04 09:09:33][INFO][utils:160] - NumExpr defaulting to 8 threads. + ^[c[2024-04-04 09:09:53,311] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 884.11 | optimizer_gradients: 6.43 | optimizer_step: 23.44 + [2024-04-04 09:09:53,312] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 6567.68 | bwd_microstep: 17950.36 | bwd_inner_microstep: 17711.20 | bwd_allreduce_microstep: 239.11 | step_microstep: 1139.27 + [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 6567.66 | bwd: 17950.35 | bwd_inner: 17711.19 | bwd_allreduce: 239.11 | step: 1139.29 + [Rank 0] (after 1 iterations) memory (MB) | allocated: 18244.640625 | max allocated: 41299.50146484375 | reserved: 46764.0 | max reserved: 46764.0 + iteration 1/ 317892 | consumed samples: 96 | consumed tokens: 393216 | elapsed time per iteration (ms): 25849.1 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 1.117136E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.714 | tokens per gpu per second(tgs): 633.832 | TFLOPs: 38.61 | + [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 327.85 | optimizer_gradients: 6.26 | optimizer_step: 23.60 + [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 4022.74 | bwd_microstep: 15738.67 | bwd_inner_microstep: 15556.80 | bwd_allreduce_microstep: 181.82 | step_microstep: 371.01 + [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 + iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | + ``` + +
+ +
[Polaris]: + + ```bash + # [09:31:35 AM] [foremans@x3112c0s13b0n0] ~/pol/p/a/Megatron-DeepSpeed  main !4 ?24 cu118-pt221 ✘ INT + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_llama_alcf.sh + source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh + Running on Polaris !! + + [python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 + Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env + Found ezpz! + /lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/ezpz/src/ezpz/__init__.py + Has ezpz installed. Nothing to do. + Done with ezpz. + ┌─────────────────────────────────────────────────────────────────── + │ Writing PBS vars to /home/foremans/.pbsenv + │ HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + │ NHOSTS: 2 + │ NGPU_PER_HOST: 4 GPUs per host + │ NGPUS: 8 GPUs total + └─────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Hosts]: + │ • [host:0] - x3112c0s13b0n0.hsn.cm.polaris.alcf.anl.gov + │ • [host:1] - x3112c0s13b1n0.hsn.cm.polaris.alcf.anl.gov + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [DIST INFO]: + │ • Loading job env from: /home/foremans/.pbsenv + │ • HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + │ • NHOSTS: 2 + │ • NGPU_PER_HOST: 4 + │ • NGPUS (NHOSTS x NGPU_PER_HOST): 8 + │ • WORLD_SIZE: 8 + │ • DIST_LAUNCH: mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Launch]: + │ • Use: 'launch' (=mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov) + │ to launch job + └────────────────────────────────────────────────────────────────── + DS_CONFIG: ds_stage2_mb8_gb32_pp1_bf16.json + ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Calling: setData() with ./convergence_debug_small.txt + -------------------- + Updated environment: + DATA_FILE_LIST: ./convergence_debug_small.txt + NUM_DOCS: 15 + WEIGHT_SUM: 15.0 + DFL_STEM: convergence_debug_small + DATA_CACHE_PATH: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache + -------------------- + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR=/opt/cray/pe/mpich/8.1.25/ofi/gnu/9.1 + - Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 + - WORLD_SIZE:8 + - NCCL: nccl + - MODEL_TYPE: llama-seq4096-pp1-tp2-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./convergence_debug_small.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + ! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed + [2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda (auto detect) + -------------------------------------------------- + DeepSpeed C++/CUDA extension op report + -------------------------------------------------- + NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. + -------------------------------------------------- + JIT compiled ops requires ninja + ninja .................. [OKAY] + -------------------------------------------------- + op name ................ installed .. compatible + -------------------------------------------------- + async_io ............... [NO] ....... [OKAY] + fused_adam ............. [NO] ....... [OKAY] + cpu_adam ............... [NO] ....... [OKAY] + cpu_adagrad ............ [NO] ....... [OKAY] + cpu_lion ............... [NO] ....... [OKAY] + [WARNING] Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH + evoformer_attn ......... [NO] ....... [NO] + fused_lamb ............. [NO] ....... [OKAY] + fused_lion ............. [NO] ....... [OKAY] + inference_core_ops ..... [NO] ....... [OKAY] + cutlass_ops ............ [NO] ....... [OKAY] + transformer_inference .. [NO] ....... [OKAY] + quantizer .............. [NO] ....... [OKAY] + ragged_device_ops ...... [NO] ....... [OKAY] + ragged_ops ............. [NO] ....... [OKAY] + random_ltd ............. [NO] ....... [OKAY] + [WARNING] sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.2 + [WARNING] using untested triton version (2.2.0), only 1.0.0 is known to be compatible + sparse_attn ............ [NO] ....... [NO] + spatial_inference ...... [NO] ....... [OKAY] + transformer ............ [NO] ....... [OKAY] + stochastic_transformer . [NO] ....... [OKAY] + -------------------------------------------------- + DeepSpeed general environment info: + torch install path ............... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/torch'] + torch version .................... 2.2.1 + deepspeed install path ........... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/deepspeed'] + deepspeed info ................... 0.14.0, unknown, unknown + torch cuda version ............... 11.8 + torch hip version ................ None + nvcc version ..................... 11.8 + deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 + shared memory (/dev/shm) size .... 251.61 GB + + deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc + e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion + --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- + grads-in-fp32 --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/tensorboard --log-timers-to-tensorboard - + -log-optimizer-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_ + pp1_tp2_bf16 --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend nccl --num-attention-heads 32 --s + ave-interval 200 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 8 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 2 + --global-batch-size 32 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-l + cf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokeniz + er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 + --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- + num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + + [!! NOTE] View output at: + logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + + # ... + + /eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.001280 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_doc_idx.npy + > loading sample-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_sample_idx.npy + > loading shuffle-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_shuffle_idx.npy + loaded indexed file in 0.004 seconds + total number of samples: 869423 + total number of epochs: 3 + > loading blendable dataset index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_index.npy + > loading blendable dataset sample index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_sample_index.npy + > size of blendable dataset: 10223415 samples + > finished creating GPT datasets ... + [after dataloaders are built] datetime: 2024-04-04 09:36:07 + done with setup ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (4794.78, 4795.23) + train/valid/test-data-iterators-setup ..........: (589.69, 721.20) + training ... + [before the start of training step] datetime: 2024-04-04 09:36:07 + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:539:forward] Activation Checkpointing Information + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:540:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:541:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:543:forward] ----Synchronization False + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:544:forward] ----Profiling time in checkpointing False + [2024-04-04 09:36:28,429] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1626.54 | optimizer_gradients: 19.29 | optimizer_step: 419.48 + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 11336.34 | bwd_microstep: 7134.73 | bwd_inner_microstep: 7090.02 | bwd_allreduce_microstep: 44.65 | step_microstep: 2564.02 + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 11336.33 | bwd: 7134.75 | bwd_inner: 7090.01 | bwd_allreduce: 44.66 | step: 2564.02 + iteration 1/ 317892 | consumed samples: 32 | consumed tokens: 131072 | elapsed time per iteration (ms): 21133.8 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.119983E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 1.514 | tokens per gpu per second(tgs): 775.250 | TFLOPs: 47.23 | + [Rank 1] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 35824.0 + [Rank 0] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 32994.0 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1605.55 | optimizer_gradients: 11.56 | optimizer_step: 50.92 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 + iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | + + # ... + ``` + +
+ +
From 9208eae5a07919acf145d578afc0742b08912bd9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 14 Jun 2024 17:57:31 -0500 Subject: [PATCH 206/590] Update README.md --- ALCF/README.md | 99 +++++++++++++++++++++++++++++++++++--------------- 1 file changed, 69 insertions(+), 30 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 00a70dbd42f..9e98a9b6ac7 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -50,40 +50,73 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv ``` - -
[output] Sunspot @ ALCF: + -
[output]: - ```bash - # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed][main *1 31 ?32] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv - Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed - No conda_prefix or virtual_env found in environment... - Setting up conda... - Running on SunSpot !! +
- Due to MODULEPATH changes, the following have been reloaded: - 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 - 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 - 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 - 4) mpfr/4.2.0-w7v7yjv + -
[Polaris]: - The following have been reloaded with a version change: - 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 - 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 - UMD: agama-ci-devel-803.29 successfully loaded: - UMD: graphics-compute-runtime/agama-ci-devel-803.29 + ```bash + # [05:47:13 PM][foremans@x3001c0s13b1n0][/eagle/a/f/p/ar/Megatron-DeepSpeed-D/Megatron-DeepSpeed][sequence-parallel 5 ?9] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + Using WORKING_DIR: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + Running on Polaris !! + + Lmod is automatically replacing "nvhpc/23.9" with "gcc-native/12.3". + + + Lmod is automatically replacing "PrgEnv-nvhpc/8.5.0" with "PrgEnv-gnu/8.5.0". + + + Due to MODULEPATH changes, the following have been reloaded: + 1) cray-mpich/8.1.28 + + Found conda at: /soft/applications/conda/2024-04-29/mconda3 + No VIRTUAL_ENV found in environment! + - Trying to setup from /soft/applications/conda/2024-04-29/mconda3 + - Using VENV_DIR=/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 + - Found existing venv, activating from /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 + [python] Using: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29/bin/python3 + ``` + +
+ + -
[Sunspot]: - The following have been reloaded with a version change: - 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 - - Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 - No VIRTUAL_ENV found in environment! - - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 - - Using VENV_DIR=/gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 - - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 - [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 - ``` - -
+ ```bash + # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed][main *1 31 ?32] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + Running on SunSpot !! + + Due to MODULEPATH changes, the following have been reloaded: + 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 + 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 + 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 + 4) mpfr/4.2.0-w7v7yjv + + The following have been reloaded with a version change: + 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 + 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 + UMD: agama-ci-devel-803.29 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-803.29 + + The following have been reloaded with a version change: + 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 + + Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + No VIRTUAL_ENV found in environment! + - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + - Using VENV_DIR=/gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 + ``` + +
- ## 🏃‍♂️ Running -To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): +To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): 1.
⏳ Request an interactive job with qsub -I: @@ -41,21 +28,21 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): 3.
🐍 Setup Python:
- + > **NOTE**: The following commands should be ran from [`Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed), following the `cd` command from 2. 1. Load `conda` module and activate base environment: - + ```bash PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv ``` -
[output]: - +
- + -
[Polaris]: - + ```bash # [05:47:13 PM][foremans@x3001c0s13b1n0][/eagle/a/f/p/ar/Megatron-DeepSpeed-D/Megatron-DeepSpeed] $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv @@ -63,16 +50,16 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): No conda_prefix or virtual_env found in environment... Setting up conda... Running on Polaris !! - + Lmod is automatically replacing "nvhpc/23.9" with "gcc-native/12.3". - - + + Lmod is automatically replacing "PrgEnv-nvhpc/8.5.0" with "PrgEnv-gnu/8.5.0". - - + + Due to MODULEPATH changes, the following have been reloaded: 1) cray-mpich/8.1.28 - + Found conda at: /soft/applications/conda/2024-04-29/mconda3 No VIRTUAL_ENV found in environment! - Trying to setup from /soft/applications/conda/2024-04-29/mconda3 @@ -80,21 +67,21 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): - Found existing venv, activating from /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 [python] Using: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29/bin/python3 ``` - +
-
[Aurora]: - + ```bash # [10:04:02 PM][foremans@x4415c0s2b0n0][/gecko/A/fo/p/a/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed No conda_prefix or virtual_env found in environment... Setting up conda... - + The following have been reloaded with a version change: 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/803.29 2) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.1 - + Found conda at: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 No VIRTUAL_ENV found in environment! - Trying to setup from /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 @@ -104,9 +91,9 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): ```
- + -
[Sunspot]: - + ```bash # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed] $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv @@ -114,22 +101,22 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): No conda_prefix or virtual_env found in environment... Setting up conda... Running on SunSpot !! - + Due to MODULEPATH changes, the following have been reloaded: 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 4) mpfr/4.2.0-w7v7yjv - + The following have been reloaded with a version change: 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 UMD: agama-ci-devel-803.29 successfully loaded: UMD: graphics-compute-runtime/agama-ci-devel-803.29 - + The following have been reloaded with a version change: 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 - + Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 No VIRTUAL_ENV found in environment! - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 @@ -137,19 +124,19 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 ``` - +
- + 2. 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): - + ```bash mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz python3 -m pip install -e deps/ezpz --require-virtualenv @@ -165,7 +152,7 @@ To launch on {`Polaris`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): 3. Setup [`wandb`](https://docs.wandb.ai/quickstart) > **NOTE**: this can be disabled by setting `export WANDB_DISABLED=1` - +
4.
🚀 Launch: - + In this case, train a ~ 2B Model (with 10 layers), for 1000 iterations using the data file list in: - + [`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) - + with a micro-batch-size of 2 (`MICRO_BATCH=2`), with the `torch.optim.AdamW` optimizer (`OPT=adamw`). - + **Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) can be overridden dynamically at runtime using this technique. - + ```bash PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh ``` - +
[output]: - +
The outputs should look _something_ like this, though YMMV (things change quick): - +
[Sunspot]: - + ```bash # [09:07:32 AM] [foremans@x1921c0s0b0n0] ~/q/llm.devkit/Megatron-DeepSpeed  main !1 ?27 q4-drop 26s ✘ INT $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_llama_alcf.sh @@ -223,15 +210,15 @@ will (1.) UMD: agama-ci-devel-736.9 successfully loaded: UMD: graphics-compute-runtime/agama-ci-devel-736.9 Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" - + Please check the spelling or version number. Also try "module spider ..." It is also possible your cache file is out-of-date; it may help to try: $ module --ignore_cache load "gcc/12.1.0" - + Also make sure that all modulefiles written in TCL start with the string #%Module - + Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. - + Running on SunSpot !! [python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env @@ -325,7 +312,7 @@ will (1.) deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD deepspeed wheel compiled w. ...... torch 2.1 shared memory (/dev/shm) size .... 503.18 GB - + deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay -style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 @@ -338,12 +325,12 @@ will (1.) key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 _nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - + [!! NOTE] View output at: logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - + # ... - + /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin creating memory view of numpy buffer... > finished creating indexed dataset in 0.010017 seconds @@ -364,7 +351,7 @@ will (1.) > loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy > loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. - + [after dataloaders are built] datetime: 2024-04-04 09:09:27 done with setup ... (min, max) time across ranks (ms): @@ -392,17 +379,17 @@ will (1.) [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | ``` - +
- +
[Polaris]: - + ```bash # [09:31:35 AM] [foremans@x3112c0s13b0n0] ~/pol/p/a/Megatron-DeepSpeed  main !4 ?24 cu118-pt221 ✘ INT $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_llama_alcf.sh source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh Running on Polaris !! - + [python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env Found ezpz! @@ -440,7 +427,7 @@ will (1.) ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - Calling: setData() with ./convergence_debug_small.txt + Calling: setData() with "./convergence_debug_small.txt" -------------------- Updated environment: DATA_FILE_LIST: ./convergence_debug_small.txt @@ -458,7 +445,7 @@ will (1.) - Using DATA_FILE_LIST: ./convergence_debug_small.txt ++++++++++++++++++++++++++++++++++++++++++++++++++ ! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed - [2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda (auto detect) + [2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda [auto detect] -------------------------------------------------- DeepSpeed C++/CUDA extension op report -------------------------------------------------- @@ -504,7 +491,7 @@ will (1.) nvcc version ..................... 11.8 deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 shared memory (/dev/shm) size .... 251.61 GB - + deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- @@ -517,12 +504,12 @@ will (1.) er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - + [!! NOTE] View output at: logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - + # ... - + /eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin creating memory view of numpy buffer... > finished creating indexed dataset in 0.001280 seconds @@ -568,12 +555,12 @@ will (1.) [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | - + # ... ``` - +
- +
@@ -583,17 +570,17 @@ will (1.) [^example]: | In this case, train a ~ 2B Model (with 10 layers), for 1000 iterations using the data file list in: - + [`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) - + with a micro-batch-size of 2, with the `torch.optim.AdamW` optimizer. Note that _any_ of the options in the - + [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) - + function from - + [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) - + can be overridden dynamically at runtime using this technique. --> @@ -616,7 +603,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books. $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed $ cd Megatron-DeepSpeed ``` - + > [!NOTE] > In the `conda create` command below, > you can replace `--name "${DAY}"` with @@ -631,7 +618,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books. $ export PYTHONUSERBASE="${HOME}/.local/polaris/conda/${DAY}" $ conda create --solver libmamba -c pytorch -c nvidia --name "${DAY}" "python==3.12" ``` - + 3. Install dependencies: ```bash @@ -641,7 +628,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books. $ python3 -m pip install --upgrade pip pybind11 toolong appdirs wandb sentencepiece ipython setuptools wheel ninja $ python3 -m pip install --upgrade deepspeed wandb ``` - + - [`ezpz`](https://github.com/saforem2/ezpz):
install: @@ -654,7 +641,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books.
- [**OPTIONAL**] [`NVIDIA/apex`](https://github.com/NVIDIA/apex): - +
install: ```bash @@ -664,7 +651,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books. $ module swap gcc gcc/10.3.0 $ python3 -m pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./ ``` - +
@@ -799,11 +786,19 @@ modules and launch --> +### 🚀 Submit as a batch job + +```bash +$ cd Megatron-DeepSpeed +$ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home train_llama_alcf.sh +``` + + ## 📝 Data Preprocessing
Data Pre-Processing: - + AuroraGPT is trained on the Dolma dataset (initially v0), now in the process of moving to v6. For more details on the dataset, refer to https://huggingface.co/datasets/allenai/dolma. The dolma dataset downloaded is already preprocessing to remove the duplicates (dedup) and filtering the data (mixing). For more details refer to https://github.com/allenai/dolma/tree/main/docs and https://github.com/vksastry/dolma_alcf/blob/main/ALCF/Readme.md. The data preprocessing of Dolma dataset before training consists of tokenization of the data using a specific tokenizer (LlamaTokenizer is what we are currently using), Use the below script to tokenize the entire dataset. Example shown for Polaris. @@ -822,7 +817,7 @@ cd /eagle/datasets/dolma/utils - [ ] Ensure / double check that optimizer settings from `ds_config.json` aren't being overwritten by some defaults in `megatron/arguments.py` - [ ] specifically, `momentum, beta{1, 2}, etc` - +
✅ Completed - Continue runs on Polaris @ @@ -834,12 +829,12 @@ cd /eagle/datasets/dolma/utils - [x] Then, try re-creating ( / fixing) conda with `cuda==12.1` - 😔, failed. - + - ~~‼️ Unable to save checkpoints with `torch==2.1` + `cuda==11.8`~~: - Fixed in [a57a21f](https://github.com/argonne-lcf/Megatron-DeepSpeed/commit/a57a21f6b2a8abf847f5ef599e1b1edcb5a5e1b5)
🐛 Bug - + - Training progresses OK: ```bash @@ -888,4 +883,3 @@ cd /eagle/datasets/dolma/utils
- diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d53c4cd6bf7..9fdbb08233f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,26 +1,42 @@ #!/bin/bash --login + +############################################################################### +# `ALCF/helpers.sh` +# # # Contains helper functions for launching `../train_llama_alcf.sh` # -# Example for interactive use: +# > [!NOTE] +# > On any of {Polaris, Aurora, Sunspot} @ ALCF: +# > +# > ```bash +# > $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# > $ cd Megatron-DeepSpeed +# > $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh +# > # on any of {Polaris, Sunspot, Aurora} @ ALCF +# > $ setup_python +# > ``` +############################################################################## + +#################################################################### +# `main`: +# +# This will get called automatically when # # ```bash -# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -# $ cd Megatron-DeepSpeed -# $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh -# $ setEnv # will setup conda + virtual environment +# $ source ALCF/helpers.sh # ``` - -# for debug mode, uncomment below: -# set -euxo pipefail - -#################################################################### -# `WORKING_DIR`: +# +# This will: # 1. if `${PBS_O_WORKDIR}` has nonzero value, use this # 2. else, if `${SLURM_SUBMIT_DIR}` has nonzero value, use that # 3. else, use `$(pwd)` #################################################################### -main() { +helpers_main() { + # for debug mode, run with `DEBUG=1` + if [[ -n "${DEBUG:-}" ]]; then + set -euxo + fi if [[ -n "${PBS_O_WORKDIR}" ]]; then WORKING_DIR="${PBS_O_WORKDIR}" elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then @@ -63,7 +79,7 @@ setup() { # 1. Identify machine we're on get_machine || exit # 2. Load `conda` environment - setEnv || exit + setup_python || exit # 3. Determine WORLD_SIZE, etc. from `PBS_*` vars setup_ezpz || exit # 4. Set command line arguments to pass to `"${EXEC}"` @@ -196,13 +212,12 @@ save_dotenv() { fi } - -where_am_i() { - if [[ $(hostname) == x4* ]]; then +get_machine_name() { + if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then machine="aurora" - elif [[ $(hostname) == x1* ]]; then + elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then machine="sunspot" - elif [[ $(hostname) == x3* ]]; then + elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then if [[ "${PBS_O_HOST}" == sirius* ]]; then machine="sirius" else @@ -377,7 +392,7 @@ set_lr_args() { ######################################################################### get_batch_size_on_polaris() { if [[ $(hostname) == x3* ]]; then - nhosts=$(wc -l < "${PBS_NODEFILE}") + nhosts=$(wc -l < "${HOSTFILE:-${PBS_NODEFILE}}") if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then mbs=1 elif [[ "${nhosts}" -ge 3 ]]; then @@ -571,7 +586,7 @@ ezpz_getjobenv() { # | 2. Install ezpz (if not installed) | # +---------------------------------------+ setup_ezpz() { - [ -n "${PBS_NODEFILE}" ] && ezpz_savejobenv || ezpz_getjobenv + [ -n "${HOSTFILE:-${PBS_NODEFILE}}" ] && ezpz_savejobenv || ezpz_getjobenv python3 -m ezpz.jobs && source "./.jobenv" if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" @@ -791,10 +806,8 @@ check_executable() { fi } - - ############################################################################## -# `setEnv`: +# `setup_python`: # # 1. Setup `conda` # - if `conda` nonempty, and `venv` empty, use `conda` to setup `venv`. @@ -814,38 +827,27 @@ check_executable() { # # 3. Print info about which python we're using ############################################################################## -setEnv() { - local virtual_env="${VIRTUAL_ENV:-}" - local conda_prefix="${CONDA_PREFIX:-}" - if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then - echo "No virtual environment found." - echo "Using conda from: ${conda_prefix}" - echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" - setup_venv_from_conda - elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then - echo "No conda found." - echo "Using virtual_env from: ${virtual_env}" - elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then - echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" - elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then - echo "No conda_prefix or virtual_env found in environment..." +setup_python() { + virtual_env="${VIRTUAL_ENV:-}" + conda_prefix="${CONDA_PREFIX:-}" + if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No conda_prefix OR virtual_env found in environment..." echo "Setting up conda..." - ######################## setup_conda ############################ - # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- - if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then - if [[ $(hostname) == x4* ]]; then - setup_conda_aurora - elif [[ $(hostname) == x1* ]]; then - setup_conda_sunspot - fi - # ----- [Polaris ( / Sirius ) @ ALCF] ---------------------------- - elif [[ $(hostname) == x3* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then + local machine_name=$(get_machine_name) + echo "machine name: ${machine_name}" + # if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then + if [[ "${machine_name}" == "aurora" ]]; then + setup_conda_aurora + # elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then + elif [[ "${machine_name}" == "sunspot" ]]; then + setup_conda_sunspot + # elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then + elif [[ "${machine_name}" == "polaris" ]]; then + if [[ "${PBS_O_HOST:-}" == sirius* ]]; then setup_conda_sirius else setup_conda_polaris fi - # ----- [Perlmutter @ NERSC] ------------------------------------- elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then echo "Running on Perlmutter !!" module load pytorch @@ -854,6 +856,17 @@ setEnv() { echo "Unknown hostname $(hostname)" exit 1 fi + # # ----- [Perlmutter @ NERSC] ------------------------------------- + elif [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then + echo "No virtual environment found." + echo "Using conda from: ${conda_prefix}" + echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" + setup_venv_from_conda + elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then + echo "No conda found." + echo "Using virtual_env from: ${virtual_env}" + elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then + echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" else echo "Unable to setup python environment. Exiting" exit 1 @@ -862,11 +875,86 @@ setEnv() { setup_venv_from_conda fi pystr="Using: $(which python3)" - printf "[python] %s" "$(printMagenta ${pystr})" + printf "[python] %s" "$(printMagenta "${pystr}")" printf "\n" export "PYTHON_EXEC=$(which python3)" } +# +# ############################################################################## +# # `setEnv`: +# # +# # 1. Setup `conda` +# # - if `conda` nonempty, and `venv` empty, use `conda` to setup `venv`. +# # - if `venv` nonempty, and `conda` empty, what do (???) +# # - if `venv` nonempty and `conda` nonempty, use these +# # - if `conda` empty and `venv` empty: +# # - if `hostname == x4*`, we're on Aurora +# # - if `hostname == x1*`, we're on Sunspot +# # - if `hostname == x3*`, we're on Polaris +# # - if `hostname == nid*`, we're on Perlmutter +# # - otherwise, you're on you're own +# # +# # 2. Activate (creating, if necessary) a `venv` on top of `base` conda +# # - use the $CONDA_PREFIX to create a venv in +# # `Megatron-DeepSpeed/venvs/${CONDA_PREFIX}` +# # - activate and use this +# # +# # 3. Print info about which python we're using +# ############################################################################## +# setEnv() { +# local virtual_env="${VIRTUAL_ENV:-}" +# local conda_prefix="${CONDA_PREFIX:-}" +# if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then +# echo "No virtual environment found." +# echo "Using conda from: ${conda_prefix}" +# echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" +# setup_venv_from_conda +# elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then +# echo "No conda found." +# echo "Using virtual_env from: ${virtual_env}" +# elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then +# echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" +# elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then +# echo "No conda_prefix or virtual_env found in environment..." +# echo "Setting up conda..." +# ######################## setup_conda ############################ +# # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- +# if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then +# if [[ $(hostname) == x4* ]]; then +# setup_conda_aurora +# elif [[ $(hostname) == x1* ]]; then +# setup_conda_sunspot +# fi +# # ----- [Polaris ( / Sirius ) @ ALCF] ---------------------------- +# elif [[ $(hostname) == x3* ]]; then +# if [[ "${PBS_O_HOST}" == sirius* ]]; then +# setup_conda_sirius +# else +# setup_conda_polaris +# fi +# # ----- [Perlmutter @ NERSC] ------------------------------------- +# elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then +# echo "Running on Perlmutter !!" +# module load pytorch +# source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" +# else # ------------------------------------- [Unknown] ------------------- +# echo "Unknown hostname $(hostname)" +# exit 1 +# fi +# else +# echo "Unable to setup python environment. Exiting" +# exit 1 +# fi +# if [[ -z "${virtual_env}" ]]; then +# setup_venv_from_conda +# fi +# pystr="Using: $(which python3)" +# printf "[python] %s" "$(printMagenta ${pystr})" +# printf "\n" +# export "PYTHON_EXEC=$(which python3)" +# } +# ###################################################################### # `makeHostiles`: @@ -900,11 +988,13 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" - local machine=$(where_am_i) + local machine=$(get_machine_name) if [[ ${machine} == "polaris" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + elif [[ ${machine} == "aurora" ]]; then + export DATA_PARENT="${DATA_PARENT:-/gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" else export DATA_PARENT="${DATA_PARENT:-${WORKING_DIR}/dataset}" fi @@ -1180,7 +1270,7 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } -main +helpers_main #### [DEPRECATED] ########################################################### # if [[ -z "${HOSTFILE}" ]]; then diff --git a/train_agpt.sh b/train_agpt.sh deleted file mode 100644 index a21b9880734..00000000000 --- a/train_agpt.sh +++ /dev/null @@ -1,12 +0,0 @@ -#!/bin/bash --login - -NOW="$(date "+%Y-%m-%d-%H%M%S")" -cd "${PBS_O_WORKDIR}" || exit - -OUTDIR="${PBS_O_WORKDIR}/pbslogs" -mkdir -p "${OUTDIR}" -OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" -echo "${OUTFILE}" >> "${OUTDIR}/latest" -echo "Logging job output to: ${OUTFILE}" -# export DEBUG=1 -bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" From e05842700566539d06d116e8ea5056c9d310dcf2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 18 Jun 2024 14:44:17 -0500 Subject: [PATCH 217/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 124 ++++++------------------------------------------ 1 file changed, 14 insertions(+), 110 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 9fdbb08233f..8da6cb0ffc5 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -18,8 +18,8 @@ # > ``` ############################################################################## -#################################################################### -# `main`: +############################################################################## +# `helpers_main`: # # This will get called automatically when # @@ -27,11 +27,15 @@ # $ source ALCF/helpers.sh # ``` # -# This will: -# 1. if `${PBS_O_WORKDIR}` has nonzero value, use this -# 2. else, if `${SLURM_SUBMIT_DIR}` has nonzero value, use that -# 3. else, use `$(pwd)` -#################################################################### +# - This will set `"${WORKING_DIR}"`, according to: +# 1. `${PBS_O_WORKDIR}` is nonzero, use this +# 2. else, if `${SLURM_SUBMIT_DIR}` is nonzero use this +# 3. else, use `$(pwd)` +# +# this is crucial since many of the functions below use paths +# which are defined relative to this "${WORKING_DIR}" +# (e.g. virtual environment, location of executables, etc.) +############################################################################## helpers_main() { # for debug mode, run with `DEBUG=1` if [[ -n "${DEBUG:-}" ]]; then @@ -46,7 +50,6 @@ helpers_main() { WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') echo "Using ${WORKING_DIR} as working directory..." fi - export WORKING_DIR="${WORKING_DIR}" printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" } @@ -148,7 +151,6 @@ setup_run_cmd() { if [[ "${SP}" -ge 2 ]]; then export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" fi - # [hacky]: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` if [[ -z "${NO_LLAMA:-}" ]]; then @@ -160,7 +162,6 @@ setup_run_cmd() { echo "!! Running in NO_LLAMA MODE !!" llama_flags="" fi - export run_cmd=" ${LAUNCHER} \ --${DTYPE} \ @@ -200,7 +201,7 @@ save_dotenv() { if [[ "$#" -ne 1 ]]; then estr="[error]" # echo "Expected exactly one argument, specifying outputdir. Received $#" - printf "%s Expected one argument (outdir). Received: %s" "$(printRed ${estr})" "$#" + printf "%s Expected one argument (outdir). Received: %s" "$(printRed "${estr}")" "$#" else outdir="$1" mkdir -p "${outdir}" @@ -880,82 +881,6 @@ setup_python() { export "PYTHON_EXEC=$(which python3)" } -# -# ############################################################################## -# # `setEnv`: -# # -# # 1. Setup `conda` -# # - if `conda` nonempty, and `venv` empty, use `conda` to setup `venv`. -# # - if `venv` nonempty, and `conda` empty, what do (???) -# # - if `venv` nonempty and `conda` nonempty, use these -# # - if `conda` empty and `venv` empty: -# # - if `hostname == x4*`, we're on Aurora -# # - if `hostname == x1*`, we're on Sunspot -# # - if `hostname == x3*`, we're on Polaris -# # - if `hostname == nid*`, we're on Perlmutter -# # - otherwise, you're on you're own -# # -# # 2. Activate (creating, if necessary) a `venv` on top of `base` conda -# # - use the $CONDA_PREFIX to create a venv in -# # `Megatron-DeepSpeed/venvs/${CONDA_PREFIX}` -# # - activate and use this -# # -# # 3. Print info about which python we're using -# ############################################################################## -# setEnv() { -# local virtual_env="${VIRTUAL_ENV:-}" -# local conda_prefix="${CONDA_PREFIX:-}" -# if [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then -# echo "No virtual environment found." -# echo "Using conda from: ${conda_prefix}" -# echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" -# setup_venv_from_conda -# elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then -# echo "No conda found." -# echo "Using virtual_env from: ${virtual_env}" -# elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then -# echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" -# elif [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then -# echo "No conda_prefix or virtual_env found in environment..." -# echo "Setting up conda..." -# ######################## setup_conda ############################ -# # ---- [SunSpot @ ALCF] || [Aurora @ ALCF] --------------------- -# if [[ $(hostname) == x1* || $(hostname) == x4* ]]; then -# if [[ $(hostname) == x4* ]]; then -# setup_conda_aurora -# elif [[ $(hostname) == x1* ]]; then -# setup_conda_sunspot -# fi -# # ----- [Polaris ( / Sirius ) @ ALCF] ---------------------------- -# elif [[ $(hostname) == x3* ]]; then -# if [[ "${PBS_O_HOST}" == sirius* ]]; then -# setup_conda_sirius -# else -# setup_conda_polaris -# fi -# # ----- [Perlmutter @ NERSC] ------------------------------------- -# elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then -# echo "Running on Perlmutter !!" -# module load pytorch -# source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" -# else # ------------------------------------- [Unknown] ------------------- -# echo "Unknown hostname $(hostname)" -# exit 1 -# fi -# else -# echo "Unable to setup python environment. Exiting" -# exit 1 -# fi -# if [[ -z "${virtual_env}" ]]; then -# setup_venv_from_conda -# fi -# pystr="Using: $(which python3)" -# printf "[python] %s" "$(printMagenta ${pystr})" -# printf "\n" -# export "PYTHON_EXEC=$(which python3)" -# } -# - ###################################################################### # `makeHostiles`: # Detect if `HOSTFILE` set in active environment. @@ -1017,7 +942,6 @@ setup_tokenizer_and_data() { printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" } - ############################################### # `setData`: # Ensure `DATA_FILE_LIST` is set, @@ -1040,7 +964,6 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" - else # -----------------------------------------------------------[UNKNOWN] echo "Unknown hostname. Must manually specify DATA_FILE_LIST." fi @@ -1270,25 +1193,6 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } +################ +# call helpers_main() helpers_main - -#### [DEPRECATED] ########################################################### -# if [[ -z "${HOSTFILE}" ]]; then -# makeHostfiles || exit # 4. create `deepspeed` hostfile from `$PBS_NODEFILE` -# else -# echo "!! USING CUSTOM HOSTFILE FROM: ${HOSTFILE}" -# fi -# ---------------------------------------------------------------------------- -# setDSlauncher "${HERE}" || exit # 10. set `launcher` args for `deepspeed ${launcher} ${EXEC} ${args}` -# ---------------------------------------------------------------------------- -# TORCH_DEVICE=$(python3 -c 'import ezpz as ez; print(ez.get_torch_device())') -# printf %s "Using TORCH_DEVICE=${TORCH_DEVICE}" -# if [[ "${TORCH_DEVICE}" == "cuda" ]]; then -# printf %s "Setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True" -# PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True -# fi -# ---------------------------------------------------------------------------- -# export MPICH_GPU_SUPPORT_ENABLED=1 -# export CUDA_DEVICE_MAX_CONNECTIONS=1 -# export NCCL_DEBUG=INFO -############################################################################# From 1ae3768e3431728f550a1f4cdeb12412c39020e6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 09:16:28 -0500 Subject: [PATCH 218/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 108 +++++++++++++++++++----------------------------- 1 file changed, 43 insertions(+), 65 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8da6cb0ffc5..09e59784344 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -249,7 +249,7 @@ get_machine() { echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." fi export MACHINE="${machine}" - printf "Running on: %s\n" "$(printBlue ${MACHINE})" + printf "Running on: %s\n" "$(printBlue "${MACHINE}")" } @@ -287,32 +287,6 @@ printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } -setupVenv() { - VENV_DIR="$1" - if [[ -d "${VENV_DIR}" ]]; then - echo "Found venv at: ${VENV_DIR}" - source "${VENV_DIR}/bin/activate" - else - echo "Skipping setupVenv() on $(hostname)" - fi -} - -loadCondaEnv() { - if [[ "${CONDA_EXE}" ]]; then - echo "Already inside ${CONDA_EXE}, exiting!" - else - MODULE_STR="$1" - module load "conda/${MODULE_STR}" - nargs="$#" - if [[ "${nargs}" -ge 2 ]]; then - conda activate "$2" - else - conda activate base - fi - fi -} - - ############################################################################# # `setupLauncher`: Launch with one of `{mpiexec, deepspeed}`. # @@ -340,7 +314,8 @@ setupLauncher() { # if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then else if [[ -n "${DIST_LAUNCH}" ]]; then - export LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + export LAUNCHER="${LAUNCHER}" else echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" printenv | tee ".env-${PBS_JOBID}" @@ -349,7 +324,7 @@ setupLauncher() { fi printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" - printf " %s" "$(printMagenta ${LAUNCHER})" + printf " %s" "$(printMagenta "${LAUNCHER}")" } setDSlauncher() { @@ -500,7 +475,7 @@ setParams() { export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? - export GLOBAL_BATCH_MAX=$(( $WORLD_SIZE * $MICRO_BATCH * $GRAD_ACC_STEPS / $TP / $PP / $SP )) # MAX GLOBAL BATCH SIZE + export GLOBAL_BATCH_MAX=$(( WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ @@ -573,13 +548,15 @@ make_ds_hostfile() { } ezpz_savejobenv() { - local outfile="${WORKING_DIR}/bin/savejobenv" ; mkdir -p $(dirname "${outfile}") - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${outfile}" && source "${outfile}" + file=$(mktemp) + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${file}" + source "${file}" || exit } ezpz_getjobenv() { - local outfile="${WORKING_DIR}/bin/getjobenv" ; mkdir -p $(dirname "${outfile}") - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${outfile}" && source "${outfile}" + file=$(mktemp) + curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" + source "${file}" || exit } # +---------------------------------------+ @@ -587,7 +564,11 @@ ezpz_getjobenv() { # | 2. Install ezpz (if not installed) | # +---------------------------------------+ setup_ezpz() { - [ -n "${HOSTFILE:-${PBS_NODEFILE}}" ] && ezpz_savejobenv || ezpz_getjobenv + if [[ -n "${HOSTFILE:-${PBS_NODEFILE}}" ]]; then + ezpz_savejobenv + else + ezpz_getjobenv + fi python3 -m ezpz.jobs && source "./.jobenv" if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then mkdir -p "${WORKING_DIR}/deps" @@ -611,13 +592,13 @@ setup_ezpz() { # `ezpz_test`: Run simple test to make sure all nodes in working order ####################################################################### ezpz_test() { - printf "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" + printf "%s" "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" # [ -n "${PBS_O_WORKIR}" ] && ezpz_savejobenv || ezpz_getjobenv # python3 -Wignore -m ezpz.jobs && source "${PBS_O_WORKDIR}/.jobenv" - printf "[$(printBlue 'ezpz:test_dist')] Running test: ${eztest}\n" + printf "%s" "[$(printBlue 'ezpz:test_dist')] Running test: ${eztest}\n" eztest="TRAIN_ITERS=50 ${LAUNCH_CMD} python3 -Wignore -m ezpz.test_dist" eval "${eztest}" - printf "[$(printBlue 'ezpz:test_dist')] Done with test!\n" + printf "%s" "[$(printBlue 'ezpz:test_dist')] Done with test!\n" } # +------------------------------------------------------------------------+ @@ -629,8 +610,8 @@ saveDSenv() { { echo "PATH=${PATH}" ; echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}" ; - echo "http_proxy=${http_proxy}" ; - echo "https_proxy=${https_proxy}" ; + echo "http_proxy=${http_proxy:-}" ; + echo "https_proxy=${https_proxy:-}" ; echo "CFLAGS=${CFLAGS}" ; echo "PYTHONUSERBASE=$PYTHONUSERBASE" ; } > .deepspeed_env @@ -665,21 +646,20 @@ setOutput() { export OUTPUT_LOG="${OUTPUT_DIR}/output.log" export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" echo "${OUTPUT_LOG}" >> "logs/latest" - printf "\n Please see logs at: %s\n" $(printGreen "${OUTPUT_DIR}") - printf "Checkpoints will be saved to: %s\n" $(printYellow "${CKPT_DIR}") + printf "\n Please see logs at: %s\n" "$(printGreen "${OUTPUT_DIR}")" + printf "Checkpoints will be saved to: %s\n" "$(printYellow "${CKPT_DIR}")" } +############################################# +# Build DeepSpeed config and write to .json +############################################# buildDSconfig() { - # ---- Build DeepSpeed Config --------------------------------- export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" - mkdir -p $(dirname "${DS_CONFIG}") + mkdir -p "$(dirname "${DS_CONFIG}")" echo "DS_CONFIG: ${DS_CONFIG}" - printf "ZS: %s, , MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${CPU_OPTIMIZER}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" - # working_dir="${PBS_O_WORKDIR:-${SLURM_SUBMIT_DIR:-$(pwd)}}" + printf "ZS: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" generateDSconfig "${DS_CONFIG}" - # bash "${WORKING_DIR}/ALCF/generate_ds_config.sh" "${DS_CONFIG}" - # ------------------------------------------------------------- } @@ -714,14 +694,20 @@ sumFiles() { setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ if [[ -z "${CONDA_PREFIX:-}" ]]; then + module use /soft/preview-modulefiles/24.086.0 + module load frameworks/2024.04.15.002.lua # module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua - source "${WORKING_DIR}/ALCF/sunspot-env-2024-q2.sh" + # source "${WORKING_DIR}/ALCF/sunspot-env-2024-q2.sh" fi } setup_conda_aurora() { if [[ -z "${CONDA_PREFIX:-}" ]]; then module use -a /soft/modulefiles ; module load frameworks/2024.1 + export CCL_KVS_MODE=mpi + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH fi } @@ -732,7 +718,7 @@ setup_conda_sirius() { if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell ${shell_name})" + eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell "${shell_name}")" micromamba activate 2024-04-23 else echo "Found existing python at: $(which python3)" @@ -756,23 +742,17 @@ setup_conda_polaris() { setup_venv_from_conda() { if [[ -z "${CONDA_PREFIX}" ]]; then - echo "No ${CONDA_PREFIX} found." # Exiting." + echo "!! No ${CONDA_PREFIX} found." # Exiting." # exit 1 else echo "Found conda at: ${CONDA_PREFIX}" - CONDA_NAME=$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') + CONDA_NAME=$(echo "${CONDA_PREFIX}" | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') export CONDA_NAME - # if [[ -n "${VIRTUAL_ENV}" ]]; then - # echo "Already inside virtual env at ${VENV_DIR}!" if [[ -z "${VIRTUAL_ENV}" ]]; then echo "No VIRTUAL_ENV found in environment!" echo " - Trying to setup from ${CONDA_PREFIX}" export VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" echo " - Using VENV_DIR=${VENV_DIR}" - # VENV_DIR="venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}')" - # VENV_DIR="${WORKING_DIR}/venvs/$(echo ${CONDA_PREFIX} | tr '\/' '\t' | awk '{print $NF}')" - # VENV_DIR="${WORKING_DIR}/venvs/anl_24_q2_release" - # if [[ -f "${VENV_DIR}/bin/activate" ]]; then if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then printf "\n - Creating a new virtual env on top of %s in %s" "$(printBlue "${CONDA_NAME}")" "$(printGreen "${VENV_DIR}")" mkdir -p "${VENV_DIR}" @@ -785,8 +765,6 @@ setup_venv_from_conda() { printf "\n [!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" fi fi - # else - # printf "[!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" fi } @@ -803,7 +781,7 @@ check_executable() { export EXEC_STEM="${exec_stem}" else estr="Unable to locate executable ${fp}" - printf "[ALCF.helpers:check_executable] %s" "$(printRed ${estr})" + printf "[ALCF.helpers:check_executable] %s" "$(printRed "${estr}")" fi } @@ -834,7 +812,7 @@ setup_python() { if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then echo "No conda_prefix OR virtual_env found in environment..." echo "Setting up conda..." - local machine_name=$(get_machine_name) + machine_name=$(get_machine_name) echo "machine name: ${machine_name}" # if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then if [[ "${machine_name}" == "aurora" ]]; then @@ -913,7 +891,7 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" - local machine=$(get_machine_name) + machine=$(get_machine_name) if [[ ${machine} == "polaris" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then @@ -938,8 +916,8 @@ setup_tokenizer_and_data() { setData "${dfl}" || exit fi fi - printf "[setData] DATA_FLAGS: %s\n" "$(printGreen ${DATA_FLAGS})" - printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" + printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" + printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" } ############################################### From abead32a1bfa77f26899d3391a68f29586ec55f5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 09:24:57 -0500 Subject: [PATCH 219/590] Update `ALCF/README.md` --- ALCF/README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/ALCF/README.md b/ALCF/README.md index 462adfd0340..2ec4e0ea7a3 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -14,6 +14,14 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I ``` + - Or, alternatively, you can submit [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) + directly as a batch script with + + ```bash + cd Megatron-DeepSpeed + qsub -A -q debug -l select=2 -l walltime=01:00:00:filesystems=eagle:home train_llama_alcf.sh + ``` +
2.
⬇️ Clone repo + navigate into it: From 025ff3f80225490651df2d2c2dbb7966ef1b0120 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 09:28:20 -0500 Subject: [PATCH 220/590] Update ALCF/README.md` --- ALCF/README.md | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/ALCF/README.md b/ALCF/README.md index 2ec4e0ea7a3..3870921f5b1 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -202,6 +202,15 @@ will (1.) PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh ``` + - **Note**: If no additional options specified, i.e. + + ```bash + PBS_O_WORKDIR=$(pwd) bash train_llama_alcf.sh + ``` + + then this will fallback to using the default AuroraGPT-7B architecture + with the full Dolma (v1.7) dataset. +
[output]:
From 732e567dbbdd62a706154cb4a9f759905d888b39 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:36:08 -0500 Subject: [PATCH 221/590] Add `ALCF/data-lists/aurora/*.txt` --- ALCF/data-lists/aurora/algebraic.txt | 32 +- ALCF/data-lists/aurora/arxiv.txt | 200 +- ALCF/data-lists/aurora/books.txt | 6 +- ALCF/data-lists/aurora/c4.txt | 342 +-- ALCF/data-lists/aurora/cc.txt | 2214 +++++++-------- ALCF/data-lists/aurora/dolma.txt | 2419 +++++++++++++++++ ALCF/data-lists/aurora/falcon.txt | 1000 +++---- ALCF/data-lists/aurora/megawiki.txt | 524 ++-- .../data-lists/aurora/open-web-math-train.txt | 26 +- ALCF/data-lists/aurora/pes2o.txt | 52 +- ALCF/data-lists/aurora/reddit.txt | 156 +- ALCF/data-lists/aurora/stack.txt | 52 +- ALCF/data-lists/aurora/starcoder.txt | 98 +- ALCF/data-lists/aurora/tulu.txt | 132 +- ALCF/data-lists/aurora/wiki.txt | 4 +- 15 files changed, 4838 insertions(+), 2419 deletions(-) create mode 100644 ALCF/data-lists/aurora/dolma.txt diff --git a/ALCF/data-lists/aurora/algebraic.txt b/ALCF/data-lists/aurora/algebraic.txt index febdf441fb4..dc676b1e839 100644 --- a/ALCF/data-lists/aurora/algebraic.txt +++ b/ALCF/data-lists/aurora/algebraic.txt @@ -1,16 +1,16 @@ -0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document -0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document -0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document -0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document -0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document -0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document -0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document -0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document -0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document -0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document -0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document -0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document -0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document -0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document -0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document -0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train diff --git a/ALCF/data-lists/aurora/arxiv.txt b/ALCF/data-lists/aurora/arxiv.txt index 3a822f5f44c..c6214709e8f 100644 --- a/ALCF/data-lists/aurora/arxiv.txt +++ b/ALCF/data-lists/aurora/arxiv.txt @@ -1,100 +1,100 @@ -0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document -0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document -0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document -0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document -0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document -0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document -0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document -0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document -0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document -0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document -0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document -0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document -0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document -0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document -0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document -0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document -0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document -0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document -0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document -0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document -0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document -0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document -0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document -0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document -0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document -0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document -0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document -0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document -0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document -0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document -0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document -0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document -0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document -0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document -0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document -0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document -0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document -0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document -0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document -0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document -0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document -0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document -0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document -0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document -0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document -0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document -0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document -0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document -0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document -0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document -0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document -0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document -0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document -0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document -0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document -0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document -0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document -0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document -0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document -0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document -0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document -0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document -0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document -0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document -0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document -0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document -0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document -0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document -0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document -0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document -0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document -0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document -0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document -0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document -0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document -0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document -0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document -0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document -0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document -0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document -0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document -0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document -0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document -0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document -0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document -0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document -0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document -0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document -0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document -0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document -0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document -0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document -0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document -0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document -0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document -0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document -0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document -0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document -0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document -0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv diff --git a/ALCF/data-lists/aurora/books.txt b/ALCF/data-lists/aurora/books.txt index 946013b617b..7d69ed32fd0 100644 --- a/ALCF/data-lists/aurora/books.txt +++ b/ALCF/data-lists/aurora/books.txt @@ -1,3 +1,3 @@ -0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document -0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document -0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books diff --git a/ALCF/data-lists/aurora/c4.txt b/ALCF/data-lists/aurora/c4.txt index 4bc8e358205..96c85cd3f4e 100644 --- a/ALCF/data-lists/aurora/c4.txt +++ b/ALCF/data-lists/aurora/c4.txt @@ -1,171 +1,171 @@ -0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document -0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document -0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document -0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document -0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document -0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document -0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document -0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document -0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document -0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document -0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document -0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document -0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document -0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document -0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document -0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document -0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document -0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document -0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document -0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document -0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document -0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document -0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document -0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document -0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document -0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document -0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document -0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document -0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document -0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document -0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document -0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document -0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document -0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document -0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document -0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document -0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document -0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document -0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document -0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document -0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document -0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document -0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document -0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document -0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document -0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document -0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document -0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document -0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document -0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document -0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document -0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document -0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document -0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document -0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document -0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document -0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document -0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document -0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document -0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document -0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document -0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document -0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document -0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document -0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document -0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document -0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document -0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document -0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document -0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document -0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document -0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document -0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document -0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document -0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document -0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document -0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document -0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document -0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document -0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document -0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document -0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document -0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document -0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document -0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document -0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document -0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document -0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document -0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document -0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document -0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document -0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document -0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document -0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document -0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document -0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document -0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document -0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document -0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document -0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document -0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document -0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document -0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document -0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document -0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document -0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document -0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document -0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document -0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document -0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document -0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document -0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document -0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document -0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document -0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document -0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document -0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document -0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document -0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document -0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document -0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document -0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document -0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document -0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document -0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document -0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document -0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document -0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document -0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document -0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document -0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document -0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document -0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document -0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document -0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document -0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document -0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document -0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document -0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document -0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document -0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document -0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document -0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document -0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document -0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document -0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document -0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document -0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document -0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document -0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document -0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document -0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document -0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document -0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document -0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document -0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document -0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document -0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document -0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document -0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document -0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document -0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document -0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document -0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document -0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document -0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document -0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document -0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document -0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document -0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document -7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 diff --git a/ALCF/data-lists/aurora/cc.txt b/ALCF/data-lists/aurora/cc.txt index 5d4fce7cff2..f3161a32ef5 100644 --- a/ALCF/data-lists/aurora/cc.txt +++ b/ALCF/data-lists/aurora/cc.txt @@ -1,1108 +1,1108 @@ -0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document -0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document -0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document -0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document -0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document -0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document -0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document -0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document -0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document -0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document -0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document -0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document -0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document -0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document -0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document -0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document -0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document -0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document -0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document -0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document -0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document -0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document -0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document -0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document -0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document -0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document -0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document -0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document -0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document -0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document -0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document -0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document -0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document -0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document -0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document -0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document -0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document -0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document -0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document -0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document -0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document -0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document -0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document -0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document -0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document -0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document -0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document -0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document -0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document -0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document -0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document -0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document -0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document -0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document -0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document -0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document -0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document -0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document -0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document -0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document -0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document -0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document -0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document -0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document -0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document -0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document -0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document -0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document -0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document -0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document -0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document -0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document -0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document -0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document -0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document -0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document -0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document -0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document -0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document -0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document -0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document -0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document -0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document -0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document -0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document -0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document -0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document -0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document -0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document -0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document -0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document -0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document -0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document -0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document -0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document -0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document -0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document -0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document -0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document -0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document -0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document -0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document -0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document -0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document -0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document -0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document -0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document -0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document -0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document -0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document -0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document -0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document -0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document -0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document -0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document -0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document -0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document -0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document -0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document -0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document -0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document -0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document -0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document -0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document -0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document -0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document -0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document -0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document -0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document -0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document -0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document -0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document -0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document -0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document -0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document -0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document -0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document -0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document -0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document -0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document -0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document -0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document -0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document -0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document -0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document -0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document -0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document -0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document -0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document -0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document -0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document -0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document -0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document -0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document -0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document -0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document -0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document -0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document -0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document -0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document -0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document -0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document -0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document -0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document -0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document -0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document -0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document -0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document -0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document -0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document -0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document -0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document -0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document -0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document -0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document -0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document -0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document -0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document -0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document -0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document -0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document -0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document -0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document -0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document -0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document -0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document -0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document -0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document -0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document -0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document -0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document -0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document -0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document -0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document -0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document -0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document -0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document -0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document -0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document -0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document -0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document -0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document -0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document -0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document -0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document -0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document -0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document -0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document -0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document -0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document -0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document -0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document -0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document -0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document -0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document -0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document -0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document -0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document -0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document -0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document -0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document -0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document -0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document -0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document -0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document -0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document -0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document -0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document -0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document -0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document -0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document -0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document -0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document -0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document -0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document -0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document -0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document -0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document -0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document -0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document -0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document -0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document -0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document -0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document -0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document -0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document -0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document -0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document -0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document -0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document -0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document -0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document -0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document -0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document -0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document -0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document -0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document -0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document -0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document -0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document -0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document -0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document -0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document -0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document -0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document -0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document -0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document -0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document -0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document -0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document -0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document -0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document -0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document -0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document -0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document -0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document -0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document -0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document -0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document -0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document -0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document -0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document -0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document -0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document -0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document -0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document -0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document -0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document -0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document -0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document -0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document -0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document -0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document -0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document -0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document -0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document -0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document -0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document -0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document -0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document -0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document -0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document -0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document -0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document -0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document -0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document -0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document -0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document -0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document -0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document -0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document -0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document -0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document -0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document -0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document -0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document -0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document -0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document -0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document -0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document -0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document -0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document -0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document -0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document -0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document -0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document -0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document -0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document -0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document -0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document -0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document -0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document -0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document -0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document -0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document -0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document -0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document -0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document -0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document -0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document -0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document -0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document -0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document -0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document -0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document -0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document -0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document -0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document -0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document -0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document -0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document -0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document -0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document -0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document -0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document -0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document -0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document -0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document -0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document -0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document -0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document -0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document -0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document -0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document -0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document -0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document -0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document -0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document -0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document -0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document -0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document -0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document -0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document -0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document -0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document -0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document -0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document -0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document -0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document -0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document -0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document -0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document -0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document -0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document -0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document -0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document -0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document -0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document -0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document -0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document -0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document -0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document -0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document -0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document -0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document -0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document -0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document -0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document -0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document -0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document -0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document -0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document -0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document -0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document -0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document -0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document -0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document -0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document -0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document -0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document -0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document -0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document -0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document -0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document -0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document -0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document -0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document -0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document -0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document -0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document -0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document -0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document -0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document -0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document -0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document -0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document -0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document -0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document -0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document -0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document -0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document -0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document -0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document -0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document -0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document -0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document -0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document -0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document -0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document -0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document -0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document -0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document -0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document -0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document -0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document -0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document -0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document -0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document -0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document -0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document -0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document -0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document -0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document -0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document -0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document -0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document -0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document -0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document -0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document -0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document -0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document -0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document -0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document -0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document -0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document -0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document -0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document -0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document -0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document -0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document -0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document -0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document -0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document -0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document -0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document -0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document -0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document -0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document -0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document -0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document -0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document -0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document -0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document -0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document -0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document -0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document -0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document -0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document -0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document -0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document -0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document -0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document -0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document -0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document -0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document -0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document -0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document -0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document -0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document -0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document -0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document -0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document -0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document -0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document -0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document -0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document -0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document -0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document -0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document -0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document -0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document -0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document -0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document -0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document -0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document -0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document -0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document -0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document -0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document -0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document -0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document -0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document -0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document -0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document -0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document -0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document -0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document -0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document -0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document -0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document -0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document -0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document -0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document -0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document -0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document -0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document -0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document -0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document -0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document -0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document -0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document -0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document -0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document -0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document -0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document -0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document -0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document -0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document -0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document -0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document -0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document -0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document -0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document -0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document -0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document -0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document -0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document -0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document -0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document -0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document -0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document -0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document -0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document -0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document -0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document -0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document -0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document -0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document -0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document -0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document -0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document -0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document -0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document -0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document -0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document -0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document -0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document -0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document -0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document -0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document -0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document -0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document -0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document -0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document -0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document -0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document -0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document -0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document -0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document -0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document -0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document -0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document -0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document -0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document -0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document -0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document -0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document -0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document -0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document -0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document -0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document -0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document -0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document -0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document -0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document -0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document -0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document -0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document -0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document -0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document -0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document -0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document -0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document -0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document -0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document -0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document -0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document -0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document -0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document -0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document -0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document -0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document -0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document -0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document -0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document -0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document -0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document -0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document -0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document -0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document -0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document -0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document -0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document -0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document -0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document -0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document -0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document -0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document -0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document -0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document -0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document -0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document -0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document -0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document -0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document -0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document -0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document -0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document -0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document -0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document -0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document -0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document -0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document -0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document -0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document -0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document -0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document -0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document -0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document -0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document -0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document -0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document -0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document -0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document -0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document -0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document -0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document -0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document -0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document -0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document -0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document -0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document -0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document -0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document -0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document -0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document -0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document -0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document -0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document -0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document -0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document -0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document -0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document -0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document -0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document -0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document -0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document -0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document -0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document -0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document -0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document -0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document -0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document -0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document -0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document -0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document -0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document -0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document -0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document -0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document -0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document -0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document -0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document -0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document -0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document -0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document -0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document -0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document -0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document -0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document -0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document -0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document -0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document -0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document -0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document -0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document -0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document -0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document -0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document -0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document -0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document -0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document -0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document -0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document -0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document -0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document -0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document -0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document -0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document -0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document -0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document -0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document -0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document -0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document -0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document -0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document -0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document -0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document -0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document -0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document -0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document -0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document -0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document -0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document -0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document -0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document -0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document -0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document -0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document -0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document -0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document -0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document -0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document -0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document -0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document -0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document -0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document -0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document -0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document -0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document -0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document -0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document -0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document -0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document -0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document -0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document -0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document -0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document -0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document -0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document -0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document -0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document -0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document -0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document -0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document -0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document -0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document -0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document -0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document -0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document -0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document -0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document -0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document -0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document -0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document -0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document -0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document -0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document -0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document -0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document -0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document -0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document -0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document -0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document -0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document -0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document -0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document -0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document -0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document -0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document -0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document -0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document -0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document -0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document -0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document -0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document -0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document -0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document -0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document -0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document -0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document -0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document -0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document -0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document -0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document -0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document -0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document -0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document -0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document -0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document -0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document -0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document -0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document -0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document -0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document -0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document -0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document -0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document -0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document -0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document -0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document -0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document -0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document -0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document -0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document -0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document -0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document -0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document -0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document -0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document -0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document -0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document -0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document -0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document -0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document -0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document -0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document -0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document -0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document -0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document -0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document -0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document -0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document -0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document -0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document -0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document -0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document -0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document -0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document -0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document -0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document -0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document -0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document -0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document -0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document -0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document -0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document -0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document -0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document -0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document -0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document -0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document -0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document -0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document -0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document -0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document -0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document -0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document -0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document -0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document -0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document -0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document -0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document -0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document -0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document -0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document -0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document -0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document -0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document -0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document -0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document -0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document -0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document -0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document -0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document -0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document -0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document -0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document -0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document -0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document -0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document -0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document -0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document -0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document -0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document -0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document -0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document -0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document -0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document -0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document -0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document -0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document -0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document -0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document -0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document -0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document -0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document -0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document -0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document -0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document -0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document -0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document -0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document -0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document -0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document -0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document -0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document -0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document -0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document -0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document -0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document -0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document -0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document -0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document -0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document -0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document -0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document -0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document -0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document -0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document -0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document -0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document -0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document -0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document -0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document -0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document -0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document -0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document -0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document -0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document -0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document -0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document -0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document -0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document -0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document -0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document -0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document -0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document -0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document -0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document -0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document -0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document -0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document -0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document -0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document -0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document -0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document -0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document -0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document -0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document -0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document -0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document -0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document -0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document -0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document -0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document -0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document -0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document -0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document -0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document -0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document -0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document -0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document -0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document -0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document -0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document -0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document -0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document -0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document -0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document -0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document -0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document -0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document -0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document -0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document -0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document -0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document -0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document -0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document -0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document -0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document -0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document -0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document -0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document -0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document -0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document -0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document -0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document -0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document -0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document -0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document -0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document -0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document -0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document -0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document -0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document -0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document -0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document -0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document -0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document -0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document -0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document -0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document -0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document -0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document -0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document -0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document -0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document -0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document -0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document -0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document -0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document -0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document -0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document -0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document -0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document -0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document -0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document -0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document -0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document -0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document -0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document -0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document -0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document -0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document -0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document -0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document -0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document -0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document -0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document -0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document -0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document -0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document -0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document -0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document -0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document -0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document -0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document -0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document -0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document -0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document -0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document -0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document -0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document -0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document -0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document -0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document -0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document -0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document -0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document -0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document -0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document -0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document -0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document -0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document -0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document -0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document -0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document -0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document -0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document -0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document -0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document -0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document -0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document -0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document -0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document -0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document -0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document -0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document -0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document -0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document -0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document -0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document -0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document -0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document -0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document -0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document -0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document -0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document -0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document -0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document -0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document -0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document -0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document -0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document -0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document -0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc diff --git a/ALCF/data-lists/aurora/dolma.txt b/ALCF/data-lists/aurora/dolma.txt new file mode 100644 index 00000000000..9449af64fc6 --- /dev/null +++ b/ALCF/data-lists/aurora/dolma.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train +0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv +0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 +0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc +0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon +6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika +0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train +0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o +0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit +0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange +0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder +0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu +0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki diff --git a/ALCF/data-lists/aurora/falcon.txt b/ALCF/data-lists/aurora/falcon.txt index 3aeb4a37513..27b3bbe0474 100644 --- a/ALCF/data-lists/aurora/falcon.txt +++ b/ALCF/data-lists/aurora/falcon.txt @@ -1,501 +1,501 @@ -0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document -0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document -0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document -0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document -0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document -0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document -0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document -0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document -0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document -0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document -0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document -0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document -0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document -0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document -0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document -0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document -0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document -0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document -0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document -0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document -0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document -0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document -0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document -0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document -0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document -0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document -0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document -0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document -0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document -0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document -0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document -0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document -0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document -0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document -0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document -0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document -0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document -0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document -0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document -0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document -0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document -0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document -0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document -0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document -0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document -0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document -0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document -0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document -0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document -0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document -0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document -0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document -0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document -0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document -0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document -0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document -0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document -0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document -0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document -0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document -0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document -0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document -0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document -0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document -0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document -0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document -0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document -0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document -0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document -0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document -0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document -0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document -0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document -0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document -0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document -0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document -0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document -0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document -0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document -0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document -0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document -0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document -0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document -0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document -0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document -0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document -0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document -0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document -0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document -0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document -0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document -0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document -0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document -0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document -0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document -0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document -0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document -0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document -0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document -0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document -0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document -0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document -0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document -0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document -0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document -0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document -0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document -0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document -0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document -0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document -0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document -0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document -0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document -0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document -0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document -0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document -0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document -0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document -0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document -0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document -0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document -0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document -0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document -0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document -0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document -0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document -0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document -0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document -0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document -0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document -0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document -0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document -0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document -0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document -0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document -0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document -0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document -0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document -0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document -0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document -0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document -0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document -0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document -0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document -0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document -0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document -0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document -0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document -0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document -0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document -0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document -0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document -0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document -0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document -0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document -0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document -0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document -0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document -0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document -0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document -0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document -0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document -0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document -0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document -0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document -0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document -0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document -0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document -0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document -0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document -0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document -0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document -0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document -0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document -0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document -0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document -0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document -0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document -0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document -0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document -0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document -0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document -0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document -0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document -0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document -0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document -0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document -0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document -0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document -0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document -0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document -0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document -0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document -0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document -0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document -0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document -0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document -0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document -0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document -0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document -0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document -0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document -0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document -0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document -0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document -0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document -0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document -0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document -0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document -0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document -0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document -0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document -0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document -0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document -0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document -0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document -0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document -0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document -0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document -0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document -0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document -0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document -0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document -0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document -0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document -0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document -0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document -0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document -0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document -0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document -0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document -0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document -0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document -0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document -0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document -0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document -0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document -0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document -0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document -0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document -0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document -0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document -0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document -0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document -0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document -0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document -0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document -0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document -0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document -0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document -0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document -0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document -0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document -0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document -0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document -0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document -0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document -0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document -0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document -0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document -0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document -0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document -0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document -0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document -0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document -0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document -0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document -0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document -0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document -0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document -0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document -0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document -0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document -0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document -0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document -0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document -0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document -0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document -0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document -0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document -0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document -0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document -0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document -0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document -0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document -0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document -0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document -0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document -0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document -0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document -0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document -0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document -0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document -0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document -0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document -0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document -0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document -0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document -0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document -0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document -0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document -0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document -0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document -0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document -0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document -0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document -0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document -0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document -0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document -0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document -0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document -0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document -0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document -0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document -0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document -0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document -0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document -0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document -0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document -0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document -0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document -0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document -0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document -0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document -0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document -0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document -0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document -0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document -0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document -0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document -0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document -0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document -0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document -0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document -0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document -0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document -0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document -0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document -0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document -0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document -0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document -0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document -0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document -0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document -0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document -0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document -0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document -0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document -0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document -0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document -0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document -0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document -0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document -0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document -0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document -0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document -0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document -0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document -0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document -0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document -0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document -0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document -0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document -0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document -0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document -0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document -0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document -0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document -0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document -0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document -0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document -0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document -0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document -0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document -0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document -0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document -0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document -0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document -0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document -0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document -0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document -0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document -0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document -0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document -0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document -0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document -0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document -0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document -0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document -0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document -0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document -0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document -0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document -0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document -0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document -0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document -0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document -0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document -0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document -0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document -0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document -0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document -0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document -0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document -0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document -0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document -0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document -0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document -0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document -0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document -0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document -0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document -0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document -0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document -0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document -0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document -0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document -0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document -0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document -0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document -0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document -0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document -0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document -0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document -0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document -0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document -0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document -0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document -0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document -0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document -0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document -0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document -0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document -0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document -0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document -0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document -0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document -0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document -0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document -0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document -0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document -0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document -0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document -0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document -0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document -0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document -0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document -0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document -0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document -0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document -0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document -0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document -0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document -0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document -0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document -0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document -0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document -0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document -0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document -0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document -0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document -0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document -0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document -0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document -0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document -0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document -0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document -0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document -0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document -0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document -0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document -0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document -0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document -0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document -0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document -0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document -0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document -0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document -0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document -0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document -0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document -0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document -0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document -0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document -0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document -0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document -0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document -0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document -0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document -0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document -0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document -0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document -0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document -0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document -0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document -0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document -0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document -0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document -0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document -0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon diff --git a/ALCF/data-lists/aurora/megawiki.txt b/ALCF/data-lists/aurora/megawiki.txt index 4983cd797c8..1e8e6672a2d 100644 --- a/ALCF/data-lists/aurora/megawiki.txt +++ b/ALCF/data-lists/aurora/megawiki.txt @@ -1,262 +1,262 @@ -6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document -2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document -5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document -2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document -2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document -2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document -5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document -1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document -1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document -1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document -1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document -1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document -1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document -1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document -1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document -1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document -1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document -1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document -1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document -1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document -1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document -1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document -1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document -1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document -1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document -1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document -1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document -1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document -1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document -1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document -1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document -1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document -1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document -1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document -1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document -1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document -1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document -1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document -1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document -1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document -1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document -1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document -1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document -1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document -1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document -1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document -1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document -1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document -1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document -1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document -1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document -1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document -5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document -5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document -5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document -5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document -5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document -5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document -5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document -4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document -4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document -5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document -4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document -5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document -6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document -5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document -4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document -5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document -5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document -4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document -5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document -5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document -5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document -5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document -5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document -6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document -5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document -5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document -5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document -5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document -5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document -5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document -5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document -4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document -5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document -5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document -4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document -5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document -5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document -5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document -4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document -5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document -4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document -5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document -5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document -6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document -5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document -5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document -5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document -5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document -5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document -5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document -4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document -5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document -5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document -5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document -5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document -5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document -5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document -5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document -5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document -5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document -5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document -5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document -2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document -1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document -3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document -3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document -2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document -2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document -2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document -2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document -2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document -2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document -2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document -2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document -2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document -2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document -2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document -2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document -2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document -1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document -1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document -2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document -2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document -9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document -2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document -1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document -2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document -2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document -4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document -2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document -2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document -1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document -1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document -1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document -1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document -1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document -1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document -1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document -1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document -1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document -1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document -1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document -1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document -1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document -1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document -1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document -1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document -1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document -1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document -1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document -1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document -1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document -1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document -1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document -1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document -8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document -8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document -9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document -9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document -7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document -8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document -9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document -8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document -8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document -8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document -7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document -9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document -8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document -1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document -3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document -2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document -1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document -1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document -1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document -2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document -2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document -8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document -1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document -1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document -1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document -1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document -1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document -1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document -1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document -2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document -1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document -2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document -5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document -1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document -1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document -1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document -1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document -1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document -1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document -1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document -1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document -1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document -1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document -1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document -1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document -1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document -1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document -1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document -1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document -1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document -1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document -9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document -1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document -1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document -1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document -1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document -1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document -1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document -1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document -1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document -1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document -1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document -4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document -1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document -1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document -2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document -2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document -2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document -2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document -1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document -1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document -3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document -6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document -1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document -1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document -1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document -2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document -3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document -1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document -1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document -1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document -1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document -1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document -1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document -1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document -1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document -1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document -1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document -4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document -3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document -3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document -3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document -3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document -3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document -7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika diff --git a/ALCF/data-lists/aurora/open-web-math-train.txt b/ALCF/data-lists/aurora/open-web-math-train.txt index a1a80aa1677..0bef3e40f7a 100644 --- a/ALCF/data-lists/aurora/open-web-math-train.txt +++ b/ALCF/data-lists/aurora/open-web-math-train.txt @@ -1,13 +1,13 @@ -0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document -0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document -0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document -0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document -0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document -6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document -0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document -0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document -0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document -0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document -0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document -0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document -0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train diff --git a/ALCF/data-lists/aurora/pes2o.txt b/ALCF/data-lists/aurora/pes2o.txt index 751cce20a02..c1e8429f4b8 100644 --- a/ALCF/data-lists/aurora/pes2o.txt +++ b/ALCF/data-lists/aurora/pes2o.txt @@ -1,26 +1,26 @@ -0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document -0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document -0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document -0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document -0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document -0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document -0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document -0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document -0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document -0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document -0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document -0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document -0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document -0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document -0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document -0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document -0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document -0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document -0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document -0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document -0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document -0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document -0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document -0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document -0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document -0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o diff --git a/ALCF/data-lists/aurora/reddit.txt b/ALCF/data-lists/aurora/reddit.txt index ea235a34685..e01a4a1b866 100644 --- a/ALCF/data-lists/aurora/reddit.txt +++ b/ALCF/data-lists/aurora/reddit.txt @@ -1,78 +1,78 @@ -0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document -0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document -0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document -0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document -0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document -0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document -0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document -0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document -0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document -0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document -0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document -0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document -0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document -0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document -0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document -0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document -0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document -0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document -0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document -0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document -0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document -0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document -0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document -0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document -0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document -0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document -0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document -0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document -0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document -0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document -0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document -0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document -0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document -0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document -0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document -0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document -0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document -0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document -0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document -0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document -0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document -0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document -0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document -0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document -0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document -0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document -0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document -0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document -0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document -0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document -0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document -0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document -0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document -0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document -0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document -0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document -0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document -0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document -0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document -0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document -0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document -0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document -0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document -0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document -0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document -0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document -0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document -0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document -0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document -0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document -0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document -0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document -0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document -0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document -0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document -0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document -0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document -0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit diff --git a/ALCF/data-lists/aurora/stack.txt b/ALCF/data-lists/aurora/stack.txt index bc7f3efbc75..b2defb3b370 100644 --- a/ALCF/data-lists/aurora/stack.txt +++ b/ALCF/data-lists/aurora/stack.txt @@ -1,26 +1,26 @@ -0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document -0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document -0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document -0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document -0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document -0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document -0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document -0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document -0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document -0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document -0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document -0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document -0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document -0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document -0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document -0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document -0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document -0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document -0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document -0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document -0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document -0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document -0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document -0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document -0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document -0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange diff --git a/ALCF/data-lists/aurora/starcoder.txt b/ALCF/data-lists/aurora/starcoder.txt index 401f053d944..a6436675f7b 100644 --- a/ALCF/data-lists/aurora/starcoder.txt +++ b/ALCF/data-lists/aurora/starcoder.txt @@ -1,50 +1,50 @@ -0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document -0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document -0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document -0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document -0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document -0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document -0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document -0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document -0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document -0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document -0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document -0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document -0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document -0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document -0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document -0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document -0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document -0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document -0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document -0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document -0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document -0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document -0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document -0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document -0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document -0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document -0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document -0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document -0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document -0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document -0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document -0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document -0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document -0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document -0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document -0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document -0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document -0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document -0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document -0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document -0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document -0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document -0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document -0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document -0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document -0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document -0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document -0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document -0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder diff --git a/ALCF/data-lists/aurora/tulu.txt b/ALCF/data-lists/aurora/tulu.txt index ac59c679df1..bafc809a8c9 100644 --- a/ALCF/data-lists/aurora/tulu.txt +++ b/ALCF/data-lists/aurora/tulu.txt @@ -1,66 +1,66 @@ -0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document -0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document -0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document -0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document -0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document -0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document -0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document -0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document -0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document -0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document -0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document -0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document -0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document -0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document -0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document -0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document -0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document -0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document -0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document -0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document -0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document -0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document -0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document -0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document -0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document -0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document -0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document -0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document -0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document -0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document -0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document -0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document -0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document -0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document -0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document -0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document -0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document -0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document -0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document -0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document -0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document -0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document -0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document -0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document -0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document -0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document -0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document -0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document -0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document -0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document -0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document -0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document -0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document -0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document -0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document -0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document -0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document -0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document -0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document -0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document -0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document -0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document -0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document -0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document -0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document -3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu diff --git a/ALCF/data-lists/aurora/wiki.txt b/ALCF/data-lists/aurora/wiki.txt index f0ab87cb679..65a8ac92a87 100644 --- a/ALCF/data-lists/aurora/wiki.txt +++ b/ALCF/data-lists/aurora/wiki.txt @@ -1,2 +1,2 @@ -0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document -0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document +0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki From 0320b696ca31daef25100aefa05abd24435dc177 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:51:56 -0500 Subject: [PATCH 222/590] Update `ALCF/data-lists/sunspot/*.txt` --- ALCF/data-lists/sunspot/algebraic.txt | 32 +- ALCF/data-lists/sunspot/arxiv.txt | 200 +- ALCF/data-lists/sunspot/books.txt | 6 +- ALCF/data-lists/sunspot/c4.txt | 342 +-- ALCF/data-lists/sunspot/cc.txt | 2214 +++++++-------- ALCF/data-lists/sunspot/dolma.txt | 2419 +++++++++++++++++ ALCF/data-lists/sunspot/falcon.txt | 1000 +++---- ALCF/data-lists/sunspot/megawiki.txt | 524 ++-- .../sunspot/open-web-math-train.txt | 26 +- ALCF/data-lists/sunspot/pes2o.txt | 52 +- ALCF/data-lists/sunspot/reddit.txt | 156 +- ALCF/data-lists/sunspot/stack.txt | 52 +- ALCF/data-lists/sunspot/starcoder.txt | 98 +- ALCF/data-lists/sunspot/tulu.txt | 132 +- ALCF/data-lists/sunspot/wiki.txt | 4 +- 15 files changed, 4838 insertions(+), 2419 deletions(-) create mode 100644 ALCF/data-lists/sunspot/dolma.txt diff --git a/ALCF/data-lists/sunspot/algebraic.txt b/ALCF/data-lists/sunspot/algebraic.txt index 0f25e30498d..f72bf47d742 100644 --- a/ALCF/data-lists/sunspot/algebraic.txt +++ b/ALCF/data-lists/sunspot/algebraic.txt @@ -1,16 +1,16 @@ -0.0018520780893211373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document -0.0017591050606817512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document -0.001459052794333798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document -0.0007405667281569194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document -0.00019420030110896795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document -0.0009008668715801845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document -0.00015115827957143057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document -0.0014552844319220648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document -0.0012469861325685161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document -0.00136412011372413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document -0.0007064279699221103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document -0.0008472240000687427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document -0.0001984375713341955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document -0.0005472773881697123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document -0.001815779629850992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document -0.0018313600689757324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0018520780893211373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train diff --git a/ALCF/data-lists/sunspot/arxiv.txt b/ALCF/data-lists/sunspot/arxiv.txt index c50df905037..34972accf4a 100644 --- a/ALCF/data-lists/sunspot/arxiv.txt +++ b/ALCF/data-lists/sunspot/arxiv.txt @@ -1,100 +1,100 @@ -0.0002583902668716813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document -0.0002646575141232155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document -0.0003165521247456758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document -0.0002920706460176214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document -0.00028396813182810215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document -0.00030445161883108107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document -0.00031628781276576474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document -0.0003083776568189157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document -0.0003176359471472902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document -0.0002536009369131698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document -0.0003067491424681363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document -0.0002597217257557784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document -0.0003788556450109768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document -0.0002796563272052598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document -0.00033573826524290287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document -0.00030523658022800287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document -0.00032211552192240096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document -0.0003329295675164247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document -0.0003101982186639862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document -0.00032361798234223355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document -0.0003495541581652915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document -0.0002821637448858042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document -0.00030399523537629673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document -0.0002955658968247219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document -0.00028942158502924254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document -0.00028769546171490733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document -0.0002938111057234182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document -0.0002711150403010948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document -0.00031130095874747565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document -0.0003002996118160777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document -0.0003732757901604459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document -0.00026784205751795894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document -0.0002799626521661984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document -0.00034334276069078164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document -0.0003582469803674965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document -0.00031094844818418623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document -0.0002766228384977191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document -0.00030297116159471485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document -0.00027033888377464685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document -0.00030090862368377933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document -0.00028543875802490955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document -0.00027559768459074204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document -0.0003182185533962886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document -0.0003311392971435837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document -0.00028751652060804325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document -0.000303466863212589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document -0.00033400462801277524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document -0.0002589234031777426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document -0.0002913508598466723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document -0.0002670572450004856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document -0.00032027399105647656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document -0.00032188376258379377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document -0.0003161585784100882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document -0.0003184249182974135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document -0.00030381336664000807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document -0.0003190437442184283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document -0.0002537961798200545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document -0.0003017817117223326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document -0.00028685268513240224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document -0.00031265179094451165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document -0.00034708319096986816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document -0.00026650837943080664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document -0.00034588832248507335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document -0.0002416982248399037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document -0.0003089296918222243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document -0.00029137184185700827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document -0.00026464226846800774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document -0.00030545397919456627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document -0.0003206778460448875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document -0.00030968971641110967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document -0.00023325653928600864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document -0.00030526899198338555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document -0.00035376719076633584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document -0.000290224385981026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document -0.000294650083382008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document -0.00028768858128616436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document -0.00030856965235527843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document -0.00030579942447879054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document -0.0002863101084704357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document -0.0002870032092492213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document -0.000264182727569885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document -0.0002974012367036449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document -0.00032238412143059203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document -0.00031683716893819036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document -0.00031157434937617524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document -0.0003411742735695989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document -0.00026778444816570715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document -0.0003037045797275201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document -0.00027746114370081314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document -0.00027148285946862043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document -0.00028042950114678207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document -0.0003235607816590721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document -0.0003086692227306295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document -0.00033990349455148105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document -0.00030945053208470265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document -0.00027309074552265303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document -0.00028737393506316194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document -0.0003098868328009879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document -0.0002614229162588409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document -0.0002884388407820923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0002583902668716813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv diff --git a/ALCF/data-lists/sunspot/books.txt b/ALCF/data-lists/sunspot/books.txt index 7aa37a00d2e..9502fba1f5f 100644 --- a/ALCF/data-lists/sunspot/books.txt +++ b/ALCF/data-lists/sunspot/books.txt @@ -1,3 +1,3 @@ -0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document -0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document -0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books diff --git a/ALCF/data-lists/sunspot/c4.txt b/ALCF/data-lists/sunspot/c4.txt index 9504bcbfe65..ca4836ad814 100644 --- a/ALCF/data-lists/sunspot/c4.txt +++ b/ALCF/data-lists/sunspot/c4.txt @@ -1,171 +1,171 @@ -0.0002406272620255565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document -0.0002404825539493424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document -0.00024062296575435581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document -0.00024069315766818953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document -0.00024055829162263452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document -0.00024062053397343032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document -0.0002410715545206964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document -0.00024024881846087368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document -0.0002407074700790688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document -0.00024072141428809043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document -0.00024027710230872736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document -0.0002409111299205489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document -0.00024081954058275009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document -0.00024086076794990912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document -0.00024098672620832446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document -0.00024068622303333862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document -0.00024140627024291824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document -0.0002414512033594384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document -0.00024028742594941463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document -0.00024018036089269645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document -0.0002398347365034979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document -0.00024006780153485276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document -0.00024015620270419213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document -0.0002408848259695227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document -0.0002408023185278831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document -0.00024021196580140326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document -0.00024077677271297493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document -0.00024087392454668027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document -0.0002408071293824126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document -0.00024042223828845715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document -0.0002411484752360495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document -0.00023605263746465907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document -0.00023471222158326908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document -0.00023432138580287644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document -0.00023407385623382327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document -0.00023487504174367091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document -0.0002341843704976313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document -0.00023421993170282486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document -0.00023445057969132037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document -0.0002337681680073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document -0.000234627964808109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document -0.0002338942211888584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document -0.00023403849286843386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document -0.00023405641310796305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document -0.00023349169562397965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document -0.00023381157386048856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document -0.00023388742993790587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document -0.00023363103829469813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document -0.00023421141834630477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document -0.00023420564352232565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document -0.00023367463699173143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document -0.00023344969163567033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document -0.00023372196941547188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document -0.00023399207645297834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document -0.00023357915605505856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document -0.00023337585642190864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document -0.00023385005470157914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document -0.00023301533534493465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document -0.00023377864302541782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document -0.00023323745848621437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document -0.0002330594611151835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document -0.0002334149675026783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document -0.00023198945902291534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document -0.00023023784834634142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document -0.00022985623060187217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document -0.0002292605284569516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document -0.00022926593333048894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document -0.00022922766406807777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document -0.00022898153911167426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document -0.0002292473111593315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document -0.000228804579400424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document -0.00022865485613513526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document -0.00022937426835887895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document -0.00022917388311587372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document -0.0002291660582019043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document -0.00022907895248360543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document -0.0002294617879920205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document -0.0002290452150516566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document -0.00022943405619715553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document -0.0002296271421006204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document -0.00022854791372910372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document -0.00022923123467686557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document -0.00022852404355738494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document -0.00022847798660086642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document -0.0002289604586810316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document -0.00022835479834950643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document -0.0002289149402884243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document -0.00022806655474763446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document -0.00022826296420992974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document -0.00022906829636213627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document -0.0002287628414466998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document -0.0002282673911253445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document -0.00022869309841939134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document -0.0002281540116815451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document -0.0002259755756162738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document -0.00022562331285233504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document -0.0002259061146106053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document -0.00022567670836663787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document -0.00022573165387587061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document -0.00022508514961670572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document -0.00022564642513773356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document -0.00022563088621998788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document -0.0002250438755373707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document -0.00022524465346241134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document -0.00022531737657666812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document -0.00022444687519363458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document -0.00022460397498596298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document -0.00022454218976501763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document -0.00022447528843671366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document -0.00022501666332178926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document -0.00022453752304377972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document -0.00022484451871163002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document -0.00022465678847154914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document -0.00022453180917044732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document -0.0002247278486823009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document -0.00022465794828242097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document -0.00022431000701925386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document -0.00022476020248460963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document -0.00022467531771795015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document -0.0002236391309945234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document -0.00022458764920536007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document -0.00022430877426744415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document -0.0002247047786127192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document -0.0002245298090400035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document -0.0002245648831396188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document -0.00022292894729820784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document -0.00022236668082957533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document -0.0002217622659895442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document -0.00022252452726732609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document -0.00022135333211363678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document -0.0002214571757787971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document -0.0002217188139237798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document -0.00022144214894640303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document -0.00022100172806631854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document -0.00022156392409199052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document -0.00022134830143710272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document -0.00022158598922529453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document -0.00022142932483041377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document -0.00022120980907786554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document -0.00022117917738112441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document -0.00022077089397851235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document -0.00022093265074996711 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document -0.00022091299741377004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document -0.0002205849150703338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document -0.0002210648204787979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document -0.0002214235747364102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document -0.00022083907302221787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document -0.0002206334237915964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document -0.00022065193929912214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document -0.00022079775597767288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document -0.00022091492909963518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document -0.00022095009987097293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document -0.0002208150577180165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document -0.00022085759102772088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document -0.00022073789170129016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document -0.00022049322781182384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document -0.00022083270617761285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document -0.00021982452827473632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document -0.00021899870446514259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document -0.00021890358773356361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document -0.00021875556609042841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document -0.00021861195987201226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document -0.00021856782186167455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document -0.00021912837771543515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document -0.00021900213768517756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document -0.00021871675851390374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document -0.0002180537056545586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document -0.0002188196714327129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document -0.00021851362624523464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document -0.0002183236795498736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document -7.291153618675672e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0002406272620255565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 diff --git a/ALCF/data-lists/sunspot/cc.txt b/ALCF/data-lists/sunspot/cc.txt index 0a2a0ce35bb..d771efb06ad 100644 --- a/ALCF/data-lists/sunspot/cc.txt +++ b/ALCF/data-lists/sunspot/cc.txt @@ -1,1108 +1,1108 @@ -0.0003742481815405742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document -0.00038204855962733055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document -0.00038821818392663593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document -0.00038723332988783727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document -0.00038916141142149904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document -0.00038049542523949033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document -0.0003854755539534284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document -0.00024202756466512517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document -0.0003915405155008087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document -0.0003927382151931033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document -0.0003839151202260479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document -0.00040006817468967907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document -0.00040318965964443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document -0.0003831013019452741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document -0.00039166638383204036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document -0.00039962784023961004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document -0.00039536707853602614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document -0.0004204304698247758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document -0.00041538899178693555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document -0.00039186953333675306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document -0.00038945837196504305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document -0.0003919951238929062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document -0.00044377065718528966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document -0.0004407759068603017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document -0.0002487811895843715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document -0.00039349432045556636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document -0.00041223198559462343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document -0.0004036573014830213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document -0.0003825982215521807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document -0.00040386867133151386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document -0.00024460575279105167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document -0.000269029789531335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document -0.0003573757493252864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document -0.0004600876681392076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document -0.0002605354166397086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document -0.0003882502452157999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document -0.0002466747612126512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document -0.0004024726105072402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document -0.00040820631128483644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document -0.0002691094350403538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document -0.00026916830387277267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document -0.0004204663297880574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document -0.00042379698687085554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document -0.0004502169227311871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document -0.0002661708937015295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document -0.00031239486948031334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document -0.0003109054589936201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document -0.00045873053079760646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document -0.00022904931423244635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document -0.0003813462028433663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document -0.00039188129256500874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document -0.00045124222276983765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document -0.00048138658436853695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document -0.0003944178776279866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document -0.00039941569676754006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document -0.00037952761190240494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document -0.0003944870860881476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document -0.0003891842411856621 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document -0.000387688981934861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document -0.00039197953876258005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document -0.00039007915280311206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document -0.0003995520363699188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document -0.00039230985654592406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document -0.0003929472067173851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document -0.0003924096172671473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document -0.0003881636143629905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document -0.000389790617937084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document -0.00037351762309221023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document -0.0003630196170929407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document -0.00033532465765142113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document -0.0003076088685761823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document -0.00039463850897720803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document -0.0002843816115231449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document -0.0002909175709416474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document -0.00028867170997202486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document -0.0002838644617723659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document -0.00029027869525543416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document -0.0002821339567560056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document -0.0002922988877045601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document -0.0002866955958315786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document -0.0002865271754558126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document -0.0002861247475618473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document -0.0002826681072408606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document -0.0002849746458282827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document -0.0002816966633435316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document -0.00026255342235948463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document -0.0002552895098829678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document -0.00025990194083107813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document -0.0002524062657685835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document -0.0002538577379748611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document -0.0002561415177406761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document -0.00026206253059694905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document -0.00026168095406910565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document -0.0002601305742008613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document -0.00025200823006814814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document -0.0003229951981263502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document -0.00037289448266476045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document -0.0003807825862179898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document -0.0003616333738191483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document -0.0003665117918907636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document -0.0003684186453633228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document -0.0003589330610806066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document -0.00036383861418030395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document -0.000359841363355303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document -0.00036431044063050464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document -0.0003668574090358279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document -0.000362768263620199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document -0.0003501888032771077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document -0.000352401968221528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document -0.0003541019701869794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document -0.0003628121865546891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document -0.0003752582953758773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document -0.00037902046230424966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document -0.0003777927146925147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document -0.0003760676130509053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document -0.00034046049078755405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document -0.0003338847563259091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document -0.00033294499102761794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document -0.0004912026198265864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document -0.00032064363474664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document -0.00032154190389541214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document -0.00032309660151746207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document -0.00031181143365304544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document -0.00031046092294569104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document -0.00031150165249068046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document -0.0003041314265988224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document -0.0003024834909739394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document -0.0003019936835833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document -0.000292329665283177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document -0.0002867061143144972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document -0.00028443615610701707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document -0.00028462291013755945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document -0.0002793538601205013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document -0.00027306573977044246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document -0.00027097155673336525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document -0.0002752934202112985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document -0.00043042012694697647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document -0.00047495648822986177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document -0.00047755032493473855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document -0.0004706974343933747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document -0.00046682163297771817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document -0.0004616765425874178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document -0.00030644496751628097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document -0.0002909492555358308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document -0.00027272036068261724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document -0.0004101070217315588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document -0.0003728914338834357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document -0.00036546911442305647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document -0.0003669945482407483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document -0.0003715902407424017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document -0.00035837486406683366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document -0.0003573318538685469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document -0.0003553784893071916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document -0.0004920659809912352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document -0.0004533619411303183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document -0.00045067066057818706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document -0.00044396985139270645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document -0.00043198288204468477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document -0.00043005174223738454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document -0.00041847118430776784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document -0.00042952036375796664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document -0.00043420594647324267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document -0.0003461123241053012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document -0.0003408581597849182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document -0.00033172705422182547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document -0.0003392566490686136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document -0.00033578341518385483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document -0.0003439196710518844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document -0.00034559163447085543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document -0.00033762478642902825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document -0.00033215210055107224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document -0.00033423579608014966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document -0.0004963355016025102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document -0.0004996862761456923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document -0.0005000551829325451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document -0.0005004212610098755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document -0.00027768695585500585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document -0.00028395983854338433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document -0.00027835826303062254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document -0.0002740073176010804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document -0.0002791830529274016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document -0.0002796863816194411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document -0.00026697453022672804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document -0.0002594197440280141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document -0.0003779565697649222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document -0.00041835823476586606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document -0.00043788493575265915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document -0.0002731731970096006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document -0.000276305847423402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document -0.0002704955773958623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document -0.0002629635944827518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document -0.000260070956974436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document -0.00025661553791456334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document -0.00025794727207576157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document -0.00025295733980001527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document -0.0003788106407021029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document -0.0004882344027669431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document -0.0003275324309642705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document -0.0004803401856640094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document -0.00046720138323433943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document -0.00043527810307095335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document -0.00043905395741627827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document -0.00048774175867331425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document -0.00048380704121346737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document -0.0004779011848346118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document -0.00046255587581908036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document -0.00045127922880511576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document -0.0004503891485256095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document -0.0004450142332303422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document -0.00044630282482516654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document -0.00044325014465743616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document -0.0004263874842796447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document -0.0004217530913646938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document -0.000415120314341852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document -0.00040987168279144537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document -0.00033468337266607834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document -0.0003353094464683005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document -0.0004833936821707294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document -0.00047194878988920935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document -0.0004648324126996427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document -0.0004562345003964941 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document -0.0004933203505465098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document -0.0003530166075325466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document -0.00035368548192804685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document -0.0004872620828289663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document -0.00048293889392426456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document -0.00047936768462267655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document -0.00047821013991587545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document -0.0004660610308564753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document -0.000394683430103437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document -0.00039165053441571324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document -0.0003906936040164381 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document -0.00038074803919159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document -0.0003686529291578143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document -0.00035832920428870976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document -0.00035929024535947033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document -0.0003538226556050544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document -0.0003584167868708799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document -0.0003480507542594234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document -0.0003413709023543034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document -0.00034001304759361455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document -0.00033430532902756514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document -0.00046519252660631277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document -0.0002938876402514769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document -0.00028676090994509047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document -0.00027296150117506716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document -0.00026513502621960483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document -0.0002680081327926125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document -0.00025831225828720344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document -0.00026647037295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document -0.0002525733734572654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document -0.00025831708887575375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document -0.00042487627444443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document -0.0004951213245023891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document -0.0004804051413177752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document -0.0004662397611340532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document -0.0004550138655253933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document -0.00044494909122746795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document -0.0002899112253051385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document -0.0004372879736279761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document -0.0004529568099252922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document -0.00045127826158829573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document -0.0004436558176737439 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document -0.0004419233237678378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document -0.000434589215880319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document -0.00029153613207706566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document -0.0004312458058738854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document -0.00028741854968757313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document -0.00046853200754421234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document -0.0004949145252030074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document -0.00044459683920483167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document -0.0003836095306696336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document -0.0003789760237872398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document -0.0003749227438304427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document -0.0003628558277173369 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document -0.00039468301394041474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document -0.00038874701821614864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document -0.0004158492456077867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document -0.00042360504554060077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document -0.00040386729844317623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document -0.00027595096702902474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document -0.00043638766787829135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document -0.0002218691596850179 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document -0.0004437566108089954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document -0.0003889996411609667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document -0.00043454421906537704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document -0.0004522564392830988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document -0.00041517835659357416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document -0.0002614360863446896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document -0.00037543522111463596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document -0.0004386190133514781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document -0.00046358333286115075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document -0.00043186261317942404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document -0.0002377581602097957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document -0.00025973334085074254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document -0.00040139099332000796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document -0.00043674860686687174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document -0.00040853289309329373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document -0.000242910191729688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document -0.0004431071731750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document -0.0004388092670482523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document -0.000381418866255965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document -0.0004100117296419717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document -0.00042469230366022745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document -0.00041744151905374254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document -0.00022835699906752945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document -0.0004380161085387397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document -0.00044803212381807456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document -0.00040554932796137236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document -0.0004234508646347761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document -0.00043341209652360653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document -0.00023966604734537185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document -0.000259165907316014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document -0.0004270653021833602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document -0.0004341547032162028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document -0.0004111478117275994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document -0.0004299383567984396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document -0.0004241899124590779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document -0.0004502719349364145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document -0.00038994621469645615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document -0.0003859912398894952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document -0.0004247535950310557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document -0.000386982084327716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document -0.0004196451040053251 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document -0.0004096278509782259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document -0.0004373334932695721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document -0.0004180889975240641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document -0.00042079636929672745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document -0.00038063574611812913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document -0.0003817505891515542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document -0.0004420096268860222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document -0.00039182670726410623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document -0.0003635667850372299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document -0.00041564996472055667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document -0.000400529358757286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document -0.0003939113874958451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document -0.00039066622068940996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document -0.0004290098538807143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document -0.0004240739958197099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document -0.00040775392659215333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document -0.0004091634200396925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document -0.00042299190476617914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document -0.0003701492680344151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document -0.0003807353844384635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document -0.00038813507771983156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document -0.00040072346558408346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document -0.0003603595180423597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document -0.00038799421353112465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document -0.00037575235582264926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document -0.0004239190342959713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document -0.0004606044799136546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document -0.00045107950652529253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document -0.0004391947201871058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document -0.0004457516661123035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document -0.0004301297170991686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document -0.00044661704164586694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document -0.0004438849846114837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document -0.0004444205734316823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document -0.0004190924165303394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document -0.00043942581131677875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document -0.00021568459798090663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document -0.0003814929225407199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document -0.0003217453179359235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document -0.00031719591470267974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document -0.00032434115726922137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document -0.0004079911120371051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document -0.000329492766381148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document -0.0003845916162001633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document -0.0003835208964390098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document -0.00037847334157173194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document -0.00038296039903791865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document -0.00037896336828472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document -0.00037620974396391355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document -0.00037420590727111843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document -0.000340490625886403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document -0.0003078314411035827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document -0.00034153990750656097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document -0.0003308858103982067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document -0.0003452640607156025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document -0.00033095276418403455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document -0.0003116308995860414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document -0.00032446713226408477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document -0.0003015816821912984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document -0.00031612418775706894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document -0.0003278516344971041 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document -0.00033079446736097217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document -0.00032278977146550837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document -0.00032065272988207914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document -0.0003936696452406576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document -0.0003450109536627789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document -0.0003339787189919641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document -0.0003284303856176974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document -0.00033652677276843477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document -0.0003257822443845694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document -0.0003293985569149334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document -0.0003310360260148262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document -0.0003233770986418526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document -0.0003172280092149422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document -0.0003160674744292835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document -0.00030931090289598506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document -0.0003093173886443107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document -0.00033167847081104083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document -0.00031131501311729723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document -0.00031046608876279845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document -0.00030569235942207244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document -0.00030777943671285197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document -0.00029303314290956683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document -0.0003045824546400205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document -0.00030360880677729793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document -0.00031646239964835433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document -0.0003129122300603785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document -0.00031060464956661433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document -0.000311819032500067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document -0.0002977872483902282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document -0.0003009448600922438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document -0.00028610292098537774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document -0.0002988326876216654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document -0.00028550828372819075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document -0.0002830381750875739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document -0.0002848495855927156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document -0.0002856443760308144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document -0.00027442895344188584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document -0.0002681160554049462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document -0.0003421482544126989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document -0.0004005872948449718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document -0.0003930123959320308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document -0.0003867271832275778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document -0.000380805140455254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document -0.0003814769861947819 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document -0.00038025170883282324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document -0.0003738026647867475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document -0.00018960856915036276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document -0.0003697177501953134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document -0.00036674194328136693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document -0.00036447406838697555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document -0.00036686410861101255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document -0.00035915267825103423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document -0.0003624758404026675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document -0.0002822812140180794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document -0.00030620512946920813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document -0.000294249776520589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document -0.00030238536967523434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document -0.00029509593361580754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document -0.0002906912701830899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document -0.0002921944165474959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document -0.00028358919691127954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document -0.0002813182772323272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document -0.00027442640800299205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document -0.0002747820342933984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document -0.0002747584403979717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document -0.00027499129634862444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document -0.0002712050404257197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document -0.0002616256943143254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document -0.00026769938929002815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document -0.00038396081322727017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document -0.0003863140490027991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document -0.00037702277513203237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document -0.0003633274156107032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document -0.0003587473889240435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document -0.0003507672084278415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document -0.00033776425499780385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document -0.0003377914127574796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document -0.00032948015659161326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document -0.00033245638541392985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document -0.00031080707640648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document -0.0002976903331149755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document -0.0002965121463725523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document -0.0002933849695266647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document -0.0002837035078508233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document -0.00028684569079589323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document -0.0003145192320802359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document -0.0003566937253273515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document -0.0003470199109592918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document -0.0003060245312041868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document -0.0002650817213818789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document -0.0002643604938780134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document -0.000299350876031416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document -0.0003178540797697938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document -0.000271850367887767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document -0.00031349896596549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document -0.00031749734412765755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document -0.0003791137842391209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document -0.0003742334169957992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document -0.0003705639757351107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document -0.0003126986769797042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document -0.00031038132814561196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document -0.00036464437173804883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document -0.0003569480488951322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document -0.0003541239221619106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document -0.00035315297411308053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document -0.0003572451925404141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document -0.0003514986129411253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document -0.0003521798298425866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document -0.00034553677439244716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document -0.000349004719809412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document -0.0003468247484872769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document -0.0003465822608356558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document -0.00035410983132162007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document -0.0003487908354969444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document -0.0003479024763238147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document -0.000341412530646823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document -0.00034451316273667034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document -0.0002618849993484869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document -0.00026788679978901144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document -0.00027450670773227214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document -0.0002661273129899329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document -0.00026836569676402957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document -0.00026155876975483236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document -0.0002609276830117151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document -0.0002644161630512771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document -0.00036789208972872557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document -0.00037829849439990513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document -0.0003788894943523098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document -0.0003617207777959397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document -0.0002541334487248998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document -0.0002707945538071073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document -0.00027046282716455214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document -0.0002652443167243215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document -0.0002685859923850986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document -0.00025734961751176414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document -0.000259041720872915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document -0.00025340107274823446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document -0.00025757135121837893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document -0.00025617700500574084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document -0.0002566931670562857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document -0.0002543871190716101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document -0.00024997565589481713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document -0.0002954079779456287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document -0.00034890741135252835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document -0.0003473298137731525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document -0.0003296959618486435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document -0.0003304520061604598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document -0.00032377956175729824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document -0.00031700696295168713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document -0.0003060382346081943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document -0.0003012003005056863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document -0.0002981074073993884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document -0.0002922128825950705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document -0.000348901087722931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document -0.0003408286289467841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document -0.0003410649680770183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document -0.0003358524215576502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document -0.0003343661874989231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document -0.00032810573699389156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document -0.00032261449539097497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document -0.0003162694866049203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document -0.0003158381156468853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document -0.000317376061083603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document -0.0003125788639953052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document -0.0003010105041885602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document -0.0003065865059090678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document -0.0003084275726508053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document -0.00030966560718296085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document -0.0002957728057853081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document -0.00029904164542325336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document -0.0002955358888729187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document -0.00028692976446931544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document -0.0002923476214935797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document -0.0002893691697212419 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document -0.0002855895211981585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document -0.00027968347097626246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document -0.0002810783462604979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document -0.00027794080455729715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document -0.00034784376461416953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document -0.0003488347959010943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document -0.00034790583710250724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document -0.000345913166618151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document -0.00033801936268066675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document -0.0003290591130212315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document -0.00034051399521366823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document -0.00032470943131841784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document -0.00031679540050914276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document -0.00031814596342422325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document -0.0003156466289485036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document -0.00029985010879003633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document -0.0002905176377776361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document -0.0004206836775460856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document -0.00020660449162246918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document -0.0003461727254468087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document -0.00020592870907067763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document -0.00034173505299233005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document -0.0004052437256652738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document -0.0004080650901351697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document -0.00039778184149144276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document -0.00039046311464950275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document -0.00039043444911071384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document -0.000388575704932843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document -0.00019737533145666597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document -0.00037610755595812403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document -0.00037315400127598317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document -0.00037415028580922163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document -0.00036694041707212337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document -0.00018947219857306515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document -0.00037046050826533545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document -0.0003587440768559087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document -0.00034623936498708903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document -0.0003502289592617922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document -0.00034692398063649823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document -0.000339340809421849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document -0.0003360510394816983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document -0.0003354673850814145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document -0.00032937682875877047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document -0.00032844505049317715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document -0.00028287199339908627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document -0.0002795217197003578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document -0.00028048955601883463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document -0.0002769326396439027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document -0.0002727090021299243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document -0.0002726577841024554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document -0.00026663619593455374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document -0.00026068042672138127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document -0.0002637704114326801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document -0.0002593043567100412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document -0.0002599897110113453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document -0.0002435078682758859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document -0.0002450530071379054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document -0.00024233331983743606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document -0.0002934750947999535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document -0.00033241226364044474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document -0.00032938406090272075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document -0.00032778705403953246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document -0.00032184551480398754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document -0.00031874002264945737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document -0.0003165319685666433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document -0.00031307071173376295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document -0.00031119524184911957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document -0.0003102253344576429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document -0.0003088976240383192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document -0.0002951410823077708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document -0.00029772657676757413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document -0.0003056048989909935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document -0.00031991305381648026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document -0.00030890256978362426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document -0.0003109382904091933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document -0.00031035798529690644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document -0.00030741666395911753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document -0.0002989918594861846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document -0.00029569635443989434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document -0.0002973992445667285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document -0.000293397351001072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document -0.00028737817438047954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document -0.00028252738144009747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document -0.0002805511898623541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document -0.0003718020784620472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document -0.0003499713845765235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document -0.00034283547445326676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document -0.00031464759888838765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document -0.00033188946446414833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document -0.000326084432195463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document -0.0003764568303917893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document -0.0003604955598858414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document -0.0003655654554133222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document -0.00035762304033750504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document -0.00038478883950347103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document -0.00027735714341247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document -0.00028139534607773563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document -0.00019777292251713763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document -0.000285571704874486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document -0.00028543482146244363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document -0.00019434234484256758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document -0.00027854908176986763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document -0.0002847068039566143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document -0.00028672356943064853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document -0.00027782687605808177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document -0.0002843539634105203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document -0.0002894748379090401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document -0.0002868852440186493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document -0.0002818504885373851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document -0.00028680112812941034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document -0.00019258978168723977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document -0.00028760637934715155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document -0.0002820439443912918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document -0.0002831001054410018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document -0.00029001901552467397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document -0.00027779449377883156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document -0.00019949837437516796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document -0.0002907306472984446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document -0.00027814858381318327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document -0.00019472790889161432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document -0.00020472626596924125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document -0.0002870045081974301 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document -0.00019812241927078482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document -0.0002817553333369554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document -0.00027829782796642117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document -0.00028289431732284113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document -0.0002795526296717729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document -0.00027682829988044574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document -0.0002895432402719184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document -0.0002823174903941811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document -0.00028170972351837796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document -0.00027807915877838826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document -0.00028588515681452956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document -0.00028112324090816726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document -0.00020636178289985485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document -0.00019447255290980535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document -0.0002850824220591452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document -0.00027856429520116784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document -0.0002820880676635633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document -0.00028943902215995714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document -0.0002676366291085329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document -0.00023806333809954687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document -0.00024526460430233455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document -0.00023876876664622726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document -0.00023379770334179805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document -0.00024175151269138382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document -0.00023386583242595706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document -0.00023771797150160827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document -0.0002262748967483896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document -0.0002408148346432682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document -0.00023398651720444235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document -0.00022989433874474592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document -0.00023948500543957772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document -0.0002331594076859196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document -0.00023375132439600242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document -0.00023923410909668642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document -0.00023952796315562954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document -0.0002327466076905069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document -0.00023082758956797212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document -0.0002240509275524448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document -0.00022798879995765268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document -0.000221172516774386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document -0.00021767045123534623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document -0.00021982832794804484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document -0.00021971626543789102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document -0.00022566565206920132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document -0.0002181984894194856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document -0.00021831417549554653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document -0.00021601405421187145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document -0.00022275733725519607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document -0.00021847734911973986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document -0.0002243591012664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document -0.00021688758139483833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document -0.0002182953624789215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document -0.00020475155724026002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document -0.00021498078062960065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document -0.0002157914337233064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document -0.00021781838494967963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document -0.00021723242266814558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document -0.0002176782686553837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document -0.0003486179404943968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document -0.00034882846352857634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document -0.00031400868448352596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document -0.00030273484020011963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document -0.00029895889118145404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document -0.00029770764609621714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document -0.0002990181332116852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document -0.00029653733972285996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document -0.00029624649222942476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document -0.00029625609720203576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document -0.00029731928930852147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document -0.00029011721326148513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document -0.00028849788197494655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document -0.00021601278623858145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document -0.00021319599281739178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document -0.0002153325290600083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document -0.00018566946174516558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document -0.00020736824394291617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document -0.00020857419820128004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document -0.00020058526129536423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document -0.00020745812166665217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document -0.00020652171015271702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document -0.00020643808911278608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document -0.00020040513914482103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document -0.00020598050188272898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document -0.0001969184139343296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document -0.0001972748812937012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document -0.0002038556751586195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document -0.00020245186011313464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document -0.00019950381422038783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document -0.00020837055459665258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document -0.00020371856218246096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document -0.00019537612301625791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document -0.00019914984508813857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document -0.0002053787713691309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document -0.00019082100541008637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document -0.00020397153334531813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document -0.0002021462693077317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document -0.00019609357008124035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document -0.00019693256622486236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document -0.00020007239732428112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document -0.00020467075741591954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document -0.00019584883400022932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document -0.00019135050391176972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document -0.0003362829834208298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document -0.00034013691154784095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document -0.00033215887031941976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document -0.00032681189065396707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document -0.0003149138485493094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document -0.00030179177307540077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document -0.0002923278437581119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document -0.00029470052278994486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document -0.0002994095093045731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document -0.00029033525096085037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document -0.00029390798852496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document -0.0002916230924130842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document -0.00029419886374594913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document -0.0002865469756730764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document -0.00021191292549942086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document -0.00021369664817409847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document -0.00021612485624266726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document -0.00022242192634588478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document -0.00014605095659989698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document -0.00022070626106341693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document -0.0002174420774054071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document -0.00021325858963116995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document -0.0002124322999488052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document -0.0002081218896969054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document -0.0002108710211556957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document -0.00020686867095978426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document -0.00020895752681041895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document -0.00020741922266415738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document -0.0002069112657197308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document -0.00020644627473468118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document -0.00020332991338121604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document -0.0003560895677789848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document -0.00032915779111908214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document -0.00033810613317040864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document -0.00033729626594036923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document -0.00033550342864602944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document -0.00034173474024556906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document -0.000331505340748827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document -0.0003270050330117195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document -0.00032585275329172556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document -0.0003143383203190604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document -0.00031655199110388894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document -0.00030738872158476413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document -0.00030838388352699285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document -0.0003053596995351888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document -0.00031836304739584593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document -0.000315315435873905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document -0.0003087116248965243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document -0.00030396790625537645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document -0.0003335812246032149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document -0.00034570956323095843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document -0.00034563035636675786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document -0.00033411265479076335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document -0.00034439191141692787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document -0.0003364483125496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document -0.0003299500453608033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document -0.00033163377700074837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document -0.00032638649660627673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document -0.00032616167939645234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document -0.0003205289298760723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document -0.00031939393740815355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document -0.00031593164066731296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document -0.00031928871111254405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document -0.00029670189073175004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document -0.00020517703846735904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document -0.00020128418186172073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document -0.00019662723895606717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document -0.0001981157042081407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document -0.00019703489037041608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document -0.00019079796331785068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document -0.0001909352306690079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document -0.00018824662295261396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document -0.00019864275319325954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document -0.00018818516521649587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document -0.00018875694972812844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document -0.00018231621170645482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document -0.00018349407845798273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document -0.00018088971427746906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document -0.00018296284236327237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document -0.0001876011825819916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document -0.000329052068725176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document -0.00032223616273648536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document -0.00031272564089633955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document -0.00031621609908414494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document -0.0003117213560911235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document -0.00030218064069945934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document -0.00030658916600512085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document -0.0002915863534115821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document -0.0002940280138374372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document -0.00029067860468866085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document -0.00028529228063135635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document -0.00028336893301452256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document -0.0002794668089130099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document -0.00021681361378827842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document -0.0001484664674497246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document -0.00021950558378215133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document -0.00021806860758808645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document -0.00021819568718852282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document -0.00021626925931585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document -0.0001464536143077762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document -0.00021432777088808917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document -0.000213473805865147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document -0.00021397067253964538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document -0.00020758957647437263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document -0.00020687124337683314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document -0.00020630057046511005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document -0.0002091166859352538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document -0.00020777355025615267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document -0.00020709287641496176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document -0.00020736464660577094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document -0.00020062246741862607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document -0.00020693207561942915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document -0.00021151004871893024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document -0.00019930249098689716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document -0.00021589710041231824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document -0.00021369204789905741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document -0.0002147099923936778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document -0.00021077531190389536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document -0.0002100509829113836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document -0.00021185362601571124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document -0.00020722136637339565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document -0.00020300093701169531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document -0.00019859737993313477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document -0.00019971314372100164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document -0.00019549908270269278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document -0.00019649820843534028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document -0.00019619415513498067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document -0.00019493006120377898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document -0.00019499409035775506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document -0.00019252988593634277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document -0.00019440768268686405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document -0.00018747161324755577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document -0.0001879575932372779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document -0.00019040707058357506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document -0.0001871931095090703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document -0.00020112966223017096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document -0.00020516878165311017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document -0.00020664735191740533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document -0.00021041398572882962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document -0.00020397992929690396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document -0.0002039978580295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document -0.00020592785601142126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document -0.0001990755527445265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document -0.00019729564847798732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document -0.00019958182230527032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document -0.0001985037302636386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document -0.00020204130355115716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document -0.0002000296401958085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document -0.0001983064832295463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document -0.00019663108484195617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document -0.00019510678560556523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document -0.0001873284057063206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document -0.00019311553072495885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document -0.00034652137288816547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document -0.0002813690318850024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document -0.00027697649713138685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document -0.0002755419092534421 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document -0.0002681583054440219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document -0.00026945753192750824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document -0.00026169470768245737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document -0.00026437008960810825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document -0.0002637294838228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document -0.00026491867965088836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document -0.00025504483625138986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document -0.0002545040623796586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document -0.0002546682814073622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document -0.00025545439487142615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document -0.0002626896557978271 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document -0.00025092040940402784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document -0.0002589154885863872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document -0.00024106160482721467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document -0.0002483289690087987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document -0.0002388930282784437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document -0.00024006340759273874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document -0.00023765248178029045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document -0.00023061351965578936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document -0.00024954224883546477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document -0.00017861017233018525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document -0.00017810832743667658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document -0.00017599709170759497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document -0.00017462723516505223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document -0.0002906316527068669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document -0.00033762141066247166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document -0.00017170670574152494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document -0.00017258674515137717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document -0.0002815386173173926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document -0.0002996845935618989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document -0.0002735268488987296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document -0.0002971738713071517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document -0.0002942690674002763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document -0.0003322222207729567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document -0.0003378721656198464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document -0.00018307262621851067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document -0.00033956081502775057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document -0.00031604820927876276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document -0.00028805657681088917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document -0.00026312293321215633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document -0.00034366936722921455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document -0.0002865256504406559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document -0.0003063615195861786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document -0.00028412791619666136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document -0.00028060835132727154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document -0.00032544974761560506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document -0.0002647177833217225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document -0.0003152621884896575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document -0.0003054625140336913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document -0.00031183308312292263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document -0.00018175026696621178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document -0.00017699918328872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document -0.00018222339261441908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document -0.00018348005930964137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document -0.0001810735993810541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document -0.00030846441282038914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document -0.0002972326889310354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document -0.00017433421318235594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document -0.00032799458649525895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document -0.00032482130048512673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document -0.00031943465668672475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document -0.00029615593630484517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document -0.0002893126939511001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document -0.0002849288351723284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document -0.00028383906633569267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document -0.00028072526091262615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document -0.000284239564292377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document -0.0002778903109432523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document -0.0002771644389501471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document -0.0002733316182319337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document -0.00026362539185869363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document -0.0002636325383220217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document -0.00026740622442302886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document -0.0002646771971853427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document -0.0002628566720605389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document -0.0002644760695434766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document -0.0002623837702310999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document -0.00026088722976772894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document -0.0002567065374799158 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document -0.00018857382101207726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document -0.00019036580399817203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document -0.00018348828065261222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document -0.00018491851780345073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document -0.00018904887260080187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document -0.0001875609304251801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document -0.00018393034720015817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document -0.00018419795526114903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document -0.00018699955623404795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document -0.00018276256902965128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document -0.00017698045695190812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document -0.00018104650132303642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document -0.00017758206731279688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document -0.00017131402995103497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document -0.000175944428350446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document -0.0003416745727147391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document -0.0003163259373952889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document -0.0002804489269172448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document -0.00028748272397403175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document -0.00027603318345630605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document -0.000271638824679648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document -0.0002763761210210942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document -0.00026501984873172717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document -0.00026422486894694714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document -0.0002686339100849262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document -0.0002610837453940606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document -0.000260974343729353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document -0.0002599403837029134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document -0.0002937273113238609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document -0.0003341790732600504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document -0.0002620661576600244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document -0.0003027929169239288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document -0.00031944039129326894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document -0.00019025676304139009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document -0.00018680910145009907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document -0.00034215840419416437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document -0.00018618120812119364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document -0.00018605853095599425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document -0.00018120712626096538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document -0.00018315079292495327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document -0.00018362556449041974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document -0.0001780024456718171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document -0.00033296526436178697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document -0.0001802398632282846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document -0.00017340263100798256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document -0.00017755840547238697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document -0.00018419413735260606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document -0.00017869518174591322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document -0.00017526271460129484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document -0.00017852168597981907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document -0.00017566536156787157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document -0.00017589867964432936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document -0.00017831487394075305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document -0.00017837310528935862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document -0.00018200908814216548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document -0.0001795136627511612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document -0.0003414021775300033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document -0.00017177291787788502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document -0.0003441900648571877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document -0.0003394534597060673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document -0.0003236887233114832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document -0.0001639544129688747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document -0.00019137443753211255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document -0.00018575146284680153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document -0.00019184792863440243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document -0.00018966043065679055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document -0.00017968851317035848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document -0.00018479881897661546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document -0.0001813642692683015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document -0.0001686449798983066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document -0.00018516104592230446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document -0.00031283726601066385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document -0.0003248607542883853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document -0.00031583241601202365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document -0.00031238270857730376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document -0.000307150592403979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document -0.00029443829986847044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document -0.0002942723732234677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document -0.00023514930666443422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document -0.0020776328951453444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document -0.0021768234410538883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document -0.002106973549276289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document -0.002110915756171751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document -0.0017032382109816464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document -0.0019047944877712286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document -0.0019402711744016077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document -0.0006264790011223686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document -0.0017885401938106643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003742481815405742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc diff --git a/ALCF/data-lists/sunspot/dolma.txt b/ALCF/data-lists/sunspot/dolma.txt new file mode 100644 index 00000000000..7015d885c1f --- /dev/null +++ b/ALCF/data-lists/sunspot/dolma.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train +0.0002583902668716813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv +0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0002406272620255565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 +0.0003742481815405742 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc +0.0003547982093445404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon +6.322825248625475e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika +0.001451215788905126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train +0.0012499632072059553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o +0.0005759963691850877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit +0.0009994361338078242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange +0.004474659408857016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder +0.00032927705604725614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu +0.003548077173506675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki diff --git a/ALCF/data-lists/sunspot/falcon.txt b/ALCF/data-lists/sunspot/falcon.txt index 0b2fd6d43f7..e5afb892839 100644 --- a/ALCF/data-lists/sunspot/falcon.txt +++ b/ALCF/data-lists/sunspot/falcon.txt @@ -1,501 +1,501 @@ -0.0003547982093445404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document -0.00035934014428504944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document -0.00035707704501371544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document -0.00035287930712815354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document -0.00035977166728996823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document -0.0003581675664109838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document -0.0003548617059697185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document -0.0003639582000286208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document -0.00035375839698688127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document -0.0003743722020080678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document -0.0003530399715341242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document -0.00035511875882752406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document -0.0003618733574783154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document -0.00035185243285420104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document -0.0003541503739732106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document -0.0003631679485751914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document -0.00035748045578182274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document -0.0003606490690555877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document -0.0003626383296610091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document -0.00035442644361264756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document -0.00035978370170539796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document -0.0003585562375341541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document -0.0003601958372888019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document -0.000350277765402227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document -0.0003616521184211704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document -0.0003620625543608188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document -0.0003560781983850704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document -0.0003553209610592676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document -0.00035905348643915075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document -0.00034744258805696526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document -0.00035462784035661496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document -0.00034768186175100895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document -0.0003568534635532736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document -0.00035586511544371234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document -0.0003524567827568137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document -0.0003512453770426313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document -0.0003591792726468799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document -0.0003514024529343127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document -0.0003584880112586934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document -0.00035133552916418045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document -0.0003600811981350215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document -0.0003571663974228119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document -0.00035768103378874214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document -0.00035939205561113694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document -0.00035186773916029825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document -0.0003542829672490847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document -0.0003592783642898726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document -0.0003556367340099302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document -0.00035391392271377027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document -0.00035486725707484836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document -0.00034866743396828035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document -0.0003517219808644735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document -0.00034874458549673823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document -0.000355773136961014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document -0.00035611750387841917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document -0.00035305602013916315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document -0.0003578207127071924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document -0.00035514635841943707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document -0.00034816946212866206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document -0.0003512707269761496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document -0.0003483392117980654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document -0.0003572169607204321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document -0.00035139153281660794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document -0.00035536422129036537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document -0.000352017164107143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document -0.000351889550179365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document -0.000358759689953589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document -0.0003569286079869268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document -0.0003657752958602099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document -0.00035396127934790697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document -0.0003618565071224743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document -0.00035146051531973204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document -0.00036107135765783567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document -0.00035019554279994576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document -0.00035567858879904983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document -0.0003504753174793183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document -0.00035931140831329194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document -0.0003502967866002823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document -0.0003532911801041972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document -0.0003583543013070199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document -0.0003566243489931224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document -0.0003468752314799221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document -0.0003597840618138091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document -0.00035128822484768084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document -0.00035889496943437507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document -0.000352400524650424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document -0.0003518689536768735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document -0.00035866864741303467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document -0.0003454687659106334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document -0.00035348007259317576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document -0.0003539752270940644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document -0.00035146495994081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document -0.00035397212846310423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document -0.00035208246467162587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document -0.0003490843168676626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document -0.00035299633658644394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document -0.00034868327466167065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document -0.00035941351365601583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document -0.0003545343062735255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document -0.0003528956380445978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document -0.0003553355770443352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document -0.0003644224004937743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document -0.00035234291036216907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document -0.0003596237469847771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document -0.0003531996065735989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document -0.0003547177054106099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document -0.0003575586499260483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document -0.00035262635135283667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document -0.0003624191962188944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document -0.0003488398052948616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document -0.0003598294093147917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document -0.00035583006534466323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document -0.00035403139653225103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document -0.00036134702642187156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document -0.0003573689927162834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document -0.0003577141131435527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document -0.00035208814419277406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document -0.00035996720683665625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document -0.00035415304658912596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document -0.00036353353029443546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document -0.0003537326003150983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document -0.00036053976358299083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document -0.000352380489373494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document -0.00036154661616900994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document -0.00035959332325963614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document -0.0003597954667189692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document -0.0003563108270597542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document -0.0003582891940460143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document -0.0003497728210484297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document -0.0003549834902179354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document -0.0003529828233484542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document -0.00034627483903285777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document -0.00035569006572589215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document -0.00035449377946910314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document -0.00035802844396194623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document -0.0003617277809353208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document -0.00035034118898654814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document -0.000351091193908611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document -0.0003527914342210668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document -0.00035028288369781376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document -0.00035775745592780506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document -0.0003449630690661468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document -0.0003583490698830361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document -0.0003476995746684122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document -0.0003535632505019212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document -0.00035640180641147417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document -0.000361731045691765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document -0.0003534082129597368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document -0.0003550344149828664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document -0.00035363002411364057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document -0.0003537265579677396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document -0.00034950531383577937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document -0.00035008511827347514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document -0.00035594533400871325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document -0.00035266312861335946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document -0.00035280268794863923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document -0.0003565470391528536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document -0.0003588492322689137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document -0.00035469909697832775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document -0.00034712082813410526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document -0.000348701157101807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document -0.0003500192014479944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document -0.00035120560544669755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document -0.00035403656850437445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document -0.00035852376560749366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document -0.0003534754068111774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document -0.00035591740046720765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document -0.000348522354782563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document -0.0003533533959664415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document -0.00035631425964030697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document -0.0003485886551574741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document -0.00035917652631065777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document -0.0003482975272111288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document -0.00035580661277480167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document -0.0003492290722955348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document -0.00034989284450240613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document -0.0003545677216162781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document -0.00034622286859463484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document -0.00036070626989861965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document -0.00035518365036320786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document -0.00035272907057848406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document -0.0003547343638218734 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document -0.0003496450144966242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document -0.0003537407829294287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document -0.0003489722653985685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document -0.00035057186899911295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document -0.0003507566548933051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document -0.00035630360179023747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document -0.00035631362503416367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document -0.0003490204248026821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document -0.00035761724058371226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document -0.00035037664777467137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document -0.000353402110481068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document -0.00034524163568371745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document -0.00035528523728570974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document -0.00034784916132431703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document -0.00034928476408048925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document -0.00034989205973784984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document -0.00034201664404094254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document -0.0003529676016338611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document -0.00034643433682346637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document -0.0003511666373001904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document -0.00034828669066575333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document -0.0003494625207264413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document -0.0003458957535879216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document -0.0003543020478990003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document -0.00034754384069014956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document -0.0003598856392240133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document -0.0003503335458553846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document -0.00035919595619778716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document -0.00035767737970754404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document -0.00035197152783998165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document -0.0003549609834422404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document -0.0003568184100569753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document -0.0003512652818651935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document -0.00035912648958665754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document -0.00034764526964056546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document -0.000352439784960359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document -0.00035295886560764226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document -0.0003518132693658672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document -0.00035589987915465713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document -0.00034923863317385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document -0.0003457987267929692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document -0.0003560928663480501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document -0.0003529603811204932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document -0.0003524438555443043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document -0.0003438847030263783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document -0.00035981978898461613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document -0.0003446342778566972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document -0.00035529584995236537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document -0.00034855740895831116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document -0.00034932634912802544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document -0.00035805518303064666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document -0.0003497941877073061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document -0.00035774398685405447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document -0.0003560421780316607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document -0.0003508844468369392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document -0.00035731928892270107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document -0.0003557884626314314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document -0.00034992996760289355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document -0.000360752554360921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document -0.0003452321668708545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document -0.0003591745226131023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document -0.00035256981433229084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document -0.00035378123159712034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document -0.000350464354895999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document -0.00035074625557389677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document -0.00035025894701994667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document -0.00035437902514857614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document -0.0003514684519732232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document -0.00035449717909633905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document -0.0003436816402714221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document -0.00035139158071782116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document -0.0003509424079843335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document -0.000343894618577506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document -0.0003500789770661659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document -0.0003407788080680086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document -0.0003581908175239701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document -0.0003465541618780918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document -0.00034600228792437736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document -0.00034416738982773204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document -0.0003519900340150641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document -0.000343369616864659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document -0.0003544993883274688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document -0.0003504441365073392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document -0.00034859160702727056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document -0.00035355909532647185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document -0.0003471900922691849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document -0.0003563015508709187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document -0.0003487888744148821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document -0.00034711767548688336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document -0.0003530734609369085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document -0.00035123969242560935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document -0.0003517127620891489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document -0.00035232835416868673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document -0.0003524437481912308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document -0.0003525996167005602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document -0.00035064770545242043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document -0.00035311558274981226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document -0.00034952204800569914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document -0.0003541471367344846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document -0.00035418812454561825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document -0.0003528951372900714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document -0.0003542338042975688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document -0.00034937738939942796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document -0.0003522182190878447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document -0.0003501406466507449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document -0.00034973079877492633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document -0.0003485274567713538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document -0.00034999308679368985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document -0.0003570051724707296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document -0.00034567230462019706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document -0.00035529000940160696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document -0.00034956512308671755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document -0.0003496962834028953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document -0.0003468745282493457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document -0.0003502717155809202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document -0.0003556240880896514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document -0.0003515109488424343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document -0.0003563156688192592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document -0.00035040277363989817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document -0.0003481408593290717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document -0.0003624575124332874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document -0.0003522684124250313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document -0.00035286996027653544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document -0.00034967623997256725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document -0.00035182649587602765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document -0.0003524892557026489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document -0.0003507642477451811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document -0.00036190408389835666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document -0.00035102739424880766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document -0.00035239718753257265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document -0.00035298076121821316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document -0.0003478704389752654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document -0.0003503109191567942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document -0.00035143250975654426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document -0.0003480663923069012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document -0.00035691540219998623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document -0.000348815437166351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document -0.00035202073257766225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document -0.0003491569096274706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document -0.00035277390475511834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document -0.0003524972090026609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document -0.0003504854249750236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document -0.00034740238025423914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document -0.00034968015462277606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document -0.0003493798632762674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document -0.0003488202537862122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document -0.0003525461864643725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document -0.00034903815232825664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document -0.00035536982539258216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document -0.00034858083265155483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document -0.0003505014973608067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document -0.00035327984042622104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document -0.0003503286677453136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document -0.00035835274842442816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document -0.00034970302660275595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document -0.000357929573140149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document -0.0003517238649788585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document -0.00036097027318848475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document -0.0003502734074110026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document -0.00035801510806036273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document -0.0003568006373479869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document -0.00036128108717454636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document -0.0003563436883111686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document -0.00035559725321852463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document -0.00035089656006854944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document -0.000359453964362057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document -0.00035629498059104033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document -0.0003622207707090437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document -0.0003540946784512821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document -0.0003594750565232011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document -0.0003566007415086991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document -0.0003562142599126134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document -0.0003569948186744601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document -0.00035166554847920186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document -0.00035047994419295137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document -0.0003561578193739437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document -0.00035470866838811544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document -0.00034216920464876335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document -0.0003550021513075795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document -0.0003488045105938729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document -0.0003513340720840151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document -0.0003448558566387584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document -0.0003460966026953241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document -0.0003488157616036459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document -0.0003446120387842362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document -0.000351528602987427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document -0.00035661118227454713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document -0.0003551342699877457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document -0.0003478953397924445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document -0.00034625782458988215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document -0.0003527515447405871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document -0.00034823744889805696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document -0.00034823314560254406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document -0.00035162668292961944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document -0.0003477307716074623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document -0.0003446457989477787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document -0.00034782916273767795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document -0.0003517249130302248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document -0.0003449873430908556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document -0.00034841291749669877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document -0.0003466028498941749 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document -0.0003486436831199424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document -0.0003478279234211838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document -0.0003495903653274374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document -0.00034896893881218957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document -0.000348941645312426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document -0.0003474221308416894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document -0.0003462621543839385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document -0.0003669373860863891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document -0.00034691156268163006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document -0.0003527774103765281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document -0.00034684565672734663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document -0.0003454250599604457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document -0.0003541536557159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document -0.000345735737037366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document -0.0003524669816385214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document -0.0003441817133096468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document -0.0003519093265859089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document -0.00035080085480352095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document -0.00035285227929327434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document -0.00034354836346901676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document -0.00034789770937373467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document -0.000343665920520102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document -0.0003490884931060568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document -0.00034380029463398654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document -0.00034874768005099945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document -0.0003457058510967673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document -0.00034644265227023904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document -0.00035008339858594957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document -0.0003462377193296194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document -0.0003620491787114201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document -0.000348717011044469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document -0.00034370072363913706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document -0.0003551981066775649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document -0.0003500119496799342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document -0.0003485082952669081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document -0.0003508155580978919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document -0.00035311375163251416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document -0.00034945972003423253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document -0.0003474220353789879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document -0.0003536443686585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document -0.0003560350489042953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document -0.0003493655927914396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document -0.0003528423977146383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document -0.00035255554724471217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document -0.0003479760010190111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document -0.00035458598862501956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document -0.0003458990560538315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document -0.00035157946422379875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document -0.00034736860650169996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document -0.0003529152313394119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document -0.00034586294329524465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document -0.00035707214923794877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document -0.0003509580363496512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document -0.00035244176725524474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document -0.0003467539557999047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document -0.00034919687962275546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document -0.00035094031731719953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document -0.0003484309008351352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document -0.0003485409424916253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document -0.0003499590776117838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document -0.0003492842758957848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document -0.0003529712275178912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document -0.0003566141287087449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document -0.0003649496522047409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document -0.0003563218912208234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document -0.00035614782126966145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document -0.0003531944298453266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document -0.0003535950949566616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document -0.0003544295554928795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document -0.0003519908503740376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document -0.00035752817626134463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document -0.0003515322689589972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document -0.0003486893890307115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document -0.0003446520464889867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document -0.0003509421562481707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document -0.00035335015702909084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document -0.0003490178167345008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document -0.0003520497821155174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document -0.0003549762618908944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document -0.00035072190850833103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document -0.0003542458638526423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document -0.000352419194572916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document -0.0003545102564672614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document -0.0003495437992331806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document -0.0003542843376993964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document -0.000352827529313958 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document -0.00035442506093223886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document -0.0003496970719044257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document -0.0003553096424442362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document -0.00034986845565067564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document -0.000352131055186658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document -0.0003527021708198983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document -0.00034905885414547214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document -0.0003583433842468394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document -0.00034409435202828383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document -0.00034846410520871483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document -0.0003554459991927314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document -0.00035310507471843076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document -0.000350028910786098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document -0.00035049727458009896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document -0.0003519047735925826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document -0.0003513027429919726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document -0.0003626947260354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document -0.0003500087324849783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document -0.0003618315726725285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document -0.0003535385113938023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document -0.0003487064058517615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document -0.0003618709124780938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document -0.00035040070335625915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document -0.0003506279032267829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document -0.0003498435310527524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document -0.0003554634749821431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document -0.00035091209738758963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document -0.00035034103678978573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document -0.00035398931854386146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document -0.00035495529304989485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document -0.00036067883473356603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +0.0003547982093445404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon diff --git a/ALCF/data-lists/sunspot/megawiki.txt b/ALCF/data-lists/sunspot/megawiki.txt index 9fc9ca5dab8..f7fbabc9134 100644 --- a/ALCF/data-lists/sunspot/megawiki.txt +++ b/ALCF/data-lists/sunspot/megawiki.txt @@ -1,262 +1,262 @@ -6.322825248625475e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document -2.4432314037946264e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document -5.6313888721313454e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document -2.4208171781595055e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document -2.325811856369237e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document -2.4010790356322705e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document -5.36773610843632e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document -1.360574433501002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document -1.3076540344853244e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document -1.3386534334886313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document -1.2498103719605153e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document -1.403763836949682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document -1.3636756723495417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document -1.2242489446940814e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document -1.2398255818973339e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document -1.2972616994216281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document -1.3947809855914134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document -1.3144843787829514e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document -1.1693809976572487e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document -1.3677252682893802e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document -1.3940876719849597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document -1.4222245138730965e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document -1.3201677767919704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document -1.1421717796486169e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document -1.2890514724498703e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document -1.3649507648749037e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document -1.2400732563490717e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document -1.1557681453277616e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document -1.2294483595964517e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document -1.2137484472122283e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document -1.3299663426456e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document -1.2461984216479532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document -1.4666434217609636e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document -1.1876997894686238e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document -1.2939155338964078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document -1.3859590039728515e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document -1.317917848615668e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document -1.1335281536110342e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document -1.2889923952861426e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document -1.3471671647053326e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document -1.2221720014475102e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document -1.2632647276287541e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document -1.28276219004076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document -1.36213704321643e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document -1.2414858625261553e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document -1.3173700421883744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document -1.295597796725686e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document -1.242783936442904e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document -1.2417374088427464e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document -1.2134479405400744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document -1.3090040663304255e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document -1.2713470581614905e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document -5.5750231378906594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document -5.777597358425469e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document -5.349786767471258e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document -5.675165050453583e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document -5.482611216158831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document -5.065421899890121e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document -5.384718357480146e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document -4.872037363236061e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document -4.532709250783155e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document -5.7257963030489613e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document -4.9014365579652036e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document -5.722863552770969e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document -6.149911636146833e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document -5.2178057608273506e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document -4.990228161160431e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document -5.866186875255134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document -5.004185734360719e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document -4.79401853705107e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document -5.435219965052376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document -5.035997225792266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document -5.622401774211625e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document -5.028826157387559e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document -5.596379470128795e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document -6.027824493191489e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document -5.5358270009931474e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document -5.9839051807685496e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document -5.1221077499249595e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document -5.517228560620279e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document -5.1687858285052305e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document -5.684188244145645e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document -5.212693275535878e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document -4.8551007022784084e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document -5.4888506639203145e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document -5.345098688527242e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document -4.8506420625516594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document -5.132168603397676e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document -5.719476795114223e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document -5.7448621149792696e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document -4.9068410568059265e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document -5.382937299647678e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document -4.8288432136304634e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document -5.841703200305416e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document -5.1589611587885584e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document -6.031113829732574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document -5.4558202844532094e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document -5.341852317196142e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document -5.1402942738369954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document -5.735421384377395e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document -5.473629863586958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document -5.4708993245733936e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document -4.931161863634078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document -5.104173022127248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document -5.510157161510824e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document -5.652501401782597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document -5.7273656573031666e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document -5.638363224821738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document -5.6128115396668704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document -5.00304877998141e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document -5.596120554779096e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document -5.5280923889040006e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document -5.223477917938408e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document -5.29472809986569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document -2.205682378243213e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document -1.4367563720603185e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document -3.5506193487931076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document -3.0442910855821778e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document -2.2540042508019627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document -2.6880163202623216e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document -2.534473148048727e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document -2.6560945431318916e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document -2.547470248967691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document -2.5248825388073738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document -2.5828729575000054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document -2.4026583817957736e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document -2.3930425429834413e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document -2.5037365362599724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document -2.6696745470595603e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document -2.140323051341762e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document -2.617354786691592e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document -1.538359101762691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document -1.2871029252377856e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document -2.255195411289217e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document -2.4832313897952067e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document -9.303873918189968e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document -2.179532302620228e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document -1.9750517506901206e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document -2.7740420380648435e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document -2.7813714782319335e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document -4.1595357937609806e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document -2.741365122389175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document -2.117451071361901e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document -1.7132649760565998e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document -1.7492547092602047e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document -1.7499951097392276e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document -1.6632444789170958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document -1.6678802252361607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document -1.5519208704558896e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document -1.652420992967167e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document -1.6119931034508755e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document -1.6638882076736552e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document -1.7198076782652946e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document -1.572927860565175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document -1.5194822618169918e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document -1.6677776832669846e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document -1.595612492245688e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document -1.682350633181197e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document -1.663983380609724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document -1.710187842689243e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document -1.5733697527539038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document -1.6972104757911438e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document -1.6610142847616577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document -1.61094882403031e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document -1.4789207305138325e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document -1.639299617676302e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document -1.3241204512116132e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document -8.582260726625535e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document -8.213000975576739e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document -9.549247732811947e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document -9.17242785339013e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document -7.632868223725218e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document -8.674401118222175e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document -9.124384255505347e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document -8.344222222417358e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document -8.992299957499065e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document -8.76689497361025e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document -7.973396239586015e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document -9.006935606644125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document -8.725545954955498e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document -1.215449694669174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document -3.3041720284158646e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document -2.0593512412624502e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document -1.893608946986248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document -1.737111666788535e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document -1.4915923449873955e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document -2.289370239067605e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document -2.8615335689614638e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document -8.847283630883125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document -1.8175470362373804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document -1.8152226683368038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document -1.789149655314284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document -1.7690523036477663e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document -1.8333732213753644e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document -1.8794105687718654e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document -1.721841156706417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document -2.0612008685724796e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document -1.9297370681336376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document -2.0188440409661018e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document -5.1741216329695265e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document -1.3417913926038429e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document -1.1010813016469651e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document -1.1252416134320087e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document -1.2801744104313002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document -1.3041514955795817e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document -1.3428837580879075e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document -1.320809382267804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document -1.3451566676555968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document -1.228284926657501e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document -1.2410599573923043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document -1.3815343367377182e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document -1.3895126265148832e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document -1.2306773644401741e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document -1.32981021906281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document -1.101337469221607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document -1.513094184404692e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document -1.1073759547073234e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document -1.2879348765857567e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document -9.619595770228435e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document -1.2384340836286436e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document -1.1766667232211577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document -1.2871049236196452e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document -1.2010645926497744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document -1.3971428231518597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document -1.2283733550547932e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document -1.2659530508255308e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document -1.551775613074462e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document -1.1169413343776979e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document -1.1433700593712463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document -4.964773647323492e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document -1.0995586595687313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document -1.2957393071411267e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document -2.75899247407709e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document -2.8269344597344854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document -2.329108187246831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document -2.4231761430460284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document -1.2434140512230442e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document -1.638718338352859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document -3.272953556801187e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document -6.061314500486327e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document -1.2465979731210292e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document -1.2737557327967737e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document -1.038428658075627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document -2.61666472045566e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document -3.6506873212272224e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document -1.5066359138295701e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document -1.1166290872121178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document -1.5546966228590285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document -1.2583434625014828e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document -1.3398826881300862e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document -1.2944933160515968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document -1.0971437399901365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document -1.2787922795775774e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document -1.404979227816985e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document -1.3344734431324463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document -4.886031157107555e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document -3.277261443596394e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document -3.5057957685786495e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document -3.287625301718589e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document -3.1370056372668855e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document -3.186092015785841e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document -7.271819324142512e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +6.322825248625475e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika diff --git a/ALCF/data-lists/sunspot/open-web-math-train.txt b/ALCF/data-lists/sunspot/open-web-math-train.txt index b36e9977c05..ffa745cd766 100644 --- a/ALCF/data-lists/sunspot/open-web-math-train.txt +++ b/ALCF/data-lists/sunspot/open-web-math-train.txt @@ -1,13 +1,13 @@ -0.001451215788905126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document -0.0014486847196258788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document -0.0008861032722895899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document -0.0018119590809459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document -0.0008916937917547129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document -6.960128832809415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document -0.002008403651063623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document -0.0014374900742131454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document -0.00180213596996716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document -0.001956178877532413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document -0.0008829547017667033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document -0.0008910853619157279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document -0.0018260998845299973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.001451215788905126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train diff --git a/ALCF/data-lists/sunspot/pes2o.txt b/ALCF/data-lists/sunspot/pes2o.txt index 63f805c06d5..a2a1209d8c5 100644 --- a/ALCF/data-lists/sunspot/pes2o.txt +++ b/ALCF/data-lists/sunspot/pes2o.txt @@ -1,26 +1,26 @@ -0.0012499632072059553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document -0.00125398260359913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document -0.0012541704774729071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document -0.0012527268234360602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document -0.0012532925243737164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document -0.0012456396241204315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document -0.0012589894424352072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document -0.001508020123999618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document -0.00333096950781965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document -0.0033233414614415547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document -0.003512387990689828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document -0.0035091382940513126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document -0.003514155927147005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document -0.003327108000579638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document -0.003329106196589836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document -0.003505604148738077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document -0.003324825759567855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document -0.0033248240149804913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document -0.0033385962112851358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document -0.0035043186296553615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document -0.003340469505431529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document -0.0035106889084796276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document -0.0033309469281030167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document -0.003340337858029757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document -0.003505919861097801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document -0.0003882924098240512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0012499632072059553 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o diff --git a/ALCF/data-lists/sunspot/reddit.txt b/ALCF/data-lists/sunspot/reddit.txt index 59eafce1ee0..a1de492a2f2 100644 --- a/ALCF/data-lists/sunspot/reddit.txt +++ b/ALCF/data-lists/sunspot/reddit.txt @@ -1,78 +1,78 @@ -0.0005759963691850877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document -0.0005959971675332674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document -0.0006026179290353799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document -0.0005824184320784846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document -0.0005854598548616037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document -0.0005903767055633473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document -0.0005930306490982049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document -0.000569425602700746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document -0.0005675060415179408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document -0.0005772431621253389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document -0.0005678026053826858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document -0.0005700398263483378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document -0.0005669467963528824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document -0.0005701015953324305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document -0.0005795907287413296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document -0.0005735602737531164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document -0.0005749862745842101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document -0.0005693257015931971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document -0.0005716568794795563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document -0.0005761083919774021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document -0.0005688343169797355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document -0.0005807913190929842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document -0.0005710229258078636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document -0.0005704083039826862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document -0.0005862132348308056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document -0.0005717662049559556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document -0.0005858155213694451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document -0.0005812012281792392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document -0.0005803981414588498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document -0.0005700102108287723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document -0.0005719243459052329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document -0.0005867253401661752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document -0.0005731087218860733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document -0.0005712197789109317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document -0.0005702376926310089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document -0.0005700411527742972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document -0.0005828090098178196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document -0.0005770140826168056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document -0.0005723509664597896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document -0.0005755499231836962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document -0.0005636407438471367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document -0.0005640281556500104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document -0.0005633159058766496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document -0.0005638034311151449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document -0.0005630066273073224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document -0.0005631803831128559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document -0.0005631228881679657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document -0.0005628178701487633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document -0.0005624448092256196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document -0.0005620957024062329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document -0.0005614201504177484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document -0.0005616890951464056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document -0.0005611348559279058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document -0.0005604238061828518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document -0.0005603301490194237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document -0.0005607291294548833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document -0.0005605234569930727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document -0.0005613778566640694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document -0.0005610248539992471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document -0.0005599977416780475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document -0.0005603632562116935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document -0.0005599177479509897 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document -0.0005595202318298379 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document -0.0005600975633499175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document -0.0005614075491213365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document -0.000612563885043477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document -0.0005515469909644413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document -0.0005526782014946906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document -0.0005472463408095445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document -0.0005502284746004587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document -0.0005414514790555363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document -0.0005513499500134784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document -0.0005391391454105187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document -0.0005415836910001838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document -0.0005208132468536551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document -0.0005889827143132871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document -0.0005822520817765276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document -0.0004173155230758696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0005759963691850877 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit diff --git a/ALCF/data-lists/sunspot/stack.txt b/ALCF/data-lists/sunspot/stack.txt index 297783ac228..60cf4451ab6 100644 --- a/ALCF/data-lists/sunspot/stack.txt +++ b/ALCF/data-lists/sunspot/stack.txt @@ -1,26 +1,26 @@ -0.0009994361338078242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document -0.001087156194657966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document -0.0010667737163656816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document -0.0009602877882124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document -0.0008968956271971105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document -0.0009198034843762967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document -0.0009423901016715341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document -0.0009674094553686345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document -0.0009858331322519164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document -0.0009970593645879198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document -0.0010027035193731686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document -0.0010128291154221853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document -0.0010215631382631918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document -0.0010288663771461238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document -0.0010346219929285867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document -0.00104544019940344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document -0.0010525172676724333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document -0.0010609529620775127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document -0.0010725892748610153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document -0.0010818563598181568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document -0.0010992760196793917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document -0.0011178992762079917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document -0.001124687532085676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document -0.001118303661267191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document -0.0010206825575416534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document -0.0005512280117499715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.0009994361338078242 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange diff --git a/ALCF/data-lists/sunspot/starcoder.txt b/ALCF/data-lists/sunspot/starcoder.txt index 37e6333de51..0011e33989e 100644 --- a/ALCF/data-lists/sunspot/starcoder.txt +++ b/ALCF/data-lists/sunspot/starcoder.txt @@ -1,50 +1,50 @@ -0.004474659408857016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document -0.00409944473890653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document -0.005137179939941845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document -0.005143172251066109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document -0.005206134363352808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document -0.004892747858974329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document -0.004844731352552902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document -0.005308320169123755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document -0.005124709815666577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document -0.005424710744483826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document -0.00538244648861977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document -0.0029107284679086853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document -0.0026825258998444705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document -0.0026904503191419243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document -0.002687906577174073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document -0.002850165346048818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document -0.005322698571717847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document -0.004450334290869719 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document -0.004700990083440683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document -0.003903568556500995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document -0.00390561515396931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document -0.0039046402900912262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document -0.003907454839379547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document -0.0038583224578603824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document -0.0037914116657695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document -0.003786665266798682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document -0.003792000802430658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document -0.00319266847466091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document -0.0032658716699838944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document -0.0034801959532460023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document -0.0028307012092022594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document -0.0028420360878146276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document -0.0028410455248484914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document -0.00283497183526842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document -0.002840187195459487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document -0.0028398709431369834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document -0.004364722843422023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document -0.004093255713117101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document -0.004092331079566252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document -0.004005326985579649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document -0.0036205502856964207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document -0.003625316793034984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document -0.003604743435602363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document -0.0035405823343673125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document -0.0041601413517253945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document -0.005886303658937057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document -0.003600909532810332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document -0.0034941365817168658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document -0.0004992164842980224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.004474659408857016 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder diff --git a/ALCF/data-lists/sunspot/tulu.txt b/ALCF/data-lists/sunspot/tulu.txt index 2b758025015..b2e14257844 100644 --- a/ALCF/data-lists/sunspot/tulu.txt +++ b/ALCF/data-lists/sunspot/tulu.txt @@ -1,66 +1,66 @@ -0.00032927705604725614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document -0.0002860154190878753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document -0.0002845217585425619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document -0.0002743528685497456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document -0.00026025323737738766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document -0.00023493876414603155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document -0.00029665994994226705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document -0.00031808102075993956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document -0.00031813573046011285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document -0.0002711905171855542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document -0.00028892513401817095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document -0.00030003908676979083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document -0.00026839878771944684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document -0.00029155935002690497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document -0.0002998624927624209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document -0.0003091705447974841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document -0.00026873195794309786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document -0.00027721873498527547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document -0.0002841662554024377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document -0.0002839461156551537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document -0.0002861705604659811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document -0.0002460995649635886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document -0.00019420142619795496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document -0.00021967677816173628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document -0.0002620283200480949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document -0.0002433390542188936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document -0.00021254976608350767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document -0.00022094815569522115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document -0.000342862378668244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document -0.00033784225259118157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document -0.0003367278459543952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document -0.00029843279042852765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document -0.0002926583661257988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document -0.00029320337282010673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document -0.00029281450669483455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document -0.0002915338187002653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document -0.0002864226923084572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document -0.00028643439083586396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document -0.00028253710956299054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document -0.0002810856078805806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document -0.00031474941344656715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document -0.0002139130222205655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document -0.0003084648871862831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document -0.0003309477872140129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document -0.0003360096824695161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document -0.0003355452655196557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document -0.00038119390366386037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document -0.00038078927630086064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document -0.0003386200917551554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document -0.0002158905159938882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document -0.00021621682877018768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document -0.00021553306942740535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document -0.00021581563462722296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document -0.0002157694110556169 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document -0.000215643699847159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document -0.00021532716715168094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document -0.00021531221326022472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document -0.0002831801179028896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document -0.0002514844936507595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document -0.00031638782778107964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document -0.0002749197545278445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document -0.00026159721512464495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document -0.0002630052420096968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document -0.00031106811228913666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document -0.0002852973415334161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document -3.7555372465932136e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.00032927705604725614 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu diff --git a/ALCF/data-lists/sunspot/wiki.txt b/ALCF/data-lists/sunspot/wiki.txt index 52af00d57b5..172bd0323b0 100644 --- a/ALCF/data-lists/sunspot/wiki.txt +++ b/ALCF/data-lists/sunspot/wiki.txt @@ -1,2 +1,2 @@ -0.003548077173506675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document -0.0018372203137874265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document +0.003548077173506675 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki From a51fb1168517081766d3124eeb114161bcf1ba4b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:52:07 -0500 Subject: [PATCH 223/590] Update `ALCF/data-lists/polaris/*.txt` --- ALCF/data-lists/polaris/algebraic.txt | 32 +- ALCF/data-lists/polaris/arxiv.txt | 200 +- ALCF/data-lists/polaris/books.txt | 6 +- ALCF/data-lists/polaris/c4.txt | 342 +-- ALCF/data-lists/polaris/cc.txt | 2214 +++++++-------- ALCF/data-lists/polaris/dolma.txt | 2419 +++++++++++++++++ ALCF/data-lists/polaris/falcon.txt | 1000 +++---- ALCF/data-lists/polaris/megawiki.txt | 524 ++-- .../polaris/open-web-math-train.txt | 26 +- ALCF/data-lists/polaris/pes2o.txt | 52 +- ALCF/data-lists/polaris/reddit.txt | 156 +- ALCF/data-lists/polaris/stack.txt | 52 +- ALCF/data-lists/polaris/starcoder.txt | 98 +- ALCF/data-lists/polaris/tulu.txt | 132 +- ALCF/data-lists/polaris/wiki.txt | 4 +- 15 files changed, 4838 insertions(+), 2419 deletions(-) create mode 100644 ALCF/data-lists/polaris/dolma.txt diff --git a/ALCF/data-lists/polaris/algebraic.txt b/ALCF/data-lists/polaris/algebraic.txt index 505276d3bfb..394649fcf45 100644 --- a/ALCF/data-lists/polaris/algebraic.txt +++ b/ALCF/data-lists/polaris/algebraic.txt @@ -1,16 +1,16 @@ -0.0018520780893211373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document -0.0017591050606817512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document -0.001459052794333798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document -0.0007405667281569194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document -0.00019420030110896795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document -0.0009008668715801845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document -0.00015115827957143057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document -0.0014552844319220648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document -0.0012469861325685161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document -0.00136412011372413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document -0.0007064279699221103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document -0.0008472240000687427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document -0.0001984375713341955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document -0.0005472773881697123 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document -0.001815779629850992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document -0.0018313600689757324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0018520780893211373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train diff --git a/ALCF/data-lists/polaris/arxiv.txt b/ALCF/data-lists/polaris/arxiv.txt index cae6e2da693..85e59adacd0 100644 --- a/ALCF/data-lists/polaris/arxiv.txt +++ b/ALCF/data-lists/polaris/arxiv.txt @@ -1,100 +1,100 @@ -0.0002583902668716813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document -0.0002646575141232155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document -0.0003165521247456758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document -0.0002920706460176214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document -0.00028396813182810215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document -0.00030445161883108107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document -0.00031628781276576474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document -0.0003083776568189157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document -0.0003176359471472902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document -0.0002536009369131698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document -0.0003067491424681363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document -0.0002597217257557784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document -0.0003788556450109768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document -0.0002796563272052598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document -0.00033573826524290287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document -0.00030523658022800287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document -0.00032211552192240096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document -0.0003329295675164247 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document -0.0003101982186639862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document -0.00032361798234223355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document -0.0003495541581652915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document -0.0002821637448858042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document -0.00030399523537629673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document -0.0002955658968247219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document -0.00028942158502924254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document -0.00028769546171490733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document -0.0002938111057234182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document -0.0002711150403010948 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document -0.00031130095874747565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document -0.0003002996118160777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document -0.0003732757901604459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document -0.00026784205751795894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document -0.0002799626521661984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document -0.00034334276069078164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document -0.0003582469803674965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document -0.00031094844818418623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document -0.0002766228384977191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document -0.00030297116159471485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document -0.00027033888377464685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document -0.00030090862368377933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document -0.00028543875802490955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document -0.00027559768459074204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document -0.0003182185533962886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document -0.0003311392971435837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document -0.00028751652060804325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document -0.000303466863212589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document -0.00033400462801277524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document -0.0002589234031777426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document -0.0002913508598466723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document -0.0002670572450004856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document -0.00032027399105647656 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document -0.00032188376258379377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document -0.0003161585784100882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document -0.0003184249182974135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document -0.00030381336664000807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document -0.0003190437442184283 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document -0.0002537961798200545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document -0.0003017817117223326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document -0.00028685268513240224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document -0.00031265179094451165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document -0.00034708319096986816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document -0.00026650837943080664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document -0.00034588832248507335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document -0.0002416982248399037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document -0.0003089296918222243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document -0.00029137184185700827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document -0.00026464226846800774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document -0.00030545397919456627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document -0.0003206778460448875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document -0.00030968971641110967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document -0.00023325653928600864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document -0.00030526899198338555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document -0.00035376719076633584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document -0.000290224385981026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document -0.000294650083382008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document -0.00028768858128616436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document -0.00030856965235527843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document -0.00030579942447879054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document -0.0002863101084704357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document -0.0002870032092492213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document -0.000264182727569885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document -0.0002974012367036449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document -0.00032238412143059203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document -0.00031683716893819036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document -0.00031157434937617524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document -0.0003411742735695989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document -0.00026778444816570715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document -0.0003037045797275201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document -0.00027746114370081314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document -0.00027148285946862043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document -0.00028042950114678207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document -0.0003235607816590721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document -0.0003086692227306295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document -0.00033990349455148105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document -0.00030945053208470265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document -0.00027309074552265303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document -0.00028737393506316194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document -0.0003098868328009879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document -0.0002614229162588409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document -0.0002884388407820923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0002583902668716813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv diff --git a/ALCF/data-lists/polaris/books.txt b/ALCF/data-lists/polaris/books.txt index 195aca5339f..c222c32c071 100644 --- a/ALCF/data-lists/polaris/books.txt +++ b/ALCF/data-lists/polaris/books.txt @@ -1,3 +1,3 @@ -0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document -0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document -0.006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0031025147279277244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books diff --git a/ALCF/data-lists/polaris/c4.txt b/ALCF/data-lists/polaris/c4.txt index 833b0958829..512556eafba 100644 --- a/ALCF/data-lists/polaris/c4.txt +++ b/ALCF/data-lists/polaris/c4.txt @@ -1,171 +1,171 @@ -0.0002406272620255565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document -0.0002404825539493424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document -0.00024062296575435581 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document -0.00024069315766818953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document -0.00024055829162263452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document -0.00024062053397343032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document -0.0002410715545206964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document -0.00024024881846087368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document -0.0002407074700790688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document -0.00024072141428809043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document -0.00024027710230872736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document -0.0002409111299205489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document -0.00024081954058275009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document -0.00024086076794990912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document -0.00024098672620832446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document -0.00024068622303333862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document -0.00024140627024291824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document -0.0002414512033594384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document -0.00024028742594941463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document -0.00024018036089269645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document -0.0002398347365034979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document -0.00024006780153485276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document -0.00024015620270419213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document -0.0002408848259695227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document -0.0002408023185278831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document -0.00024021196580140326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document -0.00024077677271297493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document -0.00024087392454668027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document -0.0002408071293824126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document -0.00024042223828845715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document -0.0002411484752360495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document -0.00023605263746465907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document -0.00023471222158326908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document -0.00023432138580287644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document -0.00023407385623382327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document -0.00023487504174367091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document -0.0002341843704976313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document -0.00023421993170282486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document -0.00023445057969132037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document -0.0002337681680073047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document -0.000234627964808109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document -0.0002338942211888584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document -0.00023403849286843386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document -0.00023405641310796305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document -0.00023349169562397965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document -0.00023381157386048856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document -0.00023388742993790587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document -0.00023363103829469813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document -0.00023421141834630477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document -0.00023420564352232565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document -0.00023367463699173143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document -0.00023344969163567033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document -0.00023372196941547188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document -0.00023399207645297834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document -0.00023357915605505856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document -0.00023337585642190864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document -0.00023385005470157914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document -0.00023301533534493465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document -0.00023377864302541782 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document -0.00023323745848621437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document -0.0002330594611151835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document -0.0002334149675026783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document -0.00023198945902291534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document -0.00023023784834634142 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document -0.00022985623060187217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document -0.0002292605284569516 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document -0.00022926593333048894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document -0.00022922766406807777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document -0.00022898153911167426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document -0.0002292473111593315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document -0.000228804579400424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document -0.00022865485613513526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document -0.00022937426835887895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document -0.00022917388311587372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document -0.0002291660582019043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document -0.00022907895248360543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document -0.0002294617879920205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document -0.0002290452150516566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document -0.00022943405619715553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document -0.0002296271421006204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document -0.00022854791372910372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document -0.00022923123467686557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document -0.00022852404355738494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document -0.00022847798660086642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document -0.0002289604586810316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document -0.00022835479834950643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document -0.0002289149402884243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document -0.00022806655474763446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document -0.00022826296420992974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document -0.00022906829636213627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document -0.0002287628414466998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document -0.0002282673911253445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document -0.00022869309841939134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document -0.0002281540116815451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document -0.0002259755756162738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document -0.00022562331285233504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document -0.0002259061146106053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document -0.00022567670836663787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document -0.00022573165387587061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document -0.00022508514961670572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document -0.00022564642513773356 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document -0.00022563088621998788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document -0.0002250438755373707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document -0.00022524465346241134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document -0.00022531737657666812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document -0.00022444687519363458 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document -0.00022460397498596298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document -0.00022454218976501763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document -0.00022447528843671366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document -0.00022501666332178926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document -0.00022453752304377972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document -0.00022484451871163002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document -0.00022465678847154914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document -0.00022453180917044732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document -0.0002247278486823009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document -0.00022465794828242097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document -0.00022431000701925386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document -0.00022476020248460963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document -0.00022467531771795015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document -0.0002236391309945234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document -0.00022458764920536007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document -0.00022430877426744415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document -0.0002247047786127192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document -0.0002245298090400035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document -0.0002245648831396188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document -0.00022292894729820784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document -0.00022236668082957533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document -0.0002217622659895442 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document -0.00022252452726732609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document -0.00022135333211363678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document -0.0002214571757787971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document -0.0002217188139237798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document -0.00022144214894640303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document -0.00022100172806631854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document -0.00022156392409199052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document -0.00022134830143710272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document -0.00022158598922529453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document -0.00022142932483041377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document -0.00022120980907786554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document -0.00022117917738112441 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document -0.00022077089397851235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document -0.00022093265074996711 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document -0.00022091299741377004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document -0.0002205849150703338 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document -0.0002210648204787979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document -0.0002214235747364102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document -0.00022083907302221787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document -0.0002206334237915964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document -0.00022065193929912214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document -0.00022079775597767288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document -0.00022091492909963518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document -0.00022095009987097293 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document -0.0002208150577180165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document -0.00022085759102772088 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document -0.00022073789170129016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document -0.00022049322781182384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document -0.00022083270617761285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document -0.00021982452827473632 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document -0.00021899870446514259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document -0.00021890358773356361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document -0.00021875556609042841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document -0.00021861195987201226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document -0.00021856782186167455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document -0.00021912837771543515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document -0.00021900213768517756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document -0.00021871675851390374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document -0.0002180537056545586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document -0.0002188196714327129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document -0.00021851362624523464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document -0.0002183236795498736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document -7.291153618675672e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0002406272620255565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 diff --git a/ALCF/data-lists/polaris/cc.txt b/ALCF/data-lists/polaris/cc.txt index edf6aab8c1c..75485866e61 100644 --- a/ALCF/data-lists/polaris/cc.txt +++ b/ALCF/data-lists/polaris/cc.txt @@ -1,1108 +1,1108 @@ -0.0003742481815405742 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document -0.00038204855962733055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document -0.00038821818392663593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document -0.00038723332988783727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document -0.00038916141142149904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document -0.00038049542523949033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document -0.0003854755539534284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document -0.00024202756466512517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document -0.0003915405155008087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document -0.0003927382151931033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document -0.0003839151202260479 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document -0.00040006817468967907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document -0.00040318965964443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document -0.0003831013019452741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document -0.00039166638383204036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document -0.00039962784023961004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document -0.00039536707853602614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document -0.0004204304698247758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document -0.00041538899178693555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document -0.00039186953333675306 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document -0.00038945837196504305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document -0.0003919951238929062 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document -0.00044377065718528966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document -0.0004407759068603017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document -0.0002487811895843715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document -0.00039349432045556636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document -0.00041223198559462343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document -0.0004036573014830213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document -0.0003825982215521807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document -0.00040386867133151386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document -0.00024460575279105167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document -0.000269029789531335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document -0.0003573757493252864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document -0.0004600876681392076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document -0.0002605354166397086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document -0.0003882502452157999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document -0.0002466747612126512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document -0.0004024726105072402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document -0.00040820631128483644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document -0.0002691094350403538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document -0.00026916830387277267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document -0.0004204663297880574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document -0.00042379698687085554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document -0.0004502169227311871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document -0.0002661708937015295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document -0.00031239486948031334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document -0.0003109054589936201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document -0.00045873053079760646 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document -0.00022904931423244635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document -0.0003813462028433663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document -0.00039188129256500874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document -0.00045124222276983765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document -0.00048138658436853695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document -0.0003944178776279866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document -0.00039941569676754006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document -0.00037952761190240494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document -0.0003944870860881476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document -0.0003891842411856621 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document -0.000387688981934861 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document -0.00039197953876258005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document -0.00039007915280311206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document -0.0003995520363699188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document -0.00039230985654592406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document -0.0003929472067173851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document -0.0003924096172671473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document -0.0003881636143629905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document -0.000389790617937084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document -0.00037351762309221023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document -0.0003630196170929407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document -0.00033532465765142113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document -0.0003076088685761823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document -0.00039463850897720803 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document -0.0002843816115231449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document -0.0002909175709416474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document -0.00028867170997202486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document -0.0002838644617723659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document -0.00029027869525543416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document -0.0002821339567560056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document -0.0002922988877045601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document -0.0002866955958315786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document -0.0002865271754558126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document -0.0002861247475618473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document -0.0002826681072408606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document -0.0002849746458282827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document -0.0002816966633435316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document -0.00026255342235948463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document -0.0002552895098829678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document -0.00025990194083107813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document -0.0002524062657685835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document -0.0002538577379748611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document -0.0002561415177406761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document -0.00026206253059694905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document -0.00026168095406910565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document -0.0002601305742008613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document -0.00025200823006814814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document -0.0003229951981263502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document -0.00037289448266476045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document -0.0003807825862179898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document -0.0003616333738191483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document -0.0003665117918907636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document -0.0003684186453633228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document -0.0003589330610806066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document -0.00036383861418030395 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document -0.000359841363355303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document -0.00036431044063050464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document -0.0003668574090358279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document -0.000362768263620199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document -0.0003501888032771077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document -0.000352401968221528 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document -0.0003541019701869794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document -0.0003628121865546891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document -0.0003752582953758773 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document -0.00037902046230424966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document -0.0003777927146925147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document -0.0003760676130509053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document -0.00034046049078755405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document -0.0003338847563259091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document -0.00033294499102761794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document -0.0004912026198265864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document -0.00032064363474664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document -0.00032154190389541214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document -0.00032309660151746207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document -0.00031181143365304544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document -0.00031046092294569104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document -0.00031150165249068046 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document -0.0003041314265988224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document -0.0003024834909739394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document -0.0003019936835833604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document -0.000292329665283177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document -0.0002867061143144972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document -0.00028443615610701707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document -0.00028462291013755945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document -0.0002793538601205013 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document -0.00027306573977044246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document -0.00027097155673336525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document -0.0002752934202112985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document -0.00043042012694697647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document -0.00047495648822986177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document -0.00047755032493473855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document -0.0004706974343933747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document -0.00046682163297771817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document -0.0004616765425874178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document -0.00030644496751628097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document -0.0002909492555358308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document -0.00027272036068261724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document -0.0004101070217315588 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document -0.0003728914338834357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document -0.00036546911442305647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document -0.0003669945482407483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document -0.0003715902407424017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document -0.00035837486406683366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document -0.0003573318538685469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document -0.0003553784893071916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document -0.0004920659809912352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document -0.0004533619411303183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document -0.00045067066057818706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document -0.00044396985139270645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document -0.00043198288204468477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document -0.00043005174223738454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document -0.00041847118430776784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document -0.00042952036375796664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document -0.00043420594647324267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document -0.0003461123241053012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document -0.0003408581597849182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document -0.00033172705422182547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document -0.0003392566490686136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document -0.00033578341518385483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document -0.0003439196710518844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document -0.00034559163447085543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document -0.00033762478642902825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document -0.00033215210055107224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document -0.00033423579608014966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document -0.0004963355016025102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document -0.0004996862761456923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document -0.0005000551829325451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document -0.0005004212610098755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document -0.00027768695585500585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document -0.00028395983854338433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document -0.00027835826303062254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document -0.0002740073176010804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document -0.0002791830529274016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document -0.0002796863816194411 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document -0.00026697453022672804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document -0.0002594197440280141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document -0.0003779565697649222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document -0.00041835823476586606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document -0.00043788493575265915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document -0.0002731731970096006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document -0.000276305847423402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document -0.0002704955773958623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document -0.0002629635944827518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document -0.000260070956974436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document -0.00025661553791456334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document -0.00025794727207576157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document -0.00025295733980001527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document -0.0003788106407021029 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document -0.0004882344027669431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document -0.0003275324309642705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document -0.0004803401856640094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document -0.00046720138323433943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document -0.00043527810307095335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document -0.00043905395741627827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document -0.00048774175867331425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document -0.00048380704121346737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document -0.0004779011848346118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document -0.00046255587581908036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document -0.00045127922880511576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document -0.0004503891485256095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document -0.0004450142332303422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document -0.00044630282482516654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document -0.00044325014465743616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document -0.0004263874842796447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document -0.0004217530913646938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document -0.000415120314341852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document -0.00040987168279144537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document -0.00033468337266607834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document -0.0003353094464683005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document -0.0004833936821707294 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document -0.00047194878988920935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document -0.0004648324126996427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document -0.0004562345003964941 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document -0.0004933203505465098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document -0.0003530166075325466 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document -0.00035368548192804685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document -0.0004872620828289663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document -0.00048293889392426456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document -0.00047936768462267655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document -0.00047821013991587545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document -0.0004660610308564753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document -0.000394683430103437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document -0.00039165053441571324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document -0.0003906936040164381 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document -0.00038074803919159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document -0.0003686529291578143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document -0.00035832920428870976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document -0.00035929024535947033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document -0.0003538226556050544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document -0.0003584167868708799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document -0.0003480507542594234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document -0.0003413709023543034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document -0.00034001304759361455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document -0.00033430532902756514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document -0.00046519252660631277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document -0.0002938876402514769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document -0.00028676090994509047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document -0.00027296150117506716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document -0.00026513502621960483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document -0.0002680081327926125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document -0.00025831225828720344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document -0.00026647037295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document -0.0002525733734572654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document -0.00025831708887575375 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document -0.00042487627444443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document -0.0004951213245023891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document -0.0004804051413177752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document -0.0004662397611340532 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document -0.0004550138655253933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document -0.00044494909122746795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document -0.0002899112253051385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document -0.0004372879736279761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document -0.0004529568099252922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document -0.00045127826158829573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document -0.0004436558176737439 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document -0.0004419233237678378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document -0.000434589215880319 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document -0.00029153613207706566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document -0.0004312458058738854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document -0.00028741854968757313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document -0.00046853200754421234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document -0.0004949145252030074 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document -0.00044459683920483167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document -0.0003836095306696336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document -0.0003789760237872398 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document -0.0003749227438304427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document -0.0003628558277173369 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document -0.00039468301394041474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document -0.00038874701821614864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document -0.0004158492456077867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document -0.00042360504554060077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document -0.00040386729844317623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document -0.00027595096702902474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document -0.00043638766787829135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document -0.0002218691596850179 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document -0.0004437566108089954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document -0.0003889996411609667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document -0.00043454421906537704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document -0.0004522564392830988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document -0.00041517835659357416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document -0.0002614360863446896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document -0.00037543522111463596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document -0.0004386190133514781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document -0.00046358333286115075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document -0.00043186261317942404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document -0.0002377581602097957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document -0.00025973334085074254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document -0.00040139099332000796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document -0.00043674860686687174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document -0.00040853289309329373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document -0.000242910191729688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document -0.0004431071731750582 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document -0.0004388092670482523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document -0.000381418866255965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document -0.0004100117296419717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document -0.00042469230366022745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document -0.00041744151905374254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document -0.00022835699906752945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document -0.0004380161085387397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document -0.00044803212381807456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document -0.00040554932796137236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document -0.0004234508646347761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document -0.00043341209652360653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document -0.00023966604734537185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document -0.000259165907316014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document -0.0004270653021833602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document -0.0004341547032162028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document -0.0004111478117275994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document -0.0004299383567984396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document -0.0004241899124590779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document -0.0004502719349364145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document -0.00038994621469645615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document -0.0003859912398894952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document -0.0004247535950310557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document -0.000386982084327716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document -0.0004196451040053251 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document -0.0004096278509782259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document -0.0004373334932695721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document -0.0004180889975240641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document -0.00042079636929672745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document -0.00038063574611812913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document -0.0003817505891515542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document -0.0004420096268860222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document -0.00039182670726410623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document -0.0003635667850372299 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document -0.00041564996472055667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document -0.000400529358757286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document -0.0003939113874958451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document -0.00039066622068940996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document -0.0004290098538807143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document -0.0004240739958197099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document -0.00040775392659215333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document -0.0004091634200396925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document -0.00042299190476617914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document -0.0003701492680344151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document -0.0003807353844384635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document -0.00038813507771983156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document -0.00040072346558408346 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document -0.0003603595180423597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document -0.00038799421353112465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document -0.00037575235582264926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document -0.0004239190342959713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document -0.0004606044799136546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document -0.00045107950652529253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document -0.0004391947201871058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document -0.0004457516661123035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document -0.0004301297170991686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document -0.00044661704164586694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document -0.0004438849846114837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document -0.0004444205734316823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document -0.0004190924165303394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document -0.00043942581131677875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document -0.00021568459798090663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document -0.0003814929225407199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document -0.0003217453179359235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document -0.00031719591470267974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document -0.00032434115726922137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document -0.0004079911120371051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document -0.000329492766381148 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document -0.0003845916162001633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document -0.0003835208964390098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document -0.00037847334157173194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document -0.00038296039903791865 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document -0.00037896336828472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document -0.00037620974396391355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document -0.00037420590727111843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document -0.000340490625886403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document -0.0003078314411035827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document -0.00034153990750656097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document -0.0003308858103982067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document -0.0003452640607156025 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document -0.00033095276418403455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document -0.0003116308995860414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document -0.00032446713226408477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document -0.0003015816821912984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document -0.00031612418775706894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document -0.0003278516344971041 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document -0.00033079446736097217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document -0.00032278977146550837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document -0.00032065272988207914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document -0.0003936696452406576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document -0.0003450109536627789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document -0.0003339787189919641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document -0.0003284303856176974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document -0.00033652677276843477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document -0.0003257822443845694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document -0.0003293985569149334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document -0.0003310360260148262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document -0.0003233770986418526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document -0.0003172280092149422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document -0.0003160674744292835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document -0.00030931090289598506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document -0.0003093173886443107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document -0.00033167847081104083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document -0.00031131501311729723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document -0.00031046608876279845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document -0.00030569235942207244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document -0.00030777943671285197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document -0.00029303314290956683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document -0.0003045824546400205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document -0.00030360880677729793 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document -0.00031646239964835433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document -0.0003129122300603785 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document -0.00031060464956661433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document -0.000311819032500067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document -0.0002977872483902282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document -0.0003009448600922438 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document -0.00028610292098537774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document -0.0002988326876216654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document -0.00028550828372819075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document -0.0002830381750875739 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document -0.0002848495855927156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document -0.0002856443760308144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document -0.00027442895344188584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document -0.0002681160554049462 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document -0.0003421482544126989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document -0.0004005872948449718 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document -0.0003930123959320308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document -0.0003867271832275778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document -0.000380805140455254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document -0.0003814769861947819 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document -0.00038025170883282324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document -0.0003738026647867475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document -0.00018960856915036276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document -0.0003697177501953134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document -0.00036674194328136693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document -0.00036447406838697555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document -0.00036686410861101255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document -0.00035915267825103423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document -0.0003624758404026675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document -0.0002822812140180794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document -0.00030620512946920813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document -0.000294249776520589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document -0.00030238536967523434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document -0.00029509593361580754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document -0.0002906912701830899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document -0.0002921944165474959 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document -0.00028358919691127954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document -0.0002813182772323272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document -0.00027442640800299205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document -0.0002747820342933984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document -0.0002747584403979717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document -0.00027499129634862444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document -0.0002712050404257197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document -0.0002616256943143254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document -0.00026769938929002815 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document -0.00038396081322727017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document -0.0003863140490027991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document -0.00037702277513203237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document -0.0003633274156107032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document -0.0003587473889240435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document -0.0003507672084278415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document -0.00033776425499780385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document -0.0003377914127574796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document -0.00032948015659161326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document -0.00033245638541392985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document -0.00031080707640648695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document -0.0002976903331149755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document -0.0002965121463725523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document -0.0002933849695266647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document -0.0002837035078508233 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document -0.00028684569079589323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document -0.0003145192320802359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document -0.0003566937253273515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document -0.0003470199109592918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document -0.0003060245312041868 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document -0.0002650817213818789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document -0.0002643604938780134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document -0.000299350876031416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document -0.0003178540797697938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document -0.000271850367887767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document -0.00031349896596549 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document -0.00031749734412765755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document -0.0003791137842391209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document -0.0003742334169957992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document -0.0003705639757351107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document -0.0003126986769797042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document -0.00031038132814561196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document -0.00036464437173804883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document -0.0003569480488951322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document -0.0003541239221619106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document -0.00035315297411308053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document -0.0003572451925404141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document -0.0003514986129411253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document -0.0003521798298425866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document -0.00034553677439244716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document -0.000349004719809412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document -0.0003468247484872769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document -0.0003465822608356558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document -0.00035410983132162007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document -0.0003487908354969444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document -0.0003479024763238147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document -0.000341412530646823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document -0.00034451316273667034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document -0.0002618849993484869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document -0.00026788679978901144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document -0.00027450670773227214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document -0.0002661273129899329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document -0.00026836569676402957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document -0.00026155876975483236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document -0.0002609276830117151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document -0.0002644161630512771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document -0.00036789208972872557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document -0.00037829849439990513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document -0.0003788894943523098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document -0.0003617207777959397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document -0.0002541334487248998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document -0.0002707945538071073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document -0.00027046282716455214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document -0.0002652443167243215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document -0.0002685859923850986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document -0.00025734961751176414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document -0.000259041720872915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document -0.00025340107274823446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document -0.00025757135121837893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document -0.00025617700500574084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document -0.0002566931670562857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document -0.0002543871190716101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document -0.00024997565589481713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document -0.0002954079779456287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document -0.00034890741135252835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document -0.0003473298137731525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document -0.0003296959618486435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document -0.0003304520061604598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document -0.00032377956175729824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document -0.00031700696295168713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document -0.0003060382346081943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document -0.0003012003005056863 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document -0.0002981074073993884 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document -0.0002922128825950705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document -0.000348901087722931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document -0.0003408286289467841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document -0.0003410649680770183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document -0.0003358524215576502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document -0.0003343661874989231 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document -0.00032810573699389156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document -0.00032261449539097497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document -0.0003162694866049203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document -0.0003158381156468853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document -0.000317376061083603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document -0.0003125788639953052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document -0.0003010105041885602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document -0.0003065865059090678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document -0.0003084275726508053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document -0.00030966560718296085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document -0.0002957728057853081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document -0.00029904164542325336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document -0.0002955358888729187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document -0.00028692976446931544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document -0.0002923476214935797 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document -0.0002893691697212419 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document -0.0002855895211981585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document -0.00027968347097626246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document -0.0002810783462604979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document -0.00027794080455729715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document -0.00034784376461416953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document -0.0003488347959010943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document -0.00034790583710250724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document -0.000345913166618151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document -0.00033801936268066675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document -0.0003290591130212315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document -0.00034051399521366823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document -0.00032470943131841784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document -0.00031679540050914276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document -0.00031814596342422325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document -0.0003156466289485036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document -0.00029985010879003633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document -0.0002905176377776361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document -0.0004206836775460856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document -0.00020660449162246918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document -0.0003461727254468087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document -0.00020592870907067763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document -0.00034173505299233005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document -0.0004052437256652738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document -0.0004080650901351697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document -0.00039778184149144276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document -0.00039046311464950275 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document -0.00039043444911071384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document -0.000388575704932843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document -0.00019737533145666597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document -0.00037610755595812403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document -0.00037315400127598317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document -0.00037415028580922163 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document -0.00036694041707212337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document -0.00018947219857306515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document -0.00037046050826533545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document -0.0003587440768559087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document -0.00034623936498708903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document -0.0003502289592617922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document -0.00034692398063649823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document -0.000339340809421849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document -0.0003360510394816983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document -0.0003354673850814145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document -0.00032937682875877047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document -0.00032844505049317715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document -0.00028287199339908627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document -0.0002795217197003578 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document -0.00028048955601883463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document -0.0002769326396439027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document -0.0002727090021299243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document -0.0002726577841024554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document -0.00026663619593455374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document -0.00026068042672138127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document -0.0002637704114326801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document -0.0002593043567100412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document -0.0002599897110113453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document -0.0002435078682758859 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document -0.0002450530071379054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document -0.00024233331983743606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document -0.0002934750947999535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document -0.00033241226364044474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document -0.00032938406090272075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document -0.00032778705403953246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document -0.00032184551480398754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document -0.00031874002264945737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document -0.0003165319685666433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document -0.00031307071173376295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document -0.00031119524184911957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document -0.0003102253344576429 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document -0.0003088976240383192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document -0.0002951410823077708 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document -0.00029772657676757413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document -0.0003056048989909935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document -0.00031991305381648026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document -0.00030890256978362426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document -0.0003109382904091933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document -0.00031035798529690644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document -0.00030741666395911753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document -0.0002989918594861846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document -0.00029569635443989434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document -0.0002973992445667285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document -0.000293397351001072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document -0.00028737817438047954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document -0.00028252738144009747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document -0.0002805511898623541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document -0.0003718020784620472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document -0.0003499713845765235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document -0.00034283547445326676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document -0.00031464759888838765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document -0.00033188946446414833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document -0.000326084432195463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document -0.0003764568303917893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document -0.0003604955598858414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document -0.0003655654554133222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document -0.00035762304033750504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document -0.00038478883950347103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document -0.00027735714341247454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document -0.00028139534607773563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document -0.00019777292251713763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document -0.000285571704874486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document -0.00028543482146244363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document -0.00019434234484256758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document -0.00027854908176986763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document -0.0002847068039566143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document -0.00028672356943064853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document -0.00027782687605808177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document -0.0002843539634105203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document -0.0002894748379090401 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document -0.0002868852440186493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document -0.0002818504885373851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document -0.00028680112812941034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document -0.00019258978168723977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document -0.00028760637934715155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document -0.0002820439443912918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document -0.0002831001054410018 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document -0.00029001901552467397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document -0.00027779449377883156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document -0.00019949837437516796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document -0.0002907306472984446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document -0.00027814858381318327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document -0.00019472790889161432 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document -0.00020472626596924125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document -0.0002870045081974301 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document -0.00019812241927078482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document -0.0002817553333369554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document -0.00027829782796642117 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document -0.00028289431732284113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document -0.0002795526296717729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document -0.00027682829988044574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document -0.0002895432402719184 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document -0.0002823174903941811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document -0.00028170972351837796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document -0.00027807915877838826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document -0.00028588515681452956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document -0.00028112324090816726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document -0.00020636178289985485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document -0.00019447255290980535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document -0.0002850824220591452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document -0.00027856429520116784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document -0.0002820880676635633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document -0.00028943902215995714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document -0.0002676366291085329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document -0.00023806333809954687 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document -0.00024526460430233455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document -0.00023876876664622726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document -0.00023379770334179805 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document -0.00024175151269138382 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document -0.00023386583242595706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document -0.00023771797150160827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document -0.0002262748967483896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document -0.0002408148346432682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document -0.00023398651720444235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document -0.00022989433874474592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document -0.00023948500543957772 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document -0.0002331594076859196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document -0.00023375132439600242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document -0.00023923410909668642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document -0.00023952796315562954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document -0.0002327466076905069 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document -0.00023082758956797212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document -0.0002240509275524448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document -0.00022798879995765268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document -0.000221172516774386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document -0.00021767045123534623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document -0.00021982832794804484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document -0.00021971626543789102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document -0.00022566565206920132 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document -0.0002181984894194856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document -0.00021831417549554653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document -0.00021601405421187145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document -0.00022275733725519607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document -0.00021847734911973986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document -0.0002243591012664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document -0.00021688758139483833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document -0.0002182953624789215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document -0.00020475155724026002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document -0.00021498078062960065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document -0.0002157914337233064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document -0.00021781838494967963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document -0.00021723242266814558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document -0.0002176782686553837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document -0.0003486179404943968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document -0.00034882846352857634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document -0.00031400868448352596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document -0.00030273484020011963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document -0.00029895889118145404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document -0.00029770764609621714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document -0.0002990181332116852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document -0.00029653733972285996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document -0.00029624649222942476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document -0.00029625609720203576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document -0.00029731928930852147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document -0.00029011721326148513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document -0.00028849788197494655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document -0.00021601278623858145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document -0.00021319599281739178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document -0.0002153325290600083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document -0.00018566946174516558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document -0.00020736824394291617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document -0.00020857419820128004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document -0.00020058526129536423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document -0.00020745812166665217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document -0.00020652171015271702 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document -0.00020643808911278608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document -0.00020040513914482103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document -0.00020598050188272898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document -0.0001969184139343296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document -0.0001972748812937012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document -0.0002038556751586195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document -0.00020245186011313464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document -0.00019950381422038783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document -0.00020837055459665258 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document -0.00020371856218246096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document -0.00019537612301625791 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document -0.00019914984508813857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document -0.0002053787713691309 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document -0.00019082100541008637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document -0.00020397153334531813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document -0.0002021462693077317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document -0.00019609357008124035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document -0.00019693256622486236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document -0.00020007239732428112 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document -0.00020467075741591954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document -0.00019584883400022932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document -0.00019135050391176972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document -0.0003362829834208298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document -0.00034013691154784095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document -0.00033215887031941976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document -0.00032681189065396707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document -0.0003149138485493094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document -0.00030179177307540077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document -0.0002923278437581119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document -0.00029470052278994486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document -0.0002994095093045731 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document -0.00029033525096085037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document -0.00029390798852496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document -0.0002916230924130842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document -0.00029419886374594913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document -0.0002865469756730764 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document -0.00021191292549942086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document -0.00021369664817409847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document -0.00021612485624266726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document -0.00022242192634588478 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document -0.00014605095659989698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document -0.00022070626106341693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document -0.0002174420774054071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document -0.00021325858963116995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document -0.0002124322999488052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document -0.0002081218896969054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document -0.0002108710211556957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document -0.00020686867095978426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document -0.00020895752681041895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document -0.00020741922266415738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document -0.0002069112657197308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document -0.00020644627473468118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document -0.00020332991338121604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document -0.0003560895677789848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document -0.00032915779111908214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document -0.00033810613317040864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document -0.00033729626594036923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document -0.00033550342864602944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document -0.00034173474024556906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document -0.000331505340748827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document -0.0003270050330117195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document -0.00032585275329172556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document -0.0003143383203190604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document -0.00031655199110388894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document -0.00030738872158476413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document -0.00030838388352699285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document -0.0003053596995351888 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document -0.00031836304739584593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document -0.000315315435873905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document -0.0003087116248965243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document -0.00030396790625537645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document -0.0003335812246032149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document -0.00034570956323095843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document -0.00034563035636675786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document -0.00033411265479076335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document -0.00034439191141692787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document -0.0003364483125496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document -0.0003299500453608033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document -0.00033163377700074837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document -0.00032638649660627673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document -0.00032616167939645234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document -0.0003205289298760723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document -0.00031939393740815355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document -0.00031593164066731296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document -0.00031928871111254405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document -0.00029670189073175004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document -0.00020517703846735904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document -0.00020128418186172073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document -0.00019662723895606717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document -0.0001981157042081407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document -0.00019703489037041608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document -0.00019079796331785068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document -0.0001909352306690079 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document -0.00018824662295261396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document -0.00019864275319325954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document -0.00018818516521649587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document -0.00018875694972812844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document -0.00018231621170645482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document -0.00018349407845798273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document -0.00018088971427746906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document -0.00018296284236327237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document -0.0001876011825819916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document -0.000329052068725176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document -0.00032223616273648536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document -0.00031272564089633955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document -0.00031621609908414494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document -0.0003117213560911235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document -0.00030218064069945934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document -0.00030658916600512085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document -0.0002915863534115821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document -0.0002940280138374372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document -0.00029067860468866085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document -0.00028529228063135635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document -0.00028336893301452256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document -0.0002794668089130099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document -0.00021681361378827842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document -0.0001484664674497246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document -0.00021950558378215133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document -0.00021806860758808645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document -0.00021819568718852282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document -0.00021626925931585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document -0.0001464536143077762 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document -0.00021432777088808917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document -0.000213473805865147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document -0.00021397067253964538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document -0.00020758957647437263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document -0.00020687124337683314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document -0.00020630057046511005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document -0.0002091166859352538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document -0.00020777355025615267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document -0.00020709287641496176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document -0.00020736464660577094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document -0.00020062246741862607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document -0.00020693207561942915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document -0.00021151004871893024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document -0.00019930249098689716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document -0.00021589710041231824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document -0.00021369204789905741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document -0.0002147099923936778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document -0.00021077531190389536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document -0.0002100509829113836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document -0.00021185362601571124 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document -0.00020722136637339565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document -0.00020300093701169531 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document -0.00019859737993313477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document -0.00019971314372100164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document -0.00019549908270269278 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document -0.00019649820843534028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document -0.00019619415513498067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document -0.00019493006120377898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document -0.00019499409035775506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document -0.00019252988593634277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document -0.00019440768268686405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document -0.00018747161324755577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document -0.0001879575932372779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document -0.00019040707058357506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document -0.0001871931095090703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document -0.00020112966223017096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document -0.00020516878165311017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document -0.00020664735191740533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document -0.00021041398572882962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document -0.00020397992929690396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document -0.0002039978580295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document -0.00020592785601142126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document -0.0001990755527445265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document -0.00019729564847798732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document -0.00019958182230527032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document -0.0001985037302636386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document -0.00020204130355115716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document -0.0002000296401958085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document -0.0001983064832295463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document -0.00019663108484195617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document -0.00019510678560556523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document -0.0001873284057063206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document -0.00019311553072495885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document -0.00034652137288816547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document -0.0002813690318850024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document -0.00027697649713138685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document -0.0002755419092534421 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document -0.0002681583054440219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document -0.00026945753192750824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document -0.00026169470768245737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document -0.00026437008960810825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document -0.0002637294838228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document -0.00026491867965088836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document -0.00025504483625138986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document -0.0002545040623796586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document -0.0002546682814073622 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document -0.00025545439487142615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document -0.0002626896557978271 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document -0.00025092040940402784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document -0.0002589154885863872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document -0.00024106160482721467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document -0.0002483289690087987 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document -0.0002388930282784437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document -0.00024006340759273874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document -0.00023765248178029045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document -0.00023061351965578936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document -0.00024954224883546477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document -0.00017861017233018525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document -0.00017810832743667658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document -0.00017599709170759497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document -0.00017462723516505223 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document -0.0002906316527068669 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document -0.00033762141066247166 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document -0.00017170670574152494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document -0.00017258674515137717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document -0.0002815386173173926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document -0.0002996845935618989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document -0.0002735268488987296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document -0.0002971738713071517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document -0.0002942690674002763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document -0.0003322222207729567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document -0.0003378721656198464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document -0.00018307262621851067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document -0.00033956081502775057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document -0.00031604820927876276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document -0.00028805657681088917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document -0.00026312293321215633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document -0.00034366936722921455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document -0.0002865256504406559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document -0.0003063615195861786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document -0.00028412791619666136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document -0.00028060835132727154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document -0.00032544974761560506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document -0.0002647177833217225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document -0.0003152621884896575 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document -0.0003054625140336913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document -0.00031183308312292263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document -0.00018175026696621178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document -0.00017699918328872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document -0.00018222339261441908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document -0.00018348005930964137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document -0.0001810735993810541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document -0.00030846441282038914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document -0.0002972326889310354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document -0.00017433421318235594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document -0.00032799458649525895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document -0.00032482130048512673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document -0.00031943465668672475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document -0.00029615593630484517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document -0.0002893126939511001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document -0.0002849288351723284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document -0.00028383906633569267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document -0.00028072526091262615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document -0.000284239564292377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document -0.0002778903109432523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document -0.0002771644389501471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document -0.0002733316182319337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document -0.00026362539185869363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document -0.0002636325383220217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document -0.00026740622442302886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document -0.0002646771971853427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document -0.0002628566720605389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document -0.0002644760695434766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document -0.0002623837702310999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document -0.00026088722976772894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document -0.0002567065374799158 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document -0.00018857382101207726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document -0.00019036580399817203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document -0.00018348828065261222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document -0.00018491851780345073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document -0.00018904887260080187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document -0.0001875609304251801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document -0.00018393034720015817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document -0.00018419795526114903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document -0.00018699955623404795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document -0.00018276256902965128 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document -0.00017698045695190812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document -0.00018104650132303642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document -0.00017758206731279688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document -0.00017131402995103497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document -0.000175944428350446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document -0.0003416745727147391 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document -0.0003163259373952889 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document -0.0002804489269172448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document -0.00028748272397403175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document -0.00027603318345630605 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document -0.000271638824679648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document -0.0002763761210210942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document -0.00026501984873172717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document -0.00026422486894694714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document -0.0002686339100849262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document -0.0002610837453940606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document -0.000260974343729353 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document -0.0002599403837029134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document -0.0002937273113238609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document -0.0003341790732600504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document -0.0002620661576600244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document -0.0003027929169239288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document -0.00031944039129326894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document -0.00019025676304139009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document -0.00018680910145009907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document -0.00034215840419416437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document -0.00018618120812119364 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document -0.00018605853095599425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document -0.00018120712626096538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document -0.00018315079292495327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document -0.00018362556449041974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document -0.0001780024456718171 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document -0.00033296526436178697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document -0.0001802398632282846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document -0.00017340263100798256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document -0.00017755840547238697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document -0.00018419413735260606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document -0.00017869518174591322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document -0.00017526271460129484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document -0.00017852168597981907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document -0.00017566536156787157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document -0.00017589867964432936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document -0.00017831487394075305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document -0.00017837310528935862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document -0.00018200908814216548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document -0.0001795136627511612 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document -0.0003414021775300033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document -0.00017177291787788502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document -0.0003441900648571877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document -0.0003394534597060673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document -0.0003236887233114832 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document -0.0001639544129688747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document -0.00019137443753211255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document -0.00018575146284680153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document -0.00019184792863440243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document -0.00018966043065679055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document -0.00017968851317035848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document -0.00018479881897661546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document -0.0001813642692683015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document -0.0001686449798983066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document -0.00018516104592230446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document -0.00031283726601066385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document -0.0003248607542883853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document -0.00031583241601202365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document -0.00031238270857730376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document -0.000307150592403979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document -0.00029443829986847044 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document -0.0002942723732234677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document -0.00023514930666443422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document -0.0020776328951453444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document -0.0021768234410538883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document -0.002106973549276289 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document -0.002110915756171751 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document -0.0017032382109816464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document -0.0019047944877712286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document -0.0019402711744016077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document -0.0006264790011223686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document -0.0017885401938106643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003742481815405742 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc diff --git a/ALCF/data-lists/polaris/dolma.txt b/ALCF/data-lists/polaris/dolma.txt new file mode 100644 index 00000000000..f2f98ab12c6 --- /dev/null +++ b/ALCF/data-lists/polaris/dolma.txt @@ -0,0 +1,2419 @@ +0.0018520780893211373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train +0.0002583902668716813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv +0.0031025147279277244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0002406272620255565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 +0.0003742481815405742 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc +0.0003547982093445404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon +6.322825248625475e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika +0.001451215788905126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train +0.0012499632072059553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o +0.0005759963691850877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit +0.0009994361338078242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange +0.004474659408857016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder +0.00032927705604725614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu +0.003548077173506675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki diff --git a/ALCF/data-lists/polaris/falcon.txt b/ALCF/data-lists/polaris/falcon.txt index 68aeb2f27bb..914d4803a4b 100644 --- a/ALCF/data-lists/polaris/falcon.txt +++ b/ALCF/data-lists/polaris/falcon.txt @@ -1,501 +1,501 @@ -0.0003547982093445404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document -0.00035934014428504944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document -0.00035707704501371544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document -0.00035287930712815354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document -0.00035977166728996823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document -0.0003581675664109838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document -0.0003548617059697185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document -0.0003639582000286208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document -0.00035375839698688127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document -0.0003743722020080678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document -0.0003530399715341242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document -0.00035511875882752406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document -0.0003618733574783154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document -0.00035185243285420104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document -0.0003541503739732106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document -0.0003631679485751914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document -0.00035748045578182274 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document -0.0003606490690555877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document -0.0003626383296610091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document -0.00035442644361264756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document -0.00035978370170539796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document -0.0003585562375341541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document -0.0003601958372888019 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document -0.000350277765402227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document -0.0003616521184211704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document -0.0003620625543608188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document -0.0003560781983850704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document -0.0003553209610592676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document -0.00035905348643915075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document -0.00034744258805696526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document -0.00035462784035661496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document -0.00034768186175100895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document -0.0003568534635532736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document -0.00035586511544371234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document -0.0003524567827568137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document -0.0003512453770426313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document -0.0003591792726468799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document -0.0003514024529343127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document -0.0003584880112586934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document -0.00035133552916418045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document -0.0003600811981350215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document -0.0003571663974228119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document -0.00035768103378874214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document -0.00035939205561113694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document -0.00035186773916029825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document -0.0003542829672490847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document -0.0003592783642898726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document -0.0003556367340099302 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document -0.00035391392271377027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document -0.00035486725707484836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document -0.00034866743396828035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document -0.0003517219808644735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document -0.00034874458549673823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document -0.000355773136961014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document -0.00035611750387841917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document -0.00035305602013916315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document -0.0003578207127071924 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document -0.00035514635841943707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document -0.00034816946212866206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document -0.0003512707269761496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document -0.0003483392117980654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document -0.0003572169607204321 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document -0.00035139153281660794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document -0.00035536422129036537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document -0.000352017164107143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document -0.000351889550179365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document -0.000358759689953589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document -0.0003569286079869268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document -0.0003657752958602099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document -0.00035396127934790697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document -0.0003618565071224743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document -0.00035146051531973204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document -0.00036107135765783567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document -0.00035019554279994576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document -0.00035567858879904983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document -0.0003504753174793183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document -0.00035931140831329194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document -0.0003502967866002823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document -0.0003532911801041972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document -0.0003583543013070199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document -0.0003566243489931224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document -0.0003468752314799221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document -0.0003597840618138091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document -0.00035128822484768084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document -0.00035889496943437507 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document -0.000352400524650424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document -0.0003518689536768735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document -0.00035866864741303467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document -0.0003454687659106334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document -0.00035348007259317576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document -0.0003539752270940644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document -0.00035146495994081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document -0.00035397212846310423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document -0.00035208246467162587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document -0.0003490843168676626 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document -0.00035299633658644394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document -0.00034868327466167065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document -0.00035941351365601583 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document -0.0003545343062735255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document -0.0003528956380445978 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document -0.0003553355770443352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document -0.0003644224004937743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document -0.00035234291036216907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document -0.0003596237469847771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document -0.0003531996065735989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document -0.0003547177054106099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document -0.0003575586499260483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document -0.00035262635135283667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document -0.0003624191962188944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document -0.0003488398052948616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document -0.0003598294093147917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document -0.00035583006534466323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document -0.00035403139653225103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document -0.00036134702642187156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document -0.0003573689927162834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document -0.0003577141131435527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document -0.00035208814419277406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document -0.00035996720683665625 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document -0.00035415304658912596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document -0.00036353353029443546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document -0.0003537326003150983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document -0.00036053976358299083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document -0.000352380489373494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document -0.00036154661616900994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document -0.00035959332325963614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document -0.0003597954667189692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document -0.0003563108270597542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document -0.0003582891940460143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document -0.0003497728210484297 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document -0.0003549834902179354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document -0.0003529828233484542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document -0.00034627483903285777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document -0.00035569006572589215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document -0.00035449377946910314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document -0.00035802844396194623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document -0.0003617277809353208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document -0.00035034118898654814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document -0.000351091193908611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document -0.0003527914342210668 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document -0.00035028288369781376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document -0.00035775745592780506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document -0.0003449630690661468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document -0.0003583490698830361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document -0.0003476995746684122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document -0.0003535632505019212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document -0.00035640180641147417 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document -0.000361731045691765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document -0.0003534082129597368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document -0.0003550344149828664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document -0.00035363002411364057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document -0.0003537265579677396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document -0.00034950531383577937 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document -0.00035008511827347514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document -0.00035594533400871325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document -0.00035266312861335946 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document -0.00035280268794863923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document -0.0003565470391528536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document -0.0003588492322689137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document -0.00035469909697832775 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document -0.00034712082813410526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document -0.000348701157101807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document -0.0003500192014479944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document -0.00035120560544669755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document -0.00035403656850437445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document -0.00035852376560749366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document -0.0003534754068111774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document -0.00035591740046720765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document -0.000348522354782563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document -0.0003533533959664415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document -0.00035631425964030697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document -0.0003485886551574741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document -0.00035917652631065777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document -0.0003482975272111288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document -0.00035580661277480167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document -0.0003492290722955348 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document -0.00034989284450240613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document -0.0003545677216162781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document -0.00034622286859463484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document -0.00036070626989861965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document -0.00035518365036320786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document -0.00035272907057848406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document -0.0003547343638218734 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document -0.0003496450144966242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document -0.0003537407829294287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document -0.0003489722653985685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document -0.00035057186899911295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document -0.0003507566548933051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document -0.00035630360179023747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document -0.00035631362503416367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document -0.0003490204248026821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document -0.00035761724058371226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document -0.00035037664777467137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document -0.000353402110481068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document -0.00034524163568371745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document -0.00035528523728570974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document -0.00034784916132431703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document -0.00034928476408048925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document -0.00034989205973784984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document -0.00034201664404094254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document -0.0003529676016338611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document -0.00034643433682346637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document -0.0003511666373001904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document -0.00034828669066575333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document -0.0003494625207264413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document -0.0003458957535879216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document -0.0003543020478990003 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document -0.00034754384069014956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document -0.0003598856392240133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document -0.0003503335458553846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document -0.00035919595619778716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document -0.00035767737970754404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document -0.00035197152783998165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document -0.0003549609834422404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document -0.0003568184100569753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document -0.0003512652818651935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document -0.00035912648958665754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document -0.00034764526964056546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document -0.000352439784960359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document -0.00035295886560764226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document -0.0003518132693658672 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document -0.00035589987915465713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document -0.00034923863317385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document -0.0003457987267929692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document -0.0003560928663480501 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document -0.0003529603811204932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document -0.0003524438555443043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document -0.0003438847030263783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document -0.00035981978898461613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document -0.0003446342778566972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document -0.00035529584995236537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document -0.00034855740895831116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document -0.00034932634912802544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document -0.00035805518303064666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document -0.0003497941877073061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document -0.00035774398685405447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document -0.0003560421780316607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document -0.0003508844468369392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document -0.00035731928892270107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document -0.0003557884626314314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document -0.00034992996760289355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document -0.000360752554360921 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document -0.0003452321668708545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document -0.0003591745226131023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document -0.00035256981433229084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document -0.00035378123159712034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document -0.000350464354895999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document -0.00035074625557389677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document -0.00035025894701994667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document -0.00035437902514857614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document -0.0003514684519732232 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document -0.00035449717909633905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document -0.0003436816402714221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document -0.00035139158071782116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document -0.0003509424079843335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document -0.000343894618577506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document -0.0003500789770661659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document -0.0003407788080680086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document -0.0003581908175239701 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document -0.0003465541618780918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document -0.00034600228792437736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document -0.00034416738982773204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document -0.0003519900340150641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document -0.000343369616864659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document -0.0003544993883274688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document -0.0003504441365073392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document -0.00034859160702727056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document -0.00035355909532647185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document -0.0003471900922691849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document -0.0003563015508709187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document -0.0003487888744148821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document -0.00034711767548688336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document -0.0003530734609369085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document -0.00035123969242560935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document -0.0003517127620891489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document -0.00035232835416868673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document -0.0003524437481912308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document -0.0003525996167005602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document -0.00035064770545242043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document -0.00035311558274981226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document -0.00034952204800569914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document -0.0003541471367344846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document -0.00035418812454561825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document -0.0003528951372900714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document -0.0003542338042975688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document -0.00034937738939942796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document -0.0003522182190878447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document -0.0003501406466507449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document -0.00034973079877492633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document -0.0003485274567713538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document -0.00034999308679368985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document -0.0003570051724707296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document -0.00034567230462019706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document -0.00035529000940160696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document -0.00034956512308671755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document -0.0003496962834028953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document -0.0003468745282493457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document -0.0003502717155809202 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document -0.0003556240880896514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document -0.0003515109488424343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document -0.0003563156688192592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document -0.00035040277363989817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document -0.0003481408593290717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document -0.0003624575124332874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document -0.0003522684124250313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document -0.00035286996027653544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document -0.00034967623997256725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document -0.00035182649587602765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document -0.0003524892557026489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document -0.0003507642477451811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document -0.00036190408389835666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document -0.00035102739424880766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document -0.00035239718753257265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document -0.00035298076121821316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document -0.0003478704389752654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document -0.0003503109191567942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document -0.00035143250975654426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document -0.0003480663923069012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document -0.00035691540219998623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document -0.000348815437166351 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document -0.00035202073257766225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document -0.0003491569096274706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document -0.00035277390475511834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document -0.0003524972090026609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document -0.0003504854249750236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document -0.00034740238025423914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document -0.00034968015462277606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document -0.0003493798632762674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document -0.0003488202537862122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document -0.0003525461864643725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document -0.00034903815232825664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document -0.00035536982539258216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document -0.00034858083265155483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document -0.0003505014973608067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document -0.00035327984042622104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document -0.0003503286677453136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document -0.00035835274842442816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document -0.00034970302660275595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document -0.000357929573140149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document -0.0003517238649788585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document -0.00036097027318848475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document -0.0003502734074110026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document -0.00035801510806036273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document -0.0003568006373479869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document -0.00036128108717454636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document -0.0003563436883111686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document -0.00035559725321852463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document -0.00035089656006854944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document -0.000359453964362057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document -0.00035629498059104033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document -0.0003622207707090437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document -0.0003540946784512821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document -0.0003594750565232011 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document -0.0003566007415086991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document -0.0003562142599126134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document -0.0003569948186744601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document -0.00035166554847920186 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document -0.00035047994419295137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document -0.0003561578193739437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document -0.00035470866838811544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document -0.00034216920464876335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document -0.0003550021513075795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document -0.0003488045105938729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document -0.0003513340720840151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document -0.0003448558566387584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document -0.0003460966026953241 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document -0.0003488157616036459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document -0.0003446120387842362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document -0.000351528602987427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document -0.00035661118227454713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document -0.0003551342699877457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document -0.0003478953397924445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document -0.00034625782458988215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document -0.0003527515447405871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document -0.00034823744889805696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document -0.00034823314560254406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document -0.00035162668292961944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document -0.0003477307716074623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document -0.0003446457989477787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document -0.00034782916273767795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document -0.0003517249130302248 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document -0.0003449873430908556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document -0.00034841291749669877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document -0.0003466028498941749 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document -0.0003486436831199424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document -0.0003478279234211838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document -0.0003495903653274374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document -0.00034896893881218957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document -0.000348941645312426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document -0.0003474221308416894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document -0.0003462621543839385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document -0.0003669373860863891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document -0.00034691156268163006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document -0.0003527774103765281 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document -0.00034684565672734663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document -0.0003454250599604457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document -0.0003541536557159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document -0.000345735737037366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document -0.0003524669816385214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document -0.0003441817133096468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document -0.0003519093265859089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document -0.00035080085480352095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document -0.00035285227929327434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document -0.00034354836346901676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document -0.00034789770937373467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document -0.000343665920520102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document -0.0003490884931060568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document -0.00034380029463398654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document -0.00034874768005099945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document -0.0003457058510967673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document -0.00034644265227023904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document -0.00035008339858594957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document -0.0003462377193296194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document -0.0003620491787114201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document -0.000348717011044469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document -0.00034370072363913706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document -0.0003551981066775649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document -0.0003500119496799342 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document -0.0003485082952669081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document -0.0003508155580978919 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document -0.00035311375163251416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document -0.00034945972003423253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document -0.0003474220353789879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document -0.0003536443686585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document -0.0003560350489042953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document -0.0003493655927914396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document -0.0003528423977146383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document -0.00035255554724471217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document -0.0003479760010190111 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document -0.00035458598862501956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document -0.0003458990560538315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document -0.00035157946422379875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document -0.00034736860650169996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document -0.0003529152313394119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document -0.00034586294329524465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document -0.00035707214923794877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document -0.0003509580363496512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document -0.00035244176725524474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document -0.0003467539557999047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document -0.00034919687962275546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document -0.00035094031731719953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document -0.0003484309008351352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document -0.0003485409424916253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document -0.0003499590776117838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document -0.0003492842758957848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document -0.0003529712275178912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document -0.0003566141287087449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document -0.0003649496522047409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document -0.0003563218912208234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document -0.00035614782126966145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document -0.0003531944298453266 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document -0.0003535950949566616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document -0.0003544295554928795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document -0.0003519908503740376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document -0.00035752817626134463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document -0.0003515322689589972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document -0.0003486893890307115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document -0.0003446520464889867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document -0.0003509421562481707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document -0.00035335015702909084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document -0.0003490178167345008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document -0.0003520497821155174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document -0.0003549762618908944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document -0.00035072190850833103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document -0.0003542458638526423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document -0.000352419194572916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document -0.0003545102564672614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document -0.0003495437992331806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document -0.0003542843376993964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document -0.000352827529313958 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document -0.00035442506093223886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document -0.0003496970719044257 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document -0.0003553096424442362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document -0.00034986845565067564 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document -0.000352131055186658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document -0.0003527021708198983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document -0.00034905885414547214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document -0.0003583433842468394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document -0.00034409435202828383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document -0.00034846410520871483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document -0.0003554459991927314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document -0.00035310507471843076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document -0.000350028910786098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document -0.00035049727458009896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document -0.0003519047735925826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document -0.0003513027429919726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document -0.0003626947260354396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document -0.0003500087324849783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document -0.0003618315726725285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document -0.0003535385113938023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document -0.0003487064058517615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document -0.0003618709124780938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document -0.00035040070335625915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document -0.0003506279032267829 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document -0.0003498435310527524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document -0.0003554634749821431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document -0.00035091209738758963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document -0.00035034103678978573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document -0.00035398931854386146 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document -0.00035495529304989485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document -0.00036067883473356603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +0.0003547982093445404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon diff --git a/ALCF/data-lists/polaris/megawiki.txt b/ALCF/data-lists/polaris/megawiki.txt index 4c4f47df5ff..56ec7debc7e 100644 --- a/ALCF/data-lists/polaris/megawiki.txt +++ b/ALCF/data-lists/polaris/megawiki.txt @@ -1,262 +1,262 @@ -6.322825248625475e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document -2.4432314037946264e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document -5.6313888721313454e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document -2.4208171781595055e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document -2.325811856369237e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document -2.4010790356322705e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document -5.36773610843632e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document -1.360574433501002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document -1.3076540344853244e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document -1.3386534334886313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document -1.2498103719605153e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document -1.403763836949682e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document -1.3636756723495417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document -1.2242489446940814e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document -1.2398255818973339e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document -1.2972616994216281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document -1.3947809855914134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document -1.3144843787829514e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document -1.1693809976572487e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document -1.3677252682893802e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document -1.3940876719849597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document -1.4222245138730965e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document -1.3201677767919704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document -1.1421717796486169e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document -1.2890514724498703e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document -1.3649507648749037e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document -1.2400732563490717e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document -1.1557681453277616e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document -1.2294483595964517e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document -1.2137484472122283e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document -1.3299663426456e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document -1.2461984216479532e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document -1.4666434217609636e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document -1.1876997894686238e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document -1.2939155338964078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document -1.3859590039728515e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document -1.317917848615668e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document -1.1335281536110342e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document -1.2889923952861426e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document -1.3471671647053326e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document -1.2221720014475102e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document -1.2632647276287541e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document -1.28276219004076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document -1.36213704321643e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document -1.2414858625261553e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document -1.3173700421883744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document -1.295597796725686e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document -1.242783936442904e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document -1.2417374088427464e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document -1.2134479405400744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document -1.3090040663304255e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document -1.2713470581614905e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document -5.5750231378906594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document -5.777597358425469e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document -5.349786767471258e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document -5.675165050453583e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document -5.482611216158831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document -5.065421899890121e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document -5.384718357480146e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document -4.872037363236061e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document -4.532709250783155e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document -5.7257963030489613e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document -4.9014365579652036e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document -5.722863552770969e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document -6.149911636146833e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document -5.2178057608273506e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document -4.990228161160431e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document -5.866186875255134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document -5.004185734360719e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document -4.79401853705107e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document -5.435219965052376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document -5.035997225792266e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document -5.622401774211625e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document -5.028826157387559e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document -5.596379470128795e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document -6.027824493191489e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document -5.5358270009931474e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document -5.9839051807685496e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document -5.1221077499249595e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document -5.517228560620279e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document -5.1687858285052305e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document -5.684188244145645e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document -5.212693275535878e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document -4.8551007022784084e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document -5.4888506639203145e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document -5.345098688527242e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document -4.8506420625516594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document -5.132168603397676e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document -5.719476795114223e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document -5.7448621149792696e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document -4.9068410568059265e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document -5.382937299647678e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document -4.8288432136304634e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document -5.841703200305416e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document -5.1589611587885584e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document -6.031113829732574e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document -5.4558202844532094e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document -5.341852317196142e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document -5.1402942738369954e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document -5.735421384377395e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document -5.473629863586958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document -5.4708993245733936e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document -4.931161863634078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document -5.104173022127248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document -5.510157161510824e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document -5.652501401782597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document -5.7273656573031666e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document -5.638363224821738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document -5.6128115396668704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document -5.00304877998141e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document -5.596120554779096e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document -5.5280923889040006e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document -5.223477917938408e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document -5.29472809986569e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document -2.205682378243213e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document -1.4367563720603185e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document -3.5506193487931076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document -3.0442910855821778e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document -2.2540042508019627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document -2.6880163202623216e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document -2.534473148048727e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document -2.6560945431318916e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document -2.547470248967691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document -2.5248825388073738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document -2.5828729575000054e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document -2.4026583817957736e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document -2.3930425429834413e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document -2.5037365362599724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document -2.6696745470595603e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document -2.140323051341762e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document -2.617354786691592e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document -1.538359101762691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document -1.2871029252377856e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document -2.255195411289217e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document -2.4832313897952067e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document -9.303873918189968e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document -2.179532302620228e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document -1.9750517506901206e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document -2.7740420380648435e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document -2.7813714782319335e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document -4.1595357937609806e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document -2.741365122389175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document -2.117451071361901e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document -1.7132649760565998e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document -1.7492547092602047e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document -1.7499951097392276e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document -1.6632444789170958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document -1.6678802252361607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document -1.5519208704558896e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document -1.652420992967167e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document -1.6119931034508755e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document -1.6638882076736552e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document -1.7198076782652946e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document -1.572927860565175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document -1.5194822618169918e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document -1.6677776832669846e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document -1.595612492245688e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document -1.682350633181197e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document -1.663983380609724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document -1.710187842689243e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document -1.5733697527539038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document -1.6972104757911438e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document -1.6610142847616577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document -1.61094882403031e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document -1.4789207305138325e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document -1.639299617676302e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document -1.3241204512116132e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document -8.582260726625535e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document -8.213000975576739e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document -9.549247732811947e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document -9.17242785339013e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document -7.632868223725218e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document -8.674401118222175e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document -9.124384255505347e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document -8.344222222417358e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document -8.992299957499065e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document -8.76689497361025e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document -7.973396239586015e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document -9.006935606644125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document -8.725545954955498e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document -1.215449694669174e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document -3.3041720284158646e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document -2.0593512412624502e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document -1.893608946986248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document -1.737111666788535e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document -1.4915923449873955e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document -2.289370239067605e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document -2.8615335689614638e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document -8.847283630883125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document -1.8175470362373804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document -1.8152226683368038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document -1.789149655314284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document -1.7690523036477663e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document -1.8333732213753644e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document -1.8794105687718654e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document -1.721841156706417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document -2.0612008685724796e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document -1.9297370681336376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document -2.0188440409661018e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document -5.1741216329695265e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document -1.3417913926038429e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document -1.1010813016469651e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document -1.1252416134320087e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document -1.2801744104313002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document -1.3041514955795817e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document -1.3428837580879075e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document -1.320809382267804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document -1.3451566676555968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document -1.228284926657501e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document -1.2410599573923043e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document -1.3815343367377182e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document -1.3895126265148832e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document -1.2306773644401741e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document -1.32981021906281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document -1.101337469221607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document -1.513094184404692e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document -1.1073759547073234e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document -1.2879348765857567e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document -9.619595770228435e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document -1.2384340836286436e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document -1.1766667232211577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document -1.2871049236196452e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document -1.2010645926497744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document -1.3971428231518597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document -1.2283733550547932e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document -1.2659530508255308e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document -1.551775613074462e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document -1.1169413343776979e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document -1.1433700593712463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document -4.964773647323492e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document -1.0995586595687313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document -1.2957393071411267e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document -2.75899247407709e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document -2.8269344597344854e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document -2.329108187246831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document -2.4231761430460284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document -1.2434140512230442e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document -1.638718338352859e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document -3.272953556801187e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document -6.061314500486327e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document -1.2465979731210292e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document -1.2737557327967737e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document -1.038428658075627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document -2.61666472045566e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document -3.6506873212272224e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document -1.5066359138295701e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document -1.1166290872121178e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document -1.5546966228590285e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document -1.2583434625014828e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document -1.3398826881300862e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document -1.2944933160515968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document -1.0971437399901365e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document -1.2787922795775774e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document -1.404979227816985e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document -1.3344734431324463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document -4.886031157107555e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document -3.277261443596394e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document -3.5057957685786495e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document -3.287625301718589e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document -3.1370056372668855e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document -3.186092015785841e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document -7.271819324142512e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +6.322825248625475e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika diff --git a/ALCF/data-lists/polaris/open-web-math-train.txt b/ALCF/data-lists/polaris/open-web-math-train.txt index caab74fb9f5..6d86bd35a46 100644 --- a/ALCF/data-lists/polaris/open-web-math-train.txt +++ b/ALCF/data-lists/polaris/open-web-math-train.txt @@ -1,13 +1,13 @@ -0.001451215788905126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document -0.0014486847196258788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document -0.0008861032722895899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document -0.0018119590809459816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document -0.0008916937917547129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document -6.960128832809415e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document -0.002008403651063623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document -0.0014374900742131454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document -0.00180213596996716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document -0.001956178877532413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document -0.0008829547017667033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document -0.0008910853619157279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document -0.0018260998845299973 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.001451215788905126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train diff --git a/ALCF/data-lists/polaris/pes2o.txt b/ALCF/data-lists/polaris/pes2o.txt index 7bb62d46b62..47a7eb3ffd9 100644 --- a/ALCF/data-lists/polaris/pes2o.txt +++ b/ALCF/data-lists/polaris/pes2o.txt @@ -1,26 +1,26 @@ -0.0012499632072059553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document -0.00125398260359913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document -0.0012541704774729071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document -0.0012527268234360602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document -0.0012532925243737164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document -0.0012456396241204315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document -0.0012589894424352072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document -0.001508020123999618 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document -0.00333096950781965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document -0.0033233414614415547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document -0.003512387990689828 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document -0.0035091382940513126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document -0.003514155927147005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document -0.003327108000579638 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document -0.003329106196589836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document -0.003505604148738077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document -0.003324825759567855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document -0.0033248240149804913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document -0.0033385962112851358 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document -0.0035043186296553615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document -0.003340469505431529 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document -0.0035106889084796276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document -0.0033309469281030167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document -0.003340337858029757 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document -0.003505919861097801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document -0.0003882924098240512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0012499632072059553 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o diff --git a/ALCF/data-lists/polaris/reddit.txt b/ALCF/data-lists/polaris/reddit.txt index 4f46ee0e648..ef79bbc7c82 100644 --- a/ALCF/data-lists/polaris/reddit.txt +++ b/ALCF/data-lists/polaris/reddit.txt @@ -1,78 +1,78 @@ -0.0005759963691850877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document -0.0005959971675332674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document -0.0006026179290353799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document -0.0005824184320784846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document -0.0005854598548616037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document -0.0005903767055633473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document -0.0005930306490982049 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document -0.000569425602700746 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document -0.0005675060415179408 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document -0.0005772431621253389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document -0.0005678026053826858 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document -0.0005700398263483378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document -0.0005669467963528824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document -0.0005701015953324305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document -0.0005795907287413296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document -0.0005735602737531164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document -0.0005749862745842101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document -0.0005693257015931971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document -0.0005716568794795563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document -0.0005761083919774021 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document -0.0005688343169797355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document -0.0005807913190929842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document -0.0005710229258078636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document -0.0005704083039826862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document -0.0005862132348308056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document -0.0005717662049559556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document -0.0005858155213694451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document -0.0005812012281792392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document -0.0005803981414588498 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document -0.0005700102108287723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document -0.0005719243459052329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document -0.0005867253401661752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document -0.0005731087218860733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document -0.0005712197789109317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document -0.0005702376926310089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document -0.0005700411527742972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document -0.0005828090098178196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document -0.0005770140826168056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document -0.0005723509664597896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document -0.0005755499231836962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document -0.0005636407438471367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document -0.0005640281556500104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document -0.0005633159058766496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document -0.0005638034311151449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document -0.0005630066273073224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document -0.0005631803831128559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document -0.0005631228881679657 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document -0.0005628178701487633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document -0.0005624448092256196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document -0.0005620957024062329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document -0.0005614201504177484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document -0.0005616890951464056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document -0.0005611348559279058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document -0.0005604238061828518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document -0.0005603301490194237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document -0.0005607291294548833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document -0.0005605234569930727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document -0.0005613778566640694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document -0.0005610248539992471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document -0.0005599977416780475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document -0.0005603632562116935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document -0.0005599177479509897 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document -0.0005595202318298379 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document -0.0005600975633499175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document -0.0005614075491213365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document -0.000612563885043477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document -0.0005515469909644413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document -0.0005526782014946906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document -0.0005472463408095445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document -0.0005502284746004587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document -0.0005414514790555363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document -0.0005513499500134784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document -0.0005391391454105187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document -0.0005415836910001838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document -0.0005208132468536551 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document -0.0005889827143132871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document -0.0005822520817765276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document -0.0004173155230758696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0005759963691850877 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit diff --git a/ALCF/data-lists/polaris/stack.txt b/ALCF/data-lists/polaris/stack.txt index 971329159b0..a81e55f94a7 100644 --- a/ALCF/data-lists/polaris/stack.txt +++ b/ALCF/data-lists/polaris/stack.txt @@ -1,26 +1,26 @@ -0.0009994361338078242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document -0.001087156194657966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document -0.0010667737163656816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document -0.0009602877882124873 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document -0.0008968956271971105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document -0.0009198034843762967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document -0.0009423901016715341 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document -0.0009674094553686345 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document -0.0009858331322519164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document -0.0009970593645879198 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document -0.0010027035193731686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document -0.0010128291154221853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document -0.0010215631382631918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document -0.0010288663771461238 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document -0.0010346219929285867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document -0.00104544019940344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document -0.0010525172676724333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document -0.0010609529620775127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document -0.0010725892748610153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document -0.0010818563598181568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document -0.0010992760196793917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document -0.0011178992762079917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document -0.001124687532085676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document -0.001118303661267191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document -0.0010206825575416534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document -0.0005512280117499715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.0009994361338078242 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange diff --git a/ALCF/data-lists/polaris/starcoder.txt b/ALCF/data-lists/polaris/starcoder.txt index c675b0b84df..5c28dd55b6f 100644 --- a/ALCF/data-lists/polaris/starcoder.txt +++ b/ALCF/data-lists/polaris/starcoder.txt @@ -1,50 +1,50 @@ -0.004474659408857016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document -0.00409944473890653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document -0.005137179939941845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document -0.005143172251066109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document -0.005206134363352808 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document -0.004892747858974329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document -0.004844731352552902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document -0.005308320169123755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document -0.005124709815666577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document -0.005424710744483826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document -0.00538244648861977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document -0.0029107284679086853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document -0.0026825258998444705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document -0.0026904503191419243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document -0.002687906577174073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document -0.002850165346048818 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document -0.005322698571717847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document -0.004450334290869719 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document -0.004700990083440683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document -0.003903568556500995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document -0.00390561515396931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document -0.0039046402900912262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document -0.003907454839379547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document -0.0038583224578603824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document -0.0037914116657695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document -0.003786665266798682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document -0.003792000802430658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document -0.00319266847466091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document -0.0032658716699838944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document -0.0034801959532460023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document -0.0028307012092022594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document -0.0028420360878146276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document -0.0028410455248484914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document -0.00283497183526842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document -0.002840187195459487 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document -0.0028398709431369834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document -0.004364722843422023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document -0.004093255713117101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document -0.004092331079566252 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document -0.004005326985579649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document -0.0036205502856964207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document -0.003625316793034984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document -0.003604743435602363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document -0.0035405823343673125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document -0.0041601413517253945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document -0.005886303658937057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document -0.003600909532810332 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document -0.0034941365817168658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document -0.0004992164842980224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.004474659408857016 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder diff --git a/ALCF/data-lists/polaris/tulu.txt b/ALCF/data-lists/polaris/tulu.txt index a65ae2b93a3..e7a681d6600 100644 --- a/ALCF/data-lists/polaris/tulu.txt +++ b/ALCF/data-lists/polaris/tulu.txt @@ -1,66 +1,66 @@ -0.00032927705604725614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document -0.0002860154190878753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document -0.0002845217585425619 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document -0.0002743528685497456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document -0.00026025323737738766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document -0.00023493876414603155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document -0.00029665994994226705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document -0.00031808102075993956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document -0.00031813573046011285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document -0.0002711905171855542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document -0.00028892513401817095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document -0.00030003908676979083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document -0.00026839878771944684 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document -0.00029155935002690497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document -0.0002998624927624209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document -0.0003091705447974841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document -0.00026873195794309786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document -0.00027721873498527547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document -0.0002841662554024377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document -0.0002839461156551537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document -0.0002861705604659811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document -0.0002460995649635886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document -0.00019420142619795496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document -0.00021967677816173628 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document -0.0002620283200480949 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document -0.0002433390542188936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document -0.00021254976608350767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document -0.00022094815569522115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document -0.000342862378668244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document -0.00033784225259118157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document -0.0003367278459543952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document -0.00029843279042852765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document -0.0002926583661257988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document -0.00029320337282010673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document -0.00029281450669483455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document -0.0002915338187002653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document -0.0002864226923084572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document -0.00028643439083586396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document -0.00028253710956299054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document -0.0002810856078805806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document -0.00031474941344656715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document -0.0002139130222205655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document -0.0003084648871862831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document -0.0003309477872140129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document -0.0003360096824695161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document -0.0003355452655196557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document -0.00038119390366386037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document -0.00038078927630086064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document -0.0003386200917551554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document -0.0002158905159938882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document -0.00021621682877018768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document -0.00021553306942740535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document -0.00021581563462722296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document -0.0002157694110556169 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document -0.000215643699847159 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document -0.00021532716715168094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document -0.00021531221326022472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document -0.0002831801179028896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document -0.0002514844936507595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document -0.00031638782778107964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document -0.0002749197545278445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document -0.00026159721512464495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document -0.0002630052420096968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document -0.00031106811228913666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document -0.0002852973415334161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document -3.7555372465932136e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.00032927705604725614 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu diff --git a/ALCF/data-lists/polaris/wiki.txt b/ALCF/data-lists/polaris/wiki.txt index 7759120d876..55ba7680ad0 100644 --- a/ALCF/data-lists/polaris/wiki.txt +++ b/ALCF/data-lists/polaris/wiki.txt @@ -1,2 +1,2 @@ -0.003548077173506675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document -0.0018372203137874265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document +0.003548077173506675 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /eagle/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki From 9d1070415d4ac8e8fd4e9dba1423d45f004bb000 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:53:18 -0500 Subject: [PATCH 224/590] Update `.gitignore` --- .gitignore | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/.gitignore b/.gitignore index 0a1387f56af..5079fb8bf2c 100644 --- a/.gitignore +++ b/.gitignore @@ -1,6 +1,17 @@ # User Added +.jobenv +**.e[0-9]** +**.o[0-9]** +**.e6** +**.o6** +**.e9** +**.o9** **.e1** **.o1** +*.o17* +*.e17* +*.o1 +*.e1 deps/* OUTPUTS/* ALCF/OUTPUTS/* @@ -11,16 +22,12 @@ ALCF/OUTPUTS/* **index-cache** **pbslogs** ezpz -*.o17* -*.e17* *hostfile* .deepspeed_env *.DS_Store old/* **venv** *.json -*.o1 -*.e1 outputs/ venvs/ wandb/ From ec600e538de7db0357dde97fbc3ba10525c469fc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:53:39 -0500 Subject: [PATCH 225/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 119 ++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 94 insertions(+), 25 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 09e59784344..669ba6483c3 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -101,6 +101,7 @@ setup() { # via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit + make_data || exit # 10. Print job info printJobInfo || exit # 11. Print info about loaded modules and runtime environment @@ -280,10 +281,10 @@ printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" echo "- Using $(which python3)" - echo "- WORLD_SIZE:${WORLD_SIZE}" - echo "- NCCL: ${NCCL:-nccl}" - echo "- MODEL_TYPE: ${MODEL_TYPE}" - echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST}" + echo "- WORLD_SIZE:${WORLD_SIZE-}" + echo "- BACKEND: ${BE:-}" + echo "- MODEL_TYPE: ${MODEL_TYPE:-}" + echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } @@ -327,20 +328,26 @@ setupLauncher() { printf " %s" "$(printMagenta "${LAUNCHER}")" } -setDSlauncher() { - # launcher setting - outdir=$1 - export hfds="$outdir/hostfile_deepspeed" - export hfmpi="$outdir/hostfile_mpich" - [ -f "$hfds" ] || exit - [ -f "$hfmpi" ] || exit - export LAUNCHER=${LAUNCHER:-MPICH} - if [[ $LAUNCHER == "deepspeed" ]]; then - export launcher="" - else - export launcher="--force_multi --hostfile $hfds --launcher=${LAUNCHER} --launcher_args='-hostfile ${hfmpi}'" - fi -} +# ########################################################################## +# # setDSlauncher +# # +# # When launching with `deepspeed`, this will +# # set the appropriate keyword arguments to be passed to `deepspeed` +# ########################################################################## +# setDSlauncher() { +# # launcher setting +# outdir=$1 +# export hfds="$outdir/hostfile_deepspeed" +# export hfmpi="$outdir/hostfile_mpich" +# [ -f "$hfds" ] || exit +# [ -f "$hfmpi" ] || exit +# export LAUNCHER=${LAUNCHER:-MPICH} +# if [[ $LAUNCHER == "deepspeed" ]]; then +# export launcher="" +# else +# export launcher="--force_multi --hostfile $hfds --launcher=${LAUNCHER} --launcher_args='-hostfile ${hfmpi}'" +# fi +# } set_lr_args() { LR_ARGS="--lr ${LR} --lr-decay-style cosine" @@ -663,6 +670,11 @@ buildDSconfig() { } +############################################################################### +# sumWeights +# +# This will sum the weights (first column) from each line in the passed +# `file_list`. sumWeights() { local file_list=$1 weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") @@ -677,6 +689,50 @@ sumFiles() { done } +########################################### +# make_data +# +# This will run `make` in `megatron/data` +# prior to launching, ensuring that +# `megatron/data/helpers.cpp` +# is built appropriately. +########################################### +make_data() { + python3 -m pip install pybind11 + mdir="${WORKING_DIR}/megatron/data" + cd "${mdir}" && make && cd - +} + + +install_dependencies() { + depsfile="${WORKING_DIR}/requirements/requirements.txt" + echo "Installing remaining dependencies from ${depsfile}" + python3 -m pip install -r "${depsfile}" --require-virtualenv +} + +###################################################################### +# install_deepspeed_for_xpu: Install microsoft/DeepSpeed on PVC +# +# This will: +# 1. Clone rep +# 2. Checkout appropriate branch +# 3. Install into virtual environment +###################################################################### +install_deepspeed_for_xpu() { + echo "Building + Installing DeepSpeed on $(hostname)" + outdir="${WORKING_DIR}/deps/DeepSpeed" + mkdir -p "${outdir}" + git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" + cd "${outdir}" || exit + git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git + git fetch yizhou_ds + git checkout yizhou/kernel_path + python3 -m pip install --require-virtualenv -r requirements/requirements.txt + python setup.py develop |& tee build.log +} + + + ######################################################## # Setup / activate conda environment, # NOTE: @@ -857,6 +913,10 @@ setup_python() { printf "[python] %s" "$(printMagenta "${pystr}")" printf "\n" export "PYTHON_EXEC=$(which python3)" + if [[ ! -x "$(command -v deepspeed)" ]]; then + install_deepspeed_for_xpu || exit + fi + install_dependencies || exit } ###################################################################### @@ -926,26 +986,26 @@ setup_tokenizer_and_data() { # fallback to default values if necessary. ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] - # =====[Set DATA_FILE_LIST_FALLBACK based on current machine]============== + ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/aurora/dolma.txt" elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] # shellcheck source=./data-lists/sunspot/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma_v1_7_file_list.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma.txt" elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] if [[ "${PBS_O_HOST}" == sirius* ]]; then # -------------------[SIRIUS] # shellcheck source=./data-lists/sirius/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma_v1_7_file_list.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma.txt" elif [[ "${PBS_O_HOST}" == polaris* ]]; then # ---------------[POLARIS] # shellcheck source=./data-lists/polaris/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma_v1_7_file_list.txt" + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma.txt" fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" else # -----------------------------------------------------------[UNKNOWN] echo "Unknown hostname. Must manually specify DATA_FILE_LIST." fi - # ========================================================================== + ############################################################################ # set `dfl` to `dfl_fallback` if not passed as an argument, # use this data file list to call `setData` dfl="${1:-${dfl_fallback}}" @@ -977,6 +1037,11 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" } +################################################################################ +# generateDSconfig: Create and save a deepspeed config .json +# +# This will contain the appropriate variables as set in the current environment. +################################################################################ generateDSconfig() { for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ "$PP" "$DTYPE" @@ -1139,6 +1204,9 @@ $flops_profiler EOT } +############################################### +# Helper functions for printing colored text +############################################### printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } @@ -1171,6 +1239,7 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } -################ +########################### # call helpers_main() +########################### helpers_main From 168cdda60795c09006e004dcb39a01252a4584d2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 10:58:18 -0500 Subject: [PATCH 226/590] Add `ALCF/requirements/requirements.txt` --- ALCF/requirements/requirements.txt | 8 ++++++++ 1 file changed, 8 insertions(+) create mode 100644 ALCF/requirements/requirements.txt diff --git a/ALCF/requirements/requirements.txt b/ALCF/requirements/requirements.txt new file mode 100644 index 00000000000..a32964aa09c --- /dev/null +++ b/ALCF/requirements/requirements.txt @@ -0,0 +1,8 @@ +transformers +bitsandbytes +sentencepiece +einops +xgboost +fixedint +pybind11 +six From 7df932956ed876ac1adf9d75ec254c9f571497b7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 11:01:30 -0500 Subject: [PATCH 227/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 669ba6483c3..78f4605456d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -705,7 +705,7 @@ make_data() { install_dependencies() { - depsfile="${WORKING_DIR}/requirements/requirements.txt" + depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" echo "Installing remaining dependencies from ${depsfile}" python3 -m pip install -r "${depsfile}" --require-virtualenv } From 77ffd100cfbbe89423f58d37bad443448bdab002 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 11:09:19 -0500 Subject: [PATCH 228/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 78f4605456d..512fa629428 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -913,10 +913,10 @@ setup_python() { printf "[python] %s" "$(printMagenta "${pystr}")" printf "\n" export "PYTHON_EXEC=$(which python3)" + install_dependencies || exit if [[ ! -x "$(command -v deepspeed)" ]]; then install_deepspeed_for_xpu || exit fi - install_dependencies || exit } ###################################################################### From e884f15c1f60471b6ca983a35bf648fdf6161a16 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 11:49:34 -0500 Subject: [PATCH 229/590] Update `ALCF/helpers.sh,requirements/requirements.txt}` --- ALCF/helpers.sh | 10 +++++++--- ALCF/requirements/requirements.txt | 1 + 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 512fa629428..a1698998d9c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -708,6 +708,9 @@ install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" echo "Installing remaining dependencies from ${depsfile}" python3 -m pip install -r "${depsfile}" --require-virtualenv + if [[ ! -x "$(command -v deepspeed)" ]]; then + install_deepspeed_for_xpu || exit + fi } ###################################################################### @@ -724,11 +727,15 @@ install_deepspeed_for_xpu() { mkdir -p "${outdir}" git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" cd "${outdir}" || exit + echo "!! pwd: $(pwd)" git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git git fetch yizhou_ds git checkout yizhou/kernel_path python3 -m pip install --require-virtualenv -r requirements/requirements.txt + python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv python setup.py develop |& tee build.log + cd "${WORKING_DIR}" + echo "!! pwd: $(pwd)" } @@ -914,9 +921,6 @@ setup_python() { printf "\n" export "PYTHON_EXEC=$(which python3)" install_dependencies || exit - if [[ ! -x "$(command -v deepspeed)" ]]; then - install_deepspeed_for_xpu || exit - fi } ###################################################################### diff --git a/ALCF/requirements/requirements.txt b/ALCF/requirements/requirements.txt index a32964aa09c..7e3a2ae4a53 100644 --- a/ALCF/requirements/requirements.txt +++ b/ALCF/requirements/requirements.txt @@ -3,6 +3,7 @@ bitsandbytes sentencepiece einops xgboost +numpy<2.0 fixedint pybind11 six From fb49de80a23cae4e04d3557dca7a6ee85a8f6afa Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 17:47:16 -0500 Subject: [PATCH 230/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 29 ++++++++++++++++++++++------- 1 file changed, 22 insertions(+), 7 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a1698998d9c..00b8913de56 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -79,10 +79,12 @@ helpers_main() { # 13. Setup run command to be executed. ############################################################################## setup() { + printf "%s" "$(printBlue "${AURORA_GPT_HEADER}")" # 1. Identify machine we're on get_machine || exit # 2. Load `conda` environment setup_python || exit + install_dependencies || exit # 3. Determine WORLD_SIZE, etc. from `PBS_*` vars setup_ezpz || exit # 4. Set command line arguments to pass to `"${EXEC}"` @@ -706,10 +708,15 @@ make_data() { install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" - echo "Installing remaining dependencies from ${depsfile}" - python3 -m pip install -r "${depsfile}" --require-virtualenv + echo "Ensuring all dependencies from ${depsfile} installed..." + python3 -m pip install -r "${depsfile}" --require-virtualenv 1> /dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then - install_deepspeed_for_xpu || exit + mn=$(get_machine_name) + if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then + install_deepspeed_for_xpu || exit + fi + printf "!! No 'deepspeed' command found on %s" "${mn}" + printf "!! No deepsepeed in $(which python3)" fi } @@ -731,9 +738,9 @@ install_deepspeed_for_xpu() { git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git git fetch yizhou_ds git checkout yizhou/kernel_path - python3 -m pip install --require-virtualenv -r requirements/requirements.txt - python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv - python setup.py develop |& tee build.log + python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1> /dev/null + python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1> /dev/null + python setup.py develop 1> /dev/null cd "${WORKING_DIR}" echo "!! pwd: $(pwd)" } @@ -920,7 +927,6 @@ setup_python() { printf "[python] %s" "$(printMagenta "${pystr}")" printf "\n" export "PYTHON_EXEC=$(which python3)" - install_dependencies || exit } ###################################################################### @@ -1243,6 +1249,15 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } +export AURORA_GPT_HEADER=""" + █████╗ ██╗ ██╗██████╗ ██████╗ ██████╗ █████╗ ██████╗ ██████╗ ████████╗ +██╔══██╗██║ ██║██╔══██╗██╔═══██╗██╔══██╗██╔══██╗ ██╔════╝ ██╔══██╗╚══██╔══╝ +███████║██║ ██║██████╔╝██║ ██║██████╔╝███████║█████╗██║ ███╗██████╔╝ ██║ +██╔══██║██║ ██║██╔══██╗██║ ██║██╔══██╗██╔══██║╚════╝██║ ██║██╔═══╝ ██║ +██║ ██║╚██████╔╝██║ ██║╚██████╔╝██║ ██║██║ ██║ ╚██████╔╝██║ ██║ +╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ +""" + ########################### # call helpers_main() ########################### From 727232608119f3124c63739e8b9aa5a91444d592 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 19 Jun 2024 17:49:06 -0500 Subject: [PATCH 231/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 00b8913de56..72db2c8fda9 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -79,7 +79,7 @@ helpers_main() { # 13. Setup run command to be executed. ############################################################################## setup() { - printf "%s" "$(printBlue "${AURORA_GPT_HEADER}")" + # printf "%s" "$(printBlue "${AURORA_GPT_HEADER}")" # 1. Identify machine we're on get_machine || exit # 2. Load `conda` environment From f826667bc57693de7bfde098bd602b049d94e230 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 21 Jun 2024 10:35:28 -0500 Subject: [PATCH 232/590] Update `ALCF/helpers.sh` with kvs fix on Aurora --- ALCF/helpers.sh | 50 ++++++++++++++++++++++++++++++++----------------- 1 file changed, 33 insertions(+), 17 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 72db2c8fda9..89274442b60 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -390,6 +390,18 @@ get_batch_size_on_polaris() { } +################################################# +# Fix for distributed key value store on Aurora +################################################# +use_kvs_fix_on_aurora() { + export NUMEXPR_MAX_THREADS=64 + export CCL_KVS_MODE=mpi + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH +} + + ############################################################################## # `setParams`: Set / configure run options by parsing environment. # @@ -400,6 +412,7 @@ get_batch_size_on_polaris() { # ``` ############################################################################## setParams() { + FLASH_ARG="" LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" # +----[Parallelism Settings] -------------------------------------------+ # +------[Aurora]--------||-------[SunSpot]-------------+ @@ -408,8 +421,11 @@ setParams() { export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-4} # GRADIENT_ACC_STEPS MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] + use_kvs_fix_on_aurora + export NO_FLASH_ATTN=1 # FLASH ATTN DISABLED [2024-06-20] (until resolved...) ############################################################## # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! if [[ -n "${NO_FLASH_ATTN-}" ]]; then @@ -457,6 +473,7 @@ setParams() { export TP="${TP}" export PP="${PP:-1}" export SP="${SP:-1}" + export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" @@ -748,19 +765,12 @@ install_deepspeed_for_xpu() { ######################################################## -# Setup / activate conda environment, -# NOTE: -# -# Jerome's `frameworks_2024_5_v2` seems broken ?? -# - seems to be missing `python3 -c 'from mpi4py import MPI'` ??? -# - consequently, we leave the setup below commented out (for the time -# being): -# if [[ -z "${CONDA_PREFIX-}" ]]; then -# module use -a /home/jmitche1/anl_release/2024/q2 ; module load frameworks_2024_5_v2 -# else -# echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" -# fi +# Setup / activate conda environment(s) ######################################################## + +########################### +# Setup conda on Sunspot +########################### setup_conda_sunspot() { ###### check if CONDA_PREFIX non-empty ################ if [[ -z "${CONDA_PREFIX:-}" ]]; then @@ -771,13 +781,12 @@ setup_conda_sunspot() { fi } +########################### +# Setup conda on Aurora +########################### setup_conda_aurora() { if [[ -z "${CONDA_PREFIX:-}" ]]; then module use -a /soft/modulefiles ; module load frameworks/2024.1 - export CCL_KVS_MODE=mpi - export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH fi } @@ -810,6 +819,13 @@ setup_conda_polaris() { fi } +######################## +# setup_venv_from_conda +# +# Build (if necessary) a virtual environment +# on top of the active conda and +# activate it. +# ###################### setup_venv_from_conda() { if [[ -z "${CONDA_PREFIX}" ]]; then echo "!! No ${CONDA_PREFIX} found." # Exiting." From 26b846adcf583877b9f06bd32c487c70797f1b47 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 21 Jun 2024 10:40:07 -0500 Subject: [PATCH 233/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 89274442b60..c417165989b 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -394,11 +394,14 @@ get_batch_size_on_polaris() { # Fix for distributed key value store on Aurora ################################################# use_kvs_fix_on_aurora() { - export NUMEXPR_MAX_THREADS=64 export CCL_KVS_MODE=mpi export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + ######################################################### + # if not set, CCL will complain... ? + export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" + ######################################################### } @@ -421,20 +424,21 @@ setParams() { export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-4} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] use_kvs_fix_on_aurora - export NO_FLASH_ATTN=1 # FLASH ATTN DISABLED [2024-06-20] (until resolved...) - ############################################################## + ###################################################################### # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! + export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" FLASH_ARG="--use-flash-attn-builder" fi - ############################################################## + ###################################################################### # +--------[Polaris]-----------------------------------+ elif [[ $(hostname) == x3* ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" From bc7fbc6626bac86c81b0b79f7fb8f08060e80786 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 21 Jun 2024 10:54:20 -0500 Subject: [PATCH 234/590] Update `ALCF/README.md` --- ALCF/README.md | 279 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 274 insertions(+), 5 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index 3870921f5b1..c5c7a8b0d42 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -42,7 +42,7 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): 1. Load `conda` module and activate base environment: ```bash - PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python ``` -
[output]: @@ -53,7 +53,7 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): ```bash # [05:47:13 PM][foremans@x3001c0s13b1n0][/eagle/a/f/p/ar/Megatron-DeepSpeed-D/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python Using WORKING_DIR: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed No conda_prefix or virtual_env found in environment... Setting up conda... @@ -104,7 +104,7 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): ```bash # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setEnv + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed No conda_prefix or virtual_env found in environment... Setting up conda... @@ -199,6 +199,8 @@ will (1.) can be overridden dynamically at runtime using this technique. ```bash + # for systems other than Polaris, replace "polaris/books.txt" below with: + # "{aurora,sunspot}/books.txt", PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh ``` @@ -217,10 +219,277 @@ will (1.) The outputs should look _something_ like this, though YMMV (things change quick): +
[Aurora]: + + ```bash + #[🌌][10:45:59 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?] + $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python + + #[🌌][10:46:57 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?][aurora_nre_models_frameworks-2024.1] + (aurora_nre_models_frameworks-2024.1) $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt bash train_llama_alcf.sh > train-log-$(tstamp).log 2>&1 & + + Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + Running on: aurora + Using virtual_env: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 on top of conda from: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 + [python] Using: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 + Ensuring all dependencies from /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/requirements/requirements.txt installed... + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + ┌─────────────────────────────────────────────────────────────────────┐ + │ [savejobenv]: + │ • Writing PBS vars to: /home/foremans/.pbsenv + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [HOSTS]: + │ • [host:0] - x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + │ • [host:1] - x4711c1s3b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [DIST INFO]: + │ • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + │ • NHOSTS=2 + │ • NGPU_PER_HOST=12 + │ • NGPUS=24 + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [LAUNCH]: + │ • To launch across all available GPUs, use: + │ 'launch' ( = mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov ) + └─────────────────────────────────────────────────────────────────────┘ + 2024-06-21 10:47:09,771 - numexpr.utils - INFO - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. + 2024-06-21 10:47:09,772 - numexpr.utils - INFO - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. + 2024-06-21 10:47:09,772 - numexpr.utils - INFO - NumExpr defaulting to 8 threads. + /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages/pandas/core/computation/expressions.py:21: UserWarning: Pandas requires version '2.8.4' or n> + from pandas.core.computation.check import NUMEXPR_INSTALLED + /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1/lib/python3.9/runpy.py:127: RuntimeWarning: 'ezpz.jobs' found in sys.modules after import of package 'ezpz', but prior to execution of 'ezpz.jobs'; this may result in u> + warn(RuntimeWarning(msg)) + [2024-06-21 10:47:10][INFO][jobs:366] - Caught PBS_JOBID='684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov', pbsnf=PosixPath('/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov') from env. Saving jobenv! + [2024-06-21 10:47:10][WARNING][jobs:117] - /home/foremans/PBS-jobs/684084 already in /home/foremans/PBS-jobs.log, not appending !! + [2024-06-21 10:47:10][INFO][jobs:192] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.sh + [2024-06-21 10:47:10][INFO][jobs:220] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.json + [2024-06-21 10:47:10][INFO][jobs:233] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.yaml + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to `.jobenv` file in /home/foremans/PBS-jobs/684084/.jobenv + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to `.jobenv` file in /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][WARNING][jobs:154] - To use `launch` alias, be sure to: `source /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][INFO][jobs:277] - Writing PBS env vars to /home/foremans/PBS-jobs/684084 / jobenv{.sh, .yaml, .json} + [2024-06-21 10:47:10][WARNING][jobs:281] - Run: `source ./.jobenv` in your current shell to set job variables + [2024-06-21 10:47:10][INFO][jobs:374] - + + [DIST_INFO]: + • DEVICE=xpu + • DEVICE_ID=xpu:0 + • DISTRIBUTED_BACKEND=ccl + • GPUS_PER_NODE=12 + • HOSTS=['x4711c1s2b0n0', 'x4711c1s3b0n0'] + • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + • HOSTNAME=x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + • LOCAL_RANK=0 + • MACHINE=Aurora + • NUM_NODES=2 + • NGPUS=24 + • NODE_ID=0 + • RANK=0 + • SCHEDULER=PBS + • WORLD_SIZE_TOTAL=24 + • WORLD_SIZE_IN_USE=1 + [2024-06-21 10:47:10][CRITICAL][jobs:245] - To launch across ALL GPUs in your job, use: + LAUNCH_CMD=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + creating alias launch=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + Found ezpz! + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + Done with ezpz. + Not using flash-attn!! + LR_ARGS: --lr 0.0003 --lr-decay-style cosine --lr-warmup-fraction 0.05 + DS_CONFIG: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ds-configs/ds_stage1_mb4_gb768_pp1_bf16.json + ZS: 1, MB: 4, GB: 768, PP: 1, DTYPE: bf16 + Please see logs at: logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + Checkpoints will be saved to: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Setting up tokenizer with Llama2 + Using data_file_list: ./ALCF/data-lists/aurora/books.txt + Using tokenizer: Llama2. Setting up data with ./ALCF/data-lists/aurora/books.txt + Calling: setData() with ./ALCF/data-lists/aurora/books.txt + -------------------- + Updated environment: + DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt + NUM_DOCS: 3 + WEIGHT_SUM: 0.0072042092147565125 + DFL_STEM: books + DATA_CACHE_PATH: .cache/books/index-cache + DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt + -------------------- + [setData] DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt + [setData] TOKENIZER_FLAGS: --tokenizer-type Llama2Tokenizer --tokenizer-model /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokenizer.model + Requirement already satisfied: pybind11 in ./venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages (2.12.0) + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + make: Nothing to be done for 'default'. + /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR=/opt/aurora/24.086.0/CNDA/mpich/20231026/mpich-ofi-all-icc-default-pmix-gpu-drop20231026 + - Using /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 + - WORLD_SIZE:24 + - BACKEND: ccl + - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + + Currently Loaded Modules: + 1) mpich/icc-all-pmix-gpu/20231026 3) libfabric/1.15.2.0 5) cray-libpals/1.3.3 7) gmp/6.2.1-pcxzkau 9) mpc/1.3.1-dfagrna 11) intel_compute_runtime/release/803.29 13) frameworks/2024.1 + 2) mpich-config/collective-tuning/1024 4) cray-pals/1.3.3 6) spack-pe-gcc/0.7.0-24.086.0 8) mpfr/4.2.0-w7v7yjv 10) gcc/12.2.0 12) oneapi/release/2024.1 + + + + Saving environment to checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.env + Not currently running. Continuing! + Launching with: MPICH + mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py + Using data_cache_path: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache + + mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --no-bias-dropout-fusion --no-masked-softmax-fusion --no-gradient-accumulation-fusion > + + [!! NOTE] View output at: + logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log + Connected to tcp://x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov:7919 + Launching application eafe3e80-ad2e-4cee-a3e4-d63af2a77c66 + [2024-06-21 10:47:31,610] [INFO] [comm.py:161:init_deepspeed_backend] Initialize ccl backend + [2024-06-21 10:47:31,610] [INFO] [comm.py:637:init_distributed] cdb=None + [2024-06-21 10:47:31,610] [INFO] [comm.py:652:init_distributed] Not using the DeepSpeed or dist launchers, attempting to detect MPI environment... + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=15, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=23, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=12, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=13, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=3, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=14, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=16, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=17, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=18, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=19, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=20, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=21, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=22, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=6, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=8, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=10, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=0, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:668:init_distributed] Initializing TorchBackend in DeepSpeed with backend ccl + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=1, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=2, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=4, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=5, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=7, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=9, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=11, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=2/23][local_rank=2/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=6/23][local_rank=6/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=8/23][local_rank=8/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=1/23][local_rank=1/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=15/23][local_rank=3/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=17/23][local_rank=5/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=3/23][local_rank=3/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=13/23][local_rank=1/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=14/23][local_rank=2/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=4/23][local_rank=4/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=18/23][local_rank=6/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=20/23][local_rank=8/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=5/23][local_rank=5/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=21/23][local_rank=9/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=7/23][local_rank=7/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=23/23][local_rank=11/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=9/23][local_rank=9/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=12/23][local_rank=0/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=16/23][local_rank=4/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=10/23][local_rank=10/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=11/23][local_rank=11/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=19/23][local_rank=7/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=22/23][local_rank=10/11][node=0/1] + 2024-06-21 10:47:32][INFO][dist:240] - DistInfo={ + "DEVICE": "xpu", + "DEVICE_ID": "xpu:0", + "DISTRIBUTED_BACKEND": "ccl", + "GPUS_PER_NODE": 12, + "HOSTFILE": "/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov", + "HOSTNAME": "x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov", + "HOSTS": "['x4711c1s2b0n0', 'x4711c1s3b0n0']", + "LOCAL_RANK": 0, + "MACHINE": "Aurora", + "NGPUS": 24, + "NODE_ID": 0, + "NUM_NODES": 2, + "RANK": 0, + "SCHEDULER": "PBS", + "WORLD_SIZE_IN_USE": 24, + "WORLD_SIZE_TOTAL": 24 + } + + # [...clipped...] + + [2024-06-21 10:48:48][INFO][utils:307] - > elapsed time for building blendable dataset indices: 1.19 (sec) + [2024-06-21 10:48:48][INFO][utils:307] - > saving index map files + [2024-06-21 10:48:51][INFO][utils:307] - > finished saving index map files in 3.0829622745513916 seconds + [2024-06-21 10:48:51][INFO][utils:307] - > loading blendable dataset index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_index.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading blendable dataset sample index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_sample_index.npy + [2024-06-21 10:48:52][INFO][utils:307] - > finished loading in 0.30188989639282227 seconds + [2024-06-21 10:48:52][INFO][utils:307] - >> building dataset for /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document + [2024-06-21 10:48:52][INFO][utils:307] - > building dataset index ... + [2024-06-21 10:48:52][INFO][utils:307] - reading sizes... + [2024-06-21 10:48:52][INFO][utils:307] - reading pointers... + [2024-06-21 10:48:52][INFO][utils:307] - reading document index... + [2024-06-21 10:48:52][INFO][utils:307] - creating numpy buffer of mmap... + [2024-06-21 10:48:52][INFO][utils:307] - /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document.bin + [2024-06-21 10:48:52][INFO][utils:307] - creating memory view of numpy buffer... + [2024-06-21 10:48:52][INFO][utils:307] - > finished creating indexed dataset in 0.003112 seconds + [2024-06-21 10:48:52][INFO][utils:307] - number of documents: 7386 + [2024-06-21 10:48:52][INFO][utils:307] - > dataset split: + [2024-06-21 10:48:52][INFO][utils:307] - train: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [0, 7386) total of 7386 documents + [2024-06-21 10:48:52][INFO][utils:307] - validation: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents + [2024-06-21 10:48:52][INFO][utils:307] - test: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents + [2024-06-21 10:48:52][INFO][utils:307] - > loading doc-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_doc_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading sample-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_sample_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading shuffle-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_shuffle_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - loaded indexed file in 0.008 seconds + [2024-06-21 10:48:52][INFO][utils:307] - total number of samples: 34196233 + [2024-06-21 10:48:52][INFO][utils:307] - total number of epochs: 175 + [2024-06-21 10:48:52][INFO][utils:307] - > size of blendable dataset: 245361763 samples + [2024-06-21 10:48:52][INFO][utils:307] - >>> Finished building BlendableDataset in 4.613574266433716 seconds + [2024-06-21 10:48:52][INFO][pretrain_gpt_alcf:579] - > finished creating GPT datasets. Took: 45730179865763.24219s + [2024-06-21 10:48:53][INFO][training:88] - [after dataloaders are built] datetime=2024-06-21 10:48:53 + [2024-06-21 10:48:53][INFO][training:307] - done with setup ... + [2024-06-21 10:48:53][INFO][training:313] - training ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (63763.34, 63857.25) + train/valid/test-data-iterators-setup ..........: (12936.53, 13432.64) + [2024-06-21 10:48:53][INFO][training:88] - [before the start of training step] datetime=2024-06-21 10:48:53 + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:541:forward] Activation Checkpointing Information + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:542:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:543:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:545:forward] ----Synchronization False + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:546:forward] ----Profiling time in checkpointing False + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1867.64 | optimizer_gradients: 19.65 | optimizer_step: 46.07 + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[1.887433467970254e-08, 1.887433467970254e-08], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 25341.72 | bwd_microstep: 77707.38 | bwd_inner_microstep: 75751.84 | bwd_allreduce_microstep: 1955.54 | step_microstep: 2218.38 + [2024-06-21 10:50:42,168] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 25341.72 | bwd: 77707.38 | bwd_inner: 75751.84 | bwd_allreduce: 1955.54 | step: 2218.38 + [2024-06-21 10:50:42][INFO][training:1609] - iteration= 1/ 317892 | consumed_samples= 768 | consumed_tokens= 3145728 | elapsed_time_per_iteration_ms=108893.2 | learning_rate=1.88743e-08 | global_batch_size= 768 | lm loss=11.133188 | loss_scale=1.0 | actual_seqlen= 4096 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=7.053 | tokens_per_gpu_per_second_tgs=1203.674 | [LM]-TFLOPs=49.66 | [DS]-TFLOPs=73.32 | + [2024-06-21 10:50:42][INFO][utils:190] - [Rank 0] (after 1 iterations) memory (MB) | allocated: 18243.64111328125 | max allocated: 50664.2548828125 | reserved: 54556.0 | max reserved: 54556.0 + (min, max) time across ranks (ms): + forward-backward ...............................: (106622.81, 106624.28) + optimizer ......................................: (2221.02, 2234.98) + ``` + +
+
[Sunspot]: ```bash - # [09:07:32 AM] [foremans@x1921c0s0b0n0] ~/q/llm.devkit/Megatron-DeepSpeed  main !1 ?27 q4-drop 26s ✘ INT + # [09:07:32 AM][foremans@x1921c0s0b0n0][~/q/llm.devkit/Megatron-DeepSpeed][🌱 main][$!?] $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_llama_alcf.sh source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... @@ -402,7 +671,7 @@ will (1.)
[Polaris]: ```bash - # [09:31:35 AM] [foremans@x3112c0s13b0n0] ~/pol/p/a/Megatron-DeepSpeed  main !4 ?24 cu118-pt221 ✘ INT + # [09:31:35 AM][foremans@x3112c0s13b0n0][~/pol/p/a/Megatron-DeepSpeed][🌱 main][$!?] $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_llama_alcf.sh source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh Running on Polaris !! From f94b8454c806cde4d4575b095e441ebdfb64f622 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 21 Jun 2024 10:59:35 -0500 Subject: [PATCH 235/590] Update `ALCF/README.md` --- ALCF/README.md | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index c5c7a8b0d42..f4cb1acef30 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -269,13 +269,12 @@ will (1.) [2024-06-21 10:47:10][INFO][jobs:192] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.sh [2024-06-21 10:47:10][INFO][jobs:220] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.json [2024-06-21 10:47:10][INFO][jobs:233] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.yaml - [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to `.jobenv` file in /home/foremans/PBS-jobs/684084/.jobenv - [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to `.jobenv` file in /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv - [2024-06-21 10:47:10][WARNING][jobs:154] - To use `launch` alias, be sure to: `source /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /home/foremans/PBS-jobs/684084/.jobenv + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][WARNING][jobs:154] - To use launch alias, be sure to: source /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv [2024-06-21 10:47:10][INFO][jobs:277] - Writing PBS env vars to /home/foremans/PBS-jobs/684084 / jobenv{.sh, .yaml, .json} - [2024-06-21 10:47:10][WARNING][jobs:281] - Run: `source ./.jobenv` in your current shell to set job variables + [2024-06-21 10:47:10][WARNING][jobs:281] - Run: source ./.jobenv in your current shell to set job variables [2024-06-21 10:47:10][INFO][jobs:374] - - [DIST_INFO]: • DEVICE=xpu • DEVICE_ID=xpu:0 From 06357f4cb70e94303378201e3f961d08e6ca1ec8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 21 Jun 2024 12:35:56 -0500 Subject: [PATCH 236/590] Create `alcf-startup-time` --- megatron/training.py | 1842 ++++++++++++++++++++++++++---------------- pretrain_gpt_alcf.py | 50 +- 2 files changed, 1178 insertions(+), 714 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 0693f73f3bd..77b36a0bc0b 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1,12 +1,15 @@ # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. - """Pretrain utilities.""" from datetime import datetime import math -import sys, os +from pathlib import Path +import sys +import os import time import json + +# noqa: E402 # The earliest we can measure the start time. _TRAIN_START_TIME = time.time() import torch @@ -23,8 +26,8 @@ from megatron import is_last_rank from megatron import update_num_microbatches from megatron.core import mpu, tensor_parallel -from megatron import print_rank_0, is_rank_0 -from megatron import print_rank_last +# from megatron import print_rank_0, is_rank_0 +# from megatron import print_rank_last from megatron.checkpointing import load_checkpoint from megatron.checkpointing import save_checkpoint from megatron.model import Float16Module @@ -41,7 +44,12 @@ from megatron.data.data_samplers import build_pretraining_data_loader from megatron.utils import calc_params_l2_norm from megatron.core.pipeline_parallel import get_forward_backward_func -from megatron.utils import report_memory, throughput_calculator, checkpoint_throughput_calculator, update_rotary_pos_emb +from megatron.utils import ( + report_memory, + throughput_calculator, + checkpoint_throughput_calculator, + update_rotary_pos_emb, +) from megatron.model.vision.knn_monitor import compute_feature_bank from megatron.arguments import core_transformer_config_from_args @@ -54,14 +62,17 @@ from megatron.model.transformer import ParallelTransformerLayer import ezpz as ez import logging -import os + dlp = Profile("TRAINING") -from deepspeed import comm as dist +# from deepspeed import comm as dist -RANK = ez.get_rank() -WORLD_SIZE = ez.get_world_size() -log = logging.getLogger(__name__) +RANK: int = ez.get_rank() +WORLD_SIZE: int = ez.get_world_size() +DEVICE_TYPE: str = ez.get_torch_device() +DEVICE: torch.device = torch.device(DEVICE_TYPE) + +log: logging.Logger = logging.getLogger(__name__) log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") try: @@ -73,8 +84,8 @@ def print_datetime(string): """Note that this call will sync across all ranks.""" tdist.barrier() - time_str = datetime.now().strftime('%Y-%m-%d %H:%M:%S') - log.info('[' + string + '] datetime={} '.format(time_str)) + time_str = datetime.now().strftime("%Y-%m-%d %H:%M:%S") + log.info("[" + string + "] datetime={} ".format(time_str)) def num_floating_point_operations(args, batch_size): @@ -107,17 +118,20 @@ def num_floating_point_operations(args, batch_size): ) ) -''' + +""" Since v0.9.0, deepspeed.initialize() has forbidden simultaneous setting of args.deepspeed_config (Path) and ds_config dict. -So, we use ds_config dict which is the more flexible option. -''' +So, we use ds_config dict which is the more flexible option +""" + + def _create_ds_config_dict(): args = get_args() assert args is not None - if isinstance(args.deepspeed_config, dict) : + if isinstance(args.deepspeed_config, dict): ds_config_dict = args.deepspeed_config else: - with open(args.deepspeed_config, 'r', encoding='utf-8') as config_file: + with open(args.deepspeed_config, "r", encoding="utf-8") as config_file: ds_config_dict = json.load(config_file) if args.universal_checkpoint: ds_config_dict["checkpoint"] = {"load_universal": True} @@ -125,6 +139,7 @@ def _create_ds_config_dict(): args.deepspeed_config = None return ds_config_dict + @dlp.log def pretrain( train_valid_test_dataset_provider, @@ -168,12 +183,15 @@ def pretrain( Returns: model (torch.nn.Module) """ - # Initalize and get arguments, timers, and Tensorboard writer. - initialize_megatron(extra_args_provider=extra_args_provider, - args_defaults=args_defaults, external_args=external_args) + initialize_megatron( + extra_args_provider=extra_args_provider, + args_defaults=args_defaults, + external_args=external_args, + ) # Set pytorch JIT layer fusion options and warmup JIT functions. - if get_accelerator().device_name() == 'cuda': + # if get_accelerator().device_name() == "cuda": + if DEVICE_TYPE == 'cuda' and torch.cuda.is_available(): set_jit_fusion_options() # Adjust the startup time so it reflects the largest value. @@ -181,143 +199,174 @@ def pretrain( # image ... launches. before_allreduce = time.time() global _TRAIN_START_TIME - log.info(f"time to finish initialize_megatron: {time.time() - _TRAIN_START_TIME} seconds") - start_time_tensor = get_accelerator().DoubleTensor([_TRAIN_START_TIME]) + log.info( + f"time to finish initialize_megatron: {time.time() - _TRAIN_START_TIME} seconds" + ) + # start_time_tensor = DEVICE.DoubleTensor([_TRAIN_START_TIME]) + start_time_tensor = torch.tensor([_TRAIN_START_TIME], device=DEVICE) tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) - # torch.distributed.all_reduce(start_time_tensor, - # op=torch.distributed.ReduceOp.MIN) log.info(f"allreduce call time: {time.time()-before_allreduce} seconds") _TRAIN_START_TIME = start_time_tensor.item() - log.info('time to initialize megatron (seconds)={:.3f}'.format( - time.time() - _TRAIN_START_TIME)) - print_datetime('after megatron is initialized') - + log.info( + "time to initialize megatron (seconds)={:.3f}".format( + time.time() - _TRAIN_START_TIME + ) + ) + print_datetime("after megatron is initialized") args = get_args() - if os.getenv('DLIO_PROFILER_DATASET_DIR') is not None: - extra_trace_path = os.environ['DLIO_PROFILER_DATASET_DIR'] + assert args is not None + if os.getenv("DLIO_PROFILER_DATASET_DIR") is not None: + extra_trace_path = os.environ["DLIO_PROFILER_DATASET_DIR"] else: - extra_trace_path='' + extra_trace_path = "" os.makedirs(args.trace_dir, exist_ok=True) - PerfTrace.initialize_log(f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", process_id=ez.get_rank()) + PerfTrace.initialize_log( + f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", + f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", + process_id=ez.get_rank(), + ) timers = get_timers() assert args is not None assert timers is not None + from ezpz.profile import get_context_manager + cm = get_context_manager(rank=RANK, outdir=args.save) + with cm: + if args.deepspeed: + args.deepspeed_config_dict = _create_ds_config_dict() + if ( + "curriculum_learning" in args.deepspeed_config_dict + and "enabled" in args.deepspeed_config_dict["curriculum_learning"] + ): + args.curriculum_learning_legacy = args.deepspeed_config_dict[ + "curriculum_learning" + ]["enabled"] + if args.curriculum_learning_legacy and not args.no_pipeline_parallel: + from deepspeed.runtime.data_pipeline.curriculum_scheduler import ( + CurriculumScheduler, + ) - if args.deepspeed: - args.deepspeed_config_dict = _create_ds_config_dict() - if "curriculum_learning" in args.deepspeed_config_dict and \ - "enabled" in args.deepspeed_config_dict["curriculum_learning"]: - args.curriculum_learning_legacy = args.deepspeed_config_dict[ \ - "curriculum_learning"]["enabled"] - if args.curriculum_learning_legacy and not args.no_pipeline_parallel: - from deepspeed.runtime.data_pipeline.curriculum_scheduler \ - import CurriculumScheduler - args.curriculum_scheduler = CurriculumScheduler( \ - args.deepspeed_config_dict["curriculum_learning"]) - if "compression_training" in args.deepspeed_config_dict: - args.compression_training = True - - # Model, optimizer, and learning rate. - timers('model-and-optimizer-setup', log_level=0).start(barrier=True) - model, optimizer, opt_param_scheduler = setup_model_and_optimizer( - model_provider, model_type, teacher=False, data_post_process=data_post_process, - build_train_valid_test_datasets_provider=train_valid_test_dataset_provider) - timers('model-and-optimizer-setup').stop() - print_datetime('after model, optimizer, and learning rate ' - 'scheduler are built') - - # Data stuff. - timers('train/valid/test-data-iterators-setup', log_level=0).start( - barrier=True) - if args.virtual_pipeline_model_parallel_size is not None: - all_data_iterators = [ - build_train_valid_test_data_iterators( - train_valid_test_dataset_provider) - for _ in range(len(model)) - ] - train_data_iterator = [data_iterators[0] - for data_iterators in all_data_iterators] - valid_data_iterator = [data_iterators[1] - for data_iterators in all_data_iterators] - test_data_iterator = [data_iterators[2] - for data_iterators in all_data_iterators] - else: - train_data_iterator, valid_data_iterator, test_data_iterator \ - = build_train_valid_test_data_iterators( - train_valid_test_dataset_provider) - if args.data_efficiency_curriculum_learning: - if args.deepspeed_dataloader is not None: - # We use args to pass the deepspeed_dataloader because adding - # output to setup_model_and_optimizer will break the API for other - # cases. We clear args.deepspeed_dataloader after updating - # train_data_iterator because args will be saved in checkpoint and - # attempting to save the whole deepspeed_dataloader will lead to - # "AttributeError: Can't pickle local object...". - train_data_iterator = iter(args.deepspeed_dataloader) - args.deepspeed_dataloader = None + args.curriculum_scheduler = CurriculumScheduler( + args.deepspeed_config_dict["curriculum_learning"] + ) + if "compression_training" in args.deepspeed_config_dict: + args.compression_training = True + + # Model, optimizer, and learning rate. + timers("model-and-optimizer-setup", log_level=0).start(barrier=True) + model, optimizer, opt_param_scheduler = setup_model_and_optimizer( + model_provider, + model_type, + teacher=False, + data_post_process=data_post_process, + build_train_valid_test_datasets_provider=train_valid_test_dataset_provider, + ) + timers("model-and-optimizer-setup").stop() + print_datetime("after model, optimizer, and learning rate " "scheduler are built") + # Data stuff. + timers("train/valid/test-data-iterators-setup", log_level=0).start(barrier=True) + if args.virtual_pipeline_model_parallel_size is not None: + all_data_iterators = [ + build_train_valid_test_data_iterators(train_valid_test_dataset_provider) + for _ in range(len(model)) + ] + train_data_iterator = [ + data_iterators[0] for data_iterators in all_data_iterators + ] + valid_data_iterator = [ + data_iterators[1] for data_iterators in all_data_iterators + ] + test_data_iterator = [ + data_iterators[2] for data_iterators in all_data_iterators + ] else: - train_data_iterator = None - timers('train/valid/test-data-iterators-setup').stop() - print_datetime('after dataloaders are built') - - # args.teacher_model is used as global variable to pass the teacher model - # for knowledge distillation. Users do not need to set it in the command - # line to use kd, but users do need to provide teacher model configurations - # like args.num_layers_teacher as described in setup_teacher_model() - args.teacher_model = None - if args.mos or args.kd: # Set up teacher model - args.teacher_model = setup_teacher_model(args, model_provider) - - # Print setup timing. - log.info('done with setup ...') - timers.log(['model-and-optimizer-setup', - 'train/valid/test-data-iterators-setup'], barrier=True) - + train_data_iterator, valid_data_iterator, test_data_iterator = ( + build_train_valid_test_data_iterators(train_valid_test_dataset_provider) + ) + if args.data_efficiency_curriculum_learning: + if args.deepspeed_dataloader is not None: + # We use args to pass the deepspeed_dataloader because adding + # output to setup_model_and_optimizer will break the API for other + # cases. We clear args.deepspeed_dataloader after updating + # train_data_iterator because args will be saved in checkpoint and + # attempting to save the whole deepspeed_dataloader will lead to + # "AttributeError: Can't pickle local object...". + train_data_iterator = iter(args.deepspeed_dataloader) + args.deepspeed_dataloader = None + else: + train_data_iterator = None + timers("train/valid/test-data-iterators-setup").stop() + print_datetime("after dataloaders are built") + # args.teacher_model is used as global variable to pass the teacher model + # for knowledge distillation. Users do not need to set it in the command + # line to use kd, but users do need to provide teacher model configurations + # like args.num_layers_teacher as described in setup_teacher_model() + args.teacher_model = None + if args.mos or args.kd: # Set up teacher model + args.teacher_model = setup_teacher_model(args, model_provider) + # Print setup timing. + log.info("done with setup ...") + timers.log( + ["model-and-optimizer-setup", "train/valid/test-data-iterators-setup"], + barrier=True, + ) if not args.skip_train: - log.info('training ...') - - if args.dataloader_type == 'cyclic' and args.retro_add_retriever: + log.info("training ...") + if args.dataloader_type == "cyclic" and args.retro_add_retriever: args.train_iters = args.retro_cyclic_train_iters log.info("retro cyclic train iters : %d" % args.train_iters) - iteration = 0 if args.do_train and args.train_iters > 0: - iteration = train(forward_step_func, - model, optimizer, opt_param_scheduler, - train_data_iterator, valid_data_iterator, - process_non_loss_data_func) - - print_datetime('after training is done') + iteration = train( + forward_step_func, + model, + optimizer, + opt_param_scheduler, + train_data_iterator, + valid_data_iterator, + process_non_loss_data_func, + ) + print_datetime("after training is done") # Clean the model if args.compression_training: model = [redundancy_clean(model[0], args.deepspeed_config_dict, mpu)] - if args.save and iteration != 0: save_checkpoint(iteration, model, optimizer, opt_param_scheduler) else: - log.info('skipping training (--skip-train is on) ...') - + log.info("skipping training (--skip-train is on) ...") iteration = args.iteration - config = core_transformer_config_from_args(args) if args.do_valid: - prefix = f'iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from validation set' - evaluate_and_print_results(prefix, forward_step_func, - valid_data_iterator, model, - iteration, process_non_loss_data_func, config, - verbose=True, write_to_tensorboard=not args.skip_train) - + prefix = f"iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from validation set" + evaluate_and_print_results( + prefix, + forward_step_func, + valid_data_iterator, + model, + iteration, + process_non_loss_data_func, + config, + verbose=True, + write_to_tensorboard=not args.skip_train, + ) if args.do_test: - prefix = f'iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from test set' - evaluate_and_print_results(prefix, forward_step_func, - test_data_iterator, model, - iteration, process_non_loss_data_func, config, - verbose=True, write_to_tensorboard=not args.skip_train, test=True) + prefix = f"iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from test set" + evaluate_and_print_results( + prefix, + forward_step_func, + test_data_iterator, + model, + iteration, + process_non_loss_data_func, + config, + verbose=True, + write_to_tensorboard=not args.skip_train, + test=True, + ) return model + @dlp.log def update_train_iters(args): - # For iteration-based training, we don't need to do anything if args.train_iters: return @@ -339,16 +388,15 @@ def update_train_iters(args): update_num_microbatches(0, consistency_check=False) # Constant phase # Note that we throw away any partial last batch. - iterations += (args.train_samples - consumed_samples) // \ - args.global_batch_size + iterations += (args.train_samples - consumed_samples) // args.global_batch_size args.train_iters = iterations - log.info('setting training iterations to {}'.format(args.train_iters)) + log.info("setting training iterations to {}".format(args.train_iters)) + @dlp.log -def setup_teacher_model(args, model_provider): - - log.info('***>>>>> Student model checkpoint iteration:{}'.format(args.iteration)) +def setup_teacher_model(args, model_provider): + log.info("***>>>>> Student model checkpoint iteration:{}".format(args.iteration)) iteration_stuent = args.iteration num_layers_student = args.num_layers num_experts_student = args.num_experts @@ -356,7 +404,7 @@ def setup_teacher_model(args, model_provider): num_attention_heads_student = args.num_attention_heads load_student = args.load - log.info('***>>>>> Setting up the teacher model') + log.info("***>>>>> Setting up the teacher model") args.num_layers = args.num_layers_teacher args.num_experts = args.num_experts_teacher @@ -364,7 +412,7 @@ def setup_teacher_model(args, model_provider): args.num_attention_heads = args.num_attention_heads_teacher args.load = args.load_teacher teacher_model, _, _ = load_model_weights_only(model_provider) - log.info('***>>>>> Teacher model:{}'.format(teacher_model)) + log.info("***>>>>> Teacher model:{}".format(teacher_model)) args.num_layers = num_layers_student args.num_experts = num_experts_student @@ -374,18 +422,25 @@ def setup_teacher_model(args, model_provider): args.iteration = iteration_stuent return teacher_model + + @dlp.log -def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap_with_ddp=True): +def get_model( + model_provider_func, model_type=ModelType.encoder_or_decoder, wrap_with_ddp=True +): """Build the model.""" args = get_args() assert args is not None args.model_type = model_type # Build model. - if mpu.get_pipeline_model_parallel_world_size() > 1 and \ - args.virtual_pipeline_model_parallel_size is not None: - assert model_type != ModelType.encoder_and_decoder, \ - "Interleaved schedule not supported for model with both encoder and decoder" + if ( + mpu.get_pipeline_model_parallel_world_size() > 1 + and args.virtual_pipeline_model_parallel_size is not None + ): + assert ( + model_type != ModelType.encoder_and_decoder + ), "Interleaved schedule not supported for model with both encoder and decoder" model = [] for i in range(args.virtual_pipeline_model_parallel_size): mpu.set_virtual_pipeline_model_parallel_rank(i) @@ -393,8 +448,7 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap pre_process = mpu.is_pipeline_first_stage() post_process = mpu.is_pipeline_last_stage() this_model = model_provider_func( - pre_process=pre_process, - post_process=post_process + pre_process=pre_process, post_process=post_process ) this_model.model_type = model_type model.append(this_model) @@ -405,37 +459,37 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap add_decoder = True if model_type == ModelType.encoder_and_decoder: if mpu.get_pipeline_model_parallel_world_size() > 1: - assert args.pipeline_model_parallel_split_rank is not None, \ - "Split rank needs to be specified for model with both encoder and decoder" + assert ( + args.pipeline_model_parallel_split_rank is not None + ), "Split rank needs to be specified for model with both encoder and decoder" rank = mpu.get_pipeline_model_parallel_rank() split_rank = args.pipeline_model_parallel_split_rank world_size = mpu.get_pipeline_model_parallel_world_size() pre_process = rank == 0 or rank == split_rank - post_process = (rank == (split_rank - 1)) or ( - rank == (world_size - 1)) + post_process = (rank == (split_rank - 1)) or (rank == (world_size - 1)) add_encoder = mpu.is_pipeline_stage_before_split() add_decoder = mpu.is_pipeline_stage_after_split() model = model_provider_func( pre_process=pre_process, post_process=post_process, add_encoder=add_encoder, - add_decoder=add_decoder) + add_decoder=add_decoder, + ) else: model = model_provider_func( - pre_process=pre_process, - post_process=post_process + pre_process=pre_process, post_process=post_process ) model.model_type = model_type - if not isinstance(model, list): model = [model] # Disallow training and inference with Transformer Engine # for non-GPT models args.allow_transformer_engine = all([type(m) == GPTModel for m in model]) - assert args.allow_transformer_engine or args.transformer_impl == 'local', \ - 'Transformer Engine is only approved for GPT models' + assert ( + args.allow_transformer_engine or args.transformer_impl == "local" + ), "Transformer Engine is only approved for GPT models" # Set tensor model parallel attributes if not set. # Only parameters that are already tensor model parallel have these @@ -443,16 +497,31 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap # are set for all params so the optimizer can use them. for model_module in model: for param in model_module.parameters(): - tensor_parallel.set_defaults_if_not_set_tensor_model_parallel_attributes(param) + tensor_parallel.set_defaults_if_not_set_tensor_model_parallel_attributes( + param + ) # Print number of parameters. if mpu.get_data_parallel_rank() == 0: - print(' > number of parameters on (tensor, pipeline) ' - 'model parallel rank ({}, {})={}'.format( - mpu.get_tensor_model_parallel_rank(), - mpu.get_pipeline_model_parallel_rank(), - sum([sum([p.ds_numel if hasattr(p,'ds_id') else p.nelement() for p in model_module.parameters()]) - for model_module in model])), flush=True) + print( + " > number of parameters on (tensor, pipeline) " + "model parallel rank ({}, {})={}".format( + mpu.get_tensor_model_parallel_rank(), + mpu.get_pipeline_model_parallel_rank(), + sum( + [ + sum( + [ + p.ds_numel if hasattr(p, "ds_id") else p.nelement() + for p in model_module.parameters() + ] + ) + for model_module in model + ] + ), + ), + flush=True, + ) if args.deepspeed: return model @@ -460,34 +529,47 @@ def get_model(model_provider_func, model_type=ModelType.encoder_or_decoder, wrap # GPU allocation. for model_module in model: model_module.to(get_accelerator().current_device_name()) - # Fp16 conversion. if args.fp16 or args.bf16: model = [Float16Module(model_module, args) for model_module in model] if wrap_with_ddp: - if args.DDP_impl == 'torch': + if args.DDP_impl == "torch": i = get_accelerator().current_device() - model = [torchDDP(model_module, device_ids=[i], output_device=i, - process_group=mpu.get_data_parallel_group()) - for model_module in model] - - elif args.DDP_impl == 'local': - model = [LocalDDP(model_module, - args.accumulate_allreduce_grads_in_fp32, - args.use_contiguous_buffers_in_local_ddp) - for model_module in model] + model = [ + torchDDP( + model_module, + device_ids=[i], + output_device=i, + process_group=mpu.get_data_parallel_group(), + ) + for model_module in model + ] + + elif args.DDP_impl == "local": + model = [ + LocalDDP( + model_module, + args.accumulate_allreduce_grads_in_fp32, + args.use_contiguous_buffers_in_local_ddp, + ) + for model_module in model + ] # broad cast params from data parallel src rank to other data parallel ranks if args.data_parallel_random_init: for model_module in model: model_module.broadcast_params() else: - raise NotImplementedError('Unknown DDP implementation specified: ' - '{}. Exiting.'.format(args.DDP_impl)) + raise NotImplementedError( + "Unknown DDP implementation specified: " "{}. Exiting.".format( + args.DDP_impl + ) + ) return model + @dlp.log def get_optimizer_param_scheduler(optimizer): """Build the learning rate scheduler.""" @@ -518,8 +600,7 @@ def get_optimizer_param_scheduler(optimizer): else: lr_warmup_steps = args.lr_warmup_samples else: - raise Exception( - 'either train-iters or train-samples should be provided.') + raise Exception("either train-iters or train-samples should be provided.") opt_param_scheduler = OptimizerParamScheduler( optimizer, @@ -533,14 +614,17 @@ def get_optimizer_param_scheduler(optimizer): wd_incr_steps=wd_incr_steps, wd_incr_style=args.weight_decay_incr_style, use_checkpoint_opt_param_scheduler=args.use_checkpoint_opt_param_scheduler, - override_opt_param_scheduler=args.override_opt_param_scheduler) + override_opt_param_scheduler=args.override_opt_param_scheduler, + ) return opt_param_scheduler + + @dlp.log def load_model_weights_only(model_provider_func): """Setup model and optimizer.""" args = get_args() - log.info('***>>>>> Args:{}'.format(args)) + log.info("***>>>>> Args:{}".format(args)) model = get_model(model_provider_func) @@ -549,58 +633,61 @@ def load_model_weights_only(model_provider_func): if args.deepspeed: # When loading just the model weights, ZeRO can be disabled. - if 'zero_optimization' in args.deepspeed_config_dict: - del args.deepspeed_config_dict['zero_optimization'] + if "zero_optimization" in args.deepspeed_config_dict: + del args.deepspeed_config_dict["zero_optimization"] model, optimizer, _, lr_scheduler = deepspeed.initialize( - model=model[0], - config=args.deepspeed_config_dict + model=model[0], config=args.deepspeed_config_dict ) - assert not isinstance(model, deepspeed.PipelineEngine), \ - 'Weight loading only mode is not supported in pipeline parallelism yet.' + assert not isinstance( + model, deepspeed.PipelineEngine + ), "Weight loading only mode is not supported in pipeline parallelism yet." model = [model] - print_datetime('before load checkpoint') + print_datetime("before load checkpoint") if args.load is not None: - iteration = load_checkpoint(model, optimizer, lr_scheduler, strict=True, load_only_weights=True) + iteration = load_checkpoint( + model, optimizer, lr_scheduler, strict=True, load_only_weights=True + ) - print_datetime('after load checkpoint weights') + print_datetime("after load checkpoint weights") return model, optimizer, lr_scheduler + @dlp.log -def setup_model_and_optimizer(model_provider_func, - model_type, - no_wd_decay_cond=None, - scale_lr_cond=None, - lr_mult=1.0, - teacher=False, - data_post_process=None, - build_train_valid_test_datasets_provider=None): +def setup_model_and_optimizer( + model_provider_func, + model_type, + no_wd_decay_cond=None, + scale_lr_cond=None, + lr_mult=1.0, + teacher=False, + data_post_process=None, + build_train_valid_test_datasets_provider=None, +): """Setup model and optimizer.""" args = get_args() - + assert args is not None model = get_model(model_provider_func, model_type) - # initialize the compression here student_global_steps = 0 if args.kd or args.mos: model, _, _, _ = deepspeed.initialize( - model=model[0], - args=args, - mpu=mpu if args.no_pipeline_parallel else None, - config=args.deepspeed_config_dict, - ) + model=model[0], + args=args, + mpu=mpu if args.no_pipeline_parallel else None, + config=args.deepspeed_config_dict, + ) model = [model] if args.load is not None: args.iteration = load_checkpoint(model, None, None, strict=False) else: args.iteration = 0 student_global_steps = model[0].global_steps - log.info('***>>>>> Student model, global step:{}'.format(student_global_steps)) - + log.info("***>>>>> Student model, global step:{}".format(student_global_steps)) if args.compression_training: model, _, _, _ = deepspeed.initialize( model=model[0], @@ -610,10 +697,7 @@ def setup_model_and_optimizer(model_provider_func, ) model = [model] model = [init_compression(model[0].module, args.deepspeed_config_dict, mpu)] - - unwrapped_model = unwrap_model(model, - (torchDDP, LocalDDP, Float16Module)) - + unwrapped_model = unwrap_model(model, (torchDDP, LocalDDP, Float16Module)) if args.inference: optimizer = None opt_param_scheduler = None @@ -621,15 +705,42 @@ def setup_model_and_optimizer(model_provider_func, if teacher: optimizer = None else: - optimizer = get_megatron_optimizer(model, no_wd_decay_cond, - scale_lr_cond, lr_mult) + optimizer = get_megatron_optimizer( + model, no_wd_decay_cond, scale_lr_cond, lr_mult + ) # opt_param_scheduler is the old lr_scheduler plus weight decay scheduling opt_param_scheduler = get_optimizer_param_scheduler(optimizer) - + # HAS_PYINSTRUMENT = False + # scalene_profiler = None + # pyinstrument_profiler = None + # scalene_start = time.perf_counter_ns() + # pyinstrument_start = time.perf_counter_ns() + # if RANK == 0: + # # if (scprof := os.environ.get("SCALENE_PROFILER", None)) is not None: + # if HAS_SCALENE and os.environ.get("SCALENE_PROFILER", None) is not None: + # try: + # from scalene import scalene_profiler + # HAS_SCALENE = True + # log.warning("Starting scalene profiler") + # scalene_profiler.start() + # except Exception as exc: + # log.exception('Unable to start scalene profiler') + # # if (pyiprof := os.environ.get("PYINSTRUMENT_PROFILER", None)) is not None: + # if os.environ.get("PYINSTRUMENT_PROFILER", None) is not None: + # try: + # import pyinstrument + # pyinstrument_profiler = pyinstrument.Profiler() + # pyinstrument_profiler.start() + # HAS_PYINSTRUMENT = True + # except Exception as exc: + # log.exception(f'Unable to start pyinstrument profiler.') if args.deepspeed: log.info("DeepSpeed is enabled.") pp = mpu.get_pipeline_model_parallel_world_size() - if args.data_efficiency_curriculum_learning and build_train_valid_test_datasets_provider is not None: + if ( + args.data_efficiency_curriculum_learning + and build_train_valid_test_datasets_provider is not None + ): train_ds = None # Only need to build dataset on tp rank 0 since Megatron has the # broadcast_data() function that broadcast data from tp rank 0. @@ -645,24 +756,30 @@ def setup_model_and_optimizer(model_provider_func, # We only need to build the training data here. And we follow # baseline's logic to build eval/test dataset later in # build_train_valid_test_data_iterators. - eval_iters = (args.train_iters // args.eval_interval + 1) * \ - args.eval_iters + eval_iters = ( + args.train_iters // args.eval_interval + 1 + ) * args.eval_iters test_iters = args.eval_iters - train_val_test_num_samples = [train_samples, - eval_iters * args.global_batch_size, - test_iters * args.global_batch_size] + train_val_test_num_samples = [ + train_samples, + eval_iters * args.global_batch_size, + test_iters * args.global_batch_size, + ] # Build the datasets. train_ds, _, _ = build_train_valid_test_datasets_provider( - train_val_test_num_samples) + train_val_test_num_samples + ) with Profile("deepspeed.initialize"): - model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = deepspeed.initialize( - model=model[0], - optimizer=optimizer, - args=args, - lr_scheduler=opt_param_scheduler, - training_data=train_ds, - mpu=mpu if args.no_pipeline_parallel else None, - config=args.deepspeed_config_dict, + model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = ( + deepspeed.initialize( + model=model[0], + optimizer=optimizer, + args=args, + lr_scheduler=opt_param_scheduler, + training_data=train_ds, + mpu=mpu if args.no_pipeline_parallel else None, + config=args.deepspeed_config_dict, + ) ) model.set_data_post_process_func(data_post_process) else: @@ -674,24 +791,55 @@ def setup_model_and_optimizer(model_provider_func, mpu=mpu if args.no_pipeline_parallel else None, config=args.deepspeed_config_dict, ) + # if RANK == 0: + # # if HAS_SCALENE and scalene_profiler is not None: + # # dtscalene = time.perf_counter_ns() - scalene_start + # # log.info(f"Finished with scalene profile. Took: {dtscalene:.5s}s") + # # scalene_profiler.stop() + # if HAS_PYINSTRUMENT and pyinstrument_profiler is not None: + # dtpyinstrument = time.perf_counter_ns() - pyinstrument_start + # pyinstrument_profiler.stop() + # pyinstrument_profiler.print(color=True) + # sp = getattr(args, 'save', None) + # fp = Path(os.getcwd()) if sp is None else Path(sp) + # fpath = Path(fp).joinpath( + # f'pyinstrument-profile-{ez.get_timestamp()}.html' + # ) + # log.info( + # 'Saving pyinstrument profile output to: ' + # f'{fpath.as_posix()}' + # ) + # pyinstrument_profiler.write_html(fpath) + # log.info(' '.join([ + # f'Finished with pyinstrument profiler.', + # f'Took: {dtpyinstrument:.5f}s', + # ])) if isinstance(model, deepspeed.PipelineEngine): # hack to get batch_fn from pretrain_gpt.py model.set_batch_fn(model.module._megatron_batch_fn) - - assert model.grid.get_pipe_parallel_rank() == mpu.get_pipeline_model_parallel_rank() - assert model.grid.get_slice_parallel_rank() == mpu.get_tensor_model_parallel_rank() + assert ( + model.grid.get_pipe_parallel_rank() + == mpu.get_pipeline_model_parallel_rank() + ) + assert ( + model.grid.get_slice_parallel_rank() + == mpu.get_tensor_model_parallel_rank() + ) assert model.grid.get_data_parallel_rank() == mpu.get_data_parallel_rank() model = [model] - # Compression has its own checkpoint loading path (e.g, loading both teacher and student models). So if compression is enabled, we skip the following checkpoint loading. + # Compression has its own checkpoint loading path (e.g, loading both teacher + # and student models). So if compression is enabled, we skip the following + # checkpoint loading. no_post_init_checkpoint_loading = args.kd or args.mos if not no_post_init_checkpoint_loading: if args.load is not None: timers = get_timers() - timers('load-checkpoint', log_level=0).start(barrier=True) + assert timers is not None + timers("load-checkpoint", log_level=0).start(barrier=True) args.iteration = load_checkpoint(model, optimizer, opt_param_scheduler) - timers('load-checkpoint').stop(barrier=True) - timers.log(['load-checkpoint']) + timers("load-checkpoint").stop(barrier=True) + timers.log(["load-checkpoint"]) else: args.iteration = 0 else: @@ -699,11 +847,14 @@ def setup_model_and_optimizer(model_provider_func, # We only support local DDP with multiple micro-batches. if len(model) > 1 or mpu.get_pipeline_model_parallel_world_size() > 1: - assert args.DDP_impl == 'local' + assert args.DDP_impl == "local" # get model without FP16 and/or TorchDDP wrappers - if args.iteration == 0 and len(unwrapped_model) == 1 \ - and hasattr(unwrapped_model[0], 'init_state_dict_from_bert'): + if ( + args.iteration == 0 + and len(unwrapped_model) == 1 + and hasattr(unwrapped_model[0], "init_state_dict_from_bert") + ): log.info("Initializing ICT from pretrained BERT model") unwrapped_model[0].init_state_dict_from_bert() if args.fp16: @@ -717,8 +868,9 @@ def setup_model_and_optimizer(model_provider_func, @dlp.log -def train_step(forward_step_func, data_iterator, - model, optimizer, opt_param_scheduler, config): +def train_step( + forward_step_func, data_iterator, model, optimizer, opt_param_scheduler, config +): """Single training step.""" args = get_args() timers = get_timers() @@ -729,7 +881,9 @@ def train_step(forward_step_func, data_iterator, assert isinstance(model[0], deepspeed.PipelineEngine) loss = model[0].train_batch(data_iter=data_iterator) additional_losses = model[0].get_additional_losses() - loss_key = 'lm loss' if additional_losses is None else 'loss' # use "lm loss" for backward compatibility + loss_key = ( + "lm loss" if additional_losses is None else "loss" + ) # use "lm loss" for backward compatibility loss_dict = OrderedDict({loss_key: loss}) if additional_losses is not None: loss_dict.update(additional_losses) @@ -738,14 +892,13 @@ def train_step(forward_step_func, data_iterator, # Set grad to zero. if not args.deepspeed: - if args.DDP_impl == 'local' and args.use_contiguous_buffers_in_local_ddp: + if args.DDP_impl == "local" and args.use_contiguous_buffers_in_local_ddp: for partition in model: partition.zero_grad_buffer() optimizer.zero_grad() # Forward pass. - timers('forward-backward', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("forward-backward", log_level=1).start(barrier=args.barrier_with_L1_time) forward_backward_func = get_forward_backward_func() if args.mos or args.kd: # args.teacher_forward is used as global variable to enable kd loss @@ -765,12 +918,13 @@ def train_step(forward_step_func, data_iterator, seq_length=args.seq_length, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, - forward_only=False) + forward_only=False, + ) # reset timers if necessary if config.timers is None: config.timers = timers - timers('forward-backward').stop() + timers("forward-backward").stop() if args.mos or args.kd: args.teacher_forward = False @@ -784,24 +938,23 @@ def train_step(forward_step_func, data_iterator, # Vision gradients. if args.vision_pretraining and args.vision_pretraining_type == "dino": - unwrapped_model = unwrap_model(model[0], - (torchDDP, LocalDDP, Float16Module)) + unwrapped_model = unwrap_model(model[0], (torchDDP, LocalDDP, Float16Module)) unwrapped_model.cancel_gradients_last_layer(args.curr_iteration) # Update parameters. - timers('optimizer', log_level=1).start(barrier=args.barrier_with_L1_time) + timers("optimizer", log_level=1).start(barrier=args.barrier_with_L1_time) if args.deepspeed: - increment = get_num_microbatches() * \ - args.micro_batch_size * \ - args.data_parallel_size + increment = ( + get_num_microbatches() * args.micro_batch_size * args.data_parallel_size + ) try: - model[0].step(lr_kwargs={'increment': increment}) + model[0].step(lr_kwargs={"increment": increment}) update_successful = model[0].was_step_applied() except Exception: update_successful = False else: update_successful, grad_norm, num_zeros_in_grad = optimizer.step(args, timers) - timers('optimizer').stop() + timers("optimizer").stop() # Gather params. if not args.deepspeed and update_successful: @@ -809,8 +962,7 @@ def train_step(forward_step_func, data_iterator, # Vision momentum. if args.vision_pretraining and args.vision_pretraining_type == "dino": - unwrapped_model = unwrap_model(model[0], - (torchDDP, LocalDDP, Float16Module)) + unwrapped_model = unwrap_model(model[0], (torchDDP, LocalDDP, Float16Module)) unwrapped_model.update_momentum(args.curr_iteration) # Update learning rate. @@ -821,13 +973,15 @@ def train_step(forward_step_func, data_iterator, loss_reduced = {} for key in losses_reduced[0]: losses_reduced_for_key = [x[key] for x in losses_reduced] - loss_reduced[key] = sum(losses_reduced_for_key) / len(losses_reduced_for_key) + loss_reduced[key] = sum(losses_reduced_for_key) / len( + losses_reduced_for_key + ) return loss_reduced, skipped_iter, grad_norm, num_zeros_in_grad else: if update_successful: - increment = get_num_microbatches() * \ - args.micro_batch_size * \ - args.data_parallel_size + increment = ( + get_num_microbatches() * args.micro_batch_size * args.data_parallel_size + ) opt_param_scheduler.step(increment=increment) skipped_iter = 0 else: @@ -842,180 +996,269 @@ def train_step(forward_step_func, data_iterator, loss_reduced = {} for key in losses_reduced[0]: losses_reduced_for_key = [x[key] for x in losses_reduced] - loss_reduced[key] = sum(losses_reduced_for_key) / len(losses_reduced_for_key) + loss_reduced[key] = sum(losses_reduced_for_key) / len( + losses_reduced_for_key + ) return loss_reduced, skipped_iter, grad_norm, num_zeros_in_grad return {}, skipped_iter, grad_norm, num_zeros_in_grad + @dlp.log -def training_log(loss_dict, total_loss_dict, learning_rate, iteration, - loss_scale, report_memory_flag, skipped_iter, - grad_norm, params_norm, num_zeros_in_grad, - model=None, optimizer=None): +def training_log( + loss_dict, + total_loss_dict, + learning_rate, + iteration, + loss_scale, + report_memory_flag, + skipped_iter, + grad_norm, + params_norm, + num_zeros_in_grad, + model=None, + optimizer=None, +): """Log training information such as losses, timing, ....""" args = get_args() timers = get_timers() writer = get_tensorboard_writer() wandb_metrics = {} # Advanced, skipped, and Nan iterations. - advanced_iters_key = 'advanced iterations' - skipped_iters_key = 'skipped iterations' - nan_iters_key = 'nan iterations' + advanced_iters_key = "advanced iterations" + skipped_iters_key = "skipped iterations" + nan_iters_key = "nan iterations" # Advanced iterations. if not skipped_iter: - total_loss_dict[advanced_iters_key] = total_loss_dict.get( - advanced_iters_key, 0) + 1 + total_loss_dict[advanced_iters_key] = ( + total_loss_dict.get(advanced_iters_key, 0) + 1 + ) else: if advanced_iters_key not in total_loss_dict: total_loss_dict[advanced_iters_key] = 0 # Skipped iterations. - total_loss_dict[skipped_iters_key] = total_loss_dict.get( - skipped_iters_key, 0) + skipped_iter + total_loss_dict[skipped_iters_key] = ( + total_loss_dict.get(skipped_iters_key, 0) + skipped_iter + ) # Update losses and set nan iterations got_nan = False for key in loss_dict: if not skipped_iter: - total_loss_dict[key] = total_loss_dict.get( - key, get_accelerator().FloatTensor([0.0])) + loss_dict[key] + total_loss_dict[key] = ( + total_loss_dict.get(key, get_accelerator().FloatTensor([0.0])) + + loss_dict[key] + ) else: value = loss_dict[key].float().sum().item() - is_nan = value == float('inf') or \ - value == -float('inf') or \ - value != value + is_nan = value == float("inf") or value == -float("inf") or value != value got_nan = got_nan or is_nan - total_loss_dict[nan_iters_key] = total_loss_dict.get( - nan_iters_key, 0) + int(got_nan) + total_loss_dict[nan_iters_key] = total_loss_dict.get(nan_iters_key, 0) + int( + got_nan + ) # Logging. timers_to_log = [ - 'forward-backward', - 'forward-compute', - 'backward-compute', - 'batch-generator', - 'forward-recv', - 'forward-send', - 'backward-recv', - 'backward-send', - 'forward-send-forward-recv', - 'forward-send-backward-recv', - 'backward-send-forward-recv', - 'backward-send-backward-recv', - 'forward-backward-send-forward-backward-recv', - 'layernorm-grads-all-reduce', - 'embedding-grads-all-reduce', - 'grads-all-reduce', - 'grads-reduce-scatter', - 'params-all-gather', - 'optimizer-copy-to-main-grad', - 'optimizer-unscale-and-check-inf', - 'optimizer-clip-main-grad', - 'optimizer-count-zeros', - 'optimizer-inner-step', - 'optimizer-copy-main-to-model-params', - 'optimizer'] + "forward-backward", + "forward-compute", + "backward-compute", + "batch-generator", + "forward-recv", + "forward-send", + "backward-recv", + "backward-send", + "forward-send-forward-recv", + "forward-send-backward-recv", + "backward-send-forward-recv", + "backward-send-backward-recv", + "forward-backward-send-forward-backward-recv", + "layernorm-grads-all-reduce", + "embedding-grads-all-reduce", + "grads-all-reduce", + "grads-reduce-scatter", + "params-all-gather", + "optimizer-copy-to-main-grad", + "optimizer-unscale-and-check-inf", + "optimizer-clip-main-grad", + "optimizer-count-zeros", + "optimizer-inner-step", + "optimizer-copy-main-to-model-params", + "optimizer", + ] # Calculate batch size. batch_size = ( - args.micro_batch_size - * args.data_parallel_size - * get_num_microbatches() + args.micro_batch_size * args.data_parallel_size * get_num_microbatches() ) total_iterations = ( - total_loss_dict[advanced_iters_key] - + total_loss_dict[skipped_iters_key] + total_loss_dict[advanced_iters_key] + total_loss_dict[skipped_iters_key] ) # Tensorboard values. # Timer requires all the ranks to call. - if args.log_timers_to_tensorboard and \ - (iteration % args.tensorboard_log_interval == 0): - timers.write(timers_to_log, writer, iteration, - normalizer=total_iterations) + if args.log_timers_to_tensorboard and ( + iteration % args.tensorboard_log_interval == 0 + ): + timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) if writer and (iteration % args.tensorboard_log_interval == 0): - writer.add_scalar('steps-vs-samples/y=steps,x=samples', iteration, args.consumed_train_samples) - writer.add_scalar('steps-vs-samples/y=samples,x=steps', args.consumed_train_samples, iteration) - writer.add_scalar('steps-vs-tokens/y=steps,x=tokens', iteration, args.consumed_train_tokens) - writer.add_scalar('steps-vs-tokens/y=tokens,x=steps', args.consumed_train_tokens, iteration) + writer.add_scalar( + "steps-vs-samples/y=steps,x=samples", iteration, args.consumed_train_samples + ) + writer.add_scalar( + "steps-vs-samples/y=samples,x=steps", args.consumed_train_samples, iteration + ) + writer.add_scalar( + "steps-vs-tokens/y=steps,x=tokens", iteration, args.consumed_train_tokens + ) + writer.add_scalar( + "steps-vs-tokens/y=tokens,x=steps", args.consumed_train_tokens, iteration + ) if args.log_learning_rate_to_tensorboard: wandb_metrics |= { - 'learning-rate/iteration': iteration, - 'learning-rate/learning-rate': learning_rate, + "learning-rate/iteration": iteration, + "learning-rate/learning-rate": learning_rate, } - writer.add_scalar('learning-rate/learning-rate', learning_rate, iteration) - writer.add_scalar('learning-rate/learning-rate vs samples', learning_rate, - args.consumed_train_samples) - writer.add_scalar('learning-rate/learning-rate vs tokens', learning_rate, - args.consumed_train_tokens) + writer.add_scalar("learning-rate/learning-rate", learning_rate, iteration) + writer.add_scalar( + "learning-rate/learning-rate vs samples", + learning_rate, + args.consumed_train_samples, + ) + writer.add_scalar( + "learning-rate/learning-rate vs tokens", + learning_rate, + args.consumed_train_tokens, + ) if args.log_batch_size_to_tensorboard: - writer.add_scalar('batch-size/batch-size', batch_size, iteration) - writer.add_scalar('batch-size/batch-size vs samples', batch_size, - args.consumed_train_samples) - writer.add_scalar('batch-size/batch-size vs tokens', batch_size, - args.consumed_train_tokens) + writer.add_scalar("batch-size/batch-size", batch_size, iteration) + writer.add_scalar( + "batch-size/batch-size vs samples", + batch_size, + args.consumed_train_samples, + ) + writer.add_scalar( + "batch-size/batch-size vs tokens", + batch_size, + args.consumed_train_tokens, + ) wandb_metrics |= { - 'lm-loss-training/iteration': iteration, - 'lm-loss-training/consumed_train_tokens': args.consumed_train_tokens, + "lm-loss-training/iteration": iteration, + "lm-loss-training/consumed_train_tokens": args.consumed_train_tokens, } for key in loss_dict: - wandb_metrics |= {f'lm-loss-training/{key}': loss_dict[key]} + wandb_metrics |= {f"lm-loss-training/{key}": loss_dict[key]} writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) - writer.add_scalar(f"lm-loss-training/{key}" + ' vs samples', loss_dict[key], - args.consumed_train_samples) - writer.add_scalar(f"lm-loss-training/{key}" + ' vs tokens', loss_dict[key], - args.consumed_train_tokens) + writer.add_scalar( + f"lm-loss-training/{key}" + " vs samples", + loss_dict[key], + args.consumed_train_samples, + ) + writer.add_scalar( + f"lm-loss-training/{key}" + " vs tokens", + loss_dict[key], + args.consumed_train_tokens, + ) if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: - writer.add_scalar('loss-scale/loss-scale', loss_scale, iteration) - writer.add_scalar('loss-scale/loss-scale vs samples', loss_scale, - args.consumed_train_samples) - writer.add_scalar('loss-scale/loss-scale vs tokens', loss_scale, - args.consumed_train_tokens) + writer.add_scalar("loss-scale/loss-scale", loss_scale, iteration) + writer.add_scalar( + "loss-scale/loss-scale vs samples", + loss_scale, + args.consumed_train_samples, + ) + writer.add_scalar( + "loss-scale/loss-scale vs tokens", + loss_scale, + args.consumed_train_tokens, + ) if args.log_world_size_to_tensorboard: - writer.add_scalar('world-size/world-size', args.world_size, iteration) - writer.add_scalar('world-size/world-size vs samples', args.world_size, - args.consumed_train_samples) - writer.add_scalar('world-size/world-size vs tokens', args.world_size, - args.consumed_train_tokens) + writer.add_scalar("world-size/world-size", args.world_size, iteration) + writer.add_scalar( + "world-size/world-size vs samples", + args.world_size, + args.consumed_train_samples, + ) + writer.add_scalar( + "world-size/world-size vs tokens", + args.world_size, + args.consumed_train_tokens, + ) if grad_norm is not None: - wandb_metrics |= {'training/grad-norm': grad_norm} - writer.add_scalar('grad-norm/grad-norm', grad_norm, iteration) - writer.add_scalar('grad-norm/grad-norm vs samples', grad_norm, - args.consumed_train_samples) - writer.add_scalar('grad-norm/grad-norm vs tokens', grad_norm, - args.consumed_train_tokens) + wandb_metrics |= {"training/grad-norm": grad_norm} + writer.add_scalar("grad-norm/grad-norm", grad_norm, iteration) + writer.add_scalar( + "grad-norm/grad-norm vs samples", grad_norm, args.consumed_train_samples + ) + writer.add_scalar( + "grad-norm/grad-norm vs tokens", grad_norm, args.consumed_train_tokens + ) if num_zeros_in_grad is not None: - wandb_metrics |= {'training/num-zeros': num_zeros_in_grad} - writer.add_scalar('num-zeros/num-zeros', num_zeros_in_grad, iteration) - writer.add_scalar('num-zeros/num-zeros vs samples', num_zeros_in_grad, - args.consumed_train_samples) - writer.add_scalar('num-zeros/num-zeros vs tokens', num_zeros_in_grad, - args.consumed_train_tokens) + wandb_metrics |= {"training/num-zeros": num_zeros_in_grad} + writer.add_scalar("num-zeros/num-zeros", num_zeros_in_grad, iteration) + writer.add_scalar( + "num-zeros/num-zeros vs samples", + num_zeros_in_grad, + args.consumed_train_samples, + ) + writer.add_scalar( + "num-zeros/num-zeros vs tokens", + num_zeros_in_grad, + args.consumed_train_tokens, + ) if params_norm is not None: - wandb_metrics |= {'training/params-norm': params_norm} - writer.add_scalar('params-norm/params-norm', params_norm, iteration) - writer.add_scalar('params-norm/params-norm vs samples', params_norm, - args.consumed_train_samples) - writer.add_scalar('params-norm/params-norm vs tokens', params_norm, - args.consumed_train_tokens) - if hasattr(args, 'actual_seq_length'): - writer.add_scalar('seqlen/actual_seq_length', args.actual_seq_length, - iteration) - writer.add_scalar('seqlen/actual_seq_length vs samples', args.actual_seq_length, - args.consumed_train_samples) - writer.add_scalar('seqlen/actual_seq_length vs tokens', args.actual_seq_length, - args.consumed_train_tokens) + wandb_metrics |= {"training/params-norm": params_norm} + writer.add_scalar("params-norm/params-norm", params_norm, iteration) + writer.add_scalar( + "params-norm/params-norm vs samples", + params_norm, + args.consumed_train_samples, + ) + writer.add_scalar( + "params-norm/params-norm vs tokens", + params_norm, + args.consumed_train_tokens, + ) + if hasattr(args, "actual_seq_length"): + writer.add_scalar( + "seqlen/actual_seq_length", args.actual_seq_length, iteration + ) + writer.add_scalar( + "seqlen/actual_seq_length vs samples", + args.actual_seq_length, + args.consumed_train_samples, + ) + writer.add_scalar( + "seqlen/actual_seq_length vs tokens", + args.actual_seq_length, + args.consumed_train_tokens, + ) if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - writer.add_scalar('seqlen/curriculum_seqlen', args.curriculum_seqlen, - iteration) - writer.add_scalar('seqlen/curriculum_seqlen vs samples', args.curriculum_seqlen, - args.consumed_train_samples) - writer.add_scalar('seqlen/curriculum_seqlen vs tokens', args.curriculum_seqlen, - args.consumed_train_tokens) + writer.add_scalar( + "seqlen/curriculum_seqlen", args.curriculum_seqlen, iteration + ) + writer.add_scalar( + "seqlen/curriculum_seqlen vs samples", + args.curriculum_seqlen, + args.consumed_train_samples, + ) + writer.add_scalar( + "seqlen/curriculum_seqlen vs tokens", + args.curriculum_seqlen, + args.consumed_train_tokens, + ) if args.random_ltd: - writer.add_scalar('seqlen/random_ltd_reserved_length', args.random_ltd_reserved_length, - iteration) - writer.add_scalar('seqlen/random_ltd_reserved_length vs samples', args.random_ltd_reserved_length, - args.consumed_train_samples) - writer.add_scalar('seqlen/random_ltd_reserved_length vs tokens', args.random_ltd_reserved_length, - args.consumed_train_tokens) + writer.add_scalar( + "seqlen/random_ltd_reserved_length", + args.random_ltd_reserved_length, + iteration, + ) + writer.add_scalar( + "seqlen/random_ltd_reserved_length vs samples", + args.random_ltd_reserved_length, + args.consumed_train_samples, + ) + writer.add_scalar( + "seqlen/random_ltd_reserved_length vs tokens", + args.random_ltd_reserved_length, + args.consumed_train_tokens, + ) if args.log_memory_to_tensorboard: mem_stats = torch.cuda.memory_stats() writer.add_scalar( @@ -1040,11 +1283,11 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, opt_stats = [0.0] * 8 opt_stats_2 = [0.0] * 4 for _, group in enumerate(optimizer.param_groups): - for _, param in enumerate(group['params']): - state_param = getattr(optimizer, 'state', None) + for _, param in enumerate(group["params"]): + state_param = getattr(optimizer, "state", None) if state_param is not None: - exp_avg_sq = state_param.get('exp_avg_sq', torch.tensor(0.)) - exp_avg = state_param.get('exp_avg', torch.tensor(0.)) + exp_avg_sq = state_param.get("exp_avg_sq", torch.tensor(0.0)) + exp_avg = state_param.get("exp_avg", torch.tensor(0.0)) opt_stats[0] += (torch.norm(exp_avg_sq).item()) ** 2 opt_stats[1] += (torch.norm(exp_avg_sq.sqrt()).item()) ** 2 opt_stats[2] += (torch.norm(exp_avg).item()) ** 2 @@ -1056,21 +1299,20 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, opt_stats_2[0] = max( opt_stats_2[0], abs(exp_avg_sq.max().item()), - abs(exp_avg_sq.min().item()) + abs(exp_avg_sq.min().item()), ) opt_stats_2[1] = max( - opt_stats_2[1], - exp_avg_sq.sqrt().abs_().max().item() + opt_stats_2[1], exp_avg_sq.sqrt().abs_().max().item() ) opt_stats_2[2] = max( opt_stats_2[2], abs(exp_avg.max().item()), - abs(exp_avg.min().item()) + abs(exp_avg.min().item()), ) opt_stats_2[3] = max( opt_stats_2[3], abs(param.max().item()), - abs(param.min().item()) + abs(param.min().item()), ) # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if args.zero_stage > 0: @@ -1078,328 +1320,459 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, # opt_stats = opt_stats.clone().detach() # opt_stats = get_accelerator().FloatTensor opt_stats = get_accelerator().FloatTensor(opt_stats) - torch.distributed.all_reduce(opt_stats, group=mpu.get_sequence_data_parallel_group()) + torch.distributed.all_reduce( + opt_stats, group=mpu.get_sequence_data_parallel_group() + ) # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, - group=mpu.get_sequence_data_parallel_group()) + torch.distributed.all_reduce( + opt_stats_2, + op=torch.distributed.ReduceOp.MAX, + group=mpu.get_sequence_data_parallel_group(), + ) if args.tensor_model_parallel_size > 1: # opt_stats = opt_stats.clone().detach() opt_stats = get_accelerator().FloatTensor(opt_stats) - torch.distributed.all_reduce(opt_stats, group=mpu.get_tensor_model_parallel_group()) + torch.distributed.all_reduce( + opt_stats, group=mpu.get_tensor_model_parallel_group() + ) # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, - group=mpu.get_tensor_model_parallel_group()) + torch.distributed.all_reduce( + opt_stats_2, + op=torch.distributed.ReduceOp.MAX, + group=mpu.get_tensor_model_parallel_group(), + ) if args.pipeline_model_parallel_size > 1: # opt_stats = opt_stats.clone().detach() opt_stats = get_accelerator().FloatTensor(opt_stats) - torch.distributed.all_reduce(opt_stats, group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.all_reduce( + opt_stats, group=mpu.get_pipeline_model_parallel_group() + ) # opt_stats_2 = opt_stats_2.clone().detach() opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - torch.distributed.all_reduce(opt_stats_2, op=torch.distributed.ReduceOp.MAX, - group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.all_reduce( + opt_stats_2, + op=torch.distributed.ReduceOp.MAX, + group=mpu.get_pipeline_model_parallel_group(), + ) wandb_metrics |= { - 'optimizer/learning_rate': learning_rate, - 'optimizer/iteration': args.iteration, - 'optimizer/consumed_train_tokens': args.consumed_train_tokens, - 'optimizer/variance_l2': opt_stats[0]**0.5, - 'optimizer/variance_sqrt_l2': opt_stats[1]**0.5, - 'optimizer/momentum_l2': opt_stats[2]**0.5, - 'optimizer/weight_l2': opt_stats[3]**0.5, - 'optimizer/variance_l1': opt_stats[4], - 'optimizer/variance_sqrt_l1': opt_stats[5], - 'optimizer/momentum_l1': opt_stats[6], - 'optimizer/weight_l1': opt_stats[7], - 'optimizer/variance_abs_max': opt_stats_2[0], - 'optimizer/variance_sqrt_abs_max': opt_stats_2[1], - 'optimizer/momentum_abs_max': opt_stats_2[2], - 'optimizer/weight_abs_max': opt_stats_2[3], + "optimizer/learning_rate": learning_rate, + "optimizer/iteration": args.iteration, + "optimizer/consumed_train_tokens": args.consumed_train_tokens, + "optimizer/variance_l2": opt_stats[0] ** 0.5, + "optimizer/variance_sqrt_l2": opt_stats[1] ** 0.5, + "optimizer/momentum_l2": opt_stats[2] ** 0.5, + "optimizer/weight_l2": opt_stats[3] ** 0.5, + "optimizer/variance_l1": opt_stats[4], + "optimizer/variance_sqrt_l1": opt_stats[5], + "optimizer/momentum_l1": opt_stats[6], + "optimizer/weight_l1": opt_stats[7], + "optimizer/variance_abs_max": opt_stats_2[0], + "optimizer/variance_sqrt_abs_max": opt_stats_2[1], + "optimizer/momentum_abs_max": opt_stats_2[2], + "optimizer/weight_abs_max": opt_stats_2[3], } # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if writer and is_last_rank(): - writer.add_scalar('optimizer/variance_l2 vs tokens', opt_stats[0]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_l2 vs tokens', opt_stats[1]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_l2 vs tokens', opt_stats[2]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_l2 vs tokens', opt_stats[3]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_l1 vs tokens', opt_stats[4], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_l1 vs tokens', opt_stats[5], args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_l1 vs tokens', opt_stats[6], args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_l1 vs tokens', opt_stats[7], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_abs_max vs tokens', opt_stats_2[0], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_abs_max vs tokens', opt_stats_2[1], args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_abs_max vs tokens', opt_stats_2[2], args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_abs_max vs tokens', opt_stats_2[3], args.consumed_train_tokens) - - writer.add_scalar('optimizer/variance_l2', opt_stats[0]**0.5, iteration) - writer.add_scalar('optimizer/variance_sqrt_l2', opt_stats[1]**0.5, iteration) - writer.add_scalar('optimizer/momentum_l2', opt_stats[2]**0.5, iteration) - writer.add_scalar('optimizer/weight_l2', opt_stats[3]**0.5, iteration) - writer.add_scalar('optimizer/variance_l1', opt_stats[4], iteration) - writer.add_scalar('optimizer/variance_sqrt_l1', opt_stats[5], iteration) - writer.add_scalar('optimizer/momentum_l1', opt_stats[6], iteration) - writer.add_scalar('optimizer/weight_l1', opt_stats[7], iteration) - writer.add_scalar('optimizer/variance_abs_max', opt_stats_2[0], iteration) - writer.add_scalar('optimizer/variance_sqrt_abs_max', opt_stats_2[1], iteration) - writer.add_scalar('optimizer/momentum_abs_max', opt_stats_2[2], iteration) - writer.add_scalar('optimizer/weight_abs_max', opt_stats_2[3], iteration) + writer.add_scalar( + "optimizer/variance_l2 vs tokens", + opt_stats[0] ** 0.5, + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/variance_sqrt_l2 vs tokens", + opt_stats[1] ** 0.5, + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/momentum_l2 vs tokens", + opt_stats[2] ** 0.5, + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/weight_l2 vs tokens", + opt_stats[3] ** 0.5, + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/variance_l1 vs tokens", + opt_stats[4], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/variance_sqrt_l1 vs tokens", + opt_stats[5], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/momentum_l1 vs tokens", + opt_stats[6], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/weight_l1 vs tokens", + opt_stats[7], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/variance_abs_max vs tokens", + opt_stats_2[0], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/variance_sqrt_abs_max vs tokens", + opt_stats_2[1], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/momentum_abs_max vs tokens", + opt_stats_2[2], + args.consumed_train_tokens, + ) + writer.add_scalar( + "optimizer/weight_abs_max vs tokens", + opt_stats_2[3], + args.consumed_train_tokens, + ) + + writer.add_scalar( + "optimizer/variance_l2", opt_stats[0] ** 0.5, iteration + ) + writer.add_scalar( + "optimizer/variance_sqrt_l2", opt_stats[1] ** 0.5, iteration + ) + writer.add_scalar( + "optimizer/momentum_l2", opt_stats[2] ** 0.5, iteration + ) + writer.add_scalar("optimizer/weight_l2", opt_stats[3] ** 0.5, iteration) + writer.add_scalar("optimizer/variance_l1", opt_stats[4], iteration) + writer.add_scalar("optimizer/variance_sqrt_l1", opt_stats[5], iteration) + writer.add_scalar("optimizer/momentum_l1", opt_stats[6], iteration) + writer.add_scalar("optimizer/weight_l1", opt_stats[7], iteration) + writer.add_scalar( + "optimizer/variance_abs_max", opt_stats_2[0], iteration + ) + writer.add_scalar( + "optimizer/variance_sqrt_abs_max", opt_stats_2[1], iteration + ) + writer.add_scalar( + "optimizer/momentum_abs_max", opt_stats_2[2], iteration + ) + writer.add_scalar("optimizer/weight_abs_max", opt_stats_2[3], iteration) assert args is not None + assert timers is not None if iteration % args.log_interval == 0: - elapsed_time = timers('interval-time').elapsed(barrier=True) + elapsed_time = timers("interval-time").elapsed(barrier=True) elapsed_time_per_iteration = elapsed_time / total_iterations seq_len = args.seq_length - if hasattr(args, 'actual_seq_length'): + if hasattr(args, "actual_seq_length"): seq_len = args.actual_seq_length samples_per_sec, tflops, approx_parameters_in_billions = throughput_calculator( - model, - args, - elapsed_time, - total_iterations + model, args, elapsed_time, total_iterations ) - num_flops = num_floating_point_operations(args, batch_size) - # throughput = ( - # num_floating_point_operations_so_far - arg - # ) samples_per_sec_per_replica = samples_per_sec / args.data_parallel_size tokens_per_sec = samples_per_sec * seq_len tokens_per_sec_per_replica = tokens_per_sec / args.data_parallel_size tokens_per_gpu_per_second = tokens_per_sec / args.world_size - tokens_per_gpu_per_second_per_replica = tokens_per_gpu_per_second / args.data_parallel_size + tokens_per_gpu_per_second_per_replica = ( + tokens_per_gpu_per_second / args.data_parallel_size + ) + # NOTE: [2024-06-19] + # Updated to use (more accurate) calculation according to + # `num_floating_point_operations` from NVIDIA/Megatron-LM + num_flop_lm = num_floating_point_operations(args, batch_size) + num_flop_per_sec_lm = (num_flop_lm / elapsed_time_per_iteration) + tflops_lm = (num_flop_per_sec_lm / (10 ** 12)) + tflops_lm_per_gpu = (tflops_lm / args.world_size) wandb_metrics |= { - 'throughput/iteration-time': elapsed_time_per_iteration, # 1000 ms / s - 'throughput/samples_per_sec': samples_per_sec, - 'throughput/samples_per_sec_per_replica': samples_per_sec_per_replica, - 'throughput/tokens_per_sec': tokens_per_sec, - 'throughput/tokens_per_sec_per_replica': tokens_per_sec_per_replica, - 'throughput/tokens_per_gpu_per_sec': tokens_per_gpu_per_second, - 'throughput/tokens_per_gpu_per_sec_per_replica': tokens_per_gpu_per_second_per_replica, - 'throughput/tflops': tflops, - 'throughput/flops': num_flops, - 'throughput/tflops-new': num_flops / elapsed_time_per_iteration, - 'throughput/approx_params_in_billions': approx_parameters_in_billions, - 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, - 'throughput/iteration': iteration, + "throughput/iteration-time": elapsed_time_per_iteration, # 1000 ms / s + "throughput/samples_per_sec": samples_per_sec, + "throughput/samples_per_sec_per_replica": samples_per_sec_per_replica, + "throughput/tokens_per_sec": tokens_per_sec, + "throughput/tokens_per_sec_per_replica": tokens_per_sec_per_replica, + "throughput/tokens_per_gpu_per_sec": tokens_per_gpu_per_second, + "throughput/tokens_per_gpu_per_sec_per_replica": tokens_per_gpu_per_second_per_replica, + "throughput/tflops": tflops, + "throughput/tflops-new": num_flop_lm / elapsed_time_per_iteration, + "throughput/tflops-lm": tflops_lm_per_gpu, + "throughput/approx_params_in_billions": approx_parameters_in_billions, + "throughput/elapsed_ms_per_iteration": elapsed_time_per_iteration, + "throughput/iteration": iteration, } if loss_dict is not None: wandb_metrics |= { - 'loss/iteration': iteration, - **{f'loss/{k}': v for k, v in loss_dict.items()} + "loss/iteration": iteration, + **{f"loss/{k}": v for k, v in loss_dict.items()}, } if writer and args.log_timers_to_tensorboard: - writer.add_scalar('iteration-time/iteration-time', - elapsed_time_per_iteration, iteration) - writer.add_scalar('iteration-time/iteration-time vs samples', - elapsed_time_per_iteration, args.consumed_train_samples) - writer.add_scalar('iteration-time/iteration-time vs tokens', - elapsed_time_per_iteration, args.consumed_train_tokens) - log_string = f' iteration={iteration:8d}/{args.train_iters:8d} |' + writer.add_scalar( + "iteration-time/iteration-time", elapsed_time_per_iteration, iteration + ) + writer.add_scalar( + "iteration-time/iteration-time vs samples", + elapsed_time_per_iteration, + args.consumed_train_samples, + ) + writer.add_scalar( + "iteration-time/iteration-time vs tokens", + elapsed_time_per_iteration, + args.consumed_train_tokens, + ) + # metrics_to_log = { + # 'iteration': iteration, + # 'train_iters': args.train_iters, + # 'consumed_samples': args.consumed_train_samples, + # 'consumed_tokens': args.consumed_tokens, + # } + log_string = f" iteration={iteration:8d}/{args.train_iters:8d} |" # .format( iteration, args.train_iters) log_string += ( - f' consumed_samples={args.consumed_train_samples:12d} |' - # .format(args.consumed_train_samples) + f" consumed_samples={args.consumed_train_samples:12d} |" + # .format(args.consumed_train_samples) ) - log_string += f' consumed_tokens={args.consumed_train_tokens:12d} |' + log_string += f" consumed_tokens={args.consumed_train_tokens:12d} |" # .format( args.consumed_train_tokens) log_string += ( - ' elapsed_time_per_iteration_ms=' - f'{elapsed_time_per_iteration * 1000.0:.1f} |' - # .format( elapsed_time_per_iteration * 1000.0) + " elapsed_time_per_iteration_ms=" + f"{elapsed_time_per_iteration * 1000.0:.1f} |" + # .format( elapsed_time_per_iteration * 1000.0) ) - log_string += f' learning_rate={learning_rate:.6f} |' - log_string += f' global_batch_size={batch_size:5d} |' + log_string += f" learning_rate={learning_rate:.6g} |" + log_string += f" global_batch_size={batch_size:5d} |" # if wandb is not None and getattr(wandb, 'run', None) is not None: wandb_metrics |= { - 'training/iteration': iteration, - 'training/iteration_time': elapsed_time_per_iteration, - 'training/iteration_time_vs_tokens': ( - (elapsed_time_per_iteration - / args.consumed_train_tokens) + "training/iteration": iteration, + "training/iteration_time": elapsed_time_per_iteration, + "training/iteration_time_vs_tokens": ( + elapsed_time_per_iteration / args.consumed_train_tokens ), - 'training/iteration_time_vs_samples': ( - (elapsed_time_per_iteration - / args.consumed_train_samples), + "training/iteration_time_vs_samples": ( + (elapsed_time_per_iteration / args.consumed_train_samples), ), - 'training/consumed_samples': args.consumed_train_samples, - 'training/consumed_tokens': args.consumed_train_tokens, + "training/consumed_samples": args.consumed_train_samples, + "training/consumed_tokens": args.consumed_train_tokens, } for key in total_loss_dict: - if key not in [advanced_iters_key, skipped_iters_key, - nan_iters_key]: - avg = total_loss_dict[key].item() / \ - float(max(1, total_loss_dict[advanced_iters_key])) + if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: + avg = total_loss_dict[key].item() / float( + max(1, total_loss_dict[advanced_iters_key]) + ) if avg > 0.0: - log_string += ' {}={:.6f} |'.format(key, avg) + log_string += " {}={:.6f} |".format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) if loss_scale is not None: - log_string += ' loss_scale={:.1f} |'.format(loss_scale) - wandb_metrics |= {'loss/loss_scale': loss_scale} + log_string += " loss_scale={:.1f} |".format(loss_scale) + wandb_metrics |= {"loss/loss_scale": loss_scale} if grad_norm is not None: - log_string += ' grad_norm={:.3f} |'.format(grad_norm) - wandb_metrics |= {'loss/grad_norm': grad_norm} + log_string += " grad_norm={:.3f} |".format(grad_norm) + wandb_metrics |= {"loss/grad_norm": grad_norm} if num_zeros_in_grad is not None: - log_string += ' num_zeros={:.1f} |'.format(num_zeros_in_grad) - wandb_metrics |= {'loss/num_zeros_in_grad': num_zeros_in_grad} + log_string += " num_zeros={:.1f} |".format(num_zeros_in_grad) + wandb_metrics |= {"loss/num_zeros_in_grad": num_zeros_in_grad} if params_norm is not None: - log_string += ' params_norm={:.3f} |'.format(params_norm) - wandb_metrics |= {'loss/params_norm': params_norm} + log_string += " params_norm={:.3f} |".format(params_norm) + wandb_metrics |= {"loss/params_norm": params_norm} if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - log_string += ' curriculum_seqlen={:5d} |'.format(args.curriculum_seqlen) + log_string += " curriculum_seqlen={:5d} |".format(args.curriculum_seqlen) if args.random_ltd: - log_string += ' random_ltd reserved_length={:5d} |'.format(args.random_ltd_reserved_length) - log_string += ' actual_seqlen={:5d} |'.format(seq_len) - log_string += ' number_of_skipped_iterations={:3d} |'.format( - total_loss_dict[skipped_iters_key]) - log_string += ' number_of_nan_iterations={:3d} |'.format( - total_loss_dict[nan_iters_key]) - log_string += ' samples_per_second={:.3f} |'.format(samples_per_sec) - log_string += ' tokens_per_gpu_per_second_tgs={:.3f} |'.format(tokens_per_gpu_per_second) - log_string += ' TFLOPs={:.2f} |'.format(tflops) + log_string += " random_ltd reserved_length={:5d} |".format( + args.random_ltd_reserved_length + ) + # log_string += " | ".join([ + # f"{seq_len=:5d} ", + # f"{}" + # f"number_of_skipped_iterations={:3d}", + # + # ]) + log_string += " actual_seqlen={:5d} |".format(seq_len) + log_string += " number_of_skipped_iterations={:3d} |".format( + total_loss_dict[skipped_iters_key] + ) + log_string += " number_of_nan_iterations={:3d} |".format( + total_loss_dict[nan_iters_key] + ) + log_string += " samples_per_second={:.3f} |".format(samples_per_sec) + log_string += " tokens_per_gpu_per_second_tgs={:.3f} |".format( + tokens_per_gpu_per_second + ) + log_string += " [LM]-TFLOPs={:.2f} |".format(tflops_lm_per_gpu) + log_string += " [DS]-TFLOPs={:.2f} |".format(tflops) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 # print_rank_last(log_string) log.info(log_string) - if report_memory_flag and learning_rate > 0.: + if report_memory_flag and learning_rate > 0.0: # Report memory after optimizer state has been initialized. - report_memory('(after {} iterations)'.format(iteration)) + report_memory("(after {} iterations)".format(iteration)) report_memory_flag = False - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb_metrics |= {'training/skiped_iterations': total_loss_dict[skipped_iters_key]} - wandb_metrics |= {'training/nan_iterations': total_loss_dict[nan_iters_key]} + if wandb is not None and getattr(wandb, "run", None) is not None: + wandb_metrics |= { + "training/skiped_iterations": total_loss_dict[skipped_iters_key] + } + wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} wandb.log(wandb_metrics) if timers is not None: timers.log(timers_to_log, normalizer=args.log_interval) return report_memory_flag + @dlp.log def save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler): timers = get_timers() # Extra barrier is added to make sure # all ranks report the max time. # assert timers is not None - timers('save-checkpoint', log_level=0).start(barrier=True) + timers("save-checkpoint", log_level=0).start(barrier=True) save_checkpoint(iteration, model, optimizer, opt_param_scheduler) - timers('save-checkpoint').stop(barrier=True) - checkpoint_throughput_calculator(model, timers('save-checkpoint').elapsed(reset=False)) - timers.log(['save-checkpoint']) + timers("save-checkpoint").stop(barrier=True) + checkpoint_throughput_calculator( + model, timers("save-checkpoint").elapsed(reset=False) + ) + timers.log(["save-checkpoint"]) + @dlp.log -def train(forward_step_func, model, optimizer, opt_param_scheduler, - train_data_iterator, valid_data_iterator, - process_non_loss_data_func): +def train( + forward_step_func, + model, + optimizer, + opt_param_scheduler, + train_data_iterator, + valid_data_iterator, + process_non_loss_data_func, +): """Train the model function.""" args = get_args() timers = get_timers() - + assert args is not None + assert timers is not None # Write args to tensorboard write_args_to_tensorboard() - if args.random_ltd: # random-ltd requires different randomness on each rank import random random.seed(args.seed + torch.distributed.get_rank()) - # Turn on training mode which enables dropout. for model_module in model: model_module.train() - # Tracking loss. total_loss_dict = {} - # Iterations. iteration = args.iteration - # Translate args to core configuration config = core_transformer_config_from_args(args) if not args.deepspeed: config.grad_scale_func = optimizer.scale_loss config.timers = timers - - timers('interval-time', log_level=0).start(barrier=True) - print_datetime('before the start of training step') + timers("interval-time", log_level=0).start(barrier=True) + print_datetime("before the start of training step") report_memory_flag = True if args.random_ltd: assert model[0].random_ltd_enabled() - args.random_ltd_layer_num = model[0].random_ltd_scheduler.get_random_ltd_layer_num() - - while iteration < args.train_iters and (args.train_tokens is None or \ - args.consumed_train_tokens < args.train_tokens): + args.random_ltd_layer_num = model[ + 0 + ].random_ltd_scheduler.get_random_ltd_layer_num() + while iteration < args.train_iters and ( + args.train_tokens is None or args.consumed_train_tokens < args.train_tokens + ): update_num_microbatches(args.consumed_train_samples) if args.deepspeed: # inform deepspeed of any batch size changes - global_batch_size = mpu.get_data_parallel_world_size() * \ - args.micro_batch_size * \ - get_num_microbatches() + global_batch_size = ( + mpu.get_data_parallel_world_size() + * args.micro_batch_size + * get_num_microbatches() + ) model[0].set_train_batch_size(global_batch_size) - if args.curriculum_learning_legacy and not args.no_pipeline_parallel: - curriculum_seqlen = args.curriculum_scheduler.update_difficulty( \ - args.iteration + 1) + curriculum_seqlen = args.curriculum_scheduler.update_difficulty( + args.iteration + 1 + ) if iteration == 0 or curriculum_seqlen != args.curriculum_seqlen: if args.use_rotary_position_embeddings: update_rotary_pos_emb(curriculum_seqlen) args.curriculum_seqlen = curriculum_seqlen args.curr_iteration = iteration - if os.getenv("TORCH_PROFILER_ENABLE")=='2': + if os.getenv("TORCH_PROFILER_ENABLE") == "2": from torch.profiler import profile, record_function, ProfilerActivity try: - activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA, ProfilerActivity.XPU] - except: + activities = [ + ProfilerActivity.CPU, + ProfilerActivity.CUDA, + ProfilerActivity.XPU, + ] + except Exception: log.warning("TORCH PROFILER WARNING: XPU is not supported") activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA] with profile(activities=activities) as prof: - loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ - train_step(forward_step_func, - train_data_iterator, - model, - optimizer, - opt_param_scheduler, - config) + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( + forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config, + ) prof.export_chrome_trace( f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" ) else: - loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = \ - train_step(forward_step_func, - train_data_iterator, - model, - optimizer, - opt_param_scheduler, - config) - + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( + forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config, + ) iteration += 1 args.iteration = iteration - new_samples = mpu.get_data_parallel_world_size() * \ - args.micro_batch_size * \ - get_num_microbatches() + new_samples = ( + mpu.get_data_parallel_world_size() + * args.micro_batch_size + * get_num_microbatches() + ) args.consumed_train_samples += new_samples # This actual_seq_length is used for actual consumed tokens calculation, flops calculation, and logging. args.actual_seq_length = args.seq_length if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: args.actual_seq_length = args.curriculum_seqlen if args.random_ltd: - args.random_ltd_reserved_length = model[0].random_ltd_scheduler.get_current_seq() + args.random_ltd_reserved_length = model[ + 0 + ].random_ltd_scheduler.get_current_seq() if args.random_ltd_reserved_length < args.actual_seq_length: - args.actual_seq_length = (args.actual_seq_length * (args.num_layers - args.random_ltd_layer_num) + args.random_ltd_reserved_length * args.random_ltd_layer_num) // args.num_layers + args.actual_seq_length = ( + args.actual_seq_length + * (args.num_layers - args.random_ltd_layer_num) + + args.random_ltd_reserved_length * args.random_ltd_layer_num + ) // args.num_layers if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - if hasattr(args, 'data_efficiency_curriculum_learning_numel'): - act_mbsz = args.data_efficiency_curriculum_learning_numel / args.curriculum_seqlen + if hasattr(args, "data_efficiency_curriculum_learning_numel"): + act_mbsz = ( + args.data_efficiency_curriculum_learning_numel + / args.curriculum_seqlen + ) act_token = act_mbsz * args.actual_seq_length - args.consumed_train_tokens += mpu.get_data_parallel_world_size() * \ - get_num_microbatches() * act_token + args.consumed_train_tokens += ( + mpu.get_data_parallel_world_size() + * get_num_microbatches() + * act_token + ) else: args.consumed_train_tokens += new_samples * args.actual_seq_length else: args.consumed_train_tokens += new_samples * args.actual_seq_length - # Logging. if args.deepspeed: - if hasattr(model[0].optimizer, 'cur_scale'): + if hasattr(model[0].optimizer, "cur_scale"): loss_scale = model[0].optimizer.cur_scale else: loss_scale = None @@ -1408,78 +1781,87 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, params_norm = None if args.log_params_norm: params_norm = calc_params_l2_norm(model) - report_memory_flag = training_log(loss_dict, total_loss_dict, - optimizer.param_groups[0]['lr'], - iteration, loss_scale, - report_memory_flag, skipped_iter, - grad_norm, params_norm, num_zeros_in_grad, - model, optimizer) - + report_memory_flag = training_log( + loss_dict, + total_loss_dict, + optimizer.param_groups[0]["lr"], + iteration, + loss_scale, + report_memory_flag, + skipped_iter, + grad_norm, + params_norm, + num_zeros_in_grad, + model, + optimizer, + ) # Autoresume - if args.adlr_autoresume and \ - (iteration % args.adlr_autoresume_interval == 0): - check_adlr_autoresume_termination(iteration, model, optimizer, - opt_param_scheduler) - + if args.adlr_autoresume and (iteration % args.adlr_autoresume_interval == 0): + check_adlr_autoresume_termination( + iteration, model, optimizer, opt_param_scheduler + ) # Evaluation - if args.eval_interval and iteration % args.eval_interval == 0 and \ - args.do_valid: - prefix = 'iteration {}'.format(iteration) - evaluate_and_print_results(prefix, forward_step_func, - valid_data_iterator, model, - iteration, process_non_loss_data_func, - config, False) - + if args.eval_interval and iteration % args.eval_interval == 0 and args.do_valid: + prefix = "iteration {}".format(iteration) + evaluate_and_print_results( + prefix, + forward_step_func, + valid_data_iterator, + model, + iteration, + process_non_loss_data_func, + config, + False, + ) # Checkpointing saved_checkpoint = False if args.exit_signal_handler: signal_handler = get_signal_handler() if any(signal_handler.signals_received()): - save_checkpoint_and_time(iteration, model, optimizer, - opt_param_scheduler) - print_datetime('exiting program after receiving SIGTERM.') + save_checkpoint_and_time( + iteration, model, optimizer, opt_param_scheduler + ) + print_datetime("exiting program after receiving SIGTERM.") sys.exit() - - if args.save and args.save_interval and \ - iteration % args.save_interval == 0: - save_checkpoint_and_time(iteration, model, optimizer, - opt_param_scheduler) + if args.save and args.save_interval and iteration % args.save_interval == 0: + save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler) saved_checkpoint = True - # Exiting based on duration if args.exit_duration_in_mins: train_time = (time.time() - _TRAIN_START_TIME) / 60.0 done_cuda = get_accelerator().IntTensor( - [train_time > args.exit_duration_in_mins]) - torch.distributed.all_reduce( - done_cuda, op=torch.distributed.ReduceOp.MAX) + [train_time > args.exit_duration_in_mins] + ) + torch.distributed.all_reduce(done_cuda, op=torch.distributed.ReduceOp.MAX) done = done_cuda.item() if done: if not saved_checkpoint: - save_checkpoint_and_time(iteration, model, optimizer, - opt_param_scheduler) - print_datetime('exiting program after {} minutes'.format(train_time)) + save_checkpoint_and_time( + iteration, model, optimizer, opt_param_scheduler + ) + print_datetime("exiting program after {} minutes".format(train_time)) sys.exit() - # Exiting based on iterations if args.exit_interval and iteration % args.exit_interval == 0: if args.save and not saved_checkpoint: - save_checkpoint_and_time(iteration, model, optimizer, - opt_param_scheduler) + save_checkpoint_and_time( + iteration, model, optimizer, opt_param_scheduler + ) torch.distributed.barrier() - print_datetime('exiting program at iteration {}'.format(iteration)) + print_datetime("exiting program at iteration {}".format(iteration)) sys.exit() - - return iteration + @dlp.log -def evaluate(forward_step_func, - data_iterator, - model, - process_non_loss_data_func, - config, - verbose=False): +def evaluate( + forward_step_func, + data_iterator, + model, + process_non_loss_data_func, + config, + verbose=False, +): """Evaluation.""" args = get_args() @@ -1509,8 +1891,7 @@ def evaluate(forward_step_func, while iteration < args.eval_iters: iteration += 1 if verbose and iteration % args.log_interval == 0: - log.info('Evaluating iter {}/{}'.format(iteration, - args.eval_iters)) + log.info("Evaluating iter {}/{}".format(iteration, args.eval_iters)) forward_backward_func = get_forward_backward_func() # Don't care about timing during evaluation @@ -1519,7 +1900,7 @@ def evaluate(forward_step_func, # DeepSpeed uses eval_batch() and already aggregates losses. assert isinstance(model, list) and len(model) == 1 loss = model[0].eval_batch(data_iterator) - loss_dicts = [{'lm loss' : loss}] * get_num_microbatches() + loss_dicts = [{"lm loss": loss}] * get_num_microbatches() else: loss_dicts = forward_backward_func( forward_step_func=forward_step_func, @@ -1529,7 +1910,8 @@ def evaluate(forward_step_func, seq_length=args.seq_length, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, - forward_only=True) + forward_only=True, + ) config.timers = get_timers() # Empty unused memory @@ -1540,13 +1922,19 @@ def evaluate(forward_step_func, # Reduce across processes. for loss_dict in loss_dicts: for key in loss_dict: - if 'moe' not in key: - total_loss_dict[key] = total_loss_dict.get( - key, get_accelerator().FloatTensor([0.0])) + loss_dict[key] - - args.consumed_valid_samples += mpu.get_data_parallel_world_size() \ - * args.micro_batch_size \ - * get_num_microbatches() + if "moe" not in key: + total_loss_dict[key] = ( + total_loss_dict.get( + key, get_accelerator().FloatTensor([0.0]) + ) + + loss_dict[key] + ) + + args.consumed_valid_samples += ( + mpu.get_data_parallel_world_size() + * args.micro_batch_size + * get_num_microbatches() + ) collected_non_loss_data = None if process_non_loss_data_func is not None and is_last_rank(): collected_non_loss_data = forward_backward_func( @@ -1558,7 +1946,8 @@ def evaluate(forward_step_func, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, forward_only=True, - collect_non_loss_data=True) + collect_non_loss_data=True, + ) # Move model back to the train mode. for model_module in model: @@ -1569,8 +1958,9 @@ def evaluate(forward_step_func, if args.curriculum_learning_legacy and not args.no_pipeline_parallel: # roll back to actual curriculum seqlen at the end of eval. - args.curriculum_seqlen = args.curriculum_scheduler.update_difficulty( \ - args.iteration + 1) + args.curriculum_seqlen = args.curriculum_scheduler.update_difficulty( + args.iteration + 1 + ) if args.curriculum_seqlen < args.seq_length: if args.use_rotary_position_embeddings: update_rotary_pos_emb(args.curriculum_seqlen) @@ -1578,11 +1968,20 @@ def evaluate(forward_step_func, return total_loss_dict, collected_non_loss_data + @dlp.log -def evaluate_and_print_results(prefix, forward_step_func, - data_iterator, model, - iteration, process_non_loss_data_func, config, - verbose=False, write_to_tensorboard=True, test=False): +def evaluate_and_print_results( + prefix, + forward_step_func, + data_iterator, + model, + iteration, + process_non_loss_data_func, + config, + verbose=False, + write_to_tensorboard=True, + test=False, +): """Helper function to evaluate and dump results on screen.""" args = get_args() if write_to_tensorboard: @@ -1591,40 +1990,58 @@ def evaluate_and_print_results(prefix, forward_step_func, writer = None total_loss_dict, collected_non_loss_data = evaluate( - forward_step_func, data_iterator, model, - process_non_loss_data_func, config, verbose) - string = ' validation loss at {} | '.format(prefix) + forward_step_func, + data_iterator, + model, + process_non_loss_data_func, + config, + verbose, + ) + string = " validation loss at {} | ".format(prefix) for key in total_loss_dict: string += f"{key} value={total_loss_dict[key].item():.6f}" ppl = math.exp(min(20, total_loss_dict[key].item())) string += f"{key} PPL={ppl:.6f}" # string += '{} PPL={:.6f} | '.format(key, ppl) if writer and is_last_rank(): - data_type = 'test' if test else 'validation' - writer.add_scalar(f'lm-loss-validation/{key} {data_type}', - total_loss_dict[key].item(), - iteration) - writer.add_scalar(f'lm-loss-validation/{key} {data_type} vs samples', - total_loss_dict[key].item(), - args.consumed_train_samples) - writer.add_scalar(f'lm-loss-validation/{key} {data_type} vs tokens', - total_loss_dict[key].item(), - args.consumed_train_tokens) + data_type = "test" if test else "validation" + writer.add_scalar( + f"lm-loss-validation/{key} {data_type}", + total_loss_dict[key].item(), + iteration, + ) + writer.add_scalar( + f"lm-loss-validation/{key} {data_type} vs samples", + total_loss_dict[key].item(), + args.consumed_train_samples, + ) + writer.add_scalar( + f"lm-loss-validation/{key} {data_type} vs tokens", + total_loss_dict[key].item(), + args.consumed_train_tokens, + ) if args.log_validation_ppl_to_tensorboard: - writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl', ppl, - iteration) - writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl vs samples', - ppl, args.consumed_train_samples) - writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl vs tokens', - ppl, args.consumed_train_tokens) + writer.add_scalar( + f"lm-loss-validation/{key} {data_type} ppl", ppl, iteration + ) + writer.add_scalar( + f"lm-loss-validation/{key} {data_type} ppl vs samples", + ppl, + args.consumed_train_samples, + ) + writer.add_scalar( + f"lm-loss-validation/{key} {data_type} ppl vs tokens", + ppl, + args.consumed_train_tokens, + ) if process_non_loss_data_func is not None and writer and is_last_rank(): process_non_loss_data_func(collected_non_loss_data, iteration, writer) length = len(string) + 1 - log.info('-' * length) + log.info("-" * length) log.info(string) - log.info('-' * length) + log.info("-" * length) def cyclic_iter(iter): @@ -1632,6 +2049,7 @@ def cyclic_iter(iter): for x in iter: yield x + @dlp.log def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): """Build pretraining datasets.""" @@ -1643,54 +2061,68 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): train_samples = args.train_samples else: train_samples = args.train_iters * args.global_batch_size - eval_iters = (args.train_iters // args.eval_interval + 1) * \ - args.eval_iters + eval_iters = (args.train_iters // args.eval_interval + 1) * args.eval_iters test_iters = args.eval_iters - train_val_test_num_samples = [train_samples, - eval_iters * args.global_batch_size, - test_iters * args.global_batch_size] - log.info(' > datasets target sizes (minimum size):') - log.info(' train: {}'.format(train_val_test_num_samples[0])) - log.info(' validation: {}'.format(train_val_test_num_samples[1])) - log.info(' test: {}'.format(train_val_test_num_samples[2])) + train_val_test_num_samples = [ + train_samples, + eval_iters * args.global_batch_size, + test_iters * args.global_batch_size, + ] + log.info(" > datasets target sizes (minimum size):") + log.info(" train: {}".format(train_val_test_num_samples[0])) + log.info(" validation: {}".format(train_val_test_num_samples[1])) + log.info(" test: {}".format(train_val_test_num_samples[2])) # Build the datasets. return build_train_valid_test_datasets_provider(train_val_test_num_samples) + @dlp.log -def build_train_valid_test_data_loaders( - build_train_valid_test_datasets_provider): +def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider): """Build pretraining data loaders.""" args = get_args() (train_dataloader, valid_dataloader, test_dataloader) = (None, None, None) - log.info('> building train, validation, and test datasets ...') + log.info("> building train, validation, and test datasets ...") # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: - assert args.train_samples is None, \ - 'only backward compatiblity support for iteration-based training' + assert ( + args.train_samples is None + ), "only backward compatiblity support for iteration-based training" args.consumed_train_samples = args.iteration * args.global_batch_size if args.iteration > 0 and args.consumed_valid_samples == 0: if args.train_samples is None: - args.consumed_valid_samples = (args.iteration // args.eval_interval) * \ - args.eval_iters * args.global_batch_size + args.consumed_valid_samples = ( + (args.iteration // args.eval_interval) + * args.eval_iters + * args.global_batch_size + ) # Data loader only on rank 0 of each model parallel group. - ds_sequence_parallel = mpu.get_sequence_parallel_world_size() > 1 or args.force_ds_sequence_parallel - rank_in_parallel_group = mpu.get_sequence_parallel_rank() if ds_sequence_parallel else mpu.get_tensor_model_parallel_rank() + ds_sequence_parallel = ( + mpu.get_sequence_parallel_world_size() > 1 or args.force_ds_sequence_parallel + ) + rank_in_parallel_group = ( + mpu.get_sequence_parallel_rank() + if ds_sequence_parallel + else mpu.get_tensor_model_parallel_rank() + ) if rank_in_parallel_group == 0: # Build datasets. train_ds, valid_ds, test_ds = build_train_valid_test_datasets( - build_train_valid_test_datasets_provider) + build_train_valid_test_datasets_provider + ) # Build dataloders. train_dataloader = build_pretraining_data_loader( - train_ds, args.consumed_train_samples) + train_ds, args.consumed_train_samples + ) valid_dataloader = build_pretraining_data_loader( - valid_ds, args.consumed_valid_samples) + valid_ds, args.consumed_valid_samples + ) test_dataloader = build_pretraining_data_loader(test_ds, 0) # Flags to know if we need to do training/validation/testing. @@ -1699,56 +2131,70 @@ def build_train_valid_test_data_loaders( do_test = test_dataloader is not None and args.eval_iters > 0 # Need to broadcast num_tokens and num_type_tokens. flags = get_accelerator().LongTensor( - [int(do_train), int(do_valid), int(do_test)]) + [int(do_train), int(do_valid), int(do_test)] + ) else: flags = get_accelerator().LongTensor([0, 0, 0]) # Broadcast num tokens. if ds_sequence_parallel: - torch.distributed.broadcast(flags, - mpu.get_sequence_parallel_src_rank(), - group=mpu.get_sequence_parallel_group()) + torch.distributed.broadcast( + flags, + mpu.get_sequence_parallel_src_rank(), + group=mpu.get_sequence_parallel_group(), + ) else: - torch.distributed.broadcast(flags, - mpu.get_tensor_model_parallel_src_rank(), - group=mpu.get_tensor_model_parallel_group()) + torch.distributed.broadcast( + flags, + mpu.get_tensor_model_parallel_src_rank(), + group=mpu.get_tensor_model_parallel_group(), + ) args.do_train = flags[0].item() args.do_valid = flags[1].item() args.do_test = flags[2].item() return train_dataloader, valid_dataloader, test_dataloader + @dlp.log -def build_train_valid_test_data_iterators( - build_train_valid_test_datasets_provider): +def build_train_valid_test_data_iterators(build_train_valid_test_datasets_provider): """Build pretraining data iterators.""" args = get_args() # Build loaders. - train_dataloader, valid_dataloader, test_dataloader = \ - build_train_valid_test_data_loaders( - build_train_valid_test_datasets_provider) + train_dataloader, valid_dataloader, test_dataloader = ( + build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider) + ) # Build iterators. dl_type = args.dataloader_type - assert dl_type in ['single', 'cyclic'] + assert dl_type in ["single", "cyclic"] if train_dataloader is not None: - train_data_iterator = iter(train_dataloader) if dl_type == 'single' \ - else iter(cyclic_iter(train_dataloader)) + train_data_iterator = ( + iter(train_dataloader) + if dl_type == "single" + else iter(cyclic_iter(train_dataloader)) + ) else: train_data_iterator = None if valid_dataloader is not None: - valid_data_iterator = iter(valid_dataloader) if dl_type == 'single' \ - else iter(cyclic_iter(valid_dataloader)) + valid_data_iterator = ( + iter(valid_dataloader) + if dl_type == "single" + else iter(cyclic_iter(valid_dataloader)) + ) else: valid_data_iterator = None if test_dataloader is not None: - test_data_iterator = iter(test_dataloader) if dl_type == 'single' \ - else iter(cyclic_iter(test_dataloader)) + test_data_iterator = ( + iter(test_dataloader) + if dl_type == "single" + else iter(cyclic_iter(test_dataloader)) + ) else: test_data_iterator = None diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 6e3c0c3e6fb..bf1c9eef194 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -91,7 +91,16 @@ def model_provider(pre_process=True, post_process=True): log.info('building GPT model ...') see_memory_usage("Before Building Model", force=True) args = get_args() + assert args is not None config = core_transformer_config_from_args(args) + # from ezpz.profile import get_context_manager + # cm = get_context_manager(rank=RANK, outdir=args.save) + # profiler = None + # if os.environ.get("PYINSTRUMENT_PROFILER", None) is not None: + # from ezpz.profile import PyInstrumentProfiler + # cm = PyInstrumentProfiler(rank=RANK, outdir=args.save) + # else: + # from contextlib # if wandb.run is not None and RANK == 0: # print(f"Updating WandB run: [{wandb.run.name}]({wandb.run.url})") # try: @@ -112,7 +121,6 @@ def model_provider(pre_process=True, post_process=True): dpg = mpu.get_data_parallel_group() else: dpg = None - deepspeed_zero_init = deepspeed.zero.Init if args.use_mics: deepspeed_zero_init = deepspeed.zero.MiCS_Init @@ -196,14 +204,14 @@ def model_provider(pre_process=True, post_process=True): log.error( 'Unable to `wandb.run.config.update({"args": vars(args)})`' ) - try: - wandb.run.watch( - model, - log='all', - log_graph=True, - ) - except Exception: - pass + # try: + # wandb.run.watch( + # model, + # log='all', + # log_graph=True, + # ) + # except Exception: + # pass return model @@ -430,14 +438,14 @@ def forward_step(data_iterator, model): """Forward step.""" args = get_args() timers = get_timers() - + assert args is not None + assert timers is not None # Get the batch. timers('batch-generator', log_level=2).start() tokens, labels, loss_mask, attention_mask, position_ids = get_batch( data_iterator ) timers('batch-generator').stop() - if args.data_efficiency_curriculum_learning: args.curriculum_seqlen = tokens.size()[1] if ( @@ -452,7 +460,7 @@ def forward_step(data_iterator, model): args.data_efficiency_curriculum_learning_numel = ( torch.numel(tokens) ) - + stu_output = None if args.mos or args.kd: # The forward func can return either the loss or the logits, # depending on whether passing in the labels or not. @@ -509,6 +517,9 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): t0 = time.perf_counter() args = get_args() assert args is not None + # from ezpz.profile import get_context_manager + # cm = get_context_manager(rank=RANK, outdir=args.save) + # with cm: log.info('> building train, validation, and test datasets for GPT ...') files = [] if args.data_file_list is not None: @@ -527,8 +538,12 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): if len(f.strip()) != 0: try: w, fname, c = f.split() - except: - raise Exception("Please provide the file list as 'weight, filename, corpus'") + except Exception as exc: + log.exception(exc) + raise Exception( + "Please provide the file list as " + "'weight, filename, corpus'" + ) if fname.find(".bin") != -1: fname = fname.split(".bin")[0] files.extend( @@ -603,10 +618,12 @@ def git_ds_info(): def main(): if os.getenv('TORCH_PROFILER_ENABLE') == '1': - from torch.profiler import profile, record_function, ProfilerActivity + # record_function + from torch.profiler import profile, ProfilerActivity try: activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA, ProfilerActivity.XPU] - except: + except Exception as exc: + log.exception(exc) log.warning("TORCH PROFILER WARNING: XPU is not supported") activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA] with profile(activities=activities) as prof: @@ -619,6 +636,7 @@ def main(): data_post_process=data_post_process ) args = get_args() + assert args is not None prof.export_chrome_trace( f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}.json" ) From c7a1e36e797cdfb0ebb6cc3e9eca4c010a2fa315 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 09:19:08 -0500 Subject: [PATCH 237/590] Add `ALCF/notes/deepspeed_init_time.md` --- ALCF/notes/deepspeed_init_time.md | 251 ++++++++++++++++++++++++++++++ 1 file changed, 251 insertions(+) create mode 100644 ALCF/notes/deepspeed_init_time.md diff --git a/ALCF/notes/deepspeed_init_time.md b/ALCF/notes/deepspeed_init_time.md new file mode 100644 index 00000000000..13ed2cf3f92 --- /dev/null +++ b/ALCF/notes/deepspeed_init_time.md @@ -0,0 +1,251 @@ +# DeepSpeed Initialization Time on Aurora + +## TODOs + +- [ ] Use `ZeRO={1, 2}` @ 256 Nodes of Aurora +- [ ] Figure out bottleneck in startup time on Aurora +- [ ] Use GAS=8 on Aurora +- [ ] Weight decay too high +- [ ] Save checkpoints every ~ 1 hr +- [ ] Write weekly updates and post to GitHub + +## Initialization Times + +- Search for "deepspeed.initialize" in `Megatron-DeepSpeed/logs/`: + +```bash +#[🌌][11:44:57 PM][foremans@aurora-uan-0010][…/Megatron-DeepSpeed/logs][🌱 alcf-startup-time][$!?] +$ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took +``` + +### Measurements + +| WORLD_SIZE | TIME | +| :--------: | :--------: | +| 96 | 61.073 | +| 192 | 107.74411 | +| 192 | 107.201338 | +| 192 | 107.10853 | +| 384 | 200.23095 | +| 384 | 206.49485 | +| 384 | 200.49485 | +| 768 | 413.55765 | +| 768 | 394.92617 | +| 768 | 414.725 | +| 768 | 387.987 | +| 768 | 411.72035 | +| 768 | 394.926 | +| 768 | 409.375 | +| 768 | 393.091 | +| 768 | 412.600 | +| 1536 | 789.30077 | +| 1536 | 788.86531 | +| 1536 | 792.71864 | +| 1536 | 836.98 | +| 1536 | 801.205 | +| 1536 | 836.98 | +| 1536 | 820.9538 | +| 1536 | 707.048 | +| 3072 | 1639.62374 | +| 3072 | 1591.345 | +| 3072 | 1632.12712 | +| 3072 | 1674.444 | +| 3072 | 1618.100 | + +- `WORLD_SIZE=96`: + + ```bash title="deepspeed_init_times.sh" + ws96_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125717_96_x4420c5s5b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:59:19][INFO][training:795] - 'deepspeed.initialize' took: 61.07362s + ws96_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125717_96_x4420c5s5b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:59:19][INFO][training:795] - 'deepspeed.initialize' took: 61.07362s + ws96_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125717_96_x4420c5s5b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:59:19][INFO][training:795] - 'deepspeed.initialize' took: 61.07362s + ``` + +- `WORLD_SIZE = 192`: + + ```bash + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-154948_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:52:30][INFO][training:795] - 'deepspeed.initialize' took: 107.74411s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-154948_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:52:30][INFO][training:795] - 'deepspeed.initialize' took: 107.74411s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-154948_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:52:30][INFO][training:795] - 'deepspeed.initialize' took: 107.74411s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamwschedulefree_lr0.0003_lwf0.05/20240623-163640_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:38:52][INFO][training:800] - 'deepspeed.initialize' took: 107.10853s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-160332_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:05:43][INFO][training:800] - 'deepspeed.initialize' took: 107.20138s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-160332_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:05:43][INFO][training:800] - 'deepspeed.initialize' took: 107.20138s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-160332_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:05:43][INFO][training:800] - 'deepspeed.initialize' took: 107.20138s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamwschedulefree_lr0.0003_lwf0.05/20240623-163640_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:38:52][INFO][training:800] - 'deepspeed.initialize' took: 107.10853s + ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamwschedulefree_lr0.0003_lwf0.05/20240623-163640_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:38:52][INFO][training:800] - 'deepspeed.initialize' took: 107.10853s + ``` + +- `WORLD_SIZE = 384`: + + ```bash + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-164607_384_x4402c6s7b0n0.hostmgmt2402.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:52:15][INFO][training:800] - 'deepspeed.initialize' took: 206.49485s + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-164607_384_x4402c6s7b0n0.hostmgmt2402.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:52:15][INFO][training:800] - 'deepspeed.initialize' took: 206.49485s + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-164607_384_x4402c6s7b0n0.hostmgmt2402.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:52:15][INFO][training:800] - 'deepspeed.initialize' took: 206.49485s + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-223159_384_x4706c1s6b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:37:53][INFO][training:800] - 'deepspeed.initialize' took: 200.23095s + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-223159_384_x4706c1s6b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:37:53][INFO][training:800] - 'deepspeed.initialize' took: 200.23095s + ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-223159_384_x4706c1s6b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:37:53][INFO][training:800] - 'deepspeed.initialize' took: 200.23095s + ``` + +- `WORLD_SIZE=768`: + + ```bash + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-180052_768_x4704c4s1b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 18:12:43][INFO][training:800] - 'deepspeed.initialize' took: 394.92617s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-185626_768_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 19:05:45][INFO][training:800] - 'deepspeed.initialize' took: 414.72580s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-185626_768_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 19:05:45][INFO][training:800] - 'deepspeed.initialize' took: 414.72580s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-233045_768_x4711c0s1b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 23:39:19][INFO][training:797] - 'deepspeed.initialize' took: 387.98744s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-233045_768_x4711c0s1b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 23:39:19][INFO][training:797] - 'deepspeed.initialize' took: 387.98744s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-233045_768_x4711c0s1b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 23:39:19][INFO][training:797] - 'deepspeed.initialize' took: 387.98744s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-141802_768_x4706c2s0b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:27:50][INFO][training:795] - 'deepspeed.initialize' took: 411.72035s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-141802_768_x4706c2s0b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:27:50][INFO][training:795] - 'deepspeed.initialize' took: 411.72035s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-180052_768_x4704c4s1b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 18:12:43][INFO][training:800] - 'deepspeed.initialize' took: 394.92617s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-180052_768_x4704c4s1b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 18:12:43][INFO][training:800] - 'deepspeed.initialize' took: 394.92617s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-134324_768_x4705c2s1b0n0.hostmgmt2705.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:51:19][INFO][training:795] - 'deepspeed.initialize' took: 393.09134s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-185626_768_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 19:05:45][INFO][training:800] - 'deepspeed.initialize' took: 414.72580s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-165713_768_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:06:47][INFO][training:800] - 'deepspeed.initialize' took: 389.15768s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122601_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:35:18][INFO][training:793] - 'deepspeed.initialize' took: 409.37578s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-134324_768_x4705c2s1b0n0.hostmgmt2705.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:51:19][INFO][training:795] - 'deepspeed.initialize' took: 393.09134s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-134324_768_x4705c2s1b0n0.hostmgmt2705.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:51:19][INFO][training:795] - 'deepspeed.initialize' took: 393.09134s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-141802_768_x4706c2s0b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:27:50][INFO][training:795] - 'deepspeed.initialize' took: 411.72035s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-165713_768_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:06:47][INFO][training:800] - 'deepspeed.initialize' took: 389.15768s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124517_768_x4315c4s1b0n0.hostmgmt2315.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:57:42][INFO][training:795] - 'deepspeed.initialize' took: 395.05079s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124517_768_x4315c4s1b0n0.hostmgmt2315.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:57:42][INFO][training:795] - 'deepspeed.initialize' took: 395.05079s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-165713_768_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:06:47][INFO][training:800] - 'deepspeed.initialize' took: 389.15768s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-130702_768_x4420c6s7b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:15:22][INFO][training:795] - 'deepspeed.initialize' took: 412.60004s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-130702_768_x4420c6s7b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:15:22][INFO][training:795] - 'deepspeed.initialize' took: 412.60004s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122601_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:35:18][INFO][training:793] - 'deepspeed.initialize took: 409.37578s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122601_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:35:18][INFO][training:793] - 'deepspeed.initialize took: 409.37578s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-173730_768_x4707c5s6b0n0.hostmgmt2707.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:45:33][INFO][training:800] - 'deepspeed.initialize' took: 400.74402s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124517_768_x4315c4s1b0n0.hostmgmt2315.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:57:42][INFO][training:795] - 'deepspeed.initialize' took: 395.05079s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-130702_768_x4420c6s7b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:15:22][INFO][training:795] - 'deepspeed.initialize' took: 412.60004s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-132452_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:34:32][INFO][training:795] - 'deepspeed.initialize' took: 413.55765s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-173730_768_x4707c5s6b0n0.hostmgmt2707.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:45:33][INFO][training:800] - 'deepspeed.initialize' took: 400.74402s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-173730_768_x4707c5s6b0n0.hostmgmt2707.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:45:33][INFO][training:800] - 'deepspeed.initialize' took: 400.74402s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-132452_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:34:32][INFO][training:795] - 'deepspeed.initialize' took: 413.55765s + ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-132452_768_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:34:32][INFO][training:795] - 'deepspeed.initialize' took: 413.55765s + ``` + +- `WORLD_SIZE = 1536`: + + ```bash + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-162028_1536_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:37:49][INFO][training:800] - 'deepspeed.initialize' took: 789.30077s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-162028_1536_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:37:49][INFO][training:800] - 'deepspeed.initialize' took: 789.30077s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-162028_1536_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:37:49][INFO][training:800] - 'deepspeed.initialize' took: 789.30077s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-145656_1536_x4119c5s7b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:14:35][INFO][training:795] - 'deepspeed.initialize' took: 788.86531s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-145656_1536_x4119c5s7b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:14:35][INFO][training:795] - 'deepspeed.initialize' took: 788.86531s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-145656_1536_x4119c5s7b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:14:35][INFO][training:795] - 'deepspeed.initialize' took: 788.86531s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122207_1536_x4309c6s4b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:39:56][INFO][training:793] - 'deepspeed.initialize' took: 792.71864s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122207_1536_x4309c6s4b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:39:56][INFO][training:793] - 'deepspeed.initialize' took: 792.71864s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-122207_1536_x4309c6s4b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 12:39:56][INFO][training:793] - 'deepspeed.initialize' took: 792.71864s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125001_1536_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:06:47][INFO][training:795] - 'deepspeed.initialize' took: 836.98388s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125001_1536_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:06:47][INFO][training:795] - 'deepspeed.initialize' took: 836.98388s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-175213_1536_x4702c1s4b0n0.hostmgmt2702.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 18:10:54][INFO][training:800] - 'deepspeed.initialize' took: 801.20500s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-175213_1536_x4702c1s4b0n0.hostmgmt2702.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 18:10:54][INFO][training:800] - 'deepspeed.initialize' took: 801.20500s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-184503_1536_x4702c1s4b0n0.hostmgmt2702.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 19:04:07][INFO][training:800] - 'deepspeed.initialize' took: 801.15950s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-131641_1536_x4315c4s1b0n0.hostmgmt2315.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:33:00][INFO][training:795] - 'deepspeed.initialize' took: 801.11322s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-213107_1536_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 21:46:29][INFO][training:800] - 'deepspeed.initialize' took: 820.95380s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-155216_1536_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 16:07:23][INFO][training:795] - 'deepspeed.initialize' took: 787.04806s + ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-141727_1536_x4102c7s0b0n0.hostmgmt2102.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:34:20][INFO][training:795] - 'deepspeed.initialize' took: 809.36787s + ``` + +- `WORLD_SIZE = 3072`: + + ```bash + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-144534_3072_x4309c6s2b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:18:41][INFO][training:795] - 'deepspeed.initialize' took: 1639.62374s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-144534_3072_x4309c6s2b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:18:41][INFO][training:795] - 'deepspeed.initialize' took: 1639.62374s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-144534_3072_x4309c6s2b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 15:18:41][INFO][training:795] - 'deepspeed.initialize' took: 1639.62374s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-213304_3072_x4704c0s6b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:03:15][INFO][training:800] - 'deepspeed.initialize' took: 1591.34487s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-213304_3072_x4704c0s6b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:03:15][INFO][training:800] - 'deepspeed.initialize' took: 1591.34487s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-213304_3072_x4704c0s6b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 22:03:15][INFO][training:800] - 'deepspeed.initialize' took: 1591.34487s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-170636_3072_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:37:20][INFO][training:800] - 'deepspeed.initialize' took: 1632.12712s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-170636_3072_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:37:20][INFO][training:800] - 'deepspeed.initialize' took: 1632.12712s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-170636_3072_x4415c2s3b0n0.hostmgmt2415.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 17:37:20][INFO][training:800] - 'deepspeed.initialize' took: 1632.12712s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124519_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:16:22][INFO][training:795] - 'deepspeed.initialize' took: 1674.44393s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124519_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:16:22][INFO][training:795] - 'deepspeed.initialize' took: 1674.44393s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-124519_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 13:16:22][INFO][training:795] - 'deepspeed.initialize' took: 1674.44393s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-140113_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:30:23][INFO][training:795] - 'deepspeed.initialize' took: 1618.10035s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-140113_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:30:23][INFO][training:795] - 'deepspeed.initialize' took: 1618.10035s + ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-140113_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: + [2024-06-23 14:30:23][INFO][training:795] - 'deepspeed.initialize' took: 1618.10035s + ``` From 0548bfbe4382bbd3cd00577947615b58f98a06fe Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 09:32:22 -0500 Subject: [PATCH 238/590] Update `ALCF/notes/deepspeed_init_time.md` --- ALCF/notes/deepspeed_init_time.md | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/ALCF/notes/deepspeed_init_time.md b/ALCF/notes/deepspeed_init_time.md index 13ed2cf3f92..c1050772e1b 100644 --- a/ALCF/notes/deepspeed_init_time.md +++ b/ALCF/notes/deepspeed_init_time.md @@ -52,7 +52,7 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took | 3072 | 1674.444 | | 3072 | 1618.100 | -- `WORLD_SIZE=96`: +-
WORLD_SIZE=96: ```bash title="deepspeed_init_times.sh" ws96_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-125717_96_x4420c5s5b0n0.hostmgmt2420.cm.aurora.alcf.anl.gov/output.log: @@ -63,7 +63,9 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took [2024-06-23 12:59:19][INFO][training:795] - 'deepspeed.initialize' took: 61.07362s ``` -- `WORLD_SIZE = 192`: +
+ +-
WORLD_SIZE = 192: ```bash ws192_ds_stage1_nl32_hs4096_mb4_seq4096_gb6144_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-154948_192_x4716c2s6b0n0.hostmgmt2716.cm.aurora.alcf.anl.gov/output.log: @@ -86,7 +88,9 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took [2024-06-23 16:38:52][INFO][training:800] - 'deepspeed.initialize' took: 107.10853s ``` -- `WORLD_SIZE = 384`: +
+ +-
WORLD_SIZE = 384: ```bash ws384_ds_stage1_nl32_hs4096_mb4_seq4096_gb12288_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-164607_384_x4402c6s7b0n0.hostmgmt2402.cm.aurora.alcf.anl.gov/output.log: @@ -103,7 +107,9 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took [2024-06-23 22:37:53][INFO][training:800] - 'deepspeed.initialize' took: 200.23095s ``` -- `WORLD_SIZE=768`: +
+ +-
WORLD_SIZE=768: ```bash ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb24576_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-180052_768_x4704c4s1b0n0.hostmgmt2704.cm.aurora.alcf.anl.gov/output.log: @@ -174,7 +180,9 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took [2024-06-23 13:34:32][INFO][training:795] - 'deepspeed.initialize' took: 413.55765s ``` -- `WORLD_SIZE = 1536`: +
+ +-
WORLD_SIZE = 1536: ```bash ws1536_ds_stage1_nl32_hs4096_mb4_seq4096_gb49152_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-162028_1536_x4706c2s3b0n0.hostmgmt2706.cm.aurora.alcf.anl.gov/output.log: @@ -215,7 +223,9 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took [2024-06-23 14:34:20][INFO][training:795] - 'deepspeed.initialize' took: 809.36787s ``` -- `WORLD_SIZE = 3072`: +
+ +-
WORLD_SIZE = 3072: ```bash ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-144534_3072_x4309c6s2b0n0.hostmgmt2309.cm.aurora.alcf.anl.gov/output.log: @@ -249,3 +259,5 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took ws3072_ds_stage1_nl32_hs4096_mb4_seq4096_gb98304_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240623-140113_3072_x4119c5s3b0n0.hostmgmt2119.cm.aurora.alcf.anl.gov/output.log: [2024-06-23 14:30:23][INFO][training:795] - 'deepspeed.initialize' took: 1618.10035s ``` + +
From 6a8f55c616f36dc51f9db766c61680f93068bc30 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 15:11:55 -0500 Subject: [PATCH 239/590] Update deepspeed_init_time.md --- ALCF/notes/deepspeed_init_time.md | 68 +++++++++++++++++-------------- 1 file changed, 37 insertions(+), 31 deletions(-) diff --git a/ALCF/notes/deepspeed_init_time.md b/ALCF/notes/deepspeed_init_time.md index c1050772e1b..a355a082a9b 100644 --- a/ALCF/notes/deepspeed_init_time.md +++ b/ALCF/notes/deepspeed_init_time.md @@ -20,37 +20,43 @@ $ rg --hidden "deepspeed\.initialize" **/**/*.log | grep took ### Measurements -| WORLD_SIZE | TIME | -| :--------: | :--------: | -| 96 | 61.073 | -| 192 | 107.74411 | -| 192 | 107.201338 | -| 192 | 107.10853 | -| 384 | 200.23095 | -| 384 | 206.49485 | -| 384 | 200.49485 | -| 768 | 413.55765 | -| 768 | 394.92617 | -| 768 | 414.725 | -| 768 | 387.987 | -| 768 | 411.72035 | -| 768 | 394.926 | -| 768 | 409.375 | -| 768 | 393.091 | -| 768 | 412.600 | -| 1536 | 789.30077 | -| 1536 | 788.86531 | -| 1536 | 792.71864 | -| 1536 | 836.98 | -| 1536 | 801.205 | -| 1536 | 836.98 | -| 1536 | 820.9538 | -| 1536 | 707.048 | -| 3072 | 1639.62374 | -| 3072 | 1591.345 | -| 3072 | 1632.12712 | -| 3072 | 1674.444 | -| 3072 | 1618.100 | +| NUM_NODES | WORLD_SIZE | TIME | +|:---------:|:----------:|:----------:| +| 8 | 96 | 61.073 | +| | | | +| 16 | 192 | 107.74411 | +| 16 | 192 | 107.201338 | +| 16 | 192 | 107.10853 | +| | | | +| 32 | 384 | 200.23095 | +| 32 | 384 | 206.49485 | +| 32 | 384 | 200.49485 | +| | | | +| 64 | 768 | 413.55765 | +| 64 | 768 | 394.92617 | +| 64 | 768 | 414.725 | +| 64 | 768 | 387.987 | +| 64 | 768 | 411.72035 | +| 64 | 768 | 394.926 | +| 64 | 768 | 409.375 | +| 64 | 768 | 393.091 | +| 64 | 768 | 412.600 | +| | | | +| 128 | 1536 | 789.30077 | +| 128 | 1536 | 788.86531 | +| 128 | 1536 | 792.71864 | +| 128 | 1536 | 836.98 | +| 128 | 1536 | 801.205 | +| 128 | 1536 | 836.98 | +| 128 | 1536 | 820.9538 | +| 128 | 1536 | 707.048 | +| | | | +| 256 | 3072 | 1639.62374 | +| 256 | 3072 | 1591.345 | +| 256 | 3072 | 1632.12712 | +| 256 | 3072 | 1674.444 | +| 256 | 3072 | 1618.100 | + -
WORLD_SIZE=96: From d0e3d797cda2ca21f0a0962f0496a22f40ebf0bf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 22:20:07 -0500 Subject: [PATCH 240/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 189 +++++++++++++++++++++++++++++++----------------- 1 file changed, 124 insertions(+), 65 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index c417165989b..84c389880ea 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -79,40 +79,38 @@ helpers_main() { # 13. Setup run command to be executed. ############################################################################## setup() { - # printf "%s" "$(printBlue "${AURORA_GPT_HEADER}")" # 1. Identify machine we're on get_machine || exit - # 2. Load `conda` environment + # 2. Load `conda` environment, setup virtual env setup_python || exit + # 3. Ensure necessary dependencies all installed install_dependencies || exit - # 3. Determine WORLD_SIZE, etc. from `PBS_*` vars + # 4. Determine WORLD_SIZE, etc. from `PBS_*` vars setup_ezpz || exit - # 4. Set command line arguments to pass to `"${EXEC}"` + # 5. Set command line arguments to pass to `"${EXEC}"` setParams || exit - # 5. Create `deepspeed_config.json` from runtime params from ^ + # 6. Create `deepspeed_config.json` from runtime params from ^ buildDSconfig || exit - # 6. Specify output directory for {logs, checkpoints, etc.} + # 7. Specify output directory for {logs, checkpoints, etc.} setOutput || exit - # 7. Specify additional `deepspeed` arguments (dependent on _newly created_ variables) + # 8. Specify additional `deepspeed` arguments (dependent on _newly created_ variables) setArgs || exit - # 8. Ensure executable exists in expected path + # 9. Ensure executable exists in expected path export EXEC="${EXEC:-${HERE}/pretrain_gpt_alcf.py}" check_executable "${EXEC}" dfl="${DATA_FILE_LIST:-}" - # 9. Setup data + tokenizer - # via `DATA_FILE_LIST` and `TOKENIZER_TYPE` + # 10. Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit make_data || exit - # 10. Print job info + # 11. Print job info printJobInfo || exit - # 11. Print info about loaded modules and runtime environment + # 12. Print info about loaded modules and runtime environment save_dotenv "${CKPT_DIR}" || exit - # 12. Check that were not already running, if so, exit. + # 13. Check that were not already running, if so, exit. check_and_kill_if_running || exit - # 13. Setup run command to be executed + # 14. Setup run command to be executed setup_run_cmd || exit - # make_ds_hostfile || exit } ##################################################### @@ -216,6 +214,9 @@ save_dotenv() { fi } +###################################################################### +# get_machine_name: Return current machine name, as lowercase string +###################################################################### get_machine_name() { if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then machine="aurora" @@ -304,20 +305,21 @@ printJobInfo() { # will launch with `deepspeed` instead of `mpiexec`. ############################################################################# setupLauncher() { - # outdir=$1 if [[ "$#" == 1 ]]; then local dist_launcher="$1" else local dist_launcher="${LAUNCH_WITH:-${LAUNCH_CMD:-"MPICH"}}" fi if [[ "${dist_launcher}" == "deepspeed" ]]; then + # Save {PATH, LD_LIBRARY_PATH, ...} to .deepspeed_env + saveDSenv || exit # Assert `./hostfile_deepspeed` exists - export hfds="${WORKING_DIR}/hostfile_deepspeed" && [ -f "${hfds}" ] || exit + export hfds="${WORKING_DIR}/hostfile_deepspeed" + make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" - # if [[ -n "${DIST_LAUNCH}" && ${LAUNCH_CMD:-"MPICH"} != "deepspeed" ]]; then else if [[ -n "${DIST_LAUNCH}" ]]; then - LAUNCHER="${DIST_LAUNCH} --genvall --cpu-bind depth -d 16 $(which python3) -Wignore ${EXEC}" + LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" export LAUNCHER="${LAUNCHER}" else echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" @@ -390,20 +392,6 @@ get_batch_size_on_polaris() { } -################################################# -# Fix for distributed key value store on Aurora -################################################# -use_kvs_fix_on_aurora() { - export CCL_KVS_MODE=mpi - export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH - ######################################################### - # if not set, CCL will complain... ? - export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" - ######################################################### -} - ############################################################################## # `setParams`: Set / configure run options by parsing environment. @@ -421,6 +409,7 @@ setParams() { # +------[Aurora]--------||-------[SunSpot]-------------+ if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 + export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 @@ -428,14 +417,20 @@ setParams() { MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] - use_kvs_fix_on_aurora + # use_kvs_fix_on_aurora + # update_ccl_env_vars_aurora ###################################################################### # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! - export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... - if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" + if [[ -z "${USE_FLASH_ATTN:-}" ]]; then + export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" + FLASH_ARG="--use-flash-attn-builder" + fi else - # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" + echo "Using flash-attn !!" FLASH_ARG="--use-flash-attn-builder" fi ###################################################################### @@ -447,14 +442,13 @@ setParams() { export BE="${NCCL}" # BE = NCCL # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS # NOTE: MICRO_BATCH is exported below # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else - # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" FLASH_ARG="--use-flash-attn-v2" fi echo "Setting up AWS NCCL OFI Plugin on Polaris..." @@ -469,7 +463,6 @@ setParams() { if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else - # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-v2" FLASH_ARG="--use-flash-attn-v2" fi fi @@ -501,7 +494,7 @@ setParams() { export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL - export SAVE_INTERVAL=${SAVE_INTERVAL:-200} # HOW FREQUENTLY TO SAVE CKPTS + export SAVE_INTERVAL=${SAVE_INTERVAL:-50} # HOW FREQUENTLY TO SAVE CKPTS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? @@ -509,7 +502,7 @@ setParams() { export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - export MODEL_TYPE="llama-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL + export MODEL_TYPE="llama-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ if [[ "${SP}" == 1 ]]; then export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" @@ -517,11 +510,13 @@ setParams() { export LLAMA_ARGS="" echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" fi + # -----[Learning Rate Settings]-------------------------------------------- export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args + # -----[Learning Rate Settings]-------------------------------------------- if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then TIMING_STR="\ --timing-log-level ${TIMING_LOG_LEVEL} \ @@ -534,6 +529,12 @@ setParams() { } +############################################## +# setArgs +# +# Specify additional (DeepSpeed specific) +# arguments to pass to pretrain_gpt_alcf.py +############################################## setArgs() { # ---- Set DeepSpeed arguments -------------------------------- ds_args=" " @@ -543,6 +544,9 @@ setArgs() { fi ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" + if [[ "${ZERO_STAGE}" == 3 ]]; then + ds_args="--use-mics ${ds_args}" + fi if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args=" --deepspeed-activation-checkpointing ${ds_args}" @@ -577,45 +581,67 @@ make_ds_hostfile() { sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } +#################### +# ezpz_savejobenv +# +# Parse relevant environment variables to determine: +# - num_hosts (by counting the number of lines in ${HOSTFILE:-${PBS_NODEFILE}}) +# - num_gpus_per_host (magically[^1]) +# - num_gpus = (num_hosts * num_gpus_per_host) +# +# [^1]: See: [`ezpz/bin/savejobenv`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/savejobenv) +#################### ezpz_savejobenv() { + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" "$@" file=$(mktemp) curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${file}" source "${file}" || exit } ezpz_getjobenv() { + # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" "$@" file=$(mktemp) curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" source "${file}" || exit } -# +---------------------------------------+ -# | 1. Git clone ezpz (if not found) | -# | 2. Install ezpz (if not installed) | -# +---------------------------------------+ +########################################### +# setup_ezpz +# +# 1. {save,get}jobenv +# 2. python3 -m ezpz.jobs && source "./.jobenv" +# 2. Install ezpz (if not installed) +########################################### setup_ezpz() { if [[ -n "${HOSTFILE:-${PBS_NODEFILE}}" ]]; then ezpz_savejobenv else ezpz_getjobenv fi - python3 -m ezpz.jobs && source "./.jobenv" - if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then + ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') + if [[ -z "${ezloc:-}" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" else - echo "Found ezpz!" - fi - ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') - if [[ -n "${ezloc}" ]]; then - # ezpz_savejobenv - # python3 -m ezpz.jobs && source "./.jobenv" - make_ds_hostfile || exit - else - echo "No ezpz detected. Attempting to install with $(which python3)" - python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --require-virtualenv + printf "Found ezpz @ %s\n" "${ezloc}" fi + python3 -m ezpz.jobs && source "./.jobenv" echo "Done with ezpz." + # ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') + # if [[ -n "${ezloc}" ]]; then + # # ezpz_savejobenv + # # python3 -m ezpz.jobs && source "./.jobenv" + # make_ds_hostfile || exit + # else + # echo "No ezpz detected. Attempting to install with $(which python3)" + # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --require-virtualenv + # fi + # if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then + # mkdir -p "${WORKING_DIR}/deps" + # git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" + # else + # echo "Found ezpz!" + # fi } ####################################################################### @@ -698,6 +724,7 @@ buildDSconfig() { # # This will sum the weights (first column) from each line in the passed # `file_list`. +############################################################################### sumWeights() { local file_list=$1 weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") @@ -785,6 +812,39 @@ setup_conda_sunspot() { fi } + +################################################# +# Fix for distributed key value store on Aurora +################################################# +use_kvs_fix_on_aurora() { + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + ######################################################### + # if not set, CCL will complain... ? + export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" + ######################################################### +} + +update_ccl_env_vars_aurora() { + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + # unset CCL_CONFIGURATION_PATH + export CCL_CONFIGURATION=cpu_gpu_dpcpp + export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + # export CCL_ALLREDUCE_SCALEOUT=direct + printenv | grep -E -v "^__" | grep -E "CCL|LD|CPATH|LIBRARY_PATH" + ######################################################### + # if not set, CCL will complain... ? + export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" + ######################################################### +} + ########################### # Setup conda on Aurora ########################### @@ -1073,8 +1133,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # This will contain the appropriate variables as set in the current environment. ################################################################################ generateDSconfig() { - for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" \ - "$PP" "$DTYPE" + for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE" do if [ -z "$v" ]; then echo "Please export required envs before execute $0" @@ -1272,10 +1331,10 @@ printWhite() { export AURORA_GPT_HEADER=""" █████╗ ██╗ ██╗██████╗ ██████╗ ██████╗ █████╗ ██████╗ ██████╗ ████████╗ ██╔══██╗██║ ██║██╔══██╗██╔═══██╗██╔══██╗██╔══██╗ ██╔════╝ ██╔══██╗╚══██╔══╝ -███████║██║ ██║██████╔╝██║ ██║██████╔╝███████║█████╗██║ ███╗██████╔╝ ██║ -██╔══██║██║ ██║██╔══██╗██║ ██║██╔══██╗██╔══██║╚════╝██║ ██║██╔═══╝ ██║ -██║ ██║╚██████╔╝██║ ██║╚██████╔╝██║ ██║██║ ██║ ╚██████╔╝██║ ██║ -╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ +███████║██║ ██║██████╔╝██║ ██║██████╔╝███████║█████╗██║ ███╗██████╔╝ ██║ +██╔══██║██║ ██║██╔══██╗██║ ██║██╔══██╗██╔══██║╚════╝██║ ██║██╔═══╝ ██║ +██║ ██║╚██████╔╝██║ ██║╚██████╔╝██║ ██║██║ ██║ ╚██████╔╝██║ ██║ +╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ """ ########################### From bb690e3888524098fe0089611005f6aff3988e8f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 22:21:00 -0500 Subject: [PATCH 241/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 103 ++++++++++++------------------------------- 1 file changed, 29 insertions(+), 74 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index bf1c9eef194..53da0dbbdb3 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -25,45 +25,41 @@ from megatron.utils import get_ltor_masks_and_position_ids from megatron.utils import average_losses_across_data_parallel_group, update_rotary_pos_emb from megatron.arguments import core_transformer_config_from_args -from megatron.utils import Profile, PerfTrace - -# from megatron.utils import ( -# # report_memory, -# # throughput_calculator, -# # checkpoint_throughput_calculator -# ) -# from pathlib import Path +# from megatron.utils import Profile, PerfTrace + import logging import deepspeed from deepspeed.runtime.utils import see_memory_usage -from deepspeed.accelerator.real_accelerator import get_accelerator +# from deepspeed.accelerator.real_accelerator import get_accelerator import subprocess import wandb -# import time from torch import nn import torch.nn.functional as F import ezpz as ez -import_time = time.time() - python_start_time -start_setup_time = time.time() +dt_imports = time.time() - python_start_time +t0_setup = time.time() + # ---- [SETUP COMMS] ------------------------ # if str(os.environ.get('LAUNCH_CMD', 'mpich')).lower() == 'mpich': -RANK = ez.setup_torch(backend="deepspeed", timeout=7200) -setup_time = time.time() - start_setup_time +RANK = ez.setup_torch(backend="deepspeed") # , timeout=7200) +dt_setup = time.time() - t0_setup # else: # RANK = ez.get_rank() WORLD_SIZE = ez.get_world_size() LOCAL_RANK = ez.get_local_rank() -DEVICE = ez.get_torch_device() +DEVICE_TYPE = ez.dist.get_torch_device_type() if torch.cuda.is_available(): torch.cuda.set_device(LOCAL_RANK) -# ------------------------------------------- -# --- [TURN OFF LOGGER ON ALL RANK != 0] ---- + log = logging.getLogger(__name__) -log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") -log.info(f"Import python modules in {import_time} seconds") -log.info(f"ez.setup_torch time: {setup_time} seconds") +LOG_LEVEL = str(os.environ.get("LOG_LEVEL", "INFO")).upper() +# set logging level to "INFO" on RANK == 0, "CRITICAL" on all other ranks +log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") + +log.info(f"Import python modules in {dt_imports} seconds") +log.info(f"ez.setup_torch time: {dt_setup} seconds") # ---- [SETUP WANDB FROM RANK 0] -------------- WANDB_MODE = os.environ.get('WANDB_MODE', None) @@ -80,12 +76,11 @@ ), ) ) - print('--------------------------------------------------') - print(f"Setting up W&B from: {RANK} with {project_name}") - print('--------------------------------------------------') + log.info(f"Setting up W&B from: {RANK} with {project_name}") _ = ez.setup_wandb(project_name=project_name) +@ez.dist.timeitlogit(rank=RANK) def model_provider(pre_process=True, post_process=True): """Build the model.""" log.info('building GPT model ...') @@ -93,28 +88,8 @@ def model_provider(pre_process=True, post_process=True): args = get_args() assert args is not None config = core_transformer_config_from_args(args) - # from ezpz.profile import get_context_manager - # cm = get_context_manager(rank=RANK, outdir=args.save) - # profiler = None - # if os.environ.get("PYINSTRUMENT_PROFILER", None) is not None: - # from ezpz.profile import PyInstrumentProfiler - # cm = PyInstrumentProfiler(rank=RANK, outdir=args.save) - # else: - # from contextlib - # if wandb.run is not None and RANK == 0: - # print(f"Updating WandB run: [{wandb.run.name}]({wandb.run.url})") - # try: - # wandb.run.config.update({"args": vars(args)}) - # except Exception: - # log.error( - # 'Unable to `wandb.run.config.update({"args": vars(args)})`' - # ) - # if wandb is not None and wandb.run is not None: - # assert wandb is not None and wandb.run is not None - # print(f'Updating {wandb.run.name=} at {wandb.run.url=}') - # wandb.run.config.update({'args': vars(args)}) - if RANK == 0: - git_ds_info() + # if RANK == 0: + # git_ds_info() if hasattr(mpu, 'get_sequence_data_parallel_group'): dpg = mpu.get_sequence_data_parallel_group() elif hasattr(mpu, 'get_data_parallel_group'): @@ -143,7 +118,6 @@ def model_provider(pre_process=True, post_process=True): # get_batch_pipe from within training.py # We need to call model.set_batch_fn after deepspeed.initialize model._megatron_batch_fn = get_batch_pipe - # Precompute the attention mask and store it in args. # This avoids having to pipeline it # as an activation during training. @@ -151,25 +125,21 @@ def model_provider(pre_process=True, post_process=True): attention_mask = torch.tril( torch.ones( (1, args.seq_length, args.seq_length), - device=get_accelerator().current_device_name() + device=DEVICE_TYPE, ) ).view(1, 1, args.seq_length, args.seq_length) - # Convert attention mask to binary: attention_mask = (attention_mask < 0.5) if args.fp16: attention_mask = attention_mask.half() elif args.bf16: attention_mask = attention_mask.bfloat16() - # Attention mask must be bool. args.attn_mask = attention_mask.to(torch.bool) - - # For prertaining, since sequence length is fixed, + # For pretraining, since sequence length is fixed, # cache rotary embedding in args, to avoid communicating around if args.use_rotary_position_embeddings: update_rotary_pos_emb(args.seq_length) - else: model = GPTModel( config=config, @@ -178,19 +148,18 @@ def model_provider(pre_process=True, post_process=True): pre_process=pre_process, post_process=post_process ) - num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) log.info(80 * '-') log.info(f"Number of parameters in model: {num_params}") log.info(80 * '-') see_memory_usage("After Building Model", force=True) - if wandb.run is not None: + if wandb is not None and getattr(wandb, 'run', None) is not None: + assert wandb.run is not None tbdir = args.tensorboard_dir # tbdir = args.getattr('tensorboard_dir', None) if tbdir is not None: log.info(f'Patching tensorboard from {tbdir}') wandb.tensorboard.patch(root_logdir=tbdir) - wandb.run.config.update({'num_params': num_params}) if "args" not in wandb.run.config: log.info( @@ -219,11 +188,10 @@ def get_batch(data_iterator): """Generate a batch""" args = get_args() tokenizer = get_tokenizer() - + assert args is not None and tokenizer is not None # Items and their type. keys = ['text'] datatype = torch.int64 - data = next(data_iterator) if data_iterator is not None else None # # Broadcast data. # if data_iterator is not None: @@ -231,12 +199,10 @@ def get_batch(data_iterator): # else: # data = None data_b = tensor_parallel.broadcast_data(keys, data, datatype) - # Unpack. tokens_ = data_b['text'].long() labels = tokens_[:, 1:].contiguous() tokens = tokens_[:, :-1].contiguous() - # Get the masks and postition ids. skip_mask = args.use_flash_attn or args.use_flash_attn_triton attention_mask, loss_mask, position_ids = get_ltor_masks_and_position_ids( @@ -246,33 +212,29 @@ def get_batch(data_iterator): args.reset_attention_mask, args.eod_mask_loss, skip_mask) - # For DS's sequence parallel seq_parallel_world_size = mpu.get_sequence_parallel_world_size() seq_parallel_world_rank = mpu.get_sequence_parallel_rank() - # For Megatron's sequence parallel if args.sequence_parallel: seq_parallel_world_size = mpu.get_tensor_model_parallel_world_size() seq_parallel_world_rank = mpu.get_tensor_model_parallel_rank() seq_length = tokens.size(1) - assert seq_length % seq_parallel_world_size == 0 sub_seq_length = seq_length // seq_parallel_world_size sub_seq_start = seq_parallel_world_rank * sub_seq_length sub_seq_end = (seq_parallel_world_rank + 1) * sub_seq_length - tokens = tokens[:, sub_seq_start:sub_seq_end] position_ids = position_ids[:, sub_seq_start:sub_seq_end] # For DS's sequence parallel if mpu.get_sequence_parallel_world_size() > 1: labels = labels[:, sub_seq_start:sub_seq_end] - return tokens, labels, loss_mask, attention_mask, position_ids def data_post_process(data, data_sampler_state_dict): args = get_args() + assert args is not None if args.data_efficiency_curriculum_learning: if 'seqlen_truncate' in data_sampler_state_dict['current_difficulties']: args.data_efficiency_curriculum_learning_seqlen_type = 'seqlen_truncate' @@ -304,19 +266,16 @@ def get_batch_pipe(data): """ args = get_args() tokenizer = get_tokenizer() - + assert args is not None # Items and their type. keys = ['text'] datatype = torch.int64 - # Broadcast data. data_b = tensor_parallel.broadcast_data(keys, data, datatype) - # Unpack. tokens_ = data_b['text'].long() labels = tokens_[:, 1:].contiguous() tokens = tokens_[:, :-1].contiguous() - # Get the masks and postition ids. attention_mask, loss_mask, position_ids = get_ltor_masks_and_position_ids( tokens, @@ -336,16 +295,15 @@ def get_batch_pipe(data): if labels is not None: labels = labels[:, :args.curriculum_seqlen].contiguous() loss_mask = loss_mask[:, :args.curriculum_seqlen].contiguous() - return (tokens, position_ids, attention_mask), (labels, loss_mask) def loss_func(loss_mask, moe_loss, mos_loss, output_tensor): args = get_args() + assert args is not None losses = output_tensor.float() loss_mask = loss_mask.view(-1).float() loss = torch.sum(losses.view(-1) * loss_mask) / loss_mask.sum() - # Reduce loss for logging. averaged_loss = average_losses_across_data_parallel_group([loss]) if args.mos or args.kd: @@ -389,7 +347,6 @@ def calculate_mos_loss( alpha = args.kd_alpha_ce beta = args.kd_beta_ce kd_temp = args.kd_temp - if teacher_model: with torch.no_grad(): if ( @@ -417,19 +374,16 @@ def calculate_mos_loss( f'Teacher: {tea_output.size()}, ' f'CL seq length {args.curriculum_seqlen}' ) - student_logits = F.log_softmax(stu_output / kd_temp, dim=2) # The target logits is expected to be probabilities. # If we use log_softmax, # then we need to set target_log to true # when initializing the KLDivLoss. tea_logits = F.softmax(tea_output / kd_temp, dim=2) - mos_loss = kd_temp * kd_temp * nn.KLDivLoss(reduction='batchmean')( student_logits, tea_logits ) - mos_loss = mos_loss.div(args.seq_length) * beta return mos_loss @@ -512,6 +466,7 @@ def forward_step(data_iterator, model): return output_tensor, partial(loss_func, loss_mask, moe_loss, mos_loss) +@ez.dist.timeitlogit(rank=RANK) def train_valid_test_datasets_provider(train_val_test_num_samples): """Build train, valid, and test datasets.""" t0 = time.perf_counter() From aa698daf79a1d87cd252e032a4c039b89b4eb9b2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 22:21:13 -0500 Subject: [PATCH 242/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 778c4e82fe3..1850f2e92f2 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -1,43 +1,45 @@ #!/bin/bash --login -#PBS -l walltime=06:00:00 -#PBS -A argonne_tpc -#PBS -q prod -#PBS -l select=48 -#PBS -l filesystems=eagle:home -############################################################################# -# Check if running in `DEBUG=1` mode. +############################################################################### +# Check if running in DEBUG=1 mode. # - If so, this will print each command before it is ran and exit if any of # them return a nonzero exit status. -############################################################################# +############################################################################### if [[ -n "${DEBUG-}" ]]; then # to use: `DEBUG=1 bash train_llama_alcf.sh` printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" set -euxo pipefail fi +############################################################################### +# Print (but DO NOT EXECUTE !!) each command that would be ran. +# +# Enable with: NOOP=1 PBS_O_WORKDIR=$(pwd) bash train_llama_alcf.sh +############################################################################### if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` echo "Run NOOP mode" set -o noexec # same as set -n fi +##################### +# MAIN PROGRAM LOGIC +##################### main() { - ##################### - # MAIN PROGRAM LOGIC - ##################### - #### 1. Navigate into `$PBS_O_WORKDIR` + # 1. Navigate into `$PBS_O_WORKDIR` cd "${PBS_O_WORKDIR}" || exit - #### 2. source `ALCF/helpers.sh` HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE + # 2. source `ALCF/helpers.sh` source "${HERE}/ALCF/helpers.sh" || exit - #### 3. call `setup` from `./ALCF/helpers.sh` + # 3. call `setup` from `./ALCF/helpers.sh` setup || exit - - # Take custom args + # 4. Take custom args export custom_args=" $@" + # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} export run_cmd="${run_cmd} ${custom_args}" - + # 6. Add "${run_cmd}" to output log echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" + # 7. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" set +x } From 12baf309288c2d4ed1742fd65e2b9662cbe2c09e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 22:52:57 -0500 Subject: [PATCH 243/590] Update `megatron/training.py` --- megatron/training.py | 331 ++++++++++++++++++++----------------------- 1 file changed, 151 insertions(+), 180 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 77b36a0bc0b..f4417d9fa37 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -3,7 +3,6 @@ from datetime import datetime import math -from pathlib import Path import sys import os import time @@ -69,11 +68,13 @@ RANK: int = ez.get_rank() WORLD_SIZE: int = ez.get_world_size() -DEVICE_TYPE: str = ez.get_torch_device() +# DEVICE_TYPE: str = ez.get_torch_device() +DEVICE_TYPE: str = ez.dist.get_torch_device_type() DEVICE: torch.device = torch.device(DEVICE_TYPE) log: logging.Logger = logging.getLogger(__name__) -log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") +LOG_LEVEL: str = str(os.environ.get("LOG_LEVEL", "INFO")).upper() +log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") try: import wandb @@ -151,7 +152,7 @@ def pretrain( args_defaults={}, data_post_process=None, external_args={}, -) -> torch.nn.Module: +) -> list[torch.nn.Module]: """Main training program. This function will run the followings in the order provided: @@ -203,7 +204,7 @@ def pretrain( f"time to finish initialize_megatron: {time.time() - _TRAIN_START_TIME} seconds" ) # start_time_tensor = DEVICE.DoubleTensor([_TRAIN_START_TIME]) - start_time_tensor = torch.tensor([_TRAIN_START_TIME], device=DEVICE) + start_time_tensor = torch.tensor([_TRAIN_START_TIME], dtype=torch.double, device=DEVICE_TYPE) tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) log.info(f"allreduce call time: {time.time()-before_allreduce} seconds") _TRAIN_START_TIME = start_time_tensor.item() @@ -228,87 +229,84 @@ def pretrain( timers = get_timers() assert args is not None assert timers is not None - from ezpz.profile import get_context_manager - cm = get_context_manager(rank=RANK, outdir=args.save) - with cm: - if args.deepspeed: - args.deepspeed_config_dict = _create_ds_config_dict() - if ( - "curriculum_learning" in args.deepspeed_config_dict - and "enabled" in args.deepspeed_config_dict["curriculum_learning"] - ): - args.curriculum_learning_legacy = args.deepspeed_config_dict[ - "curriculum_learning" - ]["enabled"] - if args.curriculum_learning_legacy and not args.no_pipeline_parallel: - from deepspeed.runtime.data_pipeline.curriculum_scheduler import ( - CurriculumScheduler, - ) - - args.curriculum_scheduler = CurriculumScheduler( - args.deepspeed_config_dict["curriculum_learning"] - ) - if "compression_training" in args.deepspeed_config_dict: - args.compression_training = True - - # Model, optimizer, and learning rate. - timers("model-and-optimizer-setup", log_level=0).start(barrier=True) - model, optimizer, opt_param_scheduler = setup_model_and_optimizer( - model_provider, - model_type, - teacher=False, - data_post_process=data_post_process, - build_train_valid_test_datasets_provider=train_valid_test_dataset_provider, + if args.deepspeed: + args.deepspeed_config_dict = _create_ds_config_dict() + if ( + "curriculum_learning" in args.deepspeed_config_dict + and "enabled" in args.deepspeed_config_dict["curriculum_learning"] + ): + args.curriculum_learning_legacy = args.deepspeed_config_dict[ + "curriculum_learning" + ]["enabled"] + if args.curriculum_learning_legacy and not args.no_pipeline_parallel: + from deepspeed.runtime.data_pipeline.curriculum_scheduler import ( + CurriculumScheduler, ) - timers("model-and-optimizer-setup").stop() - print_datetime("after model, optimizer, and learning rate " "scheduler are built") - # Data stuff. - timers("train/valid/test-data-iterators-setup", log_level=0).start(barrier=True) - if args.virtual_pipeline_model_parallel_size is not None: - all_data_iterators = [ - build_train_valid_test_data_iterators(train_valid_test_dataset_provider) - for _ in range(len(model)) - ] - train_data_iterator = [ - data_iterators[0] for data_iterators in all_data_iterators - ] - valid_data_iterator = [ - data_iterators[1] for data_iterators in all_data_iterators - ] - test_data_iterator = [ - data_iterators[2] for data_iterators in all_data_iterators - ] - else: - train_data_iterator, valid_data_iterator, test_data_iterator = ( - build_train_valid_test_data_iterators(train_valid_test_dataset_provider) + + args.curriculum_scheduler = CurriculumScheduler( + args.deepspeed_config_dict["curriculum_learning"] ) - if args.data_efficiency_curriculum_learning: - if args.deepspeed_dataloader is not None: - # We use args to pass the deepspeed_dataloader because adding - # output to setup_model_and_optimizer will break the API for other - # cases. We clear args.deepspeed_dataloader after updating - # train_data_iterator because args will be saved in checkpoint and - # attempting to save the whole deepspeed_dataloader will lead to - # "AttributeError: Can't pickle local object...". - train_data_iterator = iter(args.deepspeed_dataloader) - args.deepspeed_dataloader = None - else: - train_data_iterator = None - timers("train/valid/test-data-iterators-setup").stop() - print_datetime("after dataloaders are built") - # args.teacher_model is used as global variable to pass the teacher model - # for knowledge distillation. Users do not need to set it in the command - # line to use kd, but users do need to provide teacher model configurations - # like args.num_layers_teacher as described in setup_teacher_model() - args.teacher_model = None - if args.mos or args.kd: # Set up teacher model - args.teacher_model = setup_teacher_model(args, model_provider) - # Print setup timing. - log.info("done with setup ...") - timers.log( - ["model-and-optimizer-setup", "train/valid/test-data-iterators-setup"], - barrier=True, + if "compression_training" in args.deepspeed_config_dict: + args.compression_training = True + + # Model, optimizer, and learning rate. + timers("model-and-optimizer-setup", log_level=0).start(barrier=True) + model, optimizer, opt_param_scheduler = setup_model_and_optimizer( + model_provider, + model_type, + teacher=False, + data_post_process=data_post_process, + build_train_valid_test_datasets_provider=train_valid_test_dataset_provider, + ) + timers("model-and-optimizer-setup").stop() + print_datetime("after model, optimizer, and learning rate " "scheduler are built") + # Data stuff. + timers("train/valid/test-data-iterators-setup", log_level=0).start(barrier=True) + if args.virtual_pipeline_model_parallel_size is not None: + all_data_iterators = [ + build_train_valid_test_data_iterators(train_valid_test_dataset_provider) + for _ in range(len(model)) + ] + train_data_iterator = [ + data_iterators[0] for data_iterators in all_data_iterators + ] + valid_data_iterator = [ + data_iterators[1] for data_iterators in all_data_iterators + ] + test_data_iterator = [ + data_iterators[2] for data_iterators in all_data_iterators + ] + else: + train_data_iterator, valid_data_iterator, test_data_iterator = ( + build_train_valid_test_data_iterators(train_valid_test_dataset_provider) ) + if args.data_efficiency_curriculum_learning: + if args.deepspeed_dataloader is not None: + # We use args to pass the deepspeed_dataloader because adding + # output to setup_model_and_optimizer will break the API for other + # cases. We clear args.deepspeed_dataloader after updating + # train_data_iterator because args will be saved in checkpoint and + # attempting to save the whole deepspeed_dataloader will lead to + # "AttributeError: Can't pickle local object...". + train_data_iterator = iter(args.deepspeed_dataloader) + args.deepspeed_dataloader = None + else: + train_data_iterator = None + timers("train/valid/test-data-iterators-setup").stop() + print_datetime("after dataloaders are built") + # args.teacher_model is used as global variable to pass the teacher model + # for knowledge distillation. Users do not need to set it in the command + # line to use kd, but users do need to provide teacher model configurations + # like args.num_layers_teacher as described in setup_teacher_model() + args.teacher_model = None + if args.mos or args.kd: # Set up teacher model + args.teacher_model = setup_teacher_model(args, model_provider) + # Print setup timing. + log.info("done with setup ...") + timers.log( + ["model-and-optimizer-setup", "train/valid/test-data-iterators-setup"], + barrier=True, + ) if not args.skip_train: log.info("training ...") if args.dataloader_type == "cyclic" and args.retro_add_retriever: @@ -425,6 +423,7 @@ def setup_teacher_model(args, model_provider): @dlp.log +@ez.dist.timeitlogit(rank=RANK) def get_model( model_provider_func, model_type=ModelType.encoder_or_decoder, wrap_with_ddp=True ): @@ -528,7 +527,7 @@ def get_model( # GPU allocation. for model_module in model: - model_module.to(get_accelerator().current_device_name()) + model_module.to(DEVICE_TYPE) # Fp16 conversion. if args.fp16 or args.bf16: @@ -571,10 +570,11 @@ def get_model( @dlp.log +@ez.dist.timeitlogit(rank=RANK) def get_optimizer_param_scheduler(optimizer): """Build the learning rate scheduler.""" args = get_args() - + assert args is not None # Iteration-based training. if args.train_iters: if args.lr_decay_iters is None: @@ -624,13 +624,11 @@ def get_optimizer_param_scheduler(optimizer): def load_model_weights_only(model_provider_func): """Setup model and optimizer.""" args = get_args() + assert args is not None log.info("***>>>>> Args:{}".format(args)) - model = get_model(model_provider_func) - optimizer = None lr_scheduler = None - if args.deepspeed: # When loading just the model weights, ZeRO can be disabled. if "zero_optimization" in args.deepspeed_config_dict: @@ -640,24 +638,26 @@ def load_model_weights_only(model_provider_func): model=model[0], config=args.deepspeed_config_dict ) - assert not isinstance( - model, deepspeed.PipelineEngine - ), "Weight loading only mode is not supported in pipeline parallelism yet." - + assert not isinstance(model, deepspeed.PipelineEngine), ( + "Weight loading only mode is not supported in " + "pipeline parallelism yet." + ) model = [model] - print_datetime("before load checkpoint") if args.load is not None: iteration = load_checkpoint( - model, optimizer, lr_scheduler, strict=True, load_only_weights=True + model, + optimizer, + lr_scheduler, + strict=True, + load_only_weights=True ) - print_datetime("after load checkpoint weights") - return model, optimizer, lr_scheduler @dlp.log +@ez.dist.timeitlogit(rank=RANK) def setup_model_and_optimizer( model_provider_func, model_type, @@ -710,47 +710,31 @@ def setup_model_and_optimizer( ) # opt_param_scheduler is the old lr_scheduler plus weight decay scheduling opt_param_scheduler = get_optimizer_param_scheduler(optimizer) - # HAS_PYINSTRUMENT = False - # scalene_profiler = None - # pyinstrument_profiler = None - # scalene_start = time.perf_counter_ns() - # pyinstrument_start = time.perf_counter_ns() - # if RANK == 0: - # # if (scprof := os.environ.get("SCALENE_PROFILER", None)) is not None: - # if HAS_SCALENE and os.environ.get("SCALENE_PROFILER", None) is not None: - # try: - # from scalene import scalene_profiler - # HAS_SCALENE = True - # log.warning("Starting scalene profiler") - # scalene_profiler.start() - # except Exception as exc: - # log.exception('Unable to start scalene profiler') - # # if (pyiprof := os.environ.get("PYINSTRUMENT_PROFILER", None)) is not None: - # if os.environ.get("PYINSTRUMENT_PROFILER", None) is not None: - # try: - # import pyinstrument - # pyinstrument_profiler = pyinstrument.Profiler() - # pyinstrument_profiler.start() - # HAS_PYINSTRUMENT = True - # except Exception as exc: - # log.exception(f'Unable to start pyinstrument profiler.') if args.deepspeed: log.info("DeepSpeed is enabled.") - pp = mpu.get_pipeline_model_parallel_world_size() + # pp = mpu.get_pipeline_model_parallel_world_size() if ( args.data_efficiency_curriculum_learning and build_train_valid_test_datasets_provider is not None ): + log.info( + "Caught 'args.data_efficiency_curriculum_learning' " + "and 'build_train_valid_test_datasets_provider is not None'" + ) train_ds = None # Only need to build dataset on tp rank 0 since Megatron has the # broadcast_data() function that broadcast data from tp rank 0. if mpu.get_tensor_model_parallel_rank() == 0: + log.info( + f"Caught 'mpu.get_tensor_model_parallel_rank() == 0'" + ) # Number of train/valid/test samples. if args.train_samples: train_samples = args.train_samples update_train_iters(args) else: train_samples = args.train_iters * args.global_batch_size + log.info(f'{train_samples=}') # eval_iters and test_iters here are not actually used, only for # satisfying the input of build_train_valid_test_datasets_provider. # We only need to build the training data here. And we follow @@ -765,55 +749,49 @@ def setup_model_and_optimizer( eval_iters * args.global_batch_size, test_iters * args.global_batch_size, ] + log.info(f"{train_val_test_num_samples=}") # Build the datasets. train_ds, _, _ = build_train_valid_test_datasets_provider( train_val_test_num_samples ) - with Profile("deepspeed.initialize"): - model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = ( - deepspeed.initialize( - model=model[0], - optimizer=optimizer, - args=args, - lr_scheduler=opt_param_scheduler, - training_data=train_ds, - mpu=mpu if args.no_pipeline_parallel else None, - config=args.deepspeed_config_dict, - ) + # with Profile("deepspeed.initialize"): + model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = ( + deepspeed.initialize( + model=model[0], + optimizer=optimizer, + args=args, + lr_scheduler=opt_param_scheduler, + training_data=train_ds, + mpu=mpu if args.no_pipeline_parallel else None, + config=args.deepspeed_config_dict, ) + ) model.set_data_post_process_func(data_post_process) else: - model, optimizer, _, opt_param_scheduler = deepspeed.initialize( - model=model[0], - optimizer=optimizer, - args=args, - lr_scheduler=opt_param_scheduler, - mpu=mpu if args.no_pipeline_parallel else None, - config=args.deepspeed_config_dict, + log.info( + "Did NOT catch: ('args.data_efficiency_curriculum_learning' " + "and 'build_train_valid_test_datasets_provider is not None')" ) - # if RANK == 0: - # # if HAS_SCALENE and scalene_profiler is not None: - # # dtscalene = time.perf_counter_ns() - scalene_start - # # log.info(f"Finished with scalene profile. Took: {dtscalene:.5s}s") - # # scalene_profiler.stop() - # if HAS_PYINSTRUMENT and pyinstrument_profiler is not None: - # dtpyinstrument = time.perf_counter_ns() - pyinstrument_start - # pyinstrument_profiler.stop() - # pyinstrument_profiler.print(color=True) - # sp = getattr(args, 'save', None) - # fp = Path(os.getcwd()) if sp is None else Path(sp) - # fpath = Path(fp).joinpath( - # f'pyinstrument-profile-{ez.get_timestamp()}.html' - # ) - # log.info( - # 'Saving pyinstrument profile output to: ' - # f'{fpath.as_posix()}' - # ) - # pyinstrument_profiler.write_html(fpath) - # log.info(' '.join([ - # f'Finished with pyinstrument profiler.', - # f'Took: {dtpyinstrument:.5f}s', - # ])) + tds0 = time.time() + if os.environ.get("PYINSTRUMENT_PROFILER", None): + profiler = ez.profile.get_context_manager( + rank=RANK, outdir=args.save + ) + else: + from contextlib import nullcontext + profiler = nullcontext() + log.info(f"Calling 'deepspeed.initialize'...") + log.info(f"Wrapped with: {profiler=}") + with profiler: + model, optimizer, _, opt_param_scheduler = deepspeed.initialize( + model=model[0], + optimizer=optimizer, + args=args, + lr_scheduler=opt_param_scheduler, + mpu=mpu if args.no_pipeline_parallel else None, + config=args.deepspeed_config_dict, + ) + log.info(f"'deepspeed.initialize' took: {time.time() - tds0:.5f}s") if isinstance(model, deepspeed.PipelineEngine): # hack to get batch_fn from pretrain_gpt.py model.set_batch_fn(model.module._megatron_batch_fn) @@ -827,7 +805,6 @@ def setup_model_and_optimizer( ) assert model.grid.get_data_parallel_rank() == mpu.get_data_parallel_rank() model = [model] - # Compression has its own checkpoint loading path (e.g, loading both teacher # and student models). So if compression is enabled, we skip the following # checkpoint loading. @@ -844,11 +821,9 @@ def setup_model_and_optimizer( args.iteration = 0 else: model[0].global_steps = student_global_steps - # We only support local DDP with multiple micro-batches. if len(model) > 1 or mpu.get_pipeline_model_parallel_world_size() > 1: assert args.DDP_impl == "local" - # get model without FP16 and/or TorchDDP wrappers if ( args.iteration == 0 @@ -859,11 +834,9 @@ def setup_model_and_optimizer( unwrapped_model[0].init_state_dict_from_bert() if args.fp16: optimizer.reload_model_params() - # random-LTD requires converting transformer layers if args.random_ltd: model[0] = convert_to_random_ltd(model[0], ParallelTransformerLayer) - return model, optimizer, opt_param_scheduler @@ -875,6 +848,7 @@ def train_step( args = get_args() timers = get_timers() + assert args is not None and timers is not None if args.deepspeed and args.ds_pipeline_enabled: skipped_iter = 0 num_zeros_in_grad = 0 @@ -1022,6 +996,7 @@ def training_log( args = get_args() timers = get_timers() writer = get_tensorboard_writer() + assert args is not None and timers is not None and writer is not None wandb_metrics = {} # Advanced, skipped, and Nan iterations. advanced_iters_key = "advanced iterations" @@ -1438,7 +1413,6 @@ def training_log( opt_stats_2[3], args.consumed_train_tokens, ) - writer.add_scalar( "optimizer/variance_l2", opt_stats[0] ** 0.5, iteration ) @@ -1600,8 +1574,8 @@ def training_log( log_string += " tokens_per_gpu_per_second_tgs={:.3f} |".format( tokens_per_gpu_per_second ) - log_string += " [LM]-TFLOPs={:.2f} |".format(tflops_lm_per_gpu) - log_string += " [DS]-TFLOPs={:.2f} |".format(tflops) + log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) + log_string += " [DS]TFLOPs={:.2f} |".format(tflops) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 @@ -1624,8 +1598,10 @@ def training_log( @dlp.log +@ez.dist.timeitlogit(rank=RANK) def save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler): timers = get_timers() + assert timers is not None # Extra barrier is added to make sure # all ranks report the max time. # assert timers is not None @@ -1864,7 +1840,7 @@ def evaluate( ): """Evaluation.""" args = get_args() - + assert args is not None if args.vision_pretraining and args.vision_pretraining_type == "dino": compute_feature_bank(model) @@ -2051,6 +2027,7 @@ def cyclic_iter(iter): @dlp.log +@ez.dist.timeitlogit(rank=RANK) def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): """Build pretraining datasets.""" @@ -2078,15 +2055,13 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): @dlp.log +@ez.dist.timeitlogit(rank=RANK) def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider): """Build pretraining data loaders.""" - args = get_args() - + assert args is not None (train_dataloader, valid_dataloader, test_dataloader) = (None, None, None) - log.info("> building train, validation, and test datasets ...") - # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: assert ( @@ -2100,7 +2075,6 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider * args.eval_iters * args.global_batch_size ) - # Data loader only on rank 0 of each model parallel group. ds_sequence_parallel = ( mpu.get_sequence_parallel_world_size() > 1 or args.force_ds_sequence_parallel @@ -2115,7 +2089,6 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider train_ds, valid_ds, test_ds = build_train_valid_test_datasets( build_train_valid_test_datasets_provider ) - # Build dataloders. train_dataloader = build_pretraining_data_loader( train_ds, args.consumed_train_samples @@ -2124,7 +2097,6 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider valid_ds, args.consumed_valid_samples ) test_dataloader = build_pretraining_data_loader(test_ds, 0) - # Flags to know if we need to do training/validation/testing. do_train = train_dataloader is not None and args.train_iters > 0 do_valid = valid_dataloader is not None and args.eval_iters > 0 @@ -2135,7 +2107,6 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider ) else: flags = get_accelerator().LongTensor([0, 0, 0]) - # Broadcast num tokens. if ds_sequence_parallel: torch.distributed.broadcast( @@ -2152,11 +2123,11 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider args.do_train = flags[0].item() args.do_valid = flags[1].item() args.do_test = flags[2].item() - return train_dataloader, valid_dataloader, test_dataloader @dlp.log +@ez.dist.timeitlogit(rank=RANK) def build_train_valid_test_data_iterators(build_train_valid_test_datasets_provider): """Build pretraining data iterators.""" From 8eabb7a0f5fd873e3a8431c504702f164c99b7c7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 23:24:50 -0500 Subject: [PATCH 244/590] Update `megatron/training.py` --- megatron/training.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training.py b/megatron/training.py index f4417d9fa37..81c49c4ba2b 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -996,7 +996,7 @@ def training_log( args = get_args() timers = get_timers() writer = get_tensorboard_writer() - assert args is not None and timers is not None and writer is not None + assert args is not None and timers is not None wandb_metrics = {} # Advanced, skipped, and Nan iterations. advanced_iters_key = "advanced iterations" From d9fc18ebcc224d39db3d7e1fe8c82e1859940a3f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 24 Jun 2024 23:25:39 -0500 Subject: [PATCH 245/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 92 +++++++++++++++++++++---------------------------- 1 file changed, 39 insertions(+), 53 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 84c389880ea..d8798d145a7 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -18,13 +18,14 @@ # > ``` ############################################################################## -############################################################################## -# `helpers_main`: +################## +# helpers_main # -# This will get called automatically when +# This will get called automatically when running: # # ```bash -# $ source ALCF/helpers.sh +# $ cd Megatron-DeepSpeed +# $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh # ``` # # - This will set `"${WORKING_DIR}"`, according to: @@ -35,7 +36,7 @@ # this is crucial since many of the functions below use paths # which are defined relative to this "${WORKING_DIR}" # (e.g. virtual environment, location of executables, etc.) -############################################################################## +################## helpers_main() { # for debug mode, run with `DEBUG=1` if [[ -n "${DEBUG:-}" ]]; then @@ -55,28 +56,31 @@ helpers_main() { } ############################################################################## -# `setup`: All-in-one helper function. +# setup +# +# All-in-one helper function. # # - Explicitly, this will: # 1. Identify the machine we're on # 2. Setup `python` # 1. Load `conda` # 2. Setup `venv` on top of `conda` -# 3. Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) +# 3. Ensure all dependencies are installed +# 4. Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) # 1. Additionally, call `source deps/ezpz/src/ezpz/bin/savejobenv`, # which will automatically build a `alias launch=mpiexec ...` # according to the specifics of our active job. -# 4. Set runtime options -# 5. Build `deepspeed_config.json` -# 6. Build {logs, checkpoints, etc} dirs, named according to specifics of +# 5. Set runtime options +# 6. Build `deepspeed_config.json` +# 7. Build {logs, checkpoints, etc} dirs, named according to specifics of # current run -# 7. Specify additional `deepspeed` arguments -# 8. Ensure executable exists at expected path -# 9. Setup data + tokenizer via `TOKENIZER_TYPE` -# 10. Print job info -# 11. Save `.env` to `CKPT_DIR` for safe keeping -# 12. Check that we're not already running, and if so, exit. -# 13. Setup run command to be executed. +# 8. Specify additional `deepspeed` arguments +# 9. Ensure executable exists at expected path +# 10. Setup data + tokenizer via `TOKENIZER_TYPE` +# 11. Print job info +# 12. Save `.env` to `CKPT_DIR` for safe keeping +# 13. Check that we're not already running, and if so, exit. +# 14. Setup run command to be executed. ############################################################################## setup() { # 1. Identify machine we're on @@ -105,7 +109,7 @@ setup() { make_data || exit # 11. Print job info printJobInfo || exit - # 12. Print info about loaded modules and runtime environment + # 12. Save `.env` to `CKPT_DIR` for safe keeping save_dotenv "${CKPT_DIR}" || exit # 13. Check that were not already running, if so, exit. check_and_kill_if_running || exit @@ -114,7 +118,9 @@ setup() { } ##################################################### -# `setup_run_cmd`: Build run command to be executed. +# setup_run_cmd +# +# Build run command to be executed. ##################################################### setup_run_cmd() { #### Make it easy to track experiments by date ################### @@ -292,7 +298,7 @@ printJobInfo() { } ############################################################################# -# `setupLauncher`: Launch with one of `{mpiexec, deepspeed}`. +# setupLauncher: Launch with one of `{mpiexec, deepspeed}`. # # Explicitly, look for `LAUNCH_CMD` in environment and launch accordingly. # Will use `mpiexec` by default. @@ -332,26 +338,6 @@ setupLauncher() { printf " %s" "$(printMagenta "${LAUNCHER}")" } -# ########################################################################## -# # setDSlauncher -# # -# # When launching with `deepspeed`, this will -# # set the appropriate keyword arguments to be passed to `deepspeed` -# ########################################################################## -# setDSlauncher() { -# # launcher setting -# outdir=$1 -# export hfds="$outdir/hostfile_deepspeed" -# export hfmpi="$outdir/hostfile_mpich" -# [ -f "$hfds" ] || exit -# [ -f "$hfmpi" ] || exit -# export LAUNCHER=${LAUNCHER:-MPICH} -# if [[ $LAUNCHER == "deepspeed" ]]; then -# export launcher="" -# else -# export launcher="--force_multi --hostfile $hfds --launcher=${LAUNCHER} --launcher_args='-hostfile ${hfmpi}'" -# fi -# } set_lr_args() { LR_ARGS="--lr ${LR} --lr-decay-style cosine" @@ -392,7 +378,6 @@ get_batch_size_on_polaris() { } - ############################################################################## # `setParams`: Set / configure run options by parsing environment. # @@ -417,11 +402,11 @@ setParams() { MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] - # use_kvs_fix_on_aurora - # update_ccl_env_vars_aurora + # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? + update_ccl_env_vars_aurora ###################################################################### - # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! if [[ -z "${USE_FLASH_ATTN:-}" ]]; then + # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" @@ -645,7 +630,7 @@ setup_ezpz() { } ####################################################################### -# `ezpz_test`: Run simple test to make sure all nodes in working order +# ezpz_test: Run simple test to make sure all nodes in working order ####################################################################### ezpz_test() { printf "%s" "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" @@ -657,10 +642,12 @@ ezpz_test() { printf "%s" "[$(printBlue 'ezpz:test_dist')] Done with test!\n" } -# +------------------------------------------------------------------------+ -# | Save important environment variables to .deepspeed_env, which will be | -# | forwarded to ALL ranks with DeepSpeed | -# +------------------------------------------------------------------------+ +############################################################################ +# saveDSenv +# +# Save important environment variables to .deepspeed_env, which will be +# forwarded to ALL ranks with DeepSpeed +############################################################################ saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { @@ -794,7 +781,6 @@ install_deepspeed_for_xpu() { } - ######################################################## # Setup / activate conda environment(s) ######################################################## @@ -830,10 +816,10 @@ use_kvs_fix_on_aurora() { update_ccl_env_vars_aurora() { export CCL_KVS_MODE=mpi - export CCL_CONFIGURATION_PATH="" + # export CCL_CONFIGURATION_PATH="" # unset CCL_CONFIGURATION_PATH - export CCL_CONFIGURATION=cpu_gpu_dpcpp - export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + # export CCL_CONFIGURATION=cpu_gpu_dpcpp + # export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH From 1d413c6881bdd21d2f92347a65df9670b1b1081c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Jun 2024 08:59:03 -0500 Subject: [PATCH 246/590] Update `megatron/training.py` --- megatron/training.py | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 81c49c4ba2b..a396ed2b4aa 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -754,18 +754,18 @@ def setup_model_and_optimizer( train_ds, _, _ = build_train_valid_test_datasets_provider( train_val_test_num_samples ) - # with Profile("deepspeed.initialize"): - model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = ( - deepspeed.initialize( - model=model[0], - optimizer=optimizer, - args=args, - lr_scheduler=opt_param_scheduler, - training_data=train_ds, - mpu=mpu if args.no_pipeline_parallel else None, - config=args.deepspeed_config_dict, + with Profile("deepspeed.initialize"): + model, optimizer, args.deepspeed_dataloader, opt_param_scheduler = ( + deepspeed.initialize( + model=model[0], + optimizer=optimizer, + args=args, + lr_scheduler=opt_param_scheduler, + training_data=train_ds, + mpu=mpu if args.no_pipeline_parallel else None, + config=args.deepspeed_config_dict, + ) ) - ) model.set_data_post_process_func(data_post_process) else: log.info( @@ -778,9 +778,8 @@ def setup_model_and_optimizer( rank=RANK, outdir=args.save ) else: - from contextlib import nullcontext - profiler = nullcontext() - log.info(f"Calling 'deepspeed.initialize'...") + profiler = Profile("deepspeed.initialize") + log.info("Calling 'deepspeed.initialize'...") log.info(f"Wrapped with: {profiler=}") with profiler: model, optimizer, _, opt_param_scheduler = deepspeed.initialize( From 93e4a51952628c76ff60cf35fbef38e5feee99ab Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Jun 2024 08:59:24 -0500 Subject: [PATCH 247/590] Update `megatron/utils.py` --- megatron/utils.py | 326 ++++++++++++++++++++++++++++++---------------- 1 file changed, 215 insertions(+), 111 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 22e9a926512..7a1d60bab92 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -4,60 +4,112 @@ import sys import os +import time +import logging +from typing import ContextManager, Optional import torch from torch.nn.parallel import DistributedDataParallel as torchDDP from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == 'cuda': + +if get_accelerator().device_name() == "cuda": try: from apex.multi_tensor_apply import multi_tensor_applier import amp_C + HAS_APEX = True except Exception: HAS_APEX = False -from megatron import ( - get_args, - get_adlr_autoresume, - get_num_microbatches -) +from megatron import get_args, get_adlr_autoresume, get_num_microbatches from megatron.core import mpu from megatron.core.tensor_parallel import param_is_not_tensor_parallel_duplicate from megatron.model.module import param_is_not_shared from megatron.model.rotary_pos_embedding import RotaryEmbedding -_DLIO_PROFILER_EXIST=True + +import ezpz as ez + +RANK = ez.get_rank() +log = logging.getLogger(__name__) +# log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") + + +_DLIO_PROFILER_EXIST = True try: import dlio_profiler except: - _DLIO_PROFILER_EXIST=False + _DLIO_PROFILER_EXIST = False if _DLIO_PROFILER_EXIST: from dlio_profiler.logger import fn_interceptor as Profile from dlio_profiler.logger import dlio_logger as PerfTrace else: from functools import wraps + # from contextlib import nullcontext + # Profile: ContextManager = nullcontext + # + # class Profile(nullable_schema) + class Profile: def __init__(self, type="PROFILER"): + self._start = time.perf_counter() self.type = type + def log(self, func): return func + def iter(self, a): return a + + def __enter__(self): + self._start = time.perf_counter() + + def __exit__(self, *args, **kwargs): + dt = time.perf_counter() - self._start + log.info(f"{self.type} took: {dt:.6f}s") + + class dlio_logger: - def __init__(self,): + def __init__( + self, + ): self.type = None + def initialize_log(self, logfile=None, data_dir=None, process_id=-1): return + def iter(self, a): return a + PerfTrace = dlio_logger() +def get_logger( + name: str, + level: str = "INFO", + rank_zero_only: Optional[bool] = None, +) -> logging.Logger: + """Returns a `logging.Logger` object. + + If `rank_zero_only` passed, the level will be set to CRITICAL on all + non-zero ranks (and will be set to `level` on RANK==0). + """ + logger = logging.getLogger(name) + logger.setLevel(level) + if rank_zero_only and ez.get_rank() != 0: + logger.setLevel("CRITICAL") + return logger + + def update_rotary_pos_emb(seq_length): args = get_args() - rotary_dim = args.hidden_size // args.num_attention_heads \ - if args.kv_channels is None else args.kv_channels + assert args is not None + rotary_dim = ( + args.hidden_size // args.num_attention_heads + if args.kv_channels is None + else args.kv_channels + ) if args.rotary_percent < 1.0: rotary_dim = int(rotary_dim * args.rotary_percent) @@ -66,7 +118,8 @@ def update_rotary_pos_emb(seq_length): # Wang and Komatsuzaki et al # https://github.com/kingoflolz/mesh-transformer-jax/ rotary_pos_emb = RotaryEmbedding(rotary_dim, theta=args.rope_theta)(seq_length).to( - get_accelerator().current_device_name()) + get_accelerator().current_device_name() + ) args.rotary_pos_emb = rotary_pos_emb @@ -86,8 +139,9 @@ def unwrap_model(model, module_instances=(torchDDP)): def calc_params_l2_norm(model): - """Calculate l2 norm of parameters """ + """Calculate l2 norm of parameters""" args = get_args() + assert args is not None if not isinstance(model, list): model = [model] # Remove duplicate params. @@ -103,32 +157,31 @@ def calc_params_l2_norm(model): params_data.append(param.data) # Calculate norm dummy_overflow_buf = get_accelerator().IntTensor([0]) - - if get_accelerator().device_name() == 'cuda' and HAS_APEX: + + if get_accelerator().device_name() == "cuda" and HAS_APEX: norm, _ = multi_tensor_applier( amp_C.multi_tensor_l2norm, dummy_overflow_buf, [params_data], - False # no per-parameter norm + False, # no per-parameter norm ) else: - norm = torch.norm(params_data,p=2.0) + norm = torch.norm(params_data, p=2.0) norm_2 = norm * norm # Sum across all model-parallel GPUs. - torch.distributed.all_reduce(norm_2, - op=torch.distributed.ReduceOp.SUM, - group=mpu.get_model_parallel_group()) + torch.distributed.all_reduce( + norm_2, op=torch.distributed.ReduceOp.SUM, group=mpu.get_model_parallel_group() + ) return norm_2.item() ** 0.5 def average_losses_across_data_parallel_group(losses): """Reduce a tensor of losses across all GPUs.""" - averaged_losses = torch.cat( - [loss.clone().detach().view(1) for loss in losses]) - torch.distributed.all_reduce(averaged_losses, - group=mpu.get_data_parallel_group()) - averaged_losses = averaged_losses / \ - torch.distributed.get_world_size(group=mpu.get_data_parallel_group()) + averaged_losses = torch.cat([loss.clone().detach().view(1) for loss in losses]) + torch.distributed.all_reduce(averaged_losses, group=mpu.get_data_parallel_group()) + averaged_losses = averaged_losses / torch.distributed.get_world_size( + group=mpu.get_data_parallel_group() + ) return averaged_losses @@ -136,44 +189,47 @@ def average_losses_across_data_parallel_group(losses): def report_memory(name): """Simple GPU memory report.""" mega_bytes = 1024.0 * 1024.0 - string = name + ' memory (MB)' - string += ' | allocated: {}'.format( - get_accelerator().memory_allocated() / mega_bytes) - string += ' | max allocated: {}'.format( - get_accelerator().max_memory_allocated() / mega_bytes) - string += ' | reserved: {}'.format( - get_accelerator().memory_reserved() / mega_bytes) - string += ' | max reserved: {}'.format( - get_accelerator().max_memory_reserved() / mega_bytes) + string = name + " memory (MB)" + string += " | allocated: {}".format( + get_accelerator().memory_allocated() / mega_bytes + ) + string += " | max allocated: {}".format( + get_accelerator().max_memory_allocated() / mega_bytes + ) + string += " | reserved: {}".format(get_accelerator().memory_reserved() / mega_bytes) + string += " | max reserved: {}".format( + get_accelerator().max_memory_reserved() / mega_bytes + ) if mpu.get_data_parallel_rank() == 0: - print("[Rank {}] {}".format(torch.distributed.get_rank(), string), - flush=True) + log.info(f"[Rank {RANK}] {string}") + # log.info("[Rank {}] {}".format(torch.distributed.get_rank(), string)) # , flush=True) def print_params_min_max_norm(optimizer, iteration): """Print min, max, and norm of all parameters.""" index = 0 rank = torch.distributed.get_rank() - string = 'iteration, rank, index, tensor-model-parallel, min, max, norm\n' + string = "iteration, rank, index, tensor-model-parallel, min, max, norm\n" optimizer_ = optimizer.optimizer for param_group in optimizer_.param_groups: - for param in param_group['params']: + for param in param_group["params"]: index += 1 min_ = param.data.min() max_ = param.data.max() norm = torch.linalg.norm(param.data) - string += '{:7d}, {:4d}, {:4d}, {:2d}, '.format( - iteration, rank, index, int(param.tensor_model_parallel)) - string += '{:.6E}, {:.6E}, {:.6E}\n'.format(min_, max_, norm) - print(string, flush=True) + string += "{:7d}, {:4d}, {:4d}, {:2d}, ".format( + iteration, rank, index, int(param.tensor_model_parallel) + ) + string += "{:.6E}, {:.6E}, {:.6E}\n".format(min_, max_, norm) + # print(string, flush=True) + log.info(string) -def check_adlr_autoresume_termination(iteration, model, - optimizer, opt_param_scheduler): +def check_adlr_autoresume_termination(iteration, model, optimizer, opt_param_scheduler): """Check for autoresume signal and exit if it is received.""" from megatron.checkpointing import save_checkpoint - args = get_args() + assert args is not None autoresume = get_adlr_autoresume() # Add barrier to ensure consistnecy. torch.distributed.barrier() @@ -187,12 +243,14 @@ def check_adlr_autoresume_termination(iteration, model, sys.exit(0) -def get_ltor_masks_and_position_ids(data, - eod_token, - reset_position_ids, - reset_attention_mask, - eod_mask_loss, - skip_mask=False): +def get_ltor_masks_and_position_ids( + data, + eod_token, + reset_position_ids, + reset_attention_mask, + eod_mask_loss, + skip_mask=False, +): """Build masks and position id for left to right model.""" # Extract batch size and sequence length. @@ -206,8 +264,9 @@ def get_ltor_masks_and_position_ids(data, attention_mask = None if not skip_mask: - attention_mask = torch.tril(torch.ones( - (att_mask_batch, seq_length, seq_length))).view(att_mask_batch, 1, seq_length, seq_length) + attention_mask = torch.tril( + torch.ones((att_mask_batch, seq_length, seq_length)) + ).view(att_mask_batch, 1, seq_length, seq_length) # Loss mask. loss_mask = torch.ones(data.size(), dtype=torch.float, device=data.device) @@ -215,8 +274,7 @@ def get_ltor_masks_and_position_ids(data, loss_mask[data == eod_token] = 0.0 # Position ids. - position_ids = torch.arange(seq_length, dtype=torch.long, - device=data.device) + position_ids = torch.arange(seq_length, dtype=torch.long, device=data.device) position_ids = position_ids.unsqueeze(0).expand_as(data) # We need to clone as the ids will be modifed based on batch index. if reset_position_ids: @@ -225,7 +283,6 @@ def get_ltor_masks_and_position_ids(data, if reset_position_ids or reset_attention_mask: # Loop through the batches: for b in range(micro_batch_size): - # Find indecies where EOD token is. eod_index = position_ids[b, data[b] == eod_token] # Detach indecies from positions if going to modify positions. @@ -238,15 +295,16 @@ def get_ltor_masks_and_position_ids(data, i = eod_index[j] # Mask attention loss. if reset_attention_mask and not skip_mask: - attention_mask[b, 0, (i + 1):, :(i + 1)] = 0 + attention_mask[b, 0, (i + 1) :, : (i + 1)] = 0 # Reset positions. if reset_position_ids: - position_ids[b, (i + 1):] -= (i + 1 - prev_index) + position_ids[b, (i + 1) :] -= i + 1 - prev_index prev_index = i + 1 # Convert attention mask to binary: if not skip_mask: - attention_mask = (attention_mask < 0.5) + assert attention_mask is not None + attention_mask = attention_mask < 0.5 attention_mask = attention_mask.to(data.device) return attention_mask, loss_mask, position_ids @@ -254,55 +312,82 @@ def get_ltor_masks_and_position_ids(data, def print_rank_0(message): """If distributed is initialized, print only on rank 0.""" - if torch.distributed.is_initialized(): - if torch.distributed.get_rank() == 0: - print(message, flush=True) - else: - print(message, flush=True) + # if torch.distributed.is_initialized(): + # if torch.distributed.get_rank() == 0: + # # print(message, flush=True) + # print(message, flush=True) + # else: + # print(message, flush=True) + _ = log.info(f"{message}") if RANK == 0 else None + def is_last_rank(): - return torch.distributed.get_rank() == ( - torch.distributed.get_world_size() - 1) + return torch.distributed.get_rank() == (torch.distributed.get_world_size() - 1) + def print_rank_last(message): """If distributed is initialized, print only on last rank.""" if torch.distributed.is_initialized(): if is_last_rank(): - print(message, flush=True) + # print(message, flush=True) + log.info(message) else: - print(message, flush=True) + log.info(message) + def is_aml(): # Are we running inside an Azure Machine Learning (AML) environment? - return 'AZUREML_EXPERIMENT_ID' in os.environ + return "AZUREML_EXPERIMENT_ID" in os.environ + def is_rank_0(): """Check whether it is rank 0. For AML, check if it is rank 0 of a node""" if torch.distributed.is_initialized(): if torch.distributed.get_rank() == 0 or ( - is_aml() and torch.distributed.get_rank() % get_accelerator().device_count() == 0 - ): + is_aml() + and ( + torch.distributed.get_rank() + % get_accelerator().device_count() + ) == 0 + ): return True else: return False else: return True + def get_parameters_in_billions(model): - gpus_per_model = torch.distributed.get_world_size(group=mpu.get_model_parallel_group()) + gpus_per_model = torch.distributed.get_world_size( + group=mpu.get_model_parallel_group() + ) - approx_parameters_in_billions = sum([sum([p.ds_numel if hasattr(p,'ds_id') else p.nelement() for p in model_module.parameters()]) - for model_module in model]) + approx_parameters_in_billions = sum( + [ + sum( + [ + p.ds_numel if hasattr(p, "ds_id") else p.nelement() + for p in model_module.parameters() + ] + ) + for model_module in model + ] + ) + + return approx_parameters_in_billions * gpus_per_model / (1e9) - return approx_parameters_in_billions*gpus_per_model/(1e9) def throughput_calculator(model, args, iteration_time, total_iterations): - batch_size = args.micro_batch_size * get_num_microbatches() * args.data_parallel_size - approx_parameters_in_billions = None if (model is None) else get_parameters_in_billions(model) - elapsed_time_per_iter = iteration_time/total_iterations + batch_size = ( + args.micro_batch_size * get_num_microbatches() * args.data_parallel_size + ) + approx_parameters_in_billions = ( + None if (model is None) else get_parameters_in_billions(model) + ) + elapsed_time_per_iter = iteration_time / total_iterations samples_per_second = batch_size / elapsed_time_per_iter - #flops calculator + # flops calculator hidden_size = args.hidden_size num_layers = args.num_layers vocab_size = args.padded_vocab_size @@ -312,50 +397,76 @@ def throughput_calculator(model, args, iteration_time, total_iterations): # The factor of 4 is when used with activation check-pointing, # otherwise it will be 3. checkpoint_activations_factor = 3 - if hasattr(args, 'checkpoint_activations') and args.checkpoint_activations: + if hasattr(args, "checkpoint_activations") and args.checkpoint_activations: checkpoint_activations_factor = 4 - if hasattr(args, 'recompute_granularity') and (args.recompute_granularity == 'selective' or args.recompute_granularity == 'full'): + if hasattr(args, "recompute_granularity") and ( + args.recompute_granularity == "selective" + or args.recompute_granularity == "full" + ): checkpoint_activations_factor = 4 seq_len = args.seq_length - if hasattr(args, 'actual_seq_length'): + if hasattr(args, "actual_seq_length"): seq_len = args.actual_seq_length - flops_per_iteration = (24 * checkpoint_activations_factor * batch_size * seq_len * num_layers * (hidden_size**2)) * (1. + (seq_len / (6. * hidden_size)) + (vocab_size / (16. * num_layers * hidden_size))) + flops_per_iteration = ( + 24 + * checkpoint_activations_factor + * batch_size + * seq_len + * num_layers + * (hidden_size**2) + ) * ( + 1.0 + + (seq_len / (6.0 * hidden_size)) + + (vocab_size / (16.0 * num_layers * hidden_size)) + ) tflops = flops_per_iteration / (elapsed_time_per_iter * args.world_size * (10**12)) return samples_per_second, tflops, approx_parameters_in_billions + def checkpoint_throughput_calculator(model, latency_second): approx_parameters_in_billions = get_parameters_in_billions(model) - checkpoint_multiplier = 14 # fp16 weights (2), fp32 weights (4), fp32 momentum (4), fp32 variance (4) + checkpoint_multiplier = ( + 14 # fp16 weights (2), fp32 weights (4), fp32 momentum (4), fp32 variance (4) + ) checkpoint_GB = approx_parameters_in_billions * checkpoint_multiplier GB_per_second = checkpoint_GB / latency_second - print_rank_0(f"Checkpoint Save GB: {round(checkpoint_GB, 3)}, GB/Sec: {round(GB_per_second,2)}, Latency(second): {round(latency_second, 3)}") + print_rank_0( + f"Checkpoint Save GB: {round(checkpoint_GB, 3)}, GB/Sec: {round(GB_per_second,2)}, Latency(second): {round(latency_second, 3)}" + ) def get_fingerprint_header(): return f"{'min':^13} {'max':^13} {'mean':^13} {'l2 norm':^12} metadata" + def get_fingerprint(p): return f"{p.min():13.6e} {p.max():13.6e} {p.mean():13.6e} {p.norm():12.6e}" def dump_position_embed_weights(preamble, iteration, model): - # return + # return from deepspeed.utils import safe_get_full_fp32_param + tp_rank = mpu.get_tensor_model_parallel_rank() pp_rank = mpu.get_pipeline_model_parallel_rank() dp_rank = mpu.get_data_parallel_rank() get_fingerprint_header() for n, p in model[0].named_parameters(): - if 'position_embeddings' in n: + if "position_embeddings" in n: tag = "pos_embed" elif "word_embeddings" in n: tag = "word_embed" else: - continue - print(f"iter {iteration} {preamble} {tag} lp {tp_rank}/{pp_rank}/{dp_rank}: {get_fingerprint(p)} {p.shape}\n") + continue + log.info( + f"iter {iteration} {preamble} {tag} lp {tp_rank}/{pp_rank}/{dp_rank}: {get_fingerprint(p)} {p.shape}\n" + ) fp32_value = safe_get_full_fp32_param(p) - if fp32_value is not None: - print(f"iter {iteration} {preamble} {tag} hp {tp_rank}/{pp_rank}/{dp_rank}: {get_fingerprint(fp32_value)} {p.shape}\n") + if fp32_value is not None: + log.info( + f"iter {iteration} {preamble} {tag} hp {tp_rank}/{pp_rank}/{dp_rank}: {get_fingerprint(fp32_value)} {p.shape}\n" + ) + def dump_weights(preamble, iteration, model, optimizer, tensor=None): # return @@ -364,34 +475,25 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): dp_rank = mpu.get_data_parallel_rank() dp_size = mpu.get_data_parallel_world_size() fn = f"debug-bf16-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" - # only care for first and last pp stages and dp0 tp0 - #if not (mpu.is_pipeline_first_stage() or mpu.is_pipeline_last_stage()): + # if not (mpu.is_pipeline_first_stage() or mpu.is_pipeline_last_stage()): # return - - #if not (tp_rank == 0 and dp_rank == 0): + # if not (tp_rank == 0 and dp_rank == 0): # return - if tensor is not None: orig_tensor = tensor if hasattr(tensor, "_hp_param"): - numel = tensor._hp_param.numel() # // dp_size + numel = tensor._hp_param.numel() # // dp_size tensor = tensor.flatten().narrow(0, 0, numel) - - #print(fn) + # print(fn) with open(fn, "w") as fh: fh.write(f"{get_fingerprint_header()}\n") - if tensor is not None: fh.write(f"{get_fingerprint(tensor)} tensor {tensor.shape}\n") else: for n, p in model[0].named_parameters(): fh.write(f"{get_fingerprint(p)} {n} {p.shape}\n") - - return - - # until we figure out how to dump the actual fp32 values don't do this fn = f"debug-fp32-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" with open(fn, "w") as fh: @@ -399,14 +501,16 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): if tensor is not None: tensor = orig_tensor if hasattr(tensor, "_hp_param"): - fh.write(f"{get_fingerprint(tensor._hp_param)} tensor {tensor._hp_param.shape}\n") - #fh.write(f"{get_fingerprint(tensor._hp_grad)} tensor grad\n") + fh.write( + f"{get_fingerprint(tensor._hp_param)} tensor {tensor._hp_param.shape}\n" + ) + # fh.write(f"{get_fingerprint(tensor._hp_grad)} tensor grad\n") else: fh.write(f"{get_fingerprint(tensor)} tensor {tensor.shape}\n") - #fh.write(f"{get_fingerprint(tensor.grad)} tensor grad\n") - + # fh.write(f"{get_fingerprint(tensor.grad)} tensor grad\n") else: if hasattr(model[0].module.tied_modules, "embed"): p = model[0].module.tied_modules.embed.word_embeddings.weight._hp_param - fh.write(f"{get_fingerprint(p)} module.tied_modules.embed.word_embeddings.weight._hp_param {p.shape}\n") - + fh.write( + f"{get_fingerprint(p)} module.tied_modules.embed.word_embeddings.weight._hp_param {p.shape}\n" + ) From 99bddfa86f1e891ffd089ae0a3e4f8a4f1a8e4e2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Jun 2024 11:37:36 -0500 Subject: [PATCH 248/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d8798d145a7..0cc6aa90131 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -603,15 +603,19 @@ setup_ezpz() { else ezpz_getjobenv fi - ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') - if [[ -z "${ezloc:-}" ]]; then + ezdir="${WORKING_DIR}/deps/ezpz" + if [[ ! -d "${ezdir}" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" + fi + ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') + if [[ -z "${ezloc:-}" ]]; then + printf "[setup_ezpz] Installing ezpz from %s\n" "${ezdir}" + python3 -m pip install -e "${ezdir}" --require-virtualenv else - printf "Found ezpz @ %s\n" "${ezloc}" + printf "[setup_ezpz] Found ezpz @ %s\n" "${ezloc}" fi python3 -m ezpz.jobs && source "./.jobenv" - echo "Done with ezpz." # ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') # if [[ -n "${ezloc}" ]]; then # # ezpz_savejobenv From 9a8ccfdf506116da398f351c0d62b821ba7d5f05 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Jun 2024 11:45:04 -0500 Subject: [PATCH 249/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 0cc6aa90131..61e6767cc26 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -894,7 +894,7 @@ setup_venv_from_conda() { export VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" echo " - Using VENV_DIR=${VENV_DIR}" if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then - printf "\n - Creating a new virtual env on top of %s in %s" "$(printBlue "${CONDA_NAME}")" "$(printGreen "${VENV_DIR}")" + printf "\n - Creating a new virtual env on top of %s in %s\n" "$(printBlue "${CONDA_NAME}")" "$(printGreen "${VENV_DIR}")" mkdir -p "${VENV_DIR}" python3 -m venv "${VENV_DIR}" --system-site-packages source "${VENV_DIR}/bin/activate" || exit From 57ba1fbac3dbab452b121feccb29e7dbb7624c58 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Jun 2024 23:32:33 -0500 Subject: [PATCH 250/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 61e6767cc26..5bd09c712c2 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -460,9 +460,10 @@ setParams() { export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") - if [[ -z "${NGPU_PER_HOST-}" ]]; then + if [[ -z "${NGPU_PER_HOST:-}" ]]; then NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') fi + export NGPU_PER_HOST="${NGPU_PER_HOST}" export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" # +---[Llama2 7B Config]--------------------------------------------------+ # export MODEL_KEY="Llama-7B" @@ -1231,7 +1232,7 @@ generateDSconfig() { # elif [[ $ZERO_STAGE == 2 ]]; then elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then # if [[ -n "${CPU_OPTIMIZER}" ]]; then - if [[ "${CPU_OPTIMIZER}" != 0 ]]; then + if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" zero="\ \"zero_optimization\": { From 634e37b3def714aa5c3fc4a0bdee18f96ddff078 Mon Sep 17 00:00:00 2001 From: Xinyu Lian Date: Thu, 27 Jun 2024 07:43:05 +0900 Subject: [PATCH 251/590] Add steps and results for running ZeRO stage 3 withUniversal Checkpointing (#383) --- .../universal_checkpointing/README.md | 23 +++++++++++++++++- .../image/uc_stage3_char_training_loss.png | Bin 0 -> 54650 bytes .../image/uc_stage3_char_validation_loss.png | Bin 0 -> 41972 bytes 3 files changed, 22 insertions(+), 1 deletion(-) create mode 100644 examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_training_loss.png create mode 100644 examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_validation_loss.png diff --git a/examples_deepspeed/universal_checkpointing/README.md b/examples_deepspeed/universal_checkpointing/README.md index 341b0d113f5..14169c9e228 100644 --- a/examples_deepspeed/universal_checkpointing/README.md +++ b/examples_deepspeed/universal_checkpointing/README.md @@ -116,4 +116,25 @@ Repeat steps in ZeRO stage 1 training above with the following modifications to * Set ZERO_STAGE=2 * Add `--no-pipeline-parallel` flag to deepspeed options -## ZeRO stage 3 training (**Coming soon**) +## ZeRO stage 3 training +Repeat steps in ZeRO stage 1 training above with the following modifications to your job batch scripts: +* Set ZERO_STAGE=3 +* Add `--no-pipeline-parallel` flag to deepspeed options + +> **Note:** that the stage 3 universal checkpoint currently supports Data parallelism. + +Below is the visualization of the `png` files generated from ZeRO stage 3. + +
+ + + *Figure 1: Training LM loss curve for first 200 training steps of Step 1 (TP=1, PP=1, DP=4) and training steps 101 to 200 of Step 3 (TP=1, PP=1, DP=2), which was loaded using the Universal Checkpoint.* +
+ +
+ + + *Figure 2: Validation LM loss curve for first 200 training steps of Step 1 (TP=1, PP=1, DP=4) and training steps 101 to 200 of Step 3 (TP=1, PP=1, DP=2), which was loaded using the Universal Checkpoint.* +
+ + diff --git a/examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_training_loss.png b/examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_training_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..4c6758e99111638b74eaf8c328d1640a41eb2bab GIT binary patch literal 54650 zcmeFZWmr{R)Hb>{jWkF|qlkhu0@5j9qO_=_0@B^xh?Iz+pn!;qw1{-aR-^?{Is_Du z?%15Mc%JwD-tYas6F<(Mb6w{Lm%-j^%{AAYW8C*W?lGfoTvw+!#(WGx5QqP@sZ3hci4^!v+$Td?}M_UJ1+XrTx?)RNt9yr*G30@Es zxqzwfahQyUq5e&nZ#XjL5)oMZ){#+q|z{tzOg_Ca1^H)6k@mJee>d zCHbJb8BSTL{=8;!DK6C~G4+kz;2XQ=@=F6>lYecFrn)@-=(n4;mUhaFDICKUO07)t zK)Bg$1W_h|e`r;mMYxat&*cCA3(153n@M;blp=cn*qf`*v<5ofeQJuJIwc}P6G6qX zIp;@5M_24PpRVX?*h<0$NZ`i&K-TRt``|(`@(`sEBCd1i$U=xH z8rzc2&A)nv_w(n^A0Hpkg;TJEzkdC?(*F9ht2}fP_QR2VeL8n)eNTATu$!8i{_f7i zZ+&zkB;l00^2r3mB!H8OP5Mcl9zR3Y%`7#mv06Ecj^|;oUvsyoN>V8t?z@C_$+#~V zR3B^(+1E)pjLI$ySJ0igp60l>@#M+zM@%BG+UGAlkaQSrp6kzb8n4Iv`t?hZCy|Sb z3v)_6wmFoPE<=ZJzA|c~bF{|CX|9jA{iR%0L8{h+(QWw+OPHkAz8vk96)eY8tjyxO z`I&fTRg8&T=8dyOZu8%@Fyk=J_@pGliaSQ?0_OzIoFV?2ETr!1TirzzNJvQNxloLC zoM}I?K0iREDZj(5tfG>ToE#BCOf|VPn~sfPkrZ~Bx!T}n-ukE^5QA@0^3!p;jo$W0 z5oT}mZ_}G=sluj>1nY|=I};{A@1w&Q`^AeF-3H37d*?SuNJ!rG&+gY9?kS|ncu; zm9c=siNR8v>;jvC0=C84J$sx}vf$n2Ux}J#P2m&<<&IohT3XwM2U}wS29>V7j}sFK z2l|pKw@)>g71vvo^nY=Ao1H!J(=z?yoja`G3+_gzq%h%Z63=M1!Yme9cD*!oNG^PL z*<_M5O!?-it&OKoo;>06OP_O2}gI!z8B1dCtSc0eome~O?bD`wZS}2E_`jaYi*WQ-j``>#DfDF z&rUW{HZyx=BKY+QuePv7J5!pxA2U3M+HB{Gl?GxC1X*2M6A8fiqW5K!ac{osF`{ez zuGOov4<1#-_J^?R?{3W2sE^R~7t+%`HeQp>r288K#^r~&y(@$PtcyG6dUN%KYVW#D zz?y(tos4CZ2`?Mc=$3+~GB8`oC%$rT#aJZx@O z^vR2NOVw+Xd3w3yyN!*?`(uX(9F?wr zf?*gFEsVy^J8QEw(uKkT0u(MY?MiX%a@`d!mcs#u2lj^~_LPk1#nHL)=7PMZt2AYuh|A=R(~4Qjl_iKy5g;STnu_IK7&B5^MqOJPj5C^ zK>Dj!CuMQlQ|KaCteh8ARa0v;5jKJQmV2%7@Myk##B=Sr2{x83Mbd%2&)O}=xDRqq zm|iX#A~|WJio7?hA(pe$9pZe#p2dt*dEf`CV~j4A8~=hd(_gFa#&7tMgfHnB1X5*K zrD#YABV(M*ij#Nt@z>EgUg@zE`**0!{CnQ5;KJ-TZy+5luLZ1o%dz_|E8VlSoC+dl zSG2X|zn*p>h)RANsi>&%+?h@$95`J6rYX!EgYHMJBcl z1Ds#2`(I{sQ0?tkVI@$WOdvs690iPPS^aj`8=9l&cGhrvO9~BJez*8`N1qFu;yF7z zqhXc*)`#Y)fz4-@FO^`(%fl{R6AM2}w&BAxG&Dhndv%9XUC-e4eR3F!ty-#Ofj!U) zIX&deOV7rs+2v1?qn`Ni$rLFU9#?6$z`20y(b3T>Kg{Dyx0ffFq@20k7l)%D5>k

_yC;+;TlWm15z5Ijk|5Pb|9B&!qd(qG@+|q6t%2saTvw@Dg1J2NBtcd*p_rS}|fgBPSy zbMym@#=Z#(7U!8yj3uXGxRd>@ci@H zzdF*S64C4QS(lENH<*|}YT7w`7#&S`a25U+l40ww*;z4I`t1`<5zl-UD@eI=GdG^ygB3{kk^;03vv8)MrVKhnF|%OMCl0rq&f^w?DmM zU%I*&)Aj&Nh-RtB36?Dl&kp{CcYJiM+4O*!IP}4;Dl6ly!4^sODt!EymWPWg$ZPe> znXi7=En#x50^)kcB5A*j1ss6o=H?cj|DHb;*O9J}{%5F6Nc+YO&T<^UkNcJGf7`aE z;3-VdHv;?XL!XP&68+Zme50D@BB>qiVqLOLu5*1+KkEYO#(fw4*k#=L%|)MWPvz?6 zo1x!NoQf_?32BnS>KmER`+%{|9*uCnX1rQ!`Nx(z0$&ne!HQ|Fx)f+eUm3(+=rfk zN{;RYYiI?Vr*{s0A#wifPJc&X!9Sa=-kC`~`aLX|Wh_J*v3sAXEtba{_AR>-m-Y(o zR*5=Puc%eb#y?k5so(hgIpLJ)dFEoDEqkl;%n%ITwoi-!$l>JXCV+K+QAr7NXbf+H zhKcDj)fHB0R}zNku*3O;HAoasE8pG`j4){TzX!uOMn(poiG&q630Gm?QV~o>K}WDT zd|AcS^*nHhCXU1}_fU`qu>+tBN5f8*T8y%l)!D<8bFr7MT*+PD5oe}Ao;>d=F>NA3 zQTAcihB)TO!et1#J!KBY^5#+=i^HN3k8Q}36DD2fzf(YjJaOX0eIqe|LmJsyuiv~` z#R?i#M*`?K?|XN1zZf1R0}#QD9G&1Ut#o-mlTyPgHIijx{xT;|o+QBP@X>$x@S%cD>tV>QeZrsQ&nE&=Rj#F&llih6sV&b**bK#~%HUom> zCqxOzn@Eh}Q1towbtp$AYzD^0rlt{pA*nk3I|1qbyANP0h#{CrL0X>uLT!Pq77+2-~mygMR6h3}W^u%J_ zp+7FgfWP(EmqfYBx-*-=Ga%rSKBNr^4o2C`y?gh@SEs)svU|1qIXc_09TQ%>_*=S3 z)74GwC`Su<0f?A6>^)&7@*=~0$eE!-E)K)xL9k_&N>fYr_k9sX3k#mWUU~;e*dFWnDUWD&R4d4HDk)G2&`mW=zoc!3bE$p*hf#l+Cd&J4D7 zxByr?%galXM#siRQBQOHx2tq$Xy~#r2PA3OqzK|rcks!?pT)8AC3jhMS+_C4(00TFw@iv9!=Q;Dt5-Dn`)5>|m(TMrc)eEqqZJ)oQ zQ{1-q^`#sa_JDmSR#{K#DQukj+Dr$!(jJ`dHV7`;0J(x3_9;BYm@0o zDJh8%tJ1;CcV~Ai@P2eLY)La5KdzXsXFk;$^P}7;Yskm35DO34^WpxLI6f);f;*%y zu17#}%LGz%|K&inmrKJ`Mh2&{FYzud);#lUO*KAR0wQQ;a$Htc3W5XyXD97 za;AochR^Nop$7bd8#_*aU2ZwMxDdvvC@a6OX^yMva&}XA3lK$cLyi?d4L}$j&wS;~ zvm01lE30=V>{jqrUS(xnp}6SdvJnJW!F6%CqR(m3CHdL2D2Oa^wM!68qP^B;OuI9b zP%2ME$r=ffh8RY&VdxX(a~%0DhEic5Uk|IRWyzU@0}r|8+AK=AL`7+rYPMbu6;|P> zSDG!;{leoEa7-u(dI`%uG(23HpTT@(vL(szdLzqxK~iZu!9Xks3;qYHlw0^u_qf~u z;xtgb`tid6muw6%0OH4$kvbR)a0<=RA!CM$JCuoE2*OP-A3Qc;B2vrhv8(^Xd_l%o z`(;m-`V^41#tN6&_$N<70TX8F7a9e3MRZuGjMe&~c%-66J(@v)Aa10lF8_8}D-bSV zr0ff0bsVrc8e#iKKNGVw1Eh`oxPJ*1Ok~%)RWz<&SJTiS0lfCoZ}aD6A$?j_)~Cpw3R@4$*g0;l$VftD1!WC=H4Uw17$SkX#aT6x&q($WAcRZi~ivU&?V zi)`iS#$F99?NM9yuAy{o2f(gPFv(X03U!U9y>wqTjfmb_^f%I9_oG5YL_|_#y+l#q z4gu~Fn+!v`-{uLzK=DU2uqTw1lspeM21I-|7o5wSm;v}g29)3Vm8fN0V!*%ecpVZJ zSJLHTNMj9B!=hB3+_aZ+>Kw})#*C~NiMCYUimIsL6Ok+0+ZUL~KW{N>sC1jBhU^JU z-N#ob@F1x7ttLl9KpJH0R}N#fs7&X7c;Hbs;B)TW&0E7l&amTaHBK{1@_U5LQi)q*BrM&M2ZEz zTQ+uOQQjD(900x|5B9dEfW|w@5Sl_D+@B{bTGJC?AUk}9iUg2|Al#qvjOFm(#+*KV z+GsD#5Pe6EC#oT=>p2dOYz4~{j4drb#jOxI1YEDlhh+w<0Dba+c8ijV16I9$py{X} z*bsYB`UPdGgq<=hZ4O^}T@H}c2|E)K5+b9bg1+2aMMKvUOkPAAv$ykiKw?)@gfZvz z3z{KdbVG25NxB1z7*)=`3e`Q5SuDrDdGlt9v>RXT_9Pjq62P-yAxRWDjLC;aM3CWn zInG;A>*W~)B^!B@f(TIVxpMN8?N2T10wPJSU0_No!2a;?@p~YVJI}hJ_XFhGTZdt@&HlxuX7&muf^%(3uwX| zneK0|Z12t&0=uV>ca-y7=0tP%$`m%sZs?PdUnj_~6Nk8^!)16KuVA{e-_WCED*(51 z#-cqXsMs9)3J_fP{02H&^!+K|wzC>}34wj7pmZ99iqD@l?$mhGkNK`qKN z5L9R&@JZMWjss@6;JIv)l>Fnt_q-D*&X?U=C_(pn9bF_Sbb#oKKyl)uppl1xU+~=) zM&ZUI4h43*!H-H#P9hLtLr~TS%Ml4gGfUn6D%q&_yg-_?+p)V{4;ki8#A>#Z?<4%=3#hCw|0E+eij5tFAQMS^!R__J}E{)Mnt-yLXYsL^A-#5Nn=! z1Nu>1n)gtVX;XKO_L=InP8n3nFf6qW1Jaww(X+DdHVl|9y~jlYLcZ)uGu@9*b}T4d zKn0@A%*^fSWaIdkFJnRYTFJUV?>Cm)Z;gtwr_>%Pf>?k8VpLiN$%F2m!KIE2q&rUF zPB}H3v}--AfCv!IsA;1JAn2lLg}^4OW%J3&%W>4S;JFbxfj zddTgcg&vOpps8MMJ7<*tYYIibaB;)@?{?|{BeB9)tRS#Z? zWf#No-Y53~vZ3l@?f&W&fOY5XhnckBTx|duttEYJ^{bEyU#X0AB8;aV1==4C}N0I|u~{=qrMsX!tnbAR5p+w--vf2yFEA*8K`~_(Xt9 zvR-gR0NLO*30uKNHS+l@UY09Ny+kYGCjR}R?{mX^HPED1fC-O(Ig_3@KSSM{aR8_m z6-t+zs1!0_6T6M18M-YFOQFfy6O{3qrwz7|Xx3>5$)SsxPT2(^B_~qJF&`d&$1dmH z=k7wnMWJ&`A-dVjDSH+pd9R=_3-l~Bld-a1t+JtJr@+twA)5rc-wL%9G`|4CMw!I# zfkKXhYz>keaS)wbVITK^J`URz)yDv@H-TU{T+pL2yf=1;BMxZ@1g$mYDi12OK_RIg z#y6XEzdSM^1BvH^!Wrx!RJ+#8y?E~)H^_NyLX=lYxJtP|@h7Hatp~*j5AZhtO;mjp zvuJ-e|EB*ONbilw#{N-&nCh+0zXIIpxzVpb@2XEf>0?F7iPkqY@Tat05R~Z{5TjUXB^_f%vR<<716W4d`PwH zA6hheLu_61sjnO?HYe`w?Ip;bc?n7Gs3>)?>xRm;D6CikU3vn_QCrGiNzsAKQKdh} zzW^c5->DA8)Vh9352TMMX9E@yaAd z%d%XxU2jDL@7pmL{d5`r2|6wkoW!8Ww-Uh!oyd6xb?z1S{3uZ>%V;<8>LE~#D^H>WAtqP#Ec(c{@1JgRDciA zk`e~KW^Oj4X>1Cq_SQ_|tIyq9yV!@_?Hb!7zQJ)2RJTR7d?>%`b2xW7gcQ-`OBy4nShi`8OaM*kZuqEn#+s0{q7ruNO!aX3H zW3eLSA)pjw0^aJcENht@p?cGAy|ydoNAbT55;s~`*=y1`2AO*tH1%@-1L*)rGAd{J z_!491r; zIUykk#r-*NUp7MMM-a#)^-!-x*BcseAe$R}xJODv!K`Y7&}vNS;n$MiIxG&{820_95GN-SpRQ_%@m#j-$yTY$ zCzC8n2XEZS*l4SHttA34_=dr|jiDV%^zBvz?Cex@&x?~#Qbybu*4sD!U` zza|bhs!)HpXYdmzLXMyKg(T(;H3ri_LJ{NYW2CA|O&`?fhEKw^k=S7|9a1-LqI*{o zJO5g%*qCVfzkE3f_XTQWfoj^3@$rt#l!Ev*e4+7t#Rho@tAZ^lSF^sOLh}9t`wwL| z5r|GF-3?mP!~ICPa6$fpx{vS_QL95dQTPl(tIYh-|DaOWvvYr^(6sNJOs?zUt&*3D zCMKLu`FGxF4M1jn0l1grrO#pupPa*eHUXB+ce-+IB)BDdY@CCbm^jO8?WyN-LqvDoVFq7m(Ao-iX=-mt(qy=NrF64q zq7Ze0oVNnrls&%k=tSIj?H+!J+PNJxJ@V9Eo4= z@7GODP8ORqoJ7d&M=oO_;e^B6sFxhGRlaxc3AQUWIXOfD_XAuCn^r>-q{%|Y@7;|T zs~qscDXQ>c%c#e(H|%cMBs_Z-Bspr(p8l=a+_F7whul`;9K^v#y1DW#P`E2{|Lbvs zFNs}Y0P6j~!!dxt*vZY;+O$Nzz3o!Hd|A?f-va7TriVB=sBNM00D*{7UZ?mlfRJ*X|O_B|f>sq;tu@-3m#L9fM(!5C+QUylal^t8*~9 z;mm5*h(BK`6}nvtc<`w-1Kl?!0FQhnyWLjj6*HF?lAXRj&41rl5E}YPx3ZuTplb{u z4h#advM4B%aEfmXNRuJvX3ShXZ)gsoNE1RG-C{o`iQG7Qe18CH`ZRQOAcb&XJ6{Yz zN#=N5dYRk&P|^Lbf`~0cFAIiquflBTd-fGcs4AEaeKKBIC=o)}1VT+q+nD|>M_cqs z8^Uk6&P>U+O9oJPZM$N3ldaxzXw-MjElnZd)8Jo+-q$(fZlK~FmvF$5F4a~s%?}i| zs8C&_OG{5lY2M1cl3c(fX@A~--^n})zn)6rPB1AQ59x>M*K<~s8C24q(quA}W6cig z-{vcML(hBKe9voMo;ZLi=Yy;1IFfMkKOr0e`|Jq7ysyG;a|GlZqa0hqVzh-d5*)x~ zgXb+fyzzvpm#^mR-K}swp(!s#p%Ss#@`QJ5YI#B>cpeh_b>HO%nf)m!ud`{t8IA*# zGMS*xWBQ}W-*bN@=RyE3ZCpA=EhLQ080t>=U=eTx^oGYG;XBoHqwHL|7J-5 zZoeXU_IC9)1*>d1ho+o3F=Pqm0^_>yrMj^j)3uiON0jf}xsy~-St^K)7 z5t6vtLtR9r)%NqtccR47*xSk)(`bgMSY&vU^tIW@7_ZT+cs%ON%f{}^GP2mN^h4%D zhur>$u%Ap9e%u67ewKr~(p&=QRVzQ#9H@})yjw14{KMh)IDt0m&bV2ZGJk#Q7ps?V zh;KqBKGXgzNX+tjmdortAf@yGSX;q>6C|_9M1lPNSwh8gt+H#X)-YcJXqZ(gv+d9) zLSX+5kZ=F2Y;>M~PV2kU;tfO#-(<_j7-kk&-da;7ZB+9-A}`?2&b))dnp1K;n7PWs~^ zAw*I}d+**^@f#G8}g?s~*b$Jq^UUkUT5_1-uJ zoXjWh&d&X?*~e{}NH;Mr{SyrLM#>U5PqkW(&Pm}L2Jyg@PPC_PvF>hhL_%$s#ME@5 zb?w_*1@f#OoyCLW#|ZCq3n#g`RhLudT`nHv#cpr&Y`!<7VHSTlF~(>N zg(6jO@X(4rV=XQ&CL!m)+c5~Sc&6Qp(yW-Nr$>Z@D(+p#Q$5P9h%2h0MQyuaqEK>l z6joN&0cVF(4xgze|J_{-_(C4G#!TJN(!xW)NF^sy3di$LTsoI&( z33W7pR=|%!7F+fpv0fje-w)ongd{gI84r6#*0B%MmbO1v#l9CFVmeaG-hFpDuY;~$ zEUO|j9?wP`+(_9-GMZ?BC#wg^(cvSZ^!Phy)JZQcB$ANtIZ_uur54MIAK3Af^}>x( z!bkY{_#$?Lu0YvO0s|!mhn3pZtK{t#VxpqGI~z1{Y!OxE0G$N1W9Ys1{?kGGC|wL03=eoV4>VCf|dNWOPA*GWW=+ zW|xzb6IrS*AxS)=5IQ_;>kG_0{wuw1<>5C?g{Zu|+w2|0YvJ%`itq%x?EbR3=dT3p zS2>TZ#W4&rP-q;|Z#~6v=5*<7G+X(L%e99%{pMB(XqBR%!-reG^g6aSQ`1Jgvj`#( ztMob}@~O)(|EWWOv$MXxAyVek23D#3<%)0LzB%r$2NW3!KtZe#3eHeg;b0Oram@a1 z@at-dBNYHx0T1?fRT~#TYlh-ePxh7RWQmiM=<5;0w>p2TLHJQuyD?Yj1vJ(|GG7F{omviEN1e>d!47bVBf?Vs!EWGbwc z?)qy!n~$L&Lbase5S7b!jp@oa#mKc6y&!Ae{}k}2w~Ly+lSmVYn~-qMXCPll5yrs4 z-~tpOSLzOHVgO(3{n^s7tJ_l5Xx3ThJvsfs(C+NnZav>u#Q!jw_@w;SsVnaIh`g6Q zVobk4RW8rKVC`pZA1k@z!%=~|Z^_7c<_ujzLUH_;Uv4A(g=?id=KUfrGgE-h3_e{g z-+ZC4(eq;P_ZUcI;d=RlCX2&f^yzo--w$D(rqrfDh~4^0_%wG=q2y=i{{HRzINyhn zB=36}VsOm?C&?5}CP#7M-8OI#zIDI{7`!h}&IhOH?#=-45Fb|BnEy}^D2zdX<8$ZV z?685_1ysVcAs8&F+V8o~og*U(!GI!EIR7SIS}`GvYLcgi<*MFuV&Fti0s|EnhMztB z`0+fTT@Q?%h@1R9_Y^DbMB{I%GPRbZH3}%W{UL2O#%V}#6;oT z8BKQLdV3P;|4P?bZ|BTj(bNQA?mQF}9zZrE?07DYt#s2JqT2mGqig)bWEaHU$QXqD zLjqVq(4XnJJqdPJ6 zKrZ2}W;$3!?TZZB(Ex`#@F_BBkX@gbL_qxr(%~cL9SsOL5gZ(RQ%5HpP_aZ`2$ z{jpm7ZQCJKl>kGcUe&?1=jZ9G7eSnYYLUXkhzFzj!EUMN@5~3mYjkvp+x}nyqlYSB zYn+08o(|%<_t#_mrB5rF=p*=IOQPoWriPF`5v%U*Ik)$iEU7^GhoGrUvVgNQ3h2_* z!q0Ch@&`2JCwhsYW+jJ=f747ho$DiIlCM@+tgz%Z^0AX&U0o#@=z6|EVB|UROlzyk zd0O@ zrd^#?pf#+aAanh27?X@xdnrGc~vCY@Y5{kAj_~bQF}CM%SS6c;K6Y6 z83k)j9L*^;RIEqqMMcmYaBRtMd)X8M&ePo`@WO4lOI8QNe)Wy`Y1_);P#rwvzyW^D z+YdpL#vlMPh3~hWGWhW8lR%=O^H_oWnH^`iCjh;eQWTh}lZ8FopBFb)wanx!giCFr)qM z&t*h~e4Fq5dB%0@V2Pxc$AV!aOfR2RMRa0fVss&Rpj%YcOR z#%HxvUk9hq&ktQa?{#eu+ssy2KWIpIT}*ofX~Jx*cEoc!NzxI^mW4q8G&j7ywB64=sHa62VwaCnoCj<)HCPILt065u-O(7!+m2>kl2Fql8UYy+01)`#?0pb89Dc5jlwrN3RCZjMX${rzh* z3|cHofKg|F(H6$5KIIRZ4t2(3+{g>~KBw!-81fkXCH6l^k;=`xd7&f;s#oeDS2 z*xR`FL0 zXNtpPX_O8Grw`+4K$J-a$38v~Y`HAD=;%X@b|@8G_9qIkiW^Z%ZMW~cwD*&1!k zzx@o#eB$!-1Ow{w^oZU%<1Ir&RCyhkKcM zI($%eK6?QQ{{M}j(ojdKLRd9rikY@lr2l`yQ2E4qSK7`wpnM#hk@b~s#-Pg)IVmjsu21TP z^l1>nW;q$6QF6Du94T$}VA?UDTve4MH&0>rHyv0XPjRb;fsH8nJ#9fA7+49a-0iTB zPjjqbjirJ+@&nrktAz944Vp)VXQ7#tU}*B?%T5{Xx4yX)!2OsfSag)lxS~x%l{w zF+>wG$EZ;gFfcH9`^GT{7_Np9Lx~ODie@{jaiH2dM5>)tgU(l@U; zSEM=A>F^1bJgyQ1VP@Ix+!aSh$FSM$=Ku`UTwTeTCGGhyjJ~-RJn(&PZz5s|2!C>l z!eQxUU7ExX{kcdH`2CO!Ez*J3Np=@wk2B!FMJ*x__9(-YgTO<;V(hy@2!iy(q;t8F zCk*&DI`7ch=L<)vJ{nh4cdxfL=hW}(IGHy(Gprskz2ukoE&v(>;(4u0E*W84e5P9x z)Uq&IIXcv=i-V;#;kWf_n^iwex`A5FB=uO}Rmo4=3yvAT)#(cH#i;x8h6$#Mf`U83&&FD z$Cm1rK%lRS1G2oDZligL`()uk?55-fACeafT|k+4Ww6G&V1h-tHiO`uNa%jZgdg3? zLIZwfPtQN+{oUT_U`R-BD)Zynl4p>_pOQUapWAhFK!@*eE=~qFiXAsO-E3BDj0>cX z{y8Q~8HTbzP7n+te^#I zeLdhX0mN?d1wK;Kj9o$GbjGV8-8Wy+!4z z)?$O&>ILq0a2o-)p8q{qDg)Xcbk4DrVhU_jrAvOQ0#zlUpjiKMQ{MPSAzpoV_YFkx z(k0SYS#0;$tV<1--85rZrD(yF+6<9DmlYYbCGWGon^fR-Iw3Msix3LJP1dDKyvfRz zA30^p2+V}?CKt)>?EQ{|THl?+^UU}#8F_lGZ^M69H&~c|)@-_2K`Nl$-vBbTtYA>?oiHIvj zwm&1{zfJ>x5dn#c9!d_05enRfWc9nG9@G4ASv-vfsFgwt0y|ExQ+ZE7oBsI86 z%YoE@hZoFOZG3U;+@g_~K0|bpLtVeM%aA(qxuZj<ed9pF@Qzs;c)Yy1vvn zr2AL0A~Nml3=FZMp{n~!0S@Ls6it{CQoB5N4)TB3Fh>1Y_yFEPaH=f-Fjprq4gkdm zgsgSZ!FFiS%q%w`uBml(H_Se^Jam&;Gm6Q>CVv|#IOaL8hpml;{x2r*bVBZJ&?m!( z*gn79DYG=by)u==hsS&NtRg_Cl|h?AQ?N+uTEFeLR#6K2G)X7Nne>o`ojtTsS%9B` z<)yI`5_VdxdiRFp$4B`G0-4sqHXH{Dbz2GZ#_sMI=v<>Yaf0kwyTZf5w4K5I;yKXl zoxttFf19WM*ztS;@RbDK`X%EgOwu2ve4L@21fsc;_uO2K>AShM!($^1tDsB?5t>h=y z#iLW^v*~)N5dY!@pt?cud5o8hxkHgVdd<2+-Hzx2C&&-K>FF< z_Czg@nzA={?5wQG?jxjJN&S%G40%GRwYVVZrk;NgeB>=3g~cJ_@B5S2#$)T5yr1+N zO%H`zefnsv92ID~i!2p)FMt3G5{$+*@VZ6t{*6{(;l~zwH^seW_14-1%?}PU$xD&Sfu1cmW7e1nLX~(?gGMV&gHY0_tli5?W z*WNaLtZ=E}cCxJmO`hIi=gIB=G&y`YFlQoyLN4<)F8J_{U*)9-AP@;3i*m zu5h0W4JUe4ZZKQ+lm5Sck2dOwXaM`sW72PuXY>U47t~*<2YUMD2a^zhz z>(lpL5;v+Iu*5Nbh?*k$#g8TUF`=^zrzX&$GjYs^zNqIa=E!qY9RcM%o`fR$B)!zb zj=!AfRE|IiBI45DL<8QVBNiehZz8{&n$6(Jz>m~Ot>^wPGrb4NO&|5jF)FX;397h)m zgW=4uB8GN4q(>iELZ4TPIFNyl0R4SrGJ3h@>ljV}LKXA^4(J6)!vo=^|L66hR8K}; z{CJ%F=z`I3!7~Yt7*6^ibac%!Fnp(BSw)tAW@Pd-ytfOye~D151+`EX(v9u$uZKI4 ztVn=z!upr^Xc!LYFD^C59t0j4q^g6^t7e`=uNdkHE^YbPr$I;KE879}0sOO2GlCMU zpCq*CL_ndhTy6=Blh{a_Qy@F554#QA{ZZpnrvATiafPd8tynNt$D*DO81B&j?B&XW zGY9Oueu62~$ZtdU5#&2)DLZQEhBFbMh5V_H1MDy?yAL@i@L0HI9NKamaInGn5;8CI z@2D+}e~!zDOujTJIQCnK|urK%Z3;)KXAOZNmy` zAqFf1l>WX^6_}ktNLX#)K-T{bF+wZPk17v`4Q8VdE4Cx&i)X4PVIX~lHz2qSFuVqX z4Wlglr>&Zyp<%^^qej3T#)(@FlhVeJs?{(h?aHlkc2rD?wykTF^Eilc>>rBL^=-Z{D=@gG#N$ z-r<$YW%|5tE%*_Yc@i@L zK2IC@pp7mf;iisIF@=dOGyMC(pn9+2qoLitT+2y!%8V~&YrAWGr)TAN&;9pHqwyli zn;r*V?BfE4AM2ryg#fFqt=$7}=JWKbXd#=h2?B^N4z!mwENU5s8ZQ`-i{3oHV`nya zEEe4nMl;$c6l;vSguXmZNWjFXH4*#Bd9N#~suDnnBnHh)(5MfJ{27o#P=Ec=aRYv$ zqN3;-1F{=^+9~2T42TJ+7a)zDM!y3hIVGG9a^)$1Lxxg_uRDbz62JPjzS~&g-&cMN zSzB9M0fQ-7tW#VaZr291WjYYCg23*LI>m<3D7pw?FkJ!Ha$DLkm(ny~4@2Mf_> zJ`*{Ls-sF!jmRSZCuXA`oi!i|e|S{Im}En5Q5^a2 zzwV;{!iox`wU7UZcmFXs^o#2M?vtbNR!4na0RJeEqLCPTsb!PT8{PY8{U*))1k=A?J(`%PP`~JMsX&QvrQ%kDvQFj3v^3k4f2!zxp zPDFv(oe`|mp{gf&w<>wH-;6`^L3ZrhH*df}_g%by*qVk6PmxHC>^6n%`CIevFQG%t znVq|fc&`4sgs%6acZU;eGhRfr#!AFgdBN17DlQDb-tNSe(y2G_p7`LO$2^B{m{kA$kOa=XQ{8>yUx;GKq@K9uBCYg1r^l=;7FgXn%;Jxx#|JD#Q`=I8;#tT?|}` zj4C{drchHr?+elepIc#lz2ecCIUb97&OZOUfIqlBwaFX!`CRw1Ot6(J&v6Fpn$_yk zoILqmy8j-V|3}zg!cznf7##Lt4(I-S{`^@H4#eSSAd}x;RzW*1uMgfU_S_12P&!Dk z^@}g4R9Y;c-QA#zrYbXMs%d;C=j^F9Abq3HeVm@23;YVuDup&!KzGes!@nx` zA!0E-ibdkxk0OhPZhsXb%*#s(tz7SytI-A)=mnEF1R-PwEOghdN#wU`aBp&%yv^t` zD{U|<{dq3*2=2+M_o&E?r{G)BuuAp_AI16ifq3=aij7nlGJJVvq9X>Pdjm0`DnEyZ zc}I=za4uaHI46oz(3~XY-~CER7`TBb5G;uPx?+PtgY#>?vq&Cs8y{ZgKk+N)7_?zx z7nNEI#d?S=V{uf1 zXS{amD5&fFL5iUG9m6a>$#Dp)5JaSNsN)n$(P$^wY*+f0^(hwVsAe~k6t+3>=lo~+ zMG%BvkL9Y)arBv}sdS_fog5uFng$<1Z7DMDum!kL*C0 z+x3)?cVa0t?62PSe?oDikt1wesMUkaCIR3p6dD|v%0}E-kb3JO?b+Q^nKdM! zZS26fenG4$ERg!?b+d)tSsotp3W|ixwn3Y-Xxg0%yG$I2a3+2kCh~$_IFM=#N?9(3 z-boI+NmzVBG2{zOS_gvj=ciB?`t?fp&e_G^Kl3)3FT=yfduPJS1Iu9(NaTdr&84w8 zXdyNSm;`|^e}GxtAUh)PCpl*m(qc##kj8?iHxzf`>Ur*??9j`{hQNESLVHO#;D9eQ ze!{sz#0Ywz8`|BzWvFpk0H2Qaw?CHZg&#?|Di$|(?^C~R6Z*!qgmk45*#Mj@+NGxe z=QW8nn~6Y$$r9X3TWA|8x`YZPK4iMQz<(Eii8J}B|NJ(2Ml8IB2X~LMX!%W8HGtEB ze(XeH3uH+}unl{2st-{^E;eq~1q)u;p zF(wd+3c#X^eE#wUoEMe^WJ)%rhLJ5vwF0~meSbE7i3S|I^#{3OS)`aN8Djba`!70s zaakV&$J-r88zSK_Jlw&SF}RPQHw)S$0Bl0lR{Lo5?08Ry@D;MPiG=xdqN>gCKN5-c zZSDjh|Bj%r7c&lrj=%v6c(5E!0NL@J&o85ej~7)$%3lI#{Yw{w~mUcef!1tFar$3kkTE3v>@Fb z0xBXUA&p2$2ujJ&T}lcF3P^XWbV{Q%h=hVli2mRh!=AnO zec#vhsS6Y&je)HQ7vAe z*8;;UPSnn5r z1<=?;ucD$NY{N5IVIB`oM7jqR;@}WjT7`;(z9wjPA$2uhO{U|*SpNGv!`nbBj}L;o zG>ZNq@&&zr+4zAET(HkD0j@@6U0rgla~4tnX=3eyXb`Rg2R_}{3JtISTXSDrC1#lV zj*YY&B_X2%gOGatIg{y$cM8`=L`|G3XVP?~BeADTo@L2z(#mct?uLGGaNtr!=NrfJ zjDV&NL3j(U@V~3bWN)*7*+USJh>DA+6Gep+GX{OI;aTZ_yH<`2F~Jvxy=L4+?GIYXf-Zp1W4FI-ffejvc&p*!lJ7o9kS7oV_p$Va%2e$KNB>RNn{r1k_DH zY2S+lC1U8si9xFiL0H2tu;B`WL_h$C;yi44j!n)0zF|$*4d6?LFw4u!*D6+8-slz~ zdYgK`nVFCkjZ2$+xqjKI{X%j$T8{S{xkkQ~NF+f|Q24vO-eCAVKE@zy*ZAc9tN*E< z!}i*BV8MY+rw2Gl;yAQY7J&_l+K(TUDBCS30zUgczt!Qr1?4|EU|wv2l$r_)DVZbK z5|b9_8^j{I*+Rncz4u@8=wlYEhL>q3iS)TY%bDAx?Yzi=Xls3hDPX9Ih7tOINe*CW z`p;U~u(+1k0u*^WpbBFV7pG27Plqs#jS?Uz#JA3E3wvjPJ-X`|hgRj2T`S;X$F57j zfvak3Yo|{EuN@8q;^7dRdL?vHgtb-AbBJ4_N;H<1eCEXI@VGeX5eWOB#XwM*?*!z%lkW1 zb~WDbxjZ~UQcMmwU4e(*sKJF1Jfa{u|369IPl4AR(JYgRY~f=5X6UL+}tA6-%_o-xS873S^EF6ehzQ~uPxSS3?v?5ApL_4e15^rtPWr! zxkGKXo96MtSipn1#$%uja>CPjXp_49>K~osa`6H#1^^wvPP+V<5L&t_Q>xIcOAs0{+29Rr+!vpkzs> z7@uIR(m*64Ez+u^6~5{U<&l%|8HAPLq|~hai&2!zpddh2n0?x0p+xFwyb?9)pWibC z#5|iyB&cj(>uGZ3{fooQmpoQwA(~iuTpw)gsAjJ-vrc)l^_{GH^ly6#GL7}Wph})L zsKPI08LR6*xjA4&Y1~r`L8y~q7PU+NoX2Di)AUQnLZSZUIg#}ve?bjG6m*sp6!sTB z*2M!qP>AD?swn?On%w}lll}T1Y^NWnPP6|#AZ>RGQH=m+NPb(Sak&E=a>e5;%Z&3u zW1O=Df6I6Q@U=Xp7?U(PfRZ$LbaeEE%cWhMe``ZNKrnzPv>w3qMuK>W+<=R7ELjsq z%!mT{sGm03^}8qojhxY-{>>w&+q{DP#DQE0yYYqt)C`-#3Gmiskg$OPrUhW!!*6T0K(aL+V)Qw6SA^%M!)6z}e6mactS8oA`x3G7 zCO{CZ|9q2!>6s>=b5$yZ&@p!OILV%$*|0_QQ4~*pD7FH?o@XbU4bx2?S7vJ;1cN4x z3Vhsx3v&@$q6Q)~lNYgj;f+CevdF5#qm!fVVQbJRJ@2cFSi3hkU9klZ3WWm3dunjN z4rDm91L41Bj<_+s3Tt1YhosT*I&I94PYf$~wdKH?4t}>>8A$R41s68b7Wij~z)pyb zLIzIt7;sFph>6v|nbjyRWYx#H|MdY360Z^Uoc{~29ePthPY-79W4|jl3Km3C?s&lF zK-~uJ@C3_41z66c1s+72N0L_#mim7?BYh zXv9~%LWONBTCh(OXo(7ccqP!Wq>2G1F_AgD#-WX`lf!27CL+m zH)=fKq6u^JEg!aT|KW%$XyOFAxaEy~rPIfvtzsD_wx%#-fri#!{81Og8lBOk!rmvE z;WUp$4a)G3OV!MI5UF{+%+X_0a?jub`qj5j(fJN&0hq=0GkbAtOJOKOha=~)8q+I2 z_~Ap5yLH54H&vG5>R)w8qGzd6E|t_#T;FuN)bW?Tev_dXe-4c{K?o?SZa?=kxLOUf zT`mZV#<7Ckt5V?a?&V;x-l)U%jXZCjD%Of$y41HsA-v0ODSucrZJI@s^*M`DJ_4dh zqH(Be*6-$rUCK#Ncv$;7;-6ja+b=H24;1qbZyA$+0srYDHtFcwKq?spMF>diu>x(Q zJy;}7Q$!#jgO-*yKQ0?2ouEQ5%IK#qbJ7fC(?V57A=~GTAvIHkiOCfnzU5;r206N{ ziYmwx>Mnk7HHUt}(!(B&Y4;Qid z%Rux502^y;Y#}zlB1zM>X(}G%`^L-tDO^l5z2~DbZtiOGS~L!cfNUuBNFga(Xm4Vl z1KwwMb%kvP0BhDcW9hHlg~$IY?R^^2gFq+=4)Ab*q3tOT0k=Ak_J#xF#{-bLS2n+% z`}%c4W~M12|2}_Jq7(Fl*L?k89ZvvC(jt9jC5)kKrsb@u8_c)%!%cYu6X7GWByaG^+UXX+dgFs9s7CF28>{mDsR^G>_W_Di>90ErOU>5Z@qS8AfvPNx!B!#3Vprszu^xs7S5P6ULJsI{XCcQWeqju z3GB8~PHPSlt1vQmF&X~C)rFsN-W|f~ELiJ!DSmFbOm3D2j}9>Ux}1ujiiJ$x>t3p` zv>hKm`OXVa@&5y|-Vq0{4XA}E83q7wY-!25yt2aUSpo93v2YjAw*0ocAFNA!wk7 z>jd2paimI+I+J%^`P(}sz>>~N4)e*SbSR4?Nr9H6FwoIw1s-FxahWxoUx3hXuGVi= z7oDybq=0*pI7!d{%}foH1S0mJQU&cB4-i9Pvx|Xgn+$*(Aat)so!LKR$F-NK;cuKp za=?7v^-EYm9mo&Fu56IW2YRnikgTj%g3~CMc)i=6f+?aDhf!>^B}QHSiXxlc^R{^@ zj2~qGg{aG-TIqfzi_VYpT;12!@XEKuQ1FIwtLaIson!hGwzwcN1Dl5`2xY#R836*U zQhXNx#es-_{`|QXqhrJt{Wx9rsUKbVuuZ3v?M_$LCQEWe)-}wAx)L}VG`K@2Eff_a zVS)V?4TrpAlWw50H;0ch<+#5~F)G#0kGg}ce!@bYN{b1ak-y#@pC(R-P?hkJ!#}wg zLO?W|T18x$_5Lf>7-|iad5y!2tXM=EVsB!g^MB5 zNe!__Z}$-U-e0ZJn%<#R2xZ4a)#78;1KvffPGq$Y|`GQKknUx4-&xu zYogJ$Ar@$w4K9b24U9;DQ`(tds#LiTB7*!VD?f6>Y07nxiL>%z=L z1aFB{aIw$oqYM^|>ykx5^o5HGV%BfQGAysnwGl12gCc~b=&pncTaLOe8fvqt-+ku- zVR|gW_HrIxXB>nBtQ;pU->H}!a`Sq!x&I)=@ zo{Fic=MAI{Q*f~dzZ zH+9=DbT>DX_B|MXm2E<siD4aGtP^Bnv z*5ONa`#>Eyq=qiS7%9uPg0g&c&V*fOYSCI7L^IF9r2I~d%+GA=8ouC&oVFi#bK#(` zJE0~%ejh0#oki=-s}co9)?FmNmhSFy6F)6CxHA@Lxn!Ez)IxpLby%hJ`uQvXK z&#!kJuRM{`y=Vz!_A)sU3|m=+xu*-*dc$)3`x?7$;0X#`BUPC3fIfMM9<|;g zUOdcwkl#}OZPC-|e4Xy#W#3NY=c(#(2tyUCJnyo#n3j<8I^9W?Mq(0a*5Jz z1fn7o#sDQE)YfYF%d&3l6#M?}Q-jq*BTkYq@F(-!4zOs3w;f6B-qh3Foh2J^>!@J_r3KEBB{e_0KapMLb63ir)*q>Q+~ z31j9*g~Cl%)pEsG+sGa$s!@;V^)hb zTZ=}t#SAQG#;S8{uL+v&pXH_PzG+_b3wF2a%?7JQ^hB7)6yZ_noTmVshJN|LPaQ9T z=x1g_;k)#*z!|IVkIXlh-pLEVAOI4#=6L4lh85}@(kiT-XEi1n+WE?Qf-=?`YzKQ9 zKL)wLqSt-?_X?VPO)YX0^y`tJMU)D|m@DlA4(Tp1w0((x zv+{ZS=5m3*p=H(Vk(rdShyXkE$QZAuNAUETIONLy4sk``8@9?m^%5ZxqwgHKSV398 z?%eQz%&uXE2Np8lC*g;2``)F!<=Qz+RYJr$TMzUn8vI~g!3f=qWoUPtkVX${+nlPZ z1E0%cB4Ysxh-M$GhqW!$%z#iHGN*(m8-B#h5}zyY|HAwNH<`{qAnR=?{R~H?5iNo{ zKcaqQ!lQe?)26=Nq83$O@-E#hsCC3wq{bncmvb<EZ62?XpY+98U_?&FrEca(+fM6bbnk;!oRgitYv2_Y zG#<83d!Yt4XDZW3bdJVlcRt))(Ou^66@9Bmc+GpOV2iEKt%lQ!WvZnoIHZuyYEM5n^l>7v z`BZ-NfCdC`U?HI_rnrI)axCJRgSdC>qKWQ5q`-mRUakx@2TdZjbz8HqTsqXMoZav^ z?tmYi7?+B@Kj^juf5~WW7(6{{FtI|vki>Ph<%hu~uEMRC;B8P7#SmSUVy*_F){)F5 z;HI$r#jqrthLUp-a@aT{B2L=#H*g%9;6b4#UT>V*9B$bPVaT`8ZdCMyn|V~vzhU{zFl4z%+H28n)Rl}3(V!!cRm zU1q64OMfWIS(GE1o55Q0v7RV-oJA=AC<*4Y6Zh~-%y(tx#lZ*0)XzwD_xf)gleQY9 zOJXn{c%$f=&>>yd_5^EUmJIF{6u(q`RLFFZk@b_qH%IHOe0jDt>CLOL`h5c9Guw{K zhga}0%~ymIT$njsXiJn-BFL#fMvcDGw@JrkfpQk36-rOZ1z{*|pIS!*DCoNEETo|K!EAOzZy=P?T6g-_*6L6%X`~HN#SX0Xbio8V7`IO1YZsc({SbN3TGvDJS zz7e>2st09rjmu~`c%d+RgGRi`u~dXATW!NgoyTO=4T`OwQ{)6UI!9YUqg8;ocr4)}`AXB`Yf*Nw89WvO@IrB8fNUawnwFHjYY1L?6VlLc z>r!+gfoKTTP?E&S^&<6Qa7c&oEQJ2ayN4n;8=5yyymni}@ z(=()DheQrU(f%|u-u5#N*6`x5Vz(u{)6JAC-*cGHCu(;6gWjP;H?DqyQXxSR4a1Kud%3mDHs#)e>Qt*uJU}e1_RNf{vh>m9k|XYA8yDO)ibU z8XpsAYXfOLvXhG-bWb0kwr0i~=Ki=#t5BZ3Ro_#Brtd`7_vNx$uvu>4s!kT)PX=LY zcfh;6*vM*Hq=G2QJ8i{p%)z%PakxV4p{=(TrtPUl$UObX<`COcwGNja>q5)NQl@bcm^B~S? z@538o09%{@oU0nEm@5Jxp_nTx%!%dB*s1)1DeYw|GC*Y>5;?_ITJ-AjjakCdJRV-L zq5KZKYNXV(5l4kA@w08V5aH$?DxaK;O(`h?MWmqgsZ#SGB0oV3mUbi?W)_a=wjxu) z`6i&>*5T2(PvAkgKRR8K(KHopLSB#;MXNA{i3h4=T^*p))Aa>s^J1fd1 z8eEi=Kv^KzsC8*xy0Xr~x%KC>>l{I>voyY|Wm2Z8wQkg4|9kL`u&~i`FM$*mL2Z4=nPQU}Bn~dLDjd{DHJmR(*bSc_2E+ky4Ol5PCa&VIBVg z0q8v&zt!i2OrC05$pW(FJywrw&hpUuA(xaK?lq&Ki=!11Yz8tApkR|uv4Qw``T6=! z7&`@zPWMKzI1Ly_0z%HT+Wb{R>Udz{Ul~(p4sF+YYG*T1tdC{iu@i|vtQk2_ zNIwC9xGak4`jGBSFgMvW&*D$_VF<4;-=)M+nU@O(W_k@1*x~)mlvTs)6>-C_8Qjua zrv3Yc{&d0YaGLK)u+z71Jfm{chZw!KqqOjr{2s>mC#K15DWf&32INmq@33IzLE{9-vrK`YM zcK(H9m{$HVE(>^mf^BD#LY9H2t!BY0F8pl)SJ(D>KYG;l@sG31ygz021cw7VyihbP zF{K2yXw>#1_wlT#O0IaPeYHWOCv%ZJAL!u-SnEm_Y$4$81X&LV+N=apFL~o0Z%15e zVDS!D%WVgm(qn5dvjIde^tSk)IFz?{vEQ%`fskRsA+>R|8XrK5Y6A zN_tf-^W_W|Awx3YMmb_^4GEu9o2K*M8f{mr#K_ygK6!{7d4JJY^s0GP4e(h?Hp0P_ zU#SSTkc6K=>A46doM9&sVy)^N92{6w2Bc15us%w_NZHz>PG63(94DfU!Aq~Jxf^2@)#3b3vZwyMn_$}a&MD*(OP|}29 z^T7dDojiMazaxPI(%A6b)#RCe-XOwr(ZQt&sxDzJ$dZ5?D5 z3UUBWQJc!7$fH9kBs<>?Zn9EiescvNVdKVknzq?fVQ7M9dyc3#YxYMK)DLb1tzEkt zqBTU}5OeQjUO#2s_sM}KOzgswsGx#!u}=N1uOG7u&HL&6ik6)ER%+VReDhLH_B>tv zSh<4Mi}$<#+eT0V9s}@0V$;PIf##$aNaJ+$^g?dQ9V2eZ>>^r#K#B+7O`n4*nkREx z|1OXIel&}p*^NUXo_V3iSNT-#uKGUOPMCB3WhL;|GlIX4lA++bHnQA~D)+im;i6pC zSR6*5jmpxff@<3Bh_n2Q?#lNjKJ<@CG*oGOp>OleQE0QA_>}6gv9H?ImMT>Dfth0F z6~UzM+Q{3u3Jw?TSKNVxgC^I^pAbaG%80s57Z?JTn+}<3Hr^Pg9*W9H0 zh0yLq%wLUgWq+IXVNuz(>J*!2KDzGU#nH(K@HYIHO)1%j8WI2l*|0&$Ajkg58kpf} zKngw>=@|pIuEyq@`MN-$*lsU^s>T3E2e2BydZnM?YgZ$rQ{t)rb~0IoEliUVsxu~W zq@9BXTP&3-1`EQwO0VN+g(nnI}EKaWklZ?g8b z;3GP?R6|_XzKH{abat={!PqGMpwZs?1Hh3=)kT5RS1^rVjAqm04?e?59|X6UeMb z#fgF(a}X2(^s090Q*Is?0Lu;o#L2H-?A*r2#zqHQVt%Uar-Ee$8B^Oo939@q&mTBI zCZxJqYYvRmQIl@3ESxaitBvt{TWtK+`jtjb_LoM^S>1N+SW6A)%k2s+>+Zi5nCDU{ zF`^tlK$(+4=_;lUQ{;&eM9^Lo`czkVSijRa7#Gkc@qd!^+s@r|UkfW2h%0Z|$#xYl zFDRE=-F;ZwVNtttcOdHfrlL<`&)1sl!`VOgY8wY(B0nGfPY?aPr@I?F6fVdbuvgf5 zNPd2P=~E!j1sBK=r(67`0|uSMtU+KbNc~cXtxFvzTDm4ET4NgJLFrava z!E26RHyKjWFn%hTDm=|b5bJ-=+>jCB0gKXOTjqToos|+aMnA)6=|>^W_x*>ri)Nm#rB-onQn0YBs+Szi9Sm!);pdjFbu*ur$fHDJu z2?J^~R?)uawid;_mvx8*My7elJ(3E$Vz(g<2uw6f9I7#-l-?F+pr#Gc zIH|1IGsc+K^DQtjP^HjV)+9G?kX>o$yM3b{HyGe!VPE3&SNdi)ss&0Y>OMUaeTkPW z7Pb`LK4FZ4oCRPx6ChBJHP3G19qS?x#sTq^HArYJ-U?f|m zai~WLcHDpGPzIxEN+^ug=I9PA04TJ_6gK^t8%Gz&q)Bz->Mu}d+kAcC#@(D|UYSY%C;56ec*k#81syMH2qK(D zQgFHRktDZSDyA`hb+48opl1OmX&Yz&)7yf`w>P*=my_C8>Ko@yr$oVbSDUl5{65j* zR^*5lAT&w+K8SpL*PuYL3MGsY6W%=3Mi$*AI=tnwJY}OR)Hq~?<0uF{D$tjaX7H1Z zE{k)fAd84AV~0V2=&oF|!Gl=jW~&C(Bm7O8-UziLYaT`$e&vMIu%!M{?xGxmi@Ye+ zkscZuFr?SX5!+GaUNF+0wX@cF{a6NH{QVG-xyOu}VL8k~nHkT{6s~|3rX;*jg|X~v zu_Z=aMckhJ^|EQWvk@BQ5koZ0lmwGxedwq&;Q&o_WA1&A^GtBcS-Yk_sdc6r-j3do z!PmJ@0wtr0tRJMIa2jP&E2s;xJt-lRVJa4iCyp_0PB%09%ru%z3DKFs+rsW4m+2s5 ztk-ay4+TYj`8g#-)gVfsYO1#>aE*^FdhAW4wb_T1DxTVU`Ff)n`v-azt3f==3!C)T z{hN`b(W*#7yLf35m^z9E3Qo-n6c4v#9QHGOunwJ_aR3?x@_x#@maE3L8k4yR?XaEkkQx1bZ{+)WFR! zc^-CUF(3ck@A%$00D~xNp}kTSNE= zMBMz4q~kkZrI9;6mamtWOM|^nGYBb=n~(n7`^oXFNmYPWp~v*F0-v?3^){rTL8RdX zWEkXC$RVrgI@mi7HD!cFM}k-o!0U1G`&grV7ExI^3c>zN%b~SwkV0JZjr=f- zKMVZHk{zqI9sTeESLyy5ubkk^6_n8`$T|`@^#ko73vLD0rN6*9fa8ox#Y$&=s0`?XCl7FvJj7AQ$4_Gc zde-qQ?H)LgyQZbcg~|5s16vlir-P78=}>d{O`S1dq`#j z!Nj2VmLxJwPMoP@UwP`xgd;r_tt%|N7+;nXnc|HCD6V_*ElC6jY6p%wB|lO_mSqMA zk1G!Z-he;$6zz7f8p3^zficLUB;Hn6=-Ec+2&XTr+M_~pGM-*KU)@NWvX$xANe=g+ z&Gvgvmrd+fq!4IG+(V?V@6)e%Pdvixi5TQ<)Ycs{s*+gq`&HGA98SMI6JQj=8y+t> zHg;O34Gv*K|GX>_B+DqpC=pIM_HXYCgv3b%3v_z%F+&@3U3Cx1w)}WiLZtw~N6Pp{ z#fIyo@hz=FT?&z2UUupu_{#FMT>^Ar z(>#s;TFm42LcQ_9A?3N}Zqg(;4?bJeR(Dn2XQbR%=gJ3$8{JN@>2=||rB{{WWLXpE ze2U`~I*`NwK56v(5J1f$8m>O!Cyz|98NqRnRs%bnIErNP@H78M8q2yFP1yi70s5}| zy^E=x&c@brPonJ#dE$)NrD*Yvpl^Pbs0LND5(&!IOd_a-&e+kMkIv#yj99mNs%e<9 zlrOfKpRLjZX6P(=$AQZzSKNBxe)!I3=2S#4Z5UXT3O{Te9OzPE6udLb4A_oS^zX~7 z<0q0aG{7siqg)Jg0V`ja{(ZL^o)V`gB7*&`xa#^7N<5BUIFwC>z@+Ucgy4N(C5vto zRHzyt;~#;d2KUM-X4Loh)gk4A2y{fAV8(`Jp_Z^X|DX3@gFcxBfI1vd9(k zjU!KbvaCh!x(z%qq=;2ab1T_|bgW1qTg45jIQPGI89T?pPw$d1)s7}7J;X!V@mblb zgK=*Q9Lt8Fi5a<}(KC<#?0CuC8{f3*rgY zGD|S99Y+`r7QLg_ai>*;f|L?)>H?XEhx(^b5h+F53%9BcxIzFugtw9o*_(*&(Rdle6@*+slAKecDu4m_?4Zj_PKq$~y9Q z`4=7_B$R?mo6!J5PnZ$Lz}lU@)v9#0+2?I8B3VV%jd!_265f(BGhTkOAp`bGA_}-?J zt{RvabM>lG+bm^8Ji&oP7@sJ@$(p zT^Mm1g$KfH&4IIFLIISk;LaijZ}*2Px5aLBmK3u{Dg{#(mi8Q23|lcf05&^lLrGR1 z!C-4Hz4JYk0r0sWC|~^EWWaBVIooY_en`E+3IEOqRGj0;RY-G`Q;vd-EJxVbmE%tA zC2;tksNhXHN%!7ok__cgV|bT9n27v7E=3OX0iuDHG%YBJ%Q|s87QHulhBM)98Ip7@ z+ryLqF`{jAR6k_@I!MykbVk@+9iO+W zT8Y4#!9*g9yzeyQY1Fg!CHV+MfCAz={{r@p~Y zukYrP#;-v?qo+&nSIP}+PL*_6iTw(Q)UtI?V^M~*D`|e;^Ceb7x`AStRZf#Nl|_mV z0)z-l&LaXyIhgXvVUe7EORFAR&mx6adK%D-3cR6S0VMQO##R9WP=>k@cR`EoWCZr) zj3H06rY((KXldGn5F0Usq*m;5PP~) zp(_xm&EqwzIZ}t3c6u010bbLo8qLwt+Zsiz7Rt1=3K@}r&R3u>}uYGTV)lvtA`o~ktB zVH7oK)qdOe;CoRL!K6pX5Yvz$A9=DAi7Q2+E3G_n!(+F`J8ed#LXs$j4d+F`se3?w zPYOoks?Of2jpQ~{yadNvcRe6Wui7iyJ^F`hde6GM>c?;n&P1c9<^H1Efb_rI72Yj9 z1bR%wKI@iS<3xkiii?#>Mm9G-age!eC+ z4Eio}3E)||I|H+rp;b6=dlF%doVKcCEl3zj!Nr&0d6rKumSI?G3+FC(2WLpT^Ns0m z2EhI=Xi|Dg7c}qJQyrzCLZ$A@hf8Z)-+HaoP7JF31!k5=&1z%K?nRP755L?QLfiI& z9tRHtuEFudjk3e-;6gndcV}qbTO~l6t;dJo5s?LnDvnjNWRk}Y+tB|hOLEV$L z<^TaL7TC55F1ZNVyBvvw1q-B_L2w(;ErUk&n02vG$7?3}8eE|yS5`OnF-pawxsE;AcP*>P0yP|%;sglBDDf;NCKe>2+WRNu8C7@NRZiquz7YTmPLWl( zT@xGn@tSKtJq{@ORXNn27E!f5AkH~EpZD_q(X`<%G^|*-ZZ&EO6{3=W{c+dv${Slp z2HLhtpp_uc$i=AI4)fNW*;W}jOZbbLk6xKV=iC@P_~3BOV!Mi7;&F4%#}SY%LHUZu zLdg$IgzmVwpBHq$5Tv?c4&@$Qm z4y*J|`sX{^IL2*BZPTWYL; zC&~f*@7FL)lP9Vo>kswMOcQ(g?_LmQ9Ei|(J}Q+MuxsQO@6;_ZoUbE0hW+AYht#a5 z(<~vVB%MgE34Do;`U=4_Df04dr8oTKDL*cF_e43)W2+2bMt`m4lGqja6RCt;k`-O;QPY8I6BhWY^*5*WXwgY$7+heG}rMie&JmoLhsXd>;a8l8D8=Ew_s`l&n%_ z*vn2+s^HaC$bIip6M7LeVo%Fdr&uip&L`}>c1^y%`HX-xXqMgvC6^%#-Y}{OKajKf z^ti*pdd{h)^00OWuAtU&P*k%oqSS?VSyR~cc)nXxHojuu;y!-0*g1(m%}gwMk7@zLFvS=ErCUn zIl*1IjbV*-(e)`Y3*hC#Fn~?y8)61W_~}0AqwsQZ#e5Uqa9f1?+M?-Ie$pH;{WO{4 z{aKCaNFagiIfuLRg*Tq|h@FsD-WbHiY&7Sy^$NO;r3WYbFqs#~Ho+CU4HM3{`(we!Ny754RCKoS;lR2iIG)D2xk9EQcyv$QeQ7``sYogk?@yA$?i9>7wAU1lR z@f{zuRgBsmQA0|>A??K$#4N2eBJ>u?t{%`PZQFDoi$+(RCo(u$_i3vwQ2fuc$^Jf2 z7%9CUaNOU%z;52N1$Qd5;c!jJkGN>~(+URHWP}p&fG4i;6X`(2CdN*C`w1OT1_!hW zqJx|^){8!rG$yR}6=futoe=t-%NcHY+YZe=vbCn+JK3dZ(cr&$+R~QCy?Wt1YQPF< zw?@su&pQ>?VGqTkj|i~EUma&5IgBM0`!W({Q3bAsG}%D7Ky$){vos83r4g}@_-Zau zu&m74SkvUKCGEg+R4Z`7d@~HuB;V2y403n&!j<(Bq!~h9l6)H3TXGA)-+@)thc+9u z3RRh#4Mc2(YVvSicBRrrXiDJTucIcv*fI&i82`e#(oYHkR>7hs*Ih!V!tOJ-!xaN^ zZ!9ca8)j2GasMTZ@64SbN(59lT0)`LDV0Z+M9I6WTT^wNV+QyhUSPtFVoevA34 zmMaoQ&mbZr&N%4CF9-S)YYAv``a->eef*;s%lV7f{CzFBMSXaL>|t@Y6*wkWevK(( zWCg7@23ke5oCqPo23mxwDVuu||MYow`3c|6dtL%A_nPk&IcmDR4EI`*v2i*>LpqtM zodGwB2P$@a(qGD6>CfaSeCcKIWEZi3JRy6PT5s^AHPesirM`3T??ri?y#395FEj^` zz+5*xM(5n^CC<=N75lzlTYi}g%Xfi^nqcnP_k&j8>Pzf&XAzfUBO9)59YIOG0c6&^joM(N%u4Tnloi7{0qFyVur zUl!-yvf6(3JqJ5@cF!}muT)8mvg-Gq5kL`Q`G!wgpOt3VvkoasdvlJ5c2mAIPt`V7 z4;aET%dcqo!7xSd*`n{-IbB*wWjFrg{@Yh2j|bY`AIv(A+J^)XlmXW%uo@|WbVguZ z2I&x3t`@iVtNS70B)QW2ktSr`dmamgG~g^!t3-8lufb(eErl-86te%=UN`|NoGuMga)ub*vTc=`5_vX(y<4W| z`tkMM0Wdbqb$MJ1jiPfJ)bz27)LjX^O&>UZJj=(lI8zg}Z;(II2sra87(SR!!G+v=;Th@l=^vu4cBcl-GAn(4J$VS*eM zH1z21xcy7|F?49_U=2Ml0i_3H5Sv(Z^5YdctK8>WS)FOqi(KTYqxAfzrZ*pJAIEuT zybW^7(Jxg$D|ID$wxfVV#8o)sOck6ih`Sm+^`rF_gcD@i3%{Z<`sIe8_aBa-_y`lgOCv3btL8^EE=e2D@qJmPZ!mZ^%M#&!Ys}K-Fu70N^9@G-q__Ixhf)w z6W>va7MxXTPmO+Po8nzGW2G9=M|Ko}`i%e5szCNdw999{21yuRFN2E7^==Q+b#iN0 z^qg^QGRfj)yWV}f=7&TuYKqFv?Yrrcl$qBZG@D5v;G&6Y0~tb$Qnei(4PX*8NH1Uk zb7q09PeQ=7QHPNKa)>LBFg~fP6G18nL|N!Fa0+|pB;1Sd>TsEaG}alsI;bebM@vI46+-!^&P?L*wXq~S`-+}IiN`{DDclx3qF4UMkbsgwR=yAO4s*EbZ^ib?5=)l0h zZ!e|6h+3?$#m&tPD^LINj23Zb=nIcx9vGuO|IyNU#JJQxmM+5`txE`D6oZqDWR zxlH-mL>HY$Ud5vE4FnPmK^c)l*Y0+Ac*VCq7R0ny&Xo3@+`&oji(1a)>#B2V&$h0a zA?+`Ok&#l(AwV?e8w{61MkqvPf?!KYaiPB|h#rNHQWFf1+2pgDA}As{#vc7-`|5X$ z?&vnQVfqqJ<^S-iC;jG+&w{2>K@npfEQys`i%ZN2I28XQCc+@o;;vGq>>L*+tn;|LqM{dP^e3luBfEF<+cTf?GOFJ>j;BA4B(k zv9KZIZ=Z^L+Y-T2Qa-r%8RpL-4T%<6UI#@Ga;`DgFH8=Jsd({gahPh6SJQ zO0iyk^y}%%-rlXB9>v2&pR)e^digjo{gCZ@q2uF)dj*XAhSFxBe<9OU-N@DAR>Q6M zS{kbRq2XAztl?c^Q3OgyKc`6j{Au)pU5G=xqP`##!xxWMUt2<0AFjWf_2x!Hp^Mvg zH>RQYrM|49&CpTKyU%X>hpqYkH*^WxKEHe12@-~DYimE%19K7P)aTdN4{2aPITeT? zBBB=q)8XnqGr>nZG_DIhTBsyU8#-%UWgikc{;_o$wrZ^N@X*3Yv6#Vajw!z;%<^Ik zuL-5%6D6>9n38JSvk)+8t@Ea6=*u!ubnkKrv%?Z0QsdTGZ{79lLKaUfMZR}feeADb z$I;`osm z_tb^#Q4y^Idk#v%dKP)pE^=@EV`6q{lrrrad;abnM4Vq#(nsFq;XB9OzJlRSu9*@c ztkk9mmZyvv#D1vW24>dn)I%jp3(;4y_vAlvWNbtitl%sv3)th_%=GzoD5(98E?D4w z@laJs*Y|_c4(s51Hj0d5&T@l+%&&H)XR{uMW|zIY#QFU)$s<$Jd%kd#8anzxLCR|! zolVCGN%Mi@WM+PK&1e+wY(g%0I8&RZB!{oE`riL!qEYu`h?%C;<3})D{WHht=-aAV z_BDA*JJ2FtzGN7o-&p#Z=j)G0re|jlADAbO<5ZTGqQUfDkb}43{$+4bTl<9w0q~8= z<|)zG*IwZ@%m-(S+f6O&9Z6&+0K8lTkrK&zSDDA}YRsHj?+vL!Ga)%o&rw zIP-^5$QdVgu$ue(Crj{_N_)zM3}v&j>xAY?H?VKM=U#kLKOqo_gI1JogH6cmvpd&{ zyD)NG?An3}-Y+6n}nfkb2>Y;Lb#5=LOMA<+*|RX20P|y)8;FNEIdwnU$rt zK8l5xQ~6aRm&`l`!~Fj!Jx~t(Bp>uxyjFVq&8)Ez^dyP%2`+zio=>LCsjT*oDn%xT z4_D2Vw;S4B9%B~}uY4Fq<>fh<%QT^@H(N;}7t4=e&|+z^f}rC(ZT8h5;~12}hwWH; zNA*_K>*TML_54nNE~Z>IR1<~6kV7uB8Mp3q&M%tb316sAtHl{z+p`40K8?UJ`bxKM zTly1yh?49{HwB`qcf`9~(L?*`TWNVFhH*`SrIz3tZ6n{auAR;#;UvXqsE4Ue{qPj!ug}%}cqhsJdu zpIuA3liDHWCHBIxZ{&+=)fh&=d+&R}+{2e|T&anCA7WLj3A3lY-@qkXnU*L`;ex`u@;F3%N{~9$ab*tuqd*W6_P`XFw%C2(R*153CuXn6XY0URG zi|D?_Zl`~FfBE`oEopt`qT``l-M$IBG2`1lo)N`v)2~B~XDap`7C==+1;#*_l-x|ogeIDda>HeOZc=T|Ca zjaL%}`mH*&+OH@0S$+|+J@GyOF-Ziazq{1d!!`y3hASf96l)IgM?C1{U4iSdrG1?C zj@@CotP;!?!^)X}roAnczFVU&(H9$JqNmL6vv~AV7Q!^Y)pcR_p7VNNaekih{fX&V z$WXf)l|Q}UDvP~Eiq_c=e?g5w_uQ|DMu(N>g5g1dX)MH)y*S=4UWI#-($^3OAlr{} zXrTm|jhBRUmoD?}O&Ca8uV2LvVr^1khU^o|l+6E?{|F72GI;g5I{`cO8ZjhEnX-TD|TS`Dd1f)?yT0%hL(4nN#U5e5j z(v5V3)S*L=?t0eo`8rJ19|96RbxL5ojuRR_w&e`I)N1EW^zuw#%nT+)_)2M>tBY@Aa(Pgo!&4gcfMz zhpc4$T>ncQEv&>2(Gp;S7R^AQWKnfDkR|LM8si1exB7aqvGMW6rkg)aCGS?1#m-kg z3-8t+?g3y^iX$4YG|~0PxJ|2(0+F+3v~-qky5}iByn>yb1W8P_g%^*}9gMGceJ6eV zW$0%(Ug()!-{}7$J+I6{h2?4dK)`<5In~73EpJlg3E%N+m!9N^e%R$mDQbr%mPG{~ ze8Ugo%&&RfOQ_G%^Rro#5l3hYL4JzIa)klr!!RXM#xUaNGJVSC{qz`Kt_7RHkXC7;1t*r_NAm_jIEX-D2rUy zXiDf28~ad!$!W_1rB9==q7+qvo8UTWE6G6NMtOQ7WK+a#&o^H*`Nx(}{Q-;uJ4unW z>4CP?L2Ioh?$qQD&eX2qF(K5u2&P=Z3_v&l-FCmPJpoUQZ&yWGc`4!!nskM{3a)Y8 z3rwLg-eyeyDLe=0+eFpOWK1e*e0vq+sr@Bnwnx&;=lHz$4a57Zv4j(tm&j0G6~Aa1 zgaS_%YGrd@EPeDKF8eEvHl#L>Volxno+B3}+*9NzFKFkR8Q;tZhp(=&nwIEZaMNzk zfN^uUCV%&;pF=2>Wi7*(6f7T2R2zxrU%@JcWO)7~Bz|o2{%;7Zm3f)Z1l%xZP^xUt zADz|Fn>4IBo+YR9LVMDg6(BH}Dt-M6*AlIED_WGg+F z=?z2_h~irl99|x|>kq;9Wbs>~U}x%5&h7TFnYY%N>wZMc1aZ>td0M)VkARFt0-oMt zh@FykM;dO~o}g$3`UCO2H8gVmVRXG$<8^g`o`TqJvwWLg^#d$Amv4Dd6LKmbVl?R# zEzgH+|J`phmbgdO8&c894G#y z7t^5-US2{62jaC~$7DCBhlI-|Gp`qhtGAB^uAv6K-WN<%zizQ12M2Su3XhzegdwKV$J*Fdl>Fv6P%}-QY_lQi@pt=7ygCi1 zBi<`I78#U!a%o%Vl=?JXiB+@8VcW5|4JH;bmpHAwFF5?Z%S`U~t}YWTUa z{V#XIPybkpykH;nfBA#avTfpJHCK^sR)7plNA{UXp)~Y7&pNK|1gmDMp*+-vi(j88 zCc%v%fT;1*FdDsvtv&7TOZA~T&eW#b@AeoEn;Bt>L}(G_Li*!Xc^}ZRiC(3Y-Cp7ap#e z-g@EDx_nSoLW$83jl)H?(bmG-nL(SWZ6#apQXrS2LQ5Cyi!BvceCbR5Eh1L-u398X zR05f{LwVuk4-ty({<-S9z*dxjLf+o49B#CCavU~ta!xP3um)Kejkd-7?0#U3P((pY zSbmn$b@IVK_avA2CdU83hkAcq0=a|^quD3Fe6Op%I$fnODls9zGsoNohII2g;WH)}H zV;)a*CX1ZYMx2yI5aTJs#LFUare{(c->&({8|&U_t5qn)hMGrv@#vpmOr=U}yI;%O zcwhH#2=9DboI1v6M^i};tXCBa<~Mhr^GoAiNjWaR;by)Jmi;$(h5k}#Jm|hh=u8l{ zAA=sBx0?t~n}DDEi-N&)7rYhAd-C$cT{{x1g>`&neX2G_x3g!Zu=u>SY zlo9+B7AqAqtkxS~$e0o-ghjq8vDrxmPoGB3ZqrnK>e=f*7do>j_I)f)k`fv!P*h8< zi!$?8jaEcYVz>T$l+Uimp4Z)m5o-Q!(!#}cIM8D2zhbeZ`h3$w@gvR)O}!&JeBBZh z38g@r?*w<##Xe#CHY9A^(8OL!U_x#gu}%%|SSBLv7YEsbiW%7|G?GV+s~<|=oo#x| z;$AY}dxcV&*!Nx|zI7m%f`sICmDdd7+JSb z80lr+r86R8rzzLhEkFFdlPyGP>@V-VJD$@<9r4yxU^)ZqTh%iv&1gOV6{LtSvCb_k z+*CC}=hw#qvr*w^c`i_??CZz*`O2_Dt}tc3MCaLg64>#I9>d64Tj|*}?%N!;v{X@G zp~77&yKGUyQtZPVJsAB4K|&M_3tL(&_gro^stQ*uCYY3Z0z4*cv}5e}Bg_W>UPEwW z-T?kkT~STUa`%nV)^of?f6`A2bl0V2cb;%$!S1oSO+pdAtf?a~$tG!we2qZ$ywe>p(muZ?A1`e)Fj z&P`U3!*_C-?+tMOTY(}nv=1PGiwG|GT{9lY=)fr_cAg+&_e#CMmI z6g86&)!VI%RUQ5!iH2(~W5~fO`&DrkAIW@4?yq&r7|ofy!cy;mubeC!_d-Mdym*Wt z+XtP*6}o{3e@Iy3l3%b2w_qj;FIOrgJRf!NSQ*cw_rf3ht<&@?I$l4#GwT5&(e)9P ziZ?!_h)z=DEw2S7Jxr_|WfI9UW!a`+4GOHY8`9vXdIDc(;`eOC&SJ{^`mpX5 zzq3pc3l`V&zx?SU5gC){)@~9hJf%Bo=kuGYSnvA88HOTq`!26t-5NbGV@NCAezZl; z|Js}Q>Gzok=ctH2W236Ad>I|DObYWGz25&tiVdsP&CN22O}86;FZ@ij5{w{EvQV`kpYg1L|CJ~jTl;&+eJl++}DGw-7c-@A#cnwxr=&{kZE z`_|j&Y$*<~6w|T6hk0RoRGOHpGG8D83Ym;VUD$mxZef=S_0EEP7?j=Nv)JXMow3Pk z9$5ZhMdbwsd(Onv*tGD zqCk0IcK?e^?lVDRaSUuJKZXzj4NJ3XPf-%M1#aQ+Cu89ee&^Si5JR5IZXPZql`gL1 zZTst&tVYk;a+OePYbj+3+)ocHHi1Xt8>$SC>HxrzU z=8T$~Piar=L-h*fVnMgNq1KaZ<8TZv8e;0EYkn29zjyXe(&7H8U$XPFdlu#nv9dX(>O=H=)Nu;5V0d|yvq{OZ=p?I%pAie^Q;5K zQKpcyAo>SWngcvcPbB!)7(Rm{GQD$#mjww^s#!_%CjVh>`|(FD_B%`j)AD9Z688Ow=(gzCWYkz?_W0M2 zocfk>_!307sDkar3O1;{DtNjaEwy!E&|eSj9a%=+{pM<%fyJPd!A^}^@h@rEX6uJFFa##7(<_YZ4DL6rDaq@pG)`% zgE&a2<6C#-I4Ht72Z*SNeiVlcJXP-dog14_GM>aq)QX~U)PC$BsSlm-lu5E578UBl8=`WJ@8Yz)G5cv6g)K@BkK#E zw}Fk>NO%yEJf>44jELmHS)F{?)U21&VyZOjhuOYXgbIa!w?-+k+;%@x!g?jtUph1D zEkN?sTa>zqLw`|b`q%ShQFlQ(Y&L{-^c~{~8BMiq?UQ5NCni1ic0oVRO!sI?`btO| zQ~$uE2L61rDGJOELU-WbXLm>l`gNU~Ioe`1eLWWBMEy1NSi=cLhBz{+U3L&NYEy#y zl+VE%D|6}xT?|tPlRG&Y?6<>MQH!ZBC4s{ol=&L(UwYc+LwDu{i_f$0q+gOi&0JQ- zW7b4cSL9`u7Fko#0(X;>RHjmTo6K(|OUR1ief*n!xi+|6E+r!ebPuG6vsZOj%bXWy zqm?OH_){AqNq` zbk&lb>A0*i=9^E?Ao%Fl!#c5njNinEhW1si-5Y+A*~3OdHyX4oY&ouMY0e2UWGXiT#U~^X;sfnRgdjRK z0+%)I#A+l;Bq`Jmn-k9J#~6!kPSzEW;$NMm6ix1Qjm7=S=DfC!%g6V+xJjB#j5TtU zG5+`HE6#TTeumOF30y@nUjGXB~Z!{E{hG%%iNdE&3jmeA?zn zAcAWc%f?BsK=SL$WP)PGEvJo`3d^{AJz3eNz;gCpTxJs23)c#zTn_bAMR~>GaKutQ zXQFsBMB#+LUJ4Q!l1#il17~(MY<3Msw42_hQa&lnv}K~D`zX8|3#EfiT zTXS@7x`#U~{y$D_fV477o-W4=z4%d!*uj-xwGD~9?I$v=n1SEsqW;=}`P_FS3&tWy zw5sKEeqlf$jFSE+vQK^BMkn(0O6+(sV5F0X1TVVqc{rnY%6;>JKuT9Gryq}&B_e;$ zuRPc@dK$lM(1AiqK>dl~i^jWqWmja|R94dnevA!@TTy;ZpSz>483Oy#!Grc0Y{UL-K%AKqR;qcdyr&6l1D)iFz7F&*QwKLA{|bUK6f{Fp@a_|s zb9Jg`S}N1~)P#j+7+;%tVj1-PPDo6T{16w1kb;K`JeT?VxJ!SG9TNs4^fapfz1UzK zzWzk}PN&^nbk@N}d$4H$4}GW~uai66)@5b3rg}v8UkW0XJ=)9utX+Z}F77KxQ$XbV z$Td@c=1V{ot6Bqrl8F(wXDcI%caGxq9ORd-UF;CqpP1(vlqKM`b(d>z$pwFuZ z*5yA@VwhUynhV}|r=>;`eydrO^;U%OpkPYMArRV@RBgi@{yMR7DmH_)=S{kcOnXIbVq-Y0VRz2McKXh>!i{*-xoftIWW|=niv%=BG#gC~zo0UA` zlvsQ8ojnCgkHMB|x=)2&fb-Ecb^s3d>*qtY%NuN*r>up8UPt8C zeIF{B3wMGn;NRunZ9xm=rHO z`9^N6LvUzB>xxvI`sP2|$h$7OKP*x(ry&zzCH9d~in`^Ir<3win#{P3c++ zqADyUmrSiZ)b}&!NFqD-!sT$Lt=JX;`(2e{B;ROUN#vau{eE6fsuwL|CH$-WoqIQdCRCB0 z;Wj)UsRTbIVL%#|)8blT-cylYgy0S2ul7tSx|?(h4h3?#NoQvh4stcKRvy-H?2!qH z08^d5X?fI#&+*?isaG8C8;|lh^rZCElsfYZ8~SQ8KA^Vvs3k~iBI-p&#T^jU-u94< zaHu@bc7CS4^B!6z9_BAgJPfSp3c2}qjvONKC#SSZSbhu`STI9WMR<;?H!N@D?P3h+ z!Nhm=2D62lzZnzMLrmW@mr=0TJtQ2fy0jqv^nfbgT&ex^+zlI9Z3&GQL!k$nQS(ZO z@Ekr=?`GyrV;sy}QtoS72DBC{Jj4-sGr`_XUZu*835>kb@2)Nm>zCUrOsy2SUW=dK zGaKudWNMDtK#^$=;+fdI!~zrWQMtZKd>N_?+I~Uc@uzcB?kq4Hpiqb zcpmzcKdYN+-f&HLwjl@Y_^BZ&8Sk*`GbL`dn@9NQzZ637$(lS1wKUqb1P?!D70cAo6C8b`7{~4U1aG+tuq=4Z2 zoLEwj!(rcVvhTw)pYxn_-dX2&S5aNhPTyH$cL#-AM$uv6J8pga99=ToUiNVWziRXM zPukG%I|pFkG_PM}cR4S^c%w0)Z+IOV`8wP(l@W8YkMuPd!K=2`+<4WFAM)AzGi0S! z@1UyAws)LE0hdl?gp8uqqBO8=I5vv`zXS7lL2L&-)JG+W#foljKuX2&JGI(al$kaC zM1c2JvO4M9_igJN3>OrVh8u^+B275~+W9EOW~Y5CqGX>RLp|S=CS=Y6WU`ky!$g% zU#$96^{voY{OSork5cHf&~hT0C>hx=6xV_k1~W&x=G)R`HWQwUVU>5TGw_rjGR?gs zQWCauZ64r+!=5S!Zw9;@!OtU4{3w;zSl0iR)^CSYY*!3xx9S@VN?GQ&wtqcy^p&3l z?KlQAhRAP$gbe?!pc}0xR$ClAa$oN1 zKx~J}_b>LnJ0-e-vLaT~F7GgNv$a3z{gAQf5&FDLhP9Ol@8M?KsAzEMw#+H;gw;Oc zu|ITbh}Fe7HXg3Qmn_9H7VV=H=b(>eWU2Zg!%k^PQ`l5s8@oL8wZ~VXP|=MJ=jm85 z5tO4S71U^!|Mhu`zj}|n;!+Fx?65Z%ks3XW>5fc)B$4>jA-BXDt$=IR!c8kIB{4Sw zEWsI~y@xPLI2BD7;Avz|6~0+LI+J}vwDoJbNIHe~zd$XszkNr#Yj?6XFcna88^|9Rb`?o(`m1GRxJIwyK}doSbC_} z6GC0g*%IqA*=m(P`5|K>*}Ub`dOh#cdc2iN zM5R;v*k=yA))rxuF58RFpiDDzpu+^-=Kv%2+bkWwzhX8qsPXZKNNpm^R9JD4`g56` z@3MPGkp@Zh|C2T0@g8)l%WGt|l~|Quda%uPKCIF4=K4O625l?38NvvsbmmSl@@6?H zWm7u~5AW^?w~B=Y3Yj<*)0};wIWggt;>-g!CBk-JZ8KOn2ZAVl|QX;IU0fK($J^ ze?lea=#Ms+_z0OW_W5so5P9JTKGdpX4bOd zTBc4x&Bt{&t0$vFO89Q&!<*mB-v;5bW74>iTLfNnxsQnrf{~>bTJ)C`4$h8Of4>h! zY=}wk*ZO6c-^912sv@+01hp846`Rc_?Z%op@F$>uH+~*p{C#+HMWm@wQ?)VIFHCxnoz(`@qOTZ&zU=L*+J@XksO&0 z?7TBil~yyRIs#-?1$qOR587&(yw+N}8xdszm-{rBpBpa)B5yMd;2&(7TCb~^btdbf z&Vu@YOY+x=MEfEs0^US^oA*}-{fG$-%;`D_RJvHaVK+i0SdSA+B@%)0(va$!?kRp3 zZA``C7aPoGk3?rp^!bgggvBDJmSxQ*YA-_soROS*2K_^V*3~9|K3j)irR)z$qD6F( z{$11>yL*h3#C%TrOW$Ajc?f97k;}%wrL8btufEQ#n-8OTL!ySu-rdhHo4c})aQVy_ zLVWaF$0}264{PNq##_Q)U2ZOy0$IVT$(UWqGscSsc{dj%8&CQ-E&@W;{+n{52DC0m z4fDuta18si`S8+GXy~hJy+>ZETD$}U4;PJEXja&gHSQ`u@vfmO?U=f8z3O&c&-8de z5)NeyRnYsUB$>8K?4aorouQnF?@GVqZkS%FrG_v@Hcu+cg7#XzlW>|!ue2cD^zB`2vQJ7No~UIX11@-JSH0ZEiG z6YBWhKy+O4?yj-70!YoTR8jV6Ch&cvoddH_9ze!!X#$H9*%;r9DSTizTWS8JPV1}N zao0AaB*Gy`CRoho)79!v;yzw)o00RqV$khpJKKa|ej zsb?ZUUYV%*s^`30ZFsq!(RR1KGVpoJ^1RVg_$aL+n@{$1Mkg5se(>C11vaL%9*0mt z#kVj@oIoVvqc8Ch<^{nOc8VG+7%YF%6i*6Jfn!o~5T>zUf?|KA@#S!|zHukTj$2n! zu#D(Y_erOha@r<5Iz5MZi9X4V={nCQQ8As1x#DIjqrnZDe=c$D6LIb3?IjNy8_UWb>J)1Cr93Wgb2Tr&>njC$H%WHl zY&eUwI0B}p3TA_-0x7d_YMR}teW*-Jo6SVa=8+81#kbnzwS+6=c=)?+q`>foZr4G_ zgN5K9jQTfxR_Gan>usRH$nktNUN-l+%aq{K<@=>}>`znz$`Xc#h7l<#DUVKW|5z=6 zNxW=mLPfSp?jE9AMqK;eNJwVf4wg0z78fh(o&PlJB5)H+Bgpo-ubLu!2dJb0g_YPy zbbo2&x-PdElt^v^?`VOa5?)u-!$lxf!ctObK>Hv2-*U1C+8K3qR7j>RB$YjEe7rU8 zww7ptq=y44*wXQ&(GyKg673@W=^0tpe;3)%jCl@;&=4RAC!puY@)0bTEsO|_acv4< zoEeme4LvrQr)&gCDJYOKHd^Qb;KmFguqV%PqyYR0v~3jQ12gWdF#`!BxEo+598c3w z{aY*TkJU#mkfjQ8;mrB}3v<#$?a=@BPoJ5Tw*E*nJ&IwHCjIc2oP*S}sbF)8ILYi_ zbYUHHh~^R%JXNXN^yTR!`avh#wL5f*|3xq?KT6{va?of4Zv206zyI?Q{{MVJc6t(( z0S>@8Nd3JIzz@UpZ-E2M#&U${U$|c3-#?iyiAc^5!2War|FP&S!soVUjMP+nY2ZR8 z9DxcUe>U>DcF|t|fD7ZP5|ak;74CAYQI@Nd>ZWkf#<$}ffSfIUaESz}0xtLy7Z(8} zR~|I&lTa9WI@%Ld&Qw$@Kzm)1H+SLupn!S0$jFjX@#Uwy7VJStN*?#@{we*|0 z%)Gz9-)+0J3)$z5<}Zx|jv-+J4<9}HU@c-hGEHN^O>uKE*|Y>OD6;bM=)imdN+B3| zFL`g) ziMs^Ub1kaMq6nSv5d@=x6^BG3q`9x%9h;a~0wPJlz>Z`I#m6Sw{yg-#oLPuUPX2hg z+eD9KeJ*|E*+cjfW_|sNQ9bXGKWn$t9$EvyZKfA%tE?;L5?011)VH=QL zS~2S>i+B=A*MaPB2L~c#N#zw|NGDl_1%85u=&(J-ha|&Mg<}9zN*fY(J|Maa-o*Tr z{GQrh3arn*@tmcoZi@pcX1I5G?Y5a4j|_A5NLW+WHIjQc_FZ%J6vIFB<=QY^XaQr? z)cc3Jle~`}k${hy_wgykH~^;Zk{>R2i0DFXRZyrHea!Xy1T96m^aCI-sRHx?W9IDQ zBOCTRsh>MKmTGBgj=o3z!xA1I?$|}=hJaatiHTvrieXY30kFrMWV^cW)xi^H$PtVH zUBj6a>i`w*TDiJEML$g=xOB_p>cEu-Pxc0=oYwlHks#qF^{+0UJWrXFVgN8&VzaE5 z`=j;e?j$cji9%i>*WjmXPgexJBsUounG%?23EcbYDFbU8u0NS0yGWe+R^KUUepQX8 zb6>f0J_R%|bAkI&298YpC+gpOp#>)Q7J6PAwpE>h0Q29ks%)YoQ}eA8x(^wreS&+? z&cOl^egC%|JP+eay;U<@a!xo~*i^%Z_Y3$1$spEt(mvqk4(eU?1O1);ca9}k)zjp_ z`UL&yL{6TIt$|bJ7KA9iQuN^kmq@l%-j5%xfH{8wN+iE^(HqV9jijB~S`rWffTmc~ z3_N%v-&v|1g!3izbJ>>NO zt;h}Gi#ap~7$vCN7^z;Ym!nh||HTua6AHvGLodbdPL!-3nq=G%T;lv@+cg7$y3gEc z*70KpN^sjS=KU?*lfQExfj??LTS);pOr3f+J{(-!XD?p{gVO-WPS7sk76A1nK!V2B z^!4l4h^iU8Hh`!DQ56Zx1-!`2sw#3q@uR4EyTVC<1BvMg0a6?jU_yWrPH^PW;3+me zS;`z*lsLNiY5($*ajA~O#!zOt*srv8pPO}`W=EQ$+BK^41iivefk}OTznn{NRgCJ)1|5zxXU-YEKlEf^-ixt^*oAd2k7n(V&r!;6soIYgUzY2aySw8(~ZEdY) z8F)#+{3KKudnBByR-jjj1DgH^#ST@28@W6h=qL=`7zyZstkwN(U9j{aX(ES}?mO;t zdc|N7puuxqP5!S5PfvXWnBZxfh>sB%F}tT1fGxEM&@ioB)w6yNifXb(PmtFGOKYUi z_SHhqO<9m#HFdh@0XFi|filBZmrhprGan@IDi%zPWKl{Yt$HmoYisWwA0OLZtTLfW zq!WT~sNm$}r2pfKuheTW3~P7y$66d@2Zx83raE)2|I-P7G@>$?Xq#OwontX6>}j=hk}`kZR7(2Yn(yzS~e7Jo@J-SO5(Ua93Co={%b#e*^Vy*eW zf;6ODoHhaMRt+{%N^uWeVWRQ74Q=fp01F19!4jD#>#bC-C!UmSO>Iu-2e56kIBtWmaB86?p{8?u?PA05-@o<5 z_S;E-X0MgZ(x(3MVlaItYr3Z&G7bV3E1=iKOHzYT6VQnvBOa2;+1=kSZ(u-+jMQ0; zjdbeb$65>c^|(m6U(Ii7vH7f?fWC@k6g$khOfufTj|w@=*j6v5efSFqJuoE17uYre z(c zdNTK{$=RikK|06?JRSgOBE9j&6}?Sr{~f8H11BI*wgBe|EBUYo{AhMr88NUU0Cq<* zwE&!(Fy&8klx1XH#dSPVBc^|zSf@{4EVkhbbO%dzsrICgsA=Q zA)qLCG%SF!4p1yZ6PA%`pB^3_)x{)u7>r8KtG8YU%P?a3LE#FcJAdBCBLhz;Yg#in zNzVKJufAoao|d+@>0p{TrNCQKq^Ks;e7fKhnP1#U6)Ao~SyX@U%aEQG?D=9MU!XX4 zavIhPuU-41VgfJzw`RF;-=kq@(;3_738>$!J)sOOj!dnSoAi3%NY>Ln5P;D%tMD|C zp1PZvnH5Kdg1JRb@djBO*x>-7n?}SLYtrxoS7#(WI4!y#3Bh+}s+Yee+qHuE;`?tk z6AnsTa-vCxJbq$L-waY`KyxzzvEA*kUx?3sUd;W~FUoe?1~Cf zz>IBq6K@a%NV}ckx8N(a#=bNR1%AgAFxcHdwn%*NpkihRSY?8+qiJ$F0K50BH(hLj z{;;O&9s6fM#u$)bZGKlM*?mAX!;pUAGt1Lf5THwKrXOz4R2TQV(@A>MfRaKKiOmAa zrAtWt@P*x7hXOQdJeqD4MIyQ4;jyuR7x~(YARla7jW^={m<|dY3J`}TjeNw9Jj5HX z4phXhk4C{bkcqj7xIIW!X6=OHICW7XmP@>^J!V~JY=Ow^hi}?#02yuqQtq?yVsCU% zBLNR++MgtFw(?m$DkcU60*;g(nFrqk-l}@em0IpuJE%cEf-FX@D=sb$qfq-Jpr;v6 zm%S;`E3~bm7PO%R*5Z%c?@@{4Y{^mj3`vnkd^p)|0XaI-o6LIrDFm%j$%3}6VPp^X zep?rt0Bh%4M`rcNGUMlNBQ=r?3TzQnWy@8o*nX4=$bvF!GOxI}s|~P8b5-IyX?zZP zSTo1puAWZnJK@uLY%tc}T>L?zSVfP|^NWf)fkH6IULZ&|2(JYhS%FqI4YDo^_KKvJ zj6gyQK|%@fx@>8I6-bG!Qb3XpLF$s%{G>SOK>FO}^_w>^2zk;H?B9$isf+=KZZ8gJG>!QCIH*dcON0v;l%Qlli573=^_K0gk`p;`k59z#!wz zk=&Z?vfdKAxw@L2Rd7J%YL&+aG%kMFspO>!ciX9)bZEmDG9Jr&mKu@F5ORZW3oLrr zI&XYS(4Sf_OftR~9k^YPf42QO&Z@7kA3a2#h9>uckSufvsd6v6BkFbGta}A(&U=<^ z%ulu2=*M6Tj1lmYi#Tr}<2^(Rpss6wTgQY8SWh58fRqFB;1ai2PRKy{&YH#-E z1g~?BsDRtF)jN!!6AuF!=OQ?-6wtdGc3x0hTRVD4K*i4nK7l8aI>3DzeiV6RYhx1t z%1vN$%fBmK3dF%q(lb3h9I&kh>SSZEg$T0zwtk_7Pi19g1lgY-aBx5wZ997)Eu*|M z>tO@LdJ*n)FJ8RBj!XnohKt-i_6U!QYs6CfMgnpRWPt#(N(~JS;IaIHSn)EU&zWq> zn|v-(9QouflAd=iEM7m6{|mTBUa*6hr^QhZHWP&%5YSiE#1XsnZk@jkr}M6ds9VPu z1jW?Tf-J!MA>Y9bcX?hPPlC|Zntpp_flSB`_MJMO#{#new*sYo&ERj539`;lcm0tD z3i31PH-**t`8kx}mLlQhA+ugoQ1Inae*R-t2zZ*0s2%|?jCueAYh8WO{lp?Kw}Xoa z3sE^~_$QySBSV9Y7d%^J3XxY>7#sN#xLFVzQ%_Dk8qPqp|9!=|6e$GGhC@+K_p0XW z=m7Q@;DGnQwr9(ZYgcNQd3guzhvV^=X7f!4zaBsgFKHREOpRs9!Gv9W0b1kDrltP` z*AG^E!DtmxH?7lFvV1Eiq=S}^hi9B{>7?!mGBAmrPl z5D6Bb41iR^zqxF>r3R&p|FlgfaElREqhCz`I^9cL4C+HU1-a=GGEQl#iw8>kW0(LZ zhb#*Dm8U_rgWMznE-MVM8!|!Sh7*W_bQ|EQe*`$vUR$7}kBnqj`;k6K8t@&Qz`ZH6 zNJdFXdZcxO(S;&IQ*r$Pju;&wfU@)2%?fksmNEUULvBtW?+6k{V&z<7upZmMRcDq@ zgBU_q0+0b)+WGS=eAb#K9&K*nn}EGI{|}ls+2;TN literal 0 HcmV?d00001 diff --git a/examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_validation_loss.png b/examples_deepspeed/universal_checkpointing/assets/image/uc_stage3_char_validation_loss.png new file mode 100644 index 0000000000000000000000000000000000000000..30d6f72eb879d5a698fcd25545da8f59cd0fddd4 GIT binary patch literal 41972 zcmdqJbySu8_cnOw4ryr=kdjb9N`+g_Uq=uXsm}^mo9O!@+|5(@w&h95MO`4H1WQwglD|5sQP}@W*yD} z^KFwtE_L$cY*ya{_JhBn&*;A;lFMZtY^)RwtZX(7wp&RZ&CVS3mGSk3kjvq-g>jKQ zZNOp;Lw>J{C6q;8BIl6DVTB)ZQO~k5;nVnR3ZFGGZXIN8WtznhuKMpU+?9 zK@lKtenCtf^v`APMEw8$GJ6WLvM5YU%w;B}_?aeu1`iL98vno5ItTWT9z`m~(x|v+ zQAfwehi7CkRXff-DEf`rQf&Bferf6Q-s#l0Z|3b0BwBNhCI?4~Z#3>l1)QNpdT-(5~l&I~4@`Rure=5kTy zRiopB~ZVos#v$v0JYm=L94rJlgE0%rukUijN zh1NTgo3{C6hsYr1!+7;XiAnoxqpy(n&N9Vdcw}T~-^#eK)JydKenfS(Xh=wi;KQCo ztqB5BQry%&#Vl$6_cKr7uG-q$TbDlX+q-Rye^?vN#jEr9%~RfB+8nU#S>f_4-JBOFZ@v+IH6@bS&(E)TwzjtT1OyiZc)7S*YF$_7S6B7dMK2}cOL}f9z+??o zSgCsX_*l#~c;D905GbFGAYqtaSqazsaG#Bp6_u2fw2ek5iOr;?q(oTtrL+zXl6^4x z6cHOs)O7TNK`QXfxZ%t1@1KOcw#`s2zrKC!EVvn48cKkV9|XH3Q_4H!!eym5xuYMfe4cH1yaiGrdc2Fhb=HeTBQ7#TjC_4bxvtP50B?6cDR{G>!Qe6k75H|Lj^ z!(jc)dtY#moqd1uy_!T^T->1IVYC8TzxZCR65UN7x9AmGfd^qQ8^|{nXs{SJ2Q+p_ zKvYS)=_@LA^JeD;leAw=htJW`5fW#7{;UP3Ts3Oz>mz#3sn%BSxz!u@+RE_cWOvZX z&8_2TcNOErGaZ6PtoQDZY?FEqCuiqYGMN+2#ZzLG@4+T=0y61FFJ8PzvEdgm zV`w#J^wD~t4Neh;?BYm-MV2X6jd?rn8RRl&!5feGWF`A-<=_Pq|GJu*sDtNlxEz~LY5b4>+?s3hH@C5g zYG{zc#>Sq9q0k9gVDnwQ>b(EUu>X|^5iBZ4Q(;0Swo|v z9T0Yfta_=8Yh7@TkB?C?F{|0~ks;=EadbkMxVX4A0jClH$3L$;w6ZeqPZL6+vNfZe zm%7-5g{k1nS_|*xCT3*BCnl0N_#IV@s$zu6vHj?O849Vua;YnRdw)O0D*&csaS0A- z{OR9=%hJ-!aP)B71yM1CB@js3zJ8?yJV9A{0QV4m;BUy|<9PJO@7|_ z=hblUH({xS{EywF{Ew+%zlH<{W1Jl{pGQx;gj4n+O(NP_XpcvFR3|lA$7ffpkevQeOR%un!+T97+~) zLVAp~?0%XJi+Z@zCvbt8`6*<9OFTS;uSFd&HAlI(;S1$WO4`)^b_5g`!phLWaG)eP2XmPS;1cCbgsI!KTUh_;|`q@14Akvl7#0{_+zz?vHnW41C*a`m1iy zNWvgmR$X2FD?#Rr>p@CZ)*xZHT=brWg@u8#va$!P@}kwk!b0(f4_AF*Q&SwpF^WH{ zs;aVrNfs#gTm6#p&z}}M2+$QFA(-QHb6M^l9@enwJAZ$hnyP?9X7zPsq+{bjgI5yr zIRVoqSZ_I!o$&Ok_iou1wwaB`T`=>J#r zdvjpmKlkCeX&mj*`&aVC3WT(W85tRUezV`c87`(#HA@59YL8L4gx+asksWVsy>+QO z>9TSK#1ASNX{UJXt|tti0B`4mEtL6W-DzeH-7!!(W$4qgEGi{{2i1ze^Fc7 z-+vbwW`>}-T&+VuFZ}8-oPz_g+X!9pXfU%JnQ1taoqceO+5vrMs;Q_XO$JO=+CGLY z(c#S~?U#nklIO30Ud-v;caBzxCOG`U^pIfO0a8mIG40LpKl<^r=O(>yP{}gQ116r&0 z|GWQd`SWwt%Dvjb<^*Z|lcwXJ1zGGV=f@kR$RHT#>3iY6AJjPAIyu!(fiU_J=s>Y7 zYoQ>&mWUf+4e8IZB3cNk&MUpt5b72n^I2t{0kq+(tg0G@QRU_3MV(jK3}e#M(=W0K z%gM{Lk>d;c9Xdt-R({H8Xp!Y7&BCZYtSZMvbxA~o8s23yl1Gp!?SKEL)S;g=Bs;G1 zUod@wqM{LSJ6YSh>q_3}!A1>F?*_m9c2EnFf z3C*q^+UR#g0Z?;orv8)Hw{IUGAy{N>ZS8hL3IHTFDneDp*Ra{z;W${lgzzHmcc{5( z>mwo}fNimn0)Z?OR+&5?-t_sF}LKNJyxszrWbc zH##~RVH>!_bSOBU=pDdZ5ai!>9A{O(&dhuYo3zsQ>)YO~a0^|32OArk+s5_%tGfVO z`{X3bWc3*s;JyjQoCOyJo%@E~^78ULUw?R-xd_etEv*A(dc<6 zF^&f^S*yk=2`n;qq>!-7l_UMg(UA;-3&;wwTh^qcq!BP|1&o06GylhvWf){E(tOSX zFvD*<9^LkiNl>#rda-}`xHp-1vcda0qE$GX7~5{0zppg?HWBgeoo#eAuz6&1?5fV+ z)zpYNiKSD_Np^L0!DL_P9;tRP`m;8S1&Tliq}*ZnUr-&W0WvQR&MdS?F7>~>ez^bL zaTu^GGKGykdve`Rnd})0DsnN)o}vNVL)wS~pz8+l{)FvVJn1&Jyu5s|ac#xus?1jQ zyEu?2%s}^`6Sg6LtDZqOrBziTUovmT z0#z83w^+nx+NrnRF3 zyYxZh`w{C+J6J*4DVx z)YQiH9wb0k`Sz4^gh90;qvX6B^!hc!kHIW5OH0eb!ovBUq{|dHQz~>AJjwNS`mYg( z?rlsg!9v!!tPnn#t_c=*{b2?RDCV`zDh!(y2nT?vr5^*Y5g-ld>L~<+oWS!lgX+gG z`&|nM3sJ?Eu-PMkej=AK_-L*6asxs?0Ya6hM?Q*&kl3EPLHb;(vyL1hBHZeyf{w(^yvoR z15a!m92Ke{_h1lmurVFE1~g z#L=4Vv7sSZU~0pVsi>%^%zzk8R#*|i8sLGJg5X+!_o+3zQZh2ly^VpL5}uoJaIzOc zYFHsHF`#Hv9YSEpscYaDcqo@Re;00{5%z!UdWRF#$@pkG|6`2Y#XDOA36A zw`rG=YklG55S?4|DfiHS2e3r+fh+7yz`^mn`X*c-W9{EWx zj{o z5(MSVrGON$1})I0fWN;#Q49kAM*n04y1(ru1qz_4MaH5zvvlW*i|WaJ8P^}g6ciNf zDb#Zx43o8$U|~CeFmjYAQ{rP!Hu~NW78aiW@k8&YKSoe=Gy!CIjx%ceuWv~JEhIFb zpVjwY_>(`zdr!C1bUBN(Rjc?g?e^|y1mt~?TAd*7r%SjK^6>C*;J)|`V`7%_Vr^2| z&<0`*>Kb+RuD2sk)?g%@X@eAM)9>2%6WRegaPW&1#yB<|;R57Ill0_&_UzdmsdM-P z4;eLRW{~{Iki7w*o)3LeP|yxB{^{uGsMefneq$31vD|uqe!9jPOXg%-{#~Og9vSnC z8X7xL`rtbkC&EC{($Uc=^F4SHElSEF9S1r3-E<>?&eqKbNsG)4gOLW@zZ*4QAuCRD zx-0JL>RMA*w+Jv(%zK9&C?m%vy+D9kK!CWMJE0H&DjTA8fO%m<5&T2i&LXj}B4jFA}Q(;C{d zT{f%RW13`UOgiK53N)$!X`G#%P1d>dLYjzxKqs%Lhyz@*6>t|q>azjf3tqd{2Ap|& zH7n2pwl%POETBEXAf*AxGhIw=f@2EGZWw$d*L`E$>u84qgjn_fHB55Tn!x!M3>!N; zz@Av@>gpWhe_u2=%OIP}@S`;w;4K;&8ZpmJ#7k5q-rX^u9r$HT#@AR<;yiS zH7!F!9*tqaFnwbv04 zm|`Dca)Y!AA;byj2BF4<6lF$%;( z1yC%Jx&nerBZEOcBtd=qBY~+{up4-G^oX646IpADLF&LKFgRv_u{Z#RoY#if=;-NF zYJQYmd#jQ46r!FvXfjA19~l`bdo;)_Y(Gv6vghKSg!_6J9QJw85rQG6J$wGV+;L9Y zz|b(c2DcRIHTWbXollu0%`7Z1RaI3h?Z)0Ke6QM@;@XEzs3pk?k}W9|O^E3PQEDlHP5fBkxVa(XZ~G%`XFC_R`VNelA~d0sgb zmxP>!B>~W63$W)L&mL9|$FeQkD8jOWk2B+Dn))+LvHxMo})3wk1iOK5W<_jl0jxvhUI|`?sWXeR@iZt5wwcV_qs*wQ@Z!=wZcA&0U^|~H3SD@ zfkB;{Z7t55SFh+1^%(RlMB(n~?Ue&Dm&ct+!yV>w5d=vw;PHqY1nl6FfB+fr4`gV7 z-#9kIKB&dgTnzy}_>@un5;J?f%5zQ#Sx=I?nTm*6SXc_*yyyoaGEgk?YMEXv|N6eIrw1Pdtv1;9P77@~NZAKC zL0NhEg#>wAs=dV%bTe0Ji!_VZ_JXJ`KO2KL{*jYwE1M38><>eV5Ahrd?s~Nb#--g5^h|u?f9#{$<#qeh8sYl z{}qi)?7j*SeiVio@OLCyfw0=~(HS5TF0A^CZv%Qaeg3ovG@rn#g-|GJLahZU#1xy< z>oI;WGjBIDGYbL?q*MC=HlzjMQV>J0Ld^pQU3&8v62F*Sx)k*1$WftJ|SI{jNefR*X?SxfF{t zY?L?yQx6kH)dgbwP`RZta7Kf-y#OYNg3L#lnI z>vZv9|8}=vNwTJjnoFD-vkEokKpOdFV+1M};X$Tl1 zGM5h;eHP%_*Pg=n;^EL)7Pykv3okZ?dz40VC@3GN)LvQE-oDmE9L(RN~&je!;6RTf38QydhxK zo9dGMCBRDL^i7Utjky^p2{@4Hf#k%HWddyl1`!WCw`FXss^3As*Z@OYTl?|%YBec> zn#f4y_{9s+26=f8Cu~^UnG(8#o0>SUU2|b2f05qOBz^C#F~$m9{zLN%Zqr^}31VX5 z))gL4j6$b^ARJ5#!0;SA2aC0L`ZoIedwbhpmm;E%Gg|o{J3$;(35rq7i8UuDr_6Kf z0it_Ag#vC$3FOm-#YL?j9v3cL;Ns)MPUvn2>VV`9geD;+S&(e6TFWArxnxS{NjNz+ zMvbsUASGos!JZUof%>0;J`p3GBL>V79)8oP{8X39jSG}Nzk$@&aS&-iQAaT zgu|F$d9iu#<+T+;F8!t#_NV!GkF-h@?o(}YCYF^6#xdL=h1eYe=mK)f_IJnT;O-@J z$aYq<*c4P$VQ^bH5PKon<$4{x60t|{$s%C5m04LWniQ+#GdZn)%n~%=8yLn7UP7=7 zgy5K>uL1t`^7UQnh^E{=I1q4`1T6#=R9|1;;Q|o$@wW;b$O2#^FiBp5HF)@NS;t#a zv2DfWO>XYfxHx0{@V&n|mR7ygi4U4CC6QwWlgrMW-uw)I&(Du2E^hD%e4Nbv?WMYnKGMtfl=>9q4wEZ=Xl@<%0W~@LN}E6I{FY)1SEn~HH9x2mXTQ|AWr7|gwizd zR0LrEauO*Up(;B?Osve+t4!XJ+T2pi%CH!*DAnhsSa_cdRgB;nKDAGuQRhxCr@Oq+}k_&{2cdk?3wzKnSJRf^^ zw-!^R?^9VB>htFr2ku~d303g*HjhLg$oH6E*?=pk%*Rmx=@caXn7;9d#Sa*pK*59M9^+jUsGQn zPxH@x??XXiREr>ST)Y~TuH(>4S6W9$i|_pI?CJv7!c0_@SIF8Lmw};3+vZH5KJWg) z){@Y}+ootsX;W$>+kgD1{OL*KP^~L%G((Ge7O9-syd2sRv`=BXQEaGBC`FRaj>TjH zxhQg48wwz16w4c*W!R%eJWKhYtdmA}FLeh-w}zfVtzQFujfkkvZY!`W&Tvd(jEb%X zb84z8&~+ePeAnD+j1H8}Rs5n2} zCF2-VeH@&l8qY1#{mq&1!$UDXV>+$}1y+4}X<}&KCm|2lJ@u zkux!c!x{hx#N^|B)>|gD)TO4dF<4z;|)mBk^?VOzbzT z@d+i!-W1Ym(c{b)l=9Nj>a$l7`WA_`v%Ul|&nvjfB}$(qspMYF!eIcT z1|CRy*h^jh1S~8XjN&f%D3C^QjL}y*+U{#EfwD`kosP|?ooMdf$`^o@<^VOt-WL{bzZnB_3 zV`#sPQ#Lk!oV-*>;OmQrQY;WE$^jzC&0h)S?4k|{cYy%Z+Jqx3Bjc;3wPE>{l7j5) zXwb~qq?_!#sU!o>h!>ZZ1T8w&>>M=i0l(B{}A#DCT3H|&WgQ|`#Q1T;Wjf&L%1v}muDbmFe;X2Kyhm}0H4^H zz~3LMKjSaf$k(sEGyEJJ7XgWNl$qy3CNJdz5}Pi5onwlK*fRoBM+}IWjT<+3d2?&# z1+Vc)oc=xYmuhG*=7`bt=H@?%i;LrJR5zHsSyNv>|K}&`H#`HCCr>tStE;y~l5O=F zD??RENa&UyBh-q;pL6O;5gVzii;6nj2Q9Wh@29XBH%kU)N$9JzpR%eZIEJ<1QE_UK z8OZ^i3>UPN%lKkLV&AM zD>h@j7l(&SollRc`!njvo0lcQzMr+JxkMQ+@n4GtzsN=svv{?NDLk{=MN;(UvdmNqt-H#MYt3BZ#&D&|Sa zyi2XiMaDF%*t2RVEFuvKXbL8bo{r(QUj->quq-OY>H;-&wA$-&s3I_wlsIrKma2Z- zq+;n?UVa$;S~ungJ}r`{!Q-Lq;zH4MyxtsATG0*en>lmaZ)F{S6Gp*F>P{Zn{>Dlh zD@=)2 z-QnM;vJVVRNujB%avypVY#Uu4#%p#N-D;44;2`3y$V*FyEf(79_S#aC+xi5TyEjZf`>$5grs{1E>Qrl08Xu z9FQ@}vdZe|(fs*y#vsQD@N)zyw?rRyh3b7U25!Roc)ErLvS5%o-S*LYjuK}oeC(cq zfn;?Gk&e~fPV|HbjGwyeZGOJ)qa}g949WAB$w_Rm$x!q;Jbv8iiJ?CM7!<4kzL3V0 zEW9;`gs^d#z$h9yTzGGZzAC#IN*1}5qq=4e*|9y7Zeh*pN~_~EJEYNxcg*Q<-}@aA z(;L5rnh{cfj%f#{(Tk7azrKB=oy1U}-1p$o-+gry1p0(Inb{?B**5oQ4o0SKKPesTpa=mL%W{_tO zU7(HiV`T*x2!z`Dj{OQvMwz~fm6Fs~ue7XBHYOZ{n*#2kU%Y*%WnG>-5{nhf`T2GWq4(4gRbiuJpaWpr_c=mj*eQ1c7hsW=tB9{p;C+?@btS>h@ zqUS=|7tGpmj}rW^@a{-d&-ovaz_w}afBBK;f#IFJmm4j?{(?3`LlTlLF7B=$)^MAXx)CTUELa5XM&Pux=f(jthvr>D1NGEHL-6AMx_HF z?iLpw$4ls&x4h3RowwZ5U$$IxFN|EP*hG@7(tb5tr=XsIO}AkBN(*%meHLDQgbylc zyYS%=t~}@!i9W%EvN-b(O@!+t{_UAa!sS6x2R(Nm9$*o$97%jFP-y0W(6_{$dl-c% zdODjo!A%ljmuNKR@_Hm0{~>l_@CgO9?Yy!Z{XY_thyu==5H4ww6EU9JwhQzBv}E+k z!)s*@g?PHu{39QUw)xpqXBYnI=LjAa3nG*Y@4mI$b?mjxL$Gj9#pdCC$z62{tvPRR zU8(Go0@r4}u{;$^5S@cp`X=;j>wz21|NI&L^J`$SUcD|gjRzTEN?I8`ZEY}%nN^KRz<1x=J4zN-l#*vZ8ODk#rEJl4zKhjJzuHlHXcNZoRC zqe39l!RGn^-YI~*2;!f-0sX5*avBZY?}pwMM1lvzpbre>Ti~irC!w!MiOFU3YkL*X-~9hcsbZzWrZE3=t7Z$xm)@b6-g1&xj0_Cz|{&$$D)^H{O*yz=WD= zd1O}#02ln+qZf@t8;R5$?54s zehe4|St*i7+evdB0t6VYtK!8Dk%N=h63?J7*q>@5aq~{Ry_ASlG zh?evA`~#4ADd82coiHaq+5M6cxD9;b&u>X`QPF`!prW9>f?Rn3I)e1}QH&6ow+U=r zOzupgLw7&iYp#TEKHL)Ule6m`F1L!d;^cgr&o2O`Nj%-LUjK_XY)2~ZJA>;{-@g8p z-P9=k{{8?4>dBMe+v{UQ>C*mMPjKEut6-(236sk}!HoulLoZ;{)VWZp^ax1oDIlTQp4dK|n_^Ka|;o9Hz(?XZaADlZ3kZ7fo&L z?#0gC3-mse8IqnK+>CyI_oO4G4md8w zVD8J8Tj{B>stm-8lX&2DE7rn*TC>KVEg6NJ$l;L+HqN_;^q`3aeez6w#gX5!MlXwF z@g_GSg7dx^hLpEjdn>sX0~i@&fKbl={Cb7%#-=Jzz{Q>T(Irr z$TjiB3-}|Ktqjw zwEOyl6Vy=YoUE9baQAR-WUq;aygVfsW##<*#&*E1{haFz1zI5~D)Ggia=TUv0v!}j z+G2e$iz0_^87G^Ou{-GvJ$~0yAds@cNi_Fz^*UsVZ`k3mk@1Gbz2kQc5=mG;nC?L%g zgGVEI<#bXtJ`?(SO3tn34|Uv*4^IDjd3o2~p(@(ZGCl-DfWA*gS(z`CC|ZAhy;VV0 z_6#bDN$KffWGu6(wNEv<6f7)44#iN4GA?inMl1Vm?JSrmdXXH}Y`YM33|>0Oi|K^^ zu~BPO_==aD?0@Prvn6*{@Ue0wnxdovqM1WqQQ%|KN$yL6%3xw~1jFk*7oZ|sSX&Fj z!1966mRWO~PYv>AS3(vx8JPeoC|UP1@x>9^^AlGh*v0&=^>J~;k0-Y!Az$Nay5)8u zOL_r~{o%g&Wr<@EaN>jjIBA=hAj1sC2e@o;xZRrkMVE`}7ArZcq$CkEQY`=x;WH@9 z8~Gl{!^=w-<5CgJSFaW*#Q(B+u%V=8o7wNz(kHppL6h_f8ou#*E~1ngr7`f1Mq@)c?IGr&Z9@PHTAxF4d_ReRUDB;JYQ;ar0l`2u`^e1caaYI zMTUk7zfcj%M-a;=FgL!*X^VXM@*=4Fs>MqnDsyYU4Q*dQ*czOFm#9j+n7|r3y4$l> zjUK-pp_mp@T}?_aEO^y9)k*jC?5`u(?TDc#ftPLCtWgEBBeLM+kDo;&-et{(d^f`U>&yJmv;P)g-N=mRrp~N1NIU6Y712kIChXu;& z7v{W%-X{}9kgU=JbE^TWE>a1>MCE7ZpNbvF=;aXkFV`W+G*y zpo_rB&EaVdtQ!6P95N|$!j<9Kquob#`M)bB|HCWM*Pzgtn7B@2Rt$)*w3#{bhU0^! zR2Dr?8)m>w*YK3`!ZVe9=o@PP?2URV#ffHM#eYy;2J!sVF%8;fMT=|cx9 z321vPnwp8GzuRi`1x(Xk8k%j*nxIggo*IQkMV8+xI&1tTsajjR(cc|6T2_nn?9ZW9 zGN?K5+|I#>38X;f0`=cK=_c5y664w^UVZ+>f!8+p;g`@$8ooYYZzB{CnBXD%s$~Mq zQTgxw6Y8dMuo<&24ggz}IjdCmf|5ou^Cd1(QCy8osV=ZzIh~zc1aa~*2*wIPc#DL) zEQFv&n$Vvd`XRD+;-+e1e2of}A@qA%q`ncvZU$efxb&I6e@$$)5gN08XJfN}a`fdI@; zSJLIN*SJtNF&`}yCETGab4x_SpK~9z6$h2gQCVr zzCL8J?5yLm0yA~jmoy4+`}_K^z<`#b_xtXk;r|I$98Uf3RMG8~iG~JCb+yOgN~-C? z@0tAjGJ-s=zqaO@LFEgAy)b`vDoHWuf3))H@OWdraTC9#jJh#q{S7#n%%R2~6&Dxw z_AMtgUW6j$QE(Y2&vt`!P!2XRsCim$jISr)aZtr1K6_?@j?voRm*=F}Y#?eP#AA5t z5k{!tkx?tfjCXwe>bBZIiKS&*-#Rb@E#Nu@K@Nq!7;fi zO-QKdbwT@xsDXfO@B3I$suO@}Y* zN#K=c1&a-sa+5r{6iBHyskXr_V)o?8Q|KpfPW|}bEAv&0_80Y_GS9-2#iOC3oeCkN z;B7?Bg)`=bd!L;i=hSrau_^eBj)^#JPTM{PrzLzT zu?HA7Li@}U(_Y1fym&<$#%LKR8mrcc*d;z(}Ka7g@J-5@G@|85Ux^9e{y7T zadj24?x%a#)$x3Rgl=W{2ip8menNqx*0z`g4h`phYDXQCPTF3ccASsBFL*US`YGnr z3DwJ4QD+I8(aEBp_!rSbxm>fvk9*?%g3HBPf8_~xLVZ+3c=!S&%$A{SdBhh}St)!1 zrgAz-4?ZX-;DJb}^+{D-y$gH;aQ?1sAC1&oCbgTtDK&MG$eLPQc=!|=OSDU?*BZao22P-R^}UEtez| z&7m`s*wfRq|JB~y)dNtkp!*Lo`pqN$6%VM0gu?R#pdS|cQ83=D`N+X31dIk27#;ge zxuhsXrHB*t=BfKsCg+D@JOo>mBiUW4XKX7j;zPstpjnkatnJueE$N>_WSUfH!&jA_ zF6AWV9!N(Q!QHFRMa8W+7+taaum52MEX%jpDcB_IAt7;e(~l0{(|>=dP8jA&Nl_7s z7yJ}`ch%IgOG^`=r#89P1gu4yvyDVIZh!|DQrH60cmTelBI@T-0+pg8_d=sHB`ChS zTws1lOlxHMvI-9Viyt{Iuo#6;nutfosuGiRg^4d@nfiV$)1^)AljSO2oqKhz^nYWfgBuS6M0Ak^i?GKjE1J#_gU}H#cIGtTFbuJ0q#GZ4>c}|JWtK z&cJczcPU@H26H=Bph?!uj1}T7;&KC%qB*!zptW5A?DZWR@ys{rp;@BFX#o@JF^bX! z&=3Mhs1p2FCkNw1|dMUX5-sac(}HYkUVR46cF!%0PKfMD8(*%x@A!A>!@e)4SR z;iG&!h@J0V9!072hv8=>BD?&YoH_aKzfS|{RrJr>3j>i##|c7? z%hP`z_Tdix#XW_-w~9DsI(`xrE<{MKhrg^}<3Vi@$eV~5(A$St1z#zBNBfU`$_lzN zm~9RH=+pjvNXUy&!WN&rb8qUjfJLX4{I>{A8ug^6qvaySVZ{qFP-HTqz*Vth>>f&roz5zGl< zLWAqNlfz}>p5n5o^}C}`2|u5hAz$_)1U}mHjynUF6Z!Br_;6(UW4K<_4Q{SHf=Wb8 z+kxktY_JvKP7Mt&Uv8F-(l59D&o!Z3wMQOlz9uY)e0MK0K1CqyaOFHc9n`q?00To% zpZPW$2VV3$@&9}XA3pRVf}U)E&m9AmQ26icWW9meajg)=L`l8<_sJ{&yu#Z=x_wkH zyiAM#pV=jZ!3bTKC!zcK{uV3M!nf}_|Ax>+3FKDd%+oYBan+x>3d4Bt-|@oztajkc zZ#*pH<6X5uBCQrJXqbcFjDCg=F0pc4W^rcuxB^HK-oB24ON%!V((cFqfr0(kP08ZC z(E@St?!UK%{RUH$n`)4V@nUN5uLik)7l7P57u78*n0V@e)001Pt7qMgKKqo}omi~luK zE}#xQn|Bts8SX%#$khvik=4jo!g)i%Elt9B_UfN~r0^2HL*f3J@8zEFWG`>`^``!N z?@6>gsN|>;PL_DJDh#R_Vq~?+SCA`mGm~)TSWF=?2t}Nk-V?=MClj?i9 z(p?ZERi0@8Cg!$D4n&{-4U!0+?_aw}M`s4MDge3!kcw1iN}vHq4Eid(8qT#moShMC z6=EO<$n;)sv6QH;Rz%^(r<~alxnTUGaJl6bqC|bxlLOP?^VPGzbMPWxJL}cUWa8eK z=-sRmqm%Q_=W!hY{aq}#G@N&=h5c?av^OeZ@nFNECrZdcyn+I=nwpw38V@x!rRZ5* zv2xe>ZEmi%#Y>=xV9-rWOXE6~r4A!Zz@lcxyT8t3RniW@ue3CJbo4Pkww0*>^zRa4 zUs_%!aN79nn_f>D8a8VDSZ0s-ihRy+cdn2WlUnPN78~-kqn_t&{*+{lV&njvKtdIO z;^PHM%22RW=YmBYo^J7OqzU9wkZp*e@%f@+kuhCF@Owhq$_KeD&6 zFGQ*Y9+}=xzZspkj^R0T@f9Q@4^fJGEcDoW#@c-pjJ4oJBnCwo>%ZWRD3DoQ9~S5g z13M0gHN)cKrtKVxxN3|^-P}e24G?9Qt%n;01e5DCcP#3x8Rb2q zMxlr+br-%j2}wx2hBJOga@nrPwj@Rg@rnJHACwe)fKJILk6!=)1TQ7ul+-8OQqM=lNP{ z|KI(sr0WyT_>Z0zLjz+JGt=~P+8=&p+x?RhyaC)j zg-2dMCpNGgvn(?H<@xmA~ZijFA>nzyB1l%U|OKt z;J&?h;)t_2BYa%^5=c zwRK8dRBRj^MziKUI#zfgDWVtp7SCc6@Cl=pes$+Gm&1{g=BHU(nJt`dEh`O z4J@JKroAr&R(P0jabLMnMiqN7LSqH!;?eKEA)F2Ic7nkHp6;RuJsDPGQt$A@h_P!d_NPQ9LDh3KX5y%5W zKye1|LA2}Y7m(xJ*T6bfEXTTs9ceyqVIMa>5Vy^B?P4>YBI~g~t{ii-GR(VoB;s=B zVtlz52oIBJ|N2$?`Rkg z3naSgAYo!L$q*wgGm{kBR?$mfByEGoa=-#ZcP3{1QxoLV3?GdRb_8`Ze3g0IC+!G-QyHoYs2z+BFJjp02RV9CiIrOtLW9fzrs) zcG`Afe@pD~%=0!f+_pqRvUzY}4^`S~^BGkWgDdS2$QAPV6qGEq9s)MU0&}yWh{n2j zg!#5&s`b1eF|!iqf{I%ONUli0i0Jo0#mB`tLBAGQ!sS4h^g9a-J%A@qfTxrQ4mJ#c z3&ZuY+F@mROQ&zSm=iYzQPU`%#^-K#4xeFAXEU5p{+n+~s`F8)`_{%&` zMPx{bM2Qd?3n4=p5{XPHk)nu1D09YSNP{s#L?Vey6^b+vWu`J`mNED9)b$y@>-W3w zzwUKk>sqVzE?&cV9_O+5vG?=&*bv23cXw}rNc8D>eCpNmBtoJ!b?95^($>TO+M)CZ zPE%RK0tIT0NZ}Zy8#s6c!8iwHv2doNP=ru%RJ7ZGS6wjyn#vRd@H1HE%gI`@*kf4#Z5xKeOjMa5DMJsXMKU!fa@Ti)@;zcsIK;Ot#!a`>01wOrXF zPD)Qp>xHWInm5=UFznhOdm!8%@Qz*IxczW_&XqI`d0y(;vQ113m*x)e+HNfR!OMo* zYV^F}F{!bsDKI`heqjm~MHtJd+}+(t`%!jzGeEh)$e1}|_S4%BTss-YDrf0;zJI5< z4YxJp!>!|F4yR8C!SqN$>VD;O6F7nD_wRLAym5I#@2b;gho;_lt{S*FIp!?3rCjHh zhdl8`?+X{%Sg)+C6rBAcvTfV8tylywz-*v6prxUaf~|+R8v=I9?z89GSgW6R_xJuZ zb@v`AuO)8B2%Z2E;oJhx;|-P|%)&uP8xCk`(U6GRJRHMSb#+?Dnl=izc2uy< zK3iBwj1Qjiu=}@)jkY&Drin)R3#Ft)XzoG>XJ=D!_VMW(9lePKVcYV^u%AbRHfAq#t+koy3DW-hSQbydWP!mDJloM_ z2Ty1D@hf`;`1z4g*tVYRtp@9b+CWBy$eYkp)+QbyxDB7u`->K4{!(z5a0rXu&R;Om zuiHrHK!H3@aVK$W4o4p+D~0(B1T_+@H8%{OsqGh>p8fBc8KJy$pZwbF5ZlV?sxG@P$r@%Ji1zy7iN@_=9Xgv`)0U1? z*2h-u9~`v*Sx0j9k4ZmVq5#3&36Qw2L2Z{O};uuhD$X2MSv5R7(KcJ>Yj!VDwJu594$t=mC`eiXy ze!v@>wKt}TYY)Xs-I$1QC2S(z;|Bp73j5HJZRlAjgl7?nOq(Y0{i)p?==Ei{*s7lH zTT1`)@AMMGGeRz)qmB&+`woc|Y3UoWeqhqTN2;gc@YBfld|u-BB1Qx!XV?DI&;Fg5 zeq4v%5toxW$bAR*reD*m@~?&M->>h?8S@Zj-9CI%gxaM<>UxO1Ywu@qoa!c`iN$Cx z-1^c_NKjC$;=HYGn#^6144svi?Y&c@r!+LOdJT{6x#t}IgKq<|gE(cFlhEv-Wvo5w zXc=!WULhgy`Rjfwimop1{>E_+A_L;R%jH-1kl7Wh^WA|Qs$LkI9==diJN2%uNJ{O2 z_kLSP35o8T2b(V+gLaKbT{w((k!TpXwl;OK2^x>&C^3$-uqz%`t!s#O@RU!gF))RW zh7GT_Co6ie9Ctqe9@G?=bYb19fo&bRwQ=8Nq`ZWj3&)6i_wJ$c6@~rcor<;5nc@p) zgETY0*Ne0E4;734XBVYM)eH`q2pADVv+mseKkBttex6?j!N_`bX^uRyvYEdh6lAK$^dI^+0V6ac%y< z=ma4l4E(oCZ-Di~LwdK`&&+UzgFThGq?2yEwCal(pjR zV~~f)7LRn1z58dI>h!Iv8~8foI*pUsfab#0Z4ZC;SG_UiYWE z?&;gZ?MC@?2}EkD6C4>?_W93Fonq3dVj+FzySfKQo z4ujx(f5c^|A_Q)6M7H8pAq7|46I=ZZyg=z#6 zA6vu$!uNf7tlb|Ew)N`}HrR;w7w5CBbOO-lPy7fnYJyG#4+sHmgj~?THY5RbgLLQ; z`|EOYa)<-y-<6poM2ZL18C>Pmgx3!Iwl;YQqe_>3ym~_uw7Pcs8D}`HyZW^1n>hc) zP7N`xV?pWNpsID2U)hMZa;e+!24w4fV0VCeN>DOezi)s4lyEi!kraYaSmm;U3p&?wWK+5GGh+LOOUcGv->c0vKV_rQ> zYgh6iJ~mrvQ^v=tD$FL!WWG$EVC1Jd}AFC^Q^7dbuq9i`PY zPeh$9aI)ixn4FqA07vCy{~LyK&#AOdXIM1Fn^MBF`^82EG;*iiX@P zYL2h>YcUzqDEdFCsm8J&Q(oIpER(h#H|;KplK?9SgN(VJ_{qnG+01Vt$!57UXQT9? zE49xCfmoDU;mkp8re`IxL6uQV$hqRQJAJ1NKd=$ek0YHtI+}ar*H_m2mOHD+7lcH~ ziSY9;EK2~Uoh?9I0VnBa=Nv_ z(r@oDu9DPy7+h=3rJ2;UQPAT}ZS6{#93$JO9qk@{!sK4cq5h82Z!$MJ^sG*&||hsqiPKO#|JKQ?;$?c~H~MjDBVnpE)^ z$*;!hJc=8y1RumDBBk8jPaAXRe17~S8w<9Df8&zk=W|$?YL0P@^i?=e5!h5n&eZQn z-@pgH9b;!KMvkgweFFX&`2)dH0Dq8pUfkT=PPMWN(AO1a3uH2$51k9nrbRu_X!+E< z^9-H`Z-8v@c%=46MH|WQ-62M9;(4%*^6%yFd^+OC6mh0U4=Dc1NCK%ve05f(8q za(+!C;tOV6MoLfgE~l`I!7ZzMOdsM@_Aah|je}siN1F>3IgyHriik=EFyE<1l2s&S zWfA1!%z}{+ZI;%=AtO!FTt=$u_JPA@zS}G;C=X-X<=A4mk+^*aTzP805zo3Pk-_S=}md^4vjhid&fh5VDToIbUajA=N# zxTZzLHs6g@*0fPp?Lx~aW4_)^tdu~ekAE-586bvhgQl_ZM#KcdX>R{bNq|0YvGQQ= zQf(jY+H=}_+RDT_$=H9oXB%!eL?5~zA0L+{*jU7WBBm93%{jJz|85J4qx`j%(#J{L zGfq1Eofx)Oi4u|^U;h%?`uGx^z+Rw3Kp5UpJioXAPRVE@#k&P(PeK_3lUn+&K)|^yK!XwIj~0s1#}d` zq-cu}M4*5Ggk)u9U5{r#-8i%t2tY<)IB*$x4l5521Hf8)+Sq6*Nnu{XstT}M{1Z*A%IpXF5r6cb(>u^W;vI!eHD`@mW zVfnDp9ok}4VFCe@Bg!LiY62K114I%mlOf!7`?Z~mc1w_($a}6Nt=~}-p`sUlPB&mh zg4_|Z`cNb(aWBswKFmxscu?dc7cFD62a|_O#oIH&mX-cseSLl2CzTNei3AE{!$1M3 zi9YV}NES2CZ|=^pZmkX$6=kn(_|9q^B=2t(=tR!3t0xJ$_doUnQR`1)FzY63(a_eW z#o4!abo9h;ubuCS@8b#H4nY{fI_NBOXF$^&0^D_8Js-ut6#RRK_=fcq+`x%l*z|~x zzQgDvyS@@%G>5}`BDH$6q@1fwv%g0Ogxz`~WAL9O_*T*_a~5l?PMA|s3S)+{6Lxu% zPWf7ERAgJv&-kt_^AkijG^*4N9(>q1hjipRo}&SPF<|K4xH}m*EaFLqCGV&r^$6Kx zD|Xqo!3)NP(^VAg>|y8t4{5Ts>ahEVC0%g;uea5qnIq}EpfT9CG(j@S%zu1n{0cmjq}PokRiy!AJcGe4)JC5GzzAsre^Yt+>dAH{XU5!Z6?cc0`~=CeztK z+2AdG=uGeI;0cfrn5BBKA@2{@ET>V(nhMbyLb>7VbhlkE8OyT=wV zDb!Uwr~0YJ#>d~h$FUqGMzeHm>SZ+_boixLPuPv8YQZO3H|@hB+10aIA<6qnuqO7U z$e816gV>j1Zrw8aH3N1$kj&?c|M0W3Qv%vS5H??5LIqwE>g3l^il9>AZ(*&Y+J5*R z0dj&$$RDRD8?c01BmME@yfxxfjdbyk6ycxPP*mV{f$P?nnIC+c5DgSGPp_*BP)x%+ zts(h-n7?tSzL%E*f#L&0>V{3?hUS|Zg0z8aOreC84t>rvG`|SUl!#uUAF<_PZUj+hv;8s0cJA1fs2*)x=;LHYR-!0_XJ zAKt%qzU2|y*L=$pdGoY}GE9szl2=dhJXTm>ort|OIod6$$3GH|vp>fLpbRp)V;#x3 zQ0vDLW@-S5Y=T#koRARC#k_tHp2*zY9^hSr{0(2ZI_XB#a3a1q<{$XxzKohrk{M=2>$zSIg+)bs6~LLP`Fs0sMNxGM_*LXixvx2IZ_1Tq+LVGEwi%CvaOd&|-XOH`;`G?xhSaHc-IW55x^ zW3qiu?PZ$_g{Y=QuL}l)9_`xLxk|N1KhJ&Y85?_Q}wWrM839~<}P?J^@ z%eHgU)#Lk2FQ&x2<)E-#Q%@8uTK^L6Q@9){5E%n2NFsr9!Ne_71smg=W(cm(w=f}o zKd0)&9kxssXDV40oP$3X*sf8W`o-R2z6_N_XWdSvEB6@+ytg=9RLFH6`TZK*H;z6e&C*8H+ zr2V2b9RYgfVTbD8C(4wE>8oEpY52sBXlG7*(Hn8bJ1s^bErkWAANwr~_N3(II&`LB z{|-Tj_J>LuRfD24&GE#QcXT+HbK{$kWx;fnI?>Ggto%o6fZ>eH$0y(TmKUkiG~RG> zUL6hzJECz zB-?T@o%*}H=Z*~tHMXhgsn-LoiUafGXTA92hTFydLy#d(;3B0KcPy#0SWhXi$_Qvk zZ640Co2TW(MnhZ3UsQ1J0@o}4@AP{&l|9T)zV{$Fq44b5CZFMuBVvo`_4~CrO5(j7 zz8bKU#Ft4*f8HteyVrotkp z;*@G3?8A=P{dH`4h4UYU;4iTL=}(iW^pOrSNFkMdTAj6j#6*}HHZKa7$Vs>OIjdLr z-c+%Ky2rFVlRoOE>niWm`wuTeLJn6iDoPD~i#)vpyU*F!?>nroUR}PWxb`Puc~Bpc z(0Kc5Nk#&)8}aLB5Ln8)2@&KA?-_=GcF!>>!8GHlfjv;2Kw%?V$Q*3*vHs|2rEyt& z*~hQFri%GaECU6jV+=0)7=KYdJFrpuzT-r0;vT*e|4~eu4yPDb8I6&7qu=#<0LSHU zKE+L-mDhqx@G0XE+LlNZiIO;Gln?~t&se8|LOI?NnT!`}CJwNbuJsF{&Rsp|`mx1g zuxIqoEZQ&>vae?IMTgtp@#4@~`lny0a|d>zxr$oikvkBLMR+#Wn!c=_FP zi;(f}7iG6ap8nL~74C=iDj~Eqd2hPSc1m&&U0IShj(5~Y-c~*+nxQZ3ZC2F&P#>(t zVC+5GR<5~hoN1T)9|4>gUkiokKz1)+Oz*wBySn&Lbt0ZEL3>_i3WyZVP7J*(U!WlA z%W{N|caSpzM&MnQw|&CCTLk+C!yTWn+?*Vp#+9iTpaOo>+ZqskQMRZ>UG+)Yd#%gK zzM%mvy!*6J>7TpNNWUK#T`=v4(IrH>g$4xvk3k+ctgbCm5%{v1nONdk^Fo(K>KxCA z03>F4+Xc6>ex3Mo61z6fsmmsFfCU3J9EMW>B=Hr!E@6)Ppc%*8uVd=&?K37en-MzZ z;A*qVUx{mjeC3PLq3LZ_b`EEF)=%B^`lqS3;zCvQ|JWf-&7Vc?F^SC8q(ug57ab_?`%}lF4oSY`G|Hbx+qXT zQ$meiQBmP368#j93L)dfqoS=`XWD1$-LuK;+31!B<%g>SUrG&qiq8xa3W(0RWljcm ziw2fDTntlizGgJkm@(9r$<$g6p6s1$6>UF0oLYFcDGNW%YA;w!?gHZAPZboO&JO>u zkTL08O<2?C*xr;qgiPX9x)WxX(Gv&X-}Wp%&L-P3e%Ve0Cam`}*uz?`?crCwR&L%g zH_>BdE&!oeNPHRn+a=k$=~Wii=>-aNnZ15j8I1p3+XIeNz35Nnlrz;z60^~AnvmsP zdl2Ee=LgeYg+Ju_yWDKjvLuKqTE>T1b28$*dv-hw&FZEnK6Xl1?}nKc4ik~RNUIEe zm)sNhIVB@;(ypC5Gm6C#8)PJciMTGg)~pzEE!lrSE={Vv!`T$!7n48rt&qh4PnvzxdaZ1|o9-ko) zzfp^ted9PF8^mlb0PcD*al{`;2!QeeA#xL|!~5P{c_~x-zTNHhcKd;5<7T3q#iV@% zOP1)JI<|^IAk~W~2FdRYG;VMvByLZJj@1ofCspsZq-G_*rBVDmR#WGuI)ll+C>wR45O_2G4=vrKH!puo}QTo zvH=iJf!?VgW1s6h>*Un-hzVt2Gyu?srvZl;6ouQjDMgCIOi6#LtRSQ+*)-ZoUhX3R^IyDiIk<*O@DcIcSq^*D{SlM00q3X; z@5G6(?=9XFv0`2Q?5KlMB~Civx08_wk}Aei;X=~F@*fUGHbDuG8rJ1phV(DH2zlPEd&?K+-~nhziH&80##cA{C7Q=FI9`#;KiHN!VN zqLPHzeQnX4`%(IZWf6H-X>J}Cv~L~NNF3LJ=(~;61)OUtr>Cbc5TcnbkDk7l6R%=*P9Sx~cqKtlIT<)O;x4xB8voFD1TD%o(M|p- z*DC350_pIs_(4wd;*!^1H4QgBq5MJzk(p*<8LkOvKa6yu<6L3;= zJ0x*)h~aCf4q&Bl^sI{O6Mc7N*DpNQHWd{V_yGco`k`>)3RHo3;BpK0ZwnBWAhsx< zyXrXUa84Kzu{Y$%yLjT^5i~_}*fh@^j534AW`*Q%c@SfTOs#wZ2@pMK$8V6`W_&$n zV8Dvz7shPd-Ox-Z>mjv@H3QLE#^ROtX27uyU}X^3o6vWG-?-xsH9}6m^kjl z)O_yl?Ss7umB0JCXEfF~=8LrjPlp~Xo8ndzSX;2niJ5*RV&(!t#K(n2oyd#g8O9zE z#V9UiFBxJkQkhI@ftSc*^IyhvMTxm;z;-xN(Q|$;2kB$ceGY1OM|z2@Yxg%C%N$rYTI9NMaXE;s%KGiu zMN*B6j`joVli==q(f^C2Q#m{0Z-F=LSzov8}UnTUWf>v>F<4`eXnAMnPh$8 zVer%{Z)*B+UzBnF`(-gbverLEwGCVs0${y>c>f8qNMJp{jc+|R6TnKIGL%%6sC4g0ILlI6L(6!zV18+wWcp&=!XWyVHEuiMs?fvCpxN)#><_t?`z zNMwEX?2g<_I)MzMo#8T5GY+PN0M}->YX6fHB%RWS*R zQa{|ffS@3?t*o$ca^d(Mgm#ztGY5gDUx;#V(cejwuVOoMzOpvP$&mKS3Rz+3H{x$J zy<{lJ1TwDStd7QWD8}6&PE=uBAZ|Uuh|PM?Wurx#ZN(LI0P<3tbs~0c7bMCC3?AKY zN?+~jZErE??qgF6X*pC(nw^~VBf4w=T;p_i_BWbpr}D4ZQ8e-o1ocF!YR)d}JziX9 zTk>$YuJRzFI&~C1fVq&PFMojs6^??6F~FyWWC)b^TgD-yz?3L0bZcfNiv>0dw?1zvxBVDL6 z(BcTqeqmO=yyfqer333JK?HOnD8o9~X5P4B*ZHx$kg%yOto|?_d%3WI+-CJbpg*fo z<@yhCAp5slGB`6kX)_1wKld|q|G`LS2`gphHKtdD>d!!|4VC{|X4=^PzSu}^Hj z6uJyr;$#`?;eY^RPb56r9Hs5tcHLZZSQLWFxuo}hl%dU&vOa!neLX!~e=n0$jw${x z(!WWk{ls^87}j;JnVvlGK)Gq_t8l)@+4$gLLI{pt!_yk4^W@sOw; zeZ;kXfjEdBm|9M#4VT*mr@%Wm#%l{?EQPg)GbRAd0h(e z(K7V9p3&{p+NBw9$xmD5uPjF$|J<~9wtH3kP;;DzCz=R^p{Aa^wy*l}7gh&mix5WH zOAapdAEgGB(&i=JHu(x}P3zvT7Q(xJ$@aXKbEKV#soeVHpek)upz?0!5|%bWhqm+H z+8^SHmlHm?{(4-seKn=4BnOtiRCl8-^Onl*2~~wXrTS2flcjH)cO^q$bYq&d!ae_QG_aGZ|a?QG^Lf@VX?k{Lpd+R z{^Z*tknsR5IW*Y<`dm^vGCAE=6SkT@-*I#^nU|zMJ@d&(p#!P)4IAt65w(^7eVff% z&&k4rKTv(SEKeCupOn3=O7R-c^|j6*n%x_J5U;ppeHo09g8KHqf1&8MN)dDBBcr!w z$jNfwsuu+tt}wvE1!S22dYV7O;oW}@2F^ai(%are>9G@l5Q>+bKn*G&7)jy%e zPIIcrF=S)5Ktej&?aZa-UIuNle)*|vvwC_}fOtG|q>i`$ai1Oq-G3idG_r(kc6lvi z!R&4dzmPt)*KOpc#w<#^lwD-p zsvnJBXInF6;dQ^CRww2-t$Ul*Xp|Vm9IX&;!}Un^PmuYg3k8B~?JGMA_6^tT9y#D> zE4at_R&+uPx#4AnQmV}#H6)(hv2eIZUu6^6`y-2DahNr;{NNQK#yyVT4bFCN_OAPF z-SW(Fm#SKb_xft5zuxCk^=OndqVQ4X9BotPt>&VgH-CgHl^A?<{Ld};z#d!`K4#>O4gtg z*rP?P*OrZpwN=`te3?N>km+kUIttNJ3bNA&AY(W%R=4x@w$pvX_(SBhfKU-DiNS={ z`pJf<2i$#_pF(4dTx zPhcwS5)Mm!DwltEhGHB-O-Bb0lqzwTLm)z6pyc+ zr@DG_r`Mz4bg5eg3k$sT$tSOeRmd&mdEH32N0paOKsE!08Ym}F@f0pRhc*b2W8!~s zDEzmzn8}_)5cP!&Lcr0}%y#JhMc2NQJFhW3SKGLkTDi666D7YPmn&OfbH3V}S9XR1 zkDRwnMAHeh;-ym}BJ@M;iJ6O#klI)qOg;Gwx;l~6!=U(-7CRaHgO~>jLGI2$*G@*b zjvFdn?s6g@2!B45x`ml1j*?xaBK-Op)9U9lS!j?~FT_fvfL z0y}d%w(6{Hop1pdD+NAZc$}Dnf`X`tsUg3;{ttujQn3y#a3)Zu_c_r2*~BzCmdEs= z>#V=@#Dlb#Uh=$0yVWPBMd+=9{LTz8n{cYS%r*Yx+ar)cPH#z~rz9s%d>9_{@#9BP zk1nKi0$^JrRgCfCReDh4kr99M{ZH+c=PEd}z={hnI`eiOA^C4y0`qp#_g;2!4<-kt zi?6YnPY!GiXFPcMhib2^Gec9!YL#rkZLj5rsdgt+q3W4V`AB8`zfUO*4GqN90N|CV zRg!nNazVKb@48Xf8Ow+1=@JvHjuHPccjmgGMV|K1nIhx#v)9?IM^-j@ah}xUy0;(& zYG*`v+w)Z0FPlWV{~9=dG&}9;9Tw@lnxXgNt4N>(2}twQtSlPLEW6l-#mI-gZ!c!v z!39h@SP=|e5A=HY+N3Dc3e`aE*MdSqS>_#L^#7}6h%mr==f8zS_f>u$8yd0o=jn4U z9KXzE!7x#CXN@^9#)Fmf3Jv+pv5yy2X_W@v?m3%`53%jNg}Ylz+^aK&HVjqeMa zC;v6qvRk?;fzV?6v%=8qo_JOs^8U@hUtVC3xB5)$U2^)WBFz|)sQ zV#Q&yCs=6NMP7r+GIcD8p(-d+P+ z@Ywl~E~a40$DXJ-lz?K2Ur0Qr@ldnP;rU?kiectY!l$IQO`VBX>%(5KWM(VMbTFs5 zQqsjx1S$~&up`mCOo2n%nk)153}zsyr5Z5Dph3QPFZ31lE3Ov02xd3x(7p9sE=&GGm+57sbTzyDZd*#8mujf{&%SKYZ#-%YUqsoq@+ zN4xVD%>KIFyW}60(~v^ABo>&v4cwTLXHmeKLu7;^Cp$|f0KiydtJ z3au8^TCRv)reE$iv7vA=6QDe@Ti{a9TJ)c0vXG0?V%=;48Qk-S8`HbS6vX#AS(Z(T!on47m?QiT&1LxL_MD$sZ*(xdUc|A8`ePL|+(`Ns6r;Osl z*9~{`ODAXfXOCW`y_AvK+b#7YI5{}y69420Hv(&d4aLxt*!pR+rDvZVmCDIdR(V!j z6O${!TJhrM$&sj!DO)M1JpO7PSlIliWd26j4LPm{Zdvo&H14d&xr`4|vLCBn`yUPo za|)<6H9xR@ip(oV^}69nk!kfm0&AZ(KPp=|@Frhqf8bUD6HN~?fYq~$9`ACSstwsY zIi^TbGzl8mU!rIf$i4rt*z|?Jfnzs{qSwY0%-AAUb+k6!g1pm)xa{gn@;31?-R6y*!CY377zh0uI5;+>)ii%Uik)28r0nrY+v zAJ-4vc%-&bvWA?!Q~COsy6S<+-!bL2+cFNPZ6Ou{eUtVPK2^%n{$#Dw?Uz2#Yi9Qi zWx4W~Z4Rwheo1X1e&*2?W*V(JjqcP95Ou9z0k7&0IxMPDuIjn@J_V^Rv=To{{G#o* zdjDZP?%RCeZX6>Orq?nmLa31-cr@sYm$SO}0?GKh>-|_G@AJqH%ymu#MkaORiH{TPONOTpI7H z=dEhH%rz0tvx^6mk=m(^?1E`Uk9R2Gr210d?!PYgn%va{-usLG-SL#A`QXW+YXwv< zh>v8xp0GrFHKU@d>N6Zl-EC#9UcFUGqJXkCj0p65=GQ`MOrH-ulQa}_qeoRb#HiPW z%k7*!GtS7AT~BLOS=2sRtulwtNJo48^w`?JN2Z*-O4f1PgaKX5zLE7ia1;kvuJ7Nd zgD^4S919-Bbu^sfyp@xz`Wttr-gjhA{Ncd9O)I^(+fXuT zsrPs>dGou?77R3!N+wr=^|j;HPdEYIZ(H_L*sUs0ZTrsb+wDmn@#Bm`>IMeIxIevT zveG8MnXs$8q<1l?Dt7pL;Xpe*?=JN#!H2aUu0Ov4x2vInAbQ zQ2Cdw8sY0}Hscnq-96aiPC17(-XkYk&2qIo{kFE7WbP=*KYE>Btf@5S>1xpMKA_ix zmhy=jDTpEePTs=z1CbB1%Z&weBLBS%_cPgrpYFB?IUSEC-Xe^fw;nLEt?njd;2UhK$YN{xrgchb$n8gy@-9Vk{&y?rjP^{!K21Y>vZ?H<4-WdIO{l#Gb{{)l zF!=WRavxoLCnJ03zVMpuoYfw~3>?dm*7CD+dZ%ulw83*1!pJrBs2VnsB~O`}+;+E4 zM&GEH%^TTb1*^WY%SwhbneV-AaHeo5v@pY0oNb?%dQxCc#qk2aZl#sf*2dHzb9Rb=$;x!RJH#YNT87(J2#tnnu2H zdL9N#NSh|t=6@@T`C54pJ6N5zdoKc3%qi@W8e zlb4+ILTUSv!BZ#8+UfEllM>?Rcd;#tId(HUn%kZ{Mn@YevNfdE#W3+(rE2uz-s^k9 zBpVwCi{o|YrJIWeXC}vNe7_bYD__+$-*Vw0P3)(u%o3@FDMp5IGaJkF*P>R&b0u$E z6kl$yd3%Ul6t_=?LSXz^Ia}M!?DQ=UixZO6EU9u|y|+Jj{zyK{m5m9}Dt_1hw)`2| z+nCy~$|D&+`k1ezz;N$#YDpE92WT_4E`G;AV6J@^PvuRc7^ec_AO;gQ64Ox<5;&1j z+N0`BNl7VlxgHlM;NVd)j+m~M(4il-AB6UKd(6sRt1xRk!Fe;3KC15YGo%MFBJz3nlET~7))|i;&A#gBAm`-VZvK3}_j|EX zPEPn&<}AZ~Mpo5@OX5DoR`Z=bHKQEfb3W8{AAVBw=3hQtQy?*W!99*gtFp#ZmWO1* zu)EXiZU6UO+;K&@975MzxGz~$R49YvGIld%ZvyG7J>-kXmt%ZaN3Gmn<$oFe=pa*C zy3!Lc2snUef9Q6CMTHrk{XUh?xB1P#iCLnLY_%HBUA-9a_peuBVnp^}LxVV3RnLjq z1p&S*EyLt;kr&ioT%#hzo#APzSPRh2@_%&ioLseG&iOR4S0OKR?%h@Ck`!er*?o)j zy;r!mV-;#kjA7n`+*Y zA79u+}Zamn$>rtN;X+zv9_=r(yB8Aq`f)13t1K{=OEPwW>o zb%!4IPlb?zm}vWM{*Ydo_|zlslNu8Etu5p?N7bWGx&~*q^uBF472YOcTN7`_r%F^?|srS z6O+PAgNFfh_~fut^%2KG%(f9;SX^wq@}<7wqHE#A*&j(~FKQ)9cx1e~ZZ~JTIcR9P~oTh@$X{_?$D*(o-k zzqi;lTiL~NI4brl=b1eI!{%1^o?VkKo~>=DUG}%ydyHM^ty86OT*4uflOjRo;Ug1! z%6;bU#7ciq8F!uXrQE*8%$`e8FJXA8_R+`6gW@fhuRSQHzI&3Q6m+1IUcGS^t^>a)RgA&JTD1z{uZ+PhBR(zow>@#<;-3G-g+>@d9$H$SLU;{cSrxk zovFFvxPMZ4^{~e3IR!HB)w#BPFy#=_`&wI{zI=ejP7UU|E{s~Oy_0TM`qInANG+sb zdF95RFZcE-DyFuUrGJUDlWhN^Q7$yc!JAw8*>w4*_G*Q(6_cjW)guf#$0BoK7<av`n6@^Xtt5 zE$_|UyQ4Rq|G>16*P7W{ruaI1s;hlWxEHhOjf{;q%R1hE)6j!s@$qL4Yyf$(Y+KB= zdE#-u>w~qVj@9^Wq5oP%?sM+i&7`HvQv-KzaJK*XB7AtoXusE6lcj*APw#XijtvAQ zyyeK#Vm+zC_ler%!pk!Gbgl?d`!>lDdIKGLJ-W2QZTtsK{frE+8>QA6q}D|z9~{3u zzV^tx!Y*OhXR#>)V;+pK+!q&;A%99?Q(Z;ot9G z+uXmBHmH8?`=tlc-`eCa?w!bY?riEQj@V&I<>YJ-vJTcK6a6i~Z75 z9rpcJCqrMA-|As#y71d!{Gmu`9%uDDOPautk@tgsQ$* zhfrS{n>9roX00Ln7yfDcthbS|6QS8f^49vVsL693A_WqrK;H0ee#F19GHY(wk3ZCC zX@uxfh<^#V5kqOP{xa3t>siGA^UM5;+wFS}8cK4HubB>wjRjO&KfBDOU-LhaFXOL{ z70-u;LxDOEy?U36h5!4*yrA4ky!HPVpYvuh1#43+bPJ2{+>F_W_&v57y1D96@Ro84VaMt3*MNA6Fx;p}QGNxDzCKdYt0&-WOlSXjI4mu3R z{8Ca`Svfq27++vrO^$ZICVVO&t=ORIj#imNKSRbyrvmMYP9a^%X(|x;Xr{ej$%k5v=OrhVy zubw-WDN1jgsh0}=Yw>*By}H(0GvUfUr|&8sq&;W|agI|QgiTh%pn%#+a+CdDo#F` z9c?>uyxg-edkl4Y9dtC}@VwQ3Zr#7%1VBpDJFRf81DH?>gYh$? zQ%rvbqEv6)z55NJcu1MB8z>=700zoh86 zII8{H+9AhRamkF1A8!EQ^sdrZ8sUBWHb-b0JLN?yM>sL*)=7zr-;9oCEt1tX@KMYD z7M%Ba*e%oD{`m>Sa8!ZqtNh3slU0-Z`};Fl3j9YD&yTyF!))Gf#ZH_@R2H{67MFb; zE!msztJGY7rsQlx;6=kMw02@~iA!E(!h%9XB`6s%;0II3nRb^?l0zz+j-_`je7<{O zX=x8gU^TV1hS^8Qx?Uz=KZs38sDmNl#kExhl6l_O$Cgc(n?_4AF(wN46t5m7lx>l_ztc!jHw}q?^Z3h_11$2 zA4Xbp_ra$`kmR7qITqf2F`oDNynG36Qc7uQVtcz;(Zv(IZ5_?h(|4TBZ;`^x;nAsq z%n@+kAB%&@a&g;5M3P#s;(&r-Yd_PYRQib}Fd{i|0+0f5S2D-p@=ZOKyV!qoH`pyn z970YOAUb4%mGU*T_T*cW#171RKT&ed1}-Yh{QY|L_uyb%+QFN89pE;k!+nMcx^XbW zNZY<9HiQY@xzeszK8on?oR)L-ACDNl|3Nx!xL(-WMl^cPS11>p6@R|!f zrp|u!XsfJjT-F1xoZ~0y9U0>z?eMq&=%<$Vq=)$9zSCgROC&SXpiWF@sjh}XIo0~T z&n&JYVjKx3e!pP&Zt3sK4_sTC9~a|Y{YK_&An4nM)tr&Bt!U!)%UEmf6WiQ7zkg|0o7;H} z$}Ys0O|2Ks5198{dCR6+^mUsPiHd}S|6a5!?VyGpn)AX8sk2_#D9=L$6^2j#z03&4WeFD(NXHe&}eL?VengyfS?X4KDp7TJ1$=Sg?ZnupaARTkr!>X7an!6E=2s zjZ>$#!9PX36mqf?(zDAY|C2X5AG@});Ch#@ZF>45umCJo=<kd)%x(&U$L+n*wqYkTlJD;8qP_TLwFjc zG)+qaFY@3+BTyOrkLCUZ-NbvkVLR;QfSi1p<;zv~@yR>16`O~%&ADCsBFVzS!Z3$Z z2=g01thU#be)V{O7+sXxeC|?~(fMR`4(1&S3Q2fL)J~x8m*isHd9pka((%Clep|@b zhrbnV$Mc?rpsvsBTnmg|05EzNf4<{Da6~4ov^-XXJfzKw_fuB#HAnn0_N~bIx1ozc zg~aLJDt~m3?u9X=&kp`>F1hpBe$)iV)0^4##Pmd8Np3_!~r}XFMib2HM#Q(g>=iDyk zNV#+ERftLx3n7{(xZ0LJNXs5^Elj{y-TvR1qU`@WNhFbr-o>yQB^>G}a>sK_a`mFt z(7w^y-b}tpO(yYSXe&Rz^G^rBRNe!p+tApU@%S-;_=R5-mbQDtWb=?$A|na}G6G-) z$DbjAtn+$XsI~_<6kK?vFR*`~kRE zk6K5TudOW?gW}Zf^T!gZ`#Nx@&wfgy!sP;MkTuSKSKSLv`o zz(Dlo{Q?>{Acg#XP7!GzHkmfDO3N!(_JL}$38Oo6M}6MV06%~IrtX)u2E1cGc42V& z{Xy+3`8o;=xc-({R<(L=yi6=U;`Y6J12@xdJ02l6o{@i$a2BJso)1G~SoG7P(olK0dPiq^+vwXf{S^x0i%xBMT!vK|T zl&!6!V++eLzQ|WaCrQCBUhHUcrUKc*cUQ@z%qLvhngvc9-9_yw} zZ=ttGsK>B%f?+fT(*9>m94K1FsNq;3#r{qwRKme<2=5zwI4jrv_(|HnMB?-%bLaYX z^6qQP&JV8s-d8mJr~7IY(%!_9?Rtq-NEW`~Su8r;l;hFVgA9VmO^BGcb=$T8$m;JP zE0Q`@AN{hVsj(5`(F375C1ESw4CyQEqyz}X4;QL`Y^?EZ1kco;KQzR&`EV|&v~Cr} zefy-a@v5v$V#vm3HE5=Xx?aj5&?_wsrDu8}EHiFVO2lt(U}Lknyl7#_j*mrb!$|EP z9Tg)H9~Wam8qmvkvOIYWXM#lpQ?f*PS3z{ky%&EblrQw(BrMh4rFho#CCC^AlEN|j z>b{)Ziqztl(K&Vw4#My9?*03QckgmMDupz|C53%SK1;J?z1UXfN_@r&>+~A<_a{hL zPR`EC6w7ik;vtiA8!|ASruhgxH3*?2U8?(SZ3QPP{%k=`_Y;p}`P$!QNP6R??A`)x zdHljMd~h*Hc9G6BVW79Dm{{7QM~JS6~|z@p19Ug>~yBYGL~_~t?cUg6JK6sI-|}B z#W`l$zdBY?_U`vKJe|n8X1$a#r;k__7@A5PjxJuDwwyQ&COmcN=+DVpw>A`QTbOT$ zF#==I1m)z~`V(=rbaCd^h+Q1evtqh`7}I`vaY?s{h!6wn>S~kY1TdUVAKCWWsso z%%H;iU5oL#LC-4q@l%+G9qms+(?2ru4sn1m03-W)X?Hit48L}%Zi1%8LB_+UC+YoL zHCi&_oc$O_T03@5&Y5Z&e)vMHjIUg}lkEru5Mm_2UFJ_(O)l@oHj2p*Pjsc9U8=&2b9RfC91YStF|IzV}>TUYQX= zgnL-5wr2n<5x@2pp%3g#>A1lkKQLpV+N%Sz9%pcaS{KcmyvmWZlfHNec3&G-*ts(t zYZ``PT|8)ct5`CvVjUO^@ogTPFPYvEJ~lP^y?5{!nVCRnefHZ^VH{tD0LhHgc_Y{3 z{ly`{dZG?b=014CQLVAA383SO(ER$u>Ijeq$RIl{JL2)yotK&s;13|vMSOdhkr9B& zEWjKdM(l%Zog$}g_-wH`wCe%z2yjOPfswKXmk5<7Pi{fre)Hu^ zEYQMtD=uKeg&0LhOe~B@_{N|2U~d)oATbD#h$_pmudt>PRbB*W4~(aYJeF$@Mj<1Y zq+dhq_7jRqDPGjSaZ!FEowI)v#sLm}m*$4-Qmc>|jxlGu=t5DZ88hTQgqqKrU5>6xe0|0zMO b`0@vM{o1p=6N7z8q(cV}Yvik0U;n=VO}o5N literal 0 HcmV?d00001 From 527957e38b6e8b8bcf49f3a7f1f9cbac93c982a9 Mon Sep 17 00:00:00 2001 From: nvmdava <86460216+nvmdava@users.noreply.github.com> Date: Thu, 27 Jun 2024 21:22:11 +0800 Subject: [PATCH 252/590] Add Zero Bubble Pipeline Parallelism H1 Schedule (#396) * Zero Bubble for Megatron-Deepspeed * Remove schedule variant * Integrate zbpp into the megatron * clean up redundant lines * Rename variables * edit README * Undo readme edit * Minor change * Add an example to example directory --------- Co-authored-by: ufotalent Co-authored-by: Wan Xinyi --- examples_deepspeed/zero_bubble_pp/README.md | 51 +++ .../zero_bubble_pp/benchmark.png | Bin 0 -> 119682 bytes .../zero_bubble_pp/bw_split.png | Bin 0 -> 59010 bytes examples_deepspeed/zero_bubble_pp/zbh1.png | Bin 0 -> 225315 bytes .../zero_bubble_pp/zbh1_pretrain_gpt_1.3b.sh | 367 ++++++++++++++++++ megatron/arguments.py | 10 + .../deepspeed_zbh1_engine.py | 110 ++++++ .../deepspeed_zbh1_schedule.py | 148 +++++++ megatron/core/tensor_parallel/layers.py | 11 +- .../core/tensor_parallel/weight_grad_store.py | 34 ++ megatron/initialize.py | 10 + 11 files changed, 740 insertions(+), 1 deletion(-) create mode 100644 examples_deepspeed/zero_bubble_pp/README.md create mode 100644 examples_deepspeed/zero_bubble_pp/benchmark.png create mode 100644 examples_deepspeed/zero_bubble_pp/bw_split.png create mode 100644 examples_deepspeed/zero_bubble_pp/zbh1.png create mode 100644 examples_deepspeed/zero_bubble_pp/zbh1_pretrain_gpt_1.3b.sh create mode 100644 megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py create mode 100644 megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py create mode 100644 megatron/core/tensor_parallel/weight_grad_store.py diff --git a/examples_deepspeed/zero_bubble_pp/README.md b/examples_deepspeed/zero_bubble_pp/README.md new file mode 100644 index 00000000000..2290860783e --- /dev/null +++ b/examples_deepspeed/zero_bubble_pp/README.md @@ -0,0 +1,51 @@ +# Zero Bubble Pipeline Parallelism Tutorials + +This folder contains examples and tutorials to enable Zero Bubble Pipeline Parallelism ([Paper Link](https://arxiv.org/abs/2401.10241)). The key idea is to breaking a backward pass into a $B$ pass and $W$ pass. $B$ on one stage will only depend on the $B$ on its next stage, compared to depending on both $B$ and $W$ of in 1F1B. + +![BW Split](./bw_split.png) + +Currently supported zero bubble schedules: +* ZB-H1 + +## ZB-H1 + +![alt text](zbh1.png) + +As shown in the above image, the ZB-H1 schedule cuts pipeline buble of 1F1B to 1/3. + +### ZB-H1 and Its Variation +There're two versions of ZB-H1 implemented in Megatron-Deepspeed: an official version (the 2nd schedule in the above image) which does a uniform B-W split, and another variation (the 3rd schedule in image) that does B-W split only when necessary. We provide the variation version as the default implementation. + +In practice the variation version is more friendly to a synchonized communication implementation and combined usage with tensor parallelism. However it changes the ordering of applying weight update of different microbatches (E.g. for Device 4 in the image above, the ordering of applying weight update is 4->5->6->7->1->2->3->8), hence might result in slightly different loss curve. + + +### How to use + +Simply add the following flag to the options to enable ZB-H1: + +``` +--enable-zbh1-pipeline +``` +The default implementation is the variation version of ZB-H1 mentioned in [Previous Section](#zb-h1). + +If you want the bit-to-bit exact semantics when compared to 1F1B, you can use the following flag. It might be a bit slower than the default implementation. + +``` +--enable-zbh1-exact-semantics +``` + +### ZB-H1 Toy Example + +Here is a toy example for using **ZB-H1** inside DeepSpeed repo. + +Firstly you'll need to prepare some sample training data and change the `data_path` in `zbh1_pretrain_gpt_1.3b.sh`. Then under this folder, Run + +``` +bash zbh1_pretrain_gpt_1.3b.sh +``` + +## Benchmarks + +The implementation has been checked and verified on various setups such as ZeRO Stage 1, activation recomputation, flash attention, tensor parallel, data parallel and bf16. By approximate measure, ~10% acceleration was observed when microbatch count is twice the number of pipeline stages: + +![alt text](benchmark.png) \ No newline at end of file diff --git a/examples_deepspeed/zero_bubble_pp/benchmark.png b/examples_deepspeed/zero_bubble_pp/benchmark.png new file mode 100644 index 0000000000000000000000000000000000000000..be46817d75d3d7a9fcc4144a7fd330f7459e5b89 GIT binary patch literal 119682 zcmce;b9iLk(=HsF9otSOwlzs6n%JJ$wr$&)iEUexOfa!+n;o2f&U?=J=Xrnc_uqG2 zd$;!L-nCY(b=O^0t2*+#f)p|WJ^~mR7_yACgfbWyEEgCU#BVqV&?iX*fij>caA##H zF|e8`!V}ON2~$lOGkJM1deCb)Fz|3oFsOf;fKGhS2?hpP01gHTdItaJvjT|!dliU3! zsVRZmtKQEKl27UF9jZG2Pe05rjEv_yOyshUBxr`6uc!s=WBgFdkH%)=hTV;uCe+|* zSx`mnf!DT;iQRQbmPO9gnXL&J8NF0rcIA@4i2pV*3!xN&V2$mY8H82ZN1(d3>NhEm zMgfq5|K}r+mRcl|D=Ltk_dmA*2ee5A04@ZvT@u0I6aLd7f{9TeK``8xN%)`JfWfa* zh(taTBX&p!K>f#qGz$KEQLOQ{T-<+d699FIULfHe53L94@_#(ghM{&P9KHVAA0k$V zFrUiHXZ4Qkc5&PtQplk-+6ii(E0(tumL&m@2GUYEIDBp*TWj#jd@<-B&}ujcI;gV5 zu-|Mmg1`Mu5igKNYP5LH%Vs(T-g>}@ax|Uic5qufnoQ|iVDuM}~QpE5T#s7nQiHnLHT2v zrNi0cSvOf`LlL2!Eaz4*!wuCpSkK(&_?Etm!$V7^T{BP9EY(ye~ zA6VnN|JJaC0R+F>1;e9mwIw@_)D|O~IcUTK(E_~IYCH%11Uuh=SG|ED&Of~0mUm#T z)SZ8P`w5!^WhvJLn^6ut<~P3-Rokz-=rHDuhq8v5c*Kb znWfLyL+|(IF~ht!fetUQ54Gt748B#X@M72xnHl_wa>-yGjp;qYA)7{iz>Hy5G2Lvf zF`XmC>AyICLjg0`Xb^vu=UJn>Lq0rDx*KO2ebj6{bA3OCfgT^lD&Y^2V)4_Th&BA* zvj!D6@P|zP?^KG$%oFm^+b+#=m)IvA`MeIW4_S>WW`mYCOA&7P1QAR$4{{l{x4R59 zN~Hz!&PTi?21W2WM$sUN%*I0&-|a6-6>=kTT0K|$+M7RCirNd4ET(h$gLytPSXPq= zQAvy;bdR^mYILz<|E+g?^u7|K&q{R}w)>+7&c__RZZF@pM?n>HMq-}()|tK1+O@We zSJyLYNx%zECZ_q-3K!BEm7 z@aLQD)id7qzm0Og%2%pZ*OdD3SE#y%YuEoMw3j_HXGou@eLZh9W#B=T{sWu-kNVnN zM|1wIJ^@3h8vSm1-j<1aLBCy>aC;y!E?d8H9@2>l^7&kk@{97R|I6*dt2|#Er%+p( z%9ig&)ywFW7__=9Ps1~5V2^a#z2jW69Ux*1YARoN?N^+4Z>{ol zy7lZS!x2Y~eh>M`iqoMR@px*Lya z!Q9zw`d6aQhuX9kiTk`bPs#KvyZy|PYQAKXliR=$?@OMS?NZ)*3F`(WnhE@Rk8h=I z<16gb6?7nykyghC2Zx(YDLcOXW(TX4rm(0+N0oP+&yY+Sok8(d#Fo@2tFZNB!dMUb zG?qz(DulwC?k}^T#D9-(Lf@c-;aytpo`1e?tHh^%yOZly;7`HVm^OzSHi1reyF_Z& z+fMUO`~Bte49V!hr>&Y-@$@vAwVglOB>Z_lh%LW4r;>pS6u;Wx32EXhc+zm(>&Tx|p;Wwej&~!U z`RFz+U9MfLQzMkeUknmLbno(T%x6%$&wiB8YB`PaY#{jD)Zc+#;=12@dX@ZPaNH;_ z>?I{eC3lTGzjMZI_crf;95e<9>RNLP`nP9tr^#mLcp$oC~&hkq=ce+Nm zoq9|FWI*QpKRw55yGN0)YF?{yyw2;T?hdD5o|>&MH~JdZ^{NWwf%<_WO2(nuC6 z9WU-zbF)h#g%A}pZL)ckASUv$2;Se@MCBcswR+SAPe=oCr46TR{Iiz)br3e&B-a|S z^pQ{o4r8sYyYAl-2XzG~dHj$L$osUKY+(NleUxa%5X>5$J+h;Ul(YVy@4b~vHMv5* zb~8zwf6aGzKU!B>?LYZ1)9oC@izf5A`9J!RH{3>Rt_63(-7Z&KP1kC8?=*hPEoXDv ze^VWfIvo9><6*bj_0uKvsjyGoyZJ3THIG4~B!e-Y)GWcyi1boH4>m&F-+sdhv2$5Y z?jQLxzJlq5_`U@KT_xMcnU}0*$GHPTA^}E@n$7Kg@2|s%IBp+}#qO&C0o=#W&EPC}zP&Zi9Kr;(h=Yf&uUwlO52fB3rmsGg` zPK-Xxr6qmOLJ7R3D#7!Wg%;C|pcm(;{-lGmjtAVZVOTluPJ%!9jM>7w#a6NI=d0n3 znhtKy$8^uIPtmF0o=!h9&*9zfm+9aUgY7#|nno#v*4#t9`6c}U9&PX$Iw7W$iNI_U z*rBL{Bk{XmW6G=CS6&q!swW`VQwIsUT-qjN=XymeV?%siB*Me%4TaD8KS~YSvUnlv9 z1YFi9U?rBDK}yZ}bqPI=^C`T`{X)-hFkRZc@BM zhV&5aFVg@Ecgys4Jv>H-x*N(Rv*-9c=jMBz%Gx1ZkL89{v;=`}4iXmRKQt{cb&CAo z81`lkb$c`>wDYM8xL9$bsY$!;>Rzbk}jbTz7kNbx>#nMs$0n~*`n zK51|S*~-o4DKWE9JbdHM+f`6A;%HDycs${!uS@*%&E9Y%M(48%^z~Ksr`niW z8@)}*q;8A>ZC6Hq^nnG=7k$^mb5G|??HobJTe-oK+}U3WO1mczS3M6=LlXvo2r{TT zNzF#D=3jM8z0V~-PH@QeubBuo_^+^NRXKL+S;I19`liJQSTPV}<^LNpOGXo**X8v@ zAZUHb7TJEB?773^w%%=5MaBEOA;0BaUw~QPsn)FSycStpN=^(3^uTd}K_y2U-tP1R zTWTSFJmRO`sL<};t*H$d5Sc0X+%`ZZ<}1l_Nf41Eb=Vo;-5{^BdeFXFeCOFM*6*|$ zM$PR_@j$@zZ1%3Ch3 zJp7ZacRS^zI+Zb{sJbPP{p>wg|WV}+LWy|Ol5K!0aadWF-i~nOh!|tJj$;2;r#{REIefga} zr{i@xlrdZ&nXkV_=`SPG+bmlkOw%L=U6E}&lEX{jQ)7#!L6zCwp>{Z#m1Nrf-S4|_F>`azUMwGp z&Gp?wP=j7X7mD%*wrly{7#jy^QvI&BO-A-#iqP=Qo3INw zQ-o|F=&m+bM}c~EEC zkxpe>F=slR$mL`4Y@nI%)y>>pTa@=~)RufeAN`}2XVCbAa7z2lQyPNO5D^c$p^{$!Au-FM4wbL@VEcQc6>a=bF5P(NQLMm~OVK=XhikPsrI zGNP02qX>KlJ`D;pGEVGic_oQBKrPnMyh&(tzRV1Ghg%`!efI0 zLn#Crm zOfu4Hx3SrIW%%4fL8n4FhMD$d3?tGP_b!DS;;yx=sfFW&&4ax__#e>ap>(*v7S&Xz z-O*0e{onxyeu4i`{BPFgClNTKybg=MLimYz=41*h(+6*T_-eMGHSPy5YYFoI#&3~% zjOz#GFQ?9fm@@y9+xr#d(GvWu$PT6Y7i#@K7d}ECG+V#cz8c#7NB21Z$`}*;TN7(V zhnD7lGdupzPX!)J1^{gDD^L%^{>7dDpOB>ge~Yx8=nH(5tR-aS{%=+zS}cfjH!)^BH;szW-rdq)98_hu{$rqfa?9iQy&4{+V62ufq#+^~<3xrIFV=XvLl~l9r!(Q<&Lo;mPi{4FD*k(~s>rex&?PfD!rqbf% ztWkYa&NaIwGU;SrQ?o(^y`D+Wawkdmp?>LQT49;bgb}h)Q2u8nl7g3lw8%##$^W1A z0}g=W1!<8?Y8>u=nlycMphgNZGyh|#L`(ss>fecMV*k_a9KHnrs#TPmDE%|DBnNf+hi8!-q239uvpgxmhBTf zEqHluo8#VMnG!`oKp~r7-6dN%Dv5rbA^5m{xpHv|ok|#j*w9`kkK^vapY*w3GAcbj zFA<~IrmS+=To>mnjbq6)waUd&TM`!|cZZYwUiTIis^#?RRmT(_swhQ1KL(xXgM53a zxQLVb#J_HzE}UVMP7Jb?gRH#s(F6dxIGfFPrcNmt1YwKqzR9%GaWWp14)3oE?d&B5 zn|C-~Y|e{0#I|B*?GBeQW4&<`SrqPbix=Vq>?Vz95wGLTKl(;@s&%-Jrk)`wW^+7m z4<>TA!c8p{t95%kCh9)g@33st!}6Gq?4Kl$ z?5S2V_bvqN@;JWz=DyITl*u};{L*bait=24)}JTI@Bi^up;oCKA+fnrqt}_R-X9E0 z!5kURc^D3+x7Fjj)qUrFwpjhUrlfV#3sL6bWKIe&uvhjhm&?c5kR0dn<}qtMQv{u- zNnDFZBHriWEBQ}F4eSUv*A<(SKC zi_+AYXgqhiuJvftZUjj({48M{9F+_FMzEq{R8+!TtK>w_g+l(V=>ZYT=sTuvU*vM*BjsOaGJEP-blo zu{#FUrnaw}Poq;L-3GqTeL{_;@~3m%bJ4|!;`IxSyKP<%x2*;(5+W9$*b0waCjAA- zGop?q;k5ca{n+T|OKcd4Jt#4uqS<7TxFr@g1cpK2y#G6Jv`#ktODN?)1_u1ha8Ac& zufIRY74^D54%F&nEuDW%N@LV&Q2P0G=6h+Qd8!cQb_f`=ElrBWT>*w6KdKC}(>x9! zCME>R7(dW@sYWsH^i`~A0C9S#sbO^gR=?NL_M^&@^!shQ6Kke;3_TDOg4U-~aoOMQ z!I{c^w@2Ir1uIA62)$ZtLX(OO5^~-eUFQPyYuqocZPre_yxx^Ym~u4r((7gSu+(vUCdV%qWR}%&%xnI>2Q1Xt_#h9Eot8+{iUfVWmbb zb%@2LdS-V3-1RI9+!Cbu8`y>^c`DPICz^-w+qJmEpFF|Num_$~AS$d<7VAHz1r&ep zRsUpKVYqP#JkC)(WNL!qs@9=E$7Z}$1+iqCnY6x@M)Lyujt288;up;U3HyZ_)LwxX zL`nH6?rH<}$Mityv*bks%c$g$_z-@T#dF_q`G5kY!fy)=iU$&Q>5MrlH^WOqNY3{D zA2ZEu{WR6R_bqXvc|M^G-EVw{J8o}XCkc)<>1b1*AWbiNl3B_?pwk3US=`{z4&lBO4B)AGNSn-U zEugM9KXOYq&%9;ZVFdYzn&XC}Et9ZM;gypaMv!0hx+36Pune4{TfD|ye3sjwDOVR5 z(rh$aD;yGy59zk-3;-7+u>|gy>yf7I0|NtfsR(|8BcaV&t@c~Q#Kd`D^g4WAZjIt8 z{y_0D{)sT1LQ_{TVh%+n;4<7w0?k&#CkmVIhVj^$--#jc7>ASD+#uj$LsYg+V4w_PLfIXS zM;&dkU0bX*h!@mPV8B$D1;GolL=Z0ozySnf866CT9Q1QWbXWBzqw3Y#q1%~kru6z< z9zjs>awe`Op(;Muu?w9d0Pl9Nsboc=m*>6sN^n9N#XO1fM4ewKe~M+(2VmM4ss=_p z`}5oG!qSM+81Uy+}RM3rL9aI^d@^~DL{|+vlnJLW2+61%HBc?*Izhy$ZM4({R zd)}F0k~Fe>mw%+*Y2C(Q)|Wv4u?wS(_>n{@QhJ%Th?Oo8Nxq#x`(icb_ zei6*rXsVbm6stPck+mMp*cAsyZihysjB|)O03C+O}K;cP%3TU#ae4UGYCkMY-ypcya+RIDLEYM`>BP_?wYo$)g1pgL+k8 zbFeodLTF~A*Eb+x_n%`r9u~%aLBwv>DRRb3-9s%JoS$2PMR-Ocb}rxBMwV*yq7#V` z4s3r#@lb3|p+^n|vNG1{b%w=vv#^O`fcOEf#C53e)#UmCSP=%&v%(C)Z**b%$dVzZGdq*oyZ z?Nln`d!?mg@;gI_^E`3)wAnwO&zFnQq+?eROFDrxfb3?6f~Li0&2@Ww#zt%K_GmIs zAoGZ<$QfeJ{5`Trt^B%b>}a?6Pd3-pI)NWr9&dv=L?5LhB^R<>>mVc9?}syP_3Di- z``$|6coMI8pbu{nb^BG}S_~K56Nu%w+iiI{7(9c_EV~)#x@-$(`K(9u(R}$HaPaAJ zs}uNGtuv-lJe6HrFpOJol&Juvn8t?XBIo1q{uIXf%*XFvNWMHJd-i@++N2A`Wsfg#afTxNG5m_x$I$ZnLn|r;kKJ@CkG*!Z%9y*ZyVL!0wTH>B%cCr}Zeqn4wm}1HF?^%j z+57o2SDMPaMqpNZBIG+WC~X<0nB{^uoh#UFd{XpdOa+tfQ7&iMabbcv&_oKANZ3vF zJf4I%L=qK1?ML`fS&dmf8trai5W*AaV(?5nVTLSj{W}rl=6i2sNray5uH!9TLwXJ} zH?(^0btKcoiAzr=GPHIlUTt*y&U9tg`J6NZ)#(Es5fqPypHq8jlG@g3;z^PRosXuF zE|wrKOY)oJu^3uByxu}@qgUF1kLUP81sXfxKLuAk9z#e?KeGkBxnN|{z(Bckp?Tyv z571Jo?Rl*jMgnJ5=bF5aE~S!9^Lbn`#|D~BW*Nzut(tUdA`LK_;d+{gP@6%*Al2mz zlaZ%4;<|I$EdR9&pjpUdH_w8P{5uE2=Pf_rr19qfBC{b6?MOpW&~O1rE;H252=s8b8d@UHzNeQ@o%!TTfR@Z&lUvE9`a-< z;`4yscN{C|F9>5{ALRKK`&KrAVC+zdjZ*}?=QDoi=?p@I`>1JFjhR@1JY92D0!<9@1RuJ+etJbt%P z9adUs4wfOqiHZ2`$-a@F^(pulQ2o9!ujj$E%M@5Z5lndBmxJ^)DM*>&s|da5?K;J} zEocrHKIuOAz-}G$ovrFS0)dZgzCy){6m@XKfwMHuWX8wj(TDM@6~0ffC9!VQiv0D@ zQv9n|Zy14_R>P5mFxRVOtc1Vg!#|C9k7>>p{XvmEvv_b%?~7C*%wr@56)x|yPIHLK z`>g?0yU>=u-K)JGKVDH?&g>R>fv+W}rYoAz7$!AKa^(_v)RahuNz7z8brd?M5MNZ$_!X-`I zN%WF6<)Ow=);hz2RhHb)vS%7sSpIv}pgV2VSK7D{rq!A)=Wof67uR}M)^0@hFT>nv z&j)BSI?9fDBnfm6KD@F`ib86}Z!yDWcHHNJwu(dt7rYiRmWS0&>Gr(en>!v@@fxLQ zru2+pmdqCY=P4%v%6MEx)Dwh&x+Sq2{)74$qO&d^{$vx+j9d$_1<@kG~w4XdOinx&+B__KZF304=sK+w} zAXe)DhMk_>=3=#Z0ar!f9`l?DRrXdOhP7z5%w%UEGy*D~+>9R1ZVb01)JofF{NtL= zakmPNJIh}l5k(pjKYrM#HGIx%4NDIVBZ|2W zWjbQHlBS_pE=2p%52mdcSU83^ZO_)y@>2v&GQ4|<1BORV;x|DaQZ6PTzltSgL_Z^D zLx?4#5V8Gzz4{s$vC@2O*|lz|=igF9ORQ@gbWs(lJ^HNi*TMdBWaYN?G)Ro=itRM{jG5htq||Bf{o*p4`^JfYT+J zd^)ND!XMRAKW1k=Cb)inODkslKA795lpXM;K7fz)?si6Q+_lLj)B6YAN11iwFF3e* zeJ2Zlo>pKV06%%9MLta0_;!CxTi>297e53PO1hBQ|9gf7hY&3+h#CH^X(}*#I9)oZ zvndxxxL|X^GZtvKywSM@>_Av|*yB3c8O}c3(w}i7!R5d55T}~a_TnmMbOGf*EG=Pp zBu-R0Bz^)QyZ?}WdMH`~ShB(2z*Pb=xU!;jW2RuE#up>SskG&%CI|#(C^_;Hnw-NS zO=P(^ym$NHM8p4P3HG9FEqRMPso%ZltT23gWCT<@o6t3PwajFvYq{Y)tiUSKc8)%k z0I&yGH)H8YWQI*lf(55emytwsp5TdrKJqCjfNR;laEVTv54di$f_!bv< z`>aZ;UxQukjhPU;pwwt(@wX)||<&dN_KeSyA(aOq^!m6ESY1L0Hcel3R~uQSiI=8 zpx&{^#-q~f1cqqyO#_0uUJttu5J30W+w1PYFSQQ`0lSC~b1kowI;*cT>W;;WsX4n- z_{)pyA%MLyLVYhp7}F{eL$ko?lum?Qe8F;FF8NWjojuKl#e|Ci$MF$gKs~rp!g>Ui zpsc9QQMyk~_Vx{w@j=(6TBWY;na5o7cQ*((36mt2%hjgR?#yE-&!A_*w55YkU;f)P zbYr(MQYZBg3Bhl_zy-f)vTcDYf|EArN8#5Bb?;HuM^@m?5m4$MiA_>1Tk(9c(b+2Z zG2jB9n;62q+@`mKr%T7f#f3#h1!%cLfBA@fzLfno*~1Gn5CK$;I$-40_izUF$R?7 zDhNZu3nF7KY2UH=3oXSIfSS}Yp%cOpV-7!&L&RZ`u>V(rsgtaqBu`s~Hl-q=q~RQ$ zs7FX#GZHT}wVqR1)pN>+_)*>A*e(cD37nrN0IP>#-cmxPMDFIsDf*$UX(Wh)}j=5mG+ndrMMl;=}NG zZ@cQz3NX2p5;_>^-N-7%M;4BW2&5zlbZb5vm39g=q{gx#+-ijy10D1iw-i-~uU)zPgt!#E z7#^S`UEM64faArnH<4=8{`gC6ycg>6W|Sfo6iS8KNIkBVu{3kuGbhz*KuqyGUn8;P z9nWbaY7vVy@wM)mp3V<0$eLU>dZ+2xB^WR6Xij#FpnAOgUS`_X^^QOG|{R1V(fDOZS~) z1OKU)V!4x2O6|7+lB9zL73;WgX1nV`AV%zym0D^B6YK9Gog6XMtl0O@33QWA3Gc^i zr_^s`h2u?gcTyS!JW1dx-`%7|X1&hjmtu-bNV+M)xuJB)@yK$&7` zG5?q=sx)xZlchvqX@|ma9&iNHC9(1}rbKVx1KEO+dO79I-Sw74VrHk4bv5ex>GbEq)oA6IV$F~+%* zzziHZRvqJn$X1@vG7f@-q9IL5s$9uV<1@1VEp^D>)K5ZDM@q<&g7bucUk|(1HPjXC z*dR(o@gnNu()_)`RCa&U&v%}` zNlXSJnW1@?rBW5k{8w@E7otPh9M*WXqR?Lq$q|Jn<9XWG{v|G2V-+RJb-(A>v_KMMksla0qaw&chtlXD3EkWAqhK)|9Bnau z+qFWtn`5{rdDwK|@eD}`6TGRUeA{!ix*-28FPb$Bh~GYnUYf^AWacTqfy|YlLUqVx zyDE+5i-yHH9wXKJW#02kg{l}E3sdlyHo9M8rU_j;8i&0W0KR2O$%s+~98Pnpy_A>l z;gY<=%M^UMZ*UvmFaY87x0)g%K3!OZjcb(C`Tig6c##6F1$f&xgQI(~LW@O~|_S_y_I2uQbnHuXVzx?g`V8)0eZ+`6&C zaD3-TzCKfr#O}^N*sPYF{9VRId9l9Jhr6QA_Sbw;dHaK$rTLC^`qd5Cw)w#3=w!dp zb@Hw~JM5r(Yqlv7#Dbz)6dG2;Bt>a*5P(YLLK#qsdb7l6>CW`=i0&=o8`M3$Wb;=P zLILhM!9wz3FbJ$<0+tiw1C6AfF75fxxsq-eB+$oHd*FRzX>$K|$O+tO98;53O6`J- zEsDxd4ShsQx{TZf+rOhgFboS($`Z*ThWIsC;`XJKB&t`h=zW*{slgs3X`Msms{k+_ z^mQ_DTx%iDJASSXVffc7Nh0>b1zBDDY_z*pA|dq8fCeaNt?@RWerfjAeqa6w|2-H^ zd4GF566M@7zwfnjyf9YOZ-w*6DT1lVd@dIz+5@py;;L3BtLOpSw$encF-SV2mI^Gb#{ifN^)m))`LJ;>;sQi~fcirf_V|;W zr0==$S=IoDh_~>JY~M}MH^^}S=8s71iefybd(oTlW6M`>XJDCoL^dGRmlyHr2fJDa=d`AmnDK})g;hC9 zRTHWC0gi|Ni!av{y&@xT(TE>Whd=}`>^hQZx`j#R*kJ6t=XJD^3Mhy)p|UKCFwK+^ z6Ck`^Uk6c=GhD?thh&F@XG+RxK1KM4Zlax5wJhPJd@iftJ+jwEJdl@E(@>S0u&)O^ zLl)5!F1c%B8=G6bcO;m0FHX91$!NN_c=%VAI)kjn48 zo&rSx*bknFJ}0~rU{)wP1ePsjG$ZaPs{?3#$C3T$$6%VA2+ja43oJ55aHf$CbOhq~ z(Rh-084lxkx6l>m3>^5AML_VVIERVz9_I4#4??0IZzS40G>eJGfCAJbHct%l^d>3+ zmh3fiGR#M=#P-uV`VRJZrze)JXVm~%9S>&tH!V+ z!Jw9_(@o`x7Q`Kf$ql~CIg6d}YphhS&T5irb|ki?)QY+f%if4)?y;Z_#PRcL#riA= z%%0)vZJPWJ-p=%3aiter>F7NO!+PEaU#MPMY%TPFV$q9J8M}dg6SqWMYCy71KADGM z|3L}&#lc3qxS!p8iqX!c{nXNN+Z*$yUKQLc9-ZQ=Aro-R|8t|EH2J_7$*#a+EbM9! zl&;dg-0+%9zY&*SZ66G*&s6i#G$FQPj?!2v@rT-&Vixh?54t28Gx9NQHvpx+=aXUs zVwj9_NsPw~@o;vAlTh5LcpmGA#Qhs@eLk=%~u5(J(Yu zqxS2dc3A~_2>hbK3EF)cH6N7RGGS)qK_L>Nhj~S;Z_reFf)c!``p|}EBn5lwK)yuv zUeF{`DPayGXt3XVy*%mjIt2mCaQBZFZATtTFcTX)x1C*6o;;?PUm^18)Wtl0#5{Iy zSBK zQeV=y>pb^XMTLFik)tWYMu}wu+M5_!2v0dJs-9E)3duoTQ?0lkKTA`sLr#BQucHG) z10o1{+K)7K6v_zd9Byhn)|;55S4>3gFQcF75$_rXid|;@PC{t;h0}eNKS?k=`qAky zh2$Yt=5^-V`n8s;*{V}5rI{f;vQ!*`WwXOG*=AC8|B`YgrF_FA6;!w~nZ1^pObff7~!t7n^<_O}(YQ8J>K(~Fow!pIA`9+U_v(ITPEq4)u zyx`{9LBvV7dA%6g2l;*RIoVjo%yfx;+fM#Oowq zV9H~5>S4;~>hLy03v9de_lUaE`hozGdGvve!LJ_?268jL6>XK-f*zzkLj^(?S=`49 z)qCRy(VT6=iC=;{X@p&+4(f6qJa9*M8SbdBhwSh?2ay1g_Vj8-ssO}{%& zePg?ab7&5Ea4bWlD2o+JuFR+@J&!BR>u_&OX{@+4dDu^)lC7wURD6**$^q5$)d;>2`0-67>j8hm3#5jhPBLN`lkl7>1!IzP&bnl~ zpk`0B@xSXMo*R@lMzSH( zV{(+Yts?7VeLYNiaf1nHt>_g9A2;Q$3Y`wZe*NQXT8FVVG~dg!5_I zu!lu|RO;o&mhpB+^G0li$`W{5W2;@exv;_#=c;n`->M$j1^W?O;a^*m*2q_evem7kz)K0X*PGVHK?0`_51>q`ix+yu zrQGA4qxt#0n!q1S(I@y*_hq~|&!5b;gQdRTUPOz}^X>Q~xkkn;GzxR>pWbj=nJrK; zaiO3l3Xc7=4GxuPxmDO_8eH|X=Au%Aaij5$BvOKY-qY`6<0=YrGYs#^ND9`8}EX#Wwp zOC0#qtDFFsn+u*c`&!lYIjVd~d99d1HuGQ~q+mV%b&my7@e;;UY^e?xyTtoN+tzWH zQ-$Qd71i>~2O2ZluFt<3q`UD9E(2wS4%~JxRDPf>Lv!1?8j^&gI|)6)#@%s?2CnEP zcigsK?x&z#PYTIX4^tCYTfW7^ltOS>a3(nVlM$qr(d*?u{qGpp`9vuM1lCi8U8a7n z>QIV3n>>Ee-v(6)FB`T{e>gcdj%`!)rSw zDHAW&XEz*)sreT4DD}-Da)48|MxQG(fPsY?Px#C;ll`>{JxbgF zE8|1c7B77;i_=<8v6BE530tUcrsT^lL*5+~sRa2#&8 z+ew~BYPD`MIMTuK{`xSu%Fj07m?Un`MRwxREqOUeVle0-035gX5?I_BI}UX*ye$A_wjF!}!=i z!_d32{g8UFI;Oa2zrz8del(kJDUp-%je&7!#hb|_YhOXs8*e;~tBV&arhnTw1#RX) zUjHshhP{RyN6b%^#`0WpCW370XtRa>MQ93M4x|6NuA`<4po6S z`RsPw&u!WaVe_Y{{50k>05|0LY?^JS9=egTe#iW7M9kJ__uWqD-I15;1#P0eb}Rm; z0~~KTXwREvZ)Kng>v|~v{f}^W!EAbYKG#a| z9~2StXwfRgE%-Tno^p&vECm|eAo0$-8D{us>O4LeB_Q5cV@|V4fT9OE!q!y7??TN!adZ&OA*@UaLrK17Y5)1kK?l4za^Q7XMX0DZk zWZ=zD7d+mVRk*(%=Peu&Jo}m(Z5tm_m2WW@5d|*a8v*E#6LhU~9 zK3$tG%I69-6GPGKr&He;t-CujIXupN^9kj2YOAjq)OudL0^WyBCs%{M`4Z7XL-Rrl z^D$!H*mH9(XQKhw)fnRDa}(JFh$cITUR9O>!kP2%vI1;7Kt>QoA6MjPQjEx-4)2!ha0$W!Xu-}sK$NA7`#ImJtJX9gcHx=Vf71GQ9nOQ z1|hIZgJ|0X0qpQ3b}_yOj6FFEdVA-OnuLRwbVB%$#KqtFHz%%}vkp=Or=Lkt^4Io~ zztT2Zu+X#YctVmynz+Ul2pgt9L?_cI+KmM#i|Z+hA5I;oLP!1`{po~e`C{8G`m+TS z1DA|X1Yy)F<`9Tlg#i1!{9`&?=%6dHvgh2fBJoQ&x#wIFNbUsh>u!3s_})kinp4Se zaZGTGoQt_=Ml^CKTwQ`|o!%rW8ei%XGM%=}VWsUf*@@a`yRU$R$j=^eL06K5e zeN6;?1b!>G4XtCF>2U(&%SwO)a|w*|DtAiV00i#Le*KU*?KXbJU# z8YsT1uTRUWO`e)&NIYFMFf%kf6I_$hq@)m<^H|nXAFQ@J`<93Y7~W9fC_tG%SL!t% zI_ibpE-&APE0FIQk?K}O5(hm0q%%1_o`=@e&Uhdf#AP0t0pcpsc{!l&}o zd#nG%i=0Z9M3EKT`*2w1Km!#7J_+@>h;83z-l5f^x9H-fpH}$81BUWhZh^w-)xm6t zVcr89iZ>Z!IwuP-njKGrPnFOqRB-+P&jN0q$t3a7aFXV}hCDFR}WvU!5OC@A_-iCCn6^Ojg1y}b>OQ0&=^Fxa$|z&vn)pPdR_ zW%gu1SqZS8gxBZx1d!I~mBE%OyOl0X14a5G+7;#-Y5u1zc;kiH?+r@$o*=S( zDEep;N~Fxs37g&XY{?5t_DQ|+MPEd#!}UqWHdoY=8i(HfIWO0~TNx%R_iF~H)Fnh{ zG&b|YTTweZ3$x~X?$OcXM&=3r;!kl{dd&u|L>zPy@O}Wf^isVk0ZHQhNMP1yOBhrz58Wb;GKnV>6In<&)rVNfv5)Ty}IB{~D!RxFXc?G2*q5Y%cYhgv1Kd;%GiE*hN)avVtSq zx3IoLDRy}z&G7PLo5;anlEka*FpdzTlBw*XCYu*d7;ixO=rGMu*01qX>#c4T$x>=o zsXu55d^ft`j9_ES8!VguDRjpA@ZHnc9mQL<4^ELbn{5rcvL{Lv>H_S5>3cyyoKzvX zj6?VJd3~fJCXVhAfid0JmyZpdk*=4s2^@#NvBBMVz=u(B{GOd*l0(L>NAp5OrmVfTkb@$ zPrT9l_KMRBDo^QND2JKUc&{7Mi=e7s0jweVGffRND}kHMQKHp70=M<iAT5Xx;e6Vw|NGGVwPHzO-fDNb%&H*0XZC8=Ug#3|cmf-Ua?2VAg7WU zR$cD}IaZ0Z#d@1d&g8aM;P9`c=W=UTdk2ds(tyUCEycLDN5RUR^jh|v_jw{^|JCk2 zI{o2rrRv8Zh$dk35sMoXB<;E&JI8VF6x#~b@|A_;_T`!D8yDB^(nY(-ly4pj!0RtT99E+{*1ke(L{$t8?_~VXV%jq^PfWaR=4;00mFk0NO`yHIFFPm zsW$WJ<8f?q3%ow_T8leK%R8Kc(q3AjjUu!g-WZbT@h2o(Rxn-tUI1iDu9r{By4ghW zdBT=ybwGeJ!ZM0gxF_PsUda4(#(G!t{Ds>?gcbRy--FXVS9##{Pg9@6lgZw8#E{j| zdbZ}rPbxJ!Quu_SXjaVlPgV;(N~1(GX_cS$T8O-Wk{P}-9O~|a$rMtgV%dYI;WS$A zuMsAdTI~pX%VT3eVb6TI8WNI<4e`uce;EJE;Y?0{2IF{IQ|p>6lX9$tqCGN*DBzNc^8VUhwL z<{_mn3qaAzr6t((Ih>rXEkb`UAhBA~oU0`2t6pPrW9wq_kA8M%59Op#=|=4RtU8Fe zqjR%DSv|O@$rd$`#U>kqrO#w>=iP5i(*M;$Wdr2LPhurKGAcf21_tJWIQgv*a@1W6 zzK#L7!f3KvkJhh769BkD2&^Kex8$Z-rR|cP?6jyTvADpIXrz z{T}|5Dd689TWNm$ObRhc-Lqp$2oiCd&YvnpT#S~XKvhJ!W^2Rejm_5x1B$Fx?*<7sR1|!r{dSH+0HB<^HQY4>3Mr_hsQHbVZ|L+ZejnX z!XmyfgcCh%&x?MwmSCqgyg9GEDMI1Y(0z!qsh@hF1e$u{GRcwRc$6s;n}*pgFpU;Q zAVKzqDNh1H0m~Dp3BWNf`KK5UoGHj6ZPsCw*)e>!eOBe-fnx9Q@ggw0Qjsv_wtMds znKPd=wsm)*xlhhvnQuVBPFJ_^;Bwu?omoUdv|c|Unn46R)jQwwAlF4R|Dmd@aq&{} zkHpaKmm@#ww#)G2+RItw8|w{C{$CQ)NS0W%s`}MDKki@Bo+=b&B{8TCILU*66wV$} za2x#yO&oO#=`^03ffZW3#z#khJi@V;=1mrwZ#4cy2T&?3mRrp87x4ZtILQ1B9r5Q% ztL@J(%beH*xSbC<_UEAD^QCatGB1UX>kSBlE@n-?7ucmwW3Ja`+MWM6J zO#!GOKV^qw3(HmW%eH2x1J1?zvv&#;XkLAfYdPvq&Ra4YFFYHMdB8ns zl`ETM#C=|$Zj67@I`U>N|1?Wj)awt85QW+dRDjdJn%!V7lwq^Jyl2<0OX9Mn;YmkV zxi1Xsy4j;nz3b3jE4kpRv$~ux=Weu)Ap<;(+|F9dM~d-*47rNOD1sisZ=i+}?B;6C zw;eH6eebGqEktdqPns)69(MCv7l2t;5n+du{LqK9=>QXzD*y3Y$w}4Vy7K15U?lAu zleeJ^EvIUW$+_5ta-~)=hEf#bXsqT$g<8juUSO}b7^3M~9@2d+m)&CR6l@Qg*{fM& ztkIY0V$B#DUC)1Ts2Jdk1fcxaJMDZd&~Mvy~nB2wJZe0h<~O!o;7| zIHTE$IsG-^c0v@Sra109JV2C|KuERsYduc-=;y<)i9uL2rBpLQB;`#_YRi--M|0K7@my<3-Y`WmTIyuI z7;aALmEXtm8ka1sp9t=c=OjCq9$L*+^RP6+a*^Dl=-o;L)`2)p+H)`HE4)KcLq%b=QN-hxW zTSFB=OzI6p6h@LOOjKrSInS;@D|ILzaTACjK%Se1+`1Zbe;hP^ELU%A$bu>QkW-F? z%LFk3<~i~%bAxkNsmVgiwnkGhdQ!D+4ge^0%{%yPt}Stm#@?Na9!9) zf91dny_**JCRVB?JVL{1z!yb)%930elM=A%Y&S15TV26tYJ(A>mCSfu4;#-Bt=eJD zeU@2miRI?goR-6~^CiioORt2FcN^2)MZ{p`MBs2puyoGxwTGWs&~iiML$y8#?%V28 z`GLtA=}jkqkd*(U>r%=n%$&loDJRGV!oUJbq0teW4{FSLICKdX#y5eIrX4-Bz>muU zr1Te;y?=>}O%lPlco)8p6>s1c{A;46SCiddifJyX#WOc`&Pbo)jKJy?B{V`Z@i{)b z)iNUKA+zCg4(rL2WbYr_l9fz?uu-pw5tn(WzrI zY%$M(W|N;O+5}K&Lo%;}diab$-@oqkg``u)xN(svSf|-#SXZcnF5e6l?z`XC%c>&eayBT&B`g)gZ9&K`^P^dEBZ(qa%`P`z zY+No;;mi|+teFu1(8YotsWxue*Hjt>ksKm4dAX}(oQjlH6do31dbf}KL5?+*zIilA zK}Ntu<^~~!`bf4X8zA-{r&?wri4T$}3mUw8mw%yoRk zu-X#eFl-5yn^uP_Z7lhLBU43VqtMa74{LU+u@0t389hZ8I< zHzzj~McQt!<2OWiEsvYCF;&eStQT^L6oO6w@6_~T_c_K_AXk~l;dHs<+iDblh6t|* zcNrS(yc;GWhD@yr_2%VPFHYo6n9i65CL(-Pad7pQ2t3m1t?su3Wv%XYFxmP`e{gaI zLO;|HP1l=r%rsHTCj@(_bb-=FqFj~jXIYuO^8$Ar5%o@A*)sEsGdTc9lUCFJ|)y+o9cjfsRLGOBaW`$8?v=VOgJWvslL4`MB%Fe8|Qm zNEW3_l}el_w!NSD%>;9Z_SSwE^!u4CT(W2w)sGyN?9F!o zY-B!BgCuf5HkLn1bN>sr0(;+wI+QE1Zp`>c|6}&;A|TM~F^oWgKss4uWivpr00gek zygY*C`_W55^kL!0@Q~_aMp=DL7~}BPP*rs0z%-a#VN+G=iXSb54r|DWk1^7Rf!MH3 z`PPZ_;m-E^0KRn+qu3N}Wqekc?kj}7wuy4gEL7SN9!pP*>?O z?XCG}7pX4+OA`kFxpYuRI_2r6!@hl0MAPqe&*P?psIGEx@&jPi%E9j;UZYfwwp6 zDH081HgI)CV7~H0k>K#__sY0gfipEgFLZ}TfifFt%st!Q>6-X_!;YbLhw#>oMaa1f z{GtuA^8IQby@an@{pI!(=qCuKJQ>Xv15_0~55I@w#sDR^tLZ%;x+s&z)Hx{gaVwKU z&e!;Y=0Gh+MKGl54x8UOofElX8=?Vq;exQcp8ib9aXR1eZUreg}J2Tm79EOPwm%I94Ybym+(_t7k$ex86b zv6%s-?p(xea-(POZ!3!Z#7!HsrqkH8`|>^{RQ0AxL@T;mezzwJl|sjOj6?NY1bKC& z{Tf6(f=Y2@K(XN7`dYK@8MFt z-zi9-#R-qLn?DhY1;q_6$zr{w7LoeOMwD29~N8j=X3f$Q`W+wHPxvdsQTK5Ii`~>x3FqM?S^4b z3Wq`X8{L_Oyq8+He6Yhgq*tnlCPs1h!kzj(8=~(#hOki!{_L=Tlxc2(O`nF{*!zM( zCoff}Kj!h4XZ5Y@3B>mPa0^`9^EqZL(6d=B0zh+ZZhR_VksPyDa&cTA`5$CujXt=2h)LVUq2S!HL&B=ZL6;RCY% zZ|2G_Qfj;l6>20iWB4rX?3y${popQ1MJxUMbzxfcayW*&5()!a=c{2h6}a=r7x0=F zUFW=9YDv84=~as`{ctpZkbn6El2;NHXrVz5C5@%#IM#qY?u- zY?bscm27aF70$^dO@bcXk*$#~3+0lLUZpp*ARcx4<#}RCJb>6pKZ0GV^RrSkW4m&1 zkB=X}N!L4i7K*=1u#d^Gtm7qV!T0 z=vOb1i(+~StyxclMW%u}`-+!-tRr0|Sw<*ux?sKG4nzR&tINB}xPbx3gn~a6b5_o*GkG#uQ(4sen(= z?zE$Huge9VY{YMqMdws=l7b){vS55C($Bu_IcI!T_Ul(>QCr|KQLYQ{i278_P~kIV z2fS|gEU#NwULSUe?XDPCm?2cmSTBQEwt`JMfo z?)r}jWCmRpL?Yd9wUc}^7AgeMQLI87ZPlWfk^+`vpm(q-itX@8A=BT$JYi4QxZvNd z(l>P8A10sqvwGF=8VL0&8DMMlvfa*0GUUacur<^O-_|RxVsU`-v%p99l`VA1&{!;g z$uO@yxQeCX#m!MPzdrq^emt6I^m&0A&0SS%|nQWSLqxm5UXxZZhT=ky~{Vk7e%S9p1~?9a1A-PK|APde81QuB

lPI$i@8bs{&A=$;a`Mr)1IKgk6+X zDd>V2RV1R{xLS>X;!st-`#V=2D2tBAXMa7tHCPXeKpcOGaUV824T>Ue1)n-P1v(u) ziLV;iL*C6Tlt}4~!ZrhC>ZA36WqNO{M!!>(^YfDg34|(RD!-DKm04}IA)3^e^okJz z>#fB_=z704Q%H%r&2Z1*(;EjdyHB{6G3>m1z#g#j*s^){T5F1*Z`SfIjL8wTpBbNZ*eJ^3L2dV=MT zEI&INGyCtWT<4OE8Gqq8xs2}VbmE})Jqi2@-)P1k@3p%aV2!j^wRg1^U*^2_E3s!{ zz=9!^6VG-Bcha({Mz~!Mz2fY!VuU7X!rn|TLTKI@jIuT&Ce~*rUS0pe>Ku^Qx0z}HK(_)Szi%(GammokVi2kTvkbm<zI-e?PM(d#tHE2fXsr{|039>xHaVBW#vVGvsLm)LrLOv`h2Qg_E;7uFLt*uB9lHP zrqrn@Q`9@BU0HQixsgF*FC%A{LEbl0V|qsZBS;-M{5ASfi1f0l**K9i47mj}nr6A1 z*K9bVTnp&G+&EqG_2K@L&8wuw&ZlxUJIiA2bLclX|W z5Nb9i$YC~T6>Pa7wS*cY@a9MMd;`-M%j~pIgjd`9k|*&~w`Yv}V&1Y0wEs%5w?Myu zW-6pwC17*V^CN(`29sN_6YrJ^$NQ+Vr{TbKL<^o@hFhz2FHjBK&+aHU zQwlywW47N*Qc9^4da}e^*p^JIU88fB-k*EIW~IBV!@679ae}Ckgs`6PaRTK3>{z6K z8KN$3H z!Gp|w;eu@Tq}HlE3g1yr8G2EVkr{-n`{^^%+AY6@;l>AJNmlALg&X%|L0`ov+Ys4Q zp+W9n`LKPNn_?n#8@nm*f##A@f@DDvAkWFqX7eCBIKyD_w$OpqZzB}QFEt4s2q7Ly z;w8>#WPFEwa~|szOIMKvV}-GBna`k*M{J|m;mV*`>$LYFkK962(OJ8m zK^)c z=p~QoeLHz)B!n7%hYG*$rzSk?nSe^Oq5QKv6Ek#?3%po7G31bO(LFk37rBM+)8UW@ z!8(J~AzWn-n|6oX7%kseKG~->#ZMut*YRxV?pcr_VvyU_>8`d`Cvnp`T7To+KGOg; zKnwtwJ3C^-)ydg$xw+oL%q?8h|7-0eIr|p zDB|ZQ<@rJMR=&N+@Xq$%NXGP+qSYD-!MrqEob`wO%%q9`WSN3tW^JqUk%9 zZ#d~Qd`Iqs+{Hw#zaFVFz#{Lo-El3DQjIN?I&0gk;sHR2f6Bc&%=q1P-m7zi;>cw{ zyVq7(#?*z7Zv1R#V|t;H&XHwem@vI{1l6rkwVG^@PD{BVxPr*rwS zRu>zcWZJ}b3S}}uSJa=n&KaErATCvWgw|SYnGI7|3KowUjoK;}JDnb`^u+}K+wEll-tmCMVYIUK?)ktoW?A2Z-iMdZcIS_4q}FT#-cFJo?wxi=A|GIep}r z^wtZ;)ilD*dd(y|dS6}|c~|4yS(~`>G!-dhls()f>dZvM1R!Y8ECqB;3GBdtvUQqg zxhx5F5q|-{nRRIp8Ncg&@EIeW*r za*l`hNKcusK7r7OTC{L4^+69K8lBUMy z6nFqIPs~+~SQL(!2sj>+ z+(0*4{*6Q)0O|Zcgggjg4aPI@PhKJkRvQN!&jzJ64{ftWqd?Q0>C}0w_iFm9EanSx z(cNYO(aQfJuKT^>J(bP!Hae1$!aG0l)k@REXt+i{ z!zJ__N7S44(-%djtbqQOMNkKt#5FwWC=UyeV{yQaIdnjiF?jNJvR{&4gGKVSfSdq6M#Zhx_`yPl%k&TW^mL zFTHw0*$n~)tr|u>R*dABi&|ay1ca{%=VEkRySaVE49b4F^VRmWBu6;R-+~Xm_<$J=C+?8YQG|T~^Fixs&|mW+`{=<4dqrBeWOJW8^zwdhnz1FBd(mkpd52#y$lrZA0Op2-# zIf#(T1A_m{IR4it6#x_bgAK-DgChMKxK$PfxCt&a98d!JZ@TaQ45yp5L!$K;VCU{{L`ae}0bw z2iznq7-Re2W9DBj1Z4lf#{(jK{`xqKKdf9|(w+aNK>y2yPZo@94k8>r_+K@eDF569 zDf<5psQ(&JUGKF{SetSN->~0oor0o?yVo?aom9x0T3cx+YB*@lfiD{O$89Va331{k zDC8I8q9w_hkI7`hXhTOq#zh^P%FkyK(;h8Wn*Tnks-S`Kbe}cE_SX*nXB5EUQ~#~+ zym^7XtA`eC_^UJU`D4Peiz*?1GrUv4@XC29YY_jv&Hou)Fn`R&r)S{n-v+r3GBC)` z1Y<1!Wo-Yq?vubjLlc8d>2Dqi)D7@Z=E$KIfAvscfBZ^TRW;h*yuqR$2)UY@tO5St zT8{F^hQR|O68_dPU4($;E&GwK{ENZ;m&YLX5%}YPDJ{zW>Nf8Iw>eLUu>|$EmP3RJ za3CW`2`Czn|4Y}vghBBA)`AZ!{IEZ{bSFB(FzF2fwEk3rSok`rg!9WUDh!st{Wru- zup;*M_6X;EOGK~`A3g{q5!JPy=JDAvTBTZ!cpp9q7i^kB0xB{v@B^2Cl%boj^q}0NLY}&l3j@UEveJ zUpgNyHniJa=f=m+0NVCvfOGTU5QakHyY$QR@p20|hy^Gra729W0F9i6 zO0k4N+OLo(4dOEOWh8+*fm+$o(Gh9Fb4fWE1wf-iy{)>rx>86bp(yhJIQExT0#aDuZ2IJJ#&$HB#VG^(PvTFVQs@b)b#tV9ljxE#>3Hc z3He5}^<8$75PY7`o1QCo^AYCO?jGMcxBy1Sr!V0uiD|S68xBsbKMv+S=qLR{;tHfs z(x|?b81Qm6PjoJ??L>s(XLS z$rtF&70KnABmY7ohw!utaXekDM)~f|gz>cwSn8`bT4_cR^0u|L)nVf=pNu6l-|P~6nq6#Fvpc{@R_LC^ECzx!sxiA}zC2jQvnV)cR zaOfbsHI$VLr9$+;_yD~52VjVNYBd_i`&y^Fn-df` zNym%z*TY}A%XDa;q$n__P2f%6OCTO^-b%GpGq7O~VbVA}sy!BBu<18u6C2I$ ze-~wIqV+}P($*Qcx$qxy>b$f>xYZ+x!tJL$J!H`&Pn*H(5^=c}n5Keits^VClqLOg zb249V6blPRy;~k%j&O5B#3}qS(8X0}D>r8f5!rQw{*G`szu)a)t;qXjENOhj!F(y( z%^!wsm>$`^Z4i;D%J?_M99X!93p^4bFQ;XT)Uwbb=&{vga`VJi zL6RzSAcwV`D+|)3Qi*IiBp0euvw_&7psYc%PB!`ee6_4K_>j^(lE*7NC7RQmj=J~D znqWH@!BQtaA7rRTp@?~aA`D^8XK`kT!G{}1USK4*IGlv?!B?~X*dMc_1B_uJmA&$O z$zn2t7w8c*53GsB&6nf3Qhg zb^ZI$+8oG_Hc`ken>GgaFE9bv^>4gCa*UFFVan$kP_oV&tkq^pJ-*|rzxaowP2&_; z>RRMN-KDVUl2x{1S;$%h?|X37Up59n)=~&(-YtAPsB%5xaX`K5?^+Nhpws3!k#F8z z`u;xiWbU+3u0i%())4Jz@<%;1NCVXXQ$UJ3Wgr(@?+V~=`y7=5;yxePERt>)vIPAW zDz(ER1QwR-&BNNw=1LAGGkb{y>*c2b2?w6$5X|VBw(HBnhUQ2;zxG663!#a}W}Q;iK8=xHpl$GGculjfeBQp4-RiV!fC{ zbrpHD0>tYlV9g<&qolJ)1RR$@;!Zsf)nK?q*_?DcY{nvhg?%vri%`4$Ne5(>-7el7 z=6^rASmTmCZ|tG;x8gR6eV-nS<~I%wRV#hAj^d5V3V?P(x;t9eNU{87RmBAfvC?qF zqu~gBXa0Nt*XM<>6~Y#x3b}d~IChJ5e*JT*J@Wj;93k_~fD1paaF?O3lY{d%bs~``Rk?W*+RoK9g(JSRpJU z#@V-Y#f(kV36vuZ;x0gcaKrJ+`R!o(v>#DvDjMI^x<>G8q^n0(2oRn{Zbrdp zvxs&?l?*$m2K1_Q+0AB){GgEo8Q}13%fw}g`6Lr*1Q}>7{3nU^!R@-cI)wt@03kZ8 zAki?aknG>9z%2FhbmI`PtO;>2LKnXwl^p(qlz~t#^j9&SUjvpU8cZq#apWrCR4p>N zC%Hsmcs!o#0C*U{^{%@nq!WP}koH0W%}d3m?EqYG6H%9@kG($thvg%?FZi}GZuTPh zQMa~R#+(15doqnTRsaU=^7gvp64>Elj+1S0p>_@N%O4@}T$a}99vj|X%@KY;T7wpe z#DHuaV&U=}&W4~a%sMsMknBw~SOazGcg zdbBo(my_fcV$dr}pT-wYhiB+1M7%rAzF*|cZWpS8HELF}R;VKVTs)OibasBPFw4KL z(5_ZVShPA_X;w^wj!aXjkVlkDCKZcZZZO~2z-Q?v3I2`!&TF|$T!Z;=W`S||Evkl0 z0hNLNmtHYJpDi)Ok8WQ`Kw724>%AjE@Z<3fkk$#0O$Ufq(T^qXr9`;1+if>j{DMm*@pYFG1EJ@iy$1mc028twGC>M^jae}DYrq#!(iMHF{G;*E+5Hvy!YV7a|0L$1$_Cb zhLdG6$H`3I;*oEN|2zVh$uI(IJbFQhP2&M5ehWME{GEsc*!f-HRd=Tl>ke%Y%XZrO zWk8?f^P`jMb$)av8(NdY4b`+P@e0UG9GRrgSd|gAl}3Brk1$smBK{sZlXH~q@fKAo zo>%@I))JL7r*xTqv-5I9$!P$5wU_KujaZc)o1ol~stgOmn5dQvFP0!WR-An2cq3oK6>Y z2cZ=VU`^)BlFu=|gya0G)b0#M?M!FFS`cD6(e`>xpwr|^FcXlj0`S<;v2o;*6s_U} z^VRxdXU(bgwqFjxf`EBKxk7{Uuw?l>Z{+vGMSpYxItjV(<`;<=Qqi=uwB2a;96Aw| zG9q6=A=UvrrejLA0ahU#8oSYn6jo}(k@(~2`7T&OZbt<2q6*6wU?Sg6B$vc0=l{p8 zL|^3sA`%=65$@s+3V$FdePhZ( z&X?<}Tt?3HBTK*Q?5pqSo0Gp~Jf1k5(O@Wj2$q<{G1;B0E!Si+U&w+iDhQHEMHq-^ z$lreDhzjn5QZ`#zX`%e6pRhXD+KyZ2#6;l1$>1~&9vx2Q4)Z@8k$b`*S`2#URBx~+ z!TC{$n9P>l`S1!9dh}1ZuD4o8M^qV($4Vo3LIs1+`BU0QAL#W37ni0|mB$<>qrYR& zDXBRroavcx)1uKAX{UR}4Dg^D00sN;E?s z$PF4W;W-HZQm*6{-w*+zfHqoEuknn8?+0T1#B^A~?{+WpwdMZ9beSc#$XcjI>5tkv z^+LHYp?*oPr*LhU4G~(~x6AgCJj<=sWOQ)@pIEEcz@D?kt=Hw;I$?a!*L=}D&gMs_x%~bND<=2IAAfI4*ujTTx&QQI5Y(;+e&Nu(_Zmd z&5DeG1sfdutzuF1)`hM33>91S8&@m>8HJs_-j;N_L0WQ}#TIB*S3F{>2sa$>~ z7%eR=X5?zSGK=kQ>_}VPe7ZgU1(^8|woT}tr2P>+aGQ~ASao-L?#@<}2`-x`q4*^L zFJm%OfVh6Rxd6G3_-Fq|le^gr{Rxz2WieYAx(!Fb0D^3J^EAaH0LG_Bo4r=kdPy(M z6J~h53>@Ce5$wg~eW{XoPzDT?G@3zhl@#&^DK;n;bg5=;xJmBRY+zIdd37~w$>Tdl zAQ>!F$o!i5Y*RD%@4Cdtj1R3*bXr!AaX$9Y!NUPpPdC`O>Z?r~64vs;7_NG&ch79} zSji{`yo{XwGI@d~Yos`U2{TpcT|emust#4E5f?-ohtupa_?iyNx${;;x7Eb! zv@)Fbr9W)6{IG657ehKDOAAClMOj(^d&wnxOw;0XSUOL0L}c!73>G$4h7OeUP(Tjuasg<#Y!o!|wn!Lo1A;{og+tJiL+qdxIYTjw^fu9t;JJV* zP>TLqUTIl0)}mZBa9G4QKw>85qu92{bip_Hek@p`Sv%dN`47C$jZ5gup>-0(rzirk zSOy~9Pus@KF4w#9CcP8+$XCpYB^tp|{@^gGV8CwXWU00=95ne**pcD%6&e-rHppP3 zMiRucXh~ezosu(rL*nY$U);*j61f!N^68-DohGu<*NL@I8{ypC2pa_;RC1R`3X^W-ZZy8s*y) zkbtANy^NTGNvF%$urMvFS^<;yHO_N3kJD@+RxaO&S2D;~-t%gM8DE)J6n-czno}Lg z@_R1yL(D9$mvl%oQDIJ5na@AT2N0OsZk(0wi z=iMPx%LIYNte{h9dE4{VW=1`ng2W;at+nVrT2CO2{_*)a6jV@2ex9@oYq}Xj?YR%L z8%`W%LaoLAGaV?4`5vqXmIT!-9paz0@8$Q9G!hP{m-8C57ecnJ%Q8T5(ANhBt%>$kp-~CHi2MDxP-~4o4~QYMf((Dkzr79 zs3DS8cTyxs-4YYHmXc}Xe8b>`wH60-eCq6@mhU~)@drb?FMfr$-_yN_9kU@w%2e<~ zGBhBLzkW=*Bn_l{^~DGchgUbO0@RL*{0S>ns)asa8qdhshrYm08Qa3dHO1&C>w+R8 zfcs1q)T;=A>H33oyrbPkX{!Rqfv_Kq$7cWyN*G><%QsBxdUA93=i4L2l!7T{p~(Me zgAv<6Kt-j|J=2tYw7|+PMlsd}#9ER{kBe)K5^sMw)B#672BwmddJsVvA=L4-z9yWWTecz3;W2?Bu5-?oX9k8+=;r*P^NUUw40! zHHeo)bNztqJM+BB!bAar>P{FBG2m`A>HM#2t?Y|;%}$NxCs1EaS=~tL(h2ygS`Dm> zL)`m%zKcpGh}u+euoSC0o`nXCa89i~J$DGDLWaf*sL-hi%Lmv6)DW(BO4dv57?Xf_ zGz3UuZY%`iJRLjWCDMGCi&#C#ScqN^q$h(jpS8>zG4;0td!|(5p(1X=^c#S1u~2!3 z(<3cd+Z#*%?nYro$B<)g7+G~sxoue;24K-CN702z`^fal@h!&s#Hds>b+eqvap_7v z2GKB0A1^n%FFgjK^cVsfV6$@mxnFx!6@81;tLLsi9LsKo17H(Cpib#lb!ZScYg0NC zr-M078CM2T)!-qII0a5@2g7mw&|vCS;G|gf=Di=oI-)B;#Yl#Pe*CP1;_w)fpQ@>z zz`tHe2|YDZ*C&B}afP@X71KH{(`c3_$9%ezjxz?L4Jcf&kB^U=Sn%}yj6AEMo9AIm zK|%)!+eLq(0K{1k(LXmQs{=0;#Y?HNBjSRU4M>uBy}li2VqtZV^S2t~)NDCKSQ6y0%wjEIj0iTfww!V-bTT z*Y2B5x6H&g&&ZmuUV=VCBP#6H0-C!m&3H;}!I<2U32a1<&Q4ZZ$^>7el2p$gp6%9T z2)Jh}C8_%kA`dslKRYbH1cdf}np(r8S44WB^3B{R#0)diPpjf?ENZ-%t?!e#4UIqu z%*>AN0$jriTN#CRmHZwRA|bCTB3dyU)ZMYS@FNrwF1KjDkP;G$5G`oC$V!KYTVAAM zcli{3l`giNk%|?3sr2xi;&mxyppgu6C>c|cUdbE(`wKYqnVIs4c1U&!8D(Is392zw zc7c()--I*xcC-(RcU_45;sd3*@J5LiA+C^H>;tei%>D`pS_Re~N)>%_za z4jP}W5Aj{jXRln_d=Jmqu1XP*&ZXT~h}bmZnYw2m`RqMUPfvN1XZS7U?XDjrfYqad zInPBnHD2Rkdd&j@d#2BL+<%TS{v##pE3Gru2RS`-OuhrZKT5||)8}ZibG>^6znd7~ z8JY*-;f6D9nY?B+u1&qCFXm{yT-9>@`FlxMXn_8Rf@wr7V;okn2)x^BChv6KSI}2w z9?{JHSUq5jiBi(R?F`lyAdqHTHbzFYUGS!HIKZZ1(=2fmW$M1-Sb>sH{n!+T9#jC1 z0%Jo-Gj$^>ZD+7a*{}ktckr-N8CR~4^KbIo4!Th1%-Lo6Aj9GbxGi<5e^OB<^Lylo z*H!r~eu#}@vs${!o_y{9QINyB%WSz)UwIt=#e~Af<|4A9{CE_mblwh^%c_J;vA+hA zS%#3=zZV&womYC*IrD`tv~a_ipON@OHpA+BfWEI9D4HwZ9UASIc|Z_TM4Pk_lnwS= zfTSjmWSlxj&~Uzu3ltS`u0r)AupN*EPq@s!`xQ`Wh78#gm`7SqT&UNRm5;?__-vKX z4ueXna;>muqN=B(7@{g%{+b_2I$HiU3cgCES(3e8&~Qi5lw3g*$1H(?1{FqB!fsB` zBQZ;@{F4Pue<0!jAU+jcB_xU{?6}|}WMe&7l5p4{o2jS=osRD$v58ILjg4vN$lPI7 zV64{Jv1Luz^RCFyT8u3DkEx?TL=y6bHWu1x%XPTB(Cic!%tGX_Dggrl2wcJ_LNNTi z8sqG>2Xxn&K6@xdRC%1ZMw?Wd^DcA;NV4aJ9`)$uhaSW&s;F{Ft)gUszcxCK+o|ZD zUMnudBMpaQ5697%_P9L|h4|4fS*Nm=(4n0rRQfc^$7F(%dc$u`=Q%WwuvTbsMFsT6 z7T64>cG~pxNf7b2(Vvk?8PM{m?gXuwH>=5k)L~zEE8Xde-Ks2__%F=nI zc!d~6m;utbR=1N%4L7zlcGQ7ZWmz$WgN5Qd5sf!RKolYFyG81A)qm-2ig$HDpfIQx z%afY|o6SULiM(`m{f&-AynkYy0+;B{4-*`NjMAq#1?~E%44>OBZWeEgs06FZ7a8?M zlmkY@nVW?T#?ZCou}GDc^uxBWcsMw?`Fe^-ZYJJ zxt_oE7N^(d5vh64S0l0c1$+0OAIleQ!#9Q(D#KN=zCjPfP6_Mm?{*qVGYyO)vy zRUW?`Pf0_%0d3|2OH{TVwis2+WJKa?QfaISXJSFlZGo6or4t%S0~1MKy~1y)N+r46 z-$SrIzW=`HB>6U9riy-(0zQkXFyXxq=+HX=izwjaU9%#MWF>5=++a8sr8~Q8y~9I2 z9gL@Xgc-vICprVwA{97(IczJS)uCfUlg%PvCg*lI_J8;^9gL<+uhAsi)2P7i(Iz{Y?H!Kx->xVM>gb2+?UQ~%^q_zHor3w_NLjdV^> zrlPd0q5Dr)7LkICvwWk1rfGM$-zXir3MBwr-dekJC2mz^!X?tI7fVQIb_QyE1Yj?`-n_) z#^8%uUz*7bV7+lCst+T18DVp7Yt+BC@KNUzfBPl5zYPAck3VO#kCm$K!6W zwL#%Zt4FA0WH*@N7e($kFu}3}W^?vgfd(8!w!&h=$=QZ#LOgVazv$vwz~|CpMHd8L z%?eB!&nbst@K~gI1GKb=bOXu;dn?Uj~UfvmPyAih|tOHWfejdXECssE;z3IYnblmms0Qtcz}knJ2(DHf?lrJC36 z0fat{-Gd8{Fb?nJ=HXAMu)jj=3P5j4T!ko2)DjJjrCrPg>&R0{hvoS+*y|r|y+ddn@T3TV-m7rC8DhjITyN84&REKW z(9Il#em#X25$F7f=#NJVznDtvhr zCM9(V8@!CiqVEZMkY<;y7C>@?v9Na+?0(i&XbEY=ZWz6rRIj@Mc8}T2ZPE<~SD7y4 zjd!>dyT&CNP@wQ7mhSV8TwiX$ytzDw1_8d`J#70R6r@5=-0 zrP<^^2BnC1OBAI4DW8E+gz^uAOwy5)lh;z#DJd!KIl5&Dm`eR6%L$C84XAI>tl%1< z3fZ|?ys~dw9WL~Ed)BniVS2$7+?qqu`rEHE3qBmTjZD}p!k{B*JHN`Ljlq9RJJ1eZ zuYugfdbUvc>Q}St8?T!okDrsDd=V7!GW~=;RsI=D;u6d@vt4l>sI_FJ=r*NI09ahz z;VbbWgezM*c;Th`GHNXQN@(opL7ZlXT&Yhk50k&6R%F7z>lOP{M{IpWEz#w)qg1Ku zbzJAp^N{JC6vzE@ui3_H9| z#{%0VHCZ0?ZcI}~P!^ph(J)~Oj6P31?$(}SzYrH>dia7?CK`jKt^d6)jaCFX&G98mAW*l1$}Z^W)qw#QUb!>Zr@D9_ZgKrpo1{iaMRfxz>+a zuwIoJwBk7|)&?BRw|Y3%x@U_~7eHVFe@4GDm)m^X89SuHT>oIDILS=CXb!U979|?y zqf17WvSw#O5}DbD=|$Z4a5i;o=d-=EicPg!CMG}7>xO$sC&CIoSg%x#U~+oM&r`0Y zjHYGT|HQ94mFm+eszdMSBs|6Oqlf-F5wpo0X5WY8hH8JqXHYsJMV(o6xhks3kZnQ+Vw0)0XbcGVlL$r~* zq+-@dZsV)x3d?5Q4Bq4{)=9G^qCunEN_|EZ-DkoU;}4WC{_KPZIIrv8Tid9VUu+>; zPFW7MILxsdAkWYE3Vm%8WWXDA*ehVPo@lpi^y_gCvGJJQoZxuzCuHSMVI@a|rU{am zM7x$bvW2D@VkHxM)!NMI#x#6B*N7slBj?! z7e=-msgZf$vI?6_@ezN}{yLu=C#6RwNj%Ytm@HLSRzWIRNEP^g($ZN2Z>Q?+6)6yK zwp9FC8O;=Vd@8ZyHTT1pYXB`!TQSycL>m=`v%4gSXpMl`+hhZ>jAjg|=?4%GQIT)U z66x%*szL&8L$XlxY$ihCf@DQ~SEnm4D7XjqvOhunC|oahvLeR)hD`;CJu<5Kg_lTh z9DxO>rAEBf71}?STm1Ami{IUyDU}PCx~jCEG`m|P+(W_%Xd}1UJYCu}Go{|`|Ekww zCJQ|P`37pVXXGob+J$4?*(Pm6bZE?DfUlV)ybKdd|NiL6{9A`K5$bTZV%6?`q)!&V zv^=DQF*EdIzp+43j?yffCh<2 zav4!_sZuYmKet@35&`-Jx5qGh_bTBro(=_Wkm?>U`z$#tiRLwyhSLB;j8=Iy-&7S zA$#&MbWX^Vw-z4@kpy~QwXS58$MHw=k4sUq_Zmur>H>T_n&|2dnkZ#_Z@={nuWT5` z_9CTC@eW~`Sp50MVEo5`8wJ} zyEhHMB~qH7d&wO7=A)>_UBgYT-g~=GVY;nXFcT#FT->dVBk=!d;s5iX0xw3e0&;BR zN6Hpj)DR&&2D)_jwd=Wp>cqjusbnLnv|{(Xvs%pOjWYbnv;j>6IA}`OmX}4JLach( zA-ebHb0S$QCVRf$moMi??n&vMj=+1`HUH)M4#n}nzf%#*0{oCkdIG+pIcr+C!ET{I zG1bn|WFVSyp~k#+B{Spi1fUp;h{06hmw`XO!{{^WH`E~pWf(;)GAOQp z7z-ENX}M|(?Ob@wD&7g?zebx1oo9^ED z(_nmr6BH7fAnw@i3KLXt>pO-dH;rVU(cG*cCnRl!qDRdA38h9$AU$ab9UATyiwTGc zxZ{$*DbJNq!$MMr(>Eee;<0uKWpkeu)XDTV{CadxG@l;eQ(0$&n`?E5_60Tee|%B@ zb#DU7pLEr75q^KSMvOh4wB=Fp|94-BYH{zn_9h*`5n$RkP$WEcO=H||^}c~KZ0SiL z^YtAzF4#qrC?!6e{A9K{8p{+xhz8Epo7-FPMB;r1+Tgl+y4$~5=Jy0J9c7XNK%WW> zf-sAf3=``IUKcw*x3==z34jRoX&@&5+k+yx-hfyEAb9ZmbGti4%w@v~Hn7dXxc#B# ze>U&^(@n*&gJXg>fL}tKSEWE?GHU!cL-;uDEMgPl%=elwYnSDt%;;=i{fVWfe@1W#vwcj10@WB}JGrw8aX?~}Ez zPsBX>1_l6skLm^_DY%ZtZ_u1lg}?Zo#&-^cB|m+@cL)?8a{y)#Mkg1xTWxQ1Dfkcij3^Mm37gz=eYW4+Eck)+plcEXhBf#~$|^fWLniOY5tXmxE7Lix@H zT#1-<>msUBi?Z!PI7f$vtsxt1HuQn^) zEAl*;XP;I~Vj(T~DHDbvMp~0}dbTBh5d~?8CZup!ZUXBoxSfzipV#`aClsC62w+kD zAIU}9tqXtr_yNFW=XQrZdO1Key6#TW)*)%myOpgKGf-JGX>v&H*bc_h!d6?%fmN7q zWRPN@Pw`R&D6I&Z_(Rc1gA9O%`6OQa=1I}2MvFqqkJAbSH(fSnK)&%m53~O~*n&a< zTffreY_`$=Y2pg}oWjmN# zb3@9stD9Z7Ee4{YLqI9x?&1;@_!SzTMSr}d>+<9z`qb@cnb6j0a3Wp6V;v`&Q*>X@ zv={MBz$b!J;-8?)73(Aw_Tk0CEE1gq`Vw%36}L~U_Ufh*!-?@V=f6fJa1q;et%-%f ztW@kV@X^Ade-eB@k(^V@`QvkSlz#AlNUFbtY5y;bAOEqb%fq!Jh4Fs|LwkVqAEGt? z^_?r2I-q+X{&u=qXF+PS%)0Z%Sw5AaJNS)S$SCI;Z0FFnG>Pi|jDy-b`-Z`vsrvW* zjq3y?`-`atVusWXV=R42_ znU+Fgj#oR#1AAOFfFKu$&#%FDG@8m|i#)oBcnm4^%T}!4OXAIq0lK$v0|0)qV<0XA zldF6=psl6R%m-up0BZ0zx7&{sI}K=5-5D4A@Cu|IWPN;np&y)O z`i`-;R4GGHM>iEfN?yU&kt*ZhTmAz=`=bS%5lz0z%3;KlK~_L$Z3NI6)&YkE^%&Ib zckxN6!SRT|mxDDx4!T|pr!#|?<7*;8#;j3huls z836lWHT)|28Hv3_#iA#ar>I9(0E!SNyS@tZTZ96lT9lKPY$l+RC7N4X1zh^zTLg*` z`P8T&#GF=HqYMF1YDGmkjfNe525B+PxDj-cdT)?K4Pg=x*1*RbFh5ZwBU%<ub-3)UJ<7{L@kO&jwDBsDM8D#X9S$JuiqZ85ZUa+A}Ns1+DrA zIZDU}k{PKCn^L%L=~sbYjd-R*>LQpB8XZ@qYrrs*u~NY*ZGHLrWiT_lx5d<>F9MfV zEGhq)!lP` zfB9CdV{0!?u*w(3?Ip__1jo5?=E2bwMa+YCgK%$8BNIjPq_m_fp?B^3s022g;ewj4a9L#n++5!pEolvPwq0Sy(i-{W7#y`9s@ zmKLCTuvt+!#Ah{3#@mylAdi&v-WtL^qzr~?w3)F1y@^^FQxBKT^f>=GaD4yA*7|?G z00gRt9JBBJwR2e7=X|BB!zH>~CLA@UXb5=F-xT6nCNN$C5y;{m*gO+>r+)9xB{6ED zy|Z)1(SgxOfk|M-hhf3not#CDP6wF;KtM=6G~-ncM5y6CK#&-nLkZ-6+IEUsQp_6# zGp@jyg^WtXnA}BJ_}OU<$XoXqT$28MGejYY2=_YY6VA(66e21LHRvMjP6oaCX$LDd zG%`W@9_=b)vBSmsdI@nu-Gv#c2XWXCxhFeNUM`3cExG1&Sk|6_?0R8Z*fLCjs^|g) zB}>oh4&;J`s$Q>vI#2)(%~#Bg8bR&RHh3~i^|ql2K$67(T``mfwbE?T^C4%9YTD!J zj7jTQ-22-z{EO&y)Jn~f(^n%mh8){w-Vt$H#VrLCqjmVJx`zrb+^WW?At9D zQffU@X+$21!L64ki@EiFDrw~@TpFqKCjj`tij=%Q%zGWz9fDj~GT*1d_q;$kv%(%@ z0mbX;h?&+VEjl@77Al0`9#ncoV>aOv|MuPnV$fma(Jw~eGRj>-jWo#|EdrSON3*Nt z>3Scc0O~3B6e@+~_cXPj-!;H)7bD{T06+ymV!+s(f%9%_97sFd>W31U!69Udi}`fM zrT>UXLjvkLF=neKGBPs76wa<3O9V6$sKlv4!5%S^YdXu)DuBFrvm2Kgym*Q4 z<+D;ZN|g#7%NG{^)qP(xpY`}pi?vV!Dn+~%F6vV^`mg_h8vP<1O~XsaT$>PeT4Cf= z#;$l4!Hc|CXRy6QQH7Y#<&E>yvz*(d7?zci2G%9}vfL2d9*oPi-n{`xoz-4JUIdiQh&2*Zl%0Z0@_-j1}=S)mNB1o0nO4CNL2&f^IPR`?O;K+&%t zAoY=r(@OF|&ep-f0oo?<`@jBxe|_*S1&TBl9^18+M_@hxy%0uNgbxfmDDw2FyVQ~S znQGW6wCuq47W+bR&#_SZcPU8VaMEkABb;5SS^vx(!|@Z0C5Se!mP^*mXi#Ge3yn^L z9!_ub`$N=KyTj?@+4hU9_tHObFC!y^7rFE1+4qn_3@^kJZ(OemDQ#k;4sVmgh~b;F zmh+ZSMHoR_EW}ZO86WnakdWT{=b(&Qc2z%%?|fkVKsy!pVot{NSsm%^qduI^5h0%4 zw)M(1aDNICjPnZK#cfKKL4wl4ef0*0pww3m0+*OH&fe1Q`cpfTdqV5{kISkpCQ z{y+E7|Gu|YNv>@J$pOY3h*D=te{X#Qn7yjWBOS}P_d!$>FXR&ymgYbS6HBLD!JPd!4WlT`Tgnqx z**f5e?9+Y2o#7OWh?yf9Zkk)G(+C=?C#V<{w}c!PKu=&d6#sIxr8@fGZ^nQgmVNr& zka6fSde2c*eglV6J!msUAv+5H=4tqAu?13K2xGWwhy2}@Wjk4(fRbm_#Q%_T{P%D4 zzg`D+K81S_*px~9-TS?z#Q5IR)5C&VjrlKb6e99p-g4x^LDy+aC*Q96UUymGgQLNq zfgz0jrtzEb|A9;Y=F9#6GfYLbN8$+Di1l71oJ2+~7U%WemfV(=NFyE`Zo7Fl{U)cU zzuGS~2>CwkA%Y$wdV&GG={?8fJ|bY}d2*(&?Z&r7Lm?gjWB5O; z>HU?GFgIIB!14#{uO}oVbc&Dtmn|K*WA3)bQs{~hPqq!R5!K-* zTMUm!BFe6s&RKBJ-4%I-p@*yM6)@jB%Z28Bw7*Ti|BN<+LVB z-x7UupJ3(d>uZhz3y*S5NW18Lb@Wus6^giQ%7)ePqh(uO?M@Yf8~pD2!~|SqRIUyX z@dsc$XliOI5l8PbSXuy&kxB@I{!$N`dXRT8oruqtEPCu?GiY%BygO;051GCO=lNOXDMK3UDZMBWv4A#1}i z|JnLh>f#SQt~=l9aTcWYZxeXBH9$FgLBi&@(%+^V3KSaO@4ZzL6|Cz+s^%P)mux$E zS&u_Y1OdalQz}7>525M@t9{mf5qRJ!i&{8dppV3H)PB!w4Sew*?sfin#agUGkS$Te;W_}} zs;%DMVl-y^vnuv0%X&Ow9_Q(``eD4hvG@e6a=difrCawEaD(t5UkxiKYh^ zA9mb{O`W zUJXVE93^)Ez@Q~E>!<&KV;4M)G`CC_>UJl=MWEcD0bro5Z8n>7hM-&~P-^mS+<=>! z21@-|{vI+;D;3a-NM?on>*Vs1#puT@ zvGe+EF%r-Zu>cSiT#QleCNf;)8|b|LybdYMyp4c+T)R0Zv-%Z+jNjk34{kB+ia4U? zraQXWHRa6jg>$XQ`&r2g9ohUZ97MP??;>ZDiYISZTG1cS4sIf~q=q;(Z<%yO1&Z7c zdT$P%`N;rFXb7(D+a}IK=3BdU$K2d2vn67dp+ZlJ-aXvE%M%DV z2hGA}HTGB<*wZyg0cQc<)i>Wa0Y=+U$KpR&>7->xtu&t>;^ve{y?-*8v|nDgz)Q#+ z1Oagbc#jYfyv3X602TJ9g#kT^x^oTaKD_TPnI}otlxZGFhRMZ7cy6y&{r&URe>Lws z?c@kFqOIWQ`Ov{a&m_tm9M4e+Qb*7i{D|v!SYZkHnx@Rga+8$GJ3RJ9n^&LbVzeA4 zzkWbAlz2#^jI}M_>kzsW48kK@=9@0`cC~JhYVF#CJN6Q?TDA!|a4-<~{cs37_uanT z*-!qmY*+Pqd%jtB1~(i(Ybl3h{h*|8BjX*p#Be*hGAb!BI$2HEtlF;fn4WVu!QI~r zGwZSCnzs#UmUMVu6<7-NT$tM-W(vODJvtpWC_L1@`~eXpTsJL@+4%~lSd47zV*$Jo?iBYTDr~JehT` zO_S}H)v_r+w)peN%c!>dZBY#dMl(8g{q~L`=l9;9ryV1M{mA|mwxwaBt-E>>2mKlB zQ==&>7Ww61o_@lH6>@)l`Yf{JC23uK(9{PMhGH5f_m($GlFDE&cJg(W(9aeR z0X;LYaR48FU=|eat=G;tge#K=3)VB7;|{=o2uwc`|_4~j6JGBW`5zRw?Tzj&TX-#ws1-Pme%3Z^!b2zokA z7pZRbW-pa|p$Dmt`P7p>2>g~|+^`HT#&s@|245_g-V&LYYvb*oWPE zf$!7CLn=jBZ*jHN$ZPT3T*LZT%5eFYjd zTMOW08AdYMk+z}(zA2<9u}I*}%L*-**(V9!pgMlUkdKT>)qL|>7W0HIkZH1nDr*B| z&I}*{fbbV0CW^no^oJ!h8p=;tv~s}kppi{b6t>`l7~LCdlzdo`A>rmw6ZuT7!l3m7 z2KPwPw^!;Gz8aK4Dj4oT=EYA2$!Eor`SOMG?;{l;JERYQyzH$cD9<7CK2|SIjzDCmX9KaM+#k~pMrT0_|Y zRz#K&g-4%P2eAV1soGdFOD^EwK9h??m16Ta=-cxFhc_9_z(F8W-43$Xm*)iN85Ili zKxFae+u5s+Mqk;^NL6l7!C9CuV7~bngY~9lhIIX?QmnIGv*6%u4|OfVWM6* z87HJwc=_W`crMU!x@4BQUjqOFY@4br87D#ip?WRUwM5SkAaPNXA ze`l=6o>dmApo;O}^LL?vdbW0p#N9wgyWwX@B-0@)6ImFL_qs#R6$yHmfWWYR{fc34 zeHXR|8&#Z#Aq@t`1r)QN?z9z5J9FVs>X3Mt3>Ya77a5v9eRD;?V#*Q|^n%+_?QWkl z5MWr$P77d~4-xw{cD`~uc)HQ|J&y4_JJ%9Jc7OY7VQ5ib`&wSys@77TocgxWyx^pq z1tKV-E4l&XG|0a=n0#Dr_Uxh6uCAV&4p`2Pe6W%OGemvjexQ+YE&guz_@D;_l+@_3 zALX?i5?|sbrzL;*-hBzbP^Y;d-VKF4k4>u-!3kp#Lo(l{IK((pRPIi%S;92US=7jg zH|nrhz~Lum4fIF32p=bXKDB(R+d{&oW7cZ!&p1rQ5D141s1?Y)&t)JK@3VoYG;D>G zY3jMnr_QlRFvd#|Yq63}7RG@Y1h}n-UIyLn^|;oIFx-Y7!3#f}U8Twon$PnUuhLHv zxC))&FK;;L?8eXq?vAVHSTbeaOUd28$gZ)GH3h~&C6?uZ@Y=W+%Edy9E%L`!=a#+7 z(K{R8&=*}6DdjtJBqLL^f%R8Re=5pxcQQYDvqJ9n>X>ynMC`Zv9)(q8<{Q!&R1`U^5Tqj__hO|P%4OB!1Ro7bhdhikom^>DNRS>q{rma8+Ho-B2_SvB zXk7ukyj~YpD?+8ST`g_Ccb9m~x;s=EPTryHW_%XKkpg^lbaeAOMc|IJ$t5tD3-O~P z88+H2P|es2D$>gVJ};U6G-Kw~5F@b#S}uisQz#XARYQQt@?mbla*+)C zP%_L84-fbBz?SgTx$~un`1@_WfOnw4(1>icFB1#Bvl!??KqFXPpmI{!MtO%{+vygv zOdGgH8DQjb_52`s^LY*@UJ=BcSDE3~iYqhX2d)|%y5hAd@}52kZ?ld)VK|G=UFaizc-b!}kf zaVbxG|8cm@W<_Z0ymeZ~U2Q&z!Y5ZtrI9bM(6&}?QTwDg87CfGozO`rZj%^FA7e!k z8w`-S57TKy!U7())!o(Gyxh$hVH;l}B13fKAkEy~edPZ4*}zI)o?~GXVgT6(`(W}i zoeSmjKEj8Zl6n_x*ejDT2X!Lyw3uGhE1`KOTSQtMucC?}z++TQsr`hC!54!A%9v@s zr5fboYOvhb!=Wj*%a(sq8WMcg1MyTpes!bT)H9EE%Lv6T_@y*CYFZUtqzqb$2{|{86@7`U|HuN+ za1RKwLe6MXm8PU3Wh0QZVe}`W0l+ul5Z|xjUR$teS3zN3buDn+Js*ZUaj5`vXFb^i z@O7iDdR|4axg2R9Af_(FK`N~&7gF7ge~u|pT>Jl!YKP+qj!b0I!Mgg?jltvMa`)UE zySgr!*W7EYGF37}MQ1u>Ib0-ZPu%!bVqG(cF8~j}7$xd55lRz>)-6;2heG{kZyEgo z0JX4t!i}UzH<%KfT?7`v1mhH`iaGZAw$}xuD%^|S`~~V7kVNkyd-S;^m6lLMh!L`!{L*_t0yC76svqkz#U+m$vT6_7Q($TbwtY>!wP z%)sw+!4&?fTs1$wLqa?j?!Dyx5tq&Ka-gqnq)&s{)kNPgYK)MqM?kXvjBGr8%v7F> zN-JBma)WzewwhY;a^toz4bOOs_2KW-t3ijOv=5nH4uKQ_^bZX6d-Vt72Kdy)wtLxQ zLW3@A$%{=K8k!ug~cfj*LGhxwX1Wq=>t~vMY#y zKA9=xOO2m5nQ|J9PAW?qF}EX_(kMQ~z%+qbb=o3d4&o4hW;Q%~LrlQ4%yvjXJgFIG z^5=Y}6yLXlMH(T2d}7wMFp6jY*mae7>se%?q|195^y5eGUz^S9-%b%Y)o0UyXHhog zzSW4xLeJQQ4ZDhf`?K?!60&+4+Xk6yBw!DysN&V*FD1Q`zzG zYX7sxUCvlbQEilWZpzCQuRB?NFe-zWxYBExTM_**8v$Np4Ao>S*~?;hAWxZ6KaFVELH#oLZO>Xl zXT+USFrBQyk}|gII0ot3F+48h)AQPy%?x5K1Y4hCm-N8u*RKH}?iO0b_!EWAiT}E{ zyz08>BSKB2^8IG1xHWc@N>qSOyu@P&TVQ?&%@$C~a0yKUz8b(;wh0b!> z`Xw2Udz0h0O@?Wyx#R|OZ>yj^v_7REhBo?sBEahFXMM(aw0<|V5Mi>q?~qST_B;fxKxqf0=OMwCs^t3ir2xnNBZsmoe%W$J z>hFZQUQWGjE7qej0dx{%Hjy;=p2_Ug&VuyBmp|NZ=42^M~8xyFyFua(#?w3FoeiVv3nTrj&Z6 zA))_izp}A3)F)E947aPZ8xF=l?SAH-cFM{0;l;`a+RphWAv1Vuu79Om5I#_e?v17~ zhuW0Lje4bfJ&5^gLRUM?d#<>DYHDr%{@|_9Hbl#SI~<;6O+xz0OD?VH)ZM|tLk9hb z8uF0F$r$wdOc8k6SBJ7%MP!^6}HZC+yb#rNLs^jh@84WCNm={^CcrYR0{ zp0dK8R&651#QF=QSo)((3XgT>zy<7`mDa|1Gop5yA}Kj&IJ@;<%5>9Irm}e7I+<|_ z)CV~jUZ9F?_6d>mPc}U6Oi;VSA`C*jWD;PWJr~_f*i}l-xE~XFw6g8$@1@o0e6CaH zbc|@P^=1|TO+}vC2Dpr^uR@-)ZqD!^bXEM|d`bq{sADmKMq>mpZs zMFQiN5(d**U^{!u@~iB1`?JmTs@Od-2g8Z@g*SROruav0KXWZ?(a42KZXG`oto~RN zyHwjHt;1q|e!5A~*4e2ON$ilszN(K$SnEe6C}b*{WT>H=i+2Z=DpgfjD{!!#bp z3|_M{yobFmXzY!Xl~gv9?kbZW4FS`v#LcSMX9h8mqC=;K*x@j}{X&%deL`$+R1{uH zP(7oyjdMcDn@ltINyE35>zijz$;^@22j5nv&j3BXI4{cK?n%>{p&3FVz^=Vov|pM zaxi*fb0yDFvBoh)E8f$dOXD&vU~G9`l55FO1{|&M(^<#DmGs62s_4IA8_-&Q7TAbb z*e9~pY_>7d{j9y%x`d?vuf&fiby~!7((zWv=0FTcI)&ZEXV7q@3;m@)nltKWJDV9I zMnmy=(uzh&ki$zqtw}ok^3dP9NwKwuktMuGF@8Qx@-9>V{PNS!FDsKD6zc7XAxXIs z-xNT!8Si#ytjO@8yzz0==oB5%)-fuJ}3mU8_ z`;pTM^_@MdJwcbD>@8GGQZ>gT8!u;mUOMeI0oimw#Hu2Yid%6mPuZoV}sjZ1jRpMHC4dHutmbctV85yKR`qs^@ zci+uZycE*0n@5H9B22h+F@|Sv9?~p6O{s_S1wJHt9;6Lm2g0g7YqTAY&lW4CZMe68 z9xRXyo;Dv8Z|TWc9P{Fp-8s zB&ur^x0$pH1?&4uwc4E+Aygt%_Nq~WNW~ml^9@sr*mxHUs|<}^zi(0RQ@P7D>tj!g zjh2|~@32&qmtRqe&Uw~7PS%3D+c|1?o7GQ^QH`c29%cpa_r4xKPKjVGKf-}{ru|Pf zDYqX_=_8!4t!#BeXfJ6`E)P@7_|A-X*{t7wE2!u4V+RQfO35 ztH@#TJpqYnXldu67VN8@~qm%KatYngKEx z0ktiv&?n9;C^JaP5N~SNss106MpTnTJdXOL*C@)|TmjiWl2 zo7T<0wj`;4Z(8l=fFV&>QWZ`_kvubfj?AgK%s<4;Zfe0kf!bVkZ-9 zcSBqLg9$gCTc?Bc`}bUk$+)B2C=O(WjY^tLTDrsE@z)oFU@eVsi`W9q#b(b3yQL2^ zz2zFsa_?o7XY0pSj48(RuZ-^{WLS-a$vN^^Va#G$v$R4jY~PY3VDL<_8)r9z6y&vT zlV??wW`*=pebd;qm{fjocl?(GDwWQR&Jr7_sf```&Iqs3$EO2VmTa3j*_U9D-)D9< z+7|2+PZYOc?I)3cXJFH8Ynf{Z5MXw1eeaTAy8f(b7PTQou5!q^mX`adahf@azXT;+ zvh>FYBSI}}@wBQxYT0RotV!^n2)bTUyZOFPyCAxkt!v0gQq~+&X(;qF;6!NW>!K2@ zTZ(qDR_m{_2Ga6Rf+*Ti(&Cz`XwW!U>GO5M7-i(mS~!X{G4;4+nemU&y@MA8v?JX=!x6%uG~YpP znQEj!!lbUA!C*lwteCg#LX)dRlUbF13_g#p##*Y_ zbeuW@0Z(QxSTGPvlRmO0v$oD-QUxE6_iq8u+5j{1>!&;|U+=oAUYptbdKasMfg?p? zqQacqJcVKyM!cw(>iJv;Ahso0uCAmj43pYze}3TZM^CtVQWnNL3xwYs3O2J2z?fX3 z(~%m_=~67{v8dv&PGwa(g~k~O7yi-ZW*FZhqsge!Dh30#7^U~yaqX!uO&%%Ehr13ZOe)Eb_{VzU3Adxa zMIKlbbtQ77+&6{T4#Q|Er2wjCdmFpFO346?bDph`KBLN)_hN!kOcF8??m@2=OBLH0 zF<+Qg|AEDWVOsi7wRIj~3E#f@Nzxm~=?QU0wyG(3bq~J-)Y?LOz_Xl^GH%GX`*YDLuQEXbRNDA)bTbeAxOA6?) zO{fVjmcZ?>=UOovPiHGQaYfe;0!9IJ)hBibsR~lEsW22;mq6gRRzE6X;ff$j@sL_Em7JkhF*JG~f zWkO4|KBdrxLoLd}2OwlhTz;zr9E_Ey5VnZp(naoi z0M-WLz~580SbF}`VV)+yvV~*2Vr|)VYas7JCnV>;hsgA^*-wS^$>ENLjv#4OuKw^? z3$~e!ohVavkFl!M?Q6IgnPVDfv05Jx-gRKv|# zb2Xc$okUt({wQTWHqBz8@-Mm*OElo zC?@qsTTK*)rk2u%zrB;*7$#=AH-GE7g-qz8oA{5o7{pHbSf-FN26$ZU?c=q0IXS^2 zZGDZ`S|P1nqF@CPhqXh7V8WkI)CXW_*&`igR+T`>t(B!mhhXXG%IzN<95A)2$eMrB zI-D?W$fhxsIAL-KV`n~nTo6FNrjNKnf6PagjeGH7+}rUv10LJ~YCa+ZDKutY7)(S! zM1Wij6}_x#1XYonq^OkGE(Hd~MHKy{z_tJSCQpU;;1)|WP10s7x$w7x1A)T@#aCA= z9zTAjCcHZzBh&wC%?hNvZhuT5^$WSGLJzgTTSONF<0Yn?EsHDD+7Y ztyX(uoG4!;$o-pBe|#225EFWUFZ_U5Ece7{r+KhDu~pDcn?kw)DVx76+uN;u4LC{B zr$4c%&?FWU(<4svdf{4<6&~}07I{RnKG?>^2;HS$%BzXO)3F5g-bC3;DyKn&EHg4L zRe#M7mak}Znp?~U_UYD0l2X9oAvn}ZqEnO+wsB=(9GLx*%Atj+oUl~22yg8}MqdgA zJpC6XqHmZldU_lC#(0Ky;3^zhX7fqM;Pu@M-p>>x_R6kxhlC~Jk?7REd4;r1mm2V8 zsz>I*=~wF{?@~(I$QLT{p_Zub5bU?$RG4mGU&=@Y@dmFzWD^%}DVjpz{$KM`fB}9(ZN@x%$9%GYwRpQ2Ff(WOU zQ!mFDl~Y#cfd=KTFto(loBS z42w%{AU^dG1hx+22077hZ!AW-mM5$#ulwe8%1SNPv68Fc6{xmHQ!Rxc=@x#d*dcD6 z0Ru*r-{l0Snz%7mJ2^5*Y}l|p%#87e*yW|YEL;?1S+`AjzM>>^s$(Q9pahP0^1dVy z_DRNQtWf+3);5ofl< z-ttu@Ox$+kHrr-+6s4{6o&hg8Kj$RYnljW3EXGrW)ZEXDGrIx$2! znDyC$eOqSBmc&K1VbFL74@8`%(-L3)&J3ecBA&z@PP%rVk;dwNPVBPJV~F+sru4R8GNBsFdH5 zB5(VI%xwnxwn~U2FW|B2^InZ!45r0?8Ld`XA)1$5h=6q5-g7AVUpqWd|ITQ*p_%s_ zN=szX>iVUe&ep$In=5j+*N{?wKue4Dp%VYPF7z_K#okGV=K_y(t z-f)EO!^#zdH-GOjnu18~=RHridAa1+_(d*E^d5692@)@YbD(BQWL|2Eav~qmAzWmE zi4lk+70?meG@3$*PRg%wYZuXS2TjhDikf`2uON)Hg1uts8jQ%Hc+Rj=zskzkBK?_c zwDgMH@472e*f!#vCkF1efTIM84>60$nS;CfI9k3l zbvrU|kJCa{wbBZnS;Juc>I(%7_*V;@-P2QT-X4}^Y=&t~1+B@aWOD@G4{5K$&Hs={ z>9;rb8)i}&BQEPQce}>~)b>bA(OZ4bT@8?3h17;7pT}|-+<*|FOQ0BYHcYv{(Z4Wq zIjSIR#=JK}m$9+W6@D1I`v5ZR-Tj%n&jU6t_iz=1YSInKtR_WRGiZEgvt+NwEcM*k zKKVWD@>=e{rM=5l=63nzT|n;j71~=mMj17U#i$Nmpr^<0!GygK;?T^j{9}M%9pUo$ z-S(%W{$I6~_^s3YdYa*$v`*^nDpZJ5a4viO?xpGW+0&m%PumB@wA>9{3T#&&5!=Ty z9)}Vu?MA{9J7*#is3CJN9+Ac5-H=JF?IFEpgu+Fce9T7GC8JtY6G(YTcIeVkBs%$W zz307=UT-5!4a4q-hNzYOl?!x6ki&h3?9i>T%=DRbL_+9ZsMBgyhP{!C2P`lS$H&vF z-rOf2CwGJhHYVAG@h#T24U$fu_ z>yPlOra!%FhMh#~y-9L`R*{x$zTIPpI_&Uz^6Ysyngb8L1RhUfnIde+DXRmBj=DnB zLTk9+n<_-FK$y=`TuO%bq-A@?Ln$0%tFtob8bm*aZOT2;H`#*Ue7*CZrzt0N9zC%m zsr8`xt%(~Ge76|uYjpK=WhfZn18E#n2AU=$@h#mlY4YUe>Nya^ogR(z?$U|WJmipeW-KM9B7sQMc!F>RsCjdUqS^b3F%Iilx~oc zMx?vDyGuekBoC=H(%mT`Eg;?99f#((3`N%mPWUG237Y{+L(SiP$2_Er8BlCX?~RIH!&k=1rswP@|9#vzxydL7o~Zl&?IP2j(};?|%x$OnQ8-$7xH^*x2Et z*+OdQccTbdA{=Ye4>2eFkE$n4#9Qxciy0jcTJ01%KSQ4%7G4R?|7z<~&KOLi?b&k^ zI2u-5xoz4U(wIsdfMhwgA;y&|R(~4b*;RgEKF(S;8_Tg6&9rK8+{PJ5J~kc}Hz^1j zuC`})cZtu07b7)UUHCfPWG;5vzA~ODnzrgsht6xCP{nZQ++Pn@CEZWc1kfM+IoWpBa#t>L??3#2 zineq5c9@urSGiSU!IP{-kt!vC?4@o<0+U=*?4}Hl!t2S;6j;*CX;~{$2!+wA3IoVd zh>+$rvtFuBO;349o7=>IQkCxHu8W4o6HE~nfXae%^#7^E!cz0v#T$5Cs);y1QGY4? z#wFTIg>A)zJa+bc9%9DF_knQJ_5xzoJbmA9SfPlShW|mmg~^ta9=%#C5|Q^6UGQoy z%dQh2S@4o{W3^Ez=C@IDQdscn4e~|I)DK#{+L>8niTWy%VtL?tixPlaRC59vN2jV- z{7&BtWv?w&40@*YvNJPiql2w8IAvmN!l~D-uM^EjgQ6qTtie!b1@_%TZDS}GlDizs zuECCKtGCwxZ=n&w1RgdVHxM-j61b?C(7m64o}`YHSA3Jz$YISWQ&+xjbau5z&zR&% zgTMGH(3fB!)J+s5{7CwYpQy?c`vE7N+33Yr2DY+U_~*t${uOQ)qvT7JA2U_k z<@onzGv^5mF+R14y<%eLFdX8vS|P>EYFoqgno#?BKAHDmsrk@&U=c}sFj<*+H0&0C z7Vyp{;jR3XVTH$3cGRMGi4&u2!gl(P?}eGJz?j>wq9p?mQlfr$B8_p>4U) zro=Vt$v4SuKlr5{Yu(Ew{={2y%ipWf<^K8BFw=|^X2*N zmzF#gV1m=E*^xAj9CHrYo%$?Y)^dMsvEUTDn%w-Zw4#4Gg3mjyDU1ESY{&{l^|;f! zbRZ#>O#xo2Qv_b2aW=@LjxOX=Xn5e30+Xty~C*e6AtjGtv^hTrB%)Lm$-&xtU*2NE%>1*oS z%MCjFJTu9X3=2yWFOMdW%>9lwS!^_;QK^s9t5=S~?747s z<2ymk%8F%|a8(1*nk5fT+WOJ#E^5lXQKq1*ap9vve%}M(9a#mI(x@^Qy1Xr8cgDLYIun{PMtY$z;r$6dJ>Ac%_`M^dlv3+vmT=s_ zNea80EoDjf;$G>m z=1XSv+)6y8Q7P8!-bm_9DRoh8)#pa7+;ALT@MnNa>!yXFHPuY)qi_0cUH>B9gh5@t zSWL!*BsB5u&G8>Km+^*S2%c85ZZ)RKfKSD-kMkY*b^Gq!7sox~p)IR{Cj_~kG*=>Ngz`rsF$CDrBt^j#XYmAV0hecvZj zJ~m`R;Cc+Q&jV%BduJIa$LblhF`6AoXL=eht_eF?|C!_I4-UFbla0<=boP`AblI%J zR8ffmKS26uzYU+t9f3@H(v{O>V#ms2&AxsmWdUdy?SYp%ne_S)LkQaBoZRny6xvLcA1>J8<8 zVR)L&u0#$SHV?Og ze@Ig)b|t7uFYD6}HOv?hFr;cs543r@1ee&JnYoyRY`Te!+JWMYxn^E;m$$x^NT&cI2ve#DAE`1c^fNQ_T$at*x3$fDxXJ* z>Im%FV8088h6rLS0^I^3nb}ikD^87oK$}^~YA%QMfmjYoY?u9=aas@Oz}3s)!~7Is zq+FTRiS1kbtZ+U=LM}(&fhP+t0d-kl&<(DL1!b{2N_@ zO!`uD1iyEvRcsBb34ZG?z&O<$FVLQuz|g@cwh}}{2iFzm!9BL_r~K}FA;Dp#=(%95 zDHJ}BH%GZFhVw4-&^QTF-z@5A*k$VxXVCH!`ntbe}Ve0j378A-(BNpc7*qZoABc(cSyy!|z1mL5_}@R^0^w9zO`(l#Mef`;0Q z9t=`{d+Wb_00O699PLO28eK~G6>1f5?X(c|fBJ`j@^=~K6P6#9RD$ZrSfU^123tF& zI{R}2jmnL|6h;l~^0d3|2m&j#7;BKHZ_dO_X)6i+>yI&l1pDGmQ`jt=hxwqFvh~2M z^nblO81z`F77-;3218cQpU(7<+p!u@Q~;lp-FjaG$y8YQBvd?<{G~Uuex2KCV0ON|McD^bqTgPV1}|-4H|p(wD4k@Y za2hJT{!d?|1tVHym-fwc$bTK%f4dxi|4+d-L~-y=pZ@=TC-K$7?|UndHt@iDU#;Wd z7XSk`fDf(Kv$Ynsg0Abc?Xfq?fC~$rmn*G>2UuX`gaXDiu*22>Qjp5+rZ_;iIFA2r z2CV^mGu69*o^*A($#^jw1Mt3Ed-05ad6~dMhheoAAxO4An$r*Z%!Gu5_L{+x-3Q#` zfC4cIF!3pPa)@WsCK+A{eLnYBo1pa_SWPV&CIM**KH3yGQ)j_^v5%Z)VDy|jK)V39 zuY+;s8aR%`xkk9>sGO&TN`)88Ur3Y-K6OC~s{W6A{_pn%JORv8?%+BK0L$+n% zMu-{Wah$cP1+D4U5BGON(mAUsf~u;xqd!JWZ0eSS00t}tTL26N`at4!1}4f5p?>?p zVfwpms}dKi&NG@?)#gf|jVF;_V}n&${a>rl-&e7J{`~y-)CTi(z+Sz!L1reon$&X9 zic^WlBY{5(_1Y=OlOE(`#9Y?}%{icJ%XF^p9R#>73BiyOtHh;e!1}a=Wp@$A*un{r z2w)8a4daWoHaZ3dl4>4Q|NM;q`3vb30=VsKc)pI_2nWLjsJ#AGaseDaGReWJKo&Np zS_uew@a?ZOVs+S%j_ppY0Sq!R>EUQHVUSdZgtk$soWW^)MEvh7%fGEM2pF&js#Qg2 z4*(Xlt!ZldIwSj-lFxn(VYdhH;Zxojh?%rT5%C1|tej8kWPny!6aH<$!2+aOF3=^& zG4B{#FQTM46e$OeiL7TYGH^>7^tgbr$x|y>hJidrA-sYEV|9m@j-Z(y^q+nC@D^Iyq7K?B&7xRFw%d0>> ztD9rdQ~0;GyHQX*5Ks*O&X*$ZotZsC_)_1Wg8y_=){5MWtLQ+qzCVtiqLxW6E18>K z1-5Yxi;01arlSrdAa|HVTmkxmWv;{=cr-x=DNuAzK#fUyh_X7ebS6 z+raT@lP1GcC~S?#CAi`4Vpia0uM!|o8s0*WvCPiFNxo;Z4{Qa>|jWEz1O3CZ`tL44j}Zk zT&5F%_)(L`?#Fp^dpgXhR?)c;z2uez4nP=<&yX37nD>{%3TrTP#3#Ah`ahYXNY4V5 z59VqIvEDON!}!n&*Mc<(fDOR1;1GdR39d)-r)#W#yEXs&W~fWOlwfM5bJ%n$Te&c| z5zQ1CAse3{ean%?zA#oNs0?VsMl66G1G=K=RLk1N$?fU!39lGDbYvd@rVlJrYF8Lq z!w5xjqQ+iJe$SOlG$rJY2G5F6CKVmUhCMs7X9mE@9#ytxL5!9BqZnYBf*ICn5XHBl z_F_Ku{Dxhe_F2*ThX}TWvxJXN#Lm6%&L_!g0Mo7z>n6BEF|b0S^9gMn#SpVJYL$~p zY4lh5Dj8fIZr{c;&xEdHbVju*KPTVbwPPhS_+k>wX&Ji#l*3LVT(M!V<1QRK z9V438jpt)a?fz!y$B9qxx9kM&&SKFej&-|~Sg`43lUWh z2fhP4R^^BZB9?CrXnK(UOfOOpkm-l-l|U?%r25>pjbP@FK?bJCQ8GVafCAG~t&BRf zpCnHgZQG^+xUGbNU|3-ReyKu+-;=lW?ceGv|8vp&Ta@?ti(r*@m8k|iAE_&?7ghHb za0o`r$XjxKJg6`kCfN(vt7 z>^?_&y8LuWc+j*H%&B_UXhRld9~JF8XvMsLVkT$S--+i7=RI*bfM2V&rnQ2K`RB5S z#HT9fR`TVW%lgf92khmVnXT?2PZNBZjIHO?AC^ti4CDTh3;eI=3PA{V#C>E%#_UXO z%#4+j_GyJ99|MIRU*m|L3>9?FAxfgkVsZ##Bh+I523xkkH0-h#W9S_0US-!sW>52E zFaD56ME&RunWtgs7R#vaieZ!Cj`Xy>slKLr#hlQ1(9r)UTto^sWbRj1fJhlf7p`qe zc*)?K#)#u>oKTq*c8I$>Ak3esIrMpLj~6CN%*tIjuxtZJ9v5F)>+Ok@qn`5}^G%1^ zjVdKB<7YPhI(CMAB9}S8C>G+uk#R9O*Ua!rcFK?)g0!5Ie?|5G4DbK`3n?yw$gb!k zHc92d_6sW)Jh`pb$!-;xk0I+p;Pi)jK#5ln)CY57aQg2ZKJI`Oual|kY5Xwg&0IAx zhb8T_!&3fMMO4YGPiVm)pBYd*M!9G1Rrxdlk+CfYY-gA;6GoOs%AI|H*LxG#;f}5H zL$O!MDgvKPlgkpp*ipmL7M0Q|vy=aF$=_9wAWuEfi%!;Iy)O`2@+Np=p-fM>YO;30 z9f@rYFJ@xm7T`q5(+Iq<@7zJp@x@n+{~fXa<9|T&L}t8cT4ZYO?vQwq9iDU?kDa_8 zXog4~=jJm{yK2b;p$t195PW`isom+)55}q2hr@OCNG{iS=Lyq7lr!&W_BzEZHY7Ru zG`vLob_)wSS(2VCqiUY7ii9r=M(R;#Xg5C$x0`)IC3=R8tJ_qz9dypyIVB-x?Ey`!hWh`JFH+eBYrjpQ4;+e2I>8L476kd?L(}zDcdCa}bq--KIEFAAx zuYl?P%(_h}6F2KnVHS{Vw7RQIS5HAW1t&LKCv{sw{U27{c12_i zU2}*nuKFeOogvg(9N7%=Su* z<~xQhc5limfQA05K&Z~o!D|2M0)YHr!$m&mHq8~ub8;GRGKc#>r`?n!-uSw8@8jm1 z%9VvMk$%4u3#Ai&K!yd8w&Je?Tg&;4ADaJ58u({g|FQ~xg>qyjtu7ea$FAud4e!Br ztZgL=OEX}t&e|7x1ENGOjYkwI*SRB>Q!>|>HhZ3ZGL`B~5F#~iVV;PTG+d`lCe_Oq zz7aLILVCi#8Nl{VHG;7*&Ly~o-sd~jVwwwzNgJPl<>2sPvksJ7Mml@jK{qpY>m-1> z@$|dPJ*EI46Mx69Xx~p?%NY?9g7{h*^;5;T31A&qpR2(#E=}#yW&VdSpUY3tZaZUs z!~#SzHN>}6V7IcA!=$nU+^g5KqN;nNIbrpVAfB->FhBwzvJT^~OnMFvx2JL@#+iBT zLkXIWt6(^x^H-WpKt}_UlhfWAJR4D$F@83d!EB4f4_0UO5;IE>Y*QiJPQHM!~h^6GcEFPyX221O_J_?ZB`?~ z<9W$sV3@s^3%=f421r{YLaQ?0kB5=KQsqd_P-?`GRSnFa;HSnGwR zs{lmIEw~q)qT1e9MzIe^KheNMCQPvVe?u(mw?$C11@d6P3|Q#K`as5)PXnTb8pOik zuUr-rbgY`gKcTt*LrnVj6$??0RMrpGRvoa7c|TK^zc1G9!ToqGnLuY&)e%5~K@VIN z$K^f5HybH7;#un4H-d{VUcEIGnF7y!{A>Boh2b4j9|InTM1wGs`Z;7&yrENlot2Jm zz&hV(4|6v*qGQ1es+FbKay`7fCm}A5@VTP*7^e#)GHBSI2B3#i=ohR8{ajuDMVtu8 zIH#z>8=IW>)z1_*sYU~m%JeWb6Pq%O`{F|GdcfuIpRV&k&EuPI`!H2!lQlcT*Y`Uv zW-={--RdBX@86ZF{}F3^4)vW4n+Qs*Mr53)B7SYL54QL1Dm@hb3L^QBy^hcF0b4lj z9ShQtRSYgcXy98(%u6SA9I07l0P1l*n7g{aIZPqEWtD*q!9J8<6L0|1sd6kXz`g;B zm`t)<-;DfJKr?<%z`8DG`YUe6@6jE2Hfo7@S=~Cg@e^w|VV^&NQjmTWn@RQ2clw__ z3#gxmE69>eJ|cPQf|yT1sI$Sks#53>{01CP5`M1gg(gz*r?|fFF8O--T2xRepb20IU)^q&c}R;+V`sJD zB7UVWOVxk9tQjKT|LGdIoj<5C*|CHO^i!*cqpKxob);6*RT_AI|F#5Z!nS!vOIwxf zTV<6(DM18%+hEZML3*zl_=Nw7h>fl(;BGD-luM}Z86Jbay%D8fQ>e6Luy!=xHs2l~ zD+%u*Y)yv|)n;o;Liq|7+o#TTpnT4Aijwfa`jK81$clykl|wICf%d(S?(s&|8Kq5& zcvL>=c6=bj4|6~211habMl>_2Eyh{A0|TcmpG zO8-d?1dq*2gm4@y&n!b>x`ff2g+3s~j%G^?dsnr|XW8*W=G}LT!7Y2zd;+%L8*93M(}Dk6q6PxZN8Q;Bbk3$rgt>xA zWguivtGX^)_{(s>7lD!&#tJ9lXka|^FILjOMG_)FB`o~#{x9VjaKmCeDl{U(r9uDU z2t}3}nmwSQpa0K~ih?gmLpdG!WZWAfc-ec~YsQHmh7c^6iTOH)v2x$P$&~)ZUu!jw zvhsJ}<3K@?*otGukUaThwjL*vTeL7velB> zRWM(QUi_znY*PU;v=&tFPE;Ibib!bnuXeE9MI_dA(+vW53qgY_0@sUS140{UH6 z1F`S|&{`gaoD}PZZ4iz|wExNI2hj5*U?W-dc>t2(IC#je6<9ki0PaUBfsV{>(F+R7 z4*iQO$A&%tCkk3ully2D^Fi6A-qpNI*#j`UfEomXjfh8yW$;XTZ?V+}e3~aVkSstj z0F+p0m3$-uCQ+&S_p-;`&e21G^A@Z)nG@2N`592qG!?71D<^R02>QrUY+C9;n7wuj z^gQX_gNEIbO9M6_3IQ#3eK4gK(1wmDL(Q@M3g@d} z5KofHkXXPL;>StzDoLGYF%;iyaKFFp8q{#HytVQG5VKHh#A3x3L%^)jV&|3-ht-kG z@Vpo($iwH5%HZD-I7=+mytcZVx$|(&k#y3Uei4s^^)}&cyj-LsC>_+f@Be_7XUiq> zn;bMvT^j8*%W9nm3Wh`P&bDJ4T>=hgD^CHd{R;ezYQE{CJLJz^RlV!+ZREb*Y^|ki z3WJvE_spEW1gVtnJD&EL8zW1pH!*FG@h~brD+0YgagS^!&@Q(@SNwe&92&fG=6s7+ zBRckna9nhx5Kx(KwOVN6@>d6R{x+%^Od8p<-AT0VJTQ&cdwv^id3->I!*bh9bHNbI z^jeQs0p>WZ7SF3KN_9v0peN>=uelw!jwUp0j0Tg(A60B{HGz(~KT%BJb-xnENI61X z?2(x55vHyJ%FRKK)7B)aB(9aar>?*>1Ax+Z11M(Mevl)um=)->wy-E{uSt@A6gULg zOVfVMY-(&dKAU-!V!uYKFOKWfcR6G3C3TwYf3X0RcKK5on%p_mALlcZTx!!6n<7mJZ$7MD|SU_p#?e>Fvo;mEk zySbp9un!fKf}govxlRa_|4-jerFf3x(iy~|c@h80@$>)~ z8jo`X3MVWtx{y;?OosVx6`vOlZw$r5aV6bqd7#ccxHhi0S|61YL!L&r{m}8=S@<59 z_#%-}W3**kK`cWQvQE-%GMwsG?&e>w5x@1e+q>OgQF5lDi@1D z>4V^4Yvcp^>K0&l+z3@(D`(_joE{y~ReNoOoLi`HTO9RIseRlQA$v)~I3u!cH)<7^ zW8(Qd>*2BBVH*j%kNYqkC2&6_1y!9i*0KO2yHC%?ZE6 z!)*6lCF5|Uno95aJ+F7gV3=2Hdb8mMN7x51-eQFtlO~s-=%*Sr2c5aj#XE8WPOW)% zf%a6E8&ilZty)c~W8g+7K3>Ok9j>fiVKXwDtf)AVkS6gWjBL)YrwO$tnA-EWN? z&MVN!xoS7MF5k|UIXSB* zo;EXZXP0*81_h5(V?**2c(B|&S<4=Q-!TVx`0(JpoF1^E8Nv`f3$C}C5xMW~;lah3 z0Vt*zxdBayD(xpbpMGNq5|$r`zn_(GqT@p6^uBck=bA=32{ei2l)I9Z*{@4d5R?4^ zeX_IanX}prj=fFyYc~0m>QdBPR4c2S^R7`CR1`)#mA%`?zg54I_`MhINqLVx_Dycs zncB-BvM58&vv$-9<)wCG)3tVg?m4qbsI6nf!BhjE{jtYg&n~OS#_`s`FHp`LZEMen zZgaS(lzSEsB+^MNlg6!7D?7L`@L)H6Qhhm!fmUm|ovHnHBu5Z^3Xk~0eG-S=N_Dll zW11y*nJG*rPi3L*VfokV;Mzr{8-Keo?ix#xdycU@R-4&gDXQuneysIzmj&19wc4HE z)faG4$yGYt^2&wMhwz?Fhm|d<`a4&;dvU)%i%MSui}*8ClF6m^M(04AJL?6ywHsu- z%)52l?UOFMte}rCho}91_G}C$W;p&(EJ)#7JpOXHQL0u?sH4uUP~~cNa>vY`He#}Z z5v-hPap&{m!%p99((>KfpG7hXF8VX8_kPg#K(udrEiKLV+%n{c@tavYb#)k zoSR@!Z$OBiF~ztn_7!?=KHhgGkZ1xpD5YhFm*`Tn%DuMHAV12;J8m6Nll}es_m;VC z(&L;8G&I2OVd@V5Tx|^AvGPJ>PoIL>K{EN{xIPSA68W0=UUAQ|Bi9c56OU2 z_n996C*8p~PaIP5SbKu^u+N?{z@tr4CSKNh(?dh z67+*L8x^^UUj}K&MvZjIcp%rG?%sA;k0Kt$Gjzq2DbR zlDpl5I_r*Ir36i1g6HeZQr;+i_eAE=l}qD4gc|g`BNo`8&l*nt?0K=X;J)>1*nYF& zcyzrV9N8sBnZZqhyOVj+_350WL&=<5)|h!BT{r@I|MAkQ939M2gHSs5! zK`ZEn#9RMQv@X!lcfMW&gH@e-E&3YMaj=ukZFmbX;A7kw(;G zj0PG}7$M@MJ6G7>#npUFTDeDCdx$-d^P)bGM7SdUq#WpvR1!Jp8Tnewl-l*?9MsKL zvlIBe+=n0Y!N+WW$dkP*6Ru?OLskC)JP3lD$G$p($$V-lQ51Q!0+D&rd)ltcPdmD& z-IKb#Kv}=Dwt|Z?38&?_n;^YiWBQw~_PmeJS~8Z11`@mns;4XtW8{(SrXcZW+>U%M zwhTHiDmqKRs!xS1b)9bvax6L;w_L7-Yb~8z>rP2I>;h?8VvQGndxCEKl?7>Um}lolTHZT8F`=gj zv0~K2@?Z6EU^XD>=ucEp*>u}{fyOcWDYz1JCf$Nuz!DWvNF6LuBouD)^Bn?6U<4eM4R{yMlbCZHlzCQ)cd2c;1(7qBQNOFv|AU~ z{*xv3^60n)(>wd{8+t1Vmtwvu*2;Kb7`lx%m#rU1$ZG)~^=ta8txM=1w>G0TF};%J zj$3Bu@!T;GPy~wfcLx8yDOvCY!b)HG36`E+U!3@+RUXU_a;IC6ZIy7`dcHZo)t&(( zURKP}(*_%Q7<5`54YRecPPz$f6j=DW=mSgZfIt+|$Wl<8A(ADAzKg_})PSe;Rf1{% z9@L^=Cv9M~YseSw(?KFCRSQkSuR4j)0=b-NAP|fjI*Zu0rmwgWFC)kHUj0w5y z2uEHRCK(w(eFZI8y6xl1T3*G{gj!jBIDVv*Z7z!5TbD1cL41BhOpf4@&p;&w8^wDf z4(6(%qhCIa6iK1$$$!s$89^xW6_=Ug2PzVt{ej z3?;T%lu2%Nd|XzAK11TX&qh9Qt_?vjw|l#gq`WEct}vuLE!CvO-iODqK}y+y+2D@> zKhB*4r-L{Py$S5?yvrh2UMT|sV;=uS!hxHz{%4-X6TA3ot}`e|Q=DXeLD*w^bsl^@ zJf@uE#m(SLv~FgRxio5C6N6D;ET>x>V)S#G6obW0G6@5)Lxp_M5a~_(4(6N37(<-r^4q&xzrK8K zoYt7g53<1=Vy(}KRpx%l1}K z%Nd5>&#lU-RdPwY^jABqI$Y75(lUU8ZLdaQ6sQRWr;Qfx@ zc@LMn`W?`Evn8T!i0s`W7K z`gY^Et<@lC2#dyC;@Jc}GNLqB?Ub2P(NN`iDOvcB>#AnxX+OTEiv1{erD-E^%)}rd zXuRz9K-ZG7hljm69b!#_P!={WeFRLpQ9|e3oq(Un6C%0#7r#8yZS~-(mSVYT5qRwdnqzPc3cr7%-1hsrUG%Y@~a^hgMc;Pa&Fl_J4=uyC)W{?zEl#cY9M z+d;5p0*m}Za32nz$IrAFWb|^`F^|rV8tKZu5l2=Vg)g!CbGGh=lt{#R$uois0>cZZ|+T@wp0Rca1?({r=L z9wwH9_SRRh{0lCOMY)kA!pLgK2L(5}4OQFCs>D;ZUkb&`k&K7yYSnwK_mgGu1fJFI zF3itN7RwtCsl-T0OQV3yz)Ud_BivUbld8klj9@>?Wr}&zk z_>|#L>JABrj2Wc%n6BF_+^D~GmnI;-9#dn6%pU*9fKM2}+Gj2YLK1K|d|7i?dS-z! z;-wyO4ptG-TekP!BWYIIuxAu1YR1iS{8yBW=71xA*NG{oDhd3y90`}*9Rl6{lA3Hi zuI5U*$!#gP9u0kZFIG?@DlEyfx>VzlML<>lsQFJ5#WH>1ycr)ZcP>A$^<}?=?HucD z9c?Uz1w$T?~c(-^r^{9(UY*MMXLl+Rn#|cH4@^wnBANYwfY}6J- zWT)!PZdOrvz??%E|Ng|%B*ChmALQQ2TRUv_`Q_1PcHPkAbF@}t$eHuYOJ}NyG8GPj-DHu785bTlEF9CA7 zON3jRV)^TMf)=Q*jXyrZulPUl-lB%5w^1t^lksHD=6)2iNhM)Dg()KQ+}1ZCiYYuh57C#n;q=oUH{{lrU4P%!F5 zs<0I%JZFMSoMYiO$8pEJ&)K_YGFg{IgIO}rA>$tG;onYw013(sy>N!1Vxy*O<(HWP}B^s9oZKiufB4$kAAff!ij$(da zrhcOKV*{@|?CRZY*7Am+a1GscwMWO*@V&c~b*<(#O6Va@nH1B|`*?wZF?w z5f&Jvio`ghHo*>n=odr!*H$VDYB_Dz^1i1q_VksFz(;s&$lT*RBx?15&Y?A?GT{m- zEKiBM9@m&fX~i&hORWrs463KixJj6y2ebLTeI?~`@UmdkdH;9;#jVh!-$mp5C=|2Q zcd8BS0EW=&pg`YBJGUUZ#5>l)gB0{69)y^QgZsNRgW;XUS9Kw7xHIA~Yygy$J zw`!tZGaT1Dn-v(L)6MTJXk1RgWU$kDA(O#u!~@-ePASBwTRB$e$i{YSyMLD&4jQhh z5IB@cuC`#?NohoAn7p)i<)d(ANQBFVvULLWtyY z7G=&PoZ7m22-o5$Rn1*Cd8AZ}zg}Kh>LXI7n26h(25FDV{KZ#w3C)gn;+-=7=Z?IH zA|Amd80%qzOfLmfsV0c|+TnuR2Az4FZqx%8xv0llAP7O$g=nDA+3Wx0(0y7n9q;MS zWj1=D;>{uow{XQQaYtAna8uqt`QFXzg6nWzEEMPU}A^$3!o2Nocq7O!|g6> zel9xy=`@a+sIcfk(5TGYx++rNtGf4A!P* zpu%m3IG03gpO>LPlNHSFv16Fs?~ps7=DD}2k|D>OfX0jOz$Dxn?zQJTa7)#8Ga@g1 zopRJ3w|}+a5s_{|WI#1txoB#Ye*Yo5I*&5nuda&HIml)mWZ@=!k~HLg58_qIA;Bl( z9Lu9kE?7D$21{@7zuGLyg38K*np|AVoif96TI=7ek`AkMO`OyHNKmDgsib^_A#S*@ zk@ou`cuyIxWH39mZRrWKRN%?$pbH*gYNfO~T0~w5B8j^MRS1_0h$qeYw=^$_Ci=*v4id57~rJpCT zkJ=7!6@jWj(_EL<$Zd`o>Q94_A?Fp<%Fr<1uqO*onvh8~lLmWceWQEeSq*5JCY9Bv z{uFEAJXkAB4nkOZwtd>Hn**-&3pIG2cTaiSmBshGTeh(=7$LmJE`OavPW}=J)|0|| zF?U2k?vGX@zvWoM(JdLryqm|XTRk))eq;0gluez2uEpyxEY$azIe!j)2pMsc8P_)E zEUCNAE-bA@OKsaDxDny?LVLPoUcHCVyCj=7=O@!t$qN#>GLz3WsYPWC&(z1&qyh*+{7x zr`576;&mamb#D7}1wto0FYNk&=orroGVbDniB5Z}edH-~hOT2wm5;Nb1uwiNI7xxy z3xPv9!zeNg?M*WqivYJl&u$I@@)v(bGkFXH4^2UC#$q)sG_=~-I@w`+Jyt~$90;YzqbN9V$zzY<#*_G)pM6(o{q+Fr#!NEL) zz%{=T*u61AL-2HhvwOx6+#@)Q={^Ncu^B$9(rWt8qptQ+PU6Y2BX3*-));juXC6^C zit3H6ntno}GFIMrH&ZZ>E?U2b+XfnZihiGuWOk@Vd~suWa2&(K{5Jz0Dn%mw@Fk#w zg26yp7D6~Fs}EShZ6*`v#OZu>;pN8@YdOtacWb~5EWTSb-A#Bjk3Oz8CzY&b2K3d` zXCkZ2dB%E)^U#DYoxAIE2lW}N>dd`(1(_1na`FXM%U43-S}Ls4ncPWS?~n#l<9m{* zg-YgA>BS6Mk*7KbrXIak=}!l>9Xy}gb_pI;DT1mmH_Sy|O>0GJ8X^+XYAp#Du#tjc zfC*z-zsz#~=926D{DcLRgj#@FWGRgwCaEX_?@6QknaE&RbLN$_0We$V5?7|dDJUC9 z8LoU|ay`;zKqtJq_dyNNs%)f=?<^GA^{`2UPdzpNn`9ddKSgf2vZmJaz+G-NpNfn( z%H_dqy&~=4E;at(Ip#PMP0Vu8FQH%@lW%7c$YZW(uYGrV*_+{`CWrc?P~N6c=}n{M zmAbuh>AYs1GD_ZY&?31DP2qtOx-+&sw;qCGp?6OQmyXp;*+yH-JlXd^%Pbs*)Enb# zL)qSe4b^dL6kbe%Nl}LQ*BS-Mn@TE?I^87|+z~u5_r}dL-jK#nWI%_$;Ttct2 zExX=YC^O{I019Puv9ZV>EDy`sir#Em#riJFTjrYc*D=BHpXz7W7XXi@0x$72hr6Ya z+WbfVf?H@M%?Tmb>n7zE)@QESz>hzfJ+0HP7FwfO7;Z9{Y@~fj5raF4k)R17FOk3{ zuA@$-RIws`YO-!8!NSMqOlME@;`|#0m`*;c_>0|$GV|<#Dub(Jo?)V1peRD#7vExM zdF)_d33*kJUpO61;6^0px+E@`P$rN=B&7YFRgika14a;-(7JM0K7Y!UuRi3NBTCAN zs0e0uIVpa*d-_QAE18{F{Oo|W6A#QL802jtu(fJq%io2+NM^_gK^eB4*&5hv!F;${ zOfF3AaI+mS1tU);rQ1VtvKwu}%s=A!%B#S`X0~}@+?w+^=HvBLl+l^ zg)kd~%!MbtxN`HMr_^ZKMZl+#Q%J*YCxm+ej_6Ozrt;Xb%L2jft+MRKu8mkjTdf)T zcEg!|<1hLkZf62tiejQSPjwDC=2DeQl@~T>(UIZPtzI2hC$FB$^#CK${wXQG1}q*}q)qQ?*LK1_@^-mP`E2a_lcRqZ!&kh<1t#!m9#sRfYni% z^H$r4k`AXa1IEmk1u}O-X?S}r5s$uBjSPBi!sY45ntpQg$nWHkgeUNxc);*Ucwp$g?5l6{U0w<<6JNeU>v-1-+KQp)hs{8nIZ`wSl+O zNT$Ur%&h+z`?89S*$li@8#I9=6{tk`#nW3gn(H3$nD&Wtgj1-6?C8OfZC*|XOo{5P zN8KMXJl)Qqs{!ALc=Wpc&$XGZnsuyp>H44P7gQ>)Wej>=`ZWeK^>2R`72vSBxo&~` zaHKR-z1ika>=A*7b1V~EXr;lu86BN=?B0L0-F6< zeZ+|kiiSza+GZ{!l?uloAtWddOY`2qAIu}KpC@;peOMTy^FY8q**87jScpZhLrVLQ z3dshKM+KiHCSt-s&z93JDDab|b*HsKSn{4Vjm-avZ53=E?)Kue{HQbRa|69?44pY` z*&Ix4y=S4%H#{sMN>k@=w(nvlpCz7Z-OQPYm@K{zyzSZH=uKf{MyS^5wj=pw9?awS z0MGr6dxCWtsTUd9?<*eK7ptG0f=hp1#kJu_>!#4%+5%^Vc=C4IDTY}~sA`_dBo(!f z^Suc}1QE|f5UY{=tC51>gFe>Xi`fU1b$_GBs$E!Ua=m`ni_XC(*061^W<#>rGNn@} z0}XH#gjv)nGG9I_rcvlS2k#a_urNZ~JDo{?6ov4H6@k%=2BQdrPfRcdB2AECV3L2T zq4oC_^+d1;ppVUA1$b00pjf5-oW@gbTfW#GkpE|&;uBcv;L!KzTD+NH)FXKMPdlh! z{m9$qW(O4WZXGxa*VnH>8_Tcv0c=-brlmn56EDU@iPkPxgbC3r2Am;$*z}9S+NvSC zc~{XT{f;;Ijv(3E7m4$8U|JXZit4JO^c}&K#@CqJeaga$r`hRxpx&9Wxl17LE;06b zBH(~EivhsRoS7iEMZ$oN41sd-##pvM3VWiDESwObC!5vQ`Q$?*lFOp@{)kE9ec9LG z=hYftaEy)J_!J5cBAl4p1blkm@P3ewG<#RCJ3x-~8?hPLu>lhBlj6sB#)He%c zGB}Crzkbt?7az9p=w(?>EH|-R?~G}KhHcE=)ag%ux))W`%Q0s=!J)HcQ1i$$`cl$- zvA=M&J1p?>_?uGJ1Kp?A(}kM_Tu!=)?xD|;G%sLkap==X7q&p>;PIB>PEdGo6wM1j zX)zAudFEWN5XZNn@%){Jylk3E0sJOin}B9+bI8l?lCywYpw3@APNT*=2Cy0wU_KB0?GZX}Z2 zs@bUh_)0e+y)5T!SH=j=q_PIFk@MMqN*6|d|B9hN)G*XAxt75XNx((DNa+IeO?#`f zs)fsP;$Hf>lK*bFQW&FdvEbFS!i3H_^;d{uX1Hc23?6^Y2dtx#!PDKz!riKRGKcO> ze1YqJXsi@A1JQhv`Y6Ak3ltHBH}&za8m=>(BnR&kL6uosETP6e4SwfmIv zc^yVV&DqSkc|JBNPDkYpH55*G4M{Tsc~&mD5bI3enBmHZ(7=j0I8x9NdbiZVH}dM3 zmgFj(v<475+MH(8jzWTln|SPnlK+^cLA>#Jklv=fhA(jtbrEGX1c;Z)W*lwT_Re>L z8dVk-&5p@TMFXLiA{kvde{Lpa8~kkE?A}j(@9)3o-6{k1^$ibH?*br0o{zeJV_4p5 z(EItUK_qNMB4G;D2EVjbyC&Ipa(*JWDz^=nAxL_`b0d)&aCBrVH1`Q{nY3wgl`ctj z`*d@-vKW4M)%GO=@WlVfzdxc>UbJis&?mLJNUMD0xDg_5)rV3DzKUXl;qxdLOUy>v zX`$Jk#n~NtzSB}U6rN@YDLTnPS3y1;!al?9a_)&FqEn#w`oyjoFw3!a$vQK?wKf}V z3}|AJAZ)fYBDaSp&JSnaYGZ~)76c>IA?zV)h@+xbmUwP%1IHnhgtx&0dcRvb{IO0q z7)a{i(bq_(X<5*K{bJ>+`|MST#ec$WeAw5|HE5qH931f`=I+x-S-)_rsbhplsO^@g zES=VbKsM8=@O9e;>3(YKe}{h*r)^YWr2l#a+}VdrICTX~--vNUwPaRJ7~#e16LKc} z!yWUh2<%O^goHH&Gn}$uo`DA!@uuRpfG4PEz$RdW(V4|+fOzwiX!H;)yY8}%L}U5(%5aJ#T4hLBxo zOAyCX(767I1f876`tF%JB>8-uS84_vi-7;;HYH!^R>41e*WNRv%eaIFefS?I4=t`D z5P4&jOYww&gMJ$1!%E zlw|9B#y=Qse!gku98gUMs+m-h9-#J7UMnplum5=2;|<1Najp^f{YD+v^@C)ezBWBfKQalVN-aW@xNWJ#9nw)NdGnixdQF~tG`>S`PB;&7-5j&|JC|TI}a@L3Ie0% zGXK}9bn%LfB)PR>|N2y!k8L+C-#Wxar-kBwPU}t!mCnro!_Oe%Lvr#A`T9R>jlb-k z2b-7wP&#n`Lo@l$T|oPvyWl_X0x;4)=Z z0n0Ej{n5_cjQ$ljG0MpW36c2s=oaALd1H!eCLt+FTBlISjEn!~9cZSm$XOA=iOdbk zInd1xLQ+Nh4a4t8S2NBCXJe_igZ#>IrqSsQoa@myrAWp}@6ErWr1+*K1vp(hQFmJb zmy6+57{pKXV4Tx4$3{Az%&Hf%(C9K__DtHN%6jv_@8V(Z!zCJe^UAY zW9%k?ErCG>Cg4BeaU9xvphYvHdQ{!6|Lz-z`1ZjO*qi~&!DT8hF33T%7z5xVvkKtp z0Xr0hxzQ0_Rl3sMG`|@>v%x>W`9oM`ohI&Ow?F!b(hHd?=24gVN`gIxD**&|Gzrxj zjUhJY-O*dZ(fl^WBy9xI9QuHwWBLg8-qQNQCrhhX|8U0V*4of?G=Hu$rQRWf?Vo<1 z!m;=7lEm~d$+bquR!*3M-|$=|_feaVAOCcN(*~>-dy0DO0XTQ6`4oqTTMv7tADO`k zXyKZK^L*;4Dz%69%}_zob=%kyo=wkb839I)^Kn&=yVItC>p}@hc2^vS8lT?efxdao zUpju6QtrF!5qMq_*e2sBuK1&*eUIDW{i&HtKGN-m>q!C z+CUTdzs{RpLwe7!2TwD>3LRjiqS$?CFI)|P@dLm^04US=gde!CntiT<4$@$oFGm0Y z0@VLSax-A~mq#_7V^HuuF*yw2azsm6Y`^)Y=py#+oG36Y{fS;(C(U1#P>NuC@zwysdfnD=(vQeYjx0JM@Xi+}uF1=qz@x@Rgz; z___`qNo~cii|3HM5p$ z%UW zTkh^bB8JO7R)+IEFpXqYu4mpeB}F=CA@D#YKI_T-9<+%67_fDQQ}fv#^!q3-e(jei zIi&fd(*wD@X7_wdXPX#dhI!SOtA1iUgG1lg7xpP^>~*T28X>*Blc^TR1=+?VJ3Glg1Q;1gb)yVC@czTipjl#6D0 z`r+4CaK+iZM+yqk>50)dECE)gEtov%?{<1WbJNrmLHnbrRwc($jOCWDp!KihyYPQ7 z(r_$v3fk=*Y*waKVl3Cby^_Bzeo>A%g()pSP!$pG_PvwQin_Q>Fay#MBHFu7+x_(S zu4yoA;h88V4VTYLYNown%kG6kdqZ8eVR6FGn?4?X3{}{esfEq^IfHtiRkcfgdCsp; zR#M?AQ{e0J@?5mT{LO9C6UlS`RjGKXjWZIO`nJi*IJDCdB~R9uCR)=d zmVery8yFgY(20`#Z8=#tc61w_qErv{nsITDwsVG|#~W#Gdp&lq;sBtn^`VQ_my+#9 zQ<+xVG)*b*{FF6M!1rL4K zfr^o%Y%Pf6s;q-=6mxh~txHSt_Hs`m-!AC9kaFRdb|#)p-{0>`C`{3J&6?PMWY-{S zzdWqdSl5kp!}N)BoLf%i@o^z6L;B){eB>3`HDKXL%Q6yU+k%vX-Cr$d2F!41E}y-Z z?#kI%2u$|NyT`O&_{O(_FpqKRzY5?HOIO*1M%9W;PMkDZbv1?Zba(i?Jp`YF9a^a7~}PIeo_S!)^jvvvKExyIRqvRcBZS-C9ib zb{wod?x=(`x8LeFy0Er`2Ep9$bqUSBUn7sm#$hp+m>hYm82#Ex#~ci`P^Jpo9?_7uWi zDXKS-v%@`$Nz8S2^X69G${_`BPyExU+CL{UL07Z7#E)IhVmIp__;n}14<0=~jY~s_ zpLdke#Bo%RY-V@e?+Byty`2ib8W;f2sJ(e}M-3 zn2XgP1^SV?Cu`4LiG!NboSkYfaaaR_=#2?UGQ|b$#}+2PNirr~s2gt0Y2d_e1Z&sW z&I6d3abN0m{4Of!=2Z@sc`_T#A4Fe&rHPAM3jkiuk5nPJahDVL%fK;V_oV{L4*XM( z%{jH~K{L}y+N-$lg0^(CAOOJyK*3rAd5k;N0hqnqxQn7%+UNxuB?~T?oV{-0G*LJe&;U>K z`;rA)w;2VDv}->a7gVOuy%ar2z~dFLVG$IO^E%zVfv*-3cjFeNeX@{auzTk!_cBhF z(Udkc&8$o0&iFR)KyCSiEZuCt%}5I!8xiexd7uB)FYs$1-c1po$j!5DaUOj&@++WC z8;Hyu5!d@za1Ov0EQ#t6e>-1#JAUrh?P=5V>xZdkhORQ)BkApyxNNVCQ5L*_BkEwiY1f!Cd+W zS*yk@^;5Qez6_Xfp5B|PGL9-^Rmsr}Z5OVx$=70j_QWbR$+|a2xTIq8LB_SIxoXGA zsOr&2^iI#qZlso42|Mx+Fprj;jK~FzTj|ipmfDKU&@T8lir!?=&G=ZEJ===}bbWuq zm1iG*{mOPV@cLpZw*YG~3KawBkPC5lXIw7(!wb993(0aQGd7lG0?jB#r%ixAQGQQb z@`w|APg^3EMUD}Fz?UI%1B=j`c;yt6ec_&rT(el7c{*jc&0XT^BERB}SaI(?5o?;I zMDsw!W=9_xDV@{Wefuz0d}hVYD)JWhvnfpxs?GB5i{=Yl4EYa{FsYLlhgS$BD$bd6 zI!f{6KWR4cS{+t?{3`11R86WD-#T3LN@psu>|~B$Dk2RH zCSp0I1VSGFIx&Y&C|GrJbreEM(0n+LTGTB*;jPTl)aqnqA*45v7X`qDrJ(aLS4%sM zeRWuf^V6}sb2Lsnt)a{vgS3?$FLRZz}0nkhk*M!j7EG0mQ8)| z+-~9Lr9?l>?qtH8bL6(R_Q9<X;G5j1KzE`24pU*Ql*=cj+W zKCC^|u1Z)qBtTfGN!7hrY!8FilliLCYh|~_enH-NMFCK!npN4g3w1%kbxvy^ky^RJ zP8)fG<9XSc382Gd(5LQaFIJkj?vqAAjV9p}NW~i(Nw?aTVH1GmBpFK9a6-sNvt(eU zWP&yFxyx$N@>io;8@(>h>wI`+1oQN9MgiG1={g&VWOX<6X!O}3hPD5GE?FOMk{kaD zyS{&ef$I{vPKobUD;SAc$<#2LdnLp35DA^7Hq~AtjZwcJpIkNHZj47PcSkyQRf&TC zO(g&i)?9gJdaalh+Vic2j9y@BiCrtHdd;E^^X`N5%G;c8X!m%=>!#iLGLbCbq`k(kRCB#T1UL4?|01?MX5lA18WIuEJ%EQXgv}YU~zk9aHJ{ zNpno5*V7t5n-(G<=xi;0Vb6_fvWVP|sQG4uXo1Q{9nv(7lNIAxl%AZaTpVm}^sS`J z!#3E{Bp=onL44dLgm8%~KEFs!M~f0X1BbT-0ZMP8vMUDr@wER*4qZ?-QKNH*iY^7dp7T6yV5c_|pHyj9>0+x??+>c9LTXxw z9A;!)vmDNgR49B;tqNVxBnSX@wU2J2^Nq;UskhBdVaDhRh7B-Gr%?C$LSG)EFGRxx3XR*d zO$K)6Jd9TBMGB^GRocGrWzG0dF5xDL4m%3Jt#95{(a}olEepak&CIOg% z!Z(I5?`N$=1#HNx^-PAkZdU>9l>hfl706uCwO=NC!QLn$LA!;|);85-zjGS*+m*w_ zf=ka>`mpl>_R0Uv0stW1%^XfZG`nP=0a#JgmXJ=qFZH-7*Js!TzyOnVj4_k;-Vxc* z=klKoPad9Gk-e{3vp>E&BrChK&qevXv7xqHRIA|4xJ>+)C z9Gyv&uMa`!Jhciym4vD??lCJyH&aoz%tuGH6rE@UU)`>~|09X*a2{X&d`=WTr@~6~ z>2%>FoRe}yCx^NG?4T9B-xQz!suFTHfyep$^i}YfdB3`hVv3PjqbsO0Q&yYCoqXBo zb?b~|+UXO|@X~3!gTk^OU4=z;guA7E^K4ejwet`#H{iF{O?z(Q^2^7gz;(-Cw(9(+ zotBn=NSx}=i-Y_{%CWOL!ydM?ZASO-DgTfd`2@BwE1CJ1WXIc!H8<>zx*y`;fiQP( z&%yyN$91U?S=H>17gNuCQ|twB_3_AJh6tWvmH4?hCk|~1SwgdmelAmOX8s7oM*Hn- zoSyF&yCi(+WBZsp;tvtd-)=$`PC@dmTq=hL0tzd1hkruJEz|^*P)`r7?8qzf+aaz6 z4-!;e3mU&r7w8)|c7J5F_C#u&cYUjyX|;6WcXvmXt%M$T0-mADi>e6=4P zicd&EpAsTg#-sE`tx<2HfPyqF2KJ;nb6rSNt6vJ~Xg6K#kBMujo$7Sl0Z(9(T>ckU zu|mu>!h0;b#Wr^v0s7bEnF)Ep${H$7hC;EH$l-)6UCR#`ve-qVp>o^2`qrg@r5bNk z$);=3>d0&%#c0T#;Uu`ae}z|>&R#@fo z*e8Z`+uG5a^^Z)SeBP2Fe|~$~CqT;-;;}J$9G-N;hSNbINoWabk5IsVn1I|C3V$dr zmw-o6y_zzgzc(a)h6xwVYtiX3*C2|eqWf;-Dw|Ho91LSR^-xq;C3FzX1~nYpKc~skN!0;$;M8G-SidgBmb3y z{je-e8ZDm%4ZOD$ECMsKk|Vg2d4+EB#%{q=C?4@ngf~JjZ;4@D?#_Cd5(KR@cHnyj z_vSHedS2|PRw8%osE9dT(w!RvcsN?RBP^@+`<-KEotUkv(Y{13Z&Z#d%jhtYF*cYi z4%8{rZixbfE8A+)bdC?=vo!&t0cqolEX@+^3@Gc4TP45sh|e2PPMV%gVQASA z5*=KPd6o$eixo;#NKpq3lRaZ2F&IrlSs3NW6MwhqXSedo-453~9n(t+QB&m5=W93B zOXgD6HdAM3-M2+wwUWDNJWw%c9ebw#N_-^AcN%nsJ%hro#bcoD_YFQ}xaq}#C4JI- z#b?bjiI~dtx4A|X>?iqC73La;kc-07@;$)^=A6GqKKCq*j-5DOBX{T3V-2v_&yJ^P zWVbHlgLoA*7xNhS6h$AsM0=06uA5mh5$8AOiT2-nPTCgUlk=KCH>U^e`H_5gIam%X z7rHu4GyahCWx%+t@Gh*I=D(X3c@lU3tbl0#Cg{efaoUg4TTFkQ%)Y}zR_2)0@ct*p zDP)kvf`Y&t#dCe=MM&Z_#BCN}@#?grh1^cN$0(dMjYRk4Li7)bX=Vzsry) zQVp>yn2&v~OmlOW`3B#5<^F zibjOXj5>C+3$p?*u#gzs`XzP^D`c&6SY4R<=zOP7e}RhC62oCO`=nH2^N`@aDaWh9 zrg*^-!Mxa(#0QOReQDU%+Ag^hOqhip60v(2m!@~lAMy&jbr!{>WJWZg4}ap`9=*%d zWwr4AHvt)IhVxLfBZmrv3xU%VLTuG+K+hC(1s5&-f?c`W_x_pD^pnmSamfL8+?%w- z0p6O`tdqu`Cu7d;hlfaR3Kf44ZU_;{J5KrGyWDly%XE~4g)o_|A9}cwlhuuAVN*DP zy{&``&hBX*S1E)J(9&cWINi+O_O=+wwKBPywY=+05MZ8C;_H_9J=3QufdJE9ryY^c zTiS~ejwN>+N;l83&G1dN-ozfFpysh$23u8K3HvYW50qI+uU48-O;2JGW^o6Jk{f%f zNsP@-jQ2>XDy+aV{Pq+H6U$c7T)OW7tdibNg?CNQAf*$YO`B^`pXoA9Bu8r!N&&Da8Bx&z>K+EI(V$FeDg8C|A85<;a-z zd!ka+5~!y8(yNjFS+ZwNMT1s~srD9Sn8nn31pmd&Urx*FV?rLUvp zf0UX=-E(I$F6>dWjg6YNNIo)Z4xa~iMjaA3!HqJ$wCcdN8`>Z^HYVnpcF8SxSQyK4smT>Da79#eTz2^iHNL}0Oo4}`&#CewJZ zb{*f*B*7_?{?*2OPWp~6s0mH8c}%D$_ZFelk4L_*^`BX44ww%Z`|6w=lVy%%m30r5 z0g84(Uc&nzH?34056M}1nJBce0P^*UDba8%TEyOOU0W$wyuc@*G%#Kw;MT(_bFyqRGQlH}yD~tUnvFM3WUm$O^Nl2@6S|sxn1oF1T4-XWddTuY2YA zkS^i$)n420k=2jbwhJVR33!E&@r?xECf}7=mHYsTHpONqw%;yr?d>0?5_Bu?EQBezcGid;8Wd)x zv06RRKC)i15gyqM>=yeAul6C%8)nWbK_7i` zKmN&|tytyx5ukVP3iJP7~DV!@4tj zji~$NDLWqVi#gAmECLpqHLH|XZu+HX9o1)!DouCdWveU2(MHb>zFfg=VM-SWPkm++ z3p0C^e2?-Qq44;YPf%J3y*4JcvB!m2Kah{sbTlkrY@>Lc{GPf^olnf}&}Y-0>f@0w zNMBviEg9Dtw|S9&AmeQyQL(Oeu`n38 zQJ!$Gz6)G+UtHn$UHlfva@Gn{$PkC_bQ#JU`DA7J_&2;~`K#w9#jxw|EjNTiMgq-k zPsBHN#)G-LYGj#Vzieqd_F~p*O%C^*ZW4IjjmK)6W`+`I6wGP_2KTQ;tk7x0mz2L+ zuT&`npPHwb6utdag!)2e%)42CC>Qml)tmZJ_bu>ZJ-HDOewL>`!%-mrm3u%`z8J{$ zey(t6`IYDMVVpBj2>%`i$`A;8}pS-w9oasUPgfvR( z3XVA8c=kig5?}YM&0)}o+xZ5`8%Cn1B3J7qulT$pSMDzWS9v)Jj(Ju-N|&e+*s|W@ zIve7hDd)BmxsM$qxW7KI4?>QTNqW~U7OT`&q$8@H`Cem`7}5H5D3iP3Kv1}5bUkWI z(@tll?v+C8pTNpiv);Cl`(q4;x83?|Mv*4-H6+)s4b8ad2XNOPc=Kle ztn3+vZ$*I}=Q^SQQds~)H@=gcF--N<6WWaB{Sm=01L(`qam@@YZ=$(b;^K1m|7f( z!bxqKXJ%_m39VYJC3-3X>0>qoRSwdOf%cuIvN^7H@K{@ohmzo&&Ay%=_(*OrQ|35? zm=;#cQH*MOr^5JE)MLFSL22+q@%KUW4%^8Tp^3s)@uT9YrGd;>TDW+kQi@A>_~+Q? zzYda_Jg%$V6s+kZbMczPyE!0c8yAZI*c+@nX}z2M*wU}TL1{2Ep^^HyI3j3c(T;jp z?fr}jU3yXn*dev9gP0tiJ`tlu1q08M?1V@ehj*lI)S~N{u-?r|&oehsxS}BG#(Vkd z>+@D@HtQt8gQkLwgneexvy2y&ny)s|FdMrB9WyRt&;<9*`6mK0N(KQNnQ%J=~4k zTdcA?PANi-Couk2s3A4AN%q!rH5rks^ryJ*$(eM?#^`scQZ#Y)`)Xf;JP^ZMgjGegm9~ac@Y{QzPq;%MGt| z1&xyu#(FvQ&W!0~?0jNK5L@(WbZgy~;_*8?va<@M1hvn@n`5;g0U*i5WFT8cyD$+T z6pz-1I(k45tG=453e3~-I($P`rn*ixSB3pPD(;Ep)L9qQ zs?xtk$w6eFksarfJerhb3CH9~Mpc9`Dy&$H7Ckkz@bbW!6@=1s1=l4O5y{dPw*9b{ z=Czs81smIbU2$wDjk!@tIY|z@OtBxYCI&jnH)P)~B3W@+B@*nnkAt_mv#^(xK{v*e zQ;L1@zfV@yrQ&G2)3_0BHnf0GU_!xS7C@c0YH3M&_p@u7fKFY^ruYsJzrO$ZW$hkA zFfr_sv*c*;FtP{t%9o-=$kIm31p3}$KAUMHKEnImOGVD?E4?RwN-Xq3maD91 zlbGYM&$=H8>+Nd8Ml?QY`nywFAHvQZU0P22R#W0=u2os;TDWyZ<~Hj&V!>En3Mi5v zm$;(&y$n~%R{Bz=Ypg5H2cew{4;vk;zEl0ib(0!znBV~`eWQ6L%J=OKXKyCQ;@~+6 zAA6@xEVH5W$ZG&i>QGsv4m;wh-4mkJt#fE$;O&iMPg{Ub#yoaL99n4Rms;%yWHMS~ z>y6OiR?FkMCj>&m8K)p)z}0#ExGfyzjQ!=jq$hD(a*>8ejFyaO7>y-NvlA z4S(Q4Z6Lb^m5l`|_i)wHxD8jwEJ6O2QJcM$5v+IgVre6X(`DwJTYvHd>R#g5nNh5n zMTvq!dU>M~P(q;fj_jsucEHM{k*`j9l!ZI=T3qv~O|XOqBU-BC?QInu7q7A$1certr? z?HI&9O-Q9VWoIaa6eX#l>{|)pi8Gf_ec@OFg~aD}*2}*lTP?!@gc}qaz+h^oZvt$` zh^$T$n^vTz`y0?;BJaBJ^NX8H^p6gwIC%QCS|&nsP_Phk-U9Vs!)wr(J6aj?oNF3# zvkyYDWyv`yG;B4Y!ysa0BrmL7 zhQ-4Ncb4fbYca?;Vec(|w+h^0Aff6*RyDXmod`#h=;)7&-_@nM=ag00UNU^$eq{(n zH#KG*h!i)t>JcIiTD--4K{YU@S0bv6)IQ03<#w#Fv-Qwx=Mti^k>R%xt|Ir5PoC+G ziQbzXx*k?UR07KUs_{w>)EQtIZ#$Wr&1GXEDk#ZK`7)}Fi8D$=SY(JY`S7hh1ogbgh*o8v-gaDC!9Y$E321{fu+l5O`f_=EKqMIr3oA*oufVf zR7>h1d6Psvg4ZjnO&a@A))AS+mAGwkVo;|3!bKf8)DwhIux6*H@K`|hF&P#fiPfl$ zPwfz=@J- zlXnLjb;DkR79!*QTlV*3Zob7a=Ih`59>|gZI#RZA;|V#a*qRK|vWOSqsurl^Oaghc zVZT?HjeUK^vZx%_-25FIQ2BEeV#wo%>NB&geYA$U8#=-$wUo{VJC))2v|08g5;mQz zAEj|o-d94*OMP5mP^D2*Cg2HGMRy94%Xt_ zwt}M*NR`+#LWpQVlxmRXdGMXO+prhk^~?(O0QvfIN2;hDV%X#M7O9VVVU-LNqi_hK z;6Vj_hLAL}oh%xau3P)jI_7**ol2JrYD3^4ro@_pUONw!MpW$(r=Y^2T=V z^!4ETTrGLOc`OjCkUYd(m(pR!#0!;%fhedb#?%tZxWsZ_xw~Q|9X-1_^?L0+*2%0; zyq7iX*!J}@bGfjENQwK#xZO;ez~a)1?uq$fy5N41pnK5!vu%a=^S3{KR^=FDtwE&i zh2zpkV-ZtTRF&>vpO_Y6Yz}=Xz$RZQG^MGLYyW)r#he(y_{!1v4-cx);=9LM%`-Nh zpE3`cJoab3iIs6>v7w3H@x)OoIn_#tY~u1ElSt3x&B6He$jebVoW?yk}Q5yvP`~Wdg4oQoF)~L`UCsrEd-$f zk#Ih^okf=rZ}*~h70^|+QbrJ0IGE!#@W&Gk~@_AJw@JC9^`DA zdyBOpY3yD)H%GP{**|oCvs#PBKpn(}v(=dQEn?EH zc1s0CqmebcTZfW6027hwfx*kCoOKmV$?gfIgH7f@T_IXsQ^mx(#dB7EIt2YtvOEQh zh1Pxagz!eR>|u@!=sty8@k%2&va^V>)*OH4tWfAcqWUDZ%b2MjkHZFEjR z1*!%?n)o*yURAKXMf)kfv`A}A_8ok`nouaDxsMhr>_E|0d?K5Us!)>Anmds}yQSL5}M`@exx zx6_91Mhqj9S^1y*q3`Iup%pyB%J}$9S?mjLWTNlYSTj*H zZ>axkO$D~r|Mk^5ag99*l~HpCYy~ru8vEZr;mw5$ zY|Qr=C;#Ux;7GG1OM0Lz3;>tVr1f@-*Xba$wWdHP``a&x5fEn{-VHYS?=xI}d#4mX zlv)%6NWo&MG1PoZT}Bh1HaNMHfbgwzXKavX71!Rao8WJgsZXGWH}Phxo3S9*-{;o9 zO|Zl@elId4{I~k)e>s8Z2Z3L`@k6ivHv7`j;<@O)eW81u$>(4G;lF)^mdkDgff8Y8 z*k2#Vp##ThpInW)oXG$F$NupvpE{Ev?e7&LfZu_L- z4u+N63y=6Fix0C}8+_l89G0E_-&g-1ugB&0q^e$of=KLFNAaP7FBk}@#6G>vvP+Yu zcU8XC=0Q#tCrGRG&7F&utIH#eZ0rm{dgIR(U9@)thdhCgvX7*&UZLqww#=6|%J>_i z+d|XP`Ga^P+V?x1+ny0jSP00&Q&%)&c|M^e<*I0U&fNi=*Oy;A%58S`} z)SNSpM1?1Fw&Y)2mdi`ktObrEWCw`;>bimN?bKhPgEU1A(*3*h7lPvk*`|SiJuo+q z^}LB;vodSM|L*+YPJxxER!IKkz~SJXK6w!|y8Yb&6rcX(f4dxkV>WPHi*@qfJuja3 zX{AukszN|n{L3rwi5wg+B-{Mgo96F7-$V$W1Yy=k^#A4-KwLg48o^F~`Qd22y=Sl8 z9KfVR6J2hc|LOx6Tpnk8CHz;%E{A6u4DXqu?0<&;Kg0hQSK$BqPyd`39UUE_D*IpF zUoW|A_xF0!`ES1VZd@6FY(Y2&ujF2fokct5E?h&9cu)EO4^Whiq&c3lPxDU8+nH5uOoi^QEq;nLa z=o|E`59`s?lU{Sxtxb=VzA>o8%A3KBdmezDrnOb~@<^c+R5N*{K1 zX)4rmy&UHkmU2p(Nc8&*iKMgd`!t=`)!Fp*iV9bWlkk|-g=&|txGYCLc-mNjE=Mi* zP|k9l%6b|9;e5-WvF@M`F(0FWdKuJy$|va4%A;Zjd#2-`Wsmu-XG)U{?@@=Hm=6A7 zZJm0R&4ue_fJHUWOiN1;rJS6p1Ce47+8zBwa@lksGthP8#2)IrJ>_)V@ALV~w?Exc zT&}_DhJ5>~Ygk*TV%lZUqKlqc!Te^X*dCZKULyn_(|d#5SCG^22jtm}mUQ+Cg7gZ@ zAD{6-FjWQo+cR9W(hC}rXigo?GiY~wB=P2%JyAt!?Us+zD|#!j$I{^&{TY4%mr>*R z1FtEBuhJ3!#&r!I6>!8@y7-+tI(dXXFzm42phNL*n$Eh82WwDTh7tUZVx;CRk_#&+ z_Tm#TfUi8+>UUg?%*wSiK6SCXeqcSFsm-ZFYBCQOoaM6~(5YV^E|Lo3lh3R#P?vH% z#rkzbP;H2?Db$5d-)PXPka`Hsj#yfY5`QpSQcrQY&2kSqO7l8j8^~w_r4O}5i8d#& zXg(~65q3SzL*FGNV|{uuRMJ>&XcLe7&ET}BA!FOgQ4+}4256ktw9d#5*Jz2tAtta$ zttf#%Yg^0@3_W8}LP+#>?sPiyCJW`COO(Nz(uhi`!*29SMJT@G7D@e83{80ZA%uq- z_aQ%P*s;lU0RXqvB74D2W4-`KP#w1w&lW%wF9-E^=lwoI$xtAQ5uB0F)Jp=v#4Z(26eD=()YVjFNVGhZ@)~Z z%}ZnCaml-O$#zC-F8K>C2e}+J-Sl9J%M9c5!K+q1)SNa_?yk7}3 z>z1WCqnoiy;f?`H!x7^15r}^%)-K61o2|Fl54WOh4_?9~HpF)Fc@)OIFPWZwv@bES z;A(w%ij0gpOmyH4yWdTH_Wp!~ZClv3=Jhm@PSxg`_=ko4Y(jAzjx@7(vn2*iPmfx) zZJeo1ELfy;-}Qnk)%POx)W6d5=nHe&Xy#-DbGV1ii#v`pl?rfwajVnWi}RJm7hdPC z*^(h+9+sb0n^O^blL2EjUGJ-u=o0agqwSfQ;W6v}Za5CFGnqiC&6qcJu}wRR z01~=7UQVt*$OyP1;^Gu&#FK_~;c9DoSEtWL|s*u6rOnp)CxB(b*r6rgWYzN zH2OZ^g$mxco*+N(&Nk~u6T_Rf)2%P?a|!^!WOD+}9N!yn|BCj_HDHgm=xuO~(t?X8 ztXY$f7R|cNeYJS_G7F>O?%>%iQqH)8J0ISeO&xfc0!`lgfT7iqEoIR4u5+8__Iw;&*plI-{;7i#c09M$;PIuRJnqw|m?IzE6QEXFRjh70oJth=HMA1u%KG)jo$n+}$r z?UNULo(PcL#+5^8zoT$Kq4jI;8}KcGWWmOPoS`qI(&S%92qriw$3PmCZ2fXysv*TK z8vc==(8_#08|O`AwS{mDf05p%C|$eW?OY=CCc)X$5{)LA_HL!b0oOZkaY!EFZDbq1 z2eXdXHPA$49@`K2=3PIbJuaAWIAFl%6hVWgXJz``i3wS7vcvi`K?q$Cqc3W>cB{p) z(MB8zvhnn_O*J|qPB*zN>2`<-EXT>}0CKjwIQgnG7UMed+WyVrb&ct88017y4eq%} zq&8k<)Wq5VIlBg3q*5||nk0T+F{)icju0f&;NP8io*#v3dEeU@y{-LjK+7QBCK~N- zurO>1X~H1GbU%N2Hg!uf@z}tHLV=d(Xo@l^JjFz)mEw*Z-4hA7sCVd zTJFqpCG#Fil2CF(nv5I@cHwn{T}}F z&9oU@HE$Gk{1_vN$su;TX{=C7=_)I_f?DlzP$9z0wXmQtriizUdHOxH%{=QnKv zvQs&CX3~i|luoy5anX~9sAU%9{?Dh4?YFi+9vtW6?dr|q!%K`hl#)J%=cbj$u<6#g zOjwS>H0=I-Rg{!gG!}#nt)zB|b7A#6%<-+(%C~45E%d&lvm>qj*t|@~s){SF64Z{` zOzy21XAovjCnod&47Hbld0wZLF!~%@(j=eb&#fcG4v>3$)^U6}IUUV+A)BZvxd~?< z3P(zE*Q#zn0CY3C)xX4jCKa*_p~}Bb7z=Cqbg~fB6*q;s;QaG7|9A8f5uu(3O|L_Z z-th>U*q21zOxhygeC6$^MqwfCXIrJ zPlKSP9|%55V3nk(&;OC#rIlEqOd2LVZZCWp^l2&VA#yK!d<0Bqd7?-}t zY`*bh-hSEf>pc(!s!fZJ^S@~tprfa3Z9c$((9q3b>OAr1dMa*6Qv#F?zVgcm4Hl3F zUK=?Co!(Cmri*iO-h3JC)k|FhMTWJ|I~$H?QI92KnR65$RKA};3s+goCyFeA?pYDE zgmHGy;YM0w9?%E==S|xWe1(8C1@c8-MBD_af{OD${DIV2<5Gq)GC#XF4e$eb$y7$- z8@RyeWG?6&03QbjT8C2!1_^k9AoVqNZZpRj3oOWg5 z)n=_)oubYy&74y_9W8q882hdytcIpL6uOdKfn1!8Z1p3~Z2>(1!e)^<=rE3r~g^J+2O4aT0co zd7aK~@Pm>lX*=_iK%$K^>y|k2PffoV@|C;v{a%a4?oeyp^cS3JIdku3jfX}jfQT$? z^rFjv7Q9R2Q!S{(kTSe6U3a~M2h=7XTk`m;KsyV20YaIX6%wPl3=m*NYAOiRccONi zaU002KC>U)%KW?H;=O&jfv_0_>@u7CJscOz7RuKxf{2VcUh??2oTsYO1sX*ARe@K*p-dB33vAFkxY-v+tw&F5@x5 zCZOf_5ADBoWF5=RVGdDD*R1yb{FV-*SSqnzyMcp;tNjjj4t0Udj*PNwXNXX!=-*id5EyfA^}!>KIac?k*T_a_W_* zKU{6jc>FNUZ60M48#oa!D^3tEP_$X9(=;z$G|dAR>xVP>Xj-+|CUX}O$?)b;Akbaggz+ZUk5s5mn`DpOIUQu^oQ z%FPujFxfb+T9RIbdRF)o<@hu9N4DX*CkLzCaIG$gfoJj8*wKKAak=cc`J=6mjBTTG zcQ0%-t+xieq@t=P?_l3ecG%xXtrwo7xWtG2)~tEwIcY)j+V!8t??^xk`w4DvHyi8s zhgt?NML~-IAbdJ)&%mT^U|vBEFKxPa#CP~L{}?=ceW@c-Q~Tgty3rZA0aB`F(V8*(PH&p3d+Z3em@M5E{vT&y>>$jqfwwS}i5uV%K%d6+dDk-Ny}3A08{-c+ zpm@%5AZY{tE&O+N3qi;tP^i0NUIM)u%c8O{*~G;O<4plD*Vf^a8D>s!vYl(0(h;#t zj={bsAM@t;kQ5`G35g0xdH`^`RB6%IB9U!gf}Dfgj^mLYU7GcX7K*WW>Vwt_eK3GT zXrMPW4;A37G<}CNsf~BZ-dwh%W*ou1i>B2xG|zHO)Lic*gQne&UlL0i$BJedY+C3$ zlZi1avrW&e%_BEqvX6K@I%^EH>h4Li8*CcklTNgQ%|5q>e$$qauCzcP_fl<|ha~}6 zJfWxL6r}KCIrY`i+0RVQiZY8^+ZEgyUV}EyJllp;!d&u7^IGUim19$J!y;Wy%cH=d zD3t1V`7`@;VHb_H355DnX$*tIieqSI9$>vMx99OscS&=+)BR$JLsOJw_CGWe}wfU z+o)#m5XI915iyHDDSx$mK(Gs*mr$h%j!ss7nQ?t<=dmaol+hU0eJ4DPzl>pXtv zaV)-p8iMd}@@(y8u$5o@&Jc#6CW8J?v;H-H59D+yEUvXP5^dEol_gwjCnC0w{BDr( zBle5VdtEh);E$t5T{1^DTCE!sCr$6q7RpsExF+CNPD4z%27=|LRMQ?W-YK75c3}KWrz#dk+)n0djg=ni4DKG51MZi7@(ygaCQ?!I;OCw z($&XJB4&^5dBvo|PfMW$|F{m~c+H1Hipjj@`*zL1CFL%oOd$f~0-JZo%clxeT2k8p zy8b5lK=tQ2EA&p$uqFT126uv+A{BgNe8BOISejV%38`U=`=^NhyL$wK zJ5zKiP77UI!dncl@A=ZGTybvMBmBPx{qX~5mzwW$~X z@VNm@zZRs7dkSRU28)N`O|iL5jgZ)fCz}&Q65A8UZa$Ad|HK2XBHd##tDj)kV)+w5 z(#rE2rW>*ZJ8O-iV_rWL6S~9Geizw@e?HZNCmwB;fvr6%ZKq5%_>f&ksGjY&{}txO z6JOJ*(45g$#J0HUoCZ8vM>T*7ka~O$fdBjkmFyEeJL+`VO@a9HGEA*%%_8&zWMppyv62=V)b9XRG@)0r5d+}7lSP^=SoR5boVw%L zcAH7h>+Jx~nQiA!5dKdQ@b?Y8CywXExP$IH!b7!%b#EesDoyRDn&4hge-Ve}cQB(I zsi0=nX+9rO7s~35d|84RRF8L>3X8*+JoUDCe>Pgw`#dcZa z0fngB;%W3;nqZj$?5Acjf#Gk5PuTBdG3DaY_6x$x zgT)Wih|UGYp+eTYFq1L*e$7W(rDi`(^0eklwL=}Y=d`li;`F^o3 z+ay#HesJ6QPCEQv&rFMS!7igax+VUkq}jAQ6{`|gC;cUE)a@T>ea4Sre()L*r_C-m zZiFGDY~KKQUn6znftDGuuh@27KiXiLbVdS`9wjjo6xesWIJV5J@Yqk)Y)#I#()=2G z|6j|!6bqHSV~vQSF!X`h3YQkzet=i~dl3ZA?%8OdOZBrR`@hD!{1Gkdi26 z2BCI7x!Qd~t_a3 zQ0aq?@N22ZGG5(lu8G=n z+^Bm@4Qc+P$Dj`h1t$7-RJbt88Q2)nV@^)(lKi z$%RW?8u6|33=80}volY8ctMO~2pHZ4m@arxtIfKcJ7wR!frw#Jpl^y{?X+MYetqFI zWq5*vA)XoPGD|;z-&qs!zk#b~mF}$^x&ldv;j_%{Ko_xSS_=lg(?fp6u{=(NNC*mVp!0OOe zV59DP^AEbe7A@a`Mce){#IKL`ua9TH1jorBH{Q$sB}NeR>K*MVd zfj|G6L%9DL$V#^lMptM4=J>~J*=Y~!mFHgzemE9jU;oATNZ8}AkAE-;x(a#s?)e_s zzg>VJY}tjCp%R~ zfB3Ecclci0m6Vi}uf5{G5om9XufNOB9Ernk`1BXt=@?!ClYk2dT+w9q^lhkiP2yXO zf9!SsXNCN)pNJ_l75a3xSwey0UskSBf` zh9@W$n&?|jv$yuE9c$uRyA0+1r&a9whr}7m1 z(T)KUSLT4}DezbU#w#hziiserwmp(ovRF-ZqDRku^3^$!!-}%1h?}O2?V4`!A}QvVF)@8U$;^f=CR6*u22J)EhzWhX6G~K z4cGOIt~73knNtM4gFc;zwVUw1K#L<)NeE`aLvz%J$7jJyfbe% z-S7gwHt2`f3Wh&ldtU9hSg2Hs6{@%#vkLtt0u{^GcwT*jjVi+&u(nJ9|@X!)zVaKXiIlVz-!@i*u{ zV~#Ap;sraagg={=dy|E!n5NjFZSU_;@2~bdE>(jOmk9d#+UCWPBF%Ed`9|Q)TW^CI z3j6ng)!s5PiOA+{rm&a)T=0s9OE%60OXiwQD`HwThJ$NeFq6C06dw$ z;=fZ+^ZKCgZ9l@|`oy`q5;i^?hu5A^gfs$OUNa408AU_iEz`cirYpd3%K@=L`spAI zX+|yo^#mJIRl>orIy)jLUp`6Q$Oz$T-BaODYar0;y0#pG`S8YK+S3@R{UnFeAa zU~D!v`vEIIFlA3#xIG4&2K&?CFs8M>qiM%hFjc)9bPN0KPEYXeyE88$gl!)s;{r+H zbm6CX;*rw8{= zAG?&89n4{>qzg^KH91_ip!OoLX%CyhkIEbZ^uEBIt=O>2>at+?;T)J&mPo!pel)ky z$&?Gr5TJe8?A_UIADR(R`O&7%?xl#=mQT7E=<@rW#}koJ9nYBC#y`$dgbw6YndbLh zzCk0fkYn722p-#G8k>Wkg;~^NmnPJj4IDRH4Z-NSP{sHWsZs9pb&pWLg)^4v;WCcA zB*=R{qdmGnfKT80o&A^@V4)Z{vD!go>NeC${W0!)0>U8RQ~r6v5V%Q#RvsmXj+F^p zyFHlHvm^kIW644rPE)n0&#@gCtmDus%p@Ov=CS|G>+3}`_OkhEZm`k9?-){_;}E%@ zz`(Z2)ry2s#x+5cNmbOe`#$L_w9OZyxrjfRnSr~t)>^GLQGJfzKavmE==ui}$+!zH zd*K@eiMFit5pCOVeQ>b(-Fk&qO?dhoo4x#Px4seFyH`hdcp4IL$@!^5rW?7E*uf>t$&Dx;FZ_BLZsNhhEjrg5-k^ zU9kb5*dTJGbiUymSt9JEEgpYK>%v%r#nC|cQ-k`@O5&{fXd)oY6iN5&8`qmoHX8)N zsHLy=#%l65yY3z5k<~cXd3D-30249T3Fp>S-4U27)GlB3YmeOSwta29k>$O8&d$O8 z#RGD6!gh%e6Dbylcurf4Hz23jki1AjeHI$3S$Au-{Iu#@>o3`qSK9BIjCMS-0bpVx9_B=w*xkB$dyHv91*78mqFr+P?ArR=X{U@3n6= zqPRIUSPNVzmVJmDybkhl6WCR#1RX@{>3Gfjzm5Pm`|4}6nPY=5I-|{M?X?$^XY87f zohQosJz zi)Q&K^ORWrDU~5?_q@L@{Kg$ zwp?bErPq?}sX}n2hELXVgm*x;+E0Y0N|Ew>Es(kYjG_>5l)%XLFen=Ffr@1DoQktN z)F3e9BouY-?2hw}E&7T4SCbHza>EvRB5a@KWw3O=5eqe{Oz^;t5L7+%Zq!(ROjMKb z0)Kgk9=^`W`SLJlk~3P+HLNU_T8LT*3GDlLpk6NX_X0&cE=|&8(rOeB?B<)k1imhN z$yNRV=i)$ht!nKQm&Ehfv`V$`@>k___OS!kw*I(Q(QwnYLW*c!(IlMDRjgEi^r6zr zBK_VDE?}tk6|EUi|E;#@JHGXhh0_?vrc}9ylmwT5ozmJ{TgzMB9%R3fZsV*Xbq)9W zI2bxqC442WJ|*jG`2hAy*k!dJlIsb(t*ksm!2f&P$wQQz>UqL-NfT>n2ud29peXTk zE8y~<()~{=<$jq$g~}qJw5{e5Ey5@YW2dV!lZs+dyH)#~-$c`4Z^S*=+WJvOZwG0D zu$5jN{(e+T)tnPv!`(b%XpPBhA%{kmLlfpD3-*9vO$7DKvAN59|8|rM#Zy;Whp8lP zgAo)6wQxw@uiS(*0-U%HimD?vhDHV}Rh;JFqD9B&ymu2@UgNP+JOhZT6yer%i_0`v z=dxe8FyF5$j}AsDpDl(Q)D2ZT{k#y?Ouqm6L4OM<+cc@(xdf!v!gH&^_x7u zghmx8c@r@jH}0=0me+~9O_68Wj#n8YrYSm_+zP|-xSJ32i!Pj8%NyR9Oy35$b^7JH zl>N1Ppi9S-x^yW_tn|KQSuxg+=j}a3ePs?a4VPeP%wM@^lt_iobSH5eJ>lP545=g& zw5}qXLUrotR!x0q1~4B_gNG$`hH~5^(`p@By0E)^5BzQ)RDzF|EYa0TZE?fiVnVrA zt^Mj8|Av5uUd2Dm1yP?HJ`hp#P;mV2Pb0QpfG}EI`vJcAfAA%4kd91EOkdIl-?M#= z^5!b%Wd;8J6n#S%kOE;va4a);SeDT@m7fj!Nlc0U+~+ zZJlVscqvh@$huD_2r_}Gm$$^StVNgV4Yf|Axwt#o#ZiXs4=K2f)SrG^{91=-2?7Up}dn16tmn+SbNN`63=e zz49Ac>V+W)Uv+$cd~y!!2Z2Cz<B+|QezGjOCK-`hk4%sYJh%JK7Xky`Zw zcd_%RQ=9daEn#Qdq_V-#oX})}yY!KY_kLi^U1Hz;abV#9wq%l+ShorOsPI0rapMJo z*7LBhr@ZQzs8uISUos(J_h6-}XgogYlj-iyds)V@^u3(rRy=Up0oy7M52H3n*4F4Z(I` z66Z((7w_(kXXuSBXXywdRH0aypGoSZ*|RYmMp%lh&>59-CMM> zhABteP^*3T(55N8KjJtwbodxxa#$I@%|C{V$k7+p>mJ=aHXBPUo;^t7J$n4vlAiYT z-oAN-;NMoISNER#)a~W1Ob#uRVW$eYT=pACJv}Bc+MF#`@6}|9I&&7XpTT_KKdUGP zd41M_&fwC&9Nc;#uwEe}AH(Z5|I@VB=@UgSU&sfdRe=D3TMD4ZEonsfF!%;W3{&QU zUAb`?Ii|ORY8VM~85&iUVKjfufxH-c_s7GRSn-!HRc{->-C;MJS9(A~@JFzZz~;EU z8qp^G2z;WX{ey(XFfYmfxR@i;yx3ZLeOv??%#pPnEY;a9>#bFMHWWthx94~ua^V4r z*${-xIt2-UOhz+HtEwQq<#K*^HqECodk%rwNzPm;_t~$ zY0~den_~sDG^n2iFcAV?Xxg;k&h!#j{VuZq)GlVypyZjXPkeRs`a0fJsXAWm zlMTh18PZj|A;_h>w@hmwxc=oen`S{5q@r-_MY_1#%whBsy@sD|{BntlZ-$jr$rYpT zXSoc3osq1bgLWKi42S7^dI)!`zU8&x{Sy>u=m~o@H}Wr=JG1Qe=L914jl8L;Vh29O zBkNiX<=?zA2z++y!PAv{agH2 zXRC-)b5ZK!V%fOy5;*@rzRte4j>eKl(0BP*a&B#%0K0zZ82ldhW&#IysyaeDE62to< zwxG&#&rZ6R7VGdsg;%v@I(R#PslxYgFN3NDZcAmp(AA!_enp8%mtVX-t;r`ZuE7zm z)+<5sbv|PsC%Q$>D5{|oAo~>;sn@Ib$kMs(Ou1-fci zx)aY4kkx#KFCr@%u(Q?SR-&E5(ww|z|2@>@SXFr78VsBwe=Yv!rY6L(MO1buE=B1> z9dPWq=1=V;?_k`eP)jl8g7TxoX@skkp#pbzPPwYbkL-UAb!g!%s&`XTjscVl>G^CkMZraR{Niw{S5GtFw_Q(D zLaP^S(#IT0&XxG=Fa5U5PtHTd)gGw^RWPFZ{8@eQo?kRkJTmTnXT=ghDPYmP+IPW} z^ZuBhajbB7fb4;I>MENWsnGRCdsY82>9tM~lu7)DnD-%QCH53;?CrdMl%<>@j;3~O zzDq`v%7j5wdqZNE=xBNQ7cq`OrtGSqMen5yYHnS8acH@Oza*(Yh1u_iBg;7_Ll)5)}ui4=(dyLWQYE{_mcznsivdNaB=mjUT8fqQ$7A6 z9`^3R^JQ+C$p!Ge7%!L=GN`}?dDjnepCBa78o`>RHc7SM}{ z`1oG^axHIIVvFq75}lNa239eeT9lS2Gc-Iep>Pd7F|iF5JiT)+r#%YmzwejE&*53J zZcY^X-1^L3#E+Fhf6}+U%XX=t*FQ_CvAW>x#qA?Jd@;skTA#z(x9k}0>wJB(Y}mLMl7ZW|H?1Iq7(k$LoE z<>ct6&;+aq;{tDNG})N5vR!xS{yA_GH1WnaGhoQ6?_93S@p72O-8({4Mz zJY0C{BDX5=HE5%3r{avy5=I8^E*|VGP<+^7`27RJ^41u#GJR*POWJ;8F9Y5zU4K^~ z*@jQIZ)+MDI~p{-s}nTcab|~wShXmO-NA1yB>y1DfQ;Fm<+)Old*@-jIdZ&rPc!+J zcdNwl@Ek6AbY+XbDd#tzek=zwGJ7cn_J9K(q*hj0$jlN~o#*A2Dt7yo>&W-r)v>a1 zxd1K7@`90k@#h?S$*|iYr}ea zeM*PRNC*_junKVoqCph5F^daygHL`F>j5L@)n>ZyxY)wP7o9scYX&1j&TjcVsgMS% zH(V_8^~vsJxY%TBC)#P0bl8pGtK1YK8HW!zc zf%~EJIg9@i2n{Uicbu;skJrQB{%)D)G*&t$Z;|22n=Yi9V*y)%&1`6nx%1AX-+GU( zdQGvX*W#`*92GFDh){Dt?=XU>sbN8`6Un7}yFHLbZTgG^jHvlk1o;pla~rFEXcLWV zhwn`px@1R%2ufZj$Qz2>#2Y8ZHeCe{aKn?4BURW~<4zH?DjesJyA={2bxIf>uBZ$Y z5hr(MR_JCuFy0X;3AJvQ%O$^w%CQ>N``$74NtB-Mp0_}Wq5u&QH|n7Oq3a@Vs=)q5 zBYsQ_FJn-s?jVgY-N%WCuJJLxdO;gK>kFUP3_?D6!|%C{_xsxXa$4my4IqQ@@+>#a zHWa9)nwe=~GRq)tVC!Yp5XK^T5uVJ{@K7SkrHKT}@$!eJKV3nrXHE7R+*PvCS-UiAGvc_C*G zju-jfwE7$HBo%-sFuuPBPc*K<6n4msl)nZ`ZdwAcWb^+5o@m|ke89xSlmM4J_}@}h zE^QEM@TPS^_2I~qpiEYx;f?GE;E%j7{hlPm*@hrIdz4k~f8COQ{*#R>x@<FK;?m z3{ykx{*xzK`b*-f&zuXaKBBtW9#u<&5&v{)<3zJvvT?D=Fxsk{WpHWZdrw0GwHNFH z&BJt-o_E`2(odwpt&Z|Mw;NZ&R0baC&8yJgL9F=aOxOz;q=?BmVD9 zQ&u@J?4lXW_e&WAb{CB-UdKF>eZ^4sx!+*y`c&v3tDuV3Zjt4gmoHAtB+NJH1KV8HRhS83E!3b#`52ROBM5ZA$@j zj1L~v*-buR(E%gOz~Yu#Sh{h!&VCxmRIOdDt_}_kAQYHWt1!AZNam?K>e&9w+zrG+ zgHbpze^?57zj{?>l|`Y-n5l*fh7%z^m6gYL4XZ5J)>8WYDJM|?WJtxum2kh%dg+-P z3b#G0VxGk#ms=}@dt4C`JoVvR7=alKhP2FZXdiZ^amY%AIL=>g4h*(sE)cV+4W5L6 z_-u(Yrloq4Aq9PZyR8<@XDof`gb4qmk8s_X?0#|=u3xWnKicYYcYc4)qCeC-s|Z;s zZ*{$EH``nwH&0PjPZL+giWn5PGPVNqNu#U)3eP9O!r7~iD&j2nFiYonKC4mdAnfVGWM zq3D^l6sVa}8kAf~ZenW4y#iCNXw*qkhyUf7rBn^uI~6`kUb(egU`PBw9Ylvp)vTi; zC+Zy`Afpycab-#LVqoDZr9G_$nP2BhL+jdY@%<=(O-$80az6*ijsQt^!Ur3Gs~_f) zWRE&3JROXbw>ufe=0*p-_;fE#+|L{2a~~z=-gR;gvT>Wk$8WWw&$2lB2-Usn_fO_y zxVsG}n60#JMZBE$pi=!~csIFOb)AnHy&2iE{gh~|9?D^Pe6wYmXlt55^Wk)1Z+0Fv ziiplp)3fJCRFj8;y(Sk_o?OWH)oww?;ZsT{t$+pSZq25c(=Scc)lth1Wz}<&Uqe_D z9liBk!cON39pPhgeed~^9RW{Izn(e4lp<4weeTdqHrcaq$tsGf%-1N$edhpP^7FS( zcI;djraZQ1D}CsQ$hHgFz z@$1(4Ly z(m~*UTyssI`tO06a@|MB-yc6wa3hY!3`{%1IKjVau_Nf)E*WMST1&)0- zE!=H0k>DT+2UaYL&{gx;1zC^Ff)iy%@S1aZ;Pl3ZflZk9`_Zfsh-#us*nJOldVcpKyx{0gAnvQq=Xpt#ThBTKvK!tHm|L7nWJuO% zu)M0&FwT}Z|C$@edr5N2Td{wU*o*-{{hHw2YvwK}XB#Eiep+C|N)I-MvOiL&(PbIdSHXW;TLXT@5Ub@U@ zz)9^RF$g(8h4z~J4lT&!AS3?}78XM+|4~5R6vymGg~RMq2m})LCNO3P3y6z35phQc z2XO`M=1swXL;LaBk9>m)q_qNYWvC^XFXNSs-`sM!>zTLYFXl4#`k4$O<8%Dixu$#; zJuh4RTWGXFHau{-g@B~ssajjmWY9cnKAH0osJ!8a29LNvPxuiqFpCNwi^Rik9cU~#+KhW0 z-1Ob>ToVr>3Wu#zmn$SQBL!8Vm69BsagUo%zPze!bRLX9D5%5Sg3x(7y`aokj1~5t zA>d*yrIq^H(AJhtU-*^YPxPYx8v{}RUFjB}HcC~?thrdnSD;)28HK{;uUx;@cm^@r>j9tTz?D>tLc;e`V9{^|5%x0dFTnK6#Ezw;TUIzFe5=TKU^_k>X6ZU)az^(^? zgYbri!s^5Zm?%rx5WwABZsW0tc8;K%A+7#K$%0k`l794=Mi{}gC}2X~5lPD-+w~LV z#9hbV=7Kvc^r|gzW3)t3f5$#KQ>P>m6GP?E0GRYXnSK(1B6Zpik3Gx1!gds>e78;F z&1KZM+?I7?5;gS6Jh3n1vjl`r-m))M{y~LU)610yFKgYW9pSL~jR@!y#Zg8VDmel{ zTOB0dATQ}xsjI!b^Z=D1KWya&a@w43PQ6NXW&~8U^qClPRo8g#g;>Bv#*uuavO$1u zax;teecnaGAM3A#6-->{57!DXiuy%0Uv0%FlPJes^ls`*mRBpCeN;9;P@l*WwtAn% z(imiEBwT(g9XwOs)G3^j2~WFFfHkpit2eAobd3cr?9e8%AxAg&P{)qS7vnk$q3dy4 zI-xs$O(&G47EuZETo08FbiWmLp81(bSg>o-q!K=5NRU&iGd)3U^mQfk)z3Guu|H|` z`?SoVaWqxqo~J-?wpG$}Cq;bkJ84xi^C!h5<6GuBnDWS1bKkI!3wfIhguy1rdM9@9 zH74<6h~&egI6bR8keQKZ zp0Cdp_uQLAZmb^#K9#DeLOj4@2zL)jcXg#mcyi#U{5!9=_H)GLLO- zZ7nD$K!jyUhju$w04j2}#YZb%z@aj;aIBk6#8Xx5^av2m*dOG6G*>;;|k*twe4x-FZ=Fz(}vQ- znv!4G-aS7&c~t#bP^b3dQODTTCpG9?yJFGzbR!Whskyi!M}2f(s*rQrxrAeycnyKW zs4LuETOaS#fA?TNN8H~GE;5?(d9(YKy=7a_mdv_iV zXcA2;cfNjl(bcvZsJ;%}jGH`-?Y$)&BSyO|c#o9l%Utr8Z?W>VbvGh4jD($Eb|`E@ zLNcoTQ;sh$q2luoXSZMi*!_kE@B^pzZf$$_G{QR3WwyBKriPaivs}}zYv{0yp3uzO z(;LWn^a)BAy6bjmE+^7(BLG+E6f4^Y=YI4SMw*C6;+JRpNvlK>7x!<)#I}f0U4H_B zb}C|HH*!7=3 z_aG6upbI7*#Em!y4ml-NvbURTmb8~Rta?8+`pM>jN?N(_<(jY})6OM2Kr5&q!=F{9 zfH%W;XrH*Ixi3WTn|o|xGI33NC*4-h{o(tr=T5l_f`tak%$};nZNpyaY>>*T(o-fy z(NXJx8iga>yxX`xl{22`a4~7ajlboNHQILGOYYH%j7_nB|571eRC9Mn{UxM8R83-W zDV=>aL3h&{9-HH+zucF4x|)I7ICNR=qaAj;=$LYozYb}wNX%XXE4T~uawam z(5KSV%yHCK*3j=AwMcChd3t1_y|J}AD3bM?Z4SX!Q;GC5U@^hSKwc9_(&*kF1KqE< zyzK9oHHC&ISd$rjSZ2L>fVD1)^F<$p(upXqTJ9(nB$Z}(gl!?SU3krdIOSP}X|B^< ztjz5$=URN0=lOASjT@br0p1bI7Dp0JWlq~!aC*mGNjZlzYXSt=HlypdsboPQjyh3} z&@l6|;a$16ZOa_v=9G_wC1o1@hgx%D7!~hvH(4WhHF(#w+9EDnaGt%zSwcumj z*=EN7Kr8OdD{zeHJ)MdBlC?3hsa}7WZ~-{5zGiY>Dv#NYH1+~b)IB#_-7te#nqZ7z z3Fts83Pqx`IhCJWN}}`Q+vWg5VpMaLmkWqHiD!7Rhv(j<8SL``r>pMoWq{pAt;wFK zp6*6Hj9UBb-flA_?>)0jX^YY>RHvGc#f-H?SKYC3F^I^bvVxMC*16+p`Hx;|^^OD} zVN}?n5@FOjvh!8GcF0qZHar{}M(2OpHkXI*A8*3zXJ0gAPRgZvvy;qohzI_n{93gm zBgP>l+yTarPnh~cZ8%^BlJFb579Xa=u2mRFfFV!9s+QRXf{~cS z!J1CxdR>?;*#H-+R=_keX(ylM1R<|2;JRB|*QePu6Xc@dP(hAF<5UPv4!$60@pT9j-zQdYNA} ztrgH~`RZg_@U52r&~4+FcS{%Y7g|_*Cj30RB@a0TRa!A?+vAoY99pd2H}qXiM}r42 z+p%v3v2N?We~53+DXn8JR^NVLvB_rk$JR+47bK;dW8<7ZLGBO@s^?%I54~EmtdIQ) zXa3T%EZ(cBB0f!u(1*qQF%YG$M6MW1E0XQoSQ?7pX}9SbeuRMo#4^t8Y#Hocatvy$82ZDJMaZ~yfp64l{lR?xWI#e z6Y~svP$l0fH0|AnqA!I;=jzhD3lGqCa`W_-{yAeuAH#*lONK?Js*LvhZ!BpyQIa3o zJ*&q{DEpS#?MM(IIWc3` zMsp&1@r>MkbANfng_dAky6BSm?u-`TYsX|%Li*yJiUs0Id)PV5+2c7zSEX&^>*iTV zof~x$V@d3cs{Lh)R$;(L-kUu=*@k+ggsvjsEI(l$LO8p(yo%O#QPw}gi)yecaNJUm zdo|pYU^9}7IljU<+h8|LKfKC0S;gY?yAap;!}Ecstp_q%@q>f$^^TpKkSDr?@vAdP zT~D?w8SP7s+KsZACdZCu@>W_x%E!Lp-CL7cqwV;2+3g+Xw>&x~j}!OV=8xyYaKP5R zK_l!fck9AKhW%~mxZ-DH>tEgI_-BctUI+FU=+n{{w#$AgS0>~xNhllS`_qhYVPB_t zgF7W9U6Hn<_K$tHn1b)9_wZaxT`O)`y5Nd8W00iM@|e74iMJ;Yh{SvCk zhzM=ru4y70pf?qeMNcI6&^KB3(+Q&h<@=*s(ZOM?l0;?wi@ZcrV2^;UOGW7#vtp>c z#j97Dy=Ah~S_c!wrrMG^kf`tsZD?l{{rGAe2>4-syeYd`PXU{|aNK^#C=@|GHdABs zo=TBp*&GMf42P3XG#!U7wsRu!tl>;z{;u;5=ktPZ?CTv2hx5BiiW>z}>eC<3&Re*DY6y*9TEO6>Sou?L z50tFJ;e;RzsU#bCZ`>gHEdTtO)~Lw#7Y+w%o2C|ks?IjNVcaoxzW&(D-ir>ro^3E6 z$r@2BHx|=dWApGWxgp_8gL^(8Q;Ggx#6AE8kz^=#aCYRaUnUp$9gg>}y#`VgX= z%;*!9R#EmQUQK>~32&w~;A$<)FBURXV3@p5ZjV=EgZd%}(YswwC)_Y@p(_RDje<>q zA|~fuH&gulAY1RBCsL#zWDJ>PpD_Kx5Xrnoi}JJ z@6&36JmO+(aym{yE9ZEVH(KXy1usu!Z{-2H^~(43M#5yJ?0h;pv>nxoDIi<9z5ox;3GWVCs%--~Y_cjY+30X%WcvvLf^HWmtdlq^V*& zYgT_isw)UD*7|S^dKv=i`pFu4qnrn95-OxLpCp^eqKj^ic(t6k$L82=Na`7vP9c|< zWe;KsOHzchWca7aADSC(Z@LwCN9lNugT2xDvVytxV&@r0uu8~^&&qmKQ%ecCW|*kz zs!muTEAr`nb1ak7Ntrp`E3f4q^8Qp3)z!-PKe&yWGe1rSNYLN)epPNRRi|q+9>+4j z#MSKFbrH>1u>Z?%W%hX14)2%$hIN9=X1;l8SfgLH^(STJ`@4K+<})Q1R>IgPXi!FM zt7z|amf44%>;l^QYujDO_dC@Jg@Wfg70<1A=c5vyO%{J$j}y-je^aUu^@qq=Z~e4& zM6BE}xPrrn%Q-+7{`~<1_B^l4=plT`Hy z%wNqAQzheN*mN(K7xxFo{rS*Ig>^NA1ZY=6Jl#ke&^(hCz)`=~yo8KD1z@@~FB28Tmhs#dF@!}(g)E%s`|uURc!OEN{2 zat3wl(abJ7e$3B;62DaMd(U-}qzKkK60xsrOWf3ZV0V`7r9NH`n?pz_-b=c22OTyu zMC%l%X5TSBxnHFlN5GkRdY)O%ieu8F&60Vy&9R8+;Y6Z*)k!>kU-!$c-I~)aK)b*+ z(AsE*+~JxV0CiYL%C3S@-n&@rx@9wL2;Cf+F&;ud*(af|t?n2F+Fk?Yw(T1RZ%Zu! zRND&-hdvNTJti`UV*L%G`HY%er@-}b$iQA$0}8DYF- zK|KP?XYYxbUH&~7N8$+T?cYw_*3}oJEZ)60-tu&o*MEo zmX|inPzCI4VkIdd$cx*&Q+ox;!*+zJm$hzfrj2;-{k7@DtdjYj8TXT`J~q6xX?b7V zgAuS9{>|oKXFkA>-}|#dQo!K(a&nlTCjUVht{F0Y$*fHsH|eU$IxUecENxm~d!|m0 zg!krGzOyT6;&X^h9^_K6!!?cyS=i*h7_rD>AJu-aH8a^S>svY`!8HvMk^2(} zhlu>@;|TjgHUBB4(x|-xUMW-F3NbM;d+)7Unw!DE&thO&fg&M?6L69mH}=_v?;If40<*nGwFE}Myvx_uw`ZfSpulpx%%~N8Y>@1vLnpm=Z~ziG6+AsXLH@Z1*ki() zuP#w^AdZ$_RtzVXD9Go;gxQ&WHI~EDJ^5u@pFFwmj&@jwLeK;?wl+_4LtR6Cl9JW$ zUJrb;%b+nXV-F9P*+!IPQq+Z3*C3;}1=506NUuDSUuU|g^MmbpHWLm9a;{9TwFN1g zqh~7g({8(%c>l%@yQ(3J!{!Gy1H{|BOCMh!9eYJ z8$a@%d>6MY(WkJmGSDjN|2ntI*yks*{%QHsiULAcefK;9YO6h3y5M(IwJcKijX_;M zUm;be}QT5*La9MlGztgO=OzCEBZ+51H zVPdO6ldCmYfv>e7xsyLhJYnx9_(RuxfHgLdS6@KPXQ#Bs9@Wz76iKj~I&2ndb3Nzq zA)>`SgiR?xloU_4C*3dB5L5X~7Crhiwd0oOG+Zs7JG#Ado@>HZl*!X|ac&oI9VKR_ zN6q$VPReZCxc>Zy6=$E9J<@C$Jh)^tULri_D*WsCz7=CstqpIrZz{xJICg*HHC8P$ zdLv!qI8vNR=P2QH{;@X|_k88Q$UDEt#Kh=~t3@|nk>bR-T!gi3j2ebb@yB;%sKv8t z30ku|SyLaBFZ|1T-#YFYpA3ItNg4C+T6tN^=Fum3*?Ou|6W&TlgkJ=yRcdZOOr~u* z+7MgQ^dq$H{BJsMN31CCA83))=CKv760zLPo7?_z3B#J1!YOiI{{`wPpBWFD6ytnG zc_lq7b=w#j%#Z^To!Q(=E^!VxuX&ZYfThGon^seepP)4^8%dMA`7^=ksq0#W(@!ZB z|0Z0P>gMX*20AzCG&>*~L^+L_D<(};n<9EsZ@s{W_b2ms&sh4|v2Tqa+A*&7b{vRc zRFD#xJX&hB{bIHA0vtwldAT|B5y3s7#cMMWitWMuK-~6QY;NyEt~tG`Hy$Ah+IEpP zW9KbyB2jBg)dE;{0^E@B_|!#azb(M`N&W-sW3P^%qM37PtAGe>4Ru=P1?v5|ir=-` zJ4TV_qppe3at3L=RnA!}(u`4xfr3xo60KhSVKTk;tqEQ^UYn!SM8&72@kA1qlmQ(a zwD^xd3eZ9fReB$^{b1Y){zbfiP$1Oac=DjF#qv2DI^kziL94CJukdgPunc$jQDs#1 z-5d;X)C88sJfwq#`_4OoWY(ER7YYs?PSteLC>>l`hUhUY651EGW4PT`Gn$rtvn?o( zn6D0W4rP>l4~4oF@^yqL{C^%MyG&L!%01SF1_)%O8#Ypr|KTd|^lAR*`XmDENO-)j z6!udW1X{f&r6)~Aj=zupmhYOj(0^)C<#1MnA1<3K9d&;Ap)`>tSOKUdp-`LpG$gOt+K)7fSpZZkx` z#mM?UTe$F&)b~oRN!G5xcf$P&XUcw3+4)`O(h0C00}px)MlP`OuDcci5J3ulVgHzcxu^?aLFREB>&>YQyM? zztP2>SaUna=wi>oyv`R~6kptgvC#OyOV)Sp>=t_)jhTSuum$9Dwh1{*s$5U6yh@E=(#eMM_kRTMVx9)W`T`Vk5odIbECfgcorC$5{J&~m7t2?Yg-q9`k+?S{IUeDgj*$7Dz3wN^hLf;%ik z_TTT&AqGhfsPQed&CR>)@P;$q;;;v`8{>y_l39SC@b)12Rx)y1d)c;`k~Dpmj`~K6 zo@HtL`R(uZ7t<@PE7lJmemLpd+!VXR-#@W>;EpB9fQ0?$l|`Q__&`%77=G91->)X5 zPax4GDq0lEzh3e2C|I2M*Q}oYrymBSOz7kPbm0S6QSk|k4ff@`jYos}uUBeXo6!G5 z7A7X7Om{0Xf#P2y0s#gF%*p@L0SziXerlSPd8$gpzvcFU+t`#~{YUC3pJ31lAFl?J z!haPYQ2zbK|6}g0ppyg=mUs03mYN2Y%+UVYe`juL4ihL>!HoX;=HIG{0-CD)ZwGih zyr2`N8(oY49TZace>*_?|J42eGIjSToAwupR|-{{rI{`F2Iju|d&xS3!E)aoz44qD z74MTgtgo-n7u8oMU$>&YxkTk3lNrbVG`MjvPHo2WUmK4?HXt13*-gH?;X!{|OlXP; z)eBU>oUL!NN-81Y;RpNs`$fyiy=FK5F;#tfV2)$@`5gHF76_S!H#hV}x~;!Q(Fd-L z^x2-)(fG7{jX=(dsmZ=EXebJmC+w`~pO#81lN_0)6 zoQ$JH&lyqiH0V%2yXu)2-h23GVl2?+x(@Fg2IJkGV}-BIT(J7M#sAsd1#eK|unjE= zOCeQ0r9&QVKh48hHz$KVR7)POa!7G9G%OqgIv*XaaM_xmvJwxwMR;Gr<8*VT-XmW- zcc@%$*e~1ql)yRBsymi;oQhUl;HBSV_Xd?W@MtCAqcrUz-2W^k(hN6V)XH|&%kx>+ zMd$!`djz?i^gD|$b=P;!XYcIPpUw2_JY0@`L>rjRn0=ezxeqLtHP_I?&@e!suqP3| zDA%NByskkGP3w~=%yDP?TNih^_iAy?g#7i9bxDFxuU&B4)ql_dvqL` z8EFU9eCJZ2fdptUqPN0PqKI~0cfL|yUY_)Jb0F=N7Hf7=M3n*`wYg5n;B{%M3BJVN4v7qicu*Ag$z4m&zJn3$qpCcEZGgizpvS{JN`Iz@&C zA^6^Wm-8*J3E8wA4;(%5a)xO#L@ww6(%`p8#xu9x1SV zw{!SZ32|3vGCa64!^Lro8nrL(({-K<1miQiH_Vi&nvw(CDqs?AVWB~ODB^(a8lnbV z7@jKXj?H%b_Bt5Vi%b!DazCnS=A~U<(rwM86rziL`C{QP`tn1EddshGe|}gBzsC6D zplh_)#e-1t<}I@Ojb5J4ZTcN$8sTd9}>ULUL@!JeaN=j7U3ku5C zx~N*@OQAb7UI;DReIy+(Zs)RdgcVA=GhOp7)$G;!Xr9CL=L&-|>$u|kE*$eLn1!Un z)-QjqF}FWdO%Q84X10P12?q*lj8Q&l22LT@g1{fc=Sb#o-r<*$8MzP!5tE|3cFFSy*A>K3CI2q;p zZj%NvcRPWR)v?0&<#_Yk)3x07`)@y$e(>17oh0Ej)%vNpnWk>!gt5o!LCsUE%kz_( zH+PJm186hC6`uGQuGR7Nn7}%Jxg@H=I=`Dg*Bhm-YE`L_p}@2twH zyuV6{=!4cTvvzUC!N`5cVkr@FSYvZNw3;|<66=y8SFN!BLBTF8TZ95 z+CJRmcHqW+y3x+bFPx>$zdC4h)THuKuiOKCIs^SZ`ygS!3&Tmj*_}Q7&+Jj8+vA@h}a+Yec zZmwo#DPdPgy#BY>^*T}1;_m*9j@m)v%mu4w>-oj9@biP*CH|e)FPPOoa2S^75Fk)* z7&l&C(41FZ3iFaPMVLQYDBke8xHy`X+&ChYPS42LtMYEVe8;5T%vDe{U7MaiMOH@G z{hP-kP$ryQI8fKi%S%%-j7o%&@%n3jU7S|tmo~@mcj{OdfnP$?IXl|<-S33eC1~Cn z6!lH|BhHm}cl}WwtK!l4`r+WH^QMSqRJa@%?onvF%dUT$+PQyzo7GNCzb#F9g*$kq zG|J;(GoaC-elM0_Wx!Wv!P%ZO3YM*s>=ZiNhzS+`sR{u_L|47pp1=gMGD6c zsl9Im$ibZ!e%y?l^;~(dMu4!I>JB3jg25h^GmM-NsZogI65t5@s!cdi^JRYPGld8u z<=nUU-p%0YaKFC9qhnjo*j>(j;t6)_i8OnQe(}M9bJo*)TV}3<`ehiR@#2)+vgadd zO2CV<07)*2-6fxhtveyZd#*nm$Z-`C2mQ3L)}l)?+UcZqbuy=JXE2te11m&<85$-5 zb6jTnig8|@IKN4Ue*S*4yL zfrys)&d0Ez$6}JEM%m(ZV>9d6Y~~h@#Og0pwN@jkl1W-~eSc^406(0jmb3WByt`Gp zQL=7aXq{0$y7`=InwqEPr}c9AMC`f+O)eZZp(l*zZRexR)LxpE_YF3K$%t;-w*+FX zx{wJ9sJH~*O1hNtc{S_}dwNEloGntNS8N;~8<*Z}WLjIfwTpkj@L6Qv`H2qK)d)!uI z-i=Ct!*7U?65!yIy}SNHn-?vxcS%}i?Kbb>wICd^T0-=uZ#>cyRgbwfaTNBBdUE;c z{O)S?*wDg_a1Xa<*wf8q{DNK;5)1{qad5)GI`XE0b)>-sonCTYWa3!Rel$Xe;`F?u z*XV(z3Aeda_kDq}Ek)3i zsJTVD#ea8$yKx~(pWdepwwmBH!ZRDQhg;7mYP^a14ra1>Ot)Z}IQVwx_3m`lEG>*_ zDeM}p#e1UWy`?Z+Ccbk;7#lOl^(H16OkuG2-(TbdRAxIOsp%gb*EHV~V`-!Fdc==H z)u8r4j8tLX9V60%5o-?T1+aIy5JD*g#pB@D#S2m#%M~ zHyk$-Qzlc=`uqEve)HXvczwrpJE(OpmV`E_QS$QGVwgxZpzi6C5^&{qYqOdmuR4aP zy>`>~-$_d$OrIeN48pK6-Xi)LQ8RIxj0cw3V`YpsyPIcoa#+qO`K^2G9O{_rGtuA6 zWiDvv>tg+!sIYTNCkWEhd%itY6XVU!OU>8}vX-FU95(|ZQBi+?1^<>8sm|#4$)t_T zti!*jj^^pGpW9E4eY*L4rtZaRN^Si_VY~(Jxj%iY>U&`+{c1J zLcgcFe=FYqDdYq~rFXNyPW{mb1{s{dSbZ=0IzkKHFcRS=qZ@Rl4}~txiK|66wq_hG zVrcMMT~hlcPwFa6?CJ#MDflii^81fR^}T;z2uILa{SX-sv9SL>^ zl5oH;bhT8sy@l_YQIAEdo=i)-jcwiBg(H6Jkr;veY`QkFO5|I2cee=c)W{zGhf*HK zQoCGMveE03s&p0^VG#Tw#y$|_3ce83WhrD z1xCZ5z;Wg}2@EdvThtPDzni=U&&;IJrU%cc8zi5vYrAc=U`ocK2t1x}?vCZuNBDdU z#3l_nVQ202zHZbzRr9Q*1-+eQxhqf>6ZiFJ*PV#1;wZ*nm^*w#FK9o>yvC*@%lc%R zEtM}M_1L|B5JUO>z?bwD0^Zk{q<)v5q+Z>iE8R1xGk$Tpzqa+k;ix;&q^`dHm7m{} zscK)Y2@&k8obou%rz=ja^q4^d2{V7-#btc6EzxTezcU)wp4o}ZxYS?9Jd)@f-8zJ+ zQqk~I&p+5aEa>viryepPAXm&TT$9YAiDBi$07?W70ij{BR z1+pS!1s>V>b$*9&3G}oL;8D$&J+a+6R1!{7mg%k@`^h|9?{QhZ$6-f@QoFP|pW$eg z_Io9~!BNWTywk}ITY!rRHZR|~d%Pz3rTv}?Ylcy~1*?VdKH4`t(?Uh1<+lkMgiBqF z*tfvJJSPd|$mZfsjY2Ru+@N&-_RqH1Y;6Acu5-b0(`>JyyVLB=+t*?C&%7^BR&XV@ zY9HypqqjUKm@IECx>oX=+}E5Q4zF{;_ul)`{dmUvwBii&>#Q6_K{~0VbzM+c zBgQi4@r%*^A$9cYwBRg#J+0$e;VU0-P!v5_Cz;8=?61DIab$ZkR?2CpoGtMJPC8c?T>uGFVF$Xv@kpDvBbjp}WYlD>H zI*pIuMKcD>I4WQoA?~rixi6h9A5Ly|6NbrS_GIa;Sf$sO+|(eHNHXatibr$LjRJ8M z+ZF0~Zyb`Gn|tog+L4Wl1s-Zp0Jj>F&&S5006RQ9tjSrT zUiDr$4o(%4;wg&dG`aEPrcsgP#UfbBBO-1Cu{iuT+ON5(i>hf)ra#|BRVevoBs$&J zYo#^9RF_6lW2#>BJ>S#xH)p{|NO-2_~d3`Y7hQMcQ3tApDcZ`be%8qjGG)oYgLc zO>Csvv-#%h1v>eL6?v9zX5FZKZFC9Q%Wik(4@dL$_6*5I!#&G84L#%K$KD#92vY~V@B$=9yI_N>dKhkZ!@LFM$Jy<|Xf zoBtmp_-neKTwqCqrH5qJCr}i5k7BGZCT4!T9R41|2zMEvFMnavrniCaI64w3oBNRY)sg zJ6%)eD_!jb4)i~({YNuCV!+6isSc2wK(vB5l<8U{hv`zwKaxa2*z4`|`71^6pn*Px z(d3z=|EQn~>a#14+g2H1dtzYzMRlSX?SGnTn1V^5heiE+IIu2F-e`_%bKWm@iJ?Em zCnBdxs^^;Kz>ZB2HOu;k{h%=c?H7%NM}R_`L=g;3Ozrle!I6OH04BJ?Xs5nRg|xJX|Eh3jXQ}ACC7$Niwq!opCwbH{k!+ z%b=<<_g-;Uh2+kc)?mKd{5aoho;_UL?X0dAf&L1aCqN%TbFv7PDRE@e` zX%{jbsuXcr6oK8oO@-uDd3wbk5)(grU7Y%)qTJ4rrF=@~c+zae(Q$Ba@ch8BmleAO z>Mhj~%~3wz@r2(M&rJx{U-6=Etlg#c6K=l?nFTQ$s=9iA&~DI5ig$xXh3T?7GY%EDN z>+U;dwM*gnMrdat<$?5HbNF(Ak^AUC+c-gw-Q|oTNL=Gv)iwxw(xZkIwxdNlXUD-m zw6axVYJTSFL;~W#=zv26jU@=oPFZ=xxBHl5icgY4@`iK)-?O=)6#gzC^huq{vNK9l8w!p zJ}1j+O&uLKe;iPWdypu2)NFJ+b~T!}p?@~wX{Vzm-JafyiIf-tIiL@`6mKp}SCoVBa&=vzs3U@TgYpA!M>d0!>{@*8J@2<{()vl2tA%4o3DTWdT*nC&M-BJof zQ>%Me+J0cygrE)7*AnV+h3e{qvFkr|@|!^37w|(duDW3$EFxhiPlITfkx^LvbOQ&;q=V zLV6(v>OH)2_9*Z=V}L7M33+`Jj0C0>7|W68s|2}nC=aMw=&~9_4{ch)i2L#5ho`6K z1M&M9$~XVgR}DU(rKV1H|KZPkHW3^TJ`1pj*mOitmJWxZLl)m82~71u#_u;2Y<*rJ zHpeSE#@dBA1bW22^6TJf(EI8Z8p`=Aj1(AXg$F!L0AdhG1I-d@-qQxUN0$6aC-QX* z^Sf<2S!Tb2XHS&SI|kGk zJpB@9IY-moZOU35LoXt5W<{a_kyz7WbLfFYZ_)18xLJYi)k`kSUV}4|3}~ad<^T-d zNpn0ryv0nSiYv>c2TnT7RO2dl;m4_V+cO7#6F?PLeiNAqmu``M?*3qfEBSw8R4XW> zRmf6;2}P@@fc;0#P812VM!O#zil74)@fWKoQ0YVanQTH?XV!{CJ^R39(Zm1f~ zD^?vi4)=@o)AwJJ-&m2z{0vX|J6^-Wk`aLuQi&T7xL!Q>J6ZSJ{Kg632Muv&;^Iu zs`+@7t3*Vj6pW0)u)A zl2bnLKF1aoo+y;JVML4&n-WhHdYorV#BnYOrg=e)cb)=Qft0Np_!bQ)J3buv*RnLL z<~#-->&19_g~-%ZG+}(}-o#9$8ywEo&<;*Oc)vsX0QXIbA zlgRJZIhUe$dH%H{C3OTB7x$%Sth=pb+}ojg4=zWyAox!Olylf@n94_DMO z5e=x>fJG$)uptTzOY-_}dM@`RoIU*(cN3c;h>uzPZlb`Ww`}PB(#uLf)@N-kuNeN1 zH@VwX3U-A-tnjlkO_-6;wdS)h`d!xj70o)D8%_aWyUV(K) ze#At|^90&=ffzdaYaFFDkoT-X#Kdk#cAI)D4)1=v`z^guaJ z3ladg?js5fC%TcGDFh{t!tBOO&}j(|WU4lJVJcTq3DndaZC~Yso!qF^t{uXS4}V?d z(4XMIhGSEE#K2f1Dj-Ja@uS8j1LK7TWQw7V?psvzT+y79T^rH%8n!xi-7_TI3xtdzBu7~4rLcWwV+%D7-El=oI~*8 z(UI_rE~rN^S2^GtRps8UfgYofhj1}*ar$QZ^Yimk$sP@$zK6^jU-DrM{a#f4pdaWM z1ia&tK0h;f{~V8H$3Q}iukP_#FpxguxUWHZA>m-Yz0KfeGp~J_KwKtv8mUkjIM>{|h5{aM(ej1pdnn ztsl3&7^s7SPQ1s+O7i zUSc8jMGC;koy#PzfZl$}0qy?^2xkWSQ$HFwLqh^#evlt%&_WhvmRZ)h3I2eEjb;4lD|tJy}jj7_ul~vn+6M z)o_GW0{Vm9jSXFO9+2ig+QVv}ZK(B52nrWWescagBwDdf;L0Ik{yiu^uK0B9SNJ$y znd7DOj?zqDW|9InHK{jSgIPwPhs#!iu&Z$m-_n%W)UNE7#%5Yp@B7Q=yT#*W*0mxv z(VTiN?d7DrgS=jsUd-xANte_6>2e{u`vu28rl`La8&TtsIIf*%I%K`f__oHbms@B5 zwRBXn7~3A-?bm5IBDzb+3c-nUMxTEwbz%Q zMF(pgZcTryllui~33@){?9fgPH4f`R>K~pUp?RbT|_$s#Dy&o2$_jkKVqtJuWaRdon&`b+9q^EqkLQ zjO?AGcv_{ymSp;+ew|~9x?sv%HO91=ld_c!;ey6zlYJ+@mOXTLIhJ&7r&gUhBB<=9 z%AGbFzTckhcXEBs{UB?9gjMTkb7r{C>KE`^JxLN}mcehA*rMO_+x2DTs9qLc%-=Sh zs2bjtP%7>+4qfIp{MAjC?wE1+L|gU%QWe3jQlPBqn5Od zB1C7mV7KK+5@K~negu>eH(V?mk=Ls-?kGuep035|G$QN$?aK5Ox ze|pl@jzyl$kR3+Je|XqCSz-GPuU93ngFUWjq)<O&4ysDRx?-4mTvw|iAaEVg?$`yY5$+KfvJl;O20MHlN+^>Wx+ zM0@-Et~q|D)tH<&YQwHmR1R2I!F;UV&C{5Tcgzxxtn(dl3-rg!orV=|VfAwCshNix z3d)L~jvaCuHxE1rvaURcECd*1!TN*IuWf4^oyGmFDUSwnP$@Ec&&-V7-A`K;@pon$ zBP%D$4Tf7vUN@#3d`wVH5%ugiq_nY4ukEZ%K$W)I>3ZQ{P3}E;JiL&?C&=&iL!Zan zcjSGt#EZ{nKLXPeVmKd+O}9|?ZZopX>pALAYxKg^*-FN=PdY?$f)#-F+cvjV8 z=t0kKv&&*Tuj@J8#9ZF|C**?n^)1Z1WZZvQkEasxZSDk!4!rlVK45n4_ zfu67(r5XYLshGZ`%c_LnZS&^uXs>Tk#wHrDu4r0y#Ypi_5|#B9k9uz_{aWrhdpzA( zpI~FHOw9eDp6#qwGvh|MWA8!{gL|3%j#7F;aP3eIZid>WC*Q_{N52~`5)u=f_1v?p zE6rL0^-MekLHX85&(QNVNI#QxPQMo9Ldzao8xz(m)x=8k%W$(z$B0ddiFLW9CTyJy zJL-vgEgtufs&6Z;_6j*nB%M8A6}(-cRKYer6C=yt@<7qf!p0iB|M{&agqYJf&HTDR z*{t_j{d3a|`}Yae?7`Md+e$Xp$B(kU^j8)pyf>(Pm2y8ZLE!jnW&-Igucx-wKPk?G ztYf$4vVD3B25w3YXRCH3OQ$XSyciwJKYnG5cJ$3y+{5>af~SNZrU%*>u|GHh&^dNC z(MCC~_}J%{JRUD#^jG}BOp8NFv^ex6(=+cB*+hxqOA;Vc1j=RJhm!jX=8w_fREc+b z^G-Z_Ot%ZQ3RC)`uwXS2_((FZ`dBSJrzRSc+0m5kew8K+Z&P8 zVu1r()s!&$7JAy2PT%_@4L0urRcYoVe36*Ov~k>8Gn3uqFib-Dg7uB@s@P&>MPfuX z2X0Ur|041?vh!3xyd58>Ve>a5%*1QYs*@zSWyoiD`_2yCZnE&KZ`dU5X9q-a-99?3 zarNlBCFX=Pt0+PWRSMCnkN^-Lx3U=l+N199Y4%8olspI(=(}dO0uV{*_zL7$9#SEgkii9{`Y%p z4Gkhoad>V>(-i1-s*(#XjdUZU5876r-XVGb6?I?3It>qUbPLM2(MLhO+Uwe_@d> zS|l%g;f|QOTkpEygX9Oeh<9kTEZCL=SbT6yBU;Q;TbxDWI}Hd@B#*NXiY(&guLEJ) zmc%ww0_mfoXQW|DS6?xOi)JlY`aTKqu@BM>Bf$kQi}x z8J{#iBxWz|x55YC#s^9LY^*aHsjzBPH-v(GtOK(Nab)>7hH5Y@#q4l;3yvo1aq+!z zgK+$1kQ8t_3CpBV0vZ}YIGt5z=k-!8H0DJNR8bn7Q)J+i0UTJl9F4&1FwImB91Sjb zdRSD7-7gBu=m!G)14x=q8l31R^kyT6x&Y*0Pm`7aKaM<&4Lz(=CeU=MM?(Q2>zOu2 zV`}P;NKZ(!zSoE#=*2w1t}*ywZG(n0z=i`FloNH@+_sU#kvQ4Kmqph4g$V`5-aswH z_OBZDzw7j3@CJRIT#$mp28i3hZolZNqE>BbB-1+GXO~y^6 zBnql03^pii|NBGa9l3v`h+?QK@LzC(%P2Xt9gZh#eCBI&6-JUvA=)y8^Naxjz*N1P2dSSu~B9(qXx zJ2i%Z$7y|l<_hn0^YZn8!>70% zMYk!VoTz;S?{kukHHP>7pr)ONUXjRM(9Nx%I0$@%FDR#>{gL+3#v0UnBVMj~ zbJOnj_1@vJp{4QEPO-6ea}JtBh1E7p5%QjS&rtJy5X+KK3!q5^esf+CP8ZE$m~K`d zh@2z;h@o%wVH-c$x4x?HnSRVI=fDdYh6ej z*UY2K9q~?$W0&T~2o{)mx=0k3D`RpRmC;#+JPkTFX&|X33hxu^rhulrn0y!wyh&EY zCm<;|0gGG|(>DiO7CD;gN0VBLLo&n(=VQBv>7`MK=$W1CXc?uS{Bo=L&9c@_$@GLJ ziDTc&yKJ+ss6REYsZ*Z&h{Z2T97|!*&)`n)n8K2 zWOB`JBjtc4ya@{FGSMalV)Y|V5IMhPJmab*ZK{PF&84(_P})Bl9|Mse!;p(GQrOP`A6yj6)oY}mr2>8K0h^tO;m)G--Hjzf@MAx8zMV{lQt_y0dY*jGilAVeouh^UgTY|BkMp^7EuY3 z8=CwwxKWp7Lg3SJA^u#dUo$4 zkw27QEMz5YJzxbwZ!1;%$?|uv6)|wY!`9^ZyDy3?r@&Y+;l%8jn2-NFyJi68CRUU0 zm%Q}&o>(sw!8_IK*_VcTRPG$}ZufpTD2i_U5;@ZIl=R%v);gVPN$6P0?&bO%vlf5% zQjz7`FOk)cmYsI#Z^TyH?Y;~4!SZ_V+7$*}=ZKF$v3^k5Zcj<%n;k2o8po!^$Bjzj zpa^S&oJ$;!63b4H#SxE*Kxc6@S7{1^^IV2TMjl-eQGymiDv1zof)3rA zz(9%2zJkB0Ao!b#h8G=8j2s-;y@ia!RKbCr1it$G6HgAmI)F3;A#g!(RH)nb|Ip~E zC;;CgM4?g#%n6OJ7*5m7HT&e(JN{y=R)A3T6@%abF*9=p@Lrk`5O#k^<{Tdsv7a6G zp^t_>2BUX%(Y^HvKmcwiusDFvLg)$u;Rf*X&%e-#DC7sUzi6SK6v?;%jQF62a(R^g zH_s;~E@W|q{g~kZ_`Nx`PYgUVB8%eHKCeKIojj^-6B`4-B-53?WN;bf$HGF-dn_$x zNL(1;2y{@~{(_r+BL5FWB6D9JfZUz<0B-t%0oWby59a)M5RgJ}1Hd{QnCm?d60Ztj zx~dhurG`NeE{qT;1C3ZvAYAx=8i5pmG$(-QKIffF`#KVUx2N(T1_s#v+Y7W15I=wq zrvCj3Xlbeg4ARl@vD+`}mJvlT9st$nd&-{0_1UrsbOUTCl?OJJk@*Qd<1z# z%Zb*RfCNMYT{FGySOM!p#$0j$Cw2jvH?@l#B@|#jrY`@xZY1cT)&nl=$fSY_+2g#d zY7Rh@_|O4DRQb)Qd?r{m0@+mSYVxu8q7urE;Qx%mH+>PgaYf_qOxF2_y zx?`1}&IFttZBI^4&Ng_S9fdc7t2V{X&z+qCGvqv@WvVg`ft%yHScDKc2khdlhQ<|8 zS%!nNJ>C1ZddMz zP*BrhAuheVCAxQIifAv;sG|~q6jQcj+hZZ5G0ycu;|}Nssu;4r33(DzV+gFom<1CH z$^bWC+6N;jMENY1!vM_Z9(}-r?X7#VIbPO09w_kquz;1PK@0-J#Wn&A!Db_ItP-@|_-yixaLO})r#w$D-?*nlf7^C9rxY})EnN?tcnd&#uyK+P`ZS?!5CLS{ ze$YGl)PRkR`O(n`xLlNa3>dK#fWV!tOu8PMy!MNeQHee|IUy8d&7xrkm%pe8D9SB5 zuniu`SwX076Lfplf-D;B!@EuUfMTEii0LqG4xT9 zi8YBN2qKRea(Qu2Ag<_SHs`0Nrl!ryoE>c4 z)C<3z9P=U3YiCBB;ASo*+gsr1z>AolpRY;s zQ_NF}VXLXB1!y@_`r``=_T_>nGj$1T1&x=d%Knyu<$bLwtVTtfT`}zZ$;~NB5Hqcm z%et=kA7=x`mW(O=E>i(~gj^_bIJG2ImL69&9DY4X!jlZQ;ryqv>Srbwlg;-I4JVTh zjWMhyOv488bTWuKVrx|mk*}S$K2}P;l2%all9iF5HYjmHLbWt_=vTKwOo?F4^&U3A z2f+Q8M8<1N%oBi^BV^Z0rGRcp9ZkEY7}C_;XQvaRd+N5>S**~PD7XO%55eWPkUfK1 z*Ubh{e-emAjuW4hBxc*n@*^Xv>M?&$AXNCRCF0E%0Q|v^e1gZ|9?})*GoT4EM_`oz z8FaL*ZpB%z>IO>u49~xJ_+}323TyTGiWJ<-VQ}#mX4ZhYyL{q)TcRe$=C9+uqd)3jd(kP!8+|J=wGVJX9 zp@0?!VnQU8dS-YPT1)W#nZg2q-*VUOHUn~cUhdi@|L%HWw)Gc;uKYiAOiNc++X4ux z`uM;fjA_5;zyK4cfHwFeyQP_A56U|(6ujo}EI#>2s$@rgGf;!5J6-$6FfZ529(~DV)OI#$(I(_9~vImj+Xdyd^8F| zduFEjlNl;UdaZyd`C{SYxlU#vXqd;?ByN~j=uVUb_Lww$muvc;f;^(GA@?%h{IJ5w zDy5|s^(P%~ZE}=AITG`fnvl&vQ+8}*9(rFh@Ef35N=1p>b7rdYb&@)qnfyQ zpi{@kjSVge-^THoRGfdIiQT1f6Lf~x35qxi?Mrz6BK+i2YfPRc8$T>vj0g1&KJwR& zw(s4$=NK$i{K_61D7h3$G)oJ~*yLgA2vF-K1o$L-CmITcGVxaA2@VM0+QZ|+50^6V zfV0ZP2T>Lzh0(@1>rlGV1}0(Pv{Y)ND^0`4_f&uJ6YaIvmTdh4QZl$y`1tQwlO&fQ zW_wAOexw5(^2UVPbfKWeL>;h|qk6m;Ic_i$?+d{$9`~Nr{J0{efZG7#%o)tsEt3l+ z(xkcm0uKxbx#bsm3Wax(!SK0BaT=DU(H5QHE-mQ7aSX^xGv9_-^Af70W_ovn97G!R z#w{uiLK%INow^IPTH*;sUxGzNNC*S&5v4&$v4f)uONvNFXazo|8>m>}lSsT2LFe)Q z@-mmCJC^$y%K{D=8C!rx6ekoUACGGAE(io*6GwJ~}B2*wjsM6Fu3A~FmThl~E zpsRCpSKgA5GWtglwt!oD`LfPl@S{0Hst9a%tAWuPgXx?w`!?ZUi4b|caq)+t(vJ+O zrZ7>}0Cd8l2aQZ28f~@c_c;hyb!ERWrmr*5NWy^0Xt22}3mpFLRKH`p50yv$j=RnX z=3N8oo?`FY1-idMGQlbnQLo*&YZmp$N>RX)3uw8AJ!qZv4dnYv#&Y=4a-fQWicyj7 zd{bD;;l)T*1ULG?MC0?FBMCrX2po>a7i3~$;(Hpe+TIrM9v!~q6=Ibmk>f47L?mT{EvZ*s4uiqc z{nBHMW9<*JJ&yj`El6Ep5nuB}bHYf`1oBpl#KgyG#Yc@(iZaVo`QP^04S6hZ&2-z{ zx9Y&(o_-J5=}q&+Jn^u6rso?Xt6HHOnwvVNVk_t}L^p@+%WKt>s=j4E+P&L?PQ`8I zi+;3yY>j4pb!h2?3g}36Zs_o`qZ-wFowhnL4!sM_${t_m-Ji>jLKFBs`~DV&!F3F) z*{R~1zKO@TNDRZtM$W4Bf}TyZq_4&2ErO>uZ@*leo7-~<41E2SQ?8%STK(;x0s$J- z+hFk+vnM{GKX7LKEIi}ZQq8ixy{(n49wXh5e7*(l?RtO<>y@5Hte z^qtN>?hVmup$U#)uE1Q+w@k1RQX6D>(lYA_9J8Gcbkl z7I1&~;HS*t&AQ8hkB8(Gu@P_ye62M0$Oi3b8;eX}H+RWG`iV-&RqerDu<6mv_*k#` zf~#Hx?7H@6hg%Q(K78p1x3w|Y*BAx972n^9D>*@RRj`oR04XfEUlmR#myD8fDkEKX$3}E_PL6~hHW21ioazj3nl+(+&=BXU|Yn)nc7mJa0-)lF? zZfY{2ZqR_jLUQ$aLxmcBbUYT>|DEDoY1f+ypd6FN6`U)duC!#FFlc0{kUA}cbX5~? zI9uJ7WBz-+k*ojRJEg6IPCweg?X2|Y@5eMP(HTqy;YIiB`2H$~FU*qVK%fzb`eG1s zwc3Y%Ef+~2s7QzM1O`*+yIpv(x|ry}7Z=bDeSW?Xi(~2(P9^@?q|xj7`iN8~P=996 zrNre?NDO-*CMjJ2ICR<3IfBDe=lmNGPR6uZVgumlaY!3|mZTRyAGZYz&9!rW_8Ls0Vn+4+(G{d#vDCHKwmCC@!KTvoqJp@Zv>W3ni>h(S>HKAwO5 zch67_VU?UiZL(vn6og47$`c)3D?0H;7a-{HBt_Ng)<@i=P z%A7#w+K_>;@74(>Yw6fNz>03h*Be1?R}7 z%vyC$>U-9DdU~`;B2MOuId4F588|-|Qr}e0F$YJNqJ^%9gS}9(Os`z*8pyu@28YEt zTwq@5n~ zKOaIX4T{&r4WZ)o6IzXimoJ%w? z;qGo|GDvVQ-8oz}FT8N3UEU`D8ws2|-1xkFaARvGm7yt9#uz<2WXW6J3Ai#AI-EIlH<- zBnqI4vWM-Mm&!DZf?F9EmpicX3tr zn*$SDrDHj8rpsYmzYDWv27tvDXVc52*O3)TQw0~*y9zIR<0|Mh13jayGgQk;M#pUe z=HYf-YAH{$2hBu+oKpJQ<+BGDipS(FQa&)n3VrEPsx0XU{`7@!VS{gh+@4b|+AvuO z6eF6A-YF`5M$9d0C&=HBVD84TxBBH-vDd&*v6-w}wcW%;btCpIO8&u2H>bjFTF%1e zyo{GC%AgKfz5^c^DTh4Rp=}!c^mxa&dJp!&K>4rlu#)tdG$bNtfk`j%^j=O?3e@U%BN?v zQ?=cy?6)^w-AI0t(e~7QSFHHBqf4U7ttyyR_FF} z5f+HgMS5S?(CAEEa(g~ieRxCgVSh?no6c~@O+D2&`vTw6T~g}LVqZz>g(iGkN$tG zy=7Qc;rA^n2uMq(bST{*Eg=mGNJ~q1$)=^4#-ZFzN2*zuj1g4YnIX?4!&L^QL~jJiH2atDuafS1VNN;}8^t zH(uDi<_J>DPU6}2j*_~dUk<2~%FAoh!J=tNvgL5KIWR2osGLSWb!Z>W{~9h4+mAc7 zG4JmkGc$5Cpq4wNg@{IG5!x1jvxR*8l&GcX7BhfeOD^mJmEXBY{0;*9g+G`Us+pN4 z-Sn?lH47&*$~tHV6FtvoQmu~^bkSdZ61|%3F8Av@`Qr9FsiwU02WJw)6{&s+xhg91 zXI(%4W`5SnOs&L|OtFyMy!pK!r-M2~JNgNM1T^me+fMO{bLpmZ z&dWQMmz+s~2`ba3Aq04T9`E|7?#F8N(o-(+oHXbNWa?K&k8WEKZSbclkks;kmNhgV5etPug=giJ^e@v)#L6 zReNWc56t_QaI+xbSBa7nwu4BP6F02q zRZ!vgoCDO`%PR%SGAmns#OK2F!u3;3O0L3Y78Nl1;BwzPwp^}^A~DwQM68Dif5UZe zHPgP}e54nF+@^OQC?*@nT%__lTk*ja-RvdnpO|i!w~j~aISW^0dVR`s_FMh+dAFE~Wwv8|DMSScSQBgjT{E7xjBNftxq@TGawSs-x~MyZ7p4^EHYs=B8t znvtcA?z-BShnMr3#13LN4ZbT-lwB}JC#h2kYOwby)zs>u5^MB}oT*-w2Hbo(po&%+ zN%lfnlqkEqn^pWgEX3s=@<%yJfmZM=4$IK&&vI%w*oxjg!lk)v`QC?$qfJS#UPo|s6|)2#APLJopP%aX%TvhysgGmU zM{Ot}MHzNj0FY4n{fG78cUcbbJF(R*Bev_B}W zCMO8n697;qp2z@IkK`Xf^u7uBhrL83Xy9ZqIPJXRRyTo?T5CD~cW?)P9Eo{Yfi=#S zgm$?#B9Xy;#wzSr;1Ds3GmxECCP+B!Of$h5xFZnn_82h;4)9&`*dVVEBrc||o!*1P zp`-O$I85@2RJiVNb&Pt!VAl3&CzDElWBYP%u3+OyXE5PWROVhi3Ea<4M15!#M=3}z zbd2`TV)(Z@#4{G@G<*uI6+tHqJX&lHHNk)ooW5c57>=-LY+b~g!J#egVrGmgt9M@? zPU)mp-`eBC)O8h&l4c_SmCPL~0TdHw_{{5`h_fsHJ49w?zie{9sTIlWejS0|fc4v@ zX&CK|kzy8(x|j>aui(U-F^i>FK`nD&O>n>839*nj4m_p1vgngp7^tz5e`?ksrxU3)#ltjpn93qJ%SCT)kg-0V!oyhN4@j)n50 z*eCd3jlX{J@*1>Vy_G+E!%H}Fz@pPy;cD{vif4tcP?{kv%`$#%M6;h%;*;6V4-(jW zO%*%3(lI&Uo0V4k$NL(^gEIH)5za~XzRfC*C44)!<1P*|9yFlbIoIS3Ey%17dqqNG z=d|j!KdcRsoie<3^g`R$7e~#@%vzOix83Z#7v-A*#mtBr(Z&{R_BfXso8TA__YK44 z=@LvJ5godxJF9xRpJVrCr(=yF5dBwMNW^l?HxRYTk>2z3#Hs{rZbtbo;zP@7;U_$& z+uH534?Hx89W?ASCJ0Y>S=;pTBh)1_Jk8o+L3XmkM4WO!9iC zd4;iXEJeuk;_fDNd7o@~zO}uF%)?auIwmb{!k+FdF0!E0Y4bNjCslyWa(yCeTJW>M zwDI3vQ~F@t*Hqc=vX9u)B-dH)uBpE|e1(r_T+NyqTRbzU(^M_Ckh`^{jTSKNN%kg% z-K26|BE7Up4{&t~07UKDLQP&5S zXtG3+_>YgtT}?vl-W0-G+7fn@Wny&22|)m!U{4~P8DWH}bb z%o>lMlLKO98yEI}2&SM$oUx@fMnj%7e|^>XbRe1DtlTv{%{{bQ#J6Ba+L3%cjb5ns z7~fjTs9w3)YPaa#~NFQGH+DcggJcy@B5-kMeTcDVCOAaUn#bOeH0*|1b>i0xrS^eZeIJ(a3z(xpymDPk_Asa;Y#!6VqW8F1k_hvSALq* zJ0=5g&nw5Rne&d9jeR)U@t?XA! z{CBqEZI;tchORpDlh}3w@fYeGs%m~8u?X~yaNfFezb9;b2AviwCyTO#+-$Uqzg$kc zhm1TA$S!xae%IV)@%dU@uv;fh)NummoMqg*nNxGFIs2n&? z`Cs!htq_#_?F!zY4fvNF-y7EE(=vk5cmt*d1E?93Er@;VIQucD&|0AxyOai%Q~+}H z(jh81u(7Aj`?>J((@pIktL?$6cV?Z}c24ad{~iSh*KY9g=;(#4;m2{CaqVhqX+cQ3 z-Z;#MB`RFbe?8P?73k%0g(#$ac0IKD~t{Xbx@!%v$Mv-Le&2R2KCrYKv7-I1bD+&Zg?+SKMXde)}A zg`t|5!xDbT*y2@aJhzQzo~98?;5kll8q?eG!exXW-`N7E+eQBglaa%)Mgi;5JeotR zHFlwBY@#T9=4Q@jPG2r?l@M%787CA~Xo`cxq|ytJ2gIcg=F zW&&pjIm+XX5akACIyEgVrs4kSyQUa)c#M<3f00oU&reRs(f&o`OdukMreZD!;c`7V ztN<&fR;tsU2yCy`A}PfBRo~nV7O3CDMJO?A4!ML?YjT@?Y%Nr)sFoa8ePl09x?Ui< ze&@Dxfw7bI{$QMRXEPgBthuYF?*ZZE@BJ!?h*#Rv6;kO6->PPglZ2o0tV!nQ-s-=e z6|-^`($3L%g#@fMp2H-lm*#vvxjOQl#4VaZzy0$Rf>%)|@6UqV=?h<{1bMNPil(S_ zve83lsPAEPFzkb?Jo+KFKh}Svy|ZV7+j{wPPn~AbG|Z)$bT1`zs3v@S`qr|0(%w4I zthH>X89$y`X)VgS+Fp-Fp)ue$nv>4uK3SMPc@YFPyjK z*=J@ZMkz_X=)!6?&sq25cjzu@^UOwU`t-4n@c!REHZ2RpZgry4i(7G}#+nT9a5z|tHUq$@8p;_K-Gl}@28=I`5% z-U7GLs_vFwioh*7g5d2tJTV`us`Uy-twE>eO@?;vO3QD$)sKNZT&u7 zUvnsxeOw|mard=dg*(qw_tD92)hxYcO)ZPN`0lzW_x`K|$djtE&1IS}7li1ST4L0b-%zM_5oRJX@z)Ves2|*&{E* zOyk5|VEn7|SD)-X)v1BosX`^Wt5H%Gsz;MZmHFQde%q?Yr6X>5bxcW*Z}CUGfz2cD zx4rGUmGII}{bM?TJ!^a~cYdThp(M3Xy57}#@Yfz%Tyxzis>WnnImZ3~zJWvj>H_mE%y&Rm-+Flm+GEvu1wwkj%dv*lj%ROjaxN ztIEL5q9UpP)nK9JIIzNVauzuilC>$F4xdU z5%8I|;m=d5t8PYG6G$5GM{0^c(lZ`#ln6?g1XxiM>dcfUoHX-@F7I7*Q~8CCCC4zs z4`BVr76;?XRFJ{xGYPW-4cnZSQnpL8*aFrr9r|vP{l}%s&H6^1R8O%wh+qEyc3btm-L^l-(Q^; z94XHa4Gpgf`yP{U8xBo0HxjO~n}K`4#sua;*xp~*p^q8)99lK)wmh6Bl-y~Dz>x6F zlBs^(%eNhCx6deKuFh(Dc47)?857M~^vNzDb9KE>-Dz!EkeOMbn@L8@v#jTje!UNpT$u3BTe1Z zN^x)(|6{V?U>@C*s=fV5gvkutuBLg?7BvhFHb}q!u!#5tuK&gxTz?rPQ)%AFl1RTJ z{Eryy>)~Ea|HpVukORs80dgQJs{ANGP^l*ZvLx0bvx72BP$dEr*O2`WU?Kq|9{=lV z(6l{pm&gW7t+}y-K5XxY7d@bRm5getMxjFumG$8H4HZv{f;7vN(gfpcq&JD+zUdn- z^ZCK}V?qBP)aHRc40@^Jj(q0-Uv2>i=whdaMWF(^NN`~-4DKo=sEv)9tbA1pb;cK{ z@b$ABaKL~%WPoRbx>w*r)gQR3`k&&V=``6FDjJ7{jE=t~l8b9RGn~Ej%4Q0FcsC)Z z{v7++CrjmEmW^t5>|5MRQtzP9X@@gFEOB7RnHf4?{!SDRT^gCr7);zh_ohbpJls--VT_ z55GI5r^0#{Dxmg;RBM5unCUFQd22nE!({mS_Li^lEpZfa)j!uPDsXut@q|H{ak!IN z?;h7gfjQY!KS+&Ze!}U<>lBI=2_LEtxe(o38wq202en4}^3F@~h0$nlF>sK2*MfEO z_cTN{Xn!uxq(!s6ql(=2=)B?2A5)64dbDDcs)Oy0aJh1ce>dr1FA9Vr(go7^V$THN39ufn zKnDdREi5G^C2DiBp?~A+@ADnc?O*GYgt)dG*?0p3LvG7iy?KKKH)cf3Ym3gA&=x$7 z^17<0z5TKb94?%n5_(dTk(@?ERQggS+NLkT2=Kx1Y3^~p0rNi6>lEqw`kMTk^zm|Y zcsiE@Cl-wUz4>+dn&>y6Jqc(NVWJx2lIBJ{ZZSHG^HhH1l0F&)a%V5NFcGisacpnc zfwBe&v{Jkb6(z-kt5;`_64jOjh6fY5OdRgF=}A>+<N#z0xB35Ep$Lhh{0#hhIorv`g;!MntoKK->Y~JdhoLZV7kE_`ER(2YY+uGOXnmL$`|*yL#Y0G9_FhQ4h}opT0dcCnahOahB$M9u6L7RMv|a zZ~u2^G&MOJ>|(jP0h*`-WB9XfQbihZjq>Ucjn5t!%O1es7MoTsekx5{^1@N&vA zU{m-^DbV`N!Ou#?fuEJFH8V}P!P0{`pIPy10+$<#capL~Rzpc(58ZqT6W}~TcEmW9 zg~Zy__@0M_K@t{E)XSop@h61_6h8G~VRSUcv*_wn;*GOnvGT(r9)H^&B-GW_HNiKj z;z&aPFR<6|nPl~s@zG-7Bk`^UQx9_GE;i!B1bt>L>xD{VYCT@KKc*TAT9v>M0(P9b z?c`|O5?|A38jfIjsW>^JJK9SQ(RMFGjwopgc=`15z@i?f+5FKc--6!tfR)yC7#w3D z_zlK9CJk(GQ)d5r0pxKGWvfu@eTTzX7G!5+x6{hvCHjyIT`{6yZv0=aNWwH=COZvv zJxv0)qMdXoS~Qph?mF4&YzzX8=MziE_}ummfB?3On^gE)Q(#^o1aVD&uM?yPi!72r zS|>*f-wExXUo5L1^c{-^n*oJ2xSF6=eFCUfAlF6gbrjzR-$zGBhmK>7EAC?eI!U3( z)8*lURAf?8NEIC&U8HO*eNI$BAoQ5lLj*5{#}8{F@?9+=Ym-~8EgP6b?^-a*mFQz{ zl*)0CQw8+@=8v)T)NuD$p@$<*l_OhN@i(okon&@`bONWsX(ZfV|Nrh&e2o?@vz?-7 z6rd)zouXX~;bkJ9VIjqCA}txxtomHWIj51P(0p4K8b!Io-7p$4x8tAD_Nrfb;2XfA zvJFh9Gb2Qr&WT>fy^Lq3pgjU*!3Uw6O~BzHPq7>nMaX1-gt)jgD=9*E25(F7#;z=3 z4ehadp#y0$6htRQeh25rtDC2gIv-X4Tus>H_B?*&o8pK>!kN=XcX;c}6A32i+R@Q4 zukM~Z=a)nU5}Y?sGo-V?)YHIDDrw1tsDa8F^`1I>8Co1UQKxIH@*uNzb zi4C1H*yc?EaFvu8>io8S<1{6jHCR}WwU>2F$#wau|YNWZ=wB5whqJw+WP+us$o ze-CbR-*lyh97>0p4GyYT)1mC&6Vg?Z|XY z#KDZfeqTz+VH~tNoX-EWy5{TU^-M8a`Pz7lhe@mEV5ThHmBlZ9>Zf${ZTm%Z1nquP zhzywQHyB_M+*ET}B$>dK^Pv=}5ObPFn$T2|V@1A%8NOjx)f$)?<8!}eCGQyPXcY@U zwWHgGrJ&{B23;RuKVQK4NbIy)E!3i+BBr$JeFFs`@Vc@mm6xCzq{xJ&^mrU8Tm~x= z7lxwA4Q#(gDi*47aB?zjLrZ8B#VOh;@|wsTuVEahPtM0C{OaMX1|nXQ;HTTyjD!$T z4c}m$#gvfiN}}wSEq-1~3E|H?BzZ+5^mOl#i5*5NsEL943imzw)BV}YID01b`e!ZE zsanpgaWoJO0Pb=m9MR@3IaE{NHs!q5mbMmeiyMOx2NU?7Q&7&7YFK35Co&-=uy{uZ zKfbM0+>8*u&)4y%dJ&<=D<>}A6`IE~c`;vEvF&cQ!zdQ+PErEgdjkCYg5;Mg3?)(J z)=q$x^h%X7Za9b{o4>VuJ8AFYOhh`V;kI6UbFE3EEi*pQkBV~vO(G?Tz9PhHuw52) zQBAL*g;BJ4x6b$fYvyrSFUej}qo-&b;UyPy0h;4zl<{&oVEXt5E8m{MUL?|(MqNv=pm z1O2=XA{o*o5*YFKV)yj-a&D?uoctSRtVo+MFxQ%A+e7<{?`yOyJiql6UouLetN~`-u1ID z8m_tab#4@C;GRz0s)l|u-nxF<#2f%%K|4i%;W)vpaTGuk-BpS6A z)PX1tr%aRUd_J&9wNS^~=*r>hNc+W$7mFtqG=inSvk}mDjg0WKIVY%^+z#h!&~6gr zav!QgfR)-a4ure&U8i?AxDYM4m;d}loBT= zFK;7{WKIEe9y0hH@>ruoT^|ODnwNSv6X)+X8+QyhlPuw_Yr^YayoM zdY;5%)?zmyq0Z$p{>5q`4b+(kGV|4cJDIl9!sSifTP|}JK4`#ou{o$EByIV<2@#&iSIWd>*aSU15kXGYykS5q4P44hfXf1GVQM z;0+GMVW)Wp;*<2`8EZ|3;;AR0-Miqg7x;YI$1 z-TjZP7&B%V8#T?uCX4s3a&C6Tx4W&ndoJIcXfHUVS3EIitbYitDRsji(@i}YkEE7Z+0B${ z^H|RkkdTyk$OLw4j^$brAKrKei4out=28i&#L2FLnlxov1+z0_;xLEY&mS-uLYQF5 zgdvQiN|D4M2AG*x?&8cq!x6qyW1% zwj`h=6b6}d~ql*%ZcQH3 zGJ*^-=#{7I>+9>-%94yvUWAl5B9b@pKmIg)$wUFOi*dc3Y=bc*YFNqVUINNt^im6( zeVc0wApG~2{5GMtwa2xGGj%Hmnq}bBd?e6By6q$E zQlqn^`V-hd4JRSsyG*&W6^5_fM- z-x~%`btpmTIS1zC#a!%KD^fhfm7QDr>zXv5U`M0PY&i@C8-dfu@k8;AzrN2zA59?| zKDZ#vV&{-dIE$UrP>kT!4YKx%$j{732l(p29TEqJN>lE*JD%jteBeX0FXveFs@Q0Q0@v7$So?@?T zOE^9y@W6x_Q9RNc6eK!xnF<0R^d5UPC=Ln?;>q9?Uhz|ODL&*cNP@D-eml9n`|?zy zB)@01^xsK|holb^+tCmaF%(FIzrA8%#Nb63s7dCsnqCjbOVl>VASI?WoWZ9?M@8j4 z!J_5BhqXSkzUIKUFoPC?2_V^u%cdO;YbJ-Z%_*C7c65~cv~JvywBGJ99NNa@(kI-= zBzRf4%v7)$It?epmruEv0QEkXU&oIhKgjheE@F$+K-_V*sAcqN7((u3(GR za1YheHQ=q85=Yo2w6NJr3ZQ}@A6TT#IH>lyv$GQuQa+?*)$(wu<*LyDZ%#ME(bV`D zcYf;9AVj=hP~LEsP+^FG_G)u$Bj>3wJ22w0{5+6Y%?7cL^$=r2jVL}SiWM}E;8JLY z+=|w`cz#7Qy=ov7hziUXt=`K@ziR==r?2i%!(juf{@a$TxH#N$7mv-f`|l83?o00O zq4b7C77%;gC32Vy)*;i@ylO=M)lj(YbqK0Kh)wj<<0=e$_E_xJO|Xf%%sK5hrPO%T z4nFf_tUSZV$7lP|2sv!`E%eaCuoQDEnfbdlFc8D!@ObaS#s+^z={0*~hh#^j1)HFT zYsv&3{x|k25mTs*fhsEtqMnILm{F>rGoHVEnt+CTq1){rI?-n5y>U90Fg-#g8W7xS z)S}{*mpqzXfNlPi$8*`eUdNTWy;nt3a~^o>msl0gJT+C36ydDA7zMdgmK1qD4E?v3BDjr(J=UQ|UsU2nWpc_WH0U}4LaE(WkvdTmw?HL&z{tlnI> zEN-Yk1m2{=GiAEEEtU;sWyhcrDwu!(1&_In?62f%I|$JUipaMObD^u}zj!H8wSCe! zw4EMoJ%-l-Y7QnNBy9HLmU0Je6wjxay4H*8R$R0u$&~~iyTlYBpy7;oqR$-WJBV|A zfBk+^>J}%oa9s$lI7TMN-FzVF$2%PiDYXN+#RkxD#gFW~Y|Ca5i`bv1gNo{F1X4I6 zm-X=AkRu$%sy>K;5=eU5d=o_^q@2U#=jP`TrqKEs!vh|Z->*0xk9l;RU~aq)pLrXb zECyHExn(9J(btRiUgw{AGIa*!JIT-oz*+@^oyozRY2KV2m=&EE=Dp-cd&O$o%}5nN zFfqoikxj^J=;s_kgy-K}qSab5Pm)FkrBouxVI)Xm^s+~RcLGc(6z_Cf4RZ;Xc*pw- z48+hmJRNy-btiDU%=c}u)D5y*YYc`%X&LaxUe4>g>i4lzQRrn@gP_RRudiG@;md4| zg_>nGE*B><3VA3li?NA~^VzmM=m*fru1AkT!Afn8XbiD`J0O}AJp}|3G;MkWc=P4O zcwqK^0Whl(Gr6OEY;D=sO*8-yA3+fj&JGSm+uB1x?^ziD54N?AdmtaEM=;DhY^={3 z4g%Up3P8K{m5*us1L@^(ut6UP=BR(@#L-k6UG95%tWi-?fnYlBLDmJ(?0^!%1jLjJ zT+(mwxi_w@l%x^EfWllV#uBM~om$Tk7G(U!46~b?n!?gJ{?j$V5OZ}>#4zfM5EQ&; zB)2FmTlldzY*yS_cklmxmW|3Ie>i+as&BI({*t$ zp`12lQX<`k^S_C^F`UGIt`}IW$l#tN+I!`ogG;UqJ6ryioA2wyt$bqcDzK;qs2l2i z+!5y&6BA>p)3f>CpZR0DMPd;LvK;{9#4~G$1c%R+>-!26jc7)D(Lo23E}tFhsy%+}nDey9;gm@L@z?OfQQwf$J5Q2HrJ93ltXKnp%KNc5tD zM*%*>7XR0aP-%8=^({i*!Mp!O?rlP8jvXQj4(L4R1+R%#au61L<%p{H(a;|9t0wBLAEQ z7HrsuRmk`$AW8lV@^CZ{cYWH%|1TMEHt;M?7B)`u+1irb^Lw5x-GL$^3uy9;4NB+9 zDOXbB$!36JCV+i{7qGa@XS20l{t+g=N(@K~RH6dEs$_vVZwLcr`Cy5nRKOPU`kmy! zdU66TryDH6OA!)4xMb)=3r{~*Y$nsdAD~o&K8)HQG<=|Oj!(E~^k6C&++;VD@o!il zcLk=07(>J-;JJD_SgG9`TA#?$jyG^=|t zI_duhbQ1lbHPK}dvo}t%#LNJ`37_TUU~k@Kh;AXD)MN#`*Z){#K2=!0dVRW~Rcn1( zxXoobTkg&p^xiny3s92D_-wpLQvPkgB;xk`Fgk||9Au5o$7x3Fhj3TNE6%pI>bFZu zDafT{VT~z8MMYoP(T0W(`FSj7M>EA1-{b zq^>M9YOj+cNH*cCSBv!;hR<2_EAJ`w42^uVCSuHPjJ)6me#4F-pb@u~8_OrZ7Z*3x zx&}pMmP@rjSR5WDPgyfUl-2O{Mzi7XR;0abM_lA6h8S}qn-hYLaLJ94{C*v;$2^9f zT(k|`e=#GDFcdRV{RcA=l@XlKC=ob`d_Zqeo!?fBmW6#a)=%?JM*-GFT<`j1&1%{D zISa~G`v6n{)?6$gn!a!FpHnz9JJup@kemz=aX?(LytVadbWMzl5#snX7DmW1-!w(r zql3pT;C@{mSptxz#A70f0b>#9bqtI<5@|Bh;HX8(Sy?~YR=Zvv?0#C5?kse1*cJh) zqgNZQ8oeAZbQ#bVAPH`;5*MisHs}suHTU|pC0i{YjJEhf?9cya5_3Tbrt~VVez`nbu31U4%9_5}s;8-Fo-`kc^6Whme9KghP>j%tz&bN~ zwnl*@75~oBu{ylQTu1o{$oqKmS+vE>=Jw$}4>W!cCmdKSo+&u#O+nOKXBb?nGKrvk z;HqLtX~w`N<}kPp>Q`ZC{YshqqA&#)IqW0hHCyvalZ@ z7v$Z(ek3bMPO!N95kY&kJCO$fjNsk*pp&uE;IrfgjamDZ*+1%ZKEqM+wtpQQmn`t% z;oyoN&A8GKS#;|E==5m|1VRLuX|!1@G$uFYW?MY?nL$gZxV3oD9XHTFprO8=;9GEv zGO{iugFhC(GFj38cmc@nlza-Xe`umkk}W7#($okK4jv8%f`1ik=9!>nm(A5EP@*`J zAk>)?CjH9}OuB{w+OjEWa2_Zye$pg@(Kn0!J!vkvjviVN`ikBjP=$U>NFpL4c%L;; zE!33`8FTp%(+y+37AR&1M1D|G`tiqfjRdc$EN6xJcidtF-4=lG3GP9!R)-1GJ~W~q zo|r&fck{%{IvJ@1;`Enc-AhoNWp7or>ivU4u(5~=KW_GqRWj)j6JXRowLQ+zuH^oM ze7luk#JJ$Qz>1dB4<0=GnN4$sPM*ED;cd8BZ^vF(UR@Ld6B*QtXtrR=xxa@n6Dx+( zXg9VI77#{A=-XCLkX;i*kPTZTq}eiuHLdTQ0VsOvd{yAZlxlBgIUIMO^#yX(M)b|e zqZ#>4`U3q#No?7Ni9s48oir<0f&nG>W`or5ZudFMo$O!FvTJOszf+UATwW(O`wDU* zCIqXswU*s2m1d-Vje~?zC}6k6%!|H9ra&{^1ISk8^INY5EFSAcujRS9xe;BC5V>Kn zxZ^&8#qC&aj}#f7jO`t;hBnmqIkXnHl@d?8b%y?Gb@}-qfrwNX;QN}tc2uGl7Z*!N za!7=JBEnT33o}JFDC5Wb+BGR+cQo;XM)qW~-eckG)V1u9P{$7sIg!C>IC{Cz4h#%f z3e9mLP452c7HvMhTtT{+$H6!^)Xh@et-t1SM2l`G7+-5EIV8k_vTJ4$s*Lr9dR~&g zBw&4ZM^g<_0JW(>o4lTn_rR-mO?Cztlne@d;?>+mwnRj?VLE)Dbdqy-Shr$4oD!@X z2U?t&@oVzlX_^s>$uk4i;Xm6J*m5;6-u&6&l2Exys(*r-Tpo5hRaug5C&!bj_3qmb;$@qS`pz~ZKVOEgKIn5TWf z#~n}z1~efeZISGMy&L@@vhz2a2NpT0FS#11y>8m>L3jiWxgw!QDh6}aFazu?3gUZx zwy9RSr|#}<14j$sr*n(aV{r!}{NSh8I^MC&gk+}GESLt_NH-QK&2D$h2!32K9TXtx z(j}V1i4>GPrk{bNH5)<5Z#>#2q9df^TFj(Xdr)IB(*?KD6VcL4|dAX;Z@nb|U9LzM-4`6dPDw)SxLPA0z zm79x;>tcYV6$cx!1EiGUL{hY9Xr&s`d|)_J(Fw-Y)gjIm(>RQ=Ugd;K0EBz0Up_yw z*Ekx|F8oB7A=Mlo;`v%31+hnN6cL716?CY^!;71HbmD!~)S$n2FUFrS-J){G1Ns

-LX;;_8{bTPdT~Y4MXdYcMPL_6a3XF0rufID41W~7%Y&&i5L0Q@ zTGD+)lwIF?#JqwvpUqg~w_0Jtcp;HPNl6J}l%#|N^?8+x)i>X^fu7Ig=+&vMW%W_w zr*)~M1SJecAYhRoaWR#K>*8NX;dLlJtW5hz8>ZVnCVH!wEwP5B?`9olTB6;ES7ly9 zg$Ap}+zPU3pDCUCe&x^@ycq&f1vv}{jn~vYqDce|gR!~iazeO~CV{J{#=LT`yUDJjKgjIq?@T(Fs$nb)@m^9>A5iuDw= z`bJbg)*5{kS6&KTKdW3d)o%UvP&Fb%4BBlC%h})5j3>U1ZJ+>XZ*g>^1`IcKBouVb z^dTyLE)RR5UsE1y((HQS_u9A^v+Q$224NM*N^JvHa~1j7`8|8&jK_pR zmT$Xr#dF>){<4EQ!x0#ofdo)(E@qt#Uz%eRNL{S|d@IhXTVsxE%wcfqYHUH#C9p;Q z#iFP1&iVNHaf8DStIa45f3Ll$lPHQ&lQw0{_|T-686Vy^0%6htMf5QGNkT~ zr49;f-vsG5E@x(|vakVYPuTxD*Zl{T-&t&q;{qG2zZi0Xo|mz0Oy7sLtF9H!;; zA$4D=Q3Wau`-gz@Pr3e4ANK1g(x(nKF@v{KGiBQNkCTQI6oXV>FCFQ2DSyfBA?Ojh zK0Q7tTj!D!Eyp{wcKHsBwBDoIAnT|O))R!w5^-AwSwQ*Ni&t+`T-QQ4*>ZER2j%PP_G^ohrIH2R$3?EN3Up5{aBLGOB=PVPUSK` zh{@zq%Wna;CP0UrEJmymLGYmwW1~`z=C~S4YsXux;?54=5K>%<7)EWZsL*6iD_D)s zEa)27=Z6dUI@kh5pqq7P_cDUa%MTGC5&NCY(a#qSKVLoA;3>!B##8){&!9$2PKd{4>3D7$E(^@oUZ4fEmb0LBzBF$eb3pt7JDAxSsMBIg zI0nB~St!i1*D&3t4Cq|)HyrZ0gXLNU5GITqb$kZ@y{01v14<5jR%Lt!0xtR}_mEI5 zu1Kke$nSBJT0xh9ut?ws3&wNG#|&9Nc{o2)(I=iJrKA6i$Y{`3_>Hk_t5%3|0^b{m zhA2DRDl^-{aW5Hm zZ|6YfXnl0-RK*hYqAJUMeKr^YjYzNk`Q<=I_eVae-MQKP?$HxzX~mM=bxA%beCM;$ zca$X+_V~(r^W|5pxeJPWNi}M_W+HRF%GcUjE{|p_+}&FGaMp7pN9+U&B!WOFS!DZ# z`QJuE2;=9x(&{NJYLGyAscGylqx>`?{Dc?)58sT3ilID^@v@YxxS|;q_S3tSqouF2 zUDC}G($dl$nZx=;=jV0=z=( z`JeAsuXn4SXSEi8`u$rX@p83q;wQt)P@GSn=yZu%RAu^0))l^o?s%G&x$CDsE^2a> z{(cFdSeu2g_BdB`+&@Zc4l7a`;+zBXK>&z*Bsg5vuBL7C%XLs(52XkS3JONm3Y6Yw zp8`DpOuj02$kDp;Q{@-w6pz1~!QCqB6slqfdCFZgN?nwxp`>4kKDAX$Qt8fliuICPecB{1dSd?w9RGfa*k0t186L!#xuMf| zaN5fplP)dPSsIkdX|>$<#O%-fj=_<7w+@h`bd$F!DPYm&u%WliP*fZ*i}?KGy1PK8 zSw+K15|;(}3phAa&P$*N$Q(-`ORyK*s;g-gVX<5Sx@q^o(Bu zu77tRXv|Z*sMt6xn8*WN(u;H~KYhYMPv*4#`*?R;wEuo?Vgj4JRuqpb%l2;D>Oa|# zB;YNHvfw4!J*}hua=JkFHCqn3p63@tM3^!Xdoe1SmuCrWd$>hAz&237!|xVtZ^v$I zTjle8?V;L{pu4dZgIxZ50wlCMl?`>|oYZjnWW1r`5>~Q#Zn8s#*)5f!3${@eC2?7I?J;1TNh?;r|WmeEid-kyx!TlOH&^o?lg1QRm9RK~j4L|NK)#W4jnT5VUbUuwe=S=w{+RCG-NS#8^Kq!O2Be5xK6p`Sx{RdI z{CM{r@h4rmw8eo*MV0X|vJFmJ3FFtJ#hcyH4|yMwh4k9I1`%K?V=1srLcJm)Mz-sR z)C*!uiUyf0z!b~>)Ew!0B2bYBYSkh@>&ZisgUo|#{hq-+twx>0P&H2C)teIo2ajjp! zhvG9W1C5{q&ci*&lE^;LCxcV5@v>$<2@LgqpCo^CzKB;BIGch4eacP8DD94Ny6F;M z=`=v#Sy)E>0SFn&?YEWu3*=FfIWk`p#_~S;!NL2@6dm5*Q>gLa|&v*WL&krkY5Qqs3wk&+eY#XOFs!u=N*`VhaQzU%Y+BHr5pf#q% zi!G(>et+5%8qat%(fFZEr@2x#%Vwnoz&5Wkf5|Bg2VQV-3XY4IYB75hRnX2Knl)Q!4^%c;4|m4Ypw5uO?t6vk;?Dz%Wg>%K6V5-PZ<84N+hcP zdJ&7U3cdc*Q8E+BczT;iB^&wIeAVL`86i7Z8C{kSjgC`;onfp5ng#4K0 zRJeg50y@E4$cdhxUzEn-P}wh_&Mq4=tw2Q@Oy$Sdnbs;hy^Z&5-mVZ%3UE45>#L(R>A<`vChcrlcH%Kbo9nwgu2nf>M-7O&^ z-7Os=ozEVx>-XHxhxb2t?=Sp@VVHBDd+&3twbwe1Vv`4~$Ab%*Ev`i;JryJMV5 zZ*RKkhaVLU&aR1xnDxn+B_p&kO2q>X&U9rpnKU(;EJ&8M>+Nd>d%H8-rw?>JXE@%j zc7pI$&*BAZzM5+-Oa60R9B$f^^<6ZU-|ax4u6{{LNy$j0A_uZ+4Ndx*V87NKHXoU- z`?!6R!mK^Nu;2ltdBsDJx_g=I6MX>2Mw|V#i~wkjx_Mom(n`muAolT^oTmJ!E*yam z%8`l&c6JTJ5wrZoH-&1>oD}Uk387|S$^ss)x93=~;kP)QD#6i9*I&IdiAyUDAk#W@ zs>}qkJm>4*?5q9Snot%UEyp0;DeHlyYJx94m5}VAB1VA8$1_szqtX3ZUmPGdI zL2i!IynhnC^o^}uSy|~_o4kWkyVvF{=Sw}3{cjqUBEF~gr;EpSQ>C+C-4pz7T+m-_ z*$?gv)$L2Otai@zk9$E?YNd!&@Pauw3w^j621HM(y#lZ zICH8rSw|)RrjtfdTZc--KgZKlO10qWw0xuUT_H4vgpUEpwE|Yadt4sh-nVx*SY*7V zh7s<$w67svBDy-CiERi_Hfe&|S}u4|oix6hz?U9WP#GBCV%bGcjp}yz3%hF_Uir(U zak1@r51vdTosof5fv4rec}_)7chBO|(z9MAP;Mu^@TA~#{NnA6K;rahFK~mxC+UgF6ZRu}F0gJe)uJ%lP^+k24FDO|d!nYH) z$CL68IK*)ld zOEu4H?~WVUwd`Xkmka7Us0iaN6^#`m(&~=JIwV&1{7?I9_OS>|Sr=9mUWbi-!c2{v z{wSL;%g2~24s~Q%Vb$z27JhpI*L_{hnN{=q$&SI+3ZTnUJOWzU#fj&D&a5g8r*ikPo;Tbg`N zx5f&FikaQ1dKoR|@Ls3v9g3NLg+ZP3_32h@`itq8eXk~P#6aH!HYg@$z5M|$`8bzh zdhs^PQB90)@ZK652Vv|9_xaxRuzd&nx#HO{K*mWx4$Jz_jES?bE2@NRAI_{i2&s%n#JhpTUGsm^It;mu`=c(i|R5aexbcE7)w5p{ZVskISxbaXLx zZ2q_83x+k#Sdz7pId*Z30**{Xu=^Y3GertXPxRi}x9^x1YhG=PY1`}l**A{+!6wRb z?eDts0|$c)Up4@V8zBIgwUSiU^w#m4s`KH)CXVOh0YoYMC*U*NuJ6SmKOM#%E{>%6wb_1V`JMe9>S2PD#c&Xb!+4SG7vb|zA>PF=`0gh zm^%4435I;fv`YiWC>ZknYb%|~H))LQm$6QS^-w|gyIFwgulW-2F|Q74&7O;MVfKSd z81ExaQ2KMeKh1|Ai^)(;LYXWRc1k=RoaOR+ksol(N)N2rt|=o71?vZ>)#9FEpdwL; zzQd)}o4U8{j0{fCwch)hK>2;y>8~`2KH~BjyUwZM%nu5^wm%FbjL8q^1))5ZB-fqC z&Wb*IJcpbQ)U!nle5c>y10DKhj8QX(U0{VxmrWLxegp>P%$RxPDOB5)>^1=;(2E^_ z!;RW~bq%l8l+>ZOC4|q_ByUS23I>=hKSSVA8XOm*UPpm8*G$@){7$4)rlkj)-;+qa zDL*F%Dt5P|Cub*f%iN;=?xMimm|SUcLq>asO~X;>sB#o5`gRqLt7}!qy}=rKg4NmE zw@60aMPBLiA~jp9SOdk5n%*XpGKo^y8`jF+l+EUkSznc!UFRx%R_v|f7=05zVnJSx z4?oC0)%f3>eIn9&?eUD_#%1Pv_I-0Qki^B{dg3&La*yFZ4e_3wu#M^wIsQ7$H~Y zEoO?tHLvkW9_Ir*{C2~Ur{yOLx)DWbnfW^XI`5i8K?ly%uvlbza|#SEP_H9w0hm3U zAel?5`^p1k{Ao}fkDYCw)q>tggt?Zk%f6QKWtc)P`HLui*|Cc%NhM(+;i!Yu&0MWp8q-6O~YbDhl&7 zq-A+6n)lthq-qiw{F9kRA>Um7E-{d;`v%l~{?|)(|m?I={lB6oqfLhCxS?-fWeuDSm}>Cr57n5Io7z zOHT{0@HE7Q->kZ)eODiOrXKoUsyVofD5oNTK7Ubd=h)~^p4&|q@`K|kp;PVPDH8*% zJ5%i`q99rdZ4jt4qD(o?*Y4V#E2~4ttox#g0$yijgnXctHY*n`zuL|UnADE6NR{!v z873hGmGkBI2o|`@hN7IUeHHo@4chU;#86gm6Z$NoAQKZ)wVcH=4^?a0-jq;-v8ea% z-l*pOgMk0}GLgFZL0K|!U@%(l1QNotR`>mj%qPgqnq^vZ-%x$&EcNtm{uYEC9Im>W z!l8=bA(bj+b_62oUGWzgG;wuvi$@mzX(aT@?C`$N5E-XMa(y$w$YO3H-`Fd#yZOm~ z+-2E|c4T6}dNbLp<&s<|A$jvav&-lWnok11A8Q8dQ>MPSO_iVURmeV<`>dHc)PPf5 zOE=5N^PLN2Rb0JL2S7Hb?0_b!clN!Mdzqj`bI6YKZQKUBTK*o+t)VO9y^^qB_rZ@`Id}_ngWV?LyPx{PVwe%x5 zwtKVBpLN94*HKBHnq&=TG_e%Fb=od)XicfFM(v()@p?A8Y-74x5BarX#D2*OX!P#e zyZ*av{!1hEQaEzbvTu*TZ{yZXDyt^iWjS4A>KiWQcTQt+f5nyf7C}a1B(<>-+*You zqZW2leCBuSi(yrG&jMPkUQJXAur7P5XLmvgxvB&_>2{^}LJoZ{Hx4sS=O(B+ZtG{G zBR;-ov}(K@R4rX{)^5`YH4ir(jOanqnAP^ZjxAZ^nEr8QbYe z6<=>M{%Ii0rEJ5?eJ7;xFTlmagNZ}GKY2gx)G+OI*mB(7@@}Hv-|u!bQ;_VrGNq{Q zQj0(+=m<+qTy#@MeAZ7>FICTcH7`iYh+}gWu=L>Z`-}wrTaocoWD>IWXZCWka&nyC zhJJ7KytDN1VEG$%`|Fz@vog`M2%>mw4ntwFFr@ytT_v^0{Vbdm;f#Zd=NNjyWf1pTQ@yBm~#QjWaZ`M?>is5kg2{Ff}Y+HQsI_b48Kh5 zBo4S1+c)g0Py=*sUak21g_X2FuKAS0?*jqr98q(C;Y@mF_%c9Tum`NdtLtT-<2PA3 z{jUE76IUJh)hZ%$)O=w#Ze;1tDgY?o zbN9DQk0NM#v6seVhSzqz&o(9$;Bk!@5e<^A>+hZclIB)BgS0tj_q zsB0A4eCUp5!0;-rEH{{%;53dDm;L09B z3GbLTwp-J2ZoBjXt@@LJvrUsnF5lMfY?CfaQQt7Xs)RSe7rmT4kU@pLs9!p5{ATJ)v887 zq5AlW(LTfDY)AF{-5ZNK_ffsxvMx!_j9=g3$$?I*bjy>((?F79&EH`*?T+)5iw0+InpW0-p#R7EI9)y0>h;(EdW*=uVX9P@+buv*F(mKWb3^3KI^BUI3)6GJg9F%ioXv+P zw;L5P3ZDQr4Ju=rqlP|0NCUrSvnC}CFl9k$CIXz~tTFOo(xK~#=le6t8O5JvS$x&X z3tI1gEId*?U&mYj^Ncbo*dDo1yHtlJ)MLy!HGfe42gHDMAtp>sS0XSiq%!EcHM{TK zE`9@3UX|Ezxe2_~Jgw*JlcEIVYTW*oQk+`sIg$gyuHA0>J*2y|_+MHtWc#sDbf`fZ zn{c+UG_e=OB(jwTXLiaiXy&5D{ji(+Q$z%^``FKGNDC@B35Lx6 zU>EId9l4WSubRP7V{Y4n22*AZ9K?!fdC5Qz!)W@Gg7yu1hNnVyq2AE>$|Q0a1!~_R zf3Bxzg_MnK|Cd)O6_#vM*%@>R9iWu5#U7BV1PRBf)H3-Ko}Rt{j)LzwQkQ;qz8@vP zBMTYj1jXWL)fYWc7EIp&AKv3z_T&(h=HY9P>s(@+&e?gnouK-uIRW$zF;Nx7F2OsS$xAiDGdf{Sg)TQyeKacyrS~$A(JX11WS|yWc z2IVIPNXc5?^h!tWRoqU9?VLXhPs|$D=>2fK7@JR9j#Fqg*{VBjD(T z=l*!5cn`oP&~C1R%6S}>VCJpRR#FITlxg{02pvt_`Lw-#Xw2!#0_`sAz2CRTGJybR zLrs14C^Wlkjm71S{QPr7u-9{cqSgihf>dByV>pAR!oUl3TaHt4P}%<`*5Q6VA3^2M zC>f5;bpW~(n@Ne40~DY%j?OB`>$Q|E)Jw#F&B%&x3M8pPR2cUj+LG}`jHD%k9qTUjnRDXWq;C^(# z&tf#+;q(Y2jl^P@B)ZeZJXg<8ey;v|^1FIGrAYU3*KDrOc6rAA&1Fd1RIBb^kO1G- zmf{)G>_gXJM7TbKoB*7A+)gf-D$zDekoO9wGnr_g{!6CE8!Ij#oyD6Rhx z|KU(9(gA#-4Ww}A+7pTHWZ+uHXm-Lp+>QZYrb*j5q8Y&201gjj+itq_+PFLIabR`^ zbd+EMTGwAy;Dw@{77L@0-@5>01OQDZUHs@EO-nyJd04uQCS+41_)z}R?}i7xr@iOe zE>UjZE@uR1*>Cta>-%ssz`!mOqwbmM)=Axq2S}};1Kicgt^-_iZy1n%kP>XEF-rEb zNuBOm+jFKuoO=U5mtfH3OdLg~udcPEUkZq9V@UT@dXQi6b{=$roVD#SlqG+JYz<&z zg?-K?(PMIRbCaGaJ1*3-wwTKP?N^FS>pnTcb)x<(1EGThM`$y?f0&tx|54+}XYF@( z8R+ko*aF@3a*Y&dI21#WFeXz-PXTy_Rj+n7z$Cy#Y!&1rHD^(9CQq+Jh$;7{%X_3Z zXMaN9>< zXJH#RscivzB}seyiCx}_STTaT<4PnPH-&(!Nv(7dFU}#*Spa;K3mB9LmxD71`I6FP z>Y-_3ThSn{&+7Q#Af|b4cD8WlEiggQc=%W)8Q@P%k4u5h!r|{dLMkfb`MTPhx@6Ws zIKn1Vs*BA{3$#MC5*U%92%%AWSWQ3m99$v+H3(?0NJ~nOWQiIxA$>}xc>S3t=sVCv z$XOf95&|+hqL?@+ub`1MnZXGULTw<8J1c_bY;Rwh-jeQlvCy~Djf{u&{3AQy6FS5# zlm>)lUQfrR!eow+fYDDKk1K_#TDhVmvm1+e+q(h~S)N@^E$b^DG~!0ucgT>agO!@c zF}DCDWMtZQ#N=YXLL;&rwJ8MCl{8S+q<@Z2<@Q*p`Bw|jhz&q`8uL&<@gs-0uw!;Y z_~U0<_^UBcP5FXj8|NUCE(e%bfQ1J_XLYcN?`Q&7H_XlZc^#);21M~as?nxP{>b1a zI{*0RI%eL@)?CRXo!0?S{t#lpvz@&E9MoG-TCmwGk^xRZILPbR-GtFKVk&sb? zy3kRHSgTy`I!QcB2OmvKI=d2}PUXXTV8 z0yEX&H<+_eUQd4A9bntwKg>H8%rzyyk)z&50Ptgi0UhVeXnupp)u`Vl3=Xvj5JHk= z;n5-QfwkP=-7&OW(FWCAeF?!&4iu}#&mGW&RTx)qK0DCccs{@mhmQ_oXs6+D^%LLI zgLP0*Y_P>VF9GU+mQETliFlNBf2}ose3z<0F)g(Y3!5Jv9#U)LryqLc{JtfgScHpW zev{2cZdc*gBiBooQ2$th0V+dvTQ(l`(3oh-w6e3r*LS0QD+Z08G(s%oRNqPEdU&AF zwWSDjN-np4jaYR;m-ixXNZh)F#)i|_Z?CR&q}^B=38BYvACq*vtvTI#(+1_UYqgmE zIQ3`sO^(DcCze-sb9^Jw4$R-?gJ;geolv<3f{z?nkdOE8v-z)L&HH0Hv7+DPNxVD% zH3fRC(LCjUf+gP}$m8_OcOyR*pr_mV$cjIn{uhLsAQp%6^Ys{~fYCpGJQ6HK2;q>U z98A1uX20OvEZPT5;p;}W*8 zOjR}mzzAk=+r@=V4oQ@B0c6BK+MEtPEX11nQi=|qHUtzf3u8c9LVUp{QI?ulfcILm z+QZFBoExBXfw;-KE~*+Wpoaj~=sDSB+Go?+f8ybiz->$7|EC);-9OL;IZ6-Ro&_@2i0I4x_UdwgOfo!zP(Oe*Ba;{HTfMFc>AGwSgoPwS}v0!lF z&OsRQX;3USJ)`dFix&hweVD`?V#4Nt)B`GI*yiSiSvf3-z}2vYX??{2M&kcc*nVvUpm%@( z8>xvNM4R0(Ha2dl)<5Y@wCNEY^)pzDWyRUVgv7+B1@blJ;A2+O;2HY#f_M~|N}y!* zA4;TnC=aL|ASDG_pa6JAV!oo?y}R9aJ=q@EWCBv)EV3-MHp0d6+~ay5n#J1o_V((| z30DPA-&~C~jlQm~X0h6HZNw9^zt2`THfs1CxwULR1!(F#fq777q6&Z#?gxF4z#d`Z z9mj;NfoBF#vNGE4qQxD=3<3{b-Tq`bl>t>iI&a%&V)xU{@5Py#f3|^MitLzDoT6PM z0(rDRxOC{Rt@_@2%Z$_5&L{%rr~l!s!uBcbD+i7$P@CTO)a5|;3ka1Iqz#jC5TBS~@T^&zD%QHWtAJ2HyrFfuEL0K5_f zkARvL<1+{^DyTyO$S^P~OVyUqIxZtLEC!7nnD68@)Dl?#Tix~Y^aPDDnUx^YFqWaT z%n&+^Z>a>ZCa~9*GGC2At3NhDTo0juQ-e}pXJ@WW5Q9aNqE^$dZfy|!r&u|MYtkpz zxdU739a*p;5(38o&YKH13_(bp?G)~w@6Jz}s#AclP0+Nfz4{;w>|&Z95jETf9VG(ssmh&t_QzqEfYwek?am9R*82Iu2e*wJkCX*Gh@mSW*P=q7aT zgZa95qZEP#YcMs5G_#VQ3^laa*x1ywt>9 zXI=Jdj_6F-5U%>c={L$<{JC(nI(jg5O{+-i0-->6AU&MtPS8V+z?EA;Ikxc%i%7STTt1N$M#U>_xpTqeE*5(e?zv{GvgieO24_?@)#6YTdGk~&n0C}?;~$%54VZ$RhJLs^ z@Gh-@_r6bnCQoPbGAN=5E1r&mgvV25f!Co#t9-s%Q^BAG(1`%qF5J31nTrPOT&_1Z zTHNd~p0m*}D_yXg_-VmPS}%I0ETZJ*<;7BP*WddbM&my}oz{0^HL4|^=)p!*I5N-) zFa)z>NJOv)vl3vQw68p2l<7<`RhL3tBN8leaKWh31Je2ibviUjun9~uvCKb(QIAg@ z1Ew>9pd>H{Bi}M?=)u1KzcPJnAbLDz`)~C46d3KN|835U3p28~r6U^4u$K^lyEFxWXlPxf&+}DBX$aza5wLo~X^A2UP#e@A{ zjGAz*9rl76z6jf9CY$0R*9_wj*fMH(LlAiG=15Td^E&ZjORLRqhIlr%2EKp`;tMt$ zFIJkq0(?cG5+7XTe?w0KZb}~TeZfgCa0}b>?QvJ?sQ!-(5o>-;v)8iJ9VkFfPMCq0 zfuEnh1LHrgE#Wzs(@!~TfISba-v4vfSz&DA!~<_)cj`<7=o@IZ`Ck=z7GN2Z?I(V6 zfIp(q1a=;QSRo`Ngf&;B6bVH(FfhG^>;Xvky@uUFkRp&XVOHl60;fqah&0Wr-m;th z$ZY>8DH%PV74U~}U=@Ej!cGZ(5MEfc*PfDJk3Ia1axp-q*tH1Fnw?z4bGE6`B`6pX zL>q0l$4gw6!v;;BTdNNtN>X3QQTgc{6zre zGy$Q^r5d03j55OyIZ~KMf}S^%Z!V*biagXj4j1!On4$>S-Oo@g8K%ELjnF z1_+WkEcayf4>1qWEcVAwXa~+T0%R~y|P8j79u&a{Ow8m3t)@{@b zjMY0WJZ%0+N@QbSP~}m!=4)}e#hh2&v63Q>eu?icCTP^*ZfMOH9UUzgu;vR|AP3tb zW>buhDrE?1W}AfpKFzWGNOO7$tNL{R4=l|TK?B8A<&a907$mHKQTHCr* z{YFu9U{#>}MM)r$uKwE(NB7IPJ@@RI^fjx~3w@^zeez$_phe1VtD()uXRh39?lkV; z#|KM3JVzo9u^p@KLx}*smbVWON22Z}#a6>F0w1LK)H@6<8q(L?-wD!3JLdm&HoUO8C{hOA#ErsHRu(y18PDIwqpj~F@LYOD)ySp32l=&te zOLG4@x`X`leTTev^QZob#2=2STd<$`kJ@r==_noYq@thqmK2LL)@6Q9O%ZhbxEjYl zvbqUKadodLNvqg<)ciF!T79R9+w$pc3${N`pJV<*Np217Kfx#eePyO&{P5E1^w{0Ea+>Lb2>X zQh4n4yLAOdZ5Dm_rdB&I4?Yhl(5sD9dXdbIBum=00@c)(_p~35uiC_2pqN4WEID5-_h`LsgA}W5}H^NShjEF!`%M4x1 z!gv(pcN}>+`1)zOnDj&^MhFK`VPX`l-G?K=s>DfjqIdLjE?Fp(L*nx1Q zD#7R$DiCJqO2DmfvI&$#NSpuyeZ_CNjz36^rTZygQeRB0MJ$pxIOZaF3=WBJ^h zMd08s5g6YE1jb0?H|rK$3W`Sxq9iw#tS&>u8KD{QTt}j^Cf_Q%zD67=z`@ZE%1MfA zP@q7LlvLK05J@~(U74Vb@#2v_q!x=SZ_`2o%c1l~{s|%bNUy>JeP|IDpV2hfE~9b@ z@KT1+ob8RH4S`facnlwfXAdO%zU45xmk5Dt`AW#^3PEDyeZNY7Ol%O~^fC;RF4><) zO{t_Z>z60`&_l1hpXSYuMhOsxoq|-Q1pk{mxe=rG`k(hQKs*kF;Pq^Jzh9L zK^-6%moH#&$Lfkdkot~k^fM=#bA}$^NFGx6KH|kEcUR#08&Gc;K3Q%T-yXaF|mopzx!p^q+~Vvl^&KDZmzD=v4|NU*;f3 z`6>5Ple*A@`{lxphKm2>Ic=__bx2GcvM@q+I*qV}tLmCIN(SNc@t|zU=LzFgs%!{r z+HdAOhN*jav(vlua`F9BY_Df<{ZmfPgK*s><$cSjza3rWh+~6y5JO|wX?4UqXYihf zIfhr1ft@QAuI70K?=>h}bbk_ZflKNCGg$WP-!WV|=HUpc?DSpVE1c#}4-r(0KBjW1 zSfY(d`@UD=h}Yu`AVCtA@(b0;0pL3P?5-rr4T_v~bcm5Oe=dAQhoV>3SF;(H3t7Zu ztrAeDE>D;1lUW0~E<7BZpln=5z~LcV6gy{814J~&Kc4{uY7_+nt#8guDthe{)Qu!| zx3_0q*YYNf?B{fmQKK72BwQhPkyooP3oE zn-B-7>-%?K(9r};X7BTTab3WYXXMOBumrV2 zQXYFa`sI0v$Tix*hZpy>q%zcFPo<)$1LBz|6a0LA6FQfdm$Sb<3AHtvN~r#9{elio zS$`qM{p_#Q5Fm+0|E9?c2ozQ9RwH5LRE$92U@ug{mS@+mXGtdJkkkT5_!5xAvF02M zDdFISHffdVNeG&~g_8%M*J4noGGQQQd*|xfyQNvIB`$xww=Apj6%I)fEr&?+vK*6s z9%pEsKDxtxfYjf{z?It|%n_+1f$OUloSLl%_h!6fcrr0l*Otuy&^a=fWENs+1=SW$ z!&n63c~lInK3!LC{M@-wRDy8aCG@bIbQ)ip4KtZg|x`j$W%N1UC92Y(4Y*iKiQk&}(}V}v6T zyt&F1E_l>R=IiYEpbl`Bf7XlLHQczjh8@FecC1v2n?I%0#W!G8dryrPYxihoW@ZL^ zRk+p2r)2m3%n=(C{YjPfEnU25&q`~$ctI*-erI;a3Z_)-#+^sdj<{pR2rh`Y<{+Sk zx`ZxKWk%8vEe~#Q$EuK`nU5552&Hc(DOamA4~kg;rBZMs4E5iPF+^t#ip=P9HI_z8 zxsI$zmP!qpoH?+1$|~vT=yH|X1Y{!{j#a>U7l`)-`9-W^$A_X!oEew#M{2WaXYfYazselK!`r82gy2Jx4 zsHqF1u{(EOF}R~X(9ouOe0DE$G1oI2(h2=`vwzCaFfGUu^jbrippqc1oiG z?c6cEM2M-QIUR^wA6;}bw7Bo;H7FT^4HAw3X629?`k$$M%ZKJaH~O^~o2R&X+Pg^$00x6)0UeWeYn*OJ~?&&^LTND+!(cI zwCYKXAoWBCeoI{H3F&49_LedvlY`6|OGj+cy~f^3#YMrVgYCrvwdi?e`neRh&)>zjz&q-qqZVbpOy9Lmi|8;Wm1}CU zAMRxT+qxx7Z7`gzc`71@d0BChl-xYn2-RqpQcbJ- zRh77l?0t0gKP0Du549xDuqos%E8MDfooHsV=9dSD<^mY#m@Q_=W*FWkL}X-^Z{tc`U=~@%ofjU$@{> z={b1OQrCVbd%4Mzl=@)oZ&8web4q)wy`i&`MzyQv2i|evAI&X+Z@(L41Yg7YB;fjk zGWNAL_SJAaxvUcRoMs@qXr1WjdW5XPJx|H~JJSp+-u64_ag%c1P5WJK;pK3=YsCc* z+SQg!bAb}~8{L((ELC$EIfX}dQRirNgn&B)y}`xdibPZd?6*aMcc1IOR+{+78Mu2- z<{9Sd`$L?wi*j=%+erot$Tk`hQ9V!w-ITmqYIAE86qQ`H$96`qA*V4~zZXX|oyz62 zzW6_orecf1!8bq?(0f>E#YM5jIyK;bwWJGF_Yz2e7oeQ}iZ(FtHF&uYnnMDCQvZ~N zZ%UQ&zLcJ?B7vMk6PDX3iHmW>I;9{689e z%`|PG*bZW&8^@Ws7#8@uJ#6xX+a-x`VUrh{5%M%JkQTn&hJQVW6HMN!4w#U*@3U}0 zpfSYpasu`~ju`bB7{7!l(iyna2qAS_%iM;7D-8~Hn~&@)KO+SC5_ff4T*iP$8a5QP zgykhp5;E$a31A>(xMso-XsA(Oq}I+rASHER;2b2^qr%D8uGrSd_fOx=^xnC)v7g&s zqeff5V-thfe1P&T(D*yMAR!^)!&s}c5{SGUvK$z~q!L*KRrO&(L(njI0^BGR3!Xu5 zjq`_OB$zI|gtH@hdiq3M+TZgZf1Y>8$~cnDohk?1sQ2wgAqh6DpA%rz7u9qXIVnp? zNoBlzx4z_dEFYpD;AZFQdPbmaplWT-BusZQZ&zXYW7OdMck$WxUCKRPLn*kRN?em1 z0R}34!fW4Y2>qeA3a^&Pu!xJK z9A4R5NIk);5UW3Qs`!^6>y zl_RfxE)G82_hQ7Dr*6?6EcIpggbx`XUu0#1rGE)En6=h^3B{oWi~H6TGIQDFufCRl z>+iD?cGy2s=G3@`#LUrjw2il>ZR~&k^A*5A&btc`mtw~fEiFsubu+fAPT2pBu4^R_rA5%?$cA}2KNMi*ybnl!ZTrm4^1_A9!=33v>SWwvC82SXn zm$BK>$19~AJg+)B5+qZ3-3tnO%>%BFOUM2mYc$1nmV0}1VaG|rp$6dLprL{Yzy^o< zcHbmKk_SAZ2yC*{&3>+C&QF_`_FuLB=_j+CTK4k^xO!GYM8AuVzKtR}?JxWK|KLa$}R+jiVv6f~>4fVp|nn8ZdTC^iRb3u*Jaq zHDDrjnt~Jo^-l!w&|$^Qb_4{JbODh{$M-F3pR6;V3mQ#_#@Ff+2|G6L=65cclr?}y z{;2FvP6ITw08YEM+hsA56dvKTGPvuYvMmzclAzW@c`hE2&-Xh|TqvCSV=8gr*4 zv+qFWAdc8qFK|uDS>0vz#X1G;H#I(&j$ZGy&ho~A%v=zkZmo@KL>xtFL`!KZ=ph|x zOHub`gCPq;^k#>)q(30XZS&;AkH;ymt)Uc=^vWF+*cHq^?$AL@*}xS>l;OI;v-B01 zq&9K0efZ)sDgCsLW&;LQu5;AZ@3j2!BR5&P5hDC_Gj*k@`!f}3!H@Sr&MAs}^g-E= zB4uav$(Tn?ez0BnViS#Y88a1e2#%3m$W_Z&tdf!--^jf{+j8TA|V&`Eh@ zpDs7K6ucN`Vq}z+m;Z%v)PC#Ov}rjd>yF`dreJKB;?(?y!Rtd>fbfeWB48swxo~g2 zhHI^vy$_MX!h!0;>t|V;mb_X8Pg`f$zgg<(?d^4tT8Q!>d=@KjTUj;L*^FD~ck8ZO z`cUN!n2(^RW3N<$VcQp@^m{YmLLLeO{WP z8giQvm{SY{1{bH*M>B3%GVf2~bbqxt&9!bxUPqo-QG0_Rs16Q~RT~Q{;}snI6NJ*U zBhHzAVr+%MSA}=w&Oe^GfPQRTY;KDuI6cO?7eg5n$LL-aJ7Ry)WW6splQ)4*-EUDI z8A))5_NWPNKN;{xK{OO3jSBu8Y6^15RxG}@rnDYiR*yLEqoVY(Iv(~bI#LbdhErMh zJuL?wOV=LP=*7G*Pe`3|q&i`NNGreN8F;E&X~DY;Kl!eytzALb| z(Vd5dg^8TAG`|K*=tzchV;IhkZXLtbyml0~^GY>u2b$Ly*_p-<$KpXMCu|92ZoxqO0NX zpg)cJ3w#4+>am%;7{%-1zX9+5dCp9{9nb?Ecz2MqI5imC<@h%$ys{G#yQk;%ouDjF z59jru{~##p(SFCr&CUH(?O}|n0qu*BSM^(kH83kn%c?gKfgT9?iz1f5onNi}S>ELm zhJ`D5t0ZNWWX_O_!pW<2_10eQ^&Y4O0p?&yCZH)8r#x>o__}TpX0axE^IbLV zGUg(;@E~8~2RRf3*nv{&|Ib?>l#T;W_WyqLe`nx-XW;)oGf?aNGc_NC)CLfRO#N!m+Owa+ z0dPHBs4^EC9p+HsHxQ57wwFP3CYHA!0Z{??WLrQ`2Yn4&-1Z5{$kG!Ne;pAWLY(^a z5EOB*F7};{oA=A9{I9xDZKq3h8x))A-v-(}R`Ig4E7!?^xUuI+fJ(F*Vf)Q#36P&G z(gb4WU+(t5+kSvRj6n3Wb&iw!XoYGIn2d%G&Ha6VG9S_JFsKFEKXrmY!026b4$g<= z$J^U1|C`-Z)kTzFP$AwwnPMecWy?nQ^?R2o+D3D$k_r#+cC!p*;>h2;&4Gn`eo3XJ zEN+shP-5y%9S?E9OC{8r!!*={b@EAmc0y{G6aL`7v)x`R=7-A6#Hwyr9DK_xP z7BBpOkgtdpVwLbtBT@0d^VA~%H{Vw##$Md}4Z!^34sDEPf4>%8_48#ZSy8kXHlinC!6#%XT_rQW-Df7mM?@{GQvfySSu?0S};WR^G& zF(184q$!^)v^Z7hrGNfhJcN(U@e14u$%ZUG1Dd3?wEN-o)Fw0%BM2KswmiD?u~Dr* z*TUKwu*R>K+_wGaXyR|V$5F@MznirpAO^i1{aefQ8<1Z9ohd8XXj%uX8R_D8_i@cY08L4r=Y9Fbv@fB*>QpWb#soY^Npp+eB@AorrS&ZdM(U98h2D4MOw zW@N|30}IJhvE;Nx{(SiTj>5OCiaxc6WHG}pU%q6bLe zQSUX#RK+^6fd>sor!)OGJOt&hK=bP6#}=#p%~MMRMS>GAMLPgab(3aN<}SX0wK&q{ z*=x;HA=Um;I(z@%SeGL5O+5u4`=lB8(TaHLdl{I!g*FG$^{Wgd_at5k`!)T^jLT10iEqK>V5@)<>l%E?)bbLwVwA9$KFTHKEaCk~tK zD%ABsvp$6sO;hmP7B33tx7gNzl+-9iABup_jKxH;ej>rrJP`T&hPU_9_DPJ*MUB|Sjctor+bG@0c<@IL4M;i&*$D? z*pq?H5~SOak;2R1aQuJ9Ufo>E@U6Qcn)OTc@2t;>A@g6am?!u6+3fSaOP93nd+~Wj zn4@Fkt=4-!p+OD@cz`K5tw%W!l7eFs{>}jx75E2yt}_Y z{;Qrnuoa1`_~+mr$xkqIh0f2l_{ zaNOSV@U9(WMMP`E`U7Vqje*5{;{99mr>|bM>dD>6G+Gcd!%e%T~3}~ zwsTc?oTnsXGwZJ}FBjV{Q(dq7?(+WrBhkU3i|@JLS(&{2UTxKDj!%4gTdywk=c|qT zwlk$}Iva2_=;CZ%Ih%<1B|QzEpuU-uZ<1l~1K?1@Ggjp^KHb39s0F~dXZ%J_~fHi@=3-1l3AyoCU8JBrmKdD{_RE%D+WOi*Zt1dRzsl?1+$etU z6cG8~>-W5uynoD>voWscHn{ld7ssV~P^%j_isKOmJOXRk%l4nw_y1d}ALIrc%C?QI z{<;5jzpSm=$5oQkr%awS=@09d=j*fNFRxm;_x6Xm_4^h_z2Xf~Wo79ROjwk9CP0V- z)XsMhUH?5dc{ zz*1nV#^ufZ-uC}qEbhM&bu)4PzOCo0KC&|g0|yjl0_P%HLCNsK6eeJ)o_Ff*Mtzn& zsuw=(DS7)zNvE=SV^_kV7S2^~f+w`^4(UxP1>VoQJ8F&Zn{(d-jyfK6Ve~5ycAq)} zT(ZoHdhod4{#yCX)#3W!T*dw3<*lbLb-tdN%CgS>;M2zo)=%Qu1zhJJ!V9V>LG@5j z!%=d}U$!dvZ~#~chfHT)l5ze1 z!LPuKx6*e>!1LtTDrRuStQ{36@AwNi-UQ4~DUGi>-n_~AnYNNo*>a__(p4c=mP1Ys zaywq#TDqDmRJCmW3_C-It};)R$pZ5iYzt!be0=isiz#fZDB)4Ckm-KS*R#Ndy1?1{ zZ3P#dYU8`bbT5f(DhsGFeTvih?{WFkr9awVXIV+JRIKs7Kh20~o)|C}`SZ)q1y=?c zdk%2x@A&{c@bKF9ztiLER@SAj-+prL{M(z${r9#RF(oP~+`RGs*Q5B_rrZyF4wXB! zt<`U*Yya)kpC78Dgj#0L;XGlp>QP4P;@mppzveS%&g?T=Xb3E|dL#@Fsf+ELX|OGP zjsMc~i_WhWT(k9;aJ=2yNS6Z=jCryqz(S1$TvB&1eL6S)YX17TU!N{r%RTW~&LzF< zZ5m_cDzClbr(Zn>?&z0(Q;_odgr~6j%d<~buH80$>(y)x<%A``xVh7j7dJL8 z4*V#!`f+wxUJ(7`azwBS3j3^P6V%rnjwr$&**tTu#Bs+LH@6~oPiVq zm8-2}dsPI5s*N&lTLY|?;IWSUu}e)}eo;glMdUz6oPWvdz~cyXIGN34@~+6}aPK=oGeF5zi#`{~%VYY*UkL7WQCD zdOG%y5=6n>D1*(HXjL)8f<;2gPf>3h3*bghA-0D+6Ec9M^%KBvL$`W{pw=Qxjip!g z93ZXQzkz)@`uLdUSvYrK_zm;-haCl6K5{D^voBh~j z@!=6vZVkVs&epIedP!ti((77=%MXOA?7Thq1wa{l)y#2U^1srfpO^8EOa9vb`4^D) z-C+k-P%yF85C`1Opd+kUH4g7raYnczaceE5$GTt)cGa{d|2ZvHi0Z^({O!)J((PVq z-0L!gafYittBYd=Bd#Y$2)HF8cnUwf=J%8zb-pEq6WbB{ zzCRf}hx`SvA%(&5vD47G=_$d!$$uD&mwR&hj{DB|`C@Wxa}2oQBklKX-TJEU64*#= zW6p=X?_|PaBMha(rW?F&ok2EkF;ka>0?hXPaZ}U$lLq7spcwoxL>q=9uff(4qWz}u zm$mF{)=lW6Z5{)T0#y2t`(y;McP@z)V(){=hySqt#VaZk83H!;Sm75#;zodkoc~5J z*%ECBv66ZAo-!E_+n(R# zXldk6Z3q3qwvf_ni!s3!^`Ou7c&mDlpM3EVVMP5gV!!~s!fM6@T1NnS)NzPAeg^Vy zW4%~4xR&47dT49V&mkYS$=RVN{iL=jTc9ud7v|t~e66o>?+Ks;1x=xF7=*_WEe3wt ziFCw*5u=0(?#3t)ru6RmCq+KT!>C5$faUObvT6&}$%+9mix>*7lp&b|^bSw%^oH^7h${ zwkwV|R(BNNQ2R|a1Xgd#w)i!PEfhrnb)S+TLvBQkzio)^_sc@(J`cj!zxb<2o!dIn zB)^eO5R(zvkSs!F22^$x4G=#uHLH0qne{N;9wKoJe#Jnon#J#HATVHq|?oOi@_yA5N@Jg&s+{&`rk{c3Md7Fh@ z`7@@h4fq;ZG}%oFDk5AW9>Qk}ITq{|w#-jh4$PFyr_5Dm9gD;k8WwfTWaen*NoKgy z9eFE-B8zo11{QXWzb%waaTePu7xHD6a7%IWGmATgJ(?fCUBKhSMMcV%!0Vd$PO@ey za>|`xY{X3#-5uYZteuz~U$hBpYidVXH(8fn>TIBGSo7fUxO5VCnz^I858tn1VDty| z>hCiqF+4I^Gw`0IAJ^Z1dkKHAxUW9(p~v`*XoOEs&QQYAVs1N@%|*ep#>JxwzUAdr;x+1q(&m#I zmYRcQv{jWg-kSWCZ>uu3}5=JufcwCLU?*xc5gMq8+jhSN2{+w&%C^c*TQo5ZZ<@q62X!aCNymxeK#cvNCm4 zbWU`zJFGp<+=t!iZg{VRuAv`O_Z;Sr@~ew&x~{yp_}(&J?O&ZiP7!rDBp8ie^NK#O zK2;woA5CDt`Q3Yx{xtF5@muQ|Z^z@0;d@U<<)Bl0sIXcfoH^A<)i6KFUqd$0d7!mZ zK)^yHP(YhOFF?!u(8hZqA19=ys^F~PS|{=pbyb8QAY?QU8L*0i4jwMub&P(*;<30t z5jQvT+<*B?rU0)nTQbCOoHApm66uC$g+min75#-FBSo13!O+UEcmNku>K8Uwlfn62 zOROE!-sIri9@*eklutHdMh*3jcb;}0cP)nmlbI(#c{X@IWC6yHHD#MH01; zwIFsjZq~WrAp#bs`L`CxHTq_7&qk&ulisQ?FjN%TfebgjfFqRagrAduhr`R|#l-qY zbKF{d@5pk@%!J|H;;qh&|5N#MCb&gnESZe)(MS@IXjU|bVLm*S%INs_3f%U{G5 z2z8+@TJCw>q_q*BI6hh^@)Gjn%rxeQ9_9smTHA~sbS+3LTjdXRn%roND?P0=-IUz=Im?~K zmb`T6w0!CbTk@R!Z-NG~c`f&CI+`d<6(@ejPZXx}S^czXt!MGH&@PlQUSZvB^0S(1 zCbQxLh6Cx8xgwksE}!ZYJL;`e*Q@4~b(EFcBVFjqMavHsGChW_THDx+Y=j~eCl)nC ztxkArJ(kw>UDr{64E3&~&km?%r??_9@=8zT98~!dyOVWP|=(=j`V$_ScCGMN|4T ze{?oY)L$&`#SABB5Ad~qBmnOhm{#hiE?~3fvXVRLJrovpR_z*?D(+6l?C92XC3GH~ zP(O0thJ%)zs}bHY?~e{H6D}XwX4&3sk~^3@oA2m=-Dw^M9=M%T9$UcNf~6XlC9$jM zBVH08_ZhYarW1!}bP!#nE>Sn{@zmK>vTfnIZ$8*}?6g$qp0!M{X~379eh+Npj9XfE z{E*vs&ObdqeWRe=Kd#{NU%3VuvNj-kc$Y2x@Y&8$#hwE2|FpHju_XRw+|WwYQgTHm zynh6CGky63F~y`?@b>0~4FYvr2KR1$%P|YPg)c)S>|nBlR>)~D&m;STC+x=eMmvn> zkrUE`^Sb;2Tc;b-%9e2k{JeXx8>>nDmXQIY{Om)4eF-)LgZS)y`FyZGA22ZR9Ps}! z^8GgF>;LF~_5SB&0ioI@7#Kg8gs^~;>zDIPNKYl@&G)IywA6b^m|SEi5kUoTzgYz( z;TrjK;{|Q)Bq!Y^ZCx*~t;Y6CA8jX8*eWmHM>TONlAN3Y&T;$eAE*2Mha|JtU#p^jJO0I2{Z{4f97-V*U?xpQ| zne1={!{eB9a{ngdxcf%t-uK>WUslj zvE91=s}ACSk@*SLhWs+*gniWox>UqaVBJY^er9G*WW)LHY+zszm30#;5{@>J&Q_gQ zx0_-S#np{-(WTex7tQyKk$iA)WVPBr0eoHGv)RG?caFY~g7__U zhqEM7)m?|DXmuFeuGjk%@p8^L>#Yvi9`CL3ULNDmmmBT6^pAc;@yR;wW}coMoxM>^ zAmCMT#foqc*T35xE;PgvbZ1c!DcC{P);f#ZtOUte(x37Y`AFt`!u+gf5Yy9IoMu{; zZ|`sZxq;M-jEthYb%KAlQPXG$KZ{Ce!hEqv@5dDr$ip8U&q3*OaHdzAow}ZnB6k)B zhKA{yDx3CqG|}9*;`ZjZ{Rn*P#EPulU1lF^Hr+Dr1kk~C)T$~f-&OR6cWQF`Ul5MB z95G$~&zkgqtJf-#@Ahb#s;<{VW%CB5#&4uH+ZjR6G@)_vDROdJs0_E8L3G1ipEQ&D zXA!bh-7W(T-eFHB+n%24?)SF9P>w^L|I!-MVE!?b;l(0VRKhj#pd^blGO6=NoHabx z?JpID-gU69S^uKtpL5I6Mue(rJTxP~aM(!+MJ8kIJ~%R-PS3XLeP>ZCIW~7HC5#@P zo1VVjaM|i?+gASd{+8O^u2prUV|p?xL=zI0{BLgTb37pJPms|>t>N&d%ZHa0xL(hn zFLS@8EnrTEG_B7UE7!gv@cv=4>3Tlod%r=*tc{EiPv6~!|DTrszmw)KfpSD~qT}M? z`h#KtcW`hJ2t&8`U$XW9#}Qntpsvo~^YKRR$l~!}y>JNrkImTlIUR%Sj$P-m8%nn6 z^YO<2=i%YO?U^s#M7G^Ct=H3;*9jZkNd8~7HOCm-q__uC?yhpJ33rNzh-a@bS%>qT zd+L8*$^YW|8;HPj<5}Q>^6gF>$sRpz+P3+2nNp>FE$-dD?!x8J1eJ`GR8$tnQ<;v& zQ{z+OmS52>V6MW#3H`ZDRaFY}A- zhT0W|7yMu8HPHATP$f5W@{gufin?I2jmGtojqCi}NgS_(!g0vF#zH}VC#(!fuFW3e z;;NVWx(}D9K&QniaxfJM!3qAQ^BNPSS)oPY_83)H%+-iHUv4(v(FxijpLUgoupcm+ zCM*=7WrGGEhoD{sn{Y=w#f?`-Jv>XAP!b(VB|qsAFhZ5;Dd;oaoXc_kmxX#Hm^7Lt zI?%0)BU1kILwNHz{zn}m66cK|v@v)eR&oJUrW~VN-~tu3zL|m|2XwMI#KFGwtl%f;Wq5=<-&yT6oWK*f{*N^`dg0|9-=e!?(^i$C?m^4(!2nUy7LX6ygBST z*L1=(2E*@<$+X(5j-&}(5uN>U!9{Mv72yW^l%C^tLco-!NkohRB_`1O6qTv7k9*&b>a?(-rlF)XBS4D9 z&mLT}Mx{KyvWp?1`_dT?o!KUZiM?>ZouX4A3~uF4nA{1Tf1#%+DDhU-NmXm*bA35EQ>1PAyu5vlblL?n!an86;Sk1qR|($x&wXc_ z3NFfk9_ag&OhfuGD*ja*bs3T0{A>F3h08ux(!Nmgly_LD+dCOj?mHj;zbcEClj0~^YU|Axh&lWflY*gn)smrgtTVg@KFro9)lfniYSM z&4}QK&QEHH(^zD&BDY+)W!)#yQ3Tg$@E3Mp5H1U>C)5v7ibYw$=XM_q&%+Kh2^WOk z65-rgprYMJKF=A2r3Xs>EU}i~3Z&Ei;P_39I^HQZA{FNsvCl+)3Fgc_<^xa2tx7JQ zD`5BCJAIh!3qs(QjYp}8m*ADn&l4~_8s>|m=ie8c&E49;5qa;fZ*sm}e~MtG_^jfm zKGrNZAqJzDj;EeoIpL08_n17|IlivGgAZ>fN|Rs2TKjNB_s!+Mac^0Ze=z3yB5jhL zEc~cmtn6Oy!P{Sx;sJ#hG{^h%2LUBSnJH$jMAyEPbO^NcYSCOYYW6oFclFAVCL$-y z;V5!&SgGHY4rj+5WVcgsX3|npQnIqLvavObUEx|~sZxznqEV|M;^4fUwfk%w`exJf zYers#l4W7An9mO;>v;CT3VI@?9p-9pC88fh;yC2e53v_QMDus^glOhmoHZpT{90X> zcl%p%M@mBCa=EEm=U5kx(u_hfpJZI4(c`&PLEg=q<#Gv^INgog#UYeP zsu1wJ)%_mn+&!h;ZMAnRY&bFP;BY3qBuep^oO~SuAhxsJyYKqN+COnCV=#$O%Y+lAn0Y`g@ z87x&+0uj`O7uE}2HTjD$Dao_Ba?PcLV~oPZyydsuB2K8XPMI6`b&+822!*f-$v5cu zrPd|Aba*E(5C*TMZ8%WXYCbieEE*d#mz`#H^3Ho<>dZ+gVc$npw|2`M5U0p#{D`r= zNXGoM=Hwfji`<1ylR;fxccAQwRY+cKWXapHLBmp|nSptoo2a<>V<=s4aJ2*jh@x~s zT|w9E78l2LaDSDFMw6S-{P!V^?#SPl_*kElXLQjkI0Immj}rRBN7|Eof9hh32Vfub z!ChC?p8yk<%ey5;gKyd}`zpx42&Yz&(C_|K?Bu(GVI=3>W{ z959m_eKLy$lpBnDn!0djgx;|0sjj|~ef8mB53n~il35{|N@KBDK_7cBd1W}SD>9cN zTt9&HSq4rr3EHJhPN99zlp27LQQ&^@Mm#|{>n8v3jFm47u-&%A9V((uik&+&y2p&n zrA$WILoQP3m`GapTXwO0FWcW^3jh!V+|^85A-4@Ue3wgeue21tfcLLJTrMx8Sb$wh ztKPKI`mj?rdi&PzzFIh+RFJ*&&*?f2EJ#gH^C(t&>x#zn{(*?RUQ`<^-l&E`(yL}j zXZ83-FSpgHpAzVT2V1CAQb!82@Q0<%uBTze%YoCZDVY{D?8~R!dYO&arXZx3fs7|| z9?eha5akL3`?8Nch10+L$ zhm`ktB(E%oLcU#w;_>G1vvjs9z=K_L;0E%yL9axORRNVi*Tz+(kOU`1x^BBUxNa?a zNgcwv?$BH^Ii`av2d7Ze2cZU8}}8%f!gUv_{#Z7_isCbu^I+|$pNfziSPo4!ec=!_Upl00AY97Ff?eDS}a>E!cV$`!Xu z-iKDGmI{xqE8;3-uRaw(2<+91yvi&4Q3i}hWBZGT-F)S7BE&Xd8ry1S2M!9NswJ150O6Q}Te1l*&L7){`8rqA^W~J<0yV-lB z8*n8VA$v*#+oEZ%xX5g|;j|qH{B>*xcZ|n&vwkTC1N3=Mw3_%Gtl0d@#-<&3f4oLhWt>%b|nxSVXO_lyTV+X>P;At3&tPQ0mWC$XesSYdkmUe^sfxp#Fomy+AyC zKF2OJx2w%)YfTQe5Ri?==6W0eQ+#mU)|mHtEq2wxe$>P#`JL`O;jN2(n7dpB{tgym z2#?eFm?_Wm@Hge}WFx8!p*J|gZkXc4`z)N3n2xaf&eAF>gwqqP2~kyV5CU2k1q!&le~$TS%9eg+Z056iR2NLVgYX$caOLP^Z_ z6SQe3Bwd`*E$yS;!Nvs2fTrjM6`AtGLoyVpl55qI1Q-*Ylrd;2%Y$C9)n z0FT$?#*74yWm`l;#ZSDqd@Ut(>#bUcCk*N0{25%>`wZi;aASfOlT;rcAdW0LhAU#b&TBk=i+iv%2@GM+pQuvx3fS2pMwld z9&g}>%J0#b^&rbH9ShFDEp!w?V33&B$4??*ChlMktJ zmH<)FYJ#nak9cybGK#V2vtI`&p_MS~$84p3QG^#<6-gJ(>)uN}1w-VmXdJ^I3iP!_ z3U>UjxdU^b*aDWBCT%xR*}SLsN5>jup=;iL+RWSB;l@X3r;g=j3&GLu7cAAgwG0VU z@Q6Cm*QIs45+Ro}|kMh_CJ57z0Xl@=$mFf8`; z^}Aw75AoobqhMq74bc7JE10r>!Q;!sreeT|6 zLABg=Q{k#+X{EN_ec`mIdLBo8|7t?_`qpn4&+Roc@CyG`*ON-vPT&Ug;r$W&;j!PS zsk|T@YrFGpn%TqgiovdLV2=*!22&cT;WRT@>x}oOz4DX;B#bmU#h?La zB_eVEg>xw34@HrCJWe$0rKgqx_VKO+?p=F=!Q>>Ad?CL&B5?|ZVk;J5D2vscyXulT zkH*tzu{)QF2Y~m7&<**prwc#SEh)?*D=Y+1(-%co_5kcr7`oL|0aQx`c7@WV@TV-y%->2}MNh-MER z*dHa9e6Lj$y^o9^BRbHKzD1>4DUM_lI-sB#AJ%}x{gO@@nUvAqt?f&pBC5qX8JD45 zel|;PumP~lrf3$8O2die93UY1HifezjJd-+Nl~{{fIi~R;IYdg(Upgx>r`NX&4Q1- zoNJ7hs4@!B>m;&r|0|KwRiqGR5d*lk6AyKw-3vnC4a`1RE}a`6@jrDei@i15hI%fX zrkxQ@{TO-?tUg}hgMMDi9$V{GvS3+QAUEfHI4Mury@sKm5HxVd)0Yvst7~e58S4(n zF%ig)WJsr5tr|jE_97MrZQ z7Lnbf%@#2sJF~}4=4SbA8Eu^~5m$jLRS42V6bn^a6{mC_?i?%KKz=RBSd5Zk`P^pN zv_8nH_b#(DS)L=;-fw+F(qaPnowi7e@mn4KK<{T({U|FNyXE)6ZDPI2R8>iU%k~V- zXZk;mMCt|Z+=L$Evg1xx>toe#sjByJo)Xf$qSGQw78zNV7V{A{5VnRHeTE@+^BiWk)G)(tw5dicPcGa1k3zt$ChvA+KYxSwmO8g=O~HcLtQEQeLSM8 zc&8GeivPOaHzRF$c)qV9nF-D4; z<0+$%L~I`oLGOR(YOX|UUpi&l6V;+W3HS2wnozlelKZV!@tHXzBlr_nXA+d+z990} zMp~r~daO0nB5Rd`EvAY}=u=|Kvj}URQ2{6b&KQSDTtXwI5}eSQ{Q|+aAu1Kr=Zxn0 zo_k`<)5|sE2n4HwmYC4^RhMqA1%CaUQBe3I#BH_8O__pGj*o{6?O)fvQv2R|?#zZl zIPvCD2t3=sl>5S^J|W9kvLtbLi5m`GR2vmQyp+?__v6R2$l&yXJe{+IR02wg>gvXH z#x-H~Q+t5uj@J8gaGyH%TD{z2&pHv|Py6)WBakTuj?ZI_Th$v>Z7#^kI*i7kNw!xS zqR-l0vN-Y1%X`_2 zK2FZ`1J`rr&l)*>?df|Vcr2B2N@SFmyuTq~2LbNG(W*=B!@JarYL6a6W2q)2zn-q= z6z3C(=GJAr|7LhVRCdR{8XEYR6 zgyOv=P$*`7R%JB%%gDv0YTu6z@_!lkT=gewY@L^2TrbRzqkNEb=KMNnIUj$#M*|lP3 zNcIBJb)Zbsd|HkW_J$(+FVqH2PducoSl8+f$TFE7uD5#=Q4@=&dD4psAV6x+E`J-a z0r3VqaM}k|NPu4&C<2LhX}FTbbN!BgF|wh$f&5bFfz89}s*Nv1u13Ws@I&WP0vE15 zXY)GzLU^jd$o3}&gnZDX3N z9A(2KY5SCK{oo{)g%oA*>hnWIx6Gz}_n7cS+G&E$pSL)!iQS{MVRjK4Tspn>_adrr zeYME+V3-p07h6@(>eM&ciKDZ)5my|#m!4je;cZdmKQM#Ltm=Q}8bBBe05Exvu7B6A zTkAHGFOqc@vbU(3sG9A6?=l~L)v2#epQUd3wI`3c8+nrU~8E)!{hjf-59vem^hJv7U z^=6-)8|Id&zkRw9}|7g>udZ}df5b`YSvp|Mn$En{IgnifY=2mBYoWl5d}$gBo}!2v7nl4qfAMl_l@a& zF}SkrFLZPpXQb(XJ5EdqxD^Y+(OH*dE!`^x`m523O*DkIFy2MRQ%- ziRc!OujGW0gtRKITF+;yG%rgIq<9f>XHH?FaQgBRy#bKUoT59MO=fTXA2uLBcM*r4 z-tBm5{WVQT!CO4H`MLf^fL=1y zUX(8= zVPQ8~##L(~j+9g~+Q+-X@gLgX0UR`vV59rA6;}ut0p*@9)?0h3(2~}fCH{;;Cn6h- z=P(meQ7R=nXhvE+q%ADBo>jG{d4qH2Uj|`Y_v{Sg`9ySQwmSzMhvKvpLRP7Fy0dUE z<|@f%urt^eJ#YOHh31lxLOF~Bu$_Be_dAITgpU{@3E7K$#0pB1 zZ))COb|+It0m1>52(qepPNRqovR2^sO(UR4)6W>pn+WTQ(43cx$(I;W z82ss=jBgWv2(ciCz*N!>TBulu>1yzD>`5HW)njR%BR zV!6u>Q#gJ;Ke$>`pCO(;*+-H%DMQ=_FXzk2+nL_SBbE=shzbK)V!% z$c}Tn0LF7d1tlrQsNOtR(EGkobjywbx~|Kln6!D}UD;HtT^~n=72$_aZ_1P&X8`2t^_lFV;Z0=xMu{4e%D8C%UZy007EK0e#3 z*PyKr-fenZc4{iB{>e+8c4bJG&X$8D&90|g*r&KDP(OEu59p%#27U4bk^k!w+ef?i zoBOZFtF7*5iSfNuDto){N$I|K$fVMy-o2k-1G^TJsjNdXmTo*93YiD0drnH#j<*VO zt>~`d0gh&3mx_=?YbXbrY)Qq3wx>^_u?m${_Tv5`w4^EmwkkSXx}VeMFo%NU9ZX~M(FvGrWwEZ z2wr-fxt%uw+^pGkY1Fu~3S= z37eTnLkX(Y8XY!AL+k51-naTP2NHy0)=&{H1}4kXxko>k+DXP1P@DMpy3EHG^D3&a z+hzU=c8^u)#hFIl4#%Gam@EM;0&1jg?6liV62rtRP#}Pd@!SohvR6FvEw+I_?DP-e zy8Q;Axmc1X!?>ea?63>PJNKO=RQz{E`X&7LbNwyfUto85k+zt;PFcEdQvvjmW*aSK z2i3fGdUI9B0XC2avc1*($T`tc5Lf}gahz)KWNLSXZ)=|MyJ0d()@isO*IX=XP0INF z?T%>Fy8cD0b}o0qXVAjJdP(0JD>_cgU?G8Lm=`1AvE_8<$@Z>&UFa(0*sM0`478AR zgTyQ_m1)h*?>j8E#oYo#UhUpghY;c6=t3{fgK;vpe890ZEb*eZ^%I#1X!bhU~a8t&5G1B86+ii(O~m#(DfI#GhcygQ>AG(jyV@?>&Z1hs z1=oqu(kgf@?CtJGeW0178GJ&K^zLB6XBHOA;`1FWzI<7D8By2lIMns|xd002^}2C# z8F0o?pdm5L)iF)$ZGaG%zpfD?f*zw1btQr2-j#v&VjA93dm6EDYj29cU|$(l8a)h(Z02=bDjQmJ~jEn95RJHEoPwBgD^Z$hvC16J6UhU%FW zzZ&vPLqa1^ft0d^nxG`9m-|8;-*Gt<#HrmN7q_hvLDON?h|EWdqy;$h-adhAw-k&U z(Dn9_kc}R|cn%S1vT$YjO2LSA*OqfY;WjZTl{*i~V8}F{|>G9q^b#%4LY4CK-Ia3Hn z|BV2u0B_7HJwU3ur)wj)o1DzI%>_6PrLdDuX*@7WDG<^j1euD69?l8PM6+*X z{IXPHXIO(L{M{HBrWUK}{MY{yEQt=bELSMl`-`AGD8X2_R0lg*u5Fq!LoE?oJ27g4 zxJB;jp4{TGB%EP`)5E>47x%2Zj{Kn(dbg>Z^Wz{g??RKUedwscAB(jm2IB=9TApJj zz)=6!fE)>uUDElnDCv09uCb>vNwaiy@W_HS_&t`&E|1BnNxw2pj1B z%)zrt0N++NINaUM%}vEwW$xRxb8yCLrw4F+Tsot}^j0LB7SwQ}d6_}`B?Wt46}7_W zqPj1MY*%{jXkB0CRFg{Im5=vtT9Ix0M(}OZh!aTc2Vb>9vg156AkF}ZDbUCsrG2E8Z7MPaRNKA^lANO=_RsQHkrJd?6Y|9tBYvcu6+>b zfePCB^7cacy_4|dYo<{3QAr?a1Plw+{nH?ewm{(9d$%=}DN1eaZnw9KP&djq=d@|& z9TH%;*uv}=fOTYqLa|hmbR_PNN+B3%@&LivTv)nLsB%`3eK3*Yv34`iB~fFP9WouO zIBz+idRC<%qQEq>WR#D(BMf)5C2fMkx0sxkoBWqpfbH*WreE)P)AuY{dc^Iy0ta-UOf74jcpq-k0_5Vzi#@vaf-NP^&ON&PY|CM56Zbc_Fa4 z!Cd|H0fu=oM+6pa+VqN1&_*8IVLH_>vmyrKvSY@^Afrw0h3nwu<`2C{xrxhu2M3wa zT&R^?+UQ9dQ;4uIsb2K#f&o{5Q+YB()oBKljXC7|Jku+Iom&Ysi8B_<>1+D!c%&Tr zyq^aJ7e+GKDQt2Ngu%D22Z^c%#U8**U*N}QvONlukF+MbJKO6OQ|0|+qaDaJ#f_Y@ zhh1+axEKI(>yi6~dn|*FS05iA??34CiHA#{F%)IB}iq;?sb`YD>UgyC4BBuTPlKdgZ4s&mIvozj1h@GfRRSIL6 z(;3xdcGMe+xz>nOFEQ!J`=xxOT_1X-rj;y9C#?G4@i1B;qw1P@y~81=aOMsOwhD07nL zVCFAwCsB)tCHd%s04*{=25Z>h+3*lCncu|f*gR_S3&__+GeyP7IPld)H~Hs5Z-een z7*6eDKZ4%01Q(^SxL#AFOwRB)91sFJ!tuIZlBMs#AypW?jEtdsG%s$-x|!4S&*iw+ ziwTV<;vpl@i=$NFfIouy{?~7V({YM(0kU!WllkfvbNswjW#s-sA-N7^yo9*|!u{_` zn$IBN($>472tY9IWiX#t={~x0=Z#2X0kX@cyF!Ffw30a?iZn$zrFGhB43EJc@&ZM6 z0Paf>!-SbqMp%ImX=yn%O3Yw>zSH*$_`G?pP9L{Y6gTgtF3KD1U>vOsxC2B_&o<5d%W$HXR8AvBKpppk6fY20aj5(mSC3N{>J?-FL5ed52#=CFsh62>Z`+=~_fr8!## z4IOCY$zyW;=`%-0boiH43?TStvC&kRgQNE~k!B=c*3$XmtJKH_m$T2yOz}~+&wCwR z(#~cup7*kmLZ@Tp%;H&9Cp?$Sx}BKWi~`?%SY4QU{cnB7i64$Z@~%(K1V`{Am+4-g zWgnN1H(FV}o9ukuUqD^G*4*MOck?OPd;a;0Yv2B9=2|ANm;0G~F~eBL3M|K?OqaG_ zVsN4=UX7#66NsI>Vk7Oo_zX@pBQ{`Uj4*m0e-@pIdtCW-y!|r{dy{@b zKWI3ekowxK7fim32}Eci(IhX2HT_brl_!Tg(zdM?lDr?YN}FT8`B4MuX?W5Ep!UP2 zv3|eEe)u^9XI;2xRdy|$=(ZQ_y7vnZp(zaU(!8Pn%!3W4dCrK6? zF|HcbE~G;9^*qA;+Nd>2w01u!e{rd!dn?g8r&TUjI^*)vHT_&Y#T|ndgupdip^?mQ zI$d(0jSs6NxQG&oS*rH-E6HC){iT$qB=MKMpX2@6#zSAm<=%IvA(Qi*@ml}3h;sxo z)1Nj!i6DVE-m-KU4;ZM)q=PIY2B9b?8~8R|z?)>ry;##i2sfRowdSjEHU?Dw=5vBb z05$0=N3Ws9wh}j~o($lkILWiE7R5~`IfA?U!Bz-=N-mQTCZFQzSZNr`+Dd}I`l+go z88h|u4WW4-hq`QO(EoLl@@6O)z9Y)_txh^@u0mekTIz73U+|lTd>PugHWNTUF`57w zg)V_5oyh#|ZQ0a9C5)f#s&Q$v%$f4>^eZ@p?yNK1+Myc8 zHpM@oIOelVDR3)A*M1*x_2h&1vZZPBv9fr3^LY{ecmv+u-qnOmqpE7@*RSxt9t*9r zeWhoi}>Pf!Ws<;#RNk!YkTE($Q3>h^)`zr-sgG%SrLOB6KBN1qJ(Y3P zcISLpqtRmy(|u!Bij92Q(3QvZEeK`EiTSNTfJ|i0wSl|L0=D)80t>&vwOMoi<|97_ zA#@D?G*roC8AtNetM|LCg68`oGO!3Z<`IxG>P91!&U4V9Hl$x5*<5kGCbB8r1YEADX;Er8u6# z%m|>pp%bb$3mw)}30u&fjav|qtp1a!#}&hpOUm5J2Pqv(VDP<;M}^-H6_pvkyQox# z%h+tWVI@6#Ylo7j?l={a$-@PBN?pk10u!w*O!foMrBG%R3NUv1BP=z)86331cK^~p z_k6Ih`{?m*x0!>ooNE1+i&Q8h%Xb#Avl3%UwEvy+D2wCNP_RcjjmZAx$Aw1iOO5M( z=-^v@e!ihW_nJJ!?Gb_$E?-BtGOb@7W{T}yiiex)A%q^lt3(*}o{5R`SrumXUT-0l zuNqB|9}cvZK9AK@BK-Q^oy}#!pC&-YYE3;DfNrN?u!xE-LaD5lDcxTmY~gD#EcrDN zS%!kFvSHCO%cK^6DyCQHvgIwxX{!=+_2G;7c$btGo0gsJ^PsHS?{_=Vy!pd4Em_-r z<+FnGP^8OjRNZ=*)?<}oa(4t;;Lw^Mn@2tSny%yde30zZ0gPtXXff62pX<0-T;qK? zCNfQ9=Q!Dw&}6m2WOc-_atG97`$5=?s8ea+z5N~ zTbJJKFwy}V&PN?JWe5PEqOD(bZp-7cBC=h1n>aWwh<(8()RWR8v7khFlA*8}FDlAJK? zz2zd$siKCw^qL0SA8@f)da;dWnuT_g~{`}W~&N0Oyc(4>@tvP7!x^Q zxwGB-A}Fr}ios54RhvyZT*3VH49eHDru~~f+0QHpz0MvY67z>uwgo>1L2~OI2{ExS z7WsbOw_=CM3tCUvT06y)G`O-Nw;eYMIUZNVVyI+J4_EomFD1nr{DBppD31#JJW1c; z#%Eavw8Z`SeGw_N&I50YkM}p+Xvp%HjA;zy9W z%I5{0q_b2sMQ>tX`Dl^*LK>vFTnyw>nw-C{%EWsKmME=81juK<+yjQnKS8fiM0%|0 zXgm%CXr(B5U@~%Wa#ya*C-OANjCIB9k3O42<_S&;DpHn~zYgar`qFUwU=juDRV??! z;IMf(`-oEK1vEY^Ehe_Udmcn|392Py1c8W{b{&6CD9!! z5Pk%P4QgW2$lW}uv%RPb7+odaF{v7ah7|!L=jG+vJt)V**?Kg{wprbBW=<(3BkJYp z={KoY0c@`DMA{S#(NHO#5}3I2YA5A$&`nPJj&SR#~ejk_r$En;uG7v>$y~$krF(ejb$q5Rf}`$G>ihL8quD+WdnlROxV$raAoh?1NrP zrk7?tG<+p;x+^2(M9Zn@-E4KX@dW>Qhk{#=>dTwNM}}%h#64uISy8qVE8Q0sVzEpA_z`Ak?Z7J1uSHs znys#Mvv1{Ss+c6QV3JwC6CPZ5%j4Lm>y|Zs@MVW^7bCiFQWQD=rsylVOL}u`FXy*| zrT|(HP`6JFVx$y*Fhz0@&Mh^kEs~bo0cEzURWY?U#|!e16Az@h)}5~ZYKo=eNra-Lo-y5^2ShV>86l z0`VlCBPpps(^FW)zTJ7LZ2}W=Qv*+&xkjqHv_#dJgm@lYoci6oS?O9|x(KI)5d$Gj ze+TSR-xK+wA^Ao%szC)>v&no4v3Y4W*qukKm4=T;>o=ng5+pZJGE1K6B*vOQpJ&z9 z+U`nB25N;gu<_4-gI|acxPWnH1eo2JFd6sXF=@B5At3#PgtxOR|RomARs zCJEE-^Xat?I{2A9h>=G^x`K(_4_}bcp!qpQW=X{Kz6@)DZf!3n(`-vzz+@S_zw=em?xaJ zmuLd}&+SuG2%Fg^vF3~gE5{fNCYJAqRDr^^Ms1p4Qv7k6M~~? z4Awz10jJ)1+-RKf)hF_S0|I#tky(eU7xIwSQk-E(>n=_`O?X_zHT~Gm32+vXF6a%D zmWhs9L#4!)@%Hh`LR+=|ODE}{kjrdtni*BTTYj&(ZWs=^$po&uciKB)yBM`q(3GnY zLDH9mdZ5IB3b&$oR4N;v z^Rl8hzW&Is^OPp9D0l6Wyeb06JFAHq0K+8T?B@WCzVtW4^!wz(QS=N{uLNq3yd+VL zwun{YH@+kd$-R>p6k<3~TwCEX90n(mcH$&H$YdyhlSIMS?}jE??s*-k3DASFL2~;ZA*nxeQ_A8xh{$C6=o=nzKuaU8)5E zJx;5&;O&H?q1j_&?2+JIA!!8IVmjSIVj%Wi6Z8)Iim~a6dnes#pr%R*9b8CWaSlmR zg53#=&q3h>dpc!a^FIM3O`N5~e`EptlZpSo80r5Xsznz_tRk|i11Z*5f^Gx&@A5|# zzd|JuJ`Tuq6GjS`af1nm_H`Ce2?Ve~sz{xY$Zj?m6X0|M+j(eOTn$KZH1Hg{*ZJju zkMmwHhG6d|WIhvNmxu2GqIEyd(oy#HV?4??xY&H2>ttGBB?=HYnvDjpwD-(Yf+j1_ z;4BdH#4t#dh=f@nV{~rvzxD2!v3&roT8LAY8Eel4_o-0ZGVAQ($LU>iOX$R-a79n$ zwty_lhcR)-020gkDpa+egA5?bVEKL>bumYVv;ox3<8{JG{C2_+%+ z@HJb@4P=F6O|2;EJhWVP2}~E~9)koluE1i_9PBq3Pn6*LN+kPUmY+Dam6nz+FE6jJ zuj6)lHn=ZVl2xeH>hU`{-Pc@wI%hz9Mtn2={kpaQAOwJw+%v!xri{#FtYgUgu0Ndn zH|e|S>Z{R++o9-IISj*01%FXKdZ=AAv$r;DV&V`M7A0JU(++X*k*xNoB&w`)f3xAMrT3E?2Jvkzg>E5Y@x+!0Ihp(kJ!wdC?K+mtmm3<> zY|f{~=C9u?kYW#~nAWG)W4fyU3vyy%l5gffGfN+2d3o#T525@~w>eABY%G^^If9m> zTox015o@N2<4|i0YVB#=+Igy1sGyGmD?P)%7$$jY6?Ia1g9Ps{CSPSccNqY-P~B44 zz|-00EFBgtZ$TZtL2FYkXpgmtGICT(y`iJnF=ovhkAXMW>B{SEq!DnuAD_P#SGTf_ zc2%=l$|x#}%gP{NuhEckgjzGK?Q}ehQM6n+pF?NpbjiG2ET~Idh6odaIJf1^;|p;!SyILRhdfAo({i5!2I(1%1KCw51_C_!=V8HrbXAY zF3apbfJb>fNYRZB4~kwyWL-!WY8iw7ti8%I4cw*6%AtngNC|~X&v!#%Ao$S#KtIT# zMUqHDfXkfdFU=ifa9GWZO(11{dbV5{d=(;eyu9(Y(M>See$S>0;# z5&RH(?ECfPi;Rp6@ZhhDYqr#Y(yM{)^{c<)hS#DGge$eG^A7x0_d{VU{P#iRyk#x| zeg#dG-i`kibGrKTzhzDVV<~w#QU6Iad2o3E)7katdA56~6#ew|+-oZmAommeEp9&D z!v2k8=l=jErucskm_!YgnPyhOJ1*n7@K2r1Z{!!9N>GlJk7$1N>D3Oo$}58}TuDV> z|FQkwz@=p7hdRf>GhfsqPN7T&GFN}{ysJZM&7{t=McNnDQ|+)B@lInPQPrxkbTS>5 z3$Eu)4Q0Xsgi^`y1~FKyWDKJJz@DtE|HhumWsvY{QoPcWZb6_XX%XalIsb_)O)z$U zU%GcoUn#4hS$TS#6DXT68HB~<$m=s_@35&N<9(muq@8TNh>7oo<}~i6jW$tucN8G1 zA^12rfjz4ba;i*a%(Fcz{KH()WVKG2?dGg)kirvyTo6?^mg`Yq`8)ILg?)7`d(i5~ zEHX9b45TBt@2;}6A`MRlfq(u?%eLp{ zR8)Jd+!78R?8d#|ad~}dCCrCPpb0jeHvaW1jRq{mA_3--Ngtblq!#b-XZi`p+cxGi z8o)X!z~Qz!9yR5mEdQN19U?Zq@D!?WLuGZxOfWE-K@u@*U`=F73(QY>#9**aE}hNSj>># zlmRd-%}Z5`3bi)jY2fN^2sKFYH%=c8iU}k+ViwVd%OVWH@_>(Y@nbz!9Dn{qMj+sf zUU**0kb}wn8*URFcES}^Gt5{f=MJYeUvGsy#2oL=uCH-K+>!H~cgctiZSlu&gVa7& znH+jb9{h!B$;VK!d(-Xp@6r>51pd5H4wN^7Q|e6}7~JKb>W5of-lFj}KMl-MHq$t> z-E!y9%pugwU<@NaefJ;Jt% zdj7DrWfg}NQ9ru~{0aPBW7=%ZYP}6v7&cS-@?|c7yHN1YU4LD~rMHKpPua4fT1s+R zHPbe&0JF9-SQG*Xi60<1g?%Q}!Qq@~e7LI=4KD_R* zfEtU&D&)&xS%xAkHt-+m%9o4@jbXQ0q*Ue!_@8wamFKU2OGw;as{lVa!8li!+P3}D zYBr0qU}4^QHyZycm#1t`=>Arzad#0?11ZWcb(Jm9SEpyKLZ8Jo$@uZVQ%*b)Rn;19 za?TzF$a2$z1`?t$`?&)F{2awh4{w1MOnCp3Y+AsH^7wP{=>k@)hG~DoCh6c(RV`T% z>MZ%&yLcUo-NHra6d@!Lh08UNapL~-El)eYIP*X0rbga{^uQBKtfCfhaxlD*M71LZ zo9Assx0dKEj)_dJQkfDg@@A-f7I@G>F=1sKtWi0{Sp}pj3Bg3IkT9GZzpk*AOFh(T z5svd=aE0Z}CS zgbM+OyWHXKc=^)mXQ@J?%kKGj9xmY~g1b@oWjKkb!qbkd1$Y8S4`O6{$@Xd#L#AOijq_{&>dyY(0Q}GL&*K7u!OX7z}$)95e zm-qYX6Z<#TE4ihH?_q)jbpzL1n*|k9jKqE947l9ZN;%--Ll(zkf<{VGauo`dzRc!} z4`>Gr4YgSW5^*r8Ngd_qe0l~sf%{G|7NCXlarp)NQandtklRnIjDk{~r};kG$V-*p zZ_D=C$Q$`_j?Y!Sf6R`W77U&=yiL|ap-H@d+t$>qd}qNDKe`#A!+enCp(|WHJWVzV zjjh;uhyG1lLJlwwW7;RZL_+z&$>DDi6k&)$E? zCJxuXWRn>g)&`1{xJoR1Od$Wt#<|*;mPT=*s(8thIDXKxngUdtKYMWGoqD|o)Qz6VUn~dusexO z@n_lfJiKDAw|wfyrath5_gIe-N$4j!9UBV|0wJ^~AnwdiovpTRk)5%oRf;kSfy0dK zTgS(?JoO?YecO*b&ngJHYUM9mvUCPMVDz=uhXFz*vrt@U)b$$(&O`agAG9oA&$mrB zKX3Im*q5lE{l&o=NzUb2pz`_{wqRSn*3~$@=g{_GQi9@Knp?VBu~@6VG3%T^RI@jy zbay@h>{X|N0BGIS5w14#&bNYYBmtg1>op{*2-458-sQK6HerKM;6gys$g9UW zU+<<~52qD11bkvSU?I>9ZAKl}eb7MCckoZ|`s*$BEjSez)ZVsb6DFEXUg~$b>5p-d zpEn;Fa3Gvh+)rA1CB9#V-5iIb^;Za#J8)?(f))x?T!Izr$^zknnlwxf*)k;|C$aQA zh(1?-$u#BCG4<6J52DwAFXBh!N{avL%`aw%eWiPZbsbshmF(LNW@1Yo1PS+=b^1E; zUVLJ+_2|R%u_&+nWQ3NxVjX)-b+e~8EXmp-^$fH#>agyICBB3@WmGhuXA&=mnl1yC zFh>4k1@B$-L7y{0_`@ubtXXFAI!TAf{ADKw4!!bX^|Nd5TCL!AgQ-fNYZ&H=J6W)? zws#o3>JF4x?I>Z+EibQXIngSDJj3)kMVd4pZ>igD@VH2M;+*MSn3&q#n;PfKK`v-m zTys(Thv2Lr$*#)5$gqD%CyWQdgBh(7Xl?!ykdb>R=*6~L@3!YFW(bMEHIhh$n!J9I z==P48bsW?!%&rx^EgdEOw>2beDhRxm7c5P0=J<$pK!;d+!VFtxx8QrN6dChg$+-12 zar=klVqkHWV;xb}F4vFuCv;^2Ey_(@&$Ev1?Pd$5C7F@4ID8-Pv2&iR_yn&(mBHd( zE|sXl5*hP3DGo~!#rhJt~#bg&JF5*fAl&;kRcn$73y-M0{a2BwFZ z*m~J`)T#x+umVajNUKF$g}PHTGny4=gZ(kw?E3euQLs%(6_0uibflbZq7s2CXF>tP zHC9OIP_e%xevO8SBqp%1a7#)`_OE>a8fZXP-T@t~LzDaB30rU1+)`@p3*!~Zw7S`x zTuYmJ!J#wBm?^mUxsV`-Xr!c{z}D1a23BMNRSACwH&S-u|52miPyZf|5jZt5EHF zi@*MX_{Xd&f#Zv1%8dvxF4qHHh-)yog;Qe$zuG3Sd ze8e+nJ_PevY1z&WxU*%V;c#(uI9AH&94^5Ck!u*>rlh0_k3>~cgZOs1N$E>g$koJ( z{b|p;uhWXLW{RX(xT&AdKR3gzK#*~lPcVW!*NPwDzq&6ZKhh6>`He7|PWCll$3M#6 zmhz*s7xn5;>eZe!(7GKArgQU#ixq@MpyH~2zA8t2K4gwS)NgrY<;x=zg+jyS+~C!C z#WaT1WHIJT8i1l@g>)U}DAwbQaHLYMH=BMeVMvIdCSWElqeg#y>MKd3XNJI4m=x1# zNJgOZ4DNP?pyO5}!)YgKDv)+S@4{cf_&2FlA=|czW58zL<~QbaHG3YI z_PG!>x~8z#*9!eAnSPob!x zUTJxr<2HsRX!q^2J(>h_c_1B&0%THR0#T4+$;&Y$xNkm^_>=T&smBu-oqr^rRj7f?yG!Gf~TndJp}+ z>xICh(rIuTio`rytP0QzKl>|mvba57F+u-1tRIz9g(-_`5U%(EMlDr`;Q$FEA$xbU z{)w_tw$#rikATa0Y|C3@zoyUVE=7xe!*H@zGE2a4+<+l%AXaO}w=fnd{0Xh|?W!oe zU_M{ztZ{XL*VB)eOo>G%taL zQYaJjukh0x8Y}`+QA=Q+Volo9uKOJB59Yf#ddH)PKRXP=@Zrf%1 zWwqr}BT#s0C>R(jN|6|q;)nXto{Sq1~2(ORe(Y*t!7aXlYR2h95O{ZI~Y^-~` zN`F{QV5RXinEYz5Qbj?@0e&0_zJOJ*D?r(fGM`JYEHa!7;1hShC>FG6WOE;C-4kd5zt4;3_GK z;MauDHwuN!&u+fQGQHCuBG09usCdIlN;<5mblG}@Thrh2WNB|tukwhuT*Gl>zU&VM z{dbYI<-N7B{0W#mZ`0ljLFR6LWPAzNI(TtU(Eqq|gEQ=}#YxWC%p#RaoqjLn_I%nt z^8rlc&z&7!e}zOfbwc%+ z_Jh}qpSDO5R_`%4PgR#UI+N_}jWdelYLkWr{+NsfX4y$E3_?PNa;d-n&w{SXU1QwoQ?O)D0L_ zY`7QnTYa2BAD72kCA1k@=(ayA05GnJmhi?J9_Qy_@8nwBH~5V3621l_&pAUTgJiIW zu+H1j8IV~|wagWMhlAt$QbXjrY~2#8l_tl@78s-OY*q&OsLDZ5&+M{WCbQ?}3|Bpq z@>n?*x<+gPmfnwbLd)zUcC+SWM7{)*!P9_grnf>3AJ4aZHf#1i{NyB6?%Vmchcozr zAP5)CCCO!0t1+9TM-pIb@t)=?V+EP{4y(OnEfp0snbmqz+4qOh6g6@RioUHwX#e+Z zOJ!x$zGtZn7FgXp2b0glw8X^wTT0r!lCx^X#=$wX6>@Zx@~%^SeEgD+{f||rK|;}J zrmbGj)>8{-k6TBo&;_5btE*wSrHSs#BvqZ0z{{hfUvqxEWiF;or~Xl2o{`EX^i_tV z&d>qvur>Q}kn_;e_1xBWcIMA4`R?I8$s*{iy@Wb=$?0^TLmjd@Q=84+&)0~3E^M51 z0mlkLxv-EwUoWz!QVeZ4t#eK0Sn<}LU1e;_Ybs1{5mi*XF6Rzt?vnZUY)j+gt$uXL z`O_SKLs8+x^)VUi(imd*JkQoOB?^KVDI#grIX45Zy_sdB%OMd(Fi>V2QY{UVhw60D zM?Q35Vwo31V?DAv9@<@NT0K#ZQT=A3mdw5hb_(nb(Yd}4IRVSqm*d`T@DyDY9rj_Z zkgMYIGk5p|;_aCezE)2My`kN0X~1a8iIU$2l*d5AD<*OjHb+gej;S0!-p-;DeNA)B zae9e4@a~`jz_J3*OUufH+sixrK8~`O<4B%tY4D=!43aI0Nithp7>+ig8a3%EOg?&c z46%#oI*#+qn@rnx_mea`gOU90);vs>$ZG3c_Gq(Ja%E(s=;|mk`YbHYK!BY?D`!t7 z8|M7PTE|xYz+DD+?%kK&od*YU;70JHSepq=kvHRL+^pw$a|XlXX!N%=DFV}6H|XQ{ zmSWP+35ZDelZr%v^xLwTel=tdIj;}M?vs*ljkMaI#OW6m-eRGq;Ya8XyTn$(~&#(>Yp!CwsSq2d}iaUo$oOUdG~*Z zO-2$0U5)xiPC7i(R8*PP$UAoC%wxCBm)T{xv^RnBC&iSx^PZ$vk3|nl8Wc<8w}CJq z_`Mb}!S+$&%9+b=;eFn^c$OWsW(A<-;SJ z(${By%GdpzcKIbP1n*_8!1p#|JVcS z-r7_;f&E8x!3yU=dghK`2$KZj@WcQc{C8PBE!zp+pX9qPY#*c{{M0usIb+9QhPZ3oYfTK^y~KM*o-p9D=d`n5xc1Vc)D^zRSDcR6>UMN*-CG;6 zxJ5lO$mt{<`$7`uC+7BS=fv(1Nw0?9vx=xyB0E^N&FcqTHhEA50{TI?85<9K*&js< zi}uJRX02G_4K2$MQE1ac!&VFq%Q@L(G9@^d zb2Sp`m}Uj3Yy$q&!sFXGqP++XW=crqAHnOLy`P<;atB;fr`fMl(nz6~Fj(ELu{u0lANk;z!GaxRa-` zm@D(MI=5^HX|Ioo>T4U(#MKcQ{RVc;M(lV%h|dhTFq6qxc5Nix#^q_F9GyLJn6kPB zm@+IZY;(*DAfgS2Vs5JUu1qU<1^09I_xDY6ycf?7qE=Y&E&D#!&q#fuF{v3D7!ck~ z`-5Q%h9WIEb^!VDk^S_jhuW?iK~dPOHkH+-a8SJFGcV6Mcw0@@Yx^S!=$|h7bMh0a zwE?PgULTE18N)`lUGP<|FhG{lBZnzpeU+{0%hMG{)OKLSN{-6<_m7Fw7Pj5Ep=%eT zx;NknPaMctAiK22RNwuUT6bw^$MxN6M2E>UhxR#(QKpq?e1Pc`By-#Y>hs2X<}t`| z)uqX4M`$&=kq|dZ#=G(*GE^uBdy^%k(2i}Al+^Jw>KNAsclTEMU3Jj$jt3faMZmIX zoO-9%P*pbNf4F!)YECh_>lS)iG)J1lQjJaq;4`5do*Pv91&J_Z3G-~)_C|#^_9XA) z*(&#w(vwYWRCvM=;d))`uw`m~qh=>b7S{@9kBMc8(W;oMKc=a)vz`3y&!de6)3Jzc z(?x}su+KC^{NT{HEI)r@i?)HGYqZewaV_L02X~;_@ghu|vYf9;0&(;Dt!T7VPeNVU z9@(Xgj51g-v`I2@2xGB_s`ed~rA!*CqP`W^vkcPMr@Gka{2B-=xgu$Sq5Fsru5wwR zFIB$c*f`@?*O!=)9l)aRpv>`Y%@3|Rg-h=s;ZCLBas>lHMF`J^bA}qsOL7w~Z#py` zZO0WBUZ#2iG4tLMpNZjGWOSKZeEdbY*z_72K!BKFqEQm_1jJ9O;aSX`WjHym1xBAP z?Cy9tjEz9|B^$vUD{Hxkc~EZBQ5KjoQIit;1Lu(bG+>FxliS6{)P>MUf=>SA8x==J zdU#k59}F(i`RGw__}cu)@p(T1M9H@uMliE208&k=S<83=Od>Eh*VWWp?DXY7?RoF( zcIzx_I?)o$hYAQo!{_Q@I*jx*ELL|2LsUDufw2+ZkDlR8-;ZEf@9*bi@%{+m@o7in z@PJDrQv#BzaG5Po{0j*!rt^g(&KLKF7$AW_&&C#VDlfP2;&i6!&H28xR_312Z_+mE2>7Igm<7TXojcK*wmmxL&KR!wVZpMYFJ$V**Z;tzIbR$ ztKFPP4lC%xJUH6V?T}%YP9jEdGjdb@u!Mc(tzf{YWSL*iG}C@-_4PImWv+j`?E}8h z1}YGcdS$a|cuyphzxLb4(VX8UR||;@eQoRswMrBGRy+PRT_pM zY2?M0vyd$tf4v+sHvLa$v+iuV(;0FXpuxsNcs+I_=|v2@fa9auA53HKLXr(HGX7N| z23XUO3qGfvvdw6D@BPYa^)dTP;(0%PYyCT0N5jvL zGqUqlt-yta_|>JdsijeMa$AKh1~;jBz*Sx_82BFq>P#f*o6NrSBG`$;^L8Fy&ump! zQ4(J)$Y1TEE|XA$N_D|aoO%anV{cn?ZI{tDC@-Wy-?IRx=aPLr%8;@W=^U3)a~ZFf zy8_>?gCs+a42_%y7~{%=u9 z@qL6=QzTN&!xj;`Z^L-p)x0D_ofq0rEJCcbJ{?LDj*Y78xIAt{3v1bH8?m}|P9xqo zou2F{f-7Ag$G|6hp7(h4oyo)O3wnhO2tWjOd7xY-OGradTy_2y%c1hHQZ;2B1rDukR6Q{Wekd>f=Jfd@(mK9j8Dz^-g*?X^)p7IUkJSQtlkT+ht5Rm& zi5#n0c`zWUpmBdId)u2;)FV1BJ3H?s|L}OsQKKAmyWJIVY~_w#qKFp0Q@rvy?b#_f zU-Znu=J=_Rys_q*vZuj~yqqac2)r*Ki{Sx8jGH7HGas2buycOA$V~giSZRuGs3){T zgE6vd_zHh9;~V*B|8JPfQyPG|JR-PvuY1D3W5tTj-G46Qrfq>AIiaHCSIbxxa)_N6 z)FR)5s3R|50V<&0Q$HxmjG8yzYVbb_^e$QqUD+)%BAxC6^UM*xM>zzPPPTeokGBF2 zp(QPa_TxDr)JPc3h*DzONan*h5{uJ1Y9p7yY@I{;({A1pAH$Yyjo97sfc zCKm56sOnyD!`xYAbLnJti)=_oVP0u7osZFWzkhsm4>uQ$C3`SWXxg?u07mdfk>@lf z)pBSefK~Hb9XOIlH>Mc;Zf^tFDI|c4qFlmo+{Wyg7`YbUo`%C zy4L5-ZlPT=R}z)Yb(g+|#XQ4;Mq>5NgJ~~*w=a1;g(nxfUnz0THQ8!SJ6g5Ba-__+ z^=eU0$4~*w2_}vk^y1rCy#DM%M1bgo!E5NKAXrel3!@`nGkyWdWa<(bK7&qZZL(!7 zGe7DoCXdF{@E`3zzzs0K`&g+K9v^}k*}(jGzv?Vf12YIc@&6KEo%N&N)8~;#`eJ1z zmP)%f6IOZQ?7Nn$H;Sfj9x^Ag<8QYw?gZScIU#Dc9yL1!{r7#EscNa&#J#V{o+Af>op1X<@IIb-(;zh97$TLsO z2=&CAXlHR}62LzHG(C?0t0_Ev6e(c_0SRf8`IVwLggHSnOS673X1ujVtQ*OX zg-DO?IE9-Ni=P5>On6Rl`vhmoVNQ?Wo8k#A?KR}Q!#{|s4gN{)N+^=eR zKCH`wIFQ%ziwIPnoTi2aLZh$0k4~xF@P%U~PZ42Va2-s~Sh}YcYRPO^x~7Us#mLmn z4K$_URP)eC$-{nbi_3Uq!+_^ml@*3p?9!kCSx;&CU6p=`^z)NGKmj6badd(X3f+Zd zPVuD7n5R*5zaHMH%EB&x974l?p8OgmMMVA#vlOP}l6H|*DL(lRuA+^XnEkv`68edA z`t&rNctyosL*@F9)}KyeqlNYxT#obaV-GNR+BI1 zD`;WLi!sy8Hz!6<_f3MK5I=+^Ufm9g`+4a7+a{$+F3dC&e;yte*;?VzlOnn z=Kg5YKhPS|eVF6Q0M&ZJAnzAYKT8C@M~usCMfeU5VzNx5XEBvUG^O~nwEk|aaVD&F0m8k^`JAG5GXKm^n5P|wcVC2hh5Q=S4? zH}#h_qN7R~`k$rdE|Vp~zuIkd4nHb;_QZt)Bp$hHvKf4Ldr`lgxfAjKO>X%>d~w;t zx!?nrAeiZ_Z9`Dpl}zW+3jKm)BAkBzh^Cjsqaw0={l=j^D+{^(a$m#JoMTsH1la9Q zG$7xqGcMEyJhh4}!sZidM>w|oxirxC?(sKQYe$plX*2@6-$>VAtA_>?m~FHKKS_Y_ zy0^Rui^J1U6DbJzKtsPF{sac>2{0Y`rw8x;-$v4X;-nMB{`r~z^cuijWz_G8pfw2p z`*->Kwmr0vv6Uz2D?(vJ|8BEwBA|F?*}7cjf4fJ(A0{mD6k*ktD-yBAh~MBcnXbT_795!D)WyaoR)A^TI)NI!pE_%p zdzsj=6J>Yb1J?3h*}A|uil&R8f?CGKCGpj^ z3F}hBinyj*{^WWS{w*~Xm33`9q;VIl%^t8!9^F9FBAL{cN%(*hx*drx2g*a__dYW4!x*W}*rkg=RZ&&l{`2^Jq}-xAgkR}p%0D;``CgXDvV@mJbfKqOveK}LBLcvl_uYq| zqvWd@=_uo_YKJE@|MnqpRQB!g%dI)$UG6LqUk;f#95y63NA5A&92vLBGO_>?XP>jK z!Lx=~Anoe8-Nr|F^5@K-AGTah1q%-{b$sJVxkn5SHMWV;fT5#kjgMXZYeBS%<`{?6 zp^+oZX&VgbS7mTEX(frb0y}aFD(J$xJ zQ*$_fR|<}MpyyXoT#eJ+3+8jwjK?>Q!|YHP0P(zk3M6roj`}4$T!V9UXDV}X=LIlZ zeU_vUBdE_9Mq5k+zQNL*Ho;@#cVml6W^Io+%*_g?hY1SkBzT>Umo0X#G;y}WnLH2k zo2$Wp6w`PcwHmHi%dJ;auBLLb{(Ny|JbncIX{O!E;cZXx%0l^E4cY3!c{_5os?nV; zz8J5P*oqe|eyv4cX|190eV=hyxRCvDLffKc%bQAdj?$#BRBP2Z;lrCN9vwU4K9;G@ z>=|raI5nz(w;J`@uiSHp#AfrWZ+!Nn8#rO9xmjuG)C`NpEXvvdDq9du4OC)HwCg~PF8wt0W8`o~O~ zhR8RWV_bMAh{H`62*g$;$tIv%e4B{~sT-+y`$YGBBZSQ%JkRMY@o} zmYeoIW}0Hv)I-xp>xzl*rr`sb8iUc@NC~Jzs3ap&N4wzfgG)_qNLeM^kYuOgtr6x~ z>wyXds#wh9Wi}Qu5sfSC84@@OX99I`I%(z`Qb#J1+MX#d(4DYp;xx=<7XGNwc??06 zb;<0^;n2+iA;$u=)BMAGy(z?p(-cvW;&3)YcbojGHW$WxxZ~QxBaYSEXzLn-JA={X zm5rpAzxFyB7SfX|jJwSFjVl(LedKRD!(14njxK%z@ocL2$R83P&XT#zN2&3%-!Qn^ z6PDw2T%|Jz_MpOBcdjK&9@ZGHhfXTq+H(pddx3!oRxEd16)^Vh@(*!J?sGsiaHB>q z4nCA#*h6P~;DU{i`=p7+vCULFx~)2w{#ZayJ{lS9gk0sjUc9!cmWS?mwc$1f@{N6J zemW;_{`rh<>*wl-lb(1SkZT=)r8D+JuX^!Opj#hIRB{AxzRcBzbaFT-9IQK>lgv&& zNZ3#kj|T<;2?z)XfyGWv%$z?|S5nb>T(ztLV2QP>j?M{LpHV2|#|Kqiz`aPR=eO@7$Y*UMr1h2j^>)9&5yR%#o$oZk76l0Nhy8lgvF6GtCum3 zJ|KJtlkH}8KR#>2apr8UhFwyIZ8K7XJzM+}04B@ccJQ)riT8vzQq>S5H1K%3(y5t9 zELO%s{(x(*h=YLS2z2p!3Mi5h0!gGrGcyq*)yaqnp*zZrR*d$sC)0!&a(mNk8rl#b&7t~gdR<98f5hf*ius7?&;=16tGBdk7 zM=;trRB|!GX}tL*g8K{DSEF2RM0b{x{+A}$k*`MG$1&ynJ_F@2_M@G6fDvSPOw5>= zBut#eghQ6jZ+)`Cn9HDe3D8hJ)1T8LNeO6U>F`|7JyWO=txJ>M_cNlNh4;^`I2yk@ zTUwss;Xm*h2qEkx@b#*T{LK|p|)sI|LhZ^Fuz#5zLLnAIX--Snx#!?k}Q&F2Uqi)m^%m>EZZ#qTv-4-VU)}^-M$c{CR z#J@Q(o=0YCb)V^NCfFtte2K_Fhfgn3NMvnuFiC7|TI-@#`y`MXN3KNEo0BH=y}_n3 z8Fp0)%7PD-;OLcjWv9xnNI$A2`Mf?9e^ondoAqAU$T0>(Zy5a-z->9v1_ut*SqUG ztAUJnLTi;MSQe7FB{%mQtu-9;7>2J%Qy5TEM@}kVqEKHlJzSrUG`oR@xg53}yPFi< zqIt^(OFa{nfjq1;*j9;)a=W&1)}ZXxd7r7_}mF`cYdN#xqBSu7VhZRuag3>Qcfiahf5=0$OQ zT0Y}^Er|Hfe$oMuGz2o?-YQONYd#bw`7N89saJwQNQZ^ z^TS=LOmv3*m+uTcV|CXcN_GF-sgm`b<~lY zBDPZc)kk}?A3IocEl-@C@9WlD_Pf38V?qszvOBJPT2ijyG`C0pMuZ88NFdnQ0oKDy zkbw;JUnk|BCqQM3AGhIF*k>V0wWz3Q$4$*fMOpa*5cUsHS&}K#8@2j^CMLzi1G6eC zmJDgTd%MU;NUNuTrtC2p+U^HQ(@gl;oA$#v;^N}A-5-zWx<2$?1J~o7X=R;?Qx1^U z7P`Q8OLeM;7n=`{Td7`wcXGZ#3)?xb<7dg5kU4` z&pF@UAyn>LUhm=3+{6vcbtaE=7i`u|8pSPrJ?D~AeR_Y^Rzh_|_gM{{7A{ZoIc7$M zw9Km$=F*u>@jpBxd-Ffs8Xm^Mcc~^(zznRo{|aM;q@g!6T5DtdoD6@gLn!*H zg?g=>w_bB4gJ8Axx$>oV4b&uvHwH&Hv^Sm*Y=zOVa_S)AnYK17HgJhz>Sv#y6T2?8 zGf2Xzp2Jpiz)BEgudtO!m-)?J*IfQx*{n3U9W42_RdOeWyWZFltRA%99a*7W2-Hfe z+xv|FalRs>ul*4V!DeQ}rRp?_#(-N_RXz1#l2M9$X&@K|gDG#^grn)iQ^nzY&s8_o z_3Wy35scF~hE_&#GPKTs;TZgpPcoRW)p@&!$C7xOu zqWmWR_O!c8LFtd0SBJ@A&FlrB;dn6_EwPqV=l%F51ntb7P^S)~PI*h?uOc}gsFxI= z(8Ua*v2xUX(X=)ci8sduI2P+n?zj8?eg1;6l>f*Ac*ySbI7Hjm z`Ge_mvs|Uw?9+V1dHx7!Z}psDJethNSfylslY1;zt`J*%7yf144q_5y$9cNs`;q58 zCn`l_@_nD&S)nlcq}s^)jT!VeZa;Wb+qnfceAF>2gWj@n#itLgCE5DjUfXZQf-^U* zwN2WrswzmP<%4Ekoo)xh3Vn+Rye61b;Pdg5($!N3Yq#tVUMr*U{&m8ea7%Qi+gvtW zK(hkj-3GC6dy_#r<_GaCq&fC?T;zk^BRy@cvk1DQh3sd-X^Avd41FK>kd zqa5%0)+grvto7}*cgcav|A(=A46p1<^M_waDy~=+r;>_o+qPA)ZQHhO+p5?+LB$o@ z_W#tc`{}2rXS!!z<~q6d3D#MAuXV4_{jGkeRt}rgX!9Y~(PXaG1u#P2H{jBR&T;H- z6yzRMs%!e znbFm|-)sGle~ecq_l1n4kU|wIeg?zqj!#nUSSPm!V@Pzbl7RpHoHD=Ymiq8~yH~H3 z%WahDm&A&zOm`j_lEry)l~7)N2`-TWv$;;}D2XVOV;!zESGv88)J=0v>R(X z_qC5z<`YdXd#)tT2O|^Z(!JNb4x1DFJ>Z2z?xMS1@??)sel=HVG8Eq;sZUPx)P|5V zOIdUr4Gmg*2~o$yT(Dn;VBi}LJ}D1|2(*7*QFE=>YCQ4tGv4{2#tU&NO*o;epsEg1 znOw}VZ^m0QPW;8~vR-@u$Dw}PqTE4fUXoB?)!-ybQvN5q>g6|)_$%4i>+iw%dVWq}o%&EI#j^r_Dy=$pGi#nltH7Co7h36a}o~f*R$gQ*G9~_<%qw#BF-HK=kBr?BN(FJ^DoP<6oMtHy>C&p2Dio0 zZ{e(mUFfYV>sJ)qjsMh!mEZ|}^PIK=C+8M^4gYP?FQK=%d{T$0O7fsBMqobOMzsrg z(t&@J`>UuLU7q4e7lyF&5i^xS|GvyA#?};qy($D*6Vc1}Pn?Ytule?A9tJ&7$KO;n z6-P5&?K-IR$S!+R=aBq)BzV3g&=-{rlPVpOeqABXUd7m4%RJEMb{EYZaFu8qW_?&I znq2BR|8T|yIxilpOmc=QjpLOOe_XECrFl(Iktyb(FOAnwdh1A+Ka!Q_bL*hczOhiO zRWf}ZV@~~rgnyWkdIqnu1&S?eC#4G*5yWyu&|@dwnMM__#r4TC-_$HkD``-L4KCK>2xW)5UQ=RJRgoDpjC`KjZnUPLY(!k1 zeyCFx&huH_*9^UH$DQ8cOM1#e5*n1-_fu^2J37v#MoB&zq#Op6v9*R7YUF)M~XEuER0)p`>QXo7?4lt=`@>2v2uT z0RSw|cHMR#X;e7_z2dKryCt2T_u>=T*D3gOPB!m8S5In6YDzM)R?BN6TLfAf%KLL) zH#*TMEO5oK;x~GPec-MRsTK?W`nK`M`$Lb9%fm&w_uyFOvV)H2)!K1qlzaD>rcR2eEX?+_Z#1sw}Q_h?+*(6L)4KebSEF z*%A?rQwIg>5)toP@$qAy=wNX19=`Xf5r@KjuMpQY#C2qi_yKr-+8m=a{qFscWk7#K zu`OoX2IJy~`4xT-gHKgZ{Rj(t3hj1xP?+9{7gIP~@JL!$|v|_kU?M#IK=x=+3UZOl7jR*Gu-W1&E zU@oqbE{~xiQkUKYSWfkkCjAYt^4*0j8kky zL1Q@c*|!AB?nekXj5S}LD2yxM&u1x%E@~Cf;oGFuuC0Yv5=yW};j!30CABFpDlfEN zHQtYXjmymQyJ%A>13+>4*$cD`7-&ipL&mV-Msaj0PajK!!=|Dcl%*J(->6px; z|K<97cI{OnUmA|_h4INdvY`cJ{u4mfxR5@M`Yypw*-mnnYStoZ;lDlVwM zdpMt^v6&Gr(&pGb0rM+E33g~ZFF}pBaCkgjb#!#(OlPQD*8D^HfDphKm|yjDy~p+$ z;1Vnx%)fY>J@(R)*9?;4TMTV=i{SGJ#?t+KT*3COaC*(4?MizG*fiFyrr2$>>j8M0 z3qFJ>e0B^$WpQz_TD_zAscfoUCsnl-n3hpmo#TB^KCdv*=C-(QYVd-@_kGQe(Izwu zD%iv{sKEaUu`!=hWHIhLoK%Qo-qxkZL$l-T54#m}itd-M8)bITV016>wYV|&<48X2 zH~AUv!2>0^>lfBkN<@nBdb9Z?hVh=i%H6b6c3^atArc^Fx@>6;&hSns?|x-j(iT zl}8;ZPH*izCMh~$|Ctg`w$cG(U)|L9x(hr>li_CLr5@|%UK$rbB&As6(%*u zlq3=ZLs>bQ((1q~Uc{VN`+d5+KWIXJ+_Cc@x0h&0d&5+0AVPKmG)Aen zmzT5$XIq0`=8oEhHyF7_k;u_N$;NJ=^>tji4J<61zmer0urpo zs^dwVGZj~m@~P(iF{2>FU41Hz=fGp(p6c;gDX}X_#hwK2l-=t$}y`7u!;AH*F z!W8q3wXa*ZaPCam{ZE8-6lV3YNZ>xdwDZtag-d*pXL|_@`sKn(( z0>*ic;R?I^2R|HI@H}~V~KkRUL3bz*ekNJ}J zyKFINFrJipl?d={RP+K>pI@uJNDjIct5Scet;*#lh{7Y-~H?8spu&5O&u3^n-Os9mM9xtN@< zYF=+yq}szVat%F5WV$c6+AF6ayYK_>@ju@QAah%{r9-HH+>YxKMN1pfWiYvy0v`{y zFe%owa|MB0Le!MH#u9fk<0k2X9ZW3S%Ls;@QMwUlb_Zgeq#PmT7#FX02Ol0w@a&NB z$~oO}OIi-pDU@-zkFLkkNKL^j+sMk$j^b2g^_tVD71|W5DQ2`>&%F^oJ=;^f839{$ z7Kn_ta05XiEw?qQEJQ33X1Y#I?wHFj1>rAa<<(__9;J&@cwXC2WA5m< zmHPr_aGDJ~q?9i`vOKk9zSv0bP|WeEW8I^c*{%R-iyW^Rb^FH`Sp+>DqFrGr`z7`z zouEDXYt$23zjVrfUvMXmBj;T-^2tGrL(zM3Q9%oDT@m8t~^&no!8wv z$2mCGHQ$;F)taGAn)P1xAc zqWE5op3C-7#lWn!avQ$s=ylJVNoi5xA)&rY6y2pN&85bxQ%VMGV0w0bP{i~7Q3Nhc z$0>ua*!biLpS{qk##$)q#Z_K92L$Pg8;15uU< zqd#4#-PhNL;5&D2Y_`Zdyk1we-riTViyEa$VUi++9RzgDSdZhZxlP^Y=aYW(`C?9V zshC+6Jq0>?YOGFSqw$uUhF9uuU4#jckv03!@bIv7ohb58j{top(_uK{f@eBdgO`WI zx0j-iBn@In);{?a#W)x@T0|x{v0u5Es+(_pv1u`po%><*5xRyTm=kD^Zz&fL zB;o_$Q}5%vFW$2BCrhhL*1NHMZHrV(kA8^N^-7g@ay?r!=~w$3Q4tlS*bF7=TW|i8&iWke}oiwBk8Y}iyF&sE|czOqu43j^amg7 ziX&aI4sbzeUJ@Vgp2e_9@C|cHA$ErWItW_o2|Fq0ajd?#K@=3+L24`7hJE zNJQ;orlHX7O(rNQ=yV*#KHJLRF@6E487zXw6q7&7N!)N=mN+$<*k4OtxVdr{Gs4u@ zmJt!sV&?|nWM`>A-CsXc;A+fn*1mtvm;-K>{sc!R*xP0M$zuJ1%j5Z#R7P5Qa5FO# z=R)|u`zaFh|ItsK(h#q!e_sMdweBo!bb0L90Q}v!LhAG$pGQe^Tv=sO>GF%(!MedN0m|pq~7lR*|m%cUnaxJf#qc zl8+##bMb!S!+K67SMsCdpalSNANq4b*L?v@TPP2+u1De;exdRs)y$$glVuD9V7bpw zj#mj4Y%PlTEx}*y90ZYLPzgtFpv!<)4U@}MuBR6Ql2uz-8R=!c(CK3qATP|9`_)ZdVeK>DEsh zgFDK<*9!H}pil}`PkZBDi^NJB>T z(_Kt#B&|nZ5`Y$i!#?K;d9hbsu>jpSZ82LrjOX!1g7Q>^)q0H$818<<5$@dn6{-~7 z=imA%MXi&P2oUdsPDoVt7IcxXPzSwHG}e7mP{1w8z$?~$uWR@F?z8@o7(^A*OXIH` zpMn;mLvohuYgC*(y#EYF3|vU6d2wtT z?=+fu#4a!Vjy;?R(MMexN;h3IAx~5Wdy9K(pFqUKIsxUg{}#T^*~~$5-y(iF>O_zm zVq2zKb;1NcQ8J1YsD>7Og+^$sBv7NJw$p|a1;-peIN7#+HVvEf-jaNfTVnxrFs|wY z-VC)Yuc#Q&EbIpMGRPgai%uRP7ejGw7NHEVw zVRo#G*1eVKcVHngR*8LbTY>=8PHiM~juY;;h8FUL8HfL5B0$5Iq& zqCi>{+Zha_p*>4ONSFRO$LD!BXF&nnm_dSY7N-1z#8vnz*YlZf-2M>wDm`u4X_C5b zhY(Hmn(0IbiJieDtPI6vS3bBQ$MLo0xRvK0EJNr{zdh^i)aLcO7d>z=LfRp-l?a+sk&}yyOPZ~KY@Ej`0z0??>jSlR z*G37v+Bar^nx58rix-*!?&*<{uMD1IfT>m$dwYgz@D`VN=8RgcTS~gRvWk)Q*XK=j zD;C(<>FkynoH&h1`c8a_^nKx+7oIl$kbXNRj}ol}M$f}o5%TR;H-ZEHE$GOTswc?r z!diK~MAhPLMt2BrUMAK=u<*to4;|QZ1*pHw1OS04nX<9j7s6Jlxbg?(@ekI#pR>kx zyrbO&X0=luELX>F!BZl0F6h5N+9Ke1J%+6K02Q+9Ecj$`h`WtAJTf-E}yL&x3L1W-hR^ow!S*Jg5${h86mk}vKB zc>5u(NfANwetQ3dLm^)zWpLJuy>2(fQv5?zZ6*3P;>(hx5NO&NGkI7aVodL5;95GH z?B@|B-9`HeiX1ZkiLsgrFU*A*wEAtpg2`ShI=?=_7#I(hqZd{d!_Qt?)xg}tdsBA^{@VfN~e!Pfg$GV zV6QTC3FOuk^fXfa*9dre&aA9<{J^_0e9qjT%_DpyL1Q)2L%RT4n6H+1CZVrjvzN;zId?J(|G$%hN4U)ymkEu5dT-F{50*C&VLrP7dVrlv zUD>z0zob-c&A)|G(v+dhw-dbCX*^cXq7}U6E2b4KZeaLDB_*s@g-O5zd`-oMH{(8f zi(1l_hYL68t)PnKZRJdEoIZ!NVVhuWhc}0W9sDh;Nj=+eBo|I3Ob`RQV`_;3koJSx zI;s6L@zK$Wt)VJ(8(MrW6@_EZGx1r=H0b!#c-=Sck9?%XZwwTNj(M=b8|QZGLXBDh zaT?DP*qLxf5Go;I3taW-3Oez+jo!q7)_=xTjfoxrFwJVF$D#ei7wxCeDO(+ z_>h%in{>BS#7Dx>)a?M=y*(L;?yX$diIqFAgP@Y#4%YCy6jKoQW{;9^F>sXND?~ zc*B7agazpcS!|Ci49keE$@JEP7gc7l0#x>X;5 z0w0aH(c|~G7msMRw`t%$RUI^^XXABi92d8I{+qdoSzK@0H7faRp ziv?7*Tfr42>5eqZ^20c2?jG5*-UFv><|<>2{@A(Nnd%Pon@OXF#@g*aseV^X^Ukc; z2PA?3;|P5f`74rO8oqg9RR=gwW(fH-__Un&qQh-j^No!t7U%Rc1wG1 zXsAT?FrJ6zo0pWGPbVw^>(Wkg+A7V5Jy{(o7*qXGxNXqQ%bp^Q! z-g(?=Ab^RVhr4a{Fmk}|t>ZUivrpfzcWu$zU=$`;<~k%Wg^>Fb=mNYfeGFx-9c~d> z0ik-_6P6ToL&z~=Kl{2aPuhXAq7HJ&AB^?pyE@; zl&p~`km%M%LU|I=LWjop(C8P%^%bQDDiZ&$Gf}fp<#c`9;K!dREG63SrA-#e;F=TU z@?UB`mUMkO=8AJCR3&FaEF5bp$%Nh*=n7y-y!{fVgFO42DKRSw1%=Rg7=%VXkV_#I@ zS?jf9@E0gC6B$$KlkDWux3KQV?Fd_^i}jL+HH!lk6&3fUd8B^Pi5y=1I}%C6 zkY3kcZ!|s?ms-p?GbYUUX>#CSC6$cozeuVu+@3wckq8-QSkKGIiWcscj&D7NNAWM) z9uM|p!S!A*vv&)_4~M+^?$j|3ns~MQ}ff_kHD1of{QTIbrl~+RX+<^e5+^+LN)B#R7#SR2zvAi4T%y8I#{O(&tw|>?G;D3CkKiowIwoN;n z$*ZXDgh`IE>3U061q60H01HCQ&x%+Ms=8iH4|Ed^Zr83(=Xg%6UGIUTk-=WTV62{= zkB48aj>`(befJaU7Vp=q;SJY4V7|jjx7EvU{s6s~utP!V^!)@`p5@oZk}AP3+Sg0O zzo3|Jf-rP{>aNWuPihv+sKT+EU7>;eE2T22_*Y6L5Cx=E{1Nzwe7dj-OhY09U-N_X z^(AIgYD)tp(Rl`tA@UkA42_e{Rzt3n?<>TY)>>QTu`PD=tiKg3ezM@#!OMo@`kotw zpdzQs3i2S;NkZdEmSKCY2zF7*lf)rq9}+7asrfag>_9%%e{v^k)saox2U&&S4(h)#JHrs8|Dk|L+ z8y1D}6E_Tqp|o6(b8yV!4ZS@^{ilnPPi+vz={g^B{Nr%On2BE!s&#R=%+Ud&N_Z4x zYfTwxVC`_2%yvVaBG7SH0)b6h{XYf&fu?4Q|3@@M2$9QMbWTWGJbXbZg4!Ng|O1>`-bVwJgc6 zyY+J1YB!0)CaerJeW+Gig%_Aa5yIS`<1+H2oSp7%{SLT$eq9Hghd?3v-4OkC#7@(q zU6aj2#!zav{vo;R%#{<^dFRNHuzXtm48LwweL8c4^{w&j($?piyfjf354gcpeF>Fq z4Z}5n*c$0!H{7qJXFH*yf2sW`2E7~qXS4!5e0BT0E!9vi$w316drp^Up%g~dXwtL$WG;R= z0_awu|Bbmq%lOM&p*aCYSDlPrM|ulHFtfC3(z)G@H7&f^AIFZ?+-hlN8F&wa`S&8j zP8RoOeU6ga8HR#<<`(`(IHgtkPjIT-NX>I)Uh`wyyizC);a}`kPp$sVwe&gqO>a*- z|L2soL08S*QaP=2bF0!gW{`dotpZ%uuI75RxU8)$B(1sfrpeLLTm20`jeW%qVV)Ph zrE&<;G0$9Lu`EzoJ2TX#Y`-xAxUDu@fC2GmG^|*vq+ejq~?x~xop%}MQNab zgC~9%_BP2UviCOHx+$JK@Zu(|O;B?s-B)Av``~9@r-wq04lOx_JOFk<0ihGtfEMLK zoBVzptpD+&Rh0jfC~%FliFmj>CJxy(u9vaQ4Iyptd2Ty=y(tSo+oGUI+d%SS!R1f(Re|4SG;oA$}iKN%o zKf#ag!1|T=+cpVI&totbB8{{}*Ab!FYQJkjKeoKu!@>UEvf>n)>rZ|mM##(7Pl3#W ztd*JLj}_kO7gWF7OD3=v;wTj5et4=TK5_7p-a)E<#yyT=QI7JMLVWN*|l%VFu zgzioR5paEF>%DKlCzw#NNdu-|!5eK%7<8ng?I*7hXU|oeBYM~IjWw3*5WKgP{sU8G zasGv=a^_jGSu6jmy1H@ztz1yo5~`UWgk1JVb+^dau~FlWk>a!JXRo+SVYP&{PI4v& zKEV(Y{`us=$1~IBg9^3?{JBS%l7i~Am^KMhcNZ_qKpi$xlji;h1ZLsgH<5X$`4Bu38|ulr3=2RyG|c!-xsKc_FTi23>? zx#b`hL;%$k@JJb-uc0(PUy0;jUwH?Ade$c0Ez4OBU_tTQ|I9TwXj8e6Uh+D?k@SS} zo#WMqx5^4gs$5)UT4uq!c2^G#dD!6=Sa^ERhcSv)hGH0F;${9E(ZBT@!(=A2Y-|E#Ha+S+x@VpBW%t<_6mL0YpiunQ64 z999h@-Xc)Wl`SP}TWFuvfXO5B* z78MWv#9^6oTysH5;jmtD-M5>%#I-F>exWgTvy;R6j|J;0vHc+NOS13U&A6ao7NTRM~`! zT3mlnY4efT%4Ci3ZU*3R#EEb-t}#P*{4Yzz=}DW01WMAAp662bDDWFJaFa*N=c!WX`uo`%1Nq9% zHX&cV@j)W9mu9$EMdnX}Qvt)t=!WbYE0U+zV@lQPTX8I4fjHpAdF5ZE6mBxOt>5O; z8KmCzEayQqtjht!C>q4_ik|AjSgHqiP1a>}5YAN5f9J!9q(*WK;%v=ftD zVeNBV((u^T3+**jX|WcrH#oK+0qO2GcLNVvdg57;`I)T9*3PPlWBC{?!#+nP)VKBX zZ*TYjer$xA9lI#5)_&x0XznU5s!+6kBu>6+c)OI;q-Xe!b{kqlWl*raV-515$w;AY zq8tAam+xo$ON7^xW{$fYaTovX{~0mDS~44VqI2zRtzg1*9LZCL3!PC-j0mD;;m$RK zn_81YU&2RjFMpZ}vTNQD*^x@+O^kncZ!B0U9x zQ^2NeNyT5Y&%gVC&zOULw`j(~#TEMxb3mxC%^z8beeS6IAKy|#+ShPWrJw-pud3|7 zn|3bYfd0~}KdTy8NBg%fG?cFb8VIn=9UTe&;|$jQ4XC+nS^>)c@hyR&hw;psQPJ@K z5r+O`{VcY43+K{K@gLu^iUc_PD`sc?{@pJ~@Mp$-4E9}QO(_EY^|zq^uAJf&f-JP8 zYUrXa%T!QGOMP1ZVoWyfMe_Zb$0SZ2wGn;5si*{~GHV75gY1ID831jqAU;>hzY2-; z7!tIk4uF6&evDqL&{lYnp+v;aEETqe$aXtpq#(1EYcOP7stc(_Lg$tN;aaTxgw1UV zEs?NA#m=>^Kn_}}+ifn^Qe7XuYHlQT_qTE4(>^b3_qj3Fv;zeQ&#~0icxOnn#;R@8 z1S@r^?L3F*W&A~6l?x-a`N%$6!GIQr?5JoRM%yk?7S>>N3$jYA46OqOYyL_vF0)av zFJaZJM;<{lt8gtT-AJ~?{P!9S{U%o?U^Mb9FsKst5@~AGI zpT^i3x*X-j7>{%1 z1cimvEWP}j2eC-{nsLz-*^*#2GJ01^=f9Osuo?C?eKYJ6HO?QcA7a);+}{bCRj1F_ z^QrhHn#9WZRaos zlx0pLLLw%&GSN-8^EV`IGV|^_{{)@Z)nNrwvcb%W=bs4Si+*K}MOPNE;LVJH$^Dj< z3FJ1M!rqX}Xb0H0$YBJj5s>Dn5QnT=G}Kh4ej+ zF8%(IKsNEBM8rSOT%o=Q(2rW|W_6WMKLZe6accg(V(K@X1a6p0XEGs)pRF`ej;2?Q z)JObATu2UNmxw-{a(_@SR=Z09mSzo2_(6fltxWJSpeZ8Z#07W4<2q^%X6PR-sWm*8 z^z2i&#|;K?4y^%Su9|2(2P{OVH#r#)Lw>`zYNi)3wfa z!rXJ!0FU{bRK^D+hlnNDv#gkI?R#!mx}}&*#0jIppH1St0mPl zvF2&;3gqzfP?6j>Oi|aZ~c<1W4`Ql;&<^W@H0LwXP zpw3yp;StAvovIKMa{uXFczB}q8C|gz{;p85Y0}0{_w;j^^Xos8C7AFD3E=S)ysoWt zVMY|9KL0)YD2?#k)w#b`2?U^7DZT{2j0xQ$Rc$wA4Gk=NgyalmWo3_tbL}3zw>cw8 zR?AaWbXIvp%T0!b~*|nN=~P8K+%dHs@3TDD3#V5s3HY5`KXx} zb$JwYWIFda>|N#^$h5sXxuBzy!eg_Rq5N-4kQ;rEWZ&u? zl|9$A1FpYc=oj_C3w_1W@+v2nhbXEla@L(a_)O@ZB+UzY^nhnXf8@X&Hti5`xC_sR+hRomoC1%x;nj!=IpggFELt7%75VKsFNx1v>^lHicz+d)?F;xRG#t_Ffn1S$jCRV%OV9t$-=1=<& zFop=nfUJ(p z?@PPYk8srSII|U0-Mz|_=RPuLJV-~CD2mzGHnH|N(ks-A1=0A(_QAFz9;dUTA^|?d zs5n{`%bPhpgfyAJ)!Lc%0=Mj=`dM{wd~5S^*9BBK786U_^B6XBAtT#?W?+_&nYE## zKO0_#Jh9M`*QzIn*=DjGQkhnmnIkTmt783mp%wm5oj3qq(9t$lDL+s}@=A zJEe?!E2Z`{Z|T07O8Z-OdnDh%mR3HC2bhh7%DI#+q(+cSUq4sPT6l9s?{ajJC>7W5vL9#U&IYjwKPGch&VY<3eLhYV#!taf=k%od0S z`1@lqo6qMvMNp1!ur=}8e^3H+`2!w~7b-;0qdKo5UyJ)-awY3|-yecPH;ARknKfBB znoJ6su@sN2N@g({gFqIgns(xUIg?vRkcQACmv4(E=P zi$UbUbkCaoz*Ni;?*D=v>`5i_E!mCdTboQQ185lg$4b{xUnz5jiVv~|dN7iF}kuaof*S~kien?|+Z^c+G2U?4) z+@I_tBPP=*gK}oKFPyy$JTa1HQbO)ZysuQo#|Xyok9xmSBYn#z6=|$i933oRo0S&f zQ|EfJO?oSd7og0;-m_#9STg5R7r>x-if_?K-N84r^M0<*S;|v(NdXbe1Wh*-<1a0& z3}Sp;;mT#mom=Fmv5%+WrWGHH>()XIQ}o>aMWg5E?qU#}%;_d2DLr_kSZ2+={5wj} z*BbrR{}^UzjF9ykH}fEEtC#?L`D{v}t$bPda;I5nM!~5<;u(nvQ~5dCAwhijU{J)U zeP{@KVvG-@r@SvHYOXIgj46e^g1MeF&S*EnZ55IR(Y~XUpNFN{`a}$M0B3-j7!Yd>{Oi zFycA4Wj7*`$e#nmlF6QrTOq{Vd!n4kBhx6o3+;~B3fv!b zef7FdR3BEG-yqjoR|jB)`A-gtkL&Zlc$*>~2atgg!otTx7v%MCPOdACr%XYJzV+zZ zB&Vy}=JW(KkfT*)(|>A5MyJn@uer2!QLPv;zcvAhc%m9pDD9{lyigi6lV6rYxmpFdlbT3p4kh(?vQ9ImK@6T zF5#ZOG)H<6W#+63Nc01fA=1GT3^q4NrJL(}#v?5lqc4*FA~P#+7TIAPp^Tj}Ru`w8 z`TZu-lFN2U0sdfMubG7cfp<<{RkbQ)eGMJ^U1+eHt}_DHQ=&Z6VPGBf8ble9g3}8J65*b8@XL#8*Vku_%LH1 z#i@tXt>F3C!#ksTs?gnzbeK;T&F1&>F9lnr8?GNg25ZXZfGvW*R!Qk!3S!C?;H(cCmz1wu9K>B4jv)e$05e`d2Ks zZ?Rb%FWk*tjIuS-HkT6A$Xx`6@R>|y$x^F$>z%A&6rz~zvvrH$=W98Aee?s4;!ogX zVtkBo2?|~Q^Uz9}nIEICz`WHm>kxPxg^*25U~#n|hDSqA?$6<>JPaA7Uy)H!cehub z+hL@1gfrWN5mo)^*_PlC~>wK_;E8%p#`xi$(I! zx}2EOT7c!uud#QX(y3U5RKo8zYdxkcl@4ciZqIcZ8@*1p>oyZHszdNS&hXMQvq!h3 zEPIDLAuJ!q$~-g)5!Va~&DH$lYq_z-hi(!LtlH5WZSBAVz@KE*r{>da!HNf%01i>S zHcWZVmXjQJvkdd%M?Mk-P15oaka>nef-n&gb2K=zkt^suXC~iB7tI07hL@8`p3Is) zFiox#23Z&dIYF6(N48qW)kJmUKC+@0c3zxIw+klNk)^v4WHBN(?Y0tz#REgt@9bo(%Q&z~6#-`Kt0S8Fps}#m z)A&!;_l9|n1xLZ1oZ#R*bh}8?zQyc>pJY`c1tPe)^8G%_ZbqT12eTx6wu90}SksYH zg_NRxTutl{ov6TEvJSP4rD1U$Aia-_5AsSE*!Disq@5&trJ|$I!Y~FEC{?&cIBP@h za|Y#KE}f<&q$LHI)-#%~V@T?HUa!v*l?5G!DYRF`q6U z_HX}o|5oC3!pSa*A!mwLKSWaPaJkz4Fj6)NnbeHXFB;{Pz{o9C_88Q(!C6&QvGj%W zL{+p9NBjA1O=S|hWSzdK_-AGHH6AqNwqGFaa4$8A(B+9|?n&^_C8ovT*NtE-=*@lVA+u}1NL zd^06Zq_|&e^AH;KRIXf#z{&7Py&ImF{C;+{0>Q6L8zKBs` zt#cJb)=t)`h23%*E^pDu{)Iy-B{DZ>lSqbR zV_v_d(UHr+3Cq&99eI3a3=VhQ#|iBYk+DhLB2LAJ*0d>mQ0?7M`~rY0uzg7#fmYa22s{yj z)>sS2*TV|1!wdb}(&mNNO!`BvZHs-Dd=|vT(azY?fgcf0eT=#802YwDjNDk-P~L{) zn};aO*=l=+zP)trtLaQ-?^rprs``+_wPN=QTVBt|FcjKp4SN>B-QzV_;-Y@evpwCZ zQ5lTV+O(9e%Zll_=ntYboS>EzGK$a*%(~l*CkUeQc9SE+D3uP$=@B8bR2>52&YPp` z{<8DzzG+F+5q+e}wYTzut)3_O z{GkRfGmOaie%cHaM1bYo2x}{_8bJ=pTa)f(SZcg7xX?6g5Cbh5l0R^>x>Uu{{G=M^ zR%d@^ULe8~=rf8{T^@Eier|0I&^eBb7;S*Q&fn`usp9nhCK|{LY`=MJSWCHl5b@VE ze5&^6*_fd%QCvbHJ=3Jv=uZw>SQrYa27JquJgf7mnx|HWGxbJ=N)2#F3fVNddaA+c zM5R$pw6I<@2Jaa4#D!TT%soB1n=S6~=^!VCTKoQY0K)mNIxrB$30~ld{c#15Brv%v z6ns{z6~i>sEWV0+U_Oum3}`{}g2wdEMs8>_JH*Lg0qZTk+ze(TCST1|#fFBFd@_Tb za5q8@P~0DzdLyV)15B;0{uSWC5VKEOir?LB%W6fb$aEC8y zOcq2~jVSRpYhgE(-df}bB&{Q5?ruvR zR__OvY4Rx0clbxu)IDK09i1l-^VTcW+A!paau4u5BN|`&lH~*qK8$*-c^&zgtbfnl zy{nuTOQ|Xb!vz{pclh}h@1CuC{_LmqTr{7-en9Xkx!Ii#(rhJj-h6-g;M#MAl)ev) zIHj};)HgCM_N&V3_IkKt6=})7d3=-9wNM?G&sBav;H{oqbUP5!`Pw>KUz09A(zaYG z5?(xDujefvBbtqO_j;N~>$&VWf%9!$H8IxCb;%~y1@$R?^imm9zCRr}G}9<`+^sMB zT=2D@rvg53by*RSa@elMbv+y$b)Gm_ouo1!aOCSjrs9)oy|;;n(MTO&eB=tu+xT_pUwmp?BEtijPIt(<0E}nWk)ApK z4mMUd00Mr*udLRKWBTJO-{Y;ac-r%BCLLG^*$P{P*i0)=o8IO(I|&R+#q6`C%3@yk zcL8CUNtJ~GKR!gQH(1z;&0M@bqBgg(br!EHnZ7uHXh2I)%gtC);_)Sx=N77J4j;j6 zzT3?ZbtVjPN=Q6I$#nFujtkGM+Adm zV^(tRh3H*M@l~xO-sWInKGcXb=F}Qc+@!8!fvI}Eb6Uw1D|#LRcQW7TEgh3j zCaY*(^6Y2Qk@M=9&`D|k5un{~C-+66d&WFSaaWf#JR0v8iBR+Ijy83pjK6Ka7{94h z&1qhRBYD9$#Pw8T zN9pCqyUbK>QEtx`CNluW2TWrdSSwID-GA0khdXiOdYJ(5!;jJwR~GIzvhG`khqB@A z9_Tb#Xj{GanFA8q{%mW>@q;(oH))NjGLp(>4}erCSDhLvVL@cMtCF8xIZ%4#5cohv4q+?jGFTlH~~xeyeGdkuF+rxdw!c&Z&%z2 zmT)5;jpQfCngHZu!ZPFaj{vtFb0#kpo#5kD+M4PX(7~&gI^Nt5i!}{pqA^^l4gUG> z-?*#c6?vp}=QRg`G(R^VeptNZeA0>`v!aRCDz7LH+^y`DSgy2I|G8hXwbY%vYrY|F z`uwAA#$`ZeO=qT0R-^g~Vb-8GGCW-Bp_z0$_?EW_!IpuCxs0a7PD?t8BeQN5YuN3< zf=0+YIodMP4C@3D`HkRSnjm-Ydiz7lyVM~0=+ThZIvRB-Thffe)9XSSNK zu(PfZ|GZxdEysCaKfX}s!{BDI-Q3LVG4i-~k&TI7S6MA_2r5k&90m4LMc3e{zBP0= z*!c1)$db-5G%RIvlPc|liqqEmdC-+Gnj`GY7F`hwcM{bL&V@p*=sW(-B8|&g!OCg5 zC`sY)r$&_|?-Z1-PR(9Uy`fBr%#H~18@OV%pILx4nixFHx<6g&acFHwVww_3x(9Ru za%$gGGv>2!2~ZRSegp*Qf4Y<|1A{6&J2IC-&}#wK-3Nk~7~~XjITk z`nbUVb>KvAqWhQ~P@~_Q2VL@XIZmQvb9lo9Tp9Bh^qr=@NiH_ZVH3p*lo$ZC{gzEH2+Jp8P1$tQkx0gD(-aQ$GHTC~I` zsKbAnCfGxpyX|~9*2Dj?zF$4Gni}HXV@F8hLS3GEe+)}Cn@kq>F2|8th7o5OI|3@R z?zRrvrOg%OLT+C}-&u9Qf`wQ(&M}(Pke2d+gd^@N;qkmX#wp`N<}O6LeD6`ZW|XVs zdY|uMTglKOIvN^rciNYV5tbrdS?AG#XdO6VRCMiv%n*q{EN+wcrsU)tNM0Pw4Q1+u z&TpH-6F=5Nb+vWzelqAL=ton%m-xcI2ZnA*@I^;%^io`EmS^z`UJ!Z$hbJ$%43sFCvt~m3qh0&KwH+w+rAN(y~%v}Y>^j? z2}Uks62hJl$pv&<-OK%>bp8A!uP1D2|GT6_taQ_9mv+7JakFhhTw?tF-oc}M7ZAQ^ z?y8hXl&N$nXnZgLL+ zOEWfRgE*E!VNjo$o&D5Ppi?4~^$PPN(+~l*B6>i*l8blnfB~E=FILKzLecY%0jh-T zkss8{UnqAwCuD&JSmjF_OeUDx{gjBT5ltKIetvn!JgB+Zue+P4I_@`vTfzEZeLvqQ zn&x~Z@&pUW4%!`W;&?hW-;aw^c&^Nr0qy-8j&4q_gRB9^~(^+Y-gFKsm(PeA!Sz1#y?d&Ft-Os+O2j z<_gO{3!FJ>A;H9QTw^wBefSUtQ@LWu{G{)Gdqa_G1uPa5vcu^i`Xy^hXVs3Aj|H~wwXl#)GA8S-{Hm#W4}fq2 zO7B`?mE%!$SwiLPq+jJpPVsDZ&2WFZ$YQ1W@eXeP=K;`w=N+v?1G_513`2GgK7)RH zK)`vze5Rpi8X}QM2IG((_^T6tGf<}DK$*Kf!w*B1hgvXX`kRp)x{}*;j>S>$r}Ok* zqP93ijlbq{SeXX2G`qN%50x5+h?lR?OBd69_-(m6YqxXvQ@QP7HZKGO6~5jF8l;DHE!Rbx zWZJ46`%1sp3{?j>N=KOtQ87goQKog6vvPQQ;w~? zIXgMY!zr2!xOKjz3vrX#y=!Ii%6aqCZRJ5DU{qiqEh05`o&<`?w1oDZNDEwn4qYW^ zp+H~TZ{MXN_!Ph_46-!}2Sq0PmyqtBDc{7YGbOhp?9_qh z=#}DXVOE`;_JVhXy|p0G9U`l%tqsYzf4{6%b>+hToE7vT%3(Ry?rh`p@hEDmm;TTh z37zAE{iTA&?Z)Ps2ASNdD3CvQta$z@<$SkZf&nv9w>VhzcAUDAk^{s~t~IDGv{yr1*&^CweCCsP>rzvVgKq?8)`4rn7n*!C&+9^UbNNv=^MSMWdW; z?Qa0j=Y2;wKc;PFN=+p8k>ooo8T!_`NjYSh>E(PCf8mQGRVuq(mw#ql! zaG`^+u)MBAi<2Y5VtZB1@RvWjrPG|`(N&82ep+M=cR$;S=~X6NP&3%gH<-$VE1-^q zO=gHO9m5%(YfzI-L7qh-J7~o$f)&6v&6ZzN6#4C{~`D zTjJH(?QhJmkC|rv<~Uu>5b}y zn-I$sJZqKfUgEiuC&6n&ab;*ZToqKtLqC)c3Cnr`*PNumiMVq>_+qjA#{gIw9iqavQ#j;R; znSo&om+63DGUMUs{&+Ic_!;!5TB4QDs^!Ff31mJOhGnJ?Q4LT%qES>} zzDuAg-sw7SdFDCHF-xrdw2Q=?d%J-wTcTd^mO)zTVD88Ea&Hq&at+^-tYd;%h*|Uf zQNwg$liPe7zdEX}Gi^rYNc3XhhIYm|1yM zI0MN=;}+ffhS?8p2W=VN8xi=$}e*I@*p%DAH+GK3JIfLq8#QzB96L_FVX^M?-e#JDc)f-1z@GS=l4I}zEH>IY_np#|(n63q)3pH1I~c3e$-*pi##r`K zQCW)Z#%zKj#bf@Cp6z|AO=w*ptYgEo9pO{}(af5o6N4Y_MhyRlmQUS?4LJT3@`pvfN0`N;W@u&!*$K^>DnKNiVTM7h9 zSE;I7q0+MU!ik*JW0XELP8S~dnFP4zI{9NVvtqWCS&AtI~6X<=M89^nUY#duoXJFm6tA2O+wK*Q!do|)qCkhzkEt-FZ5gv3a+ zqvK)z*t4a?L=2|Dp4DC2BUNQm7#rMXDtmOSUZSkM^>sfQ1Es75jxd#C9Vg9mhb|WsW*z?-VFH*|prw>pVo6Xe|~iBY3zL#Xs$r zx|9-1Dn#41-@Ki7Ja+aDgU3k;BlcIg9bQ~WcH$XR86Lw3nWgJ1ZG#3Enkff_u^E&lyEZDQcuD9r6xBU z$5J#BBlB{A!@NSaLOU$m#eRIWx4yX<@x-J^pBrQ=9LwKPH5=>DtaIt(MyDz{K+z72 zSj%_xB}3mg%1Pmm<%ex&?SG>u?OwyO6vXLxn4aX$&w_e;dn?HSnN#_7(N>FTp(eLp zl=g1j&YvS3*ECpmxV3drxj@PJJt;7PVzjm~`5AY@n4?eKmHI1HU6w>`cUAES;e!w8 zlHbJf*i;E7$+gm}`blNL?4jKqeDdIYr_f$sM8n>%P5$`TuR;gg!H>`FSR@}Hq@UZT z^A4D|1Jl$Ya%($LMExCdT3?BBi#*lun0B&9WoI0+HOG&<+ll}b7he}r?e^Cg`#Wag zC8_)MPJP%WAJP~CyB@C7RG%aeMUK|VoGX5 zN{Y61Fp6Fbm3tH`yljxzn7tzcOQ5x{4ja;0$Zm(%Aj;Z`w7tBrEsmTX#n?w7Yq_F2 zOO!>!q*;A0Xh$4Sn37eyKn1i5166QYkWNdA|MaP4Y9UY01y$h%-x` z?bJ9jW!ki(wkg|TletOusJp~Swr%Yk$|)@){YgyZA(IBA)5)6wEpbfdSe#Ht>$lIZ z#8C`yh`89=93!^1hSC%S)rXys+?*cb=BL|ZzoRi%E6n4^CKXl&lr^Cj#p!5S7Rs?Q z>fQ!A378>=k9R+O|6nB-+fA;)gQWV{j`iK=eJo2fv4Z6}y2`-#Aoj~i3&npqX`3k1 zMNmDV_b;Y)1|e+TP*$7=<~dG+)nUOnHP&@wf(^3O$<<7CBbAsn)gZf#2U#O~Y`&J? zyo3}`K~I&iS0z(wp`&IVopeCDwZ4)eoqw}>Q(^YKx4z`-uZ2Jq7Z zjgJVb`NdOxK|p>#@;`l?jegaUTJ8HV z_J{xGEv;pwBme20|2n3%8=MWPJ_3i{__`Y0F>4mzUTXs-W|Xrpsix{+Y0&chT#0>w ze=lA9J9ZA6Ci<-;t|Q|>mP$+CVP?&3ZDPy|!t}FEw^ef8wE5|Uti~!4w_J)pd{({{ z=$j(n9dKw^M1wdg_Ms+L-!amshfyc{aBOAVZ?#{m&kF5NO(Z}Pw;)nS5Z2lhk%F<2 z(vWXQ(N}IchfLZpkRdvQYxGm2-0dkW}huYBIHc!DDC@_TaGA!HIZ9;&fHi!);%QbEKle!_V;49_Sv*Q3HlRWQPf2J z>mTy^(V^c^!eBrUecOCH0=H!NS{z@(Yxc*au0DBw-Wy_z?l{pkedE#(d)mx z{;zk=ro9{!m%&8CKMnYwu@G#r9Z%j3CksGLlSNNKfXopA{Dg+?E!5a#pNO!gC#@*_ ztr`Fp>55toK-ttQjJr}Ifx2oz&yVzkgzMW=ysG&Mf%-$=G*31E_}%&30+`c*IuZcN z4|}m#=TPBL-r#sN{?kx&pfjFEGmXpbBOCc_M@PrUuh4&Z2rz>g7_o9{UCt`0A58Bo zTn}!3mbTp8jrHKsOu*S~{*CVXtw;C*QOKj_lWU+{Y**U8fB$}Wxepl#uK#rio3){~ zwz8?kN=j)E=(T>Le1CPQS!ueqx*-RY>9v?C!Qpgvb$53LB5OtZY9Z1eZnt}V-$2Tw z^8m|ss8XyE2p|#+Z+QQ>VAWI^nPpo*jns%{W^@{@i(bR1SCyh8xUT7s%LfU%ZgcDR zd@GP>G#UZ=frQs08l<56p_=TTNEKdVlTWfqA2;^{bv2uX*8$&E8hGh;ul`E0a^jza zycoumdps!7cv~iQ+ndY#f}F&*?r6rzHx3MO!qaqC5SB-MG|LTE%JXW@MBZDG8|+ti zgrL;K(h!c}`^T$&8^)EikX@T$*Yn5g6;+e6%j=1;A2lMDRhLbAPk%U?wya=S1;uVl zOQti`OP}GeM^7GhLu5^ijEZ$Cuqv!>EhO(F55HotSgP^qy?Tq!;c}!^TwhiFP-g_N zhY}-&=PPaeg+_R|fvTSKfSazG+w#%*U{-=Zj(#jxcL%o0Az}EOCxzASOz7J(qH)?E z&ZY}>#^ro6GMS8!l;RS%)5gVom3blH3SGdRz?~rQ5wf#m9`8>VcX(9f4oxPKj=Ns& zO-ZIgdf62yTHZB@9AYWzB!0ZT?Q>X8HJspV9|RDlzmuiaNQg=u%?kI+y?8HJC^^DJ zW{c@cDgO3_1x^Zfah#FI&Jg6pa+~Mo6Opy)bG^5o;gONmX?dBI@+s^NThr&~vxZS# zr(Nh1xtB`1bXOT3cI@`ekHfrWHSYC^-<&(n`5Q z?dGdVgVeTKH*|NezZ5$BetvdlK3z|#;t}R1D4Pug)jX~a>$+~)Urdp|)4G(OTjLTl zSgu8*zkf3+s;!;!--;uvq3&Krf zwOKBv>t+9gK*%wGSK>}L(rnf{?HfGlf^Ijvyqm~rudy_=GcJ+ty66* z0ymi|bQVfp#YwhhH+Mf@J{IBL3Ydg^cFqbY<}9;(E9HuEFqs^Fg0DI zdks{oPiME-ip^JiA$_uK0VSXiQE?zjAz;y5@7)pz7n?MIib{7>>!tUcOWvg)TMz$m z>Q~cX7yZj}IeMcm<#-GH+VuF==XY6I?VX*w-?=yGq*`?kV!~lDB{G?f{|*6#V9@OV zg`6=zTJ{dj7*efFLl^qEz+X+gK(0|IRiKT|A=Y88`YSX01!^CDj0YD zRjul`YD<@2TM_Y0$S>cnUw{wRNL&42rnoeQ=8hke-DO1>*!jSlVqzWjZ3>H)J0peTTuL@P`t66?nX^l2R|7KX>3JCmsIe9ENT zF1C00z2>%>V4LU%wZ)cpfJ?AS3jr{o=1YBodArfz$<1Z z*g(y+jRX>iO-Y^2crp`qBGoe|6!zLz8!u!kfCsqxc0qjDi803m8(DkvpnN$+WICOy zq2b#&EHS!1>nTJ;FC#Wene6MyX(rVv>Pf00yU+u0e!34%S);_6)ufb#XTrZF7YlXO zfh++OL2hvqrBu{>j?l%Latd{Gyd!^G|}@8Uc29}*j|pxuZtW&U7czk*<*1ngfCgm)Ext?%EwYEjHbe+awJmg7(J z%#a7>$?UF|Nb@)A@Ox&fBEXU_i+!H>du0ubiGX1aj`!hgEaFI%c2uOFY$(9o8LjBGbbdVvJzY<~Z@ zk8iIUeo@CQ`Au_bI%}>aQvhfjKAsqJ~4%@%)9os<~eH&6~^R zZt;ACcLm5@061d&spr709o$)_-*I&F(RCz^`$pLH(;$%sti_h`=5R7I?9Wx@Le-^| zz~8=t42jtj{jT)Y!>P|e{KR;Fsqp#>fuMk(;Pv6cHDE;xKcM~u z4BEA|wFZ~tk`JEPs%6hWml)Mj^@y@uqX{5k<0B(uYDx;VqBZN<<|gU~l%h7+U8nu& zq1)>Zo!&MsQEa>r9; z(PFv;(S>2SYt1m|c-&wk@#{^)K%A}J%j2)LfHUF&f=YbRpp)R^isLnwbABRJB_$- zaXemLn(9DhObnHBWsCZtQmWIW_(LXMIPpap*xEqI2|Bd;yw;_AxC6or)$kT*o9LMh z?@(M!hEo_Y-jS=yXJBf*Q{9{<&dSb`FV}8#MP!a1jmZtF=}~4>^I~&4o7|oF!9YMy z|NcDNtKNy#LGx+RuVh^g^i~O84qd=4%wS?>rbd0hcrW(cWCV}HaR)G+)z#J28i5>! z7Y0qWTvp2QRH3R`2mPgHrTN2tX$w3qSMa4F#Q>qyixELC>Rt!Ze&oP9_U3=xHVG#8|tw^zQO}tUQG&V~<$@Wjf zo-ZOcy>Wh&`V=vJ-*q~Tf2Xif5fg)RS7(ok1f{7F6gYpLLJ-_QfuIOAbv#us)?h{% zdy0FQV$NpkVy@7D|3UHtXp=ZoF1#O5QE~p5dQ>CDel~(w1aq6O-M~^3A!re((X1e8 z`Ke=M#Ad2MIWjWR>+YgXK1Znt(k(J7O0{AsR~kE0b-L1Kgr^8Eq_b1#z|z*XEC)l? z1-Xk26otiV9lr0mzmol&FcvLK3S6i#q%S&yGh-14X127o2DZfuAc2%Sl+%D$Sxh}W z-5+aKT4&1aj#WTnRk-X=!xg&%yihPI5*?`rvnkm(ZxQ5vcmfLM`SGMhYZ?^E{;lkTfkHPlnt~PiyyRHDSsRj#fxCaF--CU@1FuTbEa5V zH8s(Z0NQ032)@R<$qiX*FT!iLJxy(F~9~h8c_xy~C6eg0849wo>+@RHN zSY4Pq9(tM?cveUdI8H$(lkR+2p}B=;m`2z2w0j~ur=e0p4#pWwWEYf9Am9Gc%y!v% ziX05@dAf@(%j=MXW*#w$!2KWw9J|#c2zo4H6~AcP`AV}HZR|Zu5YSLHudL4V z)}Fa2#1Ga%#={ItLq`+MY_bf|;OpL-OPwZ<{E5>OXyDt|JhTPjDkX$+U@B)jM*v}w zLI3CXenL^*@p1S8p->dK8)_xIUm3WX71GJpb~$0xE{f|+PYJ;VX*7RQTZ>~D5shPX zm;fx@ULPY4B%2{ui@S?m4$+)t3S2gQl2lHo?l+(V{M?M0E!Yrr2^9(EkdP6Xhl>i? z;+cuHCwR$tYG@c}cc2C>%R}>PUE!mM&iORAyQC;|&+w>VjmQt|NX0PJl56PTpW|sW zpLInB{9cX=s?;hiWwRr)pXEZLFDVx5UR zOr`vB3f-7b3YoR?aK74Ntg&lnEStd#KnoID&}+3-C%%6DY7T}PScY=6Ay2TglbQXC zEWAf_(U4A~)^u$K77|YBEkh5{HP8!mx*UlZ2xmqeaocjU8gxP}-u^meLpzwNKrg@E zMAtRoh%eU(FniIAo1|at8H67@=uv`McrI@gcTVAI8vNf<7-8vjL3Ck!fn$)4jR&xR zpoSfw69W=shL#t=f|_!(|L7oBt`BoWBfqzf;8G#MKhvPP$iDe``YuqE(^V*d!VnKL z3=A2T(CgraVbW5;L{M3R+4rm46!@~;OoR)wDu9D2872b=fZs#obNcl{rAoonND4D@ z6^J?v|IPF!mqGHv(|m3>6Pa{MrSWBQ&5AqreHD;X^6#36=CzVdPl4Rm(Q%}mdFJd; z5>gEqZl|}MOm=a#r`HjM-Z~Kj!-7Ew6bbnfZ1NeMg1e#M-AP&xhX$3%3b{Jx8ZG+V z2Fc`x0A_vLJrko#`j)HJ6gH-hb+Bwc25lo9)4md#SVve~u)it9U`YyUbZqk<9NF^nDs zy4?}pXG;%nP3kIr|Fcflo0BzH-7>w`%XT@Y>T9!hhLD$dl~)N>rlN*R_KNDZk8g9D zqoKu54;V+*6^bOcX7kELd__!$gY}%J^K>#zvG$;~6ez#lU7ZM}h ztOs9l36K6mfGp1~co)9NYZ3L(d$o5hCC*32ZtazJ6WxA`V2%WS$^7Q2IFl>_>IKW} zzhv&KSalT{>CMK8nVdScIX>zwB`(#c^rS>_UtuiM!1B$Kkwl~w z%08Bl3NNR&w*ph)_ z?*e(JJNDps@{+G;(!cPL5m2cX3xT0`o6Aab-(Lu+{~I9Zw{Ke+RznTSL26>Bwn!== zope2#q)^mNKjp~P``~-Q0pgq?PHY!xR}$gAOI0m#x;b4JVI-V&Of`N+b7+qlNz zys*>>8a2_|F z!;jaJ)rcgq*=1p=b#kV~|1w8lnAkxcON4ZpaTP_CY}71(1&Fv{czCXSm=TDF22Y_k zl1+^rqK)*kB<^eiVE*#SWLoYWUQanJiR4C3z|x#TfE6XM&I9B~h;h2y~~cJvHE#|k-K{E5t(ZZ4&m^a*K)8sk}YK$Cbfa#pUQTY&;P1YiSU zUYgo=0)*U4|d8?SHSa$l*z(nk`X2N-1Z#GY<03AuZ+uhZ1*xR)DsJ~1d z8sA#BoDvRq@a8_4F zM`@oaFYb;`R^K>O8<)K2b?06(2l)zZpQy!Xul z#E4VgMcdE_2tA@&%vo@?Mz3J@x3}H>ah<}N@1Hoq8OG$~i>U%M9Vb!SiR`%YWyHn# zVbLwM4{hM1%BfWI8=VGJl!V$RZiqx9Gg$KKyVlnY_NB9BVm689?F))=rV7Oe9J$u3 zzbCy|vvO%*m~)s3EG9@4eH0kJuWOpUMIHdaa3l{Sr%agmr_s0go~RGHC#PWVzJEqS zxHOem3@B6vVF4*IsSb;$)rb-cwgUs>z8X0aaYdR$>vm(XiaLB%qc-6y8KNwUX1_=q z0S5zvA7gD2;O6GJjKxDyNSKAY_=fb-Z1uGAlc!G=V??MfWdj>bp2r|-D_1rzoJyja zcM-tyA3Kct+S}W^NEFU`*)-6cTUx7DW5j1{u!y%nc>nM-Gf??}no2%Y3)rA2qtKwB zAQG<_aNLg48+o_9r~$(PA6{}N1_<=ZwTRe4ivdZmQB<7bNrbY-q6+;aQ}GOb)Vp%g zV!xhl3gz7~Egadp(YfH98RLNzbAiJ(Rlk1_$<%2o^WE9rdqh(Jmoen!DzT3gVac{E?D3nGyVODwvSEWSxkOS4udz$H@m z3y%+(4k?%_%#wJTKu%M!a9_7~zbl;BXU*f?UZNr*RUfh8=X0}e+HGv)a@sdDGmF1H z22LF1%ZbxO)Ki;Ki9~t9j2GoH%%=)&XaUEu^TXZ}HK_(G1f`w$kE;&>rzCQRP%bk! z!BhfHgm+Pj*k#wpE4td}fPo*`Gg%uTAiMX~gK(^3yQEk9@$Qo9r?#h%w8B8AL>!eq zqNkLURHc*<5KNi7Bl^8LN13k(6ZfXa6zP%9w}v`8u!({2&Ev?Bca{j5xt#&Ny(GN6 zQi02N5($<{v`!2m^9BJY<4W$8r)%d+JH)+!tYi2A2_R_nnaRArFm7|~Iah-j7nKEZ zM<^-WHBUnos>v!VNLmS209hNALW~ z@b9RY93>b4%yhcj*426qTf4CXUR#?!W`#uLm>&Pkw1oSvR0M(VJYHbirNQ_IS0 zbS43q&K?BR-&*Q$)eS?DFR-?~3JbMoDuVU(A!#2z~v$o_?GomM&t~ zp$op%cf=p`S=|cBmPj2%qT?D>_<`@~c9_esDPck(5c|e3CgcVAIQ)bugK&e@E)jGsx12 zH;^N8qL*?C+c7cg6eRkLfQ@@98o;rsviw8#+j6l(4G_r9RxoirKS|JWgHBD!V}4Z> z#8`e%geBjVO!HH|D}MX;8&*4q8DUPcq~(kO|nDeT6xD~&9qa*7T4N+iO;!Kq<0 zTc0pugANL-{&K^@TC)*XcuBYyELhrj7Rteb*xb(eP>K$+;zsyOn@PX&f-Ll|-4{9T zu<+I?j1M5cn~dz6l;Ad(6lYe2f_>W8eO8zdz7-bB>tKG^=6&XOE0p2Ic5nu~+^^LE zL{q!22*s#(D+o_vodA@_<+!SFF7@ulpC-$iv^^5&wo|fI!lR>bKUAp{U!(L<0Jc+d zpDC0q%~c@6$>0Eg;Ijnr`^OI~O8LkfZfBp~6_|*ml>(;__V;+Ign`rP7Y1dy;QP9T`#cQ2xNAtkxJ7+58__2B|N4*C?+N*jfPSV zoBJPKGcQEE+=^UReSCLr&gN;tak;UO85~C%5w_KvQ@K4@X#G!9biDeFl0fG|Qc~~g z%X*XL;00k3>rq4Mn2bQjkfd#Mshp&A!5lzZ6NCYWVmx z9!ILMz7@_og3$u(o2AkVyNKI%kOeq#aX#)4pFIp6%`P@-$xe4LkcwTt{d+9{I7)eM zWbU@16!!B1_0n6t-CDRyj{QIiZo)N|TG@1-z3_Na0SpbC1ouQXAQ8sa@y-6+Mj>X#4UbpEr!`XdKyKIctf`r6iWiu z<$@uLrC!yAg_5IxV8CLWG3Eqy{l`fsiq`fo0r%ckf}2f&&)CeCOip`B`w`12W9~A5 z74Hi}1fXJp?6Hf8FiKLd_hl$;%7d8AXV^Tps71zv$sN|R0$m-$?TB+_lIj6e>z#{r z(`pr!(8y*jWjmLI=ed2RX*(p;nV>$G-#`Y@>guZP;b+^@;BS4o5bT`PLTGAwkBZ?yhi?a8{YaoBi0HNKbk4v29 zxh~!*G8X^AEd1)T%YX=Cr&gsYVr~Y~1Yk%7gX`?oT8ZVwZGLTSMeIdF0wM=J4R-=o zIF=|RWJiYR-ZdCiOBLniXf#?VrzYUWYIXKGPDGtZn_L{&ZRJ2j0m|@$*X>k7unl1Z zUa(~0*7FNWA6xy3BbUqd1m7L4J5=60^BGQyoG^lg&%^oH4YYSYV{LooQNe&1i0#fN zkkojxAK3Akig{>iMp2eKCsID-@?fC`0eQ;4T*%JF?P~vdMF9&|Jc(ahrOB2#iIzJ> zRNC!cT*#*t~OKEQ(L?& z{$MM=LUCGf)DS?Nu`F9#7FR&Hh5P)1@dnM&JcwW?4B5Xpd#4*kORSJtXG{q;&^u)R z0nm*2qA5TGd$!u=li8%v}AS z^zAc$ejsA~n3hH{ysm*=)Au4S4oa*zw9})bqZ0Ty6$2JeDH$n|sF*lP<@$*Ot%iQi zr0C?>(Dcs$;UpsipR2S!MmaH|P0eF+T?-5GG6Bp= zA|Ag3zCKqF@3R|o2qo#{jGDQeMd?P9K0X~;YVMHNiC2)iR8uFsm=U&+V8M5p%I}}7 z&D2v{T}@e&CR(DOrJHGLN=KG_lj4WPB|M!r ze?9lO8p-6oAm;?O+p?;ZoktF2zk^bMvGn300+Rm{QTXfQ0>Gu6xQfJd_!9ww6J3wU zk(2jz?frlKXMhYzuZ7JKCN%Qx+V~vRsPIz}xoGq~zwZK?p4xeJU+sT=*KaI(cKdh< z+W_tpHL6HS09ZA7tH@5Z>tv$Izs%`hKh9RYvap!HT-zR2)q9;v&Jz3qoxO|rLdM>> z?l7PKQ$v6LTlSz4ss5o9>?|K{*oH1$O}Jw{*=a^y8V*=C*DLvvyPf4p4_b=Pf-g4QcLE` z{+B{vxYf>x`Yin5n}gcfFc8AP5st*s_Kc4po&k8Lrs}Qa7b1N*e6_O{Mp)ATXtvwe zD4*d6l~GN)_xal!Pt;FQwb;M6OHLo0SiL$O(i(Lu&rb9|Z~cEPv)&9?k-W`JP&%jV zzlz3x{2lPOPyi`Gi^4kplaw?7TRu^O!0YlS${MiIWw|uC%v*n@2mSY%IMVnKv|9u*k@_^l=$z?vs_=f=p8o+=Co`*%ZfB%>N@|M{wAaMKtBWwV5*XyS4{J+Zj zznz&fVL&RZPV*Q>{xrY>7;t!Z;;{XPh`wIfOE7j_-Z1S?17=?Ux5Iw3ZMTKh`%ng z{}qV8bnE{+0^uV_VSKmR?7u%#+R_K3t*uSQ%?r&)eQPalX<2lc5>33O>Edz?L@H~{ zmn0s%*QOKm5>csC@+=Zrtdd7{)z$3{&Vb62#URjDk_lHMcZsKojMwF@Dgy(9*zZJD z*!aWigywmlhez8R-T0(Ta(#MV3v6fg%J@Q9w50 zr3j-Qz+ZV>{$YdX^f{}YF!fg)+s>GvC^ob8caYQ7;2hFE|3Q`dx$$02w*gT{G9|?C z5ZwP~4F1QxsV4}gn&ms+ck%OAPrb{rH8oSbdwiCx&`a^*N!0<=KXPF=DddtPMj;V! z7t8bo_ZP7vx`BW#zvyg+NX`9K#E%^FQq1fBiUH!iUKspnE=n zC3)Bj5YLngRoK}cdF7RqW`W50D>wuMI^pSnXTxKz)$imZyiYODk74 zO)Z&!)&;%Mgs0~tC+B(dIcVkM@%guHua#u1fNAEI{TOf1$M7}DT z^SR7=Bs__6o~{Rp&kWP-PGJ)0SqIbq{9) z3y9PYBkjI|GDq%p2Ngy)7QBqL6QwpSbS)3jsntZgbfDX?Ujl5W7C<3kPlbn&o?WPa z>~yn@<9jlYT%ei?5b=G--|(X$EIzj?pO@UpXjLE2!jJT?JYCiNJpEZ~d;WPyr?iF} zTsbASxV z_PRZ1$y&F{d@78tDY#UY`!ROm5Ojq%V&J`4Yj=oj{np>y%o_R#oOi|Jx@nYu$H>UY zW!cGO2f7ky3jXKTLao;!e2GRwfu*z$=P};!VC{EYr7chOd7cY;X8J81{#W66Zb;4E;LXww4<-3am7odc z8J!8x6eOOf?OVYXpm>r$L91ZJ_wV5Mb#--iKkYKT{qQS$1UUau${d-%^v|a)^Q?Sk zK5;f9#CwibxQtwt{-|@vI~w(afSs2;nl8q^ayeKY0|hw$!vef4jeZPs0*&UhZC)nM zPAQn?{mfM=fJfEI0;&GGcpe8oU6$R*ZmNoli$A3wrC~0X)IPw>VN@@0B2hm6^WwV6 z`XpG_nw{syl>0hkx_=-Vl65zBR=EN=pLp)1hseLXwES^%{J`5uEk2hBKhfA4sYfx~ zxUHKd@>Kw|L_ru8iAu{+~~X7Ad}IvNP=p-V*jd zNF&g+X_XCZO$++5He>gnyXFb`KmzTg+~jiT|FP-Rsu<9KE7F$2^dF_f^*z3N45Ttl z4AlF7SeZzmGt~xEf*4N{c(@GM}Krtrw6-Oow1m$@B<9hJv&7ltrC%)-RHVHrC z)XIY?PsjA4?k2-fMq}S1Jnu`!POnw%l>F%(#Kqa5=)YmK6`iO#>G)JB9POj~o-1ny zk+_uHfY(4?GC4P|plF6Iit?s0YlO}eF6h40%TY($I{pq?ul+0{_c7;x-vj`%N=9G`Z zN!~)OSzP{(k!<9z7^+2Oa(2J$_QYOcp3dS_q-kxzFsmKSYOrDW&_7Q?HMCp_O+w^)b_CDBCF|na`(GS-l9Fb}iMMGR z3`MRM;zhcsYqYHoU-J*-qlAsdjTpCX675a`8Rr8A^Dg##L&M`Cq*XpCQQ}*wl@n%9 zWs4*QnT%pC9GZCQD@n}XVo@xgyQ&o_ikTsoYY7N+Mk8idaRh8&NQ->8Ch#?gqIYwI z7$pi(`;H>d`ML$4gCt*B1?wF{HB^7H9SDz5S-Fk&J|(nb%(O!C=&E7833;EjnnEVm zTxM)TF?(WOENg)lOVdbX!_CCSS;}0h)=Ca+X5jA743~0`v#{ zzrJMbls^@||B&U=1BJK=4Mj;Z@t-mHcYmK~c?ch#1kVi_q_VK~43(Nn%wUi=3YDc_ zN7fx7pPgvT<91Hy^b}(<^y~k`tN7r4&^UqX9^;b{I!I(e+btBOCo(VSo%?u9@>+fZ zMH()aagw8rD!8DvsufRA{!>;~C}AkBVL^tQa7+KB;V{&5f*$+KenZAg?G>D|PU)rA zK29%8GAZLwMnUxFy<8hgWKX|UL}AapmYNm$S{LxoP~hUoa0SHqgjbid5JPl%I%7ey zLdCn%?|Y;_@rXM}x=Zd0c3^)xLN*xP$(-6m|0xr1R_+|jHNK1NxJ7;r8a8?oIt$iD zRZ8CxO}aj(Imb$)F5dz_k6G`k=%z{}oS5H_(tu+x#$tN>KYYDqP+Q;k##86dy30mfWGr;uM!9494Z1t@%ysoAE!9EdwLY)M0VLEH z-2eCuzT*nD99}dpz$nQJPAU9b7LAK&9mD+lkvGLFM5&-KuN@&FHmIOoJhw*^%-viB)s3`RX4;XRehI}_}1&N2s<;k;|ahq%kH-M|fm zve|7H(kJu_5A#7R6>NTgmuyR4!AB|amFb?LblO~3GPj;@9UgqKRRR#2v+yy#70lXi z)31MVmZu}Uv1ese_Anpgs=vkdIX^-!87$wAr-e#S6xcPf`=Nm^K|w+J`T11{+OePB z&bqkfxZ3Y#WN;(*W7;lwUZ&(|;4Ps+M}wq90v>nuGR0XFPwkWb#p;cg%G$=mkM@RX zUoNrZJI4ArXL9GaYgANID(lF;D~jn_SZK)kFoWiX2WA9Bh-c5wf0~B(dczyKHmRtb z>7H9vNy$ZxjvRlg%WC*3r<4GPsv5%W7w4Za#hdQ|D+MktZIKJl-_cyN2LyTTo|W znp6(Tj>u0=u^g7Nqs>+cgfl?mK1iQDEoE4RrpXaBG$sth&wBq=?!%(_!d#PUZeg<+ z5i2x?9gLDdy+{b)qI0OF<^YWEZ&n*NSZ@R?EJ}w{PG>g>~dRB?o&0#YIgw3{O)W=KyQ) z$^v}s@;_2W?a!`X*yG3PsLOZu)Nk%hl&)VgA|5zMY3(_=Oq_o)p9%f^^~+o-%H@Kw zQm4!)M~_ipPGe)w+@-Lj(D`&8Af=R!6DoQzKa=17)V=|?+EnDJW#W4z(^gQ_ewqtH zlVFeaK-v-n0&5Gu_gk~Rph^~?E{`&ut(nz66&4h+?hRI-RbCQ@Ck>2{-(M1I?H$*$rrq3FI?}N#qGKQ^=qwFN;ZKhOuG~{ zHD7LZU-=n18QE%!Pn4nFr6bk#uWZ*9uDi?q9erM;EQ9-P1TQlN zEk_HhZBevermv)HLo*J)zJ@e$r1Z%yIsbDdi_1AMJHBtI9_$fVXS8i;HTAvu24~Sqjfcui$ zk_%FA*mDFMI|k=@ja;Ds1`2FBW8zUxNe6t188(z&7hn%c_0rL@(gYqC>2_<&ZMceP z&CJ`M{~_xDm{tyhGjBJS-_hgb@+Fkl}2*47VOp{F(Iuc2atu<6Nt>1_dom z%;u(7zSin_!JZ15ssZzTrUXRgFm9P%YdfqLs3!TxxvC6d{DDfdoveg|>(>P~1rh~u zp|!{R3o|0CH_QCxI^G%97)or>^R)u2iNMeaq8x!jA{_@Fqo$ET#U-1%bI;crZyTE1 zP&ZT2kx%ceE}N&eTtsdilJ-Bc8X#-GBVmN{Vlnv@V#Str!}&C$<$M}QpLE*w-#Sa) zd#ox!G-v444CnE#;~fv`wT|Rwtr)Qp$d2xx5Rov(Ye|qSbT@{qKm$ohAFZs5`QJj8 z?bBGhyXO^?Wa1~ve-1}b(%-ayuReSmkH;Z!hb|-$$B^~GWua7UHMOwnt(N7rwg3EL z`?`#}kIy|&kAwwu^(2dp+D~U*t0oL5W+i21WOxpcXTAr^W!!FfUIrJtKc`&rW5MmO z0dO>%+7W|`63sX8$4`uhMLcc8G}ny#HBSXS{A&*Gghku%(98JuzHe5?aN+<7=X9w{IPeI?>a6+Xr1PdWV9CnlB1q&7_vGfs!D5z%hb^BM^d$8zCTmUI`{3v6K1(KCj+*F{!}MJIsEXXbrP2X5 z7GxXP!u6t#cKR7%nHa>ie zqGoPN#=SgR5O@)!H)Tk^C*TtqVX@Y1O15^kaeErAh&tOFQ}b40!;{wsnMLUbBAzNU z%xPj7PHwE>aMn1V;GLIIbfG%|<-Xdbzr8x|JMJnt|M z?sLat(>Q9@wtqZM4^~v1&C}#Gzr+dsVnyw2eSIddMq5}p0R_x zYq%^(!5wdTFqx**Q-YV5pTAlxxcx3x`)xRaANcX?$+yl(Tw)`FziKWsufqMbvMHv` zOv_BIT(v6uhqHyU^Sr>r+U*QIbMU(p*UP%63o}qeHS7gE9${MrwdBARAvF&|4;}TW zvAYHvUn5=D$~TQI7&i(;jr=l1f6neP)nWr5VI4f1^2rU0wf3>&|9DcPe^0;Jzi)C)pwW(Gbr4Oz#LPlD+TY?@L$crfJRwua0Dwk{J+T9vKdNF=o zWlVQG=`PGrnRqZ>l7ILWd9+`ySIE#&Z5{OoOiAaKL|?;wrYMRT@(@cSQH}Bu~}=-+ecr*NhmKU$IH;9b0_M%@^?Phil3}n<-^3vV%tcAGainZZcpiQw{orEN(5|*v-EdIJI?c7&;AGrWK8~ zEiSqSa+wK5S8Xtu1}f658v~mzi&=C8X@Ce2-AO`(!RH^vseSN~S|y1%O8!a5z>2C& zZf;h_7Udel&rBSEvAS)XSDWcj0Q-j>9^X3W=(Yc@rV;bUb6v7QYhP2tx) z#Y5W2nCX(bQSYDmuVq)8^xsC%PSd8{PTw$is1Op&LIqBK`lYiD@SGGcyGfV+^KgtX z+e*;v>~J~-GqWncBymr=4(n7ofEUn+?rONyJ6BFcM`XG9`o-GJ$QMue!Ho1>$G)}9m^V#0+%BUwbwX~Dxd0J!$Ito~C)BIGS zRFB(cR>s~bMS-4*ZSm%b5o(r3jIH-&f_U-9@U0Lxst~z!>byI@$Un!7fX~+-W_zCL70`n(Kvy*~lbEV+pvPQ!KWd+}mAKBt>3zE&kr%l)_x zLM8+KK^)!m+UR(OE9P<`T8PybktT<6Jdxbu;>XWaR*$aI%OT%oJtOqBA;-McDT;2- zM3Ls4S7t%DZr^jFd9JT?KxH>YZ`PA2PyLjE)M{}^Dw^t-_fR|#@*_=O87oQg>>Olp zo``e#mcTo*`>B-`MusxDGQD0>nC}o9rd?kXd?>^!Wu+qz497eE(HP|3&6DUSa(UOA&<*P3{JWkuYHxaUck~ z%L~{vUGP&Oi&D$-6yk4l%}@`c2eLlZj;2*WpP>N>xts__?ZWBx$%g1RdWMrlt9ovN zGneIaHV**SO?Ovo*+{PRZ{L{7$}8{9jvaQ{JJy?cl;Zjx+?u*bDz~7Z>{;6w5&>pv z!)j5f{X=Y^upm3Wn!(Hv0lOs{>wO4*xz!P`5q^Y##KD4+GTnC$e4THR8xytAffWhKmGSSXhI$0RUs(_U^o_k0} zM*7s8)5)?7JLR`eEaM`-ROC%2e$tc_l_~^u4&FX(IQ{c)agu+1jFOdlLRT0VI^|La zk#)9;wGVAC9cX4~5^xYl_ZLy+minh-oOc2sj`~Qaoch$8zg4!^H(xG1r#NWV-8xu- zGSRF9--lhq>6))i%omadst_1!VG`tmxCf@D%7k<=0FHh8yQsYO_XCYw{rkV$)rIhT ziVLN)##Q?L5sXA8F?IzBzfhvdC*v-M32``~$vuo5>i%R+wI9XUZ)a*4z{ewt9o8c{ zz*5dQ8&HaW6zw7EsMLhKP(^^6$z-t{$M0EHXjk-QFYSewPfoTH(|L{l2@R}K)6rfL ztb0}mQ93ygo}ARU;L0;*JrQR7o!Q&qEel^(n4ofXNO20`J)v5dTcWl6Lc{yQgG_Qo z|6d>I(rrN2Pnm#2fJ1-NM~qRzH*gnOB8nTt)2!bclK&|pAkMqyEYkh&}{`m*92Q4)}`V9=0pT^?+qn_+oDAfMAIxxJ$0mDR@J+K0Il(K`K zFk9RHEOJH6w;6^P80=#6o__u=F#;fV-`*?g_k1Hi{Z}8DpLIzD6B>wvt>v&k*2CoP zDfog~x@)l$GBfK5cB`9Bi_Z44Z>o4zUAXA?4;xNLyXSNp#!X1Ma8xfRD$`sW=&PSV z38yPFyD`Hj3KO1cHv$BLSWcMb>Cl`9&elxSo0|2IdJ0A&&2b1KBV!NE_9>KW=MF)` zzY2wk=h^x13xr(KFcF zd$*e2GYu&ayf)wUPxay(n*9#=s&$i)(`U%30sY$8d~}2s(m)*8jeuibUO8>my`_oR zo;8b3{C@-1@t-Wv7eZ{x{GKW|4IL!PWbQQN^NXQ7k7ICs1O%aR5kmFpD9a@)wR7~bvLh0ZXB8|2tDuTh{3gG zCqJ6FshS*rX{5`?J2e4>Q*Z+RC>;reT6R^nq5~r5sDvG#%TMV~Isnq@bm7C&g3`{% z>D0{H@?F05EN!km%szKppwPc!U7L$*iK%~CYJDMbaLux;+`E z7=|@Lhu;5%uY5WjildZql0lFy(uF!MOQ3d_?tQg>v+LE!N|jTZ zQ8`}ogIwW2bWY-D-1S?jSIlR!DuzH{GZO_TOhF!zK(wq>-w1)$emgiQJRiPKp@XaD?OAQ{@8l7gaupsnM1c58gnJpHDzhW10geHr;O zIpyuG9xS|B`ocg&#Dh-%(9|;ffeK$jTI#mxAa;+7O8spI`3Hdh5sA=AW*SUO-i&@| z`Bw(as&oz7AF}i+EVf=M8q|o{s(h@@l{P2{nY3=d_#`iKs}TEF&&$@HP6{62p+{H_ ztnsY$Z#!v=10`s+bsX)hSL}l9{2xU`B#K!DNGucd4fr@X|IYApX%a2mU3s?MpX{U3 z+DKk-WI`yI+Qq_UBDo8P#>c{gKjoT#g9WbdzSqOXh%Z02sG(#fG^6T|9wS%e>9Mf^ z?|ZqkEd+m$^Q#miuZNIGBcANeOi8Bq@)!* z%M2d)dC2K@kD%;kt+O6DjNm=Kdta}_MY5tZUta+)Q@W?Sv}$Ug(jWW_M1Nt}E6{9< zwEB!Po6el>6RRzI3!hxE{ai0jF{`1i{j#>1w&}Q7*_7zUA~uH%nKQhMT;yhHnr~`s z??k%%2uv5~8q+OTwXx}R7^I>AYHouPNh|CZPnqA{AOF4sn09SYKWYQc+W^)KnbLksC`45{8r1wZHUdpBwkYOrpI9Q{l~hcvi?)=OH_-SR(bz9$^kL>VE&C?PjVDQ6$;_iei#hSTX)@Ez}OxcSZBSB8|lgJ@QZ%8oDUXC{B<7uw0ATSm_w?cI#LK(Fw^6)j$+e2!aPy z?%tr?&ms~(M}JrUs@=1EHkiI0)13S=_7)20o>KA|&83ir5h6@b4O?!pygb0u{ipLz zCoyF5FqR`|9Dp8j7_!l|dz4b77+u1T+N8K!bE=HTIzrJ^QbHZ|pB3g$!^+Cp47{A7o!J{ioT7VD zzoYeeV_Rmop3TXH)Vy>*ll=mmC+iYtF75#9O zYOaS$_4nt|Ny1lvL?vL(iP>3>DAQ>ze-F2C{&d0C>v3@&^2VGP!xV_5E##{MpFA9p zfF7b_ke0}}+Jqz2)^2W&k{DCb!=$R^@U*Rg7d{e7hxqEbp!d^E(PDC~Iguar6{CIZ zlVcYx2a-aaagObLoxEXB+jY|#VM?6;_GD}4Kvx4wy;;B>mdpE#A~L8sU z=9v~t!ZQDH2vO=c&(rT+Pgm#bx?X)1_XG%8hm#p3W_LUhmO}`x>I7cO1SUCB(&vIc zUK$FR)i3C0ki$tIYOXn|XZ}=Aa*n#5b_{i_BEnW<*iizVzdS5N+vPZ5SIAAe+fd3D zC&-j{j1@6JG;`s=M0>%?@?~5Uqd`w@xS1!!rf#(3r2r$x?Qzj0mPpPkRVH2jETEEe zE8Em`Qt)1N75t+^C?jl4uZSLc_qGr=ckG18f54JHchUJ?qTroap#{}{L>xo$Nzr)^ zq;EC7&cxttr97~-Y{k0=oDL1;{Yb<)cC7OB)sN`Y)DL|mczCYMqFsFq`4;NOB@hqu z(+u0{0nrvXMERXL?~5m{$#}l)1X#MzPT{He)MloH z!^B@-FBE0kcboG6?OGY+u@@lUenOQP6+^oYAu{s$dBS84Hr9v?YL=rv&$Q5X$^Bv7 zb<^!F8p@=u)af7$phMqq+l}@oyN|7p;^6TyIIQdLo&oP3f@3!3k?**^>91xH1`o7X zj^YZ8gBb@oAb69h$DAgsL2~WuVF3?;FlAPDn5XO!D8tI0^muQsH3$D#%>iG?;=~s@ z(pp{m2P+(`4JPXYgz*}XTS#HDE&owWj|VFrV!VB!h86BbR3TYpdT&}|5E~0SP95At zga?qbx(^wvuD`5Z(fe?EOi=7zsCOG6vFWaH=X4SPl}DeE-7knvQjZ*%o<^~(c7M&vaR6 zX`2Jtl?1jw1Pd?n4+Jr7q%V+I3L(0SyISnDqm1cs)S_T5|C^TOO0-^nUNnVJ*EdpL zCA-B?%IMO$6^5g+f;`JTCD~yQRmm@;Z20X=R zWOkls(B%mpDSe;8291`Xf`~3XeHwwo;b~qJ4f-wd^Ep*|FR3k(9Dqj7O`c>Rl$1{Nk zfcsX8BBb+!w;1Zk`_{bb3_nxB$gC@BeCe1hp2@h^UNVR~hOyOoCTvq{;Lm-@gzBYN z@=i=|PzlW!z88w4Un?4S`tM6?=z~!?ke!LHx3>@k2=VjPW$N_U$mpW)jY9gO^;UvQ zeHEB1<2@eTa8n=~rAfzQzPCyH=7q>}gFOU9)-h1CTh8dQQp3Y@h^7ogo7J~o2|LUF zVV|$9#4<^(s@p#WzfEUpX<^8BuL23k|8(^u4`9FBp>f2^54Ifnwb^yZ^ajv(_yv7c z7T4@fPpDe0f=_&P60+nf3MmZ(U~M!-jn@}rU%4NPSuAB;jl~ElCtBlg9)#4s@dm5; zQtnyVg!2u;QPY2UKZ1@GcZ=>`96h1C+YJ6fwdzl_7=9e9EtE0^lx2%bp zY7Nl&=|$G&4Mb(gKNd;w!GJ>wLk)F3T0BTrgIy&oWNLW+0?o6PwtJTHGovt8lWRpT z!EuDp{>iC9!*Ret-KT4v4mTpS($Xv%UB&keAq%io__PM4x3h~{QM--ogKn>Eg34)V z?0ek#3VMd)0UU2Iern>I`K zcB&hV!gy%5%EYaBaJeu-b}Jmfc}QocVv zpIW6~2qGH8C^G%L;xH@gzjS_Poh<0l0lNiFm&rK$xzI*EMcX-Y^KXo~tR@Wyljn5}+^MU2yecYfdW=SQ zETrgRGY1||A7F{f@Ocl1lDEVW+_0$pb`DHBT~xp8uf29rv^{gP!dr7v4j7D*{-9xm zL_OPgrAbFmZ7y+nu;{{hMk^_ySHv(E`SVP_GZaXO}(Hl z`dlq*w4ceCrC+bln6B}x8miR%R5_($)P)$^AD?6$%r*P(hBkhG)m;y!mp;_x;pdZw zWFY=`=JS5}=f?X-V8xJ`dR$%+?U=wLlui`1(c$Upg}5(sHa9mHBoXe-a1%IraA1P^ z*0~oibA|K$aKf4onkewxBc$6czo_LnZlAJMexh7nc3%U{l*I=*#8(Oo%*@P85yE7tkmBLGlW!L74X&W@VdlO^ zcr1xNqIl1J9VhR8B*^#W2K&7sg-luR3Q+6YE;mbWIx$V<`;B+WRf8(a@|JOK@(vEf zl&o``9Z2FZ?L2OvTGCyF=T;k)YbfT&4byDVtxJ!);Pi8rV~hUm`ED(HAD{l{w)w0w zAAx=Z9TAX-1cm_&G;ep)8G*6nlbu;tNNmW3Sf0BZdcoYjg)(eqa~TUBB|vzDTNhLMK(&e z)l4BH>R7zLxY*>th0XIcKjr;6Ouqdp-LgEoMh9JnU6l8`uV(|3!2n_BlN$oAF`~OH zU)2k|XT3b+C5~N(#G(Hmp`uI%u}_J3qIJ$lSl6$PPRA~YZX0~02PPT#z2ItECQ9eGdeSl^K!~UBz5{JqFI$)((C1ZXjRu1ici6GTk zYTQ}x2Pqr|B4J46X7Q0-WQ%n2m!2B47!)jrnRC=kP34P6K0jQ>2Bvz|nOIT;y|fr=%vZ&wKztG-29m%l#vq+3VtTfX1J*Ckhr z^^QF?r73zgAlp?)?NpevG3Y3$)T&#r3pw=JrKEZ+asK(dqv(*PXQy<5$(da&NIh*D z0aRbFu_xn4h3%cx_sHcgvw>;}2+BgVojAGWV=mKWoCGnE;b3A1wN-0W>qwxSh!pBx z_fJ^S$2K>Dm--k09vlbG*R2xs$|{YgE1vFP(sHL8>hS4+d*P4m{49f{Hr_miq~wDdu(icQXQU-%lNIi6SOF^?*HWvl zdpso&naGFQ1(V{kK*?B8P@@j^2f04^WWR`+k0BtT@23LwIifC}2#}fr4mp=`b+YmW z65-FWD|C#N(mZE~=ooK%?M!NHlbxO^NB38jho`3}_Z9PM4k8W-YQQ%Cf)#TS1D9V} zLu)UZV!(Hw(+Wjtm`q3bq5sq{ne>THqF%^i>k<4+>e~sNZ)Y2x_X>NzNuj)&vDG*7!or|gc_9*Fz?)#6|*V<=-Vq7XZ=2|0Lx;;B=i=xA%9*2=f< zE4(-`l9ub~cs`@V`+rPkly7Zs`tQv!oq}1HuF;ymT%CkHzIBd>16iCyxfgLVykQKF z)!ekFUPH?sdOP3t{F^B?9slTB&dTvjs3#*`3D595d)IjuGcO8lFc?iO~HO?*?DpT;|qglzQ9z?&X4R9X3U9PfT zeYhZ4DbcOi#LL{zCEP7n;%t)F#pX>M0FQmVj42#q(DcntOiaEz=7nefj=0!mLpz@m zB2Pc|i3yo>LrNaCvaFpr9H-4Sh2dHS6MdK_?eL_pyvMn)vsf+R<01{=U7w?}>E&^9%C?J2(Yp zW&t_}nx|ixL@iD8QI*-x_TQ8^j@c&FGFucL%{YRwdPQ6oTHHmy8d?aCUS+#DOS0W`1-T1xA;(DPaTbz~! zCB$;r_iX70e)7a2r4^_?>WaL=iG*z37HZnH+6b}Pe-NS6XDfd&C0Gwfb$tnqb#p=4wJ0}ky`1U=xeVqUV4wD6X3Eo-=Api7_mE+ zKL0@EqUszE>w*WWPxm|=Ts|5huMfx#c$j4fYAWg~e<@k$FKxIGZL<*cB-wLp1@Eq` z^cm3Y1;9$tM(RAJ46npr*cW-MTT^)w>Go?_8J7s(@R>8bP3{ z{v;+vYZP%3-%S9Mc)?AigbOKGl=Mdxh6^DlwY_L7 zp-#i3@j2|xmJ(fBZ{ZdO0xx(6=tF};ce#6gCmRBzas32ILv1VI)~EJ<6e#`3xNdYTO;)AT z=d|lDn_|1o>(lp9jJHGOxd$6K#6|~tXDQ708EUkj12#`i&feyL_|>TT@_D(J_%|s} z8(=%v@;**W{cWE6vX(Yf8fj_DfIzpy7$XQH3D2#n4d)=lM}fpH)O?3|v^}}M`zH{2 z%o0JU9%detuBL}u^048Hfd@Wah3*jojk|g@sN{~6QN;FfkIN-`+mMw;eB=Uoy+>Uz6rMj;bAP2 zsnrasMJ-nWiXgkbA5Fv+NQNMaztDW?9;tSoFS2^EkYydNXkZq2l0h9=U?e3sGY`Ak zXuZzfT#O7B6g&lQ^$QjA+tvx_l~H|4ZgkUYz4e}zER7J4?@Uv?Jk4=8AI zH6Lt+*SQtN-20SbuZ(;4B864=cbU*87k{3?`1bu|?5)5ewUhAaHvRLPl?6AQx}B#N zYgt+ZGen2gws+H|+44l!_gS?C;Yn~>sm_t+b|sa{y)THqQiQAIknaz&z4cXPlWi+R zaAE0jY@GW5jVvS_a0+WIs6X_#bC+h0$RO7xYp-hZ`FHM^<@#aN`oogy*op3ABm#jq zUqKLe!C3j&8cpSCSdXYgMmFABV`pzj3L5cPqmBD*o^Ij&C0WAX-~NrN%It$$ORi@; zW36E*E-m5{tq+&WznhcJO-8fTGu5VsKa(kTU4Na)=BF%q3ETL0`T6Cs--pM@&C=q% z{ft4jc^gv!i=y0eq4BuZ!~-{oUzT>v(_maLUg0Wa-G1;n=XFU{;puGM)#Z3|lZHMi z)fx+nsi;skon7Ac@TpylcWq*FdeyyQ;r-w;=Td2^pV>0Bk!+6UHrkYenIMvFa+o{Q zQ#~J9PT23i8N0j!#}V{s;3o%kRl=f)3L$>FK_%sQY}B;X?6A0J_I(?jGGnV{2cNHU zzaZ6Z3qPG$0uedKWtdx>pt_d$?bM51WmRpQ_RqIgAt7H8ir-oBW15Gg&!F|K_tra- z-9*nK9`t&$jo03XO<%HfYIM15F8AG#XwgvA;7b}6GK)nH$W^FN5aNr_gLt+;JA}aV z2<3T*2ZkMssZH~MYz$+u!Z_su^9^)lpfC?Iue1tpeqB6z`0O$U#Q>Qx_* zL+ICeWxhxvmvfQ!U_7~O$2S2c-dQ@zmzK&UxC|^r8suJmc;8eDv{77EP)`tQ#^0Fm zZdgMRidCG0PZ=UYC@|g(r0FEmHN%LZefDM>(mNpqY$IROCDK4~FcgkYFKo!i>QAD! zPY9y~Q5w1(mR&l#*rMA@rK)2s^Hk0bwMI%o2^%NNWc_U{2HLsV|ut$2dh+kz!gfMOINVj9QH4)$^>ri}?zH_*rNH{*0!Hvz71>WN%G zDi-Y%Vu=29st;&VDHAtjNDdP|`;y|)D|H`Zu$32Q)n)$Y;KNfbB6^1F;G2(P)dkVS z`BO}cn2FE(m5|`b$CunOGOaG1MULnJf1yVZBjn*{IcV9H*QF8s>;rcF27N_JC)WWVWagdCm^!uWv=wrrmNS2 zMZ?$|t%ya;G2-iM{*M#z@=l_xi3uZ>b89idXe~2sv)jc!SKufjnSQppke8zJyjz84 z>)xt3%4uO9VfK#66Y)TMEgSnjH4h~QhgLQ-MobIjZfyeZeV>j?@3SM%Xs5bzivfaT z8(%Du>!D$68`c+1Ck$hqCsM1!7XhANoE~=nd22epAg_mNx!&8uHQz`Ye|-Wc;Cniu^1-7+ZT_%u zB9-U7ko@6d@uhhefBQXjF!#^*=e>TV!fF!1BYR!$fW!ms&0 zeN1z+y(tjYH7k_!kd{yGa{`1gvE?X|bqD$Tv%VSx(HlU7#b>(xB}@`HDhM22L+HRn zshH*T#1PTP&Ru}&9}}QCH)Q3|%F4Qoy&oS7D*=9ZY7F&?;`Z3l(&tdb)5H~Nm5qX) z`Ec75RYMaFtGW1j?BZ|Lj_}VgR$cJHwa@8cLie;gdxt#d-Z`czDQB5%tyQuNjUO%E~n!@e?!GI2as_B&*71d)0 z_9)@cyX|UF88MH+UpA|M!=}HhO(tuI3__a~nYawz+YEhC4!Ql5i6 z&=9e?34R#AZ*tc-$uJnVmmZdu{bC_r-Wl+hsg*>RvL{^d^Fkq#kLlbBOH#flOD{nV z{kh^ge>NL0G&I-446G>_D{;PL-NlH_8*7?dRA;cKD=q!oG$Mv6SO+Sq zk?%h_i!B(Vagi2HQ$62U5M@kWY7P>suh&Rg3K*u!Gj||KjY9bh{rj4~&kvLza4gPS z`1qQbARcg4eik1yj(R3dop?M`bieuhmm{Ylru|H$*J(Lo5ygiQtulJ3tZxCiP{2(d zuj>jnn{$dy^YQRL96Q2?d-aU{D_CV1#0L~L9Mt-joH@Zx`d)cZ&Al+!ho4w~=Oo*x z5!;D&qbFRUHWio{;6@U+Zb^dW}B~f>>AgRR3gC8-Qj0g6#aOOD`KxkcPfuu zpj~WqWM>vV{M4wcLHG%+^ja!tjNdrX%alJfbn>5u zyY{~zE6E7Gi_HfwQYoxVwQo_O5po0ZEu{chsTh$*TgWkzTQLEm*$Q((g+?+4@qKgD z0+QeM`U=^kz-596Kfgr+WN|rbf_e3KB*wx~g2;`3+|?jYuff8$RxuxmPE&Luym7DR zpA4k;VJ^oAZbq9~*G{j%Z!8*MBWn#boe`3G#V9N&uM7(%W8_N`5P-M|OcpPn<~`eNUKRmNh3y=_fm!ioe^TmDuGumR(c zAa$>LJ98>Fw`pE7XZ~!=ZmG0tUkrA4rYDd#7A4xVnyT?-U~f8b*wrK(vmm%UBzrXf zmF#C<#ovgfsnz4+YQD**uPfAT$a`=*{Y4J?>^JV)e2G)-F5;7?4Yj@Z$opAnJ*H)w zgu&>G`w~k^=%aMd(^|542U=SN8A{GgB~+_~6$4?S>t`42k%Jn!Xc6K*PyxED z54UP}<6@qPX>b4X1U4|{n`7Ka_I+-wmcQ@l_$i@B zw1|%!cAEP08=+*aN&V=Deuy>vZ_)gz^Y=!+{LQfLhqMp#&HwgouvdSmwNFO~e6@z3 zFZvf(5>_DgHN6rb$#m=g3$Gztxm%Lw7gd#^T^&eUl5m^}mdAMl8q{(a*b{Vr8&Vm*XkTxH zU(@3PhtKe2%VzP479^CcYsxe!ZZK^R!}*ikEv(&|YCu<$!7dv|3A=4j|t+>=6kTr1O?F? zsLm#p2A&_YCQ5E+XMh*o8(Bi1Ro}Tbx&)dJiz7560(k? z+`@y8rS*UxrjJ;PRhx@1rYrk?$~jT+Q4n}^`zg*+j|%Yr4>w6_)w2}cD;-m`#7g_+ z%2a=wACg@N8r(x)n)$QBXVOWeG@bkLpEoQjA#O{yvXMrT;$VOr<#WxyGZV5%xvrT1 zri)WhK|rg{)mt$1!~H*ngB0KQ3xY$NJ<3M*&i+eK@aM)WB`A_eiL_5xfivFWb5?@}OKc(s4KTiFrESti1S4ct- zQ!gv$4D90}3>%BWMZ&|zCsQ#PnJJw+>P$)bYxrT%VtXBD+ zAnq%v8$_|M2p{b=$gIHNY?s`=lf0UNEm~(~C*gws$Tum3cc@4m)7}Rx5-XYu7ukK) zG+8mSRhqF;IiD?#aUQj1L~N6}$M{M#h|M!c< zI&Ej;R^1CG9p2DfdHR~sAO9c1-Z9FyW?R=x+qSXND{b3eY1_6|+O}=mwr$(CZ+`de zQ?+l^sa-#3YokSs-sb3m9&bEv#QBKXaJ})5liP`kP@zlg|AlIyhbrrzxDMbyNuEsE zUycYzqnjmSTdVGD=l^4fV)a9484Zi*0L9(JbIKdOFmeqiZe-=(HVZbKybd_;;KKAy z!$UhqQCJJ0PL4sSWzrRp4UWOd9}QWcA#vhdp7KdO}RAku$k+7XSJS+BmmLrisd=b}Bx$^h@PV8~B_|1f-n^stNn zj93l~8Ye~n5s<~k{DWnYoec8+&upm-k{`wkdn%2i;(vg*{DsI!`x#7&jRm{^6XS`< zR_32^GGhX*jp9FylYAgD(mx?6*5~T{=LpFELB_}!LvNxW68s0YOflI{2&JntbpJVm z_#fsC7kPc4b=ZG|#wa5Hgs@$zfQXM(M`myl%LcT75-s zDlrQy0#nUDfYjz!0w)5?#Di)5I{~XDanHXXaVHE5A#N4?Q?;iTlzn^E}8!Vv9P^^WU@K3 zV-O*bNHdkh6@;yb?)h&17tCcNqKK6rhc)(_--MR{3>J4o8xAoXq`J)w z^FT2cQ)i>gsZo~A0(l&`k+Pgu9=_4mbn}FTYu!j1@`NA;&gTE7xx6&8|BJ~|NX5rz z<3J>L=U*yJ8}jR$a*P)W+d)!0JLj3H#oGN2{+X7`cMFs<<;xTPfp!Le>6~C0c0zti ze$D?T);8e@qT!yy>S33u;U*46p>eh^XYv*linllp{>aoIvW&N(BT(Q8n&?_R+H=rB zde%@46%eqNc-LG0@y5msSorKl5$rXp<*_a3xc!$akSwOPm-;7Q&;XV)d&+^_QLonl zC&W#;OK&iiz9?=ovpnB7$4!J_gu7(;6!W*qMrdgfaqK_ZOgyTe)U6vKq^%~JIV|^ImosGs9mw@#av2Dhnxmo8}S(Mf@Ck>gyLe)2mf~t|z zGY{ir_cw0kb_t3)voa-EZy8pXj!8(4)BXApzBN(=p<#wmZYzMt!#`zJk1p0UPsv*_ zLr9pbs~`JuO%gR?;K+;u=fYx6JIsT0=NB4#I4sgeSMrcj@O0-iqokXeWN!G8>W}?H zUb$4iy$w#y`>i*3pq@WUW|z#&zqNi+v8r}xUQ?V`a(JlDfP@oPytwmsr8lzm=-!=j zylK!vS)}r^p!UEK-w^e7VE2&{^xs)5B@qyayv*dREHs^oKBMqtIB7*kcQ&fEw1Se9 zZMue95sToyvByDgQ7+^>qJocI%2u%NlVCJ7or>1WU3e6$cqdvQW!S?>X#7;(K8m+l zL%^g}LqioZr-7f~V&W8$VY5@MnZZS7HsfdY)hS-`D*oHvnW>BmE>lad>5Jux?FrAD>2K;B#h8xF&D z5}GV#k-7%F<(6sXuV}_!tL*HcpWtjL2K9MA-qjbkC-X#$|4wPH#pJ$w9oB*!va%Po$+DsSgXTvm3k-TLy*b0?h`a!2SPFcZY8mG8#?^` zLZUT|_q|yWI}ADsgPLQ#>(&_*(Po>m+2(LHV`VHo_)PN7X45FsS=j3;_i)I>HQL_i z#er5s{WSSNmqGvWqU6_P1K3aocxp>(D8(nZbk~KsvM(ut(16b4#k^ft5P%IFS0ts2 zC7s)zT}k*+c=4OjD6RLg@tNfu9sTpzGvadKlK>w&<5u0rQ}J!>Hbt#Q(mUgSiAWETJa(9iK$%#lEhaNxKr{RNibza-Qwe;_k>7DYa(I-1N znx9pFjLr&~vt;B8G)HtHkN=oWsFiajBXcI+ZKppX)jFu4RU( z1?!8ax7sbCHjgJew^@Vv@&9Al6Z1cMZ#0A#3Fb@ee7lXGt5q!K zOjsPh(zUHGxr7cfsg`S}(=f4|8dygD>dHV@yl3cjhQ0AW9|UavJDTO!zoA)PIwfa5 znY>u0%T%_F*MSqI3>qz`Sq{&0?}tp_PlJ{v;Z`f{?@>g$&}~$^uS>gsYgpX8cKJg} zwjXFYTze&5XhNi2^&2qsMQ@)vE`2!&(gbA&@QB<9aRapvBfB(VWqzz_xSA$ivdO*Q zi5Nq(*iP96B{?dRp3#Z+Od(JL{k%n$21RBv|y2>b~3L$4N6sCM`D&HzVV{=`)*$?=r`qTdlj^6lBBrGaCDh>Hzh}V@A|!DiSVTg z4)2kMGgKmth|N1Fci0Ncn^3} zzJ{X+H=HQNG+_fvE_|_6<8Up_G~{)VA_?ZJzMS>fp5zncaMxQ_Zu`tNQsVg{Pk3Af z6IeZ@nV4`#HqOVbojJ2x2B539&&{iuC$3bEG(6VZ|GGC{3#^e!YwE zkzB?5W~?Kddm_|(g#w&gK-~v%3h07K<$D)W;g7(c0k~`iNP4FYi!}*nla1#SK6`4a z_$j7$)}Klc(zq{-@Tk?@wiQ33pHPggs>G zNre@46^*Bb_<`d=R95|s6^Y{agX3~>;X%%C*P|`+Zz5AjODBrUM)SrnK6$~GK)!U3 zqrYNsYl$AOuKdy^keYG;>Jv?ED$Z20(Q`~*;b8qNTOF^z=wEqt8Z4iH@7-|0yFtpD z@=OL}5{V)K=800#FwIg0J*2`fOw|wE1fGb^CPAUmBHHwo*4vt_R$1)!2Cx(-y`DGG zMJQ`5o~sg>TQGB7|Av7Uhk`ZPY-c5<2K6b=S7ZH<0IlwAJkBqhzwVN|9VO0%YG1>v z&4mOYT`#d&JugyI{$j`HSP0_GMr1o63)+g-?c57ish@qbcK zLa=?L95j%mb+epD)jn+}1|<^O817~(uk5BFP&jubBZ+t%9Z^5jE&a)hkTK~?J9Uc*(T9%&f=FNGD)62+`i@yHB zK@n$pw&Mx#a+ccZk0;R6RGy+e@Y$v-)A|%S8`Pz%AUuH~S_a139MgpE@k*=1b#av1 zo2xD53K^rPu#~Dc@@M{k%CUK`jVmo-gRTbmzyw6ge#Gbbj+}^rYWGL)eplZK>Y|NY zCk1mzT(MpIS?$OQ3N)0B;VJ?eeUP~Pc!+y3&SGkX@L`V`#YdU?fkyYlO^2&39uHtq zX_C=+1a*laRT|b^bU88&cBiLW?>1U5778kxN-AjSdbY73?zn8uy4ihZZ(h9bi-NPM zYQ~LZ9y>dO&)bqxMV0H#k?HK=XEMX$IHhuUYLjW4xI4Q2k8jO}Uj!UTcvL3W&_+ zkGFNj!p)R872J)yUJup_d@oj;n|~Pz{Q@B=gu9@k{qwWey;&c|dpvv4Qfdq1`DEe# zkW|pN=CR}I@&N4P$+(q=ut7JB{KNRrWN{m#KuUp)Nf0D$edv4V1>PmlbE9^D+$xOE z%%OofUh6N4g6-%#p<^2?OIK`hQG~bZk))Lp2WtKOQ%2W_PC80*vRL~3Jninj*?>-GvAk=$anxr2 zz~i0TKdkHLJIOHEe~V^-cbbRtm%7eN$$# zUMyrjfx8e+kVAs*pYn!?(SmMmw@9`D_KfozYpUDTtwnX9Q-U&0q@_h&5mt%+B00f5 z4+gpZO_wNIU@xermt67aB7neWwKg&{Q$k7spep0* zU*5l=oD*iDpgft+u~C}`3qG_k^K7PA?%0%-%8vuBWM4~|wM^l>y5jw9h$R@x10w#JJ_E~#!C!N0rQTg+ReCMj?te?!?gVA8+>z42grem2Oi04Qd^T1sbL&9q=vPbc6| zD(lcVhGsM|Ydc^LHxKA|vCi&y8d@?W)r~RK7YplE0A|78@Cv`P&;RQw9ddwO)ihNO zj*c7@^gqX6#ZmZ@S4;2_}8+;yaJSV3okpKa*o-Cf}M# zsk6omJQ6@cC~2d}9@}-)!4rI2+1Z6BQd$ZO2G;O~#veFNQ)t18gdCiXv?SGr+Nrf%fu2^$CoI_Q_Xl`HSn*LB(to9rKhPgFpsSSLZA$oM=`of7}c_G%lO4B zuc@iYR`GkGC#VtoXe!fUDDxpTpP*a!bgq!w{SiUcFB-}?9R!h%!^8(=N-`Giw09YN?!2qLk>_)fFy+ixMv9TpzvKGj37za2My(9gWK z->LA4-zHNjt+u;;Edwo07UT&%8-uF?QUAj;7V&KZj6 zmp-4#nmia`6)?7(J=|YTz=3WU)fKc6SulT`G09d;J!d{mmCnn%=YtFoPK59MgrFA^ zou6Z#h7q5oy3wQvhr#D1Wu-Z;#2vOJ(r+kg%|$CpTe~pqUN>e)PvcHoDX2(U0SR4p zGkO8*gG@uK9JSk6(D@+3!c;TDq+|_%g|QcWg&d@di#dpZN$#_rRks3>KY1N-RnD)F znAbox?hTi7$RMFCJ!aScjYU$3->|}rX{?xNR1p#hF zZ7PXc0)BY^k|4y250EW=-BWZFtOTAehq%19NEK4fPS)(vtQ`nT0_cdXd?1~e_SB0k$5 zuhuJF0Fug>D6}HDiMw;&<`WyYuWmg$v9$TN3d}YAS+IG=CMF+u?BE)~AzNY)iCdVF zQ51x&K3!4sbWYmH9r0)O3(629$QR4$Xvy0{c@5bWE}vhRI!fS6&ZV44Ohe+*GLQTt ze@}u#@}=Nj2L0DM)@|27-EerGXga>G=8T3bmCNn!$M(eaU?f(K5&Q`cgVh^nY`!YZ z`F@hsgDXo#+YMBnEj%o&tgI|=KkxX5D-pqG)BWkJp??4Qyge+7(8b6@f!AQXj>l;U z#{?*Yym4aVeUI(=FobQBBOsTZNKQ3hKd0mUdWcZ`Iu7o`ZajhUfuk0Y^M2n4{&45 zgET@4^-Ka5MD#hb01T<@{`1#5s374}_SBI8nNUrz!V`NiMz#~C0h9NO&*L~G^@Wh}4ReOY&O17!o2~ktCk46;Cgz$r%X`!?;cWfh1F-I+ zN4NBHnqs&G)ukk&M2V?G?hg)q#a26cIm<5TMe=dDI4>Q9*3+=jXZ9Boo8LyTiiyfyy)CTvb0XrzL&+55cO*V9^EqFoY%UkXm@g;Cr5c_?_jhijFHPU24vLJNL zXN`wX`;TFZWnl*}SM2&@zshI#VFxj+E_GU@ z$(wY(OCBOb62r8!O74Rcd`+9yJ6CkbrZV)ETLnC6YjRy2ueJ&)4iQR8ZFsv&Om+40 zx7F*M;Qy!#Mw57v=L;VS*|WG;)~BsQ0~QH0&@kN94L6KLl?rE(o9QhPDJUw=_I!Q5 z3r=o0OmqK|!r^*+Kmxt?~h7Jd@=Q9ID=7!Fz*bGM9`0 zPh%$iZb2y-a_TvfSSaupKipT#-=kqnLll^)g zFp6ybxmwxoay^vVc)nEi<>-{2_N9o3R+JlX4Xva0%e-&w9C>=NB{(}f?MvcMS zuh-F`j3@Zq`T$({yS!p17j^wxQo1##b#(5ZiADME;t62`LwxhJ>s@xE-8xF()L5;7 z$*5FLSl|be5AV%gFR}p{*k12If)Y?0*{I$$M9Z8aAjpI*+C-#SA&ZMS6m?pRvV9Q?95AD)s4{(xyRDrDops+>Siuxc%&VwXroHYEI?Y38b5f$O3CKiRyI1Yk)>ers zx2&6v$XiwFjZi*6_I`Zsc>v81GcGJPTQJd1zY~8TOVF8S@2mszeKLwBY?q8tedO7m zcVBVOCQ})$R_iPXMK52KKc)M(@`3~s4DY?mwj2$M^Ve%Cf(nz7K1aFR`>OwD0sOF` zev^H6Reu&tQ#Gsga#o6+IB!Q6NO>+bj(>@du@FssC!8{Rra5vhhc;o#m+;!cl6^vp zPv-`NSF``=yy!oytN%l;0g^^C{w#Pq|4mY`!dD*NL*uQ%gens+#@VGUUFI_2EIW4Y zBFvb4dZ^17(nvpd7yabgY_V8B9ka>V`!dmfPg@L#28%^!M8y)5g1Aq=?pK|_Z)#tA z3tZK9<*Ndw^SN@fDjU^iLl}A{`4ME(^+BM{5*ilFb?8rQ9vKvYDH#Mup4X;--G#30 z8CuUWESK-P7DY~Cn4z)ucCCEyw~zBjwm*(~XraKFaX6;kY_k#R zGS&WFss`V&bRE=FRQ;uMdmir5?*O=pPgJbmXqkhM(ja=>N9Qo)Iar`)$1NNInd(0z`iv>sa?m!nwy0Aw76S5PAsK zZI3dp~0ko?xf+jyVmS_5%&1;LVvez-n)$7EU{mEnUe9wYH5z{bEdzT|ww^BS99H(L9gtHrFE!I=cS*m>bkLUqSA-XXh@r@+)pXj^Q4?YPqA z_V<*>br>x=14Sr=9N%jXtwnP~jNn)qZ-p%e9(h9vmurnUaLpfKnO(pZD&@Y%OFND3OHhG&tmclTepmjJU^fJS2T*yAyUl;=BS za+#d3j}OleZf~4h%Y=RO?ib+^J+6nQsw%iZVo_Q0k@m}`^(D_%u&x96k2}B&pX<)& z8+1=1h?$w^^)KN%`ws4DdK^10XmP@+etX7;!&-ykrq7?kpd8xhY`*AvR)?xs7#?n94l zpf)HdxBHQcGGg(#{a~9)b_KVRH=CE_b}qe)3_YwdI4lT}lxa}V-`{bM%0^3(RH3Gw z#%B<8w}elAf51-t4Nve$5a&jvMSA z%1sc4!3L%JgTeoS#VLpfs#;yHe3>L(UQ43KVkxYlcw)>@H^| z=S*kmo#D2QXYKD8<7n+GK7J~E7FR}R_cp43B_;XgW;S_Ozeh8I{_&oEy?xta+%H2Y z0r#gS)!APj_8ZupG110w?3t*Z{DFd{CPli@v}?ttr73^_XrK0~mk?|OHLcf_mpAd`$cZCL zouu~HSiju$m-P%jmc5|J_bDN+OEfIz&l+;?8d@D3-XBVl7enk~S!`?Uo-&f2(XZc+ z(nI&j0bZA`gFAm;G@a{rxPt(iqGdu9K_2T6N@KHjiq5wOPl8j?BaS2lo)fJAt8XG0 zmb&h%2;IfYWVUH7m6QDPRdQkNDd{CH-=TNEdx`LyGFFEEs7CfEPL>^g@cyHGce#t$ zA582by#ffY3{krNI4e4!C)-kSOw;{RD2oqDw#;C%6*lbc;(F!ONX&-fO%qmMSjU>1A4H%q_uQ|JNyNzYK{l#Dbh z$+`aa@US2Sg~xUH%cw1ex@|nwPaggCP%yo96)LR_WBlU!{w_hvV7<|{>8_FD9cdFHitu%9`acKNj2r zrbSn15mNSg4^mIiPwnRZG3?lJ@`9`2Z(7W9NXvOSx(hwT6_i;3rc=z*d+Q3f$HUL{ zX4y2RRLWU^liYVVY_09DdqtooQ3B+vEVGj&x$>fMv%J+`#ChSm&&PH4cYgf5dB8Y+ z-Xo{2UNq@EPD{xU1Nkoy3T!czcQYE_LFknD*qz3!b!dqEv~J;rp{|+(dRk2b3(5rq@rMhVCzhOqcc!{hYG_eI)$-Vt9#GCi$-5^Gwj_V<|Uy`+M&(zNY@NIP_Rjr;_u_dLE%*g;=J@ZqJ)_L^w2$D1jR&_f!Cmygset z)#h@&2qfut%E%kDBq#A-5 zKJ8eJ=;VtPU&3woeQ9U^+XRVl`Fk@=Et^!Grwd-Vs||Luy#bC3j|BL^D<1=zs;(|# zXk5x_*&D@46cj9JfLL0DmBo4;;-%Zu_yd?&zBG|3JjvZ zcQ#SP5B36;v?rBKHpy@OjPF&utX>0&}rI1 zSW73&5c#1(g5yav=bP=UGDI)N&v={N>oiJ0$Br!}%gf6yKSy+LpswxyGLxgDqZ1PY zqoK%T-cN%jmDyj(f#i4|d+BB^RdfAiBk?bE>VMlO`dR2?J#& z6YYZ8DccjhHAGPi#o{q61$mCqhTPgrhBfK3Tvj>5cC(PF0g9aE(RU0q%H7;C##UMU z0HxR=W7{OS8(LuNQ91AI2|Oo&bH00-LRyR`VdoU8beU#awaf8VMQJpu&S&Q00vjK} z__HQP%cV**eo(#mrxd9s9y2ne)ys-HPEt+T9TlQo)P1hd{7VUm%h#M%xsdq+?h@6@ z`|NG)DNGB}X}@QS{xGP#=v9$G&r(n@<6KiyAjS$LUpuCFzU0rS4SVV6W^y>xULfJ+7r^@QjVt1nz3jQiuLjqEIdAdkhf(EnY~ zVhY9wX(}qqV8gy)%g=d6a3$=)$jB;bo+_9=r{VS#voxp1VNg}LubBpJ4V3-tlEYD! zDeqrUUq;6lMI=h(_mAjcJdmu^ z=DI5m?Wo1om-Fa@$I@Y}poH1HOvTew^^kNpq%l-X;d_@DL1r2JHh+UsE|LMvOgf9L zgH}AW!tHHGKvtT`WT}hsxc|LdoPMxpjLbW2Xj)xr=O9ol>Y8%D^Igp(RBEn9LNdhE)0 z?Vkzsn3y!SqFRVf4|=|==QT>+&f63(MZEiH$a|JclBBrufF`-^mMRt;4-{YjF5?tq2#*e&*X!s>Cy*E0LXyWIp*a#j^U)8UsQ5j7 z9K#}dm4TYQ-JhqBBF%Qpl90_;X8L2`=4}Mkw{1;DePw*y>r+7((wbb9$@ndz;-^_% z=Vt_6l&jJe*d%zaeJ*4 z&M8W$Q!xLxrD}+b+0(w$cFwYbsoyqMcD|o8E$iLqTnuZ{>Gv;0MgwGv5)m38^9M`C5fDktj}4>ys&}VHQb9T#!FIISI+d{gyKVb7CR}^>$;q zZbZ}V@3F{$*S|SP;%#-=2}#LM7d!yBID)IqlZO)KF9m)VasoF^pD;S9l2Q_X(CnkC ztIFmaqnE_uNC<$yY8Ms&ppejq!}UQ@I)m2S1Lvj}Ai)x~Y{8B>Jo7)Y()9NYEC@Vt zO3SNNM9Z1tB#ux@>zDl3z~*IL5i7`)L*KKqYUg>)el}>Y@mX8p*0w+SSWPNtD6q}~q$hgFMJt@uuI3vJIz$)l_x%f7Q&fG%Cwl)w=V7e<+Z z0b$$v^DJ6O#i&8kAi-QiYI~_#ODcr%x~_Kw8lTS`()l9*Atp2?DKmRM|5v`QqkD8M zCijrx6Gq~a6~^vf2uqs=fA+*xV6R*8o^ODe=R2*LIf=-p0sIDjzPy7qD+44( z<`Y{x&60Kay~l3$3$owGgk%6O=?8k=%3DZy-j^0PB0yw(NR*6 zQx?iKZD8;CY7l00H?!xe zmpCTG0y$ON&r2|Q*^Rh3;ohn+>)9RD{rU2P8*Ue_?t9Ivk=v2rr^~iWSmr{o(k+5T zHyPY2t!9uB6US3w*(3fXlVfy8e1-yt*_nwvy-eufBNQ|vSS%qB)H%UtJT^9$8Qt`f zyZ*5FsodprR275NidWahp&@d2m=Vl9Y8&UOU%_>m%XHhcZL#TDGW#YL+@R|5E%S;o zpEybW?vP0Q-yxx)gsZDDuaD!&<>}O;JR=?J&KF{eaz~=F*?r;hR@m5e8o(SN)s+OZ?* zAQ|Y?Bqf$Q{N(^Lq2JnDF#?w^a@@84K`MxYPob`)@NtM`HsL_4+3ZIQD>Ve zEeV6FsK-*jzIbgB4tl`Y;rZKA$~Ak!$H-pCA0u$b5~@x+hJJp>i~jwQl)h4Qoq&xr zjD(v@co!@&J&u*4;|w83^V0kDek?_2xu2Cr^M187RG^3l6v<{su}M(vZj=Hr?uQq)wfLkj&9wdh)oe6K$G8 z*ZDA4FdCmks{wUCK~>eVH+=0^mA5NwRc=nxT*TP+?fZUf&W%s$*>cf; zT3?jy64zPi3&7C7ls-14P z4Bi}dlNQF>wucNGjwJopl$67n^oFOLKdo^xT~9y+3+u~v7-*0i(b>GxQe9%Kf4W<> ziqUmYtOhR@R7bw@O(wn5WtVVH^ZQB%U*dcgR>ZHDY6^M2(q)-v-_ldUmbT;px)Q$2 zt9J_g$_bD6wr`Tv2o;V$p*tv(GuIV*vkgAenVCDCpDlk$c`0s_hjbw)p8DhSKFl*v zJGYh$iuL>IYOF-0IxOrK3u|=p+Jmb&MCsyVukm=V-Pja1w5ao^hE(u?8MkhN!hK*j z$X^ zWZQKFOg3m)>7cY->yORs9L)#)v_$xdXeaPgKt>Tg_5tGThuG4C+J5~7ef9B_3odT3 ze}8W?iN^)uD}CE(2V9+{@iE6QYtXW|E-poQ)1Vz}UziL0!SAx!c#i)3US6it@q9Ja>G3YC%$#S{XGTw` zw{g)rMk{!UIJL^Z*!Zi#*dEJN2m~dHPj_;j>eAAvF!2lr|KoH~#w`;#&-1KDoHWCl zi6V^yNkMd!4r z+k&$r;US@}uFj&m6%8%ny5n6YWWK}xEO9)U4t?QxgG1*>-JOe1G-bSw6`;FA*!UJv zn!wlfaW@`}?~bvp)?sKDGXJsqkzJ|T%8;=8b1C(<7>Ro|a zIKmQ$cd)e3E5z&!_LHdXxCuZEr^#HT``v!i=AFCSdYi6knxeLr7l6hNdPi_)@x`i_ zz1==^C1?3d6w}tDMgc!=ND^p6>>>_V!Vz2~5yUC678K6p2n0)h8%=en6>(GdF;`)X zyI*uD>WQ5QaFw`)IuHPm)DuL$n4nw^ry2^jqh{>DHs3U(C(-3wjxR0tgQW|g#N^uC z9=;kU2Sm3@$TQ?dhA7;htaJMeJ=(1XqQDK(0&f#3ZrAzIji!$^*?Gwk?Nv=OyY#i4 zJEgEstSD<&L%osFX91EO#vhT-gApk*3!zDMSo2$6LimZ?t^+OQN3CuLCgFwVtz?L7 zoj*R~Fr7dM9yMl^k-~*`;B>D$nM9}nHK7B=WKS=Ps&mS|onJlTf%7CJHk#RPE}I84 zZkCkjg@zl)5yKlW@wQ$HE(Ml;y~YK7>9(cp?Ec?(C=@4m=PW9 z%J1;Ye#6i2^KWoz9?OD&T1EQUG7dhXa^X}y?S{Ck^!bXt{`rCoIxZy&yG)A(>j7PiHl0<%=Rp(xf2=Wg!9HD?4Y!5TbS_iGOU;y z=APUQfhhi30hDfQ*g=hWh!^*=*W$uym@Z&xd znS2K-`rwqKpIz_z4eq4m;}k}R!ATndP#u-^^6qmxaIJuo zvJ+R@dGp+7hX!WX`Hl+Z^==-b6%)&N@BusuJ0awM0d?$y(DnQ~BUExSo9X-6TP1%{ zp?cn2-(a7Oe^Y3ohWMVUmf7cq6uE25DItxm@~aF!sh1O59=w*_jc6}x6^iyHA)(zn6^7dsO4--pnN+}&VaPV7HQEP1wR50B@{J>7$jZo|!Q{2C=30x>&MFzS#pW|)2fNSCql-Zn{ z*p4J%w39p`AwzNKC@pZ&1Kn~_H169CdQg*Nr!`;A!=GTwY})_h(aAl!;079xB!i=j(oklN<(&XAs53>2;k z(tW5rDy56f7byv8X|vFkX?j1od-TwLRW(3#?;tDhA8A#$T01LxW#T{-6crJ4nV?P+ z$1(g#*ZqC?jnskv!93z=c}z0Wad4w1%Z!~5Ltm?Ec8Jx5*-P4MCelOiHfB++R}7rI zilZjpKl72Nf;W;gUJIw|8Fc!+h( zyl9rQ`rYUVl2rggg(4>OD`ip6@OICiWmxk%poG6aD-rke@lvebI3(O3@t8IDI$Y=+ z1}z|A%FA{^Xl*<7L$^@cwV13xswz@H70r=DOyHxR3T&+%P%QOWgxrq+D;K#ZJbA-a zb4~_400-WzrsehJmy$DqseJ8i#{?@f zI?BZe27O=<1m%3`3z5*9Db@Yk3LBG;M68>X*TJz?Ez!eG95`kyv3ZQVE1u?NT>xG>=+EG(8X*=7ng}1y+K` ziPKdLB6-Y10B%(JVxrE4+|`m7Rso3rC0M7nH;pojjUDD%&_DPYr9{@EHhSH-^=vSK z&{2VAUS+pkz?`ze%WNtbVSKakfG*tzebxN$Fx%#cr>`5{HNk=_Heeh}Y7OT2jR;jj zDh~b*t#B^oMkM>+)3gW|F4ThgaNLA2oaIc2dcWkb2_m7-7R{2Zo!2nnXIPWQSd<|A zME8R0kgTVGt|x9Vsf7d4aLj!<=nHTT`g61c@mrm7E~Ng678j(3Tv+L_#W6;Y+X!NY zFNeFDtwbC5)%`lfV7S0w&ELmt<^vb4w1-T?}D>%*P99p7{DL^!M}t(2)YozK`?;+-#>c_Ljdxb=?2vd zyo~=p9{KME98EzmDpzxhvthITLxcZ3ZA(GaV>&5mx1^el`9Jje-wporMz~mjA{Ndk zrzZX;|3BSphC$Rw{TM@5eAe*X4^)%EWp4aJDUHo;p{39s447#F}I6J=YxXcq96Q5jb4M*zSuLB?2BQ(+s&Pe&{xW zOYFAqjMR~f(C|a9mvMY#)cp!~qT+6q0mpIMxDv9O3%MB*1K zsj7&Dh-sZhx2Id621iR;8l%as5N9*I;EZ?5RHMTPJk5$E954!Mm$S9^>gAO%wv8|t z;mKb9V%9xz{d%z8hXC~3pq;nuLaNIzNgc1Zte;8k;z}KWGl*>+SHt75lqF#ymv{zUD{nx-eLCK9BN+L-v&FcY^}MRFQ1S|rTTU?X*(}J zT5(TSe7c{Mm--I$!=Cymdi5Of>>6?>rjic*G(rbnpR^k876+1LLa!~iYyI@I3ghSxDfgU0e85~5xf zG2J8%a|rcL8C@u?;{=l((1?LC)li}VjL_&DrEuiW)9BpKzQtw(JpZ#&oq#AFX45RxXI=aF_vg|~q|SOZ#s4ac|5>8`_31VhXyhVS z1UM!Ae=DQ^`T7(v;H8~J#bHkI|3)W!5!oVu>N{o7C-HB;LNBr3KSCX5W{_p{Z?td4_&Z?D9_Vaw9r0J+P8aila-o_~Z@mMG1ZVTU;X z)&cWq&Tf2AMK(VlBq0r$1a{s!oc;UDN7A^=1vMJ~Rq%O*XGOtLVb zf>=DZXnGDe%+@>l?w0G5B1hEJ7~lKYS2b}X@LER(p{!uFo`D4ral2AySm-T-!Fr z{;(CTtl(`rjxbe2Cf2Qq2nsod!5wt@d*=S$$1^N_Y|>clZXZeyz6aRvqKEH(h=<-X z5wLg_G+n`(Gb$tX_&GR)5^Be$lAkVd2+fo)FW-!K_*4&nS61hrgC3#>qMud8%+2UH z__FU~&*FmO^@lWyx)u#mUB~)B-y-M};`SXKC35~K4*|HAT+wv)b--%rjpCwrcJywv z+W(kko2F55jY2i0ys*N8vE^D|7Ujo!vSmUP?4*T-fzXMJ~X^~@ncV_E%H>bIC@g6>~&GVRUQ7jDV2Gn!42(Uxm)W!ET zMW2MihH{drJBR=R5|ZND+FD7Slv-WhLykLOvP-1Mp7zr^+T-X{nH&IX(Pp{DCTD05 z1DgwuKG>nVcT)%(#5Nr*q-9YxB^sAGzBFH3t25@f^bt_#KvDh_-QH zRvsL}oMqq;B{gPh?$wM>UHcLBZZYX!Rd2r z)8kTIj@veQg*>ExeZ^dKultOB6NH%spQ~Ti7SEJDFo0W${S63aE1WQF|K@<5jM_m}k>v1I1 zuuiEw(APWk@o+k&;0AmuxpBhG(#{<|%abR|oM!p>+47>+)i7Cns1znHuI{!{FK?>C zsqL={!;n#xDn~Y~)w`nQkz00HyR;dLh@O?OBp52;6f1)}DVDSXwyNSFC@HO+IaQK0 z8Ly6igTlyA5FcN~M9dr5Su>;T0gStLCK8yhXN)4Jmq}VGclJNe{pM6Ev2t~%qIw>a zY{6td2Cg}BygK#QaG~g#41BDV0Lx^$Ke&^~Kl&nfr* zFp0`uZL22VwRi~GCwAAn0iK=poMo)IviS4Pu1iBk;xcjNfT0HQ3TjhZFD9>UELXc{ z%4N6r)2K!h&-Cs3@Z7ZH{>*0a_@s(`n+YfRr)U=Q--;zMlOosw6-$Xu9)1(Wwo-|1 ze3TJm?mCl}s`A-fl_C?P{fEmvC9`y;(NB2ms7){J<8JZy5 zMQ875$%}rCBV^CUoXjQ+N=j#;e;#Rc8YUaXfEAz<5!m>pr*Rjj%X{mEKC;g4famjm zvu@qC^%R6>*IiwG$=dTFtzdwxtbwwBvY@KlMvIXR%a`qs7$v07of=3^EuV z;(9jrl;n9W3*cPnhY20_5M#13hK{;i>n`?wKnAcP?+(OD-l%DpV7WhZ-d_^BB469i zfY|R~4T$Xu)=;ZC=&_dbxaQucmGsq~lt{sa`UiaV!44u^5XADTb#6AES^+Yk?Xrm8 zasQCsU$TCdpAn-q;zyTP(FAus+)Yv0&;Yz?MWQD02=*xpZoU^60?&{7UMNEf$)KKIYdTSqPwDznGM_UJv1;) z!rNw8=Pp9xretOGT{qOwh#yeOeA`<2MMdY{VkjKY45=}h z2!Y38_=~BD^^6Dmf5CHU*=K?hu28s0Va<6ueBR^7laCn>shmn*?Nm1;Sq+z1CFhK ziFt*G>yz~uR*j#TI3{!b(>B5HhaaR%HKYs5XJ3!^)KgpunMv9eGj8OXQ{X`w)l{6@ zo6VWSB)?+Ie?wl5mqNdtgpWO5vG&|DJZtUw0-bL^IMoapd~Sk%&;5=0ZT5HL=}+8% z$eca(JoJCU$NWY67d}SJW8W5TpjNslu<^Rg4#F!vBI3azoSThH@WD0nenV|a z-3vf{gF>{_-jYqw1Swv*USeCx!PDdV)N)&AwU+Y9d2k74Xpk*5ANq|?91K-;eO4hz zWo(k+??7WHKuR^KnT>{m84Y^~qbf(@?cYLUOx3IHD4K%z(Gba_6-|TZLL{8sBwZ-R zjkPj+!AlihYUi{*LP*8bp~GTuxa^QC54%@d@zEE6ppo`mmritr8TI*93uvmZa25WfNX@Q3s$|#l_Jiz3oHd6`vb_D@wb)!%;S(O z`)R`c@qxe#ADM!gc<+y-gQBV?lJ{!oD}~evnJh%#pi}n>HvSw-1?L8)D3;OfpBrHE zLAJK_G%LqQiKXcJSeGso^@|J1+z&)IZgyQUzgU}RPk3Hl%oXHsCV8{lo<@40G(aWw z-QRv>&s=!g5tk7Wh2LYd3yIBKY_$cPJD`2NRWv5vKhKj55@tjf)oN4$6xi*07~)Jif~@rX&jjv|sLy!Ci$_O>zj7F6yz z>w9G!5&zlLtWhTZ_oilv_Jk*HL8DfyH;-1JVZpDzmTU#|DOA!gth;P^MQm!!KNfry ztv7irjbg6D%W*khuD7|388g+|AB@Jj9j?1M+;tqKBqvp?)%nk)qvrLnCVO=s}OiH3%z&cm|j4>V=ATP&Ltq;nM( zcbg(?igTXM7wV*6Z~XeDG9DGXbd6Z}A_ya-tBt&}tH3hON_)_-j+gahC3{O-+c6VO z9uxD^G^Fgp<;I9?n@ZK9V$%Y5ojo{i4VqLM*UKwb?=CTH`Ho_>=hfp&;YhvJ#OX=P zrg3spA~F!MR*7{f{`8^Py6l&=P3X>8hM6g0qY%wMg=v zUcDXU@P6T{-$e$2j)cr~!_HHw?4&e2gOj8$VoG$$)!$?$MD=8l88!=-VgC=jB~tJ| z;4Qlbh08<`GGZU@gLKVzq1lARQFRsTbp{ly=F=%axcbT3o72s*dIBah`lBMe<&6&}ov#_Gm-abJ=5hwZ9-{z1A0ws-LvcIodk1hifl_DfA zc9XtE*b0coih^MjVk$HnqqGtMoTJjb&bUI3r4@eTjY*WdI+^TMOW2W=u$Jm|37Z{j zR9U}wW;E3^|7u4!_{Jdl-?3tb;LjD=q?vFtLnqkN72ZwK$C7F`^(#>Dhm^`N0ZQlaH;s-D`Q z3NG1}%Qj`;c{a08w&uM&aUxKY^40ucjXwQ3J>A*04Ot(CH}CE-74|I3nT8ytE#(bD z8v+OrWq<{VA;aV&rY#j=q+#?=SDgs8TXly$JQ$r?y&lY|>)}yzt%v<=B8riEg8FQ+ z^6hQT$7iPvoyBT--)ZNi)&7~=J9B94cSzVI`_SJ9EW6E>-A2s|GRw`<>1Y##utXe6 z&HL5-l5Ee@fI+<6&jvhjN#-PA+V}HKf{*L%#L^bkro887$x&BTP0iI-%~xyP-eh~` z-9B&@b!f@?eA#X@Wkqup3O}juFM`g4M5}5M70it4#b4TeGY{-B)ijr^k zcai=VbXo-Th*ni`;W;RDRf}>ziqH=RP;RpnlIuV#pw&u*@A{?l#jr+BjTz)7m#mzl zs$c2PL%2r1>;3H>3=E#oD(&1n0j0fsAqVa{@`~Kf56-*2HGj8&6fF_yNbN z0X9>%IH1cM6`2@5WW=Fz`#M9Vk;(t);##an@bYFqgWwssWo$t z;bd+VpkaZsumK78XTG1YrMXyym>;jG)bFVstg&xAQ}@8 zRB_IfY?A;7?a6$%3Yv%bbH7nNHW|wYp#OAh7;89KSh9td>iIA+a6aOQT9lr%o&u6% zwrl=7a?IM#itFTyhX*3axW;?T)xqw)OqgGptzFNRko>sec2G;Uewa z`?2s-L-@VZ2%Q_sC=5N_`aifZBnI=$3C~hq0gDhaX(DHN93N3X8_~811!s{U4>hg_ zTx3W>f6^dI8Wxbj&XZ}o8yYa{o&%wV^`!edxMZ@8RcTF{>V`Ku^xE)ni31pwxUtrhd?x` z;}Q}RdjlgFgdo#MG)}FJazA`D>^Q`AHs{qDVK_|iv|HeHi%T|JE}$7=qH-JxWV(^v zvm$;O!5=bi{%uE@9=vV>0%P#x=L%l_Z}f}aAN{EVL30%WxS1K$hbnxSU(A^U@E0pm z(Vm1xe}8K8cuSAe@nyiVSk=MUr0unq0ez56_3@(&i8@lOZ0o_dDB&gmf)NZ1njphG zvb%7VfD#RLqvvPxHZyjc3$|bsqypxNA5B22iZSWi21~-q{Q1lP$?w4d60y!n%6?Wv znQ2rr>FYtdh4nUl?v?H+$B8T^ufO%&9;T#x2qQ80Fx+?;HRg%RX6_Y#s9ceZV-kdj zAGk!}L;?!M_{3=8YX-=hM(8jqr~W+b5X>_##Kd4;OBm0{lh?yFUb-!qj$bk z<+5en^}SZ9Nc_Y=p5OarQA*K3`t%U^|?93A_9j%@=jRG-Q-*roO1p zL)M-?QJc;i2nw(Vb+>v3Zpx2?6kT|`^d2Dl&+u<#-&EY7>5J%?70VjBMUScG0{B=l z#oXU0tiE>d#}^YlNhF6me_cb;6M}Dw$)BoQiSfUHkQNae!X1OSnjqYw1!sIN+Fl@B zDCJwvQ*5+fhCt5PYO5RZyZ;=+8^I--a8iwx>S_TT&+lz*-YE%bkOhSc$6h(4r~XEa zu0_8;!B9#Jw30eKHD#BkUMD~2l4V1+wwi1c3!`1^Z&0gR*l0+Il@l@}spO$5Lc}5C zj+)jqW95mq$0nv+z+bu-tAo3~$;>UPDL3unJn!ppdPR?*$v!83V>da0^(d%9`x6eC zcDa*q{9n;7wmScyU4j<>J{2(Xw_(}m8Hk;UDQU&|fv&t&EA zU&hYPtQ{-TS4n)xP8fRba~gL%%e(a4b>{~rsW%thyBo^$FlYx&YL z7r??-D4w)Q03v`-&=p&saq;``%c}!XK#04|3rlwME%3Lo^JACW*3#jVR}W}xR`l(= zOOD3=U3Dm;IE9tMdY~C=(BNZzyxO^yE3_a&O!4UGsp=PhXLW~7&qX1ualFRyX_gO% z%2z>1fEMFGuvvbQh=J%d1aH7eb9<)6Sgu-PW?Z<~apdst*f!@3TJJy^$@N2JJgzO* z_W0IuA!?(?6He0s)DV{MbF>wsZ@1LE4iE_$4c7x(*4<(0#(#}#!40Amn2T9LWJ!@0?e2~PmCkx{bXDGxhbH0C@q)TYMK`NA z5tRm&W9aq4Gv*P_EvXsgn$}Rt9hp}=A4-b+Aa!fG)Y*)l1A!|J>2M+`1GeZOO%R94 z8j5qXd&{&LZXfH&Hq1EdmW~G+eAet@SBVZJ92xWIR9<&iA8-3CY>h*D*i}j*M$WKN zm!?ShUTuLOWSF&ivuMJCN>B=s= zMbKo2Xq=AM>*xt`l;T`;JT3ILI^AQ+Y~Mh}or0_DUR{BnxZrEkbBQB%-83j$^7^jT zEXg>naJX*NmEso2#6=<^dU_;qY#z^L-SG{Vt*8L{f+5e_7t&A4xf0osYU15HuCV(h zQea{ztcJs0aTFaEB)Rz24YFLqGaeHecNbe1a8C6N;XFjzf$QD(t0aL$T0%@pt;KueG^t`p8r#Lr+iJ@F>2f1qt^M~{ za#S{SY5TZG~7qf72$RUObVNVginD;?IUEJcYo&^0jF>CgdFth`l! zbFn7G>O3;Bu^%EsjH*j=&})S8wafS8mKlnbl*2A6$)5GkX&5eF{CM#0;$vjCKUTGG zU7LH}K{LC5^jtU#3iuc0r{7xeW7zdBeILrFI7JMOhYN=ZS^MhBT=smN5q#}sz+8;5 zT-B-5NQXjYxkzr=bY1v8lQ1llk3cHzB+SX0MNXa3wv7}c5#h^m?#z%$OsuA3`s@BU zb$nebVNAQlqUGqmj1|}0bqAJKWMI&+vjG+f0-U%63f$^u`m?|KDS~Wb2O!RlQ=P$9 z!4O%L)jJ zTV2%As^W0k?Pm9Ioq(R_!~9(ak@eP8O%fgUF`+ii;_3NS9s}F3n55uMttMWw6w`i? z>_p3TUd{e7Q4Vlj;va!lH=8rj-3mOxCyVj7ueDQcb!U!>S*l2(CvVf4{pL_i@q(&1 zoo=e4rW=@$NTdFq%c1)`mT@JVwWa{zVC;YX9I!_m)F@-h-mq5>xCZ`F%HjSN&n@*qle-&bSuw)K5pS;qY+&N?6zQQ0ovb7g1@O*2WxOpg2r$h--8(b}{+cF6DV-fc4( z?%vqWINY==nfR6k3WY0GEHP?^?6#D=Q0S>DbhKQlv%~^jMgzhTKt|L0p^J}Ml^pp? z970=~NP47CkhbeRT+bZlsdZcD?zZmxX^nwL(G9onI0fFI__jVf!wKb2-&rUL2v#_MnHHCQ<$0+23AC&$$b zd^*xAJjT%G#}wqA)BiRPp^sZNtg%x$Cg8&`@y zALBZ7UWTY6wvOMPb3#>QB_y9bAAVP^mFWj_`4^q5bp)Xzc{yu6=3Y8!G|Y*fz2T8V znQjxX1VxUD;Mp?7+IxzIP zQC~=qb_!A_Bxo*^>|JLcG)fcb-c&oQ_W{TCMTr1HBit;T;Xsbu>Oju$tYCC@iMzdU ze}6JJ+$EVp^UUQ0gZtaG9e7Cc4zc$+cY11$jm_yz!?y{g=i! zN%Cq))Od^##&F%v&&l!|qK;m6v1_fs5ilNPAr8Jy&l~BI@%cUfWV3uZt}vqNTBT@e|F_lEB;X z^QXEBpvLk*1z!pw5@gMWF0bQrtkA=IhyM^c{&;;-eYz+~=z3IVnc&vCcM;FHMjZTe5ZkWnkzcwD9nFI`Iu=U;IPxQ8CHi05sJuZsd&!U zrUy&op&njMj$v&Jeu^p`iEeP5_(1}=W%P8eU=jC!Dk(^4+)Ce3=(v@6PFWNF_!ad5 zv3gB~ypX9-daLxbeyQ+;OQ6%_km+>RkTZ|Ntr+Xwy43U`ZH{HJz`YDW#a#!fU}xD;|!Y%om6@75c?Q6n1|^H5jw{)b3BSZbjm7Lkz!N zL7$U;tDz~A>&~amVUJcO|-Htg;G3JUKl%Erj4~lS<0yq9?Vxr6$z`m2HE=k+Y5J&r$A-$ z@#Thf+tre_7g_3h@QMTP*G2*O8OyId6bR|eK=0*Ov+u5zgT{HZKk#@th)V>s+A$Wl z=-WmHQ53W*Mq!@N;o4r!g(ji(!cxXz&h`7M`{yr-B^C%c_Y&2P5LbfDZfj_R&RG2C z;LcE4ojxCEx+3xAi^HTfb2N2233c%dkJ&qG#3$_EPUF?CBY-&PKJT>7UrIE>I+3Ds zldY(3za`?VI8=Rwi~sxqF4wfQFko(WSO5@Kx~no0y)h&A{S#9^PH1=_48;yEneoe( zMs}P=Zi96EhP^Pcl!%y+kWfa=dkHmhGes2cYlHemX4#m@aKIA zFR(J(bJm)77II1`Awk3%!n?{0_AFxJ>#Kz7qceKwI6RC3mZ1sa`^=gNZ=M>gzxTpN zfpxbeL)mpKJ=Hixk{_k;^2;yrAKDpU{2~AVbA%owDksbK>15>dq1Rq|(}9hVRIaq( zpE`7s{v2&627OqT%t;ix16ULtK+*mw@cqRF>fOcqgVh% zlV6VpS_zW^DoG72_$P`0pv2K{@-Qr52+rCH(G-5yDb=N~qPXa(cL$yxMv;b1TBISf z2Zz2NlaTNO3rw9|o4#f@#{CO15WbB&m&<~Z`?~<%CY@R%CqDIu5hO9aAoEt`X8lFL z507L^8V$G5Q!rKv3*OJ$Ytjzn3nY%3bm5a!K`v`Zb~`8)sf52V&gP0(gYa~pu>ek! z=#DI?I$`j)2Z=7BkBeIA{G~IkCgcp~rAWM(U-z}ICN1qv)zH`v{tP114!kAMYbHYF zVH2Sx_gxWe=@)uj^aNm;;km>8_Z~|$N^YOhca;!^MJikPW&?$ZhMum^+EmQD^|;sGP>cz|{-ZxoKs=$9%=AocHyF zoo=Gsp?5XKq!8QVS~DiS9Xl>qtjjuKZo3qin$Dy`z>nhNg>Y>76Nd-saFIYyW6k8? z2-^>s`%0G$e3pDC0Xx?U-&Zi=V?Bb6L3^#5+{r`c=AfpY*s9^z)lqnu{i;%_2~r>e z{XwnSOvTEIj(!KR?{o>%S8pXil8dRIucN4#kd?4;%Urp^Ut(-z1T@m@?k>|Cs(Lk* zAFSbVv9(^|?`A)puQYu1ep0g*R903}C4x6R=`X=sv`r#-Z$k0OHnep+9`$4s93~J6 zFj`#-cuA#@fP|yPWzHEb-9xD;+TZ$GaxmNS`Gov(kYSfc{FAR3mRpyEhmO|F1bNRi zk%YE}`^8;W4Va08#q|`+N!F=SVXf$GepVJsUBKmM3|*Zg!`o05@&uvs+$1Il8Ww%L zBuNkVUfM*o)u<|Bpeea~?(QmVc3)4eGyV~Uv*R}#nGYi=R*ER3sh<}}S8f>^oNQuR z^Bp+KqUbD8mwB?;6anCOgM~jl=E)iIk`Y9a2!ad4{YC1p$fnt2u_uzV#i&d8-8|@A ztp&bsIAqI~uxZjHIk&{Dj2_R+?uXx0DjmZ^3TrSCmIX73uBF!KJ_7|-?5SXX>6se; zo~>XR2;ieFAIv0um?|bkCFQy;@ssxtG!(UStI{F4fd^tYBKw)r1}h&k(OKLxBD1$B%2{ zb(_%zZ?p!7C&*jgqH1{408R(tK&*<;#hw^YdMbu$!eTFr6ZzdbUp)k)w3N|`)@v|& zW@PnCuLs2GS8GS_WW$LkC{PitlQIM4f_2805^PXm(|7%Tk?YpF@fndBQ|9yR{CEmM zJ)NXILqn`U2a4-jnNW!<-)XZ&L-72yLDMKfA9&_hQ|UN0ZRa%iaNd;d$#h(^sVT3c z8phY2%GStUC?OK*L!0^sa_$Xsx zd_9Jy3%ErP>qpmd?Xxihx5>z5*7v?_K?wnWb>T)H76toS>2w1|_6cfSx}DbzUEe|P zk^Y10ZH4+6Eq*NyJ%+{1^7-t~0roLCfro@aUoEu%=GNH+2*Mi}&H8R=QA$m*iQ*8=}_v0_!e1Dyq z63>-2A)Zqir$7<}5+Zp1Bk)0w=37*G?A2$YQdWBAuNKyMo3{(W!aBdglFhqq!N=uR z3h^H(dO1$Gn;isid9YxgA6OFR+BF$R&5LH@jMx<~r=^)y4PPm4WVw#A z-EXo0^a1ywlBdj2D3WO;h#5JzQ+%!bxM8(?mQjD*AfTD5w~^~;l5?Y&guBTlIgcU~ zltOZHz85xJqy6OTl4>qwp-eE7s>_Na-BL71TQD6;944)6o5Z70p@7Bj_gzq4$?|U8 z``xAw^nHsrjhOk5Fj;~mQQ54r<7|Tk7-OH-mv(!$a62NW^K%k+(@F2M5RkUq<%{Nc z(j2}V*`AB$N()gzr!fhlX>8&UguW~qaj|RjCOLwES_Ibzkg}n@=1vdnkBEV%uyH09 zbG?K!|1eW8Gn~wCq2lgpq795+HvF~S)tE5E3KISlM57XsathWj;WHq~GU8EC(-?)@ z*KsIAZ_XVQ^Jh12YNdfmF;peaXVLI6b-lYFxp*lICcZ=G^DpZ!{F;b^^U8~9Jwo3w z80;1sB??X|n^#8u=1`sD;)dzbAJ0-;XPgRkXw>y1fGEXFn~a&t2^GjEs#)h}4mIXC zqRK_2rBN8BswYnGP5Sho`fQ)uvYqL0`ts=V5^`AC8YlcUPSEgXe?pd#S z#B)t@67!+0Dka!IbXfS-AJ-u8<=Ne&%$%qsyHEya9x-|-`1bBH{uvV*Udnd|Ti9MK zf>OLVYA&36VO4|yoqE7+V z!ZfHP6yu;8x@z5nFS*EMpDO;5;r;b|rCO;41cf4O(XM(&;_>39*!G)@P0gg^mYB@s z3lBpso|USu;a%V5kb=_r-+DJ(Iv7wAK%8}^?KzppoSTzWyGJ!p4X9VTBpoNU83~jK5Y_4 zJq(n7ZFB*r4$Z+rzQ=(@vPc#VV(Zh3x@*-&zi(n{9j1hOXC!VC7o7DbX6}q=ixxZL z#H~81DLm={gyLgs&ogn$vhF;cOb%XPFZsk&eh!TMw)v7Yu?}}&H&~xMOdvho0v|h{ zz+o#@&kxQ|QSZumoC0ELu*ZQ|nz|E7wcg;7B$2jRt?nA<59EBcH(ka_5}~)_=3Z4i zM>qGB%*pQ%j=$K*oVK^&-2xbHA&MW)M3u}9==L8kw#syU8VVn4?JA*>bcF{3a%zh7 zTbTDBv8)-z+;?Nm;2=K&d^=?$qQsvT&K`)OkpEin zXjd^u(93)`2wo1)u*^XhoS!;f7zD6)M)m^=3{3bfkVWl5^K%3pBKc0|PsZ_>p5-CFV(qTdSS?z{i{f%S)-Wh6# zA%nTv$qAE^vhTo2N>GmjiBrwC4k-}z=T8yjvhwY0=8Ps15HFrl40CGpS|1fqjfSn` zii`#4mhGDh9Ej9?0cZD0I~>lwvA!{zPI8TLwe9J%f%enYxfU<;in9OU=1C)D0HM?KWr`EVcSpLO z(RhIQXJ9vFivv! zCm0KH?I=lA2ZPmnU384c-~F&@Nhc#Gr^jo^dEa*FYlHdv)UT?+oIRN2r0UMb>#mBm zWGXc~DJ@9XYPl9VNF$AIR6%sHpLkJ zgmHoRjwFz!n5i#LACR*K7%+~tb~JOP!sf>q*qsxpT$)FX*3Q5{NL|l1H4|HLzm#{c z(H_7rO}r{T;vcqW*5w&x;16InF1ZVxL_)^JiUEMPE$6pOkE^;Yj*r=iyV+Za-X*Pz z%MC;?<&`W*>9**m$!!-FxI5ZXy(Ndy%7BPFam5J_S>D)rW-JN+Zk;2zsmY0R`cW#9 z=T*RUjK6b;9`u2hdH6{JT^_ZV6QgkI@t-vuM-4=(;TkIJ=UI6XCo{B_qvC1^E2En$ z-NKKtuOnJ&5vRKXm!#muj(Hx+nOj$gcl1 zh{U8b2v!#}Zv3l#zdyaG)Oj)@Zp*TZOb<(c#0)}q4r1wbSs%Ojdz;t&d zeC?FF-!R~WA7-y;nr_xhxc>HW+37Il)pMJe)Z;m&jbrb?9p8?<$+GiF%;7#?$?h0s z*xXA&Py?eyu z)3%+W>vJ*F;tbXw6kq4yq{KvC!` z2OS5&GF@XWcag(0U~WqPXyCSomjCfsF<# zj}gVp028XR5-yj}ecs7ONMt(K62V`97?b1#GXXte0k6t?ruFfi$>VumX0giJ9ePje z1uH^26GuY0e$Eg8IqBEj11q68Nl@3}Fb43rC7!u_Xr1#&-ASflb&~AYld!b>5?GS0 zm!(OdrT-riMRdHXu^`0=@ffb+43_T_j!q#6iZXMi4V24^{+ zszcsg?6t|M%DHe6ysa+@JNls9x`KK?e>}nJ0QQ={%2TgC=oef4T)1TzUQt5I&u#h9 zTs>n*%S?LMU3j(^iWD|%jV;!Lj>=01>aEXX_0g1Rkj~)ylMX!dc%lXqpEKV=-rhW& zyvYxkn29{3rcUlmSVGvQkN7EbdU&tDhKh={@7C`MMQIHhQal^FJjZ?tK}ce;oL858 zox;VK=C7_uxo7o|=*Z(b4n6Fq3gweZlUk6KU`WJ6nnxHDxV^gt5w0DSkenqMytRKRVLIqqFYkF|mfvFMn z7Ky10f68@hg>!`F6eOk|654mr zzOo{G90E-qg3FFzQd5(Y-TQ*xJ|ypJzJ&PZfBt@c;sI`!%Ro>f^#wt7gd~IILdPzW zN={c4C8;RD=>=Cw0+UJ!HH1`6-v*~{w4i2WKW#B88+P0fA`%k1!T^cE0F;hM&1;XC zuFZM50@xW5At4wFX~G%4kV!!=i_hbcQ(iNS(h4|C7AeMi%$$_|W`?S-&R>4ueC*)F zej^fM60&aZ!cntYqP|>rpAa(C6a{nhq}f2z;E#o*zthvH%WA^J%AxZTNK}HhaA@T1 z8aJm$#{3bZQouh14CiNUviu19p3jSDOaVyOWbRT>h-hz)F}Hy`BDY#_u325)h9dBF zo#$EX>LL3l)Z8aTTJw>w&Z)PT6K)hHy$6lV@g2U4#Gw0gPseixUp9!EI$AobeKL~) za(fn_qoBYLOg?{48%nr??ygGQ!jA_pBaQokR<>Y@bz$*WRPb@?FF7vae5fpSAJ=2P z#mC)FzS5u)5p9PrOPhv@vStD$o6yK`B8of>ra35qL`MrwB<)Q8Z<8Y2&!kneaLgPT zxgip4ijA7QH0!jj#1ExytfnTD=9OhCotpy#V6JIYD8kPSe=YYMOx9n~E2C#d_7ss7 zv$MjEk4&J}HZNDD;pUll9gq6S!Am6QPS!DCh7d-P~osm zlRkfQJ4+VKi8)4$;GxyrC~lDcD<|j*W~@;pVCnB5ZksMOk-sB4m>SRWaKlT%V&SE% z94dl?g$}7&)8yL1UxQNjm%h+No;ijHA=5L;xYOvbv%hNCAPi+Z%nWK^`T2TVSSxQhja&4{GXaO6P7x+j8jaecoGT@jo1N>83=A^+jM@Y;mudK=bNNp;49u* z#UQZJBAP83$DahC4Or~#L;~FyU#uIWrA>(NE&|{*aOgN5?4sCasXla`?ppBL28WEF zXTE%Ay{QQZ{kD0F#6hA4F5m{El9OkBUdZS<_(R3&v%~&6*+#T<7oVgQw(lphyPy!B z@qCG31TM&*84$G!B>j;aod0qnmI z60wDIp>xBRN_B)2Bst6LccbXmz~;RXx6=vQW4dcQ$hmJ6WrG<3lIqIHIrhK=h9Y?7 zo%P0YvydqwqF~@`TYi;?g`xc~#?CS-&R|Q|f#3v(;O-vWA-KD{ySqbhcXxLhcZZrg)gNWDsoHQ!$aDl&?6t#m5D2cvgpTkH16eB7u>wc6Yk=y{R zqZEG#lu_}Z#Bos_58#x69Ppx7hg-%cb~x;~^yGbCoF@2NG*w4)25ExgM@}L1nrdXIQPZJQ+AVt7 z56GP%IKe+?VGP9hDeeF4cYmaeyyyALed|L0cuj`}*&$I24twi%HiyLitUs7A|D`z34*S0DS1U)?Fdt1NIZpt3?o4n|#vhoj3Bi=@RGA z`vv;S ze*F_c+GjbL=kfV^Wbc1%`1fztNI-2{!mphF`~Uh+N7|;0{Zz*0zl$(P824Wf`kxy9 zJtd1MP}_BqfoZP)@ihPa7Xf*k;Jm2M@?8P%hlKKE1PJvf$zB=dO_5v{_p``RdX@@k z*5O72EKSnKI72k3_A`|mr9>$tr(K&4aoQAtf*4&1HMd?07UgM2=uxdXYr4tw1r5qO zu1H>;r>5i!q>>_s82bt-EwXtg?l`-uxIpf1^_elZX^hoV&YN?#O|Bcfr|sT#(1Tv? z@zpy`_(GtGzJ3TS^*xPEVOS_qr8erUt~=N55K~!zQ}m1=L1pV)j>uD)Le?`WR7CIk z2`^9q4ZrpiG?TSZXr&GZC3ElXoXMqUh!|9Cxp(8ZFwh}qS3Hfv`^r&@unyjnrnSD3 zG0u8Pgp1GoBJ*71DpP*m#Lf>*>U@4j0XbcKl4+)GOjnuhDmE-`dgCS4l4 z#3n+4k8UVUY2}wxFpNqhqdGzYL5~z4_(=1JX@8g4quY(p=UoGO{isc1*LJ#(i>A{q z{Y4pTh79|n?-0m+E??|Ww35zmPCATU?ZODT7v@9|!nurExI-KtG-C^OnG>}Arx2ZJJu|nwYlkzL6mIOyO z)Vzt|iT89LaL4kB=N^A?Cw7*y(cNi`iPJ!b#HH{d56?0uK@kx|^;y^qdDs^J+e4&5 ziw;m;fI+kY&k*xwxKy!IdPV`}#a={Q&RWkxu2dYP+&9nsa5v=#tqgr!hf)zmXZ4n4 zT45a3h^IuqY=tQG)&7Kd?@N^zDe$|nou^CeRGeE0vPOJ8wywqDJxTaX>6K!k_xPjr z(%p1)z1#PtORpssmuG>Rjl;4UG3BK*lX`pEFtgQqh`pP~ZSV&*0r?Xgo+IAVL+0V% zmyh-w^s>G1n!PX?4;%Perl}rCG4RcmtlrEqz*WAz@<6`6?YyuAqdf0dGIxrO;@gMyc zk$xMd2?;!V$%b&lv*Ldm4gPt?eWrhhh4IkXkNAIi!z>6vH*A%^?N;euKZkKK&~r#b z!2clrmp5#NA4cL_nGW>Cu9C#R6X{_%Mnq!DSHjRB%~FYAG?jhD5Cx3(++0BfY9pHp z8t8P8Ad8aOf#RbwS%VNnNI0{iHv~PuZk2v|8IlOp!Xtr!73>uzPPq>;rs$&!{Oq`dg#c<<&{(qk_@R2jdKEO?Yf>J45L{y47p( z)$2c_p>{Lylk{NP>VU*A;u;Ycu`;zk-Q0X;>I%ea;(`Vc6@H8#eH_^gRycV>JeT&f z&Jz3E`MN(XxqtcAF%0?`1$X(vEY(vz4RAcV#nNqBHuS}WOtX(Mesv=b_Xtz|_OnU_ z5Xx?bHNiy?;O1A#N4Plht5A(h^6NF$Cu~t~(bXZhnxa}nDmhx|ccETjTRBf|fbQ0< z{23Mgn~stQ*rjHVCjt4~DiovU+2ND@a;E+c3r)ME|vVWD+vFaIW+ z$_V3KNQLty^y=!WtE&qnrS7dtAMepq*L_^lw#SYc=DF;-ZD4t-7kKpoE_*+mWw2T* zEmw8CuWa4C__M?On5Uy&(Np$w`cqdRf^UXdxFLesv=%VWTW{?BG?v6-Ci7_3Y?iV{ zT}QE8KC#EFu9kY%l1r}`m7AWvk{rzt$gnB2s}mTYV7Wz^`|W#xQ)c8@jexe}U0r*j z4lN99M1&@=R%u`Wmn=q>+>^R$tlA~7>yR}iEoaTpo-Qi#8(E19Oq8TETmm~PDZCdH zTTeWBvFx|@sg91_&o2N~49WU>Y)wt3v5K$FrIO&SLoGl?UT6_U5^vo}Xt&xZHI9el zG;@+;MB=V{1ev&gOB%V6)Ws@PQNzLV*+s1!(V=Wzs|pm($;8j1i{vRq;&BOVfugqV zfm$XRKUzvEhL!a3DikA5npKk((-R#w^8P(`_P3h%=w!&%&=Cojyb zaH5U(Ks#bK)Rgrrx)%)!_ zD`DqmrX#JYqG6qt+RX_}U~7zskU76+M<=Ki@SRiC&qF6nOIjDkz}Br@(#n{2MfY^? z?xyJ-P&AJbRF{@#qX=GKU`QHU*eO?qG?o2XPT3KJVpz?|7db#6&_FYK_+bFL_D>IT zye%ES3eP89%kFT<*%|nojos5|ew}D<)%P@L8w%=Gpp#_DmLZ+7q^01np6n+lS8ouo zH@7ItF1v5^U7Ukwmo^uyYT3fMA2;8GY(i?$lMqW00$8ff&%Am*0HQZA2fK}>b5QC< z^bFiB%4u>!!5xaWjgr*|+{veo-oND4M0jGn;tmpXXt^i>dgj;VSPy(x?G3BCx%x&` z%%|kKIzm2PX)6?-u1{ zSio^7BsD3T^*BFwcfpFMEW}nxjLO&~Iq`G31R4HaBA3J>zhILP5E;4H)y^i_$1lbQ zLX>uNB+^~#wU%Bq55)^9JhB9zAxkX87c|=IAI*k_2r#Xt@V+=M=p`MiTpNh|#qo&7IE$p{ts1E7 z1l8b__6wfw=eur}&bm5sNu+`nTF|(>VO6I-N161r$2XL;XI+CDLz#1p7mTMf^R3J;vzXt0cG@E7Gx{f?`wCMH9ncfs@Js&6auc>%F7P1Q(kXkJeR?++_@UR* zx!Scmgb7lS4~o}g(Ej4kIj9`ZdYlS3oM54S&wQ>(S5H zi3UdYkSip8_KE-Zy>(>I(xflgR5XSH4iPxXLDUVr=T59r#%`nHzd z7g(;{(*w8)iV>u$)`uJx-gbFPo2wPFvI#NuWxH1O-uhOYQcOUZQZ>zCr{@EUdq$I+ z+bJC@t8{E`&HE$2(KV8z5?#D*#M~e48V<71ZNc=DNT?oq&sdO#?8|?QD}rex$J<>K zY3Zl_jv#R$q94-aUjDK#2fQ|he!rmv z2a`!a(2GmKv1KCTe8r%l=2f}c3%JGM*xfr;~Vxgs1m%iS@L$JK1)ua4Rt ztCg-}Wzk%>Wj+US1G5T_atHo>a6F)L)G_Ly;VfYFo^F#YVr!f}lq+^H5ro z+7`W5A{#Dm-)RDjEVxy*@1|`B<$iZ7#4SPn6e(T2H%0W&O`(Lf{E9}3@J!Z;yn|BC z&q*hIpJ176YEV%wE$b+sTim4OmC3~0rJ0!jT{2G^SqKfSpE*0t$9Ed%C~p2q^dU~J zz(3?sdUEPqX-+|}bq{OsjnBR&>op74C^3dDaE!kG#E)jR=l$GVVo?#b;BFMRBI2aM zear9nU~CTI8_!OSg|SSk4C~b=Y|j){eKik+r`UN$sC)or6 zvx(JsB;s(%`9brf!4)aLbL3Uj9EsBR%X5MY{?(}cL$4WMGU{yn!fZLY zZr|&>()?3v8i}Xn9E-=zM$~i&gcIEKdI2oyzj2IYb+g-K5*@G$${hGu@QZVe=| z!?1h8p)t<6dH#aUGS46~>=JvF!f|kGenGkPi`07wiI8xMj;!9geqP@PkvujVc5eiv zderFROnE+mP1@|W1+%nSsW-b{cxqjuCO(rAXz|5W)OnV}W-;43K^{m9K1zznxe@TG zJlVE$9Jt4c$*rilh+yo6@W*#rRCV9@u#S?FN?ZuzHYIpFRZhv`sIu41N5TNi6c}Yw z4VhVUw%{qZ!aiH$kO&2_S*Y++v8S!Nyp~{6?`_}rQQ+WIVvtD%}r)%uf4&HpFcSqi_zqxo~N_>x|yg1iblI8DdcfzV{o!@1Te!3e|#o^$w>-X z$L|@#BUADY&F`(X#2-_l5Ckkq>11{4#}KR61?Iy{>Kp6(wFgaiw{Cs70<` z&HW~#Wiw{odHR?vH4oV>hyY{~Q$>QEjmhG{JcV;^73k`+&e2mz1C^IzR4$2)%_L#eNMx+q0vay?V8IE@40r=(*RR%lFZr6W`&UaOZT{qB^qIexKzyHPdv;y9al`Lf0x*Ar!1mL@i zuOy9#$IS^cKimTQ-Rk$arLWLLye8PH7Z=0bETd06(_WCKNuCvEBeH+YBX>iv*TW>VM%F~?X6FLIUdg4*izTSOorutcM?r}`!EPpwD`BQn{)y_ z3(VORcoLKzpTqH7$>{01`{w4;e&hS4|J#f7bZ^ts=_CV>Te;;8Ctq8v zKqsUWGiHe_F8fXgVAYIC)1l6FekrOaRdP0>uvO3hw1P7C**XLK`^BX{Dbb&e8V1tZl8Hn(f{b){t@0fB(DL6Y$Zs9@*ogDSuUsK8 z9>7R^&JdUc)f5b5k$Nqbf&SkvN<&P7M=VBzn$2G$aVKfUN@vAnbA`xUB^{Y|Vd_cS z%Nq@-gOAX>zvNn>DidD@EXs~w2t%~t&nzx4NQm#tDg2e3FxuN1NNyg== z&SnKbz^PJo;%>47hP!O*Ybq8fsu=k9Eb4%DI@pYRwXI=sItDYamr2Sv$2+7UZ+gB4 z&Qu)+lijtzGMz2VDTblt`69|WFT7?oALCax5u2r$s8%%|Fe=FA6`ZFpl38xMkQk_T zbCPl|8?wQZ@(thB!dvoJ*+P|NZ+4?KucM5uW@<&laO#?-8dz01!l0-TBl@!wukEW` zKHFe;4PK2cTN-<(i&Pd*l)I?F_ZgX5CN}-~$*OJcZs&dO`X!*E^-W3YQUQSAShwgF z?XdqJECAgiHsQ16H)*w?9;C4dr~?_;<|PSsc2s6B+JyofADj>24Wdp13FPkzF5T-cM|&|pj8+(jBPyQPkXbZzzm~k*-YszCSC`jN zH^gcLwEV(eB#!OVZ;fPKSeA~A#4Z`Hk>fRw?R)%NVgH-!_`9cXJpD^C>dYV5JsvX` zUR@ur%gt`jpxakdqSvWxF6R>>!5n!mT_^tMBhofNGTCt~&qd$chu5h-2heH$>2fS0 z5N$bdp18QCrKYgZXkvvaAE)ZH@{NNid?ub)Sng}T+ocY3Fit9?KEKn1gx3Z6U*EIZ zi?_um#^aCqGFkmD4@VOjlhY`WxCJnk%ZH1ZMyGl6dOO|vl$xg}2=jeGUJlpqyy>nu zT-iI$C@iaj+CtmFmOU|r&2hLX8}WE#xpt&AdNhRnx6wDmwnM-Ar(z16;kpC^jeKef zSY`wXnP3Is^teMKA?N&0F=8x7?T3r5Y-{@qViCW1slKk6*acIP4dHnuaJLhQCQq~yN7CUh8{a( zwL57tg+_#+|HPeAQCF5KRztv7psX~!-YK^&1X@jZtkINPg%*!^;vM|z_(f61PQp&l zOQQAFl>ss};#;7Zgtb~~s*tMkdP$>EW2%bE5+`uB!g990o**T%(iTo-<`I4v%-%Pp z_~c>6!nM0lZZ3Mk`7&vsc(Z^I7lI8M|=M_-RI_x5NFivAk&c6r?1DK_X1g;;9p7ICH=Urxn6R z`q6HO+_mN@)Q7`Sg=gT5jN4K0B^k}PqdXWZ>}ZGqt~1zlC`j*|R!H{jRLx)ROm-)S z+gv|IUAj=Pl}Uu)#0z}aE)Q2(0apM1VSUgB?ld8b$Kr}waxcqI6Xn+rbOHjfmswrt zF+n6r#CEU!nS02*Z{!PFt;SYf!x_MiOtc}pw9kTGti|o69f2$5N%L*Rhy(!uI}fR} z?3)yQm2-h@$M&W|IK`}AOdsi!RD8|Kl$0$ED>j9HP|7Q+`xgpNkZV@DQ<_biFG5F) zkHX6(;-uUZYca%Ohw>$Ly-{Iuo!^Sgc=78AIDIC|0VJUWi7 zNqxXerK_54ETP9Xn2GnJu6rC7Ti|@AZC|^ST*0romkv)@rk%PmfcqjEF?hYzilUb7 z56G3FYOv+N7OIT6I@B=EU6d>(lpaL4%m1YBoRnDDyUhb4-At49t zeh$k`U7?_Xq`}Nsx~Reoxc83A=k(Bj*wWS)RW==m)5BDcs^e|5qKZe`(RrTV*%4C|_mI+CuihX0?FfkHOjx0`DJU60p&}lw zV2dsOU#9niu?Z2rli(e-iVnu_S=;aEQd>-cSbX)=(($B!v7!W%Vzn1hQ4?|efcsW5HAhO z+i~Gy7>m#28z^b}$WrOS?N%Xt7#jooR2D;bXVq#EDf(53qiFXBI1Or2qA2<7-kp#9sj*BIeP*KsA0Q#+7zfT;|Ms%~c+F5+= zWrQcxS2)l^X~`C!NF5n-8dIG#L4CDi1{uFE$RS-?beX*Ja71s8eokxeBqV~26Ts^(z5pFl<&LMc)V+Z zgOgBj+uvz$ZW@`VN1*z0i*O5G9bn=Cp6=*5eS*y6_ouM;Mxr^#^|C%i(#V0Swf>p{ zddt*yL6pMZ=!I=|+vW010ey@=_D)$4>o7_E*}7D9VBYkivZ>ttvK$({Wlq*9FW?5i-d@4E97(4h3#) zx8N;I_YzYvaY!`LiCc%mQ3}$sIE0zDLp5gmzY=12=&@d~)7|P^)&AmJ6#g&1#ZBtk zPg?uma4-@j|KB(m2oWrS$bN(L*Oemu4d*l+rLKoN>MaU5$@AI0sh5kNn~0lED*^Ig zP=E+MgwrU8#5wZu-DWoY%J}i>Jif=xXz_!vks|}B3$yQxg+O|WJgdT#uGd#-y|Soh z#-UQl&(xcV`Z`Lyr#+r_il;)y(T}KCWC~wvHaakgL=MP8hQrt@0V_&{C$qxlF=*AcS;+YyizyKeQZIVVK_rT8DnnO(s|Bi$ zl%>3vhx4D*cdT#O6tRw1tLAyWb$6eUOF_F2>ohEbC!cN)TVH9|iN8kXaq3>O9NV^U zI(@2o0GUPXpbuI|x6p6h1krv0d-gi_{)rfBvkaxJKYPcPbWZD+rNro1`A_x>1a|s= zuwPcGnVS6|wi&q3XmhD3`N|S z5;CgWsB)JGmSo?;Y#h9Q&xI2yT1-SC9m}tZD5mD}I4VigN0E`;A;8?QhHA*a^;DEg z8trVUruf}S!aZVk-_ar|;R-v04Y4rs(wnEF@iOM5*`)&b;qUcL8-SxQmzWjD^qACAHyD2^RfsEiMhqb^LWg2oT;%PcT6dm%AIP6x3 zYg}PWpuj=D_{h64Q-tgf&AVUjTd}TOP0Y@GIR2DqcsCFJwM63I-4Y!e%~ynhF*r0% zjTP-ijeU0*QtgrBI|Q!t_|E?K3(?;Cm#%yBB1hXb%=o}&oY5-kb<97(B|~!}lYR-= z?CT3#{7%cXA8n@BxW0rRXtO7hmdClZJ<+Z_Y-b50pJ%AXKD!pjA$Az2j7V>4Hw345 ziMfZ9u@DArv(n45?e7~})E?ZQA_&B}p^mw%Pe4DlBROPFIz%=36LvhWB#;6Mas1db ztl$d`f~{8T-QcN7p3mCO1c9ebW@T|p_HPC+%%iUL;jydGRM+<&MKfR(!gmmu5 zDi`$u#8XfJBUFqNmk`BD8tQkWS)fZEGYz-#od$4bM`Q07BB9^o+ap;gzGF5Sr@28| ze5;7dGMTLRZp-Fv*a48=s*jK&3G2DE`Dy-sR_-%iZUYtGZC3N@u7AmGhYqP-9_#IO z2-*k8l)R8Qd84`kALtXG8auNeji!bpEzB}j#^ReCvi#21O&iA)8iB-AL?4pBBoUF}8!SKNSu{B;~ zUd%hz?thY7oaO!#xh37p=X2-b$N1Rs8bG=I%GXq!)>om z{;D6QTmR$zr0310VVU3h;IP-X+~+a3TT?TLQZ_V(ENX5?tf91ewCNb0B*Tx%F&v(z zV-T(I$DQCe7s^1G<7BbH1E|@tK*vw*3Thq$?PsKgsQ#|Eyir0qHbOli>tGhjmguWT zuIt@bw=k2<($$b)VEaZwEGwV=ClGFhQmzAc>){OT;2PA>(304nhv%k0G+kc}!smhw z?(WXx5r%mGcNPo6Qr6dLlBfG+GSTELT^{|H?Cj#3hj7RwqXx9f1zGl#C|xRyW~kEM zCO}v$UfKrJvPdBcDHQ`PzmJ$yx4bA96? zN?Y&4nvGck8cs)9_QSCRpZDJ$>;4pbX~U@kPjc6FPc}&htK4E6?R+1SI*QEU=8zsc zT6!gIjjKDv0bp>5cEUv1#LOi0q@9t$BudBwjg94(t+y)FH=DA!5og5L*Q>xafJ`T7NMGl-8Yj8nF@P&m&Z|f2QX-fOKVks)6PntQ> z0Iwi;gYRc(L?dzMd}%OEi{XOj8Ho^U#g#qlB;~}zO_l340i??fF1rm6F;?ipEdtFq zFrRfvsKwYA{gn`aouGMuTZGOR*V6kc{;MPR`*t_IqLo<&2)zRpTBrB*k5K$_ER+_r z)dn+I#E-$yP_f-6v?|M0ZTpo0lA8AGQYqW=EMJe^2dotZ`M@4Es$xsRvWk9cjls5s zsgkRI0AVO*!(o*CB5AMe$3}4q0u%~`lfUB`hkHiEC#UFwg=l6inlVQ_!_XYf7mSbz zUbrr<;zj}KPx&C)04t)O4l3iVJwuJI+f)jWT~%N z!Kn^L$Rd7q9TqW;j=q~;05Mxoa1s2mE-io{W=kyH%czR%p=hl<)lu@Wvk7hR{SVaIYUe62fCn^vCW{!a3vE5D~ zrKUm?t}BxDxn<;>h!U|s)N1ZeyPU06V|=(QIyEze^TxU8xE?jlihx7kCp4ilCW$5& zEd9$h$iBlZ)@h?}xvF8GoSCS}Z&M*fr$JoH#|uZ}5jEp?D^^@2dOmt0NdgGy#glRtT| zuXmeJ(v>eg6>xlYhcmoYYmiair3rzrCEmJr;9({N%x@i*lThwK!{E-|On-4Nahd*v zG~7PBnN?v6zb0)7Qu6+aEk7Dz3>fzQ_O~E1Q9Y;h?ce+c6;~PsTPI!a_LUFmscskK zuy?=7RTWeAD;?BH%nl_-E3IBg+;l>E6NVXsA;kfyH6_I zPb$p06nW3^RaBBmh@>bBN*;4+6>C;mK&mzh-Lw9Ih(vUPBpNbh z<9E57Le=!o*1Im8Zgv|)Qc-0&OwCKh4>~(y{V1Y@x$z1_$2jhn?D?eg>{GVQpT4`3 z4;BT}Ulm{O94RY~RAMXZndn`gzge!HBnWddriKq#}8sfGQhoBTMsyKR*al4NKr1Ok`U(!~P~du72* z{#lL45si$uuqs1S!#Rcu4#p-=gRKkQ!cwWxn|u)a@*WBkwb#tM%d^0$EZ+g(Xt#P* z^!WYJZx6UC;M)rez((S8kFNc(UMpbPNL*6=Wi1?+o0QZ>bl}W8HV_(H@mkk$&3+U} zvvHrOSJh`!poIr!?%J-bqf_Ei_le(%Ql!jP@X-)~r{z>fZnDM^b;`Bf`{B51zxrL% z2_`asFJ`5F-))Dv>;gS$wRrY@~_<*mG(KbzJ-i7KJ$mu!;@*3*vH}d zI9hoqanz>pGaC+`#jo3$Ot?(CmZoAAAg`#1xQP0R)Z)Ns+X9u6iIYr(9Or||sgBj& zmrS}WmXc49uMB-UWX=gEZ}c^+k-bg{;$$~1@YY^~fY!?9Q54Uvs{xREVs7q*_68`0 z$>KNXvGwiU@3nt~3ML~V;d&5F{?7?_L9=Wyo8sGsx;!4+Oco`-4G}FZe9ni@+x@cf zNSrNbt_NDnPETk19niOK1SV9e*VfdO$3{OjiyRgTFs{`Z{sis1KLoPYKB@Nvppc4n z-9GSy#oXOTzIQ_Q;&RyO^!mLazN8XT&k#f6SW7MOfBdkRSfmuwWj?fW*NC%{jJbYl zSltZ#g0uN!jh0qz%el3@65X<*b;GrlkD+JOKm^jn9{7FZiR3T>W7Y(cm%#m9n-Si# zinf`W)w2ot`shJdiT8Jiw?gvd1~u$ZR#VlFOSg8ZTn{|_bShnL@y8QgOBK?ravYrw zdq(a`cDLhQHI$}Rhq0BHu5n!iQ!~Y|(8aFx^5nFAOn?@{U9KeyLp8I>(n+F1s#Fw2 z9BN<&$L3LGMN8#SsK z9Cs(Eb(3p!TFh9oh02js8y$z=1-Fu+8dXkD?Rw}4Cl^@BUC)nE)$GSs*;i+T%>u77 z`os0&=Z7BhTF=(FisgG1;?|dkg@~;>MxCtkdL*tmzalLf(_QJ$Hjbc8Tt!Pn2&9ThY zWGVl(f`Noaj2?ylF5Dgm?Obdq)(u91ti6<1{v3VQY#zJGJDg-_c&G}xsd#*PoaONg z$9O}vxHL>*L;fcoAraBzFHkXs+75PBPRV?aq;oz)Axt^q?zQY}>^Rw+%^2qs!kL&% zP^)wWoRm`>TG4KH_i(Mx2$zW!rnK2`1E4Dx#R5ljSmDjkZTHB+jX0RyA(ygtXkGXQ^goCdz%OJMt#dFE~U65 zGqA3XtE`Qwu8ncxQRHYmPdm>kf`pt`QkrSm!p!Tktvd{*g=B9puW?Wwi(C$|m^k8@ zU?|{$FA|eRLV!oZ#vWDbzQ2)9T|4nGSFRpcOzW=NRTa_VdlQcpI=6vHZ_?oQ>Xnr` zP1*D|kva(8-z+P;%S~VN_ECzP6^cY9-r-}0E`Gq<0lMA$Y#uN|RL^^vKH%|pjgKl; zkczuwGtu}Cj!p@O!F4MJhujk~SFg_)HbWI@H-@sU`$WlEgtBKL$i5KIG?^`qgbr^w zw9imrE*;p@={|KR;DqL|0~8~#L)TeX??HTk@vZ^W93?MSmnl4XeSB2fc z6ARHL{g9i`x{D4PX;%Se$K<7Nja#QJ>hg3q!-N5bY{OcRNHnbrU5`t4qAIa*R{LhH#meExKW91e6U`kO;#gjEmK*B8Kb!mzILuHFl^m>| ztQX1HFu!4}BTM%o?)+jCgVeC82jXfg2ZTiCJKV$`Ex37wm9jo&{)9+4Dzz@$1z<}C z&7(oPOA=(~GM8Hm^{U)g8lmvLn3p0-h)=rc_elr2CeWoSBrHJgOtuRA(XpgyrVf#~ zb(*{Y=5B~QX`bpaEv)!L%>hwyRXG`xo>XkBc43{73r+!heDri z_61k3cKM1I1m|HuUEC!V@Ofx%7Zb{ia|>qQS6SNv#?>9nP#d2_yQS?v@N?MiA>$rI zxNq!(#rgL^6wOdx zxb~K}fm&avQWY*&hlx^qZbNGXW8H#9xd!fj=kER%Fj`N7HetoWdxx!S(V#_C`T?#5 z^9^%Mr?}S{)Tu`R6I|IK0}SyTU>@fCQ|75Zy>+|C^{)G-CnXOhZVn6x&h}&>eboK9 z^%Lpk{@nfLytOs^vrrzqU+X7hWLShaIQu8nal>GdJl_XR71zu8*6B0NMp{~0-}(hb zQt`1QN`gwplju-B@nG`&0Yw636MyUVTT;LzJJh!G6cj_oE)C-gkJe7oKY= z8h2{l3&cjJ2(;x-cp|N{U8P!Vasad#CmFLin1p8P$>0)F=Ma)YtB~*Z4Dcn8iwzk^ zH+r_`h`+gK@G@)L?A8;Ze{L)wOmVv(9GJI8oFp!%Ze%QULc_s&Dd&(Diho>urdU2f z?RpW4Y95+P5nSG?JDyq>6&*@i)=6Qvp~7LOzb@?v?+<#BOV_Z16C4zeGBv*ZdiXUW zhM~ElU+ntqO}_GY{PBUFQO$pd@bEBHjWZ@)C`r9lPt$u&Jo4>S!1w+7=z^Nqus*~p zXe~Av>}MweR&7aF2bXUc;(8vN#B0;j6QQ)2?>(9fTkRtx7L>9-s}w#&dQ=u zW5Pai|EcHopl+PcF-^JWEz-2xN!~%c&Z0KlqLH<(P@^27nqiImQL~Z#CKZPkjv`)! zvt|eStYI8T|~m~ma;=hsoB8?{Mp>uAx&}CB$cJPr(Gp^-BH3-XmXs!V^l2E zW98`4=@e3q1U;3u51Kemn`a2}_O9qD48FVa71+miIVBeB>^qg1Ek`?OjMXWLY~8~M z-Pgl}b}uKGJYR;x$0XyoNV@*YaYlOg=t*x-;74fm0rcq#X#5H9raTK z!yboG!W4?k;6F@+}%tGp1Ll`sVf(~E`oT{m zfeojY5dwBBE$nl?2b(9`7gPo7VMz;2>bZg_r^s&t>dDHd#3a|JG51i)MS4|M5tH5! z=ZlqsGO&QDA-F+?5((l0XnFlQu$U?0(UP6hPd6DS4>$_plD@|ITvf-;GY&f^)kS;C zVN;#z#fZZMj4LgjHZN`_94zu#%(sz{vY{fK=;yQf)7b~Mk$~mrOFUta#Lu5)an?13*lbYKwl1 z4V=oSWCgPM#|MLD(o8c}b(~4rO_R%*Zgyo$&g}pQ1}Ufj60-7)Dqm_;YJO_pQP;=F zWI@ggxYkQW<(%1MF1r&?mRkEc_l777dhYxWQ3WX(dBTTLql8uHiik=>N+abxmpexL(XVA>r68@a)aKpAEtb4Px?z4WjZZHMrTh0=-L^| z3T;o6h6o_xquULabEeA?F}+ezqIFMl%3X+S{TXTT3dKW7K)34wYp?6dhG}P#t9nFva zN)dk}3%*!5`x+=dnl4j>(P!frgUjV#AVC=F=X~S0sTa)Iv!0X8=JF8Z?nPH&wdz1$ z>v&y}BmdGnGV~*$aJJ{8^YXU$NsJROXCyAKcb_pfeAKGhrcKOpQ(KR@MET$x&BwCG zL%RlNo;8BN+(g%5k=&J7R^k?h)|&EicC5&*O~YNRurBiC43(`En}H*|E6#YnNi)1RvaNX?BPo#%Pz zIt3G8H=}!EEZajh!vvWgWx|lyE+GpV_i{NM)Lo4j)QuZiBQ2bYX*_S3;8DD z?v9D8!AS#9eA_~B^zItYTk?%PbY*Y#h)y5^KjSRN_KkU6@ zbY)Ms|J$98)k#MkcWm3XZQHhO+qUhbW81cECwKRA#xu_IKlj&n+?RLUao+7@*REBV zYptqT-}#xP9(#~k&u_QQ3%?RoDhyb;Y+EP>f5kY7iikh25+FxidrvVoye5X3rBhU{ z){|W*q%$b&3lQdUl9H+#Q7^bu*x&(`fAoOG4+Mth=EB9%P;z3yj3@4F zfSPCqnWYDI8{|I^QQ80&n0{r2?JLM=$nYU_6|&i7g0qRiCZ?(+=yXo#dohRUhN+gt z0^FNrWN)B=Nj z!k`){mGJxY@UwS(Ez^moy*iLja^5XNh8737DPiW4(5bM-&e*fI9Wsn{2Alg7U8eih zn)c&IuHl=w3(Hht9q$-?!CLR%zpm})ohEEGu76K`TnN zCpzDU-5(s8Eg;OtYr#KW%7tC?8yT-1hk_xgOi|$R6)H!(O|w^eB?f&@!U5;x_J^3W zvaL~mYwMQ+Rx|w8L*#waQa*W^Rk<%ftXS<1wg_0Lz()ZVeH4VS)=89ei0DKROsv!0 zKP#Hnn_841)6b87tmH2iI0}k9L3RzjCQ?HApBXejcnia$p;pM*&-X-b`FcTGKQDaE z853+^%UuJ{plK%oZMp|6s){e=#?6Z;me)DpVW;C!&h-!p>d8?Ywi5NHc5Iy(+ znDbb$^q zX)Ccu0!n%;)jX9ntU=+fu1pJa+&*LbcrfWK%^NCUT)Cl|6PRWJjK75U^I6R$$H+E^{A)R6` zT$V*9p^#o|))v0q80XrLY}SspN&#mP{|YJ1P3M!Sq#S{!^gQnHeGBPqtqi@N++ z%Zf4r2oc=zt9sm9>a5wZkD2V3qzzjM<}DG5xpRD!{7VgRnR^EbTB*TH;HS&y!2hO(fK9!jZM{0qI|%= zxdE;4215K(;y*rG+kXvYh`Q*h`JZ3>yA1afBy~AzeOn-z;6GLXD2Jiptuvn2W=;No z)co%X{%gpg%|Op9w?k$A?=ASRazZHDQb$5g+e?u+*$9?##?{0~wNq&cs$CZGB6Dg9 zq8aMgq-=;yV#u<|8(xE{{kxIYZM+;kxpQvK=WCaqB9GnC2NktHP8fgs_u~!cxYlVk zs%Z!Z4;tI4lw%F!uC5pD!3#ZeXB?79JS*(R_wBw5ao2nrl= z@J>VIZB}0TXu#1@)l!o{+QP|dFYe8$v)aXs31pfoE&Ue!E-uM9Fq<}9~@hQoprJO=*azPu$(RI z3c|!kX!)vB!rK`Z+=P?P;8~qPU0#qHy@GdwG@>fH$!24ZoxXNrx`8R|1-UR-+aEH!-!v8mWu8;WA zn`*H;)bbDa%YU`S6G2;f)@n{2?r(jWVg!OVYmC4<dN-?p1K= z@@2-zbf~{w_f;B1B z6JB|X@pNv->jReB^Yy2j`+TXwW#^-%t}bqG>R{te(_aRNxgX>WexZFoQ4Nr{B{c`z z?X2b{@C9O*DBW?zWE?LD7PHmu@nSDuKW%)0J@(sQY5_bBBN3wk_p)tyq27z(!+MR~ zAk?Spc4%4Els!*6%ikIfINx}aQN3l}_FTD>Q&CX?#D~Pm#$ClxDY=`}C<{iSwlKL} zp3IkB&jTF3fE>ZP)adDshpi5PEaEB%SXi8vYt3VwDM1vvtCOpYRvYbBo1Nj|;X1v6 z1x`|2=0Ki*ePXXval-{L{E?m4S3J*zQE52t8+^%NH=zBX^whsz_iAoiejGbC>b#o~ zJ5n_69e^3f)B1QEwWG>9eO1bTm08sF62nuPjvpX7Kv3O{@3D1Zzi`H%bT zKR!-rg7|XG1SKCl+3aS$yvo@f7W<4waefeY%CYOFqG;J)(@AmB>0iNxw;va#Nqf#V zfmODikoM8{^7c-FO%q{z-E4jPuKV^5b@_R!^?8z(-SN4eRCl?mu-W)nvGoFQV*S(H zZ0Q607S|uhFzL+=$p9Q>sHB;j)SrQ^t*wt6Uf<(9Ztt1$q|hUL!YEsbDLG*=;Q&X? zj+e{tX;qyn9}l+SgHnRwSUx@^vxU$Y%va@glP|p-Ud@2?k!%DGJwuM~ylTa}MB!dL z*F*7@oS+if+w(Q7z?CxpvB%y( z70BQZGpQyffVf;_lp;auHh{UXgMN#Ni80tW==@l7I$sH(N%LHbpQkF@r%(QZZD(t% z8VQ}N6SdUk^?6-23E){j12RRd+g);OdOWVjNpUV0r~!}=6h(irCvGq}@P}$cclS4) z5yQHc{hxl@ixR}d#8(~xoKtF&eua$JHGm=CAKF4~Ts>cJ#r{#oG2y@`mM0(}z-9ij z&0KV-|4*|AnA4-5KwSgCes*s7BCHRgE$bHt5Hqv0GV*rAI|NQ^4nns>TiDggmLt*%7h$qyRX*t&EK(fr zb9R%QgUN=kD>mwC|7=2xh)2o=3~rC@mr;)#%}3p(lCN4P6g1LP9`v z1G{!?FfI3hp*$SV9v-AwX|`EXZ2S!Q@WpaH%0WdLXMS*B8ExXAb=z9kTu3|>`De5L zN&$>d&sx&&{*7~Tq;}-}^)V0cK0hBnqlRz3;d?WM(mzj*H;@m~QMJza?^@gAf112M zzw&kjA|EKupg8=qF#gACjB$Tg7ZR4Ill<-Ro1Fo8MXxN|PW<&6gn$>YY0pT?U&hZH zXa$10yySly0hoaQ9f^ON+5e8j|Bl4}d?x*`P5l4PCcJstRO}`=zjow*XYTKVK={@P ztNuTy{ogN^dPL>x{m?{gDdS|FafBVy`#x(%w@n=0o4l4CfNqME>!$ft z>s+omh$bGJEhnchCe64@U18uT zl5^U=pqHiFm!$nEmCu^HSxApVPdy{0lNlh}cA?Q%t4t1>fjBXjt^(lnLz}_^_SyE$ zSPbO8x(KcYHOf(<%V=`DgE1Pi2e~WpQ421HN4E~jorXuHR)P%JB5a#d_R7S7kMvuk z)&Q3&lfy-kkpJwmn&#^pV>mmxU5rm_o?xnwO0}X8k=i?7S3zZD&984g8;&G}lA-`f z!ds&Vj%Y-*ZZvzfG~&|T=L~wm5#0Wq1}kMyGv4*;#er`si1nzMA{yWRiLg|NLWvdD zW(^oxLN#?ut76g?JkD2^)h*}mDE-ao+3i`AxA8`)kGI1SM(A9AOjHv@RkJSSq8S2} zTp(38YY%dEU2XVz_!zNMr}I1Pl?Qmadf$%%WQfv<(@qc*xz`snMxL90Te1OK&Dxq0 z20gbtkanQfwYqE|eR}>R_Lp^%^sirlWe%u+1{6Ra_@8dQ|K7i;QUUxS|69esmizy2 z#v;(0JOPTs^Q(yST~CP|`%?(+@p?}_`1|!|yMi2Pth9ZPfi%Wb8A2^CzQB>bj=h-S zTmPWq_ye=oT>BkihqYc_K(xNj6?mSIM(e`rhAn~5uT%foYu5+U_lsGxS&x!3n*GfkgewBMqw^9k}& z@S*--4x#x&pxZNKmg(Hq;Mj}mKuj@`&G(_-j)A{1U%!m4%YsX~GM|E4c6>a>)|65^ zVaF%ogFp=HNvCr!PJhqm9PrH^IL2~YQ}>(i7QZA-Mhw0qpb;0jwot-}&#i0V%UR@n+IEYr@y7=CiaS;>VT*HF} ztfZW71l1TkQ1ycLswrpnA6 z+w17)0EAa~g5gK0)A#s`2SZs&NJ$+}J7J7z=L2b0>^Eu@19O1vJ1VvMW63nL6RCy$ zZ&LP|!6p}*A08r}5B}G?{xoZp0&T#!glL`aEwDxwI@=%jF#E3&tWM_U>Y}bx!!sqC zeQX8PIUBQP36ED4>oRd!O-2zNq*MT+CP{&qv$;_a;ykgg!EotzA~vhpUNiedEB0bl zFpVM%7Hh6xK7$B+kYtY(BC|O6_>)JI$ztJJq1^-vn~avBf_u($FCkQJlf623B>Ol|jC8&?0A} zdDo0($ALZ>oY3_jYbG1$_jYaQqmXu;Rb;Dr3?w$!8M18;4kvl&?M^72ZZCNkN!mYE z)Me5%4oqWv)?#b+`4%W4Vl;5ae+!kj=LIs@n3&0iv)Pd~?$0O8wd(G&8r%;H*NZUE zs_wOBbImr=J}+sE6t2IX%NoZahGuv!saxlwuMly_o1Xm!3O-C^5M`dcq%fIE>5Yqv ztEHXAb7)X(naE#)pW+(x$M%!>mo&tWLc-R=Qw?xjx7o-NX(h>A?(&Fz@lmlnk;&k; z-4n3xYX4dRp-R))-29a3b{P^DaVw=P=QvzvIR0{1Cv_=%0qLwPN70X$PVnlv@_vB8 zYeF9}fe4zx=Wm8A!`|#wJ(acKsdzIinzyOqa=!Tg1%9?TTQI?E{8I4H`rF zPkRxW(JCBE^z<}uy7sb~o6&u$;Pgv>G~L@Xlp8@moxbpvl8!+gvvwm@JM{Cz#`FlO zkJ&0O=bLlEAve57g4iW?%V38M1C+qiaK;wVs-;!i7z-wG{c+(1OJ6C^vGQ$#@TAFbF7SgP0aiAtRcld+w3uD31kfhrxcLibmk z7ZzAW*QHN351q<(3VyFOD4l9uLfH?YOz{eExQrTJ_d_*)wD!c!D1AmNaXF=BR!^&T zVAX>*xpHQj&RSo(8$b`C)X9_HJEByYecSclsmMHaKVL6QJ(ptNuew4@+lho!;(zf? z70fO+ZCGo%n%n!US1L`W6X%R6QA7Sw1?p25idD`|11E>Er;b2GneCtN_M>Q|1G0)2 z#3xtBn?Q0xOcv0-n2I$YBI5BrFUJ_e`Buqtg!n>R!`E zXtJT&ME(6OuKeZEJ=$a|X?>GP8dPU#L9>~SUwbyquy zQ?IK`%j@eRAkw^W>gq{z-ZVj~q9@hGtr5xdN(#W;aMjm;K#x^BA=hkc-6?V?!yoLs zJrH4C{=To!!6q?0zC2TJ*dh%n(=wA!f0bi@I(tKMCAMcOZGLswvH!@K;pfb5%R{{8 zqYK;Jbu7Fz(PNHtVjm)-vw+`*c&unPh1a89E$E<_2W{F>r|pjOR^vH(akycJ%gSKN zzvKTkgADS~^V&hhvO@Xp?1aGUgp-(a=2u&|DUua~I+8h6hq>1Cjh4W)B76z-Pn%Yf z*2Tbxw0I4(1MW?e9w?PMtyssLDf**X*>z$BEgXO)c)U*=IAuef!NZ=3tcUru~IpHc(4w zBt*(r2{3s3_@P8*@>m_)nDx?&zey4b?9WHNl@&_|@9Dq06x}<{#k;rW)@Spx-<`FE zU1(^>rdpQ{`tjX+`c!2jXamGML)l{S>N%S~v?*_z^d*hm$?Vn_wY=V+M0V9EO~y|c zf9`OE{4apnY~`unKm_S`jOlo&p9TJ88V(Z5D2JHq!_9t#nk3DdnETRi&V=Xk}L)15S%iscP22=S=ym%BgW(OLCm7gJC);SPuE z?jkshV(rf_xyo%rqPhTC&JTo>^dA`Ak;7BY*E-ndlM7{8^d`m|mns`t_xiB#Hoa>3 zcP=}&&-y9o5zacao?f|@Q0dTvZR?02>019h4ZUVlmQx>RZDG+roJaBdLi z5WND{Bgu@08;xgC@+W545R9uyR2E^edi;J`y9nv`(ZjOX+&=%QmO}h~Lx&G_rMSR` z?M3rlhB4jx>KBlVWc=n`?W+NSYD4r{hMK=D9`plQmGTx%|AOf7@H&{s=NUG?ji-W* zq`qjfMqj<@K?Whj zN@=;3jwVg6UmHk-N)hLA&ZiyN3yUEzr;7#>r|5gvcZx+()Mi7MDB^t%eFUh}wWJ4r zWUnt-kL8;UIQ&GGZTqfqbK3E22k$tQ)u&7re7-^nL93rvMU^w2XOB23&rY0;I7Px| zV`U=Z3af38nY?~?vhyeJ8yj9jf5Mv1IbgC(Pbby&)6TAY(SKXsiwgk`(d>_9-`@6> zG&Nt>R}yba$|}{Kmq!LaZk9H&KVKKM00gvM(vr?KSf+v92;JCDnsF~Bj~d8b^i#A` zvlE)D^HQs{E?F@Bm^!k<{zEPGx2YW_EiZ&yu01szNL#3l=S#Nl99L6fpROjotuR3z zZhJ!IS0=lXeByOxu8YG(Qk=FxTv0r?;$<7|6Ez2XI7;^DD^l4!A|yaCDbG3#>5@T1 zgL%u*oPI>vD-0X3BtOnI`y<$ugk=N*K!zZ9TuY`?i^m($c>Nh{ejnYM*_RC6{;FJQ z8PvwfIq4|KE${QjB{_-HvCukV+fRG^AV9P0rE7(N0L0NP>*Suh+Fke{Jgw51B~~hi zNn~)c!k%}5nmiBWJwa$d)?>>_jEPNooTGWfHFNMJRPdk5Tt~L2jgc2`e-na@~>2hMqDl zL&)~) zs>;Xh!Yn;>V>&xZ+A-;El@wG$AJTa?cQ#o% zZPUvt%J``%^jZDb)hid}OIG(YB)ArIM>vPhm3J1RXbiOM{&*FEuw1K3VQ3d33LBRT zxl$}H?s|yrBP0A-%bJ6;>nhKC8I7ZHCU>SnKO4S|NvX7bUP9sVQ2-+hdP|iRjaoJM z6uH`$FE3xZVf`H|8O>CF%GkD8c9x0)cO!da8!;(yH>T+E1}**Yc>(hXb5H4yz+#yv z_J*VmbxfdfxYE+VWIT`3?ImRv9+}_Tx;GF8PcZ6A;&wlj%4jlELow1d1xzcYgEomm zZJ|xe%CFnH7lujuB}vo$xR#9Sb>FSV?6p#)NUCSBcBaKEzx4WCujqQ+7rXz2GE~1htD6+;7i_>-2B<|e zE5aQ&u`k zNkHdi^5AxVM}+ajPNTCf;Lww+&!+vn<7O~eFhJfn-?KQw6_0^7j_JsCuBwvBJ8H&H zQUw?8{RNU3286Qe`KEaGB$8|C;;u(-q=A*SU0>56x1zJAJhbG6l5y?=r~Wy)7KP@lj+ZvGkeB8os3s| zsgsNt-wJM)8d!(4i}(Q7C~&ZSoHuqLRIadVJdEQ7^B095Sm!uBZaF@YZ*Jd z3}?DsK^?8@VYePl%;^fFC{H|50q7nJ9zcAl^&Evanp0gKO6}9L2}_I{rKHwQJN6j4SH>MYr3)#N!o=3^*h!Z z?zubVwW6crF|bke64M2@r?!a~&O$DSN3+H-*h4MFGF8PLuCjfYGZ;q(O3}Vo!gJIK z8t5O(Jg#NW9s4yv?8YryzagW*tIN!d+=r@ieK9&Huqv;#4%OrkM+kg?P8uB-{Z`3F z)Do}#-ZE2W4%;SY)hMR(YIjSve&W(8l!TvAd^1q1B2P!`^y`o03Dlrx zPZU!Du%F7SO+OlHgf`pZW%78rE6VgV9gM{6ykDwy8Qmi>-5#`?WI<7A$Alf^`h1qE z?_sgs8v>|6u8ObBDqBD~fI4nQ$S8VKt=g|aJ$(<@@^*9K7tC8kU>abQfpt@ZBR1?N zIi5}eP&9u-!@rDiM5nE>JY#=81yL2c44ymvK{s#-Q3;N2YHF&fy&i{dn$z2D{r1^6 z*dGjbYg7m0g3kNC!*i&S35V{pGKDk&i6gKhDLZSec`JBq?~1rWE7+gC(MA>JgN~fk zQMj}ED1*mF;_z#wV}^b-$d(?w2sdcUzO6LWDN(nFe4TjeT9WmUB=XbmNv^& zysmZn_5u<$fuQaFC?Yk-uwkF7LNw`<*SY5tkIdMruZ}*dI=d-kC2GUD-P9__dI2(T z&f_Xt24T{IwrKaxt>^ZB@w>F|@akd<`Z+m$bAAO#!fE3^#cMlS8oh^mld(s~upjSx z@J+qn*@aJf#ZHk_fGmT_i!Z1~`s}JfgjZ!_#fHYSL4%i?Vpy!-qlSZsg$BV7$jH@f zio+#ay4h9(tl(oD(ncze?HV1wS0%<2C-Wt5&~m)CM~H}HRjPjHti1X+fokYH-_Oi9 zTmJ{h2YD^KQ$J=gwE0lJWsSK02&Ku3WfIB-Ic){gRe@A4nc0>Ot+W$ z^HtIeV^@_Q2m=mj9vi4N1A%Z+RoB;!*6mqGjZNuTi%Aiy5AM%#zjK)ECiaPQE#t^+ z5=>YmgwO_DJf$x!Gc3|bPy2F(K5So>awK3Hv?^CZj1*N;8Y{1H44W{MG_3UN@GP8! zmXQAlde6X*pX66S4{%rYt(t7QoBI0r%NGE_AZKbV0G1@`HzRSD-Sfu{&cipr1IuMP z6p2g_8e3*l`f~^`YO9RJ?qKUusjQQ?!Ddq`dYaGpM=!V^7RKyOS5c;`>15( zO;^cgQ33yk\&)KU$mqa}PR7PK%i$mK?l#V9Ivt>zQExoC-t6wdeX5Yq5Os!aFn`qM(?BEsb19EHj5fZRPyb!I~l%dg!Q zSGD}%O#KxJOSk_)941VJuixvkw|$Gl9bVbw?q}B9qT$^ryVJzb1P#uSJ=QgYRa)x# zuVVcgRU$2y0%|mMg6B<-yN5_BF64#+2-J+ju})D*P zbyR4cnwz-SI*#tYG33gi9K_ERg8DgqQ*ASxXJ}}CEOiVH`2M~dRFR;wjZ4VOjCg%q z$^nJ#QK;C+nh)1>)xb8rVo)En>4M)Iq=k%A$davPZ(C$sx8^M-4#&x8#wJEUPJdU) zHn^L3z#VflQ+sdaUSHSDqyG@nY0wR1-t0PbgB1U-SYy<9h&ZzQG{EN%!toIIoZ)jf zl@!(e_Qujb+7(^xkGJRXvV6J|%C0wC_kY1pnFuIAc%-1SRa>o6t-9Wk85ZD6(%D^V zt;9#WI)4(3n5gzlg(ki9{BFDKd>xo!qHNysWiVU#r5zfhq~Vb8D0VNU&Wi!*sjIy| zk}!_v=gB>e0Ls%S4PTjxWW|AFtC*AWNjI6bO-r;WmhmPRKBP|@!h^@3CBqv^xs>p^ z;FUxcA%I*UTO|<<@@PbSUVD)y#sASdlxc(djYvVPL=wV));b*6<$(Pt?v|v+SD4H6 zhpE?UsCOYED#ug0nC1In_3a0};Q`++3OgBZoj==I+vFy+5+sc2a=f+|PO9|f%=>){ zRo7bC!wUW9NgV>Tf-xM0@@On^#69s+Kxjm0-7v#*E&9Qm<#j+4C)d>s0q@BUY3Xw7 z1;P-zfLwK3nL#r&LQrsxM}R~yrrOBI$6A*dx`QMMFj7HBQuWq4kbggbbH@KL$y!Q`g(oyjLol`|!M8;)jjAQy%yE?dRL0 zpDU)SnE?9Hxv>?7mFryqgN$}vqs*Gi=ra|kd;d%jGvc9X_|_(u-lLj`L0hV|k)A!osLfHoa{4 zsF%7Xh0ccoud<9lN1IHmAczVH#grD|Zou9wAAh1x3t(!EKKAIFKIee&9_WltKmB6UIzqu87-%vorUg1(`P%9+`> zqK_(l9A4sDQP_IK9jap{B`S~z;9VA(fN%f@x?hfsZFsoij7}L>rPtrI1cf%9uV%Kh zn#uF$L4hfOx|-NTRaxIb|c@g!yhH@WCHJ-?&?) zjaml|;%#xLdn~6pCJwXcQIgI|Bq2KaAzsMFH7z5eueyvBv|PUD86O_N6{uD*mIg*Z zo}nHyIRS|h{CZT-ymZ-1ndd?-K6PD{FoHQj4pW?zjz-8S+k7+N?xIeWuF@jFbb6`Z zt9&=(hYp1tFtQKD(vodCiCAA<0FZ{)$grmx-Xa3<7@jg)b+m`Jy5ED6Gh+95YFFHl zt5%vxx`b9;{zGNKbXb-3esOgn@m@>DIDvGdu@HqWrUE`8pN?)NFyiGmJdtcS6u z?1j$sda86ay8TdM4~K-2`Ep^4aE{A~yVeB1#j% zs&|I-QCx{Q-E8*A_L+X@fq0(lokITp*%{yN%^ju%c0nI54DAepMiX{j2eFww9=Q}r zPZ{5wyZgwilk*AT%R$DZY{8tVPG*?Qm~tEnhlpv4mC#rvc<=43lyC75G26KY1MVxn zJK_{BHzl1HqFV{m#^=d(*4~tIQK+<-&QIhgSI38N_% zK?Y1{M^fs8A|z)qz-jFc*5&qSiC=rM-nLssFoc zJkl(Qs~md0DmXhw7mCCvkWs|;sJ49GgQ>4J1fQ*XKJjR~s}QT5Tbze-$tot(dY(qN zJZxo=|ury4g4aCcAQ}>CC3ZEyA5Gs!`FpMou`DwidY)S@kLA=#Nf-5 z4Uknk4GVz{LZT?LUAmdCKd9n}>~ybAjGdq%P$sGFz5Kcu=7DU;ae!C%DwsG%-4NvJ z;W+XXA|76^CGGmCuF-GOq<}=sr{Pftn(C<~-&W1;gxlnS<-$|+dS;=CdT-h(k~qzX z16y*?0bLHc;E(KF_0=gaDsjYAoiE3xx8;qaL`y>Nqi<+$ZRlp~dKtjtM;)V?LC1Fg zs%EbwD<>DmmUaVrJ?F(4QRiIA=7b+9Tc%!y(2bvmz5WtAxcfO-(LxBNy3W9V})WVA`RJw-;mdC{$vPQ;TDlZl2~dKOw- zQt=&zE>{T^1{rHhMNmj^tJJ#@w$?S@)osXTCx>dmgxl~jI$^p0Ril}LA51ru^9q*Q zCpsN#f0k-473xqD)dSCj2Rq51yDw}7?DuWHX zokQU1rdWW=<8{HDArDo;p8gJyQ1Sk96B}>($#8OmBWmq78G^JZFVs~;L?mhqYaAiF zk*%lX`1rV_q~s^deRv7B$BWEu69%))SH!#GvLTJv87%>e#$fvKWV%VT)iPDyZUDJj zFSX~`{^I|@RprL~C$7psaWnE^B>aNZ<#YgJ52Is#(qOM2PG+-Kr-K>^I)i>u&HS8S zfQ19ZSN`|0ef-|J?8*1L^K~(rBP5Z_(bQHAFfrp?X`Lxh*Y<7(4NH~x4zG8bCJ}!A z;LXpAn~iJ85fj|P5Hw$&&Kf2&Eyau*4 zF7ss>)pcf@8+yOJ8d8VSMqGly75iM=LvL#`*SP_hmH6G&P~Ucs+!}dsH^a~E;sEbt z&*&;@#kL%th=4_T;?vIcP<}rq(vDwG5p%cmO{=&BHs>vQHAr9xa@Lq%L{(7hYicKk zz|~@?Uo-Qzcnsy;mDH`#ZJQvSI`RjByPs;=)K_gE6c*UPV9*U_$bw1ocy!B&EmgB4 zn&yAt=ImWkf!R`9@|i7?;Ic!P!0Jd8eFF*T^?>+JH+wqIs*vMMj67ykwqWTnx|t*S z`-0=Ol>LMfn^V_cWEmTmsdPQh=}5_w2fh*Raw(bE&saLGM$tNsd?qgU9@lG;bNfmd znPh457Sk04u)hfVkSf%E6(&F{GNXH0$*wBoEZ}wl4@q-@qW}HtWycmry^maBbu z2IJSO(OX)bm58`%=law<=r?DwSbhsqTpv|)l5Q^a@of6 zVJnps3C&js`GS~m;<2$`8Aw|v?RIhn>J)zy`&JUqP9nMNf6YnL-z2M2;Gsn1o@!-KjZgGAsrjn7&T+*9EOi+CqtLM_(Nd2uj$ zGp#HtC%c$QiP!?Qa3LfedT3Rpn8#`8Tv%k31nlp8HQ=CQW%*Ti7$St1Xqb2evA#gU zH1tYgO4GT#okJ=uQkG|=DtASa<35;0ahRkN zzJyXZ5Q&je_@-=E92b86{$iN#5F9&c1N?0=HDs}{A^89t(l2t$x?||>SGav+W#0)oSUn*Kv|DT9)=HGi-KZ&xSdBs~VYIMz7mTJwk7ArNT zUcOFrB%fb+xG-+@Y0@r*TjdPaBMoo#l zEq2f9nqA#4WQ*u3E;|P_l3SdQA;4^*)04e;tnY|1Hxu4l`sI(K=th{C0oa!k#U=iM zEXqh~>AF;DlbEDM?=&)MyJWVOys~XEy z#b&u&rEV(OkhV6bS=&|1FA#R*ETP3|bEE0w67<2vi#*dxZh_qj!%4uDFc z?{_#V-9RK}7=U|ZHuW#=5xcUy(SKo&KEbd(skOhjKJA|_0hvx#Z&o-<(O;nGz$9tW zuIt*5POxs5Q?uOgV!W15yek%rWvkE-hxL=n%3e7?g-LBNF4^@a0HvwRD>aws_n$UP z8{T0CdTvi6^w%`yIBe#AU&yc5Jnj%G)Rz>#@NgBq3;`UR6W#x$m_qOnc8(hRTEZdZ zoHT)URPpVg%2(#MVO@x1hQD={%6&}^4U;-${cu3TGrE{kCuMw~PLYIJyl(q;V|JJc zFT(y;EtL;F_|~n4kTS~_A#7bUYtng1xfHVUB}|2rNb~rf4BWn?Z{ImRN>-amj=o#W zjx-d2EbG{cB&(khZ)(=tlI+Q~K0qDR!Nf`k)~a4y>>FSwE{)H>Jdt`7jk`Q}d7bfU z>pew(7&fSX@4?_vy_rb*IRd|cC@~Ci^o~IQR#AI5iYanckZ`~N@S8GHl6mZSBeoMk z3ujPg-xhNg15@OZ>(2ew?0t0-vLVge0Uy~))vxgn&8Gyn>U{4$6qPHHAmL90jzjb@G)iS+EC$V;`eIhDYHP79bQ%_Y-~fo`PJ^X?nt5ZH9o#a3++b~R z6%ukgMw%2b>c>gxxZM~y^LTb>B{qMpguEVci}-T9*cK#Gla>&PI&*(6rp|90HW#O+ zY6~PIo#FzxN0iw+klWER??Bb6Mh;R8ZJ9Nu{@WmthxYc^g8x2 z;9k_P+CYB#2^%}hJC9bW!s6k-0KD+IrEXcB9)^>t8Y(KsVFi(BET#YeG-%!SaZy$# zqsK~ZQ(}(u0U&JwfbjG1o_a|W^vpDl&0?rOs$%mb{B^B30r3JsZ6P~EL`LmxXY=Ez z!dV|amZ~U+@3^82nVfXqpta$+3G91dP`cf4V3C*U5S5F^HRiB~D=U)S@&o3Odul8I zSmWSW5x*zPhG+4b)C67%OCx7>ph&S708W*DqYz&gyg9DKi>YvHe748E4437urAjRA}1Iy60l);i0d!9PT>-4YIX@i;rz8}*I<=CbEU-E7EoZJc6$6 z?YL`p-UCdf*%(^lnu@ZsJ+)^Qy_qLBZ3|p8C#evd-mgnOJP=@5Ydk&RQ#FF25$Yzg z&`RaAqso6-_O@2wGiU7r!r1{Q`MVNi>6Likz#5db2?Y#l7f|*?bjDlRE{l_Dx9_Q3 z8&TX#RaO3b{^{xIVN|U@35c~k&K0!8Z_nx`Gyz(-4SQkaomer->v!mL^N$y|hI1?j z+PLE+Tw8!w`!H+M;ZWn}^UKVl2hF%yRm;Qsq+RUl&AEC>&2$A#$C^N^br&y<|?V@ zSQQ278v-rE{PhuJc{>oJ0)MwV@cPv(r8c(t+i}5v&Hv60{5vj1*e6qV8qqwuEB`6~ z4rg(lOcr#zm};$4h+`+baLh(-GR?&!t}{-9f0rA4!h(2cv{wb9h{$+1;;TAEYInzY zC&*zawmO-LQKs#(Z*a{al$^PNB+vFimOlf%#nRI7;h4!xj#uMQHbqNdx!7XoEwMrV1tZogn$V(_MwUp0VQ^bL$fqVWW3F7KwGX=(C zx$x9>8Dv4hGiWQnM@-z78YpWcIMT)J;=(FR^II7>^EAKH*P&-9ffwrO{K;1rH~5_= zODEZdsb!{(8bjFC~uQ_*z%Aa|Awbk-s>nyd6YV4LK62>Rfg4H%?c_>``9%|M&QU?TsR z=F(Uu>C0)s-rhI_V;c;R$U+AR)%!F!_ z4kqakUao14ay9>*sQue&^Hq0%Gxq15s#FPGfKhg8S$cS|<`34;)VJ`7Y;f@2AV9)DYX z%La>a1j+iYvDdy2p?EqR5(7M7`E(%34Y;+XG`1$cyX3M#=uzEs?`KA(e)94L{mTGV)hx0kH9vkl$ zXf?7kv!~;NVi`c;^Hrq(8{N{@SgRbPK|@m$JCU7v>Uo~fxz7kPS-sMyIysj-c#-Za zDIYpoR-x0Fd}PdCP9N%?cScWoq>8_!6?f@6RX&F!y+o`c^OMV1jh3TPPf)dIRjN>D z-yu_XjVN1sy9SU|y@^@g!e+ft!=`Gdyo8jB7WH06W}4NnX2Q+mQFsZ;v$Zp;*)R3L zt)QR7#_6yW6lKOenL^#dUE1ap>7L#1X(D8@j^n*2*GbD^W2xmZ6>`*Uio#h}*OzAq z^xoOY{>frcU{A{nJ!*h}AWjQ$WthM=k4Ew`?PVGVCqif-WxpOhB z5>+FL22(fVz4irD&-dX5>EdSEBC=~ik@rG4kynE_(t9<9yIooz^!}@iqo8uH#R5HI z4EJOhH#olx8sbL0G!dmS**2)|r*+EUk1I@iV>c+44m}AJ@lBVgX1IHo5CV4c7P`ph zR&Pujq66RG1p*~a^{eU$P0q|Cf#~Y3(sJhM_-jYez)=>|AAP^xI2)Ya5$8Dg;l|Nk zKF>11(h&t#2FhC93>^)P?6hxVk(Ie?@p-CfqlOn9qDo$}Nl9JaNpf`D*232X zV-bo66N^d$#V!+AXeLV3#Tm7}y19Zb=xEUsTyfTqsm^XS3KTL1X9nGTxLZwy2d^1>RWHMGR)dz`q+a zUj3fv@}!3*eL%rZHSv%UNM94;3TFr!c0e0;!4MgYVx&2A5OMx zTz`g>v{6pvNM;dBVG`X%m2!mru)NV@{#J;fgwGR)EVg_IR^@8dAlJ_Wq83*z!$l{S zK@F3uwfY+ZI4s0s=C$|utdEHFOb9nK_!I&P?8qPt-Y)~wahp4{)WJS2bYvK}&*0kG zZt+h&>nR@uJUG?W>@y4!P9jH}{7FcPTg-~zW@whd@ypY(*vGjjHCKG2?6^G7ST*U{ z-crfj#$M)mMtakI&MhPZ6-%}!Fg?QiAqy@~QYRM?7=HCxlc`4hiz*ud{@N}jfI zgRJAU&*Li6O0IgZnpw=ojJi{w7g@yQ`@e(xM5-Tn0YhL~q5!X$kN9nrUAx% z>zh6e)3f!5h5L%j4IKOh{udCxH~9bj18yBycp?1aEoTYPn7;6i$9A!8{;W zgT^2<=8vc$cHcy(`hbKl<4Cv&0MygHe~w^)W=c{+dtI*24c z&rd{D(l_bocU{r4ORHazTgv^>ro&qx&J}u>$9LGbJfVo2n|0F+s1KSw4PeoK`M&6Ee(Oa zGRBNM?R@|?VbUA8vL~_a16G?VKgq~|1bG~ipEIa`KIDp3#VgFVK_QpwU|21xS5IdJ z6^bb8^S{*V3&bKVPnQo7;slUj5%$q5@enm2aJFpc=L!%tk$i>d?4!(_>8o2~&yl!% zlVj@j&y>u>$P)#h>qTJIK8m)V#(O3a{J+?H$M;H~z1z1t9e2>_*tYGY)9Ki@ZQJPB zwmoCpwr$(VjC1C1J`41Er0At{&yn! zZ_y$M1PL1TN+JM~vMD`NQcvMk?!q~oj!naF zqAUiQdU4}$wP>xIW=qUQ7e|}^Wmt-&-B_g2ugy8OTFe~~SSb8Ly4qq+%w(=D5QUWk zm@|(`dj2}pdFld6oZ8Bzno1;x$V&WSjB&_TNC=@aF^!My4i}*oo7ck@GF}P%ptHzQ zUJ*eXyF&JcVh@lt`BWCtGLDzf zEAraL0U=oht=(_Sf!okw@zLq|Et4?QY@>cZf3D)>Lt}!G;zLlHOYjjn7r5-e-_u=( zm@gRTyyfthdk7Mjl_4cEV{s?}aIUAOAgTzS8rTrDGHorvkm14ze8CN1RlNMmk5`0! zTMEj^gb>J{mr+JFZ7skZ%Wc1)pRkt16+r6~>`*Bf3EaSXr{>3pMofo$<%pVrM}#F^ zso%k!N4!E@9{5KH3s%V&qf(MA7Krwc;7e>yv4VM5K3qR6yAqf|ZR{-@@y!SMbr0aDTJ4}7@x%YdHbY;asWlF_5&QS zkQNie;Q4f23859;>Z0LUT3VLM?P9H|xo+jMzK8~Vz44&1Txl8u6`7QkloFGZ*>$^p z1xyfLZtA#4>8DwMHacor$}G;~oZ~;Pc1QS(7g=uv0DTsKlI5;NJJT+K>*AuK1ILqr ziOh4Fjj;;Jx*{-KA>3UgxQ#hg@JxQBo#oXz3e4W3*hl>BpUmI?dx+S220 zKdfS+G;X5;;MEH01dYag5yk*QDBI}ntDxhrwM|*8stt6&B1TI7Gt@9wXenpXQwD|4 zB|%66>+qTLId$nBNs;m6`6eenFP*j(1K`oOxiqqrGFsF(wyD#q%rGE@#m*$`l$zf(4)D%yvitoaanI3IO3Zk zPw)P@0wk2FYsf`TO;mVWLwhYOaT3=!ebny(J2FLrL{i}2jKUfaoNg7j_9L5 z#IXiQlJYlQaL=pCA62*%6&E<2Eku%Iox)2O?<;|g9wH){@9`=rrs`X)x7h_$)c2dS z7#{7T+@D>sq@2PF7eiCS!Mdy+abF$-pU1*R6?p8>x^9}yM(M@@tZVOiQ5Pulh69Jx zYmHlKW_6;6_Yu<19j(m~ox=HZq%Wd#Jl35V(oiE|B-BiA2dU-5@kX286~kzhyk=^5 z?4Q7Nv?kLJE{DN!kT0%TBiedGMA{3AXwL@!$x7Wkt+Lj5#jh5V*|=P@fr+IzS;zVL zJAJ#ZlyB4`68l7Wup0nqwaNQ=CFz7UMP(Sbt#`w9-x<{#{)@dzTk(B`r)ZJsu-`_A zA!htAHhFyXk`rr^0}qeBZqH>Vs#OlqNpN3rmaItQ8Cl92(SpFBM~p200-xbJqZIbl znG7+!%N&sBt*!}fgHSJQq|t=J`#>@;H1-^CB<2WBt|gTT3m5B_MDMoL;jU+8IkzI% zy5crix-NBPv2*|R`6Rwd`Ta^N0672fc0$ zW%s)Po1+Td8noAv8<-yylIxg>f3*a6?Kxe#p1BkrnHR6W5n1toZ}OV%_Oc@K62r=A z9npNak6Db>o3NJjvz#$DO&%w5*NXhgG2?p_{F$$Z=neP6c>nJ#fHWv)pQhK$qm-tE zQfCrf@(-uc=HoEVLzHi$5i)lZOMn)?uB@t&#%p{Z6YRxW-|i+EC6yFO1M63zrr3Ph z8G)7i6COGNL;r#hD3@fF_Iup0bfbC!v=)im<1}uBwvML}@-SXHvS3#V!O`#D7;Pq- zZ|?cptyFQ4MiVx`O3!!RM$+q;((jho<33>yeh>1N78DeKpsejk_AS?=4VND;>ncYd=eW{C$PBWR z3M#E`bJ()UkBkfG`Kz8ZsqDZYM~S>F=g_1}#!U){i} zjoQ~|#ExlEVT381HU6qb{i`?|b8*Y_nWH5B8(E;hNod?Iv+1`;)&Qziw(SRH6vjmo zLozj5HM*?hqcnaF;UM+6Tb4w^h^SU_2&nz8wN7By73Ue|*&4ZBZ>knT(&e*dsMWK3 zv4rYd_!&KOnY9ASSX$oh%k4t3>Va}n$s<1BX3Dz7d(hRq7C^b{PJHl!(ykN&wB&^J zjo1JA5zmO?zR~D)RWZ%-yFKLYT9w)5&&{=)HA9-J0HH!$v+*(hCS2K=M!m8ETYTIf zI+2P5{0M71m;l&KN{#Wx(V`!y%=8mPj5ay(KPItWq5F}ax=Jhcf$1{L4>2jJ9Jbd_ zSjo@bWvJyZ-6D6{nNEA}nzwy2oD4+!;FnIf#oaP@b^JWNVJ>z1{kic|0aKe{g%T@i z3lg7?OLQ$QW-FZD@hipYL;4Jz(=6DJI5Zzu_Ce@$u;CBs8C1sE(@uM23UJsQA!pt~V&z^{nlpz<2FAX(_f=NQI*WH|9TOu%a+5)G zgFHr5KRyiuC8L|YXvXa1gQLhShR*&IzUvbc%g>Lc&zE_6{MR~+zHIgO; zj&_*P%UDfc<3e4BPmP7i>$|2xWX0(U{HiCmTG3bHprken^P~$y)DM}XPLBgE2XJN^ z1}>)u)CF>hlS=a6|22Ur$f&>J{AU6){b()JC~8A(Gy!s0yN$OP{Jv^1e=Zw|Cr?&M z*m66{w4#6jMfr6U;AA4`w>#7}Q3eh8Rxt$*~R;oH5PqRPYj-=Hx z9pU(1tRqup4XeCPRIWUfy}A>qo__o?^i3CZ9Z%fY@`fm`u^g3pqrN$Wo>U#t0I!-TcdWo!u6 z?$Sqa5IneJ2)xO<=#v3_wU}+us>N_8A-dhb>7R|0Efy0kZrMv?feVI^xz zlUJN;?GSjgiSx&HVlJ|{tc2z5gS15NYL>e-TZvd$@dE@!o(}P$BbTQx^L4#Dn{gWT zvhj-Au9``e=t40wRASm5MpPa>TN>>L7m^&U$n;UPy=Wh^BafmpWouoEi_FTL6ihad zl2;eV3{4&}Zf5w))U-~LD%OG`dlH!BsH7nj0*AlaWNy5UugjNMUb6?_AS5Dxr0bP4 zdq-SG<3)!u%49fcMXX|^w4xt;8xW-2=&+WWF((V1Gnu%72>Sdyen5KvI*yun#h2rj zG}-z5$Sl91T=?|{(R17MinT1W~Q&maL zOUX_@YGFYB{#(1nn6cE{{j2rcgV#-q8>_IwM@+8G`)P7{p8mn95^|B$QbS>%XIhzt z_VWl)W5G^{Nj2LI5f5enGu~-}Z9*zW_<_I9R&a7cP&duRs^*C#t#+67^NZ8CVe50} zt5)9Ec+)eKuj|fRl^y54hc&Jz8Gs5kmbB3!c{_FJDE2G!Ph4YR0^}1@gUQq(xG|mMS11!#0`;a9_$p&28 zDl|pyTBEH)S{;Ml>k|j#7E&Ese~HZ5PKz^y3#YQ-!*$Q_sXXKKbJUJIcD|O7#Me^^ z^PjE7Y-Y?UbIz?u1pm}ouanI{71p8l4vkn^`3+--l?V)0{mpi_X6}OS1!|VWYBvjW z(--@hWEuYFd;aX_BX%9HJ*QlWax3p>{Gl7W6^*EQjaEa-DDL!(M`3Q~Bw8)B_Wf&} zzo=JKS+Bk8L--z^9~yUdpZVTKUedFf2g2=y2eh{=N% z;B#|x{b$@Gk!@FhFrJEfR1Y?Tyb!pga{W8{B z3L@WQ9taMUWx8GBg4*DV!5hED;`=u|O z1KF-5KD&~p$t6|U9#&4J*-mHiVYj#x2XYr0xS|MqJPaqhJKR1j$E zs05c4DNJdXF(Rq?nxV8ufbzsWyM^V4xm4)?Xi?k zOEIaz%uwwV=7)VeHRmWaM!+K)Qce5wtq5!Qcr$>MBsX3`!D~84--}-h3~_UTp2B$~ zAe81tpx44HGW`?Cinq;}oJqc~jxe~)qiQYNG@+vrTpj0eL}7dwvo1Xozs{MdS7{%+ zu&^jzRszT=*Qh-nlh%eD;0Wa6u zyx(7TOyCjm-pp@Qm`x_jIRjH+5pDi3Tfs61c8p(F8lTT3wZ&=74FRjGp_qqAAGU+L<6KJK!+ zTF+EOyyri%*_>8E_*-&98j5?h8~$Y>2+1FCy*nh%<;IA|duhl&Z_}dTwmEN6xU2gK z;;mu1d3gi{Uzsx!WZf?+$R?Tw%JhNG zC;G~X1}DNcRtwtqBS@b1>Pa#VU9&US#}tKUL$jfju-{?nQurFG9X=-MyZY5sN$!Ux zt=9I2?Km*4)|w+OR0$8sqBuNu^ETS{V^A$u*_zqf(`N|yp0{(j(u*C+Q1a*X=9$rc zEiK-}=;hUM09cO^&>DUVPV|U3n-A_6ziu>!rG-I8Q!>)h6EX@RjccPCd}D;>5;u~P z+jiMtM2o-B-$ALI&wuMCEN(8xVIw*XX&*Hx~S_6K}E%cGy^j?@}DHCx5ax#Huo0NJ8}-Zw zl9Ryx2+3c%%kPxHD&;)s(K6{+w#+ak-tpiO9Z^!up%y+OU z0AUlzCXv|7`CGnt?}l_9fAEvwxE@yX8(VRWyrZW` zMA24MmE;O1^!$KC&uV-<(T&9w9CKcI;a}8Z4TXZb{=}iW-Hxxdp{@`OSeZ9LTaG4B ze!c*^oX?k$;EczLXpzaJyFWmQD4rWmOKv<_CfQzSb+%C$jCu9-EPRySlNp?-C@`R@73&cNVw@=yFTA- zWarv}Ssp`?*ui5pVKe?fv}Mc2DRdt^cfg_7pL`kl?8B;_OrQ5Not{iSFZGZ4-SN3A z>O}%P$B_oiQW%VQjk5|P9wZ6Kzao>!;7HOMK?Nug9*CwYOf zag}&PZz$t&X=;x1)(wOB%kj2)62dgia|qinJ9b2qDKv^cTEiYHTqj7W`-?? z)VUr88j0Ax#ht%oMgUsc_iLw!k&d@nOw>w4GQ;8Xl2LN56)bR8epWVmT5TmOfB%sP z7OA=(N}p=gp6YdLUUvh-5-}%YzZuNWQ}JJq&0~!nA5WQ*bsh$}25GgkQ`zoPo~ys0 zBtWN;r3&rAMPG}5FL%0RFI(T)(-=Ejw50Yr3Wpp|;V_bHGMuqa*inxn!f{LsC3=%> z)sp_BU~G(hx5DT4n)=d#I&NVAxe3BZWB2Hs(_3?1BJ(it>8XrZ8@oCU0%Y&D;Z>H? zPJd`6WH}#6`Q*4#uf1_9oK0og7YmBsK_->M$XIyrl<+&ieR~5Y;P%?4-}1scT)q%z zDFv_z8pthKIMi}D+FW~gKy@)!38S6MUYQgTg)ED)(>K}XDwyBw(Gq*AKEkJ*Vvvn-3(&7~Fo%ciSPC_lUjM-LST?7}Q6ZP%?wZ+#J#A^_17vzQS4=Gr zQmt5grf};IU#t7l)F~Z)x15BmC5cZ3{u>beMTIkr{j|M zbM^&28(rIF$CabbC(HNAE46YzQnKz@I_(a}P3ljkQQKFmpNbO_ys+%hK-_aEQV zKERI}BRJTMT|cJ3Qa-gdWxOBGG$F7s73)6O#~;$-C8n?Fr`o57a(jxF^h0Q#UGjg7zqthEXeI zPy4mQyGYrb?h#r)qEl_aBgi4`NLC+1c zP8S@Ug%#h20e8wG^Rb@$N5b8X^9A9Zu^_R+*uGx=s7; z!lc|%N`I=}-|vX+T9ZB)U%vbNJ1^@gFqdrEt5;#l=V3v68R7hK1>IQmfiVo1E+g$L zb|yBsR8|VlQy%rDyYn$AVr1)W4WkR+kUkVcm(u|H-QEJM7&a&7g+_!J7JFdi<=8I9 zT20|$_EeYsG^MK>9He+~3r{u%<$_tT1a3xyV$~Y^PMU_v`{~VzO9#XCDLKZ~Ilywg z^^iG+-Bem?rdhB0b1l1aS$^Qexdx>!}S)P-{Hq5 zajR-vCH+MdLKo{+y?KAt$!~N-=8l)*6+|u8^U#(2!*^>#mv2R4hBhBIxmw*BxhbLC z;tu(ZWrjCf_{N)pG<2lE@z9%ha)vBzCB^5Jqwo|cdGCB_dA?SKM@Xv@P|)R^+ldW+ z^ZV9k%SQ`(FwdKaacy;6oKv<>+Tk4rs+ZR(q3%hU9t83evd_(KE)24~SJPL6Yw~FP zs~>PFCJHJ#LeQg8WNnKtdh1(<%>w>Zm zV%nu3qjjcofj#&02}<-H=RR=sIlF3$$Cp8RW-%OqaB;pD*WBaXn%YkP+D^%nOMW5H zTIOjt9`#v4s9?f}nEG};wo!um_Smq=PNNfKpIq-!snULGGeJD8n`s1L+}s?xJWmZ} zWw6}b&Uo4cHjn2k;hn01HTy0b9_UeBAWrD%J1-KL>Gj)mmufkVjs_gu5KmU$iY-p? zk@5BZ&E8P3c7xFwn&(D~gUNBv-euRri-wAh1FN_HGMCTCo9P25hxPi}BY}GW(t3SO zz4uWr{ni^fwH7?Urnh{>?D^ZNE0ZO^7Syr84y-rqcMsqRgu%o~88>k32p_QCXyfsC z`;{0uL(>7iND2`Q3CWkCL)FR5Zf`a^W!`lKBEMrixCLFm-Y!8EZX-Y_aGXjRCHn6) z=0w=H&B%OKF2m%{x9kJ#u?8pWtB(PQfY!o6YS3T`^HnJN%{<;iy7hEb)d^PR9RcrZ zjvzgvwA2l`ithQkE0LQWgxjb_XfunkqSqX+Mu}b@ch(UGKmoD!Z|(06SR1$aIvx95 zjCgOCobk54jTVUw)K|z2ZRz$md&8e_?oEwdu2#hsb}KP-W_d9eyXkZuEy}!nY))k=axka|dqqgq~)*xAk7AAomVgH7jV)I&b z5V7JWudkhqmq>#~4N*En1r3=WIe3%BJ+W74KMl4>T^fo8s!+} z1cO3l(EO*{>h_02Q&3lQ8DpB zqRTie+v#*Z{|d8B2!?oi=3v%|ITc|_kI8woKcjb&HHa*tOZpEV^6O3~uPZ0z+kebq>>{d0-ZU>^cKkK`D|{myD9 zs?}=(s4kd5<9s<|6fL`lcAqJS@{tSe1>CL5H5+u1=(-OY>rhpI3rE+Qu0FT<{#dbU zEY^U4!Cf%(qmX_oZDvN=0_ONl_fxWdIP4AcSMkA=p->~2##$9$y|QB!LKh6#Dw7Ib#vjXi1F_CP$rCOWJE*IJz~UV6jnw~GN)i}eF&>AMP} z)b%x|p`jRp_8AqN{a&w+e?M^GDtCt0Zl|+<18bCDy?|T6#40*Y58?Ek%U^@DgJ(n5 z`Pz>ZRRo`JRX&DKotZ~GHW!*m_M=Po%v6am(~7bd(EJBAJlH+44Dt4kj3YwW$$+E0 zz}DGl9lF&{z5b9`Ig3*fx~^51Vbyo%F%ugDYwi)M%9j^nbe+RJW9~QZ1TR~f=iB(E z_axSUsg6eN+wpUc=mZY(xnYB78RkpLidBjMIqwB^2o9jT(g|ww80H zi#6x3cmtwavp#e#4zCv7$?siYl*Mku<=E5GhX(bWk?KEqsw4z-XQN#)zvDBqwypWL z+VD9mU!kQ!VJVNXhztRjwl)Lw3X4!a9(zf3E#z?{BiWvMaeF!D*!g9c$Jn)az>Eg+ zmcb)fBUh!I*A{h^rCxK83^sl8B=25=jYu8})Lp6O9j1l;{Zng2fGO+iOVDO#9Q&3>rfGc?t}smujD>pIwQ$smP_aQ)0JE-w|Vm3~R`r}5x> z#;8m)kxBJe+U{62=X5nGG%)i^!A|#mHmol>M0c!_;0r#6r%Om!p$2y6$=ev^%OOEj zfIyF;M6T%M3bg=|E!&9(XNz@0K{$C@!vn`IZ|R$@k5;FP)jKko%HM6pcSRM>o%KH} z7r#wnUNX3P<>@RX`ums0cJKwkjJ#Q{k}9N^cXH)#P!InOs+l`@2k5M5zP(r0i9~|X ztRm=X&S%G01mQ1Jph6ce%@lF6Cdjfi^ua;cR9N2PF?6|IawiY~Nm$a}-iP~VIC;pa z15S9zkWmjy3Y0o>*>?4KwuA`9`1~*0IFiQNzvl!aPg_fCBfnbY4hG`_fwMs(8MIf` z!olG-yD!7{S@67syzaK2EAnwY#J{~CjOTd_MEumo_7m42f$knCDs41;7?WYtK$2jQ z&U6YVZBY3ZY#?oE1W;{rU zDxS9vZmQO5_jCgs)&uWI=rrcZA{$5R6!yTCH9UQvm5}o7@)7Ug{W~)%cJe-15y~kbZBEslCTc4}xcKaK4-E=-&WzT6a<<*<*2N6= ziOur!ReaBFcf@zVGaOf1w!si4{52Qdn>R~jMi*_Pans4w<}y7UD=hc7D8kIn4k1l? zdT>ZTYUi4tS3W~s6K|%n6T4(T^?<9Zwr?}_#an$Oq>M`Lby_Ceu$ z{zx{U8rS)_P3CC!e7+4@4eC$8=i$Ql+MjtiTawA*c}~(EaMv#U!PXpC-IUS`Q zryWfBdB8Iw7XUg>-yD)cQpHrp(=M7GO}Eo}G0yM%v}ddEX~sTkU382_m31n)*sjT&o&k7ebO6m(-_Tn( zh2kZNzifoxI~Dc9ro|uw{Jt$bOhVypsqUMstep37S57v*=qCnYEMUsw=raGQ=qPvR z&B|N)980(`yG?leiq4_h8chDnv}CysMj@}JfWXk>5b7u| zx5=;F9;k-7%{Y705zJJ{ZP?9S;X;ruzzHu>1}j-^IOMe3MD_Js^&A{pj4~*Na|0}7 z@H24poqi^1REZ|Sb<-5!EhL4%+Xeb6BB-ms!b7rj>EWBW#DvX+kY0O(UuqxS7 z#M(3Urx-$o2OQbV3pa4;OVNGeXIK46&8Ae2#$}W>B#p~Ue5U0F>O!vf8=?`H3J;+E zHTXmt2`oUda$+L;y~SeK3Rfm^0ERa;C|I}(%VcNgU$7*B5=e(x{pgI|`2I@U=8SOS zAY6VP+zwHVim_8b(c<;uUJuceUzubv`NFnBNlz)D|MzlZpxHH@U_|dkDiku*$l^k5 zrZbO^7s`NdX!~@H^_7b7t`PMQ3LV@g9vsgBIV&xxvcqlXx0+dLsY$z-9E#lCbrm%| z6l_ugiKskALPtR2LRvw#1zH6px3cD(kFr#bsH8NFNa$gCS?el>4Fe@)qDRXpHh|Dd z@WjL)T#4EIm3+vM4A+$WpY~0T>6G>-9@M@^hn3A)=q#NtzR8T>YPfjb&XAjAIxQZB z^NFnDdELw9t6^7|h1Th^usd6~93+j__U%nul$jfkr}&Mq)am&{ZOZQjf?Ru4=Ss~W zINIeCNY|-e3h;=N-ay+YOa?c3f zNT>$?(N2%<-CphEQ(ohdVfC7Hk>){WF!gqkI!4zotJ?p%o|19{J;Ixp#Wv7C_<(iSw^Pok#T42`^m}Fr^&NgfXjM zf7b)$aq5DTC+13y3k>(ew~BUI^~6LTr)?*>3)Ns3N4ufNqWne;@FoWJ9x9D@WyC1U zPYnK4>>n8(-bZ3_J*nAnuP~zavr4g_61A}w9;(YTyfiA;B>QNe@@oq%R)yD}P67=O z4C!^WXP#UtAX$aZZAInOaa)Qw8PFGW&%C1rO~$t1!x@tO`z^og#`;LI-`#Y}tK~)4 zEmoC;NXg&j#93O^Pusw=&RI%r@_{Tw+wBb3deZDP~P%b7;II?f~`r=a%d&0cy%U=K$Fn31;GdWXl%Sqf~9!wvzvOm{sq z*8ujY@lmNls-f!RE>v*0LY{qu&B@EWoP3x|j zBHLGu)#&}q>4{`Fo-LQ3$$1VPfN%+ryj6|ktr7n_R%448W^Y~fMmBQMw?5BhDSJ+% zdfREiPA2~8enZ-0EdjsZCaa+IcB(nfsLtS+q6H_?*#@wfRyH!(2R&sI8yGK~ex7+$ zgK0dMsO9FX&36RS6$7bT)d4~K1^-Bb?2WE&LbqPx>eHAe$FXCQU&@A#hhrx=)U`zU zR4y7P1hzeY(Q()@qj}zjGt!S(KJTCQqPspnUWEpPRR0{$O3*i~SmJx%6MFN4RwTf6 z2bnvwWxsp3qpHh(goNp*f~S9`?h<${1biu`f^o)ed&3hBYV7Lv4O|8 zpGUp;=~0(o{!%e?HE}sN2S_t-yz;#<8^2ui+5gLyb?_atzW@rn)g%R`d$oW0SINRx zHG#7;L5-pNv{h3_D@hg}aY_67QU4ID8NV=6jrtuUc&YQp*Uv!0`;s|AJxKV$c9O^Q z{JOnF^(Pywd}}?6iFz#Co5b|2wLz@1E!W-7@=e}t5)9W}9vI#irz`#*GDGlNqVd1G zf&TG|R~0G`t!JT`9>6;6t7l-fUnYvsX7%$Ay34GwA}Ji0InHW0FBEYeb+EPvtm@)1 z=tiH5BYK|rbzbtpH1!%7>DsJ_kG|FieB_D!ow_ZaH%`+`!G1V2kMh2NucmoaP%IM+ zDVOAmicubMU1dMdP7RC9sz%@7y!|hJErLICS6+Bu%?&I#ogaF9FxLBUwc~cJc|tc# z_--cM^uZFM3{oJ-De&dn?g`SRwtQ@9@FqeXy}#$K`_p;%Xtr2k->s-=-d5YqS@CEr zJ$78B3IY)^`q^yKKDWP_y}t~QkR;G)K&H|=+^={|VFsw7^I~Lry4zeYDjvL@({y!r zTpJ@MyO;yLc`abs>u@b^_|Ztv9GM1(aMo?VFzMWGXI%EVFQqXYtdT*|78OV~+Z)5*@`7eYO7UbCMJ58m_b@vts>m=R4byj{ zc5NSTiSh?2pT~EZuv}wt?xtg3E69IN32WL;62@=jV?WQmisjAYCWAKDh6a8^Ig$QS zE~s>sc|5e54)68n34lasJvh`*uF!mF+Y$_W(Hk3{q@bhge5)PX5qN&Vu7YBn_)gz! z+v5+BcT*z$267Z4wZq+6e%SI3eup5#W5SAV77slKtxlUiJ74W4HSRWH&C}Q?TW<2B z`Cu!rnQz)du@`MHm|USf`Dx!O>`QIt;8*^WK78V4f11}}ysS~%R)9bgqtPF96 z#{Y!k5sr+NyCWv_f&9uLBaBxXQ~}N-$e4(eP9-p`{?MsnTeO_Kz;O*c-IjpZ(_5=H zi$)uf=D#K~?YFJm8}dcrA&Czp2pjs`IZ})(Ycqx4hs0Trj z3uo;3zagAX_O*SQfrBKQrpmaGhw#T1NP4f#epLQuyR0r`5uFK`3TI(s)NEG9Ny~4! z`Q(}`W3;+njep*Gmb}S~`GL%aQnG)YgQ8KM`fd4Zzg#Gk&qAw6QKWMeKT|%nyb2uy z(;%AzHD=xQN60uV8^ceeu1AYSv|qBXE*JQ}pLy0ur6c|xVz*~|Xvh0>St}>84xbtI zyr=sW{pi(yfmlZ;n5seZ?(IS9=-I_@(VYrvk1}jW=AAq4jaS|R?bm)#6AHzBf2$5( zuUJ80_?55wBg{4(6UHnnB$OO9yl?|@!r{Trzx8dC9q`qDlf&)Q-omn+*=4qIfX`6p zOu)}EQ+*#D)}TixGs-KnrYKuvnNUsgQto+XrfbCBa1L0TazfdibGV!QSYs~#{B`Jf zp;J?PzGjm;6whuvej&qiA|`4X1J}tgypp1#tbe!;KEC-C9^D~0n6M2=*xvXQvn#b9 zb_ct=jQxTi{8oF>%ru$$s%<2_yBe+)@hZWLy8?usf$*ZI?;L8eMb>((+Tb#UsauZ(ZVX|UgHsKeok*}aJ)3hC0v9gONA`|)xkJoj3a zXTy@$e6{Us093hJ7i9i$Zb{mNabSp2sCp6iKFEuN7SXte{3GHf-scnF?^$cra!<7l zS@8XSl|j03t9xhH_bZt(x>7mjti1wfl^$7n<@9ZE<14~YYQebn(ZqAdmeQK4id7UN zDH)mFjOtXxHSwK=gGJUo@tkwb8HLL(V^a8G7j>EWH)HVab>z*RR9Q~u+FJ*_k6<~g zGCgRmp714SzvN~;u(Bw+F=5=cZ3rWcL{$~0mVxm+D{8bRY zxBdbMsp%Wip^R*sR#qvb>P$#bq!U%w0``^N&tQK}r=aeev;(6(+r5A6Q`IZmQo+5m zaskQLKSknK!JijKXl-GlS5qaUi;5GUwV%y=KBcrf`huBkBEw}Qqz2=T^U4;OZPR&Q z-i}{_#F^+S*4&nObKJ4d5S6gt_2uPj2{6;ZF51RpzjHNSsXnZcW0H-xjyet`J}fAe z(U=$;>v~Vg3qV(r1R(d;rv}RTwEeWGUrA}CY4+|6qm6U;_)f2C$S=S0MUA-{&OIF2 z@gD8Y7EuGgBV%eOWh8m^; zOF?0=LG41o@tQ6Kp#XGWMp@sdYs)}1`j2us$J5NsX~}*pfyMH`Iv>se&V!&du6#pD zX|bI+Sz?ZViU`bb-)n;hgCF~3{$(TT(>WauTX{HGr#~!(KPPTO57x!mXwxd6jmf;Qzku4?Hs`#4)!4*C}j_o^40XM$a znR-c291?sP3e}=)1ohul((>bHqq=gFuX>p#kk+DQEF{>c{9?9sWHGmf}#k?R)2 z#2ZzO=2%Qg8%)KQ=-F<7k2*uniE@GGfpCW>UZC8JVai-p4ScMJ7~^?kgr~Se%Stto zOQjiJ@SClq%#!PTpYBYTT{FyYglVrDp!i|M5waOO6#sXHL`xN6jRLmC4Z1g!4D5FF{b953+|FZf>; z6ue^9W6zNCVPvTFe;i$ukvY!OL}d4g(0@?N&*Ib)&<^9IMFX09_&CRRKoJ$!V?7ZgQFJ(u5I^)b$p zh2`-dpd)!{mNUE*U10l@HJmN0XO$mcGTwwV9N#I$5`&K7QNAljJU^(0!n#N=HcsO} zs?3tf(zB@jgBW|)(>PwI_61cSMDkVIELyRX7-;ZLCWi__WAzR5bFdW2MkP^GavOx2 zGqs@R#Po!fgy_>ms7|Xv=8cFDcnM_^P|ej)W7lYd*8k#Y-FW{ z9E!M6AyixegFoW9FFuUo@u8e@A3_cJ$E!@}0&9IV)`XK|`k|i7cy2`aZ1c2WM7OlQ zffp7mF(Sz14prHJShKzQtVER^@9gblx2M*dyixct5>uk^nJE$w0zLY+(<$}m4!0%x zKCfM*~-rzFQBSd=2-Z`F7(ZMY8!0na8_zh==pyp9eR3lnUzp)nB%@Iqtx#|7rh zLOnYaf$!5%zF=e5OK$&HIkp~O8_YZ4)OX47PwUs7Lxzk<0oN5_8`BvA8Z6ZOoIk%L z`7Lc>@v^s}>q^RW6RRKD{TwvF$g5(rYG0>a)!Sc%TdM)bu3NxVs z2da(OA^w2>`Z9^>x<3+jZu-}+T$Eoe_F{rHxSiFhs>pvc?ZN4xHQ2HJOI%#MEr)!r zOgh`dkO5Q{@5JOBE)Onb_swp}7$cnD*3*TDX^==X*1zGBCF8;I&Zep}zxZmcGKB0h*xgD*%*`3N*7N95 zqAFi8z&rlUvtK#GE{4=!#Y%BQOQx~Fj1&LntIZ{Mi`?q0V^9t%i?Y+jHNt@#&duS# zl*I++{fZSXYsZDqpT5~-!Ue6a;>c6>cY(+q=Lu8)Z?v80C)XdPdr@gl%8GWsaR`d^ z$hh1=36debNk(=#0`Ua+otf-01x2M-`l(WZPhhC)R{(Suq-#-nY# zRJ}qxKx#NfO|rlon@DEA?|EG|6_Na6x;oF#({PPLXa`~BimBOL$~pKL=5PL)ce$lG zQ}f%nWpW@HA(L17a1n02Iu}W_!tI|xPeVcWG07fL3^Icc>%m^~qBHdZpT|QYX6RK% zWCl{M$>!Jmk{DDslqiP(N;R=nkL}4y3{Gd969a^ERs?REJ312nN2QG={8ejg*sM_U zABDC)4W!2v!XoHD|5HxL56piOK!UjcivJ%INB)sg08C8z%KyEwe_{eaZ%p3Mzfd-` z8|KHhBwDJ9tWgHQer76y3Rz~mRe@9J03_Qo^1M&f3_?blM*h8U-qyW3p1g@|P01Dz zN~|T~zO2U%_~X>s|C@Qo7}a)&DQ+faW|S3y@r~wUhWZ-G!Hl%$uw8kBa?+yl?l2 zQ~+F#i|s$ZSmon{n3(4K24aAi_L0!hxT-u|(X#L3(dnK%y?oYDhg?wg-%z9WkAzb# zcvUN@;*?`NXLJtbNTTlP+Wj^ms6;AO;doj^!?h!6dJ`2F3WoFC-&R!*{WUC3dT*v- zBpE3Dzx;%o2slMH{7X4R8)@al;HyOJ`=|w5f0dKda0Ldm7Z8pw#CEIjWO5{EE`b#m zLP7od+7mdYc?0A9CS<6nr7l*n=Wi<&7T7iP_v9!{<@&W{acI%tI~iMyvzRT-LPEW#>m9A{t$6Ga3Lx{n60^(YdUJF0+GX>F;^p1l64g0?RL5$w zeP!@wKDQR4~3|F#%8=vu{HO%av$C2=q{-=HA6gTf(*8-C( zQ6=r-2O}fow2TK3?A}1_?CyXnITJon<8-Xa;{r+{Znq8P9?gImw>&<0NT{V&P050~ z?53qNMVeg&4ED54@!BUsIY}>9b+qP}nwq0GetuEWPZQFLu zx%a)9b!R=zA27e1bkp`}d5w3squf_ofvJNq@81$Brz)ACt8+EUZA zbhE~%%dnn!wu2wLP>edr{)=~Kn|7zvYYUCz@n{NOB=4eEfQ2kdhr9drX~4s#4noH1 zXL!Hj3l*7MRxx_h*52b5T4F~ARi$JV>&13cpPIG~>-QQ5$FHw!Fxm;)HEKO5o{>4^Yn7@aqwG;bPn}ySE8Y^zICgjlfonzut|!`vJRI8& zs~U|p5&`$_DT)NYslY8RcDXKEloeo}tyr&?>|z`@ncDf>vGc96pUGl;*bBqIUM}2- z%btxqP*G9o@!4&$D7R=I&p?o)Z#zs_vQlLmnZ>fw>2$xFMHI@6o0VmUGS}sv)atT( zxPJ39t5|SXB}@c^k6O{v5X{R|qp$e3##6GwbRrrPaZj~ojq5a4b)A`WKiS>8W8(6K zV9fGdus_}Pwmdx$3PTZCaJZz6@4O>? ztkHd_^Ku^2R3~Ewe`b1u*$&|W4ZN#Da#;G`3GVKiAhOL%SvH(z;YTS}B_lAQ?8Umu zI;9AK-Kl~mueMW)H{&7F)y5fB5n@~vJJ|2KObsjkvU8=NUBk7jFA#$!Yq1;jQWi8F&?JDso zce~g3k*VE64h_+)qeN~JIrxzUyk55pn_a#-cg%i9&%^p3-TyCEdv(r$5tB2Nu41sv zj8J|DZ2yRGQN1E9v=DhLDk!tFd48%*r?nQ+4k%f} zz`!7~x!W$Utk`X}yPUB}?PRq*YR;_v0IgYnTL_roFC3wusf~YKM#UPp)oKimioRwe3Iiy&Pmm?VDcK+8!Jdb zv2*yR*}DkM9*?C0;c4or$hYKG*z|7uvHjD1Qe{?Umpce$=E|Yoe$qhNd~^W-!`#u= z^h}nu%P7!Tf&6PJ2(Kd}WFE5|bG#GgTlwr`WLiu}DlLcQA1P7}bnT>PVcMSiG8His=F>nMWa;uty))ful-^bN)H~0l0i(AxS|-*6_R!H zjf`_D*PgdOO7Ics5ugxp{~Xfct2Fy6;JyF+tZA@q&yezcCV>(Qu5<5i8b=y{LKv`v z!8|PwwpRYW_+!$bL@wQn1dfT9VLFf#H1wl0eOn=AQ0JSdG<1pFZWp7HJQT+l>cCFJ zXZIL8XoX73x>DythCE8N#IB{Jd|37A5?p3x zis?57W8XF;F#|c0mBJ|st>SN2xm7bEGBbn>-vI$_Wqton-Df@9+7#Db*&TKYyxzwPh z;4lY}UJJ9Q<;~n&X!Hyqt|Tla)*>6nL2DTf838sn)=^Tv18BxgXIOVhpYP!HHXh+Tt#hz zM5x^y!{s}~Zu?Kz(n_`F<61G>WQ~H+<20p=q0{n}y8n0s;af9@!@~4 z%klw0_KVu9B|&o*dLe4Ab%V}>BN4Lq^`)ad_Ny)o7l(TnyWE-X^r?zFR;npAr$`*$ zt2nxQZ+{XeChW6g`o6JS(64Hi#@|5Qwl?{;!tP`WKa97ZdPkngntiV_fKQ%cOj&x& zPtZzOnZ{hcT-?Tom9b~0uX)XnoPI_s#?LDwk)5jOQVJlXl2-o<7hd{k-FxqU#)U`p zyv<3IQ*xQ0=F=EV`Ga!@yIkTk2zj*Y9rgYzhRaU3M|Z*L_dX~*&nb&mJJ#gK2mI&p zx7S*?4~y3e$qS8j5(ANNY6YHjWvY2idA8UKrq3OFM_)FD$vvZ)@7=DI`3aj(hwta( zgQZMa=hLdPQnbb47ZU~DxPc`&^YQP&%3{>`PQ z3(}Qe6>5KpfFEo$XlQFoMYgv=#Bo>WHW3$5>|Ec`f z)$Tb2U)bhQRGxn}A)?3ob-3{3eVn`F{S>qE%ZY#E6OS+U;Gn(*r?}O5OcBS8mE$VL z^^!)bVl0-=8k%}rh5IF>EspQ`^|R&p;G6FgWtEzq((lLb1|Xz7qU(Ei^klW%^4c`C za(;d;016?4!M7JicmmRKUQubkp21?JnI*0FMCMO9mNMH)%? z^#q6h^5V%;RRu!EpWA1|_==7XFE`k!WGdl_YW|hhGQxexf4dzaEgU7)mlP8}r|`1>{oknf?W^t8;pYzY z0j?3`Cv7XLsqY#syyX;(w>Z&`@H_f~uHGM{WbYUbee6z}20^z!mpTyMC@SB6O3kJV zFlU7EvNymYo~ZJ@f^|WquWB3%}NL&BK*Sg3h~<#6cPl4KFIZB zw#}*OKW?ss=m?1Y)7Kz3J}C@bTo2=J+nq1IJ1^8{u1IIbL{8+>zdO8BIP3l;N5E(l zy8~^m8V@G?A{JB>tWs?}1tZ4c+6VPjqHd%rvUkdRH^FC=-5~Y7AqfnkxROo^k?0*6 zAH>3qq}8h?8DudWU#fW36;)MK{O*zteBcU9@adL}J381*4|3$R=;r~S%`h1B{&l)_JA_3I=H z5I60@lPs6`diPpfSCuu`dmXI8LBfm?kuB^NAdF=C1vc1t;@B=u8u;_~vmbAYDoVhj z?iQ|Nycjbb7|hbjGKuIUD(BFRz#B9^6Lqo^G?wPhLS;2mdzdHpC%n(%wF!Ckk4@0} zq5t1J>#;3kehtQ7<$^ ze}#z#s1O#_<=2&zm~Os0Fu?z_t9sD^;!6_>gA1WD*?77Q0(g_p&mD+LnQ10b_E9o8@bk8Io?mk-s|0x5Fi22TVC3zIWJW}I19@PXt5y$lVKcaRkG zwVCr4D{>qbr=rGqsz_=w(~ohx?^Y4CawCOQM}^00JquRmZ>w}#@o?P2q@uCBM%GnI z!FjBU)N&BJ_kI2~O8>aV1a2vX%<6}nGXJG)I9_2zK)z>DOg_X!MDjK(6Z6|eeo5Jk zr1h#~eRv%$jEahvhlxs6pGB4tb}zz2aqkYDX!YjpI(4fDhbTRJtcaTYn)-Y8BWHBW zV5a4rNO_oQA`g9sx~Bp2Zsu3%f+A|og3E*?-KcDcBHp7hsU0_AIs<~r;-9>*6JxP- zo0@{crRX*?Mh;R!UezIX)WN<25yH-o5Q(YZiU80ob-2w`g~{MD(i#g}sh9z1#IG&1 z`cW@oSH0w-^Ob*g__f2T0&~`nyEao-JR}`#qBiLC&5jA{RdF*=ZQF?{e?6QlMg7V) zj~a>PV-gzAFRa2<933;_H`k^{E3xSq!ry(o6Gds=Ref)P88L}HpDsBL>jyfKqR9jE zCm_-Xbq%NqvbA}i$7{yY@vS+6k%0?YlpgLY)x33DoKiQ=l z;NChAZ*YbkwAsQ8Q)LK?)GtHo#<}(9Qxfy%p8nUD}8 zbOwXUpaAy}^B0Tv2zjPpDpKAfB@%JOcHCSWps_0ItHQ>#` ztbMnuy{kCtj}1m3>v?@OfkME&O0Ktf?WENi*?TV~0sLu>(7-VL0Lo^2Ax^iueSk6s z;vAyW$+U=Y9z5S?El+iu`<>w?$V)ck9b+ELhT6JeeIBHJ#;z#7@2kP*f6xtxgyeIR z#i*6$mA2gve&d(E?SjI`;&33f5J%3<)2C!em)Na%Z}3(w;H{`G)o8WARb7nRsRq)9 z%uFrlipq_L0U08^biE%Bwu+zPuyABTBL8|5;gziP5mHYHnol9(M0362#RQEKOr2V7 z{I**CMPB+8uu~b2`nJwiWHe$MdgZtBQ$$YMzvI-j1R^3*6$aXT7z~G4X3PL0+fv!~ zo!9LwbQSt9yljRFhmOLiN;*1vT?-aDJ*`MmoIi;o7_L_+3@K~D+V<;Q3QYYpBh@TX zm>2U*M;f%Gq@uzEzmM#ty`0&~hA~*?F+MT0JC;YO?Iut)gKfD@i;@WqT}jz!RZ+X? zumkP&DwDm_3f4^3GjBIq7NtNqs2h%e>vCxbpNA_~oXHQvU0TJlKY776GW^Df@NRYE zHh%W$R(;Fd(;))NxzDIY$Ig-lVf2?U1Sf_%dFB!}pI{*o-NN_~JsW;GsVuSr zgT~i8Jf=soPIAOv3n&M~F&cn%T*C|QSbc0ivH9^Gi&q%$Dk#-a6yI??IevwLe^Ndd zsV^GRRBA5HdFKZC>#^?C>NPVrOd{if{x-_VtrUFKK0>->2Tz@q+=e}Ta}tJ$eU zbTbr5=csRPA*R8nRu7YJx0XnaLIMq25%9Tn@O-%){9~0$xfhF4Y`#`e1kI&5T-XDaz zwaDdiveJ@TO|LRks8was{_Guo%6=`k*zK?m&FA`kw@JCCD*embF%bn4G;;RF$}$Vk zG%U6;<~H$$8C*xF9dJI?uiZ~U!tt!PUC5Z(CaVEQD!<~;Cy`d7-)A_Tb&QaaN!Cv9 z=Y`EG8NBDuC*U9TatZ$fdqBCZSM=jZh|}HIV=QTVJm^3iC()VK3Dx{;d;O{p=~mD= zQTKU-4mFzEy4qiNu;&hg=FPOk?`U^YXR872)te%qrc)gw+q5wU>JohSuuCKMrQl)b z3@5>~)?~9HmCMP0;$MHhP3$`yogS^S*o zreI#J1vCBUOQvi?iZD5%4T*&X%|_?7xyjO#6TX8J6`R6DYvNoDD=vE#JC%%hS~k*v zM2{^dU5CfSbqO(xabIF?=jWl#IwUTBe}I*2BRXDXNKmFhtY!Fw%*v1{Kt%AyG)EhB zE5ecM+b?P^IKB6fK#uzey*M#>E1L1yIi@V^RlCNpmU_3 zaPHp2eUNRQk4=}waTz3TVoww{5OYIMiqT3e{xAP)*9O^j-)wRKo=;FKg;e_L{ffaq zVplW2BgI5-Lg#Gor@*f(#CTzbmK(O8v2EcjRuxV-9Le0g1ACmLdxnQZV;quD> z4T#2IG~CX9@N~Wj&;zK@KtN=dwngV8paVxh+|@80X`1KgJc;Zx)x2GzDmKH}e*eqj zoGO-NWX=}Uh&jDnf67$zFi1sfyQjVLYq)gf>m|v0q$Sx3MIb9VmIs(7)1y|qV9oVL zBuYyJAk`m0*``uA&W_L?Ls9`plC&*$R{och9eks%4rHjYzCir<_`f$^(E+ixcm z^q=qNJ*M36+#fWk3z^M!z3z`cf>lSP6Jzqy49|L>Q}iF_5;L|Le^cQXp~$NBJAuOO zp`_+4!9RYR(9T)7#k5&$U-I#KP5InBw%u{`JG(AnMm1M~+(+_<$Z)TeQX|USTC`kM z2vor5(wY^CcF z$R79icC``Mh}biRP+HK(VynVV4LMAJ6iefc1*_Mhxu0}#8W(bk-S^QPrqVxGCfylT z&I6o)v5WhdvJ~{Lsixrb;_XHJBZ;vt-Dt!rG=`wF-tuNUj`thhu7NNCiJl`ip6#u~ zX_(I1+mB-nqCQbN%_D|(a;YlmRK#NT4P(@o&??Uj$rtwvN%$S= zGUp~SS8tXU9wlJO6r<_+s`NZ^;-%F3l!s(~e9)~?!^{PL75cK)L%Jk+hF9qMCTNlE zN)Cn-)%l-5FmQ_4(V@6KeaSy6=9-U;+hHs6-!6gzX{}2Mzo-NAxe$cKiCKwR*}|u1 zjd5h-ch!XxzQ5j|F8?_)Wn5s%!1Hz3j!d6tMOSBwo%pXQN7&<0b$~4eK>WE5%^Vnv zh(2v*K%|2*PG+K`^SO*kp4cwqYSp4`Ccx-&D^A4^4XcPAVQ2lM_E*T>5vANK{_XwKQ$dbD(_) zAfgKc1L5(PZ*E=w9e-dE^g&F*pBXFmPhYnvB`48Ax>l(O-vo1OBQ6@+fNU|+*9TY>oMawZ3;P zR1zFqo%>-;hlyvM1+nL0Aq6H^B0fI%npJ(@weJJVee3Y+4DO?MKKCRjdm~GnwZ`ME zlUVBwe7pyzfBg~eFw-m>PF*?3Ay*f^B!QXE=Sd4)Vo)OkAUY*$)jnIYWPd<8vq0W# z^t=UVj#sa@Y`cC0Qdwc8YYdJrP2%|4{ia^_vXR21Gpr90H~4D~qap{Q0Nk0)&u1n@ zst63?6WQ1ai7KT=bq;c6jU%cL;ck8S7rQlTnSMs?ldQ=B>%#10npwHe>z^N=YC4Sv zV)m)<8Ym^Zj_C+@OGw>6bRtL^qVvlT83=wm}h__rIz_`Y~Dqd6~-_S^J za`Kc+w~9q}=lAn#tIQ^Z03-4lA=s>C^Ph!=otrh47oQWIg<)7|^o%3F?-42X-!AUF zCaFnDyU$fsG@nm%sqAHymDF~r)~mR`7xor5X_}noizL{=Cp0=+FbL;%-d?!aJq}{n z_HQU9?4eki7%&MH6ckE74?osb4LjzGg!CL}X%noT8zgbwK6ZZKViZn&ZCCw^`D9Yz zKxT80ZD?8ey1yvOHRaqVf;XGxIvM!$%Bd9|X&0$_yVK1@7r!aZ9p zmW*xI{^b%k53n=r_6_vW@n_l!AqPD9HK&EgFL`{SXT1 zy*IsmlCC&j4@v6Q%U5e{s0*1fOCExYd3aQuwUDp_t|6%l%KOwAiSa4aZ{7A>S7@hl zZy?kYbgnrU3t#4U27@~KM(INnJ$?rVM?cz)-&;!%i~}kfqgMES9k-Rg$yleyMfp{p zwVhIaJjrNq_C9*^7|+tzZgrtQ>r$y1B9m&mpb8t#tvL33p;f*!oQuFI3=E`$u*p6dp6aHSQsEB zXV+Z(%ATt43M10S3PC7nU&8gCe`JmHcjMtN9`$zvqY?oJ0C6@T$O|1AS8-8LEh(%BtX_L3y074bKFj? z`Rly{qQmV{bn{`m%L^WdyY|X3DLFekre6P<*L(4{`QdSQUleFurdlRBbH^(uO;RGs zas863$-#pb7KOTivXpYC(2X-2wYUT9V6QI-e0{C;U_1gA88K57qqymvfO=fYiC=C( zS*u|$q9^{H6A7E=MGmI%4D$(Egqpn0<=#<~m3>XIH8Z3D*F-o{HmI76y)zJn=yIbw znZgP@y1lQGfdc;%cV_L_-fqJ!FMo*BaS|-&&U{E!IB>Xtj+8EoRK`RsLhDdJ20Oz! zV&QC74Hs!jtONZpf}5Plo`L-8p-pFmK+;jG_WL6qMN8AaDD>@36yz7z5v04NN%z|Z4?8W@eqn@#2 zd(_lq?8)hF&@c*)GWU+%EYu-G1R>_zIm<=ZMgqA%lHXV^MZp;rj)pe9s+S;C3V-)pz(D+q{NtuN zPlM?&o@tXs25n4hD~X(_=sAR=r-TvSb$J&j@95Bkgv$(%$tf&JO(4yqptJW6e#4p> zak!70Mqnh)VFLTmKs-0lk;8nruOU7DnetSc>rqV<9m_pl+^|zF4Q8W=*`_$0Pa`y1 zBgaxv(G}4LBZ?3F=Ber&7;=|g&gF5)*88!WWfL#V6E`F>g0&8caq5erpk`8_tn*A^ zhij>&Gl$kb+?x3pK0iqZ1jvXmlfs0HNA)HCx`r8Tbk9o?vw2Rg+8ZiMd{619%a;rr z^qI0N@7YT*=K8){oZjZ!tZ+ZC8nu3nN2qOk-E^@-eT}D- zq?E1xSUU5*t&juO%DldxuUtPT)6E_0a0phXA?SK(+!pt06Frt(P+baoyM=*)fz64J zy9HvsmWy`2Cw7CHW27Q_d|t6ysSEw$G?{i=P!dlCslb8v;XHV5PiFWC`J>6E@KSBj z@nB%tX2LUd>jT1#m7Jte@?c3$>W!EgtlXXcf3U?x*Sq{Q^`qp<4p8I$I+s1v-GrWn z&~Jal#hTFkW2f6gCgzDjYF7!F!zDT- zSV*Gn>tDV{m1@0)L(wV5np;_?RO<9?7Z5J6a#K+!bfHX~x5cMIr0eN?sOHxk4^g3! zK<|HS+Wa*ZGABl4M2yhqRwex<0t;{U#&2D)sH_^eVBApOdJBm>;dPj$B6L8>;BaT( z9m$bZT~ncWBw7Y0u#2&}g<)w4^eZCP)I(R*c52&@*3*F{fX=7h^LR7Ux~@RipwWrW zj=&v?WRlL2IrF!RXN0ocf<0$QdILDwv!}AO^6&NGc+W0f%zSw$=)Vq3L`I82>g4wQ zQ}_BTN4B2k&O(||QsBUVh<>D?M9HMCSTu3wc|^?23Qv4L*4wBqYOWZ`Dl9@@Z5)g_`xQAqjw;w^u&1%wU>Q~si*wS9oJ?h&)Z($7XuN#2_i*2(9#N;foG(pyjg`$iiB_r^F<5y>*jh#l{}ijUAdH|S!@2-Erp1{#z|ppJq_sUKNY zU{Yr%UDkC>yne9J%R&UcyefJnrG!E0pxT1R2l{Xtnc1p8LR(ZtA)EFDtIIvb2G-@m z%`x4Fz^IWgiZ=vf7r>H!tziy$n#EowN6v)nS`d2{!=I613ILY2@w^^Eo~Gr()X@Th z$=XKkaM)34`dllvwq=w{P@dWHH|)(lwpT<1x$DJZ6fbY)?1j;Z&PJ(Ko0?&~GsA^5 zH{6zYPj8gS^lDM6-s%#2oX!i3vl)Xq3E`(Fvp9=}zp>g~codjrAGNtC@v=WBOt>T` zDz(6JL&F;70!u+q;*Nu<0;>@7$TP`U{X^X&}y3 z^pnpQtAy9vQLR43q?IZmDIw;J(rM&D_)DX(7-%NYwP9;K9vLnZ4YE7W!Qr}1D)l&{ zgU=@EuAyxsVgx=-#zlNfBHki{C`Udb$Dic<22__wMFbPZB~u1_;ZtcU8!GFltoHU_ zR5j6jjX&NQ2riM$ zwq#0A@#%tPBU*d~h3B5-g^kSqa7Hj+||6-(=YuWaMDY@QicgtjPO6k+T zbto4Chi}^p#YWgTd0NrSWpL7#bSFX0s87bNiXu3Zlo~dzI3n3YLu!9DI0@dMSzcQgt@R#vn8fNhA^Lk<%jmz$wfTNrc+o7q%jNhm6l>KeDYIPj0rhF6 z&o`~akCBnBk^aJe_6vN{nEFd>{N=gtniv||vjrt*1{=Ug^Zl-9xS#v|q}e8Va`;96 z=9TQHF@~HgOf8-O-s+NClvy6g4`%0s3~_km213y7S|p#=WTBpU ztE!}?HLZ1$gxhp>)Jj@9Qrc-#A~~$Nn1g`a>#^&W` zPDkZyz01w*dTn8fx?;};`C8A=Xk;>Nsf_=XUc+2w5dYu5Dq=2e#ni@xmhXG|9U$7T zr(*W&6tSBcSS!bU+9Z~x7Rg3aK_+uybZp3x41ZjD)9dKJVL5LCWi8(Cha1`TTwd;X ztKEgd_k;rAa#*g?Ca$CSzjRf#EsjwFSOGcN8n_4p93VBLp;XNBGU~fNd0t84vh%p@ z^7TpNXt5d5749JAusB?I2rA(5Oq&C%l5&n;w$|e-`*VDF2#dpxP?2=CrL>^;&p?QjA!&94q$z0-?b?tk z7CMGrrWz^7$p(0|&UW{o8q0d4)a{WYY3y9U&QAUupyIENqR4PJD=jVPX!^s)UGH=B zQRJz0^*x~YPpRxaEE@28Hi^VtlA^;MU>@WS&&77AbcH6MTQ}>R2xBY zhF@RB$S$MjXk;cPGX-N@zeiE1(ExJ`LYMJclNmdXU=Daz`vQYt)qPY#&i&Z;J9!A= zWjSHX3yZwTd-_QYE{cH6J_9-sRmZ$^7_BVW>c9!H9tt|s=0J`@4eJ{rIr?Mf zA*{G+^>qlo1{6Y5r-(&cCiGFR{pwu;(XfFnR{oSIj>GnLOU_#Jj+b&rxyh~P&pLy& z;|=h@h>V-%Jjs^*+Hva-?{bD_KV?6cnWp;mSiJz0^--VmiO#o>2K+=c(CGzfQL*0L z`~+Xm18e`P%4hvti&;9a_iGZiZKH^2(=9Z5md$wL`Nes08*aj~MQYr(N(wh=ZWc9n z$GvbO8q^G_{A%#ab)sL&5%zevMC3)C{$}@#YT}o0Pluo(C)C*=2B*O!#$|C}gIS%` zrQ!%80&Kwik@S0gXDm(f?+9>vS+O`i%Pa@S_1n+Y-Eo1>1_AiB@T8nYQ7&uuZU&T^ zFyn}GAmTbu&w|e9%LB(kMvl{?7D_HgkNH@^5BrZ3Krb|PBBYGBbjfefq;#q+Pppsf z#9;6H(Mg{&m)G$SQhKk$`K(|ryJ&c+L+^%mqotQ0FSFipRbs|ifHuD#E-xm2=4>~N88VZ`1FkkpM`?KtHp}4 zGV@R+R%BA=)6E3CDJQ(DX>hf&N?1%@ZH`Om>XGW#7SZ&%YWtTidcn*{BtKw8CC0!~1enaD`} z_OJY$Hh!PFa8{EsF+qR-$wXKPGVf#3M$okG4mC3)@xA*88^2psif3d>kcx=QkJk&W zt-g1P`8hg9Ce%gIYG`zTk&?RYX7?u-1Dq+R^Df$K#x}!g z={~mq4b}5~swshJW(0emMW9Jyhcv(vxU@vhXWy=1Xq&g3EkBs@QGy%ERw<)Uk~h7b@$5FT&JHCK2@BbyU+xkzgRaq{bxi& z^C$mmrp`xdZmdor0DW^Pl=5I7b6*qzFYAtqzNLrZFxJNh?>!Cr(StTIsRlJ7?CG?t zOUIC#J{*i{A+qloM7s9+M7T0?)p5U5@>;OauB?q;6#E9JwBI~2aBjFm4~80{?BQ40 zf&aMkz2h2BmWdrmvc-Wh`Te>YcxI}sH$opBHAOj1b6zp_uI(?{YW%!4rC<%$!X({;n!&EBcjrBc$24CXC;s|;aR82GMQ6A?J&Jb`p@0`Z;07RyB7sEs zdBN6aY8^Q+{{DMLcbuM2SVe#7+`6a9`(4mz%D-0<@*jON3jjJULOxv8(3YWx`UTbr ze;GKZ%7ts#xRMQ*?4Q-azw-jqGXrsWHZC6y?}5n#r=H^62!*jojtVAw1%hiMYdIY9 zJS1iEJw7oLvJ9=M2e|Mqh=BieZGf*ldGIqqR6x=&++$Gb(SKa+Kl|lR#E+NdTR<=eqo~F-AGGlJ-!}<3+Ur5YZ!z@%7QZd~KfmmMX#ori zSkej{%Rv+G|JVcj9LQPm(-SCyt}izP{Rk}y8mu}Af-K2DU(-w99~u&qnsa#sXietA zxPgl8ijB2{2R*FWTC!Ob=G-~rEm8#0u`rvNHFvEF=4oN(!VTo&Gufz<27yAIBpcJn z9hnN6s}!;B)`G8XE*UV-mJpRJ4GMIkV~Ae0qS%q#BE0eR2v-jQ!W}*u`TpMaC+yV( zp=};H=G@T;!{zHi#20nzg^+F*%C@>EXzHx&R++oaxRAkZDp{l8zuxpyYm7xg#M$8y zH45UJN<(Ci5p=-{P(-Tsv5;{0~!R6qP@i_Kwp_;32xX>ji?UV$TWQ>FqGiTxoKnV_!oF8z9h zV~`@JIb1Bex1 zx!U>C$2s_AByfy$X1A+)-)DYyHc_NVl1v0Tg9qjj`nv)Qmquf_rQQ+oC=!3g%^jWG z{?Ec$7`5lKp%~BrW2=a{47ds6ekL3!2SKx64xXqX>J}_^)=-O^x-zt+DkMn0x3{D| zF5NPczM=}dKt+A3@y4&e4vdUsjCTnI&-}ZN1*pch!S+>OTo`Ew&ws)qDF%@0W0cc*Tk1HuDG`*BG0>A6uD!4z(7)wnc<$C+#y{fnriFjDp~>FY3R z$rb_up@tn0rT*JJFhlAuKq5KY7mHH_yQD3ENhU4BN~;l0g&qSXJ3BJYG0w5q0dTX4 zauBB*k7o-(TAU~%AKHvtlO*bCV30t-maHgH;T|h9dbQFEmweL94>hK`7BE$quc>xD z9q0>yY zm;xRL6!$1^{M|5(GZ08u&n6)t_kB7W=D#L_?*?$Zt)k`)PqeT4{yt%lt!ky%_3k(v z4CiZWwpW?A%xs2vH!8h7Bpeyw-LBUTwky*p=m+zfoWsVIy<49-GGZ`!%^tBo1`~|e z{^$Lhz^}&*-{XE1jWW~$4-?>;l5=?%dS4ZAD1m*30zLjQK({b5)JiwzLOyHp_M<0e ze{5u8#JQ-gr6QKYO15+YvXIFhu{lJIE~{tcgwMAro%wosujKQ6d3u9XoQaxd$R`~y zT!}RGGxi2PG(dIX%pV#g^9<)yFZ}(V@26VG^BaG`R(F*=KX0oQwtzlO=SE2_OZg<) z^c)_JM%(~*g=)xeSz6l8fpISLv-*N9UMKu}EOQl&Zs9 zVA0dm*e#!3gYPM%aukVT@o~J*d}9CKV~6+9 z0TjMnM>er7w>=6u!nP0!UisBMuUol|@Zr!;Gwx6aX8Dr)HQrjTmzVk+Fi>|Q`c-hP ztZ3B9_V)<_-!sDWhMGo-`|a_*LIDS%wi*8VJoboqR}Jl6apHCd&|_CK5tse0h}Ib) zw`jz35UA89^c>XfT;<&v|8zjxuct#{2=W*gMHQd44ontZ??ws+ms530e(a`S_qL*E z21P13#L%6KIBMoE9*9c^BG?RkHUo^??HDrVu^hU+can~ZVP>*dc)j>C2f}U?2XV5# zhQ4l|CvKft^EfBC%yUtP**lDt@5k)hTs=nG@yBi{4xT5i&!D3gAE7<6^}#BTG%zIc zz@pA|cml^9fq;54rmS%?$OsfotdEEnM&D5T9lCsYeN18xosM7*;${_u&CBSH>!0Nf z*Bo>WaO6OASGEIM+<~+#8?oa*3SfcZ0F)eKEnrgsh}wfnq`B+(MB3{6Va*S^-|v?Q zKth&N$Jp4|z`!8&lFsJs{UWcf+hMOiWN&ZJZObXO=l*lc+gJTz000^#lgrVbE!=gZ zFG1I?g7jagTJ2HL)8RCHWhe-*Wn`o)fKoT=v%p(#tZQ$4%&;o)Ql>&KwP$clv`Uyf z__MUEYTblWDPMct+WNXU$t}7tCXJL89Vi8>6ls8zm{8yG zf}j1Ws+eStN!&sW^g-VV_jXVxvuH1R+Vy? z+*$Hsers$vzNM=80~cMqoy**} zK>*rH&Nz+~$LS5!eOj#izaUCC> zXr$Dk*cxs-Ja94*NHmp1WJd>73IS6HVCPCblG{wvsV^7(b6JYV>bQzL$GyNxc4enM z5A@{X;sW3IaPPJC)qgHSEDi^k*%h`TBPT}|Fe7l+NA03XM1`?G9eGLe`hCPMTV*R)LPg8i1Ot7l8 z5CsbiYbw6dtn?5r6#s$kr)lwCSIFR%aOK}2U_oD4czlRV`L%2P`FHwk!)KTj6BG+X)-Y!?Czh*ZVwM7^$XOmj*RrdSndh^%>yOfkJrjv z^e@5w#^VVwF+pWzM%A?>{m)tPT3D`@DkocQ*+EXCxeWIz=`|5i_mw1q0j?ro-vUI3BWuMrxJapx9A-#gCi}D%WD3E3xKoK zcsR_4P}HV>W+BfawEz8AxZ3{W=IvQ)YB12QlIYl59ZP}uLN$*@Rkh)uW{%8YhCBjFXIxZd4aLjgzpGYz2@L zGieKYb+06OC(QXH(|r8&%~zS0B1@7Eq(vGOy{M57uZ-jw@e%RZ{qd|v&)+C#Y^zQ} zOjF0L?t~Z{f%@z1e!JU`pSuBkFV&Y%N`%#_*6Uy4$Z+}Hj*8x(!z2bQqD%C5vRLoV z7W7>2@GIaix_{i@aTG|c+g?ANeQ}z-dwcFL9p_K~nI@k<#lhJ~+E>)wTRXf?nnlDz z!QC$YHUw$%vI{TmQCDJsM^fH~itx^Gc9w3rx&C65y<;!v?r8q4YB&@SCvJb4JD+Z% z7u8!@z3amFqY~gs);C(@d&j;R)Tbrracku^RQx);lXNu zwJ9m5@wlZ@^ZyxsooOJZs&OLTt}Y6|>fA3LJks5R6l2K)VHO5KKo>KfHKm8ahsrP4 z`iYuG0JFsLfrMR>)u7%;ZuEyKd-q*stRh@?{7*NA%gq*f&u8#y*TCvQwHwx=3#VWUcL=0#% z-Fo!TnR^-S8aG~HPnczxP<84q2XJM6VK6{vg2Ep&=!_=jmatf}>&NdFhqvK}v0ADy zd%nM5t!2I18`r(>?7HPo9k)}lRxy&^dDymtbMEPpDV3`79Df{`AI|8B_SxwPVPAp5p!ZYM^4{5w zdSggn!DTDeVZEv2jL6bxn?kf)5fI6*v)E;2-djitHYBfODYUrAAD=r}8`5!0RVa*8 zR=t~GU-xZSx!Y$v1NfHl(nn}+9Y?7j`>SjDUvd(eRdM??G8Bm%Vkp^)e|IRrt%~iJ zQ$fSR&-uzf>`in0gmvYIGu{KVDm$wD^xv^U#iKqZzjw3Y(s8+GCt;Nfsgq6SW9A~C z9O-M8NSsGR{wOmlnlXaH@%CYF1}L5 zK!GPt9N}w9o@pZ6Q`&jMXL2?&DwFg1w<%QN*|w}r!S3HO6e7^-SWjn5x*AiYH&^1P z&DpBoB`H`CPvRZdNoYn&4WH)dS}7b6)!B4Q=fYFH!ZcXiweR#5{MBpZl&h>H4~)AU z=10F<}HEH5nKm zbb5)fpsYC`(OvNlaiGS=#LT*EY5I`g%#Pj|0o=v7A#sQ!yvy~dQOT4I;sib;g_F+m zrqu2?4+pY{F_)KpBtH!7uOfI#h_`W0H)IuSJlM*#q}yybGqGtoJp;UI#+S1_;roZA z-&%9!lbr9ym&J9NGrt6eNR}yGO|zM0)6&@YSyj`u zFzZCyQS!EJo|H;2Q$HMptK?S<6_&df?EBV0cSJY3tQ<^*Qc*it3HFuwl#sKQ?GE(~ zlr*RehOz%y{U85h^$Y$Ye#v`bxxefT!=URLxbFNxX!&a7@I+Kpo9lFPSWExB6{xD~ zeUne6^_KiQ^XrdI7EfjT{Ldy>Er5LVTYXvP9=$dA z2AuV3z7029=_et@`S-j~Db!5S=Y@ocD#Ef3qp)wJpYFBBS|+nejYEXxg5Sx3dJY_a z9%=~9tUgBg9x(DDki+QIZ+_GjP26u)q-ai(D5*&)U-{TsETT(P(GZ8v&b^w0U+NT9+4rfKe9+ixDyQdR2>|K znkFcG$;$dDo2pW-#7;R&u9LKE^NxU5Z?Opkw6a2ZD2S9kxGlDa)G`q_)94`)4B069 zOUahfY1$^ZdfRw4G!98OIfn_iW^%d;?k9o5Ave3WcMzfk6T$){#c6{k?lN%TOjvZ1 zq-}&MuhbPuu$dHrun2UzVFmzRt@NKZy3Ka{%a6@`XHm)s3DmR~$ec&$- zERN`m6*J$w(p+%l4sm3PZbaP+(&m#)U6>#08O)DgQ~42z-+=vkrR+|dxjTmUYx)c7 zDWmTk9Ler-xondW_mu_sSnkjvU~B%l{MN^*(Hk)j$}_Y>a@1P=+vnreR~ zVuVKUeucJ&d=JHn7=x28h#WjQ29WQ`W<0vFO-q_7Zx=7M+CjsJlpPr&ik?*4W;w(~ z%c8Up5Wv#6(Y+^o$b>NWKk&CFd;Ha31`A2)h8l1EIX_IPtCMt){&=#a3hNgJ>{&&z z7AP9KK-fw;LSftUB5b=Lw5G1}koEgD4ts_Cs=L%Ij1FtIhw9+sJITlcTwgfnw{6K& z@(<*GiQkn3g=t$>hdA}QD3I=tc7rnuLohjY-;q5~3pnIKfn)twQz-;z`Cm|dVd$>2 z5fE$f1s2a7zAVssCK*rbxBNA24`?d;akQ>^zb0M3=D6u%Z~B%gzu^dFW@%bdScuK; zjaSh3*yp|u0Y%K>xIZz)v~Y?q+r+Rsl`AP8g-hpk_fu8Z>n7^%<)SBQ>g~T`^?7!% zW)4S^nC&hBoTS^sPJ--1oOf#gQpK{w*T-P_X{MNL1RMJAlez-Zf~6`ksk{duvxih> zgfeZ|g%7Auxb#~Y9)#aCCF3d?f+8Th!K4|}kXy=*lNhuE682)kzA0133G2A4PZp@= z=CEGHjQGc~(LjCX@FEt$L?$YJ_G>UPFj>n%rVuWxMou00WSC!F_!>MOs3GN96-K#} z{eTse&YFXf4l*dKVX~jOv zHdn`?B!6)bEnf*%Y!?b82}ZI%k;_VBj*}sk7K?9G*BBb9!n_OW+5Q!9uZF&*g!D12 zVYBMwdkV!PaSEqOydZO{s<#RAAr(hze35NxnWl=ppfmLhQ6>dVdTi%aNsGAQo$?nL zlb6Ta_ZF7uXrp3?k5blWM2oGA)=jSA#6$;b)AP1cG}hPVXRdjl>R5~8ku@a^=T}k) z0^tEg4$T2##5^qLml!jCp>+vOs+Fu>Tu=%ry*Q`70C1))1$6ZfF0qLyYKZMlmg2l* z+9ym$e&YQ)Ra9P1q%t^YCS-5IPEu~|YOdpb6Iy=cd}_~~N*zD#zA_sZ*JCJjV5Z&7 zwXr8UcA^==-AyjV)#!N~rya}8@4NNRa%xD{BFhPH+)0x`Ws;OSrdI6Sy#-a!9_xJ{ z_`Zk;W*KTulFnyiWQ$~?bNvGqe@?5N<6!9=Bo1M8O5?(WKGMELWGJ1!UWkyf5mVzZ z_>V=a2nnG(wFE|9R;}C3;7@wkgHYh~E?Dk`q##PLe_@FK94j0C%#rY!0xi5GbJ%TK zl(#!?bi7~3s9c9H{$X% zC=8Rn@5FW;8?$PvW(nodVnb@O+H<3(PouCOVy9adyQ1px_9a!7K#1EY*ewM(Xmb6qWNr=ouW@v zXfJDUQ6!O4BJxsRBmdoz@xrUii>$i>aTcVqCY_v8fVT6ueeC$145qx`$iUfF*{;r} zc_U`F(l?wxv*WelBq`<>t4E9)s{{=ltQ;`$o)SkS?cZSan|4m4!_{sNZHgP0TdMm= z0H-kKmSZ#M2=P#oea8AeH2ATk>-Tr2zM-ChqSnA*fnZ0(yF!`#>dF>?II~WJWyNef z-8EMdUR~lq-)JTkJm1W`s7T`$+0wo|nX7sZLwG?1k<)DBhGfQjyr^mNdJ_dTV6NY~ zd2Pnrw;0UUugaX@p&0s+%lEiK71QVX*ps+%QpA-~CN9NxrzV+r^GQvD#~G7{6pfds ztK-e~hViEb>krM{wL6y?Txvlus*@O@QxG#vy`pDIz8KB->YtKYXXhmEIE|{^vsF!z z)_ZI0FNR0O%|oU|(4H}wLt)?k`Nqm5J{&rbW)rf&W4`GLi+gq*R06$V`E>M{@`7y(;toZ8 zs}SEDG9GE;lwt{lg}m3{c)|noF%p}9&aULlF?-8$(ZyOQ&+jbbde?1R$wpnQsb-1E zjC!;x9$l=%$?Kt#pBHywrpz@AVjzo5Mzh_L_ z4?1YCmW=~C;Z~~!Cd*#IeeE}5Pu6EXA?#sxRNStudn~Dh-h=xxnAgvWOw&lv+A7~? zdtr>%5HzL_?CX?pBTtGMOQ-@7!$xReR&ewc7K|bmra9{+9rPIsNeEvpes8=6N?RKY z_GwdUt0I-3pjpeQAy-J0@rhsMLV!ckdTA_$tc=uGt!n4dBJ^T% zDUhwbcYpiu{5OxSx2ozY`NC-(0a+blC;_dgNLl}7rU~#z4t`?VjW(A0Ac|V2+=#_ZtO~?~lo~$x7 z-;y_=?r2Bu(6>tn2=w$~MB*d%Yi_)Hwhvv*7Ypa;VaZ;+62BzC{yv9K6I8HgrijFK zgNqW_1s6sC_(77z?rP4eZ(bK-8p5sIVV$pjBu$x$nUS=y{z|Q$*6T+woqt)A25zc( zeKz7g)Ye|1AOe)&=Y@ZR%2$T`U#R?&N3ACPJB@y&@&pbwc1hw*<|_LVggQH)%Ivz$ zu==$2-L!Ud61~b|nAr~2oXr9jTK#Y-IEt4hqlV*My5{7_nV%4{-j{%)tr>}Q zvd$mm7;wS5(p}%zyGQ4r@tZ!LB_YpM3?G`!Kq<{MzxQc*#3^+iDXlMa+Zq4-60NHe z-`R}dEiodK&UeMt&>OcLQ7)68e7TUVyY5jhC;z${-5nE#zHfGW=rsRwH}9&-S*1Y= z6^<2$Ww;o>D1JQl;U`^T-nazk>Q zUBr7QiI1@UaL(gntz%KcP+IC-U+lg2iYc^}g+aMvL==!q6&0CRW$JI+dpUh=h32V- z&ykc4jR(xuY0Bd@8P=E`Jao!Q+B@#y~$nb#z|{A1qLLpR9X38=JPuLPAeifGVZY>H2=6(rWZ$=AcOLb+Y}7 zmkmP8^CTR(~-I4dh=`>Vjd_qm95aSn?V9lkgM)RMC{`q@6-rX5{yIO7g zlSJw;J#EBHgQR6xe>)c+}o{J^}sjygM7fvo^*!3d0s*@x_4=tg9qP9zUGhSRBE)p7LI%2Szk`j_GY+dkv>ih zm8;V3SxO~}L%b|buSv)%fPYp4sth6G;{U2Mh=*QR&)97bAUp20wD3Ihdfr#qC^Z54 zfnJbXZRIyNHyj+c=dQjjoDRDsuhy)VgJ&M+d?{~$$lwL0d%`CHulf47X16P2&y%@Q z&8OJ&Ep-VPJLBF^^tP*P@hEGb&NgCN;(<>8kIU&i-n2iLZQAh-_t#$^>)Sp&rtN~c zAI1oO;waBT+|57**}9z65okVL6z7;T8;&tn@Aa5mvoSFKtDfrtnGzkWK{{<}gl2~l|Frj7*{IJ+7j@0`Bqvj*U>~Jk zqY7V)!SUmFD6<~po?XaEY*2W2CW&%-33>n1C!TVi{?&I^v;Qxoys8s-BFU7QW^lre-$OQO2R6Jo2#H+ce*oKM4%AJN z+92@{YB<3!a)h9py`iV{vMF*L56nm_gOPvzT%|H@SgWwM`*JbY)g`F2?XT9qPODKG z54SDDzrT5PvC{xk(4!a!B7E{mGBv0fy$kpcB??nwJCtt8V zp2BfVmDI@E>0V{pR1!)akW-9B<^u}rwf975IoI@3W~yFub9a(MYweoWNFH&~iJJVj zZFo$*6VSl?2n?6}CI6U;&X7V*`g}AQ8JTNyRaMpbytwXWMp;?e3(XG$@Q2Rz-Cg5= zizF5^l(z3Xv}>9xxsnMkuP+GRO+eqgY$Jp&M1q_mtUo#mI{M9#u3cVf4?se3i+LYv zVHTZ|ogOn+rrzusn8OaEn|`-KmNY_gvet_0qj%nBwz;Qeg}Wr@wn~OB74K(Xo-ZX zFC|5g3Fs^Q=p*l-#K_GoEe*0FW(1AsaRs8Gk!+)&VnlBEsXZ|K3nd>4kBgWs5T?*{ z%l2H;RYd=rbcfr`sh=zStCSL{d&u{nYn*0lSgbO-`zJaN>W8(tH|U??W~@1dQ=A%h zDOUF@lj}N)^BT`Kke%5eg+FshbBf=`!tl4#hfkAuKegtFd(AHBY3ReEMqo*d)62S@ ziLE;~A8$_(?BC0zaV%$kwJ?gQ26{D$|*XHaP|4bhJ{HphLUVre9>_EVF(5+7a=md*`w(* zMV$;4cRsX8rx`Vb>$=f~DwKKV`BtHnlDKZ%e9Qet=IX~F9mZd(!vf~+gt;H(H-7&d zkn5k>)}EJqdrbqd@nsF3rvAR1Ddc}TMmGfjdM=>VunO%!R-<}b zAY}Zo#YXqb2_x^zJu&q4j+rvO7k%Z|;_tZ9;5d)R)#L8B$04zS{xA-(_C7D6B&b;V z-6GoM^LyKSHQMhiTqX*Wl9J5K%s4o`SMKB1%PgO6Mn0C7_U|bXrY^nebG+;Q2+I65 zXNd^Q+op!>tDtG2xU+kgMvu z`@j}&r81ReMiz^GEk|ideQZ<+Wt$iU)dFiyL!k?npkb9qUft)-#y^ZcuvwjQNgguj zJYb%&LaD0_IQOVB?kK=*f$@L+JWu$bj-(6=%9Y-XzY*OSKNq}y57qzfsQ;!{9nEAJC4#D*dSmDDF}H1S zjDJim>}Mf)x0bI0xP+x*qXL++TyUk}6195BF-K}*W-df^tnQ;1UzuooJc0^p*UHc| zX03~*NmJt8&7>)K&SRi->-)>`D4fz5er{)KJ)w>k6jD>nL0TH@+~JW<4xu21h3dY8 z_MZNVkihPdbIMmxAxEWB08xxaz?&1humDxV$ikIP+0R3G*IzX9Mm8iRW^zTm&d zi51Gfcm2@4sD_YMrI(5nJDF5bveV%`xAV&`681LRFD*~qhWhdC_Vfc!sJYFDd!X2s z_Y(;!V8SE8H`oyg0^;C1*mnqAY55J9EGTN#PKytMp{gxegw$FB0!lsKDXC$gROO>1 z1$^a%B-WYMxj7kVlZA%rEZ4j)0k-=TPUfrRq}9~7G$t2+fitb09^a2*0V)wZMx5#9 zLqmo62d%O$dO+m-U{!t*o~`&p29q`^Nenk=zfCYFVCZdnX z262}zD{kq$S%Otrz4zPF{KRNFtyd!)9=pwjiam|hS;jZwnwthQdh-U=pl!fXS*D;Fm-=+lRq4%+g8J}Rb8xu zPWL-<^HLnf2v7>EMYMEwQ5u)400;iAM(r`qcNb?ZM)m`rdOA8P5iNm!bIzxNG(I;L zyXmibj@Nh``zeKXp+p2ug)Nm^WPt22uQKXnz}aQP>|tBBNxU^j7=}ysS$l3u3{F`Z ztg@m}Pw95?l_@lJ*P_R|_NSEgelGV}sr^)PY9lInC>Ff+s5&sSEYA`KaV~WAz)>3U zcG6}CqVCs~O_7>zd}AIr;pTWTyPoZyhwSIaJ?|O)ut?-g^4ZN1sDZk-&_PS&xZ;ql zc7W->M7_2KiA)rMJ>3y)b;hu7-po`OLf3|6m^S1-*GQNy$Q`}cCcn4b17WUE5!Q?T z(?n2`5B@VQdQ38Vq;T+Jmj1?;FG5Y~7cRe`_j%@aeG`c+$Op9WqvJcs9m(6B248~B z*b0n7s`N>lX;%gyzTLl0J!HM~2VyI?eOHl`!$MwXQ622aP%uphryksseJh0z#lkjY za61!6%hjiC^6oWh_`SdG^Mi#JbJLM4!W5fsjDVYb!piaC5;B%5I9oGm@Swf5N-zUo z&~ty{>1?PVBaJ0pJb=-J()r}w@0XWNS(+u9C|{tgUu42pohE93Qs~>;o^`4g-e9yUtcel36JK>&L%RuO>p^yq)K+44`vo?4Wg+^Oq5q~&QHrr`_68gVw zaE|}Lv)iw)>dHi$KT`o=A=YFs}As9bu8;;~Pa1@N&_-=ofcO>so{O47fBv z^`?`=CLuIwMM}j}E&{ss?U3*_MH!`<^?lb%H|DyP?{wupK|xGdH5&}~u@b-h43=nq ziqui5UQO~6G)%8Fe8C4fG0;BAC4vkuWh8Yx&qUGAj_%e__r`(;D)I?ESg66ZyN&co z{RXaQE(@t93#oQa7BR}eZtm89wP2a)$d$~jB13$)2XU~hRLIV$y6bj&{O6D!?zW3>8xk|pP)#i9&{NkgX z4Ef)KH4j`r1$&apa}QdmM086C-VcJ?(7z{`6Ztl14v%?Ji?pUyd+NY?rIh22x7Tv@ ztm|XtimCm6#(afZF+^ydW{m?fSR?Q^bD4~GxHsR~x(&91;`w{k&UhmUuFA7ivM#go z_1#CM?8QjJ?C+<3a%J*Z&Sei;xb`c=1URqX0;r@6xoRiv6-IYH2W-#}VHm6Ms{TG}5! z2!ntliY!l+fQy=~o*~`y-}1GhFD!}}IK3VQtF*CTXy>Kx&WPPGv5{P&d!JcOR`Xrb zFh`h~(;DA0hn?t5MnSdiH;1iHxB5t#?>|81>+mL&t3@6r zu>*v|r>o8FZ@)c$F<2irIO8ujV)U{=P=122HE_PX7rZz^+&7lbNm08TQsf-PD;0{q z7Wnj#hTUv6%akEgm!;6z(p zPOX+tHJ#xv@1A|-cDG+F`J$Z6HliXN1B2>aman>0>kRhzJ@w^xEX<9m>b7+q#x^v5L$lX%qQ-!XuAhr4 zX)x#cL^PYPM3#OqW2%Pp0 z@a&CQ=+mTQ_Z{N^0&?4IwraNEhuF*c3!<#FxWM)$NtyXdo7a;yp(nSwT-(!HGgIDT zzDs%>jVcbEs%LcTRwVbf{=Ui8<)ucHyarp?j)D_W_WkqKI0Zi+Hflm^yHp zvl=~3yjyk8R0=I$ee9$VtyO6_jeKj<8zN_-K3 z(bGc*;Z4%#75=;^y*{0Si}w=OkLrpUT&Pev$Zw5uQ=^>zd{coj-Lqi5s=30sd&**= z63W?qTDk;U@<5tf^Bfw@V)Q1#pl~7JYJI5~z=&I(A-KHvUz7*y&R9NgrY1y#Qh2{6 zs#9Xz(JZkh`gn*#Z~r_d-@})zzooyf`wGa$uyLS63_QEK~|opvW%Z<504 zbDP7;LdwgzLBm7kNsYsbUH)1q%#V$Ookj5RS% zt`AzGJm7QtG=z<*TfG$Ch;btn)1kf>M{?S#U8u`!J-Y&e>czpMu-(a0$h@0VzKlVx z4KG}3nEx=o#5}%P5(ZT*CMufEKgx_XEgV2Jgw#>HP*Zk}GaH`DoZHvPvK%r!h#bA6 zPt*5J!ZG`i-Mlh$$#T%EMHTE@YTtPsJ3OP)k-wyj(-O5(K>?Q3(oI>wRs*=@I%ulL zye8^+-OQDL_+omGk-;U7+K&7;OGvh{iHh&=@W~n9LnG1+n0LH{YL^@tPe-lxxNz*i z>m7|YK8$(Qmh>)16Qffq?tqzRpJ?CQiQ9Or`R#*4J}RBD!NVmZWbA{`m`RFor~LV$ zNeYU3X`e6lc+r8G2@0w(-R{EIxQWpdD$n{(la%-Ac5RP8VHq7gouK+mC6P6xQ2qTq z&I&jKLAG9!S5Bp+&SBVFvDyzEF1XiN-aWSDtJ}RB9AXC{MFi7sn{xeu$J>e~-QC1P z{l759b^?e*aRnvpOzm{ac`I&%?+eDF;Lr~uw{xm7V#)G=_BTTbrkw0FU|wX{gzj+o zU~ij&aOOt%Ukmk+!e13t=b2!uX|*8vfXlh@s6;<6B1`6l3xYOG&EyI^@>bCo$wXW9 z-TEb?)$M9YWsoS(L4*fMvOxAyszF$F$$vPGe){A94;+y;206h-v~w zT8WTVJ~Ne&Q)lb>+NQi~D-;Xpu&&eZNsUn8BWG%Ak>?mUCod2=xbn|bUQ?RDVf>cN zkLjtOQOs@Od%Ks|AGDX^l(}?9HfN8lY_U<&ywP3#ISQWoIw$2yLDWmC+8h>Ev1kTL zF|d1sBWZf@8x__wNq!8IIc2hCkga5|n&<*!0c4w_zI!J8bCP^n7wOsdQ&=t7u_6q#0m-_6}7r3#SN4;ic?_Dm+4Oc1{R6pll zxRNr?L8giIYoOpf*#16z_@Q+UG2ITz$k>#yuFdt6`9Vi|XHrPI_{+5|On`N&2J^is zG6BAU9Q+{hhfa|IEw2jEgn0rgA(HHh+pB%e8C8qBP6T^&G)e)$lR|8`+G6~PG7HmX z_`;I+EeyYd0ZCm4jHwBEjhVR0 zF_eVeBQX2VEGJ=dR4tn|Q1&vE>RAuOGQW?yyKmfdI&N0U_uwv{sW0iG)9oInTpfZu zeZ<5o%wC-kbT#UJ>DVc}PRX3}L%`5)jJx1^_o2UX zxG_*HT^Agt`E`TyOwwxJ3F3Q;Py2EJ=d5k>AZZCgFSa$DUGIfC*Nk3djrs+iH6mjXWZ7!* zQdyvGOg)xTHFY>bCQiU2t1wc&-4)Pyquh%hfYQ^9UC|7*X|IVgxo;#7E3S!Z%+$op zG$<)4-;D0YFV4`Q6+>%m*cH9-n2%{mZgB zYYFZi^0?WCR7{zbJC$6SvFM)3@ORtO@yeE6e>qg<7%UMn{$9Jh)4kMMJB>Z1n3j<` zG3_)`5?$AeeXo4lWm^l3>)4aRvXi#or?b5?wyw`2b&8eCaLzwAn>DerixN}GdA4l7 zwQ4mVnH|j5{Ho`G$DnAN+kD3nc)n_PcUXp=Wt*}I3GNVhd))jKEM%#cG|=@QZzid| zCYT-Cmq*V}eT(CTxleUVZZ8OfrrD9PWS5#K3i4?KdwBCujThTX7(E~UTFxQgaHeW| ze;g>yxGbWy`21+4#9T@|&99n*^eew*{Ii3XD9@e~3?aoc)!Ey)%Sy}nYj3}amOu#t zvNRv%1~cIB&1byby6^s6tI`$I4MM!owQ~uSyUAcjLiG=8_|Ho78wMjY>+o(w%loHQ z{%>pXUdmTbI$hpD^snpC1X%3X;Utoo|G9Yo$G>`#1LiVb&`A8h4C&v$(ZLK@gL_|e z?qAJ;5A~bcw8@MX^50Bu1N(gtEAQ5T!e8Gc9h#x!fZ4n(_`loo|9$B5@ORyyG&@78 z|MB?$c}3sxKW~c7sw(m~s|L=7^`yr+s?h%WrY69fK8Ay6vi|RV|NADxLjnwz37$I4 zU*DAZeJtYK^;nFHObUnmd1#Z<(e2&3&)I5oy=uA6C@KguCCTBhGhRiz5mvDaH17h3 zX>t{p7UgYqz_D}GoTpOmSM(rO*t$)vzdZXdyZ5>Qtb}gc(STmj!303%XK$`?2(N9k zztULEchPiI_HLf~=^0(Q6`ubjXW{#G0r3}A0CME_1)L3_x5NiAY8YSvr70J_mbaWwNlDYkDEr@2)B?#=mZrSWQ`)BomZ-Ubk@IB~CCWjZJO z+knftLx541JOH(=ESu5#gch&)R+slR9#N;RX*XRZRI*Z&UN$FAM?hCYgkE5s)q0++ z=N|tVT)RhuXOcUgQ&#}A4V{~0dRPx|Oq0arRMa`5x2~<6nD@^sK)?(g>RN(OAEe#v zn0oiv^4rDSZ}=v?q+F>HMITg;*!B|fU{ljp$4U-}aA4f9?;vAw+LP*okbfg%`R`B5 z@_C&u?lE{hh9A(kUdN2cEu|x#a^Iii*+@R=*gpMbUC5#My(=nB+Ow{)nfLAvf;V;L zb6?qq#r*u|^k|XIr@B#i4?zMb4Rh2f&_aIVvfq|y4Z30TX?5Evic5?XU2--J6NyUW z)u^*vY?eKv$m6pErb0EcJUk@#(XUOed2PaJ%)+5Bf>5={B>)$wqWqJk`{~1iyIYL{ ztP(_dR4%fSDS&#W7xr{dPvS!2LTIH#3ZECw_|bfMze90+si73A8ILs}R+9A5%WjA& z_puQO6j2G18G!Sx>hyH&|L^+%%8`dp7`Vlv41knFv8*657Z=xoV+`K7*LWIlV%1gx zKJ+N6*43vEI;YF^<4*eTa`87OGKB-z(*SN`my69_#q^XWqv7~QhrPKX&EoT6CW^l- zq3;5pCb~g!)XMgX2++`WsrpRXYk>y+k%Q`3A68iro0fWpx3>-FL;PaL1_lZ=&qCF# zi0a{aBgo0gUt6~Jgmz8X{&B$G!+`(j*h&2M?KCQj62c~N)q0+Jrc4`Gnbl+*#A|j* z>ZQLyQKEA6>h%QMo(fc5PnPS&qq1C9y!(NI6ioXk2ZjF;+5G#O{KNqc##^FJx)TJC zOZ?VP=)-eB7~7MpBYdKMi%cSx5%*XipWU>!P%8oPNLek9R!H=4Z?}a!ox*SPT2Pto zSre!Q|Cg}<`?8^MPXDv6aABFr^WK*B>BMDSKf|{5C!jV+r(5X>*dsKws$)Z;38i+3 zm&fgBGQZ75R&TgjeA$Qo(%gSEn)3_++u1qGDcdE*>9pjBm4Fb`czWGPj@<-Z??#t( z9oJT%{bCY5ix+CIJS#7 z++QA3uEXPxOCOIXQ~c)@<_o5S47hZf653yX86mXuG{7IPRT~Zd^9Z74_^ST#$1pUo zf0@L05G@qnE_EhplYn#c1c5v2)t6dxuU6dlrGoqIGg`_v-A>0%@BFzQJH*ErvjmS>r%dD-m(IL z_xV}IxW{b%-NqH%XenTP?pVsb$1F*yVrfo%DDC#FV!s&*7mcGMWbv{O8|;lI+JN}g zGjBOS6e`U{(Bsz!hkmq^Uk=n=nNTKU2bhV;2N$JXKSBl3#iA01!TvR^9sXc#u+FZw z+3IKzl(x$6dssA`+vt=Ff8{9+VIZrHo)yAjL_ZGiD*ryDYChLvm{rsqxBp7+n`r#@ zQ9PYhb1aD#TWsB{O%4BcHVm_Gp{hC%ifZw*?STS?Z{qHa7Z;i4R4koYl34P19si8p zTGMIs{xL(ct(xof5%ZlSjgXYX%!K9;=lBMFPm;Hp{m-{?rT{wEw`$lRn64x;xJ<#e z`jyrW6x*_w{SVm#Or}0YJ3qkl8E)|zdNMbYA-*>rH_=s$Nm&1DBhl_~sZV{RP6BMn z^OG+QXpxvlxX@NGWLzoBvqG#fRPM&J3B>@z>>g}DTpiVxG6*jFqGd3}5L)_1PmQ@t zum5YQjJSH^G_?hVir>?bQcAF~4ueSw^cQJnWAw!NSmthgA~mY-$p|OT2~KW2W}1mr zNNUjX)7`0qlaaA{=H_}B(BF9u%tZp2IVCGYnNU?>zGq|LHTy1sIF(teBQzZF)?a+L zb$PQHaDgnMle^^%1a|L|(CGZfKyu}M+qvWVviq$5m(#Cq6_S4UjyUd%h@jHH*^)pK z#y7QUDMIg4p1&CI*KEKraxV((lK*1Ba(NIy{yF_4o%?T|c8ap2!q|+*e8u=L2K>tm z*dvCofA5X`#eg$dU>Hi&aiVbQ0qgq5EzU>-NMV`0?i;WE`ExYN2w;zh|HonZ|8*$7 zQAgm8{+XKZQX8pcXDJ+Z#Oe5G42qS zg)vC`#lRlFS-gd!)7%~}IsmHlMaso&rW4=p&wncyDO0%(s?zDrM{57!%VwR|nyz*k ze`c~-ZaNvJZ?{~kF+1~ay&B$%KG>5<*KBdMGd116hu*rPorpK{00bIVi~|D#W{OqI zwVR#Msg-|L1BLPD&X0tYp_f3rF6h-*{j3xOi#%nz%;%A$;L+nf&m*;aX7-x3S)1M6dfKv z`nLPcKy~`NZUbCP%S(Ize zJL?>2kawjf&fY%O!C5p_F=z|9ILToqlM@8#t-dB4dk^mw! z_w|7%7$}QZ;C%)h=7XDC@aK3OcHvaekO`u7)lesi4F?c^``__n@fb4fa^WE)mCB~` zMI^J%U76$G;ss@uoMwrwjY7J=~iRDe!b(lI4poSbC{+##O8N zG_OT558#h)Q_owpyC zZvPpn`Jv~#YN>jh$LIV%QL*ixR2M#g9a>;8V{1CQ@LH%8ZGj_#f`?GCOn`txtv8-?9S`@oRC*~zqoT6 zy-*w>zpoTiyMaq)w-IYH8TYiPIp9pn5Fmn(J2(K+x8-IV`b(0E1z_>m?+pExdOGuY zd+Ja3`E>dHYoILOiy@KZl}=qjK|#@l^48{NOMPfoJkRA?^lY(ezXDq_9VaI?y37U5 z@y|?QIfrUM2EplYz7vv=+xr*@ARXZ?UIBpJV<2>e`(pOorPpUFtje3qO-83b5{%Dz=g0XnP0biWIiIJk zdXy}P4RqBm-px(GVi(Z>i?t+hIc5Eq%h@LKZAMVPf#kJmS{vGpj*F#~gYqk4Et)G= zB}67u8ky_{0nlJ{4AF?jKM6uf$dH!1+7%FNchZXRs7tmYHH|1x`epx!`t0TbT?GK( z4BVv@buV)hbc6yvv(h;MTH?X>httTx;_xCfdgUsm8r8<5T1U89yxeO<@;{I%7{=5b z@q-`r3q9X%$dl!Mk;(J2r>ZtK4t4h>FY_cpDN$Xpt zQg+N#LkgE+@RXu2DSWfks2YbCiPLCZlyH83-T9kQYcD_5y7d%y{lZ#C&QA*xQIB6e zK;luMGLfB8*e`TKOT*{T@yn>1RWm(K{G&T)p z>NZio0|xwy^+cJZzQ<-_Z2~Et;#mz*?#2}RCEl%T2e0N&cjr%by`Sb%&Lx5J8=a9zrIkb0NZc zG@SD&7$71AT3r{LJctt&mNVRV1F3bh{kNbsO<6~CnXEYhI=rM4=KdESh61g`djN^* z7G<4?AlW!_!F21MN+~xqYf~&P0Q%lLxR(EbN~-QatNn^os8YOmekzUIDj7aL`3oLp)@ z%@cgHX8HkcKm&go$bBh-!i(INrCKynPvb+U&ec@G$U4%<64wi|vlJB|dp!{Z^gsSB z&>!x;wt;u1_Cv73=0=86xMR!d?k06L2ZwCtPhHqthEe?G?T6x?OpapC zGki5c$eRzU4nbPScLoSd8#bdKz*`sFhHKD`LjbNJ6J7?54?; zY85Q4YP?Mg5QkT>*cb*sT|EV5gHY1hgr*F)bp~!#F7k56%UmVIZn1yo<=NhDjo1l^ z^!x$TD#Lf#=d5M=2P0kR&*G4&?KyqbTw^w-L|l6Yis-1%ZMY;7L9?M2{RCs}2Zv3G zDpH@71p5=HuaJnLO6}=k0jXPUmkLjtAnRM|-T7J-D!o;Zc(NBImm!07#(NfxOPd>3 zH=ZC;;XUM&-@rOzi{mW#E$Vrm;HK~G5s~th93$s#v;1i5w%~T`U}AJnesufZ^~ols zxpG(_m(ShJ*^sUEx)nG`p6cZBCfOvOQeN=gE$-%kak{{0cytvJl}&~S=dIz4CwBmm zVQBm9CBb||%p2@u!4Qs=0ibAbxnU36Dg%{;X8pMA!r^|&MLv&A2@U~wCL5?)dBtk_&fS#RjNSfZf7+Zy8K^!_V_5>3y*jDYl({%= zw9o1G_8yH@TNVflP-de~^#=$QYu<{eSL(V^05v?6wiWkNu%g;ep+hN@j=E4YCxHa? zpqH9}B+A&z-j*KKa0#kaR41i9wDy1vpmqA(x4(8>Z92t(m{#*Y1w^O!XiiQz6UdD= ztz2({xCSSz+lABBpU*WlgY~acy6<62@`P-oec&+gFj9TXw1feTU75ro&95bfdz4C?n(-__N4tv>#_?Dm7ia|t$&WB7h_1f_E;?F_YvcKWkn4-x()pP4Dk1z_W^8AX(K`&0HUNAbGT=%T4nA?lZTQoFwrY;K zNv+tZnnfBso=k1Pv~9Om&;N!1iMkqX-kYQx%wk*8i}uu-_8acDKE^=$DM`Ev?;vhP z6=6h=n7j26S;xPr<#b_Tt1*SJyeVdjBpHu}WFz0RWP@4X*&@JWNPGydZJGEayVq`sy4|`U zb_{n?N=k$+ff%eRu9-3EQ@>Cpm;8ZlPFldbwU89N;9Y`zWvH?;%RR&b>bsEL1k)zPJddCUu-AvtlpCziYWP<6&q8=dnej3MD zJRyFTqZlvJ^6thIf&VTl8s+lsDI?3Xd(Fu=+joo7%oS)67IVj&EP;=Wm`z?}MB8mu zS+-q!V93#PCB!SX=M3;3Ed(6(;`F}6Q^$JWQE6>#AOd7w*uTgwV|!rJaPk8ZAM%qu z6?fV}S+k%U(=7u(emFuh#xn{C0V%->&+-%%e2jLV_3a#P0jrfKd*<*!m=KThw5R*5 zSwQ-9Vo$B(_b%8mS>^F=LGk@~p{AQ!+xZu_O@d?eAGGO7eZ=0nEHrrdr4So3ro;%< z9efkoZC8Pq64+z_TYLce)D+>hbr1JvStGDvivw^FqfFITVo23uB&)iiQx%KAB@l!&R1_ii637rUJtV7bjZ2c@ifo8 z?<9h1%bRD~Qp8B?-A9kR2(#i>)U?E(c6@>tN@eMqZkulaWtz0N44{)TZLFHwWr?^Q zFH%%i-hM8%s_z-DMh+sCLy=mOe$teZ`sjoqTn95pEDjFk6Y(l{(#8eG4Bo5<( z&?b04@6S|XKEHnF!5`yDR!nob_i8^6&bFP77-h482%R;FJ`}~`(4UMLeAhX0eX2S& z(UWKhhZ@&7tX@_<*DWi0ofjb->!H^Ke^y8zeC;u-jaj0!gL#{~tUjz%w)%Wi0bFlbUM5nmcKlr+Gm%JS+Plom zQcV?6fwOxxPFY$?+Q-8$nK(GqG>6`BL8WM$pH7=^LFqnt*=k`v$BKs0R|OR;%mMg5 zqk^Gf8cb2zI?Zl6HC%oD3Qs<>xGLs%Wt5o;p^vIu4K5&W87Gkar-C2AINs#rREqN` z2VyWuxw9^g2))`)STM z2HIzT4YV$cZL@V(PG`Ho(nYzx` z?E>*sui`X4>oLoA)d!xY+sRZbCN}KLlby&L`LaIlkF|n)XPCfC|0)<-=(MiZD7i4I z^PaE3u+to+I>k$InOK@{ZTNuO`me6mW9#`aas+W zW+k_R$E=pGY}nx}`se`!i%+M*c_)U>bc`j-Xa3S)Kq;7-O z?VBK3QV1bO24k2rGnW^FY8Ik~eKM*QsD3NSV$T>5mUKU_o<8sxT-Z>Xn(BW}qmBVc zE-7+NmDc@v<{0&+3JRc_&wYxStPZie;%D=2wR_Abk&@4C?W-;oBfAT}8V8m{0>SDM zvOZ^O|EPPNj!sJBDy#bhSz)Qr(vzz!|MgqoVd$t$|!F#zE`lpXq`QdtB^NSS|h;l0y z3jlbmfGi9JMH86M{t#Sf*fCrd_46KK9Cv)mAY=6WhlQm@qMc!UhZ-1-juY*P9Fk5J z9s-5GCGU5~L5%Y7cQq`5`?v4leZ?bemNONWN`@Y0Mo5CCC)MWfy!Z%lFTW_hoK(n% z?ePAng)L2J)k480*K{Fi&iXG?hTj|@j-eCNfh`&@Dp@(7SV3}XzfL0vqg6;Z&wu(; zzTA99*Z|0A7Arm|%WK=&^xU0rt}D9UgB?RE%&cbNvDAUzzg3bi5W)v~<9LwTAm<7L z)H*rqs;12K=YI6F-Py=T*TPQI7JFUc1Sgdh571x6OM0m9aFGzhyhS75j=!?HfB$~( zPv-CyY>0=1K5D-P>`E#pk~>Vk<2rW<0C$+bKGHCVMj9fa%^m-DwVi58xLn$ z!bL%XF#lHz*rUjX?)S%{Kpttu|NG#(nIK znnAY~@^xIsTZ9ocdqL#!L|KvMHmZiW0*Z=C1LD09vE2+OIP}p=8V-f~RHfh974X=K zHs%HW5jLeaomX_ulaF^8pzdI^GvSj zSt99Tvu~;xznSE=>&_E?a4AL^+$lVPUHj$$i1bCn^AJj=)7woAGYt{(R6%!PQ@)); zyseVn=UVlPDt41mJq5_KskYyFS1UKf$J*Y&vB61q@z`#;nYR0mE;CYBvPs%F6nB?9 zys)eI?MI$B^VcVK&Vw~Yg_Z@REE=fsy6zcm>$m7Kj1~xLx0cXP^91cuRc#zRcS+-* zXQwSEDJD+SNaF_UG@n_N*&f%n=(@}HDnGF`#ahup4;ifeEUv-cqCM0@Z>^#eRr=yj zt<6`SG^8-=1fZ^s16v7HDDs_z{9krzI5}VX@8alajXkibJiR7rsNFMQvn!=HfQ;=8 z+V-yjiWkON4KVXIAPtTeNb0O!EGD;VPAe$IRSyVQ>l=H77zKd{WSx~F_zKmtU;9c< zYY(_J-_(sPJ1gHjoh*e_xTEL!L5lp4T}o`9R1H&??Mx$x!Lm-)*}D5Afhuv( zvYTPTCO{LS!t>s$a9eJhzqg#gzTb8RFFl<5M`g&G3ifm@`7kQu{d0UR1_!LW$sG9aXWN?7L8#TZUupi`}yr z6y4Ezx@DH%ms#;9yT(vA-oNL5Yt1Z}=;|{3XuV3%H@Jmk_{ok%B|3XZHP2%5_bT(N z&*RCupV*)G#g=IWR#dJgB$-cLgFYJV1v>>-#k|hrlb%Vds>7%8{hD!kkVI`o6HQLj z-~FuYr`r75HW$llWKwSpD~@`ShRyqDvZk@DymbGBU$&P;%)5hr@xr60)68f z1-)W;LO4ememTh)+mZcb3Dq!gxh%ENOuHh4*ilYtwkwhkaa!4jFMD`zg>(kHBQ4*! z!o2;nGnJJ%1D2Npi)*5OL`C5$9{z^)0@4z4-3D;UVw=VMQ;avFQ3Yp!7y}S9U9X}{ zQVEne2m)0s2K)vbTJ5(m`lzvSpxwqW32b?KD!;ReS?7jti%S{!V}_sOZ2&`);Q%|p zcOH69u^|7i?j}|x+RMsP-FBi|cvB)S)Iuons~x8@j*ySa{DBphVF%Mp`4-*?!9a-7 zcSn6rQOqAY9qSlTm#aOTzZ&g9Csb3M7StU2aKjE*&WL<@t}AUAh1+-B-*s(mK0!lk z@fg9bUfPeU<<@arNf*t{;*nA^=74uv0Aom`peP7PY`FLv8#Groi2zt2HtwZ0tNeT3 zB6Nk+g86twqHuxvU;oe_w}=fJL!xe9*8VNZznaY13C)cOig~``KEbwP`UmpZ8T$6W zJz{@+zm~7FlG7x|q=S&-$TSJn&}*z2Tqzl-M$y0JYkBaQ)9a_^0>F zAOrm1(fd{Jzr8>fd#2tlgbw0Lfvv`70eTUuT3uqF*#*ffOs=Tu;&laovD`L_&FEWJ z$4{gX5{oaH*mhT(b3akX| z`W^!eilXN`O>QnYks}tqD{b=z#Q7!9d{Q1s#M|Md1=#L2(KTBco!{Eqy>-#4jw|W* z(FYaL#volwY)L6B4@|LSqNEy~l)!w)yE2MkBBehZvV@Cv-8i-X;U|&=KlsX+!P2;G z`k#3w`JiJUERw;95F0}sV2;6-La{Xh$D4v3L~_stnm3f%w%<6eDXb1mAJh`kEy^z1 zhLz+Z9^78nhDz>SGwp+vBdtbArmBpiV}4|rk2Y5|>r755(ynHrna?V%WdFqk$XejD zl^DMWawiy8=qUSJpf!f2gXmMtR`0^!O@Dy30>GLp#uc1D6Z)Ui>A!td0wSYo2eV;7 zeEj#)*{^HZ37wddB|$ZT#a2|6iWQEq!jNN_-P!mJ1E|JXX?Bgv(pJ F{tt@NtzQ5D literal 0 HcmV?d00001 diff --git a/examples_deepspeed/zero_bubble_pp/zbh1_pretrain_gpt_1.3b.sh b/examples_deepspeed/zero_bubble_pp/zbh1_pretrain_gpt_1.3b.sh new file mode 100644 index 00000000000..cf5705d9736 --- /dev/null +++ b/examples_deepspeed/zero_bubble_pp/zbh1_pretrain_gpt_1.3b.sh @@ -0,0 +1,367 @@ +#!/bin/bash +dir=`pwd` +############################################################################### +### Main configs +## GPT-3 models use 2K sequence length/context window +seq_len=2048 + + +## The "GPT-3 XXX" below are configs from GPT-3 paper +## https://arxiv.org/abs/2005.14165, choose based on +## your desired model size or build your own configs + + +## init_std is standard deviation for weight initialization. Usually larger +## model needs lower std. We used a heuristic equation of sqrt(1/3/hidden_size) +## from the MT-NLG 530B work (https://arxiv.org/pdf/2201.11990.pdf) + + +## We changed min_lr to a lower number (1.0e-6), which we found is able to +## provide better zero-shot eval results. + + +## GPT-3 Small 125M +# model_size=0.125 +# num_layers=12 +# hidden_size=768 +# num_attn_heads=12 +# global_batch_size=256 +# lr=6.0e-4 +# min_lr=1.0e-6 +# init_std=0.02 + + +## GPT-3 Medium 350M +# model_size=0.35 +# num_layers=24 +# hidden_size=1024 +# num_attn_heads=16 +# global_batch_size=256 +# lr=3.0e-4 +# min_lr=1.0e-6 +# init_std=0.018 + + +## GPT-3 Large 760M +# model_size=0.76 +# num_layers=24 +# hidden_size=1536 +# num_attn_heads=16 +# global_batch_size=256 +# lr=2.5e-4 +# min_lr=1.0e-6 +# init_std=0.015 + + +## GPT-3 XL 1.3B +model_size=1.3 +num_layers=24 +hidden_size=2048 +num_attn_heads=16 +global_batch_size=16 +lr=2.0e-4 +min_lr=1.0e-6 +init_std=0.013 + + +## GPT-3 2.7B +# model_size=2.7 +# num_layers=32 +# hidden_size=2560 +# num_attn_heads=32 +# global_batch_size=512 +# lr=1.6e-4 +# min_lr=1.0e-6 +# init_std=0.011 + + +## GPT-3 6.7B +# model_size=6.7 +# num_layers=32 +# hidden_size=4096 +# num_attn_heads=32 +# global_batch_size=1024 +# lr=1.2e-4 +# min_lr=1.0e-6 +# init_std=0.009 + + +## GPT-3 13B +# model_size=13 +# num_layers=40 +# hidden_size=5120 +# num_attn_heads=40 +# global_batch_size=1024 +# lr=1.0e-4 +# min_lr=1.0e-6 +# init_std=0.008 + + +## GPT-3 175B +# model_size=175 +# num_layers=96 +# hidden_size=12288 +# num_attn_heads=96 +# global_batch_size=1536 +# lr=0.6e-4 +# min_lr=1.0e-6 +# init_std=0.005 +############################################################################### +### Training duration configs +## The main termination condition, original GPT-3 paper trains for 300B tokens. +train_tokens_in_billion=300 +train_tokens=$((${train_tokens_in_billion} * 1000000000)) + + +## train_samples is another termination condition and also affect the number of +## data samples to be indexed. Since we want to reach the train_tokens +## above, and data efficiency techniques may change num tokens in some samples, +## so we just set this config large enough to make sure we have enough +## processed data and don't terminate by train_samples. +train_samples=$(( 300 * 1000000000 * 2 / ${seq_len} )) + + +## Another wall-clock time termination condition in minutes. Set it large +## enough to avoid undesired early termination. +exit_duration=30000000 +############################################################################### +### lr configs +## lr warmup and decay duration. +## Original GPT-3 paper uses 375M warmup tokens and 260B cosine decay tokens. +## Here we increase the warmup tokens to 3B since when batch size warmup is not +## used, there are more tokens per step. Thus we need to increase warmup tokens +## to make sure there are enough warmup steps, which is important for training +## stability. +lr_warmup_tokens_in_million=3000 +lr_warmup_tokens=$((${lr_warmup_tokens_in_million} * 1000000)) +## Here we changed the LR decay tokens to align with total train tokens, since +## related works (e.g., https://arxiv.org/abs/2203.15556) find that setting the +## learning rate schedule to match the number of training tokens results in the +## best final model quality +lr_decay_tokens_in_billion=${train_tokens_in_billion} +lr_decay_tokens=$((${lr_decay_tokens_in_billion} * 1000000000)) +lr_decay_style="cosine" +############################################################################### +### Parallelism configs +## Model parallelism, 1 is no MP +mp_size=1 + + +## Pipeline parallelism. To disable PP, set pp_size to 1 and no_pp to true. +## Note that currently both curriculum learning and random-LTD are NOT +## compatible with pipeline parallelism. +pp_size=8 +no_pp="false" + + +## ZeRO-based data parallelism, stage=0 will disable ZeRO +zero_stage=0 + + +## Total number of GPUs. ds_ssh is from DeepSpeed library. +num_gpus=$(($(ds_ssh nvidia-smi --query-gpu=name --format=csv,noheader | wc -l)-2)) +num_gpus_pernode=$(nvidia-smi --query-gpu=name --format=csv,noheader | wc -l) +num_node=$(( ${num_gpus} / ${num_gpus_pernode} )) + + +## Data parallel size. +dp_size=$(( ${num_gpus} / ${pp_size} / ${mp_size} )) + + +## Micro batch size per GPU +## Make sure that batch_size <= global_batch_size*pp_size*mp_size/num_gpus +## Reduce it manually if GPU OOM +# batch_size=$(( ${global_batch_size} / ${dp_size} )) +batch_size=1 +############################################################################### +### Misc configs +log_interval=1 +eval_iters=10100 +eval_interval=10100 +# num_save controls how frequent to save checkpoint. num_save=20 means that a +# checkpoint will be saved every 5% of training. For longer training you would +# want larger num_save to save more frequently, and vice versa. +num_save=1 +# estimated_train_iter=$((${train_tokens} / ${seq_len} / ${global_batch_size})) +# save_interval=$((${estimated_train_iter} / ${num_save})) +save_interval=10100 + + +## Activation checkpointing saves GPU memory, but reduces training speed +activation_checkpoint="false" +# activation_checkpoint="false" + + +## Whether or not log optimizer states (norms, max abs values) to tensorboard. +## This is not required for training and might save GPU memory when turned off. +log_optimizer_state="true" +############################################################################### +### Output and data configs +current_time=$(date "+%Y.%m.%d_%H.%M.%S") +host="${HOSTNAME}" +seed=1234 +num_workers=0 + + +## Public the Pile dataset, can be downloaded at +## https://mystic.the-eye.eu/public/AI/pile_neox/ or +## https://the-eye.eu/public/AI/pile_neox/ Change data_home to where you +## store the pile_text_document.bin and pile_text_document.idx. +data_home="/code" +data_path="${data_home}/gpt_data/my-gpt2_text_document" + + +vocab_path="gpt2-vocab.json" +if [ ! -f "$vocab_path" ]; then + wget https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json +fi +merge_path="gpt2-merges.txt" +if [ ! -f "$merge_path" ]; then + wget https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt +fi + + +prescale_grad="true" +jobname="gpt_${model_size}B_tok${train_tokens_in_billion}B" +jobname="${jobname}_lr${lr}_min${min_lr}_w${lr_warmup_tokens_in_million}M_d${lr_decay_tokens_in_billion}B_${lr_decay_style}" +jobname="${jobname}_gbs${global_batch_size}_mbs${batch_size}_g${num_gpus}" +if [[ $zero_stage -gt 0 ]]; then + jobname="${jobname}_z${zero_stage}" + prescale_grad="false" +fi +if [[ $mp_size -gt 1 ]]; then + jobname="${jobname}_mp${mp_size}" +fi +if [ "${no_pp}" = "false" ]; then + jobname="${jobname}_pp${pp_size}" +fi +jobname="${jobname}_seed${seed}_rebase" + + +username=$(whoami) +output_home="/blob/users/${username}/project/data_efficient_gpt" +log_path="${output_home}/log/" +checkpoint_path="${output_home}/checkpoint/${jobname}" +## Microsoft internal constraint: because tensorboard is logged by last rank, +## it's better to put the path in NFS instead of Blob. +tensorboard_dir="/vc_data/users/${username}/project/data_efficient_gpt/tensorboard/" +tensorboard_path="${tensorboard_dir}${jobname}_${host}_${current_time}" +mkdir -p ${log_path} +mkdir -p ${checkpoint_path} +mkdir -p ${tensorboard_path} +############################################################################### +data_options=" \ + --vocab-file ${vocab_path} \ + --merge-file ${merge_path} \ + --data-path ${data_path} \ + --data-impl mmap" + + +## If CL is used, make sure to set "--split" the same as what you used during +## offline data analysis&indexing. +megatron_options=" \ + --override-opt_param-scheduler \ + --adam-beta1 0.9 \ + --adam-beta2 0.95 \ + --tensor-model-parallel-size ${mp_size} \ + --init-method-std ${init_std} \ + --lr-decay-tokens ${lr_decay_tokens} \ + --lr-warmup-tokens ${lr_warmup_tokens} \ + --micro-batch-size ${batch_size} \ + --exit-duration-in-mins ${exit_duration} \ + --global-batch-size ${global_batch_size} \ + --num-layers ${num_layers} \ + --hidden-size ${hidden_size} \ + --num-attention-heads ${num_attn_heads} \ + --seq-length ${seq_len} \ + --max-position-embeddings ${seq_len} \ + --train-tokens ${train_tokens} \ + --train-samples ${train_samples} \ + --lr ${lr} \ + --min-lr ${min_lr} \ + --lr-decay-style ${lr_decay_style} \ + --split 949,50,1 \ + --log-interval ${log_interval} \ + --eval-interval ${eval_interval} \ + --eval-iters ${eval_iters} \ + --save-interval ${save_interval} \ + --weight-decay 0.1 \ + --clip-grad 1.0 \ + --hysteresis 2 \ + --num-workers ${num_workers} \ + --fp16 \ + --seed ${seed} \ + --load ${checkpoint_path} \ + --save ${checkpoint_path} \ + --no-async-tensor-model-parallel-allreduce \ + --tensorboard-queue-size 1 \ + --log-timers-to-tensorboard \ + --log-batch-size-to-tensorboard \ + --log-validation-ppl-to-tensorboard \ + --tensorboard-dir ${tensorboard_path}" + + +if [ "${activation_checkpoint}" = "true" ]; then +megatron_options="${megatron_options} \ + --checkpoint-activations" +fi + + +if [ "${log_optimizer_state}" = "true" ]; then +megatron_options="${megatron_options} \ + --log-optimizer-states-to-tensorboard" +fi + + +config_json="ds_config_gbs${global_batch_size}_mbs${batch_size}_log${log_interval}_zero${zero_stage}.json" +template_json="../rebase/ds_config_gpt_TEMPLATE.json" +sed "s/GBSIZE/${global_batch_size}/" ${template_json} \ + | sed "s/MBSIZE/${batch_size}/" \ + | sed "s/LOG_INTERVAL/${log_interval}/" \ + | sed "s/ZERO_STAGE/${zero_stage}/" \ + | sed "s/PRESCALE_GRAD/${prescale_grad}/" \ + > ${config_json} + + +deepspeed_options=" \ + --deepspeed \ + --deepspeed_config ${config_json} \ + --zero-stage ${zero_stage} \ + --enable-zbh1-pipeline \ + --enable-zbh1-exact-semantics \ + --pipeline-model-parallel-size ${pp_size}" + + +if [[ "${no_pp}" = "true" ]]; then +deepspeed_options="${deepspeed_options} \ + --no-pipeline-parallel" +fi + + +if [ "${activation_checkpoint}" = "true" ]; then +deepspeed_options="${deepspeed_options} \ + --deepspeed-activation-checkpointing" +fi + + +## When saving checkpoint to a storage with cache, their could be consistency +## issue of the pointer to latest checkpoint. Here we find the correct pointer +## and broadcast it to all nodes. +iteration_file="$checkpoint_path/latest_checkpointed_iteration.txt" +iteration_file_2="$checkpoint_path/latest" +iteration=0 +for (( node = 0; node <= num_node-1; node++ )) +do + if $(ssh -q worker-"$node" "test -f \"$iteration_file\""); then + local_iteration=$(ssh -q worker-"$node" cat $iteration_file) + iteration=$(( ${local_iteration} > ${iteration} ? ${local_iteration} : ${iteration} )) + fi +done +if [[ $iteration -gt 0 ]]; then + iteration_2="global_step${iteration}" + ds_ssh "echo $iteration > $iteration_file" + ds_ssh "echo $iteration_2 > $iteration_file_2" +fi + + +deepspeed ${dir}/../../pretrain_gpt.py ${megatron_options} ${data_options} ${deepspeed_options} 2>&1 | tee log_zbh1_exact.txt \ No newline at end of file diff --git a/megatron/arguments.py b/megatron/arguments.py index dad993be04d..e7182c317ed 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -73,6 +73,12 @@ def validate_args(args, defaults={}): assert args.world_size % args.tensor_model_parallel_size == 0, 'world size'\ ' ({}) is not divisible by tensor model parallel size ({})'.format( args.world_size, args.tensor_model_parallel_size) + # Zero bubble pipeline is defined on deepspeed's scheduler + if args.enable_zbh1_pipeline: + assert args.deepspeed, 'Use DeepSpeed to use zero-bubble H1 pipeline' + assert args.sequence_parallel == False, "Sequence Parallel not tested, proceed at own will by removing this line" + if args.enable_zbh1_exact_semantics: + assert args.enable_zbh1_pipeline, 'Exact semantics require ZBH1 pipeline enabled' # Pipeline model parallel size. args.pipeline_model_parallel_size = min( args.pipeline_model_parallel_size, @@ -835,6 +841,10 @@ def _add_training_args(parser): 'uniformly divided recompute unit, ' '2) block: the number of individual Transformer layers ' 'to recompute within each pipeline stage.') + group.add_argument('--enable-zbh1-pipeline', action='store_true', + help='Activate zero bubble pipeline parallelism schedule method') + group.add_argument('--enable-zbh1-exact-semantics', action='store_true', + help='Use an exact semantics for zbh1 schedule, might be slower than the default.') # deprecated # HACK: added back arguments because DeepSpeed still relies on the old diff --git a/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py b/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py new file mode 100644 index 00000000000..ba451d97055 --- /dev/null +++ b/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py @@ -0,0 +1,110 @@ +from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore + +from deepspeed.runtime.pipe.engine import PipelineEngine +from deepspeed.utils.timer import BACKWARD_MICRO_TIMER, \ + BACKWARD_GLOBAL_TIMER, BACKWARD_INNER_MICRO_TIMER, BACKWARD_INNER_GLOBAL_TIMER +from deepspeed.runtime.utils import PartitionedTensor +from deepspeed.accelerator import get_accelerator + +import torch +from torch.cuda.amp import custom_bwd +from packaging import version + + +from megatron.core.parallel_state import ( + get_tensor_model_parallel_world_size, + get_tensor_model_parallel_group, + get_global_memory_buffer, +) + +def _exec_backward_only_pass(self, buffer_id): + assert self.optimizer is not None, "must provide optimizer during " \ + "init in order to use backward" + + self.mem_status('BEFORE BWD ONLY', reset_max=True) + from megatron.core.tensor_parallel.layers import LinearWithGradAccumulationAndAsyncCommunication + WeightGradStore.set_combine_bw(False) + # The last stage just runs backward on the loss using DeepSpeed's typical + # mechanisms. + if self.is_last_stage(): + super(PipelineEngine, self).backward(self.loss) + WeightGradStore.flush() + self.mem_status('AFTER BWD ONLY') + + WeightGradStore.set_combine_bw(True) + return + + outputs = self.pipe_buffers['outputs'][buffer_id] + + if self.wall_clock_breakdown(): + self.timers(BACKWARD_MICRO_TIMER).start() + self.timers(BACKWARD_GLOBAL_TIMER).start() + self.timers(BACKWARD_INNER_MICRO_TIMER).start() + self.timers(BACKWARD_INNER_GLOBAL_TIMER).start() + + # Reconstruct if we previously partitioned the output. We must be + # careful to also restore the computational graph of the tensors we partitioned. + if self.is_pipe_partitioned: + if self.is_grad_partitioned: + if self.pipe_partition_output_meta_cache is None: + self.pipe_partition_output_meta_cache = outputs[0].to('cpu') + part_output = PartitionedTensor.from_meta(meta=self.pipe_partition_output_meta_cache, + local_part=outputs[1], + group=self.grid.get_slice_parallel_group()) + self.pipe_buffers['output_tensors'][buffer_id].data = part_output.full() + outputs = (self.pipe_buffers['output_tensors'][buffer_id], *outputs[2:]) + else: + # Already restored from partition + self.pipe_buffers['output_tensors'][buffer_id].data = outputs[0] + outputs = (self.pipe_buffers['output_tensors'][buffer_id], *outputs[1:]) + + grad_tensors = self.grad_layer + if self.is_grad_partitioned: + if self.grad_partition_grad_layer_meta_cache is None: + self.grad_partition_grad_layer_meta_cache = self.grad_layer[0].to('cpu') + part_grad = PartitionedTensor.from_meta(meta=self.grad_partition_grad_layer_meta_cache, + local_part=self.grad_layer[1], + group=self.grid.get_slice_parallel_group()) + grad_tensors = (part_grad.full(), *grad_tensors[2:]) + part_grad = None + + if self.using_bf16_optimizer and not self.is_last_stage(): + # manually call because we don't call optimizer.backward() + self.optimizer.clear_lp_grads() + + # This handles either a single tensor or tuple of tensors. + + if isinstance(outputs, tuple): + out_tensors = [t for t in outputs if t.is_floating_point()] + assert len(out_tensors) == len(grad_tensors) + torch.autograd.backward(tensors=out_tensors, grad_tensors=grad_tensors) + else: + torch.autograd.backward(tensors=(outputs, ), grad_tensors=(grad_tensors, )) + + + WeightGradStore.flush() + + if self.using_bf16_optimizer and not self.is_last_stage(): + # manually call because we don't call optimizer.backward() + self.optimizer.update_hp_grads(clear_lp_grads=False) + + # Free up the memory from the output of forward() + self.pipe_buffers['output_tensors'][buffer_id] = None + self.pipe_buffers['outputs'][buffer_id] = None + grad_tensors = None + + WeightGradStore.set_combine_bw(True) + + if self.wall_clock_breakdown(): + self.timers(BACKWARD_INNER_MICRO_TIMER).stop() + self.timers(BACKWARD_INNER_GLOBAL_TIMER).stop() + self.timers(BACKWARD_MICRO_TIMER).stop() + self.timers(BACKWARD_GLOBAL_TIMER).stop() + +def _exec_weight_pass(self): + if self.using_bf16_optimizer: + # manually call because we don't call optimizer.backward() + self.optimizer.clear_lp_grads() + WeightGradStore.pop() + if self.using_bf16_optimizer: + self.optimizer.update_hp_grads(clear_lp_grads=False) \ No newline at end of file diff --git a/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py b/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py new file mode 100644 index 00000000000..651aadce729 --- /dev/null +++ b/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py @@ -0,0 +1,148 @@ +from deepspeed.runtime.pipe.schedule import PipeSchedule, PipeInstruction, BufferOpInstruction, \ + LoadMicroBatch, RecvActivation, SendActivation, RecvGrad, SendGrad, \ + ForwardPass, BackwardPass, ReduceGrads, ReduceTiedGrads, OptimizerStep +from megatron import get_args + +class ZeroBubbleH1Pipeline(PipeSchedule): + """A schedule for training a batch using hybrid parallelism. + + Pipeline parallelism is extracted through gradient accumulation and thus + convergence follows that of a data parallel approach with the same batch + size. + """ + + def steps(self): + num_warmup_microbatches = self.stages - self.stage_id + + forward = 0 + backward = 0 + weight = 0 + + # F section + for _ in range(num_warmup_microbatches - 1): + if forward == self.micro_batches: + continue + forward_id = self.get_buffer_id(forward) + forward += 1 + + cmds = [] + if not self.is_first_stage: + cmds.append(RecvActivation(forward_id)) + if self.is_first_stage or self.is_last_stage: + cmds.append(LoadMicroBatch(forward_id)) + cmds.append(ForwardPass(forward_id)) + if not self.is_last_stage: + cmds.append(SendActivation(forward_id)) + yield cmds + + # FB section + for _ in range(self.stage_id): + if forward == self.micro_batches: + continue + forward_id = self.get_buffer_id(forward) + backward_id = self.get_buffer_id(backward) + forward += 1 + backward += 1 + + cmds = [] + if not self.is_first_stage: + cmds.append(RecvActivation(forward_id)) + if self.is_first_stage or self.is_last_stage: + cmds.append(LoadMicroBatch(forward_id)) + cmds.append(ForwardPass(forward_id)) + if not self.is_last_stage: + cmds.append(RecvGrad(backward_id)) + cmds.append(SendActivation(forward_id)) + cmds.append(BackwardOnlyPass(backward_id)) + if not self.is_first_stage: + cmds.append(SendGrad(backward_id)) + yield cmds + + # FBW section + while forward < self.micro_batches: + forward_id = self.get_buffer_id(forward) + backward_id = self.get_buffer_id(backward) + forward += 1 + backward += 1 + weight += 1 + + cmds = [] + if not self.is_first_stage: + cmds.append(RecvActivation(forward_id)) + if self.is_first_stage or self.is_last_stage: + cmds.append(LoadMicroBatch(forward_id)) + cmds.append(ForwardPass(forward_id)) + if not self.is_last_stage: + cmds.append(RecvGrad(backward_id)) + cmds.append(SendActivation(forward_id)) + if self.is_first_stage: + cmds.append(BackwardPass(backward_id)) + elif forward == self.micro_batches: + cmds.append(BackwardOnlyPass(backward_id)) + cmds.append(SendGrad(backward_id)) + cmds.append(WeightPass()) + else: + if get_args().enable_zbh1_exact_semantics: + cmds.append(BackwardOnlyPass(backward_id)) + cmds.append(SendGrad(backward_id)) + cmds.append(WeightPass()) + else: + cmds.append(BackwardPass(backward_id)) + cmds.append(SendGrad(backward_id)) + yield cmds + + #BW section + while backward < self.micro_batches: + backward_id = self.get_buffer_id(backward) + backward += 1 + weight += 1 + + cmds = [] + if not self.is_last_stage: + cmds.append(RecvGrad(backward_id)) + if self.is_first_stage: + cmds.append(BackwardPass(backward_id)) + else: + cmds.append(BackwardOnlyPass(backward_id)) + cmds.append(SendGrad(backward_id)) + cmds.append(WeightPass()) + yield cmds + + #W section + while weight < self.micro_batches: + weight += 1 + yield [WeightPass()] + + yield [ReduceTiedGrads(), ReduceGrads(), OptimizerStep()] + + def get_buffer_id(self, microbatch_id): + num_warmup_microbatches = self.stages - self.stage_id + return microbatch_id % num_warmup_microbatches + + +##Additional Instruction classes +class BackwardOnlyPass(BufferOpInstruction): + """Compute a backward pass and accumulate gradients. + + Roughly: + + .. code-block:: python + + outputs = buffers['outputs'][buffer_id] + gradients = buffers['gradients'][buffer_id] + torch.autograd.backward(tensors=outputs, + grad_tensors=gradients, inputs = input_tensor) + """ + pass + +class WeightPass(PipeInstruction): + """Compute a weight pass and accumulate gradients. + + Roughly: + + .. code-block:: python + + torch.autograd.backward(tensors=outputs, + grad_tensors=gradients, inputs = model.parameters()) + """ + pass diff --git a/megatron/core/tensor_parallel/layers.py b/megatron/core/tensor_parallel/layers.py index 020d25915ab..2245113c9cf 100644 --- a/megatron/core/tensor_parallel/layers.py +++ b/megatron/core/tensor_parallel/layers.py @@ -16,6 +16,8 @@ from torch.cuda.amp import custom_fwd, custom_bwd +from megatron import get_args + from megatron.core.model_parallel_config import ModelParallelConfig from megatron.core.parallel_state import ( @@ -233,6 +235,11 @@ def __init__(self, sequence_length, embedding_dim): def forward(self, position_ids): return self.local_embeddings(position_ids - self.offset) +def gradientUpdateFunction(total_input, grad_output, weight): + if weight.grad == None: + weight.grad = grad_output.t().matmul(total_input) + else: + weight.grad += grad_output.t().matmul(total_input) class LinearWithGradAccumulationAndAsyncCommunication(torch.autograd.Function): """See linear_with_grad_accumulation_and_async_allreduce""" @@ -359,7 +366,9 @@ def backward(ctx, grad_output): # grad_weight = None # else: # grad_weight = grad_output.t().matmul(total_input) - grad_weight = grad_output.t().matmul(total_input) + from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore + WeightGradStore.put(total_input, grad_output, weight, gradientUpdateFunction) + grad_weight = None grad_bias = grad_output.sum(dim=0) if use_bias else None if ctx.sequence_parallel: diff --git a/megatron/core/tensor_parallel/weight_grad_store.py b/megatron/core/tensor_parallel/weight_grad_store.py new file mode 100644 index 00000000000..bbd1aea533c --- /dev/null +++ b/megatron/core/tensor_parallel/weight_grad_store.py @@ -0,0 +1,34 @@ +import queue + +class WeightGradStore: + + cache = [] + weight_grad_queue = queue.Queue() + combine_bw = True + + @classmethod + def set_combine_bw(cls, combine_bw): + # For the following backward pass, combine W with B and skip next W. + cls.combine_bw = combine_bw + + @classmethod + def put(cls, total_input, grad_output, weight, func): + if cls.combine_bw == True: + func(total_input, grad_output, weight) + return + # Store the weight gradient computation of linear layers. + cls.cache.append((total_input, grad_output, weight, func)) + + @classmethod + def flush(cls): + # Collect all stored computations during backward as a W. + cls.weight_grad_queue.put(cls.cache) + cls.cache = [] + + @classmethod + def pop(cls): + # Execute a single W. + assert cls.weight_grad_queue.qsize() > 0 + stored_grads = cls.weight_grad_queue.get() + for total_input, grad_output, weight, func in stored_grads: + func(total_input, grad_output, weight) \ No newline at end of file diff --git a/megatron/initialize.py b/megatron/initialize.py index 85c37a57675..2a9cb3d90fe 100644 --- a/megatron/initialize.py +++ b/megatron/initialize.py @@ -15,6 +15,8 @@ from megatron import get_args from megatron import get_tensorboard_writer from megatron.core import mpu, tensor_parallel +from megatron.core.pipeline_parallel.deepspeed_zbh1_engine import _exec_backward_only_pass, _exec_weight_pass +from megatron.core.pipeline_parallel.deepspeed_zbh1_schedule import BackwardOnlyPass, WeightPass, ZeroBubbleH1Pipeline from megatron.arguments import (parse_args, validate_args) from megatron.checkpointing import load_args_from_checkpoint from megatron.global_vars import set_global_variables @@ -216,6 +218,14 @@ def _initialize_distributed(): get_accelerator().set_device(device) # only do so when device_count > 0 + if args.enable_zbh1_pipeline: + deepspeed.runtime.pipe.schedule.TrainSchedule = ZeroBubbleH1Pipeline + deepspeed.runtime.pipe.engine.PipelineEngine._INSTRUCTION_MAP.update( + { + BackwardOnlyPass: _exec_backward_only_pass, + WeightPass: _exec_weight_pass, + } + ) # Call the init process if args.deepspeed or args.ds_inference: deepspeed.init_distributed() From f2d7589720c216b8e0e676e83bc5a0bc3ab71eca Mon Sep 17 00:00:00 2001 From: Xinyu Chen Date: Thu, 27 Jun 2024 21:38:47 +0800 Subject: [PATCH 253/590] Fix ParallelMLP and enable accelerator test (#403) --- tests/models/test_gpt_embedding.py | 12 +++++++----- tests/transformer/test_parallel_mlp.py | 27 +++++++++++++------------- 2 files changed, 21 insertions(+), 18 deletions(-) diff --git a/tests/models/test_gpt_embedding.py b/tests/models/test_gpt_embedding.py index 700990adc2b..28f8d52f15a 100644 --- a/tests/models/test_gpt_embedding.py +++ b/tests/models/test_gpt_embedding.py @@ -7,6 +7,8 @@ from megatron.core.transformer.transformer_config import TransformerConfig from megatron.core.models.gpt.gpt_embedding import GPTEmbedding +from deepspeed.accelerator import get_accelerator +device_name = get_accelerator().device_name() @pytest.fixture def gpt_embedding(transformer_config): @@ -36,12 +38,12 @@ def test_cpu_forward(self, gpt_embedding: GPTEmbedding): assert embeddings.shape[1] == input_ids.shape[0] assert embeddings.shape[2] == gpt_embedding.config.hidden_size - def test_gpu_forward(self, gpt_embedding: GPTEmbedding): - gpt_embedding.cuda() - input_ids = torch.tensor([0, 1, 2, 3], dtype=torch.int64).repeat((2, 1)).cuda() - position_ids = torch.tensor([0, 1, 2, 3], dtype=torch.int64).repeat((2, 1)).cuda() + def test_accelerator_forward(self, gpt_embedding: GPTEmbedding): + gpt_embedding.to(device_name) + input_ids = torch.tensor([0, 1, 2, 3], dtype=torch.int64).repeat((2, 1)).to(device_name) + position_ids = torch.tensor([0, 1, 2, 3], dtype=torch.int64).repeat((2, 1)).to(device_name) embeddings = gpt_embedding(input_ids, position_ids) - assert embeddings.device.type == 'cuda' + assert embeddings.device.type == device_name assert embeddings.shape[0] == gpt_embedding.max_sequence_length assert embeddings.shape[1] == input_ids.shape[0] assert embeddings.shape[2] == gpt_embedding.config.hidden_size diff --git a/tests/transformer/test_parallel_mlp.py b/tests/transformer/test_parallel_mlp.py index f43dc0b4678..f9fa365c686 100644 --- a/tests/transformer/test_parallel_mlp.py +++ b/tests/transformer/test_parallel_mlp.py @@ -6,6 +6,8 @@ from megatron.core.transformer.parallel_mlp import ParallelMLP +from deepspeed.accelerator import get_accelerator +device_name = get_accelerator().device_name() @pytest.fixture def mlp(transformer_config): @@ -19,28 +21,27 @@ def test_constructor(self, mlp): num_weights = sum([p.numel() for p in mlp.parameters()]) assert num_weights == 1212 - def test_cpu_forward(self, mlp): + def test_cpu_forward(self, mlp, transformer_config): # [sequence length, micro batch size, hidden size] - hidden_states = torch.ones((32, 2, mlp.config.hidden_size)) + hidden_states = torch.ones((32, 2, transformer_config.hidden_size)) output, output_bias = mlp(hidden_states) assert output.shape[0] == 32 assert output.shape[1] == 2 - assert output.shape[2] == mlp.config.hidden_size - assert output_bias.shape[0] == mlp.config.hidden_size + assert output.shape[2] == transformer_config.hidden_size + assert output_bias == None assert output.dtype == torch.float32 - @pytest.mark.skipif(not torch.cuda.is_available(), reason="CUDA not available") - def test_gpu_forward(self, mlp): - mlp.cuda() + @pytest.mark.skipif(not get_accelerator().is_available(), reason="accelerator not available") + def test_accelerator_forward(self, mlp, transformer_config): + mlp.to(device_name) # [sequence length, batch size, hidden size] - hidden_states = torch.ones((32, 2, mlp.config.hidden_size)) - hidden_states = hidden_states.cuda() + hidden_states = torch.ones((32, 2, transformer_config.hidden_size)) + hidden_states = hidden_states.to(device_name) output, output_bias = mlp(hidden_states) assert output.shape[0] == 32 assert output.shape[1] == 2 - assert output.shape[2] == mlp.config.hidden_size - assert output_bias.shape[0] == mlp.config.hidden_size + assert output.shape[2] == transformer_config.hidden_size + assert output_bias == None assert output.dtype == torch.float32 - assert output.device.type == 'cuda' - assert output_bias.device.type == 'cuda' + assert output.device.type == device_name From ea4b67a2bb5e572bdbe9c4fd3aa0e539e8e17110 Mon Sep 17 00:00:00 2001 From: Xinyu Chen Date: Thu, 27 Jun 2024 22:15:14 +0800 Subject: [PATCH 254/590] Fix test_deallocate_output_tensor (#404) --- tests/pipeline_parallel/test_schedules.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tests/pipeline_parallel/test_schedules.py b/tests/pipeline_parallel/test_schedules.py index a6bac5b2a31..64bd7c3ac64 100644 --- a/tests/pipeline_parallel/test_schedules.py +++ b/tests/pipeline_parallel/test_schedules.py @@ -21,7 +21,9 @@ def test_get_forward_backward_func(): def test_deallocate_output_tensor(): out = torch.tensor([[1, 2, 3], [4, 5, 6]]) schedule.deallocate_output_tensor(out) - assert(out.nelement() == 1) + assert(out.nelement() == 6) + schedule.deallocate_output_tensor(out, True) + assert(out.nelement() == 1) def test_forward_backward_func_without_pipeline_parallel(mocker): from megatron.core.pipeline_parallel import get_forward_backward_func From 7388c1abbb7e920b1371ec79f7d3e4362a144fee Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 29 Jun 2024 09:22:45 -0500 Subject: [PATCH 255/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5bd09c712c2..8704e0d65eb 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1284,6 +1284,31 @@ $flops_profiler EOT } +##################### +# train +##################### +train() { + # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] + cd "${PBS_O_WORKDIR}" || exit + HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE + # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] + source "${HERE}/ALCF/helpers.sh" || exit + # 3. call `setup` from `./ALCF/helpers.sh` + export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" + setup || exit + # 4. Take custom args + export custom_args=" $@" + # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} + export run_cmd="${run_cmd} ${custom_args}" + # 6. Add "${run_cmd}" to output log + echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" + # 7. Tell user where to find output + printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} + eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" + set +x +} + ############################################### # Helper functions for printing colored text ############################################### From 08f5a996903affd36b664c52cbbb8b7e770abcdd Mon Sep 17 00:00:00 2001 From: Costin Eseanu <19394655+costin-eseanu@users.noreply.github.com> Date: Mon, 1 Jul 2024 07:05:31 -0700 Subject: [PATCH 256/590] Fixed missing BookCorpus dataset. (#407) Co-authored-by: Costin Eseanu --- .../ds_pretrain_gpt_1.3B_seq_parallel_32k.sh | 26 +++++++++++++------ .../preprocess_bookcorpus.py | 4 +++ megatron/arguments.py | 4 +-- 3 files changed, 24 insertions(+), 10 deletions(-) create mode 100644 examples_deepspeed/sequence_parallel/preprocess_bookcorpus.py diff --git a/examples_deepspeed/sequence_parallel/ds_pretrain_gpt_1.3B_seq_parallel_32k.sh b/examples_deepspeed/sequence_parallel/ds_pretrain_gpt_1.3B_seq_parallel_32k.sh index da028dc7314..24bfa544d61 100644 --- a/examples_deepspeed/sequence_parallel/ds_pretrain_gpt_1.3B_seq_parallel_32k.sh +++ b/examples_deepspeed/sequence_parallel/ds_pretrain_gpt_1.3B_seq_parallel_32k.sh @@ -187,14 +187,6 @@ host="${HOSTNAME}" seed=1234 num_workers=0 -data_path="BookCorpusDataset_text_document" -if [ ! -f "BookCorpusDataset_text_document.bin" ]; then - wget https://the-eye.eu/public/AI/pile_neox/data/BookCorpusDataset_text_document.bin -fi -if [ ! -f "BookCorpusDataset_text_document.idx" ]; then - wget https://the-eye.eu/public/AI/pile_neox/data/BookCorpusDataset_text_document.idx -fi - vocab_path="gpt2-vocab.json" if [ ! -f "$vocab_path" ]; then wget https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json @@ -204,6 +196,24 @@ if [ ! -f "$merge_path" ]; then wget https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt fi + +data_path="BookCorpusDataset_text_document" +if [ ! -f "BookCorpusDataset_text_document.bin" ]; then + # Download the Bookcorpus dataset and convert to json + python preprocess_bookcorpus.py + + # Process the dataset + python ${dir}/../../tools/preprocess_data.py \ + --input ${data_path}.json \ + --output-prefix "BookCorpusDataset" \ + --vocab-file $vocab_path \ + --merge-file $merge_path \ + --dataset-impl mmap \ + --tokenizer-type GPT2BPETokenizer \ + --workers 32 \ + --append-eod +fi + prescale_grad="true" jobname="gpt_${model_size}B_tok${train_tokens_in_billion}B" jobname="${jobname}_lr${lr}_min${min_lr}_w${lr_warmup_tokens_in_million}M_d${lr_decay_tokens_in_billion}B_${lr_decay_style}" diff --git a/examples_deepspeed/sequence_parallel/preprocess_bookcorpus.py b/examples_deepspeed/sequence_parallel/preprocess_bookcorpus.py new file mode 100644 index 00000000000..c35a13ea68f --- /dev/null +++ b/examples_deepspeed/sequence_parallel/preprocess_bookcorpus.py @@ -0,0 +1,4 @@ +from datasets import load_dataset + +train_data = load_dataset('bookcorpus/bookcorpus', split='train') +train_data.to_json("BookCorpusDataset_text_document.json", lines=True) diff --git a/megatron/arguments.py b/megatron/arguments.py index e7182c317ed..88898548502 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -101,8 +101,8 @@ def validate_args(args, defaults={}): args.ds_sequence_parallel_size assert args.world_size % model_parallel_size == 0, 'world size ({}) is not'\ ' divisible by tensor parallel size ({}) times pipeline parallel ' \ - 'size ({})'.format(args.world_size, args.tensor_model_parallel_size, - args.pipeline_model_parallel_size) + 'size ({}) times seqence parallel size ({})'.format(args.world_size, args.tensor_model_parallel_size, + args.pipeline_model_parallel_size, args.ds_sequence_parallel_size) args.data_parallel_size = args.world_size // model_parallel_size if args.rank == 0: print('using world size: {}, data-parallel-size: {}, ' From c3a13be721da0d0de16c338d0d665b0f7d13d14f Mon Sep 17 00:00:00 2001 From: Xinyu Chen Date: Tue, 2 Jul 2024 01:22:08 +0800 Subject: [PATCH 257/590] Set proper arguments when constructing models in unit tests (#408) --- tests/models/test_gpt_embedding.py | 4 ++++ tests/models/test_gpt_model.py | 21 +++++++++++++++++---- tests/transformer/test_parallel_mlp.py | 13 +++++++++++++ 3 files changed, 34 insertions(+), 4 deletions(-) diff --git a/tests/models/test_gpt_embedding.py b/tests/models/test_gpt_embedding.py index 28f8d52f15a..8a188b0cd25 100644 --- a/tests/models/test_gpt_embedding.py +++ b/tests/models/test_gpt_embedding.py @@ -3,15 +3,19 @@ import pytest import torch +import types from megatron.core.transformer.transformer_config import TransformerConfig from megatron.core.models.gpt.gpt_embedding import GPTEmbedding +from megatron.global_vars import set_args from deepspeed.accelerator import get_accelerator device_name = get_accelerator().device_name() @pytest.fixture def gpt_embedding(transformer_config): + args = types.SimpleNamespace(params_dtype=torch.float32, embed_layernorm=False) + set_args(args) embedding = GPTEmbedding(config=transformer_config, vocab_size=100, max_sequence_length=4) return embedding diff --git a/tests/models/test_gpt_model.py b/tests/models/test_gpt_model.py index b854ecd918a..b6888af5cc7 100644 --- a/tests/models/test_gpt_model.py +++ b/tests/models/test_gpt_model.py @@ -3,18 +3,25 @@ import pytest import torch +import types from megatron.core.transformer.transformer_config import TransformerConfig from megatron.core.models.gpt.gpt_model import GPTModel +from megatron.global_vars import set_args +from deepspeed.accelerator import get_accelerator +device_name = get_accelerator().device_name() @pytest.fixture def gpt_model(transformer_config): + args = types.SimpleNamespace(params_dtype=torch.float32, embed_layernorm=False) + set_args(args) language_model = GPTModel(config=transformer_config, vocab_size=100, max_sequence_length=4) return language_model class TestGPTModel: + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_constructor(self, gpt_model: GPTModel): assert isinstance(gpt_model, GPTModel) @@ -23,6 +30,7 @@ def test_constructor(self, gpt_model: GPTModel): num_weights = sum([p.numel() for p in gpt_model.parameters()]) assert num_weights == 5040 + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_set_input_tensor(self, gpt_model: GPTModel): config: TransformerConfig = gpt_model.config sequence_length = gpt_model.max_sequence_length @@ -37,17 +45,18 @@ def test_set_input_tensor(self, gpt_model: GPTModel): assert gpt_model.decoder.input_tensor.shape[1] == micro_batch_size assert gpt_model.decoder.input_tensor.shape[2] == config.hidden_size + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_post_process_forward(self, gpt_model: GPTModel): config: TransformerConfig = gpt_model.config sequence_length = gpt_model.max_sequence_length micro_batch_size = 2 - gpt_model.cuda() + gpt_model.to(device_name) data = list(range(sequence_length)) - input_ids = torch.tensor(data, dtype=torch.int64).repeat((micro_batch_size, 1)).cuda() - position_ids = torch.tensor(data, dtype=torch.int64).repeat((micro_batch_size, 1)).cuda() - attention_mask = torch.ones((1, 1, sequence_length, sequence_length), dtype=bool).cuda() + input_ids = torch.tensor(data, dtype=torch.int64).repeat((micro_batch_size, 1)).to(device_name) + position_ids = torch.tensor(data, dtype=torch.int64).repeat((micro_batch_size, 1)).to(device_name) + attention_mask = torch.ones((1, 1, sequence_length, sequence_length), dtype=bool).to(device_name) logits = gpt_model.forward(input_ids=input_ids, position_ids=position_ids, attention_mask=attention_mask) @@ -55,15 +64,19 @@ def test_post_process_forward(self, gpt_model: GPTModel): assert logits.shape[1] == sequence_length assert logits.shape[2] == gpt_model.vocab_size + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_no_post_process_forward(self, gpt_model: GPTModel): pass + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_no_preprocess_forward(self, gpt_model: GPTModel): pass + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_state_dict_for_save_checkpoint(self, gpt_model: GPTModel): pass + @pytest.mark.xfail(device_name=='hpu', reason="TELayerNorm is not defined in HPU") def test_load_state_dict(self, gpt_model: GPTModel): pass diff --git a/tests/transformer/test_parallel_mlp.py b/tests/transformer/test_parallel_mlp.py index f9fa365c686..ac5e4b0cd5f 100644 --- a/tests/transformer/test_parallel_mlp.py +++ b/tests/transformer/test_parallel_mlp.py @@ -3,14 +3,27 @@ import pytest import torch +import types from megatron.core.transformer.parallel_mlp import ParallelMLP +from megatron.global_vars import set_args from deepspeed.accelerator import get_accelerator device_name = get_accelerator().device_name() @pytest.fixture def mlp(transformer_config): + mlp_args = types.SimpleNamespace( + swiglu=False, + openai_gelu=True, + onnx_safe=False, + bias_gelu_fusion=False, + transformer_impl="", + cache_fp8_weight=False, + fp8_interval=False, + cache_fp8_weight_fwd=False + ) + set_args(mlp_args) return ParallelMLP(transformer_config) From b511a2e3b9a02c6d1b2e07c69b17e8b53f2875e1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 5 Jul 2024 09:22:28 -0500 Subject: [PATCH 258/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 354 ++++++++++++++++++++++++------------------------ 1 file changed, 178 insertions(+), 176 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8704e0d65eb..adc2ad742fc 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -83,37 +83,39 @@ helpers_main() { # 14. Setup run command to be executed. ############################################################################## setup() { - # 1. Identify machine we're on + # Identify machine we're on get_machine || exit - # 2. Load `conda` environment, setup virtual env + # Load `conda` environment, setup virtual env setup_python || exit - # 3. Ensure necessary dependencies all installed + # Ensure necessary dependencies all installed install_dependencies || exit - # 4. Determine WORLD_SIZE, etc. from `PBS_*` vars - setup_ezpz || exit - # 5. Set command line arguments to pass to `"${EXEC}"` + # Determine WORLD_SIZE, etc. from `PBS_*` vars + ezpz_setup "$@" || exit + # Ensure `ezpz` installed into python environment + ezpz_install || exit + # Set command line arguments to pass to `"${EXEC}"` setParams || exit - # 6. Create `deepspeed_config.json` from runtime params from ^ + # Create `deepspeed_config.json` from runtime params from ^ buildDSconfig || exit - # 7. Specify output directory for {logs, checkpoints, etc.} + # Specify output directory for {logs, checkpoints, etc.} setOutput || exit - # 8. Specify additional `deepspeed` arguments (dependent on _newly created_ variables) + # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) setArgs || exit - # 9. Ensure executable exists in expected path - export EXEC="${EXEC:-${HERE}/pretrain_gpt_alcf.py}" - check_executable "${EXEC}" + # Ensure executable exists in expected path + # export EXEC="${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" + check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" dfl="${DATA_FILE_LIST:-}" - # 10. Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` + # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit make_data || exit - # 11. Print job info + # Print job info printJobInfo || exit - # 12. Save `.env` to `CKPT_DIR` for safe keeping + # Save `.env` to `CKPT_DIR` for safe keeping save_dotenv "${CKPT_DIR}" || exit - # 13. Check that were not already running, if so, exit. + # Check that were not already running, if so, exit. check_and_kill_if_running || exit - # 14. Setup run command to be executed + # Setup run command to be executed setup_run_cmd || exit } @@ -324,6 +326,7 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else + if [[ -n "${DIST_LAUNCH}" ]]; then LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" export LAUNCHER="${LAUNCHER}" @@ -333,7 +336,6 @@ setupLauncher() { echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" fi fi - printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" printf " %s" "$(printMagenta "${LAUNCHER}")" } @@ -592,31 +594,40 @@ ezpz_getjobenv() { } ########################################### -# setup_ezpz +# ezpz_setup # -# 1. {save,get}jobenv -# 2. python3 -m ezpz.jobs && source "./.jobenv" -# 2. Install ezpz (if not installed) +# 1. source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/utils.sh) +# 2. call `setup_alcf` (provided by `ezpz/utils.sh` from 1.) ########################################### -setup_ezpz() { - if [[ -n "${HOSTFILE:-${PBS_NODEFILE}}" ]]; then - ezpz_savejobenv - else - ezpz_getjobenv - fi +ezpz_setup() { + # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh + source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" && setup_alcf "$@" || exit + # setup_alcf "$@" + # file=$(mktemp) + # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" + # source "${file}" || exit +} + +########################################### +# ezpz_install +# +# Install ezpz (if not installed) +########################################### +ezpz_install() { ezdir="${WORKING_DIR}/deps/ezpz" if [[ ! -d "${ezdir}" ]]; then mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" fi + # source "${ezdir}/utils.sh" && setup_alcf "$@" || exit ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -z "${ezloc:-}" ]]; then - printf "[setup_ezpz] Installing ezpz from %s\n" "${ezdir}" + printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" python3 -m pip install -e "${ezdir}" --require-virtualenv else - printf "[setup_ezpz] Found ezpz @ %s\n" "${ezloc}" + printf "[ezpz_install] Found ezpz @ %s\n" "${ezloc}" fi - python3 -m ezpz.jobs && source "./.jobenv" + # python3 -m ezpz.jobs && source "./.jobenv" # ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') # if [[ -n "${ezloc}" ]]; then # # ezpz_savejobenv @@ -746,6 +757,11 @@ make_data() { } +############################################################################## +# install_dependencies +# +# Ensure all dependencies installed from `ALCF/requirements/requirements.txt` +############################################################################## install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" echo "Ensuring all dependencies from ${depsfile} installed..." @@ -761,7 +777,9 @@ install_dependencies() { } ###################################################################### -# install_deepspeed_for_xpu: Install microsoft/DeepSpeed on PVC +# install_deepspeed_for_xpu +# +# Install microsoft/DeepSpeed on PVC # # This will: # 1. Clone rep @@ -916,18 +934,18 @@ setup_venv_from_conda() { check_executable() { fp=$1 if [[ -f "${fp}" ]]; then - export EXEC="${EXEC}" + export EXEC="${fp}" # ----[1.5 Keep track of stem from file path]------------------------- exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") export EXEC_STEM="${exec_stem}" else estr="Unable to locate executable ${fp}" - printf "[ALCF.helpers:check_executable] %s" "$(printRed "${estr}")" + printf "[ALCF.helpers:check_executable] %s\n" "$(printRed "${estr}")" fi } ############################################################################## -# `setup_python`: +# setup_python # # 1. Setup `conda` # - if `conda` nonempty, and `venv` empty, use `conda` to setup `venv`. @@ -1062,9 +1080,10 @@ setup_tokenizer_and_data() { } ############################################### -# `setData`: -# Ensure `DATA_FILE_LIST` is set, -# fallback to default values if necessary. +# setData +# +# Ensure `DATA_FILE_LIST` is set, +# fallback to default values if necessary. ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# @@ -1094,7 +1113,6 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ndocs=$(wc -l < "${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") - # dcp="${OUTPUT_PREFIX:-$(get_output_prefix)}/.cache/${dfl_stem}/index-cache" dcp=".cache/${dfl_stem}/index-cache" export DATA_FILE_LIST="${dfl}" export NUM_DOCS="${ndocs}" @@ -1111,30 +1129,26 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" printf "DATA_FLAGS: %s\n" "${DATA_FLAGS}" echo "--------------------" - # fi - # export DATA_FLAGS="${DATA_FLAGS}" - # export TOKENIZER_FLAGS="${TOKENIZER_FLAGS}" - # printf "[setData] DATA_FLAGS: %s\n" "$(printGreen ${DATA_FLAGS})" - # printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta ${TOKENIZER_FLAGS})" } ################################################################################ -# generateDSconfig: Create and save a deepspeed config .json +# generateDSconfig +# +# Create and save a deepspeed config .json # # This will contain the appropriate variables as set in the current environment. ################################################################################ generateDSconfig() { - for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE" - do - if [ -z "$v" ]; then - echo "Please export required envs before execute $0" - exit 1 - fi - done if [ $# -ne 1 ]; then echo "Usage: $0 config_file" exit 1 fi + for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE"; do + if [ -z "$v" ]; then + echo "Please export required envs before execute $0" + exit 1 + fi + done # \"optimizer\": { # \"type\": \"AdamW\", # \"params\": { @@ -1176,103 +1190,100 @@ generateDSconfig() { \"output_file\": null }" if [[ $DTYPE == "bf16" ]]; then - dtype="\ - \"communication_data_type\": \"bf16\", - \"fp16\": { - \"enabled\": false, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": true, - \"loss_scale\": 1.0 - }," + dtype="\ + \"communication_data_type\": \"bf16\", + \"fp16\": { + \"enabled\": false, + \"loss_scale\": 0, + \"loss_scale_window\": 1000, + \"hysteresis\": 2, + \"min_loss_scale\": 1 + }, + \"bfloat16\": { + \"enabled\": true, + \"loss_scale\": 1.0 + }," elif [[ $DTYPE == "fp16" ]]; then - dtype="\ - \"communication_data_type\": \"fp16\", - \"fp16\": { - \"enabled\": true, - \"loss_scale\": 0, - \"loss_scale_window\": 1000, - \"hysteresis\": 2, - \"min_loss_scale\": 1 - }, - \"bfloat16\": { - \"enabled\": false, - \"loss_scale\": 1.0 - }," + dtype="\ + \"communication_data_type\": \"fp16\", + \"fp16\": { + \"enabled\": true, + \"loss_scale\": 0, + \"loss_scale_window\": 1000, + \"hysteresis\": 2, + \"min_loss_scale\": 1 + }, + \"bfloat16\": { + \"enabled\": false, + \"loss_scale\": 1.0 + }," else - dtype="\"communication_data_type\": \"fp32\"," + dtype="\"communication_data_type\": \"fp32\"," fi - if [ "$ZERO_STAGE" == 3 ]; then - zero="\ - \"zero_optimization\": { - \"stage\": 3, - \"reduce_scatter\": false, - \"mics_shard_size\": 4, - \"mics_hierarchical_params_gather\": true, - \"stage3_max_live_parameters\": 3e9, - \"stage3_max_reuse_distance\": 3e9, - \"stage3_param_persistence_threshold\": 1e5, - \"stage3_prefetch_bucket_size\": 5e7, - \"contiguous_gradients\": true, - \"overlap_comm\": true, - \"reduce_bucket_size\": 90000000, - \"sub_group_size\": 1e9, - \"offload_optimizer\": { - \"device\": \"none\", - \"buffer_count\": 4, - \"pipeline_read\": false, - \"pipeline_write\": false, - \"pin_memory\": true - } - }," + if [[ "${ZERO_STAGE}" == 3 ]]; then + # \"mics_shard_size\": 2, + zero="\ + \"zero_optimization\": { + \"stage\": 3, + \"reduce_scatter\": false, + \"mics_hierarchical_params_gather\": true, + \"stage3_max_live_parameters\": 3e9, + \"stage3_max_reuse_distance\": 3e9, + \"stage3_param_persistence_threshold\": 1e5, + \"stage3_prefetch_bucket_size\": 5e7, + \"contiguous_gradients\": true, + \"overlap_comm\": true, + \"reduce_bucket_size\": 90000000, + \"sub_group_size\": 1e9, + \"offload_optimizer\": { + \"device\": \"none\", + \"buffer_count\": 4, + \"pipeline_read\": false, + \"pipeline_write\": false, + \"pin_memory\": true + } + }," # elif [[ $ZERO_STAGE == 2 ]]; then - elif [ "${ZERO_STAGE}" == 2 ] || [ "${ZERO_STAGE}" == 1 ]; then - # if [[ -n "${CPU_OPTIMIZER}" ]]; then - if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then - echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" - zero="\ - \"zero_optimization\": { - \"stage\": $ZERO_STAGE, - \"offload_optimizer\": { - \"device\": \"cpu\" - } - }," - else - zero="\ - \"zero_optimization\": { - \"stage\": $ZERO_STAGE - }," - fi - # elif [[ $ZERO_STAGE == 1 ]]; then - if [[ "${PP}" -gt 1 ]]; then - extra="\ - \"data_types\": { - \"grad_accum_dtype\": \"fp32\" - }, - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," - else - # echo 'please add the config for zero_stage 1 without pipeline-parallelism' - extra="\ - \"comms_logger\": { - \"enabled\": true, - \"verbose\": false, - \"prof_all\": true, - \"debug\": false - }," - fi + elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then + # if [[ -n "${CPU_OPTIMIZER}" ]]; then + if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then + echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" + zero="\ + \"zero_optimization\": { + \"stage\": $ZERO_STAGE, + \"offload_optimizer\": { + \"device\": \"cpu\" + } + }," + else + zero="\ + \"zero_optimization\": { + \"stage\": $ZERO_STAGE + }," + fi + if [[ "${PP}" -gt 1 ]]; then + extra="\ + \"data_types\": { + \"grad_accum_dtype\": \"fp32\" + }, + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," + else + extra="\ + \"comms_logger\": { + \"enabled\": true, + \"verbose\": false, + \"prof_all\": true, + \"debug\": false + }," + fi else - echo 'Please add the correct config set!!!' + echo 'Please add the correct config set!!!' fi -# flops_profiler must at the end because no ',' is allowed at the end cat < "$1" { $common @@ -1284,30 +1295,30 @@ $flops_profiler EOT } -##################### -# train -##################### -train() { - # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] - cd "${PBS_O_WORKDIR}" || exit - HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE - # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] - source "${HERE}/ALCF/helpers.sh" || exit - # 3. call `setup` from `./ALCF/helpers.sh` - export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" - setup || exit - # 4. Take custom args - export custom_args=" $@" - # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} - export run_cmd="${run_cmd} ${custom_args}" - # 6. Add "${run_cmd}" to output log - echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" - # 7. Tell user where to find output - printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" - # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} - eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" - set +x -} +# ##################### +# # train +# ##################### +# train() { +# # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] +# cd "${PBS_O_WORKDIR}" || exit +# HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE +# # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] +# source "${HERE}/ALCF/helpers.sh" || exit +# # 3. call `setup` from `./ALCF/helpers.sh` +# # export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" +# setup || exit +# # 4. Take custom args +# export custom_args=" $@" +# # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} +# export run_cmd="${run_cmd} ${custom_args}" +# # 6. Add "${run_cmd}" to output log +# echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +# # 7. Tell user where to find output +# printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" +# # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} +# eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" +# set +x +# } ############################################### # Helper functions for printing colored text @@ -1344,15 +1355,6 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } -export AURORA_GPT_HEADER=""" - █████╗ ██╗ ██╗██████╗ ██████╗ ██████╗ █████╗ ██████╗ ██████╗ ████████╗ -██╔══██╗██║ ██║██╔══██╗██╔═══██╗██╔══██╗██╔══██╗ ██╔════╝ ██╔══██╗╚══██╔══╝ -███████║██║ ██║██████╔╝██║ ██║██████╔╝███████║█████╗██║ ███╗██████╔╝ ██║ -██╔══██║██║ ██║██╔══██╗██║ ██║██╔══██╗██╔══██║╚════╝██║ ██║██╔═══╝ ██║ -██║ ██║╚██████╔╝██║ ██║╚██████╔╝██║ ██║██║ ██║ ╚██████╔╝██║ ██║ -╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚═╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝ -""" - ########################### # call helpers_main() ########################### From 330f9f29e6d4caf9c1f85709598d9f733e8e9ff4 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Mon, 8 Jul 2024 22:25:27 +0300 Subject: [PATCH 259/590] use split/squeeze instead of slice for performance (#409) * use split/squeeze instead of slice for performance GPU may not have perf difference but HPU perf improves with this * add copyrights --- megatron/model/transformer.py | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index e75f13a24f4..be8ae6ef4bf 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Transformer.""" @@ -660,9 +661,10 @@ def repeat_kv(self, hidden_states, n_rep): head_dim) def split_tensor(self, mixed_x_layer): - query_layer = mixed_x_layer[:, :, :, :-2, :].reshape(mixed_x_layer.shape[:2] + (-1, self.hidden_size_per_attention_head)) - key_layer = mixed_x_layer[:, :, :, -2, :] - value_layer = mixed_x_layer[:, :, :, -1, :] + query_layer, key_layer, value_layer = torch.split(mixed_x_layer, [self.num_key_value_groups, 1, 1], dim=-2) + query_layer = query_layer.reshape(mixed_x_layer.shape[:2] + (self.num_attention_heads_per_partition, self.hidden_size_per_attention_head)) + key_layer = torch.squeeze(key_layer, -2) + value_layer = torch.squeeze(value_layer, -2) return query_layer, key_layer, value_layer From af06d1449a0dac732deb172568553254f30ff657 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Mon, 8 Jul 2024 22:58:32 +0300 Subject: [PATCH 260/590] improve performance by keeping attention_mask on device and run ops further on device (#411) * improve performance by keeping attention_mask on device and run ops further on device * add copyrights --- megatron/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 97294070af7..9033d6402ae 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """General utilities.""" @@ -178,7 +179,7 @@ def get_ltor_masks_and_position_ids(data, attention_mask = None if not skip_mask: attention_mask = torch.tril(torch.ones( - (att_mask_batch, seq_length, seq_length))).view(att_mask_batch, 1, seq_length, seq_length) + (att_mask_batch, seq_length, seq_length), device=data.device)).view(att_mask_batch, 1, seq_length, seq_length) # Loss mask. loss_mask = torch.ones(data.size(), dtype=torch.float, device=data.device) @@ -218,7 +219,6 @@ def get_ltor_masks_and_position_ids(data, # Convert attention mask to binary: if not skip_mask: attention_mask = (attention_mask < 0.5) - attention_mask = attention_mask.to(data.device) return attention_mask, loss_mask, position_ids From 561ddc102bcbd30ce6da788be1fd05d6a937142f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 12:57:47 -0500 Subject: [PATCH 261/590] Fix micro batch size on Polaris --- ALCF/helpers.sh | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 8704e0d65eb..4d2120178e9 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -357,10 +357,13 @@ set_lr_args() { # # - In particular, it seems that different node counts allow for different # `MICRO_BATCH` sizes. +# # Explicitly: -# - [NHOSTS <= 3]: `MICRO_BATCH=1` -# - [NHOSTS >= 4]: `MICRO_BATCH=2` -# - [NHOSTS >= 8]: `MICRO_BATCH=4` +# +# - [1 <= NHOSTS <= 2]: `MICRO_BATCH=1` +# - [3 <= NHOSTS <= 7]: `MICRO_BATCH=2` +# - [8 <= NHOSTS]: `MICRO_BATCH=4` +# # are the largest batch sizes that fit in memory at various node counts. ######################################################################### get_batch_size_on_polaris() { @@ -368,7 +371,7 @@ get_batch_size_on_polaris() { nhosts=$(wc -l < "${HOSTFILE:-${PBS_NODEFILE}}") if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then mbs=1 - elif [[ "${nhosts}" -ge 3 ]]; then + elif [[ "${nhosts}" -ge 3 && "${nhosts}" -le 7 ]]; then mbs=2 elif [[ "${nhosts}" -ge 8 ]]; then mbs=4 From 9ee09feb2905d25e3d916b778b2d424e7cb5db3f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 12:59:22 -0500 Subject: [PATCH 262/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 38 ++++++++++++++++++++++++-------------- 1 file changed, 24 insertions(+), 14 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index adc2ad742fc..9868f8a6fea 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -381,7 +381,9 @@ get_batch_size_on_polaris() { ############################################################################## -# `setParams`: Set / configure run options by parsing environment. +# setParams +# +# Set / configure run options by parsing environment. # # - any of the declared options below can be overridden # dynamically at runtime, e.g. to run with a `MICRO_BATCH` size of 2: @@ -392,8 +394,8 @@ get_batch_size_on_polaris() { setParams() { FLASH_ARG="" LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" - # +----[Parallelism Settings] -------------------------------------------+ - # +------[Aurora]--------||-------[SunSpot]-------------+ + # ---- [Parallelism Settings] -------------------------------------------+ + # ------ [Aurora] -------||------ [SunSpot] ------------- if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" @@ -405,7 +407,7 @@ setParams() { ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? - update_ccl_env_vars_aurora + # update_ccl_env_vars_aurora ###################################################################### if [[ -z "${USE_FLASH_ATTN:-}" ]]; then # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! @@ -838,20 +840,28 @@ use_kvs_fix_on_aurora() { } update_ccl_env_vars_aurora() { - export CCL_KVS_MODE=mpi - # export CCL_CONFIGURATION_PATH="" - # unset CCL_CONFIGURATION_PATH - # export CCL_CONFIGURATION=cpu_gpu_dpcpp - # export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" - export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH - # export CCL_ALLREDUCE_SCALEOUT=direct - printenv | grep -E -v "^__" | grep -E "CCL|LD|CPATH|LIBRARY_PATH" + # export CCL_KVS_MODE=mpi + # # export CCL_CONFIGURATION_PATH="" + # # unset CCL_CONFIGURATION_PATH + # # export CCL_CONFIGURATION=cpu_gpu_dpcpp + # # export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + # export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + # export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + # export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + # # export CCL_ALLREDUCE_SCALEOUT=direct + # printenv | grep -E -v "^__" | grep -E "CCL|LD|CPATH|LIBRARY_PATH" ######################################################### # if not set, CCL will complain... ? export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" ######################################################### + # Sam: [2024-06-29] + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export CCL_CONFIGURATION=cpu_gpu_dpcpp + export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH } ########################### From 76209f4df845000f492e1fe0ccdc6b92875f7a67 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 14:00:36 -0500 Subject: [PATCH 263/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 61 +++++++++++++++++++------------------------------ 1 file changed, 24 insertions(+), 37 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 9868f8a6fea..5a9ec6e6a69 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -85,14 +85,14 @@ helpers_main() { setup() { # Identify machine we're on get_machine || exit - # Load `conda` environment, setup virtual env - setup_python || exit - # Ensure necessary dependencies all installed - install_dependencies || exit + # # Load `conda` environment, setup virtual env + # setup_python || exit + # # Ensure necessary dependencies all installed + # install_dependencies || exit + # # Ensure `ezpz` installed into python environment + # ezpz_install || exit # Determine WORLD_SIZE, etc. from `PBS_*` vars ezpz_setup "$@" || exit - # Ensure `ezpz` installed into python environment - ezpz_install || exit # Set command line arguments to pass to `"${EXEC}"` setParams || exit # Create `deepspeed_config.json` from runtime params from ^ @@ -595,20 +595,6 @@ ezpz_getjobenv() { source "${file}" || exit } -########################################### -# ezpz_setup -# -# 1. source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/utils.sh) -# 2. call `setup_alcf` (provided by `ezpz/utils.sh` from 1.) -########################################### -ezpz_setup() { - # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh - source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" && setup_alcf "$@" || exit - # setup_alcf "$@" - # file=$(mktemp) - # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" - # source "${file}" || exit -} ########################################### # ezpz_install @@ -621,7 +607,6 @@ ezpz_install() { mkdir -p "${WORKING_DIR}/deps" git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" fi - # source "${ezdir}/utils.sh" && setup_alcf "$@" || exit ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -z "${ezloc:-}" ]]; then printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" @@ -629,22 +614,24 @@ ezpz_install() { else printf "[ezpz_install] Found ezpz @ %s\n" "${ezloc}" fi - # python3 -m ezpz.jobs && source "./.jobenv" - # ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') - # if [[ -n "${ezloc}" ]]; then - # # ezpz_savejobenv - # # python3 -m ezpz.jobs && source "./.jobenv" - # make_ds_hostfile || exit - # else - # echo "No ezpz detected. Attempting to install with $(which python3)" - # python3 -m pip install -e "${WORKING_DIR}/deps/ezpz" --require-virtualenv - # fi - # if [[ ! -d "${WORKING_DIR}/deps/ezpz" ]]; then - # mkdir -p "${WORKING_DIR}/deps" - # git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" - # else - # echo "Found ezpz!" - # fi +} + +########################################### +# ezpz_setup +# +# 1. source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/utils.sh) +# 2. call `setup_alcf` (provided by `ezpz/utils.sh` from 1.) +########################################### +ezpz_setup() { + # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh + source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" || exit # && setup_alcf "$@" || exit + setup_python + ezpz_install + setup_alcf "$@" + # setup_alcf "$@" + # file=$(mktemp) + # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" + # source "${file}" || exit } ####################################################################### From 541ebf1be5819ba7ca075cae057d3d0de5eb2506 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 14:56:25 -0500 Subject: [PATCH 264/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 52 +++++++++++++++++++++++++++++++++---------------- 1 file changed, 35 insertions(+), 17 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5a9ec6e6a69..fe071b9fcba 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -85,13 +85,19 @@ helpers_main() { setup() { # Identify machine we're on get_machine || exit - # # Load `conda` environment, setup virtual env - # setup_python || exit - # # Ensure necessary dependencies all installed - # install_dependencies || exit - # # Ensure `ezpz` installed into python environment - # ezpz_install || exit - # Determine WORLD_SIZE, etc. from `PBS_*` vars + ########################################################################## + # ezpz_setup will: + # 1. Setup python + # - load base conda + # - (if necessary) create virtual environment on top of base conda + # - activate virtual environment from ^ + # 2. Install ezpz (if needed) + # 3. Parse PBS_* environment variables to determine: + # - NHOSTS (by counting number of lines in $PBS_NODEFILE) + # - NGPU_PER_HOST (by magic) + # - NGPUS (= NHOSTS * NGPU_PER_HOST) + # 4. Use these (^) to build our launch command + ########################################################################## ezpz_setup "$@" || exit # Set command line arguments to pass to `"${EXEC}"` setParams || exit @@ -124,6 +130,7 @@ setup() { # # Build run command to be executed. ##################################################### + # --train-tokens ${TRAIN_TOKENS} \ setup_run_cmd() { #### Make it easy to track experiments by date ################### year="$(date "+%Y")" @@ -171,6 +178,7 @@ setup_run_cmd() { echo "!! Running in NO_LLAMA MODE !!" llama_flags="" fi + # --train-iters ${TRAIN_ITER} \ export run_cmd=" ${LAUNCHER} \ --${DTYPE} \ @@ -181,8 +189,8 @@ setup_run_cmd() { --seq-length ${SEQ} \ --num-layers ${NLAYERS} \ --hidden-size ${HIDDEN} \ - --train-iters ${TRAIN_ITER} \ --tensorboard-dir ${TBDIR} \ + --train-iters ${TRAIN_ITERS} \ --eval-iters ${EVAL_ITERS} \ --distributed-backend ${BE} \ --num-attention-heads ${HEADS} \ @@ -482,7 +490,6 @@ setParams() { export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN - export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL export SAVE_INTERVAL=${SAVE_INTERVAL:-50} # HOW FREQUENTLY TO SAVE CKPTS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS @@ -490,10 +497,23 @@ setParams() { export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$(( WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP )) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + # export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS + if [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then + export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} + export TRAIN_ITERS=$(( TRAIN_TOKENS / SEQ / GLOBAL_BATCH )) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$(( TRAIN_TOKENS / 10**9 ))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + else + export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" + fi export MODEL_TYPE="llama-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL - # +----[ADDITIONAL LLAMA SPECIFIC ARGUMENTS]------------------------------ + # NOTE: [2024-07-10] ##################################################### + # - [sam]: For whatever reason, it seems that using + # sequence-parallelism (SP) > 1 is INCOMPATIBLE with + # rotary-position-embeddings (ROPE). + # + # For this reason, we only use the default LLAMA_ARGS when SP=0. + ########################################################################## if [[ "${SP}" == 1 ]]; then export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" else @@ -503,7 +523,6 @@ setParams() { # -----[Learning Rate Settings]-------------------------------------------- export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP - # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-320000} # LR DECAY ITERS export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args # -----[Learning Rate Settings]-------------------------------------------- @@ -623,15 +642,14 @@ ezpz_install() { # 2. call `setup_alcf` (provided by `ezpz/utils.sh` from 1.) ########################################### ezpz_setup() { + # setup_alcf "$@" + # file=$(mktemp) + # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" || exit # && setup_alcf "$@" || exit setup_python ezpz_install setup_alcf "$@" - # setup_alcf "$@" - # file=$(mktemp) - # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" - # source "${file}" || exit } ####################################################################### From d76331f84061e8e8d94853178bdf775fdd6ececa Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 16:08:51 -0500 Subject: [PATCH 265/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 102 ++++++++++++++++++++++++------------------------ 1 file changed, 51 insertions(+), 51 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index fe071b9fcba..4b69c67c2f1 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,21 +1,17 @@ #!/bin/bash --login ############################################################################### -# `ALCF/helpers.sh` -# +# [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh) # # Contains helper functions for launching `../train_llama_alcf.sh` # -# > [!NOTE] -# > On any of {Polaris, Aurora, Sunspot} @ ALCF: -# > -# > ```bash -# > $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -# > $ cd Megatron-DeepSpeed -# > $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh -# > # on any of {Polaris, Sunspot, Aurora} @ ALCF -# > $ setup_python -# > ``` +# To use, on any of {Polaris, Aurora, Sunspot} @ ALCF: +# +# ```bash +# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +# $ cd Megatron-DeepSpeed +# $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup +# ``` ############################################################################## ################## @@ -29,7 +25,8 @@ # ``` # # - This will set `"${WORKING_DIR}"`, according to: -# 1. `${PBS_O_WORKDIR}` is nonzero, use this +# +# 1. if `${PBS_O_WORKDIR}` is nonzero, use this # 2. else, if `${SLURM_SUBMIT_DIR}` is nonzero use this # 3. else, use `$(pwd)` # @@ -38,7 +35,7 @@ # (e.g. virtual environment, location of executables, etc.) ################## helpers_main() { - # for debug mode, run with `DEBUG=1` + # NOTE: for debug mode, run with `DEBUG=1` if [[ -n "${DEBUG:-}" ]]; then set -euxo fi @@ -61,44 +58,44 @@ helpers_main() { # All-in-one helper function. # # - Explicitly, this will: -# 1. Identify the machine we're on -# 2. Setup `python` -# 1. Load `conda` -# 2. Setup `venv` on top of `conda` -# 3. Ensure all dependencies are installed -# 4. Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) -# 1. Additionally, call `source deps/ezpz/src/ezpz/bin/savejobenv`, -# which will automatically build a `alias launch=mpiexec ...` -# according to the specifics of our active job. -# 5. Set runtime options -# 6. Build `deepspeed_config.json` -# 7. Build {logs, checkpoints, etc} dirs, named according to specifics of -# current run -# 8. Specify additional `deepspeed` arguments -# 9. Ensure executable exists at expected path -# 10. Setup data + tokenizer via `TOKENIZER_TYPE` -# 11. Print job info -# 12. Save `.env` to `CKPT_DIR` for safe keeping -# 13. Check that we're not already running, and if so, exit. -# 14. Setup run command to be executed. +# - Identify the machine we're on +# - Setup `python` +# 1. Load `conda` +# 2. Setup `venv` on top of `conda` +# - Ensure all dependencies are installed +# - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) +# 1. Additionally, call `source deps/ezpz/src/ezpz/bin/savejobenv`, +# which will automatically build a `alias launch=mpiexec ...` +# according to the specifics of our active job. +# - Set runtime options +# - Build `deepspeed_config.json` +# - Build {logs, checkpoints, etc} dirs, named according to specifics of +# current run +# - Specify additional `deepspeed` arguments +# - Ensure executable exists at expected path +# - Setup data + tokenizer via `TOKENIZER_TYPE` +# - Print job info +# - Save `.env` to `CKPT_DIR` for safe keeping +# - Check that we're not already running, and if so, exit. +# - Setup run command to be executed. ############################################################################## setup() { # Identify machine we're on get_machine || exit ########################################################################## # ezpz_setup will: - # 1. Setup python - # - load base conda - # - (if necessary) create virtual environment on top of base conda - # - activate virtual environment from ^ - # 2. Install ezpz (if needed) - # 3. Parse PBS_* environment variables to determine: - # - NHOSTS (by counting number of lines in $PBS_NODEFILE) - # - NGPU_PER_HOST (by magic) - # - NGPUS (= NHOSTS * NGPU_PER_HOST) - # 4. Use these (^) to build our launch command - ########################################################################## + # 1. Setup python + # - load base conda + # - (if necessary) create virtual environment on top of base conda + # - activate virtual environment from ^ + # 2. Install ezpz (if needed) + # 3. Parse PBS_* environment variables to determine: + # - NHOSTS (by counting number of lines in $PBS_NODEFILE) + # - NGPU_PER_HOST (by magic) + # - NGPUS (= NHOSTS * NGPU_PER_HOST) + # 4. Use these (^) to build our launch command ezpz_setup "$@" || exit + ########################################################################## # Set command line arguments to pass to `"${EXEC}"` setParams || exit # Create `deepspeed_config.json` from runtime params from ^ @@ -108,7 +105,6 @@ setup() { # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) setArgs || exit # Ensure executable exists in expected path - # export EXEC="${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" dfl="${DATA_FILE_LIST:-}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` @@ -130,7 +126,6 @@ setup() { # # Build run command to be executed. ##################################################### - # --train-tokens ${TRAIN_TOKENS} \ setup_run_cmd() { #### Make it easy to track experiments by date ################### year="$(date "+%Y")" @@ -178,7 +173,6 @@ setup_run_cmd() { echo "!! Running in NO_LLAMA MODE !!" llama_flags="" fi - # --train-iters ${TRAIN_ITER} \ export run_cmd=" ${LAUNCHER} \ --${DTYPE} \ @@ -211,13 +205,11 @@ setup_run_cmd() { ${ds_args} \ ${gpt_args[*]} " - # ${LLAMA_ARGS} \ } save_dotenv() { if [[ "$#" -ne 1 ]]; then estr="[error]" - # echo "Expected exactly one argument, specifying outputdir. Received $#" printf "%s Expected one argument (outdir). Received: %s" "$(printRed "${estr}")" "$#" else outdir="$1" @@ -231,7 +223,15 @@ save_dotenv() { } ###################################################################### -# get_machine_name: Return current machine name, as lowercase string +# get_machine_name: +# +# Return current machine name, as lowercase string +# +# Example: +# ```bash +# $ machine_name=$(get_machine_name) +# $ echo "machine_name: ${machine_name}" +# ``` ###################################################################### get_machine_name() { if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then From d017b4c55dd678843044ca4ae57ead4dad3f37a2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 22:09:58 -0500 Subject: [PATCH 266/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 4b69c67c2f1..632ac58157c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -615,17 +615,26 @@ ezpz_getjobenv() { } +#################### +# ezpz_clone +# ################## +ezpz_clone() { + ezdir="${WORKING_DIR}/deps/ezpz" + # if [[ ! -d "${ezdir}" ]]; then + if [[ -d "${ezdir}" ]]; then + echo "Found ezpz in ${ezdir}" + else + mkdir -p "${WORKING_DIR}/deps" + git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" + fi +} + ########################################### # ezpz_install # # Install ezpz (if not installed) ########################################### ezpz_install() { - ezdir="${WORKING_DIR}/deps/ezpz" - if [[ ! -d "${ezdir}" ]]; then - mkdir -p "${WORKING_DIR}/deps" - git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" - fi ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -z "${ezloc:-}" ]]; then printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" @@ -646,6 +655,7 @@ ezpz_setup() { # file=$(mktemp) # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh + ezpz_clone source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" || exit # && setup_alcf "$@" || exit setup_python ezpz_install From bac8aab13ef823a86cd227abb477204cfacd264d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 10 Jul 2024 22:29:29 -0500 Subject: [PATCH 267/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 632ac58157c..e3bab445325 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -96,6 +96,7 @@ setup() { # 4. Use these (^) to build our launch command ezpz_setup "$@" || exit ########################################################################## + install_dependencies # Set command line arguments to pass to `"${EXEC}"` setParams || exit # Create `deepspeed_config.json` from runtime params from ^ @@ -810,9 +811,9 @@ install_deepspeed_for_xpu() { git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" cd "${outdir}" || exit echo "!! pwd: $(pwd)" - git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git - git fetch yizhou_ds - git checkout yizhou/kernel_path + # git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git + # git fetch yizhou_ds + # git checkout yizhou/kernel_path python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1> /dev/null python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1> /dev/null python setup.py develop 1> /dev/null From ec3f1f402e0e634f84458cba84c7e92967cf3107 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Thu, 11 Jul 2024 18:26:16 +0300 Subject: [PATCH 268/590] Improve RoPE perf by using cached sin/cos tensors (#410) * improve RoPE perf by using cached sin/cos tensors * add copyrights --- megatron/model/rotary_pos_embedding.py | 22 ++++++++++++++++++---- 1 file changed, 18 insertions(+), 4 deletions(-) diff --git a/megatron/model/rotary_pos_embedding.py b/megatron/model/rotary_pos_embedding.py index 4d4497e0cd9..0a7acb9efc2 100644 --- a/megatron/model/rotary_pos_embedding.py +++ b/megatron/model/rotary_pos_embedding.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # coding=utf-8 # The following code has been taken from https://github.com/NVIDIA/NeMo/blob/ \ @@ -11,6 +12,10 @@ __all__ = ['RotaryEmbedding', 'apply_rotary_pos_emb'] +# sin, cos tensors cached for all devices +cos_cached = None +sin_cached = None + class RotaryEmbedding(nn.Module): def __init__(self, dim, theta=10000): super().__init__() @@ -47,10 +52,19 @@ def apply_rotary_pos_emb(t, freqs): check https://kexue.fm/archives/8265 for detailed formulas """ rot_dim = freqs.shape[-1] - # ideally t_pass is empty so rotary pos embedding is applied to all tensor t - t, t_pass = t[..., :rot_dim], t[..., rot_dim:] + t_pass = None + if t.shape[-1] != rot_dim: + # ideally t_pass is empty so rotary pos embedding is applied to all tensor t + t, t_pass = t[..., :rot_dim], t[..., rot_dim:] + global cos_cached, sin_cached + if cos_cached is None or sin_cached is None or t.shape[0] != cos_cached.shape[0]: + freqs_ = freqs[:t.shape[0]] + cos_cached = freqs_.cos().to(t.dtype) + sin_cached = freqs_.sin().to(t.dtype) # first part is cosine component # second part is sine component, need to change signs with _rotate_half method - t = (t * freqs.cos().to(t.dtype)) + (_rotate_half(t) * freqs.sin().to(t.dtype)) - return t if t_pass.shape[-1] == 0 else torch.cat((t, t_pass), dim=-1) + t = (t * cos_cached) + (_rotate_half(t) * sin_cached) + if t_pass is None: + return t + return torch.cat((t, t_pass), dim=-1) From 911cc5ce7955ea1fed096ade5db454b9114ae13d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 12 Jul 2024 06:56:20 -0500 Subject: [PATCH 269/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 261 +++++++++--------------------------------------- 1 file changed, 47 insertions(+), 214 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index e3bab445325..ba74d559b85 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -12,7 +12,14 @@ # $ cd Megatron-DeepSpeed # $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup # ``` -############################################################################## +# +# and this will, automatically: +# +# 1. Setup python (conda + virtual environment) +# +# 2. Parse `$PBS_*` env vars to build appropriate `alias launch='mpiexec ...'` +# command for launching across all GPUs in our active PBS job. +############################################################################### ################## # helpers_main @@ -104,7 +111,7 @@ setup() { # Specify output directory for {logs, checkpoints, etc.} setOutput || exit # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) - setArgs || exit + set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" dfl="${DATA_FILE_LIST:-}" @@ -140,8 +147,8 @@ setup_run_cmd() { export TODAY="${today}" export STARTED_AT="${started_at}" ################################################################## - # to launch with DeepSpeed instead of mpiexec, run with: - # $ LAUNCH_WITH=deepspeed bash train_llama_alcf.sh + # NOTE: to launch with DeepSpeed instead of mpiexec: + # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` ################################################################## setupLauncher "${LAUNCH_WITH:-MPICH}" || exit TBDIR="${CKPT_DIR}/tensorboard" @@ -163,8 +170,10 @@ setup_run_cmd() { if [[ "${SP}" -ge 2 ]]; then export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" fi - # [hacky]: to disable Llama-type architectures, toggle via: + ################################################################## + # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` + ################################################################## if [[ -z "${NO_LLAMA:-}" ]]; then llama_flags="${LLAMA_ARGS}\ --num-key-value-heads ${NUM_KV_HEAD} \ @@ -411,7 +420,7 @@ setParams() { export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] @@ -540,12 +549,12 @@ setParams() { ############################################## -# setArgs +# set_args # # Specify additional (DeepSpeed specific) # arguments to pass to pretrain_gpt_alcf.py ############################################## -setArgs() { +set_args() { # ---- Set DeepSpeed arguments -------------------------------- ds_args=" " ds_args=" --deepspeed ${ds_args}" @@ -591,78 +600,49 @@ make_ds_hostfile() { sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } -#################### -# ezpz_savejobenv + +########################################### +# ezpz_setup # -# Parse relevant environment variables to determine: -# - num_hosts (by counting the number of lines in ${HOSTFILE:-${PBS_NODEFILE}}) -# - num_gpus_per_host (magically[^1]) -# - num_gpus = (num_hosts * num_gpus_per_host) +# 1. Clone [`saforem2/ezpz`](https://github.com/saforem2/ezpz) (if necessary) +# to `"${WORKING_DIR}/deps/ezpz/"` # -# [^1]: See: [`ezpz/bin/savejobenv`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/savejobenv) -#################### -ezpz_savejobenv() { - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/savejobenv" "$@" - file=$(mktemp) - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/savejobenv > "${file}" - source "${file}" || exit -} - -ezpz_getjobenv() { - # source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/getjobenv" "$@" - file=$(mktemp) - curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" - source "${file}" || exit -} - - -#################### -# ezpz_clone -# ################## -ezpz_clone() { +# 2. Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) +# - This provides `{ezpz_setup_python, ezpz_setup_alcf}` (called below) +# +# 3. Call `ezpz_setup_python`: +# - This will setup conda + virtual enviroment +# +# 4. Call `ezpz_setup_alcf`: +# - This will parse `$PBS_*` variables and build launch cmd +# +# 3. Call `_ezpz_install`: +# - Install ezpz from "${WORKING_DIR}/depz/ezpz/" +########################################### +ezpz_setup() { + # setup_alcf "$@" + # file=$(mktemp) + # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" + # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh ezdir="${WORKING_DIR}/deps/ezpz" - # if [[ ! -d "${ezdir}" ]]; then if [[ -d "${ezdir}" ]]; then echo "Found ezpz in ${ezdir}" else - mkdir -p "${WORKING_DIR}/deps" - git clone https://github.com/saforem2/ezpz "${WORKING_DIR}/deps/ezpz" + mkdir -p $(dirname "${ezdir}") + git clone https://github.com/saforem2/ezpz "${ezdir}" fi -} - -########################################### -# ezpz_install -# -# Install ezpz (if not installed) -########################################### -ezpz_install() { - ezloc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') - if [[ -z "${ezloc:-}" ]]; then + source "${ezdir}/src/ezpz/bin/utils.sh" || exit + ezpz_setup_python + ezpz_setup_alcf "$@" + ezpz_pip_loc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') + if [[ -z "${ezpz_pip_loc:-}" ]]; then printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" python3 -m pip install -e "${ezdir}" --require-virtualenv else - printf "[ezpz_install] Found ezpz @ %s\n" "${ezloc}" + printf "[ezpz_install] Found ezpz @ %s\n" "${ezpz_pip_loc}" fi } -########################################### -# ezpz_setup -# -# 1. source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/utils.sh) -# 2. call `setup_alcf` (provided by `ezpz/utils.sh` from 1.) -########################################### -ezpz_setup() { - # setup_alcf "$@" - # file=$(mktemp) - # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" - # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh - ezpz_clone - source "${WORKING_DIR}/deps/ezpz/src/ezpz/bin/utils.sh" || exit # && setup_alcf "$@" || exit - setup_python - ezpz_install - setup_alcf "$@" -} - ####################################################################### # ezpz_test: Run simple test to make sure all nodes in working order ####################################################################### @@ -880,80 +860,6 @@ update_ccl_env_vars_aurora() { export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH } -########################### -# Setup conda on Aurora -########################### -setup_conda_aurora() { - if [[ -z "${CONDA_PREFIX:-}" ]]; then - module use -a /soft/modulefiles ; module load frameworks/2024.1 - fi -} - -######################## -# Setup conda on Sirius -######################## -setup_conda_sirius() { - if [[ -z "${CONDA_PREFIX-}" && -z "${VIRTUAL_ENV-}" ]]; then - export MAMBA_ROOT_PREFIX=/lus/tegu/projects/PolarisAT/foremans/micromamba - shell_name=$(echo "${SHELL}" | tr "\/" "\t" | awk '{print $NF}') - eval "$("${MAMBA_ROOT_PREFIX}/bin/micromamba" shell hook --shell "${shell_name}")" - micromamba activate 2024-04-23 - else - echo "Found existing python at: $(which python3)" - fi -} - -######################## -# Setup conda on Polaris -######################## -setup_conda_polaris() { - # unset MPICH_GPU_SUPPORT_ENABLED - ###### check if CONDA_PREFIX non-empty ################ - if [[ -z "${CONDA_PREFIX-}" ]]; then - # if so, load the default conda/2024-04-29 - # module and activate base environment - module use /soft/modulefiles ; module load conda ; conda activate base - else - echo "Caught CONDA_PREFIX=${CONDA_PREFIX}" - fi -} - -######################## -# setup_venv_from_conda -# -# Build (if necessary) a virtual environment -# on top of the active conda and -# activate it. -# ###################### -setup_venv_from_conda() { - if [[ -z "${CONDA_PREFIX}" ]]; then - echo "!! No ${CONDA_PREFIX} found." # Exiting." - # exit 1 - else - echo "Found conda at: ${CONDA_PREFIX}" - CONDA_NAME=$(echo "${CONDA_PREFIX}" | tr '\/' '\t' | sed -E 's/mconda3|\/base//g' | awk '{print $NF}') - export CONDA_NAME - if [[ -z "${VIRTUAL_ENV}" ]]; then - echo "No VIRTUAL_ENV found in environment!" - echo " - Trying to setup from ${CONDA_PREFIX}" - export VENV_DIR="${WORKING_DIR}/venvs/${CONDA_NAME}" - echo " - Using VENV_DIR=${VENV_DIR}" - if [[ ! -f "${VENV_DIR}/bin/activate" ]]; then - printf "\n - Creating a new virtual env on top of %s in %s\n" "$(printBlue "${CONDA_NAME}")" "$(printGreen "${VENV_DIR}")" - mkdir -p "${VENV_DIR}" - python3 -m venv "${VENV_DIR}" --system-site-packages - source "${VENV_DIR}/bin/activate" || exit - elif [[ -f "${VENV_DIR}/bin/activate" ]]; then - echo " - Found existing venv, activating from $(printBlue "${VENV_DIR}")" - source "${VENV_DIR}/bin/activate" - else - printf "\n [!! %s]: Unable to locate %s\n" "$(printRed "ERROR")" "$(printMagenta "${VENV_DIR}/bin/activate")" - fi - fi - fi - -} - ########################################################## # Check that we can find the `.py` file we wish to launch ########################################################## @@ -970,79 +876,6 @@ check_executable() { fi } -############################################################################## -# setup_python -# -# 1. Setup `conda` -# - if `conda` nonempty, and `venv` empty, use `conda` to setup `venv`. -# - if `venv` nonempty, and `conda` empty, what do (???) -# - if `venv` nonempty and `conda` nonempty, use these -# - if `conda` empty and `venv` empty: -# - if `hostname == x4*`, we're on Aurora -# - if `hostname == x1*`, we're on Sunspot -# - if `hostname == x3*`, we're on Polaris -# - if `hostname == nid*`, we're on Perlmutter -# - otherwise, you're on you're own -# -# 2. Activate (creating, if necessary) a `venv` on top of `base` conda -# - use the $CONDA_PREFIX to create a venv in -# `Megatron-DeepSpeed/venvs/${CONDA_PREFIX}` -# - activate and use this -# -# 3. Print info about which python we're using -############################################################################## -setup_python() { - virtual_env="${VIRTUAL_ENV:-}" - conda_prefix="${CONDA_PREFIX:-}" - if [[ -z "${conda_prefix}" && -z "${virtual_env}" ]]; then - echo "No conda_prefix OR virtual_env found in environment..." - echo "Setting up conda..." - machine_name=$(get_machine_name) - echo "machine name: ${machine_name}" - # if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then - if [[ "${machine_name}" == "aurora" ]]; then - setup_conda_aurora - # elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then - elif [[ "${machine_name}" == "sunspot" ]]; then - setup_conda_sunspot - # elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then - elif [[ "${machine_name}" == "polaris" ]]; then - if [[ "${PBS_O_HOST:-}" == sirius* ]]; then - setup_conda_sirius - else - setup_conda_polaris - fi - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then - echo "Running on Perlmutter !!" - module load pytorch - source "${SLURM_SUBMIT_DIR}/venvs/perlmutter/pytorch-2.1.0-cu12/bin/activate" - else # ------------------------------------- [Unknown] ------------------- - echo "Unknown hostname $(hostname)" - exit 1 - fi - # # ----- [Perlmutter @ NERSC] ------------------------------------- - elif [[ -n "${conda_prefix}" && -z "${virtual_env}" ]]; then - echo "No virtual environment found." - echo "Using conda from: ${conda_prefix}" - echo "Setting up venv from ${CONDA_PROMPT_MODIFIER:-}" - setup_venv_from_conda - elif [[ -n "${virtual_env}" && -z "${conda_prefix}" ]]; then - echo "No conda found." - echo "Using virtual_env from: ${virtual_env}" - elif [[ -n "${virtual_env}" && -n "${conda_prefix}" ]]; then - echo "Using virtual_env: ${virtual_env} on top of conda from: ${conda_prefix}" - else - echo "Unable to setup python environment. Exiting" - exit 1 - fi - if [[ -z "${virtual_env}" ]]; then - setup_venv_from_conda - fi - pystr="Using: $(which python3)" - printf "[python] %s" "$(printMagenta "${pystr}")" - printf "\n" - export "PYTHON_EXEC=$(which python3)" -} ###################################################################### # `makeHostiles`: From 354e42067eac846fdda036527fb020ac6bdd4a92 Mon Sep 17 00:00:00 2001 From: Xinyu Chen Date: Fri, 12 Jul 2024 22:46:48 +0800 Subject: [PATCH 270/590] Extend test utilities to support more accelerators (#418) * Extend test utilities to support more accelerators * Add Intel Copyright --- tests/models/test_gpt_embedding.py | 1 + tests/models/test_gpt_model.py | 1 + tests/pipeline_parallel/test_schedules.py | 2 ++ tests/transformer/test_parallel_mlp.py | 1 + tests/unit_tests/test_utilities.py | 16 ++++++++++------ 5 files changed, 15 insertions(+), 6 deletions(-) diff --git a/tests/models/test_gpt_embedding.py b/tests/models/test_gpt_embedding.py index 8a188b0cd25..199f29dede5 100644 --- a/tests/models/test_gpt_embedding.py +++ b/tests/models/test_gpt_embedding.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. import pytest diff --git a/tests/models/test_gpt_model.py b/tests/models/test_gpt_model.py index b6888af5cc7..cf322908b3d 100644 --- a/tests/models/test_gpt_model.py +++ b/tests/models/test_gpt_model.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. import pytest diff --git a/tests/pipeline_parallel/test_schedules.py b/tests/pipeline_parallel/test_schedules.py index 64bd7c3ac64..72c2372ba47 100644 --- a/tests/pipeline_parallel/test_schedules.py +++ b/tests/pipeline_parallel/test_schedules.py @@ -1,3 +1,5 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. + import torch from tests.test_utilities import Utils from megatron.core import ModelParallelConfig diff --git a/tests/transformer/test_parallel_mlp.py b/tests/transformer/test_parallel_mlp.py index ac5e4b0cd5f..098f18a9d64 100644 --- a/tests/transformer/test_parallel_mlp.py +++ b/tests/transformer/test_parallel_mlp.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. import pytest diff --git a/tests/unit_tests/test_utilities.py b/tests/unit_tests/test_utilities.py index b35c77b58d2..68c6e6b55c5 100644 --- a/tests/unit_tests/test_utilities.py +++ b/tests/unit_tests/test_utilities.py @@ -1,21 +1,25 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. + import os import torch import megatron.core.parallel_state as ps +from deepspeed.accelerator import get_accelerator + class Utils: - world_size = torch.cuda.device_count() - rank = int(os.environ['LOCAL_RANK']) + world_size = int(os.getenv("WORLD_SIZE", '1')) + rank = int(os.getenv('LOCAL_RANK', '0')) @staticmethod def initialize_distributed(): print(f'Initializing torch.distributed with rank: {Utils.rank}, world_size: {Utils.world_size}') - torch.cuda.set_device(Utils.rank % torch.cuda.device_count()) + get_accelerator().set_device(Utils.rank % get_accelerator().device_count()) init_method = 'tcp://' master_ip = os.getenv('MASTER_ADDR', 'localhost') master_port = os.getenv('MASTER_PORT', '6000') init_method += master_ip + ':' + master_port - torch.distributed.init_process_group(backend='nccl', world_size=Utils.world_size, rank=Utils.rank, init_method=init_method) + torch.distributed.init_process_group(backend=get_accelerator().communication_backend_name(), world_size=Utils.world_size, rank=Utils.rank, init_method=init_method) @staticmethod def destroy_model_parallel(): @@ -23,8 +27,8 @@ def destroy_model_parallel(): torch.distributed.barrier() @staticmethod - def initialize_model_parallel(tensor_model_parallel_size = 1, pipeline_model_parallel_size = 1, virtual_pipeline_model_parallel_size = None, pipeline_model_parallel_split_rank = None): + def initialize_model_parallel(tensor_model_parallel_size = 1, pipeline_model_parallel_size = 1, sequence_parallel_size = 1, virtual_pipeline_model_parallel_size = None, pipeline_model_parallel_split_rank = None): ps.destroy_model_parallel() if not torch.distributed.is_initialized(): Utils.initialize_distributed() - ps.initialize_model_parallel(tensor_model_parallel_size, pipeline_model_parallel_size, virtual_pipeline_model_parallel_size, pipeline_model_parallel_split_rank) \ No newline at end of file + ps.initialize_model_parallel(tensor_model_parallel_size, pipeline_model_parallel_size, sequence_parallel_size, virtual_pipeline_model_parallel_size, pipeline_model_parallel_split_rank) \ No newline at end of file From 73252c0211fad1b4808e46f986341702b962e154 Mon Sep 17 00:00:00 2001 From: inkcherry Date: Fri, 12 Jul 2024 23:27:27 +0800 Subject: [PATCH 271/590] clear document (#395) --- examples_deepspeed/universal_checkpointing/README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/examples_deepspeed/universal_checkpointing/README.md b/examples_deepspeed/universal_checkpointing/README.md index 14169c9e228..a0a1df5b8a1 100644 --- a/examples_deepspeed/universal_checkpointing/README.md +++ b/examples_deepspeed/universal_checkpointing/README.md @@ -77,6 +77,8 @@ Please see the corresponding [pull request](https://github.com/microsoft/Megatro Combining sequence parallelism with data parallelism is another good use case for universal checkpointing, see [sp pull request](https://github.com/microsoft/DeepSpeed/pull/4752) for example and visualization of matching loss values. +Notes: The model weights using the ```--no-pipeline-parallel``` parameter and the model weights not using the ```--no-pipeline-parallel``` parameter are currently not supported for mutual conversion. + ### TensorBoard Log Analysis The Universal Checkpointing example includes a TensorBoard analysis script that will generate `csv` files and `png` plots across the unviersal checkpointing training steps for comparison of training and validation loss curves. From 0971e684246368ac74395162ec7c5df95635f87f Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Mon, 15 Jul 2024 18:44:23 +0300 Subject: [PATCH 272/590] add PyTorch profiler support (#414) * Update arguments.py * Update training.py * Create profiler.py * add copyrights * Update profiler.py * add copyrights * Update help * add copyrights --- megatron/arguments.py | 19 +++++++++++++++ megatron/initialize.py | 1 + megatron/profiler.py | 54 ++++++++++++++++++++++++++++++++++++++++++ megatron/training.py | 6 +++++ 4 files changed, 80 insertions(+) create mode 100644 megatron/profiler.py diff --git a/megatron/arguments.py b/megatron/arguments.py index 88898548502..de348504c7d 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """Megatron arguments.""" @@ -44,6 +45,7 @@ def parse_args(extra_args_provider=None, ignore_unknown_args=False): parser = _add_inference_args(parser) parser = _add_transformer_engine_args(parser) parser = _add_retro_args(parser) + parser = _add_profiler_args(parser) # Custom arguments. if extra_args_provider is not None: @@ -1541,3 +1543,20 @@ def _add_distillation_args(parser): help='Directory containing a teacher model checkpoint.') return parser + + +def _add_profiler_args(parser): + group = parser.add_argument_group(title='profiling configuration') + + group.add_argument("--profile", + type=str, + default=None, + choices=['pt', 'pt-full'], + help="Enable profiling, pt-full gives call stack compared to pt") + + group.add_argument("--profile_steps", + type=str, + default='2,3', + help="Which steps to profile. Format: ,") + + return parser diff --git a/megatron/initialize.py b/megatron/initialize.py index 2a9cb3d90fe..538f7fc456f 100644 --- a/megatron/initialize.py +++ b/megatron/initialize.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """Megatron initialization.""" diff --git a/megatron/profiler.py b/megatron/profiler.py new file mode 100644 index 00000000000..c98096482ab --- /dev/null +++ b/megatron/profiler.py @@ -0,0 +1,54 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. + +import torch + +on_step_begin = [] +on_step_end = [] + +def trigger(phase): + [f() for f in phase] + +def setup_profiler(args, device): + if args.profile is None: + return + + start_step, end_step = map(int, args.profile_steps.split(',')) + active_steps = end_step - start_step + 1 + cur_step = 0 + + def on_step_begin_fn(): + nonlocal cur_step + cur_step = cur_step + 1 + on_step_begin.append(on_step_begin_fn) + + def when(cond, clbk): + def fn(): + if cond(): + clbk() + return fn + + def is_start_step(): + return cur_step == start_step + + def is_end_step(): + return cur_step == end_step + + def is_capture_step(): + return cur_step >= start_step and cur_step <= end_step + + if args.profile.startswith('pt'): + schedule = torch.profiler.schedule(wait=0, warmup=0, active=active_steps, repeat=1) + activities = [torch.profiler.ProfilerActivity.CPU] + activities.extend([torch.profiler.ProfilerActivity.HPU] if device.startswith("hpu") else []) + activities.extend([torch.profiler.ProfilerActivity.CUDA] if device.startswith("cuda") else []) + full = args.profile == 'pt-full' + + profiler = torch.profiler.profile( + schedule=schedule, + activities=activities, + on_trace_ready=torch.profiler.tensorboard_trace_handler(args.tensorboard_dir, use_gzip=True), + with_stack=full) + + on_step_begin.append(when(is_start_step, profiler.start)) + on_step_end.append(when(is_capture_step, profiler.step)) + on_step_end.append(when(is_end_step, profiler.stop)) diff --git a/megatron/training.py b/megatron/training.py index 19b8a6c71f6..c422c6fc553 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain utilities.""" @@ -43,6 +44,7 @@ from megatron.utils import report_memory, throughput_calculator, checkpoint_throughput_calculator, update_rotary_pos_emb from megatron.model.vision.knn_monitor import compute_feature_bank from megatron.arguments import core_transformer_config_from_args +from megatron.profiler import setup_profiler, trigger, on_step_begin, on_step_end import deepspeed from deepspeed.accelerator import get_accelerator @@ -1161,6 +1163,8 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, # Write args to tensorboard write_args_to_tensorboard() + setup_profiler(args, get_accelerator().device_name()) + if args.random_ltd: # random-ltd requires different randomness on each rank import random @@ -1191,6 +1195,7 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, while iteration < args.train_iters and (args.train_tokens is None or \ args.consumed_train_tokens < args.train_tokens): + trigger(on_step_begin) update_num_microbatches(args.consumed_train_samples) if args.deepspeed: # inform deepspeed of any batch size changes @@ -1311,6 +1316,7 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, torch.distributed.barrier() print_datetime('exiting program at iteration {}'.format(iteration)) sys.exit() + trigger(on_step_end) return iteration From 2ac4fb09027dacd6da7122545759e8c794455beb Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 15 Jul 2024 16:19:19 -0500 Subject: [PATCH 273/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 101 +++++++++++++++++++++++++++++++++--------------- 1 file changed, 69 insertions(+), 32 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index ba74d559b85..5e867679493 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -346,7 +346,7 @@ setupLauncher() { else if [[ -n "${DIST_LAUNCH}" ]]; then - LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" + LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" export LAUNCHER="${LAUNCHER}" else echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" @@ -397,6 +397,51 @@ get_batch_size_on_polaris() { echo "${mbs}" } +get_grad_acc_steps_on_aurora() { + if [[ "$#" == 0 ]]; then + local hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" + elif [[ "$#" == 1 ]]; then + local hf="$1" + else + echo "Expected exactly 0 or 1 arguments, received: $#" + exit 1 + fi + nhosts=$(wc -l < "${hf}") + # NOTE: + # | nhosts | nhosts | GAS | + # |:----------------:|:----------:|:-----:| + # | n <= 32 | [0, 32] | 8 | + # | 32 < n < 64 | (32, 64) | 4 | + # | 64 <= n < 128 | [64, 128) | 2 | + # | 128 <= n < inf | [128, inf) | 1 | + if [[ 128 -le "${nhosts}" ]]; then + gas=1 + elif [[ 64 -le "${nhosts}" && "${nhosts}" -lt 128 ]]; then + gas=2 + # 32 <= nhosts < 64 + elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 64 ]]; then + gas=4 + # nhosts <= 16 + # elif [[ "${nhosts}" -lt 32 ]]; then + # gas=8 + else + gas=8 + fi + # if [[ "${nhosts}" > 512 ]]; then + # gas=1 + # # elif [[ 256 <= "${nhosts}" && "${nhosts}" <= 512 ]]; then + # elif [[ 256 -le "${nhosts}" && "${nhosts}" -le 512 ]]; then + # elif [[ 128 < "${nhosts}" && "${nhosts}" <= 256 "${nhosts}" ]]; then + # gas=2 + # # elif [[ 128 < "${nhosts}" && "${nhosts}" < 256 ]]; then + # elif [[ 64 <= "${nhosts}" && "${nhosts}" <= 128 -lt "${nhosts}" ]]; then + # gas=2 + # else + # gas=8 + # fi + echo "${gas}" +} + ############################################################################## # setParams @@ -420,8 +465,11 @@ setParams() { export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS + # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS + export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora)}" MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + export CCL_PROCESS_LAUNCHER=pmix + export CCL_ATL_TRANSPORT=mpi ###################################################################### # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? @@ -479,7 +527,7 @@ setParams() { export SP="${SP:-1}" export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" - export OPT="${OPT:-adamw}" + export OPT="${OPT:-adamwschedulefree}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") if [[ -z "${NGPU_PER_HOST:-}" ]]; then @@ -762,15 +810,15 @@ make_data() { ############################################################################## install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" - echo "Ensuring all dependencies from ${depsfile} installed..." + echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." python3 -m pip install -r "${depsfile}" --require-virtualenv 1> /dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then mn=$(get_machine_name) - if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then - install_deepspeed_for_xpu || exit - fi - printf "!! No 'deepspeed' command found on %s" "${mn}" - printf "!! No deepsepeed in $(which python3)" + # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then + # install_deepspeed_for_xpu || exit + # fi + printf "[install_dependencies] No 'deepspeed' command found on %s" "${mn}" + printf "[install_dependencies] !! No deepsepeed in $(which python3)" fi } @@ -785,38 +833,18 @@ install_dependencies() { # 3. Install into virtual environment ###################################################################### install_deepspeed_for_xpu() { + # python3 -m pip install "torch==2.1.0.post2" torchvision==0.16.0.post2 torchaudio==2.1.0.post2 intel-extension-for-pytorch==2.1.30.post0 oneccl_bind_pt==2.1.300+xpu --extra-index-url "https://pytorch-extension.intel.com/release-whl/stable/xpu/us/" echo "Building + Installing DeepSpeed on $(hostname)" outdir="${WORKING_DIR}/deps/DeepSpeed" mkdir -p "${outdir}" git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" cd "${outdir}" || exit - echo "!! pwd: $(pwd)" - # git remote add yizhou_ds https://github.com/YizhouZ/DeepSpeed.git - # git fetch yizhou_ds - # git checkout yizhou/kernel_path + echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1> /dev/null python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1> /dev/null python setup.py develop 1> /dev/null cd "${WORKING_DIR}" - echo "!! pwd: $(pwd)" -} - - -######################################################## -# Setup / activate conda environment(s) -######################################################## - -########################### -# Setup conda on Sunspot -########################### -setup_conda_sunspot() { - ###### check if CONDA_PREFIX non-empty ################ - if [[ -z "${CONDA_PREFIX:-}" ]]; then - module use /soft/preview-modulefiles/24.086.0 - module load frameworks/2024.04.15.002.lua - # module use /soft/preview-modulefiles/24.086.0 ; module load frameworks/2024.04.15.002.lua - # source "${WORKING_DIR}/ALCF/sunspot-env-2024-q2.sh" - fi + echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" } @@ -1182,6 +1210,15 @@ EOT ############################################### # Helper functions for printing colored text ############################################### +RESET="\e[0m" +BLACK="\e[1;30m" +RED="\e[1;31m" +GREEN="\e[1;32m" +YELLOW="\e[1;33m" +BLUE="\e[1;34m" +CYAN="\e[1;35m" +WHITE="\e[1;36m" + printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" } From 4876eb89f1447d7557f6b3dc8694b6ee8844c177 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Jul 2024 08:12:17 -0500 Subject: [PATCH 274/590] Update `ALCF/helpers.sh` on Polaris --- ALCF/helpers.sh | 37 +++++++++++++++++++++++++++---------- 1 file changed, 27 insertions(+), 10 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5e867679493..76cca787ff4 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -66,24 +66,36 @@ helpers_main() { # # - Explicitly, this will: # - Identify the machine we're on +# # - Setup `python` # 1. Load `conda` # 2. Setup `venv` on top of `conda` +# # - Ensure all dependencies are installed +# # - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) -# 1. Additionally, call `source deps/ezpz/src/ezpz/bin/savejobenv`, -# which will automatically build a `alias launch=mpiexec ...` -# according to the specifics of our active job. +# - Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) +# - This provides `{ezpz_setup_python, ezpz_setup_alcf}` (called below) +# # - Set runtime options +# # - Build `deepspeed_config.json` +# # - Build {logs, checkpoints, etc} dirs, named according to specifics of # current run +# # - Specify additional `deepspeed` arguments +# # - Ensure executable exists at expected path +# # - Setup data + tokenizer via `TOKENIZER_TYPE` +# # - Print job info +# # - Save `.env` to `CKPT_DIR` for safe keeping +# # - Check that we're not already running, and if so, exit. +# # - Setup run command to be executed. ############################################################################## setup() { @@ -344,9 +356,13 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - if [[ -n "${DIST_LAUNCH}" ]]; then - LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" + local mn=$(get_machine_name) + if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then + LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" + else + LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" + fi export LAUNCHER="${LAUNCHER}" else echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" @@ -980,12 +996,13 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # shellcheck source=./data-lists/sunspot/books.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma.txt" elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] - if [[ "${PBS_O_HOST}" == sirius* ]]; then # -------------------[SIRIUS] - # shellcheck source=./data-lists/sirius/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma.txt" - elif [[ "${PBS_O_HOST}" == polaris* ]]; then # ---------------[POLARIS] - # shellcheck source=./data-lists/polaris/books.txt + polaris_match=$(echo $PBS_NODEFILE | grep 'polaris') + if [[ -n "${polaris_match}" ]]; then + # shellcheck source=./data-lists/polaris/dolma.txt dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma.txt" + else + # shellcheck source=./data-lists/sirius/dolma.txt + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma.txt" fi elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" From 7385e3bd12e51d0805a108538c0d146839c7936d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Jul 2024 09:28:43 -0500 Subject: [PATCH 275/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 76cca787ff4..789ce933f41 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -543,7 +543,7 @@ setParams() { export SP="${SP:-1}" export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" - export OPT="${OPT:-adamwschedulefree}" + export OPT="${OPT:-adamw}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") if [[ -z "${NGPU_PER_HOST:-}" ]]; then From 5f5bbd40d022eb0062178b2deb8ea10004b13fb2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 16 Jul 2024 10:26:44 -0500 Subject: [PATCH 276/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 53da0dbbdb3..04018d7918d 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -158,8 +158,12 @@ def model_provider(pre_process=True, post_process=True): tbdir = args.tensorboard_dir # tbdir = args.getattr('tensorboard_dir', None) if tbdir is not None: - log.info(f'Patching tensorboard from {tbdir}') - wandb.tensorboard.patch(root_logdir=tbdir) + try: + log.info(f'Patching tensorboard from {tbdir}') + wandb.tensorboard.patch(root_logdir=tbdir) + except ValueError as exc: + log.exception(exc) + log.warning('Continuing without patching tensorboard!') wandb.run.config.update({'num_params': num_params}) if "args" not in wandb.run.config: log.info( From 73029edfa88f9b6fd2d1aed271a964ae87131dbb Mon Sep 17 00:00:00 2001 From: billishyahao Date: Wed, 17 Jul 2024 07:19:06 +0800 Subject: [PATCH 277/590] [Wandb] Refine wandb logging function (#416) * Refine wandb logging function * Address comments * enable user to specify wandb local save dir * Update and fix comments * Update --- .../pretrain_llama2_distributed.sh | 13 +- megatron/__init__.py | 1 + megatron/arguments.py | 6 + megatron/global_vars.py | 47 ++++ megatron/timers.py | 2 +- megatron/training.py | 230 +++++++++++------- 6 files changed, 204 insertions(+), 95 deletions(-) diff --git a/examples_deepspeed/pretrain_llama2_distributed.sh b/examples_deepspeed/pretrain_llama2_distributed.sh index f275ea636a5..4c790e8c191 100755 --- a/examples_deepspeed/pretrain_llama2_distributed.sh +++ b/examples_deepspeed/pretrain_llama2_distributed.sh @@ -41,6 +41,17 @@ GRAD_CLIP=1 # activation_checkpoint="true" activation_checkpoint="false" +LOG_TO_WANDB=0 +WANDB_ARGS= +if [ $LOG_TO_WANDB -eq 1 ] +then +WANDB_ARGS="\ + --wandb-project pretrain-llama2 \ + --wandb-exp-name exp0 \ + --wandb-save-dir ${BASE_PATH}/wandb \ + " +fi + # Below configuration required for llama model as per llama paper # --no-query-key-layer-scaling \ # --attention-dropout 0 \ @@ -53,7 +64,6 @@ activation_checkpoint="false" ###################################### - cat < $DS_CONFIG { "train_batch_size" : $GLOBAL_BATCH_SIZE, @@ -132,4 +142,5 @@ torchrun $DISTRIBUTED_ARGS \ --normalization rmsnorm \ --disable-bias-linear \ --num-key-value-heads $NUM_KV_HEADS \ + $WANDB_ARGS \ $ds_args diff --git a/megatron/__init__.py b/megatron/__init__.py index d92a279ec67..4a7fe2cd828 100644 --- a/megatron/__init__.py +++ b/megatron/__init__.py @@ -9,6 +9,7 @@ from .global_vars import update_num_microbatches from .global_vars import get_tokenizer from .global_vars import get_tensorboard_writer +from .global_vars import get_wandb_writer from .global_vars import get_adlr_autoresume from .global_vars import get_timers from .initialize import initialize_megatron diff --git a/megatron/arguments.py b/megatron/arguments.py index de348504c7d..30c3d669d70 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -748,6 +748,12 @@ def _add_logging_args(parser): group.add_argument('--log-world-size-to-tensorboard', action='store_true', help='Enable world size logging to tensorboard.') + group.add_argument('--wandb-project', type=str, default='', + help='The wandb project name. Ignore wandb by default.') + group.add_argument('--wandb-exp-name', type=str, default='', + help='The wandb experiment name.') + group.add_argument('--wandb-save-dir', type=str, default='', + help='Path to save the wandb results locally.') return parser diff --git a/megatron/global_vars.py b/megatron/global_vars.py index cb284b3c34f..3f9d6fd66b4 100644 --- a/megatron/global_vars.py +++ b/megatron/global_vars.py @@ -16,6 +16,7 @@ _GLOBAL_NUM_MICROBATCHES_CALCULATOR = None _GLOBAL_TOKENIZER = None _GLOBAL_TENSORBOARD_WRITER = None +_GLOBAL_WANDB_WRITER = None _GLOBAL_ADLR_AUTORESUME = None _GLOBAL_TIMERS = None _GLOBAL_SIGNAL_HANDLER = None @@ -56,6 +57,12 @@ def get_tensorboard_writer(): return _GLOBAL_TENSORBOARD_WRITER +def get_wandb_writer(): + """Return wandb writer. It can be None so no need + to check if it is initialized.""" + return _GLOBAL_WANDB_WRITER + + def get_adlr_autoresume(): """ADLR autoresume object. It can be None so no need to check if it is initialized.""" @@ -91,6 +98,7 @@ def set_global_variables(args): _build_num_microbatches_calculator(args) _ = _build_tokenizer(args) _set_tensorboard_writer(args) + _set_wandb_writer(args) _set_adlr_autoresume(args) _set_timers(args) @@ -152,6 +160,45 @@ def _set_tensorboard_writer(args): 'no TensorBoard logs will be written.', flush=True) +def _set_wandb_writer(args): + """Set wandb writer.""" + global _GLOBAL_WANDB_WRITER + _ensure_var_is_not_initialized(_GLOBAL_WANDB_WRITER, + 'wandb writer') + getattr(args, 'wandb_project', '') + getattr(args, 'wandb_exp_name', '') + + if args.rank == (args.world_size - 1): + if args.wandb_project == '' or \ + args.wandb_exp_name == '': + print('WARNING: WANDB writing requested but no legit wandb ' + 'project or experiment name provided, ' + 'therefore WANDB logs will be written ' + 'according to random generated project or experiment name.', flush=True) + + try: + import wandb + except (ImportError, ModuleNotFoundError): + print('WARNING: WANDB writing requested but is not ' + 'available (try to pip install wandb to solve it), ' + 'no WANDB logs will be written.', flush=True) + return + + if args.wandb_save_dir: + save_dir = args.wandb_save_dir + else: + # Defaults to the save dir. + save_dir = os.path.join(args.save, 'wandb') + wandb_kwargs = { + 'dir': save_dir, + 'name': args.wandb_exp_name, + 'project': args.wandb_project, + 'config': vars(args)} + os.makedirs(wandb_kwargs['dir'], exist_ok=True) + wandb.init(**wandb_kwargs) + _GLOBAL_WANDB_WRITER = wandb + + def _set_adlr_autoresume(args): """Initialize ADLR autoresume.""" global _GLOBAL_ADLR_AUTORESUME diff --git a/megatron/timers.py b/megatron/timers.py index 384c7c37a3c..53ff9be98ed 100644 --- a/megatron/timers.py +++ b/megatron/timers.py @@ -303,7 +303,7 @@ def write(self, names, writer, iteration, normalizer=1.0, assert normalizer > 0.0 name_to_min_max_time = self._get_global_min_max_time( names, reset, barrier, normalizer) - if writer is not None: + if writer.is_enabled(): for name in name_to_min_max_time: _, max_time = name_to_min_max_time[name] writer.add_scalar(name + '-time', max_time, iteration) diff --git a/megatron/training.py b/megatron/training.py index c422c6fc553..697d62f7b8f 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -8,16 +8,22 @@ import sys import time import json +try: + import wandb +except (ImportError, ModuleNotFoundError): + wandb = None # The earliest we can measure the start time. _TRAIN_START_TIME = time.time() import torch from collections import OrderedDict from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP +from enum import Enum from megatron import get_args from megatron import get_signal_handler from megatron import get_timers from megatron import get_tensorboard_writer +from megatron import get_wandb_writer from megatron import get_current_global_batch_size from megatron import get_num_microbatches from megatron import is_last_rank @@ -54,12 +60,6 @@ from deepspeed import comm as dist -try: - import wandb -except (ImportError, ModuleNotFoundError): - wandb = None - - def print_datetime(string): """Note that this call will sync across all ranks.""" torch.distributed.barrier() @@ -787,6 +787,65 @@ def train_step(forward_step_func, data_iterator, return {}, skipped_iter, grad_norm, num_zeros_in_grad +class InteropLoggingTool(Enum): + TENSORBOARD = 1 + WANDB = 2 + + +class interop_tool_logger: + def __init__(self, tb_writer=None, wandb_writer=None): + self.tb_writer = tb_writer + self.wandb_writer = wandb_writer + self.custom_x_axis = [] + self.custom_y_axis = {} + self.args = get_args() + if not hasattr(self.args, "logger_iteration"): + self.args.logger_iteration = 1 + + def is_enabled(self): + return self.tb_writer or self.wandb_writer + + def add_scalar(self, key, scalar_value, step, custom_step_name=None, \ + tool_list=[InteropLoggingTool.TENSORBOARD, InteropLoggingTool.WANDB]): + if self.tb_writer and \ + InteropLoggingTool.TENSORBOARD in tool_list: + self.tb_writer.add_scalar(key, scalar_value, step) + + if self.wandb_writer and \ + InteropLoggingTool.WANDB in tool_list: + if not custom_step_name: + self.wandb_writer.log({key: scalar_value}, step=step) + if self.args.logger_iteration < step: + # Updating iteration + self.args.logger_iteration = step + + else: + if custom_step_name not in self.custom_x_axis: + self.custom_x_axis.append(custom_step_name) + wandb.define_metric(custom_step_name) + + if key not in self.custom_y_axis: + self.custom_y_axis[key] = custom_step_name + wandb.define_metric(key, step_metric=custom_step_name) + + self.wandb_writer.log({key: scalar_value, custom_step_name: step}, \ + step=self.args.logger_iteration) + + + def add_scalar_to_tb(self, key, scalar_value, step): + return self.add_scalar(key, scalar_value, step, None, [InteropLoggingTool.TENSORBOARD]) + + def add_scalar_to_wandb(self, key, scalar_value, step, custom_step_name=None): + return self.add_scalar(key, scalar_value, step, custom_step_name, [InteropLoggingTool.WANDB]) + + def add_images(self, key, img_tensor, step=None): + if self.tb_writer: + self.tb_writer.add_images(key, img_tensor, step) + + if self.wandb_writer: + self.wandb_writer.log({key: wandb.Image(img_tensor)}, step) + + def training_log(loss_dict, total_loss_dict, learning_rate, iteration, loss_scale, report_memory_flag, skipped_iter, grad_norm, params_norm, num_zeros_in_grad, @@ -794,8 +853,10 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, """Log training information such as losses, timing, ....""" args = get_args() timers = get_timers() - writer = get_tensorboard_writer() - + writer = interop_tool_logger(tb_writer=get_tensorboard_writer(), \ + wandb_writer=get_wandb_writer()) + x_axis_samples = 'Samples' + x_axis_tokens = 'Tokens' # Advanced, skipped, and Nan iterations. advanced_iters_key = 'advanced iterations' skipped_iters_key = 'skipped iterations' @@ -866,80 +927,80 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, (iteration % args.tensorboard_log_interval == 0): timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) - if writer and (iteration % args.tensorboard_log_interval == 0): - writer.add_scalar('steps-vs-samples/y=steps,x=samples', iteration, args.consumed_train_samples) + if writer.is_enabled() and (iteration % args.tensorboard_log_interval == 0): + writer.add_scalar('steps-vs-samples/y=steps,x=samples', iteration, args.consumed_train_samples, x_axis_samples) writer.add_scalar('steps-vs-samples/y=samples,x=steps', args.consumed_train_samples, iteration) - writer.add_scalar('steps-vs-tokens/y=steps,x=tokens', iteration, args.consumed_train_tokens) + writer.add_scalar('steps-vs-tokens/y=steps,x=tokens', iteration, args.consumed_train_tokens, x_axis_tokens) writer.add_scalar('steps-vs-tokens/y=tokens,x=steps', args.consumed_train_tokens, iteration) if args.log_learning_rate_to_tensorboard: writer.add_scalar('learning-rate/learning-rate', learning_rate, iteration) writer.add_scalar('learning-rate/learning-rate vs samples', learning_rate, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('learning-rate/learning-rate vs tokens', learning_rate, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.log_batch_size_to_tensorboard: writer.add_scalar('batch-size/batch-size', batch_size, iteration) writer.add_scalar('batch-size/batch-size vs samples', batch_size, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('batch-size/batch-size vs tokens', batch_size, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) for key in loss_dict: writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) writer.add_scalar(f"lm-loss-training/{key}" + ' vs samples', loss_dict[key], - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar(f"lm-loss-training/{key}" + ' vs tokens', loss_dict[key], - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: writer.add_scalar('loss-scale/loss-scale', loss_scale, iteration) writer.add_scalar('loss-scale/loss-scale vs samples', loss_scale, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('loss-scale/loss-scale vs tokens', loss_scale, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.log_world_size_to_tensorboard: writer.add_scalar('world-size/world-size', args.world_size, iteration) writer.add_scalar('world-size/world-size vs samples', args.world_size, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('world-size/world-size vs tokens', args.world_size, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if grad_norm is not None: writer.add_scalar('grad-norm/grad-norm', grad_norm, iteration) writer.add_scalar('grad-norm/grad-norm vs samples', grad_norm, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('grad-norm/grad-norm vs tokens', grad_norm, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if num_zeros_in_grad is not None: writer.add_scalar('num-zeros/num-zeros', num_zeros_in_grad, iteration) writer.add_scalar('num-zeros/num-zeros vs samples', num_zeros_in_grad, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('num-zeros/num-zeros vs tokens', num_zeros_in_grad, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if params_norm is not None: writer.add_scalar('params-norm/params-norm', params_norm, iteration) writer.add_scalar('params-norm/params-norm vs samples', params_norm, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('params-norm/params-norm vs tokens', params_norm, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if hasattr(args, 'actual_seq_length'): writer.add_scalar('seqlen/actual_seq_length', args.actual_seq_length, iteration) writer.add_scalar('seqlen/actual_seq_length vs samples', args.actual_seq_length, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('seqlen/actual_seq_length vs tokens', args.actual_seq_length, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: writer.add_scalar('seqlen/curriculum_seqlen', args.curriculum_seqlen, iteration) writer.add_scalar('seqlen/curriculum_seqlen vs samples', args.curriculum_seqlen, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('seqlen/curriculum_seqlen vs tokens', args.curriculum_seqlen, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.random_ltd: writer.add_scalar('seqlen/random_ltd_reserved_length', args.random_ltd_reserved_length, iteration) writer.add_scalar('seqlen/random_ltd_reserved_length vs samples', args.random_ltd_reserved_length, - args.consumed_train_samples) + args.consumed_train_samples, x_axis_samples) writer.add_scalar('seqlen/random_ltd_reserved_length vs tokens', args.random_ltd_reserved_length, - args.consumed_train_tokens) + args.consumed_train_tokens, x_axis_tokens) if args.log_memory_to_tensorboard: mem_stats = torch.cuda.memory_stats() writer.add_scalar( @@ -1002,19 +1063,19 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, group=mpu.get_pipeline_model_parallel_group()) # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) - if writer and is_last_rank(): - writer.add_scalar('optimizer/variance_l2 vs tokens', opt_stats[0]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_l2 vs tokens', opt_stats[1]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_l2 vs tokens', opt_stats[2]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_l2 vs tokens', opt_stats[3]**0.5, args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_l1 vs tokens', opt_stats[4], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_l1 vs tokens', opt_stats[5], args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_l1 vs tokens', opt_stats[6], args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_l1 vs tokens', opt_stats[7], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_abs_max vs tokens', opt_stats_2[0], args.consumed_train_tokens) - writer.add_scalar('optimizer/variance_sqrt_abs_max vs tokens', opt_stats_2[1], args.consumed_train_tokens) - writer.add_scalar('optimizer/momentum_abs_max vs tokens', opt_stats_2[2], args.consumed_train_tokens) - writer.add_scalar('optimizer/weight_abs_max vs tokens', opt_stats_2[3], args.consumed_train_tokens) + if writer.is_enabled() and is_last_rank(): + writer.add_scalar('optimizer/variance_l2 vs tokens', opt_stats[0]**0.5, args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/variance_sqrt_l2 vs tokens', opt_stats[1]**0.5, args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/momentum_l2 vs tokens', opt_stats[2]**0.5, args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/weight_l2 vs tokens', opt_stats[3]**0.5, args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/variance_l1 vs tokens', opt_stats[4], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/variance_sqrt_l1 vs tokens', opt_stats[5], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/momentum_l1 vs tokens', opt_stats[6], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/weight_l1 vs tokens', opt_stats[7], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/variance_abs_max vs tokens', opt_stats_2[0], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/variance_sqrt_abs_max vs tokens', opt_stats_2[1], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/momentum_abs_max vs tokens', opt_stats_2[2], args.consumed_train_tokens, x_axis_tokens) + writer.add_scalar('optimizer/weight_abs_max vs tokens', opt_stats_2[3], args.consumed_train_tokens, x_axis_tokens) writer.add_scalar('optimizer/variance_l2', opt_stats[0]**0.5, iteration) writer.add_scalar('optimizer/variance_sqrt_l2', opt_stats[1]**0.5, iteration) @@ -1046,34 +1107,29 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, tokens_per_sec_per_replica = tokens_per_sec / args.data_parallel_size tokens_per_gpu_per_second = tokens_per_sec / args.world_size tokens_per_gpu_per_second_per_replica = tokens_per_gpu_per_second / args.data_parallel_size - if wandb is not None and getattr(wandb, 'run', None) is not None: - assert wandb.run is not None - wandb_metrics = { - 'throughput/iteration-time': elapsed_time_per_iteration, # 1000 ms / s - 'throughput/samples_per_sec': samples_per_sec, - 'throughput/samples_per_sec_per_replica': samples_per_sec_per_replica, - 'throughput/tokens_per_sec': tokens_per_sec, - 'throughput/tokens_per_sec_per_replica': tokens_per_sec_per_replica, - 'throughput/tokens_per_gpu_per_sec': tokens_per_gpu_per_second, - 'throughput/tokens_per_gpu_per_sec_per_replica': tokens_per_gpu_per_second_per_replica, - 'throughput/tflops': tflops, - 'throughput/approx_params_in_billions': approx_parameters_in_billions, - 'throughput/elapsed_ms_per_iteration': elapsed_time_per_iteration, - 'throughput/iteration': iteration, - } + + if writer.is_enabled(): + writer.add_scalar_to_wandb('throughput/iteration-time', elapsed_time_per_iteration, iteration) # 1000 ms / s + writer.add_scalar_to_wandb('throughput/samples_per_sec', samples_per_sec, iteration) + writer.add_scalar_to_wandb('throughput/samples_per_sec_per_replica', samples_per_sec_per_replica, iteration) + writer.add_scalar_to_wandb('throughput/tokens_per_sec', tokens_per_sec, iteration) + writer.add_scalar_to_wandb('throughput/tokens_per_sec_per_replica', tokens_per_sec_per_replica, iteration) + writer.add_scalar_to_wandb('throughput/tokens_per_gpu_per_sec', tokens_per_gpu_per_second, iteration) + writer.add_scalar_to_wandb('throughput/tokens_per_gpu_per_sec_per_replica', tokens_per_gpu_per_second_per_replica, iteration) + writer.add_scalar_to_wandb('throughput/tflops', tflops, iteration) + writer.add_scalar_to_wandb('throughput/approx_params_in_billions', approx_parameters_in_billions, iteration) + writer.add_scalar_to_wandb('throughput/elapsed_ms_per_iteration', elapsed_time_per_iteration, iteration) if loss_dict is not None: - wandb_metrics |= { - f'loss/{k}': v for k, v in loss_dict.items() - } - wandb_metrics |= {'loss/iteration': iteration} - if writer: + for k, v in loss_dict.items(): + writer.add_scalar_to_wandb(f'loss/{k}', v, iteration) + if args.log_timers_to_tensorboard: writer.add_scalar('iteration-time/iteration-time', elapsed_time_per_iteration, iteration) writer.add_scalar('iteration-time/iteration-time vs samples', - elapsed_time_per_iteration, args.consumed_train_samples) + elapsed_time_per_iteration, args.consumed_train_samples, x_axis_samples) writer.add_scalar('iteration-time/iteration-time vs tokens', - elapsed_time_per_iteration, args.consumed_train_tokens) + elapsed_time_per_iteration, args.consumed_train_tokens, x_axis_tokens) log_string = ' iteration {:8d}/{:8d} |'.format( iteration, args.train_iters) log_string += ' consumed samples: {:12d} |'.format( @@ -1084,21 +1140,7 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, elapsed_time_per_iteration * 1000.0) log_string += ' learning rate: {:.3E} |'.format(learning_rate) log_string += ' global batch size: {:5d} |'.format(batch_size) - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb_metrics |= { - 'training/iteration': iteration, - 'training/iteration_time': elapsed_time_per_iteration, - 'training/iteration_time_vs_tokens': ( - (elapsed_time_per_iteration - / args.consumed_train_tokens) - ), - 'training/iteration_time_vs_samples': ( - (elapsed_time_per_iteration - / args.consumed_train_samples), - ), - 'training/consumed_samples': args.consumed_train_samples, - 'training/consumed_tokens': args.consumed_train_tokens, - } + for key in total_loss_dict: if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: @@ -1107,8 +1149,6 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, if avg > 0.0: log_string += ' {}: {:.6E} |'.format(key, avg) total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) - if wandb is not None and getattr(wandb, 'run', None) is not None: - wandb.log(wandb_metrics) if loss_scale is not None: log_string += ' loss scale: {:.1f} |'.format(loss_scale) if grad_norm is not None: @@ -1433,9 +1473,11 @@ def evaluate_and_print_results(prefix, forward_step_func, """Helper function to evaluate and dump results on screen.""" args = get_args() if write_to_tensorboard: - writer = get_tensorboard_writer() + writer = interop_tool_logger(tb_writer=get_tensorboard_writer(), wandb_writer=get_wandb_writer()) else: - writer = None + writer = interop_tool_logger() + x_axis_samples = 'Samples' + x_axis_tokens = 'Tokens' total_loss_dict, collected_non_loss_data = evaluate( forward_step_func, data_iterator, model, @@ -1445,26 +1487,28 @@ def evaluate_and_print_results(prefix, forward_step_func, string += '{} value: {:.6E} | '.format(key, total_loss_dict[key].item()) ppl = math.exp(min(20, total_loss_dict[key].item())) string += '{} PPL: {:.6E} | '.format(key, ppl) - if writer and is_last_rank(): + if writer.is_enabled() and is_last_rank(): data_type = 'test' if test else 'validation' writer.add_scalar(f'lm-loss-validation/{key} {data_type}', total_loss_dict[key].item(), iteration) writer.add_scalar(f'lm-loss-validation/{key} {data_type} vs samples', total_loss_dict[key].item(), - args.consumed_train_samples) + args.consumed_train_samples, + x_axis_samples) writer.add_scalar(f'lm-loss-validation/{key} {data_type} vs tokens', total_loss_dict[key].item(), - args.consumed_train_tokens) + args.consumed_train_tokens, + x_axis_tokens) if args.log_validation_ppl_to_tensorboard: writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl', ppl, iteration) writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl vs samples', - ppl, args.consumed_train_samples) + ppl, args.consumed_train_samples, x_axis_samples) writer.add_scalar(f'lm-loss-validation/{key} {data_type} ppl vs tokens', - ppl, args.consumed_train_tokens) + ppl, args.consumed_train_tokens, x_axis_tokens) - if process_non_loss_data_func is not None and writer and is_last_rank(): + if process_non_loss_data_func is not None and writer.is_enabled() and is_last_rank(): process_non_loss_data_func(collected_non_loss_data, iteration, writer) length = len(string) + 1 From fc989b8844b34489fa8c71aa048f2529ce5e397a Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Wed, 17 Jul 2024 20:21:02 +0300 Subject: [PATCH 278/590] add kill switch file support to gracefully exit training at runtime (#412) * Update arguments.py * Update training.py * Update utils.py * add copyrights * add copyrights * add copyrights * Update arguments.py help * Update arguments.py * Update training.py * Update utils.py * Update arguments.py --- megatron/arguments.py | 3 +++ megatron/training.py | 19 +++++++++++++++++-- megatron/utils.py | 8 ++++++++ 3 files changed, 28 insertions(+), 2 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 30c3d669d70..af39f0b0e30 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -680,6 +680,9 @@ def _add_network_size_args(parser): help='Untie embeddings and output weights.'), group.add_argument('--embedding-weights-in-fp32', action='store_true', help='Cast word embedding weights to fp32 before embedding fwd.'), + group.add_argument('--kill-switch-file', type=str, default=None, + help='Location of kill switch file. ' + 'If found will automatically exit the program at runtime.') return parser diff --git a/megatron/training.py b/megatron/training.py index 697d62f7b8f..6ba26f39440 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -43,7 +43,7 @@ from megatron.optimizer_param_scheduler import OptimizerParamScheduler from megatron.model import DistributedDataParallel as LocalDDP from megatron.utils import check_adlr_autoresume_termination -from megatron.utils import unwrap_model +from megatron.utils import unwrap_model, found_kill_switch from megatron.data.data_samplers import build_pretraining_data_loader from megatron.utils import calc_params_l2_norm from megatron.core.pipeline_parallel import get_forward_backward_func @@ -128,6 +128,13 @@ def pretrain(train_valid_test_dataset_provider, # Initalize and get arguments, timers, and Tensorboard writer. initialize_megatron(extra_args_provider=extra_args_provider, args_defaults=args_defaults, external_args=external_args) + + args = get_args() + + if found_kill_switch(): + print_datetime(f"Detected kill switch at {args.kill_switch_file}. Exiting") + sys.exit() + # Set pytorch JIT layer fusion options and warmup JIT functions. if get_accelerator().device_name() == 'cuda': set_jit_fusion_options() @@ -144,7 +151,6 @@ def pretrain(train_valid_test_dataset_provider, time.time() - _TRAIN_START_TIME)) print_datetime('after megatron is initialized') - args = get_args() timers = get_timers() if args.deepspeed: @@ -1358,6 +1364,15 @@ def train(forward_step_func, model, optimizer, opt_param_scheduler, sys.exit() trigger(on_step_end) + # Exiting based on kill switch file + if found_kill_switch(): + if args.save and not saved_checkpoint: + save_checkpoint_and_time(iteration, model, optimizer, + opt_param_scheduler) + torch.distributed.barrier() + print_datetime(f"Detected kill switch at {args.kill_switch_file}, " + f"iteration={iteration}. Exiting") + sys.exit() return iteration diff --git a/megatron/utils.py b/megatron/utils.py index 9033d6402ae..cbb7aa64268 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -381,3 +381,11 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): p = model[0].module.tied_modules.embed.word_embeddings.weight._hp_param fh.write(f"{get_fingerprint(p)} module.tied_modules.embed.word_embeddings.weight._hp_param {p.shape}\n") + +def found_kill_switch(): + args = get_args() + if args.kill_switch_file is not None and os.path.exists(args.kill_switch_file): + return True + else: + return False + From 7d23e33633e27cd7798fc798c7c268c7fa286120 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Thu, 18 Jul 2024 19:47:04 +0300 Subject: [PATCH 279/590] add support to run custom Hf tokenizer for training and dataset pre-processing (#421) * Update arguments.py * Update tokenizer.py * Update preprocess_data.py --- megatron/arguments.py | 2 ++ megatron/tokenizer/tokenizer.py | 12 +++++++++--- tools/preprocess_data.py | 8 +++++++- 3 files changed, 18 insertions(+), 4 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index af39f0b0e30..49b3d8e4c66 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1299,6 +1299,8 @@ def _add_data_args(parser): help='What type of tokenizer to use.') group.add_argument('--tokenizer-model', type=str, default=None, help='Sentencepiece tokenizer model.') + group.add_argument('--trust-remote-code', action='store_true', default=False, + help='To run HFTokenizer model from local path.') group.add_argument('--data-impl', type=str, default='infer', choices=['mmap', 'infer'], help='Implementation of indexed datasets.') diff --git a/megatron/tokenizer/tokenizer.py b/megatron/tokenizer/tokenizer.py index 43c251bab10..7edef90b6fe 100644 --- a/megatron/tokenizer/tokenizer.py +++ b/megatron/tokenizer/tokenizer.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """Megatron tokenizers.""" @@ -40,7 +41,9 @@ def build_tokenizer(args): tokenizer = _NullTokenizer(args.vocab_size) elif args.tokenizer_type == 'HFTokenizer': assert args.tokenizer_model is not None - tokenizer = _HFTokenizer(args.tokenizer_model,args.seq_length) + tokenizer = _HFTokenizer(args.tokenizer_model, + args.seq_length, + args.trust_remote_code) else: raise NotImplementedError('{} tokenizer is not ' 'implemented.'.format(args.tokenizer_type)) @@ -540,10 +543,13 @@ def additional_special_tokens_ids(self): class _HFTokenizer(AbstractTokenizer): """HF Tokenizer""" - def __init__(self, tokenizer_name_or_path,max_seq_len): + def __init__(self, tokenizer_name_or_path, max_seq_len, trust_remote_code): name = tokenizer_name_or_path super().__init__(name) - self.tokenizer = AutoTokenizer.from_pretrained(tokenizer_name_or_path,padding_side="right",use_fast=False) + self.tokenizer = AutoTokenizer.from_pretrained(tokenizer_name_or_path, + padding_side="right", + trust_remote_code=trust_remote_code, + use_fast=False) DEFAULT_PAD_TOKEN = "[PAD]" DEFAULT_EOS_TOKEN = "" diff --git a/tools/preprocess_data.py b/tools/preprocess_data.py index 399f93c10e7..c34d9dcb3bb 100644 --- a/tools/preprocess_data.py +++ b/tools/preprocess_data.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """Processing large data for pretraining.""" @@ -193,10 +194,15 @@ def get_args(): group.add_argument('--tokenizer-type', type=str, required=True, choices=['BertWordPieceLowerCase','BertWordPieceCase', 'GPT2BPETokenizer', 'SentencePieceTokenizer', - 'GPTSentencePieceTokenizer', 'NullTokenizer'], + 'GPTSentencePieceTokenizer', 'HFTokenizer', + 'NullTokenizer'], help='What type of tokenizer to use.') group.add_argument('--tokenizer-model', type=str, default=None, help='YTTM tokenizer model.') + group.add_argument('--seq-length', type=int, default=None, + help='Maximum sequence length to process.') + group.add_argument('--trust-remote-code', action='store_true', + help='To run HFTokenizer model from local path.') group.add_argument('--vocab-file', type=str, default=None, help='Path to the vocab file') group.add_argument('--vocab-size', default=786, From 13f2673caab46152ee7ba8051b8b0a9602270775 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Fri, 19 Jul 2024 04:02:32 +0300 Subject: [PATCH 280/590] improve repeat_kv GQA perf (#419) --- megatron/model/transformer.py | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index be8ae6ef4bf..e79abea3cf2 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -654,11 +654,16 @@ def repeat_kv(self, hidden_states, n_rep): slen, batch, num_key_value_heads_per_partition, head_dim = hidden_states.shape if n_rep == 1: return hidden_states - hidden_states = hidden_states[:, :, :, None, :].expand( - slen, batch, num_key_value_heads_per_partition, n_rep, head_dim) - return hidden_states.reshape(slen, batch, - num_key_value_heads_per_partition * n_rep, - head_dim) + elif num_key_value_heads_per_partition == 1: + # If no of KV heads is 1 then just perform expand operation + # instead of unsqueeze, expand and reshape to match query states. + return hidden_states.expand(slen, batch, n_rep, head_dim) + else: + hidden_states = hidden_states[:, :, :, None, :].expand( + slen, batch, num_key_value_heads_per_partition, n_rep, head_dim) + return hidden_states.reshape(slen, batch, + num_key_value_heads_per_partition * n_rep, + head_dim) def split_tensor(self, mixed_x_layer): query_layer, key_layer, value_layer = torch.split(mixed_x_layer, [self.num_key_value_groups, 1, 1], dim=-2) From 3af2e255ed9cc1b6b596cd232e12a84f8cbdeea5 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Fri, 19 Jul 2024 04:03:30 +0300 Subject: [PATCH 281/590] acquire device when required (#420) * Update module.py * Update preprocess_data.py * add copyrights * add copyrights * Update tokenizer.py * add copyrights --- megatron/model/module.py | 16 ++++++++++++---- tools/preprocess_data.py | 2 +- 2 files changed, 13 insertions(+), 5 deletions(-) diff --git a/megatron/model/module.py b/megatron/model/module.py index 963ad2d29d4..08cf7cb5537 100644 --- a/megatron/model/module.py +++ b/megatron/model/module.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """Megatron Module""" @@ -10,10 +11,9 @@ from megatron.core import mpu, tensor_parallel -_FLOAT_TYPES = [get_accelerator().FloatTensor(0).dtype] -_HALF_TYPES = [get_accelerator().HalfTensor(0).dtype] -_BF16_TYPES = [get_accelerator().BFloat16Tensor(0).dtype] - +_FLOAT_TYPES = None +_HALF_TYPES = None +_BF16_TYPES = None def param_is_not_shared(param): @@ -131,6 +131,9 @@ def conversion_helper(val, conversion): def fp32_to_float16(val, float16_convertor): """Convert fp32 `val` to fp16/bf16""" + global _FLOAT_TYPES + if _FLOAT_TYPES is None: + _FLOAT_TYPES = [get_accelerator().FloatTensor(0).dtype] def half_conversion(val): val_typecheck = val if isinstance(val_typecheck, (Parameter, Variable)): @@ -143,6 +146,11 @@ def half_conversion(val): def float16_to_fp32(val): """Convert fp16/bf16 `val` to fp32""" + global _HALF_TYPES, _BF16_TYPES + if _HALF_TYPES is None: + _HALF_TYPES = [get_accelerator().HalfTensor(0).dtype] + if _BF16_TYPES is None: + _BF16_TYPES = [get_accelerator().BFloat16Tensor(0).dtype] def float_conversion(val): val_typecheck = val if isinstance(val_typecheck, (Parameter, Variable)): diff --git a/tools/preprocess_data.py b/tools/preprocess_data.py index c34d9dcb3bb..6e117db31a5 100644 --- a/tools/preprocess_data.py +++ b/tools/preprocess_data.py @@ -235,7 +235,7 @@ def get_args(): print("Are you sure you don't want to split sentences?") # some default/dummy values for the tokenizer - args.rank = 1 + args.rank = 0 args.make_vocab_size_divisible_by = 128 args.tensor_model_parallel_size = 1 args.vocab_extra_ids = 0 From b38bcb651570c7f62cc30c7cbf2ffa315d992a64 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Jul 2024 08:56:12 -0500 Subject: [PATCH 282/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 108 +++++++++++++++++++++++++++++++----------------- 1 file changed, 71 insertions(+), 37 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 789ce933f41..2b8eafb4620 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1,5 +1,4 @@ #!/bin/bash --login - ############################################################################### # [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh) # @@ -21,6 +20,27 @@ # command for launching across all GPUs in our active PBS job. ############################################################################### + +############################################################################### +# Check if running in DEBUG=1 mode. +# - If so, this will print each command before it is ran and exit if any of +# them return a nonzero exit status. +############################################################################### +if [[ -n "${DEBUG-}" ]]; then # to use: `DEBUG=1 bash train_llama_alcf.sh` + printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" + set -euxo pipefail +fi + +############################################################################### +# Print (but DO NOT EXECUTE !!) each command that would be ran. +# +# Enable with: NOOP=1 PBS_O_WORKDIR=$(pwd) bash train_llama_alcf.sh +############################################################################### +if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` + echo "Run NOOP mode" + set -o noexec # same as set -n +fi + ################## # helpers_main # @@ -413,6 +433,39 @@ get_batch_size_on_polaris() { echo "${mbs}" } +_get_num_hosts_from_hostfile() { + if [[ "$#" == 1 ]]; then + if [[ -f "$1" ]]; then + nhosts=$(wc -l < "$1") + echo "${nhosts}" + else + exit 1 + fi + else + exit 1 + fi +} + + +########################################### +# get_grad_acc_steps_on_aurora +# +# NOTE: +# We use different numbers of gradient +# accumulation steps (GAS) depending +# on the number of hosts in our job. +# +# Each host has: +# +# [2 tiles] x [6 xpus / tile] = 12 xpus +# +# | nnhosts | nhosts | GAS | +# |:-------------:|:---------:|:-----:| +# | 64 <= n < inf | [64, inf) | 1 | +# | 32 <= n < 64 | [32, 64) | 2 | +# | 16 <= n < 32 | [16, 32) | 4 | +# | 0 <= n < 16 | [0, 16) | 8 | +########################################### get_grad_acc_steps_on_aurora() { if [[ "$#" == 0 ]]; then local hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" @@ -422,39 +475,16 @@ get_grad_acc_steps_on_aurora() { echo "Expected exactly 0 or 1 arguments, received: $#" exit 1 fi - nhosts=$(wc -l < "${hf}") - # NOTE: - # | nhosts | nhosts | GAS | - # |:----------------:|:----------:|:-----:| - # | n <= 32 | [0, 32] | 8 | - # | 32 < n < 64 | (32, 64) | 4 | - # | 64 <= n < 128 | [64, 128) | 2 | - # | 128 <= n < inf | [128, inf) | 1 | - if [[ 128 -le "${nhosts}" ]]; then + local nhosts=$(wc -l < "${hf}") + if [[ 64 -le "${nhosts}" ]]; then gas=1 - elif [[ 64 -le "${nhosts}" && "${nhosts}" -lt 128 ]]; then - gas=2 - # 32 <= nhosts < 64 elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 64 ]]; then + gas=2 + elif [[ 16 -le "${nhosts}" && "${nhosts}" -lt 32 ]]; then gas=4 - # nhosts <= 16 - # elif [[ "${nhosts}" -lt 32 ]]; then - # gas=8 else gas=8 fi - # if [[ "${nhosts}" > 512 ]]; then - # gas=1 - # # elif [[ 256 <= "${nhosts}" && "${nhosts}" <= 512 ]]; then - # elif [[ 256 -le "${nhosts}" && "${nhosts}" -le 512 ]]; then - # elif [[ 128 < "${nhosts}" && "${nhosts}" <= 256 "${nhosts}" ]]; then - # gas=2 - # # elif [[ 128 < "${nhosts}" && "${nhosts}" < 256 ]]; then - # elif [[ 64 <= "${nhosts}" && "${nhosts}" <= 128 -lt "${nhosts}" ]]; then - # gas=2 - # else - # gas=8 - # fi echo "${gas}" } @@ -475,7 +505,9 @@ setParams() { LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- - if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then + # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then + local mn=$(get_machine_name) + if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" export CCL=${CCL:-ccl} # CCL @@ -483,6 +515,7 @@ setParams() { export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora)}" + echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 export CCL_PROCESS_LAUNCHER=pmix export CCL_ATL_TRANSPORT=mpi @@ -497,7 +530,6 @@ setParams() { if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else - # LLAMA_ARGS="${LLAMA_ARGS} --use-flash-attn-builder" FLASH_ARG="--use-flash-attn-builder" fi else @@ -506,7 +538,8 @@ setParams() { fi ###################################################################### # +--------[Polaris]-----------------------------------+ - elif [[ $(hostname) == x3* ]]; then + # elif [[ $(hostname) == x3* ]]; then + elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" TP=${TP:-1} # TP = 2 export NCCL=${NCCL:-nccl} # NCCL @@ -525,7 +558,8 @@ setParams() { echo "Setting up AWS NCCL OFI Plugin on Polaris..." source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # +--------[Perlmutter]---------------------------------+ - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + # elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then +elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then TP="${TP:-2}" export NCCL="${NCCL:-nccl}" export BE="${NCCL}" @@ -671,17 +705,17 @@ make_ds_hostfile() { # 1. Clone [`saforem2/ezpz`](https://github.com/saforem2/ezpz) (if necessary) # to `"${WORKING_DIR}/deps/ezpz/"` # -# 2. Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) +# 2. Source [`ezpz/src/ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) # - This provides `{ezpz_setup_python, ezpz_setup_alcf}` (called below) # -# 3. Call `ezpz_setup_python`: +# 3. Call `ezpz_setup_python` (from `ezpz/bin/utils.sh`): # - This will setup conda + virtual enviroment # -# 4. Call `ezpz_setup_alcf`: +# 4. Call `ezpz_setup_alcf` (from `ezpz/bin/utils.sh`): # - This will parse `$PBS_*` variables and build launch cmd # -# 3. Call `_ezpz_install`: -# - Install ezpz from "${WORKING_DIR}/depz/ezpz/" +# 3. Call `_ezpz_install` (from `Megatron-DeepSpeed/ALCF/helpers.sh`): +# - Install ezpz from `"${WORKING_DIR}/depz/ezpz/"` ########################################### ezpz_setup() { # setup_alcf "$@" From 0999de294871cb30662bdade4230afa3eaed2590 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Jul 2024 09:01:53 -0500 Subject: [PATCH 283/590] Update `ALCF/requirements/requirements.txt` --- ALCF/requirements/requirements.txt | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/ALCF/requirements/requirements.txt b/ALCF/requirements/requirements.txt index 7e3a2ae4a53..e0a969358d8 100644 --- a/ALCF/requirements/requirements.txt +++ b/ALCF/requirements/requirements.txt @@ -1,9 +1,17 @@ +hjson +ninja +psutil +py-cpuinfo +pydantic +tqdm transformers bitsandbytes sentencepiece einops xgboost -numpy<2.0 fixedint pybind11 six +numpy<2 +schedulefree +packaging>=20.0 From 6ad3a99c87132f807cb0e78f0e6347a23b110637 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 19 Jul 2024 13:29:45 -0500 Subject: [PATCH 284/590] Fix opt hyperparams in `ALCF/helpers.sh` --- ALCF/helpers.sh | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 2b8eafb4620..32531ec27a9 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -220,6 +220,10 @@ setup_run_cmd() { --${DTYPE} \ ${DEFAULTS} \ --optimizer ${OPT} \ + --adam-beta1=${ADAM_BETA1} \ + --adam-beta2=${ADAM_BETA2} \ + --adam-eps=${ADAM_EPS} \ + --weight-decay=${WEIGHT_DECAY} \ --save ${CKPT_DIR} \ --load ${CKPT_DIR} \ --seq-length ${SEQ} \ @@ -578,6 +582,10 @@ elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" + export ADAM_BETA1="${ADAM_BETA1:-0.9}" + export ADAM_BETA2="${ADAM_BETA2:-0.95}" + export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l < "${HOSTFILE}") if [[ -z "${NGPU_PER_HOST:-}" ]]; then From 08b93769c9a6df70ceb2c74ff3eaaebee25fe074 Mon Sep 17 00:00:00 2001 From: Logan Adams <114770087+loadams@users.noreply.github.com> Date: Fri, 19 Jul 2024 15:04:39 -0700 Subject: [PATCH 285/590] Add basic compilation test (#426) --- .github/workflows/python.yml | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) create mode 100644 .github/workflows/python.yml diff --git a/.github/workflows/python.yml b/.github/workflows/python.yml new file mode 100644 index 00000000000..d519b65e2e7 --- /dev/null +++ b/.github/workflows/python.yml @@ -0,0 +1,35 @@ +name: python + +on: + workflow_dispatch: + pull_request: + branches: + '**' + schedule: + - cron: "0 0 * * *" + +concurrency: + group: ${{ github.workflow }}-${{ github.ref }} + cancel-in-progress: true + +jobs: + unit-tests: + strategy: + matrix: + pyVersion: ["3.7", "3.8", "3.9", "3.10"] + fail-fast: false + + runs-on: ubuntu-22.04 + container: + image: deepspeed/gh-builder:py${{ matrix.pyVersion }} + + steps: + - uses: actions/checkout@v4 + + - name: environment + run: | + which python + python --version + - name: Install Megatron-DeepSpeed + run: | + pip3 install . From 3afd267e1e50b1410beb606c5625cc232a55417a Mon Sep 17 00:00:00 2001 From: Logan Adams <114770087+loadams@users.noreply.github.com> Date: Fri, 19 Jul 2024 15:09:44 -0700 Subject: [PATCH 286/590] Update yml to be valid (#427) --- .github/workflows/python.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/python.yml b/.github/workflows/python.yml index d519b65e2e7..2f9765af241 100644 --- a/.github/workflows/python.yml +++ b/.github/workflows/python.yml @@ -32,4 +32,4 @@ jobs: python --version - name: Install Megatron-DeepSpeed run: | - pip3 install . + pip3 install . From 019dc3c6da91d43f3a53c0451e29e2fa7dd3e447 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 20 Jul 2024 00:16:05 -0500 Subject: [PATCH 287/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 66 +++++++++++-------------------------------------- 1 file changed, 15 insertions(+), 51 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 32531ec27a9..cccb2d81903 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -21,26 +21,6 @@ ############################################################################### -############################################################################### -# Check if running in DEBUG=1 mode. -# - If so, this will print each command before it is ran and exit if any of -# them return a nonzero exit status. -############################################################################### -if [[ -n "${DEBUG-}" ]]; then # to use: `DEBUG=1 bash train_llama_alcf.sh` - printf "\e[1;31m%s\e[0m\n" "!! RUNNING IN DEBUG MODE !!" - set -euxo pipefail -fi - -############################################################################### -# Print (but DO NOT EXECUTE !!) each command that would be ran. -# -# Enable with: NOOP=1 PBS_O_WORKDIR=$(pwd) bash train_llama_alcf.sh -############################################################################### -if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` - echo "Run NOOP mode" - set -o noexec # same as set -n -fi - ################## # helpers_main # @@ -57,7 +37,7 @@ fi # 2. else, if `${SLURM_SUBMIT_DIR}` is nonzero use this # 3. else, use `$(pwd)` # -# this is crucial since many of the functions below use paths +# this is _crucial_ since many of the functions below use paths # which are defined relative to this "${WORKING_DIR}" # (e.g. virtual environment, location of executables, etc.) ################## @@ -137,7 +117,7 @@ setup() { ########################################################################## install_dependencies # Set command line arguments to pass to `"${EXEC}"` - setParams || exit + setParams "$@" || exit # Create `deepspeed_config.json` from runtime params from ^ buildDSconfig || exit # Specify output directory for {logs, checkpoints, etc.} @@ -320,7 +300,6 @@ get_machine() { check_and_kill_if_running() { - # kill $(ps aux | grep -E "$USER.+(mpi|main.py)" | grep -v grep | awk '{print $2}') RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') if [[ -n "${RUNNING_PIDS}" ]]; then echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}"; @@ -381,7 +360,7 @@ setupLauncher() { export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else if [[ -n "${DIST_LAUNCH}" ]]; then - local mn=$(get_machine_name) + mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" else @@ -450,7 +429,6 @@ _get_num_hosts_from_hostfile() { fi } - ########################################### # get_grad_acc_steps_on_aurora # @@ -472,14 +450,14 @@ _get_num_hosts_from_hostfile() { ########################################### get_grad_acc_steps_on_aurora() { if [[ "$#" == 0 ]]; then - local hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" + hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" elif [[ "$#" == 1 ]]; then - local hf="$1" + hf="$1" else echo "Expected exactly 0 or 1 arguments, received: $#" exit 1 fi - local nhosts=$(wc -l < "${hf}") + nhosts=$(wc -l < "${hf}") if [[ 64 -le "${nhosts}" ]]; then gas=1 elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 64 ]]; then @@ -510,7 +488,7 @@ setParams() { # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then - local mn=$(get_machine_name) + mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" @@ -518,7 +496,9 @@ setParams() { export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS - export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora)}" + gas=$(get_grad_acc_steps_on_aurora "$@") + export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" + # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 export CCL_PROCESS_LAUNCHER=pmix @@ -734,9 +714,10 @@ ezpz_setup() { if [[ -d "${ezdir}" ]]; then echo "Found ezpz in ${ezdir}" else - mkdir -p $(dirname "${ezdir}") + mkdir -p "$(dirname "${ezdir}")" git clone https://github.com/saforem2/ezpz "${ezdir}" fi + # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh source "${ezdir}/src/ezpz/bin/utils.sh" || exit ezpz_setup_python ezpz_setup_alcf "$@" @@ -876,7 +857,7 @@ install_dependencies() { # install_deepspeed_for_xpu || exit # fi printf "[install_dependencies] No 'deepspeed' command found on %s" "${mn}" - printf "[install_dependencies] !! No deepsepeed in $(which python3)" + printf "[install_dependencies] !! No deepsepeed in %s" "$(which python3)" fi } @@ -1032,25 +1013,8 @@ setup_tokenizer_and_data() { ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# - if [[ $(hostname) == x4* ]]; then # -----------------------------[AURORA] - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/aurora/dolma.txt" - elif [[ $(hostname) == x1* ]]; then # ----------------------------[SUNSPOT] - # shellcheck source=./data-lists/sunspot/books.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sunspot/dolma.txt" - elif [[ $(hostname) == x3* ]]; then # -------------------[POLARIS / SIRIUS] - polaris_match=$(echo $PBS_NODEFILE | grep 'polaris') - if [[ -n "${polaris_match}" ]]; then - # shellcheck source=./data-lists/polaris/dolma.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/polaris/dolma.txt" - else - # shellcheck source=./data-lists/sirius/dolma.txt - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/sirius/dolma.txt" - fi - elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then # [PERLMUTTER] - dfl_fallback="${SLURM_SUBMIT_DIR}/genslm-subsample.txt" - else # -----------------------------------------------------------[UNKNOWN] - echo "Unknown hostname. Must manually specify DATA_FILE_LIST." - fi + mn=$(get_machine_name) + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" ############################################################################ # set `dfl` to `dfl_fallback` if not passed as an argument, # use this data file list to call `setData` From 54bd60875d1f4ef14c5a86e893882edd1070eb9e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 20 Jul 2024 00:16:56 -0500 Subject: [PATCH 288/590] Track grad_norm in `megatron/training.py` --- megatron/training.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training.py b/megatron/training.py index a396ed2b4aa..9289c64ec80 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -941,7 +941,7 @@ def train_step( # Update learning rate. if args.deepspeed: skipped_iter = 0 - grad_norm = None + grad_norm = model[0].get_global_grad_norm() num_zeros_in_grad = None loss_reduced = {} for key in losses_reduced[0]: From 969f4c5a5e031c05898f3f988f664801a697087e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 20 Jul 2024 00:17:14 -0500 Subject: [PATCH 289/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 49 +++++++++++++++++++++++------------------------- 1 file changed, 23 insertions(+), 26 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 9dc0f1d9469..b68a266c592 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -1,30 +1,27 @@ #!/bin/bash --login +##################################### +# AuroraGPT-7B +# +# Main production script for training +# AuroraGPT-7B @ ALCF +##################################### -NOW="$(date "+%Y-%m-%d-%H%M%S")" +# 1. Navigate into `$PBS_O_WORKDIR` cd "${PBS_O_WORKDIR}" || exit - -HOSTNAME=$(hostname) -if [[ "${HOSTNAME}" == x3* ]]; then - MACHINE="polaris" -elif [[ "${HOSTNAME}" == x1* ]]; then - MACHINE="sunspot" -elif [[ "${HOSTNAME}" == x4* ]]; then - MACHINE="aurora" -fi - -OUTDIR="${PBS_O_WORKDIR}/pbslogs" && mkdir -p "${OUTDIR}" -OUTFILE="${OUTDIR}/${PBS_JOBID}-${NOW}.log" - -echo "+---------------------------------------------------------+" -echo "| Running on: ${MACHINE}" -echo "| Detected ${nhosts} hosts. Running with micro batch: ${MBS}" -echo "| Logging job output to: ${OUTFILE}" -echo "+---------------------------------------------------------+" - -export TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=6000 -echo "${OUTFILE}" >> "${OUTDIR}/latest" -# export DEBUG=1 -# export MICRO_BATCH="${MICRO_BATCH:-${MBS}}" -export DATA_FILE_LIST="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/${MACHINE}/dolma_v1_7_file_list.txt}" -bash "${PBS_O_WORKDIR}/train_llama_alcf.sh" |& tee "${OUTFILE}" +HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE +# 2. source `ALCF/helpers.sh` +source "${HERE}/ALCF/helpers.sh" || exit +# 3. call `setup` from `./ALCF/helpers.sh` +setup "$@" || exit +# # 4. Take custom args +# export custom_args=" $@" +# 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} +export run_cmd="${run_cmd}" +# 6. Add "${run_cmd}" to output log +echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +# 7. Tell user where to find output +printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" +# 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} +eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" +set +x From 95506566328f1933e98ab1bdff80a4a30976b176 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 22 Jul 2024 08:12:09 -0500 Subject: [PATCH 290/590] Update `train_llama_alcf.sh` --- train_llama_alcf.sh | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/train_llama_alcf.sh b/train_llama_alcf.sh index 1850f2e92f2..259552ffbcf 100644 --- a/train_llama_alcf.sh +++ b/train_llama_alcf.sh @@ -20,6 +20,8 @@ if [[ -v NOOP ]]; then # to use: `NOOP=1 bash train_llama_alcf.sh` set -o noexec # same as set -n fi +XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" + ##################### # MAIN PROGRAM LOGIC ##################### @@ -30,7 +32,7 @@ main() { # 2. source `ALCF/helpers.sh` source "${HERE}/ALCF/helpers.sh" || exit # 3. call `setup` from `./ALCF/helpers.sh` - setup || exit + setup "$@" || exit # 4. Take custom args export custom_args=" $@" # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} @@ -40,7 +42,7 @@ main() { # 7. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} - eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" + eval "${run_cmd}" |& grep -E -v "${XPU_IGNORE_STRING}" |& tee -a "${OUTPUT_LOG}" set +x } From 5d96d64f96be27b62cf4a10d8440aeed9694b7ab Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 22 Jul 2024 08:33:52 -0500 Subject: [PATCH 291/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index b68a266c592..8bdf776ed9c 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -7,6 +7,7 @@ # AuroraGPT-7B @ ALCF ##################################### + # 1. Navigate into `$PBS_O_WORKDIR` cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE @@ -14,14 +15,10 @@ HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE source "${HERE}/ALCF/helpers.sh" || exit # 3. call `setup` from `./ALCF/helpers.sh` setup "$@" || exit -# # 4. Take custom args -# export custom_args=" $@" -# 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} export run_cmd="${run_cmd}" -# 6. Add "${run_cmd}" to output log echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" # 7. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" +XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} -eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" -set +x +eval "${run_cmd}" |& grep -E -v "${XPU_IGNORE_STRING}" |& tee -a "${OUTPUT_LOG}" From 8822a5ced6ce74d926fbe0f49cdca6bb3389bef8 Mon Sep 17 00:00:00 2001 From: Lev Kurilenko <113481193+lekurile@users.noreply.github.com> Date: Mon, 29 Jul 2024 14:40:40 -0700 Subject: [PATCH 292/590] Update/add GPT/Llama universal checkpointing scripts (#391) This PR adds a Llama universal checkpointing example to examples_deepspeed/universal_checkpointing. It also includes changes to the README, some minor changes, and an update to the TensorBoard analysis script. --- .../universal_checkpointing/README.md | 22 +-- .../llama/run_llama_bf16.sh | 175 +++++++++++++++++ .../llama/run_tb_analysis_llama.sh | 27 +++ .../llama/run_universal_llama_bf16.sh | 176 ++++++++++++++++++ .../{ => megatron_gpt}/run_bf16.sh | 2 +- .../{ => megatron_gpt}/run_fp16.sh | 2 +- .../run_tb_analysis_gpt.sh} | 2 - .../run_tb_analysis_gpt_plot_only.sh | 31 +++ .../{ => megatron_gpt}/run_universal_bf16.sh | 2 +- .../{ => megatron_gpt}/run_universal_fp16.sh | 2 +- .../tb_analysis/arguments.py | 2 + .../tb_analysis/tb_analysis_script.py | 41 +++- .../tb_analysis/uc_analysis.py | 2 +- .../tb_analysis/utils.py | 26 ++- 14 files changed, 486 insertions(+), 26 deletions(-) create mode 100644 examples_deepspeed/universal_checkpointing/llama/run_llama_bf16.sh create mode 100755 examples_deepspeed/universal_checkpointing/llama/run_tb_analysis_llama.sh create mode 100644 examples_deepspeed/universal_checkpointing/llama/run_universal_llama_bf16.sh rename examples_deepspeed/universal_checkpointing/{ => megatron_gpt}/run_bf16.sh (99%) rename examples_deepspeed/universal_checkpointing/{ => megatron_gpt}/run_fp16.sh (99%) rename examples_deepspeed/universal_checkpointing/{run_tb_analysis.sh => megatron_gpt/run_tb_analysis_gpt.sh} (96%) create mode 100755 examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt_plot_only.sh rename examples_deepspeed/universal_checkpointing/{ => megatron_gpt}/run_universal_bf16.sh (99%) rename examples_deepspeed/universal_checkpointing/{ => megatron_gpt}/run_universal_fp16.sh (99%) diff --git a/examples_deepspeed/universal_checkpointing/README.md b/examples_deepspeed/universal_checkpointing/README.md index a0a1df5b8a1..281d320e996 100644 --- a/examples_deepspeed/universal_checkpointing/README.md +++ b/examples_deepspeed/universal_checkpointing/README.md @@ -10,12 +10,12 @@ This folder contains example scripts that demonstrate how to use Universal Check For ZeRO stage 1, we provide bash scripts for bf16 and fp16 training examples corresponding to the steps 1 and 3 above. The step 1 scripts launch a training run of TP=PP=DP=2 of 200 iterations that creates a checkpoint every 100 iterations. The step 3 scripts load a universal checkpoint of iteration 100 and resume training with TP=PP=2 and DP=1 for an additional 100 iterations. Users can modify these scripts to try out other save and resume 3D combinations (e.g., save TP=PP=DP=1 and resume TP=PP=DP=2). Tensorboard logs are created by both step 1 and 3 scripts to enable visual inspection of how well the loss curves of the initial and resumed training runs match, especially at iteration 101. 1. bf16: - * run_bf16.sh: step 1 - * run_universal_bf16.sh: step 3 + * megatron_gpt/run_bf16.sh: step 1 + * megatron_gpt/run_universal_bf16.sh: step 3 2. fp16: - * run_fp16.sh: step 1 - * run_universal_fp16.sh: step 3 + * megatron_gpt/run_fp16.sh: step 1 + * megatron_gpt/run_universal_fp16.sh: step 3 Please note that these scripts should be run from the root folder of the repo (i.e., two levels above this README). For illustration, here are the commands for running the bf16 example. @@ -41,22 +41,22 @@ NOTE: Make sure to update your `BASE_DATA_PATH` path in the `run_[bf16/fp16].sh` ### Step 1: Create ZeRO checkpoint ```bash - bash examples_deepspeed/universal_checkpointing/run_bf16.sh + bash examples_deepspeed/universal_checkpointing/megatron_gpt/run_bf16.sh ``` -By default the script will create the checkpoints in folder `z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_toy` +By default the script will create the checkpoints in folder `z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_sp1_toy` ### Step 2: Convert ZeRO checkpoint of iteration 100 to Universal format Assuming the DeepSpeed source code is cloned into the home folder, the following command will generate universal checkpoint for iteration 100. ```bash python ${HOME}/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py \ - --input_folder z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_toy/global_step100 \ - --output_folder z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_toy/global_step100_universal + --input_folder z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_sp1_toy/global_step100 \ + --output_folder z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_sp1_toy/global_step100_universal ``` Note that we chose to create the universal checkpoint in the same checkpoint folder as the ZeRO checkpoint. This maintains the normal checkpoint folder structure expected by the Megatron-DeepSpeed code, which makes it easy to load universal checkpoints with little/no script or code changes. For clarity, we show below the contents of the checkpoint folder after creation of the universal checkpoint. Note that the conversion script creates `global_step100_universal` folder and `latest_universal` file. ```bash -ls -l z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_toy/ +ls -l z1_uni_ckpt/checkpoints/gpt2/z1/bf16/tp2_pp2_dp2_sp1_toy/ total 48 drwxr-xr-x 2 user group 4096 Oct 21 08:51 global_step100 drwxr-xr-x 3 user group 4096 Oct 21 09:28 global_step100_universal @@ -69,7 +69,7 @@ drwxr-xr-x 2 user group 4096 Oct 21 09:01 global_step200 ### Step 3: Resume training with Universal checkpoint of iteration 100 ```bash -bash examples_deepspeed/universal_checkpointing/run_universal_bf16.sh +bash examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_bf16.sh ``` This resumption script effects the loading of universal checkpoint rather than the ZeRO checkpoint in the folder by passing `--universal-checkpoint` command line flag to the main training script (i.e., `pretrain_gpt.py`). @@ -85,7 +85,7 @@ The Universal Checkpointing example includes a TensorBoard analysis script that After Step 3 is completed, the script may be executed as follows: ```bash -bash examples_deepspeed/universal_checkpointing/run_tb_analysis.sh z1_uni_ckpt +bash examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt.sh z1_uni_ckpt ``` The script will output the following `csv` files: diff --git a/examples_deepspeed/universal_checkpointing/llama/run_llama_bf16.sh b/examples_deepspeed/universal_checkpointing/llama/run_llama_bf16.sh new file mode 100644 index 00000000000..72e79d4f1f9 --- /dev/null +++ b/examples_deepspeed/universal_checkpointing/llama/run_llama_bf16.sh @@ -0,0 +1,175 @@ +#!/bin/bash +set -ex + +DIR=`pwd` +###################################### +# Change the below configurations here +BASE_PATH=dataset +DS_CONFIG=${BASE_PATH}/deepspeed.json +DATASET=${BASE_PATH}/my-gpt2_text_document +TOKENIZER_PATH=${BASE_PATH}/llama-7b/tokenizer.model # offical llama tokenizer.model + +GPUS_PER_NODE=8 +MASTER_ADDR=localhost +MASTER_PORT=6000 +NNODES=1 +NODE_RANK=0 + +HIDDEN_SIZE=2048 # e.g. llama-13b: 5120 +FFN_HIDDEN_SIZE=5504 # e.g. llama-13b: 13824 +NUM_LAYERS=24 # e.g. llama-13b: 40 +NUM_HEADS=16 # e.g. llama-13b: 40 +SEQ=2048 + +LR_WARMUP_STEPS=2000 +WEIGHT_DECAY=0.1 +GRAD_CLIP=1 + +## Activation checkpointing saves GPU memory, but reduces training speed +# activation_checkpoint="true" +activation_checkpoint="false" + +ZERO_STAGE=1 +DTYPE="bf16" + +# 3D parallelism of training +TP=2 +PP=2 +DP=2 +SP=1 +WORLD_SIZE=$((TP*PP*DP*SP)) +GLOBAL_BATCH=32 +MICRO_BATCH=$((GLOBAL_BATCH/WORLD_SIZE)) +TRAIN_ITERS=250000 +LR=3e-4 +MIN_LR=3e-5 + +# Debug +DEBUG_MODE=1 +if [[ $DEBUG_MODE == 1 ]]; then + EXIT_INTERVAL=200 + SIZE_TAG="toy" +else + EXIT_INTERVAL=$TRAIN_ITERS + SIZE_TAG="big" +fi + +# 3D parallelism of checkpoint to load +LOAD_TP=$TP +LOAD_PP=$PP +LOAD_DP=$DP +LOAD_SP=$SP +RUN_TAG="save" + + +EXP_DIR="z${ZERO_STAGE}_uni_ckpt" +CHECKPOINT_PATH=${EXP_DIR}/checkpoints/llama/z${ZERO_STAGE}/$DTYPE/tp${TP}_pp${PP}_dp${DP}_sp${SP}_${SIZE_TAG} +LOAD_CHECKPOINT_PATH=${EXP_DIR}/checkpoints/llama/z${ZERO_STAGE}/$DTYPE/tp${LOAD_TP}_pp${LOAD_PP}_dp${LOAD_DP}_sp${LOAD_SP}_${SIZE_TAG} +LOG_DIR="${EXP_DIR}/tensorboard/llama/$DTYPE/tp${TP}_pp${PP}_dp${DP}_sp${SP}_hd${HIDDEN}_nl${LAYERS}_gbsz${GLOBAL_BATCH}_mbsz${MICRO_BATCH}_z${ZERO_STAGE}_LR_${LR}_${MIN_LR}_${DTYPE}_${SIZE_TAG}_${RUN_TAG}" +mkdir -p $LOG_DIR + +# Below configuration required for llama model as per llama paper +# --no-query-key-layer-scaling \ +# --attention-dropout 0 \ +# --hidden-dropout 0 \ +# --use-rotary-position-embeddings \ +# --untie-embeddings-and-output-weights \ +# --swiglu \ +# --normalization rmsnorm \ +# --disable-bias-linear \ +###################################### + +cat < $DS_CONFIG +{ + "train_batch_size" : $GLOBAL_BATCH, + "train_micro_batch_size_per_gpu": $MICRO_BATCH, + "steps_per_print": 1, + + "zero_optimization": { + "stage": $ZERO_STAGE + }, + + "bf16": { + "enabled": true + }, + + "wall_clock_breakdown" : false +} +EOT + +ds_args="" +ds_args=" --deepspeed ${ds_args}" +ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" +ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" + +if [ "${activation_checkpoint}" = "true" ]; then + ds_args="--deepspeed-activation-checkpointing ${ds_args}" + + ## old argument for recomputing the transformer layer + # ds_args="--checkpoint-activations ${ds_args}" + + ## new argument for recomputing the transformer layer + ds_args="--recompute-granularity full --recompute-method uniform ${ds_args}" + ## new argument for recomputing only the attention layer + # ds_args="--recompute-granularity selective ${ds_args}" +fi + +if [[ ${ZERO_STAGE} -gt 1 ]]; then +ds_args="${ds_args} \ + --no-pipeline-parallel" +fi + +options="\ + --tensor-model-parallel-size $TP \ + --pipeline-model-parallel-size $PP \ + --ds-sequence-parallel-size $SP \ + --num-layers $NUM_LAYERS \ + --hidden-size $HIDDEN_SIZE \ + --ffn-hidden-size $FFN_HIDDEN_SIZE \ + --num-attention-heads $NUM_HEADS \ + --micro-batch-size $MICRO_BATCH \ + --global-batch-size $GLOBAL_BATCH \ + --seq-length $SEQ \ + --max-position-embeddings $SEQ \ + --train-iters $TRAIN_ITERS \ + --save ${CHECKPOINT_PATH} \ + --load ${LOAD_CHECKPOINT_PATH} \ + --data-path $DATASET \ + --data-impl mmap \ + --tokenizer-type GPTSentencePieceTokenizer \ + --tokenizer-model $TOKENIZER_PATH \ + --split 949,50,1 \ + --distributed-backend nccl \ + --lr $LR \ + --lr-decay-style cosine \ + --min-lr $MIN_LR \ + --weight-decay $WEIGHT_DECAY \ + --clip-grad $GRAD_CLIP \ + --lr-warmup-iters $LR_WARMUP_STEPS \ + --optimizer adam \ + --adam-beta1 0.9 \ + --adam-beta2 0.95 \ + --log-interval 1 \ + --save-interval 100 \ + --eval-interval 10 \ + --eval-iters 40 \ + --exit-interval ${EXIT_INTERVAL} \ + --${DTYPE} \ + --no-query-key-layer-scaling \ + --attention-dropout 0 \ + --hidden-dropout 0 \ + --use-rotary-position-embeddings \ + --untie-embeddings-and-output-weights \ + --swiglu \ + --normalization rmsnorm \ + --disable-bias-linear \ + --tensorboard-dir $LOG_DIR \ + $ds_args +" + +WORKER_STR="--num_nodes 1 --num_gpus $WORLD_SIZE" +run_cmd="deepspeed --master_port 29700 $WORKER_STR ${DIR}/pretrain_gpt.py $@ ${options}" + +echo ${options} +echo ${run_cmd} +eval ${run_cmd} diff --git a/examples_deepspeed/universal_checkpointing/llama/run_tb_analysis_llama.sh b/examples_deepspeed/universal_checkpointing/llama/run_tb_analysis_llama.sh new file mode 100755 index 00000000000..b807fb97a75 --- /dev/null +++ b/examples_deepspeed/universal_checkpointing/llama/run_tb_analysis_llama.sh @@ -0,0 +1,27 @@ +#!/bin/bash +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +OUTPUT_PATH=$1 + +if [ "$OUTPUT_PATH" == "" ]; then + OUTPUT_PATH="z1_uni_ckpt" +fi + +# Training Loss +python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py \ + --tb_dir $OUTPUT_PATH \ + --tb_event_key "lm-loss-training/lm loss" \ + --plot_name "uc_char_training_loss.png" \ + --plot_title "Llama 7B Universal Checkpointing - Training Loss" \ + +# Validation Loss +python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py \ + --tb_dir $OUTPUT_PATH \ + --tb_event_key "lm-loss-validation/lm loss validation" \ + --csv_name "val_" \ + --plot_name "uc_char_validation_loss.png" \ + --plot_title "Llama 7B Universal Checkpointing - Validation Loss" \ + --plot_y_label "Validation LM Loss" \ diff --git a/examples_deepspeed/universal_checkpointing/llama/run_universal_llama_bf16.sh b/examples_deepspeed/universal_checkpointing/llama/run_universal_llama_bf16.sh new file mode 100644 index 00000000000..334fa3eaf6e --- /dev/null +++ b/examples_deepspeed/universal_checkpointing/llama/run_universal_llama_bf16.sh @@ -0,0 +1,176 @@ +#!/bin/bash +set -ex + +DIR=`pwd` +###################################### +# Change the below configurations here +BASE_PATH=dataset +DS_CONFIG=${BASE_PATH}/deepspeed.json +DATASET=${BASE_PATH}/my-gpt2_text_document +TOKENIZER_PATH=${BASE_PATH}/llama-7b/tokenizer.model # offical llama tokenizer.model + +GPUS_PER_NODE=8 +MASTER_ADDR=localhost +MASTER_PORT=6000 +NNODES=1 +NODE_RANK=0 + +HIDDEN_SIZE=2048 # e.g. llama-13b: 5120 +FFN_HIDDEN_SIZE=5504 # e.g. llama-13b: 13824 +NUM_LAYERS=24 # e.g. llama-13b: 40 +NUM_HEADS=16 # e.g. llama-13b: 40 +SEQ=2048 + +LR_WARMUP_STEPS=2000 +WEIGHT_DECAY=0.1 +GRAD_CLIP=1 + +## Activation checkpointing saves GPU memory, but reduces training speed +# activation_checkpoint="true" +activation_checkpoint="false" + +ZERO_STAGE=1 +DTYPE="bf16" + +# 3D parallelism of training +TP=2 +PP=2 +DP=1 +SP=1 +WORLD_SIZE=$((TP*PP*DP*SP)) +GLOBAL_BATCH=32 +MICRO_BATCH=$((GLOBAL_BATCH/WORLD_SIZE)) +TRAIN_ITERS=250000 +LR=3e-4 +MIN_LR=3e-5 + +# Debug +DEBUG_MODE=1 +if [[ $DEBUG_MODE == 1 ]]; then + EXIT_INTERVAL=200 + SIZE_TAG="toy" +else + EXIT_INTERVAL=$TRAIN_ITERS + SIZE_TAG="big" +fi + +# 3D parallelism of checkpoint to load +LOAD_TP=2 +LOAD_PP=2 +LOAD_DP=2 +LOAD_SP=1 +RUN_TAG="uni_load${LOAD_TP}_${LOAD_PP}_${LOAD_DP}_${LOAD_SP}" + + +EXP_DIR="z${ZERO_STAGE}_uni_ckpt" +CHECKPOINT_PATH=${EXP_DIR}/checkpoints/llama/z${ZERO_STAGE}/$DTYPE/tp${TP}_pp${PP}_dp${DP}_sp${SP}_${SIZE_TAG} +LOAD_CHECKPOINT_PATH=${EXP_DIR}/checkpoints/llama/z${ZERO_STAGE}/$DTYPE/tp${LOAD_TP}_pp${LOAD_PP}_dp${LOAD_DP}_sp${LOAD_SP}_${SIZE_TAG} +LOG_DIR="${EXP_DIR}/tensorboard/llama/$DTYPE/tp${TP}_pp${PP}_dp${DP}_sp${SP}_hd${HIDDEN}_nl${LAYERS}_gbsz${GLOBAL_BATCH}_mbsz${MICRO_BATCH}_z${ZERO_STAGE}_LR_${LR}_${MIN_LR}_${DTYPE}_${SIZE_TAG}_${RUN_TAG}" +mkdir -p $LOG_DIR + +# Below configuration required for llama model as per llama paper +# --no-query-key-layer-scaling \ +# --attention-dropout 0 \ +# --hidden-dropout 0 \ +# --use-rotary-position-embeddings \ +# --untie-embeddings-and-output-weights \ +# --swiglu \ +# --normalization rmsnorm \ +# --disable-bias-linear \ +###################################### + +cat < $DS_CONFIG +{ + "train_batch_size" : $GLOBAL_BATCH, + "train_micro_batch_size_per_gpu": $MICRO_BATCH, + "steps_per_print": 1, + + "zero_optimization": { + "stage": $ZERO_STAGE + }, + + "bf16": { + "enabled": true + }, + + "wall_clock_breakdown" : false +} +EOT + +ds_args="" +ds_args=" --deepspeed ${ds_args}" +ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" +ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" + +if [ "${activation_checkpoint}" = "true" ]; then + ds_args="--deepspeed-activation-checkpointing ${ds_args}" + + ## old argument for recomputing the transformer layer + # ds_args="--checkpoint-activations ${ds_args}" + + ## new argument for recomputing the transformer layer + ds_args="--recompute-granularity full --recompute-method uniform ${ds_args}" + ## new argument for recomputing only the attention layer + # ds_args="--recompute-granularity selective ${ds_args}" +fi + +if [[ ${ZERO_STAGE} -gt 1 ]]; then +ds_args="${ds_args} \ + --no-pipeline-parallel" +fi + +options="\ + --tensor-model-parallel-size $TP \ + --pipeline-model-parallel-size $PP \ + --ds-sequence-parallel-size $SP \ + --num-layers $NUM_LAYERS \ + --hidden-size $HIDDEN_SIZE \ + --ffn-hidden-size $FFN_HIDDEN_SIZE \ + --num-attention-heads $NUM_HEADS \ + --micro-batch-size $MICRO_BATCH \ + --global-batch-size $GLOBAL_BATCH \ + --seq-length $SEQ \ + --max-position-embeddings $SEQ \ + --train-iters $TRAIN_ITERS \ + --save ${CHECKPOINT_PATH} \ + --load ${LOAD_CHECKPOINT_PATH} \ + --data-path $DATASET \ + --data-impl mmap \ + --tokenizer-type GPTSentencePieceTokenizer \ + --tokenizer-model $TOKENIZER_PATH \ + --split 949,50,1 \ + --distributed-backend nccl \ + --lr $LR \ + --lr-decay-style cosine \ + --min-lr $MIN_LR \ + --weight-decay $WEIGHT_DECAY \ + --clip-grad $GRAD_CLIP \ + --lr-warmup-iters $LR_WARMUP_STEPS \ + --optimizer adam \ + --adam-beta1 0.9 \ + --adam-beta2 0.95 \ + --log-interval 1 \ + --save-interval 100 \ + --eval-interval 10 \ + --eval-iters 40 \ + --exit-interval ${EXIT_INTERVAL} \ + --${DTYPE} \ + --no-query-key-layer-scaling \ + --attention-dropout 0 \ + --hidden-dropout 0 \ + --use-rotary-position-embeddings \ + --untie-embeddings-and-output-weights \ + --swiglu \ + --normalization rmsnorm \ + --disable-bias-linear \ + --tensorboard-dir $LOG_DIR \ + --universal-checkpoint \ + $ds_args +" + +WORKER_STR="--num_nodes 1 --num_gpus $WORLD_SIZE" +run_cmd="deepspeed --master_port 29700 $WORKER_STR ${DIR}/pretrain_gpt.py $@ ${options}" + +echo ${options} +echo ${run_cmd} +eval ${run_cmd} diff --git a/examples_deepspeed/universal_checkpointing/run_bf16.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_bf16.sh similarity index 99% rename from examples_deepspeed/universal_checkpointing/run_bf16.sh rename to examples_deepspeed/universal_checkpointing/megatron_gpt/run_bf16.sh index 09539542226..07cbc30e725 100755 --- a/examples_deepspeed/universal_checkpointing/run_bf16.sh +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_bf16.sh @@ -3,7 +3,7 @@ DIR=`pwd` DATETIME=`date +'date_%y-%m-%d_time_%H-%M-%S'` -BASE_DATA_PATH=datasets +BASE_DATA_PATH=dataset DATASET=${BASE_DATA_PATH}/my-gpt2_text_document VOCAB_PATH=${BASE_DATA_PATH}/gpt2-vocab.json MERGE_PATH=${BASE_DATA_PATH}/gpt2-merges.txt diff --git a/examples_deepspeed/universal_checkpointing/run_fp16.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_fp16.sh similarity index 99% rename from examples_deepspeed/universal_checkpointing/run_fp16.sh rename to examples_deepspeed/universal_checkpointing/megatron_gpt/run_fp16.sh index 691fa8a8e6e..2f1b9940794 100755 --- a/examples_deepspeed/universal_checkpointing/run_fp16.sh +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_fp16.sh @@ -3,7 +3,7 @@ DIR=`pwd` DATETIME=`date +'date_%y-%m-%d_time_%H-%M-%S'` -BASE_DATA_PATH=datasets +BASE_DATA_PATH=dataset DATASET=${BASE_DATA_PATH}/my-gpt2_text_document VOCAB_PATH=${BASE_DATA_PATH}/gpt2-vocab.json MERGE_PATH=${BASE_DATA_PATH}/gpt2-merges.txt diff --git a/examples_deepspeed/universal_checkpointing/run_tb_analysis.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt.sh similarity index 96% rename from examples_deepspeed/universal_checkpointing/run_tb_analysis.sh rename to examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt.sh index 7aa988a0a03..3a17d667507 100755 --- a/examples_deepspeed/universal_checkpointing/run_tb_analysis.sh +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt.sh @@ -16,7 +16,6 @@ python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_scrip --tb_event_key "lm-loss-training/lm loss" \ --plot_name "uc_char_training_loss.png" \ --plot_title "Megatron-GPT Universal Checkpointing - Training Loss" \ - --use_sns # Validation Loss python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py \ @@ -26,4 +25,3 @@ python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_scrip --plot_name "uc_char_validation_loss.png" \ --plot_title "Megatron-GPT Universal Checkpointing - Validation Loss" \ --plot_y_label "Validation LM Loss" \ - --use_sns diff --git a/examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt_plot_only.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt_plot_only.sh new file mode 100755 index 00000000000..0c3ea5399c6 --- /dev/null +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_tb_analysis_gpt_plot_only.sh @@ -0,0 +1,31 @@ +#!/bin/bash +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +OUTPUT_PATH=$1 + +if [ "$OUTPUT_PATH" == "" ]; then + OUTPUT_PATH="z1_uni_ckpt" +fi + +# Training Loss +python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py \ + --tb_dir $OUTPUT_PATH \ + --tb_event_key "lm-loss-training/lm loss" \ + --plot_name "uc_char_training_loss.png" \ + --plot_title "Megatron-GPT Universal Checkpointing - Training Loss" \ + --plot_only \ + --csv_dir "/workspace/uc/megatron/loss_csv" \ + +# Validation Loss +python3 examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py \ + --tb_dir $OUTPUT_PATH \ + --tb_event_key "lm-loss-validation/lm loss validation" \ + --csv_name "val_" \ + --plot_name "uc_char_validation_loss.png" \ + --plot_title "Megatron-GPT Universal Checkpointing - Validation Loss" \ + --plot_y_label "Validation LM Loss" \ + --plot_only \ + --csv_dir "/workspace/uc/megatron/val_csv" \ diff --git a/examples_deepspeed/universal_checkpointing/run_universal_bf16.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_bf16.sh similarity index 99% rename from examples_deepspeed/universal_checkpointing/run_universal_bf16.sh rename to examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_bf16.sh index ef0e134cfc9..4134b9df482 100755 --- a/examples_deepspeed/universal_checkpointing/run_universal_bf16.sh +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_bf16.sh @@ -3,7 +3,7 @@ DIR=`pwd` DATETIME=`date +'date_%y-%m-%d_time_%H-%M-%S'` -BASE_DATA_PATH=datasets +BASE_DATA_PATH=dataset DATASET=${BASE_DATA_PATH}/my-gpt2_text_document VOCAB_PATH=${BASE_DATA_PATH}/gpt2-vocab.json MERGE_PATH=${BASE_DATA_PATH}/gpt2-merges.txt diff --git a/examples_deepspeed/universal_checkpointing/run_universal_fp16.sh b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_fp16.sh similarity index 99% rename from examples_deepspeed/universal_checkpointing/run_universal_fp16.sh rename to examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_fp16.sh index 1e207e422ba..bb3a538951b 100755 --- a/examples_deepspeed/universal_checkpointing/run_universal_fp16.sh +++ b/examples_deepspeed/universal_checkpointing/megatron_gpt/run_universal_fp16.sh @@ -3,7 +3,7 @@ DIR=`pwd` DATETIME=`date +'date_%y-%m-%d_time_%H-%M-%S'` -BASE_DATA_PATH=datasets +BASE_DATA_PATH=dataset DATASET=${BASE_DATA_PATH}/my-gpt2_text_document VOCAB_PATH=${BASE_DATA_PATH}/gpt2-vocab.json MERGE_PATH=${BASE_DATA_PATH}/gpt2-merges.txt diff --git a/examples_deepspeed/universal_checkpointing/tb_analysis/arguments.py b/examples_deepspeed/universal_checkpointing/tb_analysis/arguments.py index 3dacb45d4ee..ca80872ca02 100644 --- a/examples_deepspeed/universal_checkpointing/tb_analysis/arguments.py +++ b/examples_deepspeed/universal_checkpointing/tb_analysis/arguments.py @@ -17,3 +17,5 @@ parser.add_argument("--skip_csv", action='store_true', help="Skip generation of csv files") parser.add_argument("--use_sns", action='store_true', help="Use the SNS library to format plot") parser.add_argument("--csv_name", required=False, default="", type=str, help="Unique name for CSV files") +parser.add_argument("--plot_only", action='store_true', help="Plot only using csv files") +parser.add_argument("--csv_dir", required=False, type=str, help="Directory for csv files") diff --git a/examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py b/examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py index 337f6540ab5..fbf9b6dd284 100644 --- a/examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py +++ b/examples_deepspeed/universal_checkpointing/tb_analysis/tb_analysis_script.py @@ -6,9 +6,10 @@ import os import re import pandas as pd +import csv import matplotlib.pyplot as plt from tensorboard.backend.event_processing.event_accumulator import EventAccumulator -from utils import get_analyzer, find_files +from utils import get_analyzer, find_files_prefix, find_files_suffix from arguments import parser args = parser.parse_args() @@ -18,8 +19,8 @@ sns.set() def main(): - target_affix = 'events.out.tfevents' - tb_log_paths = find_files(args.tb_dir, target_affix) + target_prefix = 'events.out.tfevents' + tb_log_paths = find_files_prefix(args.tb_dir, target_prefix) analyzer = get_analyzer(args.analyzer) @@ -41,6 +42,8 @@ def main(): df = pd.DataFrame({"step": x, "value": y}) df.to_csv(f"{args.csv_name}{analyzer.get_csv_filename()}.csv") + plt.grid(True) + if not args.skip_plot: plt.legend() plt.title(args.plot_title) @@ -48,5 +51,35 @@ def main(): plt.ylabel(args.plot_y_label) plt.savefig(args.plot_name) +def plot_csv(): + target_suffix = 'csv' + csv_log_files = find_files_suffix(args.csv_dir, target_suffix) + + analyzer = get_analyzer(args.analyzer) + + for csv_file in csv_log_files: + analyzer.set_names(csv_file) + + x, y = [], [] + with open(csv_file, 'r') as file: + reader = csv.reader(file) + for row in reader: + if row[1] == 'step': + continue + x.append(int(row[1])) # Assuming the first column contains x values + y.append(float(row[2])) # Assuming the second column contains y values + + plt.plot(x, y, label=f'{analyzer.get_label_name()}') + + plt.grid(True) + plt.legend() + plt.title(args.plot_title) + plt.xlabel(args.plot_x_label) + plt.ylabel(args.plot_y_label) + plt.savefig(args.plot_name) + if __name__ == "__main__": - main() + if args.plot_only: + plot_csv() + else: + main() diff --git a/examples_deepspeed/universal_checkpointing/tb_analysis/uc_analysis.py b/examples_deepspeed/universal_checkpointing/tb_analysis/uc_analysis.py index f5809c3dc1d..20d46ff6a89 100644 --- a/examples_deepspeed/universal_checkpointing/tb_analysis/uc_analysis.py +++ b/examples_deepspeed/universal_checkpointing/tb_analysis/uc_analysis.py @@ -19,7 +19,7 @@ def set_names(self, path_name): tp, pp, dp, sp = match.groups() self._label_name = f"Training Run: TP: {tp}, PP: {pp}, DP: {dp}" - self._csv_name = f"uc_out_tp_{tp}_pp_{pp}_dp_{dp}_sp_{sp}" + self._csv_name = f"uc_out_tp{tp}_pp{pp}_dp{dp}_sp{sp}" def get_label_name(self): return self._label_name diff --git a/examples_deepspeed/universal_checkpointing/tb_analysis/utils.py b/examples_deepspeed/universal_checkpointing/tb_analysis/utils.py index 4bbbb3f2f04..db6624bbc45 100644 --- a/examples_deepspeed/universal_checkpointing/tb_analysis/utils.py +++ b/examples_deepspeed/universal_checkpointing/tb_analysis/utils.py @@ -7,13 +7,13 @@ from uc_analysis import UniversalCheckpointingAnalysis -def find_files(directory, file_affix): +def find_files_prefix(directory, file_prefix): """ - Searches for files with a specific affix in a directory using os.walk(). + Searches for files with a specific prefix in a directory using os.walk(). Args: directory (str): The path to the directory to search. - file_affix (str): The desired file affix. + file_prefix (str): The desired file prefix. Returns: list: A list of paths to matching files. @@ -21,10 +21,28 @@ def find_files(directory, file_affix): matching_paths = [] for root, _, files in os.walk(directory): for filename in files: - if root not in matching_paths and filename.lower().startswith(file_affix.lower()): + if root not in matching_paths and filename.lower().startswith(file_prefix.lower()): matching_paths.append(os.path.join(root)) return matching_paths +def find_files_suffix(directory, file_suffix): + """ + Searches for files with a specific suffix in a directory using os.walk(). + + Args: + directory (str): The path to the directory to search. + file_suffix (str): The desired file suffix. + + Returns: + list: A list of paths to matching files. + """ + matching_paths = [] + for root, _, files in os.walk(directory): + for filename in files: + if root not in matching_paths and filename.lower().endswith(file_suffix.lower()): + matching_paths.append(os.path.join(filename)) + return matching_paths + def get_analyzer(analyzer_name): if analyzer_name == 'universal_checkpointing': return UniversalCheckpointingAnalysis() From bcbe75f620b0964fe2285b236eaa9d1825aff141 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 31 Jul 2024 14:52:56 -0500 Subject: [PATCH 293/590] Update README.md --- ALCF/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/README.md b/ALCF/README.md index f4cb1acef30..b0eb99deb6b 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -42,7 +42,7 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): 1. Load `conda` module and activate base environment: ```bash - export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python + export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup ``` -

[output]: From 1bfc35c4e279b4edcd8291cb99dd3992e02bc8ba Mon Sep 17 00:00:00 2001 From: Jinghan Yao Date: Thu, 1 Aug 2024 15:35:11 -0700 Subject: [PATCH 294/590] fixing the bug of flash_attn import and the wrong gather index when using flash_attn_cuda in sequence parallel (#406) Co-authored-by: Jinghan Yao --- megatron/model/transformer.py | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index e79abea3cf2..74671905821 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -37,9 +37,12 @@ try: # FlashAttention (1.x) from flash_attn.flash_attn_interface import flash_attn_unpadded_func - from flash_attn.flash_attn_triton import flash_attn_func except ImportError: flash_attn_unpadded_func = None + +try: + from flash_attn.flash_attn_triton import flash_attn_func +except ImportError: flash_attn_func = None try: @@ -599,7 +602,11 @@ def __init__(self, config, layer_number, if self.enable_ds_sequence_parallel: assert dist_attn_supported, 'Distributed attention is not supported in this DeepSpeed version' assert args.num_attention_heads % parallel_state.get_sequence_parallel_world_size() == 0 - self.dist_attn = DistributedAttention(local_attn, parallel_state.get_sequence_parallel_group()) + self.dist_attn = DistributedAttention( + local_attn, + parallel_state.get_sequence_parallel_group(), + gather_idx=1 if args.use_flash_attn_v1 or args.use_flash_attn_v2 else 0) + # flash_attn_cuda assumes [b, s, nh, hd] layout, we need to make sure all2all gathers into the correct sequence dimension. else: if self.use_flash_attn: self.core_attention_flash = local_attn From 53b241f992f9b3dd7917bc36472f60cb118f8303 Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Mon, 5 Aug 2024 00:20:18 +0800 Subject: [PATCH 295/590] add fused_rms_norm support on XPU device (#431) --- megatron/model/__init__.py | 7 ++++++- megatron/model/fused_rmsnorm.py | 14 ++++++++++++++ 2 files changed, 20 insertions(+), 1 deletion(-) create mode 100644 megatron/model/fused_rmsnorm.py diff --git a/megatron/model/__init__.py b/megatron/model/__init__.py index 2306749fcb4..00c73223315 100644 --- a/megatron/model/__init__.py +++ b/megatron/model/__init__.py @@ -1,11 +1,16 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. +import torch +import intel_extension_for_pytorch from deepspeed.accelerator.real_accelerator import get_accelerator if get_accelerator().device_name() == 'cuda': from .fused_layer_norm import MixedFusedLayerNorm as LayerNorm from apex.normalization import MixedFusedRMSNorm as RMSNorm else: - from .rmsnorm import RMSNorm + if hasattr(torch.xpu, "IpexRmsNorm"): + from .fused_rmsnorm import RMSNorm + else: + from .rmsnorm import RMSNorm from torch.nn import LayerNorm from .distributed import DistributedDataParallel diff --git a/megatron/model/fused_rmsnorm.py b/megatron/model/fused_rmsnorm.py new file mode 100644 index 00000000000..a3b9927c874 --- /dev/null +++ b/megatron/model/fused_rmsnorm.py @@ -0,0 +1,14 @@ +import torch +from torch.nn.parameter import Parameter +import intel_extension_for_pytorch as ipex # noqa + +# Taken from facebookresearch/llama +class RMSNorm(torch.nn.Module): + def __init__(self, dim: int, eps: float = 1e-6): + super().__init__() + self.eps = eps + self.weight = Parameter(torch.ones(dim)) + + def forward(self, x): + output = torch.xpu.IpexRmsNorm(x, self.weight.shape, self.weight, self.eps) + return output From 61350c55478fba29ecf40940a629a3e7ce008a05 Mon Sep 17 00:00:00 2001 From: Jinghan Yao Date: Wed, 7 Aug 2024 16:11:57 -0700 Subject: [PATCH 296/590] pass batch_dim_idx to deepspeed sequence parallel distributed attention for supporting batch size larger than 1 (#433) Co-authored-by: Jinghan Yao --- megatron/model/transformer.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index 74671905821..036c11566a3 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -817,12 +817,14 @@ def forward(self, hidden_states, attention_mask, # value_layer = apply_rotary_pos_emb(value_layer, k_pos_emb) if self.enable_ds_sequence_parallel: + batch_dim_idx = 1 if self.use_flash_attn: if not self.use_flash_attn_triton: query_layer, key_layer, value_layer = [rearrange(x, 's b ... -> b s ...').contiguous() for x in (query_layer, key_layer, value_layer)] + batch_dim_idx = 0 - context_layer = self.dist_attn(query_layer, key_layer, value_layer) + context_layer = self.dist_attn(query_layer, key_layer, value_layer, batch_dim_idx) if not self.use_flash_attn_triton: context_layer = rearrange(context_layer, 'b s h d -> s b (h d)').contiguous() From f132876c41915483d07e6424addc9b5e4857ec4c Mon Sep 17 00:00:00 2001 From: billishyahao Date: Sat, 10 Aug 2024 23:45:27 +0800 Subject: [PATCH 297/590] [LLaMa] Adding support converting checkpoint from mds to hf (#432) * add support converting checkpoint from hf to mds * Fix PP issue * update --- .../finetune_hf_llama/ds_config.json | 8 +- .../finetune_hf_llama/finetune_llama.sh | 23 +- megatron/global_vars.py | 1 + tools/hf2megads_weight_converter.py | 306 +++++++++++++++--- 4 files changed, 280 insertions(+), 58 deletions(-) diff --git a/examples_deepspeed/finetune_hf_llama/ds_config.json b/examples_deepspeed/finetune_hf_llama/ds_config.json index 9c0b332473e..85f439ce472 100755 --- a/examples_deepspeed/finetune_hf_llama/ds_config.json +++ b/examples_deepspeed/finetune_hf_llama/ds_config.json @@ -1,11 +1,5 @@ { "train_batch_size" : 256, "train_micro_batch_size_per_gpu": 16, - "steps_per_print": 100, - "zero_optimization": { - "stage": 0 - }, - "bf16": { - "enabled": true - } + "steps_per_print": 1 } diff --git a/examples_deepspeed/finetune_hf_llama/finetune_llama.sh b/examples_deepspeed/finetune_hf_llama/finetune_llama.sh index c48ea11b93d..ab8bfdf419b 100644 --- a/examples_deepspeed/finetune_hf_llama/finetune_llama.sh +++ b/examples_deepspeed/finetune_hf_llama/finetune_llama.sh @@ -1,8 +1,8 @@ DS_CONFIG=./examples_deepspeed/finetune_hf_llama/ds_config.json -DATASET_PATH=./alpaca_data.json +DATASET_PATH=./examples_deepspeed/finetune_hf_llama/alpaca_data.json # dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json -HF_LLAMA_PATH=/data/llama-7b/ +HF_LLAMA_PATH=/data/llama-2-7b-hf/ # weights link: https://huggingface.co/huggyllama/llama-7b MICRO_BATCH_SIZE=16 @@ -44,11 +44,20 @@ cat < $DS_CONFIG EOT -covert_args="deepspeed tools/hf2megads_weight_converter.py \ +covert_hf2mds_args="deepspeed tools/hf2megads_weight_converter.py \ --hf-ckpt-num-shards 2 \ ---origin-hf-ckpt-dir $HF_LLAMA_PATH \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ --save $MEGA_DS_LLAMA_PATH" +covert_mds2hf_args="deepspeed tools/hf2megads_weight_converter.py \ +--hf-ckpt-num-shards 2 \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ +--to-hf-ckpt \ +--load $MEGA_DS_LLAMA_PATH \ +--save $HF_LLAMA_PATH'-hf-out' " + finetune_args="deepspeed finetune_llama.py \ --load $MEGA_DS_LLAMA_PATH" @@ -98,8 +107,10 @@ comm_args="--tensor-model-parallel-size $TP \ --no-gradient-accumulation-fusion \ --repeated-dataloader" -if [ "$1" = "convert" ]; then - task_args="$covert_args" +if [ "$1" = "convert_hf2mds" ]; then + task_args="$covert_hf2mds_args" +elif [ "$1" = "convert_mds2hf" ]; then + task_args="$covert_mds2hf_args" else task_args="$finetune_args" fi diff --git a/megatron/global_vars.py b/megatron/global_vars.py index 3f9d6fd66b4..ccd0a4e21ec 100644 --- a/megatron/global_vars.py +++ b/megatron/global_vars.py @@ -175,6 +175,7 @@ def _set_wandb_writer(args): 'project or experiment name provided, ' 'therefore WANDB logs will be written ' 'according to random generated project or experiment name.', flush=True) + return try: import wandb diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py index bfbde1fd055..12468963c54 100755 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -3,9 +3,11 @@ import sys import os sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) +import torch.distributed from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP from megatron import print_rank_0, get_tokenizer, get_args from megatron.core import mpu +from megatron.core import tensor_parallel from megatron.core.utils import divide from megatron.model import GPTModelPipe, Float16Module from megatron.utils import unwrap_model @@ -13,20 +15,30 @@ from megatron.arguments import core_transformer_config_from_args from megatron.initialize import initialize_megatron from megatron.optimizer import get_megatron_optimizer -from megatron.checkpointing import save_checkpoint +from megatron.checkpointing import save_checkpoint, load_checkpoint from megatron.training import get_optimizer_param_scheduler from deepspeed.runtime.utils import see_memory_usage import deepspeed +import copy +from pathlib import Path + def add_extra_args(parser): """Text generation arguments.""" group = parser.add_argument_group(title='hf2mega') - group.add_argument("--hf-ckpt-num-shards", type=int, help='num of llama ckpt.') - group.add_argument("--origin-hf-ckpt-dir", + group.add_argument("--hf-ckpt-dir", type=str, default="", - help="the original path of the llama-hf ckpt") + help="the llama-hf ckpt") + group.add_argument("--hf-ckpt-num-shards", type=int, default=-1, help='num of llama ckpt.') + group.add_argument("--load-mode", type=str, + default=None, + choices=['torchbin', 'safetensor', 'auto'], + help="load ckpt format: pytorch.bin or model.safetensor or auto.") + group.add_argument("--to-hf-ckpt", action="store_true", + help="by default convert from hf to megads" + "if set, convert reversely from megads to hf ckpt.") return parser @@ -55,6 +67,49 @@ def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): return loaded +def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards): + from safetensors import safe_open + # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. + hf_model = {} + print_rank_0( + f"----------------------------hf weight list----------------------------") + + for wid in range(1, hf_ckpt_num_of_shards + 1): + if hf_ckpt_num_of_shards == 1: + ckpt_path = f"{hf_ckpt_dir}/model.safetensors" + else: + ckpt_path = f"{hf_ckpt_dir}/model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.safetensors" + + with safe_open(ckpt_path, framework="pt", device="cpu") as f: + for k in f.keys(): + print_rank_0(f"name: {k}, shape: {f.get_tensor(k).shape}") + assert k not in hf_model + hf_model[k] = f.get_tensor(k).clone() + + return hf_model + + +def load_and_print_hf_weight_auto(hf_ckpt_dir, no_init=True): + from transformers import AutoConfig, AutoModelForCausalLM + from transformers.modeling_utils import no_init_weights + + if no_init: + hf_config = AutoConfig.from_pretrained(hf_ckpt_dir, trust_remote_code=True) + with no_init_weights(): + hf_model = AutoModelForCausalLM.from_config(hf_config, trust_remote_code=True, torch_dtype=torch.bfloat16) + else: + hf_model = {} + hf_auto_model = AutoModelForCausalLM.from_pretrained(hf_ckpt_dir, trust_remote_code=True, torch_dtype=torch.bfloat16) + print_rank_0( + f"----------------------------hf weight list----------------------------") + + for name, param in hf_auto_model.named_parameters(): + hf_model[name] = param.clone() + print_rank_0(name) + + return hf_model + + def print_distinct_weights(model): print_rank_0( f"----------------------------mega-ds weight list----------------------------") @@ -70,16 +125,19 @@ def print_distinct_weights(model): class refactor: - def __init__(self, model, loaded, args, config): + def __init__(self, ds_model, hf_model, args, config): tokenizer = get_tokenizer() # align layer number - self.model = model - self.loaded = loaded + self.ds_model = ds_model + self.hf_model = hf_model + self.hf_dict = {} # for handling pp case when converting mds => hf self.config = config self.offset_num = 2 self.mega_emb_wnum = 1 self.mega_norm_wnum = args.num_layers + 2 + self.num_attention_heads = args.num_attention_heads + self.num_key_value_heads = args.num_key_value_heads self.mega_lm_head_wnum = self.mega_norm_wnum + 1 self.token_vocab = tokenizer.vocab_size self.padded_vocab_size = args.padded_vocab_size @@ -95,7 +153,7 @@ def _embedding_refactor(self, pname, p): hf_name = "lm_head.weight" elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": hf_name = "model.embed_tokens.weight" - hf_w = self.loaded[hf_name] + hf_w = self.hf_model[hf_name] assert hf_w.shape[0] == self.token_vocab per_partition_vocab_size, start_index, end_index = compute_partition_range( self.padded_vocab_size, self.tp_rank, self.tp_size) @@ -112,24 +170,28 @@ def _embedding_refactor(self, pname, p): ) return new_w + + + def _direct_refactor(self, pname, p, hf_layer=None, subname=None): if pname == f"{self.mega_norm_wnum}.weight": hf_name = "model.norm.weight" elif subname in ["input_layernorm.weight", "post_attention_layernorm.weight"]: hf_name = f"model.layers.{hf_layer}.{subname}" - new_w = hf_w = self.loaded[hf_name] + new_w = hf_w = self.hf_model[hf_name] self.record_mapping_info( f"mega-ds:{pname,p.data.shape}<--hf{hf_name,} {hf_w.shape}") return new_w + def _qkv_refactor(self, pname, p, hf_layer): hf_wq_name = f"model.layers.{hf_layer}.self_attn.q_proj.weight" hf_wk_name = f"model.layers.{hf_layer}.self_attn.k_proj.weight" hf_wv_name = f"model.layers.{hf_layer}.self_attn.v_proj.weight" - wq = self.loaded[hf_wq_name] - wk = self.loaded[hf_wk_name] - wv = self.loaded[hf_wv_name] + wq = self.hf_model[hf_wq_name] + wk = self.hf_model[hf_wk_name] + wv = self.hf_model[hf_wv_name] hidden_size = wq.shape[0] per_partition_size, start_index, end_index = compute_partition_range( @@ -159,8 +221,8 @@ def _qkv_refactor(self, pname, p, hf_layer): def _mlphto4h_dense_refactor(self, pname, p, hf_layer): hf_w_gate_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" hf_w_up_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" - w_gate = self.loaded[hf_w_gate_name] - w_up = self.loaded[hf_w_up_name] + w_gate = self.hf_model[hf_w_gate_name] + w_up = self.hf_model[hf_w_up_name] hidden_size = w_gate.shape[0] per_partition_size, start_index, end_index = compute_partition_range( @@ -184,7 +246,7 @@ def _attn_dense_refactor(self, pname, p, hf_layer, subname): else: hf_name = f"model.layers.{hf_layer}.mlp.down_proj.weight" - hf_w = self.loaded[hf_name] + hf_w = self.hf_model[hf_name] hidden_size = hf_w.shape[1] per_partition_size, start_index, end_index = compute_partition_range( hidden_size, self.tp_rank, self.tp_size) @@ -200,7 +262,7 @@ def _mlphto4h1_refactor(self, pname, p, hf_layer, subname): hf_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" else: hf_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" - hf_w = self.loaded[hf_name] + hf_w = self.hf_model[hf_name] hidden_size = hf_w.shape[0] per_partition_size, start_index, end_index = compute_partition_range( hidden_size, self.tp_rank, self.tp_size) @@ -212,10 +274,11 @@ def _mlphto4h1_refactor(self, pname, p, hf_layer, subname): ) return new_w - def refactor(self): + def transform_from_hf_to_megds(self): assert self.is_refactored == False new_w = None - for pname, p in self.model.named_parameters(): + for pname, p in self.ds_model.named_parameters(): + if pname in [ f"{self.mega_emb_wnum}.word_embeddings.weight", f"{self.mega_lm_head_wnum}.lm_head.weight" @@ -253,6 +316,123 @@ def refactor(self): new_w = None self.is_refactored = True + + def _embedding_refactor_to_hf(self, pname, ds_w): + if pname == f"{self.mega_lm_head_wnum}.lm_head.weight": + hf_w = self.hf_model.lm_head.weight + hf_w_name = "lm_head.weight" + elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": + hf_w = self.hf_model.model.embed_tokens.weight + hf_w_name = "model.embed_tokens.weight" + + with torch.no_grad(): + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) + + self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank[:hf_w.shape[0], :]) + + def _direct_refactor_to_hf(self, pname, ds_w, hf_layer=None, subname=None): + if pname in [f"{self.mega_norm_wnum}.weight"]: + hf_w = self.hf_model.model.norm.weight + hf_w_name = "model.norm.weight" + elif subname in ["input_layernorm.weight"]: + hf_w = self.hf_model.model.layers[hf_layer].input_layernorm.weight + hf_w_name = f"model.layers.{hf_layer}.input_layernorm.weight" + elif subname in ["post_attention_layernorm.weight"]: + hf_w = self.hf_model.model.layers[hf_layer].post_attention_layernorm.weight + hf_w_name = f"model.layers.{hf_layer}.post_attention_layernorm.weight" + + self.hf_dict[hf_w_name] = copy.deepcopy(ds_w) + + def _attn_dense_refactor_to_hf(self, pname, ds_w, hf_layer, subname): + if subname == "self_attention.dense.weight": + hf_w = self.hf_model.model.layers[hf_layer].self_attn.o_proj.weight + hf_w_name = f"model.layers.{hf_layer}.self_attn.o_proj.weight" + elif subname == "mlp.dense_4h_to_h.weight": + hf_w = self.hf_model.model.layers[hf_layer].mlp.down_proj.weight + hf_w_name = f"model.layers.{hf_layer}.mlp.down_proj.weight" + + with torch.no_grad(): + ds_w_all_rank = tensor_parallel.mappings._gather_along_last_dim(ds_w) + + self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank) + + def _mlphto4h_dense_refactor_to_hf(self, pname, ds_w, hf_layer): + hf_g_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" + hf_u_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" + + with torch.no_grad(): + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) + + ds_w_shape = ds_w_all_rank.shape + ds_w_all_rank = ds_w_all_rank.reshape(self.tp_size, 2, -1, ds_w_shape[-1]) + self.hf_dict[hf_g_name] = copy.deepcopy(ds_w_all_rank[:, 0, :, :].reshape(-1, ds_w_shape[-1])) + self.hf_dict[hf_u_name] = copy.deepcopy(ds_w_all_rank[:, 1, :, :].reshape(-1, ds_w_shape[-1])) + + + def _qkv_refactor_to_hf(self, pname, ds_w, hf_layer): + with torch.no_grad(): + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) + + hf_q = self.hf_model.model.layers[hf_layer].self_attn.q_proj.weight + hf_k = self.hf_model.model.layers[hf_layer].self_attn.k_proj.weight + hf_v = self.hf_model.model.layers[hf_layer].self_attn.v_proj.weight + hf_q_name = f"model.layers.{hf_layer}.self_attn.q_proj.weight" + hf_k_name = f"model.layers.{hf_layer}.self_attn.k_proj.weight" + hf_v_name = f"model.layers.{hf_layer}.self_attn.v_proj.weight" + oldshape = hf_q.shape + hidden_size = oldshape[-1] + hidden_size_per_attention_head = divide(hidden_size, + self.config.num_attention_heads) + num_attention_heads_per_partition = divide(self.config.num_attention_heads, + self.tp_size) + newshape = (self.tp_size, num_attention_heads_per_partition, 3, hidden_size_per_attention_head, hidden_size) + ds_w_out = ds_w_all_rank.reshape(*newshape) + self.hf_dict[hf_q_name] = copy.deepcopy(ds_w_out[:, :, 0, :, :].reshape(-1, oldshape[-1])) + self.hf_dict[hf_k_name] = copy.deepcopy(ds_w_out[:, :, 1, :, :].reshape(-1, oldshape[-1])) + self.hf_dict[hf_v_name] = copy.deepcopy(ds_w_out[:, :, 2, :, :].reshape(-1, oldshape[-1])) + + + def transform_from_megads_to_hf(self): + use_gqa = True if self.num_attention_heads != self.num_key_value_heads else False + + for pname, p in self.ds_model.named_parameters(): + if pname in [ + f"{self.mega_emb_wnum}.word_embeddings.weight", + f"{self.mega_lm_head_wnum}.lm_head.weight", + ]: + self._embedding_refactor_to_hf(pname, p) + elif pname in [ + f"{self.mega_norm_wnum}.weight", + ]: + self._direct_refactor_to_hf(pname, p) + else: + mobj = self.decoder_pat.match(pname) + layer_num = int(mobj.group(1)) + subname = mobj.group(2) + hf_layer = layer_num - self.offset_num + if subname in ["self_attention.query_key_value.weight"]: + if not use_gqa: + self._qkv_refactor_to_hf(pname, p, hf_layer) + else: + #TODO(billishyahao): Not impl yet ... + assert False + elif subname in ["mlp.dense_h_to_4h.weight"]: + self._mlphto4h_dense_refactor_to_hf(pname, p, hf_layer) + elif subname in [ + "self_attention.dense.weight", + "mlp.dense_4h_to_h.weight" + ]: + self._attn_dense_refactor_to_hf(pname, p, hf_layer, subname) + elif subname in [ + "input_layernorm.weight", + "post_attention_layernorm.weight", + ]: + self._direct_refactor_to_hf(pname, p, hf_layer, subname) + else: + print(f"Unrecognized weight type: {pname}") + raise ValueError(f"Unrecognized weight type: {pname}") + self.is_refactored = True + def record_mapping_info(self, record_msg): self.refactor_weight_list.append(record_msg) @@ -272,7 +452,18 @@ def inorder_show_record(self): torch.distributed.barrier() -def convert_hf_to_mega_ds(): +def load_hf_weights(args, no_init): + if args.load_mode == 'torchbin': + assert no_init == False, "only work with init" + return load_and_print_hf_weight(args.hf_ckpt_dir, args.hf_ckpt_num_shards) + elif args.load_mode == 'safetensor': + assert no_init == False, "only work with init" + return load_and_print_hf_weight_from_safetensor(args.hf_ckpt_dir, args.hf_ckpt_num_shards) + elif args.load_mode == 'auto': + return load_and_print_hf_weight_auto(args.hf_ckpt_dir, no_init) + + +def convert_ckpt(): """Build the model.""" args = get_args() print_rank_0(f'building model ...') @@ -286,49 +477,74 @@ def convert_hf_to_mega_ds(): enabled=args.zero_stage == 3, mpu=mpu): if args.deepspeed and not args.no_pipeline_parallel: - model = GPTModelPipe(config, num_tokentypes=0, parallel_output=True) + ds_model = GPTModelPipe(config, num_tokentypes=0, parallel_output=True) else: raise NotImplementedError("Not implemented") see_memory_usage(f"After Building Model", force=True) if torch.distributed.get_rank() < 2: - print(f"{torch.distributed.get_rank()} {model}") - - # load and initialize HF weight dict - # print hf weights list & mega-ds weights list - hf_ckpt_dir = args.origin_hf_ckpt_dir - hf_ckpt_num_of_shards = args.hf_ckpt_num_shards - loaded = load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards) - print_distinct_weights(model) - - # refactor weight from hf to mega-ds - - cur_refactor = refactor(model, loaded, args, config) - cur_refactor.refactor() - cur_refactor.inorder_show_record() + print(f"{torch.distributed.get_rank()} {ds_model}") - del loaded + # 'torchbin', 'safetensor', 'auto' + hf_model = load_hf_weights(args, no_init=args.to_hf_ckpt) - unwrapped_model = unwrap_model([model], (torchDDP, LocalDDP, Float16Module)) - optimizer = get_megatron_optimizer(unwrapped_model) - opt_param_scheduler = get_optimizer_param_scheduler(optimizer) + # print_distinct_weights(hf_model) #init model and save print_rank_0(f"before deepspeed init") ds_engine, _, _, _ = deepspeed.initialize( - model=model, - optimizer=optimizer, + model=ds_model, + optimizer=None, args=args, - lr_scheduler=opt_param_scheduler, + lr_scheduler=None, mpu=mpu if args.no_pipeline_parallel else None) print_rank_0(f"after deepspeed init") - print_rank_0(f"mega-ds checkpoint will be saved in {args.save}") - save_checkpoint(0, [ds_engine], optimizer, opt_param_scheduler) - print_rank_0(f"save checkpoint completed") + if args.to_hf_ckpt: + load_checkpoint([ds_engine], None, None, load_only_weights=True) + print_rank_0(f"completed to load deepspeed actual checkpoint") + + # refactor weight from hf to mega-ds and vice versa + + cur_refactor = refactor(ds_model, hf_model, args, config) + if args.to_hf_ckpt: + cur_refactor.transform_from_megads_to_hf() + else: + cur_refactor.transform_from_hf_to_megds() + # cur_refactor.inorder_show_record() + + if args.to_hf_ckpt: + save_path = args.save + if not os.path.exists(save_path): + Path(save_path).mkdir(parents=True, exist_ok=True) + ckpt_per_pp_path = os.path.join(save_path, f"model_pp{mpu.get_pipeline_model_parallel_rank()}.pt") + torch.save(cur_refactor.hf_dict, ckpt_per_pp_path) + if torch.distributed.is_initialized(): + torch.distributed.barrier() + + print_rank_0(f"hf checkpoint will be saved in {save_path}/release ") + if mpu.is_pipeline_last_stage(): + ## doing checkpoint merging and saving... + # hf_model.tie_weights() + + all_wei = {} + for pprank in range(mpu.get_pipeline_model_parallel_world_size()): + ckpt_per_pp_path = os.path.join(save_path, f"model_pp{pprank}.pt") + partial_wei = torch.load(ckpt_per_pp_path) + all_wei = all_wei | partial_wei + + hf_model.load_state_dict(all_wei) + + # mega-ds checkpoint will be saved in args.save + hf_model.save_pretrained(os.path.join(save_path, "release"), safe_serialization=True) + else: + print_rank_0(f"mega-ds checkpoint will be saved in {args.save}") + save_checkpoint(0, [ds_engine], None, None) + + print_rank_0(f"save checkpoint completed") if __name__ == "__main__": initialize_megatron(extra_args_provider=add_extra_args) - convert_hf_to_mega_ds() + convert_ckpt() From cdf51944deecad75572ebfd7cddfb0abfd1b9987 Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Thu, 15 Aug 2024 04:04:01 +0800 Subject: [PATCH 298/590] add device check when import ipex (#436) --- megatron/model/__init__.py | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/megatron/model/__init__.py b/megatron/model/__init__.py index 00c73223315..64158ef99af 100644 --- a/megatron/model/__init__.py +++ b/megatron/model/__init__.py @@ -1,8 +1,10 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. import torch -import intel_extension_for_pytorch from deepspeed.accelerator.real_accelerator import get_accelerator + +if get_accelerator().device_name() == 'xpu': + import intel_extension_for_pytorch if get_accelerator().device_name() == 'cuda': from .fused_layer_norm import MixedFusedLayerNorm as LayerNorm from apex.normalization import MixedFusedRMSNorm as RMSNorm From b7b2d5ef330f43729b406630e6c5d38e873d7398 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Mon, 19 Aug 2024 19:05:27 +0530 Subject: [PATCH 299/590] fix TFLOPs calculation (#371) * fix TFLOPs calculation when GQA used, we observe right TFLOPs after this fix. when GQA is not used, huge difference in TFLOPs is solved with selective recompute . some other minor difference will also be observed as logits macs also added. * add copyrights --- megatron/utils.py | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index cbb7aa64268..15160b16447 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -275,22 +275,38 @@ def throughput_calculator(model, args, iteration_time, total_iterations): #flops calculator hidden_size = args.hidden_size + num_attention_heads = args.num_attention_heads + head_dim = hidden_size // num_attention_heads + ffn_hidden_size = args.ffn_hidden_size num_layers = args.num_layers vocab_size = args.padded_vocab_size + gqa = args.num_attention_heads // args.num_key_value_heads + ffn_multiplier = 3 if args.swiglu else 2 + macs_per_flops = 2 # General TFLOPs formula (borrowed from Equation 3 in Section 5.1 of # https://arxiv.org/pdf/2104.04473.pdf). - # The factor of 4 is when used with activation check-pointing, - # otherwise it will be 3. - checkpoint_activations_factor = 3 - if hasattr(args, 'checkpoint_activations') and args.checkpoint_activations: - checkpoint_activations_factor = 4 - if hasattr(args, 'recompute_granularity') and (args.recompute_granularity == 'selective' or args.recompute_granularity == 'full'): - checkpoint_activations_factor = 4 + # correction has been made to TFLOPs formula due to incorrect behavior + # observed with selective recompute when GQA not used and for all with GQA seq_len = args.seq_length if hasattr(args, 'actual_seq_length'): seq_len = args.actual_seq_length - flops_per_iteration = (24 * checkpoint_activations_factor * batch_size * seq_len * num_layers * (hidden_size**2)) * (1. + (seq_len / (6. * hidden_size)) + (vocab_size / (16. * num_layers * hidden_size))) + + pre_and_post_mha_gemm_macs = batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len + mha_bgemm_macs = batch_size * num_layers * 2 * head_dim * num_attention_heads * (seq_len**2) + ffn_gemm_macs = batch_size * num_layers * ffn_multiplier * ffn_hidden_size * hidden_size * seq_len + logit_lmhead_gemm_macs = batch_size * vocab_size * hidden_size * seq_len + + fwd_macs = pre_and_post_mha_gemm_macs + mha_bgemm_macs + ffn_gemm_macs + logit_lmhead_gemm_macs + bwd_macs = 2 * fwd_macs + fwd_bwd_macs = fwd_macs + bwd_macs + + if (hasattr(args, 'checkpoint_activations') and args.checkpoint_activations) or (hasattr(args, 'recompute_granularity') and args.recompute_granularity == 'full'): + fwd_bwd_macs += fwd_macs + if hasattr(args, 'recompute_granularity') and args.recompute_granularity == 'selective': + fwd_bwd_macs += mha_bgemm_macs + + flops_per_iteration = fwd_bwd_macs * macs_per_flops tflops = flops_per_iteration / (elapsed_time_per_iter * args.world_size * (10**12)) return samples_per_second, tflops, approx_parameters_in_billions From b7c17ca07531c7091d54dd33200822b7a1bb7a18 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 23 Aug 2024 12:22:36 -0500 Subject: [PATCH 300/590] Move `ALCF/mds_to_hf.py` to `mds_to_hf.py` --- ALCF/mds_to_hf.py | 91 -------------------------------- mds_to_hf.py | 130 ++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 130 insertions(+), 91 deletions(-) delete mode 100644 ALCF/mds_to_hf.py create mode 100644 mds_to_hf.py diff --git a/ALCF/mds_to_hf.py b/ALCF/mds_to_hf.py deleted file mode 100644 index a3367882740..00000000000 --- a/ALCF/mds_to_hf.py +++ /dev/null @@ -1,91 +0,0 @@ -# Usage : python mds_to_hf.py --mds_checkpoint --output_dir -# Tips : Do not run on login node. -# This script currently only takes care of tp=1. Takes a AuroraGPT Llama model trained with Megatron-DeepSpeed and converts to LLamaCausalForLM architecture from HuggingFace. - -import argparse -import torch -import pdb -import os -from transformers import LlamaConfig, LlamaForCausalLM, LlamaTokenizer - -def repeat_kv_wt(x,np): - return torch.repeat_interleave(x, dim=0, repeats=np) - -def Update_llama_config(Llama_config, mds_args): - if mds_args['swiglu']: - Llama_config.hidden_act = "silu" - Llama_config.hidden_size = mds_args['hidden_size'] - Llama_config.intermediate_size = mds_args['ffn_hidden_size'] - Llama_config.max_position_embeddings = mds_args['max_position_embeddings'] - Llama_config.num_attention_heads = mds_args['num_attention_heads'] - Llama_config.num_hidden_layers = mds_args['num_layers'] - Llama_config.num_key_value_heads = mds_args['num_key_value_heads'] - Llama_config.rms_norm_eps = mds_args['layernorm_epsilon'] - Llama_config.rope_theta = mds_args['rope_theta'] - Llama_config.vocab_size = mds_args['padded_vocab_size'] - return Llama_config - - -if __name__ == "__main__": - parser = argparse.ArgumentParser() - parser.add_argument('--mds_checkpoint', required=True) - parser.add_argument('--output_dir', required=True) - args = parser.parse_args() - - # make output_dir if it does not exits. - if not os.path.exists(args.output_dir): - os.makedirs(args.output_dir) - - filename = str(args.mds_checkpoint) - if not filename.split("/")[-1].startswith('mp_rank') and not filename.split("/")[-1].endswith('.pt'): - assert ("Provide the right file path, The file should be of format mp_rank_*.pt") - print(f"loading mds checkpoint {filename}") - - mds_model = torch.load(args.mds_checkpoint,map_location=torch.device('cpu')) - Llama_model = LlamaForCausalLM.from_pretrained("meta-llama/Llama-2-7b-hf",cache_dir='/eagle/datascience/vsastry/huggingface') - - Llama_config = Llama_model.config - Updated_Llama_config = Update_llama_config(Llama_config, mds_model['args'].__dict__) - # save the updated config.json file - Updated_Llama_config.to_json_file(os.path.join(args.output_dir,'config.json')) - - state_dict = {} - dim = mds_model['args'].__dict__['kv_channels'] - inv_freq = 1.0 / (mds_model['args'].__dict__['rope_theta'] ** (torch.arange(0,dim, 2).float() / dim)) - hidden_size = mds_model['args'].__dict__['hidden_size'] - kv_dim = mds_model['args'].__dict__['kv_channels'] * mds_model['args'].__dict__['num_key_value_heads'] - kv_groups = mds_model['args'].__dict__['num_attention_heads'] // mds_model['args'].__dict__['num_key_value_heads'] - for layer_i in range(Updated_Llama_config.__dict__['num_hidden_layers']): - # SELF ATTENTION layers. - # get the q, k, v weights separately. Keeping k and v at the GQA head dim, since the transformers/models/llama/modelling_utils will take care of it. - fused_qkv = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.query_key_value.weight"] - state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = fused_qkv[0:hidden_size] - state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = fused_qkv[hidden_size:hidden_size+kv_dim] - #state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size:hidden_size+kv_dim], kv_groups) - state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim] - #state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim],kv_groups) - state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.dense.weight"] - - # MLP Layers - fused_mlp = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_h_to_4h.weight"] - chunked_mlp = torch.chunk(fused_mlp,2,dim=0) - state_dict[f"model.layers.{layer_i}.mlp.gate_proj.weight"] = chunked_mlp[0] - state_dict[f"model.layers.{layer_i}.mlp.up_proj.weight"] = chunked_mlp[1] - state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_4h_to_h.weight"] - - #LayerNorm weights and RoPe - state_dict[f"model.layers.{layer_i}.input_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.input_layernorm.weight"] - state_dict[f"model.layers.{layer_i}.post_attention_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.post_attention_layernorm.weight"] - - state_dict[f"model.layers.{layer_i}.self_attn.rotary_emb.inv_freq"] = inv_freq - - # Get the non-encoder layer weights. - state_dict["model.embed_tokens.weight"] = mds_model['module']['language_model']['embedding']['word_embeddings']['weight'] - state_dict["model.norm.weight"] = mds_model['module']['language_model']['encoder']['final_layernorm.weight'] - state_dict["lm_head.weight"] = mds_model['module']['language_model']['output_layer']['weight'] - - # Save the model in the hf output path. - torch.save(state_dict, os.path.join(args.output_dir,"pytorch_model.bin")) - - - diff --git a/mds_to_hf.py b/mds_to_hf.py new file mode 100644 index 00000000000..9f2d87cc486 --- /dev/null +++ b/mds_to_hf.py @@ -0,0 +1,130 @@ +# Usage : python mds_to_hf.py --mds_checkpoint --output_dir +# Tips : Do not run on login node. +# This script currently only takes care of tp=1. Takes a AuroraGPT Llama model trained with Megatron-DeepSpeed and converts to LLamaCausalForLM architecture from HuggingFace. + +import argparse +import torch +import pdb +import os +from transformers import LlamaConfig, LlamaForCausalLM, LlamaTokenizer + + +def repeat_kv_wt(x, np): + return torch.repeat_interleave(x, dim=0, repeats=np) + + +def Update_llama_config(Llama_config, mds_args): + if mds_args["swiglu"]: + Llama_config.hidden_act = "silu" + Llama_config.hidden_size = mds_args["hidden_size"] + Llama_config.intermediate_size = mds_args["ffn_hidden_size"] + Llama_config.max_position_embeddings = mds_args["max_position_embeddings"] + Llama_config.num_attention_heads = mds_args["num_attention_heads"] + Llama_config.num_hidden_layers = mds_args["num_layers"] + Llama_config.num_key_value_heads = mds_args["num_key_value_heads"] + Llama_config.rms_norm_eps = mds_args["layernorm_epsilon"] + Llama_config.rope_theta = mds_args["rope_theta"] + Llama_config.vocab_size = mds_args["padded_vocab_size"] + return Llama_config + + +if __name__ == "__main__": + from pathlib import Path + + parser = argparse.ArgumentParser() + parser.add_argument("--mds_checkpoint", required=True) + parser.add_argument("--output_dir", required=True) + args = parser.parse_args() + + # make output_dir if it does not exits. + if not os.path.exists(args.output_dir): + os.makedirs(args.output_dir) + + filename = str(args.mds_checkpoint) + if not filename.split("/")[-1].startswith("mp_rank") and not filename.split("/")[ + -1 + ].endswith(".pt"): + assert "Provide the right file path, The file should be of format mp_rank_*.pt" + print(f"loading mds checkpoint {filename}") + + cache_dir = Path(os.getcwd()).joinpath(".cache", "hugging_face") + cache_dir.mkdir(exist_ok=True, parents=True) + + mds_model = torch.load(args.mds_checkpoint, map_location=torch.device("cpu")) + Llama_model = LlamaForCausalLM.from_pretrained( + "meta-llama/Llama-2-7b-hf", cache_dir=cache_dir.as_posix() + ) + Llama_config = Llama_model.config + Updated_Llama_config = Update_llama_config(Llama_config, mds_model["args"].__dict__) + # save the updated config.json file + Updated_Llama_config.to_json_file(os.path.join(args.output_dir, "config.json")) + + state_dict = {} + dim = mds_model["args"].__dict__["kv_channels"] + inv_freq = 1.0 / ( + mds_model["args"].__dict__["rope_theta"] + ** (torch.arange(0, dim, 2).float() / dim) + ) + hidden_size = mds_model["args"].__dict__["hidden_size"] + kv_dim = ( + mds_model["args"].__dict__["kv_channels"] + * mds_model["args"].__dict__["num_key_value_heads"] + ) + kv_groups = ( + mds_model["args"].__dict__["num_attention_heads"] + // mds_model["args"].__dict__["num_key_value_heads"] + ) + for layer_i in range(Updated_Llama_config.__dict__["num_hidden_layers"]): + # SELF ATTENTION layers. + # get the q, k, v weights separately. Keeping k and v at the GQA head dim, since the transformers/models/llama/modelling_utils will take care of it. + fused_qkv = mds_model["module"]["language_model"]["encoder"][ + f"layers.{layer_i}.self_attention.query_key_value.weight" + ] + state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = fused_qkv[ + 0:hidden_size + ] + state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = fused_qkv[ + hidden_size : hidden_size + kv_dim + ] + # state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size:hidden_size+kv_dim], kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = fused_qkv[ + hidden_size + kv_dim : hidden_size + 2 * kv_dim + ] + # state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim],kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = mds_model[ + "module" + ]["language_model"]["encoder"][f"layers.{layer_i}.self_attention.dense.weight"] + # MLP Layers + fused_mlp = mds_model["module"]["language_model"]["encoder"][ + f"layers.{layer_i}.mlp.dense_h_to_4h.weight" + ] + chunked_mlp = torch.chunk(fused_mlp, 2, dim=0) + state_dict[f"model.layers.{layer_i}.mlp.gate_proj.weight"] = chunked_mlp[0] + state_dict[f"model.layers.{layer_i}.mlp.up_proj.weight"] = chunked_mlp[1] + state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = mds_model[ + "module" + ]["language_model"]["encoder"][f"layers.{layer_i}.mlp.dense_4h_to_h.weight"] + # LayerNorm weights and RoPe + state_dict[f"model.layers.{layer_i}.input_layernorm.weight"] = mds_model[ + "module" + ]["language_model"]["encoder"][f"layers.{layer_i}.input_layernorm.weight"] + state_dict[ + f"model.layers.{layer_i}.post_attention_layernorm.weight" + ] = mds_model["module"]["language_model"]["encoder"][ + f"layers.{layer_i}.post_attention_layernorm.weight" + ] + + state_dict[f"model.layers.{layer_i}.self_attn.rotary_emb.inv_freq"] = inv_freq + + # Get the non-encoder layer weights. + state_dict["model.embed_tokens.weight"] = mds_model["module"]["language_model"][ + "embedding" + ]["word_embeddings"]["weight"] + state_dict["model.norm.weight"] = mds_model["module"]["language_model"]["encoder"][ + "final_layernorm.weight" + ] + state_dict["lm_head.weight"] = mds_model["module"]["language_model"][ + "output_layer" + ]["weight"] + # Save the model in the hf output path. + torch.save(state_dict, os.path.join(args.output_dir, "pytorch_model.bin")) From 4f9f1f63775daaa4c29de780a8319cd152ed0cea Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Sat, 24 Aug 2024 22:53:40 +0800 Subject: [PATCH 301/590] fix nan issue when running megatron-deepspeed (#434) --- megatron/training.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 6ba26f39440..e7d7aed8094 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -672,7 +672,6 @@ def train_step(forward_step_func, data_iterator, timers = get_timers() if args.deepspeed and args.ds_pipeline_enabled: - skipped_iter = 0 num_zeros_in_grad = 0 assert isinstance(model[0], deepspeed.PipelineEngine) loss = model[0].train_batch(data_iter=data_iterator) @@ -682,6 +681,8 @@ def train_step(forward_step_func, data_iterator, if additional_losses is not None: loss_dict.update(additional_losses) grad_norm = model[0].get_global_grad_norm() + update_successful = model[0].was_step_applied() + skipped_iter = 0 if update_successful else 1 return loss_dict, skipped_iter, grad_norm, num_zeros_in_grad # Set grad to zero. @@ -760,7 +761,7 @@ def train_step(forward_step_func, data_iterator, # Update learning rate. if args.deepspeed: - skipped_iter = 0 + skipped_iter = 0 if update_successful else 1 grad_norm = None num_zeros_in_grad = None From 8e9d97337a8533aef7b2f3f72dbd4248fdd1d4d2 Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Mon, 26 Aug 2024 23:49:19 +0800 Subject: [PATCH 302/590] enable empty cache on XPU device (#438) --- megatron/training.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index e7d7aed8094..79f39ccc2ef 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -725,7 +725,7 @@ def train_step(forward_step_func, data_iterator, # Empty unused memory. if args.empty_unused_memory_level >= 1: - torch.cuda.empty_cache() + get_accelerator().empty_cache() # Reduce gradients. if not args.deepspeed: @@ -782,7 +782,7 @@ def train_step(forward_step_func, data_iterator, # Empty unused memory. if args.empty_unused_memory_level >= 2: - torch.cuda.empty_cache() + get_accelerator().empty_cache() if mpu.is_pipeline_last_stage(ignore_virtual=True): # Average loss across microbatches. @@ -1438,7 +1438,7 @@ def evaluate(forward_step_func, # Empty unused memory if args.empty_unused_memory_level >= 1: - torch.cuda.empty_cache() + get_accelerator().empty_cache() if mpu.is_pipeline_last_stage(ignore_virtual=True): # Reduce across processes. From 543543acae44d15f4d3f83936a34c1dbd1e32a7b Mon Sep 17 00:00:00 2001 From: billishyahao Date: Wed, 28 Aug 2024 01:21:26 +0800 Subject: [PATCH 303/590] [wandb] disable wandb more gracefully (#422) Co-authored-by: Logan Adams <114770087+loadams@users.noreply.github.com> --- megatron/global_vars.py | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/megatron/global_vars.py b/megatron/global_vars.py index ccd0a4e21ec..ec6abdbda8e 100644 --- a/megatron/global_vars.py +++ b/megatron/global_vars.py @@ -165,15 +165,13 @@ def _set_wandb_writer(args): global _GLOBAL_WANDB_WRITER _ensure_var_is_not_initialized(_GLOBAL_WANDB_WRITER, 'wandb writer') - getattr(args, 'wandb_project', '') - getattr(args, 'wandb_exp_name', '') if args.rank == (args.world_size - 1): - if args.wandb_project == '' or \ - args.wandb_exp_name == '': + if getattr(args, 'wandb_project', '') == '' and \ + getattr(args, 'wandb_exp_name', '') == '': print('WARNING: WANDB writing requested but no legit wandb ' 'project or experiment name provided, ' - 'therefore WANDB logs will be written ' + 'therefore no WANDB logs will be written ' 'according to random generated project or experiment name.', flush=True) return From 1280f59c1a65e50d4e174e4195e14f173301a497 Mon Sep 17 00:00:00 2001 From: billishyahao Date: Wed, 28 Aug 2024 01:22:20 +0800 Subject: [PATCH 304/590] [Bug] Fix crash when logging optimizer state to tb (#417) --- megatron/training.py | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/megatron/training.py b/megatron/training.py index 79f39ccc2ef..0aeaabeba54 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1032,6 +1032,12 @@ def training_log(loss_dict, total_loss_dict, learning_rate, iteration, if args.log_optimizer_states_to_tensorboard and optimizer is not None: opt_stats = [0.0] * 8 opt_stats_2 = [0.0] * 4 + + #TODO(billishyahao): Remove me after bf16_optimizer promotes its state. + if not hasattr(optimizer, "state"): + assert hasattr(optimizer, "optimizer"), f"Optimizer must have optimizer property." + optimizer.state = optimizer.optimizer.state + for _, group in enumerate(optimizer.param_groups): for _, param in enumerate(group['params']): opt_stats[0] += (torch.norm(optimizer.state[param]['exp_avg_sq']).item())**2 From 5001600600d5ccc2a44cc0246fa4b0432dc5784b Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 30 Aug 2024 03:11:03 -0500 Subject: [PATCH 305/590] fixed data loader issue for TP>1 PP>1 --- megatron/data/blendable_dataset.py | 5 ++++- megatron/data/gpt_dataset.py | 12 +++++++----- 2 files changed, 11 insertions(+), 6 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index a4872613f96..ba2e00b1efe 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -96,7 +96,10 @@ def _build_indices(): print_rank_0("Data index creation unsuccessful, exiting.") exit() ''' - MPI.COMM_WORLD.Barrier() + torch.distributed.barrier(group=mpu.get_data_parallel_group()) + torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.barrier(group=mpu.get_data_parallel_group()) + start_time = time.time() print_rank_0(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 686ccecd70d..0cf97356a41 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -143,7 +143,7 @@ def __getitem__(self, idx): print_rank_0(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] - + train_datasets = [] valid_datasets = [] test_datasets = [] @@ -152,7 +152,7 @@ def __getitem__(self, idx): @dlp.log def build_corpus_datasets(dataset_type='train'): start_time = time.time() - print_rank_0(" >>> Building {dataset_type} corpus datasets ...") + print_rank_0(f" >>> Building {dataset_type} corpus datasets ...") datasets = [] corpus_builders = {} corpus_weights = {} @@ -163,7 +163,7 @@ def build_corpus_datasets(dataset_type='train'): datasets_train_valid_test_num_samples[i], seq_length, seed, skip_warmup, return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] - for i in range(torch.distributed.get_rank(), len(weights), torch.distributed.get_world_size()): + for i in range(torch.distributed.get_rank()//mpu.get_tensor_model_parallel_world_size(), len(weights), torch.distributed.get_world_size()//mpu.get_tensor_model_parallel_world_size()): dataset_builders[i].Build() print_rank_0(f" >>> Finished building individual datasets in {time.time() - start_time} seconds") start_concating_time = time.time() @@ -195,8 +195,10 @@ def build_corpus_datasets(dataset_type='train'): # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them - print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") - MPI.COMM_WORLD.Barrier() + print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") + torch.distributed.barrier(group=mpu.get_data_parallel_group()) + torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.barrier(group=mpu.get_data_parallel_group()) print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") print_rank_0(f" >>> Starting to build BlendableDataset") # Blend. From 38b2505b08ef0a92a766fb6622bf0f5a2514002a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 30 Aug 2024 10:09:38 -0500 Subject: [PATCH 306/590] Update `ALCF/data-lists/aurora/*.txt` --- ALCF/data-lists/aurora/algebraic.txt | 32 +- ALCF/data-lists/aurora/arxiv.txt | 200 +- ALCF/data-lists/aurora/books.txt | 6 +- ALCF/data-lists/aurora/c4.txt | 342 +- ALCF/data-lists/aurora/cc.txt | 2214 ++++---- ALCF/data-lists/aurora/dolma.txt | 4838 ++++++++--------- .../aurora/dolma_v1_7_file_list.txt | 4838 ++++++++--------- ALCF/data-lists/aurora/falcon.txt | 1000 ++-- ALCF/data-lists/aurora/megawiki.txt | 524 +- .../data-lists/aurora/open-web-math-train.txt | 26 +- ALCF/data-lists/aurora/pes2o.txt | 52 +- ALCF/data-lists/aurora/reddit.txt | 156 +- ALCF/data-lists/aurora/stack.txt | 52 +- ALCF/data-lists/aurora/starcoder.txt | 98 +- ALCF/data-lists/aurora/tulu.txt | 132 +- ALCF/data-lists/aurora/wiki.txt | 4 +- 16 files changed, 7257 insertions(+), 7257 deletions(-) diff --git a/ALCF/data-lists/aurora/algebraic.txt b/ALCF/data-lists/aurora/algebraic.txt index dc676b1e839..d3685cb42be 100644 --- a/ALCF/data-lists/aurora/algebraic.txt +++ b/ALCF/data-lists/aurora/algebraic.txt @@ -1,16 +1,16 @@ -0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train -0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train -0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train -0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train -0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train -0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train -0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train -0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train -0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train -0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train -0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train -0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train -0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train -0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train -0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train -0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train +0.0018520780893211373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train diff --git a/ALCF/data-lists/aurora/arxiv.txt b/ALCF/data-lists/aurora/arxiv.txt index c6214709e8f..c18c2befd28 100644 --- a/ALCF/data-lists/aurora/arxiv.txt +++ b/ALCF/data-lists/aurora/arxiv.txt @@ -1,100 +1,100 @@ -0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv -0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv -0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv -0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv -0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv -0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv -0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv -0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv -0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv -0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv -0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv -0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv -0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv -0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv -0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv -0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv -0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv -0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv -0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv -0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv -0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv -0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv -0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv -0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv -0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv -0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv -0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv -0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv -0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv -0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv -0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv -0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv -0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv -0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv -0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv -0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv -0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv -0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv -0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv -0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv -0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv -0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv -0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv -0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv -0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv -0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv -0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv -0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv -0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv -0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv -0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv -0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv -0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv -0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv -0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv -0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv -0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv -0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv -0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv -0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv -0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv -0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv -0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv -0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv -0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv -0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv -0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv -0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv -0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv -0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv -0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv -0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv -0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv -0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv -0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv -0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv -0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv -0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv -0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv -0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv -0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv -0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv -0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv -0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv -0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv -0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv -0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv -0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv -0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv -0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv -0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv -0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv -0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv -0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv -0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv -0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv -0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv -0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv -0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv -0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv +0.0002583902668716813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv diff --git a/ALCF/data-lists/aurora/books.txt b/ALCF/data-lists/aurora/books.txt index 7d69ed32fd0..6f37023596b 100644 --- a/ALCF/data-lists/aurora/books.txt +++ b/ALCF/data-lists/aurora/books.txt @@ -1,3 +1,3 @@ -0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books -0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books -0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0031025147279277244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books diff --git a/ALCF/data-lists/aurora/c4.txt b/ALCF/data-lists/aurora/c4.txt index 96c85cd3f4e..7ad92c60863 100644 --- a/ALCF/data-lists/aurora/c4.txt +++ b/ALCF/data-lists/aurora/c4.txt @@ -1,171 +1,171 @@ -0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 -0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 -0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 -0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 -0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 -0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 -0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 -0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 -0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 -0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 -0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 -0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 -0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 -0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 -0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 -0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 -0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 -0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 -0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 -0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 -0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 -0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 -0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 -0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 -0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 -0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 -0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 -0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 -0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 -0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 -0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 -0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 -0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 -0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 -0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 -0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 -0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 -0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 -0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 -0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 -0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 -0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 -0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 -0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 -0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 -0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 -0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 -0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 -0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 -0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 -0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 -0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 -0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 -0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 -0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 -0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 -0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 -0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 -0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 -0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 -0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 -0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 -0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 -0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 -0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 -0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 -0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 -0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 -0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 -0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 -0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 -0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 -0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 -0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 -0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 -0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 -0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 -0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 -0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 -0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 -0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 -0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 -0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 -0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 -0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 -0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 -0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 -0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 -0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 -0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 -0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 -0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 -0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 -0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 -0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 -0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 -0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 -0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 -0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 -0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 -0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 -0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 -0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 -0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 -0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 -0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 -0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 -0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 -0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 -0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 -0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 -0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 -0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 -0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 -0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 -0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 -0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 -0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 -0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 -0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 -0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 -0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 -0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 -0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 -0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 -0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 -0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 -0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 -0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 -0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 -0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 -0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 -0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 -0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 -0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 -0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 -0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 -0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 -0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 -0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 -0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 -0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 -0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 -0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 -0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 -0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 -0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 -0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 -0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 -0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 -0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 -0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 -0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 -0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 -0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 -0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 -0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 -0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 -0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 -0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 -0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 -0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 -0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 -0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 -0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 -0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 -0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 -0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 -0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 -0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 -7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 +0.0002406272620255565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 diff --git a/ALCF/data-lists/aurora/cc.txt b/ALCF/data-lists/aurora/cc.txt index f3161a32ef5..174bae9d6a8 100644 --- a/ALCF/data-lists/aurora/cc.txt +++ b/ALCF/data-lists/aurora/cc.txt @@ -1,1108 +1,1108 @@ -0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc -0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc -0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc -0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc -0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc -0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc -0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc -0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc -0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc -0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc -0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc -0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc -0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc -0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc -0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc -0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc -0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc -0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc -0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc -0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc -0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc -0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc -0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc -0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc -0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc -0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc -0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc -0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc -0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc -0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc -0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc -0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc -0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc -0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc -0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc -0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc -0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc -0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc -0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc -0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc -0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc -0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc -0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc -0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc -0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc -0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc -0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc -0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc -0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc -0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc -0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc -0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc -0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc -0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc -0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc -0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc -0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc -0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc -0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc -0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc -0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc -0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc -0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc -0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc -0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc -0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc -0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc -0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc -0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc -0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc -0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc -0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc -0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc -0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc -0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc -0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc -0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc -0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc -0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc -0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc -0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc -0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc -0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc -0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc -0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc -0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc -0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc -0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc -0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc -0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc -0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc -0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc -0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc -0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc -0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc -0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc -0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc -0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc -0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc -0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc -0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc -0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc -0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc -0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc -0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc -0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc -0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc -0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc -0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc -0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc -0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc -0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc -0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc -0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc -0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc -0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc -0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc -0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc -0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc -0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc -0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc -0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc -0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc -0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc -0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc -0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc -0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc -0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc -0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc -0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc -0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc -0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc -0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc -0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc -0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc -0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc -0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc -0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc -0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc -0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc -0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc -0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc -0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc -0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc -0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc -0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc -0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc -0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc -0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc -0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc -0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc -0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc -0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc -0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc -0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc -0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc -0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc -0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc -0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc -0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc -0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc -0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc -0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc -0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc -0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc -0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc -0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc -0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc -0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc -0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc -0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc -0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc -0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc -0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc -0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc -0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc -0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc -0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc -0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc -0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc -0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc -0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc -0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc -0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc -0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc -0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc -0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc -0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc -0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc -0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc -0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc -0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc -0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc -0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc -0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc -0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc -0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc -0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc -0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc -0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc -0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc -0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc -0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc -0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc -0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc -0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc -0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc -0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc -0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc -0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc -0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc -0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc -0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc -0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc -0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc -0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc -0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc -0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc -0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc -0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc -0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc -0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc -0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc -0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc -0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc -0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc -0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc -0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc -0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc -0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc -0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc -0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc -0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc -0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc -0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc -0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc -0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc -0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc -0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc -0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc -0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc -0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc -0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc -0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc -0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc -0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc -0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc -0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc -0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc -0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc -0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc -0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc -0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc -0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc -0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc -0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc -0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc -0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc -0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc -0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc -0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc -0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc -0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc -0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc -0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc -0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc -0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc -0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc -0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc -0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc -0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc -0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc -0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc -0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc -0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc -0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc -0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc -0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc -0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc -0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc -0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc -0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc -0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc -0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc -0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc -0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc -0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc -0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc -0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc -0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc -0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc -0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc -0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc -0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc -0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc -0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc -0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc -0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc -0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc -0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc -0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc -0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc -0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc -0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc -0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc -0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc -0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc -0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc -0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc -0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc -0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc -0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc -0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc -0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc -0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc -0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc -0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc -0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc -0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc -0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc -0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc -0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc -0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc -0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc -0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc -0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc -0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc -0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc -0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc -0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc -0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc -0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc -0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc -0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc -0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc -0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc -0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc -0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc -0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc -0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc -0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc -0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc -0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc -0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc -0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc -0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc -0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc -0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc -0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc -0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc -0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc -0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc -0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc -0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc -0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc -0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc -0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc -0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc -0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc -0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc -0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc -0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc -0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc -0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc -0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc -0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc -0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc -0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc -0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc -0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc -0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc -0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc -0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc -0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc -0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc -0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc -0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc -0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc -0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc -0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc -0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc -0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc -0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc -0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc -0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc -0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc -0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc -0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc -0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc -0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc -0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc -0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc -0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc -0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc -0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc -0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc -0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc -0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc -0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc -0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc -0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc -0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc -0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc -0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc -0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc -0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc -0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc -0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc -0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc -0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc -0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc -0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc -0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc -0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc -0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc -0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc -0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc -0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc -0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc -0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc -0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc -0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc -0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc -0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc -0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc -0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc -0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc -0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc -0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc -0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc -0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc -0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc -0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc -0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc -0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc -0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc -0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc -0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc -0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc -0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc -0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc -0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc -0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc -0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc -0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc -0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc -0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc -0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc -0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc -0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc -0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc -0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc -0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc -0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc -0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc -0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc -0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc -0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc -0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc -0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc -0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc -0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc -0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc -0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc -0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc -0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc -0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc -0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc -0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc -0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc -0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc -0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc -0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc -0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc -0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc -0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc -0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc -0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc -0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc -0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc -0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc -0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc -0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc -0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc -0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc -0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc -0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc -0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc -0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc -0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc -0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc -0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc -0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc -0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc -0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc -0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc -0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc -0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc -0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc -0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc -0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc -0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc -0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc -0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc -0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc -0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc -0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc -0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc -0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc -0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc -0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc -0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc -0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc -0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc -0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc -0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc -0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc -0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc -0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc -0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc -0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc -0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc -0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc -0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc -0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc -0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc -0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc -0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc -0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc -0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc -0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc -0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc -0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc -0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc -0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc -0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc -0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc -0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc -0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc -0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc -0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc -0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc -0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc -0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc -0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc -0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc -0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc -0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc -0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc -0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc -0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc -0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc -0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc -0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc -0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc -0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc -0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc -0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc -0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc -0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc -0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc -0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc -0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc -0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc -0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc -0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc -0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc -0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc -0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc -0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc -0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc -0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc -0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc -0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc -0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc -0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc -0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc -0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc -0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc -0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc -0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc -0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc -0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc -0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc -0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc -0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc -0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc -0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc -0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc -0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc -0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc -0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc -0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc -0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc -0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc -0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc -0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc -0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc -0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc -0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc -0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc -0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc -0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc -0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc -0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc -0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc -0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc -0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc -0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc -0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc -0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc -0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc -0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc -0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc -0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc -0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc -0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc -0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc -0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc -0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc -0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc -0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc -0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc -0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc -0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc -0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc -0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc -0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc -0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc -0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc -0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc -0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc -0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc -0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc -0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc -0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc -0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc -0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc -0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc -0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc -0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc -0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc -0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc -0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc -0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc -0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc -0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc -0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc -0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc -0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc -0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc -0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc -0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc -0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc -0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc -0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc -0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc -0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc -0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc -0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc -0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc -0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc -0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc -0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc -0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc -0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc -0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc -0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc -0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc -0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc -0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc -0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc -0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc -0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc -0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc -0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc -0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc -0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc -0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc -0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc -0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc -0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc -0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc -0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc -0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc -0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc -0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc -0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc -0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc -0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc -0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc -0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc -0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc -0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc -0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc -0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc -0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc -0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc -0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc -0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc -0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc -0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc -0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc -0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc -0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc -0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc -0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc -0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc -0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc -0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc -0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc -0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc -0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc -0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc -0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc -0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc -0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc -0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc -0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc -0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc -0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc -0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc -0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc -0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc -0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc -0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc -0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc -0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc -0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc -0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc -0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc -0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc -0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc -0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc -0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc -0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc -0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc -0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc -0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc -0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc -0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc -0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc -0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc -0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc -0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc -0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc -0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc -0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc -0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc -0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc -0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc -0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc -0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc -0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc -0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc -0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc -0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc -0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc -0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc -0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc -0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc -0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc -0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc -0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc -0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc -0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc -0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc -0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc -0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc -0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc -0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc -0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc -0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc -0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc -0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc -0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc -0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc -0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc -0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc -0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc -0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc -0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc -0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc -0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc -0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc -0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc -0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc -0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc -0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc -0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc -0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc -0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc -0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc -0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc -0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc -0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc -0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc -0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc -0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc -0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc -0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc -0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc -0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc -0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc -0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc -0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc -0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc -0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc -0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc -0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc -0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc -0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc -0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc -0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc -0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc -0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc -0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc -0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc -0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc -0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc -0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc -0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc -0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc -0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc -0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc -0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc -0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc -0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc -0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc -0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc -0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc -0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc -0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc -0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc -0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc -0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc -0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc -0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc -0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc -0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc -0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc -0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc -0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc -0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc -0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc -0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc -0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc -0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc -0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc -0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc -0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc -0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc -0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc -0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc -0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc -0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc -0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc -0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc -0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc -0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc -0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc -0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc -0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc -0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc -0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc -0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc -0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc -0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc -0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc -0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc -0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc -0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc -0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc -0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc -0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc -0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc -0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc -0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc -0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc -0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc -0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc -0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc -0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc -0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc -0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc -0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc -0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc -0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc -0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc -0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc -0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc -0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc -0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc -0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc -0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc -0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc -0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc -0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc -0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc -0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc -0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc -0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc -0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc -0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc -0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc -0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc -0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc -0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc -0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc -0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc -0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc -0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc -0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc -0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc -0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc -0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc -0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc -0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc -0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc -0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc -0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc -0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc -0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc -0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc -0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc -0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc -0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc -0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc -0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc -0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc -0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc -0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc -0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc -0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc -0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc -0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc -0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc -0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc -0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc -0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc -0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc -0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc -0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc -0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc -0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc -0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc -0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc -0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc -0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc -0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc -0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc -0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc -0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc -0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc -0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc -0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc -0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc -0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc -0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc -0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc -0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc -0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc -0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc -0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc -0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc -0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc -0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc -0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc -0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc -0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc -0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc -0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc -0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc -0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc -0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc -0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc -0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc -0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc -0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc -0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc -0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc -0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc -0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc -0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc -0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc -0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc -0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc -0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc -0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc -0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc -0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc -0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc -0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc -0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc -0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc -0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc -0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc -0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc -0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc -0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc -0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc -0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc -0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc -0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc -0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc -0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc -0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc -0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc -0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc -0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc -0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc -0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc -0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc -0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc -0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc -0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc -0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc -0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc -0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc -0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc -0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc -0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc -0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc -0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc -0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc -0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc -0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc -0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc -0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc -0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc -0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc -0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc -0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc -0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc -0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc -0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc -0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc -0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc -0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc -0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc -0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc -0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc -0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc -0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc -0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc -0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc -0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc -0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc -0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc -0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc -0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc -0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc -0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc -0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc -0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc -0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc -0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc -0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc -0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc -0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc -0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc -0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc -0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc -0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc -0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc -0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc -0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc -0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc -0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc -0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc -0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc -0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc -0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc -0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc -0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc -0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc -0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc -0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc -0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc -0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc -0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc -0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc -0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc -0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc -0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc -0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc -0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc -0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc -0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc -0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc -0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc -0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc -0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc -0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc -0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc -0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc -0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc -0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc -0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc -0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc -0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc -0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc -0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc -0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc -0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc -0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc -0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc +0.0003742481815405742 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc diff --git a/ALCF/data-lists/aurora/dolma.txt b/ALCF/data-lists/aurora/dolma.txt index 9449af64fc6..4aba801e009 100644 --- a/ALCF/data-lists/aurora/dolma.txt +++ b/ALCF/data-lists/aurora/dolma.txt @@ -1,2419 +1,2419 @@ -0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train -0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train -0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train -0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train -0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train -0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train -0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train -0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train -0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train -0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train -0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train -0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train -0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train -0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train -0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train -0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train -0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv -0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv -0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv -0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv -0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv -0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv -0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv -0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv -0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv -0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv -0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv -0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv -0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv -0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv -0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv -0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv -0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv -0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv -0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv -0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv -0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv -0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv -0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv -0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv -0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv -0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv -0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv -0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv -0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv -0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv -0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv -0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv -0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv -0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv -0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv -0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv -0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv -0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv -0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv -0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv -0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv -0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv -0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv -0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv -0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv -0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv -0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv -0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv -0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv -0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv -0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv -0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv -0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv -0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv -0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv -0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv -0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv -0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv -0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv -0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv -0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv -0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv -0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv -0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv -0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv -0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv -0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv -0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv -0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv -0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv -0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv -0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv -0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv -0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv -0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv -0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv -0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv -0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv -0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv -0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv -0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv -0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv -0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv -0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv -0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv -0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv -0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv -0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv -0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv -0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv -0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv -0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv -0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv -0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv -0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv -0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv -0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv -0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv -0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv -0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv -0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books -0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books -0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books -0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 -0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 -0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 -0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 -0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 -0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 -0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 -0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 -0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 -0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 -0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 -0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 -0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 -0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 -0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 -0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 -0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 -0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 -0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 -0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 -0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 -0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 -0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 -0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 -0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 -0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 -0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 -0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 -0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 -0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 -0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 -0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 -0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 -0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 -0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 -0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 -0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 -0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 -0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 -0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 -0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 -0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 -0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 -0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 -0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 -0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 -0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 -0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 -0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 -0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 -0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 -0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 -0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 -0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 -0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 -0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 -0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 -0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 -0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 -0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 -0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 -0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 -0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 -0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 -0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 -0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 -0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 -0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 -0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 -0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 -0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 -0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 -0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 -0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 -0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 -0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 -0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 -0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 -0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 -0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 -0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 -0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 -0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 -0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 -0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 -0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 -0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 -0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 -0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 -0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 -0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 -0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 -0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 -0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 -0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 -0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 -0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 -0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 -0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 -0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 -0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 -0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 -0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 -0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 -0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 -0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 -0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 -0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 -0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 -0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 -0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 -0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 -0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 -0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 -0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 -0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 -0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 -0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 -0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 -0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 -0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 -0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 -0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 -0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 -0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 -0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 -0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 -0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 -0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 -0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 -0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 -0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 -0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 -0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 -0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 -0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 -0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 -0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 -0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 -0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 -0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 -0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 -0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 -0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 -0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 -0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 -0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 -0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 -0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 -0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 -0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 -0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 -0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 -0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 -0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 -0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 -0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 -0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 -0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 -0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 -0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 -0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 -0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 -0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 -0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 -0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 -0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 -0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 -0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 -0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 -7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 -0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc -0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc -0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc -0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc -0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc -0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc -0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc -0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc -0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc -0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc -0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc -0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc -0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc -0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc -0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc -0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc -0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc -0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc -0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc -0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc -0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc -0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc -0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc -0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc -0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc -0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc -0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc -0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc -0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc -0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc -0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc -0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc -0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc -0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc -0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc -0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc -0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc -0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc -0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc -0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc -0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc -0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc -0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc -0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc -0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc -0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc -0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc -0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc -0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc -0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc -0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc -0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc -0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc -0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc -0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc -0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc -0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc -0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc -0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc -0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc -0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc -0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc -0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc -0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc -0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc -0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc -0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc -0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc -0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc -0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc -0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc -0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc -0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc -0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc -0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc -0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc -0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc -0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc -0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc -0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc -0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc -0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc -0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc -0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc -0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc -0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc -0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc -0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc -0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc -0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc -0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc -0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc -0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc -0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc -0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc -0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc -0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc -0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc -0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc -0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc -0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc -0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc -0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc -0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc -0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc -0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc -0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc -0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc -0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc -0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc -0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc -0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc -0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc -0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc -0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc -0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc -0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc -0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc -0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc -0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc -0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc -0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc -0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc -0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc -0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc -0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc -0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc -0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc -0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc -0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc -0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc -0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc -0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc -0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc -0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc -0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc -0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc -0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc -0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc -0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc -0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc -0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc -0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc -0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc -0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc -0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc -0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc -0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc -0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc -0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc -0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc -0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc -0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc -0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc -0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc -0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc -0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc -0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc -0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc -0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc -0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc -0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc -0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc -0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc -0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc -0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc -0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc -0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc -0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc -0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc -0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc -0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc -0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc -0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc -0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc -0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc -0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc -0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc -0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc -0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc -0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc -0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc -0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc -0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc -0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc -0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc -0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc -0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc -0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc -0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc -0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc -0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc -0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc -0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc -0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc -0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc -0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc -0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc -0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc -0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc -0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc -0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc -0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc -0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc -0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc -0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc -0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc -0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc -0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc -0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc -0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc -0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc -0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc -0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc -0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc -0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc -0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc -0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc -0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc -0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc -0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc -0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc -0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc -0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc -0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc -0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc -0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc -0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc -0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc -0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc -0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc -0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc -0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc -0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc -0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc -0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc -0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc -0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc -0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc -0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc -0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc -0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc -0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc -0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc -0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc -0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc -0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc -0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc -0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc -0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc -0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc -0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc -0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc -0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc -0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc -0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc -0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc -0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc -0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc -0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc -0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc -0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc -0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc -0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc -0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc -0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc -0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc -0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc -0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc -0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc -0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc -0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc -0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc -0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc -0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc -0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc -0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc -0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc -0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc -0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc -0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc -0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc -0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc -0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc -0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc -0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc -0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc -0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc -0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc -0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc -0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc -0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc -0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc -0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc -0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc -0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc -0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc -0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc -0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc -0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc -0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc -0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc -0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc -0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc -0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc -0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc -0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc -0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc -0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc -0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc -0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc -0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc -0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc -0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc -0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc -0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc -0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc -0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc -0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc -0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc -0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc -0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc -0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc -0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc -0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc -0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc -0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc -0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc -0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc -0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc -0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc -0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc -0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc -0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc -0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc -0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc -0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc -0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc -0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc -0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc -0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc -0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc -0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc -0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc -0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc -0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc -0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc -0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc -0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc -0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc -0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc -0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc -0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc -0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc -0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc -0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc -0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc -0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc -0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc -0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc -0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc -0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc -0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc -0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc -0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc -0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc -0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc -0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc -0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc -0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc -0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc -0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc -0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc -0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc -0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc -0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc -0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc -0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc -0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc -0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc -0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc -0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc -0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc -0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc -0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc -0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc -0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc -0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc -0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc -0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc -0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc -0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc -0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc -0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc -0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc -0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc -0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc -0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc -0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc -0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc -0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc -0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc -0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc -0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc -0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc -0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc -0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc -0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc -0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc -0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc -0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc -0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc -0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc -0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc -0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc -0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc -0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc -0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc -0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc -0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc -0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc -0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc -0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc -0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc -0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc -0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc -0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc -0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc -0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc -0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc -0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc -0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc -0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc -0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc -0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc -0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc -0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc -0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc -0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc -0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc -0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc -0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc -0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc -0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc -0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc -0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc -0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc -0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc -0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc -0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc -0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc -0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc -0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc -0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc -0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc -0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc -0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc -0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc -0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc -0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc -0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc -0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc -0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc -0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc -0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc -0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc -0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc -0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc -0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc -0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc -0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc -0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc -0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc -0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc -0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc -0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc -0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc -0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc -0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc -0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc -0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc -0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc -0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc -0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc -0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc -0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc -0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc -0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc -0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc -0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc -0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc -0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc -0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc -0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc -0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc -0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc -0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc -0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc -0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc -0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc -0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc -0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc -0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc -0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc -0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc -0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc -0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc -0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc -0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc -0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc -0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc -0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc -0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc -0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc -0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc -0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc -0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc -0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc -0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc -0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc -0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc -0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc -0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc -0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc -0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc -0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc -0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc -0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc -0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc -0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc -0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc -0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc -0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc -0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc -0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc -0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc -0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc -0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc -0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc -0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc -0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc -0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc -0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc -0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc -0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc -0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc -0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc -0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc -0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc -0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc -0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc -0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc -0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc -0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc -0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc -0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc -0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc -0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc -0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc -0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc -0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc -0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc -0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc -0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc -0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc -0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc -0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc -0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc -0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc -0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc -0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc -0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc -0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc -0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc -0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc -0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc -0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc -0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc -0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc -0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc -0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc -0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc -0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc -0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc -0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc -0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc -0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc -0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc -0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc -0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc -0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc -0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc -0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc -0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc -0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc -0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc -0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc -0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc -0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc -0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc -0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc -0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc -0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc -0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc -0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc -0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc -0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc -0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc -0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc -0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc -0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc -0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc -0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc -0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc -0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc -0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc -0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc -0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc -0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc -0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc -0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc -0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc -0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc -0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc -0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc -0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc -0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc -0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc -0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc -0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc -0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc -0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc -0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc -0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc -0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc -0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc -0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc -0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc -0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc -0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc -0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc -0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc -0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc -0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc -0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc -0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc -0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc -0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc -0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc -0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc -0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc -0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc -0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc -0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc -0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc -0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc -0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc -0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc -0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc -0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc -0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc -0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc -0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc -0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc -0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc -0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc -0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc -0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc -0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc -0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc -0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc -0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc -0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc -0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc -0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc -0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc -0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc -0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc -0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc -0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc -0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc -0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc -0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc -0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc -0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc -0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc -0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc -0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc -0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc -0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc -0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc -0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc -0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc -0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc -0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc -0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc -0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc -0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc -0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc -0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc -0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc -0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc -0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc -0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc -0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc -0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc -0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc -0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc -0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc -0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc -0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc -0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc -0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc -0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc -0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc -0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc -0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc -0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc -0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc -0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc -0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc -0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc -0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc -0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc -0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc -0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc -0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc -0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc -0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc -0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc -0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc -0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc -0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc -0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc -0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc -0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc -0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc -0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc -0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc -0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc -0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc -0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc -0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc -0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc -0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc -0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc -0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc -0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc -0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc -0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc -0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc -0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc -0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc -0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc -0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc -0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc -0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc -0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc -0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc -0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc -0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc -0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc -0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc -0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc -0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc -0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc -0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc -0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc -0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc -0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc -0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc -0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc -0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc -0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc -0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc -0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc -0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc -0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc -0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc -0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc -0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc -0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc -0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc -0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc -0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc -0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc -0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc -0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc -0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc -0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc -0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc -0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc -0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc -0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc -0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc -0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc -0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc -0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc -0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc -0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc -0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc -0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc -0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc -0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc -0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc -0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc -0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc -0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc -0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc -0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc -0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc -0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc -0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc -0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc -0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc -0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc -0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc -0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc -0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc -0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc -0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc -0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc -0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc -0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc -0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc -0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc -0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc -0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc -0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc -0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc -0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc -0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc -0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc -0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc -0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc -0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc -0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc -0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc -0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc -0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc -0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc -0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc -0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc -0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc -0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc -0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc -0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc -0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc -0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc -0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc -0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc -0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc -0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc -0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc -0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc -0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc -0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc -0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc -0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc -0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc -0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc -0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc -0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc -0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc -0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc -0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc -0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc -0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc -0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc -0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc -0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc -0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc -0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc -0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc -0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc -0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc -0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc -0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc -0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc -0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc -0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc -0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc -0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc -0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc -0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc -0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc -0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc -0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc -0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc -0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc -0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc -0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc -0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc -0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc -0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc -0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc -0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc -0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc -0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc -0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc -0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc -0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc -0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc -0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc -0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc -0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc -0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc -0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc -0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc -0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc -0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc -0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc -0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc -0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc -0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc -0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc -0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc -0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc -0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc -0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc -0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc -0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc -0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc -0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc -0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc -0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc -0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc -0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc -0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc -0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc -0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc -0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc -0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc -0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc -0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc -0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc -0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc -0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc -0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc -0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc -0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc -0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc -0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc -0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc -0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc -0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc -0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc -0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc -0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc -0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc -0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc -0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc -0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc -0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc -0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc -0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc -0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc -0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc -0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc -0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc -0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc -0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc -0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc -0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc -0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc -0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc -0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc -0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc -0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc -0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc -0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc -0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc -0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc -0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc -0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc -0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc -0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc -0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc -0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc -0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc -0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc -0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc -0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc -0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc -0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc -0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc -0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc -0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc -0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc -0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc -0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc -0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc -0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc -0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc -0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc -0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc -0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc -0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc -0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc -0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc -0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc -0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc -0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc -0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc -0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc -0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc -0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc -0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc -0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc -0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc -0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc -0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc -0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc -0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc -0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc -0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc -0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc -0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc -0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc -0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc -0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc -0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc -0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc -0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc -0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc -0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc -0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc -0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc -0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc -0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc -0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc -0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc -0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc -0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc -0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc -0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc -0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc -0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc -0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc -0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc -0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc -0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc -0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc -0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc -0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc -0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc -0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc -0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc -0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc -0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc -0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc -0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc -0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc -0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc -0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc -0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc -0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc -0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc -0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc -0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc -0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc -0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc -0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc -0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc -0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc -0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc -0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc -0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc -0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc -0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc -0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc -0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc -0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc -0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc -0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc -0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc -0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc -0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc -0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc -0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc -0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc -0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc -0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc -0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc -0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc -0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc -0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc -0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc -0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc -0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc -0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc -0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc -0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc -0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc -0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc -0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc -0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc -0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc -0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc -0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc -0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc -0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc -0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc -0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc -0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc -0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc -0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc -0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon -0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon -0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon -0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon -0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon -0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon -0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon -0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon -0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon -0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon -0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon -0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon -0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon -0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon -0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon -0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon -0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon -0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon -0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon -0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon -0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon -0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon -0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon -0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon -0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon -0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon -0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon -0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon -0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon -0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon -0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon -0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon -0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon -0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon -0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon -0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon -0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon -0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon -0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon -0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon -0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon -0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon -0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon -0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon -0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon -0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon -0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon -0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon -0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon -0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon -0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon -0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon -0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon -0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon -0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon -0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon -0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon -0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon -0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon -0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon -0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon -0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon -0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon -0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon -0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon -0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon -0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon -0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon -0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon -0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon -0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon -0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon -0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon -0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon -0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon -0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon -0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon -0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon -0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon -0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon -0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon -0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon -0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon -0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon -0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon -0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon -0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon -0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon -0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon -0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon -0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon -0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon -0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon -0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon -0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon -0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon -0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon -0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon -0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon -0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon -0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon -0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon -0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon -0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon -0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon -0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon -0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon -0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon -0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon -0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon -0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon -0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon -0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon -0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon -0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon -0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon -0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon -0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon -0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon -0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon -0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon -0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon -0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon -0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon -0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon -0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon -0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon -0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon -0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon -0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon -0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon -0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon -0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon -0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon -0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon -0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon -0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon -0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon -0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon -0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon -0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon -0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon -0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon -0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon -0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon -0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon -0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon -0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon -0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon -0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon -0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon -0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon -0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon -0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon -0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon -0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon -0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon -0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon -0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon -0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon -0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon -0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon -0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon -0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon -0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon -0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon -0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon -0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon -0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon -0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon -0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon -0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon -0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon -0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon -0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon -0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon -0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon -0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon -0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon -0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon -0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon -0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon -0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon -0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon -0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon -0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon -0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon -0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon -0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon -0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon -0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon -0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon -0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon -0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon -0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon -0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon -0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon -0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon -0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon -0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon -0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon -0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon -0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon -0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon -0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon -0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon -0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon -0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon -0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon -0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon -0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon -0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon -0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon -0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon -0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon -0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon -0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon -0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon -0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon -0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon -0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon -0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon -0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon -0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon -0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon -0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon -0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon -0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon -0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon -0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon -0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon -0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon -0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon -0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon -0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon -0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon -0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon -0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon -0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon -0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon -0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon -0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon -0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon -0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon -0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon -0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon -0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon -0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon -0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon -0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon -0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon -0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon -0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon -0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon -0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon -0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon -0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon -0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon -0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon -0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon -0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon -0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon -0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon -0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon -0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon -0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon -0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon -0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon -0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon -0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon -0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon -0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon -0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon -0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon -0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon -0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon -0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon -0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon -0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon -0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon -0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon -0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon -0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon -0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon -0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon -0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon -0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon -0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon -0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon -0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon -0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon -0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon -0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon -0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon -0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon -0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon -0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon -0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon -0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon -0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon -0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon -0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon -0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon -0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon -0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon -0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon -0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon -0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon -0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon -0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon -0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon -0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon -0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon -0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon -0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon -0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon -0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon -0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon -0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon -0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon -0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon -0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon -0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon -0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon -0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon -0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon -0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon -0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon -0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon -0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon -0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon -0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon -0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon -0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon -0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon -0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon -0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon -0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon -0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon -0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon -0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon -0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon -0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon -0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon -0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon -0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon -0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon -0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon -0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon -0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon -0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon -0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon -0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon -0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon -0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon -0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon -0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon -0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon -0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon -0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon -0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon -0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon -0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon -0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon -0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon -0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon -0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon -0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon -0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon -0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon -0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon -0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon -0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon -0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon -0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon -0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon -0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon -0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon -0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon -0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon -0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon -0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon -0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon -0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon -0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon -0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon -0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon -0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon -0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon -0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon -0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon -0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon -0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon -0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon -0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon -0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon -0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon -0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon -0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon -0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon -0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon -0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon -0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon -0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon -0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon -0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon -0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon -0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon -0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon -0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon -0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon -0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon -0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon -0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon -0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon -0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon -0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon -0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon -0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon -0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon -0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon -0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon -0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon -0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon -0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon -0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon -0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon -0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon -0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon -0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon -0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon -0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon -0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon -0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon -0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon -0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon -0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon -0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon -0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon -0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon -0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon -0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon -0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon -0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon -0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon -0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon -0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon -0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon -0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon -0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon -0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon -0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon -0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon -0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon -0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon -0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon -0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon -0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon -0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon -0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon -0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon -0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon -0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon -0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon -0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon -0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon -0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon -0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon -0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon -0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon -0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon -0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon -0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon -0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon -0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon -0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon -0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon -0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon -0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon -0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon -0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon -0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon -0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon -0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon -0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon -0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon -0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon -0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon -0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon -0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon -0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon -0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon -0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon -0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon -0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon -0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon -0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon -0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon -0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon -0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon -6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika -2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika -5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika -2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika -2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika -2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika -5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika -1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika -1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika -1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika -1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika -1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika -1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika -1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika -1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika -1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika -1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika -1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika -1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika -1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika -1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika -1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika -1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika -1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika -1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika -1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika -1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika -1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika -1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika -1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika -1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika -1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika -1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika -1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika -1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika -1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika -1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika -1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika -1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika -1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika -1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika -1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika -1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika -1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika -1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika -1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika -1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika -1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika -1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika -1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika -1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika -1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika -5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika -5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika -5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika -5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika -5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika -5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika -5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika -4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika -4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika -5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika -4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika -5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika -6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika -5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika -4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika -5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika -5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika -4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika -5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika -5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika -5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika -5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika -5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika -6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika -5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika -5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika -5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika -5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika -5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika -5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika -5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika -4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika -5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika -5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika -4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika -5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika -5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika -5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika -4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika -5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika -4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika -5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika -5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika -6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika -5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika -5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika -5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika -5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika -5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika -5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika -4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika -5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika -5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika -5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika -5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika -5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika -5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika -5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika -5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika -5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika -5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika -5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika -2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika -1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika -3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika -3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika -2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika -2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika -2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika -2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika -2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika -2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika -2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika -2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika -2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika -2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika -2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika -2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika -2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika -1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika -1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika -2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika -2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika -9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika -2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika -1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika -2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika -2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika -4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika -2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika -2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika -1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika -1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika -1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika -1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika -1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika -1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika -1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika -1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika -1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika -1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika -1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika -1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika -1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika -1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika -1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika -1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika -1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika -1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika -1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika -1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika -1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika -1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika -1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika -1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika -8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika -8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika -9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika -9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika -7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika -8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika -9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika -8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika -8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika -8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika -7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika -9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika -8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika -1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika -3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika -2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika -1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika -1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika -1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika -2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika -2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika -8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika -1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika -1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika -1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika -1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika -1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika -1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika -1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika -2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika -1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika -2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika -5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika -1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika -1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika -1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika -1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika -1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika -1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika -1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika -1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika -1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika -1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika -1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika -1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika -1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika -1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika -1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika -1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika -1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika -1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika -9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika -1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika -1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika -1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika -1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika -1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika -1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika -1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika -1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika -1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika -1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika -4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika -1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika -1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika -2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika -2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika -2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika -2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika -1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika -1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika -3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika -6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika -1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika -1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika -1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika -2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika -3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika -1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika -1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika -1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika -1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika -1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika -1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika -1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika -1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika -1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika -1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika -4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika -3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika -3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika -3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika -3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika -3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika -7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika -0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train -0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train -0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train -0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train -0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train -6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train -0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train -0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train -0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train -0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train -0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train -0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train -0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train -0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o -0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o -0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o -0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o -0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o -0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o -0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o -0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o -0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o -0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o -0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o -0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o -0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o -0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o -0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o -0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o -0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o -0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o -0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o -0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o -0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o -0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o -0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o -0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o -0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o -0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o -0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit -0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit -0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit -0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit -0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit -0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit -0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit -0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit -0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit -0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit -0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit -0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit -0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit -0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit -0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit -0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit -0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit -0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit -0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit -0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit -0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit -0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit -0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit -0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit -0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit -0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit -0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit -0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit -0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit -0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit -0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit -0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit -0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit -0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit -0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit -0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit -0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit -0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit -0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit -0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit -0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit -0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit -0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit -0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit -0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit -0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit -0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit -0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit -0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit -0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit -0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit -0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit -0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit -0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit -0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit -0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit -0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit -0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit -0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit -0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit -0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit -0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit -0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit -0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit -0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit -0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit -0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit -0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit -0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit -0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit -0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit -0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit -0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit -0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit -0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit -0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit -0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit -0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit -0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange -0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange -0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange -0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange -0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange -0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange -0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange -0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange -0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange -0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange -0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange -0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange -0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange -0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange -0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange -0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange -0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange -0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange -0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange -0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange -0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange -0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange -0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange -0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange -0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange -0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange -0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder -0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder -0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder -0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder -0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder -0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder -0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder -0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder -0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder -0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder -0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder -0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder -0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder -0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder -0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder -0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder -0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder -0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder -0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder -0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder -0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder -0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder -0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder -0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder -0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder -0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder -0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder -0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder -0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder -0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder -0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder -0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder -0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder -0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder -0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder -0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder -0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder -0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder -0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder -0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder -0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder -0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder -0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder -0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder -0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder -0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder -0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder -0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder -0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder -0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu -0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu -0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu -0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu -0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu -0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu -0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu -0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu -0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu -0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu -0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu -0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu -0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu -0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu -0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu -0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu -0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu -0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu -0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu -0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu -0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu -0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu -0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu -0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu -0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu -0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu -0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu -0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu -0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu -0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu -0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu -0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu -0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu -0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu -0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu -0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu -0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu -0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu -0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu -0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu -0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu -0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu -0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu -0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu -0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu -0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu -0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu -0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu -0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu -0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu -0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu -0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu -0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu -0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu -0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu -0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu -0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu -0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu -0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu -0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu -0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu -0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu -0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu -0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu -0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu -3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu -0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki -0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki +0.0018520780893211373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document algebraic-stack-train +0.0017591050606817512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document algebraic-stack-train +0.001459052794333798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document algebraic-stack-train +0.0007405667281569194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document algebraic-stack-train +0.00019420030110896795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document algebraic-stack-train +0.0009008668715801845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document algebraic-stack-train +0.00015115827957143057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document algebraic-stack-train +0.0014552844319220648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document algebraic-stack-train +0.0012469861325685161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document algebraic-stack-train +0.00136412011372413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document algebraic-stack-train +0.0007064279699221103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document algebraic-stack-train +0.0008472240000687427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document algebraic-stack-train +0.0001984375713341955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document algebraic-stack-train +0.0005472773881697123 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document algebraic-stack-train +0.001815779629850992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document algebraic-stack-train +0.0018313600689757324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document algebraic-stack-train +0.0002583902668716813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document arxiv +0.0002646575141232155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document arxiv +0.0003165521247456758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document arxiv +0.0002920706460176214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document arxiv +0.00028396813182810215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document arxiv +0.00030445161883108107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document arxiv +0.00031628781276576474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document arxiv +0.0003083776568189157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document arxiv +0.0003176359471472902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document arxiv +0.0002536009369131698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document arxiv +0.0003067491424681363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document arxiv +0.0002597217257557784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document arxiv +0.0003788556450109768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document arxiv +0.0002796563272052598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document arxiv +0.00033573826524290287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document arxiv +0.00030523658022800287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document arxiv +0.00032211552192240096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document arxiv +0.0003329295675164247 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document arxiv +0.0003101982186639862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document arxiv +0.00032361798234223355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document arxiv +0.0003495541581652915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document arxiv +0.0002821637448858042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document arxiv +0.00030399523537629673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document arxiv +0.0002955658968247219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document arxiv +0.00028942158502924254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document arxiv +0.00028769546171490733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document arxiv +0.0002938111057234182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document arxiv +0.0002711150403010948 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document arxiv +0.00031130095874747565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document arxiv +0.0003002996118160777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document arxiv +0.0003732757901604459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document arxiv +0.00026784205751795894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document arxiv +0.0002799626521661984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document arxiv +0.00034334276069078164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document arxiv +0.0003582469803674965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document arxiv +0.00031094844818418623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document arxiv +0.0002766228384977191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document arxiv +0.00030297116159471485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document arxiv +0.00027033888377464685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document arxiv +0.00030090862368377933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document arxiv +0.00028543875802490955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document arxiv +0.00027559768459074204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document arxiv +0.0003182185533962886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document arxiv +0.0003311392971435837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document arxiv +0.00028751652060804325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document arxiv +0.000303466863212589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document arxiv +0.00033400462801277524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document arxiv +0.0002589234031777426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document arxiv +0.0002913508598466723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document arxiv +0.0002670572450004856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document arxiv +0.00032027399105647656 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document arxiv +0.00032188376258379377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document arxiv +0.0003161585784100882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document arxiv +0.0003184249182974135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document arxiv +0.00030381336664000807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document arxiv +0.0003190437442184283 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document arxiv +0.0002537961798200545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document arxiv +0.0003017817117223326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document arxiv +0.00028685268513240224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document arxiv +0.00031265179094451165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document arxiv +0.00034708319096986816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document arxiv +0.00026650837943080664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document arxiv +0.00034588832248507335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document arxiv +0.0002416982248399037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document arxiv +0.0003089296918222243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document arxiv +0.00029137184185700827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document arxiv +0.00026464226846800774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document arxiv +0.00030545397919456627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document arxiv +0.0003206778460448875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document arxiv +0.00030968971641110967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document arxiv +0.00023325653928600864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document arxiv +0.00030526899198338555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document arxiv +0.00035376719076633584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document arxiv +0.000290224385981026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document arxiv +0.000294650083382008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document arxiv +0.00028768858128616436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document arxiv +0.00030856965235527843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document arxiv +0.00030579942447879054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document arxiv +0.0002863101084704357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document arxiv +0.0002870032092492213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document arxiv +0.000264182727569885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document arxiv +0.0002974012367036449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document arxiv +0.00032238412143059203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document arxiv +0.00031683716893819036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document arxiv +0.00031157434937617524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document arxiv +0.0003411742735695989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document arxiv +0.00026778444816570715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document arxiv +0.0003037045797275201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document arxiv +0.00027746114370081314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document arxiv +0.00027148285946862043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document arxiv +0.00028042950114678207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document arxiv +0.0003235607816590721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document arxiv +0.0003086692227306295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document arxiv +0.00033990349455148105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document arxiv +0.00030945053208470265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document arxiv +0.00027309074552265303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document arxiv +0.00028737393506316194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document arxiv +0.0003098868328009879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document arxiv +0.0002614229162588409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document arxiv +0.0002884388407820923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document arxiv +0.0031025147279277244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books +0.003102019887362634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books +0.0009996745994661548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0002406272620255565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document c4 +0.0002404825539493424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document c4 +0.00024062296575435581 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document c4 +0.00024069315766818953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document c4 +0.00024055829162263452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document c4 +0.00024062053397343032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document c4 +0.0002410715545206964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document c4 +0.00024024881846087368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document c4 +0.0002407074700790688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document c4 +0.00024072141428809043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document c4 +0.00024027710230872736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document c4 +0.0002409111299205489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document c4 +0.00024081954058275009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document c4 +0.00024086076794990912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document c4 +0.00024098672620832446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document c4 +0.00024068622303333862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document c4 +0.00024140627024291824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document c4 +0.0002414512033594384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document c4 +0.00024028742594941463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document c4 +0.00024018036089269645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document c4 +0.0002398347365034979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document c4 +0.00024006780153485276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document c4 +0.00024015620270419213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document c4 +0.0002408848259695227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document c4 +0.0002408023185278831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document c4 +0.00024021196580140326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document c4 +0.00024077677271297493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document c4 +0.00024087392454668027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document c4 +0.0002408071293824126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document c4 +0.00024042223828845715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document c4 +0.0002411484752360495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document c4 +0.00023605263746465907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document c4 +0.00023471222158326908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document c4 +0.00023432138580287644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document c4 +0.00023407385623382327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document c4 +0.00023487504174367091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document c4 +0.0002341843704976313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document c4 +0.00023421993170282486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document c4 +0.00023445057969132037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document c4 +0.0002337681680073047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document c4 +0.000234627964808109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document c4 +0.0002338942211888584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document c4 +0.00023403849286843386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document c4 +0.00023405641310796305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document c4 +0.00023349169562397965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document c4 +0.00023381157386048856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document c4 +0.00023388742993790587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document c4 +0.00023363103829469813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document c4 +0.00023421141834630477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document c4 +0.00023420564352232565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document c4 +0.00023367463699173143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document c4 +0.00023344969163567033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document c4 +0.00023372196941547188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document c4 +0.00023399207645297834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document c4 +0.00023357915605505856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document c4 +0.00023337585642190864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document c4 +0.00023385005470157914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document c4 +0.00023301533534493465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document c4 +0.00023377864302541782 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document c4 +0.00023323745848621437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document c4 +0.0002330594611151835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document c4 +0.0002334149675026783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document c4 +0.00023198945902291534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document c4 +0.00023023784834634142 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document c4 +0.00022985623060187217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document c4 +0.0002292605284569516 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document c4 +0.00022926593333048894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document c4 +0.00022922766406807777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document c4 +0.00022898153911167426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document c4 +0.0002292473111593315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document c4 +0.000228804579400424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document c4 +0.00022865485613513526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document c4 +0.00022937426835887895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document c4 +0.00022917388311587372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document c4 +0.0002291660582019043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document c4 +0.00022907895248360543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document c4 +0.0002294617879920205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document c4 +0.0002290452150516566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document c4 +0.00022943405619715553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document c4 +0.0002296271421006204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document c4 +0.00022854791372910372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document c4 +0.00022923123467686557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document c4 +0.00022852404355738494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document c4 +0.00022847798660086642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document c4 +0.0002289604586810316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document c4 +0.00022835479834950643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document c4 +0.0002289149402884243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document c4 +0.00022806655474763446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document c4 +0.00022826296420992974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document c4 +0.00022906829636213627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document c4 +0.0002287628414466998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document c4 +0.0002282673911253445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document c4 +0.00022869309841939134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document c4 +0.0002281540116815451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document c4 +0.0002259755756162738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document c4 +0.00022562331285233504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document c4 +0.0002259061146106053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document c4 +0.00022567670836663787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document c4 +0.00022573165387587061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document c4 +0.00022508514961670572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document c4 +0.00022564642513773356 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document c4 +0.00022563088621998788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document c4 +0.0002250438755373707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document c4 +0.00022524465346241134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document c4 +0.00022531737657666812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document c4 +0.00022444687519363458 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document c4 +0.00022460397498596298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document c4 +0.00022454218976501763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document c4 +0.00022447528843671366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document c4 +0.00022501666332178926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document c4 +0.00022453752304377972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document c4 +0.00022484451871163002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document c4 +0.00022465678847154914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document c4 +0.00022453180917044732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document c4 +0.0002247278486823009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document c4 +0.00022465794828242097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document c4 +0.00022431000701925386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document c4 +0.00022476020248460963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document c4 +0.00022467531771795015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document c4 +0.0002236391309945234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document c4 +0.00022458764920536007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document c4 +0.00022430877426744415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document c4 +0.0002247047786127192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document c4 +0.0002245298090400035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document c4 +0.0002245648831396188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document c4 +0.00022292894729820784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document c4 +0.00022236668082957533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document c4 +0.0002217622659895442 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document c4 +0.00022252452726732609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document c4 +0.00022135333211363678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document c4 +0.0002214571757787971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document c4 +0.0002217188139237798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document c4 +0.00022144214894640303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document c4 +0.00022100172806631854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document c4 +0.00022156392409199052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document c4 +0.00022134830143710272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document c4 +0.00022158598922529453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document c4 +0.00022142932483041377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document c4 +0.00022120980907786554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document c4 +0.00022117917738112441 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document c4 +0.00022077089397851235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document c4 +0.00022093265074996711 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document c4 +0.00022091299741377004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document c4 +0.0002205849150703338 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document c4 +0.0002210648204787979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document c4 +0.0002214235747364102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document c4 +0.00022083907302221787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document c4 +0.0002206334237915964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document c4 +0.00022065193929912214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document c4 +0.00022079775597767288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document c4 +0.00022091492909963518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document c4 +0.00022095009987097293 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document c4 +0.0002208150577180165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document c4 +0.00022085759102772088 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document c4 +0.00022073789170129016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document c4 +0.00022049322781182384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document c4 +0.00022083270617761285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document c4 +0.00021982452827473632 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document c4 +0.00021899870446514259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document c4 +0.00021890358773356361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document c4 +0.00021875556609042841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document c4 +0.00021861195987201226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document c4 +0.00021856782186167455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document c4 +0.00021912837771543515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document c4 +0.00021900213768517756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document c4 +0.00021871675851390374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document c4 +0.0002180537056545586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document c4 +0.0002188196714327129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document c4 +0.00021851362624523464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document c4 +0.0002183236795498736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document c4 +7.291153618675672e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document c4 +0.0003742481815405742 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document cc +0.00038204855962733055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document cc +0.00038821818392663593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document cc +0.00038723332988783727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document cc +0.00038916141142149904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document cc +0.00038049542523949033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document cc +0.0003854755539534284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document cc +0.00024202756466512517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document cc +0.0003915405155008087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document cc +0.0003927382151931033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document cc +0.0003839151202260479 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document cc +0.00040006817468967907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document cc +0.00040318965964443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document cc +0.0003831013019452741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document cc +0.00039166638383204036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document cc +0.00039962784023961004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document cc +0.00039536707853602614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document cc +0.0004204304698247758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document cc +0.00041538899178693555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document cc +0.00039186953333675306 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document cc +0.00038945837196504305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document cc +0.0003919951238929062 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document cc +0.00044377065718528966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document cc +0.0004407759068603017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document cc +0.0002487811895843715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document cc +0.00039349432045556636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document cc +0.00041223198559462343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document cc +0.0004036573014830213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document cc +0.0003825982215521807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document cc +0.00040386867133151386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document cc +0.00024460575279105167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document cc +0.000269029789531335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document cc +0.0003573757493252864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document cc +0.0004600876681392076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document cc +0.0002605354166397086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document cc +0.0003882502452157999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document cc +0.0002466747612126512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document cc +0.0004024726105072402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document cc +0.00040820631128483644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document cc +0.0002691094350403538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document cc +0.00026916830387277267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document cc +0.0004204663297880574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document cc +0.00042379698687085554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document cc +0.0004502169227311871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document cc +0.0002661708937015295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document cc +0.00031239486948031334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document cc +0.0003109054589936201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document cc +0.00045873053079760646 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document cc +0.00022904931423244635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document cc +0.0003813462028433663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document cc +0.00039188129256500874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document cc +0.00045124222276983765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document cc +0.00048138658436853695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document cc +0.0003944178776279866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document cc +0.00039941569676754006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document cc +0.00037952761190240494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document cc +0.0003944870860881476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document cc +0.0003891842411856621 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document cc +0.000387688981934861 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document cc +0.00039197953876258005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document cc +0.00039007915280311206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document cc +0.0003995520363699188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document cc +0.00039230985654592406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document cc +0.0003929472067173851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document cc +0.0003924096172671473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document cc +0.0003881636143629905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document cc +0.000389790617937084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document cc +0.00037351762309221023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document cc +0.0003630196170929407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document cc +0.00033532465765142113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document cc +0.0003076088685761823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document cc +0.00039463850897720803 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document cc +0.0002843816115231449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document cc +0.0002909175709416474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document cc +0.00028867170997202486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document cc +0.0002838644617723659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document cc +0.00029027869525543416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document cc +0.0002821339567560056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document cc +0.0002922988877045601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document cc +0.0002866955958315786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document cc +0.0002865271754558126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document cc +0.0002861247475618473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document cc +0.0002826681072408606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document cc +0.0002849746458282827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document cc +0.0002816966633435316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document cc +0.00026255342235948463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document cc +0.0002552895098829678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document cc +0.00025990194083107813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document cc +0.0002524062657685835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document cc +0.0002538577379748611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document cc +0.0002561415177406761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document cc +0.00026206253059694905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document cc +0.00026168095406910565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document cc +0.0002601305742008613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document cc +0.00025200823006814814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document cc +0.0003229951981263502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document cc +0.00037289448266476045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document cc +0.0003807825862179898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document cc +0.0003616333738191483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document cc +0.0003665117918907636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document cc +0.0003684186453633228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document cc +0.0003589330610806066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document cc +0.00036383861418030395 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document cc +0.000359841363355303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document cc +0.00036431044063050464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document cc +0.0003668574090358279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document cc +0.000362768263620199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document cc +0.0003501888032771077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document cc +0.000352401968221528 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document cc +0.0003541019701869794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document cc +0.0003628121865546891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document cc +0.0003752582953758773 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document cc +0.00037902046230424966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document cc +0.0003777927146925147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document cc +0.0003760676130509053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document cc +0.00034046049078755405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document cc +0.0003338847563259091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document cc +0.00033294499102761794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document cc +0.0004912026198265864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document cc +0.00032064363474664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document cc +0.00032154190389541214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document cc +0.00032309660151746207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document cc +0.00031181143365304544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document cc +0.00031046092294569104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document cc +0.00031150165249068046 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document cc +0.0003041314265988224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document cc +0.0003024834909739394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document cc +0.0003019936835833604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document cc +0.000292329665283177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document cc +0.0002867061143144972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document cc +0.00028443615610701707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document cc +0.00028462291013755945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document cc +0.0002793538601205013 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document cc +0.00027306573977044246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document cc +0.00027097155673336525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document cc +0.0002752934202112985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document cc +0.00043042012694697647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document cc +0.00047495648822986177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document cc +0.00047755032493473855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document cc +0.0004706974343933747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document cc +0.00046682163297771817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document cc +0.0004616765425874178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document cc +0.00030644496751628097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document cc +0.0002909492555358308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document cc +0.00027272036068261724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document cc +0.0004101070217315588 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document cc +0.0003728914338834357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document cc +0.00036546911442305647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document cc +0.0003669945482407483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document cc +0.0003715902407424017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document cc +0.00035837486406683366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document cc +0.0003573318538685469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document cc +0.0003553784893071916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document cc +0.0004920659809912352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document cc +0.0004533619411303183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document cc +0.00045067066057818706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document cc +0.00044396985139270645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document cc +0.00043198288204468477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document cc +0.00043005174223738454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document cc +0.00041847118430776784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document cc +0.00042952036375796664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document cc +0.00043420594647324267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document cc +0.0003461123241053012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document cc +0.0003408581597849182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document cc +0.00033172705422182547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document cc +0.0003392566490686136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document cc +0.00033578341518385483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document cc +0.0003439196710518844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document cc +0.00034559163447085543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document cc +0.00033762478642902825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document cc +0.00033215210055107224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document cc +0.00033423579608014966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document cc +0.0004963355016025102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document cc +0.0004996862761456923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document cc +0.0005000551829325451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document cc +0.0005004212610098755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document cc +0.00027768695585500585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document cc +0.00028395983854338433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document cc +0.00027835826303062254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document cc +0.0002740073176010804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document cc +0.0002791830529274016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document cc +0.0002796863816194411 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document cc +0.00026697453022672804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document cc +0.0002594197440280141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document cc +0.0003779565697649222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document cc +0.00041835823476586606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document cc +0.00043788493575265915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document cc +0.0002731731970096006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document cc +0.000276305847423402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document cc +0.0002704955773958623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document cc +0.0002629635944827518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document cc +0.000260070956974436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document cc +0.00025661553791456334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document cc +0.00025794727207576157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document cc +0.00025295733980001527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document cc +0.0003788106407021029 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document cc +0.0004882344027669431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document cc +0.0003275324309642705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document cc +0.0004803401856640094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document cc +0.00046720138323433943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document cc +0.00043527810307095335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document cc +0.00043905395741627827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document cc +0.00048774175867331425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document cc +0.00048380704121346737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document cc +0.0004779011848346118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document cc +0.00046255587581908036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document cc +0.00045127922880511576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document cc +0.0004503891485256095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document cc +0.0004450142332303422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document cc +0.00044630282482516654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document cc +0.00044325014465743616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document cc +0.0004263874842796447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document cc +0.0004217530913646938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document cc +0.000415120314341852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document cc +0.00040987168279144537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document cc +0.00033468337266607834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document cc +0.0003353094464683005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document cc +0.0004833936821707294 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document cc +0.00047194878988920935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document cc +0.0004648324126996427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document cc +0.0004562345003964941 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document cc +0.0004933203505465098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document cc +0.0003530166075325466 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document cc +0.00035368548192804685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document cc +0.0004872620828289663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document cc +0.00048293889392426456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document cc +0.00047936768462267655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document cc +0.00047821013991587545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document cc +0.0004660610308564753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document cc +0.000394683430103437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document cc +0.00039165053441571324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document cc +0.0003906936040164381 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document cc +0.00038074803919159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document cc +0.0003686529291578143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document cc +0.00035832920428870976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document cc +0.00035929024535947033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document cc +0.0003538226556050544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document cc +0.0003584167868708799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document cc +0.0003480507542594234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document cc +0.0003413709023543034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document cc +0.00034001304759361455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document cc +0.00033430532902756514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document cc +0.00046519252660631277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document cc +0.0002938876402514769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document cc +0.00028676090994509047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document cc +0.00027296150117506716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document cc +0.00026513502621960483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document cc +0.0002680081327926125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document cc +0.00025831225828720344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document cc +0.00026647037295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document cc +0.0002525733734572654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document cc +0.00025831708887575375 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document cc +0.00042487627444443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document cc +0.0004951213245023891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document cc +0.0004804051413177752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document cc +0.0004662397611340532 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document cc +0.0004550138655253933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document cc +0.00044494909122746795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document cc +0.0002899112253051385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document cc +0.0004372879736279761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document cc +0.0004529568099252922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document cc +0.00045127826158829573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document cc +0.0004436558176737439 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document cc +0.0004419233237678378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document cc +0.000434589215880319 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document cc +0.00029153613207706566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document cc +0.0004312458058738854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document cc +0.00028741854968757313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document cc +0.00046853200754421234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document cc +0.0004949145252030074 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document cc +0.00044459683920483167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document cc +0.0003836095306696336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document cc +0.0003789760237872398 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document cc +0.0003749227438304427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document cc +0.0003628558277173369 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document cc +0.00039468301394041474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document cc +0.00038874701821614864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document cc +0.0004158492456077867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document cc +0.00042360504554060077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document cc +0.00040386729844317623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document cc +0.00027595096702902474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document cc +0.00043638766787829135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document cc +0.0002218691596850179 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document cc +0.0004437566108089954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document cc +0.0003889996411609667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document cc +0.00043454421906537704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document cc +0.0004522564392830988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document cc +0.00041517835659357416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document cc +0.0002614360863446896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document cc +0.00037543522111463596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document cc +0.0004386190133514781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document cc +0.00046358333286115075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document cc +0.00043186261317942404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document cc +0.0002377581602097957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document cc +0.00025973334085074254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document cc +0.00040139099332000796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document cc +0.00043674860686687174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document cc +0.00040853289309329373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document cc +0.000242910191729688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document cc +0.0004431071731750582 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document cc +0.0004388092670482523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document cc +0.000381418866255965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document cc +0.0004100117296419717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document cc +0.00042469230366022745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document cc +0.00041744151905374254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document cc +0.00022835699906752945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document cc +0.0004380161085387397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document cc +0.00044803212381807456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document cc +0.00040554932796137236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document cc +0.0004234508646347761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document cc +0.00043341209652360653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document cc +0.00023966604734537185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document cc +0.000259165907316014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document cc +0.0004270653021833602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document cc +0.0004341547032162028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document cc +0.0004111478117275994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document cc +0.0004299383567984396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document cc +0.0004241899124590779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document cc +0.0004502719349364145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document cc +0.00038994621469645615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document cc +0.0003859912398894952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document cc +0.0004247535950310557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document cc +0.000386982084327716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document cc +0.0004196451040053251 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document cc +0.0004096278509782259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document cc +0.0004373334932695721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document cc +0.0004180889975240641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document cc +0.00042079636929672745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document cc +0.00038063574611812913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document cc +0.0003817505891515542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document cc +0.0004420096268860222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document cc +0.00039182670726410623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document cc +0.0003635667850372299 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document cc +0.00041564996472055667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document cc +0.000400529358757286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document cc +0.0003939113874958451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document cc +0.00039066622068940996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document cc +0.0004290098538807143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document cc +0.0004240739958197099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document cc +0.00040775392659215333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document cc +0.0004091634200396925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document cc +0.00042299190476617914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document cc +0.0003701492680344151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document cc +0.0003807353844384635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document cc +0.00038813507771983156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document cc +0.00040072346558408346 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document cc +0.0003603595180423597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document cc +0.00038799421353112465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document cc +0.00037575235582264926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document cc +0.0004239190342959713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document cc +0.0004606044799136546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document cc +0.00045107950652529253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document cc +0.0004391947201871058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document cc +0.0004457516661123035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document cc +0.0004301297170991686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document cc +0.00044661704164586694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document cc +0.0004438849846114837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document cc +0.0004444205734316823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document cc +0.0004190924165303394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document cc +0.00043942581131677875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document cc +0.00021568459798090663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document cc +0.0003814929225407199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document cc +0.0003217453179359235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document cc +0.00031719591470267974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document cc +0.00032434115726922137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document cc +0.0004079911120371051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document cc +0.000329492766381148 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document cc +0.0003845916162001633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document cc +0.0003835208964390098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document cc +0.00037847334157173194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document cc +0.00038296039903791865 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document cc +0.00037896336828472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document cc +0.00037620974396391355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document cc +0.00037420590727111843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document cc +0.000340490625886403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document cc +0.0003078314411035827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document cc +0.00034153990750656097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document cc +0.0003308858103982067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document cc +0.0003452640607156025 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document cc +0.00033095276418403455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document cc +0.0003116308995860414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document cc +0.00032446713226408477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document cc +0.0003015816821912984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document cc +0.00031612418775706894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document cc +0.0003278516344971041 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document cc +0.00033079446736097217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document cc +0.00032278977146550837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document cc +0.00032065272988207914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document cc +0.0003936696452406576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document cc +0.0003450109536627789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document cc +0.0003339787189919641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document cc +0.0003284303856176974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document cc +0.00033652677276843477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document cc +0.0003257822443845694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document cc +0.0003293985569149334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document cc +0.0003310360260148262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document cc +0.0003233770986418526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document cc +0.0003172280092149422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document cc +0.0003160674744292835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document cc +0.00030931090289598506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document cc +0.0003093173886443107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document cc +0.00033167847081104083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document cc +0.00031131501311729723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document cc +0.00031046608876279845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document cc +0.00030569235942207244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document cc +0.00030777943671285197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document cc +0.00029303314290956683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document cc +0.0003045824546400205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document cc +0.00030360880677729793 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document cc +0.00031646239964835433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document cc +0.0003129122300603785 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document cc +0.00031060464956661433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document cc +0.000311819032500067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document cc +0.0002977872483902282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document cc +0.0003009448600922438 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document cc +0.00028610292098537774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document cc +0.0002988326876216654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document cc +0.00028550828372819075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document cc +0.0002830381750875739 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document cc +0.0002848495855927156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document cc +0.0002856443760308144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document cc +0.00027442895344188584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document cc +0.0002681160554049462 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document cc +0.0003421482544126989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document cc +0.0004005872948449718 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document cc +0.0003930123959320308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document cc +0.0003867271832275778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document cc +0.000380805140455254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document cc +0.0003814769861947819 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document cc +0.00038025170883282324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document cc +0.0003738026647867475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document cc +0.00018960856915036276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document cc +0.0003697177501953134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document cc +0.00036674194328136693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document cc +0.00036447406838697555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document cc +0.00036686410861101255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document cc +0.00035915267825103423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document cc +0.0003624758404026675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document cc +0.0002822812140180794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document cc +0.00030620512946920813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document cc +0.000294249776520589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document cc +0.00030238536967523434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document cc +0.00029509593361580754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document cc +0.0002906912701830899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document cc +0.0002921944165474959 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document cc +0.00028358919691127954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document cc +0.0002813182772323272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document cc +0.00027442640800299205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document cc +0.0002747820342933984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document cc +0.0002747584403979717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document cc +0.00027499129634862444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document cc +0.0002712050404257197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document cc +0.0002616256943143254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document cc +0.00026769938929002815 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document cc +0.00038396081322727017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document cc +0.0003863140490027991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document cc +0.00037702277513203237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document cc +0.0003633274156107032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document cc +0.0003587473889240435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document cc +0.0003507672084278415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document cc +0.00033776425499780385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document cc +0.0003377914127574796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document cc +0.00032948015659161326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document cc +0.00033245638541392985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document cc +0.00031080707640648695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document cc +0.0002976903331149755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document cc +0.0002965121463725523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document cc +0.0002933849695266647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document cc +0.0002837035078508233 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document cc +0.00028684569079589323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document cc +0.0003145192320802359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document cc +0.0003566937253273515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document cc +0.0003470199109592918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document cc +0.0003060245312041868 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document cc +0.0002650817213818789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document cc +0.0002643604938780134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document cc +0.000299350876031416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document cc +0.0003178540797697938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document cc +0.000271850367887767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document cc +0.00031349896596549 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document cc +0.00031749734412765755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document cc +0.0003791137842391209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document cc +0.0003742334169957992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document cc +0.0003705639757351107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document cc +0.0003126986769797042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document cc +0.00031038132814561196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document cc +0.00036464437173804883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document cc +0.0003569480488951322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document cc +0.0003541239221619106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document cc +0.00035315297411308053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document cc +0.0003572451925404141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document cc +0.0003514986129411253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document cc +0.0003521798298425866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document cc +0.00034553677439244716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document cc +0.000349004719809412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document cc +0.0003468247484872769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document cc +0.0003465822608356558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document cc +0.00035410983132162007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document cc +0.0003487908354969444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document cc +0.0003479024763238147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document cc +0.000341412530646823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document cc +0.00034451316273667034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document cc +0.0002618849993484869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document cc +0.00026788679978901144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document cc +0.00027450670773227214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document cc +0.0002661273129899329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document cc +0.00026836569676402957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document cc +0.00026155876975483236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document cc +0.0002609276830117151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document cc +0.0002644161630512771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document cc +0.00036789208972872557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document cc +0.00037829849439990513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document cc +0.0003788894943523098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document cc +0.0003617207777959397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document cc +0.0002541334487248998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document cc +0.0002707945538071073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document cc +0.00027046282716455214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document cc +0.0002652443167243215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document cc +0.0002685859923850986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document cc +0.00025734961751176414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document cc +0.000259041720872915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document cc +0.00025340107274823446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document cc +0.00025757135121837893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document cc +0.00025617700500574084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document cc +0.0002566931670562857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document cc +0.0002543871190716101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document cc +0.00024997565589481713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document cc +0.0002954079779456287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document cc +0.00034890741135252835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document cc +0.0003473298137731525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document cc +0.0003296959618486435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document cc +0.0003304520061604598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document cc +0.00032377956175729824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document cc +0.00031700696295168713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document cc +0.0003060382346081943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document cc +0.0003012003005056863 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document cc +0.0002981074073993884 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document cc +0.0002922128825950705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document cc +0.000348901087722931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document cc +0.0003408286289467841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document cc +0.0003410649680770183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document cc +0.0003358524215576502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document cc +0.0003343661874989231 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document cc +0.00032810573699389156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document cc +0.00032261449539097497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document cc +0.0003162694866049203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document cc +0.0003158381156468853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document cc +0.000317376061083603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document cc +0.0003125788639953052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document cc +0.0003010105041885602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document cc +0.0003065865059090678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document cc +0.0003084275726508053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document cc +0.00030966560718296085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document cc +0.0002957728057853081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document cc +0.00029904164542325336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document cc +0.0002955358888729187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document cc +0.00028692976446931544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document cc +0.0002923476214935797 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document cc +0.0002893691697212419 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document cc +0.0002855895211981585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document cc +0.00027968347097626246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document cc +0.0002810783462604979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document cc +0.00027794080455729715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document cc +0.00034784376461416953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document cc +0.0003488347959010943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document cc +0.00034790583710250724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document cc +0.000345913166618151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document cc +0.00033801936268066675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document cc +0.0003290591130212315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document cc +0.00034051399521366823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document cc +0.00032470943131841784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document cc +0.00031679540050914276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document cc +0.00031814596342422325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document cc +0.0003156466289485036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document cc +0.00029985010879003633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document cc +0.0002905176377776361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document cc +0.0004206836775460856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document cc +0.00020660449162246918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document cc +0.0003461727254468087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document cc +0.00020592870907067763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document cc +0.00034173505299233005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document cc +0.0004052437256652738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document cc +0.0004080650901351697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document cc +0.00039778184149144276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document cc +0.00039046311464950275 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document cc +0.00039043444911071384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document cc +0.000388575704932843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document cc +0.00019737533145666597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document cc +0.00037610755595812403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document cc +0.00037315400127598317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document cc +0.00037415028580922163 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document cc +0.00036694041707212337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document cc +0.00018947219857306515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document cc +0.00037046050826533545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document cc +0.0003587440768559087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document cc +0.00034623936498708903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document cc +0.0003502289592617922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document cc +0.00034692398063649823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document cc +0.000339340809421849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document cc +0.0003360510394816983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document cc +0.0003354673850814145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document cc +0.00032937682875877047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document cc +0.00032844505049317715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document cc +0.00028287199339908627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document cc +0.0002795217197003578 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document cc +0.00028048955601883463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document cc +0.0002769326396439027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document cc +0.0002727090021299243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document cc +0.0002726577841024554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document cc +0.00026663619593455374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document cc +0.00026068042672138127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document cc +0.0002637704114326801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document cc +0.0002593043567100412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document cc +0.0002599897110113453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document cc +0.0002435078682758859 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document cc +0.0002450530071379054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document cc +0.00024233331983743606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document cc +0.0002934750947999535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document cc +0.00033241226364044474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document cc +0.00032938406090272075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document cc +0.00032778705403953246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document cc +0.00032184551480398754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document cc +0.00031874002264945737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document cc +0.0003165319685666433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document cc +0.00031307071173376295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document cc +0.00031119524184911957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document cc +0.0003102253344576429 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document cc +0.0003088976240383192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document cc +0.0002951410823077708 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document cc +0.00029772657676757413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document cc +0.0003056048989909935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document cc +0.00031991305381648026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document cc +0.00030890256978362426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document cc +0.0003109382904091933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document cc +0.00031035798529690644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document cc +0.00030741666395911753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document cc +0.0002989918594861846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document cc +0.00029569635443989434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document cc +0.0002973992445667285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document cc +0.000293397351001072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document cc +0.00028737817438047954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document cc +0.00028252738144009747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document cc +0.0002805511898623541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document cc +0.0003718020784620472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document cc +0.0003499713845765235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document cc +0.00034283547445326676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document cc +0.00031464759888838765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document cc +0.00033188946446414833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document cc +0.000326084432195463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document cc +0.0003764568303917893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document cc +0.0003604955598858414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document cc +0.0003655654554133222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document cc +0.00035762304033750504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document cc +0.00038478883950347103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document cc +0.00027735714341247454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document cc +0.00028139534607773563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document cc +0.00019777292251713763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document cc +0.000285571704874486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document cc +0.00028543482146244363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document cc +0.00019434234484256758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document cc +0.00027854908176986763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document cc +0.0002847068039566143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document cc +0.00028672356943064853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document cc +0.00027782687605808177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document cc +0.0002843539634105203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document cc +0.0002894748379090401 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document cc +0.0002868852440186493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document cc +0.0002818504885373851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document cc +0.00028680112812941034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document cc +0.00019258978168723977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document cc +0.00028760637934715155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document cc +0.0002820439443912918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document cc +0.0002831001054410018 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document cc +0.00029001901552467397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document cc +0.00027779449377883156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document cc +0.00019949837437516796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document cc +0.0002907306472984446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document cc +0.00027814858381318327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document cc +0.00019472790889161432 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document cc +0.00020472626596924125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document cc +0.0002870045081974301 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document cc +0.00019812241927078482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document cc +0.0002817553333369554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document cc +0.00027829782796642117 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document cc +0.00028289431732284113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document cc +0.0002795526296717729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document cc +0.00027682829988044574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document cc +0.0002895432402719184 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document cc +0.0002823174903941811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document cc +0.00028170972351837796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document cc +0.00027807915877838826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document cc +0.00028588515681452956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document cc +0.00028112324090816726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document cc +0.00020636178289985485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document cc +0.00019447255290980535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document cc +0.0002850824220591452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document cc +0.00027856429520116784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document cc +0.0002820880676635633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document cc +0.00028943902215995714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document cc +0.0002676366291085329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document cc +0.00023806333809954687 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document cc +0.00024526460430233455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document cc +0.00023876876664622726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document cc +0.00023379770334179805 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document cc +0.00024175151269138382 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document cc +0.00023386583242595706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document cc +0.00023771797150160827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document cc +0.0002262748967483896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document cc +0.0002408148346432682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document cc +0.00023398651720444235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document cc +0.00022989433874474592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document cc +0.00023948500543957772 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document cc +0.0002331594076859196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document cc +0.00023375132439600242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document cc +0.00023923410909668642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document cc +0.00023952796315562954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document cc +0.0002327466076905069 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document cc +0.00023082758956797212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document cc +0.0002240509275524448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document cc +0.00022798879995765268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document cc +0.000221172516774386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document cc +0.00021767045123534623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document cc +0.00021982832794804484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document cc +0.00021971626543789102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document cc +0.00022566565206920132 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document cc +0.0002181984894194856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document cc +0.00021831417549554653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document cc +0.00021601405421187145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document cc +0.00022275733725519607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document cc +0.00021847734911973986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document cc +0.0002243591012664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document cc +0.00021688758139483833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document cc +0.0002182953624789215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document cc +0.00020475155724026002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document cc +0.00021498078062960065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document cc +0.0002157914337233064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document cc +0.00021781838494967963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document cc +0.00021723242266814558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document cc +0.0002176782686553837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document cc +0.0003486179404943968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document cc +0.00034882846352857634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document cc +0.00031400868448352596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document cc +0.00030273484020011963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document cc +0.00029895889118145404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document cc +0.00029770764609621714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document cc +0.0002990181332116852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document cc +0.00029653733972285996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document cc +0.00029624649222942476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document cc +0.00029625609720203576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document cc +0.00029731928930852147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document cc +0.00029011721326148513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document cc +0.00028849788197494655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document cc +0.00021601278623858145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document cc +0.00021319599281739178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document cc +0.0002153325290600083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document cc +0.00018566946174516558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document cc +0.00020736824394291617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document cc +0.00020857419820128004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document cc +0.00020058526129536423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document cc +0.00020745812166665217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document cc +0.00020652171015271702 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document cc +0.00020643808911278608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document cc +0.00020040513914482103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document cc +0.00020598050188272898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document cc +0.0001969184139343296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document cc +0.0001972748812937012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document cc +0.0002038556751586195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document cc +0.00020245186011313464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document cc +0.00019950381422038783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document cc +0.00020837055459665258 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document cc +0.00020371856218246096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document cc +0.00019537612301625791 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document cc +0.00019914984508813857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document cc +0.0002053787713691309 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document cc +0.00019082100541008637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document cc +0.00020397153334531813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document cc +0.0002021462693077317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document cc +0.00019609357008124035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document cc +0.00019693256622486236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document cc +0.00020007239732428112 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document cc +0.00020467075741591954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document cc +0.00019584883400022932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document cc +0.00019135050391176972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document cc +0.0003362829834208298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document cc +0.00034013691154784095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document cc +0.00033215887031941976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document cc +0.00032681189065396707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document cc +0.0003149138485493094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document cc +0.00030179177307540077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document cc +0.0002923278437581119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document cc +0.00029470052278994486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document cc +0.0002994095093045731 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document cc +0.00029033525096085037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document cc +0.00029390798852496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document cc +0.0002916230924130842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document cc +0.00029419886374594913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document cc +0.0002865469756730764 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document cc +0.00021191292549942086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document cc +0.00021369664817409847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document cc +0.00021612485624266726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document cc +0.00022242192634588478 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document cc +0.00014605095659989698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document cc +0.00022070626106341693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document cc +0.0002174420774054071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document cc +0.00021325858963116995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document cc +0.0002124322999488052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document cc +0.0002081218896969054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document cc +0.0002108710211556957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document cc +0.00020686867095978426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document cc +0.00020895752681041895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document cc +0.00020741922266415738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document cc +0.0002069112657197308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document cc +0.00020644627473468118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document cc +0.00020332991338121604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document cc +0.0003560895677789848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document cc +0.00032915779111908214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document cc +0.00033810613317040864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document cc +0.00033729626594036923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document cc +0.00033550342864602944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document cc +0.00034173474024556906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document cc +0.000331505340748827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document cc +0.0003270050330117195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document cc +0.00032585275329172556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document cc +0.0003143383203190604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document cc +0.00031655199110388894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document cc +0.00030738872158476413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document cc +0.00030838388352699285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document cc +0.0003053596995351888 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document cc +0.00031836304739584593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document cc +0.000315315435873905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document cc +0.0003087116248965243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document cc +0.00030396790625537645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document cc +0.0003335812246032149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document cc +0.00034570956323095843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document cc +0.00034563035636675786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document cc +0.00033411265479076335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document cc +0.00034439191141692787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document cc +0.0003364483125496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document cc +0.0003299500453608033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document cc +0.00033163377700074837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document cc +0.00032638649660627673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document cc +0.00032616167939645234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document cc +0.0003205289298760723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document cc +0.00031939393740815355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document cc +0.00031593164066731296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document cc +0.00031928871111254405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document cc +0.00029670189073175004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document cc +0.00020517703846735904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document cc +0.00020128418186172073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document cc +0.00019662723895606717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document cc +0.0001981157042081407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document cc +0.00019703489037041608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document cc +0.00019079796331785068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document cc +0.0001909352306690079 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document cc +0.00018824662295261396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document cc +0.00019864275319325954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document cc +0.00018818516521649587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document cc +0.00018875694972812844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document cc +0.00018231621170645482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document cc +0.00018349407845798273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document cc +0.00018088971427746906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document cc +0.00018296284236327237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document cc +0.0001876011825819916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document cc +0.000329052068725176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document cc +0.00032223616273648536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document cc +0.00031272564089633955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document cc +0.00031621609908414494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document cc +0.0003117213560911235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document cc +0.00030218064069945934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document cc +0.00030658916600512085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document cc +0.0002915863534115821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document cc +0.0002940280138374372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document cc +0.00029067860468866085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document cc +0.00028529228063135635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document cc +0.00028336893301452256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document cc +0.0002794668089130099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document cc +0.00021681361378827842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document cc +0.0001484664674497246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document cc +0.00021950558378215133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document cc +0.00021806860758808645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document cc +0.00021819568718852282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document cc +0.00021626925931585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document cc +0.0001464536143077762 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document cc +0.00021432777088808917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document cc +0.000213473805865147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document cc +0.00021397067253964538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document cc +0.00020758957647437263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document cc +0.00020687124337683314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document cc +0.00020630057046511005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document cc +0.0002091166859352538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document cc +0.00020777355025615267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document cc +0.00020709287641496176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document cc +0.00020736464660577094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document cc +0.00020062246741862607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document cc +0.00020693207561942915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document cc +0.00021151004871893024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document cc +0.00019930249098689716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document cc +0.00021589710041231824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document cc +0.00021369204789905741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document cc +0.0002147099923936778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document cc +0.00021077531190389536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document cc +0.0002100509829113836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document cc +0.00021185362601571124 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document cc +0.00020722136637339565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document cc +0.00020300093701169531 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document cc +0.00019859737993313477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document cc +0.00019971314372100164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document cc +0.00019549908270269278 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document cc +0.00019649820843534028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document cc +0.00019619415513498067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document cc +0.00019493006120377898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document cc +0.00019499409035775506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document cc +0.00019252988593634277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document cc +0.00019440768268686405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document cc +0.00018747161324755577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document cc +0.0001879575932372779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document cc +0.00019040707058357506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document cc +0.0001871931095090703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document cc +0.00020112966223017096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document cc +0.00020516878165311017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document cc +0.00020664735191740533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document cc +0.00021041398572882962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document cc +0.00020397992929690396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document cc +0.0002039978580295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document cc +0.00020592785601142126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document cc +0.0001990755527445265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document cc +0.00019729564847798732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document cc +0.00019958182230527032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document cc +0.0001985037302636386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document cc +0.00020204130355115716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document cc +0.0002000296401958085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document cc +0.0001983064832295463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document cc +0.00019663108484195617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document cc +0.00019510678560556523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document cc +0.0001873284057063206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document cc +0.00019311553072495885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document cc +0.00034652137288816547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document cc +0.0002813690318850024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document cc +0.00027697649713138685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document cc +0.0002755419092534421 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document cc +0.0002681583054440219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document cc +0.00026945753192750824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document cc +0.00026169470768245737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document cc +0.00026437008960810825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document cc +0.0002637294838228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document cc +0.00026491867965088836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document cc +0.00025504483625138986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document cc +0.0002545040623796586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document cc +0.0002546682814073622 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document cc +0.00025545439487142615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document cc +0.0002626896557978271 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document cc +0.00025092040940402784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document cc +0.0002589154885863872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document cc +0.00024106160482721467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document cc +0.0002483289690087987 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document cc +0.0002388930282784437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document cc +0.00024006340759273874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document cc +0.00023765248178029045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document cc +0.00023061351965578936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document cc +0.00024954224883546477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document cc +0.00017861017233018525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document cc +0.00017810832743667658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document cc +0.00017599709170759497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document cc +0.00017462723516505223 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document cc +0.0002906316527068669 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document cc +0.00033762141066247166 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document cc +0.00017170670574152494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document cc +0.00017258674515137717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document cc +0.0002815386173173926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document cc +0.0002996845935618989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document cc +0.0002735268488987296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document cc +0.0002971738713071517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document cc +0.0002942690674002763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document cc +0.0003322222207729567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document cc +0.0003378721656198464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document cc +0.00018307262621851067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document cc +0.00033956081502775057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document cc +0.00031604820927876276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document cc +0.00028805657681088917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document cc +0.00026312293321215633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document cc +0.00034366936722921455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document cc +0.0002865256504406559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document cc +0.0003063615195861786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document cc +0.00028412791619666136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document cc +0.00028060835132727154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document cc +0.00032544974761560506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document cc +0.0002647177833217225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document cc +0.0003152621884896575 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document cc +0.0003054625140336913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document cc +0.00031183308312292263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document cc +0.00018175026696621178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document cc +0.00017699918328872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document cc +0.00018222339261441908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document cc +0.00018348005930964137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document cc +0.0001810735993810541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document cc +0.00030846441282038914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document cc +0.0002972326889310354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document cc +0.00017433421318235594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document cc +0.00032799458649525895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document cc +0.00032482130048512673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document cc +0.00031943465668672475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document cc +0.00029615593630484517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document cc +0.0002893126939511001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document cc +0.0002849288351723284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document cc +0.00028383906633569267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document cc +0.00028072526091262615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document cc +0.000284239564292377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document cc +0.0002778903109432523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document cc +0.0002771644389501471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document cc +0.0002733316182319337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document cc +0.00026362539185869363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document cc +0.0002636325383220217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document cc +0.00026740622442302886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document cc +0.0002646771971853427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document cc +0.0002628566720605389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document cc +0.0002644760695434766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document cc +0.0002623837702310999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document cc +0.00026088722976772894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document cc +0.0002567065374799158 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document cc +0.00018857382101207726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document cc +0.00019036580399817203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document cc +0.00018348828065261222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document cc +0.00018491851780345073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document cc +0.00018904887260080187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document cc +0.0001875609304251801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document cc +0.00018393034720015817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document cc +0.00018419795526114903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document cc +0.00018699955623404795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document cc +0.00018276256902965128 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document cc +0.00017698045695190812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document cc +0.00018104650132303642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document cc +0.00017758206731279688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document cc +0.00017131402995103497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document cc +0.000175944428350446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document cc +0.0003416745727147391 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document cc +0.0003163259373952889 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document cc +0.0002804489269172448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document cc +0.00028748272397403175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document cc +0.00027603318345630605 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document cc +0.000271638824679648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document cc +0.0002763761210210942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document cc +0.00026501984873172717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document cc +0.00026422486894694714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document cc +0.0002686339100849262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document cc +0.0002610837453940606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document cc +0.000260974343729353 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document cc +0.0002599403837029134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document cc +0.0002937273113238609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document cc +0.0003341790732600504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document cc +0.0002620661576600244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document cc +0.0003027929169239288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document cc +0.00031944039129326894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document cc +0.00019025676304139009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document cc +0.00018680910145009907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document cc +0.00034215840419416437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document cc +0.00018618120812119364 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document cc +0.00018605853095599425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document cc +0.00018120712626096538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document cc +0.00018315079292495327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document cc +0.00018362556449041974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document cc +0.0001780024456718171 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document cc +0.00033296526436178697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document cc +0.0001802398632282846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document cc +0.00017340263100798256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document cc +0.00017755840547238697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document cc +0.00018419413735260606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document cc +0.00017869518174591322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document cc +0.00017526271460129484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document cc +0.00017852168597981907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document cc +0.00017566536156787157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document cc +0.00017589867964432936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document cc +0.00017831487394075305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document cc +0.00017837310528935862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document cc +0.00018200908814216548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document cc +0.0001795136627511612 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document cc +0.0003414021775300033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document cc +0.00017177291787788502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document cc +0.0003441900648571877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document cc +0.0003394534597060673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document cc +0.0003236887233114832 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document cc +0.0001639544129688747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document cc +0.00019137443753211255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document cc +0.00018575146284680153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document cc +0.00019184792863440243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document cc +0.00018966043065679055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document cc +0.00017968851317035848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document cc +0.00018479881897661546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document cc +0.0001813642692683015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document cc +0.0001686449798983066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document cc +0.00018516104592230446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document cc +0.00031283726601066385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document cc +0.0003248607542883853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document cc +0.00031583241601202365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document cc +0.00031238270857730376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document cc +0.000307150592403979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document cc +0.00029443829986847044 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document cc +0.0002942723732234677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document cc +0.00023514930666443422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document cc +0.0020776328951453444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document cc +0.0021768234410538883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document cc +0.002106973549276289 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document cc +0.002110915756171751 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document cc +0.0017032382109816464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document cc +0.0019047944877712286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document cc +0.0019402711744016077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document cc +0.0006264790011223686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document cc +0.0017885401938106643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document cc +0.0003547982093445404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon +6.322825248625475e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika +0.001451215788905126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train +0.0012499632072059553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o +0.0005759963691850877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit +0.0009994361338078242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange +0.004474659408857016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder +0.00032927705604725614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu +0.003548077173506675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki diff --git a/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt b/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt index 2bb2ea805e7..2cc52b55f60 100644 --- a/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt +++ b/ALCF/data-lists/aurora/dolma_v1_7_file_list.txt @@ -1,2419 +1,2419 @@ -0.0018520780893211373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document -0.0017591050606817512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document -0.001459052794333798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document -0.0007405667281569194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document -0.00019420030110896795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document -0.0009008668715801845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document -0.00015115827957143057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document -0.0014552844319220648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document -0.0012469861325685161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document -0.00136412011372413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document -0.0007064279699221103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document -0.0008472240000687427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document -0.0001984375713341955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document -0.0005472773881697123 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document -0.001815779629850992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document -0.0018313600689757324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document -0.0002583902668716813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document -0.0002646575141232155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document -0.0003165521247456758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document -0.0002920706460176214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document -0.00028396813182810215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document -0.00030445161883108107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document -0.00031628781276576474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document -0.0003083776568189157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document -0.0003176359471472902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document -0.0002536009369131698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document -0.0003067491424681363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document -0.0002597217257557784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document -0.0003788556450109768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document -0.0002796563272052598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document -0.00033573826524290287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document -0.00030523658022800287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document -0.00032211552192240096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document -0.0003329295675164247 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document -0.0003101982186639862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document -0.00032361798234223355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document -0.0003495541581652915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document -0.0002821637448858042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document -0.00030399523537629673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document -0.0002955658968247219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document -0.00028942158502924254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document -0.00028769546171490733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document -0.0002938111057234182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document -0.0002711150403010948 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document -0.00031130095874747565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document -0.0003002996118160777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document -0.0003732757901604459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document -0.00026784205751795894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document -0.0002799626521661984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document -0.00034334276069078164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document -0.0003582469803674965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document -0.00031094844818418623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document -0.0002766228384977191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document -0.00030297116159471485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document -0.00027033888377464685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document -0.00030090862368377933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document -0.00028543875802490955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document -0.00027559768459074204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document -0.0003182185533962886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document -0.0003311392971435837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document -0.00028751652060804325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document -0.000303466863212589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document -0.00033400462801277524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document -0.0002589234031777426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document -0.0002913508598466723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document -0.0002670572450004856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document -0.00032027399105647656 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document -0.00032188376258379377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document -0.0003161585784100882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document -0.0003184249182974135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document -0.00030381336664000807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document -0.0003190437442184283 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document -0.0002537961798200545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document -0.0003017817117223326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document -0.00028685268513240224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document -0.00031265179094451165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document -0.00034708319096986816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document -0.00026650837943080664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document -0.00034588832248507335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document -0.0002416982248399037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document -0.0003089296918222243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document -0.00029137184185700827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document -0.00026464226846800774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document -0.00030545397919456627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document -0.0003206778460448875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document -0.00030968971641110967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document -0.00023325653928600864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document -0.00030526899198338555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document -0.00035376719076633584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document -0.000290224385981026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document -0.000294650083382008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document -0.00028768858128616436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document -0.00030856965235527843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document -0.00030579942447879054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document -0.0002863101084704357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document -0.0002870032092492213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document -0.000264182727569885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document -0.0002974012367036449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document -0.00032238412143059203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document -0.00031683716893819036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document -0.00031157434937617524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document -0.0003411742735695989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document -0.00026778444816570715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document -0.0003037045797275201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document -0.00027746114370081314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document -0.00027148285946862043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document -0.00028042950114678207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document -0.0003235607816590721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document -0.0003086692227306295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document -0.00033990349455148105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document -0.00030945053208470265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document -0.00027309074552265303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document -0.00028737393506316194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document -0.0003098868328009879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document -0.0002614229162588409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document -0.0002884388407820923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document -0.0031025147279277244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document -0.003102019887362634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document -0.0009996745994661548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document -0.0002406272620255565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document -0.0002404825539493424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document -0.00024062296575435581 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document -0.00024069315766818953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document -0.00024055829162263452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document -0.00024062053397343032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document -0.0002410715545206964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document -0.00024024881846087368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document -0.0002407074700790688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document -0.00024072141428809043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document -0.00024027710230872736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document -0.0002409111299205489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document -0.00024081954058275009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document -0.00024086076794990912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document -0.00024098672620832446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document -0.00024068622303333862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document -0.00024140627024291824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document -0.0002414512033594384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document -0.00024028742594941463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document -0.00024018036089269645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document -0.0002398347365034979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document -0.00024006780153485276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document -0.00024015620270419213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document -0.0002408848259695227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document -0.0002408023185278831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document -0.00024021196580140326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document -0.00024077677271297493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document -0.00024087392454668027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document -0.0002408071293824126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document -0.00024042223828845715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document -0.0002411484752360495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document -0.00023605263746465907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document -0.00023471222158326908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document -0.00023432138580287644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document -0.00023407385623382327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document -0.00023487504174367091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document -0.0002341843704976313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document -0.00023421993170282486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document -0.00023445057969132037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document -0.0002337681680073047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document -0.000234627964808109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document -0.0002338942211888584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document -0.00023403849286843386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document -0.00023405641310796305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document -0.00023349169562397965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document -0.00023381157386048856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document -0.00023388742993790587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document -0.00023363103829469813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document -0.00023421141834630477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document -0.00023420564352232565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document -0.00023367463699173143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document -0.00023344969163567033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document -0.00023372196941547188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document -0.00023399207645297834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document -0.00023357915605505856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document -0.00023337585642190864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document -0.00023385005470157914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document -0.00023301533534493465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document -0.00023377864302541782 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document -0.00023323745848621437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document -0.0002330594611151835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document -0.0002334149675026783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document -0.00023198945902291534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document -0.00023023784834634142 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document -0.00022985623060187217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document -0.0002292605284569516 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document -0.00022926593333048894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document -0.00022922766406807777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document -0.00022898153911167426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document -0.0002292473111593315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document -0.000228804579400424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document -0.00022865485613513526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document -0.00022937426835887895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document -0.00022917388311587372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document -0.0002291660582019043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document -0.00022907895248360543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document -0.0002294617879920205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document -0.0002290452150516566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document -0.00022943405619715553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document -0.0002296271421006204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document -0.00022854791372910372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document -0.00022923123467686557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document -0.00022852404355738494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document -0.00022847798660086642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document -0.0002289604586810316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document -0.00022835479834950643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document -0.0002289149402884243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document -0.00022806655474763446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document -0.00022826296420992974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document -0.00022906829636213627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document -0.0002287628414466998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document -0.0002282673911253445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document -0.00022869309841939134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document -0.0002281540116815451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document -0.0002259755756162738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document -0.00022562331285233504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document -0.0002259061146106053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document -0.00022567670836663787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document -0.00022573165387587061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document -0.00022508514961670572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document -0.00022564642513773356 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document -0.00022563088621998788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document -0.0002250438755373707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document -0.00022524465346241134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document -0.00022531737657666812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document -0.00022444687519363458 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document -0.00022460397498596298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document -0.00022454218976501763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document -0.00022447528843671366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document -0.00022501666332178926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document -0.00022453752304377972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document -0.00022484451871163002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document -0.00022465678847154914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document -0.00022453180917044732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document -0.0002247278486823009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document -0.00022465794828242097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document -0.00022431000701925386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document -0.00022476020248460963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document -0.00022467531771795015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document -0.0002236391309945234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document -0.00022458764920536007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document -0.00022430877426744415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document -0.0002247047786127192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document -0.0002245298090400035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document -0.0002245648831396188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document -0.00022292894729820784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document -0.00022236668082957533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document -0.0002217622659895442 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document -0.00022252452726732609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document -0.00022135333211363678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document -0.0002214571757787971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document -0.0002217188139237798 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document -0.00022144214894640303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document -0.00022100172806631854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document -0.00022156392409199052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document -0.00022134830143710272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document -0.00022158598922529453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document -0.00022142932483041377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document -0.00022120980907786554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document -0.00022117917738112441 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document -0.00022077089397851235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document -0.00022093265074996711 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document -0.00022091299741377004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document -0.0002205849150703338 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document -0.0002210648204787979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document -0.0002214235747364102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document -0.00022083907302221787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document -0.0002206334237915964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document -0.00022065193929912214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document -0.00022079775597767288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document -0.00022091492909963518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document -0.00022095009987097293 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document -0.0002208150577180165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document -0.00022085759102772088 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document -0.00022073789170129016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document -0.00022049322781182384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document -0.00022083270617761285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document -0.00021982452827473632 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document -0.00021899870446514259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document -0.00021890358773356361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document -0.00021875556609042841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document -0.00021861195987201226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document -0.00021856782186167455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document -0.00021912837771543515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document -0.00021900213768517756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document -0.00021871675851390374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document -0.0002180537056545586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document -0.0002188196714327129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document -0.00021851362624523464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document -0.0002183236795498736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document -7.291153618675672e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document -0.0003742481815405742 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document -0.00038204855962733055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document -0.00038821818392663593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document -0.00038723332988783727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document -0.00038916141142149904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document -0.00038049542523949033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document -0.0003854755539534284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document -0.00024202756466512517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document -0.0003915405155008087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document -0.0003927382151931033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document -0.0003839151202260479 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document -0.00040006817468967907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document -0.00040318965964443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document -0.0003831013019452741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document -0.00039166638383204036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document -0.00039962784023961004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document -0.00039536707853602614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document -0.0004204304698247758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document -0.00041538899178693555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document -0.00039186953333675306 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document -0.00038945837196504305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document -0.0003919951238929062 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document -0.00044377065718528966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document -0.0004407759068603017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document -0.0002487811895843715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document -0.00039349432045556636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document -0.00041223198559462343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document -0.0004036573014830213 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document -0.0003825982215521807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document -0.00040386867133151386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document -0.00024460575279105167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document -0.000269029789531335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document -0.0003573757493252864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document -0.0004600876681392076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document -0.0002605354166397086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document -0.0003882502452157999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document -0.0002466747612126512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document -0.0004024726105072402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document -0.00040820631128483644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document -0.0002691094350403538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document -0.00026916830387277267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document -0.0004204663297880574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document -0.00042379698687085554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document -0.0004502169227311871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document -0.0002661708937015295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document -0.00031239486948031334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document -0.0003109054589936201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document -0.00045873053079760646 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document -0.00022904931423244635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document -0.0003813462028433663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document -0.00039188129256500874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document -0.00045124222276983765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document -0.00048138658436853695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document -0.0003944178776279866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document -0.00039941569676754006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document -0.00037952761190240494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document -0.0003944870860881476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document -0.0003891842411856621 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document -0.000387688981934861 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document -0.00039197953876258005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document -0.00039007915280311206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document -0.0003995520363699188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document -0.00039230985654592406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document -0.0003929472067173851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document -0.0003924096172671473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document -0.0003881636143629905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document -0.000389790617937084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document -0.00037351762309221023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document -0.0003630196170929407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document -0.00033532465765142113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document -0.0003076088685761823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document -0.00039463850897720803 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document -0.0002843816115231449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document -0.0002909175709416474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document -0.00028867170997202486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document -0.0002838644617723659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document -0.00029027869525543416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document -0.0002821339567560056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document -0.0002922988877045601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document -0.0002866955958315786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document -0.0002865271754558126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document -0.0002861247475618473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document -0.0002826681072408606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document -0.0002849746458282827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document -0.0002816966633435316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document -0.00026255342235948463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document -0.0002552895098829678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document -0.00025990194083107813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document -0.0002524062657685835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document -0.0002538577379748611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document -0.0002561415177406761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document -0.00026206253059694905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document -0.00026168095406910565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document -0.0002601305742008613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document -0.00025200823006814814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document -0.0003229951981263502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document -0.00037289448266476045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document -0.0003807825862179898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document -0.0003616333738191483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document -0.0003665117918907636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document -0.0003684186453633228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document -0.0003589330610806066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document -0.00036383861418030395 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document -0.000359841363355303 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document -0.00036431044063050464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document -0.0003668574090358279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document -0.000362768263620199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document -0.0003501888032771077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document -0.000352401968221528 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document -0.0003541019701869794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document -0.0003628121865546891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document -0.0003752582953758773 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document -0.00037902046230424966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document -0.0003777927146925147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document -0.0003760676130509053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document -0.00034046049078755405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document -0.0003338847563259091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document -0.00033294499102761794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document -0.0004912026198265864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document -0.00032064363474664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document -0.00032154190389541214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document -0.00032309660151746207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document -0.00031181143365304544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document -0.00031046092294569104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document -0.00031150165249068046 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document -0.0003041314265988224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document -0.0003024834909739394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document -0.0003019936835833604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document -0.000292329665283177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document -0.0002867061143144972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document -0.00028443615610701707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document -0.00028462291013755945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document -0.0002793538601205013 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document -0.00027306573977044246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document -0.00027097155673336525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document -0.0002752934202112985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document -0.00043042012694697647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document -0.00047495648822986177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document -0.00047755032493473855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document -0.0004706974343933747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document -0.00046682163297771817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document -0.0004616765425874178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document -0.00030644496751628097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document -0.0002909492555358308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document -0.00027272036068261724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document -0.0004101070217315588 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document -0.0003728914338834357 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document -0.00036546911442305647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document -0.0003669945482407483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document -0.0003715902407424017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document -0.00035837486406683366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document -0.0003573318538685469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document -0.0003553784893071916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document -0.0004920659809912352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document -0.0004533619411303183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document -0.00045067066057818706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document -0.00044396985139270645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document -0.00043198288204468477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document -0.00043005174223738454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document -0.00041847118430776784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document -0.00042952036375796664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document -0.00043420594647324267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document -0.0003461123241053012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document -0.0003408581597849182 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document -0.00033172705422182547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document -0.0003392566490686136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document -0.00033578341518385483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document -0.0003439196710518844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document -0.00034559163447085543 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document -0.00033762478642902825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document -0.00033215210055107224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document -0.00033423579608014966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document -0.0004963355016025102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document -0.0004996862761456923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document -0.0005000551829325451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document -0.0005004212610098755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document -0.00027768695585500585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document -0.00028395983854338433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document -0.00027835826303062254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document -0.0002740073176010804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document -0.0002791830529274016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document -0.0002796863816194411 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document -0.00026697453022672804 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document -0.0002594197440280141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document -0.0003779565697649222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document -0.00041835823476586606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document -0.00043788493575265915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document -0.0002731731970096006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document -0.000276305847423402 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document -0.0002704955773958623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document -0.0002629635944827518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document -0.000260070956974436 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document -0.00025661553791456334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document -0.00025794727207576157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document -0.00025295733980001527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document -0.0003788106407021029 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document -0.0004882344027669431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document -0.0003275324309642705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document -0.0004803401856640094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document -0.00046720138323433943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document -0.00043527810307095335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document -0.00043905395741627827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document -0.00048774175867331425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document -0.00048380704121346737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document -0.0004779011848346118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document -0.00046255587581908036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document -0.00045127922880511576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document -0.0004503891485256095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document -0.0004450142332303422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document -0.00044630282482516654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document -0.00044325014465743616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document -0.0004263874842796447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document -0.0004217530913646938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document -0.000415120314341852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document -0.00040987168279144537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document -0.00033468337266607834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document -0.0003353094464683005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document -0.0004833936821707294 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document -0.00047194878988920935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document -0.0004648324126996427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document -0.0004562345003964941 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document -0.0004933203505465098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document -0.0003530166075325466 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document -0.00035368548192804685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document -0.0004872620828289663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document -0.00048293889392426456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document -0.00047936768462267655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document -0.00047821013991587545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document -0.0004660610308564753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document -0.000394683430103437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document -0.00039165053441571324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document -0.0003906936040164381 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document -0.00038074803919159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document -0.0003686529291578143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document -0.00035832920428870976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document -0.00035929024535947033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document -0.0003538226556050544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document -0.0003584167868708799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document -0.0003480507542594234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document -0.0003413709023543034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document -0.00034001304759361455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document -0.00033430532902756514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document -0.00046519252660631277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document -0.0002938876402514769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document -0.00028676090994509047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document -0.00027296150117506716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document -0.00026513502621960483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document -0.0002680081327926125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document -0.00025831225828720344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document -0.00026647037295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document -0.0002525733734572654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document -0.00025831708887575375 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document -0.00042487627444443476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document -0.0004951213245023891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document -0.0004804051413177752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document -0.0004662397611340532 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document -0.0004550138655253933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document -0.00044494909122746795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document -0.0002899112253051385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document -0.0004372879736279761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document -0.0004529568099252922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document -0.00045127826158829573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document -0.0004436558176737439 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document -0.0004419233237678378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document -0.000434589215880319 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document -0.00029153613207706566 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document -0.0004312458058738854 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document -0.00028741854968757313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document -0.00046853200754421234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document -0.0004949145252030074 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document -0.00044459683920483167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document -0.0003836095306696336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document -0.0003789760237872398 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document -0.0003749227438304427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document -0.0003628558277173369 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document -0.00039468301394041474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document -0.00038874701821614864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document -0.0004158492456077867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document -0.00042360504554060077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document -0.00040386729844317623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document -0.00027595096702902474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document -0.00043638766787829135 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document -0.0002218691596850179 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document -0.0004437566108089954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document -0.0003889996411609667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document -0.00043454421906537704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document -0.0004522564392830988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document -0.00041517835659357416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document -0.0002614360863446896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document -0.00037543522111463596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document -0.0004386190133514781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document -0.00046358333286115075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document -0.00043186261317942404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document -0.0002377581602097957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document -0.00025973334085074254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document -0.00040139099332000796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document -0.00043674860686687174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document -0.00040853289309329373 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document -0.000242910191729688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document -0.0004431071731750582 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document -0.0004388092670482523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document -0.000381418866255965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document -0.0004100117296419717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document -0.00042469230366022745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document -0.00041744151905374254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document -0.00022835699906752945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document -0.0004380161085387397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document -0.00044803212381807456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document -0.00040554932796137236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document -0.0004234508646347761 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document -0.00043341209652360653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document -0.00023966604734537185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document -0.000259165907316014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document -0.0004270653021833602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document -0.0004341547032162028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document -0.0004111478117275994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document -0.0004299383567984396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document -0.0004241899124590779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document -0.0004502719349364145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document -0.00038994621469645615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document -0.0003859912398894952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document -0.0004247535950310557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document -0.000386982084327716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document -0.0004196451040053251 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document -0.0004096278509782259 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document -0.0004373334932695721 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document -0.0004180889975240641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document -0.00042079636929672745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document -0.00038063574611812913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document -0.0003817505891515542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document -0.0004420096268860222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document -0.00039182670726410623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document -0.0003635667850372299 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document -0.00041564996472055667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document -0.000400529358757286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document -0.0003939113874958451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document -0.00039066622068940996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document -0.0004290098538807143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document -0.0004240739958197099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document -0.00040775392659215333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document -0.0004091634200396925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document -0.00042299190476617914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document -0.0003701492680344151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document -0.0003807353844384635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document -0.00038813507771983156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document -0.00040072346558408346 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document -0.0003603595180423597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document -0.00038799421353112465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document -0.00037575235582264926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document -0.0004239190342959713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document -0.0004606044799136546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document -0.00045107950652529253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document -0.0004391947201871058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document -0.0004457516661123035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document -0.0004301297170991686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document -0.00044661704164586694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document -0.0004438849846114837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document -0.0004444205734316823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document -0.0004190924165303394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document -0.00043942581131677875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document -0.00021568459798090663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document -0.0003814929225407199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document -0.0003217453179359235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document -0.00031719591470267974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document -0.00032434115726922137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document -0.0004079911120371051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document -0.000329492766381148 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document -0.0003845916162001633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document -0.0003835208964390098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document -0.00037847334157173194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document -0.00038296039903791865 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document -0.00037896336828472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document -0.00037620974396391355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document -0.00037420590727111843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document -0.000340490625886403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document -0.0003078314411035827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document -0.00034153990750656097 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document -0.0003308858103982067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document -0.0003452640607156025 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document -0.00033095276418403455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document -0.0003116308995860414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document -0.00032446713226408477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document -0.0003015816821912984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document -0.00031612418775706894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document -0.0003278516344971041 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document -0.00033079446736097217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document -0.00032278977146550837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document -0.00032065272988207914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document -0.0003936696452406576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document -0.0003450109536627789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document -0.0003339787189919641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document -0.0003284303856176974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document -0.00033652677276843477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document -0.0003257822443845694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document -0.0003293985569149334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document -0.0003310360260148262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document -0.0003233770986418526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document -0.0003172280092149422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document -0.0003160674744292835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document -0.00030931090289598506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document -0.0003093173886443107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document -0.00033167847081104083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document -0.00031131501311729723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document -0.00031046608876279845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document -0.00030569235942207244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document -0.00030777943671285197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document -0.00029303314290956683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document -0.0003045824546400205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document -0.00030360880677729793 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document -0.00031646239964835433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document -0.0003129122300603785 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document -0.00031060464956661433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document -0.000311819032500067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document -0.0002977872483902282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document -0.0003009448600922438 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document -0.00028610292098537774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document -0.0002988326876216654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document -0.00028550828372819075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document -0.0002830381750875739 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document -0.0002848495855927156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document -0.0002856443760308144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document -0.00027442895344188584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document -0.0002681160554049462 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document -0.0003421482544126989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document -0.0004005872948449718 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document -0.0003930123959320308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document -0.0003867271832275778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document -0.000380805140455254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document -0.0003814769861947819 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document -0.00038025170883282324 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document -0.0003738026647867475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document -0.00018960856915036276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document -0.0003697177501953134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document -0.00036674194328136693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document -0.00036447406838697555 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document -0.00036686410861101255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document -0.00035915267825103423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document -0.0003624758404026675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document -0.0002822812140180794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document -0.00030620512946920813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document -0.000294249776520589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document -0.00030238536967523434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document -0.00029509593361580754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document -0.0002906912701830899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document -0.0002921944165474959 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document -0.00028358919691127954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document -0.0002813182772323272 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document -0.00027442640800299205 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document -0.0002747820342933984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document -0.0002747584403979717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document -0.00027499129634862444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document -0.0002712050404257197 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document -0.0002616256943143254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document -0.00026769938929002815 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document -0.00038396081322727017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document -0.0003863140490027991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document -0.00037702277513203237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document -0.0003633274156107032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document -0.0003587473889240435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document -0.0003507672084278415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document -0.00033776425499780385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document -0.0003377914127574796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document -0.00032948015659161326 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document -0.00033245638541392985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document -0.00031080707640648695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document -0.0002976903331149755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document -0.0002965121463725523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document -0.0002933849695266647 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document -0.0002837035078508233 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document -0.00028684569079589323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document -0.0003145192320802359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document -0.0003566937253273515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document -0.0003470199109592918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document -0.0003060245312041868 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document -0.0002650817213818789 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document -0.0002643604938780134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document -0.000299350876031416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document -0.0003178540797697938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document -0.000271850367887767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document -0.00031349896596549 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document -0.00031749734412765755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document -0.0003791137842391209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document -0.0003742334169957992 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document -0.0003705639757351107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document -0.0003126986769797042 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document -0.00031038132814561196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document -0.00036464437173804883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document -0.0003569480488951322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document -0.0003541239221619106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document -0.00035315297411308053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document -0.0003572451925404141 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document -0.0003514986129411253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document -0.0003521798298425866 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document -0.00034553677439244716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document -0.000349004719809412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document -0.0003468247484872769 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document -0.0003465822608356558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document -0.00035410983132162007 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document -0.0003487908354969444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document -0.0003479024763238147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document -0.000341412530646823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document -0.00034451316273667034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document -0.0002618849993484869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document -0.00026788679978901144 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document -0.00027450670773227214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document -0.0002661273129899329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document -0.00026836569676402957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document -0.00026155876975483236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document -0.0002609276830117151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document -0.0002644161630512771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document -0.00036789208972872557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document -0.00037829849439990513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document -0.0003788894943523098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document -0.0003617207777959397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document -0.0002541334487248998 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document -0.0002707945538071073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document -0.00027046282716455214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document -0.0002652443167243215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document -0.0002685859923850986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document -0.00025734961751176414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document -0.000259041720872915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document -0.00025340107274823446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document -0.00025757135121837893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document -0.00025617700500574084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document -0.0002566931670562857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document -0.0002543871190716101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document -0.00024997565589481713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document -0.0002954079779456287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document -0.00034890741135252835 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document -0.0003473298137731525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document -0.0003296959618486435 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document -0.0003304520061604598 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document -0.00032377956175729824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document -0.00031700696295168713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document -0.0003060382346081943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document -0.0003012003005056863 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document -0.0002981074073993884 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document -0.0002922128825950705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document -0.000348901087722931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document -0.0003408286289467841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document -0.0003410649680770183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document -0.0003358524215576502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document -0.0003343661874989231 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document -0.00032810573699389156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document -0.00032261449539097497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document -0.0003162694866049203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document -0.0003158381156468853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document -0.000317376061083603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document -0.0003125788639953052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document -0.0003010105041885602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document -0.0003065865059090678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document -0.0003084275726508053 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document -0.00030966560718296085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document -0.0002957728057853081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document -0.00029904164542325336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document -0.0002955358888729187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document -0.00028692976446931544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document -0.0002923476214935797 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document -0.0002893691697212419 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document -0.0002855895211981585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document -0.00027968347097626246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document -0.0002810783462604979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document -0.00027794080455729715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document -0.00034784376461416953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document -0.0003488347959010943 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document -0.00034790583710250724 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document -0.000345913166618151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document -0.00033801936268066675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document -0.0003290591130212315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document -0.00034051399521366823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document -0.00032470943131841784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document -0.00031679540050914276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document -0.00031814596342422325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document -0.0003156466289485036 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document -0.00029985010879003633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document -0.0002905176377776361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document -0.0004206836775460856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document -0.00020660449162246918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document -0.0003461727254468087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document -0.00020592870907067763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document -0.00034173505299233005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document -0.0004052437256652738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document -0.0004080650901351697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document -0.00039778184149144276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document -0.00039046311464950275 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document -0.00039043444911071384 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document -0.000388575704932843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document -0.00019737533145666597 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document -0.00037610755595812403 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document -0.00037315400127598317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document -0.00037415028580922163 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document -0.00036694041707212337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document -0.00018947219857306515 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document -0.00037046050826533545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document -0.0003587440768559087 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document -0.00034623936498708903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document -0.0003502289592617922 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document -0.00034692398063649823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document -0.000339340809421849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document -0.0003360510394816983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document -0.0003354673850814145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document -0.00032937682875877047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document -0.00032844505049317715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document -0.00028287199339908627 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document -0.0002795217197003578 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document -0.00028048955601883463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document -0.0002769326396439027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document -0.0002727090021299243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document -0.0002726577841024554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document -0.00026663619593455374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document -0.00026068042672138127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document -0.0002637704114326801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document -0.0002593043567100412 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document -0.0002599897110113453 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document -0.0002435078682758859 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document -0.0002450530071379054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document -0.00024233331983743606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document -0.0002934750947999535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document -0.00033241226364044474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document -0.00032938406090272075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document -0.00032778705403953246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document -0.00032184551480398754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document -0.00031874002264945737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document -0.0003165319685666433 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document -0.00031307071173376295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document -0.00031119524184911957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document -0.0003102253344576429 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document -0.0003088976240383192 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document -0.0002951410823077708 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document -0.00029772657676757413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document -0.0003056048989909935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document -0.00031991305381648026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document -0.00030890256978362426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document -0.0003109382904091933 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document -0.00031035798529690644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document -0.00030741666395911753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document -0.0002989918594861846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document -0.00029569635443989434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document -0.0002973992445667285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document -0.000293397351001072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document -0.00028737817438047954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document -0.00028252738144009747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document -0.0002805511898623541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document -0.0003718020784620472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document -0.0003499713845765235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document -0.00034283547445326676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document -0.00031464759888838765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document -0.00033188946446414833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document -0.000326084432195463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document -0.0003764568303917893 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document -0.0003604955598858414 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document -0.0003655654554133222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document -0.00035762304033750504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document -0.00038478883950347103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document -0.00027735714341247454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document -0.00028139534607773563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document -0.00019777292251713763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document -0.000285571704874486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document -0.00028543482146244363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document -0.00019434234484256758 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document -0.00027854908176986763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document -0.0002847068039566143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document -0.00028672356943064853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document -0.00027782687605808177 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document -0.0002843539634105203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document -0.0002894748379090401 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document -0.0002868852440186493 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document -0.0002818504885373851 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document -0.00028680112812941034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document -0.00019258978168723977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document -0.00028760637934715155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document -0.0002820439443912918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document -0.0002831001054410018 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document -0.00029001901552467397 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document -0.00027779449377883156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document -0.00019949837437516796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document -0.0002907306472984446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document -0.00027814858381318327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document -0.00019472790889161432 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document -0.00020472626596924125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document -0.0002870045081974301 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document -0.00019812241927078482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document -0.0002817553333369554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document -0.00027829782796642117 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document -0.00028289431732284113 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document -0.0002795526296717729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document -0.00027682829988044574 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document -0.0002895432402719184 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document -0.0002823174903941811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document -0.00028170972351837796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document -0.00027807915877838826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document -0.00028588515681452956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document -0.00028112324090816726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document -0.00020636178289985485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document -0.00019447255290980535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document -0.0002850824220591452 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document -0.00027856429520116784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document -0.0002820880676635633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document -0.00028943902215995714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document -0.0002676366291085329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document -0.00023806333809954687 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document -0.00024526460430233455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document -0.00023876876664622726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document -0.00023379770334179805 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document -0.00024175151269138382 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document -0.00023386583242595706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document -0.00023771797150160827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document -0.0002262748967483896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document -0.0002408148346432682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document -0.00023398651720444235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document -0.00022989433874474592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document -0.00023948500543957772 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document -0.0002331594076859196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document -0.00023375132439600242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document -0.00023923410909668642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document -0.00023952796315562954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document -0.0002327466076905069 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document -0.00023082758956797212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document -0.0002240509275524448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document -0.00022798879995765268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document -0.000221172516774386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document -0.00021767045123534623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document -0.00021982832794804484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document -0.00021971626543789102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document -0.00022566565206920132 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document -0.0002181984894194856 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document -0.00021831417549554653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document -0.00021601405421187145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document -0.00022275733725519607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document -0.00021847734911973986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document -0.0002243591012664014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document -0.00021688758139483833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document -0.0002182953624789215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document -0.00020475155724026002 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document -0.00021498078062960065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document -0.0002157914337233064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document -0.00021781838494967963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document -0.00021723242266814558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document -0.0002176782686553837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document -0.0003486179404943968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document -0.00034882846352857634 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document -0.00031400868448352596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document -0.00030273484020011963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document -0.00029895889118145404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document -0.00029770764609621714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document -0.0002990181332116852 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document -0.00029653733972285996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document -0.00029624649222942476 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document -0.00029625609720203576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document -0.00029731928930852147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document -0.00029011721326148513 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document -0.00028849788197494655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document -0.00021601278623858145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document -0.00021319599281739178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document -0.0002153325290600083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document -0.00018566946174516558 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document -0.00020736824394291617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document -0.00020857419820128004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document -0.00020058526129536423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document -0.00020745812166665217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document -0.00020652171015271702 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document -0.00020643808911278608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document -0.00020040513914482103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document -0.00020598050188272898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document -0.0001969184139343296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document -0.0001972748812937012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document -0.0002038556751586195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document -0.00020245186011313464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document -0.00019950381422038783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document -0.00020837055459665258 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document -0.00020371856218246096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document -0.00019537612301625791 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document -0.00019914984508813857 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document -0.0002053787713691309 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document -0.00019082100541008637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document -0.00020397153334531813 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document -0.0002021462693077317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document -0.00019609357008124035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document -0.00019693256622486236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document -0.00020007239732428112 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document -0.00020467075741591954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document -0.00019584883400022932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document -0.00019135050391176972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document -0.0003362829834208298 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document -0.00034013691154784095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document -0.00033215887031941976 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document -0.00032681189065396707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document -0.0003149138485493094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document -0.00030179177307540077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document -0.0002923278437581119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document -0.00029470052278994486 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document -0.0002994095093045731 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document -0.00029033525096085037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document -0.00029390798852496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document -0.0002916230924130842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document -0.00029419886374594913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document -0.0002865469756730764 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document -0.00021191292549942086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document -0.00021369664817409847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document -0.00021612485624266726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document -0.00022242192634588478 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document -0.00014605095659989698 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document -0.00022070626106341693 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document -0.0002174420774054071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document -0.00021325858963116995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document -0.0002124322999488052 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document -0.0002081218896969054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document -0.0002108710211556957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document -0.00020686867095978426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document -0.00020895752681041895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document -0.00020741922266415738 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document -0.0002069112657197308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document -0.00020644627473468118 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document -0.00020332991338121604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document -0.0003560895677789848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document -0.00032915779111908214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document -0.00033810613317040864 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document -0.00033729626594036923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document -0.00033550342864602944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document -0.00034173474024556906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document -0.000331505340748827 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document -0.0003270050330117195 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document -0.00032585275329172556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document -0.0003143383203190604 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document -0.00031655199110388894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document -0.00030738872158476413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document -0.00030838388352699285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document -0.0003053596995351888 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document -0.00031836304739584593 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document -0.000315315435873905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document -0.0003087116248965243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document -0.00030396790625537645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document -0.0003335812246032149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document -0.00034570956323095843 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document -0.00034563035636675786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document -0.00033411265479076335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document -0.00034439191141692787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document -0.0003364483125496565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document -0.0003299500453608033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document -0.00033163377700074837 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document -0.00032638649660627673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document -0.00032616167939645234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document -0.0003205289298760723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document -0.00031939393740815355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document -0.00031593164066731296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document -0.00031928871111254405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document -0.00029670189073175004 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document -0.00020517703846735904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document -0.00020128418186172073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document -0.00019662723895606717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document -0.0001981157042081407 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document -0.00019703489037041608 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document -0.00019079796331785068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document -0.0001909352306690079 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document -0.00018824662295261396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document -0.00019864275319325954 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document -0.00018818516521649587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document -0.00018875694972812844 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document -0.00018231621170645482 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document -0.00018349407845798273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document -0.00018088971427746906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document -0.00018296284236327237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document -0.0001876011825819916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document -0.000329052068725176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document -0.00032223616273648536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document -0.00031272564089633955 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document -0.00031621609908414494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document -0.0003117213560911235 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document -0.00030218064069945934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document -0.00030658916600512085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document -0.0002915863534115821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document -0.0002940280138374372 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document -0.00029067860468866085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document -0.00028529228063135635 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document -0.00028336893301452256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document -0.0002794668089130099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document -0.00021681361378827842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document -0.0001484664674497246 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document -0.00021950558378215133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document -0.00021806860758808645 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document -0.00021819568718852282 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document -0.00021626925931585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document -0.0001464536143077762 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document -0.00021432777088808917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document -0.000213473805865147 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document -0.00021397067253964538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document -0.00020758957647437263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document -0.00020687124337683314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document -0.00020630057046511005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document -0.0002091166859352538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document -0.00020777355025615267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document -0.00020709287641496176 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document -0.00020736464660577094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document -0.00020062246741862607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document -0.00020693207561942915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document -0.00021151004871893024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document -0.00019930249098689716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document -0.00021589710041231824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document -0.00021369204789905741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document -0.0002147099923936778 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document -0.00021077531190389536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document -0.0002100509829113836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document -0.00021185362601571124 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document -0.00020722136637339565 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document -0.00020300093701169531 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document -0.00019859737993313477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document -0.00019971314372100164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document -0.00019549908270269278 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document -0.00019649820843534028 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document -0.00019619415513498067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document -0.00019493006120377898 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document -0.00019499409035775506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document -0.00019252988593634277 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document -0.00019440768268686405 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document -0.00018747161324755577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document -0.0001879575932372779 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document -0.00019040707058357506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document -0.0001871931095090703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document -0.00020112966223017096 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document -0.00020516878165311017 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document -0.00020664735191740533 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document -0.00021041398572882962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document -0.00020397992929690396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document -0.0002039978580295561 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document -0.00020592785601142126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document -0.0001990755527445265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document -0.00019729564847798732 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document -0.00019958182230527032 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document -0.0001985037302636386 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document -0.00020204130355115716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document -0.0002000296401958085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document -0.0001983064832295463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document -0.00019663108484195617 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document -0.00019510678560556523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document -0.0001873284057063206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document -0.00019311553072495885 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document -0.00034652137288816547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document -0.0002813690318850024 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document -0.00027697649713138685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document -0.0002755419092534421 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document -0.0002681583054440219 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document -0.00026945753192750824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document -0.00026169470768245737 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document -0.00026437008960810825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document -0.0002637294838228 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document -0.00026491867965088836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document -0.00025504483625138986 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document -0.0002545040623796586 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document -0.0002546682814073622 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document -0.00025545439487142615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document -0.0002626896557978271 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document -0.00025092040940402784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document -0.0002589154885863872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document -0.00024106160482721467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document -0.0002483289690087987 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document -0.0002388930282784437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document -0.00024006340759273874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document -0.00023765248178029045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document -0.00023061351965578936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document -0.00024954224883546477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document -0.00017861017233018525 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document -0.00017810832743667658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document -0.00017599709170759497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document -0.00017462723516505223 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document -0.0002906316527068669 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document -0.00033762141066247166 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document -0.00017170670574152494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document -0.00017258674515137717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document -0.0002815386173173926 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document -0.0002996845935618989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document -0.0002735268488987296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document -0.0002971738713071517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document -0.0002942690674002763 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document -0.0003322222207729567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document -0.0003378721656198464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document -0.00018307262621851067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document -0.00033956081502775057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document -0.00031604820927876276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document -0.00028805657681088917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document -0.00026312293321215633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document -0.00034366936722921455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document -0.0002865256504406559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document -0.0003063615195861786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document -0.00028412791619666136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document -0.00028060835132727154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document -0.00032544974761560506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document -0.0002647177833217225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document -0.0003152621884896575 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document -0.0003054625140336913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document -0.00031183308312292263 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document -0.00018175026696621178 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document -0.00017699918328872 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document -0.00018222339261441908 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document -0.00018348005930964137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document -0.0001810735993810541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document -0.00030846441282038914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document -0.0002972326889310354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document -0.00017433421318235594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document -0.00032799458649525895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document -0.00032482130048512673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document -0.00031943465668672475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document -0.00029615593630484517 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document -0.0002893126939511001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document -0.0002849288351723284 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document -0.00028383906633569267 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document -0.00028072526091262615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document -0.000284239564292377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document -0.0002778903109432523 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document -0.0002771644389501471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document -0.0002733316182319337 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document -0.00026362539185869363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document -0.0002636325383220217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document -0.00026740622442302886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document -0.0002646771971853427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document -0.0002628566720605389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document -0.0002644760695434766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document -0.0002623837702310999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document -0.00026088722976772894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document -0.0002567065374799158 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document -0.00018857382101207726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document -0.00019036580399817203 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document -0.00018348828065261222 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document -0.00018491851780345073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document -0.00018904887260080187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document -0.0001875609304251801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document -0.00018393034720015817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document -0.00018419795526114903 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document -0.00018699955623404795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document -0.00018276256902965128 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document -0.00017698045695190812 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document -0.00018104650132303642 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document -0.00017758206731279688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document -0.00017131402995103497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document -0.000175944428350446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document -0.0003416745727147391 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document -0.0003163259373952889 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document -0.0002804489269172448 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document -0.00028748272397403175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document -0.00027603318345630605 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document -0.000271638824679648 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document -0.0002763761210210942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document -0.00026501984873172717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document -0.00026422486894694714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document -0.0002686339100849262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document -0.0002610837453940606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document -0.000260974343729353 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document -0.0002599403837029134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document -0.0002937273113238609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document -0.0003341790732600504 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document -0.0002620661576600244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document -0.0003027929169239288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document -0.00031944039129326894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document -0.00019025676304139009 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document -0.00018680910145009907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document -0.00034215840419416437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document -0.00018618120812119364 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document -0.00018605853095599425 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document -0.00018120712626096538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document -0.00018315079292495327 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document -0.00018362556449041974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document -0.0001780024456718171 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document -0.00033296526436178697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document -0.0001802398632282846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document -0.00017340263100798256 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document -0.00017755840547238697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document -0.00018419413735260606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document -0.00017869518174591322 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document -0.00017526271460129484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document -0.00017852168597981907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document -0.00017566536156787157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document -0.00017589867964432936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document -0.00017831487394075305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document -0.00017837310528935862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document -0.00018200908814216548 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document -0.0001795136627511612 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document -0.0003414021775300033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document -0.00017177291787788502 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document -0.0003441900648571877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document -0.0003394534597060673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document -0.0003236887233114832 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document -0.0001639544129688747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document -0.00019137443753211255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document -0.00018575146284680153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document -0.00019184792863440243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document -0.00018966043065679055 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document -0.00017968851317035848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document -0.00018479881897661546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document -0.0001813642692683015 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document -0.0001686449798983066 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document -0.00018516104592230446 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document -0.00031283726601066385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document -0.0003248607542883853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document -0.00031583241601202365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document -0.00031238270857730376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document -0.000307150592403979 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document -0.00029443829986847044 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document -0.0002942723732234677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document -0.00023514930666443422 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document -0.0020776328951453444 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document -0.0021768234410538883 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document -0.002106973549276289 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document -0.002110915756171751 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document -0.0017032382109816464 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document -0.0019047944877712286 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document -0.0019402711744016077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document -0.0006264790011223686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document -0.0017885401938106643 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document -0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document -0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document -0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document -0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document -0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document -0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document -0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document -0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document -0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document -0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document -0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document -0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document -0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document -0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document -0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document -0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document -0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document -0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document -0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document -0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document -0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document -0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document -0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document -0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document -0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document -0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document -0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document -0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document -0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document -0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document -0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document -0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document -0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document -0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document -0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document -0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document -0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document -0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document -0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document -0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document -0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document -0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document -0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document -0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document -0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document -0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document -0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document -0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document -0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document -0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document -0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document -0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document -0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document -0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document -0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document -0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document -0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document -0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document -0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document -0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document -0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document -0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document -0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document -0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document -0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document -0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document -0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document -0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document -0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document -0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document -0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document -0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document -0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document -0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document -0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document -0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document -0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document -0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document -0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document -0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document -0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document -0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document -0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document -0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document -0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document -0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document -0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document -0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document -0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document -0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document -0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document -0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document -0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document -0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document -0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document -0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document -0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document -0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document -0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document -0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document -0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document -0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document -0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document -0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document -0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document -0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document -0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document -0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document -0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document -0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document -0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document -0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document -0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document -0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document -0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document -0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document -0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document -0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document -0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document -0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document -0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document -0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document -0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document -0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document -0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document -0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document -0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document -0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document -0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document -0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document -0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document -0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document -0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document -0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document -0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document -0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document -0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document -0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document -0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document -0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document -0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document -0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document -0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document -0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document -0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document -0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document -0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document -0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document -0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document -0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document -0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document -0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document -0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document -0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document -0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document -0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document -0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document -0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document -0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document -0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document -0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document -0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document -0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document -0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document -0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document -0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document -0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document -0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document -0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document -0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document -0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document -0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document -0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document -0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document -0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document -0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document -0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document -0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document -0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document -0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document -0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document -0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document -0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document -0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document -0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document -0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document -0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document -0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document -0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document -0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document -0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document -0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document -0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document -0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document -0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document -0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document -0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document -0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document -0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document -0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document -0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document -0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document -0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document -0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document -0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document -0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document -0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document -0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document -0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document -0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document -0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document -0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document -0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document -0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document -0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document -0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document -0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document -0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document -0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document -0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document -0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document -0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document -0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document -0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document -0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document -0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document -0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document -0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document -0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document -0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document -0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document -0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document -0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document -0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document -0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document -0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document -0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document -0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document -0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document -0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document -0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document -0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document -0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document -0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document -0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document -0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document -0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document -0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document -0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document -0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document -0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document -0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document -0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document -0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document -0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document -0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document -0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document -0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document -0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document -0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document -0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document -0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document -0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document -0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document -0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document -0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document -0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document -0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document -0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document -0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document -0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document -0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document -0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document -0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document -0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document -0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document -0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document -0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document -0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document -0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document -0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document -0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document -0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document -0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document -0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document -0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document -0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document -0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document -0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document -0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document -0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document -0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document -0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document -0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document -0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document -0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document -0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document -0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document -0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document -0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document -0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document -0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document -0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document -0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document -0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document -0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document -0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document -0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document -0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document -0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document -0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document -0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document -0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document -0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document -0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document -0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document -0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document -0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document -0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document -0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document -0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document -0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document -0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document -0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document -0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document -0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document -0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document -0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document -0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document -0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document -0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document -0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document -0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document -0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document -0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document -0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document -0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document -0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document -0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document -0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document -0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document -0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document -0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document -0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document -0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document -0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document -0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document -0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document -0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document -0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document -0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document -0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document -0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document -0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document -0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document -0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document -0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document -0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document -0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document -0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document -0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document -0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document -0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document -0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document -0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document -0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document -0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document -0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document -0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document -0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document -0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document -0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document -0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document -0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document -0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document -0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document -0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document -0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document -0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document -0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document -0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document -0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document -0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document -0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document -0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document -0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document -0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document -0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document -0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document -0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document -0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document -0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document -0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document -0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document -0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document -0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document -0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document -0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document -0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document -0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document -0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document -0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document -0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document -0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document -0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document -0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document -0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document -0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document -0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document -0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document -0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document -0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document -0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document -0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document -0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document -0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document -0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document -0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document -0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document -0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document -0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document -0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document -0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document -0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document -0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document -0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document -0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document -0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document -0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document -0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document -0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document -0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document -0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document -0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document -0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document -0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document -0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document -0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document -0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document -0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document -0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document -0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document -0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document -0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document -0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document -0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document -0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document -0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document -0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document -0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document -0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document -0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document -0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document -0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document -0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document -0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document -0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document -0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document -0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document -0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document -0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document -0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document -0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document -0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document -0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document -0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document -0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document -0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document -0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document -0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document -0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document -0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document -0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document -0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document -0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document -0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document -0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document -0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document -0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document -0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document -0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document -0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document -0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document -0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document -0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document -0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document -0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document -0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document -0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document -0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document -0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document -0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document -0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document -0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document -0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document -0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document -0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document -0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document -0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document -0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document -6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document -2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document -5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document -2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document -2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document -2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document -5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document -1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document -1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document -1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document -1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document -1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document -1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document -1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document -1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document -1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document -1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document -1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document -1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document -1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document -1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document -1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document -1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document -1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document -1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document -1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document -1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document -1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document -1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document -1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document -1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document -1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document -1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document -1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document -1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document -1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document -1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document -1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document -1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document -1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document -1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document -1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document -1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document -1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document -1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document -1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document -1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document -1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document -1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document -1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document -1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document -1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document -5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document -5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document -5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document -5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document -5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document -5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document -5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document -4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document -4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document -5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document -4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document -5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document -6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document -5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document -4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document -5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document -5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document -4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document -5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document -5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document -5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document -5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document -5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document -6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document -5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document -5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document -5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document -5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document -5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document -5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document -5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document -4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document -5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document -5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document -4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document -5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document -5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document -5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document -4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document -5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document -4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document -5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document -5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document -6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document -5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document -5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document -5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document -5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document -5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document -5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document -4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document -5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document -5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document -5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document -5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document -5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document -5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document -5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document -5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document -5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document -5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document -5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document -2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document -1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document -3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document -3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document -2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document -2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document -2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document -2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document -2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document -2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document -2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document -2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document -2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document -2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document -2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document -2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document -2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document -1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document -1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document -2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document -2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document -9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document -2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document -1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document -2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document -2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document -4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document -2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document -2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document -1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document -1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document -1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document -1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document -1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document -1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document -1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document -1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document -1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document -1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document -1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document -1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document -1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document -1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document -1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document -1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document -1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document -1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document -1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document -1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document -1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document -1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document -1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document -1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document -8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document -8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document -9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document -9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document -7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document -8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document -9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document -8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document -8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document -8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document -7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document -9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document -8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document -1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document -3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document -2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document -1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document -1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document -1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document -2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document -2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document -8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document -1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document -1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document -1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document -1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document -1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document -1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document -1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document -2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document -1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document -2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document -5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document -1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document -1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document -1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document -1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document -1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document -1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document -1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document -1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document -1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document -1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document -1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document -1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document -1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document -1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document -1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document -1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document -1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document -1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document -9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document -1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document -1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document -1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document -1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document -1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document -1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document -1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document -1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document -1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document -1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document -4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document -1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document -1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document -2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document -2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document -2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document -2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document -1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document -1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document -3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document -6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document -1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document -1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document -1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document -2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document -3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document -1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document -1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document -1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document -1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document -1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document -1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document -1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document -1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document -1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document -1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document -4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document -3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document -3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document -3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document -3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document -3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document -7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document -0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document -0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document -0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document -0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document -0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document -6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document -0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document -0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document -0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document -0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document -0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document -0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document -0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document -0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document -0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document -0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document -0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document -0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document -0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document -0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document -0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document -0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document -0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document -0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document -0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document -0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document -0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document -0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document -0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document -0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document -0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document -0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document -0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document -0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document -0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document -0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document -0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document -0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document -0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document -0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document -0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document -0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document -0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document -0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document -0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document -0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document -0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document -0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document -0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document -0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document -0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document -0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document -0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document -0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document -0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document -0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document -0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document -0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document -0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document -0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document -0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document -0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document -0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document -0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document -0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document -0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document -0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document -0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document -0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document -0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document -0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document -0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document -0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document -0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document -0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document -0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document -0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document -0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document -0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document -0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document -0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document -0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document -0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document -0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document -0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document -0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document -0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document -0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document -0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document -0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document -0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document -0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document -0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document -0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document -0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document -0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document -0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document -0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document -0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document -0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document -0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document -0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document -0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document -0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document -0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document -0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document -0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document -0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document -0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document -0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document -0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document -0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document -0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document -0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document -0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document -0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document -0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document -0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document -0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document -0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document -0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document -0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document -0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document -0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document -0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document -0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document -0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document -0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document -0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document -0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document -0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document -0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document -0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document -0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document -0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document -0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document -0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document -0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document -0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document -0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document -0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document -0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document -0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document -0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document -0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document -0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document -0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document -0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document -0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document -0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document -0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document -0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document -0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document -0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document -0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document -0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document -0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document -0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document -0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document -0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document -0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document -0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document -0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document -0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document -0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document -0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document -0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document -0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document -0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document -0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document -0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document -0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document -0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document -0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document -0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document -0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document -0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document -0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document -0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document -0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document -0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document -0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document -0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document -0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document -0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document -0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document -0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document -0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document -0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document -0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document -0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document -0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document -0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document -0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document -0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document -0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document -0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document -0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document -0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document -0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document -0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document -0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document -0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document -0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document -0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document -0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document -0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document -0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document -0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document -0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document -0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document -0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document -0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document -0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document -0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document -0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document -0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document -0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document -0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document -0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document -0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document -0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document -0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document -0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document -0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document -0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document -0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document -0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document -0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document -0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document -0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document -0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document -0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document -0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document -0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document -0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document -0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document -0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document -0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document -0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document -0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document -0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document -0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document -0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document -0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document -0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document -0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document -0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document -0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document -0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document -0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document -0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document -0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document -0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document -0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document -0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document -0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document -3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document -0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document -0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document +0.0018520780893211373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0000_text_document +0.0017591050606817512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0001_text_document +0.001459052794333798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0002_text_document +0.0007405667281569194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0003_text_document +0.00019420030110896795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0004_text_document +0.0009008668715801845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0005_text_document +0.00015115827957143057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0006_text_document +0.0014552844319220648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0007_text_document +0.0012469861325685161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0008_text_document +0.00136412011372413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0009_text_document +0.0007064279699221103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0010_text_document +0.0008472240000687427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0011_text_document +0.0001984375713341955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0012_text_document +0.0005472773881697123 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0013_text_document +0.001815779629850992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0014_text_document +0.0018313600689757324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/algebraic-stack-train-0015_text_document +0.0002583902668716813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0000_text_document +0.0002646575141232155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0001_text_document +0.0003165521247456758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0002_text_document +0.0002920706460176214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0003_text_document +0.00028396813182810215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0004_text_document +0.00030445161883108107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0005_text_document +0.00031628781276576474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0006_text_document +0.0003083776568189157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0007_text_document +0.0003176359471472902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0008_text_document +0.0002536009369131698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0009_text_document +0.0003067491424681363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0010_text_document +0.0002597217257557784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0011_text_document +0.0003788556450109768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0012_text_document +0.0002796563272052598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0013_text_document +0.00033573826524290287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0014_text_document +0.00030523658022800287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0015_text_document +0.00032211552192240096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0016_text_document +0.0003329295675164247 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0017_text_document +0.0003101982186639862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0018_text_document +0.00032361798234223355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0019_text_document +0.0003495541581652915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0020_text_document +0.0002821637448858042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0021_text_document +0.00030399523537629673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0022_text_document +0.0002955658968247219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0023_text_document +0.00028942158502924254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0024_text_document +0.00028769546171490733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0025_text_document +0.0002938111057234182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0026_text_document +0.0002711150403010948 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0027_text_document +0.00031130095874747565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0028_text_document +0.0003002996118160777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0029_text_document +0.0003732757901604459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0030_text_document +0.00026784205751795894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0031_text_document +0.0002799626521661984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0032_text_document +0.00034334276069078164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0033_text_document +0.0003582469803674965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0034_text_document +0.00031094844818418623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0035_text_document +0.0002766228384977191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0036_text_document +0.00030297116159471485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0037_text_document +0.00027033888377464685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0038_text_document +0.00030090862368377933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0039_text_document +0.00028543875802490955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0040_text_document +0.00027559768459074204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0041_text_document +0.0003182185533962886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0042_text_document +0.0003311392971435837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0043_text_document +0.00028751652060804325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0044_text_document +0.000303466863212589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0045_text_document +0.00033400462801277524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0046_text_document +0.0002589234031777426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0047_text_document +0.0002913508598466723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0048_text_document +0.0002670572450004856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0049_text_document +0.00032027399105647656 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0050_text_document +0.00032188376258379377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0051_text_document +0.0003161585784100882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0052_text_document +0.0003184249182974135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0053_text_document +0.00030381336664000807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0054_text_document +0.0003190437442184283 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0055_text_document +0.0002537961798200545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0056_text_document +0.0003017817117223326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0057_text_document +0.00028685268513240224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0058_text_document +0.00031265179094451165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0059_text_document +0.00034708319096986816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0060_text_document +0.00026650837943080664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0061_text_document +0.00034588832248507335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0062_text_document +0.0002416982248399037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0063_text_document +0.0003089296918222243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0064_text_document +0.00029137184185700827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0065_text_document +0.00026464226846800774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0066_text_document +0.00030545397919456627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0067_text_document +0.0003206778460448875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0068_text_document +0.00030968971641110967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0069_text_document +0.00023325653928600864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0070_text_document +0.00030526899198338555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0071_text_document +0.00035376719076633584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0072_text_document +0.000290224385981026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0073_text_document +0.000294650083382008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0074_text_document +0.00028768858128616436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0075_text_document +0.00030856965235527843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0076_text_document +0.00030579942447879054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0077_text_document +0.0002863101084704357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0078_text_document +0.0002870032092492213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0079_text_document +0.000264182727569885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0080_text_document +0.0002974012367036449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0081_text_document +0.00032238412143059203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0082_text_document +0.00031683716893819036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0083_text_document +0.00031157434937617524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0084_text_document +0.0003411742735695989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0085_text_document +0.00026778444816570715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0086_text_document +0.0003037045797275201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0087_text_document +0.00027746114370081314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0088_text_document +0.00027148285946862043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0089_text_document +0.00028042950114678207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0090_text_document +0.0003235607816590721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0091_text_document +0.0003086692227306295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0092_text_document +0.00033990349455148105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0093_text_document +0.00030945053208470265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0094_text_document +0.00027309074552265303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0095_text_document +0.00028737393506316194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0096_text_document +0.0003098868328009879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0097_text_document +0.0002614229162588409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0098_text_document +0.0002884388407820923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/arxiv-0099_text_document +0.0031025147279277244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document +0.003102019887362634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document +0.0009996745994661548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document +0.0002406272620255565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0000_text_document +0.0002404825539493424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0001_text_document +0.00024062296575435581 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0002_text_document +0.00024069315766818953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0003_text_document +0.00024055829162263452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0004_text_document +0.00024062053397343032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0005_text_document +0.0002410715545206964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0006_text_document +0.00024024881846087368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0007_text_document +0.0002407074700790688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0008_text_document +0.00024072141428809043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0009_text_document +0.00024027710230872736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0010_text_document +0.0002409111299205489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0011_text_document +0.00024081954058275009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0012_text_document +0.00024086076794990912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0013_text_document +0.00024098672620832446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0014_text_document +0.00024068622303333862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0015_text_document +0.00024140627024291824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0016_text_document +0.0002414512033594384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0017_text_document +0.00024028742594941463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0018_text_document +0.00024018036089269645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0019_text_document +0.0002398347365034979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0020_text_document +0.00024006780153485276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0021_text_document +0.00024015620270419213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0022_text_document +0.0002408848259695227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0023_text_document +0.0002408023185278831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0024_text_document +0.00024021196580140326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0025_text_document +0.00024077677271297493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0026_text_document +0.00024087392454668027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0027_text_document +0.0002408071293824126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0028_text_document +0.00024042223828845715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0029_text_document +0.0002411484752360495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0030_text_document +0.00023605263746465907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0031_text_document +0.00023471222158326908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0032_text_document +0.00023432138580287644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0033_text_document +0.00023407385623382327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0034_text_document +0.00023487504174367091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0035_text_document +0.0002341843704976313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0036_text_document +0.00023421993170282486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0037_text_document +0.00023445057969132037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0038_text_document +0.0002337681680073047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0039_text_document +0.000234627964808109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0040_text_document +0.0002338942211888584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0041_text_document +0.00023403849286843386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0042_text_document +0.00023405641310796305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0043_text_document +0.00023349169562397965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0044_text_document +0.00023381157386048856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0045_text_document +0.00023388742993790587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0046_text_document +0.00023363103829469813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0047_text_document +0.00023421141834630477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0048_text_document +0.00023420564352232565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0049_text_document +0.00023367463699173143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0050_text_document +0.00023344969163567033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0051_text_document +0.00023372196941547188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0052_text_document +0.00023399207645297834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0053_text_document +0.00023357915605505856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0054_text_document +0.00023337585642190864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0055_text_document +0.00023385005470157914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0056_text_document +0.00023301533534493465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0057_text_document +0.00023377864302541782 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0058_text_document +0.00023323745848621437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0059_text_document +0.0002330594611151835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0060_text_document +0.0002334149675026783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0061_text_document +0.00023198945902291534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0062_text_document +0.00023023784834634142 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0063_text_document +0.00022985623060187217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0064_text_document +0.0002292605284569516 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0065_text_document +0.00022926593333048894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0066_text_document +0.00022922766406807777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0067_text_document +0.00022898153911167426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0068_text_document +0.0002292473111593315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0069_text_document +0.000228804579400424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0070_text_document +0.00022865485613513526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0071_text_document +0.00022937426835887895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0072_text_document +0.00022917388311587372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0073_text_document +0.0002291660582019043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0074_text_document +0.00022907895248360543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0075_text_document +0.0002294617879920205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0076_text_document +0.0002290452150516566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0077_text_document +0.00022943405619715553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0078_text_document +0.0002296271421006204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0079_text_document +0.00022854791372910372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0080_text_document +0.00022923123467686557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0081_text_document +0.00022852404355738494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0082_text_document +0.00022847798660086642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0083_text_document +0.0002289604586810316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0084_text_document +0.00022835479834950643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0085_text_document +0.0002289149402884243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0086_text_document +0.00022806655474763446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0087_text_document +0.00022826296420992974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0088_text_document +0.00022906829636213627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0089_text_document +0.0002287628414466998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0090_text_document +0.0002282673911253445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0091_text_document +0.00022869309841939134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0092_text_document +0.0002281540116815451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0093_text_document +0.0002259755756162738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0094_text_document +0.00022562331285233504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0095_text_document +0.0002259061146106053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0096_text_document +0.00022567670836663787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0097_text_document +0.00022573165387587061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0098_text_document +0.00022508514961670572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0099_text_document +0.00022564642513773356 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0100_text_document +0.00022563088621998788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0101_text_document +0.0002250438755373707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0102_text_document +0.00022524465346241134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0103_text_document +0.00022531737657666812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0104_text_document +0.00022444687519363458 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0105_text_document +0.00022460397498596298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0106_text_document +0.00022454218976501763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0107_text_document +0.00022447528843671366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0108_text_document +0.00022501666332178926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0109_text_document +0.00022453752304377972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0110_text_document +0.00022484451871163002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0111_text_document +0.00022465678847154914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0112_text_document +0.00022453180917044732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0113_text_document +0.0002247278486823009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0114_text_document +0.00022465794828242097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0115_text_document +0.00022431000701925386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0116_text_document +0.00022476020248460963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0117_text_document +0.00022467531771795015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0118_text_document +0.0002236391309945234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0119_text_document +0.00022458764920536007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0120_text_document +0.00022430877426744415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0121_text_document +0.0002247047786127192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0122_text_document +0.0002245298090400035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0123_text_document +0.0002245648831396188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0124_text_document +0.00022292894729820784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0125_text_document +0.00022236668082957533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0126_text_document +0.0002217622659895442 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0127_text_document +0.00022252452726732609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0128_text_document +0.00022135333211363678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0129_text_document +0.0002214571757787971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0130_text_document +0.0002217188139237798 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0131_text_document +0.00022144214894640303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0132_text_document +0.00022100172806631854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0133_text_document +0.00022156392409199052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0134_text_document +0.00022134830143710272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0135_text_document +0.00022158598922529453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0136_text_document +0.00022142932483041377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0137_text_document +0.00022120980907786554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0138_text_document +0.00022117917738112441 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0139_text_document +0.00022077089397851235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0140_text_document +0.00022093265074996711 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0141_text_document +0.00022091299741377004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0142_text_document +0.0002205849150703338 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0143_text_document +0.0002210648204787979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0144_text_document +0.0002214235747364102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0145_text_document +0.00022083907302221787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0146_text_document +0.0002206334237915964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0147_text_document +0.00022065193929912214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0148_text_document +0.00022079775597767288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0149_text_document +0.00022091492909963518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0150_text_document +0.00022095009987097293 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0151_text_document +0.0002208150577180165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0152_text_document +0.00022085759102772088 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0153_text_document +0.00022073789170129016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0154_text_document +0.00022049322781182384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0155_text_document +0.00022083270617761285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0156_text_document +0.00021982452827473632 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0157_text_document +0.00021899870446514259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0158_text_document +0.00021890358773356361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0159_text_document +0.00021875556609042841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0160_text_document +0.00021861195987201226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0161_text_document +0.00021856782186167455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0162_text_document +0.00021912837771543515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0163_text_document +0.00021900213768517756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0164_text_document +0.00021871675851390374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0165_text_document +0.0002180537056545586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0166_text_document +0.0002188196714327129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0167_text_document +0.00021851362624523464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0168_text_document +0.0002183236795498736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0169_text_document +7.291153618675672e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/c4-0170_text_document +0.0003742481815405742 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0000_text_document +0.00038204855962733055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0001_text_document +0.00038821818392663593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0002_text_document +0.00038723332988783727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0003_text_document +0.00038916141142149904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0004_text_document +0.00038049542523949033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0005_text_document +0.0003854755539534284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0006_text_document +0.00024202756466512517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0007_text_document +0.0003915405155008087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0008_text_document +0.0003927382151931033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0009_text_document +0.0003839151202260479 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0010_text_document +0.00040006817468967907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0011_text_document +0.00040318965964443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0012_text_document +0.0003831013019452741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0013_text_document +0.00039166638383204036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0014_text_document +0.00039962784023961004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0015_text_document +0.00039536707853602614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0016_text_document +0.0004204304698247758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0017_text_document +0.00041538899178693555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0018_text_document +0.00039186953333675306 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0019_text_document +0.00038945837196504305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0020_text_document +0.0003919951238929062 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0021_text_document +0.00044377065718528966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0022_text_document +0.0004407759068603017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0023_text_document +0.0002487811895843715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0024_text_document +0.00039349432045556636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0025_text_document +0.00041223198559462343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0026_text_document +0.0004036573014830213 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0027_text_document +0.0003825982215521807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0028_text_document +0.00040386867133151386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0029_text_document +0.00024460575279105167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0030_text_document +0.000269029789531335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0031_text_document +0.0003573757493252864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0032_text_document +0.0004600876681392076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0033_text_document +0.0002605354166397086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0034_text_document +0.0003882502452157999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0035_text_document +0.0002466747612126512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0036_text_document +0.0004024726105072402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0037_text_document +0.00040820631128483644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0038_text_document +0.0002691094350403538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0039_text_document +0.00026916830387277267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0040_text_document +0.0004204663297880574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0041_text_document +0.00042379698687085554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0042_text_document +0.0004502169227311871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0043_text_document +0.0002661708937015295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0044_text_document +0.00031239486948031334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0045_text_document +0.0003109054589936201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0046_text_document +0.00045873053079760646 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0047_text_document +0.00022904931423244635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0048_text_document +0.0003813462028433663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0049_text_document +0.00039188129256500874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0050_text_document +0.00045124222276983765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0051_text_document +0.00048138658436853695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0052_text_document +0.0003944178776279866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0053_text_document +0.00039941569676754006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0054_text_document +0.00037952761190240494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0055_text_document +0.0003944870860881476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0056_text_document +0.0003891842411856621 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0057_text_document +0.000387688981934861 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0058_text_document +0.00039197953876258005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0059_text_document +0.00039007915280311206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0060_text_document +0.0003995520363699188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0061_text_document +0.00039230985654592406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0062_text_document +0.0003929472067173851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0063_text_document +0.0003924096172671473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0064_text_document +0.0003881636143629905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0065_text_document +0.000389790617937084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0066_text_document +0.00037351762309221023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0067_text_document +0.0003630196170929407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0068_text_document +0.00033532465765142113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0069_text_document +0.0003076088685761823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0070_text_document +0.00039463850897720803 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0071_text_document +0.0002843816115231449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0072_text_document +0.0002909175709416474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0073_text_document +0.00028867170997202486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0074_text_document +0.0002838644617723659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0075_text_document +0.00029027869525543416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0076_text_document +0.0002821339567560056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0077_text_document +0.0002922988877045601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0078_text_document +0.0002866955958315786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0079_text_document +0.0002865271754558126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0080_text_document +0.0002861247475618473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0081_text_document +0.0002826681072408606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0082_text_document +0.0002849746458282827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0083_text_document +0.0002816966633435316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0084_text_document +0.00026255342235948463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0085_text_document +0.0002552895098829678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0086_text_document +0.00025990194083107813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0087_text_document +0.0002524062657685835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0088_text_document +0.0002538577379748611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0089_text_document +0.0002561415177406761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0090_text_document +0.00026206253059694905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0091_text_document +0.00026168095406910565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0092_text_document +0.0002601305742008613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0093_text_document +0.00025200823006814814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0094_text_document +0.0003229951981263502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0095_text_document +0.00037289448266476045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0096_text_document +0.0003807825862179898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0097_text_document +0.0003616333738191483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0098_text_document +0.0003665117918907636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0099_text_document +0.0003684186453633228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0100_text_document +0.0003589330610806066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0101_text_document +0.00036383861418030395 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0102_text_document +0.000359841363355303 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0103_text_document +0.00036431044063050464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0104_text_document +0.0003668574090358279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0105_text_document +0.000362768263620199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0106_text_document +0.0003501888032771077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0107_text_document +0.000352401968221528 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0108_text_document +0.0003541019701869794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0109_text_document +0.0003628121865546891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0110_text_document +0.0003752582953758773 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0111_text_document +0.00037902046230424966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0112_text_document +0.0003777927146925147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0113_text_document +0.0003760676130509053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0114_text_document +0.00034046049078755405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0115_text_document +0.0003338847563259091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0116_text_document +0.00033294499102761794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0117_text_document +0.0004912026198265864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0118_text_document +0.00032064363474664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0119_text_document +0.00032154190389541214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0120_text_document +0.00032309660151746207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0121_text_document +0.00031181143365304544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0122_text_document +0.00031046092294569104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0123_text_document +0.00031150165249068046 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0124_text_document +0.0003041314265988224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0125_text_document +0.0003024834909739394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0126_text_document +0.0003019936835833604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0127_text_document +0.000292329665283177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0128_text_document +0.0002867061143144972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0129_text_document +0.00028443615610701707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0130_text_document +0.00028462291013755945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0131_text_document +0.0002793538601205013 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0132_text_document +0.00027306573977044246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0133_text_document +0.00027097155673336525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0134_text_document +0.0002752934202112985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0135_text_document +0.00043042012694697647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0136_text_document +0.00047495648822986177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0137_text_document +0.00047755032493473855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0138_text_document +0.0004706974343933747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0139_text_document +0.00046682163297771817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0140_text_document +0.0004616765425874178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0141_text_document +0.00030644496751628097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0142_text_document +0.0002909492555358308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0143_text_document +0.00027272036068261724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0144_text_document +0.0004101070217315588 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0145_text_document +0.0003728914338834357 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0146_text_document +0.00036546911442305647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0147_text_document +0.0003669945482407483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0148_text_document +0.0003715902407424017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0149_text_document +0.00035837486406683366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0150_text_document +0.0003573318538685469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0151_text_document +0.0003553784893071916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0152_text_document +0.0004920659809912352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0153_text_document +0.0004533619411303183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0154_text_document +0.00045067066057818706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0155_text_document +0.00044396985139270645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0156_text_document +0.00043198288204468477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0157_text_document +0.00043005174223738454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0158_text_document +0.00041847118430776784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0159_text_document +0.00042952036375796664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0160_text_document +0.00043420594647324267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0161_text_document +0.0003461123241053012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0162_text_document +0.0003408581597849182 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0163_text_document +0.00033172705422182547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0164_text_document +0.0003392566490686136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0165_text_document +0.00033578341518385483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0166_text_document +0.0003439196710518844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0167_text_document +0.00034559163447085543 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0168_text_document +0.00033762478642902825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0169_text_document +0.00033215210055107224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0170_text_document +0.00033423579608014966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0171_text_document +0.0004963355016025102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0172_text_document +0.0004996862761456923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0173_text_document +0.0005000551829325451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0174_text_document +0.0005004212610098755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0175_text_document +0.00027768695585500585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0176_text_document +0.00028395983854338433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0177_text_document +0.00027835826303062254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0178_text_document +0.0002740073176010804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0179_text_document +0.0002791830529274016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0180_text_document +0.0002796863816194411 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0181_text_document +0.00026697453022672804 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0182_text_document +0.0002594197440280141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0183_text_document +0.0003779565697649222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0184_text_document +0.00041835823476586606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0185_text_document +0.00043788493575265915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0186_text_document +0.0002731731970096006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0187_text_document +0.000276305847423402 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0188_text_document +0.0002704955773958623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0189_text_document +0.0002629635944827518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0190_text_document +0.000260070956974436 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0191_text_document +0.00025661553791456334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0192_text_document +0.00025794727207576157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0193_text_document +0.00025295733980001527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0194_text_document +0.0003788106407021029 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0195_text_document +0.0004882344027669431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0196_text_document +0.0003275324309642705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0197_text_document +0.0004803401856640094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0198_text_document +0.00046720138323433943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0199_text_document +0.00043527810307095335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0200_text_document +0.00043905395741627827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0201_text_document +0.00048774175867331425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0202_text_document +0.00048380704121346737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0203_text_document +0.0004779011848346118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0204_text_document +0.00046255587581908036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0205_text_document +0.00045127922880511576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0206_text_document +0.0004503891485256095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0207_text_document +0.0004450142332303422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0208_text_document +0.00044630282482516654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0209_text_document +0.00044325014465743616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0210_text_document +0.0004263874842796447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0211_text_document +0.0004217530913646938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0212_text_document +0.000415120314341852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0213_text_document +0.00040987168279144537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0214_text_document +0.00033468337266607834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0215_text_document +0.0003353094464683005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0216_text_document +0.0004833936821707294 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0217_text_document +0.00047194878988920935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0218_text_document +0.0004648324126996427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0219_text_document +0.0004562345003964941 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0220_text_document +0.0004933203505465098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0221_text_document +0.0003530166075325466 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0222_text_document +0.00035368548192804685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0223_text_document +0.0004872620828289663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0224_text_document +0.00048293889392426456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0225_text_document +0.00047936768462267655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0226_text_document +0.00047821013991587545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0227_text_document +0.0004660610308564753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0228_text_document +0.000394683430103437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0229_text_document +0.00039165053441571324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0230_text_document +0.0003906936040164381 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0231_text_document +0.00038074803919159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0232_text_document +0.0003686529291578143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0233_text_document +0.00035832920428870976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0234_text_document +0.00035929024535947033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0235_text_document +0.0003538226556050544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0236_text_document +0.0003584167868708799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0237_text_document +0.0003480507542594234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0238_text_document +0.0003413709023543034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0239_text_document +0.00034001304759361455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0240_text_document +0.00033430532902756514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0241_text_document +0.00046519252660631277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0242_text_document +0.0002938876402514769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0243_text_document +0.00028676090994509047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0244_text_document +0.00027296150117506716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0245_text_document +0.00026513502621960483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0246_text_document +0.0002680081327926125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0247_text_document +0.00025831225828720344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0248_text_document +0.00026647037295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0249_text_document +0.0002525733734572654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0250_text_document +0.00025831708887575375 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0251_text_document +0.00042487627444443476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0252_text_document +0.0004951213245023891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0253_text_document +0.0004804051413177752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0254_text_document +0.0004662397611340532 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0255_text_document +0.0004550138655253933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0256_text_document +0.00044494909122746795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0257_text_document +0.0002899112253051385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0258_text_document +0.0004372879736279761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0259_text_document +0.0004529568099252922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0260_text_document +0.00045127826158829573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0261_text_document +0.0004436558176737439 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0262_text_document +0.0004419233237678378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0263_text_document +0.000434589215880319 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0264_text_document +0.00029153613207706566 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0265_text_document +0.0004312458058738854 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0266_text_document +0.00028741854968757313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0267_text_document +0.00046853200754421234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0268_text_document +0.0004949145252030074 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0269_text_document +0.00044459683920483167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0270_text_document +0.0003836095306696336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0271_text_document +0.0003789760237872398 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0272_text_document +0.0003749227438304427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0273_text_document +0.0003628558277173369 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_head-0274_text_document +0.00039468301394041474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0000_text_document +0.00038874701821614864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0001_text_document +0.0004158492456077867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0002_text_document +0.00042360504554060077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0003_text_document +0.00040386729844317623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0004_text_document +0.00027595096702902474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0005_text_document +0.00043638766787829135 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0006_text_document +0.0002218691596850179 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0007_text_document +0.0004437566108089954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0008_text_document +0.0003889996411609667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0009_text_document +0.00043454421906537704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0010_text_document +0.0004522564392830988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0011_text_document +0.00041517835659357416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0012_text_document +0.0002614360863446896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0013_text_document +0.00037543522111463596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0014_text_document +0.0004386190133514781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0015_text_document +0.00046358333286115075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0016_text_document +0.00043186261317942404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0017_text_document +0.0002377581602097957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0018_text_document +0.00025973334085074254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0019_text_document +0.00040139099332000796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0020_text_document +0.00043674860686687174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0021_text_document +0.00040853289309329373 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0022_text_document +0.000242910191729688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0023_text_document +0.0004431071731750582 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0024_text_document +0.0004388092670482523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0025_text_document +0.000381418866255965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0026_text_document +0.0004100117296419717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0027_text_document +0.00042469230366022745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0028_text_document +0.00041744151905374254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0029_text_document +0.00022835699906752945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0030_text_document +0.0004380161085387397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0031_text_document +0.00044803212381807456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0032_text_document +0.00040554932796137236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0033_text_document +0.0004234508646347761 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0034_text_document +0.00043341209652360653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0035_text_document +0.00023966604734537185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0036_text_document +0.000259165907316014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0037_text_document +0.0004270653021833602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0038_text_document +0.0004341547032162028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0039_text_document +0.0004111478117275994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0040_text_document +0.0004299383567984396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0041_text_document +0.0004241899124590779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0042_text_document +0.0004502719349364145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0043_text_document +0.00038994621469645615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0044_text_document +0.0003859912398894952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0045_text_document +0.0004247535950310557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0046_text_document +0.000386982084327716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0047_text_document +0.0004196451040053251 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0048_text_document +0.0004096278509782259 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0049_text_document +0.0004373334932695721 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0050_text_document +0.0004180889975240641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0051_text_document +0.00042079636929672745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0052_text_document +0.00038063574611812913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0053_text_document +0.0003817505891515542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0054_text_document +0.0004420096268860222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0055_text_document +0.00039182670726410623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0056_text_document +0.0003635667850372299 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0057_text_document +0.00041564996472055667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0058_text_document +0.000400529358757286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0059_text_document +0.0003939113874958451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0060_text_document +0.00039066622068940996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0061_text_document +0.0004290098538807143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0062_text_document +0.0004240739958197099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0063_text_document +0.00040775392659215333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0064_text_document +0.0004091634200396925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0065_text_document +0.00042299190476617914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0066_text_document +0.0003701492680344151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0067_text_document +0.0003807353844384635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0068_text_document +0.00038813507771983156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0069_text_document +0.00040072346558408346 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0070_text_document +0.0003603595180423597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0071_text_document +0.00038799421353112465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0072_text_document +0.00037575235582264926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0073_text_document +0.0004239190342959713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0074_text_document +0.0004606044799136546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0075_text_document +0.00045107950652529253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0076_text_document +0.0004391947201871058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0077_text_document +0.0004457516661123035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0078_text_document +0.0004301297170991686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0079_text_document +0.00044661704164586694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0080_text_document +0.0004438849846114837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0081_text_document +0.0004444205734316823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0082_text_document +0.0004190924165303394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0083_text_document +0.00043942581131677875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0084_text_document +0.00021568459798090663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0085_text_document +0.0003814929225407199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0086_text_document +0.0003217453179359235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0087_text_document +0.00031719591470267974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0088_text_document +0.00032434115726922137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0089_text_document +0.0004079911120371051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0090_text_document +0.000329492766381148 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0091_text_document +0.0003845916162001633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0092_text_document +0.0003835208964390098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0093_text_document +0.00037847334157173194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0094_text_document +0.00038296039903791865 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0095_text_document +0.00037896336828472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0096_text_document +0.00037620974396391355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0097_text_document +0.00037420590727111843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0098_text_document +0.000340490625886403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0099_text_document +0.0003078314411035827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0100_text_document +0.00034153990750656097 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0101_text_document +0.0003308858103982067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0102_text_document +0.0003452640607156025 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0103_text_document +0.00033095276418403455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0104_text_document +0.0003116308995860414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0105_text_document +0.00032446713226408477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0106_text_document +0.0003015816821912984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0107_text_document +0.00031612418775706894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0108_text_document +0.0003278516344971041 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0109_text_document +0.00033079446736097217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0110_text_document +0.00032278977146550837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0111_text_document +0.00032065272988207914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0112_text_document +0.0003936696452406576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0113_text_document +0.0003450109536627789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0114_text_document +0.0003339787189919641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0115_text_document +0.0003284303856176974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0116_text_document +0.00033652677276843477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0117_text_document +0.0003257822443845694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0118_text_document +0.0003293985569149334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0119_text_document +0.0003310360260148262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0120_text_document +0.0003233770986418526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0121_text_document +0.0003172280092149422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0122_text_document +0.0003160674744292835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0123_text_document +0.00030931090289598506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0124_text_document +0.0003093173886443107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0125_text_document +0.00033167847081104083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0126_text_document +0.00031131501311729723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0127_text_document +0.00031046608876279845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0128_text_document +0.00030569235942207244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0129_text_document +0.00030777943671285197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0130_text_document +0.00029303314290956683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0131_text_document +0.0003045824546400205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0132_text_document +0.00030360880677729793 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0133_text_document +0.00031646239964835433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0134_text_document +0.0003129122300603785 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0135_text_document +0.00031060464956661433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0136_text_document +0.000311819032500067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0137_text_document +0.0002977872483902282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0138_text_document +0.0003009448600922438 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0139_text_document +0.00028610292098537774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0140_text_document +0.0002988326876216654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0141_text_document +0.00028550828372819075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0142_text_document +0.0002830381750875739 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0143_text_document +0.0002848495855927156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0144_text_document +0.0002856443760308144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0145_text_document +0.00027442895344188584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0146_text_document +0.0002681160554049462 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0147_text_document +0.0003421482544126989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0148_text_document +0.0004005872948449718 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0149_text_document +0.0003930123959320308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0150_text_document +0.0003867271832275778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0151_text_document +0.000380805140455254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0152_text_document +0.0003814769861947819 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0153_text_document +0.00038025170883282324 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0154_text_document +0.0003738026647867475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0155_text_document +0.00018960856915036276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0156_text_document +0.0003697177501953134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0157_text_document +0.00036674194328136693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0158_text_document +0.00036447406838697555 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0159_text_document +0.00036686410861101255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0160_text_document +0.00035915267825103423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0161_text_document +0.0003624758404026675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0162_text_document +0.0002822812140180794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0163_text_document +0.00030620512946920813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0164_text_document +0.000294249776520589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0165_text_document +0.00030238536967523434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0166_text_document +0.00029509593361580754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0167_text_document +0.0002906912701830899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0168_text_document +0.0002921944165474959 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0169_text_document +0.00028358919691127954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0170_text_document +0.0002813182772323272 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0171_text_document +0.00027442640800299205 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0172_text_document +0.0002747820342933984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0173_text_document +0.0002747584403979717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0174_text_document +0.00027499129634862444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0175_text_document +0.0002712050404257197 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0176_text_document +0.0002616256943143254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0177_text_document +0.00026769938929002815 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0178_text_document +0.00038396081322727017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0179_text_document +0.0003863140490027991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0180_text_document +0.00037702277513203237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0181_text_document +0.0003633274156107032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0182_text_document +0.0003587473889240435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0183_text_document +0.0003507672084278415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0184_text_document +0.00033776425499780385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0185_text_document +0.0003377914127574796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0186_text_document +0.00032948015659161326 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0187_text_document +0.00033245638541392985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0188_text_document +0.00031080707640648695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0189_text_document +0.0002976903331149755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0190_text_document +0.0002965121463725523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0191_text_document +0.0002933849695266647 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0192_text_document +0.0002837035078508233 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0193_text_document +0.00028684569079589323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0194_text_document +0.0003145192320802359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0195_text_document +0.0003566937253273515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0196_text_document +0.0003470199109592918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0197_text_document +0.0003060245312041868 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0198_text_document +0.0002650817213818789 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0199_text_document +0.0002643604938780134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0200_text_document +0.000299350876031416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0201_text_document +0.0003178540797697938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0202_text_document +0.000271850367887767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0203_text_document +0.00031349896596549 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0204_text_document +0.00031749734412765755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0205_text_document +0.0003791137842391209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0206_text_document +0.0003742334169957992 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0207_text_document +0.0003705639757351107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0208_text_document +0.0003126986769797042 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0209_text_document +0.00031038132814561196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0210_text_document +0.00036464437173804883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0211_text_document +0.0003569480488951322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0212_text_document +0.0003541239221619106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0213_text_document +0.00035315297411308053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0214_text_document +0.0003572451925404141 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0215_text_document +0.0003514986129411253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0216_text_document +0.0003521798298425866 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0217_text_document +0.00034553677439244716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0218_text_document +0.000349004719809412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0219_text_document +0.0003468247484872769 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0220_text_document +0.0003465822608356558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0221_text_document +0.00035410983132162007 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0222_text_document +0.0003487908354969444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0223_text_document +0.0003479024763238147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0224_text_document +0.000341412530646823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0225_text_document +0.00034451316273667034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0226_text_document +0.0002618849993484869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0227_text_document +0.00026788679978901144 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0228_text_document +0.00027450670773227214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0229_text_document +0.0002661273129899329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0230_text_document +0.00026836569676402957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0231_text_document +0.00026155876975483236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0232_text_document +0.0002609276830117151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0233_text_document +0.0002644161630512771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0234_text_document +0.00036789208972872557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0235_text_document +0.00037829849439990513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0236_text_document +0.0003788894943523098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0237_text_document +0.0003617207777959397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0238_text_document +0.0002541334487248998 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0240_text_document +0.0002707945538071073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0241_text_document +0.00027046282716455214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0242_text_document +0.0002652443167243215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0243_text_document +0.0002685859923850986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0244_text_document +0.00025734961751176414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0245_text_document +0.000259041720872915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0246_text_document +0.00025340107274823446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0247_text_document +0.00025757135121837893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0248_text_document +0.00025617700500574084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0249_text_document +0.0002566931670562857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0250_text_document +0.0002543871190716101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0251_text_document +0.00024997565589481713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0252_text_document +0.0002954079779456287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0253_text_document +0.00034890741135252835 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0254_text_document +0.0003473298137731525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0255_text_document +0.0003296959618486435 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0256_text_document +0.0003304520061604598 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0257_text_document +0.00032377956175729824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0258_text_document +0.00031700696295168713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0259_text_document +0.0003060382346081943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0260_text_document +0.0003012003005056863 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0261_text_document +0.0002981074073993884 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0262_text_document +0.0002922128825950705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0263_text_document +0.000348901087722931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0264_text_document +0.0003408286289467841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0265_text_document +0.0003410649680770183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0266_text_document +0.0003358524215576502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0267_text_document +0.0003343661874989231 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0268_text_document +0.00032810573699389156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0269_text_document +0.00032261449539097497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0270_text_document +0.0003162694866049203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0271_text_document +0.0003158381156468853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0272_text_document +0.000317376061083603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0273_text_document +0.0003125788639953052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0274_text_document +0.0003010105041885602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0275_text_document +0.0003065865059090678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0276_text_document +0.0003084275726508053 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0277_text_document +0.00030966560718296085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0278_text_document +0.0002957728057853081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0279_text_document +0.00029904164542325336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0280_text_document +0.0002955358888729187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0281_text_document +0.00028692976446931544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0282_text_document +0.0002923476214935797 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0283_text_document +0.0002893691697212419 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0284_text_document +0.0002855895211981585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0285_text_document +0.00027968347097626246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0286_text_document +0.0002810783462604979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0287_text_document +0.00027794080455729715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0288_text_document +0.00034784376461416953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0289_text_document +0.0003488347959010943 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0290_text_document +0.00034790583710250724 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0291_text_document +0.000345913166618151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0292_text_document +0.00033801936268066675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0293_text_document +0.0003290591130212315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0294_text_document +0.00034051399521366823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0295_text_document +0.00032470943131841784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0296_text_document +0.00031679540050914276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0297_text_document +0.00031814596342422325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0298_text_document +0.0003156466289485036 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0299_text_document +0.00029985010879003633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0300_text_document +0.0002905176377776361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0301_text_document +0.0004206836775460856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0302_text_document +0.00020660449162246918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0303_text_document +0.0003461727254468087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0304_text_document +0.00020592870907067763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0305_text_document +0.00034173505299233005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0306_text_document +0.0004052437256652738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0307_text_document +0.0004080650901351697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0308_text_document +0.00039778184149144276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0309_text_document +0.00039046311464950275 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0310_text_document +0.00039043444911071384 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0311_text_document +0.000388575704932843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0312_text_document +0.00019737533145666597 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0313_text_document +0.00037610755595812403 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0314_text_document +0.00037315400127598317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0315_text_document +0.00037415028580922163 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0316_text_document +0.00036694041707212337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0317_text_document +0.00018947219857306515 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0318_text_document +0.00037046050826533545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0319_text_document +0.0003587440768559087 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0320_text_document +0.00034623936498708903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0321_text_document +0.0003502289592617922 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0322_text_document +0.00034692398063649823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0323_text_document +0.000339340809421849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0324_text_document +0.0003360510394816983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0325_text_document +0.0003354673850814145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0326_text_document +0.00032937682875877047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0327_text_document +0.00032844505049317715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0328_text_document +0.00028287199339908627 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0329_text_document +0.0002795217197003578 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0330_text_document +0.00028048955601883463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0331_text_document +0.0002769326396439027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0332_text_document +0.0002727090021299243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0333_text_document +0.0002726577841024554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0334_text_document +0.00026663619593455374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0335_text_document +0.00026068042672138127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0336_text_document +0.0002637704114326801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0337_text_document +0.0002593043567100412 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0338_text_document +0.0002599897110113453 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0339_text_document +0.0002435078682758859 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0340_text_document +0.0002450530071379054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0341_text_document +0.00024233331983743606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0342_text_document +0.0002934750947999535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0343_text_document +0.00033241226364044474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0344_text_document +0.00032938406090272075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0345_text_document +0.00032778705403953246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0346_text_document +0.00032184551480398754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0347_text_document +0.00031874002264945737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0348_text_document +0.0003165319685666433 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0349_text_document +0.00031307071173376295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0350_text_document +0.00031119524184911957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0351_text_document +0.0003102253344576429 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0352_text_document +0.0003088976240383192 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0353_text_document +0.0002951410823077708 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0354_text_document +0.00029772657676757413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0355_text_document +0.0003056048989909935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0356_text_document +0.00031991305381648026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0357_text_document +0.00030890256978362426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0358_text_document +0.0003109382904091933 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0359_text_document +0.00031035798529690644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0360_text_document +0.00030741666395911753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0361_text_document +0.0002989918594861846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0362_text_document +0.00029569635443989434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0363_text_document +0.0002973992445667285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0364_text_document +0.000293397351001072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0365_text_document +0.00028737817438047954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0366_text_document +0.00028252738144009747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0367_text_document +0.0002805511898623541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0368_text_document +0.0003718020784620472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0369_text_document +0.0003499713845765235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0370_text_document +0.00034283547445326676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0371_text_document +0.00031464759888838765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0372_text_document +0.00033188946446414833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0373_text_document +0.000326084432195463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0374_text_document +0.0003764568303917893 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0375_text_document +0.0003604955598858414 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0376_text_document +0.0003655654554133222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0377_text_document +0.00035762304033750504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0378_text_document +0.00038478883950347103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_middle-0379_text_document +0.00027735714341247454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0000_text_document +0.00028139534607773563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0001_text_document +0.00019777292251713763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0002_text_document +0.000285571704874486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0003_text_document +0.00028543482146244363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0004_text_document +0.00019434234484256758 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0005_text_document +0.00027854908176986763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0006_text_document +0.0002847068039566143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0007_text_document +0.00028672356943064853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0008_text_document +0.00027782687605808177 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0009_text_document +0.0002843539634105203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0010_text_document +0.0002894748379090401 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0011_text_document +0.0002868852440186493 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0012_text_document +0.0002818504885373851 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0013_text_document +0.00028680112812941034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0014_text_document +0.00019258978168723977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0015_text_document +0.00028760637934715155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0016_text_document +0.0002820439443912918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0017_text_document +0.0002831001054410018 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0018_text_document +0.00029001901552467397 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0019_text_document +0.00027779449377883156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0020_text_document +0.00019949837437516796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0021_text_document +0.0002907306472984446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0022_text_document +0.00027814858381318327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0023_text_document +0.00019472790889161432 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0024_text_document +0.00020472626596924125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0025_text_document +0.0002870045081974301 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0026_text_document +0.00019812241927078482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0027_text_document +0.0002817553333369554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0028_text_document +0.00027829782796642117 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0029_text_document +0.00028289431732284113 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0030_text_document +0.0002795526296717729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0031_text_document +0.00027682829988044574 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0032_text_document +0.0002895432402719184 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0033_text_document +0.0002823174903941811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0034_text_document +0.00028170972351837796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0035_text_document +0.00027807915877838826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0036_text_document +0.00028588515681452956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0037_text_document +0.00028112324090816726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0038_text_document +0.00020636178289985485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0039_text_document +0.00019447255290980535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0040_text_document +0.0002850824220591452 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0041_text_document +0.00027856429520116784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0042_text_document +0.0002820880676635633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0043_text_document +0.00028943902215995714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0044_text_document +0.0002676366291085329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0045_text_document +0.00023806333809954687 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0046_text_document +0.00024526460430233455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0047_text_document +0.00023876876664622726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0048_text_document +0.00023379770334179805 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0049_text_document +0.00024175151269138382 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0050_text_document +0.00023386583242595706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0051_text_document +0.00023771797150160827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0052_text_document +0.0002262748967483896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0053_text_document +0.0002408148346432682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0054_text_document +0.00023398651720444235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0055_text_document +0.00022989433874474592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0056_text_document +0.00023948500543957772 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0057_text_document +0.0002331594076859196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0058_text_document +0.00023375132439600242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0059_text_document +0.00023923410909668642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0060_text_document +0.00023952796315562954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0061_text_document +0.0002327466076905069 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0062_text_document +0.00023082758956797212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0063_text_document +0.0002240509275524448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0064_text_document +0.00022798879995765268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0065_text_document +0.000221172516774386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0066_text_document +0.00021767045123534623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0067_text_document +0.00021982832794804484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0068_text_document +0.00021971626543789102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0069_text_document +0.00022566565206920132 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0070_text_document +0.0002181984894194856 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0071_text_document +0.00021831417549554653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0072_text_document +0.00021601405421187145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0073_text_document +0.00022275733725519607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0074_text_document +0.00021847734911973986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0075_text_document +0.0002243591012664014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0076_text_document +0.00021688758139483833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0077_text_document +0.0002182953624789215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0078_text_document +0.00020475155724026002 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0079_text_document +0.00021498078062960065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0080_text_document +0.0002157914337233064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0081_text_document +0.00021781838494967963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0082_text_document +0.00021723242266814558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0083_text_document +0.0002176782686553837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0084_text_document +0.0003486179404943968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0085_text_document +0.00034882846352857634 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0086_text_document +0.00031400868448352596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0087_text_document +0.00030273484020011963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0088_text_document +0.00029895889118145404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0089_text_document +0.00029770764609621714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0090_text_document +0.0002990181332116852 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0091_text_document +0.00029653733972285996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0092_text_document +0.00029624649222942476 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0093_text_document +0.00029625609720203576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0094_text_document +0.00029731928930852147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0095_text_document +0.00029011721326148513 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0096_text_document +0.00028849788197494655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0097_text_document +0.00021601278623858145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0098_text_document +0.00021319599281739178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0099_text_document +0.0002153325290600083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0100_text_document +0.00018566946174516558 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0101_text_document +0.00020736824394291617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0102_text_document +0.00020857419820128004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0103_text_document +0.00020058526129536423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0104_text_document +0.00020745812166665217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0105_text_document +0.00020652171015271702 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0106_text_document +0.00020643808911278608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0107_text_document +0.00020040513914482103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0108_text_document +0.00020598050188272898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0109_text_document +0.0001969184139343296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0110_text_document +0.0001972748812937012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0111_text_document +0.0002038556751586195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0112_text_document +0.00020245186011313464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0113_text_document +0.00019950381422038783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0114_text_document +0.00020837055459665258 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0115_text_document +0.00020371856218246096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0116_text_document +0.00019537612301625791 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0117_text_document +0.00019914984508813857 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0118_text_document +0.0002053787713691309 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0119_text_document +0.00019082100541008637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0120_text_document +0.00020397153334531813 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0121_text_document +0.0002021462693077317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0122_text_document +0.00019609357008124035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0123_text_document +0.00019693256622486236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0124_text_document +0.00020007239732428112 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0125_text_document +0.00020467075741591954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0126_text_document +0.00019584883400022932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0127_text_document +0.00019135050391176972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0128_text_document +0.0003362829834208298 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0129_text_document +0.00034013691154784095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0130_text_document +0.00033215887031941976 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0131_text_document +0.00032681189065396707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0132_text_document +0.0003149138485493094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0133_text_document +0.00030179177307540077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0134_text_document +0.0002923278437581119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0135_text_document +0.00029470052278994486 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0136_text_document +0.0002994095093045731 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0137_text_document +0.00029033525096085037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0138_text_document +0.00029390798852496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0139_text_document +0.0002916230924130842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0140_text_document +0.00029419886374594913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0141_text_document +0.0002865469756730764 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0142_text_document +0.00021191292549942086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0143_text_document +0.00021369664817409847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0144_text_document +0.00021612485624266726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0145_text_document +0.00022242192634588478 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0146_text_document +0.00014605095659989698 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0147_text_document +0.00022070626106341693 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0148_text_document +0.0002174420774054071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0149_text_document +0.00021325858963116995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0150_text_document +0.0002124322999488052 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0151_text_document +0.0002081218896969054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0152_text_document +0.0002108710211556957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0154_text_document +0.00020686867095978426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0155_text_document +0.00020895752681041895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0156_text_document +0.00020741922266415738 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0157_text_document +0.0002069112657197308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0158_text_document +0.00020644627473468118 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0159_text_document +0.00020332991338121604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0160_text_document +0.0003560895677789848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0161_text_document +0.00032915779111908214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0162_text_document +0.00033810613317040864 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0163_text_document +0.00033729626594036923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0164_text_document +0.00033550342864602944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0165_text_document +0.00034173474024556906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0166_text_document +0.000331505340748827 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0167_text_document +0.0003270050330117195 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0168_text_document +0.00032585275329172556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0169_text_document +0.0003143383203190604 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0170_text_document +0.00031655199110388894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0171_text_document +0.00030738872158476413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0172_text_document +0.00030838388352699285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0173_text_document +0.0003053596995351888 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0174_text_document +0.00031836304739584593 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0175_text_document +0.000315315435873905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0176_text_document +0.0003087116248965243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0177_text_document +0.00030396790625537645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0178_text_document +0.0003335812246032149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0179_text_document +0.00034570956323095843 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0180_text_document +0.00034563035636675786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0181_text_document +0.00033411265479076335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0182_text_document +0.00034439191141692787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0183_text_document +0.0003364483125496565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0184_text_document +0.0003299500453608033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0185_text_document +0.00033163377700074837 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0186_text_document +0.00032638649660627673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0187_text_document +0.00032616167939645234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0188_text_document +0.0003205289298760723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0189_text_document +0.00031939393740815355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0190_text_document +0.00031593164066731296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0191_text_document +0.00031928871111254405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0192_text_document +0.00029670189073175004 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0193_text_document +0.00020517703846735904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0194_text_document +0.00020128418186172073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0195_text_document +0.00019662723895606717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0196_text_document +0.0001981157042081407 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0197_text_document +0.00019703489037041608 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0198_text_document +0.00019079796331785068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0199_text_document +0.0001909352306690079 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0200_text_document +0.00018824662295261396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0201_text_document +0.00019864275319325954 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0202_text_document +0.00018818516521649587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0203_text_document +0.00018875694972812844 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0204_text_document +0.00018231621170645482 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0205_text_document +0.00018349407845798273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0206_text_document +0.00018088971427746906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0207_text_document +0.00018296284236327237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0208_text_document +0.0001876011825819916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0209_text_document +0.000329052068725176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0210_text_document +0.00032223616273648536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0211_text_document +0.00031272564089633955 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0212_text_document +0.00031621609908414494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0213_text_document +0.0003117213560911235 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0214_text_document +0.00030218064069945934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0215_text_document +0.00030658916600512085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0216_text_document +0.0002915863534115821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0217_text_document +0.0002940280138374372 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0218_text_document +0.00029067860468866085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0219_text_document +0.00028529228063135635 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0220_text_document +0.00028336893301452256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0221_text_document +0.0002794668089130099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0222_text_document +0.00021681361378827842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0223_text_document +0.0001484664674497246 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0224_text_document +0.00021950558378215133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0225_text_document +0.00021806860758808645 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0226_text_document +0.00021819568718852282 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0227_text_document +0.00021626925931585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0228_text_document +0.0001464536143077762 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0229_text_document +0.00021432777088808917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0230_text_document +0.000213473805865147 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0231_text_document +0.00021397067253964538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0232_text_document +0.00020758957647437263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0233_text_document +0.00020687124337683314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0234_text_document +0.00020630057046511005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0235_text_document +0.0002091166859352538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0236_text_document +0.00020777355025615267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0237_text_document +0.00020709287641496176 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0238_text_document +0.00020736464660577094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0239_text_document +0.00020062246741862607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0240_text_document +0.00020693207561942915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0241_text_document +0.00021151004871893024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0242_text_document +0.00019930249098689716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0243_text_document +0.00021589710041231824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0244_text_document +0.00021369204789905741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0245_text_document +0.0002147099923936778 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0246_text_document +0.00021077531190389536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0247_text_document +0.0002100509829113836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0248_text_document +0.00021185362601571124 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0249_text_document +0.00020722136637339565 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0250_text_document +0.00020300093701169531 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0251_text_document +0.00019859737993313477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0252_text_document +0.00019971314372100164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0253_text_document +0.00019549908270269278 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0254_text_document +0.00019649820843534028 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0255_text_document +0.00019619415513498067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0256_text_document +0.00019493006120377898 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0257_text_document +0.00019499409035775506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0258_text_document +0.00019252988593634277 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0259_text_document +0.00019440768268686405 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0260_text_document +0.00018747161324755577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0261_text_document +0.0001879575932372779 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0262_text_document +0.00019040707058357506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0263_text_document +0.0001871931095090703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0264_text_document +0.00020112966223017096 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0265_text_document +0.00020516878165311017 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0266_text_document +0.00020664735191740533 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0267_text_document +0.00021041398572882962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0268_text_document +0.00020397992929690396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0269_text_document +0.0002039978580295561 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0270_text_document +0.00020592785601142126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0271_text_document +0.0001990755527445265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0272_text_document +0.00019729564847798732 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0273_text_document +0.00019958182230527032 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0274_text_document +0.0001985037302636386 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0275_text_document +0.00020204130355115716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0276_text_document +0.0002000296401958085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0277_text_document +0.0001983064832295463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0278_text_document +0.00019663108484195617 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0279_text_document +0.00019510678560556523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0280_text_document +0.0001873284057063206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0281_text_document +0.00019311553072495885 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0282_text_document +0.00034652137288816547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0283_text_document +0.0002813690318850024 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0284_text_document +0.00027697649713138685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0285_text_document +0.0002755419092534421 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0286_text_document +0.0002681583054440219 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0287_text_document +0.00026945753192750824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0288_text_document +0.00026169470768245737 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0289_text_document +0.00026437008960810825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0290_text_document +0.0002637294838228 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0291_text_document +0.00026491867965088836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0292_text_document +0.00025504483625138986 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0293_text_document +0.0002545040623796586 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0294_text_document +0.0002546682814073622 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0295_text_document +0.00025545439487142615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0296_text_document +0.0002626896557978271 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0297_text_document +0.00025092040940402784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0298_text_document +0.0002589154885863872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0299_text_document +0.00024106160482721467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0300_text_document +0.0002483289690087987 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0301_text_document +0.0002388930282784437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0302_text_document +0.00024006340759273874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0303_text_document +0.00023765248178029045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0304_text_document +0.00023061351965578936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0305_text_document +0.00024954224883546477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0306_text_document +0.00017861017233018525 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0307_text_document +0.00017810832743667658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0308_text_document +0.00017599709170759497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0309_text_document +0.00017462723516505223 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0310_text_document +0.0002906316527068669 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0311_text_document +0.00033762141066247166 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0312_text_document +0.00017170670574152494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0313_text_document +0.00017258674515137717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0314_text_document +0.0002815386173173926 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0315_text_document +0.0002996845935618989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0316_text_document +0.0002735268488987296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0317_text_document +0.0002971738713071517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0318_text_document +0.0002942690674002763 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0319_text_document +0.0003322222207729567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0320_text_document +0.0003378721656198464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0321_text_document +0.00018307262621851067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0322_text_document +0.00033956081502775057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0323_text_document +0.00031604820927876276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0324_text_document +0.00028805657681088917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0325_text_document +0.00026312293321215633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0326_text_document +0.00034366936722921455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0327_text_document +0.0002865256504406559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0328_text_document +0.0003063615195861786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0329_text_document +0.00028412791619666136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0330_text_document +0.00028060835132727154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0331_text_document +0.00032544974761560506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0332_text_document +0.0002647177833217225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0333_text_document +0.0003152621884896575 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0334_text_document +0.0003054625140336913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0335_text_document +0.00031183308312292263 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0336_text_document +0.00018175026696621178 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0337_text_document +0.00017699918328872 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0338_text_document +0.00018222339261441908 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0339_text_document +0.00018348005930964137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0340_text_document +0.0001810735993810541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0341_text_document +0.00030846441282038914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0342_text_document +0.0002972326889310354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0343_text_document +0.00017433421318235594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0344_text_document +0.00032799458649525895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0345_text_document +0.00032482130048512673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0346_text_document +0.00031943465668672475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0347_text_document +0.00029615593630484517 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0348_text_document +0.0002893126939511001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0349_text_document +0.0002849288351723284 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0350_text_document +0.00028383906633569267 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0351_text_document +0.00028072526091262615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0352_text_document +0.000284239564292377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0353_text_document +0.0002778903109432523 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0354_text_document +0.0002771644389501471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0355_text_document +0.0002733316182319337 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0356_text_document +0.00026362539185869363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0357_text_document +0.0002636325383220217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0358_text_document +0.00026740622442302886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0359_text_document +0.0002646771971853427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0360_text_document +0.0002628566720605389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0361_text_document +0.0002644760695434766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0362_text_document +0.0002623837702310999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0363_text_document +0.00026088722976772894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0364_text_document +0.0002567065374799158 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0365_text_document +0.00018857382101207726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0366_text_document +0.00019036580399817203 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0367_text_document +0.00018348828065261222 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0368_text_document +0.00018491851780345073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0369_text_document +0.00018904887260080187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0370_text_document +0.0001875609304251801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0371_text_document +0.00018393034720015817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0372_text_document +0.00018419795526114903 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0373_text_document +0.00018699955623404795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0374_text_document +0.00018276256902965128 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0375_text_document +0.00017698045695190812 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0376_text_document +0.00018104650132303642 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0377_text_document +0.00017758206731279688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0378_text_document +0.00017131402995103497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0379_text_document +0.000175944428350446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0380_text_document +0.0003416745727147391 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0381_text_document +0.0003163259373952889 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0382_text_document +0.0002804489269172448 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0383_text_document +0.00028748272397403175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0384_text_document +0.00027603318345630605 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0385_text_document +0.000271638824679648 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0386_text_document +0.0002763761210210942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0387_text_document +0.00026501984873172717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0388_text_document +0.00026422486894694714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0389_text_document +0.0002686339100849262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0390_text_document +0.0002610837453940606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0391_text_document +0.000260974343729353 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0392_text_document +0.0002599403837029134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0393_text_document +0.0002937273113238609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0394_text_document +0.0003341790732600504 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0395_text_document +0.0002620661576600244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0396_text_document +0.0003027929169239288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0397_text_document +0.00031944039129326894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0398_text_document +0.00019025676304139009 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0399_text_document +0.00018680910145009907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0400_text_document +0.00034215840419416437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0401_text_document +0.00018618120812119364 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0402_text_document +0.00018605853095599425 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0403_text_document +0.00018120712626096538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0404_text_document +0.00018315079292495327 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0405_text_document +0.00018362556449041974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0406_text_document +0.0001780024456718171 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0407_text_document +0.00033296526436178697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0408_text_document +0.0001802398632282846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0409_text_document +0.00017340263100798256 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0410_text_document +0.00017755840547238697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0411_text_document +0.00018419413735260606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0412_text_document +0.00017869518174591322 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0413_text_document +0.00017526271460129484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0414_text_document +0.00017852168597981907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0415_text_document +0.00017566536156787157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0416_text_document +0.00017589867964432936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0417_text_document +0.00017831487394075305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0418_text_document +0.00017837310528935862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0419_text_document +0.00018200908814216548 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0420_text_document +0.0001795136627511612 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0421_text_document +0.0003414021775300033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0422_text_document +0.00017177291787788502 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0423_text_document +0.0003441900648571877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0424_text_document +0.0003394534597060673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0425_text_document +0.0003236887233114832 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0426_text_document +0.0001639544129688747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0427_text_document +0.00019137443753211255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0428_text_document +0.00018575146284680153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0429_text_document +0.00019184792863440243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0430_text_document +0.00018966043065679055 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0431_text_document +0.00017968851317035848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0432_text_document +0.00018479881897661546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0433_text_document +0.0001813642692683015 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0434_text_document +0.0001686449798983066 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0435_text_document +0.00018516104592230446 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0436_text_document +0.00031283726601066385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0437_text_document +0.0003248607542883853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0438_text_document +0.00031583241601202365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0439_text_document +0.00031238270857730376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0440_text_document +0.000307150592403979 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0441_text_document +0.00029443829986847044 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0442_text_document +0.0002942723732234677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0443_text_document +0.00023514930666443422 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_en_tail-0444_text_document +0.0020776328951453444 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0000_text_document +0.0021768234410538883 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0001_text_document +0.002106973549276289 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0002_text_document +0.002110915756171751 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0003_text_document +0.0017032382109816464 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_head-0004_text_document +0.0019047944877712286 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0000_text_document +0.0019402711744016077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0001_text_document +0.0006264790011223686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_middle-0002_text_document +0.0017885401938106643 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/cc_news_tail-0000_text_document +0.0003547982093445404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document +0.00035934014428504944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document +0.00035707704501371544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document +0.00035287930712815354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document +0.00035977166728996823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document +0.0003581675664109838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document +0.0003548617059697185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document +0.0003639582000286208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document +0.00035375839698688127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document +0.0003743722020080678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document +0.0003530399715341242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document +0.00035511875882752406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document +0.0003618733574783154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document +0.00035185243285420104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document +0.0003541503739732106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document +0.0003631679485751914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document +0.00035748045578182274 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document +0.0003606490690555877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document +0.0003626383296610091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document +0.00035442644361264756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document +0.00035978370170539796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document +0.0003585562375341541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document +0.0003601958372888019 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document +0.000350277765402227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document +0.0003616521184211704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document +0.0003620625543608188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document +0.0003560781983850704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document +0.0003553209610592676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document +0.00035905348643915075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document +0.00034744258805696526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document +0.00035462784035661496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document +0.00034768186175100895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document +0.0003568534635532736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document +0.00035586511544371234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document +0.0003524567827568137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document +0.0003512453770426313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document +0.0003591792726468799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document +0.0003514024529343127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document +0.0003584880112586934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document +0.00035133552916418045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document +0.0003600811981350215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document +0.0003571663974228119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document +0.00035768103378874214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document +0.00035939205561113694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document +0.00035186773916029825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document +0.0003542829672490847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document +0.0003592783642898726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document +0.0003556367340099302 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document +0.00035391392271377027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document +0.00035486725707484836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document +0.00034866743396828035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document +0.0003517219808644735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document +0.00034874458549673823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document +0.000355773136961014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document +0.00035611750387841917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document +0.00035305602013916315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document +0.0003578207127071924 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document +0.00035514635841943707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document +0.00034816946212866206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document +0.0003512707269761496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document +0.0003483392117980654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document +0.0003572169607204321 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document +0.00035139153281660794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document +0.00035536422129036537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document +0.000352017164107143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document +0.000351889550179365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document +0.000358759689953589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document +0.0003569286079869268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document +0.0003657752958602099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document +0.00035396127934790697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document +0.0003618565071224743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document +0.00035146051531973204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document +0.00036107135765783567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document +0.00035019554279994576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document +0.00035567858879904983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document +0.0003504753174793183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document +0.00035931140831329194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document +0.0003502967866002823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document +0.0003532911801041972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document +0.0003583543013070199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document +0.0003566243489931224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document +0.0003468752314799221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document +0.0003597840618138091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document +0.00035128822484768084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document +0.00035889496943437507 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document +0.000352400524650424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document +0.0003518689536768735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document +0.00035866864741303467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document +0.0003454687659106334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document +0.00035348007259317576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document +0.0003539752270940644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document +0.00035146495994081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document +0.00035397212846310423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document +0.00035208246467162587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document +0.0003490843168676626 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document +0.00035299633658644394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document +0.00034868327466167065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document +0.00035941351365601583 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document +0.0003545343062735255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document +0.0003528956380445978 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document +0.0003553355770443352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document +0.0003644224004937743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document +0.00035234291036216907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document +0.0003596237469847771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document +0.0003531996065735989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document +0.0003547177054106099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document +0.0003575586499260483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document +0.00035262635135283667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document +0.0003624191962188944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document +0.0003488398052948616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document +0.0003598294093147917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document +0.00035583006534466323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document +0.00035403139653225103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document +0.00036134702642187156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document +0.0003573689927162834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document +0.0003577141131435527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document +0.00035208814419277406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document +0.00035996720683665625 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document +0.00035415304658912596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document +0.00036353353029443546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document +0.0003537326003150983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document +0.00036053976358299083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document +0.000352380489373494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document +0.00036154661616900994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document +0.00035959332325963614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document +0.0003597954667189692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document +0.0003563108270597542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document +0.0003582891940460143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document +0.0003497728210484297 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document +0.0003549834902179354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document +0.0003529828233484542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document +0.00034627483903285777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document +0.00035569006572589215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document +0.00035449377946910314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document +0.00035802844396194623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document +0.0003617277809353208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document +0.00035034118898654814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document +0.000351091193908611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document +0.0003527914342210668 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document +0.00035028288369781376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document +0.00035775745592780506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document +0.0003449630690661468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document +0.0003583490698830361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document +0.0003476995746684122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document +0.0003535632505019212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document +0.00035640180641147417 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document +0.000361731045691765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document +0.0003534082129597368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document +0.0003550344149828664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document +0.00035363002411364057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document +0.0003537265579677396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document +0.00034950531383577937 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document +0.00035008511827347514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document +0.00035594533400871325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document +0.00035266312861335946 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document +0.00035280268794863923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document +0.0003565470391528536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document +0.0003588492322689137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document +0.00035469909697832775 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document +0.00034712082813410526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document +0.000348701157101807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document +0.0003500192014479944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document +0.00035120560544669755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document +0.00035403656850437445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document +0.00035852376560749366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document +0.0003534754068111774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document +0.00035591740046720765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document +0.000348522354782563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document +0.0003533533959664415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document +0.00035631425964030697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document +0.0003485886551574741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document +0.00035917652631065777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document +0.0003482975272111288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document +0.00035580661277480167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document +0.0003492290722955348 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document +0.00034989284450240613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document +0.0003545677216162781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document +0.00034622286859463484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document +0.00036070626989861965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document +0.00035518365036320786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document +0.00035272907057848406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document +0.0003547343638218734 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document +0.0003496450144966242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document +0.0003537407829294287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document +0.0003489722653985685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document +0.00035057186899911295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document +0.0003507566548933051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document +0.00035630360179023747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document +0.00035631362503416367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document +0.0003490204248026821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document +0.00035761724058371226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document +0.00035037664777467137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document +0.000353402110481068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document +0.00034524163568371745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document +0.00035528523728570974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document +0.00034784916132431703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document +0.00034928476408048925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document +0.00034989205973784984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document +0.00034201664404094254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document +0.0003529676016338611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document +0.00034643433682346637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document +0.0003511666373001904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document +0.00034828669066575333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document +0.0003494625207264413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document +0.0003458957535879216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document +0.0003543020478990003 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document +0.00034754384069014956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document +0.0003598856392240133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document +0.0003503335458553846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document +0.00035919595619778716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document +0.00035767737970754404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document +0.00035197152783998165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document +0.0003549609834422404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document +0.0003568184100569753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document +0.0003512652818651935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document +0.00035912648958665754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document +0.00034764526964056546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document +0.000352439784960359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document +0.00035295886560764226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document +0.0003518132693658672 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document +0.00035589987915465713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document +0.00034923863317385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document +0.0003457987267929692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document +0.0003560928663480501 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document +0.0003529603811204932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document +0.0003524438555443043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document +0.0003438847030263783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document +0.00035981978898461613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document +0.0003446342778566972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document +0.00035529584995236537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document +0.00034855740895831116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document +0.00034932634912802544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document +0.00035805518303064666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document +0.0003497941877073061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document +0.00035774398685405447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document +0.0003560421780316607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document +0.0003508844468369392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document +0.00035731928892270107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document +0.0003557884626314314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document +0.00034992996760289355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document +0.000360752554360921 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document +0.0003452321668708545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document +0.0003591745226131023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document +0.00035256981433229084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document +0.00035378123159712034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document +0.000350464354895999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document +0.00035074625557389677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document +0.00035025894701994667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document +0.00035437902514857614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document +0.0003514684519732232 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document +0.00035449717909633905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document +0.0003436816402714221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document +0.00035139158071782116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document +0.0003509424079843335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document +0.000343894618577506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document +0.0003500789770661659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document +0.0003407788080680086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document +0.0003581908175239701 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document +0.0003465541618780918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document +0.00034600228792437736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document +0.00034416738982773204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document +0.0003519900340150641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document +0.000343369616864659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document +0.0003544993883274688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document +0.0003504441365073392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document +0.00034859160702727056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document +0.00035355909532647185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document +0.0003471900922691849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document +0.0003563015508709187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document +0.0003487888744148821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document +0.00034711767548688336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document +0.0003530734609369085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document +0.00035123969242560935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document +0.0003517127620891489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document +0.00035232835416868673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document +0.0003524437481912308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document +0.0003525996167005602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document +0.00035064770545242043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document +0.00035311558274981226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document +0.00034952204800569914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document +0.0003541471367344846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document +0.00035418812454561825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document +0.0003528951372900714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document +0.0003542338042975688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document +0.00034937738939942796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document +0.0003522182190878447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document +0.0003501406466507449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document +0.00034973079877492633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document +0.0003485274567713538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document +0.00034999308679368985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document +0.0003570051724707296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document +0.00034567230462019706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document +0.00035529000940160696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document +0.00034956512308671755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document +0.0003496962834028953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document +0.0003468745282493457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document +0.0003502717155809202 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document +0.0003556240880896514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document +0.0003515109488424343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document +0.0003563156688192592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document +0.00035040277363989817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document +0.0003481408593290717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document +0.0003624575124332874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document +0.0003522684124250313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document +0.00035286996027653544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document +0.00034967623997256725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document +0.00035182649587602765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document +0.0003524892557026489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document +0.0003507642477451811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document +0.00036190408389835666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document +0.00035102739424880766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document +0.00035239718753257265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document +0.00035298076121821316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document +0.0003478704389752654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document +0.0003503109191567942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document +0.00035143250975654426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document +0.0003480663923069012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document +0.00035691540219998623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document +0.000348815437166351 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document +0.00035202073257766225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document +0.0003491569096274706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document +0.00035277390475511834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document +0.0003524972090026609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document +0.0003504854249750236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document +0.00034740238025423914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document +0.00034968015462277606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document +0.0003493798632762674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document +0.0003488202537862122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document +0.0003525461864643725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document +0.00034903815232825664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document +0.00035536982539258216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document +0.00034858083265155483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document +0.0003505014973608067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document +0.00035327984042622104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document +0.0003503286677453136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document +0.00035835274842442816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document +0.00034970302660275595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document +0.000357929573140149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document +0.0003517238649788585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document +0.00036097027318848475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document +0.0003502734074110026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document +0.00035801510806036273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document +0.0003568006373479869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document +0.00036128108717454636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document +0.0003563436883111686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document +0.00035559725321852463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document +0.00035089656006854944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document +0.000359453964362057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document +0.00035629498059104033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document +0.0003622207707090437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document +0.0003540946784512821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document +0.0003594750565232011 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document +0.0003566007415086991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document +0.0003562142599126134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document +0.0003569948186744601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document +0.00035166554847920186 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document +0.00035047994419295137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document +0.0003561578193739437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document +0.00035470866838811544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document +0.00034216920464876335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document +0.0003550021513075795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document +0.0003488045105938729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document +0.0003513340720840151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document +0.0003448558566387584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document +0.0003460966026953241 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document +0.0003488157616036459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document +0.0003446120387842362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document +0.000351528602987427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document +0.00035661118227454713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document +0.0003551342699877457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document +0.0003478953397924445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document +0.00034625782458988215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document +0.0003527515447405871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document +0.00034823744889805696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document +0.00034823314560254406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document +0.00035162668292961944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document +0.0003477307716074623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document +0.0003446457989477787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document +0.00034782916273767795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document +0.0003517249130302248 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document +0.0003449873430908556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document +0.00034841291749669877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document +0.0003466028498941749 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document +0.0003486436831199424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document +0.0003478279234211838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document +0.0003495903653274374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document +0.00034896893881218957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document +0.000348941645312426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document +0.0003474221308416894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document +0.0003462621543839385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document +0.0003669373860863891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document +0.00034691156268163006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document +0.0003527774103765281 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document +0.00034684565672734663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document +0.0003454250599604457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document +0.0003541536557159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document +0.000345735737037366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document +0.0003524669816385214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document +0.0003441817133096468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document +0.0003519093265859089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document +0.00035080085480352095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document +0.00035285227929327434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document +0.00034354836346901676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document +0.00034789770937373467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document +0.000343665920520102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document +0.0003490884931060568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document +0.00034380029463398654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document +0.00034874768005099945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document +0.0003457058510967673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document +0.00034644265227023904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document +0.00035008339858594957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document +0.0003462377193296194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document +0.0003620491787114201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document +0.000348717011044469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document +0.00034370072363913706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document +0.0003551981066775649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document +0.0003500119496799342 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document +0.0003485082952669081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document +0.0003508155580978919 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document +0.00035311375163251416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document +0.00034945972003423253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document +0.0003474220353789879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document +0.0003536443686585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document +0.0003560350489042953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document +0.0003493655927914396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document +0.0003528423977146383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document +0.00035255554724471217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document +0.0003479760010190111 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document +0.00035458598862501956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document +0.0003458990560538315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document +0.00035157946422379875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document +0.00034736860650169996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document +0.0003529152313394119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document +0.00034586294329524465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document +0.00035707214923794877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document +0.0003509580363496512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document +0.00035244176725524474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document +0.0003467539557999047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document +0.00034919687962275546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document +0.00035094031731719953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document +0.0003484309008351352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document +0.0003485409424916253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document +0.0003499590776117838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document +0.0003492842758957848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document +0.0003529712275178912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document +0.0003566141287087449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document +0.0003649496522047409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document +0.0003563218912208234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document +0.00035614782126966145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document +0.0003531944298453266 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document +0.0003535950949566616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document +0.0003544295554928795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document +0.0003519908503740376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document +0.00035752817626134463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document +0.0003515322689589972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document +0.0003486893890307115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document +0.0003446520464889867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document +0.0003509421562481707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document +0.00035335015702909084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document +0.0003490178167345008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document +0.0003520497821155174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document +0.0003549762618908944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document +0.00035072190850833103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document +0.0003542458638526423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document +0.000352419194572916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document +0.0003545102564672614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document +0.0003495437992331806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document +0.0003542843376993964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document +0.000352827529313958 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document +0.00035442506093223886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document +0.0003496970719044257 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document +0.0003553096424442362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document +0.00034986845565067564 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document +0.000352131055186658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document +0.0003527021708198983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document +0.00034905885414547214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document +0.0003583433842468394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document +0.00034409435202828383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document +0.00034846410520871483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document +0.0003554459991927314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document +0.00035310507471843076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document +0.000350028910786098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document +0.00035049727458009896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document +0.0003519047735925826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document +0.0003513027429919726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document +0.0003626947260354396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document +0.0003500087324849783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document +0.0003618315726725285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document +0.0003535385113938023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document +0.0003487064058517615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document +0.0003618709124780938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document +0.00035040070335625915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document +0.0003506279032267829 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document +0.0003498435310527524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document +0.0003554634749821431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document +0.00035091209738758963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document +0.00035034103678978573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document +0.00035398931854386146 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document +0.00035495529304989485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document +0.00036067883473356603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document +6.322825248625475e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document +2.4432314037946264e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document +5.6313888721313454e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document +2.4208171781595055e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document +2.325811856369237e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document +2.4010790356322705e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document +5.36773610843632e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document +1.360574433501002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document +1.3076540344853244e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document +1.3386534334886313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document +1.2498103719605153e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document +1.403763836949682e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document +1.3636756723495417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document +1.2242489446940814e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document +1.2398255818973339e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document +1.2972616994216281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document +1.3947809855914134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document +1.3144843787829514e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document +1.1693809976572487e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document +1.3677252682893802e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document +1.3940876719849597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document +1.4222245138730965e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document +1.3201677767919704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document +1.1421717796486169e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document +1.2890514724498703e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document +1.3649507648749037e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document +1.2400732563490717e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document +1.1557681453277616e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document +1.2294483595964517e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document +1.2137484472122283e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document +1.3299663426456e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document +1.2461984216479532e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document +1.4666434217609636e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document +1.1876997894686238e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document +1.2939155338964078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document +1.3859590039728515e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document +1.317917848615668e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document +1.1335281536110342e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document +1.2889923952861426e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document +1.3471671647053326e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document +1.2221720014475102e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document +1.2632647276287541e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document +1.28276219004076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document +1.36213704321643e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document +1.2414858625261553e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document +1.3173700421883744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document +1.295597796725686e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document +1.242783936442904e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document +1.2417374088427464e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document +1.2134479405400744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document +1.3090040663304255e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document +1.2713470581614905e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document +5.5750231378906594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document +5.777597358425469e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document +5.349786767471258e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document +5.675165050453583e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document +5.482611216158831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document +5.065421899890121e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document +5.384718357480146e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document +4.872037363236061e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document +4.532709250783155e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document +5.7257963030489613e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document +4.9014365579652036e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document +5.722863552770969e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document +6.149911636146833e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document +5.2178057608273506e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document +4.990228161160431e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document +5.866186875255134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document +5.004185734360719e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document +4.79401853705107e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document +5.435219965052376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document +5.035997225792266e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document +5.622401774211625e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document +5.028826157387559e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document +5.596379470128795e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document +6.027824493191489e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document +5.5358270009931474e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document +5.9839051807685496e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document +5.1221077499249595e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document +5.517228560620279e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document +5.1687858285052305e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document +5.684188244145645e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document +5.212693275535878e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document +4.8551007022784084e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document +5.4888506639203145e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document +5.345098688527242e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document +4.8506420625516594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document +5.132168603397676e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document +5.719476795114223e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document +5.7448621149792696e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document +4.9068410568059265e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document +5.382937299647678e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document +4.8288432136304634e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document +5.841703200305416e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document +5.1589611587885584e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document +6.031113829732574e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document +5.4558202844532094e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document +5.341852317196142e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document +5.1402942738369954e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document +5.735421384377395e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document +5.473629863586958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document +5.4708993245733936e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document +4.931161863634078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document +5.104173022127248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document +5.510157161510824e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document +5.652501401782597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document +5.7273656573031666e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document +5.638363224821738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document +5.6128115396668704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document +5.00304877998141e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document +5.596120554779096e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document +5.5280923889040006e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document +5.223477917938408e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document +5.29472809986569e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document +2.205682378243213e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document +1.4367563720603185e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document +3.5506193487931076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document +3.0442910855821778e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document +2.2540042508019627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document +2.6880163202623216e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document +2.534473148048727e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document +2.6560945431318916e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document +2.547470248967691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document +2.5248825388073738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document +2.5828729575000054e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document +2.4026583817957736e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document +2.3930425429834413e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document +2.5037365362599724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document +2.6696745470595603e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document +2.140323051341762e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document +2.617354786691592e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document +1.538359101762691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document +1.2871029252377856e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document +2.255195411289217e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document +2.4832313897952067e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document +9.303873918189968e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document +2.179532302620228e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document +1.9750517506901206e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document +2.7740420380648435e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document +2.7813714782319335e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document +4.1595357937609806e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document +2.741365122389175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document +2.117451071361901e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document +1.7132649760565998e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document +1.7492547092602047e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document +1.7499951097392276e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document +1.6632444789170958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document +1.6678802252361607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document +1.5519208704558896e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document +1.652420992967167e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document +1.6119931034508755e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document +1.6638882076736552e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document +1.7198076782652946e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document +1.572927860565175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document +1.5194822618169918e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document +1.6677776832669846e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document +1.595612492245688e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document +1.682350633181197e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document +1.663983380609724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document +1.710187842689243e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document +1.5733697527539038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document +1.6972104757911438e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document +1.6610142847616577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document +1.61094882403031e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document +1.4789207305138325e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document +1.639299617676302e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document +1.3241204512116132e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document +8.582260726625535e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document +8.213000975576739e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document +9.549247732811947e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document +9.17242785339013e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document +7.632868223725218e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document +8.674401118222175e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document +9.124384255505347e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document +8.344222222417358e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document +8.992299957499065e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document +8.76689497361025e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document +7.973396239586015e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document +9.006935606644125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document +8.725545954955498e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document +1.215449694669174e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document +3.3041720284158646e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document +2.0593512412624502e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document +1.893608946986248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document +1.737111666788535e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document +1.4915923449873955e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document +2.289370239067605e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document +2.8615335689614638e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document +8.847283630883125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document +1.8175470362373804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document +1.8152226683368038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document +1.789149655314284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document +1.7690523036477663e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document +1.8333732213753644e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document +1.8794105687718654e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document +1.721841156706417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document +2.0612008685724796e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document +1.9297370681336376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document +2.0188440409661018e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document +5.1741216329695265e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document +1.3417913926038429e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document +1.1010813016469651e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document +1.1252416134320087e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document +1.2801744104313002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document +1.3041514955795817e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document +1.3428837580879075e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document +1.320809382267804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document +1.3451566676555968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document +1.228284926657501e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document +1.2410599573923043e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document +1.3815343367377182e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document +1.3895126265148832e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document +1.2306773644401741e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document +1.32981021906281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document +1.101337469221607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document +1.513094184404692e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document +1.1073759547073234e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document +1.2879348765857567e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document +9.619595770228435e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document +1.2384340836286436e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document +1.1766667232211577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document +1.2871049236196452e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document +1.2010645926497744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document +1.3971428231518597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document +1.2283733550547932e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document +1.2659530508255308e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document +1.551775613074462e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document +1.1169413343776979e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document +1.1433700593712463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document +4.964773647323492e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document +1.0995586595687313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document +1.2957393071411267e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document +2.75899247407709e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document +2.8269344597344854e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document +2.329108187246831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document +2.4231761430460284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document +1.2434140512230442e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document +1.638718338352859e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document +3.272953556801187e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document +6.061314500486327e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document +1.2465979731210292e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document +1.2737557327967737e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document +1.038428658075627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document +2.61666472045566e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document +3.6506873212272224e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document +1.5066359138295701e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document +1.1166290872121178e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document +1.5546966228590285e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document +1.2583434625014828e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document +1.3398826881300862e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document +1.2944933160515968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document +1.0971437399901365e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document +1.2787922795775774e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document +1.404979227816985e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document +1.3344734431324463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document +4.886031157107555e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document +3.277261443596394e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document +3.5057957685786495e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document +3.287625301718589e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document +3.1370056372668855e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document +3.186092015785841e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document +7.271819324142512e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document +0.001451215788905126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document +0.0014486847196258788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document +0.0008861032722895899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document +0.0018119590809459816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document +0.0008916937917547129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document +6.960128832809415e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document +0.002008403651063623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document +0.0014374900742131454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document +0.00180213596996716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document +0.001956178877532413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document +0.0008829547017667033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document +0.0008910853619157279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document +0.0018260998845299973 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document +0.0012499632072059553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document +0.00125398260359913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document +0.0012541704774729071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document +0.0012527268234360602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document +0.0012532925243737164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document +0.0012456396241204315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document +0.0012589894424352072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document +0.001508020123999618 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document +0.00333096950781965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document +0.0033233414614415547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document +0.003512387990689828 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document +0.0035091382940513126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document +0.003514155927147005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document +0.003327108000579638 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document +0.003329106196589836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document +0.003505604148738077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document +0.003324825759567855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document +0.0033248240149804913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document +0.0033385962112851358 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document +0.0035043186296553615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document +0.003340469505431529 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document +0.0035106889084796276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document +0.0033309469281030167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document +0.003340337858029757 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document +0.003505919861097801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document +0.0003882924098240512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document +0.0005759963691850877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document +0.0005959971675332674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document +0.0006026179290353799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document +0.0005824184320784846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document +0.0005854598548616037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document +0.0005903767055633473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document +0.0005930306490982049 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document +0.000569425602700746 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document +0.0005675060415179408 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document +0.0005772431621253389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document +0.0005678026053826858 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document +0.0005700398263483378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document +0.0005669467963528824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document +0.0005701015953324305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document +0.0005795907287413296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document +0.0005735602737531164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document +0.0005749862745842101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document +0.0005693257015931971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document +0.0005716568794795563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document +0.0005761083919774021 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document +0.0005688343169797355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document +0.0005807913190929842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document +0.0005710229258078636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document +0.0005704083039826862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document +0.0005862132348308056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document +0.0005717662049559556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document +0.0005858155213694451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document +0.0005812012281792392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document +0.0005803981414588498 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document +0.0005700102108287723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document +0.0005719243459052329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document +0.0005867253401661752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document +0.0005731087218860733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document +0.0005712197789109317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document +0.0005702376926310089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document +0.0005700411527742972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document +0.0005828090098178196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document +0.0005770140826168056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document +0.0005723509664597896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document +0.0005755499231836962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document +0.0005636407438471367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document +0.0005640281556500104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document +0.0005633159058766496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document +0.0005638034311151449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document +0.0005630066273073224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document +0.0005631803831128559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document +0.0005631228881679657 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document +0.0005628178701487633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document +0.0005624448092256196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document +0.0005620957024062329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document +0.0005614201504177484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document +0.0005616890951464056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document +0.0005611348559279058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document +0.0005604238061828518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document +0.0005603301490194237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document +0.0005607291294548833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document +0.0005605234569930727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document +0.0005613778566640694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document +0.0005610248539992471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document +0.0005599977416780475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document +0.0005603632562116935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document +0.0005599177479509897 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document +0.0005595202318298379 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document +0.0005600975633499175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document +0.0005614075491213365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document +0.000612563885043477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document +0.0005515469909644413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document +0.0005526782014946906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document +0.0005472463408095445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document +0.0005502284746004587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document +0.0005414514790555363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document +0.0005513499500134784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document +0.0005391391454105187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document +0.0005415836910001838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document +0.0005208132468536551 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document +0.0005889827143132871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document +0.0005822520817765276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document +0.0004173155230758696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document +0.0009994361338078242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document +0.001087156194657966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document +0.0010667737163656816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document +0.0009602877882124873 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document +0.0008968956271971105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document +0.0009198034843762967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document +0.0009423901016715341 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document +0.0009674094553686345 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document +0.0009858331322519164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document +0.0009970593645879198 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document +0.0010027035193731686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document +0.0010128291154221853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document +0.0010215631382631918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document +0.0010288663771461238 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document +0.0010346219929285867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document +0.00104544019940344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document +0.0010525172676724333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document +0.0010609529620775127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document +0.0010725892748610153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document +0.0010818563598181568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document +0.0010992760196793917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document +0.0011178992762079917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document +0.001124687532085676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document +0.001118303661267191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document +0.0010206825575416534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document +0.0005512280117499715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document +0.004474659408857016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document +0.00409944473890653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document +0.005137179939941845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document +0.005143172251066109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document +0.005206134363352808 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document +0.004892747858974329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document +0.004844731352552902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document +0.005308320169123755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document +0.005124709815666577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document +0.005424710744483826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document +0.00538244648861977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document +0.0029107284679086853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document +0.0026825258998444705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document +0.0026904503191419243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document +0.002687906577174073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document +0.002850165346048818 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document +0.005322698571717847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document +0.004450334290869719 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document +0.004700990083440683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document +0.003903568556500995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document +0.00390561515396931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document +0.0039046402900912262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document +0.003907454839379547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document +0.0038583224578603824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document +0.0037914116657695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document +0.003786665266798682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document +0.003792000802430658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document +0.00319266847466091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document +0.0032658716699838944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document +0.0034801959532460023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document +0.0028307012092022594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document +0.0028420360878146276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document +0.0028410455248484914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document +0.00283497183526842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document +0.002840187195459487 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document +0.0028398709431369834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document +0.004364722843422023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document +0.004093255713117101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document +0.004092331079566252 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document +0.004005326985579649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document +0.0036205502856964207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document +0.003625316793034984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document +0.003604743435602363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document +0.0035405823343673125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document +0.0041601413517253945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document +0.005886303658937057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document +0.003600909532810332 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document +0.0034941365817168658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document +0.0004992164842980224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document +0.00032927705604725614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document +0.0002860154190878753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document +0.0002845217585425619 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document +0.0002743528685497456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document +0.00026025323737738766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document +0.00023493876414603155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document +0.00029665994994226705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document +0.00031808102075993956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document +0.00031813573046011285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document +0.0002711905171855542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document +0.00028892513401817095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document +0.00030003908676979083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document +0.00026839878771944684 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document +0.00029155935002690497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document +0.0002998624927624209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document +0.0003091705447974841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document +0.00026873195794309786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document +0.00027721873498527547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document +0.0002841662554024377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document +0.0002839461156551537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document +0.0002861705604659811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document +0.0002460995649635886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document +0.00019420142619795496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document +0.00021967677816173628 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document +0.0002620283200480949 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document +0.0002433390542188936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document +0.00021254976608350767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document +0.00022094815569522115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document +0.000342862378668244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document +0.00033784225259118157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document +0.0003367278459543952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document +0.00029843279042852765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document +0.0002926583661257988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document +0.00029320337282010673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document +0.00029281450669483455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document +0.0002915338187002653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document +0.0002864226923084572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document +0.00028643439083586396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document +0.00028253710956299054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document +0.0002810856078805806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document +0.00031474941344656715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document +0.0002139130222205655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document +0.0003084648871862831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document +0.0003309477872140129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document +0.0003360096824695161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document +0.0003355452655196557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document +0.00038119390366386037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document +0.00038078927630086064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document +0.0003386200917551554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document +0.0002158905159938882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document +0.00021621682877018768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document +0.00021553306942740535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document +0.00021581563462722296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document +0.0002157694110556169 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document +0.000215643699847159 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document +0.00021532716715168094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document +0.00021531221326022472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document +0.0002831801179028896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document +0.0002514844936507595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document +0.00031638782778107964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document +0.0002749197545278445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document +0.00026159721512464495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document +0.0002630052420096968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document +0.00031106811228913666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document +0.0002852973415334161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document +3.7555372465932136e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document +0.003548077173506675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document +0.0018372203137874265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document diff --git a/ALCF/data-lists/aurora/falcon.txt b/ALCF/data-lists/aurora/falcon.txt index 27b3bbe0474..997038fb383 100644 --- a/ALCF/data-lists/aurora/falcon.txt +++ b/ALCF/data-lists/aurora/falcon.txt @@ -1,501 +1,501 @@ -0.0003547982093445404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon -0.00035934014428504944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon -0.00035707704501371544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon -0.00035287930712815354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon -0.00035977166728996823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon -0.0003581675664109838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon -0.0003548617059697185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon -0.0003639582000286208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon -0.00035375839698688127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon -0.0003743722020080678 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon -0.0003530399715341242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon -0.00035511875882752406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon -0.0003618733574783154 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon -0.00035185243285420104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon -0.0003541503739732106 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon -0.0003631679485751914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon -0.00035748045578182274 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon -0.0003606490690555877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon -0.0003626383296610091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon -0.00035442644361264756 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon -0.00035978370170539796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon -0.0003585562375341541 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon -0.0003601958372888019 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon -0.000350277765402227 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon -0.0003616521184211704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon -0.0003620625543608188 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon -0.0003560781983850704 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon -0.0003553209610592676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon -0.00035905348643915075 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon -0.00034744258805696526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon -0.00035462784035661496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon -0.00034768186175100895 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon -0.0003568534635532736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon -0.00035586511544371234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon -0.0003524567827568137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon -0.0003512453770426313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon -0.0003591792726468799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon -0.0003514024529343127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon -0.0003584880112586934 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon -0.00035133552916418045 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon -0.0003600811981350215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon -0.0003571663974228119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon -0.00035768103378874214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon -0.00035939205561113694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon -0.00035186773916029825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon -0.0003542829672490847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon -0.0003592783642898726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon -0.0003556367340099302 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon -0.00035391392271377027 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon -0.00035486725707484836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon -0.00034866743396828035 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon -0.0003517219808644735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon -0.00034874458549673823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon -0.000355773136961014 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon -0.00035611750387841917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon -0.00035305602013916315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon -0.0003578207127071924 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon -0.00035514635841943707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon -0.00034816946212866206 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon -0.0003512707269761496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon -0.0003483392117980654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon -0.0003572169607204321 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon -0.00035139153281660794 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon -0.00035536422129036537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon -0.000352017164107143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon -0.000351889550179365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon -0.000358759689953589 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon -0.0003569286079869268 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon -0.0003657752958602099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon -0.00035396127934790697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon -0.0003618565071224743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon -0.00035146051531973204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon -0.00036107135765783567 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon -0.00035019554279994576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon -0.00035567858879904983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon -0.0003504753174793183 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon -0.00035931140831329194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon -0.0003502967866002823 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon -0.0003532911801041972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon -0.0003583543013070199 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon -0.0003566243489931224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon -0.0003468752314799221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon -0.0003597840618138091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon -0.00035128822484768084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon -0.00035889496943437507 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon -0.000352400524650424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon -0.0003518689536768735 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon -0.00035866864741303467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon -0.0003454687659106334 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon -0.00035348007259317576 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon -0.0003539752270940644 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon -0.00035146495994081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon -0.00035397212846310423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon -0.00035208246467162587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon -0.0003490843168676626 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon -0.00035299633658644394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon -0.00034868327466167065 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon -0.00035941351365601583 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon -0.0003545343062735255 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon -0.0003528956380445978 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon -0.0003553355770443352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon -0.0003644224004937743 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon -0.00035234291036216907 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon -0.0003596237469847771 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon -0.0003531996065735989 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon -0.0003547177054106099 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon -0.0003575586499260483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon -0.00035262635135283667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon -0.0003624191962188944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon -0.0003488398052948616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon -0.0003598294093147917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon -0.00035583006534466323 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon -0.00035403139653225103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon -0.00036134702642187156 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon -0.0003573689927162834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon -0.0003577141131435527 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon -0.00035208814419277406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon -0.00035996720683665625 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon -0.00035415304658912596 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon -0.00036353353029443546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon -0.0003537326003150983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon -0.00036053976358299083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon -0.000352380489373494 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon -0.00036154661616900994 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon -0.00035959332325963614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon -0.0003597954667189692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon -0.0003563108270597542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon -0.0003582891940460143 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon -0.0003497728210484297 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon -0.0003549834902179354 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon -0.0003529828233484542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon -0.00034627483903285777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon -0.00035569006572589215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon -0.00035449377946910314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon -0.00035802844396194623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon -0.0003617277809353208 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon -0.00035034118898654814 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon -0.000351091193908611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon -0.0003527914342210668 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon -0.00035028288369781376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon -0.00035775745592780506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon -0.0003449630690661468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon -0.0003583490698830361 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon -0.0003476995746684122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon -0.0003535632505019212 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon -0.00035640180641147417 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon -0.000361731045691765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon -0.0003534082129597368 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon -0.0003550344149828664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon -0.00035363002411364057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon -0.0003537265579677396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon -0.00034950531383577937 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon -0.00035008511827347514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon -0.00035594533400871325 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon -0.00035266312861335946 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon -0.00035280268794863923 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon -0.0003565470391528536 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon -0.0003588492322689137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon -0.00035469909697832775 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon -0.00034712082813410526 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon -0.000348701157101807 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon -0.0003500192014479944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon -0.00035120560544669755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon -0.00035403656850437445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon -0.00035852376560749366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon -0.0003534754068111774 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon -0.00035591740046720765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon -0.000348522354782563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon -0.0003533533959664415 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon -0.00035631425964030697 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon -0.0003485886551574741 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon -0.00035917652631065777 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon -0.0003482975272111288 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon -0.00035580661277480167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon -0.0003492290722955348 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon -0.00034989284450240613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon -0.0003545677216162781 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon -0.00034622286859463484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon -0.00036070626989861965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon -0.00035518365036320786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon -0.00035272907057848406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon -0.0003547343638218734 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon -0.0003496450144966242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon -0.0003537407829294287 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon -0.0003489722653985685 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon -0.00035057186899911295 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon -0.0003507566548933051 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon -0.00035630360179023747 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon -0.00035631362503416367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon -0.0003490204248026821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon -0.00035761724058371226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon -0.00035037664777467137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon -0.000353402110481068 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon -0.00034524163568371745 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon -0.00035528523728570974 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon -0.00034784916132431703 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon -0.00034928476408048925 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon -0.00034989205973784984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon -0.00034201664404094254 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon -0.0003529676016338611 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon -0.00034643433682346637 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon -0.0003511666373001904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon -0.00034828669066575333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon -0.0003494625207264413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon -0.0003458957535879216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon -0.0003543020478990003 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon -0.00034754384069014956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon -0.0003598856392240133 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon -0.0003503335458553846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon -0.00035919595619778716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon -0.00035767737970754404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon -0.00035197152783998165 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon -0.0003549609834422404 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon -0.0003568184100569753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon -0.0003512652818651935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon -0.00035912648958665754 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon -0.00034764526964056546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon -0.000352439784960359 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon -0.00035295886560764226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon -0.0003518132693658672 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon -0.00035589987915465713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon -0.00034923863317385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon -0.0003457987267929692 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon -0.0003560928663480501 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon -0.0003529603811204932 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon -0.0003524438555443043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon -0.0003438847030263783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon -0.00035981978898461613 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon -0.0003446342778566972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon -0.00035529584995236537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon -0.00034855740895831116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon -0.00034932634912802544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon -0.00035805518303064666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon -0.0003497941877073061 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon -0.00035774398685405447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon -0.0003560421780316607 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon -0.0003508844468369392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon -0.00035731928892270107 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon -0.0003557884626314314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon -0.00034992996760289355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon -0.000360752554360921 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon -0.0003452321668708545 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon -0.0003591745226131023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon -0.00035256981433229084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon -0.00035378123159712034 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon -0.000350464354895999 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon -0.00035074625557389677 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon -0.00035025894701994667 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon -0.00035437902514857614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon -0.0003514684519732232 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon -0.00035449717909633905 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon -0.0003436816402714221 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon -0.00035139158071782116 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon -0.0003509424079843335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon -0.000343894618577506 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon -0.0003500789770661659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon -0.0003407788080680086 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon -0.0003581908175239701 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon -0.0003465541618780918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon -0.00034600228792437736 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon -0.00034416738982773204 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon -0.0003519900340150641 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon -0.000343369616864659 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon -0.0003544993883274688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon -0.0003504441365073392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon -0.00034859160702727056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon -0.00035355909532647185 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon -0.0003471900922691849 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon -0.0003563015508709187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon -0.0003487888744148821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon -0.00034711767548688336 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon -0.0003530734609369085 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon -0.00035123969242560935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon -0.0003517127620891489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon -0.00035232835416868673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon -0.0003524437481912308 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon -0.0003525996167005602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon -0.00035064770545242043 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon -0.00035311558274981226 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon -0.00034952204800569914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon -0.0003541471367344846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon -0.00035418812454561825 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon -0.0003528951372900714 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon -0.0003542338042975688 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon -0.00034937738939942796 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon -0.0003522182190878447 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon -0.0003501406466507449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon -0.00034973079877492633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon -0.0003485274567713538 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon -0.00034999308679368985 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon -0.0003570051724707296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon -0.00034567230462019706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon -0.00035529000940160696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon -0.00034956512308671755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon -0.0003496962834028953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon -0.0003468745282493457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon -0.0003502717155809202 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon -0.0003556240880896514 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon -0.0003515109488424343 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon -0.0003563156688192592 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon -0.00035040277363989817 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon -0.0003481408593290717 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon -0.0003624575124332874 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon -0.0003522684124250313 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon -0.00035286996027653544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon -0.00034967623997256725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon -0.00035182649587602765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon -0.0003524892557026489 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon -0.0003507642477451811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon -0.00036190408389835666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon -0.00035102739424880766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon -0.00035239718753257265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon -0.00035298076121821316 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon -0.0003478704389752654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon -0.0003503109191567942 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon -0.00035143250975654426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon -0.0003480663923069012 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon -0.00035691540219998623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon -0.000348815437166351 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon -0.00035202073257766225 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon -0.0003491569096274706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon -0.00035277390475511834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon -0.0003524972090026609 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon -0.0003504854249750236 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon -0.00034740238025423914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon -0.00034968015462277606 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon -0.0003493798632762674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon -0.0003488202537862122 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon -0.0003525461864643725 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon -0.00034903815232825664 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon -0.00035536982539258216 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon -0.00034858083265155483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon -0.0003505014973608067 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon -0.00035327984042622104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon -0.0003503286677453136 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon -0.00035835274842442816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon -0.00034970302660275595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon -0.000357929573140149 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon -0.0003517238649788585 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon -0.00036097027318848475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon -0.0003502734074110026 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon -0.00035801510806036273 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon -0.0003568006373479869 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon -0.00036128108717454636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon -0.0003563436883111686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon -0.00035559725321852463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon -0.00035089656006854944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon -0.000359453964362057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon -0.00035629498059104033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon -0.0003622207707090437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon -0.0003540946784512821 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon -0.0003594750565232011 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon -0.0003566007415086991 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon -0.0003562142599126134 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon -0.0003569948186744601 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon -0.00035166554847920186 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon -0.00035047994419295137 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon -0.0003561578193739437 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon -0.00035470866838811544 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon -0.00034216920464876335 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon -0.0003550021513075795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon -0.0003488045105938729 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon -0.0003513340720840151 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon -0.0003448558566387584 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon -0.0003460966026953241 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon -0.0003488157616036459 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon -0.0003446120387842362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon -0.000351528602987427 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon -0.00035661118227454713 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon -0.0003551342699877457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon -0.0003478953397924445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon -0.00034625782458988215 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon -0.0003527515447405871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon -0.00034823744889805696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon -0.00034823314560254406 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon -0.00035162668292961944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon -0.0003477307716074623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon -0.0003446457989477787 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon -0.00034782916273767795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon -0.0003517249130302248 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon -0.0003449873430908556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon -0.00034841291749669877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon -0.0003466028498941749 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon -0.0003486436831199424 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon -0.0003478279234211838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon -0.0003495903653274374 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon -0.00034896893881218957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon -0.000348941645312426 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon -0.0003474221308416894 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon -0.0003462621543839385 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon -0.0003669373860863891 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon -0.00034691156268163006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon -0.0003527774103765281 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon -0.00034684565672734663 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon -0.0003454250599604457 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon -0.0003541536557159006 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon -0.000345735737037366 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon -0.0003524669816385214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon -0.0003441817133096468 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon -0.0003519093265859089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon -0.00035080085480352095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon -0.00035285227929327434 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon -0.00034354836346901676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon -0.00034789770937373467 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon -0.000343665920520102 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon -0.0003490884931060568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon -0.00034380029463398654 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon -0.00034874768005099945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon -0.0003457058510967673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon -0.00034644265227023904 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon -0.00035008339858594957 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon -0.0003462377193296194 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon -0.0003620491787114201 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon -0.000348717011044469 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon -0.00034370072363913706 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon -0.0003551981066775649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon -0.0003500119496799342 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon -0.0003485082952669081 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon -0.0003508155580978919 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon -0.00035311375163251416 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon -0.00034945972003423253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon -0.0003474220353789879 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon -0.0003536443686585001 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon -0.0003560350489042953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon -0.0003493655927914396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon -0.0003528423977146383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon -0.00035255554724471217 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon -0.0003479760010190111 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon -0.00035458598862501956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon -0.0003458990560538315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon -0.00035157946422379875 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon -0.00034736860650169996 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon -0.0003529152313394119 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon -0.00034586294329524465 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon -0.00035707214923794877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon -0.0003509580363496512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon -0.00035244176725524474 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon -0.0003467539557999047 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon -0.00034919687962275546 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon -0.00035094031731719953 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon -0.0003484309008351352 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon -0.0003485409424916253 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon -0.0003499590776117838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon -0.0003492842758957848 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon -0.0003529712275178912 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon -0.0003566141287087449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon -0.0003649496522047409 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon -0.0003563218912208234 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon -0.00035614782126966145 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon -0.0003531944298453266 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon -0.0003535950949566616 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon -0.0003544295554928795 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon -0.0003519908503740376 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon -0.00035752817626134463 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon -0.0003515322689589972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon -0.0003486893890307115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon -0.0003446520464889867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon -0.0003509421562481707 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon -0.00035335015702909084 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon -0.0003490178167345008 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon -0.0003520497821155174 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon -0.0003549762618908944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon -0.00035072190850833103 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon -0.0003542458638526423 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon -0.000352419194572916 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon -0.0003545102564672614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon -0.0003495437992331806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon -0.0003542843376993964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon -0.000352827529313958 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon -0.00035442506093223886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon -0.0003496970719044257 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon -0.0003553096424442362 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon -0.00034986845565067564 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon -0.000352131055186658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon -0.0003527021708198983 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon -0.00034905885414547214 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon -0.0003583433842468394 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon -0.00034409435202828383 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon -0.00034846410520871483 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon -0.0003554459991927314 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon -0.00035310507471843076 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon -0.000350028910786098 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon -0.00035049727458009896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon -0.0003519047735925826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon -0.0003513027429919726 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon -0.0003626947260354396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon -0.0003500087324849783 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon -0.0003618315726725285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon -0.0003535385113938023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon -0.0003487064058517615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon -0.0003618709124780938 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon -0.00035040070335625915 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon -0.0003506279032267829 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon -0.0003498435310527524 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon -0.0003554634749821431 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon -0.00035091209738758963 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon -0.00035034103678978573 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon -0.00035398931854386146 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon -0.00035495529304989485 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon -0.00036067883473356603 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon +0.0003547982093445404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0000_text_document falcon +0.00035934014428504944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0001_text_document falcon +0.00035707704501371544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0002_text_document falcon +0.00035287930712815354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0003_text_document falcon +0.00035977166728996823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0004_text_document falcon +0.0003581675664109838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0005_text_document falcon +0.0003548617059697185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0006_text_document falcon +0.0003639582000286208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0007_text_document falcon +0.00035375839698688127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0008_text_document falcon +0.0003743722020080678 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0009_text_document falcon +0.0003530399715341242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0010_text_document falcon +0.00035511875882752406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0011_text_document falcon +0.0003618733574783154 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0012_text_document falcon +0.00035185243285420104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0013_text_document falcon +0.0003541503739732106 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0014_text_document falcon +0.0003631679485751914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0015_text_document falcon +0.00035748045578182274 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0016_text_document falcon +0.0003606490690555877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0017_text_document falcon +0.0003626383296610091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0018_text_document falcon +0.00035442644361264756 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0019_text_document falcon +0.00035978370170539796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0020_text_document falcon +0.0003585562375341541 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0021_text_document falcon +0.0003601958372888019 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0022_text_document falcon +0.000350277765402227 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0023_text_document falcon +0.0003616521184211704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0024_text_document falcon +0.0003620625543608188 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0025_text_document falcon +0.0003560781983850704 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0026_text_document falcon +0.0003553209610592676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0027_text_document falcon +0.00035905348643915075 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0028_text_document falcon +0.00034744258805696526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0029_text_document falcon +0.00035462784035661496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0030_text_document falcon +0.00034768186175100895 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0031_text_document falcon +0.0003568534635532736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0032_text_document falcon +0.00035586511544371234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0033_text_document falcon +0.0003524567827568137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0034_text_document falcon +0.0003512453770426313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0035_text_document falcon +0.0003591792726468799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0036_text_document falcon +0.0003514024529343127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0037_text_document falcon +0.0003584880112586934 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0038_text_document falcon +0.00035133552916418045 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0039_text_document falcon +0.0003600811981350215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0040_text_document falcon +0.0003571663974228119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0041_text_document falcon +0.00035768103378874214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0042_text_document falcon +0.00035939205561113694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0043_text_document falcon +0.00035186773916029825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0044_text_document falcon +0.0003542829672490847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0045_text_document falcon +0.0003592783642898726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0046_text_document falcon +0.0003556367340099302 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0047_text_document falcon +0.00035391392271377027 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0048_text_document falcon +0.00035486725707484836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0049_text_document falcon +0.00034866743396828035 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0050_text_document falcon +0.0003517219808644735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0051_text_document falcon +0.00034874458549673823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0052_text_document falcon +0.000355773136961014 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0053_text_document falcon +0.00035611750387841917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0054_text_document falcon +0.00035305602013916315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0055_text_document falcon +0.0003578207127071924 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0056_text_document falcon +0.00035514635841943707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0057_text_document falcon +0.00034816946212866206 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0058_text_document falcon +0.0003512707269761496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0059_text_document falcon +0.0003483392117980654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0060_text_document falcon +0.0003572169607204321 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0061_text_document falcon +0.00035139153281660794 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0062_text_document falcon +0.00035536422129036537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0063_text_document falcon +0.000352017164107143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0064_text_document falcon +0.000351889550179365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0065_text_document falcon +0.000358759689953589 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0066_text_document falcon +0.0003569286079869268 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0067_text_document falcon +0.0003657752958602099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0068_text_document falcon +0.00035396127934790697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0069_text_document falcon +0.0003618565071224743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0070_text_document falcon +0.00035146051531973204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0071_text_document falcon +0.00036107135765783567 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0072_text_document falcon +0.00035019554279994576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0073_text_document falcon +0.00035567858879904983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0074_text_document falcon +0.0003504753174793183 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0075_text_document falcon +0.00035931140831329194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0076_text_document falcon +0.0003502967866002823 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0077_text_document falcon +0.0003532911801041972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0078_text_document falcon +0.0003583543013070199 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0079_text_document falcon +0.0003566243489931224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0080_text_document falcon +0.0003468752314799221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0081_text_document falcon +0.0003597840618138091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0082_text_document falcon +0.00035128822484768084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0083_text_document falcon +0.00035889496943437507 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0084_text_document falcon +0.000352400524650424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0085_text_document falcon +0.0003518689536768735 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0086_text_document falcon +0.00035866864741303467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0087_text_document falcon +0.0003454687659106334 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0088_text_document falcon +0.00035348007259317576 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0089_text_document falcon +0.0003539752270940644 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0090_text_document falcon +0.00035146495994081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0091_text_document falcon +0.00035397212846310423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0092_text_document falcon +0.00035208246467162587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0093_text_document falcon +0.0003490843168676626 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0094_text_document falcon +0.00035299633658644394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0095_text_document falcon +0.00034868327466167065 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0096_text_document falcon +0.00035941351365601583 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0097_text_document falcon +0.0003545343062735255 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0098_text_document falcon +0.0003528956380445978 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0099_text_document falcon +0.0003553355770443352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0100_text_document falcon +0.0003644224004937743 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0101_text_document falcon +0.00035234291036216907 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0102_text_document falcon +0.0003596237469847771 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0103_text_document falcon +0.0003531996065735989 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0104_text_document falcon +0.0003547177054106099 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0105_text_document falcon +0.0003575586499260483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0106_text_document falcon +0.00035262635135283667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0107_text_document falcon +0.0003624191962188944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0108_text_document falcon +0.0003488398052948616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0109_text_document falcon +0.0003598294093147917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0110_text_document falcon +0.00035583006534466323 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0111_text_document falcon +0.00035403139653225103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0112_text_document falcon +0.00036134702642187156 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0113_text_document falcon +0.0003573689927162834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0114_text_document falcon +0.0003577141131435527 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0115_text_document falcon +0.00035208814419277406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0116_text_document falcon +0.00035996720683665625 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0117_text_document falcon +0.00035415304658912596 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0118_text_document falcon +0.00036353353029443546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0119_text_document falcon +0.0003537326003150983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0120_text_document falcon +0.00036053976358299083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0121_text_document falcon +0.000352380489373494 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0122_text_document falcon +0.00036154661616900994 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0123_text_document falcon +0.00035959332325963614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0124_text_document falcon +0.0003597954667189692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0125_text_document falcon +0.0003563108270597542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0126_text_document falcon +0.0003582891940460143 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0127_text_document falcon +0.0003497728210484297 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0128_text_document falcon +0.0003549834902179354 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0129_text_document falcon +0.0003529828233484542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0130_text_document falcon +0.00034627483903285777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0131_text_document falcon +0.00035569006572589215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0132_text_document falcon +0.00035449377946910314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0133_text_document falcon +0.00035802844396194623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0134_text_document falcon +0.0003617277809353208 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0135_text_document falcon +0.00035034118898654814 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0136_text_document falcon +0.000351091193908611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0137_text_document falcon +0.0003527914342210668 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0138_text_document falcon +0.00035028288369781376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0139_text_document falcon +0.00035775745592780506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0140_text_document falcon +0.0003449630690661468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0141_text_document falcon +0.0003583490698830361 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0142_text_document falcon +0.0003476995746684122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0143_text_document falcon +0.0003535632505019212 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0144_text_document falcon +0.00035640180641147417 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0145_text_document falcon +0.000361731045691765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0146_text_document falcon +0.0003534082129597368 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0147_text_document falcon +0.0003550344149828664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0148_text_document falcon +0.00035363002411364057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0149_text_document falcon +0.0003537265579677396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0150_text_document falcon +0.00034950531383577937 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0151_text_document falcon +0.00035008511827347514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0152_text_document falcon +0.00035594533400871325 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0153_text_document falcon +0.00035266312861335946 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0154_text_document falcon +0.00035280268794863923 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0155_text_document falcon +0.0003565470391528536 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0156_text_document falcon +0.0003588492322689137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0157_text_document falcon +0.00035469909697832775 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0158_text_document falcon +0.00034712082813410526 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0159_text_document falcon +0.000348701157101807 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0160_text_document falcon +0.0003500192014479944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0161_text_document falcon +0.00035120560544669755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0162_text_document falcon +0.00035403656850437445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0163_text_document falcon +0.00035852376560749366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0164_text_document falcon +0.0003534754068111774 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0165_text_document falcon +0.00035591740046720765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0166_text_document falcon +0.000348522354782563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0167_text_document falcon +0.0003533533959664415 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0168_text_document falcon +0.00035631425964030697 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0169_text_document falcon +0.0003485886551574741 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0170_text_document falcon +0.00035917652631065777 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0171_text_document falcon +0.0003482975272111288 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0172_text_document falcon +0.00035580661277480167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0173_text_document falcon +0.0003492290722955348 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0174_text_document falcon +0.00034989284450240613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0175_text_document falcon +0.0003545677216162781 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0176_text_document falcon +0.00034622286859463484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0177_text_document falcon +0.00036070626989861965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0178_text_document falcon +0.00035518365036320786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0179_text_document falcon +0.00035272907057848406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0180_text_document falcon +0.0003547343638218734 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0181_text_document falcon +0.0003496450144966242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0182_text_document falcon +0.0003537407829294287 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0183_text_document falcon +0.0003489722653985685 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0184_text_document falcon +0.00035057186899911295 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0185_text_document falcon +0.0003507566548933051 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0186_text_document falcon +0.00035630360179023747 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0187_text_document falcon +0.00035631362503416367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0188_text_document falcon +0.0003490204248026821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0189_text_document falcon +0.00035761724058371226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0190_text_document falcon +0.00035037664777467137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0191_text_document falcon +0.000353402110481068 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0192_text_document falcon +0.00034524163568371745 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0193_text_document falcon +0.00035528523728570974 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0194_text_document falcon +0.00034784916132431703 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0195_text_document falcon +0.00034928476408048925 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0196_text_document falcon +0.00034989205973784984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0197_text_document falcon +0.00034201664404094254 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0198_text_document falcon +0.0003529676016338611 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0199_text_document falcon +0.00034643433682346637 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0200_text_document falcon +0.0003511666373001904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0201_text_document falcon +0.00034828669066575333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0202_text_document falcon +0.0003494625207264413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0203_text_document falcon +0.0003458957535879216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0204_text_document falcon +0.0003543020478990003 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0205_text_document falcon +0.00034754384069014956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0206_text_document falcon +0.0003598856392240133 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0207_text_document falcon +0.0003503335458553846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0208_text_document falcon +0.00035919595619778716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0209_text_document falcon +0.00035767737970754404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0210_text_document falcon +0.00035197152783998165 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0211_text_document falcon +0.0003549609834422404 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0212_text_document falcon +0.0003568184100569753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0213_text_document falcon +0.0003512652818651935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0214_text_document falcon +0.00035912648958665754 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0215_text_document falcon +0.00034764526964056546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0216_text_document falcon +0.000352439784960359 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0217_text_document falcon +0.00035295886560764226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0218_text_document falcon +0.0003518132693658672 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0219_text_document falcon +0.00035589987915465713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0220_text_document falcon +0.00034923863317385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0221_text_document falcon +0.0003457987267929692 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0222_text_document falcon +0.0003560928663480501 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0223_text_document falcon +0.0003529603811204932 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0224_text_document falcon +0.0003524438555443043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0225_text_document falcon +0.0003438847030263783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0226_text_document falcon +0.00035981978898461613 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0227_text_document falcon +0.0003446342778566972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0228_text_document falcon +0.00035529584995236537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0229_text_document falcon +0.00034855740895831116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0230_text_document falcon +0.00034932634912802544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0231_text_document falcon +0.00035805518303064666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0232_text_document falcon +0.0003497941877073061 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0233_text_document falcon +0.00035774398685405447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0234_text_document falcon +0.0003560421780316607 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0235_text_document falcon +0.0003508844468369392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0236_text_document falcon +0.00035731928892270107 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0237_text_document falcon +0.0003557884626314314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0238_text_document falcon +0.00034992996760289355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0239_text_document falcon +0.000360752554360921 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0240_text_document falcon +0.0003452321668708545 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0241_text_document falcon +0.0003591745226131023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0242_text_document falcon +0.00035256981433229084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0243_text_document falcon +0.00035378123159712034 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0244_text_document falcon +0.000350464354895999 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0245_text_document falcon +0.00035074625557389677 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0246_text_document falcon +0.00035025894701994667 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0247_text_document falcon +0.00035437902514857614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0248_text_document falcon +0.0003514684519732232 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0249_text_document falcon +0.00035449717909633905 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0250_text_document falcon +0.0003436816402714221 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0251_text_document falcon +0.00035139158071782116 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0252_text_document falcon +0.0003509424079843335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0253_text_document falcon +0.000343894618577506 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0254_text_document falcon +0.0003500789770661659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0255_text_document falcon +0.0003407788080680086 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0256_text_document falcon +0.0003581908175239701 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0257_text_document falcon +0.0003465541618780918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0258_text_document falcon +0.00034600228792437736 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0259_text_document falcon +0.00034416738982773204 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0260_text_document falcon +0.0003519900340150641 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0261_text_document falcon +0.000343369616864659 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0262_text_document falcon +0.0003544993883274688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0263_text_document falcon +0.0003504441365073392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0264_text_document falcon +0.00034859160702727056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0265_text_document falcon +0.00035355909532647185 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0266_text_document falcon +0.0003471900922691849 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0267_text_document falcon +0.0003563015508709187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0268_text_document falcon +0.0003487888744148821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0269_text_document falcon +0.00034711767548688336 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0270_text_document falcon +0.0003530734609369085 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0271_text_document falcon +0.00035123969242560935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0272_text_document falcon +0.0003517127620891489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0273_text_document falcon +0.00035232835416868673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0274_text_document falcon +0.0003524437481912308 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0275_text_document falcon +0.0003525996167005602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0276_text_document falcon +0.00035064770545242043 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0277_text_document falcon +0.00035311558274981226 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0278_text_document falcon +0.00034952204800569914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0279_text_document falcon +0.0003541471367344846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0280_text_document falcon +0.00035418812454561825 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0281_text_document falcon +0.0003528951372900714 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0282_text_document falcon +0.0003542338042975688 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0283_text_document falcon +0.00034937738939942796 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0284_text_document falcon +0.0003522182190878447 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0285_text_document falcon +0.0003501406466507449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0286_text_document falcon +0.00034973079877492633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0287_text_document falcon +0.0003485274567713538 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0288_text_document falcon +0.00034999308679368985 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0289_text_document falcon +0.0003570051724707296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0290_text_document falcon +0.00034567230462019706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0291_text_document falcon +0.00035529000940160696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0292_text_document falcon +0.00034956512308671755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0293_text_document falcon +0.0003496962834028953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0294_text_document falcon +0.0003468745282493457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0295_text_document falcon +0.0003502717155809202 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0296_text_document falcon +0.0003556240880896514 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0297_text_document falcon +0.0003515109488424343 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0298_text_document falcon +0.0003563156688192592 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0299_text_document falcon +0.00035040277363989817 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0300_text_document falcon +0.0003481408593290717 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0301_text_document falcon +0.0003624575124332874 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0302_text_document falcon +0.0003522684124250313 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0303_text_document falcon +0.00035286996027653544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0304_text_document falcon +0.00034967623997256725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0305_text_document falcon +0.00035182649587602765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0306_text_document falcon +0.0003524892557026489 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0307_text_document falcon +0.0003507642477451811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0308_text_document falcon +0.00036190408389835666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0309_text_document falcon +0.00035102739424880766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0310_text_document falcon +0.00035239718753257265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0311_text_document falcon +0.00035298076121821316 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0312_text_document falcon +0.0003478704389752654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0313_text_document falcon +0.0003503109191567942 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0314_text_document falcon +0.00035143250975654426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0315_text_document falcon +0.0003480663923069012 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0316_text_document falcon +0.00035691540219998623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0317_text_document falcon +0.000348815437166351 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0318_text_document falcon +0.00035202073257766225 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0319_text_document falcon +0.0003491569096274706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0320_text_document falcon +0.00035277390475511834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0321_text_document falcon +0.0003524972090026609 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0322_text_document falcon +0.0003504854249750236 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0323_text_document falcon +0.00034740238025423914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0324_text_document falcon +0.00034968015462277606 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0325_text_document falcon +0.0003493798632762674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0326_text_document falcon +0.0003488202537862122 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0327_text_document falcon +0.0003525461864643725 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0328_text_document falcon +0.00034903815232825664 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0329_text_document falcon +0.00035536982539258216 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0330_text_document falcon +0.00034858083265155483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0331_text_document falcon +0.0003505014973608067 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0332_text_document falcon +0.00035327984042622104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0333_text_document falcon +0.0003503286677453136 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0334_text_document falcon +0.00035835274842442816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0335_text_document falcon +0.00034970302660275595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0336_text_document falcon +0.000357929573140149 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0337_text_document falcon +0.0003517238649788585 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0338_text_document falcon +0.00036097027318848475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0339_text_document falcon +0.0003502734074110026 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0340_text_document falcon +0.00035801510806036273 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0341_text_document falcon +0.0003568006373479869 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0342_text_document falcon +0.00036128108717454636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0343_text_document falcon +0.0003563436883111686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0344_text_document falcon +0.00035559725321852463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0345_text_document falcon +0.00035089656006854944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0346_text_document falcon +0.000359453964362057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0347_text_document falcon +0.00035629498059104033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0348_text_document falcon +0.0003622207707090437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0349_text_document falcon +0.0003540946784512821 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0350_text_document falcon +0.0003594750565232011 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0351_text_document falcon +0.0003566007415086991 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0352_text_document falcon +0.0003562142599126134 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0353_text_document falcon +0.0003569948186744601 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0354_text_document falcon +0.00035166554847920186 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0355_text_document falcon +0.00035047994419295137 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0356_text_document falcon +0.0003561578193739437 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0357_text_document falcon +0.00035470866838811544 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0358_text_document falcon +0.00034216920464876335 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0359_text_document falcon +0.0003550021513075795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0360_text_document falcon +0.0003488045105938729 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0361_text_document falcon +0.0003513340720840151 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0362_text_document falcon +0.0003448558566387584 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0363_text_document falcon +0.0003460966026953241 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0364_text_document falcon +0.0003488157616036459 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0365_text_document falcon +0.0003446120387842362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0366_text_document falcon +0.000351528602987427 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0367_text_document falcon +0.00035661118227454713 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0368_text_document falcon +0.0003551342699877457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0369_text_document falcon +0.0003478953397924445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0370_text_document falcon +0.00034625782458988215 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0371_text_document falcon +0.0003527515447405871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0372_text_document falcon +0.00034823744889805696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0373_text_document falcon +0.00034823314560254406 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0374_text_document falcon +0.00035162668292961944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0375_text_document falcon +0.0003477307716074623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0376_text_document falcon +0.0003446457989477787 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0377_text_document falcon +0.00034782916273767795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0378_text_document falcon +0.0003517249130302248 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0379_text_document falcon +0.0003449873430908556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0380_text_document falcon +0.00034841291749669877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0381_text_document falcon +0.0003466028498941749 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0382_text_document falcon +0.0003486436831199424 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0383_text_document falcon +0.0003478279234211838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0384_text_document falcon +0.0003495903653274374 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0385_text_document falcon +0.00034896893881218957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0386_text_document falcon +0.000348941645312426 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0387_text_document falcon +0.0003474221308416894 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0388_text_document falcon +0.0003462621543839385 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0389_text_document falcon +0.0003669373860863891 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0390_text_document falcon +0.00034691156268163006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0391_text_document falcon +0.0003527774103765281 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0392_text_document falcon +0.00034684565672734663 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0393_text_document falcon +0.0003454250599604457 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0394_text_document falcon +0.0003541536557159006 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0395_text_document falcon +0.000345735737037366 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0396_text_document falcon +0.0003524669816385214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0397_text_document falcon +0.0003441817133096468 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0398_text_document falcon +0.0003519093265859089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0399_text_document falcon +0.00035080085480352095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0400_text_document falcon +0.00035285227929327434 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0401_text_document falcon +0.00034354836346901676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0402_text_document falcon +0.00034789770937373467 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0403_text_document falcon +0.000343665920520102 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0404_text_document falcon +0.0003490884931060568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0405_text_document falcon +0.00034380029463398654 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0406_text_document falcon +0.00034874768005099945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0407_text_document falcon +0.0003457058510967673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0408_text_document falcon +0.00034644265227023904 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0409_text_document falcon +0.00035008339858594957 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0410_text_document falcon +0.0003462377193296194 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0411_text_document falcon +0.0003620491787114201 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0412_text_document falcon +0.000348717011044469 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0413_text_document falcon +0.00034370072363913706 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0414_text_document falcon +0.0003551981066775649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0415_text_document falcon +0.0003500119496799342 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0416_text_document falcon +0.0003485082952669081 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0417_text_document falcon +0.0003508155580978919 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0418_text_document falcon +0.00035311375163251416 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0419_text_document falcon +0.00034945972003423253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0420_text_document falcon +0.0003474220353789879 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0421_text_document falcon +0.0003536443686585001 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0422_text_document falcon +0.0003560350489042953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0423_text_document falcon +0.0003493655927914396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0424_text_document falcon +0.0003528423977146383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0425_text_document falcon +0.00035255554724471217 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0426_text_document falcon +0.0003479760010190111 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0427_text_document falcon +0.00035458598862501956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0428_text_document falcon +0.0003458990560538315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0429_text_document falcon +0.00035157946422379875 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0430_text_document falcon +0.00034736860650169996 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0431_text_document falcon +0.0003529152313394119 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0432_text_document falcon +0.00034586294329524465 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0433_text_document falcon +0.00035707214923794877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0434_text_document falcon +0.0003509580363496512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0435_text_document falcon +0.00035244176725524474 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0436_text_document falcon +0.0003467539557999047 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0437_text_document falcon +0.00034919687962275546 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0438_text_document falcon +0.00035094031731719953 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0439_text_document falcon +0.0003484309008351352 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0440_text_document falcon +0.0003485409424916253 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0441_text_document falcon +0.0003499590776117838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0442_text_document falcon +0.0003492842758957848 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0443_text_document falcon +0.0003529712275178912 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0444_text_document falcon +0.0003566141287087449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0445_text_document falcon +0.0003649496522047409 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0446_text_document falcon +0.0003563218912208234 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0447_text_document falcon +0.00035614782126966145 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0448_text_document falcon +0.0003531944298453266 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0449_text_document falcon +0.0003535950949566616 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0450_text_document falcon +0.0003544295554928795 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0451_text_document falcon +0.0003519908503740376 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0452_text_document falcon +0.00035752817626134463 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0453_text_document falcon +0.0003515322689589972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0454_text_document falcon +0.0003486893890307115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0455_text_document falcon +0.0003446520464889867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0456_text_document falcon +0.0003509421562481707 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0457_text_document falcon +0.00035335015702909084 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0458_text_document falcon +0.0003490178167345008 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0459_text_document falcon +0.0003520497821155174 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0460_text_document falcon +0.0003549762618908944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0461_text_document falcon +0.00035072190850833103 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0462_text_document falcon +0.0003542458638526423 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0463_text_document falcon +0.000352419194572916 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0464_text_document falcon +0.0003545102564672614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0465_text_document falcon +0.0003495437992331806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0466_text_document falcon +0.0003542843376993964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0467_text_document falcon +0.000352827529313958 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0468_text_document falcon +0.00035442506093223886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0469_text_document falcon +0.0003496970719044257 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0470_text_document falcon +0.0003553096424442362 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0471_text_document falcon +0.00034986845565067564 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0472_text_document falcon +0.000352131055186658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0473_text_document falcon +0.0003527021708198983 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0474_text_document falcon +0.00034905885414547214 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0475_text_document falcon +0.0003583433842468394 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0476_text_document falcon +0.00034409435202828383 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0477_text_document falcon +0.00034846410520871483 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0478_text_document falcon +0.0003554459991927314 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0479_text_document falcon +0.00035310507471843076 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0480_text_document falcon +0.000350028910786098 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0481_text_document falcon +0.00035049727458009896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0482_text_document falcon +0.0003519047735925826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0483_text_document falcon +0.0003513027429919726 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0484_text_document falcon +0.0003626947260354396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0485_text_document falcon +0.0003500087324849783 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0486_text_document falcon +0.0003618315726725285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0487_text_document falcon +0.0003535385113938023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0488_text_document falcon +0.0003487064058517615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0489_text_document falcon +0.0003618709124780938 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0490_text_document falcon +0.00035040070335625915 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0491_text_document falcon +0.0003506279032267829 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0492_text_document falcon +0.0003498435310527524 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0493_text_document falcon +0.0003554634749821431 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0494_text_document falcon +0.00035091209738758963 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0495_text_document falcon +0.00035034103678978573 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0496_text_document falcon +0.00035398931854386146 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0497_text_document falcon +0.00035495529304989485 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0498_text_document falcon +0.00036067883473356603 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/falcon-0499_text_document falcon diff --git a/ALCF/data-lists/aurora/megawiki.txt b/ALCF/data-lists/aurora/megawiki.txt index 1e8e6672a2d..635eba3d906 100644 --- a/ALCF/data-lists/aurora/megawiki.txt +++ b/ALCF/data-lists/aurora/megawiki.txt @@ -1,262 +1,262 @@ -6.322825248625475e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika -2.4432314037946264e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika -5.6313888721313454e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika -2.4208171781595055e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika -2.325811856369237e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika -2.4010790356322705e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika -5.36773610843632e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika -1.360574433501002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika -1.3076540344853244e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika -1.3386534334886313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika -1.2498103719605153e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika -1.403763836949682e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika -1.3636756723495417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika -1.2242489446940814e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika -1.2398255818973339e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika -1.2972616994216281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika -1.3947809855914134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika -1.3144843787829514e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika -1.1693809976572487e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika -1.3677252682893802e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika -1.3940876719849597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika -1.4222245138730965e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika -1.3201677767919704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika -1.1421717796486169e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika -1.2890514724498703e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika -1.3649507648749037e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika -1.2400732563490717e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika -1.1557681453277616e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika -1.2294483595964517e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika -1.2137484472122283e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika -1.3299663426456e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika -1.2461984216479532e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika -1.4666434217609636e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika -1.1876997894686238e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika -1.2939155338964078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika -1.3859590039728515e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika -1.317917848615668e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika -1.1335281536110342e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika -1.2889923952861426e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika -1.3471671647053326e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika -1.2221720014475102e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika -1.2632647276287541e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika -1.28276219004076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika -1.36213704321643e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika -1.2414858625261553e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika -1.3173700421883744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika -1.295597796725686e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika -1.242783936442904e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika -1.2417374088427464e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika -1.2134479405400744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika -1.3090040663304255e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika -1.2713470581614905e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika -5.5750231378906594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika -5.777597358425469e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika -5.349786767471258e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika -5.675165050453583e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika -5.482611216158831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika -5.065421899890121e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika -5.384718357480146e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika -4.872037363236061e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika -4.532709250783155e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika -5.7257963030489613e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika -4.9014365579652036e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika -5.722863552770969e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika -6.149911636146833e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika -5.2178057608273506e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika -4.990228161160431e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika -5.866186875255134e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika -5.004185734360719e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika -4.79401853705107e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika -5.435219965052376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika -5.035997225792266e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika -5.622401774211625e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika -5.028826157387559e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika -5.596379470128795e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika -6.027824493191489e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika -5.5358270009931474e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika -5.9839051807685496e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika -5.1221077499249595e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika -5.517228560620279e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika -5.1687858285052305e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika -5.684188244145645e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika -5.212693275535878e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika -4.8551007022784084e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika -5.4888506639203145e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika -5.345098688527242e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika -4.8506420625516594e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika -5.132168603397676e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika -5.719476795114223e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika -5.7448621149792696e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika -4.9068410568059265e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika -5.382937299647678e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika -4.8288432136304634e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika -5.841703200305416e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika -5.1589611587885584e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika -6.031113829732574e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika -5.4558202844532094e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika -5.341852317196142e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika -5.1402942738369954e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika -5.735421384377395e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika -5.473629863586958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika -5.4708993245733936e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika -4.931161863634078e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika -5.104173022127248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika -5.510157161510824e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika -5.652501401782597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika -5.7273656573031666e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika -5.638363224821738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika -5.6128115396668704e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika -5.00304877998141e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika -5.596120554779096e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika -5.5280923889040006e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika -5.223477917938408e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika -5.29472809986569e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika -2.205682378243213e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika -1.4367563720603185e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika -3.5506193487931076e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika -3.0442910855821778e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika -2.2540042508019627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika -2.6880163202623216e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika -2.534473148048727e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika -2.6560945431318916e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika -2.547470248967691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika -2.5248825388073738e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika -2.5828729575000054e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika -2.4026583817957736e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika -2.3930425429834413e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika -2.5037365362599724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika -2.6696745470595603e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika -2.140323051341762e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika -2.617354786691592e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika -1.538359101762691e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika -1.2871029252377856e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika -2.255195411289217e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika -2.4832313897952067e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika -9.303873918189968e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika -2.179532302620228e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika -1.9750517506901206e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika -2.7740420380648435e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika -2.7813714782319335e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika -4.1595357937609806e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika -2.741365122389175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika -2.117451071361901e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika -1.7132649760565998e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika -1.7492547092602047e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika -1.7499951097392276e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika -1.6632444789170958e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika -1.6678802252361607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika -1.5519208704558896e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika -1.652420992967167e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika -1.6119931034508755e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika -1.6638882076736552e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika -1.7198076782652946e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika -1.572927860565175e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika -1.5194822618169918e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika -1.6677776832669846e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika -1.595612492245688e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika -1.682350633181197e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika -1.663983380609724e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika -1.710187842689243e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika -1.5733697527539038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika -1.6972104757911438e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika -1.6610142847616577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika -1.61094882403031e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika -1.4789207305138325e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika -1.639299617676302e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika -1.3241204512116132e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika -8.582260726625535e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika -8.213000975576739e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika -9.549247732811947e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika -9.17242785339013e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika -7.632868223725218e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika -8.674401118222175e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika -9.124384255505347e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika -8.344222222417358e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika -8.992299957499065e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika -8.76689497361025e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika -7.973396239586015e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika -9.006935606644125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika -8.725545954955498e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika -1.215449694669174e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika -3.3041720284158646e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika -2.0593512412624502e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika -1.893608946986248e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika -1.737111666788535e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika -1.4915923449873955e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika -2.289370239067605e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika -2.8615335689614638e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika -8.847283630883125e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika -1.8175470362373804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika -1.8152226683368038e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika -1.789149655314284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika -1.7690523036477663e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika -1.8333732213753644e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika -1.8794105687718654e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika -1.721841156706417e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika -2.0612008685724796e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika -1.9297370681336376e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika -2.0188440409661018e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika -5.1741216329695265e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika -1.3417913926038429e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika -1.1010813016469651e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika -1.1252416134320087e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika -1.2801744104313002e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika -1.3041514955795817e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika -1.3428837580879075e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika -1.320809382267804e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika -1.3451566676555968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika -1.228284926657501e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika -1.2410599573923043e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika -1.3815343367377182e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika -1.3895126265148832e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika -1.2306773644401741e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika -1.32981021906281e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika -1.101337469221607e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika -1.513094184404692e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika -1.1073759547073234e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika -1.2879348765857567e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika -9.619595770228435e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika -1.2384340836286436e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika -1.1766667232211577e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika -1.2871049236196452e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika -1.2010645926497744e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika -1.3971428231518597e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika -1.2283733550547932e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika -1.2659530508255308e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika -1.551775613074462e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika -1.1169413343776979e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika -1.1433700593712463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika -4.964773647323492e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika -1.0995586595687313e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika -1.2957393071411267e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika -2.75899247407709e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika -2.8269344597344854e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika -2.329108187246831e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika -2.4231761430460284e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika -1.2434140512230442e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika -1.638718338352859e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika -3.272953556801187e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika -6.061314500486327e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika -1.2465979731210292e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika -1.2737557327967737e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika -1.038428658075627e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika -2.61666472045566e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika -3.6506873212272224e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika -1.5066359138295701e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika -1.1166290872121178e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika -1.5546966228590285e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika -1.2583434625014828e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika -1.3398826881300862e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika -1.2944933160515968e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika -1.0971437399901365e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika -1.2787922795775774e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika -1.404979227816985e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika -1.3344734431324463e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika -4.886031157107555e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika -3.277261443596394e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika -3.5057957685786495e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika -3.287625301718589e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika -3.1370056372668855e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika -3.186092015785841e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika -7.271819324142512e-06 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika +6.322825248625475e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0000_text_document megawika +2.4432314037946264e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0001_text_document megawika +5.6313888721313454e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0002_text_document megawika +2.4208171781595055e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0003_text_document megawika +2.325811856369237e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0004_text_document megawika +2.4010790356322705e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0005_text_document megawika +5.36773610843632e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0006_text_document megawika +1.360574433501002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0007_text_document megawika +1.3076540344853244e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0008_text_document megawika +1.3386534334886313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0009_text_document megawika +1.2498103719605153e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0010_text_document megawika +1.403763836949682e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0011_text_document megawika +1.3636756723495417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0012_text_document megawika +1.2242489446940814e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0013_text_document megawika +1.2398255818973339e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0014_text_document megawika +1.2972616994216281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0015_text_document megawika +1.3947809855914134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0016_text_document megawika +1.3144843787829514e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0017_text_document megawika +1.1693809976572487e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0018_text_document megawika +1.3677252682893802e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0019_text_document megawika +1.3940876719849597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0020_text_document megawika +1.4222245138730965e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0021_text_document megawika +1.3201677767919704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0022_text_document megawika +1.1421717796486169e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0023_text_document megawika +1.2890514724498703e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0024_text_document megawika +1.3649507648749037e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0025_text_document megawika +1.2400732563490717e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0026_text_document megawika +1.1557681453277616e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0027_text_document megawika +1.2294483595964517e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0028_text_document megawika +1.2137484472122283e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0029_text_document megawika +1.3299663426456e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0030_text_document megawika +1.2461984216479532e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0031_text_document megawika +1.4666434217609636e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0032_text_document megawika +1.1876997894686238e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0033_text_document megawika +1.2939155338964078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0034_text_document megawika +1.3859590039728515e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0035_text_document megawika +1.317917848615668e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0036_text_document megawika +1.1335281536110342e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0037_text_document megawika +1.2889923952861426e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0038_text_document megawika +1.3471671647053326e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0039_text_document megawika +1.2221720014475102e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0040_text_document megawika +1.2632647276287541e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0041_text_document megawika +1.28276219004076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0042_text_document megawika +1.36213704321643e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0043_text_document megawika +1.2414858625261553e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0044_text_document megawika +1.3173700421883744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0045_text_document megawika +1.295597796725686e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0046_text_document megawika +1.242783936442904e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0047_text_document megawika +1.2417374088427464e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0048_text_document megawika +1.2134479405400744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0049_text_document megawika +1.3090040663304255e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0050_text_document megawika +1.2713470581614905e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0051_text_document megawika +5.5750231378906594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0052_text_document megawika +5.777597358425469e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0053_text_document megawika +5.349786767471258e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0054_text_document megawika +5.675165050453583e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0055_text_document megawika +5.482611216158831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0056_text_document megawika +5.065421899890121e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0057_text_document megawika +5.384718357480146e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0058_text_document megawika +4.872037363236061e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0059_text_document megawika +4.532709250783155e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0060_text_document megawika +5.7257963030489613e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0061_text_document megawika +4.9014365579652036e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0062_text_document megawika +5.722863552770969e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0063_text_document megawika +6.149911636146833e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0064_text_document megawika +5.2178057608273506e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0065_text_document megawika +4.990228161160431e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0066_text_document megawika +5.866186875255134e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0067_text_document megawika +5.004185734360719e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0068_text_document megawika +4.79401853705107e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0069_text_document megawika +5.435219965052376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0070_text_document megawika +5.035997225792266e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0071_text_document megawika +5.622401774211625e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0072_text_document megawika +5.028826157387559e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0073_text_document megawika +5.596379470128795e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0074_text_document megawika +6.027824493191489e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0075_text_document megawika +5.5358270009931474e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0076_text_document megawika +5.9839051807685496e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0077_text_document megawika +5.1221077499249595e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0078_text_document megawika +5.517228560620279e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0079_text_document megawika +5.1687858285052305e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0080_text_document megawika +5.684188244145645e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0081_text_document megawika +5.212693275535878e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0082_text_document megawika +4.8551007022784084e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0083_text_document megawika +5.4888506639203145e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0084_text_document megawika +5.345098688527242e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0085_text_document megawika +4.8506420625516594e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0086_text_document megawika +5.132168603397676e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0087_text_document megawika +5.719476795114223e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0088_text_document megawika +5.7448621149792696e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0089_text_document megawika +4.9068410568059265e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0090_text_document megawika +5.382937299647678e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0091_text_document megawika +4.8288432136304634e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0092_text_document megawika +5.841703200305416e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0093_text_document megawika +5.1589611587885584e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0094_text_document megawika +6.031113829732574e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0095_text_document megawika +5.4558202844532094e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0096_text_document megawika +5.341852317196142e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0097_text_document megawika +5.1402942738369954e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0098_text_document megawika +5.735421384377395e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0099_text_document megawika +5.473629863586958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0100_text_document megawika +5.4708993245733936e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0101_text_document megawika +4.931161863634078e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0102_text_document megawika +5.104173022127248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0103_text_document megawika +5.510157161510824e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0104_text_document megawika +5.652501401782597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0105_text_document megawika +5.7273656573031666e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0106_text_document megawika +5.638363224821738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0107_text_document megawika +5.6128115396668704e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0108_text_document megawika +5.00304877998141e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0109_text_document megawika +5.596120554779096e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0110_text_document megawika +5.5280923889040006e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0111_text_document megawika +5.223477917938408e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0112_text_document megawika +5.29472809986569e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0113_text_document megawika +2.205682378243213e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0114_text_document megawika +1.4367563720603185e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0115_text_document megawika +3.5506193487931076e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0116_text_document megawika +3.0442910855821778e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0117_text_document megawika +2.2540042508019627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0118_text_document megawika +2.6880163202623216e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0119_text_document megawika +2.534473148048727e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0120_text_document megawika +2.6560945431318916e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0121_text_document megawika +2.547470248967691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0122_text_document megawika +2.5248825388073738e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0123_text_document megawika +2.5828729575000054e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0124_text_document megawika +2.4026583817957736e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0125_text_document megawika +2.3930425429834413e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0126_text_document megawika +2.5037365362599724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0127_text_document megawika +2.6696745470595603e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0128_text_document megawika +2.140323051341762e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0129_text_document megawika +2.617354786691592e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0130_text_document megawika +1.538359101762691e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0131_text_document megawika +1.2871029252377856e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0132_text_document megawika +2.255195411289217e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0133_text_document megawika +2.4832313897952067e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0134_text_document megawika +9.303873918189968e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0135_text_document megawika +2.179532302620228e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0136_text_document megawika +1.9750517506901206e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0137_text_document megawika +2.7740420380648435e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0138_text_document megawika +2.7813714782319335e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0139_text_document megawika +4.1595357937609806e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0140_text_document megawika +2.741365122389175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0141_text_document megawika +2.117451071361901e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0142_text_document megawika +1.7132649760565998e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0143_text_document megawika +1.7492547092602047e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0144_text_document megawika +1.7499951097392276e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0145_text_document megawika +1.6632444789170958e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0146_text_document megawika +1.6678802252361607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0147_text_document megawika +1.5519208704558896e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0148_text_document megawika +1.652420992967167e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0149_text_document megawika +1.6119931034508755e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0150_text_document megawika +1.6638882076736552e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0151_text_document megawika +1.7198076782652946e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0152_text_document megawika +1.572927860565175e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0153_text_document megawika +1.5194822618169918e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0154_text_document megawika +1.6677776832669846e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0155_text_document megawika +1.595612492245688e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0156_text_document megawika +1.682350633181197e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0157_text_document megawika +1.663983380609724e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0158_text_document megawika +1.710187842689243e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0159_text_document megawika +1.5733697527539038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0160_text_document megawika +1.6972104757911438e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0161_text_document megawika +1.6610142847616577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0162_text_document megawika +1.61094882403031e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0163_text_document megawika +1.4789207305138325e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0164_text_document megawika +1.639299617676302e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0165_text_document megawika +1.3241204512116132e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0166_text_document megawika +8.582260726625535e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0167_text_document megawika +8.213000975576739e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0168_text_document megawika +9.549247732811947e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0169_text_document megawika +9.17242785339013e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0170_text_document megawika +7.632868223725218e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0171_text_document megawika +8.674401118222175e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0172_text_document megawika +9.124384255505347e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0173_text_document megawika +8.344222222417358e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0174_text_document megawika +8.992299957499065e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0175_text_document megawika +8.76689497361025e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0176_text_document megawika +7.973396239586015e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0177_text_document megawika +9.006935606644125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0178_text_document megawika +8.725545954955498e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0179_text_document megawika +1.215449694669174e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0180_text_document megawika +3.3041720284158646e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0181_text_document megawika +2.0593512412624502e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0182_text_document megawika +1.893608946986248e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0183_text_document megawika +1.737111666788535e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0184_text_document megawika +1.4915923449873955e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0185_text_document megawika +2.289370239067605e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0186_text_document megawika +2.8615335689614638e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0187_text_document megawika +8.847283630883125e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0188_text_document megawika +1.8175470362373804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0189_text_document megawika +1.8152226683368038e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0190_text_document megawika +1.789149655314284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0191_text_document megawika +1.7690523036477663e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0192_text_document megawika +1.8333732213753644e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0193_text_document megawika +1.8794105687718654e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0194_text_document megawika +1.721841156706417e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0195_text_document megawika +2.0612008685724796e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0196_text_document megawika +1.9297370681336376e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0197_text_document megawika +2.0188440409661018e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0198_text_document megawika +5.1741216329695265e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0199_text_document megawika +1.3417913926038429e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0200_text_document megawika +1.1010813016469651e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0201_text_document megawika +1.1252416134320087e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0202_text_document megawika +1.2801744104313002e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0203_text_document megawika +1.3041514955795817e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0204_text_document megawika +1.3428837580879075e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0205_text_document megawika +1.320809382267804e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0206_text_document megawika +1.3451566676555968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0207_text_document megawika +1.228284926657501e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0208_text_document megawika +1.2410599573923043e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0209_text_document megawika +1.3815343367377182e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0210_text_document megawika +1.3895126265148832e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0211_text_document megawika +1.2306773644401741e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0212_text_document megawika +1.32981021906281e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0213_text_document megawika +1.101337469221607e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0214_text_document megawika +1.513094184404692e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0215_text_document megawika +1.1073759547073234e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0216_text_document megawika +1.2879348765857567e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0217_text_document megawika +9.619595770228435e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0218_text_document megawika +1.2384340836286436e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0219_text_document megawika +1.1766667232211577e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0220_text_document megawika +1.2871049236196452e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0221_text_document megawika +1.2010645926497744e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0222_text_document megawika +1.3971428231518597e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0223_text_document megawika +1.2283733550547932e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0224_text_document megawika +1.2659530508255308e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0225_text_document megawika +1.551775613074462e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0226_text_document megawika +1.1169413343776979e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0227_text_document megawika +1.1433700593712463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0228_text_document megawika +4.964773647323492e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0229_text_document megawika +1.0995586595687313e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0230_text_document megawika +1.2957393071411267e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0231_text_document megawika +2.75899247407709e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0232_text_document megawika +2.8269344597344854e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0233_text_document megawika +2.329108187246831e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0234_text_document megawika +2.4231761430460284e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0235_text_document megawika +1.2434140512230442e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0236_text_document megawika +1.638718338352859e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0237_text_document megawika +3.272953556801187e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0238_text_document megawika +6.061314500486327e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0239_text_document megawika +1.2465979731210292e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0240_text_document megawika +1.2737557327967737e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0241_text_document megawika +1.038428658075627e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0242_text_document megawika +2.61666472045566e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0243_text_document megawika +3.6506873212272224e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0244_text_document megawika +1.5066359138295701e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0245_text_document megawika +1.1166290872121178e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0246_text_document megawika +1.5546966228590285e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0247_text_document megawika +1.2583434625014828e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0248_text_document megawika +1.3398826881300862e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0249_text_document megawika +1.2944933160515968e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0250_text_document megawika +1.0971437399901365e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0251_text_document megawika +1.2787922795775774e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0252_text_document megawika +1.404979227816985e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0253_text_document megawika +1.3344734431324463e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0254_text_document megawika +4.886031157107555e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0255_text_document megawika +3.277261443596394e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0256_text_document megawika +3.5057957685786495e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0257_text_document megawika +3.287625301718589e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0258_text_document megawika +3.1370056372668855e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0259_text_document megawika +3.186092015785841e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0260_text_document megawika +7.271819324142512e-06 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/megawika-0261_text_document megawika diff --git a/ALCF/data-lists/aurora/open-web-math-train.txt b/ALCF/data-lists/aurora/open-web-math-train.txt index 0bef3e40f7a..e0dfc30bd77 100644 --- a/ALCF/data-lists/aurora/open-web-math-train.txt +++ b/ALCF/data-lists/aurora/open-web-math-train.txt @@ -1,13 +1,13 @@ -0.001451215788905126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train -0.0014486847196258788 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train -0.0008861032722895899 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train -0.0018119590809459816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train -0.0008916937917547129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train -6.960128832809415e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train -0.002008403651063623 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train -0.0014374900742131454 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train -0.00180213596996716 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train -0.001956178877532413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train -0.0008829547017667033 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train -0.0008910853619157279 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train -0.0018260998845299973 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train +0.001451215788905126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0000_text_document open-web-math-train +0.0014486847196258788 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0001_text_document open-web-math-train +0.0008861032722895899 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0002_text_document open-web-math-train +0.0018119590809459816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0003_text_document open-web-math-train +0.0008916937917547129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0004_text_document open-web-math-train +6.960128832809415e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0005_text_document open-web-math-train +0.002008403651063623 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0006_text_document open-web-math-train +0.0014374900742131454 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0007_text_document open-web-math-train +0.00180213596996716 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0008_text_document open-web-math-train +0.001956178877532413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0009_text_document open-web-math-train +0.0008829547017667033 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0010_text_document open-web-math-train +0.0008910853619157279 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0011_text_document open-web-math-train +0.0018260998845299973 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/open-web-math-train-0012_text_document open-web-math-train diff --git a/ALCF/data-lists/aurora/pes2o.txt b/ALCF/data-lists/aurora/pes2o.txt index c1e8429f4b8..3d0cdbe4790 100644 --- a/ALCF/data-lists/aurora/pes2o.txt +++ b/ALCF/data-lists/aurora/pes2o.txt @@ -1,26 +1,26 @@ -0.0012499632072059553 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o -0.00125398260359913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o -0.0012541704774729071 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o -0.0012527268234360602 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o -0.0012532925243737164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o -0.0012456396241204315 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o -0.0012589894424352072 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o -0.001508020123999618 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o -0.00333096950781965 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o -0.0033233414614415547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o -0.003512387990689828 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o -0.0035091382940513126 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o -0.003514155927147005 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o -0.003327108000579638 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o -0.003329106196589836 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o -0.003505604148738077 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o -0.003324825759567855 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o -0.0033248240149804913 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o -0.0033385962112851358 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o -0.0035043186296553615 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o -0.003340469505431529 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o -0.0035106889084796276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o -0.0033309469281030167 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o -0.003340337858029757 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o -0.003505919861097801 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o -0.0003882924098240512 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o +0.0012499632072059553 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0000_text_document pes2o +0.00125398260359913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0001_text_document pes2o +0.0012541704774729071 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0002_text_document pes2o +0.0012527268234360602 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0003_text_document pes2o +0.0012532925243737164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0004_text_document pes2o +0.0012456396241204315 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0005_text_document pes2o +0.0012589894424352072 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0006_text_document pes2o +0.001508020123999618 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0007_text_document pes2o +0.00333096950781965 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0008_text_document pes2o +0.0033233414614415547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0009_text_document pes2o +0.003512387990689828 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0010_text_document pes2o +0.0035091382940513126 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0011_text_document pes2o +0.003514155927147005 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0012_text_document pes2o +0.003327108000579638 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0013_text_document pes2o +0.003329106196589836 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0014_text_document pes2o +0.003505604148738077 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0015_text_document pes2o +0.003324825759567855 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0016_text_document pes2o +0.0033248240149804913 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0017_text_document pes2o +0.0033385962112851358 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0018_text_document pes2o +0.0035043186296553615 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0019_text_document pes2o +0.003340469505431529 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0020_text_document pes2o +0.0035106889084796276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0021_text_document pes2o +0.0033309469281030167 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0022_text_document pes2o +0.003340337858029757 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0023_text_document pes2o +0.003505919861097801 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0024_text_document pes2o +0.0003882924098240512 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/pes2o-0025_text_document pes2o diff --git a/ALCF/data-lists/aurora/reddit.txt b/ALCF/data-lists/aurora/reddit.txt index e01a4a1b866..ebc1e15ada8 100644 --- a/ALCF/data-lists/aurora/reddit.txt +++ b/ALCF/data-lists/aurora/reddit.txt @@ -1,78 +1,78 @@ -0.0005759963691850877 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit -0.0005959971675332674 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit -0.0006026179290353799 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit -0.0005824184320784846 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit -0.0005854598548616037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit -0.0005903767055633473 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit -0.0005930306490982049 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit -0.000569425602700746 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit -0.0005675060415179408 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit -0.0005772431621253389 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit -0.0005678026053826858 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit -0.0005700398263483378 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit -0.0005669467963528824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit -0.0005701015953324305 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit -0.0005795907287413296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit -0.0005735602737531164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit -0.0005749862745842101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit -0.0005693257015931971 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit -0.0005716568794795563 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit -0.0005761083919774021 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit -0.0005688343169797355 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit -0.0005807913190929842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit -0.0005710229258078636 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit -0.0005704083039826862 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit -0.0005862132348308056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit -0.0005717662049559556 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit -0.0005858155213694451 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit -0.0005812012281792392 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit -0.0005803981414588498 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit -0.0005700102108287723 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit -0.0005719243459052329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit -0.0005867253401661752 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit -0.0005731087218860733 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit -0.0005712197789109317 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit -0.0005702376926310089 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit -0.0005700411527742972 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit -0.0005828090098178196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit -0.0005770140826168056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit -0.0005723509664597896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit -0.0005755499231836962 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit -0.0005636407438471367 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit -0.0005640281556500104 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit -0.0005633159058766496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit -0.0005638034311151449 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit -0.0005630066273073224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit -0.0005631803831128559 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit -0.0005631228881679657 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit -0.0005628178701487633 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit -0.0005624448092256196 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit -0.0005620957024062329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit -0.0005614201504177484 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit -0.0005616890951464056 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit -0.0005611348559279058 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit -0.0005604238061828518 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit -0.0005603301490194237 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit -0.0005607291294548833 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit -0.0005605234569930727 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit -0.0005613778566640694 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit -0.0005610248539992471 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit -0.0005599977416780475 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit -0.0005603632562116935 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit -0.0005599177479509897 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit -0.0005595202318298379 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit -0.0005600975633499175 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit -0.0005614075491213365 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit -0.000612563885043477 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit -0.0005515469909644413 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit -0.0005526782014946906 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit -0.0005472463408095445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit -0.0005502284746004587 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit -0.0005414514790555363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit -0.0005513499500134784 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit -0.0005391391454105187 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit -0.0005415836910001838 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit -0.0005208132468536551 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit -0.0005889827143132871 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit -0.0005822520817765276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit -0.0004173155230758696 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit +0.0005759963691850877 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0000_text_document reddit +0.0005959971675332674 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0001_text_document reddit +0.0006026179290353799 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0002_text_document reddit +0.0005824184320784846 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0003_text_document reddit +0.0005854598548616037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0004_text_document reddit +0.0005903767055633473 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0005_text_document reddit +0.0005930306490982049 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0006_text_document reddit +0.000569425602700746 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0007_text_document reddit +0.0005675060415179408 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0008_text_document reddit +0.0005772431621253389 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0009_text_document reddit +0.0005678026053826858 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0010_text_document reddit +0.0005700398263483378 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0011_text_document reddit +0.0005669467963528824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0012_text_document reddit +0.0005701015953324305 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0013_text_document reddit +0.0005795907287413296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0014_text_document reddit +0.0005735602737531164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0015_text_document reddit +0.0005749862745842101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0016_text_document reddit +0.0005693257015931971 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0017_text_document reddit +0.0005716568794795563 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0018_text_document reddit +0.0005761083919774021 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0019_text_document reddit +0.0005688343169797355 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0020_text_document reddit +0.0005807913190929842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0021_text_document reddit +0.0005710229258078636 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0022_text_document reddit +0.0005704083039826862 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0023_text_document reddit +0.0005862132348308056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0024_text_document reddit +0.0005717662049559556 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0025_text_document reddit +0.0005858155213694451 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0026_text_document reddit +0.0005812012281792392 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0027_text_document reddit +0.0005803981414588498 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0028_text_document reddit +0.0005700102108287723 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0029_text_document reddit +0.0005719243459052329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0030_text_document reddit +0.0005867253401661752 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0031_text_document reddit +0.0005731087218860733 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0032_text_document reddit +0.0005712197789109317 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0033_text_document reddit +0.0005702376926310089 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0034_text_document reddit +0.0005700411527742972 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0035_text_document reddit +0.0005828090098178196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0036_text_document reddit +0.0005770140826168056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0037_text_document reddit +0.0005723509664597896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0038_text_document reddit +0.0005755499231836962 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0039_text_document reddit +0.0005636407438471367 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0040_text_document reddit +0.0005640281556500104 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0041_text_document reddit +0.0005633159058766496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0042_text_document reddit +0.0005638034311151449 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0043_text_document reddit +0.0005630066273073224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0044_text_document reddit +0.0005631803831128559 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0045_text_document reddit +0.0005631228881679657 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0046_text_document reddit +0.0005628178701487633 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0047_text_document reddit +0.0005624448092256196 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0048_text_document reddit +0.0005620957024062329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0049_text_document reddit +0.0005614201504177484 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0050_text_document reddit +0.0005616890951464056 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0051_text_document reddit +0.0005611348559279058 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0052_text_document reddit +0.0005604238061828518 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0053_text_document reddit +0.0005603301490194237 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0054_text_document reddit +0.0005607291294548833 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0055_text_document reddit +0.0005605234569930727 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0056_text_document reddit +0.0005613778566640694 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0057_text_document reddit +0.0005610248539992471 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0058_text_document reddit +0.0005599977416780475 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0059_text_document reddit +0.0005603632562116935 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0060_text_document reddit +0.0005599177479509897 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0061_text_document reddit +0.0005595202318298379 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0062_text_document reddit +0.0005600975633499175 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0063_text_document reddit +0.0005614075491213365 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0064_text_document reddit +0.000612563885043477 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0065_text_document reddit +0.0005515469909644413 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0066_text_document reddit +0.0005526782014946906 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0067_text_document reddit +0.0005472463408095445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0068_text_document reddit +0.0005502284746004587 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0069_text_document reddit +0.0005414514790555363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0070_text_document reddit +0.0005513499500134784 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0071_text_document reddit +0.0005391391454105187 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0072_text_document reddit +0.0005415836910001838 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0073_text_document reddit +0.0005208132468536551 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0074_text_document reddit +0.0005889827143132871 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0075_text_document reddit +0.0005822520817765276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0076_text_document reddit +0.0004173155230758696 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/reddit-0077_text_document reddit diff --git a/ALCF/data-lists/aurora/stack.txt b/ALCF/data-lists/aurora/stack.txt index b2defb3b370..d99516f5fb3 100644 --- a/ALCF/data-lists/aurora/stack.txt +++ b/ALCF/data-lists/aurora/stack.txt @@ -1,26 +1,26 @@ -0.0009994361338078242 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange -0.001087156194657966 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange -0.0010667737163656816 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange -0.0009602877882124873 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange -0.0008968956271971105 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange -0.0009198034843762967 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange -0.0009423901016715341 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange -0.0009674094553686345 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange -0.0009858331322519164 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange -0.0009970593645879198 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange -0.0010027035193731686 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange -0.0010128291154221853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange -0.0010215631382631918 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange -0.0010288663771461238 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange -0.0010346219929285867 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange -0.00104544019940344 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange -0.0010525172676724333 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange -0.0010609529620775127 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange -0.0010725892748610153 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange -0.0010818563598181568 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange -0.0010992760196793917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange -0.0011178992762079917 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange -0.001124687532085676 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange -0.001118303661267191 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange -0.0010206825575416534 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange -0.0005512280117499715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange +0.0009994361338078242 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0000_text_document stackexchange +0.001087156194657966 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0001_text_document stackexchange +0.0010667737163656816 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0002_text_document stackexchange +0.0009602877882124873 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0003_text_document stackexchange +0.0008968956271971105 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0004_text_document stackexchange +0.0009198034843762967 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0005_text_document stackexchange +0.0009423901016715341 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0006_text_document stackexchange +0.0009674094553686345 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0007_text_document stackexchange +0.0009858331322519164 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0008_text_document stackexchange +0.0009970593645879198 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0009_text_document stackexchange +0.0010027035193731686 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0010_text_document stackexchange +0.0010128291154221853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0011_text_document stackexchange +0.0010215631382631918 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0012_text_document stackexchange +0.0010288663771461238 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0013_text_document stackexchange +0.0010346219929285867 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0014_text_document stackexchange +0.00104544019940344 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0015_text_document stackexchange +0.0010525172676724333 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0016_text_document stackexchange +0.0010609529620775127 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0017_text_document stackexchange +0.0010725892748610153 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0018_text_document stackexchange +0.0010818563598181568 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0019_text_document stackexchange +0.0010992760196793917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0020_text_document stackexchange +0.0011178992762079917 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0021_text_document stackexchange +0.001124687532085676 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0022_text_document stackexchange +0.001118303661267191 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0023_text_document stackexchange +0.0010206825575416534 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0024_text_document stackexchange +0.0005512280117499715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/stackexchange-0025_text_document stackexchange diff --git a/ALCF/data-lists/aurora/starcoder.txt b/ALCF/data-lists/aurora/starcoder.txt index a6436675f7b..2a5be0cf72d 100644 --- a/ALCF/data-lists/aurora/starcoder.txt +++ b/ALCF/data-lists/aurora/starcoder.txt @@ -1,50 +1,50 @@ -0.004474659408857016 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder -0.00409944473890653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder -0.005137179939941845 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder -0.005143172251066109 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder -0.005206134363352808 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder -0.004892747858974329 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder -0.004844731352552902 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder -0.005308320169123755 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder -0.005124709815666577 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder -0.005424710744483826 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder -0.00538244648861977 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder -0.0029107284679086853 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder -0.0026825258998444705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder -0.0026904503191419243 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder -0.002687906577174073 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder -0.002850165346048818 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder -0.005322698571717847 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder -0.004450334290869719 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder -0.004700990083440683 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder -0.003903568556500995 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder -0.00390561515396931 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder -0.0039046402900912262 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder -0.003907454839379547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder -0.0038583224578603824 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder -0.0037914116657695 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder -0.003786665266798682 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder -0.003792000802430658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder -0.00319266847466091 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder -0.0032658716699838944 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder -0.0034801959532460023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder -0.0028307012092022594 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder -0.0028420360878146276 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder -0.0028410455248484914 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder -0.00283497183526842 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder -0.002840187195459487 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder -0.0028398709431369834 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder -0.004364722843422023 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder -0.004093255713117101 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder -0.004092331079566252 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder -0.004005326985579649 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder -0.0036205502856964207 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder -0.003625316793034984 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder -0.003604743435602363 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder -0.0035405823343673125 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder -0.0041601413517253945 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder -0.005886303658937057 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder -0.003600909532810332 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder -0.0034941365817168658 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder -0.0004992164842980224 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder +0.004474659408857016 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0000_text_document starcoder +0.00409944473890653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0001_text_document starcoder +0.005137179939941845 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0002_text_document starcoder +0.005143172251066109 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0003_text_document starcoder +0.005206134363352808 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0004_text_document starcoder +0.004892747858974329 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0005_text_document starcoder +0.004844731352552902 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0006_text_document starcoder +0.005308320169123755 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0007_text_document starcoder +0.005124709815666577 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0008_text_document starcoder +0.005424710744483826 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0009_text_document starcoder +0.00538244648861977 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0010_text_document starcoder +0.0029107284679086853 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0011_text_document starcoder +0.0026825258998444705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0012_text_document starcoder +0.0026904503191419243 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0013_text_document starcoder +0.002687906577174073 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0014_text_document starcoder +0.002850165346048818 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0015_text_document starcoder +0.005322698571717847 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0016_text_document starcoder +0.004450334290869719 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0017_text_document starcoder +0.004700990083440683 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0018_text_document starcoder +0.003903568556500995 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0019_text_document starcoder +0.00390561515396931 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0020_text_document starcoder +0.0039046402900912262 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0021_text_document starcoder +0.003907454839379547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0022_text_document starcoder +0.0038583224578603824 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0023_text_document starcoder +0.0037914116657695 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0024_text_document starcoder +0.003786665266798682 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0025_text_document starcoder +0.003792000802430658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0026_text_document starcoder +0.00319266847466091 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0027_text_document starcoder +0.0032658716699838944 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0028_text_document starcoder +0.0034801959532460023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0029_text_document starcoder +0.0028307012092022594 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0030_text_document starcoder +0.0028420360878146276 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0031_text_document starcoder +0.0028410455248484914 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0032_text_document starcoder +0.00283497183526842 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0033_text_document starcoder +0.002840187195459487 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0034_text_document starcoder +0.0028398709431369834 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0035_text_document starcoder +0.004364722843422023 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0036_text_document starcoder +0.004093255713117101 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0037_text_document starcoder +0.004092331079566252 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0038_text_document starcoder +0.004005326985579649 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0039_text_document starcoder +0.0036205502856964207 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0040_text_document starcoder +0.003625316793034984 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0041_text_document starcoder +0.003604743435602363 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0042_text_document starcoder +0.0035405823343673125 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0043_text_document starcoder +0.0041601413517253945 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0044_text_document starcoder +0.005886303658937057 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0045_text_document starcoder +0.003600909532810332 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0046_text_document starcoder +0.0034941365817168658 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0047_text_document starcoder +0.0004992164842980224 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/starcoder-0048_text_document starcoder diff --git a/ALCF/data-lists/aurora/tulu.txt b/ALCF/data-lists/aurora/tulu.txt index bafc809a8c9..46b3a91a40c 100644 --- a/ALCF/data-lists/aurora/tulu.txt +++ b/ALCF/data-lists/aurora/tulu.txt @@ -1,66 +1,66 @@ -0.00032927705604725614 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu -0.0002860154190878753 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu -0.0002845217585425619 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu -0.0002743528685497456 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu -0.00026025323737738766 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu -0.00023493876414603155 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu -0.00029665994994226705 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu -0.00031808102075993956 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu -0.00031813573046011285 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu -0.0002711905171855542 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu -0.00028892513401817095 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu -0.00030003908676979083 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu -0.00026839878771944684 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu -0.00029155935002690497 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu -0.0002998624927624209 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu -0.0003091705447974841 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu -0.00026873195794309786 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu -0.00027721873498527547 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu -0.0002841662554024377 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu -0.0002839461156551537 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu -0.0002861705604659811 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu -0.0002460995649635886 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu -0.00019420142619795496 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu -0.00021967677816173628 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu -0.0002620283200480949 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu -0.0002433390542188936 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu -0.00021254976608350767 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu -0.00022094815569522115 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu -0.000342862378668244 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu -0.00033784225259118157 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu -0.0003367278459543952 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu -0.00029843279042852765 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu -0.0002926583661257988 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu -0.00029320337282010673 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu -0.00029281450669483455 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu -0.0002915338187002653 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu -0.0002864226923084572 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu -0.00028643439083586396 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu -0.00028253710956299054 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu -0.0002810856078805806 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu -0.00031474941344656715 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu -0.0002139130222205655 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu -0.0003084648871862831 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu -0.0003309477872140129 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu -0.0003360096824695161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu -0.0003355452655196557 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu -0.00038119390366386037 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu -0.00038078927630086064 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu -0.0003386200917551554 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu -0.0002158905159938882 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu -0.00021621682877018768 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu -0.00021553306942740535 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu -0.00021581563462722296 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu -0.0002157694110556169 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu -0.000215643699847159 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu -0.00021532716715168094 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu -0.00021531221326022472 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu -0.0002831801179028896 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu -0.0002514844936507595 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu -0.00031638782778107964 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu -0.0002749197545278445 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu -0.00026159721512464495 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu -0.0002630052420096968 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu -0.00031106811228913666 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu -0.0002852973415334161 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu -3.7555372465932136e-05 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu +0.00032927705604725614 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0000_text_document tulu +0.0002860154190878753 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0001_text_document tulu +0.0002845217585425619 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0002_text_document tulu +0.0002743528685497456 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0003_text_document tulu +0.00026025323737738766 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0004_text_document tulu +0.00023493876414603155 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0005_text_document tulu +0.00029665994994226705 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0006_text_document tulu +0.00031808102075993956 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0007_text_document tulu +0.00031813573046011285 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0008_text_document tulu +0.0002711905171855542 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0009_text_document tulu +0.00028892513401817095 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0010_text_document tulu +0.00030003908676979083 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0011_text_document tulu +0.00026839878771944684 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0012_text_document tulu +0.00029155935002690497 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0013_text_document tulu +0.0002998624927624209 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0014_text_document tulu +0.0003091705447974841 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0015_text_document tulu +0.00026873195794309786 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0016_text_document tulu +0.00027721873498527547 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0017_text_document tulu +0.0002841662554024377 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0018_text_document tulu +0.0002839461156551537 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0019_text_document tulu +0.0002861705604659811 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0020_text_document tulu +0.0002460995649635886 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0021_text_document tulu +0.00019420142619795496 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0022_text_document tulu +0.00021967677816173628 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0023_text_document tulu +0.0002620283200480949 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0024_text_document tulu +0.0002433390542188936 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0025_text_document tulu +0.00021254976608350767 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0026_text_document tulu +0.00022094815569522115 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0027_text_document tulu +0.000342862378668244 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0028_text_document tulu +0.00033784225259118157 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0029_text_document tulu +0.0003367278459543952 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0030_text_document tulu +0.00029843279042852765 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0031_text_document tulu +0.0002926583661257988 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0032_text_document tulu +0.00029320337282010673 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0033_text_document tulu +0.00029281450669483455 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0034_text_document tulu +0.0002915338187002653 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0035_text_document tulu +0.0002864226923084572 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0036_text_document tulu +0.00028643439083586396 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0037_text_document tulu +0.00028253710956299054 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0038_text_document tulu +0.0002810856078805806 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0039_text_document tulu +0.00031474941344656715 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0040_text_document tulu +0.0002139130222205655 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0041_text_document tulu +0.0003084648871862831 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0042_text_document tulu +0.0003309477872140129 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0043_text_document tulu +0.0003360096824695161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0044_text_document tulu +0.0003355452655196557 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0045_text_document tulu +0.00038119390366386037 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0046_text_document tulu +0.00038078927630086064 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0047_text_document tulu +0.0003386200917551554 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0048_text_document tulu +0.0002158905159938882 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0049_text_document tulu +0.00021621682877018768 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0050_text_document tulu +0.00021553306942740535 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0051_text_document tulu +0.00021581563462722296 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0052_text_document tulu +0.0002157694110556169 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0053_text_document tulu +0.000215643699847159 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0054_text_document tulu +0.00021532716715168094 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0055_text_document tulu +0.00021531221326022472 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0056_text_document tulu +0.0002831801179028896 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0057_text_document tulu +0.0002514844936507595 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0058_text_document tulu +0.00031638782778107964 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0059_text_document tulu +0.0002749197545278445 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0060_text_document tulu +0.00026159721512464495 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0061_text_document tulu +0.0002630052420096968 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0062_text_document tulu +0.00031106811228913666 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0063_text_document tulu +0.0002852973415334161 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0064_text_document tulu +3.7555372465932136e-05 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/tulu_flan-0065_text_document tulu diff --git a/ALCF/data-lists/aurora/wiki.txt b/ALCF/data-lists/aurora/wiki.txt index 65a8ac92a87..c70a54f5987 100644 --- a/ALCF/data-lists/aurora/wiki.txt +++ b/ALCF/data-lists/aurora/wiki.txt @@ -1,2 +1,2 @@ -0.003548077173506675 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki -0.0018372203137874265 /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki +0.003548077173506675 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0000_text_document wiki +0.0018372203137874265 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/wiki-0001_text_document wiki From ea0c3c762b2c1fef35898d8952a6129a31f1645d Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Fri, 30 Aug 2024 15:09:11 -0500 Subject: [PATCH 307/590] fixed dftracer compatibility --- megatron/utils.py | 73 ++++++++++++++++++++++++++++------------------- 1 file changed, 44 insertions(+), 29 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 7a1d60bab92..19eef703998 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -36,12 +36,22 @@ _DLIO_PROFILER_EXIST = True +_DFTRACER_EXIST=True + +try: + import dftracer +except: + _DFTRACER_EXIST=True + try: import dlio_profiler except: _DLIO_PROFILER_EXIST = False -if _DLIO_PROFILER_EXIST: + +if _DFTRACER_EXIST: + from dftracer.logger import dftracer as PerfTrace, dft_fn as Profile, DFTRACER_ENABLE as DFTRACER_ENABLE +elif _DLIO_PROFILER_EXIST: from dlio_profiler.logger import fn_interceptor as Profile from dlio_profiler.logger import dlio_logger as PerfTrace else: @@ -50,40 +60,45 @@ # Profile: ContextManager = nullcontext # # class Profile(nullable_schema) - - class Profile: - def __init__(self, type="PROFILER"): - self._start = time.perf_counter() - self.type = type - - def log(self, func): + class Profile(object): + def __init__(self, cat, name=None, epoch=None, step=None, image_idx=None, image_size=None): + return + def log(self, func): + return func + def log_init(self, func): + return func + def iter(self, func, iter_name="step"): return func - - def iter(self, a): - return a - def __enter__(self): - self._start = time.perf_counter() - - def __exit__(self, *args, **kwargs): - dt = time.perf_counter() - self._start - log.info(f"{self.type} took: {dt:.6f}s") - - - class dlio_logger: - def __init__( - self, - ): + return + def __exit__(self, type, value, traceback): + return + def update(self, epoch=None, step=None, image_idx=None, image_size=None, args={}): + return + def flush(self): + return + def reset(self): + return + def log_static(self, func): + return + class dftracer(object): + def __init__(self,): self.type = None - def initialize_log(self, logfile=None, data_dir=None, process_id=-1): return + def get_time(self): + return + def enter_event(self): + return + def exit_event(self): + return + def log_event(self, name, cat, start_time, duration, string_args=None): + return + def finalize(self): + return - def iter(self, a): - return a - - PerfTrace = dlio_logger() - + PerfTrace = dftracer() + DFTRACER_ENABLE = False def get_logger( name: str, From 50e27296e10559c462160a3e965551ea92d49ac2 Mon Sep 17 00:00:00 2001 From: Varuni Date: Sat, 31 Aug 2024 14:11:23 +0000 Subject: [PATCH 308/590] hf cp conversion and inference scripts added --- mds_to_hf.py | 176 ++++++++++++++++++++-------------------------- test_agptllama.py | 34 +++++++++ 2 files changed, 110 insertions(+), 100 deletions(-) create mode 100644 test_agptllama.py diff --git a/mds_to_hf.py b/mds_to_hf.py index 9f2d87cc486..d91513ed8b0 100644 --- a/mds_to_hf.py +++ b/mds_to_hf.py @@ -1,130 +1,106 @@ -# Usage : python mds_to_hf.py --mds_checkpoint --output_dir -# Tips : Do not run on login node. -# This script currently only takes care of tp=1. Takes a AuroraGPT Llama model trained with Megatron-DeepSpeed and converts to LLamaCausalForLM architecture from HuggingFace. +# Usage : python mds_to_hf.py --mds_checkpoint --output_dir --cache-dir /flare/Aurora_deployment/vsastry +# Tips : Do not run on login node. +# This script currently only takes care of tp=1. Takes a AuroraGPT Llama model trained with Megatron-DeepSpeed and converts to LLamaCausalForLM architecture from HuggingFace. import argparse import torch -import pdb import os from transformers import LlamaConfig, LlamaForCausalLM, LlamaTokenizer - -def repeat_kv_wt(x, np): +def repeat_kv_wt(x,np): return torch.repeat_interleave(x, dim=0, repeats=np) - def Update_llama_config(Llama_config, mds_args): - if mds_args["swiglu"]: + if mds_args['swiglu']: Llama_config.hidden_act = "silu" - Llama_config.hidden_size = mds_args["hidden_size"] - Llama_config.intermediate_size = mds_args["ffn_hidden_size"] - Llama_config.max_position_embeddings = mds_args["max_position_embeddings"] - Llama_config.num_attention_heads = mds_args["num_attention_heads"] - Llama_config.num_hidden_layers = mds_args["num_layers"] - Llama_config.num_key_value_heads = mds_args["num_key_value_heads"] - Llama_config.rms_norm_eps = mds_args["layernorm_epsilon"] - Llama_config.rope_theta = mds_args["rope_theta"] - Llama_config.vocab_size = mds_args["padded_vocab_size"] + Llama_config.hidden_size = mds_args['hidden_size'] + Llama_config.intermediate_size = mds_args['ffn_hidden_size'] + Llama_config.max_position_embeddings = mds_args['max_position_embeddings'] + Llama_config.num_attention_heads = mds_args['num_attention_heads'] + Llama_config.num_hidden_layers = mds_args['num_layers'] + Llama_config.num_key_value_heads = mds_args['num_key_value_heads'] + Llama_config.rms_norm_eps = mds_args['layernorm_epsilon'] + Llama_config.rope_theta = mds_args['rope_theta'] + Llama_config.vocab_size = mds_args['padded_vocab_size'] + if mds_args['fp16'] == True: + Llama_config.torch_dtype = 'float16' + elif mds_args['bf16'] == True: + Llama_config.torch_dtype = 'bfloat16' return Llama_config if __name__ == "__main__": - from pathlib import Path - parser = argparse.ArgumentParser() - parser.add_argument("--mds_checkpoint", required=True) - parser.add_argument("--output_dir", required=True) + parser.add_argument('--mds_checkpoint', required=True) + parser.add_argument('--output_dir', required=True) + parser.add_argument('--cache_dir', required=True) args = parser.parse_args() # make output_dir if it does not exits. if not os.path.exists(args.output_dir): os.makedirs(args.output_dir) - filename = str(args.mds_checkpoint) - if not filename.split("/")[-1].startswith("mp_rank") and not filename.split("/")[ - -1 - ].endswith(".pt"): - assert "Provide the right file path, The file should be of format mp_rank_*.pt" + filename = str(args.mds_checkpoint) + if not filename.split("/")[-1].startswith('mp_rank') and not filename.split("/")[-1].endswith('.pt'): + assert ("Provide the right file path, The file should be of format mp_rank_*.pt") print(f"loading mds checkpoint {filename}") - - cache_dir = Path(os.getcwd()).joinpath(".cache", "hugging_face") - cache_dir.mkdir(exist_ok=True, parents=True) - - mds_model = torch.load(args.mds_checkpoint, map_location=torch.device("cpu")) - Llama_model = LlamaForCausalLM.from_pretrained( - "meta-llama/Llama-2-7b-hf", cache_dir=cache_dir.as_posix() - ) - Llama_config = Llama_model.config - Updated_Llama_config = Update_llama_config(Llama_config, mds_model["args"].__dict__) - # save the updated config.json file - Updated_Llama_config.to_json_file(os.path.join(args.output_dir, "config.json")) + + mds_model = torch.load(args.mds_checkpoint,map_location=torch.device('cpu')) + Llama_model = LlamaForCausalLM.from_pretrained("meta-llama/Llama-2-7b-hf",cache_dir=args.cache_dir) + + Llama_config = Llama_model.config + Updated_Llama_config = Update_llama_config(Llama_config, mds_model['args'].__dict__) + # save the updated config.json file + Updated_Llama_config.to_json_file(os.path.join(args.output_dir,'config.json')) state_dict = {} - dim = mds_model["args"].__dict__["kv_channels"] - inv_freq = 1.0 / ( - mds_model["args"].__dict__["rope_theta"] - ** (torch.arange(0, dim, 2).float() / dim) - ) - hidden_size = mds_model["args"].__dict__["hidden_size"] - kv_dim = ( - mds_model["args"].__dict__["kv_channels"] - * mds_model["args"].__dict__["num_key_value_heads"] - ) - kv_groups = ( - mds_model["args"].__dict__["num_attention_heads"] - // mds_model["args"].__dict__["num_key_value_heads"] - ) - for layer_i in range(Updated_Llama_config.__dict__["num_hidden_layers"]): + dim = mds_model['args'].__dict__['kv_channels'] + inv_freq = 1.0 / (mds_model['args'].__dict__['rope_theta'] ** (torch.arange(0,dim, 2).float() / dim)) + hidden_size = mds_model['args'].__dict__['hidden_size'] + kv_dim = mds_model['args'].__dict__['kv_channels'] * mds_model['args'].__dict__['num_key_value_heads'] + kv_groups = mds_model['args'].__dict__['num_attention_heads'] // mds_model['args'].__dict__['num_key_value_heads'] + nkvheads = mds_model['args'].__dict__['num_key_value_heads'] + for layer_i in range(Updated_Llama_config.__dict__['num_hidden_layers']): # SELF ATTENTION layers. - # get the q, k, v weights separately. Keeping k and v at the GQA head dim, since the transformers/models/llama/modelling_utils will take care of it. - fused_qkv = mds_model["module"]["language_model"]["encoder"][ - f"layers.{layer_i}.self_attention.query_key_value.weight" - ] - state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = fused_qkv[ - 0:hidden_size - ] - state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = fused_qkv[ - hidden_size : hidden_size + kv_dim - ] - # state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size:hidden_size+kv_dim], kv_groups) - state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = fused_qkv[ - hidden_size + kv_dim : hidden_size + 2 * kv_dim - ] - # state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim],kv_groups) - state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = mds_model[ - "module" - ]["language_model"]["encoder"][f"layers.{layer_i}.self_attention.dense.weight"] - # MLP Layers - fused_mlp = mds_model["module"]["language_model"]["encoder"][ - f"layers.{layer_i}.mlp.dense_h_to_4h.weight" - ] - chunked_mlp = torch.chunk(fused_mlp, 2, dim=0) + # get the q, k, v weights separately. Keeping k and v at the GQA head dim, since the transformers/models/llama/modelling_utils will take care of it. + fused_qkv = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.query_key_value.weight"] + fused_reshape = fused_qkv.view(nkvheads,(kv_groups+2)*dim,hidden_size) + ex_q = fused_reshape[:,:kv_groups*dim,:] + con_q = ex_q.contiguous().view(-1, fused_reshape.size(2)) + + ex_k = fused_reshape[:,kv_groups*dim:(kv_groups+1)*dim,:] + con_k = ex_k.contiguous().view(-1, fused_reshape.size(2)) + + ex_v = fused_reshape[:,(kv_groups+1)*dim:(kv_groups+2)*dim,:] + con_v = ex_v.contiguous().view(-1, fused_reshape.size(2)) + + state_dict[f"model.layers.{layer_i}.self_attn.q_proj.weight"] = con_q + state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = con_k + #state_dict[f"model.layers.{layer_i}.self_attn.k_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size:hidden_size+kv_dim], kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = con_v + #state_dict[f"model.layers.{layer_i}.self_attn.v_proj.weight"] = repeat_kv_wt(fused_qkv[hidden_size+kv_dim:hidden_size+2*kv_dim],kv_groups) + state_dict[f"model.layers.{layer_i}.self_attn.o_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.self_attention.dense.weight"] + + # MLP Layers + fused_mlp = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_h_to_4h.weight"] + chunked_mlp = torch.chunk(fused_mlp,2,dim=0) state_dict[f"model.layers.{layer_i}.mlp.gate_proj.weight"] = chunked_mlp[0] state_dict[f"model.layers.{layer_i}.mlp.up_proj.weight"] = chunked_mlp[1] - state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = mds_model[ - "module" - ]["language_model"]["encoder"][f"layers.{layer_i}.mlp.dense_4h_to_h.weight"] - # LayerNorm weights and RoPe - state_dict[f"model.layers.{layer_i}.input_layernorm.weight"] = mds_model[ - "module" - ]["language_model"]["encoder"][f"layers.{layer_i}.input_layernorm.weight"] - state_dict[ - f"model.layers.{layer_i}.post_attention_layernorm.weight" - ] = mds_model["module"]["language_model"]["encoder"][ - f"layers.{layer_i}.post_attention_layernorm.weight" - ] + state_dict[f"model.layers.{layer_i}.mlp.down_proj.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.mlp.dense_4h_to_h.weight"] + + #LayerNorm weights and RoPe + state_dict[f"model.layers.{layer_i}.input_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.input_layernorm.weight"] + state_dict[f"model.layers.{layer_i}.post_attention_layernorm.weight"] = mds_model['module']['language_model']['encoder'][f"layers.{layer_i}.post_attention_layernorm.weight"] state_dict[f"model.layers.{layer_i}.self_attn.rotary_emb.inv_freq"] = inv_freq - # Get the non-encoder layer weights. - state_dict["model.embed_tokens.weight"] = mds_model["module"]["language_model"][ - "embedding" - ]["word_embeddings"]["weight"] - state_dict["model.norm.weight"] = mds_model["module"]["language_model"]["encoder"][ - "final_layernorm.weight" - ] - state_dict["lm_head.weight"] = mds_model["module"]["language_model"][ - "output_layer" - ]["weight"] - # Save the model in the hf output path. - torch.save(state_dict, os.path.join(args.output_dir, "pytorch_model.bin")) + # Get the non-encoder layer weights. + state_dict["model.embed_tokens.weight"] = mds_model['module']['language_model']['embedding']['word_embeddings']['weight'] + state_dict["model.norm.weight"] = mds_model['module']['language_model']['encoder']['final_layernorm.weight'] + state_dict["lm_head.weight"] = mds_model['module']['language_model']['output_layer']['weight'] + + # Save the model in the hf output path. + torch.save(state_dict, os.path.join(args.output_dir,"pytorch_model.bin")) + + + diff --git a/test_agptllama.py b/test_agptllama.py new file mode 100644 index 00000000000..e1d207fa278 --- /dev/null +++ b/test_agptllama.py @@ -0,0 +1,34 @@ +import torch +import intel_extension_for_pytorch as ipex +from transformers import GPT2Tokenizer, AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, LlamaTokenizer, LlamaForCausalLM +def batch_encode(prompts, tokenizer, prompt_len=512): + input_tokens = tokenizer.batch_encode_plus(prompts, return_tensors="pt", padding="max_length", max_length=len(prompts)) + for t in input_tokens: + if torch.is_tensor(input_tokens[t]): + input_tokens[t] = input_tokens[t].to("xpu") + #input_tokens[t] = input_tokens[t].to(torch.cuda.current_device()) + return input_tokens + + +def generate_prompt(model, tokenizer, prompts): + + input_tokens = batch_encode(prompts, tokenizer) + print(input_tokens) + generate_kwargs = dict(max_new_tokens=30, do_sample=False) + output_ids = model.generate(**input_tokens, **generate_kwargs) + print(output_ids) + outputs = tokenizer.batch_decode(output_ids, skip_special_tokens=True) + + return outputs + +if __name__ == '__main__': + + model = LlamaForCausalLM.from_pretrained("/flare/Aurora_deployment/vsastry/hf_new_cp/") + model.to("xpu") # model.cuda() + model.seqlen = 4096 + + # get llama tokenizer + tokenizer = LlamaTokenizer.from_pretrained("/flare/Aurora_deployment/AuroraGPT/datasets/dolma/utils/tokenizer.model") + tokenizer.pad_token = tokenizer.eos_token + output = generate_prompt(model, tokenizer, prompts=["What is the language spoken in Mexico ?"]) + print(output) From a0ac7503d99560ba2f968c816812fabf81b3608a Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Tue, 3 Sep 2024 14:55:24 -0500 Subject: [PATCH 309/590] added requirements.txt --- requirements.txt | 3 +++ 1 file changed, 3 insertions(+) create mode 100644 requirements.txt diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 00000000000..4f758bed7d9 --- /dev/null +++ b/requirements.txt @@ -0,0 +1,3 @@ +pydftracer +wandb +git+https://github.com/saforem2/ezpz@main From 0d6e3793a1fc06eded9764ef15ad12bcc0281101 Mon Sep 17 00:00:00 2001 From: Polisetty V R K Jyothendra Varma Date: Wed, 4 Sep 2024 11:12:52 +0530 Subject: [PATCH 310/590] Enable Sequence Parallelism (#429) --- megatron/arguments.py | 3 ++- megatron/core/tensor_parallel/layers.py | 4 +++- megatron/model/fused_layer_norm.py | 14 ++++++++++++-- megatron/model/gpt_model.py | 8 ++++++-- megatron/model/language_model.py | 5 +++-- megatron/model/rmsnorm.py | 17 +++++++++++++++-- megatron/model/transformer.py | 13 +++++++++---- 7 files changed, 50 insertions(+), 14 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 49b3d8e4c66..9228da6ee9f 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -399,7 +399,8 @@ def validate_args(args, defaults={}): args.async_tensor_model_parallel_allreduce = False if not args.use_dataset_only: - if os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": + if deepspeed.accelerator.get_accelerator().device_name() == "cuda" \ + and os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": if args.sequence_parallel: raise RuntimeError( "Using sequence parallelism requires setting the environment variable " diff --git a/megatron/core/tensor_parallel/layers.py b/megatron/core/tensor_parallel/layers.py index 2245113c9cf..67a78853aa6 100644 --- a/megatron/core/tensor_parallel/layers.py +++ b/megatron/core/tensor_parallel/layers.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. # Parts of the code here are adapted from PyTorch @@ -450,7 +451,8 @@ def linear_with_grad_accumulation_and_async_allreduce( ] if not linear_with_grad_accumulation_and_async_allreduce.warned: - if os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": + if get_accelerator().device_name() == "cuda" \ + and os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": if sequence_parallel: warnings.warn( "When using sequence parallelism it is recommended to set the " diff --git a/megatron/model/fused_layer_norm.py b/megatron/model/fused_layer_norm.py index 2f3b89014b9..d1ef034397c 100644 --- a/megatron/model/fused_layer_norm.py +++ b/megatron/model/fused_layer_norm.py @@ -1,9 +1,11 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. """This code is copied fron NVIDIA apex: https://github.com/NVIDIA/apex with some changes. """ +from deepspeed.accelerator.real_accelerator import get_accelerator import numbers import torch from torch.nn.parameter import Parameter @@ -13,6 +15,7 @@ import inspect from megatron.core.utils import make_viewless_tensor +from megatron import get_args try: from apex.contrib.layer_norm.layer_norm import FastLayerNormFN @@ -56,8 +59,15 @@ def __init__(self, normalized_shape, eps=1e-5, normalized_shape = (normalized_shape,) self.normalized_shape = torch.Size(normalized_shape) self.eps = eps - self.weight = Parameter(torch.Tensor(*normalized_shape)) - self.bias = Parameter(torch.Tensor(*normalized_shape)) + init_device = None + if get_accelerator().device_name() == 'hpu': + init_device = get_accelerator().current_device_name() + self.weight = Parameter(torch.empty(*normalized_shape, + device=init_device, + dtype=get_args().params_dtype)) + self.bias = Parameter(torch.empty(*normalized_shape, + device=init_device, + dtype=get_args().params_dtype)) self.reset_parameters() self.no_persist_layer_norm = no_persist_layer_norm self.sequence_parallel = sequence_parallel diff --git a/megatron/model/gpt_model.py b/megatron/model/gpt_model.py index 8968c96655d..e5e60c43ee4 100644 --- a/megatron/model/gpt_model.py +++ b/megatron/model/gpt_model.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """GPT-2 model.""" @@ -393,9 +394,12 @@ def _to_float16(inputs): if args.normalization == 'layernorm': self.specs.append(LayerSpec(LayerNorm, args.hidden_size, - eps=args.layernorm_epsilon)) + eps=args.layernorm_epsilon, + sequence_parallel=args.sequence_parallel)) else: - self.specs.append(LayerSpec(RMSNorm, args.hidden_size, args.layernorm_epsilon)) + self.specs.append(LayerSpec(RMSNorm, args.hidden_size, + args.layernorm_epsilon, + sequence_parallel=args.sequence_parallel)) def _logits_helper(embedding, lm_output): """A wrapper to massage inputs/outputs from pipeline. """ diff --git a/megatron/model/language_model.py b/megatron/model/language_model.py index ec2ae1877ad..3b8e4e0da1c 100644 --- a/megatron/model/language_model.py +++ b/megatron/model/language_model.py @@ -1,3 +1,4 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Transformer based language model.""" @@ -256,8 +257,8 @@ def forward(self, input_ids, position_ids, tokentype_ids=None): # Dropout. if self.sequence_parallel: - # already partition sequence, do not need scatter_to_sequence_parallel_region - # embeddings = tensor_parallel.scatter_to_sequence_parallel_region(embeddings) + # already partition sequence, do not need scatter_to_sequence_parallel_region ? + embeddings = tensor_parallel.scatter_to_sequence_parallel_region(embeddings) with tensor_parallel.get_cuda_rng_tracker().fork(): embeddings = self.embedding_dropout(embeddings) else: diff --git a/megatron/model/rmsnorm.py b/megatron/model/rmsnorm.py index 60e89781710..4860d81716b 100644 --- a/megatron/model/rmsnorm.py +++ b/megatron/model/rmsnorm.py @@ -1,4 +1,10 @@ +# Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. + +from deepspeed.accelerator import get_accelerator +from megatron import get_args + import torch +from torch.nn import init from torch.nn.parameter import Parameter # Taken from facebookresearch/llama @@ -6,11 +12,18 @@ class RMSNorm(torch.nn.Module): def __init__(self, dim: int, eps: float = 1e-6): super().__init__() self.eps = eps - self.weight = Parameter(torch.ones(dim)) + init_device = None + if get_accelerator().device_name() == 'hpu': + init_device = get_accelerator().current_device_name() + self.weight = Parameter(torch.empty(dim, + device=init_device, + dtype=get_args().params_dtype)) + init.ones_(self.weight) + setattr(self.weight, 'sequence_parallel', sequence_parallel) def _norm(self, x): return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) def forward(self, x): output = self._norm(x.float()).type_as(x) - return output * self.weight \ No newline at end of file + return output * self.weight diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index 036c11566a3..74e977103f6 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -931,7 +931,8 @@ def __init__(self, config, config.hidden_size, eps=config.layernorm_epsilon) else: - self.input_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon) + self.input_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel) # Self attention. self.self_attention = ParallelAttention( config, @@ -957,7 +958,8 @@ def __init__(self, config, config.hidden_size, eps=config.layernorm_epsilon) else: - self.post_attention_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon) + self.post_attention_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel) # Cross attention. if self.layer_type in (LayerType.decoder, LayerType.retro_decoder, @@ -977,7 +979,9 @@ def __init__(self, config, apply_layernorm_1p=args.apply_layernorm_1p, mem_efficient_ln=args.mem_efficient_ln) else: - self.post_inter_attention_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon) + self.post_inter_attention_layernorm = RMSNorm(config.hidden_size, + config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel) # MLP self.num_experts = num_experts @@ -1780,7 +1784,8 @@ def build_layer(layer_number, n_e): config.hidden_size, eps=config.layernorm_epsilon) else: - self.final_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon) + self.final_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel) def _get_layer(self, layer_number): return self.layers[layer_number] From de7f22f35361e0eff628cb1d61d26ae94234876b Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Wed, 4 Sep 2024 10:00:00 -0500 Subject: [PATCH 311/590] Update utils.py --- megatron/utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/utils.py b/megatron/utils.py index 19eef703998..67dc366d50c 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -80,7 +80,7 @@ def flush(self): def reset(self): return def log_static(self, func): - return + return func class dftracer(object): def __init__(self,): self.type = None From 3edba7f37256f4463df2b04c6e29ba0a12683fbf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Sep 2024 09:45:44 -0500 Subject: [PATCH 312/590] Add `--train-range-to-skip` to `megatron/arguments.py` --- megatron/arguments.py | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/megatron/arguments.py b/megatron/arguments.py index 9a5e4b8da7e..534f6095a5c 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -991,6 +991,13 @@ def _add_training_args(parser): group.add_argument('--use-dataset-only', type=bool, required=False, default=False, help='If set to True, only use the megatron dataset for external trainer ') group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') + group.add_argument( + "--train-range-to-skip", + action="extend", + nargs="+", + type=int, + help=("Range of iters to skip during training. Must be in pairs."), + ) group.add_argument('--train-iters-to-skip', action="extend", nargs="+", type=str, help=( "Specific train iterations to skip when training. " From 76a259b9ce704d90a3a7e84938d51b8eae39a714 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Sep 2024 09:50:55 -0500 Subject: [PATCH 313/590] Add logic for `--trin-range-to-skip` to `megatron/training.py` --- megatron/training.py | 74 +++++++++++++++++++++++++++++--------------- 1 file changed, 49 insertions(+), 25 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 9289c64ec80..c32ba290004 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1639,6 +1639,7 @@ def train( model_module.train() # Tracking loss. total_loss_dict = {} + loss_dict = {} # Iterations. iteration = args.iteration # Translate args to core configuration @@ -1654,6 +1655,19 @@ def train( args.random_ltd_layer_num = model[ 0 ].random_ltd_scheduler.get_random_ltd_layer_num() + ranges_to_skip = None + if args.train_range_to_skip is not None: + assert ( + len(args.train_range_to_skip) % 2 == 0 + ), f"""Expected --train-range-to-skip to have an even number of values. + Received: {len(args.train_range_to_skip)} + """ + ranges_to_skip = list( + zip( + args.train_range_to_skip[::2], + args.train_range_to_skip[1::2], + ) + ) while iteration < args.train_iters and ( args.train_tokens is None or args.consumed_train_tokens < args.train_tokens ): @@ -1675,18 +1689,39 @@ def train( update_rotary_pos_emb(curriculum_seqlen) args.curriculum_seqlen = curriculum_seqlen args.curr_iteration = iteration - if os.getenv("TORCH_PROFILER_ENABLE") == "2": - from torch.profiler import profile, record_function, ProfilerActivity - try: - activities = [ - ProfilerActivity.CPU, - ProfilerActivity.CUDA, - ProfilerActivity.XPU, - ] - except Exception: - log.warning("TORCH PROFILER WARNING: XPU is not supported") - activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA] - with profile(activities=activities) as prof: + if ranges_to_skip is not None and any( + [i <= (iteration + 1) <= j for (i, j) in ranges_to_skip] + ): + log.info(f"Caught {iteration + 1} in 'ranges_to_skip', skipping!") + loss_dict = {} + skipped_iter = 1 + grad_norm = None + num_zeros_in_grad = None + else: + if os.getenv("TORCH_PROFILER_ENABLE") == "2": + from torch.profiler import profile, record_function, ProfilerActivity + try: + activities = [ + ProfilerActivity.CPU, + ProfilerActivity.CUDA, + ProfilerActivity.XPU, + ] + except Exception: + log.warning("TORCH PROFILER WARNING: XPU is not supported") + activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA] + with profile(activities=activities) as prof: + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( + forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config, + ) + prof.export_chrome_trace( + f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" + ) + else: loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( forward_step_func, train_data_iterator, @@ -1695,18 +1730,6 @@ def train( opt_param_scheduler, config, ) - prof.export_chrome_trace( - f"{args.trace_dir}/torch-trace-{RANK}-of-{WORLD_SIZE}-step{iteration}.json" - ) - else: - loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( - forward_step_func, - train_data_iterator, - model, - optimizer, - opt_param_scheduler, - config, - ) iteration += 1 args.iteration = iteration new_samples = ( @@ -1792,7 +1815,8 @@ def train( saved_checkpoint = False if args.exit_signal_handler: signal_handler = get_signal_handler() - if any(signal_handler.signals_received()): + # if any(signal_handler.signals_received()): + if signal_handler is not None and any(signal_handler.signals_received()): save_checkpoint_and_time( iteration, model, optimizer, opt_param_scheduler ) From fd1ac6d73b9c24f8d594c21db0ff2defaa0f4e20 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 10 Sep 2024 08:05:12 -0500 Subject: [PATCH 314/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 224 +++++++++++++++++++++++++++--------------------- 1 file changed, 124 insertions(+), 100 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 7720a919d9b..aa7e682a772 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -20,7 +20,6 @@ # command for launching across all GPUs in our active PBS job. ############################################################################### - ################## # helpers_main # @@ -75,7 +74,7 @@ helpers_main() { # # - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) # - Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -# - This provides `{ezpz_setup_python, ezpz_setup_alcf}` (called below) +# - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) # # - Set runtime options # @@ -113,11 +112,11 @@ setup() { # - NGPU_PER_HOST (by magic) # - NGPUS (= NHOSTS * NGPU_PER_HOST) # 4. Use these (^) to build our launch command - ezpz_setup "$@" || exit + ezpz_setup || exit ########################################################################## install_dependencies # Set command line arguments to pass to `"${EXEC}"` - setParams "$@" || exit + setParams || exit # Create `deepspeed_config.json` from runtime params from ^ buildDSconfig || exit # Specify output directory for {logs, checkpoints, etc.} @@ -138,7 +137,7 @@ setup() { # Check that were not already running, if so, exit. check_and_kill_if_running || exit # Setup run command to be executed - setup_run_cmd || exit + setup_run_cmd "$@" || exit } ##################################################### @@ -147,11 +146,17 @@ setup() { # Build run command to be executed. ##################################################### setup_run_cmd() { + ############################## + # take in additional arguments + # and append them directly to + # the end of the `run_cmd` + custom_args="$@" + ############################## #### Make it easy to track experiments by date ################### year="$(date "+%Y")" month="$(date "+%m")" day="$(date "+%Y-%m-%d")" - today="$(date "+%Y-%m-%d")" # kept for backwards compatibility + today="$(date "+%Y-%m-%d")" # kept for backwards compatibility started_at="$(date "+%Y-%m-%d-%H%M%S")" export YEAR="${year}" export MONTH="${month}" @@ -229,7 +234,8 @@ setup_run_cmd() { ${TIMING_STR} \ ${TOKENIZER_FLAGS} \ ${ds_args} \ - ${gpt_args[*]} + ${gpt_args[*]} \ + ${custom_args} " } @@ -243,7 +249,7 @@ save_dotenv() { module list dotenv_file="${outdir}/.env" echo "Saving environment to ${dotenv_file}" - printenv | grep -v "LS_COLORS" > "${dotenv_file}" + printenv | grep -v "LS_COLORS" >"${dotenv_file}" export DOTENV_FILE="${dotenv_file}" fi } @@ -298,29 +304,26 @@ get_machine() { printf "Running on: %s\n" "$(printBlue "${MACHINE}")" } - check_and_kill_if_running() { RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') - if [[ -n "${RUNNING_PIDS}" ]]; - then echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}"; + if [[ -n "${RUNNING_PIDS}" ]]; then + echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}" else echo "Not currently running. Continuing!" fi } - setupSrun() { if [[ $(hostname) == login* || $(hostname) == nid* ]]; then export NHOSTS="${SLURM_NNODES:-1}" export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" - export NGPUS="$(( NHOSTS * NGPU_PER_HOST ))" + export NGPUS="$((NHOSTS * NGPU_PER_HOST))" export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" else echo "Skipping setupSrun() on $(hostname)" fi } - printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" @@ -377,7 +380,6 @@ setupLauncher() { printf " %s" "$(printMagenta "${LAUNCHER}")" } - set_lr_args() { LR_ARGS="--lr ${LR} --lr-decay-style cosine" if [[ -n "${LR_DECAY_ITERS:-}" ]]; then @@ -390,7 +392,6 @@ set_lr_args() { export LR_ARGS="${LR_ARGS}" } - ######################################################################### # `get_batch_size_on_polaris`: Identify MICRO_BATCH to use on Polaris. # @@ -407,8 +408,8 @@ set_lr_args() { ######################################################################### get_batch_size_on_polaris() { if [[ $(hostname) == x3* ]]; then - nhosts=$(wc -l < "${HOSTFILE:-${PBS_NODEFILE}}") - if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then + nhosts=$(wc -l <"${HOSTFILE:-${PBS_NODEFILE}}") + if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then mbs=1 elif [[ "${nhosts}" -ge 3 && "${nhosts}" -le 7 ]]; then mbs=2 @@ -420,9 +421,9 @@ get_batch_size_on_polaris() { } _get_num_hosts_from_hostfile() { - if [[ "$#" == 1 ]]; then + if [[ "$#" == 1 ]]; then if [[ -f "$1" ]]; then - nhosts=$(wc -l < "$1") + nhosts=$(wc -l <"$1") echo "${nhosts}" else exit 1 @@ -460,7 +461,7 @@ get_grad_acc_steps_on_aurora() { echo "Expected exactly 0 or 1 arguments, received: $#" exit 1 fi - nhosts=$(wc -l < "${hf}") + nhosts=$(wc -l <"${hf}") if [[ 64 -le "${nhosts}" ]]; then gas=1 elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 64 ]]; then @@ -473,6 +474,33 @@ get_grad_acc_steps_on_aurora() { echo "${gas}" } +set_ccl_vars_on_aurora() { + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export CCL_CONFIGURATION=cpu_gpu_dpcpp + # export CCL_ROOT=/tmp/oneccl/ + # export LD_LIBRARY_PATH=${CCL_ROOT}/lib:$LD_LIBRARY_PATH + # export CPATH=${CCL_ROOT}/include:$CPATH + # export LIBRARY_PATH=${CCL_ROOT}/lib:$LIBRARY_PATH + export CCL_KVS_CONNECTION_TIMEOUT=3600 + export FI_CXI_RX_MATCH_MODE=hybrid + export CCL_BCAST=double_tree + + export ZE_ENABLE_PCI_ID_DEVICE_ORDER=1 + export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich + export FI_PROVIDER=cxi # Required by Aurora mpich + export PALS_PMI=pmix # Required by Aurora mpich + export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + export TORCH_LLM_ALLREDUCE=1 + export CCL_SYCL_ESIMD=1 + export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) + export CCL_ENABLE_SYCL_KERNELS=1 + export CCL_WORKER_AFFINITY=5,13,21,29,37,45,57,65,73,81,89,97 + export CCL_ZE_CACHE_OPEN_IPC_HANDLES_THRESHOLD=32768 + export FI_CXI_DEFAULT_CQ_SIZE=1048576 + export FI_CXI_RX_MATCH_MODE=hybrid + export CCL_BCAST=double_tree +} ############################################################################## # setParams @@ -493,20 +521,24 @@ setParams() { # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then - TP=${TP:-1} # TP = 1 + TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" - export CCL=${CCL:-ccl} # CCL - export BE="${CCL}" # COMMUNICATION BACKEND = CCL - export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 + export CCL=${CCL:-ccl} # CCL + export BE="${CCL}" # COMMUNICATION BACKEND = CCL + export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS - gas=$(get_grad_acc_steps_on_aurora "$@") + gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 - export CCL_PROCESS_LAUNCHER=pmix - export CCL_ATL_TRANSPORT=mpi - ###################################################################### + MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + #### [sam: 08/17/2024] ########################################## + # Use best set of CCL env vars from Gordon Bell runs on Aurora + set_ccl_vars_on_aurora + ################################################################# + #### [sam: 06/20/2024] ############################################### + # export CCL_PROCESS_LAUNCHER=pmix + # export CCL_ATL_TRANSPORT=mpi # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? # update_ccl_env_vars_aurora @@ -528,12 +560,12 @@ setParams() { # elif [[ $(hostname) == x3* ]]; then elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? - export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS # NOTE: MICRO_BATCH is exported below # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" @@ -546,7 +578,7 @@ setParams() { source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # +--------[Perlmutter]---------------------------------+ # elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then -elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then + elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then TP="${TP:-2}" export NCCL="${NCCL:-nccl}" export BE="${NCCL}" @@ -567,45 +599,45 @@ elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then export OPT="${OPT:-adamw}" export ADAM_BETA1="${ADAM_BETA1:-0.9}" export ADAM_BETA2="${ADAM_BETA2:-0.95}" - export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" - NHOSTS=$(wc -l < "${HOSTFILE}") + NHOSTS=$(wc -l <"${HOSTFILE}") if [[ -z "${NGPU_PER_HOST:-}" ]]; then NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') fi export NGPU_PER_HOST="${NGPU_PER_HOST}" - export WORLD_SIZE="${WORLD_SIZE:-$(( NHOSTS * NGPU_PER_HOST ))}" + export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" # +---[Llama2 7B Config]--------------------------------------------------+ # export MODEL_KEY="Llama-7B" - export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS - export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS - export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE - export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION - export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE + export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS - export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN - export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL - export SAVE_INTERVAL=${SAVE_INTERVAL:-50} # HOW FREQUENTLY TO SAVE CKPTS - export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS - export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? - export GLOBAL_BATCH_MAX=$(( WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP )) # MAX GLOBAL BATCH SIZE - export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE + export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS + export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN + export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL + export SAVE_INTERVAL=${SAVE_INTERVAL:-50} # HOW FREQUENTLY TO SAVE CKPTS + export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS + export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? + export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE + export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT # export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS if [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} - export TRAIN_ITERS=$(( TRAIN_TOKENS / SEQ / GLOBAL_BATCH )) - printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$(( TRAIN_TOKENS / 10**9 ))" + export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" else export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" fi - export MODEL_TYPE="llama-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL + export MODEL_TYPE="llama-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # NOTE: [2024-07-10] ##################################################### # - [sam]: For whatever reason, it seems that using # sequence-parallelism (SP) > 1 is INCOMPATIBLE with @@ -620,8 +652,8 @@ elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" fi # -----[Learning Rate Settings]-------------------------------------------- - export LR=${LR:-0.0003} # LEARNING_RATE - export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + export LR=${LR:-0.0003} # LEARNING_RATE + export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args # -----[Learning Rate Settings]-------------------------------------------- @@ -636,7 +668,6 @@ elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then fi } - ############################################## # set_args # @@ -648,7 +679,7 @@ set_args() { ds_args=" " ds_args=" --deepspeed ${ds_args}" if [[ $PP == 1 ]]; then - ds_args=" --no-pipeline-parallel ${ds_args}" + ds_args=" --no-pipeline-parallel ${ds_args}" fi ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" @@ -676,20 +707,18 @@ set_args() { export gpt_args } - make_ds_hostfile() { export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" # ---- Make MPICH hostfile ---------------- hf="${HOSTFILE:-${PBS_NODEFILE}}" export hostfile_mpich=hostfile_mpich - cat "${hf}" > "${hostfile_mpich}" + cat "${hf}" >"${hostfile_mpich}" # ---- Make DeepSpeed hostfile ------------------- export hostfile_deepspeed=hostfile_deepspeed - cat "${hf}" > "${hostfile_deepspeed}" + cat "${hf}" >"${hostfile_deepspeed}" sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } - ########################################### # ezpz_setup # @@ -697,12 +726,12 @@ make_ds_hostfile() { # to `"${WORKING_DIR}/deps/ezpz/"` # # 2. Source [`ezpz/src/ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -# - This provides `{ezpz_setup_python, ezpz_setup_alcf}` (called below) +# - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) # # 3. Call `ezpz_setup_python` (from `ezpz/bin/utils.sh`): # - This will setup conda + virtual enviroment # -# 4. Call `ezpz_setup_alcf` (from `ezpz/bin/utils.sh`): +# 4. Call `ezpz_setup_job` (from `ezpz/bin/utils.sh`): # - This will parse `$PBS_*` variables and build launch cmd # # 3. Call `_ezpz_install` (from `Megatron-DeepSpeed/ALCF/helpers.sh`): @@ -723,7 +752,7 @@ ezpz_setup() { # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh source "${ezdir}/src/ezpz/bin/utils.sh" || exit ezpz_setup_python - ezpz_setup_alcf "$@" + ezpz_setup_job "$@" ezpz_pip_loc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') if [[ -z "${ezpz_pip_loc:-}" ]]; then printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" @@ -750,21 +779,20 @@ ezpz_test() { # saveDSenv # # Save important environment variables to .deepspeed_env, which will be -# forwarded to ALL ranks with DeepSpeed +# forwarded to ALL ranks with DeepSpeed ############################################################################ saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { - echo "PATH=${PATH}" ; - echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}" ; - echo "http_proxy=${http_proxy:-}" ; - echo "https_proxy=${https_proxy:-}" ; - echo "CFLAGS=${CFLAGS}" ; - echo "PYTHONUSERBASE=$PYTHONUSERBASE" ; - } > .deepspeed_env + echo "PATH=${PATH}" + echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}" + echo "http_proxy=${http_proxy:-}" + echo "https_proxy=${https_proxy:-}" + echo "CFLAGS=${CFLAGS}" + echo "PYTHONUSERBASE=$PYTHONUSERBASE" + } >.deepspeed_env } - get_output_prefix() { # ---- Specify output location -------------------------------- pre="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" @@ -792,7 +820,7 @@ setOutput() { export OUTPUT_DIR="${OUTPUT_DIR}" && mkdir -p "${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" - echo "${OUTPUT_LOG}" >> "logs/latest" + echo "${OUTPUT_LOG}" >>"logs/latest" printf "\n Please see logs at: %s\n" "$(printGreen "${OUTPUT_DIR}")" printf "Checkpoints will be saved to: %s\n" "$(printYellow "${CKPT_DIR}")" } @@ -809,7 +837,6 @@ buildDSconfig() { generateDSconfig "${DS_CONFIG}" } - ############################################################################### # sumWeights # @@ -844,7 +871,6 @@ make_data() { cd "${mdir}" && make && cd - } - ############################################################################## # install_dependencies # @@ -853,7 +879,7 @@ make_data() { install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." - python3 -m pip install -r "${depsfile}" --require-virtualenv 1> /dev/null + python3 -m pip install -r "${depsfile}" --require-virtualenv 1>/dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then mn=$(get_machine_name) # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then @@ -882,14 +908,13 @@ install_deepspeed_for_xpu() { git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" cd "${outdir}" || exit echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" - python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1> /dev/null - python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1> /dev/null - python setup.py develop 1> /dev/null + python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1>/dev/null + python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1>/dev/null + python setup.py develop 1>/dev/null cd "${WORKING_DIR}" echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" } - ################################################# # Fix for distributed key value store on Aurora ################################################# @@ -946,7 +971,6 @@ check_executable() { fi } - ###################################################################### # `makeHostiles`: # Detect if `HOSTFILE` set in active environment. @@ -955,7 +979,7 @@ check_executable() { ###################################################################### makeHostfiles() { if [[ -n "${HOSTFILE}" ]]; then - printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" + printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" else make_ds_hostfile fi @@ -996,8 +1020,8 @@ setup_tokenizer_and_data() { else export DATA_FLAGS="" export TOKENIZER_TYPE="Llama2" - tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ export TOKENIZER_FLAGS="--tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" @@ -1014,7 +1038,7 @@ setup_tokenizer_and_data() { # Ensure `DATA_FILE_LIST` is set, # fallback to default values if necessary. ############################################### -setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] +setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# mn=$(get_machine_name) dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" @@ -1023,7 +1047,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # use this data file list to call `setData` dfl="${1:-${dfl_fallback}}" printf "Calling: setData() with %s\n" "${dfl}" - ndocs=$(wc -l < "${dfl}") + ndocs=$(wc -l <"${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") dcp=".cache/${dfl_stem}/index-cache" @@ -1032,7 +1056,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export WEIGHT_SUM="${ws}" export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" - export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" + export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" echo "--------------------" echo "Updated environment:" printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" @@ -1053,8 +1077,8 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ################################################################################ generateDSconfig() { if [ $# -ne 1 ]; then - echo "Usage: $0 config_file" - exit 1 + echo "Usage: $0 config_file" + exit 1 fi for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE"; do if [ -z "$v" ]; then @@ -1134,7 +1158,7 @@ generateDSconfig() { dtype="\"communication_data_type\": \"fp32\"," fi if [[ "${ZERO_STAGE}" == 3 ]]; then - # \"mics_shard_size\": 2, + # \"mics_shard_size\": 2, zero="\ \"zero_optimization\": { \"stage\": 3, @@ -1157,7 +1181,7 @@ generateDSconfig() { } }," # elif [[ $ZERO_STAGE == 2 ]]; then - elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then + elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then # if [[ -n "${CPU_OPTIMIZER}" ]]; then if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" @@ -1197,7 +1221,7 @@ generateDSconfig() { else echo 'Please add the correct config set!!!' fi -cat < "$1" + cat <"$1" { $common $zero @@ -1243,7 +1267,7 @@ GREEN="\e[1;32m" YELLOW="\e[1;33m" BLUE="\e[1;34m" CYAN="\e[1;35m" -WHITE="\e[1;36m" +# WHITE="\e[1;36m" printBlack() { printf "\e[1;30m%s\e[0m\n" "$@" From 6f27f5dce68e0903e36d9faa47906ba89c6074ea Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 10 Sep 2024 08:06:09 -0500 Subject: [PATCH 315/590] Update `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 8bdf776ed9c..a6a2db72abb 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -7,18 +7,28 @@ # AuroraGPT-7B @ ALCF ##################################### - # 1. Navigate into `$PBS_O_WORKDIR` cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE + # 2. source `ALCF/helpers.sh` source "${HERE}/ALCF/helpers.sh" || exit + # 3. call `setup` from `./ALCF/helpers.sh` setup "$@" || exit export run_cmd="${run_cmd}" echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" -# 7. Tell user where to find output + +# 4. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + +# 5. Ignore the following strings on Intel XPU devices +# (otherwise they'll clutter up logs) XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" -# 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} + +# if [[ $(ezpz_get_machine_name) == "aurora" ]]; then +# module unload mpich && module load mpich +# fi +# +# 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} eval "${run_cmd}" |& grep -E -v "${XPU_IGNORE_STRING}" |& tee -a "${OUTPUT_LOG}" From 6df33ad7686449be378a3fdede119e019b7ded01 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 11 Sep 2024 14:54:49 -0500 Subject: [PATCH 316/590] fix: `--override-opt_param-scheduler` if `OVERRIDE_CKPT_OPT_PARAM=1` e.g.: ```bash $ PBS_O_WORKDIR=$(pwd) LR=0.00020 OVERRIDE_CKPT_OPT_PARAM=1 bash train_aGPT_7B.sh --train-range-to-skip 43000 47000 --override-opt_param-scheduler ``` will override the `lr_scheduler` params from the checkpoint and instead use the specified value, `LR=0.00020` instead. --- ALCF/helpers.sh | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index aa7e682a772..bc31c322ce7 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -181,9 +181,12 @@ setup_run_cmd() { --no-masked-softmax-fusion \ --no-gradient-accumulation-fusion \ --accumulate-allreduce-grads-in-fp32 \ - --use-checkpoint-opt_param-scheduler \ --log-timers-to-tensorboard \ --log-optimizer-states-to-tensorboard" + OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" + if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then + DEFAULTS="${DEFAULTS} --use-checkpoint-opt_param-scheduler" + fi if [[ "${SP}" -ge 2 ]]; then export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" fi From a1ede68a5ee9a6ce11ad1adf90b57acef4859f69 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 11 Sep 2024 19:45:26 -0500 Subject: [PATCH 317/590] Remove duplicate `--profile` arg --- megatron/arguments.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 154f53e2bc9..b3ed06353e2 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1012,7 +1012,7 @@ def _add_training_args(parser): dest='gradient_accumulation_fusion') group.add_argument('--use-dataset-only', type=bool, required=False, default=False, help='If set to True, only use the megatron dataset for external trainer ') - group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') + # group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') group.add_argument( "--train-range-to-skip", action="extend", From 6b32cff40aecd5cc13ad0d0e7414ea91560034d9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 11 Sep 2024 19:46:11 -0500 Subject: [PATCH 318/590] debug: `sequence_parallel` issue in `RMSNorm` ?? --- megatron/model/fused_rmsnorm.py | 4 +- megatron/model/rmsnorm.py | 15 +- megatron/model/transformer.py | 1513 +++++++++++++++++++------------ 3 files changed, 958 insertions(+), 574 deletions(-) diff --git a/megatron/model/fused_rmsnorm.py b/megatron/model/fused_rmsnorm.py index a3b9927c874..0ee372a7eb2 100644 --- a/megatron/model/fused_rmsnorm.py +++ b/megatron/model/fused_rmsnorm.py @@ -2,9 +2,11 @@ from torch.nn.parameter import Parameter import intel_extension_for_pytorch as ipex # noqa + # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__(self, dim: int, eps: float = 1e-6): + # def __init__(self, dim: int, eps: float = 1e-6): + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False): super().__init__() self.eps = eps self.weight = Parameter(torch.ones(dim)) diff --git a/megatron/model/rmsnorm.py b/megatron/model/rmsnorm.py index 4860d81716b..6feb575f680 100644 --- a/megatron/model/rmsnorm.py +++ b/megatron/model/rmsnorm.py @@ -7,19 +7,20 @@ from torch.nn import init from torch.nn.parameter import Parameter + # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__(self, dim: int, eps: float = 1e-6): + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False): super().__init__() self.eps = eps init_device = None - if get_accelerator().device_name() == 'hpu': - init_device = get_accelerator().current_device_name() - self.weight = Parameter(torch.empty(dim, - device=init_device, - dtype=get_args().params_dtype)) + if get_accelerator().device_name() == "hpu": + init_device = get_accelerator().current_device_name() + self.weight = Parameter( + torch.empty(dim, device=init_device, dtype=get_args().params_dtype) + ) init.ones_(self.weight) - setattr(self.weight, 'sequence_parallel', sequence_parallel) + setattr(self.weight, "sequence_parallel", sequence_parallel) def _norm(self, x): return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) diff --git a/megatron/model/transformer.py b/megatron/model/transformer.py index 74e977103f6..592ff2855b3 100644 --- a/megatron/model/transformer.py +++ b/megatron/model/transformer.py @@ -25,6 +25,7 @@ try: from deepspeed.sequence.layer import DistributedAttention + dist_attn_supported = True except ImportError: dist_attn_supported = False @@ -70,28 +71,31 @@ hyperparameters: transformer hyperparameters """ + class DropPath(MegatronModule): """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). """ - def __init__(self, drop_prob=0.): + def __init__(self, drop_prob=0.0): super(DropPath, self).__init__() self.drop_prob = drop_prob def forward(self, hidden_state): - if self.drop_prob == 0. or not self.training: + if self.drop_prob == 0.0 or not self.training: return hidden_state keep_prob = 1 - self.drop_prob # work with diff dim tensors, not just 2D ConvNets # hidden_state: [s, b, h] shape = (1,) + (hidden_state.shape[1],) + (1,) * (hidden_state.ndim - 2) - random_tensor = keep_prob + \ - torch.rand(shape, dtype=hidden_state.dtype, device=hidden_state.device) + random_tensor = keep_prob + torch.rand( + shape, dtype=hidden_state.dtype, device=hidden_state.device + ) random_tensor.floor_() # binarize output = hidden_state.div(keep_prob) * random_tensor return output + class ParallelMLP(MegatronModule): """MLP. @@ -120,7 +124,7 @@ def __init__(self, config, moe=False, enable_expert_tensor_parallelism=False): gather_output=False, skip_bias_add=True, moe=moe, - enable_expert_tensor_parallelism=enable_expert_tensor_parallelism + enable_expert_tensor_parallelism=enable_expert_tensor_parallelism, ) self.bias_gelu_fusion = False @@ -132,13 +136,17 @@ def __init__(self, config, moe=False, enable_expert_tensor_parallelism=False): elif args.onnx_safe: self.activation_func = erf_gelu elif args.swiglu: + def swiglu(x): x = torch.chunk(x, 2, dim=-1) return F.silu(x[0]) * x[1] + self.activation_func = swiglu elif args.squared_relu: + def squared_relu(x): return torch.pow(F.relu(x), 2) + self.activation_func = squared_relu else: self.bias_gelu_fusion = args.bias_gelu_fusion @@ -153,7 +161,7 @@ def squared_relu(x): bias=self.add_bias, input_is_parallel=True, moe=moe, - enable_expert_tensor_parallelism=enable_expert_tensor_parallelism + enable_expert_tensor_parallelism=enable_expert_tensor_parallelism, ) def forward(self, hidden_states): @@ -175,10 +183,12 @@ def forward(self, hidden_states): output, output_bias = self.dense_4h_to_h(intermediate_parallel) return output, output_bias + class SwitchMLP(MegatronModule): """ Routes input to one of N MLP "experts" """ + def __init__(self, config): super(SwitchMLP, self).__init__() args = get_args() @@ -195,29 +205,29 @@ def forward(self, hidden_states): route = self.router(hidden_states) route = torch.nn.functional.softmax(route, dim=2) max_prob, max_ind = torch.max(route, dim=2) - max_prob = torch.unsqueeze(max_prob, 2) # [s b 1] + max_prob = torch.unsqueeze(max_prob, 2) # [s b 1] # TODO (rprenger) TODO this could be made easier to read # Converting [s, b, h] to [s*b, h]. # Each vector could be routed differently - hidden_states = hidden_states.view(-1, hidden_states.size(2)) # [s*b h] - max_prob = max_prob.view(-1, max_prob.size(2)) # [s*b 1] - max_ind = max_ind.view(-1) # [s*b] + hidden_states = hidden_states.view(-1, hidden_states.size(2)) # [s*b h] + max_prob = max_prob.view(-1, max_prob.size(2)) # [s*b 1] + max_ind = max_ind.view(-1) # [s*b] output_total = torch.empty_like(hidden_states) output_bias_total = torch.empty_like(hidden_states) - #TODO (rprenger) This does each expert in serial, but it could be parallelized + # TODO (rprenger) This does each expert in serial, but it could be parallelized for expert_num, expert in enumerate(self.experts): local_indices = (max_ind == expert_num).nonzero() - hidden = hidden_states[local_indices,:] + hidden = hidden_states[local_indices, :] output, output_bias = expert(hidden) output_bias = output_bias.expand_as(output) - output_total[local_indices,:] = output - output_bias_total[local_indices,:] = output_bias + output_total[local_indices, :] = output + output_bias_total[local_indices, :] = output_bias - output_total = output_total*max_prob - output_bias_total = output_bias_total*max_prob + output_total = output_total * max_prob + output_bias_total = output_bias_total * max_prob output_total = output_total.view(s, b, h) output_bias_total = output_bias_total.view(s, b, h) @@ -226,8 +236,7 @@ def forward(self, hidden_states): class CoreAttention(MegatronModule): - def __init__(self, layer_number, config, - attn_mask_type=AttnMaskType.padding): + def __init__(self, layer_number, config, attn_mask_type=AttnMaskType.padding): super(CoreAttention, self).__init__() self.fp16 = config.fp16 self.bf16 = config.bf16 @@ -246,14 +255,19 @@ def __init__(self, layer_number, config, seq_parallel_world_size = 1 if parallel_state.sequence_parallel_is_initialized(): seq_parallel_world_size = parallel_state.get_sequence_parallel_world_size() - world_size = seq_parallel_world_size if seq_parallel_world_size > 1 else parallel_state.get_tensor_model_parallel_world_size() + world_size = ( + seq_parallel_world_size + if seq_parallel_world_size > 1 + else parallel_state.get_tensor_model_parallel_world_size() + ) - self.hidden_size_per_partition = core.utils.divide(projection_size, - world_size) + self.hidden_size_per_partition = core.utils.divide(projection_size, world_size) self.hidden_size_per_attention_head = core.utils.divide( - projection_size, config.num_attention_heads) + projection_size, config.num_attention_heads + ) self.num_attention_heads_per_partition = core.utils.divide( - config.num_attention_heads, world_size) + config.num_attention_heads, world_size + ) coeff = None self.norm_factor = math.sqrt(self.hidden_size_per_attention_head) @@ -262,49 +276,56 @@ def __init__(self, layer_number, config, self.norm_factor *= coeff self.scale_mask_softmax = FusedScaleMaskSoftmax( - self.fp16, self.bf16, + self.fp16, + self.bf16, self.attn_mask_type, config.masked_softmax_fusion, attention_mask_func, self.attention_softmax_in_fp32, - coeff) + coeff, + ) # Dropout. Note that for a single iteration, this layer will generate # different outputs on different number of parallel partitions but # on average it should not be partition dependent. self.attention_dropout = torch.nn.Dropout(config.attention_dropout) - def forward(self, query_layer, key_layer, - value_layer, attention_mask): + def forward(self, query_layer, key_layer, value_layer, attention_mask): # =================================== # Raw attention scores. [b, np, s, s] # =================================== # [b, np, sq, sk] - output_size = (query_layer.size(1), - query_layer.size(2), - query_layer.size(0), - key_layer.size(0)) + output_size = ( + query_layer.size(1), + query_layer.size(2), + query_layer.size(0), + key_layer.size(0), + ) # [sq, b, np, hn] -> [sq, b * np, hn] - query_layer = query_layer.view(output_size[2], - output_size[0] * output_size[1], -1) + query_layer = query_layer.view( + output_size[2], output_size[0] * output_size[1], -1 + ) # [sk, b, np, hn] -> [sk, b * np, hn] - key_layer = key_layer.view(output_size[3], - output_size[0] * output_size[1], -1) + key_layer = key_layer.view(output_size[3], output_size[0] * output_size[1], -1) # preallocting input tensor: [b * np, sq, sk] matmul_input_buffer = parallel_state.get_global_memory_buffer().get_tensor( - (output_size[0]*output_size[1], output_size[2], output_size[3]), - query_layer.dtype, "mpu") + (output_size[0] * output_size[1], output_size[2], output_size[3]), + query_layer.dtype, + "mpu", + ) # Raw attention scores. [b * np, sq, sk] matmul_result = torch.baddbmm( matmul_input_buffer, - query_layer.transpose(0, 1), # [b * np, sq, hn] + query_layer.transpose(0, 1), # [b * np, sq, hn] key_layer.transpose(0, 1).transpose(1, 2), # [b * np, hn, sk] - beta=0.0, alpha=(1.0/self.norm_factor)) + beta=0.0, + alpha=(1.0 / self.norm_factor), + ) # change view to [b, np, sq, sk] attention_scores = matmul_result.view(*output_size) @@ -314,8 +335,7 @@ def forward(self, query_layer, key_layer, # =========================== # attention scores and attention mask [b, np, sq, sk] - attention_probs = self.scale_mask_softmax(attention_scores, - attention_mask) + attention_probs = self.scale_mask_softmax(attention_scores, attention_mask) # This is actually dropping out entire tokens to attend to, which might # seem a bit unusual, but is taken from the original Transformer paper. @@ -333,18 +353,22 @@ def forward(self, query_layer, key_layer, # [sk, b, np, hn] --> [b, np, sq, hn] # context layer shape: [b, np, sq, hn] - output_size = (value_layer.size(1), - value_layer.size(2), - query_layer.size(0), - value_layer.size(3)) + output_size = ( + value_layer.size(1), + value_layer.size(2), + query_layer.size(0), + value_layer.size(3), + ) # change view [sk, b * np, hn] - value_layer = value_layer.view(value_layer.size(0), - output_size[0] * output_size[1], -1) + value_layer = value_layer.view( + value_layer.size(0), output_size[0] * output_size[1], -1 + ) # change view [b * np, sq, sk] - attention_probs = attention_probs.view(output_size[0] * output_size[1], - output_size[2], -1) + attention_probs = attention_probs.view( + output_size[0] * output_size[1], output_size[2], -1 + ) # matmul: [b * np, sq, hn] context_layer = torch.bmm(attention_probs, value_layer.transpose(0, 1)) @@ -356,8 +380,9 @@ def forward(self, query_layer, key_layer, context_layer = context_layer.permute(2, 0, 1, 3).contiguous() # [sq, b, np, hn] --> [sq, b, hp] - new_context_layer_shape = context_layer.size()[:-2] + \ - (self.hidden_size_per_partition,) + new_context_layer_shape = context_layer.size()[:-2] + ( + self.hidden_size_per_partition, + ) context_layer = context_layer.view(*new_context_layer_shape) return context_layer @@ -373,12 +398,24 @@ class FlashSelfAttention(torch.nn.Module): attention_dropout: The dropout rate to apply to the attention (default: 0.0) """ - def __init__(self, causal=False, softmax_scale=None, attention_dropout=0.0, - device=None, dtype=None): + + def __init__( + self, + causal=False, + softmax_scale=None, + attention_dropout=0.0, + device=None, + dtype=None, + ): super().__init__() - assert flash_attn_unpadded_func is not None or flash_attn_varlen_func is not None or flash_attn_builder is not None, \ - ('Please install FlashAttention first, e.g., with pip install flash-attn or implement your own flash attention') - assert rearrange is not None, 'Please install einops first, e.g., with pip install einops' + assert ( + flash_attn_unpadded_func is not None + or flash_attn_varlen_func is not None + or flash_attn_builder is not None + ), "Please install FlashAttention first, e.g., with pip install flash-attn or implement your own flash attention" + assert ( + rearrange is not None + ), "Please install einops first, e.g., with pip install einops" self.causal = causal self.softmax_scale = softmax_scale self.dropout_p = attention_dropout @@ -389,14 +426,18 @@ def __init__(self, causal=False, softmax_scale=None, attention_dropout=0.0, self.use_flash_attn_builder_v2 = False self.use_flash_attn = False if args.use_flash_attn_builder: - if hasattr(flash_attn_builder, 'flash_attn_func'): + if hasattr(flash_attn_builder, "flash_attn_func"): self.flash_attn_func = flash_attn_builder.flash_attn_func self.use_flash_attn_builder_v1 = True else: self.flash_attn_func = flash_attn_builder.flash_attn_func_v2 self.use_flash_attn_builder_v2 = True else: - self.flash_attn_func = flash_attn_varlen_func if args.use_flash_attn_v2 else flash_attn_unpadded_func + self.flash_attn_func = ( + flash_attn_varlen_func + if args.use_flash_attn_v2 + else flash_attn_unpadded_func + ) self.use_flash_attn = True def forward(self, q, k, v): @@ -406,42 +447,67 @@ def forward(self, q, k, v): q, k, v: The tensor containing the query, key, and value. (B, S, H, D) """ - assert all((i.dtype in [torch.float16, torch.bfloat16] for i in (q,k,v))) + assert all((i.dtype in [torch.float16, torch.bfloat16] for i in (q, k, v))) assert all((get_accelerator().on_accelerator(i) for i in (q, k, v))) batch_size, seqlen_q = q.shape[0], q.shape[1] seqlen_k = k.shape[1] if self.use_flash_attn: - q, k, v = [rearrange(x, 'b s ... -> (b s) ...') for x in [q, k, v]] - cu_seqlens_q = torch.arange(0, (batch_size + 1) * seqlen_q, step=seqlen_q, dtype=torch.int32, - device=q.device) + q, k, v = [rearrange(x, "b s ... -> (b s) ...") for x in [q, k, v]] + cu_seqlens_q = torch.arange( + 0, + (batch_size + 1) * seqlen_q, + step=seqlen_q, + dtype=torch.int32, + device=q.device, + ) elif self.use_flash_attn_builder_v1: - q, k, v = [rearrange(x, 'b s h d -> b h s d').contiguous() for x in [q, k, v]] + q, k, v = [ + rearrange(x, "b s h d -> b h s d").contiguous() for x in [q, k, v] + ] else: # use_flash_attn_builder_v2 - q, k, v = [rearrange(x, 'b s h d -> b h s d') for x in [q, k, v]] + q, k, v = [rearrange(x, "b s h d -> b h s d") for x in [q, k, v]] if self.training: # during training q,k,v always have same seqlen assert seqlen_k == seqlen_q is_causal = self.causal - cu_seqlens_k = cu_seqlens_q if get_accelerator().device_name() == 'cuda' else None + cu_seqlens_k = ( + cu_seqlens_q if get_accelerator().device_name() == "cuda" else None + ) dropout_p = self.dropout_p else: # turn off FA causal mask after first inference autoregressive iteration # only on first autoregressive step q,k,v have same seqlen is_causal = seqlen_q == seqlen_k - cu_seqlens_k = torch.arange(0, (batch_size + 1) * seqlen_k, step=seqlen_k, dtype=torch.int32, - device=q.device) if get_accelerator().device_name() == 'cuda' else None + cu_seqlens_k = ( + torch.arange( + 0, + (batch_size + 1) * seqlen_k, + step=seqlen_k, + dtype=torch.int32, + device=q.device, + ) + if get_accelerator().device_name() == "cuda" + else None + ) dropout_p = 0 if self.use_flash_attn: output = self.flash_attn_func( - q, k, v, cu_seqlens_q, cu_seqlens_k, seqlen_q, seqlen_k, + q, + k, + v, + cu_seqlens_q, + cu_seqlens_k, + seqlen_q, + seqlen_k, dropout_p, - softmax_scale=self.softmax_scale, causal=is_causal + softmax_scale=self.softmax_scale, + causal=is_causal, ) else: # use_flash_attn_builder @@ -450,15 +516,16 @@ def forward(self, q, k, v): ) if self.use_flash_attn: - output = rearrange(output, '(b s) ... -> b s ...', b=batch_size) + output = rearrange(output, "(b s) ... -> b s ...", b=batch_size) elif self.use_flash_attn_builder_v1: - output = rearrange(output, 'b h s d -> b s h d').contiguous() + output = rearrange(output, "b h s d -> b s h d").contiguous() else: # use_flash_attn_builder_v2: - output = rearrange(output, 'b h s d -> b s h d') + output = rearrange(output, "b h s d -> b s h d") return output + class FlashSelfAttentionTriton(torch.nn.Module): """Implement the scaled dot product attention with softmax. Arguments @@ -469,11 +536,22 @@ class FlashSelfAttentionTriton(torch.nn.Module): attention_dropout: The dropout rate to apply to the attention (default: 0.0) """ - def __init__(self, causal=False, softmax_scale=None, attention_dropout=0.0, - device=None, dtype=None): + + def __init__( + self, + causal=False, + softmax_scale=None, + attention_dropout=0.0, + device=None, + dtype=None, + ): super().__init__() - assert flash_attn_func is not None, ('Triton version of FlashAttention is not installed.') - assert rearrange is not None, 'Please install einops first, e.g., with pip install einops' + assert ( + flash_attn_func is not None + ), "Triton version of FlashAttention is not installed." + assert ( + rearrange is not None + ), "Please install einops first, e.g., with pip install einops" self.causal = causal self.softmax_scale = softmax_scale self.dropout_p = attention_dropout @@ -487,13 +565,13 @@ def forward(self, q, k, v): assert q.dtype in [torch.float16, torch.bfloat16] assert q.is_cuda - q, k, v = [rearrange(x, 's b ... -> b s ...').contiguous() - for x in (q, k, v)] - + q, k, v = [rearrange(x, "s b ... -> b s ...").contiguous() for x in (q, k, v)] + output = flash_attn_func(q, k, v, None, self.causal) - output = rearrange(output, 'b s h d -> s b (h d)').contiguous() + output = rearrange(output, "b s h d -> s b (h d)").contiguous() return output + class ParallelAttention(MegatronModule): """Parallel self-attention layer abstract class. @@ -501,9 +579,13 @@ class ParallelAttention(MegatronModule): and returns output of the same size. """ - def __init__(self, config, layer_number, - attention_type=AttnType.self_attn, - attn_mask_type=AttnMaskType.padding): + def __init__( + self, + config, + layer_number, + attention_type=AttnType.self_attn, + attn_mask_type=AttnMaskType.padding, + ): super(ParallelAttention, self).__init__() args = get_args() self.layer_number = max(1, layer_number) @@ -513,12 +595,18 @@ def __init__(self, config, layer_number, self.sequence_parallel = config.sequence_parallel self.num_attention_heads = config.num_attention_heads self.num_key_value_heads = config.num_key_value_heads - self.use_gqa = (self.num_attention_heads != self.num_key_value_heads) - - self.use_flash_attn = (args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2 or \ - args.use_flash_attn_builder) \ - and attention_type == AttnType.self_attn \ + self.use_gqa = self.num_attention_heads != self.num_key_value_heads + + self.use_flash_attn = ( + ( + args.use_flash_attn_v1 + or args.use_flash_attn_triton + or args.use_flash_attn_v2 + or args.use_flash_attn_builder + ) + and attention_type == AttnType.self_attn and self.attn_mask_type == AttnMaskType.causal + ) self.use_flash_attn_triton = args.use_flash_attn_triton if self.use_flash_attn: global flash_attn_builder @@ -528,38 +616,53 @@ def __init__(self, config, layer_number, flash_attn_builder = None if args.use_flash_attn_v1: - assert flash_attn_unpadded_func != None, "Cannot import FlashAttention v1 " + assert ( + flash_attn_unpadded_func != None + ), "Cannot import FlashAttention v1 " if args.use_flash_attn_v2: - assert flash_attn_varlen_func != None, "Cannot import FlashAttention v2 " + assert ( + flash_attn_varlen_func != None + ), "Cannot import FlashAttention v2 " if args.use_flash_attn_triton: assert flash_attn_func != None, "Cannot import FlashAttention triton " if args.use_flash_attn_builder: - assert flash_attn_builder != None, "Cannot find FlashAttention op builder " + assert ( + flash_attn_builder != None + ), "Cannot find FlashAttention op builder " - assert attention_type == AttnType.self_attn, ('FlashAttention code path only supports ' - 'self-attention for now') - assert self.attn_mask_type == AttnMaskType.causal, ('FlashAttention code path only ' - 'supports causal mask for now') + assert attention_type == AttnType.self_attn, ( + "FlashAttention code path only supports " "self-attention for now" + ) + assert self.attn_mask_type == AttnMaskType.causal, ( + "FlashAttention code path only " "supports causal mask for now" + ) if rearrange is None: - raise ImportError('einops is not installed, please install with pip install einops') + raise ImportError( + "einops is not installed, please install with pip install einops" + ) projection_size = config.kv_channels * config.num_attention_heads # Per attention head and per partition values. world_size = parallel_state.get_tensor_model_parallel_world_size() self.hidden_size_per_attention_head = core.utils.divide( - projection_size, config.num_attention_heads) + projection_size, config.num_attention_heads + ) self.num_attention_heads_per_partition = core.utils.divide( - config.num_attention_heads, world_size) + config.num_attention_heads, world_size + ) # Per GQA head and per partition values self.num_key_value_heads_per_partition = core.utils.divide( - config.num_key_value_heads, world_size) + config.num_key_value_heads, world_size + ) self.num_key_value_groups = core.utils.divide( - config.num_attention_heads, config.num_key_value_heads) + config.num_attention_heads, config.num_key_value_heads + ) kv_projection_size = config.kv_channels * config.num_key_value_heads assert self.hidden_size_per_attention_head == core.utils.divide( - kv_projection_size, config.num_key_value_heads) + kv_projection_size, config.num_key_value_heads + ) # Strided linear layer. if attention_type == AttnType.self_attn: @@ -569,7 +672,8 @@ def __init__(self, config, layer_number, config=config, init_method=config.init_method, bias=args.add_bias_linear, - gather_output=False) + gather_output=False, + ) else: assert attention_type == AttnType.cross_attn self.query = tensor_parallel.ColumnParallelLinear( @@ -578,8 +682,8 @@ def __init__(self, config, layer_number, config=config, init_method=config.init_method, bias=config.add_bias_linear, - gather_output=False) - + gather_output=False, + ) self.key_value = tensor_parallel.ColumnParallelLinear( config.hidden_size, @@ -587,32 +691,48 @@ def __init__(self, config, layer_number, config=config, init_method=config.init_method, bias=config.add_bias_linear, - gather_output=False) + gather_output=False, + ) # Currently FlashAttention only works with causal mask if self.use_flash_attn_triton: - local_attn = FlashSelfAttentionTriton(causal=True, attention_dropout=args.attention_dropout) + local_attn = FlashSelfAttentionTriton( + causal=True, attention_dropout=args.attention_dropout + ) elif self.use_flash_attn: - local_attn = FlashSelfAttention(causal=True, attention_dropout=config.attention_dropout) + local_attn = FlashSelfAttention( + causal=True, attention_dropout=config.attention_dropout + ) else: local_attn = CoreAttention(self.layer_number, config, self.attn_mask_type) - self.enable_ds_sequence_parallel = parallel_state.get_sequence_parallel_world_size() > 1 \ - or args.force_ds_sequence_parallel + self.enable_ds_sequence_parallel = ( + parallel_state.get_sequence_parallel_world_size() > 1 + or args.force_ds_sequence_parallel + ) if self.enable_ds_sequence_parallel: - assert dist_attn_supported, 'Distributed attention is not supported in this DeepSpeed version' - assert args.num_attention_heads % parallel_state.get_sequence_parallel_world_size() == 0 + assert ( + dist_attn_supported + ), "Distributed attention is not supported in this DeepSpeed version" + assert ( + args.num_attention_heads + % parallel_state.get_sequence_parallel_world_size() + == 0 + ) self.dist_attn = DistributedAttention( - local_attn, - parallel_state.get_sequence_parallel_group(), - gather_idx=1 if args.use_flash_attn_v1 or args.use_flash_attn_v2 else 0) + local_attn, + parallel_state.get_sequence_parallel_group(), + gather_idx=1 if args.use_flash_attn_v1 or args.use_flash_attn_v2 else 0, + ) # flash_attn_cuda assumes [b, s, nh, hd] layout, we need to make sure all2all gathers into the correct sequence dimension. else: if self.use_flash_attn: self.core_attention_flash = local_attn else: self.core_attention = local_attn - self.checkpoint_core_attention = config.recompute_granularity == 'selective' + self.checkpoint_core_attention = ( + config.recompute_granularity == "selective" + ) # Output. self.dense = tensor_parallel.RowParallelLinear( @@ -622,29 +742,38 @@ def __init__(self, config, layer_number, init_method=config.output_layer_init_method, bias=args.add_bias_linear, input_is_parallel=True, - skip_bias_add=True) - + skip_bias_add=True, + ) - def _checkpointed_attention_forward(self, query_layer, key_layer, - value_layer, attention_mask, - rotary_pos_emb=None): + def _checkpointed_attention_forward( + self, query_layer, key_layer, value_layer, attention_mask, rotary_pos_emb=None + ): """Forward method with activation checkpointing.""" + def custom_forward(*inputs): query_layer = inputs[0] key_layer = inputs[1] value_layer = inputs[2] attention_mask = inputs[3] - output_ = self.core_attention(query_layer, key_layer, - value_layer, attention_mask) + output_ = self.core_attention( + query_layer, key_layer, value_layer, attention_mask + ) return output_ - q_pos_emb, k_pos_emb = (None, None) if rotary_pos_emb is None \ - else rotary_pos_emb + q_pos_emb, k_pos_emb = ( + (None, None) if rotary_pos_emb is None else rotary_pos_emb + ) hidden_states = tensor_parallel.checkpoint( custom_forward, - False, query_layer, key_layer, value_layer, attention_mask, - q_pos_emb, k_pos_emb) + False, + query_layer, + key_layer, + value_layer, + attention_mask, + q_pos_emb, + k_pos_emb, + ) return hidden_states @@ -655,7 +784,8 @@ def _allocate_memory(self, inference_max_sequence_len, batch_size): self.num_attention_heads_per_partition, self.hidden_size_per_attention_head, dtype=self.params_dtype, - device=get_accelerator().current_device_name()) + device=get_accelerator().current_device_name(), + ) def repeat_kv(self, hidden_states, n_rep): slen, batch, num_key_value_heads_per_partition, head_dim = hidden_states.shape @@ -667,22 +797,36 @@ def repeat_kv(self, hidden_states, n_rep): return hidden_states.expand(slen, batch, n_rep, head_dim) else: hidden_states = hidden_states[:, :, :, None, :].expand( - slen, batch, num_key_value_heads_per_partition, n_rep, head_dim) - return hidden_states.reshape(slen, batch, - num_key_value_heads_per_partition * n_rep, - head_dim) - + slen, batch, num_key_value_heads_per_partition, n_rep, head_dim + ) + return hidden_states.reshape( + slen, batch, num_key_value_heads_per_partition * n_rep, head_dim + ) + def split_tensor(self, mixed_x_layer): - query_layer, key_layer, value_layer = torch.split(mixed_x_layer, [self.num_key_value_groups, 1, 1], dim=-2) - query_layer = query_layer.reshape(mixed_x_layer.shape[:2] + (self.num_attention_heads_per_partition, self.hidden_size_per_attention_head)) + query_layer, key_layer, value_layer = torch.split( + mixed_x_layer, [self.num_key_value_groups, 1, 1], dim=-2 + ) + query_layer = query_layer.reshape( + mixed_x_layer.shape[:2] + + ( + self.num_attention_heads_per_partition, + self.hidden_size_per_attention_head, + ) + ) key_layer = torch.squeeze(key_layer, -2) value_layer = torch.squeeze(value_layer, -2) return query_layer, key_layer, value_layer - def forward(self, hidden_states, attention_mask, - encoder_output=None, inference_params=None, - rotary_pos_emb=None): + def forward( + self, + hidden_states, + attention_mask, + encoder_output=None, + inference_params=None, + rotary_pos_emb=None, + ): # hidden_states: [sq, b, h] # ================================================= @@ -694,15 +838,20 @@ def forward(self, hidden_states, attention_mask, inf_max_seq_len = inference_params.max_sequence_len inf_max_batch_size = inference_params.max_batch_size inference_key_memory = self._allocate_memory( - inf_max_seq_len, inf_max_batch_size) + inf_max_seq_len, inf_max_batch_size + ) inference_value_memory = self._allocate_memory( - inf_max_seq_len, inf_max_batch_size) + inf_max_seq_len, inf_max_batch_size + ) inference_params.key_value_memory_dict[self.layer_number] = ( - inference_key_memory, inference_value_memory) + inference_key_memory, + inference_value_memory, + ) is_first_step = True else: - inference_key_memory, inference_value_memory = \ + inference_key_memory, inference_value_memory = ( inference_params.key_value_memory_dict[self.layer_number] + ) # ===================== # Query, Key, and Value @@ -713,43 +862,45 @@ def forward(self, hidden_states, attention_mask, mixed_x_layer, _ = self.query_key_value(hidden_states) # [sq, b, ((nq + 2 * nkv) * hn)] --> [sq, b, nkv, (nq // nkv + 2), hn] - new_tensor_shape = mixed_x_layer.size()[:-1] + \ - (-1, (self.num_key_value_groups + 2), - self.hidden_size_per_attention_head) + new_tensor_shape = mixed_x_layer.size()[:-1] + ( + -1, + (self.num_key_value_groups + 2), + self.hidden_size_per_attention_head, + ) mixed_x_layer = mixed_x_layer.view(*new_tensor_shape) # [sq, b, nkv, (nq // nkv + 2), hn] --> 3 [sq, b, np, hn] - (query_layer, - key_layer, - value_layer) = self.split_tensor(mixed_x_layer) + (query_layer, key_layer, value_layer) = self.split_tensor(mixed_x_layer) # Repeat kv if self.use_gqa: key_layer = self.repeat_kv(key_layer, self.num_key_value_groups) - value_layer = self.repeat_kv(value_layer, - self.num_key_value_groups) + value_layer = self.repeat_kv(value_layer, self.num_key_value_groups) else: - assert not self.use_gqa, 'GQA + cross-attn not tested yet' + assert not self.use_gqa, "GQA + cross-attn not tested yet" # Attention heads [sk, b, h] --> [sk, b, (np * 2 * hn)] mixed_kv_layer, _ = self.key_value(encoder_output) # [sk, b, (np * 2 * hn)] --> [sk, b, np, 2 * hn] - new_tensor_shape = mixed_kv_layer.size()[:-1] + \ - (self.num_attention_heads_per_partition, - 2 * self.hidden_size_per_attention_head) + new_tensor_shape = mixed_kv_layer.size()[:-1] + ( + self.num_attention_heads_per_partition, + 2 * self.hidden_size_per_attention_head, + ) mixed_kv_layer = mixed_kv_layer.view(*new_tensor_shape) # [sk, b, np, 2 * hn] --> 2 [sk, b, np, hn] - (key_layer, - value_layer) = tensor_parallel.split_tensor_along_last_dim(mixed_kv_layer, 2) + (key_layer, value_layer) = tensor_parallel.split_tensor_along_last_dim( + mixed_kv_layer, 2 + ) # Attention head [sq, b, h] --> [sq, b, hp] query_layer, _ = self.query(hidden_states) # [sq, b, hp] --> [sq, b, np, hn] - new_tensor_shape = query_layer.size()[:-1] + \ - (self.num_attention_heads_per_partition, - self.hidden_size_per_attention_head) + new_tensor_shape = query_layer.size()[:-1] + ( + self.num_attention_heads_per_partition, + self.hidden_size_per_attention_head, + ) query_layer = query_layer.view(*new_tensor_shape) # ================================== @@ -761,7 +912,7 @@ def forward(self, hidden_states, attention_mask, if isinstance(rotary_pos_emb, tuple): rotary_pos_emb = rotary_pos_emb else: - rotary_pos_emb = ((rotary_pos_emb,) * 2) + rotary_pos_emb = (rotary_pos_emb,) * 2 if inference_params: batch_start = inference_params.batch_size_offset @@ -771,15 +922,16 @@ def forward(self, hidden_states, attention_mask, sequence_end = sequence_start + key_layer.size(0) assert sequence_end <= inference_key_memory.size(0) # Copy key and values. - inference_key_memory[sequence_start:sequence_end, - batch_start:batch_end, ...] = key_layer - inference_value_memory[sequence_start:sequence_end, - batch_start:batch_end, ...] = value_layer - key_layer = inference_key_memory[ - :sequence_end, batch_start:batch_end, ...] + inference_key_memory[ + sequence_start:sequence_end, batch_start:batch_end, ... + ] = key_layer + inference_value_memory[ + sequence_start:sequence_end, batch_start:batch_end, ... + ] = value_layer + key_layer = inference_key_memory[:sequence_end, batch_start:batch_end, ...] value_layer = inference_value_memory[ - :sequence_end, batch_start:batch_end, ...] - + :sequence_end, batch_start:batch_end, ... + ] # adjust the key rotary positional embedding if rotary_pos_emb is not None: @@ -801,7 +953,6 @@ def forward(self, hidden_states, attention_mask, k_pos_emb = k_pos_emb[:sequence_end, :, :, :] rotary_pos_emb = (q_pos_emb, k_pos_emb) - # ================================== # core attention computation # ================================== @@ -820,37 +971,55 @@ def forward(self, hidden_states, attention_mask, batch_dim_idx = 1 if self.use_flash_attn: if not self.use_flash_attn_triton: - query_layer, key_layer, value_layer = [rearrange(x, 's b ... -> b s ...').contiguous() - for x in (query_layer, key_layer, value_layer)] + query_layer, key_layer, value_layer = [ + rearrange(x, "s b ... -> b s ...").contiguous() + for x in (query_layer, key_layer, value_layer) + ] batch_dim_idx = 0 - context_layer = self.dist_attn(query_layer, key_layer, value_layer, batch_dim_idx) + context_layer = self.dist_attn( + query_layer, key_layer, value_layer, batch_dim_idx + ) if not self.use_flash_attn_triton: - context_layer = rearrange(context_layer, 'b s h d -> s b (h d)').contiguous() + context_layer = rearrange( + context_layer, "b s h d -> s b (h d)" + ).contiguous() else: - context_layer = self.dist_attn(query_layer, key_layer, value_layer, attention_mask) + context_layer = self.dist_attn( + query_layer, key_layer, value_layer, attention_mask + ) else: if self.use_flash_attn: if not self.use_flash_attn_triton: - query_layer, key_layer, value_layer = [rearrange(x, 's b ... -> b s ...').contiguous() - for x in (query_layer, key_layer, value_layer)] + query_layer, key_layer, value_layer = [ + rearrange(x, "s b ... -> b s ...").contiguous() + for x in (query_layer, key_layer, value_layer) + ] if self.sequence_parallel: - context_layer = self.core_attention_flash(query_layer, key_layer, value_layer) + context_layer = self.core_attention_flash( + query_layer, key_layer, value_layer + ) else: with tensor_parallel.get_cuda_rng_tracker().fork(): - context_layer = self.core_attention_flash(query_layer, key_layer, value_layer) + context_layer = self.core_attention_flash( + query_layer, key_layer, value_layer + ) if not self.use_flash_attn_triton: - context_layer = rearrange(context_layer, 'b s h d -> s b (h d)').contiguous() + context_layer = rearrange( + context_layer, "b s h d -> s b (h d)" + ).contiguous() else: if self.checkpoint_core_attention: context_layer = self._checkpointed_attention_forward( - query_layer, key_layer, value_layer, attention_mask) + query_layer, key_layer, value_layer, attention_mask + ) else: context_layer = self.core_attention( - query_layer, key_layer, value_layer, attention_mask) + query_layer, key_layer, value_layer, attention_mask + ) # ================= # Output. [sq, b, h] @@ -873,22 +1042,21 @@ def bias_dropout_add(x, bias, residual, prob, training): def get_bias_dropout_add(training): def _bias_dropout_add(x, bias, residual, prob): return bias_dropout_add(x, bias, residual, prob, training) + return _bias_dropout_add @torch.jit.script -def bias_dropout_add_fused_train(x: torch.Tensor, - bias: Optional[torch.Tensor], - residual: torch.Tensor, - prob: float) -> torch.Tensor: +def bias_dropout_add_fused_train( + x: torch.Tensor, bias: Optional[torch.Tensor], residual: torch.Tensor, prob: float +) -> torch.Tensor: return bias_dropout_add(x, bias, residual, prob, True) @torch.jit.script -def bias_dropout_add_fused_inference(x: torch.Tensor, - bias: Optional[torch.Tensor], - residual: torch.Tensor, - prob: float) -> torch.Tensor: +def bias_dropout_add_fused_inference( + x: torch.Tensor, bias: Optional[torch.Tensor], residual: torch.Tensor, prob: float +) -> torch.Tensor: return bias_dropout_add(x, bias, residual, prob, False) @@ -899,10 +1067,15 @@ class ParallelTransformerLayer(MegatronModule): output of the same size. """ - def __init__(self, config, - layer_number, layer_type=LayerType.encoder, - self_attn_mask_type=AttnMaskType.padding, - drop_path_rate=0., num_experts=1): + def __init__( + self, + config, + layer_number, + layer_type=LayerType.encoder, + self_attn_mask_type=AttnMaskType.padding, + drop_path_rate=0.0, + num_experts=1, + ): # retriever=None): args = get_args() @@ -910,110 +1083,130 @@ def __init__(self, config, self.layer_number = layer_number self.layer_type = layer_type - self.apply_residual_connection_post_layernorm \ - = config.apply_residual_connection_post_layernorm + self.apply_residual_connection_post_layernorm = ( + config.apply_residual_connection_post_layernorm + ) self.bf16 = config.bf16 self.fp32_residual_connection = config.fp32_residual_connection # Layernorm on the input data. - if args.normalization == 'layernorm': - if get_accelerator().device_name() == 'cuda': + if args.normalization == "layernorm": + if get_accelerator().device_name() == "cuda": self.input_layernorm = LayerNorm( config.hidden_size, eps=config.layernorm_epsilon, no_persist_layer_norm=args.no_persist_layer_norm, sequence_parallel=config.sequence_parallel, apply_layernorm_1p=args.apply_layernorm_1p, - mem_efficient_ln=args.mem_efficient_ln) + mem_efficient_ln=args.mem_efficient_ln, + ) else: self.input_layernorm = LayerNorm( - config.hidden_size, - eps=config.layernorm_epsilon) + config.hidden_size, eps=config.layernorm_epsilon + ) else: - self.input_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, - sequence_parallel=config.sequence_parallel) + self.input_layernorm = RMSNorm( + config.hidden_size, + config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel, + ) + # self.input_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon_ # Self attention. self.self_attention = ParallelAttention( config, layer_number, attention_type=AttnType.self_attn, - attn_mask_type=self_attn_mask_type) + attn_mask_type=self_attn_mask_type, + ) self.hidden_dropout = config.hidden_dropout self.bias_dropout_fusion = config.bias_dropout_fusion self.drop_path = DropPath(drop_path_rate) if drop_path_rate > 0.0 else None # Layernorm on the attention output - if args.normalization == 'layernorm': - if get_accelerator().device_name() == 'cuda': + if args.normalization == "layernorm": + if get_accelerator().device_name() == "cuda": self.post_attention_layernorm = LayerNorm( config.hidden_size, eps=config.layernorm_epsilon, no_persist_layer_norm=not config.persist_layer_norm, sequence_parallel=config.sequence_parallel, apply_layernorm_1p=args.apply_layernorm_1p, - mem_efficient_ln=args.mem_efficient_ln) + mem_efficient_ln=args.mem_efficient_ln, + ) else: self.post_attention_layernorm = LayerNorm( - config.hidden_size, - eps=config.layernorm_epsilon) + config.hidden_size, eps=config.layernorm_epsilon + ) else: - self.post_attention_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, - sequence_parallel=config.sequence_parallel) + self.post_attention_layernorm = RMSNorm( + config.hidden_size, + config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel, + ) # Cross attention. - if self.layer_type in (LayerType.decoder, - LayerType.retro_decoder, - LayerType.retro_decoder_with_retriever, - LayerType.retro_encoder): + if self.layer_type in ( + LayerType.decoder, + LayerType.retro_decoder, + LayerType.retro_decoder_with_retriever, + LayerType.retro_encoder, + ): self.inter_attention = ParallelAttention( - config, - layer_number, - attention_type=AttnType.cross_attn) + config, layer_number, attention_type=AttnType.cross_attn + ) # Layernorm on the attention output. - if args.normalization == 'layernorm': + if args.normalization == "layernorm": self.post_inter_attention_layernorm = LayerNorm( config.hidden_size, eps=config.layernorm_epsilon, no_persist_layer_norm=not config.persist_layer_norm, sequence_parallel=config.sequence_parallel, apply_layernorm_1p=args.apply_layernorm_1p, - mem_efficient_ln=args.mem_efficient_ln) + mem_efficient_ln=args.mem_efficient_ln, + ) else: - self.post_inter_attention_layernorm = RMSNorm(config.hidden_size, - config.layernorm_epsilon, - sequence_parallel=config.sequence_parallel) + self.post_inter_attention_layernorm = RMSNorm( + config.hidden_size, + config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel, + ) # MLP self.num_experts = num_experts if args.num_experts_switch is not None: - self.mlp = SwitchMLP(config) # Megatron-LM's MoE + self.mlp = SwitchMLP(config) # Megatron-LM's MoE else: - if self.num_experts <= 1: # dense, not MoE + if self.num_experts <= 1: # dense, not MoE self.mlp = ParallelMLP(config) - else: # DeepSpeed's MoE + else: # DeepSpeed's MoE enable_expert_tensor_parallelism = args.enable_expert_tensor_parallelism - self.mlp = MoE(args.hidden_size, - ParallelMLP(config, - moe=True, - enable_expert_tensor_parallelism=enable_expert_tensor_parallelism), - num_experts=self.num_experts, - ep_size=args.moe_expert_parallel_size, - k=args.topk, - use_residual=(args.mlp_type == 'residual'), - capacity_factor=args.moe_train_capacity_factor, - eval_capacity_factor=args.moe_eval_capacity_factor, - min_capacity=args.moe_min_capacity, - drop_tokens=args.moe_token_dropping, - use_tutel=args.use_tutel, - enable_expert_tensor_parallelism=enable_expert_tensor_parallelism, - top2_2nd_expert_sampling=args.moe_top2_2nd_expert_sampling) + self.mlp = MoE( + args.hidden_size, + ParallelMLP( + config, + moe=True, + enable_expert_tensor_parallelism=enable_expert_tensor_parallelism, + ), + num_experts=self.num_experts, + ep_size=args.moe_expert_parallel_size, + k=args.topk, + use_residual=(args.mlp_type == "residual"), + capacity_factor=args.moe_train_capacity_factor, + eval_capacity_factor=args.moe_eval_capacity_factor, + min_capacity=args.moe_min_capacity, + drop_tokens=args.moe_token_dropping, + use_tutel=args.use_tutel, + enable_expert_tensor_parallelism=enable_expert_tensor_parallelism, + top2_2nd_expert_sampling=args.moe_top2_2nd_expert_sampling, + ) # Set bias+dropout+add fusion grad_enable execution handler. - TORCH_MAJOR = int(torch.__version__.split('.')[0]) - TORCH_MINOR = int(torch.__version__.split('.')[1]) + TORCH_MAJOR = int(torch.__version__.split(".")[0]) + TORCH_MINOR = int(torch.__version__.split(".")[1]) use_nvfuser = TORCH_MAJOR > 1 or (TORCH_MAJOR == 1 and TORCH_MINOR >= 10) - self.bias_dropout_add_exec_handler = \ - nullcontext if use_nvfuser else torch.enable_grad + self.bias_dropout_add_exec_handler = ( + nullcontext if use_nvfuser else torch.enable_grad + ) if args.retro_add_retriever: retro_args = get_retro_args() @@ -1031,23 +1224,24 @@ def __init__(self, config, pre_process=True, post_process=False, ) - self._retriever_key = 'retriever' + self._retriever_key = "retriever" else: self.retriever = None - def default_decoder_cross_attention(self, - encoder_output, - enc_dec_attn_mask, - layernorm_input, - layernorm_output, - bias_dropout_add_func): - '''Cross attention for a standard encoder-decoder model.''' + def default_decoder_cross_attention( + self, + encoder_output, + enc_dec_attn_mask, + layernorm_input, + layernorm_output, + bias_dropout_add_func, + ): + """Cross attention for a standard encoder-decoder model.""" # Attention. - attention_output, attention_bias = \ - self.inter_attention(layernorm_output, - enc_dec_attn_mask, - encoder_output=encoder_output) + attention_output, attention_bias = self.inter_attention( + layernorm_output, enc_dec_attn_mask, encoder_output=encoder_output + ) # Residual connection. if self.apply_residual_connection_post_layernorm: @@ -1061,21 +1255,17 @@ def default_decoder_cross_attention(self, # Bias-dropout-add. with self.bias_dropout_add_exec_handler(): layernorm_input = bias_dropout_add_func( - attention_output, - attention_bias, - residual, - self.hidden_dropout) + attention_output, attention_bias, residual, self.hidden_dropout + ) # Layer norm. layernorm_output = self.post_inter_attention_layernorm(layernorm_input) return layernorm_input, layernorm_output - def retro_encoder_cross_attention(self, - retriever_output, - layernorm_input, - layernorm_output, - bias_dropout_add_func): + def retro_encoder_cross_attention( + self, retriever_output, layernorm_input, layernorm_output, bias_dropout_add_func + ): """Cross attention for Retro encoder. Notation: @@ -1087,16 +1277,15 @@ def retro_encoder_cross_attention(self, r : Number of retrieved tokens (neighbors + continuation). """ - ns, bs, d = layernorm_output.shape # [r, bs * l * k, d] + ns, bs, d = layernorm_output.shape # [r, bs * l * k, d] # Divide sequence dimension into chunks. - chunked_outputs = layernorm_output.reshape(self.retro_retrieved_length, - -1, - self.retro_num_neighbors, - d) - chunked_outputs_before_layer_norm = \ - layernorm_input.reshape(self.retro_retrieved_length, -1, - self.retro_num_neighbors, d) # [r, bs*l, k, d] + chunked_outputs = layernorm_output.reshape( + self.retro_retrieved_length, -1, self.retro_num_neighbors, d + ) + chunked_outputs_before_layer_norm = layernorm_input.reshape( + self.retro_retrieved_length, -1, self.retro_num_neighbors, d + ) # [r, bs*l, k, d] # Per-chunk attention. layernorm_inputs = [] @@ -1104,51 +1293,55 @@ def retro_encoder_cross_attention(self, for k in range(self.retro_num_neighbors): # Attention. - chunked_output = chunked_outputs[:,:,k].contiguous() - attention_output, attention_bias = \ - self.inter_attention( - chunked_output, # Q (neighbor embedding) - None, - encoder_output=retriever_output) # K, V (hidden act) + chunked_output = chunked_outputs[:, :, k].contiguous() + attention_output, attention_bias = self.inter_attention( + chunked_output, # Q (neighbor embedding) + None, + encoder_output=retriever_output, + ) # K, V (hidden act) # Residual connection. if self.apply_residual_connection_post_layernorm: residual = chunked_output else: - residual = chunked_outputs_before_layer_norm[:,:,k] + residual = chunked_outputs_before_layer_norm[:, :, k] # Re-enable torch grad to enable fused optimization. with torch.enable_grad(): layernorm_input = bias_dropout_add_func( attention_output, - None if attention_bias is None else attention_bias.expand_as(residual), + ( + None + if attention_bias is None + else attention_bias.expand_as(residual) + ), residual, - self.hidden_dropout) + self.hidden_dropout, + ) layernorm_inputs.append(layernorm_input) # Layer norm. - layernorm_output = \ - self.post_inter_attention_layernorm(layernorm_input) + layernorm_output = self.post_inter_attention_layernorm(layernorm_input) layernorm_outputs.append(layernorm_output) # Concatenate layer norms. # layernorm_input : [r, k * bs * l, d] # layernorm_output : [r, k * bs * l, d] - layernorm_input = \ - torch.stack(layernorm_inputs, dim=1).reshape(ns, bs, d) - layernorm_output = \ - torch.stack(layernorm_outputs, dim=1).reshape(ns, bs, d) + layernorm_input = torch.stack(layernorm_inputs, dim=1).reshape(ns, bs, d) + layernorm_output = torch.stack(layernorm_outputs, dim=1).reshape(ns, bs, d) return layernorm_input, layernorm_output - def retro_decoder_cross_attention(self, - retriever_input, - retriever_output, - retriever_attn_mask, - layernorm_input, - layernorm_output, - inference_params, - bias_dropout_add_func): + def retro_decoder_cross_attention( + self, + retriever_input, + retriever_output, + retriever_attn_mask, + layernorm_input, + layernorm_output, + inference_params, + bias_dropout_add_func, + ): """Cross attention for Retro decoder. Notation: @@ -1169,22 +1362,27 @@ def retro_decoder_cross_attention(self, first_ns = ns % self.retro_chunk_length if first_ns > 0: raise Exception("test this case.") - first_chunk, rest_chunk = \ - layernorm_output[:first_ns], layernorm_output[first_ns:] + first_chunk, rest_chunk = ( + layernorm_output[:first_ns], + layernorm_output[first_ns:], + ) first_chunk = torch.nn.functional.pad( first_chunk, (0, 0, 0, 0, 0, self.retro_chunk_length - first_ns), - 'constant', - 0) - chunked_output = \ - torch.cat((first_chunk, rest_chunk), dim=0) # [l * m, bs, d] + "constant", + 0, + ) + chunked_output = torch.cat( + (first_chunk, rest_chunk), dim=0 + ) # [l * m, bs, d] else: - chunked_output = layernorm_output # [l * m, bs, d] - chunked_output = chunked_output \ - .reshape(l, self.retro_chunk_length, bs, d) \ - .permute(1, 2, 0, 3) \ - .reshape(self.retro_chunk_length, bs * l, d) \ + chunked_output = layernorm_output # [l * m, bs, d] + chunked_output = ( + chunked_output.reshape(l, self.retro_chunk_length, bs, d) + .permute(1, 2, 0, 3) + .reshape(self.retro_chunk_length, bs * l, d) .contiguous() + ) # Get Encoder Output retriever_output = self.retriever( @@ -1192,9 +1390,11 @@ def retro_decoder_cross_attention(self, attention_mask=retriever_attn_mask, retriever_output=chunked_output, retriever_attn_mask=retriever_attn_mask, - inference_params=inference_params) # [r, k * bs * l , d] + inference_params=inference_params, + ) # [r, k * bs * l , d] retriever_output = retriever_output.reshape( - self.retro_retrieved_length * self.retro_num_neighbors, bs * l, d) # [r * k, bs * l, d] + self.retro_retrieved_length * self.retro_num_neighbors, bs * l, d + ) # [r * k, bs * l, d] # Chunks. pad = (ns - 1) % self.retro_chunk_length @@ -1202,18 +1402,20 @@ def retro_decoder_cross_attention(self, padded_chunks = torch.nn.functional.pad( attending_chunks, (0, 0, 0, 0, 0, self.retro_chunk_length - 1), - 'constant', 0) - padded_chunked_output = padded_chunks \ - .reshape(l, self.retro_chunk_length, bs, d) \ - .permute(1, 2, 0, 3) + "constant", + 0, + ) + padded_chunked_output = padded_chunks.reshape( + l, self.retro_chunk_length, bs, d + ).permute(1, 2, 0, 3) padded_chunked_output = padded_chunked_output.reshape( - self.retro_chunk_length, bs * l, d).contiguous() + self.retro_chunk_length, bs * l, d + ).contiguous() # Encoder output. - attention_output, attention_bias = \ - self.inter_attention(padded_chunked_output, - None, - encoder_output=retriever_output) + attention_output, attention_bias = self.inter_attention( + padded_chunked_output, None, encoder_output=retriever_output + ) # Residual connection. if self.apply_residual_connection_post_layernorm: @@ -1225,17 +1427,27 @@ def retro_decoder_cross_attention(self, with torch.enable_grad(): layernorm_input = bias_dropout_add_func( attention_output, - None if attention_bias is None else attention_bias.expand_as(attention_output), + ( + None + if attention_bias is None + else attention_bias.expand_as(attention_output) + ), torch.zeros_like(attention_output), - self.hidden_dropout) - layernorm_input = layernorm_input \ - .reshape(self.retro_chunk_length, bs, l, d) \ - .permute(2, 0, 1, 3) # [l, m, bs, d] - layernorm_input = layernorm_input.reshape(self.retro_chunk_length * l, bs, d) + self.hidden_dropout, + ) + layernorm_input = layernorm_input.reshape( + self.retro_chunk_length, bs, l, d + ).permute( + 2, 0, 1, 3 + ) # [l, m, bs, d] + layernorm_input = layernorm_input.reshape( + self.retro_chunk_length * l, bs, d + ) layernorm_input = torch.nn.functional.pad( - layernorm_input, - (0, 0, 0, 0, pad, 0), - 'constant', 0)[:ns] # [ns, b, d] + layernorm_input, (0, 0, 0, 0, pad, 0), "constant", 0 + )[ + :ns + ] # [ns, b, d] layernorm_input = layernorm_input + residual # Layer norm post the decoder attention @@ -1243,26 +1455,31 @@ def retro_decoder_cross_attention(self, return retriever_output, layernorm_input, layernorm_output - def forward(self, hidden_states, attention_mask=None, - encoder_output=None, enc_dec_attn_mask=None, - retriever_input=None, - retriever_output=None, - retriever_attn_mask=None, - inference_params=None, - rotary_pos_emb=None, - aggregated_moe_loss=None): + def forward( + self, + hidden_states, + attention_mask=None, + encoder_output=None, + enc_dec_attn_mask=None, + retriever_input=None, + retriever_output=None, + retriever_attn_mask=None, + inference_params=None, + rotary_pos_emb=None, + aggregated_moe_loss=None, + ): # hidden_states: [s, b, h] # Layer norm at the beginning of the transformer layer. layernorm_output = self.input_layernorm(hidden_states) # Self attention. - attention_output, attention_bias = \ - self.self_attention( - layernorm_output, - attention_mask, - inference_params=inference_params, - rotary_pos_emb=rotary_pos_emb) + attention_output, attention_bias = self.self_attention( + layernorm_output, + attention_mask, + inference_params=inference_params, + rotary_pos_emb=rotary_pos_emb, + ) # Residual connection. if self.apply_residual_connection_post_layernorm: @@ -1287,14 +1504,14 @@ def forward(self, hidden_states, attention_mask=None, attention_bias = attention_bias.expand_as(residual) with self.bias_dropout_add_exec_handler(): layernorm_input = bias_dropout_add_func( - attention_output, - attention_bias, - residual, - self.hidden_dropout) + attention_output, attention_bias, residual, self.hidden_dropout + ) else: - out = torch.nn.functional.dropout(attention_output + attention_bias, - p=self.hidden_dropout, - training=self.training) + out = torch.nn.functional.dropout( + attention_output + attention_bias, + p=self.hidden_dropout, + training=self.training, + ) layernorm_input = residual + self.drop_path(out) # Layer norm post the self attention. @@ -1304,23 +1521,25 @@ def forward(self, hidden_states, attention_mask=None, if self.layer_type == LayerType.encoder: pass elif self.layer_type == LayerType.decoder: - layernorm_input, layernorm_output = \ - self.default_decoder_cross_attention( - encoder_output, - enc_dec_attn_mask, - layernorm_input, - layernorm_output, - bias_dropout_add_func) + layernorm_input, layernorm_output = self.default_decoder_cross_attention( + encoder_output, + enc_dec_attn_mask, + layernorm_input, + layernorm_output, + bias_dropout_add_func, + ) elif self.layer_type == LayerType.retro_encoder: - layernorm_input, layernorm_output = \ - self.retro_encoder_cross_attention( - retriever_output, - layernorm_input, - layernorm_output, - bias_dropout_add_func) - elif self.layer_type in (LayerType.retro_decoder, - LayerType.retro_decoder_with_retriever): - retriever_output, layernorm_input, layernorm_output = \ + layernorm_input, layernorm_output = self.retro_encoder_cross_attention( + retriever_output, + layernorm_input, + layernorm_output, + bias_dropout_add_func, + ) + elif self.layer_type in ( + LayerType.retro_decoder, + LayerType.retro_decoder_with_retriever, + ): + retriever_output, layernorm_input, layernorm_output = ( self.retro_decoder_cross_attention( retriever_input, retriever_output, @@ -1328,14 +1547,19 @@ def forward(self, hidden_states, attention_mask=None, layernorm_input, layernorm_output, inference_params, - bias_dropout_add_func) + bias_dropout_add_func, + ) + ) else: - raise Exception("Unsupported layer type, '%s'." % - self.layer_type.name) + raise Exception("Unsupported layer type, '%s'." % self.layer_type.name) # MLP. - moe_loss = torch.tensor(0.0, device=layernorm_output.device, dtype=layernorm_output.dtype) - mlp_bias = torch.tensor(0.0, device=layernorm_output.device, dtype=layernorm_output.dtype) + moe_loss = torch.tensor( + 0.0, device=layernorm_output.device, dtype=layernorm_output.dtype + ) + mlp_bias = torch.tensor( + 0.0, device=layernorm_output.device, dtype=layernorm_output.dtype + ) if self.num_experts == 1: mlp_output, mlp_bias = self.mlp(layernorm_output) @@ -1357,10 +1581,8 @@ def forward(self, hidden_states, attention_mask=None, mlp_bias = mlp_bias.expand_as(residual) with self.bias_dropout_add_exec_handler(): output = bias_dropout_add_func( - mlp_output, - mlp_bias, - residual, - self.hidden_dropout) + mlp_output, mlp_bias, residual, self.hidden_dropout + ) # Jit compiled function creates 'view' tensor. This tensor # potentially gets saved in the MPU checkpoint function context, @@ -1368,16 +1590,16 @@ def forward(self, hidden_states, attention_mask=None, # won't result in memory savings (like the data loader, or # p2p_communication), it serves to document the origin of this # 'view' tensor. - output = core.utils.make_viewless_tensor(inp = output, - requires_grad = output.requires_grad, - keep_graph = True) + output = core.utils.make_viewless_tensor( + inp=output, requires_grad=output.requires_grad, keep_graph=True + ) else: if mlp_bias is not None: mlp_output = mlp_output + mlp_bias - out = torch.nn.functional.dropout(mlp_output, - p=self.hidden_dropout, - training=self.training) + out = torch.nn.functional.dropout( + mlp_output, p=self.hidden_dropout, training=self.training + ) output = residual + self.drop_path(out) if self.layer_type == LayerType.retro_decoder_with_retriever: @@ -1406,25 +1628,47 @@ class ParallelTransformerLayerPipe(ParallelTransformerLayer): If no mask is provided, the module will query `self._args.attn_mask` for the mask and only return `super().forward(...)` """ - def __init__(self, config, - layer_number, layer_type=LayerType.encoder, - self_attn_mask_type=AttnMaskType.padding, - drop_path_rate=0., num_experts=1, - input_aggregated_moe_loss=False, return_aggregated_moe_loss=False): + + def __init__( + self, + config, + layer_number, + layer_type=LayerType.encoder, + self_attn_mask_type=AttnMaskType.padding, + drop_path_rate=0.0, + num_experts=1, + input_aggregated_moe_loss=False, + return_aggregated_moe_loss=False, + ): self.input_aggregated_moe_loss = input_aggregated_moe_loss self.return_aggregated_moe_loss = return_aggregated_moe_loss - super().__init__(config, layer_number, layer_type, self_attn_mask_type, drop_path_rate, num_experts) + super().__init__( + config, + layer_number, + layer_type, + self_attn_mask_type, + drop_path_rate, + num_experts, + ) def forward(self, inputs, **kwargs): assert torch.is_tensor(inputs) or isinstance(inputs, tuple) - if not hasattr(self, '_args'): + if not hasattr(self, "_args"): self._args = get_args() - rotary_pos_emb = self._args.rotary_pos_emb if self._args.use_rotary_position_embeddings else None + rotary_pos_emb = ( + self._args.rotary_pos_emb + if self._args.use_rotary_position_embeddings + else None + ) if torch.is_tensor(inputs) or len(inputs) == 1: - assert not self.input_aggregated_moe_loss, f'Expecting an input tuple of size >= 2' + assert ( + not self.input_aggregated_moe_loss + ), f"Expecting an input tuple of size >= 2" # No attention mask forwarded, search for args.attn_mask hidden_states, attention_mask = inputs, self._args.attn_mask - output, moe_loss = super().forward(hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb) + output, moe_loss = super().forward( + hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb + ) return (output, moe_loss) if self.return_aggregated_moe_loss else output elif len(inputs) in (2, 3): # Attention mask and aggregated_moe can both be activations. @@ -1437,22 +1681,28 @@ def forward(self, inputs, **kwargs): hidden_states, attention_mask = inputs[0], inputs[1] return_attention_mask = True else: - hidden_states, attention_mask, aggregated_moe_loss = inputs[0], inputs[1], inputs[2] + hidden_states, attention_mask, aggregated_moe_loss = ( + inputs[0], + inputs[1], + inputs[2], + ) # Forward aggregated_moe_loss to ParallelTransformerLayer for further accumulation if self.input_aggregated_moe_loss: - kwargs.update({'aggregated_moe_loss': aggregated_moe_loss}) + kwargs.update({"aggregated_moe_loss": aggregated_moe_loss}) - output, moe_loss = super().forward(hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb) + output, moe_loss = super().forward( + hidden_states, attention_mask, **kwargs, rotary_pos_emb=rotary_pos_emb + ) - ret = (output, ) + ret = (output,) if return_attention_mask: - ret += (attention_mask, ) + ret += (attention_mask,) if self.return_aggregated_moe_loss: - ret += (moe_loss, ) + ret += (moe_loss,) return ret else: - raise RuntimeError('Received more inputs than understood.') + raise RuntimeError("Received more inputs than understood.") class NoopTransformerLayer(MegatronModule): @@ -1475,15 +1725,20 @@ def __init__(self, layer_number): super().__init__() self.layer_number = layer_number - def forward(self, hidden_states, attention_mask, - encoder_output=None, enc_dec_attn_mask=None, - inference_params=None): + def forward( + self, + hidden_states, + attention_mask, + encoder_output=None, + enc_dec_attn_mask=None, + inference_params=None, + ): return hidden_states.clone() def _get_num_layers(args, model_type, is_decoder=False): """Compute the number of transformer layers resident on the current rank.""" - is_encoder_and_decoder_model = (model_type == ModelType.encoder_and_decoder) + is_encoder_and_decoder_model = model_type == ModelType.encoder_and_decoder if model_type == ModelType.retro_encoder: num_layers = args.retro_encoder_layers elif parallel_state.get_pipeline_model_parallel_world_size() > 1: @@ -1496,27 +1751,34 @@ def _get_num_layers(args, model_type, is_decoder=False): # the same whether or not a standalone embedding stage is used. num_ranks_in_encoder = ( args.pipeline_model_parallel_split_rank - 1 - if args.standalone_embedding_stage else - args.pipeline_model_parallel_split_rank + if args.standalone_embedding_stage + else args.pipeline_model_parallel_split_rank + ) + num_ranks_in_decoder = ( + args.transformer_pipeline_model_parallel_size - num_ranks_in_encoder + ) + assert args.encoder_num_layers % num_ranks_in_encoder == 0, ( + "encoder_num_layers (%d) must be divisible by number of ranks given to encoder (%d)" + % (args.encoder_num_layers, num_ranks_in_encoder) + ) + assert args.decoder_num_layers % num_ranks_in_decoder == 0, ( + "decoder_num_layers (%d) must be divisible by number of ranks given to decoder (%d)" + % (args.decoder_num_layers, num_ranks_in_decoder) ) - num_ranks_in_decoder = args.transformer_pipeline_model_parallel_size - num_ranks_in_encoder - assert args.encoder_num_layers % num_ranks_in_encoder == 0, \ - 'encoder_num_layers (%d) must be divisible by number of ranks given to encoder (%d)' % (args.encoder_num_layers, num_ranks_in_encoder) - assert args.decoder_num_layers % num_ranks_in_decoder == 0, \ - 'decoder_num_layers (%d) must be divisible by number of ranks given to decoder (%d)' % (args.decoder_num_layers, num_ranks_in_decoder) if parallel_state.is_pipeline_stage_before_split(): num_layers = ( 0 if args.standalone_embedding_stage - and parallel_state.get_pipeline_model_parallel_rank() == 0 else - args.encoder_num_layers // num_ranks_in_encoder + and parallel_state.get_pipeline_model_parallel_rank() == 0 + else args.encoder_num_layers // num_ranks_in_encoder ) else: num_layers = args.decoder_num_layers // num_ranks_in_decoder else: assert args.num_layers == args.encoder_num_layers - assert args.num_layers % args.transformer_pipeline_model_parallel_size == 0, \ - 'num_layers must be divisible by transformer_pipeline_model_parallel_size' + assert ( + args.num_layers % args.transformer_pipeline_model_parallel_size == 0 + ), "num_layers must be divisible by transformer_pipeline_model_parallel_size" # When a standalone embedding stage is used, all transformer layers # are divided among pipeline rank >= 1, while on pipeline rank 0, @@ -1525,8 +1787,8 @@ def _get_num_layers(args, model_type, is_decoder=False): num_layers = ( 0 if args.standalone_embedding_stage - and parallel_state.get_pipeline_model_parallel_rank() == 0 else - args.num_layers // args.transformer_pipeline_model_parallel_size + and parallel_state.get_pipeline_model_parallel_rank() == 0 + else args.num_layers // args.transformer_pipeline_model_parallel_size ) else: if not is_decoder: @@ -1536,14 +1798,15 @@ def _get_num_layers(args, model_type, is_decoder=False): return num_layers -def _get_layer_type(model_type, default_layer_type, retro_layer_numbers, - layer_number): +def _get_layer_type(model_type, default_layer_type, retro_layer_numbers, layer_number): args = get_args() if args.retro_add_retriever and layer_number in retro_layer_numbers: if model_type == ModelType.retro_decoder: - return LayerType.retro_decoder_with_retriever \ - if layer_number == retro_layer_numbers[0] \ - else LayerType.retro_decoder + return ( + LayerType.retro_decoder_with_retriever + if layer_number == retro_layer_numbers[0] + else LayerType.retro_decoder + ) elif model_type == ModelType.retro_encoder: return LayerType.retro_encoder else: @@ -1552,15 +1815,22 @@ def _get_layer_type(model_type, default_layer_type, retro_layer_numbers, return default_layer_type -def get_num_experts_per_layer(num_experts: list, num_layers: int, expert_interval: int, offset: int = 0) -> list: - assert len(num_experts) == 1 or len(num_experts) == num_layers // expert_interval, \ - 'num_experts must be either a single value or a list of the same length as the number of MoE layers' +def get_num_experts_per_layer( + num_experts: list, num_layers: int, expert_interval: int, offset: int = 0 +) -> list: + assert ( + len(num_experts) == 1 or len(num_experts) == num_layers // expert_interval + ), "num_experts must be either a single value or a list of the same length as the number of MoE layers" if len(num_experts) == 1: num_experts = num_experts * (num_layers // expert_interval) experts_per_layer = [] for i in range(num_layers): layer_num = i + 1 + offset - n_e = num_experts[(layer_num-1) // expert_interval] if layer_num % expert_interval == 0 else 1 + n_e = ( + num_experts[(layer_num - 1) // expert_interval] + if layer_num % expert_interval == 0 + else 1 + ) experts_per_layer.append(n_e) return experts_per_layer @@ -1568,14 +1838,18 @@ def get_num_experts_per_layer(num_experts: list, num_layers: int, expert_interva class ParallelTransformer(MegatronModule): """Transformer class.""" - def __init__(self, config, - model_type, layer_type=LayerType.encoder, - self_attn_mask_type=AttnMaskType.padding, - post_layer_norm=True, - pre_process=True, - post_process=True, - drop_path_rate=0.0, - num_experts=[1]): + def __init__( + self, + config, + model_type, + layer_type=LayerType.encoder, + self_attn_mask_type=AttnMaskType.padding, + post_layer_norm=True, + pre_process=True, + post_process=True, + drop_path_rate=0.0, + num_experts=[1], + ): super(ParallelTransformer, self).__init__() args = get_args() @@ -1598,14 +1872,15 @@ def __init__(self, config, self.recompute_granularity = config.recompute_granularity self.recompute_method = config.recompute_method self.recompute_num_layers = config.recompute_num_layers - self.distribute_saved_activations = \ + self.distribute_saved_activations = ( config.distribute_saved_activations and not config.sequence_parallel + ) self.sequence_parallel = config.sequence_parallel # Transformer Engine Init. self.transformer_engine_rope_available = False - if self.transformer_impl == 'transformer_engine': + if self.transformer_impl == "transformer_engine": global transformer_engine import transformer_engine from importlib.metadata import version @@ -1637,45 +1912,53 @@ def __init__(self, config, self.num_microbatches_in_previous_step = -1 self.microbatch_count = 0 - self.checkpoint_core_attention = config.recompute_granularity == 'selective' + self.checkpoint_core_attention = config.recompute_granularity == "selective" # Number of layers. - self.num_layers = _get_num_layers(args, model_type, - layer_type==LayerType.decoder) + self.num_layers = _get_num_layers( + args, model_type, layer_type == LayerType.decoder + ) self.drop_path_rates = [ - rate.item() for rate in - torch.linspace(0, self.drop_path_rate, config.num_layers)] + rate.item() + for rate in torch.linspace(0, self.drop_path_rate, config.num_layers) + ] self.retro_layer_numbers = None if model_type == ModelType.retro_decoder: retro_layer_start = 6 if config.num_layers <= 15 else 9 - self.retro_layer_numbers = \ - np.arange(retro_layer_start, args.num_layers + 1, 3).tolist() + self.retro_layer_numbers = np.arange( + retro_layer_start, args.num_layers + 1, 3 + ).tolist() if model_type == ModelType.retro_encoder: self.retro_layer_numbers = [1] # Transformer layers. if args.retro_add_retriever: - assert self.recompute_granularity != 'full', \ - "Full recompute not supported for Retro." - assert args.transformer_impl == 'local', \ - "Transformer engine does not support Retro layers." + assert ( + self.recompute_granularity != "full" + ), "Full recompute not supported for Retro." + assert ( + args.transformer_impl == "local" + ), "Transformer engine does not support Retro layers." + def build_layer(layer_number, n_e): - if args.transformer_impl == 'local': + if args.transformer_impl == "local": current_layer_type = _get_layer_type( - model_type, layer_type, self.retro_layer_numbers, - layer_number) + model_type, layer_type, self.retro_layer_numbers, layer_number + ) return ParallelTransformerLayer( config, layer_number, layer_type=current_layer_type, self_attn_mask_type=self_attn_mask_type, drop_path_rate=self.drop_path_rates[layer_number - 1], - num_experts=n_e) + num_experts=n_e, + ) else: - assert config.num_attention_heads == config.num_key_value_heads, \ - 'Transformer_engine does not support GQA' + assert ( + config.num_attention_heads == config.num_key_value_heads + ), "Transformer_engine does not support GQA" return transformer_engine.pytorch.TransformerLayer( config.hidden_size, config.ffn_hidden_size, @@ -1702,16 +1985,22 @@ def build_layer(layer_number, n_e): layer_type="encoder", drop_path_rate=self.drop_path_rates[layer_number - 1], set_parallel_mode=True, - fuse_qkv_params=True) + fuse_qkv_params=True, + ) if config.virtual_pipeline_model_parallel_size is not None: - assert config.num_layers % config.virtual_pipeline_model_parallel_size == 0, \ - 'num_layers_per_stage must be divisible by ' \ - 'virtual_pipeline_model_parallel_size' + assert ( + config.num_layers % config.virtual_pipeline_model_parallel_size == 0 + ), ( + "num_layers_per_stage must be divisible by " + "virtual_pipeline_model_parallel_size" + ) assert args.model_type != ModelType.encoder_and_decoder # Number of layers in each model chunk is the number of layers in the stage, # divided by the number of model chunks in a stage. - self.num_layers = self.num_layers // config.virtual_pipeline_model_parallel_size + self.num_layers = ( + self.num_layers // config.virtual_pipeline_model_parallel_size + ) # With 8 layers, 2 stages, and 4 model chunks, we want an assignment of # layers to stages like (each list is a model chunk): # Stage 0: [0] [2] [4] [6] @@ -1721,12 +2010,14 @@ def build_layer(layer_number, n_e): # Stage 0: [0, 1] [4, 5] # Stage 1: [2, 3] [6, 7] offset = parallel_state.get_virtual_pipeline_model_parallel_rank() * ( - config.num_layers // config.virtual_pipeline_model_parallel_size) + \ - (parallel_state.get_pipeline_model_parallel_rank() * self.num_layers) + config.num_layers // config.virtual_pipeline_model_parallel_size + ) + (parallel_state.get_pipeline_model_parallel_rank() * self.num_layers) else: # Each stage gets a contiguous set of layers. - if args.model_type == ModelType.encoder_and_decoder and \ - parallel_state.get_pipeline_model_parallel_world_size() > 1: + if ( + args.model_type == ModelType.encoder_and_decoder + and parallel_state.get_pipeline_model_parallel_world_size() > 1 + ): pipeline_rank = parallel_state.get_pipeline_model_parallel_rank() if layer_type == LayerType.encoder: offset = pipeline_rank * self.num_layers @@ -1734,7 +2025,9 @@ def build_layer(layer_number, n_e): num_ranks_in_enc = args.pipeline_model_parallel_split_rank offset = (pipeline_rank - num_ranks_in_enc) * self.num_layers else: - offset = parallel_state.get_pipeline_model_parallel_rank() * self.num_layers + offset = ( + parallel_state.get_pipeline_model_parallel_rank() * self.num_layers + ) if self.num_layers == 0: # When a standalone embedding stage is used (e.g., @@ -1746,11 +2039,13 @@ def build_layer(layer_number, n_e): # this, we assign a 'no-op' layer on these ranks, which will # disconnect the input tensor from the output tensor. self.num_layers = 1 - self.layers = torch.nn.ModuleList([ NoopTransformerLayer(1) ]) + self.layers = torch.nn.ModuleList([NoopTransformerLayer(1)]) else: # Build the layers self.layers = [] - experts_per_layer = get_num_experts_per_layer(num_experts, self.num_layers, args.expert_interval, offset) + experts_per_layer = get_num_experts_per_layer( + num_experts, self.num_layers, args.expert_interval, offset + ) for i in range(self.num_layers): layer_num = i + 1 + offset n_e = experts_per_layer[i] @@ -1761,41 +2056,54 @@ def build_layer(layer_number, n_e): if model_type == ModelType.retro_encoder: for layer in self.layers: if layer.self_attention.use_flash_attn: - layer.self_attention.core_attention_flash.dropout_p = \ + layer.self_attention.core_attention_flash.dropout_p = ( torch.nn.Dropout(args.retro_encoder_attention_dropout) + ) else: - layer.self_attention.core_attention.attention_dropout.p =\ + layer.self_attention.core_attention.attention_dropout.p = ( args.retro_encoder_attention_dropout + ) layer.hidden_dropout = args.retro_encoder_hidden_dropout if self.post_process and self.post_layer_norm: # Final layer norm before output. - if args.normalization == 'layernorm': - if get_accelerator().device_name() == 'cuda': + if args.normalization == "layernorm": + if get_accelerator().device_name() == "cuda": self.final_layernorm = LayerNorm( config.hidden_size, eps=config.layernorm_epsilon, no_persist_layer_norm=args.no_persist_layer_norm, sequence_parallel=config.sequence_parallel, apply_layernorm_1p=args.apply_layernorm_1p, - mem_efficient_ln=args.mem_efficient_ln) + mem_efficient_ln=args.mem_efficient_ln, + ) else: self.final_layernorm = LayerNorm( - config.hidden_size, - eps=config.layernorm_epsilon) + config.hidden_size, eps=config.layernorm_epsilon + ) else: - self.final_layernorm = RMSNorm(config.hidden_size, config.layernorm_epsilon, - sequence_parallel=config.sequence_parallel) + self.final_layernorm = RMSNorm( + config.hidden_size, + config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel, + ) def _get_layer(self, layer_number): return self.layers[layer_number] - def _checkpointed_forward(self, hidden_states, attention_mask, - encoder_output, enc_dec_attn_mask, - rotary_pos_emb, is_first_microbatch): + def _checkpointed_forward( + self, + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + rotary_pos_emb, + is_first_microbatch, + ): args = get_args() """Forward method with activation checkpointing.""" + def custom(start, end): def custom_forward(*args, **kwargs): x_, *args = args @@ -1807,11 +2115,14 @@ def custom_forward(*args, **kwargs): x_, moe_loss = output else: x_ = output - moe_loss = torch.tensor(0.0, device=x_.device, dtype=x_.dtype, requires_grad=True) + moe_loss = torch.tensor( + 0.0, device=x_.device, dtype=x_.dtype, requires_grad=True + ) moe_losses.append(moe_loss) return (x_, *moe_losses) + return custom_forward - + if args.deepspeed and args.deepspeed_activation_checkpointing: moe_losses = [] # Make sure memory is freed. @@ -1819,9 +2130,18 @@ def custom_forward(*args, **kwargs): l = 0 while l < self.num_layers: hidden_states, *local_moe_losses = tensor_parallel.checkpoint( - custom(l, l + self.checkpoint_num_layers), False, - hidden_states, attention_mask, encoder_output, enc_dec_attn_mask, - None, None, None, None, rotary_pos_emb) + custom(l, l + self.checkpoint_num_layers), + False, + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + None, + None, + None, + None, + rotary_pos_emb, + ) moe_losses.extend(local_moe_losses) l += self.checkpoint_num_layers @@ -1829,66 +2149,105 @@ def custom_forward(*args, **kwargs): else: moe_losses = [] te_forward_kwargs = {} - if self.transformer_impl == 'transformer_engine': - te_forward_kwargs['is_first_microbatch'] = is_first_microbatch + if self.transformer_impl == "transformer_engine": + te_forward_kwargs["is_first_microbatch"] = is_first_microbatch if self.transformer_engine_rope_available: - te_forward_kwargs['rotary_pos_emb'] = rotary_pos_emb + te_forward_kwargs["rotary_pos_emb"] = rotary_pos_emb - if self.recompute_method == 'uniform': + if self.recompute_method == "uniform": # Uniformly divide the total number of Transformer layers and # checkpoint the input activation of each divided chunk. # A method to further reduce memory usage reducing checkpoints. l = 0 while l < self.num_layers: - if self.transformer_impl == 'transformer_engine': - hidden_states, *local_moe_losses = transformer_engine.pytorch.distributed.checkpoint( - custom(l, l + self.recompute_num_layers), - self.distribute_saved_activations, - tensor_parallel.get_cuda_rng_tracker, - mpu.get_tensor_model_parallel_group(), - hidden_states, attention_mask, encoder_output, - enc_dec_attn_mask, **te_forward_kwargs) + if self.transformer_impl == "transformer_engine": + hidden_states, *local_moe_losses = ( + transformer_engine.pytorch.distributed.checkpoint( + custom(l, l + self.recompute_num_layers), + self.distribute_saved_activations, + tensor_parallel.get_cuda_rng_tracker, + mpu.get_tensor_model_parallel_group(), + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + **te_forward_kwargs, + ) + ) else: hidden_states, *local_moe_losses = tensor_parallel.checkpoint( custom(l, l + self.recompute_num_layers), self.distribute_saved_activations, - hidden_states, attention_mask, - encoder_output, enc_dec_attn_mask, - None, None, None, None, rotary_pos_emb) + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + None, + None, + None, + None, + rotary_pos_emb, + ) moe_losses.extend(local_moe_losses) l += self.recompute_num_layers - elif self.recompute_method == 'block': + elif self.recompute_method == "block": # Checkpoint the input activation of only a set number of individual # Transformer layers and skip the rest. # A method fully use the device memory removing redundant re-computation. for l in range(self.num_layers): if l < self.recompute_num_layers: - if self.transformer_impl == 'transformer_engine': - hidden_states, *local_moe_losses = transformer_engine.pytorch.distributed.checkpoint( - custom(l, l + 1), - self.distribute_saved_activations, - tensor_parallel.get_cuda_rng_tracker, - mpu.get_tensor_model_parallel_group(), - hidden_states, attention_mask, encoder_output, - enc_dec_attn_mask, **te_forward_kwargs) + if self.transformer_impl == "transformer_engine": + hidden_states, *local_moe_losses = ( + transformer_engine.pytorch.distributed.checkpoint( + custom(l, l + 1), + self.distribute_saved_activations, + tensor_parallel.get_cuda_rng_tracker, + mpu.get_tensor_model_parallel_group(), + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + **te_forward_kwargs, + ) + ) else: - hidden_states, *local_moe_losses = tensor_parallel.checkpoint( - custom(l, l + 1), - self.distribute_saved_activations, - hidden_states, attention_mask, - encoder_output, enc_dec_attn_mask, - None, None, None, None, rotary_pos_emb) + hidden_states, *local_moe_losses = ( + tensor_parallel.checkpoint( + custom(l, l + 1), + self.distribute_saved_activations, + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + None, + None, + None, + None, + rotary_pos_emb, + ) + ) else: - if self.transformer_impl == 'transformer_engine': + if self.transformer_impl == "transformer_engine": hidden_states, *local_moe_losses = custom(l, l + 1)( - hidden_states, attention_mask, encoder_output, - enc_dec_attn_mask, **te_forward_kwargs) + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + **te_forward_kwargs, + ) else: hidden_states, *local_moe_losses = custom(l, l + 1)( - hidden_states, attention_mask, - encoder_output, enc_dec_attn_mask, - None, None, None, None, rotary_pos_emb) - + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + None, + None, + None, + None, + rotary_pos_emb, + ) + moe_losses.extend(local_moe_losses) else: raise ValueError("Invalid activation recompute method.") @@ -1904,19 +2263,25 @@ def set_input_tensor(self, input_tensor): forward_step_func""" self.input_tensor = input_tensor - def forward(self, hidden_states, attention_mask, - encoder_output=None, enc_dec_attn_mask=None, - retriever_input=None, - retriever_output=None, - retriever_attn_mask=None, - inference_params=None, - rotary_pos_emb=None): + def forward( + self, + hidden_states, + attention_mask, + encoder_output=None, + enc_dec_attn_mask=None, + retriever_input=None, + retriever_output=None, + retriever_attn_mask=None, + inference_params=None, + rotary_pos_emb=None, + ): # hidden_states: [s, b, h] # Checks. if inference_params: - assert self.recompute_granularity is None, \ - 'inference does not work with activation checkpointing' + assert ( + self.recompute_granularity is None + ), "inference does not work with activation checkpointing" # TODO: Below old DeepSpeed code are commented because it's unsure whether # it is still relevant. @@ -1971,64 +2336,77 @@ def forward(self, hidden_states, attention_mask, with rng_context: # The fp8_autocast context manager is a no-op when enabled=True # The if...else serves to short circuit name resolution for fp8_autocast - with transformer_engine.pytorch.fp8_autocast( - enabled=self.use_fp8, - fp8_recipe=self.fp8_recipe, - fp8_group=self.fp8_group - ) if self.use_fp8 else nullcontext(): + with ( + transformer_engine.pytorch.fp8_autocast( + enabled=self.use_fp8, + fp8_recipe=self.fp8_recipe, + fp8_group=self.fp8_group, + ) + if self.use_fp8 + else nullcontext() + ): # Determine if the current iteration is first microbatch if self.num_microbatches_in_previous_step != get_num_microbatches(): - self.microbatch_count = 0 # Reset count on new batch size rampup interval + self.microbatch_count = ( + 0 # Reset count on new batch size rampup interval + ) self.num_microbatches_in_previous_step = get_num_microbatches() - is_first_microbatch = self.microbatch_count % get_num_microbatches() == 0 + is_first_microbatch = ( + self.microbatch_count % get_num_microbatches() == 0 + ) # Forward pass. moe_losses = [] if self.checkpoint_activations: - hidden_states, moe_losses = self._checkpointed_forward(hidden_states, - attention_mask, - encoder_output, - enc_dec_attn_mask, - rotary_pos_emb, - is_first_microbatch) - elif self.recompute_granularity == 'full': - hidden_states, moe_losses = self._checkpointed_forward(hidden_states, - attention_mask, - encoder_output, - enc_dec_attn_mask, - rotary_pos_emb, - is_first_microbatch) + hidden_states, moe_losses = self._checkpointed_forward( + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + rotary_pos_emb, + is_first_microbatch, + ) + elif self.recompute_granularity == "full": + hidden_states, moe_losses = self._checkpointed_forward( + hidden_states, + attention_mask, + encoder_output, + enc_dec_attn_mask, + rotary_pos_emb, + is_first_microbatch, + ) else: forward_kwargs = { - 'encoder_output': encoder_output, - 'enc_dec_attn_mask': enc_dec_attn_mask, - 'inference_params': inference_params, + "encoder_output": encoder_output, + "enc_dec_attn_mask": enc_dec_attn_mask, + "inference_params": inference_params, } - if self.transformer_impl == 'transformer_engine': - forward_kwargs['is_first_microbatch'] = is_first_microbatch - forward_kwargs['checkpoint_core_attention'] = self.checkpoint_core_attention + if self.transformer_impl == "transformer_engine": + forward_kwargs["is_first_microbatch"] = is_first_microbatch + forward_kwargs["checkpoint_core_attention"] = ( + self.checkpoint_core_attention + ) if self.transformer_engine_rope_available: - forward_kwargs['rotary_pos_emb'] = rotary_pos_emb + forward_kwargs["rotary_pos_emb"] = rotary_pos_emb else: - forward_kwargs['rotary_pos_emb'] = rotary_pos_emb - forward_kwargs['retriever_input'] = retriever_input - forward_kwargs['retriever_output'] = retriever_output - forward_kwargs['retriever_attn_mask'] = retriever_attn_mask + forward_kwargs["rotary_pos_emb"] = rotary_pos_emb + forward_kwargs["retriever_input"] = retriever_input + forward_kwargs["retriever_output"] = retriever_output + forward_kwargs["retriever_attn_mask"] = retriever_attn_mask for index in range(self.num_layers): layer = self._get_layer(index) hidden_states = layer( - hidden_states, - attention_mask, - **forward_kwargs) + hidden_states, attention_mask, **forward_kwargs + ) # First Retro decoder layer returns both hidden_states # and retriever_output. Make retriever_output available # to subsequence Retro layers. if isinstance(hidden_states, tuple): - assert (len(hidden_states) == 2 or len(hidden_states) == 3) + assert len(hidden_states) == 2 or len(hidden_states) == 3 if len(hidden_states) == 2: if not self.ds_inference: hidden_states, moe_loss = hidden_states @@ -2054,6 +2432,7 @@ def forward(self, hidden_states, attention_mask, return (hidden_states, *moe_losses) + class LMHeadPipe(MegatronModule): """ Arguments: @@ -2067,11 +2446,13 @@ class LMHeadPipe(MegatronModule): def __init__(self, hidden_size, vocab_size, config): args = get_args() super(LMHeadPipe, self).__init__() - self.lm_head = tensor_parallel.ColumnParallelLinear(input_size=hidden_size, - output_size=vocab_size, - bias=False, - config=config, - init_method=config.init_method,) + self.lm_head = tensor_parallel.ColumnParallelLinear( + input_size=hidden_size, + output_size=vocab_size, + bias=False, + config=config, + init_method=config.init_method, + ) def forward(self, inputs, **kwargs): assert torch.is_tensor(inputs) or isinstance(inputs, tuple) @@ -2080,10 +2461,10 @@ def forward(self, inputs, **kwargs): else: hidden_states = inputs - if not hasattr(self, '_args'): + if not hasattr(self, "_args"): self._args = get_args() - if hasattr(self._args, 'attn_mask'): + if hasattr(self._args, "attn_mask"): attention_mask = None else: attention_mask = inputs[1] @@ -2091,7 +2472,7 @@ def forward(self, inputs, **kwargs): logits, _ = self.lm_head(hidden_states) # If cmd args has attn_mask, we don't forward it as an activation. - if hasattr(self._args, 'attn_mask'): + if hasattr(self._args, "attn_mask"): return logits else: return logits, attention_mask From 12f6f8e5563e5fa1d21838c2a273e8cd5e096c0c Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Thu, 12 Sep 2024 05:11:58 +0000 Subject: [PATCH 319/590] fix check --- megatron/utils.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 67dc366d50c..8a9f3e78581 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -41,7 +41,7 @@ try: import dftracer except: - _DFTRACER_EXIST=True + _DFTRACER_EXIST=False try: import dlio_profiler @@ -80,7 +80,7 @@ def flush(self): def reset(self): return def log_static(self, func): - return func + return class dftracer(object): def __init__(self,): self.type = None From 5ac877a880447a7b0028d4959dd290e56eafe512 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 12 Sep 2024 09:17:33 -0500 Subject: [PATCH 320/590] Update `megatron/training_log_alcf.py` --- megatron/training_log_alcf.py | 72 ++++++++++++++++++++++++----------- 1 file changed, 50 insertions(+), 22 deletions(-) diff --git a/megatron/training_log_alcf.py b/megatron/training_log_alcf.py index 7229549e24a..dcd872971d7 100644 --- a/megatron/training_log_alcf.py +++ b/megatron/training_log_alcf.py @@ -39,7 +39,7 @@ # from megatron.optimizer_param_scheduler import OptimizerParamScheduler # from megatron.profiler import on_step_begin, on_step_end, setup_profiler, trigger # from megatron.utils import check_adlr_autoresume_termination -from megatron.utils import found_kill_switch, unwrap_model +# from megatron.utils import found_kill_switch, unwrap_model import ezpz as ez # from megatron.utils import calc_params_l2_norm @@ -78,6 +78,37 @@ class InteropLoggingTool(Enum): log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") +def num_floating_point_operations(args, batch_size): + # Group Query Attention. + # if not args.group_query_attention: + if not args.num_key_value_heads: + args.num_key_value_heads = args.num_attention_heads + # args.num_query_groups = args.num_attention_heads + # MoE. + # num_experts_routed_to = 1 if args.num_experts is None else args.moe_router_topk + num_experts_routed_to = 1 if args.num_experts is None else args.topk + gated_linear_multiplier = 3 / 2 if args.swiglu else 1 + return ( + 12 + * batch_size + * args.seq_length + * args.num_layers + * args.hidden_size + * args.hidden_size + * ( + 1 + + ( + (args.ffn_hidden_size / args.hidden_size) + * num_experts_routed_to + * gated_linear_multiplier + ) + + (args.num_key_value_heads / args.num_attention_heads) + + (args.seq_length / args.hidden_size) + + (args.padded_vocab_size / (2 * args.num_layers * args.hidden_size)) + ) + ) + + def training_log( loss_dict, total_loss_dict, @@ -116,12 +147,10 @@ def training_log( ) # Update losses and set nan iterations got_nan = False + _zero = torch.tensor([0.0]).to(DEVICE) for key in loss_dict: if not skipped_iter: - total_loss_dict[key] = ( - total_loss_dict.get(key, get_accelerator().FloatTensor([0.0])) - + loss_dict[key] - ) + total_loss_dict[key] = total_loss_dict.get(key, _zero) + loss_dict[key] else: value = loss_dict[key].float().sum().item() is_nan = value == float("inf") or value == -float("inf") or value != value @@ -170,7 +199,7 @@ def training_log( # Tensorboard values. # Timer requires all the ranks to call. if args.log_timers_to_tensorboard and ( - iteration % args.tensorboard_log_interval == 0 + iteration % args.tensorboard_log_interval == 0 and writer is not None ): timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) if writer and (iteration % args.tensorboard_log_interval == 0): @@ -389,18 +418,15 @@ def training_log( abs(param.max().item()), abs(param.min().item()), ) - # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if args.zero_stage > 0: # ZeRO partiions optimizer states - # opt_stats = opt_stats.clone().detach() - # opt_stats = get_accelerator().FloatTensor - opt_stats = get_accelerator().FloatTensor(opt_stats) + # opt_stats = get_accelerator().FloatTensor(opt_stats) + opt_stats = torch.tensor(opt_stats).to(DEVICE) torch.distributed.all_reduce( opt_stats, group=mpu.get_sequence_data_parallel_group() ) # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - # opt_stats_2 = opt_stats_2.clone().detach() - opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, @@ -408,13 +434,13 @@ def training_log( ) if args.tensor_model_parallel_size > 1: - # opt_stats = opt_stats.clone().detach() - opt_stats = get_accelerator().FloatTensor(opt_stats) + opt_stats = torch.tensor(opt_stats).to(DEVICE) + # opt_stats = get_accelerator().FloatTensor(opt_stats) torch.distributed.all_reduce( opt_stats, group=mpu.get_tensor_model_parallel_group() ) - # opt_stats_2 = opt_stats_2.clone().detach() - opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, @@ -422,18 +448,19 @@ def training_log( ) if args.pipeline_model_parallel_size > 1: - # opt_stats = opt_stats.clone().detach() - opt_stats = get_accelerator().FloatTensor(opt_stats) + # opt_stats = get_accelerator().FloatTensor(opt_stats) + opt_stats = torch.tensor(opt_stats).to(DEVICE) torch.distributed.all_reduce( opt_stats, group=mpu.get_pipeline_model_parallel_group() ) - # opt_stats_2 = opt_stats_2.clone().detach() - opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) + opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_pipeline_model_parallel_group(), ) + wandb_metrics |= { "optimizer/learning_rate": learning_rate, "optimizer/iteration": args.iteration, @@ -452,7 +479,8 @@ def training_log( "optimizer/weight_abs_max": opt_stats_2[3], } # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) - if writer and is_last_rank(): + # if writer and is_last_rank(): + if writer is not None and RANK == 0: writer.add_scalar( "optimizer/variance_l2 vs tokens", opt_stats[0] ** 0.5, @@ -638,7 +666,7 @@ def training_log( ) if avg > 0.0: log_string += " {}={:.6f} |".format(key, avg) - total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) + total_loss_dict[key] = torch.tensor([0.0]).to(DEVICE) if loss_scale is not None: log_string += " loss_scale={:.1f} |".format(loss_scale) wandb_metrics |= {"loss/loss_scale": loss_scale} From b3e0f6f17bba765b9137204b6aac042f94d79fd6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 12 Sep 2024 21:46:42 -0500 Subject: [PATCH 321/590] Update `megatron/training.py` --- megatron/training.py | 1106 +++--------------------------------------- 1 file changed, 75 insertions(+), 1031 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index ad5b1b908c1..98413860492 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1,79 +1,77 @@ # Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain utilities.""" +import time + +# The earliest we can measure the start time. +_TRAIN_START_TIME = time.time() +from collections import OrderedDict from datetime import datetime +import json +import logging import math -import sys import os +import sys import time -import json -# noqa: E402 -# The earliest we can measure the start time. -_TRAIN_START_TIME = time.time() +import deepspeed +from deepspeed.accelerator import get_accelerator +from deepspeed.compression.compress import init_compression, redundancy_clean +from deepspeed.runtime.data_pipeline.data_routing.helper import ( + convert_to_random_ltd, +) +import ezpz as ez import torch import torch.distributed as tdist -from collections import OrderedDict from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP -from enum import Enum - -from megatron import get_args -from megatron import get_signal_handler -from megatron import get_timers -from megatron import get_tensorboard_writer -from megatron import get_wandb_writer -from megatron import get_current_global_batch_size -from megatron import get_num_microbatches -from megatron import is_last_rank -from megatron import update_num_microbatches -from megatron.core import mpu, tensor_parallel -# from megatron import print_rank_0, is_rank_0 -# from megatron import print_rank_last -from megatron.checkpointing import load_checkpoint -from megatron.checkpointing import save_checkpoint -from megatron.model import Float16Module -from megatron.model import GPTModel +from megatron import ( + get_args, + get_current_global_batch_size, + get_num_microbatches, + get_signal_handler, + get_tensorboard_writer, + get_timers, + is_last_rank, + update_num_microbatches, +) +from megatron.arguments import core_transformer_config_from_args +from megatron.checkpointing import load_checkpoint, save_checkpoint +from megatron.core import mpu, tensor_parallel from megatron.core.enums import ModelType +from megatron.core.pipeline_parallel import get_forward_backward_func +from megatron.data.data_samplers import build_pretraining_data_loader +from megatron.initialize import ( + initialize_megatron, + set_jit_fusion_options, + write_args_to_tensorboard, +) +from megatron.model import Float16Module, GPTModel +from megatron.model import DistributedDataParallel as LocalDDP +from megatron.model.transformer import ParallelTransformerLayer +from megatron.model.vision.knn_monitor import compute_feature_bank from megatron.optimizer import get_megatron_optimizer -from megatron.initialize import initialize_megatron -from megatron.initialize import write_args_to_tensorboard -from megatron.initialize import set_jit_fusion_options from megatron.optimizer_param_scheduler import OptimizerParamScheduler -from megatron.model import DistributedDataParallel as LocalDDP -from megatron.utils import check_adlr_autoresume_termination -from megatron.utils import unwrap_model, found_kill_switch -from megatron.data.data_samplers import build_pretraining_data_loader -from megatron.utils import calc_params_l2_norm -from megatron.core.pipeline_parallel import get_forward_backward_func +from megatron.training_log import training_log from megatron.utils import ( - report_memory, - throughput_calculator, + PerfTrace, + Profile, + calc_params_l2_norm, + check_adlr_autoresume_termination, checkpoint_throughput_calculator, + found_kill_switch, + # num_floating_point_operations, + # report_memory, + # throughput_calculator, + unwrap_model, update_rotary_pos_emb, ) -from megatron.model.vision.knn_monitor import compute_feature_bank -from megatron.arguments import core_transformer_config_from_args -from megatron.profiler import setup_profiler, trigger, on_step_begin, on_step_end -from megatron.training_log import training_log as training_log_microsoft -from megatron.training_log_alcf import training_log as training_log - -from megatron.utils import PerfTrace, Profile - -try: - import wandb -except (ImportError, ModuleNotFoundError): - wandb = None - -import deepspeed -from deepspeed.accelerator import get_accelerator -from deepspeed.compression.compress import init_compression, redundancy_clean -from deepspeed.runtime.data_pipeline.data_routing.helper import convert_to_random_ltd -from megatron.model.transformer import ParallelTransformerLayer -import ezpz as ez -import logging +# noqa: E402 +# The earliest we can measure the start time. +_TRAIN_START_TIME = time.time() +# noqa dlp = Profile("TRAINING") @@ -89,6 +87,11 @@ LOG_LEVEL: str = str(os.environ.get("LOG_LEVEL", "INFO")).upper() log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") +# try: +# import wandb +# except (ImportError, ModuleNotFoundError): +# wandb = None + def print_datetime(string): """Note that this call will sync across all ranks.""" @@ -97,37 +100,6 @@ def print_datetime(string): log.info("[" + string + "] datetime={} ".format(time_str)) -def num_floating_point_operations(args, batch_size): - # Group Query Attention. - # if not args.group_query_attention: - if not args.num_key_value_heads: - args.num_key_value_heads = args.num_attention_heads - # args.num_query_groups = args.num_attention_heads - # MoE. - # num_experts_routed_to = 1 if args.num_experts is None else args.moe_router_topk - num_experts_routed_to = 1 if args.num_experts is None else args.topk - gated_linear_multiplier = 3 / 2 if args.swiglu else 1 - return ( - 12 - * batch_size - * args.seq_length - * args.num_layers - * args.hidden_size - * args.hidden_size - * ( - 1 - + ( - (args.ffn_hidden_size / args.hidden_size) - * num_experts_routed_to - * gated_linear_multiplier - ) - + (args.num_key_value_heads / args.num_attention_heads) - + (args.seq_length / args.hidden_size) - + (args.padded_vocab_size / (2 * args.num_layers * args.hidden_size)) - ) - ) - - """ Since v0.9.0, deepspeed.initialize() has forbidden simultaneous setting of args.deepspeed_config (Path) and ds_config dict. So, we use ds_config dict which is the more flexible option @@ -198,9 +170,8 @@ def pretrain( args_defaults=args_defaults, external_args=external_args, ) - args = get_args() - + assert args is not None if found_kill_switch(): print_datetime(f"Detected kill switch at {args.kill_switch_file}. Exiting") sys.exit() @@ -231,8 +202,6 @@ def pretrain( ) ) print_datetime("after megatron is initialized") - args = get_args() - assert args is not None if os.getenv("DLIO_PROFILER_DATASET_DIR") is not None: extra_trace_path = os.environ["DLIO_PROFILER_DATASET_DIR"] else: @@ -660,7 +629,7 @@ def load_model_weights_only(model_provider_func): model = [model] print_datetime("before load checkpoint") if args.load is not None: - iteration = load_checkpoint( + _ = load_checkpoint( model, optimizer, lr_scheduler, strict=True, load_only_weights=True ) print_datetime("after load checkpoint weights") @@ -736,7 +705,7 @@ def setup_model_and_optimizer( # Only need to build dataset on tp rank 0 since Megatron has the # broadcast_data() function that broadcast data from tp rank 0. if mpu.get_tensor_model_parallel_rank() == 0: - log.info(f"Caught 'mpu.get_tensor_model_parallel_rank() == 0'") + log.info("Caught 'mpu.get_tensor_model_parallel_rank() == 0'") # Number of train/valid/test samples. if args.train_samples: train_samples = args.train_samples @@ -911,7 +880,7 @@ def train_step( # Empty unused memory. if args.empty_unused_memory_level >= 1: - get_accelerator().empty_cache() + torch.cuda.empty_cache() # Reduce gradients. if not args.deepspeed: @@ -948,7 +917,7 @@ def train_step( # Update learning rate. if args.deepspeed: - skipped_iter = 0 if update_successful else 1 + skipped_iter = 0 grad_norm = model[0].get_global_grad_norm() num_zeros_in_grad = None loss_reduced = {} @@ -970,7 +939,7 @@ def train_step( # Empty unused memory. if args.empty_unused_memory_level >= 2: - get_accelerator().empty_cache() + torch.cuda.empty_cache() if mpu.is_pipeline_last_stage(ignore_virtual=True): # Average loss across microbatches. @@ -984,854 +953,6 @@ def train_step( return {}, skipped_iter, grad_norm, num_zeros_in_grad -# <<<<<<< HEAD -# @dlp.log -# def training_log( -# loss_dict, -# total_loss_dict, -# learning_rate, -# iteration, -# loss_scale, -# report_memory_flag, -# skipped_iter, -# grad_norm, -# params_norm, -# num_zeros_in_grad, -# model=None, -# optimizer=None, -# ): -# """Log training information such as losses, timing, ....""" -# args = get_args() -# timers = get_timers() -# writer = get_tensorboard_writer() -# assert args is not None and timers is not None -# wandb_metrics = {} -# ======= -# class InteropLoggingTool(Enum): -# TENSORBOARD = 1 -# WANDB = 2 -# -# -# class interop_tool_logger: -# def __init__(self, tb_writer=None, wandb_writer=None): -# self.tb_writer = tb_writer -# self.wandb_writer = wandb_writer -# self.custom_x_axis = [] -# self.custom_y_axis = {} -# self.args = get_args() -# if not hasattr(self.args, "logger_iteration"): -# self.args.logger_iteration = 1 -# -# def is_enabled(self): -# return self.tb_writer or self.wandb_writer -# -# def add_scalar(self, key, scalar_value, step, custom_step_name=None, \ -# tool_list=[InteropLoggingTool.TENSORBOARD, InteropLoggingTool.WANDB]): -# if self.tb_writer and \ -# InteropLoggingTool.TENSORBOARD in tool_list: -# self.tb_writer.add_scalar(key, scalar_value, step) -# -# if self.wandb_writer and \ -# InteropLoggingTool.WANDB in tool_list: -# if not custom_step_name: -# self.wandb_writer.log({key: scalar_value}, step=step) -# if self.args.logger_iteration < step: -# # Updating iteration -# self.args.logger_iteration = step -# -# else: -# if custom_step_name not in self.custom_x_axis: -# self.custom_x_axis.append(custom_step_name) -# wandb.define_metric(custom_step_name) -# -# if key not in self.custom_y_axis: -# self.custom_y_axis[key] = custom_step_name -# wandb.define_metric(key, step_metric=custom_step_name) -# -# self.wandb_writer.log({key: scalar_value, custom_step_name: step}, \ -# step=self.args.logger_iteration) -# -# -# def add_scalar_to_tb(self, key, scalar_value, step): -# return self.add_scalar(key, scalar_value, step, None, [InteropLoggingTool.TENSORBOARD]) -# -# def add_scalar_to_wandb(self, key, scalar_value, step, custom_step_name=None): -# return self.add_scalar(key, scalar_value, step, custom_step_name, [InteropLoggingTool.WANDB]) -# -# def add_images(self, key, img_tensor, step=None): -# if self.tb_writer: -# self.tb_writer.add_images(key, img_tensor, step) -# -# if self.wandb_writer: -# self.wandb_writer.log({key: wandb.Image(img_tensor)}, step) - - -# def training_log(loss_dict, total_loss_dict, learning_rate, iteration, -# loss_scale, report_memory_flag, skipped_iter, -# grad_norm, params_norm, num_zeros_in_grad, -# model=None, optimizer=None): -# """Log training information such as losses, timing, ....""" -# args = get_args() -# timers = get_timers() -# writer = interop_tool_logger(tb_writer=get_tensorboard_writer(), \ -# wandb_writer=get_wandb_writer()) -# x_axis_samples = 'Samples' -# x_axis_tokens = 'Tokens' -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 -# # Advanced, skipped, and Nan iterations. -# advanced_iters_key = "advanced iterations" -# skipped_iters_key = "skipped iterations" -# nan_iters_key = "nan iterations" -# # Advanced iterations. -# if not skipped_iter: -# total_loss_dict[advanced_iters_key] = ( -# total_loss_dict.get(advanced_iters_key, 0) + 1 -# ) -# else: -# if advanced_iters_key not in total_loss_dict: -# total_loss_dict[advanced_iters_key] = 0 -# # Skipped iterations. -# total_loss_dict[skipped_iters_key] = ( -# total_loss_dict.get(skipped_iters_key, 0) + skipped_iter -# ) -# # Update losses and set nan iterations -# got_nan = False -# for key in loss_dict: -# if not skipped_iter: -# total_loss_dict[key] = ( -# total_loss_dict.get(key, get_accelerator().FloatTensor([0.0])) -# + loss_dict[key] -# ) -# else: -# value = loss_dict[key].float().sum().item() -# is_nan = value == float("inf") or value == -float("inf") or value != value -# got_nan = got_nan or is_nan -# total_loss_dict[nan_iters_key] = total_loss_dict.get(nan_iters_key, 0) + int( -# got_nan -# ) -# -# # Logging. -# timers_to_log = [ -# "forward-backward", -# "forward-compute", -# "backward-compute", -# "batch-generator", -# "forward-recv", -# "forward-send", -# "backward-recv", -# "backward-send", -# "forward-send-forward-recv", -# "forward-send-backward-recv", -# "backward-send-forward-recv", -# "backward-send-backward-recv", -# "forward-backward-send-forward-backward-recv", -# "layernorm-grads-all-reduce", -# "embedding-grads-all-reduce", -# "grads-all-reduce", -# "grads-reduce-scatter", -# "params-all-gather", -# "optimizer-copy-to-main-grad", -# "optimizer-unscale-and-check-inf", -# "optimizer-clip-main-grad", -# "optimizer-count-zeros", -# "optimizer-inner-step", -# "optimizer-copy-main-to-model-params", -# "optimizer", -# ] -# -# # Calculate batch size. -# batch_size = ( -# args.micro_batch_size * args.data_parallel_size * get_num_microbatches() -# ) -# total_iterations = ( -# total_loss_dict[advanced_iters_key] + total_loss_dict[skipped_iters_key] -# ) -# -# # Tensorboard values. -# # Timer requires all the ranks to call. -# <<<<<<< HEAD -# if args.log_timers_to_tensorboard and ( -# iteration % args.tensorboard_log_interval == 0 -# ): -# timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) -# if writer and (iteration % args.tensorboard_log_interval == 0): -# writer.add_scalar( -# "steps-vs-samples/y=steps,x=samples", iteration, args.consumed_train_samples -# ) -# writer.add_scalar( -# "steps-vs-samples/y=samples,x=steps", args.consumed_train_samples, iteration -# ) -# writer.add_scalar( -# "steps-vs-tokens/y=steps,x=tokens", iteration, args.consumed_train_tokens -# ) -# writer.add_scalar( -# "steps-vs-tokens/y=tokens,x=steps", args.consumed_train_tokens, iteration -# ) -# if args.log_learning_rate_to_tensorboard: -# wandb_metrics |= { -# "learning-rate/iteration": iteration, -# "learning-rate/learning-rate": learning_rate, -# } -# writer.add_scalar("learning-rate/learning-rate", learning_rate, iteration) -# writer.add_scalar( -# "learning-rate/learning-rate vs samples", -# learning_rate, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "learning-rate/learning-rate vs tokens", -# learning_rate, -# args.consumed_train_tokens, -# ) -# if args.log_batch_size_to_tensorboard: -# writer.add_scalar("batch-size/batch-size", batch_size, iteration) -# writer.add_scalar( -# "batch-size/batch-size vs samples", -# batch_size, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "batch-size/batch-size vs tokens", -# batch_size, -# args.consumed_train_tokens, -# ) -# wandb_metrics |= { -# "lm-loss-training/iteration": iteration, -# "lm-loss-training/consumed_train_tokens": args.consumed_train_tokens, -# } -# ======= -# if args.log_timers_to_tensorboard and \ -# (iteration % args.tensorboard_log_interval == 0): -# timers.write(timers_to_log, writer, iteration, -# normalizer=total_iterations) -# if writer.is_enabled() and (iteration % args.tensorboard_log_interval == 0): -# writer.add_scalar('steps-vs-samples/y=steps,x=samples', iteration, args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('steps-vs-samples/y=samples,x=steps', args.consumed_train_samples, iteration) -# writer.add_scalar('steps-vs-tokens/y=steps,x=tokens', iteration, args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('steps-vs-tokens/y=tokens,x=steps', args.consumed_train_tokens, iteration) -# if args.log_learning_rate_to_tensorboard: -# writer.add_scalar('learning-rate/learning-rate', learning_rate, iteration) -# writer.add_scalar('learning-rate/learning-rate vs samples', learning_rate, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('learning-rate/learning-rate vs tokens', learning_rate, -# args.consumed_train_tokens, x_axis_tokens) -# if args.log_batch_size_to_tensorboard: -# writer.add_scalar('batch-size/batch-size', batch_size, iteration) -# writer.add_scalar('batch-size/batch-size vs samples', batch_size, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('batch-size/batch-size vs tokens', batch_size, -# args.consumed_train_tokens, x_axis_tokens) -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 -# for key in loss_dict: -# wandb_metrics |= {f"lm-loss-training/{key}": loss_dict[key]} -# writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) -# <<<<<<< HEAD -# writer.add_scalar( -# f"lm-loss-training/{key}" + " vs samples", -# loss_dict[key], -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# f"lm-loss-training/{key}" + " vs tokens", -# loss_dict[key], -# args.consumed_train_tokens, -# ) -# if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: -# writer.add_scalar("loss-scale/loss-scale", loss_scale, iteration) -# writer.add_scalar( -# "loss-scale/loss-scale vs samples", -# loss_scale, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "loss-scale/loss-scale vs tokens", -# loss_scale, -# args.consumed_train_tokens, -# ) -# if args.log_world_size_to_tensorboard: -# writer.add_scalar("world-size/world-size", args.world_size, iteration) -# writer.add_scalar( -# "world-size/world-size vs samples", -# args.world_size, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "world-size/world-size vs tokens", -# args.world_size, -# args.consumed_train_tokens, -# ) -# if grad_norm is not None: -# wandb_metrics |= {"training/grad-norm": grad_norm} -# writer.add_scalar("grad-norm/grad-norm", grad_norm, iteration) -# writer.add_scalar( -# "grad-norm/grad-norm vs samples", grad_norm, args.consumed_train_samples -# ) -# writer.add_scalar( -# "grad-norm/grad-norm vs tokens", grad_norm, args.consumed_train_tokens -# ) -# if num_zeros_in_grad is not None: -# wandb_metrics |= {"training/num-zeros": num_zeros_in_grad} -# writer.add_scalar("num-zeros/num-zeros", num_zeros_in_grad, iteration) -# writer.add_scalar( -# "num-zeros/num-zeros vs samples", -# num_zeros_in_grad, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "num-zeros/num-zeros vs tokens", -# num_zeros_in_grad, -# args.consumed_train_tokens, -# ) -# if params_norm is not None: -# wandb_metrics |= {"training/params-norm": params_norm} -# writer.add_scalar("params-norm/params-norm", params_norm, iteration) -# writer.add_scalar( -# "params-norm/params-norm vs samples", -# params_norm, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "params-norm/params-norm vs tokens", -# params_norm, -# args.consumed_train_tokens, -# ) -# if hasattr(args, "actual_seq_length"): -# writer.add_scalar( -# "seqlen/actual_seq_length", args.actual_seq_length, iteration -# ) -# writer.add_scalar( -# "seqlen/actual_seq_length vs samples", -# args.actual_seq_length, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "seqlen/actual_seq_length vs tokens", -# args.actual_seq_length, -# args.consumed_train_tokens, -# ) -# if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: -# writer.add_scalar( -# "seqlen/curriculum_seqlen", args.curriculum_seqlen, iteration -# ) -# writer.add_scalar( -# "seqlen/curriculum_seqlen vs samples", -# args.curriculum_seqlen, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "seqlen/curriculum_seqlen vs tokens", -# args.curriculum_seqlen, -# args.consumed_train_tokens, -# ) -# if args.random_ltd: -# writer.add_scalar( -# "seqlen/random_ltd_reserved_length", -# args.random_ltd_reserved_length, -# iteration, -# ) -# writer.add_scalar( -# "seqlen/random_ltd_reserved_length vs samples", -# args.random_ltd_reserved_length, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "seqlen/random_ltd_reserved_length vs tokens", -# args.random_ltd_reserved_length, -# args.consumed_train_tokens, -# ) -# ======= -# writer.add_scalar(f"lm-loss-training/{key}" + ' vs samples', loss_dict[key], -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar(f"lm-loss-training/{key}" + ' vs tokens', loss_dict[key], -# args.consumed_train_tokens, x_axis_tokens) -# if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: -# writer.add_scalar('loss-scale/loss-scale', loss_scale, iteration) -# writer.add_scalar('loss-scale/loss-scale vs samples', loss_scale, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('loss-scale/loss-scale vs tokens', loss_scale, -# args.consumed_train_tokens, x_axis_tokens) -# if args.log_world_size_to_tensorboard: -# writer.add_scalar('world-size/world-size', args.world_size, iteration) -# writer.add_scalar('world-size/world-size vs samples', args.world_size, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('world-size/world-size vs tokens', args.world_size, -# args.consumed_train_tokens, x_axis_tokens) -# if grad_norm is not None: -# writer.add_scalar('grad-norm/grad-norm', grad_norm, iteration) -# writer.add_scalar('grad-norm/grad-norm vs samples', grad_norm, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('grad-norm/grad-norm vs tokens', grad_norm, -# args.consumed_train_tokens, x_axis_tokens) -# if num_zeros_in_grad is not None: -# writer.add_scalar('num-zeros/num-zeros', num_zeros_in_grad, iteration) -# writer.add_scalar('num-zeros/num-zeros vs samples', num_zeros_in_grad, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('num-zeros/num-zeros vs tokens', num_zeros_in_grad, -# args.consumed_train_tokens, x_axis_tokens) -# if params_norm is not None: -# writer.add_scalar('params-norm/params-norm', params_norm, iteration) -# writer.add_scalar('params-norm/params-norm vs samples', params_norm, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('params-norm/params-norm vs tokens', params_norm, -# args.consumed_train_tokens, x_axis_tokens) -# if hasattr(args, 'actual_seq_length'): -# writer.add_scalar('seqlen/actual_seq_length', args.actual_seq_length, -# iteration) -# writer.add_scalar('seqlen/actual_seq_length vs samples', args.actual_seq_length, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('seqlen/actual_seq_length vs tokens', args.actual_seq_length, -# args.consumed_train_tokens, x_axis_tokens) -# if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: -# writer.add_scalar('seqlen/curriculum_seqlen', args.curriculum_seqlen, -# iteration) -# writer.add_scalar('seqlen/curriculum_seqlen vs samples', args.curriculum_seqlen, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('seqlen/curriculum_seqlen vs tokens', args.curriculum_seqlen, -# args.consumed_train_tokens, x_axis_tokens) -# if args.random_ltd: -# writer.add_scalar('seqlen/random_ltd_reserved_length', args.random_ltd_reserved_length, -# iteration) -# writer.add_scalar('seqlen/random_ltd_reserved_length vs samples', args.random_ltd_reserved_length, -# args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('seqlen/random_ltd_reserved_length vs tokens', args.random_ltd_reserved_length, -# args.consumed_train_tokens, x_axis_tokens) -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 -# if args.log_memory_to_tensorboard: -# mem_stats = torch.cuda.memory_stats() -# writer.add_scalar( -# "mem-reserved-bytes", -# mem_stats["reserved_bytes.all.current"], -# iteration, -# ) -# writer.add_scalar( -# "mem-allocated-bytes", -# mem_stats["allocated_bytes.all.current"], -# iteration, -# ) -# writer.add_scalar( -# "mem-allocated-count", -# mem_stats["allocation.all.current"], -# iteration, -# ) -# if iteration % args.tensorboard_log_interval == 0: -# # This logging write various optimizer states to tensorboard. This -# # feature may consume extra GPU memory thus is set at false by default. -# if args.log_optimizer_states_to_tensorboard and optimizer is not None: -# opt_stats = [0.0] * 8 -# opt_stats_2 = [0.0] * 4 -# -# #TODO(billishyahao): Remove me after bf16_optimizer promotes its state. -# if not hasattr(optimizer, "state"): -# assert hasattr(optimizer, "optimizer"), f"Optimizer must have optimizer property." -# optimizer.state = optimizer.optimizer.state -# -# for _, group in enumerate(optimizer.param_groups): -# for _, param in enumerate(group["params"]): -# state_param = getattr(optimizer, "state", None) -# if state_param is not None: -# exp_avg_sq = state_param.get("exp_avg_sq", torch.tensor(0.0)) -# exp_avg = state_param.get("exp_avg", torch.tensor(0.0)) -# opt_stats[0] += (torch.norm(exp_avg_sq).item()) ** 2 -# opt_stats[1] += (torch.norm(exp_avg_sq.sqrt()).item()) ** 2 -# opt_stats[2] += (torch.norm(exp_avg).item()) ** 2 -# opt_stats[3] += (torch.norm(param).item()) ** 2 -# opt_stats[4] += torch.norm(exp_avg_sq, p=1).item() -# opt_stats[5] += torch.norm(exp_avg_sq.sqrt(), p=1).item() -# opt_stats[6] += torch.norm(exp_avg, p=1).item() -# opt_stats[7] += torch.norm(param, p=1).item() -# opt_stats_2[0] = max( -# opt_stats_2[0], -# abs(exp_avg_sq.max().item()), -# abs(exp_avg_sq.min().item()), -# ) -# opt_stats_2[1] = max( -# opt_stats_2[1], exp_avg_sq.sqrt().abs_().max().item() -# ) -# opt_stats_2[2] = max( -# opt_stats_2[2], -# abs(exp_avg.max().item()), -# abs(exp_avg.min().item()), -# ) -# opt_stats_2[3] = max( -# opt_stats_2[3], -# abs(param.max().item()), -# abs(param.min().item()), -# ) -# # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) -# if args.zero_stage > 0: -# # ZeRO partiions optimizer states -# # opt_stats = opt_stats.clone().detach() -# # opt_stats = get_accelerator().FloatTensor -# opt_stats = get_accelerator().FloatTensor(opt_stats) -# torch.distributed.all_reduce( -# opt_stats, group=mpu.get_sequence_data_parallel_group() -# ) -# # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) -# # opt_stats_2 = opt_stats_2.clone().detach() -# opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) -# torch.distributed.all_reduce( -# opt_stats_2, -# op=torch.distributed.ReduceOp.MAX, -# group=mpu.get_sequence_data_parallel_group(), -# ) -# -# if args.tensor_model_parallel_size > 1: -# # opt_stats = opt_stats.clone().detach() -# opt_stats = get_accelerator().FloatTensor(opt_stats) -# torch.distributed.all_reduce( -# opt_stats, group=mpu.get_tensor_model_parallel_group() -# ) -# # opt_stats_2 = opt_stats_2.clone().detach() -# opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) -# torch.distributed.all_reduce( -# opt_stats_2, -# op=torch.distributed.ReduceOp.MAX, -# group=mpu.get_tensor_model_parallel_group(), -# ) -# -# if args.pipeline_model_parallel_size > 1: -# # opt_stats = opt_stats.clone().detach() -# opt_stats = get_accelerator().FloatTensor(opt_stats) -# torch.distributed.all_reduce( -# opt_stats, group=mpu.get_pipeline_model_parallel_group() -# ) -# # opt_stats_2 = opt_stats_2.clone().detach() -# opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) -# torch.distributed.all_reduce( -# opt_stats_2, -# op=torch.distributed.ReduceOp.MAX, -# group=mpu.get_pipeline_model_parallel_group(), -# ) -# wandb_metrics |= { -# "optimizer/learning_rate": learning_rate, -# "optimizer/iteration": args.iteration, -# "optimizer/consumed_train_tokens": args.consumed_train_tokens, -# "optimizer/variance_l2": opt_stats[0] ** 0.5, -# "optimizer/variance_sqrt_l2": opt_stats[1] ** 0.5, -# "optimizer/momentum_l2": opt_stats[2] ** 0.5, -# "optimizer/weight_l2": opt_stats[3] ** 0.5, -# "optimizer/variance_l1": opt_stats[4], -# "optimizer/variance_sqrt_l1": opt_stats[5], -# "optimizer/momentum_l1": opt_stats[6], -# "optimizer/weight_l1": opt_stats[7], -# "optimizer/variance_abs_max": opt_stats_2[0], -# "optimizer/variance_sqrt_abs_max": opt_stats_2[1], -# "optimizer/momentum_abs_max": opt_stats_2[2], -# "optimizer/weight_abs_max": opt_stats_2[3], -# } -# # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) -# <<<<<<< HEAD -# if writer and is_last_rank(): -# writer.add_scalar( -# "optimizer/variance_l2 vs tokens", -# opt_stats[0] ** 0.5, -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_sqrt_l2 vs tokens", -# opt_stats[1] ** 0.5, -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/momentum_l2 vs tokens", -# opt_stats[2] ** 0.5, -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/weight_l2 vs tokens", -# opt_stats[3] ** 0.5, -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_l1 vs tokens", -# opt_stats[4], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_sqrt_l1 vs tokens", -# opt_stats[5], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/momentum_l1 vs tokens", -# opt_stats[6], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/weight_l1 vs tokens", -# opt_stats[7], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_abs_max vs tokens", -# opt_stats_2[0], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_sqrt_abs_max vs tokens", -# opt_stats_2[1], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/momentum_abs_max vs tokens", -# opt_stats_2[2], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/weight_abs_max vs tokens", -# opt_stats_2[3], -# args.consumed_train_tokens, -# ) -# writer.add_scalar( -# "optimizer/variance_l2", opt_stats[0] ** 0.5, iteration -# ) -# writer.add_scalar( -# "optimizer/variance_sqrt_l2", opt_stats[1] ** 0.5, iteration -# ) -# writer.add_scalar( -# "optimizer/momentum_l2", opt_stats[2] ** 0.5, iteration -# ) -# writer.add_scalar("optimizer/weight_l2", opt_stats[3] ** 0.5, iteration) -# writer.add_scalar("optimizer/variance_l1", opt_stats[4], iteration) -# writer.add_scalar("optimizer/variance_sqrt_l1", opt_stats[5], iteration) -# writer.add_scalar("optimizer/momentum_l1", opt_stats[6], iteration) -# writer.add_scalar("optimizer/weight_l1", opt_stats[7], iteration) -# writer.add_scalar( -# "optimizer/variance_abs_max", opt_stats_2[0], iteration -# ) -# writer.add_scalar( -# "optimizer/variance_sqrt_abs_max", opt_stats_2[1], iteration -# ) -# writer.add_scalar( -# "optimizer/momentum_abs_max", opt_stats_2[2], iteration -# ) -# writer.add_scalar("optimizer/weight_abs_max", opt_stats_2[3], iteration) -# ======= -# if writer.is_enabled() and is_last_rank(): -# writer.add_scalar('optimizer/variance_l2 vs tokens', opt_stats[0]**0.5, args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/variance_sqrt_l2 vs tokens', opt_stats[1]**0.5, args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/momentum_l2 vs tokens', opt_stats[2]**0.5, args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/weight_l2 vs tokens', opt_stats[3]**0.5, args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/variance_l1 vs tokens', opt_stats[4], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/variance_sqrt_l1 vs tokens', opt_stats[5], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/momentum_l1 vs tokens', opt_stats[6], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/weight_l1 vs tokens', opt_stats[7], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/variance_abs_max vs tokens', opt_stats_2[0], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/variance_sqrt_abs_max vs tokens', opt_stats_2[1], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/momentum_abs_max vs tokens', opt_stats_2[2], args.consumed_train_tokens, x_axis_tokens) -# writer.add_scalar('optimizer/weight_abs_max vs tokens', opt_stats_2[3], args.consumed_train_tokens, x_axis_tokens) -# -# writer.add_scalar('optimizer/variance_l2', opt_stats[0]**0.5, iteration) -# writer.add_scalar('optimizer/variance_sqrt_l2', opt_stats[1]**0.5, iteration) -# writer.add_scalar('optimizer/momentum_l2', opt_stats[2]**0.5, iteration) -# writer.add_scalar('optimizer/weight_l2', opt_stats[3]**0.5, iteration) -# writer.add_scalar('optimizer/variance_l1', opt_stats[4], iteration) -# writer.add_scalar('optimizer/variance_sqrt_l1', opt_stats[5], iteration) -# writer.add_scalar('optimizer/momentum_l1', opt_stats[6], iteration) -# writer.add_scalar('optimizer/weight_l1', opt_stats[7], iteration) -# writer.add_scalar('optimizer/variance_abs_max', opt_stats_2[0], iteration) -# writer.add_scalar('optimizer/variance_sqrt_abs_max', opt_stats_2[1], iteration) -# writer.add_scalar('optimizer/momentum_abs_max', opt_stats_2[2], iteration) -# writer.add_scalar('optimizer/weight_abs_max', opt_stats_2[3], iteration) -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 -# -# assert args is not None -# assert timers is not None -# if iteration % args.log_interval == 0: -# elapsed_time = timers("interval-time").elapsed(barrier=True) -# elapsed_time_per_iteration = elapsed_time / total_iterations -# seq_len = args.seq_length -# if hasattr(args, "actual_seq_length"): -# seq_len = args.actual_seq_length -# samples_per_sec, tflops, approx_parameters_in_billions = throughput_calculator( -# model, args, elapsed_time, total_iterations -# ) -# samples_per_sec_per_replica = samples_per_sec / args.data_parallel_size -# tokens_per_sec = samples_per_sec * seq_len -# tokens_per_sec_per_replica = tokens_per_sec / args.data_parallel_size -# tokens_per_gpu_per_second = tokens_per_sec / args.world_size -# <<<<<<< HEAD -# tokens_per_gpu_per_second_per_replica = ( -# tokens_per_gpu_per_second / args.data_parallel_size -# ) -# # NOTE: [2024-06-19] -# # Updated to use (more accurate) calculation according to -# # `num_floating_point_operations` from NVIDIA/Megatron-LM -# num_flop_lm = num_floating_point_operations(args, batch_size) -# num_flop_per_sec_lm = (num_flop_lm / elapsed_time_per_iteration) -# tflops_lm = (num_flop_per_sec_lm / (10 ** 12)) -# tflops_lm_per_gpu = (tflops_lm / args.world_size) -# wandb_metrics |= { -# "throughput/iteration-time": elapsed_time_per_iteration, # 1000 ms / s -# "throughput/samples_per_sec": samples_per_sec, -# "throughput/samples_per_sec_per_replica": samples_per_sec_per_replica, -# "throughput/tokens_per_sec": tokens_per_sec, -# "throughput/tokens_per_sec_per_replica": tokens_per_sec_per_replica, -# "throughput/tokens_per_gpu_per_sec": tokens_per_gpu_per_second, -# "throughput/tokens_per_gpu_per_sec_per_replica": tokens_per_gpu_per_second_per_replica, -# "throughput/tflops": tflops, -# "throughput/tflops-new": num_flop_lm / elapsed_time_per_iteration, -# "throughput/tflops-lm": tflops_lm_per_gpu, -# "throughput/approx_params_in_billions": approx_parameters_in_billions, -# "throughput/elapsed_ms_per_iteration": elapsed_time_per_iteration, -# "throughput/iteration": iteration, -# } -# if loss_dict is not None: -# wandb_metrics |= { -# "loss/iteration": iteration, -# **{f"loss/{k}": v for k, v in loss_dict.items()}, -# } -# if writer and args.log_timers_to_tensorboard: -# writer.add_scalar( -# "iteration-time/iteration-time", elapsed_time_per_iteration, iteration -# ) -# writer.add_scalar( -# "iteration-time/iteration-time vs samples", -# elapsed_time_per_iteration, -# args.consumed_train_samples, -# ) -# writer.add_scalar( -# "iteration-time/iteration-time vs tokens", -# elapsed_time_per_iteration, -# args.consumed_train_tokens, -# ) -# # metrics_to_log = { -# # 'iteration': iteration, -# # 'train_iters': args.train_iters, -# # 'consumed_samples': args.consumed_train_samples, -# # 'consumed_tokens': args.consumed_tokens, -# # } -# log_string = f" iteration={iteration:8d}/{args.train_iters:8d} |" -# # .format( iteration, args.train_iters) -# log_string += ( -# f" consumed_samples={args.consumed_train_samples:12d} |" -# # .format(args.consumed_train_samples) -# ) -# log_string += f" consumed_tokens={args.consumed_train_tokens:12d} |" -# # .format( args.consumed_train_tokens) -# log_string += ( -# " elapsed_time_per_iteration_ms=" -# f"{elapsed_time_per_iteration * 1000.0:.1f} |" -# # .format( elapsed_time_per_iteration * 1000.0) -# ) -# log_string += f" learning_rate={learning_rate:.6g} |" -# log_string += f" global_batch_size={batch_size:5d} |" -# # if wandb is not None and getattr(wandb, 'run', None) is not None: -# wandb_metrics |= { -# "training/iteration": iteration, -# "training/iteration_time": elapsed_time_per_iteration, -# "training/iteration_time_vs_tokens": ( -# elapsed_time_per_iteration / args.consumed_train_tokens -# ), -# "training/iteration_time_vs_samples": ( -# (elapsed_time_per_iteration / args.consumed_train_samples), -# ), -# "training/consumed_samples": args.consumed_train_samples, -# "training/consumed_tokens": args.consumed_train_tokens, -# } -# ======= -# tokens_per_gpu_per_second_per_replica = tokens_per_gpu_per_second / args.data_parallel_size -# -# if writer.is_enabled(): -# writer.add_scalar_to_wandb('throughput/iteration-time', elapsed_time_per_iteration, iteration) # 1000 ms / s -# writer.add_scalar_to_wandb('throughput/samples_per_sec', samples_per_sec, iteration) -# writer.add_scalar_to_wandb('throughput/samples_per_sec_per_replica', samples_per_sec_per_replica, iteration) -# writer.add_scalar_to_wandb('throughput/tokens_per_sec', tokens_per_sec, iteration) -# writer.add_scalar_to_wandb('throughput/tokens_per_sec_per_replica', tokens_per_sec_per_replica, iteration) -# writer.add_scalar_to_wandb('throughput/tokens_per_gpu_per_sec', tokens_per_gpu_per_second, iteration) -# writer.add_scalar_to_wandb('throughput/tokens_per_gpu_per_sec_per_replica', tokens_per_gpu_per_second_per_replica, iteration) -# writer.add_scalar_to_wandb('throughput/tflops', tflops, iteration) -# writer.add_scalar_to_wandb('throughput/approx_params_in_billions', approx_parameters_in_billions, iteration) -# writer.add_scalar_to_wandb('throughput/elapsed_ms_per_iteration', elapsed_time_per_iteration, iteration) -# if loss_dict is not None: -# for k, v in loss_dict.items(): -# writer.add_scalar_to_wandb(f'loss/{k}', v, iteration) -# -# if args.log_timers_to_tensorboard: -# writer.add_scalar('iteration-time/iteration-time', -# elapsed_time_per_iteration, iteration) -# writer.add_scalar('iteration-time/iteration-time vs samples', -# elapsed_time_per_iteration, args.consumed_train_samples, x_axis_samples) -# writer.add_scalar('iteration-time/iteration-time vs tokens', -# elapsed_time_per_iteration, args.consumed_train_tokens, x_axis_tokens) -# log_string = ' iteration {:8d}/{:8d} |'.format( -# iteration, args.train_iters) -# log_string += ' consumed samples: {:12d} |'.format( -# args.consumed_train_samples) -# log_string += ' consumed tokens: {:12d} |'.format( -# args.consumed_train_tokens) -# log_string += ' elapsed time per iteration (ms): {:.1f} |'.format( -# elapsed_time_per_iteration * 1000.0) -# log_string += ' learning rate: {:.3E} |'.format(learning_rate) -# log_string += ' global batch size: {:5d} |'.format(batch_size) -# -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 -# for key in total_loss_dict: -# if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: -# avg = total_loss_dict[key].item() / float( -# max(1, total_loss_dict[advanced_iters_key]) -# ) -# if avg > 0.0: -# log_string += " {}={:.6f} |".format(key, avg) -# total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) -# if loss_scale is not None: -# log_string += " loss_scale={:.1f} |".format(loss_scale) -# wandb_metrics |= {"loss/loss_scale": loss_scale} -# if grad_norm is not None: -# log_string += " grad_norm={:.3f} |".format(grad_norm) -# wandb_metrics |= {"loss/grad_norm": grad_norm} -# if num_zeros_in_grad is not None: -# log_string += " num_zeros={:.1f} |".format(num_zeros_in_grad) -# wandb_metrics |= {"loss/num_zeros_in_grad": num_zeros_in_grad} -# if params_norm is not None: -# log_string += " params_norm={:.3f} |".format(params_norm) -# wandb_metrics |= {"loss/params_norm": params_norm} -# if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: -# log_string += " curriculum_seqlen={:5d} |".format(args.curriculum_seqlen) -# if args.random_ltd: -# log_string += " random_ltd reserved_length={:5d} |".format( -# args.random_ltd_reserved_length -# ) -# # log_string += " | ".join([ -# # f"{seq_len=:5d} ", -# # f"{}" -# # f"number_of_skipped_iterations={:3d}", -# # -# # ]) -# log_string += " actual_seqlen={:5d} |".format(seq_len) -# log_string += " number_of_skipped_iterations={:3d} |".format( -# total_loss_dict[skipped_iters_key] -# ) -# log_string += " number_of_nan_iterations={:3d} |".format( -# total_loss_dict[nan_iters_key] -# ) -# log_string += " samples_per_second={:.3f} |".format(samples_per_sec) -# log_string += " tokens_per_gpu_per_second_tgs={:.3f} |".format( -# tokens_per_gpu_per_second -# ) -# log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) -# log_string += " [DS]TFLOPs={:.2f} |".format(tflops) -# total_loss_dict[advanced_iters_key] = 0 -# total_loss_dict[skipped_iters_key] = 0 -# total_loss_dict[nan_iters_key] = 0 -# # print_rank_last(log_string) -# log.info(log_string) -# if report_memory_flag and learning_rate > 0.0: -# # Report memory after optimizer state has been initialized. -# report_memory("(after {} iterations)".format(iteration)) -# report_memory_flag = False -# if wandb is not None and getattr(wandb, "run", None) is not None: -# wandb_metrics |= { -# "training/skiped_iterations": total_loss_dict[skipped_iters_key] -# } -# wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} -# wandb.log(wandb_metrics) -# if timers is not None: -# timers.log(timers_to_log, normalizer=args.log_interval) -# -# return report_memory_flag - - @dlp.log @ez.dist.timeitlogit(rank=RANK) def save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler): @@ -1866,31 +987,24 @@ def train( assert timers is not None # Write args to tensorboard write_args_to_tensorboard() - - setup_profiler(args, get_accelerator().device_name()) if args.random_ltd: # random-ltd requires different randomness on each rank import random random.seed(args.seed + torch.distributed.get_rank()) - # Turn on training mode which enables dropout. for model_module in model: model_module.train() - # Tracking loss. total_loss_dict = {} loss_dict = {} - # Iterations. iteration = args.iteration - # Translate args to core configuration config = core_transformer_config_from_args(args) if not args.deepspeed: config.grad_scale_func = optimizer.scale_loss config.timers = timers - timers("interval-time", log_level=0).start(barrier=True) print_datetime("before the start of training step") report_memory_flag = True @@ -1915,7 +1029,6 @@ def train( while iteration < args.train_iters and ( args.train_tokens is None or args.consumed_train_tokens < args.train_tokens ): - trigger(on_step_begin) update_num_microbatches(args.consumed_train_samples) if args.deepspeed: # inform deepspeed of any batch size changes @@ -1925,7 +1038,6 @@ def train( * get_num_microbatches() ) model[0].set_train_batch_size(global_batch_size) - if args.curriculum_learning_legacy and not args.no_pipeline_parallel: curriculum_seqlen = args.curriculum_scheduler.update_difficulty( args.iteration + 1 @@ -1945,7 +1057,7 @@ def train( num_zeros_in_grad = None else: if os.getenv("TORCH_PROFILER_ENABLE") == "2": - from torch.profiler import profile, record_function, ProfilerActivity + from torch.profiler import profile, ProfilerActivity try: activities = [ @@ -2015,7 +1127,6 @@ def train( args.consumed_train_tokens += new_samples * args.actual_seq_length else: args.consumed_train_tokens += new_samples * args.actual_seq_length - # Logging. if args.deepspeed: if hasattr(model[0].optimizer, "cur_scale"): @@ -2046,7 +1157,6 @@ def train( check_adlr_autoresume_termination( iteration, model, optimizer, opt_param_scheduler ) - # # Evaluation if args.eval_interval and iteration % args.eval_interval == 0 and args.do_valid: prefix = "iteration {}".format(iteration) @@ -2060,7 +1170,6 @@ def train( config, False, ) - # Checkpointing saved_checkpoint = False if args.exit_signal_handler: @@ -2072,11 +1181,9 @@ def train( ) print_datetime("exiting program after receiving SIGTERM.") sys.exit() - if args.save and args.save_interval and iteration % args.save_interval == 0: save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler) saved_checkpoint = True - # Exiting based on duration if args.exit_duration_in_mins: train_time = (time.time() - _TRAIN_START_TIME) / 60.0 @@ -2092,7 +1199,6 @@ def train( ) print_datetime("exiting program after {} minutes".format(train_time)) sys.exit() - # Exiting based on iterations if args.exit_interval and iteration % args.exit_interval == 0: if args.save and not saved_checkpoint: @@ -2102,22 +1208,6 @@ def train( torch.distributed.barrier() print_datetime("exiting program at iteration {}".format(iteration)) sys.exit() - - trigger(on_step_end) - - # Exiting based on kill switch file - if found_kill_switch(): - if args.save and not saved_checkpoint: - save_checkpoint_and_time( - iteration, model, optimizer, opt_param_scheduler - ) - torch.distributed.barrier() - print_datetime( - f"Detected kill switch at {args.kill_switch_file}, " - f"iteration={iteration}. Exiting" - ) - sys.exit() - return iteration @@ -2184,7 +1274,7 @@ def evaluate( # Empty unused memory if args.empty_unused_memory_level >= 1: - get_accelerator().empty_cache() + torch.cuda.empty_cache() if mpu.is_pipeline_last_stage(ignore_virtual=True): # Reduce across processes. @@ -2251,17 +1341,12 @@ def evaluate_and_print_results( test=False, ): """Helper function to evaluate and dump results on screen.""" - from megatron.training_log import interop_tool_logger - args = get_args() + assert args is not None if write_to_tensorboard: - writer = interop_tool_logger( - tb_writer=get_tensorboard_writer(), wandb_writer=get_wandb_writer() - ) + writer = get_tensorboard_writer() else: - writer = interop_tool_logger() - x_axis_samples = "Samples" - x_axis_tokens = "Tokens" + writer = None total_loss_dict, collected_non_loss_data = evaluate( forward_step_func, @@ -2275,43 +1360,9 @@ def evaluate_and_print_results( for key in total_loss_dict: string += f"{key} value={total_loss_dict[key].item():.6f}" ppl = math.exp(min(20, total_loss_dict[key].item())) - # <<<<<<< HEAD - # string += f"{key} PPL={ppl:.6f}" - # # string += '{} PPL={:.6f} | '.format(key, ppl) - # if writer and is_last_rank(): - # data_type = "test" if test else "validation" - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type}", - # total_loss_dict[key].item(), - # iteration, - # ) - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type} vs samples", - # total_loss_dict[key].item(), - # args.consumed_train_samples, - # ) - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type} vs tokens", - # total_loss_dict[key].item(), - # args.consumed_train_tokens, - # ) - # if args.log_validation_ppl_to_tensorboard: - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type} ppl", ppl, iteration - # ) - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type} ppl vs samples", - # ppl, - # args.consumed_train_samples, - # ) - # writer.add_scalar( - # f"lm-loss-validation/{key} {data_type} ppl vs tokens", - # ppl, - # args.consumed_train_tokens, - # ) - # ======= - string += "{} PPL: {:.6E} | ".format(key, ppl) - if writer.is_enabled() and is_last_rank(): + string += f"{key} PPL={ppl:.6f}" + # string += '{} PPL={:.6f} | '.format(key, ppl) + if writer is not None and is_last_rank(): data_type = "test" if test else "validation" writer.add_scalar( f"lm-loss-validation/{key} {data_type}", @@ -2322,13 +1373,11 @@ def evaluate_and_print_results( f"lm-loss-validation/{key} {data_type} vs samples", total_loss_dict[key].item(), args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( f"lm-loss-validation/{key} {data_type} vs tokens", total_loss_dict[key].item(), args.consumed_train_tokens, - x_axis_tokens, ) if args.log_validation_ppl_to_tensorboard: writer.add_scalar( @@ -2338,21 +1387,14 @@ def evaluate_and_print_results( f"lm-loss-validation/{key} {data_type} ppl vs samples", ppl, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( f"lm-loss-validation/{key} {data_type} ppl vs tokens", ppl, args.consumed_train_tokens, - x_axis_tokens, ) - # >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 - if ( - process_non_loss_data_func is not None - and writer.is_enabled() - and is_last_rank() - ): + if process_non_loss_data_func is not None and writer and is_last_rank(): process_non_loss_data_func(collected_non_loss_data, iteration, writer) length = len(string) + 1 @@ -2375,6 +1417,7 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): args = get_args() # Number of train/valid/test samples. + assert args is not None if args.train_samples: train_samples = args.train_samples else: @@ -2473,6 +1516,7 @@ def build_train_valid_test_data_iterators(build_train_valid_test_datasets_provid """Build pretraining data iterators.""" args = get_args() + assert args is not None # Build loaders. train_dataloader, valid_dataloader, test_dataloader = ( From 2113dbc9fc8b38aea09786380ce8846894c0eb25 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 13 Sep 2024 09:16:20 -0500 Subject: [PATCH 322/590] Update `megatron/utils.py` --- megatron/training.py | 147 ++++++----- megatron/training_log.py | 543 +++++++++++++++------------------------ megatron/utils.py | 247 +++++++----------- 3 files changed, 399 insertions(+), 538 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 98413860492..90a1250648d 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -61,17 +61,17 @@ check_adlr_autoresume_termination, checkpoint_throughput_calculator, found_kill_switch, - # num_floating_point_operations, - # report_memory, - # throughput_calculator, unwrap_model, update_rotary_pos_emb, ) -# noqa: E402 -# The earliest we can measure the start time. -_TRAIN_START_TIME = time.time() -# noqa +from megatron.profiler import ( + setup_profiler, + trigger, + on_step_begin, + on_step_end, +) + dlp = Profile("TRAINING") @@ -87,11 +87,6 @@ LOG_LEVEL: str = str(os.environ.get("LOG_LEVEL", "INFO")).upper() log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") -# try: -# import wandb -# except (ImportError, ModuleNotFoundError): -# wandb = None - def print_datetime(string): """Note that this call will sync across all ranks.""" @@ -415,6 +410,8 @@ def get_model( ): """Build the model.""" args = get_args() + accelerator = get_accelerator() + assert accelerator is not None assert args is not None args.model_type = model_type @@ -521,7 +518,7 @@ def get_model( if wrap_with_ddp: if args.DDP_impl == "torch": - i = get_accelerator().current_device() + i = accelerator.current_device() model = [ torchDDP( model_module, @@ -808,6 +805,7 @@ def setup_model_and_optimizer( log.info("Initializing ICT from pretrained BERT model") unwrapped_model[0].init_state_dict_from_bert() if args.fp16: + assert optimizer is not None optimizer.reload_model_params() # random-LTD requires converting transformer layers if args.random_ltd: @@ -822,8 +820,8 @@ def train_step( """Single training step.""" args = get_args() timers = get_timers() - - assert args is not None and timers is not None + accelerator = get_accelerator() + assert args is not None and timers is not None and accelerator is not None if args.deepspeed and args.ds_pipeline_enabled: num_zeros_in_grad = 0 assert isinstance(model[0], deepspeed.PipelineEngine) @@ -860,11 +858,13 @@ def train_step( if args.timing_log_level < 2: config.timers = None + num_microbatches = get_num_microbatches() + assert num_microbatches is not None losses_reduced = forward_backward_func( forward_step_func=forward_step_func, data_iterator=data_iterator, model=model, - num_microbatches=get_num_microbatches(), + num_microbatches=num_microbatches, seq_length=args.seq_length, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, @@ -879,8 +879,8 @@ def train_step( args.teacher_forward = False # Empty unused memory. - if args.empty_unused_memory_level >= 1: - torch.cuda.empty_cache() + if args.empty_unused_memory_level >= 1 and accelerator is not None: + accelerator.empty_cache() # Reduce gradients. if not args.deepspeed: @@ -917,8 +917,14 @@ def train_step( # Update learning rate. if args.deepspeed: - skipped_iter = 0 + skipped_iter = 0 if update_successful else 1 grad_norm = model[0].get_global_grad_norm() + # XXX: [saforem2]: ---------------------------------------------------- + # Is `num_zeros_in_grad` worth calculating (/ implementing) ?? + # the `Megatron`-specific implementation is at: + # [megatron.optimizer.clip_grads.count_zeros_fp32](./optimizer/clip_grads.py) + # For now, explicitly set to None + # --------------------------------------------------------------------- num_zeros_in_grad = None loss_reduced = {} for key in losses_reduced[0]: @@ -927,29 +933,28 @@ def train_step( losses_reduced_for_key ) return loss_reduced, skipped_iter, grad_norm, num_zeros_in_grad + if update_successful: + increment = ( + get_num_microbatches() * args.micro_batch_size * args.data_parallel_size + ) + opt_param_scheduler.step(increment=increment) + skipped_iter = 0 else: - if update_successful: - increment = ( - get_num_microbatches() * args.micro_batch_size * args.data_parallel_size - ) - opt_param_scheduler.step(increment=increment) - skipped_iter = 0 - else: - skipped_iter = 1 + skipped_iter = 1 - # Empty unused memory. - if args.empty_unused_memory_level >= 2: - torch.cuda.empty_cache() - - if mpu.is_pipeline_last_stage(ignore_virtual=True): - # Average loss across microbatches. - loss_reduced = {} - for key in losses_reduced[0]: - losses_reduced_for_key = [x[key] for x in losses_reduced] - loss_reduced[key] = sum(losses_reduced_for_key) / len( - losses_reduced_for_key - ) - return loss_reduced, skipped_iter, grad_norm, num_zeros_in_grad + # Empty unused memory. + if args.empty_unused_memory_level >= 2 and accelerator is not None: + accelerator.empty_cache() + + if mpu.is_pipeline_last_stage(ignore_virtual=True): + # Average loss across microbatches. + loss_reduced = {} + for key in losses_reduced[0]: + losses_reduced_for_key = [x[key] for x in losses_reduced] + loss_reduced[key] = sum(losses_reduced_for_key) / len( + losses_reduced_for_key + ) + return loss_reduced, skipped_iter, grad_norm, num_zeros_in_grad return {}, skipped_iter, grad_norm, num_zeros_in_grad @@ -983,10 +988,12 @@ def train( """Train the model function.""" args = get_args() timers = get_timers() - assert args is not None - assert timers is not None + accelerator = get_accelerator() + assert args is not None and timers is not None and accelerator is not None # Write args to tensorboard write_args_to_tensorboard() + assert accelerator is not None + setup_profiler(args, accelerator.device_name()) if args.random_ltd: # random-ltd requires different randomness on each rank import random @@ -1002,6 +1009,7 @@ def train( iteration = args.iteration # Translate args to core configuration config = core_transformer_config_from_args(args) + num_skipped_iters = 0 if not args.deepspeed: config.grad_scale_func = optimizer.scale_loss config.timers = timers @@ -1029,6 +1037,7 @@ def train( while iteration < args.train_iters and ( args.train_tokens is None or args.consumed_train_tokens < args.train_tokens ): + trigger(on_step_begin) update_num_microbatches(args.consumed_train_samples) if args.deepspeed: # inform deepspeed of any batch size changes @@ -1051,10 +1060,12 @@ def train( [i <= (iteration + 1) <= j for (i, j) in ranges_to_skip] ): log.info(f"Caught {iteration + 1} in 'ranges_to_skip', skipping!") - loss_dict = {} + # total_loss_dict = {"skipped iterations": } skipped_iter = 1 + total_loss_dict["skipped iterations"] += skipped_iter grad_norm = None num_zeros_in_grad = None + num_skipped_iters += 1 else: if os.getenv("TORCH_PROFILER_ENABLE") == "2": from torch.profiler import profile, ProfilerActivity @@ -1187,9 +1198,7 @@ def train( # Exiting based on duration if args.exit_duration_in_mins: train_time = (time.time() - _TRAIN_START_TIME) / 60.0 - done_cuda = get_accelerator().IntTensor( - [train_time > args.exit_duration_in_mins] - ) + done_cuda = accelerator.IntTensor([train_time > args.exit_duration_in_mins]) torch.distributed.all_reduce(done_cuda, op=torch.distributed.ReduceOp.MAX) done = done_cuda.item() if done: @@ -1208,6 +1217,19 @@ def train( torch.distributed.barrier() print_datetime("exiting program at iteration {}".format(iteration)) sys.exit() + trigger(on_step_end) + # Exiting based on kill switch file + if found_kill_switch(): + if args.save and not saved_checkpoint: + save_checkpoint_and_time( + iteration, model, optimizer, opt_param_scheduler + ) + torch.distributed.barrier() + print_datetime( + f"Detected kill switch at {args.kill_switch_file}, " + f"iteration={iteration}. Exiting" + ) + sys.exit() return iteration @@ -1222,7 +1244,8 @@ def evaluate( ): """Evaluation.""" args = get_args() - assert args is not None + accelerator = get_accelerator() + assert args is not None and accelerator is not None if args.vision_pretraining and args.vision_pretraining_type == "dino": compute_feature_bank(model) @@ -1244,6 +1267,10 @@ def evaluate( total_loss_dict = {} + num_microbatches = get_num_microbatches() + assert num_microbatches is not None + forward_backward_func = get_forward_backward_func() + with torch.no_grad(): iteration = 0 while iteration < args.eval_iters: @@ -1251,20 +1278,19 @@ def evaluate( if verbose and iteration % args.log_interval == 0: log.info("Evaluating iter {}/{}".format(iteration, args.eval_iters)) - forward_backward_func = get_forward_backward_func() # Don't care about timing during evaluation config.timers = None if args.deepspeed and args.ds_pipeline_enabled: # DeepSpeed uses eval_batch() and already aggregates losses. assert isinstance(model, list) and len(model) == 1 loss = model[0].eval_batch(data_iterator) - loss_dicts = [{"lm loss": loss}] * get_num_microbatches() + loss_dicts = [{"lm loss": loss}] * num_microbatches else: loss_dicts = forward_backward_func( forward_step_func=forward_step_func, data_iterator=data_iterator, model=model, - num_microbatches=get_num_microbatches(), + num_microbatches=num_microbatches, seq_length=args.seq_length, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, @@ -1274,7 +1300,7 @@ def evaluate( # Empty unused memory if args.empty_unused_memory_level >= 1: - torch.cuda.empty_cache() + accelerator.empty_cache() if mpu.is_pipeline_last_stage(ignore_virtual=True): # Reduce across processes. @@ -1282,16 +1308,14 @@ def evaluate( for key in loss_dict: if "moe" not in key: total_loss_dict[key] = ( - total_loss_dict.get( - key, get_accelerator().FloatTensor([0.0]) - ) + total_loss_dict.get(key, accelerator.FloatTensor([0.0])) + loss_dict[key] ) args.consumed_valid_samples += ( mpu.get_data_parallel_world_size() * args.micro_batch_size - * get_num_microbatches() + * num_microbatches ) collected_non_loss_data = None if process_non_loss_data_func is not None and is_last_rank(): @@ -1299,7 +1323,7 @@ def evaluate( forward_step_func=forward_step_func, data_iterator=data_iterator, model=model, - num_microbatches=get_num_microbatches(), + num_microbatches=num_microbatches, seq_length=args.seq_length, micro_batch_size=args.micro_batch_size, decoder_seq_length=args.decoder_seq_length, @@ -1312,7 +1336,7 @@ def evaluate( model_module.train() for key in total_loss_dict: - total_loss_dict[key] /= args.eval_iters * get_num_microbatches() + total_loss_dict[key] /= args.eval_iters * num_microbatches if args.curriculum_learning_legacy and not args.no_pipeline_parallel: # roll back to actual curriculum seqlen at the end of eval. @@ -1443,7 +1467,8 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider): """Build pretraining data loaders.""" args = get_args() - assert args is not None + accelerator = get_accelerator() + assert args is not None and accelerator is not None (train_dataloader, valid_dataloader, test_dataloader) = (None, None, None) log.info("> building train, validation, and test datasets ...") # Backward compatibility, assume fixed batch size. @@ -1486,11 +1511,9 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider do_valid = valid_dataloader is not None and args.eval_iters > 0 do_test = test_dataloader is not None and args.eval_iters > 0 # Need to broadcast num_tokens and num_type_tokens. - flags = get_accelerator().LongTensor( - [int(do_train), int(do_valid), int(do_test)] - ) + flags = accelerator.LongTensor([int(do_train), int(do_valid), int(do_test)]) else: - flags = get_accelerator().LongTensor([0, 0, 0]) + flags = accelerator.LongTensor([0, 0, 0]) # Broadcast num tokens. if ds_sequence_parallel: torch.distributed.broadcast( diff --git a/megatron/training_log.py b/megatron/training_log.py index e4f74b24c38..be2eacd0595 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -1,154 +1,51 @@ # Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. +""" +training_log.py +""" -"""Pretrain utilities.""" - -from enum import Enum - -# from deepspeed.accelerator import get_accelerator -# from deepspeed.compression.compress import redundancy_clean -import torch -import os import logging +import os -from megatron import get_args -from megatron import get_timers -from megatron import get_tensorboard_writer -from megatron import get_wandb_writer -from megatron import get_num_microbatches -from megatron.core import mpu - -# from megatron import is_rank_0, print_rank_0 -# from megatron import print_rank_last -# from megatron.arguments import core_transformer_config_from_args -# from megatron.checkpointing import load_checkpoint -# from megatron.checkpointing import save_checkpoint -# from megatron.core import mpu, tensor_parallel -# from megatron.core.enums import ModelType -# from megatron.core.pipeline_parallel import get_forward_backward_func -# from megatron.data.data_samplers import build_pretraining_data_loader -# from megatron.initialize import initialize_megatron -# from megatron.initialize import write_args_to_tensorboard -# from megatron.initialize import set_jit_fusion_options -# from megatron.model import Float16Module -# from megatron.model import GPTModel -# from megatron.model import DistributedDataParallel as LocalDDP -# from megatron.model.transformer import ParallelTransformerLayer -# from megatron.model.vision.knn_monitor import compute_feature_bank -# from megatron.optimizer import get_megatron_optimizer -# from megatron.optimizer_param_scheduler import OptimizerParamScheduler -# from megatron.profiler import on_step_begin, on_step_end, setup_profiler, trigger -# from megatron.utils import check_adlr_autoresume_termination -from megatron.utils import found_kill_switch +from deepspeed import get_accelerator import ezpz as ez +import torch -# from megatron.utils import calc_params_l2_norm +from megatron.core import mpu +from megatron.global_vars import ( + get_args, + get_num_microbatches, + get_tensorboard_writer, + get_timers, +) from megatron.utils import ( - # checkpoint_throughput_calculator, + Profile, + is_last_rank, report_memory, throughput_calculator, - # update_rotary_pos_emb, + num_floating_point_operations, ) -try: - import wandb -except (ImportError, ModuleNotFoundError): - wandb = None -# The earliest we can measure the start time. -# _TRAIN_START_TIME = time.time() - - -log = logging.getLogger(__name__) - - -class InteropLoggingTool(Enum): - TENSORBOARD = 1 - WANDB = 2 - RANK: int = ez.get_rank() -LOCAL_RANK: int = ez.get_local_rank() WORLD_SIZE: int = ez.get_world_size() DEVICE_TYPE: str = ez.dist.get_torch_device_type() -DEVICE_ID: str = f"{DEVICE_TYPE}:{LOCAL_RANK}" DEVICE: torch.device = torch.device(DEVICE_TYPE) log: logging.Logger = logging.getLogger(__name__) LOG_LEVEL: str = str(os.environ.get("LOG_LEVEL", "INFO")).upper() log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") +try: + import wandb +except (ImportError, ModuleNotFoundError): + wandb = None -class interop_tool_logger: - def __init__(self, tb_writer=None, wandb_writer=None): - self.tb_writer = tb_writer - self.wandb_writer = wandb_writer - self.custom_x_axis = [] - self.custom_y_axis = {} - self.args = get_args() - assert self.args is not None - if not hasattr(self.args, "logger_iteration"): - self.args.logger_iteration = 1 - assert self.args.logger_iteration is not None - - def is_enabled(self): - return self.tb_writer or self.wandb_writer - - def add_scalar( - self, - key, - scalar_value, - step, - custom_step_name=None, - tool_list=[InteropLoggingTool.TENSORBOARD, InteropLoggingTool.WANDB], - ): - if self.tb_writer and InteropLoggingTool.TENSORBOARD in tool_list: - self.tb_writer.add_scalar(key, scalar_value, step) - - if ( - wandb is not None - and self.wandb_writer - and InteropLoggingTool.WANDB in tool_list - ): - assert self.args is not None - assert self.args.logger_iteration is not None - if not custom_step_name: - self.wandb_writer.log({key: scalar_value}, step=step) - if self.args.logger_iteration < step: - # Updating iteration - self.args.logger_iteration = step - - else: - if custom_step_name not in self.custom_x_axis: - self.custom_x_axis.append(custom_step_name) - wandb.define_metric(custom_step_name) - - if key not in self.custom_y_axis: - self.custom_y_axis[key] = custom_step_name - wandb.define_metric(key, step_metric=custom_step_name) - - self.wandb_writer.log( - {key: scalar_value, custom_step_name: step}, - step=self.args.logger_iteration, - ) - - def add_scalar_to_tb(self, key, scalar_value, step): - return self.add_scalar( - key, scalar_value, step, None, [InteropLoggingTool.TENSORBOARD] - ) - - def add_scalar_to_wandb(self, key, scalar_value, step, custom_step_name=None): - return self.add_scalar( - key, scalar_value, step, custom_step_name, [InteropLoggingTool.WANDB] - ) - - def add_images(self, key, img_tensor, step=None): - if self.tb_writer: - self.tb_writer.add_images(key, img_tensor, step) - if wandb is not None and self.wandb_writer: - self.wandb_writer.log({key: wandb.Image(img_tensor)}, step) +dlp = Profile("TRAINING_LOG") +@dlp.log def training_log( loss_dict, total_loss_dict, @@ -165,12 +62,11 @@ def training_log( ): """Log training information such as losses, timing, ....""" args = get_args() + accelerator = get_accelerator() timers = get_timers() - writer = interop_tool_logger( - tb_writer=get_tensorboard_writer(), wandb_writer=get_wandb_writer() - ) - x_axis_samples = "Samples" - x_axis_tokens = "Tokens" + writer = get_tensorboard_writer() + assert args is not None and timers is not None and accelerator is not None + wandb_metrics = {} # Advanced, skipped, and Nan iterations. advanced_iters_key = "advanced iterations" skipped_iters_key = "skipped iterations" @@ -189,10 +85,12 @@ def training_log( ) # Update losses and set nan iterations got_nan = False - _zero = torch.tensor([0.0]).to(DEVICE) for key in loss_dict: if not skipped_iter: - total_loss_dict[key] = total_loss_dict.get(key, _zero) + loss_dict[key] + total_loss_dict[key] = ( + total_loss_dict.get(key, accelerator.FloatTensor([0.0])) + + loss_dict[key] + ) else: value = loss_dict[key].float().sum().item() is_nan = value == float("inf") or value == -float("inf") or value != value @@ -230,12 +128,10 @@ def training_log( "optimizer", ] - assert args is not None and timers is not None # Calculate batch size. batch_size = ( args.micro_batch_size * args.data_parallel_size * get_num_microbatches() ) - total_iterations = ( total_loss_dict[advanced_iters_key] + total_loss_dict[skipped_iters_key] ) @@ -245,39 +141,35 @@ def training_log( if args.log_timers_to_tensorboard and ( iteration % args.tensorboard_log_interval == 0 ): - timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) # type: ignore - if writer.is_enabled() and (iteration % args.tensorboard_log_interval == 0): + timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) + if writer and (iteration % args.tensorboard_log_interval == 0): writer.add_scalar( - "steps-vs-samples/y=steps,x=samples", - iteration, - args.consumed_train_samples, - x_axis_samples, + "steps-vs-samples/y=steps,x=samples", iteration, args.consumed_train_samples ) writer.add_scalar( "steps-vs-samples/y=samples,x=steps", args.consumed_train_samples, iteration ) writer.add_scalar( - "steps-vs-tokens/y=steps,x=tokens", - iteration, - args.consumed_train_tokens, - x_axis_tokens, + "steps-vs-tokens/y=steps,x=tokens", iteration, args.consumed_train_tokens ) writer.add_scalar( "steps-vs-tokens/y=tokens,x=steps", args.consumed_train_tokens, iteration ) if args.log_learning_rate_to_tensorboard: + wandb_metrics |= { + "learning-rate/iteration": iteration, + "learning-rate/learning-rate": learning_rate, + } writer.add_scalar("learning-rate/learning-rate", learning_rate, iteration) writer.add_scalar( "learning-rate/learning-rate vs samples", learning_rate, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "learning-rate/learning-rate vs tokens", learning_rate, args.consumed_train_tokens, - x_axis_tokens, ) if args.log_batch_size_to_tensorboard: writer.add_scalar("batch-size/batch-size", batch_size, iteration) @@ -285,27 +177,28 @@ def training_log( "batch-size/batch-size vs samples", batch_size, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "batch-size/batch-size vs tokens", batch_size, args.consumed_train_tokens, - x_axis_tokens, ) + wandb_metrics |= { + "lm-loss-training/iteration": iteration, + "lm-loss-training/consumed_train_tokens": args.consumed_train_tokens, + } for key in loss_dict: + wandb_metrics |= {f"lm-loss-training/{key}": loss_dict[key]} writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) writer.add_scalar( f"lm-loss-training/{key}" + " vs samples", loss_dict[key], args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( f"lm-loss-training/{key}" + " vs tokens", loss_dict[key], args.consumed_train_tokens, - x_axis_tokens, ) if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: writer.add_scalar("loss-scale/loss-scale", loss_scale, iteration) @@ -313,13 +206,11 @@ def training_log( "loss-scale/loss-scale vs samples", loss_scale, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "loss-scale/loss-scale vs tokens", loss_scale, args.consumed_train_tokens, - x_axis_tokens, ) if args.log_world_size_to_tensorboard: writer.add_scalar("world-size/world-size", args.world_size, iteration) @@ -327,55 +218,46 @@ def training_log( "world-size/world-size vs samples", args.world_size, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "world-size/world-size vs tokens", args.world_size, args.consumed_train_tokens, - x_axis_tokens, ) if grad_norm is not None: + wandb_metrics |= {"training/grad-norm": grad_norm} writer.add_scalar("grad-norm/grad-norm", grad_norm, iteration) writer.add_scalar( - "grad-norm/grad-norm vs samples", - grad_norm, - args.consumed_train_samples, - x_axis_samples, + "grad-norm/grad-norm vs samples", grad_norm, args.consumed_train_samples ) writer.add_scalar( - "grad-norm/grad-norm vs tokens", - grad_norm, - args.consumed_train_tokens, - x_axis_tokens, + "grad-norm/grad-norm vs tokens", grad_norm, args.consumed_train_tokens ) if num_zeros_in_grad is not None: + wandb_metrics |= {"training/num-zeros": num_zeros_in_grad} writer.add_scalar("num-zeros/num-zeros", num_zeros_in_grad, iteration) writer.add_scalar( "num-zeros/num-zeros vs samples", num_zeros_in_grad, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "num-zeros/num-zeros vs tokens", num_zeros_in_grad, args.consumed_train_tokens, - x_axis_tokens, ) if params_norm is not None: + wandb_metrics |= {"training/params-norm": params_norm} writer.add_scalar("params-norm/params-norm", params_norm, iteration) writer.add_scalar( "params-norm/params-norm vs samples", params_norm, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "params-norm/params-norm vs tokens", params_norm, args.consumed_train_tokens, - x_axis_tokens, ) if hasattr(args, "actual_seq_length"): writer.add_scalar( @@ -385,13 +267,11 @@ def training_log( "seqlen/actual_seq_length vs samples", args.actual_seq_length, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "seqlen/actual_seq_length vs tokens", args.actual_seq_length, args.consumed_train_tokens, - x_axis_tokens, ) if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: writer.add_scalar( @@ -401,13 +281,11 @@ def training_log( "seqlen/curriculum_seqlen vs samples", args.curriculum_seqlen, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "seqlen/curriculum_seqlen vs tokens", args.curriculum_seqlen, args.consumed_train_tokens, - x_axis_tokens, ) if args.random_ltd: writer.add_scalar( @@ -419,13 +297,11 @@ def training_log( "seqlen/random_ltd_reserved_length vs samples", args.random_ltd_reserved_length, args.consumed_train_samples, - x_axis_samples, ) writer.add_scalar( "seqlen/random_ltd_reserved_length vs tokens", args.random_ltd_reserved_length, args.consumed_train_tokens, - x_axis_tokens, ) if args.log_memory_to_tensorboard: mem_stats = torch.cuda.memory_stats() @@ -444,70 +320,56 @@ def training_log( mem_stats["allocation.all.current"], iteration, ) - if iteration % args.tensorboard_log_interval == 0: # This logging write various optimizer states to tensorboard. This # feature may consume extra GPU memory thus is set at false by default. if args.log_optimizer_states_to_tensorboard and optimizer is not None: opt_stats = [0.0] * 8 opt_stats_2 = [0.0] * 4 - - # TODO(billishyahao): Remove me after bf16_optimizer promotes its state. - if not hasattr(optimizer, "state"): - assert hasattr( - optimizer, "optimizer" - ), "Optimizer must have optimizer property." - optimizer.state = optimizer.optimizer.state - for _, group in enumerate(optimizer.param_groups): for _, param in enumerate(group["params"]): - opt_stats[0] += ( - torch.norm(optimizer.state[param]["exp_avg_sq"]).item() - ) ** 2 - opt_stats[1] += ( - torch.norm(optimizer.state[param]["exp_avg_sq"].sqrt()).item() - ) ** 2 - opt_stats[2] += ( - torch.norm(optimizer.state[param]["exp_avg"]).item() - ) ** 2 - opt_stats[3] += (torch.norm(param).item()) ** 2 - opt_stats[4] += torch.norm( - optimizer.state[param]["exp_avg_sq"], p=1 - ).item() - opt_stats[5] += torch.norm( - optimizer.state[param]["exp_avg_sq"].sqrt(), p=1 - ).item() - opt_stats[6] += torch.norm( - optimizer.state[param]["exp_avg"], p=1 - ).item() - opt_stats[7] += torch.norm(param, p=1).item() - opt_stats_2[0] = max( - opt_stats_2[0], - abs(optimizer.state[param]["exp_avg_sq"].max().item()), - abs(optimizer.state[param]["exp_avg_sq"].min().item()), - ) - opt_stats_2[1] = max( - opt_stats_2[1], - optimizer.state[param]["exp_avg_sq"].sqrt().abs_().max().item(), - ) - opt_stats_2[2] = max( - opt_stats_2[2], - abs(optimizer.state[param]["exp_avg"].max().item()), - abs(optimizer.state[param]["exp_avg"].min().item()), - ) - opt_stats_2[3] = max( - opt_stats_2[3], abs(param.max().item()), abs(param.min().item()) - ) + state_param = getattr(optimizer, "state", None) + if state_param is not None: + exp_avg_sq = state_param.get("exp_avg_sq", torch.tensor(0.0)) + exp_avg = state_param.get("exp_avg", torch.tensor(0.0)) + opt_stats[0] += (torch.norm(exp_avg_sq).item()) ** 2 + opt_stats[1] += (torch.norm(exp_avg_sq.sqrt()).item()) ** 2 + opt_stats[2] += (torch.norm(exp_avg).item()) ** 2 + opt_stats[3] += (torch.norm(param).item()) ** 2 + opt_stats[4] += torch.norm(exp_avg_sq, p=1).item() + opt_stats[5] += torch.norm(exp_avg_sq.sqrt(), p=1).item() + opt_stats[6] += torch.norm(exp_avg, p=1).item() + opt_stats[7] += torch.norm(param, p=1).item() + opt_stats_2[0] = max( + opt_stats_2[0], + abs(exp_avg_sq.max().item()), + abs(exp_avg_sq.min().item()), + ) + opt_stats_2[1] = max( + opt_stats_2[1], exp_avg_sq.sqrt().abs_().max().item() + ) + opt_stats_2[2] = max( + opt_stats_2[2], + abs(exp_avg.max().item()), + abs(exp_avg.min().item()), + ) + opt_stats_2[3] = max( + opt_stats_2[3], + abs(param.max().item()), + abs(param.min().item()), + ) # print('step {} rank {} before sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) if args.zero_stage > 0: # ZeRO partiions optimizer states - # opt_stats = get_accelerator().FloatTensor(opt_stats) - opt_stats = torch.tensor(opt_stats).to(DEVICE) + # opt_stats = opt_stats.clone().detach() + # opt_stats = get_accelerator().FloatTensor + opt_stats = accelerator.FloatTensor(opt_stats) torch.distributed.all_reduce( opt_stats, group=mpu.get_sequence_data_parallel_group() ) # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) + # opt_stats_2 = opt_stats_2.clone().detach() + opt_stats_2 = accelerator.FloatTensor(opt_stats_2) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, @@ -515,13 +377,13 @@ def training_log( ) if args.tensor_model_parallel_size > 1: - opt_stats = torch.tensor(opt_stats).to(DEVICE) - # opt_stats = get_accelerator().FloatTensor(opt_stats) + # opt_stats = opt_stats.clone().detach() + opt_stats = accelerator.FloatTensor(opt_stats) torch.distributed.all_reduce( opt_stats, group=mpu.get_tensor_model_parallel_group() ) - # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) + # opt_stats_2 = opt_stats_2.clone().detach() + opt_stats_2 = accelerator.FloatTensor(opt_stats_2) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, @@ -529,95 +391,97 @@ def training_log( ) if args.pipeline_model_parallel_size > 1: - # opt_stats = get_accelerator().FloatTensor(opt_stats) - opt_stats = torch.tensor(opt_stats).to(DEVICE) + # opt_stats = opt_stats.clone().detach() + opt_stats = accelerator.FloatTensor(opt_stats) torch.distributed.all_reduce( opt_stats, group=mpu.get_pipeline_model_parallel_group() ) - # opt_stats_2 = get_accelerator().FloatTensor(opt_stats_2) - opt_stats_2 = torch.tensor(opt_stats_2).to(DEVICE) + # opt_stats_2 = opt_stats_2.clone().detach() + opt_stats_2 = accelerator.get_accelerator().FloatTensor(opt_stats_2) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, group=mpu.get_pipeline_model_parallel_group(), ) - + wandb_metrics |= { + "optimizer/learning_rate": learning_rate, + "optimizer/iteration": args.iteration, + "optimizer/consumed_train_tokens": args.consumed_train_tokens, + "optimizer/variance_l2": opt_stats[0] ** 0.5, + "optimizer/variance_sqrt_l2": opt_stats[1] ** 0.5, + "optimizer/momentum_l2": opt_stats[2] ** 0.5, + "optimizer/weight_l2": opt_stats[3] ** 0.5, + "optimizer/variance_l1": opt_stats[4], + "optimizer/variance_sqrt_l1": opt_stats[5], + "optimizer/momentum_l1": opt_stats[6], + "optimizer/weight_l1": opt_stats[7], + "optimizer/variance_abs_max": opt_stats_2[0], + "optimizer/variance_sqrt_abs_max": opt_stats_2[1], + "optimizer/momentum_abs_max": opt_stats_2[2], + "optimizer/weight_abs_max": opt_stats_2[3], + } # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) - # if writer.is_enabled() and is_last_rank(): - if writer.is_enabled() and RANK == 0: + if writer and is_last_rank(): writer.add_scalar( "optimizer/variance_l2 vs tokens", opt_stats[0] ** 0.5, args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/variance_sqrt_l2 vs tokens", opt_stats[1] ** 0.5, args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/momentum_l2 vs tokens", opt_stats[2] ** 0.5, args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/weight_l2 vs tokens", opt_stats[3] ** 0.5, args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/variance_l1 vs tokens", opt_stats[4], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/variance_sqrt_l1 vs tokens", opt_stats[5], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/momentum_l1 vs tokens", opt_stats[6], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/weight_l1 vs tokens", opt_stats[7], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/variance_abs_max vs tokens", opt_stats_2[0], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/variance_sqrt_abs_max vs tokens", opt_stats_2[1], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/momentum_abs_max vs tokens", opt_stats_2[2], args.consumed_train_tokens, - x_axis_tokens, ) writer.add_scalar( "optimizer/weight_abs_max vs tokens", opt_stats_2[3], args.consumed_train_tokens, - x_axis_tokens, ) - writer.add_scalar( "optimizer/variance_l2", opt_stats[0] ** 0.5, iteration ) @@ -643,6 +507,8 @@ def training_log( ) writer.add_scalar("optimizer/weight_abs_max", opt_stats_2[3], iteration) + assert args is not None + assert timers is not None if iteration % args.log_interval == 0: elapsed_time = timers("interval-time").elapsed(barrier=True) elapsed_time_per_iteration = elapsed_time / total_iterations @@ -659,114 +525,126 @@ def training_log( tokens_per_gpu_per_second_per_replica = ( tokens_per_gpu_per_second / args.data_parallel_size ) - - if writer.is_enabled(): - writer.add_scalar_to_wandb( - "throughput/iteration-time", elapsed_time_per_iteration, iteration - ) # 1000 ms / s - writer.add_scalar_to_wandb( - "throughput/samples_per_sec", samples_per_sec, iteration - ) - writer.add_scalar_to_wandb( - "throughput/samples_per_sec_per_replica", - samples_per_sec_per_replica, - iteration, - ) - writer.add_scalar_to_wandb( - "throughput/tokens_per_sec", tokens_per_sec, iteration - ) - writer.add_scalar_to_wandb( - "throughput/tokens_per_sec_per_replica", - tokens_per_sec_per_replica, - iteration, - ) - writer.add_scalar_to_wandb( - "throughput/tokens_per_gpu_per_sec", - tokens_per_gpu_per_second, - iteration, - ) - writer.add_scalar_to_wandb( - "throughput/tokens_per_gpu_per_sec_per_replica", - tokens_per_gpu_per_second_per_replica, - iteration, + # NOTE: [2024-06-19] + # Updated to use (more accurate) calculation according to + # `num_floating_point_operations` from NVIDIA/Megatron-LM + num_flop_lm = num_floating_point_operations(args, batch_size) + num_flop_per_sec_lm = num_flop_lm / elapsed_time_per_iteration + tflops_lm = num_flop_per_sec_lm / (10**12) + tflops_lm_per_gpu = tflops_lm / args.world_size + wandb_metrics |= { + "throughput/iteration-time": elapsed_time_per_iteration, # 1000 ms / s + "throughput/samples_per_sec": samples_per_sec, + "throughput/samples_per_sec_per_replica": samples_per_sec_per_replica, + "throughput/tokens_per_sec": tokens_per_sec, + "throughput/tokens_per_sec_per_replica": tokens_per_sec_per_replica, + "throughput/tokens_per_gpu_per_sec": tokens_per_gpu_per_second, + "throughput/tokens_per_gpu_per_sec_per_replica": tokens_per_gpu_per_second_per_replica, + "throughput/tflops": tflops, + "throughput/tflops-new": num_flop_lm / elapsed_time_per_iteration, + "throughput/tflops-lm": tflops_lm_per_gpu, + "throughput/approx_params_in_billions": approx_parameters_in_billions, + "throughput/elapsed_ms_per_iteration": elapsed_time_per_iteration, + "throughput/iteration": iteration, + } + if loss_dict is not None: + wandb_metrics |= { + "loss/iteration": iteration, + **{f"loss/{k}": v for k, v in loss_dict.items()}, + } + if writer and args.log_timers_to_tensorboard: + writer.add_scalar( + "iteration-time/iteration-time", elapsed_time_per_iteration, iteration ) - writer.add_scalar_to_wandb("throughput/tflops", tflops, iteration) - writer.add_scalar_to_wandb( - "throughput/approx_params_in_billions", - approx_parameters_in_billions, - iteration, + writer.add_scalar( + "iteration-time/iteration-time vs samples", + elapsed_time_per_iteration, + args.consumed_train_samples, ) - writer.add_scalar_to_wandb( - "throughput/elapsed_ms_per_iteration", + writer.add_scalar( + "iteration-time/iteration-time vs tokens", elapsed_time_per_iteration, - iteration, + args.consumed_train_tokens, ) - if loss_dict is not None: - for k, v in loss_dict.items(): - writer.add_scalar_to_wandb(f"loss/{k}", v, iteration) - - if args.log_timers_to_tensorboard: - writer.add_scalar( - "iteration-time/iteration-time", - elapsed_time_per_iteration, - iteration, - ) - writer.add_scalar( - "iteration-time/iteration-time vs samples", - elapsed_time_per_iteration, - args.consumed_train_samples, - x_axis_samples, - ) - writer.add_scalar( - "iteration-time/iteration-time vs tokens", - elapsed_time_per_iteration, - args.consumed_train_tokens, - x_axis_tokens, - ) - log_string = " iteration {:8d}/{:8d} |".format(iteration, args.train_iters) - log_string += " consumed samples: {:12d} |".format(args.consumed_train_samples) - log_string += " consumed tokens: {:12d} |".format(args.consumed_train_tokens) - log_string += " elapsed time per iteration (ms): {:.1f} |".format( - elapsed_time_per_iteration * 1000.0 + # metrics_to_log = { + # 'iteration': iteration, + # 'train_iters': args.train_iters, + # 'consumed_samples': args.consumed_train_samples, + # 'consumed_tokens': args.consumed_tokens, + # } + log_string = f" iteration={iteration:8d}/{args.train_iters:8d} |" + # .format( iteration, args.train_iters) + log_string += ( + f" consumed_samples={args.consumed_train_samples:12d} |" + # .format(args.consumed_train_samples) ) - log_string += " learning rate: {:.3E} |".format(learning_rate) - log_string += " global batch size: {:5d} |".format(batch_size) - + log_string += f" consumed_tokens={args.consumed_train_tokens:12d} |" + # .format( args.consumed_train_tokens) + log_string += ( + " elapsed_time_per_iteration_ms=" + f"{elapsed_time_per_iteration * 1000.0:.1f} |" + # .format( elapsed_time_per_iteration * 1000.0) + ) + log_string += f" learning_rate={learning_rate:.6g} |" + log_string += f" global_batch_size={batch_size:5d} |" + # if wandb is not None and getattr(wandb, 'run', None) is not None: + wandb_metrics |= { + "training/iteration": iteration, + "training/iteration_time": elapsed_time_per_iteration, + "training/iteration_time_vs_tokens": ( + elapsed_time_per_iteration / args.consumed_train_tokens + ), + "training/iteration_time_vs_samples": ( + (elapsed_time_per_iteration / args.consumed_train_samples), + ), + "training/consumed_samples": args.consumed_train_samples, + "training/consumed_tokens": args.consumed_train_tokens, + } for key in total_loss_dict: if key not in [advanced_iters_key, skipped_iters_key, nan_iters_key]: avg = total_loss_dict[key].item() / float( max(1, total_loss_dict[advanced_iters_key]) ) if avg > 0.0: - log_string += " {}: {:.6E} |".format(key, avg) - # total_loss_dict[key] = get_accelerator().FloatTensor([0.0]) - total_loss_dict[key] = torch.tensor([0.0]).to(DEVICE) + log_string += " {}={:.6f} |".format(key, avg) + total_loss_dict[key] = accelerator.FloatTensor([0.0]) if loss_scale is not None: - log_string += " loss scale: {:.1f} |".format(loss_scale) + log_string += " loss_scale={:.1f} |".format(loss_scale) + wandb_metrics |= {"loss/loss_scale": loss_scale} if grad_norm is not None: - log_string += " grad norm: {:.3f} |".format(grad_norm) + log_string += " grad_norm={:.3f} |".format(grad_norm) + wandb_metrics |= {"loss/grad_norm": grad_norm} if num_zeros_in_grad is not None: - log_string += " num zeros: {:.1f} |".format(num_zeros_in_grad) + log_string += " num_zeros={:.1f} |".format(num_zeros_in_grad) + wandb_metrics |= {"loss/num_zeros_in_grad": num_zeros_in_grad} if params_norm is not None: - log_string += " params norm: {:.3f} |".format(params_norm) + log_string += " params_norm={:.3f} |".format(params_norm) + wandb_metrics |= {"loss/params_norm": params_norm} if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - log_string += " curriculum seqlen: {:5d} |".format(args.curriculum_seqlen) + log_string += " curriculum_seqlen={:5d} |".format(args.curriculum_seqlen) if args.random_ltd: - log_string += " random ltd reserved length: {:5d} |".format( + log_string += " random_ltd reserved_length={:5d} |".format( args.random_ltd_reserved_length ) - log_string += " actual seqlen: {:5d} |".format(seq_len) - log_string += " number of skipped iterations: {:3d} |".format( + # log_string += " | ".join([ + # f"{seq_len=:5d} ", + # f"{}" + # f"number_of_skipped_iterations={:3d}", + # + # ]) + log_string += " actual_seqlen={:5d} |".format(seq_len) + log_string += " number_of_skipped_iterations={:3d} |".format( total_loss_dict[skipped_iters_key] ) - log_string += " number of nan iterations: {:3d} |".format( + log_string += " number_of_nan_iterations={:3d} |".format( total_loss_dict[nan_iters_key] ) - log_string += " samples per second: {:.3f} |".format(samples_per_sec) - log_string += " tokens per gpu per second (tgs): {:.3f} |".format( + log_string += " samples_per_second={:.3f} |".format(samples_per_sec) + log_string += " tokens_per_gpu_per_second_tgs={:.3f} |".format( tokens_per_gpu_per_second ) - log_string += " TFLOPs: {:.2f} |".format(tflops) + log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) + log_string += " [DS]TFLOPs={:.2f} |".format(tflops) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 @@ -776,6 +654,13 @@ def training_log( # Report memory after optimizer state has been initialized. report_memory("(after {} iterations)".format(iteration)) report_memory_flag = False - timers.log(timers_to_log, normalizer=args.log_interval) + if wandb is not None and getattr(wandb, "run", None) is not None: + wandb_metrics |= { + "training/skiped_iterations": total_loss_dict[skipped_iters_key] + } + wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} + wandb.log(wandb_metrics) + if timers is not None: + timers.log(timers_to_log, normalizer=args.log_interval) return report_memory_flag diff --git a/megatron/utils.py b/megatron/utils.py index c1b43624d37..31c8e205085 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -7,17 +7,21 @@ import os import time import logging -from typing import ContextManager, Optional +from typing import Optional +# from ezpz.dist import get_rank import torch from torch.nn.parallel import DistributedDataParallel as torchDDP from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == "cuda": +ACCELERATOR = get_accelerator() +assert ACCELERATOR is not None + +if ACCELERATOR.device_name() == "cuda": try: - from apex.multi_tensor_apply import multi_tensor_applier - import amp_C + from apex.multi_tensor_apply import multi_tensor_applier # type: ignore + import amp_C # type:ignore HAS_APEX = True except Exception: @@ -38,19 +42,15 @@ _DLIO_PROFILER_EXIST = True try: - import dlio_profiler -except: + import dlio_profiler # type: ignore +except Exception: _DLIO_PROFILER_EXIST = False if _DLIO_PROFILER_EXIST: - from dlio_profiler.logger import fn_interceptor as Profile - from dlio_profiler.logger import dlio_logger as PerfTrace + from dlio_profiler.logger import fn_interceptor as Profile # type:ignore + from dlio_profiler.logger import dlio_logger as PerfTrace # type:ignore else: from functools import wraps - # from contextlib import nullcontext - # Profile: ContextManager = nullcontext - # - # class Profile(nullable_schema) class Profile: def __init__(self, type="PROFILER"): @@ -70,7 +70,6 @@ def __exit__(self, *args, **kwargs): dt = time.perf_counter() - self._start log.info(f"{self.type} took: {dt:.6f}s") - class dlio_logger: def __init__( self, @@ -87,9 +86,9 @@ def iter(self, a): def get_logger( - name: str, - level: str = "INFO", - rank_zero_only: Optional[bool] = None, + name: str, + level: str = "INFO", + rank_zero_only: Optional[bool] = None, ) -> logging.Logger: """Returns a `logging.Logger` object. @@ -105,7 +104,8 @@ def get_logger( def update_rotary_pos_emb(seq_length): args = get_args() - assert args is not None + accelerator = get_accelerator() + assert args is not None and accelerator is not None rotary_dim = ( args.hidden_size // args.num_attention_heads if args.kv_channels is None @@ -119,7 +119,7 @@ def update_rotary_pos_emb(seq_length): # Wang and Komatsuzaki et al # https://github.com/kingoflolz/mesh-transformer-jax/ rotary_pos_emb = RotaryEmbedding(rotary_dim, theta=args.rope_theta)(seq_length).to( - get_accelerator().current_device_name() + accelerator.current_device_name() ) args.rotary_pos_emb = rotary_pos_emb @@ -189,21 +189,22 @@ def average_losses_across_data_parallel_group(losses): def report_memory(name): """Simple GPU memory report.""" + accelerator = get_accelerator() + assert accelerator is not None mega_bytes = 1024.0 * 1024.0 string = name + " memory (MB)" - string += " | allocated: {}".format( - get_accelerator().memory_allocated() / mega_bytes - ) + string += " | allocated: {}".format(accelerator.memory_allocated() / mega_bytes) string += " | max allocated: {}".format( - get_accelerator().max_memory_allocated() / mega_bytes - ) - string += " | reserved: {}".format(get_accelerator().memory_reserved() / mega_bytes) - string += " | max reserved: {}".format( - get_accelerator().max_memory_reserved() / mega_bytes + accelerator.max_memory_allocated() / mega_bytes ) + reserved = accelerator.memory_reserved() + max_reserved = accelerator.max_memory_reserved() + if reserved is not None: + string += " | reserved: {}".format(reserved / mega_bytes) + if max_reserved is not None: + string += " | max reserved: {}".format(max_reserved / mega_bytes) if mpu.get_data_parallel_rank() == 0: log.info(f"[Rank {RANK}] {string}") - # log.info("[Rank {}] {}".format(torch.distributed.get_rank(), string)) # , flush=True) def print_params_min_max_norm(optimizer, iteration): @@ -222,19 +223,19 @@ def print_params_min_max_norm(optimizer, iteration): iteration, rank, index, int(param.tensor_model_parallel) ) string += "{:.6E}, {:.6E}, {:.6E}\n".format(min_, max_, norm) - # print(string, flush=True) log.info(string) def check_adlr_autoresume_termination(iteration, model, optimizer, opt_param_scheduler): """Check for autoresume signal and exit if it is received.""" from megatron.checkpointing import save_checkpoint + args = get_args() assert args is not None autoresume = get_adlr_autoresume() # Add barrier to ensure consistnecy. torch.distributed.barrier() - if autoresume.termination_requested(): + if autoresume is not None and autoresume.termination_requested(): if args.save: save_checkpoint(iteration, model, optimizer, opt_param_scheduler) print_rank_0(">>> autoresume termination request found!") @@ -265,18 +266,9 @@ def get_ltor_masks_and_position_ids( attention_mask = None if not skip_mask: -# <<<<<<< HEAD -# attention_mask = torch.tril( -# torch.ones((att_mask_batch, seq_length, seq_length)) -# ).view(att_mask_batch, 1, seq_length, seq_length) -# ======= attention_mask = torch.tril( - torch.ones( - (att_mask_batch, seq_length, seq_length), - device=data.device - ) + torch.ones((att_mask_batch, seq_length, seq_length), device=data.device) ).view(att_mask_batch, 1, seq_length, seq_length) -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 # Loss mask. loss_mask = torch.ones(data.size(), dtype=torch.float, device=data.device) @@ -304,27 +296,21 @@ def get_ltor_masks_and_position_ids( for j in range(eod_index.size()[0]): i = eod_index[j] # Mask attention loss. - if reset_attention_mask and not skip_mask: + if ( + reset_attention_mask + and not skip_mask + and attention_mask is not None + ): attention_mask[b, 0, (i + 1) :, : (i + 1)] = 0 # Reset positions. if reset_position_ids: position_ids[b, (i + 1) :] -= i + 1 - prev_index prev_index = i + 1 - # # Convert attention mask to binary: - # if not skip_mask: - # <<<<<<< HEAD - # assert attention_mask is not None - # attention_mask = attention_mask < 0.5 - # attention_mask = attention_mask.to(data.device) - # ======= - # attention_mask = (attention_mask < 0.5) - # >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 - # Convert attention mask to binary: if not skip_mask: assert attention_mask is not None - attention_mask = (attention_mask < 0.5) + attention_mask = attention_mask < 0.5 return attention_mask, loss_mask, position_ids @@ -364,10 +350,7 @@ def is_rank_0(): if torch.distributed.is_initialized(): if torch.distributed.get_rank() == 0 or ( is_aml() - and ( - torch.distributed.get_rank() - % get_accelerator().device_count() - ) == 0 + and (torch.distributed.get_rank() % get_accelerator().device_count()) == 0 ): return True else: @@ -396,6 +379,37 @@ def get_parameters_in_billions(model): return approx_parameters_in_billions * gpus_per_model / (1e9) +def num_floating_point_operations(args, batch_size): + # Group Query Attention. + # if not args.group_query_attention: + if not args.num_key_value_heads: + args.num_key_value_heads = args.num_attention_heads + # args.num_query_groups = args.num_attention_heads + # MoE. + # num_experts_routed_to = 1 if args.num_experts is None else args.moe_router_topk + num_experts_routed_to = 1 if args.num_experts is None else args.topk + gated_linear_multiplier = 3 / 2 if args.swiglu else 1 + return ( + 12 + * batch_size + * args.seq_length + * args.num_layers + * args.hidden_size + * args.hidden_size + * ( + 1 + + ( + (args.ffn_hidden_size / args.hidden_size) + * num_experts_routed_to + * gated_linear_multiplier + ) + + (args.num_key_value_heads / args.num_attention_heads) + + (args.seq_length / args.hidden_size) + + (args.padded_vocab_size / (2 * args.num_layers * args.hidden_size)) + ) + ) + + def throughput_calculator(model, args, iteration_time, total_iterations): batch_size = ( args.micro_batch_size * get_num_microbatches() * args.data_parallel_size @@ -419,55 +433,47 @@ def throughput_calculator(model, args, iteration_time, total_iterations): # General TFLOPs formula (borrowed from Equation 3 in Section 5.1 of # https://arxiv.org/pdf/2104.04473.pdf). -# <<<<<<< HEAD -# # The factor of 4 is when used with activation check-pointing, -# # otherwise it will be 3. -# checkpoint_activations_factor = 3 -# if hasattr(args, "checkpoint_activations") and args.checkpoint_activations: -# checkpoint_activations_factor = 4 -# if hasattr(args, "recompute_granularity") and ( -# args.recompute_granularity == "selective" -# or args.recompute_granularity == "full" -# ): -# checkpoint_activations_factor = 4 -# ======= # correction has been made to TFLOPs formula due to incorrect behavior # observed with selective recompute when GQA not used and for all with GQA -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 seq_len = args.seq_length if hasattr(args, "actual_seq_length"): seq_len = args.actual_seq_length -# <<<<<<< HEAD -# flops_per_iteration = ( -# 24 -# * checkpoint_activations_factor -# * batch_size -# * seq_len -# * num_layers -# * (hidden_size**2) -# ) * ( -# 1.0 -# + (seq_len / (6.0 * hidden_size)) -# + (vocab_size / (16.0 * num_layers * hidden_size)) -# ) -# ======= - - pre_and_post_mha_gemm_macs = batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len - mha_bgemm_macs = batch_size * num_layers * 2 * head_dim * num_attention_heads * (seq_len**2) - ffn_gemm_macs = batch_size * num_layers * ffn_multiplier * ffn_hidden_size * hidden_size * seq_len + pre_and_post_mha_gemm_macs = ( + batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len + ) + mha_bgemm_macs = ( + batch_size * num_layers * 2 * head_dim * num_attention_heads * (seq_len**2) + ) + ffn_gemm_macs = ( + batch_size + * num_layers + * ffn_multiplier + * ffn_hidden_size + * hidden_size + * seq_len + ) logit_lmhead_gemm_macs = batch_size * vocab_size * hidden_size * seq_len - fwd_macs = pre_and_post_mha_gemm_macs + mha_bgemm_macs + ffn_gemm_macs + logit_lmhead_gemm_macs + fwd_macs = ( + pre_and_post_mha_gemm_macs + + mha_bgemm_macs + + ffn_gemm_macs + + logit_lmhead_gemm_macs + ) bwd_macs = 2 * fwd_macs fwd_bwd_macs = fwd_macs + bwd_macs - if (hasattr(args, 'checkpoint_activations') and args.checkpoint_activations) or (hasattr(args, 'recompute_granularity') and args.recompute_granularity == 'full'): + if (hasattr(args, "checkpoint_activations") and args.checkpoint_activations) or ( + hasattr(args, "recompute_granularity") and args.recompute_granularity == "full" + ): fwd_bwd_macs += fwd_macs - if hasattr(args, 'recompute_granularity') and args.recompute_granularity == 'selective': + if ( + hasattr(args, "recompute_granularity") + and args.recompute_granularity == "selective" + ): fwd_bwd_macs += mha_bgemm_macs flops_per_iteration = fwd_bwd_macs * macs_per_flops -# >>>>>>> 0d6e3793a1fc06eded9764ef15ad12bcc0281101 tflops = flops_per_iteration / (elapsed_time_per_iter * args.world_size * (10**12)) return samples_per_second, tflops, approx_parameters_in_billions @@ -517,58 +523,6 @@ def dump_position_embed_weights(preamble, iteration, model): ) -# def dump_weights(preamble, iteration, model, optimizer, tensor=None): -# # return -# tp_rank = mpu.get_tensor_model_parallel_rank() -# pp_rank = mpu.get_pipeline_model_parallel_rank() -# dp_rank = mpu.get_data_parallel_rank() -# dp_size = mpu.get_data_parallel_world_size() -# fn = f"debug-bf16-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" -# # only care for first and last pp stages and dp0 tp0 -# # if not (mpu.is_pipeline_first_stage() or mpu.is_pipeline_last_stage()): -# # return -# # if not (tp_rank == 0 and dp_rank == 0): -# # return -# if tensor is not None: -# orig_tensor = tensor -# if hasattr(tensor, "_hp_param"): -# numel = tensor._hp_param.numel() # // dp_size -# tensor = tensor.flatten().narrow(0, 0, numel) -# # print(fn) -# with open(fn, "w") as fh: -# fh.write(f"{get_fingerprint_header()}\n") -# if tensor is not None: -# fh.write(f"{get_fingerprint(tensor)} tensor {tensor.shape}\n") -# else: -# for n, p in model[0].named_parameters(): -# fh.write(f"{get_fingerprint(p)} {n} {p.shape}\n") -# return -# # until we figure out how to dump the actual fp32 values don't do this -# fn = f"debug-fp32-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" -# with open(fn, "w") as fh: -# fh.write(f"{get_fingerprint_header()}\n") -# if tensor is not None: -# tensor = orig_tensor -# if hasattr(tensor, "_hp_param"): -# fh.write( -# f"{get_fingerprint(tensor._hp_param)} tensor {tensor._hp_param.shape}\n" -# ) -# # fh.write(f"{get_fingerprint(tensor._hp_grad)} tensor grad\n") -# else: -# fh.write(f"{get_fingerprint(tensor)} tensor {tensor.shape}\n") -# # fh.write(f"{get_fingerprint(tensor.grad)} tensor grad\n") -# else: -# if hasattr(model[0].module.tied_modules, "embed"): -# p = model[0].module.tied_modules.embed.word_embeddings.weight._hp_param -# <<<<<<< HEAD -# fh.write( -# f"{get_fingerprint(p)} module.tied_modules.embed.word_embeddings.weight._hp_param {p.shape}\n" -# ) -# ======= -# fh.write(f"{get_fingerprint(p)} module.tied_modules.embed.word_embeddings.weight._hp_param {p.shape}\n") -# - - def dump_weights(preamble, iteration, model, optimizer, tensor=None): # return tp_rank = mpu.get_tensor_model_parallel_rank() @@ -578,19 +532,19 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): fn = f"debug-bf16-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" # only care for first and last pp stages and dp0 tp0 - #if not (mpu.is_pipeline_first_stage() or mpu.is_pipeline_last_stage()): + # if not (mpu.is_pipeline_first_stage() or mpu.is_pipeline_last_stage()): # return - #if not (tp_rank == 0 and dp_rank == 0): + # if not (tp_rank == 0 and dp_rank == 0): # return if tensor is not None: orig_tensor = tensor if hasattr(tensor, "_hp_param"): - numel = tensor._hp_param.numel() # // dp_size + numel = tensor._hp_param.numel() # // dp_size tensor = tensor.flatten().narrow(0, 0, numel) - #print(fn) + # print(fn) with open(fn, "w") as fh: fh.write(f"{get_fingerprint_header()}\n") @@ -600,7 +554,6 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): for n, p in model[0].named_parameters(): fh.write(f"{get_fingerprint(p)} {n} {p.shape}\n") - # # # until we figure out how to dump the actual fp32 values don't do this # fn = f"debug-fp32-{iteration}-pp{pp_rank}-tp{tp_rank}-dp{dp_rank}-{preamble}.txt" # with open(fn, "w") as fh: From 7f71572ab0d916dcd0989e9fad014391c04912e1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 13 Sep 2024 12:24:06 -0500 Subject: [PATCH 323/590] Update `megatron/training_log.py` --- megatron/training_log.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index be2eacd0595..cd6638e17d8 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -397,7 +397,7 @@ def training_log( opt_stats, group=mpu.get_pipeline_model_parallel_group() ) # opt_stats_2 = opt_stats_2.clone().detach() - opt_stats_2 = accelerator.get_accelerator().FloatTensor(opt_stats_2) + opt_stats_2 = accelerator.FloatTensor(opt_stats_2) torch.distributed.all_reduce( opt_stats_2, op=torch.distributed.ReduceOp.MAX, From 7cb9c1137f3bee4e84071c2b58fffdd301b44c54 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 14 Sep 2024 21:30:23 -0500 Subject: [PATCH 324/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 336 ++++++++++++++++++++++--------------------- 1 file changed, 171 insertions(+), 165 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 04018d7918d..4a6c3453da6 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -2,7 +2,9 @@ """Pretrain GPT""" import time +from typing import Callable, Type from mpi4py import MPI + comm = MPI.COMM_WORLD comm.Barrier() python_start_time = time.time() @@ -14,6 +16,7 @@ import math from functools import partial from megatron import get_args + # from megatron import print_rank_0 from megatron import get_timers from megatron import get_tokenizer @@ -23,14 +26,19 @@ from megatron.model import GPTModel, GPTModelPipe from megatron.training import pretrain from megatron.utils import get_ltor_masks_and_position_ids -from megatron.utils import average_losses_across_data_parallel_group, update_rotary_pos_emb +from megatron.utils import ( + average_losses_across_data_parallel_group, + update_rotary_pos_emb, +) from megatron.arguments import core_transformer_config_from_args + # from megatron.utils import Profile, PerfTrace import logging import deepspeed from deepspeed.runtime.utils import see_memory_usage + # from deepspeed.accelerator.real_accelerator import get_accelerator import subprocess import wandb @@ -38,7 +46,8 @@ from torch import nn import torch.nn.functional as F import ezpz as ez -dt_imports = time.time() - python_start_time + +dt_imports = time.time() - python_start_time t0_setup = time.time() # ---- [SETUP COMMS] ------------------------ @@ -62,19 +71,12 @@ log.info(f"ez.setup_torch time: {dt_setup} seconds") # ---- [SETUP WANDB FROM RANK 0] -------------- -WANDB_MODE = os.environ.get('WANDB_MODE', None) -DISABLE_WANDB = ( - WANDB_MODE is not None and str(WANDB_MODE).lower() == 'disabled' -) +WANDB_MODE = os.environ.get("WANDB_MODE", None) +DISABLE_WANDB = WANDB_MODE is not None and str(WANDB_MODE).lower() == "disabled" if RANK == 0 and not DISABLE_WANDB: - project_name = ( - os.environ.get( - 'WB_PROJECT', # look for WB_PROJECT in env - os.environ.get( - 'WANDB_PROJECT', # look for WANDB_PROJECT in env - 'AuroraGPT' - ), - ) + project_name = os.environ.get( + "WB_PROJECT", # look for WB_PROJECT in env + os.environ.get("WANDB_PROJECT", "AuroraGPT"), # look for WANDB_PROJECT in env ) log.info(f"Setting up W&B from: {RANK} with {project_name}") _ = ez.setup_wandb(project_name=project_name) @@ -83,16 +85,16 @@ @ez.dist.timeitlogit(rank=RANK) def model_provider(pre_process=True, post_process=True): """Build the model.""" - log.info('building GPT model ...') + log.info("building GPT model ...") see_memory_usage("Before Building Model", force=True) args = get_args() assert args is not None config = core_transformer_config_from_args(args) # if RANK == 0: # git_ds_info() - if hasattr(mpu, 'get_sequence_data_parallel_group'): + if hasattr(mpu, "get_sequence_data_parallel_group"): dpg = mpu.get_sequence_data_parallel_group() - elif hasattr(mpu, 'get_data_parallel_group'): + elif hasattr(mpu, "get_data_parallel_group"): dpg = mpu.get_data_parallel_group() else: dpg = None @@ -100,20 +102,14 @@ def model_provider(pre_process=True, post_process=True): if args.use_mics: deepspeed_zero_init = deepspeed.zero.MiCS_Init with deepspeed_zero_init( - data_parallel_group=dpg, - remote_device=( - None if args.remote_device == 'none' else args.remote_device - ), - config_dict_or_path=args.deepspeed_config_dict, - enabled=args.zero_stage == 3, - mpu=mpu + data_parallel_group=dpg, + remote_device=(None if args.remote_device == "none" else args.remote_device), + config_dict_or_path=args.deepspeed_config_dict, + enabled=args.zero_stage == 3, + mpu=mpu, ): if args.deepspeed and not args.no_pipeline_parallel: - model = GPTModelPipe( - config=config, - num_tokentypes=0, - parallel_output=True - ) + model = GPTModelPipe(config=config, num_tokentypes=0, parallel_output=True) # This is a hack to give us a reference to # get_batch_pipe from within training.py # We need to call model.set_batch_fn after deepspeed.initialize @@ -129,7 +125,7 @@ def model_provider(pre_process=True, post_process=True): ) ).view(1, 1, args.seq_length, args.seq_length) # Convert attention mask to binary: - attention_mask = (attention_mask < 0.5) + attention_mask = attention_mask < 0.5 if args.fp16: attention_mask = attention_mask.half() elif args.bf16: @@ -146,37 +142,33 @@ def model_provider(pre_process=True, post_process=True): num_tokentypes=0, parallel_output=True, pre_process=pre_process, - post_process=post_process + post_process=post_process, ) num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) - log.info(80 * '-') + log.info(80 * "-") log.info(f"Number of parameters in model: {num_params}") - log.info(80 * '-') + log.info(80 * "-") see_memory_usage("After Building Model", force=True) - if wandb is not None and getattr(wandb, 'run', None) is not None: + if wandb is not None and getattr(wandb, "run", None) is not None: assert wandb.run is not None tbdir = args.tensorboard_dir # tbdir = args.getattr('tensorboard_dir', None) if tbdir is not None: try: - log.info(f'Patching tensorboard from {tbdir}') + log.info(f"Patching tensorboard from {tbdir}") wandb.tensorboard.patch(root_logdir=tbdir) except ValueError as exc: log.exception(exc) - log.warning('Continuing without patching tensorboard!') - wandb.run.config.update({'num_params': num_params}) + log.warning("Continuing without patching tensorboard!") + wandb.run.config.update({"num_params": num_params}) if "args" not in wandb.run.config: log.info( f"Updating WandB run.config: [{wandb.run.name}]({wandb.run.get_url()})" ) try: - wandb.run.config.update( - {"args": dict(sorted(vars(args).items()))} - ) + wandb.run.config.update({"args": dict(sorted(vars(args).items()))}) except Exception: - log.error( - 'Unable to `wandb.run.config.update({"args": vars(args)})`' - ) + log.error('Unable to `wandb.run.config.update({"args": vars(args)})`') # try: # wandb.run.watch( # model, @@ -194,7 +186,7 @@ def get_batch(data_iterator): tokenizer = get_tokenizer() assert args is not None and tokenizer is not None # Items and their type. - keys = ['text'] + keys = ["text"] datatype = torch.int64 data = next(data_iterator) if data_iterator is not None else None # # Broadcast data. @@ -204,7 +196,7 @@ def get_batch(data_iterator): # data = None data_b = tensor_parallel.broadcast_data(keys, data, datatype) # Unpack. - tokens_ = data_b['text'].long() + tokens_ = data_b["text"].long() labels = tokens_[:, 1:].contiguous() tokens = tokens_[:, :-1].contiguous() # Get the masks and postition ids. @@ -215,7 +207,8 @@ def get_batch(data_iterator): args.reset_position_ids, args.reset_attention_mask, args.eod_mask_loss, - skip_mask) + skip_mask, + ) # For DS's sequence parallel seq_parallel_world_size = mpu.get_sequence_parallel_world_size() seq_parallel_world_rank = mpu.get_sequence_parallel_rank() @@ -240,24 +233,37 @@ def data_post_process(data, data_sampler_state_dict): args = get_args() assert args is not None if args.data_efficiency_curriculum_learning: - if 'seqlen_truncate' in data_sampler_state_dict['current_difficulties']: - args.data_efficiency_curriculum_learning_seqlen_type = 'seqlen_truncate' - current_seqlen = data_sampler_state_dict['current_difficulties']['seqlen_truncate'] + if "seqlen_truncate" in data_sampler_state_dict["current_difficulties"]: + args.data_efficiency_curriculum_learning_seqlen_type = "seqlen_truncate" + current_seqlen = data_sampler_state_dict["current_difficulties"][ + "seqlen_truncate" + ] if current_seqlen < args.seq_length: - data['text'] = data['text'][:, :(current_seqlen+1)].contiguous() - elif 'seqlen_reshape' in data_sampler_state_dict['current_difficulties']: - args.data_efficiency_curriculum_learning_seqlen_type = 'seqlen_reshape' - current_seqlen = data_sampler_state_dict['current_difficulties']['seqlen_reshape'] + data["text"] = data["text"][:, : (current_seqlen + 1)].contiguous() + elif "seqlen_reshape" in data_sampler_state_dict["current_difficulties"]: + args.data_efficiency_curriculum_learning_seqlen_type = "seqlen_reshape" + current_seqlen = data_sampler_state_dict["current_difficulties"][ + "seqlen_reshape" + ] if current_seqlen < args.seq_length: - orig_num_token = torch.numel(data['text']) - reshape_len = (data['text'].size()[1] // (current_seqlen+1)) * (current_seqlen+1) - data['text'] = torch.cat((data['text'][:, :reshape_len].contiguous().view(-1, current_seqlen+1), - data['text'][:, -(current_seqlen+1):]), 0).contiguous() - num_row = math.ceil(orig_num_token / (current_seqlen+1)) - num_row = min(num_row, data['text'].size()[0]) + orig_num_token = torch.numel(data["text"]) + reshape_len = (data["text"].size()[1] // (current_seqlen + 1)) * ( + current_seqlen + 1 + ) + data["text"] = torch.cat( + ( + data["text"][:, :reshape_len] + .contiguous() + .view(-1, current_seqlen + 1), + data["text"][:, -(current_seqlen + 1) :], + ), + 0, + ).contiguous() + num_row = math.ceil(orig_num_token / (current_seqlen + 1)) + num_row = min(num_row, data["text"].size()[0]) if num_row > 1 and num_row % 2 != 0: num_row -= 1 - data['text'] = data['text'][:num_row, :].contiguous() + data["text"] = data["text"][:num_row, :].contiguous() else: args.data_efficiency_curriculum_learning_seqlen_type = None return data @@ -272,12 +278,12 @@ def get_batch_pipe(data): tokenizer = get_tokenizer() assert args is not None # Items and their type. - keys = ['text'] + keys = ["text"] datatype = torch.int64 # Broadcast data. data_b = tensor_parallel.broadcast_data(keys, data, datatype) # Unpack. - tokens_ = data_b['text'].long() + tokens_ = data_b["text"].long() labels = tokens_[:, 1:].contiguous() tokens = tokens_[:, :-1].contiguous() # Get the masks and postition ids. @@ -286,19 +292,17 @@ def get_batch_pipe(data): tokenizer.eod, args.reset_position_ids, args.reset_attention_mask, - args.eod_mask_loss) - if ( - args.curriculum_learning_legacy - and args.curriculum_seqlen < tokens.size()[1] - ): + args.eod_mask_loss, + ) + if args.curriculum_learning_legacy and args.curriculum_seqlen < tokens.size()[1]: # seqlen-based curriculum learning # tokens, position_ids, labels, loss_mask # have size [batch size, seqlen] - tokens = tokens[:, :args.curriculum_seqlen].contiguous() - position_ids = position_ids[:, :args.curriculum_seqlen].contiguous() + tokens = tokens[:, : args.curriculum_seqlen].contiguous() + position_ids = position_ids[:, : args.curriculum_seqlen].contiguous() if labels is not None: - labels = labels[:, :args.curriculum_seqlen].contiguous() - loss_mask = loss_mask[:, :args.curriculum_seqlen].contiguous() + labels = labels[:, : args.curriculum_seqlen].contiguous() + loss_mask = loss_mask[:, : args.curriculum_seqlen].contiguous() return (tokens, position_ids, attention_mask), (labels, loss_mask) @@ -315,37 +319,32 @@ def loss_func(loss_mask, moe_loss, mos_loss, output_tensor): loss = loss + moe_loss + mos_loss if args.mos: return loss, { - 'total loss': loss, - 'lm loss': averaged_loss[0], - 'moe loss': moe_loss, - 'mos loss': mos_loss + "total loss": loss, + "lm loss": averaged_loss[0], + "moe loss": moe_loss, + "mos loss": mos_loss, } elif args.kd: return loss, { - 'total loss': loss, - 'lm loss': averaged_loss[0], - 'moe loss': moe_loss, - 'kd loss': mos_loss + "total loss": loss, + "lm loss": averaged_loss[0], + "moe loss": moe_loss, + "kd loss": mos_loss, } log.info( - f'>>> total loss: {loss}, ' - f'lm loss {averaged_loss[0]}, ' - f'kd loss {mos_loss}' + f">>> total loss: {loss}, " + f"lm loss {averaged_loss[0]}, " + f"kd loss {mos_loss}" ) else: if max(args.num_experts) <= 1: - return loss, {'lm loss': averaged_loss[0]} + return loss, {"lm loss": averaged_loss[0]} loss = loss + moe_loss - return loss, {'lm loss': averaged_loss[0], 'moe loss': moe_loss} + return loss, {"lm loss": averaged_loss[0], "moe loss": moe_loss} def calculate_mos_loss( - args, - stu_output, - teacher_model, - tokens, - position_ids, - attention_mask + args, stu_output, teacher_model, tokens, position_ids, attention_mask ): mos_loss = 0 alpha = args.kd_alpha_ce @@ -354,29 +353,25 @@ def calculate_mos_loss( if teacher_model: with torch.no_grad(): if ( - args.curriculum_learning_legacy and - args.curriculum_seqlen < args.seq_length + args.curriculum_learning_legacy + and args.curriculum_seqlen < args.seq_length ): assert args.curriculum_seqlen is not None curriculum_seqlen = args.curriculum_seqlen tokens = tokens[:, :curriculum_seqlen].contiguous() position_ids = position_ids[:, :curriculum_seqlen].contiguous() csl = curriculum_seqlen - attention_mask = ( - attention_mask[:, :, :csl, :csl].contiguous() - ) + attention_mask = attention_mask[:, :, :csl, :csl].contiguous() # No need to truncate labels # as we do not need it for the teacher logits tea_output, tea_other_losses = teacher_model( - tokens, - position_ids, - attention_mask + tokens, position_ids, attention_mask ) assert stu_output.size() == tea_output.size(), ( - 'teacher and student output should match in size. ' - f'Student: {stu_output.size()}, ' - f'Teacher: {tea_output.size()}, ' - f'CL seq length {args.curriculum_seqlen}' + "teacher and student output should match in size. " + f"Student: {stu_output.size()}, " + f"Teacher: {tea_output.size()}, " + f"CL seq length {args.curriculum_seqlen}" ) student_logits = F.log_softmax(stu_output / kd_temp, dim=2) # The target logits is expected to be probabilities. @@ -384,67 +379,81 @@ def calculate_mos_loss( # then we need to set target_log to true # when initializing the KLDivLoss. tea_logits = F.softmax(tea_output / kd_temp, dim=2) - mos_loss = kd_temp * kd_temp * nn.KLDivLoss(reduction='batchmean')( - student_logits, - tea_logits + mos_loss = ( + kd_temp + * kd_temp + * nn.KLDivLoss(reduction="batchmean")(student_logits, tea_logits) ) mos_loss = mos_loss.div(args.seq_length) * beta return mos_loss -def forward_step(data_iterator, model): +# ForwardStepOutput = Type[tuple[torch.Tensor | None, Callable[[torch.Tensor], torch.Tensor | None]]] + + +def _return_none(_: torch.Tensor) -> torch.Tensor | None: + return None + + +def forward_step(data_iterator, model) -> tuple[torch.Tensor | None, Callable]: """Forward step.""" args = get_args() timers = get_timers() assert args is not None assert timers is not None # Get the batch. - timers('batch-generator', log_level=2).start() - tokens, labels, loss_mask, attention_mask, position_ids = get_batch( - data_iterator - ) - timers('batch-generator').stop() + timers("batch-generator", log_level=2).start() + tokens, labels, loss_mask, attention_mask, position_ids = get_batch(data_iterator) + timers("batch-generator").stop() + ranges_to_skip = None + if args.train_range_to_skip is not None: + assert ( + len(args.train_range_to_skip) % 2 == 0 + ), f"""Expected --train-range-to-skip to have an even number of values. + Received: {len(args.train_range_to_skip)} + """ + ranges_to_skip = list( + zip( + args.train_range_to_skip[::2], + args.train_range_to_skip[1::2], + ) + ) + if ranges_to_skip is not None and any( + [i <= (args.iteration + 1) <= j for (i, j) in ranges_to_skip] + ): + log.info( + f"Caught {args.iteration} in 'forward_step', {tokens.shape()=}, {args.consumed_train_tokens=}'" + ) + # log.info(f"Caught {args.iteration + 1} in 'ranges_to_skip', skipping!" + # return (None, _return_none) + return ( + torch.tensor([0.0], device=tokens.device), + lambda _: torch.Tensor([0.0], device=tokens.device), + # lambda _: return torch.Tensor([0.0], deviec=tokens.device), + ) if args.data_efficiency_curriculum_learning: args.curriculum_seqlen = tokens.size()[1] - if ( - hasattr( - args, - 'data_efficiency_curriculum_learning_seqlen_type') - and ( - args.data_efficiency_curriculum_learning_seqlen_type - == 'seqlen_reshape' - ) + if hasattr(args, "data_efficiency_curriculum_learning_seqlen_type") and ( + args.data_efficiency_curriculum_learning_seqlen_type == "seqlen_reshape" ): - args.data_efficiency_curriculum_learning_numel = ( - torch.numel(tokens) - ) + args.data_efficiency_curriculum_learning_numel = torch.numel(tokens) stu_output = None if args.mos or args.kd: # The forward func can return either the loss or the logits, # depending on whether passing in the labels or not. stu_output, other_losses = model(tokens, position_ids, attention_mask) - if ( - args.curriculum_learning_legacy - and args.curriculum_seqlen < args.seq_length - ): + if args.curriculum_learning_legacy and args.curriculum_seqlen < args.seq_length: assert args.curriculum_seqlen is not None - labels = labels[:, :args.curriculum_seqlen].contiguous() + labels = labels[:, : args.curriculum_seqlen].contiguous() output_tensor = tensor_parallel.vocab_parallel_cross_entropy( - stu_output.contiguous().float(), - labels + stu_output.contiguous().float(), labels ) else: output_tensor, other_losses = model( - tokens, - position_ids, - attention_mask, - labels=labels + tokens, position_ids, attention_mask, labels=labels ) - if ( - args.curriculum_learning_legacy and - args.curriculum_seqlen < args.seq_length - ): - loss_mask = loss_mask[:, :args.curriculum_seqlen].contiguous() + if args.curriculum_learning_legacy and args.curriculum_seqlen < args.seq_length: + loss_mask = loss_mask[:, : args.curriculum_seqlen].contiguous() moe_losses = [] for moe_loss in other_losses: @@ -462,7 +471,7 @@ def forward_step(data_iterator, model): args.teacher_model[0], tokens, position_ids, - attention_mask + attention_mask, ) # Output_tensor stores the standard loss, @@ -479,7 +488,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): # from ezpz.profile import get_context_manager # cm = get_context_manager(rank=RANK, outdir=args.save) # with cm: - log.info('> building train, validation, and test datasets for GPT ...') + log.info("> building train, validation, and test datasets for GPT ...") files = [] if args.data_file_list is not None: log.info(f"Reading datasets from {args.data_file_list}") @@ -492,7 +501,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): # - `/path/to/data_text_document` is the path to the text document # - `corpus` is the corpus (~ source, can be made up) where that # document came from (i.e. `books`, `arxiv`, etc.) - with open(args.data_file_list, 'r') as flist: + with open(args.data_file_list, "r") as flist: for f in flist.readlines(): if len(f.strip()) != 0: try: @@ -505,17 +514,11 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): ) if fname.find(".bin") != -1: fname = fname.split(".bin")[0] - files.extend( - [ - float(w), # weight - fname, # filename - c # corpus - ] - ) + files.extend([float(w), fname, c]) # weight # filename # corpus elif len(args.data_path) == 1 and os.path.isdir(args.data_path[0]): path = args.data_path[0] + "/" for f in os.listdir(path): - if (os.path.isfile(path + f) and f.find(".bin") != -1): + if os.path.isfile(path + f) and f.find(".bin") != -1: files.append(1) files.append(path + f.split(".bin")[0]) else: @@ -540,11 +543,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): def command_exists(cmd): - result = subprocess.Popen( - f'type {cmd}', - stdout=subprocess.PIPE, - shell=True - ) + result = subprocess.Popen(f"type {cmd}", stdout=subprocess.PIPE, shell=True) return result.wait() == 0 @@ -552,17 +551,18 @@ def git_ds_info(): if RANK != 0: return from deepspeed.env_report import main as ds_report + ds_report() # Write out version/git info git_hash_cmd = "git rev-parse --short HEAD" git_branch_cmd = "git rev-parse --abbrev-ref HEAD" - if command_exists('git'): + if command_exists("git"): try: result = subprocess.check_output(git_hash_cmd, shell=True) - git_hash = result.decode('utf-8').strip() + git_hash = result.decode("utf-8").strip() result = subprocess.check_output(git_branch_cmd, shell=True) - git_branch = result.decode('utf-8').strip() + git_branch = result.decode("utf-8").strip() except subprocess.CalledProcessError: git_hash = "unknown" git_branch = "unknown" @@ -570,21 +570,26 @@ def git_ds_info(): git_hash = "unknown" git_branch = "unknown" print( - f'**** Git info for Megatron: ' - f'git_hash={git_hash} git_branch={git_branch} ****' + f"**** Git info for Megatron: " + f"git_hash={git_hash} git_branch={git_branch} ****" ) def main(): - if os.getenv('TORCH_PROFILER_ENABLE') == '1': + if os.getenv("TORCH_PROFILER_ENABLE") == "1": # record_function from torch.profiler import profile, ProfilerActivity + try: - activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA, ProfilerActivity.XPU] + activities = [ + ProfilerActivity.CPU, + ProfilerActivity.CUDA, + ProfilerActivity.XPU, + ] except Exception as exc: log.exception(exc) - log.warning("TORCH PROFILER WARNING: XPU is not supported") - activities=[ProfilerActivity.CPU, ProfilerActivity.CUDA] + log.warning("TORCH PROFILER WARNING: XPU is not supported") + activities = [ProfilerActivity.CPU, ProfilerActivity.CUDA] with profile(activities=activities) as prof: model = pretrain( train_valid_test_datasets_provider, @@ -592,7 +597,7 @@ def main(): ModelType.encoder_or_decoder, forward_step, # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, - data_post_process=data_post_process + data_post_process=data_post_process, ) args = get_args() assert args is not None @@ -606,7 +611,7 @@ def main(): ModelType.encoder_or_decoder, forward_step, # args_defaults={'tokenizer_type': 'GPT2BPETokenizer'}, - data_post_process=data_post_process + data_post_process=data_post_process, ) # try: # from megatron.text_generation import generate_and_post_process @@ -641,6 +646,7 @@ def main(): # data_post_process=data_post_process) import sys import deepspeed.comm as dist + model = main() dist.log_summary() if wandb.run is not None: From e83de19527830ade3a49f3fcfb698739a84c4ac9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 14 Sep 2024 21:33:05 -0500 Subject: [PATCH 325/590] Update `megatron/training_log.py` --- megatron/training_log.py | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index cd6638e17d8..24b8015264d 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -645,6 +645,12 @@ def training_log( ) log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) log_string += " [DS]TFLOPs={:.2f} |".format(tflops) + if wandb is not None and getattr(wandb, "run", None) is not None: + wandb_metrics |= { + "training/skiped_iterations": total_loss_dict[skipped_iters_key] + } + wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} + wandb.log(wandb_metrics) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 @@ -654,12 +660,6 @@ def training_log( # Report memory after optimizer state has been initialized. report_memory("(after {} iterations)".format(iteration)) report_memory_flag = False - if wandb is not None and getattr(wandb, "run", None) is not None: - wandb_metrics |= { - "training/skiped_iterations": total_loss_dict[skipped_iters_key] - } - wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} - wandb.log(wandb_metrics) if timers is not None: timers.log(timers_to_log, normalizer=args.log_interval) From 29756d6e1e97464e8a7fbe1eca4cf832fa2a916a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 14 Sep 2024 21:34:33 -0500 Subject: [PATCH 326/590] Warn if mismatch b/w iters in `megatron/checkpointing.py` --- megatron/checkpointing.py | 4 +++- megatron/training.py | 17 ++++++++++++++--- 2 files changed, 17 insertions(+), 4 deletions(-) diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index b7f4b30bde8..a4f82ec9d3a 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -239,7 +239,9 @@ def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): """Save a model checkpoint.""" args = get_args() assert args is not None - iteration = args.iteration + args_iter = args.iteration + if args_iter != iteration: + log.warning(f"{args.iteration=} != {iteration} passed to 'save_checkpoint'") save_lr_state_dict() diff --git a/megatron/training.py b/megatron/training.py index 90a1250648d..668aea930cf 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1004,7 +1004,7 @@ def train( model_module.train() # Tracking loss. total_loss_dict = {} - loss_dict = {} + loss_dict = {"skipped_iter": 0} # Iterations. iteration = args.iteration # Translate args to core configuration @@ -1061,11 +1061,22 @@ def train( ): log.info(f"Caught {iteration + 1} in 'ranges_to_skip', skipping!") # total_loss_dict = {"skipped iterations": } + # loss_dict skipped_iter = 1 - total_loss_dict["skipped iterations"] += skipped_iter - grad_norm = None + # grad_norm = None num_zeros_in_grad = None num_skipped_iters += 1 + increment = ( + get_num_microbatches() * args.micro_batch_size * args.data_parallel_size + ) + model[0].skipped_steps += 1 + model[0].global_steps += 1 + model[0].micro_steps += 1 + model[0].global_samples += model[0].train_batch_size() + # model[0].step(lr_kwargs={"increment": increment}) + # grad_norm = model[0].get_global_grad_norm() + # update_successful = model[0].was_step_applied() + opt_param_scheduler.step(increment=increment) else: if os.getenv("TORCH_PROFILER_ENABLE") == "2": from torch.profiler import profile, ProfilerActivity From 1a7f03b67a260e2e7326fe3a0f12709a8908d616 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 15 Sep 2024 20:39:24 -0500 Subject: [PATCH 327/590] fix: `try/except` for non tensors in `megatron/training_log.py` --- megatron/training_log.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index 24b8015264d..3eb96c392d9 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -92,7 +92,10 @@ def training_log( + loss_dict[key] ) else: - value = loss_dict[key].float().sum().item() + try: + value = loss_dict[key].float().sum().item() + except AttributeError: + value = loss_dict[key] is_nan = value == float("inf") or value == -float("inf") or value != value got_nan = got_nan or is_nan total_loss_dict[nan_iters_key] = total_loss_dict.get(nan_iters_key, 0) + int( From 828f6a944627335336f7f6d2e348dd3989ffb021 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Sep 2024 08:17:17 -0500 Subject: [PATCH 328/590] fix: Correctly draw `grad_acc_steps` batches of data when skipping step --- megatron/training.py | 31 +++++++++++++++++++++++-------- 1 file changed, 23 insertions(+), 8 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 668aea930cf..8ffac6cb9c2 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -822,6 +822,8 @@ def train_step( timers = get_timers() accelerator = get_accelerator() assert args is not None and timers is not None and accelerator is not None + grad_norm = None + num_zeros_in_grad = None if args.deepspeed and args.ds_pipeline_enabled: num_zeros_in_grad = 0 assert isinstance(model[0], deepspeed.PipelineEngine) @@ -919,6 +921,10 @@ def train_step( if args.deepspeed: skipped_iter = 0 if update_successful else 1 grad_norm = model[0].get_global_grad_norm() + # Empty unused memory. + if args.empty_unused_memory_level >= 2 and accelerator is not None: + accelerator.empty_cache() + # XXX: [saforem2]: ---------------------------------------------------- # Is `num_zeros_in_grad` worth calculating (/ implementing) ?? # the `Megatron`-specific implementation is at: @@ -1002,6 +1008,7 @@ def train( # Turn on training mode which enables dropout. for model_module in model: model_module.train() + grad_norm = None # Tracking loss. total_loss_dict = {} loss_dict = {"skipped_iter": 0} @@ -1060,12 +1067,23 @@ def train( [i <= (iteration + 1) <= j for (i, j) in ranges_to_skip] ): log.info(f"Caught {iteration + 1} in 'ranges_to_skip', skipping!") - # total_loss_dict = {"skipped iterations": } - # loss_dict skipped_iter = 1 - # grad_norm = None - num_zeros_in_grad = None num_skipped_iters += 1 + num_zeros_in_grad = None + gas = args.deepspeed_config_dict["gradient_accumulation_steps"] + for microstep in range(gas): + _batch = next(train_data_iterator) + _tokens = _batch["text"] + if ( + iteration < 10 + and os.environ.get("DUMP_SKIPPED_ITERS", None) + and RANK == 0 + ): + log.info(f"{_tokens.shape}, {len(train_data_iterator)=}") + log.info( + f"{iteration=} [{microstep}/{gas}]: ({_tokens.shape})\n{_tokens[:10]=}" + ) + increment = ( get_num_microbatches() * args.micro_batch_size * args.data_parallel_size ) @@ -1073,9 +1091,6 @@ def train( model[0].global_steps += 1 model[0].micro_steps += 1 model[0].global_samples += model[0].train_batch_size() - # model[0].step(lr_kwargs={"increment": increment}) - # grad_norm = model[0].get_global_grad_norm() - # update_successful = model[0].was_step_applied() opt_param_scheduler.step(increment=increment) else: if os.getenv("TORCH_PROFILER_ENABLE") == "2": @@ -1085,7 +1100,7 @@ def train( activities = [ ProfilerActivity.CPU, ProfilerActivity.CUDA, - ProfilerActivity.XPU, + ProfilerActivity.XPU, # type:ignore ] except Exception: log.warning("TORCH PROFILER WARNING: XPU is not supported") From 295fcb3d57a40ec513a521aa8814d99a5c8827b8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Sep 2024 08:18:32 -0500 Subject: [PATCH 329/590] Update `pretrain_gpt_alcf.py` Remve `--train-range-to-skip` logic from `pretrain_gpt_alcf.py` and remove redundant code. --- pretrain_gpt_alcf.py | 44 +++++++++----------------------------------- 1 file changed, 9 insertions(+), 35 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 4a6c3453da6..12a05c52998 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -2,13 +2,12 @@ """Pretrain GPT""" import time -from typing import Callable, Type +from typing import Callable from mpi4py import MPI comm = MPI.COMM_WORLD comm.Barrier() python_start_time = time.time() -from pathlib import Path import os from rich import print @@ -189,6 +188,14 @@ def get_batch(data_iterator): keys = ["text"] datatype = torch.int64 data = next(data_iterator) if data_iterator is not None else None + + if ( + args.iteration < 10 + and RANK == 0 + and os.environ.get("DUMP_TOKENS", None) + and data is not None + ): + log.info(f"{args.iteration=}: {data['text'][:10]=}") # # Broadcast data. # if data_iterator is not None: # data = next(data_iterator) @@ -388,13 +395,6 @@ def calculate_mos_loss( return mos_loss -# ForwardStepOutput = Type[tuple[torch.Tensor | None, Callable[[torch.Tensor], torch.Tensor | None]]] - - -def _return_none(_: torch.Tensor) -> torch.Tensor | None: - return None - - def forward_step(data_iterator, model) -> tuple[torch.Tensor | None, Callable]: """Forward step.""" args = get_args() @@ -405,32 +405,6 @@ def forward_step(data_iterator, model) -> tuple[torch.Tensor | None, Callable]: timers("batch-generator", log_level=2).start() tokens, labels, loss_mask, attention_mask, position_ids = get_batch(data_iterator) timers("batch-generator").stop() - ranges_to_skip = None - if args.train_range_to_skip is not None: - assert ( - len(args.train_range_to_skip) % 2 == 0 - ), f"""Expected --train-range-to-skip to have an even number of values. - Received: {len(args.train_range_to_skip)} - """ - ranges_to_skip = list( - zip( - args.train_range_to_skip[::2], - args.train_range_to_skip[1::2], - ) - ) - if ranges_to_skip is not None and any( - [i <= (args.iteration + 1) <= j for (i, j) in ranges_to_skip] - ): - log.info( - f"Caught {args.iteration} in 'forward_step', {tokens.shape()=}, {args.consumed_train_tokens=}'" - ) - # log.info(f"Caught {args.iteration + 1} in 'ranges_to_skip', skipping!" - # return (None, _return_none) - return ( - torch.tensor([0.0], device=tokens.device), - lambda _: torch.Tensor([0.0], device=tokens.device), - # lambda _: return torch.Tensor([0.0], deviec=tokens.device), - ) if args.data_efficiency_curriculum_learning: args.curriculum_seqlen = tokens.size()[1] if hasattr(args, "data_efficiency_curriculum_learning_seqlen_type") and ( From 598c0920b70a138ef808b0ca3637131ad7579588 Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Sat, 21 Sep 2024 06:42:59 +0800 Subject: [PATCH 330/590] grad_wei can't be NoneType when running with DeepSpeed, for zero3 will divided the gradient (#428) --- megatron/core/tensor_parallel/layers.py | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/megatron/core/tensor_parallel/layers.py b/megatron/core/tensor_parallel/layers.py index 67a78853aa6..3dd3299ae08 100644 --- a/megatron/core/tensor_parallel/layers.py +++ b/megatron/core/tensor_parallel/layers.py @@ -286,6 +286,7 @@ def forward(ctx, input, weight, bias, gradient_accumulation_fusion, @staticmethod @custom_bwd def backward(ctx, grad_output): + args = get_args() input, weight = ctx.saved_tensors use_bias = ctx.use_bias @@ -367,9 +368,13 @@ def backward(ctx, grad_output): # grad_weight = None # else: # grad_weight = grad_output.t().matmul(total_input) - from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore - WeightGradStore.put(total_input, grad_output, weight, gradientUpdateFunction) - grad_weight = None + if args.enable_zbh1_pipeline: + from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore + WeightGradStore.put(total_input, grad_output, weight, gradientUpdateFunction) + grad_weight = None + else: + grad_weight = grad_output.t().matmul(total_input) + grad_bias = grad_output.sum(dim=0) if use_bias else None if ctx.sequence_parallel: From cf80e6bb75c56cd8a26627e5364e4290da736bd8 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Mon, 23 Sep 2024 16:21:00 +0000 Subject: [PATCH 331/590] added sophia --- megatron/arguments.py | 10 ++ megatron/optimizer/__init__.py | 9 ++ megatron/optimizer/sophia.py | 202 +++++++++++++++++++++++++++++++++ 3 files changed, 221 insertions(+) create mode 100644 megatron/optimizer/sophia.py diff --git a/megatron/arguments.py b/megatron/arguments.py index b3ed06353e2..2a0ac606ce4 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -780,6 +780,15 @@ def _add_regularization_args(parser): help='Weight decay increment function.') group.add_argument('--clip-grad', type=float, default=1.0, help='Gradient clipping based on global L2 norm.') + group.add_argument('--sophiag-beta1', type=float, default=0.9, + help='First coefficient for computing running averages ' + 'of gradient and its hessian') + group.add_argument('--sophiag-beta2', type=float, default=0.95, + help='Second coefficient for computing running averages ' + 'of gradient and its hessian') + group.add_argument('--sophiag-rho', type=float, default=0.01, + help='SophiaG clipping threshhold') + group.add_argument('--adam-beta1', type=float, default=0.9, help='First coefficient for computing running averages ' 'of gradient and its square') @@ -946,6 +955,7 @@ def _add_training_args(parser): choices=[ 'adam', 'adamw', + 'sophiag', 'sgd', 'ds.fusedlamb', 'ipex.lamb', diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 48f2737a062..99145ff4f49 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -315,6 +315,15 @@ def optimizer_hook(p): weight_decay=args.weight_decay, momentum=args.sgd_momentum ) + elif str(args.optimizer).lower() == 'sophiag': + from .sophia import SophiaG + optimizer = SophiaG( + param_groups, + lr=args.lr, + betas=(args.sophiag_beta1, args.sophiag_beta2), + rho = args.sophiag_rho, + weight_decay=args.weight_decay + ) else: raise TypeError(f'{args.optimizer} optimizer is not supported.') if args.deepspeed: diff --git a/megatron/optimizer/sophia.py b/megatron/optimizer/sophia.py new file mode 100644 index 00000000000..4c4e0747907 --- /dev/null +++ b/megatron/optimizer/sophia.py @@ -0,0 +1,202 @@ +import math +import torch +from torch import Tensor +from torch.optim.optimizer import Optimizer +from typing import List, Optional + + +#SOphiaG implementation from https://github.com/Liuhong99/Sophia/blob/main/sophia.py, copy pasted here because no pip and not sure about submodules + +class SophiaG(Optimizer): + def __init__(self, params, lr=1e-4, betas=(0.965, 0.99), rho = 0.04, + weight_decay=1e-1, *, maximize: bool = False, + capturable: bool = False): + if not 0.0 <= lr: + raise ValueError("Invalid learning rate: {}".format(lr)) + if not 0.0 <= betas[0] < 1.0: + raise ValueError("Invalid beta parameter at index 0: {}".format(betas[0])) + if not 0.0 <= betas[1] < 1.0: + raise ValueError("Invalid beta parameter at index 1: {}".format(betas[1])) + if not 0.0 <= rho: + raise ValueError("Invalid rho parameter at index 1: {}".format(rho)) + if not 0.0 <= weight_decay: + raise ValueError("Invalid weight_decay value: {}".format(weight_decay)) + defaults = dict(lr=lr, betas=betas, rho=rho, + weight_decay=weight_decay, + maximize=maximize, capturable=capturable) + super(SophiaG, self).__init__(params, defaults) + + def __setstate__(self, state): + super().__setstate__(state) + for group in self.param_groups: + group.setdefault('maximize', False) + group.setdefault('capturable', False) + state_values = list(self.state.values()) + step_is_tensor = (len(state_values) != 0) and torch.is_tensor(state_values[0]['step']) + if not step_is_tensor: + for s in state_values: + s['step'] = torch.tensor(float(s['step'])) + + @torch.no_grad() + def update_hessian(self): + for group in self.param_groups: + beta1, beta2 = group['betas'] + for p in group['params']: + if p.grad is None: + continue + state = self.state[p] + + if len(state) == 0: + state['step'] = torch.zeros((1,), dtype=torch.float, device=p.device) \ + if self.defaults['capturable'] else torch.tensor(0.) + state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format) + state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) + + if 'hessian' not in state.keys(): + state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) + + state['hessian'].mul_(beta2).addcmul_(p.grad, p.grad, value=1 - beta2) + + + @torch.no_grad() + def step(self, closure=None, bs=5120): + loss = None + if closure is not None: + with torch.enable_grad(): + loss = closure() + + for group in self.param_groups: + params_with_grad = [] + grads = [] + exp_avgs = [] + state_steps = [] + hessian = [] + beta1, beta2 = group['betas'] + + for p in group['params']: + if p.grad is None: + continue + params_with_grad.append(p) + + if p.grad.is_sparse: + raise RuntimeError('Hero does not support sparse gradients') + grads.append(p.grad) + state = self.state[p] + # State initialization + if len(state) == 0: + state['step'] = torch.zeros((1,), dtype=torch.float, device=p.device) \ + if self.defaults['capturable'] else torch.tensor(0.) + state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format) + state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) + + if 'hessian' not in state.keys(): + state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) + + exp_avgs.append(state['exp_avg']) + state_steps.append(state['step']) + hessian.append(state['hessian']) + + if self.defaults['capturable']: + bs = torch.ones((1,), dtype=torch.float, device=p.device) * bs + + sophiag(params_with_grad, + grads, + exp_avgs, + hessian, + state_steps, + bs=bs, + beta1=beta1, + beta2=beta2, + rho=group['rho'], + lr=group['lr'], + weight_decay=group['weight_decay'], + maximize=group['maximize'], + capturable=group['capturable']) + + return loss + +def sophiag(params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + hessian: List[Tensor], + state_steps: List[Tensor], + capturable: bool = False, + *, + bs: int, + beta1: float, + beta2: float, + rho: float, + lr: float, + weight_decay: float, + maximize: bool): + + if not all(isinstance(t, torch.Tensor) for t in state_steps): + raise RuntimeError("API has changed, `state_steps` argument must contain a list of singleton tensors") + + + func = _single_tensor_sophiag + + func(params, + grads, + exp_avgs, + hessian, + state_steps, + bs=bs, + beta1=beta1, + beta2=beta2, + rho=rho, + lr=lr, + weight_decay=weight_decay, + maximize=maximize, + capturable=capturable) + +def _single_tensor_sophiag(params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + hessian: List[Tensor], + state_steps: List[Tensor], + *, + bs: int, + beta1: float, + beta2: float, + rho: float, + lr: float, + weight_decay: float, + maximize: bool, + capturable: bool): + + for i, param in enumerate(params): + grad = grads[i] if not maximize else -grads[i] + exp_avg = exp_avgs[i] + hess = hessian[i] + step_t = state_steps[i] + + if capturable: + assert param.is_cuda and step_t.is_cuda and bs.is_cuda + + if torch.is_complex(param): + grad = torch.view_as_real(grad) + exp_avg = torch.view_as_real(exp_avg) + hess = torch.view_as_real(hess) + param = torch.view_as_real(param) + + # update step + step_t += 1 + + # Perform stepweight decay + param.mul_(1 - lr * weight_decay) + + # Decay the first and second moment running average coefficient + exp_avg.mul_(beta1).add_(grad, alpha=1 - beta1) + + if capturable: + step_size = lr + step_size_neg = step_size.neg() + + ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None,1) + param.addcmul_(exp_avg.sign(), ratio, value=step_size_neg) + else: + step_size_neg = - lr + + ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None,1) + param.addcmul_(exp_avg.sign(), ratio, value=step_size_neg) From 8be7f4895968574d44ae06785fdbe115baa1c13c Mon Sep 17 00:00:00 2001 From: YiSheng5 Date: Fri, 4 Oct 2024 18:01:49 +0800 Subject: [PATCH 332/590] fix init issue for rms_norm in squence_parallel (#448) --- megatron/model/fused_rmsnorm.py | 10 ++++++++-- megatron/model/rmsnorm.py | 5 +++-- 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/megatron/model/fused_rmsnorm.py b/megatron/model/fused_rmsnorm.py index a3b9927c874..ab1c42e24e9 100644 --- a/megatron/model/fused_rmsnorm.py +++ b/megatron/model/fused_rmsnorm.py @@ -1,13 +1,19 @@ +from megatron import get_args + import torch from torch.nn.parameter import Parameter +from torch.nn import init import intel_extension_for_pytorch as ipex # noqa # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__(self, dim: int, eps: float = 1e-6): + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel=False): super().__init__() self.eps = eps - self.weight = Parameter(torch.ones(dim)) + self.weight = Parameter(torch.ones(dim, + dtype=get_args().params_dtype)) + self.sequence_parallel = sequence_parallel + setattr(self.weight, 'sequence_parallel', self.sequence_parallel) def forward(self, x): output = torch.xpu.IpexRmsNorm(x, self.weight.shape, self.weight, self.eps) diff --git a/megatron/model/rmsnorm.py b/megatron/model/rmsnorm.py index 4860d81716b..7bcaec37eff 100644 --- a/megatron/model/rmsnorm.py +++ b/megatron/model/rmsnorm.py @@ -9,7 +9,7 @@ # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__(self, dim: int, eps: float = 1e-6): + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel=False): super().__init__() self.eps = eps init_device = None @@ -19,7 +19,8 @@ def __init__(self, dim: int, eps: float = 1e-6): device=init_device, dtype=get_args().params_dtype)) init.ones_(self.weight) - setattr(self.weight, 'sequence_parallel', sequence_parallel) + self.sequence_parallel = sequence_parallel + setattr(self.weight, 'sequence_parallel', self.sequence_parallel) def _norm(self, x): return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) From 444849272a9a39048baee8e6baf98d41d0b0d76e Mon Sep 17 00:00:00 2001 From: ranzhejiang Date: Tue, 8 Oct 2024 22:02:46 +0800 Subject: [PATCH 333/590] enable profiler for specific ranks (#451) --- megatron/arguments.py | 6 ++++++ megatron/profiler.py | 4 +++- 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 9228da6ee9f..6580dba80a8 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1570,5 +1570,11 @@ def _add_profiler_args(parser): type=str, default='2,3', help="Which steps to profile. Format: ,") + + group.add_argument("--profile-ranks", + type=int, + nargs='+', + default=None, + help="Which ranks to profile. Format: 0 1 2 3") return parser diff --git a/megatron/profiler.py b/megatron/profiler.py index c98096482ab..aeab1448466 100644 --- a/megatron/profiler.py +++ b/megatron/profiler.py @@ -36,7 +36,9 @@ def is_end_step(): def is_capture_step(): return cur_step >= start_step and cur_step <= end_step - if args.profile.startswith('pt'): + if args.profile.startswith('pt') and ( + args.profile_ranks is None or torch.distributed.get_rank() in args.profile_ranks + ): schedule = torch.profiler.schedule(wait=0, warmup=0, active=active_steps, repeat=1) activities = [torch.profiler.ProfilerActivity.CPU] activities.extend([torch.profiler.ProfilerActivity.HPU] if device.startswith("hpu") else []) From 53941565b264a8eb6e8dde5ca13aaad8d426b2f8 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Sat, 12 Oct 2024 12:41:17 +0000 Subject: [PATCH 334/590] shuffle concate dataset index --- ALCF/test_blendable_dataset.py | 40 ++++++++++++++++++- .../pipeline_parallel/p2p_communication.py | 20 +++++----- megatron/data/gpt_dataset.py | 14 +++---- 3 files changed, 56 insertions(+), 18 deletions(-) diff --git a/ALCF/test_blendable_dataset.py b/ALCF/test_blendable_dataset.py index a3cabddd293..c1198621421 100644 --- a/ALCF/test_blendable_dataset.py +++ b/ALCF/test_blendable_dataset.py @@ -1,5 +1,6 @@ #!/usr/bin/env python import time +import json start_time = time.time() from mpi4py import MPI import os @@ -37,7 +38,7 @@ def print_rank_0(msg): os.makedirs(args.trace_dir, exist_ok=True) - +corpus_all = [] data_file_list = args.data_file_list print_rank_0(f"Reading data from {args.data_file_list}") files = [] @@ -51,6 +52,9 @@ def print_rank_0(msg): files.append(float(w)) files.append(fname) files.append(c) + if c not in corpus_all: + corpus_all.append(c) + splits_string="100,0,0" weights = np.array(weights) @@ -82,6 +86,40 @@ def print_rank_0(msg): print_rank_0(f"Total number of samples: {len(train_ds)}") print_rank_0(f"Weights set: {weights[:min(8, num_datasets)]}") + +def get_sample_info(blendable_dataset, idx): + # corpus dataset + cd = blendable_dataset.dataset_index[idx] + # index within the corpus dataset + cds = blendable_dataset.dataset_sample_index[idx] + # dataset index within each corpus + fcd = blendable_dataset.datasets[cd].dataset_index[cds] + # sample index within the dataset + fcds = blendable_dataset.datasets[cd].dataset_sample_index[cds] + # corresponding data file + prefix = blendable_dataset.datasets[cd].dataset_builders[fcd].prefix + corpus = blendable_dataset.datasets[cd].dataset_builders[fcd].corpus + #v = blendable_dataset[idx]['text'] + #norm = np.linalg.norm(v) + return prefix, corpus, fcds + +num_batches = args.train_iters +print(f"global_batch_size: {args.global_batch_size}") +print(f"number of batches: {num_batches}") + +fout = open("samples_list.jsonl", "w") +if comm.rank == 0: + for i in range(num_batches): + ns_corpus = {} + for c in corpus_all: + ns_corpus[c] = 0 + for j in range(args.global_batch_size): + prefix, corpus, idx = get_sample_info(train_ds, i*args.global_batch_size+j) + ns_corpus[corpus] +=1 + fout.write(f"\u007b 'batch': {i}, 'sample': {j}, 'corpus': '{corpus}', 'prefix': '{prefix}', 'dataset_sample_index': {idx} \u007d\n") + fout.write(f"\u007b 'batch': {i}, 'histogram': {ns_corpus} \u007d \n") +comm.Barrier() +exit() start_build_dataloader = time.time() print_rank_0(f"Starting to build the data loader") rank_in_parallel_group = mpu.get_sequence_parallel_rank() diff --git a/megatron/core/pipeline_parallel/p2p_communication.py b/megatron/core/pipeline_parallel/p2p_communication.py index b23f6c84b3f..64f75ad0053 100644 --- a/megatron/core/pipeline_parallel/p2p_communication.py +++ b/megatron/core/pipeline_parallel/p2p_communication.py @@ -16,7 +16,8 @@ from megatron.core import ModelParallelConfig from deepspeed.accelerator import get_accelerator - +from megatron.utils import Profile +Profile("PIPELINE") # Types Shape = Union[List[int], torch.Size] @@ -329,6 +330,7 @@ def _ring_exchange_wrapper(**kwargs): return tensor_recv_prev, tensor_recv_next, reqs +@dlp.log def recv_forward(tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: """ Receive tensor from previous rank in pipeline (forward receive). @@ -353,7 +355,7 @@ def recv_forward(tensor_shape: Shape, config.timers('forward-recv').stop() return input_tensor - +@dlp.log def recv_backward(tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: """Receive tensor from next rank in pipeline (backward receive). @@ -376,7 +378,7 @@ def recv_backward(tensor_shape: Shape, config.timers('backward-recv').stop() return output_tensor_grad - +@dlp.log def send_forward(output_tensor: torch.Tensor, config: ModelParallelConfig) -> None: """Send tensor to next rank in pipeline (forward send). @@ -397,7 +399,7 @@ def send_forward(output_tensor: torch.Tensor, if config.timers is not None: config.timers('forward-send').stop() - +@dlp.log def send_backward(input_tensor_grad: torch.Tensor, config: ModelParallelConfig) -> None: """Send tensor to previous rank in pipeline (backward send). @@ -417,7 +419,7 @@ def send_backward(input_tensor_grad: torch.Tensor, if config.timers is not None: config.timers('backward-send').stop() - +@dlp.log def send_forward_recv_backward(output_tensor: torch.Tensor, tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: @@ -441,7 +443,7 @@ def send_forward_recv_backward(output_tensor: torch.Tensor, config.timers('forward-send-backward-recv').stop() return output_tensor_grad - +@dlp.log def send_backward_recv_forward(input_tensor_grad: torch.Tensor, tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: @@ -465,7 +467,7 @@ def send_backward_recv_forward(input_tensor_grad: torch.Tensor, config.timers('backward-send-forward-recv').stop() return input_tensor - +@dlp.log def send_forward_recv_forward(output_tensor: torch.Tensor, recv_prev: bool, tensor_shape: Shape, @@ -491,7 +493,7 @@ def send_forward_recv_forward(output_tensor: torch.Tensor, return input_tensor, wait_handles return input_tensor - +@dlp.log def send_backward_recv_backward(input_tensor_grad: torch.Tensor, recv_next: bool, tensor_shape: Shape, @@ -517,7 +519,7 @@ def send_backward_recv_backward(input_tensor_grad: torch.Tensor, return output_tensor_grad, wait_handles return output_tensor_grad - +@dlp.log def send_forward_backward_recv_forward_backward( output_tensor: torch.Tensor, input_tensor_grad: torch.Tensor, diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 0cf97356a41..c801a6a5ae5 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -114,8 +114,10 @@ def _build_indices(): print_rank_0('> elapsed time for building concat dataset indices: ' '{:.2f} (sec)'.format(time.time() - start_time)) return dataset_index, dataset_sample_index - + self.dataset_index, self.dataset_sample_index = _build_indices() + np_rng = np.random.RandomState(seed=dataset_builders[0].seed) + self.shuffle_index=np_rng.shuffle(range(self.num_samples)) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," @@ -125,13 +127,9 @@ def __len__(self): @dlp.log def __getitem__(self, idx): - if idx >= self.num_samples: - print_rank_0(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") - id = np.random.randint(self.num_samples) - else: - id = idx - i = self.dataset_index[idx] - j = self.dataset_sample_index[idx] + id_shuffle = self.shuffle_index[idx] + i = self.dataset_index[id_shuffle] + j = self.dataset_sample_index[id_shuffle] if self.dataset_builders[i].build: return self.dataset_builders[i].dataset[j] else: From 573b668ef20f4f5060937511169dfcd2c877a74c Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Sat, 12 Oct 2024 09:27:04 -0500 Subject: [PATCH 335/590] fixed bugs --- megatron/core/pipeline_parallel/p2p_communication.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/core/pipeline_parallel/p2p_communication.py b/megatron/core/pipeline_parallel/p2p_communication.py index 64f75ad0053..78e43e7fed1 100644 --- a/megatron/core/pipeline_parallel/p2p_communication.py +++ b/megatron/core/pipeline_parallel/p2p_communication.py @@ -17,7 +17,7 @@ from megatron.core import ModelParallelConfig from deepspeed.accelerator import get_accelerator from megatron.utils import Profile -Profile("PIPELINE") +dlp = Profile("PIPELINE") # Types Shape = Union[List[int], torch.Size] From 41ff0598b1a1c150a155404de408ef0575f6d6ee Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 12 Oct 2024 13:25:19 -0500 Subject: [PATCH 336/590] Update `ALCF/helpers.sh`, `train_aGPT_7B.sh` --- ALCF/helpers.sh | 382 ++++++++++++++++++++++++++++++----------------- train_aGPT_7B.sh | 13 +- 2 files changed, 252 insertions(+), 143 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index bc31c322ce7..109916dc37d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -125,9 +125,9 @@ setup() { set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` - tok="${TOKENIZER_TYPE:-Llama2}" + tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit make_data || exit # Print job info @@ -140,6 +140,7 @@ setup() { setup_run_cmd "$@" || exit } + ##################################################### # setup_run_cmd # @@ -150,7 +151,8 @@ setup_run_cmd() { # take in additional arguments # and append them directly to # the end of the `run_cmd` - custom_args="$@" + # custom_args="$@" + custom_args=("$@") ############################## #### Make it easy to track experiments by date ################### year="$(date "+%Y")" @@ -168,78 +170,122 @@ setup_run_cmd() { # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` ################################################################## setupLauncher "${LAUNCH_WITH:-MPICH}" || exit - TBDIR="${CKPT_DIR}/tensorboard" - mkdir -p "${TBDIR}" export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" printf "\n" echo "Using data_cache_path: ${data_cache_path}" - export DEFAULTS="\ - --split 100,0,0 \ - --log-interval 1 \ - --no-bias-gelu-fusion \ - --no-bias-dropout-fusion \ - --no-masked-softmax-fusion \ - --no-gradient-accumulation-fusion \ - --accumulate-allreduce-grads-in-fp32 \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard" - OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" + TRAIN_SPLIT="${TRAIN_SPLIT:-100}" + VAL_SPLIT="${VAL_SPLIT:-0}" + TEST_SPLIT="${TEST_SPLIT:-0}" + LOG_INTERVAL="${LOG_INTERVAL:-1}" + DEFAULTS=( + "--split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT}" + "--log-interval ${LOG_INTERVAL}" + "--no-bias-gelu-fusion" + "--no-bias-dropout-fusion" + "--no-masked-softmax-fusion" + "--no-gradient-accumulation-fusion" + "--accumulate-allreduce-grads-in-fp32" + ) + # export DEFAULTS="\ + # --split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT} \ + # --log-interval ${LOG_INTERVAL} \ + # --no-bias-gelu-fusion \ + # --no-bias-dropout-fusion \ + # --no-masked-softmax-fusion \ + # --no-gradient-accumulation-fusion \ + # --accumulate-allreduce-grads-in-fp32" + # OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then - DEFAULTS="${DEFAULTS} --use-checkpoint-opt_param-scheduler" + DEFAULTS+=("--use-checkpoint-opt_param-scheduler") fi - if [[ "${SP}" -ge 2 ]]; then - export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" + if [[ "${SP}" -gt 1 ]]; then + DEFAULTS+=( + "--ds-sequence-parallel-size ${SP}" + "--force-ds-sequence-parallel" + ) fi ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` ################################################################## - if [[ -z "${NO_LLAMA:-}" ]]; then - llama_flags="${LLAMA_ARGS}\ - --num-key-value-heads ${NUM_KV_HEAD} \ - --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - " + LLAMA_ARGS="" + if [[ "${SP}" == 1 ]]; then + export LLAMA_ARGS="${LLAMA_ARGS} " else - echo "!! Running in NO_LLAMA MODE !!" - llama_flags="" + export LLAMA_ARGS="" + echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" + fi + if [[ -z "${NO_LLAMA:-}" ]]; then + llama_flags=( + "--swiglu" + "--hidden-dropout 0" + "--attention-dropout 0" + "--normalization rmsnorm" + "--disable-bias-linear" + "--no-query-key-layer-scaling" + "--use-rotary-position-embeddings" + "--untie-embeddings-and-output-weights" + "--num-key-value-heads ${NUM_KV_HEAD}" + "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" + ) + fi + + TENSORBARD_ARGS=() + if [[ -z "${USE_TENSORBARD:-}" ]]; then + TBDIR="${CKPT_DIR}/tensorboard" + mkdir -p "${TBDIR}" + # --log-timers-to-tensorboard \ + # --log-optimizer-states-to-tensorboard" + # --tensorboard-dir ${TBDIR} \ + TENSORBARD_ARGS+=( + "--log-timers-to-tensorboard" + "--log-optimizer-states-to-tensorboard" + "--tensorboard-dir ${TBDIR}" + ) fi - export run_cmd=" - ${LAUNCHER} \ - --${DTYPE} \ - ${DEFAULTS} \ - --optimizer ${OPT} \ - --adam-beta1=${ADAM_BETA1} \ - --adam-beta2=${ADAM_BETA2} \ - --adam-eps=${ADAM_EPS} \ - --weight-decay=${WEIGHT_DECAY} \ - --save ${CKPT_DIR} \ - --load ${CKPT_DIR} \ - --seq-length ${SEQ} \ - --num-layers ${NLAYERS} \ - --hidden-size ${HIDDEN} \ - --tensorboard-dir ${TBDIR} \ - --train-iters ${TRAIN_ITERS} \ - --eval-iters ${EVAL_ITERS} \ - --distributed-backend ${BE} \ - --num-attention-heads ${HEADS} \ - --save-interval ${SAVE_INTERVAL} \ - --eval-interval ${EVAL_INTERVAL} \ - --max-position-embeddings ${SEQ} \ - --micro-batch-size ${MICRO_BATCH} \ - --tensor-model-parallel-size ${TP} \ - --global-batch-size ${GLOBAL_BATCH} \ - --pipeline-model-parallel-size ${PP} \ - --data-cache-path ${data_cache_path} \ - ${DATA_FLAGS} \ - ${LR_ARGS} \ - ${llama_flags} \ - ${FLASH_ARG} \ - ${TIMING_STR} \ - ${TOKENIZER_FLAGS} \ - ${ds_args} \ - ${gpt_args[*]} \ - ${custom_args} - " + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" + export ADAM_BETA1="${ADAM_BETA1:-0.9}" + export ADAM_BETA2="${ADAM_BETA2:-0.95}" + export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + export run_cmd=( + "${LAUNCHER}" + "--${DTYPE}" + "${DEFAULTS[@]}" + "--optimizer ${OPT}" + "--save ${CKPT_DIR}" + "--load ${CKPT_DIR}" + "--seq-length ${SEQ}" + "--num-layers ${NLAYERS}" + "--hidden-size ${HIDDEN}" + "--train-iters ${TRAIN_ITERS}" + "--eval-iters ${EVAL_ITERS}" + "--distributed-backend ${BE}" + "--adam-beta1 ${ADAM_BETA1:-0.9}" + "--adam-beta2 ${ADAM_BETA2:-0.95}" + "--adam-eps ${ADAM_EPS:-0.00001}" + "--clip-grad ${CLIP_GRAD:-1.0}" + "--weight-decay ${WEIGHT_DECAY:-0.1}" + "--num-attention-heads ${HEADS}" + "--save-interval ${SAVE_INTERVAL}" + "--eval-interval ${EVAL_INTERVAL}" + "--max-position-embeddings ${SEQ}" + "--micro-batch-size ${MICRO_BATCH}" + "--tensor-model-parallel-size ${TP}" + "--global-batch-size ${GLOBAL_BATCH}" + "--pipeline-model-parallel-size ${PP}" + "--data-cache-path ${data_cache_path}" + "--data-file-list ${DATA_FILE_LIST:-${dfl_fallback}}" + "${TENSORBARD_ARGS[@]}" + "${DATA_FLAGS}" + "${LR_ARGS}" + "${llama_flags[@]}" + "${FLASH_ARG}" + "${TIMING_STR}" + "${TOKENIZER_FLAGS}" + "${ds_args[@]}" + "${gpt_args[@]}" + "${custom_args[@]}" + ) } save_dotenv() { @@ -518,7 +564,6 @@ set_ccl_vars_on_aurora() { ############################################################################## setParams() { FLASH_ARG="" - LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then @@ -600,9 +645,9 @@ setParams() { export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" - export ADAM_BETA1="${ADAM_BETA1:-0.9}" - export ADAM_BETA2="${ADAM_BETA2:-0.95}" - export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + # export ADAM_BETA1="${ADAM_BETA1:-0.9}" + # export ADAM_BETA2="${ADAM_BETA2:-0.95}" + # export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l <"${HOSTFILE}") @@ -648,24 +693,19 @@ setParams() { # # For this reason, we only use the default LLAMA_ARGS when SP=0. ########################################################################## - if [[ "${SP}" == 1 ]]; then - export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - else - export LLAMA_ARGS="" - echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" - fi # -----[Learning Rate Settings]-------------------------------------------- export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args # -----[Learning Rate Settings]-------------------------------------------- - if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then + # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then + if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then TIMING_STR="\ - --timing-log-level ${TIMING_LOG_LEVEL} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ - " + --timing-log-level ${TIMING_LOG_LEVEL}" + # --log-timers-to-tensorboard \ + # --log-optimizer-states-to-tensorboard \ + # " else TIMING_STR="" fi @@ -679,19 +719,31 @@ setParams() { ############################################## set_args() { # ---- Set DeepSpeed arguments -------------------------------- - ds_args=" " - ds_args=" --deepspeed ${ds_args}" - if [[ $PP == 1 ]]; then - ds_args=" --no-pipeline-parallel ${ds_args}" + ds_args=( + "--deepspeed" + ) + if [[ "${PP:-1}" == 1 ]]; then + ds_args+=("--no-pipeline-parallel") fi - ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" - ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" + ds_args+=("--deepspeed_config=${DS_CONFIG}") + ds_args+=("--zero-stage=$ZERO_STAGE") if [[ "${ZERO_STAGE}" == 3 ]]; then - ds_args="--use-mics ${ds_args}" + ds_args+=("--use-mics") fi + # ds_args=" " + # ds_args=" --deepspeed ${ds_args}" + # if [[ $PP == 1 ]]; then + # ds_args=" --no-pipeline-parallel ${ds_args}" + # fi + # ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" + # ds_args="--zero-stage=$ZERO_STAGE ${ds_args}" + # if [[ "${ZERO_STAGE}" == 3 ]]; then + # ds_args="--use-mics ${ds_args}" + # fi if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - ds_args=" --deepspeed-activation-checkpointing ${ds_args}" + ds_args+=("--deepspeed-activation-checkpointing") + # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" # --checkpoint-activations \ # --deepspeed-activation-checkpointing fi @@ -804,7 +856,8 @@ get_output_prefix() { pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" if [[ -n "${TOKENIZER_TYPE:-}" ]]; then - pre="${pre}_tok${TOKENIZER_TYPE}" + _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa + pre="${pre}_tok${_tok}" fi if [[ -n "${LR_DECAY_ITERS}" ]]; then pre="${pre}_ldi${LR_DECAY_ITERS}" @@ -832,7 +885,7 @@ setOutput() { # Build DeepSpeed config and write to .json ############################################# buildDSconfig() { - export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" + # export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" mkdir -p "$(dirname "${DS_CONFIG}")" echo "DS_CONFIG: ${DS_CONFIG}" @@ -893,31 +946,6 @@ install_dependencies() { fi } -###################################################################### -# install_deepspeed_for_xpu -# -# Install microsoft/DeepSpeed on PVC -# -# This will: -# 1. Clone rep -# 2. Checkout appropriate branch -# 3. Install into virtual environment -###################################################################### -install_deepspeed_for_xpu() { - # python3 -m pip install "torch==2.1.0.post2" torchvision==0.16.0.post2 torchaudio==2.1.0.post2 intel-extension-for-pytorch==2.1.30.post0 oneccl_bind_pt==2.1.300+xpu --extra-index-url "https://pytorch-extension.intel.com/release-whl/stable/xpu/us/" - echo "Building + Installing DeepSpeed on $(hostname)" - outdir="${WORKING_DIR}/deps/DeepSpeed" - mkdir -p "${outdir}" - git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" - cd "${outdir}" || exit - echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" - python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1>/dev/null - python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1>/dev/null - python setup.py develop 1>/dev/null - cd "${WORKING_DIR}" - echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" -} - ################################################# # Fix for distributed key value store on Aurora ################################################# @@ -1003,9 +1031,11 @@ setup_tokenizer_and_data() { fi echo "Setting up tokenizer with ${tok}" echo "Using data_file_list: ${dfl}" + _data_flags=() + _tokenizer_flags=() if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" - export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" + _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") machine=$(get_machine_name) if [[ ${machine} == "polaris" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" @@ -1019,18 +1049,25 @@ setup_tokenizer_and_data() { export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - export DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" + _data_flags+=( + "--data-path ${DATA_PATH}" + "--vocab-file ${VOCAB_FILE}" + "--merge-file ${MERGE_FILE}" + ) else - export DATA_FLAGS="" - export TOKENIZER_TYPE="Llama2" + export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - export TOKENIZER_FLAGS="--tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" - if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then - echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" - setData "${dfl}" || exit - fi + _tokenizer_flags+=( + "--tokenizer-type ${TOKENIZER_TYPE}" + "--tokenizer-model ${TOKENIZER_MODEL}" + ) + # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then + echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST:-}" + setData "${dfl}" || exit fi + export DATA_FLAGS="${_data_flags[*]}" + export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" } @@ -1059,7 +1096,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export WEIGHT_SUM="${ws}" export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" - export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" + # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" echo "--------------------" echo "Updated environment:" printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" @@ -1089,16 +1126,6 @@ generateDSconfig() { exit 1 fi done - # \"optimizer\": { - # \"type\": \"AdamW\", - # \"params\": { - # \"lr\": ${LR}, - # \"beta1\": 0.9, - # \"beta2\": 0.95, - # \"eps\": 1e-5, - # \"weight_decay\": 1e-1 - # } - # }, # \"scheduler\": { # \"type\": \"WarmupLR\", # \"params\": { @@ -1113,6 +1140,7 @@ generateDSconfig() { \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, + \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, \"gradient_clipping\": 1.0, \"activation_checkpointing\": { @@ -1160,6 +1188,20 @@ generateDSconfig() { else dtype="\"communication_data_type\": \"fp32\"," fi + if [[ "${OPT:-adamw}" == "ds.adamw" ]]; then + optimizer="\ + \"optimizer\": { + \"type\": \"AdamW\", + \"params\": { + \"lr\": ${LR}, + \"beta1\": 0.9, + \"beta2\": 0.95, + \"eps\": 1e-5, + \"weight_decay\": 1e-1 + }," + else + optimizer="" + fi if [[ "${ZERO_STAGE}" == 3 ]]; then # \"mics_shard_size\": 2, zero="\ @@ -1185,8 +1227,7 @@ generateDSconfig() { }," # elif [[ $ZERO_STAGE == 2 ]]; then elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then - # if [[ -n "${CPU_OPTIMIZER}" ]]; then - if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then + if [[ -z "${CPU_OPTIMIZER:-}" ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" zero="\ \"zero_optimization\": { @@ -1215,9 +1256,8 @@ generateDSconfig() { else extra="\ \"comms_logger\": { - \"enabled\": true, + \"enabled\": ${COMMS_LOGGER:-false}, \"verbose\": false, - \"prof_all\": true, \"debug\": false }," fi @@ -1227,6 +1267,7 @@ generateDSconfig() { cat <"$1" { $common +$optimizer $zero $dtype $extra @@ -1304,6 +1345,73 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } +reset_env() { + custom_vars=( + NO_FLASH_ATTN + TP + PP + SP + FLASH_ARG + OPT + ADAM_BETA1 + ADAM_BETA2 + ADAM_EPS + WEIGHT_DECAY + HEADS + NLAYERS + HIDDEN + NUM_KV_HEAD + FFN_HIDDEN_SIZE + SEQ + ZERO_STAGE + MICRO_BATCH + EVAL_ITERS + EVAL_INTERVAL + TIMING_LOG_LEVEL + ACT_CKPT_NUM_LAYERS + USE_ACTIVATION_CHECKPOINTING + GLOBAL_BATCH_MAX + GLOBAL_BATCH + TRAIN_TOKENS + TRAIN_ITERS + MODEL_TYPE + LLAMA_ARGS + LR + LR_WARMUP_FRAC + LR_DECAY_ITERS + LR_ARGS + CPU_OPTIMIZER + DS_CONFIG + OUTPUT_DIR + OUTPUT_LOG + CKPT_DIR + ds_args + EXEC + EXEC_STEM + DATA_FLAGS + TOKENIZER_TYPE + TOKENIZER_MODEL + TOKENIZER_FLAGS + DATA_FILE_LIST + NUM_DOCS + WEIGHT_SUM + DFL_STEM + DATA_CACHE_PATH + DOTENV_FILE + YEAR + MONTH + DAY + TODAY + STARTED_AT + LAUNCHER + data_cache_path + DEFAULTS + ) + printf "Unsetting custom vars: %s\n" "${custom_vars[*]}" + unset "${custom_vars[@]}" +} + + ########################### # call helpers_main() ########################### diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index a6a2db72abb..286740fc89f 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -16,19 +16,20 @@ source "${HERE}/ALCF/helpers.sh" || exit # 3. call `setup` from `./ALCF/helpers.sh` setup "$@" || exit -export run_cmd="${run_cmd}" -echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +# export run_cmd="${run_cmd}" +echo "${run_cmd[@]}" | tee -a "${OUTPUT_LOG}" # 4. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" -# 5. Ignore the following strings on Intel XPU devices -# (otherwise they'll clutter up logs) -XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" +# # 5. Ignore the following strings on Intel XPU devices +# # (otherwise they'll clutter up logs) +# XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" # if [[ $(ezpz_get_machine_name) == "aurora" ]]; then # module unload mpich && module load mpich # fi # # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} -eval "${run_cmd}" |& grep -E -v "${XPU_IGNORE_STRING}" |& tee -a "${OUTPUT_LOG}" +# eval "${run_cmd[@]}" |& tee -a "${OUTPUT_LOG}" +eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" From 9de83a9684742081e8dabd82f59ede7c7fdfffb3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 12 Oct 2024 15:56:18 -0500 Subject: [PATCH 337/590] Fix `shuffle_idx` in `megatron/data/gpt_dataset.py` --- megatron/data/gpt_dataset.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index c801a6a5ae5..a8457609db2 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -117,6 +117,8 @@ def _build_indices(): self.dataset_index, self.dataset_sample_index = _build_indices() np_rng = np.random.RandomState(seed=dataset_builders[0].seed) + self.shuffle_index = np.arange(self.num_samples) + np_rng.shuffle(self.shuffle_index) self.shuffle_index=np_rng.shuffle(range(self.num_samples)) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," From d7a2594a87fd0ffcd8eb3a6f102af809cf397ab2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 12 Oct 2024 16:24:02 -0500 Subject: [PATCH 338/590] Fix `shuffle_idx` in `megatron/data/gpt_dataset.py` --- megatron/data/gpt_dataset.py | 1 - 1 file changed, 1 deletion(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index a8457609db2..8501324752e 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -119,7 +119,6 @@ def _build_indices(): np_rng = np.random.RandomState(seed=dataset_builders[0].seed) self.shuffle_index = np.arange(self.num_samples) np_rng.shuffle(self.shuffle_index) - self.shuffle_index=np_rng.shuffle(range(self.num_samples)) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," From 3e33a6a66d0dae2328e02cad1c760b862c0064dc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 13 Oct 2024 10:15:30 -0500 Subject: [PATCH 339/590] Update `ALCF/helpers.sh`, `train_aGPT_7B.sh` --- ALCF/helpers.sh | 432 +++++++++++++++++++++++++++++++---------------- train_aGPT_7B.sh | 13 +- 2 files changed, 297 insertions(+), 148 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index bc31c322ce7..5df9a2c7a58 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -120,14 +120,15 @@ setup() { # Create `deepspeed_config.json` from runtime params from ^ buildDSconfig || exit # Specify output directory for {logs, checkpoints, etc.} + setup_checkpoint || exit setOutput || exit # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` - tok="${TOKENIZER_TYPE:-Llama2}" + tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit make_data || exit # Print job info @@ -140,6 +141,7 @@ setup() { setup_run_cmd "$@" || exit } + ##################################################### # setup_run_cmd # @@ -150,7 +152,8 @@ setup_run_cmd() { # take in additional arguments # and append them directly to # the end of the `run_cmd` - custom_args="$@" + # custom_args="$@" + custom_args=("$@") ############################## #### Make it easy to track experiments by date ################### year="$(date "+%Y")" @@ -168,78 +171,122 @@ setup_run_cmd() { # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` ################################################################## setupLauncher "${LAUNCH_WITH:-MPICH}" || exit - TBDIR="${CKPT_DIR}/tensorboard" - mkdir -p "${TBDIR}" export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" printf "\n" echo "Using data_cache_path: ${data_cache_path}" - export DEFAULTS="\ - --split 100,0,0 \ - --log-interval 1 \ - --no-bias-gelu-fusion \ - --no-bias-dropout-fusion \ - --no-masked-softmax-fusion \ - --no-gradient-accumulation-fusion \ - --accumulate-allreduce-grads-in-fp32 \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard" - OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" + TRAIN_SPLIT="${TRAIN_SPLIT:-100}" + VAL_SPLIT="${VAL_SPLIT:-0}" + TEST_SPLIT="${TEST_SPLIT:-0}" + LOG_INTERVAL="${LOG_INTERVAL:-1}" + DEFAULTS=( + "--split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT}" + "--log-interval ${LOG_INTERVAL}" + "--no-bias-gelu-fusion" + "--no-bias-dropout-fusion" + "--no-masked-softmax-fusion" + "--no-gradient-accumulation-fusion" + "--accumulate-allreduce-grads-in-fp32" + ) + # export DEFAULTS="\ + # --split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT} \ + # --log-interval ${LOG_INTERVAL} \ + # --no-bias-gelu-fusion \ + # --no-bias-dropout-fusion \ + # --no-masked-softmax-fusion \ + # --no-gradient-accumulation-fusion \ + # --accumulate-allreduce-grads-in-fp32" + # OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then - DEFAULTS="${DEFAULTS} --use-checkpoint-opt_param-scheduler" + DEFAULTS+=("--use-checkpoint-opt_param-scheduler") fi - if [[ "${SP}" -ge 2 ]]; then - export DEFAULTS="${DEFAULTS} --ds-sequence-parallel-size ${SP} --force-ds-sequence-parallel" + if [[ "${SP}" -gt 1 ]]; then + DEFAULTS+=( + "--ds-sequence-parallel-size ${SP}" + "--force-ds-sequence-parallel" + ) fi ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` ################################################################## - if [[ -z "${NO_LLAMA:-}" ]]; then - llama_flags="${LLAMA_ARGS}\ - --num-key-value-heads ${NUM_KV_HEAD} \ - --ffn-hidden-size ${FFN_HIDDEN_SIZE} \ - " + LLAMA_ARGS="" + if [[ "${SP}" == 1 ]]; then + export LLAMA_ARGS="${LLAMA_ARGS} " else - echo "!! Running in NO_LLAMA MODE !!" - llama_flags="" + export LLAMA_ARGS="" + echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" fi - export run_cmd=" - ${LAUNCHER} \ - --${DTYPE} \ - ${DEFAULTS} \ - --optimizer ${OPT} \ - --adam-beta1=${ADAM_BETA1} \ - --adam-beta2=${ADAM_BETA2} \ - --adam-eps=${ADAM_EPS} \ - --weight-decay=${WEIGHT_DECAY} \ - --save ${CKPT_DIR} \ - --load ${CKPT_DIR} \ - --seq-length ${SEQ} \ - --num-layers ${NLAYERS} \ - --hidden-size ${HIDDEN} \ - --tensorboard-dir ${TBDIR} \ - --train-iters ${TRAIN_ITERS} \ - --eval-iters ${EVAL_ITERS} \ - --distributed-backend ${BE} \ - --num-attention-heads ${HEADS} \ - --save-interval ${SAVE_INTERVAL} \ - --eval-interval ${EVAL_INTERVAL} \ - --max-position-embeddings ${SEQ} \ - --micro-batch-size ${MICRO_BATCH} \ - --tensor-model-parallel-size ${TP} \ - --global-batch-size ${GLOBAL_BATCH} \ - --pipeline-model-parallel-size ${PP} \ - --data-cache-path ${data_cache_path} \ - ${DATA_FLAGS} \ - ${LR_ARGS} \ - ${llama_flags} \ - ${FLASH_ARG} \ - ${TIMING_STR} \ - ${TOKENIZER_FLAGS} \ - ${ds_args} \ - ${gpt_args[*]} \ - ${custom_args} - " + if [[ -z "${NO_LLAMA:-}" ]]; then + llama_flags=( + "--swiglu" + "--hidden-dropout 0" + "--attention-dropout 0" + "--normalization rmsnorm" + "--disable-bias-linear" + "--no-query-key-layer-scaling" + "--use-rotary-position-embeddings" + "--untie-embeddings-and-output-weights" + "--num-key-value-heads ${NUM_KV_HEAD}" + "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" + ) + fi + + TENSORBARD_ARGS=() + if [[ -z "${USE_TENSORBARD:-}" ]]; then + TBDIR="${CKPT_DIR}/tensorboard" + mkdir -p "${TBDIR}" + # --log-timers-to-tensorboard \ + # --log-optimizer-states-to-tensorboard" + # --tensorboard-dir ${TBDIR} \ + TENSORBARD_ARGS+=( + "--log-timers-to-tensorboard" + "--log-optimizer-states-to-tensorboard" + "--tensorboard-dir ${TBDIR}" + ) + fi + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" + export ADAM_BETA1="${ADAM_BETA1:-0.9}" + export ADAM_BETA2="${ADAM_BETA2:-0.95}" + export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + export run_cmd=( + "${LAUNCHER}" + "--${DTYPE}" + "${DEFAULTS[@]}" + "--optimizer ${OPT}" + "--save ${CKPT_DIR}" + "--load ${CKPT_DIR}" + "--seq-length ${SEQ}" + "--num-layers ${NLAYERS}" + "--hidden-size ${HIDDEN}" + "--train-iters ${TRAIN_ITERS}" + "--eval-iters ${EVAL_ITERS}" + "--distributed-backend ${BE}" + "--adam-beta1 ${ADAM_BETA1:-0.9}" + "--adam-beta2 ${ADAM_BETA2:-0.95}" + "--adam-eps ${ADAM_EPS:-0.00001}" + "--clip-grad ${CLIP_GRAD:-1.0}" + "--weight-decay ${WEIGHT_DECAY:-0.1}" + "--num-attention-heads ${HEADS}" + "--save-interval ${SAVE_INTERVAL}" + "--eval-interval ${EVAL_INTERVAL}" + "--max-position-embeddings ${SEQ}" + "--micro-batch-size ${MICRO_BATCH}" + "--tensor-model-parallel-size ${TP}" + "--global-batch-size ${GLOBAL_BATCH}" + "--pipeline-model-parallel-size ${PP}" + "--data-cache-path ${data_cache_path}" + "--data-file-list ${DATA_FILE_LIST:-${dfl_fallback}}" + "${TENSORBARD_ARGS[@]}" + "${DATA_FLAGS}" + "${LR_ARGS}" + "${llama_flags[@]}" + "${FLASH_ARG}" + "${TIMING_STR}" + "${TOKENIZER_FLAGS}" + "${ds_args[@]}" + "${gpt_args[@]}" + "${custom_args[@]}" + ) } save_dotenv() { @@ -518,7 +565,6 @@ set_ccl_vars_on_aurora() { ############################################################################## setParams() { FLASH_ARG="" - LLAMA_ARGS="--attention-dropout 0 --hidden-dropout 0" # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then @@ -600,9 +646,9 @@ setParams() { export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" - export ADAM_BETA1="${ADAM_BETA1:-0.9}" - export ADAM_BETA2="${ADAM_BETA2:-0.95}" - export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} + # export ADAM_BETA1="${ADAM_BETA1:-0.9}" + # export ADAM_BETA2="${ADAM_BETA2:-0.95}" + # export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l <"${HOSTFILE}") @@ -648,24 +694,19 @@ setParams() { # # For this reason, we only use the default LLAMA_ARGS when SP=0. ########################################################################## - if [[ "${SP}" == 1 ]]; then - export LLAMA_ARGS="${LLAMA_ARGS} --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear" - else - export LLAMA_ARGS="" - echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" - fi # -----[Learning Rate Settings]-------------------------------------------- export LR=${LR:-0.0003} # LEARNING_RATE export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS set_lr_args # -----[Learning Rate Settings]-------------------------------------------- - if [[ "${TIMING_LOG_LEVEL}" -ge 1 ]]; then + # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then + if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then TIMING_STR="\ - --timing-log-level ${TIMING_LOG_LEVEL} \ - --log-timers-to-tensorboard \ - --log-optimizer-states-to-tensorboard \ - " + --timing-log-level ${TIMING_LOG_LEVEL}" + # --log-timers-to-tensorboard \ + # --log-optimizer-states-to-tensorboard \ + # " else TIMING_STR="" fi @@ -679,19 +720,31 @@ setParams() { ############################################## set_args() { # ---- Set DeepSpeed arguments -------------------------------- - ds_args=" " - ds_args=" --deepspeed ${ds_args}" - if [[ $PP == 1 ]]; then - ds_args=" --no-pipeline-parallel ${ds_args}" + ds_args=( + "--deepspeed" + ) + if [[ "${PP:-1}" == 1 ]]; then + ds_args+=("--no-pipeline-parallel") fi - ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" - ds_args=" --zero-stage=$ZERO_STAGE ${ds_args}" + ds_args+=("--deepspeed_config=${DS_CONFIG}") + ds_args+=("--zero-stage=$ZERO_STAGE") if [[ "${ZERO_STAGE}" == 3 ]]; then - ds_args="--use-mics ${ds_args}" + ds_args+=("--use-mics") fi + # ds_args=" " + # ds_args=" --deepspeed ${ds_args}" + # if [[ $PP == 1 ]]; then + # ds_args=" --no-pipeline-parallel ${ds_args}" + # fi + # ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" + # ds_args="--zero-stage=$ZERO_STAGE ${ds_args}" + # if [[ "${ZERO_STAGE}" == 3 ]]; then + # ds_args="--use-mics ${ds_args}" + # fi if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - ds_args=" --deepspeed-activation-checkpointing ${ds_args}" + ds_args+=("--deepspeed-activation-checkpointing") + # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" # --checkpoint-activations \ # --deepspeed-activation-checkpointing fi @@ -804,7 +857,8 @@ get_output_prefix() { pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" if [[ -n "${TOKENIZER_TYPE:-}" ]]; then - pre="${pre}_tok${TOKENIZER_TYPE}" + _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa + pre="${pre}_tok${_tok}" fi if [[ -n "${LR_DECAY_ITERS}" ]]; then pre="${pre}_ldi${LR_DECAY_ITERS}" @@ -822,9 +876,21 @@ setOutput() { OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" export OUTPUT_DIR="${OUTPUT_DIR}" && mkdir -p "${OUTPUT_DIR}" export OUTPUT_LOG="${OUTPUT_DIR}/output.log" - export CKPT_DIR="checkpoints/${OUTPUT_PREFIX}" echo "${OUTPUT_LOG}" >>"logs/latest" printf "\n Please see logs at: %s\n" "$(printGreen "${OUTPUT_DIR}")" +} + +get_checkpoint_dir() { + if [[ -n "${CKPT_DIR:-}" ]]; then + echo "${CKPT_DIR}" + else + echo "checkpoints/$(get_output_prefix)" + fi +} + +setup_checkpoint() { + ckpt_dir=$(get_checkpoint_dir) + export CKPT_DIR="${ckpt_dir}" printf "Checkpoints will be saved to: %s\n" "$(printYellow "${CKPT_DIR}")" } @@ -832,7 +898,7 @@ setOutput() { # Build DeepSpeed config and write to .json ############################################# buildDSconfig() { - export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" + # export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" mkdir -p "$(dirname "${DS_CONFIG}")" echo "DS_CONFIG: ${DS_CONFIG}" @@ -893,31 +959,6 @@ install_dependencies() { fi } -###################################################################### -# install_deepspeed_for_xpu -# -# Install microsoft/DeepSpeed on PVC -# -# This will: -# 1. Clone rep -# 2. Checkout appropriate branch -# 3. Install into virtual environment -###################################################################### -install_deepspeed_for_xpu() { - # python3 -m pip install "torch==2.1.0.post2" torchvision==0.16.0.post2 torchaudio==2.1.0.post2 intel-extension-for-pytorch==2.1.30.post0 oneccl_bind_pt==2.1.300+xpu --extra-index-url "https://pytorch-extension.intel.com/release-whl/stable/xpu/us/" - echo "Building + Installing DeepSpeed on $(hostname)" - outdir="${WORKING_DIR}/deps/DeepSpeed" - mkdir -p "${outdir}" - git clone https://github.com/microsoft/DeepSpeed.git "${outdir}" - cd "${outdir}" || exit - echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" - python3 -m pip install --require-virtualenv -r requirements/requirements.txt 1>/dev/null - python3 -m pip install xgboost "numpy<2" --force-reinstall --upgrade --require-virtualenv 1>/dev/null - python setup.py develop 1>/dev/null - cd "${WORKING_DIR}" - echo "[install_deepspeed_for_xpu] !! pwd: $(pwd)" -} - ################################################# # Fix for distributed key value store on Aurora ################################################# @@ -1003,9 +1044,11 @@ setup_tokenizer_and_data() { fi echo "Setting up tokenizer with ${tok}" echo "Using data_file_list: ${dfl}" + _data_flags=() + _tokenizer_flags=() if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" - export TOKENIZER_FLAGS="--tokenizer-type GPT2BPETokenizer" + _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") machine=$(get_machine_name) if [[ ${machine} == "polaris" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" @@ -1019,18 +1062,25 @@ setup_tokenizer_and_data() { export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - export DATA_FLAGS="--data-path ${DATA_PATH} --vocab-file ${VOCAB_FILE} --merge-file ${MERGE_FILE}" + _data_flags+=( + "--data-path ${DATA_PATH}" + "--vocab-file ${VOCAB_FILE}" + "--merge-file ${MERGE_FILE}" + ) else - export DATA_FLAGS="" - export TOKENIZER_TYPE="Llama2" + export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - export TOKENIZER_FLAGS="--tokenizer-type Llama2Tokenizer --tokenizer-model ${TOKENIZER_MODEL}" - if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then - echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST-}" - setData "${dfl}" || exit - fi + _tokenizer_flags+=( + "--tokenizer-type ${TOKENIZER_TYPE}" + "--tokenizer-model ${TOKENIZER_MODEL}" + ) + # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then + echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST:-}" + setData "${dfl}" || exit fi + export DATA_FLAGS="${_data_flags[*]}" + export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" } @@ -1059,7 +1109,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export WEIGHT_SUM="${ws}" export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" - export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" + # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" echo "--------------------" echo "Updated environment:" printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" @@ -1071,6 +1121,30 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] echo "--------------------" } +generateDSconfig_new() { + cat < "${CONFIG_JSON}" + { + "train_batch_size" : $GLOBAL_BATCH, + "train_micro_batch_size_per_gpu": $MICRO_BATCH, + "steps_per_print": 1, + + "zero_optimization": { + "stage": $ZERO_STAGE + }, + + "bf16": { + "enabled": true + }, + + "data_types": { + "grad_accum_dtype": "fp32" + }, + + "wall_clock_breakdown" : false + } +EOT +} + ################################################################################ # generateDSconfig # @@ -1089,16 +1163,6 @@ generateDSconfig() { exit 1 fi done - # \"optimizer\": { - # \"type\": \"AdamW\", - # \"params\": { - # \"lr\": ${LR}, - # \"beta1\": 0.9, - # \"beta2\": 0.95, - # \"eps\": 1e-5, - # \"weight_decay\": 1e-1 - # } - # }, # \"scheduler\": { # \"type\": \"WarmupLR\", # \"params\": { @@ -1113,13 +1177,17 @@ generateDSconfig() { \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, + \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, \"gradient_clipping\": 1.0, - \"activation_checkpointing\": { - \"partition_activations\": true, - \"contiguous_memory_optimization\": true - }, \"wall_clock_breakdown\": false," + if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then + activation_checkpointing="\ + \"activation_checkpointing\": { + \"partition_activations\": true, + \"contiguous_memory_optimization\": true + }," + fi flops_profiler="\ \"flops_profiler\": { \"enabled\": true, @@ -1160,6 +1228,20 @@ generateDSconfig() { else dtype="\"communication_data_type\": \"fp32\"," fi + if [[ "${OPT:-adamw}" == "ds.adamw" ]]; then + optimizer="\ + \"optimizer\": { + \"type\": \"AdamW\", + \"params\": { + \"lr\": ${LR}, + \"beta1\": 0.9, + \"beta2\": 0.95, + \"eps\": 1e-5, + \"weight_decay\": 1e-1 + }," + else + optimizer="" + fi if [[ "${ZERO_STAGE}" == 3 ]]; then # \"mics_shard_size\": 2, zero="\ @@ -1185,8 +1267,7 @@ generateDSconfig() { }," # elif [[ $ZERO_STAGE == 2 ]]; then elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then - # if [[ -n "${CPU_OPTIMIZER}" ]]; then - if [[ "${CPU_OPTIMIZER:-0}" != 0 ]]; then + if [[ -z "${CPU_OPTIMIZER:-}" ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" zero="\ \"zero_optimization\": { @@ -1215,9 +1296,8 @@ generateDSconfig() { else extra="\ \"comms_logger\": { - \"enabled\": true, + \"enabled\": ${COMMS_LOGGER:-false}, \"verbose\": false, - \"prof_all\": true, \"debug\": false }," fi @@ -1227,6 +1307,7 @@ generateDSconfig() { cat <"$1" { $common +$optimizer $zero $dtype $extra @@ -1304,6 +1385,73 @@ printWhite() { printf "\e[1;37m%s\e[0m\n" "$@" } +reset_env() { + custom_vars=( + NO_FLASH_ATTN + TP + PP + SP + FLASH_ARG + OPT + ADAM_BETA1 + ADAM_BETA2 + ADAM_EPS + WEIGHT_DECAY + HEADS + NLAYERS + HIDDEN + NUM_KV_HEAD + FFN_HIDDEN_SIZE + SEQ + ZERO_STAGE + MICRO_BATCH + EVAL_ITERS + EVAL_INTERVAL + TIMING_LOG_LEVEL + ACT_CKPT_NUM_LAYERS + USE_ACTIVATION_CHECKPOINTING + GLOBAL_BATCH_MAX + GLOBAL_BATCH + TRAIN_TOKENS + TRAIN_ITERS + MODEL_TYPE + LLAMA_ARGS + LR + LR_WARMUP_FRAC + LR_DECAY_ITERS + LR_ARGS + CPU_OPTIMIZER + DS_CONFIG + OUTPUT_DIR + OUTPUT_LOG + CKPT_DIR + ds_args + EXEC + EXEC_STEM + DATA_FLAGS + TOKENIZER_TYPE + TOKENIZER_MODEL + TOKENIZER_FLAGS + DATA_FILE_LIST + NUM_DOCS + WEIGHT_SUM + DFL_STEM + DATA_CACHE_PATH + DOTENV_FILE + YEAR + MONTH + DAY + TODAY + STARTED_AT + LAUNCHER + data_cache_path + DEFAULTS + ) + printf "Unsetting custom vars: %s\n" "${custom_vars[*]}" + unset "${custom_vars[@]}" +} + + ########################### # call helpers_main() ########################### diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index a6a2db72abb..286740fc89f 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -16,19 +16,20 @@ source "${HERE}/ALCF/helpers.sh" || exit # 3. call `setup` from `./ALCF/helpers.sh` setup "$@" || exit -export run_cmd="${run_cmd}" -echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +# export run_cmd="${run_cmd}" +echo "${run_cmd[@]}" | tee -a "${OUTPUT_LOG}" # 4. Tell user where to find output printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" -# 5. Ignore the following strings on Intel XPU devices -# (otherwise they'll clutter up logs) -XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" +# # 5. Ignore the following strings on Intel XPU devices +# # (otherwise they'll clutter up logs) +# XPU_IGNORE_STRING="CCL_WARN|\ -\ INFO\ \-\ |real_accelerator\.py|numexpr\.utils|async_io|libaio" # if [[ $(ezpz_get_machine_name) == "aurora" ]]; then # module unload mpich && module load mpich # fi # # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} -eval "${run_cmd}" |& grep -E -v "${XPU_IGNORE_STRING}" |& tee -a "${OUTPUT_LOG}" +# eval "${run_cmd[@]}" |& tee -a "${OUTPUT_LOG}" +eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" From 43cde2b8f10171735dce68b899bd4a43f3158bfc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 13 Oct 2024 10:15:49 -0500 Subject: [PATCH 340/590] Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 12a05c52998..3686c6ceeb7 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -1,6 +1,7 @@ # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain GPT""" + import time from typing import Callable from mpi4py import MPI @@ -103,7 +104,7 @@ def model_provider(pre_process=True, post_process=True): with deepspeed_zero_init( data_parallel_group=dpg, remote_device=(None if args.remote_device == "none" else args.remote_device), - config_dict_or_path=args.deepspeed_config_dict, + config_dict_or_path=args.deepspeed_config, # _dict, enabled=args.zero_stage == 3, mpu=mpu, ): From 9f097339de3f9dacd856b4649fb696c5d0591010 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 13 Oct 2024 10:16:22 -0500 Subject: [PATCH 341/590] Update `megatron/data/{blendable,gpt,indexed}_dataset.py` --- megatron/data/blendable_dataset.py | 24 +- megatron/data/gpt_dataset.py | 885 +++++++++++++++++++---------- megatron/data/indexed_dataset.py | 217 ++++--- 3 files changed, 713 insertions(+), 413 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index ba2e00b1efe..590a379971d 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -6,14 +6,20 @@ import os import time +import logging import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0 +# from megatron import print_rank_0 from megatron.core import mpu from megatron.utils import Profile, PerfTrace from mpi4py import MPI + +from megatron.utils import get_logger + +log = get_logger(__name__, rank_zero_only=True) + dlp = Profile("DATASET") class BlendableDataset(torch.utils.data.Dataset): @dlp.log @@ -43,7 +49,7 @@ def _build_indices(): helpers.build_blending_indices(dataset_index, dataset_sample_index, weights, num_datasets, self.size, torch.distributed.get_rank() == 0) - print_rank_0('> elapsed time for building blendable dataset indices: ' + log.info('> elapsed time for building blendable dataset indices: ' '{:.2f} (sec)'.format(time.time() - start_time)) return dataset_index, dataset_sample_index @@ -68,7 +74,7 @@ def _build_indices(): ' dataset, building indices on rank 0 ...', flush=True) dataset_index, dataset_sample_index = _build_indices() try: - print_rank_0(" > saving index map files") + log.info(" > saving index map files") start_time = time.time() os.makedirs(os.path.dirname(index_path), exist_ok=True) with open(desc_path, 'wt') as fd: @@ -76,7 +82,7 @@ def _build_indices(): np.save(index_path, dataset_index, allow_pickle=True) np.save(sample_index_path, dataset_sample_index, allow_pickle=True) - print_rank_0(f" > finished saving index map files in {time.time() - start_time} seconds") + log.info(f" > finished saving index map files in {time.time() - start_time} seconds") except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_path})') print('or a file in it. This is set with the --data-cache-path argument. Please') @@ -93,7 +99,7 @@ def _build_indices(): torch.distributed.get_world_size() // torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group()) // torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): - print_rank_0("Data index creation unsuccessful, exiting.") + log.info("Data index creation unsuccessful, exiting.") exit() ''' torch.distributed.barrier(group=mpu.get_data_parallel_group()) @@ -101,13 +107,13 @@ def _build_indices(): torch.distributed.barrier(group=mpu.get_data_parallel_group()) start_time = time.time() - print_rank_0(f'> loading blendable dataset index: {index_path}') + log.info(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_index.size == self.size - print_rank_0(f'> loading blendable dataset sample index: {sample_index_path}') + log.info(f'> loading blendable dataset sample index: {sample_index_path}') self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size - print_rank_0(f'> finished loading in {time.time() - start_time} seconds') + log.info(f'> finished loading in {time.time() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() @@ -119,7 +125,7 @@ def _build_indices(): raise RuntimeError('BlendedDataset size is improperly bounded') except IndexError: pass - print_rank_0('> size of blendable dataset: ' + log.info('> size of blendable dataset: ' '{} samples'.format(self.size)) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 8501324752e..0a01ea31caa 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -9,67 +9,96 @@ import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0, is_rank_0, get_args +from megatron import is_rank_0, get_args from megatron.core import mpu -from megatron.data import helpers +from megatron.data import helpers # type:ignore from megatron.data.blendable_dataset import BlendableDataset -from megatron.data.dataset_utils import get_datasets_weights_and_num_samples, get_datasets_corpuses_weights_and_num_samples +from megatron.data.dataset_utils import ( + get_datasets_weights_and_num_samples, + get_datasets_corpuses_weights_and_num_samples, +) from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset -from megatron.utils import PerfTrace, Profile +from megatron.utils import PerfTrace, Profile, get_logger from mpi4py import MPI dlp = Profile("DATASET") +log = get_logger(__name__, rank_zero_only=True) + + @dlp.log -def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - train_data_prefix=None, - valid_data_prefix=None, - test_data_prefix=None, - return_doc_ids=False, *, - data_cache_path=None): +def build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + train_data_prefix=None, + valid_data_prefix=None, + test_data_prefix=None, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" if data_prefix: - print_rank_0("Single data path provided for train, valid & test") + log.debug("Single data path provided for train, valid & test") # Single dataset. if len(data_prefix) == 1: - return _build_train_valid_test_datasets(data_prefix[0], - data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + return _build_train_valid_test_datasets( + data_prefix[0], + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) # Blending dataset. # Parse the values. - output = get_datasets_corpuses_weights_and_num_samples(data_prefix, - train_valid_test_num_samples) + output = get_datasets_corpuses_weights_and_num_samples( + data_prefix, train_valid_test_num_samples + ) prefixes, corpuses, weights, datasets_train_valid_test_num_samples = output corpus_list = sorted(set(corpuses)) train_num_samples, valid_num_samples, test_num_samples = map( - sum, - zip(*datasets_train_valid_test_num_samples) + sum, zip(*datasets_train_valid_test_num_samples) ) class DatasetBuilder: - ''' + """ This is for building individual dataset from each dataset file - ''' + """ + @dlp.log - def __init__(self, prefix, corpus, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - return_doc_ids, - data_cache_path=data_cache_path, name='train'): + def __init__( + self, + prefix, + corpus, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + return_doc_ids, + data_cache_path=data_cache_path, + name="train", + ): self.prefix = prefix self.data_impl = data_impl self.splits_string = splits_string - if name == 'train': + if name == "train": self.num_samples = num_samples[0] - elif name == 'valid': + elif name == "valid": self.num_samples = num_samples[1] else: self.num_samples = num_samples[2] @@ -84,11 +113,21 @@ def __init__(self, prefix, corpus, data_impl, splits_string, self.desc = prefix + f"{self.num_samples}" + f"{seq_length}" + f"{seed}" self.build = False self.corpus = corpus + @dlp.log def Build(self): - self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, - self.num_samples_train_valid_test, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, - data_cache_path=self.data_cache_path) + self.dataset = _build_train_valid_test_datasets_single( + self.prefix, + self.data_impl, + self.splits_string, + self.num_samples_train_valid_test, + self.seq_length, + self.seed, + self.skip_warmup, + self.name, + self.return_doc_ids, + data_cache_path=self.data_cache_path, + ) self.build = True return self.dataset @@ -98,21 +137,27 @@ def __init__(self, dataset_builders): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) - self.indices=np.zeros((self.num_samples, 2), dtype=np.uint64) - self.desc="ConcatDataset:" - m = 0 + self.indices = np.zeros((self.num_samples, 2), dtype=np.uint64) + self.desc = "ConcatDataset:" + # m = 0 num_samples_list = np.array([d.num_samples for d in dataset_builders]) self.num_samples = np.sum(num_samples_list) + def _build_indices(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) - helpers.build_concat_indices(dataset_index, dataset_sample_index, - num_samples_list, - self.num_datasets, - torch.distributed.get_rank()==0) - print_rank_0('> elapsed time for building concat dataset indices: ' - '{:.2f} (sec)'.format(time.time() - start_time)) + helpers.build_concat_indices( + dataset_index, + dataset_sample_index, + num_samples_list, + self.num_datasets, + torch.distributed.get_rank() == 0, + ) + log.debug( + "> elapsed time for building concat dataset indices: " + "{:.2f} (sec)".format(time.time() - start_time) + ) return dataset_index, dataset_sample_index self.dataset_index, self.dataset_sample_index = _build_indices() @@ -122,7 +167,12 @@ def _build_indices(): for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," - self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" + self.desc += ( + f"-{self.num_samples}" + + f"-{dataset_builders[0].seq_length}" + + f"{dataset_builders[0].seed}" + ) + def __len__(self): return self.num_samples @@ -135,227 +185,340 @@ def __getitem__(self, idx): return self.dataset_builders[i].dataset[j] else: return self.dataset_builders[i].Build()[j] - - # Predetermine whether need to build the specific dataset or not. + # Predetermine whether need to build the specific dataset or not. start_time = time.time() - print_rank_0(" >>> Started building datasets in distributed way ... ") + log.debug(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] - + train_datasets = [] valid_datasets = [] test_datasets = [] # Build individual datasets. @dlp.log - def build_corpus_datasets(dataset_type='train'): + def build_corpus_datasets(dataset_type="train"): start_time = time.time() - print_rank_0(f" >>> Building {dataset_type} corpus datasets ...") + log.debug(f" >>> Building {dataset_type} corpus datasets ...") datasets = [] corpus_builders = {} corpus_weights = {} for c in corpus_list: corpus_builders[c] = [] corpus_weights[c] = 0.0 - dataset_builders = [DatasetBuilder(prefixes[i], corpuses[i], data_impl, splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] - for i in range(torch.distributed.get_rank()//mpu.get_tensor_model_parallel_world_size(), len(weights), torch.distributed.get_world_size()//mpu.get_tensor_model_parallel_world_size()): + dataset_builders = [ + DatasetBuilder( + prefixes[i], + corpuses[i], + data_impl, + splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, + seed, + skip_warmup, + return_doc_ids, + data_cache_path, + dataset_type, + ) + for i in range(len(weights)) + ] + for i in range( + torch.distributed.get_rank() + // mpu.get_tensor_model_parallel_world_size(), + len(weights), + torch.distributed.get_world_size() + // mpu.get_tensor_model_parallel_world_size(), + ): dataset_builders[i].Build() - print_rank_0(f" >>> Finished building individual datasets in {time.time() - start_time} seconds") + log.debug( + f" >>> Finished building individual datasets in {time.time() - start_time} seconds" + ) start_concating_time = time.time() for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] total = 0 - print_rank_0(" > number of samples for each corpus ") - corpus_weights_achieved={} + log.debug(" > number of samples for each corpus ") + corpus_weights_achieved = {} for c in corpus_list: datasets.append(BuildConcatDataset(corpus_builders[c])) total += datasets[-1].num_samples - corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples - print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") - - print_rank_0(f" > total number of samples: {total}") - print_rank_0(f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds") - print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") + corpus_weights_achieved[c] = ( + float(datasets[-1].num_samples) / train_num_samples + ) + log.debug( + f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})" + ) + + log.debug(f" > total number of samples: {total}") + log.debug( + f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds" + ) + log.debug( + f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds" + ) return datasets, [corpus_weights_achieved[c] for c in corpus_list] + train_weights = None if a > 0: - train_datasets, train_weights = build_corpus_datasets('train') - + train_datasets, train_weights = build_corpus_datasets("train") + valid_weights = None if b > 0: - valid_datasets, valid_weights = build_corpus_datasets('valid') - - if c > 0: - test_datasets, test_weights = build_corpus_datasets('test') + valid_datasets, valid_weights = build_corpus_datasets("valid") + test_weights = None + if c > 0: + test_datasets, test_weights = build_corpus_datasets("test") # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them - print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") + log.debug( + f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds" + ) torch.distributed.barrier(group=mpu.get_data_parallel_group()) torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) - print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") - print_rank_0(f" >>> Starting to build BlendableDataset") + log.debug( + f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds" + ) + log.debug(" >>> Starting to build BlendableDataset") # Blend. start_time = time.time() blending_train_dataset = None - if train_datasets: - blending_train_dataset = BlendableDataset(train_datasets, train_weights, train_num_samples, - data_cache_path=data_cache_path) + if train_datasets and train_weights: + blending_train_dataset = BlendableDataset( + train_datasets, + train_weights, + train_num_samples, + data_cache_path=data_cache_path, + ) blending_valid_dataset = None - if valid_datasets: - blending_valid_dataset = BlendableDataset(valid_datasets, valid_weights, valid_num_samples, - data_cache_path=data_cache_path) + if valid_datasets and valid_weights: + blending_valid_dataset = BlendableDataset( + valid_datasets, + valid_weights, + valid_num_samples, + data_cache_path=data_cache_path, + ) blending_test_dataset = None - if test_datasets: - blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, - data_cache_path=data_cache_path) + if test_datasets and test_weights: + blending_test_dataset = BlendableDataset( + test_datasets, + test_weights, + test_num_samples, + data_cache_path=data_cache_path, + ) end_time = time.time() - print_rank_0(f" >>> Finished building BlendableDataset in {end_time - start_time} seconds") - return (blending_train_dataset, blending_valid_dataset, - blending_test_dataset) + log.debug( + f" >>> Finished building BlendableDataset in {end_time - start_time} seconds" + ) + return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) else: - print_rank_0("Separate data paths provided for train, valid & test. Split string will be ignored.") + log.debug( + "Separate data paths provided for train, valid & test. Split string will be ignored." + ) train_dataset, valid_dataset, test_dataset = None, None, None # Single dataset. if train_data_prefix is not None: - train_dataset = build_dataset("train", train_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[0], - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + train_dataset = build_dataset( + "train", + train_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[0], + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) if valid_data_prefix is not None: - valid_dataset = build_dataset("valid", valid_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[1], - seq_length, seed, False, - data_cache_path=data_cache_path) - + valid_dataset = build_dataset( + "valid", + valid_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[1], + seq_length, + seed, + False, + data_cache_path=data_cache_path, + ) if test_data_prefix is not None: - test_dataset = build_dataset("test", test_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[2], - seq_length, seed, False, - data_cache_path=data_cache_path) + test_dataset = build_dataset( + "test", + test_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[2], + seq_length, + seed, + False, + data_cache_path=data_cache_path, + ) return (train_dataset, valid_dataset, test_dataset) + @dlp.log -def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - return_doc_ids=False, *, - data_cache_path=None): +def _build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + log.debug(" > dataset split:") def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' - 'documents'.format(splits[index], splits[index + 1], - splits[index + 1] - splits[index])) - print_split_stats('train', 0) - print_split_stats('validation', 1) - print_split_stats('test', 2) + log.debug(" {}:".format(name)) + log.debug( + " document indices in [{}, {}) total of {} " "documents".format( + splits[index], splits[index + 1], splits[index + 1] - splits[index] + ) + ) + + print_split_stats("train", 0) + print_split_stats("validation", 1) + print_split_stats("test", 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange(start=splits[index], stop=splits[index + 1], - step=1, dtype=np.int32) - dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, seed, - return_doc_ids, - data_cache_path=data_cache_path) + documents = np.arange( + start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 + ) + dataset = GPTDataset( + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, + seed, + return_doc_ids, + data_cache_path=data_cache_path, + ) return dataset - train_dataset = build_dataset(0, 'train') - valid_dataset = build_dataset(1, 'valid') - test_dataset = build_dataset(2, 'test') + train_dataset = build_dataset(0, "train") + valid_dataset = build_dataset(1, "valid") + test_dataset = build_dataset(2, "test") return (train_dataset, valid_dataset, test_dataset) + @dlp.log -def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, name, - return_doc_ids=False, *, - data_cache_path=None): +def _build_train_valid_test_datasets_single( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + name, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" # Each rank print out information - print_rank_0(f" >> building dataset for {data_prefix}") + log.debug(f" >> building dataset for {data_prefix}") # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + log.debug(" > dataset split:") def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' - 'documents'.format(splits[index], splits[index + 1], - splits[index + 1] - splits[index])) - print_split_stats('train', 0) - print_split_stats('validation', 1) - print_split_stats('test', 2) + log.debug(" {}:".format(name)) + log.debug( + " document indices in [{}, {}) total of {} " "documents".format( + splits[index], splits[index + 1], splits[index + 1] - splits[index] + ) + ) + + print_split_stats("train", 0) + print_split_stats("validation", 1) + print_split_stats("test", 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange(start=splits[index], stop=splits[index + 1], - step=1, dtype=np.int32) - dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, seed, - return_doc_ids, - data_cache_path=data_cache_path) + documents = np.arange( + start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 + ) + dataset = GPTDataset( + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, + seed, + return_doc_ids, + data_cache_path=data_cache_path, + ) return dataset - if name.find("train")!=-1: - return build_dataset(0, 'train') - if name.find("valid")!=-1: - return build_dataset(1, 'valid') - if name.find("test")!=-1: - return build_dataset(2, 'test') + + if name.find("train") != -1: + return build_dataset(0, "train") + if name.find("valid") != -1: + return build_dataset(1, "valid") + if name.find("test") != -1: + return build_dataset(2, "test") + @dlp.log -def build_dataset(dataset_name, data_prefix, data_impl, - splits_string, num_samples, - seq_length, seed, skip_warmup, - *, - data_cache_path=None): +def build_dataset( + dataset_name, + data_prefix, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + *, + data_cache_path=None, +): dataset = None if len(data_prefix) == 1: - dataset = _build_dataset(dataset_name, data_prefix[0], data_impl, - splits_string, num_samples, seq_length, - seed, skip_warmup, - data_cache_path=data_cache_path) + dataset = _build_dataset( + dataset_name, + data_prefix[0], + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) else: # Blending dataset. # Parse the values. @@ -366,73 +529,108 @@ def build_dataset(dataset_name, data_prefix, data_impl, # Build individual datasets. datasets = [] for i in range(len(prefixes)): - ds = _build_dataset(dataset_name, prefixes[i], data_impl, - splits_string, dataset_num_samples[i], - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + ds = _build_dataset( + dataset_name, + prefixes[i], + data_impl, + splits_string, + dataset_num_samples[i], + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) if ds: datasets.append(ds) if datasets: - dataset = BlendableDataset(datasets, weights, num_samples, - data_cache_path=data_cache_path) + dataset = BlendableDataset( + datasets, weights, num_samples, data_cache_path=data_cache_path + ) return dataset + @dlp.log -def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - *, - data_cache_path=None): +def _build_dataset( + dataset_name, + data_prefix, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + *, + data_cache_path=None, +): """ Build dataset. This method is called when individual train, valid, test datasets are provided """ # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] - print_rank_0(' {}:'.format(dataset_name)) - print_rank_0(' document indices in [0, {}) total of {} ' - 'documents'.format(total_num_of_documents, total_num_of_documents)) - - documents = np.arange(start=0, stop=total_num_of_documents, - step=1, dtype=np.int32) - - dataset = GPTDataset(dataset_name, data_prefix, documents, indexed_dataset, - splits_string, num_samples, seq_length, seed, - data_cache_path=data_cache_path) + log.debug(" {}:".format(dataset_name)) + log.debug( + " document indices in [0, {}) total of {} " "documents".format( + total_num_of_documents, total_num_of_documents + ) + ) + + documents = np.arange(start=0, stop=total_num_of_documents, step=1, dtype=np.int32) + + dataset = GPTDataset( + dataset_name, + data_prefix, + documents, + indexed_dataset, + splits_string, + num_samples, + seq_length, + seed, + data_cache_path=data_cache_path, + ) return dataset + @dlp.log def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): """Build indexed dataset.""" - print_rank_0(' > building dataset index ...') + log.debug(" > building dataset index ...") start_time = time.time() - indexed_dataset = make_indexed_dataset(data_prefix, - data_impl, - skip_warmup) - print_rank_0(' > finished creating indexed dataset in {:4f} ' - 'seconds'.format(time.time() - start_time)) - print_rank_0(' number of documents: {}'.format( - indexed_dataset.sizes.shape[0])) + indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) + log.debug( + " > finished creating indexed dataset in {:4f} " "seconds".format( + time.time() - start_time + ) + ) + log.debug(" number of documents: {}".format(indexed_dataset.sizes.shape[0])) return indexed_dataset class GPTDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, name, data_prefix, documents, indexed_dataset, - splits_string, num_samples, seq_length, seed, - return_doc_ids=False, *, - data_cache_path=None): - + def __init__( + self, + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + num_samples, + seq_length, + seed, + return_doc_ids=False, + *, + data_cache_path=None, + ): self.name = name self.indexed_dataset = indexed_dataset self.return_doc_ids = return_doc_ids @@ -442,20 +640,29 @@ def __init__(self, name, data_prefix, documents, indexed_dataset, assert np.max(documents) < indexed_dataset.sizes.shape[0] # Build index mappings. - self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = \ - _build_index_mappings(self.name, data_prefix, - documents, self.indexed_dataset.sizes, - splits_string, num_samples, seq_length, seed, - data_cache_path=data_cache_path) - + self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = ( + _build_index_mappings( + self.name, + data_prefix, + documents, + self.indexed_dataset.sizes, + splits_string, + num_samples, + seq_length, + seed, + data_cache_path=data_cache_path, + ) + ) def __len__(self): # -1 is due to data structure used to retieve the index: # sample i --> [sample_idx[i], sample_idx[i+1]) return self.sample_idx.shape[0] - 1 + @dlp.log def __getitem__(self, idx): args = get_args() + assert args is not None orig_idx = idx # Get the shuffled index. try: @@ -464,21 +671,24 @@ def __getitem__(self, idx): if is_rank_0(): import json from rich import print_json + print(exc) print( - '\n'.join( - ['-------------------------------------------------', - f'Trying to access {idx=} from self.shuffle_idx,', - f'but {len(self.shuffle_idx)=}', - '-------------------------------------------------'] + "\n".join( + [ + "-------------------------------------------------", + f"Trying to access {idx=} from self.shuffle_idx,", + f"but {len(self.shuffle_idx)=}", + "-------------------------------------------------", + ] ) ) print_json( json.dumps( { - 'doc_idx': len(self.doc_idx), - 'sample_idx': len(self.sample_idx), - 'shuffle_idx': len(self.shuffle_idx), + "doc_idx": len(self.doc_idx), + "sample_idx": len(self.sample_idx), + "shuffle_idx": len(self.shuffle_idx), }, indent=4, ) @@ -492,45 +702,57 @@ def __getitem__(self, idx): doc_ids = [] if doc_index_f == doc_index_l: doc_ids.append(self.doc_idx[doc_index_f]) - sample = self.indexed_dataset.get(self.doc_idx[doc_index_f], - offset=offset_f, - length=offset_l - offset_f + 1) + sample = self.indexed_dataset.get( + self.doc_idx[doc_index_f], + offset=offset_f, + length=offset_l - offset_f + 1, + ) else: # Otherwise, get the rest of the initial document. doc_ids.append(self.doc_idx[doc_index_f]) - sample_list = [self.indexed_dataset.get(self.doc_idx[doc_index_f], - offset=offset_f)] + sample_list = [ + self.indexed_dataset.get(self.doc_idx[doc_index_f], offset=offset_f) + ] # Loop over all in between documents and add the entire document. for i in range(doc_index_f + 1, doc_index_l): doc_ids.append(self.doc_idx[i]) sample_list.append(self.indexed_dataset.get(self.doc_idx[i])) # And finally add the relevant portion of last document. doc_ids.append(self.doc_idx[doc_index_l]) - sample_list.append(self.indexed_dataset.get( - self.doc_idx[doc_index_l], - length=offset_l + 1)) + sample_list.append( + self.indexed_dataset.get(self.doc_idx[doc_index_l], length=offset_l + 1) + ) sample = np.concatenate(sample_list) - text_name = 'text' + text_name = "text" if args.use_dataset_only: - text_name = 'input_ids' + text_name = "input_ids" sample_dict = {text_name: np.array(sample, dtype=np.int64)} if args.return_data_index: - sample_dict.update({'index': np.array([orig_idx], dtype=np.int64)}) + sample_dict.update({"index": np.array([orig_idx], dtype=np.int64)}) - if self.return_doc_ids: # for retro preprocessing - sample_dict.update({'doc_ids': np.array(doc_ids, dtype=np.int64)}) + if self.return_doc_ids: # for retro preprocessing + sample_dict.update({"doc_ids": np.array(doc_ids, dtype=np.int64)}) if args.use_dataset_only: - sample_dict.update({'labels': np.array(sample, dtype=np.int64)}) + sample_dict.update({"labels": np.array(sample, dtype=np.int64)}) return sample_dict + @dlp.log -def _build_index_mappings(name, data_prefix, documents, sizes, - splits_string, num_samples, seq_length, seed, - *, - data_cache_path): +def _build_index_mappings( + name, + data_prefix, + documents, + sizes, + splits_string, + num_samples, + seq_length, + seed, + *, + data_cache_path, +): """Build doc-idx, sample-idx, and shuffle-idx. doc-idx: is an array (ordered) of documents to be used in training. sample-idx: is the start document index and document offset for each @@ -538,10 +760,11 @@ def _build_index_mappings(name, data_prefix, documents, sizes, shuffle-idx: maps the sample index into a random index into sample-idx. """ args = get_args() + assert args is not None # Number of tokens in each epoch and number of required epochs. tokens_per_epoch = _num_tokens(documents, sizes) num_epochs = _num_epochs(tokens_per_epoch, seq_length, num_samples) - if args.train_data_exact_num_epochs is not None and name == 'train': + if args.train_data_exact_num_epochs is not None and name == "train": num_epochs = args.train_data_exact_num_epochs # rng state @@ -556,13 +779,13 @@ def _build_index_mappings(name, data_prefix, documents, sizes, desc += f"Sequence length {seq_length}\n" desc += f"Random seed {seed}\n" desc += f"Split {splits_string}\n" - desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() + desc_hash = hashlib.md5(desc.encode("utf-8")).hexdigest() desc_filename = desc_hash + ".dsc" - doc_idx_filename = desc_hash + '_doc_idx.npy' - sample_idx_filename = desc_hash + '_sample_idx.npy' - shuffle_idx_filename = desc_hash + '_shuffle_idx.npy' + doc_idx_filename = desc_hash + "_doc_idx.npy" + sample_idx_filename = desc_hash + "_sample_idx.npy" + shuffle_idx_filename = desc_hash + "_shuffle_idx.npy" - if name == 'train': + if name == "train": # force to use certain index files if args.train_desc_path is not None: desc_filename = args.train_desc_path @@ -577,15 +800,15 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # duplication, then look in data-cache-path if specified, # If nothing is found, use the last path looked in build_indices = True - prefixes = [os.path.join(os.path.dirname(data_prefix), 'index-cache')] + prefixes = [os.path.join(os.path.dirname(data_prefix), "index-cache")] if data_cache_path is not None: prefixes.append(data_cache_path) for prefix in prefixes: idx_path = { - 'desc': os.path.join(prefix, desc_filename), - 'doc': os.path.join(prefix, doc_idx_filename), - 'sample': os.path.join(prefix, sample_idx_filename), - 'shuffle': os.path.join(prefix, shuffle_idx_filename) + "desc": os.path.join(prefix, desc_filename), + "doc": os.path.join(prefix, doc_idx_filename), + "sample": os.path.join(prefix, sample_idx_filename), + "shuffle": os.path.join(prefix, shuffle_idx_filename), } for f in idx_path.values(): if not os.path.isfile(f): @@ -594,15 +817,17 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Found our files! build_indices = False break - data_cache_dir = os.path.dirname(idx_path['desc']) + data_cache_dir = os.path.dirname(idx_path["desc"]) data_cache_success = True # Build the indexed mapping if not exist. if build_indices: - # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the - # ranks will first create the document files, and then read it. + # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the + # ranks will first create the document files, and then read it. # There will not be contension effects going on either - print_rank_0(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") + log.warning( + f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}" + ) # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -611,64 +836,80 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # not mean anything. if num_epochs == 1: separate_last_epoch = False - print_rank_0(' > only one epoch required, setting ' - 'separate_last_epoch to False') + log.debug( + " > only one epoch required, setting " "separate_last_epoch to False" + ) else: # Get the number of samples for the last epoch num_samples_from_epochs_minus_one = ( - (num_epochs - 1) * tokens_per_epoch - 1) // seq_length - last_epoch_num_samples = num_samples - \ - num_samples_from_epochs_minus_one - assert last_epoch_num_samples >= 0, \ - 'last epoch number of samples should be non-negative.' + (num_epochs - 1) * tokens_per_epoch - 1 + ) // seq_length + last_epoch_num_samples = num_samples - num_samples_from_epochs_minus_one + assert ( + last_epoch_num_samples >= 0 + ), "last epoch number of samples should be non-negative." num_samples_per_epoch = (tokens_per_epoch - 1) // seq_length - assert last_epoch_num_samples <= (num_samples_per_epoch + 1), \ - 'last epoch number of samples exceeded max value.' + assert last_epoch_num_samples <= ( + num_samples_per_epoch + 1 + ), "last epoch number of samples exceeded max value." # If we have less than 80% of the samples for the last epoch, # seperate out the epoch and treat it differently. # Note: the 80% number is just based on common sense and can # be adjusted if needed. - separate_last_epoch = (last_epoch_num_samples < - int(0.80 * num_samples_per_epoch)) + separate_last_epoch = last_epoch_num_samples < int( + 0.80 * num_samples_per_epoch + ) if separate_last_epoch: - string = ' > last epoch number of samples ({}) is smaller '\ - 'than 80% of number of samples per epoch ({}), '\ - 'setting separate_last_epoch to True' + string = ( + " > last epoch number of samples ({}) is smaller " + "than 80% of number of samples per epoch ({}), " + "setting separate_last_epoch to True" + ) else: - string = ' > last epoch number of samples ({}) is larger '\ - 'than 80% of number of samples per epoch ({}), '\ - 'setting separate_last_epoch to False' - print_rank_0(string.format(last_epoch_num_samples, - num_samples_per_epoch)) - + string = ( + " > last epoch number of samples ({}) is larger " + "than 80% of number of samples per epoch ({}), " + "setting separate_last_epoch to False" + ) + log.debug(string.format(last_epoch_num_samples, num_samples_per_epoch)) try: os.makedirs(data_cache_dir, exist_ok=True) # description - with open(idx_path['desc'], 'wt') as fd: + with open(idx_path["desc"], "wt") as fd: fd.write(desc) # doc-idx. start_time = time.time() - doc_idx = _build_doc_idx(documents, num_epochs, np_rng, - separate_last_epoch) - np.save(idx_path['doc'], doc_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save doc-idx mapping ' - '(seconds): {:4f}'.format(time.time() - start_time)) + doc_idx = _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch) + np.save(idx_path["doc"], doc_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save doc-idx mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # sample-idx. start_time = time.time() # Use C++ implementation for speed. # First compile and then import. from megatron.data import helpers + assert doc_idx.dtype == np.int32 assert sizes.dtype == np.int32 - sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, - num_epochs, tokens_per_epoch, torch.distributed.get_rank()==0) - np.save(idx_path['sample'], sample_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save sample-idx mapping ' - '(seconds): {:4f}'.format(time.time() - start_time)) + sample_idx = helpers.build_sample_idx( + sizes, + doc_idx, + seq_length, + num_epochs, + tokens_per_epoch, + torch.distributed.get_rank() == 0, + ) + np.save(idx_path["sample"], sample_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save sample-idx mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # shuffle-idx. start_time = time.time() # -1 is due to data structure used to retieve the index: @@ -677,35 +918,46 @@ def _build_index_mappings(name, data_prefix, documents, sizes, num_samples_ = num_samples_from_epochs_minus_one else: num_samples_ = sample_idx.shape[0] - 1 - shuffle_idx = _build_shuffle_idx(num_samples_, - sample_idx.shape[0] - 1, np_rng) - np.save(idx_path['shuffle'], shuffle_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save shuffle-idx mapping' - ' (seconds): {:4f}'.format(time.time() - start_time)) + shuffle_idx = _build_shuffle_idx( + num_samples_, sample_idx.shape[0] - 1, np_rng + ) + np.save(idx_path["shuffle"], shuffle_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save shuffle-idx mapping" + " (seconds): {:4f}".format(time.time() - start_time) + ) except OSError: - print(f'There was an error trying to create the data cache directory ({data_cache_dir})') - print('or a file in it. This defaults to a directory "index-cache" within the directory') - print('the data files are in and can be set with the --data-cache-path argument. Please') - print('ensure you have write access to this directory or specify one that you do have') - print('write access to.') + print( + f"There was an error trying to create the data cache directory ({data_cache_dir})" + ) + print( + 'or a file in it. This defaults to a directory "index-cache" within the directory' + ) + print( + "the data files are in and can be set with the --data-cache-path argument. Please" + ) + print( + "ensure you have write access to this directory or specify one that you do have" + ) + print("write access to.") data_cache_success = False # Load mappings. start_time = time.time() - print_rank_0(f" > loading doc-idx mapping from {idx_path['doc']}") - doc_idx = np.load(idx_path['doc'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading doc-idx mapping from {idx_path['doc']}") + doc_idx = np.load(idx_path["doc"], allow_pickle=True, mmap_mode="r") - print_rank_0(f" > loading sample-idx mapping from {idx_path['sample']}") - sample_idx = np.load(idx_path['sample'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading sample-idx mapping from {idx_path['sample']}") + sample_idx = np.load(idx_path["sample"], allow_pickle=True, mmap_mode="r") - print_rank_0(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") - shuffle_idx = np.load(idx_path['shuffle'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") + shuffle_idx = np.load(idx_path["shuffle"], allow_pickle=True, mmap_mode="r") - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( - time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( - sample_idx.shape[0])) - print_rank_0(' total number of epochs: {}'.format(num_epochs)) + log.debug( + " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) + ) + log.debug(" total number of samples: {}".format(sample_idx.shape[0])) + log.debug(" total number of epochs: {}".format(num_epochs)) return doc_idx, sample_idx, shuffle_idx, desc, desc_hash @@ -729,25 +981,26 @@ def _num_epochs(tokens_per_epoch, seq_length, num_samples): if ((total_tokens - 1) // seq_length) >= num_samples: return num_epochs + @dlp.log def _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch): """Build an array with length = number-of-epochs * number-of-dcuments. Each index is mapped to a corresponding document.""" if not separate_last_epoch or num_epochs == 1: - doc_idx = np.mgrid[0:num_epochs, 0:len(documents)][1] + doc_idx = np.mgrid[0:num_epochs, 0 : len(documents)][1] doc_idx[:] = documents doc_idx = doc_idx.reshape(-1) doc_idx = doc_idx.astype(np.int32) np_rng.shuffle(doc_idx) return doc_idx - doc_idx_first = _build_doc_idx(documents, num_epochs-1, np_rng, False) + doc_idx_first = _build_doc_idx(documents, num_epochs - 1, np_rng, False) doc_idx_last = _build_doc_idx(documents, 1, np_rng, False) return np.concatenate((doc_idx_first, doc_idx_last)) + @dlp.log -def _build_sample_idx(sizes, doc_idx, seq_length, - num_epochs, tokens_per_epoch): +def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): """Sample index mapping is a 2D array with sizes [number-of-samples + 1, 2] where [..., 0] contains the index into `doc_idx` and [..., 1] is the @@ -781,7 +1034,7 @@ def _build_sample_idx(sizes, doc_idx, seq_length, # Note that -1 here is for the same reason we have -1 in # `_num_epochs` calculations. if remaining_seq_length <= 0: - doc_offset += (remaining_seq_length + doc_length - 1) + doc_offset += remaining_seq_length + doc_length - 1 remaining_seq_length = 0 else: # Otherwise, start from the begining of the next document. @@ -794,24 +1047,28 @@ def _build_sample_idx(sizes, doc_idx, seq_length, return sample_idx + @dlp.log def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" - print_rank_0(' > building shuffle index with split [0, {}) and [{}, {}) ' - '...'.format(num_samples, num_samples, total_size)) + log.debug( + " > building shuffle index with split [0, {}) and [{}, {}) " "...".format( + num_samples, num_samples, total_size + ) + ) dtype_ = np.uint32 if total_size >= (np.iinfo(np.uint32).max - 1): dtype_ = np.int64 - shuffle_idx_first = np.arange(start=0, stop=num_samples, - step=1, dtype=dtype_) + shuffle_idx_first = np.arange(start=0, stop=num_samples, step=1, dtype=dtype_) np_rng.shuffle(shuffle_idx_first) if num_samples == total_size: return shuffle_idx_first - shuffle_idx_last = np.arange(start=num_samples, stop=total_size, - step=1, dtype=dtype_) + shuffle_idx_last = np.arange( + start=num_samples, stop=total_size, step=1, dtype=dtype_ + ) np_rng.shuffle(shuffle_idx_last) return np.concatenate((shuffle_idx_first, shuffle_idx_last)) diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index 1eb9b7842bb..8479f33fab3 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -15,17 +15,24 @@ from functools import lru_cache import os + +# import logging import shutil import struct from itertools import accumulate import numpy as np import torch -from megatron import print_rank_0 -from megatron.utils import Profile + +# from megatron import print_rank_0 +from megatron.utils import Profile, get_logger + +log = get_logger(__name__) + dlp = Profile("DATASET") + def __best_fitting_dtype(vocab_size=None): if vocab_size is not None and vocab_size < 65500: return np.uint16 @@ -34,28 +41,32 @@ def __best_fitting_dtype(vocab_size=None): def get_available_dataset_impl(): - return ['lazy', 'cached', 'mmap'] + return ["lazy", "cached", "mmap"] def infer_dataset_impl(path): if IndexedDataset.exists(path): - with open(index_file_path(path), 'rb') as f: + with open(index_file_path(path), "rb") as f: magic = f.read(8) if magic == IndexedDataset._HDR_MAGIC: - return 'cached' + return "cached" elif magic == MMapIndexedDataset.Index._HDR_MAGIC[:8]: - return 'mmap' + return "mmap" else: return None else: print(f"Dataset does not exist: {path}") - print("Path should be a basename that both .idx and .bin can be appended to get full filenames.") + print( + "Path should be a basename that both .idx and .bin can be appended to get full filenames." + ) return None def make_builder(out_file, impl, vocab_size=None): - if impl == 'mmap': - return MMapIndexedDatasetBuilder(out_file, dtype=__best_fitting_dtype(vocab_size)) + if impl == "mmap": + return MMapIndexedDatasetBuilder( + out_file, dtype=__best_fitting_dtype(vocab_size) + ) else: return IndexedDatasetBuilder(out_file) @@ -63,22 +74,24 @@ def make_builder(out_file, impl, vocab_size=None): def make_dataset(path, impl, skip_warmup=False): if not IndexedDataset.exists(path): print(f"Dataset does not exist: {path}") - print("Path should be a basename that both .idx and .bin can be appended to get full filenames.") + print( + "Path should be a basename that both .idx and .bin can be appended to get full filenames." + ) return None - if impl == 'infer': + if impl == "infer": impl = infer_dataset_impl(path) - if impl == 'lazy' and IndexedDataset.exists(path): + if impl == "lazy" and IndexedDataset.exists(path): return IndexedDataset(path) - elif impl == 'cached' and IndexedDataset.exists(path): + elif impl == "cached" and IndexedDataset.exists(path): return IndexedCachedDataset(path) - elif impl == 'mmap' and MMapIndexedDataset.exists(path): + elif impl == "mmap" and MMapIndexedDataset.exists(path): return MMapIndexedDataset(path, skip_warmup) print(f"Unknown dataset implementation: {impl}") return None def dataset_exists(path, impl): - if impl == 'mmap': + if impl == "mmap": return MMapIndexedDataset.exists(path) else: return IndexedDataset.exists(path) @@ -114,11 +127,11 @@ def code(dtype): def index_file_path(prefix_path): - return prefix_path + '.idx' + return prefix_path + ".idx" def data_file_path(prefix_path): - return prefix_path + '.bin' + return prefix_path + ".bin" def create_doc_idx(sizes): @@ -131,38 +144,41 @@ def create_doc_idx(sizes): class IndexedDataset(torch.utils.data.Dataset): """Loader for IndexedDataset""" - _HDR_MAGIC = b'TNTIDX\x00\x00' + + _HDR_MAGIC = b"TNTIDX\x00\x00" def __init__(self, path): super().__init__() self.path = path self.data_file = None self.read_index(path) + @dlp.log def read_index(self, path): - with open(index_file_path(path), 'rb') as f: + with open(index_file_path(path), "rb") as f: magic = f.read(8) assert magic == self._HDR_MAGIC, ( - 'Index file doesn\'t match expected format. ' - 'Make sure that --dataset-impl is configured properly.' + "Index file doesn't match expected format. " + "Make sure that --dataset-impl is configured properly." ) version = f.read(8) - assert struct.unpack('= self._len: - raise IndexError('index out of range') + raise IndexError("index out of range") def __del__(self): if self.data_file: @@ -176,7 +192,7 @@ def __getitem__(self, idx): if isinstance(idx, int): i = idx self.check_index(i) - tensor_size = self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + tensor_size = self.sizes[self.dim_offsets[i] : self.dim_offsets[i + 1]] a = np.empty(tensor_size, dtype=self.dtype) self.data_file.seek(self.data_offsets[i] * self.element_size) self.data_file.readinto(a) @@ -185,7 +201,7 @@ def __getitem__(self, idx): start, stop, step = idx.indices(len(self)) if step != 1: raise ValueError("Slices into indexed_dataset must be contiguous") - sizes = self.sizes[self.dim_offsets[start]:self.dim_offsets[stop]] + sizes = self.sizes[self.dim_offsets[start] : self.dim_offsets[stop]] size = sum(sizes) a = np.empty(size, dtype=self.dtype) self.data_file.seek(self.data_offsets[start] * self.element_size) @@ -205,8 +221,8 @@ def size(self, index): @staticmethod def exists(path): - return ( - os.path.exists(index_file_path(path)) and os.path.exists(data_file_path(path)) + return os.path.exists(index_file_path(path)) and os.path.exists( + data_file_path(path) ) @property @@ -215,7 +231,6 @@ def supports_prefetch(self): class IndexedCachedDataset(IndexedDataset): - def __init__(self, path): super().__init__(path) self.cache = None @@ -224,6 +239,7 @@ def __init__(self, path): @property def supports_prefetch(self): return True + @dlp.log def prefetch(self, indices): if all(i in self.cache_index for i in indices): @@ -240,7 +256,7 @@ def prefetch(self, indices): for i in indices: self.cache_index[i] = ptx size = self.data_offsets[i + 1] - self.data_offsets[i] - a = self.cache[ptx: ptx + size] + a = self.cache[ptx : ptx + size] self.data_file.seek(self.data_offsets[i] * self.element_size) self.data_file.readinto(a) ptx += size @@ -255,10 +271,10 @@ def __getitem__(self, idx): if isinstance(idx, int): i = idx self.check_index(i) - tensor_size = self.sizes[self.dim_offsets[i]:self.dim_offsets[i + 1]] + tensor_size = self.sizes[self.dim_offsets[i] : self.dim_offsets[i + 1]] a = np.empty(tensor_size, dtype=self.dtype) ptx = self.cache_index[i] - np.copyto(a, self.cache[ptx: ptx + a.size]) + np.copyto(a, self.cache[ptx : ptx + a.size]) return a elif isinstance(idx, slice): # Hack just to make this work, can optimizer later if necessary @@ -278,15 +294,17 @@ class IndexedDatasetBuilder(object): np.float32: 4, np.float64: 8, } + @dlp.log def __init__(self, out_file, dtype=np.int32): - self.out_file = open(out_file, 'wb') + self.out_file = open(out_file, "wb") self.dtype = dtype self.data_offsets = [0] self.dim_offsets = [0] self.sizes = [] self.element_size = self.element_sizes[self.dtype] self.doc_idx = [0] + @dlp.log def add_item(self, tensor): bytes = self.out_file.write(np.array(tensor.numpy(), dtype=self.dtype)) @@ -297,6 +315,7 @@ def add_item(self, tensor): def end_document(self): self.doc_idx.append(len(self.sizes)) + @dlp.log def merge_file_(self, another_file): index = IndexedDataset(another_file) @@ -315,7 +334,7 @@ def merge_file_(self, another_file): self.doc_idx.extend((doc_offset + index.doc_idx)[1:]) - with open(data_file_path(another_file), 'rb') as f: + with open(data_file_path(another_file), "rb") as f: while True: data = f.read(1024) if data: @@ -325,21 +344,22 @@ def merge_file_(self, another_file): def finalize(self, index_file): self.out_file.close() - index = open(index_file, 'wb') - index.write(b'TNTIDX\x00\x00') - index.write(struct.pack(' Date: Sun, 13 Oct 2024 10:16:44 -0500 Subject: [PATCH 342/590] Update `ALCF/requirements/requirements.txt` --- ALCF/requirements/requirements.txt | 2 -- 1 file changed, 2 deletions(-) diff --git a/ALCF/requirements/requirements.txt b/ALCF/requirements/requirements.txt index 78d1789d2c6..03541ba514a 100644 --- a/ALCF/requirements/requirements.txt +++ b/ALCF/requirements/requirements.txt @@ -15,6 +15,4 @@ six numpy<2 schedulefree packaging>=20.0 -pydftracer wandb -# git+https://github.com/saforem2/ezpz@main From 5e9eed0443e5b448e924cd28d5696e82e475a9da Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 13 Oct 2024 10:16:54 -0500 Subject: [PATCH 343/590] Update `megatron/utils.py` --- megatron/utils.py | 94 ++++++++++++++++++++++++++++++----------------- 1 file changed, 60 insertions(+), 34 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index d00f4cd0efb..3d5eef46723 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -4,24 +4,14 @@ import sys import os -import time import logging -from typing import ContextManager, Optional +from typing import Optional import torch from torch.nn.parallel import DistributedDataParallel as torchDDP from deepspeed.accelerator import get_accelerator -if get_accelerator().device_name() == "cuda": - try: - from apex.multi_tensor_apply import multi_tensor_applier - import amp_C - - HAS_APEX = True - except Exception: - HAS_APEX = False - from megatron import get_args, get_adlr_autoresume, get_num_microbatches from megatron.core import mpu from megatron.core.tensor_parallel import param_is_not_tensor_parallel_duplicate @@ -30,70 +20,104 @@ import ezpz as ez +ACCELERATOR = get_accelerator() +assert ACCELERATOR is not None + +if ACCELERATOR.device_name() == "cuda": + try: + from apex.multi_tensor_apply import multi_tensor_applier # type:ignore + import amp_C # type:ignore + + HAS_APEX = True + except Exception: + HAS_APEX = False + RANK = ez.get_rank() log = logging.getLogger(__name__) -# log.setLevel("INFO") if RANK == 0 else log.setLevel("CRITICAL") - +log.setLevel(os.environ.get("LOG_LEVEL", ("INFO" if RANK == 0 else "CRITICAL"))) _DLIO_PROFILER_EXIST = True -_DFTRACER_EXIST=True +_DFTRACER_EXIST = True try: - import dftracer + import dftracer # type:ignore except Exception: - _DFTRACER_EXIST=False + _DFTRACER_EXIST = False try: - import dlio_profiler + import dlio_profiler # type:ignore except Exception: _DLIO_PROFILER_EXIST = False if _DFTRACER_EXIST: - from dftracer.logger import dftracer as PerfTrace, dft_fn as Profile, DFTRACER_ENABLE as DFTRACER_ENABLE + from dftracer.logger import ( # type:ignore + dftracer as PerfTrace, + dft_fn as Profile, + DFTRACER_ENABLE as DFTRACER_ENABLE, + ) elif _DLIO_PROFILER_EXIST: - from dlio_profiler.logger import fn_interceptor as Profile - from dlio_profiler.logger import dlio_logger as PerfTrace + from dlio_profiler.logger import fn_interceptor as Profile # type:ignore + from dlio_profiler.logger import dlio_logger as PerfTrace # type:ignore else: from functools import wraps - # from contextlib import nullcontext - # Profile: ContextManager = nullcontext - # - # class Profile(nullable_schema) + class Profile(object): - def __init__(self, cat, name=None, epoch=None, step=None, image_idx=None, image_size=None): - return - def log(self, func): + def __init__( + self, cat, name=None, epoch=None, step=None, image_idx=None, image_size=None + ): + return + + def log(self, func): return func - def log_init(self, func): + + def log_init(self, func): return func - def iter(self, func, iter_name="step"): + + def iter(self, func, iter_name="step"): return func + def __enter__(self): return + def __exit__(self, type, value, traceback): return - def update(self, epoch=None, step=None, image_idx=None, image_size=None, args={}): + + def update( + self, epoch=None, step=None, image_idx=None, image_size=None, args={} + ): return + def flush(self): return + def reset(self): return + def log_static(self, func): return + class dftracer(object): - def __init__(self,): + def __init__( + self, + ): self.type = None + def initialize_log(self, logfile=None, data_dir=None, process_id=-1): return + def get_time(self): return + def enter_event(self): return + def exit_event(self): return + def log_event(self, name, cat, start_time, duration, string_args=None): return + def finalize(self): return @@ -103,8 +127,8 @@ def finalize(self): def get_logger( name: str, - level: str = "INFO", - rank_zero_only: Optional[bool] = None, + level: Optional[str] = None, + rank_zero_only: Optional[bool] = True, ) -> logging.Logger: """Returns a `logging.Logger` object. @@ -112,7 +136,9 @@ def get_logger( non-zero ranks (and will be set to `level` on RANK==0). """ logger = logging.getLogger(name) - logger.setLevel(level) + logger.setLevel( + str(level if level is not None else os.environ.get("LOG_LEVEL", "INFO")).upper() + ) if rank_zero_only and ez.get_rank() != 0: logger.setLevel("CRITICAL") return logger From 3dcb2974da465b3d3a061215694464787737108b Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Sun, 13 Oct 2024 22:36:10 -0500 Subject: [PATCH 344/590] fixed bugs and added commandline option --- megatron/arguments.py | 2 ++ megatron/data/gpt_dataset.py | 10 ++++++---- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 9a5e4b8da7e..82e000923fc 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1252,6 +1252,8 @@ def _add_data_args(parser): group.add_argument('--data-file-list', type=str, default=None, help='The file with the list of dataset and weights') + group.add_argument('--shuffle-sample', action='stored_true', help="Whether to shuffle the samples within in the dataset files") + group.add_argument('--split', type=str, default='969, 30, 1', help='Comma-separated list of proportions for training,' ' validation, and test split. For example the split ' diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index c801a6a5ae5..8c32be7d8e3 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -94,7 +94,7 @@ def Build(self): class BuildConcatDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, dataset_builders): + def __init__(self, dataset_builders, shuffle=False): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) @@ -117,7 +117,9 @@ def _build_indices(): self.dataset_index, self.dataset_sample_index = _build_indices() np_rng = np.random.RandomState(seed=dataset_builders[0].seed) - self.shuffle_index=np_rng.shuffle(range(self.num_samples)) + self.shuffle_index = np.arange(self.num_samples) + if shuffle: + np_rng.shuffle(self.shuffle_index) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," @@ -146,7 +148,7 @@ def __getitem__(self, idx): valid_datasets = [] test_datasets = [] # Build individual datasets. - + args = get_args() @dlp.log def build_corpus_datasets(dataset_type='train'): start_time = time.time() @@ -172,7 +174,7 @@ def build_corpus_datasets(dataset_type='train'): print_rank_0(" > number of samples for each corpus ") corpus_weights_achieved={} for c in corpus_list: - datasets.append(BuildConcatDataset(corpus_builders[c])) + datasets.append(BuildConcatDataset(corpus_builders[c], args.shuffle_sample)) total += datasets[-1].num_samples corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") From 43fc2feaace3781bb399e32b2f6e827920622c61 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Sun, 13 Oct 2024 23:00:21 -0500 Subject: [PATCH 345/590] fixed typo --- megatron/arguments.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 82e000923fc..307e725e515 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1252,7 +1252,7 @@ def _add_data_args(parser): group.add_argument('--data-file-list', type=str, default=None, help='The file with the list of dataset and weights') - group.add_argument('--shuffle-sample', action='stored_true', help="Whether to shuffle the samples within in the dataset files") + group.add_argument('--shuffle-sample', action='store_true', help="Whether to shuffle the samples within in the dataset files") group.add_argument('--split', type=str, default='969, 30, 1', help='Comma-separated list of proportions for training,' From d50239f1c5470fd2b2495affc01773e063942e37 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 14 Oct 2024 14:20:56 -0500 Subject: [PATCH 346/590] added support for blending samples across different files in the same corpus --- megatron/arguments.py | 4 +++- megatron/data/gpt_dataset.py | 36 ++++++++++++++++++++++++++++-------- 2 files changed, 31 insertions(+), 9 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 9ab3e40953d..9b0e6ccb1ac 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1291,7 +1291,9 @@ def _add_data_args(parser): group.add_argument('--data-file-list', type=str, default=None, help='The file with the list of dataset and weights') - group.add_argument('--shuffle-sample', action='store_true', help="Whether to shuffle the samples within in the dataset files") + group.add_argument('--shuffle-sample-in-corpus', action='store_true', help="Whether to shuffle the samples within in the dataset files") + + group.add_argument('--blend-sample-in-corpus', action='store_true', help="Whether to blend different files in the same corpus") group.add_argument('--split', type=str, default='969, 30, 1', help='Comma-separated list of proportions for training,' diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index c412d02b31c..38df5562675 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -131,19 +131,35 @@ def Build(self): self.build = True return self.dataset - class BuildConcatDataset(torch.utils.data.Dataset): + class BuildCorpusDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, dataset_builders, shuffle=False): + def __init__(self, dataset_builders): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) self.indices = np.zeros((self.num_samples, 2), dtype=np.uint64) - self.desc = "ConcatDataset:" + self.desc = "CorpusDataset:" # m = 0 num_samples_list = np.array([d.num_samples for d in dataset_builders]) self.num_samples = np.sum(num_samples_list) + args = get_args() - def _build_indices(): + @dlp.log + def _build_indices_blended(): + start_time = time.time() + dataset_index = np.zeros(self.num_samples, dtype=np.int64) + dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) + weights = num_samples_list / self.num_samples + helpers.build_blending_indices( + dataset_index, dataset_sample_index, + weights, self.num_datasets, self.num_samples, + torch.distributed.get_rank() == 0) + log.debug('> elapsed time for building blendable dataset indices for corpus {self.dataset_builders[0].corpus}: ' + '{:.2f} (sec)'.format(time.time() - start_time)) + return dataset_index, dataset_sample_index + + + def _build_indices_concat(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) @@ -159,11 +175,15 @@ def _build_indices(): "{:.2f} (sec)".format(time.time() - start_time) ) return dataset_index, dataset_sample_index - - self.dataset_index, self.dataset_sample_index = _build_indices() + + if args.blend_sample_in_corpus: + self.dataset_index, self.dataset_sample_index = _build_indices_blended() + else: + self.dataset_index, self.dataset_sample_index = _build_indices_concat() + np_rng = np.random.RandomState(seed=dataset_builders[0].seed) self.shuffle_index = np.arange(self.num_samples) - if shuffle: + if args.shuffle_sample_in_corpus: np_rng.shuffle(self.shuffle_index) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," @@ -243,7 +263,7 @@ def build_corpus_datasets(dataset_type="train"): log.debug(" > number of samples for each corpus ") corpus_weights_achieved = {} for c in corpus_list: - datasets.append(BuildConcatDataset(corpus_builders[c], args.shuffle_sample)) + datasets.append(BuildCorpusDataset(corpus_builders[c])) total += datasets[-1].num_samples corpus_weights_achieved[c] = ( float(datasets[-1].num_samples) / train_num_samples From 45ff652dc473fa07ec674543199b2a7785e7f9a6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 14 Oct 2024 23:28:00 -0500 Subject: [PATCH 347/590] Discard changes to megatron/data/gpt_dataset.py --- megatron/data/gpt_dataset.py | 921 ++++++++++++----------------------- 1 file changed, 325 insertions(+), 596 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index f62b567ddc4..0cf97356a41 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -5,114 +5,71 @@ import hashlib import os import time -import logging import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import is_rank_0, get_args +from megatron import print_rank_0, is_rank_0, get_args from megatron.core import mpu -from megatron.data import helpers # type:ignore +from megatron.data import helpers from megatron.data.blendable_dataset import BlendableDataset -from megatron.data.dataset_utils import ( - get_datasets_weights_and_num_samples, - get_datasets_corpuses_weights_and_num_samples, -) +from megatron.data.dataset_utils import get_datasets_weights_and_num_samples, get_datasets_corpuses_weights_and_num_samples from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset -from megatron.utils import PerfTrace, Profile, get_logger +from megatron.utils import PerfTrace, Profile from mpi4py import MPI -try: - import ezpz as ez - RANK = ez.get_rank() -except Exception: - RANK = torch.distributed.get_rank() - -# NOTE: [logging]----------------------------------------------------------- -# - Set logging level to "INFO" on RANK == 0, "CRITICAL" on all other ranks -log = logging.getLogger(__name__) -LOG_LEVEL = str(os.environ.get("LOG_LEVEL", "INFO")).upper() -log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") -# -------------------------------------------------------------------------- - dlp = Profile("DATASET") -log = get_logger(__name__, rank_zero_only=True) - - @dlp.log -def build_train_valid_test_datasets( - data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples, - seq_length, - seed, - skip_warmup, - train_data_prefix=None, - valid_data_prefix=None, - test_data_prefix=None, - return_doc_ids=False, - *, - data_cache_path=None, -): +def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, + train_valid_test_num_samples, + seq_length, seed, skip_warmup, + train_data_prefix=None, + valid_data_prefix=None, + test_data_prefix=None, + return_doc_ids=False, *, + data_cache_path=None): """Build train, valid, and test datasets.""" if data_prefix: - log.debug("Single data path provided for train, valid & test") + print_rank_0("Single data path provided for train, valid & test") # Single dataset. if len(data_prefix) == 1: - return _build_train_valid_test_datasets( - data_prefix[0], - data_impl, - splits_string, - train_valid_test_num_samples, - seq_length, - seed, - skip_warmup, - data_cache_path=data_cache_path, - ) + return _build_train_valid_test_datasets(data_prefix[0], + data_impl, splits_string, + train_valid_test_num_samples, + seq_length, seed, skip_warmup, + data_cache_path=data_cache_path) # Blending dataset. # Parse the values. - output = get_datasets_corpuses_weights_and_num_samples( - data_prefix, train_valid_test_num_samples - ) + output = get_datasets_corpuses_weights_and_num_samples(data_prefix, + train_valid_test_num_samples) prefixes, corpuses, weights, datasets_train_valid_test_num_samples = output corpus_list = sorted(set(corpuses)) train_num_samples, valid_num_samples, test_num_samples = map( - sum, zip(*datasets_train_valid_test_num_samples) + sum, + zip(*datasets_train_valid_test_num_samples) ) class DatasetBuilder: - """ + ''' This is for building individual dataset from each dataset file - """ - + ''' @dlp.log - def __init__( - self, - prefix, - corpus, - data_impl, - splits_string, - num_samples, - seq_length, - seed, - skip_warmup, - return_doc_ids, - data_cache_path=data_cache_path, - name="train", - ): + def __init__(self, prefix, corpus, data_impl, splits_string, + num_samples, seq_length, seed, skip_warmup, + return_doc_ids, + data_cache_path=data_cache_path, name='train'): self.prefix = prefix self.data_impl = data_impl self.splits_string = splits_string - if name == "train": + if name == 'train': self.num_samples = num_samples[0] - elif name == "valid": + elif name == 'valid': self.num_samples = num_samples[1] else: self.num_samples = num_samples[2] @@ -127,413 +84,279 @@ def __init__( self.desc = prefix + f"{self.num_samples}" + f"{seq_length}" + f"{seed}" self.build = False self.corpus = corpus - @dlp.log def Build(self): - self.dataset = _build_train_valid_test_datasets_single( - self.prefix, - self.data_impl, - self.splits_string, - self.num_samples_train_valid_test, - self.seq_length, - self.seed, - self.skip_warmup, - self.name, - self.return_doc_ids, - data_cache_path=self.data_cache_path, - ) + self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, + self.num_samples_train_valid_test, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, + data_cache_path=self.data_cache_path) self.build = True return self.dataset class BuildConcatDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, dataset_builders, shuffle=False): + def __init__(self, dataset_builders): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) - self.indices = np.zeros((self.num_samples, 2), dtype=np.uint64) - self.desc = "ConcatDataset:" - # m = 0 + self.indices=np.zeros((self.num_samples, 2), dtype=np.uint64) + self.desc="ConcatDataset:" + m = 0 num_samples_list = np.array([d.num_samples for d in dataset_builders]) self.num_samples = np.sum(num_samples_list) - def _build_indices(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) - helpers.build_concat_indices( - dataset_index, - dataset_sample_index, - num_samples_list, - self.num_datasets, - torch.distributed.get_rank() == 0, - ) - log.debug( - "> elapsed time for building concat dataset indices: " - "{:.2f} (sec)".format(time.time() - start_time) - ) + helpers.build_concat_indices(dataset_index, dataset_sample_index, + num_samples_list, + self.num_datasets, + torch.distributed.get_rank()==0) + print_rank_0('> elapsed time for building concat dataset indices: ' + '{:.2f} (sec)'.format(time.time() - start_time)) return dataset_index, dataset_sample_index - + self.dataset_index, self.dataset_sample_index = _build_indices() - np_rng = np.random.RandomState(seed=dataset_builders[0].seed) - self.shuffle_index = np.arange(self.num_samples) - if shuffle: - np_rng.shuffle(self.shuffle_index) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," - self.desc += ( - f"-{self.num_samples}" - + f"-{dataset_builders[0].seq_length}" - + f"{dataset_builders[0].seed}" - ) - + self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" def __len__(self): return self.num_samples @dlp.log def __getitem__(self, idx): - id_shuffle = self.shuffle_index[idx] - i = self.dataset_index[id_shuffle] - j = self.dataset_sample_index[id_shuffle] + if idx >= self.num_samples: + print_rank_0(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") + id = np.random.randint(self.num_samples) + else: + id = idx + i = self.dataset_index[idx] + j = self.dataset_sample_index[idx] if self.dataset_builders[i].build: return self.dataset_builders[i].dataset[j] else: return self.dataset_builders[i].Build()[j] + - # Predetermine whether need to build the specific dataset or not. + # Predetermine whether need to build the specific dataset or not. start_time = time.time() - log.debug(" >>> Started building datasets in distributed way ... ") + print_rank_0(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] - + train_datasets = [] valid_datasets = [] test_datasets = [] # Build individual datasets. - args = get_args() + @dlp.log - def build_corpus_datasets(dataset_type="train"): + def build_corpus_datasets(dataset_type='train'): start_time = time.time() - log.debug(f" >>> Building {dataset_type} corpus datasets ...") + print_rank_0(f" >>> Building {dataset_type} corpus datasets ...") datasets = [] corpus_builders = {} corpus_weights = {} for c in corpus_list: corpus_builders[c] = [] corpus_weights[c] = 0.0 - dataset_builders = [ - DatasetBuilder( - prefixes[i], - corpuses[i], - data_impl, - splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, - seed, - skip_warmup, - return_doc_ids, - data_cache_path, - dataset_type, - ) - for i in range(len(weights)) - ] - for i in range( - torch.distributed.get_rank() - // mpu.get_tensor_model_parallel_world_size(), - len(weights), - torch.distributed.get_world_size() - // mpu.get_tensor_model_parallel_world_size(), - ): + dataset_builders = [DatasetBuilder(prefixes[i], corpuses[i], data_impl, splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, seed, skip_warmup, + return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] + for i in range(torch.distributed.get_rank()//mpu.get_tensor_model_parallel_world_size(), len(weights), torch.distributed.get_world_size()//mpu.get_tensor_model_parallel_world_size()): dataset_builders[i].Build() - log.debug( - f" >>> Finished building individual datasets in {time.time() - start_time} seconds" - ) + print_rank_0(f" >>> Finished building individual datasets in {time.time() - start_time} seconds") start_concating_time = time.time() for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] total = 0 - log.debug(" > number of samples for each corpus ") - corpus_weights_achieved = {} + print_rank_0(" > number of samples for each corpus ") + corpus_weights_achieved={} for c in corpus_list: - datasets.append(BuildConcatDataset(corpus_builders[c], args.shuffle_sample)) + datasets.append(BuildConcatDataset(corpus_builders[c])) total += datasets[-1].num_samples - corpus_weights_achieved[c] = ( - float(datasets[-1].num_samples) / train_num_samples - ) - log.debug( - f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})" - ) - - log.debug(f" > total number of samples: {total}") - log.debug( - f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds" - ) - log.debug( - f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds" - ) + corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples + print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") + + print_rank_0(f" > total number of samples: {total}") + print_rank_0(f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds") + print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") return datasets, [corpus_weights_achieved[c] for c in corpus_list] - train_weights = None if a > 0: - train_datasets, train_weights = build_corpus_datasets("train") - valid_weights = None + train_datasets, train_weights = build_corpus_datasets('train') + if b > 0: - valid_datasets, valid_weights = build_corpus_datasets("valid") - test_weights = None - if c > 0: - test_datasets, test_weights = build_corpus_datasets("test") + valid_datasets, valid_weights = build_corpus_datasets('valid') + + if c > 0: + test_datasets, test_weights = build_corpus_datasets('test') # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them - log.debug( - f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds" - ) + print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") torch.distributed.barrier(group=mpu.get_data_parallel_group()) torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) - log.debug( - f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds" - ) - log.debug(" >>> Starting to build BlendableDataset") + print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") + print_rank_0(f" >>> Starting to build BlendableDataset") # Blend. start_time = time.time() blending_train_dataset = None - if train_datasets and train_weights: - blending_train_dataset = BlendableDataset( - train_datasets, - train_weights, - train_num_samples, - data_cache_path=data_cache_path, - ) + if train_datasets: + blending_train_dataset = BlendableDataset(train_datasets, train_weights, train_num_samples, + data_cache_path=data_cache_path) blending_valid_dataset = None - if valid_datasets and valid_weights: - blending_valid_dataset = BlendableDataset( - valid_datasets, - valid_weights, - valid_num_samples, - data_cache_path=data_cache_path, - ) + if valid_datasets: + blending_valid_dataset = BlendableDataset(valid_datasets, valid_weights, valid_num_samples, + data_cache_path=data_cache_path) blending_test_dataset = None - if test_datasets and test_weights: - blending_test_dataset = BlendableDataset( - test_datasets, - test_weights, - test_num_samples, - data_cache_path=data_cache_path, - ) + if test_datasets: + blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, + data_cache_path=data_cache_path) end_time = time.time() - log.debug( - f" >>> Finished building BlendableDataset in {end_time - start_time} seconds" - ) - return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) + print_rank_0(f" >>> Finished building BlendableDataset in {end_time - start_time} seconds") + return (blending_train_dataset, blending_valid_dataset, + blending_test_dataset) else: - log.debug( - "Separate data paths provided for train, valid & test. Split string will be ignored." - ) + print_rank_0("Separate data paths provided for train, valid & test. Split string will be ignored.") train_dataset, valid_dataset, test_dataset = None, None, None # Single dataset. if train_data_prefix is not None: - train_dataset = build_dataset( - "train", - train_data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples[0], - seq_length, - seed, - skip_warmup, - data_cache_path=data_cache_path, - ) + train_dataset = build_dataset("train", train_data_prefix, data_impl, + splits_string, + train_valid_test_num_samples[0], + seq_length, seed, skip_warmup, + data_cache_path=data_cache_path) if valid_data_prefix is not None: - valid_dataset = build_dataset( - "valid", - valid_data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples[1], - seq_length, - seed, - False, - data_cache_path=data_cache_path, - ) + valid_dataset = build_dataset("valid", valid_data_prefix, data_impl, + splits_string, + train_valid_test_num_samples[1], + seq_length, seed, False, + data_cache_path=data_cache_path) + if test_data_prefix is not None: - test_dataset = build_dataset( - "test", - test_data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples[2], - seq_length, - seed, - False, - data_cache_path=data_cache_path, - ) + test_dataset = build_dataset("test", test_data_prefix, data_impl, + splits_string, + train_valid_test_num_samples[2], + seq_length, seed, False, + data_cache_path=data_cache_path) return (train_dataset, valid_dataset, test_dataset) - @dlp.log -def _build_train_valid_test_datasets( - data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples, - seq_length, - seed, - skip_warmup, - return_doc_ids=False, - *, - data_cache_path=None, -): +def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, + train_valid_test_num_samples, + seq_length, seed, skip_warmup, + return_doc_ids=False, *, + data_cache_path=None): """Build train, valid, and test datasets.""" # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, + data_impl, + skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - log.debug(" > dataset split:") + print_rank_0(' > dataset split:') def print_split_stats(name, index): - log.debug(" {}:".format(name)) - log.debug( - " document indices in [{}, {}) total of {} " "documents".format( - splits[index], splits[index + 1], splits[index + 1] - splits[index] - ) - ) - - print_split_stats("train", 0) - print_split_stats("validation", 1) - print_split_stats("test", 2) + print_rank_0(' {}:'.format(name)) + print_rank_0(' document indices in [{}, {}) total of {} ' + 'documents'.format(splits[index], splits[index + 1], + splits[index + 1] - splits[index])) + print_split_stats('train', 0) + print_split_stats('validation', 1) + print_split_stats('test', 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange( - start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 - ) - dataset = GPTDataset( - name, - data_prefix, - documents, - indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, - seed, - return_doc_ids, - data_cache_path=data_cache_path, - ) + documents = np.arange(start=splits[index], stop=splits[index + 1], + step=1, dtype=np.int32) + dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, seed, + return_doc_ids, + data_cache_path=data_cache_path) return dataset - train_dataset = build_dataset(0, "train") - valid_dataset = build_dataset(1, "valid") - test_dataset = build_dataset(2, "test") + train_dataset = build_dataset(0, 'train') + valid_dataset = build_dataset(1, 'valid') + test_dataset = build_dataset(2, 'test') return (train_dataset, valid_dataset, test_dataset) - @dlp.log -def _build_train_valid_test_datasets_single( - data_prefix, - data_impl, - splits_string, - train_valid_test_num_samples, - seq_length, - seed, - skip_warmup, - name, - return_doc_ids=False, - *, - data_cache_path=None, -): +def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_string, + train_valid_test_num_samples, + seq_length, seed, skip_warmup, name, + return_doc_ids=False, *, + data_cache_path=None): """Build train, valid, and test datasets.""" # Each rank print out information - log.debug(f" >> building dataset for {data_prefix}") + print_rank_0(f" >> building dataset for {data_prefix}") # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, + data_impl, + skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - log.debug(" > dataset split:") + print_rank_0(' > dataset split:') def print_split_stats(name, index): - log.debug(" {}:".format(name)) - log.debug( - " document indices in [{}, {}) total of {} " "documents".format( - splits[index], splits[index + 1], splits[index + 1] - splits[index] - ) - ) - - print_split_stats("train", 0) - print_split_stats("validation", 1) - print_split_stats("test", 2) + print_rank_0(' {}:'.format(name)) + print_rank_0(' document indices in [{}, {}) total of {} ' + 'documents'.format(splits[index], splits[index + 1], + splits[index + 1] - splits[index])) + print_split_stats('train', 0) + print_split_stats('validation', 1) + print_split_stats('test', 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange( - start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 - ) - dataset = GPTDataset( - name, - data_prefix, - documents, - indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, - seed, - return_doc_ids, - data_cache_path=data_cache_path, - ) + documents = np.arange(start=splits[index], stop=splits[index + 1], + step=1, dtype=np.int32) + dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, seed, + return_doc_ids, + data_cache_path=data_cache_path) return dataset - - if name.find("train") != -1: - return build_dataset(0, "train") - if name.find("valid") != -1: - return build_dataset(1, "valid") - if name.find("test") != -1: - return build_dataset(2, "test") - + if name.find("train")!=-1: + return build_dataset(0, 'train') + if name.find("valid")!=-1: + return build_dataset(1, 'valid') + if name.find("test")!=-1: + return build_dataset(2, 'test') @dlp.log -def build_dataset( - dataset_name, - data_prefix, - data_impl, - splits_string, - num_samples, - seq_length, - seed, - skip_warmup, - *, - data_cache_path=None, -): +def build_dataset(dataset_name, data_prefix, data_impl, + splits_string, num_samples, + seq_length, seed, skip_warmup, + *, + data_cache_path=None): dataset = None if len(data_prefix) == 1: - dataset = _build_dataset( - dataset_name, - data_prefix[0], - data_impl, - splits_string, - num_samples, - seq_length, - seed, - skip_warmup, - data_cache_path=data_cache_path, - ) + dataset = _build_dataset(dataset_name, data_prefix[0], data_impl, + splits_string, num_samples, seq_length, + seed, skip_warmup, + data_cache_path=data_cache_path) else: # Blending dataset. # Parse the values. @@ -544,108 +367,73 @@ def build_dataset( # Build individual datasets. datasets = [] for i in range(len(prefixes)): - ds = _build_dataset( - dataset_name, - prefixes[i], - data_impl, - splits_string, - dataset_num_samples[i], - seq_length, - seed, - skip_warmup, - data_cache_path=data_cache_path, - ) + ds = _build_dataset(dataset_name, prefixes[i], data_impl, + splits_string, dataset_num_samples[i], + seq_length, seed, skip_warmup, + data_cache_path=data_cache_path) if ds: datasets.append(ds) if datasets: - dataset = BlendableDataset( - datasets, weights, num_samples, data_cache_path=data_cache_path - ) + dataset = BlendableDataset(datasets, weights, num_samples, + data_cache_path=data_cache_path) return dataset - @dlp.log -def _build_dataset( - dataset_name, - data_prefix, - data_impl, - splits_string, - num_samples, - seq_length, - seed, - skip_warmup, - *, - data_cache_path=None, -): +def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, + num_samples, seq_length, seed, skip_warmup, + *, + data_cache_path=None): """ Build dataset. This method is called when individual train, valid, test datasets are provided """ # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, + data_impl, + skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] - log.debug(" {}:".format(dataset_name)) - log.debug( - " document indices in [0, {}) total of {} " "documents".format( - total_num_of_documents, total_num_of_documents - ) - ) - - documents = np.arange(start=0, stop=total_num_of_documents, step=1, dtype=np.int32) - - dataset = GPTDataset( - dataset_name, - data_prefix, - documents, - indexed_dataset, - splits_string, - num_samples, - seq_length, - seed, - data_cache_path=data_cache_path, - ) + print_rank_0(' {}:'.format(dataset_name)) + print_rank_0(' document indices in [0, {}) total of {} ' + 'documents'.format(total_num_of_documents, total_num_of_documents)) - return dataset + documents = np.arange(start=0, stop=total_num_of_documents, + step=1, dtype=np.int32) + dataset = GPTDataset(dataset_name, data_prefix, documents, indexed_dataset, + splits_string, num_samples, seq_length, seed, + data_cache_path=data_cache_path) + + return dataset @dlp.log def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): """Build indexed dataset.""" - log.debug(" > building dataset index ...") + print_rank_0(' > building dataset index ...') start_time = time.time() - indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) - log.debug( - " > finished creating indexed dataset in {:4f} " "seconds".format( - time.time() - start_time - ) - ) - log.debug(" number of documents: {}".format(indexed_dataset.sizes.shape[0])) + indexed_dataset = make_indexed_dataset(data_prefix, + data_impl, + skip_warmup) + print_rank_0(' > finished creating indexed dataset in {:4f} ' + 'seconds'.format(time.time() - start_time)) + print_rank_0(' number of documents: {}'.format( + indexed_dataset.sizes.shape[0])) return indexed_dataset class GPTDataset(torch.utils.data.Dataset): @dlp.log - def __init__( - self, - name, - data_prefix, - documents, - indexed_dataset, - splits_string, - num_samples, - seq_length, - seed, - return_doc_ids=False, - *, - data_cache_path=None, - ): + def __init__(self, name, data_prefix, documents, indexed_dataset, + splits_string, num_samples, seq_length, seed, + return_doc_ids=False, *, + data_cache_path=None): + self.name = name self.indexed_dataset = indexed_dataset self.return_doc_ids = return_doc_ids @@ -655,29 +443,20 @@ def __init__( assert np.max(documents) < indexed_dataset.sizes.shape[0] # Build index mappings. - self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = ( - _build_index_mappings( - self.name, - data_prefix, - documents, - self.indexed_dataset.sizes, - splits_string, - num_samples, - seq_length, - seed, - data_cache_path=data_cache_path, - ) - ) + self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = \ + _build_index_mappings(self.name, data_prefix, + documents, self.indexed_dataset.sizes, + splits_string, num_samples, seq_length, seed, + data_cache_path=data_cache_path) + def __len__(self): # -1 is due to data structure used to retieve the index: # sample i --> [sample_idx[i], sample_idx[i+1]) return self.sample_idx.shape[0] - 1 - @dlp.log def __getitem__(self, idx): args = get_args() - assert args is not None orig_idx = idx # Get the shuffled index. try: @@ -686,24 +465,21 @@ def __getitem__(self, idx): if is_rank_0(): import json from rich import print_json - print(exc) print( - "\n".join( - [ - "-------------------------------------------------", - f"Trying to access {idx=} from self.shuffle_idx,", - f"but {len(self.shuffle_idx)=}", - "-------------------------------------------------", - ] + '\n'.join( + ['-------------------------------------------------', + f'Trying to access {idx=} from self.shuffle_idx,', + f'but {len(self.shuffle_idx)=}', + '-------------------------------------------------'] ) ) print_json( json.dumps( { - "doc_idx": len(self.doc_idx), - "sample_idx": len(self.sample_idx), - "shuffle_idx": len(self.shuffle_idx), + 'doc_idx': len(self.doc_idx), + 'sample_idx': len(self.sample_idx), + 'shuffle_idx': len(self.shuffle_idx), }, indent=4, ) @@ -717,57 +493,45 @@ def __getitem__(self, idx): doc_ids = [] if doc_index_f == doc_index_l: doc_ids.append(self.doc_idx[doc_index_f]) - sample = self.indexed_dataset.get( - self.doc_idx[doc_index_f], - offset=offset_f, - length=offset_l - offset_f + 1, - ) + sample = self.indexed_dataset.get(self.doc_idx[doc_index_f], + offset=offset_f, + length=offset_l - offset_f + 1) else: # Otherwise, get the rest of the initial document. doc_ids.append(self.doc_idx[doc_index_f]) - sample_list = [ - self.indexed_dataset.get(self.doc_idx[doc_index_f], offset=offset_f) - ] + sample_list = [self.indexed_dataset.get(self.doc_idx[doc_index_f], + offset=offset_f)] # Loop over all in between documents and add the entire document. for i in range(doc_index_f + 1, doc_index_l): doc_ids.append(self.doc_idx[i]) sample_list.append(self.indexed_dataset.get(self.doc_idx[i])) # And finally add the relevant portion of last document. doc_ids.append(self.doc_idx[doc_index_l]) - sample_list.append( - self.indexed_dataset.get(self.doc_idx[doc_index_l], length=offset_l + 1) - ) + sample_list.append(self.indexed_dataset.get( + self.doc_idx[doc_index_l], + length=offset_l + 1)) sample = np.concatenate(sample_list) - text_name = "text" + text_name = 'text' if args.use_dataset_only: - text_name = "input_ids" + text_name = 'input_ids' sample_dict = {text_name: np.array(sample, dtype=np.int64)} if args.return_data_index: - sample_dict.update({"index": np.array([orig_idx], dtype=np.int64)}) + sample_dict.update({'index': np.array([orig_idx], dtype=np.int64)}) - if self.return_doc_ids: # for retro preprocessing - sample_dict.update({"doc_ids": np.array(doc_ids, dtype=np.int64)}) + if self.return_doc_ids: # for retro preprocessing + sample_dict.update({'doc_ids': np.array(doc_ids, dtype=np.int64)}) if args.use_dataset_only: - sample_dict.update({"labels": np.array(sample, dtype=np.int64)}) + sample_dict.update({'labels': np.array(sample, dtype=np.int64)}) return sample_dict - @dlp.log -def _build_index_mappings( - name, - data_prefix, - documents, - sizes, - splits_string, - num_samples, - seq_length, - seed, - *, - data_cache_path, -): +def _build_index_mappings(name, data_prefix, documents, sizes, + splits_string, num_samples, seq_length, seed, + *, + data_cache_path): """Build doc-idx, sample-idx, and shuffle-idx. doc-idx: is an array (ordered) of documents to be used in training. sample-idx: is the start document index and document offset for each @@ -775,11 +539,10 @@ def _build_index_mappings( shuffle-idx: maps the sample index into a random index into sample-idx. """ args = get_args() - assert args is not None # Number of tokens in each epoch and number of required epochs. tokens_per_epoch = _num_tokens(documents, sizes) num_epochs = _num_epochs(tokens_per_epoch, seq_length, num_samples) - if args.train_data_exact_num_epochs is not None and name == "train": + if args.train_data_exact_num_epochs is not None and name == 'train': num_epochs = args.train_data_exact_num_epochs # rng state @@ -794,13 +557,13 @@ def _build_index_mappings( desc += f"Sequence length {seq_length}\n" desc += f"Random seed {seed}\n" desc += f"Split {splits_string}\n" - desc_hash = hashlib.md5(desc.encode("utf-8")).hexdigest() + desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() desc_filename = desc_hash + ".dsc" - doc_idx_filename = desc_hash + "_doc_idx.npy" - sample_idx_filename = desc_hash + "_sample_idx.npy" - shuffle_idx_filename = desc_hash + "_shuffle_idx.npy" + doc_idx_filename = desc_hash + '_doc_idx.npy' + sample_idx_filename = desc_hash + '_sample_idx.npy' + shuffle_idx_filename = desc_hash + '_shuffle_idx.npy' - if name == "train": + if name == 'train': # force to use certain index files if args.train_desc_path is not None: desc_filename = args.train_desc_path @@ -815,15 +578,15 @@ def _build_index_mappings( # duplication, then look in data-cache-path if specified, # If nothing is found, use the last path looked in build_indices = True - prefixes = [os.path.join(os.path.dirname(data_prefix), "index-cache")] + prefixes = [os.path.join(os.path.dirname(data_prefix), 'index-cache')] if data_cache_path is not None: prefixes.append(data_cache_path) for prefix in prefixes: idx_path = { - "desc": os.path.join(prefix, desc_filename), - "doc": os.path.join(prefix, doc_idx_filename), - "sample": os.path.join(prefix, sample_idx_filename), - "shuffle": os.path.join(prefix, shuffle_idx_filename), + 'desc': os.path.join(prefix, desc_filename), + 'doc': os.path.join(prefix, doc_idx_filename), + 'sample': os.path.join(prefix, sample_idx_filename), + 'shuffle': os.path.join(prefix, shuffle_idx_filename) } for f in idx_path.values(): if not os.path.isfile(f): @@ -832,17 +595,15 @@ def _build_index_mappings( # Found our files! build_indices = False break - data_cache_dir = os.path.dirname(idx_path["desc"]) + data_cache_dir = os.path.dirname(idx_path['desc']) data_cache_success = True # Build the indexed mapping if not exist. if build_indices: - # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the - # ranks will first create the document files, and then read it. + # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the + # ranks will first create the document files, and then read it. # There will not be contension effects going on either - log.warning( - f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}" - ) + print_rank_0(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -851,80 +612,64 @@ def _build_index_mappings( # not mean anything. if num_epochs == 1: separate_last_epoch = False - log.debug( - " > only one epoch required, setting " "separate_last_epoch to False" - ) + print_rank_0(' > only one epoch required, setting ' + 'separate_last_epoch to False') else: # Get the number of samples for the last epoch num_samples_from_epochs_minus_one = ( - (num_epochs - 1) * tokens_per_epoch - 1 - ) // seq_length - last_epoch_num_samples = num_samples - num_samples_from_epochs_minus_one - assert ( - last_epoch_num_samples >= 0 - ), "last epoch number of samples should be non-negative." + (num_epochs - 1) * tokens_per_epoch - 1) // seq_length + last_epoch_num_samples = num_samples - \ + num_samples_from_epochs_minus_one + assert last_epoch_num_samples >= 0, \ + 'last epoch number of samples should be non-negative.' num_samples_per_epoch = (tokens_per_epoch - 1) // seq_length - assert last_epoch_num_samples <= ( - num_samples_per_epoch + 1 - ), "last epoch number of samples exceeded max value." + assert last_epoch_num_samples <= (num_samples_per_epoch + 1), \ + 'last epoch number of samples exceeded max value.' # If we have less than 80% of the samples for the last epoch, # seperate out the epoch and treat it differently. # Note: the 80% number is just based on common sense and can # be adjusted if needed. - separate_last_epoch = last_epoch_num_samples < int( - 0.80 * num_samples_per_epoch - ) + separate_last_epoch = (last_epoch_num_samples < + int(0.80 * num_samples_per_epoch)) if separate_last_epoch: - string = ( - " > last epoch number of samples ({}) is smaller " - "than 80% of number of samples per epoch ({}), " - "setting separate_last_epoch to True" - ) + string = ' > last epoch number of samples ({}) is smaller '\ + 'than 80% of number of samples per epoch ({}), '\ + 'setting separate_last_epoch to True' else: - string = ( - " > last epoch number of samples ({}) is larger " - "than 80% of number of samples per epoch ({}), " - "setting separate_last_epoch to False" - ) - log.debug(string.format(last_epoch_num_samples, num_samples_per_epoch)) + string = ' > last epoch number of samples ({}) is larger '\ + 'than 80% of number of samples per epoch ({}), '\ + 'setting separate_last_epoch to False' + print_rank_0(string.format(last_epoch_num_samples, + num_samples_per_epoch)) + try: os.makedirs(data_cache_dir, exist_ok=True) # description - with open(idx_path["desc"], "wt") as fd: + with open(idx_path['desc'], 'wt') as fd: fd.write(desc) # doc-idx. start_time = time.time() - doc_idx = _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch) - np.save(idx_path["doc"], doc_idx, allow_pickle=True) - log.debug( - " > elasped time to build and save doc-idx mapping " - "(seconds): {:4f}".format(time.time() - start_time) - ) + doc_idx = _build_doc_idx(documents, num_epochs, np_rng, + separate_last_epoch) + np.save(idx_path['doc'], doc_idx, allow_pickle=True) + print_rank_0(' > elasped time to build and save doc-idx mapping ' + '(seconds): {:4f}'.format(time.time() - start_time)) # sample-idx. start_time = time.time() # Use C++ implementation for speed. # First compile and then import. from megatron.data import helpers - assert doc_idx.dtype == np.int32 assert sizes.dtype == np.int32 - sample_idx = helpers.build_sample_idx( - sizes, - doc_idx, - seq_length, - num_epochs, - tokens_per_epoch, - torch.distributed.get_rank() == 0, - ) - np.save(idx_path["sample"], sample_idx, allow_pickle=True) - log.debug( - " > elasped time to build and save sample-idx mapping " - "(seconds): {:4f}".format(time.time() - start_time) - ) + sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, + num_epochs, tokens_per_epoch, torch.distributed.get_rank()==0) + np.save(idx_path['sample'], sample_idx, allow_pickle=True) + print_rank_0(' > elasped time to build and save sample-idx mapping ' + '(seconds): {:4f}'.format(time.time() - start_time)) # shuffle-idx. start_time = time.time() # -1 is due to data structure used to retieve the index: @@ -933,46 +678,35 @@ def _build_index_mappings( num_samples_ = num_samples_from_epochs_minus_one else: num_samples_ = sample_idx.shape[0] - 1 - shuffle_idx = _build_shuffle_idx( - num_samples_, sample_idx.shape[0] - 1, np_rng - ) - np.save(idx_path["shuffle"], shuffle_idx, allow_pickle=True) - log.debug( - " > elasped time to build and save shuffle-idx mapping" - " (seconds): {:4f}".format(time.time() - start_time) - ) + shuffle_idx = _build_shuffle_idx(num_samples_, + sample_idx.shape[0] - 1, np_rng) + np.save(idx_path['shuffle'], shuffle_idx, allow_pickle=True) + print_rank_0(' > elasped time to build and save shuffle-idx mapping' + ' (seconds): {:4f}'.format(time.time() - start_time)) except OSError: - print( - f"There was an error trying to create the data cache directory ({data_cache_dir})" - ) - print( - 'or a file in it. This defaults to a directory "index-cache" within the directory' - ) - print( - "the data files are in and can be set with the --data-cache-path argument. Please" - ) - print( - "ensure you have write access to this directory or specify one that you do have" - ) - print("write access to.") + print(f'There was an error trying to create the data cache directory ({data_cache_dir})') + print('or a file in it. This defaults to a directory "index-cache" within the directory') + print('the data files are in and can be set with the --data-cache-path argument. Please') + print('ensure you have write access to this directory or specify one that you do have') + print('write access to.') data_cache_success = False # Load mappings. start_time = time.time() - log.debug(f" > loading doc-idx mapping from {idx_path['doc']}") - doc_idx = np.load(idx_path["doc"], allow_pickle=True, mmap_mode="r") + print_rank_0(f" > loading doc-idx mapping from {idx_path['doc']}") + doc_idx = np.load(idx_path['doc'], allow_pickle=True, mmap_mode='r') - log.debug(f" > loading sample-idx mapping from {idx_path['sample']}") - sample_idx = np.load(idx_path["sample"], allow_pickle=True, mmap_mode="r") + print_rank_0(f" > loading sample-idx mapping from {idx_path['sample']}") + sample_idx = np.load(idx_path['sample'], allow_pickle=True, mmap_mode='r') - log.debug(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") - shuffle_idx = np.load(idx_path["shuffle"], allow_pickle=True, mmap_mode="r") + print_rank_0(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") + shuffle_idx = np.load(idx_path['shuffle'], allow_pickle=True, mmap_mode='r') - log.debug( - " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) - ) - log.debug(" total number of samples: {}".format(sample_idx.shape[0])) - log.debug(" total number of epochs: {}".format(num_epochs)) + print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( + time.time() - start_time)) + print_rank_0(' total number of samples: {}'.format( + sample_idx.shape[0])) + print_rank_0(' total number of epochs: {}'.format(num_epochs)) return doc_idx, sample_idx, shuffle_idx, desc, desc_hash @@ -996,26 +730,25 @@ def _num_epochs(tokens_per_epoch, seq_length, num_samples): if ((total_tokens - 1) // seq_length) >= num_samples: return num_epochs - @dlp.log def _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch): """Build an array with length = number-of-epochs * number-of-dcuments. Each index is mapped to a corresponding document.""" if not separate_last_epoch or num_epochs == 1: - doc_idx = np.mgrid[0:num_epochs, 0 : len(documents)][1] + doc_idx = np.mgrid[0:num_epochs, 0:len(documents)][1] doc_idx[:] = documents doc_idx = doc_idx.reshape(-1) doc_idx = doc_idx.astype(np.int32) np_rng.shuffle(doc_idx) return doc_idx - doc_idx_first = _build_doc_idx(documents, num_epochs - 1, np_rng, False) + doc_idx_first = _build_doc_idx(documents, num_epochs-1, np_rng, False) doc_idx_last = _build_doc_idx(documents, 1, np_rng, False) return np.concatenate((doc_idx_first, doc_idx_last)) - @dlp.log -def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): +def _build_sample_idx(sizes, doc_idx, seq_length, + num_epochs, tokens_per_epoch): """Sample index mapping is a 2D array with sizes [number-of-samples + 1, 2] where [..., 0] contains the index into `doc_idx` and [..., 1] is the @@ -1049,7 +782,7 @@ def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): # Note that -1 here is for the same reason we have -1 in # `_num_epochs` calculations. if remaining_seq_length <= 0: - doc_offset += remaining_seq_length + doc_length - 1 + doc_offset += (remaining_seq_length + doc_length - 1) remaining_seq_length = 0 else: # Otherwise, start from the begining of the next document. @@ -1062,28 +795,24 @@ def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): return sample_idx - @dlp.log def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" - log.debug( - " > building shuffle index with split [0, {}) and [{}, {}) " "...".format( - num_samples, num_samples, total_size - ) - ) + print_rank_0(' > building shuffle index with split [0, {}) and [{}, {}) ' + '...'.format(num_samples, num_samples, total_size)) dtype_ = np.uint32 if total_size >= (np.iinfo(np.uint32).max - 1): dtype_ = np.int64 - shuffle_idx_first = np.arange(start=0, stop=num_samples, step=1, dtype=dtype_) + shuffle_idx_first = np.arange(start=0, stop=num_samples, + step=1, dtype=dtype_) np_rng.shuffle(shuffle_idx_first) if num_samples == total_size: return shuffle_idx_first - shuffle_idx_last = np.arange( - start=num_samples, stop=total_size, step=1, dtype=dtype_ - ) + shuffle_idx_last = np.arange(start=num_samples, stop=total_size, + step=1, dtype=dtype_) np_rng.shuffle(shuffle_idx_last) return np.concatenate((shuffle_idx_first, shuffle_idx_last)) From 52a406ce7005f7be260462688020f325190a48b8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 14 Oct 2024 23:51:07 -0500 Subject: [PATCH 348/590] Consistent logging in `megatron/data/*.py` --- megatron/data/blendable_dataset.py | 14 +- megatron/data/gpt_dataset.py | 907 ++++++++++++++++++----------- megatron/data/indexed_dataset.py | 16 - 3 files changed, 590 insertions(+), 347 deletions(-) mode change 100755 => 100644 megatron/data/gpt_dataset.py diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 590a379971d..979e9a174e7 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -49,8 +49,10 @@ def _build_indices(): helpers.build_blending_indices(dataset_index, dataset_sample_index, weights, num_datasets, self.size, torch.distributed.get_rank() == 0) - log.info('> elapsed time for building blendable dataset indices: ' - '{:.2f} (sec)'.format(time.time() - start_time)) + log.info( + "> elapsed time for building blendable dataset indices: " + f"{time.perf_counter() - start_time:.2f} (sec)" + ) return dataset_index, dataset_sample_index desc = "Blendable dataset\n\n" @@ -74,7 +76,7 @@ def _build_indices(): ' dataset, building indices on rank 0 ...', flush=True) dataset_index, dataset_sample_index = _build_indices() try: - log.info(" > saving index map files") + log.debug(" > saving index map files") start_time = time.time() os.makedirs(os.path.dirname(index_path), exist_ok=True) with open(desc_path, 'wt') as fd: @@ -105,7 +107,7 @@ def _build_indices(): torch.distributed.barrier(group=mpu.get_data_parallel_group()) torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) - + start_time = time.time() log.info(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') @@ -113,7 +115,7 @@ def _build_indices(): log.info(f'> loading blendable dataset sample index: {sample_index_path}') self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size - log.info(f'> finished loading in {time.time() - start_time} seconds') + log.info(f'> finished loading in {time.time() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() @@ -139,4 +141,4 @@ def __getitem__(self, idx): return { "dataset_idx" : dataset_idx, **self.datasets[dataset_idx][sample_idx], - } + } diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py old mode 100755 new mode 100644 index 0cf97356a41..c412d02b31c --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -9,67 +9,96 @@ import numpy as np import torch from deepspeed.accelerator import get_accelerator -from megatron import print_rank_0, is_rank_0, get_args +from megatron import is_rank_0, get_args from megatron.core import mpu -from megatron.data import helpers +from megatron.data import helpers # type:ignore from megatron.data.blendable_dataset import BlendableDataset -from megatron.data.dataset_utils import get_datasets_weights_and_num_samples, get_datasets_corpuses_weights_and_num_samples +from megatron.data.dataset_utils import ( + get_datasets_weights_and_num_samples, + get_datasets_corpuses_weights_and_num_samples, +) from megatron.data.dataset_utils import get_train_valid_test_split_ from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset -from megatron.utils import PerfTrace, Profile +from megatron.utils import PerfTrace, Profile, get_logger from mpi4py import MPI dlp = Profile("DATASET") +log = get_logger(__name__, rank_zero_only=True) + + @dlp.log -def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - train_data_prefix=None, - valid_data_prefix=None, - test_data_prefix=None, - return_doc_ids=False, *, - data_cache_path=None): +def build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + train_data_prefix=None, + valid_data_prefix=None, + test_data_prefix=None, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" if data_prefix: - print_rank_0("Single data path provided for train, valid & test") + log.debug("Single data path provided for train, valid & test") # Single dataset. if len(data_prefix) == 1: - return _build_train_valid_test_datasets(data_prefix[0], - data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + return _build_train_valid_test_datasets( + data_prefix[0], + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) # Blending dataset. # Parse the values. - output = get_datasets_corpuses_weights_and_num_samples(data_prefix, - train_valid_test_num_samples) + output = get_datasets_corpuses_weights_and_num_samples( + data_prefix, train_valid_test_num_samples + ) prefixes, corpuses, weights, datasets_train_valid_test_num_samples = output corpus_list = sorted(set(corpuses)) train_num_samples, valid_num_samples, test_num_samples = map( - sum, - zip(*datasets_train_valid_test_num_samples) + sum, zip(*datasets_train_valid_test_num_samples) ) class DatasetBuilder: - ''' + """ This is for building individual dataset from each dataset file - ''' + """ + @dlp.log - def __init__(self, prefix, corpus, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - return_doc_ids, - data_cache_path=data_cache_path, name='train'): + def __init__( + self, + prefix, + corpus, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + return_doc_ids, + data_cache_path=data_cache_path, + name="train", + ): self.prefix = prefix self.data_impl = data_impl self.splits_string = splits_string - if name == 'train': + if name == "train": self.num_samples = num_samples[0] - elif name == 'valid': + elif name == "valid": self.num_samples = num_samples[1] else: self.num_samples = num_samples[2] @@ -84,279 +113,413 @@ def __init__(self, prefix, corpus, data_impl, splits_string, self.desc = prefix + f"{self.num_samples}" + f"{seq_length}" + f"{seed}" self.build = False self.corpus = corpus + @dlp.log def Build(self): - self.dataset = _build_train_valid_test_datasets_single(self.prefix, self.data_impl, self.splits_string, - self.num_samples_train_valid_test, self.seq_length, self.seed, self.skip_warmup, self.name, self.return_doc_ids, - data_cache_path=self.data_cache_path) + self.dataset = _build_train_valid_test_datasets_single( + self.prefix, + self.data_impl, + self.splits_string, + self.num_samples_train_valid_test, + self.seq_length, + self.seed, + self.skip_warmup, + self.name, + self.return_doc_ids, + data_cache_path=self.data_cache_path, + ) self.build = True return self.dataset class BuildConcatDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, dataset_builders): + def __init__(self, dataset_builders, shuffle=False): self.dataset_builders = dataset_builders self.num_datasets = len(dataset_builders) self.num_samples = np.sum([d.num_samples for d in dataset_builders]) - self.indices=np.zeros((self.num_samples, 2), dtype=np.uint64) - self.desc="ConcatDataset:" - m = 0 + self.indices = np.zeros((self.num_samples, 2), dtype=np.uint64) + self.desc = "ConcatDataset:" + # m = 0 num_samples_list = np.array([d.num_samples for d in dataset_builders]) self.num_samples = np.sum(num_samples_list) + def _build_indices(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) - helpers.build_concat_indices(dataset_index, dataset_sample_index, - num_samples_list, - self.num_datasets, - torch.distributed.get_rank()==0) - print_rank_0('> elapsed time for building concat dataset indices: ' - '{:.2f} (sec)'.format(time.time() - start_time)) + helpers.build_concat_indices( + dataset_index, + dataset_sample_index, + num_samples_list, + self.num_datasets, + torch.distributed.get_rank() == 0, + ) + log.debug( + "> elapsed time for building concat dataset indices: " + "{:.2f} (sec)".format(time.time() - start_time) + ) return dataset_index, dataset_sample_index - + self.dataset_index, self.dataset_sample_index = _build_indices() + np_rng = np.random.RandomState(seed=dataset_builders[0].seed) + self.shuffle_index = np.arange(self.num_samples) + if shuffle: + np_rng.shuffle(self.shuffle_index) for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," - self.desc += f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" + f"{dataset_builders[0].seed}" + self.desc += ( + f"-{self.num_samples}" + + f"-{dataset_builders[0].seq_length}" + + f"{dataset_builders[0].seed}" + ) + def __len__(self): return self.num_samples @dlp.log def __getitem__(self, idx): - if idx >= self.num_samples: - print_rank_0(f"WARNING: index overflow encountered {idx} > {self.num_samples} for {self.dataset_builders[0].corpus}; will randomly pick one sample") - id = np.random.randint(self.num_samples) - else: - id = idx - i = self.dataset_index[idx] - j = self.dataset_sample_index[idx] + id_shuffle = self.shuffle_index[idx] + i = self.dataset_index[id_shuffle] + j = self.dataset_sample_index[id_shuffle] if self.dataset_builders[i].build: return self.dataset_builders[i].dataset[j] else: return self.dataset_builders[i].Build()[j] - - # Predetermine whether need to build the specific dataset or not. + # Predetermine whether need to build the specific dataset or not. start_time = time.time() - print_rank_0(" >>> Started building datasets in distributed way ... ") + log.debug(" >>> Started building datasets in distributed way ... ") a, b, c = [int(d) for d in splits_string.split(",")] - + train_datasets = [] valid_datasets = [] test_datasets = [] # Build individual datasets. - + args = get_args() @dlp.log - def build_corpus_datasets(dataset_type='train'): + def build_corpus_datasets(dataset_type="train"): start_time = time.time() - print_rank_0(f" >>> Building {dataset_type} corpus datasets ...") + log.debug(f" >>> Building {dataset_type} corpus datasets ...") datasets = [] corpus_builders = {} corpus_weights = {} for c in corpus_list: corpus_builders[c] = [] corpus_weights[c] = 0.0 - dataset_builders = [DatasetBuilder(prefixes[i], corpuses[i], data_impl, splits_string, - datasets_train_valid_test_num_samples[i], - seq_length, seed, skip_warmup, - return_doc_ids,data_cache_path, dataset_type) for i in range(len(weights))] - for i in range(torch.distributed.get_rank()//mpu.get_tensor_model_parallel_world_size(), len(weights), torch.distributed.get_world_size()//mpu.get_tensor_model_parallel_world_size()): + dataset_builders = [ + DatasetBuilder( + prefixes[i], + corpuses[i], + data_impl, + splits_string, + datasets_train_valid_test_num_samples[i], + seq_length, + seed, + skip_warmup, + return_doc_ids, + data_cache_path, + dataset_type, + ) + for i in range(len(weights)) + ] + for i in range( + torch.distributed.get_rank() + // mpu.get_tensor_model_parallel_world_size(), + len(weights), + torch.distributed.get_world_size() + // mpu.get_tensor_model_parallel_world_size(), + ): dataset_builders[i].Build() - print_rank_0(f" >>> Finished building individual datasets in {time.time() - start_time} seconds") + log.debug( + f" >>> Finished building individual datasets in {time.time() - start_time} seconds" + ) start_concating_time = time.time() for i, d in zip(range(len(weights)), dataset_builders): corpus_builders[d.corpus].append(d) corpus_weights[d.corpus] += weights[i] total = 0 - print_rank_0(" > number of samples for each corpus ") - corpus_weights_achieved={} + log.debug(" > number of samples for each corpus ") + corpus_weights_achieved = {} for c in corpus_list: - datasets.append(BuildConcatDataset(corpus_builders[c])) + datasets.append(BuildConcatDataset(corpus_builders[c], args.shuffle_sample)) total += datasets[-1].num_samples - corpus_weights_achieved[c] = float(datasets[-1].num_samples)/train_num_samples - print_rank_0(f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})") - - print_rank_0(f" > total number of samples: {total}") - print_rank_0(f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds") - print_rank_0(f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds") + corpus_weights_achieved[c] = ( + float(datasets[-1].num_samples) / train_num_samples + ) + log.debug( + f" {c}: {datasets[-1].num_samples} w={corpus_weights_achieved[c]} (expected: {corpus_weights[c]})" + ) + + log.debug(f" > total number of samples: {total}") + log.debug( + f" >>> Finished concatenating datasets in {time.time() - start_concating_time} seconds" + ) + log.debug( + f" >>> Finished building {dataset_type} corpus datasets in {time.time() - start_time} seconds" + ) return datasets, [corpus_weights_achieved[c] for c in corpus_list] + train_weights = None if a > 0: - train_datasets, train_weights = build_corpus_datasets('train') - + train_datasets, train_weights = build_corpus_datasets("train") + valid_weights = None if b > 0: - valid_datasets, valid_weights = build_corpus_datasets('valid') - - if c > 0: - test_datasets, test_weights = build_corpus_datasets('test') + valid_datasets, valid_weights = build_corpus_datasets("valid") + test_weights = None + if c > 0: + test_datasets, test_weights = build_corpus_datasets("test") # This barrier is critical to make sure that all the datasets are built once # and the metadata were written to the cache folder before other ranks touch them - print_rank_0(f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds") + log.debug( + f" >>> Rank 0 - finished building datasets in {time.time() - start_time} seconds" + ) torch.distributed.barrier(group=mpu.get_data_parallel_group()) torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) - print_rank_0(f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds") - print_rank_0(f" >>> Starting to build BlendableDataset") + log.debug( + f" >>> Finished building datasets (all ranks) in distributed way in {time.time() - start_time} seconds" + ) + log.debug(" >>> Starting to build BlendableDataset") # Blend. start_time = time.time() blending_train_dataset = None - if train_datasets: - blending_train_dataset = BlendableDataset(train_datasets, train_weights, train_num_samples, - data_cache_path=data_cache_path) + if train_datasets and train_weights: + blending_train_dataset = BlendableDataset( + train_datasets, + train_weights, + train_num_samples, + data_cache_path=data_cache_path, + ) blending_valid_dataset = None - if valid_datasets: - blending_valid_dataset = BlendableDataset(valid_datasets, valid_weights, valid_num_samples, - data_cache_path=data_cache_path) + if valid_datasets and valid_weights: + blending_valid_dataset = BlendableDataset( + valid_datasets, + valid_weights, + valid_num_samples, + data_cache_path=data_cache_path, + ) blending_test_dataset = None - if test_datasets: - blending_test_dataset = BlendableDataset(test_datasets, test_weights, test_num_samples, - data_cache_path=data_cache_path) + if test_datasets and test_weights: + blending_test_dataset = BlendableDataset( + test_datasets, + test_weights, + test_num_samples, + data_cache_path=data_cache_path, + ) end_time = time.time() - print_rank_0(f" >>> Finished building BlendableDataset in {end_time - start_time} seconds") - return (blending_train_dataset, blending_valid_dataset, - blending_test_dataset) + log.debug( + f" >>> Finished building BlendableDataset in {end_time - start_time} seconds" + ) + return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) else: - print_rank_0("Separate data paths provided for train, valid & test. Split string will be ignored.") + log.debug( + "Separate data paths provided for train, valid & test. Split string will be ignored." + ) train_dataset, valid_dataset, test_dataset = None, None, None # Single dataset. if train_data_prefix is not None: - train_dataset = build_dataset("train", train_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[0], - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + train_dataset = build_dataset( + "train", + train_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[0], + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) if valid_data_prefix is not None: - valid_dataset = build_dataset("valid", valid_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[1], - seq_length, seed, False, - data_cache_path=data_cache_path) - + valid_dataset = build_dataset( + "valid", + valid_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[1], + seq_length, + seed, + False, + data_cache_path=data_cache_path, + ) if test_data_prefix is not None: - test_dataset = build_dataset("test", test_data_prefix, data_impl, - splits_string, - train_valid_test_num_samples[2], - seq_length, seed, False, - data_cache_path=data_cache_path) + test_dataset = build_dataset( + "test", + test_data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples[2], + seq_length, + seed, + False, + data_cache_path=data_cache_path, + ) return (train_dataset, valid_dataset, test_dataset) + @dlp.log -def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, - return_doc_ids=False, *, - data_cache_path=None): +def _build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + log.debug(" > dataset split:") def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' - 'documents'.format(splits[index], splits[index + 1], - splits[index + 1] - splits[index])) - print_split_stats('train', 0) - print_split_stats('validation', 1) - print_split_stats('test', 2) + log.debug(" {}:".format(name)) + log.debug( + " document indices in [{}, {}) total of {} " "documents".format( + splits[index], splits[index + 1], splits[index + 1] - splits[index] + ) + ) + + print_split_stats("train", 0) + print_split_stats("validation", 1) + print_split_stats("test", 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange(start=splits[index], stop=splits[index + 1], - step=1, dtype=np.int32) - dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, seed, - return_doc_ids, - data_cache_path=data_cache_path) + documents = np.arange( + start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 + ) + dataset = GPTDataset( + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, + seed, + return_doc_ids, + data_cache_path=data_cache_path, + ) return dataset - train_dataset = build_dataset(0, 'train') - valid_dataset = build_dataset(1, 'valid') - test_dataset = build_dataset(2, 'test') + train_dataset = build_dataset(0, "train") + valid_dataset = build_dataset(1, "valid") + test_dataset = build_dataset(2, "test") return (train_dataset, valid_dataset, test_dataset) + @dlp.log -def _build_train_valid_test_datasets_single(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - seq_length, seed, skip_warmup, name, - return_doc_ids=False, *, - data_cache_path=None): +def _build_train_valid_test_datasets_single( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + seq_length, + seed, + skip_warmup, + name, + return_doc_ids=False, + *, + data_cache_path=None, +): """Build train, valid, and test datasets.""" # Each rank print out information - print_rank_0(f" >> building dataset for {data_prefix}") + log.debug(f" >> building dataset for {data_prefix}") # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + log.debug(" > dataset split:") def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' - 'documents'.format(splits[index], splits[index + 1], - splits[index + 1] - splits[index])) - print_split_stats('train', 0) - print_split_stats('validation', 1) - print_split_stats('test', 2) + log.debug(" {}:".format(name)) + log.debug( + " document indices in [{}, {}) total of {} " "documents".format( + splits[index], splits[index + 1], splits[index + 1] - splits[index] + ) + ) + + print_split_stats("train", 0) + print_split_stats("validation", 1) + print_split_stats("test", 2) def build_dataset(index, name): dataset = None if splits[index + 1] > splits[index]: - documents = np.arange(start=splits[index], stop=splits[index + 1], - step=1, dtype=np.int32) - dataset = GPTDataset(name, data_prefix, documents, indexed_dataset, - splits_string, - train_valid_test_num_samples[index], - seq_length, seed, - return_doc_ids, - data_cache_path=data_cache_path) + documents = np.arange( + start=splits[index], stop=splits[index + 1], step=1, dtype=np.int32 + ) + dataset = GPTDataset( + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + train_valid_test_num_samples[index], + seq_length, + seed, + return_doc_ids, + data_cache_path=data_cache_path, + ) return dataset - if name.find("train")!=-1: - return build_dataset(0, 'train') - if name.find("valid")!=-1: - return build_dataset(1, 'valid') - if name.find("test")!=-1: - return build_dataset(2, 'test') + + if name.find("train") != -1: + return build_dataset(0, "train") + if name.find("valid") != -1: + return build_dataset(1, "valid") + if name.find("test") != -1: + return build_dataset(2, "test") + @dlp.log -def build_dataset(dataset_name, data_prefix, data_impl, - splits_string, num_samples, - seq_length, seed, skip_warmup, - *, - data_cache_path=None): +def build_dataset( + dataset_name, + data_prefix, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + *, + data_cache_path=None, +): dataset = None if len(data_prefix) == 1: - dataset = _build_dataset(dataset_name, data_prefix[0], data_impl, - splits_string, num_samples, seq_length, - seed, skip_warmup, - data_cache_path=data_cache_path) + dataset = _build_dataset( + dataset_name, + data_prefix[0], + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) else: # Blending dataset. # Parse the values. @@ -367,73 +530,108 @@ def build_dataset(dataset_name, data_prefix, data_impl, # Build individual datasets. datasets = [] for i in range(len(prefixes)): - ds = _build_dataset(dataset_name, prefixes[i], data_impl, - splits_string, dataset_num_samples[i], - seq_length, seed, skip_warmup, - data_cache_path=data_cache_path) + ds = _build_dataset( + dataset_name, + prefixes[i], + data_impl, + splits_string, + dataset_num_samples[i], + seq_length, + seed, + skip_warmup, + data_cache_path=data_cache_path, + ) if ds: datasets.append(ds) if datasets: - dataset = BlendableDataset(datasets, weights, num_samples, - data_cache_path=data_cache_path) + dataset = BlendableDataset( + datasets, weights, num_samples, data_cache_path=data_cache_path + ) return dataset + @dlp.log -def _build_dataset(dataset_name, data_prefix, data_impl, splits_string, - num_samples, seq_length, seed, skip_warmup, - *, - data_cache_path=None): +def _build_dataset( + dataset_name, + data_prefix, + data_impl, + splits_string, + num_samples, + seq_length, + seed, + skip_warmup, + *, + data_cache_path=None, +): """ Build dataset. This method is called when individual train, valid, test datasets are provided """ # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) total_num_of_documents = indexed_dataset.sizes.shape[0] - print_rank_0(' {}:'.format(dataset_name)) - print_rank_0(' document indices in [0, {}) total of {} ' - 'documents'.format(total_num_of_documents, total_num_of_documents)) - - documents = np.arange(start=0, stop=total_num_of_documents, - step=1, dtype=np.int32) - - dataset = GPTDataset(dataset_name, data_prefix, documents, indexed_dataset, - splits_string, num_samples, seq_length, seed, - data_cache_path=data_cache_path) + log.debug(" {}:".format(dataset_name)) + log.debug( + " document indices in [0, {}) total of {} " "documents".format( + total_num_of_documents, total_num_of_documents + ) + ) + + documents = np.arange(start=0, stop=total_num_of_documents, step=1, dtype=np.int32) + + dataset = GPTDataset( + dataset_name, + data_prefix, + documents, + indexed_dataset, + splits_string, + num_samples, + seq_length, + seed, + data_cache_path=data_cache_path, + ) return dataset + @dlp.log def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): """Build indexed dataset.""" - print_rank_0(' > building dataset index ...') + log.debug(" > building dataset index ...") start_time = time.time() - indexed_dataset = make_indexed_dataset(data_prefix, - data_impl, - skip_warmup) - print_rank_0(' > finished creating indexed dataset in {:4f} ' - 'seconds'.format(time.time() - start_time)) - print_rank_0(' number of documents: {}'.format( - indexed_dataset.sizes.shape[0])) + indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) + log.debug( + " > finished creating indexed dataset in {:4f} " "seconds".format( + time.time() - start_time + ) + ) + log.debug(" number of documents: {}".format(indexed_dataset.sizes.shape[0])) return indexed_dataset class GPTDataset(torch.utils.data.Dataset): @dlp.log - def __init__(self, name, data_prefix, documents, indexed_dataset, - splits_string, num_samples, seq_length, seed, - return_doc_ids=False, *, - data_cache_path=None): - + def __init__( + self, + name, + data_prefix, + documents, + indexed_dataset, + splits_string, + num_samples, + seq_length, + seed, + return_doc_ids=False, + *, + data_cache_path=None, + ): self.name = name self.indexed_dataset = indexed_dataset self.return_doc_ids = return_doc_ids @@ -443,20 +641,29 @@ def __init__(self, name, data_prefix, documents, indexed_dataset, assert np.max(documents) < indexed_dataset.sizes.shape[0] # Build index mappings. - self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = \ - _build_index_mappings(self.name, data_prefix, - documents, self.indexed_dataset.sizes, - splits_string, num_samples, seq_length, seed, - data_cache_path=data_cache_path) - + self.doc_idx, self.sample_idx, self.shuffle_idx, self.desc, self.desc_hash = ( + _build_index_mappings( + self.name, + data_prefix, + documents, + self.indexed_dataset.sizes, + splits_string, + num_samples, + seq_length, + seed, + data_cache_path=data_cache_path, + ) + ) def __len__(self): # -1 is due to data structure used to retieve the index: # sample i --> [sample_idx[i], sample_idx[i+1]) return self.sample_idx.shape[0] - 1 + @dlp.log def __getitem__(self, idx): args = get_args() + assert args is not None orig_idx = idx # Get the shuffled index. try: @@ -465,21 +672,24 @@ def __getitem__(self, idx): if is_rank_0(): import json from rich import print_json + print(exc) print( - '\n'.join( - ['-------------------------------------------------', - f'Trying to access {idx=} from self.shuffle_idx,', - f'but {len(self.shuffle_idx)=}', - '-------------------------------------------------'] + "\n".join( + [ + "-------------------------------------------------", + f"Trying to access {idx=} from self.shuffle_idx,", + f"but {len(self.shuffle_idx)=}", + "-------------------------------------------------", + ] ) ) print_json( json.dumps( { - 'doc_idx': len(self.doc_idx), - 'sample_idx': len(self.sample_idx), - 'shuffle_idx': len(self.shuffle_idx), + "doc_idx": len(self.doc_idx), + "sample_idx": len(self.sample_idx), + "shuffle_idx": len(self.shuffle_idx), }, indent=4, ) @@ -493,45 +703,57 @@ def __getitem__(self, idx): doc_ids = [] if doc_index_f == doc_index_l: doc_ids.append(self.doc_idx[doc_index_f]) - sample = self.indexed_dataset.get(self.doc_idx[doc_index_f], - offset=offset_f, - length=offset_l - offset_f + 1) + sample = self.indexed_dataset.get( + self.doc_idx[doc_index_f], + offset=offset_f, + length=offset_l - offset_f + 1, + ) else: # Otherwise, get the rest of the initial document. doc_ids.append(self.doc_idx[doc_index_f]) - sample_list = [self.indexed_dataset.get(self.doc_idx[doc_index_f], - offset=offset_f)] + sample_list = [ + self.indexed_dataset.get(self.doc_idx[doc_index_f], offset=offset_f) + ] # Loop over all in between documents and add the entire document. for i in range(doc_index_f + 1, doc_index_l): doc_ids.append(self.doc_idx[i]) sample_list.append(self.indexed_dataset.get(self.doc_idx[i])) # And finally add the relevant portion of last document. doc_ids.append(self.doc_idx[doc_index_l]) - sample_list.append(self.indexed_dataset.get( - self.doc_idx[doc_index_l], - length=offset_l + 1)) + sample_list.append( + self.indexed_dataset.get(self.doc_idx[doc_index_l], length=offset_l + 1) + ) sample = np.concatenate(sample_list) - text_name = 'text' + text_name = "text" if args.use_dataset_only: - text_name = 'input_ids' + text_name = "input_ids" sample_dict = {text_name: np.array(sample, dtype=np.int64)} if args.return_data_index: - sample_dict.update({'index': np.array([orig_idx], dtype=np.int64)}) + sample_dict.update({"index": np.array([orig_idx], dtype=np.int64)}) - if self.return_doc_ids: # for retro preprocessing - sample_dict.update({'doc_ids': np.array(doc_ids, dtype=np.int64)}) + if self.return_doc_ids: # for retro preprocessing + sample_dict.update({"doc_ids": np.array(doc_ids, dtype=np.int64)}) if args.use_dataset_only: - sample_dict.update({'labels': np.array(sample, dtype=np.int64)}) + sample_dict.update({"labels": np.array(sample, dtype=np.int64)}) return sample_dict + @dlp.log -def _build_index_mappings(name, data_prefix, documents, sizes, - splits_string, num_samples, seq_length, seed, - *, - data_cache_path): +def _build_index_mappings( + name, + data_prefix, + documents, + sizes, + splits_string, + num_samples, + seq_length, + seed, + *, + data_cache_path, +): """Build doc-idx, sample-idx, and shuffle-idx. doc-idx: is an array (ordered) of documents to be used in training. sample-idx: is the start document index and document offset for each @@ -539,10 +761,11 @@ def _build_index_mappings(name, data_prefix, documents, sizes, shuffle-idx: maps the sample index into a random index into sample-idx. """ args = get_args() + assert args is not None # Number of tokens in each epoch and number of required epochs. tokens_per_epoch = _num_tokens(documents, sizes) num_epochs = _num_epochs(tokens_per_epoch, seq_length, num_samples) - if args.train_data_exact_num_epochs is not None and name == 'train': + if args.train_data_exact_num_epochs is not None and name == "train": num_epochs = args.train_data_exact_num_epochs # rng state @@ -557,13 +780,13 @@ def _build_index_mappings(name, data_prefix, documents, sizes, desc += f"Sequence length {seq_length}\n" desc += f"Random seed {seed}\n" desc += f"Split {splits_string}\n" - desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() + desc_hash = hashlib.md5(desc.encode("utf-8")).hexdigest() desc_filename = desc_hash + ".dsc" - doc_idx_filename = desc_hash + '_doc_idx.npy' - sample_idx_filename = desc_hash + '_sample_idx.npy' - shuffle_idx_filename = desc_hash + '_shuffle_idx.npy' + doc_idx_filename = desc_hash + "_doc_idx.npy" + sample_idx_filename = desc_hash + "_sample_idx.npy" + shuffle_idx_filename = desc_hash + "_shuffle_idx.npy" - if name == 'train': + if name == "train": # force to use certain index files if args.train_desc_path is not None: desc_filename = args.train_desc_path @@ -578,15 +801,15 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # duplication, then look in data-cache-path if specified, # If nothing is found, use the last path looked in build_indices = True - prefixes = [os.path.join(os.path.dirname(data_prefix), 'index-cache')] + prefixes = [os.path.join(os.path.dirname(data_prefix), "index-cache")] if data_cache_path is not None: prefixes.append(data_cache_path) for prefix in prefixes: idx_path = { - 'desc': os.path.join(prefix, desc_filename), - 'doc': os.path.join(prefix, doc_idx_filename), - 'sample': os.path.join(prefix, sample_idx_filename), - 'shuffle': os.path.join(prefix, shuffle_idx_filename) + "desc": os.path.join(prefix, desc_filename), + "doc": os.path.join(prefix, doc_idx_filename), + "sample": os.path.join(prefix, sample_idx_filename), + "shuffle": os.path.join(prefix, shuffle_idx_filename), } for f in idx_path.values(): if not os.path.isfile(f): @@ -595,15 +818,17 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # Found our files! build_indices = False break - data_cache_dir = os.path.dirname(idx_path['desc']) + data_cache_dir = os.path.dirname(idx_path["desc"]) data_cache_success = True # Build the indexed mapping if not exist. if build_indices: - # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the - # ranks will first create the document files, and then read it. + # Since this function will be called by all the rank in the very beginning. Therefore, we assume that all the + # ranks will first create the document files, and then read it. # There will not be contension effects going on either - print_rank_0(f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}") + log.warning( + f" > WARNING: could not find index map files, building on rank {torch.distributed.get_rank()}" + ) # For the last epoch, decide whether include the entire epoch # in the global shuffle or not. @@ -612,64 +837,80 @@ def _build_index_mappings(name, data_prefix, documents, sizes, # not mean anything. if num_epochs == 1: separate_last_epoch = False - print_rank_0(' > only one epoch required, setting ' - 'separate_last_epoch to False') + log.debug( + " > only one epoch required, setting " "separate_last_epoch to False" + ) else: # Get the number of samples for the last epoch num_samples_from_epochs_minus_one = ( - (num_epochs - 1) * tokens_per_epoch - 1) // seq_length - last_epoch_num_samples = num_samples - \ - num_samples_from_epochs_minus_one - assert last_epoch_num_samples >= 0, \ - 'last epoch number of samples should be non-negative.' + (num_epochs - 1) * tokens_per_epoch - 1 + ) // seq_length + last_epoch_num_samples = num_samples - num_samples_from_epochs_minus_one + assert ( + last_epoch_num_samples >= 0 + ), "last epoch number of samples should be non-negative." num_samples_per_epoch = (tokens_per_epoch - 1) // seq_length - assert last_epoch_num_samples <= (num_samples_per_epoch + 1), \ - 'last epoch number of samples exceeded max value.' + assert last_epoch_num_samples <= ( + num_samples_per_epoch + 1 + ), "last epoch number of samples exceeded max value." # If we have less than 80% of the samples for the last epoch, # seperate out the epoch and treat it differently. # Note: the 80% number is just based on common sense and can # be adjusted if needed. - separate_last_epoch = (last_epoch_num_samples < - int(0.80 * num_samples_per_epoch)) + separate_last_epoch = last_epoch_num_samples < int( + 0.80 * num_samples_per_epoch + ) if separate_last_epoch: - string = ' > last epoch number of samples ({}) is smaller '\ - 'than 80% of number of samples per epoch ({}), '\ - 'setting separate_last_epoch to True' + string = ( + " > last epoch number of samples ({}) is smaller " + "than 80% of number of samples per epoch ({}), " + "setting separate_last_epoch to True" + ) else: - string = ' > last epoch number of samples ({}) is larger '\ - 'than 80% of number of samples per epoch ({}), '\ - 'setting separate_last_epoch to False' - print_rank_0(string.format(last_epoch_num_samples, - num_samples_per_epoch)) - + string = ( + " > last epoch number of samples ({}) is larger " + "than 80% of number of samples per epoch ({}), " + "setting separate_last_epoch to False" + ) + log.debug(string.format(last_epoch_num_samples, num_samples_per_epoch)) try: os.makedirs(data_cache_dir, exist_ok=True) # description - with open(idx_path['desc'], 'wt') as fd: + with open(idx_path["desc"], "wt") as fd: fd.write(desc) # doc-idx. start_time = time.time() - doc_idx = _build_doc_idx(documents, num_epochs, np_rng, - separate_last_epoch) - np.save(idx_path['doc'], doc_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save doc-idx mapping ' - '(seconds): {:4f}'.format(time.time() - start_time)) + doc_idx = _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch) + np.save(idx_path["doc"], doc_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save doc-idx mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # sample-idx. start_time = time.time() # Use C++ implementation for speed. # First compile and then import. from megatron.data import helpers + assert doc_idx.dtype == np.int32 assert sizes.dtype == np.int32 - sample_idx = helpers.build_sample_idx(sizes, doc_idx, seq_length, - num_epochs, tokens_per_epoch, torch.distributed.get_rank()==0) - np.save(idx_path['sample'], sample_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save sample-idx mapping ' - '(seconds): {:4f}'.format(time.time() - start_time)) + sample_idx = helpers.build_sample_idx( + sizes, + doc_idx, + seq_length, + num_epochs, + tokens_per_epoch, + torch.distributed.get_rank() == 0, + ) + np.save(idx_path["sample"], sample_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save sample-idx mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # shuffle-idx. start_time = time.time() # -1 is due to data structure used to retieve the index: @@ -678,35 +919,46 @@ def _build_index_mappings(name, data_prefix, documents, sizes, num_samples_ = num_samples_from_epochs_minus_one else: num_samples_ = sample_idx.shape[0] - 1 - shuffle_idx = _build_shuffle_idx(num_samples_, - sample_idx.shape[0] - 1, np_rng) - np.save(idx_path['shuffle'], shuffle_idx, allow_pickle=True) - print_rank_0(' > elasped time to build and save shuffle-idx mapping' - ' (seconds): {:4f}'.format(time.time() - start_time)) + shuffle_idx = _build_shuffle_idx( + num_samples_, sample_idx.shape[0] - 1, np_rng + ) + np.save(idx_path["shuffle"], shuffle_idx, allow_pickle=True) + log.debug( + " > elasped time to build and save shuffle-idx mapping" + " (seconds): {:4f}".format(time.time() - start_time) + ) except OSError: - print(f'There was an error trying to create the data cache directory ({data_cache_dir})') - print('or a file in it. This defaults to a directory "index-cache" within the directory') - print('the data files are in and can be set with the --data-cache-path argument. Please') - print('ensure you have write access to this directory or specify one that you do have') - print('write access to.') + print( + f"There was an error trying to create the data cache directory ({data_cache_dir})" + ) + print( + 'or a file in it. This defaults to a directory "index-cache" within the directory' + ) + print( + "the data files are in and can be set with the --data-cache-path argument. Please" + ) + print( + "ensure you have write access to this directory or specify one that you do have" + ) + print("write access to.") data_cache_success = False # Load mappings. start_time = time.time() - print_rank_0(f" > loading doc-idx mapping from {idx_path['doc']}") - doc_idx = np.load(idx_path['doc'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading doc-idx mapping from {idx_path['doc']}") + doc_idx = np.load(idx_path["doc"], allow_pickle=True, mmap_mode="r") - print_rank_0(f" > loading sample-idx mapping from {idx_path['sample']}") - sample_idx = np.load(idx_path['sample'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading sample-idx mapping from {idx_path['sample']}") + sample_idx = np.load(idx_path["sample"], allow_pickle=True, mmap_mode="r") - print_rank_0(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") - shuffle_idx = np.load(idx_path['shuffle'], allow_pickle=True, mmap_mode='r') + log.debug(f" > loading shuffle-idx mapping from {idx_path['shuffle']}") + shuffle_idx = np.load(idx_path["shuffle"], allow_pickle=True, mmap_mode="r") - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( - time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( - sample_idx.shape[0])) - print_rank_0(' total number of epochs: {}'.format(num_epochs)) + log.debug( + " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) + ) + log.debug(" total number of samples: {}".format(sample_idx.shape[0])) + log.debug(" total number of epochs: {}".format(num_epochs)) return doc_idx, sample_idx, shuffle_idx, desc, desc_hash @@ -730,25 +982,26 @@ def _num_epochs(tokens_per_epoch, seq_length, num_samples): if ((total_tokens - 1) // seq_length) >= num_samples: return num_epochs + @dlp.log def _build_doc_idx(documents, num_epochs, np_rng, separate_last_epoch): """Build an array with length = number-of-epochs * number-of-dcuments. Each index is mapped to a corresponding document.""" if not separate_last_epoch or num_epochs == 1: - doc_idx = np.mgrid[0:num_epochs, 0:len(documents)][1] + doc_idx = np.mgrid[0:num_epochs, 0 : len(documents)][1] doc_idx[:] = documents doc_idx = doc_idx.reshape(-1) doc_idx = doc_idx.astype(np.int32) np_rng.shuffle(doc_idx) return doc_idx - doc_idx_first = _build_doc_idx(documents, num_epochs-1, np_rng, False) + doc_idx_first = _build_doc_idx(documents, num_epochs - 1, np_rng, False) doc_idx_last = _build_doc_idx(documents, 1, np_rng, False) return np.concatenate((doc_idx_first, doc_idx_last)) + @dlp.log -def _build_sample_idx(sizes, doc_idx, seq_length, - num_epochs, tokens_per_epoch): +def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): """Sample index mapping is a 2D array with sizes [number-of-samples + 1, 2] where [..., 0] contains the index into `doc_idx` and [..., 1] is the @@ -782,7 +1035,7 @@ def _build_sample_idx(sizes, doc_idx, seq_length, # Note that -1 here is for the same reason we have -1 in # `_num_epochs` calculations. if remaining_seq_length <= 0: - doc_offset += (remaining_seq_length + doc_length - 1) + doc_offset += remaining_seq_length + doc_length - 1 remaining_seq_length = 0 else: # Otherwise, start from the begining of the next document. @@ -795,24 +1048,28 @@ def _build_sample_idx(sizes, doc_idx, seq_length, return sample_idx + @dlp.log def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" - print_rank_0(' > building shuffle index with split [0, {}) and [{}, {}) ' - '...'.format(num_samples, num_samples, total_size)) + log.debug( + " > building shuffle index with split [0, {}) and [{}, {}) " "...".format( + num_samples, num_samples, total_size + ) + ) dtype_ = np.uint32 if total_size >= (np.iinfo(np.uint32).max - 1): dtype_ = np.int64 - shuffle_idx_first = np.arange(start=0, stop=num_samples, - step=1, dtype=dtype_) + shuffle_idx_first = np.arange(start=0, stop=num_samples, step=1, dtype=dtype_) np_rng.shuffle(shuffle_idx_first) if num_samples == total_size: return shuffle_idx_first - shuffle_idx_last = np.arange(start=num_samples, stop=total_size, - step=1, dtype=dtype_) + shuffle_idx_last = np.arange( + start=num_samples, stop=total_size, step=1, dtype=dtype_ + ) np_rng.shuffle(shuffle_idx_last) return np.concatenate((shuffle_idx_first, shuffle_idx_last)) diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index ec2997f7b8d..e2a0c4751ff 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -16,7 +16,6 @@ from functools import lru_cache import os -# import logging import shutil import struct from itertools import accumulate @@ -29,21 +28,6 @@ log = get_logger(__name__) - -try: - import ezpz as ez - RANK = ez.get_rank() -except Exception: - RANK = torch.distributed.get_rank() - -# NOTE: [logging]----------------------------------------------------------- -# - Set logging level to "INFO" on RANK == 0, "CRITICAL" on all other ranks -log = logging.getLogger(__name__) -LOG_LEVEL = str(os.environ.get("LOG_LEVEL", "INFO")).upper() -log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") -# -------------------------------------------------------------------------- - - dlp = Profile("DATASET") From 63b1901b6127cd71f5b54877fad211714499120f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 16 Oct 2024 11:24:09 -0500 Subject: [PATCH 349/590] Update `megatron/data/gpt_dataset.py` --- megatron/data/gpt_dataset.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index c412d02b31c..0a3d898d63f 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -168,6 +168,10 @@ def _build_indices(): for i in range(self.num_datasets): self.desc += dataset_builders[i].prefix + "," + log.info( + f"[BuildConcatDataset] Caught {shuffle=} across" + f" {self.num_samples} samples" + ) self.desc += ( f"-{self.num_samples}" + f"-{dataset_builders[0].seq_length}" From 7ef26bf922262eb80ff58fa481dbc6b2ff84d5ad Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 16 Oct 2024 11:24:37 -0500 Subject: [PATCH 350/590] Use `time.perf_counter` in `megatron/data/blendable_dataset.py` --- megatron/data/blendable_dataset.py | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index 979e9a174e7..ab164fdc489 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -41,7 +41,7 @@ def __init__(self, datasets, weights, size, *, # Build indicies. @dlp.log def _build_indices(): - start_time = time.time() + start_time = time.perf_counter() dataset_index = np.zeros(self.size, dtype=np.int64) dataset_sample_index = np.zeros(self.size, dtype=np.int64) @@ -77,14 +77,14 @@ def _build_indices(): dataset_index, dataset_sample_index = _build_indices() try: log.debug(" > saving index map files") - start_time = time.time() + start_time = time.perf_counter() os.makedirs(os.path.dirname(index_path), exist_ok=True) with open(desc_path, 'wt') as fd: fd.write(desc) np.save(index_path, dataset_index, allow_pickle=True) np.save(sample_index_path, dataset_sample_index, allow_pickle=True) - log.info(f" > finished saving index map files in {time.time() - start_time} seconds") + log.info(f" > finished saving index map files in {time.perf_counter() - start_time} seconds") except OSError: print(f'There was an error trying to create the data cache directory ({data_cache_path})') print('or a file in it. This is set with the --data-cache-path argument. Please') @@ -108,14 +108,14 @@ def _build_indices(): torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) - start_time = time.time() + start_time = time.perf_counter() log.info(f'> loading blendable dataset index: {index_path}') self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_index.size == self.size log.info(f'> loading blendable dataset sample index: {sample_index_path}') self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') assert self.dataset_sample_index.size == self.size - log.info(f'> finished loading in {time.time() - start_time} seconds') + log.info(f'> finished loading in {time.perf_counter() - start_time} seconds') else: self.dataset_index, self.dataset_sample_index = _build_indices() From deb95cd7aa5f677c13b7bee4c0491c04d59d81dd Mon Sep 17 00:00:00 2001 From: Xinyu Lian Date: Thu, 17 Oct 2024 15:02:26 -0500 Subject: [PATCH 351/590] fix init issue for silently ignoring the deepspeed config (#452) --- megatron/initialize.py | 1 + 1 file changed, 1 insertion(+) diff --git a/megatron/initialize.py b/megatron/initialize.py index 538f7fc456f..90acf496ee0 100644 --- a/megatron/initialize.py +++ b/megatron/initialize.py @@ -185,6 +185,7 @@ def setup_deepspeed_random_and_activation_checkpointing(args): deepspeed.checkpointing.configure( mpu, + deepspeed_config=args.deepspeed_config, partition_activations=args.partition_activations, contiguous_checkpointing=args.contigious_checkpointing, num_checkpoints=num_layers, From 68da2dbd0a1a1ab9500cd389a0cfd82d2032ebd6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 17 Oct 2024 17:25:03 -0500 Subject: [PATCH 352/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 312 ++++++++++++++++++++++++++---------------------- 1 file changed, 167 insertions(+), 145 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5df9a2c7a58..bc7753aa718 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -9,7 +9,7 @@ # ```bash # $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed # $ cd Megatron-DeepSpeed -# $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup +# $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup # ``` # # and this will, automatically: @@ -174,48 +174,10 @@ setup_run_cmd() { export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" printf "\n" echo "Using data_cache_path: ${data_cache_path}" - TRAIN_SPLIT="${TRAIN_SPLIT:-100}" - VAL_SPLIT="${VAL_SPLIT:-0}" - TEST_SPLIT="${TEST_SPLIT:-0}" - LOG_INTERVAL="${LOG_INTERVAL:-1}" - DEFAULTS=( - "--split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT}" - "--log-interval ${LOG_INTERVAL}" - "--no-bias-gelu-fusion" - "--no-bias-dropout-fusion" - "--no-masked-softmax-fusion" - "--no-gradient-accumulation-fusion" - "--accumulate-allreduce-grads-in-fp32" - ) - # export DEFAULTS="\ - # --split ${TRAIN_SPLIT},${VAL_SPLIT},${TEST_SPLIT} \ - # --log-interval ${LOG_INTERVAL} \ - # --no-bias-gelu-fusion \ - # --no-bias-dropout-fusion \ - # --no-masked-softmax-fusion \ - # --no-gradient-accumulation-fusion \ - # --accumulate-allreduce-grads-in-fp32" - # OVERRIDE_CKPT_OPT_PARAM="${OVERRIDE_CKPT_OPT_PARAM:-}" - if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then - DEFAULTS+=("--use-checkpoint-opt_param-scheduler") - fi - if [[ "${SP}" -gt 1 ]]; then - DEFAULTS+=( - "--ds-sequence-parallel-size ${SP}" - "--force-ds-sequence-parallel" - ) - fi ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` ################################################################## - LLAMA_ARGS="" - if [[ "${SP}" == 1 ]]; then - export LLAMA_ARGS="${LLAMA_ARGS} " - else - export LLAMA_ARGS="" - echo "NOT USING ROTARY EMBEDDINGS! LLAMA_ARGS=${LLAMA_ARGS}" - fi if [[ -z "${NO_LLAMA:-}" ]]; then llama_flags=( "--swiglu" @@ -230,63 +192,89 @@ setup_run_cmd() { "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" ) fi + # min_lr=$(python3 -c 'print(f"{2 / (10 ** 5):.8f}")') + # "--min-lr ${LR:-${min_lr}}" # 2e-5 + lr_flags=( + "--lr ${LR:-0.0002}" + "--lr-decay-style ${LR_DECAY_STYLE:-cosine}" + "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 + "--lr-warmup-fraction ${LR_WARMUP_FRAC:-0.05}" + ) + if [[ -n "${LR_DECAY_ITERS:-}" ]]; then + lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") + fi - TENSORBARD_ARGS=() - if [[ -z "${USE_TENSORBARD:-}" ]]; then + tb_flags=() + if [[ -z "${NO_TENSORBOARD:-}" ]]; then TBDIR="${CKPT_DIR}/tensorboard" mkdir -p "${TBDIR}" - # --log-timers-to-tensorboard \ - # --log-optimizer-states-to-tensorboard" - # --tensorboard-dir ${TBDIR} \ - TENSORBARD_ARGS+=( + tb_flags+=( "--log-timers-to-tensorboard" "--log-optimizer-states-to-tensorboard" "--tensorboard-dir ${TBDIR}" ) fi dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" - export ADAM_BETA1="${ADAM_BETA1:-0.9}" - export ADAM_BETA2="${ADAM_BETA2:-0.95}" - export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} - export run_cmd=( - "${LAUNCHER}" + + train_args=() + if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then + train_args+=("--use-checkpoint-opt_param-scheduler") + fi + # "--init-method-std ${INIT_METHOD_STD:-0.0006}" + # "--weight-decay ${WEIGHT_DECAY:-0.1}" + # --accumulate-allreduce-grads-in-fp32" + train_args+=( + "${lr_flags[@]}" + "${custom_args[@]}" + "${llama_flags[@]}" + "${DATA_FLAGS}" + "${FLASH_ARG}" + "${TIMING_STR}" + "${TOKENIZER_FLAGS}" + "${tb_flags[@]}" + "${ds_args[@]}" + "${gpt_args[@]}" "--${DTYPE}" - "${DEFAULTS[@]}" + "--shuffle-sample" + "--no-bias-gelu-fusion" + "--no-bias-dropout-fusion" + "--no-masked-softmax-fusion" + "--no-gradient-accumulation-fusion" "--optimizer ${OPT}" - "--save ${CKPT_DIR}" - "--load ${CKPT_DIR}" + "--tensor-model-parallel-size ${TP}" + "--pipeline-model-parallel-size ${PP}" + "--max-position-embeddings ${SEQ}" + "--micro-batch-size ${MICRO_BATCH}" + "--ds-sequence-parallel-size ${SP}" + "--global-batch-size ${GLOBAL_BATCH}" + "--split ${TRAIN_SPLIT:-950},${VAL_SPLIT:-50},${TEST_SPLIT:-0}" + "--timing-log-level ${TIMING_LOG_LEVEL:-1}" + "--eval-interval ${EVAL_INTERVAL:-50}" + "--eval-iters ${EVAL_ITERS:-40}" + "--save-interval ${SAVE_INTERVAL:-50}" + "--log-interval ${LOG_INTERVAL:-1}" + "--save ${SAVE:-${CKPT_DIR}}" + "--load ${LOAD:-${CKPT_DIR}}" "--seq-length ${SEQ}" "--num-layers ${NLAYERS}" "--hidden-size ${HIDDEN}" "--train-iters ${TRAIN_ITERS}" - "--eval-iters ${EVAL_ITERS}" "--distributed-backend ${BE}" "--adam-beta1 ${ADAM_BETA1:-0.9}" "--adam-beta2 ${ADAM_BETA2:-0.95}" "--adam-eps ${ADAM_EPS:-0.00001}" "--clip-grad ${CLIP_GRAD:-1.0}" - "--weight-decay ${WEIGHT_DECAY:-0.1}" "--num-attention-heads ${HEADS}" - "--save-interval ${SAVE_INTERVAL}" - "--eval-interval ${EVAL_INTERVAL}" - "--max-position-embeddings ${SEQ}" - "--micro-batch-size ${MICRO_BATCH}" - "--tensor-model-parallel-size ${TP}" - "--global-batch-size ${GLOBAL_BATCH}" - "--pipeline-model-parallel-size ${PP}" "--data-cache-path ${data_cache_path}" "--data-file-list ${DATA_FILE_LIST:-${dfl_fallback}}" - "${TENSORBARD_ARGS[@]}" - "${DATA_FLAGS}" - "${LR_ARGS}" - "${llama_flags[@]}" - "${FLASH_ARG}" - "${TIMING_STR}" - "${TOKENIZER_FLAGS}" - "${ds_args[@]}" - "${gpt_args[@]}" - "${custom_args[@]}" ) + cache_dir="${PBS_O_WORKDIR}/.cache/" + mkdir -p "${cache_dir}" + targs_cache="${cache_dir}/train_args.txt" + for arg in "${train_args[@]}"; do echo "${arg}" >> "${targs_cache}" ; done + export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}"|sort)") + printf "Training Arguments: %s\n" "${TRAIN_ARGS[@]}" + export run_cmd=("${LAUNCHER}" "${train_args[@]}") } save_dotenv() { @@ -430,17 +418,20 @@ setupLauncher() { printf " %s" "$(printMagenta "${LAUNCHER}")" } -set_lr_args() { - LR_ARGS="--lr ${LR} --lr-decay-style cosine" - if [[ -n "${LR_DECAY_ITERS:-}" ]]; then - LR_ARGS="${LR_ARGS} --lr-decay-iters ${LR_DECAY_ITERS}" - fi - if [[ -n "${LR_WARMUP_FRAC}" ]]; then - LR_ARGS="${LR_ARGS} --lr-warmup-fraction ${LR_WARMUP_FRAC}" - fi - echo "LR_ARGS: ${LR_ARGS}" - export LR_ARGS="${LR_ARGS}" -} +# set_lr_args() { +# export LR=${LR:-0.0002} # LEARNING_RATE +# export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP +# export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS +# LR_ARGS="--lr ${LR} --lr-decay-style cosine" +# if [[ -n "${LR_DECAY_ITERS:-}" ]]; then +# LR_ARGS="${LR_ARGS} --lr-decay-iters ${LR_DECAY_ITERS}" +# fi +# if [[ -n "${LR_WARMUP_FRAC}" ]]; then +# LR_ARGS="${LR_ARGS} --lr-warmup-fraction ${LR_WARMUP_FRAC}" +# fi +# echo "LR_ARGS: ${LR_ARGS}" +# export LR_ARGS="${LR_ARGS}" +# } ######################################################################### # `get_batch_size_on_polaris`: Identify MICRO_BATCH to use on Polaris. @@ -495,12 +486,14 @@ _get_num_hosts_from_hostfile() { # # [2 tiles] x [6 xpus / tile] = 12 xpus # -# | nnhosts | nhosts | GAS | -# |:-------------:|:---------:|:-----:| -# | 64 <= n < inf | [64, inf) | 1 | -# | 32 <= n < 64 | [32, 64) | 2 | -# | 16 <= n < 32 | [16, 32) | 4 | -# | 0 <= n < 16 | [0, 16) | 8 | +# | nnhosts | nhosts | GAS | +# |:---------------:|:----------:|:-----:| +# | 256 <= n < inf | [256, inf) | 1 | +# | 128 <= n < 256 | [128, 256) | 2 | +# | 32 <= n < 128 | [32, 128) | 4 | +# | 16 <= n < 32 | [16, 32) | 8 | +# | 0 <= n < 16 | [0, 16) | 16 | +# ########################################### get_grad_acc_steps_on_aurora() { if [[ "$#" == 0 ]]; then @@ -508,18 +501,21 @@ get_grad_acc_steps_on_aurora() { elif [[ "$#" == 1 ]]; then hf="$1" else + echo "Usage: get_grad_acc_steps_on_aurora" echo "Expected exactly 0 or 1 arguments, received: $#" exit 1 fi nhosts=$(wc -l <"${hf}") - if [[ 64 -le "${nhosts}" ]]; then + if [[ "${nhosts}" -gt 256 ]]; then gas=1 - elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 64 ]]; then + elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then gas=2 - elif [[ 16 -le "${nhosts}" && "${nhosts}" -lt 32 ]]; then + elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 128 ]]; then gas=4 - else + elif [[ 16 -le "${nhosts}" && "${nhosts}" -lt 32 ]]; then gas=8 + else + gas=16 fi echo "${gas}" } @@ -580,7 +576,7 @@ setParams() { export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - MICRO_BATCH=${MICRO_BATCH:-4} # MICRO_BATCH = 4 + MICRO_BATCH=${MICRO_BATCH:-1} # MICRO_BATCH = 4 #### [sam: 08/17/2024] ########################################## # Use best set of CCL env vars from Gordon Bell runs on Aurora set_ccl_vars_on_aurora @@ -604,9 +600,7 @@ setParams() { echo "Using flash-attn !!" FLASH_ARG="--use-flash-attn-builder" fi - ###################################################################### - # +--------[Polaris]-----------------------------------+ - # elif [[ $(hostname) == x3* ]]; then + # [Polaris] elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" TP=${TP:-1} # TP = 2 @@ -625,30 +619,25 @@ setParams() { fi echo "Setting up AWS NCCL OFI Plugin on Polaris..." source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit - # +--------[Perlmutter]---------------------------------+ - # elif [[ $(hostname) == login* || $(hostname) == nid* ]]; then + # [Perlmutter] elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then TP="${TP:-2}" export NCCL="${NCCL:-nccl}" export BE="${NCCL}" export DTYPE="${DTYPE:-bf16}" - MICRO_BATCH="${MICRO_BATCH:-8}" + MICRO_BATCH="${MICRO_BATCH:-1}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi fi - # +----------------------------------------------------------------------+ export TP="${TP}" export PP="${PP:-1}" export SP="${SP:-1}" export FLASH_ARG="${FLASH_ARG}" export DTYPE="${DTYPE:-bf16}" export OPT="${OPT:-adamw}" - # export ADAM_BETA1="${ADAM_BETA1:-0.9}" - # export ADAM_BETA2="${ADAM_BETA2:-0.95}" - # export ADAM_EPS="${ADAM_EPS:-0.00001}" # 1 * 10^{-5} export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" NHOSTS=$(wc -l <"${HOSTFILE}") @@ -667,18 +656,19 @@ setParams() { # +---[Run Settings]------------------------------------------------------+ export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - export MICRO_BATCH=${MICRO_BATCH:-8} # MICRO BATCH SIZE + export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS - export EVAL_ITERS="${EVAL_ITERS:-10}" # NUMBER OF EVAL ITERS TO RUN - export EVAL_INTERVAL="${EVAL_INTERVAL:-50000}" # HOW FREQUENTLY TO RUN EVAL - export SAVE_INTERVAL=${SAVE_INTERVAL:-50} # HOW FREQUENTLY TO SAVE CKPTS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-1} # USE ACTIVATION CHECKPOINTING ? + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - # export TRAIN_ITER=${TRAIN_ITER:-317892} # NUMBER OF TRAIN ITERS - if [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then + if [[ -n "${TRAIN_TOKENS:-}" ]]; then + export TRAIN_TOKENS="${TRAIN_TOKENS}" + export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" @@ -694,22 +684,20 @@ setParams() { # # For this reason, we only use the default LLAMA_ARGS when SP=0. ########################################################################## + # # -----[Learning Rate Settings]-------------------------------------------- + # export LR=${LR:-0.0002} # LEARNING_RATE + # export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS + # set_lr_args # -----[Learning Rate Settings]-------------------------------------------- - export LR=${LR:-0.0003} # LEARNING_RATE - export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP - export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS - set_lr_args - # -----[Learning Rate Settings]-------------------------------------------- + # # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then - if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then - TIMING_STR="\ - --timing-log-level ${TIMING_LOG_LEVEL}" - # --log-timers-to-tensorboard \ - # --log-optimizer-states-to-tensorboard \ - # " - else - TIMING_STR="" - fi + # TIMING_STR="\ + # --timing-log-level ${TIMING_LOG_LEVEL}" + # # " + # else + # TIMING_STR="" + # fi } ############################################## @@ -741,7 +729,7 @@ set_args() { # if [[ "${ZERO_STAGE}" == 3 ]]; then # ds_args="--use-mics ${ds_args}" # fi - if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then + if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args+=("--deepspeed-activation-checkpointing") # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" @@ -904,6 +892,7 @@ buildDSconfig() { echo "DS_CONFIG: ${DS_CONFIG}" printf "ZS: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" generateDSconfig "${DS_CONFIG}" + cat "${DS_CONFIG}" | jq . } ############################################################################### @@ -1176,27 +1165,19 @@ generateDSconfig() { \"train_batch_size\": $GLOBAL_BATCH, \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, + \"gradient_clipping\": 1.0, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, \"gradient_clipping\": 1.0, \"wall_clock_breakdown\": false," - if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then - activation_checkpointing="\ - \"activation_checkpointing\": { - \"partition_activations\": true, - \"contiguous_memory_optimization\": true - }," - fi - flops_profiler="\ - \"flops_profiler\": { - \"enabled\": true, - \"profile_step\": 2, - \"module_depth\": -1, - \"top_modules\": 1, - \"detailed\": true, - \"output_file\": null - }" + # if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then + # activation_checkpointing="\ + # \"activation_checkpointing\": { + # \"partition_activations\": true, + # \"contiguous_memory_optimization\": true + # }," + # fi if [[ $DTYPE == "bf16" ]]; then dtype="\ \"communication_data_type\": \"bf16\", @@ -1228,7 +1209,7 @@ generateDSconfig() { else dtype="\"communication_data_type\": \"fp32\"," fi - if [[ "${OPT:-adamw}" == "ds.adamw" ]]; then + if [[ "${OPT:-}" == "ds.adamw" ]]; then optimizer="\ \"optimizer\": { \"type\": \"AdamW\", @@ -1238,6 +1219,24 @@ generateDSconfig() { \"beta2\": 0.95, \"eps\": 1e-5, \"weight_decay\": 1e-1 + }, + }," + elif [[ "${OPT:-}" == "ds.onebitlamb" ]]; then + optimizer="\ + \"optimizer\": { + \"type\": \"OneBitLamb\", + \"params\": { + \"lr\": 11e-3, + \"max_coeff\": 0.3, + \"min_coeff\": 0.01, + \"freeze_step\": 1000, + \"cuda_aware\": false, + \"comm_backend_name\": \"${BE}\", + \"coeff_beta\": 0.9, + \"factor_max\": 4.0, + \"factor_min\": 0.5, + \"factor_threshold\": 0.1 + } }," else optimizer="" @@ -1267,7 +1266,7 @@ generateDSconfig() { }," # elif [[ $ZERO_STAGE == 2 ]]; then elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then - if [[ -z "${CPU_OPTIMIZER:-}" ]]; then + if [[ -n "${CPU_OPTIMIZER:-}" ]]; then echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" zero="\ \"zero_optimization\": { @@ -1304,6 +1303,15 @@ generateDSconfig() { else echo 'Please add the correct config set!!!' fi + flops_profiler="\ + \"flops_profiler\": { + \"enabled\": true, + \"profile_step\": 2, + \"module_depth\": -1, + \"top_modules\": 1, + \"detailed\": true, + \"output_file\": null + }" cat <"$1" { $common @@ -1388,6 +1396,7 @@ printWhite() { reset_env() { custom_vars=( NO_FLASH_ATTN + USE_FLASH_ATTN TP PP SP @@ -1415,7 +1424,6 @@ reset_env() { TRAIN_TOKENS TRAIN_ITERS MODEL_TYPE - LLAMA_ARGS LR LR_WARMUP_FRAC LR_DECAY_ITERS @@ -1447,10 +1455,24 @@ reset_env() { data_cache_path DEFAULTS ) + # LLAMA_ARGS printf "Unsetting custom vars: %s\n" "${custom_vars[*]}" unset "${custom_vars[@]}" } +convert_ckpt_to_universal() { + if [[ "$#" -ne 1 ]]; then + echo "Usage: convert_ckpt_to_universal ckpt_dir" + echo "Expected one argument (ckpt_dir), received: $#" + exit 1 + fi + ckptdir=$1 + gs=$(cat "${ckptdir}/latest_checkpointed_iteration.txt") + src="${ckptdir}/global_step${gs}" + dst="${ckptdir}/global_step${gs}_universal" + convert_script="${PBS_O_WORKDIR}/deps/DeepSpeed/checkpoint/ds_to_universal.py" + python3 "${convert_script}" --input_folder "${src}" --output_folder "${dst}" +} ########################### # call helpers_main() From 6acc370a41440098031227e881bd9f1f23aa369e Mon Sep 17 00:00:00 2001 From: ranzhejiang Date: Fri, 18 Oct 2024 18:31:05 +0800 Subject: [PATCH 353/590] fix moe tflops (#445) --- megatron/utils.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/megatron/utils.py b/megatron/utils.py index 15160b16447..68d06f0dae4 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -281,6 +281,7 @@ def throughput_calculator(model, args, iteration_time, total_iterations): num_layers = args.num_layers vocab_size = args.padded_vocab_size gqa = args.num_attention_heads // args.num_key_value_heads + num_experts_routed_to = args.topk ffn_multiplier = 3 if args.swiglu else 2 macs_per_flops = 2 @@ -294,7 +295,7 @@ def throughput_calculator(model, args, iteration_time, total_iterations): pre_and_post_mha_gemm_macs = batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len mha_bgemm_macs = batch_size * num_layers * 2 * head_dim * num_attention_heads * (seq_len**2) - ffn_gemm_macs = batch_size * num_layers * ffn_multiplier * ffn_hidden_size * hidden_size * seq_len + ffn_gemm_macs = batch_size * num_layers * ffn_multiplier * ffn_hidden_size * hidden_size * seq_len * num_experts_routed_to logit_lmhead_gemm_macs = batch_size * vocab_size * hidden_size * seq_len fwd_macs = pre_and_post_mha_gemm_macs + mha_bgemm_macs + ffn_gemm_macs + logit_lmhead_gemm_macs From 9e015cc164512e463559a2a6846679676aa3b17f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 18 Oct 2024 10:00:26 -0500 Subject: [PATCH 354/590] Remove duplicate `gradient_accumulation_steps` in DS config --- ALCF/helpers.sh | 1 - 1 file changed, 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index bc7753aa718..ade65a8d09b 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -1165,7 +1165,6 @@ generateDSconfig() { \"train_batch_size\": $GLOBAL_BATCH, \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"steps_per_print\": 1, - \"gradient_clipping\": 1.0, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, From 58dc2d7c86d8c5b562871450ab7e2377d9db022f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 21 Oct 2024 09:04:19 -0500 Subject: [PATCH 355/590] Update default EVAL args --- ALCF/helpers.sh | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index ade65a8d09b..cedac99f96d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -247,10 +247,10 @@ setup_run_cmd() { "--micro-batch-size ${MICRO_BATCH}" "--ds-sequence-parallel-size ${SP}" "--global-batch-size ${GLOBAL_BATCH}" - "--split ${TRAIN_SPLIT:-950},${VAL_SPLIT:-50},${TEST_SPLIT:-0}" + "--split ${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" "--timing-log-level ${TIMING_LOG_LEVEL:-1}" - "--eval-interval ${EVAL_INTERVAL:-50}" - "--eval-iters ${EVAL_ITERS:-40}" + "--eval-interval ${EVAL_INTERVAL:-100}" + "--eval-iters ${EVAL_ITERS:-20}" "--save-interval ${SAVE_INTERVAL:-50}" "--log-interval ${LOG_INTERVAL:-1}" "--save ${SAVE:-${CKPT_DIR}}" From 277d308db336a9c5000c098f265d8d193f4b8bae Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 21 Oct 2024 09:04:37 -0500 Subject: [PATCH 356/590] Catch eval metrics in `megatron/training.py` --- megatron/training.py | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 8ffac6cb9c2..d39f21c1289 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -26,6 +26,7 @@ import torch.distributed as tdist from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP +import wandb from megatron import ( get_args, get_current_global_batch_size, @@ -316,7 +317,7 @@ def pretrain( config = core_transformer_config_from_args(args) if args.do_valid: prefix = f"iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from validation set" - evaluate_and_print_results( + _ = evaluate_and_print_results( prefix, forward_step_func, valid_data_iterator, @@ -329,7 +330,7 @@ def pretrain( ) if args.do_test: prefix = f"iteration {iteration} on {args.eval_iters * args.global_batch_size}-sample draw from test set" - evaluate_and_print_results( + _ = evaluate_and_print_results( prefix, forward_step_func, test_data_iterator, @@ -924,7 +925,6 @@ def train_step( # Empty unused memory. if args.empty_unused_memory_level >= 2 and accelerator is not None: accelerator.empty_cache() - # XXX: [saforem2]: ---------------------------------------------------- # Is `num_zeros_in_grad` worth calculating (/ implementing) ?? # the `Megatron`-specific implementation is at: @@ -1406,6 +1406,16 @@ def evaluate_and_print_results( config, verbose, ) + key = "test" if test else "val" + if wandb is not None and wandb.run is not None: + wandb.log({ + f"{key}/iteration": iteration, + **{f"{key}/{k}": v for k, v in total_loss_dict.items()}, + **{ + f"{key}/ppl_{k}": math.exp(min(20, v.item())) + for k, v in total_loss_dict.items() + }, + }) string = " validation loss at {} | ".format(prefix) for key in total_loss_dict: string += f"{key} value={total_loss_dict[key].item():.6f}" @@ -1451,6 +1461,7 @@ def evaluate_and_print_results( log.info("-" * length) log.info(string) log.info("-" * length) + return total_loss_dict def cyclic_iter(iter): From af4cba12b810e420b8c0039e384d0b06c8842861 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 21 Oct 2024 09:05:06 -0500 Subject: [PATCH 357/590] Save git branch to env in `train_aGPT_7B.sh` --- train_aGPT_7B.sh | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/train_aGPT_7B.sh b/train_aGPT_7B.sh index 286740fc89f..1350ea0f2a4 100644 --- a/train_aGPT_7B.sh +++ b/train_aGPT_7B.sh @@ -1,4 +1,7 @@ #!/bin/bash --login +#PBS -q lustre_scaling +#PBS -A Aurora_Deployment +#PBS -j oe ##################################### # AuroraGPT-7B @@ -10,6 +13,8 @@ # 1. Navigate into `$PBS_O_WORKDIR` cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE +GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + # 2. source `ALCF/helpers.sh` source "${HERE}/ALCF/helpers.sh" || exit From 8a8472c7bd83a3c9bc13f27b54c099ab1cde98b6 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Mon, 21 Oct 2024 19:35:57 +0000 Subject: [PATCH 358/590] fixed print out bug --- megatron/data/gpt_dataset.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 38df5562675..9ff27032772 100755 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -154,8 +154,8 @@ def _build_indices_blended(): dataset_index, dataset_sample_index, weights, self.num_datasets, self.num_samples, torch.distributed.get_rank() == 0) - log.debug('> elapsed time for building blendable dataset indices for corpus {self.dataset_builders[0].corpus}: ' - '{:.2f} (sec)'.format(time.time() - start_time)) + log.debug(f"> elapsed time for building blendable dataset indices for corpus {self.dataset_builders[0].corpus}: " + "{:.2f} (sec)".format(time.time() - start_time)) return dataset_index, dataset_sample_index From 6cb727dde1ef92eedfafc24882cdf51a73a7203b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 21 Oct 2024 15:43:58 -0500 Subject: [PATCH 359/590] Fix `args.shuffle` in `megatron/data/gpt_dataset.py` --- megatron/data/gpt_dataset.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index 8872d709c7b..d09f08d63a4 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -189,7 +189,7 @@ def _build_indices_concat(): self.desc += dataset_builders[i].prefix + "," log.info( - f"[BuildConcatDataset] Caught {shuffle=} across" + f"[BuildConcatDataset] Caught {args.shuffle_sample_in_corpus=} across" f" {self.num_samples} samples" ) self.desc += ( From 5d10179125c884516a70e76d0a1b36f59b4e1e4c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 23 Oct 2024 22:36:50 -0500 Subject: [PATCH 360/590] Update `--{shuffle,blend}-sample-in-corpus` arg in `ALCF/helpers.sh` --- ALCF/helpers.sh | 98 +++++++++++++++++++++++++------------------------ 1 file changed, 50 insertions(+), 48 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index cedac99f96d..603bc88fdb2 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -141,7 +141,6 @@ setup() { setup_run_cmd "$@" || exit } - ##################################################### # setup_run_cmd # @@ -180,24 +179,24 @@ setup_run_cmd() { ################################################################## if [[ -z "${NO_LLAMA:-}" ]]; then llama_flags=( - "--swiglu" + "--swiglu" "--hidden-dropout 0" "--attention-dropout 0" - "--normalization rmsnorm" - "--disable-bias-linear" + "--normalization rmsnorm" + "--disable-bias-linear" "--no-query-key-layer-scaling" - "--use-rotary-position-embeddings" - "--untie-embeddings-and-output-weights" + "--use-rotary-position-embeddings" + "--untie-embeddings-and-output-weights" "--num-key-value-heads ${NUM_KV_HEAD}" "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" ) fi # min_lr=$(python3 -c 'print(f"{2 / (10 ** 5):.8f}")') # "--min-lr ${LR:-${min_lr}}" # 2e-5 + # "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 lr_flags=( "--lr ${LR:-0.0002}" "--lr-decay-style ${LR_DECAY_STYLE:-cosine}" - "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 "--lr-warmup-fraction ${LR_WARMUP_FRAC:-0.05}" ) if [[ -n "${LR_DECAY_ITERS:-}" ]]; then @@ -221,8 +220,7 @@ setup_run_cmd() { train_args+=("--use-checkpoint-opt_param-scheduler") fi # "--init-method-std ${INIT_METHOD_STD:-0.0006}" - # "--weight-decay ${WEIGHT_DECAY:-0.1}" - # --accumulate-allreduce-grads-in-fp32" + # "--shuffle-sample" train_args+=( "${lr_flags[@]}" "${custom_args[@]}" @@ -235,44 +233,48 @@ setup_run_cmd() { "${ds_args[@]}" "${gpt_args[@]}" "--${DTYPE}" - "--shuffle-sample" + "--shuffle-sample-in-corpus" + "--blend-sample-in-corpus" + "--accumulate-allreduce-grads-in-fp32" "--no-bias-gelu-fusion" "--no-bias-dropout-fusion" "--no-masked-softmax-fusion" "--no-gradient-accumulation-fusion" - "--optimizer ${OPT}" - "--tensor-model-parallel-size ${TP}" - "--pipeline-model-parallel-size ${PP}" - "--max-position-embeddings ${SEQ}" - "--micro-batch-size ${MICRO_BATCH}" - "--ds-sequence-parallel-size ${SP}" - "--global-batch-size ${GLOBAL_BATCH}" - "--split ${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" - "--timing-log-level ${TIMING_LOG_LEVEL:-1}" - "--eval-interval ${EVAL_INTERVAL:-100}" - "--eval-iters ${EVAL_ITERS:-20}" - "--save-interval ${SAVE_INTERVAL:-50}" - "--log-interval ${LOG_INTERVAL:-1}" - "--save ${SAVE:-${CKPT_DIR}}" - "--load ${LOAD:-${CKPT_DIR}}" - "--seq-length ${SEQ}" - "--num-layers ${NLAYERS}" - "--hidden-size ${HIDDEN}" - "--train-iters ${TRAIN_ITERS}" - "--distributed-backend ${BE}" - "--adam-beta1 ${ADAM_BETA1:-0.9}" - "--adam-beta2 ${ADAM_BETA2:-0.95}" - "--adam-eps ${ADAM_EPS:-0.00001}" - "--clip-grad ${CLIP_GRAD:-1.0}" - "--num-attention-heads ${HEADS}" - "--data-cache-path ${data_cache_path}" - "--data-file-list ${DATA_FILE_LIST:-${dfl_fallback}}" + "--optimizer=${OPT}" + "--tensor-model-parallel-size=${TP}" + "--pipeline-model-parallel-size=${PP}" + "--max-position-embeddings=${SEQ}" + "--micro-batch-size=${MICRO_BATCH}" + "--ds-sequence-parallel-size=${SP}" + "--global-batch-size=${GLOBAL_BATCH}" + "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" + "--timing-log-level=${TIMING_LOG_LEVEL:-1}" + "--eval-interval=${EVAL_INTERVAL:-100}" + "--eval-iters=${EVAL_ITERS:-20}" + "--save-interval=${SAVE_INTERVAL:-50}" + "--log-interval=${LOG_INTERVAL:-1}" + "--save=${SAVE:-${CKPT_DIR}}" + "--load=${LOAD:-${CKPT_DIR}}" + "--seq-length=${SEQ}" + "--num-layers=${NLAYERS}" + "--hidden-size=${HIDDEN}" + "--train-iters=${TRAIN_ITERS}" + "--distributed-backend=${BE}" + "--weight-decay=${WEIGHT_DECAY:-0.1}" + "--adam-beta1=${ADAM_BETA1:-0.9}" + "--adam-beta2=${ADAM_BETA2:-0.95}" + "--adam-eps=${ADAM_EPS:-0.00001}" + "--clip-grad=${CLIP_GRAD:-1.0}" + "--num-attention-heads=${HEADS}" + "--data-cache-path=${data_cache_path}" + "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" ) + # "--adam-eps ${ADAM_EPS:-0.00001}" cache_dir="${PBS_O_WORKDIR}/.cache/" mkdir -p "${cache_dir}" targs_cache="${cache_dir}/train_args.txt" - for arg in "${train_args[@]}"; do echo "${arg}" >> "${targs_cache}" ; done - export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}"|sort)") + for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done + export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") printf "Training Arguments: %s\n" "${TRAIN_ARGS[@]}" export run_cmd=("${LAUNCHER}" "${train_args[@]}") } @@ -506,7 +508,7 @@ get_grad_acc_steps_on_aurora() { exit 1 fi nhosts=$(wc -l <"${hf}") - if [[ "${nhosts}" -gt 256 ]]; then + if [[ "${nhosts}" -ge 256 ]]; then gas=1 elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then gas=2 @@ -567,7 +569,7 @@ setParams() { mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 - export SAVE_INTERVAL="${SAVE_INTERVAL:-20}" + export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" export CCL=${CCL:-ccl} # CCL export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 @@ -845,7 +847,7 @@ get_output_prefix() { pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" if [[ -n "${TOKENIZER_TYPE:-}" ]]; then - _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa + _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa pre="${pre}_tok${_tok}" fi if [[ -n "${LR_DECAY_ITERS}" ]]; then @@ -1111,7 +1113,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] } generateDSconfig_new() { - cat < "${CONFIG_JSON}" + cat <"${CONFIG_JSON}" { "train_batch_size" : $GLOBAL_BATCH, "train_micro_batch_size_per_gpu": $MICRO_BATCH, @@ -1164,11 +1166,11 @@ generateDSconfig() { common="\ \"train_batch_size\": $GLOBAL_BATCH, \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, + \"gradient_clipping\": 1.0, \"steps_per_print\": 1, \"gradient_accumulation_steps\": $GRAD_ACC_STEPS, \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, - \"gradient_clipping\": 1.0, \"wall_clock_breakdown\": false," # if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then # activation_checkpointing="\ @@ -1178,8 +1180,8 @@ generateDSconfig() { # }," # fi if [[ $DTYPE == "bf16" ]]; then + # \"communication_data_type\": \"bf16\", dtype="\ - \"communication_data_type\": \"bf16\", \"fp16\": { \"enabled\": false, \"loss_scale\": 0, @@ -1214,9 +1216,9 @@ generateDSconfig() { \"type\": \"AdamW\", \"params\": { \"lr\": ${LR}, - \"beta1\": 0.9, - \"beta2\": 0.95, - \"eps\": 1e-5, + \"beta1\": ${ADAM_BETA1}, + \"beta2\": ${ADAM_BETA2}, + \"eps\": ${ADAM_EPS}, \"weight_decay\": 1e-1 }, }," From 160d6a69717689226798ae8e378ea4c7b2866ad4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 30 Oct 2024 21:25:53 -0500 Subject: [PATCH 361/590] fix: `GRAD_ACC_STEPS` when `NHOSTS == 256` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 603bc88fdb2..fac2f4d8ecb 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -508,7 +508,7 @@ get_grad_acc_steps_on_aurora() { exit 1 fi nhosts=$(wc -l <"${hf}") - if [[ "${nhosts}" -ge 256 ]]; then + if [[ "${nhosts}" -gt 256 ]]; then gas=1 elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then gas=2 From ce7d55356ff3bd7f02f864845dda13ecb8c02ff0 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 6 Nov 2024 21:57:02 -0600 Subject: [PATCH 362/590] =?UTF-8?q?=F0=9F=9A=A7=20`ALCF/ds=5Fto=5Funiversa?= =?UTF-8?q?l.py`?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ALCF/ds_to_universal.py | 693 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 693 insertions(+) create mode 100755 ALCF/ds_to_universal.py diff --git a/ALCF/ds_to_universal.py b/ALCF/ds_to_universal.py new file mode 100755 index 00000000000..da7fcb708e9 --- /dev/null +++ b/ALCF/ds_to_universal.py @@ -0,0 +1,693 @@ +#!/usr/bin/env python + +# Copyright (c) Microsoft Corporation. +# SPDX-License-Identifier: Apache-2.0 + +# DeepSpeed Team + +from functools import partial +from itertools import chain +import argparse +import glob +import itertools +import math +from concurrent.futures import ProcessPoolExecutor +import os +import re +import shutil +import torch +import tqdm +# from pprint import pprint + +from deepspeed.checkpoint import DeepSpeedCheckpoint +from deepspeed.checkpoint import ( + OPTIMIZER_STATE_DICT, + ZERO_STAGE, + BASE_OPTIMIZER_STATE, + SINGLE_PARTITION_OF_FP32_GROUPS, + PARAM_GROUPS, + PARAM_SLICE_MAPPINGS, + PARAM_SHAPES, + PARAM, + CAT_DIM, + PARAM_N_SUB_PARAMS, + SUB_PARAM_SHAPE, + VOCAB_TENSOR, + UNIVERSAL_CHECKPOINT_INFO, + UNIVERSAL_CHECKPOINT_VERSION_KEY, + UNIVERSAL_CHECKPOINT_VERSION_VALUE, + VOCABULARY_PARAMETER_PATTERNS, + PIPELINE_REPLICATED_PARAMETER_PATTERNS, + TP_REPLICATED_PARAMETER_PATTERNS, + PARAMETER_TO_AVERAGE_PATTERNS, + PARAMETER_WITH_ROW_PARALLELISM_PATTERNS, + PARAMETER_WITH_2_SUB_PARAMS_CAT_DIM_0, + PARAMETER_WITH_SUB_PARAMS, + SubparamShape, +) + + +def parse_arguments(): + parser = argparse.ArgumentParser() + parser.add_argument( + "--input_folder", + type=str, + required=True, + help="Input DeepSpeed Checkpoint folder", + ) + parser.add_argument( + "--output_folder", + type=str, + required=False, + default=None, + help="Output DeepSpeed checkpoint folder", + ) + parser.add_argument( + "--num_extract_workers", + default=4, + type=int, + help="How many parallel processes to extract zero shards", + ) + parser.add_argument( + "--num_merge_workers", + default=2, + type=int, + help="How many parallel processes to merge tp slices (more memory intensive, use much fewer than --num_extract_workers))", + ) + parser.add_argument( + "--keep_temp_folder", + action="store_true", + help="Preserve temporary folder of intermediate checkpoint slice files. Useful for debugging.", + ) + parser.add_argument( + "--no_strict", + dest="strict", + action="store_false", + help="Do not perform validity checks on converted checkpoint.", + ) + parser.add_argument( + "--inject_missing_state", + action="store_true", + help="Inject missing checkpoint state into the checkpoint if it is absent.", + ) + args = parser.parse_args() + print(f"args = {args}") + return args + + +def atoi(text): + return int(text) if text.isdigit() else text + + +def natural_keys(text): + """ + alist.sort(key=natural_keys) sorts in human order + http://nedbatchelder.com/blog/200712/human_sorting.html + (See Toothy's implementation in the comments) + """ + return [atoi(c) for c in re.split(r"(\d+)", text)] + + +def _create_checkpoint_paths(base_folder, iteration, tp_degree, pp_degree): + path_list = [] + iter_folder = f"iter_{iteration:07d}" + for i in range(0, tp_degree): + path_list.append([]) + for j in range(0, pp_degree): + rank_folder = ( + f"mp_rank_{i:02d}" if pp_degree == 1 else f"mp_rank_{i:02d}_{j:03d}" + ) + ckpt_path = os.path.join(rank_folder, "model_optim_rng.pt") + path_list[i].append(os.path.join(base_folder, iter_folder, ckpt_path)) + + return path_list + + +def _save_checkpoint(file_path, chkpt_sd): + dir, _ = os.path.split(file_path) + os.makedirs(dir, exist_ok=True) + torch.save(chkpt_sd, file_path) + + +def extract_zero_shards(dir, ds_checkpoint, indices_3D): + pp_index, tp_index, dp_index = indices_3D + sd = ds_checkpoint.get_zero_checkpoint_state( + pp_index=pp_index, + tp_index=tp_index, + dp_index=dp_index, + strip_tensor_paddings=False, + ) + + # pprint(f"Processing {dp_index=} {pp_index=}, {tp_index=}") + + optim_sd = sd[OPTIMIZER_STATE_DICT] + param_slice_mappings = optim_sd[PARAM_SLICE_MAPPINGS] + universal_checkpoint_info = ds_checkpoint.get_checkpoint_info( + UNIVERSAL_CHECKPOINT_INFO + ) + pipeline_replicated_params = universal_checkpoint_info.get( + PIPELINE_REPLICATED_PARAMETER_PATTERNS, [] + ) + # print(f'{pipeline_replicated_params=}') + + # dict + state_groups = optim_sd[BASE_OPTIMIZER_STATE]["state"] + # list + fp32_groups = optim_sd[SINGLE_PARTITION_OF_FP32_GROUPS] + param_groups_cnt = len(state_groups) + + for param_group_id in range(param_groups_cnt): + flat_state = dict( + exp_avg=state_groups[param_group_id]["exp_avg"], + exp_avg_sq=state_groups[param_group_id]["exp_avg_sq"], + fp32=fp32_groups[param_group_id], + ) + + if "step" in state_groups[param_group_id]: + flat_state["step"] = state_groups[param_group_id]["step"] + + for name, fragment_mapping in param_slice_mappings[param_group_id].items(): + if pp_index > 0 and any( + re.match(pattern, name) for pattern in pipeline_replicated_params + ): + # Skip tied weights that are replicated in first and last pp stages + continue + + # pprint(f"dpt{dp_index}{pp_index}{tp_index} {param_group_id} {name} => {fragment_mapping.start}:{fragment_mapping.numel}") + for state_key in flat_state.keys(): + dump_param_fragment( + dir, + tp_index, + dp_index, + state_key, + flat_state[state_key], + name, + fragment_mapping.start, + fragment_mapping.numel, + ) + + +def extract_zero_shards_stage3( + optim_files, param_shapes, dp_degree, temp_dir, dp_index +): + state_dict = torch.load(optim_files[dp_index], map_location="cpu") + + flat_state = dict( + exp_avg=state_dict[OPTIMIZER_STATE_DICT]["optimizer_state_dict"]["state"][0][ + "exp_avg" + ], + exp_avg_sq=state_dict[OPTIMIZER_STATE_DICT]["optimizer_state_dict"]["state"][0][ + "exp_avg_sq" + ], + fp32=state_dict[OPTIMIZER_STATE_DICT]["fp32_flat_groups"][0], + ) + + offset = 0 + for name, shape in param_shapes.items(): + unpartitioned_numel = shape.numel() + partitioned_numel, _ = _zero_partitioned_param_info( + unpartitioned_numel, dp_degree + ) + padding_free_numel = min( + partitioned_numel, abs(unpartitioned_numel - dp_index * partitioned_numel) + ) + for state_key in flat_state.keys(): + dump_param_fragment( + temp_dir, + 0, + dp_index, + state_key, + flat_state[state_key], + name, + offset, + padding_free_numel, + ) + offset += partitioned_numel + + +cnt = 0 + + +def dp_index_to_str(dp_index): + return f"{dp_index:0>2d}" + + +def dump_param_fragment( + dir, tp_index, dp_index, state_name, state_flat_tensor, param_name, offset, numel +): + global cnt # temp hack + + param_base_path = os.path.join(dir, param_name, str(tp_index)) + os.makedirs(param_base_path, exist_ok=True) + + cnt += 1 + + path = os.path.join(param_base_path, f"{state_name}.{dp_index_to_str(dp_index)}") + + # print(f"{param_name}: {offset}: {numel} => {path}") + + # State might be a python int or a tensor + if state_name != "step" and torch.is_tensor(state_flat_tensor): + state_flat_tensor = state_flat_tensor.narrow(0, offset, numel).clone() + _save_checkpoint(path, state_flat_tensor) + + +def _merge_zero_shards(param_base_path, state, tp_degree, slice_shape=None): + slices = [] + for tp_index in range(tp_degree): + prefix_path = os.path.join(param_base_path, str(tp_index), f"{state}") + paths = glob.glob(f"{prefix_path}.*") + + if len(paths) == 0: + continue + + pattern = re.compile(f"{prefix_path}\\.([0-9]+)") + dp_indices = set() + for p in paths: + m = pattern.match(p) + if m: + dp_indices.add(int(m.group(1))) + else: + raise ValueError(f"Cannot parse dp_rank from {p}") + + paths = [ + f"{prefix_path}.{dp_index_to_str(dp_index)}" + for dp_index in sorted(list(dp_indices)) + ] + shards = [torch.load(p) for p in paths] + + if state == "step": + assert all( + v == shards[0] for v in shards + ), "All shards must have the same step value" + slice = shards[0] + else: + if slice_shape is None: + slice = torch.cat(shards, dim=0) + else: + slice = torch.cat(shards, dim=0).reshape(slice_shape) + + slices.append(slice) + return slices + + +def merge_tp_slices(ds_checkpoint, dir, slice_dir, tp_degree, name_and_shape): + name, shape = name_and_shape + slice_base_path = os.path.join(slice_dir, name) + param_base_path = os.path.join(dir, name) + + universal_checkpoint_info = ds_checkpoint.get_checkpoint_info( + UNIVERSAL_CHECKPOINT_INFO + ) + replicated_parameters = universal_checkpoint_info.get( + TP_REPLICATED_PARAMETER_PATTERNS, [] + ) + parameters_to_average = universal_checkpoint_info.get( + PARAMETER_TO_AVERAGE_PATTERNS, [] + ) + parameters_with_row_parallelism = universal_checkpoint_info.get( + PARAMETER_WITH_ROW_PARALLELISM_PATTERNS, [] + ) + vocabulary_parameters = universal_checkpoint_info.get( + VOCABULARY_PARAMETER_PATTERNS, [] + ) + parameters_with_2_sub_params_cat_dim_0 = universal_checkpoint_info.get( + PARAMETER_WITH_2_SUB_PARAMS_CAT_DIM_0, [] + ) + parameter_with_sub_params = universal_checkpoint_info.get( + PARAMETER_WITH_SUB_PARAMS, [] + ) + + unmatched_patterns = set( + replicated_parameters + + parameters_to_average + + parameters_with_row_parallelism + + vocabulary_parameters + + parameters_with_2_sub_params_cat_dim_0 + ) + unmatched_patterns.update( + chain.from_iterable( + SubparamShape(**s).patterns for s in parameter_with_sub_params + ) + ) + + def get_matched_pattern(patterns_, name_): + matched_ = [pattern_ for pattern_ in patterns_ if re.match(pattern_, name_)] + assert ( + len(matched_) <= 1 + ), f"Got more than one matching patterns={matched_} for {name_}" + if matched_: + pattern_ = matched_[0] + unmatched_patterns.discard(pattern_) + return pattern_ + return None + + def get_matched_sub_params_pattern(name_): + for subparam_shape_dict in parameter_with_sub_params: + subparam_shape = SubparamShape(**subparam_shape_dict) + for pattern_ in subparam_shape.patterns: + if re.match(pattern_, name_): + unmatched_patterns.discard(pattern_) + return subparam_shape + return None + + matched_sub_params_shape = get_matched_sub_params_pattern(name) + + step_merged = _merge_zero_shards(slice_base_path, "step", tp_degree, shape) + if step_merged: + _save_checkpoint(os.path.join(param_base_path, f"step.pt"), step_merged[0]) + + for state in ("fp32", "exp_avg", "exp_avg_sq"): + slices = _merge_zero_shards(slice_base_path, state, tp_degree, shape) + final_path = os.path.join(param_base_path, f"{state}.pt") + + # print(f"Expected shape: {shape}") + # print(f"Fragment sizes:", list(frag.shape for frag in slices)) + ckpt_dict = {} + if get_matched_pattern(replicated_parameters, name): + if len(slices) > 1: + assert all([slices[0].equal(other_slice) for other_slice in slices[1:]]) + param = slices[0] + # print(f'replicate {name} using first slice') + elif get_matched_pattern(parameters_to_average, name): + param = sum(slices) / len(slices) + # print(f'merge {name} using average') + elif get_matched_pattern(parameters_with_2_sub_params_cat_dim_0, name): + cat_dim = 0 + chunked_slices = [torch.chunk(s, 2, dim=cat_dim) for s in slices] + merged_chunks_0 = torch.cat([s[0] for s in chunked_slices], dim=cat_dim) + merged_chunks_1 = torch.cat([s[1] for s in chunked_slices], dim=cat_dim) + param = torch.cat([merged_chunks_0, merged_chunks_1], dim=cat_dim) + ckpt_dict[CAT_DIM] = cat_dim + ckpt_dict[PARAM_N_SUB_PARAMS] = 2 + elif matched_sub_params_shape: + merged_chunks = [] + partition_dim = matched_sub_params_shape.partition_dim + + sub_dim_sizes = matched_sub_params_shape.shape[partition_dim] + if not isinstance(sub_dim_sizes, tuple): + sub_dim_sizes = (sub_dim_sizes,) + + partition_shape = [ + sum(d) if isinstance(d, tuple) else d + for d in matched_sub_params_shape.shape + ] + partition_shape = [ + d // tp_degree if i == partition_dim else d + for i, d in enumerate(partition_shape) + ] + slices = [s.view(partition_shape) for s in slices] + + offset = 0 + for sub_dim_size in sub_dim_sizes: + part_sub_dim_size = sub_dim_size // tp_degree + merged_chunks.append( + torch.cat( + [ + s.narrow(partition_dim, offset, part_sub_dim_size) + for s in slices + ], + dim=partition_dim, + ) + ) + offset += part_sub_dim_size + param = torch.cat(merged_chunks, dim=partition_dim) + ckpt_dict[SUB_PARAM_SHAPE] = matched_sub_params_shape + else: + cat_dim = ( + 1 if get_matched_pattern(parameters_with_row_parallelism, name) else 0 + ) + # print(f"merge {name} with CAT DIM: {cat_dim}") + param = torch.cat(slices, dim=cat_dim) + ckpt_dict[CAT_DIM] = cat_dim + + if get_matched_pattern(vocabulary_parameters, name): + # print(f"Before {param.shape=}") + # strip padding + original_vocab_size = universal_checkpoint_info["original_vocab_size"] + param = param[:original_vocab_size, :] + ckpt_dict[VOCAB_TENSOR] = True + # print(f"After {param.shape=}") + + # print(f"Final shape: {param.shape}") + ckpt_dict[PARAM] = param + _save_checkpoint(final_path, ckpt_dict) + + return unmatched_patterns + + +def merge_zero3_slices(dp_degree, dir, slice_dir, name): + slice_base_path = os.path.join(slice_dir, name) + param_base_path = os.path.join(dir, name) + + for state in ("fp32", "exp_avg", "exp_avg_sq"): + slices = _merge_zero_shards(slice_base_path, state, 1) + final_path = os.path.join(param_base_path, f"{state}.pt") + _save_checkpoint(final_path, slices[0]) + + +def _do_parallel_work(do_work, work_chunks, num_workers): + results = [] + if num_workers > 1: + with ProcessPoolExecutor(max_workers=num_workers) as executor: + future_list = [executor.submit(do_work, work) for work in work_chunks] + for f in tqdm.tqdm(future_list): + results.append(f.result()) + else: + # No parallel pass for unit testing + # We can't create child processes in tests + for work in tqdm.tqdm(work_chunks): + results.append(do_work(work)) + return results + + +def _extract_zero_shard_files(args, ds_checkpoint, temp_dir): + _3d_range_list = list( + itertools.product( + range(ds_checkpoint.pp_degree), + range(ds_checkpoint.tp_degree), + range(ds_checkpoint.dp_degree), + ) + ) + # pprint(f'{_3d_range_list=}') + + do_work = partial(extract_zero_shards, temp_dir, ds_checkpoint) + _do_parallel_work(do_work, _3d_range_list, args.num_extract_workers) + + +def _extract_zero_shard_files_stage3( + args, optim_files, param_shapes, dp_degree, temp_dir +): + do_work = partial( + extract_zero_shards_stage3, optim_files, param_shapes, dp_degree, temp_dir + ) + _do_parallel_work(do_work, list(range(dp_degree)), args.num_extract_workers) + + +def _merge_tp_slice_files(args, ds_checkpoint, slice_shapes, temp_dir): + zero_output_folder = os.path.join(args.output_folder, "zero") + do_work = partial( + merge_tp_slices, + ds_checkpoint, + zero_output_folder, + temp_dir, + ds_checkpoint.tp_degree, + ) + unmatched_patterns_lists = _do_parallel_work( + do_work, list(slice_shapes.items()), args.num_merge_workers + ) + + # verify that all patterns were used + # if a pattern was not used by any of the workers, then it was not used at all -> assert/alert + sets = [set(lst) for lst in unmatched_patterns_lists] + unmatched_patterns = list(set.intersection(*sets)) + if args.strict: + assert ( + not unmatched_patterns + ), f"Unused patterns={unmatched_patterns} while merging tp slices" + elif unmatched_patterns: + print(f"Warning: Unused patterns={unmatched_patterns} while merging tp slices") + + +def _merge_zero3_slice_files(args, param_shapes, dp_degree, temp_dir): + zero_output_folder = os.path.join(args.output_folder, "zero") + do_work = partial(merge_zero3_slices, dp_degree, zero_output_folder, temp_dir) + _do_parallel_work(do_work, param_shapes.keys(), args.num_merge_workers) + + +def _zero_partitioned_param_info(unpartitioned_numel, world_size): + remainder = unpartitioned_numel % world_size + padding_numel = (world_size - remainder) if remainder else 0 + partitioned_numel = math.ceil(unpartitioned_numel / world_size) + return partitioned_numel, padding_numel + + +def _parse_model_states_stage3(files): + return torch.load(files[0], map_location=torch.device("cpu"))[PARAM_SHAPES] + + +def _save_optimizer_state(args, ds_checkpoint): + sharded_states = [ + BASE_OPTIMIZER_STATE, + PARAM_SLICE_MAPPINGS, + SINGLE_PARTITION_OF_FP32_GROUPS, + ] + sd = ds_checkpoint.get_zero_checkpoint_state( + pp_index=0, tp_index=0, dp_index=0, strip_tensor_paddings=False + ) + + optim_sd = sd[OPTIMIZER_STATE_DICT] + output_sd = {k: v for k, v in optim_sd.items() if k not in sharded_states} + output_sd[PARAM_GROUPS] = optim_sd[BASE_OPTIMIZER_STATE][PARAM_GROUPS] + zero_output_folder = os.path.join(args.output_folder, "zero") + output_file_path = os.path.join(zero_output_folder, f"optimizer_state.pt") + _save_checkpoint(output_file_path, output_sd) + + +def _save_optimizer_state_stage3(args, optim_files): + sd = torch.load(optim_files[0], map_location=torch.device("cpu")) + output_sd = sd[OPTIMIZER_STATE_DICT] + output_sd[PARAM_GROUPS] = output_sd[OPTIMIZER_STATE_DICT][PARAM_GROUPS] + zero_output_folder = os.path.join(args.output_folder, "zero") + output_file_path = os.path.join(zero_output_folder, f"optimizer_state.pt") + _save_checkpoint(output_file_path, output_sd) + + +def _get_optim_files(checkpoint_dir): + return _get_checkpoint_files(checkpoint_dir, "*_optim_states.pt") + + +def _get_model_state_files(checkpoint_dir): + return _get_checkpoint_files(checkpoint_dir, "*_model_states.pt") + + +def _get_checkpoint_files(checkpoint_dir, glob_pattern): + ckpt_files = sorted( + glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys + ) + + if len(ckpt_files) == 0: + raise FileNotFoundError( + f"can't find {glob_pattern} files in directory '{checkpoint_dir}'" + ) + + return ckpt_files + + +def _get_zero_stage(optim_files): + state_dict = torch.load(optim_files[0], map_location=torch.device("cpu")) + optimizer_state = state_dict[OPTIMIZER_STATE_DICT] + zero_stage = optimizer_state.get(ZERO_STAGE, 1) + return zero_stage + + +def _inject_missing_state(ds_checkpoint): + if UNIVERSAL_CHECKPOINT_INFO not in ds_checkpoint.global_state: + sd = torch.load( + ds_checkpoint.mp_rank_files[0], map_location=torch.device("cpu") + ) + if UNIVERSAL_CHECKPOINT_INFO not in sd: + ds_checkpoint.global_state[UNIVERSAL_CHECKPOINT_INFO] = {} + ds_checkpoint.global_state[UNIVERSAL_CHECKPOINT_INFO][ + UNIVERSAL_CHECKPOINT_VERSION_KEY + ] = UNIVERSAL_CHECKPOINT_VERSION_VALUE + + +def _check_for_required_state(ds_checkpoint): + universal_checkpoint_info = ds_checkpoint.get_checkpoint_info( + UNIVERSAL_CHECKPOINT_INFO + ) + assert ( + universal_checkpoint_info is not None + ), f"Required {UNIVERSAL_CHECKPOINT_INFO} state is missing in checkpoint. Verify that client creates this state." + + +def main(args): + print(f"Convert DeepSpeed Checkpoint to Universal Checkpoint") + + print( + f"Converting DeepSpeed checkpoint in {args.input_folder} to Universal checkpoint in {args.output_folder}" + ) + + optim_files = _get_optim_files(args.input_folder) + zero_stage = _get_zero_stage(optim_files) + + if zero_stage <= 2: + ds_checkpoint = DeepSpeedCheckpoint(args.input_folder) + if args.inject_missing_state: + _inject_missing_state(ds_checkpoint) + else: + _check_for_required_state(ds_checkpoint) + + iteration = ds_checkpoint.get_iteration() + # _create_latest_file(args.output_folder, iteration) + checkpoint_paths = _create_checkpoint_paths( + args.output_folder, + iteration, + ds_checkpoint.tp_degree, + ds_checkpoint.pp_degree, + ) + + slice_shapes = [] + for mp_rank_file in ds_checkpoint.mp_rank_files: + mp_sd = torch.load(mp_rank_file, map_location=torch.device("cpu")) + slice_shapes += mp_sd[PARAM_SHAPES] + + # fix back to normal flat dict, merge duplicates for tp>1 + slice_shapes = dict((k, v) for d in slice_shapes for k, v in d.items()) + temp_dir = os.path.join(args.output_folder, "tmp") + + print("1. Extracting ZeRO fragments") + _extract_zero_shard_files(args, ds_checkpoint, temp_dir) + + print("2. Merging slices .....") + _merge_tp_slice_files(args, ds_checkpoint, slice_shapes, temp_dir) + + print("3. Saving common optimizer states") + _save_optimizer_state(args, ds_checkpoint) + + if not args.keep_temp_folder: + shutil.rmtree(temp_dir, ignore_errors=True) + + # Copy mp* files into output folder + for f in glob.glob(os.path.join(args.input_folder, "mp*")): + shutil.copy2(f, args.output_folder) + + else: + model_files = _get_model_state_files(args.input_folder) + param_shapes = _parse_model_states_stage3(model_files) + param_shapes = {k: v for d in param_shapes for k, v in d.items()} + dp_degree = len(model_files) + + temp_dir = os.path.join(args.output_folder, "tmp") + + print("*** 1. Extracting ZeRO fragments") + _extract_zero_shard_files_stage3( + args, optim_files, param_shapes, dp_degree, temp_dir + ) + + print("*** 2. Merging slices .....") + _merge_zero3_slice_files(args, param_shapes, dp_degree, temp_dir) + + print("*** 3. Saving common optimizer states") + _save_optimizer_state_stage3(args, optim_files) + + if not args.keep_temp_folder: + shutil.rmtree(temp_dir, ignore_errors=True) + + # Copy *model_states files into output folder + for f in glob.glob(os.path.join(args.input_folder, "*model_states.pt")): + shutil.copy2(f, args.output_folder) + + # Update latest to output folder + checkpoint_root_folder, step_folder = os.path.split(args.output_folder) + latest_file = os.path.join(checkpoint_root_folder, "latest_universal") + with open(latest_file, "w") as f: + f.write(step_folder) + + print("*** Done!") + + +if __name__ == "__main__": + args = parse_arguments() + main(args) From 8e0bff815700b6e7c9b2ff1d19a87141150d136b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Nov 2024 12:19:53 -0600 Subject: [PATCH 363/590] docs: Add `ALCF/notes/checkpoints.md` --- ALCF/notes/checkpoints.md | 207 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 207 insertions(+) create mode 100644 ALCF/notes/checkpoints.md diff --git a/ALCF/notes/checkpoints.md b/ALCF/notes/checkpoints.md new file mode 100644 index 00000000000..f5acd425a73 --- /dev/null +++ b/ALCF/notes/checkpoints.md @@ -0,0 +1,207 @@ +# Converting Checkpoints + +## Megatron $\rightarrow$ 🤗 HuggingFace + +On Aurora, + +- Setup: + + ```bash + CKPT_ROOT="/flare/Aurora_deployment/AuroraGPT-Testing/foremans/rollback-41k8/Megatron-DeepSpeed-41800/checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05"; + + LAST_STEP=$(cat "${CKPT_ROOT}/latest_checkpointed_iteration.txt") + GLOBAL_STEP="${GLOBAL_STEP:-${LAST_STEP}}" + + SRC="${CKPT_ROOT}/global_step${GLOBAL_STEP}" + + OUTPUT_PARENT="/flare/Aurora_deployment/AuroraGPT-Checkpoints/production-checkpoints/aGPT-7B/HF" + DST="${OUTPUT_PARENT}/global_step${GLOBAL_STEP}_hf" + + printf "SRC: %s\n DST: %s\n" "${SRC}" "${DST}" + ``` + +- Convert: + + ```bash + python3 Megatron-DeepSpeed/mds_to_hf.py \ + --mds_checkpoint "${SRC}/mp_rank_00_model_states.pt" \ + --output_dir "${DST}" \ + --cache_dir "./.cache" + ``` + + + + + + + + + + [DST] --> + + + + + + + + + + + +## Use in 🤗 `transformers` + +```python +from pathlib import Path +import time +from rich import print +from typing import Optional +from transformers import LlamaForCausalLM, AutoTokenizer + +def load_model(ckpt_dir: str, step: Optional[int] = None): + if step is None: + fp = Path(ckpt_dir) + else: + fp = Path(ckpt_dir).joinpath(f"global_step{step}_hf") + print(f"Loading ckpt from: {fp}") + if fp.exists(): + model = LlamaForCausalLM.from_pretrained(fp.as_posix()) + print(f"{model=}") + return model + + raise FileNotFoundError(f"Unable to locate checkpoint at: {fp}") + + +def eval_model( + model: torch.nn.Module, + max_length: int = 64, + prompt: Optional[str] = None, + tokenizer: Optional[AutoTokenizer] = None, +) -> str: + prompt = "What is it like in there?" if prompt is None else prompt + tokenizer = ( + AutoTokenizer.from_pretrained("meta-llama/Llama-2-7B-hf") + if tokenizer is None else tokenizer + ) + output = ( + tokenizer.batch_decode( + model.generate( + **tokenizer(prompt, return_tensors="pt"), + max_length=max_length, + ), + clean_up_tokenization_spaces=True, + skip_special_tokens=True, + )[0] + ) + return output + + +def loop_over_checkpoints( + steps_list: list[int], + ckpt_dir: str, + max_length: int = 128, + prompt: Optional[str] = None, +): + for step in steps_list: + t0 = time.perf_counter() + prompt = "What is it like in there?" if prompt is None else prompt + print(f"\n Loading model from checkpoint at global step: {step}") + outputs = eval_model( + load_model(step, ckpt_dir), + max_length=max_length, + prompt=prompt, + ) + print(f"{outputs}") + print(f"\ntook: {time.perf_counter() - t0:.6f}s\n") +``` + +```python +>>> ckpt_dir = "/flare/Aurora_deployment/AuroraGPT-Checkpoints/production-checkpoints/aGPT-7B/HF/" +>>> tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-2-7B-hf") +>>> model = load_model(76300, ckpt_dir) +Loading ckpt from: +/flare/Aurora_deployment/AuroraGPT-Checkpoints/production-checkpoints/aGPT-7B/HF/global_step76300_hf +model=LlamaForCausalLM( + (model): LlamaModel( + (embed_tokens): Embedding(32000, 4096) + (layers): ModuleList( + (0-31): 32 x LlamaDecoderLayer( + (self_attn): LlamaSdpaAttention( + (q_proj): Linear(in_features=4096, out_features=4096, bias=False) + (k_proj): Linear(in_features=4096, out_features=1024, bias=False) + (v_proj): Linear(in_features=4096, out_features=1024, bias=False) + (o_proj): Linear(in_features=4096, out_features=4096, bias=False) + (rotary_emb): LlamaRotaryEmbedding() + ) + (mlp): LlamaMLP( + (gate_proj): Linear(in_features=4096, out_features=11008, bias=False) + (up_proj): Linear(in_features=4096, out_features=11008, bias=False) + (down_proj): Linear(in_features=11008, out_features=4096, bias=False) + (act_fn): SiLU() + ) + (input_layernorm): LlamaRMSNorm((4096,), eps=1e-05) + (post_attention_layernorm): LlamaRMSNorm((4096,), eps=1e-05) + ) + ) + (norm): LlamaRMSNorm((4096,), eps=1e-05) + (rotary_emb): LlamaRotaryEmbedding() + ) + (lm_head): Linear(in_features=4096, out_features=32000, bias=False) +) + +>>> print( +... eval_model( +... model, +... max_length=128, +... prompt="What is it like in there?", +... tokenizer=tokenizer +... ) +... ) +Setting `pad_token_id` to `eos_token_id`:None for open-end generation. +Starting from v4.46, the `logits` model output will have the same type as the model (except at train time, whereit will always be FP32) +What is it like in there? +I've been in there a few times. It's a pretty cool place. +I've been in there a few times. It's a pretty cool place. +I've been in there a few times. It's a pretty cool place. +I've been in there a few times. It's a pretty cool place. +I've been in there a few times. It's a pretty cool place. +I've been in there a few times. It's a pretty cool place. +I've been in +``` + +## Helper Script + +```bash +convert_mds_to_hf() { + if [[ "$#" -eq 3 ]]; then + GLOBAL_STEP=$1 + CKPT_ROOT=$2 + OUTPUT_PARENT=$3 + elif [[ "$#" -eq 2 ]]; then + GLOBAL_STEP=$1 + CKPT_ROOT=$2 + OUPUT_PARENT=$(pwd) + elif [[ "$#" -eq 1 ]]; then + GLOBAL_STEP=$1 + CKPT_ROOT="/flare/Aurora_deployment/AuroraGPT-Testing/foremans/rollback-41k8/Megatron-DeepSpeed-41800/checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/"; + OUPUT_PARENT=$(pwd) + else + echo "Expected exactly 1, 2, or 3 arguments (global_step, src, dst, respectively)" + exit + fi + SRC="${CKPT_ROOT}/global_step${GLOBAL_STEP}" + DST="${OUTPUT_PARENT}/global_step${GLOBAL_STEP}_hf" + if [[ -d "${SRC}" ]]; then + echo "Converting checkpoint @ global step ${GLOBAL_STEP}" + echo "\tsrc = ${SRC}\n" + echo "\tdst = ${DST}\n" + python3 mds_to_hf.py \ + --mds_checkpoint "${SRC}/mp_rank_00_model_states.pt" \ + --output_dir "${DST}" \ + --cache_dir "./.cache" + else + echo "Unable to locate directory ${SRC}. Exiting" + exit 1 + fi +} +``` From bd8c246bcfbdca4833e57c598e37e889218a1c8a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Nov 2024 12:21:59 -0600 Subject: [PATCH 364/590] feat: Enable `--use-flash-attn-builder` by default on Aurora --- ALCF/helpers.sh | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index fac2f4d8ecb..252e641fe1c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -578,7 +578,12 @@ setParams() { export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - MICRO_BATCH=${MICRO_BATCH:-1} # MICRO_BATCH = 4 + MICRO_BATCH=${MICRO_BATCH:-1} + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-builder" + fi #### [sam: 08/17/2024] ########################################## # Use best set of CCL env vars from Gordon Bell runs on Aurora set_ccl_vars_on_aurora @@ -590,18 +595,18 @@ setParams() { # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? # update_ccl_env_vars_aurora ###################################################################### - if [[ -z "${USE_FLASH_ATTN:-}" ]]; then - # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! - export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... - if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" - else - FLASH_ARG="--use-flash-attn-builder" - fi - else - echo "Using flash-attn !!" - FLASH_ARG="--use-flash-attn-builder" - fi + # if [[ -z "${USE_FLASH_ATTN:-}" ]]; then + # # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! + # export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... + # if [[ -n "${NO_FLASH_ATTN-}" ]]; then + # echo "Not using flash-attn!!" + # else + # FLASH_ARG="--use-flash-attn-builder" + # fi + # else + # echo "Using flash-attn !!" + # FLASH_ARG="--use-flash-attn-builder" + # fi # [Polaris] elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" From 26f2e71f1aa81222bce6d6643b3e57c8fc1ea6dc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Nov 2024 13:33:17 -0600 Subject: [PATCH 365/590] Update python.yml --- .github/workflows/python.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/python.yml b/.github/workflows/python.yml index 2f9765af241..8e1dd8f7714 100644 --- a/.github/workflows/python.yml +++ b/.github/workflows/python.yml @@ -16,7 +16,7 @@ jobs: unit-tests: strategy: matrix: - pyVersion: ["3.7", "3.8", "3.9", "3.10"] + pyVersion: ["3.9", "3.10", "3.11", "3.12"] fail-fast: false runs-on: ubuntu-22.04 From 48b3c81fc54a3d0bd45b9a7c52cdf432f2bbbfc6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Nov 2024 13:49:05 -0600 Subject: [PATCH 366/590] Update python.yml --- .github/workflows/python.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/python.yml b/.github/workflows/python.yml index 8e1dd8f7714..7f5ffafc24a 100644 --- a/.github/workflows/python.yml +++ b/.github/workflows/python.yml @@ -16,7 +16,7 @@ jobs: unit-tests: strategy: matrix: - pyVersion: ["3.9", "3.10", "3.11", "3.12"] + pyVersion: ["3.10", "3.12"] fail-fast: false runs-on: ubuntu-22.04 From 0a997bb0e5b810b6a74df7278c7a2e9cbd261af4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Nov 2024 14:08:21 -0600 Subject: [PATCH 367/590] Update python.yml --- .github/workflows/python.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/python.yml b/.github/workflows/python.yml index 7f5ffafc24a..6746ff98dc0 100644 --- a/.github/workflows/python.yml +++ b/.github/workflows/python.yml @@ -16,7 +16,7 @@ jobs: unit-tests: strategy: matrix: - pyVersion: ["3.10", "3.12"] + pyVersion: ["3.10"] fail-fast: false runs-on: ubuntu-22.04 From 1a360040d59e3d7b78147acfd933624391dce78d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 16 Nov 2024 10:26:23 -0600 Subject: [PATCH 368/590] Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 44 +++++++++++++++++++++++++++++++++++--------- 1 file changed, 35 insertions(+), 9 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 252e641fe1c..54c7b104de5 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -194,10 +194,13 @@ setup_run_cmd() { # min_lr=$(python3 -c 'print(f"{2 / (10 ** 5):.8f}")') # "--min-lr ${LR:-${min_lr}}" # 2e-5 # "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 + export LR="${LR:-0.0002}" + export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" + export LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" lr_flags=( - "--lr ${LR:-0.0002}" - "--lr-decay-style ${LR_DECAY_STYLE:-cosine}" - "--lr-warmup-fraction ${LR_WARMUP_FRAC:-0.05}" + "--lr ${LR}" + "--lr-decay-style ${LR_DECAY_STYLE}" + "--lr-warmup-fraction ${LR_WARMUP_FRAC}" ) if [[ -n "${LR_DECAY_ITERS:-}" ]]; then lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") @@ -225,9 +228,9 @@ setup_run_cmd() { "${lr_flags[@]}" "${custom_args[@]}" "${llama_flags[@]}" - "${DATA_FLAGS}" "${FLASH_ARG}" - "${TIMING_STR}" + "${TIMING_STR:-}" + "${DATA_FLAGS}" "${TOKENIZER_FLAGS}" "${tb_flags[@]}" "${ds_args[@]}" @@ -316,6 +319,8 @@ get_machine_name() { else machine="polaris" fi + elif [[ $(hostname) == sophia* ]]; then + machine="sophia" elif [[ $(hostname) == nid* ]]; then machine="perlmutter" else @@ -325,6 +330,7 @@ get_machine_name() { } get_machine() { + machine=$(hostname) if [[ $(hostname) == x4* ]]; then machine="aurora" elif [[ $(hostname) == x1* ]]; then @@ -335,6 +341,8 @@ get_machine() { else machine="polaris" fi + elif [[ $(hostname) == sophia* ]]; then + machine="sophia" elif [[ $(hostname) == nid* ]]; then machine="perlmutter" else @@ -366,7 +374,7 @@ setupSrun() { printJobInfo() { echo "++++++++++++++++++++++++++++++++++++++++++++++++++" - echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT}}" + echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" echo "- Using $(which python3)" echo "- WORLD_SIZE:${WORLD_SIZE-}" echo "- BACKEND: ${BE:-}" @@ -406,6 +414,8 @@ setupLauncher() { mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" + elif [[ "${mn}" == "sophia" ]]; then + LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" else LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" fi @@ -626,6 +636,22 @@ setParams() { fi echo "Setting up AWS NCCL OFI Plugin on Polaris..." source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit + # ---- [Sophia] ---------------------- + elif [[ "${mn}" == sophia* ]]; then + # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL + export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-v2" + fi + # echo "Setting up AWS NCCL OFI Plugin on Polaris..." + # source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # [Perlmutter] elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then TP="${TP:-2}" @@ -1046,7 +1072,7 @@ setup_tokenizer_and_data() { export TOKENIZER_TYPE="GPT2" _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") machine=$(get_machine_name) - if [[ ${machine} == "polaris" ]]; then + if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" @@ -1075,7 +1101,7 @@ setup_tokenizer_and_data() { echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST:-}" setData "${dfl}" || exit fi - export DATA_FLAGS="${_data_flags[*]}" + export DATA_FLAGS="${_data_flags[*]:-}" export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" @@ -1113,7 +1139,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" printf "DFL_STEM: %s\n" "${DFL_STEM}" printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" - printf "DATA_FLAGS: %s\n" "${DATA_FLAGS}" + printf "DATA_FLAGS: %s\n" "${DATA_FLAGS:-}" echo "--------------------" } From b8007f4334eb335d5d97cd0cdfd3dadfbe096ec1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 25 Nov 2024 15:57:59 -0600 Subject: [PATCH 369/590] fix: `GRAD_ACC_STEPS` on 32 nodes of Aurora --- ALCF/helpers.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 54c7b104de5..a52fc2bbb1d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -522,9 +522,9 @@ get_grad_acc_steps_on_aurora() { gas=1 elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then gas=2 - elif [[ 32 -le "${nhosts}" && "${nhosts}" -lt 128 ]]; then + elif [[ 32 -lt "${nhosts}" && "${nhosts}" -lt 129 ]]; then gas=4 - elif [[ 16 -le "${nhosts}" && "${nhosts}" -lt 32 ]]; then + elif [[ 16 -le "${nhosts}" && "${nhosts}" -le 32 ]]; then gas=8 else gas=16 From 854cfc771119ce94d3cb74d4e3d6b53b164bf802 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 23 Dec 2024 13:55:02 -0600 Subject: [PATCH 370/590] chore: Fix printing in `ALCF/helpers.sh` --- ALCF/helpers.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a52fc2bbb1d..64ab51cc0a7 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -922,8 +922,8 @@ buildDSconfig() { # export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" mkdir -p "$(dirname "${DS_CONFIG}")" - echo "DS_CONFIG: ${DS_CONFIG}" - printf "ZS: %s, MB: %s, GB: %s, PP: %s, DTYPE: %s" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" + printf "DS_CONFIG: %s\n" "${DS_CONFIG}" + printf "ZS=%s, MB=%s, GB=%s, PP=%s, DTYPE=%s\n" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" generateDSconfig "${DS_CONFIG}" cat "${DS_CONFIG}" | jq . } From 7a6a5bd6b14da005d39dbe348906b45c27656281 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 23 Dec 2024 13:55:45 -0600 Subject: [PATCH 371/590] chore: Alphabetize optimizer(s) in `megatron/arguments.py` --- megatron/arguments.py | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 9b0e6ccb1ac..cef72acf51f 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -954,20 +954,22 @@ def _add_training_args(parser): default='adam', choices=[ 'adam', + 'adam8bit', 'adamw', - 'sophiag', - 'sgd', - 'ds.fusedlamb', - 'ipex.lamb', - 'ipex.fusedlamb', + 'adamwschedulefree', 'apex.adam', 'apex.sgd', - 'adamwschedulefree', - 'sgdschedulefree', + 'ds.fusedlamb', + 'ds.onebitlamb', 'galoreadamw', - 'adam8bit', 'galoreadamw8bit', - 'galoreadamw8bitperlayer' + 'galoreadamw8bitperlayer', + 'ipex.fusedlamb', + 'ipex.lamb', + 'shampoo', + 'sgd', + 'sgdschedulefree', + 'sophiag' ], help='Optimizer function' ) From 37194a6bb0139be9aa28ad8ee106f3f9ff46a167 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 23 Dec 2024 13:56:09 -0600 Subject: [PATCH 372/590] chore: Update `megatron/optimizer/__init__.py` --- megatron/optimizer/__init__.py | 205 +++++++++++++++++++-------------- 1 file changed, 121 insertions(+), 84 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 99145ff4f49..37107a44e71 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -127,6 +127,8 @@ def get_megatron_optimizer( param_groups ) + optimizer = None + # ---- CPU Optimizer -------------------------------------- if args.cpu_optimizer: assert args.optimizer == 'adam', 'CPU offloading is for Adam' if args.cpu_torch_adam: @@ -141,52 +143,73 @@ def get_megatron_optimizer( betas=(args.adam_beta1, args.adam_beta2), eps=args.adam_eps, ) - - elif args.optimizer.lower() == "galore_adamw": - from galore_torch import GaLoreAdamW, GaLoreAdamW8bit - # redefine way to call galore_adamw - optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) - elif args.optimizer.lower() == "galore_adamw": - # redefine way to call galore_adamw - optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) - # implement adafactor - elif args.optimizer.lower() == "adafactor": - import transformers - args.beta1 = None if args.beta1 == 0.0 else args.beta1 - optimizer = transformers.optimization.Adafactor( + # ---- Adam -------------------------------------- + elif args.optimizer == 'adam': + if args.ds_fused_adam: + # global Adam + from deepspeed.ops.adam import FusedAdam + Adam = FusedAdam + else: + Adam = torch.optim.Adam + optimizer = Adam( param_groups, lr=args.lr, - eps=(1e-30, 1e-3), - clip_threshold=1.0, - decay_rate=-0.8, - beta1=args.beta1, weight_decay=args.weight_decay, - relative_step=False, - scale_parameter=False, - warmup_init=False, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps ) - # low-rank adafactor - elif args.optimizer.lower() == "galore_adafactor": - args.beta1 = None if args.beta1 == 0.0 else args.beta1 - optimizer = GaLoreAdafactor( + # ---- apex.Adam -------------------------------------------- + elif str(args.optimizer).lower() == 'apex.adam': + assert get_accelerator().device_name() == 'cuda' + from apex.optimizers import FusedAdam as Adam + optimizer = Adam( param_groups, lr=args.lr, - eps=(1e-30, 1e-3), - clip_threshold=1.0, - decay_rate=-0.8, - beta1=args.beta1, weight_decay=args.weight_decay, - relative_step=False, - scale_parameter=False, - warmup_init=False, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps ) - # 8-bit Adam + # ---- Adam8Bit -------------------------------------- elif args.optimizer.lower() == "adam8bit": import bitsandbytes as bnb optimizer = bnb.optim.Adam8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + # ---- AdamW -------------------------------------- + elif str(args.optimizer).lower() == 'adamw': + optimizer = torch.optim.AdamW( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps + ) + # ---- AdamW: ScheduleFree ------------------------------------- + elif str(args.optimizer).lower() == 'adamwschedulefree': + import schedulefree + optimizer = schedulefree.AdamWScheduleFree( + param_groups, + lr=args.lr, + weight_decay=args.weight_decay, + betas=(args.adam_beta1, args.adam_beta2), + eps=args.adam_eps, + warmup_steps=args.lr_warmup_iters, + foreach=args.schedulefree_for_each, + ) + # ---- AdamW: Galore ------------------------------------------ + elif args.optimizer.lower() == "galore_adamw": + from galore_torch import GaLoreAdamW + # redefine way to call galore_adamw + optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) + # elif args.optimizer.lower() == "galore_adamw": + # from galore_torch import GaLoreAdamW + # # redefine way to call galore_adamw + # optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) + # ---- AdamW: GaloRe 8Bit -------------------------------------- elif args.optimizer.lower() == "galore_adamw8bit": + from galore_torch import GaLoreAdamW8bit optimizer = GaLoreAdamW8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + # ---- AdamW8bitPerLayer: GaloRE ---------------------------- elif args.optimizer.lower() == 'galore_adamw8bit_per_layer': + from galore_torch import GaLoreAdamW8bit # TODO: seems scheduler call twice in one update step, need to check, for now double the num_training_steps, warmup_steps and update_proj_gap optimizer_dict = {} for p in model.parameters(): @@ -219,45 +242,48 @@ def optimizer_hook(p): if p.requires_grad: p.register_post_accumulate_grad_hook(optimizer_hook) layer_wise_flag = True - elif str(args.optimizer) == 'ipex.lamb': - from intel_extension_for_pytorch.optim._lamb import Lamb - optimizer = Lamb( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps, - ) - elif str(args.optimizer) == 'ipex.fusedlamb': - from intel_extension_for_pytorch.optim._lamb import Lamb - optimizer = Lamb( + # ---- AdaFactor -------------------------------------- + elif args.optimizer.lower() == "adafactor": + import transformers + args.beta1 = None if args.beta1 == 0.0 else args.beta1 + optimizer = transformers.optimization.Adafactor( param_groups, lr=args.lr, + eps=(1e-30, 1e-3), + clip_threshold=1.0, + decay_rate=-0.8, + beta1=args.beta1, weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps, - fused=True, + relative_step=False, + scale_parameter=False, + warmup_init=False, ) - elif str(args.optimizer).lower() == 'ds.fusedlamb': - from deepspeed.ops.lamb import FusedLamb - optimizer = FusedLamb( + # ---- GaLore: Adafactor adafactor ------------------------------------ + elif args.optimizer.lower() == "galore_adafactor": + from galore_torch import GaLoreAdafactor + args.beta1 = None if args.beta1 == 0.0 else args.beta1 + optimizer = GaLoreAdafactor( param_groups, lr=args.lr, + eps=(1e-30, 1e-3), + clip_threshold=1.0, + decay_rate=-0.8, + beta1=args.beta1, weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps, + relative_step=False, + scale_parameter=False, + warmup_init=False, ) - elif str(args.optimizer).lower() == 'adamwschedulefree': - import schedulefree - optimizer = schedulefree.AdamWScheduleFree( + # ---- Apex: sgd --------------------------------------------- + elif str(args.optimizer).lower() == 'apex.sgd': + from apex.optimizers import FusedSGD as SGD + optimizer = SGD( param_groups, lr=args.lr, weight_decay=args.weight_decay, - betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps, - warmup_steps=args.lr_warmup_iters, - foreach=args.schedulefree_for_each, + momentum=args.sgd_momentum ) + # ---- ScheduleFree: SGD ------------------------------- elif str(args.optimizer).lower() == 'sgdschedulefree': import schedulefree optimizer = schedulefree.SGDScheduleFree( @@ -268,45 +294,54 @@ def optimizer_hook(p): warmup_steps=args.lr_warmup_iters, foreach=args.schedulefree_for_each, ) - elif str(args.optimizer).lower() == 'apex.adam': - assert get_accelerator().device_name() == 'cuda' - from apex.optimizers import FusedAdam as Adam - optimizer = Adam( + # ---- Lamb: Ipex -------------------------------------------- + elif str(args.optimizer) == 'ipex.lamb': + from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( param_groups, lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps - ) - elif str(args.optimizer).lower() == 'apex.sgd': - from apex.optimizers import FusedSGD as SGD - optimizer = SGD( - param_groups, - lr=args.lr, - weight_decay=args.weight_decay, - momentum=args.sgd_momentum + eps=args.adam_eps, ) - elif str(args.optimizer).lower() == 'adamw': - optimizer = torch.optim.AdamW( + # ---- Lamb(Fused): Ipex ---------------------------------------- + elif str(args.optimizer) == 'ipex.fusedlamb': + from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( param_groups, lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps + eps=args.adam_eps, + fused=True, ) - elif args.optimizer == 'adam': - if args.ds_fused_adam: - # global Adam - from deepspeed.ops.adam import FusedAdam - Adam = FusedAdam - else: - Adam = torch.optim.Adam - optimizer = Adam( + # ---- Lamb(Fused): DeepSpeed ------------------------------------------ + elif str(args.optimizer).lower() == 'ds.fusedlamb': + from deepspeed.ops.lamb import FusedLamb + optimizer = FusedLamb( param_groups, lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps + eps=args.adam_eps, + ) + # ---- Shampoo ---------------------------------------- + elif args.optimizer == 'shampoo': + from distributed_shampoo.distributed_shampoo import DistributedShampoo + from distributed_shampoo.shampoo_types import AdamGraftingConfig + optimizer = DistributedShampoo( + model.parameters(), + lr=0.001, + betas=(0.9, 0.999), + epsilon=1e-12, + weight_decay=1e-05, + max_preconditioner_dim=8192, + precondition_frequency=100, + use_decoupled_weight_decay=True, + grafting_config=AdamGraftingConfig( + beta2=0.999, + epsilon=1e-08, + ), ) elif args.optimizer == 'sgd': optimizer = torch.optim.SGD( @@ -326,8 +361,10 @@ def optimizer_hook(p): ) else: raise TypeError(f'{args.optimizer} optimizer is not supported.') + assert optimizer is not None if args.deepspeed: return optimizer + # Determine whether the params have main-grad field. params_have_main_grad = False if args.use_contiguous_buffers_in_local_ddp: From 3413625eae82072e43cb4125fd2f8cd43282bda4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 23 Dec 2024 13:56:29 -0600 Subject: [PATCH 373/590] chore: Remove blankline in `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 1 - 1 file changed, 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 3686c6ceeb7..c196b19a420 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -1,7 +1,6 @@ # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain GPT""" - import time from typing import Callable from mpi4py import MPI From 1504bd011fcda3728248eb017151d29d81f5668e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 23 Dec 2024 13:57:02 -0600 Subject: [PATCH 374/590] feat: Use logging in `tools/hf2megads_weight_converter.py` --- tools/hf2megads_weight_converter.py | 84 +++++++++++++++++------------ 1 file changed, 50 insertions(+), 34 deletions(-) diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py index 12468963c54..76d40ce2528 100755 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -2,15 +2,18 @@ import re import sys import os +import ezpz as ez + sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) import torch.distributed from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP -from megatron import print_rank_0, get_tokenizer, get_args +from megatron import get_tokenizer, get_args +from megatron.tokenizer.tokenizer import _vocab_size_with_padding from megatron.core import mpu from megatron.core import tensor_parallel from megatron.core.utils import divide from megatron.model import GPTModelPipe, Float16Module -from megatron.utils import unwrap_model +from megatron.utils import unwrap_model, get_logger from megatron.model import DistributedDataParallel as LocalDDP from megatron.arguments import core_transformer_config_from_args from megatron.initialize import initialize_megatron @@ -22,6 +25,8 @@ import copy from pathlib import Path +RANK = ez.setup_torch('deepspeed') +log = get_logger(__name__, rank_zero_only=True) def add_extra_args(parser): @@ -52,7 +57,7 @@ def compute_partition_range(hidden_size, local_rank, tp_size): def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. loaded = {} - print_rank_0( + log.info( f"----------------------------hf weight list----------------------------") for wid in range(1, hf_ckpt_num_of_shards + 1): @@ -60,7 +65,7 @@ def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): f"{hf_ckpt_dir}/pytorch_model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.bin", map_location=torch.device('cpu')) for k in d: - print_rank_0(k) + log.info(k) assert k not in loaded loaded[k] = d[k].clone() del d @@ -71,7 +76,7 @@ def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards) from safetensors import safe_open # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. hf_model = {} - print_rank_0( + log.info( f"----------------------------hf weight list----------------------------") for wid in range(1, hf_ckpt_num_of_shards + 1): @@ -82,7 +87,7 @@ def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards) with safe_open(ckpt_path, framework="pt", device="cpu") as f: for k in f.keys(): - print_rank_0(f"name: {k}, shape: {f.get_tensor(k).shape}") + log.info(f"name: {k}, shape: {f.get_tensor(k).shape}") assert k not in hf_model hf_model[k] = f.get_tensor(k).clone() @@ -100,18 +105,18 @@ def load_and_print_hf_weight_auto(hf_ckpt_dir, no_init=True): else: hf_model = {} hf_auto_model = AutoModelForCausalLM.from_pretrained(hf_ckpt_dir, trust_remote_code=True, torch_dtype=torch.bfloat16) - print_rank_0( + log.info( f"----------------------------hf weight list----------------------------") for name, param in hf_auto_model.named_parameters(): hf_model[name] = param.clone() - print_rank_0(name) + log.info(name) return hf_model def print_distinct_weights(model): - print_rank_0( + log.info( f"----------------------------mega-ds weight list----------------------------") for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): if mpu.get_pipeline_model_parallel_rank() == pipe_rank: @@ -154,7 +159,11 @@ def _embedding_refactor(self, pname, p): elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": hf_name = "model.embed_tokens.weight" hf_w = self.hf_model[hf_name] - assert hf_w.shape[0] == self.token_vocab + log.info(f"{hf_w.shape[0]=}") + log.info(f"{self.token_vocab=}") + if hf_w.shape[0] != self.padded_vocab_size: + torch.distributed.breakpoint(0) + assert hf_w.shape[0] == self.padded_vocab_size per_partition_vocab_size, start_index, end_index = compute_partition_range( self.padded_vocab_size, self.tp_rank, self.tp_size) end_index = min(end_index, self.token_vocab) @@ -170,9 +179,6 @@ def _embedding_refactor(self, pname, p): ) return new_w - - - def _direct_refactor(self, pname, p, hf_layer=None, subname=None): if pname == f"{self.mega_norm_wnum}.weight": hf_name = "model.norm.weight" @@ -203,16 +209,26 @@ def _qkv_refactor(self, pname, p, hf_layer): new_w = torch.zeros((per_partition_size * 3, wq.shape[1]), dtype=wq.dtype) + # >>> pname + # '2.self_attention.query_key_value.weight' + # >>> p.shap^U + # '2.self_attention.query_key_value.weight' + # >>> xp = p + # >>> xp.shape + # torch.Size([6144, 4096]) for i in range(num_attention_heads_per_partition): - current_index = start_index + i * hidden_size_per_attention_head - next_index = current_index + hidden_size_per_attention_head - new_w_index = i * (3 * hidden_size_per_attention_head) - new_w[new_w_index: new_w_index + (3 * hidden_size_per_attention_head), :] = \ - torch.cat([ - wq[current_index: next_index, :], - wk[current_index: next_index, :], - wv[current_index: next_index, :] - ], dim=0) + try: + current_index = start_index + i * hidden_size_per_attention_head + next_index = current_index + hidden_size_per_attention_head + new_w_index = i * (3 * hidden_size_per_attention_head) + new_w[new_w_index: new_w_index + (3 * hidden_size_per_attention_head), :] = \ + torch.cat([ + wq[current_index: next_index, :], + wk[current_index: next_index, :], + wv[current_index: next_index, :] + ], dim=0) + except Exception: + torch.distributed.breakpoint(0) self.record_mapping_info( f"mega-ds:{pname,p.data.shape}<--hf{hf_wq_name,hf_wk_name,hf_wv_name,} cat q,k,v [{current_index}:{next_index},:] of q,k,v{wq.shape}" ) @@ -275,7 +291,7 @@ def _mlphto4h1_refactor(self, pname, p, hf_layer, subname): return new_w def transform_from_hf_to_megds(self): - assert self.is_refactored == False + assert not self.is_refactored new_w = None for pname, p in self.ds_model.named_parameters(): @@ -288,6 +304,7 @@ def transform_from_hf_to_megds(self): new_w = self._direct_refactor(pname, p) else: mobj = self.decoder_pat.match(pname) + assert mobj is not None layer_num = int(mobj.group(1)) subname = mobj.group(2) hf_layer = layer_num - self.offset_num @@ -316,7 +333,7 @@ def transform_from_hf_to_megds(self): new_w = None self.is_refactored = True - + def _embedding_refactor_to_hf(self, pname, ds_w): if pname == f"{self.mega_lm_head_wnum}.lm_head.weight": hf_w = self.hf_model.lm_head.weight @@ -327,7 +344,7 @@ def _embedding_refactor_to_hf(self, pname, ds_w): with torch.no_grad(): ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) - + self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank[:hf_w.shape[0], :]) def _direct_refactor_to_hf(self, pname, ds_w, hf_layer=None, subname=None): @@ -438,7 +455,7 @@ def record_mapping_info(self, record_msg): def inorder_show_record(self): assert self.is_refactored - print_rank_0( + log.info( f"----------------------------mapping list----------------------------") # print dp rank0 tp rank0 records. for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): @@ -466,7 +483,7 @@ def load_hf_weights(args, no_init): def convert_ckpt(): """Build the model.""" args = get_args() - print_rank_0(f'building model ...') + log.info(f'building model ...') see_memory_usage(f"Before Building Model", force=True) config = core_transformer_config_from_args(args) @@ -491,18 +508,18 @@ def convert_ckpt(): # print_distinct_weights(hf_model) #init model and save - print_rank_0(f"before deepspeed init") + log.info(f"before deepspeed init") ds_engine, _, _, _ = deepspeed.initialize( model=ds_model, optimizer=None, args=args, lr_scheduler=None, mpu=mpu if args.no_pipeline_parallel else None) - print_rank_0(f"after deepspeed init") + log.info(f"after deepspeed init") if args.to_hf_ckpt: load_checkpoint([ds_engine], None, None, load_only_weights=True) - print_rank_0(f"completed to load deepspeed actual checkpoint") + log.info(f"completed to load deepspeed actual checkpoint") # refactor weight from hf to mega-ds and vice versa @@ -523,7 +540,7 @@ def convert_ckpt(): if torch.distributed.is_initialized(): torch.distributed.barrier() - print_rank_0(f"hf checkpoint will be saved in {save_path}/release ") + log.info(f"hf checkpoint will be saved in {save_path}/release ") if mpu.is_pipeline_last_stage(): ## doing checkpoint merging and saving... # hf_model.tie_weights() @@ -539,10 +556,9 @@ def convert_ckpt(): # mega-ds checkpoint will be saved in args.save hf_model.save_pretrained(os.path.join(save_path, "release"), safe_serialization=True) else: - print_rank_0(f"mega-ds checkpoint will be saved in {args.save}") + log.info(f"mega-ds checkpoint will be saved in {args.save}") save_checkpoint(0, [ds_engine], None, None) - - print_rank_0(f"save checkpoint completed") + log.info(f"save checkpoint completed") if __name__ == "__main__": From f1b9e8d8133b964a186eb88eeef4675eea88e345 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 29 Dec 2024 11:28:42 -0600 Subject: [PATCH 375/590] docs: Add `ALCF/notes/universal_checkpoint_bug.md` --- ALCF/notes/universal_checkpoint_bug.md | 372 +++++++++++++++++++++++++ 1 file changed, 372 insertions(+) create mode 100644 ALCF/notes/universal_checkpoint_bug.md diff --git a/ALCF/notes/universal_checkpoint_bug.md b/ALCF/notes/universal_checkpoint_bug.md new file mode 100644 index 00000000000..d7d17cb52f3 --- /dev/null +++ b/ALCF/notes/universal_checkpoint_bug.md @@ -0,0 +1,372 @@ +# 🐛 Universal Checkpoint Conversion Bug (with `DP=768`) + +## Table of Contents + +- [📓 Summary](#summary) +- [🚧 Issue](#issue) +- [🤔 Why is this Happening?](#-why-is-this-happening) +- [🔍 Running with Debugger](#-running-with-debugger) +- [🧰 Fix](#-fix) +- [✅ Confirm Fix Works](#-confirm-fix-works) +- [👻 Bug Doesn't Appear for Smaller Checkpoints](#-bug-doesnt-appear-for-smaller-checkpoints) + + +## 📓 Summary + +- Everything works _as is_ for checkpoints created on small scales (small DP ?) + - Explicitly confrm this + (see [👻 Bug Doesn't appear for smaller checkpoints](#bug-doesnt-appear-for-smaller-checkpoints)) + by: + + 1. Generate checkpoint from using 4 nodes of Aurora + 2. Convert this checkpoint to universal without issue + +- Trying to repeat this exact same process, but using a c checkpoint created + with `DP=768` (12 nodes of Aurora) fails with a `RuntimeError` + + +- ❌ `RuntimeError` when trying to convert checkpoint created with `DP=768` (12 nodes of Aurora) + - Description of bug + - 🧰 Proposed fix + - ✅ Confirmation of Fix + + +## 🚧 Issue + +Trying to convert a checkpoint[^parallel-config] created with `DP=768` (12 nodes of Aurora) to +universal checkpoint format, I encountered the following `RuntimeError`: + +```python +RuntimeError: narrow(): length must be non-negative. +``` + +[^parallel-config]: In the example described here, + we the checkpoint was created with: + + ```yaml + data_parallel_size: 768 + ds_sequence_parallel_size: 1 + no_pipeline_parallel: true + pipeline_model_parallel_size: 1 + pipeline_model_parallel_split_rank: null + sequence_parallel: false + tensor_model_parallel_size: 1 + transformer_pipeline_model_parallel_size: 1 + ``` + +The full command and traceback are included below: + +```bash +#[01:06:34 AM][x4705c5s4b0n0][/f/A/f/p/a/t/2/Megatron-DeepSpeed][🌱 docs-ucp-bug][?] +$ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05 ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder +``` + +Output: + +```bash +On branch master +Your branch is up to date with 'origin/master'. + +nothing to commit, working tree clean +/flare/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed +global step: 95600 +[2024-12-29 01:06:45,089] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +[2024-12-29 01:06:45,434] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +args = Namespace(input_folder='checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600', output_folder='checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600_universal', num_extract_workers=4, num_merge_workers=2, keep_temp_folder=True, strict=True, inject_missing_state=False) +Convert DeepSpeed Checkpoint to Universal Checkpoint +Converting DeepSpeed checkpoint in checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600 to Universal checkpoint in checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600_universal +*** 1. Extracting ZeRO fragments +100%|████████████████████████▋| 767/768 [01:49<00:00, 7.01it/s] +``` + +before crashing with the following traceback: + +```python +concurrent.futures.process._RemoteTraceback: +Traceback (most recent call last): + File "/opt/aurora/24.180.1/frameworks/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/concurrent/futures/process.py", line 246, in _process_worker + r = call_item.fn(*call_item.args, **call_item.kwargs) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 114, in extract_zero_shards + sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/deepspeed_checkpoint.py", line 124, in get_zero_checkpoint_state + return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 62, in get_state_for_rank + self._strip_tensor_paddings(sd) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 110, in _strip_tensor_paddings + group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +RuntimeError: narrow(): length must be non-negative. +``` + + + +## 🤔 Why is this Happening? + +The `extract_zero_shards` function [here] in `ds_to_universal.py` + +[`extract_zero_shards`](/ALCF/ds_to_universal.py#L132) function, the +`raw_length` is calculated as `raw_length = state_value.numel() - +group_paddings[key]`. + +This is causing the `raw_length` to be negative in some +cases. This is causing the following error: + + +```python +def extract_zero_shards(dir, ds_checkpoint, indices_3D): + pp_index, tp_index, dp_index = indices_3D + sd = ds_checkpoint.get_zero_checkpoint_state( + pp_index=pp_index, + tp_index=tp_index, + dp_index=dp_index, + strip_tensor_paddings=False, + ) +``` + + +In particular, in the call + +```python +sd = ds_checkpoint.get_zero_checkpoint_state( + pp_index=pp_index, tp_index=tp_index, dp_index=dp_index +) +``` + +this then calls the `DeepSpeedCheckpoint.get_zero_checkpoint_state` method +[here](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/deepspeed_checkpoint.py#L123-L127) + +```python +def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: + return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, + tp_index=tp_index, + dp_index=dp_index, + keys_to_ignore=[PARAM_SHAPES]) +``` + +This is a call to the `ZeROCheckpoint.get_state_for_rank` method from +[here](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/zero_checkpoint.py#L53-L73) + +Now, this `get_state_for_rank` function accepts an argument +`strip_tensor_paddings` which is true by default. + +Calling the `strip_tensor_paddings` method (with `strip_tensor_paddings=True`), we hit: + +```python +for state_file in state_file_list: + # ...clipped... + if strip_tensor_paddings: + self._strip_tensor_paddings(sd) # <-- this is where the error is raised +``` + +and we hit the `RuntimeError` when calling this +`self._strip_tensor_paddings(sd)` method. + +Stepping into the `self._strip_tensor_paddings` method, we see that it calls +[this block](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/zero_checkpoint.py#L108-L110) +which calculates the `raw_length` as: + +```python +def _strip_tensor_paddings(self, sd): + param_group_states = self._get_param_group_states(sd) + if param_group_states is None: + return + + group_paddings = self._get_optimizer_state(sd, GROUP_PADDINGS) + if group_paddings is None: + return + + for key, group_state in param_group_states.items(): + if group_paddings[key] == 0: + continue + for state_name, state_value in group_state.items(): + if state_name != "step" and torch.is_tensor(state_value): + # 🐛 see debugger output below + raw_length = state_value.numel() - group_paddings[key] # <-- this is negative + group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() + else: + group_state[state_name] = state_value +``` + +## 🔍 Running with Debugger + +Running with: + +```bash +python3 -m pudb deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py \ + --input_folder "${ckpt_dir}/global_step${gs}" \ + --output_folder "${ckpt_dir}/global_step${gs}_universal" \ + --keep_temp_folder \ + --num_extract_workers 1 \ + --num_merge_workers 1 +``` + +
Traceback + +```python +Traceback (most recent call last): + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/__init__.py", line 176, in _runscript + dbg._runscript(mainpyfile) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/debugger.py", line 529, in _runscript + self.run(statement) + File "/opt/aurora/24.180.1/frameworks/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/bdb.py", line 598, in run + exec(cmd, globals, locals) + File "", line 1, in + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 549, in + main(args) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 499, in main + _extract_zero_shard_files(args, ds_checkpoint, temp_dir) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 370, in _extract_zero_shard_files + _do_parallel_work(do_work, _3d_range_list, args.num_extract_workers) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 359, in _do_parallel_work + results.append(do_work(work)) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 114, in extract_zero_shards + sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/deepspeed_checkpoint.py", line 124, in get_zero_checkpoint_state + return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 62, in get_state_for_rank + self._strip_tensor_paddings(sd) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 110, in _strip_tensor_paddings + group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +RuntimeError: narrow(): length must be non-negative. +``` + +
+ +Stepping through this command: + +```python +>>> group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +Traceback (most recent call last): + File "", line 1, in +RuntimeError: narrow(): length must be non-negative. + +>>> raw_length +-676 +>>> state_value.numel() +348 +>>> group_paddings[key] +1024 +>>> key +1 +>>> state_name +'exp_avg' +>>> key +1 +``` + +## 🧰 Fix + +Naively, the first thing I tried was to see if I could just skip this +`strip_tensor_paddings` step by setting `strip_tensor_paddings=False` in the +call to `get_zero_checkpoint_state` in the `extract_zero_shards` function, as +shown below: + +```diff +diff --git a/deepspeed/checkpoint/ds_to_universal.py b/deepspeed/checkpoint/ds_to_universal.py +index f7b75eee..cbbbef6b 100755 +--- a/deepspeed/checkpoint/ds_to_universal.py ++++ b/deepspeed/checkpoint/ds_to_universal.py +@@ -111,7 +111,7 @@ def _save_checkpoint(file_path, chkpt_sd): + + def extract_zero_shards(dir, ds_checkpoint, indices_3D): + pp_index, tp_index, dp_index = indices_3D +- sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) ++ sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index, strip_tensor_paddings=False) + + # pprint(f"Processing {dp_index=} {pp_index=}, {tp_index=}") +``` + +But unfortunately, the `DeepSpeedCheckpoint` doesn't allow the +`strip_tensor_paddings` argument to be passed through to the +`get_zero_checkpoint_state` method, requiring an additional modification: + +```diff +warning: Empty last update token. +diff --git a/deepspeed/checkpoint/deepspeed_checkpoint.py b/deepspeed/checkpoint/deepspeed_checkpoint.py +index 31997177..a2ef5d0d 100644 +--- a/deepspeed/checkpoint/deepspeed_checkpoint.py ++++ b/deepspeed/checkpoint/deepspeed_checkpoint.py +@@ -120,11 +120,12 @@ class DeepSpeedCheckpoint(object): + self.global_state[ITERATION_KEY] = sd.get(ITERATION_KEY, 0) + self.global_state[ARGS_KEY] = sd.get(ARGS_KEY, None) + +- def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: +- return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, ++ def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index, strip_tensor_paddings: bool = True) -> dict: ++ return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, # type:ignore + tp_index=tp_index, + dp_index=dp_index, +- keys_to_ignore=[PARAM_SHAPES]) ++ keys_to_ignore=[PARAM_SHAPES], ++ strip_tensor_paddings=strip_tensor_paddings) + + def get_zero_files(self, pp_index, tp_index, dp_index) -> list: + return self.zero_checkpoint.get_files_for_rank(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) +``` + +### ✅ Confirm Fix Works + +```bash +$ cd deps/DeepSpeed && git status && git checkout 'saforem2/ucp-bug' && PAGER='' git diff deepspeed/checkpoint/ && cd - && ckpt_dir=checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05 ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder +On branch saforem2/ucp-bug +nothing to commit, working tree clean +Already on 'saforem2/ucp-bug' +/flare/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed +global step: 95600 +[2024-12-29 01:58:23,658] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +[2024-12-29 01:58:30,635] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +args = Namespace(input_folder='checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600', output_folder='checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600_universal', num_extract_workers=4, num_merge_workers=2, keep_temp_folder=True, strict=True, inject_missing_state=False) +Convert DeepSpeed Checkpoint to Universal Checkpoint +Converting DeepSpeed checkpoint in checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600 to Universal checkpoint in checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05/global_step95600_universal +*** 1. Extracting ZeRO fragments +100%|██████████████████████████| 768/768 [04:15<00:00, 3.00it/s] +*** 2. Merging slices ..... +100%|██████████████████████████| 195/195 [01:11<00:00, 2.74it/s] +*** 3. Saving common optimizer states +*** Done! +took: 0h:09m:00s +``` + +## 👻 Bug Doesn't Appear for Smaller Checkpoints + +As a sanity check, we can explicitly test: + +1. Create checkpoint on 4 nodes of Aurora + + ```bash + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt NLAYERS=1 SAVE_INTERVAL=10 bash train_aGPT_7B.sh + # ...irrelevant output clipped... + [2024-12-29 00:41:41.185272][INFO][utils.py:368] successfully saved checkpoint at iteration 690 to checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash + ``` + +2. Convert to universal checkpoint using `DeepSpeed` master (**unchanged**), and + confirm that it works without issue. + +Using the checkpoint generated in 1., we use +[DeepSpeed / deepspeed / checkpoints / `ds_to_universal.py`](https://github.com/microsoft/DeepSpeed/blob/master/deepspeed/checkpoint/ds_to_universal.py) +**as is**: + +```bash +#[🐍 aurora_nre_models_frameworks-2024.2.1_u1](👻 aurora_nre_models_frameworks-2024.2.1_u1) +#[12:46:03 AM][x4705c5s4b0n0][/f/A/f/p/a/t/2/Megatron-DeepSpeed][🌱 docs-ucp-bug][?] +$ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder +On branch master # on master +Your branch is up to date with 'origin/master'. # no local changes +nothing to commit, working tree clean +/flare/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed +global step: 690 # <-- ckpt from 1. +[2024-12-29 00:46:14,466] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +[2024-12-29 00:46:14,832] [INFO] [real_accelerator.py:222:get_accelerator] Setting ds_accelerator to xpu (auto detect) +args = Namespace(input_folder='checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash/global_step690', output_folder='checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash/global_step690_universal', num_extract_workers=4, num_merge_workers=2, keep_temp_folder=True, strict=True, inject_missing_state=False) +Convert DeepSpeed Checkpoint to Universal Checkpoint +Converting DeepSpeed checkpoint in checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash/global_step690 to Universal checkpoint in checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash/global_step690_universal +*** 1. Extracting ZeRO fragments +100%|██████████████████████████████████| 48/48 [00:19<00:00, 2.44it/s] +*** 2. Merging slices ..... +100%|██████████████████████████████████| 9/9 [00:05<00:00, 1.68it/s] +*** 3. Saving common optimizer states +*** Done! +took: 0h:01m:40s +``` From 246e82bdfe22df5b3cddb644d5e64b46736c8164 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 29 Dec 2024 12:16:27 -0600 Subject: [PATCH 376/590] Update universal_checkpoint_bug.md --- ALCF/notes/universal_checkpoint_bug.md | 290 +++++++++++++------------ 1 file changed, 146 insertions(+), 144 deletions(-) diff --git a/ALCF/notes/universal_checkpoint_bug.md b/ALCF/notes/universal_checkpoint_bug.md index d7d17cb52f3..d25afd41169 100644 --- a/ALCF/notes/universal_checkpoint_bug.md +++ b/ALCF/notes/universal_checkpoint_bug.md @@ -2,33 +2,35 @@ ## Table of Contents -- [📓 Summary](#summary) -- [🚧 Issue](#issue) -- [🤔 Why is this Happening?](#-why-is-this-happening) -- [🔍 Running with Debugger](#-running-with-debugger) -- [🧰 Fix](#-fix) -- [✅ Confirm Fix Works](#-confirm-fix-works) -- [👻 Bug Doesn't Appear for Smaller Checkpoints](#-bug-doesnt-appear-for-smaller-checkpoints) + +1. [📓 Summary](#-summary) +1. [🚧 Issue](#-issue) + 1. [🔍 Running with Debugger](#-running-with-debugger) + 1. [🤔 Why is this Happening?](#-why-is-this-happening) + 1. [🧰 Proposed Fix](#-proposed-fix) + 1. [✅ Confirm Fix Works](#-confirm-fix-works) + 1. [👻 Bug Doesn't Appear for Smaller Checkpoints](#-bug-doesnt-appear-for-smaller-checkpoints) ## 📓 Summary -- Everything works _as is_ for checkpoints created on small scales (small DP ?) +- ✅ Everything works _as is_ for checkpoints created on small scales (small DP ?) - Explicitly confrm this - (see [👻 Bug Doesn't appear for smaller checkpoints](#bug-doesnt-appear-for-smaller-checkpoints)) + (see [👻 Bug Doesn't appear for smaller checkpoints](#-bug-doesnt-appear-for-smaller-checkpoints)) by: 1. Generate checkpoint from using 4 nodes of Aurora 2. Convert this checkpoint to universal without issue -- Trying to repeat this exact same process, but using a c checkpoint created - with `DP=768` (12 nodes of Aurora) fails with a `RuntimeError` - +- ❌ Trying to repeat this exact same process, but using a checkpoint created + with `DP=768` (12 nodes of Aurora) fails with a `RuntimeError`. -- ❌ `RuntimeError` when trying to convert checkpoint created with `DP=768` (12 nodes of Aurora) - - Description of bug - - 🧰 Proposed fix - - ✅ Confirmation of Fix + We then walk through: + - [🚧 A Description of bug](#-issue) + - [🔍 Running with Debugger](#-running-with-debugger) + - [🤔 Why is this Happening?](#-why-is-this-happening) + - [🧰 Proposed fix](#-proposed-fix) + - [✅ Confirmation of Fix](#-confirm-fix-works) ## 🚧 Issue @@ -102,31 +104,76 @@ Even more interesting, this only seems to happen for checkpoints created using more than 18 nodes of Aurora (which would correspond to a `data_parallel_size = 216`) --> -## 🤔 Why is this Happening? +### 🔍 Running with Debugger -The `extract_zero_shards` function [here] in `ds_to_universal.py` +Running with: -[`extract_zero_shards`](/ALCF/ds_to_universal.py#L132) function, the -`raw_length` is calculated as `raw_length = state_value.numel() - -group_paddings[key]`. +```bash +python3 -m pudb deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py \ + --input_folder "${ckpt_dir}/global_step${gs}" \ + --output_folder "${ckpt_dir}/global_step${gs}_universal" \ + --keep_temp_folder \ + --num_extract_workers 1 \ + --num_merge_workers 1 +``` -This is causing the `raw_length` to be negative in some -cases. This is causing the following error: +
Traceback +```python +Traceback (most recent call last): + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/__init__.py", line 176, in _runscript + dbg._runscript(mainpyfile) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/debugger.py", line 529, in _runscript + self.run(statement) + File "/opt/aurora/24.180.1/frameworks/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/bdb.py", line 598, in run + exec(cmd, globals, locals) + File "", line 1, in + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 549, in + main(args) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 499, in main + _extract_zero_shard_files(args, ds_checkpoint, temp_dir) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 370, in _extract_zero_shard_files + _do_parallel_work(do_work, _3d_range_list, args.num_extract_workers) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 359, in _do_parallel_work + results.append(do_work(work)) + File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 114, in extract_zero_shards + sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/deepspeed_checkpoint.py", line 124, in get_zero_checkpoint_state + return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 62, in get_state_for_rank + self._strip_tensor_paddings(sd) + File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 110, in _strip_tensor_paddings + group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +RuntimeError: narrow(): length must be non-negative. +``` + +
+ +Stepping through this command: ```python -def extract_zero_shards(dir, ds_checkpoint, indices_3D): - pp_index, tp_index, dp_index = indices_3D - sd = ds_checkpoint.get_zero_checkpoint_state( - pp_index=pp_index, - tp_index=tp_index, - dp_index=dp_index, - strip_tensor_paddings=False, - ) +>>> group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +Traceback (most recent call last): + File "", line 1, in +RuntimeError: narrow(): length must be non-negative. + +>>> raw_length +-676 +>>> state_value.numel() +348 +>>> group_paddings[key] +1024 +>>> key +1 +>>> state_name +'exp_avg' +>>> key +1 ``` +### 🤔 Why is this Happening? -In particular, in the call +The problematic line occurs here in [deepspeed / checkpoint / `ds_to_universal.py#L114`](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/ds_to_universal.py#L114), shown below: ```python sd = ds_checkpoint.get_zero_checkpoint_state( @@ -134,8 +181,8 @@ sd = ds_checkpoint.get_zero_checkpoint_state( ) ``` -this then calls the `DeepSpeedCheckpoint.get_zero_checkpoint_state` method -[here](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/deepspeed_checkpoint.py#L123-L127) +The `extract_zero_shards` function tries calling the `DeepSpeedCheckpoint.get_zero_checkpoint_state` method +here [deepspeed / `checkpoint.py#L123-127`](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/deepspeed_checkpoint.py#L123-L127), which looks like: ```python def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: @@ -145,8 +192,8 @@ def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: keys_to_ignore=[PARAM_SHAPES]) ``` -This is a call to the `ZeROCheckpoint.get_state_for_rank` method from -[here](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/zero_checkpoint.py#L53-L73) +This (^) then calls the `ZeROCheckpoint.get_state_for_rank` method from +[deepspeed / checkpoint / `zero_checkpoint.py#L53-73`](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/zero_checkpoint.py#L53-L73). Now, this `get_state_for_rank` function accepts an argument `strip_tensor_paddings` which is true by default. @@ -165,7 +212,7 @@ and we hit the `RuntimeError` when calling this Stepping into the `self._strip_tensor_paddings` method, we see that it calls [this block](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/zero_checkpoint.py#L108-L110) -which calculates the `raw_length` as: +(shown below) which calculates the `raw_length` as: ```python def _strip_tensor_paddings(self, sd): @@ -189,125 +236,77 @@ def _strip_tensor_paddings(self, sd): group_state[state_name] = state_value ``` -## 🔍 Running with Debugger - -Running with: - -```bash -python3 -m pudb deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py \ - --input_folder "${ckpt_dir}/global_step${gs}" \ - --output_folder "${ckpt_dir}/global_step${gs}_universal" \ - --keep_temp_folder \ - --num_extract_workers 1 \ - --num_merge_workers 1 -``` - -
Traceback +which, when `raw_length` is negative, causes: ```python -Traceback (most recent call last): - File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/__init__.py", line 176, in _runscript - dbg._runscript(mainpyfile) - File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/site-packages/pudb/debugger.py", line 529, in _runscript - self.run(statement) - File "/opt/aurora/24.180.1/frameworks/aurora_nre_models_frameworks-2024.2.1_u1/lib/python3.10/bdb.py", line 598, in run - exec(cmd, globals, locals) - File "", line 1, in - File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 549, in - main(args) - File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 499, in main - _extract_zero_shard_files(args, ds_checkpoint, temp_dir) - File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 370, in _extract_zero_shard_files - _do_parallel_work(do_work, _3d_range_list, args.num_extract_workers) - File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 359, in _do_parallel_work - results.append(do_work(work)) - File "deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py", line 114, in extract_zero_shards - sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) - File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/deepspeed_checkpoint.py", line 124, in get_zero_checkpoint_state - return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, - File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 62, in get_state_for_rank - self._strip_tensor_paddings(sd) - File "/lus/flare/projects/Aurora_deployment/foremans/projects/argonne-lcf/tmp/2024-12-28-154515/Megatron-DeepSpeed/deps/DeepSpeed/deepspeed/checkpoint/zero_checkpoint.py", line 110, in _strip_tensor_paddings - group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() +group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() RuntimeError: narrow(): length must be non-negative. ``` -
+It wasn't immediately obvious what this `strip_tensor_paddings` argument represents +(or even what the method is doing, to be honest), so I didn't have much insight +into why this would only be happening for checkpoints created at larger scales. -Stepping through this command: -```python ->>> group_state[state_name] = torch.narrow(state_value, 0, 0, raw_length).clone() -Traceback (most recent call last): - File "", line 1, in -RuntimeError: narrow(): length must be non-negative. +### 🧰 Proposed Fix ->>> raw_length --676 ->>> state_value.numel() -348 ->>> group_paddings[key] -1024 ->>> key -1 ->>> state_name -'exp_avg' ->>> key -1 -``` - -## 🧰 Fix - -Naively, the first thing I tried was to see if I could just skip this +Naively, the first (and easiest) thing to try was to see if I could just skip this `strip_tensor_paddings` step by setting `strip_tensor_paddings=False` in the -call to `get_zero_checkpoint_state` in the `extract_zero_shards` function, as -shown below: - -```diff -diff --git a/deepspeed/checkpoint/ds_to_universal.py b/deepspeed/checkpoint/ds_to_universal.py -index f7b75eee..cbbbef6b 100755 ---- a/deepspeed/checkpoint/ds_to_universal.py -+++ b/deepspeed/checkpoint/ds_to_universal.py -@@ -111,7 +111,7 @@ def _save_checkpoint(file_path, chkpt_sd): - - def extract_zero_shards(dir, ds_checkpoint, indices_3D): - pp_index, tp_index, dp_index = indices_3D -- sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) -+ sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index, strip_tensor_paddings=False) - - # pprint(f"Processing {dp_index=} {pp_index=}, {tp_index=}") -``` +call to (1) `get_zero_checkpoint_state` in the (2) `extract_zero_shards` function. + +Unfortunately, since (1) `DeepSpeedCheckpoint.get_zero_checkpoint_state()` +**DOES NOT** take in the `strip_tensor_paddings` argument, +there is no way to pass this along to the (2) `ZeROCheckpoint.get_state_for_rank()` call. + +So, our proposed fix requires two modifications: + +1. Modify `DeepSpeedCheckpoint.get_zero_checkpoint_state` signature from [here](https://github.com/microsoft/DeepSpeed/blob/cc03c76d57f41752d8cfb84c2e45b8e0da8083da/deepspeed/checkpoint/deepspeed_checkpoint.py#L123) to accept the `strip_tensor_paddings` argument: + + ```diff + warning: Empty last update token. + diff --git a/deepspeed/checkpoint/deepspeed_checkpoint.py b/deepspeed/checkpoint/deepspeed_checkpoint.py + index 31997177..a2ef5d0d 100644 + --- a/deepspeed/checkpoint/deepspeed_checkpoint.py + +++ b/deepspeed/checkpoint/deepspeed_checkpoint.py + @@ -120,11 +120,12 @@ class DeepSpeedCheckpoint(object): + self.global_state[ITERATION_KEY] = sd.get(ITERATION_KEY, 0) + self.global_state[ARGS_KEY] = sd.get(ARGS_KEY, None) + + - def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: + - return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, + + def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index, strip_tensor_paddings: bool = True) -> dict: + + return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, # type:ignore + tp_index=tp_index, + dp_index=dp_index, + - keys_to_ignore=[PARAM_SHAPES]) + + keys_to_ignore=[PARAM_SHAPES], + + strip_tensor_paddings=strip_tensor_paddings) + + def get_zero_files(self, pp_index, tp_index, dp_index) -> list: + return self.zero_checkpoint.get_files_for_rank(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) + ``` -But unfortunately, the `DeepSpeedCheckpoint` doesn't allow the -`strip_tensor_paddings` argument to be passed through to the -`get_zero_checkpoint_state` method, requiring an additional modification: - -```diff -warning: Empty last update token. -diff --git a/deepspeed/checkpoint/deepspeed_checkpoint.py b/deepspeed/checkpoint/deepspeed_checkpoint.py -index 31997177..a2ef5d0d 100644 ---- a/deepspeed/checkpoint/deepspeed_checkpoint.py -+++ b/deepspeed/checkpoint/deepspeed_checkpoint.py -@@ -120,11 +120,12 @@ class DeepSpeedCheckpoint(object): - self.global_state[ITERATION_KEY] = sd.get(ITERATION_KEY, 0) - self.global_state[ARGS_KEY] = sd.get(ARGS_KEY, None) - -- def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index) -> dict: -- return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, -+ def get_zero_checkpoint_state(self, pp_index, tp_index, dp_index, strip_tensor_paddings: bool = True) -> dict: -+ return self.zero_checkpoint.get_state_for_rank(pp_index=pp_index, # type:ignore - tp_index=tp_index, - dp_index=dp_index, -- keys_to_ignore=[PARAM_SHAPES]) -+ keys_to_ignore=[PARAM_SHAPES], -+ strip_tensor_paddings=strip_tensor_paddings) - - def get_zero_files(self, pp_index, tp_index, dp_index) -> list: - return self.zero_checkpoint.get_files_for_rank(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) -``` +1. With this in place, we can now try setting `strip_tensor_paddings = False` in the call shown below: + + ```diff + diff --git a/deepspeed/checkpoint/ds_to_universal.py b/deepspeed/checkpoint/ds_to_universal.py + index f7b75eee..cbbbef6b 100755 + --- a/deepspeed/checkpoint/ds_to_universal.py + +++ b/deepspeed/checkpoint/ds_to_universal.py + @@ -111,7 +111,7 @@ def _save_checkpoint(file_path, chkpt_sd): + + def extract_zero_shards(dir, ds_checkpoint, indices_3D): + pp_index, tp_index, dp_index = indices_3D + - sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index) + + sd = ds_checkpoint.get_zero_checkpoint_state(pp_index=pp_index, tp_index=tp_index, dp_index=dp_index, strip_tensor_paddings=False) + ``` ### ✅ Confirm Fix Works +We've added the proposed changes above to the `saforem2/ucp-bug` branch. + +We can confirm explicitly that the proposed fix works by retrying the conversion: + ```bash $ cd deps/DeepSpeed && git status && git checkout 'saforem2/ucp-bug' && PAGER='' git diff deepspeed/checkpoint/ && cd - && ckpt_dir=checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr0.00020_lwf0.05 ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder On branch saforem2/ucp-bug @@ -329,9 +328,12 @@ Converting DeepSpeed checkpoint in checkpoints/ws768_ds_stage1_nl32_hs4096_mb4_s took: 0h:09m:00s ``` +fixed! + ## 👻 Bug Doesn't Appear for Smaller Checkpoints -As a sanity check, we can explicitly test: +As a sanity check, we can explicitly test that everything works +_as is_ when converting smaller checkpoints to universal format. 1. Create checkpoint on 4 nodes of Aurora From 03da5714511319ba6a82be554732b32c55e6b014 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 Jan 2025 14:20:27 -0600 Subject: [PATCH 377/590] feat: Add `ALCF/examples/finetune_llama3/*` --- ALCF/examples/finetune_llama3/README.md | 24 +++ ALCF/examples/finetune_llama3/ds_config.json | 12 ++ .../finetune_llama3/ds_config_empty.json | 12 ++ .../finetune_llama3/finetune_llama.sh | 169 ++++++++++++++++++ 4 files changed, 217 insertions(+) create mode 100644 ALCF/examples/finetune_llama3/README.md create mode 100755 ALCF/examples/finetune_llama3/ds_config.json create mode 100755 ALCF/examples/finetune_llama3/ds_config_empty.json create mode 100644 ALCF/examples/finetune_llama3/finetune_llama.sh diff --git a/ALCF/examples/finetune_llama3/README.md b/ALCF/examples/finetune_llama3/README.md new file mode 100644 index 00000000000..7c81518fa4a --- /dev/null +++ b/ALCF/examples/finetune_llama3/README.md @@ -0,0 +1,24 @@ +## Example of Finetuning LLAMA-7B from Hugging Face Weights + +### Dataset +You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). + +### Pre-trained Weights +The pre-trained weights can be found at [Hugging Face - LLAMA-7B](https://huggingface.co/huggyllama/llama-7b). + +### Usage: + +#### 1. Converting Hugging Face Model Weights to Megatron-Deepspeed Model +```bash +bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh convert_hf2mds +``` +This command writes the Hugging Face model weights into the Megatron-Deepspeed model and saves it. You can adjust the parallel configuration in the script.```convert_mds2hf``` can convert a Megatron-Deepspeed model into the Hugging Face format + +#### 2. Fine-tuning Process +```bash +bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh +``` +Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). + + + diff --git a/ALCF/examples/finetune_llama3/ds_config.json b/ALCF/examples/finetune_llama3/ds_config.json new file mode 100755 index 00000000000..096ba9d2de9 --- /dev/null +++ b/ALCF/examples/finetune_llama3/ds_config.json @@ -0,0 +1,12 @@ +{ + "train_batch_size" : 96, + "train_micro_batch_size_per_gpu": 8, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": 0 + }, + "bf16": { + "enabled": true + } +} diff --git a/ALCF/examples/finetune_llama3/ds_config_empty.json b/ALCF/examples/finetune_llama3/ds_config_empty.json new file mode 100755 index 00000000000..f3e52853e19 --- /dev/null +++ b/ALCF/examples/finetune_llama3/ds_config_empty.json @@ -0,0 +1,12 @@ +{ + "train_batch_size" : 96, + "train_micro_batch_size_per_gpu": 8, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": 0 + }, + "bf16": { + "enabled": true + } +} diff --git a/ALCF/examples/finetune_llama3/finetune_llama.sh b/ALCF/examples/finetune_llama3/finetune_llama.sh new file mode 100644 index 00000000000..f0ba9499a8a --- /dev/null +++ b/ALCF/examples/finetune_llama3/finetune_llama.sh @@ -0,0 +1,169 @@ +DS_CONFIG=./examples_deepspeed/finetune_hf_llama3p3/ds_config.json +DS_CONFIG_EMPTY=./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json +DATASET_PATH=./examples_deepspeed/finetune_hf_llama3p3/alpaca_data.json +# dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json +HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct +# HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct +# HF_LLAMA_PATH=/home/foremans/.llama/checkpoints/Llama3.3-70B-Instruct +# HF_LLAMA_PATH=./Llama-2-7B-hf +PBS_O_WORKDIR=$(pwd) +source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) && ezpz_setup_env +ezpz_setup_env +# head -n 1 "${PBS_NODEFILE}" > nodefile-0 +#ezpz_setup_job nodefile-0 +# HF_LLAMA_PATH=/data/llama-2-7b-hf/ +# weights link: https://huggingface.co/huggyllama/llama-7b + +MICRO_BATCH_SIZE=8 +ZERO_STAGE=0 +# ${NGPUS}" +TP=4 +PP=1 +WORLD_SIZE="${NGPUS}" +# WORLD_SIZE="${TP}" +GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * WORLD_SIZE / (TP * PP))) +# require to align with weight dimensions +# HIDDEN_SIZE=4096 +# FFN_HIDDEN_SIZE=11008 +# NUM_LAYERS=32 +# NUM_HEADS=32 +HIDDEN_SIZE=8192 +FFN_HIDDEN_SIZE=28672 +NUM_LAYERS=80 +NUM_HEADS=64 +NUM_KV_HEADS=8 +SEQ_LENGTH=64 +###################################### + +MEGA_DS_LLAMA_PATH=./"llama-3p3-70B-mega-ds-T${TP}P${PP}" + +# Below configuration required for llama model as per llama paper +# --no-query-key-layer-scaling \ +# --attention-dropout 0 \ +# --hidden-dropout 0 \ +# --use-rotary-position-embeddings \ +# --untie-embeddings-and-output-weights \ +# --swiglu \ +# --normalization rmsnorm \ +# --disable-bias-linear \ +###################################### +cat < $DS_CONFIG +{ + "train_batch_size" : $GLOBAL_BATCH_SIZE, + "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": $ZERO_STAGE + }, + "bf16": { + "enabled": true + } +} +EOT + +cat < $DS_CONFIG_EMPTY +{ + "train_batch_size" : $GLOBAL_BATCH_SIZE, + "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": $ZERO_STAGE + }, + "bf16": { + "enabled": true + } +} +EOT + +if [ "$1" = "convert_hf2mds" ]; then + DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json" +elif [ "$1" = "convert_mds2hf" ]; then + DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json" +else + DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config.json" +fi + +# --hf-ckpt-num-shards 2 \ +covert_hf2mds_args="launch python3 tools/hf2megads_weight_converter.py \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ +--save $MEGA_DS_LLAMA_PATH" + +# --hf-ckpt-num-shards 2 \ +covert_mds2hf_args="launch python3 tools/hf2megads_weight_converter.py \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ +--to-hf-ckpt \ +--load $MEGA_DS_LLAMA_PATH \ +--save $HF_LLAMA_PATH'-hf-out' " + +finetune_args="launch python3 finetune_llama.py \ +--load $MEGA_DS_LLAMA_PATH" + +comm_args="--tensor-model-parallel-size $TP \ +--pipeline-model-parallel-size $PP \ +--lr-warmup-iters 2000 \ +--weight-decay 0.1 \ +--clip-grad 1 \ +--num-layers $NUM_LAYERS \ +--hidden-size $HIDDEN_SIZE \ +--num-attention-heads $NUM_HEADS \ +--finetune \ +--ffn-hidden-size $FFN_HIDDEN_SIZE \ +--num-key-value-heads $NUM_KV_HEADS \ +--attention-dropout 0 \ +--hidden-dropout 0 \ +--no-query-key-layer-scaling \ +--disable-bias-linear \ +--normalization rmsnorm \ +--use-rotary-position-embeddings \ +--untie-embeddings-and-output-weights \ +--swiglu \ +--seq-length $SEQ_LENGTH \ +--max-position-embeddings $SEQ_LENGTH \ +--micro-batch-size $MICRO_BATCH_SIZE \ +--global-batch-size $GLOBAL_BATCH_SIZE \ +--train-iters 3500 \ +--lr 2e-5 \ +--tensorboard-dir tensorboard_output \ +--lr-decay-iters 320000 \ +--lr-decay-style cosine \ +--log-interval 1 \ +--eval-iters 100 \ +--eval-interval 100 \ +--data-path $DATASET_PATH \ +--save-interval 1500 \ +--split 100,0,0 \ +--bf16 \ +--zero-stage $ZERO_STAGE \ +--tokenizer-type HFTokenizer \ +--tokenizer-model meta-llama/Llama-2-7B-hf \ +--deepspeed_config $DS_CONFIG_PATH \ +--deepspeed \ +--distributed-backend ccl \ +--num-workers 0 \ +--no-masked-softmax-fusion \ +--no-bias-gelu-fusion \ +--no-bias-dropout-fusion \ +--no-gradient-accumulation-fusion \ +--repeated-dataloader" + +# --tokenizer-type HFTokenizer \ +# --tokenizer-model 'file:///flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct' \ +# --tokenizer-model meta-llama/Llama-3-70B-Instruct \ +# --tokenizer-model ${HOME}/.llama/checkpoints/Llama3.3-70B-Instruct \ +# +if [ "$1" = "convert_hf2mds" ]; then + task_args="$covert_hf2mds_args" +elif [ "$1" = "convert_mds2hf" ]; then + task_args="$covert_mds2hf_args" +else + task_args="$finetune_args" +fi + +full_cmd="$task_args $comm_args" + +eval "$full_cmd" | tee finetune-llama-$(tstamp).log +wait $! From 19bdff098c49c832cea1e1e47e9299f28c4b632e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 Jan 2025 18:32:31 -0600 Subject: [PATCH 378/590] docs: Update `ALCF/examples/finetune_llama3/*` --- ALCF/examples/finetune_llama3/README.md | 11 +++++++++++ ALCF/examples/finetune_llama3/ds_config.json | 2 +- ALCF/examples/finetune_llama3/ds_config_empty.json | 2 +- ALCF/examples/finetune_llama3/finetune_llama.sh | 13 +++++++------ 4 files changed, 20 insertions(+), 8 deletions(-) diff --git a/ALCF/examples/finetune_llama3/README.md b/ALCF/examples/finetune_llama3/README.md index 7c81518fa4a..c027459a546 100644 --- a/ALCF/examples/finetune_llama3/README.md +++ b/ALCF/examples/finetune_llama3/README.md @@ -1,3 +1,14 @@ +# Finetune Llama3 from Hugging Face Checkpoint + +```bash +git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +cd Megatron-DeepSpeed +curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json +bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds +``` + + + ## Example of Finetuning LLAMA-7B from Hugging Face Weights ### Dataset diff --git a/ALCF/examples/finetune_llama3/ds_config.json b/ALCF/examples/finetune_llama3/ds_config.json index 096ba9d2de9..293e12494fb 100755 --- a/ALCF/examples/finetune_llama3/ds_config.json +++ b/ALCF/examples/finetune_llama3/ds_config.json @@ -1,5 +1,5 @@ { - "train_batch_size" : 96, + "train_batch_size" : 48, "train_micro_batch_size_per_gpu": 8, "steps_per_print": 1, "gradient_accumulation_steps": 1, diff --git a/ALCF/examples/finetune_llama3/ds_config_empty.json b/ALCF/examples/finetune_llama3/ds_config_empty.json index f3e52853e19..5abe8463016 100755 --- a/ALCF/examples/finetune_llama3/ds_config_empty.json +++ b/ALCF/examples/finetune_llama3/ds_config_empty.json @@ -1,5 +1,5 @@ { - "train_batch_size" : 96, + "train_batch_size" : 48, "train_micro_batch_size_per_gpu": 8, "steps_per_print": 1, "gradient_accumulation_steps": 1, diff --git a/ALCF/examples/finetune_llama3/finetune_llama.sh b/ALCF/examples/finetune_llama3/finetune_llama.sh index f0ba9499a8a..8604d448754 100644 --- a/ALCF/examples/finetune_llama3/finetune_llama.sh +++ b/ALCF/examples/finetune_llama3/finetune_llama.sh @@ -1,6 +1,6 @@ -DS_CONFIG=./examples_deepspeed/finetune_hf_llama3p3/ds_config.json -DS_CONFIG_EMPTY=./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json -DATASET_PATH=./examples_deepspeed/finetune_hf_llama3p3/alpaca_data.json +DS_CONFIG=./ALCF/examples/finetune_llama3/ds_config.json +DS_CONFIG_EMPTY=./ALCF/examples/finetune_llama3/ds_config_empty.json +DATASET_PATH=./ALCF/examples/finetune_llama3/alpaca_data.json # dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct # HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct @@ -78,11 +78,11 @@ cat < $DS_CONFIG_EMPTY EOT if [ "$1" = "convert_hf2mds" ]; then - DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" elif [ "$1" = "convert_mds2hf" ]; then - DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config_empty.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" else - DS_CONFIG_PATH="./examples_deepspeed/finetune_hf_llama3p3/ds_config.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config.json" fi # --hf-ckpt-num-shards 2 \ @@ -150,6 +150,7 @@ comm_args="--tensor-model-parallel-size $TP \ --no-gradient-accumulation-fusion \ --repeated-dataloader" +# --tokenizer-model meta-llama/Llama-2-7B-hf \ # --tokenizer-type HFTokenizer \ # --tokenizer-model 'file:///flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct' \ # --tokenizer-model meta-llama/Llama-3-70B-Instruct \ From e2cb2093f94c368169271da0354b7f255eae59ae Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 Jan 2025 20:03:49 -0600 Subject: [PATCH 379/590] chore: Update `tools/hf2megads_weight_converter.py` --- tools/hf2megads_weight_converter.py | 138 ++++++++++++++-------------- 1 file changed, 69 insertions(+), 69 deletions(-) mode change 100755 => 100644 tools/hf2megads_weight_converter.py diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py old mode 100755 new mode 100644 index 76d40ce2528..74a2f8c0db3 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -1,19 +1,17 @@ +import ezpz import torch import re import sys import os -import ezpz as ez - sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) import torch.distributed from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP from megatron import get_tokenizer, get_args -from megatron.tokenizer.tokenizer import _vocab_size_with_padding from megatron.core import mpu from megatron.core import tensor_parallel from megatron.core.utils import divide from megatron.model import GPTModelPipe, Float16Module -from megatron.utils import unwrap_model, get_logger +from megatron.utils import unwrap_model from megatron.model import DistributedDataParallel as LocalDDP from megatron.arguments import core_transformer_config_from_args from megatron.initialize import initialize_megatron @@ -25,8 +23,9 @@ import copy from pathlib import Path -RANK = ez.setup_torch('deepspeed') -log = get_logger(__name__, rank_zero_only=True) + +RANK = ezpz.setup_torch('deepspeed') +logger = ezpz.get_logger(__name__) def add_extra_args(parser): @@ -57,7 +56,7 @@ def compute_partition_range(hidden_size, local_rank, tp_size): def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. loaded = {} - log.info( + logger.info( f"----------------------------hf weight list----------------------------") for wid in range(1, hf_ckpt_num_of_shards + 1): @@ -65,7 +64,7 @@ def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): f"{hf_ckpt_dir}/pytorch_model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.bin", map_location=torch.device('cpu')) for k in d: - log.info(k) + logger.info(k) assert k not in loaded loaded[k] = d[k].clone() del d @@ -76,7 +75,7 @@ def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards) from safetensors import safe_open # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. hf_model = {} - log.info( + logger.info( f"----------------------------hf weight list----------------------------") for wid in range(1, hf_ckpt_num_of_shards + 1): @@ -87,7 +86,7 @@ def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards) with safe_open(ckpt_path, framework="pt", device="cpu") as f: for k in f.keys(): - log.info(f"name: {k}, shape: {f.get_tensor(k).shape}") + logger.info(f"name: {k}, shape: {f.get_tensor(k).shape}") assert k not in hf_model hf_model[k] = f.get_tensor(k).clone() @@ -105,18 +104,18 @@ def load_and_print_hf_weight_auto(hf_ckpt_dir, no_init=True): else: hf_model = {} hf_auto_model = AutoModelForCausalLM.from_pretrained(hf_ckpt_dir, trust_remote_code=True, torch_dtype=torch.bfloat16) - log.info( + logger.info( f"----------------------------hf weight list----------------------------") for name, param in hf_auto_model.named_parameters(): hf_model[name] = param.clone() - log.info(name) + logger.info(name) return hf_model def print_distinct_weights(model): - log.info( + logger.info( f"----------------------------mega-ds weight list----------------------------") for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): if mpu.get_pipeline_model_parallel_rank() == pipe_rank: @@ -159,11 +158,7 @@ def _embedding_refactor(self, pname, p): elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": hf_name = "model.embed_tokens.weight" hf_w = self.hf_model[hf_name] - log.info(f"{hf_w.shape[0]=}") - log.info(f"{self.token_vocab=}") - if hf_w.shape[0] != self.padded_vocab_size: - torch.distributed.breakpoint(0) - assert hf_w.shape[0] == self.padded_vocab_size + assert hf_w.shape[0] == self.token_vocab per_partition_vocab_size, start_index, end_index = compute_partition_range( self.padded_vocab_size, self.tp_rank, self.tp_size) end_index = min(end_index, self.token_vocab) @@ -179,6 +174,9 @@ def _embedding_refactor(self, pname, p): ) return new_w + + + def _direct_refactor(self, pname, p, hf_layer=None, subname=None): if pname == f"{self.mega_norm_wnum}.weight": hf_name = "model.norm.weight" @@ -199,38 +197,43 @@ def _qkv_refactor(self, pname, p, hf_layer): wk = self.hf_model[hf_wk_name] wv = self.hf_model[hf_wv_name] - hidden_size = wq.shape[0] - per_partition_size, start_index, end_index = compute_partition_range( - hidden_size, self.tp_rank, self.tp_size) - hidden_size_per_attention_head = divide(hidden_size, + query_hidden_size = wq.shape[0] + kv_hidden_size = wk.shape[0] + + per_partition_size, start_qindex, end_index = compute_partition_range( + query_hidden_size, self.tp_rank, self.tp_size) + _,start_kvindex, _= compute_partition_range( + kv_hidden_size, self.tp_rank, self.tp_size) + + hidden_size_per_attention_head = divide(query_hidden_size, self.config.num_attention_heads) num_attention_heads_per_partition = divide(self.config.num_attention_heads, self.tp_size) - new_w = torch.zeros((per_partition_size * 3, wq.shape[1]), dtype=wq.dtype) - - # >>> pname - # '2.self_attention.query_key_value.weight' - # >>> p.shap^U - # '2.self_attention.query_key_value.weight' - # >>> xp = p - # >>> xp.shape - # torch.Size([6144, 4096]) - for i in range(num_attention_heads_per_partition): - try: - current_index = start_index + i * hidden_size_per_attention_head - next_index = current_index + hidden_size_per_attention_head - new_w_index = i * (3 * hidden_size_per_attention_head) - new_w[new_w_index: new_w_index + (3 * hidden_size_per_attention_head), :] = \ - torch.cat([ - wq[current_index: next_index, :], - wk[current_index: next_index, :], - wv[current_index: next_index, :] - ], dim=0) - except Exception: - torch.distributed.breakpoint(0) + num_kv_heads_per_partition= divide(self.config.num_key_value_heads, + self.tp_size) + qkv_size=(num_attention_heads_per_partition+2*num_kv_heads_per_partition)*hidden_size_per_attention_head + num_qheads_per_group=divide(self.config.num_attention_heads,self.config.num_key_value_heads) + num_groups =divide(num_attention_heads_per_partition,num_qheads_per_group) + new_w = torch.zeros((qkv_size, wq.shape[1]), dtype=wq.dtype) + + for i in range(num_groups): + query_current_index=start_qindex+i*num_qheads_per_group*hidden_size_per_attention_head + query_next_index=query_current_index+num_qheads_per_group*hidden_size_per_attention_head + kv_current_index=start_kvindex+i*hidden_size_per_attention_head + kv_next_kvindex=kv_current_index+hidden_size_per_attention_head + + new_w_index=i* (num_qheads_per_group+2)*hidden_size_per_attention_head + + new_w[new_w_index:new_w_index+(num_qheads_per_group+2)*hidden_size_per_attention_head,:]=\ + torch.cat([ + wq[query_current_index:query_next_index,:], + wk[kv_current_index:kv_next_kvindex,:], + wv[kv_current_index:kv_next_kvindex,:] + ],dim=0) + self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_wq_name,hf_wk_name,hf_wv_name,} cat q,k,v [{current_index}:{next_index},:] of q,k,v{wq.shape}" + f"mega-ds:{pname,p.data.shape}<--hf{hf_wq_name,hf_wk_name,hf_wv_name,} cat q,k,v [{query_current_index}:{query_next_index},:] of q,k,v{wq.shape}" ) return new_w @@ -291,7 +294,7 @@ def _mlphto4h1_refactor(self, pname, p, hf_layer, subname): return new_w def transform_from_hf_to_megds(self): - assert not self.is_refactored + assert self.is_refactored == False new_w = None for pname, p in self.ds_model.named_parameters(): @@ -304,7 +307,6 @@ def transform_from_hf_to_megds(self): new_w = self._direct_refactor(pname, p) else: mobj = self.decoder_pat.match(pname) - assert mobj is not None layer_num = int(mobj.group(1)) subname = mobj.group(2) hf_layer = layer_num - self.offset_num @@ -333,7 +335,7 @@ def transform_from_hf_to_megds(self): new_w = None self.is_refactored = True - + def _embedding_refactor_to_hf(self, pname, ds_w): if pname == f"{self.mega_lm_head_wnum}.lm_head.weight": hf_w = self.hf_model.lm_head.weight @@ -344,7 +346,7 @@ def _embedding_refactor_to_hf(self, pname, ds_w): with torch.no_grad(): ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) - + self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank[:hf_w.shape[0], :]) def _direct_refactor_to_hf(self, pname, ds_w, hf_layer=None, subname=None): @@ -400,17 +402,18 @@ def _qkv_refactor_to_hf(self, pname, ds_w, hf_layer): hidden_size = oldshape[-1] hidden_size_per_attention_head = divide(hidden_size, self.config.num_attention_heads) - num_attention_heads_per_partition = divide(self.config.num_attention_heads, - self.tp_size) - newshape = (self.tp_size, num_attention_heads_per_partition, 3, hidden_size_per_attention_head, hidden_size) + # MHA & GQA + group = divide(self.config.num_attention_heads, self.config.num_key_value_heads) + newshape = (self.config.num_key_value_heads, group + 2, hidden_size_per_attention_head, hidden_size) ds_w_out = ds_w_all_rank.reshape(*newshape) - self.hf_dict[hf_q_name] = copy.deepcopy(ds_w_out[:, :, 0, :, :].reshape(-1, oldshape[-1])) - self.hf_dict[hf_k_name] = copy.deepcopy(ds_w_out[:, :, 1, :, :].reshape(-1, oldshape[-1])) - self.hf_dict[hf_v_name] = copy.deepcopy(ds_w_out[:, :, 2, :, :].reshape(-1, oldshape[-1])) + query_weight, key_weight, value_weight = torch.split(ds_w_out, [group, 1, 1], dim=1) + self.hf_dict[hf_q_name] = copy.deepcopy(query_weight.reshape(-1, hidden_size)) + self.hf_dict[hf_k_name] = copy.deepcopy(key_weight.reshape(-1, hidden_size)) + self.hf_dict[hf_v_name] = copy.deepcopy(value_weight.reshape(-1, hidden_size)) + del query_weight, key_weight, value_weight def transform_from_megads_to_hf(self): - use_gqa = True if self.num_attention_heads != self.num_key_value_heads else False for pname, p in self.ds_model.named_parameters(): if pname in [ @@ -428,11 +431,7 @@ def transform_from_megads_to_hf(self): subname = mobj.group(2) hf_layer = layer_num - self.offset_num if subname in ["self_attention.query_key_value.weight"]: - if not use_gqa: - self._qkv_refactor_to_hf(pname, p, hf_layer) - else: - #TODO(billishyahao): Not impl yet ... - assert False + self._qkv_refactor_to_hf(pname, p, hf_layer) elif subname in ["mlp.dense_h_to_4h.weight"]: self._mlphto4h_dense_refactor_to_hf(pname, p, hf_layer) elif subname in [ @@ -455,7 +454,7 @@ def record_mapping_info(self, record_msg): def inorder_show_record(self): assert self.is_refactored - log.info( + logger.info( f"----------------------------mapping list----------------------------") # print dp rank0 tp rank0 records. for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): @@ -483,7 +482,7 @@ def load_hf_weights(args, no_init): def convert_ckpt(): """Build the model.""" args = get_args() - log.info(f'building model ...') + logger.info(f'building model ...') see_memory_usage(f"Before Building Model", force=True) config = core_transformer_config_from_args(args) @@ -508,18 +507,18 @@ def convert_ckpt(): # print_distinct_weights(hf_model) #init model and save - log.info(f"before deepspeed init") + logger.info(f"before deepspeed init") ds_engine, _, _, _ = deepspeed.initialize( model=ds_model, optimizer=None, args=args, lr_scheduler=None, mpu=mpu if args.no_pipeline_parallel else None) - log.info(f"after deepspeed init") + logger.info(f"after deepspeed init") if args.to_hf_ckpt: load_checkpoint([ds_engine], None, None, load_only_weights=True) - log.info(f"completed to load deepspeed actual checkpoint") + logger.info(f"completed to load deepspeed actual checkpoint") # refactor weight from hf to mega-ds and vice versa @@ -540,7 +539,7 @@ def convert_ckpt(): if torch.distributed.is_initialized(): torch.distributed.barrier() - log.info(f"hf checkpoint will be saved in {save_path}/release ") + logger.info(f"hf checkpoint will be saved in {save_path}/release ") if mpu.is_pipeline_last_stage(): ## doing checkpoint merging and saving... # hf_model.tie_weights() @@ -556,9 +555,10 @@ def convert_ckpt(): # mega-ds checkpoint will be saved in args.save hf_model.save_pretrained(os.path.join(save_path, "release"), safe_serialization=True) else: - log.info(f"mega-ds checkpoint will be saved in {args.save}") + logger.info(f"mega-ds checkpoint will be saved in {args.save}") save_checkpoint(0, [ds_engine], None, None) - log.info(f"save checkpoint completed") + + logger.info(f"save checkpoint completed") if __name__ == "__main__": From a86878837d4cc8c1f29c700f9e3adc99c2def2f8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 14 Jan 2025 21:08:38 -0600 Subject: [PATCH 380/590] feat: Add `ALCF/examples/finetune_llama3p2_1B/*` --- ALCF/examples/finetune_llama3p2_1B/README.md | 35 ++++ .../finetune_llama3p2_1B/ds_config.json | 12 ++ .../finetune_llama3p2_1B/ds_config_empty.json | 12 ++ .../finetune_llama3p2_1B/finetune_llama.sh | 185 ++++++++++++++++++ 4 files changed, 244 insertions(+) create mode 100644 ALCF/examples/finetune_llama3p2_1B/README.md create mode 100755 ALCF/examples/finetune_llama3p2_1B/ds_config.json create mode 100755 ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json create mode 100644 ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh diff --git a/ALCF/examples/finetune_llama3p2_1B/README.md b/ALCF/examples/finetune_llama3p2_1B/README.md new file mode 100644 index 00000000000..c027459a546 --- /dev/null +++ b/ALCF/examples/finetune_llama3p2_1B/README.md @@ -0,0 +1,35 @@ +# Finetune Llama3 from Hugging Face Checkpoint + +```bash +git clone https://github.com/argonne-lcf/Megatron-DeepSpeed +cd Megatron-DeepSpeed +curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json +bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds +``` + + + +## Example of Finetuning LLAMA-7B from Hugging Face Weights + +### Dataset +You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). + +### Pre-trained Weights +The pre-trained weights can be found at [Hugging Face - LLAMA-7B](https://huggingface.co/huggyllama/llama-7b). + +### Usage: + +#### 1. Converting Hugging Face Model Weights to Megatron-Deepspeed Model +```bash +bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh convert_hf2mds +``` +This command writes the Hugging Face model weights into the Megatron-Deepspeed model and saves it. You can adjust the parallel configuration in the script.```convert_mds2hf``` can convert a Megatron-Deepspeed model into the Hugging Face format + +#### 2. Fine-tuning Process +```bash +bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh +``` +Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). + + + diff --git a/ALCF/examples/finetune_llama3p2_1B/ds_config.json b/ALCF/examples/finetune_llama3p2_1B/ds_config.json new file mode 100755 index 00000000000..bad063119dc --- /dev/null +++ b/ALCF/examples/finetune_llama3p2_1B/ds_config.json @@ -0,0 +1,12 @@ +{ + "train_batch_size" : 64, + "train_micro_batch_size_per_gpu": 8, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": 0 + }, + "bf16": { + "enabled": true + } +} diff --git a/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json b/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json new file mode 100755 index 00000000000..ca14161f24c --- /dev/null +++ b/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json @@ -0,0 +1,12 @@ +{ + "train_batch_size" : 64, + "train_micro_batch_size_per_gpu": 8, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": 0 + }, + "bf16": { + "enabled": true + } +} diff --git a/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh b/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh new file mode 100644 index 00000000000..85a911dfb32 --- /dev/null +++ b/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh @@ -0,0 +1,185 @@ +DS_CONFIG=./ALCF/examples/finetune_llama3p2_1B/ds_config.json +DS_CONFIG_EMPTY=./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json +DATASET_PATH=./ALCF/examples/finetune_llama3p2_1B/alpaca_data.json +# dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json +# HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct +# HF_LLAMA_PATH=/home/foremans/.llama/checkpoints/Llama3.3-70B-Instruct +# HF_LLAMA_PATH=./Llama-2-7B-hf +PBS_O_WORKDIR=$(pwd) +source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) && ezpz_setup_env +ezpz_setup_env +# head -n 1 "${PBS_NODEFILE}" > nodefile-0 +#ezpz_setup_job nodefile-0 +# HF_LLAMA_PATH=/data/llama-2-7b-hf/ +# weights link: https://huggingface.co/huggyllama/llama-7b +# + +machine_name=$(ezpz_get_machine_name) +if [[ "${machine_name}" == "aurora" ]]; then + BACKEND="ccl" + HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct +elif [[ "${machine_name}" == "polaris" ]]; then + BACKEND="nccl" + HF_LLAMA_PATH="Llama-3.2-1B" +fi + +HF_CONFIG="${HF_LLAMA_PATH}/config.json" + +MICRO_BATCH_SIZE=8 +ZERO_STAGE=0 +# ${NGPUS}" +TP=1 +PP=1 +WORLD_SIZE="${NGPUS}" +# WORLD_SIZE="${TP}" +GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * WORLD_SIZE / (TP * PP))) +# require to align with weight dimensions +# HIDDEN_SIZE=4096 +# FFN_HIDDEN_SIZE=11008 +# NUM_LAYERS=32 +# NUM_HEADS=32 +HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.hidden_size') +FFN_HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.intermediate_size') +NUM_LAYERS=$(cat "${HF_CONFIG}" | jq -r '.num_hidden_layers') +NUM_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') +NUM_KV_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') +SEQ_LENGTH=512 +###################################### + +MEGA_DS_LLAMA_PATH=./"llama-3p2-1B-mega-ds-T${TP}P${PP}" + +# Below configuration required for llama model as per llama paper +# --no-query-key-layer-scaling \ +# --attention-dropout 0 \ +# --hidden-dropout 0 \ +# --use-rotary-position-embeddings \ +# --untie-embeddings-and-output-weights \ +# --swiglu \ +# --normalization rmsnorm \ +# --disable-bias-linear \ +###################################### +cat < $DS_CONFIG +{ + "train_batch_size" : $GLOBAL_BATCH_SIZE, + "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": $ZERO_STAGE + }, + "bf16": { + "enabled": true + } +} +EOT + +cat < $DS_CONFIG_EMPTY +{ + "train_batch_size" : $GLOBAL_BATCH_SIZE, + "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, + "steps_per_print": 1, + "gradient_accumulation_steps": 1, + "zero_optimization": { + "stage": $ZERO_STAGE + }, + "bf16": { + "enabled": true + } +} +EOT + +if [ "$1" = "convert_hf2mds" ]; then + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json" +elif [ "$1" = "convert_mds2hf" ]; then + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json" +else + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config.json" +fi + +# --hf-ckpt-num-shards 2 \ +covert_hf2mds_args="launch python3 tools/hf2megads_weight_converter.py \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ +--save $MEGA_DS_LLAMA_PATH" + +# --hf-ckpt-num-shards 2 \ +covert_mds2hf_args="launch python3 tools/hf2megads_weight_converter.py \ +--hf-ckpt-dir $HF_LLAMA_PATH \ +--load-mode auto \ +--to-hf-ckpt \ +--load $MEGA_DS_LLAMA_PATH \ +--save $HF_LLAMA_PATH'-hf-out' " + +finetune_args="launch python3 finetune_llama.py \ +--load $MEGA_DS_LLAMA_PATH" + +comm_args="--tensor-model-parallel-size $TP \ +--pipeline-model-parallel-size $PP \ +--lr-warmup-iters 2000 \ +--weight-decay 0.1 \ +--clip-grad 1 \ +--num-layers $NUM_LAYERS \ +--hidden-size $HIDDEN_SIZE \ +--num-attention-heads $NUM_HEADS \ +--finetune \ +--ffn-hidden-size $FFN_HIDDEN_SIZE \ +--num-key-value-heads $NUM_KV_HEADS \ +--attention-dropout 0 \ +--hidden-dropout 0 \ +--no-query-key-layer-scaling \ +--disable-bias-linear \ +--normalization rmsnorm \ +--use-rotary-position-embeddings \ +--untie-embeddings-and-output-weights \ +--swiglu \ +--seq-length $SEQ_LENGTH \ +--max-position-embeddings $SEQ_LENGTH \ +--micro-batch-size $MICRO_BATCH_SIZE \ +--global-batch-size $GLOBAL_BATCH_SIZE \ +--train-iters 3500 \ +--lr 2e-5 \ +--tensorboard-dir tensorboard_output \ +--lr-decay-iters 320000 \ +--lr-decay-style cosine \ +--log-interval 1 \ +--eval-iters 100 \ +--eval-interval 100 \ +--data-path $DATASET_PATH \ +--save-interval 1500 \ +--split 100,0,0 \ +--bf16 \ +--zero-stage $ZERO_STAGE \ +--tokenizer-type HFTokenizer \ +--tokenizer-model meta-llama/Llama-2-7B-hf \ +--deepspeed_config $DS_CONFIG_PATH \ +--deepspeed \ +--distributed-backend $BACKEND \ +--num-workers 0 \ +--no-masked-softmax-fusion \ +--no-bias-gelu-fusion \ +--no-bias-dropout-fusion \ +--no-gradient-accumulation-fusion \ +--repeated-dataloader" + +# --tokenizer-model meta-llama/Llama-2-7B-hf \ +# --tokenizer-type HFTokenizer \ +# --tokenizer-model 'file:///flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct' \ +# --tokenizer-model meta-llama/Llama-3-70B-Instruct \ +# --tokenizer-model ${HOME}/.llama/checkpoints/Llama3.3-70B-Instruct \ +# +if [ "$1" = "convert_hf2mds" ]; then + task_args="$covert_hf2mds_args" +elif [ "$1" = "convert_mds2hf" ]; then + task_args="$covert_mds2hf_args" +else + task_args="$finetune_args" +fi + +full_cmd="$task_args $comm_args" + +OUTFILE="finetune-llama-$(tstamp).log" +printf "task_args: %s\n" $task_args[@] | tee "${OUTFILE}" +printf "comm_args: %s\n" $comm_args[@] | tee -a "${OUTFILE}" +printf "full_cmd: %s\n" $full_cmd[@] | tee -a "${OUTFILE}" +eval "$full_cmd" | tee -a "${OUTFILE}" +wait $! From babef03d5d2c436a89376246ba795225078a78af Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 15 Jan 2025 08:34:09 -0600 Subject: [PATCH 381/590] feat: Update `ALCF/examples/finetune_llama3/*` --- ALCF/examples/finetune_llama3/README.md | 34 +- ALCF/examples/finetune_llama3/ds_config.json | 8 +- .../finetune_llama3/ds_config_empty.json | 8 +- .../finetune_llama3/finetune_llama.sh | 216 +++--- megatron/checkpointing.py | 1 + tools/hf2megads_weight_converter.py | 627 +++++++++++------- 6 files changed, 565 insertions(+), 329 deletions(-) diff --git a/ALCF/examples/finetune_llama3/README.md b/ALCF/examples/finetune_llama3/README.md index c027459a546..b860a640736 100644 --- a/ALCF/examples/finetune_llama3/README.md +++ b/ALCF/examples/finetune_llama3/README.md @@ -1,35 +1,39 @@ # Finetune Llama3 from Hugging Face Checkpoint -```bash -git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -cd Megatron-DeepSpeed -curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json -bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds -``` +1. Download HF Checkpoint: + ```bash + MODEL="Llama-3.2-1B" + HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download "meta-llama/${MODEL}" --local-dir "${MODEL}" + ``` +1. Convert HF --> MDS: -## Example of Finetuning LLAMA-7B from Hugging Face Weights + ```bash + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed + cd Megatron-DeepSpeed + curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json + TP=1 PP=1 ZERO_STAGE=1 MODEL_NAME=Llama-3.2-1B bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds + ``` -### Dataset -You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). +## Dataset -### Pre-trained Weights -The pre-trained weights can be found at [Hugging Face - LLAMA-7B](https://huggingface.co/huggyllama/llama-7b). +You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). -### Usage: +### Usage #### 1. Converting Hugging Face Model Weights to Megatron-Deepspeed Model + ```bash bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh convert_hf2mds ``` + This command writes the Hugging Face model weights into the Megatron-Deepspeed model and saves it. You can adjust the parallel configuration in the script.```convert_mds2hf``` can convert a Megatron-Deepspeed model into the Hugging Face format #### 2. Fine-tuning Process + ```bash bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh ``` -Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). - - +Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). diff --git a/ALCF/examples/finetune_llama3/ds_config.json b/ALCF/examples/finetune_llama3/ds_config.json index 293e12494fb..64d559a27fb 100755 --- a/ALCF/examples/finetune_llama3/ds_config.json +++ b/ALCF/examples/finetune_llama3/ds_config.json @@ -3,8 +3,14 @@ "train_micro_batch_size_per_gpu": 8, "steps_per_print": 1, "gradient_accumulation_steps": 1, + "optimizer": { + "type": "Adam", + "params": { + "lr": 1e-4 + } + }, "zero_optimization": { - "stage": 0 + "stage": 1 }, "bf16": { "enabled": true diff --git a/ALCF/examples/finetune_llama3/ds_config_empty.json b/ALCF/examples/finetune_llama3/ds_config_empty.json index 5abe8463016..e8ddb534a84 100755 --- a/ALCF/examples/finetune_llama3/ds_config_empty.json +++ b/ALCF/examples/finetune_llama3/ds_config_empty.json @@ -3,8 +3,14 @@ "train_micro_batch_size_per_gpu": 8, "steps_per_print": 1, "gradient_accumulation_steps": 1, + "optimizer": { + "type": "Adam", + "params": { + "lr": 1e-4 + } + }, "zero_optimization": { - "stage": 0 + "stage": 1 }, "bf16": { "enabled": true diff --git a/ALCF/examples/finetune_llama3/finetune_llama.sh b/ALCF/examples/finetune_llama3/finetune_llama.sh index 8604d448754..d3509caccf9 100644 --- a/ALCF/examples/finetune_llama3/finetune_llama.sh +++ b/ALCF/examples/finetune_llama3/finetune_llama.sh @@ -1,41 +1,74 @@ +PBS_O_WORKDIR=$(pwd) +source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) && ezpz_setup_env + DS_CONFIG=./ALCF/examples/finetune_llama3/ds_config.json DS_CONFIG_EMPTY=./ALCF/examples/finetune_llama3/ds_config_empty.json -DATASET_PATH=./ALCF/examples/finetune_llama3/alpaca_data.json +DATASET_PATH="./dataset/alpaca_data.json" +# DATASET_PATH=./ALCF/examples/finetune_llama3/alpaca_data.json # dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json -HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct # HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct # HF_LLAMA_PATH=/home/foremans/.llama/checkpoints/Llama3.3-70B-Instruct # HF_LLAMA_PATH=./Llama-2-7B-hf -PBS_O_WORKDIR=$(pwd) -source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) && ezpz_setup_env -ezpz_setup_env # head -n 1 "${PBS_NODEFILE}" > nodefile-0 #ezpz_setup_job nodefile-0 # HF_LLAMA_PATH=/data/llama-2-7b-hf/ # weights link: https://huggingface.co/huggyllama/llama-7b +# + +MODEL_NAME="${MODEL_NAME:-"Llama-3.3-70B-Instruct"}" -MICRO_BATCH_SIZE=8 -ZERO_STAGE=0 -# ${NGPUS}" -TP=4 -PP=1 -WORLD_SIZE="${NGPUS}" -# WORLD_SIZE="${TP}" +machine_name=$(ezpz_get_machine_name) +if [[ "${machine_name}" == "aurora" ]]; then + BACKEND="ccl" + HF_LLAMA_PATH="${MODEL_NAME}" + # HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct +elif [[ "${machine_name}" == "polaris" ]]; then + BACKEND="nccl" + # HF_LLAMA_PATH="Llama-3.2-1B" + # HF_LLAMA_PATH=Llama-3.3-70B-Instruct +fi + +HF_LLAMA_PATH="${MODEL_NAME}" +HF_CONFIG="${HF_LLAMA_PATH}/config.json" + +MICRO_BATCH_SIZE="${MICRO_BATCH_SIZE:-8}" +ZERO_STAGE="${ZERO_STAGE:-0}" +TP="${TP:-1}" +PP="${PP:-1}" +WORLD_SIZE="${WORLD_SIZE:-${NGPUS}}" GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * WORLD_SIZE / (TP * PP))) # require to align with weight dimensions # HIDDEN_SIZE=4096 # FFN_HIDDEN_SIZE=11008 # NUM_LAYERS=32 # NUM_HEADS=32 -HIDDEN_SIZE=8192 -FFN_HIDDEN_SIZE=28672 -NUM_LAYERS=80 -NUM_HEADS=64 -NUM_KV_HEADS=8 -SEQ_LENGTH=64 +echo "Model: ${MODEL_NAME}" +echo "HF_CONFIG: ${HF_CONFIG}" +cat "${HF_CONFIG}" | jq . +HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.hidden_size') +FFN_HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.intermediate_size') +NUM_LAYERS=$(cat "${HF_CONFIG}" | jq -r '.num_hidden_layers') +NUM_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') +NUM_KV_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_key_value_heads') +MAX_SEQ_LENGTH=$(cat "${HF_CONFIG}" | jq -r '.max_position_embeddings') +SEQ_LENGTH=512 ###################################### -MEGA_DS_LLAMA_PATH=./"llama-3p3-70B-mega-ds-T${TP}P${PP}" +printf "GLOBAL_BATCH_SIZE: %s\n" $GLOBAL_BATCH_SIZE +printf "MICRO_BATCH_SIZE: %s\n" $MICRO_BATCH_SIZE +printf "ZERO_STAGE: %s\n" $ZERO_STAGE +printf "TP: %s\n" $TP +printf "PP: %s\n" $PP +printf "WORLD_SIZE: %s\n" $WORLD_SIZE +printf "HIDDEN_SIZE: %s\n" $HIDDEN_SIZE +printf "FFN_HIDDEN_SIZE: %s\n" $FFN_HIDDEN_SIZE +printf "NUM_LAYERS: %s\n" $NUM_LAYERS +printf "NUM_HEADS: %s\n" $NUM_HEADS +printf "NUM_KV_HEADS: %s\n" $NUM_KV_HEADS +printf "SEQ_LENGTH: %s\n" $SEQ_LENGTH + +MEGA_DS_LLAMA_PATH="converted_hf_ckpts/${MODEL_NAME}-MDS-GBS${GLOBAL_BATCH_SIZE}-ZS${ZERO_STAGE}-TP${TP}-PP${PP}" +mkdir -p $(dirname $MEGA_DS_LLAMA_PATH) # Below configuration required for llama model as per llama paper # --no-query-key-layer-scaling \ @@ -47,12 +80,18 @@ MEGA_DS_LLAMA_PATH=./"llama-3p3-70B-mega-ds-T${TP}P${PP}" # --normalization rmsnorm \ # --disable-bias-linear \ ###################################### -cat < $DS_CONFIG +cat <$DS_CONFIG { "train_batch_size" : $GLOBAL_BATCH_SIZE, "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, "steps_per_print": 1, "gradient_accumulation_steps": 1, + "optimizer": { + "type": "Adam", + "params": { + "lr": 1e-4 + } + }, "zero_optimization": { "stage": $ZERO_STAGE }, @@ -62,12 +101,18 @@ cat < $DS_CONFIG } EOT -cat < $DS_CONFIG_EMPTY +cat <$DS_CONFIG_EMPTY { "train_batch_size" : $GLOBAL_BATCH_SIZE, "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, "steps_per_print": 1, "gradient_accumulation_steps": 1, + "optimizer": { + "type": "Adam", + "params": { + "lr": 1e-4 + } + }, "zero_optimization": { "stage": $ZERO_STAGE }, @@ -78,78 +123,81 @@ cat < $DS_CONFIG_EMPTY EOT if [ "$1" = "convert_hf2mds" ]; then - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" elif [ "$1" = "convert_mds2hf" ]; then - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config_empty.json" else - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config.json" + DS_CONFIG_PATH="./ALCF/examples/finetune_llama3/ds_config.json" fi # --hf-ckpt-num-shards 2 \ covert_hf2mds_args="launch python3 tools/hf2megads_weight_converter.py \ ---hf-ckpt-dir $HF_LLAMA_PATH \ +--hf-ckpt-dir ${HF_LLAMA_PATH} \ --load-mode auto \ ---save $MEGA_DS_LLAMA_PATH" +--save ${MEGA_DS_LLAMA_PATH}" # --hf-ckpt-num-shards 2 \ covert_mds2hf_args="launch python3 tools/hf2megads_weight_converter.py \ ---hf-ckpt-dir $HF_LLAMA_PATH \ +--hf-ckpt-dir ${HF_LLAMA_PATH} \ --load-mode auto \ --to-hf-ckpt \ ---load $MEGA_DS_LLAMA_PATH \ ---save $HF_LLAMA_PATH'-hf-out' " +--load ${MEGA_DS_LLAMA_PATH} \ +--save ${HF_LLAMA_PATH}'-hf-out' " finetune_args="launch python3 finetune_llama.py \ ---load $MEGA_DS_LLAMA_PATH" - -comm_args="--tensor-model-parallel-size $TP \ ---pipeline-model-parallel-size $PP \ ---lr-warmup-iters 2000 \ ---weight-decay 0.1 \ ---clip-grad 1 \ ---num-layers $NUM_LAYERS \ ---hidden-size $HIDDEN_SIZE \ ---num-attention-heads $NUM_HEADS \ ---finetune \ ---ffn-hidden-size $FFN_HIDDEN_SIZE \ ---num-key-value-heads $NUM_KV_HEADS \ ---attention-dropout 0 \ ---hidden-dropout 0 \ ---no-query-key-layer-scaling \ ---disable-bias-linear \ ---normalization rmsnorm \ ---use-rotary-position-embeddings \ ---untie-embeddings-and-output-weights \ ---swiglu \ ---seq-length $SEQ_LENGTH \ ---max-position-embeddings $SEQ_LENGTH \ ---micro-batch-size $MICRO_BATCH_SIZE \ ---global-batch-size $GLOBAL_BATCH_SIZE \ ---train-iters 3500 \ ---lr 2e-5 \ ---tensorboard-dir tensorboard_output \ ---lr-decay-iters 320000 \ ---lr-decay-style cosine \ ---log-interval 1 \ ---eval-iters 100 \ ---eval-interval 100 \ ---data-path $DATASET_PATH \ ---save-interval 1500 \ ---split 100,0,0 \ ---bf16 \ ---zero-stage $ZERO_STAGE \ ---tokenizer-type HFTokenizer \ ---tokenizer-model meta-llama/Llama-2-7B-hf \ ---deepspeed_config $DS_CONFIG_PATH \ ---deepspeed \ ---distributed-backend ccl \ ---num-workers 0 \ ---no-masked-softmax-fusion \ ---no-bias-gelu-fusion \ ---no-bias-dropout-fusion \ ---no-gradient-accumulation-fusion \ ---repeated-dataloader" +--load ${MEGA_DS_LLAMA_PATH}" + +comm_args+=( + "--tensor-model-parallel-size=${TP}" + "--pipeline-model-parallel-size=${PP}" + "--lr-warmup-iters=2000" + "--weight-decay=0.1" + "--clip-grad=1" + "--num-layers=${NUM_LAYERS}" + "--hidden-size=${HIDDEN_SIZE}" + "--num-attention-heads=${NUM_HEADS}" + "--finetune" + "--ffn-hidden-size=${FFN_HIDDEN_SIZE}" + "--num-key-value-heads=${NUM_KV_HEADS}" + "--attention-dropout=0" + "--hidden-dropout=0" + "--no-query-key-layer-scaling" + "--disable-bias-linear" + "--normalization=rmsnorm" + "--use-rotary-position-embeddings" + "--untie-embeddings-and-output-weights" + "--swiglu" + "--seq-length=${SEQ_LENGTH}" + "--max-position-embeddings=${MAX_SEQ_LENGTH}" + "--micro-batch-size=${MICRO_BATCH_SIZE}" + "--global-batch-size=${GLOBAL_BATCH_SIZE}" + "--train-iters=3500" + "--lr=2e-5" + "--tensorboard-dir=tensorboard_output" + "--lr-decay-iters=320000" + "--lr-decay-style=cosine" + "--log-interval=1" + "--eval-iters=100" + "--eval-interval=100" + "--data-path=${DATASET_PATH}" + "--save-interval=1500" + "--split=100,0,0" + "--bf16" + "--zero-stage=${ZERO_STAGE}" + "--tokenizer-type=HFTokenizer" + "--tokenizer-model=meta-llama/${MODEL_NAME}" + "--deepspeed_config=${DS_CONFIG_PATH}" + "--deepspeed" + "--distributed-backend=$BACKEND" + "--num-workers=0" + "--no-masked-softmax-fusion" + "--no-bias-gelu-fusion" + "--no-bias-dropout-fusion" + "--no-gradient-accumulation-fusion" + "--repeated-dataloader" +) +# "--optimizer=adamw" # --tokenizer-model meta-llama/Llama-2-7B-hf \ # --tokenizer-type HFTokenizer \ # --tokenizer-model 'file:///flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct' \ @@ -157,14 +205,16 @@ comm_args="--tensor-model-parallel-size $TP \ # --tokenizer-model ${HOME}/.llama/checkpoints/Llama3.3-70B-Instruct \ # if [ "$1" = "convert_hf2mds" ]; then - task_args="$covert_hf2mds_args" + task_args="$covert_hf2mds_args" elif [ "$1" = "convert_mds2hf" ]; then - task_args="$covert_mds2hf_args" + task_args="$covert_mds2hf_args" else - task_args="$finetune_args" + task_args="$finetune_args" fi -full_cmd="$task_args $comm_args" +full_cmd="$task_args ${comm_args[*]}" -eval "$full_cmd" | tee finetune-llama-$(tstamp).log +OUTFILE="finetune-llama-$(tstamp).log" +printf "full_cmd: %s\n" "${full_cmd}" | tee -a "${OUTFILE}" +eval "$full_cmd" | tee -a "${OUTFILE}" wait $! diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index a4f82ec9d3a..f420795f522 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -586,6 +586,7 @@ def save_lr_state_dict() -> None: lr_state_dict_fp = Path(args.save).joinpath( "lr_state_dict.yaml" ) + lr_state_dict_fp.parent.mkdir(parents=True, exist_ok=True) log.info(f"Saving lr_state_dict to {lr_state_dict_fp.as_posix()}") with lr_state_dict_fp.open('w') as f: yaml.dump( diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py index 74a2f8c0db3..f84f6faf4e7 100644 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -3,46 +3,61 @@ import re import sys import os + sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) import torch.distributed -from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP + +# from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP from megatron import get_tokenizer, get_args from megatron.core import mpu from megatron.core import tensor_parallel from megatron.core.utils import divide -from megatron.model import GPTModelPipe, Float16Module -from megatron.utils import unwrap_model -from megatron.model import DistributedDataParallel as LocalDDP +from megatron.model import GPTModelPipe # , Float16Module + +# from megatron.utils import unwrap_model +# from megatron.model import DistributedDataParallel as LocalDDP from megatron.arguments import core_transformer_config_from_args from megatron.initialize import initialize_megatron -from megatron.optimizer import get_megatron_optimizer + +# from megatron.optimizer import get_megatron_optimizer from megatron.checkpointing import save_checkpoint, load_checkpoint -from megatron.training import get_optimizer_param_scheduler + +# from megatron.training import get_optimizer_param_scheduler from deepspeed.runtime.utils import see_memory_usage import deepspeed import copy from pathlib import Path -RANK = ezpz.setup_torch('deepspeed') +try: + RANK = ezpz.setup_torch('deepspeed') +except Exception: + RANK = 0 logger = ezpz.get_logger(__name__) def add_extra_args(parser): """Text generation arguments.""" group = parser.add_argument_group(title='hf2mega') - group.add_argument("--hf-ckpt-dir", - type=str, - default="", - help="the llama-hf ckpt") - group.add_argument("--hf-ckpt-num-shards", type=int, default=-1, help='num of llama ckpt.') - group.add_argument("--load-mode", type=str, - default=None, - choices=['torchbin', 'safetensor', 'auto'], - help="load ckpt format: pytorch.bin or model.safetensor or auto.") - group.add_argument("--to-hf-ckpt", action="store_true", - help="by default convert from hf to megads" - "if set, convert reversely from megads to hf ckpt.") + group.add_argument( + '--hf-ckpt-dir', type=str, default='', help='the llama-hf ckpt' + ) + group.add_argument( + '--hf-ckpt-num-shards', type=int, default=-1, help='num of llama ckpt.' + ) + group.add_argument( + '--load-mode', + type=str, + default=None, + choices=['torchbin', 'safetensor', 'auto'], + help='load ckpt format: pytorch.bin or model.safetensor or auto.', + ) + group.add_argument( + '--to-hf-ckpt', + action='store_true', + help='by default convert from hf to megads' + 'if set, convert reversely from megads to hf ckpt.', + ) return parser @@ -57,12 +72,14 @@ def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. loaded = {} logger.info( - f"----------------------------hf weight list----------------------------") + '----------------------------hf weight list----------------------------' + ) for wid in range(1, hf_ckpt_num_of_shards + 1): d = torch.load( - f"{hf_ckpt_dir}/pytorch_model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.bin", - map_location=torch.device('cpu')) + f'{hf_ckpt_dir}/pytorch_model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.bin', + map_location=torch.device('cpu'), + ) for k in d: logger.info(k) assert k not in loaded @@ -71,22 +88,26 @@ def load_and_print_hf_weight(hf_ckpt_dir, hf_ckpt_num_of_shards): return loaded -def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards): +def load_and_print_hf_weight_from_safetensor( + hf_ckpt_dir, hf_ckpt_num_of_shards +): from safetensors import safe_open + # Optimization point: We can selectively load specific 'shared' data to reduce CPU memory usage. hf_model = {} logger.info( - f"----------------------------hf weight list----------------------------") + '----------------------------hf weight list----------------------------' + ) for wid in range(1, hf_ckpt_num_of_shards + 1): if hf_ckpt_num_of_shards == 1: - ckpt_path = f"{hf_ckpt_dir}/model.safetensors" + ckpt_path = f'{hf_ckpt_dir}/model.safetensors' else: - ckpt_path = f"{hf_ckpt_dir}/model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.safetensors" + ckpt_path = f'{hf_ckpt_dir}/model-{wid:05d}-of-{hf_ckpt_num_of_shards:05d}.safetensors' - with safe_open(ckpt_path, framework="pt", device="cpu") as f: + with safe_open(ckpt_path, framework='pt', device='cpu') as f: for k in f.keys(): - logger.info(f"name: {k}, shape: {f.get_tensor(k).shape}") + logger.info(f'name: {k}, shape: {f.get_tensor(k).shape}') assert k not in hf_model hf_model[k] = f.get_tensor(k).clone() @@ -94,47 +115,67 @@ def load_and_print_hf_weight_from_safetensor(hf_ckpt_dir, hf_ckpt_num_of_shards) def load_and_print_hf_weight_auto(hf_ckpt_dir, no_init=True): - from transformers import AutoConfig, AutoModelForCausalLM + from contextlib import nullcontext + from transformers import AutoModelForCausalLM from transformers.modeling_utils import no_init_weights - if no_init: - hf_config = AutoConfig.from_pretrained(hf_ckpt_dir, trust_remote_code=True) - with no_init_weights(): - hf_model = AutoModelForCausalLM.from_config(hf_config, trust_remote_code=True, torch_dtype=torch.bfloat16) - else: - hf_model = {} - hf_auto_model = AutoModelForCausalLM.from_pretrained(hf_ckpt_dir, trust_remote_code=True, torch_dtype=torch.bfloat16) - logger.info( - f"----------------------------hf weight list----------------------------") + ctx = nullcontext() if no_init else no_init_weights() + with ctx: + hf_auto_model = AutoModelForCausalLM.from_pretrained( + hf_ckpt_dir, trust_remote_code=True, torch_dtype=torch.bfloat16 + ) + logger.info( + '----------------------------hf weight list----------------------------' + ) - for name, param in hf_auto_model.named_parameters(): - hf_model[name] = param.clone() - logger.info(name) + hf_model = {} + for name, submodule in hf_auto_model.named_children(): + for pname, param in submodule.named_parameters(): + logger.info(f'[{name}.{pname}] shape={param.shape}') + hf_model[f'{name}.{pname}'] = param.clone() + torch.distributed.barrier() return hf_model def print_distinct_weights(model): logger.info( - f"----------------------------mega-ds weight list----------------------------") + '----------------------------mega-ds weight list----------------------------' + ) for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): if mpu.get_pipeline_model_parallel_rank() == pipe_rank: - if mpu.get_data_parallel_rank() == 0 and mpu.get_tensor_model_parallel_rank( - ) == 0: + if ( + mpu.get_data_parallel_rank() == 0 + and mpu.get_tensor_model_parallel_rank() == 0 + ): for pname, p in model.named_parameters(): - print(pname) + logger.info(f'{pname}, {p.shape}') torch.distributed.barrier() else: torch.distributed.barrier() +def get_hf_tokenizer(name: str): + from transformers import AutoTokenizer + + tokenizer = AutoTokenizer.from_pretrained(name) + return tokenizer + + class refactor: def __init__(self, ds_model, hf_model, args, config): - tokenizer = get_tokenizer() + self.tokenizer = get_tokenizer() + if args.tokenizer_type == 'HFTokenizer': + self.hf_tokenizer = get_hf_tokenizer(args.tokenizer_model) + self.token_vocab = len(self.hf_tokenizer) + else: + self.hf_tokenizer = None + assert self.tokenizer is not None + self.token_vocab = self.tokenizer.vocab_size # align layer number self.ds_model = ds_model self.hf_model = hf_model - self.hf_dict = {} # for handling pp case when converting mds => hf + self.hf_dict = {} # for handling pp case when converting mds => hf self.config = config self.offset_num = 2 @@ -143,56 +184,80 @@ def __init__(self, ds_model, hf_model, args, config): self.num_attention_heads = args.num_attention_heads self.num_key_value_heads = args.num_key_value_heads self.mega_lm_head_wnum = self.mega_norm_wnum + 1 - self.token_vocab = tokenizer.vocab_size self.padded_vocab_size = args.padded_vocab_size self.more_padded = self.padded_vocab_size - self.token_vocab self.tp_size = mpu.get_tensor_model_parallel_world_size() self.tp_rank = mpu.get_tensor_model_parallel_rank() - self.decoder_pat = re.compile("(\d+)\.(.+)") + self.decoder_pat = re.compile('(\d+)\.(.+)') self.refactor_weight_list = [] self.is_refactored = False def _embedding_refactor(self, pname, p): - if pname == f"{self.mega_lm_head_wnum}.lm_head.weight": - hf_name = "lm_head.weight" - elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": - hf_name = "model.embed_tokens.weight" - hf_w = self.hf_model[hf_name] + if pname == f'{self.mega_lm_head_wnum}.lm_head.weight': + hf_name = 'lm_head.weight' + elif pname == f'{self.mega_emb_wnum}.word_embeddings.weight': + hf_name = 'model.embed_tokens.weight' + else: + hf_name = pname + # try: + hf_w = self.hf_model.get(hf_name, None) + assert hf_w is not None + # except Exception: + # if ezpz.get_world_size() > 1: + # from ezpz.utils import breakpoint + # breakpoint(0) + # else: + # import pudb; pudb.set_trace() + # torch.distributed.barrier() + # try: assert hf_w.shape[0] == self.token_vocab - per_partition_vocab_size, start_index, end_index = compute_partition_range( - self.padded_vocab_size, self.tp_rank, self.tp_size) + # except Exception: + # from ezpz.utils import breakpoint + # breakpoint(0) + # torch.distributed.barrier(0) + per_partition_vocab_size, start_index, end_index = ( + compute_partition_range( + self.padded_vocab_size, self.tp_rank, self.tp_size + ) + ) end_index = min(end_index, self.token_vocab) real_partition_vocab_size = end_index - start_index - new_w = torch.zeros((per_partition_vocab_size, hf_w.shape[1]), dtype=hf_w.dtype) + new_w = torch.zeros( + (per_partition_vocab_size, hf_w.shape[1]), dtype=hf_w.dtype + ) new_w[:real_partition_vocab_size, :] = hf_w[start_index:end_index, :] if self.tp_rank == self.tp_size - 1 and self.more_padded > 0: - new_w[-self.more_padded:] = hf_w[:self.token_vocab].mean(dim=0, keepdim=True) + new_w[-self.more_padded :] = hf_w[: self.token_vocab].mean( + dim=0, keepdim=True + ) self.record_mapping_info( - f"mega-ds: {pname,p.data.shape}<--hf: {hf_name,} [{start_index}:{end_index},:] of {hf_w.shape}" + f'mega-ds: {pname, p.data.shape}<--hf: {(hf_name,)} [{start_index}:{end_index},:] of {hf_w.shape}' ) return new_w - - - def _direct_refactor(self, pname, p, hf_layer=None, subname=None): - if pname == f"{self.mega_norm_wnum}.weight": - hf_name = "model.norm.weight" - elif subname in ["input_layernorm.weight", "post_attention_layernorm.weight"]: - hf_name = f"model.layers.{hf_layer}.{subname}" + if pname == f'{self.mega_norm_wnum}.weight': + hf_name = 'model.norm.weight' + elif subname in [ + 'input_layernorm.weight', + 'post_attention_layernorm.weight', + ]: + hf_name = f'model.layers.{hf_layer}.{subname}' + else: + hf_name = pname new_w = hf_w = self.hf_model[hf_name] self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_name,} {hf_w.shape}") + f'mega-ds:{pname, p.data.shape}<--hf{(hf_name,)} {hf_w.shape}' + ) return new_w - def _qkv_refactor(self, pname, p, hf_layer): - hf_wq_name = f"model.layers.{hf_layer}.self_attn.q_proj.weight" - hf_wk_name = f"model.layers.{hf_layer}.self_attn.k_proj.weight" - hf_wv_name = f"model.layers.{hf_layer}.self_attn.v_proj.weight" + hf_wq_name = f'model.layers.{hf_layer}.self_attn.q_proj.weight' + hf_wk_name = f'model.layers.{hf_layer}.self_attn.k_proj.weight' + hf_wv_name = f'model.layers.{hf_layer}.self_attn.v_proj.weight' wq = self.hf_model[hf_wq_name] wk = self.hf_model[hf_wk_name] wv = self.hf_model[hf_wv_name] @@ -201,228 +266,299 @@ def _qkv_refactor(self, pname, p, hf_layer): kv_hidden_size = wk.shape[0] per_partition_size, start_qindex, end_index = compute_partition_range( - query_hidden_size, self.tp_rank, self.tp_size) - _,start_kvindex, _= compute_partition_range( - kv_hidden_size, self.tp_rank, self.tp_size) - - hidden_size_per_attention_head = divide(query_hidden_size, - self.config.num_attention_heads) - num_attention_heads_per_partition = divide(self.config.num_attention_heads, - self.tp_size) - - num_kv_heads_per_partition= divide(self.config.num_key_value_heads, - self.tp_size) - qkv_size=(num_attention_heads_per_partition+2*num_kv_heads_per_partition)*hidden_size_per_attention_head - num_qheads_per_group=divide(self.config.num_attention_heads,self.config.num_key_value_heads) - num_groups =divide(num_attention_heads_per_partition,num_qheads_per_group) - new_w = torch.zeros((qkv_size, wq.shape[1]), dtype=wq.dtype) + query_hidden_size, self.tp_rank, self.tp_size + ) + _, start_kvindex, _ = compute_partition_range( + kv_hidden_size, self.tp_rank, self.tp_size + ) - for i in range(num_groups): - query_current_index=start_qindex+i*num_qheads_per_group*hidden_size_per_attention_head - query_next_index=query_current_index+num_qheads_per_group*hidden_size_per_attention_head - kv_current_index=start_kvindex+i*hidden_size_per_attention_head - kv_next_kvindex=kv_current_index+hidden_size_per_attention_head + hidden_size_per_attention_head = divide( + query_hidden_size, self.config.num_attention_heads + ) + num_attention_heads_per_partition = divide( + self.config.num_attention_heads, self.tp_size + ) - new_w_index=i* (num_qheads_per_group+2)*hidden_size_per_attention_head + num_kv_heads_per_partition = divide( + self.config.num_key_value_heads, self.tp_size + ) + qkv_size = ( + num_attention_heads_per_partition + 2 * num_kv_heads_per_partition + ) * hidden_size_per_attention_head + num_qheads_per_group = divide( + self.config.num_attention_heads, self.config.num_key_value_heads + ) + num_groups = divide( + num_attention_heads_per_partition, num_qheads_per_group + ) + new_w = torch.zeros((qkv_size, wq.shape[1]), dtype=wq.dtype) - new_w[new_w_index:new_w_index+(num_qheads_per_group+2)*hidden_size_per_attention_head,:]=\ - torch.cat([ - wq[query_current_index:query_next_index,:], - wk[kv_current_index:kv_next_kvindex,:], - wv[kv_current_index:kv_next_kvindex,:] - ],dim=0) + for i in range(num_groups): + query_current_index = ( + start_qindex + + i * num_qheads_per_group * hidden_size_per_attention_head + ) + query_next_index = ( + query_current_index + + num_qheads_per_group * hidden_size_per_attention_head + ) + kv_current_index = ( + start_kvindex + i * hidden_size_per_attention_head + ) + kv_next_kvindex = kv_current_index + hidden_size_per_attention_head + + new_w_index = ( + i * (num_qheads_per_group + 2) * hidden_size_per_attention_head + ) + + new_w[ + new_w_index : new_w_index + + (num_qheads_per_group + 2) * hidden_size_per_attention_head, + :, + ] = torch.cat( + [ + wq[query_current_index:query_next_index, :], + wk[kv_current_index:kv_next_kvindex, :], + wv[kv_current_index:kv_next_kvindex, :], + ], + dim=0, + ) self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_wq_name,hf_wk_name,hf_wv_name,} cat q,k,v [{query_current_index}:{query_next_index},:] of q,k,v{wq.shape}" + f'mega-ds:{pname, p.data.shape}<--hf{hf_wq_name, hf_wk_name, hf_wv_name} ' + f'cat q,k,v [{query_current_index}:{query_next_index},:] of q,k,v{wq.shape}' ) return new_w def _mlphto4h_dense_refactor(self, pname, p, hf_layer): - hf_w_gate_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" - hf_w_up_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" + hf_w_gate_name = f'model.layers.{hf_layer}.mlp.gate_proj.weight' + hf_w_up_name = f'model.layers.{hf_layer}.mlp.up_proj.weight' w_gate = self.hf_model[hf_w_gate_name] w_up = self.hf_model[hf_w_up_name] hidden_size = w_gate.shape[0] per_partition_size, start_index, end_index = compute_partition_range( - hidden_size, self.tp_rank, self.tp_size) - new_w = torch.zeros((per_partition_size * 2, - w_gate.shape[1]), - dtype=w_gate.dtype) - new_w[:per_partition_size * 2, :] = \ - torch.cat([ - w_gate[start_index:end_index, :], - w_up[start_index:end_index, :] - ], dim=0) + hidden_size, self.tp_rank, self.tp_size + ) + new_w = torch.zeros( + (per_partition_size * 2, w_gate.shape[1]), dtype=w_gate.dtype + ) + new_w[: per_partition_size * 2, :] = torch.cat( + [w_gate[start_index:end_index, :], w_up[start_index:end_index, :]], + dim=0, + ) self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_w_gate_name,hf_w_up_name} cat gate,up [{start_index}:{end_index},:] of gate,up{w_gate.shape}" + f'mega-ds:{pname, p.data.shape}<--hf{hf_w_gate_name, hf_w_up_name} cat gate,up [{start_index}:{end_index},:] of gate,up{w_gate.shape}' ) return new_w def _attn_dense_refactor(self, pname, p, hf_layer, subname): - if subname == "self_attention.dense.weight": - hf_name = f"model.layers.{hf_layer}.self_attn.o_proj.weight" + if subname == 'self_attention.dense.weight': + hf_name = f'model.layers.{hf_layer}.self_attn.o_proj.weight' else: - hf_name = f"model.layers.{hf_layer}.mlp.down_proj.weight" + hf_name = f'model.layers.{hf_layer}.mlp.down_proj.weight' hf_w = self.hf_model[hf_name] hidden_size = hf_w.shape[1] per_partition_size, start_index, end_index = compute_partition_range( - hidden_size, self.tp_rank, self.tp_size) - new_w = torch.zeros((hf_w.shape[0], per_partition_size), dtype=hf_w.dtype) + hidden_size, self.tp_rank, self.tp_size + ) + new_w = torch.zeros( + (hf_w.shape[0], per_partition_size), dtype=hf_w.dtype + ) new_w[:, :per_partition_size] = hf_w[:, start_index:end_index] self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_name,} [:,{start_index}:{end_index}] of {hf_w.shape}" + f'mega-ds:{pname, p.data.shape}<--hf{(hf_name,)} [:,{start_index}:{end_index}] of {hf_w.shape}' ) return new_w def _mlphto4h1_refactor(self, pname, p, hf_layer, subname): - if subname == "mlp.dense_h_to_4h1.weight": - hf_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" + if subname == 'mlp.dense_h_to_4h1.weight': + hf_name = f'model.layers.{hf_layer}.mlp.gate_proj.weight' else: - hf_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" + hf_name = f'model.layers.{hf_layer}.mlp.up_proj.weight' hf_w = self.hf_model[hf_name] hidden_size = hf_w.shape[0] per_partition_size, start_index, end_index = compute_partition_range( - hidden_size, self.tp_rank, self.tp_size) - new_w = torch.zeros((per_partition_size, hf_w.shape[1]), dtype=hf_w.dtype) + hidden_size, self.tp_rank, self.tp_size + ) + new_w = torch.zeros( + (per_partition_size, hf_w.shape[1]), dtype=hf_w.dtype + ) new_w[:per_partition_size, :] = hf_w[start_index:end_index, :] self.record_mapping_info( - f"mega-ds:{pname,p.data.shape}<--hf{hf_name,} [{start_index}:{end_index},:] of {hf_w.shape}" + f'mega-ds:{pname, p.data.shape}<--hf{(hf_name,)} [{start_index}:{end_index},:] of {hf_w.shape}' ) return new_w def transform_from_hf_to_megds(self): - assert self.is_refactored == False + assert not self.is_refactored new_w = None for pname, p in self.ds_model.named_parameters(): - if pname in [ - f"{self.mega_emb_wnum}.word_embeddings.weight", - f"{self.mega_lm_head_wnum}.lm_head.weight" + f'{self.mega_emb_wnum}.word_embeddings.weight', + f'{self.mega_lm_head_wnum}.lm_head.weight', ]: new_w = self._embedding_refactor(pname, p) - elif pname == f"{self.mega_norm_wnum}.weight": + elif pname == f'{self.mega_norm_wnum}.weight': new_w = self._direct_refactor(pname, p) else: mobj = self.decoder_pat.match(pname) + assert mobj is not None layer_num = int(mobj.group(1)) subname = mobj.group(2) hf_layer = layer_num - self.offset_num - if subname in ["self_attention.query_key_value.weight"]: + if subname in ['self_attention.query_key_value.weight']: new_w = self._qkv_refactor(pname, p, hf_layer) - elif subname in ["mlp.dense_h_to_4h.weight"]: + elif subname in ['mlp.dense_h_to_4h.weight']: new_w = self._mlphto4h_dense_refactor(pname, p, hf_layer) elif subname in [ - "self_attention.dense.weight", - "mlp.dense_4h_to_h.weight" + 'self_attention.dense.weight', + 'mlp.dense_4h_to_h.weight', ]: - new_w = self._attn_dense_refactor(pname, p, hf_layer, subname) + new_w = self._attn_dense_refactor( + pname, p, hf_layer, subname + ) elif subname in [ - "mlp.dense_h_to_4h1.weight", - "mlp.dense_h_to_4h2.weight" + 'mlp.dense_h_to_4h1.weight', + 'mlp.dense_h_to_4h2.weight', ]: new_w = self._mlphto4h1_refactor() elif subname in [ - "input_layernorm.weight", - "post_attention_layernorm.weight" + 'input_layernorm.weight', + 'post_attention_layernorm.weight', ]: new_w = self._direct_refactor(pname, p, hf_layer, subname) else: - raise ValueError("Unrecognized weight type") + raise ValueError('Unrecognized weight type') p.data.copy_(new_w) new_w = None self.is_refactored = True - def _embedding_refactor_to_hf(self, pname, ds_w): - if pname == f"{self.mega_lm_head_wnum}.lm_head.weight": + if pname == f'{self.mega_lm_head_wnum}.lm_head.weight': hf_w = self.hf_model.lm_head.weight - hf_w_name = "lm_head.weight" - elif pname == f"{self.mega_emb_wnum}.word_embeddings.weight": + hf_w_name = 'lm_head.weight' + elif pname == f'{self.mega_emb_wnum}.word_embeddings.weight': hf_w = self.hf_model.model.embed_tokens.weight - hf_w_name = "model.embed_tokens.weight" + hf_w_name = 'model.embed_tokens.weight' + else: + hf_w = self.hf_model[pname] + hf_w_name = pname with torch.no_grad(): - ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) - - self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank[:hf_w.shape[0], :]) + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim( + ds_w + ) + + self.hf_dict[hf_w_name] = copy.deepcopy( + ds_w_all_rank[: hf_w.shape[0], :] + ) def _direct_refactor_to_hf(self, pname, ds_w, hf_layer=None, subname=None): - if pname in [f"{self.mega_norm_wnum}.weight"]: + if pname in [f'{self.mega_norm_wnum}.weight']: hf_w = self.hf_model.model.norm.weight - hf_w_name = "model.norm.weight" - elif subname in ["input_layernorm.weight"]: + hf_w_name = 'model.norm.weight' + elif subname in ['input_layernorm.weight']: hf_w = self.hf_model.model.layers[hf_layer].input_layernorm.weight - hf_w_name = f"model.layers.{hf_layer}.input_layernorm.weight" - elif subname in ["post_attention_layernorm.weight"]: - hf_w = self.hf_model.model.layers[hf_layer].post_attention_layernorm.weight - hf_w_name = f"model.layers.{hf_layer}.post_attention_layernorm.weight" + hf_w_name = f'model.layers.{hf_layer}.input_layernorm.weight' + elif subname in ['post_attention_layernorm.weight']: + hf_w = self.hf_model.model.layers[ + hf_layer + ].post_attention_layernorm.weight + hf_w_name = ( + f'model.layers.{hf_layer}.post_attention_layernorm.weight' + ) self.hf_dict[hf_w_name] = copy.deepcopy(ds_w) def _attn_dense_refactor_to_hf(self, pname, ds_w, hf_layer, subname): - if subname == "self_attention.dense.weight": + if subname == 'self_attention.dense.weight': hf_w = self.hf_model.model.layers[hf_layer].self_attn.o_proj.weight - hf_w_name = f"model.layers.{hf_layer}.self_attn.o_proj.weight" - elif subname == "mlp.dense_4h_to_h.weight": + hf_w_name = f'model.layers.{hf_layer}.self_attn.o_proj.weight' + elif subname == 'mlp.dense_4h_to_h.weight': hf_w = self.hf_model.model.layers[hf_layer].mlp.down_proj.weight - hf_w_name = f"model.layers.{hf_layer}.mlp.down_proj.weight" + hf_w_name = f'model.layers.{hf_layer}.mlp.down_proj.weight' with torch.no_grad(): - ds_w_all_rank = tensor_parallel.mappings._gather_along_last_dim(ds_w) + ds_w_all_rank = tensor_parallel.mappings._gather_along_last_dim( + ds_w + ) self.hf_dict[hf_w_name] = copy.deepcopy(ds_w_all_rank) def _mlphto4h_dense_refactor_to_hf(self, pname, ds_w, hf_layer): - hf_g_name = f"model.layers.{hf_layer}.mlp.gate_proj.weight" - hf_u_name = f"model.layers.{hf_layer}.mlp.up_proj.weight" - + hf_g_name = f'model.layers.{hf_layer}.mlp.gate_proj.weight' + hf_u_name = f'model.layers.{hf_layer}.mlp.up_proj.weight' + with torch.no_grad(): - ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) - + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim( + ds_w + ) + ds_w_shape = ds_w_all_rank.shape - ds_w_all_rank = ds_w_all_rank.reshape(self.tp_size, 2, -1, ds_w_shape[-1]) - self.hf_dict[hf_g_name] = copy.deepcopy(ds_w_all_rank[:, 0, :, :].reshape(-1, ds_w_shape[-1])) - self.hf_dict[hf_u_name] = copy.deepcopy(ds_w_all_rank[:, 1, :, :].reshape(-1, ds_w_shape[-1])) + ds_w_all_rank = ds_w_all_rank.reshape( + self.tp_size, 2, -1, ds_w_shape[-1] + ) + self.hf_dict[hf_g_name] = copy.deepcopy( + ds_w_all_rank[:, 0, :, :].reshape(-1, ds_w_shape[-1]) + ) + self.hf_dict[hf_u_name] = copy.deepcopy( + ds_w_all_rank[:, 1, :, :].reshape(-1, ds_w_shape[-1]) + ) - def _qkv_refactor_to_hf(self, pname, ds_w, hf_layer): with torch.no_grad(): - ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim(ds_w) + ds_w_all_rank = tensor_parallel.mappings._gather_along_first_dim( + ds_w + ) hf_q = self.hf_model.model.layers[hf_layer].self_attn.q_proj.weight hf_k = self.hf_model.model.layers[hf_layer].self_attn.k_proj.weight hf_v = self.hf_model.model.layers[hf_layer].self_attn.v_proj.weight - hf_q_name = f"model.layers.{hf_layer}.self_attn.q_proj.weight" - hf_k_name = f"model.layers.{hf_layer}.self_attn.k_proj.weight" - hf_v_name = f"model.layers.{hf_layer}.self_attn.v_proj.weight" + hf_q_name = f'model.layers.{hf_layer}.self_attn.q_proj.weight' + hf_k_name = f'model.layers.{hf_layer}.self_attn.k_proj.weight' + hf_v_name = f'model.layers.{hf_layer}.self_attn.v_proj.weight' oldshape = hf_q.shape hidden_size = oldshape[-1] - hidden_size_per_attention_head = divide(hidden_size, - self.config.num_attention_heads) + hidden_size_per_attention_head = divide( + hidden_size, self.config.num_attention_heads + ) # MHA & GQA - group = divide(self.config.num_attention_heads, self.config.num_key_value_heads) - newshape = (self.config.num_key_value_heads, group + 2, hidden_size_per_attention_head, hidden_size) + group = divide( + self.config.num_attention_heads, self.config.num_key_value_heads + ) + newshape = ( + self.config.num_key_value_heads, + group + 2, + hidden_size_per_attention_head, + hidden_size, + ) ds_w_out = ds_w_all_rank.reshape(*newshape) - query_weight, key_weight, value_weight = torch.split(ds_w_out, [group, 1, 1], dim=1) - self.hf_dict[hf_q_name] = copy.deepcopy(query_weight.reshape(-1, hidden_size)) - self.hf_dict[hf_k_name] = copy.deepcopy(key_weight.reshape(-1, hidden_size)) - self.hf_dict[hf_v_name] = copy.deepcopy(value_weight.reshape(-1, hidden_size)) + query_weight, key_weight, value_weight = torch.split( + ds_w_out, [group, 1, 1], dim=1 + ) + self.hf_dict[hf_q_name] = copy.deepcopy( + query_weight.reshape(-1, hidden_size) + ) + self.hf_dict[hf_k_name] = copy.deepcopy( + key_weight.reshape(-1, hidden_size) + ) + self.hf_dict[hf_v_name] = copy.deepcopy( + value_weight.reshape(-1, hidden_size) + ) del query_weight, key_weight, value_weight - def transform_from_megads_to_hf(self): - for pname, p in self.ds_model.named_parameters(): if pname in [ - f"{self.mega_emb_wnum}.word_embeddings.weight", - f"{self.mega_lm_head_wnum}.lm_head.weight", + f'{self.mega_emb_wnum}.word_embeddings.weight', + f'{self.mega_lm_head_wnum}.lm_head.weight', ]: self._embedding_refactor_to_hf(pname, p) elif pname in [ - f"{self.mega_norm_wnum}.weight", + f'{self.mega_norm_wnum}.weight', ]: self._direct_refactor_to_hf(pname, p) else: @@ -430,23 +566,25 @@ def transform_from_megads_to_hf(self): layer_num = int(mobj.group(1)) subname = mobj.group(2) hf_layer = layer_num - self.offset_num - if subname in ["self_attention.query_key_value.weight"]: + if subname in ['self_attention.query_key_value.weight']: self._qkv_refactor_to_hf(pname, p, hf_layer) - elif subname in ["mlp.dense_h_to_4h.weight"]: + elif subname in ['mlp.dense_h_to_4h.weight']: self._mlphto4h_dense_refactor_to_hf(pname, p, hf_layer) elif subname in [ - "self_attention.dense.weight", - "mlp.dense_4h_to_h.weight" + 'self_attention.dense.weight', + 'mlp.dense_4h_to_h.weight', ]: - self._attn_dense_refactor_to_hf(pname, p, hf_layer, subname) + self._attn_dense_refactor_to_hf( + pname, p, hf_layer, subname + ) elif subname in [ - "input_layernorm.weight", - "post_attention_layernorm.weight", + 'input_layernorm.weight', + 'post_attention_layernorm.weight', ]: self._direct_refactor_to_hf(pname, p, hf_layer, subname) else: - print(f"Unrecognized weight type: {pname}") - raise ValueError(f"Unrecognized weight type: {pname}") + print(f'Unrecognized weight type: {pname}') + raise ValueError(f'Unrecognized weight type: {pname}') self.is_refactored = True def record_mapping_info(self, record_msg): @@ -455,12 +593,15 @@ def record_mapping_info(self, record_msg): def inorder_show_record(self): assert self.is_refactored logger.info( - f"----------------------------mapping list----------------------------") + f'----------------------------mapping list----------------------------' + ) # print dp rank0 tp rank0 records. for pipe_rank in range(mpu.get_pipeline_model_parallel_world_size()): if mpu.get_pipeline_model_parallel_rank() == pipe_rank: - if mpu.get_data_parallel_rank( - ) == 0 and mpu.get_tensor_model_parallel_rank() == 0: + if ( + mpu.get_data_parallel_rank() == 0 + and mpu.get_tensor_model_parallel_rank() == 0 + ): for record in self.refactor_weight_list: print(record) torch.distributed.barrier() @@ -470,58 +611,71 @@ def inorder_show_record(self): def load_hf_weights(args, no_init): if args.load_mode == 'torchbin': - assert no_init == False, "only work with init" - return load_and_print_hf_weight(args.hf_ckpt_dir, args.hf_ckpt_num_shards) + assert no_init == False, 'only work with init' + return load_and_print_hf_weight( + args.hf_ckpt_dir, args.hf_ckpt_num_shards + ) elif args.load_mode == 'safetensor': - assert no_init == False, "only work with init" - return load_and_print_hf_weight_from_safetensor(args.hf_ckpt_dir, args.hf_ckpt_num_shards) - elif args.load_mode == 'auto': + assert no_init == False, 'only work with init' + return load_and_print_hf_weight_from_safetensor( + args.hf_ckpt_dir, args.hf_ckpt_num_shards + ) + elif args.load_mode == 'auto': return load_and_print_hf_weight_auto(args.hf_ckpt_dir, no_init) def convert_ckpt(): """Build the model.""" args = get_args() - logger.info(f'building model ...') - see_memory_usage(f"Before Building Model", force=True) + assert args is not None + logger.info('building model ...') + see_memory_usage('Before Building Model', force=True) config = core_transformer_config_from_args(args) with deepspeed.zero.Init( - data_parallel_group=mpu.get_data_parallel_group(), - remote_device=None if args.remote_device == 'none' else args.remote_device, - config_dict_or_path=args.deepspeed_config, - enabled=args.zero_stage == 3, - mpu=mpu): + data_parallel_group=mpu.get_data_parallel_group(), + remote_device=None + if args.remote_device == 'none' + else args.remote_device, + config_dict_or_path=args.deepspeed_config, + enabled=args.zero_stage == 3, + mpu=mpu, + ): if args.deepspeed and not args.no_pipeline_parallel: - ds_model = GPTModelPipe(config, num_tokentypes=0, parallel_output=True) + ds_model = GPTModelPipe( + config, num_tokentypes=0, parallel_output=True + ) else: - raise NotImplementedError("Not implemented") + raise NotImplementedError('Not implemented') - see_memory_usage(f"After Building Model", force=True) + see_memory_usage('After Building Model', force=True) if torch.distributed.get_rank() < 2: - print(f"{torch.distributed.get_rank()} {ds_model}") + print(f'{torch.distributed.get_rank()} {ds_model}') # 'torchbin', 'safetensor', 'auto' hf_model = load_hf_weights(args, no_init=args.to_hf_ckpt) # print_distinct_weights(hf_model) + # optimizer = None + # if getattr(args, 'optimizer', None) is not None: + # optimizer = get_megatron_optimizer(ds_model) - #init model and save - logger.info(f"before deepspeed init") - ds_engine, _, _, _ = deepspeed.initialize( + # init model and save + logger.info('before deepspeed init') + ds_engine, optimizer, _, _ = deepspeed.initialize( model=ds_model, - optimizer=None, + # optimizer=optimizer, args=args, lr_scheduler=None, - mpu=mpu if args.no_pipeline_parallel else None) - logger.info(f"after deepspeed init") + mpu=mpu if args.no_pipeline_parallel else None, + ) + logger.info('after deepspeed init') if args.to_hf_ckpt: load_checkpoint([ds_engine], None, None, load_only_weights=True) - logger.info(f"completed to load deepspeed actual checkpoint") + logger.info('completed to load deepspeed actual checkpoint') # refactor weight from hf to mega-ds and vice versa - cur_refactor = refactor(ds_model, hf_model, args, config) if args.to_hf_ckpt: cur_refactor.transform_from_megads_to_hf() @@ -533,34 +687,49 @@ def convert_ckpt(): save_path = args.save if not os.path.exists(save_path): Path(save_path).mkdir(parents=True, exist_ok=True) - ckpt_per_pp_path = os.path.join(save_path, f"model_pp{mpu.get_pipeline_model_parallel_rank()}.pt") + ckpt_per_pp_path = os.path.join( + save_path, f'model_pp{mpu.get_pipeline_model_parallel_rank()}.pt' + ) torch.save(cur_refactor.hf_dict, ckpt_per_pp_path) if torch.distributed.is_initialized(): torch.distributed.barrier() - logger.info(f"hf checkpoint will be saved in {save_path}/release ") + logger.info(f'hf checkpoint will be saved in {save_path}/release ') if mpu.is_pipeline_last_stage(): ## doing checkpoint merging and saving... # hf_model.tie_weights() all_wei = {} for pprank in range(mpu.get_pipeline_model_parallel_world_size()): - ckpt_per_pp_path = os.path.join(save_path, f"model_pp{pprank}.pt") + ckpt_per_pp_path = os.path.join( + save_path, f'model_pp{pprank}.pt' + ) partial_wei = torch.load(ckpt_per_pp_path) all_wei = all_wei | partial_wei hf_model.load_state_dict(all_wei) # mega-ds checkpoint will be saved in args.save - hf_model.save_pretrained(os.path.join(save_path, "release"), safe_serialization=True) + hf_model.save_pretrained( + os.path.join(save_path, 'release'), safe_serialization=True + ) else: - logger.info(f"mega-ds checkpoint will be saved in {args.save}") - save_checkpoint(0, [ds_engine], None, None) - - logger.info(f"save checkpoint completed") + logger.info(f'mega-ds checkpoint will be saved in {args.save}') + args.iteration = 0 + # try: + save_checkpoint( + iteration=0, + model=[ds_engine], + optimizer=optimizer, + opt_param_scheduler=None, + ) + # except Exception: + # import pudb; pudb.set_trace() + + logger.info('save checkpoint completed') -if __name__ == "__main__": +if __name__ == '__main__': initialize_megatron(extra_args_provider=add_extra_args) convert_ckpt() From 7727f9372ad75381d82922442200c4c1afbaa47e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 15 Jan 2025 08:41:51 -0600 Subject: [PATCH 382/590] Update README.md --- ALCF/examples/finetune_llama3/README.md | 48 +++++++++++++++++++++---- 1 file changed, 41 insertions(+), 7 deletions(-) diff --git a/ALCF/examples/finetune_llama3/README.md b/ALCF/examples/finetune_llama3/README.md index b860a640736..9f80d3ba946 100644 --- a/ALCF/examples/finetune_llama3/README.md +++ b/ALCF/examples/finetune_llama3/README.md @@ -1,24 +1,56 @@ # Finetune Llama3 from Hugging Face Checkpoint -1. Download HF Checkpoint: +1. **Clone + navigate into repo**: + + ```bash + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed + cd Megatron-DeepSpeed + ``` + +1. **Setup environment**: + + ```bash + PBS_O_WORKDIR=$(pwd) source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) + ezpz_setup_env + +1. **Install Dependencies**: + + ```bash + python3 -m pip install deepspeed --require-virtualenv + python3 -m pip install -e "git+https://github.com/saforem2/ezpz#egg=ezpz" --require-virtualenv + python3 -m pip install -e . + ``` + +1. **Download data**: + + ```bash + curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o dataset/alpaca_data.json + ``` + + (from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json)) + +1. **Download HF Checkpoint**: ```bash MODEL="Llama-3.2-1B" HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download "meta-llama/${MODEL}" --local-dir "${MODEL}" ``` -1. Convert HF --> MDS: + - _might_ require updating `huggingface_hub, hf_transfer`: + + ```bash + python3 -m pip install --upgrade "huggingface_hub[hf_transfer,cli]" hf_transfer` + ``` + +1. **Convert HF --> MDS**: ```bash - git clone https://github.com/argonne-lcf/Megatron-DeepSpeed - cd Megatron-DeepSpeed - curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json TP=1 PP=1 ZERO_STAGE=1 MODEL_NAME=Llama-3.2-1B bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds ``` -## Dataset +
Old: -You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). +From original README: ### Usage @@ -37,3 +69,5 @@ bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh ``` Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). + +
From 13666a1278c390a9129e74fa6e44eb06ea760ae5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 15 Jan 2025 08:42:45 -0600 Subject: [PATCH 383/590] Remove redundant `ALCF/examples/finetune_llama3p2_1B/*` --- ALCF/examples/finetune_llama3p2_1B/README.md | 35 ---- .../finetune_llama3p2_1B/ds_config.json | 12 -- .../finetune_llama3p2_1B/ds_config_empty.json | 12 -- .../finetune_llama3p2_1B/finetune_llama.sh | 185 ------------------ 4 files changed, 244 deletions(-) delete mode 100644 ALCF/examples/finetune_llama3p2_1B/README.md delete mode 100755 ALCF/examples/finetune_llama3p2_1B/ds_config.json delete mode 100755 ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json delete mode 100644 ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh diff --git a/ALCF/examples/finetune_llama3p2_1B/README.md b/ALCF/examples/finetune_llama3p2_1B/README.md deleted file mode 100644 index c027459a546..00000000000 --- a/ALCF/examples/finetune_llama3p2_1B/README.md +++ /dev/null @@ -1,35 +0,0 @@ -# Finetune Llama3 from Hugging Face Checkpoint - -```bash -git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -cd Megatron-DeepSpeed -curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o alpaca_data.json -bash ALCF/examples/finetune_llama3/finetune_llama.sh convert_hf2mds -``` - - - -## Example of Finetuning LLAMA-7B from Hugging Face Weights - -### Dataset -You can access the dataset from [here](https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json). - -### Pre-trained Weights -The pre-trained weights can be found at [Hugging Face - LLAMA-7B](https://huggingface.co/huggyllama/llama-7b). - -### Usage: - -#### 1. Converting Hugging Face Model Weights to Megatron-Deepspeed Model -```bash -bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh convert_hf2mds -``` -This command writes the Hugging Face model weights into the Megatron-Deepspeed model and saves it. You can adjust the parallel configuration in the script.```convert_mds2hf``` can convert a Megatron-Deepspeed model into the Hugging Face format - -#### 2. Fine-tuning Process -```bash -bash examples_deepspeed/finetune_hf_llama/finetune_llama.sh -``` -Execute this command to initiate the finetuning process. The task originates from [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca.git). - - - diff --git a/ALCF/examples/finetune_llama3p2_1B/ds_config.json b/ALCF/examples/finetune_llama3p2_1B/ds_config.json deleted file mode 100755 index bad063119dc..00000000000 --- a/ALCF/examples/finetune_llama3p2_1B/ds_config.json +++ /dev/null @@ -1,12 +0,0 @@ -{ - "train_batch_size" : 64, - "train_micro_batch_size_per_gpu": 8, - "steps_per_print": 1, - "gradient_accumulation_steps": 1, - "zero_optimization": { - "stage": 0 - }, - "bf16": { - "enabled": true - } -} diff --git a/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json b/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json deleted file mode 100755 index ca14161f24c..00000000000 --- a/ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json +++ /dev/null @@ -1,12 +0,0 @@ -{ - "train_batch_size" : 64, - "train_micro_batch_size_per_gpu": 8, - "steps_per_print": 1, - "gradient_accumulation_steps": 1, - "zero_optimization": { - "stage": 0 - }, - "bf16": { - "enabled": true - } -} diff --git a/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh b/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh deleted file mode 100644 index 85a911dfb32..00000000000 --- a/ALCF/examples/finetune_llama3p2_1B/finetune_llama.sh +++ /dev/null @@ -1,185 +0,0 @@ -DS_CONFIG=./ALCF/examples/finetune_llama3p2_1B/ds_config.json -DS_CONFIG_EMPTY=./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json -DATASET_PATH=./ALCF/examples/finetune_llama3p2_1B/alpaca_data.json -# dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json -# HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct -# HF_LLAMA_PATH=/home/foremans/.llama/checkpoints/Llama3.3-70B-Instruct -# HF_LLAMA_PATH=./Llama-2-7B-hf -PBS_O_WORKDIR=$(pwd) -source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh) && ezpz_setup_env -ezpz_setup_env -# head -n 1 "${PBS_NODEFILE}" > nodefile-0 -#ezpz_setup_job nodefile-0 -# HF_LLAMA_PATH=/data/llama-2-7b-hf/ -# weights link: https://huggingface.co/huggyllama/llama-7b -# - -machine_name=$(ezpz_get_machine_name) -if [[ "${machine_name}" == "aurora" ]]; then - BACKEND="ccl" - HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct -elif [[ "${machine_name}" == "polaris" ]]; then - BACKEND="nccl" - HF_LLAMA_PATH="Llama-3.2-1B" -fi - -HF_CONFIG="${HF_LLAMA_PATH}/config.json" - -MICRO_BATCH_SIZE=8 -ZERO_STAGE=0 -# ${NGPUS}" -TP=1 -PP=1 -WORLD_SIZE="${NGPUS}" -# WORLD_SIZE="${TP}" -GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * WORLD_SIZE / (TP * PP))) -# require to align with weight dimensions -# HIDDEN_SIZE=4096 -# FFN_HIDDEN_SIZE=11008 -# NUM_LAYERS=32 -# NUM_HEADS=32 -HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.hidden_size') -FFN_HIDDEN_SIZE=$(cat "${HF_CONFIG}" | jq -r '.intermediate_size') -NUM_LAYERS=$(cat "${HF_CONFIG}" | jq -r '.num_hidden_layers') -NUM_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') -NUM_KV_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') -SEQ_LENGTH=512 -###################################### - -MEGA_DS_LLAMA_PATH=./"llama-3p2-1B-mega-ds-T${TP}P${PP}" - -# Below configuration required for llama model as per llama paper -# --no-query-key-layer-scaling \ -# --attention-dropout 0 \ -# --hidden-dropout 0 \ -# --use-rotary-position-embeddings \ -# --untie-embeddings-and-output-weights \ -# --swiglu \ -# --normalization rmsnorm \ -# --disable-bias-linear \ -###################################### -cat < $DS_CONFIG -{ - "train_batch_size" : $GLOBAL_BATCH_SIZE, - "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, - "steps_per_print": 1, - "gradient_accumulation_steps": 1, - "zero_optimization": { - "stage": $ZERO_STAGE - }, - "bf16": { - "enabled": true - } -} -EOT - -cat < $DS_CONFIG_EMPTY -{ - "train_batch_size" : $GLOBAL_BATCH_SIZE, - "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, - "steps_per_print": 1, - "gradient_accumulation_steps": 1, - "zero_optimization": { - "stage": $ZERO_STAGE - }, - "bf16": { - "enabled": true - } -} -EOT - -if [ "$1" = "convert_hf2mds" ]; then - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json" -elif [ "$1" = "convert_mds2hf" ]; then - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config_empty.json" -else - DS_CONFIG_PATH="./ALCF/examples/finetune_llama3p2_1B/ds_config.json" -fi - -# --hf-ckpt-num-shards 2 \ -covert_hf2mds_args="launch python3 tools/hf2megads_weight_converter.py \ ---hf-ckpt-dir $HF_LLAMA_PATH \ ---load-mode auto \ ---save $MEGA_DS_LLAMA_PATH" - -# --hf-ckpt-num-shards 2 \ -covert_mds2hf_args="launch python3 tools/hf2megads_weight_converter.py \ ---hf-ckpt-dir $HF_LLAMA_PATH \ ---load-mode auto \ ---to-hf-ckpt \ ---load $MEGA_DS_LLAMA_PATH \ ---save $HF_LLAMA_PATH'-hf-out' " - -finetune_args="launch python3 finetune_llama.py \ ---load $MEGA_DS_LLAMA_PATH" - -comm_args="--tensor-model-parallel-size $TP \ ---pipeline-model-parallel-size $PP \ ---lr-warmup-iters 2000 \ ---weight-decay 0.1 \ ---clip-grad 1 \ ---num-layers $NUM_LAYERS \ ---hidden-size $HIDDEN_SIZE \ ---num-attention-heads $NUM_HEADS \ ---finetune \ ---ffn-hidden-size $FFN_HIDDEN_SIZE \ ---num-key-value-heads $NUM_KV_HEADS \ ---attention-dropout 0 \ ---hidden-dropout 0 \ ---no-query-key-layer-scaling \ ---disable-bias-linear \ ---normalization rmsnorm \ ---use-rotary-position-embeddings \ ---untie-embeddings-and-output-weights \ ---swiglu \ ---seq-length $SEQ_LENGTH \ ---max-position-embeddings $SEQ_LENGTH \ ---micro-batch-size $MICRO_BATCH_SIZE \ ---global-batch-size $GLOBAL_BATCH_SIZE \ ---train-iters 3500 \ ---lr 2e-5 \ ---tensorboard-dir tensorboard_output \ ---lr-decay-iters 320000 \ ---lr-decay-style cosine \ ---log-interval 1 \ ---eval-iters 100 \ ---eval-interval 100 \ ---data-path $DATASET_PATH \ ---save-interval 1500 \ ---split 100,0,0 \ ---bf16 \ ---zero-stage $ZERO_STAGE \ ---tokenizer-type HFTokenizer \ ---tokenizer-model meta-llama/Llama-2-7B-hf \ ---deepspeed_config $DS_CONFIG_PATH \ ---deepspeed \ ---distributed-backend $BACKEND \ ---num-workers 0 \ ---no-masked-softmax-fusion \ ---no-bias-gelu-fusion \ ---no-bias-dropout-fusion \ ---no-gradient-accumulation-fusion \ ---repeated-dataloader" - -# --tokenizer-model meta-llama/Llama-2-7B-hf \ -# --tokenizer-type HFTokenizer \ -# --tokenizer-model 'file:///flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct' \ -# --tokenizer-model meta-llama/Llama-3-70B-Instruct \ -# --tokenizer-model ${HOME}/.llama/checkpoints/Llama3.3-70B-Instruct \ -# -if [ "$1" = "convert_hf2mds" ]; then - task_args="$covert_hf2mds_args" -elif [ "$1" = "convert_mds2hf" ]; then - task_args="$covert_mds2hf_args" -else - task_args="$finetune_args" -fi - -full_cmd="$task_args $comm_args" - -OUTFILE="finetune-llama-$(tstamp).log" -printf "task_args: %s\n" $task_args[@] | tee "${OUTFILE}" -printf "comm_args: %s\n" $comm_args[@] | tee -a "${OUTFILE}" -printf "full_cmd: %s\n" $full_cmd[@] | tee -a "${OUTFILE}" -eval "$full_cmd" | tee -a "${OUTFILE}" -wait $! From 6b5eed5aa136bc65378ae3c4e502725318ea8b50 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 16 Jan 2025 15:30:04 -0600 Subject: [PATCH 384/590] chore: Add `DummyOptimizer` to `tools/hf2megads_weight_converter.py` --- tools/hf2megads_weight_converter.py | 63 +++++++++++++++++++++-------- 1 file changed, 46 insertions(+), 17 deletions(-) diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py index f84f6faf4e7..fc1f82cb2ff 100644 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -7,6 +7,8 @@ sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) import torch.distributed +from dataclasses import dataclass + # from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP from megatron import get_tokenizer, get_args from megatron.core import mpu @@ -19,7 +21,7 @@ from megatron.arguments import core_transformer_config_from_args from megatron.initialize import initialize_megatron -# from megatron.optimizer import get_megatron_optimizer +from megatron.optimizer import get_megatron_optimizer from megatron.checkpointing import save_checkpoint, load_checkpoint # from megatron.training import get_optimizer_param_scheduler @@ -36,6 +38,14 @@ logger = ezpz.get_logger(__name__) +class DummyOptimizer: + def __init__(self, model: torch.nn.Module, lr: float): + self.model = model + self.lr = lr + self.state_dict = {} + + + def add_extra_args(parser): """Text generation arguments.""" group = parser.add_argument_group(title='hf2mega') @@ -662,15 +672,32 @@ def convert_ckpt(): # init model and save logger.info('before deepspeed init') - ds_engine, optimizer, _, _ = deepspeed.initialize( - model=ds_model, - # optimizer=optimizer, - args=args, - lr_scheduler=None, - mpu=mpu if args.no_pipeline_parallel else None, - ) + optimizer = None + dummy_optimizer = DummyOptimizer(model=ds_model, lr=args.lr) + if args.optimizer is None: + ds_engine, ds_opt, _, _ = deepspeed.initialize( + model=ds_model, + # optimizer=optimizer, + args=args, + lr_scheduler=None, + mpu=mpu if args.no_pipeline_parallel else None, + ) + if ds_opt is not None: + optimizer = ds_opt + else: + ds_engine, _, _, _ = deepspeed.initialize( + model=ds_model, + # optimizer=DummyOptimizer(model=ds_model, lr=0.0), + args=args, + lr_scheduler=None, + mpu=mpu if args.no_pipeline_parallel else None, + ) + #optimizer = torch.optim.AdamW(ds_engine ) + #optimizer = get_megatron_optimizer(ds_model) logger.info('after deepspeed init') + optimizer = dummy_optimizer if optimizer is None else optimizer + if args.to_hf_ckpt: load_checkpoint([ds_engine], None, None, load_only_weights=True) logger.info('completed to load deepspeed actual checkpoint') @@ -717,15 +744,17 @@ def convert_ckpt(): else: logger.info(f'mega-ds checkpoint will be saved in {args.save}') args.iteration = 0 - # try: - save_checkpoint( - iteration=0, - model=[ds_engine], - optimizer=optimizer, - opt_param_scheduler=None, - ) - # except Exception: - # import pudb; pudb.set_trace() + try: + save_checkpoint( + iteration=0, + model=[ds_engine], + optimizer=optimizer, + opt_param_scheduler=None, + ) + except Exception: + from ezpz.utils import breakpoint + breakpoint(0) + torch.distributed.barrier() logger.info('save checkpoint completed') From 2f9e19dd22a74dd0495c2c16fa47cce24a6b03c1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 16 Jan 2025 15:52:42 -0600 Subject: [PATCH 385/590] fix: `NO_FLASH_ATTN` on Polaris in `ALCF/helpers.sh` --- ALCF/helpers.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 64ab51cc0a7..a0e2e78bb24 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -518,7 +518,7 @@ get_grad_acc_steps_on_aurora() { exit 1 fi nhosts=$(wc -l <"${hf}") - if [[ "${nhosts}" -gt 256 ]]; then + if [[ "${nhosts}" -ge 256 ]]; then gas=1 elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then gas=2 @@ -629,7 +629,7 @@ setParams() { # NOTE: MICRO_BATCH is exported below # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" - if [[ -n "${NO_FLASH_ATTN-}" ]]; then + if [[ -n "${NO_FLASH_ATTN:-}" ]]; then echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" From 0636aea53e663867405970124fab3052af4dcb7b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 17 Jan 2025 09:00:29 -0600 Subject: [PATCH 386/590] Add {sunspot, sophia} in `ALCF/examples/finetune_llama3/*` --- ALCF/examples/finetune_llama3/finetune_llama.sh | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/ALCF/examples/finetune_llama3/finetune_llama.sh b/ALCF/examples/finetune_llama3/finetune_llama.sh index d3509caccf9..00d56823aac 100644 --- a/ALCF/examples/finetune_llama3/finetune_llama.sh +++ b/ALCF/examples/finetune_llama3/finetune_llama.sh @@ -4,6 +4,12 @@ source <(curl -s https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main DS_CONFIG=./ALCF/examples/finetune_llama3/ds_config.json DS_CONFIG_EMPTY=./ALCF/examples/finetune_llama3/ds_config_empty.json DATASET_PATH="./dataset/alpaca_data.json" + +if [[ ! -f "${DATASET_PATH}" ]]; then + echo "Downloading alpaca_data.json to dataset/alpaca_data.json..." + curl https://raw.githubusercontent.com/tatsu-lab/stanford_alpaca/refs/heads/main/alpaca_data.json -o dataset/alpaca_data.json +fi + # DATASET_PATH=./ALCF/examples/finetune_llama3/alpaca_data.json # dataset link: https://github.com/tatsu-lab/stanford_alpaca/blob/main/alpaca_data.json # HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct @@ -18,14 +24,10 @@ DATASET_PATH="./dataset/alpaca_data.json" MODEL_NAME="${MODEL_NAME:-"Llama-3.3-70B-Instruct"}" machine_name=$(ezpz_get_machine_name) -if [[ "${machine_name}" == "aurora" ]]; then +if [[ "${machine_name}" == "aurora" || "${machine_name}" == "sunspot" ]]; then BACKEND="ccl" - HF_LLAMA_PATH="${MODEL_NAME}" - # HF_LLAMA_PATH=/flare/Aurora_deployment/meta-llama/70B/Llama-3.3-70B-Instruct -elif [[ "${machine_name}" == "polaris" ]]; then +elif [[ "${machine_name}" == "polaris" || "${machine_name}" == "sophia" ]]; then BACKEND="nccl" - # HF_LLAMA_PATH="Llama-3.2-1B" - # HF_LLAMA_PATH=Llama-3.3-70B-Instruct fi HF_LLAMA_PATH="${MODEL_NAME}" From adeca53bfe67a69f1cb2b6af314439e8db5b08f7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sat, 18 Jan 2025 22:48:45 -0600 Subject: [PATCH 387/590] fix: Call `set_ccl_vars_on_aurora` only if `WORLD_SIZE > 1` --- ALCF/helpers.sh | 28 +++++----------------------- 1 file changed, 5 insertions(+), 23 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index a0e2e78bb24..627f5838a10 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -594,29 +594,6 @@ setParams() { else FLASH_ARG="--use-flash-attn-builder" fi - #### [sam: 08/17/2024] ########################################## - # Use best set of CCL env vars from Gordon Bell runs on Aurora - set_ccl_vars_on_aurora - ################################################################# - #### [sam: 06/20/2024] ############################################### - # export CCL_PROCESS_LAUNCHER=pmix - # export CCL_ATL_TRANSPORT=mpi - # !XXX: USE KEY VALUE STORE FIX ON AURORA [2024-06-20] - # use_kvs_fix_on_aurora # <-- why are these different from those in update_ccl_env_vars_aurora ?? - # update_ccl_env_vars_aurora - ###################################################################### - # if [[ -z "${USE_FLASH_ATTN:-}" ]]; then - # # NOTE: if NO_FLASH_ATTN is NON-empty; then NO FLASH ATTN !! - # export NO_FLASH_ATTN=1 # disabled on [2024-06-20] waiting on fix... - # if [[ -n "${NO_FLASH_ATTN-}" ]]; then - # echo "Not using flash-attn!!" - # else - # FLASH_ARG="--use-flash-attn-builder" - # fi - # else - # echo "Using flash-attn !!" - # FLASH_ARG="--use-flash-attn-builder" - # fi # [Polaris] elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" @@ -679,6 +656,11 @@ setParams() { fi export NGPU_PER_HOST="${NGPU_PER_HOST}" export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" + if [[ "${WORLD_SIZE}" -gt 1 && "${mn}" == "aurora" ]]; then + #### [sam: 08/17/2024] ########################################## + # Use best set of CCL env vars from Gordon Bell runs on Aurora + set_ccl_vars_on_aurora + fi # +---[Llama2 7B Config]--------------------------------------------------+ # export MODEL_KEY="Llama-7B" export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS From 4d0077c1c817aeef257ecb245a946caa2e11ba7e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 26 Jan 2025 18:30:08 -0600 Subject: [PATCH 388/590] Update README.md --- ALCF/README.md | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index b0eb99deb6b..e88cebc8ebe 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,7 +1,7 @@ # Megatron-DeepSpeed @ ALCF > [!IMPORTANT] -> [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) is the main entry point for launching +> [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) is the main entry point for launching > distributed training on {Polaris, Aurora, Sunspot} @ ALCF. ## 🏃‍♂️ Running @@ -14,12 +14,12 @@ To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I ``` - - Or, alternatively, you can submit [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) + - Or, alternatively, you can submit [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) directly as a batch script with ```bash cd Megatron-DeepSpeed - qsub -A -q debug -l select=2 -l walltime=01:00:00:filesystems=eagle:home train_llama_alcf.sh + qsub -A -q debug -l select=2 -l walltime=01:00:00:filesystems=eagle:home train_aGPT_7B.sh ```
@@ -201,13 +201,13 @@ will (1.) ```bash # for systems other than Polaris, replace "polaris/books.txt" below with: # "{aurora,sunspot}/books.txt", - PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_llama_alcf.sh + PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_aGPT_7B.sh ``` - **Note**: If no additional options specified, i.e. ```bash - PBS_O_WORKDIR=$(pwd) bash train_llama_alcf.sh + PBS_O_WORKDIR=$(pwd) bash train_aGPT_7B.sh ``` then this will fallback to using the default AuroraGPT-7B architecture @@ -226,7 +226,7 @@ will (1.) $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python #[🌌][10:46:57 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?][aurora_nre_models_frameworks-2024.1] - (aurora_nre_models_frameworks-2024.1) $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt bash train_llama_alcf.sh > train-log-$(tstamp).log 2>&1 & + (aurora_nre_models_frameworks-2024.1) $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt bash train_aGPT_7B.sh > train-log-$(tstamp).log 2>&1 & Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed Running on: aurora @@ -489,7 +489,7 @@ will (1.) ```bash # [09:07:32 AM][foremans@x1921c0s0b0n0][~/q/llm.devkit/Megatron-DeepSpeed][🌱 main][$!?] - $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_llama_alcf.sh + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_aGPT_7B.sh source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... UMD: agama-ci-devel-736.9 successfully loaded: @@ -671,7 +671,7 @@ will (1.) ```bash # [09:31:35 AM][foremans@x3112c0s13b0n0][~/pol/p/a/Megatron-DeepSpeed][🌱 main][$!?] - $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_llama_alcf.sh + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_aGPT_7B.sh source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh Running on Polaris !! @@ -870,8 +870,8 @@ will (1.) --> @@ -998,7 +998,7 @@ export PBS_O_WORKDIR="$(pwd)" && DATA_FILE_LIST=./ALCF/data-lists/polaris/books. ### Running - The (shell) script used to launch pre-training is: - - [`train_llama_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_llama_alcf.sh) + - [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) - This shell script will set the appropriate environment variables, load the correct conda modules and launch @@ -1015,7 +1015,7 @@ modules and launch $ cd Megatron-DeepSpeed # 4. Launch: $ export PBS_O_WORKDIR=$(pwd) - $ bash train_llama_alcf_polaris.sh + $ bash train_aGPT_7B_polaris.sh ```
[Output] @@ -1075,7 +1075,7 @@ modules and launch ```bash $ cd Megatron-DeepSpeed -$ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home train_llama_alcf.sh +$ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home train_aGPT_7B.sh ``` From e7990d5fea031a811d26785cb2e18c47391e4996 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Tue, 28 Jan 2025 19:16:59 +0000 Subject: [PATCH 389/590] added adopt optimizer --- megatron/arguments.py | 14 +- megatron/model/__init__.py | 14 +- megatron/optimizer/__init__.py | 34 +- megatron/optimizer/adopt.py | 559 +++++++++++++++++++++++++++++++++ 4 files changed, 601 insertions(+), 20 deletions(-) create mode 100644 megatron/optimizer/adopt.py diff --git a/megatron/arguments.py b/megatron/arguments.py index cef72acf51f..1c3ac06c424 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -798,9 +798,18 @@ def _add_regularization_args(parser): group.add_argument('--adam-eps', type=float, default=1e-08, help='Term added to the denominator to improve' 'numerical stability') + group.add_argument('--dshampooadamw-max-preconditioner-dim', type=int, default=8192, + help='Max preconditioner dim') + group.add_argument('--dshampooadamw-precondition-freq', type=int, default=100, + help='dshampoo preconditioner frequency') + group.add_argument('--dshampooadamw-use-decoupled-weight-decay', type=bool, default=True, + help='dshampoo use decoupled weight decay') + group.add_argument('--dshampooadamw-eps', type=float, default=1e-12, + help='dshampoo epsilon') group.add_argument('--sgd-momentum', type=float, default=0.9, help='Momentum factor for sgd') + return parser @@ -966,10 +975,11 @@ def _add_training_args(parser): 'galoreadamw8bitperlayer', 'ipex.fusedlamb', 'ipex.lamb', - 'shampoo', + 'dshampooadamw', 'sgd', 'sgdschedulefree', - 'sophiag' + 'sophiag', + 'adopt' ], help='Optimizer function' ) diff --git a/megatron/model/__init__.py b/megatron/model/__init__.py index f2beea06e81..6b3923aa977 100644 --- a/megatron/model/__init__.py +++ b/megatron/model/__init__.py @@ -13,13 +13,13 @@ if accelerator is not None and accelerator.device_name() == "cuda": from .fused_layer_norm import MixedFusedLayerNorm as LayerNorm - try: - from apex.normalization import MixedFusedRMSNorm as RMSNorm # type:ignore - - HAS_APEX = True - except Exception: - HAS_APEX = False - from .rmsnorm import RMSNorm +# try: +# from apex.normalization import MixedFusedRMSNorm as RMSNorm # type:ignore +# + # HAS_APEX = True + # except Exception: + # HAS_APEX = False + from .rmsnorm import RMSNorm else: if hasattr(torch.xpu, "IpexRmsNorm"): from .fused_rmsnorm import RMSNorm diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 37107a44e71..27527beef51 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -326,23 +326,24 @@ def optimizer_hook(p): eps=args.adam_eps, ) # ---- Shampoo ---------------------------------------- - elif args.optimizer == 'shampoo': + elif str(args.optimizer).lower() == 'dshampooadamw': from distributed_shampoo.distributed_shampoo import DistributedShampoo from distributed_shampoo.shampoo_types import AdamGraftingConfig optimizer = DistributedShampoo( - model.parameters(), - lr=0.001, - betas=(0.9, 0.999), - epsilon=1e-12, - weight_decay=1e-05, - max_preconditioner_dim=8192, - precondition_frequency=100, - use_decoupled_weight_decay=True, + param_groups, + lr=args.lr, + betas=(args.adam_beta1, args.adam_beta2), + epsilon=args.dshampooadamw_eps, + weight_decay=args.weight_decay, + max_preconditioner_dim=args.dshampooadamw_max_preconditioner_dim, + precondition_frequency=args.dshampooadamw_precondition_freq, + use_decoupled_weight_decay=args.dshampooadamw_use_decoupled_weight_decay, grafting_config=AdamGraftingConfig( - beta2=0.999, - epsilon=1e-08, + beta2=args.adam_beta2, + epsilon=args.adam_eps, ), ) + #--- sgd ------------------------------------------------- elif args.optimizer == 'sgd': optimizer = torch.optim.SGD( param_groups, @@ -350,6 +351,7 @@ def optimizer_hook(p): weight_decay=args.weight_decay, momentum=args.sgd_momentum ) + #--- sophiag ----------------------------------------------- elif str(args.optimizer).lower() == 'sophiag': from .sophia import SophiaG optimizer = SophiaG( @@ -359,6 +361,16 @@ def optimizer_hook(p): rho = args.sophiag_rho, weight_decay=args.weight_decay ) + #--- adopt ----------------------------------------------- + elif str(args.optimizer).lower() == 'adopt': + from .adopt import ADOPT + optimizer = ADOPT( + param_groups, + lr=args.lr, + betas=(args.adam_beta1, args.adam_beta2), + eps = args.adam_eps, + weight_decay=args.weight_decay + ) else: raise TypeError(f'{args.optimizer} optimizer is not supported.') assert optimizer is not None diff --git a/megatron/optimizer/adopt.py b/megatron/optimizer/adopt.py new file mode 100644 index 00000000000..300cf3aad3a --- /dev/null +++ b/megatron/optimizer/adopt.py @@ -0,0 +1,559 @@ +# mypy: allow-untyped-decorators +# mypy: allow-untyped-defs +from typing import cast, Callable, List, Optional, Tuple, Union + +import torch +from torch import Tensor + +from torch.optim.optimizer import ( + _capturable_doc, + _default_to_fused_or_foreach, + _device_dtype_check_for_fused, + _differentiable_doc, + _disable_dynamo_if_unsupported, + _foreach_doc, + _fused_doc, + _get_capturable_supported_devices, + _get_scalar_dtype, + _get_value, + _maximize_doc, + _stack_if_compiling, + _use_grad_for_differentiable, + _view_as_real, + DeviceDict, + Optimizer, + ParamsT, +) + + +__all__ = ["ADOPT", "adopt"] + +#code from https://github.com/iShohei220/adopt/blob/main/src/adopt/adopt.py, copy pasted here because no pip and not sure about submodules +class ADOPT(Optimizer): + def __init__( + self, + params: ParamsT, + lr: Union[float, Tensor] = 1e-3, + betas: Tuple[float, float] = (0.9, 0.9999), + eps: float = 1e-6, + clip_lambda: Optional[Callable[[int], float]] = lambda step: step**0.25, + weight_decay: float = 0.0, + decouple: bool = False, + *, + foreach: Optional[bool] = None, + maximize: bool = False, + capturable: bool = False, + differentiable: bool = False, + fused: Optional[bool] = None, + ): + if isinstance(lr, Tensor): + if foreach and not capturable: + raise ValueError( + "lr as a Tensor is not supported for capturable=False and foreach=True" + ) + if lr.numel() != 1: + raise ValueError("Tensor lr must be 1-element") + if not 0.0 <= lr: + raise ValueError(f"Invalid learning rate: {lr}") + if not 0.0 <= eps: + raise ValueError(f"Invalid epsilon value: {eps}") + if not 0.0 <= betas[0] < 1.0: + raise ValueError(f"Invalid beta parameter at index 0: {betas[0]}") + if not 0.0 <= betas[1] < 1.0: + raise ValueError(f"Invalid beta parameter at index 1: {betas[1]}") + if not 0.0 <= weight_decay: + raise ValueError(f"Invalid weight_decay value: {weight_decay}") + + self.clip_lambda = clip_lambda + + defaults = dict( + lr=lr, + betas=betas, + eps=eps, + weight_decay=weight_decay, + decouple=decouple, + maximize=maximize, + foreach=foreach, + capturable=capturable, + differentiable=differentiable, + fused=fused, + ) + super().__init__(params, defaults) + + if fused: + # TODO: support fused + raise RuntimeError("`fused` is not currently supported") + + if differentiable: + raise RuntimeError("`fused` does not support `differentiable`") + self._step_supports_amp_scaling = True + # TODO(crcrpar): [low prec params & their higher prec copy] + # Support AMP with FP16/BF16 model params which would need + # higher prec copy of params to do update math in higher prec to + # alleviate the loss of information. + if foreach: + raise RuntimeError("`fused` and `foreach` cannot be `True` together.") + + def __setstate__(self, state): + super().__setstate__(state) + for group in self.param_groups: + group.setdefault("maximize", False) + group.setdefault("foreach", None) + group.setdefault("capturable", False) + group.setdefault("differentiable", False) + fused = group.setdefault("fused", None) + for p in group["params"]: + p_state = self.state.get(p, []) + if len(p_state) != 0 and not torch.is_tensor(p_state["step"]): + step_val = float(p_state["step"]) + p_state["step"] = ( + torch.tensor( + step_val, + dtype=_get_scalar_dtype(is_fused=fused), + device=p.device, + ) + if group["capturable"] or group["fused"] + else torch.tensor(step_val, dtype=_get_scalar_dtype()) + ) + + def _init_group( + self, + group, + params_with_grad, + grads, + exp_avgs, + exp_avg_sqs, + state_steps, + ): + has_complex = False + for p in group["params"]: + if p.grad is not None: + has_complex |= torch.is_complex(p) + params_with_grad.append(p) + if p.grad.is_sparse: + raise RuntimeError( + "ADOPT does not support sparse gradients" + ) + grads.append(p.grad) + + state = self.state[p] + # Lazy state initialization + if len(state) == 0: + if group["fused"]: + _device_dtype_check_for_fused(p) + # note(crcrpar): [special device hosting for step] + # Deliberately host `step` on CPU if both capturable and fused are off. + # This is because kernel launches are costly on CUDA and XLA. + state["step"] = ( + torch.zeros( + (), + dtype=_get_scalar_dtype(is_fused=group["fused"]), + device=p.device, + ) + if group["capturable"] or group["fused"] + else torch.tensor(0.0, dtype=_get_scalar_dtype()) + ) + # Exponential moving average of gradient values + state["exp_avg"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + # Exponential moving average of squared gradient values + state["exp_avg_sq"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + + exp_avgs.append(state["exp_avg"]) + exp_avg_sqs.append(state["exp_avg_sq"]) + + if group["differentiable"] and state["step"].requires_grad: + raise RuntimeError( + "`requires_grad` is not supported for `step` in differentiable mode" + ) + + # Foreach without capturable does not support a tensor lr + if ( + group["foreach"] + and torch.is_tensor(group["lr"]) + and not group["capturable"] + ): + raise RuntimeError( + "lr as a Tensor is not supported for capturable=False and foreach=True" + ) + + state_steps.append(state["step"]) + return has_complex + + @_use_grad_for_differentiable + def step(self, closure=None): + """Perform a single optimization step. + + Args: + closure (Callable, optional): A closure that reevaluates the model + and returns the loss. + """ + self._cuda_graph_capture_health_check() + + loss = None + if closure is not None: + with torch.enable_grad(): + loss = closure() + + for group in self.param_groups: + params_with_grad: List[Tensor] = [] + grads: List[Tensor] = [] + exp_avgs: List[Tensor] = [] + exp_avg_sqs: List[Tensor] = [] + state_steps: List[Tensor] = [] + beta1, beta2 = group["betas"] + + has_complex = self._init_group( + group, + params_with_grad, + grads, + exp_avgs, + exp_avg_sqs, + state_steps, + ) + + adopt( + params_with_grad, + grads, + exp_avgs, + exp_avg_sqs, + state_steps, + has_complex=has_complex, + beta1=beta1, + beta2=beta2, + lr=group["lr"], + clip_lambda=self.clip_lambda, + weight_decay=group["weight_decay"], + decouple=group["decouple"], + eps=group["eps"], + maximize=group["maximize"], + foreach=group["foreach"], + capturable=group["capturable"], + differentiable=group["differentiable"], + fused=group["fused"], + grad_scale=getattr(self, "grad_scale", None), + found_inf=getattr(self, "found_inf", None), + ) + + return loss + + +def _single_tensor_adopt( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + exp_avg_sqs: List[Tensor], + state_steps: List[Tensor], + grad_scale: Optional[Tensor], + found_inf: Optional[Tensor], + *, + has_complex: bool, + beta1: float, + beta2: float, + lr: Union[float, Tensor], + clip_lambda: Optional[Callable[[int], float]], + weight_decay: float, + decouple: bool, + eps: float, + maximize: bool, + capturable: bool, + differentiable: bool, +): + assert grad_scale is None and found_inf is None + + if torch.jit.is_scripting(): + # this assert is due to JIT being dumb and not realizing that the ops below + # have overloads to handle both float and Tensor lrs, so we just assert it's + # a float since most people using JIT are using floats + assert isinstance(lr, float) + + for i, param in enumerate(params): + grad = grads[i] if not maximize else -grads[i] + exp_avg = exp_avgs[i] + exp_avg_sq = exp_avg_sqs[i] + step_t = state_steps[i] + + # If compiling, the compiler will handle cudagraph checks, see note [torch.compile x capturable] + if not torch._utils.is_compiling() and capturable: + capturable_supported_devices = _get_capturable_supported_devices() + assert ( + param.device.type == step_t.device.type + and param.device.type in capturable_supported_devices + ), f"If capturable=True, params and state_steps must be on supported devices: {capturable_supported_devices}." + + step = step_t if capturable or differentiable else _get_value(step_t) + + if weight_decay != 0 and not decouple: + grad = grad.add(param, alpha=weight_decay) + + if torch.is_complex(param): + grad = torch.view_as_real(grad) + if exp_avg is not None: + exp_avg = torch.view_as_real(exp_avg) + if exp_avg_sq is not None: + exp_avg_sq = torch.view_as_real(exp_avg_sq) + param = torch.view_as_real(param) + + if step == 0: + exp_avg_sq.addcmul_(grad, grad.conj()) + # update step + step_t += 1 + continue + + if weight_decay != 0 and decouple: + param.add_(param, alpha=-lr*weight_decay) + + denom = torch.clamp(exp_avg_sq.sqrt(), eps) + normed_grad = grad.div(denom) + if clip_lambda is not None: + clip = clip_lambda(step) + normed_grad.clamp_(-clip, clip) + + exp_avg.lerp_(normed_grad, 1 - beta1) + + param.add_(exp_avg, alpha=-lr) + exp_avg_sq.mul_(beta2).addcmul_(grad, grad.conj(), value=1 - beta2) + + # update step + step_t += 1 + + +def _multi_tensor_adopt( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + exp_avg_sqs: List[Tensor], + state_steps: List[Tensor], + grad_scale: Optional[Tensor], + found_inf: Optional[Tensor], + *, + has_complex: bool, + beta1: float, + beta2: float, + lr: Union[float, Tensor], + clip_lambda: Optional[Callable[[int], float]], + weight_decay: float, + decouple: bool, + eps: float, + maximize: bool, + capturable: bool, + differentiable: bool, +): + if len(params) == 0: + return + + if isinstance(lr, Tensor) and not capturable: + raise RuntimeError( + "lr as a Tensor is not supported for capturable=False and foreach=True" + ) + + # If compiling, the compiler will handle cudagraph checks, see note [torch.compile x capturable] + if not torch._utils.is_compiling() and capturable: + capturable_supported_devices = _get_capturable_supported_devices( + supports_xla=False + ) + assert all( + p.device.type == step.device.type + and p.device.type in capturable_supported_devices + for p, step in zip(params, state_steps) + ), f"If capturable=True, params and state_steps must be on supported devices: {capturable_supported_devices}." + + assert grad_scale is None and found_inf is None + + assert not differentiable, "_foreach ops don't support autograd" + + grouped_tensors = Optimizer._group_tensors_by_device_and_dtype( + [params, grads, exp_avgs, exp_avg_sqs, state_steps] # type: ignore[list-item] + ) + for ( + device_params_, + device_grads_, + device_exp_avgs_, + device_exp_avg_sqs_, + device_state_steps_, + ), _ in grouped_tensors.values(): + device_params = cast(List[Tensor], device_params_) + device_grads = cast(List[Tensor], device_grads_) + device_exp_avgs = cast(List[Tensor], device_exp_avgs_) + device_exp_avg_sqs = cast(List[Tensor], device_exp_avg_sqs_) + device_state_steps = cast(List[Tensor], device_state_steps_) + + # Handle complex parameters + if has_complex: + _view_as_real( + device_params, device_grads, device_exp_avgs, device_exp_avg_sqs + ) + + if maximize: + device_grads = torch._foreach_neg(device_grads) # type: ignore[assignment] + + if weight_decay != 0 and not decouple: + # Re-use the intermediate memory (device_grads) already allocated for maximize + if maximize: + torch._foreach_add_(device_grads, device_params, alpha=weight_decay) + else: + device_grads = torch._foreach_add( # type: ignore[assignment] + device_grads, device_params, alpha=weight_decay + ) + + if device_state_steps[0] == 0: + torch._foreach_addcmul_(device_exp_avg_sqs, device_grads, device_grads) + + # Update steps + # If steps are on CPU, foreach will fall back to the slow path, which is a for-loop calling t.add(1) over + # and over. 1 will then be wrapped into a Tensor over and over again, which is slower than if we just + # wrapped it once now. The alpha is required to assure we go to the right overload. + if not torch._utils.is_compiling() and device_state_steps[0].is_cpu: + torch._foreach_add_( + device_state_steps, torch.tensor(1.0, device="cpu"), alpha=1.0 + ) + else: + torch._foreach_add_(device_state_steps, 1) + + continue + + if weight_decay != 0 and decouple: + torch._foreach_add_(device_params, device_params, alpha=-lr*weight_decay) + + exp_avg_sq_sqrt = torch._foreach_sqrt(device_exp_avg_sqs) + torch._foreach_maximum_(exp_avg_sq_sqrt, eps) + + normed_grad = torch._foreach_div(device_grads, exp_avg_sq_sqrt) + if clip_lambda is not None: + clip = clip_lambda(device_state_steps[0]) + torch._foreach_maximum_(normed_grad, -clip) + torch._foreach_minimum_(normed_grad, clip) + + torch._foreach_lerp_(device_exp_avgs, normed_grad, 1 - beta1) + + torch._foreach_add_(device_params, device_exp_avgs, alpha=-lr) + torch._foreach_mul_(device_exp_avg_sqs, beta2) + torch._foreach_addcmul_( + device_exp_avg_sqs, device_grads, device_grads, value=1 - beta2 + ) + + # Update steps + # If steps are on CPU, foreach will fall back to the slow path, which is a for-loop calling t.add(1) over + # and over. 1 will then be wrapped into a Tensor over and over again, which is slower than if we just + # wrapped it once now. The alpha is required to assure we go to the right overload. + if not torch._utils.is_compiling() and device_state_steps[0].is_cpu: + torch._foreach_add_( + device_state_steps, torch.tensor(1.0, device="cpu"), alpha=1.0 + ) + else: + torch._foreach_add_(device_state_steps, 1) + + +def _fused_adopt( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + exp_avg_sqs: List[Tensor], + state_steps: List[Tensor], + grad_scale: Optional[Tensor], + found_inf: Optional[Tensor], + *, + has_complex: bool, # Needed for consistency. + beta1: float, + beta2: float, + lr: Union[float, Tensor], + clip_lambda: Optional[Callable[[int], float]], + weight_decay: float, + decouple: bool, + eps: float, + maximize: bool, + capturable: bool, # Needed for consistency. + differentiable: bool, +) -> None: + raise NotImplementedError + + +@_disable_dynamo_if_unsupported(single_tensor_fn=_single_tensor_adopt) +def adopt( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + exp_avg_sqs: List[Tensor], + state_steps: List[Tensor], + # kwonly args with defaults are not supported by functions compiled with torchscript issue #70627 + # setting this as kwarg for now as functional API is compiled by torch/distributed/optim + foreach: Optional[bool] = None, + capturable: bool = False, + differentiable: bool = False, + fused: Optional[bool] = None, + grad_scale: Optional[Tensor] = None, + found_inf: Optional[Tensor] = None, + has_complex: bool = False, + *, + beta1: float, + beta2: float, + lr: Union[float, Tensor], + clip_lambda: Optional[Callable[[int], float]], + weight_decay: float, + decouple: bool, + eps: float, + maximize: bool, +): + r"""Functional API that performs ADOPT algorithm computation. + + """ + # Respect when the user inputs False/True for foreach or fused. We only want to change + # the default when neither have been user-specified. Note that we default to foreach + # and pass False to use_fused. This is not a mistake--we want to give the fused impl + # bake-in time before making it the default, even if it is typically faster. + if fused is None and foreach is None: + _, foreach = _default_to_fused_or_foreach( + params, differentiable, use_fused=False + ) + # Do not flip on foreach for the unsupported case where lr is a Tensor and capturable=False. + if foreach and isinstance(lr, Tensor) and not capturable: + foreach = False + if fused is None: + fused = False + if foreach is None: + foreach = False + + # this check is slow during compilation, so we skip it + # if it's strictly needed we can add this check back in dynamo + if not torch._utils.is_compiling() and not all( + isinstance(t, torch.Tensor) for t in state_steps + ): + raise RuntimeError( + "API has changed, `state_steps` argument must contain a list of singleton tensors" + ) + + if foreach and torch.jit.is_scripting(): + raise RuntimeError("torch.jit.script not supported with foreach optimizers") + if fused and torch.jit.is_scripting(): + raise RuntimeError("torch.jit.script not supported with fused optimizers") + + if fused and not torch.jit.is_scripting(): + func = _fused_adopt + elif foreach and not torch.jit.is_scripting(): + func = _multi_tensor_adopt + else: + func = _single_tensor_adopt + + func( + params, + grads, + exp_avgs, + exp_avg_sqs, + state_steps, + has_complex=has_complex, + beta1=beta1, + beta2=beta2, + lr=lr, + clip_lambda=clip_lambda, + weight_decay=weight_decay, + decouple=decouple, + eps=eps, + maximize=maximize, + capturable=capturable, + differentiable=differentiable, + grad_scale=grad_scale, + found_inf=found_inf, + ) From 0948f84df64102e4877e069d25c4f432b16c4d04 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Tue, 28 Jan 2025 19:53:03 +0000 Subject: [PATCH 390/590] adopt optimizer --- megatron/optimizer/adopt.py | 1 + 1 file changed, 1 insertion(+) diff --git a/megatron/optimizer/adopt.py b/megatron/optimizer/adopt.py index 300cf3aad3a..4e5b3722dc0 100644 --- a/megatron/optimizer/adopt.py +++ b/megatron/optimizer/adopt.py @@ -28,6 +28,7 @@ __all__ = ["ADOPT", "adopt"] + #code from https://github.com/iShohei220/adopt/blob/main/src/adopt/adopt.py, copy pasted here because no pip and not sure about submodules class ADOPT(Optimizer): def __init__( From 1c04f64dd82d349343fe861fbd34728eb51c743d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 28 Jan 2025 14:36:45 -0600 Subject: [PATCH 391/590] fix: Resolve merge commit --- .../finetune_llama3/finetune_llama.sh | 34 ++++++++++++------- 1 file changed, 22 insertions(+), 12 deletions(-) diff --git a/ALCF/examples/finetune_llama3/finetune_llama.sh b/ALCF/examples/finetune_llama3/finetune_llama.sh index 00d56823aac..a87e8b3c347 100644 --- a/ALCF/examples/finetune_llama3/finetune_llama.sh +++ b/ALCF/examples/finetune_llama3/finetune_llama.sh @@ -26,8 +26,13 @@ MODEL_NAME="${MODEL_NAME:-"Llama-3.3-70B-Instruct"}" machine_name=$(ezpz_get_machine_name) if [[ "${machine_name}" == "aurora" || "${machine_name}" == "sunspot" ]]; then BACKEND="ccl" + HF_LLAMA_PATH="${MODEL_NAME}" + FLASH_ARG="--use-flash-attn-builder" elif [[ "${machine_name}" == "polaris" || "${machine_name}" == "sophia" ]]; then BACKEND="nccl" + FLASH_ARG="--use-flash-attn-v2" + # HF_LLAMA_PATH="Llama-3.2-1B" + # HF_LLAMA_PATH=Llama-3.3-70B-Instruct fi HF_LLAMA_PATH="${MODEL_NAME}" @@ -38,7 +43,8 @@ ZERO_STAGE="${ZERO_STAGE:-0}" TP="${TP:-1}" PP="${PP:-1}" WORLD_SIZE="${WORLD_SIZE:-${NGPUS}}" -GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * WORLD_SIZE / (TP * PP))) +GAS="${GRAD_ACC_STEPS:-${GAS:-1}}" +GLOBAL_BATCH_SIZE=$((MICRO_BATCH_SIZE * GAS * WORLD_SIZE / (TP * PP))) # require to align with weight dimensions # HIDDEN_SIZE=4096 # FFN_HIDDEN_SIZE=11008 @@ -53,7 +59,7 @@ NUM_LAYERS=$(cat "${HF_CONFIG}" | jq -r '.num_hidden_layers') NUM_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_attention_heads') NUM_KV_HEADS=$(cat "${HF_CONFIG}" | jq -r '.num_key_value_heads') MAX_SEQ_LENGTH=$(cat "${HF_CONFIG}" | jq -r '.max_position_embeddings') -SEQ_LENGTH=512 +SEQ_LENGTH=2048 ###################################### printf "GLOBAL_BATCH_SIZE: %s\n" $GLOBAL_BATCH_SIZE @@ -69,8 +75,9 @@ printf "NUM_HEADS: %s\n" $NUM_HEADS printf "NUM_KV_HEADS: %s\n" $NUM_KV_HEADS printf "SEQ_LENGTH: %s\n" $SEQ_LENGTH -MEGA_DS_LLAMA_PATH="converted_hf_ckpts/${MODEL_NAME}-MDS-GBS${GLOBAL_BATCH_SIZE}-ZS${ZERO_STAGE}-TP${TP}-PP${PP}" -mkdir -p $(dirname $MEGA_DS_LLAMA_PATH) +CKPT_DIR="converted_hf_ckpts/${MODEL_NAME}-MDS-GBS${GLOBAL_BATCH_SIZE}-ZS${ZERO_STAGE}-TP${TP}-PP${PP}" +TB_DIR="${CKPT_DIR}/tensorboard-output" +mkdir -p $(dirname $TB_DIR) # Below configuration required for llama model as per llama paper # --no-query-key-layer-scaling \ @@ -87,7 +94,7 @@ cat <$DS_CONFIG "train_batch_size" : $GLOBAL_BATCH_SIZE, "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, "steps_per_print": 1, - "gradient_accumulation_steps": 1, + "gradient_accumulation_steps": $GAS, "optimizer": { "type": "Adam", "params": { @@ -108,7 +115,7 @@ cat <$DS_CONFIG_EMPTY "train_batch_size" : $GLOBAL_BATCH_SIZE, "train_micro_batch_size_per_gpu": $MICRO_BATCH_SIZE, "steps_per_print": 1, - "gradient_accumulation_steps": 1, + "gradient_accumulation_steps": $GAS, "optimizer": { "type": "Adam", "params": { @@ -136,20 +143,21 @@ fi covert_hf2mds_args="launch python3 tools/hf2megads_weight_converter.py \ --hf-ckpt-dir ${HF_LLAMA_PATH} \ --load-mode auto \ ---save ${MEGA_DS_LLAMA_PATH}" +--save ${CKPT_DIR}" # --hf-ckpt-num-shards 2 \ covert_mds2hf_args="launch python3 tools/hf2megads_weight_converter.py \ --hf-ckpt-dir ${HF_LLAMA_PATH} \ --load-mode auto \ --to-hf-ckpt \ ---load ${MEGA_DS_LLAMA_PATH} \ +--load ${CKPT_DIR} \ --save ${HF_LLAMA_PATH}'-hf-out' " finetune_args="launch python3 finetune_llama.py \ ---load ${MEGA_DS_LLAMA_PATH}" +--load ${CKPT_DIR}" comm_args+=( + "${FLASH_ARG}" "--tensor-model-parallel-size=${TP}" "--pipeline-model-parallel-size=${PP}" "--lr-warmup-iters=2000" @@ -174,15 +182,17 @@ comm_args+=( "--micro-batch-size=${MICRO_BATCH_SIZE}" "--global-batch-size=${GLOBAL_BATCH_SIZE}" "--train-iters=3500" - "--lr=2e-5" - "--tensorboard-dir=tensorboard_output" + "--lr=${LR:-2e-5}" "--lr-decay-iters=320000" "--lr-decay-style=cosine" "--log-interval=1" + "--log-timers-to-tensorboard" + "--timing-log-level=1" + "--tensorboard-dir=${TB_DIR}" "--eval-iters=100" "--eval-interval=100" "--data-path=${DATASET_PATH}" - "--save-interval=1500" + "--save-interval=100" "--split=100,0,0" "--bf16" "--zero-stage=${ZERO_STAGE}" From c1f99b97766afb413ee27914a8f1b8a248da5839 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 28 Jan 2025 15:00:48 -0600 Subject: [PATCH 392/590] chore: Update Llama FT --- ALCF/examples/finetune_llama3/ds_config.json | 4 +- .../finetune_llama3/ds_config_empty.json | 4 +- finetune_llama.py | 58 +++++++++++++++++-- tools/hf2megads_weight_converter.py | 16 +++-- 4 files changed, 68 insertions(+), 14 deletions(-) diff --git a/ALCF/examples/finetune_llama3/ds_config.json b/ALCF/examples/finetune_llama3/ds_config.json index 64d559a27fb..f116c812386 100755 --- a/ALCF/examples/finetune_llama3/ds_config.json +++ b/ALCF/examples/finetune_llama3/ds_config.json @@ -1,6 +1,6 @@ { - "train_batch_size" : 48, - "train_micro_batch_size_per_gpu": 8, + "train_batch_size" : 6, + "train_micro_batch_size_per_gpu": 1, "steps_per_print": 1, "gradient_accumulation_steps": 1, "optimizer": { diff --git a/ALCF/examples/finetune_llama3/ds_config_empty.json b/ALCF/examples/finetune_llama3/ds_config_empty.json index e8ddb534a84..e50c73c1d7a 100755 --- a/ALCF/examples/finetune_llama3/ds_config_empty.json +++ b/ALCF/examples/finetune_llama3/ds_config_empty.json @@ -1,6 +1,6 @@ { - "train_batch_size" : 48, - "train_micro_batch_size_per_gpu": 8, + "train_batch_size" : 6, + "train_micro_batch_size_per_gpu": 1, "steps_per_print": 1, "gradient_accumulation_steps": 1, "optimizer": { diff --git a/finetune_llama.py b/finetune_llama.py index fa302a4f839..16f77b17d8e 100644 --- a/finetune_llama.py +++ b/finetune_llama.py @@ -2,11 +2,11 @@ """Finetune LLAMA, Modified from pretrain_gpt.py""" +import ezpz import torch import math from functools import partial from megatron import get_args -from megatron import print_rank_0 from megatron import get_timers from megatron import get_tokenizer from megatron.core import mpu, tensor_parallel @@ -29,14 +29,38 @@ import torch.nn.functional as F from transformers import AutoTokenizer +try: + RANK = ezpz.setup_torch('deepspeed') +except Exception: + RANK = 0 + +try: + import wandb +except (ImportError, ModuleNotFoundError): + wandb = None + +logger = ezpz.get_logger(__name__) + +# ---- [SETUP WANDB FROM RANK 0] -------------- +WANDB_MODE = os.environ.get("WANDB_MODE", None) +DISABLE_WANDB = WANDB_MODE is not None and str(WANDB_MODE).lower() == "disabled" +if RANK == 0 and not DISABLE_WANDB: + project_name = os.environ.get( + "WB_PROJECT", # look for WB_PROJECT in env + os.environ.get("WANDB_PROJECT", "AuroraGPT"), # look for WANDB_PROJECT in env + ) + logger.info(f"Setting up W&B from: {RANK} with {project_name}") + _ = ezpz.setup_wandb(project_name=project_name) + def model_provider(pre_process=True, post_process=True): """Build the model.""" - print_rank_0('building GPT model ...') - see_memory_usage(f"Before Building Model", force=True) + logger.info('building GPT model ...') + see_memory_usage("Before Building Model", force=True) args = get_args() + assert args is not None config = core_transformer_config_from_args(args) with deepspeed.zero.Init(sequence_data_parallel_group=mpu.get_sequence_data_parallel_group(), remote_device=None if args.remote_device == 'none' else args.remote_device, @@ -82,7 +106,29 @@ def model_provider(pre_process=True, post_process=True): pre_process=pre_process, post_process=post_process ) - see_memory_usage(f"After Building Model", force=True) + see_memory_usage("After Building Model", force=True) + if wandb is not None and getattr(wandb, "run", None) is not None: + assert wandb.run is not None + tbdir = args.tensorboard_dir + # tbdir = args.getattr('tensorboard_dir', None) + if tbdir is not None: + try: + logger.info(f"Patching tensorboard from {tbdir}") + wandb.tensorboard.patch(root_logdir=tbdir) # type:ignore + except ValueError as exc: + logger.exception(exc) + logger.warning("Continuing without patching tensorboard!") + num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) + wandb.run.config.update({"num_params": num_params}) + if "args" not in wandb.run.config: + logger.info( + f"Updating WandB run.config: [{wandb.run.name}]({wandb.run.get_url()})" + ) + try: + wandb.run.config.update({"args": dict(sorted(vars(args).items()))}) + except Exception: + logger.error('Unable to `wandb.run.config.update({"args": vars(args)})`') + return model @@ -220,7 +266,7 @@ def loss_func(loss_mask, moe_loss, mos_loss, output_tensor): return loss, {'total loss': loss, 'lm loss': averaged_loss[0], 'moe loss': moe_loss, 'mos loss': mos_loss} elif args.kd: return loss, {'total loss': loss, 'lm loss': averaged_loss[0], 'moe loss': moe_loss, 'kd loss': mos_loss} - print_rank_0('>>> total loss: {}, lm loss {}, kd loss {}'.format(loss, averaged_loss[0], mos_loss)) + logger.info('>>> total loss: {}, lm loss {}, kd loss {}'.format(loss, averaged_loss[0], mos_loss)) else: if max(args.num_experts) <= 1: return loss, {'lm loss': averaged_loss[0]} @@ -305,7 +351,7 @@ def prompt_train_valid_test_datasets_provider(train_val_test_num_samples): """Build train, valid, and test datasets.""" args = get_args() - print_rank_0('> building finetune prompt datasets ' + logger.info('> building finetune prompt datasets ' 'for llama ...') tokenizer = get_tokenizer() diff --git a/tools/hf2megads_weight_converter.py b/tools/hf2megads_weight_converter.py index fc1f82cb2ff..fa17f053a99 100644 --- a/tools/hf2megads_weight_converter.py +++ b/tools/hf2megads_weight_converter.py @@ -137,12 +137,20 @@ def load_and_print_hf_weight_auto(hf_ckpt_dir, no_init=True): logger.info( '----------------------------hf weight list----------------------------' ) + import gc hf_model = {} - for name, submodule in hf_auto_model.named_children(): - for pname, param in submodule.named_parameters(): - logger.info(f'[{name}.{pname}] shape={param.shape}') - hf_model[f'{name}.{pname}'] = param.clone() + with torch.no_grad(): + name, submodule = None, None + for name, submodule in hf_auto_model.named_children(): + pname, param = None, None + for pname, param in submodule.named_parameters(): + logger.info(f'[{name}.{pname}] shape={param.shape}') + hf_model[f'{name}.{pname}'] = param.clone() + del pname, param + gc.collect() + del submodule + gc.collect() torch.distributed.barrier() return hf_model From 3991f259064ccdefb90f4b024d4cf21bc8e04a2c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 28 Jan 2025 15:01:12 -0600 Subject: [PATCH 393/590] chore: Update `megatron/data/prompt_dataset.py` --- megatron/data/prompt_dataset.py | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/megatron/data/prompt_dataset.py b/megatron/data/prompt_dataset.py index 40a2949bdfe..b873e1dc3a4 100644 --- a/megatron/data/prompt_dataset.py +++ b/megatron/data/prompt_dataset.py @@ -1,7 +1,7 @@ +import ezpz # Utilizing code snippet from https://github.com/tatsu-lab/stanford_alpaca import copy -import logging from typing import Dict, Sequence import io import torch @@ -9,6 +9,10 @@ from torch.utils.data import Dataset import json + +logger = ezpz.get_logger(__name__) + + PROMPT_DICT = { "prompt_input": ( "Below is an instruction that describes a task, paired with an input that provides further context. " @@ -38,9 +42,9 @@ class SupervisedDataset(Dataset): def __init__(self, data_path: str, HFtokenizer): tokenizer = HFtokenizer.tokenizer super(SupervisedDataset, self).__init__() - logging.warning("Loading data...") + logger.warning("Loading data...") list_data_dict = jload(data_path) - logging.warning("Formatting inputs...") + logger.warning("Formatting inputs...") prompt_input, prompt_no_input = PROMPT_DICT["prompt_input"], PROMPT_DICT["prompt_no_input"] sources = [ prompt_input.format_map(example) if example.get("input", "") != "" else prompt_no_input.format_map(example) @@ -48,7 +52,7 @@ def __init__(self, data_path: str, HFtokenizer): ] targets = [f"{example['output']}{tokenizer.eos_token}" for example in list_data_dict] - logging.warning("Tokenizing inputs... This may take some time...") + logger.warning("Tokenizing inputs... This may take some time...") data_dict = preprocess(sources, targets, tokenizer) self.input_ids = data_dict["input_ids"] self.labels = data_dict["labels"] From 101d0ed0e5add6587cd73425777e5861d4fd9a78 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 31 Jan 2025 12:36:55 -0600 Subject: [PATCH 394/590] feat: Add `ALCF/examples/checkpoint_conversion/*` --- ALCF/examples/checkpoint_conversion/README.md | 72 ++++++++ .../checkpoint_conversion/hf_to_zero.py | 174 ++++++++++++++++++ 2 files changed, 246 insertions(+) create mode 100644 ALCF/examples/checkpoint_conversion/README.md create mode 100644 ALCF/examples/checkpoint_conversion/hf_to_zero.py diff --git a/ALCF/examples/checkpoint_conversion/README.md b/ALCF/examples/checkpoint_conversion/README.md new file mode 100644 index 00000000000..a020a953572 --- /dev/null +++ b/ALCF/examples/checkpoint_conversion/README.md @@ -0,0 +1,72 @@ +# Converting `AutoModel` to DeepSpeed ZeRO Checkpoint + +We would like to convert an (arbitrarily large) HuggingFace model to a ZeRO +checkpoint so that we can use it for continual pre-training with +Megatron-DeepSpeed. + +Previously, we had been using the approach from [ALCF / examples / finetune_llama3](/ALCF/examples/finetune_llama3/README.md) + + + + +| Model Name | Model Size | Model Parameters | Largest Layer Parameters | Memory Needed | +|:----------------------:|:----------:|:----------------:|:------------------------:|:-------------:| +| Llama-3.3-70B-Instruct | 70B | 69503M | 1050M | 70.45GB | + + +```bash +launch python3 \ + ALCF/examples/checkpoint_conversion/hf_to_zero.py \ + --zero-stage=3 \ + --device=cpu \ + --model='meta-llama/llama-3.3-70b-instruct' +``` + + + + + +## Estimate Memory Needs for Llama-3.3-70B-Instruct + +```bash +$ python3 -c 'from transformers import AutoModel; \ +∙ from deepspeed.runtime.zero.stage3 import estimate_zero3_model_states_mem_needs_all_live; \ +∙ model = AutoModel.from_pretrained("meta-llama/Llama-3.3-70B-Instruct"); \ +∙ estimate_zero3_model_states_mem_needs_all_live(model, num_gpus_per_node=12, num_nodes=4)' +``` + +
Output + + +```bash +Loading checkpoint shards: 100%|████████████████| 30/30 [08:28<00:00, 16.94s/it] +Estimated memory needed for params, optim states and gradients for a: +HW: Setup with 4 nodes, 12 GPUs per node. +SW: Model with 69503M total params, 1050M largest layer params. + per CPU | per GPU | Options + 436.93GB | 3.91GB | offload_param=cpu , offload_optimizer=cpu , zero_init=1 + 4660.54GB | 3.91GB | offload_param=cpu , offload_optimizer=cpu , zero_init=0 + 388.38GB | 6.61GB | offload_param=none, offload_optimizer=cpu , zero_init=1 + 4660.54GB | 6.61GB | offload_param=none, offload_optimizer=cpu , zero_init=0 + 70.45GB | 28.19GB | offload_param=none, offload_optimizer=none, zero_init=1 + 4660.54GB | 28.19GB | offload_param=none, offload_optimizer=none, zero_init=0 +took: 0h:08m:44s +``` + +
+ + +- Model States and Memory Needs for Llama-3.3-70B-Instruct: + + + | per CPU | per GPU | Options | + |:---------:|:-------:|:-------------------------------------------------------:| + | 436.93GB | 3.91GB | offload_param=cpu, offload_optimizer=cpu, zero_init=1 | + | 4660.54GB | 3.91GB | offload_param=cpu, offload_optimizer=cpu, zero_init=0 | + | 388.38GB | 6.61GB | offload_param=none, offload_optimizer=cpu, zero_init=1 | + | 4660.54GB | 6.61GB | offload_param=none, offload_optimizer=cpu, zero_init=0 | + | 70.45GB | 28.19GB | offload_param=none, offload_optimizer=none, zero_init=1 | + | 4660.54GB | 28.19GB | offload_param=none, offload_optimizer=none, zero_init=0 | + + + diff --git a/ALCF/examples/checkpoint_conversion/hf_to_zero.py b/ALCF/examples/checkpoint_conversion/hf_to_zero.py new file mode 100644 index 00000000000..8c7619eda65 --- /dev/null +++ b/ALCF/examples/checkpoint_conversion/hf_to_zero.py @@ -0,0 +1,174 @@ +from argparse import Namespace +import os +from pathlib import Path +from typing import Optional + +import ezpz +import torch +import torch.distributed +import deepspeed + +from transformers import AutoModelForCausalLM + +logger = ezpz.get_logger(__name__) + + +def parse_args(): + import argparse + + parser = argparse.ArgumentParser() + parser.add_argument( + '--model', type=str, default='meta-llama/Llama-3.2-1B-Instruct' + ) + parser.add_argument('--device', type=str, default=None, required=False) + parser.add_argument('--train-batch-size', type=int, default=1) + parser.add_argument('--zero-stage', type=int, default=3) + # add arg for output directory + parser.add_argument('--output-dir', type=str, default='.') + parser.add_argument('--kv-offload', action='store_true') + parser.add_argument('--async-kv-offload', action='store_true') + parser.add_argument('--gen-len', type=int, default=1024) + parser.add_argument('--strict', action='store_true') + return parser.parse_args() + + +def meta_to_cpu(container, dtype=None): + if isinstance(container, torch.Tensor): + return torch.empty(*container.shape, dtype=dtype or container.dtype) + elif isinstance(container, tuple): + return tuple(meta_to_cpu(x, dtype) for x in container) + elif isinstance(container, dict): + return dict((k, meta_to_cpu(v, dtype)) for k, v in container.items()) + else: + raise ValueError(f'Invalid type: {container}') + + +def get_model( + model_name: str = 'meta-llama/Llama-3.2-1B-Instruct', + dummy: Optional[bool] = None, + ignore_mismatched_sizes: bool = True, +) -> torch.nn.Module: + if dummy: + filename = Path('.').joinpath( + f'{model_name}.replace("/", "-")-hf-weights' + ) + if not filename.exists(): + from accelerate import init_empty_weights + + logger.info('Creating dummy weights') + with init_empty_weights(): + model = AutoModelForCausalLM.from_pretrained( + f'{model_name}', + ignore_mismatched_sizes=ignore_mismatched_sizes, + ) + model.save_pretrained( + filename, + state_dict=meta_to_cpu(model.state_dict(), torch.float16), + ) + return model + + model = AutoModelForCausalLM.from_pretrained( + f'{model_name}', + ignore_mismatched_sizes=ignore_mismatched_sizes, + ) + return model + + +def get_ds_config( + micro_batch_size: int = 1, + gradient_accumulation_steps: int = 2, + zero_stage: int = 3, + hidden_size: Optional[int] = None, +) -> dict: + train_batch_size = ( + micro_batch_size * ezpz.get_world_size() * gradient_accumulation_steps + ) + zero_config = { + 'stage': zero_stage, + } + if zero_stage == 3: + if hidden_size is not None: + zero_config |= { + 'stage3_prefetch_bucket_size': 2 * hidden_size * hidden_size, + 'stage3_param_persistence_threshold': hidden_size, + 'stage3_max_live_parameters': 2 * hidden_size * hidden_size, + } + zero_config |= { + 'offload_optimizer': { + 'device': 'cpu', + }, + 'offload_param': { + 'device': 'cpu', + }, + } + + return { + 'bf16': {'enabled': True}, + 'fp16': {'enabled': False}, + 'gradient_accumulation_steps': gradient_accumulation_steps, + 'optimizer': { + 'type': 'Adam', + }, + 'steps_per_print': 1, + 'train_batch_size': train_batch_size, + 'train_micro_batch_size_per_gpu': 1, + 'wall_clock_breakdown': True, + 'zero_optimization': zero_config, + } + + +def convert_checkpoint(args: Namespace): + if args.device is not None and args.device == 'cpu': + os.environ['TORCH_DEVICE'] = 'cpu' + os.environ['DS_ACCELERATOR'] = 'cpu' + + if args.zero_stage == 3: + cm = deepspeed.zero.Init() + else: + from contextlib import nullcontext + + cm = nullcontext() + + with cm: + with torch.no_grad(): + model = get_model( + args.model, ignore_mismatched_sizes=not args.strict + ) + + assert isinstance(model, torch.nn.Module) + if args.kv_offload: + model.set_kv_cache_offload( + True, + gen_len=args.gen_len, + async_kv_offload=args.async_kv_offload, + ) + + logger.info(f'model:\n{model}') + logger.info(f'{model.config=}') + ds_config = get_ds_config( + args.train_batch_size, + args.zero_stage, + hidden_size=model.config.hidden_size, + ) + output_dir = Path('zero-checkpoints').joinpath( + f'{args.model}-zs{args.zero_stage}-mb{args.train_batch_size}' + ) + + ds_engine = deepspeed.initialize(model=model, config_params=ds_config)[0] + ds_engine.module.eval() + model = ds_engine.module + logger.info(f'Saving ZeRO checkpoint to {output_dir}') + + ds_engine.save_checkpoint(output_dir) + + torch.distributed.barrier() + + +def main(): + _ = ezpz.setup_torch(backend='DDP') + args = parse_args() + convert_checkpoint(args) + + +if __name__ == '__main__': + main() From 10903ef7137eba87aaa21e44f1db14fa1ed90aa8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 31 Jan 2025 13:27:19 -0600 Subject: [PATCH 395/590] docs: Update `ALCF/examples/checkpoint_conversion/README.md` --- ALCF/examples/checkpoint_conversion/README.md | 56 +++++++++++++++++-- 1 file changed, 51 insertions(+), 5 deletions(-) diff --git a/ALCF/examples/checkpoint_conversion/README.md b/ALCF/examples/checkpoint_conversion/README.md index a020a953572..97debad17e6 100644 --- a/ALCF/examples/checkpoint_conversion/README.md +++ b/ALCF/examples/checkpoint_conversion/README.md @@ -4,15 +4,47 @@ We would like to convert an (arbitrarily large) HuggingFace model to a ZeRO checkpoint so that we can use it for continual pre-training with Megatron-DeepSpeed. -Previously, we had been using the approach from [ALCF / examples / finetune_llama3](/ALCF/examples/finetune_llama3/README.md) +Previously, we had been using the approach from [ALCF / examples / +finetune_llama3](/ALCF/examples/finetune_llama3/README.md). +In particular, this approach works by: +1. Instantiate the Megatron-DeepSpeed (MDS) model as normal (with empty + weights), from [\[here\]](/tools/hf2megads_weight_converter.py#L712) + ```python + from megatron.model import GPTModelPipe + ds_model = GPTModelPipe(config, num_tokentypes=0, parallel_output=True) + ``` -| Model Name | Model Size | Model Parameters | Largest Layer Parameters | Memory Needed | -|:----------------------:|:----------:|:----------------:|:------------------------:|:-------------:| -| Llama-3.3-70B-Instruct | 70B | 69503M | 1050M | 70.45GB | +1. Instantiate the HF model \[[here\]](/tools/hf2megads_weight_converter.py#L725) + + ```python + from transformers import AutoModel + hf_model = AutoModel.from_pretrained("meta-llama/llama-3.3-70b-instruct") + ``` + +3. Instantiate optimizer [\[here\]](/tools/hf2megads_weight_converter.py#L736) + +1. Layer by layer, copy the weights from the HF model to the MDS model + \[[here\]](/tools/hf2megads_weight_converter.py#L766) + + +Unfortunately, for very large models, this will slowly consume available host +memory until it is exhausted causing the application to crash. + +## Proposed Solution + +Our proposed solution is simple and entirely contained in [ALCF / examples / checkpoint_conversion / hf_to_zero.py](/ALCF/examples/checkpoint_conversion/hf_to_zero.py). +Explicitly: + +1. Create the HF model as normal +2. Pass it to `deepspeed.initalize(...)` to create the `DeepSpeedEngine` +3. `DeepSpeedEngine.save_checkpoint(...)` to save the checkpoint. + + +To run: ```bash launch python3 \ @@ -22,11 +54,25 @@ launch python3 \ --model='meta-llama/llama-3.3-70b-instruct' ``` +> [!WARNING] +> I believe this approach is still not finished because I expect there will be +> naming mismatches between the layers of the HF model (now saved in our ZeRO +> checkpoint) and what our MDS model expects. +> +> This requires further testing to confirm, but we are now able to successfully +> convert the 70B model to a ZeRO checkpoint. +## Estimate Memory Needs for Llama-3.3-70B-Instruct +Deepspeed provides a nice mechanism for determining the memory needs of a model. + +We provide below the summary for the Llama-3.3-70B-Instruct model of interest. + +| Model Name | Model Size | Model Parameters | Largest Layer Parameters | Memory Needed | +|:----------------------:|:----------:|:----------------:|:------------------------:|:-------------:| +| Llama-3.3-70B-Instruct | 70B | 69503M | 1050M | 70.45GB | -## Estimate Memory Needs for Llama-3.3-70B-Instruct ```bash $ python3 -c 'from transformers import AutoModel; \ From 19b3b741580737fdd9aead21d980be454b809dde Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 12 Mar 2025 10:19:27 -0500 Subject: [PATCH 396/590] Update README.md --- ALCF/README.md | 1110 ++++-------------------------------------------- 1 file changed, 73 insertions(+), 1037 deletions(-) diff --git a/ALCF/README.md b/ALCF/README.md index e88cebc8ebe..f89b6daccbf 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -6,1069 +6,106 @@ ## 🏃‍♂️ Running -To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): - -1.
⏳ Request an interactive job with qsub -I: - - ```bash - qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I - ``` - - - Or, alternatively, you can submit [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) - directly as a batch script with - - ```bash - cd Megatron-DeepSpeed - qsub -A -q debug -l select=2 -l walltime=01:00:00:filesystems=eagle:home train_aGPT_7B.sh - ``` - -
- -2.
⬇️ Clone repo + navigate into it: +1. Clone the [argonne-lcf / `Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed) repository: ```bash - git clone "https://github.com/argonne-lcf/Megatron-DeepSpeed" + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed cd Megatron-DeepSpeed ``` -
- -3.
🐍 Setup Python: - -
- - > **NOTE**: The following commands should be ran from [`Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed), following the `cd` command from 2. - - 1. Load `conda` module and activate base environment: - - ```bash - export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup - ``` - - -
[output]: - -
- - -
[Polaris]: - - ```bash - # [05:47:13 PM][foremans@x3001c0s13b1n0][/eagle/a/f/p/ar/Megatron-DeepSpeed-D/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python - Using WORKING_DIR: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed - No conda_prefix or virtual_env found in environment... - Setting up conda... - Running on Polaris !! - - Lmod is automatically replacing "nvhpc/23.9" with "gcc-native/12.3". - - - Lmod is automatically replacing "PrgEnv-nvhpc/8.5.0" with "PrgEnv-gnu/8.5.0". - - - Due to MODULEPATH changes, the following have been reloaded: - 1) cray-mpich/8.1.28 - - Found conda at: /soft/applications/conda/2024-04-29/mconda3 - No VIRTUAL_ENV found in environment! - - Trying to setup from /soft/applications/conda/2024-04-29/mconda3 - - Using VENV_DIR=/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 - - Found existing venv, activating from /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 - [python] Using: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29/bin/python3 - ``` - -
- - -
[Aurora]: - - ```bash - # [10:04:02 PM][foremans@x4415c0s2b0n0][/gecko/A/fo/p/a/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python - Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed - No conda_prefix or virtual_env found in environment... - Setting up conda... - - The following have been reloaded with a version change: - 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/803.29 2) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.1 - - Found conda at: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 - No VIRTUAL_ENV found in environment! - - Trying to setup from /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 - - Using VENV_DIR=/gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 - - Found existing venv, activating from /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 - [python] Using: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 - ``` - -
- - -
[Sunspot]: - - ```bash - # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed] - $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python - Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed - No conda_prefix or virtual_env found in environment... - Setting up conda... - Running on SunSpot !! - - Due to MODULEPATH changes, the following have been reloaded: - 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 - 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 - 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 - 4) mpfr/4.2.0-w7v7yjv - - The following have been reloaded with a version change: - 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 - 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 - UMD: agama-ci-devel-803.29 successfully loaded: - UMD: graphics-compute-runtime/agama-ci-devel-803.29 - - The following have been reloaded with a version change: - 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 - - Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 - No VIRTUAL_ENV found in environment! - - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 - - Using VENV_DIR=/gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 - - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 - [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 - ``` - -
- - - - 2. 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): - - ```bash - mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz - python3 -m pip install -e deps/ezpz --require-virtualenv - ``` - - [^venv]: Its generally a good practice to keep separate virtual Python environments different projects. - We provide a helper function, [`setup_venv_from_conda()`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/2f0154394bbdf3c64b4669f9d944645e2cdb8f2b/ALCF/helpers.sh#L440), - that helps take care of this for you. -
- This will: activate (or build, if necessary) a `venv` in your working dir, - _automatically_ matching the name of your active `conda` environment (e.g. `2024-04-29`, on Polaris_. - - 3. Setup [`wandb`](https://docs.wandb.ai/quickstart) - - > **NOTE**: this can be disabled by setting `export WANDB_DISABLED=1` - -
- - - -4.
🚀 Launch: - - In this case, train a ~ 2B Model (with 10 layers), - for 1000 iterations using the data file list in: - - [`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) - - with a micro-batch-size of 2 (`MICRO_BATCH=2`), with the `torch.optim.AdamW` optimizer (`OPT=adamw`). - - **Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) - function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) - can be overridden dynamically at runtime using this technique. - - ```bash - # for systems other than Polaris, replace "polaris/books.txt" below with: - # "{aurora,sunspot}/books.txt", - PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_aGPT_7B.sh - ``` - - - **Note**: If no additional options specified, i.e. - - ```bash - PBS_O_WORKDIR=$(pwd) bash train_aGPT_7B.sh - ``` - - then this will fallback to using the default AuroraGPT-7B architecture - with the full Dolma (v1.7) dataset. - -
[output]: - -
- - The outputs should look _something_ like this, though YMMV (things change quick): - -
[Aurora]: - - ```bash - #[🌌][10:45:59 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?] - $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python - - #[🌌][10:46:57 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?][aurora_nre_models_frameworks-2024.1] - (aurora_nre_models_frameworks-2024.1) $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt bash train_aGPT_7B.sh > train-log-$(tstamp).log 2>&1 & - - Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed - Running on: aurora - Using virtual_env: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 on top of conda from: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 - [python] Using: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 - Ensuring all dependencies from /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/requirements/requirements.txt installed... - - [notice] A new release of pip is available: 24.0 -> 24.1 - [notice] To update, run: pip install --upgrade pip - ┌─────────────────────────────────────────────────────────────────────┐ - │ [savejobenv]: - │ • Writing PBS vars to: /home/foremans/.pbsenv - └─────────────────────────────────────────────────────────────────────┘ - ┌─────────────────────────────────────────────────────────────────────┐ - │ [HOSTS]: - │ • [host:0] - x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov - │ • [host:1] - x4711c1s3b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov - └─────────────────────────────────────────────────────────────────────┘ - ┌─────────────────────────────────────────────────────────────────────┐ - │ [DIST INFO]: - │ • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov - │ • NHOSTS=2 - │ • NGPU_PER_HOST=12 - │ • NGPUS=24 - └─────────────────────────────────────────────────────────────────────┘ - ┌─────────────────────────────────────────────────────────────────────┐ - │ [LAUNCH]: - │ • To launch across all available GPUs, use: - │ 'launch' ( = mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov ) - └─────────────────────────────────────────────────────────────────────┘ - 2024-06-21 10:47:09,771 - numexpr.utils - INFO - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. - 2024-06-21 10:47:09,772 - numexpr.utils - INFO - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. - 2024-06-21 10:47:09,772 - numexpr.utils - INFO - NumExpr defaulting to 8 threads. - /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages/pandas/core/computation/expressions.py:21: UserWarning: Pandas requires version '2.8.4' or n> - from pandas.core.computation.check import NUMEXPR_INSTALLED - /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1/lib/python3.9/runpy.py:127: RuntimeWarning: 'ezpz.jobs' found in sys.modules after import of package 'ezpz', but prior to execution of 'ezpz.jobs'; this may result in u> - warn(RuntimeWarning(msg)) - [2024-06-21 10:47:10][INFO][jobs:366] - Caught PBS_JOBID='684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov', pbsnf=PosixPath('/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov') from env. Saving jobenv! - [2024-06-21 10:47:10][WARNING][jobs:117] - /home/foremans/PBS-jobs/684084 already in /home/foremans/PBS-jobs.log, not appending !! - [2024-06-21 10:47:10][INFO][jobs:192] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.sh - [2024-06-21 10:47:10][INFO][jobs:220] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.json - [2024-06-21 10:47:10][INFO][jobs:233] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.yaml - [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /home/foremans/PBS-jobs/684084/.jobenv - [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv - [2024-06-21 10:47:10][WARNING][jobs:154] - To use launch alias, be sure to: source /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv - [2024-06-21 10:47:10][INFO][jobs:277] - Writing PBS env vars to /home/foremans/PBS-jobs/684084 / jobenv{.sh, .yaml, .json} - [2024-06-21 10:47:10][WARNING][jobs:281] - Run: source ./.jobenv in your current shell to set job variables - [2024-06-21 10:47:10][INFO][jobs:374] - - [DIST_INFO]: - • DEVICE=xpu - • DEVICE_ID=xpu:0 - • DISTRIBUTED_BACKEND=ccl - • GPUS_PER_NODE=12 - • HOSTS=['x4711c1s2b0n0', 'x4711c1s3b0n0'] - • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov - • HOSTNAME=x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov - • LOCAL_RANK=0 - • MACHINE=Aurora - • NUM_NODES=2 - • NGPUS=24 - • NODE_ID=0 - • RANK=0 - • SCHEDULER=PBS - • WORLD_SIZE_TOTAL=24 - • WORLD_SIZE_IN_USE=1 - [2024-06-21 10:47:10][CRITICAL][jobs:245] - To launch across ALL GPUs in your job, use: - LAUNCH_CMD=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov - creating alias launch=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov - Found ezpz! - - [notice] A new release of pip is available: 24.0 -> 24.1 - [notice] To update, run: pip install --upgrade pip - Done with ezpz. - Not using flash-attn!! - LR_ARGS: --lr 0.0003 --lr-decay-style cosine --lr-warmup-fraction 0.05 - DS_CONFIG: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ds-configs/ds_stage1_mb4_gb768_pp1_bf16.json - ZS: 1, MB: 4, GB: 768, PP: 1, DTYPE: bf16 - Please see logs at: logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov - Checkpoints will be saved to: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05 - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - Setting up tokenizer with Llama2 - Using data_file_list: ./ALCF/data-lists/aurora/books.txt - Using tokenizer: Llama2. Setting up data with ./ALCF/data-lists/aurora/books.txt - Calling: setData() with ./ALCF/data-lists/aurora/books.txt - -------------------- - Updated environment: - DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt - NUM_DOCS: 3 - WEIGHT_SUM: 0.0072042092147565125 - DFL_STEM: books - DATA_CACHE_PATH: .cache/books/index-cache - DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt - -------------------- - [setData] DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt - [setData] TOKENIZER_FLAGS: --tokenizer-type Llama2Tokenizer --tokenizer-model /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokenizer.model - Requirement already satisfied: pybind11 in ./venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages (2.12.0) - - [notice] A new release of pip is available: 24.0 -> 24.1 - [notice] To update, run: pip install --upgrade pip - make: Nothing to be done for 'default'. - /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed - ++++++++++++++++++++++++++++++++++++++++++++++++++ - - MPICH_DIR=/opt/aurora/24.086.0/CNDA/mpich/20231026/mpich-ofi-all-icc-default-pmix-gpu-drop20231026 - - Using /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 - - WORLD_SIZE:24 - - BACKEND: ccl - - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden - - Using DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt - ++++++++++++++++++++++++++++++++++++++++++++++++++ - - Currently Loaded Modules: - 1) mpich/icc-all-pmix-gpu/20231026 3) libfabric/1.15.2.0 5) cray-libpals/1.3.3 7) gmp/6.2.1-pcxzkau 9) mpc/1.3.1-dfagrna 11) intel_compute_runtime/release/803.29 13) frameworks/2024.1 - 2) mpich-config/collective-tuning/1024 4) cray-pals/1.3.3 6) spack-pe-gcc/0.7.0-24.086.0 8) mpfr/4.2.0-w7v7yjv 10) gcc/12.2.0 12) oneapi/release/2024.1 - - - - Saving environment to checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.env - Not currently running. Continuing! - Launching with: MPICH - mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py - Using data_cache_path: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache - - mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --no-bias-dropout-fusion --no-masked-softmax-fusion --no-gradient-accumulation-fusion > - - [!! NOTE] View output at: - logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log - Connected to tcp://x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov:7919 - Launching application eafe3e80-ad2e-4cee-a3e4-d63af2a77c66 - [2024-06-21 10:47:31,610] [INFO] [comm.py:161:init_deepspeed_backend] Initialize ccl backend - [2024-06-21 10:47:31,610] [INFO] [comm.py:637:init_distributed] cdb=None - [2024-06-21 10:47:31,610] [INFO] [comm.py:652:init_distributed] Not using the DeepSpeed or dist launchers, attempting to detect MPI environment... - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=15, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=23, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=12, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=13, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=3, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=14, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=16, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=17, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=18, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=19, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=20, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=21, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=22, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=6, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=8, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=10, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=0, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:668:init_distributed] Initializing TorchBackend in DeepSpeed with backend ccl - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=1, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=2, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=4, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=5, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=7, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=9, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=11, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=2/23][local_rank=2/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=6/23][local_rank=6/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=8/23][local_rank=8/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=1/23][local_rank=1/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=15/23][local_rank=3/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=17/23][local_rank=5/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=3/23][local_rank=3/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=13/23][local_rank=1/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=14/23][local_rank=2/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=4/23][local_rank=4/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=18/23][local_rank=6/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=20/23][local_rank=8/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=5/23][local_rank=5/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=21/23][local_rank=9/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=7/23][local_rank=7/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=23/23][local_rank=11/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=9/23][local_rank=9/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=12/23][local_rank=0/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=16/23][local_rank=4/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=10/23][local_rank=10/11][node=0/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=11/23][local_rank=11/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=19/23][local_rank=7/11][node=1/1] - [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=22/23][local_rank=10/11][node=0/1] - 2024-06-21 10:47:32][INFO][dist:240] - DistInfo={ - "DEVICE": "xpu", - "DEVICE_ID": "xpu:0", - "DISTRIBUTED_BACKEND": "ccl", - "GPUS_PER_NODE": 12, - "HOSTFILE": "/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov", - "HOSTNAME": "x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov", - "HOSTS": "['x4711c1s2b0n0', 'x4711c1s3b0n0']", - "LOCAL_RANK": 0, - "MACHINE": "Aurora", - "NGPUS": 24, - "NODE_ID": 0, - "NUM_NODES": 2, - "RANK": 0, - "SCHEDULER": "PBS", - "WORLD_SIZE_IN_USE": 24, - "WORLD_SIZE_TOTAL": 24 - } - - # [...clipped...] - - [2024-06-21 10:48:48][INFO][utils:307] - > elapsed time for building blendable dataset indices: 1.19 (sec) - [2024-06-21 10:48:48][INFO][utils:307] - > saving index map files - [2024-06-21 10:48:51][INFO][utils:307] - > finished saving index map files in 3.0829622745513916 seconds - [2024-06-21 10:48:51][INFO][utils:307] - > loading blendable dataset index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_index.npy - [2024-06-21 10:48:52][INFO][utils:307] - > loading blendable dataset sample index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_sample_index.npy - [2024-06-21 10:48:52][INFO][utils:307] - > finished loading in 0.30188989639282227 seconds - [2024-06-21 10:48:52][INFO][utils:307] - >> building dataset for /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document - [2024-06-21 10:48:52][INFO][utils:307] - > building dataset index ... - [2024-06-21 10:48:52][INFO][utils:307] - reading sizes... - [2024-06-21 10:48:52][INFO][utils:307] - reading pointers... - [2024-06-21 10:48:52][INFO][utils:307] - reading document index... - [2024-06-21 10:48:52][INFO][utils:307] - creating numpy buffer of mmap... - [2024-06-21 10:48:52][INFO][utils:307] - /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document.bin - [2024-06-21 10:48:52][INFO][utils:307] - creating memory view of numpy buffer... - [2024-06-21 10:48:52][INFO][utils:307] - > finished creating indexed dataset in 0.003112 seconds - [2024-06-21 10:48:52][INFO][utils:307] - number of documents: 7386 - [2024-06-21 10:48:52][INFO][utils:307] - > dataset split: - [2024-06-21 10:48:52][INFO][utils:307] - train: - [2024-06-21 10:48:52][INFO][utils:307] - document indices in [0, 7386) total of 7386 documents - [2024-06-21 10:48:52][INFO][utils:307] - validation: - [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents - [2024-06-21 10:48:52][INFO][utils:307] - test: - [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents - [2024-06-21 10:48:52][INFO][utils:307] - > loading doc-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_doc_idx.npy - [2024-06-21 10:48:52][INFO][utils:307] - > loading sample-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_sample_idx.npy - [2024-06-21 10:48:52][INFO][utils:307] - > loading shuffle-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_shuffle_idx.npy - [2024-06-21 10:48:52][INFO][utils:307] - loaded indexed file in 0.008 seconds - [2024-06-21 10:48:52][INFO][utils:307] - total number of samples: 34196233 - [2024-06-21 10:48:52][INFO][utils:307] - total number of epochs: 175 - [2024-06-21 10:48:52][INFO][utils:307] - > size of blendable dataset: 245361763 samples - [2024-06-21 10:48:52][INFO][utils:307] - >>> Finished building BlendableDataset in 4.613574266433716 seconds - [2024-06-21 10:48:52][INFO][pretrain_gpt_alcf:579] - > finished creating GPT datasets. Took: 45730179865763.24219s - [2024-06-21 10:48:53][INFO][training:88] - [after dataloaders are built] datetime=2024-06-21 10:48:53 - [2024-06-21 10:48:53][INFO][training:307] - done with setup ... - [2024-06-21 10:48:53][INFO][training:313] - training ... - (min, max) time across ranks (ms): - model-and-optimizer-setup ......................: (63763.34, 63857.25) - train/valid/test-data-iterators-setup ..........: (12936.53, 13432.64) - [2024-06-21 10:48:53][INFO][training:88] - [before the start of training step] datetime=2024-06-21 10:48:53 - [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:541:forward] Activation Checkpointing Information - [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:542:forward] ----Partition Activations False, CPU CHECKPOINTING False - [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:543:forward] ----contiguous Memory Checkpointing False with 32 total layers - [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:545:forward] ----Synchronization False - [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:546:forward] ----Profiling time in checkpointing False - [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1867.64 | optimizer_gradients: 19.65 | optimizer_step: 46.07 - [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[1.887433467970254e-08, 1.887433467970254e-08], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 25341.72 | bwd_microstep: 77707.38 | bwd_inner_microstep: 75751.84 | bwd_allreduce_microstep: 1955.54 | step_microstep: 2218.38 - [2024-06-21 10:50:42,168] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 25341.72 | bwd: 77707.38 | bwd_inner: 75751.84 | bwd_allreduce: 1955.54 | step: 2218.38 - [2024-06-21 10:50:42][INFO][training:1609] - iteration= 1/ 317892 | consumed_samples= 768 | consumed_tokens= 3145728 | elapsed_time_per_iteration_ms=108893.2 | learning_rate=1.88743e-08 | global_batch_size= 768 | lm loss=11.133188 | loss_scale=1.0 | actual_seqlen= 4096 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=7.053 | tokens_per_gpu_per_second_tgs=1203.674 | [LM]-TFLOPs=49.66 | [DS]-TFLOPs=73.32 | - [2024-06-21 10:50:42][INFO][utils:190] - [Rank 0] (after 1 iterations) memory (MB) | allocated: 18243.64111328125 | max allocated: 50664.2548828125 | reserved: 54556.0 | max reserved: 54556.0 - (min, max) time across ranks (ms): - forward-backward ...............................: (106622.81, 106624.28) - optimizer ......................................: (2221.02, 2234.98) - ``` - -
- -
[Sunspot]: - - ```bash - # [09:07:32 AM][foremans@x1921c0s0b0n0][~/q/llm.devkit/Megatron-DeepSpeed][🌱 main][$!?] - $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_aGPT_7B.sh - source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh - Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... - UMD: agama-ci-devel-736.9 successfully loaded: - UMD: graphics-compute-runtime/agama-ci-devel-736.9 - Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" - - Please check the spelling or version number. Also try "module spider ..." - It is also possible your cache file is out-of-date; it may help to try: - $ module --ignore_cache load "gcc/12.1.0" - - Also make sure that all modulefiles written in TCL start with the string #%Module - - Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. - - Running on SunSpot !! - [python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 - Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env - Found ezpz! - /lus/gila/projects/Aurora_deployment/foremans/locations/sunspot/projects/saforem2/ezpz/src/ezpz/__init__.py - Has ezpz installed. Nothing to do. - Done with ezpz. - ┌─────────────────────────────────────────────────────────────────── - │ Writing PBS vars to /home/foremans/.pbsenv - │ HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 - │ NHOSTS: 2 - │ NGPU_PER_HOST: 12 GPUs per host - │ NGPUS: 24 GPUs total - └─────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [Hosts]: - │ • [host:0] - x1921c0s0b0n0.hostmgmt2000.cm.americas.sgi.com - │ • [host:1] - x1921c0s1b0n0.hostmgmt2000.cm.americas.sgi.com - └────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [DIST INFO]: - │ • Loading job env from: /home/foremans/.pbsenv - │ • HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 - │ • NHOSTS: 2 - │ • NGPU_PER_HOST: 12 - │ • NGPUS (NHOSTS x NGPU_PER_HOST): 24 - │ • WORLD_SIZE: 24 - │ • DIST_LAUNCH: mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001 - └────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [Launch]: - │ • Use: 'launch' (=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001) - │ to launch job - └────────────────────────────────────────────────────────────────── - DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json - ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - Calling: setData() with ./convergence_debug_small.txt - -------------------- - Updated environment: - DATA_FILE_LIST: ./convergence_debug_small.txt - NUM_DOCS: 15 - WEIGHT_SUM: 15.0 - DFL_STEM: convergence_debug_small - DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache - -------------------- - ++++++++++++++++++++++++++++++++++++++++++++++++++ - - MPICH_DIR= - - Using /home/foremans/miniconda3/envs/q4-drop/bin/python3 - - WORLD_SIZE:24 - - NCCL: nccl - - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden - - Using DATA_FILE_LIST: ./convergence_debug_small.txt - ++++++++++++++++++++++++++++++++++++++++++++++++++ - ! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed - /home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html - __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') - /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you dont plan on using image function - ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch - vision` from source? - warn( - [2024-04-04 09:07:45,585] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) - [2024-04-04 09:07:45,818] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) - -------------------------------------------------- - DeepSpeed C++/CUDA extension op report - -------------------------------------------------- - NOTE: Ops not installed will be just-in-time (JIT) compiled at - runtime if needed. Op compatibility means that your system - meet the required dependencies to JIT install the op. - -------------------------------------------------- - JIT compiled ops requires ninja - ninja .................. [OKAY] - -------------------------------------------------- - op name ................ installed .. compatible - -------------------------------------------------- - async_io ............... [NO] ....... [OKAY] - cpu_adagrad ............ [NO] ....... [OKAY] - cpu_adam ............... [NO] ....... [OKAY] - flash_attn ............. [NO] ....... [OKAY] - fused_adam ............. [NO] ....... [OKAY] - quantizer .............. [NO] ....... [OKAY] - transformer ............ [NO] ....... [OKAY] - transformer_inference .. [NO] ....... [OKAY] - utils .................. [NO] ....... [OKAY] - -------------------------------------------------- - DeepSpeed general environment info: - torch install path ............... ['/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch'] - torch version .................... 2.1.0a0+cxx11.abi - deepspeed install path ........... ['/lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/DeepSpeed/deepspeed'] - deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD - deepspeed wheel compiled w. ...... torch 2.1 - shared memory (/dev/shm) size .... 503.18 GB - - deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ - foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay - -style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 - --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/tensorboard --log-timers-to-tensorboard --log-optimizer - -states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 - --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend ccl --num-attention-heads 32 --save-interval 20 - 0 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 4 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 1 --global-bat - ch-size 96 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ - .cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/tokenizer.model --no-query- - key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z - ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 - _nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - - [!! NOTE] View output at: - logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log - - # ... - - /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin - creating memory view of numpy buffer... - > finished creating indexed dataset in 0.010017 seconds - number of documents: 1498927 - > dataset split: - train: - document indices in [0, 1498927) total of 1498927 documents - validation: - document indices in [1498927, 1498927) total of 0 documents - test: - document indices in [1498927, 1498927) total of 0 documents - > loading doc-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_doc_idx.npy - > loading sample-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_sample_idx.npy - > loading shuffle-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_shuffle_idx.npy - loaded indexed file in 0.056 seconds - total number of samples: 2318461 - total number of epochs: 8 - > loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy - > loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy - /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. - - [after dataloaders are built] datetime: 2024-04-04 09:09:27 - done with setup ... - (min, max) time across ranks (ms): - model-and-optimizer-setup ......................: (64818.18, 64858.22) - train/valid/test-data-iterators-setup ..........: (1968.10, 2288.56) - training ... - [before the start of training step] datetime: 2024-04-04 09:09:27 - [2024-04-04 09:09:27,718] [INFO] [checkpointing.py:540:forward] Activation Checkpointing Information - [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:541:forward] ----Partition Activations False, CPU CHECKPOINTING False - [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:542:forward] ----contiguous Memory Checkpointing False with 32 total layers - [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:544:forward] ----Synchronization False - [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:545:forward] ----Profiling time in checkpointing False - [2024-04-04 09:09:33][INFO][utils:145] - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. - [2024-04-04 09:09:33][INFO][utils:148] - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. - [2024-04-04 09:09:33][INFO][utils:160] - NumExpr defaulting to 8 threads. - ^[c[2024-04-04 09:09:53,311] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 884.11 | optimizer_gradients: 6.43 | optimizer_step: 23.44 - [2024-04-04 09:09:53,312] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 6567.68 | bwd_microstep: 17950.36 | bwd_inner_microstep: 17711.20 | bwd_allreduce_microstep: 239.11 | step_microstep: 1139.27 - [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 6567.66 | bwd: 17950.35 | bwd_inner: 17711.19 | bwd_allreduce: 239.11 | step: 1139.29 - [Rank 0] (after 1 iterations) memory (MB) | allocated: 18244.640625 | max allocated: 41299.50146484375 | reserved: 46764.0 | max reserved: 46764.0 - iteration 1/ 317892 | consumed samples: 96 | consumed tokens: 393216 | elapsed time per iteration (ms): 25849.1 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 1.117136E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.714 | tokens per gpu per second(tgs): 633.832 | TFLOPs: 38.61 | - [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 327.85 | optimizer_gradients: 6.26 | optimizer_step: 23.60 - [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 4022.74 | bwd_microstep: 15738.67 | bwd_inner_microstep: 15556.80 | bwd_allreduce_microstep: 181.82 | step_microstep: 371.01 - [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 - iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | - ``` - -
- -
[Polaris]: - - ```bash - # [09:31:35 AM][foremans@x3112c0s13b0n0][~/pol/p/a/Megatron-DeepSpeed][🌱 main][$!?] - $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_aGPT_7B.sh - source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh - Running on Polaris !! - - [python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 - Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env - Found ezpz! - /lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/ezpz/src/ezpz/__init__.py - Has ezpz installed. Nothing to do. - Done with ezpz. - ┌─────────────────────────────────────────────────────────────────── - │ Writing PBS vars to /home/foremans/.pbsenv - │ HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov - │ NHOSTS: 2 - │ NGPU_PER_HOST: 4 GPUs per host - │ NGPUS: 8 GPUs total - └─────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [Hosts]: - │ • [host:0] - x3112c0s13b0n0.hsn.cm.polaris.alcf.anl.gov - │ • [host:1] - x3112c0s13b1n0.hsn.cm.polaris.alcf.anl.gov - └────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [DIST INFO]: - │ • Loading job env from: /home/foremans/.pbsenv - │ • HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov - │ • NHOSTS: 2 - │ • NGPU_PER_HOST: 4 - │ • NGPUS (NHOSTS x NGPU_PER_HOST): 8 - │ • WORLD_SIZE: 8 - │ • DIST_LAUNCH: mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov - └────────────────────────────────────────────────────────────────── - ┌────────────────────────────────────────────────────────────────── - │ [Launch]: - │ • Use: 'launch' (=mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov) - │ to launch job - └────────────────────────────────────────────────────────────────── - DS_CONFIG: ds_stage2_mb8_gb32_pp1_bf16.json - ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! - Calling: setData() with "./convergence_debug_small.txt" - -------------------- - Updated environment: - DATA_FILE_LIST: ./convergence_debug_small.txt - NUM_DOCS: 15 - WEIGHT_SUM: 15.0 - DFL_STEM: convergence_debug_small - DATA_CACHE_PATH: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache - -------------------- - ++++++++++++++++++++++++++++++++++++++++++++++++++ - - MPICH_DIR=/opt/cray/pe/mpich/8.1.25/ofi/gnu/9.1 - - Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 - - WORLD_SIZE:8 - - NCCL: nccl - - MODEL_TYPE: llama-seq4096-pp1-tp2-32layers-32heads-4096hidden - - Using DATA_FILE_LIST: ./convergence_debug_small.txt - ++++++++++++++++++++++++++++++++++++++++++++++++++ - ! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed - [2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda [auto detect] - -------------------------------------------------- - DeepSpeed C++/CUDA extension op report - -------------------------------------------------- - NOTE: Ops not installed will be just-in-time (JIT) compiled at - runtime if needed. Op compatibility means that your system - meet the required dependencies to JIT install the op. - -------------------------------------------------- - JIT compiled ops requires ninja - ninja .................. [OKAY] - -------------------------------------------------- - op name ................ installed .. compatible - -------------------------------------------------- - async_io ............... [NO] ....... [OKAY] - fused_adam ............. [NO] ....... [OKAY] - cpu_adam ............... [NO] ....... [OKAY] - cpu_adagrad ............ [NO] ....... [OKAY] - cpu_lion ............... [NO] ....... [OKAY] - [WARNING] Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH - evoformer_attn ......... [NO] ....... [NO] - fused_lamb ............. [NO] ....... [OKAY] - fused_lion ............. [NO] ....... [OKAY] - inference_core_ops ..... [NO] ....... [OKAY] - cutlass_ops ............ [NO] ....... [OKAY] - transformer_inference .. [NO] ....... [OKAY] - quantizer .............. [NO] ....... [OKAY] - ragged_device_ops ...... [NO] ....... [OKAY] - ragged_ops ............. [NO] ....... [OKAY] - random_ltd ............. [NO] ....... [OKAY] - [WARNING] sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.2 - [WARNING] using untested triton version (2.2.0), only 1.0.0 is known to be compatible - sparse_attn ............ [NO] ....... [NO] - spatial_inference ...... [NO] ....... [OKAY] - transformer ............ [NO] ....... [OKAY] - stochastic_transformer . [NO] ....... [OKAY] - -------------------------------------------------- - DeepSpeed general environment info: - torch install path ............... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/torch'] - torch version .................... 2.2.1 - deepspeed install path ........... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/deepspeed'] - deepspeed info ................... 0.14.0, unknown, unknown - torch cuda version ............... 11.8 - torch hip version ................ None - nvcc version ..................... 11.8 - deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 - shared memory (/dev/shm) size .... 251.61 GB - - deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc - e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion - --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- - grads-in-fp32 --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/tensorboard --log-timers-to-tensorboard - - -log-optimizer-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_ - pp1_tp2_bf16 --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend nccl --num-attention-heads 32 --s - ave-interval 200 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 8 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 2 - --global-batch-size 32 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-l - cf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokeniz - er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 - --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- - num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - - [!! NOTE] View output at: - logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log - - # ... - - /eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin - creating memory view of numpy buffer... - > finished creating indexed dataset in 0.001280 seconds - number of documents: 1498927 - > dataset split: - train: - document indices in [0, 1498927) total of 1498927 documents - validation: - document indices in [1498927, 1498927) total of 0 documents - test: - document indices in [1498927, 1498927) total of 0 documents - > loading doc-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_doc_idx.npy - > loading sample-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_sample_idx.npy - > loading shuffle-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_shuffle_idx.npy - loaded indexed file in 0.004 seconds - total number of samples: 869423 - total number of epochs: 3 - > loading blendable dataset index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_index.npy - > loading blendable dataset sample index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_sample_index.npy - > size of blendable dataset: 10223415 samples - > finished creating GPT datasets ... - [after dataloaders are built] datetime: 2024-04-04 09:36:07 - done with setup ... - (min, max) time across ranks (ms): - model-and-optimizer-setup ......................: (4794.78, 4795.23) - train/valid/test-data-iterators-setup ..........: (589.69, 721.20) - training ... - [before the start of training step] datetime: 2024-04-04 09:36:07 - [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:539:forward] Activation Checkpointing Information - [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:540:forward] ----Partition Activations False, CPU CHECKPOINTING False - [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:541:forward] ----contiguous Memory Checkpointing False with 32 total layers - [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:543:forward] ----Synchronization False - [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:544:forward] ----Profiling time in checkpointing False - [2024-04-04 09:36:28,429] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1626.54 | optimizer_gradients: 19.29 | optimizer_step: 419.48 - [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 11336.34 | bwd_microstep: 7134.73 | bwd_inner_microstep: 7090.02 | bwd_allreduce_microstep: 44.65 | step_microstep: 2564.02 - [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 11336.33 | bwd: 7134.75 | bwd_inner: 7090.01 | bwd_allreduce: 44.66 | step: 2564.02 - iteration 1/ 317892 | consumed samples: 32 | consumed tokens: 131072 | elapsed time per iteration (ms): 21133.8 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.119983E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 1.514 | tokens per gpu per second(tgs): 775.250 | TFLOPs: 47.23 | - [Rank 1] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 35824.0 - [Rank 0] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 32994.0 - [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1605.55 | optimizer_gradients: 11.56 | optimizer_step: 50.92 - [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] - [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 - [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 - iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | - - # ... - ``` - -
- -
- -
- - - - - - - - - - +
[Overridable Options]: + + This is a simple subset of the overridable options. + + The full list (as well as their default values) can be found in [ALCF / `helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh) + + - `DTYPE`: Data type + - `DATA_FILE_LIST`: Data file list + - `FFN_HIDDEN_SIZE`: Feedforward Neural Network projection size + - `GRAD_ACC_STEPS`: Gradient accumulation steps + - `HEADS`: Number of attention heads + - `HIDDEN`: Hidden size + - `MICRO_BATCH`: Micro batch size + - `NO_FLASH_ATTN`: No Flash Attention + - `NLAYERS`: Number of layers + - `NUM_KV_HEAD`: Number of key-value heads + - `OPT`: Optimizer + - `adam` + - `adam8bit` + - `adamw` + - `adamwschedulefree` + - `apex.adam` + - `apex.sgd` + - `ds.fusedlamb` + - `ds.onebitlamb` + - `galoreadamw` + - `galoreadamw8bit` + - `galoreadamw8bitperlayer` + - `ipex.fusedlamb` + - `ipex.lamb` + - `shampoo` + - `sgd` + - `sgdschedulefree` + - `sophiag` + - `PP`: Pipeline parallelism degree + - `SEQ`: Sequence length + - `SP`: Sequence parallelism (Ulysses) degree + - `TP`: Tensor parallelism degree + - `TRAIN_TOKENS`: Number of training tokens + - `TRAIN_ITERS`: Number of training iterations + - `USE_ACTIVATION_CHECKPOINTING`: Use activation checkpointing + - `WEIGHT_DECAY`: Weight decay + - `ZERO_STAGE`: Zero stage + +
### 🚀 Submit as a batch job @@ -1079,7 +116,6 @@ $ qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=e ``` - ## 📝 Data Preprocessing
Data Pre-Processing: From 0dcc1017d9748fcb0671a39350240237ea596f28 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 12 Mar 2025 10:33:59 -0500 Subject: [PATCH 397/590] docs: Add `ALCF/notes/deprecated.md` --- ALCF/notes/deprecated.md | 1068 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 1068 insertions(+) create mode 100644 ALCF/notes/deprecated.md diff --git a/ALCF/notes/deprecated.md b/ALCF/notes/deprecated.md new file mode 100644 index 00000000000..878986c4ef7 --- /dev/null +++ b/ALCF/notes/deprecated.md @@ -0,0 +1,1068 @@ +# Deprecated + +This contains the (now deprecated) contents of the [`ALCF/README.md`](ALCF/README.md) file. + +To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): + +1.
⏳ Request an interactive job with qsub -I: + + ```bash + qsub -A -q debug -l select=2 -l walltime=01:00:00,filesystems=eagle:home -I + ``` + + - Or, alternatively, you can submit [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) + directly as a batch script with + + ```bash + cd Megatron-DeepSpeed + qsub -A -q debug -l select=2 -l walltime=01:00:00:filesystems=eagle:home train_aGPT_7B.sh + ``` + +
+ +2.
⬇️ Clone repo + navigate into it: + + ```bash + git clone "https://github.com/argonne-lcf/Megatron-DeepSpeed" + cd Megatron-DeepSpeed + ``` + +
+ +3.
🐍 Setup Python: + +
+ + > **NOTE**: The following commands should be ran from [`Megatron-DeepSpeed`](https://github.com/argonne-lcf/Megatron-DeepSpeed), following the `cd` command from 2. + + 1. Load `conda` module and activate base environment: + + ```bash + export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && ezpz_setup + ``` + + -
[output]: + +
+ + -
[Polaris]: + + ```bash + # [05:47:13 PM][foremans@x3001c0s13b1n0][/eagle/a/f/p/ar/Megatron-DeepSpeed-D/Megatron-DeepSpeed] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python + Using WORKING_DIR: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + Running on Polaris !! + + Lmod is automatically replacing "nvhpc/23.9" with "gcc-native/12.3". + + + Lmod is automatically replacing "PrgEnv-nvhpc/8.5.0" with "PrgEnv-gnu/8.5.0". + + + Due to MODULEPATH changes, the following have been reloaded: + 1) cray-mpich/8.1.28 + + Found conda at: /soft/applications/conda/2024-04-29/mconda3 + No VIRTUAL_ENV found in environment! + - Trying to setup from /soft/applications/conda/2024-04-29/mconda3 + - Using VENV_DIR=/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 + - Found existing venv, activating from /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29 + [python] Using: /eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed-DistributedDataLoading/Megatron-DeepSpeed/venvs/2024-04-29/bin/python3 + ``` + +
+ + -
[Aurora]: + + ```bash + # [10:04:02 PM][foremans@x4415c0s2b0n0][/gecko/A/fo/p/a/Megatron-DeepSpeed] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python + Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + + The following have been reloaded with a version change: + 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/803.29 2) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.1 + + Found conda at: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 + No VIRTUAL_ENV found in environment! + - Trying to setup from /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 + - Using VENV_DIR=/gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 + - Found existing venv, activating from /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 + [python] Using: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 + ``` + +
+ + -
[Sunspot]: + + ```bash + # [05:37:18 PM][foremans@x1921c0s0b0n0][/gila/A/fo/p/a/Megatron-DeepSpeed] + $ PBS_O_WORKDIR=$(pwd) source ALCF/helpers.sh && setup_python + Using WORKING_DIR: /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + No conda_prefix or virtual_env found in environment... + Setting up conda... + Running on SunSpot !! + + Due to MODULEPATH changes, the following have been reloaded: + 1) gcc/12.2.0 5) mpich-config/collective-tuning/1024 + 2) gmp/6.2.1-pcxzkau 6) mpich/icc-all-pmix-gpu/20231026 + 3) mpc/1.3.1-dfagrna 7) oneapi/eng-compiler/2024.04.15.002 + 4) mpfr/4.2.0-w7v7yjv + + The following have been reloaded with a version change: + 1) intel_compute_runtime/release/821.36 => intel_compute_runtime/release/775.20 + 2) spack-pe-gcc/0.7.0-24.086.0 => spack-pe-gcc/0.6.1-23.275.2 + UMD: agama-ci-devel-803.29 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-803.29 + + The following have been reloaded with a version change: + 1) oneapi/eng-compiler/2024.04.15.002 => oneapi/release/2024.04.15.001 + + Found conda at: /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + No VIRTUAL_ENV found in environment! + - Trying to setup from /soft/datascience/aurora_nre_models_frameworks-2024.1_preview_u1 + - Using VENV_DIR=/gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + - Found existing venv, activating from /gila/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1 + [python] Using: /lus/gila/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1_preview_u1/bin/python3 + ``` + +
+ + + + 2. 🍋 Install [`ezpz`](https://github.com/saforem2/ezpz): + + ```bash + mkdir deps && git clone https://github.com/saforem2/ezpz deps/ezpz + python3 -m pip install -e deps/ezpz --require-virtualenv + ``` + + [^venv]: Its generally a good practice to keep separate virtual Python environments different projects. + We provide a helper function, [`setup_venv_from_conda()`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/2f0154394bbdf3c64b4669f9d944645e2cdb8f2b/ALCF/helpers.sh#L440), + that helps take care of this for you. +
+ This will: activate (or build, if necessary) a `venv` in your working dir, + _automatically_ matching the name of your active `conda` environment (e.g. `2024-04-29`, on Polaris_. + + 3. Setup [`wandb`](https://docs.wandb.ai/quickstart) + + > **NOTE**: this can be disabled by setting `export WANDB_DISABLED=1` + +
+ + + +4.
🚀 Launch: + + In this case, train a ~ 2B Model (with 10 layers), + for 1000 iterations using the data file list in: + + [`ALCF/data-lists/polaris/books.txt`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/data-lists/polaris/books.txt) + + with a micro-batch-size of 2 (`MICRO_BATCH=2`), with the `torch.optim.AdamW` optimizer (`OPT=adamw`). + + **Note** that _any_ of the options in the [`setParams`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh#L140) + function from [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/7d203596dbf14e048e756c5ee6705de7dcb22283/ALCF/helpers.sh) + can be overridden dynamically at runtime using this technique. + + ```bash + # for systems other than Polaris, replace "polaris/books.txt" below with: + # "{aurora,sunspot}/books.txt", + PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt TRAIN_ITER=1000 NLAYERS=10 MICRO_BATCH=2 OPT=adamw bash train_aGPT_7B.sh + ``` + + - **Note**: If no additional options specified, i.e. + + ```bash + PBS_O_WORKDIR=$(pwd) bash train_aGPT_7B.sh + ``` + + then this will fallback to using the default AuroraGPT-7B architecture + with the full Dolma (v1.7) dataset. + +
[output]: + +
+ + The outputs should look _something_ like this, though YMMV (things change quick): + +
[Aurora]: + + ```bash + #[🌌][10:45:59 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?] + $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup_python + + #[🌌][10:46:57 AM][foremans@x4711c1s2b0n0][…/Megatron-DeepSpeed][🌱 main][$!?][aurora_nre_models_frameworks-2024.1] + (aurora_nre_models_frameworks-2024.1) $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt bash train_aGPT_7B.sh > train-log-$(tstamp).log 2>&1 & + + Using WORKING_DIR: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + Running on: aurora + Using virtual_env: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1 on top of conda from: /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1 + [python] Using: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 + Ensuring all dependencies from /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/requirements/requirements.txt installed... + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + ┌─────────────────────────────────────────────────────────────────────┐ + │ [savejobenv]: + │ • Writing PBS vars to: /home/foremans/.pbsenv + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [HOSTS]: + │ • [host:0] - x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + │ • [host:1] - x4711c1s3b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [DIST INFO]: + │ • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + │ • NHOSTS=2 + │ • NGPU_PER_HOST=12 + │ • NGPUS=24 + └─────────────────────────────────────────────────────────────────────┘ + ┌─────────────────────────────────────────────────────────────────────┐ + │ [LAUNCH]: + │ • To launch across all available GPUs, use: + │ 'launch' ( = mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov ) + └─────────────────────────────────────────────────────────────────────┘ + 2024-06-21 10:47:09,771 - numexpr.utils - INFO - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. + 2024-06-21 10:47:09,772 - numexpr.utils - INFO - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. + 2024-06-21 10:47:09,772 - numexpr.utils - INFO - NumExpr defaulting to 8 threads. + /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages/pandas/core/computation/expressions.py:21: UserWarning: Pandas requires version '2.8.4' or n> + from pandas.core.computation.check import NUMEXPR_INSTALLED + /opt/aurora/24.086.0/frameworks/aurora_nre_models_frameworks-2024.1/lib/python3.9/runpy.py:127: RuntimeWarning: 'ezpz.jobs' found in sys.modules after import of package 'ezpz', but prior to execution of 'ezpz.jobs'; this may result in u> + warn(RuntimeWarning(msg)) + [2024-06-21 10:47:10][INFO][jobs:366] - Caught PBS_JOBID='684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov', pbsnf=PosixPath('/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov') from env. Saving jobenv! + [2024-06-21 10:47:10][WARNING][jobs:117] - /home/foremans/PBS-jobs/684084 already in /home/foremans/PBS-jobs.log, not appending !! + [2024-06-21 10:47:10][INFO][jobs:192] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.sh + [2024-06-21 10:47:10][INFO][jobs:220] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.json + [2024-06-21 10:47:10][INFO][jobs:233] - Saving job env to /home/foremans/PBS-jobs/684084/jobenv.yaml + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /home/foremans/PBS-jobs/684084/.jobenv + [2024-06-21 10:47:10][INFO][jobs:137] - Saving job env to .jobenv file in /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][WARNING][jobs:154] - To use launch alias, be sure to: source /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/.jobenv + [2024-06-21 10:47:10][INFO][jobs:277] - Writing PBS env vars to /home/foremans/PBS-jobs/684084 / jobenv{.sh, .yaml, .json} + [2024-06-21 10:47:10][WARNING][jobs:281] - Run: source ./.jobenv in your current shell to set job variables + [2024-06-21 10:47:10][INFO][jobs:374] - + [DIST_INFO]: + • DEVICE=xpu + • DEVICE_ID=xpu:0 + • DISTRIBUTED_BACKEND=ccl + • GPUS_PER_NODE=12 + • HOSTS=['x4711c1s2b0n0', 'x4711c1s3b0n0'] + • HOSTFILE=/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + • HOSTNAME=x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + • LOCAL_RANK=0 + • MACHINE=Aurora + • NUM_NODES=2 + • NGPUS=24 + • NODE_ID=0 + • RANK=0 + • SCHEDULER=PBS + • WORLD_SIZE_TOTAL=24 + • WORLD_SIZE_IN_USE=1 + [2024-06-21 10:47:10][CRITICAL][jobs:245] - To launch across ALL GPUs in your job, use: + LAUNCH_CMD=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + creating alias launch=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + Found ezpz! + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + Done with ezpz. + Not using flash-attn!! + LR_ARGS: --lr 0.0003 --lr-decay-style cosine --lr-warmup-fraction 0.05 + DS_CONFIG: /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ds-configs/ds_stage1_mb4_gb768_pp1_bf16.json + ZS: 1, MB: 4, GB: 768, PP: 1, DTYPE: bf16 + Please see logs at: logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov + Checkpoints will be saved to: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Setting up tokenizer with Llama2 + Using data_file_list: ./ALCF/data-lists/aurora/books.txt + Using tokenizer: Llama2. Setting up data with ./ALCF/data-lists/aurora/books.txt + Calling: setData() with ./ALCF/data-lists/aurora/books.txt + -------------------- + Updated environment: + DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt + NUM_DOCS: 3 + WEIGHT_SUM: 0.0072042092147565125 + DFL_STEM: books + DATA_CACHE_PATH: .cache/books/index-cache + DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt + -------------------- + [setData] DATA_FLAGS: --data-file-list ./ALCF/data-lists/aurora/books.txt + [setData] TOKENIZER_FLAGS: --tokenizer-type Llama2Tokenizer --tokenizer-model /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokenizer.model + Requirement already satisfied: pybind11 in ./venvs/aurora_nre_models_frameworks-2024.1/lib/python3.9/site-packages (2.12.0) + + [notice] A new release of pip is available: 24.0 -> 24.1 + [notice] To update, run: pip install --upgrade pip + make: Nothing to be done for 'default'. + /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR=/opt/aurora/24.086.0/CNDA/mpich/20231026/mpich-ofi-all-icc-default-pmix-gpu-drop20231026 + - Using /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 + - WORLD_SIZE:24 + - BACKEND: ccl + - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./ALCF/data-lists/aurora/books.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + + Currently Loaded Modules: + 1) mpich/icc-all-pmix-gpu/20231026 3) libfabric/1.15.2.0 5) cray-libpals/1.3.3 7) gmp/6.2.1-pcxzkau 9) mpc/1.3.1-dfagrna 11) intel_compute_runtime/release/803.29 13) frameworks/2024.1 + 2) mpich-config/collective-tuning/1024 4) cray-pals/1.3.3 6) spack-pe-gcc/0.7.0-24.086.0 8) mpfr/4.2.0-w7v7yjv 10) gcc/12.2.0 12) oneapi/release/2024.1 + + + + Saving environment to checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.env + Not currently running. Continuing! + Launching with: MPICH + mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py + Using data_cache_path: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache + + mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov --genvall --cpu-bind depth -d 16 /gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/venvs/aurora_nre_models_frameworks-2024.1/bin/python3 -Wignore /lus/gecko/projects/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --no-bias-dropout-fusion --no-masked-softmax-fusion --no-gradient-accumulation-fusion > + + [!! NOTE] View output at: + logs/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/20240621-104713_24_x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov/output.log + Connected to tcp://x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov:7919 + Launching application eafe3e80-ad2e-4cee-a3e4-d63af2a77c66 + [2024-06-21 10:47:31,610] [INFO] [comm.py:161:init_deepspeed_backend] Initialize ccl backend + [2024-06-21 10:47:31,610] [INFO] [comm.py:637:init_distributed] cdb=None + [2024-06-21 10:47:31,610] [INFO] [comm.py:652:init_distributed] Not using the DeepSpeed or dist launchers, attempting to detect MPI environment... + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=15, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=23, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=12, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=13, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=3, local_rank=3, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=14, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=16, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=17, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=18, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=19, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=20, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=21, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=22, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=6, local_rank=6, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=8, local_rank=8, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=10, local_rank=10, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=0, local_rank=0, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:668:init_distributed] Initializing TorchBackend in DeepSpeed with backend ccl + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=1, local_rank=1, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=2, local_rank=2, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=4, local_rank=4, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=5, local_rank=5, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=7, local_rank=7, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=9, local_rank=9, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:31,611] [INFO] [comm.py:702:mpi_discovery] Discovered MPI settings of world_rank=11, local_rank=11, world_size=24, master_addr=10.115.79.12, master_port=29500 + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=2/23][local_rank=2/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=6/23][local_rank=6/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=8/23][local_rank=8/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=1/23][local_rank=1/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=15/23][local_rank=3/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=17/23][local_rank=5/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=3/23][local_rank=3/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=13/23][local_rank=1/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=14/23][local_rank=2/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=4/23][local_rank=4/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=18/23][local_rank=6/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=20/23][local_rank=8/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=5/23][local_rank=5/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=21/23][local_rank=9/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=7/23][local_rank=7/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=23/23][local_rank=11/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=9/23][local_rank=9/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=12/23][local_rank=0/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=16/23][local_rank=4/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=10/23][local_rank=10/11][node=0/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=11/23][local_rank=11/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=19/23][local_rank=7/11][node=1/1] + [2024-06-21 10:47:32][INFO][dist:291] - [device='xpu'][rank=22/23][local_rank=10/11][node=0/1] + 2024-06-21 10:47:32][INFO][dist:240] - DistInfo={ + "DEVICE": "xpu", + "DEVICE_ID": "xpu:0", + "DISTRIBUTED_BACKEND": "ccl", + "GPUS_PER_NODE": 12, + "HOSTFILE": "/var/spool/pbs/aux/684084.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov", + "HOSTNAME": "x4711c1s2b0n0.hostmgmt2711.cm.aurora.alcf.anl.gov", + "HOSTS": "['x4711c1s2b0n0', 'x4711c1s3b0n0']", + "LOCAL_RANK": 0, + "MACHINE": "Aurora", + "NGPUS": 24, + "NODE_ID": 0, + "NUM_NODES": 2, + "RANK": 0, + "SCHEDULER": "PBS", + "WORLD_SIZE_IN_USE": 24, + "WORLD_SIZE_TOTAL": 24 + } + + # [...clipped...] + + [2024-06-21 10:48:48][INFO][utils:307] - > elapsed time for building blendable dataset indices: 1.19 (sec) + [2024-06-21 10:48:48][INFO][utils:307] - > saving index map files + [2024-06-21 10:48:51][INFO][utils:307] - > finished saving index map files in 3.0829622745513916 seconds + [2024-06-21 10:48:51][INFO][utils:307] - > loading blendable dataset index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_index.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading blendable dataset sample index: checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/49e9529a32d0a98f1e40f4a82872b11c_sample_index.npy + [2024-06-21 10:48:52][INFO][utils:307] - > finished loading in 0.30188989639282227 seconds + [2024-06-21 10:48:52][INFO][utils:307] - >> building dataset for /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document + [2024-06-21 10:48:52][INFO][utils:307] - > building dataset index ... + [2024-06-21 10:48:52][INFO][utils:307] - reading sizes... + [2024-06-21 10:48:52][INFO][utils:307] - reading pointers... + [2024-06-21 10:48:52][INFO][utils:307] - reading document index... + [2024-06-21 10:48:52][INFO][utils:307] - creating numpy buffer of mmap... + [2024-06-21 10:48:52][INFO][utils:307] - /gecko/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document.bin + [2024-06-21 10:48:52][INFO][utils:307] - creating memory view of numpy buffer... + [2024-06-21 10:48:52][INFO][utils:307] - > finished creating indexed dataset in 0.003112 seconds + [2024-06-21 10:48:52][INFO][utils:307] - number of documents: 7386 + [2024-06-21 10:48:52][INFO][utils:307] - > dataset split: + [2024-06-21 10:48:52][INFO][utils:307] - train: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [0, 7386) total of 7386 documents + [2024-06-21 10:48:52][INFO][utils:307] - validation: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents + [2024-06-21 10:48:52][INFO][utils:307] - test: + [2024-06-21 10:48:52][INFO][utils:307] - document indices in [7386, 7386) total of 0 documents + [2024-06-21 10:48:52][INFO][utils:307] - > loading doc-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_doc_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading sample-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_sample_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - > loading shuffle-idx mapping from checkpoints/ws24_ds_stage1_nl32_hs4096_mb4_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr0.0003_lwf0.05/.cache/books/index-cache/1fa7757ef8907da21e1e1326705e7f3f_shuffle_idx.npy + [2024-06-21 10:48:52][INFO][utils:307] - loaded indexed file in 0.008 seconds + [2024-06-21 10:48:52][INFO][utils:307] - total number of samples: 34196233 + [2024-06-21 10:48:52][INFO][utils:307] - total number of epochs: 175 + [2024-06-21 10:48:52][INFO][utils:307] - > size of blendable dataset: 245361763 samples + [2024-06-21 10:48:52][INFO][utils:307] - >>> Finished building BlendableDataset in 4.613574266433716 seconds + [2024-06-21 10:48:52][INFO][pretrain_gpt_alcf:579] - > finished creating GPT datasets. Took: 45730179865763.24219s + [2024-06-21 10:48:53][INFO][training:88] - [after dataloaders are built] datetime=2024-06-21 10:48:53 + [2024-06-21 10:48:53][INFO][training:307] - done with setup ... + [2024-06-21 10:48:53][INFO][training:313] - training ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (63763.34, 63857.25) + train/valid/test-data-iterators-setup ..........: (12936.53, 13432.64) + [2024-06-21 10:48:53][INFO][training:88] - [before the start of training step] datetime=2024-06-21 10:48:53 + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:541:forward] Activation Checkpointing Information + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:542:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:543:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:545:forward] ----Synchronization False + [2024-06-21 10:48:53,396] [INFO] [checkpointing.py:546:forward] ----Profiling time in checkpointing False + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1867.64 | optimizer_gradients: 19.65 | optimizer_step: 46.07 + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[1.887433467970254e-08, 1.887433467970254e-08], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-06-21 10:50:42,167] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 25341.72 | bwd_microstep: 77707.38 | bwd_inner_microstep: 75751.84 | bwd_allreduce_microstep: 1955.54 | step_microstep: 2218.38 + [2024-06-21 10:50:42,168] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 25341.72 | bwd: 77707.38 | bwd_inner: 75751.84 | bwd_allreduce: 1955.54 | step: 2218.38 + [2024-06-21 10:50:42][INFO][training:1609] - iteration= 1/ 317892 | consumed_samples= 768 | consumed_tokens= 3145728 | elapsed_time_per_iteration_ms=108893.2 | learning_rate=1.88743e-08 | global_batch_size= 768 | lm loss=11.133188 | loss_scale=1.0 | actual_seqlen= 4096 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=7.053 | tokens_per_gpu_per_second_tgs=1203.674 | [LM]-TFLOPs=49.66 | [DS]-TFLOPs=73.32 | + [2024-06-21 10:50:42][INFO][utils:190] - [Rank 0] (after 1 iterations) memory (MB) | allocated: 18243.64111328125 | max allocated: 50664.2548828125 | reserved: 54556.0 | max reserved: 54556.0 + (min, max) time across ranks (ms): + forward-backward ...............................: (106622.81, 106624.28) + optimizer ......................................: (2221.02, 2234.98) + ``` + +
+ +
[Sunspot]: + + ```bash + # [09:07:32 AM][foremans@x1921c0s0b0n0][~/q/llm.devkit/Megatron-DeepSpeed][🌱 main][$!?] + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt bash train_aGPT_7B.sh + source-ing /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/helpers.sh + Sourcing /home/foremans/q4-drop_sunspot/llm.devkit/setenv.sh... + UMD: agama-ci-devel-736.9 successfully loaded: + UMD: graphics-compute-runtime/agama-ci-devel-736.9 + Lmod has detected the following error: The following module(s) are unknown: "gcc/12.1.0" + + Please check the spelling or version number. Also try "module spider ..." + It is also possible your cache file is out-of-date; it may help to try: + $ module --ignore_cache load "gcc/12.1.0" + + Also make sure that all modulefiles written in TCL start with the string #%Module + + Note: the module "intel_compute_runtime/release/agama-devel-647" cannot be unloaded because it was not loaded. + + Running on SunSpot !! + [python] Using: /home/foremans/miniconda3/envs/q4-drop/bin/python3 + Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env + Found ezpz! + /lus/gila/projects/Aurora_deployment/foremans/locations/sunspot/projects/saforem2/ezpz/src/ezpz/__init__.py + Has ezpz installed. Nothing to do. + Done with ezpz. + ┌─────────────────────────────────────────────────────────────────── + │ Writing PBS vars to /home/foremans/.pbsenv + │ HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 + │ NHOSTS: 2 + │ NGPU_PER_HOST: 12 GPUs per host + │ NGPUS: 24 GPUs total + └─────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Hosts]: + │ • [host:0] - x1921c0s0b0n0.hostmgmt2000.cm.americas.sgi.com + │ • [host:1] - x1921c0s1b0n0.hostmgmt2000.cm.americas.sgi.com + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [DIST INFO]: + │ • Loading job env from: /home/foremans/.pbsenv + │ • HOSTFILE: /var/spool/pbs/aux/8988430.amn-0001 + │ • NHOSTS: 2 + │ • NGPU_PER_HOST: 12 + │ • NGPUS (NHOSTS x NGPU_PER_HOST): 24 + │ • WORLD_SIZE: 24 + │ • DIST_LAUNCH: mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001 + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Launch]: + │ • Use: 'launch' (=mpiexec --verbose --envall -n 24 -ppn 12 --hostfile /var/spool/pbs/aux/8988430.amn-0001) + │ to launch job + └────────────────────────────────────────────────────────────────── + DS_CONFIG: ds_stage2_mb4_gb96_pp1_bf16.json + ZS: 2, CPU_OPTIMIZER: , MB: 4, GB: 96, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Calling: setData() with ./convergence_debug_small.txt + -------------------- + Updated environment: + DATA_FILE_LIST: ./convergence_debug_small.txt + NUM_DOCS: 15 + WEIGHT_SUM: 15.0 + DFL_STEM: convergence_debug_small + DATA_CACHE_PATH: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache + -------------------- + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR= + - Using /home/foremans/miniconda3/envs/q4-drop/bin/python3 + - WORLD_SIZE:24 + - NCCL: nccl + - MODEL_TYPE: llama-seq4096-pp1-tp1-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./convergence_debug_small.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + ! Using /home/foremans/miniconda3/envs/q4-drop/bin/deepspeed + /home/foremans/miniconda3/envs/q4-drop/bin/ds_report:4: DeprecationWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html + __import__('pkg_resources').require('deepspeed==0.12.3+6ea44d02') + /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: ''If you dont plan on using image function + ality from `torchvision.io`, you can ignore this warning. Otherwise, there might be something wrong with your environment. Did you have `libjpeg` or `libpng` installed before building `torch + vision` from source? + warn( + [2024-04-04 09:07:45,585] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) + [2024-04-04 09:07:45,818] [INFO] [real_accelerator.py:158:get_accelerator] Setting ds_accelerator to xpu (auto detect) + -------------------------------------------------- + DeepSpeed C++/CUDA extension op report + -------------------------------------------------- + NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. + -------------------------------------------------- + JIT compiled ops requires ninja + ninja .................. [OKAY] + -------------------------------------------------- + op name ................ installed .. compatible + -------------------------------------------------- + async_io ............... [NO] ....... [OKAY] + cpu_adagrad ............ [NO] ....... [OKAY] + cpu_adam ............... [NO] ....... [OKAY] + flash_attn ............. [NO] ....... [OKAY] + fused_adam ............. [NO] ....... [OKAY] + quantizer .............. [NO] ....... [OKAY] + transformer ............ [NO] ....... [OKAY] + transformer_inference .. [NO] ....... [OKAY] + utils .................. [NO] ....... [OKAY] + -------------------------------------------------- + DeepSpeed general environment info: + torch install path ............... ['/home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch'] + torch version .................... 2.1.0a0+cxx11.abi + deepspeed install path ........... ['/lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/DeepSpeed/deepspeed'] + deepspeed info ................... 0.12.3+6ea44d02, 6ea44d02, HEAD + deepspeed wheel compiled w. ...... torch 2.1 + shared memory (/dev/shm) size .... 503.18 GB + + deepspeed --hostfile /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/gila/projects/Aurora_deployment/ + foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion --lr-decay + -style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce-grads-in-fp32 + --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/tensorboard --log-timers-to-tensorboard --log-optimizer + -states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16 + --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend ccl --num-attention-heads 32 --save-interval 20 + 0 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 4 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 1 --global-bat + ch-size 96 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ + .cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/ALCF/tokenizer.model --no-query- + key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --deepspeed-activation-checkpointing --z + ero-stage=2 --deepspeed_config=ds_stage2_mb4_gb96_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint-num-layers 1 |& tee logs/ds_stage2 + _nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + + [!! NOTE] View output at: + logs/ds_stage2_nl32_hs4096_mb4_seq4096_gb96_pp1_tp1_bf16/0404090742_x1921c0s0b0n0/output.log + + # ... + + /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.010017 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_doc_idx.npy + > loading sample-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_sample_idx.npy + > loading shuffle-idx mapping from /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/bf90c74a625ac2ee4de6e1d6f7f84fbb_shuffle_idx.npy + loaded indexed file in 0.056 seconds + total number of samples: 2318461 + total number of epochs: 8 + > loading blendable dataset index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_index.npy + > loading blendable dataset sample index: /lus/gila/projects/Aurora_deployment/foremans/q4-drop_sunspot/llm.devkit/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/3a426af74008c22f9db24db811aad6b7_sample_index.npy + /home/foremans/miniconda3/envs/q4-drop/lib/python3.9/site-packages/torch/utils/data/dataloader.py:557: UserWarning: This DataLoader will create 2 worker processes in total. Our suggested max number of worker in current system is 1, which is smaller than what this DataLoader is going to create. Please be aware that excessive worker creation might get DataLoader running slow or even freeze, lower the worker number to avoid potential slowness/freeze if necessary. + + [after dataloaders are built] datetime: 2024-04-04 09:09:27 + done with setup ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (64818.18, 64858.22) + train/valid/test-data-iterators-setup ..........: (1968.10, 2288.56) + training ... + [before the start of training step] datetime: 2024-04-04 09:09:27 + [2024-04-04 09:09:27,718] [INFO] [checkpointing.py:540:forward] Activation Checkpointing Information + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:541:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:542:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:544:forward] ----Synchronization False + [2024-04-04 09:09:27,719] [INFO] [checkpointing.py:545:forward] ----Profiling time in checkpointing False + [2024-04-04 09:09:33][INFO][utils:145] - Note: detected 208 virtual cores but NumExpr set to maximum of 64, check "NUMEXPR_MAX_THREADS" environment variable. + [2024-04-04 09:09:33][INFO][utils:148] - Note: NumExpr detected 208 cores but "NUMEXPR_MAX_THREADS" not set, so enforcing safe limit of 8. + [2024-04-04 09:09:33][INFO][utils:160] - NumExpr defaulting to 8 threads. + ^[c[2024-04-04 09:09:53,311] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 884.11 | optimizer_gradients: 6.43 | optimizer_step: 23.44 + [2024-04-04 09:09:53,312] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 6567.68 | bwd_microstep: 17950.36 | bwd_inner_microstep: 17711.20 | bwd_allreduce_microstep: 239.11 | step_microstep: 1139.27 + [2024-04-04 09:09:53,313] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 6567.66 | bwd: 17950.35 | bwd_inner: 17711.19 | bwd_allreduce: 239.11 | step: 1139.29 + [Rank 0] (after 1 iterations) memory (MB) | allocated: 18244.640625 | max allocated: 41299.50146484375 | reserved: 46764.0 | max reserved: 46764.0 + iteration 1/ 317892 | consumed samples: 96 | consumed tokens: 393216 | elapsed time per iteration (ms): 25849.1 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 1.117136E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.714 | tokens per gpu per second(tgs): 633.832 | TFLOPs: 38.61 | + [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 327.85 | optimizer_gradients: 6.26 | optimizer_step: 23.60 + [2024-04-04 09:10:13,619] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 4022.74 | bwd_microstep: 15738.67 | bwd_inner_microstep: 15556.80 | bwd_allreduce_microstep: 181.82 | step_microstep: 371.01 + [2024-04-04 09:10:13,620] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 4022.73 | bwd: 15738.66 | bwd_inner: 15556.62 | bwd_allreduce: 181.81 | step: 371.02 + iteration 2/ 317892 | consumed samples: 192 | consumed tokens: 786432 | elapsed time per iteration (ms): 20298.3 | learning rate: 3.000E-04 | global batch size: 96 | lm loss: 2.537718E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 4.729 | tokens per gpu per second(tgs): 807.159 | TFLOPs: 49.17 | + ``` + +
+ +
[Polaris]: + + ```bash + # [09:31:35 AM][foremans@x3112c0s13b0n0][~/pol/p/a/Megatron-DeepSpeed][🌱 main][$!?] + $ PBS_O_WORKDIR=$(pwd) DATA_FILE_LIST=./ALCF/data-lists/polaris/books.txt OPT=adamw bash train_aGPT_7B.sh + source-ing /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/helpers.sh + Running on Polaris !! + + [python] Using: /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 + Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env + Found ezpz! + /lus/eagle/projects/datascience/foremans/tmp/Megatron-DeepSpeed/ezpz/src/ezpz/__init__.py + Has ezpz installed. Nothing to do. + Done with ezpz. + ┌─────────────────────────────────────────────────────────────────── + │ Writing PBS vars to /home/foremans/.pbsenv + │ HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + │ NHOSTS: 2 + │ NGPU_PER_HOST: 4 GPUs per host + │ NGPUS: 8 GPUs total + └─────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Hosts]: + │ • [host:0] - x3112c0s13b0n0.hsn.cm.polaris.alcf.anl.gov + │ • [host:1] - x3112c0s13b1n0.hsn.cm.polaris.alcf.anl.gov + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [DIST INFO]: + │ • Loading job env from: /home/foremans/.pbsenv + │ • HOSTFILE: /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + │ • NHOSTS: 2 + │ • NGPU_PER_HOST: 4 + │ • NGPUS (NHOSTS x NGPU_PER_HOST): 8 + │ • WORLD_SIZE: 8 + │ • DIST_LAUNCH: mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov + └────────────────────────────────────────────────────────────────── + ┌────────────────────────────────────────────────────────────────── + │ [Launch]: + │ • Use: 'launch' (=mpiexec --verbose --envall -n 8 -ppn 4 --hostfile /var/spool/pbs/aux/1822297.polaris-pbs-01.hsn.cm.polaris.alcf.anl.gov) + │ to launch job + └────────────────────────────────────────────────────────────────── + DS_CONFIG: ds_stage2_mb8_gb32_pp1_bf16.json + ZS: 2, CPU_OPTIMIZER: , MB: 8, GB: 32, PP: 1, DTYPE: bf16!!!Please see logs at logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0 + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + !! Caught USE_ACTIVATION_CHECKPOINTING=1 !! + Calling: setData() with "./convergence_debug_small.txt" + -------------------- + Updated environment: + DATA_FILE_LIST: ./convergence_debug_small.txt + NUM_DOCS: 15 + WEIGHT_SUM: 15.0 + DFL_STEM: convergence_debug_small + DATA_CACHE_PATH: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache + -------------------- + ++++++++++++++++++++++++++++++++++++++++++++++++++ + - MPICH_DIR=/opt/cray/pe/mpich/8.1.25/ofi/gnu/9.1 + - Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/python3 + - WORLD_SIZE:8 + - NCCL: nccl + - MODEL_TYPE: llama-seq4096-pp1-tp2-32layers-32heads-4096hidden + - Using DATA_FILE_LIST: ./convergence_debug_small.txt + ++++++++++++++++++++++++++++++++++++++++++++++++++ + ! Using /eagle/datascience/foremans/miniconda3/envs/cu118-pt221/bin/deepspeed + [2024-04-04 09:35:35,959] [INFO] [real_accelerator.py:191:get_accelerator] Setting ds_accelerator to cuda [auto detect] + -------------------------------------------------- + DeepSpeed C++/CUDA extension op report + -------------------------------------------------- + NOTE: Ops not installed will be just-in-time (JIT) compiled at + runtime if needed. Op compatibility means that your system + meet the required dependencies to JIT install the op. + -------------------------------------------------- + JIT compiled ops requires ninja + ninja .................. [OKAY] + -------------------------------------------------- + op name ................ installed .. compatible + -------------------------------------------------- + async_io ............... [NO] ....... [OKAY] + fused_adam ............. [NO] ....... [OKAY] + cpu_adam ............... [NO] ....... [OKAY] + cpu_adagrad ............ [NO] ....... [OKAY] + cpu_lion ............... [NO] ....... [OKAY] + [WARNING] Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH + evoformer_attn ......... [NO] ....... [NO] + fused_lamb ............. [NO] ....... [OKAY] + fused_lion ............. [NO] ....... [OKAY] + inference_core_ops ..... [NO] ....... [OKAY] + cutlass_ops ............ [NO] ....... [OKAY] + transformer_inference .. [NO] ....... [OKAY] + quantizer .............. [NO] ....... [OKAY] + ragged_device_ops ...... [NO] ....... [OKAY] + ragged_ops ............. [NO] ....... [OKAY] + random_ltd ............. [NO] ....... [OKAY] + [WARNING] sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.2 + [WARNING] using untested triton version (2.2.0), only 1.0.0 is known to be compatible + sparse_attn ............ [NO] ....... [NO] + spatial_inference ...... [NO] ....... [OKAY] + transformer ............ [NO] ....... [OKAY] + stochastic_transformer . [NO] ....... [OKAY] + -------------------------------------------------- + DeepSpeed general environment info: + torch install path ............... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/torch'] + torch version .................... 2.2.1 + deepspeed install path ........... ['/eagle/datascience/foremans/miniconda3/envs/cu118-pt221/lib/python3.12/site-packages/deepspeed'] + deepspeed info ................... 0.14.0, unknown, unknown + torch cuda version ............... 11.8 + torch hip version ................ None + nvcc version ..................... 11.8 + deepspeed wheel compiled w. ...... torch 2.2, cuda 11.8 + shared memory (/dev/shm) size .... 251.61 GB + + deepspeed --hostfile /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/hostfile_deepspeed --launcher MPICH /lus/eagle/projects/datascienc + e/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/pretrain_gpt_alcf.py --bf16 --optimizer adamw --split 100,0,0 --log-interval 1 --no-bias-gelu-fusion + --lr-decay-style cosine --no-bias-dropout-fusion --no-masked-softmax-fusion --tokenizer-type Llama2Tokenizer --no-gradient-accumulation-fusion --accumulate-allreduce- + grads-in-fp32 --use-checkpoint-opt_param-scheduler --tensorboard-dir checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/tensorboard --log-timers-to-tensorboard - + -log-optimizer-states-to-tensorboard --lr 0.0003 --save checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16 --load checkpoints/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_ + pp1_tp2_bf16 --seq-length 4096 --num-layers 32 --hidden-size 4096 --train-iters 317892 --eval-iters 10 --distributed-backend nccl --num-attention-heads 32 --s + ave-interval 200 --eval-interval 50000 --max-position-embeddings 4096 --micro-batch-size 8 --data-file-list ./convergence_debug_small.txt --tensor-model-parallel-size 2 + --global-batch-size 32 --pipeline-model-parallel-size 1 --num-key-value-heads 8 --data-cache-path /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-l + cf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache --ffn-hidden-size 11008 --tokenizer-model /home/foremans/polaris/projects/argonne-lcf/Megatron-DeepSpeed/ALCF/tokeniz + er.model --no-query-key-layer-scaling --use-rotary-position-embeddings --untie-embeddings-and-output-weights --swiglu --normalization rmsnorm --disable-bias-linear --use-flash-attn-v2 + --deepspeed-activation-checkpointing --zero-stage=2 --deepspeed_config=ds_stage2_mb8_gb32_pp1_bf16.json --no-pipeline-parallel --deepspeed --checkpoint-activations --checkpoint- + num-layers 1 |& tee logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + + [!! NOTE] View output at: + logs/ds_stage2_nl32_hs4096_mb8_seq4096_gb32_pp1_tp2_bf16/0404093534_x3112c0s13b0n0/output.log + + # ... + + /eagle/datasets/dolma/data_Llama2Tokenizer/common-crawl/cc_en_middle/cc_en_middle-0051_text_document.bin + creating memory view of numpy buffer... + > finished creating indexed dataset in 0.001280 seconds + number of documents: 1498927 + > dataset split: + train: + document indices in [0, 1498927) total of 1498927 documents + validation: + document indices in [1498927, 1498927) total of 0 documents + test: + document indices in [1498927, 1498927) total of 0 documents + > loading doc-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_doc_idx.npy + > loading sample-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_sample_idx.npy + > loading shuffle-idx mapping from /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/9217d94f3290abc2fddf9e87bff236d6_shuffle_idx.npy + loaded indexed file in 0.004 seconds + total number of samples: 869423 + total number of epochs: 3 + > loading blendable dataset index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_index.npy + > loading blendable dataset sample index: /lus/eagle/projects/datascience/foremans/locations/polaris/projects/argonne-lcf/Megatron-DeepSpeed/.cache/convergence_debug_small/index-cache/a815d51f6752c6f486d94194ce95fb87_sample_index.npy + > size of blendable dataset: 10223415 samples + > finished creating GPT datasets ... + [after dataloaders are built] datetime: 2024-04-04 09:36:07 + done with setup ... + (min, max) time across ranks (ms): + model-and-optimizer-setup ......................: (4794.78, 4795.23) + train/valid/test-data-iterators-setup ..........: (589.69, 721.20) + training ... + [before the start of training step] datetime: 2024-04-04 09:36:07 + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:539:forward] Activation Checkpointing Information + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:540:forward] ----Partition Activations False, CPU CHECKPOINTING False + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:541:forward] ----contiguous Memory Checkpointing False with 32 total layers + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:543:forward] ----Synchronization False + [2024-04-04 09:36:07,407] [INFO] [checkpointing.py:544:forward] ----Profiling time in checkpointing False + [2024-04-04 09:36:28,429] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1626.54 | optimizer_gradients: 19.29 | optimizer_step: 419.48 + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] step=1, skipped=0, lr=[0.00029999999999267505, 0.00029999999999267505], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 11336.34 | bwd_microstep: 7134.73 | bwd_inner_microstep: 7090.02 | bwd_allreduce_microstep: 44.65 | step_microstep: 2564.02 + [2024-04-04 09:36:28,430] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 11336.33 | bwd: 7134.75 | bwd_inner: 7090.01 | bwd_allreduce: 44.66 | step: 2564.02 + iteration 1/ 317892 | consumed samples: 32 | consumed tokens: 131072 | elapsed time per iteration (ms): 21133.8 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.119983E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 1.514 | tokens per gpu per second(tgs): 775.250 | TFLOPs: 47.23 | + [Rank 1] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 35824.0 + [Rank 0] (after 1 iterations) memory (MB) | allocated: 14165.525390625 | max allocated: 22332.37255859375 | reserved: 24642.0 | max reserved: 32994.0 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | optimizer_allgather: 1605.55 | optimizer_gradients: 11.56 | optimizer_step: 50.92 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] step=2, skipped=0, lr=[0.00029999999997070033, 0.00029999999997070033], mom=[(0.9, 0.999), (0.9, 0.999)] + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd_microstep: 1395.17 | bwd_microstep: 6832.48 | bwd_inner_microstep: 6789.73 | bwd_allreduce_microstep: 42.70 | step_microstep: 1867.64 + [2024-04-04 09:36:38,623] [INFO] [logging.py:96:log_dist] [Rank 0] time (ms) | fwd: 1395.15 | bwd: 6832.49 | bwd_inner: 6789.73 | bwd_allreduce: 42.71 | step: 1867.65 + iteration 2/ 317892 | consumed samples: 64 | consumed tokens: 262144 | elapsed time per iteration (ms): 10154.3 | learning rate: 3.000E-04 | global batch size: 32 | lm loss: 1.766422E+01 | loss scale: 1.0 | actual seqlen: 4096 | number of skipped iterations: 0 | number of nan iterations: 0 | samples per second: 3.151 | tokens per gpu per second(tgs): 1613.503 | TFLOPs: 98.29 | + + # ... + ``` + +
+ +
+ +
+ + + + + + + + + + + From a3424b66083018d764e474abd93f2fee19ab81a4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 12 Mar 2025 11:26:07 -0500 Subject: [PATCH 398/590] docs: Update `ALCF/README.md` --- ALCF/README.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/ALCF/README.md b/ALCF/README.md index f89b6daccbf..a6ee6a06558 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -61,6 +61,9 @@ PBS_O_WORKDIR=$(pwd) bash train_aGPT_7B.sh ``` + This will default to using the default AuroraGPT-7B architecture with the + full [Dolma (v1.7)](https://huggingface.co/datasets/allenai/dolma) dataset. +
[Overridable Options]: This is a simple subset of the overridable options. From bf50938e0a104dc8932941d750da2912a66ba825 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 12 Mar 2025 11:26:19 -0500 Subject: [PATCH 399/590] docs: Update `ALCF/notes/deprecated.md` --- ALCF/notes/deprecated.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/deprecated.md b/ALCF/notes/deprecated.md index 878986c4ef7..c6413d5b764 100644 --- a/ALCF/notes/deprecated.md +++ b/ALCF/notes/deprecated.md @@ -1,6 +1,6 @@ # Deprecated -This contains the (now deprecated) contents of the [`ALCF/README.md`](ALCF/README.md) file. +This contains the (now deprecated) contents from the [`ALCF/README.md`](ALCF/README.md) file. To launch on {`Polaris`, `Aurora`, `Sunspot`} @ [ALCF](https://alcf.anl.gov): From 6fcf7d565d6efe6cc93075007d33e81a716e913e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 12 Mar 2025 12:14:14 -0500 Subject: [PATCH 400/590] Update README.md --- ALCF/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/README.md b/ALCF/README.md index a6ee6a06558..e782b27a939 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -44,7 +44,7 @@ 1. 🍋 [saforem2 / `ezpz`](https://github.com/saforem2/ezpz): ```bash - python3 -m pip install "https://github.com/saforem2/ezpz" --require-virtualenv + python3 -m pip install "git+https://github.com/saforem2/ezpz" --require-virtualenv ``` 1. [microsoft / `DeepSpeed`](https://github.com/microsoft/DeepSpeed): From a5fea867500151ad66897e0627ec402290237307 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Thu, 20 Mar 2025 15:52:53 +0000 Subject: [PATCH 401/590] added muon --- megatron/arguments.py | 24 +++++++++++++++++++++++- megatron/checkpointing.py | 5 +++++ megatron/optimizer/__init__.py | 26 ++++++++++++++++++++++++++ 3 files changed, 54 insertions(+), 1 deletion(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 1c3ac06c424..76da8f62b3a 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -810,6 +810,27 @@ def _add_regularization_args(parser): help='Momentum factor for sgd') + + + group.add_argument('--muon-momentum', type=float, default=0.95, + help='Momentum factor for Muon ') + group.add_argument('--muon-nesterov', type=bool, default=True, + help='Whether to use Nesterov in the internal SGD') + group.add_argument('--muon-ns-steps', type=int, default=6, + help='The number of Newton-Schulz iterations to run') + group.add_argument('--muonadamw-beta1', type=float, default=0.9, + help='beta1 for internal adamw') + group.add_argument('--muonadamw-beta2', type=float, default=0.999, + help='beta2 for internal adamw') + group.add_argument('--muonadamw-eps', type=float, default=1e-12, + help='epsilon for internal adamw') + + + + + + + return parser @@ -979,7 +1000,8 @@ def _add_training_args(parser): 'sgd', 'sgdschedulefree', 'sophiag', - 'adopt' + 'adopt', + 'muon' ], help='Optimizer function' ) diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index a4f82ec9d3a..14e3c87f087 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -319,6 +319,11 @@ def state_dict_for_save_checkpoint_deepspeed(destination=None, prefix='', keep_v # Trim off the filename and mp_rank_* directory. for _ in range(3): checkpoint_name = os.path.dirname(checkpoint_name) + #try: + # model[0].save_checkpoint(checkpoint_name, client_state=state_dict) + #except Exception: + # import ezpz + # ezpz.utils.breakpoint(0) model[0].save_checkpoint(checkpoint_name, client_state=state_dict) if args.no_pipeline_parallel: diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 27527beef51..499993aaf46 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -371,6 +371,32 @@ def optimizer_hook(p): eps = args.adam_eps, weight_decay=args.weight_decay ) + #--- muon ----------------------------------------------- + elif str(args.optimizer).lower() == 'muon': + from .muon import Muon + + muon_param_groups = [ + group for group in param_groups + if any( # Use `any()` instead of `all()` to check at least one parameter + p.ndim >= 2 and not (p.ndim == 2 and p.shape[0] > 10000) + for p in group["params"] + ) + ] + + adamw_param_groups = [group for group in param_groups if group not in muon_param_groups] + + optimizer = Muon( + lr=args.lr, + wd=args.weight_decay, + all_params=param_groups, + muon_params=muon_param_groups, + momentum=args.muon_momentum, + nesterov=args.muon_nesterov, + ns_steps=args.muon_ns_steps, + adamw_params=adamw_param_groups, + adamw_betas=(args.muonadamw_beta1,args.muonadamw_beta2), + adamw_eps=args.muonadamw_eps + ) else: raise TypeError(f'{args.optimizer} optimizer is not supported.') assert optimizer is not None From f1451950e17cc3913b9e30414f497848cc4b0627 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Mon, 24 Mar 2025 21:58:33 +0000 Subject: [PATCH 402/590] added muon optimizer --- megatron/optimizer/__init__.py | 14 +--------- megatron/optimizer_param_scheduler.py | 38 +++++++++++++++++++++++++++ megatron/training.py | 3 ++- megatron/utils.py | 2 ++ 4 files changed, 43 insertions(+), 14 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 499993aaf46..f4ee7cc776d 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -374,26 +374,14 @@ def optimizer_hook(p): #--- muon ----------------------------------------------- elif str(args.optimizer).lower() == 'muon': from .muon import Muon - - muon_param_groups = [ - group for group in param_groups - if any( # Use `any()` instead of `all()` to check at least one parameter - p.ndim >= 2 and not (p.ndim == 2 and p.shape[0] > 10000) - for p in group["params"] - ) - ] - - adamw_param_groups = [group for group in param_groups if group not in muon_param_groups] optimizer = Muon( + param_groups, lr=args.lr, wd=args.weight_decay, - all_params=param_groups, - muon_params=muon_param_groups, momentum=args.muon_momentum, nesterov=args.muon_nesterov, ns_steps=args.muon_ns_steps, - adamw_params=adamw_param_groups, adamw_betas=(args.muonadamw_beta1,args.muonadamw_beta2), adamw_eps=args.muonadamw_eps ) diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 7304b257550..41a005d5952 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -120,6 +120,44 @@ def get_lr(self): lr = self.max_lr * warmup_tokens ** 0.5 / (num_tokens ** 0.5) return max(self.min_lr, lr) + # --- Infinite Decay Options --- + elif self.lr_decay_style in ["infinite_cosine", "infinite_inv_sqrt"]: + # Compute iterations after warmup. (Using steps if lr_decay_tokens is None, + # otherwise tokens.) + if self.lr_decay_tokens is None: + num_iters_ = self.num_steps - self.lr_warmup_steps + else: + num_iters_ = self.num_tokens - self.lr_warmup_tokens + + if num_iters_ <= self.cooldown_iters: + if self.lr_decay_style == "infinite_cosine": + lr = self.constant_lr + ( + (self.start_lr - self.constant_lr) + / 2.0 + * (math.cos(math.pi * num_iters_ / self.cooldown_iters) + 1) + ) + else: # infinite_inv_sqrt + def inv_f(t): + return (1 / math.sqrt(1 + (self.timescale * t))) - 1 + lr = self.start_lr + ( + (self.constant_lr - self.start_lr) + / inv_f(1) + * inv_f(num_iters_ / self.cooldown_iters) + ) + else: + num_iters_ = num_iters_ - self.cooldown_iters + if num_iters_ <= self.constant_iters: + # Stay constant for constant_iters + lr = self.constant_lr + else: + # Exponential decay from constant_lr to min_lr. + # Note: self.warmup_iter is assumed to be defined (similar to lr_warmup_steps). + end_iter_ = self.end_iter - self.warmup_iter - self.cooldown_iters - self.constant_iters + num_iters_ = num_iters_ - self.constant_iters + exp_factor = -math.log(self.min_lr / self.constant_lr) / end_iter_ + lr = self.constant_lr * math.exp(-exp_factor * num_iters_) + return max(lr, self.min_lr) + if self.lr_decay_tokens is None: num_steps_ = self.num_steps - self.lr_warmup_steps decay_steps_ = self.lr_decay_steps - self.lr_warmup_steps diff --git a/megatron/training.py b/megatron/training.py index d39f21c1289..db1c74c87a5 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -683,9 +683,10 @@ def setup_model_and_optimizer( if teacher: optimizer = None else: + optimizer = get_megatron_optimizer( model, no_wd_decay_cond, scale_lr_cond, lr_mult - ) + ) # opt_param_scheduler is the old lr_scheduler plus weight decay scheduling opt_param_scheduler = get_optimizer_param_scheduler(optimizer) if args.deepspeed: diff --git a/megatron/utils.py b/megatron/utils.py index dc1dea0b3a0..81a3f4396df 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -625,3 +625,5 @@ def found_kill_switch(): return True else: return False + + From 8bfee6fe4f3d695245843fae454b26fbb6a978be Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Mon, 24 Mar 2025 22:22:23 +0000 Subject: [PATCH 403/590] added muon optimizer --- megatron/optimizer/muon.py | 246 +++++++++++++++++++++++++++++++++++++ 1 file changed, 246 insertions(+) create mode 100644 megatron/optimizer/muon.py diff --git a/megatron/optimizer/muon.py b/megatron/optimizer/muon.py new file mode 100644 index 00000000000..c25865b6a66 --- /dev/null +++ b/megatron/optimizer/muon.py @@ -0,0 +1,246 @@ +import os +import torch +import math +import torch.distributed as dist +from torch import Tensor + +# This code snippet is a modified version adapted from the following GitHub repository: +# https://github.com/KellerJordan/Muon/blob/master/muon.py +#and https://github.com/MoonshotAI/Moonlight/blob/master/examples/toy_train.py +@torch.compile +def zeropower_via_newtonschulz5(G, steps): + """ + Newton-Schulz iteration to compute the zeroth power / orthogonalization of G. + """ + assert len(G.shape) == 2 + a, b, c = (3.4445, -4.7750, 2.0315) + X = G.bfloat16() + if G.size(0) > G.size(1): + X = X.T + # Ensure spectral norm is at most 1 + X = X / (X.norm() + 1e-7) + # Perform the NS iterations + for _ in range(steps): + A = X @ X.T + B = ( + b * A + c * A @ A + ) + X = a * X + B @ X + + if G.size(0) > G.size(1): + X = X.T + return X + + +class Muon(torch.optim.Optimizer): + """ + Muon - MomentUm Orthogonalized by Newton-schulz + """ + + + def __init__( + self, + params, + lr=1e-3, + wd=0.1, + momentum=0.95, + nesterov=True, + ns_steps=5, + adamw_betas=(0.95, 0.95), + adamw_eps=1e-8, + ): + defaults = dict( + lr=lr, + wd=wd, + momentum=momentum, + nesterov=nesterov, + ns_steps=ns_steps, + adamw_betas=adamw_betas, + adamw_eps=adamw_eps, + ) + + # Initialize the base optimizer with all parameter groups + super(Muon, self).__init__(params, defaults) + + # Process parameter groups to determine which will use Muon and which will use AdamW + for i, group in enumerate(self.param_groups): + # Ensure each group has all required parameters with defaults + for k, v in defaults.items(): + group.setdefault(k, v) + + # Mark parameters as using Muon or AdamW + group['use_muon_list'] = [] + + for p_idx, p in enumerate(group['params']): + if p.grad is None: + continue + + use_muon = p.ndim == 2 + + # If parameter is 2D but has a large dimension, it's likely an embedding or LM head, need to change this!!! + if use_muon and max(p.shape) > 10000: + use_muon = False + + group['use_muon_list'].append(use_muon) + + # Initialize parameter state + state = self.state[p] + if len(state) == 0: + state['use_muon'] = use_muon + if use_muon: + state['momentum_buffer'] = torch.zeros_like(p.grad) + else: + state['step'] = torch.tensor(0.0) + state['moment1'] = torch.zeros_like(p.grad) + state['moment2'] = torch.zeros_like(p.grad) + + def __setstate__(self, state): + """ + Handle state loading for the optimizer. + """ + super(Muon, self).__setstate__(state) + + # Ensure all parameter groups have the required defaults + for group in self.param_groups: + group.setdefault("nesterov", True) + group.setdefault("momentum", 0.95) + group.setdefault("ns_steps", 5) + group.setdefault("adamw_betas", (0.95, 0.95)) + group.setdefault("adamw_eps", 1e-8) + group.setdefault("wd", 0.1) + group.setdefault("lr", 1e-3) + group.setdefault("use_muon_list", []) + + # Convert step from float to tensor if needed + state_values = list(self.state.values()) + if state_values and 'step' in state_values[0]: + step_is_tensor = torch.is_tensor(state_values[0]['step']) + if not step_is_tensor: + for s in state_values: + if 'step' in s: + s['step'] = torch.tensor(float(s['step'])) + + def adjust_lr_for_muon(self, lr, param_shape): + """ + Adjust learning rate based on parameter shape for Muon. + """ + A, B = param_shape[:2] + # We adjust the learning rate based on the size of the parameter matrix + adjusted_ratio = 0.2 * math.sqrt(max(A, B)) + adjusted_lr = lr * adjusted_ratio + return adjusted_lr + + def step(self, closure=None): + """Perform a single optimization step.""" + loss = None + if closure is not None: + with torch.enable_grad(): + loss = closure() + + for group_idx, group in enumerate(self.param_groups): + lr = group["lr"] + wd = group["wd"] + momentum = group["momentum"] + nesterov = group["nesterov"] + ns_steps = group["ns_steps"] + adamw_betas = group["adamw_betas"] + adamw_eps = group["adamw_eps"] + + # Get use_muon_list (initialize if not present) + if 'use_muon_list' not in group or not group['use_muon_list']: + group['use_muon_list'] = [] + for p in group['params']: + if p.grad is None: + continue + # Default determination of whether to use Muon + use_muon = p.ndim == 2 + # Don't use Muon for embeddings/LM heads (approximated by large dimension check) + if use_muon and max(p.shape) > 10000: + use_muon = False + group['use_muon_list'].append(use_muon) + + # Apply optimization step to each parameter + for param_idx, p in enumerate(group['params']): + if p.grad is None: + continue + + grad = p.grad + + # Initialize state if needed + state = self.state[p] + if len(state) == 0: + # Determine if we should use Muon for this parameter + use_muon = False + if p.ndim == 2: + # Don't use Muon for embeddings/LM heads (approximated by large dimension check) + use_muon = max(p.shape) <= 10000 + + state['use_muon'] = use_muon + + if use_muon: + state['momentum_buffer'] = torch.zeros_like(grad) + else: + state['step'] = torch.tensor(0.0) + state['moment1'] = torch.zeros_like(grad) + state['moment2'] = torch.zeros_like(grad) + + # Check if we should use Muon for this parameter + use_muon = state.get('use_muon', False) + + if use_muon: + # Muon optimization + if grad.ndim > 2: + grad = grad.view(grad.size(0), -1) + + # Initialize momentum buffer if not present + if "momentum_buffer" not in state: + state["momentum_buffer"] = torch.zeros_like(grad) + + buf = state["momentum_buffer"] + buf.mul_(momentum).add_(grad) + + if nesterov: + g = grad.add(buf, alpha=momentum) + else: + g = buf + + u = zeropower_via_newtonschulz5(g, steps=ns_steps) + + # Scale update + adjusted_lr = self.adjust_lr_for_muon(lr, p.shape) + + # Apply weight decay + p.data.mul_(1 - lr * wd) + + # Apply update + p.data.add_(u, alpha=-adjusted_lr) + else: + # AdamW optimization + beta1, beta2 = adamw_betas + eps = adamw_eps + weight_decay = wd + + # Initialize AdamW state if not present + if "step" not in state: + state["step"] = torch.tensor(0.0) + state["moment1"] = torch.zeros_like(grad) + state["moment2"] = torch.zeros_like(grad) + + state["step"] += 1 + step = state["step"] + buf1 = state["moment1"] + buf2 = state["moment2"] + + buf1.lerp_(grad, 1 - beta1) + buf2.lerp_(grad.square(), 1 - beta2) + + g = buf1 / (eps + buf2.sqrt()) + + bias_correction1 = 1 - beta1**step + bias_correction2 = 1 - beta2**step + scale = bias_correction1 / bias_correction2**0.5 + + p.data.mul_(1 - lr * weight_decay) + p.data.add_(g, alpha=-lr / scale) + + return loss From 444af685164adf020a1d91b78e17fe33b7fca4be Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Mar 2025 21:48:44 -0500 Subject: [PATCH 404/590] chore: formatting in `megatron/model/__init__.py` --- megatron/model/__init__.py | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) diff --git a/megatron/model/__init__.py b/megatron/model/__init__.py index 6b3923aa977..bc5c003e7cf 100644 --- a/megatron/model/__init__.py +++ b/megatron/model/__init__.py @@ -12,14 +12,12 @@ if accelerator is not None and accelerator.device_name() == "cuda": from .fused_layer_norm import MixedFusedLayerNorm as LayerNorm - -# try: -# from apex.normalization import MixedFusedRMSNorm as RMSNorm # type:ignore -# - # HAS_APEX = True - # except Exception: - # HAS_APEX = False from .rmsnorm import RMSNorm + # try: + # from apex.normalization import FusedLayerNorm as LayerNorm # type:ignore + # HAS_APEX = True + # except Exception: + # HAS_APEX = False else: if hasattr(torch.xpu, "IpexRmsNorm"): from .fused_rmsnorm import RMSNorm @@ -27,7 +25,6 @@ from .rmsnorm import RMSNorm # noqa:E401,E402,F401 from torch.nn import LayerNorm # noqa:E401,E402,F401 - from .distributed import DistributedDataParallel # noqa:E401,E402,F401 from .bert_model import BertModel # noqa:E401,E402,F401 from .gpt_model import GPTModel, GPTModelPipe # noqa:E401,E402,F401 From 9c472b4dd5ff2712f7a74dc4f81e6325ca797a4c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 25 Mar 2025 21:49:06 -0500 Subject: [PATCH 405/590] chore: formatting `megatron/utils.py` --- megatron/utils.py | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/megatron/utils.py b/megatron/utils.py index 81a3f4396df..5f1c66640d1 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -8,6 +8,7 @@ from typing import Optional import torch +import torch.distributed from torch.nn.parallel import DistributedDataParallel as torchDDP from deepspeed.accelerator import get_accelerator @@ -201,7 +202,7 @@ def calc_params_l2_norm(model): # Calculate norm dummy_overflow_buf = get_accelerator().IntTensor([0]) - if get_accelerator().device_name() == "cuda" and HAS_APEX: + if ACCELERATOR.device_name() == "cuda" and HAS_APEX: norm, _ = multi_tensor_applier( amp_C.multi_tensor_l2norm, dummy_overflow_buf, @@ -390,9 +391,10 @@ def is_aml(): def is_rank_0(): """Check whether it is rank 0. For AML, check if it is rank 0 of a node""" if torch.distributed.is_initialized(): + assert ACCELERATOR is not None if torch.distributed.get_rank() == 0 or ( is_aml() - and (torch.distributed.get_rank() % get_accelerator().device_count()) == 0 + and (torch.distributed.get_rank() % ACCELERATOR.device_count()) == 0 ): return True else: @@ -479,7 +481,7 @@ def throughput_calculator(model, args, iteration_time, total_iterations): # correction has been made to TFLOPs formula due to incorrect behavior # observed with selective recompute when GQA not used and for all with GQA seq_len = args.seq_length - if hasattr(args, 'actual_seq_length'): + if hasattr(args, "actual_seq_length"): seq_len = args.actual_seq_length pre_and_post_mha_gemm_macs = ( batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len @@ -530,7 +532,7 @@ def checkpoint_throughput_calculator(model, latency_second): checkpoint_GB = approx_parameters_in_billions * checkpoint_multiplier GB_per_second = checkpoint_GB / latency_second print_rank_0( - f"Checkpoint Save GB: {round(checkpoint_GB, 3)}, GB/Sec: {round(GB_per_second,2)}, Latency(second): {round(latency_second, 3)}" + f"Checkpoint Save GB: {round(checkpoint_GB, 3)}, GB/Sec: {round(GB_per_second, 2)}, Latency(second): {round(latency_second, 3)}" ) @@ -583,7 +585,7 @@ def dump_weights(preamble, iteration, model, optimizer, tensor=None): # return if tensor is not None: - orig_tensor = tensor + orig_tensor = tensor # XXX: Unused ???? if hasattr(tensor, "_hp_param"): numel = tensor._hp_param.numel() # // dp_size tensor = tensor.flatten().narrow(0, 0, numel) @@ -625,5 +627,3 @@ def found_kill_switch(): return True else: return False - - From 267d6508143172c00b6531bc4c718a64318d4407 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Mon, 21 Apr 2025 15:36:52 +0000 Subject: [PATCH 406/590] added infinite schedulers --- megatron/arguments.py | 20 ++++- megatron/optimizer_param_scheduler.py | 106 ++++++++++++++++++-------- megatron/training.py | 15 ++++ 3 files changed, 110 insertions(+), 31 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 76da8f62b3a..1bfce8c93db 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1098,7 +1098,7 @@ def _add_learning_rate_args(parser): 'and initial warmup, the learing rate at each ' 'iteration would be different.') group.add_argument('--lr-decay-style', type=str, default='linear', - choices=['constant', 'linear', 'cosine', 'inverse-square-root'], + choices=['constant', 'linear', 'cosine', 'inverse-square-root', 'infinite-cosine', 'infinite-inv-square-root'], help='Learning rate decay function.') group.add_argument('--lr-decay-iters', type=int, default=None, help='number of iterations to decay learning rate over,' @@ -1121,12 +1121,30 @@ def _add_learning_rate_args(parser): group.add_argument('--lr-warmup-tokens', type=int, default=None, help='number of tokens to linearly warmup ' 'learning rate over.') + group.add_argument('--lr-constant-tokens', type=int, default=None, + help='number of tokens to keep constant ' + 'learning rate over.') + group.add_argument('--lr-constant-fraction', type=float, default=0.001, + help='fraction of lr-constant-(iters/samples) to use ' + 'for constant phase (as a float)') + + group.add_argument('--lr-cooldown-tokens', type=int, default=None, + help='number of tokens to cooldown learning rate over,' + ' If not None will override iter/sample-based decay') + group.add_argument('--lr-cooldown-fraction', type=float, default=0.65, + help='fraction -(iters/samples) to use ' + 'for cooldownt phase (as a float)') + group.add_argument('--warmup', type=int, default=None, help='Old lr warmup argument, do not use. Use one of the' '--lr-warmup-* arguments above') group.add_argument('--min-lr', type=float, default=0.0, help='Minumum value for learning rate. The scheduler' 'clip values below this threshold.') + group.add_argument('--constant-lr', type=float, default=0.00011, + help='Constant value for learning rate') + group.add_argument('--timescale', type=float, default=10.0, + help='Timescale for the steepness of the inverse square root cooldown') group.add_argument('--override-opt_param-scheduler', action='store_true', help='Reset the values of the scheduler (learning rate,' 'warmup iterations, minimum learning rate, maximum ' diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 41a005d5952..0dafa896985 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -12,6 +12,10 @@ class OptimizerParamScheduler(object): def __init__(self, optimizer, max_lr, min_lr, lr_warmup_steps, lr_decay_steps, lr_decay_style, start_wd, end_wd, wd_incr_steps, wd_incr_style, + constant_lr = 0.0, + lr_constant_steps=None, + lr_cooldown_steps=None, + timescale=None, use_checkpoint_opt_param_scheduler=True, override_opt_param_scheduler=False): args = get_args() @@ -20,18 +24,27 @@ def __init__(self, optimizer, max_lr, min_lr, self.max_lr = float(max_lr) self.min_lr = min_lr + self.constant_lr = constant_lr assert self.min_lr >= 0.0 assert self.max_lr >= self.min_lr + assert self.constant_lr <= self.max_lr self.lr_warmup_steps = lr_warmup_steps self.num_steps = 0 self.lr_decay_steps = lr_decay_steps + self.lr_cooldown_steps = lr_cooldown_steps + self.lr_constant_steps = lr_constant_steps + self.end_steps = lr_decay_steps assert self.lr_decay_steps > 0 assert self.lr_warmup_steps < self.lr_decay_steps + assert self.lr_constant_steps < self.lr_decay_steps self.lr_decay_tokens = args.lr_decay_tokens self.num_tokens = 0 self.lr_warmup_tokens = args.lr_warmup_tokens + self.lr_cooldown_tokens = args.lr_cooldown_tokens + self.lr_constant_tokens = args.lr_constant_tokens + self.end_tokens = args.train_tokens self.lr_decay_style = lr_decay_style @@ -121,41 +134,74 @@ def get_lr(self): return max(self.min_lr, lr) # --- Infinite Decay Options --- - elif self.lr_decay_style in ["infinite_cosine", "infinite_inv_sqrt"]: + elif self.lr_decay_style in ["infinite-cosine", "infinite-inv-square-root"]: # Compute iterations after warmup. (Using steps if lr_decay_tokens is None, # otherwise tokens.) + assert self.constant_lr > self.min_lr + delta_lr = (self.max_lr - self.constant_lr) if self.lr_decay_tokens is None: - num_iters_ = self.num_steps - self.lr_warmup_steps - else: - num_iters_ = self.num_tokens - self.lr_warmup_tokens - - if num_iters_ <= self.cooldown_iters: - if self.lr_decay_style == "infinite_cosine": - lr = self.constant_lr + ( - (self.start_lr - self.constant_lr) - / 2.0 - * (math.cos(math.pi * num_iters_ / self.cooldown_iters) + 1) - ) - else: # infinite_inv_sqrt - def inv_f(t): - return (1 / math.sqrt(1 + (self.timescale * t))) - 1 - lr = self.start_lr + ( - (self.constant_lr - self.start_lr) - / inv_f(1) - * inv_f(num_iters_ / self.cooldown_iters) - ) - else: - num_iters_ = num_iters_ - self.cooldown_iters - if num_iters_ <= self.constant_iters: - # Stay constant for constant_iters - lr = self.constant_lr + num_steps_ = self.num_steps - self.lr_warmup_steps + cooldown_steps_ = self.lr_cooldown_steps - self.lr_warmup_steps + if self.constant_steps is None: + raise Exception('Constant LR steps need to be provided for infinite schedulers') + if num_steps_ <= cooldown_steps_: + cooldown_ratio = float(num_steps_) / float(cooldown_steps_) + + + if self.lr_decay_style == "infinite_cosine": + coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) + lr = self.constant_lr + delta_lr * coeff + else: # infinite_inv_sqrt + def inv_f(t): + return (1 / math.sqrt(1 + (self.timescale * t))) - 1 + coeff = inv_f(1) * inv_f(cooldown_ratio) + lr = self.start_lr - delta_lr * coeff + else: + num_steps_ = num_steps_ - cooldown_steps_ + if num_steps_ <= self.lr_constant_steps: + # Stay constant for constant_iters + lr = self.constant_lr + else: # Exponential decay from constant_lr to min_lr. - # Note: self.warmup_iter is assumed to be defined (similar to lr_warmup_steps). - end_iter_ = self.end_iter - self.warmup_iter - self.cooldown_iters - self.constant_iters - num_iters_ = num_iters_ - self.constant_iters - exp_factor = -math.log(self.min_lr / self.constant_lr) / end_iter_ - lr = self.constant_lr * math.exp(-exp_factor * num_iters_) + + end_steps_ = self.end_steps - self.warmup_steps - cooldown_steps_ - self.lr_constant_steps + num_steps_ = num_steps_ - self.lr_constant_steps + exp_factor = -math.log(self.min_lr / self.constant_lr) / end_steps_ + lr = self.constant_lr * math.exp(-exp_factor * num_steps_) + + # token based decay + else: + num_tokens_ = self.num_tokens_ - self.lr_warmup_tokens + cooldown_tokens_ = self.lr_cooldown_tokens - self.lr_warmup_tokens + if self.constant_tokens is None: + raise Exception('Constant LR tokens need to be provided for infinite schedulers') + if num_tokens_ <= cooldown_tokens_: + cooldown_ratio = float(num_tokens_) / float(cooldown_tokens_) + + if self.lr_decay_style == "infinite_cosine": + coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) + lr = self.constant_lr + delta_lr * coeff + else: # infinite_inv_sqrt + def inv_f(t): + return (1 / math.sqrt(1 + (self.timescale * t))) - 1 + coeff = inv_f(1) * inv_f(cooldown_ratio) + lr = self.start_lr - delta_lr * coeff + else: + num_tokens_ = num_tokens_ - cooldown_tokens_ + + if num_tokens_ <= self.lr_constant_tokens: + # Stay constant for constant_tokens period + lr = self.constant_lr + else: + # Exponential decay from constant_lr to min_lr + + end_tokens_ = self.end_tokens - self.lr_warmup_tokens - cooldown_tokens_ - self.lr_constant_tokens + num_tokens_ = num_tokens_ - constant_period + exp_factor = -math.log(self.min_lr / self.constant_lr) / end_tokens_ + lr = self.constant_lr * math.exp(-exp_factor * num_tokens_) + + return max(lr, self.min_lr) if self.lr_decay_tokens is None: diff --git a/megatron/training.py b/megatron/training.py index db1c74c87a5..d66a504053c 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -562,12 +562,19 @@ def get_optimizer_param_scheduler(optimizer): if args.train_iters: if args.lr_decay_iters is None: args.lr_decay_iters = args.train_iters + lr_decay_steps = args.lr_decay_iters * args.global_batch_size wd_incr_steps = args.train_iters * args.global_batch_size if args.lr_warmup_fraction is not None: lr_warmup_steps = args.lr_warmup_fraction * lr_decay_steps else: lr_warmup_steps = args.lr_warmup_iters * args.global_batch_size + + if args.lr_constant_fraction is not None: + lr_constant_steps = args.lr_constant_fraction * lr_decay_steps + else: + lr_constant_steps = args.lr_constant_iters * args.global_batch_size + lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps # Sample-based training. elif args.train_samples: # We need to set training iters for later use. Technically @@ -582,6 +589,11 @@ def get_optimizer_param_scheduler(optimizer): lr_warmup_steps = args.lr_warmup_fraction * lr_decay_steps else: lr_warmup_steps = args.lr_warmup_samples + if args.lr_constant_fraction is not None: + lr_constant_steps = args.lr_constant_fraction * lr_decay_steps + else: + lr_constant_steps = args.lr_constant_samples + lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps else: raise Exception("either train-iters or train-samples should be provided.") @@ -596,6 +608,9 @@ def get_optimizer_param_scheduler(optimizer): end_wd=args.end_weight_decay, wd_incr_steps=wd_incr_steps, wd_incr_style=args.weight_decay_incr_style, + constant_lr=args.constant_lr, + lr_constant_steps=lr_constant_steps, + lr_cooldown_steps=lr_cooldown_steps, use_checkpoint_opt_param_scheduler=args.use_checkpoint_opt_param_scheduler, override_opt_param_scheduler=args.override_opt_param_scheduler, ) From 6b6c63d2e61a861d06b4f52b0347b8fefcb380c9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 22 Apr 2025 16:12:49 -0500 Subject: [PATCH 407/590] fix: Fix imports in `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 24 +++++++++++++----------- 1 file changed, 13 insertions(+), 11 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index c196b19a420..4b7c98d16eb 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -1,8 +1,12 @@ # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain GPT""" + +# ruff: noqa:F401,E402 import time from typing import Callable +import ezpz + from mpi4py import MPI comm = MPI.COMM_WORLD @@ -44,20 +48,18 @@ from torch import nn import torch.nn.functional as F -import ezpz as ez - dt_imports = time.time() - python_start_time t0_setup = time.time() # ---- [SETUP COMMS] ------------------------ # if str(os.environ.get('LAUNCH_CMD', 'mpich')).lower() == 'mpich': -RANK = ez.setup_torch(backend="deepspeed") # , timeout=7200) +RANK = ezpz.setup_torch(backend="deepspeed") # , timeout=7200) dt_setup = time.time() - t0_setup # else: -# RANK = ez.get_rank() -WORLD_SIZE = ez.get_world_size() -LOCAL_RANK = ez.get_local_rank() -DEVICE_TYPE = ez.dist.get_torch_device_type() +# RANK = ezpz.get_rank() +WORLD_SIZE = ezpz.get_world_size() +LOCAL_RANK = ezpz.get_local_rank() +DEVICE_TYPE = ezpz.get_torch_device_type() if torch.cuda.is_available(): torch.cuda.set_device(LOCAL_RANK) @@ -67,7 +69,7 @@ log.setLevel(LOG_LEVEL) if RANK == 0 else log.setLevel("CRITICAL") log.info(f"Import python modules in {dt_imports} seconds") -log.info(f"ez.setup_torch time: {dt_setup} seconds") +log.info(f"ezpz.setup_torch time: {dt_setup} seconds") # ---- [SETUP WANDB FROM RANK 0] -------------- WANDB_MODE = os.environ.get("WANDB_MODE", None) @@ -78,10 +80,10 @@ os.environ.get("WANDB_PROJECT", "AuroraGPT"), # look for WANDB_PROJECT in env ) log.info(f"Setting up W&B from: {RANK} with {project_name}") - _ = ez.setup_wandb(project_name=project_name) + _ = ezpz.setup_wandb(project_name=project_name) -@ez.dist.timeitlogit(rank=RANK) +@ezpz.timeitlogit(rank=RANK) def model_provider(pre_process=True, post_process=True): """Build the model.""" log.info("building GPT model ...") @@ -453,7 +455,7 @@ def forward_step(data_iterator, model) -> tuple[torch.Tensor | None, Callable]: return output_tensor, partial(loss_func, loss_mask, moe_loss, mos_loss) -@ez.dist.timeitlogit(rank=RANK) +@ezpz.timeitlogit(rank=RANK) def train_valid_test_datasets_provider(train_val_test_num_samples): """Build train, valid, and test datasets.""" t0 = time.perf_counter() From 7ee03d7c7ffc160401ecbf632cc8ec767103b87e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 23 Apr 2025 11:00:26 -0500 Subject: [PATCH 408/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 71 +++++++++++++++++++++++-------------------------- 1 file changed, 34 insertions(+), 37 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 627f5838a10..b5243cc5b9d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -101,18 +101,18 @@ setup() { # Identify machine we're on get_machine || exit ########################################################################## - # ezpz_setup will: - # 1. Setup python - # - load base conda - # - (if necessary) create virtual environment on top of base conda - # - activate virtual environment from ^ - # 2. Install ezpz (if needed) - # 3. Parse PBS_* environment variables to determine: - # - NHOSTS (by counting number of lines in $PBS_NODEFILE) - # - NGPU_PER_HOST (by magic) - # - NGPUS (= NHOSTS * NGPU_PER_HOST) - # 4. Use these (^) to build our launch command - ezpz_setup || exit + # # ezpz_setup will: + # # 1. Setup python + # # - load base conda + # # - (if necessary) create virtual environment on top of base conda + # # - activate virtual environment from ^ + # # 2. Install ezpz (if needed) + # # 3. Parse PBS_* environment variables to determine: + # # - NHOSTS (by counting number of lines in $PBS_NODEFILE) + # # - NGPU_PER_HOST (by magic) + # # - NGPUS (= NHOSTS * NGPU_PER_HOST) + # # 4. Use these (^) to build our launch command + # ezpz_setup || exit ########################################################################## install_dependencies # Set command line arguments to pass to `"${EXEC}"` @@ -126,7 +126,7 @@ setup() { set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit @@ -796,30 +796,27 @@ make_ds_hostfile() { # 3. Call `_ezpz_install` (from `Megatron-DeepSpeed/ALCF/helpers.sh`): # - Install ezpz from `"${WORKING_DIR}/depz/ezpz/"` ########################################### -ezpz_setup() { - # setup_alcf "$@" - # file=$(mktemp) - # curl -Ls https://raw.githubusercontent.com/saforem2/ezpz/main/src/ezpz/bin/getjobenv > "${file}" - # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh - ezdir="${WORKING_DIR}/deps/ezpz" - if [[ -d "${ezdir}" ]]; then - echo "Found ezpz in ${ezdir}" - else - mkdir -p "$(dirname "${ezdir}")" - git clone https://github.com/saforem2/ezpz "${ezdir}" - fi - # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh - source "${ezdir}/src/ezpz/bin/utils.sh" || exit - ezpz_setup_python - ezpz_setup_job "$@" - ezpz_pip_loc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') - if [[ -z "${ezpz_pip_loc:-}" ]]; then - printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" - python3 -m pip install -e "${ezdir}" --require-virtualenv - else - printf "[ezpz_install] Found ezpz @ %s\n" "${ezpz_pip_loc}" - fi -} +# ezpz_setup() { +# source <() +# ezdir="${WORKING_DIR}/deps/ezpz" +# if [[ -d "${ezdir}" ]]; then +# echo "Found ezpz in ${ezdir}" +# else +# mkdir -p "$(dirname "${ezdir}")" +# git clone https://github.com/saforem2/ezpz "${ezdir}" +# fi +# # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh +# source "${ezdir}/src/ezpz/bin/utils.sh" || exit +# ezpz_setup_python +# ezpz_setup_job "$@" +# ezpz_pip_loc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') +# if [[ -z "${ezpz_pip_loc:-}" ]]; then +# printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" +# python3 -m pip install -e "${ezdir}" --require-virtualenv +# else +# printf "[ezpz_install] Found ezpz @ %s\n" "${ezpz_pip_loc}" +# fi +# } ####################################################################### # ezpz_test: Run simple test to make sure all nodes in working order From 3c3eb45be438921bd14ce2fc2e0e49a1c1062f5e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 23 Apr 2025 11:00:40 -0500 Subject: [PATCH 409/590] feat: Add `train_alcf.sh` --- train_alcf.sh | 44 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 44 insertions(+) create mode 100644 train_alcf.sh diff --git a/train_alcf.sh b/train_alcf.sh new file mode 100644 index 00000000000..fe0fd7d7dc4 --- /dev/null +++ b/train_alcf.sh @@ -0,0 +1,44 @@ +#!/bin/bash --login +#PBS -q lustre_scaling +#PBS -A Aurora_Deployment +#PBS -j oe + +##################################### +# AuroraGPT-7B +# +# Main production script for training +# AuroraGPT-7B @ ALCF +##################################### +train_aGPT() { + + # 1. Navigate into `$PBS_O_WORKDIR` + # cd "${PBS_O_WORKDIR}" || exit + HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE + GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + + if [[ "${HERE}" != "${PBS_O_WORKDIR}" ]]; then + printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" + printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" + printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" + fi + + # 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: + source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') && ezpz_setup_env + python3 -m pip install "git+https://github.com/saforem2/ezpz" + + # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup + source "${HERE}/ALCF/helpers.sh" || exit + + # 3. call `setup` from `./ALCF/helpers.sh` + setup "$@" || exit + # export run_cmd="${run_cmd}" + echo "${run_cmd[@]}" | tee -a "${OUTPUT_LOG}" + + # 4. Tell user where to find output + printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + + # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} + eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" +} + +train_aGPT "$@" From f40b7e5e3e7e52b85aca4c6ab755c166676c7eb8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 23 Apr 2025 14:29:03 -0500 Subject: [PATCH 410/590] fix: Update `train_alcf.sh` --- train_alcf.sh | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index fe0fd7d7dc4..261c4d8a564 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -16,14 +16,16 @@ train_aGPT() { HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + # 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: + source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') + if [[ "${HERE}" != "${PBS_O_WORKDIR}" ]]; then printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" fi - # 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: - source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') && ezpz_setup_env + ezpz_setup_env python3 -m pip install "git+https://github.com/saforem2/ezpz" # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup From 952940a8944eb5975c7fc261ac656e0a324ee07b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 30 Apr 2025 08:42:39 -0500 Subject: [PATCH 411/590] fix: Fix `ALCF/helpers.sh` --- ALCF/helpers.sh | 187 ++++++++++++++++++++++-------------------------- 1 file changed, 84 insertions(+), 103 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index b5243cc5b9d..d95bfc3214e 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -20,6 +20,13 @@ # command for launching across all GPUs in our active PBS job. ############################################################################### +############################################################################### +# Source: +# [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) +source <(curl -L https://bit.ly/ezpz-utils) || exit +ezpz_setup_job >/dev/null || exit +############################################################################### + ################## # helpers_main # @@ -65,41 +72,30 @@ helpers_main() { # # - Explicitly, this will: # - Identify the machine we're on -# # - Setup `python` # 1. Load `conda` # 2. Setup `venv` on top of `conda` -# # - Ensure all dependencies are installed -# # - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) # - Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) # - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) -# # - Set runtime options -# # - Build `deepspeed_config.json` -# # - Build {logs, checkpoints, etc} dirs, named according to specifics of # current run -# # - Specify additional `deepspeed` arguments -# # - Ensure executable exists at expected path -# # - Setup data + tokenizer via `TOKENIZER_TYPE` -# # - Print job info -# # - Save `.env` to `CKPT_DIR` for safe keeping -# # - Check that we're not already running, and if so, exit. -# # - Setup run command to be executed. ############################################################################## setup() { # Identify machine we're on - get_machine || exit + mn=$(ezpz_get_machine_name) + export MACHINE="${mn}" + # get_machine || exit ########################################################################## # # ezpz_setup will: # # 1. Setup python @@ -126,7 +122,7 @@ setup() { set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit @@ -216,7 +212,7 @@ setup_run_cmd() { "--tensorboard-dir ${TBDIR}" ) fi - dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt}" train_args=() if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then @@ -297,60 +293,60 @@ save_dotenv() { fi } -###################################################################### -# get_machine_name: -# -# Return current machine name, as lowercase string +####################################################################### +## get_machine_name: +## +## Return current machine name, as lowercase string +## +## Example: +## ```bash +## $ machine_name=$(get_machine_name) +## $ echo "machine_name: ${machine_name}" +## ``` +####################################################################### +#get_machine_name() { +# if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then +# machine="aurora" +# elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then +# machine="sunspot" +# elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then +# if [[ "${PBS_O_HOST}" == sirius* ]]; then +# machine="sirius" +# else +# machine="polaris" +# fi +# elif [[ $(hostname) == sophia* ]]; then +# machine="sophia" +# elif [[ $(hostname) == nid* ]]; then +# machine="perlmutter" +# else +# machine=$(hostname) +# fi +# echo "${machine}" +#} # -# Example: -# ```bash -# $ machine_name=$(get_machine_name) -# $ echo "machine_name: ${machine_name}" -# ``` -###################################################################### -get_machine_name() { - if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then - machine="aurora" - elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then - machine="sunspot" - elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then - machine="sirius" - else - machine="polaris" - fi - elif [[ $(hostname) == sophia* ]]; then - machine="sophia" - elif [[ $(hostname) == nid* ]]; then - machine="perlmutter" - else - machine=$(hostname) - fi - echo "${machine}" -} - -get_machine() { - machine=$(hostname) - if [[ $(hostname) == x4* ]]; then - machine="aurora" - elif [[ $(hostname) == x1* ]]; then - machine="sunspot" - elif [[ $(hostname) == x3* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then - machine="sirius" - else - machine="polaris" - fi - elif [[ $(hostname) == sophia* ]]; then - machine="sophia" - elif [[ $(hostname) == nid* ]]; then - machine="perlmutter" - else - echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." - fi - export MACHINE="${machine}" - printf "Running on: %s\n" "$(printBlue "${MACHINE}")" -} +#get_machine() { +# machine=$(hostname) +# if [[ $(hostname) == x4* ]]; then +# machine="aurora" +# elif [[ $(hostname) == x1* ]]; then +# machine="sunspot" +# elif [[ $(hostname) == x3* ]]; then +# if [[ "${PBS_O_HOST}" == sirius* ]]; then +# machine="sirius" +# else +# machine="polaris" +# fi +# elif [[ $(hostname) == sophia* ]]; then +# machine="sophia" +# elif [[ $(hostname) == nid* ]]; then +# machine="perlmutter" +# else +# echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." +# fi +# export MACHINE="${machine}" +# printf "Running on: %s\n" "$(printBlue "${MACHINE}")" +#} check_and_kill_if_running() { RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') @@ -397,6 +393,10 @@ printJobInfo() { # will launch with `deepspeed` instead of `mpiexec`. ############################################################################# setupLauncher() { + shell_type=$(basename "${SHELL}") + if [[ "${shell_type}" == "bash" ]]; then + shopt -s expand_aliases + fi if [[ "$#" == 1 ]]; then local dist_launcher="$1" else @@ -410,8 +410,13 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else + # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job + # echo "ezpz_launch: $(which ezpz_launch)" + # export -f ezpz_launch + # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" + if [[ -n "${DIST_LAUNCH}" ]]; then - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" elif [[ "${mn}" == "sophia" ]]; then @@ -548,7 +553,8 @@ set_ccl_vars_on_aurora() { export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich export FI_PROVIDER=cxi # Required by Aurora mpich export PALS_PMI=pmix # Required by Aurora mpich - export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + # export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + export CCL_ATL_TRANSPORT=ofi # [SF]: Changed 04/30/2025 export TORCH_LLM_ALLREDUCE=1 export CCL_SYCL_ESIMD=1 export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) @@ -576,7 +582,7 @@ setParams() { # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" @@ -616,9 +622,9 @@ setParams() { # ---- [Sophia] ---------------------- elif [[ "${mn}" == sophia* ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" @@ -951,7 +957,7 @@ install_dependencies() { echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." python3 -m pip install -r "${depsfile}" --require-virtualenv 1>/dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then # install_deepspeed_for_xpu || exit # fi @@ -1050,7 +1056,7 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") - machine=$(get_machine_name) + machine=$(ezpz_get_machine_name) if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then @@ -1094,7 +1100,7 @@ setup_tokenizer_and_data() { ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" ############################################################################ # set `dfl` to `dfl_fallback` if not passed as an argument, @@ -1335,31 +1341,6 @@ $flops_profiler EOT } -# ##################### -# # train -# ##################### -# train() { -# # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] -# cd "${PBS_O_WORKDIR}" || exit -# HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE -# # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] -# source "${HERE}/ALCF/helpers.sh" || exit -# # 3. call `setup` from `./ALCF/helpers.sh` -# # export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" -# setup || exit -# # 4. Take custom args -# export custom_args=" $@" -# # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} -# export run_cmd="${run_cmd} ${custom_args}" -# # 6. Add "${run_cmd}" to output log -# echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" -# # 7. Tell user where to find output -# printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" -# # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} -# eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" -# set +x -# } - ############################################### # Helper functions for printing colored text ############################################### From b136aa51f72025367a1d41ff501de33be33a960e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 30 Apr 2025 10:21:48 -0500 Subject: [PATCH 412/590] fix: Replace `eval` with `bash -c` in `train_alcf.sh` --- train_alcf.sh | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 261c4d8a564..3c176fa0d07 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -19,7 +19,8 @@ train_aGPT() { # 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') - if [[ "${HERE}" != "${PBS_O_WORKDIR}" ]]; then + if [[ "${HERE}" != "${PBS_O_WORKDIR:-}" ]]; then + export PBS_O_WORKDIR="${HERE}" printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" @@ -40,7 +41,8 @@ train_aGPT() { printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} - eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" + # eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" + bash -c "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" } train_aGPT "$@" From 9a3f6bda7b36bc3c3c9ace81afe97de68e417962 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 1 May 2025 07:23:37 -0500 Subject: [PATCH 413/590] chore: Fix unset `CFLAGS` in `ALCF/helpers.sh` --- ALCF/helpers.sh | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d95bfc3214e..d75acb304e3 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -846,12 +846,12 @@ ezpz_test() { saveDSenv() { echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { - echo "PATH=${PATH}" - echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH}" + echo "PATH=${PATH:-}" + echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH:-}" echo "http_proxy=${http_proxy:-}" echo "https_proxy=${https_proxy:-}" - echo "CFLAGS=${CFLAGS}" - echo "PYTHONUSERBASE=$PYTHONUSERBASE" + echo "CFLAGS=${CFLAGS:-}" + echo "PYTHONUSERBASE=${PYTHONUSERBASE:-}" } >.deepspeed_env } From 61ce1c5af577f17be324267df910f3cf7b0f4c20 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 5 May 2025 08:20:36 -0500 Subject: [PATCH 414/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 3c176fa0d07..6dc64245409 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -26,8 +26,13 @@ train_aGPT() { printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" fi - ezpz_setup_env - python3 -m pip install "git+https://github.com/saforem2/ezpz" + ezpz_setup_env || exit + if command -v "ezpz-test"; then + printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" + else + printf "[!! %s] ezpz is not installed. Installing...\n" "$(printRed "WARNING")" + python3 -m pip install "git+https://github.com/saforem2/ezpz" + fi # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup source "${HERE}/ALCF/helpers.sh" || exit @@ -42,7 +47,13 @@ train_aGPT() { # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} # eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" - bash -c "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" + if [[ "${DEBUG:-}" ]]; then + set -x + bash -c "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" + set +x + else + bash -c "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" + fi } train_aGPT "$@" From ba01f41853a1f125a22fe24f76fa55c5363be96f Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Tue, 6 May 2025 01:09:19 +0000 Subject: [PATCH 415/590] added lr finder logic --- ALCF/helpers.sh | 4 +- megatron/arguments.py | 5 +- megatron/optimizer_param_scheduler.py | 2 +- megatron/training.py | 135 ++++++++++++++++++++++++++ 4 files changed, 142 insertions(+), 4 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 627f5838a10..f42f331af59 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -669,7 +669,7 @@ setParams() { export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export SEQ=${SEQ:-2048} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS @@ -684,7 +684,7 @@ setParams() { printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then - export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} + export TRAIN_TOKENS=${TRAIN_TOKENS:-10000000000} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" diff --git a/megatron/arguments.py b/megatron/arguments.py index 1bfce8c93db..9d9d52284d7 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1134,7 +1134,10 @@ def _add_learning_rate_args(parser): group.add_argument('--lr-cooldown-fraction', type=float, default=0.65, help='fraction -(iters/samples) to use ' 'for cooldownt phase (as a float)') - + group.add_argument('--lr-finder', action='store_true', + help='Run learning rate finder mode for 10% of training data then exit') + + group.add_argument('--warmup', type=int, default=None, help='Old lr warmup argument, do not use. Use one of the' '--lr-warmup-* arguments above') diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 0dafa896985..5933ffa048a 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -27,7 +27,6 @@ def __init__(self, optimizer, max_lr, min_lr, self.constant_lr = constant_lr assert self.min_lr >= 0.0 assert self.max_lr >= self.min_lr - assert self.constant_lr <= self.max_lr self.lr_warmup_steps = lr_warmup_steps self.num_steps = 0 @@ -138,6 +137,7 @@ def get_lr(self): # Compute iterations after warmup. (Using steps if lr_decay_tokens is None, # otherwise tokens.) assert self.constant_lr > self.min_lr + assert self.constant_lr <= self.max_lr delta_lr = (self.max_lr - self.constant_lr) if self.lr_decay_tokens is None: num_steps_ = self.num_steps - self.lr_warmup_steps diff --git a/megatron/training.py b/megatron/training.py index d66a504053c..43710fd691d 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1016,6 +1016,8 @@ def train( write_args_to_tensorboard() assert accelerator is not None setup_profiler(args, accelerator.device_name()) + + if args.random_ltd: # random-ltd requires different randomness on each rank import random @@ -1057,6 +1059,139 @@ def train( args.train_range_to_skip[1::2], ) ) + + # Learning rate finder mode + if hasattr(args, 'lr_finder') and args.lr_finder: + # Calculate number of iterations to use (10% of train_iters) + finder_iters = max(1, int(args.train_iters * 0.1)) + + # Initialize tracking variables for LR finder + lr_finder_losses = [] + lr_finder_lrs = [] + + # Initialize loss smoothing variables + avg_loss = 0.0 + best_loss = float('inf') + batch_num = 0 + beta = 0.98 # Smoothing factor + + # Set initial learning rate and calculate multiplier + init_lr = 1e-6 + max_lr = 1.0 + mult = (max_lr / init_lr) ** (1 / finder_iters) + + # Set initial learning rate + curr_lr = init_lr + for param_group in optimizer.param_groups: + param_group['lr'] = curr_lr + + # Turn on training mode which enables dropout + for model_module in model: + model_module.train() + + # Get configuration for training + config = core_transformer_config_from_args(args) + if not args.deepspeed: + config.grad_scale_func = optimizer.scale_loss + config.timers = timers + + log.info(f"Running learning rate finder for {finder_iters} iterations (10% of {args.train_iters})") + + # Main LR finder loop + for i in range(finder_iters): + # Execute training step + loss_dict, skipped_iter, grad_norm, num_zeros_in_grad = train_step( + forward_step_func, + train_data_iterator, + model, + optimizer, + opt_param_scheduler, + config, + ) + + # Skip iterations that didn't update weights + if skipped_iter: + continue + + # Get loss value (use first available loss if multiple present) + if 'lm loss' in loss_dict: + loss_val = loss_dict['lm loss'] + else: + loss_val = next(iter(loss_dict.values())) + + if isinstance(loss_val, torch.Tensor): + loss_val = loss_val.item() + + # Update batch counter + batch_num += 1 + + # Compute smoothed loss + avg_loss = beta * avg_loss + (1 - beta) * loss_val + smoothed_loss = avg_loss / (1 - beta**batch_num) + + # Stop if the loss is exploding + if batch_num > 1 and smoothed_loss > 4 * best_loss: + log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") + break + + # Record the best loss + if smoothed_loss < best_loss or batch_num == 1: + best_loss = smoothed_loss + + # Record values for plotting + lr_finder_losses.append(smoothed_loss) + lr_finder_lrs.append(curr_lr) + + # Log to wandb +# if wandb is not None and wandb.run is not None: + # wandb.log({ + # "lr_finder/learning_rate": curr_lr, + # "lr_finder/raw_loss": loss_val, + # "lr_finder/smoothed_loss": smoothed_loss, + # "lr_finder/iteration": i, + # "lr_finder/log_lr": math.log10(curr_lr) + # }) + + + # Print progress + if (i+1) % args.log_interval == 0: + log.info(f"LR Finder: iteration {i+1}/{finder_iters}, " + f"lr: {curr_lr:.8f}, loss: {smoothed_loss:.4f}") + + # Update the learning rate for the next step (bypassing scheduler) + curr_lr *= mult + for param_group in optimizer.param_groups: + param_group['lr'] = curr_lr + + # Save raw data (on rank 0 only) + if mpu.get_data_parallel_rank() == 0: + # Create the results directory if it doesn't exist + os.makedirs(f'{args.save}/lr_finder', exist_ok=True) + + # Save raw data to a simple CSV file + import csv + with open(f'{args.save}/lr_finder/lr_finder_data.csv', 'w', newline='') as f: + writer = csv.writer(f) + writer.writerow(['learning_rate', 'loss']) + for lr, loss in zip(lr_finder_lrs, lr_finder_losses): + writer.writerow([lr, loss]) + + # Also save as numpy arrays for convenience + try: + import numpy as np + np.savez( + f'{args.save}/lr_finder/lr_finder_data.npz', + learning_rates=np.array(lr_finder_lrs), + losses=np.array(lr_finder_losses) + ) + except ImportError: + pass + + log.info(f"LR finder completed. Results saved to {args.save}/lr_finder/") + + # Return after LR finder is done + return args.iteration + while iteration < args.train_iters and ( args.train_tokens is None or args.consumed_train_tokens < args.train_tokens ): From d7e12df8a4ff4bf54212a679aef0029a610284aa Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 5 May 2025 21:06:41 -0500 Subject: [PATCH 416/590] fix: Remove call to `set_ccl_vars_on_aurora` in `ALCF/helpers.sh` --- ALCF/helpers.sh | 288 +++++++++++++++++++++++++++--------------------- 1 file changed, 163 insertions(+), 125 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index d75acb304e3..fd01f24ce2c 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -20,6 +20,7 @@ # command for launching across all GPUs in our active PBS job. ############################################################################### + ############################################################################### # Source: # [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) @@ -72,30 +73,41 @@ helpers_main() { # # - Explicitly, this will: # - Identify the machine we're on +# # - Setup `python` # 1. Load `conda` # 2. Setup `venv` on top of `conda` +# # - Ensure all dependencies are installed +# # - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) # - Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) # - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) +# # - Set runtime options +# # - Build `deepspeed_config.json` +# # - Build {logs, checkpoints, etc} dirs, named according to specifics of # current run +# # - Specify additional `deepspeed` arguments +# # - Ensure executable exists at expected path +# # - Setup data + tokenizer via `TOKENIZER_TYPE` +# # - Print job info +# # - Save `.env` to `CKPT_DIR` for safe keeping +# # - Check that we're not already running, and if so, exit. +# # - Setup run command to be executed. ############################################################################## setup() { # Identify machine we're on - mn=$(ezpz_get_machine_name) - export MACHINE="${mn}" - # get_machine || exit + get_machine || exit ########################################################################## # # ezpz_setup will: # # 1. Setup python @@ -122,7 +134,7 @@ setup() { set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit @@ -137,38 +149,8 @@ setup() { setup_run_cmd "$@" || exit } -##################################################### -# setup_run_cmd -# -# Build run command to be executed. -##################################################### -setup_run_cmd() { - ############################## - # take in additional arguments - # and append them directly to - # the end of the `run_cmd` - # custom_args="$@" - custom_args=("$@") - ############################## - #### Make it easy to track experiments by date ################### - year="$(date "+%Y")" - month="$(date "+%m")" - day="$(date "+%Y-%m-%d")" - today="$(date "+%Y-%m-%d")" # kept for backwards compatibility - started_at="$(date "+%Y-%m-%d-%H%M%S")" - export YEAR="${year}" - export MONTH="${month}" - export DAY="${day}" - export TODAY="${today}" - export STARTED_AT="${started_at}" - ################################################################## - # NOTE: to launch with DeepSpeed instead of mpiexec: - # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` - ################################################################## - setupLauncher "${LAUNCH_WITH:-MPICH}" || exit - export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" - printf "\n" - echo "Using data_cache_path: ${data_cache_path}" + +get_training_args() { ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` @@ -212,7 +194,7 @@ setup_run_cmd() { "--tensorboard-dir ${TBDIR}" ) fi - dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt}" + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" train_args=() if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then @@ -269,11 +251,49 @@ setup_run_cmd() { "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" ) # "--adam-eps ${ADAM_EPS:-0.00001}" - cache_dir="${PBS_O_WORKDIR}/.cache/" - mkdir -p "${cache_dir}" - targs_cache="${cache_dir}/train_args.txt" - for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done + echo "${train_args[@]}" +} + +##################################################### +# setup_run_cmd +# +# Build run command to be executed. +##################################################### +setup_run_cmd() { + ############################## + # take in additional arguments + # and append them directly to + # the end of the `run_cmd` + # custom_args="$@" + custom_args=("$@") + ############################## + #### Make it easy to track experiments by date ################### + year="$(date "+%Y")" + month="$(date "+%m")" + day="$(date "+%Y-%m-%d")" + today="$(date "+%Y-%m-%d")" # kept for backwards compatibility + started_at="$(date "+%Y-%m-%d-%H%M%S")" + export YEAR="${year}" + export MONTH="${month}" + export DAY="${day}" + export TODAY="${today}" + export STARTED_AT="${started_at}" + ################################################################## + # NOTE: to launch with DeepSpeed instead of mpiexec: + # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` + ################################################################## + setupLauncher "${LAUNCH_WITH:-MPICH}" || exit + export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" + train_args=("$(get_training_args)") + cache_dir="${WORKING_DIR}/.cache/" + if [[ ! -d "${cache_dir}" ]]; then + mkdir -p "${cache_dir}" + fi + targs_cache="${cache_dir}/train_args_${started_at}.txt" + for arg in "${train_args[@]}"; do echo "${arg}" >"${targs_cache}"; done export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") + # printf "\n" + echo "Using data_cache_path: ${data_cache_path}" printf "Training Arguments: %s\n" "${TRAIN_ARGS[@]}" export run_cmd=("${LAUNCHER}" "${train_args[@]}") } @@ -293,60 +313,60 @@ save_dotenv() { fi } -####################################################################### -## get_machine_name: -## -## Return current machine name, as lowercase string -## -## Example: -## ```bash -## $ machine_name=$(get_machine_name) -## $ echo "machine_name: ${machine_name}" -## ``` -####################################################################### -#get_machine_name() { -# if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then -# machine="aurora" -# elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then -# machine="sunspot" -# elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then -# if [[ "${PBS_O_HOST}" == sirius* ]]; then -# machine="sirius" -# else -# machine="polaris" -# fi -# elif [[ $(hostname) == sophia* ]]; then -# machine="sophia" -# elif [[ $(hostname) == nid* ]]; then -# machine="perlmutter" -# else -# machine=$(hostname) -# fi -# echo "${machine}" -#} +###################################################################### +# get_machine_name: # -#get_machine() { -# machine=$(hostname) -# if [[ $(hostname) == x4* ]]; then -# machine="aurora" -# elif [[ $(hostname) == x1* ]]; then -# machine="sunspot" -# elif [[ $(hostname) == x3* ]]; then -# if [[ "${PBS_O_HOST}" == sirius* ]]; then -# machine="sirius" -# else -# machine="polaris" -# fi -# elif [[ $(hostname) == sophia* ]]; then -# machine="sophia" -# elif [[ $(hostname) == nid* ]]; then -# machine="perlmutter" -# else -# echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." -# fi -# export MACHINE="${machine}" -# printf "Running on: %s\n" "$(printBlue "${MACHINE}")" -#} +# Return current machine name, as lowercase string +# +# Example: +# ```bash +# $ machine_name=$(get_machine_name) +# $ echo "machine_name: ${machine_name}" +# ``` +###################################################################### +get_machine_name() { + if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then + machine="aurora" + elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then + machine="sunspot" + elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + machine="sirius" + else + machine="polaris" + fi + elif [[ $(hostname) == sophia* ]]; then + machine="sophia" + elif [[ $(hostname) == nid* ]]; then + machine="perlmutter" + else + machine=$(hostname) + fi + echo "${machine}" +} + +get_machine() { + machine=$(hostname) + if [[ $(hostname) == x4* ]]; then + machine="aurora" + elif [[ $(hostname) == x1* ]]; then + machine="sunspot" + elif [[ $(hostname) == x3* ]]; then + if [[ "${PBS_O_HOST}" == sirius* ]]; then + machine="sirius" + else + machine="polaris" + fi + elif [[ $(hostname) == sophia* ]]; then + machine="sophia" + elif [[ $(hostname) == nid* ]]; then + machine="perlmutter" + else + echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." + fi + export MACHINE="${machine}" + printf "Running on: %s\n" "$(printBlue "${MACHINE}")" +} check_and_kill_if_running() { RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') @@ -393,10 +413,6 @@ printJobInfo() { # will launch with `deepspeed` instead of `mpiexec`. ############################################################################# setupLauncher() { - shell_type=$(basename "${SHELL}") - if [[ "${shell_type}" == "bash" ]]; then - shopt -s expand_aliases - fi if [[ "$#" == 1 ]]; then local dist_launcher="$1" else @@ -410,13 +426,8 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job - # echo "ezpz_launch: $(which ezpz_launch)" - # export -f ezpz_launch - # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" - if [[ -n "${DIST_LAUNCH}" ]]; then - mn=$(ezpz_get_machine_name) + mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" elif [[ "${mn}" == "sophia" ]]; then @@ -553,8 +564,7 @@ set_ccl_vars_on_aurora() { export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich export FI_PROVIDER=cxi # Required by Aurora mpich export PALS_PMI=pmix # Required by Aurora mpich - # export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich - export CCL_ATL_TRANSPORT=ofi # [SF]: Changed 04/30/2025 + export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich export TORCH_LLM_ALLREDUCE=1 export CCL_SYCL_ESIMD=1 export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) @@ -582,7 +592,7 @@ setParams() { # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then - mn=$(ezpz_get_machine_name) + mn=$(get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" @@ -622,9 +632,9 @@ setParams() { # ---- [Sophia] ---------------------- elif [[ "${mn}" == sophia* ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" @@ -662,11 +672,14 @@ setParams() { fi export NGPU_PER_HOST="${NGPU_PER_HOST}" export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" - if [[ "${WORLD_SIZE}" -gt 1 && "${mn}" == "aurora" ]]; then - #### [sam: 08/17/2024] ########################################## - # Use best set of CCL env vars from Gordon Bell runs on Aurora - set_ccl_vars_on_aurora - fi + #### [sam: 05/05/2025] ################################################# + # NOTE: Commented out following release of frameworks-2025.0.0 + # if [[ "${WORLD_SIZE}" -gt 12 && "${mn}" == "aurora" ]]; then + # #### [sam: 08/17/2024] ########################################### + # # Use best set of CCL env vars from Gordon Bell runs on Aurora + # set_ccl_vars_on_aurora + # fi + ######################################################################## # +---[Llama2 7B Config]--------------------------------------------------+ # export MODEL_KEY="Llama-7B" export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS @@ -957,7 +970,7 @@ install_dependencies() { echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." python3 -m pip install -r "${depsfile}" --require-virtualenv 1>/dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then - mn=$(ezpz_get_machine_name) + mn=$(get_machine_name) # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then # install_deepspeed_for_xpu || exit # fi @@ -1056,7 +1069,7 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") - machine=$(ezpz_get_machine_name) + machine=$(get_machine_name) if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then @@ -1100,7 +1113,7 @@ setup_tokenizer_and_data() { ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# - mn=$(ezpz_get_machine_name) + mn=$(get_machine_name) dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" ############################################################################ # set `dfl` to `dfl_fallback` if not passed as an argument, @@ -1341,16 +1354,41 @@ $flops_profiler EOT } +# ##################### +# # train +# ##################### +# train() { +# # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] +# cd "${PBS_O_WORKDIR}" || exit +# HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE +# # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] +# source "${HERE}/ALCF/helpers.sh" || exit +# # 3. call `setup` from `./ALCF/helpers.sh` +# # export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" +# setup || exit +# # 4. Take custom args +# export custom_args=" $@" +# # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} +# export run_cmd="${run_cmd} ${custom_args}" +# # 6. Add "${run_cmd}" to output log +# echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" +# # 7. Tell user where to find output +# printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" +# # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} +# eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" +# set +x +# } + ############################################### # Helper functions for printing colored text ############################################### -RESET="\e[0m" -BLACK="\e[1;30m" -RED="\e[1;31m" -GREEN="\e[1;32m" -YELLOW="\e[1;33m" -BLUE="\e[1;34m" -CYAN="\e[1;35m" +# RESET="\e[0m" +# BLACK="\e[1;30m" +# RED="\e[1;31m" +# GREEN="\e[1;32m" +# YELLOW="\e[1;33m" +# BLUE="\e[1;34m" +# CYAN="\e[1;35m" # WHITE="\e[1;36m" printBlack() { From e8efc709508fe53acb62ac8b13c29cf8de34c076 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 5 May 2025 21:07:18 -0500 Subject: [PATCH 417/590] chore: Clean up `train_alcf.sh` --- train_alcf.sh | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 6dc64245409..2a6a1d3fce1 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -27,12 +27,13 @@ train_aGPT() { fi ezpz_setup_env || exit - if command -v "ezpz-test"; then - printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" - else - printf "[!! %s] ezpz is not installed. Installing...\n" "$(printRed "WARNING")" - python3 -m pip install "git+https://github.com/saforem2/ezpz" - fi + + # if command -v "ezpz-test"; then + # printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" + # else + # printf "[!! %s] ezpz is not installed. Installing...\n" "$(printRed "WARNING")" + # python3 -m pip install "git+https://github.com/saforem2/ezpz" + # fi # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup source "${HERE}/ALCF/helpers.sh" || exit From fa73d59cdd079111453976322013d80be6178afd Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 18 May 2025 19:06:33 -0500 Subject: [PATCH 418/590] feat: Resolve conflicts in `train_alcf.sh` --- train_alcf.sh | 64 ++++++++++++++++++++++++++++++++++----------------- 1 file changed, 43 insertions(+), 21 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 2a6a1d3fce1..6e475d22e8a 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -3,6 +3,42 @@ #PBS -A Aurora_Deployment #PBS -j oe +# 1. Navigate into `$PBS_O_WORKDIR` +# cd "${PBS_O_WORKDIR}" || exit +HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE +GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + +# 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: +# source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') + +# if [[ "${HERE}" != "${PBS_O_WORKDIR:-}" ]]; then +# export PBS_O_WORKDIR="${HERE}" +# printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" +# printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" +# printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" +# fi +# ezpz_setup_env || exit + +# source <(curl -L 'https://bit.ly/ezpz-utils') || exit +NO_COLOR=1 source ../../saforem2/ezpz/src/ezpz/bin/utils.sh || exit +NO_COLOR=1 ezpz_setup_env || exit +# ezpz_setup_env || exit +# if ! command -v "ezpz_launch"; then +# ezpz_setup_env || exit +# else +# log_message INFO "" +# fi + +if command -v "ezpz-test"; then + log_message INFO "${GREEN}✓${RESET} ezpz is already installed." + # printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" +else + log_message WARNING "${RED}✗${RESET} ezpz is not installed." + log_message INFO "Installing ezpz..." + python3 -m pip install "git+https://github.com/saforem2/ezpz" +fi + + ##################################### # AuroraGPT-7B # @@ -16,25 +52,6 @@ train_aGPT() { HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH - # 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: - source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') - - if [[ "${HERE}" != "${PBS_O_WORKDIR:-}" ]]; then - export PBS_O_WORKDIR="${HERE}" - printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" - printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" - printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" - fi - - ezpz_setup_env || exit - - # if command -v "ezpz-test"; then - # printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" - # else - # printf "[!! %s] ezpz is not installed. Installing...\n" "$(printRed "WARNING")" - # python3 -m pip install "git+https://github.com/saforem2/ezpz" - # fi - # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup source "${HERE}/ALCF/helpers.sh" || exit @@ -44,9 +61,10 @@ train_aGPT() { echo "${run_cmd[@]}" | tee -a "${OUTPUT_LOG}" # 4. Tell user where to find output - printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" + printf "Output will be saved to %s\n" "${OUTPUT_LOG}" | tee -a "${OUTPUT_LOG}" + # printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" - # 6. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} + # 5. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} # eval "${run_cmd[*]}" |& tee -a "${OUTPUT_LOG}" if [[ "${DEBUG:-}" ]]; then set -x @@ -57,4 +75,8 @@ train_aGPT() { fi } +# add trap train_aGPT "$@" +# pid=$(train_aGPT "$@") +# trap 'kill -TERM "$pid"' TERM INT +# wait "$pid" From ac0df1d8fdb154759551dbf2ae98217dca253865 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:01:17 -0500 Subject: [PATCH 419/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 22 +--------------------- 1 file changed, 1 insertion(+), 21 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 6e475d22e8a..31a8c38ef70 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -4,30 +4,10 @@ #PBS -j oe # 1. Navigate into `$PBS_O_WORKDIR` -# cd "${PBS_O_WORKDIR}" || exit HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH - -# 3. source `ezpz/bin/uitils.sh` and setup {job, python} environment: -# source <(curl 'https://raw.githubusercontent.com/saforem2/ezpz/refs/heads/main/src/ezpz/bin/utils.sh') - -# if [[ "${HERE}" != "${PBS_O_WORKDIR:-}" ]]; then -# export PBS_O_WORKDIR="${HERE}" -# printf "[!! %s] WARNING: Current working directory (%s) does not match PBS_O_WORKDIR (%s)\n" "$(printRed "WARNING")" "${HERE}" "${PBS_O_WORKDIR}" -# printf "[!! %s] This may cause issues with the job submission.\n" "$(printRed "WARNING")" -# printf "Setting PBS_O_WORKDIR to %s and continuing...\n" "${HERE}" -# fi -# ezpz_setup_env || exit - -# source <(curl -L 'https://bit.ly/ezpz-utils') || exit -NO_COLOR=1 source ../../saforem2/ezpz/src/ezpz/bin/utils.sh || exit +source <(curl -L https://bit.ly/ezpz-utils) || exit NO_COLOR=1 ezpz_setup_env || exit -# ezpz_setup_env || exit -# if ! command -v "ezpz_launch"; then -# ezpz_setup_env || exit -# else -# log_message INFO "" -# fi if command -v "ezpz-test"; then log_message INFO "${GREEN}✓${RESET} ezpz is already installed." From 48f300fadc5e090ff8f89b257ed50c9badf7366f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:02:41 -0500 Subject: [PATCH 420/590] feat: Add `ALCF/notes/AuroraGPT-70B.md` --- ALCF/notes/AuroraGPT-70B.md | 113 ++++++++++++++++++++++++++++++++++++ 1 file changed, 113 insertions(+) create mode 100644 ALCF/notes/AuroraGPT-70B.md diff --git a/ALCF/notes/AuroraGPT-70B.md b/ALCF/notes/AuroraGPT-70B.md new file mode 100644 index 00000000000..f2dfc46ba70 --- /dev/null +++ b/ALCF/notes/AuroraGPT-70B.md @@ -0,0 +1,113 @@ +# AuroraGPT-70B Performance Study + +```bash +TOTAL_PARAMS=68976648192 (68976.648M) +``` + +## Personal Experiments + +### Pipeline + ZeRO-1 + +```bash +PP="${NHOSTS}" \ + TP=1 \ + HEADS=64 \ + NLAYERS=80 \ + NUM_KV_HEAD=8 \ + FFN_HIDDEN_SIZE=28672 \ + HIDDEN=8192 \ + MICRO_BATCH=1 \ + GRAD_ACC_STEPS="${GAS}" \ + SEQ_LEN=8192 \ + DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt \ + USE_ACTIVATION_CHECKPOINTING=1 \ + bash train_alcf.sh +``` + +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | +| :----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :----------------------------------------------------------------------------- | +| 8 | 96 | 8 | 12 | 1 | 1 | 1 | 12 | True | OOM | OOM | OOM | [chocolate-meadow-2125](https://wandb.ai/aurora_gpt/AuroraGPT/runs/uhfkwmp2) | +| 8 | 96 | 8 | 6 | 2 | 1 | 1 | 6 | True | 10.26 | 0.541 | 23.084 | [glamorous-darkness-2126](https://wandb.ai/aurora_gpt/AuroraGPT/runs/tdehvqey) | +| 8 | 96 | 8 | 6 | 2 | 1 | 1 | 6 | False | 13.16 | 0.694 | 29.608 | [proud-frog-2153](https://wandb.ai/aurora_gpt/AuroraGPT/runs/ho1lwmer) | +| 8 | 96 | 8 | 6 | 2 | 1 | 2 | 12 | True | 17.75 | 0.936 | 39.927 | [stoic-dragon-2127](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2vay044x) | +| 8 | 96 | 8 | 6 | 2 | 1 | 4 | 24 | True | 28.93 | 1.525 | 65.081 | [drawn-wildflower-2128](https://wandb.ai/aurora_gpt/AuroraGPT/runs/qjaqpbxg) | +| 8 | 96 | 8 | 6 | 2 | 1 | 8 | 48 | True | 42.27 | 2.229 | 95.103 | [fresh-waterfall-2129](https://wandb.ai/aurora_gpt/AuroraGPT/runs/8m38fq95) | +| 8 | 96 | 8 | 6 | 2 | 1 | 16 | 96 | True | 54.66 | 2.882 | 122.974 | [divine-waterfall-2130](https://wandb.ai/aurora_gpt/AuroraGPT/runs/on3m4isp) | +| 8 | 96 | 8 | 6 | 2 | 1 | 32 | 192 | True | 64.02 | 3.376 | 144.037 | [wild-bee-2131](https://wandb.ai/aurora_gpt/AuroraGPT/runs/hitrbi6d) | +|   | | | | | | | | | | | | | +| 16 | 192 | 16 | 12 | 1 | 1 | 1 | 12 | | 5.6 | 0.594 | 12.695 | [fluent-surf-2118](https://wandb.ai/aurora_gpt/AuroraGPT/runs/0y250j0i) | +| 16 | 192 | 16 | 12 | 1 | 1 | 2 | 24 | | 10.25 | 1.081 | 23.064 | [dulcet-salad-2119](https://wandb.ai/aurora_gpt/AuroraGPT/runs/5f1rdn9p) | +| 16 | 192 | 16 | 12 | 1 | 1 | 4 | 48 | | 18.10 | 1.908 | 40.713 | [rose-blaze-2120](https://wandb.ai/aurora_gpt/AuroraGPT/runs/9obt1iqi) | +| 16 | 192 | 16 | 12 | 1 | 1 | 8 | 96 | | 29.49 | 3.110 | 66.342 | [azure-jazz-2121](https://wandb.ai/aurora_gpt/AuroraGPT/runs/oaft4n5p) | +| 16 | 192 | 16 | 12 | 1 | 1 | 16 | 192 | | OOM | OOM | OOM | OOM | +|   | | | | | | | | | | | | | + +- **Note**: + - `sps`: samples per second + - `tpgps`: tokens per GPU per second + +## 42B Model (Pure Tensor Parallelism) (TP=6) + +- 42B param model + +```bash +#[🐍 aurora_nre_models_frameworks-2025.0.0](👻 aurora_nre_models_frameworks-2025.0.0) +#[/f/d/f/p/a/Megatron-DeepSpeed][🌱 saforem2/dev][📦📝🤷✓] [⏱️ 4m15s] +#[06/14/25 @ 16:33:06][x4515c6s0b0n0] +; TP=6 PP=1 HEADS=$((TP * $((48 / TP)))) NLAYERS=$((TP * $((48 / TP)))) NUM_KV_HEAD=$((TP * $((8 / TP)))) FFN_HIDDEN_SIZE=$((TP * $((28672 / TP)))) HIDDEN=$((HEADS * $((8192 / HEADS)))) ZERO_STAGE=2 MICRO_BATCH=1 GRAD_ACC_STEPS=1 DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt bash train_alcf.sh +``` + +```bash +==== ARCHITECTURE ==== +NLAYERS: 48 +GAS: 1 +PP: 1 +HEADS: 48 +USE_ACTIVATION_CHECKPOINTING: 0 +FFN_HIDDEN_SIZE: 28668 +SEQ: 4096 +GBS: 16 +DP: 16 +NUM_KV_HEAD: 6 +HIDDEN: 8160 +TP: 6 +SP: 1 +MBS: 1 +====================== +``` + +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | TFLOPS | samples / s | tok / gpu / s | wandb | +| ----: | :--: | --: | --: | --: | --: | --: | --: | -----: | ----------: | ------------: | :----------------------------------------------------------------------------- | +| 8 | 96 | 1 | 16 | 6 | 1 | 1 | 16 | 19.76 | 1.740 | 74.257 | [royal-forest-2135](https://wandb.ai/aurora_gpt/AuroraGPT/runs/o3cyy1mq) | +| 8 | 96 | 1 | 16 | 6 | 1 | 2 | 32 | 22.13 | 1.949 | 83.170 | [balmy-terrain-2136](https://wandb.ai/aurora_gpt/AuroraGPT/runs/rpo1fbk9) | +| 8 | 96 | 1 | 16 | 6 | 1 | 4 | 64 | 23.60 | 2.078 | 88.676 | [eternal-wildflower-2137](https://wandb.ai/aurora_gpt/AuroraGPT/runs/faxpu0r6) | + +## From Deepak + +### Pipeline + ZeRO-1 + +| Nodes | NGPU | PP | DP | TP | MBS | GBS | TFLOPS | samples / s | +| ----: | ---: | --: | --: | --: | --: | --: | -----: | ----------: | +| 8 | 96 | 8 | 12 | 1 | 1 | 96 | OOM | OOM | +| 8 | 96 | 8 | 6 | 2 | 1 | 48 | 58.2 | 1.1 | +| 8 | 96 | 8 | 6 | 2 | 1 | 96 | 72.5 | 1.4 | +| 8 | 96 | 8 | 6 | 2 | 1 | 192 | 85.6 | 1.7 | +| 16 | 192 | 16 | 12 | 1 | 1 | 192 | 59.9 | 2.3 | +| 16 | 192 | 16 | 12 | 1 | 1 | 384 | 76.1 | 3.0 | +| 16 | 192 | 16 | 12 | 1 | 1 | 768 | OOM | OOM | +| 32 | 384 | 16 | 24 | 1 | 1 | 384 | OOM | OOM | +| 32 | 384 | 16 | 24 | 1 | 1 | 768 | OOM | OOM | + +### ZeRO-3 MiCS Performance Data + +| Nodes | DP | TP | MBS | GBS | TFLOPS | sample/s | +| :---: | :-: | :-: | :-: | :-: | :----: | :------: | +| 4 | 48 | 1 | 1 | 48 | 98.5 | 0.96 | +| 8 | 96 | 1 | 1 | 96 | 96.7 | 1.88 | +| 16 | 192 | 1 | 1 | 192 | 95.4 | 3.70 | + +| Nodes | DP | TP | MBS | GBS | TFLOPS | sample/s | +| :---: | :-: | :-: | :-: | :-: | :----: | :------: | +| 4 | 48 | 1 | 1 | 48 | 96.8 | 0.94 | +| 8 | 96 | 1 | 1 | 96 | 99.1 | 1.92 | +| 16 | 192 | 1 | 1 | 192 | 97.2 | 3.77 | From 5cd5da9f05a65c47713245251a7a67ad28db6d69 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:03:29 -0500 Subject: [PATCH 421/590] chore: Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 4b7c98d16eb..0dfd33fd67c 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -105,7 +105,7 @@ def model_provider(pre_process=True, post_process=True): with deepspeed_zero_init( data_parallel_group=dpg, remote_device=(None if args.remote_device == "none" else args.remote_device), - config_dict_or_path=args.deepspeed_config, # _dict, + config_dict_or_path=args.deepspeed_config_dict, # _dict, enabled=args.zero_stage == 3, mpu=mpu, ): From be64b022ee2aec7ba06367907c2d213ea348b84b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:03:56 -0500 Subject: [PATCH 422/590] chore: Update `megatron/training.py` --- megatron/training.py | 54 ++++++++++++++++++++++++-------------------- 1 file changed, 29 insertions(+), 25 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index d39f21c1289..27e791a6a8b 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1,6 +1,7 @@ # Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain utilities.""" + import time # The earliest we can measure the start time. @@ -190,7 +191,7 @@ def pretrain( [_TRAIN_START_TIME], dtype=torch.double, device=DEVICE_TYPE ) tdist.all_reduce(start_time_tensor, op=tdist.ReduceOp.MIN) - log.info(f"allreduce call time: {time.time()-before_allreduce} seconds") + log.info(f"allreduce call time: {time.time() - before_allreduce} seconds") _TRAIN_START_TIME = start_time_tensor.item() log.info( "time to initialize megatron (seconds)={:.3f}".format( @@ -241,7 +242,7 @@ def pretrain( build_train_valid_test_datasets_provider=train_valid_test_dataset_provider, ) timers("model-and-optimizer-setup").stop() - print_datetime("after model, optimizer, and learning rate " "scheduler are built") + print_datetime("after model, optimizer, and learning rate scheduler are built") # Data stuff. timers("train/valid/test-data-iterators-setup", log_level=0).start(barrier=True) if args.virtual_pipeline_model_parallel_size is not None: @@ -421,9 +422,9 @@ def get_model( mpu.get_pipeline_model_parallel_world_size() > 1 and args.virtual_pipeline_model_parallel_size is not None ): - assert ( - model_type != ModelType.encoder_and_decoder - ), "Interleaved schedule not supported for model with both encoder and decoder" + assert model_type != ModelType.encoder_and_decoder, ( + "Interleaved schedule not supported for model with both encoder and decoder" + ) model = [] for i in range(args.virtual_pipeline_model_parallel_size): mpu.set_virtual_pipeline_model_parallel_rank(i) @@ -442,9 +443,9 @@ def get_model( add_decoder = True if model_type == ModelType.encoder_and_decoder: if mpu.get_pipeline_model_parallel_world_size() > 1: - assert ( - args.pipeline_model_parallel_split_rank is not None - ), "Split rank needs to be specified for model with both encoder and decoder" + assert args.pipeline_model_parallel_split_rank is not None, ( + "Split rank needs to be specified for model with both encoder and decoder" + ) rank = mpu.get_pipeline_model_parallel_rank() split_rank = args.pipeline_model_parallel_split_rank world_size = mpu.get_pipeline_model_parallel_world_size() @@ -470,9 +471,9 @@ def get_model( # Disallow training and inference with Transformer Engine # for non-GPT models args.allow_transformer_engine = all([type(m) == GPTModel for m in model]) - assert ( - args.allow_transformer_engine or args.transformer_impl == "local" - ), "Transformer Engine is only approved for GPT models" + assert args.allow_transformer_engine or args.transformer_impl == "local", ( + "Transformer Engine is only approved for GPT models" + ) # Set tensor model parallel attributes if not set. # Only parameters that are already tensor model parallel have these @@ -545,8 +546,9 @@ def get_model( model_module.broadcast_params() else: raise NotImplementedError( - "Unknown DDP implementation specified: " - "{}. Exiting.".format(args.DDP_impl) + "Unknown DDP implementation specified: {}. Exiting.".format( + args.DDP_impl + ) ) return model @@ -622,7 +624,7 @@ def load_model_weights_only(model_provider_func): ) assert not isinstance(model, deepspeed.PipelineEngine), ( - "Weight loading only mode is not supported in " "pipeline parallelism yet." + "Weight loading only mode is not supported in pipeline parallelism yet." ) model = [model] print_datetime("before load checkpoint") @@ -1408,14 +1410,16 @@ def evaluate_and_print_results( ) key = "test" if test else "val" if wandb is not None and wandb.run is not None: - wandb.log({ - f"{key}/iteration": iteration, - **{f"{key}/{k}": v for k, v in total_loss_dict.items()}, - **{ - f"{key}/ppl_{k}": math.exp(min(20, v.item())) - for k, v in total_loss_dict.items() - }, - }) + wandb.log( + { + f"{key}/iteration": iteration, + **{f"{key}/{k}": v for k, v in total_loss_dict.items()}, + **{ + f"{key}/ppl_{k}": math.exp(min(20, v.item())) + for k, v in total_loss_dict.items() + }, + } + ) string = " validation loss at {} | ".format(prefix) for key in total_loss_dict: string += f"{key} value={total_loss_dict[key].item():.6f}" @@ -1510,9 +1514,9 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider log.info("> building train, validation, and test datasets ...") # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: - assert ( - args.train_samples is None - ), "only backward compatiblity support for iteration-based training" + assert args.train_samples is None, ( + "only backward compatiblity support for iteration-based training" + ) args.consumed_train_samples = args.iteration * args.global_batch_size if args.iteration > 0 and args.consumed_valid_samples == 0: if args.train_samples is None: From 08ab7ee633df7b8624e192fe83b3d7a4d1da927f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:06:52 -0500 Subject: [PATCH 423/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 31a8c38ef70..ac879e20889 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -3,7 +3,6 @@ #PBS -A Aurora_Deployment #PBS -j oe -# 1. Navigate into `$PBS_O_WORKDIR` HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH source <(curl -L https://bit.ly/ezpz-utils) || exit @@ -15,7 +14,7 @@ if command -v "ezpz-test"; then else log_message WARNING "${RED}✗${RESET} ezpz is not installed." log_message INFO "Installing ezpz..." - python3 -m pip install "git+https://github.com/saforem2/ezpz" + python3 -m pip install "git+https://github.com/saforem2/ezpz" || exit fi @@ -55,8 +54,8 @@ train_aGPT() { fi } -# add trap + +# Kill any existing MPI processes +ezpz_kill_mpi || exit + train_aGPT "$@" -# pid=$(train_aGPT "$@") -# trap 'kill -TERM "$pid"' TERM INT -# wait "$pid" From 8c13fefc1918c735ab9999f9084af4d0cd5eb883 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:11:36 -0500 Subject: [PATCH 424/590] chore: Remove tensoboard tracking in `megatron/training_log.py` --- megatron/training_log.py | 277 +-------------------------------------- 1 file changed, 6 insertions(+), 271 deletions(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index 3eb96c392d9..933badc31d0 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -141,189 +141,23 @@ def training_log( # Tensorboard values. # Timer requires all the ranks to call. - if args.log_timers_to_tensorboard and ( - iteration % args.tensorboard_log_interval == 0 - ): - timers.write(timers_to_log, writer, iteration, normalizer=total_iterations) - if writer and (iteration % args.tensorboard_log_interval == 0): - writer.add_scalar( - "steps-vs-samples/y=steps,x=samples", iteration, args.consumed_train_samples - ) - writer.add_scalar( - "steps-vs-samples/y=samples,x=steps", args.consumed_train_samples, iteration - ) - writer.add_scalar( - "steps-vs-tokens/y=steps,x=tokens", iteration, args.consumed_train_tokens - ) - writer.add_scalar( - "steps-vs-tokens/y=tokens,x=steps", args.consumed_train_tokens, iteration - ) + if iteration % args.tensorboard_log_interval == 0: if args.log_learning_rate_to_tensorboard: wandb_metrics |= { "learning-rate/iteration": iteration, "learning-rate/learning-rate": learning_rate, } - writer.add_scalar("learning-rate/learning-rate", learning_rate, iteration) - writer.add_scalar( - "learning-rate/learning-rate vs samples", - learning_rate, - args.consumed_train_samples, - ) - writer.add_scalar( - "learning-rate/learning-rate vs tokens", - learning_rate, - args.consumed_train_tokens, - ) - if args.log_batch_size_to_tensorboard: - writer.add_scalar("batch-size/batch-size", batch_size, iteration) - writer.add_scalar( - "batch-size/batch-size vs samples", - batch_size, - args.consumed_train_samples, - ) - writer.add_scalar( - "batch-size/batch-size vs tokens", - batch_size, - args.consumed_train_tokens, - ) wandb_metrics |= { "lm-loss-training/iteration": iteration, "lm-loss-training/consumed_train_tokens": args.consumed_train_tokens, + **{"lm-loss-training/" + k: v for k, v in loss_dict.items()}, } - for key in loss_dict: - wandb_metrics |= {f"lm-loss-training/{key}": loss_dict[key]} - writer.add_scalar(f"lm-loss-training/{key}", loss_dict[key], iteration) - writer.add_scalar( - f"lm-loss-training/{key}" + " vs samples", - loss_dict[key], - args.consumed_train_samples, - ) - writer.add_scalar( - f"lm-loss-training/{key}" + " vs tokens", - loss_dict[key], - args.consumed_train_tokens, - ) - if args.fp16 and loss_scale and args.log_loss_scale_to_tensorboard: - writer.add_scalar("loss-scale/loss-scale", loss_scale, iteration) - writer.add_scalar( - "loss-scale/loss-scale vs samples", - loss_scale, - args.consumed_train_samples, - ) - writer.add_scalar( - "loss-scale/loss-scale vs tokens", - loss_scale, - args.consumed_train_tokens, - ) - if args.log_world_size_to_tensorboard: - writer.add_scalar("world-size/world-size", args.world_size, iteration) - writer.add_scalar( - "world-size/world-size vs samples", - args.world_size, - args.consumed_train_samples, - ) - writer.add_scalar( - "world-size/world-size vs tokens", - args.world_size, - args.consumed_train_tokens, - ) if grad_norm is not None: wandb_metrics |= {"training/grad-norm": grad_norm} - writer.add_scalar("grad-norm/grad-norm", grad_norm, iteration) - writer.add_scalar( - "grad-norm/grad-norm vs samples", grad_norm, args.consumed_train_samples - ) - writer.add_scalar( - "grad-norm/grad-norm vs tokens", grad_norm, args.consumed_train_tokens - ) if num_zeros_in_grad is not None: wandb_metrics |= {"training/num-zeros": num_zeros_in_grad} - writer.add_scalar("num-zeros/num-zeros", num_zeros_in_grad, iteration) - writer.add_scalar( - "num-zeros/num-zeros vs samples", - num_zeros_in_grad, - args.consumed_train_samples, - ) - writer.add_scalar( - "num-zeros/num-zeros vs tokens", - num_zeros_in_grad, - args.consumed_train_tokens, - ) if params_norm is not None: wandb_metrics |= {"training/params-norm": params_norm} - writer.add_scalar("params-norm/params-norm", params_norm, iteration) - writer.add_scalar( - "params-norm/params-norm vs samples", - params_norm, - args.consumed_train_samples, - ) - writer.add_scalar( - "params-norm/params-norm vs tokens", - params_norm, - args.consumed_train_tokens, - ) - if hasattr(args, "actual_seq_length"): - writer.add_scalar( - "seqlen/actual_seq_length", args.actual_seq_length, iteration - ) - writer.add_scalar( - "seqlen/actual_seq_length vs samples", - args.actual_seq_length, - args.consumed_train_samples, - ) - writer.add_scalar( - "seqlen/actual_seq_length vs tokens", - args.actual_seq_length, - args.consumed_train_tokens, - ) - if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: - writer.add_scalar( - "seqlen/curriculum_seqlen", args.curriculum_seqlen, iteration - ) - writer.add_scalar( - "seqlen/curriculum_seqlen vs samples", - args.curriculum_seqlen, - args.consumed_train_samples, - ) - writer.add_scalar( - "seqlen/curriculum_seqlen vs tokens", - args.curriculum_seqlen, - args.consumed_train_tokens, - ) - if args.random_ltd: - writer.add_scalar( - "seqlen/random_ltd_reserved_length", - args.random_ltd_reserved_length, - iteration, - ) - writer.add_scalar( - "seqlen/random_ltd_reserved_length vs samples", - args.random_ltd_reserved_length, - args.consumed_train_samples, - ) - writer.add_scalar( - "seqlen/random_ltd_reserved_length vs tokens", - args.random_ltd_reserved_length, - args.consumed_train_tokens, - ) - if args.log_memory_to_tensorboard: - mem_stats = torch.cuda.memory_stats() - writer.add_scalar( - "mem-reserved-bytes", - mem_stats["reserved_bytes.all.current"], - iteration, - ) - writer.add_scalar( - "mem-allocated-bytes", - mem_stats["allocated_bytes.all.current"], - iteration, - ) - writer.add_scalar( - "mem-allocated-count", - mem_stats["allocation.all.current"], - iteration, - ) - if iteration % args.tensorboard_log_interval == 0: # This logging write various optimizer states to tensorboard. This # feature may consume extra GPU memory thus is set at false by default. if args.log_optimizer_states_to_tensorboard and optimizer is not None: @@ -424,92 +258,6 @@ def training_log( "optimizer/weight_abs_max": opt_stats_2[3], } # print('step {} rank {} after sync opt_stats {}, {}'.format(iteration, torch.distributed.get_rank(), opt_stats_2, opt_stats)) - if writer and is_last_rank(): - writer.add_scalar( - "optimizer/variance_l2 vs tokens", - opt_stats[0] ** 0.5, - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_sqrt_l2 vs tokens", - opt_stats[1] ** 0.5, - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/momentum_l2 vs tokens", - opt_stats[2] ** 0.5, - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/weight_l2 vs tokens", - opt_stats[3] ** 0.5, - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_l1 vs tokens", - opt_stats[4], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_sqrt_l1 vs tokens", - opt_stats[5], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/momentum_l1 vs tokens", - opt_stats[6], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/weight_l1 vs tokens", - opt_stats[7], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_abs_max vs tokens", - opt_stats_2[0], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_sqrt_abs_max vs tokens", - opt_stats_2[1], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/momentum_abs_max vs tokens", - opt_stats_2[2], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/weight_abs_max vs tokens", - opt_stats_2[3], - args.consumed_train_tokens, - ) - writer.add_scalar( - "optimizer/variance_l2", opt_stats[0] ** 0.5, iteration - ) - writer.add_scalar( - "optimizer/variance_sqrt_l2", opt_stats[1] ** 0.5, iteration - ) - writer.add_scalar( - "optimizer/momentum_l2", opt_stats[2] ** 0.5, iteration - ) - writer.add_scalar("optimizer/weight_l2", opt_stats[3] ** 0.5, iteration) - writer.add_scalar("optimizer/variance_l1", opt_stats[4], iteration) - writer.add_scalar("optimizer/variance_sqrt_l1", opt_stats[5], iteration) - writer.add_scalar("optimizer/momentum_l1", opt_stats[6], iteration) - writer.add_scalar("optimizer/weight_l1", opt_stats[7], iteration) - writer.add_scalar( - "optimizer/variance_abs_max", opt_stats_2[0], iteration - ) - writer.add_scalar( - "optimizer/variance_sqrt_abs_max", opt_stats_2[1], iteration - ) - writer.add_scalar( - "optimizer/momentum_abs_max", opt_stats_2[2], iteration - ) - writer.add_scalar("optimizer/weight_abs_max", opt_stats_2[3], iteration) - assert args is not None assert timers is not None if iteration % args.log_interval == 0: @@ -536,6 +284,7 @@ def training_log( tflops_lm = num_flop_per_sec_lm / (10**12) tflops_lm_per_gpu = tflops_lm / args.world_size wandb_metrics |= { + "throughput/iteration": iteration, "throughput/iteration-time": elapsed_time_per_iteration, # 1000 ms / s "throughput/samples_per_sec": samples_per_sec, "throughput/samples_per_sec_per_replica": samples_per_sec_per_replica, @@ -548,27 +297,12 @@ def training_log( "throughput/tflops-lm": tflops_lm_per_gpu, "throughput/approx_params_in_billions": approx_parameters_in_billions, "throughput/elapsed_ms_per_iteration": elapsed_time_per_iteration, - "throughput/iteration": iteration, } if loss_dict is not None: wandb_metrics |= { "loss/iteration": iteration, **{f"loss/{k}": v for k, v in loss_dict.items()}, } - if writer and args.log_timers_to_tensorboard: - writer.add_scalar( - "iteration-time/iteration-time", elapsed_time_per_iteration, iteration - ) - writer.add_scalar( - "iteration-time/iteration-time vs samples", - elapsed_time_per_iteration, - args.consumed_train_samples, - ) - writer.add_scalar( - "iteration-time/iteration-time vs tokens", - elapsed_time_per_iteration, - args.consumed_train_tokens, - ) # metrics_to_log = { # 'iteration': iteration, # 'train_iters': args.train_iters, @@ -646,8 +380,9 @@ def training_log( log_string += " tokens_per_gpu_per_second_tgs={:.3f} |".format( tokens_per_gpu_per_second ) - log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) - log_string += " [DS]TFLOPs={:.2f} |".format(tflops) + log_string += " TFLOPs={:.2f} |".format(tflops_lm_per_gpu) + # log_string += " [LM]TFLOPs={:.2f} |".format(tflops_lm_per_gpu) + # log_string += " [DS]TFLOPs={:.2f} |".format(tflops) if wandb is not None and getattr(wandb, "run", None) is not None: wandb_metrics |= { "training/skiped_iterations": total_loss_dict[skipped_iters_key] From 3e174cde8c95808ae7b40ab773c6347553c030a5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 16 Jun 2025 08:13:15 -0500 Subject: [PATCH 425/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 440 ++++++++++++++++++++++++++++-------------------- 1 file changed, 253 insertions(+), 187 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index fd01f24ce2c..f033a95099e 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -20,11 +20,10 @@ # command for launching across all GPUs in our active PBS job. ############################################################################### - ############################################################################### # Source: # [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -source <(curl -L https://bit.ly/ezpz-utils) || exit +source <(curl -L https://bit.ly/ezpz-utils) > /dev/null || exit ezpz_setup_job >/dev/null || exit ############################################################################### @@ -73,41 +72,30 @@ helpers_main() { # # - Explicitly, this will: # - Identify the machine we're on -# # - Setup `python` # 1. Load `conda` # 2. Setup `venv` on top of `conda` -# # - Ensure all dependencies are installed -# # - Clone + Install [`saforem2/ezpz`](https://github.com/saforem2/ezpz) # - Source [`ezpz/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) # - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) -# # - Set runtime options -# # - Build `deepspeed_config.json` -# # - Build {logs, checkpoints, etc} dirs, named according to specifics of # current run -# # - Specify additional `deepspeed` arguments -# # - Ensure executable exists at expected path -# # - Setup data + tokenizer via `TOKENIZER_TYPE` -# # - Print job info -# # - Save `.env` to `CKPT_DIR` for safe keeping -# # - Check that we're not already running, and if so, exit. -# # - Setup run command to be executed. ############################################################################## setup() { # Identify machine we're on - get_machine || exit + mn=$(ezpz_get_machine_name) + export MACHINE="${mn}" + # get_machine || exit ########################################################################## # # ezpz_setup will: # # 1. Setup python @@ -134,7 +122,7 @@ setup() { set_args || exit # Ensure executable exists in expected path check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(get_machine_name)/dolma.txt"}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit @@ -149,8 +137,38 @@ setup() { setup_run_cmd "$@" || exit } - -get_training_args() { +##################################################### +# setup_run_cmd +# +# Build run command to be executed. +##################################################### +setup_run_cmd() { + ############################## + # take in additional arguments + # and append them directly to + # the end of the `run_cmd` + # custom_args="$@" + custom_args=("$@") + ############################## + #### Make it easy to track experiments by date ################### + year="$(date "+%Y")" + month="$(date "+%m")" + day="$(date "+%Y-%m-%d")" + today="$(date "+%Y-%m-%d")" # kept for backwards compatibility + started_at="$(date "+%Y-%m-%d-%H%M%S")" + export YEAR="${year}" + export MONTH="${month}" + export DAY="${day}" + export TODAY="${today}" + export STARTED_AT="${started_at}" + ################################################################## + # NOTE: to launch with DeepSpeed instead of mpiexec: + # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` + ################################################################## + setupLauncher "${LAUNCH_WITH:-MPICH}" || exit + export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" + printf "\n" + echo "Using data_cache_path: ${data_cache_path}" ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` @@ -194,7 +212,7 @@ get_training_args() { "--tensorboard-dir ${TBDIR}" ) fi - dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(get_machine_name)/dolma.txt}" + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt}" train_args=() if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then @@ -250,51 +268,42 @@ get_training_args() { "--data-cache-path=${data_cache_path}" "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" ) + declare -A arch_map + printf "==== ARCHITECTURE ====\n" + arch_map=( + ["TP"]="${TP}" + ["PP"]="${PP}" + ["SP"]="${SP}" + ["DP"]="${DP}" + ["MBS"]="${MICRO_BATCH}" + ["GAS"]="${GRAD_ACC_STEPS}" + ["GBS"]="${GLOBAL_BATCH}" + ["NLAYERS"]="${NLAYERS}" + ["HIDDEN"]="${HIDDEN}" + ["HEADS"]="${HEADS}" + ["SEQ"]="${SEQ}" + ["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE}" + ["NUM_KV_HEAD"]="${NUM_KV_HEAD}" + ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" + # ["DTYPE"]="${DTYPE}" + # ["OPT"]="${OPT}" + # ["WEIGHT_DECAY"]="${WEIGHT_DECAY}" + # ["ADAM_BETA1"]="${ADAM_BETA1}" + # ["ADAM_BETA2"]="${ADAM_BETA2}" + # ["ADAM_EPS"]="${ADAM_EPS}" + # ["CLIP_GRAD"]="${CLIP_GRAD:-1.0}" + ) + for v in "${!arch_map[@]}"; do + printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" + done + printf "======================\n" # "--adam-eps ${ADAM_EPS:-0.00001}" - echo "${train_args[@]}" -} - -##################################################### -# setup_run_cmd -# -# Build run command to be executed. -##################################################### -setup_run_cmd() { - ############################## - # take in additional arguments - # and append them directly to - # the end of the `run_cmd` - # custom_args="$@" - custom_args=("$@") - ############################## - #### Make it easy to track experiments by date ################### - year="$(date "+%Y")" - month="$(date "+%m")" - day="$(date "+%Y-%m-%d")" - today="$(date "+%Y-%m-%d")" # kept for backwards compatibility - started_at="$(date "+%Y-%m-%d-%H%M%S")" - export YEAR="${year}" - export MONTH="${month}" - export DAY="${day}" - export TODAY="${today}" - export STARTED_AT="${started_at}" - ################################################################## - # NOTE: to launch with DeepSpeed instead of mpiexec: - # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` - ################################################################## - setupLauncher "${LAUNCH_WITH:-MPICH}" || exit - export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" - train_args=("$(get_training_args)") - cache_dir="${WORKING_DIR}/.cache/" - if [[ ! -d "${cache_dir}" ]]; then - mkdir -p "${cache_dir}" - fi - targs_cache="${cache_dir}/train_args_${started_at}.txt" - for arg in "${train_args[@]}"; do echo "${arg}" >"${targs_cache}"; done + cache_dir="${PBS_O_WORKDIR}/.cache/" + mkdir -p "${cache_dir}" + targs_cache="${cache_dir}/train_args.txt" + for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") - # printf "\n" - echo "Using data_cache_path: ${data_cache_path}" - printf "Training Arguments: %s\n" "${TRAIN_ARGS[@]}" + printf "Training Arguments: %s\n" "$(printBlue "${TRAIN_ARGS[@]}")" export run_cmd=("${LAUNCHER}" "${train_args[@]}") } @@ -313,60 +322,60 @@ save_dotenv() { fi } -###################################################################### -# get_machine_name: -# -# Return current machine name, as lowercase string +####################################################################### +## get_machine_name: +## +## Return current machine name, as lowercase string +## +## Example: +## ```bash +## $ machine_name=$(get_machine_name) +## $ echo "machine_name: ${machine_name}" +## ``` +####################################################################### +#get_machine_name() { +# if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then +# machine="aurora" +# elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then +# machine="sunspot" +# elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then +# if [[ "${PBS_O_HOST}" == sirius* ]]; then +# machine="sirius" +# else +# machine="polaris" +# fi +# elif [[ $(hostname) == sophia* ]]; then +# machine="sophia" +# elif [[ $(hostname) == nid* ]]; then +# machine="perlmutter" +# else +# machine=$(hostname) +# fi +# echo "${machine}" +#} # -# Example: -# ```bash -# $ machine_name=$(get_machine_name) -# $ echo "machine_name: ${machine_name}" -# ``` -###################################################################### -get_machine_name() { - if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then - machine="aurora" - elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then - machine="sunspot" - elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then - machine="sirius" - else - machine="polaris" - fi - elif [[ $(hostname) == sophia* ]]; then - machine="sophia" - elif [[ $(hostname) == nid* ]]; then - machine="perlmutter" - else - machine=$(hostname) - fi - echo "${machine}" -} - -get_machine() { - machine=$(hostname) - if [[ $(hostname) == x4* ]]; then - machine="aurora" - elif [[ $(hostname) == x1* ]]; then - machine="sunspot" - elif [[ $(hostname) == x3* ]]; then - if [[ "${PBS_O_HOST}" == sirius* ]]; then - machine="sirius" - else - machine="polaris" - fi - elif [[ $(hostname) == sophia* ]]; then - machine="sophia" - elif [[ $(hostname) == nid* ]]; then - machine="perlmutter" - else - echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." - fi - export MACHINE="${machine}" - printf "Running on: %s\n" "$(printBlue "${MACHINE}")" -} +#get_machine() { +# machine=$(hostname) +# if [[ $(hostname) == x4* ]]; then +# machine="aurora" +# elif [[ $(hostname) == x1* ]]; then +# machine="sunspot" +# elif [[ $(hostname) == x3* ]]; then +# if [[ "${PBS_O_HOST}" == sirius* ]]; then +# machine="sirius" +# else +# machine="polaris" +# fi +# elif [[ $(hostname) == sophia* ]]; then +# machine="sophia" +# elif [[ $(hostname) == nid* ]]; then +# machine="perlmutter" +# else +# echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." +# fi +# export MACHINE="${machine}" +# printf "Running on: %s\n" "$(printBlue "${MACHINE}")" +#} check_and_kill_if_running() { RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') @@ -413,6 +422,10 @@ printJobInfo() { # will launch with `deepspeed` instead of `mpiexec`. ############################################################################# setupLauncher() { + shell_type=$(basename "${SHELL}") + if [[ "${shell_type}" == "bash" ]]; then + shopt -s expand_aliases + fi if [[ "$#" == 1 ]]; then local dist_launcher="$1" else @@ -426,21 +439,26 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - if [[ -n "${DIST_LAUNCH}" ]]; then - mn=$(get_machine_name) - if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then - LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" - elif [[ "${mn}" == "sophia" ]]; then - LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" - else - LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" - fi - export LAUNCHER="${LAUNCHER}" - else - echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" - printenv | tee ".env-${PBS_JOBID}" - echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" - fi + LAUNCHER="ezpz-launch ${EXEC}" + # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job + # echo "ezpz_launch: $(which ezpz_launch)" + # export -f ezpz_launch + # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" + # if [[ -n "${DIST_LAUNCH}" ]]; then + # mn=$(ezpz_get_machine_name) + # if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then + # LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" + # elif [[ "${mn}" == "sophia" ]]; then + # LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" + # else + # LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" + # fi + # export LAUNCHER="${LAUNCHER}" + # else + # echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" + # printenv | tee ".env-${PBS_JOBID}" + # echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" + # fi fi printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" printf " %s" "$(printMagenta "${LAUNCHER}")" @@ -564,7 +582,8 @@ set_ccl_vars_on_aurora() { export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich export FI_PROVIDER=cxi # Required by Aurora mpich export PALS_PMI=pmix # Required by Aurora mpich - export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + # export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + export CCL_ATL_TRANSPORT=ofi # [SF]: Changed 04/30/2025 export TORCH_LLM_ALLREDUCE=1 export CCL_SYCL_ESIMD=1 export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) @@ -576,6 +595,59 @@ set_ccl_vars_on_aurora() { export CCL_BCAST=double_tree } + +get_model_arch_7B() { + # 7B + # export MODEL_KEY="AuroraGPT-7B" + export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 +} + +# get_model_arch_70B() { +# # 70B +# export MODEL_KEY="AuroraGPT-70B" +# export TP=6 +# export ZERO_STAGE=2 +# export GRAD_ACC_STEPS=1 +# export USE_ACTIVATION_CHECKPOINTING=1 +# export MICRO_BATCH=1 +# export NLAYERS=80 +# export HIDDEN=8192 +# export HEADS=96 +# export FFN_HIDDEN_SIZE=11040 +# export SEQ=8192 +# export NUM_KV_HEAD=6 +# } +# +get_model_arch_70B() { + HEADS=64 + NLAYERS=80 + NUM_KV_HEAD=8 + FFN_HIDDEN_SIZE=28672 + HIDDEN=8192 + SEQ_LEN=8192 +} + +get_model_arch_33B() { + # 33B + export MODEL_KEY="AuroraGPT-33B" + export TP=6 + export ZERO_STAGE=2 + export GRAD_ACC_STEPS=1 + # export USE_ACTIVATION_CHECKPOINTING=1 + export MICRO_BATCH=1 + export NLAYERS=78 + export HIDDEN=8268 + export HEADS=78 + export FFN_HIDDEN_SIZE=11076 + export SEQ=4096 + export NUM_KV_HEAD=6 +} + ############################################################################## # setParams # @@ -592,7 +664,8 @@ setParams() { # ---- [Parallelism Settings] -------------------------------------------+ # ------ [Aurora] -------||------ [SunSpot] ------------- # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then - mn=$(get_machine_name) + GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + mn=$(ezpz_get_machine_name) if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then TP=${TP:-1} # TP = 1 export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" @@ -632,9 +705,9 @@ setParams() { # ---- [Sophia] ---------------------- elif [[ "${mn}" == sophia* ]]; then # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" @@ -658,6 +731,13 @@ setParams() { FLASH_ARG="--use-flash-attn-v2" fi fi + # model_arch="${MODEL_ARCH:-7B}" + # printf "Using model architecture: %s\n" "$(printGreen "${model_arch}")" + if [[ "${MODEL_ARCH:-}" == "70B" ]]; then + get_model_arch_70B + else + get_model_arch_7B + fi export TP="${TP}" export PP="${PP:-1}" export SP="${SP:-1}" @@ -672,23 +752,12 @@ setParams() { fi export NGPU_PER_HOST="${NGPU_PER_HOST}" export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" - #### [sam: 05/05/2025] ################################################# - # NOTE: Commented out following release of frameworks-2025.0.0 - # if [[ "${WORLD_SIZE}" -gt 12 && "${mn}" == "aurora" ]]; then - # #### [sam: 08/17/2024] ########################################### + # if [[ "${WORLD_SIZE}" -gt 1 && "${mn}" == "aurora" ]]; then + # #### [sam: 08/17/2024] ########################################## # # Use best set of CCL env vars from Gordon Bell runs on Aurora # set_ccl_vars_on_aurora # fi - ######################################################################## - # +---[Llama2 7B Config]--------------------------------------------------+ - # export MODEL_KEY="Llama-7B" - export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS - export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS - export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE - export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION - export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS @@ -696,7 +765,9 @@ setParams() { export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE + export DP=$((WORLD_SIZE / TP / PP / SP)) # DATA PARALLELISM export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT + echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" if [[ -n "${TRAIN_TOKENS:-}" ]]; then export TRAIN_TOKENS="${TRAIN_TOKENS}" export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) @@ -970,7 +1041,7 @@ install_dependencies() { echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." python3 -m pip install -r "${depsfile}" --require-virtualenv 1>/dev/null if [[ ! -x "$(command -v deepspeed)" ]]; then - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then # install_deepspeed_for_xpu || exit # fi @@ -1069,7 +1140,7 @@ setup_tokenizer_and_data() { if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then export TOKENIZER_TYPE="GPT2" _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") - machine=$(get_machine_name) + machine=$(ezpz_get_machine_name) if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" elif [[ ${machine} == "sunspot" ]]; then @@ -1113,7 +1184,7 @@ setup_tokenizer_and_data() { ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# - mn=$(get_machine_name) + mn=$(ezpz_get_machine_name) dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" ############################################################################ # set `dfl` to `dfl_fallback` if not passed as an argument, @@ -1219,8 +1290,7 @@ generateDSconfig() { \"min_loss_scale\": 1 }, \"bfloat16\": { - \"enabled\": true, - \"loss_scale\": 1.0 + \"enabled\": true }," elif [[ $DTYPE == "fp16" ]]; then dtype="\ @@ -1273,11 +1343,32 @@ generateDSconfig() { fi if [[ "${ZERO_STAGE}" == 3 ]]; then # \"mics_shard_size\": 2, + # hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-${NGPU_PER_HOST:-1}} + # zero="\ + # \"zero_optimization\": { + # \"stage\": 3, + # \"reduce_scatter\": true, + # \"reduce_bucket_size\": 10000000, + # \"reduce_scatter\": true, + # \"zero_quantized_weights\": true, + # \"zero_hpz_partition_size\": $hpz_partition_size, + # \"zero_quantized_gradients\": true, + # \"contiguous_gradients\": true, + # \"overlap_comm\": true, + # } + # }," + # mics_shard_size="${MICS_SHARD_SIZE:-${NGPU_PER_HOST}}" + # \"mics_shard_size\": $mics_shard_size, + hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-${NGPU_PER_HOST:-1}} zero="\ \"zero_optimization\": { \"stage\": 3, \"reduce_scatter\": false, - \"mics_hierarchical_params_gather\": true, + \"zero_quantized_weights\": false, + \"zero_hpz_partition_size\": $hpz_partition_size, + \"zero_quantized_gradients\": false, + \"mics_shard_size\": 1, + \"mics_hierarchical_params_gather\": false, \"stage3_max_live_parameters\": 3e9, \"stage3_max_reuse_distance\": 3e9, \"stage3_param_persistence_threshold\": 1e5, @@ -1354,41 +1445,16 @@ $flops_profiler EOT } -# ##################### -# # train -# ##################### -# train() { -# # 1. Navigate into `$PBS_O_WORKDIR` <-- [should be Megatron-Deepspeed] -# cd "${PBS_O_WORKDIR}" || exit -# HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE -# # 2. source `ALCF/helpers.sh` <-- [should be ./ALCF/helpers.sh] -# source "${HERE}/ALCF/helpers.sh" || exit -# # 3. call `setup` from `./ALCF/helpers.sh` -# # export DATA_FILE_LIST="${HERE}/ALCF/data-lists/$(get_machine_name)/books.txt" -# setup || exit -# # 4. Take custom args -# export custom_args=" $@" -# # 5. Update ${run_cmd} (from setup ALCF/helpers.sh) with ${custom_args} -# export run_cmd="${run_cmd} ${custom_args}" -# # 6. Add "${run_cmd}" to output log -# echo "${run_cmd}" | tee -a "${OUTPUT_LOG}" -# # 7. Tell user where to find output -# printf "[!! %s] View output at:\n %s\n" "$(printBlue "NOTE")" "$(printYellow "${OUTPUT_LOG}")" | tee -a "${OUTPUT_LOG}" -# # 8. Evaluate ${run_cmd} and append outputs to ${OUTPUT_LOG} -# eval "${run_cmd}" |& tee -a "${OUTPUT_LOG}" -# set +x -# } - ############################################### # Helper functions for printing colored text ############################################### -# RESET="\e[0m" -# BLACK="\e[1;30m" -# RED="\e[1;31m" -# GREEN="\e[1;32m" -# YELLOW="\e[1;33m" -# BLUE="\e[1;34m" -# CYAN="\e[1;35m" +RESET="\e[0m" +BLACK="\e[1;30m" +RED="\e[1;31m" +GREEN="\e[1;32m" +YELLOW="\e[1;33m" +BLUE="\e[1;34m" +CYAN="\e[1;35m" # WHITE="\e[1;36m" printBlack() { From 0ddd003b00e4b419fe6f8a2d4428c527530148f8 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Jun 2025 07:45:35 -0500 Subject: [PATCH 426/590] feat: Remove `--use-mics` flag wen using `ZeRO` 3 --- ALCF/helpers.sh | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index f033a95099e..660b79427a5 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -821,9 +821,11 @@ set_args() { fi ds_args+=("--deepspeed_config=${DS_CONFIG}") ds_args+=("--zero-stage=$ZERO_STAGE") - if [[ "${ZERO_STAGE}" == 3 ]]; then - ds_args+=("--use-mics") - fi + + # if [[ "${ZERO_STAGE}" == 3 ]]; then + # ds_args+=("--use-mics") + # fi + # ds_args=" " # ds_args=" --deepspeed ${ds_args}" # if [[ $PP == 1 ]]; then @@ -1359,16 +1361,18 @@ generateDSconfig() { # }," # mics_shard_size="${MICS_SHARD_SIZE:-${NGPU_PER_HOST}}" # \"mics_shard_size\": $mics_shard_size, - hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-${NGPU_PER_HOST:-1}} + hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-1} + + # \"zero_quantized_weights\": false, + # \"zero_hpz_partition_size\": $hpz_partition_size, + # \"zero_quantized_gradients\": false, + # \"mics_shard_size\": 1, + # \"mics_hierarchical_params_gather\": false, zero="\ \"zero_optimization\": { \"stage\": 3, - \"reduce_scatter\": false, - \"zero_quantized_weights\": false, \"zero_hpz_partition_size\": $hpz_partition_size, - \"zero_quantized_gradients\": false, - \"mics_shard_size\": 1, - \"mics_hierarchical_params_gather\": false, + \"reduce_scatter\": false, \"stage3_max_live_parameters\": 3e9, \"stage3_max_reuse_distance\": 3e9, \"stage3_param_persistence_threshold\": 1e5, From 43794c14a5d634d1537e96f812513cc2dfc8d686 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Jun 2025 13:40:05 -0500 Subject: [PATCH 427/590] docs: Update `ALCF/notes/AuroraGPT-70B.md` --- ALCF/notes/AuroraGPT-70B.md | 76 +++++++++++++++++++++++++++---------- 1 file changed, 55 insertions(+), 21 deletions(-) diff --git a/ALCF/notes/AuroraGPT-70B.md b/ALCF/notes/AuroraGPT-70B.md index f2dfc46ba70..0f69a4db031 100644 --- a/ALCF/notes/AuroraGPT-70B.md +++ b/ALCF/notes/AuroraGPT-70B.md @@ -24,28 +24,62 @@ PP="${NHOSTS}" \ bash train_alcf.sh ``` -| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | -| :----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :----------------------------------------------------------------------------- | -| 8 | 96 | 8 | 12 | 1 | 1 | 1 | 12 | True | OOM | OOM | OOM | [chocolate-meadow-2125](https://wandb.ai/aurora_gpt/AuroraGPT/runs/uhfkwmp2) | -| 8 | 96 | 8 | 6 | 2 | 1 | 1 | 6 | True | 10.26 | 0.541 | 23.084 | [glamorous-darkness-2126](https://wandb.ai/aurora_gpt/AuroraGPT/runs/tdehvqey) | -| 8 | 96 | 8 | 6 | 2 | 1 | 1 | 6 | False | 13.16 | 0.694 | 29.608 | [proud-frog-2153](https://wandb.ai/aurora_gpt/AuroraGPT/runs/ho1lwmer) | -| 8 | 96 | 8 | 6 | 2 | 1 | 2 | 12 | True | 17.75 | 0.936 | 39.927 | [stoic-dragon-2127](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2vay044x) | -| 8 | 96 | 8 | 6 | 2 | 1 | 4 | 24 | True | 28.93 | 1.525 | 65.081 | [drawn-wildflower-2128](https://wandb.ai/aurora_gpt/AuroraGPT/runs/qjaqpbxg) | -| 8 | 96 | 8 | 6 | 2 | 1 | 8 | 48 | True | 42.27 | 2.229 | 95.103 | [fresh-waterfall-2129](https://wandb.ai/aurora_gpt/AuroraGPT/runs/8m38fq95) | -| 8 | 96 | 8 | 6 | 2 | 1 | 16 | 96 | True | 54.66 | 2.882 | 122.974 | [divine-waterfall-2130](https://wandb.ai/aurora_gpt/AuroraGPT/runs/on3m4isp) | -| 8 | 96 | 8 | 6 | 2 | 1 | 32 | 192 | True | 64.02 | 3.376 | 144.037 | [wild-bee-2131](https://wandb.ai/aurora_gpt/AuroraGPT/runs/hitrbi6d) | -|   | | | | | | | | | | | | | -| 16 | 192 | 16 | 12 | 1 | 1 | 1 | 12 | | 5.6 | 0.594 | 12.695 | [fluent-surf-2118](https://wandb.ai/aurora_gpt/AuroraGPT/runs/0y250j0i) | -| 16 | 192 | 16 | 12 | 1 | 1 | 2 | 24 | | 10.25 | 1.081 | 23.064 | [dulcet-salad-2119](https://wandb.ai/aurora_gpt/AuroraGPT/runs/5f1rdn9p) | -| 16 | 192 | 16 | 12 | 1 | 1 | 4 | 48 | | 18.10 | 1.908 | 40.713 | [rose-blaze-2120](https://wandb.ai/aurora_gpt/AuroraGPT/runs/9obt1iqi) | -| 16 | 192 | 16 | 12 | 1 | 1 | 8 | 96 | | 29.49 | 3.110 | 66.342 | [azure-jazz-2121](https://wandb.ai/aurora_gpt/AuroraGPT/runs/oaft4n5p) | -| 16 | 192 | 16 | 12 | 1 | 1 | 16 | 192 | | OOM | OOM | OOM | OOM | -|   | | | | | | | | | | | | | - - **Note**: - `sps`: samples per second - `tpgps`: tokens per GPU per second + +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | +| :-------: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :----------------------------------------------------------------------------- | +| 8[^2125] | 96 | 8 | 12 | 1 | 1 | 1 | 12 | True | OOM | OOM | OOM | [chocolate-meadow-2125](https://wandb.ai/aurora_gpt/AuroraGPT/runs/uhfkwmp2) | +| 8[^2126] | 96 | 8 | 6 | 2 | 1 | 1 | 6 | True | 10.26 | 0.541 | 23.084 | [glamorous-darkness-2126](https://wandb.ai/aurora_gpt/AuroraGPT/runs/tdehvqey) | +| 8[^2153] | 96 | 8 | 6 | 2 | 1 | 1 | 6 | False | 13.16 | 0.694 | 29.608 | [proud-frog-2153](https://wandb.ai/aurora_gpt/AuroraGPT/runs/ho1lwmer) | +| 8[^2127] | 96 | 8 | 6 | 2 | 1 | 2 | 12 | True | 17.75 | 0.936 | 39.927 | [stoic-dragon-2127](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2vay044x) | +| 8[^2128] | 96 | 8 | 6 | 2 | 1 | 4 | 24 | True | 28.93 | 1.525 | 65.081 | [drawn-wildflower-2128](https://wandb.ai/aurora_gpt/AuroraGPT/runs/qjaqpbxg) | +| 8[^2129] | 96 | 8 | 6 | 2 | 1 | 8 | 48 | True | 42.27 | 2.229 | 95.103 | [fresh-waterfall-2129](https://wandb.ai/aurora_gpt/AuroraGPT/runs/8m38fq95) | +| 8[^2130] | 96 | 8 | 6 | 2 | 1 | 16 | 96 | True | 54.66 | 2.882 | 122.974 | [divine-waterfall-2130](https://wandb.ai/aurora_gpt/AuroraGPT/runs/on3m4isp) | +| 8[^2131] | 96 | 8 | 6 | 2 | 1 | 32 | 192 | True | 64.02 | 3.376 | 144.037 | [wild-bee-2131](https://wandb.ai/aurora_gpt/AuroraGPT/runs/hitrbi6d) | +|   | | | | | | | | | | | | | +| 16[^2118] | 192 | 16 | 12 | 1 | 1 | 1 | 12 | True | 5.6 | 0.594 | 12.695 | [fluent-surf-2118](https://wandb.ai/aurora_gpt/AuroraGPT/runs/0y250j0i) | +| 16[^2119] | 192 | 16 | 12 | 1 | 1 | 2 | 24 | True | 10.25 | 1.081 | 23.064 | [dulcet-salad-2119](https://wandb.ai/aurora_gpt/AuroraGPT/runs/5f1rdn9p) | +| 16[^2120] | 192 | 16 | 12 | 1 | 1 | 4 | 48 | True | 18.10 | 1.908 | 40.713 | [rose-blaze-2120](https://wandb.ai/aurora_gpt/AuroraGPT/runs/9obt1iqi) | +| 16[^2121] | 192 | 16 | 12 | 1 | 1 | 8 | 96 | True | 29.49 | 3.110 | 66.342 | [azure-jazz-2121](https://wandb.ai/aurora_gpt/AuroraGPT/runs/oaft4n5p) | +| 16 | 192 | 16 | 12 | 1 | 1 | 16 | 192 | True | OOM | OOM | OOM | OOM | + + +[^2125]: [chocolate-meadow-2125](https://wandb.ai/aurora_gpt/AuroraGPT/runs/uhfkwmp2) +[^2126]: [glamorous-darkness-2126](https://wandb.ai/aurora_gpt/AuroraGPT/runs/tdehvqey) +[^2153]: [proud-frog-2153](https://wandb.ai/aurora_gpt/AuroraGPT/runs/ho1lwmer) +[^2127]: [stoic-dragon-2127](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2vay044x) +[^2128]: [drawn-wildflower-2128](https://wandb.ai/aurora_gpt/AuroraGPT/runs/qjaqpbxg) +[^2129]: [fresh-waterfall-2129](https://wandb.ai/aurora_gpt/AuroraGPT/runs/8m38fq95) +[^2130]: [divine-waterfall-2130](https://wandb.ai/aurora_gpt/AuroraGPT/runs/on3m4isp) +[^2131]: [wild-bee-2131](https://wandb.ai/aurora_gpt/AuroraGPT/runs/hitrbi6d) +[^2118]: [fluent-surf-2118](https://wandb.ai/aurora_gpt/AuroraGPT/runs/0y250j0i) +[^2119]: [dulcet-salad-2119](https://wandb.ai/aurora_gpt/AuroraGPT/runs/5f1rdn9p) +[^2120]: [rose-blaze-2120](https://wandb.ai/aurora_gpt/AuroraGPT/runs/9obt1iqi) +[^2121]: [azure-jazz-2121](https://wandb.ai/aurora_gpt/AuroraGPT/runs/oaft4n5p) + + +## ZeRO-3 + HPZ + +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | +| :----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :------------------------------------------------------------------------- | +| 4 | 48 | 1 | 1 | 1 | 1 | 1 | 48 | True | ??? | ??? | ??? | ??? | +|   | | | | | | | | | | | | | +| 6 | 72 | 1 | 1 | 1 | 1 | 1 | 72 | True | 39.05 | 1.544 | 87.849 | [charmed-fire-2203](https://wandb.ai/aurora_gpt/AuroraGPT/runs/actny2dl) | +|   | | | | | | | | | | | | | +| 8 | 96 | 1 | 1 | 1 | 1 | 1 | 96 | True | 47.60 | 2.510 | 107.087 | [glamorous-wood-2199](https://wandb.ai/aurora_gpt/AuroraGPT/runs/bgu31497) | +| 8 | 96 | 1 | 1 | 1 | 1 | 2 | 192 | True | 55.69 | 2.936 | 125.287 | [solar-plant-2200](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gnc48o99) | +| 8 | 96 | 1 | 1 | 1 | 1 | 4 | 384 | True | 59.66 | 3.146 | 134.216 | [smooth-eon-2207](https://wandb.ai/aurora_gpt/AuroraGPT/runs/awt6d825) | +| 8 | 96 | 1 | 1 | 1 | 1 | 8 | 768 | True | 62.19 | 3.279 | 139.919 | [glamorous-eon-2209](https://wandb.ai/aurora_gpt/AuroraGPT/runs/jmukcjdc) | +|   | | | | | | | | | | | | | +| 16 | 192 | 1 | 1 | 1 | 1 | 1 | 192 | True | 57.6 | 6.074 | 129.577 | [gallant-field-2208](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gb94ahi3) | +| 16 | 384 | 1 | 1 | 1 | 1 | 2 | 384 | True | 61.86 | 6.523 | 139.161 | [swept-shadow-2210](https://wandb.ai/aurora_gpt/AuroraGPT/runs/riy7y3k1) | + + +[^z3h]: Seems to get hung on 4 nodes (waited ~30 min) + ## 42B Model (Pure Tensor Parallelism) (TP=6) - 42B param model @@ -108,6 +142,6 @@ MBS: 1 | Nodes | DP | TP | MBS | GBS | TFLOPS | sample/s | | :---: | :-: | :-: | :-: | :-: | :----: | :------: | -| 4 | 48 | 1 | 1 | 48 | 96.8 | 0.94 | -| 8 | 96 | 1 | 1 | 96 | 99.1 | 1.92 | -| 16 | 192 | 1 | 1 | 192 | 97.2 | 3.77 | +| 4 | 48 | 1 | 1 | 48 | 96.8 | 0.94 | +| 8 | 96 | 1 | 1 | 96 | 99.1 | 1.92 | +| 16 | 192 | 1 | 1 | 192 | 97.2 | 3.77 | From 351cfebfcbd29bd76337c494b484866fb584ada0 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Jun 2025 15:53:51 -0500 Subject: [PATCH 428/590] docs: Update `ALCF/notes/AuroraGPT-70B.md` --- ALCF/notes/AuroraGPT-70B.md | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/ALCF/notes/AuroraGPT-70B.md b/ALCF/notes/AuroraGPT-70B.md index 0f69a4db031..7c05f7a2d26 100644 --- a/ALCF/notes/AuroraGPT-70B.md +++ b/ALCF/notes/AuroraGPT-70B.md @@ -63,20 +63,20 @@ PP="${NHOSTS}" \ ## ZeRO-3 + HPZ -| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | -| :----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :------------------------------------------------------------------------- | -| 4 | 48 | 1 | 1 | 1 | 1 | 1 | 48 | True | ??? | ??? | ??? | ??? | -|   | | | | | | | | | | | | | -| 6 | 72 | 1 | 1 | 1 | 1 | 1 | 72 | True | 39.05 | 1.544 | 87.849 | [charmed-fire-2203](https://wandb.ai/aurora_gpt/AuroraGPT/runs/actny2dl) | -|   | | | | | | | | | | | | | -| 8 | 96 | 1 | 1 | 1 | 1 | 1 | 96 | True | 47.60 | 2.510 | 107.087 | [glamorous-wood-2199](https://wandb.ai/aurora_gpt/AuroraGPT/runs/bgu31497) | -| 8 | 96 | 1 | 1 | 1 | 1 | 2 | 192 | True | 55.69 | 2.936 | 125.287 | [solar-plant-2200](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gnc48o99) | -| 8 | 96 | 1 | 1 | 1 | 1 | 4 | 384 | True | 59.66 | 3.146 | 134.216 | [smooth-eon-2207](https://wandb.ai/aurora_gpt/AuroraGPT/runs/awt6d825) | -| 8 | 96 | 1 | 1 | 1 | 1 | 8 | 768 | True | 62.19 | 3.279 | 139.919 | [glamorous-eon-2209](https://wandb.ai/aurora_gpt/AuroraGPT/runs/jmukcjdc) | -|   | | | | | | | | | | | | | -| 16 | 192 | 1 | 1 | 1 | 1 | 1 | 192 | True | 57.6 | 6.074 | 129.577 | [gallant-field-2208](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gb94ahi3) | -| 16 | 384 | 1 | 1 | 1 | 1 | 2 | 384 | True | 61.86 | 6.523 | 139.161 | [swept-shadow-2210](https://wandb.ai/aurora_gpt/AuroraGPT/runs/riy7y3k1) | - +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | +| :-----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :------------------------------------------------------------------------- | +| 4[^z3h] | 48 | 1 | 1 | 1 | 1 | 1 | 48 | True | ??? | ??? | ??? | ??? | +|   | | | | | | | | | | | | | +| 6 | 72 | 1 | 1 | 1 | 1 | 1 | 72 | True | 39.05 | 1.544 | 87.849 | [charmed-fire-2203](https://wandb.ai/aurora_gpt/AuroraGPT/runs/actny2dl) | +|   | | | | | | | | | | | | | +| 8 | 96 | 1 | 1 | 1 | 1 | 1 | 96 | True | 47.60 | 2.510 | 107.087 | [glamorous-wood-2199](https://wandb.ai/aurora_gpt/AuroraGPT/runs/bgu31497) | +| 8 | 96 | 1 | 1 | 1 | 1 | 2 | 192 | True | 55.69 | 2.936 | 125.287 | [solar-plant-2200](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gnc48o99) | +| 8 | 96 | 1 | 1 | 1 | 1 | 4 | 384 | True | 59.66 | 3.146 | 134.216 | [smooth-eon-2207](https://wandb.ai/aurora_gpt/AuroraGPT/runs/awt6d825) | +| 8 | 96 | 1 | 1 | 1 | 1 | 8 | 768 | True | 62.19 | 3.279 | 139.919 | [glamorous-eon-2209](https://wandb.ai/aurora_gpt/AuroraGPT/runs/jmukcjdc) | +|   | | | | | | | | | | | | | +| 16 | 192 | 1 | 1 | 1 | 1 | 1 | 192 | True | 57.6 | 6.074 | 129.577 | [gallant-field-2208](https://wandb.ai/aurora_gpt/AuroraGPT/runs/gb94ahi3) | +| 16 | 192 | 1 | 1 | 1 | 1 | 2 | 384 | True | 61.86 | 6.523 | 139.161 | [swept-shadow-2210](https://wandb.ai/aurora_gpt/AuroraGPT/runs/riy7y3k1) | +| 16 | 192 | 1 | 1 | 1 | 1 | 4 | 768 | True | 64.28 | 6.779 | 144.621 | [restful-oath-2212](https://wandb.ai/aurora_gpt/AuroraGPT/runs/kzmhlcju) | [^z3h]: Seems to get hung on 4 nodes (waited ~30 min) From e7157e408a0ef55cbece914191db99473b9c8a60 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 17 Jun 2025 15:57:51 -0500 Subject: [PATCH 429/590] docs: Update `ALCF/notes/AuroraGPT-70B.md` --- ALCF/notes/AuroraGPT-70B.md | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/ALCF/notes/AuroraGPT-70B.md b/ALCF/notes/AuroraGPT-70B.md index 7c05f7a2d26..530c2145780 100644 --- a/ALCF/notes/AuroraGPT-70B.md +++ b/ALCF/notes/AuroraGPT-70B.md @@ -63,6 +63,27 @@ PP="${NHOSTS}" \ ## ZeRO-3 + HPZ +- Command: + + ```bash + TP=1 \ + GRAD_ACC_STEPS=4 \ + ZERO_HPZ_PARTITION_SIZE="${NGPU_PER_HOST}" \ + PP=1 \ + ZERO_STAGE=3 \ + HEADS=64 \ + NLAYERS=80 \ + NUM_KV_HEAD=8 \ + FFN_HIDDEN_SIZE=28672 \ + HIDDEN=8192 \ + MICRO_BATCH=1 \ + SEQ_LEN=8192 \ + USE_ACTIVATION_CHECKPOINTING=1 \ + DATA_FILE_LIST="ALCF/data-lists/aurora/books.txt" \ + bash train.sh + ``` + + | Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | | :-----: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :------------------------------------------------------------------------- | | 4[^z3h] | 48 | 1 | 1 | 1 | 1 | 1 | 48 | True | ??? | ??? | ??? | ??? | From c9e5879309e7ae472eb09630d23fe83351f41195 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 09:25:20 -0500 Subject: [PATCH 430/590] chore: Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 0dfd33fd67c..20a23905008 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -105,7 +105,7 @@ def model_provider(pre_process=True, post_process=True): with deepspeed_zero_init( data_parallel_group=dpg, remote_device=(None if args.remote_device == "none" else args.remote_device), - config_dict_or_path=args.deepspeed_config_dict, # _dict, + config_dict_or_path=args.deepspeed_config_dict, enabled=args.zero_stage == 3, mpu=mpu, ): From b85d33bd5b60c111465ac93805a1f46e357132a3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 09:25:31 -0500 Subject: [PATCH 431/590] chore: Update `megatron/training_log.py` --- megatron/training_log.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index 933badc31d0..b798ba9b39a 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -150,7 +150,7 @@ def training_log( wandb_metrics |= { "lm-loss-training/iteration": iteration, "lm-loss-training/consumed_train_tokens": args.consumed_train_tokens, - **{"lm-loss-training/" + k: v for k, v in loss_dict.items()}, + **{f"lm-loss-training/{k}": v for k, v in loss_dict.items()}, } if grad_norm is not None: wandb_metrics |= {"training/grad-norm": grad_norm} From a9620b9d86901a5956898842aa15ce39f670bf7c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:13:26 -0500 Subject: [PATCH 432/590] chore: Format `megatron/data/*` --- megatron/data/autoaugment.py | 26 +- megatron/data/bert_dataset.py | 152 ++++--- megatron/data/biencoder_dataset_utils.py | 119 +++--- megatron/data/blendable_dataset.py | 83 ++-- megatron/data/data_samplers.py | 95 ++--- megatron/data/dataset_utils.py | 460 ++++++++++++--------- megatron/data/gpt_dataset.py | 53 ++- megatron/data/ict_dataset.py | 61 ++- megatron/data/image_folder.py | 117 ++++-- megatron/data/indexed_dataset.py | 4 + megatron/data/orqa_wiki_dataset.py | 105 ++--- megatron/data/prompt_dataset.py | 39 +- megatron/data/realm_dataset_utils.py | 125 +++--- megatron/data/realm_index.py | 46 ++- megatron/data/t5_dataset.py | 188 ++++++--- megatron/data/test/test_indexed_dataset.py | 85 ++-- megatron/data/vit_dataset.py | 231 ++++++----- 17 files changed, 1191 insertions(+), 798 deletions(-) diff --git a/megatron/data/autoaugment.py b/megatron/data/autoaugment.py index 585a4fa6a51..6fa7cdab07e 100644 --- a/megatron/data/autoaugment.py +++ b/megatron/data/autoaugment.py @@ -192,9 +192,7 @@ def __init__( "translateY": np.linspace(0, 150 / 331, num_levels), "rotate": np.linspace(0, 30, num_levels), "color": np.linspace(0.0, 0.9, num_levels), - "posterize": np.round(np.linspace(8, 4, num_levels), 0).astype( - np.int - ), + "posterize": np.round(np.linspace(8, 4, num_levels), 0).astype(np.int), "solarize": np.linspace(256, 0, num_levels), # range [0, 256] "contrast": np.linspace(0.0, 0.9, num_levels), "sharpness": np.linspace(0.0, 0.9, num_levels), @@ -275,21 +273,17 @@ def rotate_with_fill(img, magnitude): "color": lambda img, magnitude: ImageEnhance.Color(img).enhance( 1 + magnitude * random.choice([-1, 1]) ), - "posterize": lambda img, magnitude: ImageOps.posterize( - img, magnitude + "posterize": lambda img, magnitude: ImageOps.posterize(img, magnitude), + "solarize": lambda img, magnitude: ImageOps.solarize(img, magnitude), + "contrast": lambda img, magnitude: ImageEnhance.Contrast(img).enhance( + 1 + magnitude * random.choice([-1, 1]) ), - "solarize": lambda img, magnitude: ImageOps.solarize( - img, magnitude + "sharpness": lambda img, magnitude: ImageEnhance.Sharpness(img).enhance( + 1 + magnitude * random.choice([-1, 1]) + ), + "brightness": lambda img, magnitude: ImageEnhance.Brightness(img).enhance( + 1 + magnitude * random.choice([-1, 1]) ), - "contrast": lambda img, magnitude: ImageEnhance.Contrast( - img - ).enhance(1 + magnitude * random.choice([-1, 1])), - "sharpness": lambda img, magnitude: ImageEnhance.Sharpness( - img - ).enhance(1 + magnitude * random.choice([-1, 1])), - "brightness": lambda img, magnitude: ImageEnhance.Brightness( - img - ).enhance(1 + magnitude * random.choice([-1, 1])), "autocontrast": lambda img, magnitude: ImageOps.autocontrast(img), "equalize": lambda img, magnitude: ImageOps.equalize(img), "invert": lambda img, magnitude: ImageOps.invert(img), diff --git a/megatron/data/bert_dataset.py b/megatron/data/bert_dataset.py index 93b7c8cd32f..39162101ede 100644 --- a/megatron/data/bert_dataset.py +++ b/megatron/data/bert_dataset.py @@ -5,25 +5,31 @@ import numpy as np import torch -from megatron import ( - get_args, - get_tokenizer, - mpu, - print_rank_0 -) +from megatron import (get_args, get_tokenizer, mpu, print_rank_0) from megatron.data.dataset_utils import ( get_samples_mapping, get_a_and_b_segments, truncate_segments, create_tokens_and_tokentypes, - create_masked_lm_predictions + create_masked_lm_predictions, ) + class BertDataset(torch.utils.data.Dataset): - def __init__(self, name, indexed_dataset, data_prefix, - num_epochs, max_num_samples, masked_lm_prob, - max_seq_length, short_seq_prob, seed, binary_head): + def __init__( + self, + name, + indexed_dataset, + data_prefix, + num_epochs, + max_num_samples, + masked_lm_prob, + max_seq_length, + short_seq_prob, + seed, + binary_head, + ): # Params to store. self.name = name @@ -36,15 +42,17 @@ def __init__(self, name, indexed_dataset, data_prefix, self.indexed_dataset = indexed_dataset # Build the samples mapping. - self.samples_mapping = get_samples_mapping(self.indexed_dataset, - data_prefix, - num_epochs, - max_num_samples, - self.max_seq_length - 3, # account for added tokens - short_seq_prob, - self.seed, - self.name, - self.binary_head) + self.samples_mapping = get_samples_mapping( + self.indexed_dataset, + data_prefix, + num_epochs, + max_num_samples, + self.max_seq_length - 3, # account for added tokens + short_seq_prob, + self.seed, + self.name, + self.binary_head, + ) # Vocab stuff. tokenizer = get_tokenizer() @@ -66,26 +74,39 @@ def __getitem__(self, idx): # python randint is inclusive whereas the numpy one is exclusive. # We % 2**32 since numpy requres the seed to be between 0 and 2**32 - 1 np_rng = np.random.RandomState(seed=((self.seed + idx) % 2**32)) - train_sample = build_training_sample(sample, seq_length, - self.max_seq_length, # needed for padding - self.vocab_id_list, - self.vocab_id_to_token_dict, - self.cls_id, self.sep_id, - self.mask_id, self.pad_id, - self.masked_lm_prob, np_rng, - self.binary_head) + train_sample = build_training_sample( + sample, + seq_length, + self.max_seq_length, # needed for padding + self.vocab_id_list, + self.vocab_id_to_token_dict, + self.cls_id, + self.sep_id, + self.mask_id, + self.pad_id, + self.masked_lm_prob, + np_rng, + self.binary_head, + ) if args.return_data_index: - train_sample['index'] = np.array([idx], dtype=np.int64) + train_sample["index"] = np.array([idx], dtype=np.int64) return train_sample - - -def build_training_sample(sample, - target_seq_length, max_seq_length, - vocab_id_list, vocab_id_to_token_dict, - cls_id, sep_id, mask_id, pad_id, - masked_lm_prob, np_rng, binary_head): +def build_training_sample( + sample, + target_seq_length, + max_seq_length, + vocab_id_list, + vocab_id_to_token_dict, + cls_id, + sep_id, + mask_id, + pad_id, + masked_lm_prob, + np_rng, + binary_head, +): """Biuld training sample. Arguments: @@ -112,8 +133,7 @@ def build_training_sample(sample, # Divide sample into two segments (A and B). if binary_head: - tokens_a, tokens_b, is_next_random = get_a_and_b_segments(sample, - np_rng) + tokens_a, tokens_b, is_next_random = get_a_and_b_segments(sample, np_rng) else: tokens_a = [] for j in range(len(sample)): @@ -123,45 +143,60 @@ def build_training_sample(sample, # Truncate to `target_sequence_length`. max_num_tokens = target_seq_length - truncated = truncate_segments(tokens_a, tokens_b, len(tokens_a), - len(tokens_b), max_num_tokens, np_rng) + truncated = truncate_segments( + tokens_a, tokens_b, len(tokens_a), len(tokens_b), max_num_tokens, np_rng + ) # Build tokens and toketypes. - tokens, tokentypes = create_tokens_and_tokentypes(tokens_a, tokens_b, - cls_id, sep_id) + tokens, tokentypes = create_tokens_and_tokentypes( + tokens_a, tokens_b, cls_id, sep_id + ) # Masking. max_predictions_per_seq = masked_lm_prob * max_num_tokens (tokens, masked_positions, masked_labels, _, _) = create_masked_lm_predictions( - tokens, vocab_id_list, vocab_id_to_token_dict, masked_lm_prob, - cls_id, sep_id, mask_id, max_predictions_per_seq, np_rng) + tokens, + vocab_id_list, + vocab_id_to_token_dict, + masked_lm_prob, + cls_id, + sep_id, + mask_id, + max_predictions_per_seq, + np_rng, + ) # Padding. - tokens_np, tokentypes_np, labels_np, padding_mask_np, loss_mask_np \ - = pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, - masked_labels, pad_id, max_seq_length) + tokens_np, tokentypes_np, labels_np, padding_mask_np, loss_mask_np = ( + pad_and_convert_to_numpy( + tokens, tokentypes, masked_positions, masked_labels, pad_id, max_seq_length + ) + ) train_sample = { - 'text': tokens_np, - 'types': tokentypes_np, - 'labels': labels_np, - 'is_random': int(is_next_random), - 'loss_mask': loss_mask_np, - 'padding_mask': padding_mask_np, - 'truncated': int(truncated)} + "text": tokens_np, + "types": tokentypes_np, + "labels": labels_np, + "is_random": int(is_next_random), + "loss_mask": loss_mask_np, + "padding_mask": padding_mask_np, + "truncated": int(truncated), + } return train_sample -def pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, - masked_labels, pad_id, max_seq_length): +def pad_and_convert_to_numpy( + tokens, tokentypes, masked_positions, masked_labels, pad_id, max_seq_length +): """Pad sequences and convert them to numpy.""" # Some checks. num_tokens = len(tokens) padding_length = max_seq_length - num_tokens - assert padding_length >= 0, \ - f"num_tokens ({num_tokens}) is greater than " \ + assert padding_length >= 0, ( + f"num_tokens ({num_tokens}) is greater than " "max_seq_length ({max_seq_length})." + ) assert len(tokentypes) == num_tokens assert len(masked_positions) == len(masked_labels) @@ -171,8 +206,7 @@ def pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, tokentypes_np = np.array(tokentypes + filler, dtype=np.int64) # Padding mask. - padding_mask_np = np.array([1] * num_tokens + [0] * padding_length, - dtype=np.int64) + padding_mask_np = np.array([1] * num_tokens + [0] * padding_length, dtype=np.int64) # Lables and loss mask. labels = [-1] * max_seq_length diff --git a/megatron/data/biencoder_dataset_utils.py b/megatron/data/biencoder_dataset_utils.py index d4661cd34a2..f196d0efd41 100644 --- a/megatron/data/biencoder_dataset_utils.py +++ b/megatron/data/biencoder_dataset_utils.py @@ -6,10 +6,11 @@ from megatron import get_args, get_tokenizer, print_rank_0 from megatron.core import mpu, tensor_parallel -from megatron.data.dataset_utils import create_masked_lm_predictions, \ - pad_and_convert_to_numpy +from megatron.data.dataset_utils import create_masked_lm_predictions, pad_and_convert_to_numpy from megatron.data.data_samplers import MegatronPretrainingSampler from deepspeed.accelerator import get_accelerator + + def make_attention_mask(source_block, target_block): """ Returns a 2-dimensional (2-D) attention mask @@ -21,6 +22,7 @@ def make_attention_mask(source_block, target_block): # (source_length, target_length) return mask + def get_one_epoch_dataloader(dataset, micro_batch_size=None): """Specifically one epoch to be used in an indexing job.""" args = get_args() @@ -39,18 +41,23 @@ def get_one_epoch_dataloader(dataset, micro_batch_size=None): micro_batch_size=args.micro_batch_size, data_parallel_rank=mpu.get_data_parallel_rank(), data_parallel_size=mpu.get_data_parallel_world_size(), - drop_last=False) + drop_last=False, + ) - return torch.utils.data.DataLoader(dataset, - batch_sampler=batch_sampler, - num_workers=num_workers, - pin_memory=True) + return torch.utils.data.DataLoader( + dataset, batch_sampler=batch_sampler, num_workers=num_workers, pin_memory=True + ) def get_ict_batch(data_iterator): # Items and their type. - keys = ['query_tokens', 'query_mask', - 'context_tokens', 'context_mask', 'block_data'] + keys = [ + "query_tokens", + "query_mask", + "context_tokens", + "context_mask", + "block_data", + ] datatype = torch.int64 # Broadcast data. @@ -61,14 +68,13 @@ def get_ict_batch(data_iterator): data_b = tensor_parallel.broadcast_data(keys, data, datatype) # Unpack. - query_tokens = data_b['query_tokens'].long() - query_mask = data_b['query_mask'] < 0.5 - context_tokens = data_b['context_tokens'].long() - context_mask = data_b['context_mask'] < 0.5 - block_indices = data_b['block_data'].long() + query_tokens = data_b["query_tokens"].long() + query_mask = data_b["query_mask"] < 0.5 + context_tokens = data_b["context_tokens"].long() + context_mask = data_b["context_mask"] < 0.5 + block_indices = data_b["block_data"].long() - return query_tokens, query_mask,\ - context_tokens, context_mask, block_indices + return query_tokens, query_mask, context_tokens, context_mask, block_indices def join_str_list(str_list): @@ -90,6 +96,7 @@ class BlockSampleData(object): :param doc_idx: the index of the document from which the block comes in the original indexed dataset :param block_idx: a unique integer identifier given to every block. """ + def __init__(self, start_idx, end_idx, doc_idx, block_idx): self.start_idx = start_idx self.end_idx = end_idx @@ -97,13 +104,16 @@ def __init__(self, start_idx, end_idx, doc_idx, block_idx): self.block_idx = block_idx def as_array(self): - return np.array([self.start_idx, self.end_idx, self.doc_idx, self.block_idx]).astype(np.int64) + return np.array( + [self.start_idx, self.end_idx, self.doc_idx, self.block_idx] + ).astype(np.int64) def as_tuple(self): return self.start_idx, self.end_idx, self.doc_idx, self.block_idx class BlockSamplesMapping(object): + def __init__(self, mapping_array): # make sure that the array is compatible with BlockSampleData assert mapping_array.shape[1] == 4 @@ -118,8 +128,17 @@ def __getitem__(self, idx): return sample_data -def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epochs, - max_num_samples, max_seq_length, seed, name, use_one_sent_docs=False): +def get_block_samples_mapping( + block_dataset, + title_dataset, + data_prefix, + num_epochs, + max_num_samples, + max_seq_length, + seed, + name, + use_one_sent_docs=False, +): """Get samples mapping for a dataset over fixed size blocks. This function also requires a dataset of the titles for the source documents since their lengths must be taken into account. @@ -128,30 +147,30 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo if not num_epochs: if not max_num_samples: - raise ValueError("Need to specify either max_num_samples " - "or num_epochs") + raise ValueError("Need to specify either max_num_samples " "or num_epochs") num_epochs = np.iinfo(np.int32).max - 1 if not max_num_samples: max_num_samples = np.iinfo(np.int64).max - 1 # Filename of the index mapping indexmap_filename = data_prefix - indexmap_filename += '_{}_indexmap'.format(name) + indexmap_filename += "_{}_indexmap".format(name) if num_epochs != (np.iinfo(np.int32).max - 1): - indexmap_filename += '_{}ep'.format(num_epochs) + indexmap_filename += "_{}ep".format(num_epochs) if max_num_samples != (np.iinfo(np.int64).max - 1): - indexmap_filename += '_{}mns'.format(max_num_samples) - indexmap_filename += '_{}msl'.format(max_seq_length) - indexmap_filename += '_{}s'.format(seed) + indexmap_filename += "_{}mns".format(max_num_samples) + indexmap_filename += "_{}msl".format(max_seq_length) + indexmap_filename += "_{}s".format(seed) if use_one_sent_docs: - indexmap_filename += '_1sentok' - indexmap_filename += '.npy' + indexmap_filename += "_1sentok" + indexmap_filename += ".npy" # Build the indexed mapping if not exist. - if mpu.get_data_parallel_rank() == 0 and \ - not os.path.isfile(indexmap_filename): - print(' > WARNING: could not find index map file {}, building ' - 'the indices on rank 0 ...'.format(indexmap_filename)) + if mpu.get_data_parallel_rank() == 0 and not os.path.isfile(indexmap_filename): + print( + " > WARNING: could not find index map file {}, building " + "the indices on rank 0 ...".format(indexmap_filename) + ) # Make sure the types match the helpers input types. assert block_dataset.doc_idx.dtype == np.int64 @@ -160,10 +179,10 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo # Build samples mapping verbose = torch.distributed.get_rank() == 0 start_time = time.time() - print_rank_0(' > building samples index mapping for {} ...'.format( - name)) + print_rank_0(" > building samples index mapping for {} ...".format(name)) from megatron.data import helpers + mapping_array = helpers.build_blocks_mapping( block_dataset.doc_idx, block_dataset.sizes, @@ -173,17 +192,17 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo max_seq_length - 3, # account for added tokens seed, verbose, - use_one_sent_docs) - + use_one_sent_docs, + ) - print_rank_0(' > done building samples index mapping') + print_rank_0(" > done building samples index mapping") np.save(indexmap_filename, mapping_array, allow_pickle=True) - print_rank_0(' > saved the index mapping in {}'.format( - indexmap_filename)) + print_rank_0(" > saved the index mapping in {}".format(indexmap_filename)) # Make sure all the ranks have built the mapping - print_rank_0(' > elapsed time to build and save samples mapping ' - '(seconds): {:4f}'.format( - time.time() - start_time)) + print_rank_0( + " > elapsed time to build and save samples mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # This should be a barrier but nccl barrier assumes # device_index=rank which is not the case for model @@ -191,19 +210,19 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo counts = get_accelerator().LongTensor([1]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) assert counts[0].item() == torch.distributed.get_world_size( - group=mpu.get_data_parallel_group()) + group=mpu.get_data_parallel_group() + ) # Load indexed dataset. - print_rank_0(' > loading indexed mapping from {}'.format( - indexmap_filename)) + print_rank_0(" > loading indexed mapping from {}".format(indexmap_filename)) start_time = time.time() - mapping_array = np.load(indexmap_filename, allow_pickle=True, mmap_mode='r') + mapping_array = np.load(indexmap_filename, allow_pickle=True, mmap_mode="r") samples_mapping = BlockSamplesMapping(mapping_array) - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( - time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( - mapping_array.shape[0])) + print_rank_0( + " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) + ) + print_rank_0(" total number of samples: {}".format(mapping_array.shape[0])) return samples_mapping diff --git a/megatron/data/blendable_dataset.py b/megatron/data/blendable_dataset.py index ab164fdc489..ff7dec96227 100755 --- a/megatron/data/blendable_dataset.py +++ b/megatron/data/blendable_dataset.py @@ -21,10 +21,12 @@ log = get_logger(__name__, rank_zero_only=True) dlp = Profile("DATASET") + + class BlendableDataset(torch.utils.data.Dataset): + @dlp.log - def __init__(self, datasets, weights, size, *, - data_cache_path=None): + def __init__(self, datasets, weights, size, *, data_cache_path=None): self.datasets = datasets num_datasets = len(datasets) @@ -46,9 +48,15 @@ def _build_indices(): dataset_sample_index = np.zeros(self.size, dtype=np.int64) from megatron.data import helpers - helpers.build_blending_indices(dataset_index, dataset_sample_index, - weights, num_datasets, self.size, - torch.distributed.get_rank() == 0) + + helpers.build_blending_indices( + dataset_index, + dataset_sample_index, + weights, + num_datasets, + self.size, + torch.distributed.get_rank() == 0, + ) log.info( "> elapsed time for building blendable dataset indices: " f"{time.perf_counter() - start_time:.2f} (sec)" @@ -65,35 +73,49 @@ def _build_indices(): self.dataset_index = np.zeros(self.size, dtype=np.int64) self.dataset_sample_index = np.zeros(self.size, dtype=np.int64) if data_cache_path: - desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() + desc_hash = hashlib.md5(desc.encode("utf-8")).hexdigest() desc_path = os.path.join(data_cache_path, desc_hash + ".dsc") index_path = os.path.join(data_cache_path, desc_hash + "_index.npy") - sample_index_path = os.path.join(data_cache_path, desc_hash + "_sample_index.npy") + sample_index_path = os.path.join( + data_cache_path, desc_hash + "_sample_index.npy" + ) cache_hit = os.path.isfile(index_path) and os.path.isfile(sample_index_path) cache_success = True if torch.distributed.get_rank() == 0 and not cache_hit: - print(' > WARNING: could not find index map files for blendable' - ' dataset, building indices on rank 0 ...', flush=True) + print( + " > WARNING: could not find index map files for blendable" + " dataset, building indices on rank 0 ...", + flush=True, + ) dataset_index, dataset_sample_index = _build_indices() try: log.debug(" > saving index map files") start_time = time.perf_counter() os.makedirs(os.path.dirname(index_path), exist_ok=True) - with open(desc_path, 'wt') as fd: + with open(desc_path, "wt") as fd: fd.write(desc) np.save(index_path, dataset_index, allow_pickle=True) - np.save(sample_index_path, dataset_sample_index, - allow_pickle=True) - log.info(f" > finished saving index map files in {time.perf_counter() - start_time} seconds") + np.save( + sample_index_path, dataset_sample_index, allow_pickle=True + ) + log.info( + f" > finished saving index map files in {time.perf_counter() - start_time} seconds" + ) except OSError: - print(f'There was an error trying to create the data cache directory ({data_cache_path})') - print('or a file in it. This is set with the --data-cache-path argument. Please') - print('ensure you have write access to this directory or specify one that you do have') - print('write access to.') + print( + f"There was an error trying to create the data cache directory ({data_cache_path})" + ) + print( + "or a file in it. This is set with the --data-cache-path argument. Please" + ) + print( + "ensure you have write access to this directory or specify one that you do have" + ) + print("write access to.") cache_success = False self.dataset_index = dataset_index self.dataset_sample_index = dataset_sample_index - ''' I don't think the following piece of code is necessary any more; I commented them out now + """ I don't think the following piece of code is necessary any more; I commented them out now counts = get_accelerator().LongTensor([cache_success]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) @@ -103,33 +125,34 @@ def _build_indices(): torch.distributed.get_world_size(group=mpu.get_sequence_parallel_group())): log.info("Data index creation unsuccessful, exiting.") exit() - ''' + """ torch.distributed.barrier(group=mpu.get_data_parallel_group()) torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) torch.distributed.barrier(group=mpu.get_data_parallel_group()) start_time = time.perf_counter() - log.info(f'> loading blendable dataset index: {index_path}') - self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') + log.info(f"> loading blendable dataset index: {index_path}") + self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode="r") assert self.dataset_index.size == self.size - log.info(f'> loading blendable dataset sample index: {sample_index_path}') - self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') + log.info(f"> loading blendable dataset sample index: {sample_index_path}") + self.dataset_sample_index = np.load( + sample_index_path, allow_pickle=True, mmap_mode="r" + ) assert self.dataset_sample_index.size == self.size - log.info(f'> finished loading in {time.perf_counter() - start_time} seconds') + log.info( + f"> finished loading in {time.perf_counter() - start_time} seconds" + ) else: self.dataset_index, self.dataset_sample_index = _build_indices() - # Check size _ = self.__getitem__(self.size - 1) try: _ = self.__getitem__(self.size) - raise RuntimeError('BlendedDataset size is improperly bounded') + raise RuntimeError("BlendedDataset size is improperly bounded") except IndexError: pass - log.info('> size of blendable dataset: ' - '{} samples'.format(self.size)) - + log.info("> size of blendable dataset: " "{} samples".format(self.size)) def __len__(self): return self.size @@ -139,6 +162,6 @@ def __getitem__(self, idx): dataset_idx = self.dataset_index[idx] sample_idx = self.dataset_sample_index[idx] return { - "dataset_idx" : dataset_idx, + "dataset_idx": dataset_idx, **self.datasets[dataset_idx][sample_idx], } diff --git a/megatron/data/data_samplers.py b/megatron/data/data_samplers.py index b242101b3a6..09fb1341c63 100644 --- a/megatron/data/data_samplers.py +++ b/megatron/data/data_samplers.py @@ -19,14 +19,15 @@ def build_pretraining_data_loader(dataset, consumed_samples): args = get_args() # Megatron sampler - if args.dataloader_type == 'single': + if args.dataloader_type == "single": batch_sampler = MegatronPretrainingSampler( total_samples=len(dataset), consumed_samples=consumed_samples, micro_batch_size=args.micro_batch_size, data_parallel_rank=mpu.get_data_parallel_rank(), - data_parallel_size=mpu.get_data_parallel_world_size()) - elif args.dataloader_type == 'cyclic': + data_parallel_size=mpu.get_data_parallel_world_size(), + ) + elif args.dataloader_type == "cyclic": batch_sampler = MegatronPretrainingRandomSampler( dataset, total_samples=len(dataset), @@ -34,10 +35,12 @@ def build_pretraining_data_loader(dataset, consumed_samples): micro_batch_size=args.micro_batch_size, data_parallel_rank=mpu.get_data_parallel_rank(), data_parallel_size=mpu.get_data_parallel_world_size(), - data_sharding=args.data_sharding) + data_sharding=args.data_sharding, + ) else: - raise Exception('{} dataloader type is not supported.'.format( - args.dataloader_type)) + raise Exception( + "{} dataloader type is not supported.".format(args.dataloader_type) + ) # Torch dataloader. loader = torch.utils.data.DataLoader( @@ -46,24 +49,24 @@ def build_pretraining_data_loader(dataset, consumed_samples): num_workers=args.num_workers, pin_memory=True, multiprocessing_context=( - args.multiprocessing_context if args.num_workers > 0 - else None - ) + args.multiprocessing_context if args.num_workers > 0 else None + ), ) if args.repeated_dataloader: - loader=RepeatingLoader(loader) + loader = RepeatingLoader(loader) return loader + class MegatronPretrainingSampler: def __init__( - self, - total_samples, - consumed_samples, - micro_batch_size, - data_parallel_rank, - data_parallel_size, - drop_last=True + self, + total_samples, + consumed_samples, + micro_batch_size, + data_parallel_rank, + data_parallel_size, + drop_last=True, ): # Keep a copy of input params for later use. self.total_samples = total_samples @@ -76,18 +79,16 @@ def __init__( self.drop_last = drop_last # Sanity checks. - assert self.total_samples > 0, ( - f'no sample to consume: {self.total_samples}' - ) + assert self.total_samples > 0, f"no sample to consume: {self.total_samples}" assert self.consumed_samples < self.total_samples, ( - 'no samples left to consume: ' - f'{self.consumed_samples}, {self.total_samples}' + "no samples left to consume: " + f"{self.consumed_samples}, {self.total_samples}" ) assert self.micro_batch_size > 0 assert data_parallel_size > 0 assert self.data_parallel_rank < data_parallel_size, ( - f'data_parallel_rank should be smaller than data size: ' - f'{self.data_parallel_rank}, {data_parallel_size}' + f"data_parallel_rank should be smaller than data size: " + f"{self.data_parallel_rank}, {data_parallel_size}" ) def __len__(self): @@ -139,14 +140,14 @@ def __getitem__(self, idx): class MegatronPretrainingRandomSampler: def __init__( - self, - dataset, - total_samples, - consumed_samples, - micro_batch_size, - data_parallel_rank, - data_parallel_size, - data_sharding + self, + dataset, + total_samples, + consumed_samples, + micro_batch_size, + data_parallel_rank, + data_parallel_size, + data_sharding, ): # Keep a copy of input params for later use. self.dataset = dataset @@ -164,14 +165,12 @@ def __init__( ) # Sanity checks. - assert self.total_samples > 0, ( - f'no sample to consume: {self.total_samples}' - ) + assert self.total_samples > 0, f"no sample to consume: {self.total_samples}" assert self.micro_batch_size > 0 assert data_parallel_size > 0 assert self.data_parallel_rank < data_parallel_size, ( - f'data_parallel_rank should be smaller than data size: ' - f'{self.data_parallel_rank}, {data_parallel_size}' + f"data_parallel_rank should be smaller than data size: " + f"{self.data_parallel_rank}, {data_parallel_size}" ) def __len__(self): @@ -188,11 +187,8 @@ def __iter__(self): # data sharding and random sampling if self.data_sharding: - bucket_size = ( - self.micro_batch_size * ( - self.total_samples - // self.micro_batch_times_data_parallel_size - ) + bucket_size = self.micro_batch_size * ( + self.total_samples // self.micro_batch_times_data_parallel_size ) bucket_offset = current_epoch_samples // self.data_parallel_size start_idx = self.data_parallel_rank * bucket_size @@ -201,20 +197,17 @@ def __iter__(self): random_idx = torch.randperm(bucket_size, generator=g).tolist() idx_range = [start_idx + x for x in random_idx[bucket_offset:]] else: - full_bucket_size = ( - self.micro_batch_size * ( - self.total_samples - // self.micro_batch_size - ) + full_bucket_size = self.micro_batch_size * ( + self.total_samples // self.micro_batch_size ) full_bucket_offset = current_epoch_samples g = torch.Generator() g.manual_seed(self.epoch) - idx_range_total = ( - torch.randperm(full_bucket_size, generator=g).tolist() - ) + idx_range_total = torch.randperm(full_bucket_size, generator=g).tolist() idx_range_active = idx_range_total[full_bucket_offset:] - idx_range = idx_range_active[self.data_parallel_rank::self.data_parallel_size] + idx_range = idx_range_active[ + self.data_parallel_rank :: self.data_parallel_size + ] batch = [] # Last batch if not complete will be dropped. diff --git a/megatron/data/dataset_utils.py b/megatron/data/dataset_utils.py index 61dd3909e4d..44869f11ddc 100644 --- a/megatron/data/dataset_utils.py +++ b/megatron/data/dataset_utils.py @@ -26,45 +26,48 @@ import numpy as np import torch -from megatron import ( - get_args, - print_rank_0 -) +from megatron import (get_args, print_rank_0) from megatron.core import mpu from megatron.data.blendable_dataset import BlendableDataset from megatron.data.indexed_dataset import make_dataset as make_indexed_dataset from deepspeed.accelerator import get_accelerator -DSET_TYPE_BERT = 'standard_bert' -DSET_TYPE_ICT = 'ict' -DSET_TYPE_T5 = 't5' + +DSET_TYPE_BERT = "standard_bert" +DSET_TYPE_ICT = "ict" +DSET_TYPE_T5 = "t5" DSET_TYPES = [DSET_TYPE_BERT, DSET_TYPE_ICT, DSET_TYPE_T5] -def get_datasets_corpuses_weights_and_num_samples(data_prefix, train_valid_test_num_samples): + +def get_datasets_corpuses_weights_and_num_samples( + data_prefix, train_valid_test_num_samples +): assert len(data_prefix) % 3 == 0 num_datasets = len(data_prefix) // 3 data_new_prefix = [] corpuses = [] for i in range(num_datasets): - data_new_prefix += [data_prefix[3*i], data_prefix[3*i+1]] - corpuses.append(data_prefix[3*i+2]) - prefixes, weights, datasets_train_valid_test_num_samples = \ - get_datasets_weights_and_num_samples(data_new_prefix, - train_valid_test_num_samples) + data_new_prefix += [data_prefix[3 * i], data_prefix[3 * i + 1]] + corpuses.append(data_prefix[3 * i + 2]) + prefixes, weights, datasets_train_valid_test_num_samples = ( + get_datasets_weights_and_num_samples( + data_new_prefix, train_valid_test_num_samples + ) + ) return prefixes, corpuses, weights, datasets_train_valid_test_num_samples -def get_datasets_weights_and_num_samples(data_prefix, - train_valid_test_num_samples): + +def get_datasets_weights_and_num_samples(data_prefix, train_valid_test_num_samples): # The data prefix should be in the format of: # weight-1, data-prefix-1, weight-2, data-prefix-2, .. assert len(data_prefix) % 2 == 0 num_datasets = len(data_prefix) // 2 - weights = [0]*num_datasets - prefixes = [0]*num_datasets + weights = [0] * num_datasets + prefixes = [0] * num_datasets for i in range(num_datasets): - weights[i] = float(data_prefix[2*i]) - prefixes[i] = (data_prefix[2*i+1]).strip() + weights[i] = float(data_prefix[2 * i]) + prefixes[i] = (data_prefix[2 * i + 1]).strip() # Normalize weights weight_sum = 0.0 for weight in weights: @@ -79,14 +82,18 @@ def get_datasets_weights_and_num_samples(data_prefix, datasets_train_valid_test_num_samples = [] for weight in weights: datasets_train_valid_test_num_samples.append( - [int(math.ceil(val * weight * 1.005)) - for val in train_valid_test_num_samples]) + [ + int(math.ceil(val * weight * 1.005)) + for val in train_valid_test_num_samples + ] + ) else: # Used when separate dataset files are provided for train, # valid and test datasets_train_valid_test_num_samples = [ int(math.ceil(train_valid_test_num_samples * weight * 1.005)) - for weight in weights] + for weight in weights + ] return prefixes, weights, datasets_train_valid_test_num_samples @@ -96,11 +103,13 @@ def compile_helper(): is invoked on a single process.""" import os import subprocess + path = os.path.abspath(os.path.dirname(__file__)) - ret = subprocess.run(['make', '-C', path]) + ret = subprocess.run(["make", "-C", path]) if ret.returncode != 0: print("Making C++ dataset helpers module failed, exiting.") import sys + sys.exit(1) @@ -110,7 +119,7 @@ def get_a_and_b_segments(sample, np_rng): # Number of sentences in the sample. n_sentences = len(sample) # Make sure we always have two sentences. - assert n_sentences > 1, 'make sure each sample has at least two sentences.' + assert n_sentences > 1, "make sure each sample has at least two sentences." # First part: # `a_end` is how many sentences go into the `A`. @@ -138,7 +147,7 @@ def get_a_and_b_segments(sample, np_rng): def truncate_segments(tokens_a, tokens_b, len_a, len_b, max_num_tokens, np_rng): """Truncates a pair of sequences to a maximum sequence length.""" - #print(len_a, len_b, max_num_tokens) + # print(len_a, len_b, max_num_tokens) assert len_a > 0 if len_a + len_b <= max_num_tokens: return False @@ -183,8 +192,7 @@ def create_tokens_and_tokentypes(tokens_a, tokens_b, cls_id, sep_id): return tokens, tokentypes -MaskedLmInstance = collections.namedtuple("MaskedLmInstance", - ["index", "label"]) +MaskedLmInstance = collections.namedtuple("MaskedLmInstance", ["index", "label"]) def is_start_piece(piece): @@ -196,18 +204,23 @@ def is_start_piece(piece): return not piece.startswith("##") -def create_masked_lm_predictions(tokens, - vocab_id_list, vocab_id_to_token_dict, - masked_lm_prob, - cls_id, sep_id, mask_id, - max_predictions_per_seq, - np_rng, - max_ngrams=3, - do_whole_word_mask=True, - favor_longer_ngram=False, - do_permutation=False, - geometric_dist=False, - masking_style="bert"): +def create_masked_lm_predictions( + tokens, + vocab_id_list, + vocab_id_to_token_dict, + masked_lm_prob, + cls_id, + sep_id, + mask_id, + max_predictions_per_seq, + np_rng, + max_ngrams=3, + do_whole_word_mask=True, + favor_longer_ngram=False, + do_permutation=False, + geometric_dist=False, + masking_style="bert", +): """Creates the predictions for the masked LM objective. Note: Tokens here are vocab ids and not text tokens.""" @@ -217,7 +230,7 @@ def create_masked_lm_predictions(tokens, # on-the-fly whole word masking is possible. token_boundary = [0] * len(tokens) - for (i, token) in enumerate(tokens): + for i, token in enumerate(tokens): if token == cls_id or token == sep_id: token_boundary[i] = 1 continue @@ -227,8 +240,11 @@ def create_masked_lm_predictions(tokens, # Note that Whole Word Masking does *not* change the training code # at all -- we still predict each WordPiece independently, softmaxed # over the entire vocabulary. - if (do_whole_word_mask and len(cand_indexes) >= 1 and - not is_start_piece(vocab_id_to_token_dict[token])): + if ( + do_whole_word_mask + and len(cand_indexes) >= 1 + and not is_start_piece(vocab_id_to_token_dict[token]) + ): cand_indexes[-1].append(i) else: cand_indexes.append([i]) @@ -241,17 +257,23 @@ def create_masked_lm_predictions(tokens, masked_lm_labels = [] if masked_lm_prob == 0: - return (output_tokens, masked_lm_positions, - masked_lm_labels, token_boundary, None) - - num_to_predict = min(max_predictions_per_seq, - max(1, int(round(len(tokens) * masked_lm_prob)))) + return ( + output_tokens, + masked_lm_positions, + masked_lm_labels, + token_boundary, + None, + ) + + num_to_predict = min( + max_predictions_per_seq, max(1, int(round(len(tokens) * masked_lm_prob))) + ) ngrams = np.arange(1, max_ngrams + 1, dtype=np.int64) if not geometric_dist: # Note(mingdachen): # By default, we set the probilities to favor shorter ngram sequences. - pvals = 1. / np.arange(1, max_ngrams + 1) + pvals = 1.0 / np.arange(1, max_ngrams + 1) pvals /= pvals.sum(keepdims=True) if favor_longer_ngram: pvals = pvals[::-1] @@ -260,7 +282,7 @@ def create_masked_lm_predictions(tokens, for idx in range(len(cand_indexes)): ngram_index = [] for n in ngrams: - ngram_index.append(cand_indexes[idx:idx + n]) + ngram_index.append(cand_indexes[idx : idx + n]) ngram_indexes.append(ngram_index) np_rng.shuffle(ngram_indexes) @@ -280,9 +302,11 @@ def create_masked_lm_predictions(tokens, continue if not geometric_dist: - n = np_rng.choice(ngrams[:len(cand_index_set)], - p=pvals[:len(cand_index_set)] / - pvals[:len(cand_index_set)].sum(keepdims=True)) + n = np_rng.choice( + ngrams[: len(cand_index_set)], + p=pvals[: len(cand_index_set)] + / pvals[: len(cand_index_set)].sum(keepdims=True), + ) else: # Sampling "n" from the geometric distribution and clipping it to # the max_ngrams. Using p=0.2 default from the SpanBERT paper @@ -323,7 +347,9 @@ def create_masked_lm_predictions(tokens, masked_token = tokens[index] # 10% of the time, replace with random word else: - masked_token = vocab_id_list[np_rng.randint(0, len(vocab_id_list))] + masked_token = vocab_id_list[ + np_rng.randint(0, len(vocab_id_list)) + ] elif masking_style == "t5": masked_token = mask_id else: @@ -332,9 +358,11 @@ def create_masked_lm_predictions(tokens, output_tokens[index] = masked_token masked_lms.append(MaskedLmInstance(index=index, label=tokens[index])) - masked_spans.append(MaskedLmInstance( - index=index_set, - label=[tokens[index] for index in index_set])) + masked_spans.append( + MaskedLmInstance( + index=index_set, label=[tokens[index] for index in index_set] + ) + ) assert len(masked_lms) <= num_to_predict np_rng.shuffle(ngram_indexes) @@ -353,9 +381,11 @@ def create_masked_lm_predictions(tokens, if index in covered_indexes or index in select_indexes: continue - n = np.random.choice(ngrams[:len(cand_index_set)], - p=pvals[:len(cand_index_set)] / - pvals[:len(cand_index_set)].sum(keepdims=True)) + n = np.random.choice( + ngrams[: len(cand_index_set)], + p=pvals[: len(cand_index_set)] + / pvals[: len(cand_index_set)].sum(keepdims=True), + ) index_set = sum(cand_index_set[n - 1], []) n -= 1 @@ -395,11 +425,18 @@ def create_masked_lm_predictions(tokens, for p in masked_lms: masked_lm_positions.append(p.index) masked_lm_labels.append(p.label) - return (output_tokens, masked_lm_positions, masked_lm_labels, token_boundary, masked_spans) + return ( + output_tokens, + masked_lm_positions, + masked_lm_labels, + token_boundary, + masked_spans, + ) -def pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, - masked_labels, pad_id, max_seq_length): +def pad_and_convert_to_numpy( + tokens, tokentypes, masked_positions, masked_labels, pad_id, max_seq_length +): """Pad sequences and convert them to numpy.""" # Some checks. @@ -415,8 +452,7 @@ def pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, tokentypes_np = np.array(tokentypes + filler, dtype=np.int64) # Padding mask. - padding_mask_np = np.array([1] * num_tokens + [0] * padding_length, - dtype=np.int64) + padding_mask_np = np.array([1] * num_tokens + [0] * padding_length, dtype=np.int64) # Lables and loss mask. labels = [-1] * max_seq_length @@ -431,32 +467,44 @@ def pad_and_convert_to_numpy(tokens, tokentypes, masked_positions, return tokens_np, tokentypes_np, labels_np, padding_mask_np, loss_mask_np -def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - max_seq_length, - masked_lm_prob, short_seq_prob, seed, - skip_warmup, binary_head=False, - max_seq_length_dec=None, - dataset_type='standard_bert'): +def build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + max_seq_length, + masked_lm_prob, + short_seq_prob, + seed, + skip_warmup, + binary_head=False, + max_seq_length_dec=None, + dataset_type="standard_bert", +): if len(data_prefix) == 1: - return _build_train_valid_test_datasets(data_prefix[0], - data_impl, splits_string, - train_valid_test_num_samples, - max_seq_length, masked_lm_prob, - short_seq_prob, seed, - skip_warmup, - binary_head, - max_seq_length_dec, - dataset_type=dataset_type) + return _build_train_valid_test_datasets( + data_prefix[0], + data_impl, + splits_string, + train_valid_test_num_samples, + max_seq_length, + masked_lm_prob, + short_seq_prob, + seed, + skip_warmup, + binary_head, + max_seq_length_dec, + dataset_type=dataset_type, + ) # Blending dataset. # Parse the values. - output = get_datasets_weights_and_num_samples(data_prefix, - train_valid_test_num_samples) + output = get_datasets_weights_and_num_samples( + data_prefix, train_valid_test_num_samples + ) prefixes, weights, datasets_train_valid_test_num_samples = output train_num_samples, valid_num_samples, test_num_samples = map( - sum, - zip(*datasets_train_valid_test_num_samples) + sum, zip(*datasets_train_valid_test_num_samples) ) # Build individual datasets. @@ -465,11 +513,19 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, test_datasets = [] for i in range(len(prefixes)): train_ds, valid_ds, test_ds = _build_train_valid_test_datasets( - prefixes[i], data_impl, splits_string, + prefixes[i], + data_impl, + splits_string, datasets_train_valid_test_num_samples[i], - max_seq_length, masked_lm_prob, short_seq_prob, - seed, skip_warmup, binary_head, max_seq_length_dec, - dataset_type=dataset_type) + max_seq_length, + masked_lm_prob, + short_seq_prob, + seed, + skip_warmup, + binary_head, + max_seq_length_dec, + dataset_type=dataset_type, + ) if train_ds: train_datasets.append(train_ds) if valid_ds: @@ -480,39 +536,49 @@ def build_train_valid_test_datasets(data_prefix, data_impl, splits_string, # Blend. blending_train_dataset = None if train_datasets: - blending_train_dataset = BlendableDataset(train_datasets, weights, train_num_samples) + blending_train_dataset = BlendableDataset( + train_datasets, weights, train_num_samples + ) blending_valid_dataset = None if valid_datasets: - blending_valid_dataset = BlendableDataset(valid_datasets, weights, valid_num_samples) + blending_valid_dataset = BlendableDataset( + valid_datasets, weights, valid_num_samples + ) blending_test_dataset = None if test_datasets: - blending_test_dataset = BlendableDataset(test_datasets, weights, test_num_samples) - - return (blending_train_dataset, blending_valid_dataset, - blending_test_dataset) - - -def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, - train_valid_test_num_samples, - max_seq_length, - masked_lm_prob, short_seq_prob, seed, - skip_warmup, binary_head, - max_seq_length_dec, - dataset_type='standard_bert'): + blending_test_dataset = BlendableDataset( + test_datasets, weights, test_num_samples + ) + + return (blending_train_dataset, blending_valid_dataset, blending_test_dataset) + + +def _build_train_valid_test_datasets( + data_prefix, + data_impl, + splits_string, + train_valid_test_num_samples, + max_seq_length, + masked_lm_prob, + short_seq_prob, + seed, + skip_warmup, + binary_head, + max_seq_length_dec, + dataset_type="standard_bert", +): if dataset_type not in DSET_TYPES: raise ValueError("Invalid dataset_type: ", dataset_type) # Indexed dataset. - indexed_dataset = get_indexed_dataset_(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = get_indexed_dataset_(data_prefix, data_impl, skip_warmup) if dataset_type == DSET_TYPE_ICT: args = get_args() - title_dataset = get_indexed_dataset_(args.titles_data_path, - data_impl, - skip_warmup) + title_dataset = get_indexed_dataset_( + args.titles_data_path, data_impl, skip_warmup + ) # Get start and end indices of train/valid/train into doc-idx # Note that doc-idx is desinged to be num-docs + 1 so we can @@ -521,26 +587,32 @@ def _build_train_valid_test_datasets(data_prefix, data_impl, splits_string, splits = get_train_valid_test_split_(splits_string, total_num_of_documents) # Print stats about the splits. - print_rank_0(' > dataset split:') + print_rank_0(" > dataset split:") def print_split_stats(name, index): - print_rank_0(' {}:'.format(name)) - print_rank_0(' document indices in [{}, {}) total of {} ' - 'documents'.format(splits[index], splits[index + 1], - splits[index + 1] - splits[index])) + print_rank_0(" {}:".format(name)) + print_rank_0( + " document indices in [{}, {}) total of {} " + "documents".format( + splits[index], splits[index + 1], splits[index + 1] - splits[index] + ) + ) start_index = indexed_dataset.doc_idx[splits[index]] end_index = indexed_dataset.doc_idx[splits[index + 1]] - print_rank_0(' sentence indices in [{}, {}) total of {} ' - 'sentences'.format(start_index, end_index, - end_index - start_index)) - print_split_stats('train', 0) - print_split_stats('validation', 1) - print_split_stats('test', 2) + print_rank_0( + " sentence indices in [{}, {}) total of {} " + "sentences".format(start_index, end_index, end_index - start_index) + ) + + print_split_stats("train", 0) + print_split_stats("validation", 1) + print_split_stats("test", 2) def build_dataset(index, name): from megatron.data.bert_dataset import BertDataset from megatron.data.ict_dataset import ICTDataset from megatron.data.t5_dataset import T5Dataset + dataset = None if splits[index + 1] > splits[index]: # Get the pointer to the original doc-idx so we can set it later. @@ -594,58 +666,56 @@ def build_dataset(index, name): indexed_dataset.set_doc_idx(doc_idx_ptr) # Checks. assert indexed_dataset.doc_idx[0] == 0 - assert indexed_dataset.doc_idx.shape[0] == \ - (total_num_of_documents + 1) + assert indexed_dataset.doc_idx.shape[0] == (total_num_of_documents + 1) return dataset - train_dataset = build_dataset(0, 'train') - valid_dataset = build_dataset(1, 'valid') - test_dataset = build_dataset(2, 'test') + train_dataset = build_dataset(0, "train") + valid_dataset = build_dataset(1, "valid") + test_dataset = build_dataset(2, "test") return (train_dataset, valid_dataset, test_dataset) def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): - print_rank_0(' > building dataset index ...') + print_rank_0(" > building dataset index ...") start_time = time.time() - indexed_dataset = make_indexed_dataset(data_prefix, - data_impl, - skip_warmup) + indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) assert indexed_dataset.sizes.shape[0] == indexed_dataset.doc_idx[-1] - print_rank_0(' > finished creating indexed dataset in {:4f} ' - 'seconds'.format(time.time() - start_time)) + print_rank_0( + " > finished creating indexed dataset in {:4f} " + "seconds".format(time.time() - start_time) + ) - print_rank_0(' > indexed dataset stats:') - print_rank_0(' number of documents: {}'.format( - indexed_dataset.doc_idx.shape[0] - 1)) - print_rank_0(' number of sentences: {}'.format( - indexed_dataset.sizes.shape[0])) + print_rank_0(" > indexed dataset stats:") + print_rank_0( + " number of documents: {}".format(indexed_dataset.doc_idx.shape[0] - 1) + ) + print_rank_0(" number of sentences: {}".format(indexed_dataset.sizes.shape[0])) return indexed_dataset def get_train_valid_test_split_(splits_string, size): - """ Get dataset splits from comma or '/' separated string list.""" + """Get dataset splits from comma or '/' separated string list.""" splits = [] - if splits_string.find(',') != -1: - splits = [float(s) for s in splits_string.split(',')] - elif splits_string.find('/') != -1: - splits = [float(s) for s in splits_string.split('/')] + if splits_string.find(",") != -1: + splits = [float(s) for s in splits_string.split(",")] + elif splits_string.find("/") != -1: + splits = [float(s) for s in splits_string.split("/")] else: splits = [float(splits_string)] while len(splits) < 3: - splits.append(0.) + splits.append(0.0) splits = splits[:3] splits_sum = sum(splits) assert splits_sum > 0.0 splits = [split / splits_sum for split in splits] splits_index = [0] for index, split in enumerate(splits): - splits_index.append(splits_index[index] + - int(round(split * float(size)))) + splits_index.append(splits_index[index] + int(round(split * float(size)))) diff = splits_index[-1] - size for index in range(1, len(splits_index)): splits_index[index] -= diff @@ -653,54 +723,59 @@ def get_train_valid_test_split_(splits_string, size): assert splits_index[-1] == size return splits_index -def get_samples_mapping(indexed_dataset, - data_prefix, - num_epochs, - max_num_samples, - max_seq_length, - short_seq_prob, - seed, - name, - binary_head): + +def get_samples_mapping( + indexed_dataset, + data_prefix, + num_epochs, + max_num_samples, + max_seq_length, + short_seq_prob, + seed, + name, + binary_head, +): """Get a list that maps a sample index to a starting sentence index, end sentence index, and length""" args = get_args() - if args.train_data_exact_num_epochs is not None and name == 'train': + if args.train_data_exact_num_epochs is not None and name == "train": num_epochs = args.train_data_exact_num_epochs max_num_samples = np.iinfo(np.int64).max - 1 else: if not num_epochs: if not max_num_samples: - raise ValueError("Need to specify either max_num_samples " - "or num_epochs") + raise ValueError( + "Need to specify either max_num_samples " "or num_epochs" + ) num_epochs = np.iinfo(np.int32).max - 1 if not max_num_samples: max_num_samples = np.iinfo(np.int64).max - 1 # Filename of the index mapping indexmap_filename = data_prefix - indexmap_filename += '_{}_indexmap'.format(name) - if args.train_data_exact_num_epochs is not None and name == 'train': - indexmap_filename += '_exact{}ep'.format(num_epochs) + indexmap_filename += "_{}_indexmap".format(name) + if args.train_data_exact_num_epochs is not None and name == "train": + indexmap_filename += "_exact{}ep".format(num_epochs) else: if num_epochs != (np.iinfo(np.int32).max - 1): - indexmap_filename += '_{}ep'.format(num_epochs) + indexmap_filename += "_{}ep".format(num_epochs) if max_num_samples != (np.iinfo(np.int64).max - 1): - indexmap_filename += '_{}mns'.format(max_num_samples) - indexmap_filename += '_{}msl'.format(max_seq_length) - indexmap_filename += '_{:0.2f}ssp'.format(short_seq_prob) - indexmap_filename += '_{}s'.format(seed) - indexmap_filename += '.npy' + indexmap_filename += "_{}mns".format(max_num_samples) + indexmap_filename += "_{}msl".format(max_seq_length) + indexmap_filename += "_{:0.2f}ssp".format(short_seq_prob) + indexmap_filename += "_{}s".format(seed) + indexmap_filename += ".npy" - if name == 'train': + if name == "train": # force to use certain index files if args.train_idx_path is not None: indexmap_filename = args.train_idx_path # Build the indexed mapping if not exist. - if torch.distributed.get_rank() == 0 and \ - not os.path.isfile(indexmap_filename): - print(' > WARNING: could not find index map file {}, building ' - 'the indices on rank 0 ...'.format(indexmap_filename)) + if torch.distributed.get_rank() == 0 and not os.path.isfile(indexmap_filename): + print( + " > WARNING: could not find index map file {}, building " + "the indices on rank 0 ...".format(indexmap_filename) + ) # Make sure the types match the helpers input types. assert indexed_dataset.doc_idx.dtype == np.int64 @@ -709,10 +784,10 @@ def get_samples_mapping(indexed_dataset, # Build samples mapping verbose = torch.distributed.get_rank() == 0 start_time = time.time() - print_rank_0(' > building samples index mapping for {} ...'.format( - name)) + print_rank_0(" > building samples index mapping for {} ...".format(name)) # First compile and then import. from megatron.data import helpers + samples_mapping = helpers.build_mapping( indexed_dataset.doc_idx, indexed_dataset.sizes, @@ -722,34 +797,39 @@ def get_samples_mapping(indexed_dataset, short_seq_prob, seed, verbose, - 2 if binary_head else 1) - print_rank_0(' > done building samples index maping') + 2 if binary_head else 1, + ) + print_rank_0(" > done building samples index maping") np.save(indexmap_filename, samples_mapping, allow_pickle=True) - print_rank_0(' > saved the index mapping in {}'.format( - indexmap_filename)) + print_rank_0(" > saved the index mapping in {}".format(indexmap_filename)) # Make sure all the ranks have built the mapping - print_rank_0(' > elasped time to build and save samples mapping ' - '(seconds): {:4f}'.format( - time.time() - start_time)) + print_rank_0( + " > elasped time to build and save samples mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # This should be a barrier but nccl barrier assumes # device_index=rank which is not the case for model # parallel case - if get_accelerator().device_count() > 0: # Skip when CPU-only + if get_accelerator().device_count() > 0: # Skip when CPU-only counts = get_accelerator().LongTensor([1]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) - torch.distributed.all_reduce(counts, group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.all_reduce( + counts, group=mpu.get_pipeline_model_parallel_group() + ) assert counts[0].item() == ( - torch.distributed.get_world_size() // - torch.distributed.get_world_size(group=mpu.get_tensor_model_parallel_group())) + torch.distributed.get_world_size() + // torch.distributed.get_world_size( + group=mpu.get_tensor_model_parallel_group() + ) + ) # Load indexed dataset. - print_rank_0(' > loading indexed mapping from {}'.format( - indexmap_filename)) + print_rank_0(" > loading indexed mapping from {}".format(indexmap_filename)) start_time = time.time() - samples_mapping = np.load(indexmap_filename, allow_pickle=True, mmap_mode='r') - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( - time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( - samples_mapping.shape[0])) + samples_mapping = np.load(indexmap_filename, allow_pickle=True, mmap_mode="r") + print_rank_0( + " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) + ) + print_rank_0(" total number of samples: {}".format(samples_mapping.shape[0])) return samples_mapping diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index d09f08d63a4..18c34c5fc63 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -132,6 +132,7 @@ def Build(self): return self.dataset class BuildCorpusDataset(torch.utils.data.Dataset): + @dlp.log def __init__(self, dataset_builders): self.dataset_builders = dataset_builders @@ -151,14 +152,19 @@ def _build_indices_blended(): dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) weights = num_samples_list / self.num_samples helpers.build_blending_indices( - dataset_index, dataset_sample_index, - weights, self.num_datasets, self.num_samples, - torch.distributed.get_rank() == 0) - log.debug(f"> elapsed time for building blendable dataset indices for corpus {self.dataset_builders[0].corpus}: " - "{:.2f} (sec)".format(time.time() - start_time)) + dataset_index, + dataset_sample_index, + weights, + self.num_datasets, + self.num_samples, + torch.distributed.get_rank() == 0, + ) + log.debug( + f"> elapsed time for building blendable dataset indices for corpus {self.dataset_builders[0].corpus}: " + "{:.2f} (sec)".format(time.time() - start_time) + ) return dataset_index, dataset_sample_index - def _build_indices_concat(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) @@ -175,12 +181,16 @@ def _build_indices_concat(): "{:.2f} (sec)".format(time.time() - start_time) ) return dataset_index, dataset_sample_index - + if args.blend_sample_in_corpus: - self.dataset_index, self.dataset_sample_index = _build_indices_blended() + self.dataset_index, self.dataset_sample_index = ( + _build_indices_blended() + ) else: - self.dataset_index, self.dataset_sample_index = _build_indices_concat() - + self.dataset_index, self.dataset_sample_index = ( + _build_indices_concat() + ) + np_rng = np.random.RandomState(seed=dataset_builders[0].seed) self.shuffle_index = np.arange(self.num_samples) if args.shuffle_sample_in_corpus: @@ -222,6 +232,7 @@ def __getitem__(self, idx): test_datasets = [] # Build individual datasets. args = get_args() + @dlp.log def build_corpus_datasets(dataset_type="train"): start_time = time.time() @@ -415,7 +426,8 @@ def _build_train_valid_test_datasets( def print_split_stats(name, index): log.debug(" {}:".format(name)) log.debug( - " document indices in [{}, {}) total of {} " "documents".format( + " document indices in [{}, {}) total of {} " + "documents".format( splits[index], splits[index + 1], splits[index + 1] - splits[index] ) ) @@ -481,7 +493,8 @@ def _build_train_valid_test_datasets_single( def print_split_stats(name, index): log.debug(" {}:".format(name)) log.debug( - " document indices in [{}, {}) total of {} " "documents".format( + " document indices in [{}, {}) total of {} " + "documents".format( splits[index], splits[index + 1], splits[index + 1] - splits[index] ) ) @@ -601,9 +614,8 @@ def _build_dataset( log.debug(" {}:".format(dataset_name)) log.debug( - " document indices in [0, {}) total of {} " "documents".format( - total_num_of_documents, total_num_of_documents - ) + " document indices in [0, {}) total of {} " + "documents".format(total_num_of_documents, total_num_of_documents) ) documents = np.arange(start=0, stop=total_num_of_documents, step=1, dtype=np.int32) @@ -631,9 +643,8 @@ def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): start_time = time.time() indexed_dataset = make_indexed_dataset(data_prefix, data_impl, skip_warmup) log.debug( - " > finished creating indexed dataset in {:4f} " "seconds".format( - time.time() - start_time - ) + " > finished creating indexed dataset in {:4f} " + "seconds".format(time.time() - start_time) ) log.debug(" number of documents: {}".format(indexed_dataset.sizes.shape[0])) @@ -641,6 +652,7 @@ def get_indexed_dataset_(data_prefix, data_impl, skip_warmup): class GPTDataset(torch.utils.data.Dataset): + @dlp.log def __init__( self, @@ -1077,9 +1089,8 @@ def _build_sample_idx(sizes, doc_idx, seq_length, num_epochs, tokens_per_epoch): def _build_shuffle_idx(num_samples, total_size, np_rng): """Build the range [0, size) and shuffle.""" log.debug( - " > building shuffle index with split [0, {}) and [{}, {}) " "...".format( - num_samples, num_samples, total_size - ) + " > building shuffle index with split [0, {}) and [{}, {}) " + "...".format(num_samples, num_samples, total_size) ) dtype_ = np.uint32 diff --git a/megatron/data/ict_dataset.py b/megatron/data/ict_dataset.py index 6dac35ff9d4..9adc216bb85 100644 --- a/megatron/data/ict_dataset.py +++ b/megatron/data/ict_dataset.py @@ -9,6 +9,7 @@ from megatron.data.dataset_utils import get_indexed_dataset_ from megatron.data.realm_dataset_utils import get_block_samples_mapping + def make_attention_mask(source_block, target_block): """ Returns a 2-dimensional (2-D) attention mask @@ -20,16 +21,17 @@ def make_attention_mask(source_block, target_block): # (source_length, target_length) return mask + def get_ict_dataset(use_titles=True, query_in_block_prob=1): """Get a dataset which uses block samples mappings to get ICT/block indexing data (via get_block()) rather than for training, since it is only built with a single epoch sample mapping. """ args = get_args() - block_dataset = get_indexed_dataset_(args.data_path, 'mmap', True) - titles_dataset = get_indexed_dataset_(args.titles_data_path, 'mmap', True) + block_dataset = get_indexed_dataset_(args.data_path, "mmap", True) + titles_dataset = get_indexed_dataset_(args.titles_data_path, "mmap", True) kwargs = dict( - name='full', + name="full", block_dataset=block_dataset, title_dataset=titles_dataset, data_prefix=args.data_path, @@ -39,7 +41,7 @@ def get_ict_dataset(use_titles=True, query_in_block_prob=1): seed=1, query_in_block_prob=query_in_block_prob, use_titles=use_titles, - use_one_sent_docs=args.use_one_sent_docs + use_one_sent_docs=args.use_one_sent_docs, ) dataset = ICTDataset(**kwargs) return dataset @@ -47,9 +49,22 @@ def get_ict_dataset(use_titles=True, query_in_block_prob=1): class ICTDataset(Dataset): """Dataset containing sentences and their blocks for an inverse cloze task.""" - def __init__(self, name, block_dataset, title_dataset, data_prefix, - num_epochs, max_num_samples, max_seq_length, query_in_block_prob, - seed, use_titles=True, use_one_sent_docs=False, binary_head=False): + + def __init__( + self, + name, + block_dataset, + title_dataset, + data_prefix, + num_epochs, + max_num_samples, + max_seq_length, + query_in_block_prob, + seed, + use_titles=True, + use_one_sent_docs=False, + binary_head=False, + ): self.name = name self.seed = seed self.max_seq_length = max_seq_length @@ -61,8 +76,16 @@ def __init__(self, name, block_dataset, title_dataset, data_prefix, self.use_one_sent_docs = use_one_sent_docs self.samples_mapping = get_block_samples_mapping( - block_dataset, title_dataset, data_prefix, num_epochs, - max_num_samples, max_seq_length, seed, name, use_one_sent_docs) + block_dataset, + title_dataset, + data_prefix, + num_epochs, + max_num_samples, + max_seq_length, + seed, + name, + use_one_sent_docs, + ) self.tokenizer = get_tokenizer() self.vocab_id_list = list(self.tokenizer.inv_vocab.keys()) self.vocab_id_to_token_list = self.tokenizer.inv_vocab @@ -99,8 +122,8 @@ def __getitem__(self, idx): # still need to truncate because blocks are concluded when # the sentence lengths have exceeded max_seq_length. - query = query[:self.max_seq_length - 2] - block = list(itertools.chain(*block))[:self.max_seq_length - title_pad_offset] + query = query[: self.max_seq_length - 2] + block = list(itertools.chain(*block))[: self.max_seq_length - title_pad_offset] query_tokens, query_pad_mask = self.concat_and_pad_tokens(query) context_tokens, context_pad_mask = self.concat_and_pad_tokens(block, title) @@ -111,13 +134,13 @@ def __getitem__(self, idx): block_data = sample_data.as_array() sample = { - 'query_tokens': query_tokens, - 'query_mask': query_mask, - 'query_pad_mask': query_pad_mask, - 'context_tokens': context_tokens, - 'context_mask': context_mask, - 'context_pad_mask': context_pad_mask, - 'block_data': block_data, + "query_tokens": query_tokens, + "query_mask": query_mask, + "query_pad_mask": query_pad_mask, + "context_tokens": context_tokens, + "context_mask": context_mask, + "context_pad_mask": context_pad_mask, + "block_data": block_data, } return sample @@ -127,7 +150,7 @@ def get_block(self, start_idx, end_idx, doc_idx): block = [self.block_dataset[i] for i in range(start_idx, end_idx)] title = self.title_dataset[int(doc_idx)] - block = list(itertools.chain(*block))[:self.max_seq_length - (3 + len(title))] + block = list(itertools.chain(*block))[: self.max_seq_length - (3 + len(title))] block_tokens, block_pad_mask = self.concat_and_pad_tokens(block, title) return block_tokens, block_pad_mask diff --git a/megatron/data/image_folder.py b/megatron/data/image_folder.py index de15b29bf06..a49eda08e7e 100644 --- a/megatron/data/image_folder.py +++ b/megatron/data/image_folder.py @@ -1,6 +1,6 @@ # BSD 3-Clause License # -# Copyright (c) Soumith Chintala 2016, +# Copyright (c) Soumith Chintala 2016, # All rights reserved. # # Redistribution and use in source and binary forms, with or without @@ -28,7 +28,7 @@ # OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -# code taken from +# code taken from # https://github.com/pytorch/vision/blob/main/torchvision/datasets/folder.py # added support for classes_fraction and data_per_class_fraction @@ -40,6 +40,7 @@ from typing import Any, Callable, cast, Dict, List, Optional, Tuple import numpy as np + def has_file_allowed_extension(filename: str, extensions: Tuple[str, ...]) -> bool: """Checks if a file is an allowed extension. Args: @@ -88,10 +89,14 @@ def make_dataset( both_none = extensions is None and is_valid_file is None both_something = extensions is not None and is_valid_file is not None if both_none or both_something: - raise ValueError("Both extensions and is_valid_file cannot be None or not None at the same time") + raise ValueError( + "Both extensions and is_valid_file cannot be None or not None at the same time" + ) if extensions is not None: + def is_valid_file(x: str) -> bool: return has_file_allowed_extension(x, cast(Tuple[str, ...], extensions)) + is_valid_file = cast(Callable[[str], bool], is_valid_file) for target_class in sorted(class_to_idx.keys()): class_index = class_to_idx[target_class] @@ -106,7 +111,9 @@ def is_valid_file(x: str) -> bool: item = path, class_index local_instances.append(item) - instances.extend(local_instances[0:int(len(local_instances) * data_per_class_fraction)]) + instances.extend( + local_instances[0 : int(len(local_instances) * data_per_class_fraction)] + ) return instances @@ -140,26 +147,29 @@ class DatasetFolder(VisionDataset): """ def __init__( - self, - root: str, - loader: Callable[[str], Any], - extensions: Optional[Tuple[str, ...]] = None, - transform: Optional[Callable] = None, - target_transform: Optional[Callable] = None, - classes_fraction=1.0, - data_per_class_fraction=1.0, - is_valid_file: Optional[Callable[[str], bool]] = None, + self, + root: str, + loader: Callable[[str], Any], + extensions: Optional[Tuple[str, ...]] = None, + transform: Optional[Callable] = None, + target_transform: Optional[Callable] = None, + classes_fraction=1.0, + data_per_class_fraction=1.0, + is_valid_file: Optional[Callable[[str], bool]] = None, ) -> None: - super(DatasetFolder, self).__init__(root, transform=transform, - target_transform=target_transform) + super(DatasetFolder, self).__init__( + root, transform=transform, target_transform=target_transform + ) self.classes_fraction = classes_fraction self.data_per_class_fraction = data_per_class_fraction classes, class_to_idx = self._find_classes(self.root) - samples = self.make_dataset(self.root, - class_to_idx, - self.data_per_class_fraction, - extensions, - is_valid_file) + samples = self.make_dataset( + self.root, + class_to_idx, + self.data_per_class_fraction, + extensions, + is_valid_file, + ) if len(samples) == 0: msg = "Found 0 files in subfolders of: {}\n".format(self.root) if extensions is not None: @@ -182,11 +192,13 @@ def make_dataset( extensions: Optional[Tuple[str, ...]] = None, is_valid_file: Optional[Callable[[str], bool]] = None, ) -> List[Tuple[str, int]]: - return make_dataset(directory, - class_to_idx, - data_per_class_fraction, - extensions=extensions, - is_valid_file=is_valid_file) + return make_dataset( + directory, + class_to_idx, + data_per_class_fraction, + extensions=extensions, + is_valid_file=is_valid_file, + ) def _find_classes(self, dir: str) -> Tuple[List[str], Dict[str, int]]: """ @@ -199,7 +211,7 @@ def _find_classes(self, dir: str) -> Tuple[List[str], Dict[str, int]]: No class is a subdirectory of another. """ all_classes = [d.name for d in os.scandir(dir) if d.is_dir()] - classes = all_classes[0:int(len(all_classes) * self.classes_fraction)] + classes = all_classes[0 : int(len(all_classes) * self.classes_fraction)] classes.sort() class_to_idx = {cls_name: i for i, cls_name in enumerate(classes)} return classes, class_to_idx @@ -231,19 +243,30 @@ def __len__(self) -> int: return len(self.samples) -IMG_EXTENSIONS = ('.jpg', '.jpeg', '.png', '.ppm', '.bmp', '.pgm', '.tif', '.tiff', '.webp') +IMG_EXTENSIONS = ( + ".jpg", + ".jpeg", + ".png", + ".ppm", + ".bmp", + ".pgm", + ".tif", + ".tiff", + ".webp", +) def pil_loader(path: str) -> Image.Image: # open path as file to avoid ResourceWarning (https://github.com/python-pillow/Pillow/issues/835) - with open(path, 'rb') as f: + with open(path, "rb") as f: img = Image.open(f) - return img.convert('RGB') + return img.convert("RGB") # TODO: specify the return type def accimage_loader(path: str) -> Any: import accimage + try: return accimage.Image(path) except IOError: @@ -253,7 +276,8 @@ def accimage_loader(path: str) -> Any: def default_loader(path: str) -> Any: from torchvision import get_image_backend - if get_image_backend() == 'accimage': + + if get_image_backend() == "accimage": return accimage_loader(path) else: return pil_loader(path) @@ -283,20 +307,23 @@ class ImageFolder(DatasetFolder): """ def __init__( - self, - root: str, - transform: Optional[Callable] = None, - target_transform: Optional[Callable] = None, - classes_fraction=1.0, - data_per_class_fraction=1.0, - loader: Callable[[str], Any] = default_loader, - is_valid_file: Optional[Callable[[str], bool]] = None, + self, + root: str, + transform: Optional[Callable] = None, + target_transform: Optional[Callable] = None, + classes_fraction=1.0, + data_per_class_fraction=1.0, + loader: Callable[[str], Any] = default_loader, + is_valid_file: Optional[Callable[[str], bool]] = None, ): - super(ImageFolder, self).__init__(root, loader, IMG_EXTENSIONS if is_valid_file is None else None, - transform=transform, - target_transform=target_transform, - classes_fraction=classes_fraction, - data_per_class_fraction=data_per_class_fraction, - is_valid_file=is_valid_file) + super(ImageFolder, self).__init__( + root, + loader, + IMG_EXTENSIONS if is_valid_file is None else None, + transform=transform, + target_transform=target_transform, + classes_fraction=classes_fraction, + data_per_class_fraction=data_per_class_fraction, + is_valid_file=is_valid_file, + ) self.imgs = self.samples - diff --git a/megatron/data/indexed_dataset.py b/megatron/data/indexed_dataset.py index e2a0c4751ff..f7f6e0e1d3a 100644 --- a/megatron/data/indexed_dataset.py +++ b/megatron/data/indexed_dataset.py @@ -229,6 +229,7 @@ def supports_prefetch(self): class IndexedCachedDataset(IndexedDataset): + def __init__(self, path): super().__init__(path) self.cache = None @@ -395,12 +396,14 @@ def get_pointers_with_total(sizes, elemsize, dtype): class MMapIndexedDataset(torch.utils.data.Dataset): + class Index(object): _HDR_MAGIC = b"MMIDIDX\x00\x00" @classmethod def writer(cls, path, dtype): class _Writer(object): + def __enter__(self): self._file = open(path, "wb") @@ -626,6 +629,7 @@ def dtype(self): class MMapIndexedDatasetBuilder(object): + def __init__(self, out_file, dtype=np.int64): self._data_file = open(out_file, "wb") self._dtype = dtype diff --git a/megatron/data/orqa_wiki_dataset.py b/megatron/data/orqa_wiki_dataset.py index 4019cd764c2..45ac1aaf8b9 100644 --- a/megatron/data/orqa_wiki_dataset.py +++ b/megatron/data/orqa_wiki_dataset.py @@ -13,22 +13,24 @@ from megatron.core import tensor_parallel from megatron.data.biencoder_dataset_utils import make_attention_mask + def get_open_retrieval_wiki_dataset(): args = get_args() tokenizer = get_tokenizer() - dataset = OpenRetrievalEvidenceDataset('2018 Wikipedia from DPR codebase', - 'evidence', - args.evidence_data_path, - tokenizer, - args.retriever_seq_length) + dataset = OpenRetrievalEvidenceDataset( + "2018 Wikipedia from DPR codebase", + "evidence", + args.evidence_data_path, + tokenizer, + args.retriever_seq_length, + ) return dataset def get_open_retrieval_batch(data_iterator): # Items and their type. - keys = ['row_id', 'context', 'context_mask', 'context_types', - 'context_pad_mask'] + keys = ["row_id", "context", "context_mask", "context_types", "context_pad_mask"] datatype = torch.int64 # Broadcast data. @@ -36,14 +38,14 @@ def get_open_retrieval_batch(data_iterator): data_b = tensor_parallel.broadcast_data(keys, data, datatype) # Unpack. - row_id = data_b['row_id'].long() - context = data_b['context'].long() + row_id = data_b["row_id"].long() + context = data_b["context"].long() # TODO: make the context mask a binary one - context_mask = (data_b['context_mask'] < 0.5) + context_mask = data_b["context_mask"] < 0.5 - context_types = data_b['context_types'].long() - context_pad_mask = data_b['context_pad_mask'].long() + context_types = data_b["context_types"].long() + context_pad_mask = data_b["context_pad_mask"].long() return row_id, context, context_mask, context_types, context_pad_mask @@ -51,22 +53,27 @@ def get_open_retrieval_batch(data_iterator): def build_tokens_types_paddings_from_text(row, tokenizer, max_seq_length): """Build token types and paddings, trim if needed, and pad if needed.""" - title_ids = tokenizer.tokenize(row['title']) - context_ids = tokenizer.tokenize(row['text']) + title_ids = tokenizer.tokenize(row["title"]) + context_ids = tokenizer.tokenize(row["text"]) # Appending the title of the context at front extended_context_ids = title_ids + [tokenizer.sep_id] + context_ids - context_ids, context_types, context_pad_mask = \ - build_tokens_types_paddings_from_ids(extended_context_ids, - max_seq_length, tokenizer.cls, tokenizer.sep, tokenizer.pad) + context_ids, context_types, context_pad_mask = build_tokens_types_paddings_from_ids( + extended_context_ids, + max_seq_length, + tokenizer.cls, + tokenizer.sep, + tokenizer.pad, + ) return context_ids, context_types, context_pad_mask # noinspection DuplicatedCode -def build_tokens_types_paddings_from_ids(text_ids, max_seq_length, - cls_id, sep_id, pad_id): +def build_tokens_types_paddings_from_ids( + text_ids, max_seq_length, cls_id, sep_id, pad_id +): """Build token types and paddings, trim if needed, and pad if needed.""" enc_ids = [] tokentypes_enc = [] @@ -82,8 +89,8 @@ def build_tokens_types_paddings_from_ids(text_ids, max_seq_length, # Cap the size. if len(enc_ids) > max_seq_length - 1: - enc_ids = enc_ids[0: max_seq_length - 1] - tokentypes_enc = tokentypes_enc[0: max_seq_length - 1] + enc_ids = enc_ids[0 : max_seq_length - 1] + tokentypes_enc = tokentypes_enc[0 : max_seq_length - 1] # [SEP]. enc_ids.append(sep_id) @@ -109,40 +116,38 @@ def build_sample(row_id, context_ids, context_types, context_pad_mask): context_types = np.array(context_types, dtype=np.int64) context_mask = make_attention_mask(context_ids, context_ids) - sample = ({ - 'row_id': row_id, - 'context': context_ids, - 'context_mask': context_mask, - 'context_types': context_types, - 'context_pad_mask': context_pad_mask - }) + sample = { + "row_id": row_id, + "context": context_ids, + "context_mask": context_mask, + "context_types": context_types, + "context_pad_mask": context_pad_mask, + } return sample class OpenRetrievalEvidenceDataset(ABC, Dataset): """Open Retrieval Evidence dataset class.""" - def __init__(self, task_name, dataset_name, datapath, tokenizer, - max_seq_length): + def __init__(self, task_name, dataset_name, datapath, tokenizer, max_seq_length): # Store inputs. self.task_name = task_name self.dataset_name = dataset_name self.tokenizer = tokenizer self.max_seq_length = max_seq_length - print_rank_0(' > building {} dataset for {}:'.format(self.task_name, - self.dataset_name)) + print_rank_0( + " > building {} dataset for {}:".format(self.task_name, self.dataset_name) + ) # Process the files. print_rank_0(datapath) - self.samples, self.id2text = self.process_samples_from_single_path( - datapath) + self.samples, self.id2text = self.process_samples_from_single_path(datapath) args = get_args() if args.sample_rate < 1: # subsample k = int(len(self.samples) * args.sample_rate) self.samples = random.sample(self.samples, k) - print_rank_0(' >> total number of samples: {}'.format( - len(self.samples))) + print_rank_0(" >> total number of samples: {}".format(len(self.samples))) def __len__(self): return len(self.samples) @@ -150,26 +155,27 @@ def __len__(self): def __getitem__(self, idx): row = self.samples[idx] - context_ids, context_types, context_pad_mask = \ - build_tokens_types_paddings_from_text(row, self.tokenizer, - self.max_seq_length) + context_ids, context_types, context_pad_mask = ( + build_tokens_types_paddings_from_text( + row, self.tokenizer, self.max_seq_length + ) + ) - sample = build_sample(row['doc_id'], - context_ids, - context_types, - context_pad_mask) + sample = build_sample( + row["doc_id"], context_ids, context_types, context_pad_mask + ) return sample @staticmethod def process_samples_from_single_path(filename): - print_rank_0(' > Processing {} ...'.format(filename)) + print_rank_0(" > Processing {} ...".format(filename)) total = 0 rows = [] id2text = {} with open(filename) as tsvfile: - reader = csv.reader(tsvfile, delimiter='\t') + reader = csv.reader(tsvfile, delimiter="\t") next(reader, None) # skip the headers for row in reader: # file format: doc_id, doc_text, title @@ -177,17 +183,14 @@ def process_samples_from_single_path(filename): text = row[1] title = row[2] - rows.append({'doc_id': doc_id, - 'text': text, - 'title': title}) + rows.append({"doc_id": doc_id, "text": text, "title": title}) assert doc_id not in id2text id2text[doc_id] = (text, title) total += 1 if total % 100000 == 0: - print_rank_0(' > processed {} rows so far ...'.format( - total)) + print_rank_0(" > processed {} rows so far ...".format(total)) - print_rank_0(' >> processed {} samples.'.format(len(rows))) + print_rank_0(" >> processed {} samples.".format(len(rows))) return rows, id2text diff --git a/megatron/data/prompt_dataset.py b/megatron/data/prompt_dataset.py index b873e1dc3a4..a550a9bb373 100644 --- a/megatron/data/prompt_dataset.py +++ b/megatron/data/prompt_dataset.py @@ -1,4 +1,3 @@ - import ezpz # Utilizing code snippet from https://github.com/tatsu-lab/stanford_alpaca import copy @@ -25,17 +24,22 @@ "### Instruction:\n{instruction}\n\n### Response:" ), } + + def _make_r_io_base(f, mode: str): if not isinstance(f, io.IOBase): f = open(f, mode=mode) return f + def jload(f, mode="r"): """Load a .json file into a dictionary.""" f = _make_r_io_base(f, mode) jdict = json.load(f) f.close() return jdict + + class SupervisedDataset(Dataset): """Dataset for supervised fine-tuning.""" @@ -45,24 +49,35 @@ def __init__(self, data_path: str, HFtokenizer): logger.warning("Loading data...") list_data_dict = jload(data_path) logger.warning("Formatting inputs...") - prompt_input, prompt_no_input = PROMPT_DICT["prompt_input"], PROMPT_DICT["prompt_no_input"] + prompt_input, prompt_no_input = ( + PROMPT_DICT["prompt_input"], + PROMPT_DICT["prompt_no_input"], + ) sources = [ - prompt_input.format_map(example) if example.get("input", "") != "" else prompt_no_input.format_map(example) + prompt_input.format_map(example) + if example.get("input", "") != "" + else prompt_no_input.format_map(example) for example in list_data_dict ] - targets = [f"{example['output']}{tokenizer.eos_token}" for example in list_data_dict] + targets = [ + f"{example['output']}{tokenizer.eos_token}" for example in list_data_dict + ] logger.warning("Tokenizing inputs... This may take some time...") data_dict = preprocess(sources, targets, tokenizer) self.input_ids = data_dict["input_ids"] self.labels = data_dict["labels"] + def __len__(self): return len(self.input_ids) + def __getitem__(self, i) -> Dict[str, torch.Tensor]: return dict(input_ids=self.input_ids[i], labels=self.labels[i]) -def _tokenize_fn(strings: Sequence[str], tokenizer: transformers.PreTrainedTokenizer) -> Dict: +def _tokenize_fn( + strings: Sequence[str], tokenizer: transformers.PreTrainedTokenizer +) -> Dict: """Tokenize a list of strings.""" # +1 for alignment labels and tokens tokenized_list = [ @@ -70,14 +85,15 @@ def _tokenize_fn(strings: Sequence[str], tokenizer: transformers.PreTrainedToken text, return_tensors="pt", padding="max_length", - max_length=tokenizer.model_max_length+1, + max_length=tokenizer.model_max_length + 1, truncation=True, ) for text in strings ] input_ids = labels = [tokenized.input_ids[0] for tokenized in tokenized_list] input_ids_lens = labels_lens = [ - tokenized.input_ids.ne(tokenizer.pad_token_id).sum().item() for tokenized in tokenized_list + tokenized.input_ids.ne(tokenizer.pad_token_id).sum().item() + for tokenized in tokenized_list ] return dict( input_ids=input_ids, @@ -85,7 +101,8 @@ def _tokenize_fn(strings: Sequence[str], tokenizer: transformers.PreTrainedToken input_ids_lens=input_ids_lens, labels_lens=labels_lens, ) - + + def preprocess( sources: Sequence[str], targets: Sequence[str], @@ -93,10 +110,12 @@ def preprocess( ) -> Dict: """Preprocess the data by tokenizing.""" examples = [s + t for s, t in zip(sources, targets)] - examples_tokenized, sources_tokenized = [_tokenize_fn(strings, tokenizer) for strings in (examples, sources)] + examples_tokenized, sources_tokenized = [ + _tokenize_fn(strings, tokenizer) for strings in (examples, sources) + ] input_ids = examples_tokenized["input_ids"] labels = copy.deepcopy(input_ids) - + "Here we use padding to fill the prompt in the labels." for label, source_len in zip(labels, sources_tokenized["input_ids_lens"]): label[:source_len] = tokenizer.pad_token_id diff --git a/megatron/data/realm_dataset_utils.py b/megatron/data/realm_dataset_utils.py index bf3cd55683f..df0f6033b70 100644 --- a/megatron/data/realm_dataset_utils.py +++ b/megatron/data/realm_dataset_utils.py @@ -24,24 +24,31 @@ def get_one_epoch_dataloader(dataset, micro_batch_size=None): sampler = torch.utils.data.SequentialSampler(dataset) # importantly, drop_last must be False to get all the data. - assert False, 'DistributedBatchSampler deprecated, change the implementation' + assert False, "DistributedBatchSampler deprecated, change the implementation" from megatron.data.samplers import DistributedBatchSampler - batch_sampler = DistributedBatchSampler(sampler, - batch_size=global_batch_size, - drop_last=False, - rank=rank, - world_size=world_size) - return torch.utils.data.DataLoader(dataset, - batch_sampler=batch_sampler, - num_workers=num_workers, - pin_memory=True) + batch_sampler = DistributedBatchSampler( + sampler, + batch_size=global_batch_size, + drop_last=False, + rank=rank, + world_size=world_size, + ) + + return torch.utils.data.DataLoader( + dataset, batch_sampler=batch_sampler, num_workers=num_workers, pin_memory=True + ) def get_ict_batch(data_iterator): # Items and their type. - keys = ['query_tokens', 'query_pad_mask', - 'block_tokens', 'block_pad_mask', 'block_data'] + keys = [ + "query_tokens", + "query_pad_mask", + "block_tokens", + "block_pad_mask", + "block_data", + ] datatype = torch.int64 # Broadcast data. @@ -52,14 +59,13 @@ def get_ict_batch(data_iterator): data_b = tensor_parallel.broadcast_data(keys, data, datatype) # Unpack. - query_tokens = data_b['query_tokens'].long() - query_pad_mask = data_b['query_pad_mask'].long() - block_tokens = data_b['block_tokens'].long() - block_pad_mask = data_b['block_pad_mask'].long() - block_indices = data_b['block_data'].long() + query_tokens = data_b["query_tokens"].long() + query_pad_mask = data_b["query_pad_mask"].long() + block_tokens = data_b["block_tokens"].long() + block_pad_mask = data_b["block_pad_mask"].long() + block_indices = data_b["block_data"].long() - return query_tokens, query_pad_mask,\ - block_tokens, block_pad_mask, block_indices + return query_tokens, query_pad_mask, block_tokens, block_pad_mask, block_indices def join_str_list(str_list): @@ -81,6 +87,7 @@ class BlockSampleData(object): :param doc_idx: the index of the document from which the block comes in the original indexed dataset :param block_idx: a unique integer identifier given to every block. """ + def __init__(self, start_idx, end_idx, doc_idx, block_idx): self.start_idx = start_idx self.end_idx = end_idx @@ -88,13 +95,16 @@ def __init__(self, start_idx, end_idx, doc_idx, block_idx): self.block_idx = block_idx def as_array(self): - return np.array([self.start_idx, self.end_idx, self.doc_idx, self.block_idx]).astype(np.int64) + return np.array( + [self.start_idx, self.end_idx, self.doc_idx, self.block_idx] + ).astype(np.int64) def as_tuple(self): return self.start_idx, self.end_idx, self.doc_idx, self.block_idx class BlockSamplesMapping(object): + def __init__(self, mapping_array): # make sure that the array is compatible with BlockSampleData assert mapping_array.shape[1] == 4 @@ -109,8 +119,17 @@ def __getitem__(self, idx): return sample_data -def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epochs, - max_num_samples, max_seq_length, seed, name, use_one_sent_docs=False): +def get_block_samples_mapping( + block_dataset, + title_dataset, + data_prefix, + num_epochs, + max_num_samples, + max_seq_length, + seed, + name, + use_one_sent_docs=False, +): """Get samples mapping for a dataset over fixed size blocks. This function also requires a dataset of the titles for the source documents since their lengths must be taken into account. @@ -119,30 +138,30 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo if not num_epochs: if not max_num_samples: - raise ValueError("Need to specify either max_num_samples " - "or num_epochs") + raise ValueError("Need to specify either max_num_samples " "or num_epochs") num_epochs = np.iinfo(np.int32).max - 1 if not max_num_samples: max_num_samples = np.iinfo(np.int64).max - 1 # Filename of the index mapping indexmap_filename = data_prefix - indexmap_filename += '_{}_indexmap'.format(name) + indexmap_filename += "_{}_indexmap".format(name) if num_epochs != (np.iinfo(np.int32).max - 1): - indexmap_filename += '_{}ep'.format(num_epochs) + indexmap_filename += "_{}ep".format(num_epochs) if max_num_samples != (np.iinfo(np.int64).max - 1): - indexmap_filename += '_{}mns'.format(max_num_samples) - indexmap_filename += '_{}msl'.format(max_seq_length) - indexmap_filename += '_{}s'.format(seed) + indexmap_filename += "_{}mns".format(max_num_samples) + indexmap_filename += "_{}msl".format(max_seq_length) + indexmap_filename += "_{}s".format(seed) if use_one_sent_docs: - indexmap_filename += '_1sentok' - indexmap_filename += '.npy' + indexmap_filename += "_1sentok" + indexmap_filename += ".npy" # Build the indexed mapping if not exist. - if mpu.get_data_parallel_rank() == 0 and \ - not os.path.isfile(indexmap_filename): - print(' > WARNING: could not find index map file {}, building ' - 'the indices on rank 0 ...'.format(indexmap_filename)) + if mpu.get_data_parallel_rank() == 0 and not os.path.isfile(indexmap_filename): + print( + " > WARNING: could not find index map file {}, building " + "the indices on rank 0 ...".format(indexmap_filename) + ) # Make sure the types match the helpers input types. assert block_dataset.doc_idx.dtype == np.int64 @@ -151,10 +170,10 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo # Build samples mapping verbose = torch.distributed.get_rank() == 0 start_time = time.time() - print_rank_0(' > building samples index mapping for {} ...'.format( - name)) + print_rank_0(" > building samples index mapping for {} ...".format(name)) from megatron.data import helpers + mapping_array = helpers.build_blocks_mapping( block_dataset.doc_idx, block_dataset.sizes, @@ -164,17 +183,17 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo max_seq_length - 3, # account for added tokens seed, verbose, - use_one_sent_docs) - + use_one_sent_docs, + ) - print_rank_0(' > done building samples index mapping') + print_rank_0(" > done building samples index mapping") np.save(indexmap_filename, mapping_array, allow_pickle=True) - print_rank_0(' > saved the index mapping in {}'.format( - indexmap_filename)) + print_rank_0(" > saved the index mapping in {}".format(indexmap_filename)) # Make sure all the ranks have built the mapping - print_rank_0(' > elapsed time to build and save samples mapping ' - '(seconds): {:4f}'.format( - time.time() - start_time)) + print_rank_0( + " > elapsed time to build and save samples mapping " + "(seconds): {:4f}".format(time.time() - start_time) + ) # This should be a barrier but nccl barrier assumes # device_index=rank which is not the case for model @@ -182,19 +201,19 @@ def get_block_samples_mapping(block_dataset, title_dataset, data_prefix, num_epo counts = get_accelerator().LongTensor([1]) torch.distributed.all_reduce(counts, group=mpu.get_data_parallel_group()) assert counts[0].item() == torch.distributed.get_world_size( - group=mpu.get_data_parallel_group()) + group=mpu.get_data_parallel_group() + ) # Load indexed dataset. - print_rank_0(' > loading indexed mapping from {}'.format( - indexmap_filename)) + print_rank_0(" > loading indexed mapping from {}".format(indexmap_filename)) start_time = time.time() - mapping_array = np.load(indexmap_filename, allow_pickle=True, mmap_mode='r') + mapping_array = np.load(indexmap_filename, allow_pickle=True, mmap_mode="r") samples_mapping = BlockSamplesMapping(mapping_array) - print_rank_0(' loaded indexed file in {:3.3f} seconds'.format( - time.time() - start_time)) - print_rank_0(' total number of samples: {}'.format( - mapping_array.shape[0])) + print_rank_0( + " loaded indexed file in {:3.3f} seconds".format(time.time() - start_time) + ) + print_rank_0(" total number of samples: {}".format(mapping_array.shape[0])) return samples_mapping diff --git a/megatron/data/realm_index.py b/megatron/data/realm_index.py index 1fa4a309edc..e67e9491564 100644 --- a/megatron/data/realm_index.py +++ b/megatron/data/realm_index.py @@ -19,6 +19,7 @@ class OpenRetreivalDataStore(object): Serializable data structure for holding data for blocks -- embeddings and necessary metadata for Retriever """ + def __init__(self, embedding_path=None, load_from_path=True, rank=None): self.embed_data = dict() if embedding_path is None: @@ -32,11 +33,11 @@ def __init__(self, embedding_path=None, load_from_path=True, rank=None): self.load_from_file() block_data_name = os.path.splitext(self.embedding_path)[0] - self.temp_dir_name = block_data_name + '_tmp' + self.temp_dir_name = block_data_name + "_tmp" def state(self): return { - 'embed_data': self.embed_data, + "embed_data": self.embed_data, } def clear(self): @@ -52,11 +53,11 @@ def load_from_file(self): if not mpu.model_parallel_is_initialized() or mpu.get_data_parallel_rank() == 0: print("\n> Unpickling BlockData", flush=True) - state_dict = pickle.load(open(self.embedding_path, 'rb')) + state_dict = pickle.load(open(self.embedding_path, "rb")) if not mpu.model_parallel_is_initialized() or mpu.get_data_parallel_rank() == 0: print(">> Finished unpickling BlockData\n", flush=True) - self.embed_data = state_dict['embed_data'] + self.embed_data = state_dict["embed_data"] def add_block_data(self, row_id, block_embeds, allow_overwrite=False): """ @@ -79,12 +80,11 @@ def save_shard(self): os.makedirs(self.temp_dir_name, exist_ok=True) # save the data for each shard - with open('{}/{}.pkl'.format(self.temp_dir_name, self.rank), 'wb') \ - as writer: + with open("{}/{}.pkl".format(self.temp_dir_name, self.rank), "wb") as writer: pickle.dump(self.state(), writer) def merge_shards_and_save(self): - #Combine all the shards made using save_shard + # Combine all the shards made using save_shard shard_names = os.listdir(self.temp_dir_name) seen_own_shard = False @@ -94,31 +94,36 @@ def merge_shards_and_save(self): seen_own_shard = True continue - with open('{}/{}'.format(self.temp_dir_name, fname), 'rb') as f: + with open("{}/{}".format(self.temp_dir_name, fname), "rb") as f: data = pickle.load(f) old_size = len(self.embed_data) - shard_size = len(data['embed_data']) + shard_size = len(data["embed_data"]) # add the shard's data and check to make sure there # is no overlap - self.embed_data.update(data['embed_data']) + self.embed_data.update(data["embed_data"]) assert len(self.embed_data) == old_size + shard_size assert seen_own_shard # save the consolidated shards and remove temporary directory - with open(self.embedding_path, 'wb') as final_file: + with open(self.embedding_path, "wb") as final_file: pickle.dump(self.state(), final_file) shutil.rmtree(self.temp_dir_name, ignore_errors=True) - print("Finished merging {} shards for a total of {} embeds".format( - len(shard_names), len(self.embed_data)), flush=True) + print( + "Finished merging {} shards for a total of {} embeds".format( + len(shard_names), len(self.embed_data) + ), + flush=True, + ) class FaissMIPSIndex(object): """ Wrapper object for a BlockData which similarity search via FAISS under the hood """ + def __init__(self, embed_size, embed_data=None, use_gpu=False): self.embed_size = embed_size self.embed_data = embed_data @@ -149,12 +154,18 @@ def _set_mips_index(self): config.useFloat16 = True gpu_index = faiss.index_cpu_to_all_gpus(cpu_index, co=config) self.mips_index = faiss.IndexIDMap(gpu_index) - if not mpu.model_parallel_is_initialized() or mpu.get_data_parallel_rank() == 0: + if ( + not mpu.model_parallel_is_initialized() + or mpu.get_data_parallel_rank() == 0 + ): print(">> Initialized index on GPU", flush=True) else: # CPU index supports IDs so wrap with IDMap self.mips_index = faiss.IndexIDMap(cpu_index) - if not mpu.model_parallel_is_initialized() or mpu.get_data_parallel_rank() == 0: + if ( + not mpu.model_parallel_is_initialized() + or mpu.get_data_parallel_rank() == 0 + ): print(">> Initialized index on CPU", flush=True) # if we were constructed with a BlockData, then automatically load it @@ -215,8 +226,9 @@ def search_mips_index(self, query_embeds, top_k, reconstruct=True): if reconstruct: # get the vectors themselves - top_k_block_embeds = self.mips_index.search_and_reconstruct(\ - query_embeds, top_k) + top_k_block_embeds = self.mips_index.search_and_reconstruct( + query_embeds, top_k + ) return top_k_block_embeds else: # get distances and indices of closest vectors diff --git a/megatron/data/t5_dataset.py b/megatron/data/t5_dataset.py index e606814909b..b9c6ad3f8b4 100644 --- a/megatron/data/t5_dataset.py +++ b/megatron/data/t5_dataset.py @@ -10,15 +10,25 @@ from megatron import get_tokenizer from megatron.data.dataset_utils import ( create_masked_lm_predictions, - get_samples_mapping + get_samples_mapping, ) + class T5Dataset(torch.utils.data.Dataset): - def __init__(self, name, indexed_dataset, data_prefix, - num_epochs, max_num_samples, masked_lm_prob, - max_seq_length, max_seq_length_dec, - short_seq_prob, seed): + def __init__( + self, + name, + indexed_dataset, + data_prefix, + num_epochs, + max_num_samples, + masked_lm_prob, + max_seq_length, + max_seq_length_dec, + short_seq_prob, + seed, + ): # Params to store. self.name = name @@ -31,15 +41,17 @@ def __init__(self, name, indexed_dataset, data_prefix, self.indexed_dataset = indexed_dataset # Build the samples mapping. - self.samples_mapping = get_samples_mapping(self.indexed_dataset, - data_prefix, - num_epochs, - max_num_samples, - self.max_seq_length - 2, # account for added tokens - short_seq_prob, - self.seed, - self.name, - False) + self.samples_mapping = get_samples_mapping( + self.indexed_dataset, + data_prefix, + num_epochs, + max_num_samples, + self.max_seq_length - 2, # account for added tokens + short_seq_prob, + self.seed, + self.name, + False, + ) # Vocab stuff. tokenizer = get_tokenizer() @@ -52,7 +64,9 @@ def __init__(self, name, indexed_dataset, data_prefix, self.bos_id = tokenizer.bos_token_id self.eos_id = tokenizer.eos_token_id self.sentinel_tokens = tokenizer.additional_special_tokens_ids - assert len(self.sentinel_tokens) > 0, "Provide the argument --vocab-extra-ids 100 to the script" + assert ( + len(self.sentinel_tokens) > 0 + ), "Provide the argument --vocab-extra-ids 100 to the script" def __len__(self): return self.samples_mapping.shape[0] @@ -66,24 +80,42 @@ def __getitem__(self, idx): # Note that this rng state should be numpy and not python since # python randint is inclusive whereas the numpy one is exclusive. np_rng = np.random.RandomState(seed=(self.seed + idx)) - return build_training_sample(sample, seq_length, - self.max_seq_length, # needed for padding - self.max_seq_length_dec, - self.vocab_id_list, - self.vocab_id_to_token_dict, - self.cls_id, self.sep_id, - self.mask_id, self.pad_id, - self.masked_lm_prob, np_rng, - self.bos_id, self.eos_id, - self.sentinel_tokens) - - -def build_training_sample(sample, target_seq_length, - max_seq_length, max_seq_length_dec, - vocab_id_list, vocab_id_to_token_dict, - cls_id, sep_id, mask_id, pad_id, - masked_lm_prob, np_rng, bos_id=None, - eos_id=None, sentinel_tokens=None): + return build_training_sample( + sample, + seq_length, + self.max_seq_length, # needed for padding + self.max_seq_length_dec, + self.vocab_id_list, + self.vocab_id_to_token_dict, + self.cls_id, + self.sep_id, + self.mask_id, + self.pad_id, + self.masked_lm_prob, + np_rng, + self.bos_id, + self.eos_id, + self.sentinel_tokens, + ) + + +def build_training_sample( + sample, + target_seq_length, + max_seq_length, + max_seq_length_dec, + vocab_id_list, + vocab_id_to_token_dict, + cls_id, + sep_id, + mask_id, + pad_id, + masked_lm_prob, + np_rng, + bos_id=None, + eos_id=None, + sentinel_tokens=None, +): """Build training sample. Arguments: @@ -118,37 +150,64 @@ def build_training_sample(sample, target_seq_length, # Masking. max_predictions_per_seq = masked_lm_prob * max_num_tokens - (tokens, masked_positions, masked_labels, _, masked_spans) = create_masked_lm_predictions( - tokens, vocab_id_list, vocab_id_to_token_dict, masked_lm_prob, - cls_id, sep_id, mask_id, max_predictions_per_seq, np_rng, - max_ngrams=10, geometric_dist=True, masking_style="t5") + (tokens, masked_positions, masked_labels, _, masked_spans) = ( + create_masked_lm_predictions( + tokens, + vocab_id_list, + vocab_id_to_token_dict, + masked_lm_prob, + cls_id, + sep_id, + mask_id, + max_predictions_per_seq, + np_rng, + max_ngrams=10, + geometric_dist=True, + masking_style="t5", + ) + ) # Padding. - tokens_enc, tokens_dec_in, labels, enc_mask, \ - dec_mask, enc_dec_mask, loss_mask \ - = pad_and_convert_to_numpy(tokens, masked_positions, - masked_labels, pad_id, max_seq_length, - max_seq_length_dec, masked_spans, - bos_id, eos_id, sentinel_tokens) + tokens_enc, tokens_dec_in, labels, enc_mask, dec_mask, enc_dec_mask, loss_mask = ( + pad_and_convert_to_numpy( + tokens, + masked_positions, + masked_labels, + pad_id, + max_seq_length, + max_seq_length_dec, + masked_spans, + bos_id, + eos_id, + sentinel_tokens, + ) + ) train_sample = { - 'text_enc': tokens_enc, - 'text_dec': tokens_dec_in, - 'labels': labels, - 'loss_mask': loss_mask, - 'truncated': int(truncated), - 'enc_mask': enc_mask, - 'dec_mask': dec_mask, - 'enc_dec_mask': enc_dec_mask, + "text_enc": tokens_enc, + "text_dec": tokens_dec_in, + "labels": labels, + "loss_mask": loss_mask, + "truncated": int(truncated), + "enc_mask": enc_mask, + "dec_mask": dec_mask, + "enc_dec_mask": enc_dec_mask, } return train_sample -def pad_and_convert_to_numpy(tokens, masked_positions, - masked_labels, pad_id, - max_seq_length, max_seq_length_dec, - masked_spans=None, bos_id=None, - eos_id=None, sentinel_tokens=None): +def pad_and_convert_to_numpy( + tokens, + masked_positions, + masked_labels, + pad_id, + max_seq_length, + max_seq_length_dec, + masked_spans=None, + bos_id=None, + eos_id=None, + sentinel_tokens=None, +): """Pad sequences and convert them to numpy.""" sentinel_tokens = collections.deque(sentinel_tokens) @@ -165,7 +224,7 @@ def pad_and_convert_to_numpy(tokens, masked_positions, t5_decoder_out.extend(span.label) end_index = span.index[0] - t5_input.extend(tokens[start_index: end_index]) + t5_input.extend(tokens[start_index:end_index]) t5_input.append(flag) # the next start index is the token after the last span token @@ -213,8 +272,15 @@ def pad_and_convert_to_numpy(tokens, masked_positions, loss_mask = ([1] * num_tokens_dec) + ([0] * padding_length_dec) loss_mask = np.array(loss_mask, dtype=np.int64) - return tokens_enc, tokens_dec_in, labels, enc_mask, \ - dec_mask, enc_dec_mask, loss_mask + return ( + tokens_enc, + tokens_dec_in, + labels, + enc_mask, + dec_mask, + enc_dec_mask, + loss_mask, + ) def make_attention_mask(source_block, target_block): @@ -244,7 +310,7 @@ def make_attention_mask_3d(source_block, target_block): def make_history_mask(block): length = block.shape[0] arange = np.arange(length) - history_mask = (arange[None, ] <= arange[:, None]) + history_mask = arange[None,] <= arange[:, None] history_mask = history_mask.astype(np.int64) return history_mask @@ -252,6 +318,6 @@ def make_history_mask(block): def make_history_mask_3d(block): batch, length = block.shape arange = torch.arange(length, device=block.device) - history_mask = (arange[None, ] <= arange[:, None])[None, ] + history_mask = (arange[None,] <= arange[:, None])[None,] history_mask = history_mask.expand(batch, length, length) return history_mask diff --git a/megatron/data/test/test_indexed_dataset.py b/megatron/data/test/test_indexed_dataset.py index 12fec8d8192..4c8a71de1ef 100644 --- a/megatron/data/test/test_indexed_dataset.py +++ b/megatron/data/test/test_indexed_dataset.py @@ -61,6 +61,7 @@ def test_indexed_dataset_get(args): print(part) # print(tokenizer.detokenize(part.data.tolist())) + # def test_albert_dataset(args): # # tokenizer = FullBertTokenizer(args.vocab, do_lower_case=True) # # idataset = indexed_dataset.make_dataset(args.data, args.dataset_impl) @@ -81,34 +82,60 @@ def test_indexed_dataset_get(args): def main(): parser = argparse.ArgumentParser() - parser.add_argument('--data', type=str, help='prefix to data files') - parser.add_argument('--dataset-impl', type=str, default='infer', - choices=['lazy', 'cached', 'mmap', 'infer']) - parser.add_argument('--count', type=int, default=10, - help='Number of samples/documents to print') - - group = parser.add_argument_group(title='tokenizer') - group.add_argument('--tokenizer-type', type=str, required=True, - choices=['BertWordPieceLowerCase', - 'GPT2BPETokenizer'], - help='What type of tokenizer to use.') - group.add_argument('--vocab-file', type=str, default=None, - help='Path to the vocab file') - group.add_argument('--merge-file', type=str, default=None, - help='Path to the BPE merge file (if necessary).') - - parser.add_argument('--epochs', type=int, default=5, - help='Number of epochs to plan for') - parser.add_argument('--max-num-samples', type=int, default=None, - help='Maximum number of samples to plan for') - parser.add_argument('--masked-lm-prob', type=float, default=0.15, - help='probability of masking tokens') - parser.add_argument('--seq-length', type=int, default=512, - help='maximum sequence length') - parser.add_argument('--short-seq-prob', type=float, default=0.1, - help='probability of creating a short sequence') - parser.add_argument('--seed', type=int, default=1234, - help='random seed') + parser.add_argument("--data", type=str, help="prefix to data files") + parser.add_argument( + "--dataset-impl", + type=str, + default="infer", + choices=["lazy", "cached", "mmap", "infer"], + ) + parser.add_argument( + "--count", type=int, default=10, help="Number of samples/documents to print" + ) + + group = parser.add_argument_group(title="tokenizer") + group.add_argument( + "--tokenizer-type", + type=str, + required=True, + choices=["BertWordPieceLowerCase", "GPT2BPETokenizer"], + help="What type of tokenizer to use.", + ) + group.add_argument( + "--vocab-file", type=str, default=None, help="Path to the vocab file" + ) + group.add_argument( + "--merge-file", + type=str, + default=None, + help="Path to the BPE merge file (if necessary).", + ) + + parser.add_argument( + "--epochs", type=int, default=5, help="Number of epochs to plan for" + ) + parser.add_argument( + "--max-num-samples", + type=int, + default=None, + help="Maximum number of samples to plan for", + ) + parser.add_argument( + "--masked-lm-prob", + type=float, + default=0.15, + help="probability of masking tokens", + ) + parser.add_argument( + "--seq-length", type=int, default=512, help="maximum sequence length" + ) + parser.add_argument( + "--short-seq-prob", + type=float, + default=0.1, + help="probability of creating a short sequence", + ) + parser.add_argument("--seed", type=int, default=1234, help="random seed") args = parser.parse_args() args.rank = 0 args.make_vocab_size_divisible_by = 128 @@ -117,7 +144,7 @@ def main(): if args.dataset_impl == "infer": args.dataset_impl = indexed_dataset.infer_dataset_impl(args.data) -# test_albert_dataset(args) + # test_albert_dataset(args) test_indexed_dataset_get(args) diff --git a/megatron/data/vit_dataset.py b/megatron/data/vit_dataset.py index 82391e9157e..1dee48709b0 100644 --- a/megatron/data/vit_dataset.py +++ b/megatron/data/vit_dataset.py @@ -16,7 +16,8 @@ class GaussianBlur(object): """ Apply Gaussian Blur to the PIL image. """ - def __init__(self, p=0.5, radius_min=0.1, radius_max=2.): + + def __init__(self, p=0.5, radius_min=0.1, radius_max=2.0): self.prob = p self.radius_min = radius_min self.radius_max = radius_max @@ -37,6 +38,7 @@ class Solarization(object): """ Apply Solarization to the PIL image. """ + def __init__(self, p): self.p = p @@ -47,64 +49,77 @@ def __call__(self, img): return img -class ClassificationTransform(): +class ClassificationTransform: + def __init__(self, image_size, train=True): args = get_args() assert args.fp16 or args.bf16 self.data_type = torch.half if args.fp16 else torch.bfloat16 if train: - self.transform = T.Compose([ - T.RandomResizedCrop(image_size), - T.RandomHorizontalFlip(), - T.ColorJitter(0.4, 0.4, 0.4, 0.1), - ImageNetPolicy(), - T.ToTensor(), - T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), - T.ConvertImageDtype(self.data_type) - ]) + self.transform = T.Compose( + [ + T.RandomResizedCrop(image_size), + T.RandomHorizontalFlip(), + T.ColorJitter(0.4, 0.4, 0.4, 0.1), + ImageNetPolicy(), + T.ToTensor(), + T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), + T.ConvertImageDtype(self.data_type), + ] + ) else: - self.transform = T.Compose([ - T.Resize(image_size), - T.CenterCrop(image_size), - T.ToTensor(), - T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), - T.ConvertImageDtype(self.data_type) - ]) + self.transform = T.Compose( + [ + T.Resize(image_size), + T.CenterCrop(image_size), + T.ToTensor(), + T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), + T.ConvertImageDtype(self.data_type), + ] + ) def __call__(self, input): output = self.transform(input) return output -class InpaintingTransform(): - def __init__(self, image_size, train=True): +class InpaintingTransform: + def __init__(self, image_size, train=True): args = get_args() self.mask_factor = args.mask_factor self.mask_type = args.mask_type self.image_size = image_size self.patch_size = args.patch_dim - self.mask_size = int(self.mask_factor*(image_size[0]/self.patch_size)*(image_size[1]/self.patch_size)) + self.mask_size = int( + self.mask_factor + * (image_size[0] / self.patch_size) + * (image_size[1] / self.patch_size) + ) self.train = train assert args.fp16 or args.bf16 self.data_type = torch.half if args.fp16 else torch.bfloat16 - + if self.train: - self.transform = T.Compose([ - T.RandomResizedCrop(self.image_size), - T.RandomHorizontalFlip(), - T.ColorJitter(0.4, 0.4, 0.4, 0.1), - ImageNetPolicy(), - T.ToTensor(), - T.ConvertImageDtype(self.data_type) - ]) + self.transform = T.Compose( + [ + T.RandomResizedCrop(self.image_size), + T.RandomHorizontalFlip(), + T.ColorJitter(0.4, 0.4, 0.4, 0.1), + ImageNetPolicy(), + T.ToTensor(), + T.ConvertImageDtype(self.data_type), + ] + ) else: - self.transform = T.Compose([ - T.Resize(self.image_size, interpolation=2), - T.CenterCrop(self.image_size), - T.ToTensor(), - T.ConvertImageDtype(self.data_type) - ]) + self.transform = T.Compose( + [ + T.Resize(self.image_size, interpolation=2), + T.CenterCrop(self.image_size), + T.ToTensor(), + T.ConvertImageDtype(self.data_type), + ] + ) def gen_mask(self, image_size, mask_size, mask_type, patch_size): # output: mask as a list with indices for missing patches @@ -115,7 +130,7 @@ def gen_mask(self, image_size, mask_size, mask_type, patch_size): # drop masked patches mask = torch.zeros((image_size[0], image_size[1]), dtype=torch.float) - if mask_type == 'random': + if mask_type == "random": x = torch.randint(0, img_size_patch, ()) y = torch.randint(0, img_size_patch, ()) for i in range(mask_size): @@ -124,84 +139,108 @@ def gen_mask(self, image_size, mask_size, mask_type, patch_size): y = torch.clamp(y + action_list[r][1], min=0, max=img_size_patch - 1) x_offset = x * patch_size y_offset = y * patch_size - mask[x_offset:x_offset+patch_size, y_offset:y_offset+patch_size] = 1 + mask[ + x_offset : x_offset + patch_size, y_offset : y_offset + patch_size + ] = 1 else: - assert mask_type == 'row' + assert mask_type == "row" count = 0 for x in reversed(range(img_size_patch)): for y in reversed(range(img_size_patch)): - if (count < mask_size): + if count < mask_size: count += 1 x_offset = x * patch_size y_offset = y * patch_size - mask[x_offset:x_offset+patch_size, y_offset:y_offset+patch_size] = 1 + mask[ + x_offset : x_offset + patch_size, + y_offset : y_offset + patch_size, + ] = 1 return mask def __call__(self, input): trans_input = self.transform(input) - mask = self.gen_mask(self.image_size, self.mask_size, - self.mask_type, self.patch_size) + mask = self.gen_mask( + self.image_size, self.mask_size, self.mask_type, self.patch_size + ) mask = mask.unsqueeze(dim=0) return trans_input, mask class DinoTransform(object): + def __init__(self, image_size, train=True): args = get_args() self.data_type = torch.half if args.fp16 else torch.bfloat16 - flip_and_color_jitter = T.Compose([ - T.RandomHorizontalFlip(p=0.5), - T.RandomApply( - [T.ColorJitter(brightness=0.4, contrast=0.4, - saturation=0.2, hue=0.1)], - p=0.8 - ), - T.RandomGrayscale(p=0.2), - ]) + flip_and_color_jitter = T.Compose( + [ + T.RandomHorizontalFlip(p=0.5), + T.RandomApply( + [ + T.ColorJitter( + brightness=0.4, contrast=0.4, saturation=0.2, hue=0.1 + ) + ], + p=0.8, + ), + T.RandomGrayscale(p=0.2), + ] + ) if args.fp16 or args.bf16: - normalize = T.Compose([ - T.ToTensor(), - T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), - T.ConvertImageDtype(self.data_type) - ]) + normalize = T.Compose( + [ + T.ToTensor(), + T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), + T.ConvertImageDtype(self.data_type), + ] + ) else: - normalize = T.Compose([ - T.ToTensor(), - T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), - ]) + normalize = T.Compose( + [ + T.ToTensor(), + T.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)), + ] + ) # first global crop scale_const = 0.4 - self.global_transform1 = T.Compose([ - T.RandomResizedCrop(image_size, - scale=(scale_const, 1), - interpolation=Image.BICUBIC), - flip_and_color_jitter, - GaussianBlur(1.0), - normalize - ]) + self.global_transform1 = T.Compose( + [ + T.RandomResizedCrop( + image_size, scale=(scale_const, 1), interpolation=Image.BICUBIC + ), + flip_and_color_jitter, + GaussianBlur(1.0), + normalize, + ] + ) # second global crop - self.global_transform2 = T.Compose([ - T.RandomResizedCrop(image_size, - scale=(scale_const, 1), - interpolation=Image.BICUBIC), - flip_and_color_jitter, - GaussianBlur(0.1), - Solarization(0.2), - normalize - ]) + self.global_transform2 = T.Compose( + [ + T.RandomResizedCrop( + image_size, scale=(scale_const, 1), interpolation=Image.BICUBIC + ), + flip_and_color_jitter, + GaussianBlur(0.1), + Solarization(0.2), + normalize, + ] + ) # transformation for the local small crops self.local_crops_number = args.dino_local_crops_number - self.local_transform = T.Compose([ - T.RandomResizedCrop(args.dino_local_img_size, - scale=(0.05, scale_const), - interpolation=Image.BICUBIC), - flip_and_color_jitter, - GaussianBlur(p=0.5), - normalize - ]) + self.local_transform = T.Compose( + [ + T.RandomResizedCrop( + args.dino_local_img_size, + scale=(0.05, scale_const), + interpolation=Image.BICUBIC, + ), + flip_and_color_jitter, + GaussianBlur(p=0.5), + normalize, + ] + ) def __call__(self, image): crops = [] @@ -215,18 +254,21 @@ def __call__(self, image): def build_train_valid_datasets(data_path, image_size=224): args = get_args() - if args.vision_pretraining_type == 'classify': + if args.vision_pretraining_type == "classify": train_transform = ClassificationTransform(image_size) val_transform = ClassificationTransform(image_size, train=False) - elif args.vision_pretraining_type == 'inpaint': + elif args.vision_pretraining_type == "inpaint": train_transform = InpaintingTransform(image_size, train=False) val_transform = InpaintingTransform(image_size, train=False) - elif args.vision_pretraining_type == 'dino': + elif args.vision_pretraining_type == "dino": train_transform = DinoTransform(image_size, train=True) val_transform = ClassificationTransform(image_size, train=False) else: - raise Exception('{} vit pretraining type is not supported.'.format( - args.vit_pretraining_type)) + raise Exception( + "{} vit pretraining type is not supported.".format( + args.vit_pretraining_type + ) + ) # training dataset train_data_path = data_path[0] if len(data_path) <= 2 else data_path[2] @@ -234,16 +276,13 @@ def build_train_valid_datasets(data_path, image_size=224): root=train_data_path, transform=train_transform, classes_fraction=args.classes_fraction, - data_per_class_fraction=args.data_per_class_fraction + data_per_class_fraction=args.data_per_class_fraction, ) train_data = RandomSeedDataset(train_data) # validation dataset val_data_path = data_path[1] - val_data = ImageFolder( - root=val_data_path, - transform=val_transform - ) + val_data = ImageFolder(root=val_data_path, transform=val_transform) val_data = RandomSeedDataset(val_data) return train_data, val_data From 7af703b2e16697074fa2059247d3b48ebfe17a2b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:13:43 -0500 Subject: [PATCH 433/590] chore: Format `megatron/text_generation/*` --- megatron/text_generation/__init__.py | 5 +- megatron/text_generation/api.py | 249 +++++++++++++--------- megatron/text_generation/beam_utils.py | 10 +- megatron/text_generation/communication.py | 51 ++--- megatron/text_generation/forward_step.py | 126 ++++++----- megatron/text_generation/generation.py | 245 ++++++++++++++------- megatron/text_generation/sampling.py | 25 +-- megatron/text_generation/tokenization.py | 56 ++--- 8 files changed, 449 insertions(+), 318 deletions(-) diff --git a/megatron/text_generation/__init__.py b/megatron/text_generation/__init__.py index 77da7be30ae..73b9a41822f 100644 --- a/megatron/text_generation/__init__.py +++ b/megatron/text_generation/__init__.py @@ -1,7 +1,4 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. -from .api import ( - generate, - generate_and_post_process, - beam_search_and_post_process) +from .api import (generate, generate_and_post_process, beam_search_and_post_process) diff --git a/megatron/text_generation/api.py b/megatron/text_generation/api.py index 090b630a5f3..ea9676e49be 100644 --- a/megatron/text_generation/api.py +++ b/megatron/text_generation/api.py @@ -8,28 +8,30 @@ from megatron.core import mpu from .communication import broadcast_float_list from .generation import ( - generate_tokens_probs_and_return_on_first_stage, - score_and_return_on_first_stage, - beam_search_and_return_on_first_stage) -from .tokenization import ( - tokenize_prompts, - detokenize_generations) - -def generate_and_post_process(model, - prompts=None, - tokens_to_generate=0, - return_output_log_probs=False, - top_k_sampling=0, - top_p_sampling=0.0, - top_p_decay=0.0, - top_p_bound=0.0, - temperature=1.0, - add_BOS=False, - use_eod_token_for_early_termination=True, - stop_on_double_eol=False, - stop_on_eol=False, - prevent_newline_after_colon=False, - random_seed=-1): + generate_tokens_probs_and_return_on_first_stage, + score_and_return_on_first_stage, + beam_search_and_return_on_first_stage, +) +from .tokenization import (tokenize_prompts, detokenize_generations) + + +def generate_and_post_process( + model, + prompts=None, + tokens_to_generate=0, + return_output_log_probs=False, + top_k_sampling=0, + top_p_sampling=0.0, + top_p_decay=0.0, + top_p_bound=0.0, + temperature=1.0, + add_BOS=False, + use_eod_token_for_early_termination=True, + stop_on_double_eol=False, + stop_on_eol=False, + prevent_newline_after_colon=False, + random_seed=-1, +): """Run inference and post-process outputs, i.e., detokenize, move to cpu and convert to list.""" @@ -49,55 +51,73 @@ def generate_and_post_process(model, stop_on_double_eol=stop_on_double_eol, stop_on_eol=stop_on_eol, prevent_newline_after_colon=prevent_newline_after_colon, - random_seed=random_seed) + random_seed=random_seed, + ) # Only post-process on first stage. if mpu.is_pipeline_first_stage(): - tokens, prompts_plus_generations, prompts_plus_generations_segments = \ + tokens, prompts_plus_generations, prompts_plus_generations_segments = ( detokenize_generations(tokens, lengths, True) + ) if return_output_log_probs: output_log_probs = output_log_probs.cpu().numpy().tolist() - for i, (prob, seg) in enumerate(zip(output_log_probs, prompts_plus_generations_segments)): - output_log_probs[i] = prob[:len(seg)-1] - - return prompts_plus_generations, prompts_plus_generations_segments, \ - output_log_probs, tokens + for i, (prob, seg) in enumerate( + zip(output_log_probs, prompts_plus_generations_segments) + ): + output_log_probs[i] = prob[: len(seg) - 1] + + return ( + prompts_plus_generations, + prompts_plus_generations_segments, + output_log_probs, + tokens, + ) return None -def generate(model, - prompts=None, - tokens_to_generate=0, - return_output_log_probs=False, - top_k_sampling=0, - top_p_sampling=0.0, - top_p_decay=0.0, - top_p_bound=0.0, - temperature=1.0, - add_BOS=False, - use_eod_token_for_early_termination=True, - stop_on_double_eol=False, - stop_on_eol=False, - prevent_newline_after_colon=False, - random_seed=-1): + +def generate( + model, + prompts=None, + tokens_to_generate=0, + return_output_log_probs=False, + top_k_sampling=0, + top_p_sampling=0.0, + top_p_decay=0.0, + top_p_bound=0.0, + temperature=1.0, + add_BOS=False, + use_eod_token_for_early_termination=True, + stop_on_double_eol=False, + stop_on_eol=False, + prevent_newline_after_colon=False, + random_seed=-1, +): """Given prompts and input parameters, run inference and return: - tokens: prompts plus the generated tokens. - lengths: length of the prompt + generations. Note that we can - discard tokens in the tokens tensor that are after the - corresponding length. - output_log_probs: log probs of the tokens. + tokens: prompts plus the generated tokens. + lengths: length of the prompt + generations. Note that we can + discard tokens in the tokens tensor that are after the + corresponding length. + output_log_probs: log probs of the tokens. """ # Make sure input params are avaialble to all ranks. - values = [tokens_to_generate, - return_output_log_probs, - top_k_sampling, top_p_sampling, top_p_decay, top_p_bound, - temperature, add_BOS, use_eod_token_for_early_termination, - stop_on_double_eol, - stop_on_eol, - prevent_newline_after_colon, - random_seed] + values = [ + tokens_to_generate, + return_output_log_probs, + top_k_sampling, + top_p_sampling, + top_p_decay, + top_p_bound, + temperature, + add_BOS, + use_eod_token_for_early_termination, + stop_on_double_eol, + stop_on_eol, + prevent_newline_after_colon, + random_seed, + ] values_float_tensor = broadcast_float_list(len(values), float_list=values) tokens_to_generate = int(values_float_tensor[0].item()) return_output_log_probs = bool(values_float_tensor[1].item()) @@ -120,18 +140,22 @@ def generate(model, # Note that these tensors are broadcaseted to all ranks. if torch.distributed.get_rank() == 0: assert prompts is not None - + context_tokens_tensor, context_length_tensor = tokenize_prompts( - prompts=prompts, tokens_to_generate=tokens_to_generate, add_BOS=add_BOS) + prompts=prompts, tokens_to_generate=tokens_to_generate, add_BOS=add_BOS + ) if tokens_to_generate == 0: return score_and_return_on_first_stage( - model, context_tokens_tensor, context_length_tensor) - + model, context_tokens_tensor, context_length_tensor + ) + # Main inference function. # Note that the outputs are available on the first stage. return generate_tokens_probs_and_return_on_first_stage( - model, context_tokens_tensor, context_length_tensor, + model, + context_tokens_tensor, + context_length_tensor, return_output_log_probs=return_output_log_probs, top_k=top_k_sampling, top_p=top_p_sampling, @@ -141,48 +165,71 @@ def generate(model, use_eod_token_for_early_termination=use_eod_token_for_early_termination, stop_on_double_eol=stop_on_double_eol, stop_on_eol=stop_on_eol, - prevent_newline_after_colon=prevent_newline_after_colon) - -def beam_search_and_post_process(model, - prompts=None, - tokens_to_generate=0, - beam_size=0, - add_BOS=False, - stop_token=50256, - num_return_gen=1, - length_penalty=1, - prevent_newline_after_colon=False): + prevent_newline_after_colon=prevent_newline_after_colon, + ) + + +def beam_search_and_post_process( + model, + prompts=None, + tokens_to_generate=0, + beam_size=0, + add_BOS=False, + stop_token=50256, + num_return_gen=1, + length_penalty=1, + prevent_newline_after_colon=False, +): """Run beam search and post-process outputs, i.e., detokenize, move to cpu and convert to list.""" # Main inference. - tokens, scores = beam_search(model, - prompts=prompts, - tokens_to_generate=tokens_to_generate, - beam_size=beam_size, - add_BOS=add_BOS, - stop_token=stop_token, - num_return_gen=num_return_gen, - length_penalty=length_penalty, - prevent_newline_after_colon=prevent_newline_after_colon) + tokens, scores = beam_search( + model, + prompts=prompts, + tokens_to_generate=tokens_to_generate, + beam_size=beam_size, + add_BOS=add_BOS, + stop_token=stop_token, + num_return_gen=num_return_gen, + length_penalty=length_penalty, + prevent_newline_after_colon=prevent_newline_after_colon, + ) # Only post-process on first stage. if mpu.is_pipeline_first_stage(): - lengths = tokens.size(1)*torch.ones(beam_size, dtype=torch.int64, device=torch.cuda.current_device()) - tokens, prompts_plus_generations, prompts_plus_generations_segments = detokenize_generations(tokens, lengths, True) + lengths = tokens.size(1) * torch.ones( + beam_size, dtype=torch.int64, device=torch.cuda.current_device() + ) + tokens, prompts_plus_generations, prompts_plus_generations_segments = ( + detokenize_generations(tokens, lengths, True) + ) scores = scores.cpu().numpy().tolist() return prompts_plus_generations, prompts_plus_generations_segments, scores return None -def beam_search(model, prompts=None, tokens_to_generate=0, beam_size=0, add_BOS=False, stop_token=50256, num_return_gen=1, length_penalty=1, prevent_newline_after_colon=False): + +def beam_search( + model, + prompts=None, + tokens_to_generate=0, + beam_size=0, + add_BOS=False, + stop_token=50256, + num_return_gen=1, + length_penalty=1, + prevent_newline_after_colon=False, +): # Make sure input params are avaialble to all ranks. - values = [tokens_to_generate, - beam_size, - add_BOS, - stop_token, - num_return_gen, - length_penalty, - prevent_newline_after_colon] + values = [ + tokens_to_generate, + beam_size, + add_BOS, + stop_token, + num_return_gen, + length_penalty, + prevent_newline_after_colon, + ] values_float_tensor = broadcast_float_list(len(values), float_list=values) tokens_to_generate = int(values_float_tensor[0].item()) beam_size = int(values_float_tensor[1].item()) @@ -193,8 +240,16 @@ def beam_search(model, prompts=None, tokens_to_generate=0, beam_size=0, add_BOS= prevent_newline_after_colon = values_float_tensor[6].item() context_tokens_tensor, context_length_tensor = tokenize_prompts( - prompts=prompts, tokens_to_generate=tokens_to_generate, add_BOS=add_BOS) - - return beam_search_and_return_on_first_stage(model, context_tokens_tensor, context_length_tensor, - beam_size, stop_token=stop_token, num_return_gen=num_return_gen, length_penalty=length_penalty, - prevent_newline_after_colon=prevent_newline_after_colon) + prompts=prompts, tokens_to_generate=tokens_to_generate, add_BOS=add_BOS + ) + + return beam_search_and_return_on_first_stage( + model, + context_tokens_tensor, + context_length_tensor, + beam_size, + stop_token=stop_token, + num_return_gen=num_return_gen, + length_penalty=length_penalty, + prevent_newline_after_colon=prevent_newline_after_colon, + ) diff --git a/megatron/text_generation/beam_utils.py b/megatron/text_generation/beam_utils.py index 911a64143a8..85de99dfd9f 100644 --- a/megatron/text_generation/beam_utils.py +++ b/megatron/text_generation/beam_utils.py @@ -17,6 +17,7 @@ ## from huggingface beam search class BeamHypotheses(object): + def __init__(self, num_beams, length_penalty=1.0, early_stopping=False): """ Initialize n-best list of hypotheses. @@ -37,11 +38,13 @@ def add(self, hyp, sum_logprobs, length): """ Add a new hypothesis to the list. """ - score = sum_logprobs / length ** self.length_penalty + score = sum_logprobs / length**self.length_penalty if len(self) < self.num_beams or score > self.worst_score: self.beams.append((score, hyp)) if len(self) > self.num_beams: - sorted_scores = sorted([(s, idx) for idx, (s, _) in enumerate(self.beams)]) + sorted_scores = sorted( + [(s, idx) for idx, (s, _) in enumerate(self.beams)] + ) del self.beams[sorted_scores[0][1]] self.worst_score = sorted_scores[1][0] else: @@ -58,7 +61,6 @@ def is_done(self, best_sum_logprobs, cur_len): elif self.early_stopping: return True else: - cur_score = best_sum_logprobs / cur_len ** self.length_penalty + cur_score = best_sum_logprobs / cur_len**self.length_penalty ret = self.worst_score >= cur_score return ret - diff --git a/megatron/text_generation/communication.py b/megatron/text_generation/communication.py index dee32077f34..3479084f397 100644 --- a/megatron/text_generation/communication.py +++ b/megatron/text_generation/communication.py @@ -8,7 +8,6 @@ from megatron.core import mpu - # TODO: use functions from megatron/p2p def recv_from_prev_pipeline_rank_(recv_buffer=None): """Receive from previous pipeline stage and update the @@ -16,8 +15,10 @@ def recv_from_prev_pipeline_rank_(recv_buffer=None): if not mpu.is_pipeline_first_stage(): assert recv_buffer is not None recv_prev_op = torch.distributed.P2POp( - torch.distributed.irecv, recv_buffer, - mpu.get_pipeline_model_parallel_prev_rank()) + torch.distributed.irecv, + recv_buffer, + mpu.get_pipeline_model_parallel_prev_rank(), + ) reqs = torch.distributed.batch_isend_irecv([recv_prev_op]) for req in reqs: req.wait() @@ -25,15 +26,14 @@ def recv_from_prev_pipeline_rank_(recv_buffer=None): torch.cuda.synchronize() - # TODO: use functions from megatron/p2p def send_to_next_pipeline_rank(tensor=None): """Send output to the next pipeline stage.""" if not mpu.is_pipeline_last_stage(): assert tensor is not None send_next_op = torch.distributed.P2POp( - torch.distributed.isend, tensor, - mpu.get_pipeline_model_parallel_next_rank()) + torch.distributed.isend, tensor, mpu.get_pipeline_model_parallel_next_rank() + ) reqs = torch.distributed.batch_isend_irecv([send_next_op]) for req in reqs: req.wait() @@ -41,21 +41,18 @@ def send_to_next_pipeline_rank(tensor=None): torch.cuda.synchronize() - def _is_cuda(tensor): """Check if a tensor is not none and is cuda.""" assert tensor is not None assert tensor.is_cuda - def _is_cuda_contiguous(tensor): """Check if a tensor is not none, is cuda, and is contiguous.""" _is_cuda(tensor) assert tensor.is_contiguous() - def broadcast_from_last_pipeline_stage(size, dtype, tensor=None): """Broadcast a tensor from last pipeline stage to all ranks.""" @@ -68,9 +65,7 @@ def broadcast_from_last_pipeline_stage(size, dtype, tensor=None): if is_last_stage: _is_cuda_contiguous(tensor) else: - tensor = torch.empty(size, - dtype=dtype, - device=torch.cuda.current_device()) + tensor = torch.empty(size, dtype=dtype, device=torch.cuda.current_device()) # Get the group and corresponding source rank. src = mpu.get_pipeline_model_parallel_last_rank() group = mpu.get_pipeline_model_parallel_group() @@ -79,7 +74,6 @@ def broadcast_from_last_pipeline_stage(size, dtype, tensor=None): return tensor - def broadcast_from_last_to_first_pipeline_stage(size, dtype, tensor=None): """Broadcast tensor values from last stage into the first stage.""" @@ -94,9 +88,7 @@ def broadcast_from_last_to_first_pipeline_stage(size, dtype, tensor=None): if is_last_stage: _is_cuda_contiguous(tensor) else: - tensor = torch.empty(size, - dtype=dtype, - device=torch.cuda.current_device()) + tensor = torch.empty(size, dtype=dtype, device=torch.cuda.current_device()) src = mpu.get_pipeline_model_parallel_last_rank() group = mpu.get_embedding_group() # Broadcast from last stage into the first stage. @@ -107,7 +99,6 @@ def broadcast_from_last_to_first_pipeline_stage(size, dtype, tensor=None): return tensor - def copy_from_last_to_first_pipeline_stage(size, dtype, tensor=None): """Copy tensor values from last stage into the first stage. Note that the input tensor is updated in place.""" @@ -130,9 +121,9 @@ def copy_from_last_to_first_pipeline_stage(size, dtype, tensor=None): if is_last_stage: tensor_ = tensor.contiguous() else: - tensor_ = torch.empty(size, - dtype=dtype, - device=torch.cuda.current_device()) + tensor_ = torch.empty( + size, dtype=dtype, device=torch.cuda.current_device() + ) # Broadcast from last stage into the first stage. torch.distributed.broadcast(tensor_, src, group) # Update the first stage tensor @@ -140,46 +131,40 @@ def copy_from_last_to_first_pipeline_stage(size, dtype, tensor=None): tensor[...] = tensor_ - def broadcast_tensor(size, dtype, tensor=None, rank=0): - """ Given size and type of a tensor on all ranks and the tensor value - only on a specific rank, broadcast from that rank to all other ranks. + """Given size and type of a tensor on all ranks and the tensor value + only on a specific rank, broadcast from that rank to all other ranks. """ if torch.distributed.get_rank() == rank: _is_cuda_contiguous(tensor) else: - tensor = torch.empty(size, - dtype=dtype, - device=torch.cuda.current_device()) + tensor = torch.empty(size, dtype=dtype, device=torch.cuda.current_device()) torch.distributed.broadcast(tensor, rank) return tensor - def broadcast_list(size, dtype, list_values=None, rank=0): """Broadcast a list of values with a given type.""" tensor = None if torch.distributed.get_rank() == rank: - tensor = torch.tensor(list_values, dtype=dtype, - device=torch.cuda.current_device()) + tensor = torch.tensor( + list_values, dtype=dtype, device=torch.cuda.current_device() + ) return broadcast_tensor(size, dtype, tensor=tensor, rank=rank) - def broadcast_int_list(size, int_list=None, rank=0): """Broadcast a list of interger values.""" return broadcast_list(size, torch.int64, list_values=int_list, rank=rank) - def broadcast_float_list(size, float_list=None, rank=0): """Broadcast a list of float values.""" - return broadcast_list(size, torch.float32, list_values=float_list, - rank=rank) + return broadcast_list(size, torch.float32, list_values=float_list, rank=rank) diff --git a/megatron/text_generation/forward_step.py b/megatron/text_generation/forward_step.py index feb087cbb61..43fc45f4a7b 100644 --- a/megatron/text_generation/forward_step.py +++ b/megatron/text_generation/forward_step.py @@ -8,10 +8,7 @@ from megatron import get_args from megatron.core import mpu -from .communication import ( - send_to_next_pipeline_rank, - recv_from_prev_pipeline_rank_) - +from .communication import (send_to_next_pipeline_rank, recv_from_prev_pipeline_rank_) class InferenceParams: @@ -32,14 +29,21 @@ def swap_key_value_dict(self, batch_idx): "swap between batches" if len(self.key_value_memory_dict) == 0: raise ValueError("should not swap when dict in empty") - + for layer_number in self.key_value_memory_dict.keys(): - inference_key_memory, inference_value_memory = self.key_value_memory_dict[layer_number] - assert len(batch_idx) == inference_key_memory.shape[1] ## make sure batch size is the same + inference_key_memory, inference_value_memory = self.key_value_memory_dict[ + layer_number + ] + assert ( + len(batch_idx) == inference_key_memory.shape[1] + ) ## make sure batch size is the same new_inference_key_memory = inference_key_memory[:, batch_idx] new_inference_value_memory = inference_value_memory[:, batch_idx] self.key_value_memory_dict[layer_number] = ( - new_inference_key_memory, new_inference_value_memory) + new_inference_key_memory, + new_inference_value_memory, + ) + class ForwardStep: """Forward step function with all the communications. @@ -49,21 +53,18 @@ class ForwardStep: def __init__(self, model, max_batch_size, max_sequence_len): """Set values so we don't need to do it multiple times.""" # Make sure model is in eval mode. - assert not isinstance(model, Iterable), \ - 'interleaving schedule is not supported for inference' + assert not isinstance( + model, Iterable + ), "interleaving schedule is not supported for inference" model.eval() self.model = model # Initialize inference parameters. - self.inference_params = InferenceParams(max_batch_size, - max_sequence_len) + self.inference_params = InferenceParams(max_batch_size, max_sequence_len) # Pipelining arguments. args = get_args() - self.pipeline_size_larger_than_one = ( - args.pipeline_model_parallel_size > 1) + self.pipeline_size_larger_than_one = args.pipeline_model_parallel_size > 1 # Threshold of pipelining. - self.pipelining_batch_x_seqlen = \ - args.inference_batch_times_seqlen_threshold - + self.pipelining_batch_x_seqlen = args.inference_batch_times_seqlen_threshold def __call__(self, tokens, position_ids, attention_mask): """Invocation of the forward methods. Note that self.inference_params @@ -72,21 +73,21 @@ def __call__(self, tokens, position_ids, attention_mask): if self.pipeline_size_larger_than_one: current_batch_x_seqlen = tokens.size(0) * tokens.size(1) if current_batch_x_seqlen >= self.pipelining_batch_x_seqlen: - micro_batch_size = \ - max(1, self.pipelining_batch_x_seqlen // tokens.size(1)) - return _with_pipelining_forward_step(self.model, - tokens, - position_ids, - attention_mask, - self.inference_params, - micro_batch_size) - - return _no_pipelining_forward_step(self.model, - tokens, - position_ids, - attention_mask, - self.inference_params) - + micro_batch_size = max( + 1, self.pipelining_batch_x_seqlen // tokens.size(1) + ) + return _with_pipelining_forward_step( + self.model, + tokens, + position_ids, + attention_mask, + self.inference_params, + micro_batch_size, + ) + + return _no_pipelining_forward_step( + self.model, tokens, position_ids, attention_mask, self.inference_params + ) def _get_recv_buffer_dtype(args): @@ -96,21 +97,22 @@ def _get_recv_buffer_dtype(args): return args.params_dtype - def _allocate_recv_buffer(batch_size, sequence_length): """Receive happens between the layers with size [s, b, h].""" if mpu.is_pipeline_first_stage(): return None args = get_args() recv_size = (sequence_length, batch_size, args.hidden_size) - return torch.empty(recv_size, - dtype=_get_recv_buffer_dtype(args), - device=torch.cuda.current_device()) - + return torch.empty( + recv_size, + dtype=_get_recv_buffer_dtype(args), + device=torch.cuda.current_device(), + ) -def _forward_step_helper(model, tokens, position_ids, attention_mask, - inference_params, recv_buffer=None): +def _forward_step_helper( + model, tokens, position_ids, attention_mask, inference_params, recv_buffer=None +): """Single forward step. Update the allocate memory flag so only the first time the memory is allocated.""" batch_size = tokens.size(0) @@ -123,8 +125,9 @@ def _forward_step_helper(model, tokens, position_ids, attention_mask, # Forward pass through the model. model.set_input_tensor(recv_buffer) - output_tensor = model(tokens, position_ids, attention_mask, - inference_params=inference_params) + output_tensor = model( + tokens, position_ids, attention_mask, inference_params=inference_params + ) # Send output to the next stage. send_to_next_pipeline_rank(output_tensor) @@ -132,14 +135,19 @@ def _forward_step_helper(model, tokens, position_ids, attention_mask, return output_tensor - -def _no_pipelining_forward_step(model, tokens, position_ids, attention_mask, - inference_params, recv_buffer=None): +def _no_pipelining_forward_step( + model, tokens, position_ids, attention_mask, inference_params, recv_buffer=None +): """If recv_buffer is none, we will allocate one on the fly.""" # Run a simple forward pass. - output_tensor = _forward_step_helper(model, tokens, position_ids, - attention_mask, inference_params, - recv_buffer=recv_buffer) + output_tensor = _forward_step_helper( + model, + tokens, + position_ids, + attention_mask, + inference_params, + recv_buffer=recv_buffer, + ) # Update the sequence length offset. inference_params.sequence_len_offset += tokens.size(1) @@ -150,16 +158,15 @@ def _no_pipelining_forward_step(model, tokens, position_ids, attention_mask, return logits - -def _with_pipelining_forward_step(model, tokens, position_ids, attention_mask, - inference_params, micro_batch_size): +def _with_pipelining_forward_step( + model, tokens, position_ids, attention_mask, inference_params, micro_batch_size +): """No interleaving is supported.""" sequence_length = tokens.size(1) batch_size = tokens.size(0) # Divide the batch dimension into micro batches. - num_micro_batches, last_chunk = divmod(batch_size, - micro_batch_size) + num_micro_batches, last_chunk = divmod(batch_size, micro_batch_size) if last_chunk > 0: num_micro_batches += 1 @@ -169,7 +176,9 @@ def _with_pipelining_forward_step(model, tokens, position_ids, attention_mask, args = get_args() logits = torch.empty( (batch_size, sequence_length, args.padded_vocab_size), - dtype=torch.float32, device=torch.cuda.current_device()) + dtype=torch.float32, + device=torch.cuda.current_device(), + ) # Preallocate recv buffer. recv_buffer = _allocate_recv_buffer(micro_batch_size, sequence_length) @@ -185,9 +194,14 @@ def _with_pipelining_forward_step(model, tokens, position_ids, attention_mask, # Run a simple forward pass. if this_micro_batch_size != micro_batch_size: recv_buffer = None - output = _forward_step_helper(model, tokens2use, position_ids2use, - attention_mask, inference_params, - recv_buffer=recv_buffer) + output = _forward_step_helper( + model, + tokens2use, + position_ids2use, + attention_mask, + inference_params, + recv_buffer=recv_buffer, + ) # Adjust the batch size offset to account for the micro-batch. inference_params.batch_size_offset += this_micro_batch_size diff --git a/megatron/text_generation/generation.py b/megatron/text_generation/generation.py index 79c8b4d1af1..4a0ee1a2118 100644 --- a/megatron/text_generation/generation.py +++ b/megatron/text_generation/generation.py @@ -11,11 +11,13 @@ from .communication import ( copy_from_last_to_first_pipeline_stage, broadcast_from_last_pipeline_stage, - broadcast_from_last_to_first_pipeline_stage) + broadcast_from_last_to_first_pipeline_stage, +) from .forward_step import ForwardStep from .sampling import sample from .beam_utils import BeamHypotheses + def score_and_return_on_first_stage(model, tokens, lengths): """Function for just scoring. Arguments: @@ -38,7 +40,12 @@ def score_and_return_on_first_stage(model, tokens, lengths): raise ValueError("Length of prompt + tokens_to_generate longer than allowed") if max_prompt_length * batch_size > args.max_tokens_to_oom: - raise ValueError("Too many tokens. " + str(max_prompt_length*batch_size)+ " is greater than "+str(args.max_tokens_to_oom)) + raise ValueError( + "Too many tokens. " + + str(max_prompt_length * batch_size) + + " is greater than " + + str(args.max_tokens_to_oom) + ) # forward step. forward_step = ForwardStep(model, batch_size, max_prompt_length) @@ -52,9 +59,11 @@ def score_and_return_on_first_stage(model, tokens, lengths): output_log_probs_size = (batch_size, max_prompt_length - 1) if mpu.is_pipeline_last_stage(): - output_log_probs = torch.empty(output_log_probs_size, - dtype=torch.float32, - device=torch.cuda.current_device()) + output_log_probs = torch.empty( + output_log_probs_size, + dtype=torch.float32, + device=torch.cuda.current_device(), + ) # ============= # Run infernece @@ -81,20 +90,27 @@ def score_and_return_on_first_stage(model, tokens, lengths): # Broadcast to the first pipeline stage. # ====================================== output_log_probs = broadcast_from_last_to_first_pipeline_stage( - output_log_probs_size, torch.float32, output_log_probs) + output_log_probs_size, torch.float32, output_log_probs + ) return tokens, lengths, output_log_probs + def generate_tokens_probs_and_return_on_first_stage( - model, tokens, lengths, - return_output_log_probs=False, - top_k=0, top_p=0.0, top_p_decay=0.0, top_p_bound=0.0, - temperature=1.0, - use_eod_token_for_early_termination=True, - stop_on_double_eol=False, - stop_on_eol=False, - prevent_newline_after_colon=True - ): + model, + tokens, + lengths, + return_output_log_probs=False, + top_k=0, + top_p=0.0, + top_p_decay=0.0, + top_p_bound=0.0, + temperature=1.0, + use_eod_token_for_early_termination=True, + stop_on_double_eol=False, + stop_on_eol=False, + prevent_newline_after_colon=True, +): """Main token generation function. Arguments: model: no interleaving is supported. @@ -133,14 +149,19 @@ def generate_tokens_probs_and_return_on_first_stage( raise ValueError("Length of prompt + tokens_to_generate longer than allowed") if max_sequence_length * batch_size > args.max_tokens_to_oom: - raise ValueError("Too many tokens. " + str(max_sequence_length*batch_size)+ " is greater than "+str(args.max_tokens_to_oom)) + raise ValueError( + "Too many tokens. " + + str(max_sequence_length * batch_size) + + " is greater than " + + str(args.max_tokens_to_oom) + ) # forward step. forward_step = ForwardStep(model, batch_size, max_sequence_length) # Added termination_id to support the case that we want to terminate the # generation once that id is generated. - if hasattr(args, 'eos_id'): + if hasattr(args, "eos_id"): termination_id = args.eos_id else: termination_id = tokenizer.eod @@ -156,24 +177,29 @@ def generate_tokens_probs_and_return_on_first_stage( generated_sequence_lengths = None if mpu.is_pipeline_last_stage(): if return_output_log_probs: - output_log_probs = torch.empty(output_log_probs_size, - dtype=torch.float32, - device=torch.cuda.current_device()) - generated_sequence_lengths = torch.ones( - batch_size, dtype=torch.int64, - device=torch.cuda.current_device()) * max_sequence_length + output_log_probs = torch.empty( + output_log_probs_size, + dtype=torch.float32, + device=torch.cuda.current_device(), + ) + generated_sequence_lengths = ( + torch.ones( + batch_size, dtype=torch.int64, device=torch.cuda.current_device() + ) + * max_sequence_length + ) # Whether we have reached a termination id. - is_generation_done = torch.zeros(batch_size, dtype=torch.uint8, - device=torch.cuda.current_device()) + is_generation_done = torch.zeros( + batch_size, dtype=torch.uint8, device=torch.cuda.current_device() + ) # ============= # Run infernece # ============= with torch.no_grad(): - attention_mask, position_ids = _build_attention_mask_and_position_ids( - tokens) + attention_mask, position_ids = _build_attention_mask_and_position_ids(tokens) prev_context_length = 0 for context_length in range(min_prompt_length, max_sequence_length): @@ -181,7 +207,8 @@ def generate_tokens_probs_and_return_on_first_stage( tokens2use = tokens[:, prev_context_length:context_length] positions2use = position_ids[:, prev_context_length:context_length] attention_mask2use = attention_mask[ - ..., prev_context_length:context_length, :context_length] + ..., prev_context_length:context_length, :context_length + ] # logits will be meanigful only in the last pipeline stage. logits = forward_step(tokens2use, positions2use, attention_mask2use) @@ -189,17 +216,23 @@ def generate_tokens_probs_and_return_on_first_stage( if mpu.is_pipeline_last_stage(): if prevent_newline_after_colon: - logits[tokens2use[:, -1] == tokenizer.tokenize(':')[0], -1, tokenizer.tokenize('\n')[0]] = -1e10 # disable "\n" after ":" + logits[ + tokens2use[:, -1] == tokenizer.tokenize(":")[0], + -1, + tokenizer.tokenize("\n")[0], + ] = -1e10 # disable "\n" after ":" # Always the last stage should have an output. assert logits is not None # Sample. last_token_logits = logits[:, -1, :] - new_sample = sample(last_token_logits, - top_k=top_k, - top_p=top_p, - temperature=temperature, - vocab_size=tokenizer.vocab_size) + new_sample = sample( + last_token_logits, + top_k=top_k, + top_p=top_p, + temperature=temperature, + vocab_size=tokenizer.vocab_size, + ) if top_p > 0.0 and top_p_decay > 0.0: top_p = top_p * top_p_decay if top_p_bound > 0.0: @@ -220,18 +253,18 @@ def generate_tokens_probs_and_return_on_first_stage( # the token which we selected in the current logits, # so shift by 1. indices = torch.unsqueeze( - tokens[ - :, - (prev_context_length + 1):(context_length + 1)], - 2) - output_log_probs[:, - prev_context_length:context_length] = \ + tokens[:, (prev_context_length + 1) : (context_length + 1)], + 2, + ) + output_log_probs[:, prev_context_length:context_length] = ( torch.gather(log_probs, 2, indices).squeeze(2) + ) # Update the tokens on the first stage so the next input to # the network is correct. - copy_from_last_to_first_pipeline_stage(batch_size, torch.int64, - tokens[:, context_length]) + copy_from_last_to_first_pipeline_stage( + batch_size, torch.int64, tokens[:, context_length] + ) # Update the context length for the next token generation. prev_context_length = context_length @@ -243,23 +276,24 @@ def generate_tokens_probs_and_return_on_first_stage( # instead tokenization should be in the inference loop so stop sequences can be used if stop_on_double_eol: hit_double_eol = (new_sample == 628).byte() & started.byte() - hit_two_eols = (new_sample == 198).byte() & (tokens[:, context_length-1] == 198).byte() & started.byte() + hit_two_eols = ( + (new_sample == 198).byte() + & (tokens[:, context_length - 1] == 198).byte() + & started.byte() + ) done_token = hit_double_eol | hit_two_eols elif stop_on_eol: hit_double_eol = (new_sample == 628).byte() & started.byte() hit_eol = (new_sample == 198).byte() & started.byte() done_token = hit_double_eol | hit_eol else: - done_token = (new_sample == termination_id).byte() & \ - started.byte() + done_token = (new_sample == termination_id).byte() & started.byte() just_finished = (done_token & ~is_generation_done).bool() - generated_sequence_lengths[just_finished.view(-1)] = \ - context_length + 1 + generated_sequence_lengths[just_finished.view(-1)] = context_length + 1 is_generation_done = is_generation_done | done_token done = torch.all(is_generation_done) - done = broadcast_from_last_pipeline_stage(1, torch.uint8, - tensor=done) + done = broadcast_from_last_pipeline_stage(1, torch.uint8, tensor=done) if use_eod_token_for_early_termination and done: break @@ -267,7 +301,7 @@ def generate_tokens_probs_and_return_on_first_stage( # Update the length of based on max generated length. # =================================================== - tokens = tokens[:, :(context_length + 1)] + tokens = tokens[:, : (context_length + 1)] if mpu.is_pipeline_last_stage(): if return_output_log_probs: output_log_probs = output_log_probs[:, :context_length] @@ -277,20 +311,32 @@ def generate_tokens_probs_and_return_on_first_stage( # ====================================== generated_sequence_lengths = broadcast_from_last_to_first_pipeline_stage( - batch_size, torch.int64, generated_sequence_lengths) + batch_size, torch.int64, generated_sequence_lengths + ) if return_output_log_probs: output_log_probs_size = (batch_size, context_length) output_log_probs = broadcast_from_last_to_first_pipeline_stage( - output_log_probs_size, torch.float32, output_log_probs) + output_log_probs_size, torch.float32, output_log_probs + ) return tokens, generated_sequence_lengths, output_log_probs -def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, stop_token, num_return_gen, length_penalty, prevent_newline_after_colon=True): + +def beam_search_and_return_on_first_stage( + model, + tokens, + lengths, + beam_size, + stop_token, + num_return_gen, + length_penalty, + prevent_newline_after_colon=True, +): args = get_args() tokenizer = get_tokenizer() batch_size = tokens.size(0) - assert(batch_size == 1) + assert batch_size == 1 prompt_length = lengths.item() final_sequence_length = tokens.size(1) final_sequence_length = min(final_sequence_length, args.max_position_embeddings) @@ -305,9 +351,9 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto beam_hyp = BeamHypotheses(beam_size, length_penalty) best_batches = None done = torch.zeros(1, dtype=torch.uint8, device=torch.cuda.current_device()) - scores = torch.zeros(beam_size, - dtype=torch.float32, - device=torch.cuda.current_device()).unsqueeze(1) + scores = torch.zeros( + beam_size, dtype=torch.float32, device=torch.cuda.current_device() + ).unsqueeze(1) scores_size_tensor, tokens_size_tensor = None, None # ============= # Run infernece @@ -322,25 +368,36 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto tokens2use = tokens[:, prev_context_length:context_length] positions2use = position_ids[:, prev_context_length:context_length] attention_mask2use = attention_mask[ - ..., prev_context_length:context_length, :context_length] + ..., prev_context_length:context_length, :context_length + ] # logits will be meanigful only in the last pipeline stage. logits = forward_step(tokens2use, positions2use, attention_mask2use) if mpu.is_pipeline_last_stage(): if prevent_newline_after_colon: - logits[tokens2use[:, -1] == tokenizer.tokenize(':')[0], -1, tokenizer.tokenize('\n')[0]] = -1e10 # disable "\n" after ":" + logits[ + tokens2use[:, -1] == tokenizer.tokenize(":")[0], + -1, + tokenizer.tokenize("\n")[0], + ] = -1e10 # disable "\n" after ":" vocab_size = logits.size(2) log_probs = F.log_softmax(logits, dim=2) new_scores = log_probs[:, -1, :] + scores if context_length == prompt_length: # if this is the first one - sorted_scores, indices = torch.sort(new_scores[0,:], descending=True) + sorted_scores, indices = torch.sort( + new_scores[0, :], descending=True + ) else: - sorted_scores, indices = torch.sort(new_scores.view(-1), descending=True) - - best_beam_ids = torch.div(indices[: 2 * beam_size], vocab_size).trunc().long() - best_words = indices[:2 * beam_size] % vocab_size + sorted_scores, indices = torch.sort( + new_scores.view(-1), descending=True + ) + + best_beam_ids = ( + torch.div(indices[: 2 * beam_size], vocab_size).trunc().long() + ) + best_words = indices[: 2 * beam_size] % vocab_size best_scores = sorted_scores[: 2 * beam_size] next_beams = [] @@ -349,13 +406,15 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto ): if token_id.item() == stop_token: # if beam_token does not belong to top num_beams tokens, it should not be added - is_beam_token_worse_than_top_num_beams = beam_token_rank >= beam_size + is_beam_token_worse_than_top_num_beams = ( + beam_token_rank >= beam_size + ) if is_beam_token_worse_than_top_num_beams: continue beam_hyp.add( tokens[beam_id].clone(), beam_score, - context_length + 1 - prompt_length + context_length + 1 - prompt_length, ) else: # add next predicted token since it is not eos_token @@ -364,11 +423,15 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto if len(next_beams) == beam_size: break - if beam_hyp.is_done(best_scores.max().item(), context_length + 1 - prompt_length): - done = torch.ones(1, dtype=torch.uint8, device=torch.cuda.current_device()) + if beam_hyp.is_done( + best_scores.max().item(), context_length + 1 - prompt_length + ): + done = torch.ones( + 1, dtype=torch.uint8, device=torch.cuda.current_device() + ) best_batches = tokens.new([item[2] for item in next_beams]) - tokens = tokens[best_batches,:] + tokens = tokens[best_batches, :] tokens[:, context_length] = tokens.new([item[0] for item in next_beams]) scores = scores.new([item[1] for item in next_beams]).unsqueeze(1) @@ -379,11 +442,12 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto # Update the tokens on the first stage so the next input to # the network is correct. - copy_from_last_to_first_pipeline_stage(tokens.size(), torch.int64, - tokens) + copy_from_last_to_first_pipeline_stage(tokens.size(), torch.int64, tokens) # set inference key values to make it consistent with best beam index - best_batches = broadcast_from_last_pipeline_stage(beam_size, torch.int64, best_batches) + best_batches = broadcast_from_last_pipeline_stage( + beam_size, torch.int64, best_batches + ) forward_step.inference_params.swap_key_value_dict(best_batches) # Update the context length for the next token generation. @@ -393,7 +457,11 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto # if cannot find stop token, add open beams to hyps if not done: for beam_id in range(beam_size): - beam_hyp.add(tokens[beam_id].clone(), scores[beam_id].squeeze(), context_length + 1 - prompt_length) + beam_hyp.add( + tokens[beam_id].clone(), + scores[beam_id].squeeze(), + context_length + 1 - prompt_length, + ) # rank based on scores sorted_hyps = sorted(beam_hyp.beams, key=lambda x: x[0], reverse=True) @@ -402,14 +470,26 @@ def beam_search_and_return_on_first_stage(model, tokens, lengths, beam_size, sto tokens = [sorted_hyps[i][1] for i in range(num_return_gen)] scores = torch.stack(scores, dim=0) tokens = torch.stack(tokens, dim=0) - scores_size_tensor = torch.tensor(scores.shape, dtype=torch.int64, device=torch.cuda.current_device()) - tokens_size_tensor = torch.tensor(tokens.shape, dtype=torch.int64, device=torch.cuda.current_device()) - - scores_size_tensor = broadcast_from_last_pipeline_stage(1, torch.int64, scores_size_tensor) - tokens_size_tensor = broadcast_from_last_pipeline_stage(2, torch.int64, tokens_size_tensor) - - scores = broadcast_from_last_to_first_pipeline_stage(tuple(scores_size_tensor), torch.float32, scores) - tokens = broadcast_from_last_to_first_pipeline_stage(tuple(tokens_size_tensor), torch.int64, tokens) + scores_size_tensor = torch.tensor( + scores.shape, dtype=torch.int64, device=torch.cuda.current_device() + ) + tokens_size_tensor = torch.tensor( + tokens.shape, dtype=torch.int64, device=torch.cuda.current_device() + ) + + scores_size_tensor = broadcast_from_last_pipeline_stage( + 1, torch.int64, scores_size_tensor + ) + tokens_size_tensor = broadcast_from_last_pipeline_stage( + 2, torch.int64, tokens_size_tensor + ) + + scores = broadcast_from_last_to_first_pipeline_stage( + tuple(scores_size_tensor), torch.float32, scores + ) + tokens = broadcast_from_last_to_first_pipeline_stage( + tuple(tokens_size_tensor), torch.int64, tokens + ) return tokens, scores @@ -424,6 +504,7 @@ def _build_attention_mask_and_position_ids(tokens): eod_token=None, reset_position_ids=False, reset_attention_mask=False, - eod_mask_loss=False) + eod_mask_loss=False, + ) return attention_mask, position_ids diff --git a/megatron/text_generation/sampling.py b/megatron/text_generation/sampling.py index 370773a36c0..44b5ce5bb8b 100644 --- a/megatron/text_generation/sampling.py +++ b/megatron/text_generation/sampling.py @@ -10,13 +10,11 @@ import torch - def modify_logits_for_top_k_filtering(logits, top_k): """Set the logits for none top-k values to -inf.""" filter_ = logits < torch.topk(logits, top_k)[0][..., -1, None] - logits.masked_fill_(filter_, float('-Inf')) - + logits.masked_fill_(filter_, float("-Inf")) def modify_logits_for_top_p_filtering(logits, top_p): @@ -38,12 +36,11 @@ def modify_logits_for_top_p_filtering(logits, top_p): # Fill in the filtered part filter_ = filter_.scatter(1, sorted_indices, filter_) - logits.masked_fill_(filter_, float('-Inf')) - + logits.masked_fill_(filter_, float("-Inf")) def sample(logits, top_k=0, top_p=0.0, temperature=1.0, vocab_size=None): - """ Sample and generate a token. + """Sample and generate a token. Note: logits has the dimension [b, v] where b is the batch size and v is the vocabulary size. If vocab_size is provided, we will make sure the sample that is @@ -52,14 +49,12 @@ def sample(logits, top_k=0, top_p=0.0, temperature=1.0, vocab_size=None): """ # Check logits for consistency. - assert logits.ndim == 2, 'expected the logits to be of [b, v] shape.' - assert logits.type() == 'torch.cuda.FloatTensor', \ - 'input logits should be floats.' - + assert logits.ndim == 2, "expected the logits to be of [b, v] shape." + assert logits.type() == "torch.cuda.FloatTensor", "input logits should be floats." # Greedy is just simple argmax. if top_k == 1: - assert top_p == 0.0, 'cannot set both greedy and top-p samplings.' + assert top_p == 0.0, "cannot set both greedy and top-p samplings." samples = torch.argmax(logits, dim=-1) # Top-k or top-p sampling. @@ -71,14 +66,14 @@ def sample(logits, top_k=0, top_p=0.0, temperature=1.0, vocab_size=None): logits.div_(temperature) if top_k > 1: - assert top_p == 0.0, 'cannot set both top-k and top-p samplings.' - assert top_k <= logits.size(1), 'top-k is larger than logit size.' + assert top_p == 0.0, "cannot set both top-k and top-p samplings." + assert top_k <= logits.size(1), "top-k is larger than logit size." if vocab_size: - assert top_k < vocab_size, 'top-k is larger than vocab size.' + assert top_k < vocab_size, "top-k is larger than vocab size." modify_logits_for_top_k_filtering(logits, top_k) elif top_p > 0.0: - assert top_p <= 1.0, 'top-p should be in (0, 1].' + assert top_p <= 1.0, "top-p should be in (0, 1]." modify_logits_for_top_p_filtering(logits, top_p) # After filtering, we need to recalculate the distribution. diff --git a/megatron/text_generation/tokenization.py b/megatron/text_generation/tokenization.py index accead319a1..4b4e5c9a1f0 100644 --- a/megatron/text_generation/tokenization.py +++ b/megatron/text_generation/tokenization.py @@ -10,9 +10,7 @@ from .communication import broadcast_int_list, broadcast_tensor -def detokenize_generations(tokens_gpu_tensor, - lengths_gpu_tensor, - return_segments): +def detokenize_generations(tokens_gpu_tensor, lengths_gpu_tensor, return_segments): """Detokenize the generated tokens.""" tokenizer = get_tokenizer() @@ -25,33 +23,32 @@ def detokenize_generations(tokens_gpu_tensor, lengths = lengths_gpu_tensor.cpu().numpy().tolist() for sequence_tokens, length in zip(tokens, lengths): sequence_tokens = sequence_tokens[:length] - prompts_plus_generations.append( - tokenizer.detokenize(sequence_tokens)) + prompts_plus_generations.append(tokenizer.detokenize(sequence_tokens)) if return_segments: words = [] for token in sequence_tokens: - if args.tokenizer_type in ['SentencePieceTokenizer', - 'GPTSentencePieceTokenizer']: + if args.tokenizer_type in [ + "SentencePieceTokenizer", + "GPTSentencePieceTokenizer", + ]: word = tokenizer.decoder[token] - elif args.tokenizer_type == 'NullTokenizer': + elif args.tokenizer_type == "NullTokenizer": word = str(token) else: word = tokenizer.tokenizer.decoder[token] word = bytearray( - [tokenizer.tokenizer.byte_decoder[c] for c in word]).decode( - 'utf-8', errors='replace') + [tokenizer.tokenizer.byte_decoder[c] for c in word] + ).decode("utf-8", errors="replace") words.append(word) prompts_plus_generations_segments.append(words) if return_segments: - return tokens, prompts_plus_generations, \ - prompts_plus_generations_segments + return tokens, prompts_plus_generations, prompts_plus_generations_segments return tokens, prompts_plus_generations -def tokenize_prompts(prompts=None, tokens_to_generate=None, - add_BOS=None, rank=0): +def tokenize_prompts(prompts=None, tokens_to_generate=None, add_BOS=None, rank=0): """Tokenize prompts and make them avaiable on all ranks.""" # On all ranks set to None so we can pass them to functions @@ -64,11 +61,14 @@ def tokenize_prompts(prompts=None, tokens_to_generate=None, assert prompts is not None assert tokens_to_generate is not None # Tensor of tokens padded and their unpadded length. - prompts_tokens_cuda_long_tensor, prompts_length_cuda_long_tensor = \ + prompts_tokens_cuda_long_tensor, prompts_length_cuda_long_tensor = ( _tokenize_prompts_and_batch(prompts, tokens_to_generate, add_BOS) + ) # We need the sizes of these tensors for the boradcast - sizes_list = [prompts_tokens_cuda_long_tensor.size(0), # Batch size - prompts_tokens_cuda_long_tensor.size(1)] # Sequence lenght + sizes_list = [ + prompts_tokens_cuda_long_tensor.size(0), # Batch size + prompts_tokens_cuda_long_tensor.size(1), + ] # Sequence lenght # First, broadcast the sizes. sizes_tensor = broadcast_int_list(2, int_list=sizes_list, rank=rank) @@ -77,28 +77,30 @@ def tokenize_prompts(prompts=None, tokens_to_generate=None, # and length tensors. sizes = sizes_tensor.tolist() prompts_tokens_cuda_long_tensor = broadcast_tensor( - sizes, torch.int64, tensor=prompts_tokens_cuda_long_tensor, rank=rank) + sizes, torch.int64, tensor=prompts_tokens_cuda_long_tensor, rank=rank + ) prompts_length_cuda_long_tensor = broadcast_tensor( - sizes[0], torch.int64, tensor=prompts_length_cuda_long_tensor, - rank=rank) + sizes[0], torch.int64, tensor=prompts_length_cuda_long_tensor, rank=rank + ) return prompts_tokens_cuda_long_tensor, prompts_length_cuda_long_tensor def _tokenize_prompts_and_batch(prompts, tokens_to_generate, add_BOS): """Given a set of prompts and number of tokens to generate: - - tokenize prompts - - set the sequence length to be the max of length of prompts - plus the number of tokens we would like to generate - - pad all the sequences to this length so we can convert them - into a 2D tensor. + - tokenize prompts + - set the sequence length to be the max of length of prompts + plus the number of tokens we would like to generate + - pad all the sequences to this length so we can convert them + into a 2D tensor. """ # Tokenize all the prompts. tokenizer = get_tokenizer() if add_BOS: - prompts_tokens = [[tokenizer.eod] + tokenizer.tokenize(prompt) - for prompt in prompts] + prompts_tokens = [ + [tokenizer.eod] + tokenizer.tokenize(prompt) for prompt in prompts + ] else: prompts_tokens = [tokenizer.tokenize(prompt) for prompt in prompts] From 103467dc8d3f7be4bb742195cc7b5a35d8d4777c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:13:59 -0500 Subject: [PATCH 434/590] chore: Format `megatron/optimizer/*` --- megatron/optimizer/__init__.py | 189 ++++++--- megatron/optimizer/clip_grads.py | 51 +-- megatron/optimizer/distrib_optimizer.py | 494 +++++++++++++----------- megatron/optimizer/grad_scaler.py | 37 +- megatron/optimizer/optimizer.py | 457 ++++++++++++---------- megatron/optimizer/sophia.py | 257 +++++++----- 6 files changed, 848 insertions(+), 637 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 37107a44e71..38cb3544939 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -12,15 +12,16 @@ import ezpz as ez + RANK = ez.get_rank() def get_param_groups( - modules: Union[torch.nn.Module, Iterable[torch.nn.Module]], - no_weight_decay_cond: Callable[[str, torch.Tensor], bool], - scale_lr_cond: Callable[[str, torch.Tensor], bool], - lr_mult: Any, - use_galore: bool = False, + modules: Union[torch.nn.Module, Iterable[torch.nn.Module]], + no_weight_decay_cond: Callable[[str, torch.Tensor], bool], + scale_lr_cond: Callable[[str, torch.Tensor], bool], + lr_mult: Any, + use_galore: bool = False, ): """ Creates param groups (regularized vs non) based on: @@ -92,37 +93,58 @@ def get_param_groups( param_groups = [] if len(wd_no_scale_lr): - param_groups.append({'name': 'wd_no_scale_lr', 'params': wd_no_scale_lr, 'wd_mult': 1.0, 'lr_mult': 1.0}) + param_groups.append( + { + "name": "wd_no_scale_lr", + "params": wd_no_scale_lr, + "wd_mult": 1.0, + "lr_mult": 1.0, + } + ) if len(wd_scale_lr): - param_groups.append({'name': 'wd_scale_lr', 'params': wd_scale_lr, 'wd_mult': 1.0, 'lr_mult': lr_mult}) + param_groups.append( + { + "name": "wd_scale_lr", + "params": wd_scale_lr, + "wd_mult": 1.0, + "lr_mult": lr_mult, + } + ) if len(no_wd_no_scale_lr): - param_groups.append({'name': 'no_wd_no_scale_lr', 'params': no_wd_no_scale_lr, 'wd_mult': 0.0, 'lr_mult': 1.0}) + param_groups.append( + { + "name": "no_wd_no_scale_lr", + "params": no_wd_no_scale_lr, + "wd_mult": 0.0, + "lr_mult": 1.0, + } + ) if len(no_wd_scale_lr): - param_groups.append({'name': 'no_wd_scale_lr', 'params': no_wd_scale_lr, 'wd_mult': 0.0, 'lr_mult': lr_mult}) + param_groups.append( + { + "name": "no_wd_scale_lr", + "params": no_wd_scale_lr, + "wd_mult": 0.0, + "lr_mult": lr_mult, + } + ) return param_groups def get_megatron_optimizer( - model, - no_weight_decay_cond=None, - scale_lr_cond=None, - lr_mult=1.0 + model, no_weight_decay_cond=None, scale_lr_cond=None, lr_mult=1.0 ): args = get_args() assert args is not None # Base optimizer. - param_groups = get_param_groups( - model, - no_weight_decay_cond, - scale_lr_cond, - lr_mult - ) + param_groups = get_param_groups(model, no_weight_decay_cond, scale_lr_cond, lr_mult) if args.create_moe_param_group: from deepspeed.moe.utils import ( - split_params_into_different_moe_groups_for_optimizer + split_params_into_different_moe_groups_for_optimizer, ) + param_groups = split_params_into_different_moe_groups_for_optimizer( param_groups ) @@ -130,11 +152,12 @@ def get_megatron_optimizer( optimizer = None # ---- CPU Optimizer -------------------------------------- if args.cpu_optimizer: - assert args.optimizer == 'adam', 'CPU offloading is for Adam' + assert args.optimizer == "adam", "CPU offloading is for Adam" if args.cpu_torch_adam: cpu_adam_optimizer = torch.optim.AdamW else: from deepspeed.ops.adam import DeepSpeedCPUAdam + cpu_adam_optimizer = DeepSpeedCPUAdam optimizer = cpu_adam_optimizer( param_groups, @@ -144,10 +167,11 @@ def get_megatron_optimizer( eps=args.adam_eps, ) # ---- Adam -------------------------------------- - elif args.optimizer == 'adam': + elif args.optimizer == "adam": if args.ds_fused_adam: # global Adam from deepspeed.ops.adam import FusedAdam + Adam = FusedAdam else: Adam = torch.optim.Adam @@ -156,35 +180,40 @@ def get_megatron_optimizer( lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps + eps=args.adam_eps, ) # ---- apex.Adam -------------------------------------------- - elif str(args.optimizer).lower() == 'apex.adam': - assert get_accelerator().device_name() == 'cuda' + elif str(args.optimizer).lower() == "apex.adam": + assert get_accelerator().device_name() == "cuda" from apex.optimizers import FusedAdam as Adam + optimizer = Adam( param_groups, lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps + eps=args.adam_eps, ) # ---- Adam8Bit -------------------------------------- elif args.optimizer.lower() == "adam8bit": import bitsandbytes as bnb - optimizer = bnb.optim.Adam8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + + optimizer = bnb.optim.Adam8bit( + param_groups, lr=args.lr, weight_decay=args.weight_decay + ) # ---- AdamW -------------------------------------- - elif str(args.optimizer).lower() == 'adamw': + elif str(args.optimizer).lower() == "adamw": optimizer = torch.optim.AdamW( param_groups, lr=args.lr, weight_decay=args.weight_decay, betas=(args.adam_beta1, args.adam_beta2), - eps=args.adam_eps + eps=args.adam_eps, ) # ---- AdamW: ScheduleFree ------------------------------------- - elif str(args.optimizer).lower() == 'adamwschedulefree': + elif str(args.optimizer).lower() == "adamwschedulefree": import schedulefree + optimizer = schedulefree.AdamWScheduleFree( param_groups, lr=args.lr, @@ -198,7 +227,9 @@ def get_megatron_optimizer( elif args.optimizer.lower() == "galore_adamw": from galore_torch import GaLoreAdamW # redefine way to call galore_adamw - optimizer = GaLoreAdamW(param_groups, lr=args.lr, weight_decay=args.weight_decay) + optimizer = GaLoreAdamW( + param_groups, lr=args.lr, weight_decay=args.weight_decay + ) # elif args.optimizer.lower() == "galore_adamw": # from galore_torch import GaLoreAdamW # # redefine way to call galore_adamw @@ -206,21 +237,39 @@ def get_megatron_optimizer( # ---- AdamW: GaloRe 8Bit -------------------------------------- elif args.optimizer.lower() == "galore_adamw8bit": from galore_torch import GaLoreAdamW8bit - optimizer = GaLoreAdamW8bit(param_groups, lr=args.lr, weight_decay=args.weight_decay) + + optimizer = GaLoreAdamW8bit( + param_groups, lr=args.lr, weight_decay=args.weight_decay + ) # ---- AdamW8bitPerLayer: GaloRE ---------------------------- - elif args.optimizer.lower() == 'galore_adamw8bit_per_layer': + elif args.optimizer.lower() == "galore_adamw8bit_per_layer": from galore_torch import GaLoreAdamW8bit # TODO: seems scheduler call twice in one update step, need to check, for now double the num_training_steps, warmup_steps and update_proj_gap optimizer_dict = {} for p in model.parameters(): if p.requires_grad: if id(p) in id_galore_params: - optimizer_dict[p] = GaLoreAdamW8bit([{'params': [p], 'rank': args.rank, 'update_proj_gap': args.update_proj_gap * 2, 'scale': args.galore_scale, 'proj_type': args.proj_type}], lr=args.lr, weight_decay=args.weight_decay) + optimizer_dict[p] = GaLoreAdamW8bit( + [ + { + "params": [p], + "rank": args.rank, + "update_proj_gap": args.update_proj_gap * 2, + "scale": args.galore_scale, + "proj_type": args.proj_type, + } + ], + lr=args.lr, + weight_decay=args.weight_decay, + ) else: - optimizer_dict[p] = bnb.optim.Adam8bit([p], lr=args.lr, weight_decay=args.weight_decay) + optimizer_dict[p] = bnb.optim.Adam8bit( + [p], lr=args.lr, weight_decay=args.weight_decay + ) # get scheduler dict scheduler_dict = {} from galore_torch.peft_pretraining import training_utils + for p in model.parameters(): if p.requires_grad: scheduler_dict[p] = training_utils.get_scheculer( @@ -232,11 +281,12 @@ def get_megatron_optimizer( ) def optimizer_hook(p): - if p.grad is None: + if p.grad is None: return optimizer_dict[p].step() optimizer_dict[p].zero_grad() scheduler_dict[p].step() + # Register the hook onto every parameter for p in model.parameters(): if p.requires_grad: @@ -245,6 +295,7 @@ def optimizer_hook(p): # ---- AdaFactor -------------------------------------- elif args.optimizer.lower() == "adafactor": import transformers + args.beta1 = None if args.beta1 == 0.0 else args.beta1 optimizer = transformers.optimization.Adafactor( param_groups, @@ -261,6 +312,7 @@ def optimizer_hook(p): # ---- GaLore: Adafactor adafactor ------------------------------------ elif args.optimizer.lower() == "galore_adafactor": from galore_torch import GaLoreAdafactor + args.beta1 = None if args.beta1 == 0.0 else args.beta1 optimizer = GaLoreAdafactor( param_groups, @@ -275,17 +327,19 @@ def optimizer_hook(p): warmup_init=False, ) # ---- Apex: sgd --------------------------------------------- - elif str(args.optimizer).lower() == 'apex.sgd': + elif str(args.optimizer).lower() == "apex.sgd": from apex.optimizers import FusedSGD as SGD + optimizer = SGD( param_groups, lr=args.lr, weight_decay=args.weight_decay, - momentum=args.sgd_momentum + momentum=args.sgd_momentum, ) # ---- ScheduleFree: SGD ------------------------------- - elif str(args.optimizer).lower() == 'sgdschedulefree': + elif str(args.optimizer).lower() == "sgdschedulefree": import schedulefree + optimizer = schedulefree.SGDScheduleFree( param_groups, lr=args.lr, @@ -295,8 +349,9 @@ def optimizer_hook(p): foreach=args.schedulefree_for_each, ) # ---- Lamb: Ipex -------------------------------------------- - elif str(args.optimizer) == 'ipex.lamb': + elif str(args.optimizer) == "ipex.lamb": from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( param_groups, lr=args.lr, @@ -305,8 +360,9 @@ def optimizer_hook(p): eps=args.adam_eps, ) # ---- Lamb(Fused): Ipex ---------------------------------------- - elif str(args.optimizer) == 'ipex.fusedlamb': + elif str(args.optimizer) == "ipex.fusedlamb": from intel_extension_for_pytorch.optim._lamb import Lamb + optimizer = Lamb( param_groups, lr=args.lr, @@ -316,8 +372,9 @@ def optimizer_hook(p): fused=True, ) # ---- Lamb(Fused): DeepSpeed ------------------------------------------ - elif str(args.optimizer).lower() == 'ds.fusedlamb': + elif str(args.optimizer).lower() == "ds.fusedlamb": from deepspeed.ops.lamb import FusedLamb + optimizer = FusedLamb( param_groups, lr=args.lr, @@ -326,9 +383,10 @@ def optimizer_hook(p): eps=args.adam_eps, ) # ---- Shampoo ---------------------------------------- - elif args.optimizer == 'shampoo': + elif args.optimizer == "shampoo": from distributed_shampoo.distributed_shampoo import DistributedShampoo from distributed_shampoo.shampoo_types import AdamGraftingConfig + optimizer = DistributedShampoo( model.parameters(), lr=0.001, @@ -343,24 +401,25 @@ def optimizer_hook(p): epsilon=1e-08, ), ) - elif args.optimizer == 'sgd': + elif args.optimizer == "sgd": optimizer = torch.optim.SGD( param_groups, lr=args.lr, weight_decay=args.weight_decay, - momentum=args.sgd_momentum + momentum=args.sgd_momentum, ) - elif str(args.optimizer).lower() == 'sophiag': + elif str(args.optimizer).lower() == "sophiag": from .sophia import SophiaG + optimizer = SophiaG( param_groups, lr=args.lr, betas=(args.sophiag_beta1, args.sophiag_beta2), - rho = args.sophiag_rho, - weight_decay=args.weight_decay + rho=args.sophiag_rho, + weight_decay=args.weight_decay, ) else: - raise TypeError(f'{args.optimizer} optimizer is not supported.') + raise TypeError(f"{args.optimizer} optimizer is not supported.") assert optimizer is not None if args.deepspeed: return optimizer @@ -393,22 +452,26 @@ def optimizer_hook(p): growth_factor=2.0, backoff_factor=0.5, growth_interval=args.loss_scale_window, - hysteresis=args.hysteresis) + hysteresis=args.hysteresis, + ) # Megatron optimizer. opt_ty = ( - DistributedOptimizer if args.use_distributed_optimizer - else Float16OptimizerWithFloat16Params + DistributedOptimizer + if args.use_distributed_optimizer + else Float16OptimizerWithFloat16Params + ) + return opt_ty( + optimizer, + args.clip_grad, + args.log_num_zeros_in_grad, + params_have_main_grad, + args.use_contiguous_buffers_in_local_ddp, + args.fp16, + args.bf16, + args.params_dtype, + grad_scaler, + model, ) - return opt_ty(optimizer, - args.clip_grad, - args.log_num_zeros_in_grad, - params_have_main_grad, - args.use_contiguous_buffers_in_local_ddp, - args.fp16, - args.bf16, - args.params_dtype, - grad_scaler, - model) # FP32. return FP32Optimizer( optimizer, @@ -416,5 +479,5 @@ def optimizer_hook(p): args.log_num_zeros_in_grad, params_have_main_grad, args.use_contiguous_buffers_in_local_ddp, - model + model, ) diff --git a/megatron/optimizer/clip_grads.py b/megatron/optimizer/clip_grads.py index b5141d00590..1c731100efe 100644 --- a/megatron/optimizer/clip_grads.py +++ b/megatron/optimizer/clip_grads.py @@ -3,6 +3,7 @@ """Gradient clipping.""" import torch + try: from torch._six import inf as inf except ModuleNotFoundError: @@ -13,6 +14,7 @@ try: from apex.multi_tensor_apply import multi_tensor_applier import amp_C + HAS_APEX = True except Exception: HAS_APEX = False @@ -21,9 +23,9 @@ from megatron.core.tensor_parallel import param_is_not_tensor_parallel_duplicate -def clip_grad_norm_fp32(parameters, grads_for_norm, - max_norm, norm_type=2, - model_parallel_group=None): +def clip_grad_norm_fp32( + parameters, grads_for_norm, max_norm, norm_type=2, model_parallel_group=None +): """Clips gradient norm of an iterable of parameters whose gradients are in fp32. @@ -55,7 +57,9 @@ def clip_grad_norm_fp32(parameters, grads_for_norm, grads = [] for param in parameters: if param.grad is not None: - assert param.grad.type() == 'torch.{}.FloatTensor'.format(get_accelerator().device_name()) + assert param.grad.type() == "torch.{}.FloatTensor".format( + get_accelerator().device_name() + ) grads.append(param.grad.detach()) # Norm parameters. @@ -68,14 +72,16 @@ def clip_grad_norm_fp32(parameters, grads_for_norm, total_norm = max(grad.abs().max() for grad in grads_for_norm) total_norm_cuda = get_accelerator().FloatTensor([float(total_norm)]) # Take max across all model-parallel GPUs. - torch.distributed.all_reduce(total_norm_cuda, - op=torch.distributed.ReduceOp.MAX, - group=model_parallel_group) + torch.distributed.all_reduce( + total_norm_cuda, + op=torch.distributed.ReduceOp.MAX, + group=model_parallel_group, + ) total_norm = total_norm_cuda[0].item() else: if norm_type == 2.0: - if get_accelerator().device_name() == 'cuda' and HAS_APEX: + if get_accelerator().device_name() == "cuda" and HAS_APEX: dummy_overflow_buf = torch.cuda.IntTensor([0]) # Use apex's multi-tensor applier for efficiency reasons. # Multi-tensor applier takes a function and a list of list @@ -85,35 +91,34 @@ def clip_grad_norm_fp32(parameters, grads_for_norm, amp_C.multi_tensor_l2norm, dummy_overflow_buf, [grads_for_norm], - False # no per-parameter norm + False, # no per-parameter norm ) else: grad_norm = torch.cuda.FloatTensor([0]) else: - grad_norm = torch.norm(grads_for_norm,p=2.0) + grad_norm = torch.norm(grads_for_norm, p=2.0) # Since we will be summing across data parallel groups, # we need the pow(norm-type). - total_norm = grad_norm ** norm_type + total_norm = grad_norm**norm_type else: for grad in grads_for_norm: grad_norm = torch.norm(grad, norm_type) - total_norm += grad_norm ** norm_type + total_norm += grad_norm**norm_type # Sum across all model-parallel GPUs. - torch.distributed.all_reduce(total_norm, - op=torch.distributed.ReduceOp.SUM, - group=model_parallel_group) + torch.distributed.all_reduce( + total_norm, op=torch.distributed.ReduceOp.SUM, group=model_parallel_group + ) total_norm = total_norm.item() ** (1.0 / norm_type) # Scale. clip_coeff = max_norm / (total_norm + 1.0e-6) if clip_coeff < 1.0: - if get_accelerator().device_name() == 'cuda': + if get_accelerator().device_name() == "cuda": dummy_overflow_buf = get_accelerator().IntTensor([0]) - multi_tensor_applier(amp_C.multi_tensor_scale, - dummy_overflow_buf, - [grads, grads], - clip_coeff) + multi_tensor_applier( + amp_C.multi_tensor_scale, dummy_overflow_buf, [grads, grads], clip_coeff + ) else: for g in grads: g.detach().mul_(clip_coeff.to(g.device)) @@ -141,9 +146,9 @@ def count_zeros_fp32(parameters, model_parallel_group): total_num_zeros = num_zeros + total_num_zeros # Sum across all model-parallel GPUs. - torch.distributed.all_reduce(total_num_zeros, - op=torch.distributed.ReduceOp.SUM, - group=model_parallel_group) + torch.distributed.all_reduce( + total_num_zeros, op=torch.distributed.ReduceOp.SUM, group=model_parallel_group + ) total_num_zeros = total_num_zeros.item() diff --git a/megatron/optimizer/distrib_optimizer.py b/megatron/optimizer/distrib_optimizer.py index 10331607d99..117797efd61 100644 --- a/megatron/optimizer/distrib_optimizer.py +++ b/megatron/optimizer/distrib_optimizer.py @@ -27,14 +27,18 @@ class Range: A range represents a start and end points for indexing a shard from a full tensor. """ + def __init__(self, start, end): self.start = start self.end = end self.size = end - start - def normalize(self, start = 0): + + def normalize(self, start=0): return Range(start, start + self.size) + def __str__(self): return "%d,%d [%d]" % (self.start, self.end, self.size) + def __len__(self): return self.end - self.start @@ -103,29 +107,27 @@ def build_model_gbuf_param_range_map(cls, model, dtype, gbuf_world_range): # Param range. param_world_start, param_world_end = param_world_indexes - param_local_start = max( - 0, - param_world_start - gbuf_world_range.start) + param_local_start = max(0, param_world_start - gbuf_world_range.start) param_local_end = min( - gbuf_world_range.size, - param_world_end - gbuf_world_range.start) + gbuf_world_range.size, param_world_end - gbuf_world_range.start + ) # Add param, if within local gbuf range. if param_local_end > param_local_start: param_local_range = Range(param_local_start, param_local_end) param_world_range = param_local_range.normalize( - param_local_start + gbuf_world_range.start) - sub_param_start = max(0, gbuf_world_range.start-param_world_start) + param_local_start + gbuf_world_range.start + ) + sub_param_start = max(0, gbuf_world_range.start - param_world_start) sub_param_range = param_local_range.normalize(sub_param_start) param_range_map[param] = { - "gbuf_world" : param_world_range, - "gbuf_local" : param_local_range, - "param" : sub_param_range, + "gbuf_world": param_world_range, + "gbuf_local": param_local_range, + "param": sub_param_range, } return param_range_map - @classmethod def build_model_gbuf_range(cls, model, dtype): """ @@ -150,7 +152,7 @@ def build_model_gbuf_range(cls, model, dtype): gbuf_world_all_ranges = [] for r in range(data_parallel_world_size): gbuf_world_start = r * max_gbuf_range_size - gbuf_world_end = min(gbuf_size, gbuf_world_start+max_gbuf_range_size) + gbuf_world_end = min(gbuf_size, gbuf_world_start + max_gbuf_range_size) gbuf_world_range = Range(gbuf_world_start, gbuf_world_end) gbuf_world_all_ranges.append(gbuf_world_range) @@ -159,22 +161,21 @@ def build_model_gbuf_range(cls, model, dtype): gbuf_local_range = gbuf_world_range.normalize() # Get each param's ranges. - param_range_map = cls.build_model_gbuf_param_range_map(model, - dtype, - gbuf_world_range) + param_range_map = cls.build_model_gbuf_param_range_map( + model, dtype, gbuf_world_range + ) # Group into dict. data = { - "local" : gbuf_local_range, - "world" : gbuf_world_range, - "world_all" : gbuf_world_all_ranges, - "param_map" : param_range_map, - "max_range_size" : max_gbuf_range_size, + "local": gbuf_local_range, + "world": gbuf_world_range, + "world_all": gbuf_world_all_ranges, + "param_map": param_range_map, + "max_range_size": max_gbuf_range_size, } return data - @classmethod def build_model_gbuf_range_map(cls, model): """ @@ -182,11 +183,10 @@ def build_model_gbuf_range_map(cls, model): within a specific virtual model. """ return { - dtype : cls.build_model_gbuf_range(model, dtype) + dtype: cls.build_model_gbuf_range(model, dtype) for dtype in model._grad_buffers } - @classmethod def build_model_param_gbuf_map(cls, model_gbuf_ranges): """ @@ -200,7 +200,6 @@ def build_model_param_gbuf_map(cls, model_gbuf_ranges): param_gbuf_map[param] = (model_index, dtype) return param_gbuf_map - @classmethod def build_optimizer_group_ranges(cls, param_groups, model_gbuf_ranges): """ @@ -233,15 +232,17 @@ def build_optimizer_group_ranges(cls, param_groups, model_gbuf_ranges): # the group. The group index and order are particularly important for # saving and loading checkpoints. local_param_group_map = {} - group_ranges = [ {"params": []} for _ in param_groups ] + group_ranges = [{"params": []} for _ in param_groups] for model_gbuf_range_map in model_gbuf_ranges: for dtype, gbuf_range_map in model_gbuf_range_map.items(): for param in gbuf_range_map["param_map"]: group_index = world_param_group_map[param] group_range = group_ranges[group_index] group_range["params"].append(param) - local_param_group_map[param] = \ - (group_index, len(group_range["params"]) - 1) + local_param_group_map[param] = ( + group_index, + len(group_range["params"]) - 1, + ) # Squeeze zero-size group ranges. for group_index, group_range in enumerate(group_ranges): @@ -250,12 +251,10 @@ def build_optimizer_group_ranges(cls, param_groups, model_gbuf_ranges): return local_param_group_map, group_ranges - @classmethod - def build_model_and_main_param_groups(cls, - model_gbuf_ranges, - param_gbuf_map, - opt_group_ranges): + def build_model_and_main_param_groups( + cls, model_gbuf_ranges, param_gbuf_map, opt_group_ranges + ): """ Create main parameter groups needed for the optimizer step. @@ -293,7 +292,8 @@ def build_model_and_main_param_groups(cls, shard_float16_groups.append(shard_float16_params_this_group) shard_fp32_groups.append(shard_fp32_params_this_group) shard_fp32_from_float16_groups.append( - shard_fp32_from_float16_params_this_group) + shard_fp32_from_float16_params_this_group + ) for model_param in group_range["params"]: @@ -304,18 +304,23 @@ def build_model_and_main_param_groups(cls, param_range = gbuf_range["param_map"][model_param]["param"] # fp16, bf16 params. - if model_param.type() in ['torch.cuda.HalfTensor', - 'torch.cuda.BFloat16Tensor']: + if model_param.type() in [ + "torch.cuda.HalfTensor", + "torch.cuda.BFloat16Tensor", + ]: # Clone model -> main. - shard_model_param = model_param.detach().view(-1) \ - [param_range.start:param_range.end] + shard_model_param = model_param.detach().view(-1)[ + param_range.start : param_range.end + ] shard_main_param = shard_model_param.clone().float() tensor_parallel.copy_tensor_model_parallel_attributes( - shard_model_param, model_param) + shard_model_param, model_param + ) tensor_parallel.copy_tensor_model_parallel_attributes( - shard_main_param, model_param) - if hasattr(model_param, 'shared'): + shard_main_param, model_param + ) + if hasattr(model_param, "shared"): shard_model_param.shared = model_param.shared shard_main_param.shared = model_param.shared @@ -325,22 +330,26 @@ def build_model_and_main_param_groups(cls, shard_fp32_from_float16_params_this_group.append(shard_main_param) # fp32 params. - elif model_param.type() == 'torch.cuda.FloatTensor': - shard_model_param = model_param.view(-1) \ - [param_range.start:param_range.end] + elif model_param.type() == "torch.cuda.FloatTensor": + shard_model_param = model_param.view(-1)[ + param_range.start : param_range.end + ] model_fp32_params_this_group.append(model_param) shard_fp32_params_this_group.append(shard_model_param) tensor_parallel.copy_tensor_model_parallel_attributes( - shard_model_param, model_param) - if hasattr(model_param, 'shared'): + shard_model_param, model_param + ) + if hasattr(model_param, "shared"): shard_model_param.shared = model_param.shared else: - raise TypeError('Wrapped parameters must be one of ' - 'torch.cuda.FloatTensor, ' - 'torch.cuda.HalfTensor, or ' - 'torch.cuda.BFloat16Tensor. ' - 'Received {}'.format(param.type())) + raise TypeError( + "Wrapped parameters must be one of " + "torch.cuda.FloatTensor, " + "torch.cuda.HalfTensor, or " + "torch.cuda.BFloat16Tensor. " + "Received {}".format(param.type()) + ) # Update optimizer's params. group_range["orig_group"]["params"] = [ @@ -356,10 +365,19 @@ def build_model_and_main_param_groups(cls, shard_fp32_from_float16_groups, ) - - def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - fp16, bf16, params_dtype, grad_scaler, models): + def __init__( + self, + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + fp16, + bf16, + params_dtype, + grad_scaler, + models, + ): """ See top of class definition for argument descriptions. @@ -371,28 +389,40 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, """ super().__init__( - optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - fp16, bf16, params_dtype, grad_scaler, models) + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + fp16, + bf16, + params_dtype, + grad_scaler, + models, + ) # Verify that contiguous buffers are being used. # - Note: this should already be checked in arguments.py. assert use_contiguous_buffers_in_local_ddp - assert isinstance(optimizer, Adam), \ - "Only Adam currently supported, due to checkpointing requirements." + assert isinstance( + optimizer, Adam + ), "Only Adam currently supported, due to checkpointing requirements." # Model grad buffer ranges. self.model_gbuf_ranges = [] for model_index, model in enumerate(self.models): self.model_gbuf_ranges.append(self.build_model_gbuf_range_map(model)) - self.model_param_gbuf_map = \ - self.build_model_param_gbuf_map(self.model_gbuf_ranges) + self.model_param_gbuf_map = self.build_model_param_gbuf_map( + self.model_gbuf_ranges + ) # Optimizer ranges. - self.model_param_group_index_map, self.opt_group_ranges = \ - self.build_optimizer_group_ranges(self.optimizer.param_groups, - self.model_gbuf_ranges) - + self.model_param_group_index_map, self.opt_group_ranges = ( + self.build_optimizer_group_ranges( + self.optimizer.param_groups, self.model_gbuf_ranges + ) + ) + # Allocate main param shards. ( self.model_float16_groups, @@ -400,9 +430,9 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, self.shard_float16_groups, self.shard_fp32_groups, self.shard_fp32_from_float16_groups, - ) = self.build_model_and_main_param_groups(self.model_gbuf_ranges, - self.model_param_gbuf_map, - self.opt_group_ranges) + ) = self.build_model_and_main_param_groups( + self.model_gbuf_ranges, self.model_param_gbuf_map, self.opt_group_ranges + ) # Initialize param buffers. # - These are views on the DDP model's grad buffers, that share @@ -421,21 +451,18 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, # Typed param buffer. param_buffer = torch.tensor( - storage, - dtype = params_dtype, - device = grad_buffer.data.device) - param_buffer = param_buffer[:grad_buffer.numel_padded] + storage, dtype=params_dtype, device=grad_buffer.data.device + ) + param_buffer = param_buffer[: grad_buffer.numel_padded] current_param_buffers[dtype] = param_buffer self.param_buffers.append(current_param_buffers) # Update optimizer groups. # - Also, leverage state_dict() and load_state_dict() to # recast preexisting per-param state tensors. - self.optimizer.param_groups = \ - [ g["orig_group"] for g in self.opt_group_ranges ] + self.optimizer.param_groups = [g["orig_group"] for g in self.opt_group_ranges] self.optimizer.load_state_dict(self.optimizer.state_dict()) - def get_model_param_range_map(self, param): """ Given a model param, get the index sub-range of the param that this @@ -446,7 +473,6 @@ def get_model_param_range_map(self, param): param_range_map = gbuf_range_map["param_map"][param] return param_range_map - def get_model_parallel_group(self): """ With the distributed optimizer, the model parallel group is the @@ -454,7 +480,6 @@ def get_model_parallel_group(self): """ return None - def state_dict(self): """ The state dict contains all non-DP-rank-dependent (i.e., non-parameter- @@ -467,21 +492,18 @@ def state_dict(self): state_dict = {} # Optimizer state (do not store parameter state here). - state_dict['optimizer'] = { - k : v - for k, v in self.optimizer.state_dict().items() - if k != "state" + state_dict["optimizer"] = { + k: v for k, v in self.optimizer.state_dict().items() if k != "state" } for param_group in state_dict["optimizer"]["param_groups"]: del param_group["params"] # Grad scaler state. if self.grad_scaler: - state_dict['grad_scaler'] = self.grad_scaler.state_dict() + state_dict["grad_scaler"] = self.grad_scaler.state_dict() return state_dict - def load_state_dict(self, state_dict): """Load the state dict. @@ -518,61 +540,76 @@ def load_state_dict(self, state_dict): # the ordering of parameters within its flattened parameter state # list. inner_state_dict = self.optimizer.state_dict() - state_dict_param_groups = [{ - **group, - "params" : list(inner_state_dict["param_groups"][idx]["params"]), - } for idx, group in enumerate(state_dict["optimizer"]["param_groups"])] + state_dict_param_groups = [ + { + **group, + "params": list(inner_state_dict["param_groups"][idx]["params"]), + } + for idx, group in enumerate(state_dict["optimizer"]["param_groups"]) + ] # Allocate 'dummy' data for optimizer state (i.e., torch.empty() below) # - Real data is overwritten during load_parameter_state(). state_dict_state = [] for gbuf_range_maps in self.model_gbuf_ranges: for gbuf_range_map in gbuf_range_maps.values(): - for model_param, param_range_map in \ - gbuf_range_map["param_map"].items(): + for model_param, param_range_map in gbuf_range_map["param_map"].items(): # Get parameter ordering information (see method docstring # for details). - group_index, group_order = \ - self.model_param_group_index_map[model_param] - state_order = inner_state_dict["param_groups"] \ - [group_index]["params"][group_order] + group_index, group_order = self.model_param_group_index_map[ + model_param + ] + state_order = inner_state_dict["param_groups"][group_index][ + "params" + ][group_order] # Allocate dummy tensors. numel = len(param_range_map["gbuf_world"]) - init_shard = lambda : torch.empty( + init_shard = lambda: torch.empty( (numel,), dtype=torch.float32, - device=torch.cuda.current_device()) + device=torch.cuda.current_device(), + ) - state_dict_state.append((state_order, { - "exp_avg" : init_shard(), - "exp_avg_sq" : init_shard(), - })) + state_dict_state.append( + ( + state_order, + { + "exp_avg": init_shard(), + "exp_avg_sq": init_shard(), + }, + ) + ) # Sort by state order (see method docstring for details). - state_dict_state.sort(key = lambda s : s[0]) - state_dict_state = {s[0]:s[1] for s in state_dict_state} + state_dict_state.sort(key=lambda s: s[0]) + state_dict_state = {s[0]: s[1] for s in state_dict_state} # Optimizer. - self.optimizer.load_state_dict({ - "state" : state_dict_state, - "param_groups" : state_dict_param_groups, - }) + self.optimizer.load_state_dict( + { + "state": state_dict_state, + "param_groups": state_dict_param_groups, + } + ) # Grad scaler. - if 'grad_scaler' not in state_dict: + if "grad_scaler" not in state_dict: if self.fp16: - print_rank_0('***WARNING*** found an old checkpoint, will not ' - 'load grad scaler ...') + print_rank_0( + "***WARNING*** found an old checkpoint, will not " + "load grad scaler ..." + ) else: if self.grad_scaler: - self.grad_scaler.load_state_dict(state_dict['grad_scaler']) + self.grad_scaler.load_state_dict(state_dict["grad_scaler"]) else: - print_rank_0('***WARNING*** fould the grad scaler in the ' - 'checkpoint but it is None in the class. ' - 'Skipping loading grad scaler ...') - + print_rank_0( + "***WARNING*** fould the grad scaler in the " + "checkpoint but it is None in the class. " + "Skipping loading grad scaler ..." + ) def save_parameter_state(self, filename): """Save parameter state (i.e., parameter & optimizer tensors). @@ -604,25 +641,28 @@ def save_parameter_state(self, filename): # Compute local DP contiguous shard's size. model = self.models[model_idx] gbuf_world_numel = model._grad_buffers[dtype].numel_padded - gbuf_local_numel = int(gbuf_world_numel/data_parallel_world_size) - local_shards = {key:torch.empty((gbuf_local_numel,), - dtype=torch.float32, - device="cpu") - for key in ("param", "exp_avg", "exp_avg_sq")} + gbuf_local_numel = int(gbuf_world_numel / data_parallel_world_size) + local_shards = { + key: torch.empty( + (gbuf_local_numel,), dtype=torch.float32, device="cpu" + ) + for key in ("param", "exp_avg", "exp_avg_sq") + } # Build contiguous DP rank shards (for param + optim states). - for model_param, param_range_map in \ - gbuf_range_map["param_map"].items(): + for model_param, param_range_map in gbuf_range_map["param_map"].items(): # Main param & optimizer states. - group_index, group_order = \ - self.model_param_group_index_map[model_param] - main_param = self.optimizer.param_groups \ - [group_index]["params"][group_order] + group_index, group_order = self.model_param_group_index_map[ + model_param + ] + main_param = self.optimizer.param_groups[group_index]["params"][ + group_order + ] optim_state = self.optimizer.state[main_param] tensors = { - "param" : main_param, + "param": main_param, **optim_state, } @@ -630,19 +670,22 @@ def save_parameter_state(self, filename): gbuf_local_start = param_range_map["gbuf_local"].start gbuf_local_end = param_range_map["gbuf_local"].end for key in local_shards: - local_shards[key][gbuf_local_start:gbuf_local_end] \ - .data.copy_(tensors[key].detach().cpu()) + local_shards[key][gbuf_local_start:gbuf_local_end].data.copy_( + tensors[key].detach().cpu() + ) # Gather contiguous shards on DP rank 0. world_tensors = {} for key, send_tensor in local_shards.items(): - + # Gather tensor list. if data_parallel_rank == 0: - recv_tensors = [torch.empty((gbuf_local_numel,), - dtype=torch.float32, - device="cpu") - for _ in range(data_parallel_world_size)] + recv_tensors = [ + torch.empty( + (gbuf_local_numel,), dtype=torch.float32, device="cpu" + ) + for _ in range(data_parallel_world_size) + ] else: recv_tensors = None @@ -666,7 +709,6 @@ def save_parameter_state(self, filename): if data_parallel_rank == 0: torch.save(state, filename) - def load_parameter_state(self, filename): """Load parameter state (i.e., parameter & optimizer tensors). @@ -696,24 +738,29 @@ def load_parameter_state(self, filename): # Compute local DP contiguous shard's size. model = self.models[model_idx] gbuf_world_numel = model._grad_buffers[dtype].numel_padded - gbuf_local_numel = int(gbuf_world_numel/data_parallel_world_size) + gbuf_local_numel = int(gbuf_world_numel / data_parallel_world_size) # Contiguous local shards (received from DP rank 0). - local_shards = {key:torch.empty((gbuf_local_numel,), - dtype=torch.float32, - device="cpu") - for key in ("param", "exp_avg", "exp_avg_sq")} + local_shards = { + key: torch.empty( + (gbuf_local_numel,), dtype=torch.float32, device="cpu" + ) + for key in ("param", "exp_avg", "exp_avg_sq") + } # Scatter local shards from DP rank 0. for key, recv_tensor in local_shards.items(): - + # Scatter tensor list. if data_parallel_rank == 0: world_tensor = loaded_state[model_idx][dtype][key] - gbuf_start_idxs = \ - list(range(0, gbuf_world_numel, gbuf_local_numel)) - send_tensors = [world_tensor[i:(i+gbuf_local_numel)] - for i in gbuf_start_idxs] + gbuf_start_idxs = list( + range(0, gbuf_world_numel, gbuf_local_numel) + ) + send_tensors = [ + world_tensor[i : (i + gbuf_local_numel)] + for i in gbuf_start_idxs + ] else: send_tensors = None @@ -726,18 +773,19 @@ def load_parameter_state(self, filename): ) # Copy local contiguous shards to param/optim shards. - for model_param, param_range_map in \ - gbuf_range_map["param_map"].items(): + for model_param, param_range_map in gbuf_range_map["param_map"].items(): # Main param & optimizer states. - group_index, group_order = \ - self.model_param_group_index_map[model_param] - main_param = self.optimizer.param_groups \ - [group_index]["params"][group_order] + group_index, group_order = self.model_param_group_index_map[ + model_param + ] + main_param = self.optimizer.param_groups[group_index]["params"][ + group_order + ] optim_state = self.optimizer.state[main_param] tensors = { - "param" : main_param, + "param": main_param, **optim_state, } @@ -746,8 +794,8 @@ def load_parameter_state(self, filename): gbuf_local_end = param_range_map["gbuf_local"].end for key in local_shards: tensors[key].data.copy_( - local_shards[key][gbuf_local_start:gbuf_local_end]) - + local_shards[key][gbuf_local_start:gbuf_local_end] + ) def zero_grad(self, set_to_none=True): """ @@ -760,15 +808,15 @@ def zero_grad(self, set_to_none=True): used by this field can be safely deallocated at this point. """ for groups in ( - self.model_float16_groups, - self.model_fp32_groups, - self.shard_float16_groups, # grad empty/unused here? - self.shard_fp32_groups, # throws grad-access warning - self.shard_fp32_from_float16_groups): + self.model_float16_groups, + self.model_fp32_groups, + self.shard_float16_groups, # grad empty/unused here? + self.shard_fp32_groups, # throws grad-access warning + self.shard_fp32_from_float16_groups, + ): for group in groups: _zero_grad_group_helper(group, set_to_none) - @staticmethod def get_model_buffer_dp_views(model_buffers): """ @@ -795,24 +843,26 @@ def get_model_buffer_dp_views(model_buffers): assert buf.numel() % data_parallel_world_size == 0 shard_size = int(buf.numel() / data_parallel_world_size) - buf_views = [buf[(r*shard_size):((r+1)*shard_size)] - for r in range(data_parallel_world_size)] + buf_views = [ + buf[(r * shard_size) : ((r + 1) * shard_size)] + for r in range(data_parallel_world_size) + ] view_items.append((model_index, dtype, buf, buf_views)) return view_items - def get_model_grad_buffer_dp_views(self): - return self.get_model_buffer_dp_views([ - {dtype : mem_buffer.data} - for model in self.models - for dtype, mem_buffer in model._grad_buffers.items()]) - + return self.get_model_buffer_dp_views( + [ + {dtype: mem_buffer.data} + for model in self.models + for dtype, mem_buffer in model._grad_buffers.items() + ] + ) def get_model_param_buffer_dp_views(self): return self.get_model_buffer_dp_views(self.param_buffers) - def reduce_model_grads(self, args, timers): """ Reduce-scatter model grads. @@ -826,20 +876,23 @@ def reduce_model_grads(self, args, timers): """ # All-reduce layer-norm grads (for sequence parallelism). - timers('layernorm-grads-all-reduce', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("layernorm-grads-all-reduce", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.allreduce_layernorm_grads(args) - timers('layernorm-grads-all-reduce').stop() + timers("layernorm-grads-all-reduce").stop() # All-reduce embedding grads. - timers('embedding-grads-all-reduce', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("embedding-grads-all-reduce", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.allreduce_embedding_grads(args) - timers('embedding-grads-all-reduce').stop() + timers("embedding-grads-all-reduce").stop() # Reduce-scatter setup. - timers('grads-reduce-scatter', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("grads-reduce-scatter", log_level=1).start( + barrier=args.barrier_with_L1_time + ) data_parallel_rank = mpu.get_data_parallel_rank() data_parallel_world_size = mpu.get_data_parallel_world_size() data_parallel_group = mpu.get_data_parallel_group() @@ -851,17 +904,15 @@ def reduce_model_grads(self, args, timers): # Reduce-scatter all grads. gbuf_view_items = self.get_model_grad_buffer_dp_views() - for index, (model_index, dtype, gbuf, gbuf_views) \ - in enumerate(gbuf_view_items): + for index, (model_index, dtype, gbuf, gbuf_views) in enumerate(gbuf_view_items): torch.distributed._reduce_scatter_base( gbuf_views[data_parallel_rank], gbuf, - group = data_parallel_group, + group=data_parallel_group, ) - timers('grads-reduce-scatter').stop() - + timers("grads-reduce-scatter").stop() def gather_model_params(self, args, timers): """ @@ -872,8 +923,9 @@ def gather_model_params(self, args, timers): can be copied from the param buffer to the param. """ - timers('params-all-gather', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("params-all-gather", log_level=1).start( + barrier=args.barrier_with_L1_time + ) data_parallel_rank = mpu.get_data_parallel_rank() data_parallel_group = mpu.get_data_parallel_group() @@ -885,20 +937,19 @@ def gather_model_params(self, args, timers): # all sub-views will have consistent start/end indexes across data # parallel ranks. pbuf_view_items = self.get_model_param_buffer_dp_views() - for index, (model_index, dtype, pbuf, pbuf_views) \ - in enumerate(pbuf_view_items): + for index, (model_index, dtype, pbuf, pbuf_views) in enumerate(pbuf_view_items): - if version.parse(torch.__version__) >= version.parse('1.13'): + if version.parse(torch.__version__) >= version.parse("1.13"): torch.distributed.all_gather_into_tensor( pbuf, pbuf_views[data_parallel_rank], - group = data_parallel_group, + group=data_parallel_group, ) else: torch.distributed._all_gather_base( pbuf, pbuf_views[data_parallel_rank], - group = data_parallel_group, + group=data_parallel_group, ) # Copy from param buffer to each param. @@ -909,8 +960,7 @@ def gather_model_params(self, args, timers): param_buf_shard = param_buf[buf_start:buf_end] param.view(-1).detach().copy_(param_buf_shard) - timers('params-all-gather').stop() - + timers("params-all-gather").stop() def _collect_main_grad_data_for_unscaling(self): """ @@ -923,21 +973,20 @@ def _collect_main_grad_data_for_unscaling(self): for param in group["params"] ] - def _get_model_and_main_params_data_float16(self): """ Get aligned list of model and main params. """ model_data = [] main_data = [] - for model_group, main_group in zip(self.shard_float16_groups, - self.shard_fp32_from_float16_groups): + for model_group, main_group in zip( + self.shard_float16_groups, self.shard_fp32_from_float16_groups + ): for model_param, main_param in zip(model_group, main_group): model_data.append(model_param.data) main_data.append(main_param.data) return model_data, main_data - def _copy_model_grads_to_main_grads(self): """ Copy model grads to main grads. @@ -949,26 +998,22 @@ def _copy_model_grads_to_main_grads(self): # Utility method for copying group grads. def copy_group_grads(model_groups, shard_main_groups): - for model_group, shard_main_group in zip(model_groups, - shard_main_groups): - for model_param, shard_main_param in zip(model_group, - shard_main_group): + for model_group, shard_main_group in zip(model_groups, shard_main_groups): + for model_param, shard_main_param in zip(model_group, shard_main_group): param_range_map = self.get_model_param_range_map(model_param) param_range = param_range_map["param"] assert param_range.size == shard_main_param.nelement() model_grad = model_param.main_grad - shard_model_grad = model_grad.view(-1) \ - [param_range.start:param_range.end] + shard_model_grad = model_grad.view(-1)[ + param_range.start : param_range.end + ] shard_main_param.grad = shard_model_grad.float() # Copy model groups to shard groups. - copy_group_grads(self.model_float16_groups, - self.shard_fp32_from_float16_groups) - copy_group_grads(self.model_fp32_groups, - self.shard_fp32_groups) - + copy_group_grads(self.model_float16_groups, self.shard_fp32_from_float16_groups) + copy_group_grads(self.model_fp32_groups, self.shard_fp32_groups) def _copy_main_params_to_model_params(self): """ @@ -981,10 +1026,8 @@ def _copy_main_params_to_model_params(self): # Utility method for copying group params. def copy_group_params(shard_main_groups, model_groups): - for shard_main_group, model_group in zip(shard_main_groups, - model_groups): - for shard_main_param, model_param in zip(shard_main_group, - model_group): + for shard_main_group, model_group in zip(shard_main_groups, model_groups): + for shard_main_param, model_param in zip(shard_main_group, model_group): param_range_map = self.get_model_param_range_map(model_param) world_range = param_range_map["gbuf_world"] @@ -994,17 +1037,17 @@ def copy_group_params(shard_main_groups, model_groups): model_id, dtype = self.model_param_gbuf_map[model_param] model_param_buffer = self.param_buffers[model_id][dtype] - shard_model_param = model_param_buffer.view(-1) \ - [world_range.start:world_range.end] + shard_model_param = model_param_buffer.view(-1)[ + world_range.start : world_range.end + ] shard_model_param.data.copy_(shard_main_param) # Copy shard groups to model groups. - copy_group_params(self.shard_fp32_from_float16_groups, - self.model_float16_groups) - copy_group_params(self.shard_fp32_groups, - self.model_fp32_groups) - + copy_group_params( + self.shard_fp32_from_float16_groups, self.model_float16_groups + ) + copy_group_params(self.shard_fp32_groups, self.model_fp32_groups) def _copy_model_params_to_main_params(self): """ @@ -1017,21 +1060,20 @@ def _copy_model_params_to_main_params(self): # Utility method for copying group params. def copy_group_params(model_groups, shard_main_groups): - for model_group, shard_main_group in zip(model_groups, - shard_main_groups): - for model_param, shard_main_param in zip(model_group, - shard_main_group): + for model_group, shard_main_group in zip(model_groups, shard_main_groups): + for model_param, shard_main_param in zip(model_group, shard_main_group): param_range_map = self.get_model_param_range_map(model_param) param_range = param_range_map["param"] assert param_range.size == shard_main_param.nelement() - shard_model_param = model_param.view(-1) \ - [param_range.start:param_range.end] + shard_model_param = model_param.view(-1)[ + param_range.start : param_range.end + ] shard_main_param.data.copy_(shard_model_param) # Copy model groups to shard groups. - copy_group_params(self.model_float16_groups, - self.shard_fp32_from_float16_groups) - copy_group_params(self.model_fp32_groups, - self.shard_fp32_groups) + copy_group_params( + self.model_float16_groups, self.shard_fp32_from_float16_groups + ) + copy_group_params(self.model_fp32_groups, self.shard_fp32_groups) diff --git a/megatron/optimizer/grad_scaler.py b/megatron/optimizer/grad_scaler.py index 2babebc632f..4f5e2f1509f 100644 --- a/megatron/optimizer/grad_scaler.py +++ b/megatron/optimizer/grad_scaler.py @@ -8,6 +8,7 @@ import torch from deepspeed.accelerator import get_accelerator + class MegatronGradScaler(ABC): def __init__(self, initial_scale): @@ -36,7 +37,6 @@ def load_state_dict(self, state_dict): pass - class ConstantGradScaler(MegatronGradScaler): def update(self, found_inf): @@ -49,13 +49,18 @@ def load_state_dict(self, state_dict): pass - class DynamicGradScaler(MegatronGradScaler): - def __init__(self, initial_scale, min_scale, - growth_factor, backoff_factor, - growth_interval, hysteresis): - """"Grad scaler with dynamic scale that gets adjusted + def __init__( + self, + initial_scale, + min_scale, + growth_factor, + backoff_factor, + growth_interval, + hysteresis, + ): + """ "Grad scaler with dynamic scale that gets adjusted during training.""" super(DynamicGradScaler, self).__init__(initial_scale) @@ -82,7 +87,6 @@ def __init__(self, initial_scale, min_scale, self._growth_tracker = 0 self._hysteresis_tracker = self.hysteresis - def update(self, found_inf): # If we have an inf/nan, growth tracker is set to 0 @@ -92,8 +96,9 @@ def update(self, found_inf): self._hysteresis_tracker -= 1 # Now if we are out of hysteresis count, scale down the loss. if self._hysteresis_tracker <= 0: - self._scale = torch.max(self._scale * self.backoff_factor, - self.min_scale) + self._scale = torch.max( + self._scale * self.backoff_factor, self.min_scale + ) else: # If there is no nan/inf, increment the growth tracker. self._growth_tracker += 1 @@ -105,16 +110,14 @@ def update(self, found_inf): # and scale up the loss scale. self._scale = self._scale * self.growth_factor - def state_dict(self): state_dict = {} - state_dict['scale'] = self._scale - state_dict['growth_tracker'] = self._growth_tracker - state_dict['hysteresis_tracker'] = self._hysteresis_tracker + state_dict["scale"] = self._scale + state_dict["growth_tracker"] = self._growth_tracker + state_dict["hysteresis_tracker"] = self._hysteresis_tracker return state_dict - def load_state_dict(self, state_dict): - self._scale = state_dict['scale'].to(get_accelerator().current_device_name()) - self._growth_tracker = state_dict['growth_tracker'] - self._hysteresis_tracker = state_dict['hysteresis_tracker'] + self._scale = state_dict["scale"].to(get_accelerator().current_device_name()) + self._growth_tracker = state_dict["growth_tracker"] + self._hysteresis_tracker = state_dict["hysteresis_tracker"] diff --git a/megatron/optimizer/optimizer.py b/megatron/optimizer/optimizer.py index 760d49383e5..24ae1de15e4 100644 --- a/megatron/optimizer/optimizer.py +++ b/megatron/optimizer/optimizer.py @@ -19,6 +19,7 @@ from .clip_grads import clip_grad_norm_fp32, count_zeros_fp32 from deepspeed.accelerator import get_accelerator + def _zero_grad_group_helper(group, set_to_none): """Zero out the gradient for a group of parameters. Note: copied from torch.optim.optimizer.""" @@ -39,34 +40,32 @@ def _multi_tensor_copy_this_to_that(this, that, overflow_buf=None): We don't have a blfoat16 implementation so for now if the overflow_buf is not provided, we default back to simple loop copy to be compatible with bfloat16.""" - if get_accelerator().device_name() == 'cuda' and overflow_buf: + if get_accelerator().device_name() == "cuda" and overflow_buf: from apex.multi_tensor_apply import multi_tensor_applier import amp_C overflow_buf.fill_(0) # Scaling with factor `1.0` is equivalent to copy. - multi_tensor_applier(amp_C.multi_tensor_scale, - overflow_buf, - [this, that], - 1.0) + multi_tensor_applier(amp_C.multi_tensor_scale, overflow_buf, [this, that], 1.0) else: for this_, that_ in zip(this, that): that_.copy_(this_) - class MegatronOptimizer(ABC): - - def __init__(self, optimizer, clip_grad, - log_num_zeros_in_grad, - params_have_main_grad, - use_contiguous_buffers_in_local_ddp, - models): - + def __init__( + self, + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + models, + ): """Input optimizer is the base optimizer for example Adam.""" self.optimizer = optimizer - assert self.optimizer, 'no optimizer is provided.' + assert self.optimizer, "no optimizer is provided." # Set gradient clipping and logging params. self.clip_grad = clip_grad self.log_num_zeros_in_grad = log_num_zeros_in_grad @@ -78,18 +77,17 @@ def __init__(self, optimizer, clip_grad, self.models = models if self.use_contiguous_buffers_in_local_ddp: - assert self.params_have_main_grad, \ - "use of contiguous buffer requires that params have main grad" - + assert ( + self.params_have_main_grad + ), "use of contiguous buffer requires that params have main grad" def get_parameters(self): params = [] for param_group in self.optimizer.param_groups: - for param in param_group['params']: + for param in param_group["params"]: params.append(param) return params - def get_main_grads_for_grad_norm(self): # Filter parameters based on: @@ -102,48 +100,47 @@ def get_main_grads_for_grad_norm(self): grad = param.grad grad_not_none = grad is not None is_not_shared = param_is_not_shared(param) - is_not_tp_duplicate = tensor_parallel.param_is_not_tensor_parallel_duplicate(param) + is_not_tp_duplicate = ( + tensor_parallel.param_is_not_tensor_parallel_duplicate(param) + ) if grad_not_none and is_not_shared and is_not_tp_duplicate: grads_for_norm.append(grad) return grads_for_norm - def get_model_parallel_group(self): """Default returned here, but the distributed optimizer overrides this.""" return mpu.get_model_parallel_group() - def clip_grad_norm(self, clip_grad): params = self.get_parameters() grads_for_norm = self.get_main_grads_for_grad_norm() return clip_grad_norm_fp32( - params, grads_for_norm, clip_grad, - model_parallel_group=self.get_model_parallel_group()) - + params, + grads_for_norm, + clip_grad, + model_parallel_group=self.get_model_parallel_group(), + ) def count_zeros(self): params = self.get_parameters() - return count_zeros_fp32(params, - model_parallel_group=self.get_model_parallel_group()) - + return count_zeros_fp32( + params, model_parallel_group=self.get_model_parallel_group() + ) @abstractmethod def zero_grad(self, set_to_none=True): pass - @abstractmethod def get_loss_scale(self): """The output should be a cuda tensor of size 1.""" pass - def scale_loss(self, loss): """Simple scaling.""" return self.get_loss_scale() * loss - @abstractmethod def reload_model_params(self): """Refreshes any internal state from the current model parameters. @@ -153,17 +150,14 @@ def reload_model_params(self): with main parameters, the main parameters need to also be updated.""" pass - @abstractmethod def state_dict(self): pass - @abstractmethod def load_state_dict(self, state_dict): pass - # Promote state so it can be retrieved or set via # "optimizer_instance.state" def _get_state(self): @@ -174,7 +168,6 @@ def _set_state(self, value): state = property(_get_state, _set_state) - # Promote param_groups so it can be retrieved or set via # "optimizer_instance.param_groups" # (for example, to adjust the learning rate) @@ -186,12 +179,10 @@ def _set_param_groups(self, value): param_groups = property(_get_param_groups, _set_param_groups) - @abstractmethod def step(self, args, timers): pass - def gather_model_params(self, args, timers): """ For the case of a non-distributed-optimizer, there is nothing to @@ -199,7 +190,6 @@ def gather_model_params(self, args, timers): """ pass - def allreduce_word_embedding_grads(self, args): """ All-reduce word embedding grads. @@ -209,8 +199,10 @@ def allreduce_word_embedding_grads(self, args): pipelined model parallelism (BERT and GPT-2). """ - if mpu.is_rank_in_embedding_group(ignore_virtual=True) and \ - mpu.get_pipeline_model_parallel_world_size() > 1: + if ( + mpu.is_rank_in_embedding_group(ignore_virtual=True) + and mpu.get_pipeline_model_parallel_world_size() > 1 + ): if mpu.is_pipeline_first_stage(ignore_virtual=True): unwrapped_model = self.models[0] elif mpu.is_pipeline_last_stage(ignore_virtual=True): @@ -218,17 +210,17 @@ def allreduce_word_embedding_grads(self, args): else: # We do not support the interleaved schedule for T5 yet. unwrapped_model = self.models[0] unwrapped_model = unwrap_model( - unwrapped_model, (torchDDP, LocalDDP, Float16Module)) + unwrapped_model, (torchDDP, LocalDDP, Float16Module) + ) if unwrapped_model.share_embeddings_and_output_weights: weight = unwrapped_model.shared_embedding_or_output_weight() - if args.DDP_impl == 'local': + if args.DDP_impl == "local": grad = weight.main_grad else: grad = weight.grad torch.distributed.all_reduce(grad, group=mpu.get_embedding_group()) - def allreduce_position_embedding_grads(self, args): """ All-reduce position_embeddings grad across first (encoder) and @@ -236,69 +228,77 @@ def allreduce_position_embedding_grads(self, args): stay in sync. This should only run for T5 models with pipeline parallelism. """ - if mpu.is_rank_in_position_embedding_group() and \ - mpu.get_pipeline_model_parallel_world_size() > 1 and \ - args.pipeline_model_parallel_split_rank is not None: + if ( + mpu.is_rank_in_position_embedding_group() + and mpu.get_pipeline_model_parallel_world_size() > 1 + and args.pipeline_model_parallel_split_rank is not None + ): unwrapped_model = self.models[0] unwrapped_model = unwrap_model( - unwrapped_model, (torchDDP, LocalDDP, Float16Module)) - assert args.DDP_impl == 'local', \ - 'T5 model is only supported with local DDP mode' - grad = unwrapped_model.language_model.embedding.position_embeddings.weight.main_grad + unwrapped_model, (torchDDP, LocalDDP, Float16Module) + ) + assert ( + args.DDP_impl == "local" + ), "T5 model is only supported with local DDP mode" + grad = ( + unwrapped_model.language_model.embedding.position_embeddings.weight.main_grad + ) torch.distributed.all_reduce(grad, group=mpu.get_position_embedding_group()) - def allreduce_embedding_grads(self, args): """All-reduce both word and position embeddings.""" self.allreduce_word_embedding_grads(args) self.allreduce_position_embedding_grads(args) - def allreduce_layernorm_grads(self, args): """All-reduce layernorm grads (for sequence parallelism).""" # All-reduce layernorm parameters across model parallel nodes # when sequence parallelism is used - if mpu.get_tensor_model_parallel_world_size() > 1 and \ - args.sequence_parallel: + if mpu.get_tensor_model_parallel_world_size() > 1 and args.sequence_parallel: grads = [] for model_module in self.models: - unwrapped_model = unwrap_model( - model_module, (torchDDP, LocalDDP, Float16Module)) + unwrapped_model = unwrap_model( + model_module, (torchDDP, LocalDDP, Float16Module) + ) for param in unwrapped_model.parameters(): - if getattr(param, 'sequence_parallel', False): - grad = param.main_grad if args.DDP_impl == 'local' else param.grad + if getattr(param, "sequence_parallel", False): + grad = ( + param.main_grad if args.DDP_impl == "local" else param.grad + ) grads.append(grad.data) coalesced = _flatten_dense_tensors(grads) torch.distributed.all_reduce( - coalesced, group=mpu.get_tensor_model_parallel_group()) - for buf, synced in zip(grads, _unflatten_dense_tensors( - coalesced, grads)): + coalesced, group=mpu.get_tensor_model_parallel_group() + ) + for buf, synced in zip(grads, _unflatten_dense_tensors(coalesced, grads)): buf.copy_(synced) - def reduce_model_grads(self, args, timers): """All-reduce all grads, and all-reduce embeddings.""" # All-reduce layer-norm grads (for sequence parallelism). - timers('layernorm-grads-all-reduce', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("layernorm-grads-all-reduce", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.allreduce_layernorm_grads(args) - timers('layernorm-grads-all-reduce').stop() + timers("layernorm-grads-all-reduce").stop() # All-reduce if needed. - if args.DDP_impl == 'local': - timers('grads-all-reduce', log_level=1).start( - barrier=args.barrier_with_L1_time) + if args.DDP_impl == "local": + timers("grads-all-reduce", log_level=1).start( + barrier=args.barrier_with_L1_time + ) for model in self.models: model.allreduce_gradients() - timers('grads-all-reduce').stop() + timers("grads-all-reduce").stop() # All-reduce embedding grads. - timers('embedding-grads-all-reduce', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("embedding-grads-all-reduce", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.allreduce_embedding_grads(args) - timers('embedding-grads-all-reduce').stop() + timers("embedding-grads-all-reduce").stop() class MixedPrecisionOptimizer(MegatronOptimizer): @@ -332,15 +332,28 @@ class MixedPrecisionOptimizer(MegatronOptimizer): is used by the distributed optimizer for mapping parameters. """ - def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - fp16, bf16, params_dtype, grad_scaler, - models): + def __init__( + self, + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + fp16, + bf16, + params_dtype, + grad_scaler, + models, + ): super().__init__( - optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - models) + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + models, + ) self.fp16 = fp16 self.bf16 = bf16 @@ -349,7 +362,7 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, # None grad scaler is only supported for bf16. if self.grad_scaler is None: - assert not self.fp16, 'fp16 expects a grad scaler.' + assert not self.fp16, "fp16 expects a grad scaler." # Tensor used to determine if a nan/if has happend. # Any non-zero value indicates inf/nan. @@ -370,17 +383,14 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, if self.grad_scaler is None: self._scale_one = get_accelerator().FloatTensor([1.0]) - def get_loss_scale(self): if self.grad_scaler is None: return self._scale_one return self.grad_scaler.scale - def reload_model_params(self): self._copy_model_params_to_main_params() - def _unscale_main_grads_and_check_for_nan(self): # Collect main grads. @@ -391,37 +401,41 @@ def _unscale_main_grads_and_check_for_nan(self): # Unscale and set found inf/nan torch._amp_foreach_non_finite_check_and_unscale_( - main_grads, self.found_inf, self.grad_scaler.inv_scale) + main_grads, self.found_inf, self.grad_scaler.inv_scale + ) # Update across all model parallel instances. - torch.distributed.all_reduce(self.found_inf, - op=torch.distributed.ReduceOp.MAX, - group=self.get_model_parallel_group()) + torch.distributed.all_reduce( + self.found_inf, + op=torch.distributed.ReduceOp.MAX, + group=self.get_model_parallel_group(), + ) # Check for nan. - found_inf_flag = (self.found_inf.item() > 0) + found_inf_flag = self.found_inf.item() > 0 return found_inf_flag - @torch.no_grad() def step(self, args, timers): # Copy gradients from model params to main params. - timers('optimizer-copy-to-main-grad', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-copy-to-main-grad", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self._copy_model_grads_to_main_grads() - timers('optimizer-copy-to-main-grad').stop() + timers("optimizer-copy-to-main-grad").stop() # Do unscale, check for inf, and update grad scaler only for # the case that grad scaler is provided. if self.grad_scaler: # Unscale and check for inf/nan. - timers('optimizer-unscale-and-check-inf', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-unscale-and-check-inf", log_level=1).start( + barrier=args.barrier_with_L1_time + ) found_inf_flag = self._unscale_main_grads_and_check_for_nan() - timers('optimizer-unscale-and-check-inf').stop() + timers("optimizer-unscale-and-check-inf").stop() # We are done with scaling gradients # so we can update the loss scale. @@ -432,31 +446,34 @@ def step(self, args, timers): return False, None, None # Clip the main gradients. - timers('optimizer-clip-main-grad', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-clip-main-grad", log_level=1).start( + barrier=args.barrier_with_L1_time + ) grad_norm = None if self.clip_grad > 0.0: grad_norm = self.clip_grad_norm(self.clip_grad) - timers('optimizer-clip-main-grad').stop() + timers("optimizer-clip-main-grad").stop() # Count the zeros in the grads. - timers('optimizer-count-zeros', log_level=1).start( - barrier=args.barrier_with_L1_time) - num_zeros_in_grad = self.count_zeros() if \ - self.log_num_zeros_in_grad else None - timers('optimizer-count-zeros').stop() + timers("optimizer-count-zeros", log_level=1).start( + barrier=args.barrier_with_L1_time + ) + num_zeros_in_grad = self.count_zeros() if self.log_num_zeros_in_grad else None + timers("optimizer-count-zeros").stop() # Step the optimizer. - timers('optimizer-inner-step', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-inner-step", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.optimizer.step() - timers('optimizer-inner-step').stop() + timers("optimizer-inner-step").stop() # Update params from main params. - timers('optimizer-copy-main-to-model-params', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-copy-main-to-model-params", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self._copy_main_params_to_model_params() - timers('optimizer-copy-main-to-model-params').stop() + timers("optimizer-copy-main-to-model-params").stop() # Successful update. return True, grad_norm, num_zeros_in_grad @@ -492,14 +509,32 @@ class Float16OptimizerWithFloat16Params(MixedPrecisionOptimizer): is used by the distributed optimizer for mapping parameters. """ - def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - fp16, bf16, params_dtype, grad_scaler, models): + def __init__( + self, + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + fp16, + bf16, + params_dtype, + grad_scaler, + models, + ): super().__init__( - optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - fp16, bf16, params_dtype, grad_scaler, models) + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + fp16, + bf16, + params_dtype, + grad_scaler, + models, + ) # ====================== # main parameter stuff @@ -519,49 +554,58 @@ def __init__(self, optimizer, clip_grad, log_num_zeros_in_grad, fp32_params_this_group = [] fp32_from_float16_params_this_group = [] # For all the parameters in this group: - for i, param in enumerate(param_group['params']): + for i, param in enumerate(param_group["params"]): if param.requires_grad: # float16 params: - - if param.type() in ['torch.{}.HalfTensor'.format(get_accelerator().device_name()), - 'torch.{}.BFloat16Tensor'.format(get_accelerator().device_name())]: + if param.type() in [ + "torch.{}.HalfTensor".format(get_accelerator().device_name()), + "torch.{}.BFloat16Tensor".format( + get_accelerator().device_name() + ), + ]: float16_params_this_group.append(param) # Create a copy main_param = param.detach().clone().float() # Copy tensor model parallel attributes. - tensor_parallel.copy_tensor_model_parallel_attributes(main_param, - param) - if hasattr(param, 'shared'): + tensor_parallel.copy_tensor_model_parallel_attributes( + main_param, param + ) + if hasattr(param, "shared"): main_param.shared = param.shared # Replace the optimizer params with the new fp32 copy. - param_group['params'][i] = main_param + param_group["params"][i] = main_param fp32_from_float16_params_this_group.append(main_param) # Reset existing state dict key to the new main param. if param in self.optimizer.state: - self.optimizer.state[main_param] \ - = self.optimizer.state.pop(param) + self.optimizer.state[main_param] = self.optimizer.state.pop( + param + ) # fp32 params. - elif param.type() == 'torch.{}.FloatTensor'.format(format(get_accelerator().device_name())): + elif param.type() == "torch.{}.FloatTensor".format( + format(get_accelerator().device_name()) + ): fp32_params_this_group.append(param) - param_group['params'][i] = param + param_group["params"][i] = param else: device_name = get_accelerator().device_name() - raise TypeError('Wrapped parameters must be one of ' - 'torch.{}.FloatTensor, ' - 'torch.{}.HalfTensor, or ' - 'torch.{}.BFloat16Tensor. ' - 'Received {}'.format(device_name,device_name,device_name,param.type())) + raise TypeError( + "Wrapped parameters must be one of " + "torch.{}.FloatTensor, " + "torch.{}.HalfTensor, or " + "torch.{}.BFloat16Tensor. " + "Received {}".format( + device_name, device_name, device_name, param.type() + ) + ) self.float16_groups.append(float16_params_this_group) - self.fp32_from_float16_groups.append( - fp32_from_float16_params_this_group) + self.fp32_from_float16_groups.append(fp32_from_float16_params_this_group) self.fp32_from_fp32_groups.append(fp32_params_this_group) - def zero_grad(self, set_to_none=True): """We only need to zero the model related parameters, i.e., float16_groups & fp32_from_fp32_groups. We additionally zero @@ -575,7 +619,6 @@ def zero_grad(self, set_to_none=True): for group in self.fp32_from_fp32_groups: _zero_grad_group_helper(group, set_to_none) - def _collect_main_grad_data_for_unscaling(self): main_grads = [] @@ -591,27 +634,27 @@ def _collect_main_grad_data_for_unscaling(self): for main_param in main_group: if main_param.grad is not None: main_grads.append(main_param.grad.data) - - return main_grads + return main_grads def _get_model_and_main_params_data_float16(self): model_data = [] main_data = [] - for model_group, main_group in zip(self.float16_groups, - self.fp32_from_float16_groups): + for model_group, main_group in zip( + self.float16_groups, self.fp32_from_float16_groups + ): for model_param, main_param in zip(model_group, main_group): model_data.append(model_param.data) main_data.append(main_param.data) return model_data, main_data - def _copy_model_grads_to_main_grads(self): # This only needs to be done for the float16 group. - for model_group, main_group in zip(self.float16_groups, - self.fp32_from_float16_groups): + for model_group, main_group in zip( + self.float16_groups, self.fp32_from_float16_groups + ): for model_param, main_param in zip(model_group, main_group): - if self.params_have_main_grad and hasattr(model_param, 'main_grad'): + if self.params_have_main_grad and hasattr(model_param, "main_grad"): main_param.grad = model_param.main_grad.float() else: if model_param.grad is not None: @@ -621,8 +664,10 @@ def _copy_model_grads_to_main_grads(self): # (If using contiguous buffers, main_grad's memory should # persist and therefore should not be deallocated.) model_param.grad = None - if self.params_have_main_grad and \ - not self.use_contiguous_buffers_in_local_ddp: + if ( + self.params_have_main_grad + and not self.use_contiguous_buffers_in_local_ddp + ): model_param.main_grad = None # For fp32 grads, we need to reset the grads to main grad. @@ -637,101 +682,110 @@ def _copy_model_grads_to_main_grads(self): if not self.use_contiguous_buffers_in_local_ddp: model_param.main_grad = None - def _copy_main_params_to_model_params(self): # Only needed for the float16 params. model_data, main_data = self._get_model_and_main_params_data_float16() - _multi_tensor_copy_this_to_that(this=main_data, that=model_data, - overflow_buf=self._dummy_overflow_buf) - + _multi_tensor_copy_this_to_that( + this=main_data, that=model_data, overflow_buf=self._dummy_overflow_buf + ) def _copy_model_params_to_main_params(self): # Only needed for the float16 params. model_data, main_data = self._get_model_and_main_params_data_float16() - _multi_tensor_copy_this_to_that(this=model_data, that=main_data, - overflow_buf=self._dummy_overflow_buf) - + _multi_tensor_copy_this_to_that( + this=model_data, that=main_data, overflow_buf=self._dummy_overflow_buf + ) def state_dict(self): state_dict = {} - state_dict['optimizer'] = self.optimizer.state_dict() + state_dict["optimizer"] = self.optimizer.state_dict() if self.grad_scaler: - state_dict['grad_scaler'] = self.grad_scaler.state_dict() - state_dict['fp32_from_fp16_params'] = self.fp32_from_float16_groups + state_dict["grad_scaler"] = self.grad_scaler.state_dict() + state_dict["fp32_from_fp16_params"] = self.fp32_from_float16_groups return state_dict - def load_state_dict(self, state_dict): # Optimizer. - optimizer_key = 'optimizer' + optimizer_key = "optimizer" if optimizer_key not in state_dict: - optimizer_key = 'optimizer_state_dict' - print_rank_0('***WARNING*** loading optimizer from ' - 'an old checkpoint ...') + optimizer_key = "optimizer_state_dict" + print_rank_0( + "***WARNING*** loading optimizer from " "an old checkpoint ..." + ) self.optimizer.load_state_dict(state_dict[optimizer_key]) # Grad scaler. - if 'grad_scaler' not in state_dict: + if "grad_scaler" not in state_dict: if self.fp16: - print_rank_0('***WARNING*** found an old checkpoint, will not ' - 'load grad scaler ...') + print_rank_0( + "***WARNING*** found an old checkpoint, will not " + "load grad scaler ..." + ) else: if self.grad_scaler: - self.grad_scaler.load_state_dict(state_dict['grad_scaler']) + self.grad_scaler.load_state_dict(state_dict["grad_scaler"]) else: - print_rank_0('***WARNING*** fould the grad scaler in the ' - 'checkpoint but it is None in the class. ' - 'Skipping loading grad scaler ...') + print_rank_0( + "***WARNING*** fould the grad scaler in the " + "checkpoint but it is None in the class. " + "Skipping loading grad scaler ..." + ) # Copy data for the main params. - fp32_from_float16_params_key = 'fp32_from_fp16_params' + fp32_from_float16_params_key = "fp32_from_fp16_params" if fp32_from_float16_params_key not in state_dict: - fp32_from_float16_params_key = 'fp32_from_fp16' + fp32_from_float16_params_key = "fp32_from_fp16" for current_group, saved_group in zip( - self.fp32_from_float16_groups, - state_dict[fp32_from_float16_params_key]): + self.fp32_from_float16_groups, state_dict[fp32_from_float16_params_key] + ): for current_param, saved_param in zip(current_group, saved_group): current_param.data.copy_(saved_param.data) class FP32Optimizer(MegatronOptimizer): - def __init__(self, optimizer, clip_grad, - log_num_zeros_in_grad, - params_have_main_grad, - use_contiguous_buffers_in_local_ddp, - models): + def __init__( + self, + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + models, + ): super(FP32Optimizer, self).__init__( - optimizer, clip_grad, log_num_zeros_in_grad, - params_have_main_grad, use_contiguous_buffers_in_local_ddp, - models) + optimizer, + clip_grad, + log_num_zeros_in_grad, + params_have_main_grad, + use_contiguous_buffers_in_local_ddp, + models, + ) self._scale = get_accelerator().FloatTensor([1.0]) - def zero_grad(self, set_to_none=True): """Copied from torch.optim.optimizer""" for group in self.optimizer.param_groups: - _zero_grad_group_helper(group['params'], set_to_none) - + _zero_grad_group_helper(group["params"], set_to_none) def get_loss_scale(self): """FP32 optimizer does not do any scaling.""" return self._scale - @torch.no_grad() def step(self, args, timers): """Clip gradients (if needed) and step the base optimizer. Always return successful since there is no overflow.""" # Copy main_grads to grads. - timers('optimizer-copy-to-main-grad', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-copy-to-main-grad", log_level=1).start( + barrier=args.barrier_with_L1_time + ) if self.params_have_main_grad: for param_group in self.optimizer.param_groups: - for param in param_group['params']: + for param in param_group["params"]: param.grad = param.main_grad # Safe to de-reference model's main_grad after copying. @@ -739,40 +793,39 @@ def step(self, args, timers): # persist and therefore should not be deallocated.) if not self.use_contiguous_buffers_in_local_ddp: param.main_grad = None - timers('optimizer-copy-to-main-grad').stop() + timers("optimizer-copy-to-main-grad").stop() # Clip gradients. - timers('optimizer-clip-main-grad', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-clip-main-grad", log_level=1).start( + barrier=args.barrier_with_L1_time + ) grad_norm = None if self.clip_grad > 0.0: grad_norm = self.clip_grad_norm(self.clip_grad) - timers('optimizer-clip-main-grad').stop() + timers("optimizer-clip-main-grad").stop() # count the zeros in the grads - timers('optimizer-count-zeros', log_level=1).start( - barrier=args.barrier_with_L1_time) - num_zeros_in_grad = self.count_zeros() if \ - self.log_num_zeros_in_grad else None - timers('optimizer-count-zeros').stop() + timers("optimizer-count-zeros", log_level=1).start( + barrier=args.barrier_with_L1_time + ) + num_zeros_in_grad = self.count_zeros() if self.log_num_zeros_in_grad else None + timers("optimizer-count-zeros").stop() # Update parameters. - timers('optimizer-inner-step', log_level=1).start( - barrier=args.barrier_with_L1_time) + timers("optimizer-inner-step", log_level=1).start( + barrier=args.barrier_with_L1_time + ) self.optimizer.step() - timers('optimizer-inner-step').stop() + timers("optimizer-inner-step").stop() # No overflow for FP32 optimizer. return True, grad_norm, num_zeros_in_grad - def reload_model_params(self): pass - def state_dict(self): return self.optimizer.state_dict() - def load_state_dict(self, state_dict): self.optimizer.load_state_dict(state_dict) diff --git a/megatron/optimizer/sophia.py b/megatron/optimizer/sophia.py index 4c4e0747907..0789e0d9f58 100644 --- a/megatron/optimizer/sophia.py +++ b/megatron/optimizer/sophia.py @@ -5,12 +5,22 @@ from typing import List, Optional -#SOphiaG implementation from https://github.com/Liuhong99/Sophia/blob/main/sophia.py, copy pasted here because no pip and not sure about submodules +# SOphiaG implementation from https://github.com/Liuhong99/Sophia/blob/main/sophia.py, copy pasted here because no pip and not sure about submodules + class SophiaG(Optimizer): - def __init__(self, params, lr=1e-4, betas=(0.965, 0.99), rho = 0.04, - weight_decay=1e-1, *, maximize: bool = False, - capturable: bool = False): + + def __init__( + self, + params, + lr=1e-4, + betas=(0.965, 0.99), + rho=0.04, + weight_decay=1e-1, + *, + maximize: bool = False, + capturable: bool = False + ): if not 0.0 <= lr: raise ValueError("Invalid learning rate: {}".format(lr)) if not 0.0 <= betas[0] < 1.0: @@ -21,42 +31,57 @@ def __init__(self, params, lr=1e-4, betas=(0.965, 0.99), rho = 0.04, raise ValueError("Invalid rho parameter at index 1: {}".format(rho)) if not 0.0 <= weight_decay: raise ValueError("Invalid weight_decay value: {}".format(weight_decay)) - defaults = dict(lr=lr, betas=betas, rho=rho, - weight_decay=weight_decay, - maximize=maximize, capturable=capturable) + defaults = dict( + lr=lr, + betas=betas, + rho=rho, + weight_decay=weight_decay, + maximize=maximize, + capturable=capturable, + ) super(SophiaG, self).__init__(params, defaults) def __setstate__(self, state): super().__setstate__(state) for group in self.param_groups: - group.setdefault('maximize', False) - group.setdefault('capturable', False) + group.setdefault("maximize", False) + group.setdefault("capturable", False) state_values = list(self.state.values()) - step_is_tensor = (len(state_values) != 0) and torch.is_tensor(state_values[0]['step']) + step_is_tensor = (len(state_values) != 0) and torch.is_tensor( + state_values[0]["step"] + ) if not step_is_tensor: for s in state_values: - s['step'] = torch.tensor(float(s['step'])) - + s["step"] = torch.tensor(float(s["step"])) + @torch.no_grad() def update_hessian(self): for group in self.param_groups: - beta1, beta2 = group['betas'] - for p in group['params']: + beta1, beta2 = group["betas"] + for p in group["params"]: if p.grad is None: continue state = self.state[p] if len(state) == 0: - state['step'] = torch.zeros((1,), dtype=torch.float, device=p.device) \ - if self.defaults['capturable'] else torch.tensor(0.) - state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format) - state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) - - if 'hessian' not in state.keys(): - state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) - - state['hessian'].mul_(beta2).addcmul_(p.grad, p.grad, value=1 - beta2) - + state["step"] = ( + torch.zeros((1,), dtype=torch.float, device=p.device) + if self.defaults["capturable"] + else torch.tensor(0.0) + ) + state["exp_avg"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + state["hessian"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + + if "hessian" not in state.keys(): + state["hessian"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + + state["hessian"].mul_(beta2).addcmul_(p.grad, p.grad, value=1 - beta2) @torch.no_grad() def step(self, closure=None, bs=5120): @@ -71,99 +96,119 @@ def step(self, closure=None, bs=5120): exp_avgs = [] state_steps = [] hessian = [] - beta1, beta2 = group['betas'] + beta1, beta2 = group["betas"] - for p in group['params']: + for p in group["params"]: if p.grad is None: continue params_with_grad.append(p) - + if p.grad.is_sparse: - raise RuntimeError('Hero does not support sparse gradients') + raise RuntimeError("Hero does not support sparse gradients") grads.append(p.grad) state = self.state[p] # State initialization if len(state) == 0: - state['step'] = torch.zeros((1,), dtype=torch.float, device=p.device) \ - if self.defaults['capturable'] else torch.tensor(0.) - state['exp_avg'] = torch.zeros_like(p, memory_format=torch.preserve_format) - state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) - - if 'hessian' not in state.keys(): - state['hessian'] = torch.zeros_like(p, memory_format=torch.preserve_format) - - exp_avgs.append(state['exp_avg']) - state_steps.append(state['step']) - hessian.append(state['hessian']) - - if self.defaults['capturable']: + state["step"] = ( + torch.zeros((1,), dtype=torch.float, device=p.device) + if self.defaults["capturable"] + else torch.tensor(0.0) + ) + state["exp_avg"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + state["hessian"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + + if "hessian" not in state.keys(): + state["hessian"] = torch.zeros_like( + p, memory_format=torch.preserve_format + ) + + exp_avgs.append(state["exp_avg"]) + state_steps.append(state["step"]) + hessian.append(state["hessian"]) + + if self.defaults["capturable"]: bs = torch.ones((1,), dtype=torch.float, device=p.device) * bs - sophiag(params_with_grad, - grads, - exp_avgs, - hessian, - state_steps, - bs=bs, - beta1=beta1, - beta2=beta2, - rho=group['rho'], - lr=group['lr'], - weight_decay=group['weight_decay'], - maximize=group['maximize'], - capturable=group['capturable']) + sophiag( + params_with_grad, + grads, + exp_avgs, + hessian, + state_steps, + bs=bs, + beta1=beta1, + beta2=beta2, + rho=group["rho"], + lr=group["lr"], + weight_decay=group["weight_decay"], + maximize=group["maximize"], + capturable=group["capturable"], + ) return loss -def sophiag(params: List[Tensor], - grads: List[Tensor], - exp_avgs: List[Tensor], - hessian: List[Tensor], - state_steps: List[Tensor], - capturable: bool = False, - *, - bs: int, - beta1: float, - beta2: float, - rho: float, - lr: float, - weight_decay: float, - maximize: bool): + +def sophiag( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + hessian: List[Tensor], + state_steps: List[Tensor], + capturable: bool = False, + *, + bs: int, + beta1: float, + beta2: float, + rho: float, + lr: float, + weight_decay: float, + maximize: bool +): if not all(isinstance(t, torch.Tensor) for t in state_steps): - raise RuntimeError("API has changed, `state_steps` argument must contain a list of singleton tensors") + raise RuntimeError( + "API has changed, `state_steps` argument must contain a list of singleton tensors" + ) - func = _single_tensor_sophiag - func(params, - grads, - exp_avgs, - hessian, - state_steps, - bs=bs, - beta1=beta1, - beta2=beta2, - rho=rho, - lr=lr, - weight_decay=weight_decay, - maximize=maximize, - capturable=capturable) - -def _single_tensor_sophiag(params: List[Tensor], - grads: List[Tensor], - exp_avgs: List[Tensor], - hessian: List[Tensor], - state_steps: List[Tensor], - *, - bs: int, - beta1: float, - beta2: float, - rho: float, - lr: float, - weight_decay: float, - maximize: bool, - capturable: bool): + func( + params, + grads, + exp_avgs, + hessian, + state_steps, + bs=bs, + beta1=beta1, + beta2=beta2, + rho=rho, + lr=lr, + weight_decay=weight_decay, + maximize=maximize, + capturable=capturable, + ) + + +def _single_tensor_sophiag( + params: List[Tensor], + grads: List[Tensor], + exp_avgs: List[Tensor], + hessian: List[Tensor], + state_steps: List[Tensor], + *, + bs: int, + beta1: float, + beta2: float, + rho: float, + lr: float, + weight_decay: float, + maximize: bool, + capturable: bool +): for i, param in enumerate(params): grad = grads[i] if not maximize else -grads[i] @@ -172,8 +217,8 @@ def _single_tensor_sophiag(params: List[Tensor], step_t = state_steps[i] if capturable: - assert param.is_cuda and step_t.is_cuda and bs.is_cuda - + assert param.is_cuda and step_t.is_cuda and bs.is_cuda + if torch.is_complex(param): grad = torch.view_as_real(grad) exp_avg = torch.view_as_real(exp_avg) @@ -188,15 +233,15 @@ def _single_tensor_sophiag(params: List[Tensor], # Decay the first and second moment running average coefficient exp_avg.mul_(beta1).add_(grad, alpha=1 - beta1) - + if capturable: - step_size = lr + step_size = lr step_size_neg = step_size.neg() - ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None,1) + ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None, 1) param.addcmul_(exp_avg.sign(), ratio, value=step_size_neg) else: - step_size_neg = - lr - - ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None,1) + step_size_neg = -lr + + ratio = (exp_avg.abs() / (rho * bs * hess + 1e-15)).clamp(None, 1) param.addcmul_(exp_avg.sign(), ratio, value=step_size_neg) From e37ab7d824f856c1772a6862bcb621da24c716bc Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:14:11 -0500 Subject: [PATCH 435/590] chore: Format `megatron/mpu/*` --- megatron/mpu/tests/commons.py | 40 +- megatron/mpu/tests/test_cross_entropy.py | 80 ++-- megatron/mpu/tests/test_data.py | 35 +- megatron/mpu/tests/test_initialize.py | 35 +- megatron/mpu/tests/test_layers.py | 441 +++++++++++++++-------- megatron/mpu/tests/test_random.py | 84 +++-- 6 files changed, 463 insertions(+), 252 deletions(-) diff --git a/megatron/mpu/tests/commons.py b/megatron/mpu/tests/commons.py index 0fdb5ab0fc8..20b3789ac15 100644 --- a/megatron/mpu/tests/commons.py +++ b/megatron/mpu/tests/commons.py @@ -9,7 +9,9 @@ import mpu from deepspeed.accelerator import get_accelerator + class IdentityLayer(torch.nn.Module): + def __init__(self, size, scale=1.0): super(IdentityLayer, self).__init__() self.weight = torch.nn.Parameter(scale * torch.randn(size)) @@ -26,21 +28,27 @@ def set_random_seed(seed): mpu.model_parallel_cuda_manual_seed(seed) -def initialize_distributed(backend='nccl'): +def initialize_distributed(backend="nccl"): """Initialize torch.distributed.""" # Get local rank in case it is provided. parser = argparse.ArgumentParser() - parser.add_argument('--local_rank', type=int, default=None, - help='local rank passed from distributed launcher') + parser.add_argument( + "--local_rank", + type=int, + default=None, + help="local rank passed from distributed launcher", + ) args = parser.parse_args() local_rank = args.local_rank # Get rank and world size. - rank = int(os.getenv('RANK', '0')) - world_size = int(os.getenv("WORLD_SIZE", '1')) + rank = int(os.getenv("RANK", "0")) + world_size = int(os.getenv("WORLD_SIZE", "1")) - print('> initializing torch.distributed with local rank: {}, ' - 'rank: {}, world size: {}'.format(local_rank, rank, world_size)) + print( + "> initializing torch.distributed with local rank: {}, " + "rank: {}, world size: {}".format(local_rank, rank, world_size) + ) # Set the device id. device = rank % get_accelerator().device_count() @@ -49,22 +57,20 @@ def initialize_distributed(backend='nccl'): get_accelerator().set_device(device) # Call the init process. - init_method = 'tcp://' - master_ip = os.getenv('MASTER_ADDR', 'localhost') - master_port = os.getenv('MASTER_PORT', '6000') - init_method += master_ip + ':' + master_port + init_method = "tcp://" + master_ip = os.getenv("MASTER_ADDR", "localhost") + master_port = os.getenv("MASTER_PORT", "6000") + init_method += master_ip + ":" + master_port torch.distributed.init_process_group( - backend=backend, - world_size=world_size, - rank=rank, - init_method=init_method) + backend=backend, world_size=world_size, rank=rank, init_method=init_method + ) def print_separator(message): torch.distributed.barrier() filler_len = (78 - len(message)) // 2 - filler = '-' * filler_len - string = '\n' + filler + ' {} '.format(message) + filler + filler = "-" * filler_len + string = "\n" + filler + " {} ".format(message) + filler if torch.distributed.get_rank() == 0: print(string, flush=True) torch.distributed.barrier() diff --git a/megatron/mpu/tests/test_cross_entropy.py b/megatron/mpu/tests/test_cross_entropy.py index 7f161348ce6..86b4300cb74 100644 --- a/megatron/mpu/tests/test_cross_entropy.py +++ b/megatron/mpu/tests/test_cross_entropy.py @@ -11,33 +11,44 @@ import random import sys from deepspeed.accelerator import get_accelerator + sys.path.append("../..") -def torch_cross_entropy(batch_size, seq_length, vocab_size, - logits_scale, seed): +def torch_cross_entropy(batch_size, seq_length, vocab_size, logits_scale, seed): set_random_seed(seed) - identity = IdentityLayer((batch_size, seq_length, vocab_size), - scale=logits_scale).to(get_accelerator().device_name()) + identity = IdentityLayer( + (batch_size, seq_length, vocab_size), scale=logits_scale + ).to(get_accelerator().device_name()) logits = identity() - target = get_accelerator().LongTensor( - size=(batch_size, seq_length)).random_(0, vocab_size) - loss = F.cross_entropy(logits.view(-1, logits.size()[-1]), - target.view(-1), - reduction='none').view_as(target).mean() + target = ( + get_accelerator() + .LongTensor(size=(batch_size, seq_length)) + .random_(0, vocab_size) + ) + loss = ( + F.cross_entropy( + logits.view(-1, logits.size()[-1]), target.view(-1), reduction="none" + ) + .view_as(target) + .mean() + ) loss.backward() return loss, identity.weight.grad -def mpu_cross_entropy(batch_size, seq_length, vocab_size, - logits_scale, seed): +def mpu_cross_entropy(batch_size, seq_length, vocab_size, logits_scale, seed): set_random_seed(seed) - identity = IdentityLayer((batch_size, seq_length, vocab_size), - scale=logits_scale).to(get_accelerator().device_name()) + identity = IdentityLayer( + (batch_size, seq_length, vocab_size), scale=logits_scale + ).to(get_accelerator().device_name()) logits = identity() logits_parallel = mpu.scatter_to_tensor_model_parallel_region(logits) - target = get_accelerator().LongTensor( - size=(batch_size, seq_length)).random_(0, vocab_size) + target = ( + get_accelerator() + .LongTensor(size=(batch_size, seq_length)) + .random_(0, vocab_size) + ) loss = vocab_parallel_cross_entropy(logits_parallel, target).mean() loss.backward() return loss, identity.weight.grad @@ -46,8 +57,11 @@ def mpu_cross_entropy(batch_size, seq_length, vocab_size, def test_cross_entropy(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing cross entropy with model parallel size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing cross entropy with model parallel size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -59,21 +73,27 @@ def test_cross_entropy(tensor_model_parallel_size): vocab_size = vocab_size_per_partition * tensor_model_parallel_size seed = 1234 - loss_torch, grad_torch = torch_cross_entropy(batch_size, seq_length, - vocab_size, logits_scale, - seed) - loss_mpu, grad_mpu = mpu_cross_entropy(batch_size, seq_length, - vocab_size, logits_scale, - seed) + loss_torch, grad_torch = torch_cross_entropy( + batch_size, seq_length, vocab_size, logits_scale, seed + ) + loss_mpu, grad_mpu = mpu_cross_entropy( + batch_size, seq_length, vocab_size, logits_scale, seed + ) error = loss_torch.sub_(loss_mpu).abs().max() - print(' max error in loss on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " max error in loss on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 error = grad_torch.sub_(grad_mpu).abs().max() - print(' max error in grad on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " max error in grad on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 # Reset groups @@ -81,16 +101,16 @@ def test_cross_entropy(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") -if __name__ == '__main__': +if __name__ == "__main__": initialize_distributed() world_size = torch.distributed.get_world_size() tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test cross entropy') + print_separator("test cross entropy") test_cross_entropy(tensor_model_parallel_size) tensor_model_parallel_size *= 2 diff --git a/megatron/mpu/tests/test_data.py b/megatron/mpu/tests/test_data.py index 1e95447099f..2637bd7bdd5 100644 --- a/megatron/mpu/tests/test_data.py +++ b/megatron/mpu/tests/test_data.py @@ -9,24 +9,30 @@ import functools import operator import sys + sys.path.append("../..") def test_broadcast_data(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing broadcast_data with model parallel size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing broadcast_data with model parallel size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) torch.manual_seed(1234 + mpu.get_data_parallel_rank()) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() - key_size_t = {'key1': [7, 11], - 'key2': [8, 2, 1], - 'key3': [13], - 'key4': [5, 1, 2], - 'key5': [5, 12]} + key_size_t = { + "key1": [7, 11], + "key2": [8, 2, 1], + "key3": [13], + "key4": [5, 1, 2], + "key5": [5, 12], + } keys = list(key_size_t.keys()) data = {} @@ -34,14 +40,15 @@ def test_broadcast_data(tensor_model_parallel_size): for key in key_size_t: data[key] = torch.LongTensor(size=key_size_t[key]).random_(0, 1000) data_t[key] = data[key].clone() - data['keyX'] = torch.FloatTensor(size=(5, )).random_(0, 1000) - data_t['keyX'] = data['keyX'].clone() + data["keyX"] = torch.FloatTensor(size=(5,)).random_(0, 1000) + data_t["keyX"] = data["keyX"].clone() if mpu.get_tensor_model_parallel_rank() != 0: data = None data_utils._check_data_types(keys, data_t, torch.int64) - key_size, key_numel, \ - total_numel = data_utils._build_key_size_numel_dictionaries(keys, data) + key_size, key_numel, total_numel = data_utils._build_key_size_numel_dictionaries( + keys, data + ) for key in keys: assert key_size[key] == key_size_t[key] total_numel_t = 0 @@ -61,16 +68,16 @@ def test_broadcast_data(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") -if __name__ == '__main__': +if __name__ == "__main__": initialize_distributed() world_size = torch.distributed.get_world_size() tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test test broadcast data') + print_separator("test test broadcast data") test_broadcast_data(tensor_model_parallel_size) tensor_model_parallel_size *= 2 diff --git a/megatron/mpu/tests/test_initialize.py b/megatron/mpu/tests/test_initialize.py index e5d2be37e26..35b7608c48d 100644 --- a/megatron/mpu/tests/test_initialize.py +++ b/megatron/mpu/tests/test_initialize.py @@ -5,16 +5,21 @@ import mpu import torch import sys + sys.path.append("../..") def test_initialize_model_parallel(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing initialize_model_parallel with size {} ...'.format( - tensor_model_parallel_size)) - tensor_model_parallel_size_ = min(tensor_model_parallel_size, - torch.distributed.get_world_size()) + print( + "> testing initialize_model_parallel with size {} ...".format( + tensor_model_parallel_size + ) + ) + tensor_model_parallel_size_ = min( + tensor_model_parallel_size, torch.distributed.get_world_size() + ) assert not mpu.model_parallel_is_initialized() mpu.initialize_model_parallel(tensor_model_parallel_size_) assert mpu.model_parallel_is_initialized() @@ -43,16 +48,20 @@ def check(group, world_size, rank): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") def test_get_tensor_model_parallel_src_rank(tensor_model_parallel_size_): if torch.distributed.get_rank() == 0: - print('> testing get_tensor_model_parallel_src_rank with size {} ...'.format( - tensor_model_parallel_size_)) - tensor_model_parallel_size = min(tensor_model_parallel_size_, - torch.distributed.get_world_size()) + print( + "> testing get_tensor_model_parallel_src_rank with size {} ...".format( + tensor_model_parallel_size_ + ) + ) + tensor_model_parallel_size = min( + tensor_model_parallel_size_, torch.distributed.get_world_size() + ) assert not mpu.model_parallel_is_initialized() mpu.initialize_model_parallel(tensor_model_parallel_size) assert mpu.model_parallel_is_initialized() @@ -66,17 +75,17 @@ def test_get_tensor_model_parallel_src_rank(tensor_model_parallel_size_): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") -if __name__ == '__main__': +if __name__ == "__main__": initialize_distributed() world_size = torch.distributed.get_world_size() tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test initialize model parallel') + print_separator("test initialize model parallel") test_initialize_model_parallel(tensor_model_parallel_size) - print_separator('test model parallel source rank') + print_separator("test model parallel source rank") test_get_tensor_model_parallel_src_rank(tensor_model_parallel_size) tensor_model_parallel_size *= 2 diff --git a/megatron/mpu/tests/test_layers.py b/megatron/mpu/tests/test_layers.py index f524f944412..26f456dabf4 100644 --- a/megatron/mpu/tests/test_layers.py +++ b/megatron/mpu/tests/test_layers.py @@ -10,14 +10,20 @@ import torch import random import sys + sys.path.append("../..") device_name = get_accelerator().device_name() + + def test_parallel_embedding(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing parallel embedding with model parallel size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing parallel embedding with model parallel size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -29,8 +35,11 @@ def test_parallel_embedding(tensor_model_parallel_size): seed = 1236 set_random_seed(123) - input_data = torch.LongTensor( - size=(batch_size, seq_length)).random_(0, vocab_size).to(device_name) + input_data = ( + torch.LongTensor(size=(batch_size, seq_length)) + .random_(0, vocab_size) + .to(device_name) + ) loss_weight = torch.randn([batch_size, seq_length, hidden_size]).to(device_name) set_random_seed(seed) @@ -42,61 +51,76 @@ def test_parallel_embedding(tensor_model_parallel_size): set_random_seed(seed) embedding_parallel = layers.ParallelEmbedding( - vocab_size, hidden_size, init_method=init.normal_).to(device_name) + vocab_size, hidden_size, init_method=init.normal_ + ).to(device_name) output = embedding_parallel(input_data) loss_parallel = torch.mul(output, loss_weight).sum() loss_parallel.backward() set_random_seed(seed) embedding_vocab_parallel = layers.VocabParallelEmbedding( - vocab_size, hidden_size, init_method=init.normal_).to(device_name) + vocab_size, hidden_size, init_method=init.normal_ + ).to(device_name) output = embedding_vocab_parallel(input_data) loss_vocab_parallel = torch.mul(output, loss_weight).sum() loss_vocab_parallel.backward() torch.distributed.barrier() error = loss_parallel.sub(loss_original).abs() - print(' error in loss (parallel) on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 1.0e-12, 'error: {}'.format(error) + print( + " error in loss (parallel) on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 1.0e-12, "error: {}".format(error) torch.distributed.barrier() error = loss_vocab_parallel.sub(loss_original).abs() - print(' error in loss (vocab parallel) on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 1.0e-12, 'error: {}'.format(error) - - weight_grad_orig = torch.split(embedding_original.weight.grad, - hidden_size // tensor_model_parallel_size, - 1)[mpu.get_tensor_model_parallel_rank()] + print( + " error in loss (vocab parallel) on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 1.0e-12, "error: {}".format(error) + + weight_grad_orig = torch.split( + embedding_original.weight.grad, hidden_size // tensor_model_parallel_size, 1 + )[mpu.get_tensor_model_parallel_rank()] error = embedding_parallel.weight.grad.sub(weight_grad_orig).abs().max() - print(' error in grad (parallel) on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 1.0e-12, 'error: {}'.format(error) - - weight_grad_orig = torch.split(embedding_original.weight.grad, - vocab_size // tensor_model_parallel_size, - 0)[mpu.get_tensor_model_parallel_rank()] - error = embedding_vocab_parallel.weight.grad.sub( - weight_grad_orig).abs().max() - print(' error in grad (vocab parallel) on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 1.0e-12, 'error: {}'.format(error) + print( + " error in grad (parallel) on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 1.0e-12, "error: {}".format(error) + + weight_grad_orig = torch.split( + embedding_original.weight.grad, vocab_size // tensor_model_parallel_size, 0 + )[mpu.get_tensor_model_parallel_rank()] + error = embedding_vocab_parallel.weight.grad.sub(weight_grad_orig).abs().max() + print( + " error in grad (vocab parallel) on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 1.0e-12, "error: {}".format(error) # Reset groups mpu.destroy_model_parallel() torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") def test_initialize_affine_weight(tensor_model_parallel_size): mpu.initialize_model_parallel(tensor_model_parallel_size) if torch.distributed.get_rank() == 0: - print('> testing initialize_affine_weight with model parallel ' - 'size: {}'.format(tensor_model_parallel_size)) + print( + "> testing initialize_affine_weight with model parallel " + "size: {}".format(tensor_model_parallel_size) + ) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() seed = 12345 @@ -110,23 +134,25 @@ def test_initialize_affine_weight(tensor_model_parallel_size): # --------------- weight = torch.empty(output_size_coeff, input_size) set_random_seed(seed) - layers._initialize_affine_weight(weight, output_size, input_size, - - output_size_coeff, 0, - torch.nn.init.normal_) + layers._initialize_affine_weight( + weight, output_size, input_size, output_size_coeff, 0, torch.nn.init.normal_ + ) # Target. set_random_seed(seed) master_weight = torch.empty(output_size, input_size) torch.nn.init.normal_(master_weight) rank = mpu.get_tensor_model_parallel_rank() - my_weight = torch.split(master_weight, output_size_coeff, - dim=0)[rank].contiguous().clone() + my_weight = ( + torch.split(master_weight, output_size_coeff, dim=0)[rank].contiguous().clone() + ) # Compare. error = weight.sub(my_weight).abs().max() torch.distributed.barrier() - print(' column parallel max error (should be zero) on global rank ' - '{}: {}'.format(torch.distributed.get_rank(), error)) + print( + " column parallel max error (should be zero) on global rank " + "{}: {}".format(torch.distributed.get_rank(), error) + ) assert error < 1.0e-6 # ------------ @@ -134,22 +160,25 @@ def test_initialize_affine_weight(tensor_model_parallel_size): # ------------ weight = torch.empty(output_size, input_size_coeff) set_random_seed(seed) - mpu.layers._initialize_affine_weight(weight, output_size, input_size, - input_size_coeff, 1, - torch.nn.init.normal_) + mpu.layers._initialize_affine_weight( + weight, output_size, input_size, input_size_coeff, 1, torch.nn.init.normal_ + ) # Target. set_random_seed(seed) master_weight = torch.empty(output_size, input_size) torch.nn.init.normal_(master_weight) rank = mpu.get_tensor_model_parallel_rank() - my_weight = torch.split(master_weight, input_size_coeff, - dim=1)[rank].contiguous().clone() + my_weight = ( + torch.split(master_weight, input_size_coeff, dim=1)[rank].contiguous().clone() + ) # Compare. error = weight.sub(my_weight).abs().max() torch.distributed.barrier() - print(' row parallel max error (should be zero) on global rank ' - '{}: {}'.format(torch.distributed.get_rank(), error)) + print( + " row parallel max error (should be zero) on global rank " + "{}: {}".format(torch.distributed.get_rank(), error) + ) assert error < 1.0e-6 # Reset groups @@ -157,10 +186,11 @@ def test_initialize_affine_weight(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print(' >> passed the test :-)') + print(" >> passed the test :-)") class IdentityLayer2D(torch.nn.Module): + def __init__(self, m, n): super(IdentityLayer2D, self).__init__() self.weight = Parameter(torch.Tensor(m, n)) @@ -174,8 +204,10 @@ def test_column_parallel_linear(tensor_model_parallel_size): mpu.initialize_model_parallel(tensor_model_parallel_size) if torch.distributed.get_rank() == 0: - print('> testing ColumnParallelLinear with model parallel ' - 'size: {}'.format(tensor_model_parallel_size)) + print( + "> testing ColumnParallelLinear with model parallel " + "size: {}".format(tensor_model_parallel_size) + ) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() seed = 12345 @@ -189,7 +221,8 @@ def test_column_parallel_linear(tensor_model_parallel_size): # Network identity_layer = IdentityLayer2D(batch_size, input_size).to(device_name) linear_layer = mpu.ColumnParallelLinear( - input_size, output_size, keep_master_weight_for_test=True).to(device_name) + input_size, output_size, keep_master_weight_for_test=True + ).to(device_name) loss_weight = torch.randn([batch_size, output_size]).to(device_name) # Forward input_ = identity_layer() @@ -207,26 +240,33 @@ def test_column_parallel_linear(tensor_model_parallel_size): dLdX = torch.matmul(dLdY, A) rank = mpu.get_tensor_model_parallel_rank() - my_dLdA = torch.split(dLdA, output_size_coeff, - dim=0)[rank].contiguous().clone() + my_dLdA = torch.split(dLdA, output_size_coeff, dim=0)[rank].contiguous().clone() error = my_dLdA.sub(linear_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' error in dLdA on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdA on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 - my_dLdb = torch.split(dLdb, output_size_coeff, - dim=0)[rank].contiguous().clone() + my_dLdb = torch.split(dLdb, output_size_coeff, dim=0)[rank].contiguous().clone() error = my_dLdb.sub(linear_layer.bias.grad).abs().max() torch.distributed.barrier() - print(' error in dLdb on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdb on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 error = dLdX.sub(identity_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' error in dLdX on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdX on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 # Reset groups @@ -234,15 +274,17 @@ def test_column_parallel_linear(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print(' >> passed the test :-)') + print(" >> passed the test :-)") def test_row_parallel_linear(tensor_model_parallel_size): mpu.initialize_model_parallel(tensor_model_parallel_size) if torch.distributed.get_rank() == 0: - print('> testing RowParallelLinear with model parallel ' - 'size: {}'.format(tensor_model_parallel_size)) + print( + "> testing RowParallelLinear with model parallel " + "size: {}".format(tensor_model_parallel_size) + ) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() seed = 12345 @@ -256,7 +298,8 @@ def test_row_parallel_linear(tensor_model_parallel_size): # Network identity_layer = IdentityLayer2D(batch_size, input_size).to(device_name) linear_layer = mpu.RowParallelLinear( - input_size, output_size, keep_master_weight_for_test=True).to(device_name) + input_size, output_size, keep_master_weight_for_test=True + ).to(device_name) loss_weight = torch.randn([batch_size, output_size]).to(device_name) # Forward input_ = identity_layer() @@ -274,24 +317,32 @@ def test_row_parallel_linear(tensor_model_parallel_size): dLdX = torch.matmul(dLdY, A) rank = mpu.get_tensor_model_parallel_rank() - my_dLdA = torch.split(dLdA, input_size_coeff, - dim=1)[rank].contiguous().clone() + my_dLdA = torch.split(dLdA, input_size_coeff, dim=1)[rank].contiguous().clone() error = my_dLdA.sub(linear_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' error in dLdA on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdA on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 error = dLdb.sub(linear_layer.bias.grad).abs().max() torch.distributed.barrier() - print(' error in dLdb on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdb on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 error = dLdX.sub(identity_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' error in dLdX on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " error in dLdX on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 1.0e-6 # Reset groups @@ -299,10 +350,11 @@ def test_row_parallel_linear(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print(' >> passed the test :-)') + print(" >> passed the test :-)") class IdentityLayer3D(torch.nn.Module): + def __init__(self, m, n, k): super(IdentityLayer3D, self).__init__() self.weight = Parameter(torch.Tensor(m, n, k)) @@ -312,25 +364,33 @@ def forward(self): return self.weight -def parallel_self_attention(tensor_model_parallel_size, num_att_heads_per_partition, - hidden_size_per_att_head, dropout_prob, batch_size, - sequence_length): +def parallel_self_attention( + tensor_model_parallel_size, + num_att_heads_per_partition, + hidden_size_per_att_head, + dropout_prob, + batch_size, + sequence_length, +): mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() seed = 12345 set_random_seed(seed) - num_att_heads = num_att_heads_per_partition * \ - torch.distributed.get_world_size() + num_att_heads = num_att_heads_per_partition * torch.distributed.get_world_size() hidden_size = hidden_size_per_att_head * num_att_heads # Network - identity_layer = IdentityLayer3D(batch_size, sequence_length, - hidden_size).to(device_name) - attention_layer = mpu.BertParallelSelfAttention(hidden_size, num_att_heads, - dropout_prob).to(device_name) - loss_weight = torch.randn([batch_size, sequence_length, hidden_size]).to(device_name) + identity_layer = IdentityLayer3D(batch_size, sequence_length, hidden_size).to( + device_name + ) + attention_layer = mpu.BertParallelSelfAttention( + hidden_size, num_att_heads, dropout_prob + ).to(device_name) + loss_weight = torch.randn([batch_size, sequence_length, hidden_size]).to( + device_name + ) attention_mask = torch.randn([batch_size, 1, 1, sequence_length]).to(device_name) # Forward input_ = identity_layer() @@ -341,15 +401,23 @@ def parallel_self_attention(tensor_model_parallel_size, num_att_heads_per_partit rank = mpu.get_tensor_model_parallel_rank() mpu.destroy_model_parallel() - return rank, hidden_size, tensor_model_parallel_size, loss, \ - attention_layer, identity_layer + return ( + rank, + hidden_size, + tensor_model_parallel_size, + loss, + attention_layer, + identity_layer, + ) def test_parallel_self_attention(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing ParallelSelfAttention with model parallel ' - 'size: {}'.format(tensor_model_parallel_size)) + print( + "> testing ParallelSelfAttention with model parallel " + "size: {}".format(tensor_model_parallel_size) + ) num_att_heads_per_partition = 3 hidden_size_per_att_head = 7 @@ -357,48 +425,84 @@ def test_parallel_self_attention(tensor_model_parallel_size): batch_size = 5 sequence_length = 13 - rank_1, hideen_size_1, tensor_model_parallel_size_1, loss_1, \ - attention_layer_1, identity_layer_1 = parallel_self_attention( - 1, num_att_heads_per_partition, - hidden_size_per_att_head, dropout_prob, batch_size, sequence_length) - - rank, hidden_size, tensor_model_parallel_size, loss, \ - attention_layer, identity_layer = parallel_self_attention( - tensor_model_parallel_size, num_att_heads_per_partition, - hidden_size_per_att_head, dropout_prob, batch_size, sequence_length) + ( + rank_1, + hideen_size_1, + tensor_model_parallel_size_1, + loss_1, + attention_layer_1, + identity_layer_1, + ) = parallel_self_attention( + 1, + num_att_heads_per_partition, + hidden_size_per_att_head, + dropout_prob, + batch_size, + sequence_length, + ) + + ( + rank, + hidden_size, + tensor_model_parallel_size, + loss, + attention_layer, + identity_layer, + ) = parallel_self_attention( + tensor_model_parallel_size, + num_att_heads_per_partition, + hidden_size_per_att_head, + dropout_prob, + batch_size, + sequence_length, + ) assert hideen_size_1 == hidden_size error = loss_1.sub(loss).abs().max() torch.distributed.barrier() - print(' loss error on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " loss error on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 5.0e-6 my_lin_grad_list = torch.split( attention_layer_1.query_key_value.weight.grad, - hidden_size // tensor_model_parallel_size, 0)[rank::tensor_model_parallel_size] + hidden_size // tensor_model_parallel_size, + 0, + )[rank::tensor_model_parallel_size] my_lin_grad = torch.cat(my_lin_grad_list, dim=0) - error = my_lin_grad.sub( - attention_layer.query_key_value.weight.grad).abs().max() + error = my_lin_grad.sub(attention_layer.query_key_value.weight.grad).abs().max() torch.distributed.barrier() - print(' weight gradient error on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " weight gradient error on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 5.0e-6 - error = identity_layer_1.weight.grad.sub( - identity_layer.weight.grad).abs().max() + error = identity_layer_1.weight.grad.sub(identity_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' input gradient error on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) + print( + " input gradient error on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error < 5.0e-6 torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print(' >> passed the test :-)') + print(" >> passed the test :-)") -def parallel_transformer(tensor_model_parallel_size, num_att_heads_per_partition, - hidden_size_per_att_head, batch_size, sequence_length): +def parallel_transformer( + tensor_model_parallel_size, + num_att_heads_per_partition, + hidden_size_per_att_head, + batch_size, + sequence_length, +): mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -406,19 +510,27 @@ def parallel_transformer(tensor_model_parallel_size, num_att_heads_per_partition seed = 12345 set_random_seed(seed) - num_att_heads = num_att_heads_per_partition * \ - torch.distributed.get_world_size() + num_att_heads = num_att_heads_per_partition * torch.distributed.get_world_size() hidden_size = hidden_size_per_att_head * num_att_heads intermediate_size = 4 * hidden_size # Network - identity_layer = IdentityLayer3D(batch_size, sequence_length, - hidden_size).to(device_name) + identity_layer = IdentityLayer3D(batch_size, sequence_length, hidden_size).to( + device_name + ) transformer_layer = mpu.BertParallelTransformerLayer( - hidden_size, intermediate_size, num_att_heads, 0.0, 0.0, - torch.nn.functional.relu, 1.0e-5).to(device_name) - - loss_weight = torch.randn([batch_size, sequence_length, hidden_size]).to(device_name) + hidden_size, + intermediate_size, + num_att_heads, + 0.0, + 0.0, + torch.nn.functional.relu, + 1.0e-5, + ).to(device_name) + + loss_weight = torch.randn([batch_size, sequence_length, hidden_size]).to( + device_name + ) attention_mask = torch.randn([batch_size, 1, 1, sequence_length]).to(device_name) # Forward input_ = identity_layer() @@ -429,50 +541,83 @@ def parallel_transformer(tensor_model_parallel_size, num_att_heads_per_partition rank = mpu.get_tensor_model_parallel_rank() mpu.destroy_model_parallel() - return rank, hidden_size, tensor_model_parallel_size, loss, \ - transformer_layer, identity_layer + return ( + rank, + hidden_size, + tensor_model_parallel_size, + loss, + transformer_layer, + identity_layer, + ) def test_parallel_transformer_layer(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing ParallelTransformerLayer with model parallel ' - 'size: {}'.format(tensor_model_parallel_size)) + print( + "> testing ParallelTransformerLayer with model parallel " + "size: {}".format(tensor_model_parallel_size) + ) num_att_heads_per_partition = 3 hidden_size_per_att_head = 7 batch_size = 5 sequence_length = 13 - rank_1, hidden_size_1, tensor_model_parallel_size_1, loss_1, \ - transformer_layer_1, identity_layer_1 = parallel_transformer( - 1, num_att_heads_per_partition, - hidden_size_per_att_head, batch_size, sequence_length) - - rank, hidden_size, tensor_model_parallel_size, loss, \ - transformer_layer, identity_layer = parallel_transformer( - tensor_model_parallel_size, num_att_heads_per_partition, - hidden_size_per_att_head, batch_size, sequence_length) + ( + rank_1, + hidden_size_1, + tensor_model_parallel_size_1, + loss_1, + transformer_layer_1, + identity_layer_1, + ) = parallel_transformer( + 1, + num_att_heads_per_partition, + hidden_size_per_att_head, + batch_size, + sequence_length, + ) + + ( + rank, + hidden_size, + tensor_model_parallel_size, + loss, + transformer_layer, + identity_layer, + ) = parallel_transformer( + tensor_model_parallel_size, + num_att_heads_per_partition, + hidden_size_per_att_head, + batch_size, + sequence_length, + ) error = loss_1.sub(loss).abs().max() torch.distributed.barrier() - print(' loss error on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 5.0e-5, 'error: {}'.format(error) - - error = identity_layer_1.weight.grad.sub( - identity_layer.weight.grad).abs().max() + print( + " loss error on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 5.0e-5, "error: {}".format(error) + + error = identity_layer_1.weight.grad.sub(identity_layer.weight.grad).abs().max() torch.distributed.barrier() - print(' input gradient error on global rank {}: {}'.format( - torch.distributed.get_rank(), error)) - assert error < 5.0e-5, 'error: {}'.format(error) + print( + " input gradient error on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) + assert error < 5.0e-5, "error: {}".format(error) torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print(' >> passed the test :-)') + print(" >> passed the test :-)") -if __name__ == '__main__': +if __name__ == "__main__": torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = False @@ -480,7 +625,7 @@ def test_parallel_transformer_layer(tensor_model_parallel_size): initialize_distributed() world_size = torch.distributed.get_world_size() - print_separator('test initialize affine weight') + print_separator("test initialize affine weight") tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: test_initialize_affine_weight(tensor_model_parallel_size) @@ -488,29 +633,29 @@ def test_parallel_transformer_layer(tensor_model_parallel_size): tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test parallel embedding') + print_separator("test parallel embedding") test_parallel_embedding(tensor_model_parallel_size) tensor_model_parallel_size *= 2 - print_separator('test column-parallel linear') + print_separator("test column-parallel linear") tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: test_column_parallel_linear(tensor_model_parallel_size) tensor_model_parallel_size *= 2 - print_separator('test row-parallel linear') + print_separator("test row-parallel linear") tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: test_row_parallel_linear(tensor_model_parallel_size) tensor_model_parallel_size *= 2 - print_separator('test parallel self-attention') + print_separator("test parallel self-attention") tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: test_parallel_self_attention(tensor_model_parallel_size) tensor_model_parallel_size *= 2 - print_separator('test parallel transformer') + print_separator("test parallel transformer") tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: test_parallel_transformer_layer(tensor_model_parallel_size) diff --git a/megatron/mpu/tests/test_random.py b/megatron/mpu/tests/test_random.py index 21dcd109078..90b23d72867 100644 --- a/megatron/mpu/tests/test_random.py +++ b/megatron/mpu/tests/test_random.py @@ -5,14 +5,18 @@ import mpu import torch import sys + sys.path.append("../..") def test_set_cuda_rng_state(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing set_rng_state with size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing set_rng_state with size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -37,8 +41,11 @@ def test_set_cuda_rng_state(tensor_model_parallel_size): # State should be different. new_rng_state = get_accelerator().get_rng_state() max_diff = new_rng_state.sub(rng_state).max() - print(' max diff in rng state (should be non-zero) on global rank {}: {}'. - format(torch.distributed.get_rank(), max_diff)) + print( + " max diff in rng state (should be non-zero) on global rank {}: {}".format( + torch.distributed.get_rank(), max_diff + ) + ) assert max_diff > 0 # Reset the rng state and do the same stuff. @@ -52,14 +59,19 @@ def test_set_cuda_rng_state(tensor_model_parallel_size): # Results should be the same error = result_2.sub(result_1).abs().max() - print(' max error in generated tensors (should be zero) on ' - 'global rank {}: {}'.format(torch.distributed.get_rank(), error)) + print( + " max error in generated tensors (should be zero) on " + "global rank {}: {}".format(torch.distributed.get_rank(), error) + ) assert error < 1.0e-6 # Input state should have remained intact. error = rng_state.sub(rng_state_copy).max() - print(' max error in rng state (should be zero) on global rank {}: {}'. - format(torch.distributed.get_rank(), error)) + print( + " max error in rng state (should be zero) on global rank {}: {}".format( + torch.distributed.get_rank(), error + ) + ) assert error == 0 # Reset groups @@ -67,14 +79,17 @@ def test_set_cuda_rng_state(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") def test_cuda_rng_tracker(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing cuda rng tracker with size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing cuda rng tracker with size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -101,33 +116,38 @@ def test_cuda_rng_tracker(tensor_model_parallel_size): # Now if we interleave seed_1 and seed_2, # we should still get the same tensors get_accelerator().manual_seed(seed_1) - mpu.get_cuda_rng_tracker().add('test', seed_2) + mpu.get_cuda_rng_tracker().add("test", seed_2) torch.randn(size, out=tensor) result_11 = tensor.clone() - with mpu.get_cuda_rng_tracker().fork('test'): + with mpu.get_cuda_rng_tracker().fork("test"): torch.randn(size, out=tensor) result_21 = tensor.clone() torch.randn(size, out=tensor) result_12 = tensor.clone() - with mpu.get_cuda_rng_tracker().fork('test'): + with mpu.get_cuda_rng_tracker().fork("test"): torch.randn(size, out=tensor) result_22 = tensor.clone() diff = result_11.sub(result_21).abs().max() diff = min(diff, result_12.sub(result_22).abs().max()) - print(' max diff in generated tensors (should be non-zero) on ' - 'global rank {}: {}'.format(torch.distributed.get_rank(), diff)) + print( + " max diff in generated tensors (should be non-zero) on " + "global rank {}: {}".format(torch.distributed.get_rank(), diff) + ) assert diff > 1.0e-6 - error = max(result_11.sub(target_11).abs().max(), - result_12.sub(target_12).abs().max()) + error = max( + result_11.sub(target_11).abs().max(), result_12.sub(target_12).abs().max() + ) error = max(error, result_21.sub(target_21).abs().max()) error = max(error, result_22.sub(target_22).abs().max()) - print(' max error in generated tensors (should be zero) on ' - 'global rank {}: {}'.format(torch.distributed.get_rank(), error)) + print( + " max error in generated tensors (should be zero) on " + "global rank {}: {}".format(torch.distributed.get_rank(), error) + ) assert error < 1.0e-6 # Reset the tracker @@ -138,14 +158,17 @@ def test_cuda_rng_tracker(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") def test_model_parallel_cuda_manual_seed(tensor_model_parallel_size): if torch.distributed.get_rank() == 0: - print('> testing model parallel cuda manual seed with size {} ...'. - format(tensor_model_parallel_size)) + print( + "> testing model parallel cuda manual seed with size {} ...".format( + tensor_model_parallel_size + ) + ) mpu.initialize_model_parallel(tensor_model_parallel_size) tensor_model_parallel_size = mpu.get_tensor_model_parallel_world_size() @@ -153,8 +176,9 @@ def test_model_parallel_cuda_manual_seed(tensor_model_parallel_size): mpu.model_parallel_cuda_manual_seed(12345) assert get_accelerator().initial_seed() == 12345 with mpu.get_cuda_rng_tracker().fork(): - assert get_accelerator().initial_seed() == (12345 + 2718 + - mpu.get_tensor_model_parallel_rank()) + assert get_accelerator().initial_seed() == ( + 12345 + 2718 + mpu.get_tensor_model_parallel_rank() + ) # Reset the tracker mpu.get_cuda_rng_tracker().reset() @@ -164,28 +188,28 @@ def test_model_parallel_cuda_manual_seed(tensor_model_parallel_size): torch.distributed.barrier() if torch.distributed.get_rank() == 0: - print('>> passed the test :-)') + print(">> passed the test :-)") -if __name__ == '__main__': +if __name__ == "__main__": initialize_distributed() world_size = torch.distributed.get_world_size() tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test set rng state') + print_separator("test set rng state") test_set_cuda_rng_state(tensor_model_parallel_size) tensor_model_parallel_size *= 2 tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test cuda rng tracker') + print_separator("test cuda rng tracker") test_cuda_rng_tracker(tensor_model_parallel_size) tensor_model_parallel_size *= 2 tensor_model_parallel_size = 1 while tensor_model_parallel_size <= world_size: - print_separator('test model parallel cuda manual seed') + print_separator("test model parallel cuda manual seed") test_model_parallel_cuda_manual_seed(tensor_model_parallel_size) tensor_model_parallel_size *= 2 From b22a2e160538e2316151293752a80d27994ddb58 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:14:26 -0500 Subject: [PATCH 436/590] chore: Format `megatron/tokenizer/*` --- megatron/tokenizer/bert_tokenization.py | 70 +++++---- megatron/tokenizer/gpt2_tokenization.py | 170 +++++++++++++------- megatron/tokenizer/tokenizer.py | 198 ++++++++++++++---------- 3 files changed, 264 insertions(+), 174 deletions(-) diff --git a/megatron/tokenizer/bert_tokenization.py b/megatron/tokenizer/bert_tokenization.py index 642041e778e..19fbc7716b6 100644 --- a/megatron/tokenizer/bert_tokenization.py +++ b/megatron/tokenizer/bert_tokenization.py @@ -43,13 +43,16 @@ def validate_case_matches_checkpoint(do_lower_case, init_checkpoint): model_name = m.group(1) lower_models = [ - "uncased_L-24_H-1024_A-16", "uncased_L-12_H-768_A-12", - "multilingual_L-12_H-768_A-12", "chinese_L-12_H-768_A-12" + "uncased_L-24_H-1024_A-16", + "uncased_L-12_H-768_A-12", + "multilingual_L-12_H-768_A-12", + "chinese_L-12_H-768_A-12", ] cased_models = [ - "cased_L-12_H-768_A-12", "cased_L-24_H-1024_A-16", - "multi_cased_L-12_H-768_A-12" + "cased_L-12_H-768_A-12", + "cased_L-24_H-1024_A-16", + "multi_cased_L-12_H-768_A-12", ] is_bad_config = False @@ -71,8 +74,9 @@ def validate_case_matches_checkpoint(do_lower_case, init_checkpoint): "However, `%s` seems to be a %s model, so you " "should pass in `--do_lower_case=%s` so that the fine-tuning matches " "how the model was pre-training. If this error is wrong, please " - "just comment out this check." % (actual_flag, init_checkpoint, - model_name, case_name, opposite_flag)) + "just comment out this check." + % (actual_flag, init_checkpoint, model_name, case_name, opposite_flag) + ) def convert_to_unicode(text): @@ -122,7 +126,7 @@ def load_vocab(vocab_file): """Loads a vocabulary file into a dictionary.""" vocab = collections.OrderedDict() index = 0 - with open(vocab_file, "r", encoding = "utf-8") as reader: + with open(vocab_file, "r", encoding="utf-8") as reader: while True: token = convert_to_unicode(reader.readline()) if not token: @@ -183,27 +187,27 @@ def convert_ids_to_tokens(self, ids): @staticmethod def convert_tokens_to_string(tokens, clean_up_tokenization_spaces=True): - """ Converts a sequence of tokens (string) in a single string. """ + """Converts a sequence of tokens (string) in a single string.""" def clean_up_tokenization(out_string): - """ Clean up a list of simple English tokenization artifacts + """Clean up a list of simple English tokenization artifacts like spaces before punctuations and abreviated forms. """ out_string = ( out_string.replace(" .", ".") - .replace(" ?", "?") - .replace(" !", "!") - .replace(" ,", ",") - .replace(" ' ", "'") - .replace(" n't", "n't") - .replace(" 'm", "'m") - .replace(" 's", "'s") - .replace(" 've", "'ve") - .replace(" 're", "'re") + .replace(" ?", "?") + .replace(" !", "!") + .replace(" ,", ",") + .replace(" ' ", "'") + .replace(" n't", "n't") + .replace(" 'm", "'m") + .replace(" 's", "'s") + .replace(" 've", "'ve") + .replace(" 're", "'re") ) return out_string - text = ' '.join(tokens).replace(' ##', '').strip() + text = " ".join(tokens).replace(" ##", "").strip() if clean_up_tokenization_spaces: clean_text = clean_up_tokenization(text) return clean_text @@ -303,14 +307,16 @@ def _is_chinese_char(self, cp): # as is Japanese Hiragana and Katakana. Those alphabets are used to write # space-separated words, so they are not treated specially and handled # like the all of the other languages. - if ((cp >= 0x4E00 and cp <= 0x9FFF) or # - (cp >= 0x3400 and cp <= 0x4DBF) or # - (cp >= 0x20000 and cp <= 0x2A6DF) or # - (cp >= 0x2A700 and cp <= 0x2B73F) or # - (cp >= 0x2B740 and cp <= 0x2B81F) or # - (cp >= 0x2B820 and cp <= 0x2CEAF) or - (cp >= 0xF900 and cp <= 0xFAFF) or # - (cp >= 0x2F800 and cp <= 0x2FA1F)): # + if ( + (cp >= 0x4E00 and cp <= 0x9FFF) # + or (cp >= 0x3400 and cp <= 0x4DBF) # + or (cp >= 0x20000 and cp <= 0x2A6DF) # + or (cp >= 0x2A700 and cp <= 0x2B73F) # + or (cp >= 0x2B740 and cp <= 0x2B81F) # + or (cp >= 0x2B820 and cp <= 0x2CEAF) + or (cp >= 0xF900 and cp <= 0xFAFF) # + or (cp >= 0x2F800 and cp <= 0x2FA1F) + ): # return True return False @@ -320,7 +326,7 @@ def _clean_text(self, text): output = [] for char in text: cp = ord(char) - if cp == 0 or cp == 0xfffd or _is_control(char): + if cp == 0 or cp == 0xFFFD or _is_control(char): continue if _is_whitespace(char): output.append(" ") @@ -422,8 +428,12 @@ def _is_punctuation(char): # Characters such as "^", "$", and "`" are not in the Unicode # Punctuation class but we treat them as punctuation anyways, for # consistency. - if ((cp >= 33 and cp <= 47) or (cp >= 58 and cp <= 64) or - (cp >= 91 and cp <= 96) or (cp >= 123 and cp <= 126)): + if ( + (cp >= 33 and cp <= 47) + or (cp >= 58 and cp <= 64) + or (cp >= 91 and cp <= 96) + or (cp >= 123 and cp <= 126) + ): return True cat = unicodedata.category(char) if cat.startswith("P"): diff --git a/megatron/tokenizer/gpt2_tokenization.py b/megatron/tokenizer/gpt2_tokenization.py index 3f37e449089..380b6704575 100644 --- a/megatron/tokenizer/gpt2_tokenization.py +++ b/megatron/tokenizer/gpt2_tokenization.py @@ -15,8 +15,7 @@ """Tokenization classes for OpenAI GPT.""" -from __future__ import (absolute_import, division, print_function, - unicode_literals) +from __future__ import (absolute_import, division, print_function, unicode_literals) import sys import json @@ -38,17 +37,17 @@ def lru_cache(): logger = logging.getLogger(__name__) PRETRAINED_VOCAB_ARCHIVE_MAP = { - 'gpt2': "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json", + "gpt2": "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-vocab.json", } PRETRAINED_MERGES_ARCHIVE_MAP = { - 'gpt2': "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt", + "gpt2": "https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-merges.txt", } PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP = { - 'gpt2': 1024, + "gpt2": 1024, } -VOCAB_NAME = 'vocab.json' -MERGES_NAME = 'merges.txt' -SPECIAL_TOKENS_NAME = 'special_tokens.txt' +VOCAB_NAME = "vocab.json" +MERGES_NAME = "merges.txt" +SPECIAL_TOKENS_NAME = "special_tokens.txt" @lru_cache() @@ -63,8 +62,11 @@ def bytes_to_unicode(): And avoids mapping to whitespace/control characters the bpe code barfs on. """ _chr = unichr if sys.version_info[0] == 2 else chr - bs = list(range(ord("!"), ord("~") + 1)) + list(range(ord("¡"), ord("¬") + 1)) + \ - list(range(ord("®"), ord("ÿ") + 1)) + bs = ( + list(range(ord("!"), ord("~") + 1)) + + list(range(ord("¡"), ord("¬") + 1)) + + list(range(ord("®"), ord("ÿ") + 1)) + ) cs = bs[:] n = 0 for b in range(2**8): @@ -94,8 +96,11 @@ class GPT2Tokenizer(object): GPT-2 BPE tokenizer. Peculiarities: - Byte-level BPE """ + @classmethod - def from_pretrained(cls, pretrained_model_name_or_path, cache_dir=None, *inputs, **kwargs): + def from_pretrained( + cls, pretrained_model_name_or_path, cache_dir=None, *inputs, **kwargs + ): """ Instantiate a PreTrainedBertModel from a pre-trained model file. Download and cache the pre-trained model file if needed. @@ -107,14 +112,19 @@ def from_pretrained(cls, pretrained_model_name_or_path, cache_dir=None, *inputs, else: vocab_file = os.path.join(pretrained_model_name_or_path, VOCAB_NAME) merges_file = os.path.join(pretrained_model_name_or_path, MERGES_NAME) - special_tokens_file = os.path.join(pretrained_model_name_or_path, SPECIAL_TOKENS_NAME) + special_tokens_file = os.path.join( + pretrained_model_name_or_path, SPECIAL_TOKENS_NAME + ) if not os.path.exists(special_tokens_file): special_tokens_file = None else: - logger.info("loading special tokens file {}".format(special_tokens_file)) + logger.info( + "loading special tokens file {}".format(special_tokens_file) + ) # redirect to the cache, if necessary try: from .file_utils import cached_path + resolved_vocab_file = cached_path(vocab_file, cache_dir=cache_dir) resolved_merges_file = cached_path(merges_file, cache_dir=cache_dir) except EnvironmentError: @@ -123,45 +133,68 @@ def from_pretrained(cls, pretrained_model_name_or_path, cache_dir=None, *inputs, "We assumed '{}' was a path or url but couldn't find files {} and {} " "at this path or url.".format( pretrained_model_name_or_path, - ', '.join(PRETRAINED_VOCAB_ARCHIVE_MAP.keys()), + ", ".join(PRETRAINED_VOCAB_ARCHIVE_MAP.keys()), pretrained_model_name_or_path, - vocab_file, merges_file)) + vocab_file, + merges_file, + ) + ) return None if resolved_vocab_file == vocab_file and resolved_merges_file == merges_file: logger.info("loading vocabulary file {}".format(vocab_file)) logger.info("loading merges file {}".format(merges_file)) else: - logger.info("loading vocabulary file {} from cache at {}".format( - vocab_file, resolved_vocab_file)) - logger.info("loading merges file {} from cache at {}".format( - merges_file, resolved_merges_file)) - if pretrained_model_name_or_path in PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP: + logger.info( + "loading vocabulary file {} from cache at {}".format( + vocab_file, resolved_vocab_file + ) + ) + logger.info( + "loading merges file {} from cache at {}".format( + merges_file, resolved_merges_file + ) + ) + if ( + pretrained_model_name_or_path + in PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP + ): # if we're using a pretrained model, ensure the tokenizer wont index sequences longer # than the number of positional embeddings - max_len = PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP[pretrained_model_name_or_path] - kwargs['max_len'] = min(kwargs.get('max_len', int(1e12)), max_len) + max_len = PRETRAINED_VOCAB_POSITIONAL_EMBEDDINGS_SIZE_MAP[ + pretrained_model_name_or_path + ] + kwargs["max_len"] = min(kwargs.get("max_len", int(1e12)), max_len) # Instantiate tokenizer. - if special_tokens_file and 'special_tokens' not in kwargs: - special_tokens = open(special_tokens_file, encoding='utf-8').read().split('\n')[:-1] + if special_tokens_file and "special_tokens" not in kwargs: + special_tokens = ( + open(special_tokens_file, encoding="utf-8").read().split("\n")[:-1] + ) else: - special_tokens = kwargs.pop('special_tokens', []) + special_tokens = kwargs.pop("special_tokens", []) tokenizer = cls( resolved_vocab_file, resolved_merges_file, special_tokens=special_tokens, *inputs, - **kwargs) + **kwargs + ) return tokenizer - def __init__(self, vocab_file, merges_file, errors='replace', - special_tokens=None, max_len=None): + def __init__( + self, + vocab_file, + merges_file, + errors="replace", + special_tokens=None, + max_len=None, + ): self.max_len = max_len if max_len is not None else int(1e12) self.encoder = json.load(open(vocab_file)) self.decoder = {v: k for k, v in self.encoder.items()} self.errors = errors # how to handle errors in decoding self.byte_encoder = bytes_to_unicode() self.byte_decoder = {v: k for k, v in self.byte_encoder.items()} - bpe_data = open(merges_file, encoding='utf-8').read().split('\n')[1:-1] + bpe_data = open(merges_file, encoding="utf-8").read().split("\n")[1:-1] bpe_merges = [tuple(merge.split()) for merge in bpe_data] self.bpe_ranks = dict(zip(bpe_merges, range(len(bpe_merges)))) self.cache = {} @@ -169,7 +202,8 @@ def __init__(self, vocab_file, merges_file, errors='replace', # Should haved added re.IGNORECASE so BPE merges can happen for # capitalized versions of contractions self.pat = re.compile( - r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""") + r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""" + ) self.special_tokens = {} self.special_tokens_decoder = {} @@ -179,16 +213,17 @@ def __len__(self): return len(self.encoder) + len(self.special_tokens) def set_special_tokens(self, special_tokens): - """ Add a list of additional tokens to the encoder. - The additional tokens are indexed starting from the last index of the - current vocabulary in the order of the `special_tokens` list. + """Add a list of additional tokens to the encoder. + The additional tokens are indexed starting from the last index of the + current vocabulary in the order of the `special_tokens` list. """ if not special_tokens: self.special_tokens = {} self.special_tokens_decoder = {} return - self.special_tokens = dict((tok, len(self.encoder) + i) - for i, tok in enumerate(special_tokens)) + self.special_tokens = dict( + (tok, len(self.encoder) + i) for i, tok in enumerate(special_tokens) + ) self.special_tokens_decoder = {v: k for k, v in self.special_tokens.items()} logger.info("Special tokens {}".format(self.special_tokens)) @@ -202,7 +237,7 @@ def bpe(self, token): return token while True: - bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float('inf'))) + bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf"))) if bigram not in self.bpe_ranks: break first, second = bigram @@ -229,25 +264,27 @@ def bpe(self, token): break else: pairs = get_pairs(word) - word = ' '.join(word) + word = " ".join(word) self.cache[token] = word return word def tokenize(self, text): - """ Tokenize a string. """ + """Tokenize a string.""" bpe_tokens = [] for token in re.findall(self.pat, text): if sys.version_info[0] == 2: - token = ''.join(self.byte_encoder[ord(b)] for b in token) + token = "".join(self.byte_encoder[ord(b)] for b in token) else: - token = ''.join(self.byte_encoder[b] for b in token.encode('utf-8')) - bpe_tokens.extend(bpe_token for bpe_token in self.bpe(token).split(' ')) + token = "".join(self.byte_encoder[b] for b in token.encode("utf-8")) + bpe_tokens.extend(bpe_token for bpe_token in self.bpe(token).split(" ")) return bpe_tokens def convert_tokens_to_ids(self, tokens): - """ Converts a sequence of tokens into ids using the vocab. """ + """Converts a sequence of tokens into ids using the vocab.""" ids = [] - if isinstance(tokens, str) or (sys.version_info[0] == 2 and isinstance(tokens, unicode)): + if isinstance(tokens, str) or ( + sys.version_info[0] == 2 and isinstance(tokens, unicode) + ): if tokens in self.special_tokens: return self.special_tokens[tokens] else: @@ -262,7 +299,8 @@ def convert_tokens_to_ids(self, tokens): "Token indices sequence length is longer than the specified maximum " " sequence length for this OpenAI GPT model ({} > {}). Running this" " sequence through the model will result in indexing errors".format( - len(ids), self.max_len) + len(ids), self.max_len + ) ) return ids @@ -281,41 +319,57 @@ def encode(self, text): return self.convert_tokens_to_ids(self.tokenize(text)) def decode(self, tokens): - text = ''.join([self.decoder[token] for token in tokens]) - text = bytearray([self.byte_decoder[c] for c in text]).decode('utf-8', errors=self.errors) + text = "".join([self.decoder[token] for token in tokens]) + text = bytearray([self.byte_decoder[c] for c in text]).decode( + "utf-8", errors=self.errors + ) return text def save_vocabulary(self, vocab_path): """Save the tokenizer vocabulary and merge files to a directory.""" if not os.path.isdir(vocab_path): - logger.error("Vocabulary path ({}) should be a directory".format(vocab_path)) + logger.error( + "Vocabulary path ({}) should be a directory".format(vocab_path) + ) return vocab_file = os.path.join(vocab_path, VOCAB_NAME) merge_file = os.path.join(vocab_path, MERGES_NAME) special_tokens_file = os.path.join(vocab_path, SPECIAL_TOKENS_NAME) - with open(vocab_file, 'w', encoding='utf-8') as f: + with open(vocab_file, "w", encoding="utf-8") as f: f.write(json.dumps(self.encoder, ensure_ascii=False)) index = 0 with open(merge_file, "w", encoding="utf-8") as writer: - writer.write(u'#version: 0.2\n') - for bpe_tokens, token_index in sorted(self.bpe_ranks.items(), key=lambda kv: kv[1]): + writer.write("#version: 0.2\n") + for bpe_tokens, token_index in sorted( + self.bpe_ranks.items(), key=lambda kv: kv[1] + ): if index != token_index: - logger.warning("Saving vocabulary to {}: BPE merge indices are not consecutive." - " Please check that the tokenizer is not corrupted!".format(merge_file)) + logger.warning( + "Saving vocabulary to {}: BPE merge indices are not consecutive." + " Please check that the tokenizer is not corrupted!".format( + merge_file + ) + ) index = token_index - writer.write(' '.join(bpe_tokens) + u'\n') + writer.write(" ".join(bpe_tokens) + "\n") index += 1 index = len(self.encoder) - with open(special_tokens_file, 'w', encoding='utf-8') as writer: - for token, token_index in sorted(self.special_tokens.items(), key=lambda kv: kv[1]): + with open(special_tokens_file, "w", encoding="utf-8") as writer: + for token, token_index in sorted( + self.special_tokens.items(), key=lambda kv: kv[1] + ): if index != token_index: - logger.warning("Saving special tokens vocabulary to {}: BPE indices are not consecutive." - " Please check that the tokenizer is not corrupted!".format(special_tokens_file)) + logger.warning( + "Saving special tokens vocabulary to {}: BPE indices are not consecutive." + " Please check that the tokenizer is not corrupted!".format( + special_tokens_file + ) + ) index = token_index - writer.write(token + u'\n') + writer.write(token + "\n") index += 1 return vocab_file, merge_file, special_tokens_file diff --git a/megatron/tokenizer/tokenizer.py b/megatron/tokenizer/tokenizer.py index 92853fb30dd..0776a087411 100644 --- a/megatron/tokenizer/tokenizer.py +++ b/megatron/tokenizer/tokenizer.py @@ -14,49 +14,54 @@ def build_tokenizer(args): """Initialize tokenizer.""" if args.rank == 0: - print('> building {} tokenizer ...'.format(args.tokenizer_type), - flush=True) + print("> building {} tokenizer ...".format(args.tokenizer_type), flush=True) # Select and instantiate the tokenizer. - if args.tokenizer_type == 'BertWordPieceLowerCase': + if args.tokenizer_type == "BertWordPieceLowerCase": assert args.vocab_file is not None - tokenizer = _BertWordPieceTokenizer(vocab_file=args.vocab_file, - lower_case=True, - vocab_extra_ids=args.vocab_extra_ids) - elif args.tokenizer_type == 'BertWordPieceCase': + tokenizer = _BertWordPieceTokenizer( + vocab_file=args.vocab_file, + lower_case=True, + vocab_extra_ids=args.vocab_extra_ids, + ) + elif args.tokenizer_type == "BertWordPieceCase": assert args.vocab_file is not None - tokenizer = _BertWordPieceTokenizer(vocab_file=args.vocab_file, - lower_case=False, - vocab_extra_ids=args.vocab_extra_ids) - elif args.tokenizer_type == 'GPT2BPETokenizer': + tokenizer = _BertWordPieceTokenizer( + vocab_file=args.vocab_file, + lower_case=False, + vocab_extra_ids=args.vocab_extra_ids, + ) + elif args.tokenizer_type == "GPT2BPETokenizer": assert args.vocab_file is not None assert args.merge_file is not None tokenizer = _GPT2BPETokenizer(args.vocab_file, args.merge_file) - elif args.tokenizer_type == 'SentencePieceTokenizer': + elif args.tokenizer_type == "SentencePieceTokenizer": assert args.tokenizer_model is not None - tokenizer = _SentencePieceTokenizer(args.tokenizer_model, vocab_extra_ids=args.vocab_extra_ids) - elif args.tokenizer_type == 'GPTSentencePieceTokenizer': + tokenizer = _SentencePieceTokenizer( + args.tokenizer_model, vocab_extra_ids=args.vocab_extra_ids + ) + elif args.tokenizer_type == "GPTSentencePieceTokenizer": assert args.tokenizer_model is not None tokenizer = _GPTSentencePieceTokenizer(args.tokenizer_model) - elif args.tokenizer_type == 'Llama2Tokenizer': + elif args.tokenizer_type == "Llama2Tokenizer": assert args.tokenizer_model is not None tokenizer = _Llama2Tokenizer(args.tokenizer_model) - elif args.tokenizer_type == 'NullTokenizer': + elif args.tokenizer_type == "NullTokenizer": assert args.vocab_size is not None tokenizer = _NullTokenizer(args.vocab_size) - elif args.tokenizer_type == 'HFTokenizer': + elif args.tokenizer_type == "HFTokenizer": assert args.tokenizer_model is not None # tokenizer = _HFTokenizer(args.tokenizer_model) - tokenizer = _HFTokenizer(args.tokenizer_model, - args.seq_length, - args.trust_remote_code) + tokenizer = _HFTokenizer( + args.tokenizer_model, args.seq_length, args.trust_remote_code + ) else: - raise NotImplementedError('{} tokenizer is not ' - 'implemented.'.format(args.tokenizer_type)) - + raise NotImplementedError( + "{} tokenizer is not " "implemented.".format(args.tokenizer_type) + ) + # Add vocab size. - args.padded_vocab_size = _vocab_size_with_padding(tokenizer.vocab_size, - args) + args.padded_vocab_size = _vocab_size_with_padding(tokenizer.vocab_size, args) return tokenizer @@ -66,14 +71,15 @@ def _vocab_size_with_padding(orig_vocab_size, args): still having GPU friendly size.""" after = orig_vocab_size - multiple = args.make_vocab_size_divisible_by * \ - args.tensor_model_parallel_size + multiple = args.make_vocab_size_divisible_by * args.tensor_model_parallel_size while (after % multiple) != 0: after += 1 if args.rank == 0: - print(' > padded vocab (size: {}) with {} dummy tokens ' - '(new size: {})'.format( - orig_vocab_size, after - orig_vocab_size, after), flush=True) + print( + " > padded vocab (size: {}) with {} dummy tokens " + "(new size: {})".format(orig_vocab_size, after - orig_vocab_size, after), + flush=True, + ) return after @@ -106,33 +112,39 @@ def tokenize(self, text): pass def detokenize(self, token_ids): - raise NotImplementedError('detokenizer is not implemented for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "detokenizer is not implemented for {} " "tokenizer".format(self.name) + ) @property def cls(self): - raise NotImplementedError('CLS is not provided for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "CLS is not provided for {} " "tokenizer".format(self.name) + ) @property def sep(self): - raise NotImplementedError('SEP is not provided for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "SEP is not provided for {} " "tokenizer".format(self.name) + ) @property def pad(self): - raise NotImplementedError('PAD is not provided for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "PAD is not provided for {} " "tokenizer".format(self.name) + ) @property def eod(self): - raise NotImplementedError('EOD is not provided for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "EOD is not provided for {} " "tokenizer".format(self.name) + ) @property def mask(self): - raise NotImplementedError('MASK is not provided for {} ' - 'tokenizer'.format(self.name)) + raise NotImplementedError( + "MASK is not provided for {} " "tokenizer".format(self.name) + ) class _BertWordPieceTokenizer(AbstractTokenizer): @@ -140,25 +152,24 @@ class _BertWordPieceTokenizer(AbstractTokenizer): def __init__(self, vocab_file, lower_case=True, vocab_extra_ids=0): if lower_case: - name = 'BERT Lower Case' + name = "BERT Lower Case" else: - name = 'BERT Upper Case' + name = "BERT Upper Case" super().__init__(name) self.tokenizer = FullBertTokenizer(vocab_file, do_lower_case=lower_case) - self.cls_id = self.tokenizer.vocab['[CLS]'] - self.sep_id = self.tokenizer.vocab['[SEP]'] - self.pad_id = self.tokenizer.vocab['[PAD]'] - self.mask_id = self.tokenizer.vocab['[MASK]'] + self.cls_id = self.tokenizer.vocab["[CLS]"] + self.sep_id = self.tokenizer.vocab["[SEP]"] + self.pad_id = self.tokenizer.vocab["[PAD]"] + self.mask_id = self.tokenizer.vocab["[MASK]"] self._additional_special_tokens = [] # (dsachan) Add BOS and EOS tokens - SPECIAL_TOKENS = {'eos_token': '[EOS]', - 'bos_token': '[BOS]'} - self._bos_token = '[BOS]' + SPECIAL_TOKENS = {"eos_token": "[EOS]", "bos_token": "[BOS]"} + self._bos_token = "[BOS]" self.add_token(self._bos_token) self._bos_token_id = self.vocab.get(self._bos_token) - self._eos_token = '[EOS]' + self._eos_token = "[EOS]" self.add_token(self._eos_token) self._eos_token_id = self.vocab.get(self._eos_token) @@ -166,7 +177,8 @@ def __init__(self, vocab_file, lower_case=True, vocab_extra_ids=0): # These can be used as sentinel tokens in T5 model inputs additional_special_tokens = [] additional_special_tokens.extend( - ["".format(i) for i in range(vocab_extra_ids)]) + ["".format(i) for i in range(vocab_extra_ids)] + ) self.add_additional_special_tokens(additional_special_tokens) def add_token(self, token): @@ -203,7 +215,7 @@ def decode(self, ids): def decode_token_ids(self, token_ids): tokens = self.tokenizer.convert_ids_to_tokens(token_ids) - exclude_list = ['[PAD]', '[CLS]'] + exclude_list = ["[PAD]", "[CLS]"] non_pads = [t for t in tokens if t not in exclude_list] result = "" @@ -233,32 +245,32 @@ def mask(self): @property def bos_token(self): - """ Beginning of sentence token id """ + """Beginning of sentence token id""" return self._bos_token @property def eos_token(self): - """ End of sentence token id """ + """End of sentence token id""" return self._eos_token @property def additional_special_tokens(self): - """ All the additional special tokens you may want to use (list of strings).""" + """All the additional special tokens you may want to use (list of strings).""" return self._additional_special_tokens @property def bos_token_id(self): - """ Id of the beginning of sentence token in the vocabulary.""" + """Id of the beginning of sentence token in the vocabulary.""" return self._bos_token_id @property def eos_token_id(self): - """ Id of the end of sentence token in the vocabulary.""" + """Id of the end of sentence token in the vocabulary.""" return self._eos_token_id @property def additional_special_tokens_ids(self): - """ Ids of all the additional special tokens in the vocabulary (list of integers).""" + """Ids of all the additional special tokens in the vocabulary (list of integers).""" return [self.vocab.get(token) for token in self._additional_special_tokens] @additional_special_tokens.setter @@ -270,12 +282,13 @@ class _GPT2BPETokenizer(AbstractTokenizer): """Original GPT2 BPE tokenizer.""" def __init__(self, vocab_file, merge_file): - name = 'GPT2 BPE' + name = "GPT2 BPE" super().__init__(name) - self.tokenizer = GPT2Tokenizer(vocab_file, merge_file, errors='replace', - special_tokens=[], max_len=None) - self.eod_id = self.tokenizer.encoder['<|endoftext|>'] + self.tokenizer = GPT2Tokenizer( + vocab_file, merge_file, errors="replace", special_tokens=[], max_len=None + ) + self.eod_id = self.tokenizer.encoder["<|endoftext|>"] @property def vocab_size(self): @@ -304,10 +317,11 @@ class _SentencePieceTokenizer(AbstractTokenizer): """SentencePieceTokenizer-Megatron wrapper""" def __init__(self, model_file, vocab_extra_ids=0): - name = 'SentencePieceTokenizer' + name = "SentencePieceTokenizer" super().__init__(name) import sentencepiece + self.tokenizer = sentencepiece.SentencePieceProcessor(model_file=model_file) self._initalize(vocab_extra_ids) @@ -335,20 +349,20 @@ def _add_special_token(t): self._special_tokens[t] = self._vocab[t] self._inv_special_tokens[self._vocab[t]] = t - _add_special_token('') - self._cls_id = self._vocab[''] - _add_special_token('') - self._sep_id = self._vocab[''] - _add_special_token('') - self._eod_id = self._vocab[''] - _add_special_token('') - self._mask_id = self._vocab[''] + _add_special_token("") + self._cls_id = self._vocab[""] + _add_special_token("") + self._sep_id = self._vocab[""] + _add_special_token("") + self._eod_id = self._vocab[""] + _add_special_token("") + self._mask_id = self._vocab[""] pad_id = self.tokenizer.pad_id() try: pad_token = self.tokenizer.id_to_piece(pad_id) except IndexError: - pad_token = '' + pad_token = "" _add_special_token(pad_token) self._pad_id = self._vocab[pad_token] @@ -356,7 +370,7 @@ def _add_special_token(t): try: bos_token = self.tokenizer.id_to_piece(bos_id) except IndexError: - bos_token = '' + bos_token = "" _add_special_token(bos_token) self._bos_id = self._vocab[bos_token] @@ -364,7 +378,7 @@ def _add_special_token(t): try: eos_token = self.tokenizer.id_to_piece(eos_id) except IndexError: - eos_token = '' + eos_token = "" _add_special_token(eos_token) self._eos_id = self._vocab[eos_token] @@ -474,10 +488,14 @@ def mask(self): def additional_special_tokens_ids(self): return [self.vocab[k] for k in self._t5_tokens] + class _GPTSentencePieceTokenizer(_SentencePieceTokenizer): """SentencePieceTokenizer-Megatron wrapper""" - def __init__(self, model_file,): + def __init__( + self, + model_file, + ): super().__init__(model_file, vocab_extra_ids=0) def _initalize(self, vocab_extra_ids): @@ -517,7 +535,10 @@ def additional_special_tokens_ids(self): class _Llama2Tokenizer(_SentencePieceTokenizer): """SentencePieceTokenizer-Megatron wrapper""" - def __init__(self, model_file,): + def __init__( + self, + model_file, + ): super().__init__(model_file, vocab_extra_ids=0) def _initalize(self, vocab_extra_ids): @@ -531,7 +552,7 @@ def _initalize(self, vocab_extra_ids): assert self.tokenizer.vocab_size() == self.tokenizer.get_piece_size() def tokenize(self, s: str, bos=True, eos=False): - '''Default args for text completion, not chat/dialog.''' + """Default args for text completion, not chat/dialog.""" assert type(s) is str t = self.tokenizer.encode(s) if bos: @@ -563,18 +584,20 @@ def eod(self): def additional_special_tokens_ids(self): return None + class _NullTokenizer: + def __init__(self, vocab_size): vocab_size = int(vocab_size) self._eos_id = vocab_size - self.vocab_size = vocab_size+1 + self.vocab_size = vocab_size + 1 def tokenize(self, text): - return [int(x) for x in text.split(' ')] + return [int(x) for x in text.split(" ")] def detokenize(self, ids): text = [str(x) for x in ids] - return ' '.join(text) + return " ".join(text) @property def cls(self): @@ -599,13 +622,16 @@ def additional_special_tokens_ids(self): class _HFTokenizer(AbstractTokenizer): """HF Tokenizer""" + def __init__(self, tokenizer_name_or_path, max_seq_len, trust_remote_code): name = tokenizer_name_or_path super().__init__(name) - self.tokenizer = AutoTokenizer.from_pretrained(tokenizer_name_or_path, - padding_side="right", - trust_remote_code=trust_remote_code, - use_fast=False) + self.tokenizer = AutoTokenizer.from_pretrained( + tokenizer_name_or_path, + padding_side="right", + trust_remote_code=trust_remote_code, + use_fast=False, + ) DEFAULT_PAD_TOKEN = "[PAD]" DEFAULT_EOS_TOKEN = "" DEFAULT_BOS_TOKEN = "" From ed02e5f0b7f64c23fdf63cff955a25041258dfba Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:14:53 -0500 Subject: [PATCH 437/590] chore: Format `megatron/model/*` --- megatron/model/bert_model.py | 163 ++++--- megatron/model/biencoder_model.py | 254 ++++++---- megatron/model/classification.py | 54 ++- megatron/model/distributed.py | 93 ++-- megatron/model/enums.py | 6 +- megatron/model/fused_bias_gelu.py | 13 +- megatron/model/fused_layer_norm.py | 159 +++--- megatron/model/fused_rmsnorm.py | 9 +- megatron/model/fused_softmax.py | 6 +- megatron/model/gpt_model.py | 387 +++++++++------ megatron/model/language_model.py | 425 ++++++++-------- megatron/model/module.py | 92 ++-- megatron/model/multiple_choice.py | 50 +- megatron/model/realm_model.py | 139 +++--- megatron/model/rmsnorm.py | 5 +- megatron/model/rotary_pos_embedding.py | 22 +- megatron/model/t5_model.py | 148 +++--- megatron/model/utils.py | 36 +- megatron/model/vision/classification.py | 17 +- megatron/model/vision/dino.py | 150 +++--- megatron/model/vision/esvit_swin_backbone.py | 485 +++++++++++++------ megatron/model/vision/inpainting.py | 63 ++- megatron/model/vision/knn_monitor.py | 65 +-- megatron/model/vision/mit_backbone.py | 401 +++++++++++---- megatron/model/vision/swin_backbone.py | 371 ++++++++++---- megatron/model/vision/utils.py | 31 +- megatron/model/vision/vit_backbone.py | 59 +-- 27 files changed, 2347 insertions(+), 1356 deletions(-) diff --git a/megatron/model/bert_model.py b/megatron/model/bert_model.py index 325c0f50e81..26d7fc1f56b 100644 --- a/megatron/model/bert_model.py +++ b/megatron/model/bert_model.py @@ -29,15 +29,15 @@ def bert_extended_attention_mask(attention_mask): extended_attention_mask = attention_mask_bss.unsqueeze(1) # Convert attention mask to binary: - extended_attention_mask = (extended_attention_mask < 0.5) + extended_attention_mask = extended_attention_mask < 0.5 return extended_attention_mask + def bert_position_ids(token_ids): # Create position ids seq_length = token_ids.size(1) - position_ids = torch.arange(seq_length, dtype=torch.long, - device=token_ids.device) + position_ids = torch.arange(seq_length, dtype=torch.long, device=token_ids.device) position_ids = position_ids.unsqueeze(0).expand_as(token_ids) return position_ids @@ -61,13 +61,20 @@ def __init__(self, mpu_vocab_size, hidden_size, config, parallel_output): tensor_parallel.set_tensor_model_parallel_attributes(self.bias, True, 0, 1) self.parallel_output = parallel_output - self.dense = get_linear_layer(hidden_size, hidden_size, config.init_method, gather_params_on_init=args.zero_stage == 3) - setattr(self.dense.weight, 'sequence_parallel', config.sequence_parallel) - setattr(self.dense.bias, 'sequence_parallel', config.sequence_parallel) + self.dense = get_linear_layer( + hidden_size, + hidden_size, + config.init_method, + gather_params_on_init=args.zero_stage == 3, + ) + setattr(self.dense.weight, "sequence_parallel", config.sequence_parallel) + setattr(self.dense.bias, "sequence_parallel", config.sequence_parallel) - self.layernorm = LayerNorm(hidden_size, - eps=config.layernorm_epsilon, - sequence_parallel=config.sequence_parallel) + self.layernorm = LayerNorm( + hidden_size, + eps=config.layernorm_epsilon, + sequence_parallel=config.sequence_parallel, + ) self.gelu = torch.nn.functional.gelu if args.openai_gelu: self.gelu = openai_gelu @@ -78,21 +85,23 @@ def forward(self, hidden_states, word_embeddings_weight): hidden_states = self.dense(hidden_states) hidden_states = self.gelu(hidden_states) hidden_states = self.layernorm(hidden_states) - output = parallel_lm_logits(hidden_states, - word_embeddings_weight, - self.parallel_output, - bias=self.bias) + output = parallel_lm_logits( + hidden_states, word_embeddings_weight, self.parallel_output, bias=self.bias + ) return output -def post_language_model_processing(lm_output, pooled_output, - lm_head, binary_head, - lm_labels, - logit_weights, - fp16_lm_cross_entropy): +def post_language_model_processing( + lm_output, + pooled_output, + lm_head, + binary_head, + lm_labels, + logit_weights, + fp16_lm_cross_entropy, +): # Output. - lm_logits = lm_head( - lm_output, logit_weights) + lm_logits = lm_head(lm_output, logit_weights) binary_logits = None if binary_head is not None: @@ -100,33 +109,36 @@ def post_language_model_processing(lm_output, pooled_output, if lm_labels is None: # [s b h] => [b s h] - return lm_logits.transpose(0,1).contiguous(), binary_logits + return lm_logits.transpose(0, 1).contiguous(), binary_logits else: # [b s] => [s b] - lm_labels = lm_labels.transpose(0,1).contiguous() + lm_labels = lm_labels.transpose(0, 1).contiguous() # lm_logits : [s, b, h] and lm_labels: [s, b] if fp16_lm_cross_entropy: assert lm_logits.dtype == torch.half lm_loss = tensor_parallel.vocab_parallel_cross_entropy(lm_logits, lm_labels) else: - lm_loss = tensor_parallel.vocab_parallel_cross_entropy(lm_logits.float(), - lm_labels) + lm_loss = tensor_parallel.vocab_parallel_cross_entropy( + lm_logits.float(), lm_labels + ) # [s, b] => [b s] - lm_loss = lm_loss.transpose(0,1).contiguous() + lm_loss = lm_loss.transpose(0, 1).contiguous() return lm_loss, binary_logits class BertModel(MegatronModule): """Bert Language model.""" - def __init__(self, - config, - num_tokentypes=2, - add_binary_head=True, - parallel_output=True, - pre_process=True, - post_process=True, - return_moe_loss=False): + def __init__( + self, + config, + num_tokentypes=2, + add_binary_head=True, + parallel_output=True, + pre_process=True, + post_process=True, + return_moe_loss=False, + ): super().__init__(config=config) args = get_args() @@ -156,22 +168,27 @@ def __init__(self, self.initialize_word_embeddings() if self.post_process: - self.lm_head = BertLMHead(self.shared_embedding_or_output_weight().size(0), config.hidden_size, - config, parallel_output) - self._lm_head_key = 'lm_head' + self.lm_head = BertLMHead( + self.shared_embedding_or_output_weight().size(0), + config.hidden_size, + config, + parallel_output, + ) + self._lm_head_key = "lm_head" self.binary_head = None if self.add_binary_head: - self.binary_head = get_linear_layer(config.hidden_size, 2, - config.init_method, - args.zero_stage == 3) - self._binary_head_key = 'binary_head' + self.binary_head = get_linear_layer( + config.hidden_size, 2, config.init_method, args.zero_stage == 3 + ) + self._binary_head_key = "binary_head" def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" self.language_model.set_input_tensor(input_tensor) - def forward(self, bert_model_input, attention_mask, - tokentype_ids=None, lm_labels=None): + def forward( + self, bert_model_input, attention_mask, tokentype_ids=None, lm_labels=None + ): extended_attention_mask = bert_extended_attention_mask(attention_mask) input_ids = bert_model_input @@ -181,7 +198,7 @@ def forward(self, bert_model_input, attention_mask, input_ids, position_ids, extended_attention_mask, - tokentype_ids=tokentype_ids + tokentype_ids=tokentype_ids, ) if self.post_process and self.add_binary_head: @@ -198,9 +215,10 @@ def forward(self, bert_model_input, attention_mask, output = torch.zeros( size=(embeddings.shape[0], embeddings.shape[2]), dtype=torch.float32, - device=torch.cuda.current_device()) + device=torch.cuda.current_device(), + ) for i, (embedding, mask) in enumerate(zip(embeddings, masks)): - output[i, :] = torch.mean(embedding[1: mask - 1], dim=0) + output[i, :] = torch.mean(embedding[1 : mask - 1], dim=0) return output @@ -210,49 +228,60 @@ def forward(self, bert_model_input, attention_mask, if self.post_process: if not self.add_binary_head: lm_output, moe_losses = lm_output - lm_output = post_language_model_processing(lm_output, pooled_output, - self.lm_head, self.binary_head, - lm_labels, - self.shared_embedding_or_output_weight(), - self.fp16_lm_cross_entropy) + lm_output = post_language_model_processing( + lm_output, + pooled_output, + self.lm_head, + self.binary_head, + lm_labels, + self.shared_embedding_or_output_weight(), + self.fp16_lm_cross_entropy, + ) return *lm_output, moe_losses if self.return_moe_loss else lm_output else: return lm_output - - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.post_process: - state_dict_[self._lm_head_key] \ - = self.lm_head.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._lm_head_key] = ( + self.lm_head.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.post_process and self.add_binary_head: - state_dict_[self._binary_head_key] \ - = self.binary_head.state_dict(prefix=prefix, keep_vars=keep_vars) + state_dict_[self._binary_head_key] = self.binary_head.state_dict( + prefix=prefix, keep_vars=keep_vars + ) # Save word_embeddings. if self.post_process and not self.pre_process: - state_dict_[self._word_embeddings_for_head_key] \ - = self.word_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + state_dict_[self._word_embeddings_for_head_key] = ( + self.word_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) + state_dict[self._language_model_key], strict=strict + ) if self.post_process: - self.lm_head.load_state_dict( - state_dict[self._lm_head_key], strict=strict) + self.lm_head.load_state_dict(state_dict[self._lm_head_key], strict=strict) if self.post_process and self.add_binary_head: self.binary_head.load_state_dict( - state_dict[self._binary_head_key], strict=strict) + state_dict[self._binary_head_key], strict=strict + ) # Load word_embeddings. if self.post_process and not self.pre_process: self.word_embeddings.load_state_dict( - state_dict[self._word_embeddings_for_head_key], strict=strict) + state_dict[self._word_embeddings_for_head_key], strict=strict + ) diff --git a/megatron/model/biencoder_model.py b/megatron/model/biencoder_model.py index 3d0207458e5..e27df0156c8 100644 --- a/megatron/model/biencoder_model.py +++ b/megatron/model/biencoder_model.py @@ -15,36 +15,45 @@ from megatron.model.utils import scaled_init_method_normal from .module import MegatronModule -def get_model_provider(only_query_model=False, only_context_model=False, - biencoder_shared_query_context_model=False): + +def get_model_provider( + only_query_model=False, + only_context_model=False, + biencoder_shared_query_context_model=False, +): def model_provider(pre_process=True, post_process=True): """Build the model.""" - print_rank_0('building Bienoder model ...') - model = biencoder_model_provider(only_query_model=only_query_model, - only_context_model = only_context_model, - biencoder_shared_query_context_model = \ - biencoder_shared_query_context_model, - pre_process=pre_process, post_process=post_process) + print_rank_0("building Bienoder model ...") + model = biencoder_model_provider( + only_query_model=only_query_model, + only_context_model=only_context_model, + biencoder_shared_query_context_model=biencoder_shared_query_context_model, + pre_process=pre_process, + post_process=post_process, + ) return model return model_provider -def biencoder_model_provider(only_query_model=False, - only_context_model=False, - biencoder_shared_query_context_model=False, - pre_process=True, - post_process=True): +def biencoder_model_provider( + only_query_model=False, + only_context_model=False, + biencoder_shared_query_context_model=False, + pre_process=True, + post_process=True, +): """Build the model.""" - assert mpu.get_tensor_model_parallel_world_size() == 1 and \ - mpu.get_pipeline_model_parallel_world_size() == 1, \ - "Model parallel size > 1 not supported for ICT" + assert ( + mpu.get_tensor_model_parallel_world_size() == 1 + and mpu.get_pipeline_model_parallel_world_size() == 1 + ), "Model parallel size > 1 not supported for ICT" - print_rank_0('building BiEncoderModel...') + print_rank_0("building BiEncoderModel...") # simpler to just keep using 2 tokentypes since # the LM we initialize with has 2 tokentypes @@ -53,10 +62,10 @@ def biencoder_model_provider(only_query_model=False, parallel_output=False, only_query_model=only_query_model, only_context_model=only_context_model, - biencoder_shared_query_context_model=\ - biencoder_shared_query_context_model, + biencoder_shared_query_context_model=biencoder_shared_query_context_model, pre_process=pre_process, - post_process=post_process) + post_process=post_process, + ) return model @@ -64,14 +73,16 @@ def biencoder_model_provider(only_query_model=False, class BiEncoderModel(MegatronModule): """Bert-based module for Biencoder model.""" - def __init__(self, - num_tokentypes=1, - parallel_output=True, - only_query_model=False, - only_context_model=False, - biencoder_shared_query_context_model=False, - pre_process=True, - post_process=True): + def __init__( + self, + num_tokentypes=1, + parallel_output=True, + only_query_model=False, + only_context_model=False, + biencoder_shared_query_context_model=False, + pre_process=True, + post_process=True, + ): super(BiEncoderModel, self).__init__() args = get_args() @@ -79,10 +90,10 @@ def __init__(self, num_tokentypes=num_tokentypes, parallel_output=parallel_output, pre_process=pre_process, - post_process=post_process) + post_process=post_process, + ) - self.biencoder_shared_query_context_model = \ - biencoder_shared_query_context_model + self.biencoder_shared_query_context_model = biencoder_shared_query_context_model assert not (only_context_model and only_query_model) self.use_context_model = not only_query_model self.use_query_model = not only_context_model @@ -90,18 +101,18 @@ def __init__(self, if self.biencoder_shared_query_context_model: self.model = PretrainedBertModel(**bert_kwargs) - self._model_key = 'shared_model' + self._model_key = "shared_model" self.query_model, self.context_model = self.model, self.model else: if self.use_query_model: # this model embeds (pseudo-)queries - Embed_input in the paper self.query_model = PretrainedBertModel(**bert_kwargs) - self._query_key = 'query_model' + self._query_key = "query_model" if self.use_context_model: # this model embeds evidence blocks - Embed_doc in the paper self.context_model = PretrainedBertModel(**bert_kwargs) - self._context_key = 'context_model' + self._context_key = "context_model" def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" @@ -110,23 +121,31 @@ def set_input_tensor(self, input_tensor): # self.language_model.set_input_tensor(input_tensor) return - def forward(self, query_tokens, query_attention_mask, query_types, - context_tokens, context_attention_mask, context_types): + def forward( + self, + query_tokens, + query_attention_mask, + query_types, + context_tokens, + context_attention_mask, + context_types, + ): """Run a forward pass for each of the models and return the respective embeddings.""" if self.use_query_model: - query_logits = self.embed_text(self.query_model, - query_tokens, - query_attention_mask, - query_types) + query_logits = self.embed_text( + self.query_model, query_tokens, query_attention_mask, query_types + ) else: raise ValueError("Cannot embed query without the query model.") if self.use_context_model: - context_logits = self.embed_text(self.context_model, - context_tokens, - context_attention_mask, - context_types) + context_logits = self.embed_text( + self.context_model, + context_tokens, + context_attention_mask, + context_types, + ) else: raise ValueError("Cannot embed block without the block model.") return query_logits, context_logits @@ -134,28 +153,30 @@ def forward(self, query_tokens, query_attention_mask, query_types, @staticmethod def embed_text(model, tokens, attention_mask, token_types): """Embed a batch of tokens using the model""" - logits = model(tokens, - attention_mask, - token_types) + logits = model(tokens, attention_mask, token_types) return logits - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """Save dict with state dicts of each of the models.""" state_dict_ = {} if self.biencoder_shared_query_context_model: - state_dict_[self._model_key] = \ - self.model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + state_dict_[self._model_key] = self.model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) else: if self.use_query_model: - state_dict_[self._query_key] = \ + state_dict_[self._query_key] = ( self.query_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + prefix=prefix, keep_vars=keep_vars + ) + ) if self.use_context_model: - state_dict_[self._context_key] = \ + state_dict_[self._context_key] = ( self.context_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + prefix=prefix, keep_vars=keep_vars + ) + ) return state_dict_ @@ -163,18 +184,19 @@ def load_state_dict(self, state_dict, strict=True): """Load the state dicts of each of the models""" if self.biencoder_shared_query_context_model: print_rank_0("Loading shared query-context model") - self.model.load_state_dict(state_dict[self._model_key], \ - strict=strict) + self.model.load_state_dict(state_dict[self._model_key], strict=strict) else: if self.use_query_model: print_rank_0("Loading query model") - self.query_model.load_state_dict( \ - state_dict[self._query_key], strict=strict) + self.query_model.load_state_dict( + state_dict[self._query_key], strict=strict + ) if self.use_context_model: print_rank_0("Loading context model") - self.context_model.load_state_dict( \ - state_dict[self._context_key], strict=strict) + self.context_model.load_state_dict( + state_dict[self._context_key], strict=strict + ) def init_state_dict_from_bert(self): """Initialize the state from a pretrained BERT model @@ -188,37 +210,42 @@ def init_state_dict_from_bert(self): tracker_filename = get_checkpoint_tracker_filename(args.bert_load) if not os.path.isfile(tracker_filename): raise FileNotFoundError("Could not find BERT checkpoint") - with open(tracker_filename, 'r') as f: + with open(tracker_filename, "r") as f: iteration = int(f.read().strip()) assert iteration > 0 checkpoint_name = get_checkpoint_name(args.bert_load, iteration, False) if mpu.get_data_parallel_rank() == 0: - print('global rank {} is loading BERT checkpoint {}'.format( - torch.distributed.get_rank(), checkpoint_name)) + print( + "global rank {} is loading BERT checkpoint {}".format( + torch.distributed.get_rank(), checkpoint_name + ) + ) # Load the checkpoint. try: - state_dict = torch.load(checkpoint_name, map_location='cpu') + state_dict = torch.load(checkpoint_name, map_location="cpu") except ModuleNotFoundError: from megatron.fp16_deprecated import loss_scaler # For backward compatibility. - print_rank_0(' > deserializing using the old code structure ...') - sys.modules['fp16.loss_scaler'] = sys.modules[ - 'megatron.fp16_deprecated.loss_scaler'] - sys.modules['megatron.fp16.loss_scaler'] = sys.modules[ - 'megatron.fp16_deprecated.loss_scaler'] - state_dict = torch.load(checkpoint_name, map_location='cpu') - sys.modules.pop('fp16.loss_scaler', None) - sys.modules.pop('megatron.fp16.loss_scaler', None) + print_rank_0(" > deserializing using the old code structure ...") + sys.modules["fp16.loss_scaler"] = sys.modules[ + "megatron.fp16_deprecated.loss_scaler" + ] + sys.modules["megatron.fp16.loss_scaler"] = sys.modules[ + "megatron.fp16_deprecated.loss_scaler" + ] + state_dict = torch.load(checkpoint_name, map_location="cpu") + sys.modules.pop("fp16.loss_scaler", None) + sys.modules.pop("megatron.fp16.loss_scaler", None) except BaseException: - print_rank_0('could not load the BERT checkpoint') + print_rank_0("could not load the BERT checkpoint") sys.exit() - checkpoint_version = state_dict.get('checkpoint_version', 0) + checkpoint_version = state_dict.get("checkpoint_version", 0) # load the LM state dict into each model - model_dict = state_dict['model']['language_model'] + model_dict = state_dict["model"]["language_model"] if self.biencoder_shared_query_context_model: self.model.language_model.load_state_dict(model_dict) @@ -228,17 +255,17 @@ def init_state_dict_from_bert(self): self.query_model.language_model.load_state_dict(model_dict) # give each model the same ict_head to begin with as well if self.biencoder_projection_dim > 0: - query_proj_state_dict = \ - self.state_dict_for_save_checkpoint()\ - [self._query_key]['projection_enc'] + query_proj_state_dict = self.state_dict_for_save_checkpoint()[ + self._query_key + ]["projection_enc"] fix_query_key_value_ordering(self.query_model, checkpoint_version) if self.use_context_model: self.context_model.language_model.load_state_dict(model_dict) - if self.query_model is not None and \ - self.biencoder_projection_dim > 0: - self.context_model.projection_enc.load_state_dict\ - (query_proj_state_dict) + if self.query_model is not None and self.biencoder_projection_dim > 0: + self.context_model.projection_enc.load_state_dict( + query_proj_state_dict + ) fix_query_key_value_ordering(self.context_model, checkpoint_version) @@ -246,8 +273,13 @@ class PretrainedBertModel(MegatronModule): """BERT-based encoder for queries or contexts used for learned information retrieval.""" - def __init__(self, num_tokentypes=2, - parallel_output=True, pre_process=True, post_process=True): + def __init__( + self, + num_tokentypes=2, + parallel_output=True, + pre_process=True, + post_process=True, + ): super(PretrainedBertModel, self).__init__() args = get_args() @@ -259,7 +291,8 @@ def __init__(self, num_tokentypes=2, self.post_process = post_process init_method = init_method_normal(args.init_method_std) scaled_init_method = scaled_init_method_normal( - args.init_method_std, args.num_layers) + args.init_method_std, args.num_layers + ) self.language_model, self._language_model_key = get_language_model( num_tokentypes=num_tokentypes, @@ -268,24 +301,29 @@ def __init__(self, num_tokentypes=2, init_method=init_method, scaled_init_method=scaled_init_method, pre_process=self.pre_process, - post_process=self.post_process) + post_process=self.post_process, + ) if args.biencoder_projection_dim > 0: - self.projection_enc = get_linear_layer(args.hidden_size, - args.biencoder_projection_dim, - init_method, - gather_params_on_init=args.zero_stage == 3) - self._projection_enc_key = 'projection_enc' + self.projection_enc = get_linear_layer( + args.hidden_size, + args.biencoder_projection_dim, + init_method, + gather_params_on_init=args.zero_stage == 3, + ) + self._projection_enc_key = "projection_enc" def forward(self, input_ids, attention_mask, tokentype_ids=None): extended_attention_mask = attention_mask.unsqueeze(1) - #extended_attention_mask = bert_extended_attention_mask(attention_mask) + # extended_attention_mask = bert_extended_attention_mask(attention_mask) position_ids = bert_position_ids(input_ids) - lm_output = self.language_model(input_ids, - position_ids, - extended_attention_mask, - tokentype_ids=tokentype_ids) + lm_output = self.language_model( + input_ids, + position_ids, + extended_attention_mask, + tokentype_ids=tokentype_ids, + ) # This mask will be used in average-pooling and max-pooling pool_mask = (input_ids == self.pad_id).unsqueeze(2) @@ -301,19 +339,21 @@ def forward(self, input_ids, attention_mask, tokentype_ids=None): return pooled_output - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.biencoder_projection_dim > 0: - state_dict_[self._projection_enc_key] = \ - self.projection_enc.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._projection_enc_key] = self.projection_enc.state_dict( + prefix=prefix, keep_vars=keep_vars + ) return state_dict_ @@ -321,9 +361,11 @@ def load_state_dict(self, state_dict, strict=True): """Customized load.""" print_rank_0("loading pretrained weights") self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) + state_dict[self._language_model_key], strict=strict + ) if self.biencoder_projection_dim > 0: print_rank_0("loading projection head weights") self.projection_enc.load_state_dict( - state_dict[self._projection_enc_key], strict=strict) + state_dict[self._projection_enc_key], strict=strict + ) diff --git a/megatron/model/classification.py b/megatron/model/classification.py index f2469f9b855..32d91b00616 100644 --- a/megatron/model/classification.py +++ b/megatron/model/classification.py @@ -16,12 +16,9 @@ class Classification(MegatronModule): - def __init__(self, - config, - num_classes, - num_tokentypes=2, - pre_process=True, - post_process=True): + def __init__( + self, config, num_classes, num_tokentypes=2, pre_process=True, post_process=True + ): super().__init__(config=config, share_embeddings_and_output_weights=False) args = get_args() @@ -35,16 +32,19 @@ def __init__(self, add_pooler=True, encoder_attn_mask_type=AttnMaskType.padding, pre_process=self.pre_process, - post_process=self.post_process) + post_process=self.post_process, + ) # Multi-choice head. if self.post_process: self.classification_dropout = torch.nn.Dropout(args.hidden_dropout) - self.classification_head = get_linear_layer(args.hidden_size, - self.num_classes, - init_method, - gather_params_on_init=args.zero_stage == 3) - self._classification_head_key = 'classification_head' + self.classification_head = get_linear_layer( + args.hidden_size, + self.num_classes, + init_method, + gather_params_on_init=args.zero_stage == 3, + ) + self._classification_head_key = "classification_head" def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" @@ -60,7 +60,7 @@ def forward(self, model_input, attention_mask, tokentype_ids=None): input_ids, position_ids, extended_attention_mask, - tokentype_ids=tokentype_ids + tokentype_ids=tokentype_ids, ) if self.post_process: @@ -74,29 +74,35 @@ def forward(self, model_input, attention_mask, tokentype_ids=None): return classification_logits return lm_output - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.post_process: - state_dict_[self._classification_head_key] \ - = self.classification_head.state_dict(prefix=prefix, keep_vars=keep_vars) + state_dict_[self._classification_head_key] = ( + self.classification_head.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) + state_dict[self._language_model_key], strict=strict + ) if self.post_process: if self._classification_head_key in state_dict: self.classification_head.load_state_dict( - state_dict[self._classification_head_key], strict=strict) + state_dict[self._classification_head_key], strict=strict + ) else: - print_rank_last('***WARNING*** could not find {} in the checkpoint, ' - 'initializing to random'.format( - self._classification_head_key)) + print_rank_last( + "***WARNING*** could not find {} in the checkpoint, " + "initializing to random".format(self._classification_head_key) + ) diff --git a/megatron/model/distributed.py b/megatron/model/distributed.py index f6b1203c34a..eacb73034b0 100644 --- a/megatron/model/distributed.py +++ b/megatron/model/distributed.py @@ -12,34 +12,34 @@ from .module import MegatronModule from deepspeed.accelerator import get_accelerator + class MemoryBuffer: def __init__(self, numel, numel_padded, dtype): self.numel = numel self.numel_padded = numel_padded self.dtype = dtype - self.data = torch.zeros(self.numel_padded, - dtype=self.dtype, - device=get_accelerator().current_device_name(), - requires_grad=False) + self.data = torch.zeros( + self.numel_padded, + dtype=self.dtype, + device=get_accelerator().current_device_name(), + requires_grad=False, + ) def zero(self): """Reset the buffer to zero.""" self.data.zero_() - def get(self, shape, start_index): """Return a tensor with the input `shape` as a view into the 1-D data starting at `start_index`.""" end_index = start_index + shape.numel() - assert end_index <= self.numel, \ - 'requested tensor is out of the buffer range.' + assert end_index <= self.numel, "requested tensor is out of the buffer range." buffer_tensor = self.data[start_index:end_index] buffer_tensor = buffer_tensor.view(shape) return buffer_tensor - class DistributedDataParallelBase(MegatronModule, ABC): """Abstract class for DDP.""" @@ -48,30 +48,25 @@ def __init__(self, module): # Keep a pointer to the model. self.module = module - @abstractmethod def allreduce_gradients(self): pass - def forward(self, *inputs, **kwargs): return self.module(*inputs, **kwargs) - - def state_dict(self, prefix='', keep_vars=False): + def state_dict(self, prefix="", keep_vars=False): return self.module.state_dict(prefix=prefix, keep_vars=keep_vars) - - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): - return self.module.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) - + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): + return self.module.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) def load_state_dict(self, state_dict, strict=True): self.module.load_state_dict(state_dict, strict=strict) - class DistributedDataParallel(DistributedDataParallelBase): """DDP with contiguous buffers options to storre and accumulate gradients. This class: @@ -88,14 +83,13 @@ class DistributedDataParallel(DistributedDataParallelBase): gradients. """ - def __init__(self, module, - accumulate_allreduce_grads_in_fp32, - use_contiguous_buffers): + def __init__( + self, module, accumulate_allreduce_grads_in_fp32, use_contiguous_buffers + ): super(DistributedDataParallel, self).__init__(module) - self.accumulate_allreduce_grads_in_fp32 \ - = accumulate_allreduce_grads_in_fp32 + self.accumulate_allreduce_grads_in_fp32 = accumulate_allreduce_grads_in_fp32 self.use_contiguous_buffers = use_contiguous_buffers # If we are using fp32-accumulate-allreduce explicitly # this means we need main grads in a continous buffer. @@ -115,16 +109,20 @@ def __init__(self, module, # Simple function to define buffer type. def _get_buffer_type(param): - return torch.float if \ - self.accumulate_allreduce_grads_in_fp32 else param.dtype + return ( + torch.float + if self.accumulate_allreduce_grads_in_fp32 + else param.dtype + ) # First calculate total number of elements per type. type_num_elements = {} for param in self.module.parameters(): if param.requires_grad: dtype = _get_buffer_type(param) - type_num_elements[dtype] = type_num_elements.get(dtype, 0) \ - + param.data.nelement() + type_num_elements[dtype] = ( + type_num_elements.get(dtype, 0) + param.data.nelement() + ) # Allocate the buffer. for dtype, num_elements in type_num_elements.items(): @@ -133,13 +131,14 @@ def _get_buffer_type(param): # multiple of data_parallel_world_size. (This padding is done # due to a constraint with the reduce_scatter op, which requires # all tensors have equal size. See: optimizer.py.) - num_elements_padded = data_parallel_world_size * \ - int(math.ceil(num_elements / data_parallel_world_size)) + num_elements_padded = data_parallel_world_size * int( + math.ceil(num_elements / data_parallel_world_size) + ) # Allocate grad buffer. - self._grad_buffers[dtype] = MemoryBuffer(num_elements, - num_elements_padded, - dtype) + self._grad_buffers[dtype] = MemoryBuffer( + num_elements, num_elements_padded, dtype + ) # Assume the back prop order is reverse the params order, # store the start index for the gradients. @@ -148,7 +147,8 @@ def _get_buffer_type(param): dtype = _get_buffer_type(param) type_num_elements[dtype] -= param.data.nelement() param.main_grad = self._grad_buffers[dtype].get( - param.data.shape, type_num_elements[dtype]) + param.data.shape, type_num_elements[dtype] + ) if dtype not in self._grad_buffer_param_index_map: self._grad_buffer_param_index_map[dtype] = {} self._grad_buffer_param_index_map[dtype][param] = ( @@ -170,9 +170,9 @@ def _get_buffer_type(param): grad_acc.register_hook(self._make_param_hook(param)) self.grad_accs.append(grad_acc) - def _make_param_hook(self, param): """Create the all-reduce hook for backprop.""" + # Hook used for back-prop. def param_hook(*unused): # Add the gradient to the buffer. @@ -181,23 +181,23 @@ def param_hook(*unused): param.main_grad.add_(param.grad.data) # Now we can deallocate grad memory. param.grad = None - return param_hook + return param_hook def zero_grad_buffer(self): """Set the grad buffer data to zero. Needs to be called at the begining of each iteration.""" - assert self._grad_buffers is not None, 'buffers are not initialized.' + assert self._grad_buffers is not None, "buffers are not initialized." for _, buffer_ in self._grad_buffers.items(): buffer_.zero() - def broadcast_params(self): for param in self.module.parameters(): - torch.distributed.broadcast(param.data, - src=mpu.get_data_parallel_src_rank(), - group=mpu.get_data_parallel_group()) - + torch.distributed.broadcast( + param.data, + src=mpu.get_data_parallel_src_rank(), + group=mpu.get_data_parallel_group(), + ) def allreduce_gradients(self): """Reduce gradients across data parallel ranks.""" @@ -206,7 +206,8 @@ def allreduce_gradients(self): for _, buffer_ in self._grad_buffers.items(): buffer_.data /= mpu.get_data_parallel_world_size() torch.distributed.all_reduce( - buffer_.data, group=mpu.get_data_parallel_group()) + buffer_.data, group=mpu.get_data_parallel_group() + ) else: # Otherwise, bucketize and all-reduce buckets = {} @@ -225,7 +226,9 @@ def allreduce_gradients(self): coalesced = _flatten_dense_tensors(grads) coalesced /= mpu.get_data_parallel_world_size() torch.distributed.all_reduce( - coalesced, group=mpu.get_data_parallel_group()) - for buf, synced in zip(grads, _unflatten_dense_tensors( - coalesced, grads)): + coalesced, group=mpu.get_data_parallel_group() + ) + for buf, synced in zip( + grads, _unflatten_dense_tensors(coalesced, grads) + ): buf.copy_(synced) diff --git a/megatron/model/enums.py b/megatron/model/enums.py index bc4e4aa29a0..37568a05059 100644 --- a/megatron/model/enums.py +++ b/megatron/model/enums.py @@ -2,20 +2,24 @@ import enum + class LayerType(enum.Enum): encoder = 1 decoder = 2 retro_encoder = 3 retro_decoder = 4 retro_decoder_with_retriever = 5 - + + class AttnType(enum.Enum): self_attn = 1 cross_attn = 2 + class AttnMaskType(enum.Enum): padding = 1 causal = 2 + # For backward compatibility with old model checkpoints from megatron.core.enums import ModelType diff --git a/megatron/model/fused_bias_gelu.py b/megatron/model/fused_bias_gelu.py index 29222db024e..ac632705b46 100644 --- a/megatron/model/fused_bias_gelu.py +++ b/megatron/model/fused_bias_gelu.py @@ -11,10 +11,12 @@ # actual gelu is: # x * 0.5 * (1.0 + torch.erf(x * 0.70710678)) + @torch.jit.script def bias_gelu(bias, y): x = bias + y - return x * 0.5 * (1.0 + torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x))) + return x * 0.5 * (1.0 + torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x))) + # gradient of tanh approximation of gelu # gradient of actual gelu is: @@ -24,10 +26,14 @@ def bias_gelu_back(g, bias, y): x = bias + y tanh_out = torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x)) # sqrt(2/pi) * 3 * 0.044715 -> 0.1070322243 - ff = 0.5 * x * ((1 - tanh_out * tanh_out) * (0.79788456 + 0.1070322243 * x * x)) + 0.5 * (1 + tanh_out) - return ff*g + ff = 0.5 * x * ( + (1 - tanh_out * tanh_out) * (0.79788456 + 0.1070322243 * x * x) + ) + 0.5 * (1 + tanh_out) + return ff * g + class GeLUFunction(torch.autograd.Function): + @staticmethod # bias is an optional argument def forward(ctx, input, bias): @@ -40,4 +46,5 @@ def backward(ctx, grad_output): tmp = bias_gelu_back(grad_output, bias, input) return tmp, tmp + bias_gelu_impl = GeLUFunction.apply diff --git a/megatron/model/fused_layer_norm.py b/megatron/model/fused_layer_norm.py index d1ef034397c..d6b5cdbc8f9 100644 --- a/megatron/model/fused_layer_norm.py +++ b/megatron/model/fused_layer_norm.py @@ -19,6 +19,7 @@ try: from apex.contrib.layer_norm.layer_norm import FastLayerNormFN + HAVE_PERSIST_LAYER_NORM = True except: HAVE_PERSIST_LAYER_NORM = False @@ -32,11 +33,15 @@ class MixedFusedLayerNorm(torch.nn.Module): - def __init__(self, normalized_shape, eps=1e-5, - no_persist_layer_norm=True, - sequence_parallel=False, - apply_layernorm_1p=False, - mem_efficient_ln=True): + def __init__( + self, + normalized_shape, + eps=1e-5, + no_persist_layer_norm=True, + sequence_parallel=False, + apply_layernorm_1p=False, + mem_efficient_ln=True, + ): super(MixedFusedLayerNorm, self).__init__() self.apply_layernorm_1p = apply_layernorm_1p @@ -48,11 +53,36 @@ def __init__(self, normalized_shape, eps=1e-5, # List of hiddens sizes supported in the persistent layer norm kernel # If the hidden size is not supported, fall back to the non-persistent # kernel. - persist_ln_hidden_sizes = [1024, 1536, 2048, 2304, 3072, 3840, 4096, - 5120, 6144, 8192, 10240, 12288, 12800, 15360, 16384, 18432, 20480, - 24576, 25600, 30720, 32768, 40960, 49152, 65536] - if normalized_shape not in persist_ln_hidden_sizes or \ - not HAVE_PERSIST_LAYER_NORM: + persist_ln_hidden_sizes = [ + 1024, + 1536, + 2048, + 2304, + 3072, + 3840, + 4096, + 5120, + 6144, + 8192, + 10240, + 12288, + 12800, + 15360, + 16384, + 18432, + 20480, + 24576, + 25600, + 30720, + 32768, + 40960, + 49152, + 65536, + ] + if ( + normalized_shape not in persist_ln_hidden_sizes + or not HAVE_PERSIST_LAYER_NORM + ): no_persist_layer_norm = True if isinstance(normalized_shape, numbers.Integral): @@ -60,57 +90,76 @@ def __init__(self, normalized_shape, eps=1e-5, self.normalized_shape = torch.Size(normalized_shape) self.eps = eps init_device = None - if get_accelerator().device_name() == 'hpu': - init_device = get_accelerator().current_device_name() - self.weight = Parameter(torch.empty(*normalized_shape, - device=init_device, - dtype=get_args().params_dtype)) - self.bias = Parameter(torch.empty(*normalized_shape, - device=init_device, - dtype=get_args().params_dtype)) + if get_accelerator().device_name() == "hpu": + init_device = get_accelerator().current_device_name() + self.weight = Parameter( + torch.empty( + *normalized_shape, device=init_device, dtype=get_args().params_dtype + ) + ) + self.bias = Parameter( + torch.empty( + *normalized_shape, device=init_device, dtype=get_args().params_dtype + ) + ) self.reset_parameters() self.no_persist_layer_norm = no_persist_layer_norm self.sequence_parallel = sequence_parallel # set sequence parallelism flag on weight and bias parameters - setattr(self.weight, 'sequence_parallel', self.sequence_parallel) - setattr(self.bias, 'sequence_parallel', self.sequence_parallel) - - - def reset_parameters(self): + setattr(self.weight, "sequence_parallel", self.sequence_parallel) + setattr(self.bias, "sequence_parallel", self.sequence_parallel) - if self.apply_layernorm_1p: - init.zeros_(self.weight) - init.zeros_(self.bias) - else: - init.ones_(self.weight) - init.zeros_(self.bias) + def reset_parameters(self): - def forward(self, input): + if self.apply_layernorm_1p: + init.zeros_(self.weight) + init.zeros_(self.bias) + else: + init.ones_(self.weight) + init.zeros_(self.bias) + + def forward(self, input): + + weight = self.weight + 1 if self.apply_layernorm_1p else self.weight + # CPU path is here for unittest sake. + if not input.is_cuda: + print( + "WARNING! The input of FusedLayerNorm should be on the GPU." + "This warning should only be triggered in the FusedLayerNorm unit tests." + ) + return F.layer_norm( + input, self.normalized_shape, weight, self.bias, self.eps + ) + + if self.no_persist_layer_norm: + # Apex does not have versions yet (https://github.com/NVIDIA/apex/pull/1648), so we need to inspect + # the function manually on whether the extra arg introduced in https://github.com/NVIDIA/apex/pull/1715 exists yet + if ( + "memory_efficient" + in inspect.getfullargspec(FusedLayerNormAffineFunction.forward).args + ): + return FusedLayerNormAffineFunction.apply( + input, + weight, + self.bias, + self.normalized_shape, + self.eps, + self.mem_efficient_ln, + ) + else: + return FusedLayerNormAffineFunction.apply( + input, weight, self.bias, self.normalized_shape, self.eps + ) + else: + output = FastLayerNormFN.apply(input, weight, self.bias, self.eps) - weight = self.weight + 1 if self.apply_layernorm_1p else self.weight - # CPU path is here for unittest sake. - if not input.is_cuda: - print("WARNING! The input of FusedLayerNorm should be on the GPU." - "This warning should only be triggered in the FusedLayerNorm unit tests.") - return F.layer_norm(input, self.normalized_shape, weight, self.bias, self.eps) + # Apex's fast layer norm function outputs a 'view' tensor (i.e., has + # a populated '_base' field). This will result in schedule.py's + # deallocate_output_tensor() throwing an error, so a viewless tensor is + # created to prevent this. + output = make_viewless_tensor( + inp=output, requires_grad=input.requires_grad, keep_graph=True + ) - if self.no_persist_layer_norm: - # Apex does not have versions yet (https://github.com/NVIDIA/apex/pull/1648), so we need to inspect - # the function manually on whether the extra arg introduced in https://github.com/NVIDIA/apex/pull/1715 exists yet - if 'memory_efficient' in inspect.getfullargspec(FusedLayerNormAffineFunction.forward).args: - return FusedLayerNormAffineFunction.apply(input, weight, self.bias, self.normalized_shape, self.eps, self.mem_efficient_ln) - else: - return FusedLayerNormAffineFunction.apply(input, weight, self.bias, self.normalized_shape, self.eps) - else: - output = FastLayerNormFN.apply(input, weight, self.bias, self.eps) - - # Apex's fast layer norm function outputs a 'view' tensor (i.e., has - # a populated '_base' field). This will result in schedule.py's - # deallocate_output_tensor() throwing an error, so a viewless tensor is - # created to prevent this. - output = make_viewless_tensor(inp = output, - requires_grad = input.requires_grad, - keep_graph = True) - - return output + return output diff --git a/megatron/model/fused_rmsnorm.py b/megatron/model/fused_rmsnorm.py index d69b0822a40..f0c891d4ef6 100644 --- a/megatron/model/fused_rmsnorm.py +++ b/megatron/model/fused_rmsnorm.py @@ -8,9 +8,8 @@ # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__( - self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False - ): + + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False): super().__init__() self.eps = eps self.weight = Parameter(torch.ones(dim, dtype=get_args().params_dtype)) @@ -18,7 +17,5 @@ def __init__( setattr(self.weight, "sequence_parallel", self.sequence_parallel) def forward(self, x): - output = torch.xpu.IpexRmsNorm( - x, self.weight.shape, self.weight, self.eps - ) + output = torch.xpu.IpexRmsNorm(x, self.weight.shape, self.weight, self.eps) return output diff --git a/megatron/model/fused_softmax.py b/megatron/model/fused_softmax.py index 2fe61d40738..e81dc9bffaa 100644 --- a/megatron/model/fused_softmax.py +++ b/megatron/model/fused_softmax.py @@ -81,9 +81,7 @@ def forward(ctx, inputs, scale): scale_t = torch.tensor([scale]) - softmax_results = scaled_softmax_cuda.forward( - inputs, scale_t[0] - ) + softmax_results = scaled_softmax_cuda.forward(inputs, scale_t[0]) ctx.save_for_backward(softmax_results, scale_t) return softmax_results @@ -157,7 +155,7 @@ def is_kernel_available(self, mask, b, np, sq, sk): and self.input_in_float16 # input must be fp16 and 16 < sk <= 4096 # sk must be 16 ~ 2048 and sq % 4 == 0 # sq must be divisor of 4 - and sk % 4 == 0 # sk must be divisor of 4 + and sk % 4 == 0 # sk must be divisor of 4 and attn_batches % 4 == 0 # np * b must be divisor of 4 ): if 0 <= sk <= 4096: diff --git a/megatron/model/gpt_model.py b/megatron/model/gpt_model.py index e5e60c43ee4..a32b0475089 100644 --- a/megatron/model/gpt_model.py +++ b/megatron/model/gpt_model.py @@ -22,7 +22,7 @@ from deepspeed.pipe import PipelineModule, LayerSpec, TiedLayerSpec -try: +try: from deepspeed.checkpoint import ( VOCABULARY_PARAMETER_PATTERNS, PIPELINE_REPLICATED_PARAMETER_PATTERNS, @@ -30,29 +30,30 @@ PARAMETER_WITH_ROW_PARALLELISM_PATTERNS, PARAMETER_WITH_2_SUB_PARAMS_CAT_DIM_0, ) - DS_UNIVERSAL_CHECKPOINT_INFO = True + + DS_UNIVERSAL_CHECKPOINT_INFO = True except ImportError: - DS_UNIVERSAL_CHECKPOINT_INFO = False + DS_UNIVERSAL_CHECKPOINT_INFO = False -def post_language_model_processing(lm_output, labels, logit_weights, - parallel_output, - fp16_lm_cross_entropy): +def post_language_model_processing( + lm_output, labels, logit_weights, parallel_output, fp16_lm_cross_entropy +): # Output. Format [s b h] - output = parallel_lm_logits( - lm_output, - logit_weights, - parallel_output) + output = parallel_lm_logits(lm_output, logit_weights, parallel_output) if labels is None: # [s b h] => [b s h] - return output.transpose(0,1).contiguous() + return output.transpose(0, 1).contiguous() else: # [b s] => [s b] - labels = labels.transpose(0,1).contiguous() - cross_entropy = sequence_parallel.vocab_sequence_parallel_cross_entropy if mpu.get_sequence_parallel_world_size() > 1 \ + labels = labels.transpose(0, 1).contiguous() + cross_entropy = ( + sequence_parallel.vocab_sequence_parallel_cross_entropy + if mpu.get_sequence_parallel_world_size() > 1 else tensor_parallel.vocab_parallel_cross_entropy + ) if fp16_lm_cross_entropy: assert output.dtype == torch.half loss = cross_entropy(output, labels) @@ -60,11 +61,12 @@ def post_language_model_processing(lm_output, labels, logit_weights, loss = cross_entropy(output.float(), labels) # [s b] => [b, s] - loss = loss.transpose(0,1).contiguous() + loss = loss.transpose(0, 1).contiguous() return loss class UniversalCheckpointInfo: + def __init__(self, using_model_pipe: bool): self.using_model_pipe = using_model_pipe self.args = get_args() @@ -81,35 +83,36 @@ def _build_universal_checkpoint_info(self): if self.using_model_pipe: # Replicated (shared) parameters on the pipeline dimension - info[PIPELINE_REPLICATED_PARAMETER_PATTERNS] = self._get_pp_replicated_param_patterns() + info[PIPELINE_REPLICATED_PARAMETER_PATTERNS] = ( + self._get_pp_replicated_param_patterns() + ) if self.args.tensor_model_parallel_size > 1: # Parameter slices that should be averaged not concatenated. - info[TP_REPLICATED_PARAMETER_PATTERNS] = self._get_tp_replicated_param_patterns() + info[TP_REPLICATED_PARAMETER_PATTERNS] = ( + self._get_tp_replicated_param_patterns() + ) # Parameter that are sliced on the row dimension - info[PARAMETER_WITH_ROW_PARALLELISM_PATTERNS] = self._get_row_parallel_param_patterns() + info[PARAMETER_WITH_ROW_PARALLELISM_PATTERNS] = ( + self._get_row_parallel_param_patterns() + ) # SWIGLU parameters are first sliced on dim=0 to tp slices # Then, each tp slice is chunked into 2 to create the linear layers L1, L2 used for silu(L1(x)) * L2(x)) - info[PARAMETER_WITH_2_SUB_PARAMS_CAT_DIM_0] = self._get_swiglu_col_parallel_param_patterns() + info[PARAMETER_WITH_2_SUB_PARAMS_CAT_DIM_0] = ( + self._get_swiglu_col_parallel_param_patterns() + ) return info def _get_vocab_param_patterns(self): if self.using_model_pipe: if self.args.untie_embeddings_and_output_weights: - patterns = [ - r"\d+.word_embeddings.weight", - r"\d+.lm_head.weight" - ] + patterns = [r"\d+.word_embeddings.weight", r"\d+.lm_head.weight"] else: - patterns = [ - r"tied_modules.embed.word_embeddings.weight" - ] + patterns = [r"tied_modules.embed.word_embeddings.weight"] else: - patterns = [ - "language_model.embedding.word_embeddings.weight" - ] + patterns = ["language_model.embedding.word_embeddings.weight"] if self.args.untie_embeddings_and_output_weights: patterns.append("language_model.output_layer.weight") return patterns @@ -132,30 +135,43 @@ def _get_tp_replicated_param_patterns(self): layers_prefix + r"\d+.post_attention_layernorm.weight", ] # Add final normalization layer - final_norm_w_pattern = r"\d+.weight" if self.using_model_pipe \ + final_norm_w_pattern = ( + r"\d+.weight" + if self.using_model_pipe else "language_model.encoder.final_layernorm.weight" + ) patterns.append(final_norm_w_pattern) - if self.args.normalization == 'layernorm': - final_norm_b_pattern = r"\d+.bias" if self.using_model_pipe \ + if self.args.normalization == "layernorm": + final_norm_b_pattern = ( + r"\d+.bias" + if self.using_model_pipe else "language_model.encoder.final_layernorm.bias" + ) patterns.append(final_norm_b_pattern) # add Positional Embedding if self.args.add_position_embedding: - pos_emb_pattern = "tied_modules.embed.position_embeddings.weight" if self.using_model_pipe \ + pos_emb_pattern = ( + "tied_modules.embed.position_embeddings.weight" + if self.using_model_pipe else "language_model.embedding.position_embeddings.weight" + ) patterns.append(pos_emb_pattern) # add Linear bias if self.args.add_bias_linear: - patterns.extend([ - layers_prefix + r"\d+.self_attention.dense.bias", - layers_prefix + r"\d+.mlp.dense_4h_to_h.bias", - ]) + patterns.extend( + [ + layers_prefix + r"\d+.self_attention.dense.bias", + layers_prefix + r"\d+.mlp.dense_4h_to_h.bias", + ] + ) # add LN bias - if self.args.normalization == 'layernorm': - patterns.extend([ - layers_prefix + r"\d+.input_layernorm.bias", - layers_prefix + r"\d+.post_attention_layernorm.bias", - ]) + if self.args.normalization == "layernorm": + patterns.extend( + [ + layers_prefix + r"\d+.input_layernorm.bias", + layers_prefix + r"\d+.post_attention_layernorm.bias", + ] + ) return patterns def _get_row_parallel_param_patterns(self): @@ -180,22 +196,29 @@ def _get_swiglu_col_parallel_param_patterns(self): class GPTModel(MegatronModule): """GPT-2 Language model.""" - def __init__(self, - config, - num_tokentypes=0, - parallel_output=True, - pre_process=True, - post_process=True, - return_moe_loss=True): + def __init__( + self, + config, + num_tokentypes=0, + parallel_output=True, + pre_process=True, + post_process=True, + return_moe_loss=True, + ): args = get_args() - super().__init__(config=config, share_embeddings_and_output_weights=not args.untie_embeddings_and_output_weights) + super().__init__( + config=config, + share_embeddings_and_output_weights=not args.untie_embeddings_and_output_weights, + ) self.parallel_output = parallel_output self.pre_process = pre_process self.post_process = post_process self.fp16_lm_cross_entropy = args.fp16_lm_cross_entropy self.return_moe_loss = return_moe_loss - self.untie_embeddings_and_output_weights = args.untie_embeddings_and_output_weights + self.untie_embeddings_and_output_weights = ( + args.untie_embeddings_and_output_weights + ) self.language_model, self._language_model_key = get_language_model( config=config, @@ -204,7 +227,8 @@ def __init__(self, encoder_attn_mask_type=AttnMaskType.causal, pre_process=self.pre_process, post_process=self.post_process, - num_experts=args.num_experts) + num_experts=args.num_experts, + ) if not args.untie_embeddings_and_output_weights: self.initialize_word_embeddings() @@ -213,12 +237,19 @@ def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" self.language_model.set_input_tensor(input_tensor) - def forward(self, input_ids, position_ids, attention_mask, - retriever_input_ids=None, - retriever_position_ids=None, - retriever_attn_mask=None, - labels=None, tokentype_ids=None, inference_params=None, - curriculum_seqlen=None): + def forward( + self, + input_ids, + position_ids, + attention_mask, + retriever_input_ids=None, + retriever_position_ids=None, + retriever_attn_mask=None, + labels=None, + tokentype_ids=None, + inference_params=None, + curriculum_seqlen=None, + ): args = get_args() if curriculum_seqlen is not None: args.curriculum_seqlen = curriculum_seqlen @@ -231,7 +262,9 @@ def forward(self, input_ids, position_ids, attention_mask, labels = labels[:, :curriculum_seqlen].contiguous() # attention_mask has size [1, 1, seqlen, seqlen] - attention_mask = attention_mask[:, :, :curriculum_seqlen, :curriculum_seqlen].contiguous() + attention_mask = attention_mask[ + :, :, :curriculum_seqlen, :curriculum_seqlen + ].contiguous() else: if args.curriculum_learning_legacy: # If got a None input, need to reset curriculum_seqlen on user side @@ -244,22 +277,28 @@ def forward(self, input_ids, position_ids, attention_mask, retriever_input_ids=retriever_input_ids, retriever_position_ids=retriever_position_ids, retriever_attn_mask=retriever_attn_mask, - inference_params=inference_params) + inference_params=inference_params, + ) if self.post_process: lm_output = post_language_model_processing( - lm_output, labels, - self.language_model.output_layer.weight if self.untie_embeddings_and_output_weights else self.shared_embedding_or_output_weight(), + lm_output, + labels, + self.language_model.output_layer.weight + if self.untie_embeddings_and_output_weights + else self.shared_embedding_or_output_weight(), self.parallel_output, - self.fp16_lm_cross_entropy) + self.fp16_lm_cross_entropy, + ) return lm_output, moe_losses if self.return_moe_loss else lm_output - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): state_dict_ = {} language_model_state_dict = self.language_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + prefix=prefix, keep_vars=keep_vars + ) # MoE states need to be handled separately by DeepSpeed engine, thus # moving them to the top level dictionary if "moe_state_dict" in language_model_state_dict: @@ -268,23 +307,32 @@ def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): del language_model_state_dict["moe_state_dict"] state_dict_[self._language_model_key] = language_model_state_dict # Save word_embeddings. - if self.post_process and not self.pre_process and not self.untie_embeddings_and_output_weights: - state_dict_[self._word_embeddings_for_head_key] \ - = self.word_embeddings.state_dict(prefix=prefix, - keep_vars=keep_vars) + if ( + self.post_process + and not self.pre_process + and not self.untie_embeddings_and_output_weights + ): + state_dict_[self._word_embeddings_for_head_key] = ( + self.word_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" # Load word_embeddings. - if self.post_process and not self.pre_process and not self.untie_embeddings_and_output_weights: + if ( + self.post_process + and not self.pre_process + and not self.untie_embeddings_and_output_weights + ): self.word_embeddings.load_state_dict( - state_dict[self._word_embeddings_for_head_key], strict=strict) + state_dict[self._word_embeddings_for_head_key], strict=strict + ) # Gather MoE states and move under language model moe_state_dict = {} for key in list(state_dict.keys()): - if 'expert' in key and 'moe.gate.wg.weight' not in key: + if "expert" in key and "moe.gate.wg.weight" not in key: moe_state_dict[key] = state_dict.pop(key) if self._language_model_key in state_dict: state_dict = state_dict[self._language_model_key] @@ -303,7 +351,9 @@ def CrossEntropy(output, labels): # [b s] => [s b] labels = labels.transpose(0, 1).contiguous() - losses = tensor_parallel.vocab_parallel_cross_entropy(output.contiguous().float(), labels) + losses = tensor_parallel.vocab_parallel_cross_entropy( + output.contiguous().float(), labels + ) # [s b] => [b, s] losses = losses.transpose(0, 1).contiguous() loss_mask = loss_mask.view(-1) @@ -311,13 +361,10 @@ def CrossEntropy(output, labels): return loss -class GPTModelPipe(PipelineModule,MegatronModule): +class GPTModelPipe(PipelineModule, MegatronModule): """GPT-2 Language model.""" - def __init__(self, - config, - num_tokentypes=0, - parallel_output=True): + def __init__(self, config, num_tokentypes=0, parallel_output=True): args = get_args() self.parallel_output = parallel_output @@ -325,8 +372,9 @@ def __init__(self, config.init_method = init_method_normal(config.init_method_std) if config.output_layer_init_method is None: - config.output_layer_init_method = scaled_init_method_normal(config.init_method_std, - config.num_layers) + config.output_layer_init_method = scaled_init_method_normal( + config.init_method_std, config.num_layers + ) self.specs = [] @@ -342,48 +390,70 @@ def _to_float16(inputs): # Embedding layer if args.untie_embeddings_and_output_weights: - self.specs.append(LayerSpec(EmbeddingPipe, - args.hidden_size, - args.padded_vocab_size, - args.max_position_embeddings, - args.hidden_dropout, - config, - num_tokentypes=num_tokentypes, - embedding_weights_in_fp32=args.embedding_weights_in_fp32,)) + self.specs.append( + LayerSpec( + EmbeddingPipe, + args.hidden_size, + args.padded_vocab_size, + args.max_position_embeddings, + args.hidden_dropout, + config, + num_tokentypes=num_tokentypes, + embedding_weights_in_fp32=args.embedding_weights_in_fp32, + ) + ) else: - self.specs.append(TiedLayerSpec('embed', - EmbeddingPipe, - args.hidden_size, - args.padded_vocab_size, - args.max_position_embeddings, - args.hidden_dropout, - config, - num_tokentypes=num_tokentypes, - embedding_weights_in_fp32=args.embedding_weights_in_fp32, - tied_weight_attr='word_embeddings_weight')) - - experts_per_layer = get_num_experts_per_layer(args.num_experts, args.num_layers, args.expert_interval) + self.specs.append( + TiedLayerSpec( + "embed", + EmbeddingPipe, + args.hidden_size, + args.padded_vocab_size, + args.max_position_embeddings, + args.hidden_dropout, + config, + num_tokentypes=num_tokentypes, + embedding_weights_in_fp32=args.embedding_weights_in_fp32, + tied_weight_attr="word_embeddings_weight", + ) + ) + + experts_per_layer = get_num_experts_per_layer( + args.num_experts, args.num_layers, args.expert_interval + ) self.is_moe_model = any(n_experts > 1 for n_experts in experts_per_layer) # Currently PipelineEngine does not support more than 1 pipe and/or grad partitioned tensors that # require grads. # When using MoE, we have 2 tensors that are passed along pipeline stages and both require grads. # Therefore, verify that both pipe_partitioned / grad_partitioned are not enabled - if self.is_moe_model and args.pipeline_model_parallel_size > 1 and args.tensor_model_parallel_size > 1: - pipe_partitioned_enabled = args.deepspeed_config_dict.get('pipeline', {}).get('pipe_partitioned', False) - grad_partitioned_enabled = args.deepspeed_config_dict.get('pipeline', {}).get('grad_partitioned', False) - assert not pipe_partitioned_enabled and not grad_partitioned_enabled, \ - 'Pipe and/or Grad partitioning are not supported for MoE model' + if ( + self.is_moe_model + and args.pipeline_model_parallel_size > 1 + and args.tensor_model_parallel_size > 1 + ): + pipe_partitioned_enabled = args.deepspeed_config_dict.get( + "pipeline", {} + ).get("pipe_partitioned", False) + grad_partitioned_enabled = args.deepspeed_config_dict.get( + "pipeline", {} + ).get("grad_partitioned", False) + assert ( + not pipe_partitioned_enabled and not grad_partitioned_enabled + ), "Pipe and/or Grad partitioning are not supported for MoE model" for layer_idx in range(args.num_layers): self.specs.append( - LayerSpec(ParallelTransformerLayerPipe, - config, - layer_number=layer_idx, - self_attn_mask_type=AttnMaskType.causal, - num_experts=experts_per_layer[layer_idx], - input_aggregated_moe_loss=(self.is_moe_model and layer_idx > 0), - return_aggregated_moe_loss=self.is_moe_model)) + LayerSpec( + ParallelTransformerLayerPipe, + config, + layer_number=layer_idx, + self_attn_mask_type=AttnMaskType.causal, + num_experts=experts_per_layer[layer_idx], + input_aggregated_moe_loss=(self.is_moe_model and layer_idx > 0), + return_aggregated_moe_loss=self.is_moe_model, + ) + ) # if model has experts, add a layer to get and cache the aggregated moe loss from the # last transformer layer @@ -391,39 +461,50 @@ def _to_float16(inputs): self.specs.append(self._calculate_moe_loss) # Final layernorm after transformer layers - if args.normalization == 'layernorm': - self.specs.append(LayerSpec(LayerNorm, - args.hidden_size, - eps=args.layernorm_epsilon, - sequence_parallel=args.sequence_parallel)) + if args.normalization == "layernorm": + self.specs.append( + LayerSpec( + LayerNorm, + args.hidden_size, + eps=args.layernorm_epsilon, + sequence_parallel=args.sequence_parallel, + ) + ) else: - self.specs.append(LayerSpec(RMSNorm, args.hidden_size, - args.layernorm_epsilon, - sequence_parallel=args.sequence_parallel)) + self.specs.append( + LayerSpec( + RMSNorm, + args.hidden_size, + args.layernorm_epsilon, + sequence_parallel=args.sequence_parallel, + ) + ) def _logits_helper(embedding, lm_output): - """A wrapper to massage inputs/outputs from pipeline. """ + """A wrapper to massage inputs/outputs from pipeline.""" return parallel_lm_logits( - lm_output, - embedding.word_embeddings_weight, - self.parallel_output) + lm_output, embedding.word_embeddings_weight, self.parallel_output + ) + if args.untie_embeddings_and_output_weights: self.specs.append( LayerSpec(LMHeadPipe, args.hidden_size, args.padded_vocab_size, config) ) else: self.specs.append( - TiedLayerSpec('embed', - EmbeddingPipe, - args.hidden_size, - args.padded_vocab_size, - args.max_position_embeddings, - args.hidden_dropout, - config, - num_tokentypes=num_tokentypes, - embedding_weights_in_fp32=args.embedding_weights_in_fp32, - forward_fn=_logits_helper, - tied_weight_attr='word_embeddings_weight') + TiedLayerSpec( + "embed", + EmbeddingPipe, + args.hidden_size, + args.padded_vocab_size, + args.max_position_embeddings, + args.hidden_dropout, + config, + num_tokentypes=num_tokentypes, + embedding_weights_in_fp32=args.embedding_weights_in_fp32, + forward_fn=_logits_helper, + tied_weight_attr="word_embeddings_weight", + ) ) # Convert to fp32 if needed @@ -432,30 +513,37 @@ def _logits_helper(embedding, lm_output): # Cache losses self.moe_loss = None - self.last_lm_loss = None # detached, for display only - self.last_moe_loss = None # detached, for display only + self.last_lm_loss = None # detached, for display only + self.last_moe_loss = None # detached, for display only if args.checkpoint_activations: interval = args.checkpoint_num_layers - elif args.recompute_granularity == "full" and args.recompute_method == 'uniform': + elif ( + args.recompute_granularity == "full" and args.recompute_method == "uniform" + ): # deepspeed's pipeline doesn't support the block recompute method interval = args.recompute_num_layers else: interval = 0 from deepspeed.runtime.pipe.topology import PipeModelDataParallelTopology - topo = PipeModelDataParallelTopology(num_pp=mpu.get_pipeline_model_parallel_world_size(), - num_mp=mpu.get_tensor_model_parallel_world_size(), - num_dp=mpu.get_data_parallel_world_size()) - super().__init__(layers=self.specs, - loss_fn=self.loss_func, - topology=topo, - activation_checkpoint_interval=interval, - partition_method='type:transformer') + topo = PipeModelDataParallelTopology( + num_pp=mpu.get_pipeline_model_parallel_world_size(), + num_mp=mpu.get_tensor_model_parallel_world_size(), + num_dp=mpu.get_data_parallel_world_size(), + ) + + super().__init__( + layers=self.specs, + loss_fn=self.loss_func, + topology=topo, + activation_checkpoint_interval=interval, + partition_method="type:transformer", + ) def _calculate_moe_loss(self, inputs): - """ Calculate MoE auxiliary loss """ + """Calculate MoE auxiliary loss""" assert isinstance(inputs, tuple) and len(inputs) == 2 hidden, aggregated_moe_loss = inputs[0], inputs[1] args = get_args() @@ -476,7 +564,6 @@ def universal_checkpoint_info(self): def get_additional_losses(self): if not self.is_moe_model: return None - return OrderedDict({ - 'lm loss': self.last_lm_loss, - 'moe loss': self.last_moe_loss - }) + return OrderedDict( + {"lm loss": self.last_lm_loss, "moe loss": self.last_moe_loss} + ) diff --git a/megatron/model/language_model.py b/megatron/model/language_model.py index eebf8744ca5..901bb4c661c 100644 --- a/megatron/model/language_model.py +++ b/megatron/model/language_model.py @@ -18,17 +18,18 @@ from .utils import init_method_normal, scaled_init_method_normal, gather_and_init -def parallel_lm_logits(input_, word_embeddings_weight, parallel_output, - bias=None): +def parallel_lm_logits(input_, word_embeddings_weight, parallel_output, bias=None): """LM logits using word embedding weights.""" args = get_args() # Parallel logits. - if args.async_tensor_model_parallel_allreduce or\ - args.sequence_parallel: + if args.async_tensor_model_parallel_allreduce or args.sequence_parallel: input_parallel = input_ model_parallel = mpu.get_tensor_model_parallel_world_size() > 1 - async_grad_allreduce = args.async_tensor_model_parallel_allreduce and \ - model_parallel and not args.sequence_parallel + async_grad_allreduce = ( + args.async_tensor_model_parallel_allreduce + and model_parallel + and not args.sequence_parallel + ) else: input_parallel = tensor_parallel.copy_to_tensor_model_parallel_region(input_) async_grad_allreduce = False @@ -40,7 +41,8 @@ def parallel_lm_logits(input_, word_embeddings_weight, parallel_output, bias=bias, gradient_accumulation_fusion=args.gradient_accumulation_fusion, async_grad_allreduce=async_grad_allreduce, - sequence_parallel=args.sequence_parallel) + sequence_parallel=args.sequence_parallel, + ) # Gather if needed. if parallel_output: @@ -49,20 +51,27 @@ def parallel_lm_logits(input_, word_embeddings_weight, parallel_output, return tensor_parallel.gather_from_tensor_model_parallel_region(logits_parallel) -def get_language_model(config, num_tokentypes, add_pooler, - encoder_attn_mask_type, - add_encoder=True, - add_decoder=False, - decoder_attn_mask_type=AttnMaskType.causal, - pre_process=True, post_process=True, num_experts=[1]): +def get_language_model( + config, + num_tokentypes, + add_pooler, + encoder_attn_mask_type, + add_encoder=True, + add_decoder=False, + decoder_attn_mask_type=AttnMaskType.causal, + pre_process=True, + post_process=True, + num_experts=[1], +): """Build language model and return along with the key to save.""" args = get_args() if config.init_method is None: config.init_method = init_method_normal(config.init_method_std) if config.output_layer_init_method is None: - config.output_layer_init_method = scaled_init_method_normal(config.init_method_std, - config.num_layers) + config.output_layer_init_method = scaled_init_method_normal( + config.init_method_std, config.num_layers + ) # Language model. language_model = TransformerLanguageModel( @@ -75,9 +84,10 @@ def get_language_model(config, num_tokentypes, add_pooler, add_pooler=add_pooler, pre_process=pre_process, post_process=post_process, - num_experts=num_experts) + num_experts=num_experts, + ) # key used for checkpoints. - language_model_key = 'language_model' + language_model_key = "language_model" return language_model, language_model_key @@ -97,10 +107,14 @@ class Pooler(MegatronModule): def __init__(self, hidden_size, init_method): super(Pooler, self).__init__() args = get_args() - self.dense = get_linear_layer(hidden_size, hidden_size, init_method, gather_params_on_init=args.zero_stage == 3) + self.dense = get_linear_layer( + hidden_size, + hidden_size, + init_method, + gather_params_on_init=args.zero_stage == 3, + ) self.sequence_parallel = args.sequence_parallel - def forward(self, hidden_states, sequence_index=0): # hidden_states: [s, b, h] # sequence_index: index of the token to pool. @@ -109,8 +123,8 @@ def forward(self, hidden_states, sequence_index=0): # same pooler is run on all tensor parallel nodes if self.sequence_parallel: hidden_states = tensor_parallel.gather_from_sequence_parallel_region( - hidden_states, - tensor_parallel_output_grad=False) + hidden_states, tensor_parallel_output_grad=False + ) pooled = hidden_states[sequence_index, :, :] pooled = self.dense(pooled) @@ -136,14 +150,16 @@ class Embedding(MegatronModule): training in bf16. """ - def __init__(self, - hidden_size, - vocab_size, - max_sequence_length, - embedding_dropout_prob, - config, - num_tokentypes=0, - embedding_weights_in_fp32=False): + def __init__( + self, + hidden_size, + vocab_size, + max_sequence_length, + embedding_dropout_prob, + config, + num_tokentypes=0, + embedding_weights_in_fp32=False, + ): super(Embedding, self).__init__() self.hidden_size = hidden_size @@ -156,25 +172,32 @@ def __init__(self, self.embedding_weights_in_fp32 = embedding_weights_in_fp32 self.params_dtype = args.params_dtype self.word_embeddings = tensor_parallel.VocabParallelEmbedding( - vocab_size, self.hidden_size, config=config, init_method=config.init_method) - self._word_embeddings_key = 'word_embeddings' + vocab_size, self.hidden_size, config=config, init_method=config.init_method + ) + self._word_embeddings_key = "word_embeddings" # Position embedding (serial). self.add_position_embedding = args.add_position_embedding if self.add_position_embedding: - self._position_embeddings_key = 'position_embeddings' + self._position_embeddings_key = "position_embeddings" if args.sequence_parallel: - self.position_embeddings = tensor_parallel.layers.SequenceParallelPositionEmbedding( - max_sequence_length, self.hidden_size) + self.position_embeddings = ( + tensor_parallel.layers.SequenceParallelPositionEmbedding( + max_sequence_length, self.hidden_size + ) + ) # Initialize the position embeddings. self.init_method(self.position_embeddings.local_embeddings.weight) else: self.position_embeddings = torch.nn.Embedding( - max_sequence_length, self.hidden_size) + max_sequence_length, self.hidden_size + ) # Initialize the position embeddings. if args.perform_initialization: if args.zero_stage == 3: - gather_and_init(self.position_embeddings.weight, self.init_method) + gather_and_init( + self.position_embeddings.weight, self.init_method + ) else: self.init_method(self.position_embeddings.weight) @@ -182,10 +205,11 @@ def __init__(self, # Add this as an optional field that can be added through # method call so we can load a pretrain model without # token types and add them as needed. - self._tokentype_embeddings_key = 'tokentype_embeddings' + self._tokentype_embeddings_key = "tokentype_embeddings" if self.num_tokentypes > 0: - self.tokentype_embeddings = torch.nn.Embedding(self.num_tokentypes, - self.hidden_size) + self.tokentype_embeddings = torch.nn.Embedding( + self.num_tokentypes, self.hidden_size + ) # Initialize the token-type embeddings. if args.perform_initialization: if args.zero_stage == 3: @@ -217,13 +241,13 @@ def add_tokentype_embeddings(self, num_tokentypes): This allows us to load the model normally and then add this embedding. """ if self.tokentype_embeddings is not None: - raise Exception('tokentype embeddings is already initialized') + raise Exception("tokentype embeddings is already initialized") if torch.distributed.get_rank() == 0: - print('adding embedding for {} tokentypes'.format(num_tokentypes), - flush=True) + print( + "adding embedding for {} tokentypes".format(num_tokentypes), flush=True + ) self.num_tokentypes = num_tokentypes - self.tokentype_embeddings = torch.nn.Embedding(num_tokentypes, - self.hidden_size) + self.tokentype_embeddings = torch.nn.Embedding(num_tokentypes, self.hidden_size) # Initialize the token-type embeddings. args = get_args() self.init_method(self.tokentype_embeddings.weight) @@ -266,21 +290,21 @@ def forward(self, input_ids, position_ids, tokentype_ids=None): return embeddings - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load.""" state_dict_ = {} - state_dict_[self._word_embeddings_key] \ - = self.word_embeddings.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._word_embeddings_key] = self.word_embeddings.state_dict( + prefix=prefix, keep_vars=keep_vars + ) if self.add_position_embedding: - state_dict_[self._position_embeddings_key] \ - = self.position_embeddings.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._position_embeddings_key] = ( + self.position_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) if self.num_tokentypes > 0: - state_dict_[self._tokentype_embeddings_key] \ - = self.tokentype_embeddings.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._tokentype_embeddings_key] = ( + self.tokentype_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ @@ -294,9 +318,8 @@ def load_state_dict(self, state_dict, strict=True): # for backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if 'word_embeddings' in key: - state_dict_[key.split('word_embeddings.')[1]] \ - = state_dict[key] + if "word_embeddings" in key: + state_dict_[key.split("word_embeddings.")[1]] = state_dict[key] self.word_embeddings.load_state_dict(state_dict_, strict=strict) # Position embedding. @@ -307,9 +330,10 @@ def load_state_dict(self, state_dict, strict=True): # for backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if 'position_embeddings' in key: - state_dict_[key.split('position_embeddings.')[1]] \ - = state_dict[key] + if "position_embeddings" in key: + state_dict_[key.split("position_embeddings.")[1]] = state_dict[ + key + ] self.position_embeddings.load_state_dict(state_dict_, strict=strict) # Tokentype embedding. @@ -320,26 +344,29 @@ def load_state_dict(self, state_dict, strict=True): else: # for backward compatibility. for key in state_dict.keys(): - if 'tokentype_embeddings' in key: - state_dict_[key.split('tokentype_embeddings.')[1]] \ - = state_dict[key] + if "tokentype_embeddings" in key: + state_dict_[key.split("tokentype_embeddings.")[1]] = state_dict[ + key + ] if len(state_dict_.keys()) > 0: - self.tokentype_embeddings.load_state_dict(state_dict_, - strict=strict) + self.tokentype_embeddings.load_state_dict(state_dict_, strict=strict) else: - print('***WARNING*** expected tokentype embeddings in the ' - 'checkpoint but could not find it', flush=True) + print( + "***WARNING*** expected tokentype embeddings in the " + "checkpoint but could not find it", + flush=True, + ) class EmbeddingPipe(Embedding): def forward(self, inputs, **kwargs): - if not hasattr(self, '_args'): + if not hasattr(self, "_args"): self._args = get_args() input_ids = inputs[0] position_ids = inputs[1] - if hasattr(self._args, 'attn_mask'): + if hasattr(self._args, "attn_mask"): attention_mask = None else: attention_mask = inputs[2] @@ -348,17 +375,18 @@ def forward(self, inputs, **kwargs): tokentype_ids = inputs[3] else: tokentype_ids = None - - embeddings = super().forward(input_ids, position_ids, tokentype_ids=tokentype_ids) + + embeddings = super().forward( + input_ids, position_ids, tokentype_ids=tokentype_ids + ) # If cmd args has attn_mask, we don't forward it as an activation. - if hasattr(self._args, 'attn_mask'): + if hasattr(self._args, "attn_mask"): return embeddings else: assert False return embeddings, attention_mask - @property def word_embeddings_weight(self): """Easy accessory for the DeepSpeed pipeline engine to tie embeddings across stages.""" @@ -378,17 +406,19 @@ class TransformerLanguageModel(MegatronModule): will ignore this embedding """ - def __init__(self, - config, - encoder_attn_mask_type, - num_tokentypes=0, - add_encoder=True, - add_decoder=False, - decoder_attn_mask_type=AttnMaskType.causal, - add_pooler=False, - pre_process=True, - post_process=True, - num_experts=[1]): + def __init__( + self, + config, + encoder_attn_mask_type, + num_tokentypes=0, + add_encoder=True, + add_decoder=False, + decoder_attn_mask_type=AttnMaskType.causal, + add_pooler=False, + pre_process=True, + post_process=True, + num_experts=[1], + ): args = get_args() # TODO: passing `share_embeddings_and_output_weights=False` # will not work correctly for T5 and embeddings will not be synced. @@ -397,8 +427,8 @@ def __init__(self, assert not add_decoder super(TransformerLanguageModel, self).__init__( share_embeddings_and_output_weights=( - not args.untie_embeddings_and_output_weights - ) + not args.untie_embeddings_and_output_weights + ) ) self.pre_process = pre_process self.post_process = post_process @@ -426,20 +456,18 @@ def __init__(self, args.hidden_dropout, config, self.num_tokentypes, - args.embedding_weights_in_fp32 + args.embedding_weights_in_fp32, ) - self._embedding_key = 'embedding' + self._embedding_key = "embedding" # Rotary positional embeddings - self.use_rotary_position_embeddings = ( - args.use_rotary_position_embeddings - ) + self.use_rotary_position_embeddings = args.use_rotary_position_embeddings if args.use_rotary_position_embeddings: self.seq_length = args.seq_length rotary_dim = ( - args.hidden_size // args.num_attention_heads - if args.kv_channels is None - else args.kv_channels + args.hidden_size // args.num_attention_heads + if args.kv_channels is None + else args.kv_channels ) if args.rotary_percent < 1.0: @@ -448,10 +476,7 @@ def __init__(self, # partial rotary embeddings, which is better than full rotary # Wang and Komatsuzaki et al # https://github.com/kingoflolz/mesh-transformer-jax/ - self.rotary_pos_emb = RotaryEmbedding( - rotary_dim, - theta=args.rope_theta - ) + self.rotary_pos_emb = RotaryEmbedding(rotary_dim, theta=args.rope_theta) # Encoder (usually set to True, False if part of an encoder-decoder # architecture and in encoder-only stage). @@ -461,15 +486,16 @@ def __init__(self, # args.model_type if not args.retro_add_retriever # else ModelType.retro_decoder model_type=( - ModelType.retro_decoder if args.retro_add_retriever + ModelType.retro_decoder + if args.retro_add_retriever else args.model_type ), self_attn_mask_type=self.encoder_attn_mask_type, pre_process=self.pre_process, post_process=self.post_process, - num_experts=self.num_experts + num_experts=self.num_experts, ) - self._encoder_key = 'encoder' + self._encoder_key = "encoder" else: self.encoder = None @@ -483,9 +509,9 @@ def __init__(self, self_attn_mask_type=self.decoder_attn_mask_type, pre_process=self.pre_process, post_process=self.post_process, - num_experts=self.num_experts + num_experts=self.num_experts, ) - self._decoder_key = 'decoder' + self._decoder_key = "decoder" else: self.decoder = None @@ -493,7 +519,7 @@ def __init__(self, # Pooler. if self.add_pooler: self.pooler = Pooler(self.hidden_size, self.init_method) - self._pooler_key = 'pooler' + self._pooler_key = "pooler" if self.untie_embeddings_and_output_weights: self.output_layer = tensor_parallel.ColumnParallelLinear( @@ -503,12 +529,12 @@ def __init__(self, init_method=self.init_method, # Setting bias to False always to keep it consistent with # embedding tying that also does not have a bias. - bias=False + bias=False, ) - self._output_layer_key = 'output_layer' + self._output_layer_key = "output_layer" def set_input_tensor(self, input_tensor): - """ See megatron.model.transformer.set_input_tensor()""" + """See megatron.model.transformer.set_input_tensor()""" # This is usually handled in schedules.py but some inference code still # gives us non-lists or None if not isinstance(input_tensor, list): @@ -516,14 +542,13 @@ def set_input_tensor(self, input_tensor): if self.add_encoder and self.add_decoder: assert len(input_tensor) == 1, ( - 'input_tensor should only be length 1 ' - 'for stage with both encoder and decoder' + "input_tensor should only be length 1 " + "for stage with both encoder and decoder" ) self.encoder.set_input_tensor(input_tensor[0]) elif self.add_encoder: assert len(input_tensor) == 1, ( - 'input_tensor should only be length 1 ' - 'for stage with only encoder' + "input_tensor should only be length 1 " "for stage with only encoder" ) self.encoder.set_input_tensor(input_tensor[0]) elif self.add_decoder: @@ -534,39 +559,33 @@ def set_input_tensor(self, input_tensor): self.decoder.set_input_tensor(None) self.encoder_hidden_state = input_tensor[0] else: - raise Exception( - 'input_tensor must have either length 1 or 2' - ) + raise Exception("input_tensor must have either length 1 or 2") else: - raise Exception( - 'Stage must have at least either encoder or decoder' - ) + raise Exception("Stage must have at least either encoder or decoder") def forward( - self, - enc_input_ids, - enc_position_ids, - enc_attn_mask, - dec_input_ids=None, - dec_position_ids=None, - dec_attn_mask=None, - retriever_input_ids=None, - retriever_position_ids=None, - retriever_attn_mask=None, - enc_dec_attn_mask=None, - tokentype_ids=None, - inference_params=None, - pooling_sequence_index=0, - enc_hidden_states=None, - output_enc_hidden=False + self, + enc_input_ids, + enc_position_ids, + enc_attn_mask, + dec_input_ids=None, + dec_position_ids=None, + dec_attn_mask=None, + retriever_input_ids=None, + retriever_position_ids=None, + retriever_attn_mask=None, + enc_dec_attn_mask=None, + tokentype_ids=None, + inference_params=None, + pooling_sequence_index=0, + enc_hidden_states=None, + output_enc_hidden=False, ): args = get_args() # Encoder embedding. if self.pre_process: encoder_input = self.embedding( - enc_input_ids, - enc_position_ids, - tokentype_ids=tokentype_ids + enc_input_ids, enc_position_ids, tokentype_ids=tokentype_ids ) else: encoder_input = None @@ -574,9 +593,7 @@ def forward( # Retriever embedding. if self.add_retriever and self.pre_process: retriever_input = self.embedding( - retriever_input_ids, - retriever_position_ids, - tokentype_ids=tokentype_ids + retriever_input_ids, retriever_position_ids, tokentype_ids=tokentype_ids ) else: retriever_input = None @@ -585,10 +602,12 @@ def forward( rotary_pos_emb = None if self.use_rotary_position_embeddings: if inference_params is not None: - rotary_pos_emb = \ - self.rotary_pos_emb(inference_params.max_sequence_len) + rotary_pos_emb = self.rotary_pos_emb(inference_params.max_sequence_len) else: - if args.curriculum_learning_legacy or args.data_efficiency_curriculum_learning: + if ( + args.curriculum_learning_legacy + or args.data_efficiency_curriculum_learning + ): rotary_pos_emb = self.rotary_pos_emb(args.curriculum_seqlen) else: rotary_pos_emb = self.rotary_pos_emb(self.seq_length) @@ -602,16 +621,19 @@ def forward( retriever_input=retriever_input, retriever_attn_mask=retriever_attn_mask, inference_params=inference_params, - rotary_pos_emb=rotary_pos_emb) + rotary_pos_emb=rotary_pos_emb, + ) else: encoder_output = self.encoder_hidden_state else: - encoder_output, encoder_moe_losses = enc_hidden_states.to(encoder_input.dtype), [] + encoder_output, encoder_moe_losses = ( + enc_hidden_states.to(encoder_input.dtype), + [], + ) if self.post_process: if self.add_pooler: - pooled_output = self.pooler(encoder_output, - pooling_sequence_index) + pooled_output = self.pooler(encoder_output, pooling_sequence_index) # output_enc_hidden refers to when we just need the encoder's # output. For example, it is helpful to compute @@ -624,8 +646,7 @@ def forward( # Decoder embedding. if self.pre_process: - decoder_input = self.embedding(dec_input_ids, - dec_position_ids) + decoder_input = self.embedding(dec_input_ids, dec_position_ids) else: decoder_input = None @@ -636,25 +657,40 @@ def forward( encoder_output=encoder_output, enc_dec_attn_mask=enc_dec_attn_mask, inference_params=inference_params, - rotary_pos_emb=rotary_pos_emb) + rotary_pos_emb=rotary_pos_emb, + ) if self.add_pooler and self.post_process: - return decoder_output, encoder_output, pooled_output, decoder_moe_losses, encoder_moe_losses + return ( + decoder_output, + encoder_output, + pooled_output, + decoder_moe_losses, + encoder_moe_losses, + ) else: - return decoder_output, encoder_output, decoder_moe_losses, encoder_moe_losses + return ( + decoder_output, + encoder_output, + decoder_moe_losses, + encoder_moe_losses, + ) - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load.""" args = get_args() state_dict_ = {} moe_state_dict = {} if self.pre_process: - state_dict_[self._embedding_key] \ - = self.embedding.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._embedding_key] = ( + self.embedding.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.add_encoder: encoder_state_dict = self.encoder.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + prefix=prefix, keep_vars=keep_vars + ) if args.random_ltd: # When using random-LTD, it is required to call remove_random_ltd_state_dict # during model checkpoint saving to transfer the random-LTD-wrapped @@ -663,29 +699,37 @@ def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): # finetuning of the checkpoint there is no need to depend on random-LTD # again. from deepspeed.runtime.data_pipeline.data_routing.helper import remove_random_ltd_state_dict + encoder_state_dict = remove_random_ltd_state_dict(encoder_state_dict) # MoE states need to be handled separately by DeepSpeed engine, thus # moving them to the top level dictionary # If components other than encoder may contain MoE states, need to add # the same logic for key in list(encoder_state_dict.keys()): - if 'expert' in key and 'moe.gate.wg.weight' not in key: - moe_state_dict[self._encoder_key+key] = encoder_state_dict.pop(key) + if "expert" in key and "moe.gate.wg.weight" not in key: + moe_state_dict[self._encoder_key + key] = encoder_state_dict.pop( + key + ) state_dict_[self._encoder_key] = encoder_state_dict if self.post_process: if self.add_pooler: - state_dict_[self._pooler_key] \ - = self.pooler.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._pooler_key] = ( + self.pooler.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.untie_embeddings_and_output_weights: - state_dict_[self._output_layer_key] \ - = self.output_layer.state_dict(prefix=prefix, keep_vars=keep_vars) + state_dict_[self._output_layer_key] = self.output_layer.state_dict( + prefix=prefix, keep_vars=keep_vars + ) if self.add_decoder: - state_dict_[self._decoder_key] \ - = self.decoder.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._decoder_key] = ( + self.decoder.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) state_dict_["moe_state_dict"] = moe_state_dict return state_dict_ @@ -701,7 +745,7 @@ def load_state_dict(self, state_dict, strict=True): # for backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if '_embeddings' in key: + if "_embeddings" in key: state_dict_[key] = state_dict[key] self.embedding.load_state_dict(state_dict_, strict=strict) @@ -710,14 +754,14 @@ def load_state_dict(self, state_dict, strict=True): if self._encoder_key in state_dict: state_dict_ = state_dict[self._encoder_key] # For backward compatibility. - elif 'transformer' in state_dict: - state_dict_ = state_dict['transformer'] + elif "transformer" in state_dict: + state_dict_ = state_dict["transformer"] else: # For backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if 'transformer.' in key: - state_dict_[key.split('transformer.')[1]] = state_dict[key] + if "transformer." in key: + state_dict_[key.split("transformer.")[1]] = state_dict[key] # For backward compatibility. # Somehow this backward compatibility could be wrong: sometimes @@ -726,9 +770,10 @@ def load_state_dict(self, state_dict, strict=True): state_dict_self_attention = {} encoder_state_dict_keys = list(self.encoder.state_dict().keys()) for key in state_dict_.keys(): - if '.attention.' in key and key not in encoder_state_dict_keys: - state_dict_self_attention[key.replace(".attention.", - ".self_attention.")] = state_dict_[key] + if ".attention." in key and key not in encoder_state_dict_keys: + state_dict_self_attention[ + key.replace(".attention.", ".self_attention.") + ] = state_dict_[key] else: state_dict_self_attention[key] = state_dict_[key] state_dict_ = state_dict_self_attention @@ -737,11 +782,11 @@ def load_state_dict(self, state_dict, strict=True): if "moe_state_dict" in state_dict: for key in list(state_dict["moe_state_dict"].keys()): if self._encoder_key in key: - key_list = key.split('.') - while key_list[0] != 'encoder': + key_list = key.split(".") + while key_list[0] != "encoder": key_list.pop(0) key_list.pop(0) - actual_key = '.'.join(key_list) + actual_key = ".".join(key_list) state_dict_[actual_key] = state_dict["moe_state_dict"].pop(key) if len(state_dict["moe_state_dict"]) == 0: del state_dict["moe_state_dict"] @@ -750,18 +795,20 @@ def load_state_dict(self, state_dict, strict=True): # Pooler. if self.post_process: if self.add_pooler: - assert 'pooler' in state_dict, \ - 'could not find data for pooler in the checkpoint' - self.pooler.load_state_dict(state_dict[self._pooler_key], - strict=strict) + assert ( + "pooler" in state_dict + ), "could not find data for pooler in the checkpoint" + self.pooler.load_state_dict(state_dict[self._pooler_key], strict=strict) if self.untie_embeddings_and_output_weights: - assert 'output_layer' in state_dict, \ - 'could not find data for output_layer in the checkpoint' - self.output_layer.load_state_dict(state_dict[self._output_layer_key], - strict=strict) + assert ( + "output_layer" in state_dict + ), "could not find data for output_layer in the checkpoint" + self.output_layer.load_state_dict( + state_dict[self._output_layer_key], strict=strict + ) # Decoder. if self.add_decoder: - assert 'decoder' in state_dict, \ - 'could not find data for pooler in the checkpoint' - self.decoder.load_state_dict(state_dict[self._decoder_key], - strict=strict) + assert ( + "decoder" in state_dict + ), "could not find data for pooler in the checkpoint" + self.decoder.load_state_dict(state_dict[self._decoder_key], strict=strict) diff --git a/megatron/model/module.py b/megatron/model/module.py index 08cf7cb5537..67b0e966540 100644 --- a/megatron/model/module.py +++ b/megatron/model/module.py @@ -17,8 +17,7 @@ def param_is_not_shared(param): - return not hasattr(param, 'shared') or not param.shared - + return not hasattr(param, "shared") or not param.shared class MegatronModule(torch.nn.Module): @@ -30,28 +29,29 @@ def __init__(self, config=None, share_embeddings_and_output_weights=True): self.config = config self.share_embeddings_and_output_weights = share_embeddings_and_output_weights - - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """Use this function to override the state dict for saving checkpoints.""" return self.state_dict(prefix=prefix, keep_vars=keep_vars) - def shared_embedding_or_output_weight(self): if self.pre_process: return self.language_model.embedding.word_embeddings.weight else: if not self.share_embeddings_and_output_weights: - raise Exception('shared_embedding_or_output_weight() called for last ' - 'stage, but share_embeddings_and_output_weights is false') + raise Exception( + "shared_embedding_or_output_weight() called for last " + "stage, but share_embeddings_and_output_weights is false" + ) return self.word_embeddings.weight - def initialize_word_embeddings(self): args = get_args() if not self.share_embeddings_and_output_weights: - raise Exception('initialize_word_embeddings() was called but ' - 'share_embeddings_and_output_weights is false') + raise Exception( + "initialize_word_embeddings() was called but " + "share_embeddings_and_output_weights is false" + ) # This function just initializes the word embeddings in the final stage # when we are using pipeline parallelism. Nothing to do if we aren't @@ -73,51 +73,62 @@ def initialize_word_embeddings(self): # update is the same on both stages. if mpu.is_pipeline_last_stage() and not self.pre_process: assert not mpu.is_pipeline_first_stage() - self._word_embeddings_for_head_key = 'word_embeddings_for_head' + self._word_embeddings_for_head_key = "word_embeddings_for_head" # set word_embeddings weights to 0 here, then copy first # stage's weights using all_reduce below. self.word_embeddings = tensor_parallel.VocabParallelEmbedding( - args.padded_vocab_size, self.config.hidden_size, - config=self.config, init_method=self.config.init_method) + args.padded_vocab_size, + self.config.hidden_size, + config=self.config, + init_method=self.config.init_method, + ) self.word_embeddings.weight.data.fill_(0) self.word_embeddings.weight.shared = True # Zero out initial weights for decoder embedding. # NOTE: We don't currently support T5 with the interleaved schedule. - if not mpu.is_pipeline_first_stage(ignore_virtual=True) and \ - self.pre_process: + if not mpu.is_pipeline_first_stage(ignore_virtual=True) and self.pre_process: self.language_model.embedding.zero_parameters() if not torch.distributed.is_initialized(): if not getattr(MegatronModule, "embedding_warning_printed", False): - print("WARNING! Distributed processes aren't initialized, so " - "word embeddings in the last layer are not initialized. " - "If you are just manipulating a model this is fine, but " - "this needs to be handled manually. If you are training " - "something is definitely wrong.") + print( + "WARNING! Distributed processes aren't initialized, so " + "word embeddings in the last layer are not initialized. " + "If you are just manipulating a model this is fine, but " + "this needs to be handled manually. If you are training " + "something is definitely wrong." + ) MegatronModule.embedding_warning_printed = True return # Ensure that first and last stages have the same initial parameter # values. if mpu.is_rank_in_embedding_group(): - torch.distributed.all_reduce(self.shared_embedding_or_output_weight().data, - group=mpu.get_embedding_group()) + torch.distributed.all_reduce( + self.shared_embedding_or_output_weight().data, + group=mpu.get_embedding_group(), + ) # Ensure that encoder(first stage) and decoder(split stage) position # embeddings have the same initial parameter values # NOTE: We don't currently support T5 with the interleaved schedule. - if mpu.is_rank_in_position_embedding_group() and \ - args.pipeline_model_parallel_split_rank is not None: + if ( + mpu.is_rank_in_position_embedding_group() + and args.pipeline_model_parallel_split_rank is not None + ): # TODO: Support tokentype embedding. self.language_model.embedding.cuda() position_embeddings = self.language_model.embedding.position_embeddings - torch.distributed.all_reduce(position_embeddings.weight.data, - group=mpu.get_position_embedding_group()) + torch.distributed.all_reduce( + position_embeddings.weight.data, + group=mpu.get_position_embedding_group(), + ) def universal_checkpoint_info(self): return {} + def conversion_helper(val, conversion): """Apply conversion to val. Recursively apply conversion if `val` #is a nested tuple/list structure.""" @@ -134,6 +145,7 @@ def fp32_to_float16(val, float16_convertor): global _FLOAT_TYPES if _FLOAT_TYPES is None: _FLOAT_TYPES = [get_accelerator().FloatTensor(0).dtype] + def half_conversion(val): val_typecheck = val if isinstance(val_typecheck, (Parameter, Variable)): @@ -141,6 +153,7 @@ def half_conversion(val): if val_typecheck.dtype in _FLOAT_TYPES: val = float16_convertor(val) return val + return conversion_helper(val, half_conversion) @@ -151,6 +164,7 @@ def float16_to_fp32(val): _HALF_TYPES = [get_accelerator().HalfTensor(0).dtype] if _BF16_TYPES is None: _BF16_TYPES = [get_accelerator().BFloat16Tensor(0).dtype] + def float_conversion(val): val_typecheck = val if isinstance(val_typecheck, (Parameter, Variable)): @@ -158,8 +172,8 @@ def float_conversion(val): if val_typecheck.dtype in _BF16_TYPES + _HALF_TYPES: val = val.float() return val - return conversion_helper(val, float_conversion) + return conversion_helper(val, float_conversion) class Float16Module(MegatronModule): @@ -168,23 +182,25 @@ def __init__(self, module, args): super(Float16Module, self).__init__() if args.fp16: - self.add_module('module', module.half()) + self.add_module("module", module.half()) + def float16_convertor(val): return val.half() + elif args.bf16: - self.add_module('module', module.bfloat16()) + self.add_module("module", module.bfloat16()) + def float16_convertor(val): return val.bfloat16() + else: - raise Exception('should not be here') + raise Exception("should not be here") self.float16_convertor = float16_convertor - def set_input_tensor(self, input_tensor): return self.module.set_input_tensor(input_tensor) - def forward(self, *inputs, **kwargs): if mpu.is_pipeline_first_stage(): inputs = fp32_to_float16(inputs, self.float16_convertor) @@ -193,15 +209,13 @@ def forward(self, *inputs, **kwargs): outputs = float16_to_fp32(outputs) return outputs - - def state_dict(self, prefix='', keep_vars=False): + def state_dict(self, prefix="", keep_vars=False): return self.module.state_dict(prefix=prefix, keep_vars=keep_vars) - - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): - return self.module.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) - + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): + return self.module.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) def load_state_dict(self, state_dict, strict=True): self.module.load_state_dict(state_dict, strict=strict) diff --git a/megatron/model/multiple_choice.py b/megatron/model/multiple_choice.py index 13e1e34a9ab..07b3af88a4d 100644 --- a/megatron/model/multiple_choice.py +++ b/megatron/model/multiple_choice.py @@ -16,11 +16,7 @@ class MultipleChoice(MegatronModule): - def __init__(self, - config, - num_tokentypes=2, - pre_process=True, - post_process=True): + def __init__(self, config, num_tokentypes=2, pre_process=True, post_process=True): super(MultipleChoice, self).__init__(share_embeddings_and_output_weights=False) args = get_args() @@ -33,15 +29,19 @@ def __init__(self, add_pooler=True, encoder_attn_mask_type=AttnMaskType.padding, pre_process=self.pre_process, - post_process=self.post_process) + post_process=self.post_process, + ) # Multi-choice head. if self.post_process: self.multichoice_dropout = torch.nn.Dropout(args.hidden_dropout) - self.multichoice_head = get_linear_layer(args.hidden_size, 1, - init_method, - gather_params_on_init=args.zero_stage == 3) - self._multichoice_head_key = 'multichoice_head' + self.multichoice_head = get_linear_layer( + args.hidden_size, + 1, + init_method, + gather_params_on_init=args.zero_stage == 3, + ) + self._multichoice_head_key = "multichoice_head" def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" @@ -72,7 +72,7 @@ def forward(self, model_input, attention_mask, tokentype_ids=None): input_ids, position_ids, extended_attention_mask, - tokentype_ids=tokentype_ids + tokentype_ids=tokentype_ids, ) if self.post_process: _, pooled_output = lm_output[0], lm_output[1] @@ -85,29 +85,35 @@ def forward(self, model_input, attention_mask, tokentype_ids=None): return multichoice_logits return lm_output - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.post_process: - state_dict_[self._multichoice_head_key] \ - = self.multichoice_head.state_dict(prefix=prefix, keep_vars=keep_vars) + state_dict_[self._multichoice_head_key] = self.multichoice_head.state_dict( + prefix=prefix, keep_vars=keep_vars + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) + state_dict[self._language_model_key], strict=strict + ) if self.post_process: if self._multichoice_head_key in state_dict: self.multichoice_head.load_state_dict( - state_dict[self._multichoice_head_key], strict=strict) + state_dict[self._multichoice_head_key], strict=strict + ) else: - print_rank_last('***WARNING*** could not find {} in the checkpoint, ' - 'initializing to random'.format( - self._multichoice_head_key)) + print_rank_last( + "***WARNING*** could not find {} in the checkpoint, " + "initializing to random".format(self._multichoice_head_key) + ) diff --git a/megatron/model/realm_model.py b/megatron/model/realm_model.py index 62d92b85586..5635f04a888 100644 --- a/megatron/model/realm_model.py +++ b/megatron/model/realm_model.py @@ -14,15 +14,19 @@ from megatron.model.bert_model import bert_extended_attention_mask, bert_position_ids from deepspeed.accelerator import get_accelerator + def general_ict_model_provider(only_query_model=False, only_block_model=False): """Build the model.""" args = get_args() - assert args.ict_head_size is not None, \ - "Need to specify --ict-head-size to provide an ICTBertModel" - assert mpu.get_tensor_model_parallel_world_size() == 1 and mpu.get_pipeline_model_parallel_world_size() == 1, \ - "Model parallel size > 1 not supported for ICT" + assert ( + args.ict_head_size is not None + ), "Need to specify --ict-head-size to provide an ICTBertModel" + assert ( + mpu.get_tensor_model_parallel_world_size() == 1 + and mpu.get_pipeline_model_parallel_world_size() == 1 + ), "Model parallel size > 1 not supported for ICT" - print_rank_0('building ICTBertModel...') + print_rank_0("building ICTBertModel...") # simpler to just keep using 2 tokentypes since the LM we initialize with has 2 tokentypes model = ICTBertModel( @@ -30,24 +34,28 @@ def general_ict_model_provider(only_query_model=False, only_block_model=False): num_tokentypes=2, parallel_output=True, only_query_model=only_query_model, - only_block_model=only_block_model) + only_block_model=only_block_model, + ) return model class ICTBertModel(MegatronModule): """Bert-based module for Inverse Cloze task.""" - def __init__(self, - ict_head_size, - num_tokentypes=1, - parallel_output=True, - only_query_model=False, - only_block_model=False): + + def __init__( + self, + ict_head_size, + num_tokentypes=1, + parallel_output=True, + only_query_model=False, + only_block_model=False, + ): super(ICTBertModel, self).__init__() bert_kwargs = dict( ict_head_size=ict_head_size, num_tokentypes=num_tokentypes, - parallel_output=parallel_output + parallel_output=parallel_output, ) assert not (only_block_model and only_query_model) self.use_block_model = not only_query_model @@ -56,14 +64,16 @@ def __init__(self, if self.use_query_model: # this model embeds (pseudo-)queries - Embed_input in the paper self.query_model = IREncoderBertModel(**bert_kwargs) - self._query_key = 'question_model' + self._query_key = "question_model" if self.use_block_model: # this model embeds evidence blocks - Embed_doc in the paper self.block_model = IREncoderBertModel(**bert_kwargs) - self._block_key = 'context_model' + self._block_key = "context_model" - def forward(self, query_tokens, query_attention_mask, block_tokens, block_attention_mask): + def forward( + self, query_tokens, query_attention_mask, block_tokens, block_attention_mask + ): """Run a forward pass for each of the models and return the respective embeddings.""" query_logits = self.embed_query(query_tokens, query_attention_mask) block_logits = self.embed_block(block_tokens, block_attention_mask) @@ -73,7 +83,9 @@ def embed_query(self, query_tokens, query_attention_mask): """Embed a batch of tokens using the query model""" if self.use_query_model: query_types = get_accelerator().LongTensor(*query_tokens.shape).fill_(0) - query_ict_logits, _ = self.query_model.forward(query_tokens, query_attention_mask, query_types) + query_ict_logits, _ = self.query_model.forward( + query_tokens, query_attention_mask, query_types + ) return query_ict_logits else: raise ValueError("Cannot embed query without query model.") @@ -82,23 +94,29 @@ def embed_block(self, block_tokens, block_attention_mask): """Embed a batch of tokens using the block model""" if self.use_block_model: block_types = get_accelerator().LongTensor(*block_tokens.shape).fill_(0) - block_ict_logits, _ = self.block_model.forward(block_tokens, block_attention_mask, block_types) + block_ict_logits, _ = self.block_model.forward( + block_tokens, block_attention_mask, block_types + ) return block_ict_logits else: raise ValueError("Cannot embed block without block model.") - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """Save dict with state dicts of each of the models.""" state_dict_ = {} if self.use_query_model: - state_dict_[self._query_key] \ - = self.query_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + state_dict_[self._query_key] = ( + self.query_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.use_block_model: - state_dict_[self._block_key] \ - = self.block_model.state_dict_for_save_checkpoint( - prefix=prefix, keep_vars=keep_vars) + state_dict_[self._block_key] = ( + self.block_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) return state_dict_ @@ -106,13 +124,11 @@ def load_state_dict(self, state_dict, strict=True): """Load the state dicts of each of the models""" if self.use_query_model: print("Loading ICT query model", flush=True) - self.query_model.load_state_dict( - state_dict[self._query_key], strict=strict) + self.query_model.load_state_dict(state_dict[self._query_key], strict=strict) if self.use_block_model: print("Loading ICT block model", flush=True) - self.block_model.load_state_dict( - state_dict[self._block_key], strict=strict) + self.block_model.load_state_dict(state_dict[self._block_key], strict=strict) def init_state_dict_from_bert(self): """Initialize the state from a pretrained BERT model on iteration zero of ICT pretraining""" @@ -120,32 +136,38 @@ def init_state_dict_from_bert(self): tracker_filename = get_checkpoint_tracker_filename(args.bert_load) if not os.path.isfile(tracker_filename): raise FileNotFoundError("Could not find BERT load for ICT") - with open(tracker_filename, 'r') as f: + with open(tracker_filename, "r") as f: iteration = int(f.read().strip()) assert iteration > 0 checkpoint_name = get_checkpoint_name(args.bert_load, iteration, False) if mpu.get_data_parallel_rank() == 0: - print('global rank {} is loading checkpoint {}'.format( - torch.distributed.get_rank(), checkpoint_name)) + print( + "global rank {} is loading checkpoint {}".format( + torch.distributed.get_rank(), checkpoint_name + ) + ) try: - state_dict = torch.load(checkpoint_name, map_location='cpu') + state_dict = torch.load(checkpoint_name, map_location="cpu") except BaseException: raise ValueError("Could not load checkpoint") # load the LM state dict into each model - model_dict = state_dict['model']['language_model'] + model_dict = state_dict["model"]["language_model"] self.query_model.language_model.load_state_dict(model_dict) self.block_model.language_model.load_state_dict(model_dict) # give each model the same ict_head to begin with as well - query_ict_head_state_dict = self.state_dict_for_save_checkpoint()[self._query_key]['ict_head'] + query_ict_head_state_dict = self.state_dict_for_save_checkpoint()[ + self._query_key + ]["ict_head"] self.block_model.ict_head.load_state_dict(query_ict_head_state_dict) class IREncoderBertModel(MegatronModule): """BERT-based encoder for queries or blocks used for learned information retrieval.""" + def __init__(self, ict_head_size, num_tokentypes=2, parallel_output=True): super(IREncoderBertModel, self).__init__() args = get_args() @@ -153,52 +175,61 @@ def __init__(self, ict_head_size, num_tokentypes=2, parallel_output=True): self.ict_head_size = ict_head_size self.parallel_output = parallel_output init_method = init_method_normal(args.init_method_std) - scaled_init_method = scaled_init_method_normal(args.init_method_std, - args.num_layers) + scaled_init_method = scaled_init_method_normal( + args.init_method_std, args.num_layers + ) self.language_model, self._language_model_key = get_language_model( num_tokentypes=num_tokentypes, add_pooler=True, encoder_attn_mask_type=AttnMaskType.padding, init_method=init_method, - scaled_init_method=scaled_init_method) + scaled_init_method=scaled_init_method, + ) - self.ict_head = get_linear_layer(args.hidden_size, ict_head_size, init_method, gather_params_on_init=args.zero_stage == 3) - self._ict_head_key = 'ict_head' + self.ict_head = get_linear_layer( + args.hidden_size, + ict_head_size, + init_method, + gather_params_on_init=args.zero_stage == 3, + ) + self._ict_head_key = "ict_head" def forward(self, input_ids, attention_mask, tokentype_ids=None): extended_attention_mask = bert_extended_attention_mask( - attention_mask, next(self.language_model.parameters()).dtype) + attention_mask, next(self.language_model.parameters()).dtype + ) position_ids = bert_position_ids(input_ids) lm_output, pooled_output = self.language_model( input_ids, position_ids, extended_attention_mask, - tokentype_ids=tokentype_ids) + tokentype_ids=tokentype_ids, + ) # Output. ict_logits = self.ict_head(pooled_output) return ict_logits, None - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) - state_dict_[self._ict_head_key] \ - = self.ict_head.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) + state_dict_[self._ict_head_key] = self.ict_head.state_dict( + prefix=prefix, keep_vars=keep_vars + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) - self.ict_head.load_state_dict( - state_dict[self._ict_head_key], strict=strict) - - + state_dict[self._language_model_key], strict=strict + ) + self.ict_head.load_state_dict(state_dict[self._ict_head_key], strict=strict) diff --git a/megatron/model/rmsnorm.py b/megatron/model/rmsnorm.py index 68b792c83d8..33ad1d74827 100644 --- a/megatron/model/rmsnorm.py +++ b/megatron/model/rmsnorm.py @@ -10,9 +10,8 @@ # Taken from facebookresearch/llama class RMSNorm(torch.nn.Module): - def __init__( - self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False - ): + + def __init__(self, dim: int, eps: float = 1e-6, sequence_parallel: bool = False): super().__init__() self.eps = eps init_device = None diff --git a/megatron/model/rotary_pos_embedding.py b/megatron/model/rotary_pos_embedding.py index 0a7acb9efc2..bc53a49b1ca 100644 --- a/megatron/model/rotary_pos_embedding.py +++ b/megatron/model/rotary_pos_embedding.py @@ -10,29 +10,32 @@ from torch import einsum, nn -__all__ = ['RotaryEmbedding', 'apply_rotary_pos_emb'] +__all__ = ["RotaryEmbedding", "apply_rotary_pos_emb"] # sin, cos tensors cached for all devices cos_cached = None sin_cached = None + class RotaryEmbedding(nn.Module): + def __init__(self, dim, theta=10000): super().__init__() inv_freq = 1.0 / (theta ** (torch.arange(0, dim, 2).float() / dim)) - self.register_buffer('inv_freq', inv_freq) - if importlib.util.find_spec('einops') is None: + self.register_buffer("inv_freq", inv_freq) + if importlib.util.find_spec("einops") is None: raise RuntimeError("einops is required for Rotary Embedding") def forward(self, max_seq_len, offset=0): seq = torch.arange(max_seq_len, device=self.inv_freq.device) + offset - freqs = einsum('i , j -> i j', seq.type_as(self.inv_freq), self.inv_freq) + freqs = einsum("i , j -> i j", seq.type_as(self.inv_freq), self.inv_freq) # first part even vector components, second part odd vector components, # 2 * dim in dimension size emb = torch.cat((freqs, freqs), dim=-1) # emb [seq_length, .., dim] from einops import rearrange - return rearrange(emb, 'n d -> n 1 1 d') + + return rearrange(emb, "n d -> n 1 1 d") def _rotate_half(x): @@ -40,7 +43,8 @@ def _rotate_half(x): change sign so the last dimension becomes [-odd, +even] """ from einops import rearrange - x = rearrange(x, '... (j d) -> ... j d', j=2) + + x = rearrange(x, "... (j d) -> ... j d", j=2) x1, x2 = x.unbind(dim=-2) return torch.cat((-x2, x1), dim=-1) @@ -59,12 +63,16 @@ def apply_rotary_pos_emb(t, freqs): global cos_cached, sin_cached if cos_cached is None or sin_cached is None or t.shape[0] != cos_cached.shape[0]: - freqs_ = freqs[:t.shape[0]] + freqs_ = freqs[: t.shape[0]] cos_cached = freqs_.cos().to(t.dtype) sin_cached = freqs_.sin().to(t.dtype) # first part is cosine component # second part is sine component, need to change signs with _rotate_half method + # try: t = (t * cos_cached) + (_rotate_half(t) * sin_cached) + # except Exception: + # import ezpz + # ezpz.breakpoint(0) if t_pass is None: return t return torch.cat((t, t_pass), dim=-1) diff --git a/megatron/model/t5_model.py b/megatron/model/t5_model.py index 5a4c2c32c93..85336fee260 100644 --- a/megatron/model/t5_model.py +++ b/megatron/model/t5_model.py @@ -9,10 +9,7 @@ from megatron.model.enums import AttnMaskType from megatron.model.language_model import parallel_lm_logits, get_language_model from megatron.model import LayerNorm -from megatron.model.utils import ( - openai_gelu, - get_linear_layer -) +from megatron.model.utils import (openai_gelu, get_linear_layer) from .module import MegatronModule @@ -29,8 +26,7 @@ def attn_mask_postprocess(attn_mask): def t5_position_ids(token_ids): # Create position ids seq_length = token_ids.size(1) - position_ids = torch.arange(seq_length, dtype=torch.long, - device=token_ids.device) + position_ids = torch.arange(seq_length, dtype=torch.long, device=token_ids.device) position_ids = position_ids.unsqueeze(0).expand_as(token_ids) return position_ids @@ -54,25 +50,26 @@ def __init__(self, mpu_vocab_size, parallel_output): self.parallel_output = parallel_output def forward(self, hidden_states, word_embeddings_weight): - output = parallel_lm_logits(hidden_states, - word_embeddings_weight, - self.parallel_output, - bias=self.bias) + output = parallel_lm_logits( + hidden_states, word_embeddings_weight, self.parallel_output, bias=self.bias + ) return output class T5Model(MegatronModule): """T5 Language model.""" - def __init__(self, - config, - num_tokentypes=0, - parallel_output=True, - pre_process=True, - post_process=True, - add_encoder=True, - add_decoder=True, - return_moe_loss=False): + def __init__( + self, + config, + num_tokentypes=0, + parallel_output=True, + pre_process=True, + post_process=True, + add_encoder=True, + add_decoder=True, + return_moe_loss=False, + ): super().__init__(config=config) args = get_args() @@ -93,95 +90,122 @@ def __init__(self, encoder_attn_mask_type=AttnMaskType.padding, pre_process=self.pre_process, post_process=self.post_process, - num_experts=args.num_experts,) + num_experts=args.num_experts, + ) self.initialize_word_embeddings() if self.post_process and self.add_decoder: self.lm_head = T5LMHead( - self.shared_embedding_or_output_weight().size(0), - parallel_output) - self._lm_head_key = 'lm_head' + self.shared_embedding_or_output_weight().size(0), parallel_output + ) + self._lm_head_key = "lm_head" def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" self.language_model.set_input_tensor(input_tensor) - def forward(self, encoder_input_ids, decoder_input_ids, encoder_attn_mask, - decoder_attn_mask, encoder_decoder_attn_mask, - tokentype_ids=None, lm_labels=None, enc_hidden_states=None): + def forward( + self, + encoder_input_ids, + decoder_input_ids, + encoder_attn_mask, + decoder_attn_mask, + encoder_decoder_attn_mask, + tokentype_ids=None, + lm_labels=None, + enc_hidden_states=None, + ): # Converting the attention masks to proper parameter settings - encoder_attn_mask, decoder_attn_mask, encoder_decoder_attn_mask = t5_extended_attention_mask( - [encoder_attn_mask, decoder_attn_mask, encoder_decoder_attn_mask]) + encoder_attn_mask, decoder_attn_mask, encoder_decoder_attn_mask = ( + t5_extended_attention_mask( + [encoder_attn_mask, decoder_attn_mask, encoder_decoder_attn_mask] + ) + ) encoder_position_ids = t5_position_ids(encoder_input_ids) decoder_position_ids = t5_position_ids(decoder_input_ids) - lm_output = self.language_model(encoder_input_ids, - encoder_position_ids, - encoder_attn_mask, - decoder_input_ids, - decoder_position_ids, - decoder_attn_mask, - encoder_decoder_attn_mask, - tokentype_ids=tokentype_ids, - enc_hidden_states=enc_hidden_states) + lm_output = self.language_model( + encoder_input_ids, + encoder_position_ids, + encoder_attn_mask, + decoder_input_ids, + decoder_position_ids, + decoder_attn_mask, + encoder_decoder_attn_mask, + tokentype_ids=tokentype_ids, + enc_hidden_states=enc_hidden_states, + ) if self.post_process and self.add_decoder: decoder_output, encoder_output, dec_moe_losses, enc_moe_losses = lm_output # Output. [s, b, h] - lm_logits = self.lm_head(decoder_output, - self.shared_embedding_or_output_weight()) + lm_logits = self.lm_head( + decoder_output, self.shared_embedding_or_output_weight() + ) if lm_labels is None: # [s b h] => [b s h] - return lm_logits.transpose(0,1).contiguous() + return lm_logits.transpose(0, 1).contiguous() else: # [b s] => [s b] - lm_labels = lm_labels.transpose(0,1).contiguous() + lm_labels = lm_labels.transpose(0, 1).contiguous() if self.fp16_lm_cross_entropy: assert lm_logits.dtype == torch.half - lm_loss = tensor_parallel.vocab_parallel_cross_entropy(lm_logits, lm_labels) + lm_loss = tensor_parallel.vocab_parallel_cross_entropy( + lm_logits, lm_labels + ) else: - lm_loss = tensor_parallel.vocab_parallel_cross_entropy(lm_logits.float(), - lm_labels) + lm_loss = tensor_parallel.vocab_parallel_cross_entropy( + lm_logits.float(), lm_labels + ) # [s b] => [b s] - lm_loss = lm_loss.transpose(0,1).contiguous() - return lm_loss, dec_moe_losses, enc_moe_losses if self.return_moe_loss else lm_loss + lm_loss = lm_loss.transpose(0, 1).contiguous() + return ( + lm_loss, + dec_moe_losses, + enc_moe_losses if self.return_moe_loss else lm_loss, + ) elif self.add_decoder and not self.add_encoder: - decoder_output, _, decoder_moe_losses, _= lm_output + decoder_output, _, decoder_moe_losses, _ = lm_output return decoder_output, decoder_moe_losses return lm_output - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load when model is combined with other heads, add an extra key.""" state_dict_ = {} - state_dict_[self._language_model_key] \ - = self.language_model.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._language_model_key] = ( + self.language_model.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) if self.post_process and self.add_decoder: - state_dict_[self._lm_head_key] \ - = self.lm_head.state_dict_for_save_checkpoint(prefix=prefix, - keep_vars=keep_vars) - # Save word_embeddings. + state_dict_[self._lm_head_key] = ( + self.lm_head.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + ) + # Save word_embeddings. if self.post_process and not self.pre_process and self.add_decoder: - state_dict_[self._word_embeddings_for_head_key] \ - = self.word_embeddings.state_dict(prefix=prefix, - keep_vars=keep_vars) + state_dict_[self._word_embeddings_for_head_key] = ( + self.word_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ def load_state_dict(self, state_dict, strict=True): """Customized load.""" self.language_model.load_state_dict( - state_dict[self._language_model_key], strict=strict) + state_dict[self._language_model_key], strict=strict + ) if self.post_process and self.add_decoder: - self.lm_head.load_state_dict(state_dict[self._lm_head_key], - strict=strict) + self.lm_head.load_state_dict(state_dict[self._lm_head_key], strict=strict) # Load word embeddings. if self.post_process and not self.pre_process and self.add_decoder: self.word_embeddings.load_state_dict( - state_dict[self._word_embeddings_for_head_key], strict=strict) + state_dict[self._word_embeddings_for_head_key], strict=strict + ) diff --git a/megatron/model/utils.py b/megatron/model/utils.py index 6c94921c958..59d4e2b7b8f 100644 --- a/megatron/model/utils.py +++ b/megatron/model/utils.py @@ -10,8 +10,10 @@ from deepspeed.runtime.zero import GatheredParameters + def init_method_normal(sigma): """Init method based on N(0, sigma).""" + def init_(tensor): return torch.nn.init.normal_(tensor, mean=0.0, std=sigma) @@ -31,7 +33,7 @@ def init_(tensor): def gather_and_init(param, init_method): with GatheredParameters(param, modifier_rank=0): init_method(param) - + def attention_mask_func(attention_scores, attention_mask): args = get_args() @@ -40,7 +42,9 @@ def attention_mask_func(attention_scores, attention_mask): actual_seqlen = attention_scores.size()[2] if actual_seqlen != attention_mask_.size()[2]: # attention_mask has size [1, 1, seqlen, seqlen] - attention_mask_ = attention_mask_[:, :, :actual_seqlen, :actual_seqlen].contiguous() + attention_mask_ = attention_mask_[ + :, :, :actual_seqlen, :actual_seqlen + ].contiguous() attention_scores.masked_fill_(attention_mask_, -10000.0) else: attention_scores.masked_fill_(attention_mask, -10000.0) @@ -51,22 +55,38 @@ def get_linear_layer(rows, columns, init_method, gather_params_on_init=False): """Simple linear layer with weight initialization.""" layer = torch.nn.Linear(rows, columns) if get_args().perform_initialization: - with GatheredParameters(layer.weight, modifier_rank=0, enabled=gather_params_on_init): + with GatheredParameters( + layer.weight, modifier_rank=0, enabled=gather_params_on_init + ): init_method(layer.weight) with torch.no_grad(): - with GatheredParameters(layer.bias, modifier_rank=0, enabled=gather_params_on_init): + with GatheredParameters( + layer.bias, modifier_rank=0, enabled=gather_params_on_init + ): layer.bias.zero_() return layer + @torch.jit.script def gelu_impl(x): """OpenAI's gelu implementation.""" - return 0.5 * x * (1.0 + torch.tanh(0.7978845608028654 * x * - (1.0 + 0.044715 * x * x))) + return ( + 0.5 * x * (1.0 + torch.tanh(0.7978845608028654 * x * (1.0 + 0.044715 * x * x))) + ) + + def openai_gelu(x): return gelu_impl(x) -#This is actually Python equivalent of torch.nn.functional.gelu(), also with type hints for ONNX exporter + +# This is actually Python equivalent of torch.nn.functional.gelu(), also with type hints for ONNX exporter @torch.jit.script def erf_gelu(x): - return x * 0.5 * (torch.erf(x / 1.41421).to(dtype=x.dtype)+torch.ones_like(x).to(dtype=x.dtype)) + return ( + x + * 0.5 + * ( + torch.erf(x / 1.41421).to(dtype=x.dtype) + + torch.ones_like(x).to(dtype=x.dtype) + ) + ) diff --git a/megatron/model/vision/classification.py b/megatron/model/vision/classification.py index 2be1951eead..ee01b1f6c1e 100644 --- a/megatron/model/vision/classification.py +++ b/megatron/model/vision/classification.py @@ -10,11 +10,13 @@ from megatron.model.vision.mit_backbone import mit_b3_avg from megatron.model.module import MegatronModule + class VitClassificationModel(MegatronModule): """Vision Transformer Model.""" - def __init__(self, config, num_classes, finetune=False, - pre_process=True, post_process=True): + def __init__( + self, config, num_classes, finetune=False, pre_process=True, post_process=True + ): super(VitClassificationModel, self).__init__() args = get_args() @@ -27,9 +29,9 @@ def __init__(self, config, num_classes, finetune=False, config=config, pre_process=self.pre_process, post_process=self.post_process, - single_token_output=True + single_token_output=True, ) - + if self.post_process: if not self.finetune: self.head = VitMlpHead(self.hidden_size, self.num_classes) @@ -38,7 +40,7 @@ def __init__(self, config, num_classes, finetune=False, self.hidden_size, self.num_classes, torch.nn.init.zeros_, - gather_params_on_init=args.zero_stage == 3 + gather_params_on_init=args.zero_stage == 3, ) def set_input_tensor(self, input_tensor): @@ -57,8 +59,7 @@ def forward(self, input): class MitClassificationModel(MegatronModule): """Mix vision Transformer Model.""" - def __init__(self, num_classes, - pre_process=True, post_process=True): + def __init__(self, num_classes, pre_process=True, post_process=True): super(MitClassificationModel, self).__init__() args = get_args() @@ -71,7 +72,7 @@ def __init__(self, num_classes, def _init_weights(self, m): if isinstance(m, torch.nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, torch.nn.Linear) and m.bias is not None: torch.nn.init.constant_(m.bias, 0) diff --git a/megatron/model/vision/dino.py b/megatron/model/vision/dino.py index 1c577d2e199..7311f4000d0 100644 --- a/megatron/model/vision/dino.py +++ b/megatron/model/vision/dino.py @@ -21,9 +21,18 @@ class DINOLoss(torch.nn.Module): - def __init__(self, out_dim, ncrops, warmup_teacher_temp, teacher_temp, - warmup_teacher_temp_epochs, nepochs, student_temp=0.1, - center_momentum=0.9): + + def __init__( + self, + out_dim, + ncrops, + warmup_teacher_temp, + teacher_temp, + warmup_teacher_temp_epochs, + nepochs, + student_temp=0.1, + center_momentum=0.9, + ): super().__init__() self.student_temp = student_temp self.center_momentum = center_momentum @@ -31,11 +40,14 @@ def __init__(self, out_dim, ncrops, warmup_teacher_temp, teacher_temp, self.register_buffer("center", torch.zeros(1, out_dim)) # we apply a warm up for the teacher temperature because # a too high temperature makes the training instable at the beginning - self.teacher_temp_schedule = np.concatenate(( - np.linspace(warmup_teacher_temp, - teacher_temp, warmup_teacher_temp_epochs), - np.ones(nepochs - warmup_teacher_temp_epochs) * teacher_temp - )) + self.teacher_temp_schedule = np.concatenate( + ( + np.linspace( + warmup_teacher_temp, teacher_temp, warmup_teacher_temp_epochs + ), + np.ones(nepochs - warmup_teacher_temp_epochs) * teacher_temp, + ) + ) self.teacher_temp = teacher_temp def forward(self, student_output, teacher_output, iteration): @@ -76,10 +88,16 @@ def update_center(self, teacher_output): """ batch_center = torch.sum(teacher_output, dim=0, keepdim=True) torch.distributed.all_reduce(batch_center) - batch_center = batch_center / (len(teacher_output) * torch.distributed.get_world_size()) - self.center = self.center * self.center_momentum + batch_center * (1 - self.center_momentum) + batch_center = batch_center / ( + len(teacher_output) * torch.distributed.get_world_size() + ) + self.center = self.center * self.center_momentum + batch_center * ( + 1 - self.center_momentum + ) + class DINOHead(torch.nn.Module): + def __init__(self, in_dim, out_dim, norm_last_layer=True, nlayers=3): super().__init__() args = get_args() @@ -97,14 +115,16 @@ def __init__(self, in_dim, out_dim, norm_last_layer=True, nlayers=3): layers.append(torch.nn.Linear(hidden_dim, bottleneck_dim)) self.mlp = torch.nn.Sequential(*layers) self.apply(self._init_weights) - self.last_layer = torch.nn.utils.weight_norm(torch.nn.Linear(bottleneck_dim, out_dim, bias=False)) + self.last_layer = torch.nn.utils.weight_norm( + torch.nn.Linear(bottleneck_dim, out_dim, bias=False) + ) self.last_layer.weight_g.data.fill_(1) if norm_last_layer: self.last_layer.weight_g.requires_grad = False def _init_weights(self, m): if isinstance(m, torch.nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, torch.nn.Linear) and m.bias is not None: torch.nn.init.constant_(m.bias, 0) @@ -116,7 +136,6 @@ def forward(self, x): class MultiCropWrapper(MegatronModule): - """ Perform forward pass separately on each resolution input. The inputs corresponding to a single resolution are clubbed and single @@ -125,10 +144,11 @@ class MultiCropWrapper(MegatronModule): concatenate all the output features and run the head forward on these concatenated features. """ + def __init__(self, backbone, head): super(MultiCropWrapper, self).__init__() # disable layers dedicated to ImageNet labels classification - #backbone.fc, backbone.head = torch.nn.Identity(), torch.nn.Identity() + # backbone.fc, backbone.head = torch.nn.Identity(), torch.nn.Identity() self.backbone = backbone self.head = head @@ -136,14 +156,17 @@ def forward(self, x): # convert to list if not isinstance(x, list): x = [x] - idx_crops = torch.cumsum(torch.unique_consecutive( - torch.tensor([inp.shape[-1] for inp in x]), - return_counts=True, - )[1], 0) + idx_crops = torch.cumsum( + torch.unique_consecutive( + torch.tensor([inp.shape[-1] for inp in x]), + return_counts=True, + )[1], + 0, + ) start_idx = 0 for end_idx in idx_crops: - _out = self.backbone(torch.cat(x[start_idx: end_idx])) + _out = self.backbone(torch.cat(x[start_idx:end_idx])) if start_idx == 0: output = _out else: @@ -156,17 +179,18 @@ def forward(self, x): return output -def cosine_scheduler(base_value, final_value, epochs, niter_per_ep, - warmup_epochs=0, start_warmup_value=0): +def cosine_scheduler( + base_value, final_value, epochs, niter_per_ep, warmup_epochs=0, start_warmup_value=0 +): warmup_schedule = np.array([]) warmup_iters = warmup_epochs * niter_per_ep if warmup_epochs > 0: - warmup_schedule = \ - np.linspace(start_warmup_value, base_value, warmup_iters) + warmup_schedule = np.linspace(start_warmup_value, base_value, warmup_iters) iters = np.arange(epochs * niter_per_ep - warmup_iters) - schedule = final_value + 0.5 * (base_value - final_value) \ - * (1 + np.cos(np.pi * iters / len(iters))) + schedule = final_value + 0.5 * (base_value - final_value) * ( + 1 + np.cos(np.pi * iters / len(iters)) + ) schedule = np.concatenate((warmup_schedule, schedule)) assert len(schedule) == epochs * niter_per_ep @@ -176,47 +200,55 @@ def cosine_scheduler(base_value, final_value, epochs, niter_per_ep, def get_student_backbone_and_num_features(config, pre_process=True, post_process=True): args = get_args() - if args.vision_backbone_type == 'vit': - student = VitBackbone(config, - pre_process=pre_process, - post_process=post_process, - drop_path_rate=0.1, - single_token_output=True) + if args.vision_backbone_type == "vit": + student = VitBackbone( + config, + pre_process=pre_process, + post_process=post_process, + drop_path_rate=0.1, + single_token_output=True, + ) num_features = args.hidden_size - elif args.vision_backbone_type == 'mit': + elif args.vision_backbone_type == "mit": student = mit_b5_avg(drop_path_rate=0.1) num_features = 512 - elif args.vision_backbone_type == 'swin': + elif args.vision_backbone_type == "swin": student = get_swin() num_features = student.num_features else: - raise Exception('{} vision backbone is not supported.'.format( - args.vision_backbone_type)) - + raise Exception( + "{} vision backbone is not supported.".format(args.vision_backbone_type) + ) + return student, num_features + def get_teacher_backbone_and_num_features(config, pre_process=True, post_process=True): args = get_args() - if args.vision_backbone_type == 'vit': - teacher = VitBackbone(config, - pre_process=pre_process, - post_process=post_process, - single_token_output=True) + if args.vision_backbone_type == "vit": + teacher = VitBackbone( + config, + pre_process=pre_process, + post_process=post_process, + single_token_output=True, + ) num_features = args.hidden_size - elif args.vision_backbone_type == 'mit': + elif args.vision_backbone_type == "mit": teacher = mit_b5_avg(drop_path_rate=0.0) num_features = 512 - elif args.vision_backbone_type == 'swin': + elif args.vision_backbone_type == "swin": teacher = get_swin(is_teacher=True) num_features = teacher.num_features else: - raise Exception('{} vision backbone is not supported.'.format( - args.vision_backbone_type)) + raise Exception( + "{} vision backbone is not supported.".format(args.vision_backbone_type) + ) return teacher, num_features class DINOPretrainModel(MegatronModule): + def __init__(self, config, pre_process=True, post_process=True): super(DINOPretrainModel, self).__init__() args = get_args() @@ -235,26 +267,29 @@ def __init__(self, config, pre_process=True, post_process=True): self.post_process = post_process self.momentum_teacher = 0.996 - student_backbone, num_features = \ - get_student_backbone_and_num_features(config, pre_process, post_process) + student_backbone, num_features = get_student_backbone_and_num_features( + config, pre_process, post_process + ) self.student = MultiCropWrapper( student_backbone, - DINOHead(num_features, self.out_dim, - norm_last_layer=args.dino_norm_last_layer) + DINOHead( + num_features, self.out_dim, norm_last_layer=args.dino_norm_last_layer + ), ) self.momentum_schedule = cosine_scheduler( - self.momentum_teacher, 1, + self.momentum_teacher, + 1, args.train_iters // args.iter_per_epoch, - args.iter_per_epoch + args.iter_per_epoch, ) - teacher_backbone, num_features = \ - get_teacher_backbone_and_num_features(config, pre_process, post_process) + teacher_backbone, num_features = get_teacher_backbone_and_num_features( + config, pre_process, post_process + ) self.teacher = MultiCropWrapper( - teacher_backbone, - DINOHead(num_features, self.out_dim) + teacher_backbone, DINOHead(num_features, self.out_dim) ) self.teacher.load_state_dict(self.student.state_dict()) @@ -285,6 +320,7 @@ def cancel_gradients_last_layer(self, iteration): def update_momentum(self, iteration): with torch.no_grad(): m = self.momentum_schedule[iteration] - for param_q, param_k in zip(self.student.parameters(), self.teacher.parameters()): + for param_q, param_k in zip( + self.student.parameters(), self.teacher.parameters() + ): param_k.data.mul_(m).add_((1 - m) * param_q.detach().data) - diff --git a/megatron/model/vision/esvit_swin_backbone.py b/megatron/model/vision/esvit_swin_backbone.py index 70aee3db429..d5d167ce86f 100644 --- a/megatron/model/vision/esvit_swin_backbone.py +++ b/megatron/model/vision/esvit_swin_backbone.py @@ -23,8 +23,15 @@ class Mlp(nn.Module): - def __init__(self, in_features, hidden_features=None, - out_features=None, act_layer=nn.GELU, drop=0.): + + def __init__( + self, + in_features, + hidden_features=None, + out_features=None, + act_layer=nn.GELU, + drop=0.0, + ): super(Mlp, self).__init__() out_features = out_features or in_features hidden_features = hidden_features or in_features @@ -52,7 +59,9 @@ def window_partition(x, window_size): """ B, H, W, C = x.shape x = x.view(B, H // window_size, window_size, W // window_size, window_size, C) - windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + windows = ( + x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + ) return windows @@ -67,7 +76,9 @@ def window_reverse(windows, window_size, H, W): x: (B, H, W, C) """ B = int(windows.shape[0] / (H * W / window_size / window_size)) - x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) + x = windows.view( + B, H // window_size, W // window_size, window_size, window_size, -1 + ) x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1) return x @@ -85,26 +96,40 @@ class WindowAttention(nn.Module): proj_drop (float, optional): Dropout ratio of output. Default: 0.0 """ - def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, attn_drop=0., proj_drop=0.): + def __init__( + self, + dim, + window_size, + num_heads, + qkv_bias=True, + qk_scale=None, + attn_drop=0.0, + proj_drop=0.0, + ): super(WindowAttention, self).__init__() self.dim = dim self.window_size = window_size # Wh, Ww self.num_heads = num_heads head_dim = dim // num_heads - self.scale = qk_scale or head_dim ** -0.5 + self.scale = qk_scale or head_dim**-0.5 # define a parameter table of relative position bias self.relative_position_bias_table = nn.Parameter( - torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads)) # 2*Wh-1 * 2*Ww-1, nH + torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads) + ) # 2*Wh-1 * 2*Ww-1, nH # get pair-wise relative position index for each token inside the window coords_h = torch.arange(self.window_size[0]) coords_w = torch.arange(self.window_size[1]) coords = torch.stack(torch.meshgrid([coords_h, coords_w])) # 2, Wh, Ww coords_flatten = torch.flatten(coords, 1) # 2 Wh*Ww - relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :] # 2, Wh*Ww, Wh*Ww - relative_coords = relative_coords.permute(1, 2, 0).contiguous() # Wh*Ww, Wh*Ww, 2 + relative_coords = ( + coords_flatten[:, :, None] - coords_flatten[:, None, :] + ) # 2, Wh*Ww, Wh*Ww + relative_coords = relative_coords.permute( + 1, 2, 0 + ).contiguous() # Wh*Ww, Wh*Ww, 2 relative_coords[:, :, 0] += self.window_size[0] - 1 # shift to start from 0 relative_coords[:, :, 1] += self.window_size[1] - 1 relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1 @@ -116,7 +141,7 @@ def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, at self.proj = nn.Linear(dim, dim) self.proj_drop = nn.Dropout(proj_drop) - trunc_normal_(self.relative_position_bias_table, std=.02) + trunc_normal_(self.relative_position_bias_table, std=0.02) self.softmax = nn.Softmax(dim=-1) def forward(self, x, mask=None): @@ -126,20 +151,37 @@ def forward(self, x, mask=None): mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None """ B_, N, C = x.shape - qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) - q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple) + qkv = ( + self.qkv(x) + .reshape(B_, N, 3, self.num_heads, C // self.num_heads) + .permute(2, 0, 3, 1, 4) + ) + q, k, v = ( + qkv[0], + qkv[1], + qkv[2], + ) # make torchscript happy (cannot use tensor as tuple) q = q * self.scale - attn = (q @ k.transpose(-2, -1)) - - relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view( - self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1) # Wh*Ww,Wh*Ww,nH - relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww + attn = q @ k.transpose(-2, -1) + + relative_position_bias = self.relative_position_bias_table[ + self.relative_position_index.view(-1) + ].view( + self.window_size[0] * self.window_size[1], + self.window_size[0] * self.window_size[1], + -1, + ) # Wh*Ww,Wh*Ww,nH + relative_position_bias = relative_position_bias.permute( + 2, 0, 1 + ).contiguous() # nH, Wh*Ww, Wh*Ww attn = attn + relative_position_bias.unsqueeze(0) if mask is not None: nW = mask.shape[0] - attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0).type(attn.type()) + attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze( + 1 + ).unsqueeze(0).type(attn.type()) attn = attn.view(-1, self.num_heads, N, N) attn = self.softmax(attn) else: @@ -154,7 +196,7 @@ def forward(self, x, mask=None): return x, attn_out def extra_repr(self) -> str: - return f'dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}' + return f"dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}" def flops(self, N): # calculate flops for 1 window with token length of N @@ -194,9 +236,22 @@ class SwinTransformerBlock(nn.Module): norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm """ - def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0, - mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0., - act_layer=nn.GELU, norm_layer=nn.LayerNorm): + def __init__( + self, + dim, + input_resolution, + num_heads, + window_size=7, + shift_size=0, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop=0.0, + attn_drop=0.0, + drop_path=0.0, + act_layer=nn.GELU, + norm_layer=nn.LayerNorm, + ): super().__init__() self.dim = dim self.input_resolution = input_resolution @@ -208,50 +263,69 @@ def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0 # if window size is larger than input resolution, we don't partition windows self.shift_size = 0 self.window_size = min(self.input_resolution) - assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size" + assert ( + 0 <= self.shift_size < self.window_size + ), "shift_size must in 0-window_size" self.norm1 = norm_layer(dim) self.attn = WindowAttention( - dim, window_size=(self.window_size, self.window_size), num_heads=num_heads, - qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop) - - self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + dim, + window_size=(self.window_size, self.window_size), + num_heads=num_heads, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + attn_drop=attn_drop, + proj_drop=drop, + ) + + self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity() self.norm2 = norm_layer(dim) mlp_hidden_dim = int(dim * mlp_ratio) - self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) + self.mlp = Mlp( + in_features=dim, + hidden_features=mlp_hidden_dim, + act_layer=act_layer, + drop=drop, + ) self.H = input_resolution[0] self.W = input_resolution[1] self.attn_mask_dict = {} - def create_attn_mask(self, H, W): # calculate attention mask for SW-MSA Hp = int(np.ceil(H / self.window_size)) * self.window_size Wp = int(np.ceil(W / self.window_size)) * self.window_size img_mask = torch.zeros((1, Hp, Wp, 1)) # 1 Hp Wp 1 - h_slices = (slice(0, -self.window_size), - slice(-self.window_size, -self.shift_size), - slice(-self.shift_size, None)) - w_slices = (slice(0, -self.window_size), - slice(-self.window_size, -self.shift_size), - slice(-self.shift_size, None)) + h_slices = ( + slice(0, -self.window_size), + slice(-self.window_size, -self.shift_size), + slice(-self.shift_size, None), + ) + w_slices = ( + slice(0, -self.window_size), + slice(-self.window_size, -self.shift_size), + slice(-self.shift_size, None), + ) cnt = 0 for h in h_slices: for w in w_slices: img_mask[:, h, w, :] = cnt cnt += 1 - mask_windows = window_partition(img_mask, self.window_size) # nW, window_size, window_size, 1 + mask_windows = window_partition( + img_mask, self.window_size + ) # nW, window_size, window_size, 1 mask_windows = mask_windows.view(-1, self.window_size * self.window_size) attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2) - attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0)) + attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill( + attn_mask == 0, float(0.0) + ) return attn_mask - def forward(self, x): B, L, C = x.shape H = int(sqrt(L)) @@ -270,12 +344,16 @@ def forward(self, x): # cyclic shift if self.shift_size > 0: - shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2)) + shifted_x = torch.roll( + x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2) + ) if H in self.attn_mask_dict.keys(): attn_mask = self.attn_mask_dict[H] else: - self.attn_mask_dict[H] = self.create_attn_mask(self.H, self.W).to(x.device) + self.attn_mask_dict[H] = self.create_attn_mask(self.H, self.W).to( + x.device + ) attn_mask = self.attn_mask_dict[H] else: @@ -283,11 +361,17 @@ def forward(self, x): attn_mask = None # partition windows - x_windows = window_partition(shifted_x, self.window_size) # nW*B, window_size, window_size, C - x_windows = x_windows.view(-1, self.window_size * self.window_size, C) # nW*B, window_size*window_size, C + x_windows = window_partition( + shifted_x, self.window_size + ) # nW*B, window_size, window_size, C + x_windows = x_windows.view( + -1, self.window_size * self.window_size, C + ) # nW*B, window_size*window_size, C # W-MSA/SW-MSA - attn_windows, attn = self.attn(x_windows, attn_mask) # nW*B, window_size*window_size, C + attn_windows, attn = self.attn( + x_windows, attn_mask + ) # nW*B, window_size*window_size, C # merge windows attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C) @@ -295,7 +379,9 @@ def forward(self, x): # reverse cyclic shift if self.shift_size > 0: - x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2)) + x = torch.roll( + shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2) + ) else: x = shifted_x @@ -311,8 +397,10 @@ def forward(self, x): return x, attn def extra_repr(self) -> str: - return f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " \ - f"window_size={self.window_size}, shift_size={self.shift_size} mlp_ratio={self.mlp_ratio}" + return ( + f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " + f"window_size={self.window_size}, shift_size={self.shift_size} mlp_ratio={self.mlp_ratio}" + ) def flops(self): flops = 0 @@ -345,7 +433,7 @@ def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm): self.norm = norm_layer(4 * dim) def forward(self, x): - """ Forward function. + """Forward function. Args: x: Input feature, tensor size (B, H*W, C). H, W: Spatial resolution of the input feature. @@ -373,7 +461,6 @@ def forward(self, x): return x - def extra_repr(self) -> str: return f"input_resolution={self.input_resolution}, dim={self.dim}" @@ -402,27 +489,53 @@ class BasicLayer(nn.Module): downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None """ - def __init__(self, dim, input_resolution, depth, num_heads, window_size, - mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., - drop_path=0., norm_layer=nn.LayerNorm, downsample=None): + def __init__( + self, + dim, + input_resolution, + depth, + num_heads, + window_size, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop=0.0, + attn_drop=0.0, + drop_path=0.0, + norm_layer=nn.LayerNorm, + downsample=None, + ): super().__init__() self.dim = dim self.input_resolution = input_resolution self.depth = depth - self.blocks = nn.ModuleList([ - SwinTransformerBlock(dim=dim, input_resolution=input_resolution, - num_heads=num_heads, window_size=window_size, - shift_size=0 if (i % 2 == 0) else window_size // 2, - mlp_ratio=mlp_ratio, - qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop, attn_drop=attn_drop, - drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path, - norm_layer=norm_layer) - for i in range(depth)]) + self.blocks = nn.ModuleList( + [ + SwinTransformerBlock( + dim=dim, + input_resolution=input_resolution, + num_heads=num_heads, + window_size=window_size, + shift_size=0 if (i % 2 == 0) else window_size // 2, + mlp_ratio=mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop, + attn_drop=attn_drop, + drop_path=drop_path[i] + if isinstance(drop_path, list) + else drop_path, + norm_layer=norm_layer, + ) + for i in range(depth) + ] + ) if downsample is not None: - self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer) + self.downsample = downsample( + input_resolution, dim=dim, norm_layer=norm_layer + ) else: self.downsample = None @@ -451,7 +564,6 @@ def forward_with_attention(self, x): x = self.downsample(x) return x, attns - def extra_repr(self) -> str: return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}" @@ -465,14 +577,18 @@ def flops(self): class PatchEmbed(nn.Module): - """ Image to Patch Embedding - """ + """Image to Patch Embedding""" - def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, norm_layer=None): + def __init__( + self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, norm_layer=None + ): super().__init__() img_size = (img_size, img_size) patch_size = (patch_size, patch_size) - patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]] + patches_resolution = [ + img_size[0] // patch_size[0], + img_size[1] // patch_size[1], + ] self.img_size = img_size self.patch_size = patch_size self.patches_resolution = patches_resolution @@ -481,7 +597,9 @@ def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, norm_ self.in_chans = in_chans self.embed_dim = embed_dim - self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size) + self.proj = nn.Conv2d( + in_chans, embed_dim, kernel_size=patch_size, stride=patch_size + ) if norm_layer is not None: self.norm = norm_layer(embed_dim) else: @@ -495,16 +613,22 @@ def forward(self, x): x = self.norm(x) return x - def flops(self): Ho, Wo = self.patches_resolution - flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1]) + flops = ( + Ho + * Wo + * self.embed_dim + * self.in_chans + * (self.patch_size[0] * self.patch_size[1]) + ) if self.norm is not None: flops += Ho * Wo * self.embed_dim return flops + class SwinTransformer(nn.Module): - r""" Swin Transformer + r"""Swin Transformer A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows` - https://arxiv.org/pdf/2103.14030 Args: @@ -527,11 +651,27 @@ class SwinTransformer(nn.Module): patch_norm (bool): If True, add normalization after patch embedding. """ - def __init__(self, img_size=224, patch_size=4, in_chans=3, num_classes=1000, - embed_dim=96, depths=[2, 2, 6, 2], num_heads=[3, 6, 12, 24], - window_size=7, mlp_ratio=4., qkv_bias=True, qk_scale=None, - drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1, - norm_layer=nn.LayerNorm, ape=False, patch_norm=True, **kwargs): + def __init__( + self, + img_size=224, + patch_size=4, + in_chans=3, + num_classes=1000, + embed_dim=96, + depths=[2, 2, 6, 2], + num_heads=[3, 6, 12, 24], + window_size=7, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop_rate=0.0, + attn_drop_rate=0.0, + drop_path_rate=0.1, + norm_layer=nn.LayerNorm, + ape=False, + patch_norm=True, + **kwargs, + ): super().__init__() self.num_classes = num_classes @@ -543,33 +683,47 @@ def __init__(self, img_size=224, patch_size=4, in_chans=3, num_classes=1000, self.mlp_ratio = mlp_ratio self.patch_embed = PatchEmbed( - img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim, - norm_layer=norm_layer if self.patch_norm else None) + img_size=img_size, + patch_size=patch_size, + in_chans=in_chans, + embed_dim=embed_dim, + norm_layer=norm_layer if self.patch_norm else None, + ) num_patches = self.patch_embed.num_patches patches_resolution = self.patch_embed.patches_resolution self.patches_resolution = patches_resolution if self.ape: - self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim)) - trunc_normal_(self.absolute_pos_embed, std=.02) + self.absolute_pos_embed = nn.Parameter( + torch.zeros(1, num_patches, embed_dim) + ) + trunc_normal_(self.absolute_pos_embed, std=0.02) self.pos_drop = nn.Dropout(p=drop_rate) - dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] # stochastic depth decay rule + dpr = [ + x.item() for x in torch.linspace(0, drop_path_rate, sum(depths)) + ] # stochastic depth decay rule self.layers = nn.ModuleList() for i_layer in range(self.num_layers): - layer = BasicLayer(dim=int(embed_dim * 2 ** i_layer), - input_resolution=(patches_resolution[0] // (2 ** i_layer), - patches_resolution[1] // (2 ** i_layer)), - depth=depths[i_layer], - num_heads=num_heads[i_layer], - window_size=window_size, - mlp_ratio=self.mlp_ratio, - qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, - drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])], - norm_layer=norm_layer, - downsample=PatchMerging if (i_layer < self.num_layers - 1) else None) + layer = BasicLayer( + dim=int(embed_dim * 2**i_layer), + input_resolution=( + patches_resolution[0] // (2**i_layer), + patches_resolution[1] // (2**i_layer), + ), + depth=depths[i_layer], + num_heads=num_heads[i_layer], + window_size=window_size, + mlp_ratio=self.mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])], + norm_layer=norm_layer, + downsample=PatchMerging if (i_layer < self.num_layers - 1) else None, + ) self.layers.append(layer) self.norm = norm_layer(self.num_features) @@ -579,7 +733,7 @@ def __init__(self, img_size=224, patch_size=4, in_chans=3, num_classes=1000, def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -588,12 +742,12 @@ def _init_weights(self, m): @torch.jit.ignore def no_weight_decay(self): - return {'absolute_pos_embed'} + return {"absolute_pos_embed"} @torch.jit.ignore def no_weight_decay_keywords(self): # todo: to be implemented - return {'relative_position_bias_table'} + return {"relative_position_bias_table"} def forward(self, x): x = self.patch_embed(x) @@ -610,7 +764,6 @@ def forward(self, x): return x - def forward_feature_maps(self, x): x = self.patch_embed(x) if self.ape: @@ -626,17 +779,15 @@ def forward_feature_maps(self, x): return x, x_grid - def forward_selfattention(self, x, n=1): # n=1 return the last layer attn map; otherwise return attn maps in all layers - x = self.patch_embed(x) if self.ape: x = x + self.absolute_pos_embed x = self.pos_drop(x) - if n==1: + if n == 1: return self.forward_last_selfattention(x) else: return self.forward_all_selfattention(x) @@ -659,8 +810,9 @@ def forward_all_selfattention(self, x): return attn_out - - def forward_return_n_last_blocks(self, x, n=1, return_patch_avgpool=False, depth=[]): + def forward_return_n_last_blocks( + self, x, n=1, return_patch_avgpool=False, depth=[] + ): num_blks = sum(depth) start_idx = num_blks - n @@ -673,7 +825,6 @@ def forward_return_n_last_blocks(self, x, n=1, return_patch_avgpool=False, depth start_blk = start_idx - sum_cur sum_cur = sum_cur_new - x = self.patch_embed(x) if self.ape: x = x + self.absolute_pos_embed @@ -689,19 +840,17 @@ def forward_return_n_last_blocks(self, x, n=1, return_patch_avgpool=False, depth if i >= start_stage: for x_ in fea[start_blk:]: - if i == len(self.layers)-1: # use the norm in the last stage + if i == len(self.layers) - 1: # use the norm in the last stage x_ = self.norm(x_) - x_avg = torch.flatten(self.avgpool(x_.transpose(1, 2)), 1) # B C - # print(f'Stage {i}, x_avg {x_avg.shape}') + x_avg = torch.flatten(self.avgpool(x_.transpose(1, 2)), 1) # B C + # print(f'Stage {i}, x_avg {x_avg.shape}') output.append(x_avg) start_blk = 0 return torch.cat(output, dim=-1) - - def flops(self): flops = 0 flops += self.patch_embed.flops() @@ -709,33 +858,40 @@ def flops(self): flops += layer.flops() if dist.get_rank() == 0: print(f"GFLOPs layer_{i}: {layer.flops() / 1e9}") - flops += self.num_features * self.patches_resolution[0] * self.patches_resolution[1] // (2 ** self.num_layers) + flops += ( + self.num_features + * self.patches_resolution[0] + * self.patches_resolution[1] + // (2**self.num_layers) + ) flops += self.num_features * self.num_classes return flops - def init_weights(self, pretrained='', pretrained_layers=[], verbose=True): + def init_weights(self, pretrained="", pretrained_layers=[], verbose=True): if os.path.isfile(pretrained): - pretrained_dict = torch.load(pretrained, map_location='cpu') - logging.info(f'=> loading pretrained model {pretrained}') + pretrained_dict = torch.load(pretrained, map_location="cpu") + logging.info(f"=> loading pretrained model {pretrained}") model_dict = self.state_dict() pretrained_dict = { - k: v for k, v in pretrained_dict.items() - if k in model_dict.keys() + k: v for k, v in pretrained_dict.items() if k in model_dict.keys() } need_init_state_dict = {} for k, v in pretrained_dict.items(): need_init = ( - k.split('.')[0] in pretrained_layers - or pretrained_layers[0] is '*' - or 'relative_position_index' not in k - or 'attn_mask' not in k + k.split(".")[0] in pretrained_layers + or pretrained_layers[0] is "*" + or "relative_position_index" not in k + or "attn_mask" not in k ) if need_init: if verbose: - logging.info(f'=> init {k} from {pretrained}') + logging.info(f"=> init {k} from {pretrained}") - if 'relative_position_bias_table' in k and v.size() != model_dict[k].size(): + if ( + "relative_position_bias_table" in k + and v.size() != model_dict[k].size() + ): relative_position_bias_table_pretrained = v relative_position_bias_table_current = model_dict[k] L1, nH1 = relative_position_bias_table_pretrained.size() @@ -745,18 +901,28 @@ def init_weights(self, pretrained='', pretrained_layers=[], verbose=True): else: if L1 != L2: logging.info( - '=> load_pretrained: resized variant: {} to {}' - .format((L1, nH1), (L2, nH2)) + "=> load_pretrained: resized variant: {} to {}".format( + (L1, nH1), (L2, nH2) + ) + ) + S1 = int(L1**0.5) + S2 = int(L2**0.5) + relative_position_bias_table_pretrained_resized = ( + torch.nn.functional.interpolate( + relative_position_bias_table_pretrained.permute( + 1, 0 + ).view(1, nH1, S1, S1), + size=(S2, S2), + mode="bicubic", + ) + ) + v = relative_position_bias_table_pretrained_resized.view( + nH2, L2 + ).permute( + 1, 0 ) - S1 = int(L1 ** 0.5) - S2 = int(L2 ** 0.5) - relative_position_bias_table_pretrained_resized = torch.nn.functional.interpolate( - relative_position_bias_table_pretrained.permute(1, 0).view(1, nH1, S1, S1), - size=(S2, S2), - mode='bicubic') - v = relative_position_bias_table_pretrained_resized.view(nH2, L2).permute(1, 0) - - if 'absolute_pos_embed' in k and v.size() != model_dict[k].size(): + + if "absolute_pos_embed" in k and v.size() != model_dict[k].size(): absolute_pos_embed_pretrained = v absolute_pos_embed_current = model_dict[k] _, L1, C1 = absolute_pos_embed_pretrained.size() @@ -766,16 +932,30 @@ def init_weights(self, pretrained='', pretrained_layers=[], verbose=True): else: if L1 != L2: logging.info( - '=> load_pretrained: resized variant: {} to {}' - .format((1, L1, C1), (1, L2, C2)) + "=> load_pretrained: resized variant: {} to {}".format( + (1, L1, C1), (1, L2, C2) + ) + ) + S1 = int(L1**0.5) + S2 = int(L2**0.5) + absolute_pos_embed_pretrained = ( + absolute_pos_embed_pretrained.reshape( + -1, S1, S1, C1 + ) ) - S1 = int(L1 ** 0.5) - S2 = int(L2 ** 0.5) - absolute_pos_embed_pretrained = absolute_pos_embed_pretrained.reshape(-1, S1, S1, C1) - absolute_pos_embed_pretrained = absolute_pos_embed_pretrained.permute(0, 3, 1, 2) - absolute_pos_embed_pretrained_resized = torch.nn.functional.interpolate( - absolute_pos_embed_pretrained, size=(S2, S2), mode='bicubic') - v = absolute_pos_embed_pretrained_resized.permute(0, 2, 3, 1).flatten(1, 2) + absolute_pos_embed_pretrained = ( + absolute_pos_embed_pretrained.permute(0, 3, 1, 2) + ) + absolute_pos_embed_pretrained_resized = ( + torch.nn.functional.interpolate( + absolute_pos_embed_pretrained, + size=(S2, S2), + mode="bicubic", + ) + ) + v = absolute_pos_embed_pretrained_resized.permute( + 0, 2, 3, 1 + ).flatten(1, 2) need_init_state_dict[k] = v self.load_state_dict(need_init_state_dict, strict=False) @@ -783,27 +963,21 @@ def init_weights(self, pretrained='', pretrained_layers=[], verbose=True): def freeze_pretrained_layers(self, frozen_layers=[]): for name, module in self.named_modules(): if ( - name.split('.')[0] in frozen_layers - or '.'.join(name.split('.')[0:2]) in frozen_layers - or (len(frozen_layers) > 0 and frozen_layers[0] is '*') + name.split(".")[0] in frozen_layers + or ".".join(name.split(".")[0:2]) in frozen_layers + or (len(frozen_layers) > 0 and frozen_layers[0] is "*") ): for _name, param in module.named_parameters(): param.requires_grad = False - logging.info( - '=> set param {} requires grad to False' - .format(name) - ) + logging.info("=> set param {} requires grad to False".format(name)) for name, param in self.named_parameters(): if ( - name.split('.')[0] in frozen_layers - or (len(frozen_layers) > 0 and frozen_layers[0] is '*') - and param.requires_grad is True + name.split(".")[0] in frozen_layers + or (len(frozen_layers) > 0 and frozen_layers[0] is "*") + and param.requires_grad is True ): param.requires_grad = False - logging.info( - '=> set param {} requires grad to False' - .format(name) - ) + logging.info("=> set param {} requires grad to False".format(name)) return self @@ -815,7 +989,7 @@ def get_swin(is_teacher=False): depths = [2, 2, 6, 2] num_heads = [3, 6, 12, 24] drop_path_rate = 0.1 - elif args.swin_backbone_type == 'h3': + elif args.swin_backbone_type == "h3": embed_dim = 384 depths = [2, 2, 18, 2] num_heads = [6, 12, 24, 48] @@ -846,4 +1020,3 @@ def get_swin(is_teacher=False): ) return swin - diff --git a/megatron/model/vision/inpainting.py b/megatron/model/vision/inpainting.py index 3d7a6da415e..b01ace72f28 100644 --- a/megatron/model/vision/inpainting.py +++ b/megatron/model/vision/inpainting.py @@ -42,7 +42,7 @@ def __init__(self, config, pre_process=True, post_process=True): self.hidden_size, self.backbone.flatten_dim, torch.nn.init.zeros_, - gather_params_on_init=args.zero_stage == 3 + gather_params_on_init=args.zero_stage == 3, ) def set_input_tensor(self, input_tensor): @@ -56,13 +56,13 @@ def forward(self, input): return hidden_states decoded_output = self.linear_decoder(hidden_states) output = einops.rearrange( - decoded_output, - "b (h w) (p1 p2 c) -> b c (h p1) (w p2)", - p1=self.patch_dim, - p2=self.patch_dim, - h=self.img_h//self.patch_dim, - w=self.img_w//self.patch_dim, - ) + decoded_output, + "b (h w) (p1 p2 c) -> b c (h p1) (w p2)", + p1=self.patch_dim, + p2=self.patch_dim, + h=self.img_h // self.patch_dim, + w=self.img_w // self.patch_dim, + ) return output @@ -71,6 +71,7 @@ class MLP(torch.nn.Module): """ Linear Embedding """ + def __init__(self, input_dim=2048, embed_dim=768): super().__init__() self.proj = torch.nn.Linear(input_dim, embed_dim) @@ -99,19 +100,25 @@ def __init__(self, pre_process=True, post_process=True): self.in_channels = [64, 128, 320, 512] self.embedding_dim = 768 - c1_in_channels, c2_in_channels, c3_in_channels, c4_in_channels = self.in_channels + c1_in_channels, c2_in_channels, c3_in_channels, c4_in_channels = ( + self.in_channels + ) self.linear_c4 = MLP(input_dim=c4_in_channels, embed_dim=self.embedding_dim) self.linear_c3 = MLP(input_dim=c3_in_channels, embed_dim=self.embedding_dim) self.linear_c2 = MLP(input_dim=c2_in_channels, embed_dim=self.embedding_dim) self.linear_c1 = MLP(input_dim=c1_in_channels, embed_dim=self.embedding_dim) - self.conv_fuse = torch.nn.Conv2d(self.embedding_dim*4, self.embedding_dim, 1, 1, bias=False) + self.conv_fuse = torch.nn.Conv2d( + self.embedding_dim * 4, self.embedding_dim, 1, 1, bias=False + ) self.norm = apex.parallel.SyncBatchNorm(self.embedding_dim) self.dropout = torch.nn.Dropout2d(0.1) - - self.linear_pred = torch.nn.Conv2d(self.embedding_dim, self.flatten_dim, kernel_size=1) - + + self.linear_pred = torch.nn.Conv2d( + self.embedding_dim, self.flatten_dim, kernel_size=1 + ) + def set_input_tensor(self, input_tensor): """See megatron.model.transformer.set_input_tensor()""" pass @@ -120,20 +127,28 @@ def forward(self, input): c1, c2, c3, c4 = self.backbone(input) n, _, h, w = c4.shape - _c4 = self.linear_c4(c4).permute(0, 2, 1).reshape(n, -1, c4.shape[2], c4.shape[3]) - _c4 = resize(_c4, size=c1.size()[2:], mode='bilinear', align_corners=False) - - _c3 = self.linear_c3(c3).permute(0, 2, 1).reshape(n, -1, c3.shape[2], c3.shape[3]) - _c3 = resize(_c3, size=c1.size()[2:], mode='bilinear', align_corners=False) + _c4 = ( + self.linear_c4(c4).permute(0, 2, 1).reshape(n, -1, c4.shape[2], c4.shape[3]) + ) + _c4 = resize(_c4, size=c1.size()[2:], mode="bilinear", align_corners=False) + + _c3 = ( + self.linear_c3(c3).permute(0, 2, 1).reshape(n, -1, c3.shape[2], c3.shape[3]) + ) + _c3 = resize(_c3, size=c1.size()[2:], mode="bilinear", align_corners=False) - _c2 = self.linear_c2(c2).permute(0, 2, 1).reshape(n, -1, c2.shape[2], c2.shape[3]) - _c2 = resize(_c2, size=c1.size()[2:], mode='bilinear', align_corners=False) + _c2 = ( + self.linear_c2(c2).permute(0, 2, 1).reshape(n, -1, c2.shape[2], c2.shape[3]) + ) + _c2 = resize(_c2, size=c1.size()[2:], mode="bilinear", align_corners=False) - _c1 = self.linear_c1(c1).permute(0, 2, 1).reshape(n, -1, c1.shape[2], c1.shape[3]) + _c1 = ( + self.linear_c1(c1).permute(0, 2, 1).reshape(n, -1, c1.shape[2], c1.shape[3]) + ) _c = torch.cat([_c4, _c3, _c2, _c1], dim=1) _c = self.conv_fuse(_c) - + x = self.norm(_c) x = F.relu(x, inplace=True) x = self.dropout(x) @@ -145,8 +160,8 @@ def forward(self, input): "b (c p1 p2) h w -> b c (h p1) (w p2)", p1=self.patch_dim, p2=self.patch_dim, - h=self.img_h//self.patch_dim, - w=self.img_w//self.patch_dim, + h=self.img_h // self.patch_dim, + w=self.img_w // self.patch_dim, ) return output diff --git a/megatron/model/vision/knn_monitor.py b/megatron/model/vision/knn_monitor.py index a7d79854eb5..18ec61e7b65 100644 --- a/megatron/model/vision/knn_monitor.py +++ b/megatron/model/vision/knn_monitor.py @@ -17,8 +17,11 @@ def build_data_loader(dataset, drop_last=True, shuffle=False): world_size = mpu.get_data_parallel_world_size() rank = mpu.get_data_parallel_rank() sampler = torch.utils.data.distributed.DistributedSampler( - dataset, num_replicas=world_size, rank=rank, - drop_last=drop_last, shuffle=shuffle + dataset, + num_replicas=world_size, + rank=rank, + drop_last=drop_last, + shuffle=shuffle, ) # Data loader. Note that batch size is the per GPU batch size. @@ -43,11 +46,11 @@ def compute_feature_bank(model): train_ds = ImageFolder( root=args.data_path[0], transform=ClassificationTransform((args.img_h, args.img_w), train=False), - data_per_class_fraction=1.0 + data_per_class_fraction=1.0, ) classes = len(train_ds.classes) dataloader = build_data_loader(train_ds) - + for m in model: m.eval() @@ -59,7 +62,7 @@ def compute_feature_bank(model): feature = F.normalize(teacher_feature.float(), dim=1) feature_bank.append(feature) feature_label.append(labels) - + for m in model: m.train() @@ -67,20 +70,25 @@ def compute_feature_bank(model): feature_bank = torch.cat(feature_bank, dim=0).contiguous() feature_label = torch.cat(feature_label, dim=0).contiguous() - feature_banks = [torch.zeros_like(feature_bank) - for i in range(mpu.get_data_parallel_world_size())] - torch.distributed.all_gather(feature_banks, - feature_bank, - group=mpu.get_data_parallel_group()) + feature_banks = [ + torch.zeros_like(feature_bank) + for i in range(mpu.get_data_parallel_world_size()) + ] + torch.distributed.all_gather( + feature_banks, feature_bank, group=mpu.get_data_parallel_group() + ) - assert torch.all(torch.eq(feature_banks[mpu.get_data_parallel_rank()], - feature_bank)) + assert torch.all( + torch.eq(feature_banks[mpu.get_data_parallel_rank()], feature_bank) + ) - feature_labels = [torch.zeros_like(feature_label) - for i in range(mpu.get_data_parallel_world_size())] - torch.distributed.all_gather(feature_labels, - feature_label, - group=mpu.get_data_parallel_group()) + feature_labels = [ + torch.zeros_like(feature_label) + for i in range(mpu.get_data_parallel_world_size()) + ] + torch.distributed.all_gather( + feature_labels, feature_label, group=mpu.get_data_parallel_group() + ) # [D, N] feature_banks = torch.cat(feature_banks, dim=0).t().contiguous() @@ -107,23 +115,24 @@ def knn_predict(feature, feature_bank, feature_labels, classes, knn_k, knn_t): # [B, K] sim_weight, sim_indices = sim_matrix.topk(k=knn_k, dim=-1) # [B, K] - sim_labels = torch.gather(feature_labels.expand(feature.size(0), -1), - dim=-1, - index=sim_indices) + sim_labels = torch.gather( + feature_labels.expand(feature.size(0), -1), dim=-1, index=sim_indices + ) sim_weight = (sim_weight / knn_t).exp() # counts for each class - one_hot_label = torch.zeros(feature.size(0) * knn_k, - classes, - device=sim_labels.device) + one_hot_label = torch.zeros( + feature.size(0) * knn_k, classes, device=sim_labels.device + ) # [B*K, C] - one_hot_label = one_hot_label.scatter(dim=-1, - index=sim_labels.view(-1, 1), - value=1.0) + one_hot_label = one_hot_label.scatter( + dim=-1, index=sim_labels.view(-1, 1), value=1.0 + ) # weighted score ---> [B, C] pred_scores = torch.sum( - one_hot_label.view(feature.size(0), -1, classes) * sim_weight.unsqueeze(dim=-1), - dim=1) + one_hot_label.view(feature.size(0), -1, classes) * sim_weight.unsqueeze(dim=-1), + dim=1, + ) pred_labels = pred_scores.argsort(dim=-1, descending=True) return pred_labels diff --git a/megatron/model/vision/mit_backbone.py b/megatron/model/vision/mit_backbone.py index c67ca2c62bb..d238aab264a 100644 --- a/megatron/model/vision/mit_backbone.py +++ b/megatron/model/vision/mit_backbone.py @@ -2,7 +2,7 @@ # Copyright (c) 2021, NVIDIA Corporation. All rights reserved. # # This work is licensed under the NVIDIA Source Code License -# found in the LICENSE file in the root directory of this +# found in the LICENSE file in the root directory of this # source tree. # --------------------------------------------------------------- import math @@ -16,12 +16,15 @@ class Mlp(nn.Module): - def __init__(self, - in_features, - hidden_features=None, - out_features=None, - act_layer=nn.GELU, - drop=0.): + + def __init__( + self, + in_features, + hidden_features=None, + out_features=None, + act_layer=nn.GELU, + drop=0.0, + ): super().__init__() out_features = out_features or in_features hidden_features = hidden_features or in_features @@ -35,7 +38,7 @@ def __init__(self, def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -59,21 +62,26 @@ def forward(self, x, H, W): class Attention(nn.Module): - def __init__(self, - dim, - num_heads=8, - qkv_bias=False, - qk_scale=None, - attn_drop=0., - proj_drop=0., - sr_ratio=1): + + def __init__( + self, + dim, + num_heads=8, + qkv_bias=False, + qk_scale=None, + attn_drop=0.0, + proj_drop=0.0, + sr_ratio=1, + ): super().__init__() - assert dim % num_heads == 0, f"dim {dim} should be divided by num_heads {num_heads}." + assert ( + dim % num_heads == 0 + ), f"dim {dim} should be divided by num_heads {num_heads}." self.dim = dim self.num_heads = num_heads head_dim = dim // num_heads - self.scale = qk_scale or head_dim ** -0.5 + self.scale = qk_scale or head_dim**-0.5 self.q = nn.Linear(dim, dim, bias=qkv_bias) self.kv = nn.Linear(dim, dim * 2, bias=qkv_bias) @@ -90,7 +98,7 @@ def __init__(self, def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -105,15 +113,27 @@ def _init_weights(self, m): def forward(self, x, H, W): B, N, C = x.shape - q = self.q(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3) + q = ( + self.q(x) + .reshape(B, N, self.num_heads, C // self.num_heads) + .permute(0, 2, 1, 3) + ) if self.sr_ratio > 1: x_ = x.permute(0, 2, 1).reshape(B, C, H, W) x_ = self.sr(x_).reshape(B, C, -1).permute(0, 2, 1) x_ = self.norm(x_) - kv = self.kv(x_).reshape(B, -1, 2, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) + kv = ( + self.kv(x_) + .reshape(B, -1, 2, self.num_heads, C // self.num_heads) + .permute(2, 0, 3, 1, 4) + ) else: - kv = self.kv(x).reshape(B, -1, 2, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) + kv = ( + self.kv(x) + .reshape(B, -1, 2, self.num_heads, C // self.num_heads) + .permute(2, 0, 3, 1, 4) + ) k, v = kv[0], kv[1] attn = (q @ k.transpose(-2, -1)) * self.scale @@ -129,25 +149,47 @@ def forward(self, x, H, W): class Block(nn.Module): - def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0., - drop_path=0., act_layer=nn.GELU, norm_layer=LayerNorm, sr_ratio=1): + def __init__( + self, + dim, + num_heads, + mlp_ratio=4.0, + qkv_bias=False, + qk_scale=None, + drop=0.0, + attn_drop=0.0, + drop_path=0.0, + act_layer=nn.GELU, + norm_layer=LayerNorm, + sr_ratio=1, + ): super().__init__() self.norm1 = norm_layer(dim) self.attn = Attention( dim, - num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, - attn_drop=attn_drop, proj_drop=drop, sr_ratio=sr_ratio) + num_heads=num_heads, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + attn_drop=attn_drop, + proj_drop=drop, + sr_ratio=sr_ratio, + ) # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here - self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity() self.norm2 = norm_layer(dim) mlp_hidden_dim = int(dim * mlp_ratio) - self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) + self.mlp = Mlp( + in_features=dim, + hidden_features=mlp_hidden_dim, + act_layer=act_layer, + drop=drop, + ) self.apply(self._init_weights) def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -168,23 +210,27 @@ def forward(self, x, H, W): class OverlapPatchEmbed(nn.Module): - """ Image to Patch Embedding - """ + """Image to Patch Embedding""" def __init__(self, img_size=224, patch_size=7, stride=4, in_chans=3, embed_dim=768): super().__init__() img_size = (img_size, img_size) patch_size = (patch_size, patch_size) - self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=stride, - padding=(patch_size[0] // 2, patch_size[1] // 2)) + self.proj = nn.Conv2d( + in_chans, + embed_dim, + kernel_size=patch_size, + stride=stride, + padding=(patch_size[0] // 2, patch_size[1] // 2), + ) self.norm = LayerNorm(embed_dim) self.apply(self._init_weights) def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -207,64 +253,150 @@ def forward(self, x): class MixVisionTransformer(nn.Module): - def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, embed_dims=[64, 128, 256, 512], - num_heads=[1, 2, 4, 8], mlp_ratios=[4, 4, 4, 4], qkv_bias=False, qk_scale=None, drop_rate=0., - attn_drop_rate=0., drop_path_rate=0., norm_layer=LayerNorm, - depths=[3, 4, 6, 3], sr_ratios=[8, 4, 2, 1], output_avg=False): + + def __init__( + self, + img_size=224, + patch_size=16, + in_chans=3, + num_classes=1000, + embed_dims=[64, 128, 256, 512], + num_heads=[1, 2, 4, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=False, + qk_scale=None, + drop_rate=0.0, + attn_drop_rate=0.0, + drop_path_rate=0.0, + norm_layer=LayerNorm, + depths=[3, 4, 6, 3], + sr_ratios=[8, 4, 2, 1], + output_avg=False, + ): super().__init__() self.num_classes = num_classes self.depths = depths self.output_avg = output_avg # patch_embed - self.patch_embed1 = OverlapPatchEmbed(img_size=img_size, patch_size=7, stride=4, in_chans=in_chans, - embed_dim=embed_dims[0]) - self.patch_embed2 = OverlapPatchEmbed(img_size=img_size // 4, patch_size=3, stride=2, in_chans=embed_dims[0], - embed_dim=embed_dims[1]) - self.patch_embed3 = OverlapPatchEmbed(img_size=img_size // 8, patch_size=3, stride=2, in_chans=embed_dims[1], - embed_dim=embed_dims[2]) - self.patch_embed4 = OverlapPatchEmbed(img_size=img_size // 16, patch_size=3, stride=2, in_chans=embed_dims[2], - embed_dim=embed_dims[3]) + self.patch_embed1 = OverlapPatchEmbed( + img_size=img_size, + patch_size=7, + stride=4, + in_chans=in_chans, + embed_dim=embed_dims[0], + ) + self.patch_embed2 = OverlapPatchEmbed( + img_size=img_size // 4, + patch_size=3, + stride=2, + in_chans=embed_dims[0], + embed_dim=embed_dims[1], + ) + self.patch_embed3 = OverlapPatchEmbed( + img_size=img_size // 8, + patch_size=3, + stride=2, + in_chans=embed_dims[1], + embed_dim=embed_dims[2], + ) + self.patch_embed4 = OverlapPatchEmbed( + img_size=img_size // 16, + patch_size=3, + stride=2, + in_chans=embed_dims[2], + embed_dim=embed_dims[3], + ) # transformer encoder - dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] # stochastic depth decay rule + dpr = [ + x.item() for x in torch.linspace(0, drop_path_rate, sum(depths)) + ] # stochastic depth decay rule cur = 0 - self.block1 = nn.ModuleList([Block( - dim=embed_dims[0], num_heads=num_heads[0], mlp_ratio=mlp_ratios[0], qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[cur + i], norm_layer=norm_layer, - sr_ratio=sr_ratios[0]) - for i in range(depths[0])]) + self.block1 = nn.ModuleList( + [ + Block( + dim=embed_dims[0], + num_heads=num_heads[0], + mlp_ratio=mlp_ratios[0], + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[cur + i], + norm_layer=norm_layer, + sr_ratio=sr_ratios[0], + ) + for i in range(depths[0]) + ] + ) self.norm1 = norm_layer(embed_dims[0]) cur += depths[0] - self.block2 = nn.ModuleList([Block( - dim=embed_dims[1], num_heads=num_heads[1], mlp_ratio=mlp_ratios[1], qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[cur + i], norm_layer=norm_layer, - sr_ratio=sr_ratios[1]) - for i in range(depths[1])]) + self.block2 = nn.ModuleList( + [ + Block( + dim=embed_dims[1], + num_heads=num_heads[1], + mlp_ratio=mlp_ratios[1], + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[cur + i], + norm_layer=norm_layer, + sr_ratio=sr_ratios[1], + ) + for i in range(depths[1]) + ] + ) self.norm2 = norm_layer(embed_dims[1]) cur += depths[1] - self.block3 = nn.ModuleList([Block( - dim=embed_dims[2], num_heads=num_heads[2], mlp_ratio=mlp_ratios[2], qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[cur + i], norm_layer=norm_layer, - sr_ratio=sr_ratios[2]) - for i in range(depths[2])]) + self.block3 = nn.ModuleList( + [ + Block( + dim=embed_dims[2], + num_heads=num_heads[2], + mlp_ratio=mlp_ratios[2], + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[cur + i], + norm_layer=norm_layer, + sr_ratio=sr_ratios[2], + ) + for i in range(depths[2]) + ] + ) self.norm3 = norm_layer(embed_dims[2]) cur += depths[2] - self.block4 = nn.ModuleList([Block( - dim=embed_dims[3], num_heads=num_heads[3], mlp_ratio=mlp_ratios[3], qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[cur + i], norm_layer=norm_layer, - sr_ratio=sr_ratios[3]) - for i in range(depths[3])]) + self.block4 = nn.ModuleList( + [ + Block( + dim=embed_dims[3], + num_heads=num_heads[3], + mlp_ratio=mlp_ratios[3], + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[cur + i], + norm_layer=norm_layer, + sr_ratio=sr_ratios[3], + ) + for i in range(depths[3]) + ] + ) self.norm4 = norm_layer(embed_dims[3]) self.apply(self._init_weights) def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -339,7 +471,7 @@ def forward_features(self, x): def forward(self, x): x = self.forward_features(x) - + if self.output_avg: x = x[3].mean(dim=1) @@ -347,6 +479,7 @@ def forward(self, x): class DWConv(nn.Module): + def __init__(self, dim=768): super(DWConv, self).__init__() self.dwconv = nn.Conv2d(dim, dim, 3, 1, 1, bias=True, groups=dim) @@ -359,62 +492,140 @@ def forward(self, x, H, W): return x + class mit_b0(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b0, self).__init__( - patch_size=4, embed_dims=[32, 64, 160, 256], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[2, 2, 2, 2], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[32, 64, 160, 256], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[2, 2, 2, 2], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) class mit_b1(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b1, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[2, 2, 2, 2], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[2, 2, 2, 2], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) class mit_b2(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b2, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 4, 6, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 4, 6, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) + - class mit_b3(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b3, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 4, 18, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 4, 18, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) + class mit_b3_avg(MixVisionTransformer): + def __init__(self, drop_path_rate=0.1, **kwargs): super(mit_b3_avg, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 4, 18, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=drop_path_rate, output_avg=True) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 4, 18, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=drop_path_rate, + output_avg=True, + ) + class mit_b4(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b4, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 8, 27, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 8, 27, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) + class mit_b5(MixVisionTransformer): + def __init__(self, **kwargs): super(mit_b5, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 6, 40, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=0.1) + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 6, 40, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=0.1, + ) + class mit_b5_avg(MixVisionTransformer): + def __init__(self, drop_path_rate=0.1, **kwargs): super(mit_b5_avg, self).__init__( - patch_size=4, embed_dims=[64, 128, 320, 512], num_heads=[1, 2, 5, 8], mlp_ratios=[4, 4, 4, 4], - qkv_bias=True, norm_layer=partial(LayerNorm, eps=1e-6), depths=[3, 6, 40, 3], sr_ratios=[8, 4, 2, 1], - drop_rate=0.0, drop_path_rate=drop_path_rate, output_avg=True) - + patch_size=4, + embed_dims=[64, 128, 320, 512], + num_heads=[1, 2, 5, 8], + mlp_ratios=[4, 4, 4, 4], + qkv_bias=True, + norm_layer=partial(LayerNorm, eps=1e-6), + depths=[3, 6, 40, 3], + sr_ratios=[8, 4, 2, 1], + drop_rate=0.0, + drop_path_rate=drop_path_rate, + output_avg=True, + ) diff --git a/megatron/model/vision/swin_backbone.py b/megatron/model/vision/swin_backbone.py index 9a622c7070f..cec11f26928 100644 --- a/megatron/model/vision/swin_backbone.py +++ b/megatron/model/vision/swin_backbone.py @@ -17,8 +17,15 @@ class Mlp(nn.Module): - def __init__(self, in_features, hidden_features=None, - out_features=None, act_layer=nn.GELU, drop=0.): + + def __init__( + self, + in_features, + hidden_features=None, + out_features=None, + act_layer=nn.GELU, + drop=0.0, + ): super().__init__() out_features = out_features or in_features hidden_features = hidden_features or in_features @@ -47,7 +54,9 @@ def window_partition(x, window_size): """ B, H, W, C = x.shape x = x.view(B, H // window_size, window_size, W // window_size, window_size, C) - windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + windows = ( + x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + ) return windows @@ -63,13 +72,15 @@ def window_reverse(windows, window_size, H, W): x: (B, H, W, C) """ B = int(windows.shape[0] / (H * W / window_size / window_size)) - x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) + x = windows.view( + B, H // window_size, W // window_size, window_size, window_size, -1 + ) x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1) return x class WindowAttention(nn.Module): - r""" Window based multi-head self attention (W-MSA) module with relative position bias. + r"""Window based multi-head self attention (W-MSA) module with relative position bias. It supports both of shifted and non-shifted window. Args: @@ -82,26 +93,40 @@ class WindowAttention(nn.Module): proj_drop (float, optional): Dropout ratio of output. Default: 0.0 """ - def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, attn_drop=0., proj_drop=0.): + def __init__( + self, + dim, + window_size, + num_heads, + qkv_bias=True, + qk_scale=None, + attn_drop=0.0, + proj_drop=0.0, + ): super().__init__() self.dim = dim self.window_size = window_size # Wh, Ww self.num_heads = num_heads head_dim = dim // num_heads - self.scale = qk_scale or head_dim ** -0.5 + self.scale = qk_scale or head_dim**-0.5 # define a parameter table of relative position bias self.relative_position_bias_table = nn.Parameter( - torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads)) # 2*Wh-1 * 2*Ww-1, nH + torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads) + ) # 2*Wh-1 * 2*Ww-1, nH # get pair-wise relative position index for each token inside the window coords_h = torch.arange(self.window_size[0]) coords_w = torch.arange(self.window_size[1]) coords = torch.stack(torch.meshgrid([coords_h, coords_w])) # 2, Wh, Ww coords_flatten = torch.flatten(coords, 1) # 2, Wh*Ww - relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :] # 2, Wh*Ww, Wh*Ww - relative_coords = relative_coords.permute(1, 2, 0).contiguous() # Wh*Ww, Wh*Ww, 2 + relative_coords = ( + coords_flatten[:, :, None] - coords_flatten[:, None, :] + ) # 2, Wh*Ww, Wh*Ww + relative_coords = relative_coords.permute( + 1, 2, 0 + ).contiguous() # Wh*Ww, Wh*Ww, 2 relative_coords[:, :, 0] += self.window_size[0] - 1 # shift to start from 0 relative_coords[:, :, 1] += self.window_size[1] - 1 relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1 @@ -113,7 +138,7 @@ def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, at self.proj = nn.Linear(dim, dim) self.proj_drop = nn.Dropout(proj_drop) - trunc_normal_(self.relative_position_bias_table, std=.02) + trunc_normal_(self.relative_position_bias_table, std=0.02) self.softmax = nn.Softmax(dim=-1) def forward(self, x, mask=None): @@ -123,20 +148,37 @@ def forward(self, x, mask=None): mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None """ B_, N, C = x.shape - qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) - q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple) + qkv = ( + self.qkv(x) + .reshape(B_, N, 3, self.num_heads, C // self.num_heads) + .permute(2, 0, 3, 1, 4) + ) + q, k, v = ( + qkv[0], + qkv[1], + qkv[2], + ) # make torchscript happy (cannot use tensor as tuple) q = q * self.scale - attn = (q @ k.transpose(-2, -1)) - - relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view( - self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1) # Wh*Ww,Wh*Ww,nH - relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww + attn = q @ k.transpose(-2, -1) + + relative_position_bias = self.relative_position_bias_table[ + self.relative_position_index.view(-1) + ].view( + self.window_size[0] * self.window_size[1], + self.window_size[0] * self.window_size[1], + -1, + ) # Wh*Ww,Wh*Ww,nH + relative_position_bias = relative_position_bias.permute( + 2, 0, 1 + ).contiguous() # nH, Wh*Ww, Wh*Ww attn = attn + relative_position_bias.unsqueeze(0) if mask is not None: nW = mask.shape[0] - attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0) + attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze( + 1 + ).unsqueeze(0) attn = attn.view(-1, self.num_heads, N, N) attn = self.softmax(attn) else: @@ -150,7 +192,7 @@ def forward(self, x, mask=None): return x def extra_repr(self) -> str: - return f'dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}' + return f"dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}" def flops(self, N): # calculate flops for 1 window with token length of N @@ -167,7 +209,7 @@ def flops(self, N): class SwinTransformerBlock(nn.Module): - r""" Swin Transformer Block. + r"""Swin Transformer Block. Args: dim (int): Number of input channels. @@ -185,9 +227,22 @@ class SwinTransformerBlock(nn.Module): norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm """ - def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0, - mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0., - act_layer=nn.GELU, norm_layer=nn.LayerNorm): + def __init__( + self, + dim, + input_resolution, + num_heads, + window_size=7, + shift_size=0, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop=0.0, + attn_drop=0.0, + drop_path=0.0, + act_layer=nn.GELU, + norm_layer=nn.LayerNorm, + ): super().__init__() self.dim = dim self.input_resolution = input_resolution @@ -199,22 +254,35 @@ def __init__(self, dim, input_resolution, num_heads, window_size=7, shift_size=0 # if window size is larger than input resolution, we don't partition windows self.shift_size = 0 self.window_size = min(self.input_resolution) - assert 0 <= self.shift_size < self.window_size, "shift_size must in 0-window_size" + assert ( + 0 <= self.shift_size < self.window_size + ), "shift_size must in 0-window_size" self.norm1 = norm_layer(dim) self.attn = WindowAttention( - dim, window_size=to_2tuple(self.window_size), num_heads=num_heads, - qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop) - - self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + dim, + window_size=to_2tuple(self.window_size), + num_heads=num_heads, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + attn_drop=attn_drop, + proj_drop=drop, + ) + + self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity() self.norm2 = norm_layer(dim) mlp_hidden_dim = int(dim * mlp_ratio) - self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) + self.mlp = Mlp( + in_features=dim, + hidden_features=mlp_hidden_dim, + act_layer=act_layer, + drop=drop, + ) self.H = input_resolution[0] self.W = input_resolution[1] - self.attn_mask_dict = {} + self.attn_mask_dict = {} def create_attn_mask(self, H, W): # calculate attention mask for SW-MSA @@ -222,26 +290,33 @@ def create_attn_mask(self, H, W): Hp = int(np.ceil(H / self.window_size)) * self.window_size Wp = int(np.ceil(W / self.window_size)) * self.window_size img_mask = torch.zeros((1, Hp, Wp, 1)) # 1 Hp Wp 1 - h_slices = (slice(0, -self.window_size), - slice(-self.window_size, -self.shift_size), - slice(-self.shift_size, None)) - w_slices = (slice(0, -self.window_size), - slice(-self.window_size, -self.shift_size), - slice(-self.shift_size, None)) + h_slices = ( + slice(0, -self.window_size), + slice(-self.window_size, -self.shift_size), + slice(-self.shift_size, None), + ) + w_slices = ( + slice(0, -self.window_size), + slice(-self.window_size, -self.shift_size), + slice(-self.shift_size, None), + ) cnt = 0 for h in h_slices: for w in w_slices: img_mask[:, h, w, :] = cnt cnt += 1 - mask_windows = window_partition(img_mask, self.window_size) # nW, window_size, window_size, 1 + mask_windows = window_partition( + img_mask, self.window_size + ) # nW, window_size, window_size, 1 mask_windows = mask_windows.view(-1, self.window_size * self.window_size) attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2) - attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0)) + attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill( + attn_mask == 0, float(0.0) + ) return attn_mask - def forward(self, x): B, L, C = x.shape H = int(sqrt(L)) @@ -253,16 +328,24 @@ def forward(self, x): # cyclic shift if self.shift_size > 0: - shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2)) + shifted_x = torch.roll( + x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2) + ) else: shifted_x = x # partition windows - x_windows = window_partition(shifted_x, self.window_size) # nW*B, window_size, window_size, C - x_windows = x_windows.view(-1, self.window_size * self.window_size, C) # nW*B, window_size*window_size, C + x_windows = window_partition( + shifted_x, self.window_size + ) # nW*B, window_size, window_size, C + x_windows = x_windows.view( + -1, self.window_size * self.window_size, C + ) # nW*B, window_size*window_size, C # W-MSA/SW-MSA - attn_windows = self.attn(x_windows, mask=self.attn_mask) # nW*B, window_size*window_size, C + attn_windows = self.attn( + x_windows, mask=self.attn_mask + ) # nW*B, window_size*window_size, C # merge windows attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C) @@ -270,7 +353,9 @@ def forward(self, x): # reverse cyclic shift if self.shift_size > 0: - x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2)) + x = torch.roll( + shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2) + ) else: x = shifted_x x = x.view(B, H * W, C) @@ -282,8 +367,10 @@ def forward(self, x): return x def extra_repr(self) -> str: - return f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " \ - f"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}" + return ( + f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " + f"window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}" + ) def flops(self): flops = 0 @@ -301,7 +388,7 @@ def flops(self): class PatchMerging(nn.Module): - r""" Patch Merging Layer. + r"""Patch Merging Layer. Args: input_resolution (tuple[int]): Resolution of input feature. @@ -350,7 +437,7 @@ def flops(self): class BasicLayer(nn.Module): - """ A basic Swin Transformer layer for one stage. + """A basic Swin Transformer layer for one stage. Args: dim (int): Number of input channels. @@ -369,9 +456,23 @@ class BasicLayer(nn.Module): use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False. """ - def __init__(self, dim, input_resolution, depth, num_heads, window_size, - mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., - drop_path=0., norm_layer=nn.LayerNorm, downsample=None, use_checkpoint=False): + def __init__( + self, + dim, + input_resolution, + depth, + num_heads, + window_size, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop=0.0, + attn_drop=0.0, + drop_path=0.0, + norm_layer=nn.LayerNorm, + downsample=None, + use_checkpoint=False, + ): super().__init__() self.dim = dim @@ -380,20 +481,33 @@ def __init__(self, dim, input_resolution, depth, num_heads, window_size, self.use_checkpoint = use_checkpoint # build blocks - self.blocks = nn.ModuleList([ - SwinTransformerBlock(dim=dim, input_resolution=input_resolution, - num_heads=num_heads, window_size=window_size, - shift_size=0 if (i % 2 == 0) else window_size // 2, - mlp_ratio=mlp_ratio, - qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop, attn_drop=attn_drop, - drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path, - norm_layer=norm_layer) - for i in range(depth)]) + self.blocks = nn.ModuleList( + [ + SwinTransformerBlock( + dim=dim, + input_resolution=input_resolution, + num_heads=num_heads, + window_size=window_size, + shift_size=0 if (i % 2 == 0) else window_size // 2, + mlp_ratio=mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop, + attn_drop=attn_drop, + drop_path=drop_path[i] + if isinstance(drop_path, list) + else drop_path, + norm_layer=norm_layer, + ) + for i in range(depth) + ] + ) # patch merging layer if downsample is not None: - self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer) + self.downsample = downsample( + input_resolution, dim=dim, norm_layer=norm_layer + ) else: self.downsample = None @@ -421,7 +535,7 @@ def flops(self): class PatchEmbed(nn.Module): - r""" Image to Patch Embedding + r"""Image to Patch Embedding Args: img_size (int): Image size. Default: 224. @@ -431,11 +545,16 @@ class PatchEmbed(nn.Module): norm_layer (nn.Module, optional): Normalization layer. Default: None """ - def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None): + def __init__( + self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None + ): super().__init__() img_size = to_2tuple(img_size) patch_size = to_2tuple(patch_size) - patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]] + patches_resolution = [ + img_size[0] // patch_size[0], + img_size[1] // patch_size[1], + ] self.img_size = img_size self.patch_size = patch_size self.patches_resolution = patches_resolution @@ -444,7 +563,9 @@ def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_la self.in_chans = in_chans self.embed_dim = embed_dim - self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size) + self.proj = nn.Conv2d( + in_chans, embed_dim, kernel_size=patch_size, stride=patch_size + ) if norm_layer is not None: self.norm = norm_layer(embed_dim) else: @@ -453,8 +574,9 @@ def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_la def forward(self, x): B, C, H, W = x.shape # FIXME look at relaxing size constraints - assert H == self.img_size[0] and W == self.img_size[1], \ - f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})." + assert ( + H == self.img_size[0] and W == self.img_size[1] + ), f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})." x = self.proj(x).flatten(2).transpose(1, 2) # B Ph*Pw C if self.norm is not None: x = self.norm(x) @@ -462,14 +584,20 @@ def forward(self, x): def flops(self): Ho, Wo = self.patches_resolution - flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1]) + flops = ( + Ho + * Wo + * self.embed_dim + * self.in_chans + * (self.patch_size[0] * self.patch_size[1]) + ) if self.norm is not None: flops += Ho * Wo * self.embed_dim return flops class SwinTransformer(nn.Module): - r""" Swin Transformer + r"""Swin Transformer A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows` - https://arxiv.org/pdf/2103.14030 @@ -493,12 +621,28 @@ class SwinTransformer(nn.Module): use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False """ - def __init__(self, img_size=224, patch_size=4, in_chans=3, - embed_dim=96, depths=[2, 2, 6, 2], num_heads=[3, 6, 12, 24], - window_size=7, mlp_ratio=4., qkv_bias=True, qk_scale=None, - drop_rate=0., attn_drop_rate=0., drop_path_rate=0.3, - norm_layer=partial(nn.LayerNorm, eps=1e-6), ape=False, patch_norm=True, - use_checkpoint=False, output_avg=False, **kwargs): + def __init__( + self, + img_size=224, + patch_size=4, + in_chans=3, + embed_dim=96, + depths=[2, 2, 6, 2], + num_heads=[3, 6, 12, 24], + window_size=7, + mlp_ratio=4.0, + qkv_bias=True, + qk_scale=None, + drop_rate=0.0, + attn_drop_rate=0.0, + drop_path_rate=0.3, + norm_layer=partial(nn.LayerNorm, eps=1e-6), + ape=False, + patch_norm=True, + use_checkpoint=False, + output_avg=False, + **kwargs, + ): super().__init__() self.num_layers = len(depths) @@ -510,48 +654,62 @@ def __init__(self, img_size=224, patch_size=4, in_chans=3, self.img_size = to_2tuple(img_size) self.patch_size = to_2tuple(patch_size) self.output_avg = output_avg - + # split image into non-overlapping patches self.patch_embed = PatchEmbed( - img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim, - norm_layer=norm_layer if self.patch_norm else None) + img_size=img_size, + patch_size=patch_size, + in_chans=in_chans, + embed_dim=embed_dim, + norm_layer=norm_layer if self.patch_norm else None, + ) num_patches = self.patch_embed.num_patches patches_resolution = self.patch_embed.patches_resolution self.patches_resolution = patches_resolution # absolute position embedding if self.ape: - self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim)) - trunc_normal_(self.absolute_pos_embed, std=.02) + self.absolute_pos_embed = nn.Parameter( + torch.zeros(1, num_patches, embed_dim) + ) + trunc_normal_(self.absolute_pos_embed, std=0.02) self.pos_drop = nn.Dropout(p=drop_rate) # stochastic depth - dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] # stochastic depth decay rule + dpr = [ + x.item() for x in torch.linspace(0, drop_path_rate, sum(depths)) + ] # stochastic depth decay rule # build layers self.layers = nn.ModuleList() for i_layer in range(self.num_layers): - layer = BasicLayer(dim=int(embed_dim * 2 ** i_layer), - input_resolution=(patches_resolution[0] // (2 ** i_layer), - patches_resolution[1] // (2 ** i_layer)), - depth=depths[i_layer], - num_heads=num_heads[i_layer], - window_size=window_size, - mlp_ratio=self.mlp_ratio, - qkv_bias=qkv_bias, qk_scale=qk_scale, - drop=drop_rate, attn_drop=attn_drop_rate, - drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])], - norm_layer=norm_layer, - downsample=PatchMerging if (i_layer < self.num_layers - 1) else None, - use_checkpoint=use_checkpoint) + layer = BasicLayer( + dim=int(embed_dim * 2**i_layer), + input_resolution=( + patches_resolution[0] // (2**i_layer), + patches_resolution[1] // (2**i_layer), + ), + depth=depths[i_layer], + num_heads=num_heads[i_layer], + window_size=window_size, + mlp_ratio=self.mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])], + norm_layer=norm_layer, + downsample=PatchMerging if (i_layer < self.num_layers - 1) else None, + use_checkpoint=use_checkpoint, + ) self.layers.append(layer) self.apply(self._init_weights) def _init_weights(self, m): if isinstance(m, nn.Linear): - trunc_normal_(m.weight, std=.02) + trunc_normal_(m.weight, std=0.02) if isinstance(m, nn.Linear) and m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): @@ -560,11 +718,11 @@ def _init_weights(self, m): @torch.jit.ignore def no_weight_decay(self): - return {'absolute_pos_embed'} + return {"absolute_pos_embed"} @torch.jit.ignore def no_weight_decay_keywords(self): - return {'relative_position_bias_table'} + return {"relative_position_bias_table"} def forward(self, x): x = self.patch_embed(x) @@ -584,7 +742,7 @@ def forward(self, x): px = px.permute(0, 2, 1).contiguous() px = px.reshape(b, c, h, w) # is this a fair assumption ?? i think it's baked into the architecture - h, w = h//2, w//2 + h, w = h // 2, w // 2 outs.append(px) if self.output_avg: @@ -597,7 +755,12 @@ def flops(self): flops += self.patch_embed.flops() for i, layer in enumerate(self.layers): flops += layer.flops() - flops += self.num_features * self.patches_resolution[0] * self.patches_resolution[1] // (2 ** self.num_layers) + flops += ( + self.num_features + * self.patches_resolution[0] + * self.patches_resolution[1] + // (2**self.num_layers) + ) flops += self.num_features * self.num_classes return flops @@ -610,7 +773,10 @@ def get_swin(drop_path_rate=0.3, output_avg=False): depths = [2, 2, 18, 2] num_heads = [4, 8, 16, 32] swin = SwinTransformer( - img_size=(args.img_h, args.img_w,), + img_size=( + args.img_h, + args.img_w, + ), in_chans=3, patch_size=args.patch_dim, embed_dim=embed_dim, @@ -622,4 +788,3 @@ def get_swin(drop_path_rate=0.3, output_avg=False): ) return swin - diff --git a/megatron/model/vision/utils.py b/megatron/model/vision/utils.py index b4068912c8b..d40b5e5fa7c 100644 --- a/megatron/model/vision/utils.py +++ b/megatron/model/vision/utils.py @@ -3,25 +3,30 @@ import torch.nn.functional as F -def resize(input, - size=None, - scale_factor=None, - mode='nearest', - align_corners=None, - warning=True): +def resize( + input, + size=None, + scale_factor=None, + mode="nearest", + align_corners=None, + warning=True, +): if warning: if size is not None and align_corners: input_h, input_w = tuple(int(x) for x in input.shape[2:]) output_h, output_w = tuple(int(x) for x in size) if output_h > input_h or output_w > output_h: - if ((output_h > 1 and output_w > 1 and input_h > 1 - and input_w > 1) and (output_h - 1) % (input_h - 1) - and (output_w - 1) % (input_w - 1)): + if ( + (output_h > 1 and output_w > 1 and input_h > 1 and input_w > 1) + and (output_h - 1) % (input_h - 1) + and (output_w - 1) % (input_w - 1) + ): warnings.warn( - f'When align_corners={align_corners}, ' - 'the output would more aligned if ' - f'input size {(input_h, input_w)} is `x+1` and ' - f'out size {(output_h, output_w)} is `nx+1`') + f"When align_corners={align_corners}, " + "the output would more aligned if " + f"input size {(input_h, input_w)} is `x+1` and " + f"out size {(output_h, output_w)} is `nx+1`" + ) if isinstance(size, torch.Size): size = tuple(int(x) for x in size) return F.interpolate(input, size, scale_factor, mode, align_corners) diff --git a/megatron/model/vision/vit_backbone.py b/megatron/model/vision/vit_backbone.py index 1efef9c17a8..cadeb8faccb 100644 --- a/megatron/model/vision/vit_backbone.py +++ b/megatron/model/vision/vit_backbone.py @@ -18,6 +18,7 @@ CLASS_TOKEN_LENGTH = 8 + class VitMlpHead(MegatronModule): """Pooler layer. @@ -47,9 +48,9 @@ def forward(self, hidden_states): def isPerfectSquare(x): - if(x >= 0): + if x >= 0: sr = math.sqrt(x) - return (int(sr) * int(sr) == x) + return int(sr) * int(sr) == x return False @@ -76,9 +77,15 @@ def twod_interpolate_position_embeddings_hook( input_param = state_dict[key] input_seq_len = input_param.shape[0] - assert(isPerfectSquare(input_seq_len) or isPerfectSquare(input_seq_len - CLASS_TOKEN_LENGTH)) + assert isPerfectSquare(input_seq_len) or isPerfectSquare( + input_seq_len - CLASS_TOKEN_LENGTH + ) input_has_class_token = not isPerfectSquare(input_seq_len) - num_tok_input = input_seq_len - CLASS_TOKEN_LENGTH if input_has_class_token else input_seq_len + num_tok_input = ( + input_seq_len - CLASS_TOKEN_LENGTH + if input_has_class_token + else input_seq_len + ) num_tok_output = num_patches output_has_class_token = args.class_token_present @@ -98,9 +105,7 @@ def twod_interpolate_position_embeddings_hook( gs_new = (num_patches_per_dim_h, num_patches_per_dim_w) input_param_grid = input_param_grid.transpose(0, 1).contiguous() - input_param_grid = input_param_grid.reshape( - (1, -1, gs_input, gs_input) - ) + input_param_grid = input_param_grid.reshape((1, -1, gs_input, gs_input)) input_param_grid = input_param_grid.float() scale_factor = (gs_new[0] / gs_input, gs_new[1] / gs_input) @@ -129,14 +134,16 @@ def twod_interpolate_position_embeddings_hook( class VitBackbone(MegatronModule): """Vision Transformer Model.""" - def __init__(self, - config, - pre_process=True, - post_process=True, - class_token=True, - single_token_output=False, - post_layer_norm=True, - drop_path_rate=0.0): + def __init__( + self, + config, + pre_process=True, + post_process=True, + class_token=True, + single_token_output=False, + post_layer_norm=True, + drop_path_rate=0.0, + ): super(VitBackbone, self).__init__(share_embeddings_and_output_weights=False) args = get_args() @@ -159,7 +166,9 @@ def __init__(self, self.num_patches_per_dim_h = self.img_h // self.patch_dim self.num_patches_per_dim_w = self.img_w // self.patch_dim self.num_patches = self.num_patches_per_dim_h * self.num_patches_per_dim_w - self.seq_length = self.num_patches + (CLASS_TOKEN_LENGTH if self.class_token else 0) + self.seq_length = self.num_patches + ( + CLASS_TOKEN_LENGTH if self.class_token else 0 + ) self.flatten_dim = self.patch_dim * self.patch_dim * args.num_channels self.input_tensor = None self.position_ids = None @@ -172,19 +181,15 @@ def __init__(self, ) torch.nn.init.zeros_(self.cls_token) self.position_ids = torch.arange(self.seq_length).expand(1, -1).cuda() - + # Linear encoder - self.linear_encoder = torch.nn.Linear( - self.flatten_dim, self.hidden_size - ) + self.linear_encoder = torch.nn.Linear(self.flatten_dim, self.hidden_size) # embedding self.position_embeddings = torch.nn.Embedding( self.seq_length, self.hidden_size ) - init_method_normal(args.init_method_std)( - self.position_embeddings.weight - ) + init_method_normal(args.init_method_std)(self.position_embeddings.weight) args.class_token_present = self.class_token self.position_embeddings._register_load_state_dict_pre_hook( @@ -199,7 +204,7 @@ def __init__(self, pre_process=self.pre_process, post_process=self.post_process, post_layer_norm=self.post_layer_norm, - drop_path_rate=self.drop_path_rate + drop_path_rate=self.drop_path_rate, ) def set_input_tensor(self, input_tensor): @@ -224,8 +229,9 @@ def forward(self, input): cls_tokens = self.cls_token.expand(encoder_output.shape[0], -1, -1) concatenated_tokens = torch.cat((cls_tokens, encoder_output), dim=1) - token_embeddings = concatenated_tokens + \ - self.position_embeddings(self.position_ids[:, :concatenated_tokens.shape[1]]) + token_embeddings = concatenated_tokens + self.position_embeddings( + self.position_ids[:, : concatenated_tokens.shape[1]] + ) # [b, s, h] => [s, b, h] token_embeddings = token_embeddings.transpose(0, 1).contiguous() hidden_states = self.embedding_dropout(token_embeddings) @@ -242,4 +248,3 @@ def forward(self, input): hidden_states = hidden_states.transpose(0, 1).contiguous() return hidden_states - From 40e942afa4fc56c58748d88d27ba4fdb366959b3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:15:07 -0500 Subject: [PATCH 438/590] chore: Format `megatron/core/*` --- megatron/core/__init__.py | 7 +- megatron/core/enums.py | 1 + megatron/core/fusions/fused_bias_dropout.py | 12 +- megatron/core/fusions/fused_bias_gelu.py | 13 +- megatron/core/fusions/fused_layer_norm.py | 94 ++- megatron/core/fusions/fused_softmax.py | 6 +- megatron/core/model_parallel_config.py | 17 +- megatron/core/models/gpt/gpt_embedding.py | 30 +- megatron/core/models/gpt/gpt_model.py | 34 +- megatron/core/package_info.py | 24 +- megatron/core/parallel_state.py | 213 ++++-- .../deepspeed_zbh1_engine.py | 58 +- .../deepspeed_zbh1_schedule.py | 20 +- .../pipeline_parallel/p2p_communication.py | 322 ++++---- megatron/core/pipeline_parallel/schedules.py | 719 +++++++++++------- .../core/sequence_parallel/cross_entropy.py | 45 +- megatron/core/tensor_parallel/__init__.py | 7 +- .../core/tensor_parallel/cross_entropy.py | 41 +- megatron/core/tensor_parallel/data.py | 23 +- megatron/core/tensor_parallel/layers.py | 460 +++++++---- megatron/core/tensor_parallel/mappings.py | 61 +- megatron/core/tensor_parallel/random.py | 109 ++- megatron/core/tensor_parallel/utils.py | 94 ++- .../core/tensor_parallel/weight_grad_store.py | 3 +- megatron/core/transformer/attention.py | 104 ++- megatron/core/transformer/core_attention.py | 61 +- .../custom_layers/transformer_engine.py | 75 +- megatron/core/transformer/mlp.py | 6 +- megatron/core/transformer/module.py | 19 +- .../core/transformer/transformer_block.py | 50 +- .../core/transformer/transformer_config.py | 212 +++--- .../core/transformer/transformer_layer.py | 20 +- megatron/core/transformer/utils.py | 22 +- megatron/core/utils.py | 78 +- 34 files changed, 1923 insertions(+), 1137 deletions(-) diff --git a/megatron/core/__init__.py b/megatron/core/__init__.py index a10f04d164e..010adce64c8 100644 --- a/megatron/core/__init__.py +++ b/megatron/core/__init__.py @@ -7,9 +7,4 @@ # Alias parallel_state as mpu, its legacy name mpu = parallel_state -__all__ = [ - "parallel_state", - "tensor_parallel", - "utils", - "ModelParallelConfig" -] +__all__ = ["parallel_state", "tensor_parallel", "utils", "ModelParallelConfig"] diff --git a/megatron/core/enums.py b/megatron/core/enums.py index cf1452b23e4..46e7d3b766a 100644 --- a/megatron/core/enums.py +++ b/megatron/core/enums.py @@ -2,6 +2,7 @@ import enum + class ModelType(enum.Enum): encoder_or_decoder = 1 encoder_and_decoder = 2 diff --git a/megatron/core/fusions/fused_bias_dropout.py b/megatron/core/fusions/fused_bias_dropout.py index 5c0d49c972c..a790e392490 100644 --- a/megatron/core/fusions/fused_bias_dropout.py +++ b/megatron/core/fusions/fused_bias_dropout.py @@ -3,6 +3,7 @@ import torch from typing import Tuple, Optional + def _bias_dropout_add_func(x, bias, residual, prob, training): # type: (Tensor, Optional[Tensor], Tensor, float, bool) -> Tensor # NOTE: Previously, the argument `bias` used to be passed as @@ -16,28 +17,29 @@ def _bias_dropout_add_func(x, bias, residual, prob, training): out = residual + out return out + def get_bias_dropout_add(training, fused): def unfused_bias_dropout_add(x_with_bias, residual, prob): - x, bias = x_with_bias # unpack + x, bias = x_with_bias # unpack return _bias_dropout_add_func(x, bias, residual, prob, training) @torch.jit.script def bias_dropout_add_fused_train( x_with_bias: Tuple[torch.Tensor, Optional[torch.Tensor]], residual: torch.Tensor, - prob: float + prob: float, ) -> torch.Tensor: - x, bias = x_with_bias # unpack + x, bias = x_with_bias # unpack return _bias_dropout_add_func(x, bias, residual, prob, True) @torch.jit.script def bias_dropout_add_fused_inference( x_with_bias: Tuple[torch.Tensor, Optional[torch.Tensor]], residual: torch.Tensor, - prob: float + prob: float, ) -> torch.Tensor: - x, bias = x_with_bias # unpack + x, bias = x_with_bias # unpack return _bias_dropout_add_func(x, bias, residual, prob, False) if fused: diff --git a/megatron/core/fusions/fused_bias_gelu.py b/megatron/core/fusions/fused_bias_gelu.py index 29222db024e..ac632705b46 100644 --- a/megatron/core/fusions/fused_bias_gelu.py +++ b/megatron/core/fusions/fused_bias_gelu.py @@ -11,10 +11,12 @@ # actual gelu is: # x * 0.5 * (1.0 + torch.erf(x * 0.70710678)) + @torch.jit.script def bias_gelu(bias, y): x = bias + y - return x * 0.5 * (1.0 + torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x))) + return x * 0.5 * (1.0 + torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x))) + # gradient of tanh approximation of gelu # gradient of actual gelu is: @@ -24,10 +26,14 @@ def bias_gelu_back(g, bias, y): x = bias + y tanh_out = torch.tanh(0.79788456 * x * (1 + 0.044715 * x * x)) # sqrt(2/pi) * 3 * 0.044715 -> 0.1070322243 - ff = 0.5 * x * ((1 - tanh_out * tanh_out) * (0.79788456 + 0.1070322243 * x * x)) + 0.5 * (1 + tanh_out) - return ff*g + ff = 0.5 * x * ( + (1 - tanh_out * tanh_out) * (0.79788456 + 0.1070322243 * x * x) + ) + 0.5 * (1 + tanh_out) + return ff * g + class GeLUFunction(torch.autograd.Function): + @staticmethod # bias is an optional argument def forward(ctx, input, bias): @@ -40,4 +46,5 @@ def backward(ctx, grad_output): tmp = bias_gelu_back(grad_output, bias, input) return tmp, tmp + bias_gelu_impl = GeLUFunction.apply diff --git a/megatron/core/fusions/fused_layer_norm.py b/megatron/core/fusions/fused_layer_norm.py index ae0c3b987af..37f7e95109f 100644 --- a/megatron/core/fusions/fused_layer_norm.py +++ b/megatron/core/fusions/fused_layer_norm.py @@ -10,12 +10,14 @@ try: from apex.contrib.layer_norm.layer_norm import FastLayerNormFN + HAVE_PERSIST_LAYER_NORM = True except: HAVE_PERSIST_LAYER_NORM = False try: from apex.normalization.fused_layer_norm import FusedLayerNormAffineFunction + HAVE_FUSED_LAYER_NORM = True except: HAVE_FUSED_LAYER_NORM = False @@ -23,10 +25,14 @@ class FusedLayerNorm(torch.nn.Module): - def __init__(self, hidden_size, eps=1e-5, - persist_layer_norm=True, - sequence_parallel=False, - zero_centered_gamma=False): + def __init__( + self, + hidden_size, + eps=1e-5, + persist_layer_norm=True, + sequence_parallel=False, + zero_centered_gamma=False, + ): super().__init__() self.zero_centered_gamma = zero_centered_gamma @@ -34,15 +40,38 @@ def __init__(self, hidden_size, eps=1e-5, # List of hiddens sizes supported in the persistent layer norm kernel # If the hidden size is not supported, fall back to the non-persistent # kernel. - persist_ln_hidden_sizes = [1024, 1536, 2048, 2304, 3072, 3840, 4096, - 5120, 6144, 8192, 10240, 12288, 12800, 15360, 16384, 18432, 20480, - 24576, 25600, 30720, 32768, 40960, 49152, 65536] + persist_ln_hidden_sizes = [ + 1024, + 1536, + 2048, + 2304, + 3072, + 3840, + 4096, + 5120, + 6144, + 8192, + 10240, + 12288, + 12800, + 15360, + 16384, + 18432, + 20480, + 24576, + 25600, + 30720, + 32768, + 40960, + 49152, + 65536, + ] if hidden_size not in persist_ln_hidden_sizes or not HAVE_PERSIST_LAYER_NORM: persist_layer_norm = False if not persist_layer_norm and not HAVE_FUSED_LAYER_NORM: # TODO: Add pytorch only layer norm - raise ValueError(f'Apex must currently be installed to use megatron core.') + raise ValueError(f"Apex must currently be installed to use megatron core.") if isinstance(hidden_size, numbers.Integral): hidden_size = (hidden_size,) @@ -55,35 +84,36 @@ def __init__(self, hidden_size, eps=1e-5, self.sequence_parallel = sequence_parallel # set sequence parallelism flag on weight and bias parameters - setattr(self.weight, 'sequence_parallel', self.sequence_parallel) - setattr(self.bias, 'sequence_parallel', self.sequence_parallel) - + setattr(self.weight, "sequence_parallel", self.sequence_parallel) + setattr(self.bias, "sequence_parallel", self.sequence_parallel) - def reset_parameters(self): + def reset_parameters(self): - if self.zero_centered_gamma: - init.zeros_(self.weight) - init.zeros_(self.bias) - else: - init.ones_(self.weight) - init.zeros_(self.bias) + if self.zero_centered_gamma: + init.zeros_(self.weight) + init.zeros_(self.bias) + else: + init.ones_(self.weight) + init.zeros_(self.bias) - def forward(self, input): + def forward(self, input): - weight = self.weight + 1 if self.zero_centered_gamma else self.weight + weight = self.weight + 1 if self.zero_centered_gamma else self.weight - if self.persist_layer_norm: - output = FastLayerNormFN.apply(input, weight, self.bias, self.eps) + if self.persist_layer_norm: + output = FastLayerNormFN.apply(input, weight, self.bias, self.eps) - # Apex's fast layer norm function outputs a 'view' tensor (i.e., has - # a populated '_base' field). This will result in schedule.py's - # deallocate_output_tensor() throwing an error, so a viewless tensor is - # created to prevent this. - output = make_viewless_tensor(inp = output, - requires_grad = input.requires_grad, - keep_graph = True) + # Apex's fast layer norm function outputs a 'view' tensor (i.e., has + # a populated '_base' field). This will result in schedule.py's + # deallocate_output_tensor() throwing an error, so a viewless tensor is + # created to prevent this. + output = make_viewless_tensor( + inp=output, requires_grad=input.requires_grad, keep_graph=True + ) - else: - output = FusedLayerNormAffineFunction.apply(input, weight, self.bias, self.hidden_size, self.eps) + else: + output = FusedLayerNormAffineFunction.apply( + input, weight, self.bias, self.hidden_size, self.eps + ) - return output + return output diff --git a/megatron/core/fusions/fused_softmax.py b/megatron/core/fusions/fused_softmax.py index bd31f934d7f..91afb44c224 100644 --- a/megatron/core/fusions/fused_softmax.py +++ b/megatron/core/fusions/fused_softmax.py @@ -81,9 +81,7 @@ def forward(ctx, inputs, scale): scale_t = torch.tensor([scale]) - softmax_results = scaled_softmax_cuda.forward( - inputs, scale_t[0] - ) + softmax_results = scaled_softmax_cuda.forward(inputs, scale_t[0]) ctx.save_for_backward(softmax_results, scale_t) return softmax_results @@ -157,7 +155,7 @@ def is_kernel_available(self, mask, b, np, sq, sk): and self.input_in_float16 # input must be fp16 and 16 < sk <= 4096 # sk must be 16 ~ 2048 and sq % 4 == 0 # sq must be divisor of 4 - and sk % 4 == 0 # sk must be divisor of 4 + and sk % 4 == 0 # sk must be divisor of 4 and attn_batches % 4 == 0 # np * b must be divisor of 4 ): if 0 <= sk <= 4096: diff --git a/megatron/core/model_parallel_config.py b/megatron/core/model_parallel_config.py index 441e5a892d0..f267407fb3f 100644 --- a/megatron/core/model_parallel_config.py +++ b/megatron/core/model_parallel_config.py @@ -5,6 +5,7 @@ import torch + @dataclass class ModelParallelConfig: """Base configuration for Megatron Core @@ -70,7 +71,7 @@ class ModelParallelConfig: enable_autocast (bool): If true runs the forward step function inside torch.autocast context. Default is False. autocast_dtype (torch.dtype): dtype to pass to torch.amp.autocast when enabled. Default is pipeline_dtype. - + variable_seq_lengths (bool, default=False): Support for variable sequence lengths across microbatches. Setting this communicates the size of tensors during pipeline parallelism communication, because of this extra overhead it should only be set if the sequence length varies by microbatch within a global batch. @@ -128,7 +129,7 @@ class ModelParallelConfig: # Optimizations gradient_accumulation_fusion: bool = False async_tensor_model_parallel_allreduce: bool = False - + # Pipeline Parallel pipeline_dtype: torch.dtype = None grad_scale_func: Callable = None @@ -146,19 +147,23 @@ class ModelParallelConfig: param_sync_func: Callable = None def __post_init__(self): - """ Python dataclass method that is used to modify attributes after initialization. - See https://docs.python.org/3/library/dataclasses.html#post-init-processing for more details. + """Python dataclass method that is used to modify attributes after initialization. + See https://docs.python.org/3/library/dataclasses.html#post-init-processing for more details. """ if self.sequence_parallel: if self.tensor_model_parallel_size <= 1: - raise ValueError("Can not use sequence paralllelism without tensor parallelism") + raise ValueError( + "Can not use sequence paralllelism without tensor parallelism" + ) if self.async_tensor_model_parallel_allreduce: # sequence_parallelism already does this async self.async_tensor_model_parallel_allreduce = False if self.pipeline_model_parallel_size > 1: if self.pipeline_dtype is None: - raise ValueError("When using pipeline parallelism, pipeline_dtype must be specified") + raise ValueError( + "When using pipeline parallelism, pipeline_dtype must be specified" + ) if self.autocast_dtype is None: self.autocast_dtype = self.params_dtype diff --git a/megatron/core/models/gpt/gpt_embedding.py b/megatron/core/models/gpt/gpt_embedding.py index 0a06dd719f9..1a53465466a 100644 --- a/megatron/core/models/gpt/gpt_embedding.py +++ b/megatron/core/models/gpt/gpt_embedding.py @@ -19,7 +19,9 @@ class GPTEmbedding(MegatronModule): embedding_dropout_prob float): dropout probability for embeddings """ - def __init__(self, config: TransformerConfig, vocab_size: int, max_sequence_length: int): + def __init__( + self, config: TransformerConfig, vocab_size: int, max_sequence_length: int + ): super().__init__(config=config) self.config: TransformerConfig = config @@ -31,14 +33,16 @@ def __init__(self, config: TransformerConfig, vocab_size: int, max_sequence_leng num_embeddings=self.vocab_size, embedding_dim=self.config.hidden_size, init_method=self.config.init_method, - config=self.config + config=self.config, ) # @jcasper are these keys needed? - self._word_embeddings_key = 'word_embeddings' + self._word_embeddings_key = "word_embeddings" # Position embedding (serial). - self.position_embeddings = torch.nn.Embedding(self.max_sequence_length, self.config.hidden_size) - self._position_embeddings_key = 'position_embeddings' + self.position_embeddings = torch.nn.Embedding( + self.max_sequence_length, self.config.hidden_size + ) + self._position_embeddings_key = "position_embeddings" # Initialize the position embeddings. if self.config.perform_initialization: @@ -77,14 +81,16 @@ def forward(self, input_ids, position_ids): return embeddings - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """For easy load.""" state_dict_ = {} - state_dict_[self._word_embeddings_key] = self.word_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) - state_dict_[self._position_embeddings_key] = self.position_embeddings.state_dict( + state_dict_[self._word_embeddings_key] = self.word_embeddings.state_dict( prefix=prefix, keep_vars=keep_vars ) + state_dict_[self._position_embeddings_key] = ( + self.position_embeddings.state_dict(prefix=prefix, keep_vars=keep_vars) + ) return state_dict_ @@ -98,8 +104,8 @@ def load_state_dict(self, state_dict, strict=True): # for backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if 'word_embeddings' in key: - state_dict_[key.split('word_embeddings.')[1]] = state_dict[key] + if "word_embeddings" in key: + state_dict_[key.split("word_embeddings.")[1]] = state_dict[key] self.word_embeddings.load_state_dict(state_dict_, strict=strict) # Position embedding. @@ -109,6 +115,6 @@ def load_state_dict(self, state_dict, strict=True): # for backward compatibility. state_dict_ = {} for key in state_dict.keys(): - if 'position_embeddings' in key: - state_dict_[key.split('position_embeddings.')[1]] = state_dict[key] + if "position_embeddings" in key: + state_dict_[key.split("position_embeddings.")[1]] = state_dict[key] self.position_embeddings.load_state_dict(state_dict_, strict=strict) diff --git a/megatron/core/models/gpt/gpt_model.py b/megatron/core/models/gpt/gpt_model.py index ae51db69790..17323a88501 100644 --- a/megatron/core/models/gpt/gpt_model.py +++ b/megatron/core/models/gpt/gpt_model.py @@ -60,7 +60,9 @@ def __init__( # Embeddings. if self.pre_process: self.embedding = GPTEmbedding( - config=self.config, vocab_size=self.vocab_size, max_sequence_length=self.max_sequence_length, + config=self.config, + vocab_size=self.vocab_size, + max_sequence_length=self.max_sequence_length, ) # Transformer. @@ -81,20 +83,24 @@ def __init__( bias=False, skip_bias_add=False, gather_output=not self.parallel_output, - skip_weight_param_allocation=self.pre_process and self.share_embeddings_and_output_weights) + skip_weight_param_allocation=self.pre_process + and self.share_embeddings_and_output_weights, + ) - if self.share_embeddings_and_output_weights and (self.pre_process or self.post_process): + if self.share_embeddings_and_output_weights and ( + self.pre_process or self.post_process + ): self.initialize_last_stage_with_word_embeddings() def set_input_tensor(self, input_tensor): - """ See megatron.model.transformer.set_input_tensor()""" + """See megatron.model.transformer.set_input_tensor()""" # This is usually handled in schedules.py but some inference code still # gives us non-lists or None if not isinstance(input_tensor, list): input_tensor = [input_tensor] - assert len(input_tensor) == 1, 'input_tensor should only be length 1 for gpt' + assert len(input_tensor) == 1, "input_tensor should only be length 1 for gpt" self.decoder.set_input_tensor(input_tensor[0]) def forward( @@ -108,7 +114,9 @@ def forward( # Encoder embedding. if self.pre_process: - decoder_input = self.embedding(input_ids=input_ids, position_ids=position_ids) + decoder_input = self.embedding( + input_ids=input_ids, position_ids=position_ids + ) else: # intermediate stage of pipeline # encoder will get hidden_states from encoder.input_tensor @@ -116,7 +124,9 @@ def forward( # Run encoder. hidden_states = self.decoder( - hidden_states=decoder_input, attention_mask=attention_mask, inference_params=inference_params + hidden_states=decoder_input, + attention_mask=attention_mask, + inference_params=inference_params, ) if not self.post_process: @@ -153,7 +163,9 @@ def initialize_last_stage_with_word_embeddings(self): # when we are using pipeline parallelism and sharing word # embeddings. Nothing to do if we aren't sharing weights or aren't using # pipeline parallelism. - if not self.share_embeddings_and_output_weights or (self.pre_process and self.post_process): + if not self.share_embeddings_and_output_weights or ( + self.pre_process and self.post_process + ): return if self.post_process and not self.pre_process: @@ -181,7 +193,9 @@ def initialize_last_stage_with_word_embeddings(self): if torch.distributed.is_initialized(): if parallel_state.is_rank_in_embedding_group(): weight = self.shared_embedding_or_output_weight() - torch.distributed.all_reduce(weight.data, group=parallel_state.get_embedding_group()) + torch.distributed.all_reduce( + weight.data, group=parallel_state.get_embedding_group() + ) elif not getattr(GPTModel, "embedding_warning_printed", False): logging.getLogger(__name__).warning( @@ -194,7 +208,7 @@ def initialize_last_stage_with_word_embeddings(self): GPTModel.embedding_warning_printed = True # TODO: add distributed checkpointing - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): pass # """For easy load.""" diff --git a/megatron/core/package_info.py b/megatron/core/package_info.py index 73fbf09e7bd..3e4faeebe31 100644 --- a/megatron/core/package_info.py +++ b/megatron/core/package_info.py @@ -4,20 +4,20 @@ MAJOR = 0 MINOR = 2 PATCH = 0 -PRE_RELEASE = '' +PRE_RELEASE = "" # Use the following formatting: (major, minor, patch, pre-release) VERSION = (MAJOR, MINOR, PATCH, PRE_RELEASE) -__shortversion__ = '.'.join(map(str, VERSION[:3])) -__version__ = '.'.join(map(str, VERSION[:3])) + ''.join(VERSION[3:]) +__shortversion__ = ".".join(map(str, VERSION[:3])) +__version__ = ".".join(map(str, VERSION[:3])) + "".join(VERSION[3:]) -__package_name__ = 'megatron_core' -__contact_names__ = 'NVIDIA' -__contact_emails__ = 'nemo-toolkit@nvidia.com' # use NeMo Email -__homepage__ = 'https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/stable/' # use NeMo homepage -__repository_url__ = 'https://github.com/NVIDIA/Megatron-LM/megatron/core' -__download_url__ = 'https://github.com/NVIDIA/Megatron-LM/releases' -__description__ = 'Megatron Core - a library for efficient and scalable training of transformer based models' -__license__ = 'BSD-3' -__keywords__ = 'deep learning, machine learning, gpu, NLP, NLU, language, transformer, nvidia, pytorch, torch' +__package_name__ = "megatron_core" +__contact_names__ = "NVIDIA" +__contact_emails__ = "nemo-toolkit@nvidia.com" # use NeMo Email +__homepage__ = "https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/stable/" # use NeMo homepage +__repository_url__ = "https://github.com/NVIDIA/Megatron-LM/megatron/core" +__download_url__ = "https://github.com/NVIDIA/Megatron-LM/releases" +__description__ = "Megatron Core - a library for efficient and scalable training of transformer based models" +__license__ = "BSD-3" +__keywords__ = "deep learning, machine learning, gpu, NLP, NLU, language, transformer, nvidia, pytorch, torch" diff --git a/megatron/core/parallel_state.py b/megatron/core/parallel_state.py index 819760e98bd..cae60de40b5 100644 --- a/megatron/core/parallel_state.py +++ b/megatron/core/parallel_state.py @@ -143,30 +143,42 @@ def initialize_model_parallel( enable_ds_sequence_parallel = sequence_parallel_size > 1 if enable_ds_sequence_parallel: - assert tensor_model_parallel_size == 1 and pipeline_model_parallel_size == 1, \ - 'DeepSpeed\'s sequence parallel does not work with tensor parallel or pipeline parallel' + assert ( + tensor_model_parallel_size == 1 and pipeline_model_parallel_size == 1 + ), "DeepSpeed's sequence parallel does not work with tensor parallel or pipeline parallel" if world_size % sequence_parallel_size != 0: raise RuntimeError( f"world_size ({world_size}) is not divisible by sequence_parallel_size {sequence_parallel_size})" ) - data_parallel_size: int = world_size // (tensor_model_parallel_size * pipeline_model_parallel_size * sequence_parallel_size) + data_parallel_size: int = world_size // ( + tensor_model_parallel_size + * pipeline_model_parallel_size + * sequence_parallel_size + ) sequence_data_parallel_size: int = sequence_parallel_size * data_parallel_size num_tensor_model_parallel_groups: int = world_size // tensor_model_parallel_size num_pipeline_model_parallel_groups: int = world_size // pipeline_model_parallel_size num_data_parallel_groups: int = world_size // data_parallel_size num_sequence_parallel_groups: int = world_size // sequence_parallel_size - num_sequence_data_parallel_groups: int = world_size // sequence_parallel_size // data_parallel_size + num_sequence_data_parallel_groups: int = ( + world_size // sequence_parallel_size // data_parallel_size + ) if virtual_pipeline_model_parallel_size is not None: if not pipeline_model_parallel_size > 2: - raise RuntimeError("pipeline-model-parallel size should be greater than 2 with " "interleaved schedule") + raise RuntimeError( + "pipeline-model-parallel size should be greater than 2 with " + "interleaved schedule" + ) global _VIRTUAL_PIPELINE_MODEL_PARALLEL_RANK global _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE _VIRTUAL_PIPELINE_MODEL_PARALLEL_RANK = 0 - _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE = virtual_pipeline_model_parallel_size + _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE = ( + virtual_pipeline_model_parallel_size + ) if pipeline_model_parallel_split_rank is not None: global _PIPELINE_MODEL_PARALLEL_SPLIT_RANK @@ -178,7 +190,7 @@ def initialize_model_parallel( global _DATA_PARALLEL_GROUP global _DATA_PARALLEL_GROUP_GLOO global _DATA_PARALLEL_GLOBAL_RANKS - assert _DATA_PARALLEL_GROUP is None, 'data parallel group is already initialized' + assert _DATA_PARALLEL_GROUP is None, "data parallel group is already initialized" all_data_parallel_group_ranks = [] for i in range(pipeline_model_parallel_size): start_rank = i * num_pipeline_model_parallel_groups @@ -204,24 +216,26 @@ def initialize_model_parallel( # Build the sequence parallel groups. global _SEQUENCE_PARALLEL_GROUP - assert _SEQUENCE_PARALLEL_GROUP is None, \ - 'sequence parallel group is already initialized' + assert ( + _SEQUENCE_PARALLEL_GROUP is None + ), "sequence parallel group is already initialized" for i in range(num_sequence_parallel_groups): - ranks = range(i * sequence_parallel_size, - (i + 1) * sequence_parallel_size) + ranks = range(i * sequence_parallel_size, (i + 1) * sequence_parallel_size) group = torch.distributed.new_group(ranks) if rank in ranks: _SEQUENCE_PARALLEL_GROUP = group # Build the sequence data parallel groups. global _SEQUENCE_DATA_PARALLEL_GROUP - assert _SEQUENCE_DATA_PARALLEL_GROUP is None, \ - 'sequence data parallel group is already initialized' + assert ( + _SEQUENCE_DATA_PARALLEL_GROUP is None + ), "sequence data parallel group is already initialized" all_data_sequence_parallel_group_ranks = [] if enable_ds_sequence_parallel: for i in range(num_sequence_data_parallel_groups): - ranks = range(i * sequence_data_parallel_size, - (i + 1) * sequence_data_parallel_size) + ranks = range( + i * sequence_data_parallel_size, (i + 1) * sequence_data_parallel_size + ) group = torch.distributed.new_group(ranks) all_data_sequence_parallel_group_ranks.append(list(ranks)) if rank in ranks: @@ -231,36 +245,54 @@ def initialize_model_parallel( # Build the model-parallel groups. global _MODEL_PARALLEL_GROUP - assert _MODEL_PARALLEL_GROUP is None, 'model parallel group is already initialized' - num_model_parallel_groups = sequence_data_parallel_size if enable_ds_sequence_parallel else data_parallel_size - model_parallel_group_ranks = all_data_sequence_parallel_group_ranks if enable_ds_sequence_parallel else all_data_parallel_group_ranks + assert _MODEL_PARALLEL_GROUP is None, "model parallel group is already initialized" + num_model_parallel_groups = ( + sequence_data_parallel_size + if enable_ds_sequence_parallel + else data_parallel_size + ) + model_parallel_group_ranks = ( + all_data_sequence_parallel_group_ranks + if enable_ds_sequence_parallel + else all_data_parallel_group_ranks + ) for i in range(num_model_parallel_groups): - ranks = [parallel_group_ranks[i] for parallel_group_ranks in model_parallel_group_ranks] + ranks = [ + parallel_group_ranks[i] + for parallel_group_ranks in model_parallel_group_ranks + ] group = torch.distributed.new_group(ranks) if rank in ranks: _MODEL_PARALLEL_GROUP = group # Build the tensor model-parallel groups. global _TENSOR_MODEL_PARALLEL_GROUP - assert _TENSOR_MODEL_PARALLEL_GROUP is None, 'tensor model parallel group is already initialized' + assert ( + _TENSOR_MODEL_PARALLEL_GROUP is None + ), "tensor model parallel group is already initialized" for i in range(num_tensor_model_parallel_groups): - ranks = range(i * tensor_model_parallel_size, (i + 1) * tensor_model_parallel_size) + ranks = range( + i * tensor_model_parallel_size, (i + 1) * tensor_model_parallel_size + ) group = torch.distributed.new_group(ranks) if rank in ranks: _TENSOR_MODEL_PARALLEL_GROUP = group - # Build the pipeline model-parallel groups and embedding groups # (first and last rank in each pipeline model-parallel group). global _PIPELINE_MODEL_PARALLEL_GROUP global _PIPELINE_GLOBAL_RANKS - assert _PIPELINE_MODEL_PARALLEL_GROUP is None, 'pipeline model parallel group is already initialized' + assert ( + _PIPELINE_MODEL_PARALLEL_GROUP is None + ), "pipeline model parallel group is already initialized" global _EMBEDDING_GROUP global _EMBEDDING_GLOBAL_RANKS - assert _EMBEDDING_GROUP is None, 'embedding group is already initialized' + assert _EMBEDDING_GROUP is None, "embedding group is already initialized" global _POSITION_EMBEDDING_GROUP global _POSITION_EMBEDDING_GLOBAL_RANKS - assert _POSITION_EMBEDDING_GROUP is None, 'position embedding group is already initialized' + assert ( + _POSITION_EMBEDDING_GROUP is None + ), "position embedding group is already initialized" for i in range(num_pipeline_model_parallel_groups): ranks = range(i, world_size, num_pipeline_model_parallel_groups) group = torch.distributed.new_group(ranks) @@ -274,9 +306,19 @@ def initialize_model_parallel( position_embedding_ranks = [ranks[0]] if pipeline_model_parallel_split_rank is not None: if ranks[pipeline_model_parallel_split_rank] not in embedding_ranks: - embedding_ranks = [ranks[0], ranks[pipeline_model_parallel_split_rank], ranks[-1]] - if ranks[pipeline_model_parallel_split_rank] not in position_embedding_ranks: - position_embedding_ranks = [ranks[0], ranks[pipeline_model_parallel_split_rank]] + embedding_ranks = [ + ranks[0], + ranks[pipeline_model_parallel_split_rank], + ranks[-1], + ] + if ( + ranks[pipeline_model_parallel_split_rank] + not in position_embedding_ranks + ): + position_embedding_ranks = [ + ranks[0], + ranks[pipeline_model_parallel_split_rank], + ] else: embedding_ranks = ranks position_embedding_ranks = ranks @@ -295,8 +337,9 @@ def initialize_model_parallel( # Build the FP8 groups. global _AMAX_REDUCTION_GROUP - assert _AMAX_REDUCTION_GROUP is None, \ - 'FP8 amax reduction group is already initialized' + assert ( + _AMAX_REDUCTION_GROUP is None + ), "FP8 amax reduction group is already initialized" if use_fp8: amax_group_size: int = tensor_model_parallel_size * data_parallel_size num_amax_groups: int = world_size // amax_group_size @@ -322,84 +365,101 @@ def is_unitialized(): def model_parallel_is_initialized(): """Check if model and data parallel groups are initialized.""" - if _TENSOR_MODEL_PARALLEL_GROUP is None or _PIPELINE_MODEL_PARALLEL_GROUP is None or _DATA_PARALLEL_GROUP is None: + if ( + _TENSOR_MODEL_PARALLEL_GROUP is None + or _PIPELINE_MODEL_PARALLEL_GROUP is None + or _DATA_PARALLEL_GROUP is None + ): return False return True + def sequence_parallel_is_initialized(): """Check if sequence and data parallel groups are initialized.""" - if _SEQUENCE_PARALLEL_GROUP is None or \ - _DATA_PARALLEL_GROUP is None: + if _SEQUENCE_PARALLEL_GROUP is None or _DATA_PARALLEL_GROUP is None: return False return True + def sequence_data_parallel_is_initialized(): """Check if sequence data parallel groups are initialized.""" if _SEQUENCE_DATA_PARALLEL_GROUP is None: return False return True + def get_model_parallel_group(): """Get the model parallel group the caller rank belongs to.""" - assert _MODEL_PARALLEL_GROUP is not None, 'model parallel group is not initialized' + assert _MODEL_PARALLEL_GROUP is not None, "model parallel group is not initialized" return _MODEL_PARALLEL_GROUP def get_tensor_model_parallel_group(check_initialized=True): """Get the tensor model parallel group the caller rank belongs to.""" if check_initialized: - assert _TENSOR_MODEL_PARALLEL_GROUP is not None, 'tensor model parallel group is not initialized' + assert ( + _TENSOR_MODEL_PARALLEL_GROUP is not None + ), "tensor model parallel group is not initialized" return _TENSOR_MODEL_PARALLEL_GROUP def get_pipeline_model_parallel_group(): """Get the pipeline model parallel group the caller rank belongs to.""" - assert _PIPELINE_MODEL_PARALLEL_GROUP is not None, 'pipeline_model parallel group is not initialized' + assert ( + _PIPELINE_MODEL_PARALLEL_GROUP is not None + ), "pipeline_model parallel group is not initialized" return _PIPELINE_MODEL_PARALLEL_GROUP + def get_sequence_parallel_group(): """Get the sequence parallel group the caller rank belongs to.""" - assert _SEQUENCE_PARALLEL_GROUP is not None, \ - 'sequence parallel group is not initialized' + assert ( + _SEQUENCE_PARALLEL_GROUP is not None + ), "sequence parallel group is not initialized" return _SEQUENCE_PARALLEL_GROUP def get_sequence_data_parallel_group(): """Get the sequence parallel group the caller rank belongs to.""" - assert _SEQUENCE_DATA_PARALLEL_GROUP is not None, \ - 'sequence data parallel group is not initialized' + assert ( + _SEQUENCE_DATA_PARALLEL_GROUP is not None + ), "sequence data parallel group is not initialized" return _SEQUENCE_DATA_PARALLEL_GROUP def get_data_parallel_group(): """Get the data parallel group the caller rank belongs to.""" - assert _DATA_PARALLEL_GROUP is not None, 'data parallel group is not initialized' + assert _DATA_PARALLEL_GROUP is not None, "data parallel group is not initialized" return _DATA_PARALLEL_GROUP def get_data_parallel_group_gloo(): """Get the data parallel group-gloo the caller rank belongs to.""" - assert _DATA_PARALLEL_GROUP_GLOO is not None, \ - 'data parallel group-gloo is not initialized' + assert ( + _DATA_PARALLEL_GROUP_GLOO is not None + ), "data parallel group-gloo is not initialized" return _DATA_PARALLEL_GROUP_GLOO def get_embedding_group(): """Get the embedding group the caller rank belongs to.""" - assert _EMBEDDING_GROUP is not None, 'embedding group is not initialized' + assert _EMBEDDING_GROUP is not None, "embedding group is not initialized" return _EMBEDDING_GROUP def get_position_embedding_group(): """Get the position embedding group the caller rank belongs to.""" - assert _POSITION_EMBEDDING_GROUP is not None, 'position embedding group is not initialized' + assert ( + _POSITION_EMBEDDING_GROUP is not None + ), "position embedding group is not initialized" return _POSITION_EMBEDDING_GROUP def get_amax_reduction_group(): """Get the FP8 amax reduction group the caller rank belongs to.""" - assert _AMAX_REDUCTION_GROUP is not None, \ - 'FP8 amax reduction group is not initialized' + assert ( + _AMAX_REDUCTION_GROUP is not None + ), "FP8 amax reduction group is not initialized" return _AMAX_REDUCTION_GROUP @@ -408,26 +468,31 @@ def set_tensor_model_parallel_world_size(world_size): global _MPU_TENSOR_MODEL_PARALLEL_WORLD_SIZE _MPU_TENSOR_MODEL_PARALLEL_WORLD_SIZE = world_size + def set_sequence_parallel_world_size(world_size): """Set the sequence parallel size""" global _SEQUENCE_PARALLEL_WORLD_SIZE _SEQUENCE_PARALLEL_WORLD_SIZE = world_size + def set_sequence_data_parallel_world_size(world_size): """Set the sequence parallel size""" global _SEQUENCE_DATA_PARALLEL_WORLD_SIZE _SEQUENCE_DATA_PARALLEL_WORLD_SIZE = world_size + def set_pipeline_model_parallel_world_size(world_size): """Set the pipeline model parallel size""" global _MPU_PIPELINE_MODEL_PARALLEL_WORLD_SIZE _MPU_PIPELINE_MODEL_PARALLEL_WORLD_SIZE = world_size + def set_virtual_pipeline_model_parallel_world_size(world_size): """Set the pipeline model parallel size""" global _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE = world_size + def set_virtual_pipeline_model_parallel_world_size(world_size): """Set the virtual pipeline model parallel size""" global _VIRTUAL_PIPELINE_MODEL_PARALLEL_WORLD_SIZE @@ -441,10 +506,14 @@ def get_tensor_model_parallel_world_size(): return _MPU_TENSOR_MODEL_PARALLEL_WORLD_SIZE return torch.distributed.get_world_size(group=get_tensor_model_parallel_group()) + def get_model_parallel_world_size(): - assert get_pipeline_model_parallel_world_size() == 1, "legacy get_model_parallel_world_size is only supported if PP is disabled" + assert ( + get_pipeline_model_parallel_world_size() == 1 + ), "legacy get_model_parallel_world_size is only supported if PP is disabled" return get_tensor_model_parallel_world_size() + def get_sequence_parallel_world_size(): """Return world size for the sequence parallel group.""" global _SEQUENCE_PARALLEL_WORLD_SIZE @@ -452,6 +521,7 @@ def get_sequence_parallel_world_size(): return _SEQUENCE_PARALLEL_WORLD_SIZE return torch.distributed.get_world_size(group=get_sequence_parallel_group()) + def get_sequence_data_parallel_world_size(): """Return world size for the sequence parallel group.""" global _SEQUENCE_DATA_PARALLEL_WORLD_SIZE @@ -459,6 +529,7 @@ def get_sequence_data_parallel_world_size(): return _SEQUENCE_DATA_PARALLEL_WORLD_SIZE return torch.distributed.get_world_size(group=get_sequence_data_parallel_group()) + def get_pipeline_model_parallel_world_size(): """Return world size for the pipeline model parallel group.""" global _MPU_PIPELINE_MODEL_PARALLEL_WORLD_SIZE @@ -474,7 +545,9 @@ def set_tensor_model_parallel_rank(rank): def get_model_parallel_rank(): - assert get_pipeline_model_parallel_world_size() == 1, "legacy get_model_parallel_rank is only supported if PP is disabled" + assert ( + get_pipeline_model_parallel_world_size() == 1 + ), "legacy get_model_parallel_rank is only supported if PP is disabled" return get_tensor_model_parallel_rank() @@ -554,12 +627,18 @@ def is_pipeline_first_stage(ignore_virtual=False): def is_pipeline_last_stage(ignore_virtual=False): """Return True if in the last pipeline model-parallel stage, False otherwise.""" if not ignore_virtual: - virtual_pipeline_model_parallel_world_size = get_virtual_pipeline_model_parallel_world_size() - if virtual_pipeline_model_parallel_world_size is not None and get_virtual_pipeline_model_parallel_rank() != ( - virtual_pipeline_model_parallel_world_size - 1 + virtual_pipeline_model_parallel_world_size = ( + get_virtual_pipeline_model_parallel_world_size() + ) + if ( + virtual_pipeline_model_parallel_world_size is not None + and get_virtual_pipeline_model_parallel_rank() + != (virtual_pipeline_model_parallel_world_size - 1) ): return False - return get_pipeline_model_parallel_rank() == (get_pipeline_model_parallel_world_size() - 1) + return get_pipeline_model_parallel_rank() == ( + get_pipeline_model_parallel_world_size() - 1 + ) def is_rank_in_embedding_group(ignore_virtual=False): @@ -620,7 +699,9 @@ def is_pipeline_stage_at_split(): stage executes encoder block for a model with both encoder and decoder.""" rank = get_pipeline_model_parallel_rank() - return is_pipeline_stage_before_split(rank) and is_pipeline_stage_after_split(rank + 1) + return is_pipeline_stage_before_split(rank) and is_pipeline_stage_after_split( + rank + 1 + ) def get_virtual_pipeline_model_parallel_rank(): @@ -666,28 +747,36 @@ def get_sequence_parallel_src_rank(): def get_data_parallel_src_rank(): """Calculate the global rank corresponding to the first local rank in the data parallel group.""" - assert _DATA_PARALLEL_GLOBAL_RANKS is not None, "Data parallel group is not initialized" + assert ( + _DATA_PARALLEL_GLOBAL_RANKS is not None + ), "Data parallel group is not initialized" return _DATA_PARALLEL_GLOBAL_RANKS[0] def get_pipeline_model_parallel_first_rank(): """Return the global rank of the first process in the pipeline for the current tensor parallel group""" - assert _PIPELINE_GLOBAL_RANKS is not None, "Pipeline parallel group is not initialized" + assert ( + _PIPELINE_GLOBAL_RANKS is not None + ), "Pipeline parallel group is not initialized" return _PIPELINE_GLOBAL_RANKS[0] def get_pipeline_model_parallel_last_rank(): """Return the global rank of the last process in the pipeline for the current tensor parallel group""" - assert _PIPELINE_GLOBAL_RANKS is not None, "Pipeline parallel group is not initialized" + assert ( + _PIPELINE_GLOBAL_RANKS is not None + ), "Pipeline parallel group is not initialized" last_rank_local = get_pipeline_model_parallel_world_size() - 1 return _PIPELINE_GLOBAL_RANKS[last_rank_local] def get_pipeline_model_parallel_next_rank(): """Return the global rank that follows the caller in the pipeline""" - assert _PIPELINE_GLOBAL_RANKS is not None, "Pipeline parallel group is not initialized" + assert ( + _PIPELINE_GLOBAL_RANKS is not None + ), "Pipeline parallel group is not initialized" rank_in_pipeline = get_pipeline_model_parallel_rank() world_size = get_pipeline_model_parallel_world_size() return _PIPELINE_GLOBAL_RANKS[(rank_in_pipeline + 1) % world_size] @@ -695,7 +784,9 @@ def get_pipeline_model_parallel_next_rank(): def get_pipeline_model_parallel_prev_rank(): """Return the global rank that preceeds the caller in the pipeline""" - assert _PIPELINE_GLOBAL_RANKS is not None, "Pipeline parallel group is not initialized" + assert ( + _PIPELINE_GLOBAL_RANKS is not None + ), "Pipeline parallel group is not initialized" rank_in_pipeline = get_pipeline_model_parallel_rank() world_size = get_pipeline_model_parallel_world_size() return _PIPELINE_GLOBAL_RANKS[(rank_in_pipeline - 1) % world_size] @@ -714,13 +805,13 @@ def get_data_parallel_rank(): def _set_global_memory_buffer(): """Initialize global buffer""" global _GLOBAL_MEMORY_BUFFER - assert _GLOBAL_MEMORY_BUFFER is None, 'global memory buffer is already initialized' + assert _GLOBAL_MEMORY_BUFFER is None, "global memory buffer is already initialized" _GLOBAL_MEMORY_BUFFER = GlobalMemoryBuffer() def get_global_memory_buffer(): """Return the global GlobalMemoryBuffer object""" - assert _GLOBAL_MEMORY_BUFFER is not None, 'global memory buffer is not initialized' + assert _GLOBAL_MEMORY_BUFFER is not None, "global memory buffer is not initialized" return _GLOBAL_MEMORY_BUFFER diff --git a/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py b/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py index ba451d97055..f2206380b75 100644 --- a/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py +++ b/megatron/core/pipeline_parallel/deepspeed_zbh1_engine.py @@ -1,8 +1,7 @@ from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore from deepspeed.runtime.pipe.engine import PipelineEngine -from deepspeed.utils.timer import BACKWARD_MICRO_TIMER, \ - BACKWARD_GLOBAL_TIMER, BACKWARD_INNER_MICRO_TIMER, BACKWARD_INNER_GLOBAL_TIMER +from deepspeed.utils.timer import BACKWARD_MICRO_TIMER, BACKWARD_GLOBAL_TIMER, BACKWARD_INNER_MICRO_TIMER, BACKWARD_INNER_GLOBAL_TIMER from deepspeed.runtime.utils import PartitionedTensor from deepspeed.accelerator import get_accelerator @@ -17,24 +16,27 @@ get_global_memory_buffer, ) + def _exec_backward_only_pass(self, buffer_id): - assert self.optimizer is not None, "must provide optimizer during " \ - "init in order to use backward" + assert self.optimizer is not None, ( + "must provide optimizer during " "init in order to use backward" + ) - self.mem_status('BEFORE BWD ONLY', reset_max=True) + self.mem_status("BEFORE BWD ONLY", reset_max=True) from megatron.core.tensor_parallel.layers import LinearWithGradAccumulationAndAsyncCommunication + WeightGradStore.set_combine_bw(False) # The last stage just runs backward on the loss using DeepSpeed's typical # mechanisms. if self.is_last_stage(): super(PipelineEngine, self).backward(self.loss) WeightGradStore.flush() - self.mem_status('AFTER BWD ONLY') + self.mem_status("AFTER BWD ONLY") WeightGradStore.set_combine_bw(True) return - outputs = self.pipe_buffers['outputs'][buffer_id] + outputs = self.pipe_buffers["outputs"][buffer_id] if self.wall_clock_breakdown(): self.timers(BACKWARD_MICRO_TIMER).start() @@ -47,24 +49,28 @@ def _exec_backward_only_pass(self, buffer_id): if self.is_pipe_partitioned: if self.is_grad_partitioned: if self.pipe_partition_output_meta_cache is None: - self.pipe_partition_output_meta_cache = outputs[0].to('cpu') - part_output = PartitionedTensor.from_meta(meta=self.pipe_partition_output_meta_cache, - local_part=outputs[1], - group=self.grid.get_slice_parallel_group()) - self.pipe_buffers['output_tensors'][buffer_id].data = part_output.full() - outputs = (self.pipe_buffers['output_tensors'][buffer_id], *outputs[2:]) + self.pipe_partition_output_meta_cache = outputs[0].to("cpu") + part_output = PartitionedTensor.from_meta( + meta=self.pipe_partition_output_meta_cache, + local_part=outputs[1], + group=self.grid.get_slice_parallel_group(), + ) + self.pipe_buffers["output_tensors"][buffer_id].data = part_output.full() + outputs = (self.pipe_buffers["output_tensors"][buffer_id], *outputs[2:]) else: # Already restored from partition - self.pipe_buffers['output_tensors'][buffer_id].data = outputs[0] - outputs = (self.pipe_buffers['output_tensors'][buffer_id], *outputs[1:]) + self.pipe_buffers["output_tensors"][buffer_id].data = outputs[0] + outputs = (self.pipe_buffers["output_tensors"][buffer_id], *outputs[1:]) grad_tensors = self.grad_layer if self.is_grad_partitioned: if self.grad_partition_grad_layer_meta_cache is None: - self.grad_partition_grad_layer_meta_cache = self.grad_layer[0].to('cpu') - part_grad = PartitionedTensor.from_meta(meta=self.grad_partition_grad_layer_meta_cache, - local_part=self.grad_layer[1], - group=self.grid.get_slice_parallel_group()) + self.grad_partition_grad_layer_meta_cache = self.grad_layer[0].to("cpu") + part_grad = PartitionedTensor.from_meta( + meta=self.grad_partition_grad_layer_meta_cache, + local_part=self.grad_layer[1], + group=self.grid.get_slice_parallel_group(), + ) grad_tensors = (part_grad.full(), *grad_tensors[2:]) part_grad = None @@ -73,14 +79,13 @@ def _exec_backward_only_pass(self, buffer_id): self.optimizer.clear_lp_grads() # This handles either a single tensor or tuple of tensors. - + if isinstance(outputs, tuple): out_tensors = [t for t in outputs if t.is_floating_point()] assert len(out_tensors) == len(grad_tensors) torch.autograd.backward(tensors=out_tensors, grad_tensors=grad_tensors) else: - torch.autograd.backward(tensors=(outputs, ), grad_tensors=(grad_tensors, )) - + torch.autograd.backward(tensors=(outputs,), grad_tensors=(grad_tensors,)) WeightGradStore.flush() @@ -89,10 +94,10 @@ def _exec_backward_only_pass(self, buffer_id): self.optimizer.update_hp_grads(clear_lp_grads=False) # Free up the memory from the output of forward() - self.pipe_buffers['output_tensors'][buffer_id] = None - self.pipe_buffers['outputs'][buffer_id] = None + self.pipe_buffers["output_tensors"][buffer_id] = None + self.pipe_buffers["outputs"][buffer_id] = None grad_tensors = None - + WeightGradStore.set_combine_bw(True) if self.wall_clock_breakdown(): @@ -101,10 +106,11 @@ def _exec_backward_only_pass(self, buffer_id): self.timers(BACKWARD_MICRO_TIMER).stop() self.timers(BACKWARD_GLOBAL_TIMER).stop() + def _exec_weight_pass(self): if self.using_bf16_optimizer: # manually call because we don't call optimizer.backward() self.optimizer.clear_lp_grads() WeightGradStore.pop() if self.using_bf16_optimizer: - self.optimizer.update_hp_grads(clear_lp_grads=False) \ No newline at end of file + self.optimizer.update_hp_grads(clear_lp_grads=False) diff --git a/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py b/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py index 651aadce729..1801f1ae804 100644 --- a/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py +++ b/megatron/core/pipeline_parallel/deepspeed_zbh1_schedule.py @@ -1,8 +1,7 @@ -from deepspeed.runtime.pipe.schedule import PipeSchedule, PipeInstruction, BufferOpInstruction, \ - LoadMicroBatch, RecvActivation, SendActivation, RecvGrad, SendGrad, \ - ForwardPass, BackwardPass, ReduceGrads, ReduceTiedGrads, OptimizerStep +from deepspeed.runtime.pipe.schedule import PipeSchedule, PipeInstruction, BufferOpInstruction, LoadMicroBatch, RecvActivation, SendActivation, RecvGrad, SendGrad, ForwardPass, BackwardPass, ReduceGrads, ReduceTiedGrads, OptimizerStep from megatron import get_args + class ZeroBubbleH1Pipeline(PipeSchedule): """A schedule for training a batch using hybrid parallelism. @@ -57,7 +56,7 @@ def steps(self): if not self.is_first_stage: cmds.append(SendGrad(backward_id)) yield cmds - + # FBW section while forward < self.micro_batches: forward_id = self.get_buffer_id(forward) @@ -84,14 +83,14 @@ def steps(self): else: if get_args().enable_zbh1_exact_semantics: cmds.append(BackwardOnlyPass(backward_id)) - cmds.append(SendGrad(backward_id)) + cmds.append(SendGrad(backward_id)) cmds.append(WeightPass()) else: cmds.append(BackwardPass(backward_id)) cmds.append(SendGrad(backward_id)) yield cmds - #BW section + # BW section while backward < self.micro_batches: backward_id = self.get_buffer_id(backward) backward += 1 @@ -107,12 +106,12 @@ def steps(self): cmds.append(SendGrad(backward_id)) cmds.append(WeightPass()) yield cmds - - #W section + + # W section while weight < self.micro_batches: weight += 1 yield [WeightPass()] - + yield [ReduceTiedGrads(), ReduceGrads(), OptimizerStep()] def get_buffer_id(self, microbatch_id): @@ -133,8 +132,10 @@ class BackwardOnlyPass(BufferOpInstruction): torch.autograd.backward(tensors=outputs, grad_tensors=gradients, inputs = input_tensor) """ + pass + class WeightPass(PipeInstruction): """Compute a weight pass and accumulate gradients. @@ -145,4 +146,5 @@ class WeightPass(PipeInstruction): torch.autograd.backward(tensors=outputs, grad_tensors=gradients, inputs = model.parameters()) """ + pass diff --git a/megatron/core/pipeline_parallel/p2p_communication.py b/megatron/core/pipeline_parallel/p2p_communication.py index 78e43e7fed1..5fda68f688b 100644 --- a/megatron/core/pipeline_parallel/p2p_communication.py +++ b/megatron/core/pipeline_parallel/p2p_communication.py @@ -17,12 +17,15 @@ from megatron.core import ModelParallelConfig from deepspeed.accelerator import get_accelerator from megatron.utils import Profile + dlp = Profile("PIPELINE") # Types Shape = Union[List[int], torch.Size] -def _communicate_shapes(tensor_send_next, tensor_send_prev, - recv_prev, recv_next, config): + +def _communicate_shapes( + tensor_send_next, tensor_send_prev, recv_prev, recv_next, config +): """Communicate tensor shapes between stages. Used to communicate tensor shapes before the actual tensor communication happens. This is required when the sequence lengths across micro batches @@ -46,49 +49,63 @@ def _communicate_shapes(tensor_send_next, tensor_send_prev, send_prev_shape_tensor = None send_next_shape_tensor = None if recv_prev: - recv_prev_shape_tensor = torch.empty((3), - device=get_accelerator().current_device(), - dtype=torch.int64) + recv_prev_shape_tensor = torch.empty( + (3), device=get_accelerator().current_device(), dtype=torch.int64 + ) if recv_next: - recv_next_shape_tensor = torch.empty((3), - device=get_accelerator().current_device(), - dtype=torch.int64) + recv_next_shape_tensor = torch.empty( + (3), device=get_accelerator().current_device(), dtype=torch.int64 + ) if tensor_send_prev is not None: - send_prev_shape_tensor = torch.tensor(tensor_send_prev.size(), - device=get_accelerator().current_device(), - dtype=torch.int64) + send_prev_shape_tensor = torch.tensor( + tensor_send_prev.size(), + device=get_accelerator().current_device(), + dtype=torch.int64, + ) if tensor_send_next is not None: - send_next_shape_tensor = torch.tensor(tensor_send_next.size(), - device=get_accelerator().current_device(), - dtype=torch.int64) + send_next_shape_tensor = torch.tensor( + tensor_send_next.size(), + device=get_accelerator().current_device(), + dtype=torch.int64, + ) if config.use_ring_exchange_p2p: - torch.distributed.ring_exchange(tensor_send_prev=send_prev_shape_tensor, - tensor_recv_prev=recv_prev_shape_tensor, - tensor_send_next=send_next_shape_tensor, - tensor_recv_next=recv_next_shape_tensor, - group=get_pipeline_model_parallel_group()) + torch.distributed.ring_exchange( + tensor_send_prev=send_prev_shape_tensor, + tensor_recv_prev=recv_prev_shape_tensor, + tensor_send_next=send_next_shape_tensor, + tensor_recv_next=recv_next_shape_tensor, + group=get_pipeline_model_parallel_group(), + ) else: ops = [] if send_prev_shape_tensor is not None: send_prev_op = torch.distributed.P2POp( - torch.distributed.isend, send_prev_shape_tensor, - get_pipeline_model_parallel_prev_rank()) + torch.distributed.isend, + send_prev_shape_tensor, + get_pipeline_model_parallel_prev_rank(), + ) ops.append(send_prev_op) if recv_prev_shape_tensor is not None: recv_prev_op = torch.distributed.P2POp( - torch.distributed.irecv, recv_prev_shape_tensor, - get_pipeline_model_parallel_prev_rank()) + torch.distributed.irecv, + recv_prev_shape_tensor, + get_pipeline_model_parallel_prev_rank(), + ) ops.append(recv_prev_op) if send_next_shape_tensor is not None: send_next_op = torch.distributed.P2POp( - torch.distributed.isend, send_next_shape_tensor, - get_pipeline_model_parallel_next_rank()) + torch.distributed.isend, + send_next_shape_tensor, + get_pipeline_model_parallel_next_rank(), + ) ops.append(send_next_op) if recv_next_shape_tensor is not None: recv_next_op = torch.distributed.P2POp( - torch.distributed.irecv, recv_next_shape_tensor, - get_pipeline_model_parallel_next_rank()) + torch.distributed.irecv, + recv_next_shape_tensor, + get_pipeline_model_parallel_next_rank(), + ) ops.append(recv_next_op) if len(ops) > 0: reqs = torch.distributed.batch_isend_irecv(ops) @@ -109,36 +126,47 @@ def _communicate_shapes(tensor_send_next, tensor_send_prev, return recv_prev_shape, recv_next_shape -def _batched_p2p_ops(*, - tensor_send_prev: Optional[torch.Tensor], - tensor_recv_prev: Optional[torch.Tensor], - tensor_send_next: Optional[torch.Tensor], - tensor_recv_next: Optional[torch.Tensor], - group: torch.distributed.ProcessGroup): + +def _batched_p2p_ops( + *, + tensor_send_prev: Optional[torch.Tensor], + tensor_recv_prev: Optional[torch.Tensor], + tensor_send_next: Optional[torch.Tensor], + tensor_recv_next: Optional[torch.Tensor], + group: torch.distributed.ProcessGroup +): ops = [] if tensor_send_prev is not None: send_prev_op = torch.distributed.P2POp( - torch.distributed.isend, tensor_send_prev, + torch.distributed.isend, + tensor_send_prev, get_pipeline_model_parallel_prev_rank(), - group) + group, + ) ops.append(send_prev_op) if tensor_recv_prev is not None: recv_prev_op = torch.distributed.P2POp( - torch.distributed.irecv, tensor_recv_prev, + torch.distributed.irecv, + tensor_recv_prev, get_pipeline_model_parallel_prev_rank(), - group) + group, + ) ops.append(recv_prev_op) if tensor_send_next is not None: send_next_op = torch.distributed.P2POp( - torch.distributed.isend, tensor_send_next, + torch.distributed.isend, + tensor_send_next, get_pipeline_model_parallel_next_rank(), - group) + group, + ) ops.append(send_next_op) if tensor_recv_next is not None: recv_next_op = torch.distributed.P2POp( - torch.distributed.irecv, tensor_recv_next, + torch.distributed.irecv, + tensor_recv_next, get_pipeline_model_parallel_next_rank(), - group) + group, + ) ops.append(recv_next_op) if len(ops) > 0: reqs = torch.distributed.batch_isend_irecv(ops) @@ -146,12 +174,15 @@ def _batched_p2p_ops(*, reqs = [] return reqs -def _p2p_ops(*, - tensor_send_prev: Optional[torch.Tensor], - tensor_recv_prev: Optional[torch.Tensor], - tensor_send_next: Optional[torch.Tensor], - tensor_recv_next: Optional[torch.Tensor], - group: torch.distributed.ProcessGroup): + +def _p2p_ops( + *, + tensor_send_prev: Optional[torch.Tensor], + tensor_recv_prev: Optional[torch.Tensor], + tensor_send_next: Optional[torch.Tensor], + tensor_recv_next: Optional[torch.Tensor], + group: torch.distributed.ProcessGroup +): reqs = [] rank = get_pipeline_model_parallel_rank() if get_pipeline_model_parallel_rank() % 2 == 0: @@ -221,13 +252,17 @@ def _p2p_ops(*, reqs.append(send_prev_req) return reqs -def _communicate(*, tensor_send_next: Optional[torch.Tensor], - tensor_send_prev: Optional[torch.Tensor], - recv_prev: bool, - recv_next: bool, - tensor_shape: Shape, - config: ModelParallelConfig, - wait_on_reqs: bool = True) -> Tuple[torch.Tensor, torch.Tensor]: + +def _communicate( + *, + tensor_send_next: Optional[torch.Tensor], + tensor_send_prev: Optional[torch.Tensor], + recv_prev: bool, + recv_next: bool, + tensor_shape: Shape, + config: ModelParallelConfig, + wait_on_reqs: bool = True +) -> Tuple[torch.Tensor, torch.Tensor]: """Communicate tensors between stages. Used as helper method in other communication methods that are used in megatron/schedules.py. @@ -270,9 +305,9 @@ def _communicate(*, tensor_send_next: Optional[torch.Tensor], recv_prev_shape = tensor_shape recv_next_shape = tensor_shape else: - recv_prev_shape, recv_next_shape = \ - _communicate_shapes(tensor_send_next, tensor_send_prev, - recv_prev, recv_next, config) + recv_prev_shape, recv_next_shape = _communicate_shapes( + tensor_send_next, tensor_send_prev, recv_prev, recv_next, config + ) if recv_prev: if config.pipeline_dtype is None: @@ -282,10 +317,12 @@ def _communicate(*, tensor_send_next: Optional[torch.Tensor], "tensor_shape must be specified if recv_prev is True. " "Common tensor_shape is (seq_length, micro_batch_size, hidden_size)" ) - tensor_recv_prev = torch.empty(recv_prev_shape, - requires_grad=True, - device=get_accelerator().current_device(), - dtype=config.pipeline_dtype) + tensor_recv_prev = torch.empty( + recv_prev_shape, + requires_grad=True, + device=get_accelerator().current_device(), + dtype=config.pipeline_dtype, + ) if recv_next: if config.pipeline_dtype is None: raise RuntimeError("dtype must be provided if recv_next is True") @@ -294,16 +331,20 @@ def _communicate(*, tensor_send_next: Optional[torch.Tensor], "tensor_shape must be specified if recv_next is True. " "Common tensor_shape is (seq_length, micro_batch_size, hidden_size)" ) - tensor_recv_next = torch.empty(recv_next_shape, - requires_grad=True, - device=get_accelerator().current_device(), - dtype=config.pipeline_dtype) + tensor_recv_next = torch.empty( + recv_next_shape, + requires_grad=True, + device=get_accelerator().current_device(), + dtype=config.pipeline_dtype, + ) # Send tensors in both the forward and backward directions as appropriate. if config.use_ring_exchange_p2p: + def _ring_exchange_wrapper(**kwargs): torch.distributed.ring_exchange(**kwargs) return [] + p2p_func = _ring_exchange_wrapper elif config.batch_p2p_comm: assert wait_on_reqs @@ -311,11 +352,13 @@ def _ring_exchange_wrapper(**kwargs): else: p2p_func = _p2p_ops - reqs = p2p_func(tensor_send_prev=tensor_send_prev, - tensor_recv_prev=tensor_recv_prev, - tensor_send_next=tensor_send_next, - tensor_recv_next=tensor_recv_next, - group=get_pipeline_model_parallel_group()) + reqs = p2p_func( + tensor_send_prev=tensor_send_prev, + tensor_recv_prev=tensor_recv_prev, + tensor_send_next=tensor_send_next, + tensor_recv_next=tensor_recv_next, + group=get_pipeline_model_parallel_group(), + ) if wait_on_reqs and len(reqs) > 0: for req in reqs: @@ -331,9 +374,8 @@ def _ring_exchange_wrapper(**kwargs): @dlp.log -def recv_forward(tensor_shape: Shape, - config: ModelParallelConfig) -> torch.Tensor: - """ Receive tensor from previous rank in pipeline (forward receive). +def recv_forward(tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: + """Receive tensor from previous rank in pipeline (forward receive). See _communicate for argument details. @@ -343,21 +385,22 @@ def recv_forward(tensor_shape: Shape, input_tensor = None else: if config.timers is not None: - config.timers('forward-recv', log_level=2).start() + config.timers("forward-recv", log_level=2).start() input_tensor, _, _ = _communicate( tensor_send_next=None, tensor_send_prev=None, recv_prev=True, recv_next=False, tensor_shape=tensor_shape, - config=config) + config=config, + ) if config.timers is not None: - config.timers('forward-recv').stop() + config.timers("forward-recv").stop() return input_tensor + @dlp.log -def recv_backward(tensor_shape: Shape, - config: ModelParallelConfig) -> torch.Tensor: +def recv_backward(tensor_shape: Shape, config: ModelParallelConfig) -> torch.Tensor: """Receive tensor from next rank in pipeline (backward receive). See _communicate for argument details. @@ -366,21 +409,22 @@ def recv_backward(tensor_shape: Shape, output_tensor_grad = None else: if config.timers is not None: - config.timers('backward-recv', log_level=2).start() + config.timers("backward-recv", log_level=2).start() _, output_tensor_grad, _ = _communicate( tensor_send_next=None, tensor_send_prev=None, recv_prev=False, recv_next=True, tensor_shape=tensor_shape, - config=config) + config=config, + ) if config.timers is not None: - config.timers('backward-recv').stop() + config.timers("backward-recv").stop() return output_tensor_grad + @dlp.log -def send_forward(output_tensor: torch.Tensor, - config: ModelParallelConfig) -> None: +def send_forward(output_tensor: torch.Tensor, config: ModelParallelConfig) -> None: """Send tensor to next rank in pipeline (forward send). See _communicate for argument details. @@ -388,41 +432,44 @@ def send_forward(output_tensor: torch.Tensor, if not core.parallel_state.is_pipeline_last_stage(): if config.timers is not None: - config.timers('forward-send', log_level=2).start() + config.timers("forward-send", log_level=2).start() _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, recv_prev=False, recv_next=False, tensor_shape=None, - config=config) + config=config, + ) if config.timers is not None: - config.timers('forward-send').stop() + config.timers("forward-send").stop() + @dlp.log -def send_backward(input_tensor_grad: torch.Tensor, - config: ModelParallelConfig) -> None: +def send_backward(input_tensor_grad: torch.Tensor, config: ModelParallelConfig) -> None: """Send tensor to previous rank in pipeline (backward send). See _communicate for argument details. """ if not core.parallel_state.is_pipeline_first_stage(): if config.timers is not None: - config.timers('backward-send', log_level=2).start() + config.timers("backward-send", log_level=2).start() _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, recv_prev=False, recv_next=False, tensor_shape=None, - config=config) + config=config, + ) if config.timers is not None: - config.timers('backward-send').stop() + config.timers("backward-send").stop() + @dlp.log -def send_forward_recv_backward(output_tensor: torch.Tensor, - tensor_shape: Shape, - config: ModelParallelConfig) -> torch.Tensor: +def send_forward_recv_backward( + output_tensor: torch.Tensor, tensor_shape: Shape, config: ModelParallelConfig +) -> torch.Tensor: """Batched send and recv with next rank in pipeline. See _communicate for argument details. @@ -431,22 +478,24 @@ def send_forward_recv_backward(output_tensor: torch.Tensor, output_tensor_grad = None else: if config.timers is not None: - config.timers('forward-send-backward-recv', log_level=2).start() - _, output_tensor_grad,_ = _communicate( + config.timers("forward-send-backward-recv", log_level=2).start() + _, output_tensor_grad, _ = _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, recv_prev=False, recv_next=True, tensor_shape=tensor_shape, - config=config) + config=config, + ) if config.timers is not None: - config.timers('forward-send-backward-recv').stop() + config.timers("forward-send-backward-recv").stop() return output_tensor_grad + @dlp.log -def send_backward_recv_forward(input_tensor_grad: torch.Tensor, - tensor_shape: Shape, - config: ModelParallelConfig) -> torch.Tensor: +def send_backward_recv_forward( + input_tensor_grad: torch.Tensor, tensor_shape: Shape, config: ModelParallelConfig +) -> torch.Tensor: """Batched send and recv with previous rank in pipeline. See _communicate for argument details. @@ -455,30 +504,34 @@ def send_backward_recv_forward(input_tensor_grad: torch.Tensor, input_tensor = None else: if config.timers is not None: - config.timers('backward-send-forward-recv', log_level=2).start() + config.timers("backward-send-forward-recv", log_level=2).start() input_tensor, _, _ = _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, recv_prev=True, recv_next=False, tensor_shape=tensor_shape, - config=config) + config=config, + ) if config.timers is not None: - config.timers('backward-send-forward-recv').stop() + config.timers("backward-send-forward-recv").stop() return input_tensor + @dlp.log -def send_forward_recv_forward(output_tensor: torch.Tensor, - recv_prev: bool, - tensor_shape: Shape, - config: ModelParallelConfig, - overlap_p2p_comm: bool = False) -> torch.Tensor: +def send_forward_recv_forward( + output_tensor: torch.Tensor, + recv_prev: bool, + tensor_shape: Shape, + config: ModelParallelConfig, + overlap_p2p_comm: bool = False, +) -> torch.Tensor: """Batched recv from previous rank and send to next rank in pipeline. See _communicate for argument details. """ if config.timers is not None: - config.timers('forward-send-forward-recv', log_level=2).start() + config.timers("forward-send-forward-recv", log_level=2).start() input_tensor, _, wait_handles = _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, @@ -486,25 +539,29 @@ def send_forward_recv_forward(output_tensor: torch.Tensor, recv_next=False, tensor_shape=tensor_shape, wait_on_reqs=(not overlap_p2p_comm), - config=config) + config=config, + ) if config.timers is not None: - config.timers('forward-send-forward-recv').stop() + config.timers("forward-send-forward-recv").stop() if overlap_p2p_comm: return input_tensor, wait_handles return input_tensor + @dlp.log -def send_backward_recv_backward(input_tensor_grad: torch.Tensor, - recv_next: bool, - tensor_shape: Shape, - config: ModelParallelConfig, - overlap_p2p_comm: bool = False) -> torch.Tensor: +def send_backward_recv_backward( + input_tensor_grad: torch.Tensor, + recv_next: bool, + tensor_shape: Shape, + config: ModelParallelConfig, + overlap_p2p_comm: bool = False, +) -> torch.Tensor: """Batched recv from next rank and send to previous rank in pipeline. See _communicate for argument details. """ if config.timers is not None: - config.timers('backward-send-backward-recv', log_level=2).start() + config.timers("backward-send-backward-recv", log_level=2).start() _, output_tensor_grad, wait_handles = _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, @@ -512,35 +569,40 @@ def send_backward_recv_backward(input_tensor_grad: torch.Tensor, recv_next=recv_next, tensor_shape=tensor_shape, wait_on_reqs=(not overlap_p2p_comm), - config=config) + config=config, + ) if config.timers is not None: - config.timers('backward-send-backward-recv').stop() + config.timers("backward-send-backward-recv").stop() if overlap_p2p_comm: return output_tensor_grad, wait_handles return output_tensor_grad + @dlp.log def send_forward_backward_recv_forward_backward( - output_tensor: torch.Tensor, - input_tensor_grad: torch.Tensor, - recv_prev: bool, - recv_next: bool, - tensor_shape: Shape, - config: ModelParallelConfig) -> torch.Tensor: + output_tensor: torch.Tensor, + input_tensor_grad: torch.Tensor, + recv_prev: bool, + recv_next: bool, + tensor_shape: Shape, + config: ModelParallelConfig, +) -> torch.Tensor: """Batched send and recv with previous and next ranks in pipeline. See _communicate for argument details. """ if config.timers is not None: - config.timers('forward-backward-send-forward-backward-recv', - log_level=2).start() + config.timers( + "forward-backward-send-forward-backward-recv", log_level=2 + ).start() input_tensor, output_tensor_grad, _ = _communicate( tensor_send_next=output_tensor, tensor_send_prev=input_tensor_grad, recv_prev=recv_prev, recv_next=recv_next, tensor_shape=tensor_shape, - config=config) + config=config, + ) if config.timers is not None: - config.timers('forward-backward-send-forward-backward-recv').stop() + config.timers("forward-backward-send-forward-backward-recv").stop() return input_tensor, output_tensor_grad diff --git a/megatron/core/pipeline_parallel/schedules.py b/megatron/core/pipeline_parallel/schedules.py index c24959c64f1..cfb3bab87b5 100644 --- a/megatron/core/pipeline_parallel/schedules.py +++ b/megatron/core/pipeline_parallel/schedules.py @@ -23,6 +23,7 @@ # Types Shape = Union[List[int], torch.Size] + def get_forward_backward_func(): """Retrieves the appropriate forward_backward function given the configuration of parallel_state. @@ -97,7 +98,9 @@ def forward_step(data_iterator, model): collect_non_loss_data (optional, bool, default=False): TODO """ - pipeline_model_parallel_size = parallel_state.get_pipeline_model_parallel_world_size() + pipeline_model_parallel_size = ( + parallel_state.get_pipeline_model_parallel_world_size() + ) if pipeline_model_parallel_size > 1: if parallel_state.get_virtual_pipeline_model_parallel_world_size() is not None: forward_backward_func = forward_backward_pipelining_with_interleaving @@ -107,74 +110,77 @@ def forward_step(data_iterator, model): forward_backward_func = forward_backward_no_pipelining return forward_backward_func + def deallocate_output_tensor(out, deallocate_pipeline_outputs=False): - '''Pseudo-deallocate (i.e., set to scalar) the output tensor's '.data' field. + """Pseudo-deallocate (i.e., set to scalar) the output tensor's '.data' field. This method should be called right after the output tensor has been sent to the next pipeline stage. At this point, the output tensor is only useful for its '.grad_fn' field, and not its '.data'. - ''' + """ if (out is None) or (not deallocate_pipeline_outputs): return - assert isinstance(out, torch.Tensor), \ + assert isinstance(out, torch.Tensor), ( "expected Tensor, found %s." % type(out).__name__ - assert out._base is None, \ - "counter-productive to free a view of another tensor." + ) + assert out._base is None, "counter-productive to free a view of another tensor." out.data = torch.empty( (1,), - device = out.device, - dtype = out.dtype, + device=out.device, + dtype=out.dtype, ) -@dlp.log + +@dlp.log def custom_backward(output, grad_output): - '''Directly call C++ autograd engine. + """Directly call C++ autograd engine. To make the 'deallocate_output_tensor' (above) optimization work, the C++ autograd engine must be called directly, bypassing Pytorch's torch.autograd.backward. Pytorch's 'backward' checks that the output and grad have the same shape, while C++'s 'backward' does not. - ''' + """ - assert output.numel() == 1, \ - "output should be pseudo-'freed' in schedule, to optimize memory" - assert isinstance(output, torch.Tensor), \ - "output == '%s'." % type(output).__name__ - assert isinstance(grad_output, (torch.Tensor, type(None))), \ + assert ( + output.numel() == 1 + ), "output should be pseudo-'freed' in schedule, to optimize memory" + assert isinstance(output, torch.Tensor), "output == '%s'." % type(output).__name__ + assert isinstance(grad_output, (torch.Tensor, type(None))), ( "grad_output == '%s'." % type(grad_output).__name__ + ) # Handle scalar output if grad_output is None: assert output.numel() == 1, "implicit grad requires scalar output." grad_output = torch.ones_like( output, - memory_format = torch.preserve_format, + memory_format=torch.preserve_format, ) # Call c++ engine [ see torch/csrc/autograd/python_engine.cpp ] Variable._execution_engine.run_backward( - tensors = (output,), - grad_tensors = (grad_output,), - keep_graph = False, - create_graph = False, - inputs = tuple(), + tensors=(output,), + grad_tensors=(grad_output,), + keep_graph=False, + create_graph=False, + inputs=tuple(), allow_unreachable=True, accumulate_grad=True, ) - - @dlp.log -def forward_step(forward_step_func, - data_iterator, - model, - num_microbatches, - input_tensor, - forward_data_store, - config, - collect_non_loss_data=False, - checkpoint_activations_microbatch=None): +def forward_step( + forward_step_func, + data_iterator, + model, + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data=False, + checkpoint_activations_microbatch=None, +): """Forward step for passed-in model. If first stage, input tensor is obtained from data_iterator, otherwise @@ -183,7 +189,7 @@ def forward_step(forward_step_func, Returns output tensor.""" args = get_args() if config.timers is not None: - config.timers('forward-compute', log_level=2).start() + config.timers("forward-compute", log_level=2).start() unwrap_output_tensor = False if not isinstance(input_tensor, list): @@ -201,7 +207,9 @@ def forward_step(forward_step_func, if checkpoint_activations_microbatch is None: output_tensor, loss_func = forward_step_func(data_iterator, model) else: - output_tensor, loss_func = forward_step_func(data_iterator, model, checkpoint_activations_microbatch) + output_tensor, loss_func = forward_step_func( + data_iterator, model, checkpoint_activations_microbatch + ) if parallel_state.is_pipeline_last_stage(): if not collect_non_loss_data: @@ -217,27 +225,25 @@ def forward_step(forward_step_func, forward_data_store.append(data) if config.timers is not None: - config.timers('forward-compute').stop() + config.timers("forward-compute").stop() # If T5 model (or other model with encoder and decoder) # and in decoder stack, then send encoder_hidden_state # downstream as well. model_type = get_model_type(model) - if parallel_state.is_pipeline_stage_after_split() and \ - model_type == ModelType.encoder_and_decoder: + if ( + parallel_state.is_pipeline_stage_after_split() + and model_type == ModelType.encoder_and_decoder + ): return [output_tensor, input_tensor[-1]] if unwrap_output_tensor: return output_tensor return [output_tensor] + @dlp.log def backward_step( - input_tensor, - output_tensor, - output_tensor_grad, - model_type, - config, - model=None + input_tensor, output_tensor, output_tensor_grad, model_type, config, model=None ): """Backward step through passed-in output tensor. @@ -253,14 +259,12 @@ def backward_step( args = get_args() assert args is not None if config.timers is not None: - config.timers('backward-compute', log_level=2).start() - if (to_skip := getattr(args, 'train_iters_to_skip', None)) is not None: + config.timers("backward-compute", log_level=2).start() + if (to_skip := getattr(args, "train_iters_to_skip", None)) is not None: if config.timers is not None: - config.timers('backward-compute').stop() + config.timers("backward-compute").stop() if len(to_skip) > 0 and args.iteration in [int(i) for i in to_skip]: - print_rank_0( - f'Caught {args.iteration=} in `iters_to_skip`! Skipping!' - ) + print_rank_0(f"Caught {args.iteration=} in `iters_to_skip`! Skipping!") return [None] if args.deepspeed: assert model is not None @@ -285,7 +289,9 @@ def backward_step( if config.deallocate_pipeline_outputs: custom_backward(output_tensor[0], output_tensor_grad[0]) else: - torch.autograd.backward(output_tensor[0], grad_tensors=output_tensor_grad[0]) + torch.autograd.backward( + output_tensor[0], grad_tensors=output_tensor_grad[0] + ) # Collect the grad of the input_tensor. input_tensor_grad = [None] if input_tensor is not None: @@ -297,29 +303,33 @@ def backward_step( input_tensor_grad.append(x.grad) # Handle single skip connection if it exists (encoder_hidden_state in # model with encoder and decoder). - if parallel_state.get_pipeline_model_parallel_world_size() > 1 and \ - parallel_state.is_pipeline_stage_after_split() and \ - model_type == ModelType.encoder_and_decoder: + if ( + parallel_state.get_pipeline_model_parallel_world_size() > 1 + and parallel_state.is_pipeline_stage_after_split() + and model_type == ModelType.encoder_and_decoder + ): if output_tensor_grad[1] is not None: input_tensor_grad[-1].add_(output_tensor_grad[1]) if unwrap_input_tensor_grad: input_tensor_grad = input_tensor_grad[0] if config.timers is not None: - config.timers('backward-compute').stop() + config.timers("backward-compute").stop() return input_tensor_grad + @dlp.log -def forward_backward_no_pipelining(*, - forward_step_func, - data_iterator: Union[Iterator, List[Iterator]], - model: Union[torch.nn.Module, List[torch.nn.Module]], - num_microbatches: int, - seq_length: int, # unused - micro_batch_size: int, # unused - decoder_seq_length: int = None, # unused - forward_only: bool = False, - collect_non_loss_data: bool = False, - ): +def forward_backward_no_pipelining( + *, + forward_step_func, + data_iterator: Union[Iterator, List[Iterator]], + model: Union[torch.nn.Module, List[torch.nn.Module]], + num_microbatches: int, + seq_length: int, # unused + micro_batch_size: int, # unused + decoder_seq_length: int = None, # unused + forward_only: bool = False, + collect_non_loss_data: bool = False, +): """Run forward and backward passes with no pipeline parallelism (no inter-stage communication). @@ -330,12 +340,14 @@ def forward_backward_no_pipelining(*, """ if isinstance(model, list): - assert len(model) == 1, \ - "non-pipeline-parallel schedule does not support model chunking" + assert ( + len(model) == 1 + ), "non-pipeline-parallel schedule does not support model chunking" model = model[0] if isinstance(data_iterator, list): - assert len(data_iterator) == 1, \ - "non-pipeline-parallel schedule does not support model chunking" + assert ( + len(data_iterator) == 1 + ), "non-pipeline-parallel schedule does not support model chunking" data_iterator = data_iterator[0] config = get_model_config(model) @@ -356,45 +368,75 @@ def forward_backward_no_pipelining(*, input_tensor, output_tensor_grad = None, None with no_sync_func(): for i in dlp.iter(range(num_microbatches - 1)): - output_tensor = forward_step(forward_step_func, data_iterator, model, num_microbatches, - input_tensor, forward_data_store, config, collect_non_loss_data) + output_tensor = forward_step( + forward_step_func, + data_iterator, + model, + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data, + ) if not forward_only: - backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config, model) + backward_step( + input_tensor, + output_tensor, + output_tensor_grad, + model_type, + config, + model, + ) if args.deepspeed: model.set_gradient_accumulation_boundary(True) # Run computation for last microbatch out of context handler (want to # synchronize gradients). - output_tensor = forward_step(forward_step_func, data_iterator, model, num_microbatches, - input_tensor, forward_data_store, config, collect_non_loss_data) + output_tensor = forward_step( + forward_step_func, + data_iterator, + model, + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data, + ) if not forward_only: - backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config, model) + backward_step( + input_tensor, output_tensor, output_tensor_grad, model_type, config, model + ) return forward_data_store + @dlp.log -def forward_backward_pipelining_with_interleaving(*, - forward_step_func, - data_iterator: Union[Iterator, List[Iterator]], - model: Union[torch.nn.Module, List[torch.nn.Module]], - num_microbatches: int, - seq_length: int, - micro_batch_size: int, - decoder_seq_length: int = None, - forward_only: bool = False, - collect_non_loss_data: bool = False, - ): +def forward_backward_pipelining_with_interleaving( + *, + forward_step_func, + data_iterator: Union[Iterator, List[Iterator]], + model: Union[torch.nn.Module, List[torch.nn.Module]], + num_microbatches: int, + seq_length: int, + micro_batch_size: int, + decoder_seq_length: int = None, + forward_only: bool = False, + collect_non_loss_data: bool = False, +): """Run interleaved 1F1B schedule (model split into model chunks), with communication between pipeline stages as needed. Returns dictionary with losses if the last stage, empty dict otherwise.""" - assert isinstance(model, list), \ - "interleaved pipeline parallelism expected model chunking" - assert all(isinstance(chunk, torch.nn.Module) for chunk in model), \ - "invalid model chunking" - assert isinstance(data_iterator, list), \ - "interleaved pipeline parallelism expected each model chunk to have a data iterator" + assert isinstance( + model, list + ), "interleaved pipeline parallelism expected model chunking" + assert all( + isinstance(chunk, torch.nn.Module) for chunk in model + ), "invalid model chunking" + assert isinstance( + data_iterator, list + ), "interleaved pipeline parallelism expected each model chunk to have a data iterator" config = get_model_config(model[0]) if config.overlap_p2p_comm and config.batch_p2p_comm: @@ -403,27 +445,32 @@ def forward_backward_pipelining_with_interleaving(*, # Disable async grad reductions no_sync_func = config.no_sync_func if no_sync_func is None and all(isinstance(chunk, torchDDP) for chunk in model): + def multi_no_sync(): stack = contextlib.ExitStack() for chunk in model: stack.enter_context(chunk.no_sync()) return stack + no_sync_func = multi_no_sync if no_sync_func is None: no_sync_func = contextlib.nullcontext no_sync_context = None + def disable_grad_sync(): """Disable asynchronous grad reductions""" nonlocal no_sync_context if no_sync_context is None: no_sync_context = no_sync_func() no_sync_context.__enter__() + def enable_grad_sync(): """Enable asynchronous grad reductions""" nonlocal no_sync_context if no_sync_context is not None: no_sync_context.__exit__(None, None, None) no_sync_context = None + disable_grad_sync() # Model chunk IDs with synchronized grads @@ -439,21 +486,27 @@ def enable_grad_sync(): pipeline_parallel_rank = parallel_state.get_pipeline_model_parallel_rank() if num_microbatches % pipeline_parallel_size != 0: - msg = f'number of microbatches ({num_microbatches}) is not divisible by ' - msg += f'pipeline-model-parallel-size ({pipeline_parallel_size}) ' - msg += 'when using interleaved schedule' + msg = f"number of microbatches ({num_microbatches}) is not divisible by " + msg += f"pipeline-model-parallel-size ({pipeline_parallel_size}) " + msg += "when using interleaved schedule" raise RuntimeError(msg) model_type = get_model_type(model[0]) if model_type == ModelType.encoder_and_decoder: - raise RuntimeError("Interleaving is not supported with an encoder and decoder model.") + raise RuntimeError( + "Interleaving is not supported with an encoder and decoder model." + ) if decoder_seq_length is not None and decoder_seq_length != tensor_shape[0]: - raise RuntimeError("Interleaving is not supported with a different decoder sequence length.") + raise RuntimeError( + "Interleaving is not supported with a different decoder sequence length." + ) tensor_shape = (seq_length, micro_batch_size, config.hidden_size) if config.sequence_parallel: - tensor_shape[0] = tensor_shape[0] // parallel_state.get_tensor_model_parallel_world_size() + tensor_shape[0] = ( + tensor_shape[0] // parallel_state.get_tensor_model_parallel_world_size() + ) # Compute number of warmup and remaining microbatches. num_model_chunks = len(model) @@ -472,9 +525,13 @@ def enable_grad_sync(): num_warmup_microbatches = total_num_microbatches all_warmup_microbatches = True else: - num_warmup_microbatches = (pipeline_parallel_size - pipeline_parallel_rank - 1) * 2 + num_warmup_microbatches = ( + pipeline_parallel_size - pipeline_parallel_rank - 1 + ) * 2 num_warmup_microbatches += (num_model_chunks - 1) * pipeline_parallel_size - num_warmup_microbatches = min(num_warmup_microbatches, total_num_microbatches) + num_warmup_microbatches = min( + num_warmup_microbatches, total_num_microbatches + ) num_microbatches_remaining = total_num_microbatches - num_warmup_microbatches # Checkpoint the activations of partial Transformer layers in a number of micro-batches @@ -496,10 +553,12 @@ def enable_grad_sync(): def get_model_chunk_id(microbatch_id, forward): """Helper method to get the model chunk ID given the iteration number.""" - microbatch_id_in_group = microbatch_id % (pipeline_parallel_size * num_model_chunks) + microbatch_id_in_group = microbatch_id % ( + pipeline_parallel_size * num_model_chunks + ) model_chunk_id = microbatch_id_in_group // pipeline_parallel_size if not forward: - model_chunk_id = (num_model_chunks - model_chunk_id - 1) + model_chunk_id = num_model_chunks - model_chunk_id - 1 return model_chunk_id def is_first_microbatch_for_model_chunk(microbatch_id: int) -> bool: @@ -520,11 +579,13 @@ def is_last_microbatch_for_model_chunk(microbatch_id: int) -> bool: microbatch_group_id = microbatch_id // microbatch_group_size microbatch_id_in_group = microbatch_id % microbatch_group_size if microbatch_group_id == num_microbatch_groups - 1: - return microbatch_id_in_group % pipeline_parallel_size == pipeline_parallel_size - 1 + return ( + microbatch_id_in_group % pipeline_parallel_size + == pipeline_parallel_size - 1 + ) else: return False - def forward_step_helper(microbatch_id, checkpoint_activations_microbatch): """Helper method to run forward step with model split into chunks (run set_virtual_pipeline_model_parallel_rank() before calling @@ -539,26 +600,34 @@ def forward_step_helper(microbatch_id, checkpoint_activations_microbatch): # pipeline-parallel group. if config.param_sync_func is not None: param_sync_microbatch_id = microbatch_id + pipeline_parallel_rank - if param_sync_microbatch_id < num_microbatches and is_first_microbatch_for_model_chunk(param_sync_microbatch_id): - param_sync_chunk_id = get_model_chunk_id(param_sync_microbatch_id, forward=True) + 1 + if ( + param_sync_microbatch_id < num_microbatches + and is_first_microbatch_for_model_chunk(param_sync_microbatch_id) + ): + param_sync_chunk_id = ( + get_model_chunk_id(param_sync_microbatch_id, forward=True) + 1 + ) if 1 < param_sync_chunk_id < num_model_chunks: config.param_sync_func(model[param_sync_chunk_id].parameters()) # forward step if parallel_state.is_pipeline_first_stage(): - if len(input_tensors[model_chunk_id]) == \ - len(output_tensors[model_chunk_id]): + if len(input_tensors[model_chunk_id]) == len( + output_tensors[model_chunk_id] + ): input_tensors[model_chunk_id].append(None) input_tensor = input_tensors[model_chunk_id][-1] - output_tensor = forward_step(forward_step_func, - data_iterator[model_chunk_id], - model[model_chunk_id], - num_microbatches, - input_tensor, - forward_data_store, - config, - collect_non_loss_data, - checkpoint_activations_microbatch) + output_tensor = forward_step( + forward_step_func, + data_iterator[model_chunk_id], + model[model_chunk_id], + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data, + checkpoint_activations_microbatch, + ) output_tensors[model_chunk_id].append(output_tensor) # if forward-only, no need to save tensors for a backward pass @@ -576,7 +645,9 @@ def backward_step_helper(microbatch_id): parallel_state.set_virtual_pipeline_model_parallel_rank(model_chunk_id) # launch grad synchronization (default) - if config.grad_sync_func is None and is_last_microbatch_for_model_chunk(microbatch_id): + if config.grad_sync_func is None and is_last_microbatch_for_model_chunk( + microbatch_id + ): enable_grad_sync() synchronized_model_chunks.add(model_chunk_id) @@ -586,8 +657,9 @@ def backward_step_helper(microbatch_id): input_tensor = input_tensors[model_chunk_id].pop(0) output_tensor = output_tensors[model_chunk_id].pop(0) output_tensor_grad = output_tensor_grads[model_chunk_id].pop(0) - input_tensor_grad = \ - backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config) + input_tensor_grad = backward_step( + input_tensor, output_tensor, output_tensor_grad, model_type, config + ) # launch grad synchronization (custom grad sync) # Note: Asynchronous communication tends to slow down compute. @@ -596,8 +668,12 @@ def backward_step_helper(microbatch_id): # pipeline-parallel group. if config.grad_sync_func is not None: grad_sync_microbatch_id = microbatch_id - pipeline_parallel_rank - if grad_sync_microbatch_id >= 0 and is_last_microbatch_for_model_chunk(grad_sync_microbatch_id): - grad_sync_chunk_id = get_model_chunk_id(grad_sync_microbatch_id, forward=False) + if grad_sync_microbatch_id >= 0 and is_last_microbatch_for_model_chunk( + grad_sync_microbatch_id + ): + grad_sync_chunk_id = get_model_chunk_id( + grad_sync_microbatch_id, forward=False + ) enable_grad_sync() config.grad_sync_func(model[grad_sync_chunk_id].parameters()) synchronized_model_chunks.add(grad_sync_chunk_id) @@ -607,8 +683,7 @@ def backward_step_helper(microbatch_id): # Run warmup forward passes. parallel_state.set_virtual_pipeline_model_parallel_rank(0) - input_tensors[0].append( - p2p_communication.recv_forward(tensor_shape, config)) + input_tensors[0].append(p2p_communication.recv_forward(tensor_shape, config)) fwd_wait_handles = None bwd_wait_handles = None @@ -621,15 +696,17 @@ def backward_step_helper(microbatch_id): # Decide to checkpoint all layers' activations of the current micro-batch if max_outstanding_backprops is not None: - checkpoint_activations_microbatch = k % max_outstanding_backprops >= \ - config.num_microbatches_with_partial_activation_checkpoints + checkpoint_activations_microbatch = ( + k % max_outstanding_backprops + >= config.num_microbatches_with_partial_activation_checkpoints + ) else: checkpoint_activations_microbatch = None output_tensor = forward_step_helper(k, checkpoint_activations_microbatch) # Determine if tensor should be received from previous stage. - next_forward_model_chunk_id = get_model_chunk_id(k+1, forward=True) + next_forward_model_chunk_id = get_model_chunk_id(k + 1, forward=True) recv_prev = True if parallel_state.is_pipeline_first_stage(ignore_virtual=True): if next_forward_model_chunk_id == 0: @@ -644,46 +721,66 @@ def backward_step_helper(microbatch_id): # Send and receive tensors as appropriate (send tensors computed # in this iteration; receive tensors for next iteration). if not config.overlap_p2p_comm: - if k == (num_warmup_microbatches - 1) and not forward_only and \ - not all_warmup_microbatches: + if ( + k == (num_warmup_microbatches - 1) + and not forward_only + and not all_warmup_microbatches + ): input_tensor_grad = None recv_next = True if parallel_state.is_pipeline_last_stage(ignore_virtual=True): recv_next = False - input_tensor, output_tensor_grad = \ + input_tensor, output_tensor_grad = ( p2p_communication.send_forward_backward_recv_forward_backward( - output_tensor, input_tensor_grad, - recv_prev=recv_prev, recv_next=recv_next, - tensor_shape=tensor_shape, config=config) - output_tensor_grads[num_model_chunks-1].append(output_tensor_grad) - else: - input_tensor = \ - p2p_communication.send_forward_recv_forward( - output_tensor, recv_prev=recv_prev, + output_tensor, + input_tensor_grad, + recv_prev=recv_prev, + recv_next=recv_next, tensor_shape=tensor_shape, - config=config) + config=config, + ) + ) + output_tensor_grads[num_model_chunks - 1].append(output_tensor_grad) + else: + input_tensor = p2p_communication.send_forward_recv_forward( + output_tensor, + recv_prev=recv_prev, + tensor_shape=tensor_shape, + config=config, + ) input_tensors[next_forward_model_chunk_id].append(input_tensor) else: - input_tensor, fwd_wait_handles = \ + input_tensor, fwd_wait_handles = ( p2p_communication.send_forward_recv_forward( - output_tensor, recv_prev=recv_prev, - tensor_shape=tensor_shape, config=config, - overlap_p2p_comm=True) + output_tensor, + recv_prev=recv_prev, + tensor_shape=tensor_shape, + config=config, + overlap_p2p_comm=True, + ) + ) - if k == (num_warmup_microbatches - 1) and not forward_only and \ - not all_warmup_microbatches: + if ( + k == (num_warmup_microbatches - 1) + and not forward_only + and not all_warmup_microbatches + ): input_tensor_grad = None recv_next = True if parallel_state.is_pipeline_last_stage(ignore_virtual=True): recv_next = False - output_tensor_grad, bwd_wait_handles = p2p_communication.send_backward_recv_backward( - input_tensor_grad, recv_next=recv_next, - tensor_shape=tensor_shape, - config=config, - overlap_p2p_comm=True) + output_tensor_grad, bwd_wait_handles = ( + p2p_communication.send_backward_recv_backward( + input_tensor_grad, + recv_next=recv_next, + tensor_shape=tensor_shape, + config=config, + overlap_p2p_comm=True, + ) + ) - output_tensor_grads[num_model_chunks-1].append(output_tensor_grad) + output_tensor_grads[num_model_chunks - 1].append(output_tensor_grad) input_tensors[next_forward_model_chunk_id].append(input_tensor) deallocate_output_tensor(output_tensor, config.deallocate_pipeline_outputs) @@ -696,8 +793,8 @@ def backward_step_helper(microbatch_id): # Decide to checkpoint all layers' activations of the current micro-batch if max_outstanding_backprops is not None: checkpoint_activations_microbatch = ( - forward_k % max_outstanding_backprops >= \ - config.num_microbatches_with_partial_activation_checkpoints + forward_k % max_outstanding_backprops + >= config.num_microbatches_with_partial_activation_checkpoints ) else: checkpoint_activations_microbatch = None @@ -709,12 +806,16 @@ def backward_step_helper(microbatch_id): deallocate_output_tensor(output_tensor, config.deallocate_pipeline_outputs) - output_tensor = forward_step_helper(forward_k, checkpoint_activations_microbatch) + output_tensor = forward_step_helper( + forward_k, checkpoint_activations_microbatch + ) # Determine if current stage has anything to send in either direction, # otherwise set tensor to None. forward_model_chunk_id = get_model_chunk_id(forward_k, forward=True) - parallel_state.set_virtual_pipeline_model_parallel_rank(forward_model_chunk_id) + parallel_state.set_virtual_pipeline_model_parallel_rank( + forward_model_chunk_id + ) # Last virtual stage no activation tensor to send if parallel_state.is_pipeline_last_stage(): @@ -726,13 +827,15 @@ def backward_step_helper(microbatch_id): if parallel_state.is_pipeline_first_stage(ignore_virtual=True): # First stage is ahead of last stage by (pipeline_parallel_size - 1). next_forward_model_chunk_id = get_model_chunk_id( - forward_k - (pipeline_parallel_size - 1), forward=True) + forward_k - (pipeline_parallel_size - 1), forward=True + ) if next_forward_model_chunk_id == (num_model_chunks - 1): recv_prev = False next_forward_model_chunk_id += 1 else: - next_forward_model_chunk_id = get_model_chunk_id(forward_k + 1, - forward=True) + next_forward_model_chunk_id = get_model_chunk_id( + forward_k + 1, forward=True + ) # If last iteration, don't receive; we already received one extra # before the start of the for loop. @@ -741,14 +844,17 @@ def backward_step_helper(microbatch_id): # Send activation tensor to the next stage and receive activation tensor from the # previous stage - input_tensor, fwd_wait_handles = \ + input_tensor, fwd_wait_handles = ( p2p_communication.send_forward_recv_forward( - output_tensor, recv_prev=recv_prev, + output_tensor, + recv_prev=recv_prev, tensor_shape=tensor_shape, dtype=dtype, batch_p2p_comm=batch_p2p_comm, timers=timers, - overlap_p2p_comm=True) + overlap_p2p_comm=True, + ) + ) # assert fwd_wait_handles is not None if bwd_wait_handles is not None: @@ -760,7 +866,9 @@ def backward_step_helper(microbatch_id): input_tensor_grad = backward_step_helper(backward_k) backward_model_chunk_id = get_model_chunk_id(backward_k, forward=False) - parallel_state.set_virtual_pipeline_model_parallel_rank(backward_model_chunk_id) + parallel_state.set_virtual_pipeline_model_parallel_rank( + backward_model_chunk_id + ) # First virtual stage no activation gradient tensor to send if parallel_state.is_pipeline_first_stage(): @@ -781,14 +889,20 @@ def backward_step_helper(microbatch_id): backward_k + 1, forward=False ) - output_tensor_grad, bwd_wait_handles = p2p_communication.send_backward_recv_backward( - input_tensor_grad, recv_next=recv_next, - tensor_shape=tensor_shape, - config=config, - overlap_p2p_comm=True) + output_tensor_grad, bwd_wait_handles = ( + p2p_communication.send_backward_recv_backward( + input_tensor_grad, + recv_next=recv_next, + tensor_shape=tensor_shape, + config=config, + overlap_p2p_comm=True, + ) + ) - else: # no p2p overlap - output_tensor = forward_step_helper(forward_k, checkpoint_activations_microbatch) + else: # no p2p overlap + output_tensor = forward_step_helper( + forward_k, checkpoint_activations_microbatch + ) # Backward pass. backward_k = k @@ -800,12 +914,16 @@ def backward_step_helper(microbatch_id): # Determine if current stage has anything to send in either direction, # otherwise set tensor to None. forward_model_chunk_id = get_model_chunk_id(forward_k, forward=True) - parallel_state.set_virtual_pipeline_model_parallel_rank(forward_model_chunk_id) + parallel_state.set_virtual_pipeline_model_parallel_rank( + forward_model_chunk_id + ) if parallel_state.is_pipeline_last_stage(): output_tensor = None backward_model_chunk_id = get_model_chunk_id(backward_k, forward=False) - parallel_state.set_virtual_pipeline_model_parallel_rank(backward_model_chunk_id) + parallel_state.set_virtual_pipeline_model_parallel_rank( + backward_model_chunk_id + ) if parallel_state.is_pipeline_first_stage(): input_tensor_grad = None @@ -815,25 +933,29 @@ def backward_step_helper(microbatch_id): if parallel_state.is_pipeline_first_stage(ignore_virtual=True): # First stage is ahead of last stage by (pipeline_parallel_size - 1). next_forward_model_chunk_id = get_model_chunk_id( - forward_k - (pipeline_parallel_size - 1), forward=True) + forward_k - (pipeline_parallel_size - 1), forward=True + ) if next_forward_model_chunk_id == (num_model_chunks - 1): recv_prev = False next_forward_model_chunk_id += 1 else: - next_forward_model_chunk_id = get_model_chunk_id(forward_k + 1, - forward=True) + next_forward_model_chunk_id = get_model_chunk_id( + forward_k + 1, forward=True + ) recv_next = True if parallel_state.is_pipeline_last_stage(ignore_virtual=True): # Last stage is ahead of first stage by (pipeline_parallel_size - 1). next_backward_model_chunk_id = get_model_chunk_id( - backward_k - (pipeline_parallel_size - 1), forward=False) + backward_k - (pipeline_parallel_size - 1), forward=False + ) if next_backward_model_chunk_id == 0: recv_next = False next_backward_model_chunk_id -= 1 else: - next_backward_model_chunk_id = get_model_chunk_id(backward_k + 1, - forward=False) + next_backward_model_chunk_id = get_model_chunk_id( + backward_k + 1, forward=False + ) # If last iteration, don't receive; we already received one extra # before the start of the for loop. @@ -841,11 +963,16 @@ def backward_step_helper(microbatch_id): recv_prev = False # Communicate tensors. - input_tensor, output_tensor_grad = \ + input_tensor, output_tensor_grad = ( p2p_communication.send_forward_backward_recv_forward_backward( - output_tensor, input_tensor_grad, - recv_prev=recv_prev, recv_next=recv_next, - tensor_shape=tensor_shape, config=config) + output_tensor, + input_tensor_grad, + recv_prev=recv_prev, + recv_next=recv_next, + tensor_shape=tensor_shape, + config=config, + ) + ) deallocate_output_tensor(output_tensor, config.deallocate_pipeline_outputs) # Put input_tensor and output_tensor_grad in data structures in the @@ -853,8 +980,7 @@ def backward_step_helper(microbatch_id): if recv_prev: input_tensors[next_forward_model_chunk_id].append(input_tensor) if recv_next: - output_tensor_grads[next_backward_model_chunk_id].append( - output_tensor_grad) + output_tensor_grads[next_backward_model_chunk_id].append(output_tensor_grad) deallocate_output_tensor(output_tensor, config.deallocate_pipeline_outputs) @@ -865,11 +991,12 @@ def backward_step_helper(microbatch_id): wait_handle.wait() if all_warmup_microbatches: - output_tensor_grads[num_model_chunks-1].append( - p2p_communication.recv_backward(tensor_shape, config=config)) + output_tensor_grads[num_model_chunks - 1].append( + p2p_communication.recv_backward(tensor_shape, config=config) + ) for k in range(num_microbatches_remaining, total_num_microbatches): input_tensor_grad = backward_step_helper(k) - next_backward_model_chunk_id = get_model_chunk_id(k+1, forward=False) + next_backward_model_chunk_id = get_model_chunk_id(k + 1, forward=False) recv_next = True if parallel_state.is_pipeline_last_stage(ignore_virtual=True): if next_backward_model_chunk_id == (num_model_chunks - 1): @@ -878,8 +1005,12 @@ def backward_step_helper(microbatch_id): recv_next = False output_tensor_grads[next_backward_model_chunk_id].append( p2p_communication.send_backward_recv_backward( - input_tensor_grad, recv_next=recv_next, - tensor_shape=tensor_shape, config=config)) + input_tensor_grad, + recv_next=recv_next, + tensor_shape=tensor_shape, + config=config, + ) + ) # Launch any remaining grad reductions enable_grad_sync() @@ -894,13 +1025,16 @@ def backward_step_helper(microbatch_id): return forward_data_store -def get_tensor_shapes(*, - rank: int, - model_type: ModelType, - seq_length: int, - micro_batch_size: int, - decoder_seq_length: int, - config): + +def get_tensor_shapes( + *, + rank: int, + model_type: ModelType, + seq_length: int, + micro_batch_size: int, + decoder_seq_length: int, + config, +): # Determine right tensor sizes (based on position of rank with respect to split # rank) and model size. # Send two tensors if model is T5 and rank is in decoder stage: @@ -913,13 +1047,17 @@ def get_tensor_shapes(*, if config.sequence_parallel: seq_length = seq_length // parallel_state.get_tensor_model_parallel_world_size() - decoder_seq_length = decoder_seq_length // parallel_state.get_tensor_model_parallel_world_size() + decoder_seq_length = ( + decoder_seq_length // parallel_state.get_tensor_model_parallel_world_size() + ) if model_type == ModelType.encoder_and_decoder: if parallel_state.is_pipeline_stage_before_split(rank): tensor_shapes.append((seq_length, micro_batch_size, config.hidden_size)) else: - tensor_shapes.append((decoder_seq_length, micro_batch_size, config.hidden_size)) + tensor_shapes.append( + (decoder_seq_length, micro_batch_size, config.hidden_size) + ) tensor_shapes.append((seq_length, micro_batch_size, config.hidden_size)) else: tensor_shapes.append((seq_length, micro_batch_size, config.hidden_size)) @@ -936,6 +1074,7 @@ def recv_forward(tensor_shapes, config): input_tensors.append(p2p_communication.recv_forward(tensor_shape, config)) return input_tensors + @dlp.log def recv_backward(tensor_shapes, config): output_tensor_grads = [] @@ -943,84 +1082,98 @@ def recv_backward(tensor_shapes, config): if tensor_shape is None: output_tensor_grads.append(None) else: - output_tensor_grads.append(p2p_communication.recv_backward(tensor_shape, config)) + output_tensor_grads.append( + p2p_communication.recv_backward(tensor_shape, config) + ) return output_tensor_grads + @dlp.log def send_forward(output_tensors, tensor_shapes, config): if not isinstance(output_tensors, list): output_tensors = [output_tensors] - for (output_tensor, tensor_shape) in zip(output_tensors, tensor_shapes): + for output_tensor, tensor_shape in zip(output_tensors, tensor_shapes): if tensor_shape is None: continue p2p_communication.send_forward(output_tensor, config) + @dlp.log def send_backward(input_tensor_grads, tensor_shapes, config): if not isinstance(input_tensor_grads, list): input_tensor_grads = [input_tensor_grads] - for (input_tensor_grad, tensor_shape) in zip(input_tensor_grads, tensor_shapes): + for input_tensor_grad, tensor_shape in zip(input_tensor_grads, tensor_shapes): if tensor_shape is None: continue p2p_communication.send_backward(input_tensor_grad, config) + @dlp.log def send_forward_recv_backward(output_tensors, tensor_shapes, config): if not isinstance(output_tensors, list): output_tensors = [output_tensors] output_tensor_grads = [] - for (output_tensor, tensor_shape) in zip(output_tensors, tensor_shapes): + for output_tensor, tensor_shape in zip(output_tensors, tensor_shapes): if tensor_shape is None: output_tensor_grads.append(None) continue output_tensor_grad = p2p_communication.send_forward_recv_backward( - output_tensor, tensor_shape, config) + output_tensor, tensor_shape, config + ) output_tensor_grads.append(output_tensor_grad) return output_tensor_grads + @dlp.log def send_backward_recv_forward(input_tensor_grads, tensor_shapes, config): if not isinstance(input_tensor_grads, list): input_tensor_grads = [input_tensor_grads] input_tensors = [] - for (input_tensor_grad, tensor_shape) in zip(input_tensor_grads, tensor_shapes): + for input_tensor_grad, tensor_shape in zip(input_tensor_grads, tensor_shapes): if tensor_shape is None: input_tensors.append(None) continue input_tensor = p2p_communication.send_backward_recv_forward( - input_tensor_grad, tensor_shape, config) + input_tensor_grad, tensor_shape, config + ) input_tensors.append(input_tensor) return input_tensors + @dlp.log -def forward_backward_pipelining_without_interleaving(*, - forward_step_func, - data_iterator: Union[Iterator, List[Iterator]], - model: Union[torch.nn.Module, List[torch.nn.Module]], - num_microbatches: int, - seq_length: int, - micro_batch_size: int, - decoder_seq_length: int = None, - forward_only: bool = False, - collect_non_loss_data: bool = False, - ): +def forward_backward_pipelining_without_interleaving( + *, + forward_step_func, + data_iterator: Union[Iterator, List[Iterator]], + model: Union[torch.nn.Module, List[torch.nn.Module]], + num_microbatches: int, + seq_length: int, + micro_batch_size: int, + decoder_seq_length: int = None, + forward_only: bool = False, + collect_non_loss_data: bool = False, +): """Run non-interleaved 1F1B schedule, with communication between pipeline stages. Returns dictionary with losses if the last stage, empty dict otherwise.""" if isinstance(model, list): - assert len(model) == 1, \ - "non-interleaved pipeline parallelism does not support model chunking" + assert ( + len(model) == 1 + ), "non-interleaved pipeline parallelism does not support model chunking" model = model[0] if isinstance(data_iterator, list): - assert len(data_iterator) == 1, \ - "non-pipeline-parallel schedule does not support model chunking" + assert ( + len(data_iterator) == 1 + ), "non-pipeline-parallel schedule does not support model chunking" data_iterator = data_iterator[0] config = get_model_config(model) if config.overlap_p2p_comm: - raise ValueError("Non-interleaved pipeline parallelism does not support overlapping p2p communication") + raise ValueError( + "Non-interleaved pipeline parallelism does not support overlapping p2p communication" + ) # Disable async grad reductions no_sync_func = config.no_sync_func @@ -1029,29 +1182,31 @@ def forward_backward_pipelining_without_interleaving(*, if no_sync_func is None: no_sync_func = contextlib.nullcontext no_sync_context = None + def disable_grad_sync(): """Disable asynchronous grad reductions""" nonlocal no_sync_context if no_sync_context is None: no_sync_context = no_sync_func() no_sync_context.__enter__() + def enable_grad_sync(): """Enable asynchronous grad reductions""" nonlocal no_sync_context if no_sync_context is not None: no_sync_context.__exit__(None, None, None) no_sync_context = None + disable_grad_sync() # Compute number of warmup microbatches. - num_warmup_microbatches = \ - (parallel_state.get_pipeline_model_parallel_world_size() - - parallel_state.get_pipeline_model_parallel_rank() - 1) - num_warmup_microbatches = min( - num_warmup_microbatches, - num_microbatches) - num_microbatches_remaining = \ - num_microbatches - num_warmup_microbatches + num_warmup_microbatches = ( + parallel_state.get_pipeline_model_parallel_world_size() + - parallel_state.get_pipeline_model_parallel_rank() + - 1 + ) + num_warmup_microbatches = min(num_warmup_microbatches, num_microbatches) + num_microbatches_remaining = num_microbatches - num_warmup_microbatches # Checkpoint the activations of partial Transformer layers in a number of micro-batches # within the maximum outstanding micro-batch backpropagations. @@ -1068,18 +1223,22 @@ def enable_grad_sync(): model_type = get_model_type(model) rank = parallel_state.get_pipeline_model_parallel_rank() - recv_tensor_shapes = get_tensor_shapes(rank=rank-1, - model_type=model_type, - seq_length=seq_length, - micro_batch_size=micro_batch_size, - decoder_seq_length=decoder_seq_length, - config=config) - send_tensor_shapes = get_tensor_shapes(rank=rank, - model_type=model_type, - seq_length=seq_length, - micro_batch_size=micro_batch_size, - decoder_seq_length=decoder_seq_length, - config=config) + recv_tensor_shapes = get_tensor_shapes( + rank=rank - 1, + model_type=model_type, + seq_length=seq_length, + micro_batch_size=micro_batch_size, + decoder_seq_length=decoder_seq_length, + config=config, + ) + send_tensor_shapes = get_tensor_shapes( + rank=rank, + model_type=model_type, + seq_length=seq_length, + micro_batch_size=micro_batch_size, + decoder_seq_length=decoder_seq_length, + config=config, + ) # Input, output tensors only need to be saved when doing backward passes input_tensors = None @@ -1094,21 +1253,32 @@ def enable_grad_sync(): # Decide to checkpoint all layers' activations of the current micro-batch if max_outstanding_backprops is not None: checkpoint_activations_microbatch = ( - i % max_outstanding_backprops >= config.num_microbatches_with_partial_activation_checkpoints + i % max_outstanding_backprops + >= config.num_microbatches_with_partial_activation_checkpoints ) else: checkpoint_activations_microbatch = None input_tensor = recv_forward(recv_tensor_shapes, config) - output_tensor = forward_step(forward_step_func, data_iterator, model, num_microbatches, - input_tensor, forward_data_store, config, collect_non_loss_data, - checkpoint_activations_microbatch) + output_tensor = forward_step( + forward_step_func, + data_iterator, + model, + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data, + checkpoint_activations_microbatch, + ) send_forward(output_tensor, send_tensor_shapes, config) if not forward_only: input_tensors.append(input_tensor) output_tensors.append(output_tensor) - deallocate_output_tensor(output_tensor[0], config.deallocate_pipeline_outputs) + deallocate_output_tensor( + output_tensor[0], config.deallocate_pipeline_outputs + ) # Before running 1F1B, need to receive first forward tensor. # If all microbatches are run in warmup / cooldown phase, then no need to @@ -1118,20 +1288,27 @@ def enable_grad_sync(): # Run 1F1B in steady state. for i in range(num_microbatches_remaining): - last_iteration = (i == (num_microbatches_remaining - 1)) + last_iteration = i == (num_microbatches_remaining - 1) # Decide to checkpoint all layers' activations of the current micro-batch if max_outstanding_backprops is not None: checkpoint_activations_microbatch = ( - ((i+num_warmup_microbatches) % max_outstanding_backprops) >= \ - config.num_microbatches_with_partial_activation_checkpoints - ) + (i + num_warmup_microbatches) % max_outstanding_backprops + ) >= config.num_microbatches_with_partial_activation_checkpoints else: checkpoint_activations_microbatch = None - output_tensor = forward_step(forward_step_func, data_iterator, model, num_microbatches, - input_tensor, forward_data_store, config, collect_non_loss_data, - checkpoint_activations_microbatch) + output_tensor = forward_step( + forward_step_func, + data_iterator, + model, + num_microbatches, + input_tensor, + forward_data_store, + config, + collect_non_loss_data, + checkpoint_activations_microbatch, + ) if forward_only: send_forward(output_tensor, send_tensor_shapes, config) @@ -1140,28 +1317,38 @@ def enable_grad_sync(): input_tensor = recv_forward(recv_tensor_shapes, config) else: - output_tensor_grad = \ - send_forward_recv_backward(output_tensor, send_tensor_shapes, config) + output_tensor_grad = send_forward_recv_backward( + output_tensor, send_tensor_shapes, config + ) # Add input_tensor and output_tensor to end of list. input_tensors.append(input_tensor) output_tensors.append(output_tensor) - deallocate_output_tensor(output_tensor[0], config.deallocate_pipeline_outputs) + deallocate_output_tensor( + output_tensor[0], config.deallocate_pipeline_outputs + ) # Pop input_tensor and output_tensor from the start of the list for # the backward pass. input_tensor = input_tensors.pop(0) output_tensor = output_tensors.pop(0) - input_tensor_grad = \ - backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config, model) + input_tensor_grad = backward_step( + input_tensor, + output_tensor, + output_tensor_grad, + model_type, + config, + model, + ) if last_iteration: input_tensor = None send_backward(input_tensor_grad, recv_tensor_shapes, config) else: - input_tensor = \ - send_backward_recv_forward(input_tensor_grad, recv_tensor_shapes, config) + input_tensor = send_backward_recv_forward( + input_tensor_grad, recv_tensor_shapes, config + ) # Run cooldown backward passes. if not forward_only: @@ -1172,7 +1359,7 @@ def enable_grad_sync(): # async grad reduction in first pipeline stage. Other # pipeline stages do grad reduction during pipeline # bubble. - if i == num_warmup_microbatches-1: + if i == num_warmup_microbatches - 1: if config.grad_sync_func is None or rank == 0: enable_grad_sync() @@ -1181,8 +1368,14 @@ def enable_grad_sync(): output_tensor_grad = recv_backward(send_tensor_shapes, config) - input_tensor_grad = \ - backward_step(input_tensor, output_tensor, output_tensor_grad, model_type, config, model) + input_tensor_grad = backward_step( + input_tensor, + output_tensor, + output_tensor_grad, + model_type, + config, + model, + ) send_backward(input_tensor_grad, recv_tensor_shapes, config) diff --git a/megatron/core/sequence_parallel/cross_entropy.py b/megatron/core/sequence_parallel/cross_entropy.py index 4da51d59533..45ba29f82ff 100644 --- a/megatron/core/sequence_parallel/cross_entropy.py +++ b/megatron/core/sequence_parallel/cross_entropy.py @@ -4,9 +4,10 @@ from megatron.core.parallel_state import ( get_sequence_parallel_group, get_sequence_parallel_rank, - get_sequence_parallel_world_size + get_sequence_parallel_world_size, ) + class _VocabSequenceParallelCrossEntropy(torch.autograd.Function): @staticmethod @@ -18,16 +19,29 @@ def forward(ctx, vocab_seq_parallel_logits, target, label_smoothing=0.0): # Need softmax for backward softmax = torch.nn.functional.softmax(vocab_seq_parallel_logits, dim=-1) ctx.vocab_size = vocab_seq_parallel_logits.size(2) - loss = torch.nn.functional.nll_loss(softmax.log().view(-1, ctx.vocab_size), target.view(-1), reduction='none') - - ctx.seqlen = vocab_seq_parallel_logits.size(0) * get_sequence_parallel_world_size() + loss = torch.nn.functional.nll_loss( + softmax.log().view(-1, ctx.vocab_size), target.view(-1), reduction="none" + ) + + ctx.seqlen = ( + vocab_seq_parallel_logits.size(0) * get_sequence_parallel_world_size() + ) batch_size = vocab_seq_parallel_logits.size(1) - loss_all = torch.empty(ctx.seqlen, batch_size, dtype=vocab_seq_parallel_logits.dtype, device=vocab_seq_parallel_logits.device) - if version.parse(torch.__version__) >= version.parse('1.13'): - torch.distributed.all_gather_into_tensor(loss_all, loss, group=get_sequence_parallel_group()) + loss_all = torch.empty( + ctx.seqlen, + batch_size, + dtype=vocab_seq_parallel_logits.dtype, + device=vocab_seq_parallel_logits.device, + ) + if version.parse(torch.__version__) >= version.parse("1.13"): + torch.distributed.all_gather_into_tensor( + loss_all, loss, group=get_sequence_parallel_group() + ) else: - torch.distributed._all_gather_base(loss_all, loss, group=get_sequence_parallel_group()) + torch.distributed._all_gather_base( + loss_all, loss, group=get_sequence_parallel_group() + ) ctx.save_for_backward(softmax, target) @@ -39,12 +53,13 @@ def backward(ctx, grad_output): step_seqlen = ctx.seqlen // get_sequence_parallel_world_size() sp_rank = get_sequence_parallel_rank() - grad_output_part = grad_output[step_seqlen*sp_rank:step_seqlen*(sp_rank + 1), :] + grad_output_part = grad_output[ + step_seqlen * sp_rank : step_seqlen * (sp_rank + 1), : + ] grad_input = softmax grad_2d = grad_input.view(-1, ctx.vocab_size) - arange_1d = torch.arange(start=0, end=grad_2d.size()[0], - device=grad_2d.device) + arange_1d = torch.arange(start=0, end=grad_2d.size()[0], device=grad_2d.device) grad_2d[arange_1d, target.view(-1)] -= 1 grad_input.mul_(grad_output_part.unsqueeze(dim=-1)) @@ -52,5 +67,9 @@ def backward(ctx, grad_output): return grad_input, None, None -def vocab_sequence_parallel_cross_entropy(vocab_parallel_logits, target, label_smoothing=0.0): - return _VocabSequenceParallelCrossEntropy.apply(vocab_parallel_logits, target, label_smoothing) +def vocab_sequence_parallel_cross_entropy( + vocab_parallel_logits, target, label_smoothing=0.0 +): + return _VocabSequenceParallelCrossEntropy.apply( + vocab_parallel_logits, target, label_smoothing + ) diff --git a/megatron/core/tensor_parallel/__init__.py b/megatron/core/tensor_parallel/__init__.py index 12c1ab95bb6..45fb7c7775c 100644 --- a/megatron/core/tensor_parallel/__init__.py +++ b/megatron/core/tensor_parallel/__init__.py @@ -9,8 +9,7 @@ set_defaults_if_not_set_tensor_model_parallel_attributes, copy_tensor_model_parallel_attributes, param_is_not_tensor_parallel_duplicate, - linear_with_grad_accumulation_and_async_allreduce - + linear_with_grad_accumulation_and_async_allreduce, ) from .mappings import ( @@ -41,7 +40,7 @@ "vocab_parallel_cross_entropy", # data.py "broadcast_data", - #layers.py + # layers.py "ColumnParallelLinear", "RowParallelLinear", "VocabParallelEmbedding", @@ -54,7 +53,7 @@ "copy_to_tensor_model_parallel_region", "gather_from_tensor_model_parallel_region", "gather_from_sequence_parallel_region", -# "reduce_from_tensor_model_parallel_region", + # "reduce_from_tensor_model_parallel_region", "scatter_to_tensor_model_parallel_region", "scatter_to_sequence_parallel_region", # random.py diff --git a/megatron/core/tensor_parallel/cross_entropy.py b/megatron/core/tensor_parallel/cross_entropy.py index d0453d25eae..ba144c399e2 100644 --- a/megatron/core/tensor_parallel/cross_entropy.py +++ b/megatron/core/tensor_parallel/cross_entropy.py @@ -5,7 +5,7 @@ from megatron.core.parallel_state import ( get_tensor_model_parallel_group, get_tensor_model_parallel_rank, - get_tensor_model_parallel_world_size + get_tensor_model_parallel_world_size, ) from .utils import VocabUtility @@ -18,9 +18,11 @@ def forward(ctx, vocab_parallel_logits, target, label_smoothing=0.0): # Maximum value along vocab dimension across all GPUs. logits_max = torch.max(vocab_parallel_logits, dim=-1)[0] - torch.distributed.all_reduce(logits_max, - op=torch.distributed.ReduceOp.MAX, - group=get_tensor_model_parallel_group()) + torch.distributed.all_reduce( + logits_max, + op=torch.distributed.ReduceOp.MAX, + group=get_tensor_model_parallel_group(), + ) # Subtract the maximum value. vocab_parallel_logits = vocab_parallel_logits - logits_max.unsqueeze(dim=-1) @@ -30,7 +32,8 @@ def forward(ctx, vocab_parallel_logits, target, label_smoothing=0.0): rank = get_tensor_model_parallel_rank() world_size = get_tensor_model_parallel_world_size() vocab_start_index, vocab_end_index = get_vocab_range( - partition_vocab_size, rank, world_size) + partition_vocab_size, rank, world_size + ) # Create a mask of valid vocab ids (1 means it needs to be masked). target_mask = (target < vocab_start_index) | (target >= vocab_end_index) @@ -42,24 +45,29 @@ def forward(ctx, vocab_parallel_logits, target, label_smoothing=0.0): # [*, partition-vocab-size] and target to a 1-D tensor of size [*]. logits_2d = vocab_parallel_logits.view(-1, partition_vocab_size) masked_target_1d = masked_target.view(-1) - arange_1d = torch.arange(start=0, end=logits_2d.size()[0], - device=logits_2d.device) + arange_1d = torch.arange( + start=0, end=logits_2d.size()[0], device=logits_2d.device + ) predicted_logits_1d = logits_2d[arange_1d, masked_target_1d] predicted_logits_1d = predicted_logits_1d.clone().contiguous() predicted_logits = predicted_logits_1d.view_as(target) predicted_logits[target_mask] = 0.0 # All reduce is needed to get the chunks from other GPUs. - torch.distributed.all_reduce(predicted_logits, - op=torch.distributed.ReduceOp.SUM, - group=get_tensor_model_parallel_group()) + torch.distributed.all_reduce( + predicted_logits, + op=torch.distributed.ReduceOp.SUM, + group=get_tensor_model_parallel_group(), + ) # Sum of exponential of logits along vocab dimension across all GPUs. exp_logits = vocab_parallel_logits torch.exp(vocab_parallel_logits, out=exp_logits) sum_exp_logits = exp_logits.sum(dim=-1) - torch.distributed.all_reduce(sum_exp_logits, - op=torch.distributed.ReduceOp.SUM, - group=get_tensor_model_parallel_group()) + torch.distributed.all_reduce( + sum_exp_logits, + op=torch.distributed.ReduceOp.SUM, + group=get_tensor_model_parallel_group(), + ) # Loss = log(sum(exp(logits))) - predicted-logit. loss = torch.log(sum_exp_logits) - predicted_logits @@ -107,8 +115,7 @@ def backward(ctx, grad_output): grad_2d = grad_input.view(-1, partition_vocab_size) # Add the gradient from matching classes. - arange_1d = torch.arange(start=0, end=grad_2d.size()[0], - device=grad_2d.device) + arange_1d = torch.arange(start=0, end=grad_2d.size()[0], device=grad_2d.device) softmax_update = 1.0 - target_mask.view(-1).float() @@ -139,4 +146,6 @@ def vocab_parallel_cross_entropy(vocab_parallel_logits, target, label_smoothing= lobal_smoothing: smoothing factor, must be in range [0.0, 1.0) default is no smoothing (=0.0) """ - return _VocabParallelCrossEntropy.apply(vocab_parallel_logits, target, label_smoothing) + return _VocabParallelCrossEntropy.apply( + vocab_parallel_logits, target, label_smoothing + ) diff --git a/megatron/core/tensor_parallel/data.py b/megatron/core/tensor_parallel/data.py index 4de1a8b65b3..b9e50189189 100644 --- a/megatron/core/tensor_parallel/data.py +++ b/megatron/core/tensor_parallel/data.py @@ -19,8 +19,11 @@ def _check_data_types(keys, data, target_dtype): """Check that all the keys have the same target data type.""" for key in keys: - assert data[key].dtype == target_dtype, '{} has data type {} which '\ - 'is different than {}'.format(key, data[key].dtype, target_dtype) + assert ( + data[key].dtype == target_dtype + ), "{} has data type {} which " "is different than {}".format( + key, data[key].dtype, target_dtype + ) def _build_key_size_numel_dictionaries(keys, data, group=None, rank=-1, src_rank=-1): @@ -30,7 +33,7 @@ def _build_key_size_numel_dictionaries(keys, data, group=None, rank=-1, src_rank src_rank = get_tensor_model_parallel_src_rank() if rank < 0: rank = get_tensor_model_parallel_rank() - + """Build the size on rank 0 and broadcast.""" max_dim = _MAX_DATA_DIM sizes = [0 for _ in range(max_dim) for _ in keys] @@ -39,7 +42,7 @@ def _build_key_size_numel_dictionaries(keys, data, group=None, rank=-1, src_rank if rank == 0: offset = 0 for key in keys: - assert data[key].dim() < max_dim, 'you should increase MAX_DATA_DIM' + assert data[key].dim() < max_dim, "you should increase MAX_DATA_DIM" size = data[key].size() for i, s in enumerate(size): sizes[i + offset] = s @@ -94,7 +97,8 @@ def broadcast_data(keys, data, datatype): group = get_tensor_model_parallel_group() key_size, key_numel, total_numel = _build_key_size_numel_dictionaries( - keys, data, group=group, rank=rank, src_rank=src_rank) + keys, data, group=group, rank=rank, src_rank=src_rank + ) # Pack on rank zero. if rank == 0: @@ -102,11 +106,12 @@ def broadcast_data(keys, data, datatype): _check_data_types(keys, data, datatype) # Flatten the data associated with the keys flatten_data = torch.cat( - [data[key].contiguous().view(-1) for key in keys], dim=0).to(get_accelerator().device_name()) + [data[key].contiguous().view(-1) for key in keys], dim=0 + ).to(get_accelerator().device_name()) else: - flatten_data = torch.empty(total_numel, - device=get_accelerator().current_device_name(), - dtype=datatype) + flatten_data = torch.empty( + total_numel, device=get_accelerator().current_device_name(), dtype=datatype + ) # Broadcast torch.distributed.broadcast(flatten_data, src_rank, group=group) diff --git a/megatron/core/tensor_parallel/layers.py b/megatron/core/tensor_parallel/layers.py index 3dd3299ae08..f82fd3cd250 100644 --- a/megatron/core/tensor_parallel/layers.py +++ b/megatron/core/tensor_parallel/layers.py @@ -52,14 +52,17 @@ except ImportError: _grad_accum_fusion_available = False -_MODEL_PARALLEL_ATTRIBUTE_DEFAULTS = {'tensor_model_parallel': False, - 'partition_dim': -1, - 'partition_stride': 1} +_MODEL_PARALLEL_ATTRIBUTE_DEFAULTS = { + "tensor_model_parallel": False, + "partition_dim": -1, + "partition_stride": 1, +} + def param_is_not_tensor_parallel_duplicate(param): - return (hasattr(param, 'tensor_model_parallel') and - param.tensor_model_parallel) or ( - get_tensor_model_parallel_rank() == 0) + return ( + hasattr(param, "tensor_model_parallel") and param.tensor_model_parallel + ) or (get_tensor_model_parallel_rank() == 0) def set_tensor_model_parallel_attributes(tensor, is_parallel, dim, stride): @@ -67,15 +70,16 @@ def set_tensor_model_parallel_attributes(tensor, is_parallel, dim, stride): for attribute in _MODEL_PARALLEL_ATTRIBUTE_DEFAULTS: assert not hasattr(tensor, attribute) # Set the attributes. - setattr(tensor, 'tensor_model_parallel', is_parallel) - setattr(tensor, 'partition_dim', dim) - setattr(tensor, 'partition_stride', stride) + setattr(tensor, "tensor_model_parallel", is_parallel) + setattr(tensor, "partition_dim", dim) + setattr(tensor, "partition_stride", stride) def set_defaults_if_not_set_tensor_model_parallel_attributes(tensor): def maybe_set(attribute, value): if not hasattr(tensor, attribute): setattr(tensor, attribute, value) + for attribute in _MODEL_PARALLEL_ATTRIBUTE_DEFAULTS: maybe_set(attribute, _MODEL_PARALLEL_ATTRIBUTE_DEFAULTS[attribute]) @@ -83,51 +87,56 @@ def maybe_set(attribute, value): def copy_tensor_model_parallel_attributes(destination_tensor, source_tensor): def maybe_copy(attribute): if hasattr(source_tensor, attribute): - setattr(destination_tensor, attribute, - getattr(source_tensor, attribute)) + setattr(destination_tensor, attribute, getattr(source_tensor, attribute)) + for attribute in _MODEL_PARALLEL_ATTRIBUTE_DEFAULTS: maybe_copy(attribute) -def _initialize_affine_weight_gpu(weight, init_method, - partition_dim, stride=1): +def _initialize_affine_weight_gpu(weight, init_method, partition_dim, stride=1): """Initialize affine weight for model parallel on GPU.""" - set_tensor_model_parallel_attributes(tensor=weight, - is_parallel=True, - dim=partition_dim, - stride=stride) + set_tensor_model_parallel_attributes( + tensor=weight, is_parallel=True, dim=partition_dim, stride=stride + ) with get_cuda_rng_tracker().fork(): init_method(weight) -def _initialize_affine_weight_cpu(weight, output_size, input_size, - per_partition_size, partition_dim, - init_method, stride=1, - return_master_weight=False, - *, params_dtype=torch.float32): +def _initialize_affine_weight_cpu( + weight, + output_size, + input_size, + per_partition_size, + partition_dim, + init_method, + stride=1, + return_master_weight=False, + *, + params_dtype=torch.float32, +): """Initialize affine weight for model parallel. Build the master weight on all processes and scatter the relevant chunk.""" - set_tensor_model_parallel_attributes(tensor=weight, - is_parallel=True, - dim=partition_dim, - stride=stride) + set_tensor_model_parallel_attributes( + tensor=weight, is_parallel=True, dim=partition_dim, stride=stride + ) # Initialize master weight - master_weight = torch.empty(output_size, input_size, - dtype=torch.float, - requires_grad=False) + master_weight = torch.empty( + output_size, input_size, dtype=torch.float, requires_grad=False + ) init_method(master_weight) master_weight = master_weight.to(dtype=params_dtype) # Split and copy per_partition_per_stride_size = divide(per_partition_size, stride) - weight_list = torch.split(master_weight, per_partition_per_stride_size, - dim=partition_dim) + weight_list = torch.split( + master_weight, per_partition_per_stride_size, dim=partition_dim + ) rank = get_tensor_model_parallel_rank() world_size = get_tensor_model_parallel_world_size() my_weight_list = weight_list[rank::world_size] @@ -152,9 +161,14 @@ class VocabParallelEmbedding(torch.nn.Module): config: A megatron.core.ModelParallelConfig object """ - def __init__(self, num_embeddings: int, embedding_dim: int, *, - init_method: Callable, - config: ModelParallelConfig): + def __init__( + self, + num_embeddings: int, + embedding_dim: int, + *, + init_method: Callable, + config: ModelParallelConfig, + ): super(VocabParallelEmbedding, self).__init__() # Keep the input dimensions. self.num_embeddings = num_embeddings @@ -162,52 +176,77 @@ def __init__(self, num_embeddings: int, embedding_dim: int, *, # Set the detauls for compatibility. self.padding_idx = None self.max_norm = None - self.norm_type = 2. + self.norm_type = 2.0 self.scale_grad_by_freq = False self.sparse = False self._weight = None self.tensor_model_parallel_size = get_tensor_model_parallel_world_size() # Divide the weight matrix along the vocaburaly dimension. - self.vocab_start_index, self.vocab_end_index = \ + self.vocab_start_index, self.vocab_end_index = ( VocabUtility.vocab_range_from_global_vocab_size( - self.num_embeddings, get_tensor_model_parallel_rank(), - self.tensor_model_parallel_size) - self.num_embeddings_per_partition = self.vocab_end_index - \ - self.vocab_start_index + self.num_embeddings, + get_tensor_model_parallel_rank(), + self.tensor_model_parallel_size, + ) + ) + self.num_embeddings_per_partition = ( + self.vocab_end_index - self.vocab_start_index + ) # Allocate weights and initialize. if config.use_cpu_initialization: - self.weight = Parameter(torch.empty( - self.num_embeddings_per_partition, self.embedding_dim, - dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.num_embeddings_per_partition, + self.embedding_dim, + dtype=config.params_dtype, + ) + ) if config.perform_initialization: _initialize_affine_weight_cpu( - self.weight, self.num_embeddings, self.embedding_dim, - self.num_embeddings_per_partition, 0, init_method, - params_dtype=config.params_dtype) + self.weight, + self.num_embeddings, + self.embedding_dim, + self.num_embeddings_per_partition, + 0, + init_method, + params_dtype=config.params_dtype, + ) else: - self.weight = Parameter(torch.empty( - self.num_embeddings_per_partition, self.embedding_dim, - device=get_accelerator().current_device_name(), dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.num_embeddings_per_partition, + self.embedding_dim, + device=get_accelerator().current_device_name(), + dtype=config.params_dtype, + ) + ) if config.perform_initialization: - _initialize_affine_weight_gpu(self.weight, init_method, - partition_dim=0, stride=1) + _initialize_affine_weight_gpu( + self.weight, init_method, partition_dim=0, stride=1 + ) def forward(self, input_): if self.tensor_model_parallel_size > 1: # Build the mask. - input_mask = (input_ < self.vocab_start_index) | \ - (input_ >= self.vocab_end_index) + input_mask = (input_ < self.vocab_start_index) | ( + input_ >= self.vocab_end_index + ) # Mask the input. masked_input = input_.clone() - self.vocab_start_index masked_input[input_mask] = 0 else: masked_input = input_ # Get the embeddings. - output_parallel = F.embedding(masked_input, self.weight, - self.padding_idx, self.max_norm, - self.norm_type, self.scale_grad_by_freq, - self.sparse) + output_parallel = F.embedding( + masked_input, + self.weight, + self.padding_idx, + self.max_norm, + self.norm_type, + self.scale_grad_by_freq, + self.sparse, + ) # Mask the output embedding. if self.tensor_model_parallel_size > 1: output_parallel[input_mask, :] = 0.0 @@ -230,25 +269,33 @@ def __init__(self, sequence_length, embedding_dim): assert sequence_length % sequence_parallel_size == 0 local_sequence_length = sequence_length // sequence_parallel_size self.offset = local_sequence_length * get_tensor_model_parallel_rank() - self.local_embeddings = torch.nn.Embedding( - local_sequence_length, embedding_dim) + self.local_embeddings = torch.nn.Embedding(local_sequence_length, embedding_dim) def forward(self, position_ids): return self.local_embeddings(position_ids - self.offset) + def gradientUpdateFunction(total_input, grad_output, weight): if weight.grad == None: weight.grad = grad_output.t().matmul(total_input) else: weight.grad += grad_output.t().matmul(total_input) + class LinearWithGradAccumulationAndAsyncCommunication(torch.autograd.Function): """See linear_with_grad_accumulation_and_async_allreduce""" @staticmethod @custom_fwd - def forward(ctx, input, weight, bias, gradient_accumulation_fusion, - async_grad_allreduce, sequence_parallel): + def forward( + ctx, + input, + weight, + bias, + gradient_accumulation_fusion, + async_grad_allreduce, + sequence_parallel, + ): ctx.save_for_backward(input, weight) ctx.use_bias = bias is not None ctx.gradient_accumulation_fusion = gradient_accumulation_fusion @@ -260,19 +307,18 @@ def forward(ctx, input, weight, bias, gradient_accumulation_fusion, dim_size = list(input.size()) dim_size[0] = dim_size[0] * world_size - all_gather_buffer = \ - get_global_memory_buffer().get_tensor(dim_size, input.dtype, "mpu") + all_gather_buffer = get_global_memory_buffer().get_tensor( + dim_size, input.dtype, "mpu" + ) - if version.parse(torch.__version__) >= version.parse('1.13'): + if version.parse(torch.__version__) >= version.parse("1.13"): torch.distributed.all_gather_into_tensor( - all_gather_buffer, - input, - group=get_tensor_model_parallel_group()) + all_gather_buffer, input, group=get_tensor_model_parallel_group() + ) else: torch.distributed._all_gather_base( - all_gather_buffer, - input, - group=get_tensor_model_parallel_group()) + all_gather_buffer, input, group=get_tensor_model_parallel_group() + ) total_input = all_gather_buffer else: @@ -295,19 +341,24 @@ def backward(ctx, grad_output): dim_size = list(input.size()) dim_size[0] = dim_size[0] * world_size - all_gather_buffer = \ - get_global_memory_buffer().get_tensor(dim_size, input.dtype, "mpu") + all_gather_buffer = get_global_memory_buffer().get_tensor( + dim_size, input.dtype, "mpu" + ) - if version.parse(torch.__version__) >= version.parse('1.13'): + if version.parse(torch.__version__) >= version.parse("1.13"): handle = torch.distributed.all_gather_into_tensor( all_gather_buffer, input, - group=get_tensor_model_parallel_group(), async_op=True) + group=get_tensor_model_parallel_group(), + async_op=True, + ) else: handle = torch.distributed._all_gather_base( all_gather_buffer, input, - group=get_tensor_model_parallel_group(), async_op=True) + group=get_tensor_model_parallel_group(), + async_op=True, + ) # Here we rely on CUDA_DEVICE_MAX_CONNECTIONS=1 to ensure that the # gather is scheduled before the input gradient computation @@ -326,10 +377,12 @@ def backward(ctx, grad_output): grad_output = grad_output.contiguous() # Convert the tensor shapes to 2D for execution compatibility if len(grad_output.shape) == 3: - grad_output = grad_output.view(grad_output.shape[0] * grad_output.shape[1], - grad_output.shape[2]) - total_input = total_input.view(total_input.shape[0] * total_input.shape[1], - total_input.shape[2]) + grad_output = grad_output.view( + grad_output.shape[0] * grad_output.shape[1], grad_output.shape[2] + ) + total_input = total_input.view( + total_input.shape[0] * total_input.shape[1], total_input.shape[2] + ) else: # Somehow when DeepSpeed MoE is used, grad_output could have 4 dimensions. # TODO: May need further investigation @@ -340,20 +393,27 @@ def backward(ctx, grad_output): if ctx.async_grad_allreduce: # Asynchronous all-reduce handle = torch.distributed.all_reduce( - grad_input, group=get_tensor_model_parallel_group(), async_op=True) + grad_input, group=get_tensor_model_parallel_group(), async_op=True + ) # Here we rely on CUDA_DEVICE_MAX_CONNECTIONS=1 to ensure that the # all-reduce is scheduled before the weight gradient computation if ctx.sequence_parallel: assert not ctx.async_grad_allreduce dim_size = list(input.size()) - sub_grad_input = torch.empty(dim_size, dtype=input.dtype, - device=get_accelerator().current_device_name(), - requires_grad=False) + sub_grad_input = torch.empty( + dim_size, + dtype=input.dtype, + device=get_accelerator().current_device_name(), + requires_grad=False, + ) # reduce_scatter - handle = torch.distributed._reduce_scatter_base(sub_grad_input, grad_input, - group=get_tensor_model_parallel_group(), - async_op=True) + handle = torch.distributed._reduce_scatter_base( + sub_grad_input, + grad_input, + group=get_tensor_model_parallel_group(), + async_op=True, + ) # Here we rely on CUDA_DEVICE_MAX_CONNECTIONS=1 to ensure that the # reduce scatter is scheduled before the weight gradient computation @@ -370,7 +430,10 @@ def backward(ctx, grad_output): # grad_weight = grad_output.t().matmul(total_input) if args.enable_zbh1_pipeline: from megatron.core.tensor_parallel.weight_grad_store import WeightGradStore - WeightGradStore.put(total_input, grad_output, weight, gradientUpdateFunction) + + WeightGradStore.put( + total_input, grad_output, weight, gradientUpdateFunction + ) grad_weight = None else: grad_weight = grad_output.t().matmul(total_input) @@ -386,6 +449,7 @@ def backward(ctx, grad_output): return grad_input, grad_weight, grad_bias, None, None, None + def linear_with_grad_accumulation_and_async_allreduce( input: torch.Tensor, weight: torch.Tensor, @@ -456,26 +520,32 @@ def linear_with_grad_accumulation_and_async_allreduce( ] if not linear_with_grad_accumulation_and_async_allreduce.warned: - if get_accelerator().device_name() == "cuda" \ - and os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": + if ( + get_accelerator().device_name() == "cuda" + and os.environ.get("CUDA_DEVICE_MAX_CONNECTIONS") != "1" + ): if sequence_parallel: warnings.warn( "When using sequence parallelism it is recommended to set the " "environment variable CUDA_DEVICE_MAX_CONNECTIONS to 1 for " - "maximum speedup") + "maximum speedup" + ) linear_with_grad_accumulation_and_async_allreduce.warned = True if async_grad_allreduce: warnings.warn( "When using async grad allreduce it is recommended to set the " "environment variable CUDA_DEVICE_MAX_CONNECTIONS to 1 for " - "maximum speedup") + "maximum speedup" + ) linear_with_grad_accumulation_and_async_allreduce.warned = True return LinearWithGradAccumulationAndAsyncCommunication.apply(*args) + linear_with_grad_accumulation_and_async_allreduce.warned = False + class ColumnParallelLinear(torch.nn.Module): """Linear layer with column parallelism. @@ -511,14 +581,22 @@ class ColumnParallelLinear(torch.nn.Module): """ - def __init__(self, input_size, output_size, *, - config: ModelParallelConfig, - init_method: Callable, - bias=True, gather_output=False, stride=1, - keep_master_weight_for_test=False, - skip_bias_add=False, - skip_weight_param_allocation: bool=False, - moe=False, enable_expert_tensor_parallelism=False): + def __init__( + self, + input_size, + output_size, + *, + config: ModelParallelConfig, + init_method: Callable, + bias=True, + gather_output=False, + stride=1, + keep_master_weight_for_test=False, + skip_bias_add=False, + skip_weight_param_allocation: bool = False, + moe=False, + enable_expert_tensor_parallelism=False, + ): torch.nn.Module.__init__(self) # Keep input parameters @@ -542,44 +620,66 @@ def __init__(self, input_size, output_size, *, # Initialize weight. if not skip_weight_param_allocation: if config.use_cpu_initialization: - self.weight = Parameter(torch.empty(self.output_size_per_partition, - self.input_size, - dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.output_size_per_partition, + self.input_size, + dtype=config.params_dtype, + ) + ) if config.perform_initialization: self.master_weight = _initialize_affine_weight_cpu( - self.weight, self.output_size, self.input_size, - self.output_size_per_partition, 0, init_method, - stride=stride, return_master_weight=keep_master_weight_for_test) + self.weight, + self.output_size, + self.input_size, + self.output_size_per_partition, + 0, + init_method, + stride=stride, + return_master_weight=keep_master_weight_for_test, + ) else: - self.weight = Parameter(torch.empty( - self.output_size_per_partition, self.input_size, - device=get_accelerator().current_device_name(), dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.output_size_per_partition, + self.input_size, + device=get_accelerator().current_device_name(), + dtype=config.params_dtype, + ) + ) if config.perform_initialization: - _initialize_affine_weight_gpu(self.weight, init_method, - partition_dim=0, stride=stride) + _initialize_affine_weight_gpu( + self.weight, init_method, partition_dim=0, stride=stride + ) else: self.weight = None if bias: if config.use_cpu_initialization: - self.bias = Parameter(torch.empty( - self.output_size_per_partition, dtype=config.params_dtype)) + self.bias = Parameter( + torch.empty( + self.output_size_per_partition, dtype=config.params_dtype + ) + ) else: - self.bias = Parameter(torch.empty( - self.output_size_per_partition, - device=get_accelerator().current_device_name(), - dtype=config.params_dtype)) + self.bias = Parameter( + torch.empty( + self.output_size_per_partition, + device=get_accelerator().current_device_name(), + dtype=config.params_dtype, + ) + ) set_tensor_model_parallel_attributes(self.bias, True, 0, stride) if config.perform_initialization: # Always initialize bias to zero. with torch.no_grad(): self.bias.zero_() else: - self.register_parameter('bias', None) + self.register_parameter("bias", None) self.async_tensor_model_parallel_allreduce = ( - config.async_tensor_model_parallel_allreduce and - world_size > 1) + config.async_tensor_model_parallel_allreduce and world_size > 1 + ) self.sequence_parallel = config.sequence_parallel if self.sequence_parallel and world_size <= 1: @@ -595,7 +695,7 @@ def __init__(self, input_size, output_size, *, "to True but the custom CUDA extension fused_weight_gradient_mlp_cuda " "module is not found. To use gradient_accumulation_fusion you must " "install APEX with --cpp_ext and --cuda_ext. For example: " - "pip install --global-option=\"--cpp_ext\" --global-option=\"--cuda_ext .\" " + 'pip install --global-option="--cpp_ext" --global-option="--cuda_ext ." ' "Note that the extension requires CUDA>=11. Otherwise, you must turn off " "gradient accumulation fusion." ) @@ -607,10 +707,7 @@ def __init__(self, input_size, output_size, *, "cannot be enabled at the same time." ) - - def forward(self, - input_: torch.Tensor, - weight: Optional[torch.Tensor] = None): + def forward(self, input_: torch.Tensor, weight: Optional[torch.Tensor] = None): """Forward of ColumnParallelLinear Args: @@ -626,21 +723,27 @@ def forward(self, """ if weight is None: if self.weight is None: - raise RuntimeError("weight was not supplied to ColumnParallelLinear forward pass " - "and skip_weight_param_allocation is True.") + raise RuntimeError( + "weight was not supplied to ColumnParallelLinear forward pass " + "and skip_weight_param_allocation is True." + ) weight = self.weight else: # Check the weight passed in is the correct shape expected_shape = (self.output_size_per_partition, self.input_size) if weight.shape != expected_shape: - raise RuntimeError(f"supplied weight's shape is {tuple(weight.shape)}, " - f"not {expected_shape} as expected") + raise RuntimeError( + f"supplied weight's shape is {tuple(weight.shape)}, " + f"not {expected_shape} as expected" + ) bias = self.bias if not self.skip_bias_add else None - if self.async_tensor_model_parallel_allreduce or \ - self.sequence_parallel or \ - self.is_expert_without_slicing: # non-expert only tensor parallelism + if ( + self.async_tensor_model_parallel_allreduce + or self.sequence_parallel + or self.is_expert_without_slicing + ): # non-expert only tensor parallelism input_parallel = input_ else: input_parallel = copy_to_tensor_model_parallel_region(input_) @@ -651,7 +754,7 @@ def forward(self, bias=bias, gradient_accumulation_fusion=self.gradient_accumulation_fusion, async_grad_allreduce=self.async_tensor_model_parallel_allreduce, - sequence_parallel=self.sequence_parallel + sequence_parallel=self.sequence_parallel, ) if self.gather_output and not self.is_expert_without_slicing: # All-gather across the partitions. @@ -698,15 +801,21 @@ class RowParallelLinear(torch.nn.Module): """ - def __init__(self, input_size: int, output_size: int, *, - config: ModelParallelConfig, - init_method: Callable, - bias: bool = True, - input_is_parallel: bool = False, - stride: int = 1, - keep_master_weight_for_test: bool = False, - skip_bias_add: bool = False, - moe=False, enable_expert_tensor_parallelism=False): + def __init__( + self, + input_size: int, + output_size: int, + *, + config: ModelParallelConfig, + init_method: Callable, + bias: bool = True, + input_is_parallel: bool = False, + stride: int = 1, + keep_master_weight_for_test: bool = False, + skip_bias_add: bool = False, + moe=False, + enable_expert_tensor_parallelism=False, + ): torch.nn.Module.__init__(self) # Keep input parameters @@ -718,54 +827,75 @@ def __init__(self, input_size: int, output_size: int, *, world_size = 1 else: world_size = get_tensor_model_parallel_world_size() - self.is_expert_without_slicing = moe and world_size==1 + self.is_expert_without_slicing = moe and world_size == 1 self.input_size_per_partition = divide(input_size, world_size) self.skip_bias_add = skip_bias_add self.config = config self.gradient_accumulation_fusion = config.gradient_accumulation_fusion self.sequence_parallel = config.sequence_parallel if self.sequence_parallel and not self.input_is_parallel: - raise RuntimeError("To enable `sequence_parallel`, `input_is_parallel` must be `True`") + raise RuntimeError( + "To enable `sequence_parallel`, `input_is_parallel` must be `True`" + ) # Parameters. # Note: torch.nn.functional.linear performs XA^T + b and as a result # we allocate the transpose. # Initialize weight. if config.use_cpu_initialization: - self.weight = Parameter(torch.empty(self.output_size, - self.input_size_per_partition, - dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.output_size, + self.input_size_per_partition, + dtype=config.params_dtype, + ) + ) if config.perform_initialization: self.master_weight = _initialize_affine_weight_cpu( - self.weight, self.output_size, self.input_size, - self.input_size_per_partition, 1, init_method, - stride=stride, return_master_weight=keep_master_weight_for_test, - params_dtype=config.params_dtype) + self.weight, + self.output_size, + self.input_size, + self.input_size_per_partition, + 1, + init_method, + stride=stride, + return_master_weight=keep_master_weight_for_test, + params_dtype=config.params_dtype, + ) else: - self.weight = Parameter(torch.empty( - self.output_size, self.input_size_per_partition, - device=get_accelerator().current_device_name(), dtype=config.params_dtype)) + self.weight = Parameter( + torch.empty( + self.output_size, + self.input_size_per_partition, + device=get_accelerator().current_device_name(), + dtype=config.params_dtype, + ) + ) if config.perform_initialization: - _initialize_affine_weight_gpu(self.weight, init_method, - partition_dim=1, stride=stride) + _initialize_affine_weight_gpu( + self.weight, init_method, partition_dim=1, stride=stride + ) if bias: if config.use_cpu_initialization: - self.bias = Parameter(torch.empty(self.output_size, - dtype=config.params_dtype)) + self.bias = Parameter( + torch.empty(self.output_size, dtype=config.params_dtype) + ) else: - self.bias = Parameter(torch.empty( - self.output_size, device=get_accelerator().current_device_name(), - dtype=config.params_dtype)) - setattr(self.bias, 'sequence_parallel', self.sequence_parallel) + self.bias = Parameter( + torch.empty( + self.output_size, + device=get_accelerator().current_device_name(), + dtype=config.params_dtype, + ) + ) + setattr(self.bias, "sequence_parallel", self.sequence_parallel) if config.perform_initialization: # Always initialize bias to zero. with torch.no_grad(): self.bias.zero_() else: - self.register_parameter('bias', None) - - + self.register_parameter("bias", None) def forward(self, input_): """Forward of RowParallelLinear @@ -796,7 +926,7 @@ def forward(self, input_): # All-reduce across all the partitions. if self.sequence_parallel: output_ = reduce_scatter_to_sequence_parallel_region(output_parallel) - elif self.is_expert_without_slicing: # non-expert only tensor-parallelism + elif self.is_expert_without_slicing: # non-expert only tensor-parallelism output_ = output_parallel else: output_ = reduce_from_tensor_model_parallel_region(output_parallel) diff --git a/megatron/core/tensor_parallel/mappings.py b/megatron/core/tensor_parallel/mappings.py index ae8d63ab2c2..1027f51330e 100644 --- a/megatron/core/tensor_parallel/mappings.py +++ b/megatron/core/tensor_parallel/mappings.py @@ -15,7 +15,7 @@ def _reduce(input_): """All-reduce the input tensor across model parallel group.""" # Bypass the function if we are using only 1 GPU. - if get_tensor_model_parallel_world_size()==1: + if get_tensor_model_parallel_world_size() == 1: return input_ # All-reduce. @@ -54,13 +54,14 @@ def _split_along_first_dim(input_): # Split along first dimension. dim_size = input_.size()[0] - assert dim_size % world_size == 0, \ - "First dimension of the tensor should be divisible by tensor parallel size" + assert ( + dim_size % world_size == 0 + ), "First dimension of the tensor should be divisible by tensor parallel size" local_dim_size = dim_size // world_size rank = get_tensor_model_parallel_rank() dim_offset = rank * local_dim_size - output = input_[dim_offset:dim_offset+local_dim_size].contiguous() + output = input_[dim_offset : dim_offset + local_dim_size].contiguous() return output @@ -79,7 +80,9 @@ def _gather_along_last_dim(input_): tensor_list = [torch.empty_like(input_) for _ in range(world_size)] tensor_list[rank] = input_ - torch.distributed.all_gather(tensor_list, input_, group=get_tensor_model_parallel_group()) + torch.distributed.all_gather( + tensor_list, input_, group=get_tensor_model_parallel_group() + ) # Note: torch.cat already creates a contiguous tensor. output = torch.cat(tensor_list, dim=last_dim).contiguous() @@ -98,13 +101,16 @@ def _gather_along_first_dim(input_): dim_size = list(input_.size()) dim_size[0] = dim_size[0] * world_size - output = torch.empty(dim_size, dtype=input_.dtype, - device=get_accelerator().current_device_name()) - torch.distributed._all_gather_base(output, input_.contiguous(), - group=get_tensor_model_parallel_group()) + output = torch.empty( + dim_size, dtype=input_.dtype, device=get_accelerator().current_device_name() + ) + torch.distributed._all_gather_base( + output, input_.contiguous(), group=get_tensor_model_parallel_group() + ) return output + def _reduce_scatter_along_first_dim(input_): """Reduce-scatter the input tensor across model parallel group.""" world_size = get_tensor_model_parallel_world_size() @@ -113,15 +119,18 @@ def _reduce_scatter_along_first_dim(input_): return input_ dim_size = list(input_.size()) - assert dim_size[0] % world_size == 0, \ - "First dimension of the tensor should be divisible by tensor parallel size" - + assert ( + dim_size[0] % world_size == 0 + ), "First dimension of the tensor should be divisible by tensor parallel size" + dim_size[0] = dim_size[0] // world_size - - output = torch.empty(dim_size, dtype=input_.dtype, - device=get_accelerator().current_device_name()) - torch.distributed._reduce_scatter_base(output, input_.contiguous(), - group=get_tensor_model_parallel_group()) + + output = torch.empty( + dim_size, dtype=input_.dtype, device=get_accelerator().current_device_name() + ) + torch.distributed._reduce_scatter_base( + output, input_.contiguous(), group=get_tensor_model_parallel_group() + ) return output @@ -131,7 +140,7 @@ class _CopyToModelParallelRegion(torch.autograd.Function): @staticmethod def symbolic(graph, input_): return input_ - + @staticmethod def forward(ctx, input_): return input_ @@ -147,7 +156,7 @@ class _ReduceFromModelParallelRegion(torch.autograd.Function): @staticmethod def symbolic(graph, input_): return _reduce(input_) - + @staticmethod def forward(ctx, input_): return _reduce(input_) @@ -179,7 +188,7 @@ class _GatherFromModelParallelRegion(torch.autograd.Function): @staticmethod def symbolic(graph, input_): return _gather_along_last_dim(input_) - + @staticmethod def forward(ctx, input_): return _gather_along_last_dim(input_) @@ -206,12 +215,12 @@ def backward(ctx, grad_output): class _GatherFromSequenceParallelRegion(torch.autograd.Function): - """Gather the input from sequence parallel region and concatinate.""" + """Gather the input from sequence parallel region and concatinate.""" @staticmethod def symbolic(graph, input_, tensor_parallel_output_grad=True): return _gather_along_first_dim(input_) - + @staticmethod def forward(ctx, input_, tensor_parallel_output_grad=True): ctx.tensor_parallel_output_grad = tensor_parallel_output_grad @@ -222,8 +231,8 @@ def backward(ctx, grad_output): tensor_parallel_output_grad = ctx.tensor_parallel_output_grad # If the computation graph after the gather operation is - # in the tensor parallel mode, output gradients need to reduce - # scattered and whereas if the computation is duplicated, + # in the tensor parallel mode, output gradients need to reduce + # scattered and whereas if the computation is duplicated, # output gradients need to be scattered. if tensor_parallel_output_grad: return _reduce_scatter_along_first_dim(grad_output), None @@ -237,7 +246,7 @@ class _ReduceScatterToSequenceParallelRegion(torch.autograd.Function): @staticmethod def symbolic(graph, input_): return _reduce_scatter_along_first_dim(input_) - + @staticmethod def forward(ctx, input_): return _reduce_scatter_along_first_dim(input_) @@ -251,6 +260,7 @@ def backward(ctx, grad_output): # Helper functions. # ----------------- + def copy_to_tensor_model_parallel_region(input_): return _CopyToModelParallelRegion.apply(input_) @@ -277,4 +287,3 @@ def gather_from_sequence_parallel_region(input_, tensor_parallel_output_grad=Tru def reduce_scatter_to_sequence_parallel_region(input_): return _ReduceScatterToSequenceParallelRegion.apply(input_) - diff --git a/megatron/core/tensor_parallel/random.py b/megatron/core/tensor_parallel/random.py index dd287115a59..ea8693b79f0 100644 --- a/megatron/core/tensor_parallel/random.py +++ b/megatron/core/tensor_parallel/random.py @@ -29,7 +29,7 @@ import deepspeed # Default name for the model parallel rng tracker. -_MODEL_PARALLEL_RNG_TRACKER_NAME = 'model-parallel-rng' +_MODEL_PARALLEL_RNG_TRACKER_NAME = "model-parallel-rng" # Whether apply model parallelsim to checkpointed hidden states. _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER = None @@ -39,10 +39,15 @@ def init_checkpointed_activations_memory_buffer(): """Initializ the memory buffer for the checkpointed activations.""" args = get_args() - per_layer = args.micro_batch_size * args.max_position_embeddings * \ - args.hidden_size // args.tensor_model_parallel_size - assert args.num_layers % args.checkpoint_num_layers == 0, \ - 'number of layers is not divisible by checkpoint-num-layers' + per_layer = ( + args.micro_batch_size + * args.max_position_embeddings + * args.hidden_size + // args.tensor_model_parallel_size + ) + assert ( + args.num_layers % args.checkpoint_num_layers == 0 + ), "number of layers is not divisible by checkpoint-num-layers" num_checkpointer_layers = args.num_layers // args.checkpoint_num_layers numel = per_layer * num_checkpointer_layers dtype = torch.half @@ -50,10 +55,12 @@ def init_checkpointed_activations_memory_buffer(): dtype = torch.float global _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER - assert _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER is None, \ - 'checkpointed activations memory buffer is already allocated.' + assert ( + _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER is None + ), "checkpointed activations memory buffer is already allocated." _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER = allocate_mem_buff( - 'checkpointed activations', numel, dtype, track_usage=False) + "checkpointed activations", numel, dtype, track_usage=False + ) def reset_checkpointed_activations_memory_buffer(): @@ -61,6 +68,7 @@ def reset_checkpointed_activations_memory_buffer(): if _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER is not None: _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER.reset() + def _set_cuda_rng_state(new_state, device=-1): """Sets the random number generator state of the current GPU. @@ -70,11 +78,12 @@ def _set_cuda_rng_state(new_state, device=-1): with a single change: the input state is not cloned. Cloning caused major performance issues for +4 GPU cases. """ - if hasattr(_C, '_cuda_setRNGState') and callable(_C._cuda_setRNGState): + if hasattr(_C, "_cuda_setRNGState") and callable(_C._cuda_setRNGState): # older PyTorch def cb(): with get_accelerator().device(device): _C._cuda_setRNGState(new_state) + else: # newer PyTorch if device == -1: @@ -94,7 +103,6 @@ def cb(): get_accelerator().lazy_call(cb) - class CudaRNGStatesTracker: """Tracker for the cuda RNG states. @@ -132,11 +140,11 @@ def add(self, name, seed): """Track the rng state.""" # Check seed is not already used. if seed in self.seeds_: - raise Exception('seed {} already exists'.format(seed)) + raise Exception("seed {} already exists".format(seed)) self.seeds_.add(seed) # Check that state is not already defined. if name in self.states_: - raise Exception('cuda rng state {} already exists'.format(name)) + raise Exception("cuda rng state {} already exists".format(name)) # Get the current rng state. orig_rng_state = get_accelerator().get_rng_state() # Set the new state and store it. @@ -152,7 +160,7 @@ def fork(self, name=_MODEL_PARALLEL_RNG_TRACKER_NAME): # Check if we have added the state if name not in self.states_: print(name, self.states_) - raise Exception('cuda rng state {} is not added'.format(name)) + raise Exception("cuda rng state {} is not added".format(name)) # Store current rng state. orig_cuda_rng_state = get_accelerator().get_rng_state() # Set rng state to the desired one @@ -170,11 +178,12 @@ def fork(self, name=_MODEL_PARALLEL_RNG_TRACKER_NAME): # RNG tracker object. _CUDA_RNG_STATE_TRACKER = CudaRNGStatesTracker() + def get_cuda_rng_tracker(): """Get cuda rng tracker.""" if deepspeed.checkpointing.is_configured(): return deepspeed.checkpointing.get_cuda_rng_tracker() - + return _CUDA_RNG_STATE_TRACKER @@ -197,7 +206,7 @@ def model_parallel_cuda_manual_seed(seed): """ if deepspeed.checkpointing.is_configured(): return deepspeed.checkpointing.model_parallel_cuda_manual_seed(seed) - + # 2718 is just for fun and any POSITIVE value will work. offset = seed + 2718 tensor_model_parallel_seed = offset + get_tensor_model_parallel_rank() @@ -205,18 +214,25 @@ def model_parallel_cuda_manual_seed(seed): data_parallel_seed = seed if torch.distributed.get_rank() == 0: - print('> initializing model parallel cuda seeds on global rank {}, ' - 'model parallel rank {}, and data parallel rank {} with ' - 'model parallel seed: {} and data parallel seed: {}'.format( - torch.distributed.get_rank(), get_tensor_model_parallel_rank(), - get_data_parallel_rank(), tensor_model_parallel_seed, - data_parallel_seed), flush=True) + print( + "> initializing model parallel cuda seeds on global rank {}, " + "model parallel rank {}, and data parallel rank {} with " + "model parallel seed: {} and data parallel seed: {}".format( + torch.distributed.get_rank(), + get_tensor_model_parallel_rank(), + get_data_parallel_rank(), + tensor_model_parallel_seed, + data_parallel_seed, + ), + flush=True, + ) _CUDA_RNG_STATE_TRACKER.reset() # Set the default state. get_accelerator().manual_seed(data_parallel_seed) # and model parallel state. - _CUDA_RNG_STATE_TRACKER.add(_MODEL_PARALLEL_RNG_TRACKER_NAME, - tensor_model_parallel_seed) + _CUDA_RNG_STATE_TRACKER.add( + _MODEL_PARALLEL_RNG_TRACKER_NAME, tensor_model_parallel_seed + ) def model_parallel_reconfigure_tp_seed(seed): @@ -230,16 +246,16 @@ def model_parallel_reconfigure_tp_seed(seed): class CheckpointFunction(torch.autograd.Function): """This function is adapted from torch.utils.checkpoint with - two main changes: - 1) torch.cuda.set_rng_state is replaced with `_set_cuda_rng_state` - 2) the states in the model parallel tracker are also properly - tracked/set/reset. + two main changes: + 1) torch.cuda.set_rng_state is replaced with `_set_cuda_rng_state` + 2) the states in the model parallel tracker are also properly + tracked/set/reset. """ + @staticmethod def forward(ctx, run_function, distribute_saved_activations, *args): ctx.run_function = run_function - ctx.distribute_saved_activations \ - = distribute_saved_activations + ctx.distribute_saved_activations = distribute_saved_activations # Copy the rng states. ctx.fwd_cpu_rng_state = torch.get_rng_state() @@ -255,7 +271,8 @@ def forward(ctx, run_function, distribute_saved_activations, *args): ctx.input_0_shape = args[0].data.shape safely_set_viewless_tensor_data( args[0], - split_tensor_into_1d_equal_chunks(args[0].data, new_buffer=True)) + split_tensor_into_1d_equal_chunks(args[0].data, new_buffer=True), + ) # HACK: currently when DeepSpeed is used, we always set # distribute_saved_activations to false, and use the following older @@ -263,8 +280,7 @@ def forward(ctx, run_function, distribute_saved_activations, *args): if _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER is not None: ctx.input_0_shape = args[0].data.shape args[0].data = split_tensor_into_1d_equal_chunks(args[0].data) - args[0].data = _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER.add( - args[0].data) + args[0].data = _CHECKPOINTED_ACTIVATIONS_MEMORY_BUFFER.add(args[0].data) # Store everything. ctx.save_for_backward(*args) @@ -274,13 +290,16 @@ def forward(ctx, run_function, distribute_saved_activations, *args): @staticmethod def backward(ctx, *args): if not torch.autograd._is_checkpoint_valid(): - raise RuntimeError("Checkpointing is not compatible with .grad(), " - "please use .backward() if possible") + raise RuntimeError( + "Checkpointing is not compatible with .grad(), " + "please use .backward() if possible" + ) inputs = ctx.saved_tensors if ctx.distribute_saved_activations: safely_set_viewless_tensor_data( inputs[0], - gather_split_1d_tensor(inputs[0].data).view(ctx.input_0_shape)) + gather_split_1d_tensor(inputs[0].data).view(ctx.input_0_shape), + ) # HACK: currently when DeepSpeed is used, we always set # distribute_saved_activations to false, and use the following older # activation checkpointing mechanisms @@ -310,13 +329,20 @@ def backward(ctx, *args): if isinstance(outputs, torch.Tensor): outputs = (outputs,) - elif len(outputs) == 2 and isinstance(outputs[1], torch.Tensor) and \ - torch.equal(outputs[1], torch.tensor(0).to(get_accelerator().device_name())): + elif ( + len(outputs) == 2 + and isinstance(outputs[1], torch.Tensor) + and torch.equal( + outputs[1], torch.tensor(0).to(get_accelerator().device_name()) + ) + ): # a hacky solution to overcome issue when running old script examples/pretrain_gpt_distributed.sh outputs = (outputs[0],) torch.autograd.backward(outputs, args) - grads = tuple(inp.grad if isinstance(inp, torch.Tensor) else inp - for inp in detached_inputs) + grads = tuple( + inp.grad if isinstance(inp, torch.Tensor) else inp + for inp in detached_inputs + ) return (None, None) + grads @@ -325,6 +351,5 @@ def checkpoint(function, distribute_saved_activations, *args): This has been directly copied from torch.utils.checkpoint.""" if deepspeed.checkpointing.is_configured(): return deepspeed.checkpointing.checkpoint(function, *args) - - return CheckpointFunction.apply(function, - distribute_saved_activations, *args) + + return CheckpointFunction.apply(function, distribute_saved_activations, *args) diff --git a/megatron/core/tensor_parallel/utils.py b/megatron/core/tensor_parallel/utils.py index ee1dadef7c0..bda4877361d 100644 --- a/megatron/core/tensor_parallel/utils.py +++ b/megatron/core/tensor_parallel/utils.py @@ -7,21 +7,22 @@ from megatron.core.utils import divide from megatron.core import parallel_state + def split_tensor_along_last_dim( tensor: torch.Tensor, num_partitions: int, contiguous_split_chunks: bool = False, ) -> List[torch.Tensor]: - """ Split a tensor along its last dimension. + """Split a tensor along its last dimension. - Arguments: - tensor: input tensor. - num_partitions: number of partitions to split the tensor - contiguous_split_chunks: If True, make each chunk contiguous - in memory. + Arguments: + tensor: input tensor. + num_partitions: number of partitions to split the tensor + contiguous_split_chunks: If True, make each chunk contiguous + in memory. - Returns: - A list of Tensors + Returns: + A list of Tensors """ # Get the size and dimension. last_dim = tensor.dim() - 1 @@ -34,28 +35,33 @@ def split_tensor_along_last_dim( return tensor_list + def split_tensor_into_1d_equal_chunks(tensor, new_buffer=False): - """ Break a tensor into equal 1D chunks across tensor parallel ranks. + """Break a tensor into equal 1D chunks across tensor parallel ranks. - Returns a Tensor or View with this rank's portion of the data. + Returns a Tensor or View with this rank's portion of the data. - Arguments: - tensor: The tensor to split + Arguments: + tensor: The tensor to split - Keyword Arguments: - new_buffer (bool): If True, returns a new Tensor. - If False, returns a view into the existing Tensor. - Default is False + Keyword Arguments: + new_buffer (bool): If True, returns a new Tensor. + If False, returns a view into the existing Tensor. + Default is False """ - partition_size = torch.numel(tensor) // \ - parallel_state.get_tensor_model_parallel_world_size() + partition_size = ( + torch.numel(tensor) // parallel_state.get_tensor_model_parallel_world_size() + ) start_index = partition_size * parallel_state.get_tensor_model_parallel_rank() end_index = start_index + partition_size if new_buffer: - data = torch.empty(partition_size, dtype=tensor.dtype, - device=torch.cuda.current_device(), - requires_grad=False) + data = torch.empty( + partition_size, + dtype=tensor.dtype, + device=torch.cuda.current_device(), + requires_grad=False, + ) data.copy_(tensor.view(-1)[start_index:end_index]) else: data = tensor.view(-1)[start_index:end_index] @@ -63,38 +69,44 @@ def split_tensor_into_1d_equal_chunks(tensor, new_buffer=False): def gather_split_1d_tensor(tensor): - """ Opposite of split_tensor_into_1d_equal_chunks. Gather values from tensor - model parallel ranks. + """Opposite of split_tensor_into_1d_equal_chunks. Gather values from tensor + model parallel ranks. - Returns a new Tensor with the gathered data. + Returns a new Tensor with the gathered data. - Arguments: - tensor: A Tensor or view of this rank's portion of the data. + Arguments: + tensor: A Tensor or view of this rank's portion of the data. """ - numel_gathered = torch.numel(tensor) * \ - parallel_state.get_tensor_model_parallel_world_size() - gathered = torch.empty(numel_gathered, dtype=tensor.dtype, - device=torch.cuda.current_device(), - requires_grad=False) + numel_gathered = ( + torch.numel(tensor) * parallel_state.get_tensor_model_parallel_world_size() + ) + gathered = torch.empty( + numel_gathered, + dtype=tensor.dtype, + device=torch.cuda.current_device(), + requires_grad=False, + ) # TODO: This API is experimental in pytorch (as of Feb 2022) and # this might break in future pytorch releases. We chose this API # as opposed to torch.distributed.all_gather for efficiency reasons. # This API calls directly NCCL all-gather versus the former does # internal copies and can potentially cause slow down. - if version.parse(torch.__version__) >= version.parse('1.13'): - torch.distributed.all_gather_into_tensor(gathered, tensor, - group=parallel_state.get_tensor_model_parallel_group()) + if version.parse(torch.__version__) >= version.parse("1.13"): + torch.distributed.all_gather_into_tensor( + gathered, tensor, group=parallel_state.get_tensor_model_parallel_group() + ) else: - torch.distributed._all_gather_base(gathered, tensor, - group=parallel_state.get_tensor_model_parallel_group()) + torch.distributed._all_gather_base( + gathered, tensor, group=parallel_state.get_tensor_model_parallel_group() + ) return gathered class VocabUtility: - """ Split the vocabulary into `world_size` chunks and return the first - and last index of the vocabulary belonging to the `rank` - partition: Note that indices in [fist, last) + """Split the vocabulary into `world_size` chunks and return the first + and last index of the vocabulary belonging to the `rank` + partition: Note that indices in [fist, last) """ @@ -107,7 +119,9 @@ def vocab_range_from_per_partition_vocab_size( return index_f, index_l @staticmethod - def vocab_range_from_global_vocab_size(global_vocab_size: int, rank: int, world_size: int) -> Sequence[int]: + def vocab_range_from_global_vocab_size( + global_vocab_size: int, rank: int, world_size: int + ) -> Sequence[int]: per_partition_vocab_size = divide(global_vocab_size, world_size) return VocabUtility.vocab_range_from_per_partition_vocab_size( per_partition_vocab_size, rank, world_size diff --git a/megatron/core/tensor_parallel/weight_grad_store.py b/megatron/core/tensor_parallel/weight_grad_store.py index bbd1aea533c..fc548374048 100644 --- a/megatron/core/tensor_parallel/weight_grad_store.py +++ b/megatron/core/tensor_parallel/weight_grad_store.py @@ -1,5 +1,6 @@ import queue + class WeightGradStore: cache = [] @@ -31,4 +32,4 @@ def pop(cls): assert cls.weight_grad_queue.qsize() > 0 stored_grads = cls.weight_grad_queue.get() for total_input, grad_output, weight, func in stored_grads: - func(total_input, grad_output, weight) \ No newline at end of file + func(total_input, grad_output, weight) diff --git a/megatron/core/transformer/attention.py b/megatron/core/transformer/attention.py index 15818bddf1f..2157433b086 100644 --- a/megatron/core/transformer/attention.py +++ b/megatron/core/transformer/attention.py @@ -12,8 +12,8 @@ from megatron.core.transformer.module import MegatronModule from megatron.core.transformer.enums import AttnType, AttnMaskType from megatron.core.transformer.transformer_config import TransformerConfig -from megatron.core.transformer.custom_layers.transformer_engine import \ - TECoreAttention, TEColumnParallelLinear, TERowParallelLinear +from megatron.core.transformer.custom_layers.transformer_engine import TECoreAttention, TEColumnParallelLinear, TERowParallelLinear + class Attention(MegatronModule, ABC): """Attention layer abstract class. @@ -38,16 +38,22 @@ def __init__( # Per attention head and per partition values. world_size = parallel_state.get_tensor_model_parallel_world_size() - self.hidden_size_per_attention_head = divide(self.projection_size, self.config.num_attention_heads) - self.num_attention_heads_per_partition = divide(self.config.num_attention_heads, world_size) + self.hidden_size_per_attention_head = divide( + self.projection_size, self.config.num_attention_heads + ) + self.num_attention_heads_per_partition = divide( + self.config.num_attention_heads, world_size + ) self.core_attention = TECoreAttention( config=self.config, layer_number=self.layer_number, - attn_mask_type=self.attn_mask_type + attn_mask_type=self.attn_mask_type, ) - self.checkpoint_core_attention = self.config.recompute_granularity == 'selective' + self.checkpoint_core_attention = ( + self.config.recompute_granularity == "selective" + ) # Output. self.linear_proj = TERowParallelLinear( @@ -93,7 +99,13 @@ def get_query_key_value_tensors(self, hidden_states, key_value_states): is "self-attn" or "cross-attn". """ - def forward(self, hidden_states, attention_mask, key_value_states=None, inference_params=None): + def forward( + self, + hidden_states, + attention_mask, + key_value_states=None, + inference_params=None, + ): # hidden_states: [sq, b, h] # ================================================= @@ -106,23 +118,29 @@ def forward(self, hidden_states, attention_mask, key_value_states=None, inferenc if self.layer_number not in inference_params.key_value_memory_dict: inf_max_seq_len = inference_params.max_sequence_len inf_max_batch_size = inference_params.max_batch_size - inference_key_memory = self._allocate_memory(inf_max_seq_len, inf_max_batch_size) - inference_value_memory = self._allocate_memory(inf_max_seq_len, inf_max_batch_size) + inference_key_memory = self._allocate_memory( + inf_max_seq_len, inf_max_batch_size + ) + inference_value_memory = self._allocate_memory( + inf_max_seq_len, inf_max_batch_size + ) inference_params.key_value_memory_dict[self.layer_number] = ( inference_key_memory, inference_value_memory, ) else: - inference_key_memory, inference_value_memory = inference_params.key_value_memory_dict[ - self.layer_number - ] + inference_key_memory, inference_value_memory = ( + inference_params.key_value_memory_dict[self.layer_number] + ) # ===================== # Query, Key, and Value # ===================== # Get the query, key and value tensors based on the type of attention - # self or cross attn. - query, key, value = self.get_query_key_value_tensors(hidden_states, key_value_states) + query, key, value = self.get_query_key_value_tensors( + hidden_states, key_value_states + ) # ================================== # Adjust key and value for inference @@ -136,8 +154,12 @@ def forward(self, hidden_states, attention_mask, key_value_states=None, inferenc sequence_end = sequence_start + key.size(0) assert sequence_end <= inference_key_memory.size(0) # Copy key and values. - inference_key_memory[sequence_start:sequence_end, batch_start:batch_end, ...] = key - inference_value_memory[sequence_start:sequence_end, batch_start:batch_end, ...] = value + inference_key_memory[ + sequence_start:sequence_end, batch_start:batch_end, ... + ] = key + inference_value_memory[ + sequence_start:sequence_end, batch_start:batch_end, ... + ] = value key = inference_key_memory[:sequence_end, batch_start:batch_end, ...] value = inference_value_memory[:sequence_end, batch_start:batch_end, ...] @@ -146,7 +168,9 @@ def forward(self, hidden_states, attention_mask, key_value_states=None, inferenc # ================================== if self.checkpoint_core_attention: - core_attn_out = self._checkpointed_attention_forward(query, key, value, attention_mask) + core_attn_out = self._checkpointed_attention_forward( + query, key, value, attention_mask + ) else: core_attn_out = self.core_attention(query, key, value, attention_mask) @@ -158,29 +182,31 @@ def forward(self, hidden_states, attention_mask, key_value_states=None, inferenc return output, bias + class SelfAttention(Attention): """Self-attention layer class Self-attention layer takes input with size [s, b, h] and returns output of the same size. """ - def __init__(self, - config: TransformerConfig, - layer_number: int = 1, - attn_mask_type=AttnMaskType.padding): + + def __init__( + self, + config: TransformerConfig, + layer_number: int = 1, + attn_mask_type=AttnMaskType.padding, + ): super().__init__( - config=config, - layer_number=layer_number, - attn_mask_type=attn_mask_type + config=config, layer_number=layer_number, attn_mask_type=attn_mask_type ) self.linear_qkv = TEColumnParallelLinear( - self.config.hidden_size, - 3 * self.projection_size, - config=self.config, - init_method=self.config.init_method, - bias=self.config.add_bias_linear, - skip_bias_add=False + self.config.hidden_size, + 3 * self.projection_size, + config=self.config, + init_method=self.config.init_method, + bias=self.config.add_bias_linear, + skip_bias_add=False, ) def get_query_key_value_tensors(self, hidden_states, key_value_states=None): @@ -202,20 +228,22 @@ def get_query_key_value_tensors(self, hidden_states, key_value_states=None): return query, key, value + class CrossAttention(Attention): """Cross-attention layer class Cross-attention layer takes input with size [s, b, h] and context with size [s, b, h] and returns output of the same size. """ - def __init__(self, - config: TransformerConfig, - layer_number: int = 1, - attn_mask_type=AttnMaskType.padding): + + def __init__( + self, + config: TransformerConfig, + layer_number: int = 1, + attn_mask_type=AttnMaskType.padding, + ): super().__init__( - config=config, - layer_number=layer_number, - attn_mask_type=attn_mask_type + config=config, layer_number=layer_number, attn_mask_type=attn_mask_type ) self.linear_q = TEColumnParallelLinear( @@ -224,7 +252,7 @@ def __init__(self, config=self.config, init_method=self.config.init_method, bias=self.config.add_bias_linear, - skip_bias_add=False + skip_bias_add=False, ) self.linear_kv = TEColumnParallelLinear( @@ -233,7 +261,7 @@ def __init__(self, config=self.config, init_method=self.config.init_method, bias=self.config.add_bias_linear, - skip_bias_add=False + skip_bias_add=False, ) def get_query_key_value_tensors(self, hidden_states, key_value_states): diff --git a/megatron/core/transformer/core_attention.py b/megatron/core/transformer/core_attention.py index aa5795a794c..9c888283d47 100644 --- a/megatron/core/transformer/core_attention.py +++ b/megatron/core/transformer/core_attention.py @@ -30,7 +30,12 @@ class CoreAttention(MegatronModule): s: sequence length """ - def __init__(self, config: TransformerConfig, layer_number: int = 1, attn_mask_type=AttnMaskType.padding): + def __init__( + self, + config: TransformerConfig, + layer_number: int = 1, + attn_mask_type=AttnMaskType.padding, + ): super().__init__(config=config) self.config: TransformerConfig = config @@ -43,8 +48,12 @@ def __init__(self, config: TransformerConfig, layer_number: int = 1, attn_mask_t # Per attention head and per partition values. world_size = parallel_state.get_tensor_model_parallel_world_size() self.hidden_size_per_partition = divide(projection_size, world_size) - self.hidden_size_per_attention_head = divide(projection_size, config.num_attention_heads) - self.num_attention_heads_per_partition = divide(config.num_attention_heads, world_size) + self.hidden_size_per_attention_head = divide( + projection_size, config.num_attention_heads + ) + self.num_attention_heads_per_partition = divide( + config.num_attention_heads, world_size + ) coeff = None self.norm_factor = math.sqrt(self.hidden_size_per_attention_head) @@ -67,23 +76,38 @@ def __init__(self, config: TransformerConfig, layer_number: int = 1, attn_mask_t # on average it should not be partition dependent. self.attention_dropout = torch.nn.Dropout(self.config.attention_dropout) - def forward(self, query_layer: Tensor, key_layer: Tensor, value_layer: Tensor, attention_mask: Tensor): + def forward( + self, + query_layer: Tensor, + key_layer: Tensor, + value_layer: Tensor, + attention_mask: Tensor, + ): # =================================== # Raw attention scores. [b, n/p, s, s] # =================================== # [b, np, sq, sk] - output_size = (query_layer.size(1), query_layer.size(2), query_layer.size(0), key_layer.size(0)) + output_size = ( + query_layer.size(1), + query_layer.size(2), + query_layer.size(0), + key_layer.size(0), + ) # [sq, b, np, hn] -> [sq, b * np, hn] - query_layer = query_layer.view(output_size[2], output_size[0] * output_size[1], -1) + query_layer = query_layer.view( + output_size[2], output_size[0] * output_size[1], -1 + ) # [sk, b, np, hn] -> [sk, b * np, hn] key_layer = key_layer.view(output_size[3], output_size[0] * output_size[1], -1) # preallocting input tensor: [b * np, sq, sk] matmul_input_buffer = parallel_state.get_global_memory_buffer().get_tensor( - (output_size[0] * output_size[1], output_size[2], output_size[3]), query_layer.dtype, "mpu" + (output_size[0] * output_size[1], output_size[2], output_size[3]), + query_layer.dtype, + "mpu", ) # Raw attention scores. [b * np, sq, sk] @@ -103,7 +127,9 @@ def forward(self, query_layer: Tensor, key_layer: Tensor, value_layer: Tensor, a # =========================== # attention scores and attention mask [b, np, sq, sk] - attention_probs: Tensor = self.scale_mask_softmax(attention_scores, attention_mask) + attention_probs: Tensor = self.scale_mask_softmax( + attention_scores, attention_mask + ) # This is actually dropping out entire tokens to attend to, which might # seem a bit unusual, but is taken from the original Transformer paper. @@ -122,13 +148,22 @@ def forward(self, query_layer: Tensor, key_layer: Tensor, value_layer: Tensor, a # [sk, b, np, hn] --> [b, np, sq, hn] # context layer shape: [b, np, sq, hn] - output_size = (value_layer.size(1), value_layer.size(2), query_layer.size(0), value_layer.size(3)) + output_size = ( + value_layer.size(1), + value_layer.size(2), + query_layer.size(0), + value_layer.size(3), + ) # change view [sk, b * np, hn] - value_layer = value_layer.view(value_layer.size(0), output_size[0] * output_size[1], -1) + value_layer = value_layer.view( + value_layer.size(0), output_size[0] * output_size[1], -1 + ) # change view [b * np, sq, sk] - attention_probs = attention_probs.view(output_size[0] * output_size[1], output_size[2], -1) + attention_probs = attention_probs.view( + output_size[0] * output_size[1], output_size[2], -1 + ) # matmul: [b * np, sq, hn] context_layer = torch.bmm(attention_probs, value_layer.transpose(0, 1)) @@ -140,7 +175,9 @@ def forward(self, query_layer: Tensor, key_layer: Tensor, value_layer: Tensor, a context_layer = context_layer.permute(2, 0, 1, 3).contiguous() # [sq, b, np, hn] --> [sq, b, hp] - new_context_layer_shape = context_layer.size()[:-2] + (self.hidden_size_per_partition,) + new_context_layer_shape = context_layer.size()[:-2] + ( + self.hidden_size_per_partition, + ) context_layer = context_layer.view(*new_context_layer_shape) return context_layer diff --git a/megatron/core/transformer/custom_layers/transformer_engine.py b/megatron/core/transformer/custom_layers/transformer_engine.py index 8d5c6aa15c0..88ff1ec961d 100644 --- a/megatron/core/transformer/custom_layers/transformer_engine.py +++ b/megatron/core/transformer/custom_layers/transformer_engine.py @@ -7,21 +7,24 @@ from megatron.core.parallel_state import get_tensor_model_parallel_group from megatron.core.tensor_parallel import get_cuda_rng_tracker + class TELayerNorm(te.pytorch.module.LayerNorm): """ Wrapper for the Transformer-Engine's `LayerNorm`. """ - def __init__(self, - hidden_size: int, - eps: float = 1e-5, - sequence_parallel: bool = False, - **kwargs): + + def __init__( + self, + hidden_size: int, + eps: float = 1e-5, + sequence_parallel: bool = False, + **kwargs + ): super().__init__( - hidden_size=hidden_size, - eps=eps, - sequence_parallel=sequence_parallel + hidden_size=hidden_size, eps=eps, sequence_parallel=sequence_parallel ) + class TELinear(te.pytorch.module.Linear): """ Wrapper for the Transformer-Engine's `Linear` layer. @@ -30,15 +33,19 @@ class TELinear(te.pytorch.module.Linear): yet, the tp_group passed to TE will be None and must be set later via set_tensor_parallel_group(). """ - def __init__(self, - input_size: int, - output_size: int, - config: TransformerConfig, - parallel_mode: str, - init_method: Callable, *, - bias: bool = True, - skip_bias_add: bool = False, - **kwargs): + + def __init__( + self, + input_size: int, + output_size: int, + config: TransformerConfig, + parallel_mode: str, + init_method: Callable, + *, + bias: bool = True, + skip_bias_add: bool = False, + **kwargs + ): self.config = config # TE returns a zero length Tensor when bias=False and @@ -74,16 +81,16 @@ def forward(self, x): return out return out, None + class TEColumnParallelLinear(TELinear): """ Wrapper for the Transformer-Engine's `Linear` layer but specialized similar to megatron's `ColumnParallelLinear` layer. """ - def __init__(self, - input_size: int, - output_size: int, - config: TransformerConfig, - **kwargs): + + def __init__( + self, input_size: int, output_size: int, config: TransformerConfig, **kwargs + ): self.config = config super().__init__( input_size=input_size, @@ -93,16 +100,16 @@ def __init__(self, **kwargs ) + class TERowParallelLinear(TELinear): """ Wrapper for the Transformer-Engine's `Linear` layer but specialized similar to megatron's `RowParallelLinear` layer. """ - def __init__(self, - input_size: int, - output_size: int, - config: TransformerConfig, - **kwargs): + + def __init__( + self, input_size: int, output_size: int, config: TransformerConfig, **kwargs + ): self.config = config super().__init__( input_size=input_size, @@ -112,6 +119,7 @@ def __init__(self, **kwargs ) + class TECoreAttention(te.pytorch.transformer.DotProductAttention): """ Wrapper for the Transformer-Engine's `DotProductAttention` layer that also @@ -121,11 +129,14 @@ class TECoreAttention(te.pytorch.transformer.DotProductAttention): yet, the tp_group passed to TE will be None and must be set later via set_tensor_parallel_group(). """ - def __init__(self, - config: TransformerConfig, - layer_number: int = 1, - attn_mask_type: AttnMaskType = AttnMaskType.padding, - **kwargs): + + def __init__( + self, + config: TransformerConfig, + layer_number: int = 1, + attn_mask_type: AttnMaskType = AttnMaskType.padding, + **kwargs + ): self.config = config super().__init__( num_attention_heads=self.config.num_attention_heads, diff --git a/megatron/core/transformer/mlp.py b/megatron/core/transformer/mlp.py index 69d5a01db38..4327ca65305 100644 --- a/megatron/core/transformer/mlp.py +++ b/megatron/core/transformer/mlp.py @@ -7,8 +7,8 @@ from megatron.core.fusions.fused_bias_gelu import bias_gelu_impl from megatron.core.transformer.module import MegatronModule from megatron.core.transformer.transformer_config import TransformerConfig -from megatron.core.transformer.custom_layers.transformer_engine import \ - TERowParallelLinear, TEColumnParallelLinear +from megatron.core.transformer.custom_layers.transformer_engine import TERowParallelLinear, TEColumnParallelLinear + class MLP(MegatronModule): """ @@ -47,9 +47,11 @@ def __init__(self, config: TransformerConfig): ) if self.config.gated_linear_unit: + def glu(x): x = torch.chunk(x, 2, dim=-1) return self.config.activation_func(x[0]) * x[1] + self.activation_func = glu else: self.activation_func = self.config.activation_func diff --git a/megatron/core/transformer/module.py b/megatron/core/transformer/module.py index 43d1bccb6fa..0ded86b0674 100644 --- a/megatron/core/transformer/module.py +++ b/megatron/core/transformer/module.py @@ -16,7 +16,7 @@ def param_is_not_shared(param): - return not hasattr(param, 'shared') or not param.shared + return not hasattr(param, "shared") or not param.shared class MegatronModule(torch.nn.Module): @@ -28,7 +28,7 @@ def __init__(self, config: TransformerConfig): super().__init__() self.config = config - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): """Use this function to override the state dict for saving checkpoints.""" return self.state_dict(prefix=prefix, keep_vars=keep_vars) @@ -74,6 +74,7 @@ def float_conversion(val): class Float16Module(MegatronModule): + def __init__(self, config: TransformerConfig, module: torch.nn.Module): super(Float16Module, self).__init__(config) self.config = config @@ -81,19 +82,19 @@ def __init__(self, config: TransformerConfig, module: torch.nn.Module): self.bf16 = config.bf16 if self.fp16: - self.add_module('module', module.half()) + self.add_module("module", module.half()) def float16_convertor(val): return val.half() elif self.bf16: - self.add_module('module', module.bfloat16()) + self.add_module("module", module.bfloat16()) def float16_convertor(val): return val.bfloat16() else: - raise Exception('Either config.fp16 or config.bf16 should be True.') + raise Exception("Either config.fp16 or config.bf16 should be True.") self.float16_convertor = float16_convertor @@ -108,11 +109,13 @@ def forward(self, *inputs, **kwargs): outputs = float16_to_fp32(outputs) return outputs - def state_dict(self, destination=None, prefix='', keep_vars=False): + def state_dict(self, destination=None, prefix="", keep_vars=False): return self.module.state_dict(prefix=prefix, keep_vars=keep_vars) - def state_dict_for_save_checkpoint(self, prefix='', keep_vars=False): - return self.module.state_dict_for_save_checkpoint(prefix=prefix, keep_vars=keep_vars) + def state_dict_for_save_checkpoint(self, prefix="", keep_vars=False): + return self.module.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) def load_state_dict(self, state_dict, strict=True): self.module.load_state_dict(state_dict, strict=strict) diff --git a/megatron/core/transformer/transformer_block.py b/megatron/core/transformer/transformer_block.py index 35bd7a6fc73..be191c7ae98 100644 --- a/megatron/core/transformer/transformer_block.py +++ b/megatron/core/transformer/transformer_block.py @@ -36,12 +36,15 @@ def __init__( # required for pipeline parallel schedules self.input_tensor = None - self.checkpoint_core_attention = self.config.recompute_granularity == 'selective' + self.checkpoint_core_attention = ( + self.config.recompute_granularity == "selective" + ) # TODO: Maybe we can create a build_transformer_block method here instead self.num_layers_per_pipeline_rank = ( - self.config.num_layers // parallel_state.get_pipeline_model_parallel_world_size() + self.config.num_layers + // parallel_state.get_pipeline_model_parallel_world_size() ) self._build_layers() @@ -55,7 +58,9 @@ def _build_layers(self): # self.norm_factor *= coeff def build_layer(layer_number): return TransformerLayer( - config=self.config, layer_number=layer_number, self_attn_mask_type=self.self_attn_mask_type, + config=self.config, + layer_number=layer_number, + self_attn_mask_type=self.self_attn_mask_type, ) pipeline_rank = parallel_state.get_pipeline_model_parallel_rank() @@ -78,10 +83,15 @@ def build_layer(layer_number): total_num_layers = self.config.num_layers num_layers_per_virtual_rank = self.num_layers_per_pipeline_rank // vp_size total_virtual_chunks = total_num_layers / vp_size - offset = vp_rank * total_virtual_chunks + (pipeline_rank * num_layers_per_virtual_rank) + offset = vp_rank * total_virtual_chunks + ( + pipeline_rank * num_layers_per_virtual_rank + ) self.layers = torch.nn.ModuleList( - [build_layer(i + 1 + offset) for i in range(num_layers_per_virtual_rank)] + [ + build_layer(i + 1 + offset) + for i in range(num_layers_per_virtual_rank) + ] ) else: # Each stage gets a contiguous set of layers. @@ -92,7 +102,10 @@ def build_layer(layer_number): # @jcasper why is layer_number using 1 index? self.layers = torch.nn.ModuleList( - [build_layer(i + 1 + offset) for i in range(self.num_layers_per_pipeline_rank)] + [ + build_layer(i + 1 + offset) + for i in range(self.num_layers_per_pipeline_rank) + ] ) # # TODO: add back standalone_embedding_stage @@ -136,7 +149,7 @@ def custom_forward(*args, **kwargs): return custom_forward - if self.config.recompute_method == 'uniform': + if self.config.recompute_method == "uniform": # Uniformly divide the total number of Transformer layers and checkpoint # the input activation of each divided chunk. # A method to further reduce memory usage reducing checkpoints. @@ -151,14 +164,17 @@ def custom_forward(*args, **kwargs): l += self.recompute_num_layers - elif self.config.recompute_method == 'block': + elif self.config.recompute_method == "block": # Checkpoint the input activation of only a set number of individual # Transformer layers and skip the rest. # A method fully use the device memory removing redundant re-computation. for l in range(self.num_layers_per_pipeline_rank): if l < self.config.recompute_num_layers: hidden_states = tensor_parallel.checkpoint( - custom(l, l + 1), self.config.distribute_saved_activations, hidden_states, attention_mask, + custom(l, l + 1), + self.config.distribute_saved_activations, + hidden_states, + attention_mask, ) else: hidden_states = custom(l, l + 1)(hidden_states, attention_mask) @@ -200,7 +216,11 @@ def forward(self, hidden_states, attention_mask, inference_params=None): # likely redundant, since p2p_communication.py (likely originator) # already creates viewless tensors. That said, make_viewless_tensor() # is called here to be future-proof and corner-case-proof. - hidden_states = make_viewless_tensor(inp=hidden_states, requires_grad=True, keep_graph=True,) + hidden_states = make_viewless_tensor( + inp=hidden_states, + requires_grad=True, + keep_graph=True, + ) if self.config.sequence_parallel: rng_context = tensor_parallel.get_cuda_rng_tracker().fork() @@ -209,11 +229,15 @@ def forward(self, hidden_states, attention_mask, inference_params=None): with rng_context: # Forward pass. - if self.config.recompute_granularity == 'full': - hidden_states = self._checkpointed_forward(hidden_states=hidden_states, attention_mask=attention_mask) + if self.config.recompute_granularity == "full": + hidden_states = self._checkpointed_forward( + hidden_states=hidden_states, attention_mask=attention_mask + ) else: for layer in self.layers: - hidden_states = layer(hidden_states=hidden_states, attention_mask=attention_mask) + hidden_states = layer( + hidden_states=hidden_states, attention_mask=attention_mask + ) # Final layer norm. if self.post_process and self.post_layer_norm: diff --git a/megatron/core/transformer/transformer_config.py b/megatron/core/transformer/transformer_config.py index de21046f026..b352b34fe0d 100644 --- a/megatron/core/transformer/transformer_config.py +++ b/megatron/core/transformer/transformer_config.py @@ -9,96 +9,97 @@ from megatron.core import ModelParallelConfig from megatron.core.utils import init_method_normal, scaled_init_method_normal + @dataclass class TransformerConfig(ModelParallelConfig): """Configuration object for megatron-core transformers. - Attributes: - - # model architecture - num_layers (int): Number of transformer layers in a transformer block. - hidden_size (int): Transformer hidden size. - ffn_hidden_size (int): Transformer Feed-Forward Network hidden size. - This is set to 4*hidden_size if not provided. Defaults to None.') - num_attention_heads (int): Number of transformer attention heads. - num_key_value_heads (int): This is the number of key_value heads that should be used to implement Grouped Query Attention. If - `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if - `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. - For more details checkout [this paper](https://arxiv.org/pdf/2305.13245.pdf). - If it is not specified, will default to `num_attention_heads`. - kv_channels (int): Projection weights dimension in multi-head attention. - This is set to hidden_size // num_attention_heads if not provided. - Defaults to None. - hidden_dropout (float): Dropout probability for transformer hidden state. Defaults to 0.1. - attention_dropout (float): Post attention dropout probability. Defaults to 0.1. - fp32_residual_connection (bool): If true, move residual connections to fp32. - apply_residual_connection_post_layernorm (bool): If true, uses the original BERT residule connection ordering. - Defaults to False. - layernorm_epsilon (float): Layernorm epsilon. Defaults to 1e-5. - - layernorm_zero_centered_gamma (bool): if set to 'True', the LayerNorm is adjusted to center the gamma values - around 0. This improves numerical stability. Defaults to False. - - add_bias_linear (bool): Include a bias term in all linear layers (QKV projections, after core attention, and two - in MLP layer). Default is True. - - gated_linear_unit (bool): Use a gated linear unit for the first linear layer in the MLP. Defaults to False. - - activation_func (Callable): Activation function to use for the non-linearity in the MLP. Defaults to F.gelu. - - # initialization - init_method (Callable): Method to initialize weights. Note that bias is always set to - zero. Should be a function that takes a single Tensor and - initializes it. Defaults to - megatron.core.utils.init_method_normal(init_method_std) which is - torch.nn.init.normal_ with mean=0.0 and std=init_method_Std. - - output_layer_init_method (Callable): Method to initialize weights of the output layer of - both attention and MLP blocks. Defaults to - megatron.core.utils.scaled_init_method_normal(init_method_std) - which is torch.nn.init.normal_ with mean=0.0 and - std=init_method_std / math.sqrt(2.0 * num_layers). - - init_method_std (float): Standard deviation of the zero mean normal for the default - initialization method, not used if init_method and - output_layer_init_method are provided. Defaults to 0.02. - - # mixed-precision - apply_query_key_layer_scaling (bool): If true, scale Q * K^T by 1 / layer-number. Defaults to True. - attention_softmax_in_fp32 (bool): If true, run attention masking and softmax in fp32. - This should be true if apply_query_key_layer_scaling is true. - - # fusion - bias_gelu_fustion (bool): If true, fuses bias and gelu. Defaults to False. - masked_softmax_fusion (bool): If true, uses softmax fusion. - persist_layer_norm (bool): If true, uses the persistent fused layer norm kernel. - This kernel only supports a fixed set of hidden sizes. - Defaults to False. - bias_dropout_fusion (bool): If true, uses bias dropout fusion. - - # activation recomputation - - recompute_granularity (str): megatron-core supports 'selective' activation checkpointing where only the memory - intensive part of attention is checkpointed. These memory intensive activations - are also less compute intensive which makes activation checkpointing more efficient - for LLMs (20B+). See Reducing Activation Recomputation in Large Transformer - Models: https://arxiv.org/abs/2205.05198 for more details. 'full' will checkpoint - the entire transformer layer. Must be 'selective' or 'full'. Defaults to None. - - recompute_method (str): uniform will uniformly divide the total number of transformer layers in a transformer - block and recompute the input activation of each divided chunk at the specified - granularity. block will recompute the input activations for only a set number of - transformer layers per pipeline stage. The rest of the layers in the pipeline stage - will not have any activations recomputed. Must be 'uniform' or 'block'. Defaults to - None. - - recompute_num_layers (int): When recompute_method is uniform, recompute_num_layers is the number of transformer - layers in each uniformly divided recompute unit. When recompute_method is block, - recompute_num_layers is the number of transformer layers to recompute within each - pipeline stage. Defaults to None. - - distribute_saved_activations (bool): If true, distribute recomputed activations across the model parallel - group. Defaults to None. + Attributes: + + # model architecture + num_layers (int): Number of transformer layers in a transformer block. + hidden_size (int): Transformer hidden size. + ffn_hidden_size (int): Transformer Feed-Forward Network hidden size. + This is set to 4*hidden_size if not provided. Defaults to None.') + num_attention_heads (int): Number of transformer attention heads. + num_key_value_heads (int): This is the number of key_value heads that should be used to implement Grouped Query Attention. If + `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if + `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. + For more details checkout [this paper](https://arxiv.org/pdf/2305.13245.pdf). + If it is not specified, will default to `num_attention_heads`. + kv_channels (int): Projection weights dimension in multi-head attention. + This is set to hidden_size // num_attention_heads if not provided. + Defaults to None. + hidden_dropout (float): Dropout probability for transformer hidden state. Defaults to 0.1. + attention_dropout (float): Post attention dropout probability. Defaults to 0.1. + fp32_residual_connection (bool): If true, move residual connections to fp32. + apply_residual_connection_post_layernorm (bool): If true, uses the original BERT residule connection ordering. + Defaults to False. + layernorm_epsilon (float): Layernorm epsilon. Defaults to 1e-5. + + layernorm_zero_centered_gamma (bool): if set to 'True', the LayerNorm is adjusted to center the gamma values + around 0. This improves numerical stability. Defaults to False. + + add_bias_linear (bool): Include a bias term in all linear layers (QKV projections, after core attention, and two + in MLP layer). Default is True. + + gated_linear_unit (bool): Use a gated linear unit for the first linear layer in the MLP. Defaults to False. + + activation_func (Callable): Activation function to use for the non-linearity in the MLP. Defaults to F.gelu. + + # initialization + init_method (Callable): Method to initialize weights. Note that bias is always set to + zero. Should be a function that takes a single Tensor and + initializes it. Defaults to + megatron.core.utils.init_method_normal(init_method_std) which is + torch.nn.init.normal_ with mean=0.0 and std=init_method_Std. + + output_layer_init_method (Callable): Method to initialize weights of the output layer of + both attention and MLP blocks. Defaults to + megatron.core.utils.scaled_init_method_normal(init_method_std) + which is torch.nn.init.normal_ with mean=0.0 and + std=init_method_std / math.sqrt(2.0 * num_layers). + + init_method_std (float): Standard deviation of the zero mean normal for the default + initialization method, not used if init_method and + output_layer_init_method are provided. Defaults to 0.02. + + # mixed-precision + apply_query_key_layer_scaling (bool): If true, scale Q * K^T by 1 / layer-number. Defaults to True. + attention_softmax_in_fp32 (bool): If true, run attention masking and softmax in fp32. + This should be true if apply_query_key_layer_scaling is true. + + # fusion + bias_gelu_fustion (bool): If true, fuses bias and gelu. Defaults to False. + masked_softmax_fusion (bool): If true, uses softmax fusion. + persist_layer_norm (bool): If true, uses the persistent fused layer norm kernel. + This kernel only supports a fixed set of hidden sizes. + Defaults to False. + bias_dropout_fusion (bool): If true, uses bias dropout fusion. + + # activation recomputation + + recompute_granularity (str): megatron-core supports 'selective' activation checkpointing where only the memory + intensive part of attention is checkpointed. These memory intensive activations + are also less compute intensive which makes activation checkpointing more efficient + for LLMs (20B+). See Reducing Activation Recomputation in Large Transformer + Models: https://arxiv.org/abs/2205.05198 for more details. 'full' will checkpoint + the entire transformer layer. Must be 'selective' or 'full'. Defaults to None. + + recompute_method (str): uniform will uniformly divide the total number of transformer layers in a transformer + block and recompute the input activation of each divided chunk at the specified + granularity. block will recompute the input activations for only a set number of + transformer layers per pipeline stage. The rest of the layers in the pipeline stage + will not have any activations recomputed. Must be 'uniform' or 'block'. Defaults to + None. + + recompute_num_layers (int): When recompute_method is uniform, recompute_num_layers is the number of transformer + layers in each uniformly divided recompute unit. When recompute_method is block, + recompute_num_layers is the number of transformer layers to recompute within each + pipeline stage. Defaults to None. + + distribute_saved_activations (bool): If true, distribute recomputed activations across the model parallel + group. Defaults to None. """ @@ -145,12 +146,14 @@ class TransformerConfig(ModelParallelConfig): distribute_saved_activations: bool = None def __post_init__(self): - """ Python dataclass method that is used to modify attributes after initialization. - See https://docs.python.org/3/library/dataclasses.html#post-init-processing for more details. + """Python dataclass method that is used to modify attributes after initialization. + See https://docs.python.org/3/library/dataclasses.html#post-init-processing for more details. """ super().__post_init__() if self.fp16 and self.bf16: - raise ValueError(f'Only one of self.fp16: {self.fp16} and self.bf16 {self.bf16} should be True.') + raise ValueError( + f"Only one of self.fp16: {self.fp16} and self.bf16 {self.bf16} should be True." + ) if self.ffn_hidden_size is None: self.ffn_hidden_size = 4 * self.hidden_size @@ -166,34 +169,36 @@ def __post_init__(self): self.attention_softmax_in_fp32 = True if self.recompute_granularity is not None: - if not self.recompute_granularity in ['full', 'selective']: + if not self.recompute_granularity in ["full", "selective"]: raise ValueError( f'When using recompute_granuarlity: {self.recompute_granularity} must be "full" or "selective".' ) if self.recompute_method is not None: - if not self.recompute_method in ['block', 'uniform']: - raise ValueError(f'recompute_method: {self.recompute_method} must be "block" or "uniform".') - elif self.recompute_granularity != 'selective': + if not self.recompute_method in ["block", "uniform"]: + raise ValueError( + f'recompute_method: {self.recompute_method} must be "block" or "uniform".' + ) + elif self.recompute_granularity != "selective": raise ValueError( f'Using recompute_granularity: {self.recompute_granularity} so recompute_method must be "block" or "uniform"' ) if self.recompute_num_layers is None: raise ValueError( - f'When using recompute_granularity: {self.recompute_granularity} so recompute_num_layers must be between ' - f'1 and num_layers_per_pipeline_rank: {self.num_layers // self.pipeline_model_parallel_size}' + f"When using recompute_granularity: {self.recompute_granularity} so recompute_num_layers must be between " + f"1 and num_layers_per_pipeline_rank: {self.num_layers // self.pipeline_model_parallel_size}" ) if self.distribute_saved_activations and self.sequence_parallel_enabled: raise ValueError( - f'distribute_saved_activations: {self.distribute_saved_activations} must be false when sequence parallel is enabled: {self.sequence_parallel_enabled}' + f"distribute_saved_activations: {self.distribute_saved_activations} must be false when sequence parallel is enabled: {self.sequence_parallel_enabled}" ) if self.virtual_pipeline_model_parallel_size is not None: if not self.num_layers % self.virtual_pipeline_model_parallel_size == 0: raise ValueError( - f'num_layers: {self.num_layers} must be divisible by virtual_model_parallel_size {self.virtual_pipeline_model_parallel_size}' + f"num_layers: {self.num_layers} must be divisible by virtual_model_parallel_size {self.virtual_pipeline_model_parallel_size}" ) if self.apply_query_key_layer_scaling: @@ -201,14 +206,19 @@ def __post_init__(self): if self.bias_gelu_fusion: if not self.add_bias_linear: - raise ValueError("When bias_gelu_fusion is True, add_bias_linear must also be True.") + raise ValueError( + "When bias_gelu_fusion is True, add_bias_linear must also be True." + ) if self.activation_func != F.gelu: - raise ValueError(f'When bias_gelu_fusion is True, activation_func must be F.gelu.') + raise ValueError( + f"When bias_gelu_fusion is True, activation_func must be F.gelu." + ) if self.init_method is None: self.init_method = init_method_normal(self.init_method_std) if self.output_layer_init_method is None: - self.output_layer_init_method = scaled_init_method_normal(self.init_method_std, self.num_layers) - + self.output_layer_init_method = scaled_init_method_normal( + self.init_method_std, self.num_layers + ) diff --git a/megatron/core/transformer/transformer_layer.py b/megatron/core/transformer/transformer_layer.py index 19804e4c603..1210bf2bedc 100644 --- a/megatron/core/transformer/transformer_layer.py +++ b/megatron/core/transformer/transformer_layer.py @@ -11,6 +11,7 @@ from megatron.core.utils import make_viewless_tensor from megatron.core.transformer.custom_layers.transformer_engine import TELayerNorm + class TransformerLayer(MegatronModule): """A single transformer layer. @@ -19,7 +20,10 @@ class TransformerLayer(MegatronModule): """ def __init__( - self, config: TransformerConfig, layer_number: int = 1, self_attn_mask_type=AttnMaskType.padding, + self, + config: TransformerConfig, + layer_number: int = 1, + self_attn_mask_type=AttnMaskType.padding, ): super().__init__(config=config) self.config: TransformerConfig = config @@ -65,13 +69,17 @@ def __init__( self.bias_dropout_add_exec_handler = torch.enable_grad self.bias_dropout_add_func = get_bias_dropout_add( - self.training, - self.config.bias_dropout_fusion + self.training, self.config.bias_dropout_fusion ) # TODO: decide how to do inference_params def forward( - self, hidden_states, attention_mask, encoder_output=None, enc_dec_attn_mask=None, inference_params=None + self, + hidden_states, + attention_mask, + encoder_output=None, + enc_dec_attn_mask=None, + inference_params=None, ): # hidden_states: [s, b, h] @@ -117,6 +125,8 @@ def forward( # won't result in memory savings (like the data loader, or # p2p_communication), it serves to document the origin of this # 'view' tensor. - output = make_viewless_tensor(inp=output, requires_grad=output.requires_grad, keep_graph=True) + output = make_viewless_tensor( + inp=output, requires_grad=output.requires_grad, keep_graph=True + ) return output diff --git a/megatron/core/transformer/utils.py b/megatron/core/transformer/utils.py index 481c63331f3..7e45882893e 100644 --- a/megatron/core/transformer/utils.py +++ b/megatron/core/transformer/utils.py @@ -8,6 +8,7 @@ from deepspeed.runtime.zero import GatheredParameters + def attention_mask_func(attention_scores, attention_mask): attention_scores.masked_fill_(attention_mask, -10000.0) return attention_scores @@ -17,10 +18,14 @@ def get_linear_layer(rows, columns, init_method, gather_params_on_init=False): """Simple linear layer with weight initialization.""" layer = torch.nn.Linear(rows, columns) if get_args().perform_initialization: - with GatheredParameters(layer.weight, modifier_rank=0, enable=gather_params_on_init): + with GatheredParameters( + layer.weight, modifier_rank=0, enable=gather_params_on_init + ): init_method(layer.weight) with torch.no_grad(): - with GatheredParameters(layer.weight, modifier_rank=0, enable=gather_params_on_init): + with GatheredParameters( + layer.weight, modifier_rank=0, enable=gather_params_on_init + ): layer.bias.zero_() return layer @@ -28,7 +33,9 @@ def get_linear_layer(rows, columns, init_method, gather_params_on_init=False): @torch.jit.script def gelu_impl(x): """OpenAI's gelu implementation.""" - return 0.5 * x * (1.0 + torch.tanh(0.7978845608028654 * x * (1.0 + 0.044715 * x * x))) + return ( + 0.5 * x * (1.0 + torch.tanh(0.7978845608028654 * x * (1.0 + 0.044715 * x * x))) + ) def openai_gelu(x): @@ -38,4 +45,11 @@ def openai_gelu(x): # This is actually Python equivalent of torch.nn.functional.gelu(), also with type hints for ONNX exporter @torch.jit.script def erf_gelu(x): - return x * 0.5 * (torch.erf(x / 1.41421).to(dtype=x.dtype) + torch.ones_like(x).to(dtype=x.dtype)) + return ( + x + * 0.5 + * ( + torch.erf(x / 1.41421).to(dtype=x.dtype) + + torch.ones_like(x).to(dtype=x.dtype) + ) + ) diff --git a/megatron/core/utils.py b/megatron/core/utils.py index e672c376de2..ee20db4ec9c 100644 --- a/megatron/core/utils.py +++ b/megatron/core/utils.py @@ -26,33 +26,42 @@ def divide(numerator, denominator): ensure_divisibility(numerator, denominator) return numerator // denominator + def get_attr_wrapped_model(model, attr, allow_none=True): """Get an attribute from a wrapped model""" if isinstance(model, list): raise RuntimeError("_get_attr_wrapped_model given a list of models") if allow_none: + def condition(model, attr): return not hasattr(model, attr) + else: + def condition(model, attr): return getattr(model, attr, None) is None while condition(model, attr): if not hasattr(model, "module"): - raise RuntimeError(f"_get_attr_wrapped_model couldn't find attribute {attr}") + raise RuntimeError( + f"_get_attr_wrapped_model couldn't find attribute {attr}" + ) model = model.module return getattr(model, attr) + def get_model_type(model): - return get_attr_wrapped_model(model, 'model_type') + return get_attr_wrapped_model(model, "model_type") + def get_model_config(model): args = get_args() if args.deepspeed: - return get_attr_wrapped_model(model.module, 'config', allow_none=False) - return get_attr_wrapped_model(model, 'config', allow_none=False) + return get_attr_wrapped_model(model.module, "config", allow_none=False) + return get_attr_wrapped_model(model, "config", allow_none=False) + class GlobalMemoryBuffer: """Global buffer to avoid dynamic memory allocations. @@ -64,59 +73,67 @@ def __init__(self): def get_tensor(self, tensor_shape, dtype, name): required_len = reduce(operator.mul, tensor_shape, 1) - if self.buffer.get((name, dtype), None) is None or \ - self.buffer[(name, dtype)].numel() < required_len: - self.buffer[(name, dtype)] = \ - torch.empty(required_len, - dtype=dtype, - device=get_accelerator().current_device_name(), - requires_grad=False) + if ( + self.buffer.get((name, dtype), None) is None + or self.buffer[(name, dtype)].numel() < required_len + ): + self.buffer[(name, dtype)] = torch.empty( + required_len, + dtype=dtype, + device=get_accelerator().current_device_name(), + requires_grad=False, + ) return self.buffer[(name, dtype)][0:required_len].view(*tensor_shape) + def _kernel_make_viewless_tensor(inp, requires_grad): - '''Make a viewless tensor. + """Make a viewless tensor. View tensors have the undesirable side-affect of retaining a reference to the originally-viewed tensor, even after manually setting the '.data' field. This method creates a new tensor that links to the old tensor's data, without linking the viewed tensor, referenced via the '._base' field. - ''' + """ out = torch.empty( (1,), - dtype = inp.dtype, - device = inp.device, - requires_grad = requires_grad, + dtype=inp.dtype, + device=inp.device, + requires_grad=requires_grad, ) out.data = inp.data return out + class MakeViewlessTensor(torch.autograd.Function): - ''' + """ Autograd function to make a viewless tensor. This function should be used in cases where the computation graph needs to be propagated, but we only want a viewless tensor (e.g., ParallelTransformer's hidden_states). Call this function by passing 'keep_graph = True' to 'make_viewless_tensor()'. - ''' + """ + @staticmethod def forward(ctx, inp, requires_grad): return _kernel_make_viewless_tensor(inp, requires_grad) + @staticmethod def backward(ctx, grad_output): return grad_output, None + def make_viewless_tensor(inp, requires_grad, keep_graph): - ''' + """ Entry-point for creating viewless tensors. This method should be used, rather than calling 'MakeViewlessTensor' or '_kernel_make_viewless_tensor' directly. This method acts as a switch for determining if an autograd function or a regular method should be used to create the tensor. - ''' + """ # return tensor as-is, if not a 'view' if inp._base is None: @@ -128,11 +145,12 @@ def make_viewless_tensor(inp, requires_grad, keep_graph): else: return _kernel_make_viewless_tensor(inp, requires_grad) -def assert_viewless_tensor(tensor, extra_msg = None): - '''Assert that a tensor is not a view (i.e., its '._base' field is - not set).''' + +def assert_viewless_tensor(tensor, extra_msg=None): + """Assert that a tensor is not a view (i.e., its '._base' field is + not set).""" if isinstance(tensor, list): - [ assert_viewless_tensor(t) for t in tensor ] + [assert_viewless_tensor(t) for t in tensor] return tensor if not isinstance(tensor, torch.Tensor): return tensor @@ -143,15 +161,21 @@ def assert_viewless_tensor(tensor, extra_msg = None): ) % extra_msg return tensor + def safely_set_viewless_tensor_data(tensor, new_data_tensor): - '''Safely set tensor's '.data' field. + """Safely set tensor's '.data' field. Check first that the tensor is viewless (i.e., '._base' not set). If not, raise an exception. - ''' - assert_viewless_tensor(tensor, extra_msg = "FYI, tensor._base has shape %s, and new_data_tensor has shape %s." % ("--" if tensor._base is None else tensor._base.shape, new_data_tensor.shape)) + """ + assert_viewless_tensor( + tensor, + extra_msg="FYI, tensor._base has shape %s, and new_data_tensor has shape %s." + % ("--" if tensor._base is None else tensor._base.shape, new_data_tensor.shape), + ) tensor.data = new_data_tensor + def init_method_normal(sigma): """Init method based on N(0, sigma).""" From f4cd2e1edc99446782c2b6055b476239e5437396 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:16:06 -0500 Subject: [PATCH 439/590] chore: Format `megatron/*.py` --- megatron/__init__.py | 8 +- megatron/arguments.py | 3469 +++++++++++++++-------- megatron/checkpointing.py | 618 ++-- megatron/dist_signal_handler.py | 18 +- megatron/enums.py | 6 +- megatron/fp16_deprecated/loss_scaler.py | 55 +- megatron/global_vars.py | 126 +- megatron/indexer.py | 54 +- megatron/initialize.py | 217 +- megatron/memory.py | 71 +- megatron/microbatches.py | 126 +- megatron/optimizer_param_scheduler.py | 260 +- megatron/p2p_communication.py | 148 +- megatron/profiler.py | 29 +- megatron/text_generation_server.py | 166 +- megatron/text_generation_utils.py | 340 ++- megatron/training.py | 30 +- megatron/utils.py | 4 +- 18 files changed, 3613 insertions(+), 2132 deletions(-) diff --git a/megatron/__init__.py b/megatron/__init__.py index 4a7fe2cd828..8dd5f2d8b14 100644 --- a/megatron/__init__.py +++ b/megatron/__init__.py @@ -12,10 +12,6 @@ from .global_vars import get_wandb_writer from .global_vars import get_adlr_autoresume from .global_vars import get_timers -from .initialize import initialize_megatron +from .initialize import initialize_megatron -from .utils import (print_rank_0, - is_last_rank, - print_rank_last, - is_rank_0, - is_aml) +from .utils import (print_rank_0, is_last_rank, print_rank_last, is_rank_0, is_aml) diff --git a/megatron/arguments.py b/megatron/arguments.py index cef72acf51f..c29555b45d1 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -18,10 +18,12 @@ from megatron.core.transformer import TransformerConfig + def parse_args(extra_args_provider=None, ignore_unknown_args=False): """Parse all arguments.""" - parser = argparse.ArgumentParser(description='Megatron-LM Arguments', - allow_abbrev=False) + parser = argparse.ArgumentParser( + description="Megatron-LM Arguments", allow_abbrev=False + ) # Standard arguments. parser = _add_network_size_args(parser) @@ -63,74 +65,108 @@ def parse_args(extra_args_provider=None, ignore_unknown_args=False): args.ds_pipeline_enabled = not args.no_pipeline_parallel # Args from environment - args.rank = int(os.getenv('RANK', '0')) - args.world_size = int(os.getenv("WORLD_SIZE", '1')) + args.rank = int(os.getenv("RANK", "0")) + args.world_size = int(os.getenv("WORLD_SIZE", "1")) return args + def validate_args(args, defaults={}): # Tensor model parallel size. args.tensor_model_parallel_size = min( - args.tensor_model_parallel_size, args.world_size) - assert args.world_size % args.tensor_model_parallel_size == 0, 'world size'\ - ' ({}) is not divisible by tensor model parallel size ({})'.format( - args.world_size, args.tensor_model_parallel_size) + args.tensor_model_parallel_size, args.world_size + ) + assert ( + args.world_size % args.tensor_model_parallel_size == 0 + ), "world size" " ({}) is not divisible by tensor model parallel size ({})".format( + args.world_size, args.tensor_model_parallel_size + ) # Zero bubble pipeline is defined on deepspeed's scheduler if args.enable_zbh1_pipeline: - assert args.deepspeed, 'Use DeepSpeed to use zero-bubble H1 pipeline' - assert args.sequence_parallel == False, "Sequence Parallel not tested, proceed at own will by removing this line" + assert args.deepspeed, "Use DeepSpeed to use zero-bubble H1 pipeline" + assert ( + args.sequence_parallel == False + ), "Sequence Parallel not tested, proceed at own will by removing this line" if args.enable_zbh1_exact_semantics: - assert args.enable_zbh1_pipeline, 'Exact semantics require ZBH1 pipeline enabled' + assert ( + args.enable_zbh1_pipeline + ), "Exact semantics require ZBH1 pipeline enabled" # Pipeline model parallel size. args.pipeline_model_parallel_size = min( args.pipeline_model_parallel_size, - (args.world_size // args.tensor_model_parallel_size)) + (args.world_size // args.tensor_model_parallel_size), + ) args.transformer_pipeline_model_parallel_size = ( args.pipeline_model_parallel_size - 1 - if args.standalone_embedding_stage else - args.pipeline_model_parallel_size + if args.standalone_embedding_stage + else args.pipeline_model_parallel_size ) # Checks. if args.no_pipeline_parallel: - assert args.pipeline_model_parallel_size == 1, \ - "pipeline_model_parallel_size must be 1 if pipeline parallel is disabled" - + assert ( + args.pipeline_model_parallel_size == 1 + ), "pipeline_model_parallel_size must be 1 if pipeline parallel is disabled" + if args.ds_sequence_parallel_size > 1: - assert version.parse(deepspeed.__version__) >= version.parse("0.10.2"), "sequence parallelism requires DeepSpeed version 0.10.2+" - - model_parallel_size = args.pipeline_model_parallel_size * \ - args.tensor_model_parallel_size * \ - args.ds_sequence_parallel_size - assert args.world_size % model_parallel_size == 0, 'world size ({}) is not'\ - ' divisible by tensor parallel size ({}) times pipeline parallel ' \ - 'size ({}) times seqence parallel size ({})'.format(args.world_size, args.tensor_model_parallel_size, - args.pipeline_model_parallel_size, args.ds_sequence_parallel_size) + assert version.parse(deepspeed.__version__) >= version.parse( + "0.10.2" + ), "sequence parallelism requires DeepSpeed version 0.10.2+" + + model_parallel_size = ( + args.pipeline_model_parallel_size + * args.tensor_model_parallel_size + * args.ds_sequence_parallel_size + ) + assert args.world_size % model_parallel_size == 0, ( + "world size ({}) is not" + " divisible by tensor parallel size ({}) times pipeline parallel " + "size ({}) times seqence parallel size ({})".format( + args.world_size, + args.tensor_model_parallel_size, + args.pipeline_model_parallel_size, + args.ds_sequence_parallel_size, + ) + ) args.data_parallel_size = args.world_size // model_parallel_size if args.rank == 0: - print('using world size: {}, data-parallel-size: {}, ' - 'sequence-parallel size: {}, ' - 'tensor-model-parallel size: {}, ' - 'pipeline-model-parallel size: {} '.format( - args.world_size, args.data_parallel_size, - args.ds_sequence_parallel_size, - args.tensor_model_parallel_size, - args.pipeline_model_parallel_size), flush=True) + print( + "using world size: {}, data-parallel-size: {}, " + "sequence-parallel size: {}, " + "tensor-model-parallel size: {}, " + "pipeline-model-parallel size: {} ".format( + args.world_size, + args.data_parallel_size, + args.ds_sequence_parallel_size, + args.tensor_model_parallel_size, + args.pipeline_model_parallel_size, + ), + flush=True, + ) if args.pipeline_model_parallel_size > 1: if args.pipeline_model_parallel_split_rank is not None: - assert args.pipeline_model_parallel_split_rank < \ - args.pipeline_model_parallel_size, 'split rank needs'\ - ' to be less than pipeline model parallel size ({})'.format( - args.pipeline_model_parallel_size) + assert ( + args.pipeline_model_parallel_split_rank + < args.pipeline_model_parallel_size + ), ( + "split rank needs" + " to be less than pipeline model parallel size ({})".format( + args.pipeline_model_parallel_size + ) + ) # Deprecated arguments - assert args.batch_size is None, '--batch-size argument is no longer ' \ - 'valid, use --micro-batch-size instead' + assert args.batch_size is None, ( + "--batch-size argument is no longer " "valid, use --micro-batch-size instead" + ) del args.batch_size - assert args.warmup is None, '--warmup argument is no longer valid, use ' \ - '--lr-warmup-fraction instead' + assert args.warmup is None, ( + "--warmup argument is no longer valid, use " "--lr-warmup-fraction instead" + ) del args.warmup - assert args.model_parallel_size is None, '--model-parallel-size is no ' \ - 'longer valid, use --tensor-model-parallel-size instead' + assert args.model_parallel_size is None, ( + "--model-parallel-size is no " + "longer valid, use --tensor-model-parallel-size instead" + ) del args.model_parallel_size # HACK: below is commented because DeepSpeed still relies on the old @@ -143,7 +179,7 @@ def validate_args(args, defaults={}): # del args.checkpoint_activations if args.recompute_activations: - args.recompute_granularity = 'selective' + args.recompute_granularity = "selective" del args.recompute_activations # Set input defaults. @@ -153,10 +189,13 @@ def validate_args(args, defaults={}): # ensuring the arg is set to None. if getattr(args, key) is not None: if args.rank == 0: - print('WARNING: overriding default arguments for {key}:{v} \ - with {key}:{v2}'.format(key=key, v=defaults[key], - v2=getattr(args, key)), - flush=True) + print( + "WARNING: overriding default arguments for {key}:{v} \ + with {key}:{v2}".format( + key=key, v=defaults[key], v2=getattr(args, key) + ), + flush=True, + ) else: setattr(args, key, defaults[key]) @@ -166,19 +205,23 @@ def validate_args(args, defaults={}): if args.global_batch_size is None: args.global_batch_size = args.micro_batch_size * args.data_parallel_size if args.rank == 0: - print('setting global batch size to {}'.format( - args.global_batch_size), flush=True) + print( + "setting global batch size to {}".format(args.global_batch_size), + flush=True, + ) assert args.global_batch_size > 0 if args.num_layers_per_virtual_pipeline_stage is not None: - assert args.pipeline_model_parallel_size > 2, \ - 'pipeline-model-parallel size should be greater than 2 with ' \ - 'interleaved schedule' - assert args.num_layers % args.num_layers_per_virtual_pipeline_stage == 0, \ - 'number of layers is not divisible by number of layers per virtual ' \ - 'pipeline stage' - args.virtual_pipeline_model_parallel_size = \ - (args.num_layers // args.transformer_pipeline_model_parallel_size) // \ - args.num_layers_per_virtual_pipeline_stage + assert args.pipeline_model_parallel_size > 2, ( + "pipeline-model-parallel size should be greater than 2 with " + "interleaved schedule" + ) + assert args.num_layers % args.num_layers_per_virtual_pipeline_stage == 0, ( + "number of layers is not divisible by number of layers per virtual " + "pipeline stage" + ) + args.virtual_pipeline_model_parallel_size = ( + args.num_layers // args.transformer_pipeline_model_parallel_size + ) // args.num_layers_per_virtual_pipeline_stage else: args.virtual_pipeline_model_parallel_size = None @@ -195,32 +238,34 @@ def validate_args(args, defaults={}): if not args.accumulate_allreduce_grads_in_fp32: args.accumulate_allreduce_grads_in_fp32 = True if args.rank == 0: - print('accumulate and all-reduce gradients in fp32 for ' - 'bfloat16 data type.', flush=True) + print( + "accumulate and all-reduce gradients in fp32 for " + "bfloat16 data type.", + flush=True, + ) if args.rank == 0: - print('using {} for parameters ...'.format(args.params_dtype), - flush=True) + print("using {} for parameters ...".format(args.params_dtype), flush=True) # If we do accumulation and all-reduces in fp32, we need to have local DDP # and we should make sure use-contiguous-buffers-in-local-ddp is not off. if args.accumulate_allreduce_grads_in_fp32: - assert args.DDP_impl == 'local' + assert args.DDP_impl == "local" assert args.use_contiguous_buffers_in_local_ddp # If we use the distributed optimizer, we need to have local DDP # and we should make sure use-contiguous-buffers-in-local-ddp is on. if args.use_distributed_optimizer: - assert args.DDP_impl == 'local' + assert args.DDP_impl == "local" assert args.use_contiguous_buffers_in_local_ddp # For torch DDP, we do not use contiguous buffer # if args.DDP_impl == 'torch': - if args.DDP_impl != 'local': + if args.DDP_impl != "local": args.use_contiguous_buffers_in_local_ddp = False if args.dataloader_type is None: - args.dataloader_type = 'single' + args.dataloader_type = "single" # Consumed tokens. args.consumed_train_samples = 0 @@ -238,47 +283,53 @@ def validate_args(args, defaults={}): if args.train_iters: # If we use iteration-based training, make sure the # sample-based options are off. - assert args.train_samples is None, \ - 'expected iteration-based training' - assert args.lr_decay_samples is None, \ - 'expected iteration-based learning rate decay' - assert args.lr_warmup_samples == 0, \ - 'expected iteration-based learning rate warmup' - assert args.rampup_batch_size is None, \ - 'expected no batch-size rampup for iteration-based training' + assert args.train_samples is None, "expected iteration-based training" + assert ( + args.lr_decay_samples is None + ), "expected iteration-based learning rate decay" + assert ( + args.lr_warmup_samples == 0 + ), "expected iteration-based learning rate warmup" + assert ( + args.rampup_batch_size is None + ), "expected no batch-size rampup for iteration-based training" if args.lr_warmup_fraction is not None: - assert args.lr_warmup_iters == 0, \ - 'can only specify one of lr-warmup-fraction and lr-warmup-iters' + assert ( + args.lr_warmup_iters == 0 + ), "can only specify one of lr-warmup-fraction and lr-warmup-iters" # Sample-based training. if args.train_samples: # If we use sample-based training, make sure the # iteration-based options are off. - assert args.train_iters is None, \ - 'expected sample-based training' - assert args.lr_decay_iters is None, \ - 'expected sample-based learning rate decay' - assert args.lr_warmup_iters == 0, \ - 'expected sample-based learnig rate warmup' + assert args.train_iters is None, "expected sample-based training" + assert args.lr_decay_iters is None, "expected sample-based learning rate decay" + assert args.lr_warmup_iters == 0, "expected sample-based learnig rate warmup" if args.lr_warmup_fraction is not None: - assert args.lr_warmup_samples == 0, \ - 'can only specify one of lr-warmup-fraction ' \ - 'and lr-warmup-samples' + assert args.lr_warmup_samples == 0, ( + "can only specify one of lr-warmup-fraction " "and lr-warmup-samples" + ) if args.num_layers is not None: - assert args.encoder_num_layers is None, \ - 'cannot have both num-layers and encoder-num-layers specified' + assert ( + args.encoder_num_layers is None + ), "cannot have both num-layers and encoder-num-layers specified" args.encoder_num_layers = args.num_layers else: if not args.use_dataset_only: - assert args.encoder_num_layers is not None, \ - 'either num-layers or encoder-num-layers should be specified' + assert ( + args.encoder_num_layers is not None + ), "either num-layers or encoder-num-layers should be specified" args.num_layers = args.encoder_num_layers # Check required arguments. if not args.use_dataset_only: - required_args = ['num_layers', 'hidden_size', 'num_attention_heads', - 'max_position_embeddings'] + required_args = [ + "num_layers", + "hidden_size", + "num_attention_heads", + "max_position_embeddings", + ] for req_arg in required_args: _check_arg_is_not_none(args, req_arg) @@ -321,13 +372,14 @@ def validate_args(args, defaults={}): assert args.save_interval is not None # Mixed precision checks. if args.fp16_lm_cross_entropy: - assert args.fp16, 'lm cross entropy in fp16 only support in fp16 mode.' + assert args.fp16, "lm cross entropy in fp16 only support in fp16 mode." if args.fp32_residual_connection: - assert args.fp16 or args.bf16, \ - 'residual connection in fp32 only supported when using fp16 or bf16.' + assert ( + args.fp16 or args.bf16 + ), "residual connection in fp32 only supported when using fp16 or bf16." if not args.use_dataset_only: - if args.weight_decay_incr_style == 'constant': + if args.weight_decay_incr_style == "constant": assert args.start_weight_decay is None assert args.end_weight_decay is None args.start_weight_decay = args.weight_decay @@ -336,50 +388,61 @@ def validate_args(args, defaults={}): assert args.start_weight_decay is not None assert args.end_weight_decay is not None - TORCH_MAJOR = int(torch.__version__.split('.')[0]) - TORCH_MINOR = int(torch.__version__.split('.')[1]) + TORCH_MAJOR = int(torch.__version__.split(".")[0]) + TORCH_MINOR = int(torch.__version__.split(".")[1]) # Persistent fused layer norm. if TORCH_MAJOR < 1 or (TORCH_MAJOR == 1 and TORCH_MINOR < 11): args.no_persist_layer_norm = True if args.rank == 0: - print('Persistent fused layer norm kernel is supported from ' - 'pytorch v1.11 (nvidia pytorch container paired with v1.11). ' - 'Defaulting to no_persist_layer_norm=True') + print( + "Persistent fused layer norm kernel is supported from " + "pytorch v1.11 (nvidia pytorch container paired with v1.11). " + "Defaulting to no_persist_layer_norm=True" + ) # Activation checkpointing. if args.distribute_checkpointed_activations: - assert args.checkpoint_activations, \ - 'for distribute-checkpointed-activations to work you '\ - 'need to enable checkpoint-activations' + assert args.checkpoint_activations, ( + "for distribute-checkpointed-activations to work you " + "need to enable checkpoint-activations" + ) # Activation recomputing. if args.distribute_saved_activations: - assert args.tensor_model_parallel_size > 1, 'can distribute ' \ - 'recomputed activations only across tensor model ' \ - 'parallel groups' - assert args.recompute_granularity == 'full', \ - 'distributed recompute activations is only '\ - 'application to full recompute granularity' - assert args.recompute_method is not None, \ - 'for distributed recompute activations to work you '\ - 'need to use a recompute method ' - assert TORCH_MAJOR >= 1 and TORCH_MINOR >= 10, \ - 'distributed recompute activations are supported for pytorch ' \ - 'v1.10 and above (Nvidia Pytorch container >= 21.07). Current ' \ - 'pytorch version is v%s.%s.' % (TORCH_MAJOR, TORCH_MINOR) + assert args.tensor_model_parallel_size > 1, ( + "can distribute " + "recomputed activations only across tensor model " + "parallel groups" + ) + assert args.recompute_granularity == "full", ( + "distributed recompute activations is only " + "application to full recompute granularity" + ) + assert args.recompute_method is not None, ( + "for distributed recompute activations to work you " + "need to use a recompute method " + ) + assert TORCH_MAJOR >= 1 and TORCH_MINOR >= 10, ( + "distributed recompute activations are supported for pytorch " + "v1.10 and above (Nvidia Pytorch container >= 21.07). Current " + "pytorch version is v%s.%s." % (TORCH_MAJOR, TORCH_MINOR) + ) # Tranformer-Engine/FP8 related checking if args.fp8_e4m3 or args.fp8_hybrid: - assert args.transformer_impl == 'transformer_engine', \ - 'transformer-engine required for fp8 training and inference' + assert ( + args.transformer_impl == "transformer_engine" + ), "transformer-engine required for fp8 training and inference" - assert not (args.fp8_e4m3 and args.fp8_hybrid), \ - 'cannot train with both fp8 e4m3 and hybrid formatting' + assert not ( + args.fp8_e4m3 and args.fp8_hybrid + ), "cannot train with both fp8 e4m3 and hybrid formatting" - if args.recompute_granularity == 'selective': - assert args.recompute_method is None, \ - 'recompute method is not yet supported for ' \ - 'selective recomputing granularity' + if args.recompute_granularity == "selective": + assert args.recompute_method is None, ( + "recompute method is not yet supported for " + "selective recomputing granularity" + ) # disable sequence parallelism when tp=1 # to avoid change in numerics when @@ -399,16 +462,20 @@ def validate_args(args, defaults={}): args.async_tensor_model_parallel_allreduce = False if not args.use_dataset_only: - if deepspeed.accelerator.get_accelerator().device_name() == "cuda" \ - and os.environ.get('CUDA_DEVICE_MAX_CONNECTIONS') != "1": + if ( + deepspeed.accelerator.get_accelerator().device_name() == "cuda" + and os.environ.get("CUDA_DEVICE_MAX_CONNECTIONS") != "1" + ): if args.sequence_parallel: raise RuntimeError( "Using sequence parallelism requires setting the environment variable " - "CUDA_DEVICE_MAX_CONNECTIONS to 1") + "CUDA_DEVICE_MAX_CONNECTIONS to 1" + ) if args.async_tensor_model_parallel_allreduce: raise RuntimeError( "Using async gradient all reduce requires setting the environment " - "variable CUDA_DEVICE_MAX_CONNECTIONS to 1") + "variable CUDA_DEVICE_MAX_CONNECTIONS to 1" + ) # Disable bias gelu fusion if we are disabling bias altogether if not args.add_bias_linear: @@ -421,16 +488,21 @@ def validate_args(args, defaults={}): with open(retro_args_path) as f: retro_args = types.SimpleNamespace(**json.load(f)) retro_args.retro_return_doc_ids = args.retro_return_doc_ids - retro_args.retro_gpt_retrieved_length = \ - args.retro_num_retrieved_chunks * \ - retro_args.retro_gpt_chunk_length + retro_args.retro_gpt_retrieved_length = ( + args.retro_num_retrieved_chunks * retro_args.retro_gpt_chunk_length + ) set_retro_args(retro_args) args.curriculum_learning_legacy = False args.compression_training = False # FlashAttention - args.use_flash_attn = args.use_flash_attn_v1 or args.use_flash_attn_triton or args.use_flash_attn_v2 or args.use_flash_attn_builder + args.use_flash_attn = ( + args.use_flash_attn_v1 + or args.use_flash_attn_triton + or args.use_flash_attn_v2 + or args.use_flash_attn_builder + ) # AML if args.aml_data_download_path is not None: @@ -443,19 +515,26 @@ def validate_args(args, defaults={}): if not args.use_dataset_only: if args.num_key_value_heads is None: args.num_key_value_heads = args.num_attention_heads - assert args.num_attention_heads % args.num_key_value_heads == 0, \ - f"num_attention_heads must be divisible by num_key_value_heads (got `num_attention_heads`: {args.num_attention_heads} " \ + assert args.num_attention_heads % args.num_key_value_heads == 0, ( + f"num_attention_heads must be divisible by num_key_value_heads (got `num_attention_heads`: {args.num_attention_heads} " f"and `num_key_value_heads`: {args.num_key_value_heads})." + ) if args.num_key_value_heads != args.num_attention_heads: # if GQA - assert not args.mos, 'GQA currently does not support args.mos' - assert not args.kd, 'GQA currently does not support args.kd' + assert not args.mos, "GQA currently does not support args.mos" + assert not args.kd, "GQA currently does not support args.kd" # Print arguments. _print_args("arguments", args) retro_args = get_retro_args() if retro_args and args != retro_args: - _print_args("retro arguments", types.SimpleNamespace(**{k:v for k,v in vars(retro_args).items() if k.startswith("retro")}, rank=args.rank)) + _print_args( + "retro arguments", + types.SimpleNamespace( + **{k: v for k, v in vars(retro_args).items() if k.startswith("retro")}, + rank=args.rank, + ), + ) return args @@ -463,20 +542,19 @@ def validate_args(args, defaults={}): def _print_args(title, args): """Print arguments.""" if args.rank == 0: - print(f'------------------------ {title} ------------------------', - flush=True) + print(f"------------------------ {title} ------------------------", flush=True) str_list = [] for arg in vars(args): - dots = '.' * (48 - len(arg)) - str_list.append(' {} {} {}'.format(arg, dots, getattr(args, arg))) + dots = "." * (48 - len(arg)) + str_list.append(" {} {} {}".format(arg, dots, getattr(args, arg))) for arg in sorted(str_list, key=lambda x: x.lower()): print(arg, flush=True) - print(f'-------------------- end of {title} ---------------------', - flush=True) + print(f"-------------------- end of {title} ---------------------", flush=True) def _check_arg_is_not_none(args, arg): - assert getattr(args, arg) is not None, '{} argument is None'.format(arg) + assert getattr(args, arg) is not None, "{} argument is None".format(arg) + def core_transformer_config_from_args(args): @@ -485,985 +563,1838 @@ def core_transformer_config_from_args(args): for f in dataclasses.fields(TransformerConfig): if hasattr(args, f.name): kw_args[f.name] = getattr(args, f.name) - kw_args['persist_layer_norm'] = not args.no_persist_layer_norm - kw_args['layernorm_zero_centered_gamma'] = args.apply_layernorm_1p - kw_args['deallocate_pipeline_outputs'] = True - kw_args['pipeline_dtype'] = args.params_dtype - kw_args['batch_p2p_comm'] = not args.overlap_p2p_comm + kw_args["persist_layer_norm"] = not args.no_persist_layer_norm + kw_args["layernorm_zero_centered_gamma"] = args.apply_layernorm_1p + kw_args["deallocate_pipeline_outputs"] = True + kw_args["pipeline_dtype"] = args.params_dtype + kw_args["batch_p2p_comm"] = not args.overlap_p2p_comm if args.swiglu: - kw_args['activation_func'] = F.silu - kw_args['gated_linear_unit'] = True - kw_args['bias_gelu_fusion'] = False + kw_args["activation_func"] = F.silu + kw_args["gated_linear_unit"] = True + kw_args["bias_gelu_fusion"] = False if args.init_method_xavier_uniform: - kw_args['init_method'] = torch.nn.init.xavier_uniform_ - kw_args['scaled_init_method'] = torch.nn.init.xavier_uniform_ + kw_args["init_method"] = torch.nn.init.xavier_uniform_ + kw_args["scaled_init_method"] = torch.nn.init.xavier_uniform_ return TransformerConfig(**kw_args) + def _add_transformer_engine_args(parser): - group = parser.add_argument_group(title='Transformer-Engine') - - group.add_argument('--fp8-e4m3', action='store_true', - help='E4M3 TransformerLayer', dest='fp8_e4m3') - group.add_argument('--fp8-hybrid', action='store_true', - help='Hybrid FP8 TransformerLayer', dest='fp8_hybrid') - group.add_argument('--no-fp8-wgrad', action='store_false', - help='Execute wgrad in higher precision even for FP8 runs', dest='fp8_wgrad') - group.add_argument('--fp8-margin', type=int, default=0, - help='Scaling margin for fp8', dest='fp8_margin') - group.add_argument('--fp8-interval', type=int, default=1, - help='Scaling update interval for fp8', dest='fp8_interval') - group.add_argument('--transformer-impl', default='local', - choices=['local', 'transformer_engine'], - help='Which Transformer implementation to use.', - dest='transformer_impl') - group.add_argument('--fp8-amax-history-len', type=int, default=1, - help='Number of steps for which amax history is recorded per tensor', - dest='fp8_amax_history_len') - group.add_argument('--fp8-amax-compute-algo', default='most_recent', - choices=['most_recent', 'max'], - help='Algorithm for computing amax from history', - dest='fp8_amax_compute_algo') + group = parser.add_argument_group(title="Transformer-Engine") + + group.add_argument( + "--fp8-e4m3", action="store_true", help="E4M3 TransformerLayer", dest="fp8_e4m3" + ) + group.add_argument( + "--fp8-hybrid", + action="store_true", + help="Hybrid FP8 TransformerLayer", + dest="fp8_hybrid", + ) + group.add_argument( + "--no-fp8-wgrad", + action="store_false", + help="Execute wgrad in higher precision even for FP8 runs", + dest="fp8_wgrad", + ) + group.add_argument( + "--fp8-margin", + type=int, + default=0, + help="Scaling margin for fp8", + dest="fp8_margin", + ) + group.add_argument( + "--fp8-interval", + type=int, + default=1, + help="Scaling update interval for fp8", + dest="fp8_interval", + ) + group.add_argument( + "--transformer-impl", + default="local", + choices=["local", "transformer_engine"], + help="Which Transformer implementation to use.", + dest="transformer_impl", + ) + group.add_argument( + "--fp8-amax-history-len", + type=int, + default=1, + help="Number of steps for which amax history is recorded per tensor", + dest="fp8_amax_history_len", + ) + group.add_argument( + "--fp8-amax-compute-algo", + default="most_recent", + choices=["most_recent", "max"], + help="Algorithm for computing amax from history", + dest="fp8_amax_compute_algo", + ) return parser + def _add_inference_args(parser): - group = parser.add_argument_group(title='inference') - - group.add_argument('--inference-batch-times-seqlen-threshold', - type=int, default=512, - help='During inference, if batch-size times ' - 'sequence-length is smaller than this threshold ' - 'then we will not use pipelining, otherwise we will.') - group.add_argument('--max-tokens-to-oom', - type=int, default=12000, - help='Maximum number of tokens during inference' - 'tokens here is # in prompt + # to generate' - 'Allows us to throw an error before OOM crashes server') - group.add_argument('--output-bert-embeddings', action='store_true', - help='Output Bert embeddings (via mean pooling) from ' - 'model, rather than its binary head output or entire ' - 'hidden batch.') - group.add_argument('--bert-embedder-type', default="megatron", - choices=["megatron", "huggingface"], - help='Select either Megatron or Huggingface as the ' - 'Bert embedder.') + group = parser.add_argument_group(title="inference") + + group.add_argument( + "--inference-batch-times-seqlen-threshold", + type=int, + default=512, + help="During inference, if batch-size times " + "sequence-length is smaller than this threshold " + "then we will not use pipelining, otherwise we will.", + ) + group.add_argument( + "--max-tokens-to-oom", + type=int, + default=12000, + help="Maximum number of tokens during inference" + "tokens here is # in prompt + # to generate" + "Allows us to throw an error before OOM crashes server", + ) + group.add_argument( + "--output-bert-embeddings", + action="store_true", + help="Output Bert embeddings (via mean pooling) from " + "model, rather than its binary head output or entire " + "hidden batch.", + ) + group.add_argument( + "--bert-embedder-type", + default="megatron", + choices=["megatron", "huggingface"], + help="Select either Megatron or Huggingface as the " "Bert embedder.", + ) return parser def _add_retro_args(parser): - group = parser.add_argument_group(title='retro') - - group.add_argument('--retro-workdir', default=None, - help='Retro working directory, which contains the ' - 'preprocessed data for for pretraining. This directory ' - 'is built during preprocessing (see ' - 'tools/retro/README.md), and contains subdirectories ' - 'for the chunk database and pretraining neighbors.') - group.add_argument('--retro-add-retriever', - action='store_true', default=False, - help='Add a retriever to the transformer, for use in ' - 'pretraining a Retro model.') - group.add_argument('--retro-cyclic-train-iters', type=int, default=None, - help='Set number of training iterations for cyclic ' - 'Retro training.') - group.add_argument('--retro-encoder-layers', type=int, default=2, - help='Number of layers to use for the retrieval ' - 'encoder.') - group.add_argument('--retro-encoder-hidden-dropout', - type=float, default=0.1, help='Hidden dropout for ' - 'retrieval encoder.') - group.add_argument('--retro-encoder-attention-dropout', - type=float, default=0.1, help='Attention dropout for ' - 'retrieval encoder.') - group.add_argument("--retro-num-neighbors", type=int, default=2, - help='Number of neighbors to retrieve during ' - 'pretraining.') - group.add_argument("--retro-num-retrieved-chunks", type=int, default=2, - help='Number of chunks to retrieve from the retrieval ' - 'database.') - group.add_argument("--retro-return-doc-ids", action="store_true", - help="Turn this on when preprocessing retro data.") + group = parser.add_argument_group(title="retro") + + group.add_argument( + "--retro-workdir", + default=None, + help="Retro working directory, which contains the " + "preprocessed data for for pretraining. This directory " + "is built during preprocessing (see " + "tools/retro/README.md), and contains subdirectories " + "for the chunk database and pretraining neighbors.", + ) + group.add_argument( + "--retro-add-retriever", + action="store_true", + default=False, + help="Add a retriever to the transformer, for use in " + "pretraining a Retro model.", + ) + group.add_argument( + "--retro-cyclic-train-iters", + type=int, + default=None, + help="Set number of training iterations for cyclic " "Retro training.", + ) + group.add_argument( + "--retro-encoder-layers", + type=int, + default=2, + help="Number of layers to use for the retrieval " "encoder.", + ) + group.add_argument( + "--retro-encoder-hidden-dropout", + type=float, + default=0.1, + help="Hidden dropout for " "retrieval encoder.", + ) + group.add_argument( + "--retro-encoder-attention-dropout", + type=float, + default=0.1, + help="Attention dropout for " "retrieval encoder.", + ) + group.add_argument( + "--retro-num-neighbors", + type=int, + default=2, + help="Number of neighbors to retrieve during " "pretraining.", + ) + group.add_argument( + "--retro-num-retrieved-chunks", + type=int, + default=2, + help="Number of chunks to retrieve from the retrieval " "database.", + ) + group.add_argument( + "--retro-return-doc-ids", + action="store_true", + help="Turn this on when preprocessing retro data.", + ) # Enforce argument naming convention. for action in group._group_actions: prefix = action.dest.split("_")[0] - assert prefix == "retro", \ - "Retro args must be prefixed with '--retro-*', for consistent " \ + assert prefix == "retro", ( + "Retro args must be prefixed with '--retro-*', for consistent " "styling. Please fix '%s'." % ", ".join(action.option_strings) + ) return parser def _add_network_size_args(parser): - group = parser.add_argument_group(title='network size') - - group.add_argument('--num-layers', type=int, default=None, - help='Number of transformer layers.') - group.add_argument('--encoder-num-layers', type=int, default=None, - help='Number of encoder transformer layers.') - group.add_argument('--decoder-num-layers', type=int, default=None, - help='Number of decoder transformer layers.') - group.add_argument('--num-experts', type=int, nargs='+', default=[1,], - help='number of experts list, MoE related.') - group.add_argument('--mlp-type', type=str, default='standard', - help='Only applicable when num-experts > 1, accepts [standard, residual]') - group.add_argument('--topk', type=int, default=1, - help='Sets the k in TopK gating for MoE layers') - group.add_argument('--expert-interval', type=int, default=2, - help='Use experts in every "expert-interval" layers') - group.add_argument('--hidden-size', type=int, default=None, - help='Tansformer hidden size.') - group.add_argument('--ffn-hidden-size', type=int, default=None, - help='Transformer Feed-Forward Network hidden size. ' - 'This is set to 4*hidden-size if not provided') - group.add_argument('--num-attention-heads', type=int, default=None, - help='Number of transformer attention heads.') - group.add_argument('--num-key-value-heads', type=int, default=None, - help='Number of key_value heads that should be used to implement Grouped Query Attention.') - group.add_argument('--kv-channels', type=int, default=None, - help='Projection weights dimension in multi-head ' - 'attention. This is set to ' - ' args.hidden_size // args.num_attention_heads ' - 'if not provided.') - group.add_argument('--max-position-embeddings', type=int, default=None, - help='Maximum number of position embeddings to use. ' - 'This is the size of position embedding.') - group.add_argument('--use-rotary-position-embeddings', action='store_true', - help='Use rotary positional embeddings or not') - group.add_argument('--rotary-position-embeddings-theta', type=int, default=10000, - help='Rotary positional embeddings theta value.', - dest='rope_theta') - group.add_argument('--rotary-percent', type=float, default=1.0, - help='Percent of rotary dimension to use, default 100%') - group.add_argument('--no-position-embedding', - action='store_false', - help='Disable position embedding.', - dest='add_position_embedding') - group.add_argument('--make-vocab-size-divisible-by', type=int, default=128, - help='Pad the vocab size to be divisible by this value.' - 'This is added for computational efficieny reasons.') - group.add_argument('--normalization', type=str, default='layernorm', - choices=['layernorm', 'rmsnorm'], - help='Options for layer normalization type:' - ' layernorm' - ' rmsnorm') - group.add_argument('--layernorm-epsilon', type=float, default=1e-5, - help='Layer norm epsilon.') - group.add_argument('--apply-layernorm-1p', action='store_true', - help='Adjust LayerNorm weights such that they are centered ' - 'around zero. This improves numerical stability.') - group.add_argument('--disable-mem-efficient-ln', action='store_false', - help='Disable the memory-efficient fused LayerNorm optimization ' - 'introduced in https://github.com/NVIDIA/apex/pull/1715', dest='mem_efficient_ln') - group.add_argument('--apply-residual-connection-post-layernorm', - action='store_true', - help='If set, use original BERT residula connection ' - 'ordering.') - group.add_argument('--openai-gelu', action='store_true', - help='Use OpenAIs GeLU implementation. This option' - 'should not be used unless for backward compatibility' - 'reasons.') - group.add_argument('--squared-relu', action='store_true', - help='Use squared relu activation instead of default gelu') - group.add_argument('--swiglu', action='store_true', - help='Use gated linear units and SiLU activation instead of default gelu') - group.add_argument('--onnx-safe', type=bool, required=False, - help='Use workarounds for known problems with ' - 'Torch ONNX exporter') - group.add_argument('--bert-no-binary-head', action='store_false', - help='Disable BERT binary head.', - dest='bert_binary_head') - group.add_argument('--num-experts-switch', type=int, default=None, - help='Number of Experts in Switch Transformer (None means no Switch)') - group.add_argument('--untie-embeddings-and-output-weights', action='store_true', - help='Untie embeddings and output weights.'), - group.add_argument('--embedding-weights-in-fp32', action='store_true', - help='Cast word embedding weights to fp32 before embedding fwd.'), - group.add_argument('--kill-switch-file', type=str, default=None, - help='Location of kill switch file. ' - 'If found will automatically exit the program at runtime.') - return parser - - -def _add_logging_args(parser): - group = parser.add_argument_group(title='logging') - - group.add_argument('--log-params-norm', action='store_true', - help='If set, calculate and log parameters norm.') - group.add_argument('--log-num-zeros-in-grad', action='store_true', - help='If set, calculate and log the number of zeros in gradient.') - group.add_argument('--timing-log-level', type=int, - default=0, choices=range(0,3), - help='Granularity level to measure and report timing. ' - ' 0: report only iteration time and make sure timing ' - ' does not introduce extra overhead.' - ' 1: report timing for operations that are executed ' - ' very limited times (basically once) during ' - ' each iteration (such as gradient all-reduce) ' - ' 2: report timing for operations that migh be ' - ' executed numerous times during each iteration. ' - 'Note that setting the level to 1 or 2 might ' - 'cause increase in iteration time.') - group.add_argument('--no-barrier-with-level-1-timing', action='store_false', - help='If not set, use barrier with level 1 time ' - 'measurements. Note that this is up to the user ' - 'to make sure calling barrier with their timers ' - 'will not result in hangs. This can happen if for ' - 'example the user adds a level 1 timer that is not ' - 'called by all ranks.', - dest='barrier_with_L1_time') - group.add_argument('--timing-log-option', type=str, default='minmax', - choices=['max', 'minmax', 'all'], - help='Options for logging timing:' - ' max: report the max timing across all ranks' - ' minmax: report min and max timings across all ranks' - ' all: report timings of all ranks.') - group.add_argument('--tensorboard-log-interval', type=int, default=1, - help='Report to tensorboard interval.') - group.add_argument('--tensorboard-queue-size', type=int, default=1000, - help='Size of the tensorboard queue for pending events ' - 'and summaries before one of the ‘add’ calls forces a ' - 'flush to disk.') - group.add_argument('--log-timers-to-tensorboard', action='store_true', - help='If set, write timers to tensorboard.') - group.add_argument('--log-batch-size-to-tensorboard', action='store_true', - help='If set, write batch-size to tensorboard.') - group.add_argument('--no-log-learnig-rate-to-tensorboard', - action='store_false', - help='Disable learning rate logging to tensorboard.', - dest='log_learning_rate_to_tensorboard') - group.add_argument('--no-log-loss-scale-to-tensorboard', - action='store_false', - help='Disable loss-scale logging to tensorboard.', - dest='log_loss_scale_to_tensorboard') - group.add_argument('--log-validation-ppl-to-tensorboard', - action='store_true', - help='If set, write validation perplexity to ' - 'tensorboard.') - group.add_argument('--log-optimizer-states-to-tensorboard', - action='store_true', - help='If set, write various optimizer states to ' - 'tensorboard. This feature may consume extra GPU memory.') - group.add_argument('--log-memory-to-tensorboard', - action='store_true', - help='Enable memory logging to tensorboard.') - group.add_argument('--log-world-size-to-tensorboard', - action='store_true', - help='Enable world size logging to tensorboard.') - group.add_argument('--wandb-project', type=str, default='', - help='The wandb project name. Ignore wandb by default.') - group.add_argument('--wandb-exp-name', type=str, default='', - help='The wandb experiment name.') - group.add_argument('--wandb-save-dir', type=str, default='', - help='Path to save the wandb results locally.') - - return parser - - -def _add_regularization_args(parser): - group = parser.add_argument_group(title='regularization') - - group.add_argument('--attention-dropout', type=float, default=0.1, - help='Post attention dropout probability.') - group.add_argument('--hidden-dropout', type=float, default=0.1, - help='Dropout probability for hidden state transformer.') - group.add_argument('--weight-decay', type=float, default=0.01, - help='Weight decay coefficient for L2 regularization.') - group.add_argument('--start-weight-decay', type=float, - help='Initial weight decay coefficient for L2 regularization.') - group.add_argument('--end-weight-decay', type=float, - help='End of run weight decay coefficient for L2 regularization.') - group.add_argument('--weight-decay-incr-style', type=str, default='constant', - choices=['constant', 'linear', 'cosine'], - help='Weight decay increment function.') - group.add_argument('--clip-grad', type=float, default=1.0, - help='Gradient clipping based on global L2 norm.') - group.add_argument('--sophiag-beta1', type=float, default=0.9, - help='First coefficient for computing running averages ' - 'of gradient and its hessian') - group.add_argument('--sophiag-beta2', type=float, default=0.95, - help='Second coefficient for computing running averages ' - 'of gradient and its hessian') - group.add_argument('--sophiag-rho', type=float, default=0.01, - help='SophiaG clipping threshhold') - - group.add_argument('--adam-beta1', type=float, default=0.9, - help='First coefficient for computing running averages ' - 'of gradient and its square') - group.add_argument('--adam-beta2', type=float, default=0.999, - help='Second coefficient for computing running averages ' - 'of gradient and its square') - group.add_argument('--adam-eps', type=float, default=1e-08, - help='Term added to the denominator to improve' - 'numerical stability') - group.add_argument('--sgd-momentum', type=float, default=0.9, - help='Momentum factor for sgd') - - return parser - + group = parser.add_argument_group(title="network size") -def _add_training_args(parser): - group = parser.add_argument_group(title='training') - - group.add_argument('--micro-batch-size', type=int, default=None, - help='Batch size per model instance (local batch size). ' - 'Global batch size is local batch size times data ' - 'parallel size times number of micro batches.') - group.add_argument('--batch-size', type=int, default=None, - help='Old batch size parameter, do not use. ' - 'Use --micro-batch-size instead') - group.add_argument('--global-batch-size', type=int, default=None, - help='Training batch size. If set, it should be a ' - 'multiple of micro-batch-size times data-parallel-size. ' - 'If this value is None, then ' - 'use micro-batch-size * data-parallel-size as the ' - 'global batch size. This choice will result in 1 for ' - 'number of micro-batches.') - group.add_argument('--rampup-batch-size', nargs='*', default=None, - help='Batch size ramp up with the following values:' - ' --rampup-batch-size ' - ' ' - ' ' - 'For example:' - ' --rampup-batch-size 16 8 300000 \\ ' - ' --global-batch-size 1024' - 'will start with global batch size 16 and over ' - ' (1024 - 16) / 8 = 126 intervals will increase' - 'the batch size linearly to 1024. In each interval' - 'we will use approximately 300000 / 126 = 2380 samples.') - group.add_argument('--recompute-activations', action='store_true', - help='recompute activation to allow for training ' - 'with larger models, sequences, and batch sizes.') - group.add_argument('--recompute-granularity', type=str, default=None, - choices=['full', 'selective'], - help='Checkpoint activations to allow for training ' - 'with larger models, sequences, and batch sizes. ' - 'It is supported at two granularities 1) full: ' - 'whole transformer layer is recomputed, ' - '2) selective: core attention part of the transformer ' - 'layer is recomputed.') - group.add_argument('--distribute-saved-activations', - action='store_true', - help='If set, distribute recomputed activations ' - 'across model parallel group.') - group.add_argument('--recompute-method', type=str, default=None, - choices=['uniform', 'block'], - help='1) uniform: uniformly divide the total number of ' - 'Transformer layers and recompute the input activation of ' - 'each divided chunk at specified granularity, ' - '2) recompute the input activations of only a set number of ' - 'individual Transformer layers per pipeline stage and do the ' - 'rest without any recomputing at specified granularity' - 'default) do not apply activations recompute to any layers') - group.add_argument('--recompute-num-layers', type=int, default=1, - help='1) uniform: the number of Transformer layers in each ' - 'uniformly divided recompute unit, ' - '2) block: the number of individual Transformer layers ' - 'to recompute within each pipeline stage.') - group.add_argument('--enable-zbh1-pipeline', action='store_true', - help='Activate zero bubble pipeline parallelism schedule method') - group.add_argument('--enable-zbh1-exact-semantics', action='store_true', - help='Use an exact semantics for zbh1 schedule, might be slower than the default.') - - # deprecated - # HACK: added back arguments because DeepSpeed still relies on the old - # activation checkpointing mechanism. - group.add_argument('--checkpoint-activations', action='store_true', - help='Checkpoint activation to allow for training ' - 'with larger models, sequences, and batch sizes.') - group.add_argument('--distribute-checkpointed-activations', - action='store_true', - help='If set, distribute checkpointed activations ' - 'across model parallel group.') - group.add_argument('--checkpoint-num-layers', type=int, default=1, - help='chunk size (number of layers) for checkpointing.') - group.add_argument('--train-iters', type=int, default=None, - help='Total number of iterations to train over all ' - 'training runs. Note that either train-iters or ' - 'train-samples should be provided.') - group.add_argument('--train-samples', type=int, default=None, - help='Total number of samples to train over all ' - 'training runs. Note that either train-iters or ' - 'train-samples should be provided.') - group.add_argument('--train-tokens', type=int, default=None, - help='Total number of tokens to train over all ' - 'training runs.') - group.add_argument('--random-ltd', - action='store_true', - help='enable random layer token drop') - group.add_argument('--log-interval', type=int, default=100, - help='Report loss and timing interval.') - group.add_argument('--exit-interval', type=int, default=None, - help='Exit the program after the iteration is divisible ' - 'by this value.') - group.add_argument('--exit-duration-in-mins', type=int, default=None, - help='Exit the program after this many minutes.') - group.add_argument('--exit-signal-handler', action='store_true', - help='Dynamically save the checkpoint and shutdown the ' - 'training if SIGTERM is received') - group.add_argument('--tensorboard-dir', type=str, default=None, - help='Write TensorBoard logs to this directory.') - group.add_argument('--trace-dir', type=str, default="./trace/", - help='Write trace logs to this directory.') - group.add_argument('--no-masked-softmax-fusion', - action='store_false', - help='Disable fusion of query_key_value scaling, ' - 'masking, and softmax.', - dest='masked_softmax_fusion') - group.add_argument('--no-bias-gelu-fusion', action='store_false', - help='Disable bias and gelu fusion.', - dest='bias_gelu_fusion') - group.add_argument('--no-bias-dropout-fusion', action='store_false', - help='Disable bias and dropout fusion.', - dest='bias_dropout_fusion') - group.add_argument('--disable-moe-token-dropping', action='store_false', - help='Disable MoE expert token dropping.', - dest='moe_token_dropping') - group.add_argument('--moe-train-capacity-factor', type=float, default=1.0, - help='The capacity of the MoE expert at training time') - group.add_argument('--moe-eval-capacity-factor', type=float, default=1.0, - help='The capacity of the MoE expert at eval time.') - group.add_argument('--moe-min-capacity', type=int, default=4, - help='The minimum capacity per MoE expert regardless of the capacity_factor.') - group.add_argument('--moe-loss-coeff', type=float, default=0.1, - help='Scaling coefficient for adding MoE loss to model loss') - group.add_argument('--create-moe-param-group', action='store_true', - help='Create separate groups for MoE params.' - 'This is necessary for techniques like ZeRO.') - group.add_argument('--disable-moe-top2-2nd-expert-sampling', action='store_false', - help='Disable MoE top2 sampling of the 2nd expert. Instead of sampling, use argmax.', - dest='moe_top2_2nd_expert_sampling') - group.add_argument('--use-flash-attn', '--use-flash-attn-v1', dest='use_flash_attn_v1', action='store_true', - help='use first version FlashAttention implementation of attention. ' - 'https://arxiv.org/abs/2205.14135') - group.add_argument('--use-flash-attn-v2', action='store_true', - help='use second version FlashAttention implementation of attention. ' - 'https://arxiv.org/abs/2307.08691') - group.add_argument('--use-flash-attn-triton', action='store_true', - help='use FlashAttention implementation of attention using Triton.') - group.add_argument('--use-flash-attn-builder', action='store_true', - help='use FlashAttention op builder.') - group.add_argument('--disable-bias-linear', action='store_false', - help='Disable bias in the linear layers', - dest='add_bias_linear') - group.add_argument( - '--optimizer', - type=str, - default='adam', - choices=[ - 'adam', - 'adam8bit', - 'adamw', - 'adamwschedulefree', - 'apex.adam', - 'apex.sgd', - 'ds.fusedlamb', - 'ds.onebitlamb', - 'galoreadamw', - 'galoreadamw8bit', - 'galoreadamw8bitperlayer', - 'ipex.fusedlamb', - 'ipex.lamb', - 'shampoo', - 'sgd', - 'sgdschedulefree', - 'sophiag' + group.add_argument( + "--num-layers", type=int, default=None, help="Number of transformer layers." + ) + group.add_argument( + "--encoder-num-layers", + type=int, + default=None, + help="Number of encoder transformer layers.", + ) + group.add_argument( + "--decoder-num-layers", + type=int, + default=None, + help="Number of decoder transformer layers.", + ) + group.add_argument( + "--num-experts", + type=int, + nargs="+", + default=[ + 1, ], - help='Optimizer function' + help="number of experts list, MoE related.", ) group.add_argument( - "--schedulefree-for-each", + "--mlp-type", + type=str, + default="standard", + help="Only applicable when num-experts > 1, accepts [standard, residual]", + ) + group.add_argument( + "--topk", type=int, default=1, help="Sets the k in TopK gating for MoE layers" + ) + group.add_argument( + "--expert-interval", + type=int, + default=2, + help='Use experts in every "expert-interval" layers', + ) + group.add_argument( + "--hidden-size", type=int, default=None, help="Tansformer hidden size." + ) + group.add_argument( + "--ffn-hidden-size", + type=int, + default=None, + help="Transformer Feed-Forward Network hidden size. " + "This is set to 4*hidden-size if not provided", + ) + group.add_argument( + "--num-attention-heads", + type=int, + default=None, + help="Number of transformer attention heads.", + ) + group.add_argument( + "--num-key-value-heads", + type=int, + default=None, + help="Number of key_value heads that should be used to implement Grouped Query Attention.", + ) + group.add_argument( + "--kv-channels", + type=int, + default=None, + help="Projection weights dimension in multi-head " + "attention. This is set to " + " args.hidden_size // args.num_attention_heads " + "if not provided.", + ) + group.add_argument( + "--max-position-embeddings", + type=int, + default=None, + help="Maximum number of position embeddings to use. " + "This is the size of position embedding.", + ) + group.add_argument( + "--use-rotary-position-embeddings", action="store_true", - help=""" - Use a foreach-backed implementation of the schedulefree optimizers. - Should be significantly faster, - but will have a higher peak memory usage. - """, + help="Use rotary positional embeddings or not", ) - group.add_argument('--dataloader-type', type=str, default=None, - choices=['single', 'cyclic'], - help='Single pass vs multiple pass data loader') - group.add_argument('--ds-inference', action='store_true', - help='DeepSpeed inference engine being used') - group.add_argument('--cpu-optimizer', action='store_true', - help='Run optimizer on CPU') - group.add_argument('--cpu_torch_adam', action='store_true', - help='Use Torch Adam as optimizer on CPU.') - group.add_argument('--ds_fused_adam', action='store_true', - help='Use DeepSpeed FusedAdam as optimizer.') - group.add_argument('--no-pipeline-parallel', action='store_true', - help='Disable pipeline parallelism') - group.add_argument('--use-tutel', action='store_true', - help='Use Tutel optimization for MoE') - group.add_argument('--inference', action='store_true', - help='Very basic inference mode: not allocating optim/lr - requires ZERO_STAGE=0') - - group.add_argument('--no-async-tensor-model-parallel-allreduce', - action='store_false', - help='Disable asynchronous execution of ' - 'tensor-model-parallel all-reduce with weight ' - 'gradient compuation of a column-linear layer.', - dest='async_tensor_model_parallel_allreduce') - group.add_argument('--no-persist-layer-norm', action='store_true', - help='Disable using persistent fused layer norm kernel. ' - 'This kernel supports only a set of hidden sizes. Please ' - 'check persist_ln_hidden_sizes if your hidden ' - 'size is supported.') - group.add_argument('--sequence-parallel', action='store_true', - help='Enable Megatron-LM\'s sequence parallel optimization.') - group.add_argument('--ds-sequence-parallel-size', type=int, default=1, - help='Enable DeepSpeed\'s sequence parallel. Cannot be combined with "--sequence-parallel", which enables Megatron-LM\'s sequence parallel.') - group.add_argument('--force-ds-sequence-parallel', action='store_true', - help='use DeepSpeed sequence parallelism regardless of sequence parallel size.') - group.add_argument('--no-gradient-accumulation-fusion', - action='store_false', - help='Disable fusing gradient accumulation to weight ' - 'gradient computation of linear layers', - dest='gradient_accumulation_fusion') - group.add_argument('--use-dataset-only', type=bool, required=False, default=False, - help='If set to True, only use the megatron dataset for external trainer ') - # group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') group.add_argument( - "--train-range-to-skip", - action="extend", - nargs="+", - type=int, - help=("Range of iters to skip during training. Must be in pairs."), - ) - group.add_argument('--train-iters-to-skip', action="extend", nargs="+", type=str, - help=( - "Specific train iterations to skip when training. " - "Load the data and just perform a noop." - )) - return parser - - -def _add_initialization_args(parser): - group = parser.add_argument_group(title='initialization') - - group.add_argument('--seed', type=int, default=1234, - help='Random seed used for python, numpy, ' - 'pytorch, and cuda.') - group.add_argument('--data-parallel-random-init', action='store_true', - help='Enable random initialization of params ' - 'across data parallel ranks') - group.add_argument('--init-method-std', type=float, default=0.02, - help='Standard deviation of the zero mean normal ' - 'distribution used for weight initialization.') - group.add_argument('--init-method-xavier-uniform', action='store_true', - help='Enable Xavier uniform parameter initialization') - + "--rotary-position-embeddings-theta", + type=int, + default=10000, + help="Rotary positional embeddings theta value.", + dest="rope_theta", + ) + group.add_argument( + "--rotary-percent", + type=float, + default=1.0, + help="Percent of rotary dimension to use, default 100%", + ) + group.add_argument( + "--no-position-embedding", + action="store_false", + help="Disable position embedding.", + dest="add_position_embedding", + ) + group.add_argument( + "--make-vocab-size-divisible-by", + type=int, + default=128, + help="Pad the vocab size to be divisible by this value." + "This is added for computational efficieny reasons.", + ) + group.add_argument( + "--normalization", + type=str, + default="layernorm", + choices=["layernorm", "rmsnorm"], + help="Options for layer normalization type:" " layernorm" " rmsnorm", + ) + group.add_argument( + "--layernorm-epsilon", type=float, default=1e-5, help="Layer norm epsilon." + ) + group.add_argument( + "--apply-layernorm-1p", + action="store_true", + help="Adjust LayerNorm weights such that they are centered " + "around zero. This improves numerical stability.", + ) + group.add_argument( + "--disable-mem-efficient-ln", + action="store_false", + help="Disable the memory-efficient fused LayerNorm optimization " + "introduced in https://github.com/NVIDIA/apex/pull/1715", + dest="mem_efficient_ln", + ) + group.add_argument( + "--apply-residual-connection-post-layernorm", + action="store_true", + help="If set, use original BERT residula connection " "ordering.", + ) + group.add_argument( + "--openai-gelu", + action="store_true", + help="Use OpenAIs GeLU implementation. This option" + "should not be used unless for backward compatibility" + "reasons.", + ) + group.add_argument( + "--squared-relu", + action="store_true", + help="Use squared relu activation instead of default gelu", + ) + group.add_argument( + "--swiglu", + action="store_true", + help="Use gated linear units and SiLU activation instead of default gelu", + ) + group.add_argument( + "--onnx-safe", + type=bool, + required=False, + help="Use workarounds for known problems with " "Torch ONNX exporter", + ) + group.add_argument( + "--bert-no-binary-head", + action="store_false", + help="Disable BERT binary head.", + dest="bert_binary_head", + ) + group.add_argument( + "--num-experts-switch", + type=int, + default=None, + help="Number of Experts in Switch Transformer (None means no Switch)", + ) + group.add_argument( + "--untie-embeddings-and-output-weights", + action="store_true", + help="Untie embeddings and output weights.", + ), + group.add_argument( + "--embedding-weights-in-fp32", + action="store_true", + help="Cast word embedding weights to fp32 before embedding fwd.", + ), + group.add_argument( + "--kill-switch-file", + type=str, + default=None, + help="Location of kill switch file. " + "If found will automatically exit the program at runtime.", + ) return parser -def _add_learning_rate_args(parser): - group = parser.add_argument_group(title='learning rate') - - group.add_argument('--lr', type=float, default=None, - help='Initial learning rate. Depending on decay style ' - 'and initial warmup, the learing rate at each ' - 'iteration would be different.') - group.add_argument('--lr-decay-style', type=str, default='linear', - choices=['constant', 'linear', 'cosine', 'inverse-square-root'], - help='Learning rate decay function.') - group.add_argument('--lr-decay-iters', type=int, default=None, - help='number of iterations to decay learning rate over,' - ' If None defaults to `--train-iters`') - group.add_argument('--lr-decay-samples', type=int, default=None, - help='number of samples to decay learning rate over,' - ' If None defaults to `--train-samples`') - group.add_argument('--lr-decay-tokens', type=int, default=None, - help='number of tokens to decay learning rate over,' - ' If not None will override iter/sample-based decay') - group.add_argument('--lr-warmup-fraction', type=float, default=None, - help='fraction of lr-warmup-(iters/samples) to use ' - 'for warmup (as a float)') - group.add_argument('--lr-warmup-iters', type=int, default=0, - help='number of iterations to linearly warmup ' - 'learning rate over.') - group.add_argument('--lr-warmup-samples', type=int, default=0, - help='number of samples to linearly warmup ' - 'learning rate over.') - group.add_argument('--lr-warmup-tokens', type=int, default=None, - help='number of tokens to linearly warmup ' - 'learning rate over.') - group.add_argument('--warmup', type=int, default=None, - help='Old lr warmup argument, do not use. Use one of the' - '--lr-warmup-* arguments above') - group.add_argument('--min-lr', type=float, default=0.0, - help='Minumum value for learning rate. The scheduler' - 'clip values below this threshold.') - group.add_argument('--override-opt_param-scheduler', action='store_true', - help='Reset the values of the scheduler (learning rate,' - 'warmup iterations, minimum learning rate, maximum ' - 'number of iterations, and decay style from input ' - 'arguments and ignore values from checkpoints. Note' - 'that all the above values will be reset.') - group.add_argument('--use-checkpoint-opt_param-scheduler', action='store_true', - help='Use checkpoint to set the values of the scheduler ' - '(learning rate, warmup iterations, minimum learning ' - 'rate, maximum number of iterations, and decay style ' - 'from checkpoint and ignore input arguments.') - - return parser +def _add_logging_args(parser): + group = parser.add_argument_group(title="logging") + group.add_argument( + "--log-params-norm", + action="store_true", + help="If set, calculate and log parameters norm.", + ) + group.add_argument( + "--log-num-zeros-in-grad", + action="store_true", + help="If set, calculate and log the number of zeros in gradient.", + ) + group.add_argument( + "--timing-log-level", + type=int, + default=0, + choices=range(0, 3), + help="Granularity level to measure and report timing. " + " 0: report only iteration time and make sure timing " + " does not introduce extra overhead." + " 1: report timing for operations that are executed " + " very limited times (basically once) during " + " each iteration (such as gradient all-reduce) " + " 2: report timing for operations that migh be " + " executed numerous times during each iteration. " + "Note that setting the level to 1 or 2 might " + "cause increase in iteration time.", + ) + group.add_argument( + "--no-barrier-with-level-1-timing", + action="store_false", + help="If not set, use barrier with level 1 time " + "measurements. Note that this is up to the user " + "to make sure calling barrier with their timers " + "will not result in hangs. This can happen if for " + "example the user adds a level 1 timer that is not " + "called by all ranks.", + dest="barrier_with_L1_time", + ) + group.add_argument( + "--timing-log-option", + type=str, + default="minmax", + choices=["max", "minmax", "all"], + help="Options for logging timing:" + " max: report the max timing across all ranks" + " minmax: report min and max timings across all ranks" + " all: report timings of all ranks.", + ) + group.add_argument( + "--tensorboard-log-interval", + type=int, + default=1, + help="Report to tensorboard interval.", + ) + group.add_argument( + "--tensorboard-queue-size", + type=int, + default=1000, + help="Size of the tensorboard queue for pending events " + "and summaries before one of the ‘add’ calls forces a " + "flush to disk.", + ) + group.add_argument( + "--log-timers-to-tensorboard", + action="store_true", + help="If set, write timers to tensorboard.", + ) + group.add_argument( + "--log-batch-size-to-tensorboard", + action="store_true", + help="If set, write batch-size to tensorboard.", + ) + group.add_argument( + "--no-log-learnig-rate-to-tensorboard", + action="store_false", + help="Disable learning rate logging to tensorboard.", + dest="log_learning_rate_to_tensorboard", + ) + group.add_argument( + "--no-log-loss-scale-to-tensorboard", + action="store_false", + help="Disable loss-scale logging to tensorboard.", + dest="log_loss_scale_to_tensorboard", + ) + group.add_argument( + "--log-validation-ppl-to-tensorboard", + action="store_true", + help="If set, write validation perplexity to " "tensorboard.", + ) + group.add_argument( + "--log-optimizer-states-to-tensorboard", + action="store_true", + help="If set, write various optimizer states to " + "tensorboard. This feature may consume extra GPU memory.", + ) + group.add_argument( + "--log-memory-to-tensorboard", + action="store_true", + help="Enable memory logging to tensorboard.", + ) + group.add_argument( + "--log-world-size-to-tensorboard", + action="store_true", + help="Enable world size logging to tensorboard.", + ) + group.add_argument( + "--wandb-project", + type=str, + default="", + help="The wandb project name. Ignore wandb by default.", + ) + group.add_argument( + "--wandb-exp-name", type=str, default="", help="The wandb experiment name." + ) + group.add_argument( + "--wandb-save-dir", + type=str, + default="", + help="Path to save the wandb results locally.", + ) -def _add_checkpointing_args(parser): - group = parser.add_argument_group(title='checkpointing') - - group.add_argument('--save', type=str, default=None, - help='Output directory to save checkpoints to.') - group.add_argument('--save-interval', type=int, default=None, - help='Number of iterations between checkpoint saves.') - group.add_argument('--no-save-optim', action='store_true', default=None, - help='Do not save current optimizer.') - group.add_argument('--no-save-rng', action='store_true', default=None, - help='Do not save current rng state.') - group.add_argument('--load', type=str, default=None, - help='Directory containing a model checkpoint.') - group.add_argument('--load-tag', type=str, default=None, - help='Specific checkpoint tag to load. Ignores latest.') - group.add_argument('--no-load-optim', action='store_true', default=None, - help='Do not load optimizer when loading checkpoint.') - group.add_argument('--no-load-rng', action='store_true', default=None, - help='Do not load rng state when loading checkpoint.') - group.add_argument('--no-load-lr-state', action='store_true', - help='Do not load lr state when loading checkpoint.') - group.add_argument('--finetune', action='store_true', - help='Load model for finetuning. Do not load optimizer ' - 'or rng state from checkpoint and set iteration to 0. ' - 'Assumed when loading a release checkpoint.') - group.add_argument('--no-initialization', action='store_false', - help='Do not perform initialization when building model, ' - 'can reduce startup time when definitely loading from a ' - 'checkpoint', - dest='perform_initialization') - group.add_argument('--use-checkpoint-args', action='store_true', - help='Override any command line arguments with arguments ' - 'from the checkpoint') - group.add_argument('--exit-on-missing-checkpoint', action='store_true', - help="If '--load' is set, but checkpoint is not found " - "(e.g., path typo), then exit instead of random " - "initialization.") - group.add_argument('--universal-checkpoint', action='store_true', - help='Loading a universal format checkpoint.') return parser -def _add_mixed_precision_args(parser): - group = parser.add_argument_group(title='mixed precision') - - group.add_argument('--fp16', action='store_true', - help='Run model in fp16 mode.') - group.add_argument('--bf16', action='store_true', - help='Run model in bfloat16 mode.') - group.add_argument('--loss-scale', type=float, default=None, - help='Static loss scaling, positive power of 2 ' - 'values can improve fp16 convergence. If None, dynamic' - 'loss scaling is used.') - group.add_argument('--initial-loss-scale', type=float, default=2**32, - help='Initial loss-scale for dynamic loss scaling.') - group.add_argument('--min-loss-scale', type=float, default=1.0, - help='Minimum loss scale for dynamic loss scale.') - group.add_argument('--loss-scale-window', type=float, default=1000, - help='Window over which to raise/lower dynamic scale.') - group.add_argument('--hysteresis', type=int, default=2, - help='hysteresis for dynamic loss scaling') - group.add_argument('--fp32-residual-connection', action='store_true', - help='Move residual connections to fp32.') - group.add_argument('--no-query-key-layer-scaling', action='store_false', - help='Do not scale Q * K^T by 1 / layer-number.', - dest='apply_query_key_layer_scaling') - group.add_argument('--attention-softmax-in-fp32', action='store_true', - help='Run attention masking and softmax in fp32. ' - 'This flag is ignored unless ' - '--no-query-key-layer-scaling is specified.') - group.add_argument('--accumulate-allreduce-grads-in-fp32', - action='store_true', - help='Gradient accumulation and all-reduce in fp32.') - group.add_argument('--fp16-lm-cross-entropy', action='store_true', - help='Move the cross entropy unreduced loss calculation' - 'for lm head to fp16.') - - return parser +def _add_regularization_args(parser): + group = parser.add_argument_group(title="regularization") + group.add_argument( + "--attention-dropout", + type=float, + default=0.1, + help="Post attention dropout probability.", + ) + group.add_argument( + "--hidden-dropout", + type=float, + default=0.1, + help="Dropout probability for hidden state transformer.", + ) + group.add_argument( + "--weight-decay", + type=float, + default=0.01, + help="Weight decay coefficient for L2 regularization.", + ) + group.add_argument( + "--start-weight-decay", + type=float, + help="Initial weight decay coefficient for L2 regularization.", + ) + group.add_argument( + "--end-weight-decay", + type=float, + help="End of run weight decay coefficient for L2 regularization.", + ) + group.add_argument( + "--weight-decay-incr-style", + type=str, + default="constant", + choices=["constant", "linear", "cosine"], + help="Weight decay increment function.", + ) + group.add_argument( + "--clip-grad", + type=float, + default=1.0, + help="Gradient clipping based on global L2 norm.", + ) + group.add_argument( + "--sophiag-beta1", + type=float, + default=0.9, + help="First coefficient for computing running averages " + "of gradient and its hessian", + ) + group.add_argument( + "--sophiag-beta2", + type=float, + default=0.95, + help="Second coefficient for computing running averages " + "of gradient and its hessian", + ) + group.add_argument( + "--sophiag-rho", type=float, default=0.01, help="SophiaG clipping threshhold" + ) -def _add_distributed_args(parser): - group = parser.add_argument_group(title='distributed') - - group.add_argument('--tensor-model-parallel-size', type=int, default=1, - help='Degree of tensor model parallelism.') - group.add_argument('--enable-expert-tensor-parallelism', action='store_true', - default=False, - help="use tensor parallelism for expert layers in MoE") - group.add_argument('--pipeline-model-parallel-size', type=int, default=1, - help='Degree of pipeline model parallelism.') - group.add_argument('--pipeline-model-parallel-split-rank', - type=int, default=None, - help='Rank where encoder and decoder should be split.') - group.add_argument('--moe-expert-parallel-size', type=int, default=1, - help='Degree of the MoE expert parallelism.') - group.add_argument('--model-parallel-size', type=int, default=None, - help='Old model parallel argument, do not use. Use ' - '--tensor-model-parallel-size instead.') - group.add_argument('--num-layers-per-virtual-pipeline-stage', type=int, default=None, - help='Number of layers per virtual pipeline stage') - group.add_argument('--overlap-p2p-communication', - action='store_true', - help='overlap pipeline parallel communication with forward and backward chunks', - dest='overlap_p2p_comm') - group.add_argument('--distributed-backend', default='nccl', - choices=['nccl', 'gloo', 'ccl', 'hccl'], - help='Which backend to use for distributed training.') - group.add_argument('--distributed-timeout-minutes', type=int, default=10, - help='Timeout minutes for torch.distributed.') - group.add_argument('--DDP-impl', default='local', - choices=['local', 'torch', 'FSDP'], - help='which DistributedDataParallel implementation ' - 'to use.') - group.add_argument('--no-contiguous-buffers-in-local-ddp', - action='store_false', help='If set, dont use ' - 'contiguous buffer in local DDP.', - dest='use_contiguous_buffers_in_local_ddp') - group.add_argument('--no-scatter-gather-tensors-in-pipeline', action='store_false', - help='Use scatter/gather to optimize communication of tensors in pipeline', - dest='scatter_gather_tensors_in_pipeline') - group.add_argument('--use-ring-exchange-p2p', action='store_true', - default=False, help='If set, use custom-built ring exchange ' - 'for p2p communications. Note that this option will require ' - 'a custom built image that support ring-exchange p2p.') - group.add_argument('--local-rank', '--local_rank', type=int, default=None, - help='local rank passed from distributed launcher.') - group.add_argument('--lazy-mpu-init', type=bool, required=False, - help='If set to True, initialize_megatron() ' - 'skips DDP initialization and returns function to ' - 'complete it instead.Also turns on ' - '--use-cpu-initialization flag. This is for ' - 'external DDP manager.' ) - group.add_argument('--use-cpu-initialization', action='store_true', - default=None, help='If set, affine parallel weights ' - 'initialization uses CPU' ) - group.add_argument('--empty-unused-memory-level', default=0, type=int, - choices=[0, 1, 2], - help='Call torch.cuda.empty_cache() each iteration ' - '(training and eval), to reduce fragmentation.' - '0=off, 1=moderate, 2=aggressive.') - group.add_argument('--standalone-embedding-stage', action='store_true', - default=False, help='If set, *input* embedding layer ' - 'is placed on its own pipeline stage, without any ' - 'transformer layers. (For T5, this flag currently only ' - 'affects the encoder embedding.)') - group.add_argument('--use-distributed-optimizer', action='store_true', - help='Use distributed optimizer.') + group.add_argument( + "--adam-beta1", + type=float, + default=0.9, + help="First coefficient for computing running averages " + "of gradient and its square", + ) + group.add_argument( + "--adam-beta2", + type=float, + default=0.999, + help="Second coefficient for computing running averages " + "of gradient and its square", + ) + group.add_argument( + "--adam-eps", + type=float, + default=1e-08, + help="Term added to the denominator to improve" "numerical stability", + ) + group.add_argument( + "--sgd-momentum", type=float, default=0.9, help="Momentum factor for sgd" + ) return parser -def _add_validation_args(parser): - group = parser.add_argument_group(title='validation') - - group.add_argument('--eval-iters', type=int, default=100, - help='Number of iterations to run for evaluation' - 'validation/test for.') - group.add_argument('--eval-interval', type=int, default=1000, - help='Interval between running evaluation on ' - 'validation set.') - group.add_argument('--skip-train', action='store_true', - default=False, help='If set, bypass the training loop, ' - 'optionally do evaluation for validation/test, and exit.') +def _add_training_args(parser): + group = parser.add_argument_group(title="training") - return parser + group.add_argument( + "--micro-batch-size", + type=int, + default=None, + help="Batch size per model instance (local batch size). " + "Global batch size is local batch size times data " + "parallel size times number of micro batches.", + ) + group.add_argument( + "--batch-size", + type=int, + default=None, + help="Old batch size parameter, do not use. " "Use --micro-batch-size instead", + ) + group.add_argument( + "--global-batch-size", + type=int, + default=None, + help="Training batch size. If set, it should be a " + "multiple of micro-batch-size times data-parallel-size. " + "If this value is None, then " + "use micro-batch-size * data-parallel-size as the " + "global batch size. This choice will result in 1 for " + "number of micro-batches.", + ) + group.add_argument( + "--rampup-batch-size", + nargs="*", + default=None, + help="Batch size ramp up with the following values:" + " --rampup-batch-size " + " " + " " + "For example:" + " --rampup-batch-size 16 8 300000 \\ " + " --global-batch-size 1024" + "will start with global batch size 16 and over " + " (1024 - 16) / 8 = 126 intervals will increase" + "the batch size linearly to 1024. In each interval" + "we will use approximately 300000 / 126 = 2380 samples.", + ) + group.add_argument( + "--recompute-activations", + action="store_true", + help="recompute activation to allow for training " + "with larger models, sequences, and batch sizes.", + ) + group.add_argument( + "--recompute-granularity", + type=str, + default=None, + choices=["full", "selective"], + help="Checkpoint activations to allow for training " + "with larger models, sequences, and batch sizes. " + "It is supported at two granularities 1) full: " + "whole transformer layer is recomputed, " + "2) selective: core attention part of the transformer " + "layer is recomputed.", + ) + group.add_argument( + "--distribute-saved-activations", + action="store_true", + help="If set, distribute recomputed activations " + "across model parallel group.", + ) + group.add_argument( + "--recompute-method", + type=str, + default=None, + choices=["uniform", "block"], + help="1) uniform: uniformly divide the total number of " + "Transformer layers and recompute the input activation of " + "each divided chunk at specified granularity, " + "2) recompute the input activations of only a set number of " + "individual Transformer layers per pipeline stage and do the " + "rest without any recomputing at specified granularity" + "default) do not apply activations recompute to any layers", + ) + group.add_argument( + "--recompute-num-layers", + type=int, + default=1, + help="1) uniform: the number of Transformer layers in each " + "uniformly divided recompute unit, " + "2) block: the number of individual Transformer layers " + "to recompute within each pipeline stage.", + ) + group.add_argument( + "--enable-zbh1-pipeline", + action="store_true", + help="Activate zero bubble pipeline parallelism schedule method", + ) + group.add_argument( + "--enable-zbh1-exact-semantics", + action="store_true", + help="Use an exact semantics for zbh1 schedule, might be slower than the default.", + ) + + # deprecated + # HACK: added back arguments because DeepSpeed still relies on the old + # activation checkpointing mechanism. + group.add_argument( + "--checkpoint-activations", + action="store_true", + help="Checkpoint activation to allow for training " + "with larger models, sequences, and batch sizes.", + ) + group.add_argument( + "--distribute-checkpointed-activations", + action="store_true", + help="If set, distribute checkpointed activations " + "across model parallel group.", + ) + group.add_argument( + "--checkpoint-num-layers", + type=int, + default=1, + help="chunk size (number of layers) for checkpointing.", + ) + group.add_argument( + "--train-iters", + type=int, + default=None, + help="Total number of iterations to train over all " + "training runs. Note that either train-iters or " + "train-samples should be provided.", + ) + group.add_argument( + "--train-samples", + type=int, + default=None, + help="Total number of samples to train over all " + "training runs. Note that either train-iters or " + "train-samples should be provided.", + ) + group.add_argument( + "--train-tokens", + type=int, + default=None, + help="Total number of tokens to train over all " "training runs.", + ) + group.add_argument( + "--random-ltd", action="store_true", help="enable random layer token drop" + ) + group.add_argument( + "--log-interval", type=int, default=100, help="Report loss and timing interval." + ) + group.add_argument( + "--exit-interval", + type=int, + default=None, + help="Exit the program after the iteration is divisible " "by this value.", + ) + group.add_argument( + "--exit-duration-in-mins", + type=int, + default=None, + help="Exit the program after this many minutes.", + ) + group.add_argument( + "--exit-signal-handler", + action="store_true", + help="Dynamically save the checkpoint and shutdown the " + "training if SIGTERM is received", + ) + group.add_argument( + "--tensorboard-dir", + type=str, + default=None, + help="Write TensorBoard logs to this directory.", + ) + group.add_argument( + "--trace-dir", + type=str, + default="./trace/", + help="Write trace logs to this directory.", + ) + group.add_argument( + "--no-masked-softmax-fusion", + action="store_false", + help="Disable fusion of query_key_value scaling, " "masking, and softmax.", + dest="masked_softmax_fusion", + ) + group.add_argument( + "--no-bias-gelu-fusion", + action="store_false", + help="Disable bias and gelu fusion.", + dest="bias_gelu_fusion", + ) + group.add_argument( + "--no-bias-dropout-fusion", + action="store_false", + help="Disable bias and dropout fusion.", + dest="bias_dropout_fusion", + ) + group.add_argument( + "--disable-moe-token-dropping", + action="store_false", + help="Disable MoE expert token dropping.", + dest="moe_token_dropping", + ) + group.add_argument( + "--moe-train-capacity-factor", + type=float, + default=1.0, + help="The capacity of the MoE expert at training time", + ) + group.add_argument( + "--moe-eval-capacity-factor", + type=float, + default=1.0, + help="The capacity of the MoE expert at eval time.", + ) + group.add_argument( + "--moe-min-capacity", + type=int, + default=4, + help="The minimum capacity per MoE expert regardless of the capacity_factor.", + ) + group.add_argument( + "--moe-loss-coeff", + type=float, + default=0.1, + help="Scaling coefficient for adding MoE loss to model loss", + ) + group.add_argument( + "--create-moe-param-group", + action="store_true", + help="Create separate groups for MoE params." + "This is necessary for techniques like ZeRO.", + ) + group.add_argument( + "--disable-moe-top2-2nd-expert-sampling", + action="store_false", + help="Disable MoE top2 sampling of the 2nd expert. Instead of sampling, use argmax.", + dest="moe_top2_2nd_expert_sampling", + ) + group.add_argument( + "--use-flash-attn", + "--use-flash-attn-v1", + dest="use_flash_attn_v1", + action="store_true", + help="use first version FlashAttention implementation of attention. " + "https://arxiv.org/abs/2205.14135", + ) + group.add_argument( + "--use-flash-attn-v2", + action="store_true", + help="use second version FlashAttention implementation of attention. " + "https://arxiv.org/abs/2307.08691", + ) + group.add_argument( + "--use-flash-attn-triton", + action="store_true", + help="use FlashAttention implementation of attention using Triton.", + ) + group.add_argument( + "--use-flash-attn-builder", + action="store_true", + help="use FlashAttention op builder.", + ) + group.add_argument( + "--disable-bias-linear", + action="store_false", + help="Disable bias in the linear layers", + dest="add_bias_linear", + ) + group.add_argument( + "--optimizer", + type=str, + default="adam", + choices=[ + "adam", + "adam8bit", + "adamw", + "adamwschedulefree", + "apex.adam", + "apex.sgd", + "ds.fusedlamb", + "ds.onebitlamb", + "galoreadamw", + "galoreadamw8bit", + "galoreadamw8bitperlayer", + "ipex.fusedlamb", + "ipex.lamb", + "shampoo", + "sgd", + "sgdschedulefree", + "sophiag", + ], + help="Optimizer function", + ) + group.add_argument( + "--schedulefree-for-each", + action="store_true", + help=""" + Use a foreach-backed implementation of the schedulefree optimizers. + Should be significantly faster, + but will have a higher peak memory usage. + """, + ) + group.add_argument( + "--dataloader-type", + type=str, + default=None, + choices=["single", "cyclic"], + help="Single pass vs multiple pass data loader", + ) + group.add_argument( + "--ds-inference", + action="store_true", + help="DeepSpeed inference engine being used", + ) + group.add_argument( + "--cpu-optimizer", action="store_true", help="Run optimizer on CPU" + ) + group.add_argument( + "--cpu_torch_adam", + action="store_true", + help="Use Torch Adam as optimizer on CPU.", + ) + group.add_argument( + "--ds_fused_adam", + action="store_true", + help="Use DeepSpeed FusedAdam as optimizer.", + ) + group.add_argument( + "--no-pipeline-parallel", + action="store_true", + help="Disable pipeline parallelism", + ) + group.add_argument( + "--use-tutel", action="store_true", help="Use Tutel optimization for MoE" + ) + group.add_argument( + "--inference", + action="store_true", + help="Very basic inference mode: not allocating optim/lr - requires ZERO_STAGE=0", + ) + + group.add_argument( + "--no-async-tensor-model-parallel-allreduce", + action="store_false", + help="Disable asynchronous execution of " + "tensor-model-parallel all-reduce with weight " + "gradient compuation of a column-linear layer.", + dest="async_tensor_model_parallel_allreduce", + ) + group.add_argument( + "--no-persist-layer-norm", + action="store_true", + help="Disable using persistent fused layer norm kernel. " + "This kernel supports only a set of hidden sizes. Please " + "check persist_ln_hidden_sizes if your hidden " + "size is supported.", + ) + group.add_argument( + "--sequence-parallel", + action="store_true", + help="Enable Megatron-LM's sequence parallel optimization.", + ) + group.add_argument( + "--ds-sequence-parallel-size", + type=int, + default=1, + help="Enable DeepSpeed's sequence parallel. Cannot be combined with \"--sequence-parallel\", which enables Megatron-LM's sequence parallel.", + ) + group.add_argument( + "--force-ds-sequence-parallel", + action="store_true", + help="use DeepSpeed sequence parallelism regardless of sequence parallel size.", + ) + group.add_argument( + "--no-gradient-accumulation-fusion", + action="store_false", + help="Disable fusing gradient accumulation to weight " + "gradient computation of linear layers", + dest="gradient_accumulation_fusion", + ) + group.add_argument( + "--use-dataset-only", + type=bool, + required=False, + default=False, + help="If set to True, only use the megatron dataset for external trainer ", + ) + # group.add_argument('--profile', action='store_true', help='Enable Torch Profiler') + group.add_argument( + "--train-range-to-skip", + action="extend", + nargs="+", + type=int, + help=("Range of iters to skip during training. Must be in pairs."), + ) + group.add_argument( + "--train-iters-to-skip", + action="extend", + nargs="+", + type=str, + help=( + "Specific train iterations to skip when training. " + "Load the data and just perform a noop." + ), + ) + return parser + + +def _add_initialization_args(parser): + group = parser.add_argument_group(title="initialization") + + group.add_argument( + "--seed", + type=int, + default=1234, + help="Random seed used for python, numpy, " "pytorch, and cuda.", + ) + group.add_argument( + "--data-parallel-random-init", + action="store_true", + help="Enable random initialization of params " "across data parallel ranks", + ) + group.add_argument( + "--init-method-std", + type=float, + default=0.02, + help="Standard deviation of the zero mean normal " + "distribution used for weight initialization.", + ) + group.add_argument( + "--init-method-xavier-uniform", + action="store_true", + help="Enable Xavier uniform parameter initialization", + ) + + return parser + + +def _add_learning_rate_args(parser): + group = parser.add_argument_group(title="learning rate") + + group.add_argument( + "--lr", + type=float, + default=None, + help="Initial learning rate. Depending on decay style " + "and initial warmup, the learing rate at each " + "iteration would be different.", + ) + group.add_argument( + "--lr-decay-style", + type=str, + default="linear", + choices=["constant", "linear", "cosine", "inverse-square-root"], + help="Learning rate decay function.", + ) + group.add_argument( + "--lr-decay-iters", + type=int, + default=None, + help="number of iterations to decay learning rate over," + " If None defaults to `--train-iters`", + ) + group.add_argument( + "--lr-decay-samples", + type=int, + default=None, + help="number of samples to decay learning rate over," + " If None defaults to `--train-samples`", + ) + group.add_argument( + "--lr-decay-tokens", + type=int, + default=None, + help="number of tokens to decay learning rate over," + " If not None will override iter/sample-based decay", + ) + group.add_argument( + "--lr-warmup-fraction", + type=float, + default=None, + help="fraction of lr-warmup-(iters/samples) to use " "for warmup (as a float)", + ) + group.add_argument( + "--lr-warmup-iters", + type=int, + default=0, + help="number of iterations to linearly warmup " "learning rate over.", + ) + group.add_argument( + "--lr-warmup-samples", + type=int, + default=0, + help="number of samples to linearly warmup " "learning rate over.", + ) + group.add_argument( + "--lr-warmup-tokens", + type=int, + default=None, + help="number of tokens to linearly warmup " "learning rate over.", + ) + group.add_argument( + "--warmup", + type=int, + default=None, + help="Old lr warmup argument, do not use. Use one of the" + "--lr-warmup-* arguments above", + ) + group.add_argument( + "--min-lr", + type=float, + default=0.0, + help="Minumum value for learning rate. The scheduler" + "clip values below this threshold.", + ) + group.add_argument( + "--override-opt_param-scheduler", + action="store_true", + help="Reset the values of the scheduler (learning rate," + "warmup iterations, minimum learning rate, maximum " + "number of iterations, and decay style from input " + "arguments and ignore values from checkpoints. Note" + "that all the above values will be reset.", + ) + group.add_argument( + "--use-checkpoint-opt_param-scheduler", + action="store_true", + help="Use checkpoint to set the values of the scheduler " + "(learning rate, warmup iterations, minimum learning " + "rate, maximum number of iterations, and decay style " + "from checkpoint and ignore input arguments.", + ) + + return parser + + +def _add_checkpointing_args(parser): + group = parser.add_argument_group(title="checkpointing") + + group.add_argument( + "--save", + type=str, + default=None, + help="Output directory to save checkpoints to.", + ) + group.add_argument( + "--save-interval", + type=int, + default=None, + help="Number of iterations between checkpoint saves.", + ) + group.add_argument( + "--no-save-optim", + action="store_true", + default=None, + help="Do not save current optimizer.", + ) + group.add_argument( + "--no-save-rng", + action="store_true", + default=None, + help="Do not save current rng state.", + ) + group.add_argument( + "--load", + type=str, + default=None, + help="Directory containing a model checkpoint.", + ) + group.add_argument( + "--load-tag", + type=str, + default=None, + help="Specific checkpoint tag to load. Ignores latest.", + ) + group.add_argument( + "--no-load-optim", + action="store_true", + default=None, + help="Do not load optimizer when loading checkpoint.", + ) + group.add_argument( + "--no-load-rng", + action="store_true", + default=None, + help="Do not load rng state when loading checkpoint.", + ) + group.add_argument( + "--no-load-lr-state", + action="store_true", + help="Do not load lr state when loading checkpoint.", + ) + group.add_argument( + "--finetune", + action="store_true", + help="Load model for finetuning. Do not load optimizer " + "or rng state from checkpoint and set iteration to 0. " + "Assumed when loading a release checkpoint.", + ) + group.add_argument( + "--no-initialization", + action="store_false", + help="Do not perform initialization when building model, " + "can reduce startup time when definitely loading from a " + "checkpoint", + dest="perform_initialization", + ) + group.add_argument( + "--use-checkpoint-args", + action="store_true", + help="Override any command line arguments with arguments " + "from the checkpoint", + ) + group.add_argument( + "--exit-on-missing-checkpoint", + action="store_true", + help="If '--load' is set, but checkpoint is not found " + "(e.g., path typo), then exit instead of random " + "initialization.", + ) + group.add_argument( + "--universal-checkpoint", + action="store_true", + help="Loading a universal format checkpoint.", + ) + return parser + + +def _add_mixed_precision_args(parser): + group = parser.add_argument_group(title="mixed precision") + group.add_argument("--fp16", action="store_true", help="Run model in fp16 mode.") + group.add_argument( + "--bf16", action="store_true", help="Run model in bfloat16 mode." + ) + group.add_argument( + "--loss-scale", + type=float, + default=None, + help="Static loss scaling, positive power of 2 " + "values can improve fp16 convergence. If None, dynamic" + "loss scaling is used.", + ) + group.add_argument( + "--initial-loss-scale", + type=float, + default=2**32, + help="Initial loss-scale for dynamic loss scaling.", + ) + group.add_argument( + "--min-loss-scale", + type=float, + default=1.0, + help="Minimum loss scale for dynamic loss scale.", + ) + group.add_argument( + "--loss-scale-window", + type=float, + default=1000, + help="Window over which to raise/lower dynamic scale.", + ) + group.add_argument( + "--hysteresis", type=int, default=2, help="hysteresis for dynamic loss scaling" + ) + group.add_argument( + "--fp32-residual-connection", + action="store_true", + help="Move residual connections to fp32.", + ) + group.add_argument( + "--no-query-key-layer-scaling", + action="store_false", + help="Do not scale Q * K^T by 1 / layer-number.", + dest="apply_query_key_layer_scaling", + ) + group.add_argument( + "--attention-softmax-in-fp32", + action="store_true", + help="Run attention masking and softmax in fp32. " + "This flag is ignored unless " + "--no-query-key-layer-scaling is specified.", + ) + group.add_argument( + "--accumulate-allreduce-grads-in-fp32", + action="store_true", + help="Gradient accumulation and all-reduce in fp32.", + ) + group.add_argument( + "--fp16-lm-cross-entropy", + action="store_true", + help="Move the cross entropy unreduced loss calculation" "for lm head to fp16.", + ) + + return parser + + +def _add_distributed_args(parser): + group = parser.add_argument_group(title="distributed") + + group.add_argument( + "--tensor-model-parallel-size", + type=int, + default=1, + help="Degree of tensor model parallelism.", + ) + group.add_argument( + "--enable-expert-tensor-parallelism", + action="store_true", + default=False, + help="use tensor parallelism for expert layers in MoE", + ) + group.add_argument( + "--pipeline-model-parallel-size", + type=int, + default=1, + help="Degree of pipeline model parallelism.", + ) + group.add_argument( + "--pipeline-model-parallel-split-rank", + type=int, + default=None, + help="Rank where encoder and decoder should be split.", + ) + group.add_argument( + "--moe-expert-parallel-size", + type=int, + default=1, + help="Degree of the MoE expert parallelism.", + ) + group.add_argument( + "--model-parallel-size", + type=int, + default=None, + help="Old model parallel argument, do not use. Use " + "--tensor-model-parallel-size instead.", + ) + group.add_argument( + "--num-layers-per-virtual-pipeline-stage", + type=int, + default=None, + help="Number of layers per virtual pipeline stage", + ) + group.add_argument( + "--overlap-p2p-communication", + action="store_true", + help="overlap pipeline parallel communication with forward and backward chunks", + dest="overlap_p2p_comm", + ) + group.add_argument( + "--distributed-backend", + default="nccl", + choices=["nccl", "gloo", "ccl", "hccl"], + help="Which backend to use for distributed training.", + ) + group.add_argument( + "--distributed-timeout-minutes", + type=int, + default=10, + help="Timeout minutes for torch.distributed.", + ) + group.add_argument( + "--DDP-impl", + default="local", + choices=["local", "torch", "FSDP"], + help="which DistributedDataParallel implementation " "to use.", + ) + group.add_argument( + "--no-contiguous-buffers-in-local-ddp", + action="store_false", + help="If set, dont use " "contiguous buffer in local DDP.", + dest="use_contiguous_buffers_in_local_ddp", + ) + group.add_argument( + "--no-scatter-gather-tensors-in-pipeline", + action="store_false", + help="Use scatter/gather to optimize communication of tensors in pipeline", + dest="scatter_gather_tensors_in_pipeline", + ) + group.add_argument( + "--use-ring-exchange-p2p", + action="store_true", + default=False, + help="If set, use custom-built ring exchange " + "for p2p communications. Note that this option will require " + "a custom built image that support ring-exchange p2p.", + ) + group.add_argument( + "--local-rank", + "--local_rank", + type=int, + default=None, + help="local rank passed from distributed launcher.", + ) + group.add_argument( + "--lazy-mpu-init", + type=bool, + required=False, + help="If set to True, initialize_megatron() " + "skips DDP initialization and returns function to " + "complete it instead.Also turns on " + "--use-cpu-initialization flag. This is for " + "external DDP manager.", + ) + group.add_argument( + "--use-cpu-initialization", + action="store_true", + default=None, + help="If set, affine parallel weights " "initialization uses CPU", + ) + group.add_argument( + "--empty-unused-memory-level", + default=0, + type=int, + choices=[0, 1, 2], + help="Call torch.cuda.empty_cache() each iteration " + "(training and eval), to reduce fragmentation." + "0=off, 1=moderate, 2=aggressive.", + ) + group.add_argument( + "--standalone-embedding-stage", + action="store_true", + default=False, + help="If set, *input* embedding layer " + "is placed on its own pipeline stage, without any " + "transformer layers. (For T5, this flag currently only " + "affects the encoder embedding.)", + ) + group.add_argument( + "--use-distributed-optimizer", + action="store_true", + help="Use distributed optimizer.", + ) + + return parser + + +def _add_validation_args(parser): + group = parser.add_argument_group(title="validation") + + group.add_argument( + "--eval-iters", + type=int, + default=100, + help="Number of iterations to run for evaluation" "validation/test for.", + ) + group.add_argument( + "--eval-interval", + type=int, + default=1000, + help="Interval between running evaluation on " "validation set.", + ) + group.add_argument( + "--skip-train", + action="store_true", + default=False, + help="If set, bypass the training loop, " + "optionally do evaluation for validation/test, and exit.", + ) + + return parser + + +def _add_data_args(parser): + group = parser.add_argument_group(title="data and dataloader") + + group.add_argument( + "--aml-data-download-path", + type=str, + default=None, + help="Path to mounted input dataset", + ) + group.add_argument( + "--data-path", + nargs="*", + default=None, + help="Path to the training dataset. Accepted format:" + "1) a single data path, 2) multiple datasets in the" + "form: dataset1-weight dataset1-path dataset2-weight " + "dataset2-path ... It is used with --split when a " + "single dataset used for all three: train, valid " + "and test. It is exclusive to the other " + "--*-data-path args", + ) + group.add_argument( + "--data-file-list", + type=str, + default=None, + help="The file with the list of dataset and weights", + ) + + group.add_argument( + "--shuffle-sample-in-corpus", + action="store_true", + help="Whether to shuffle the samples within in the dataset files", + ) + + group.add_argument( + "--blend-sample-in-corpus", + action="store_true", + help="Whether to blend different files in the same corpus", + ) + + group.add_argument( + "--split", + type=str, + default="969, 30, 1", + help="Comma-separated list of proportions for training," + " validation, and test split. For example the split " + "`90,5,5` will use 90%% of data for training, 5%% for " + "validation and 5%% for test.", + ) + group.add_argument( + "--train-data-path", + nargs="*", + default=None, + help="Path to the training dataset. Accepted format:" + "1) a single data path, 2) multiple datasets in the" + "form: dataset1-weight dataset1-path dataset2-weight " + "dataset2-path ...", + ) + group.add_argument( + "--valid-data-path", + nargs="*", + default=None, + help="Path to the validation dataset. Accepted format:" + "1) a single data path, 2) multiple datasets in the" + "form: dataset1-weight dataset1-path dataset2-weight " + "dataset2-path ...", + ) + group.add_argument( + "--test-data-path", + nargs="*", + default=None, + help="Path to the test dataset. Accepted format:" + "1) a single data path, 2) multiple datasets in the" + "form: dataset1-weight dataset1-path dataset2-weight " + "dataset2-path ...", + ) + group.add_argument( + "--data-cache-path", + default=None, + help="Path to a directory to hold cached index files.", + ) -def _add_data_args(parser): - group = parser.add_argument_group(title='data and dataloader') - - group.add_argument('--aml-data-download-path', type=str, default=None, - help='Path to mounted input dataset') - group.add_argument('--data-path', nargs='*', default=None, - help='Path to the training dataset. Accepted format:' - '1) a single data path, 2) multiple datasets in the' - 'form: dataset1-weight dataset1-path dataset2-weight ' - 'dataset2-path ... It is used with --split when a ' - 'single dataset used for all three: train, valid ' - 'and test. It is exclusive to the other ' - '--*-data-path args') - group.add_argument('--data-file-list', type=str, default=None, - help='The file with the list of dataset and weights') - - group.add_argument('--shuffle-sample-in-corpus', action='store_true', help="Whether to shuffle the samples within in the dataset files") - - group.add_argument('--blend-sample-in-corpus', action='store_true', help="Whether to blend different files in the same corpus") - - group.add_argument('--split', type=str, default='969, 30, 1', - help='Comma-separated list of proportions for training,' - ' validation, and test split. For example the split ' - '`90,5,5` will use 90%% of data for training, 5%% for ' - 'validation and 5%% for test.') - group.add_argument('--train-data-path', nargs='*', default=None, - help='Path to the training dataset. Accepted format:' - '1) a single data path, 2) multiple datasets in the' - 'form: dataset1-weight dataset1-path dataset2-weight ' - 'dataset2-path ...') - group.add_argument('--valid-data-path', nargs='*', default=None, - help='Path to the validation dataset. Accepted format:' - '1) a single data path, 2) multiple datasets in the' - 'form: dataset1-weight dataset1-path dataset2-weight ' - 'dataset2-path ...') - group.add_argument('--test-data-path', nargs='*', default=None, - help='Path to the test dataset. Accepted format:' - '1) a single data path, 2) multiple datasets in the' - 'form: dataset1-weight dataset1-path dataset2-weight ' - 'dataset2-path ...') - group.add_argument('--data-cache-path', default=None, - help='Path to a directory to hold cached index files.') - - group.add_argument('--vocab-size', type=int, default=None, - help='Size of vocab before EOD or padding.') - group.add_argument('--vocab-file', type=str, default=None, - help='Path to the vocab file.') - group.add_argument('--merge-file', type=str, default=None, - help='Path to the BPE merge file.') - group.add_argument('--vocab-extra-ids', type=int, default=0, - help='Number of additional vocabulary tokens. ' - 'They are used for span masking in the T5 model') - group.add_argument('--seq-length', type=int, default=None, - help='Maximum sequence length to process.') - group.add_argument('--encoder-seq-length', type=int, default=None, - help='Maximum encoder sequence length to process.' - 'This should be exclusive of --seq-length') - group.add_argument('--decoder-seq-length', type=int, default=None, - help="Maximum decoder sequence length to process.") - group.add_argument('--retriever-seq-length', type=int, default=256, - help='Maximum sequence length for the biencoder model ' - 'for retriever') - group.add_argument('--sample-rate', type=float, default=1.0, - help='sample rate for training data. Supposed to be 0 ' - ' < sample_rate < 1') - group.add_argument('--mask-prob', type=float, default=0.15, - help='Probability of replacing a token with mask.') - group.add_argument('--short-seq-prob', type=float, default=0.1, - help='Probability of producing a short sequence.') - group.add_argument('--mmap-warmup', action='store_true', - help='Warm up mmap files.') - group.add_argument('--num-workers', type=int, default=2, - help="Dataloader number of workers.") - group.add_argument('--tokenizer-type', type=str, - default=None, - choices=['BertWordPieceLowerCase', - 'BertWordPieceCase', - 'GPT2BPETokenizer', - 'SentencePieceTokenizer', - 'GPTSentencePieceTokenizer', - 'HFTokenizer', - 'NullTokenizer', - 'Llama2Tokenizer'], - help='What type of tokenizer to use.') - group.add_argument('--tokenizer-model', type=str, default=None, - help='Sentencepiece tokenizer model.') - group.add_argument('--trust-remote-code', action='store_true', default=False, - help='To run HFTokenizer model from local path.') - group.add_argument('--data-impl', type=str, default='infer', - choices=['mmap', 'infer'], - help='Implementation of indexed datasets.') - group.add_argument('--reset-position-ids', action='store_true', - help='Reset posistion ids after end-of-document token.') - group.add_argument('--reset-attention-mask', action='store_true', - help='Reset self attention maske after ' - 'end-of-document token.') - group.add_argument('--eod-mask-loss', action='store_true', - help='Mask loss for the end of document tokens.') - group.add_argument('--train-data-exact-num-epochs', type=int, default=None, - help='When building the train dataset, force it to be ' - 'an exact number of epochs of the raw data') - group.add_argument('--return-data-index', action='store_true', - help='Return the index of data sample.') - group.add_argument('--data-efficiency-curriculum-learning', action='store_true', - help='Use DeepSpeed data efficiency library curriculum learning feature.') - group.add_argument('--train-idx-path', type=str, default=None, - help='Force to use certain index file.') - group.add_argument('--train-desc-path', type=str, default=None, - help='Force to use certain index file.') - group.add_argument('--train-doc-idx-path', type=str, default=None, - help='Force to use certain index file.') - group.add_argument('--train-sample-idx-path', type=str, default=None, - help='Force to use certain index file.') - group.add_argument('--train-shuffle-idx-path', type=str, default=None, - help='Force to use certain index file.') - group.add_argument('--repeated-dataloader', action='store_true', - help='Once all the data has been loaded, reuse the DataLoader.') - group.add_argument('--multiprocessing-context', type=str, default='fork') + group.add_argument( + "--vocab-size", + type=int, + default=None, + help="Size of vocab before EOD or padding.", + ) + group.add_argument( + "--vocab-file", type=str, default=None, help="Path to the vocab file." + ) + group.add_argument( + "--merge-file", type=str, default=None, help="Path to the BPE merge file." + ) + group.add_argument( + "--vocab-extra-ids", + type=int, + default=0, + help="Number of additional vocabulary tokens. " + "They are used for span masking in the T5 model", + ) + group.add_argument( + "--seq-length", + type=int, + default=None, + help="Maximum sequence length to process.", + ) + group.add_argument( + "--encoder-seq-length", + type=int, + default=None, + help="Maximum encoder sequence length to process." + "This should be exclusive of --seq-length", + ) + group.add_argument( + "--decoder-seq-length", + type=int, + default=None, + help="Maximum decoder sequence length to process.", + ) + group.add_argument( + "--retriever-seq-length", + type=int, + default=256, + help="Maximum sequence length for the biencoder model " "for retriever", + ) + group.add_argument( + "--sample-rate", + type=float, + default=1.0, + help="sample rate for training data. Supposed to be 0 " " < sample_rate < 1", + ) + group.add_argument( + "--mask-prob", + type=float, + default=0.15, + help="Probability of replacing a token with mask.", + ) + group.add_argument( + "--short-seq-prob", + type=float, + default=0.1, + help="Probability of producing a short sequence.", + ) + group.add_argument("--mmap-warmup", action="store_true", help="Warm up mmap files.") + group.add_argument( + "--num-workers", type=int, default=2, help="Dataloader number of workers." + ) + group.add_argument( + "--tokenizer-type", + type=str, + default=None, + choices=[ + "BertWordPieceLowerCase", + "BertWordPieceCase", + "GPT2BPETokenizer", + "SentencePieceTokenizer", + "GPTSentencePieceTokenizer", + "HFTokenizer", + "NullTokenizer", + "Llama2Tokenizer", + ], + help="What type of tokenizer to use.", + ) + group.add_argument( + "--tokenizer-model", + type=str, + default=None, + help="Sentencepiece tokenizer model.", + ) + group.add_argument( + "--trust-remote-code", + action="store_true", + default=False, + help="To run HFTokenizer model from local path.", + ) + group.add_argument( + "--data-impl", + type=str, + default="infer", + choices=["mmap", "infer"], + help="Implementation of indexed datasets.", + ) + group.add_argument( + "--reset-position-ids", + action="store_true", + help="Reset posistion ids after end-of-document token.", + ) + group.add_argument( + "--reset-attention-mask", + action="store_true", + help="Reset self attention maske after " "end-of-document token.", + ) + group.add_argument( + "--eod-mask-loss", + action="store_true", + help="Mask loss for the end of document tokens.", + ) + group.add_argument( + "--train-data-exact-num-epochs", + type=int, + default=None, + help="When building the train dataset, force it to be " + "an exact number of epochs of the raw data", + ) + group.add_argument( + "--return-data-index", + action="store_true", + help="Return the index of data sample.", + ) + group.add_argument( + "--data-efficiency-curriculum-learning", + action="store_true", + help="Use DeepSpeed data efficiency library curriculum learning feature.", + ) + group.add_argument( + "--train-idx-path", + type=str, + default=None, + help="Force to use certain index file.", + ) + group.add_argument( + "--train-desc-path", + type=str, + default=None, + help="Force to use certain index file.", + ) + group.add_argument( + "--train-doc-idx-path", + type=str, + default=None, + help="Force to use certain index file.", + ) + group.add_argument( + "--train-sample-idx-path", + type=str, + default=None, + help="Force to use certain index file.", + ) + group.add_argument( + "--train-shuffle-idx-path", + type=str, + default=None, + help="Force to use certain index file.", + ) + group.add_argument( + "--repeated-dataloader", + action="store_true", + help="Once all the data has been loaded, reuse the DataLoader.", + ) + group.add_argument("--multiprocessing-context", type=str, default="fork") return parser def _add_autoresume_args(parser): - group = parser.add_argument_group(title='autoresume') + group = parser.add_argument_group(title="autoresume") - group.add_argument('--adlr-autoresume', action='store_true', - help='Enable autoresume on adlr cluster.') - group.add_argument('--adlr-autoresume-interval', type=int, default=1000, - help='Intervals over which check for autoresume' - 'termination signal') + group.add_argument( + "--adlr-autoresume", + action="store_true", + help="Enable autoresume on adlr cluster.", + ) + group.add_argument( + "--adlr-autoresume-interval", + type=int, + default=1000, + help="Intervals over which check for autoresume" "termination signal", + ) return parser def _add_biencoder_args(parser): - group = parser.add_argument_group(title='biencoder') + group = parser.add_argument_group(title="biencoder") # network size - group.add_argument('--ict-head-size', type=int, default=None, - help='Size of block embeddings to be used in ICT and ' - 'REALM (paper default: 128)') - group.add_argument('--biencoder-projection-dim', type=int, default=0, - help='Size of projection head used in biencoder (paper' - ' default: 128)') - group.add_argument('--biencoder-shared-query-context-model', action='store_true', - help='Whether to share the parameters of the query ' - 'and context models or not') + group.add_argument( + "--ict-head-size", + type=int, + default=None, + help="Size of block embeddings to be used in ICT and " + "REALM (paper default: 128)", + ) + group.add_argument( + "--biencoder-projection-dim", + type=int, + default=0, + help="Size of projection head used in biencoder (paper" " default: 128)", + ) + group.add_argument( + "--biencoder-shared-query-context-model", + action="store_true", + help="Whether to share the parameters of the query " + "and context models or not", + ) # checkpointing - group.add_argument('--ict-load', type=str, default=None, - help='Directory containing an ICTBertModel checkpoint') - group.add_argument('--bert-load', type=str, default=None, - help='Directory containing an BertModel checkpoint ' - '(needed to start ICT and REALM)') + group.add_argument( + "--ict-load", + type=str, + default=None, + help="Directory containing an ICTBertModel checkpoint", + ) + group.add_argument( + "--bert-load", + type=str, + default=None, + help="Directory containing an BertModel checkpoint " + "(needed to start ICT and REALM)", + ) # data - group.add_argument('--titles-data-path', type=str, default=None, - help='Path to titles dataset used for ICT') - group.add_argument('--query-in-block-prob', type=float, default=0.1, - help='Probability of keeping query in block for ' - 'ICT dataset') - group.add_argument('--use-one-sent-docs', action='store_true', - help='Whether to use one sentence documents in ICT') - group.add_argument('--evidence-data-path', type=str, default=None, - help='Path to Wikipedia Evidence frm DPR paper') + group.add_argument( + "--titles-data-path", + type=str, + default=None, + help="Path to titles dataset used for ICT", + ) + group.add_argument( + "--query-in-block-prob", + type=float, + default=0.1, + help="Probability of keeping query in block for " "ICT dataset", + ) + group.add_argument( + "--use-one-sent-docs", + action="store_true", + help="Whether to use one sentence documents in ICT", + ) + group.add_argument( + "--evidence-data-path", + type=str, + default=None, + help="Path to Wikipedia Evidence frm DPR paper", + ) # training - group.add_argument('--retriever-report-topk-accuracies', nargs='+', type=int, - default=[], help="Which top-k accuracies to report " - "(e.g. '1 5 20')") - group.add_argument('--retriever-score-scaling', action='store_true', - help='Whether to scale retriever scores by inverse ' - 'square root of hidden size') + group.add_argument( + "--retriever-report-topk-accuracies", + nargs="+", + type=int, + default=[], + help="Which top-k accuracies to report " "(e.g. '1 5 20')", + ) + group.add_argument( + "--retriever-score-scaling", + action="store_true", + help="Whether to scale retriever scores by inverse " + "square root of hidden size", + ) # faiss index - group.add_argument('--block-data-path', type=str, default=None, - help='Where to save/load BlockData to/from') - group.add_argument('--embedding-path', type=str, default=None, - help='Where to save/load Open-Retrieval Embedding' - ' data to/from') + group.add_argument( + "--block-data-path", + type=str, + default=None, + help="Where to save/load BlockData to/from", + ) + group.add_argument( + "--embedding-path", + type=str, + default=None, + help="Where to save/load Open-Retrieval Embedding" " data to/from", + ) # indexer - group.add_argument('--indexer-batch-size', type=int, default=128, - help='How large of batches to use when doing indexing ' - 'jobs') - group.add_argument('--indexer-log-interval', type=int, default=1000, - help='After how many batches should the indexer ' - 'report progress') + group.add_argument( + "--indexer-batch-size", + type=int, + default=128, + help="How large of batches to use when doing indexing " "jobs", + ) + group.add_argument( + "--indexer-log-interval", + type=int, + default=1000, + help="After how many batches should the indexer " "report progress", + ) return parser @@ -1471,67 +2402,147 @@ def _add_vision_args(parser): group = parser.add_argument_group(title="vision") # general vision arguements - group.add_argument('--num-classes', type=int, default=1000, - help='num of classes in vision classificaiton task') - group.add_argument('--img-h', type=int, default=224, - help='Image height for vision classification task') - group.add_argument('--img-w', type=int, default=224, - help='Image height for vision classification task') - group.add_argument('--num-channels', type=int, default=3, - help='Number of channels in input image data') - group.add_argument('--patch-dim', type=int, default=16, - help='patch dimension') - group.add_argument('--classes-fraction', type=float, default=1.0, - help='training with fraction of classes.') - group.add_argument('--data-per-class-fraction', type=float, default=1.0, - help='training with fraction of data per class.') - group.add_argument('--no-data-sharding', action='store_false', - help='Disable data sharding.', - dest='data_sharding') - group.add_argument('--head-lr-mult', type=float, default=1.0, - help='learning rate multiplier for head during finetuning') + group.add_argument( + "--num-classes", + type=int, + default=1000, + help="num of classes in vision classificaiton task", + ) + group.add_argument( + "--img-h", + type=int, + default=224, + help="Image height for vision classification task", + ) + group.add_argument( + "--img-w", + type=int, + default=224, + help="Image height for vision classification task", + ) + group.add_argument( + "--num-channels", + type=int, + default=3, + help="Number of channels in input image data", + ) + group.add_argument("--patch-dim", type=int, default=16, help="patch dimension") + group.add_argument( + "--classes-fraction", + type=float, + default=1.0, + help="training with fraction of classes.", + ) + group.add_argument( + "--data-per-class-fraction", + type=float, + default=1.0, + help="training with fraction of data per class.", + ) + group.add_argument( + "--no-data-sharding", + action="store_false", + help="Disable data sharding.", + dest="data_sharding", + ) + group.add_argument( + "--head-lr-mult", + type=float, + default=1.0, + help="learning rate multiplier for head during finetuning", + ) # pretraining type and backbone selection` - group.add_argument('--vision-pretraining', action='store_true', - help='flag to indicate vision pretraining') - group.add_argument('--vision-pretraining-type', type=str, default='classify', - choices=['classify', 'inpaint', 'dino'], - help='pretraining objectives') - group.add_argument('--vision-backbone-type', type=str, default='vit', - choices=['vit', 'mit', 'swin'], - help='backbone types types') - group.add_argument('--swin-backbone-type', type=str, default='tiny', - choices=['tiny', 'base', 'h3'], - help='pretraining objectives') + group.add_argument( + "--vision-pretraining", + action="store_true", + help="flag to indicate vision pretraining", + ) + group.add_argument( + "--vision-pretraining-type", + type=str, + default="classify", + choices=["classify", "inpaint", "dino"], + help="pretraining objectives", + ) + group.add_argument( + "--vision-backbone-type", + type=str, + default="vit", + choices=["vit", "mit", "swin"], + help="backbone types types", + ) + group.add_argument( + "--swin-backbone-type", + type=str, + default="tiny", + choices=["tiny", "base", "h3"], + help="pretraining objectives", + ) # inpainting arguments - group.add_argument('--mask-type', type=str, default='random', - choices=['random', 'row'], - help='mask types') - group.add_argument('--mask-factor', type=float, default=1.0, - help='mask size scaling parameter') + group.add_argument( + "--mask-type", + type=str, + default="random", + choices=["random", "row"], + help="mask types", + ) + group.add_argument( + "--mask-factor", type=float, default=1.0, help="mask size scaling parameter" + ) # dino arguments - group.add_argument('--iter-per-epoch', type=int, default=1250, - help='iterations per epoch') - group.add_argument('--dino-local-img-size', type=int, default=96, - help='Image size for vision classification task') - group.add_argument('--dino-local-crops-number', type=int, default=10, - help='Number of local crops') - group.add_argument('--dino-head-hidden-size', type=int, default=2048, - help='Hidden dimension size in dino head') - group.add_argument('--dino-bottleneck-size', type=int, default=256, - help='Bottle neck dimension in dino head ') - group.add_argument('--dino-freeze-last-layer', type=float, default=1, - help='Freezing last layer weights') - group.add_argument('--dino-norm-last-layer', action='store_true', - help='Disable Norm in last layer.') - group.add_argument('--dino-warmup-teacher-temp', type=float, default=0.04, - help='warump teacher temperature') - group.add_argument('--dino-teacher-temp', type=float, default=0.07, - help='teacher temperature') - group.add_argument('--dino-warmup-teacher-temp-epochs', type=int, default=30, - help='warmup teacher temperaure epochs') + group.add_argument( + "--iter-per-epoch", type=int, default=1250, help="iterations per epoch" + ) + group.add_argument( + "--dino-local-img-size", + type=int, + default=96, + help="Image size for vision classification task", + ) + group.add_argument( + "--dino-local-crops-number", type=int, default=10, help="Number of local crops" + ) + group.add_argument( + "--dino-head-hidden-size", + type=int, + default=2048, + help="Hidden dimension size in dino head", + ) + group.add_argument( + "--dino-bottleneck-size", + type=int, + default=256, + help="Bottle neck dimension in dino head ", + ) + group.add_argument( + "--dino-freeze-last-layer", + type=float, + default=1, + help="Freezing last layer weights", + ) + group.add_argument( + "--dino-norm-last-layer", + action="store_true", + help="Disable Norm in last layer.", + ) + group.add_argument( + "--dino-warmup-teacher-temp", + type=float, + default=0.04, + help="warump teacher temperature", + ) + group.add_argument( + "--dino-teacher-temp", type=float, default=0.07, help="teacher temperature" + ) + group.add_argument( + "--dino-warmup-teacher-temp-epochs", + type=int, + default=30, + help="warmup teacher temperaure epochs", + ) return parser @@ -1539,105 +2550,187 @@ def _add_vision_args(parser): def _add_zero_args(parser): """Text generate arguments.""" - group = parser.add_argument_group('ZeRO configurations', 'configurations') + group = parser.add_argument_group("ZeRO configurations", "configurations") group.add_argument("--zero-stage", type=int, default=1.0) - group.add_argument('--zero-reduce-scatter', action='store_true', - help='Use reduce scatter if specified') - group.add_argument('--zero-contigious-gradients', action='store_true', - help='Use contigious memory optimizaiton if specified') + group.add_argument( + "--zero-reduce-scatter", + action="store_true", + help="Use reduce scatter if specified", + ) + group.add_argument( + "--zero-contigious-gradients", + action="store_true", + help="Use contigious memory optimizaiton if specified", + ) group.add_argument("--zero-reduce-bucket-size", type=int, default=0.0) group.add_argument("--zero-allgather-bucket-size", type=int, default=0.0) - group.add_argument('--remote-device', type=str, default='none', choices=['none', 'cpu', 'nvme'], - help='Remote device for ZeRO-3 initialized parameters.') - group.add_argument('--use-pin-memory', action='store_true', - help='Use pinned CPU memory for ZeRO-3 initialized model parameters.') - group.add_argument('--use-mics', action='store_true', - help='Use MiCS') + group.add_argument( + "--remote-device", + type=str, + default="none", + choices=["none", "cpu", "nvme"], + help="Remote device for ZeRO-3 initialized parameters.", + ) + group.add_argument( + "--use-pin-memory", + action="store_true", + help="Use pinned CPU memory for ZeRO-3 initialized model parameters.", + ) + group.add_argument("--use-mics", action="store_true", help="Use MiCS") return parser + def _add_memoryopt_args(parser): """Memory optimization arguments.""" - group = parser.add_argument_group('Memory optimizations', 'configurations') - group.add_argument("--scattered-embeddings", action='store_true', - help='Save memory by scattering embedding activations. ' - 'Introduces dropout differences across MP configurations.') - group.add_argument("--split-transformers", action='store_true', - help='Save memory by splitting transformer layers into two parts, ' - 'allowing for more frequent activation checkpoint savings.') - group.add_argument("--memory-centric-tiled-linear", action="store_true", - help='Save memory by tiling with deepspeed.zero.TiledLinear.') - group.add_argument("--tile-factor", type=int, default=1, - help='Make all linear layers the same size of [hidden/tile_factor, hidden/tile_factor]. ' - 'Must be enabled with --memory-centric-tiled-linear. ' - 'Example A: if tile_factor=1, the qkv layer [hidden, 3* hidden] would be converted into [1,3] tiles of size [hidden,hidden]. ' - 'Example B: if tile_factor=2, the intermediate layer [4*hidden, hidden] will be converted into [8, 2] tiles of size [hidden/2, hidden/2]. ' - 'Default is 1.') + group = parser.add_argument_group("Memory optimizations", "configurations") + group.add_argument( + "--scattered-embeddings", + action="store_true", + help="Save memory by scattering embedding activations. " + "Introduces dropout differences across MP configurations.", + ) + group.add_argument( + "--split-transformers", + action="store_true", + help="Save memory by splitting transformer layers into two parts, " + "allowing for more frequent activation checkpoint savings.", + ) + group.add_argument( + "--memory-centric-tiled-linear", + action="store_true", + help="Save memory by tiling with deepspeed.zero.TiledLinear.", + ) + group.add_argument( + "--tile-factor", + type=int, + default=1, + help="Make all linear layers the same size of [hidden/tile_factor, hidden/tile_factor]. " + "Must be enabled with --memory-centric-tiled-linear. " + "Example A: if tile_factor=1, the qkv layer [hidden, 3* hidden] would be converted into [1,3] tiles of size [hidden,hidden]. " + "Example B: if tile_factor=2, the intermediate layer [4*hidden, hidden] will be converted into [8, 2] tiles of size [hidden/2, hidden/2]. " + "Default is 1.", + ) return parser + def _add_activation_checkpoint_args(parser): - group = parser.add_argument_group('Activation Checkpointing', - 'Checkpointing Configurations') - group.add_argument('--deepspeed-activation-checkpointing', action='store_true', - help='uses activation checkpointing from deepspeed') - group.add_argument('--partition-activations', action='store_true', - help='partition Activations across GPUs before checkpointing.') - group.add_argument('--contigious-checkpointing', action='store_true', - help='Contigious memory checkpointing for activatoins.') - group.add_argument('--checkpoint-in-cpu', action='store_true', - help='Move the activation checkpoints to CPU.') - group.add_argument('--synchronize-each-layer', action='store_true', - help='does a synchronize at the beginning and end of each checkpointed layer.') - group.add_argument('--profile-backward', action='store_true', - help='Enables backward pass profiling for checkpointed layers.') + group = parser.add_argument_group( + "Activation Checkpointing", "Checkpointing Configurations" + ) + group.add_argument( + "--deepspeed-activation-checkpointing", + action="store_true", + help="uses activation checkpointing from deepspeed", + ) + group.add_argument( + "--partition-activations", + action="store_true", + help="partition Activations across GPUs before checkpointing.", + ) + group.add_argument( + "--contigious-checkpointing", + action="store_true", + help="Contigious memory checkpointing for activatoins.", + ) + group.add_argument( + "--checkpoint-in-cpu", + action="store_true", + help="Move the activation checkpoints to CPU.", + ) + group.add_argument( + "--synchronize-each-layer", + action="store_true", + help="does a synchronize at the beginning and end of each checkpointed layer.", + ) + group.add_argument( + "--profile-backward", + action="store_true", + help="Enables backward pass profiling for checkpointed layers.", + ) return parser def _add_distillation_args(parser): - group = parser.add_argument_group('Knowledge distillation', - 'Distillation Configurations') - group.add_argument('--num-layers-teacher', type=int, default=None, - help='Number of the teacher transformer layers.') - group.add_argument('--num-experts-teacher', type=int, nargs='+', default=[1,], - help='number of teacher experts list, MoE related.') - group.add_argument('--hidden-size-teacher', type=int, default=None, - help='Tansformer teacher hidden size.') - group.add_argument('--num-attention-heads-teacher', type=int, default=None, - help='Number of teacher transformer attention heads.') - group.add_argument('--mos', action='store_true', - help='Enable Mixture-of-Students via knolwedge distillation.') - group.add_argument('--kd', action='store_true', - help='Enable knolwedge distillation.') - group.add_argument('--kd-alpha-ce', default=1, type=float) - group.add_argument('--kd-beta-ce', default=1, type=float) - group.add_argument('--kd-temp', default=1.0, type=float) - group.add_argument('--reset-iteration', action='store_true', - help='Reset the iteration count.') - group.add_argument('--load-teacher', type=str, default=None, - help='Directory containing a teacher model checkpoint.') + group = parser.add_argument_group( + "Knowledge distillation", "Distillation Configurations" + ) + group.add_argument( + "--num-layers-teacher", + type=int, + default=None, + help="Number of the teacher transformer layers.", + ) + group.add_argument( + "--num-experts-teacher", + type=int, + nargs="+", + default=[ + 1, + ], + help="number of teacher experts list, MoE related.", + ) + group.add_argument( + "--hidden-size-teacher", + type=int, + default=None, + help="Tansformer teacher hidden size.", + ) + group.add_argument( + "--num-attention-heads-teacher", + type=int, + default=None, + help="Number of teacher transformer attention heads.", + ) + group.add_argument( + "--mos", + action="store_true", + help="Enable Mixture-of-Students via knolwedge distillation.", + ) + group.add_argument( + "--kd", action="store_true", help="Enable knolwedge distillation." + ) + group.add_argument("--kd-alpha-ce", default=1, type=float) + group.add_argument("--kd-beta-ce", default=1, type=float) + group.add_argument("--kd-temp", default=1.0, type=float) + group.add_argument( + "--reset-iteration", action="store_true", help="Reset the iteration count." + ) + group.add_argument( + "--load-teacher", + type=str, + default=None, + help="Directory containing a teacher model checkpoint.", + ) return parser def _add_profiler_args(parser): - group = parser.add_argument_group(title='profiling configuration') - - group.add_argument("--profile", - type=str, - default=None, - choices=['pt', 'pt-full'], - help="Enable profiling, pt-full gives call stack compared to pt") - - group.add_argument("--profile_steps", - type=str, - default='2,3', - help="Which steps to profile. Format: ,") - - group.add_argument("--profile-ranks", - type=int, - nargs='+', - default=None, - help="Which ranks to profile. Format: 0 1 2 3") + group = parser.add_argument_group(title="profiling configuration") + + group.add_argument( + "--profile", + type=str, + default=None, + choices=["pt", "pt-full"], + help="Enable profiling, pt-full gives call stack compared to pt", + ) + + group.add_argument( + "--profile_steps", + type=str, + default="2,3", + help="Which steps to profile. Format: ,", + ) + + group.add_argument( + "--profile-ranks", + type=int, + nargs="+", + default=None, + help="Which ranks to profile. Format: 0 1 2 3", + ) return parser diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index f420795f522..f3d83c5e0c9 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -18,9 +18,7 @@ from megatron import update_num_microbatches, get_tokenizer from megatron.core import mpu, tensor_parallel from .global_vars import get_args -from .utils import (unwrap_model, - print_rank_0, - is_rank_0) +from .utils import (unwrap_model, print_rank_0, is_rank_0) from .utils import PerfTrace, Profile from deepspeed.checkpoint import ( @@ -40,11 +38,12 @@ _CHECKPOINT_VERSION = None dlp = Profile("CHECKPOINT") + + def set_checkpoint_version(value): global _CHECKPOINT_VERSION if _CHECKPOINT_VERSION is not None: - assert _CHECKPOINT_VERSION == value, \ - "checkpoint versions do not match" + assert _CHECKPOINT_VERSION == value, "checkpoint versions do not match" _CHECKPOINT_VERSION = value @@ -68,51 +67,56 @@ def _compare(arg_name, old_arg_name=None, default=None): else: checkpoint_value = getattr(checkpoint_args, ckpt_arg_name) args_value = getattr(args, arg_name) - error_message = '{} value from checkpoint ({}) is not equal to the ' \ - 'input argument value ({}).'.format( - arg_name, checkpoint_value, args_value) + error_message = ( + "{} value from checkpoint ({}) is not equal to the " + "input argument value ({}).".format(arg_name, checkpoint_value, args_value) + ) assert checkpoint_value == args_value, error_message if not args.mos and not args.kd: - _compare('num_layers') - _compare('hidden_size') - _compare('num_attention_heads') - _compare('num_key_value_heads') - _compare('add_position_embedding', default=True) + _compare("num_layers") + _compare("hidden_size") + _compare("num_attention_heads") + _compare("num_key_value_heads") + _compare("add_position_embedding", default=True) if args.vocab_file: - _compare('max_position_embeddings') + _compare("max_position_embeddings") if not args.universal_checkpoint: - _compare('make_vocab_size_divisible_by') - _compare('padded_vocab_size') - _compare('tokenizer_type') + _compare("make_vocab_size_divisible_by") + _compare("padded_vocab_size") + _compare("tokenizer_type") if args.data_parallel_random_init: - _compare('data_parallel_random_init') + _compare("data_parallel_random_init") if get_checkpoint_version() < 3.0 and not args.universal_checkpoint: - _compare('tensor_model_parallel_size', - old_arg_name='model_parallel_size') + _compare("tensor_model_parallel_size", old_arg_name="model_parallel_size") if get_checkpoint_version() >= 3.0 and not args.universal_checkpoint: - _compare('tensor_model_parallel_size') - _compare('pipeline_model_parallel_size') + _compare("tensor_model_parallel_size") + _compare("pipeline_model_parallel_size") def ensure_directory_exists(filename): """Build filename's path if it does not already exists.""" dirname = os.path.dirname(filename) - os.makedirs(dirname, exist_ok = True) + os.makedirs(dirname, exist_ok=True) -def get_checkpoint_name(checkpoints_path, iteration, release=False, - pipeline_parallel=None, - tensor_rank=None, pipeline_rank=None): +def get_checkpoint_name( + checkpoints_path, + iteration, + release=False, + pipeline_parallel=None, + tensor_rank=None, + pipeline_rank=None, +): """Determine the directory name for this rank's checkpoint.""" if release: - directory = 'release' + directory = "release" else: - directory = 'iter_{:07d}'.format(iteration) + directory = "iter_{:07d}".format(iteration) # Use both the tensor and pipeline MP rank. if pipeline_parallel is None: - pipeline_parallel = (mpu.get_pipeline_model_parallel_world_size() > 1) + pipeline_parallel = mpu.get_pipeline_model_parallel_world_size() > 1 if tensor_rank is None: tensor_rank = mpu.get_tensor_model_parallel_rank() if pipeline_rank is None: @@ -122,18 +126,21 @@ def get_checkpoint_name(checkpoints_path, iteration, release=False, # optimizer, then the optimizer's path must additionally include the # data parallel rank. if not pipeline_parallel: - common_path = os.path.join(checkpoints_path, directory, - f'mp_rank_{tensor_rank:02d}') + common_path = os.path.join( + checkpoints_path, directory, f"mp_rank_{tensor_rank:02d}" + ) else: - common_path = os.path.join(checkpoints_path, directory, - f'mp_rank_{tensor_rank:02d}_{pipeline_rank:03d}') + common_path = os.path.join( + checkpoints_path, + directory, + f"mp_rank_{tensor_rank:02d}_{pipeline_rank:03d}", + ) return os.path.join(common_path, "model_optim_rng.pt") def get_distributed_optimizer_checkpoint_name(model_checkpoint_name): - return os.path.join(os.path.dirname(model_checkpoint_name), - "distrib_optim.pt") + return os.path.join(os.path.dirname(model_checkpoint_name), "distrib_optim.pt") def find_checkpoint_rank_0(checkpoints_path, iteration, release=False): @@ -146,16 +153,26 @@ def find_checkpoint_rank_0(checkpoints_path, iteration, release=False): """ # Look for checkpoint with no pipelining - filename = get_checkpoint_name(checkpoints_path, iteration, release, - pipeline_parallel=False, - tensor_rank=0, pipeline_rank=0) + filename = get_checkpoint_name( + checkpoints_path, + iteration, + release, + pipeline_parallel=False, + tensor_rank=0, + pipeline_rank=0, + ) if os.path.isfile(filename): return filename # Look for checkpoint with pipelining - filename = get_checkpoint_name(checkpoints_path, iteration, release, - pipeline_parallel=True, - tensor_rank=0, pipeline_rank=0) + filename = get_checkpoint_name( + checkpoints_path, + iteration, + release, + pipeline_parallel=True, + tensor_rank=0, + pipeline_rank=0, + ) if os.path.isfile(filename): return filename @@ -163,10 +180,10 @@ def find_checkpoint_rank_0(checkpoints_path, iteration, release=False): def get_checkpoint_tracker_filename(checkpoints_path): - """Tracker file rescords the latest chckpoint during training to restart from.""" - return os.path.join(checkpoints_path, 'latest_checkpointed_iteration.txt') + return os.path.join(checkpoints_path, "latest_checkpointed_iteration.txt") + @dlp.log def read_metadata(tracker_filename): @@ -174,18 +191,20 @@ def read_metadata(tracker_filename): # mark it as a release checkpoint. iteration = 0 release = False - with open(tracker_filename, 'r') as f: + with open(tracker_filename, "r") as f: metastring = f.read().strip() try: iteration = int(metastring) except ValueError: - release = metastring == 'release' + release = metastring == "release" if not release: - print_rank_0('ERROR: Invalid metadata file {}. Exiting'.format( - tracker_filename)) + print_rank_0( + "ERROR: Invalid metadata file {}. Exiting".format(tracker_filename) + ) sys.exit() - assert iteration > 0 or release, 'error parsing metadata file {}'.format( - tracker_filename) + assert iteration > 0 or release, "error parsing metadata file {}".format( + tracker_filename + ) # Get the max iteration retrieved across the ranks. if torch.distributed.is_initialized(): @@ -197,10 +216,14 @@ def read_metadata(tracker_filename): # If not, print a warning and chose the maximum # iteration across all ranks. if iteration != max_iter: - print('WARNING: on rank {} found iteration {} in the ' - 'metadata while max iteration across the ranks ' - 'is {}, replacing it with max iteration.'.format( - torch.distributed.get_rank(), iteration, max_iter), flush=True) + print( + "WARNING: on rank {} found iteration {} in the " + "metadata while max iteration across the ranks " + "is {}, replacing it with max iteration.".format( + torch.distributed.get_rank(), iteration, max_iter + ), + flush=True, + ) else: # When loading a checkpoint outside of training (for example, # when editing it), we might not have torch distributed @@ -208,32 +231,35 @@ def read_metadata(tracker_filename): max_iter = iteration return max_iter, release + @dlp.log def get_rng_state(): - """ collect rng state across data parallel ranks """ + """collect rng state across data parallel ranks""" args = get_args() rng_state = { - 'random_rng_state': random.getstate(), - 'np_rng_state': np.random.get_state(), - 'torch_rng_state': torch.get_rng_state(), - 'cuda_rng_state': get_accelerator().get_rng_state(), - 'rng_tracker_states': tensor_parallel.get_cuda_rng_tracker().get_states()} + "random_rng_state": random.getstate(), + "np_rng_state": np.random.get_state(), + "torch_rng_state": torch.get_rng_state(), + "cuda_rng_state": get_accelerator().get_rng_state(), + "rng_tracker_states": tensor_parallel.get_cuda_rng_tracker().get_states(), + } rng_state_list = None - if torch.distributed.is_initialized() and \ - mpu.get_data_parallel_world_size() > 1 and \ - args.data_parallel_random_init: - rng_state_list = \ - [None for i in range(mpu.get_data_parallel_world_size())] + if ( + torch.distributed.is_initialized() + and mpu.get_data_parallel_world_size() > 1 + and args.data_parallel_random_init + ): + rng_state_list = [None for i in range(mpu.get_data_parallel_world_size())] torch.distributed.all_gather_object( - rng_state_list, - rng_state, - group=mpu.get_data_parallel_group()) + rng_state_list, rng_state, group=mpu.get_data_parallel_group() + ) else: rng_state_list = [rng_state] return rng_state_list + @dlp.log def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): """Save a model checkpoint.""" @@ -249,8 +275,9 @@ def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): if not args.deepspeed: model = unwrap_model(model) - print_rank_0('saving checkpoint at iteration {:7d} to {}'.format( - iteration, args.save)) + print_rank_0( + "saving checkpoint at iteration {:7d} to {}".format(iteration, args.save) + ) # Collect rng state across data parallel ranks. rng_state = get_rng_state() @@ -260,40 +287,44 @@ def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): # Save distributed optimizer's custom parameter state. if args.use_distributed_optimizer: - optim_checkpoint_name = \ - get_distributed_optimizer_checkpoint_name(checkpoint_name) + optim_checkpoint_name = get_distributed_optimizer_checkpoint_name( + checkpoint_name + ) ensure_directory_exists(optim_checkpoint_name) optimizer.save_parameter_state(optim_checkpoint_name) # Collect args, model, RNG. - if not torch.distributed.is_initialized() \ - or mpu.get_data_parallel_rank() == 0 or args.deepspeed: + if ( + not torch.distributed.is_initialized() + or mpu.get_data_parallel_rank() == 0 + or args.deepspeed + ): # Arguments, iteration, and model. state_dict = {} - state_dict['args'] = args - state_dict['checkpoint_version'] = 3.0 - state_dict['iteration'] = iteration - state_dict['tokens'] = args.consumed_train_tokens + state_dict["args"] = args + state_dict["checkpoint_version"] = 3.0 + state_dict["iteration"] = iteration + state_dict["tokens"] = args.consumed_train_tokens state_dict[UNIVERSAL_CHECKPOINT_INFO] = _universal_checkpoint_info(model) # DeepSpeed saves the model/optimizer/scheduler if not args.deepspeed: if len(model) == 1: - state_dict['model'] = model[0].state_dict_for_save_checkpoint() + state_dict["model"] = model[0].state_dict_for_save_checkpoint() else: for i in range(len(model)): mpu.set_virtual_pipeline_model_parallel_rank(i) - state_dict['model%d' % i] = \ - model[i].state_dict_for_save_checkpoint() + state_dict["model%d" % i] = model[ + i + ].state_dict_for_save_checkpoint() # Optimizer stuff. if not args.no_save_optim: if optimizer is not None: - state_dict['optimizer'] = optimizer.state_dict() + state_dict["optimizer"] = optimizer.state_dict() if opt_param_scheduler is not None: - state_dict['opt_param_scheduler'] = \ - opt_param_scheduler.state_dict() + state_dict["opt_param_scheduler"] = opt_param_scheduler.state_dict() # RNG states. if not args.no_save_rng: @@ -305,12 +336,18 @@ def save_checkpoint(iteration, model, optimizer, opt_param_scheduler): torch.save(state_dict, checkpoint_name) if args.deepspeed: - #megatron model uses state_dict_for_save_checkpointing instead of the standard state_dict - #state_dict is used by deepspeed for module saving so it needs to point to the right function + # megatron model uses state_dict_for_save_checkpointing instead of the standard state_dict + # state_dict is used by deepspeed for module saving so it needs to point to the right function if args.no_pipeline_parallel: original_state_dict = model[0].module.state_dict - def state_dict_for_save_checkpoint_deepspeed(destination=None, prefix='', keep_vars=False): - return model[0].module.state_dict_for_save_checkpoint(prefix=prefix, keep_vars=keep_vars) + + def state_dict_for_save_checkpoint_deepspeed( + destination=None, prefix="", keep_vars=False + ): + return model[0].module.state_dict_for_save_checkpoint( + prefix=prefix, keep_vars=keep_vars + ) + model[0].module.state_dict = state_dict_for_save_checkpoint_deepspeed # Saving is a collective communication @@ -328,39 +365,47 @@ def state_dict_for_save_checkpoint_deepspeed(destination=None, prefix='', keep_v if torch.distributed.is_initialized(): torch.distributed.barrier() - print_rank_0(' successfully saved checkpoint at iteration {:7d} to {}' \ - .format(iteration, args.save)) + print_rank_0( + " successfully saved checkpoint at iteration {:7d} to {}".format( + iteration, args.save + ) + ) # And update the latest iteration if is_rank_0(): tracker_filename = get_checkpoint_tracker_filename(args.save) - with open(tracker_filename, 'w') as f: + with open(tracker_filename, "w") as f: f.write(str(iteration)) # Wait so everyone is done (not necessary) if torch.distributed.is_initialized(): torch.distributed.barrier() + @dlp.log def _transpose_first_dim(t, num_splits, num_splits_first, model): input_shape = t.size() # We use a self_attention module but the values extracted aren't # specific to self attention so should work for cross attention as well - while hasattr(model, 'module'): + while hasattr(model, "module"): model = model.module attention_module = model.language_model.encoder.layers[0].self_attention - #attention_module = model.language_model.encoder.layers[0].attention + # attention_module = model.language_model.encoder.layers[0].attention hidden_size_per_attention_head = attention_module.hidden_size_per_attention_head - num_attention_heads_per_partition = attention_module.num_attention_heads_per_partition + num_attention_heads_per_partition = ( + attention_module.num_attention_heads_per_partition + ) if num_splits_first: """[num_splits * np * hn, h] -->(view) [num_splits, np, hn, h] -->(tranpose) [np, num_splits, hn, h] - -->(view) [np * num_splits * hn, h] """ + -->(view) [np * num_splits * hn, h]""" - intermediate_shape = \ - (num_splits, num_attention_heads_per_partition, - hidden_size_per_attention_head) + input_shape[1:] + intermediate_shape = ( + num_splits, + num_attention_heads_per_partition, + hidden_size_per_attention_head, + ) + input_shape[1:] t = t.view(*intermediate_shape) t = t.transpose(0, 1).contiguous() @@ -368,12 +413,13 @@ def _transpose_first_dim(t, num_splits, num_splits_first, model): """[np * hn * num_splits, h] -->(view) [np, hn, num_splits, h] -->(tranpose) [np, num_splits, hn, h] - -->(view) [np * num_splits * hn, h] """ + -->(view) [np * num_splits * hn, h]""" - intermediate_shape = \ - (num_attention_heads_per_partition, - hidden_size_per_attention_head, num_splits) +\ - input_shape[1:] + intermediate_shape = ( + num_attention_heads_per_partition, + hidden_size_per_attention_head, + num_splits, + ) + input_shape[1:] t = t.view(*intermediate_shape) t = t.transpose(1, 2).contiguous() @@ -381,16 +427,17 @@ def _transpose_first_dim(t, num_splits, num_splits_first, model): return t + def fix_query_key_value_ordering(model, checkpoint_version): """Fix up query/key/value matrix ordering if checkpoint version is smaller than 2.0 """ if checkpoint_version < 2.0: if isinstance(model, list): - assert len(model)==1 + assert len(model) == 1 model = model[0] for name, param in model.named_parameters(): - if name.endswith(('.query_key_value.weight', '.query_key_value.bias')): + if name.endswith((".query_key_value.weight", ".query_key_value.bias")): if checkpoint_version == 0: fixed_param = _transpose_first_dim(param.data, 3, True, model) elif checkpoint_version == 1.0: @@ -399,7 +446,7 @@ def fix_query_key_value_ordering(model, checkpoint_version): print_rank_0(f"Invalid checkpoint version {checkpoint_version}.") sys.exit() param.data.copy_(fixed_param) - if name.endswith(('.key_value.weight', '.key_value.bias')): + if name.endswith((".key_value.weight", ".key_value.bias")): if checkpoint_version == 0: fixed_param = _transpose_first_dim(param.data, 2, True, model) elif checkpoint_version == 1.0: @@ -408,12 +455,15 @@ def fix_query_key_value_ordering(model, checkpoint_version): print_rank_0(f"Invalid checkpoint version {checkpoint_version}.") sys.exit() param.data.copy_(fixed_param) - print_rank_0(" succesfully fixed query-key-values ordering for" - " checkpoint version {}".format(checkpoint_version)) + print_rank_0( + " succesfully fixed query-key-values ordering for" + " checkpoint version {}".format(checkpoint_version) + ) + @dlp.log def _load_base_checkpoint(load_dir, rank0=False): - """ Load the base state_dict from the given directory + """Load the base state_dict from the given directory If rank0 is true, just loads rank 0 checkpoint, ignoring arguments. """ @@ -424,10 +474,12 @@ def _load_base_checkpoint(load_dir, rank0=False): # If no tracker file, return nothing if not os.path.isfile(tracker_filename): if not rank0: - print_rank_0('WARNING: could not find the metadata file {} '.format( - tracker_filename)) - print_rank_0(' will not load any checkpoints and will start from ' - 'random') + print_rank_0( + "WARNING: could not find the metadata file {} ".format(tracker_filename) + ) + print_rank_0( + " will not load any checkpoints and will start from " "random" + ) return None, False # Otherwise, read the tracker file and either set the iteration or @@ -440,34 +492,39 @@ def _load_base_checkpoint(load_dir, rank0=False): else: checkpoint_name = get_checkpoint_name(load_dir, iteration, release) if release: - print_rank_0(f' loading release checkpoint from {load_dir}') + print_rank_0(f" loading release checkpoint from {load_dir}") else: - print_rank_0(f' loading checkpoint from {load_dir} at iteration {iteration}') + print_rank_0( + f" loading checkpoint from {load_dir} at iteration {iteration}" + ) # Load the checkpoint. try: - state_dict = torch.load(checkpoint_name, map_location='cpu') + state_dict = torch.load(checkpoint_name, map_location="cpu") except ModuleNotFoundError: from megatron.fp16_deprecated import loss_scaler # For backward compatibility. if not rank0: - print_rank_0(' > deserializing using the old code structure ...') - sys.modules['fp16.loss_scaler'] = sys.modules[ - 'megatron.fp16_deprecated.loss_scaler'] - sys.modules['megatron.fp16.loss_scaler'] = sys.modules[ - 'megatron.fp16_deprecated.loss_scaler'] - state_dict = torch.load(checkpoint_name, map_location='cpu') - sys.modules.pop('fp16.loss_scaler', None) - sys.modules.pop('megatron.fp16.loss_scaler', None) + print_rank_0(" > deserializing using the old code structure ...") + sys.modules["fp16.loss_scaler"] = sys.modules[ + "megatron.fp16_deprecated.loss_scaler" + ] + sys.modules["megatron.fp16.loss_scaler"] = sys.modules[ + "megatron.fp16_deprecated.loss_scaler" + ] + state_dict = torch.load(checkpoint_name, map_location="cpu") + sys.modules.pop("fp16.loss_scaler", None) + sys.modules.pop("megatron.fp16.loss_scaler", None) except BaseException as e: - print_rank_0('could not load the checkpoint') + print_rank_0("could not load the checkpoint") print_rank_0(e) sys.exit() return state_dict, release + @dlp.log -def load_args_from_checkpoint(args, load_arg='load'): +def load_args_from_checkpoint(args, load_arg="load"): """Set required arguments from the checkpoint specified in the arguments. @@ -483,27 +540,35 @@ def load_args_from_checkpoint(args, load_arg='load'): load_dir = getattr(args, load_arg) if load_dir is None: - print_rank_0('No load directory specified, using provided arguments.') + print_rank_0("No load directory specified, using provided arguments.") return args state_dict, release = _load_base_checkpoint(load_dir, rank0=True) # Args. if not state_dict: - print_rank_0('Checkpoint not found to provide arguments, using provided arguments.') + print_rank_0( + "Checkpoint not found to provide arguments, using provided arguments." + ) return args - if 'args' not in state_dict: - print_rank_0('Checkpoint provided does not have arguments saved, using provided arguments.') + if "args" not in state_dict: + print_rank_0( + "Checkpoint provided does not have arguments saved, using provided arguments." + ) return args - checkpoint_args = state_dict['args'] - checkpoint_version = state_dict.get('checkpoint_version', 0) - args.iteration = state_dict['iteration'] + checkpoint_args = state_dict["args"] + checkpoint_version = state_dict.get("checkpoint_version", 0) + args.iteration = state_dict["iteration"] # One-off conversion for foundation models - if hasattr(checkpoint_args, 'disable_bias_linear'): - setattr(checkpoint_args, 'add_bias_linear', not getattr(checkpoint_args, 'disable_bias_linear')) + if hasattr(checkpoint_args, "disable_bias_linear"): + setattr( + checkpoint_args, + "add_bias_linear", + not getattr(checkpoint_args, "disable_bias_linear"), + ) def _set_arg(arg_name, old_arg_name=None, force=False): if not force and getattr(args, arg_name, None) is not None: @@ -520,33 +585,33 @@ def _set_arg(arg_name, old_arg_name=None, force=False): else: print_rank_0(f"Checkpoint did not provide arguments {arg_name}") - _set_arg('num_layers') - _set_arg('hidden_size') - _set_arg('ffn_hidden_size') - _set_arg('seq_length') - _set_arg('num_attention_heads') - _set_arg('num_key_value_heads') - _set_arg('kv_channels') - _set_arg('max_position_embeddings') - _set_arg('add_position_embedding', force=True) - _set_arg('use_rotary_position_embeddings', force=True) - _set_arg('rotary_percent', force=True) - _set_arg('add_bias_linear', force=True) - _set_arg('swiglu', force=True) - _set_arg('untie_embeddings_and_output_weights', force=True) - _set_arg('apply_layernorm_1p', force=True) - _set_arg('tokenizer_type') - _set_arg('padded_vocab_size') + _set_arg("num_layers") + _set_arg("hidden_size") + _set_arg("ffn_hidden_size") + _set_arg("seq_length") + _set_arg("num_attention_heads") + _set_arg("num_key_value_heads") + _set_arg("kv_channels") + _set_arg("max_position_embeddings") + _set_arg("add_position_embedding", force=True) + _set_arg("use_rotary_position_embeddings", force=True) + _set_arg("rotary_percent", force=True) + _set_arg("add_bias_linear", force=True) + _set_arg("swiglu", force=True) + _set_arg("untie_embeddings_and_output_weights", force=True) + _set_arg("apply_layernorm_1p", force=True) + _set_arg("tokenizer_type") + _set_arg("padded_vocab_size") if checkpoint_version < 3.0: - _set_arg('tensor_model_parallel_size', - 'model_parallel_size') + _set_arg("tensor_model_parallel_size", "model_parallel_size") else: - _set_arg('tensor_model_parallel_size', force=True) - _set_arg('pipeline_model_parallel_size', force=True) - _set_arg('virtual_pipeline_model_parallel_size', force=True) - _set_arg('num_layers_per_virtual_pipeline_stage') + _set_arg("tensor_model_parallel_size", force=True) + _set_arg("pipeline_model_parallel_size", force=True) + _set_arg("virtual_pipeline_model_parallel_size", force=True) + _set_arg("num_layers_per_virtual_pipeline_stage") return args, checkpoint_args + @dlp.log def load_lr_state_dict(strict: bool = False) -> dict: """Load {iteration, lr} from .yaml file when restoring from checkpoint.""" @@ -557,20 +622,19 @@ def load_lr_state_dict(strict: bool = False) -> dict: ) lr_state_dict = {} if lr_state_dict_fp.is_file(): - with lr_state_dict_fp.open('r') as f: + with lr_state_dict_fp.open("r") as f: lr_state_dict = yaml.safe_load(f) - args.lr = lr_state_dict['lr'] + args.lr = lr_state_dict["lr"] else: if strict: - raise FileNotFoundError( - f"{lr_state_dict_fp=}.is_file() is False" - ) + raise FileNotFoundError(f"{lr_state_dict_fp=}.is_file() is False") log.info( f"Unable to load lr_state_dict from {lr_state_dict_fp=}, " f"but strict=False. Returning empty dictionary: {lr_state_dict=}" ) return lr_state_dict + @dlp.log def save_lr_state_dict() -> None: """Save {iteration, lr} to .yaml file for safe-keeping. @@ -581,28 +645,24 @@ def save_lr_state_dict() -> None: return None args = get_args() assert args is not None - outdir = getattr(args, 'save', None) + outdir = getattr(args, "save", None) assert outdir is not None - lr_state_dict_fp = Path(args.save).joinpath( - "lr_state_dict.yaml" - ) + lr_state_dict_fp = Path(args.save).joinpath("lr_state_dict.yaml") lr_state_dict_fp.parent.mkdir(parents=True, exist_ok=True) log.info(f"Saving lr_state_dict to {lr_state_dict_fp.as_posix()}") - with lr_state_dict_fp.open('w') as f: - yaml.dump( - {'iteration': args.iteration, 'lr': args.lr}, - f - ) + with lr_state_dict_fp.open("w") as f: + yaml.dump({"iteration": args.iteration, "lr": args.lr}, f) + @dlp.log def load_checkpoint( - model, - optimizer, - opt_param_scheduler, - load_arg: str = 'load', - strict: bool = True, - load_only_weights: bool = False, - strict_lr_state_dict: bool = False + model, + optimizer, + opt_param_scheduler, + load_arg: str = "load", + strict: bool = True, + load_only_weights: bool = False, + strict_lr_state_dict: bool = False, ): """Load a model checkpoint and return the iteration. strict (bool): whether to strictly enforce that the keys in @@ -616,9 +676,9 @@ def load_checkpoint( lr_tensor = torch.tensor(args.lr, requires_grad=False, device=DEVICE) if RANK == 0: lr_state_dict = load_lr_state_dict(strict=strict_lr_state_dict) - if len(lr_state_dict.keys()) > 0 and 'lr' in lr_state_dict: + if len(lr_state_dict.keys()) > 0 and "lr" in lr_state_dict: lr_tensor = torch.tensor( - lr_state_dict['lr'], + lr_state_dict["lr"], requires_grad=False, device=DEVICE, ) @@ -626,18 +686,25 @@ def load_checkpoint( args.lr = lr_tensor.item() if args.deepspeed: if args.finetune: - loaded_dir, state_dict = model[0].load_checkpoint(load_dir, - load_module_strict=strict, load_optimizer_states=False, - load_lr_scheduler_states=False, load_module_only=True, - tag=args.load_tag) + loaded_dir, state_dict = model[0].load_checkpoint( + load_dir, + load_module_strict=strict, + load_optimizer_states=False, + load_lr_scheduler_states=False, + load_module_only=True, + tag=args.load_tag, + ) else: - loaded_dir, state_dict = model[0].load_checkpoint(load_dir, - load_module_strict=strict, tag=args.load_tag) + loaded_dir, state_dict = model[0].load_checkpoint( + load_dir, load_module_strict=strict, tag=args.load_tag + ) if loaded_dir is None: - print_rank_0('WARNING: could not find the metadata file {} '.format( - load_dir)) - print_rank_0(' will not load any checkpoints and will start from ' - 'random') + print_rank_0( + "WARNING: could not find the metadata file {} ".format(load_dir) + ) + print_rank_0( + " will not load any checkpoints and will start from " "random" + ) return 0 release = False else: @@ -656,10 +723,10 @@ def load_checkpoint( # Iteration defaults to 0. return 0 - checkpoint_name = get_checkpoint_name(load_dir, state_dict['iteration'], release) + checkpoint_name = get_checkpoint_name(load_dir, state_dict["iteration"], release) # Set checkpoint version. - set_checkpoint_version(state_dict.get('checkpoint_version', 0)) + set_checkpoint_version(state_dict.get("checkpoint_version", 0)) # Set iteration. if args.finetune or release or args.reset_iteration or load_only_weights: @@ -668,16 +735,17 @@ def load_checkpoint( model[0].global_steps = 0 else: try: - iteration = state_dict['iteration'] - if 'tokens' in state_dict: - args.consumed_train_tokens = state_dict['tokens'] + iteration = state_dict["iteration"] + if "tokens" in state_dict: + args.consumed_train_tokens = state_dict["tokens"] except KeyError: try: # Backward compatible with older checkpoints - iteration = state_dict['total_iters'] + iteration = state_dict["total_iters"] except KeyError: - print_rank_0('A metadata file exists but unable to load ' - 'iteration from checkpoint {}, exiting'.format( - checkpoint_name)) + print_rank_0( + "A metadata file exists but unable to load " + "iteration from checkpoint {}, exiting".format(checkpoint_name) + ) sys.exit() # Check arguments. @@ -685,29 +753,31 @@ def load_checkpoint( if not load_only_weights and not reset_train_valid_samples: assert args.consumed_train_samples == 0 assert args.consumed_valid_samples == 0 - if 'args' in state_dict and not args.finetune: - checkpoint_args = state_dict['args'] + if "args" in state_dict and not args.finetune: + checkpoint_args = state_dict["args"] check_checkpoint_args(checkpoint_args) - args.consumed_train_samples = getattr(checkpoint_args, - 'consumed_train_samples', 0) + args.consumed_train_samples = getattr( + checkpoint_args, "consumed_train_samples", 0 + ) update_num_microbatches(consumed_samples=args.consumed_train_samples) - args.consumed_valid_samples = getattr(checkpoint_args, - 'consumed_valid_samples', 0) + args.consumed_valid_samples = getattr( + checkpoint_args, "consumed_valid_samples", 0 + ) else: - print_rank_0('could not find arguments in the checkpoint ...') + print_rank_0("could not find arguments in the checkpoint ...") # Model. if not args.deepspeed: if len(model) == 1: - model[0].load_state_dict(state_dict['model'], strict=strict) + model[0].load_state_dict(state_dict["model"], strict=strict) else: for i in range(len(model)): mpu.set_virtual_pipeline_model_parallel_rank(i) - model[i].load_state_dict(state_dict['model%d' % i], strict=strict) + model[i].load_state_dict(state_dict["model%d" % i], strict=strict) # Fix up query/key/value matrix ordering if needed. checkpoint_version = get_checkpoint_version() - print_rank_0(f' checkpoint version {checkpoint_version}') + print_rank_0(f" checkpoint version {checkpoint_version}") fix_query_key_value_ordering(model, checkpoint_version) # Optimizer. @@ -716,30 +786,35 @@ def load_checkpoint( try: # Load state dict. if optimizer is not None: - optimizer.load_state_dict(state_dict['optimizer']) + optimizer.load_state_dict(state_dict["optimizer"]) # Load distributed optimizer's custom parameter state. if args.use_distributed_optimizer: tracker_filename = get_checkpoint_tracker_filename(load_dir) iteration, release = read_metadata(tracker_filename) - model_checkpoint_name = \ - get_checkpoint_name(load_dir, iteration, release) - optim_checkpoint_name = \ - get_distributed_optimizer_checkpoint_name( - model_checkpoint_name) + model_checkpoint_name = get_checkpoint_name( + load_dir, iteration, release + ) + optim_checkpoint_name = get_distributed_optimizer_checkpoint_name( + model_checkpoint_name + ) optimizer.load_parameter_state(optim_checkpoint_name) # Load scheduler. if opt_param_scheduler is not None: - if 'lr_scheduler' in state_dict: # backward compatbility - opt_param_scheduler.load_state_dict(state_dict['lr_scheduler']) + if "lr_scheduler" in state_dict: # backward compatbility + opt_param_scheduler.load_state_dict(state_dict["lr_scheduler"]) else: - opt_param_scheduler.load_state_dict(state_dict['opt_param_scheduler']) + opt_param_scheduler.load_state_dict( + state_dict["opt_param_scheduler"] + ) except KeyError: - print_rank_0('Unable to load optimizer from checkpoint {}. ' - 'Specify --no-load-optim or --finetune to prevent ' - 'attempting to load the optimizer state, ' - 'exiting ...'.format(checkpoint_name)) + print_rank_0( + "Unable to load optimizer from checkpoint {}. " + "Specify --no-load-optim or --finetune to prevent " + "attempting to load the optimizer state, " + "exiting ...".format(checkpoint_name) + ) sys.exit() else: if (args.fp16 or args.bf16) and optimizer is not None: @@ -748,36 +823,40 @@ def load_checkpoint( # rng states. if not release and not args.finetune and not args.no_load_rng: try: - if 'rng_state' in state_dict: + if "rng_state" in state_dict: # access rng_state for data parallel rank if args.data_parallel_random_init: - rng_state = state_dict['rng_state'][mpu.get_data_parallel_rank()] + rng_state = state_dict["rng_state"][mpu.get_data_parallel_rank()] else: - rng_state = state_dict['rng_state'][0] - random.setstate(rng_state['random_rng_state']) - np.random.set_state(rng_state['np_rng_state']) - torch.set_rng_state(rng_state['torch_rng_state']) - get_accelerator().set_rng_state(rng_state['cuda_rng_state']) + rng_state = state_dict["rng_state"][0] + random.setstate(rng_state["random_rng_state"]) + np.random.set_state(rng_state["np_rng_state"]) + torch.set_rng_state(rng_state["torch_rng_state"]) + get_accelerator().set_rng_state(rng_state["cuda_rng_state"]) # Check for empty states array - if not rng_state['rng_tracker_states']: + if not rng_state["rng_tracker_states"]: raise KeyError tensor_parallel.get_cuda_rng_tracker().set_states( - rng_state['rng_tracker_states']) + rng_state["rng_tracker_states"] + ) else: # backward compatability - random.setstate(state_dict['random_rng_state']) - np.random.set_state(state_dict['np_rng_state']) - torch.set_rng_state(state_dict['torch_rng_state']) - get_accelerator().set_rng_state(state_dict['cuda_rng_state']) + random.setstate(state_dict["random_rng_state"]) + np.random.set_state(state_dict["np_rng_state"]) + torch.set_rng_state(state_dict["torch_rng_state"]) + get_accelerator().set_rng_state(state_dict["cuda_rng_state"]) # Check for empty states array - if not state_dict['rng_tracker_states']: + if not state_dict["rng_tracker_states"]: raise KeyError tensor_parallel.get_cuda_rng_tracker().set_states( - state_dict['rng_tracker_states']) + state_dict["rng_tracker_states"] + ) except KeyError: - print_rank_0('Unable to load rng state from checkpoint {}. ' - 'Specify --no-load-rng or --finetune to prevent ' - 'attempting to load the rng state, ' - 'exiting ...'.format(checkpoint_name)) + print_rank_0( + "Unable to load rng state from checkpoint {}. " + "Specify --no-load-rng or --finetune to prevent " + "attempting to load the rng state, " + "exiting ...".format(checkpoint_name) + ) sys.exit() if args.universal_checkpoint: @@ -795,19 +874,29 @@ def load_checkpoint( # In the case of mp configuration change, we reconfigure the model-parallel-rng states s.t. each # tp-rank will have a unique state. In order to ensure that subsequent loads from universal will # not cause the model-parallel-rng states to be repeated, we add the iteration number to the base seed. - ckp_args = state_dict['args'] - if ((args.tensor_model_parallel_size != ckp_args.tensor_model_parallel_size) - or (args.pipeline_model_parallel_size != ckp_args.pipeline_model_parallel_size)): - print_rank_0(' loading universal checkpoint with modified mp configuration ' - '-> reconfigure tp seed') - tensor_parallel.model_parallel_reconfigure_tp_seed(args.seed + iteration) + ckp_args = state_dict["args"] + if ( + args.tensor_model_parallel_size != ckp_args.tensor_model_parallel_size + ) or ( + args.pipeline_model_parallel_size + != ckp_args.pipeline_model_parallel_size + ): + print_rank_0( + " loading universal checkpoint with modified mp configuration " + "-> reconfigure tp seed" + ) + tensor_parallel.model_parallel_reconfigure_tp_seed( + args.seed + iteration + ) # Some utilities want to load a checkpoint without distributed being initialized if torch.distributed.is_initialized(): torch.distributed.barrier() - print_rank_0(f' successfully loaded checkpoint from {args.load} ' - f'at iteration {iteration}') + print_rank_0( + f" successfully loaded checkpoint from {args.load} " + f"at iteration {iteration}" + ) # from .utils import dump_weights, dump_position_embed_weights # dump_weights(f'{args.universal_checkpoint=}', iteration, model, optimizer) @@ -815,9 +904,11 @@ def load_checkpoint( return iteration + @dlp.log -def load_biencoder_checkpoint(model, only_query_model=False, - only_context_model=False, custom_load_path=None): +def load_biencoder_checkpoint( + model, only_query_model=False, only_context_model=False, custom_load_path=None +): """ selectively load retrieval models for indexing/retrieving from saved checkpoints @@ -830,31 +921,34 @@ def load_biencoder_checkpoint(model, only_query_model=False, load_path = custom_load_path if custom_load_path is not None else args.load tracker_filename = get_checkpoint_tracker_filename(load_path) - with open(tracker_filename, 'r') as f: + with open(tracker_filename, "r") as f: iteration = int(f.read().strip()) - checkpoint_name = get_checkpoint_name(load_path, iteration, - args.use_distributed_optimizer, - release=False) + checkpoint_name = get_checkpoint_name( + load_path, iteration, args.use_distributed_optimizer, release=False + ) if mpu.get_data_parallel_rank() == 0: - print('global rank {} is loading checkpoint {}'.format( - torch.distributed.get_rank(), checkpoint_name)) + print( + "global rank {} is loading checkpoint {}".format( + torch.distributed.get_rank(), checkpoint_name + ) + ) - state_dict = torch.load(checkpoint_name, map_location='cpu') - ret_state_dict = state_dict['model'] + state_dict = torch.load(checkpoint_name, map_location="cpu") + ret_state_dict = state_dict["model"] if only_query_model: - ret_state_dict.pop('context_model') + ret_state_dict.pop("context_model") if only_context_model: - ret_state_dict.pop('query_model') + ret_state_dict.pop("query_model") assert len(model) == 1 model[0].load_state_dict(ret_state_dict) torch.distributed.barrier() if mpu.get_data_parallel_rank() == 0: - print(' successfully loaded {}'.format(checkpoint_name)) + print(" successfully loaded {}".format(checkpoint_name)) return model diff --git a/megatron/dist_signal_handler.py b/megatron/dist_signal_handler.py index a60204f004a..559daec6255 100644 --- a/megatron/dist_signal_handler.py +++ b/megatron/dist_signal_handler.py @@ -13,21 +13,20 @@ def get_world_size(): def get_device(local_rank=None): backend = torch.distributed.get_backend() - if backend == 'nccl': + if backend == "nccl": if local_rank is None: - device = torch.device('cuda') + device = torch.device("cuda") else: - device = torch.device(f'cuda:{local_rank}') - elif backend == 'gloo': - device = torch.device('cpu') + device = torch.device(f"cuda:{local_rank}") + elif backend == "gloo": + device = torch.device("cpu") else: raise RuntimeError return device def all_gather_item(item, dtype, group=None, async_op=False, local_rank=None): - if not torch.distributed.is_available() or \ - not torch.distributed.is_initialized(): + if not torch.distributed.is_available() or not torch.distributed.is_initialized(): return [item] device = get_device(local_rank) @@ -48,13 +47,12 @@ def all_gather_item(item, dtype, group=None, async_op=False, local_rank=None): class DistributedSignalHandler: + def __init__(self, sig=signal.SIGTERM): self.sig = sig def signals_received(self): - all_received = all_gather_item( - self._signal_received, dtype=torch.int32 - ) + all_received = all_gather_item(self._signal_received, dtype=torch.int32) return all_received def __enter__(self): diff --git a/megatron/enums.py b/megatron/enums.py index d9050462aaf..84c027c322e 100644 --- a/megatron/enums.py +++ b/megatron/enums.py @@ -15,19 +15,23 @@ import enum + class LayerType(enum.Enum): encoder = 1 decoder = 2 - + + class AttnType(enum.Enum): self_attn = 1 cross_attn = 2 + class AttnMaskType(enum.Enum): padding = 1 causal = 2 prefix = 3 + class PositionEmbeddingType(enum.Enum): rotary = 1 absolute = 2 diff --git a/megatron/fp16_deprecated/loss_scaler.py b/megatron/fp16_deprecated/loss_scaler.py index cb64aa92892..2541281320a 100755 --- a/megatron/fp16_deprecated/loss_scaler.py +++ b/megatron/fp16_deprecated/loss_scaler.py @@ -1,26 +1,31 @@ # Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved. - -"""For backward compatibility, we need the class definitions to deserialize.""" - -class LossScaler: - def __init__(self, scale=1): - self.cur_scale = scale - -class DynamicLossScaler: - def __init__(self, - init_scale=2**32, - scale_factor=2., - scale_window=1000, - min_scale=1, - delayed_shift=1, - consecutive_hysteresis=False): - self.cur_scale = init_scale - self.cur_iter = 0 - self.last_overflow_iter = -1 - self.scale_factor = scale_factor - self.scale_window = scale_window - self.min_scale = min_scale - self.delayed_shift = delayed_shift - self.cur_hysteresis = delayed_shift - self.consecutive_hysteresis = consecutive_hysteresis - + +"""For backward compatibility, we need the class definitions to deserialize.""" + + +class LossScaler: + + def __init__(self, scale=1): + self.cur_scale = scale + + +class DynamicLossScaler: + + def __init__( + self, + init_scale=2**32, + scale_factor=2.0, + scale_window=1000, + min_scale=1, + delayed_shift=1, + consecutive_hysteresis=False, + ): + self.cur_scale = init_scale + self.cur_iter = 0 + self.last_overflow_iter = -1 + self.scale_factor = scale_factor + self.scale_window = scale_window + self.min_scale = min_scale + self.delayed_shift = delayed_shift + self.cur_hysteresis = delayed_shift + self.consecutive_hysteresis = consecutive_hysteresis diff --git a/megatron/global_vars.py b/megatron/global_vars.py index 35680e12dfc..b86cdece65f 100644 --- a/megatron/global_vars.py +++ b/megatron/global_vars.py @@ -4,13 +4,17 @@ import os import sys -import torch from megatron import dist_signal_handler from megatron.tokenizer import build_tokenizer from .microbatches import build_num_microbatches_calculator from .timers import Timers +try: + import wandb +except Exception: + wandb = None + _GLOBAL_ARGS = None _GLOBAL_RETRO_ARGS = None _GLOBAL_NUM_MICROBATCHES_CALCULATOR = None @@ -21,9 +25,10 @@ _GLOBAL_TIMERS = None _GLOBAL_SIGNAL_HANDLER = None + def get_args(): """Return arguments.""" - _ensure_var_is_initialized(_GLOBAL_ARGS, 'args') + _ensure_var_is_initialized(_GLOBAL_ARGS, "args") return _GLOBAL_ARGS @@ -33,21 +38,23 @@ def get_retro_args(): def get_num_microbatches(): + assert _GLOBAL_NUM_MICROBATCHES_CALCULATOR is not None return _GLOBAL_NUM_MICROBATCHES_CALCULATOR.get() -def get_current_global_batch_size(): +def get_current_global_batch_size() -> int: + assert _GLOBAL_NUM_MICROBATCHES_CALCULATOR is not None return _GLOBAL_NUM_MICROBATCHES_CALCULATOR.get_current_global_batch_size() def update_num_microbatches(consumed_samples, consistency_check=True): - _GLOBAL_NUM_MICROBATCHES_CALCULATOR.update(consumed_samples, - consistency_check) + assert _GLOBAL_NUM_MICROBATCHES_CALCULATOR is not None + _GLOBAL_NUM_MICROBATCHES_CALCULATOR.update(consumed_samples, consistency_check) def get_tokenizer(): """Return tokenizer.""" - _ensure_var_is_initialized(_GLOBAL_TOKENIZER, 'tokenizer') + _ensure_var_is_initialized(_GLOBAL_TOKENIZER, "tokenizer") return _GLOBAL_TOKENIZER @@ -71,28 +78,27 @@ def get_adlr_autoresume(): def get_timers(): """Return timers.""" - _ensure_var_is_initialized(_GLOBAL_TIMERS, 'timers') + _ensure_var_is_initialized(_GLOBAL_TIMERS, "timers") return _GLOBAL_TIMERS def get_signal_handler(): - _ensure_var_is_initialized(_GLOBAL_SIGNAL_HANDLER, 'signal handler') + _ensure_var_is_initialized(_GLOBAL_SIGNAL_HANDLER, "signal handler") return _GLOBAL_SIGNAL_HANDLER def _set_signal_handler(): global _GLOBAL_SIGNAL_HANDLER - _ensure_var_is_not_initialized(_GLOBAL_SIGNAL_HANDLER, 'signal handler') + _ensure_var_is_not_initialized(_GLOBAL_SIGNAL_HANDLER, "signal handler") _GLOBAL_SIGNAL_HANDLER = dist_signal_handler.DistributedSignalHandler().__enter__() - def set_global_variables(args): """Set args, tokenizer, tensorboard-writer, adlr-autoresume, and timers.""" assert args is not None - _ensure_var_is_not_initialized(_GLOBAL_ARGS, 'args') + _ensure_var_is_not_initialized(_GLOBAL_ARGS, "args") set_args(args) _build_num_microbatches_calculator(args) @@ -117,19 +123,18 @@ def set_retro_args(retro_args): def _build_num_microbatches_calculator(args): - global _GLOBAL_NUM_MICROBATCHES_CALCULATOR - _ensure_var_is_not_initialized(_GLOBAL_NUM_MICROBATCHES_CALCULATOR, - 'num microbatches calculator') + _ensure_var_is_not_initialized( + _GLOBAL_NUM_MICROBATCHES_CALCULATOR, "num microbatches calculator" + ) - _GLOBAL_NUM_MICROBATCHES_CALCULATOR = build_num_microbatches_calculator( - args) + _GLOBAL_NUM_MICROBATCHES_CALCULATOR = build_num_microbatches_calculator(args) def _build_tokenizer(args): """Initialize tokenizer.""" global _GLOBAL_TOKENIZER - _ensure_var_is_not_initialized(_GLOBAL_TOKENIZER, 'tokenizer') + _ensure_var_is_not_initialized(_GLOBAL_TOKENIZER, "tokenizer") _GLOBAL_TOKENIZER = build_tokenizer(args) return _GLOBAL_TOKENIZER @@ -143,73 +148,90 @@ def rebuild_tokenizer(args): def _set_tensorboard_writer(args): """Set tensorboard writer.""" global _GLOBAL_TENSORBOARD_WRITER - _ensure_var_is_not_initialized(_GLOBAL_TENSORBOARD_WRITER, - 'tensorboard writer') - if hasattr(args, 'tensorboard_dir') and \ - args.tensorboard_dir and args.rank == (args.world_size - 1): + _ensure_var_is_not_initialized(_GLOBAL_TENSORBOARD_WRITER, "tensorboard writer") + if ( + hasattr(args, "tensorboard_dir") + and args.tensorboard_dir + and args.rank == (args.world_size - 1) + ): try: from torch.utils.tensorboard.writer import SummaryWriter - print('> setting tensorboard ...') + + print("> setting tensorboard ...") _GLOBAL_TENSORBOARD_WRITER = SummaryWriter( - log_dir=args.tensorboard_dir, - max_queue=args.tensorboard_queue_size) + log_dir=args.tensorboard_dir, max_queue=args.tensorboard_queue_size + ) except ModuleNotFoundError: - print('WARNING: TensorBoard writing requested but is not ' - 'available (are you using PyTorch 1.1.0 or later?), ' - 'no TensorBoard logs will be written.', flush=True) + print( + "WARNING: TensorBoard writing requested but is not " + "available (are you using PyTorch 1.1.0 or later?), " + "no TensorBoard logs will be written.", + flush=True, + ) def _set_wandb_writer(args): """Set wandb writer.""" global _GLOBAL_WANDB_WRITER - _ensure_var_is_not_initialized(_GLOBAL_WANDB_WRITER, - 'wandb writer') + _ensure_var_is_not_initialized(_GLOBAL_WANDB_WRITER, "wandb writer") if args.rank == (args.world_size - 1): - if getattr(args, 'wandb_project', '') == '' and \ - getattr(args, 'wandb_exp_name', '') == '': - print('WARNING: WANDB writing requested but no legit wandb ' - 'project or experiment name provided, ' - 'therefore no WANDB logs will be written ' - 'according to random generated project or experiment name.', flush=True) + if ( + getattr(args, "wandb_project", "") == "" + and getattr(args, "wandb_exp_name", "") == "" + ): + print( + "WARNING: WANDB writing requested but no legit wandb " + "project or experiment name provided, " + "therefore no WANDB logs will be written " + "according to random generated project or experiment name.", + flush=True, + ) return try: import wandb except (ImportError, ModuleNotFoundError): - print('WARNING: WANDB writing requested but is not ' - 'available (try to pip install wandb to solve it), ' - 'no WANDB logs will be written.', flush=True) + print( + "WARNING: WANDB writing requested but is not " + "available (try to pip install wandb to solve it), " + "no WANDB logs will be written.", + flush=True, + ) + wandb = None return + assert wandb is not None + if args.wandb_save_dir: save_dir = args.wandb_save_dir else: # Defaults to the save dir. - save_dir = os.path.join(args.save, 'wandb') + save_dir = os.path.join(args.save, "wandb") wandb_kwargs = { - 'dir': save_dir, - 'name': args.wandb_exp_name, - 'project': args.wandb_project, - 'config': vars(args)} - os.makedirs(wandb_kwargs['dir'], exist_ok=True) - wandb.init(**wandb_kwargs) + "dir": save_dir, + "name": args.wandb_exp_name, + "project": args.wandb_project, + "config": vars(args), + } + os.makedirs(wandb_kwargs["dir"], exist_ok=True) + wandb.init(**wandb_kwargs) # type:ignore _GLOBAL_WANDB_WRITER = wandb def _set_adlr_autoresume(args): """Initialize ADLR autoresume.""" global _GLOBAL_ADLR_AUTORESUME - _ensure_var_is_not_initialized(_GLOBAL_ADLR_AUTORESUME, 'adlr autoresume') + _ensure_var_is_not_initialized(_GLOBAL_ADLR_AUTORESUME, "adlr autoresume") if args.adlr_autoresume: if args.rank == 0: - print('enabling autoresume ...', flush=True) - sys.path.append(os.environ.get('SUBMIT_SCRIPTS', '.')) + print("enabling autoresume ...", flush=True) + sys.path.append(os.environ.get("SUBMIT_SCRIPTS", ".")) try: from userlib.auto_resume import AutoResume except BaseException: - print('ADLR autoresume is not available, exiting ...') + print("ADLR autoresume is not available, exiting ...") sys.exit() _GLOBAL_ADLR_AUTORESUME = AutoResume @@ -218,15 +240,15 @@ def _set_adlr_autoresume(args): def _set_timers(args): """Initialize timers.""" global _GLOBAL_TIMERS - _ensure_var_is_not_initialized(_GLOBAL_TIMERS, 'timers') + _ensure_var_is_not_initialized(_GLOBAL_TIMERS, "timers") _GLOBAL_TIMERS = Timers(args.timing_log_level, args.timing_log_option) def _ensure_var_is_initialized(var, name): """Make sure the input variable is not None.""" - assert var is not None, f'{name} is not initialized.' + assert var is not None, f"{name} is not initialized." def _ensure_var_is_not_initialized(var, name): """Make sure the input variable is not None.""" - assert var is None, f'{name} is already initialized.' + assert var is None, f"{name} is already initialized." diff --git a/megatron/indexer.py b/megatron/indexer.py index 45f530a7d4d..647537060bd 100644 --- a/megatron/indexer.py +++ b/megatron/indexer.py @@ -19,13 +19,15 @@ class IndexBuilder(object): Object for taking one pass over a dataset and creating a BlockData of its embeddings """ + def __init__(self): args = get_args() self.model = None self.dataloader = None self.evidence_embedder_obj = None - self.biencoder_shared_query_context_model = \ + self.biencoder_shared_query_context_model = ( args.biencoder_shared_query_context_model + ) # need to know whether we're using a REALM checkpoint (args.load) # or ICT checkpoint @@ -47,22 +49,24 @@ def load_attributes(self): if self.biencoder_shared_query_context_model: only_context_model = False - model = get_model(get_model_provider(only_context_model=\ - only_context_model, biencoder_shared_query_context_model=\ - self.biencoder_shared_query_context_model)) + model = get_model( + get_model_provider( + only_context_model=only_context_model, + biencoder_shared_query_context_model=self.biencoder_shared_query_context_model, + ) + ) - self.model = load_biencoder_checkpoint(model, - only_context_model=only_context_model) + self.model = load_biencoder_checkpoint( + model, only_context_model=only_context_model + ) assert len(self.model) == 1 self.model[0].eval() self.dataset = get_open_retrieval_wiki_dataset() - self.dataloader = iter(get_one_epoch_dataloader(self.dataset, \ - self.batch_size)) + self.dataloader = iter(get_one_epoch_dataloader(self.dataset, self.batch_size)) - self.evidence_embedder_obj = OpenRetreivalDataStore( \ - load_from_path=False) + self.evidence_embedder_obj = OpenRetreivalDataStore(load_from_path=False) def track_and_report_progress(self, batch_size): """ @@ -71,8 +75,12 @@ def track_and_report_progress(self, batch_size): self.iteration += 1 self.total_processed += batch_size * self.num_total_builders if self.is_main_builder and self.iteration % self.log_interval == 0: - print('Batch {:10d} | Total {:10d}'.format(self.iteration, - self.total_processed), flush=True) + print( + "Batch {:10d} | Total {:10d}".format( + self.iteration, self.total_processed + ), + flush=True, + ) def build_and_save_index(self): """ @@ -86,15 +94,19 @@ def build_and_save_index(self): assert len(self.model) == 1 unwrapped_model = self.model[0] - while not hasattr(unwrapped_model, 'embed_text'): + while not hasattr(unwrapped_model, "embed_text"): unwrapped_model = unwrapped_model.module while True: try: # batch also has query_tokens and query_pad_data - row_id, context_tokens, context_mask, context_types, \ - context_pad_mask = get_open_retrieval_batch( \ - self.dataloader) + ( + row_id, + context_tokens, + context_mask, + context_types, + context_pad_mask, + ) = get_open_retrieval_batch(self.dataloader) except (StopIteration, IndexError): break @@ -102,8 +114,11 @@ def build_and_save_index(self): # detach, separate fields and add to BlockData assert context_mask.dtype == torch.bool context_logits = unwrapped_model.embed_text( - unwrapped_model.context_model, context_tokens, context_mask, - context_types) + unwrapped_model.context_model, + context_tokens, + context_mask, + context_types, + ) context_logits = detach(context_logits) row_id = detach(row_id) @@ -121,8 +136,7 @@ def build_and_save_index(self): if self.is_main_builder: self.evidence_embedder_obj.merge_shards_and_save() # make sure that every single piece of data was embedded - assert len(self.evidence_embedder_obj.embed_data) == \ - len(self.dataset) + assert len(self.evidence_embedder_obj.embed_data) == len(self.dataset) self.evidence_embedder_obj.clear() # complete building the final copy diff --git a/megatron/initialize.py b/megatron/initialize.py index 90acf496ee0..b0ce5336df5 100644 --- a/megatron/initialize.py +++ b/megatron/initialize.py @@ -31,19 +31,24 @@ is_rocm_pytorch = OpBuilder.is_rocm_pytorch() -def initialize_megatron(extra_args_provider=None, args_defaults={}, - ignore_unknown_args=False, allow_no_cuda=False, external_args={}): +def initialize_megatron( + extra_args_provider=None, + args_defaults={}, + ignore_unknown_args=False, + allow_no_cuda=False, + external_args={}, +): """Set global variables, initialize distributed, and set autoresume and random seeds. - `allow_no_cuda` should not be set unless using megatron for cpu only - data processing. In general this arg should not be set unless you know + `allow_no_cuda` should not be set unless using megatron for cpu only + data processing. In general this arg should not be set unless you know what you are doing. - Returns a function to finalize distributed env initialization + Returns a function to finalize distributed env initialization (optionally, only when args.lazy_mpu_init == True) """ if not allow_no_cuda: # Make sure cuda is available. - assert get_accelerator().is_available(), 'Megatron requires accelerator.' + assert get_accelerator().is_available(), "Megatron requires accelerator." # Parse arguments args = parse_args(extra_args_provider, ignore_unknown_args) @@ -52,8 +57,8 @@ def initialize_megatron(extra_args_provider=None, args_defaults={}, if key in args: setattr(args, key, external_args[key]) - if args.use_checkpoint_args or args_defaults.get('use_checkpoint_args', False): - assert args.load is not None, '--use-checkpoints-args requires --load argument' + if args.use_checkpoint_args or args_defaults.get("use_checkpoint_args", False): + assert args.load is not None, "--use-checkpoints-args requires --load argument" load_args_from_checkpoint(args) validate_args(args, args_defaults) @@ -67,16 +72,16 @@ def finish_mpu_init(): args = get_args() # Pytorch distributed. _initialize_distributed() - + # Random seeds for reproducibility. if args.rank == 0: - print('> setting random seeds to {} ...'.format(args.seed)) + print("> setting random seeds to {} ...".format(args.seed)) _set_random_seed(args.seed, args.data_parallel_random_init) args = get_args() - if args.lazy_mpu_init: + if args.lazy_mpu_init: # TODO is this still a necessary option? - args.use_cpu_initialization=True + args.use_cpu_initialization = True # delayed initialization of DDP-related stuff # We only set basic DDP globals mpu.set_tensor_model_parallel_world_size(args.tensor_model_parallel_size) @@ -111,15 +116,19 @@ def _compile_dependencies(): # TODO: move this to ninja if is_rank_0(): start_time = time.time() - print('> compiling dataset index builder ...') + print("> compiling dataset index builder ...") from megatron.data.dataset_utils import compile_helper + compile_helper() - print('>>> done with dataset index builder. Compilation time: {:.3f} ' - 'seconds'.format(time.time() - start_time), flush=True) - - if not get_accelerator().device_name() == 'cuda': + print( + ">>> done with dataset index builder. Compilation time: {:.3f} " + "seconds".format(time.time() - start_time), + flush=True, + ) + + if not get_accelerator().device_name() == "cuda": print(">fused kernel is only supported in cuda, skip loading fused kernel") - return + return if args.use_dataset_only: return @@ -129,27 +138,36 @@ def _compile_dependencies(): # Custom kernel constraints check. seq_len = args.seq_length - attn_batch_size = \ - (args.num_attention_heads / args.tensor_model_parallel_size) * \ - args.micro_batch_size + attn_batch_size = ( + args.num_attention_heads / args.tensor_model_parallel_size + ) * args.micro_batch_size # Constraints on sequence length and attn_batch_size to enable warp based # optimization and upper triangular optimization (for causal mask) - custom_kernel_constraint = seq_len > 16 and seq_len <=4096 and \ - seq_len % 4 == 0 and attn_batch_size % 4 == 0 + custom_kernel_constraint = ( + seq_len > 16 + and seq_len <= 4096 + and seq_len % 4 == 0 + and attn_batch_size % 4 == 0 + ) # Print a warning. - if not ((args.fp16 or args.bf16) and - custom_kernel_constraint and - args.masked_softmax_fusion): + if not ( + (args.fp16 or args.bf16) + and custom_kernel_constraint + and args.masked_softmax_fusion + ): if args.rank == 0: - print('WARNING: constraints for invoking optimized' - ' fused softmax kernel are not met. We default' - ' back to unfused kernel invocations.', flush=True) - + print( + "WARNING: constraints for invoking optimized" + " fused softmax kernel are not met. We default" + " back to unfused kernel invocations.", + flush=True, + ) + # Always build on rank zero first. if is_rank_0(): start_time = time.time() - print('> compiling and loading fused kernels ...', flush=True) - if get_accelerator().device_count() > 0: # Skip when CPU-only + print("> compiling and loading fused kernels ...", flush=True) + if get_accelerator().device_count() > 0: # Skip when CPU-only fused_kernels.load(args) torch.distributed.barrier() else: @@ -161,13 +179,15 @@ def _compile_dependencies(): # the lock is released. torch.distributed.barrier() if is_rank_0(): - print('>>> done with compiling and loading fused kernels. ' - 'Compilation time: {:.3f} seconds'.format( - time.time() - start_time), flush=True) + print( + ">>> done with compiling and loading fused kernels. " + "Compilation time: {:.3f} seconds".format(time.time() - start_time), + flush=True, + ) def setup_deepspeed_random_and_activation_checkpointing(args): - '''Optional DeepSpeed Activation Checkpointing features. + """Optional DeepSpeed Activation Checkpointing features. Gives access to partition activations, contiguous memory optimizations and cpu checkpointing. Activation checkpoint requires keep track of the random states @@ -177,9 +197,13 @@ def setup_deepspeed_random_and_activation_checkpointing(args): Since they are used in places outside of activation checkpointing, we overwrite them to maintain consistency. This must be called before all the calls to mpu.model_parallel_cuda_manual_seed - ''' + """ num_layers = args.num_layers // args.checkpoint_num_layers - num_layers = num_layers if args.num_layers % args.checkpoint_num_layers == 0 else num_layers + 1 + num_layers = ( + num_layers + if args.num_layers % args.checkpoint_num_layers == 0 + else num_layers + 1 + ) if args.split_transformers: num_layers *= 2 @@ -191,7 +215,8 @@ def setup_deepspeed_random_and_activation_checkpointing(args): num_checkpoints=num_layers, checkpoint_in_cpu=args.checkpoint_in_cpu, synchronize=args.synchronize_each_layer, - profile=args.profile_backward) + profile=args.profile_backward, + ) def _initialize_distributed(): @@ -201,24 +226,28 @@ def _initialize_distributed(): if torch.distributed.is_initialized(): if args.rank == 0: - print('torch distributed is already initialized, ' - 'skipping initialization ...', flush=True) + print( + "torch distributed is already initialized, " + "skipping initialization ...", + flush=True, + ) args.rank = torch.distributed.get_rank() args.world_size = torch.distributed.get_world_size() else: if args.rank == 0: - print('> initializing torch distributed ...', flush=True) + print("> initializing torch distributed ...", flush=True) # Manually set the device ids. if device_count > 0: device = args.rank % device_count if args.local_rank is not None: - assert args.local_rank == device, \ - 'expected local-rank to be the same as rank % device-count.' + assert ( + args.local_rank == device + ), "expected local-rank to be the same as rank % device-count." else: args.local_rank = device - get_accelerator().set_device(device) # only do so when device_count > 0 + get_accelerator().set_device(device) # only do so when device_count > 0 if args.enable_zbh1_pipeline: deepspeed.runtime.pipe.schedule.TrainSchedule = ZeroBubbleH1Pipeline @@ -235,14 +264,16 @@ def _initialize_distributed(): if not torch.distributed.is_initialized(): torch.distributed.init_process_group( backend=get_accelerator().communication_backend_name(), - world_size=args.world_size, rank=args.rank, - timeout=timedelta(minutes=args.distributed_timeout_minutes)) + world_size=args.world_size, + rank=args.rank, + timeout=timedelta(minutes=args.distributed_timeout_minutes), + ) # Set the tensor model-parallel, pipeline model-parallel, and # data-parallel communicators. if device_count > 0: if mpu.model_parallel_is_initialized(): - print('model parallel is already initialized') + print("model parallel is already initialized") else: if args.ds_sequence_parallel_size > 1 and args.sequence_parallel: raise RuntimeError( @@ -251,17 +282,23 @@ def _initialize_distributed(): f"Remove --sequence_parallel to use DeepSpeed's sequence parallel." ) - mpu.initialize_model_parallel(args.tensor_model_parallel_size, - args.pipeline_model_parallel_size, - args.ds_sequence_parallel_size, - args.virtual_pipeline_model_parallel_size, - args.pipeline_model_parallel_split_rank, - use_distributed_optimizer=args.use_distributed_optimizer) + mpu.initialize_model_parallel( + args.tensor_model_parallel_size, + args.pipeline_model_parallel_size, + args.ds_sequence_parallel_size, + args.virtual_pipeline_model_parallel_size, + args.pipeline_model_parallel_split_rank, + use_distributed_optimizer=args.use_distributed_optimizer, + ) if args.rank == 0: - print(f'> initialized tensor model parallel with size ' - f'{mpu.get_tensor_model_parallel_world_size()}') - print(f'> initialized pipeline model parallel with size ' - f'{mpu.get_pipeline_model_parallel_world_size()}') + print( + f"> initialized tensor model parallel with size " + f"{mpu.get_tensor_model_parallel_world_size()}" + ) + print( + f"> initialized pipeline model parallel with size " + f"{mpu.get_pipeline_model_parallel_world_size()}" + ) if args.deepspeed and args.deepspeed_activation_checkpointing: setup_deepspeed_random_and_activation_checkpointing(args) @@ -294,7 +331,7 @@ def _set_random_seed(seed_, data_parallel_random_init=False): if get_accelerator().device_count() > 0: tensor_parallel.model_parallel_cuda_manual_seed(seed) else: - raise ValueError('Seed ({}) should be a positive integer.'.format(seed)) + raise ValueError("Seed ({}) should be a positive integer.".format(seed)) def write_args_to_tensorboard(): @@ -303,8 +340,7 @@ def write_args_to_tensorboard(): writer = get_tensorboard_writer() if writer: for arg in vars(args): - writer.add_text(arg, str(getattr(args, arg)), - global_step=args.iteration) + writer.add_text(arg, str(getattr(args, arg)), global_step=args.iteration) def _initialize_mem_buffs(): @@ -318,9 +354,11 @@ def _initialize_mem_buffs(): def set_jit_fusion_options(): """Set PyTorch JIT layer fusion options.""" # flags required to enable jit fusion kernels - TORCH_MAJOR = int(torch.__version__.split('.')[0]) - TORCH_MINOR = int(torch.__version__.split('.')[1]) - if ((TORCH_MAJOR > 1) or (TORCH_MAJOR == 1 and TORCH_MINOR >= 10)) and not is_rocm_pytorch: + TORCH_MAJOR = int(torch.__version__.split(".")[0]) + TORCH_MINOR = int(torch.__version__.split(".")[1]) + if ( + (TORCH_MAJOR > 1) or (TORCH_MAJOR == 1 and TORCH_MINOR >= 10) + ) and not is_rocm_pytorch: # nvfuser torch._C._jit_set_profiling_executor(True) torch._C._jit_set_profiling_mode(True) @@ -340,7 +378,7 @@ def set_jit_fusion_options(): def _warmup_jit_function(): - """ Compilie JIT functions before the main training steps """ + """Compilie JIT functions before the main training steps""" args = get_args() if args.bf16: dtype = torch.bfloat16 @@ -350,11 +388,20 @@ def _warmup_jit_function(): dtype = torch.float32 # Warmup fused bias+gelu - bias = torch.rand(args.ffn_hidden_size // args.tensor_model_parallel_size, - dtype=dtype, device='cuda') - input = torch.rand((args.seq_length // args.ds_sequence_parallel_size, args.micro_batch_size, - args.ffn_hidden_size // args.tensor_model_parallel_size), - dtype=dtype, device='cuda') + bias = torch.rand( + args.ffn_hidden_size // args.tensor_model_parallel_size, + dtype=dtype, + device="cuda", + ) + input = torch.rand( + ( + args.seq_length // args.ds_sequence_parallel_size, + args.micro_batch_size, + args.ffn_hidden_size // args.tensor_model_parallel_size, + ), + dtype=dtype, + device="cuda", + ) # Warmup JIT fusions with the input grad_enable state of both forward # prop and recomputation for bias_grad, input_grad in zip([True, True], [False, True]): @@ -368,15 +415,33 @@ def _warmup_jit_function(): seq_length = args.seq_length // mpu.get_tensor_model_parallel_world_size() else: seq_length = args.seq_length - input = torch.rand((seq_length // args.ds_sequence_parallel_size, args.micro_batch_size, args.hidden_size), - dtype=dtype, device='cuda') - residual = torch.rand((seq_length // args.ds_sequence_parallel_size, args.micro_batch_size, args.hidden_size), - dtype=dtype, device='cuda') - bias = torch.rand((args.hidden_size), dtype=dtype, device='cuda').expand_as(residual) + input = torch.rand( + ( + seq_length // args.ds_sequence_parallel_size, + args.micro_batch_size, + args.hidden_size, + ), + dtype=dtype, + device="cuda", + ) + residual = torch.rand( + ( + seq_length // args.ds_sequence_parallel_size, + args.micro_batch_size, + args.hidden_size, + ), + dtype=dtype, + device="cuda", + ) + bias = torch.rand((args.hidden_size), dtype=dtype, device="cuda").expand_as( + residual + ) dropout_rate = 0.1 # Warmup JIT fusions with the input grad_enable state of both forward # prop and recomputation - for input_grad, bias_grad, residual_grad in zip([False, True], [True, True], [True, True]): + for input_grad, bias_grad, residual_grad in zip( + [False, True], [True, True], [True, True] + ): input.requires_grad = input_grad bias.requires_grad = bias_grad residual.requires_grad = residual_grad diff --git a/megatron/memory.py b/megatron/memory.py index e63afa51f08..8813ed5a943 100644 --- a/megatron/memory.py +++ b/megatron/memory.py @@ -10,8 +10,7 @@ def allocate_mem_buff(name, numel, dtype, track_usage): """Allocate a memory buffer.""" - assert name not in _MEM_BUFFS, \ - 'memory buffer {} already allocated.'.format(name) + assert name not in _MEM_BUFFS, "memory buffer {} already allocated.".format(name) _MEM_BUFFS[name] = MemoryBuffer(name, numel, dtype, track_usage) return _MEM_BUFFS[name] @@ -33,20 +32,26 @@ class MemoryBuffer: `_start` index. """ + def __init__(self, name, numel, dtype, track_usage): if torch.distributed.get_rank() == 0: element_size = torch.tensor([], dtype=dtype).element_size() - print('> building the {} memory buffer with {} num elements ' - 'and {} dtype ({:.1f} MB)...'.format( - name, numel, dtype, numel*element_size/1024/1024), - flush=True) + print( + "> building the {} memory buffer with {} num elements " + "and {} dtype ({:.1f} MB)...".format( + name, numel, dtype, numel * element_size / 1024 / 1024 + ), + flush=True, + ) self.name = name self.numel = numel self.dtype = dtype - self.data = torch.empty(self.numel, - dtype=self.dtype, - device=get_accelerator().current_device_name(), - requires_grad=False) + self.data = torch.empty( + self.numel, + dtype=self.dtype, + device=get_accelerator().current_device_name(), + requires_grad=False, + ) # Index tracking the start of the free memory. self._start = 0 @@ -57,60 +62,60 @@ def __init__(self, name, numel, dtype, track_usage): self.in_use_value = 0.0 self.total_value = 0.0 - def reset(self): """Reset the buffer start index to the beginning of the buffer.""" self._start = 0 - def is_in_use(self): """Whether the current buffer hold on to any memory.""" return self._start > 0 - def numel_in_use(self): """Return number of elements in use.""" return self._start - def add(self, tensor): """Allocate a chunk of memory from the buffer to tensor and copy the values.""" - assert tensor.dtype == self.dtype, \ - 'Input tensor type {} different from buffer type {}'.format( - tensor.dtype, self.dtype) + assert ( + tensor.dtype == self.dtype + ), "Input tensor type {} different from buffer type {}".format( + tensor.dtype, self.dtype + ) # Number of elements of the input tensor. tensor_numel = torch.numel(tensor) new_start = self._start + tensor_numel - assert new_start <= self.numel, \ - 'Not enough memory left in the buffer ({} > {})'.format( - tensor_numel, self.numel - self._start) + assert ( + new_start <= self.numel + ), "Not enough memory left in the buffer ({} > {})".format( + tensor_numel, self.numel - self._start + ) # New tensor is a view into the memory. - new_tensor = self.data[self._start:new_start] + new_tensor = self.data[self._start : new_start] self._start = new_start new_tensor = new_tensor.view(tensor.shape) new_tensor.copy_(tensor) # Return a pointer to the new tensor. return new_tensor - def get_data(self): """Return the data currently in use.""" if self.track_usage: self.in_use_value += float(self._start) self.total_value += float(self.numel) - return self.data[:self._start] - + return self.data[: self._start] def print_average_usage(self): """Print memory usage average over time. We would like this value to be as high as possible.""" - assert self.track_usage, 'You need to enable track usage.' + assert self.track_usage, "You need to enable track usage." if torch.distributed.get_rank() == 0: - print(' > usage of {} memory buffer: {:.2f} %'.format( - self.name, self.in_use_value * 100.0 / self.total_value), - flush=True) - + print( + " > usage of {} memory buffer: {:.2f} %".format( + self.name, self.in_use_value * 100.0 / self.total_value + ), + flush=True, + ) class RingMemBuffer: @@ -119,14 +124,14 @@ class RingMemBuffer: def __init__(self, name, num_buffers, numel, dtype, track_usage): self.num_buffers = num_buffers self.buffers = [ - allocate_mem_buff(name+' {}'.format(i), numel, dtype, track_usage) - for i in range(num_buffers)] + allocate_mem_buff(name + " {}".format(i), numel, dtype, track_usage) + for i in range(num_buffers) + ] self._index = -1 - def get_next_buffer(self): self._index += 1 self._index = self._index % self.num_buffers buff = self.buffers[self._index] - assert not buff.is_in_use(), 'buffer is already in use.' + assert not buff.is_in_use(), "buffer is already in use." return buff diff --git a/megatron/microbatches.py b/megatron/microbatches.py index 6449d7479c9..0d338b7caf2 100644 --- a/megatron/microbatches.py +++ b/megatron/microbatches.py @@ -7,34 +7,48 @@ def build_num_microbatches_calculator(args): - # Constant num micro-batches. if args.rampup_batch_size is None: num_microbatches_calculator = ConstantNumMicroBatches( - args.global_batch_size, args.micro_batch_size, - args.data_parallel_size) + args.global_batch_size, args.micro_batch_size, args.data_parallel_size + ) if args.rank == 0: - print('setting number of micro-batches to constant {}'.format( - num_microbatches_calculator.get()), flush=True) + print( + "setting number of micro-batches to constant {}".format( + num_microbatches_calculator.get() + ), + flush=True, + ) else: - assert len(args.rampup_batch_size) == 3, 'expected the following ' \ - 'format: --rampup-batch-size ' \ - ' ' + assert len(args.rampup_batch_size) == 3, ( + "expected the following " + "format: --rampup-batch-size " + " " + ) start_batch_size = int(args.rampup_batch_size[0]) batch_size_increment = int(args.rampup_batch_size[1]) ramup_samples = int(args.rampup_batch_size[2]) if args.rank == 0: - print('will use batch size rampup starting from global batch ' - 'size {} to global batch size {} with batch size increments ' - '{} over {} samples.'.format(start_batch_size, - args.global_batch_size, - batch_size_increment, - ramup_samples), flush=True) + print( + "will use batch size rampup starting from global batch " + "size {} to global batch size {} with batch size increments " + "{} over {} samples.".format( + start_batch_size, + args.global_batch_size, + batch_size_increment, + ramup_samples, + ), + flush=True, + ) num_microbatches_calculator = RampupBatchsizeNumMicroBatches( - start_batch_size, batch_size_increment, ramup_samples, - args.global_batch_size, args.micro_batch_size, - args.data_parallel_size) + start_batch_size, + batch_size_increment, + ramup_samples, + args.global_batch_size, + args.micro_batch_size, + args.data_parallel_size, + ) return num_microbatches_calculator @@ -48,7 +62,7 @@ def __init__(self): def get(self): return self.num_micro_batches - def get_current_global_batch_size(self): + def get_current_global_batch_size(self) -> int: return self.current_global_batch_size @abstractmethod @@ -59,15 +73,14 @@ def update(self, consumed_samples, consistency_check): class ConstantNumMicroBatches(NumMicroBatchesCalculator): def __init__(self, global_batch_size, micro_batch_size, data_parallel_size): - micro_batch_times_data_parallel = micro_batch_size * \ - data_parallel_size - assert global_batch_size % micro_batch_times_data_parallel == 0, \ - 'global batch size ({}) is not divisible by micro batch size ({})' \ - ' times data parallel size ({})'.format(global_batch_size, - micro_batch_size, - data_parallel_size) - self.num_micro_batches = global_batch_size // \ - micro_batch_times_data_parallel + micro_batch_times_data_parallel = micro_batch_size * data_parallel_size + assert global_batch_size % micro_batch_times_data_parallel == 0, ( + "global batch size ({}) is not divisible by micro batch size ({})" + " times data parallel size ({})".format( + global_batch_size, micro_batch_size, data_parallel_size + ) + ) + self.num_micro_batches = global_batch_size // micro_batch_times_data_parallel assert self.num_micro_batches >= 1 self.current_global_batch_size = global_batch_size @@ -77,10 +90,17 @@ def update(self, consumed_samples, consistency_check): class RampupBatchsizeNumMicroBatches(NumMicroBatchesCalculator): - def __init__(self, start_batch_size, batch_size_increment, ramup_samples, - global_batch_size, micro_batch_size, data_parallel_size): + def __init__( + self, + start_batch_size, + batch_size_increment, + ramup_samples, + global_batch_size, + micro_batch_size, + data_parallel_size, + ): """Batch size ramp up. - Over + Over steps = (global-batch-size - start-batch-size) / batch_size_increment increment batch size from start-batch-size to global-batch-size using rampup-samples / steps @@ -97,10 +117,11 @@ def __init__(self, start_batch_size, batch_size_increment, ramup_samples, self.micro_batch_size = micro_batch_size self.data_parallel_size = data_parallel_size - self.micro_batch_times_data_parallel_size = self.micro_batch_size * \ - self.data_parallel_size + self.micro_batch_times_data_parallel_size = ( + self.micro_batch_size * self.data_parallel_size + ) assert self.micro_batch_times_data_parallel_size > 0 - + assert start_batch_size > 0 self.start_batch_size = start_batch_size @@ -110,9 +131,11 @@ def __init__(self, start_batch_size, batch_size_increment, ramup_samples, assert diff_batch_size >= 0 assert batch_size_increment > 0 self.batch_size_increment = batch_size_increment - assert diff_batch_size % batch_size_increment == 0, 'expected ' \ - 'global batch size interval ({}) to be divisible by global batch ' \ - 'size increment ({})'.format(diff_batch_size, batch_size_increment) + assert diff_batch_size % batch_size_increment == 0, ( + "expected " + "global batch size interval ({}) to be divisible by global batch " + "size increment ({})".format(diff_batch_size, batch_size_increment) + ) num_increments = diff_batch_size // self.batch_size_increment self.ramup_samples = ramup_samples @@ -122,23 +145,30 @@ def __init__(self, start_batch_size, batch_size_increment, ramup_samples, # Initialize number of microbatches. self.update(0, False) - def update(self, consumed_samples, consistency_check): - if consumed_samples > self.ramup_samples: self.current_global_batch_size = self.global_batch_size else: steps = int(consumed_samples / self.rampup_samples_per_increment) - self.current_global_batch_size = self.start_batch_size + \ - steps * self.batch_size_increment + self.current_global_batch_size = ( + self.start_batch_size + steps * self.batch_size_increment + ) assert self.current_global_batch_size <= self.global_batch_size if consistency_check: - assert self.current_global_batch_size % \ - self.micro_batch_times_data_parallel_size == 0, 'current global ' \ - 'batch size ({}) is not divisible by micro-batch-size ({}) times' \ - 'data parallel size ({})'.format(self.current_global_batch_size, - self.micro_batch_size, - self.data_parallel_size) - self.num_micro_batches = self.current_global_batch_size // \ - self.micro_batch_times_data_parallel_size + assert ( + self.current_global_batch_size + % self.micro_batch_times_data_parallel_size + == 0 + ), ( + "current global " + "batch size ({}) is not divisible by micro-batch-size ({}) times" + "data parallel size ({})".format( + self.current_global_batch_size, + self.micro_batch_size, + self.data_parallel_size, + ) + ) + self.num_micro_batches = ( + self.current_global_batch_size // self.micro_batch_times_data_parallel_size + ) diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 7304b257550..63e777b9501 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -6,14 +6,25 @@ from megatron import print_rank_0, get_args + class OptimizerParamScheduler(object): """Anneals learning rate and weight decay""" - def __init__(self, optimizer, max_lr, min_lr, - lr_warmup_steps, lr_decay_steps, lr_decay_style, - start_wd, end_wd, wd_incr_steps, wd_incr_style, - use_checkpoint_opt_param_scheduler=True, - override_opt_param_scheduler=False): + def __init__( + self, + optimizer, + max_lr, + min_lr, + lr_warmup_steps, + lr_decay_steps, + lr_decay_style, + start_wd, + end_wd, + wd_incr_steps, + wd_incr_style, + use_checkpoint_opt_param_scheduler=True, + override_opt_param_scheduler=False, + ): args = get_args() # Class values. self.optimizer = optimizer @@ -45,20 +56,20 @@ def __init__(self, optimizer, max_lr, min_lr, self.override_opt_param_scheduler = override_opt_param_scheduler self.use_checkpoint_opt_param_scheduler = use_checkpoint_opt_param_scheduler if self.override_opt_param_scheduler: - assert not self.use_checkpoint_opt_param_scheduler, 'both override and '\ - 'use-checkpoint are set.' + assert not self.use_checkpoint_opt_param_scheduler, ( + "both override and " "use-checkpoint are set." + ) # Set the learning rate self.step(0) - print_rank_0('> learning rate decay style: {}'.format(self.lr_decay_style)) - + print_rank_0("> learning rate decay style: {}".format(self.lr_decay_style)) def get_wd(self): - """ Weight decay incr functions""" + """Weight decay incr functions""" if self.num_steps > self.wd_incr_steps: return self.end_wd - if self.wd_incr_style == 'constant': + if self.wd_incr_style == "constant": assert self.start_wd == self.end_wd return self.end_wd @@ -67,37 +78,41 @@ def get_wd(self): assert incr_ratio <= 1.0 delta_wd = self.end_wd - self.start_wd - if self.wd_incr_style == 'linear': + if self.wd_incr_style == "linear": coeff = incr_ratio - elif self.wd_incr_style == 'cosine': + elif self.wd_incr_style == "cosine": coeff = 0.5 * (math.cos(math.pi * (1 - incr_ratio)) + 1.0) else: - raise Exception('{} weight decay increment style is not supported.'.format( - self.wd_incr_style)) + raise Exception( + "{} weight decay increment style is not supported.".format( + self.wd_incr_style + ) + ) return self.start_wd + coeff * delta_wd - def get_lr(self): """Learning rate decay functions from: - https://openreview.net/pdf?id=BJYwwY9ll pg. 4""" + https://openreview.net/pdf?id=BJYwwY9ll pg. 4""" # Use linear warmup for the initial part. if self.lr_warmup_tokens is None: if self.lr_warmup_steps > 0 and self.num_steps <= self.lr_warmup_steps: - if self.num_steps == self.lr_warmup_steps and \ - self.lr_decay_tokens is not None: + if ( + self.num_steps == self.lr_warmup_steps + and self.lr_decay_tokens is not None + ): # The case of step/sample-wise warmup + token-wise decay self.lr_warmup_tokens = self.num_tokens - return self.max_lr * float(self.num_steps) / \ - float(self.lr_warmup_steps) + return self.max_lr * float(self.num_steps) / float(self.lr_warmup_steps) else: if self.lr_warmup_tokens > 0 and self.num_tokens <= self.lr_warmup_tokens: - return self.max_lr * float(self.num_tokens) / \ - float(self.lr_warmup_tokens) + return ( + self.max_lr * float(self.num_tokens) / float(self.lr_warmup_tokens) + ) # If the learning rate is constant, just return the initial value. - if self.lr_decay_style == 'constant': + if self.lr_decay_style == "constant": return self.max_lr # For any steps larger than `self.lr_decay_steps`, use `self.min_lr`. @@ -109,15 +124,15 @@ def get_lr(self): return self.min_lr # If we are done with the warmup period, use the decay style. - if self.lr_decay_style == 'inverse-square-root': + if self.lr_decay_style == "inverse-square-root": if self.lr_warmup_tokens is None: warmup_steps = max(self.lr_warmup_steps, 1) num_steps = max(self.num_steps, 1) - lr = self.max_lr * warmup_steps ** 0.5 / (num_steps ** 0.5) + lr = self.max_lr * warmup_steps**0.5 / (num_steps**0.5) else: warmup_tokens = max(self.lr_warmup_tokens, 1) num_tokens = max(self.num_tokens, 1) - lr = self.max_lr * warmup_tokens ** 0.5 / (num_tokens ** 0.5) + lr = self.max_lr * warmup_tokens**0.5 / (num_tokens**0.5) return max(self.min_lr, lr) if self.lr_decay_tokens is None: @@ -132,17 +147,17 @@ def get_lr(self): assert decay_ratio <= 1.0 delta_lr = self.max_lr - self.min_lr - if self.lr_decay_style == 'linear': - coeff = (1.0 - decay_ratio) - elif self.lr_decay_style == 'cosine': + if self.lr_decay_style == "linear": + coeff = 1.0 - decay_ratio + elif self.lr_decay_style == "cosine": coeff = 0.5 * (math.cos(math.pi * decay_ratio) + 1.0) else: - raise Exception('{} decay style is not supported.'.format( - self.lr_decay_style)) + raise Exception( + "{} decay style is not supported.".format(self.lr_decay_style) + ) return self.min_lr + coeff * delta_lr - def step(self, increment, token_num=None): """Set lr for all parameters groups.""" if token_num is None: @@ -153,125 +168,116 @@ def step(self, increment, token_num=None): new_lr = self.get_lr() new_wd = self.get_wd() for group in self.optimizer.param_groups: - group['lr'] = new_lr * group.get('lr_mult', 1.0) - group['weight_decay'] = new_wd * group.get('wd_mult', 1.0) - + group["lr"] = new_lr * group.get("lr_mult", 1.0) + group["weight_decay"] = new_wd * group.get("wd_mult", 1.0) def state_dict(self): state_dict = { - 'max_lr': self.max_lr, - 'lr_warmup_steps': self.lr_warmup_steps, - 'lr_warmup_tokens': self.lr_warmup_tokens, - 'num_steps': self.num_steps, - 'num_tokens': self.num_tokens, - 'lr_decay_style': self.lr_decay_style, - 'lr_decay_steps': self.lr_decay_steps, - 'lr_decay_tokens': self.lr_decay_tokens, - 'min_lr': self.min_lr, - 'start_wd': self.start_wd, - 'end_wd': self.end_wd, - 'wd_incr_style': self.wd_incr_style, - 'wd_incr_steps': self.wd_incr_steps + "max_lr": self.max_lr, + "lr_warmup_steps": self.lr_warmup_steps, + "lr_warmup_tokens": self.lr_warmup_tokens, + "num_steps": self.num_steps, + "num_tokens": self.num_tokens, + "lr_decay_style": self.lr_decay_style, + "lr_decay_steps": self.lr_decay_steps, + "lr_decay_tokens": self.lr_decay_tokens, + "min_lr": self.min_lr, + "start_wd": self.start_wd, + "end_wd": self.end_wd, + "wd_incr_style": self.wd_incr_style, + "wd_incr_steps": self.wd_incr_steps, } return state_dict - def _check_and_set(self, cls_value, sd_value, name): """Auxiliary function for checking the values in the checkpoint and setting them.""" if self.override_opt_param_scheduler: - print_rank_0(' > overriding {} value to {}'.format(name, cls_value)) + print_rank_0(" > overriding {} value to {}".format(name, cls_value)) return cls_value if not self.use_checkpoint_opt_param_scheduler: - assert cls_value == sd_value, \ - f'OptimizerParamScheduler: class input value {cls_value} and checkpoint' \ - f'value {sd_value} for {name} do not match' - print_rank_0(' > using checkpoint value {} for {}'.format(sd_value, - name)) + assert cls_value == sd_value, ( + f"OptimizerParamScheduler: class input value {cls_value} and checkpoint" + f"value {sd_value} for {name} do not match" + ) + print_rank_0(" > using checkpoint value {} for {}".format(sd_value, name)) return sd_value - def load_state_dict(self, sd): - if 'start_lr' in sd: - max_lr_ = sd['start_lr'] + if "start_lr" in sd: + max_lr_ = sd["start_lr"] else: - max_lr_ = sd['max_lr'] - self.max_lr = self._check_and_set(self.max_lr, max_lr_, - 'learning rate') - - self.min_lr = self._check_and_set(self.min_lr, sd['min_lr'], - 'minimum learning rate') - - if 'warmup_iter' in sd: - lr_warmup_steps_ = sd['warmup_iter'] - elif 'warmup_steps' in sd: - lr_warmup_steps_ = sd['warmup_steps'] + max_lr_ = sd["max_lr"] + self.max_lr = self._check_and_set(self.max_lr, max_lr_, "learning rate") + + self.min_lr = self._check_and_set( + self.min_lr, sd["min_lr"], "minimum learning rate" + ) + + if "warmup_iter" in sd: + lr_warmup_steps_ = sd["warmup_iter"] + elif "warmup_steps" in sd: + lr_warmup_steps_ = sd["warmup_steps"] else: - lr_warmup_steps_ = sd['lr_warmup_steps'] - self.lr_warmup_steps = self._check_and_set(self.lr_warmup_steps, - lr_warmup_steps_, - 'warmup iterations') - if 'warmup_tokens' in sd: - lr_warmup_tokens_ = sd['warmup_tokens'] + lr_warmup_steps_ = sd["lr_warmup_steps"] + self.lr_warmup_steps = self._check_and_set( + self.lr_warmup_steps, lr_warmup_steps_, "warmup iterations" + ) + if "warmup_tokens" in sd: + lr_warmup_tokens_ = sd["warmup_tokens"] else: - lr_warmup_tokens_ = sd['lr_warmup_tokens'] - self.lr_warmup_tokens = self._check_and_set(self.lr_warmup_tokens, - lr_warmup_tokens_, - 'warmup tokens') - - if 'end_iter' in sd: - lr_decay_steps_ = sd['end_iter'] - elif 'decay_steps' in sd: - lr_decay_steps_ = sd['decay_steps'] + lr_warmup_tokens_ = sd["lr_warmup_tokens"] + self.lr_warmup_tokens = self._check_and_set( + self.lr_warmup_tokens, lr_warmup_tokens_, "warmup tokens" + ) + + if "end_iter" in sd: + lr_decay_steps_ = sd["end_iter"] + elif "decay_steps" in sd: + lr_decay_steps_ = sd["decay_steps"] else: - lr_decay_steps_ = sd['lr_decay_steps'] - self.lr_decay_steps = self._check_and_set(self.lr_decay_steps, lr_decay_steps_, - 'total number of iterations') - if 'decay_tokens' in sd: - lr_decay_tokens_ = sd['decay_tokens'] + lr_decay_steps_ = sd["lr_decay_steps"] + self.lr_decay_steps = self._check_and_set( + self.lr_decay_steps, lr_decay_steps_, "total number of iterations" + ) + if "decay_tokens" in sd: + lr_decay_tokens_ = sd["decay_tokens"] else: - lr_decay_tokens_ = sd['lr_decay_tokens'] - self.lr_decay_tokens = self._check_and_set(self.lr_decay_tokens, - lr_decay_tokens_, - 'decay tokens') + lr_decay_tokens_ = sd["lr_decay_tokens"] + self.lr_decay_tokens = self._check_and_set( + self.lr_decay_tokens, lr_decay_tokens_, "decay tokens" + ) - if 'decay_style' in sd: - lr_decay_style_ = sd['decay_style'] + if "decay_style" in sd: + lr_decay_style_ = sd["decay_style"] else: - lr_decay_style_ = sd['lr_decay_style'] - self.lr_decay_style = self._check_and_set(self.lr_decay_style, - lr_decay_style_, - 'learning rate decay style') + lr_decay_style_ = sd["lr_decay_style"] + self.lr_decay_style = self._check_and_set( + self.lr_decay_style, lr_decay_style_, "learning rate decay style" + ) - if 'num_iters' in sd: - num_steps = sd['num_iters'] + if "num_iters" in sd: + num_steps = sd["num_iters"] else: - num_steps = sd['num_steps'] - if 'num_tokens' in sd: - self.num_tokens = sd['num_tokens'] + num_steps = sd["num_steps"] + if "num_tokens" in sd: + self.num_tokens = sd["num_tokens"] self.step(increment=num_steps, token_num=self.num_tokens) - - if 'start_wd' in sd: - self.start_wd = self._check_and_set(self.start_wd, - sd['start_wd'], - "start weight decay") - self.end_wd = self._check_and_set(self.end_wd, - sd['end_wd'], - "end weight decay") - self.wd_incr_steps = self._check_and_set(self.wd_incr_steps, - sd['wd_incr_steps'], - "total number of weight decay iterations") - self.wd_incr_style = self._check_and_set(self.wd_incr_style, - sd['wd_incr_style'], - "weight decay incr style") - - - - - - - - + if "start_wd" in sd: + self.start_wd = self._check_and_set( + self.start_wd, sd["start_wd"], "start weight decay" + ) + self.end_wd = self._check_and_set( + self.end_wd, sd["end_wd"], "end weight decay" + ) + self.wd_incr_steps = self._check_and_set( + self.wd_incr_steps, + sd["wd_incr_steps"], + "total number of weight decay iterations", + ) + self.wd_incr_style = self._check_and_set( + self.wd_incr_style, sd["wd_incr_style"], "weight decay incr style" + ) diff --git a/megatron/p2p_communication.py b/megatron/p2p_communication.py index 770060a8899..d5755643f13 100644 --- a/megatron/p2p_communication.py +++ b/megatron/p2p_communication.py @@ -21,8 +21,9 @@ from megatron.core import mpu -def _communicate(tensor_send_next, tensor_send_prev, recv_prev, recv_next, - use_ring_exchange=False): +def _communicate( + tensor_send_next, tensor_send_prev, recv_prev, recv_next, use_ring_exchange=False +): """Communicate tensors between stages. Used as helper method in other communication methods that are used in megatron/schedules.py. @@ -49,23 +50,29 @@ def _communicate(tensor_send_next, tensor_send_prev, recv_prev, recv_next, tensor_recv_next = None tensor_shape = (args.seq_length, args.micro_batch_size, args.hidden_size) if args.scatter_gather_tensors_in_pipeline: - tensor_chunk_shape = reduce(operator.mul, tensor_shape, 1) // \ - mpu.get_tensor_model_parallel_world_size() + tensor_chunk_shape = ( + reduce(operator.mul, tensor_shape, 1) + // mpu.get_tensor_model_parallel_world_size() + ) else: tensor_chunk_shape = tensor_shape dtype = args.params_dtype if args.fp32_residual_connection: dtype = torch.float if recv_prev: - tensor_recv_prev = torch.empty(tensor_chunk_shape, - requires_grad=True, - device=get_accelerator().current_device_name(), - dtype=dtype) + tensor_recv_prev = torch.empty( + tensor_chunk_shape, + requires_grad=True, + device=get_accelerator().current_device_name(), + dtype=dtype, + ) if recv_next: - tensor_recv_next = torch.empty(tensor_chunk_shape, - requires_grad=True, - device=get_accelerator().current_device_name(), - dtype=dtype) + tensor_recv_next = torch.empty( + tensor_chunk_shape, + requires_grad=True, + device=get_accelerator().current_device_name(), + dtype=dtype, + ) # Split tensor into smaller chunks if using scatter-gather optimization. if args.scatter_gather_tensors_in_pipeline: @@ -77,32 +84,42 @@ def _communicate(tensor_send_next, tensor_send_prev, recv_prev, recv_next, # Send tensors in both the forward and backward directions as appropriate. if use_ring_exchange: - torch.distributed.ring_exchange(tensor_send_prev=tensor_send_prev, - tensor_recv_prev=tensor_recv_prev, - tensor_send_next=tensor_send_next, - tensor_recv_next=tensor_recv_next, - group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.ring_exchange( + tensor_send_prev=tensor_send_prev, + tensor_recv_prev=tensor_recv_prev, + tensor_send_next=tensor_send_next, + tensor_recv_next=tensor_recv_next, + group=mpu.get_pipeline_model_parallel_group(), + ) else: ops = [] if tensor_send_prev is not None: send_prev_op = torch.distributed.P2POp( - torch.distributed.isend, tensor_send_prev, - mpu.get_pipeline_model_parallel_prev_rank()) + torch.distributed.isend, + tensor_send_prev, + mpu.get_pipeline_model_parallel_prev_rank(), + ) ops.append(send_prev_op) if tensor_recv_prev is not None: recv_prev_op = torch.distributed.P2POp( - torch.distributed.irecv, tensor_recv_prev, - mpu.get_pipeline_model_parallel_prev_rank()) + torch.distributed.irecv, + tensor_recv_prev, + mpu.get_pipeline_model_parallel_prev_rank(), + ) ops.append(recv_prev_op) if tensor_send_next is not None: send_next_op = torch.distributed.P2POp( - torch.distributed.isend, tensor_send_next, - mpu.get_pipeline_model_parallel_next_rank()) + torch.distributed.isend, + tensor_send_next, + mpu.get_pipeline_model_parallel_next_rank(), + ) ops.append(send_next_op) if tensor_recv_next is not None: recv_next_op = torch.distributed.P2POp( - torch.distributed.irecv, tensor_recv_next, - mpu.get_pipeline_model_parallel_next_rank()) + torch.distributed.irecv, + tensor_recv_next, + mpu.get_pipeline_model_parallel_next_rank(), + ) ops.append(recv_next_op) if len(ops) > 0: reqs = torch.distributed.batch_isend_irecv(ops) @@ -114,12 +131,18 @@ def _communicate(tensor_send_next, tensor_send_prev, recv_prev, recv_next, # If using scatter-gather optimization, gather smaller chunks. if args.scatter_gather_tensors_in_pipeline: if recv_prev: - tensor_recv_prev = mpu.gather_split_1d_tensor( - tensor_recv_prev).view(tensor_shape).requires_grad_() + tensor_recv_prev = ( + mpu.gather_split_1d_tensor(tensor_recv_prev) + .view(tensor_shape) + .requires_grad_() + ) if recv_next: - tensor_recv_next = mpu.gather_split_1d_tensor( - tensor_recv_next).view(tensor_shape).requires_grad_() + tensor_recv_next = ( + mpu.gather_split_1d_tensor(tensor_recv_next) + .view(tensor_shape) + .requires_grad_() + ) return tensor_recv_prev, tensor_recv_next @@ -130,14 +153,15 @@ def recv_forward(timers=None): input_tensor = None else: if timers is not None: - timers('forward-recv').start() + timers("forward-recv").start() input_tensor, _ = _communicate( tensor_send_next=None, tensor_send_prev=None, recv_prev=True, - recv_next=False) + recv_next=False, + ) if timers is not None: - timers('forward-recv').stop() + timers("forward-recv").stop() return input_tensor @@ -147,14 +171,15 @@ def recv_backward(timers=None): output_tensor_grad = None else: if timers is not None: - timers('backward-recv').start() + timers("backward-recv").start() _, output_tensor_grad = _communicate( tensor_send_next=None, tensor_send_prev=None, recv_prev=False, - recv_next=True) + recv_next=True, + ) if timers is not None: - timers('backward-recv').stop() + timers("backward-recv").stop() return output_tensor_grad @@ -162,28 +187,30 @@ def send_forward(output_tensor, timers=None): """Send tensor to next rank in pipeline (forward send).""" if not mpu.is_pipeline_last_stage(): if timers is not None: - timers('forward-send').start() + timers("forward-send").start() _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, recv_prev=False, - recv_next=False) + recv_next=False, + ) if timers is not None: - timers('forward-send').stop() + timers("forward-send").stop() def send_backward(input_tensor_grad, timers=None): """Send tensor to previous rank in pipeline (backward send).""" if not mpu.is_pipeline_first_stage(): if timers is not None: - timers('backward-send').start() + timers("backward-send").start() _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, recv_prev=False, - recv_next=False) + recv_next=False, + ) if timers is not None: - timers('backward-send').stop() + timers("backward-send").stop() def send_forward_recv_backward(output_tensor, timers=None): @@ -192,14 +219,15 @@ def send_forward_recv_backward(output_tensor, timers=None): output_tensor_grad = None else: if timers is not None: - timers('forward-send-backward-recv').start() + timers("forward-send-backward-recv").start() _, output_tensor_grad = _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, recv_prev=False, - recv_next=True) + recv_next=True, + ) if timers is not None: - timers('forward-send-backward-recv').stop() + timers("forward-send-backward-recv").stop() return output_tensor_grad @@ -209,56 +237,60 @@ def send_backward_recv_forward(input_tensor_grad, timers=None): input_tensor = None else: if timers is not None: - timers('backward-send-forward-recv').start() + timers("backward-send-forward-recv").start() input_tensor, _ = _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, recv_prev=True, - recv_next=False) + recv_next=False, + ) if timers is not None: - timers('backward-send-forward-recv').stop() + timers("backward-send-forward-recv").stop() return input_tensor def send_forward_recv_forward(output_tensor, recv_prev, timers=None): """Batched recv from previous rank and send to next rank in pipeline.""" if timers is not None: - timers('forward-send-forward-recv').start() + timers("forward-send-forward-recv").start() input_tensor, _ = _communicate( tensor_send_next=output_tensor, tensor_send_prev=None, recv_prev=recv_prev, - recv_next=False) + recv_next=False, + ) if timers is not None: - timers('forward-send-forward-recv').stop() + timers("forward-send-forward-recv").stop() return input_tensor def send_backward_recv_backward(input_tensor_grad, recv_next, timers=None): """Batched recv from next rank and send to previous rank in pipeline.""" if timers is not None: - timers('backward-send-backward-recv').start() + timers("backward-send-backward-recv").start() _, output_tensor_grad = _communicate( tensor_send_next=None, tensor_send_prev=input_tensor_grad, recv_prev=False, - recv_next=recv_next) + recv_next=recv_next, + ) if timers is not None: - timers('backward-send-backward-recv').stop() + timers("backward-send-backward-recv").stop() return output_tensor_grad def send_forward_backward_recv_forward_backward( - output_tensor, input_tensor_grad, recv_prev, - recv_next, timers=None): + output_tensor, input_tensor_grad, recv_prev, recv_next, timers=None +): """Batched send and recv with previous and next ranks in pipeline.""" if timers is not None: - timers('forward-backward-send-forward-backward-recv').start() + timers("forward-backward-send-forward-backward-recv").start() input_tensor, output_tensor_grad = _communicate( tensor_send_next=output_tensor, tensor_send_prev=input_tensor_grad, recv_prev=recv_prev, - recv_next=recv_next) + recv_next=recv_next, + ) if timers is not None: - timers('forward-backward-send-forward-backward-recv').stop() + timers("forward-backward-send-forward-backward-recv").stop() return input_tensor, output_tensor_grad diff --git a/megatron/profiler.py b/megatron/profiler.py index aeab1448466..1d4b8b133df 100644 --- a/megatron/profiler.py +++ b/megatron/profiler.py @@ -5,26 +5,30 @@ on_step_begin = [] on_step_end = [] + def trigger(phase): [f() for f in phase] + def setup_profiler(args, device): if args.profile is None: return - start_step, end_step = map(int, args.profile_steps.split(',')) + start_step, end_step = map(int, args.profile_steps.split(",")) active_steps = end_step - start_step + 1 cur_step = 0 def on_step_begin_fn(): nonlocal cur_step cur_step = cur_step + 1 + on_step_begin.append(on_step_begin_fn) def when(cond, clbk): def fn(): if cond(): clbk() + return fn def is_start_step(): @@ -36,20 +40,29 @@ def is_end_step(): def is_capture_step(): return cur_step >= start_step and cur_step <= end_step - if args.profile.startswith('pt') and ( + if args.profile.startswith("pt") and ( args.profile_ranks is None or torch.distributed.get_rank() in args.profile_ranks ): - schedule = torch.profiler.schedule(wait=0, warmup=0, active=active_steps, repeat=1) + schedule = torch.profiler.schedule( + wait=0, warmup=0, active=active_steps, repeat=1 + ) activities = [torch.profiler.ProfilerActivity.CPU] - activities.extend([torch.profiler.ProfilerActivity.HPU] if device.startswith("hpu") else []) - activities.extend([torch.profiler.ProfilerActivity.CUDA] if device.startswith("cuda") else []) - full = args.profile == 'pt-full' + activities.extend( + [torch.profiler.ProfilerActivity.HPU] if device.startswith("hpu") else [] + ) + activities.extend( + [torch.profiler.ProfilerActivity.CUDA] if device.startswith("cuda") else [] + ) + full = args.profile == "pt-full" profiler = torch.profiler.profile( schedule=schedule, activities=activities, - on_trace_ready=torch.profiler.tensorboard_trace_handler(args.tensorboard_dir, use_gzip=True), - with_stack=full) + on_trace_ready=torch.profiler.tensorboard_trace_handler( + args.tensorboard_dir, use_gzip=True + ), + with_stack=full, + ) on_step_begin.append(when(is_start_step, profiler.start)) on_step_end.append(when(is_capture_step, profiler.step)) diff --git a/megatron/text_generation_server.py b/megatron/text_generation_server.py index 58550f2e631..00f34453a30 100644 --- a/megatron/text_generation_server.py +++ b/megatron/text_generation_server.py @@ -14,7 +14,9 @@ BEAM_NUM = 1 lock = threading.Lock() + class MegatronGenerate(Resource): + def __init__(self, model): self.model = model @@ -22,21 +24,21 @@ def __init__(self, model): def send_do_generate(): choice = torch.cuda.LongTensor([GENERATE_NUM]) torch.distributed.broadcast(choice, 0) - + @staticmethod def send_do_beam_search(): choice = torch.cuda.LongTensor([BEAM_NUM]) torch.distributed.broadcast(choice, 0) - + def put(self): args = get_args() - + if not "prompts" in request.get_json(): return "prompts argument required", 400 - + if "max_len" in request.get_json(): return "max_len is no longer used. Replace with tokens_to_generate", 400 - + if "sentences" in request.get_json(): return "sentences is no longer used. Replace with prompts", 400 @@ -46,35 +48,43 @@ def put(self): if len(prompts) == 0: return "prompts is empty", 400 - + if len(prompts) > 128: return "Maximum number of prompts is 128", 400 - - tokens_to_generate = 64 # Choosing hopefully sane default. Full sequence is slow + + tokens_to_generate = ( + 64 # Choosing hopefully sane default. Full sequence is slow + ) if "tokens_to_generate" in request.get_json(): tokens_to_generate = request.get_json()["tokens_to_generate"] if not isinstance(tokens_to_generate, int): return "tokens_to_generate must be an integer greater than 0" if tokens_to_generate < 0: - return "tokens_to_generate must be an integer greater than or equal to 0" + return ( + "tokens_to_generate must be an integer greater than or equal to 0" + ) logprobs = False if "logprobs" in request.get_json(): logprobs = request.get_json()["logprobs"] if not isinstance(logprobs, bool): return "logprobs must be a boolean value" - + if tokens_to_generate == 0 and not logprobs: return "tokens_to_generate=0 implies logprobs should be True" - + temperature = 1.0 if "temperature" in request.get_json(): temperature = request.get_json()["temperature"] if not (type(temperature) == int or type(temperature) == float): - return "temperature must be a positive number less than or equal to 100.0" + return ( + "temperature must be a positive number less than or equal to 100.0" + ) if not (0.0 < temperature <= 100.0): - return "temperature must be a positive number less than or equal to 100.0" - + return ( + "temperature must be a positive number less than or equal to 100.0" + ) + top_k = 0.0 if "top_k" in request.get_json(): top_k = request.get_json()["top_k"] @@ -82,7 +92,7 @@ def put(self): return "top_k must be an integer equal to or greater than 0 and less than or equal to 1000" if not (0 <= top_k <= 1000): return "top_k must be equal to or greater than 0 and less than or equal to 1000" - + top_p = 0.0 if "top_p" in request.get_json(): top_p = request.get_json()["top_p"] @@ -92,7 +102,7 @@ def put(self): return "cannot set both top-k and top-p samplings." if not (0 <= top_p <= 1.0): return "top_p must be less than or equal to 1.0" - + top_p_decay = 0.0 if "top_p_decay" in request.get_json(): top_p_decay = request.get_json()["top_p_decay"] @@ -102,23 +112,25 @@ def put(self): return "top_p_decay cannot be set without top_p" if not (0 <= top_p_decay <= 1.0): return "top_p_decay must be less than or equal to 1.0" - + top_p_bound = 0.0 if "top_p_bound" in request.get_json(): top_p_bound = request.get_json()["top_p_bound"] if not (type(top_p_bound) == float): - return "top_p_bound must be a positive float less than or equal to top_p" + return ( + "top_p_bound must be a positive float less than or equal to top_p" + ) if top_p == 0.0: return "top_p_bound cannot be set without top_p" if not (0.0 < top_p_bound <= top_p): return "top_p_bound must be greater than 0 and less than top_p" - + add_BOS = False if "add_BOS" in request.get_json(): add_BOS = request.get_json()["add_BOS"] if not isinstance(add_BOS, bool): return "add_BOS must be a boolean value" - + if any([len(prompt) == 0 for prompt in prompts]) and not add_BOS: return "Empty prompts require add_BOS=true" @@ -127,7 +139,7 @@ def put(self): stop_on_double_eol = request.get_json()["stop_on_double_eol"] if not isinstance(stop_on_double_eol, bool): return "stop_on_double_eol must be a boolean value" - + stop_on_eol = False if "stop_on_eol" in request.get_json(): stop_on_eol = request.get_json()["stop_on_eol"] @@ -136,7 +148,9 @@ def put(self): prevent_newline_after_colon = False if "prevent_newline_after_colon" in request.get_json(): - prevent_newline_after_colon = request.get_json()["prevent_newline_after_colon"] + prevent_newline_after_colon = request.get_json()[ + "prevent_newline_after_colon" + ] if not isinstance(prevent_newline_after_colon, bool): return "prevent_newline_after_colon must be a boolean value" @@ -145,7 +159,7 @@ def put(self): random_seed = request.get_json()["random_seed"] if not isinstance(random_seed, int): return "random_seed must be integer" - if random_seed < 0: + if random_seed < 0: return "random_seed must be a positive integer" no_log = False @@ -153,7 +167,7 @@ def put(self): no_log = request.get_json()["no_log"] if not isinstance(no_log, bool): return "no_log must be a boolean value" - + beam_width = None if "beam_width" in request.get_json(): beam_width = request.get_json()["beam_width"] @@ -164,78 +178,90 @@ def put(self): if len(prompts) > 1: return "When doing beam_search, batch size must be 1" - stop_token=50256 + stop_token = 50256 if "stop_token" in request.get_json(): stop_token = request.get_json()["stop_token"] if not isinstance(stop_token, int): return "stop_token must be an integer" - - length_penalty = 1 + + length_penalty = 1 if "length_penalty" in request.get_json(): length_penalty = request.get_json()["length_penalty"] if not isinstance(length_penalty, float): return "length_penalty must be a float" - + with lock: # Need to get lock to keep multiple threads from hitting code - + if not no_log: print("request IP: " + str(request.remote_addr)) - print(json.dumps(request.get_json()),flush=True) + print(json.dumps(request.get_json()), flush=True) print("start time: ", datetime.datetime.now()) - + try: if beam_width is not None: MegatronGenerate.send_do_beam_search() # Tell other ranks we're doing beam_search - response, response_seg, response_scores = \ + response, response_seg, response_scores = ( beam_search_and_post_process( - self.model, - prompts=prompts, - tokens_to_generate=tokens_to_generate, - beam_size = beam_width, - add_BOS=add_BOS, - stop_token=stop_token, - num_return_gen=beam_width, # Returning whole beam - length_penalty=length_penalty, - prevent_newline_after_colon=prevent_newline_after_colon + self.model, + prompts=prompts, + tokens_to_generate=tokens_to_generate, + beam_size=beam_width, + add_BOS=add_BOS, + stop_token=stop_token, + num_return_gen=beam_width, # Returning whole beam + length_penalty=length_penalty, + prevent_newline_after_colon=prevent_newline_after_colon, ) - - return jsonify({"text": response, - "segments": response_seg, - "scores": response_scores}) + ) + + return jsonify( + { + "text": response, + "segments": response_seg, + "scores": response_scores, + } + ) else: MegatronGenerate.send_do_generate() # Tell other ranks we're doing generate - response, response_seg, response_logprobs, _ = \ + response, response_seg, response_logprobs, _ = ( generate_and_post_process( - self.model, - prompts=prompts, - tokens_to_generate=tokens_to_generate, - return_output_log_probs=logprobs, - top_k_sampling=top_k, - top_p_sampling=top_p, - top_p_decay=top_p_decay, - top_p_bound=top_p_bound, - temperature=temperature, - add_BOS=add_BOS, - use_eod_token_for_early_termination=True, - stop_on_double_eol=stop_on_double_eol, - stop_on_eol=stop_on_eol, - prevent_newline_after_colon=prevent_newline_after_colon, - random_seed=random_seed) - - return jsonify({"text": response, - "segments": response_seg, - "logprobs": response_logprobs}) + self.model, + prompts=prompts, + tokens_to_generate=tokens_to_generate, + return_output_log_probs=logprobs, + top_k_sampling=top_k, + top_p_sampling=top_p, + top_p_decay=top_p_decay, + top_p_bound=top_p_bound, + temperature=temperature, + add_BOS=add_BOS, + use_eod_token_for_early_termination=True, + stop_on_double_eol=stop_on_double_eol, + stop_on_eol=stop_on_eol, + prevent_newline_after_colon=prevent_newline_after_colon, + random_seed=random_seed, + ) + ) + + return jsonify( + { + "text": response, + "segments": response_seg, + "logprobs": response_logprobs, + } + ) except ValueError as ve: return ve.args[0] print("end time: ", datetime.datetime.now()) - + class MegatronServer(object): + def __init__(self, model): - self.app = Flask(__name__, static_url_path='') + self.app = Flask(__name__, static_url_path="") api = Api(self.app) - api.add_resource(MegatronGenerate, '/api', resource_class_args=[model]) - - def run(self, url): + api.add_resource(MegatronGenerate, "/api", resource_class_args=[model]) + + def run(self, url): self.app.run(url, threaded=True, debug=False) diff --git a/megatron/text_generation_utils.py b/megatron/text_generation_utils.py index f5ee09dd09c..f0f2ed16dcb 100644 --- a/megatron/text_generation_utils.py +++ b/megatron/text_generation_utils.py @@ -33,29 +33,36 @@ from megatron.model import DistributedDataParallel as LocalDDP from megatron.model import Float16Module from deepspeed.accelerator import get_accelerator + + def get_batch(context_tokens): """Generate batch from context tokens.""" args = get_args() tokenizer = get_tokenizer() # Move to GPU. - tokens = context_tokens.view(args.micro_batch_size, -1).contiguous().to(get_accelerator().device_name()) + tokens = ( + context_tokens.view(args.micro_batch_size, -1) + .contiguous() + .to(get_accelerator().device_name()) + ) # Get the attention mask and postition ids. attention_mask, _, position_ids = get_ltor_masks_and_position_ids( tokens, tokenizer.eod, args.reset_position_ids, args.reset_attention_mask, - args.eod_mask_loss) + args.eod_mask_loss, + ) return tokens, attention_mask, position_ids -def top_k_logits(logits, top_k=0, top_p=0.0, filter_value=-float('Inf')): - """ This function has been mostly taken from huggingface conversational - ai code at - https://medium.com/huggingface/how-to-build-a-state-of-the-art- - conversational-ai-with-transfer-learning-2d818ac26313 """ +def top_k_logits(logits, top_k=0, top_p=0.0, filter_value=-float("Inf")): + """This function has been mostly taken from huggingface conversational + ai code at + https://medium.com/huggingface/how-to-build-a-state-of-the-art- + conversational-ai-with-transfer-learning-2d818ac26313""" if top_k > 0: # Remove all tokens with a probability less than the @@ -65,17 +72,14 @@ def top_k_logits(logits, top_k=0, top_p=0.0, filter_value=-float('Inf')): if top_p > 0.0: # Cconvert to 1D - sorted_logits, sorted_indices = torch.sort( - logits, descending=True, dim=-1) - cumulative_probs = torch.cumsum(F.softmax(sorted_logits, dim=-1), - dim=-1) + sorted_logits, sorted_indices = torch.sort(logits, descending=True, dim=-1) + cumulative_probs = torch.cumsum(F.softmax(sorted_logits, dim=-1), dim=-1) # Remove tokens with cumulative probability above the threshold sorted_indices_to_remove = cumulative_probs > top_p # Shift the indices to the right to keep also the first token # above the threshold - sorted_indices_to_remove[..., 1:] \ - = sorted_indices_to_remove[..., :-1].clone() + sorted_indices_to_remove[..., 1:] = sorted_indices_to_remove[..., :-1].clone() sorted_indices_to_remove[..., 0] = 0 for i in range(sorted_indices.size(0)): indices_to_remove = sorted_indices[i][sorted_indices_to_remove[i]] @@ -90,8 +94,7 @@ def generate_samples_input_from_file(model): tokenizer = get_tokenizer() # Read the sample file and open the output file. - assert args.sample_input_file is not None, \ - 'sample input file is not provided.' + assert args.sample_input_file is not None, "sample input file is not provided." if mpu.is_pipeline_first_stage() and mpu.get_tensor_model_parallel_rank() == 0: fname = open(args.sample_input_file, "r") all_raw_text = fname.readlines() @@ -99,8 +102,10 @@ def generate_samples_input_from_file(model): input_pos = 0 if args.sample_output_file is None: sample_output_file = args.sample_input_file + ".out" - print('`sample-output-file` not specified, setting ' - 'it to {}'.format(sample_output_file)) + print( + "`sample-output-file` not specified, setting " + "it to {}".format(sample_output_file) + ) else: sample_output_file = args.sample_output_file fname_out = open(sample_output_file, "w+") @@ -112,8 +117,10 @@ def generate_samples_input_from_file(model): terminate_runs = 0 raw_text_len = 0 - if mpu.is_pipeline_first_stage() \ - and mpu.get_tensor_model_parallel_rank() == 0: + if ( + mpu.is_pipeline_first_stage() + and mpu.get_tensor_model_parallel_rank() == 0 + ): raw_text = all_raw_text[input_pos] input_pos += 1 if input_pos == input_count: @@ -127,9 +134,13 @@ def generate_samples_input_from_file(model): context_length = len(context_tokens) if context_length >= (args.seq_length // 2): - print("\nContext length", context_length, - "\nPlease give smaller context (half of the " - "sequence length)!", flush=True) + print( + "\nContext length", + context_length, + "\nPlease give smaller context (half of the " + "sequence length)!", + flush=True, + ) continue else: context_tokens = tokenizer.tokenize("EMPTY TEXT") @@ -137,8 +148,9 @@ def generate_samples_input_from_file(model): input_info = [terminate_runs, raw_text_len, context_length] input_info_tensor = get_accelerator().LongTensor(input_info) - torch.distributed.all_reduce(input_info_tensor, - group=mpu.get_model_parallel_group()) + torch.distributed.all_reduce( + input_info_tensor, group=mpu.get_model_parallel_group() + ) terminate_runs = input_info_tensor[0].item() raw_text_len = input_info_tensor[1].item() context_length = input_info_tensor[2].item() @@ -148,8 +160,10 @@ def generate_samples_input_from_file(model): # For pipeline parallel we send context tokens to other stages # so they get the lengths correct - if mpu.get_tensor_model_parallel_rank() == 0 \ - and args.pipeline_model_parallel_size > 1: + if ( + mpu.get_tensor_model_parallel_rank() == 0 + and args.pipeline_model_parallel_size > 1 + ): if mpu.is_pipeline_first_stage(): src = mpu.get_pipeline_model_parallel_first_rank() group = mpu.get_pipeline_model_parallel_group() @@ -158,9 +172,11 @@ def generate_samples_input_from_file(model): else: src = mpu.get_pipeline_model_parallel_first_rank() group = mpu.get_pipeline_model_parallel_group() - context_tokens_tensor = torch.empty(context_length, - dtype=torch.int64, - device=get_accelerator().current_device_name()) + context_tokens_tensor = torch.empty( + context_length, + dtype=torch.int64, + device=get_accelerator().current_device_name(), + ) torch.distributed.broadcast(context_tokens_tensor, src, group) context_tokens = context_tokens_tensor.cpu().numpy().tolist() @@ -170,7 +186,7 @@ def generate_samples_input_from_file(model): if mpu.get_tensor_model_parallel_rank() == 0: if mpu.is_pipeline_first_stage(): - os.system('clear') + os.system("clear") print("\nContext:", raw_text, flush=True) fname_out.write("\nContext:") @@ -178,8 +194,9 @@ def generate_samples_input_from_file(model): decode_tokens, _ = decode_tokens decode_tokens = decode_tokens[0].cpu().numpy().tolist() - trim_decode_tokens = tokenizer.detokenize( - decode_tokens)[raw_text_len:] + trim_decode_tokens = tokenizer.detokenize(decode_tokens)[ + raw_text_len: + ] print("\nMegatron-LM:", trim_decode_tokens, flush=True) fname_out.write("\n\nMegatron-LM:") @@ -189,6 +206,7 @@ def generate_samples_input_from_file(model): raw_text = None context_count += 1 + # We added this function to support the tasks evaluation such as squad # and drop in the https://github.com/EleutherAI/lm-evaluation-harness # codebase. The lm-evaluation-harness code can now call this function @@ -215,8 +233,7 @@ def generate_samples_eval(model, context, max_gen_length, eos_token_id): decode_tokens, _ = decode_tokens decode_tokens = decode_tokens[0].cpu().numpy().tolist() - trim_decode_tokens = tokenizer.detokenize( - decode_tokens)[raw_text_len:] + trim_decode_tokens = tokenizer.detokenize(decode_tokens)[raw_text_len:] return trim_decode_tokens @@ -233,12 +250,14 @@ def generate_samples_interactive(model, print_frequency=24): terminate_runs = 0 raw_text_len = 0 - if mpu.is_pipeline_first_stage() \ - and mpu.get_tensor_model_parallel_rank() == 0: - os.system('clear') + if ( + mpu.is_pipeline_first_stage() + and mpu.get_tensor_model_parallel_rank() == 0 + ): + os.system("clear") raw_text = input("\nContext prompt (stop to exit) >>> ") while not raw_text: - print('Prompt should not be empty!') + print("Prompt should not be empty!") raw_text = input("\nContext prompt (stop to exit) >>> ") raw_text_len = len(raw_text) @@ -249,9 +268,13 @@ def generate_samples_interactive(model, print_frequency=24): context_length = len(context_tokens) if context_length >= (args.seq_length // 2): - print("\nContext length", context_length, - "\nPlease give smaller context (half of the " - "sequence length)!", flush=True) + print( + "\nContext length", + context_length, + "\nPlease give smaller context (half of the " + "sequence length)!", + flush=True, + ) continue else: context_tokens = tokenizer.tokenize("EMPTY TEXT") @@ -259,8 +282,9 @@ def generate_samples_interactive(model, print_frequency=24): input_info = [terminate_runs, raw_text_len, context_length] input_info_tensor = get_accelerator().LongTensor(input_info) - torch.distributed.all_reduce(input_info_tensor, - group=mpu.get_model_parallel_group()) + torch.distributed.all_reduce( + input_info_tensor, group=mpu.get_model_parallel_group() + ) terminate_runs = input_info_tensor[0].item() raw_text_len = input_info_tensor[1].item() context_length = input_info_tensor[2].item() @@ -270,8 +294,10 @@ def generate_samples_interactive(model, print_frequency=24): # For pipeline parallel we send context tokens to other stages # so they get the lengths correct - if mpu.get_tensor_model_parallel_rank() == 0 \ - and args.pipeline_model_parallel_size > 1: + if ( + mpu.get_tensor_model_parallel_rank() == 0 + and args.pipeline_model_parallel_size > 1 + ): if mpu.is_pipeline_first_stage(): src = mpu.get_pipeline_model_parallel_first_rank() group = mpu.get_pipeline_model_parallel_group() @@ -280,39 +306,43 @@ def generate_samples_interactive(model, print_frequency=24): else: src = mpu.get_pipeline_model_parallel_first_rank() group = mpu.get_pipeline_model_parallel_group() - context_tokens_tensor = torch.empty(context_length, - dtype=torch.int64, - device=torch.device(get_accelerator().device_name())) + context_tokens_tensor = torch.empty( + context_length, + dtype=torch.int64, + device=torch.device(get_accelerator().device_name()), + ) torch.distributed.broadcast(context_tokens_tensor, src, group) context_tokens = context_tokens_tensor.cpu().numpy().tolist() token_stream = get_token_stream(model, [context_tokens]) for counter, decode_tokens in enumerate(token_stream): - if counter % print_frequency != 0 \ - or mpu.get_tensor_model_parallel_rank() != 0 \ - or not mpu.is_pipeline_first_stage(): + if ( + counter % print_frequency != 0 + or mpu.get_tensor_model_parallel_rank() != 0 + or not mpu.is_pipeline_first_stage() + ): continue - os.system('clear') + os.system("clear") print("\nContext:", raw_text, flush=True) decode_tokens, _ = decode_tokens decode_tokens = decode_tokens[0].cpu().numpy().tolist() - trim_decode_tokens = tokenizer.detokenize( - decode_tokens)[raw_text_len:] + trim_decode_tokens = tokenizer.detokenize(decode_tokens)[raw_text_len:] print("\nMegatron-LM:", trim_decode_tokens, flush=True) - if mpu.is_pipeline_first_stage() \ - and mpu.get_tensor_model_parallel_rank() == 0: - os.system('clear') + if ( + mpu.is_pipeline_first_stage() + and mpu.get_tensor_model_parallel_rank() == 0 + ): + os.system("clear") print("\nContext:", raw_text, flush=True) if not isinstance(decode_tokens, list): decode_tokens, _ = decode_tokens decode_tokens = decode_tokens[0].cpu().numpy().tolist() - trim_decode_tokens = tokenizer.detokenize( - decode_tokens)[raw_text_len:] + trim_decode_tokens = tokenizer.detokenize(decode_tokens)[raw_text_len:] print("\nMegatron-LM:", trim_decode_tokens, flush=True) input("\nPress Enter to continue >>>") @@ -321,28 +351,31 @@ def generate_samples_interactive(model, print_frequency=24): context_count += 1 - -def generate_samples_unconditional(model, latencies=[], model_latencies=[], single_token_latency=[]): +def generate_samples_unconditional( + model, latencies=[], model_latencies=[], single_token_latency=[] +): args = get_args() tokenizer = get_tokenizer() num_samples = args.num_samples - context_tokens = [[tokenizer.eod] - for _ in range(args.micro_batch_size)] + context_tokens = [[tokenizer.eod] for _ in range(args.micro_batch_size)] ctr = 0 while True: get_accelerator().synchronize() start_time = time.time() - for token_stream in get_token_stream(model, - copy.deepcopy(context_tokens), model_latencies=model_latencies, single_token_latency=single_token_latency): + for token_stream in get_token_stream( + model, + copy.deepcopy(context_tokens), + model_latencies=model_latencies, + single_token_latency=single_token_latency, + ): pass get_accelerator().synchronize() latencies.append(time.time() - start_time) start_time = time.time() - if mpu.is_pipeline_last_stage() and \ - mpu.get_tensor_model_parallel_rank() == 0: - #if ctr % args.log_interval == 0: + if mpu.is_pipeline_last_stage() and mpu.get_tensor_model_parallel_rank() == 0: + # if ctr % args.log_interval == 0: # print('Avg s/batch:', # (time.time() - start_time) / min(args.log_interval, ctr + 1)) # start_time = time.time() @@ -351,10 +384,10 @@ def generate_samples_unconditional(model, latencies=[], model_latencies=[], sing length_batch = token_stream[1].cpu().numpy().tolist() assert len(length_batch) == args.micro_batch_size for tokens, length in zip(token_batch, length_batch): - tokens = tokens[1:length - 1] + tokens = tokens[1 : length - 1] text = tokenizer.detokenize(tokens) is_finished = length < args.seq_length - 1 - datum = {'text': text, 'length': length - 1, 'finished': is_finished} + datum = {"text": text, "length": length - 1, "finished": is_finished} yield datum ctr += 1 if ctr >= num_samples: @@ -369,15 +402,24 @@ def generate_samples_unconditional(model, latencies=[], model_latencies=[], sing break -def generate_and_write_samples_unconditional(model, latencies=[], single_token_latency=[], model_latencies=[]): +def generate_and_write_samples_unconditional( + model, latencies=[], single_token_latency=[], model_latencies=[] +): args = get_args() assert args.genfile is not None - with open(args.genfile, 'w') as f: - for datum in generate_samples_unconditional(model, latencies=latencies, model_latencies=model_latencies, single_token_latency=single_token_latency): - if mpu.is_pipeline_last_stage() and \ - mpu.get_tensor_model_parallel_rank() == 0: - f.write(json.dumps(datum) + '\n') + with open(args.genfile, "w") as f: + for datum in generate_samples_unconditional( + model, + latencies=latencies, + model_latencies=model_latencies, + single_token_latency=single_token_latency, + ): + if ( + mpu.is_pipeline_last_stage() + and mpu.get_tensor_model_parallel_rank() == 0 + ): + f.write(json.dumps(datum) + "\n") def pad_batch(batch, pad_id, args): @@ -391,42 +433,52 @@ def pad_batch(batch, pad_id, args): return batch, context_lengths -def get_token_stream(model, context_tokens, model_latencies=[], single_token_latency=[]): +def get_token_stream( + model, context_tokens, model_latencies=[], single_token_latency=[] +): args = get_args() tokenizer = get_tokenizer() - context_tokens, context_lengths = pad_batch(context_tokens, - tokenizer.eod, args) + context_tokens, context_lengths = pad_batch(context_tokens, tokenizer.eod, args) context_tokens_tensor = get_accelerator().LongTensor(context_tokens) context_length_tensor = get_accelerator().LongTensor(context_lengths) - torch.distributed.broadcast(context_length_tensor, - mpu.get_tensor_model_parallel_src_rank(), - group=mpu.get_tensor_model_parallel_group()) - torch.distributed.broadcast(context_tokens_tensor, - mpu.get_tensor_model_parallel_src_rank(), - group=mpu.get_tensor_model_parallel_group()) + torch.distributed.broadcast( + context_length_tensor, + mpu.get_tensor_model_parallel_src_rank(), + group=mpu.get_tensor_model_parallel_group(), + ) + torch.distributed.broadcast( + context_tokens_tensor, + mpu.get_tensor_model_parallel_src_rank(), + group=mpu.get_tensor_model_parallel_group(), + ) context_length = context_length_tensor.min().item() tokens, attention_mask, position_ids = get_batch(context_tokens_tensor) - batch_token_iterator = sample_sequence_batch(model, context_tokens_tensor, - context_length_tensor, - attention_mask, position_ids, model_latencies=model_latencies) + batch_token_iterator = sample_sequence_batch( + model, + context_tokens_tensor, + context_length_tensor, + attention_mask, + position_ids, + model_latencies=model_latencies, + ) count = 0 - t0=time.time() + t0 = time.time() for tokens, lengths in batch_token_iterator: if count > 1: - get_accelerator().synchronize() - t_elapsed = time.time() - t0 - single_token_latency.append(t_elapsed) + get_accelerator().synchronize() + t_elapsed = time.time() - t0 + single_token_latency.append(t_elapsed) get_accelerator().synchronize() - t0=time.time() - count+=1 + t0 = time.time() + count += 1 context_length += 1 if tokens is not None: yield tokens[:, :context_length], lengths @@ -440,9 +492,17 @@ def switch(val1, val2, boolean): return (1 - boolean) * val1 + boolean * val2 -def forward_step(model, tokens, position_ids, attention_mask, tokentype_ids, - layer_past=None, get_key_value=None, - forward_method_parallel_output=None, model_latencies=[]): +def forward_step( + model, + tokens, + position_ids, + attention_mask, + tokentype_ids, + layer_past=None, + get_key_value=None, + forward_method_parallel_output=None, + model_latencies=[], +): # Hidden size changes when not using recompute, need to tell p2p_communicate # functions the correct size @@ -455,19 +515,22 @@ def forward_step(model, tokens, position_ids, attention_mask, tokentype_ids, input_tensor = recv_forward() # Forward pass through the model. - unwrapped_model = unwrap_model( - model, (torchDDP, LocalDDP, Float16Module)) + unwrapped_model = unwrap_model(model, (torchDDP, LocalDDP, Float16Module)) - if hasattr(unwrapped_model, 'set_input_tensor'): + if hasattr(unwrapped_model, "set_input_tensor"): unwrapped_model.set_input_tensor(input_tensor) elif args.deepspeed or args.ds_inference: unwrapped_model.module.set_input_tensor(input_tensor) - output_tensor = model(tokens, position_ids, attention_mask, - tokentype_ids=tokentype_ids, - layer_past=layer_past, - get_key_value=get_key_value, - forward_method_parallel_output=forward_method_parallel_output) + output_tensor = model( + tokens, + position_ids, + attention_mask, + tokentype_ids=tokentype_ids, + layer_past=layer_past, + get_key_value=get_key_value, + forward_method_parallel_output=forward_method_parallel_output, + ) if get_key_value: output_tensor, layer_past = output_tensor @@ -476,15 +539,22 @@ def forward_step(model, tokens, position_ids, attention_mask, tokentype_ids, args.seq_length = orig_seq_length get_accelerator().synchronize() - model_latencies.append(time.time()-t0) + model_latencies.append(time.time() - t0) if get_key_value: return output_tensor, layer_past return output_tensor -def sample_sequence_batch(model, context_tokens, context_lengths, - attention_mask, position_ids, - maxlen=None, type_ids=None, model_latencies=[]): +def sample_sequence_batch( + model, + context_tokens, + context_lengths, + attention_mask, + position_ids, + maxlen=None, + type_ids=None, + model_latencies=[], +): args = get_args() tokenizer = get_tokenizer() @@ -495,7 +565,7 @@ def sample_sequence_batch(model, context_tokens, context_lengths, # added eos_id to support the function generate_samples_eval that passes # eos_id as an argument and needs termination when that id id found. - if hasattr(args, 'eos_id'): + if hasattr(args, "eos_id"): eos_id = args.eos_id else: eos_id = tokenizer.eod @@ -512,15 +582,20 @@ def sample_sequence_batch(model, context_tokens, context_lengths, if maxlen > (org_context_length + args.out_seq_length): maxlen = org_context_length + args.out_seq_length - lengths = torch.ones([batch_size]).long().to(get_accelerator().device_name()) * maxlen + lengths = ( + torch.ones([batch_size]).long().to(get_accelerator().device_name()) * maxlen + ) while context_length <= (maxlen): if args.recompute: - output = forward_step(model, tokens, - position_ids, - attention_mask, - tokentype_ids=type_ids, - forward_method_parallel_output=False) + output = forward_step( + model, + tokens, + position_ids, + attention_mask, + tokentype_ids=type_ids, + forward_method_parallel_output=False, + ) if mpu.is_pipeline_last_stage(): assert output is not None logits = output[:, context_length - 1, :] @@ -532,20 +607,23 @@ def sample_sequence_batch(model, context_tokens, context_lengths, if type_ids is not None: types2use = type_ids[:, :context_length] else: - tokens2use = tokens[:, context_length - 1].view( - batch_size, -1) + tokens2use = tokens[:, context_length - 1].view(batch_size, -1) positions2use = position_ids[:, context_length - 1].view( - batch_size, -1) + batch_size, -1 + ) if type_ids is not None: - types2use = type_ids[:, context_length - 1].view( - batch_size, -1) - output, layer_past = forward_step(model, tokens2use, - positions2use, - attention_mask, - layer_past=layer_past, - get_key_value=True, - tokentype_ids=types2use, - forward_method_parallel_output=False, model_latencies=model_latencies) + types2use = type_ids[:, context_length - 1].view(batch_size, -1) + output, layer_past = forward_step( + model, + tokens2use, + positions2use, + attention_mask, + layer_past=layer_past, + get_key_value=True, + tokentype_ids=types2use, + forward_method_parallel_output=False, + model_latencies=model_latencies, + ) if mpu.is_pipeline_last_stage(): assert output is not None logits = output[:, -1].view(batch_size, -1).contiguous() @@ -556,15 +634,13 @@ def sample_sequence_batch(model, context_tokens, context_lengths, else: logits = logits.float() logits /= args.temperature - logits = top_k_logits(logits, top_k=args.top_k, - top_p=args.top_p) + logits = top_k_logits(logits, top_k=args.top_k, top_p=args.top_p) log_probs = F.softmax(logits, dim=-1) prev = torch.multinomial(log_probs, num_samples=1).view(-1) started = context_lengths <= context_length - new_tokens = switch( - tokens[:, context_length].view(-1), prev, started) + new_tokens = switch(tokens[:, context_length].view(-1), prev, started) tokens[:, context_length] = new_tokens src = mpu.get_pipeline_model_parallel_last_rank() group = mpu.get_embedding_group() diff --git a/megatron/training.py b/megatron/training.py index 27e791a6a8b..2a75b3317a0 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -422,9 +422,9 @@ def get_model( mpu.get_pipeline_model_parallel_world_size() > 1 and args.virtual_pipeline_model_parallel_size is not None ): - assert model_type != ModelType.encoder_and_decoder, ( - "Interleaved schedule not supported for model with both encoder and decoder" - ) + assert ( + model_type != ModelType.encoder_and_decoder + ), "Interleaved schedule not supported for model with both encoder and decoder" model = [] for i in range(args.virtual_pipeline_model_parallel_size): mpu.set_virtual_pipeline_model_parallel_rank(i) @@ -443,9 +443,9 @@ def get_model( add_decoder = True if model_type == ModelType.encoder_and_decoder: if mpu.get_pipeline_model_parallel_world_size() > 1: - assert args.pipeline_model_parallel_split_rank is not None, ( - "Split rank needs to be specified for model with both encoder and decoder" - ) + assert ( + args.pipeline_model_parallel_split_rank is not None + ), "Split rank needs to be specified for model with both encoder and decoder" rank = mpu.get_pipeline_model_parallel_rank() split_rank = args.pipeline_model_parallel_split_rank world_size = mpu.get_pipeline_model_parallel_world_size() @@ -471,9 +471,9 @@ def get_model( # Disallow training and inference with Transformer Engine # for non-GPT models args.allow_transformer_engine = all([type(m) == GPTModel for m in model]) - assert args.allow_transformer_engine or args.transformer_impl == "local", ( - "Transformer Engine is only approved for GPT models" - ) + assert ( + args.allow_transformer_engine or args.transformer_impl == "local" + ), "Transformer Engine is only approved for GPT models" # Set tensor model parallel attributes if not set. # Only parameters that are already tensor model parallel have these @@ -623,9 +623,9 @@ def load_model_weights_only(model_provider_func): model=model[0], config=args.deepspeed_config_dict ) - assert not isinstance(model, deepspeed.PipelineEngine), ( - "Weight loading only mode is not supported in pipeline parallelism yet." - ) + assert not isinstance( + model, deepspeed.PipelineEngine + ), "Weight loading only mode is not supported in pipeline parallelism yet." model = [model] print_datetime("before load checkpoint") if args.load is not None: @@ -1514,9 +1514,9 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider log.info("> building train, validation, and test datasets ...") # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: - assert args.train_samples is None, ( - "only backward compatiblity support for iteration-based training" - ) + assert ( + args.train_samples is None + ), "only backward compatiblity support for iteration-based training" args.consumed_train_samples = args.iteration * args.global_batch_size if args.iteration > 0 and args.consumed_valid_samples == 0: if args.train_samples is None: diff --git a/megatron/utils.py b/megatron/utils.py index dc1dea0b3a0..dcda34e24a6 100644 --- a/megatron/utils.py +++ b/megatron/utils.py @@ -63,6 +63,7 @@ from functools import wraps class Profile(object): + def __init__( self, cat, name=None, epoch=None, step=None, image_idx=None, image_size=None ): @@ -98,6 +99,7 @@ def log_static(self, func): return class dftracer(object): + def __init__( self, ): @@ -479,7 +481,7 @@ def throughput_calculator(model, args, iteration_time, total_iterations): # correction has been made to TFLOPs formula due to incorrect behavior # observed with selective recompute when GQA not used and for all with GQA seq_len = args.seq_length - if hasattr(args, 'actual_seq_length'): + if hasattr(args, "actual_seq_length"): seq_len = args.actual_seq_length pre_and_post_mha_gemm_macs = ( batch_size * num_layers * (1 + (2 // gqa) + 1) * (hidden_size**2) * seq_len From 3a6ffec683362bdbf57019d47af177878d70ec84 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 18 Jun 2025 10:16:19 -0500 Subject: [PATCH 440/590] chore: Format `megatron/fused_kernels/*.py` --- megatron/fused_kernels/__init__.py | 98 +++++++++++-------- .../fused_kernels/tests/test_fused_kernels.py | 75 ++++++++++---- 2 files changed, 115 insertions(+), 58 deletions(-) diff --git a/megatron/fused_kernels/__init__.py b/megatron/fused_kernels/__init__.py index 6886c17e8b6..f96354483a0 100644 --- a/megatron/fused_kernels/__init__.py +++ b/megatron/fused_kernels/__init__.py @@ -21,45 +21,52 @@ def load(args): cc_flag = [] if torch.version.hip is None: _, bare_metal_major, bare_metal_minor = _get_cuda_bare_metal_version( - cpp_extension.CUDA_HOME) + cpp_extension.CUDA_HOME + ) if int(bare_metal_major) >= 11: - cc_flag.append('-gencode') - cc_flag.append('arch=compute_80,code=sm_80') + cc_flag.append("-gencode") + cc_flag.append("arch=compute_80,code=sm_80") if int(bare_metal_minor) >= 1: - cc_flag.append('-gencode') - cc_flag.append('arch=compute_86,code=sm_86') + cc_flag.append("-gencode") + cc_flag.append("arch=compute_86,code=sm_86") if int(bare_metal_minor) >= 4: - cc_flag.append('-gencode') - cc_flag.append('arch=compute_87,code=sm_87') + cc_flag.append("-gencode") + cc_flag.append("arch=compute_87,code=sm_87") if int(bare_metal_minor) >= 8: - cc_flag.append('-gencode') - cc_flag.append('arch=compute_89,code=sm_89') + cc_flag.append("-gencode") + cc_flag.append("arch=compute_89,code=sm_89") if int(bare_metal_major) >= 12: - cc_flag.append('-gencode') - cc_flag.append('arch=compute_90,code=sm_90') + cc_flag.append("-gencode") + cc_flag.append("arch=compute_90,code=sm_90") # Build path srcpath = pathlib.Path(__file__).parent.absolute() - buildpath = srcpath / 'build' + buildpath = srcpath / "build" _create_build_dir(buildpath) # Helper function to build the kernels. - def _cpp_extention_load_helper(name, sources, extra_cuda_flags, extra_include_paths): + def _cpp_extention_load_helper( + name, sources, extra_cuda_flags, extra_include_paths + ): if torch.version.hip is not None: - extra_cuda_cflags=['-O3'] + extra_cuda_flags + cc_flag + extra_cuda_cflags = ["-O3"] + extra_cuda_flags + cc_flag else: - extra_cuda_cflags=['-O3', - '-gencode', 'arch=compute_70,code=sm_70', - '--use_fast_math'] + extra_cuda_flags + cc_flag + extra_cuda_cflags = ( + ["-O3", "-gencode", "arch=compute_70,code=sm_70", "--use_fast_math"] + + extra_cuda_flags + + cc_flag + ) return cpp_extension.load( name=name, sources=sources, build_directory=buildpath, - extra_cflags=['-O3',], + extra_cflags=[ + "-O3", + ], extra_cuda_cflags=extra_cuda_cflags, extra_include_paths=extra_include_paths, - verbose=(args.rank == 0) + verbose=(args.rank == 0), ) # ============== @@ -67,43 +74,56 @@ def _cpp_extention_load_helper(name, sources, extra_cuda_flags, extra_include_pa # ============== if torch.version.hip is not None: - extra_include_paths=[os.path.abspath(srcpath)] + extra_include_paths = [os.path.abspath(srcpath)] else: - extra_include_paths=[] + extra_include_paths = [] if args.masked_softmax_fusion: if torch.version.hip is not None: - extra_cuda_flags = ['-D__HIP_NO_HALF_OPERATORS__=1', - '-D__HIP_NO_HALF_CONVERSIONS__=1'] + extra_cuda_flags = [ + "-D__HIP_NO_HALF_OPERATORS__=1", + "-D__HIP_NO_HALF_CONVERSIONS__=1", + ] else: - extra_cuda_flags = ['-U__CUDA_NO_HALF_OPERATORS__', - '-U__CUDA_NO_HALF_CONVERSIONS__', - '--expt-relaxed-constexpr', - '--expt-extended-lambda'] - + extra_cuda_flags = [ + "-U__CUDA_NO_HALF_OPERATORS__", + "-U__CUDA_NO_HALF_CONVERSIONS__", + "--expt-relaxed-constexpr", + "--expt-extended-lambda", + ] + # Upper triangular softmax. - sources=[srcpath / 'scaled_upper_triang_masked_softmax.cpp', - srcpath / 'scaled_upper_triang_masked_softmax_cuda.cu'] + sources = [ + srcpath / "scaled_upper_triang_masked_softmax.cpp", + srcpath / "scaled_upper_triang_masked_softmax_cuda.cu", + ] scaled_upper_triang_masked_softmax_cuda = _cpp_extention_load_helper( "scaled_upper_triang_masked_softmax_cuda", - sources, extra_cuda_flags, extra_include_paths) + sources, + extra_cuda_flags, + extra_include_paths, + ) # Masked softmax. - sources=[srcpath / 'scaled_masked_softmax.cpp', - srcpath / 'scaled_masked_softmax_cuda.cu'] + sources = [ + srcpath / "scaled_masked_softmax.cpp", + srcpath / "scaled_masked_softmax_cuda.cu", + ] scaled_masked_softmax_cuda = _cpp_extention_load_helper( - "scaled_masked_softmax_cuda", sources, extra_cuda_flags, extra_include_paths) + "scaled_masked_softmax_cuda", sources, extra_cuda_flags, extra_include_paths + ) # Softmax - sources=[srcpath / 'scaled_softmax.cpp', - srcpath / 'scaled_softmax_cuda.cu'] + sources = [srcpath / "scaled_softmax.cpp", srcpath / "scaled_softmax_cuda.cu"] scaled_softmax_cuda = _cpp_extention_load_helper( - "scaled_softmax_cuda", sources, extra_cuda_flags, extra_include_paths) + "scaled_softmax_cuda", sources, extra_cuda_flags, extra_include_paths + ) def _get_cuda_bare_metal_version(cuda_dir): - raw_output = subprocess.check_output([cuda_dir + "/bin/nvcc", "-V"], - universal_newlines=True) + raw_output = subprocess.check_output( + [cuda_dir + "/bin/nvcc", "-V"], universal_newlines=True + ) output = raw_output.split() release_idx = output.index("release") + 1 release = output[release_idx].split(".") diff --git a/megatron/fused_kernels/tests/test_fused_kernels.py b/megatron/fused_kernels/tests/test_fused_kernels.py index 74024c5020f..f1b2e02d267 100644 --- a/megatron/fused_kernels/tests/test_fused_kernels.py +++ b/megatron/fused_kernels/tests/test_fused_kernels.py @@ -9,6 +9,7 @@ from megatron.model.utils import attention_mask_func from megatron.fused_kernels import load + def test_load_fused_kernels(): try: import fused_layer_norm_cuda @@ -21,6 +22,7 @@ def test_load_fused_kernels(): print("[Fail] load_fused_kernels") raise e + def test_fused_softmax(): bert = BertModel.from_pretrained("bert-base-cased").cuda().half() tokenizer = BertTokenizer.from_pretrained("bert-base-cased") @@ -298,13 +300,22 @@ def test_masked_softmax_forward(): scale_t = torch.tensor([1.0]) for qlen in [128, 256, 1024, 2048, 4096]: for klen in [128, 256, 1024, 2048]: - inputs = torch.normal(0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device='cuda:0') - masks = torch.randint(0, 2, (batch, 1, qlen, klen), dtype=torch.bool, device='cuda:0') - softmax_results = scaled_masked_softmax_cuda.forward(inputs, masks, scale_t[0].item()) - softmax_results_torch = forward_torch_softmax(inputs, masks, scale_t[0].item()) + inputs = torch.normal( + 0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device="cuda:0" + ) + masks = torch.randint( + 0, 2, (batch, 1, qlen, klen), dtype=torch.bool, device="cuda:0" + ) + softmax_results = scaled_masked_softmax_cuda.forward( + inputs, masks, scale_t[0].item() + ) + softmax_results_torch = forward_torch_softmax( + inputs, masks, scale_t[0].item() + ) error = (softmax_results_torch - softmax_results).abs().max() assert error < 1e-3 + def test_masked_softmax_backward(): import scaled_masked_softmax_cuda @@ -313,14 +324,24 @@ def test_masked_softmax_backward(): scale_t = torch.tensor([1.0]) for qlen in [128, 256, 1024, 2048, 4096]: for klen in [128, 256, 1024, 2048]: - inputs = torch.normal(0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device='cuda:0') - backward = torch.rand_like(inputs, dtype=torch.float16, device='cuda:0') - masks = torch.randint(0, 2, (batch, 1, qlen, klen), dtype=torch.bool, device='cuda:0') - softmax_results = scaled_masked_softmax_cuda.forward(inputs, masks, scale_t[0].item()) - back_grad = scaled_masked_softmax_cuda.backward(backward, softmax_results, scale_t[0].item()) + inputs = torch.normal( + 0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device="cuda:0" + ) + backward = torch.rand_like(inputs, dtype=torch.float16, device="cuda:0") + masks = torch.randint( + 0, 2, (batch, 1, qlen, klen), dtype=torch.bool, device="cuda:0" + ) + softmax_results = scaled_masked_softmax_cuda.forward( + inputs, masks, scale_t[0].item() + ) + back_grad = scaled_masked_softmax_cuda.backward( + backward, softmax_results, scale_t[0].item() + ) inputs.requires_grad = True - softmax_results_torch = forward_torch_softmax(inputs, masks, scale_t[0].item()) + softmax_results_torch = forward_torch_softmax( + inputs, masks, scale_t[0].item() + ) softmax_results_torch.backward(backward) error = (back_grad - inputs.grad).abs().max() assert error < 1e-3 @@ -334,9 +355,15 @@ def test_allmasked_softmax_forward(): scale_t = torch.tensor([1.0]) for qlen in [128, 256, 1024, 2048, 4096]: for klen in [128, 256, 1024, 2048]: - inputs = torch.normal(0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device='cuda:0') - masks = torch.ones((batch, 1, qlen, klen), dtype=torch.bool, device='cuda:0') - softmax_results = scaled_masked_softmax_cuda.forward(inputs, masks, scale_t[0].item()) + inputs = torch.normal( + 0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device="cuda:0" + ) + masks = torch.ones( + (batch, 1, qlen, klen), dtype=torch.bool, device="cuda:0" + ) + softmax_results = scaled_masked_softmax_cuda.forward( + inputs, masks, scale_t[0].item() + ) softmax_results_torch = torch.zeros_like(inputs) error = (softmax_results_torch - softmax_results).abs().max() assert error == 0.0 @@ -350,13 +377,23 @@ def test_allmasked_softmax_backward(): scale_t = torch.tensor([1.0]) for qlen in [128, 256, 1024, 2048, 4096]: for klen in [128, 256, 1024, 2048]: - inputs = torch.normal(0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device='cuda:0') - backward = torch.rand_like(inputs, dtype=torch.float16, device='cuda:0') - masks = torch.ones((batch, 1, qlen, klen), dtype=torch.bool, device='cuda:0') - softmax_results = scaled_masked_softmax_cuda.forward(inputs, masks, scale_t[0].item()) - back_grad = scaled_masked_softmax_cuda.backward(backward, softmax_results, scale_t[0].item()) + inputs = torch.normal( + 0, 2, (batch, attn, qlen, klen), dtype=torch.float16, device="cuda:0" + ) + backward = torch.rand_like(inputs, dtype=torch.float16, device="cuda:0") + masks = torch.ones( + (batch, 1, qlen, klen), dtype=torch.bool, device="cuda:0" + ) + softmax_results = scaled_masked_softmax_cuda.forward( + inputs, masks, scale_t[0].item() + ) + back_grad = scaled_masked_softmax_cuda.backward( + backward, softmax_results, scale_t[0].item() + ) inputs.requires_grad = True - softmax_results_torch = forward_torch_softmax(inputs, masks, scale_t[0].item()) + softmax_results_torch = forward_torch_softmax( + inputs, masks, scale_t[0].item() + ) softmax_results_torch.backward(backward) error = (back_grad - inputs.grad).abs().max() assert error < 1e-3 From c030485081b990b6206cb28329f8b66a51449a28 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Thu, 3 Jul 2025 11:35:37 +0000 Subject: [PATCH 441/590] infinite schedulers and learning rate finder --- ALCF/helpers.sh | 5 +++-- mds_to_hf.py | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index f42f331af59..a8ed505c24d 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -669,7 +669,7 @@ setParams() { export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE # +---[Run Settings]------------------------------------------------------+ - export SEQ=${SEQ:-2048} # SEQ_LEN: 4096 + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS @@ -684,7 +684,7 @@ setParams() { printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then - export TRAIN_TOKENS=${TRAIN_TOKENS:-10000000000} + export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" @@ -859,6 +859,7 @@ get_output_prefix() { pre="${pre}_seq${SEQ}_gb${GLOBAL_BATCH}" pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" + pre="${pre}_ntok${TRAIN_TOKENS}" if [[ -n "${TOKENIZER_TYPE:-}" ]]; then _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa pre="${pre}_tok${_tok}" diff --git a/mds_to_hf.py b/mds_to_hf.py index d91513ed8b0..3c0b12668a7 100644 --- a/mds_to_hf.py +++ b/mds_to_hf.py @@ -45,7 +45,7 @@ def Update_llama_config(Llama_config, mds_args): assert ("Provide the right file path, The file should be of format mp_rank_*.pt") print(f"loading mds checkpoint {filename}") - mds_model = torch.load(args.mds_checkpoint,map_location=torch.device('cpu')) + mds_model = torch.load(args.mds_checkpoint,map_location=torch.device('cpu'), weights_only=False) Llama_model = LlamaForCausalLM.from_pretrained("meta-llama/Llama-2-7b-hf",cache_dir=args.cache_dir) Llama_config = Llama_model.config From fdb0965316ad1522ffea1335fbadebb65d899adb Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Fri, 11 Jul 2025 21:40:50 +0000 Subject: [PATCH 442/590] micromamba --- .../github.com/saforem2/ezpz\"/micromamba" | Bin 0 -> 16620456 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100755 "\r# setup for shell\rshell_type=\"$(basename \"${SHELL}\")\"\reval \"$(micromamba shell hook --shell \"${shell_type}\")\"\r\rmicromamba activate /lus/flare/projects/datascience_collab/foremans/micromamba/envs/pt29-2025-07\rsource <(curl -L https:/bit.ly/ezpz-utils) && ezpz_load_new_pt_modules && ezpz_setup_env\rpython3 -m pip install \"git+https:/github.com/saforem2/ezpz\"/micromamba" diff --git "a/\r# setup for shell\rshell_type=\"$(basename \"${SHELL}\")\"\reval \"$(micromamba shell hook --shell \"${shell_type}\")\"\r\rmicromamba activate /lus/flare/projects/datascience_collab/foremans/micromamba/envs/pt29-2025-07\rsource <(curl -L https:/bit.ly/ezpz-utils) && ezpz_load_new_pt_modules && ezpz_setup_env\rpython3 -m pip install \"git+https:/github.com/saforem2/ezpz\"/micromamba" "b/\r# setup for shell\rshell_type=\"$(basename \"${SHELL}\")\"\reval \"$(micromamba shell hook --shell \"${shell_type}\")\"\r\rmicromamba activate /lus/flare/projects/datascience_collab/foremans/micromamba/envs/pt29-2025-07\rsource <(curl -L https:/bit.ly/ezpz-utils) && ezpz_load_new_pt_modules && ezpz_setup_env\rpython3 -m pip install \"git+https:/github.com/saforem2/ezpz\"/micromamba" new file mode 100755 index 0000000000000000000000000000000000000000..dbe5eef86e31871cf78e0cac653f5b9e790a1195 GIT binary patch literal 16620456 zcmeF(cX;K*`8Rwm)zlT;)QFZigd$*BOc&jBQ9?|$ye@mpx(M65dmw~}Zc0FO0|^KM ziIc!DO7bHyT@Z>B0*H`+Ng#+&V!8+f457TEGvC=g`u;q5-s`!Z=epiM-jkdi%}4jS zHFImEBOBQ7?tkb3%gma!tby0E#-7G3vAx;YIyG_NO;>%6EyHk(W@AvT}>a|^}HB1>%L4+ta_W zx=pot?lyDhY&$ft?GCkh7x|0W0}uaZ?enVsR;_0ktB7e_Dgu6lzH0s`HZD^gv_qmj zSsj@6sqRn3hA-X;;7_7mdSyu*#J;^m`vJC9Gbr+BqSHyoxkm+1dl<#h?pk+4#0=E! z^k(>6QLFuhw~BT>(YDmZzbLjB&hVKcWS_dYH7KT0^mnb>uQngU#x@kM_K#p={tWxy z#lFW?a9ulx;^nH50UL{EZ?XTzx_I5@?R7hTBDRl0-UVpatlPgj9kXtiaXjtYqkgB@ z-dODaU7hvU#cv|EyXyL9As36Rcv2~?IOl1{t&M%WMh0&>|3F3|Hr}~!M>ugvAscE zOhFTH)f9L6b?RaQ?D#p_14O%H-Tu?y-J<=D@b$&J_`Wz(w5E8!h5g*juY>dIw$_}{XA@=W7=S94FZK`p# zpW9TxZsPqBc1($OFXT9da$P|ElSS+$uouMkV&vFbv_Hdk(e&1Be^takBl_Fdnc6%~ z>^p>xq4ueE4YeWKdE$LXU0=7k9p(Hg0xuP-sE7)!j z?TVBayJl=EBDPKJ*+aBfiT2PLqSl5TE@HN?%e`6McKyfJgr8WqzXjXxA@Iq%4;&=6 zkH!8YM7tXH^@{C>>*DG+2N17p9`y@v6?U$8-=cy&d=0C(c7w@;ldt2mo(f&$U@!B2zVmmJW5U&G-tt>Z7e{NUm{m;%_ zpo_)lFA>{csmsmL4cK=tcASHLDEi+L+tZ@0(&tYS_KtY}Q@sB!-u17!B6cFB03s;%$rfB4N9W_s!z{680C(vSMHHIvuU- z{ong@t^a$e|391gwEs}s{r@kFdKt=j0{TsLeoWoA+H}>%e`pV-xqL&!2EtaQ&-9?b zY@Jucc1PXk=hp4}KfT#GL&E=_!496m6!l=N=O_^E2I75sUB7?CV2& zMqMm#It>4ze^cGQHgT+1MgPxrrrZ3MZimL#gC9Zr7l{78uvKVZ(VZ>cn~V2aVr=(` zc89`Hlg-5T|ITZhx(_vpT<0NQwa))no4bh5{Gu-C4CFWlt=hGr*f<59ozPxHF$H%E zyI8bOAf{Qgr;7L4;=MlXmuRm?tD0@q2DL|0U)Oz>cAO-}asgS<-Kfs?shh+7!jG%l z_rJ4wjrh>sB4_bDx{`P=clkcIoM*Qi4O_3>GOJ;hdC>}webFM#IRMdDgj?N9z`8mb{9*#dzz#H<#?h#{Ob|@r-q# zc&lemj;a4SNBs;?e?L1+e71NM&%O1p;u*Dg#w&g{Dt=$0o+;~N)Ve8t51?MfGj8!r zTKvpW{G6qpkBjqDT&tdV^Q(BCDxPhN`(!8W?-KoDofWV7qOH*W!i#%W@iWx|+E+YV z7r!4-uj2PH_0QDm=fh&Z`Z;;9E?#}^KeffPhI$pxHN~@L@!ePdY@zOl#r0PFzM^;) zzpE%--xY1~T~fS?=Oy+0TJ($O`rv{IX66=eRJ&t`|r8>%8eddJ=^1)MLYfGsm4QRXV?Al?qzpg^!b6a z7S5h^;a~Uo)jHRo(I)cV^_R65{`iI~H@fz!6+eGy#rt0E_}sl`-{UU$a@OXb-1?tt zdi(N*vE~c`+lViNnS4+2a59sxCms}L^83XPO*46;csQTQ7l@lr{m<`eXpgKmb9|q; z)6A6TJ&|W7e?sGriwB^Y;+NNqvrRnM&lLYVu|{X|y)@&mYQ}SmW?rvp#kE8)x=+^8RvzXb=9qz?+cnd_h{DJts1{Uvu>-J`mCg>&xEMM zOzZPg&H3F;bAA0;Gp}Pb<6or7U)8L`k2QX$W?rvp#(BDCJdK+5P~4+uT0h%q;y2Oc z&o$%OLo=^iHTC?1X1#5psl)Y}`_*qW$Gt+cZg14gt3z`hFVW0*ZOw6y(&YI_bKFxk z^|@b@e_c%-HrCYhU2&tH>Ac^fxj!^(uDfP&aaro7U)OG-sq?*>x~;02?@^jOyJ+J7 zqN(S3n)yDjsavz={5KmVR)eIBK`es9;z>ljV_ElquP*4%$q*NijN)ZzD<^R>Tb9loMDZl;MpTXWpw zMEp$84=-uHk51Cmr}+1pX1d@0N#n0;=4)%>U(%efD>QZ9L36(PH1)Ypvwmi2^6#h_ z|9hJ2t50+PSyhu~cg=j;HNJ{wJ^WfTo+~x;dPGyV%`|oUp(cNz@eeff?bNKR7d3v0 zW}K&JuJ_|K_mlHA@x?zEo@qUIYtCb)SwFwjjDH(V{o6I?aa7}%YSzOgn)+{{@#i)7 zhu?|$ZYpZQugf)c8_?9j*BtjE&3gE;=6rpsspo^5I&7#}w^wM^&wpvgd6H(FJ7~r; zDGD~z`q^KTXHCt0Z!1myM>Y8?n)%+Y$$z}2ZXapR<6WBZf228ID`jekco&UG|( z*iDn?QO);PkLEfW)#Ul3=6Y||_$!(`cW8VwO`R8N;!oCGr}t~>)~?BOfoA+SYwEL5 z<9`s>%S_J?*J<)Us98VPY0mGLn)+<0Ilo71@|&9Re@$~Act?}xD$Q|!t;zE%&G=8$ z#6PY1URXm@&k>E!(TuZ2bG=`wxo_O2xzF6Di9b^lU;O()Gu;n25$AoT^>dfzeC?{a zKOdr5KO1Z6u)gMcc~*1Yx6>T=2u=RWG=84uysxP_?hTqcKdqVX)0*o&)*N>g&3gEr z<~%Oe9QPZV`5vo@|Al70Z)?VRv8E35HD3JlqM6p)VVd*)xh8&Z&AiUj_^q1jG1Amw z8O{B?`1dI`6K{U~RC8Utr@7A@qN&gC#W-iW?m9K|dRR05bu{Z|R5PzbG}qBXn)qWi z^X<^AlRY)@ABaNERL|mH7My9Fh~;2dTzt_#M^n$gXx8)hHRHTaQ@2et^;u5iM``>V z&AcAh_;woa(~R>}QHz=C_O9ms^R(vtZl|f|MVj~nHP=_0#=q3$-&l;_5^sKesyXhN znsxQ8CjXB#b=z7T(BgxNenqo>V$FEA)x;mDssD+Z>-S*IypGeXt1C3eT}I=hVs107 zljAh^nfEl~*+(;;b2RfB)Z}m0;6obYHqt za~@aGthcQ+`LEa1r}+0zW*YxbH2#dHZl7t+*Kag+xIk0Sqc!vD*NlI6&3q5gthX06 z<2hCnf3Kzv%WKBJg{BT)*UW1(&GmAT=KJDrn*4`o@=R*(=TB<#Kc~rmuqMwhHTSm- zH1Tg}^6#a&?ylG5`9w4RXEphMrpYszMX=yuOur-!ahV z4|WX=iDQVSt$(PieXyfqsN;1E^mo=iw?z6D4tG_>{Dy`H2Tn7Fy81hp?CLsoVSDeA z-cVO>_ma-w;@X534;GCv*wwM9dvKs{$?l<^`F-s}3zqcdL5lG%q1s3mx3j2$r~%_Z z3;UNG$mkwAt-nLePgJ3QpiOxDaL+*hlEg(xJG*+jL`hYWf&PxAodrF;OD81#wUz&5G07S({DoP*>Li@saM1{^6y^>1Y>2 zSlY#`$_0_jKn>%Rfu4Rt9ICyeG&`|`mW-}@U~qxaElbngBbKB3Z2R!=VEHWLfP-Q^ ziSs3f+~3tf=VIRAKznCL`_NME>>6&bTRXL^>iiT7Wq82oF7lO*jU%ZzH6&SRoW}n4 z{(+(1t}eekVo`Z{NO3)hvnM8Bt7vhkzM=U>t+GZzeotp{WN|?Z4lEQ4wykY($56*W zZ*N<#A=a^&M6sNWAu+bDVX?RB&u^P2u((Eyz6G7bU46c>xUZ{E9MS0R?p-)kseR5D z7eJ*?l(@Kh+KYgW%Ahg7uQ?J5>y>hBOo?j9Z#7i#UA@8~b?30+lD^nszCYTJL{Htw-E3#YEfuiny5- zC$hFy78bJ>WvNwI6uocWP_0fwhM1HZWw*G9hub+M z+c7Y(pr=a|wo6>wJ^fw7ZT-UoesNCwJ4AlWRF<)=t-q_ddllzL)Tqch6j2(jrMs1gTrEQoy8SiGm*QiDsF{x#ro~Fxz=`d$|dd36PLO;#kB<> zk_;^F8Z0h|;c9VJ)=Dl4E~+Xn?b>n@A0G1ijd?xdeqTW6Kwo=Le_O|ZxG=>mYc=wR z#cfjrcH>K_eMo$>49mIpEf8}n)@rfHL{S!tL`JP@xEOkd#99@%RXIvgo5dZ&r}zmisv2d`huf2-N7b|vnSgfCp%E03OViNNQ#g$rIa7FnR ziTZVS4RjmgYq3}yMz@HN)9#U%p3&F7pi51>*elmyahDKhxpp3EOS6BdIAX6jWyS46 zTuD7c)y|&zqVBZ?S$wC6^R&?Sy9UKH#F(q?ZDR2jCq{d9)~+5=QFVB+l0?~tPb-Qn zMp9fHJ^kGSvUJ5D7a5BO#Z|GWcAfSw>@()~4$LdYCH`MD9mQ$u8IotXcC8K%bc^9& z24eLM_7w|lxI-MGcJhmJEJ`K5c8VLjsM%n#Qi>am7(j8$s|}!2bmWa@VR7lzGUJvh zF28Gs=)D@8HHsA~^Z#R(S!LOsQONn9mjkrks`GVI#;YF{O_3ql-Llu{H# zv|>r23gRYPEU5O5jxK+Ajn z?;BWDJOma2?Za(56z9qBD#lZLHtZK)W9|L(i?b%iP}dV5k(bfD;;9D-#K*8ZM!zT+U>Nnr(51vi^WlVYZu=Z^59~nEFA1Aei|q)9?=ww3n#z0 zWRP`H{dLo~pq3R2c!8K^ZPZ40ZJEirVEh0Rns_`A7kF`wm)z&X_0*0BBDF%=#QGZO5cl4$;YITn zisz)-^O!hn?U_T&q;}0!y9fJPO{j(^prgynMt1a8ViY6;dRw zeLWX*@9pUmx9y%Gb$>4|t-7@>R%Y>WaYrfco=d+VYZrF$)023X6KRTzOfEB#zE3>p ziaWU&Ofd_&f|fpo6=f6`o|=~UV5tt`*`a4%2Yym4zKRWTYY;csZajDm`|?3Wd>)k$ zS9S4KC2xI(S`4+Xi=yJ<0OBq(D88Q;i%s#d;!d^nzE(V1)owGzSt@RX;^(FQMLmP! z8(&RHOkUh%dxwgN7EhVtfuct|uNHrG)%L|^XRrKQtP1TK8t7fLq>G38qHhB?BJ;^;smmCvAlLHZh)eSoGgRX&jmWZTN3e#7QS+Bx9B=9W z)M%Cw|Ga(8I+sT)&ZUekj=gjwMYYs9D~?mHto+1sBHwId?f+4iq8{?7ODj`+LY}!| zylQ_jx8-Z~ks0Nm%Py%xQH>(TG}aPUAF1k8wDP_5TuV_IRZW*Q))BGtf9lh!uiE9s zS)OI=B>p3RQ(PsMc(2vdBYp$BmT_?LAO4G1vHxxHyIQq>;RZe8H{Pak%#!^g7d~%` z`1Na_U&HWVGJdtX_*urs|0CZjrTv@YvzEvwKfhAlesNywjwkmoU)x`dad7=&e9IcA zh?wFvOI%mqKzz}^_VR1QZ_d{-7QwEm_18AeLjUSoe@){O^c!pQnQdH){#CX8s~b0? zziw^+s>bc;hqe7H8^1<>z1sc;<3aR?YvWtdcpUw;Yx`F)oOf4N%yi}}d;Ewg0) z>hGTv|G2XD`sV05wRQ}D->35S8>GkNFXHc49VfSUl<^6-ADQqxwpUcoZN4bK207RBHj4=KWp_+bqEia-asB5DcvNGkCEO;Zl5gOB6r%Q zH<7!Y(rt2YzVv2t{}kyCdC(`lg*^17yX4W3^j7kCk#vvTSS-ClZo++X3tlC+;Q_e= zA0c<)A-M-1CHLVGc>o_H58*L+1Rp1l;R$&HpCC`+DR~B;B+ua)c~q74pCWfplb(}% z@M&@%ZoE=n{{g&#JcOI%IlPhFI9-m%A~)epgjR zBM;#f@(Ave$M7n70uRWYZ_B!kkSFkvJcW;v`)9~J5xMal>0{(3JSMl!l=0)_HasDB z;1lE#JS8{2EAvm1d+>}rflrZJfy|SW2k>d~3~szyUjNQnGEW1!3pdGQcq6%aw#;LZ zTkt0G0B)0q@MiJ|?vTgu7V-q{lBe)i@(k{g=kN--agLm?Pj14i1x z;iKdpJRY-bil2Epi*)MDDM*d?w z=f~tLsONlg`*wNU3G(<^=^6Q6_dZa2&YmJa5T28}@M-d6;l}Ibb$cSbf!u?eAdz`KjCBFk6 zkpBWcLVhPaB)-}qqty z&xKErUkp#luYgaIUk}g7Z-q~h-vQ6b{~K<+RbJ=!!5hdQgq!4#z#GYb2e-(dfH#pp z4Y$ev40p(%hqsXb74DM12yZ2S3GR`<46l&C0{6*Zg9qgQfQRJoz$5Yx;4%3$JRvvs zmiMKUdm;~Qx<$V6xiWtf`Ic~-d|P-k`5d@I zzALF6Q{Vyl5PXDuF+3#yHhh%)yYPry)n|lTid`vzHA1D7lJRy(Z6XdtTQ}QwRB>C;|j68-ahr8snekiXy zk9>J}g?vT0PreeoO1=s_AYTnWLOvTFlCJ@e$k&3$n+aJ@Oaf74nziKKbkLD*0RRfc&5E5%SOAA^CE5%lVFyuLh6E*M*OfZvc-bB6tZj%qf9rAC(Tgbl)cgfFzw~}7~_sB1XSI94e`{Y-`1M+L(Bjne?L-HHo zqvSt;N8~rb$H;Gn$K*eRC*-%mQ}Um|GxBqPEw8(r{1=Ef{#jmEcfn2a2jCX@(3s3` zlcyKS`(6wASN|>5`3Kes<-8MCNw@te&I?F z)xKY*$Qx!$pC(@w-ta-W{wCZazhFa|$0ol6?vURGcgb&GSLX4^e+l==$Ke6_{qT_d zVR%ITdw5Kq!V~gm;VJnG@QnQL@SJ=KZhTm-^Skgy@+ZG8>tm67cYeKgew)a%d!^gt z!9CKO$rpWMd0s_YnlK<*tW^Nf%? z2T2dfZTKj;1&_#0_!zkXkIC~EdE9aG6yr?D-J@ju1bGTi$=zdQ{3LmNwDgQT#BrzC zG0vPkf=`pjaN}R)^_;*P$WypUp1~W*4b0af58zGY2Igy%2dGaod4~8F^0&8>*Ig_5 z^Y9A!EAT4$)9?}U=isB{e}a#Z-v=Kje;7VN{v>>o{B?Ln{^W&nzEkASz;p8F;nUtPDsK>jY=B>xEBNd6hzB46erIsPW{72!7dYVcTcyZEc(r zJ)@f&4VMNq#oGk$ePhkzWdLBEJG|lV1yO zCjUO%A^#z~h5QbRkQSz7I z5&0|dG4g-FWAeA)w>6Xc)5Q}S5{%XL0Uz7jkmUkg4(z7aep-yCjCm)H5$ z@CI@lZj#S|H*TCf^V4khj2F$Pb3Qqe7Pv=#2fRXlC)_8$A6_MY7#@)S z9zH_;Bs?U420ltY36IEMf{&5E0guVwhmVtg3Qx$F?UC2l1o?{alzbKVB>8N3My~2J zMZOl|bMp1z)8re%jenQd^H<;vOr0hgZpm;Q@ISK0TtAb&<4 zVO~?@#$1^vCy)1$K24tOD&6?JydF%{$0Q%ccpAwQoG**qK0xMgB3I|jCRgXHnLOKB z=5fgF?@4bV5BHSbN*-hU9=WP#gKTxGsAoj}#%uEXYmEGTcuf8& ze4N~PUFJ#1SAb8DuMAJgXTvAS*M?{08^Wi^w}j{9+rX#E=faII%Ik11cmw&qaFhH% zcq92CaEn~kr-}Rs#M|UY!<)&!1$W3#fVYsh!CmqWcq{pQxJSMKULhZZ`{bv=tK?_F z1M-XEBji`XL-HHo5&5m~G4e5ZOny6joIHjn5n-vr)7-UPSFZFn>JE^vo@A9xG-0dSZ6 zV0bI};c$;!)u%#!6ykmIW8qcuY3Opk3hmVmjgvaD(z{knY zg(u{f!Y9bDg{R~7o92G7ZV4xc9f72KFrUgy7oH;~^0H_0D>TjalmH<3RA zx5*!aH|bk-wSzYK+GrKk`7CzlHpvtEIc-e}Y%Y-@NS+`ExG$Kj9=`eufQ$xi{VY=SHf-b@57tPZ-G1HWAGO8+u<(xFW{}@zlM9{_rWXVkHLNNr{Goc zNq9j1GJJ&mHF!w=27HwK9e6~p>N7_EKH_8Yf5FGeKY=IYpTj4}XSK=olaenFpCn%i zo{_HxpCVryo|A6~pC;cNZY)<`&)dKo$mhaM@;%@d`2p}I@`K?v`Qh+p@}uAm`LXa8 z^5fwy`N?pPd>*_)-U;`~yWv&x3Opb`1wKOF3lGWr;SsqHkI5Io6Y?|QDfwCOjQm1) zPJS8OSiZbYM&TCu(chKpyovk-xJ}*(Zzewl?vVTN7V=@ZOMV)>mHbS&M}8i>LVg+C zC%+b6C6C|%`A^^@9V!M1D4WjQo6fOnwo3ocvOF zLaypFL4F0|Q}S!zljPUKGx8hZQ{=b6bMha-r^$Z`HyX<8_P^i_uM{wNp8a}@?GFIxqZ!{ zwfko?dGx$=hy7~lE#x-hJ@T!-@mMW?g?tCNPre(xO1?KdAU_a3LVgH5BzNJX^kNh=wh5QY;PyP

OjydKBMmxagV%frXXSA-|zE5RqoSAnPGtHCq!+3=iv z9k{Vdc^z&DH_0uyMcxFr$+v|&EEg^ilH3?SEgp ze~yuxJ4hcVZ-q~gPs1n4-5q6~De?e5O`gIVRxQ`xnj`ZxlKb!`@(A8cp2J(ne~S9F zl0OEokXt*+@l?qJ_z3wpe3ZO0SLPWbkKyCwjXTTu3Gym@l01P=kx#*=$*pFYzhSj< z{afLUv5b$%zt~9n82NbvGEYqYF5<_@M-ZQoUjm;XUj$Fdzjom7YS;B7`I?B&$ot?^ z@^!7cJv;STvJ*U9`X9%&pwvkO#UG1gnTJMd=mv)~T-dGHqU3*j#L_u#GMm%}~stKk*$>)<~5 z58zeuo8bZZZSWEDpTI-%JKz!doj;K4ZHzp8zuhQaF?o=QYdg*+!lQ zatCgbt2$WZM;$m(yARmpp<18h8S*>i5#l}a#vAedORn-)$-P(Qd;@ZoKO%4bsmvdf z2mhW|o7Xt|${jUN$YaD$kq3W}`Hi*8^*8pBc}((fPwCC%!G6*matq!QUi z+o(@M9-}@P`PKM+LQZb}rn7c^O_O_xbYq?Jd9e_0k!MfK|9+u~9etZTLcf{ZLtnM( zzb|N7_1KZM8R{#~*@ZqzPp0Ji;yg~0FZ-5^&&Z!Zo+Jp8(h4;stocVtCb{}J+Y z?v{B%^6$Y%$*+e;3vv`cRxfABsz zPM3VI2c)->{|?9X$o~PakUtIg$zOt3$=`toXW%lJyUZYiG6P4aw7)~As?dQZAVZoMJhA-^Bby)ESF#`A0E-6i+ot>pQ- zGTtNiHj*BY{}tEK2)VJLj1Sp&ls-xxtRsDlJhG+7 z=~LvJ{_(NedYdNy`Tf$3P0RH>^=auQxv{6Le8#!r%`^QBLb+b2kO zHY=aUSD$%I-e<_w^Q}kzH^lqo+x$u93CPv+Mo7K`;v@1eo|1WDa`pV2kgMmAlw3W} zWaM*k+?>4fr97^&dAUBHV!kH%f8qJlBHs=9ZF2Q&O5?^B<@4)ZA@ekl`+;BPBi$u89+uup zZaya6BlnTNqVa_MUl{)cxf*{;uEsw_uEw8}tMN~htMMCOE!R_xzd_?3x%$0Fg?tbE zp3Nse2wo*W93GHwojz8(FO85NjrfpUeeaBtpN#m3yc<47z5pJRA2nD0{5wwWBR(Nl z-!B>YZ*d=(B2S-|>nbPr;nU>KlQQ1;TKPPBsdS6{3w6DahtJA*o7{a@mDk+kO!zkLaypOL9XVTlB@YnlB@Ynk*oRUK$WKDNOYXs2$ey1M*|xBjoSmpFf7=Cm?>5d_ImFk-voaG4ch7kI8-bIQcLvA2Q#P9!`$g7BV$a~?fKrd*%xe~#w?ay8CIay3qiT#d7dT#eHvSL1ZZ z)i^!!Ce*V^9_%LP6_6*pN)O4SX6X@mu$}aTycNftBv;4H$klOka&=r|+j9NYaV_!@ z9JiTV9oHdO$92ioaXoT%+<<%!tcQ@?-b2=Blw6G`B3I)XBe!;!d17)k&V*czGb29| zlKmZ6sI6waC?RZSrJCd0dBF9oHjY z>wS6OtCFkZ2IT6vA$f#(Mda$Z3HjA;%j2fx8R|SquEvv*tMN>cr#s0y9$kU;PyXQt^0+PJDvwL9@_6LO z;(Do&+o3$JPwrhNJs{r~c}B=po{(JSiO6l_8DqzIVse!yAhU zQ6GmL=hq|eIZ%$@Cs+BaebfIR(3 zdP45}OZo(P{-N}Y{Bj&OCs)UvCRfKbb}pY6b=(GWbzF=5M*R0d&Eyfr>5#{8kNg^p zze2v`zXu_AK9zL{$u0OOdGLvhkI0>A=?Qtho2<(O`ET%h{*?Sl_#}CD(Brl9o{<~7 z%luR1FCb4&{yuz~yy;+>-)JtMM-%xQ$UjFOlYE`!I*C6Ow-cA0m|KvIHPmu3`JSq7B@JVt9`7?5358VIBk3^oFyc0f6UPXRm*Yf%|k-vew zA9+mjT~@~ZpL`toEpiL_o5;V7{5H9_hWxppnLI~+hulW~7V-;`-zEP6yp`NM9P6Ll zLH-K)81ne!_rj~>9`Xm|F7l6%KY=_U`QPB99pIznyTZrF_koX-9}J%$KMFocems1Nd>(w7{1kYDx}I=-4Z<7A zPlq>=pABy&zX;w!eg(Xh{CaqW{1$kX{HO2{@?XM7$?t}bkv{|9^7 z_X?)SH-b--pR$r%x5l33?|1hF`S->&kUMab+=e%jTks}wHO^*oe;qlV7V<^#R`LaF z%lHa;fcPr;Meq^wtKp;M+rY=jZ^CiM$s@#1kpB#xksEu-x~sMU>woM-`MuMC{mQ4{ zCi$I>vJQ>pPs1(p#*<|JCUOUElfQ#J&Ey{99r6&~LOu?6$uoE>dBe$gz9L^`eVli4 z8}UB52d|QEi97*$i1-omf2ix1+-Z~J94GhT6XYR$l01P=kx#*=$*;rs8=Ug>`%8Eu z`IGP_@MM z@ss50J~KtG?pM>~J0O3<-sSr52{*|t^qa^X^lgo|kgNN-M;@TBT6N#>J7j%ED1HfV&o3dlKVKeql)P(WtWWaMBW3&;`Pqn%$*+NrlaJ!K z3Hi?vKSBN&JS9(%XOi5g$Z=-m&G3eOX=7J#rs?)vEK=it&dOpCeC9 zo}jNctY;OGx7j# z98n&B2)D>1xI-SpJ#rs?)vEQJsq2E`6Xc1>Q+P_A!E^E)Zo1`p8L!CmWs{q5m)wH; zCcupR{%_Ga>x6oIus;9NK{GN3vK0zLjJcS4389XA-;R(6%s;p;5 zZo-YDX#8-C+=e^k4%{Pm;Q_e^kH~#^LLR^~@(^wuT^@e~x5#6-L!Q7r@)RDB8?VWF zM&vR2s#WLR$9YdF-a?+7+=H9Pl;`Eco5*eSRjZC0;J7Zu$H?Q8C-9IwLSMBie~A1E z#iz)Vk!Nt@o8@`saEsh{UDnwlchOg^8c&4r_!MsF)vEC)7{5pHKJot;yaGTtPyW}CeAy42Lc?vg< zD~~^eTjV+1AvdOE-8^z1ebuV*TNr;x@h0-bN^&&!3|bu&*W&&zm2*54*K;V!ua_sMN|NbbO6au=SGd+?mx zhnpvs#~;9L@(}Kld+4iH)h9%K0*X(OCnC?_33(3B$c;B;osE;q<2T_JxdnH~UG!C} z#-CvPKE>O}6OuddnB0Y@>WO;&)vEj^^5+!qA&=Q!o|g}|$sP1ntMXgO?@@e! zJOOzKkH{lR?lRjyx{8 zhrVi6o&b3QiZ|Ys=RG1f;R(3~&&X}K(N&(81GmUsxI^y2J#rr&kO%OHJc1|WDf+5a z^@&iQoZ=JYF}ut2O5rwn26xGGxKD1pC+i=QoA8+2f~Vv*JSTVH=KS*bUARr|ps!k0 zpA_}+DBeS!fZT^i@yBqBJb^pp8Qdc`a#{a?+=NHuG5V@i^~q76 zl;SPq$;oZF*;AgE1GmXtxJ&NAeR3Zjk_YgZJc6g>8TzVK^EKC#-^<1+|7C_k%w@jzdZg3Zjr}uhdhCMV(mau0pgs_`cne?ajz@lZ^JEe2kwx&aF5)F z2jl@fA`jsSc?8eMW4N)fJpKf3k*9EnJcE1WIXoZ_(O0die}wwS6mNWl??G}Co|9W} zb5VI-Hryt6;4Zlf_sLVNw~*XJd`zC8uUa+V1oO=(-bWr|ae4d!+#(O*4tWIk$YXdw zp1>pW6rPah@QmE}Se`GVS{}a%x5zEHLvF)8at9ueyYPrSMqjn6{yFNOQoM&eIk^uv zPb<$WfZOCD+$GP{{faz7e8`S|L~gB*`xSYNJQ;ZcH%>2)KZRT58QdYy;U2j;Ezf&E zZowmR2cD3p=&M$pF9-9@Dc(gM^V{Wld2pNDhr8qf+$Rs=A$gAT7?VecPsua%RjcOZ zVP3`=xJ7QGuUa*)M4cCk_mRgZ58xqr z2#?7lcuF3_bMgdkez!cY6mFAeaF;xX`{V)os#SH!P=|=(jZfwIO~_4nMsC54pgevX zZjoD9w+^|3c#qsiU$tud#s>0!8&bTBJTbWkPsx3FP9DI`v&!=d;Wl{$cgbV8PoBU- zavSxD$y3Cq!ii3%AKVxJ&NCeewVvl85k^Jc6g>F+3+v;O6<|@uzT` zJcGOBIou~VK1cn@O?XUh!BcV@o|8Lp^MdmDUARr|ps!lBt}?7EkK#S#3CMkTL>|Bs z@(`YpM{r}LJpLGNktcA6JcWDY89X4*;Sss<1-|ddO?XCb!Ho;c|Et@)(|xCvfAk^7wPOMQ+cM^>@f! z^i`|IpJDty$0JWj?!jYnAD)tj@SNNT<$P4D#$#+G?}rxlD^HQfAkcaS$JV#%(YMd6vZ(d$LUorC7#5H zJVRf#YW)OQKgLz%@teqFky~(w+=hGP4m==t;Ssq9Psn|EMjpbAtIOk$;1+odcgPdC zN1nn1@(do4$LOn8)j39;Q;N@#Cnq&EraUhbZj)PZm)t{NwQ4-}Ap^DV%?ia^ zhf4R!!~LXJ$+Lr{hvX6Zs#V8LP=|!#ZHzx7ci_gg@hS4;pAz__dWCZ*!#2h&gB8zTf&@I1MbTsctxJV1Ni_R%E$1A zJcmbe|Fq)zV|f5iE9l6h4*@;8Xb+p38H%SJM02>BaT?@&I0uH{gLhf`{@H z-jEOAE%_K8$#Zxt_s=Ma$a8oq_kUGfe?_0$ zkM{MI>h%?3exObZp37sn_b#*k9^98_@UcF(AMJX2T2GGoiaP$Ui_a^N2k=ndfH&k3 zyd_WJk$eD;vQ|ju5+w?>tTLipD#R@PvPEs z%zk>mDX!m_SKxs>gg5lL{b<)Y)jA{2tCPZ0`2gONkKwsIhkIw4=VL$G_2gPlMgPBb ze<`yMA-p1Q!9#frkL5l1K%d)>_Hn(H^>rWfJ#}*URPO(_cwV_YfO|`u=hc9>^tt_L z*VEAbZ(~0DoVn)_y!9D<{|v9_aR=~BpWBc2aeeK>6!W1v-tUUfw;`{Y_458jh!@Jv31Pvzd(#r0Q~G3Qc&hw>2K(C7A}-M2vd7Gb`nP73eI2k?P>4A0~_ z+&f3FFMYJ@>FJ)7;0?JC&o(#rCxB1&xD9wqpWBc2aVzS#F(2xEJA(K0xGCK8&AANV z75NxG)#vu3T~DZU@%qJc57nu_V|fTqX(v4?dM=@LWEDhwnA}ki%oSe{S)4 z_2gA}E)U_A_nCeR9?CoLRGz>G@*aFF&)~Uy3ip>a>zTt_^3r+5=Nro_@SZ$`=kgZZ zf4_O$7#_%b@RmG-`)`_`r<%dHUwyRj*GY2t^ceHHo5LHaaqs-%^BtUEyadnH@!{jd z%$MP*yaM-sWag{zKpwyw@)|snhwxNhhY#cp_*mYA=kgZZ*Yj$_)055fis0jujCbfi zHXg$R^;395-i1f<9=!2$(;vZW_TOKIM_n^NfoH!oK85>N8lS=AYmDdc^eW?Xcy_jN z@At)fn5tib`{$T>AMRahybO;nHeP{G&of?yr>7bZ;NGu{*Wj`CErh4?I(#Z`z%xB= z6W-H&3m*U0tS5pubY2~JB=5uReHg&)eHg;+eaPVUK8)b@K8)e^K1|^DK1|{EKFr|u zKICwFALej-AG`~S_s!mi65QSgA8zkM8E)@G1#a&{6>jfC0Jry{2DkSigxmYjfZO}f zgxmX&z>i(od>u*Qy+4@y)`f=;8t=j5zZvhtgNgA0Jif*F5FXxaJcDO{GCqP=?lwM# zMPd;PHy)dFAlI`;C_d#dEjk<-_gsEyL~e zt-$T`t-|f|4d{BlHMo7gA>2OSI@~_r2HZa12p)QOF1&{wcu$}A;eqCd@QQo{xBV&n z;s4yS@NsAG4WBjtPG1h6eap6GENQ-SZMP7U7M->g4`U;DVZhfVlc^DTHw{SLgL z`568U&3EAy&G+D2y=Y!9L%5yK;QQZi<|pvno)^5MuSYZZI%}BM%N!m&ZuZT)w0K_j z{ig)C?>|1=zWkyxPAW#;r9Kf4!7?=4Y+;(X~OOMPYZ6}f7)>S{u9G{ z`rLjjEqvGUywPvXeePp^E+4>mTF0Er5MKSA>16QlX?_H6XnqXu$S3f=daX$=e6Gj!;r8Eks=z~iZa>=ne@6W}<|}8L{cOOW(|i-Y_`2p? zTJX9$ZTJT@AHi$#4m?&Th7aTkd?HWbrE|>l>cXq?9=xIT^x;j-58$zU2=B`?_*g!I zdwsLcF}y0D!0YlUJd)4gd+T}S@Kp12_(<+uUc8^VyaYd5eIH&v*F3K>{8Y_X-~)LT zeuq4Ohw9hhGtGzaj^^v|w|&yw=LWo|`6hfN&9~qq&9~vXJc5_cGyBtluc3Yn|ExTL z?;%g&`^&rVP>b$JZGSL;vU9nGikp1ccxN}V43O?e+aQfC0qEx=2Q4} z@-Fu3mVxCtG-jIjz9ku>C z{JZi7JW{6#PvtH6iR!fB=gA}ZK%EYJDv#l%OU?65;DJ1ax8+@USKfnX@;*G558&m? z%zB3KKk4~q@KEz3cqAXg?f0J(cvtgNcqX60AJ%$u`0MgHe7Q}`>%zOHc>m|>m*ADl z%|7_>ro0S~RM)d=2Ge+(bX zC-BnH?B^6-lh5ESc@9tHb9hDT^sX)5|AFR9@QK`qdsmrtmfTfb;i)`;Z>)9J z;3Lh4@Le=thtDuNnssc@w_2IxTql zk7nQ6@NG07!S|MT;GsG(Jd!8yAF7kWdz$aUC-NS=l+%7-BOk(ls`X^>p5{mJ znS2cQuQ%(Nz|U5H3b(&s;9Xa|w-tSEKbG|S#i9CD%wMg30KZ#agGcIw@UFZL&(&$b zGtD>QGkFW{-=O;j59AU2J)4>P-+`|xkKrwK5_l?4;dcLf@JyfEk9Pkvbw-$<%Exg3 zM%@Q^D4)V3`3&BZ=kSSq4lmtg9@qP0@m_`U61*e#;RAUYK9yJC{>|obtMIx!fOq6I z_&^@Qr}8@7|C4##2D~nB!aMR7yf1IVr}7By-(nuO0}tgfJd!8yo;-z5mR@y@*zBuXYihU1Ru%A@LWEDm+v#{nZiT) z4E}BHXAVD3K8Lr|@op&I!$e+!XL29jSHBFOYQ6%$O7m5C>3;Kk1Nfbqufd;`hj4Ez zb3g0w<>d`{pvP^(oAMSsmbc*pc?7rjsso>BK8BYjW*-vxnp%Gfx3Awm{9yfD@dQ5I z>41g*yU47VkqOufaok6Yk4f@MHlmCxW^dF_s3=U4IszKVPTUteDRbFs6QyaQiPK7#-4 zQr-VMi=B6c=JnNvpQX+aevZ8KmtyDpCDUobU(%nC=)>2)Q}-YK(%X&K$Ho5i9^+m3 z1zOJxeu=#H*J9^Nc>=%X%cehp|3UNByNaEgUOrF5ow`u?3x9jU#_5Nb# z@o$*<4m?qR1b;~9Tb>j9KN#V>kkzB zS7^QqKj`mz|A&83^R)+you)j2?;)SSzb3CfRO}oikKu!pzQ6GMhXme}r|?wXg-7xX z-ja{t4Y~LC;yTyUd9~nM%iHih<$d_R@&Wuvc@95OK8K$zul=LA&QD!r&M}5xs`(N8 zD)|`xC%OM{v46L`4F89`3Ge*Ze4p8ZKdJd1yr=m-eDS@_>vRSmX+DRqqWNH2Tz@F9 z!FQ5(;9rr)@crc({4n_lev-WONO3)9$Q$s1ybHfkK84>WpTYknuRdCQ+~?&1e6fAZ zK1cAS{ihyaoS*Jc8dOkKwn=d+_D#_cQS9W%GVDfWP*2?LYi3^)q<1 zcxU0)uMs?ukKwJ?4_Y`sfqR;t!dw3{^D}s)P7d$A&8%k*&);F(`&aQCeb0Cap6hXa zc&x`Q!^dx$bynbwcba|`o~jeTeeFXH-qU;??y28^4>aF|XYv-jqQ`B+$8VVD6~RM! z2R>FmhUZ%60AAVN{JO*#zFcJP;S_FvU1AQe>-#`|Ry<$({bUtxzn=`@A6KUdx8F}j z@K0zyfnTWa_j>Tl__;&IM{Ig?o?m2u%&6l1jKHuHs6?jWtgMVKh!GA1I;U~!l@Kfbu_;2Mo{6e|^ zY;pZVc>w=|yaCVT5&UL(3LnV_@O$NB_(Yz=pOE|iF0TJMc>sS+-heN@zd7FszM?#Z z?1Nj)fxjctA<^F$)>;JAifFCJuz<(}};PaLA>-O-oH9vq~As@ppmFMsqH1EAoyeADk zZVA4v<}2`6{VIG{&4+ML{W^Rf%{Sl&%3JWU9=8quf0~cs{sm?q68Mt#eBnP+rw0$u zH=REGG|dm-=gKqq^gNv}e5m;`{6_f{o(~UM`1N`QA75!ahc_-WK8Jgn_g*aChwLgd zUxJ5M829Oy8!y9K@(O%#jhU~)W6cL}{|{!q22ZsQA>3EL4sTp*Iw|~^BhBly2cNBD zybu5BQF>j&&yx?~H_J2lSU!T=-;W-{|DWb3@W0Ea@ME6V?|;Gnv8(yIlf$o6XAb|5 z+>_Z9uispUzVn>_#GJGj{1%BiIGxJsW2KSif6~I?erv^V*>kQ%h%Iok| z)oH+Y*L)MciM$10P2PsDr+x%~zvese&&gx>o7$fQ{u%XCc&_;YJo}>gb-4^a)#qdQ zG3w9Z(ME?Zd|wJ)F5Z*g4yNCLuc!4y@Tulg_~x1)!*`VX|1Cc5x8x0YN1nouv3>Z@ zY=2(tpJn^-i)Dtx{d;X6{)p|vpSFGYtG54YvA>M&RRjJ}c?#b^K89~8 z_g^daceZ`_-nI|_zU{+*Z2PYl`={DI{5;!-Uv2yFTjV*svZJ}T{u{-|UHnJpeJOyq zG~a-~PxCQ6+}-^9Vk!KKn$O^NehlA5^WOi8>x|X+;VsRFa68|Cx9$4jc0DP42fKc_ zogc$@wCjJvA2w>=nkF zaPQrBEL=ew-c`Q?KTn>(NAfQG*k77`=)?b}`62w2Pn!7=JQ^6Ez-Q{r;Qy7+;maIr zI;FQ2_kVSH8D7=?RN)PI4Zevwb@-#FY5(EdYQ7EMRo;Q`BTwLm%DeF6<$d@W@*(_E z`3Qcyd;))1K7;>PK8G*)W9|Rji~GNlybNC;ufliN&tcZ!n`^!f-%;L#e^cIuA0qF- zyYd8nk-Q7PTHc4>Cm+I}laJu`o=oADt^aD?K^80SXJs4X75etZL;4QJTl6m(kLf!b z@6mTLp22&&8lS=|yBqi3f%WfUyh7j8cnF`$TkvcRGau7GVZ29Q*LX(%l<_IOX54#c zasAmw#w+wqjEC@WQ{ydowwdvm9vbh_H#eTqKVy6fpUS<(i|e0mVdg9JEsck8Z!6<1 z`e%*D@L1l1_v9H}=QxG8Tc+qz5uxObHC2HaPt32&&=g8PSwP!0mmD;r6~IaC_fUxV>*(xV>*ZxV>+ExV>)!xV>*fxV>)~+}^hl+}^h_ z+}^hd+}^ht+}^jGuKQD6vUq>&eGB0BzBTE(Z!NgJZ*BNM_br0k`F4ug22F{^9n|gTUXNn)SEf!F8t7 zhBvM?9>FW}4&1xS%*XKj8siDPr}-3~56yfRo@%}a?_Fu;`|w!v19&JO!UK6mztr?c z^y`g};l4T(_*Bnp3U6FtIy1O;neiOn(t76bEHm@oGR6B4s^5gi>a^fvJ#HKB%OiNI zP72R-&wKFDH|ra~?Y$ktTf3Uq%M>1~lf!Mt*Z*JAuLDfgslYRR9>8s<2~TzIU58H` z?!UKqE{E)3exITYe@d^@3fz0eoNpDL{?~W__vJPE^Jcyd&tEj&fP0#6!owHLdsKfcykfv4J^Dm>DB0QYn*HF!ny4fu37^LmWn%j)NXQh2<(zMjHA zqWKAYx~G|+!Nac@_m?g1L#+7#9;n}ddzz2no7(lmy*;#k_~-5V;a*GYhxg?E`>}pK zuK>QYT|a!PP6UrMpTO5V$$Z`K!apJJ!#9+V;2Yb2FAV;1&ClQ~%1g@?pYJL1GTi=L zSrb0}!VwGK!Kd)JX*`EVI~$+F2VXYcT)w!TxBBq6`taVD)Q3-ZQh$X-_b2dhS2I6_ zr*(D8jt)Gt9k}fTA1HRdeX_X^HTeGW5bp12)>DV4Uo_r;`|>8dvWJ;(!^7>3NAN)N z9eAw$iQ(DSrjx)u%@5##d<1WO&UD7`Q1dgmuXE4gq1Hc#Pqm-r4;IfQ(mk)hJ$V%# zZ)4UOz{72g*WfL!KZIwWH}iG)K;D2))oH?e@(6yM?o|i=GkFZR_bq|j`+iea!n%@t)ZGR)TxlhX8KxTODriTLT_!XWj?8aC=XB zaC=YsaC=V%aC=XNaC=WOxVz?>tYf7apIj>a4CR35>-9n5_E;o|yx zyBbg6wx7ane*m}rtb)GQKZ4u-7;gJH-1fbXqObLr;I{9>Z9jn9ez*$ycK_kF-+C1+x4$ioF8m!_Mr}s>c%5@V=LqFCyE_^JL3u5b_Q_U$<{7*?0QCU z+sWa!Nd*!7g)wiCc@CtSDKvFoYBZ6|`;PW;JY$F3)V+s*)PJK1{0PV3WVpGWY< zmd10q_i^Li`o&ITGvg(?Isx2v!VQWYyPi7St|x-qPFyQ??0OQq)-!xu? z_%p?MZw=!K+|FlP6z6S!1aGZvI^I^rdD}0+?R@yz;=JwG;f=Ml{;iAiwx7W5d{!^c z+x`gNTT|=*d~x3POK>|MZd07M{W{!JKi;-DZ~F;d^Vt`Q^R_>Nrwg?H?ThoaUxLTG zoB6O&oVWctJk|Sf{KewD?I-ZT>ZYIVP@K2@5!~K4Z>QqC?U&$oKKxQ~-uCNod*9+O z7w2t1f!q0P=itb&EmZ6*WvcQ#osE<+kOJK^Vz<| zdD|bs?S1pUU7WZ565P&*-zm=9ejRS_XZ(M~dD~Cu_Wd&|&fES7Ztth}z2dy>m*Dn( zhWi)iZNCn;_cK1AIB)w2+}_XZ`^EWiZ}Yx7f{*vr`~N|Uo-e^Gnh!gRp0C6GZ<>Dm zgGJ9L@K#&z{|7I6egx0<(fSWr^n3~4(0q94qUY=I-Z!-V!xlZC!0mo!hc9}51h@O? z9kJ;765Q@*c;uqz>u|fD@llJOPvCYxv!fS1KZ4u+^p08ddX;CuJwecEV@&N=LeWh1h?ynf3fIJ0*}9^{ejy~)-85IJ?;qp*8ae4$2)b=of3R% z_lIux=d?w4>hQPr2X5CBpT6i$0)K0N;I@;UfzAS%47IZ@)Uloyazu~K7gMr&)}!Y$M7@dQ}{XZ9Dae^ zJF~c-m&z;ftKiq}4nmmJVDj&nYB%i|fljrcGYH&PvM)&d+_b#1Nffu4E`PY z7=D<13O_-f!+$0B&Mxl%g>oN$oxB3SQy#$gIK}K+3*Ol5sD(eb(1us!5!{ov&MB^Q zY&-Ct?Z9K(=@&bVEwmnZ{u$#Dd@N7lUTEf9=Q{fExBBq6`tY~<=N0=EyMFjveYown z&UbX+Z*|~rb$;*Yz~Ab?-|Af8=)m9Vz~AazSnLdRUTyezbDb}IAa7k%?4-5>Z`peS z59O^vu~V@fc&>Ad;IX`Qak1ms4m`6Rc+Yk&DRxpl-!{Bu_Xi%@{kgQ*sn`xY*XtsJ zkL~N_vSMc-Z^L8Tfw$x-Jh1)Ci~W2{z24#8R=N*xUtYhW*s=dkSrh(l{rp}VZvR|W z2X223If2`shwsAe&jwwPya|6$-iH5I-hsaCcD4kI{S;eu}&X|DXOIusZw#%{Sq< z$lLIp&NP3|QU`vg<`ej1@-F-pc^|&ah3345@QQo{Ut2zbe?~ro?=GLiza=kSQ{4ZF z&Z`VRT=P|US6+jkFR#P@ByYm^d%M|(Hhix2bl?xFlfYk+ci~IveXI{(Q9gvPCm+EZ z@(H{ppTR4uo3Af(_(7U4U0dA$y5`I9<27G}pCPZo+v?QeS8Bcq|Es(We^%asFR8yj zErEYj-i2={@58?!AHw&QkKp(J!Q7JxJkj%=!DDsi@PXz_e<<$%Pc>hL|4v?oUoNk~ zZa2QN0yw*y}wPvB!cuP!{)d>_86d~k@6<|RCybIg}ej*i#&lpDeuDnC-1{YTK^E9 z%SZ4JUZMSmuO*+s%M0}V5BxKlFWpew|6Sx|`2O-L{EK^;_0-^pX}%8sg}e#BNZy9u zAn(8*kSFl}$h+`&=-(mg!{09-!q)bev#(u@Ehb!_}%h0Jkj&&z#rFq0)I)~g}*89!#^}M>mS0`mXF|D$tUo>-GBI= znxDh>lb3EP?*B3JGW=)qD*QZo4L;I(>hPJo3D4AN!~Y`h!2c;v;4jF#@ba4GeV`A2 zoBlg}2wzS-4e_=fTsJk;aP;X7%*baQe4zbP-n50h8nC(CQ_v*mU8mGUP1E_oaN zguDZPMV`RlrLW6f_=n|v_(1fS*HhigT%=<$JzN$QdZzS)+hxY!%zo7Xc{Oj@&e5%d_ez@jm@X}iP z`U?-_rCW;o-;|f(9eEZ0OFeE4-qU;?e!1qG@H^#g_%reje58H?&*fcs`4f8ofrs)T zye%KW6Zr(bq`qIB!TXw@!^iT{sJQ>k&DY@v$eZvJ=GjZT7hfzg_cvc&Pay{4vds;BT|P{saGjdFj z(rv~4-&kITH|16M0rDC=)Z^CS$7#L^Z)?5{Pvsr>P@cfgR=*3sPTq$W|@+!P7ufad5$F0LZFK@#4k+jL~pQ-sie5&~&{A$gQ;P=TV@aN<+_~O@_ z=QW3~EHC}Jxc}?R%kVGAtMG5gYw!c*b@(yzCj1O}8-A(01OKBuf!`_b!k>`$;jhYv z@MUf=`!<4qL_UG9FQ37^be9YvIoUHsCL4 zz6qbp+wf!5>A)AiOZBJl zGt|l9AJF>e4ln(sc#gkQrwm_BoeF$yc>upqoe;i_)?bHjsZJAqr8;f+E}Ddu;#n)W8{7Kt?CTnr)fTe_vB;vUFuBX*Jys`@HzZ$bxPynIWF&;`{u(l^(*ju z)d}EZ{W*~u{C0Kf@QFH2_z}N1KX=iB->*&te^8wm{%O6A68Pilbm9L{rw`xrA+w$V ze6CIge?*-ze94>4>vRHtk9-DyT%9@G{(PYK*Wx)oAL|_9|5B#{Uqz2wb$AW_v^sV8 z>gqHc-h#iNP6S_1osPp3`0MI);ajQGbNB%M&TY-tiwwStIwOZq;7hABgCD3)?r`s} z;yJFMjt~ElI%S7f;T3gi@YB=@9o~Siu1*VnraEnhci`)&lfVb+r0}7<2j57Y0X)NMa_ z%A4@#5@Vhp(pjIecw- z`JUprY#^_|Htz2_VPA7+0#6)2>#5^%-4|^{#8A%6uz_iUHDhzefZw$4B-c8 zK7+^dDf|}Q^BMej&ClWc>Tyf=7Wes=n)l&n$}8~w)d}F|YQ6@)MBap7BX2o8f_L<| zG5mUU5{Gx;u{wSDo$3r6p23e*XAFN>or%L|@RQV;!=F~iyRUd&)_wSC>Qvybs8e-# z4gOno>hO2|$-G_~4sXFPP$z;fuTIC|3H%Cmy6~zxJ%a^hJYQ7D>T%N$Mm#6SM zy$Ss?(^f{F!#0uKky~ev#S#=ul zH8tOauP1NAUsk6B-%Rr{d~10Me@&epd>764;op=G;ocX__u(V>^S{&g2@apamsBT* zAFjup!;hDj{#HES_o-8cpRV}|{2X}zUqPJ^ezoT7@SEgKctxEy{65V`@PEo<`0DDU z@c(JP3-?E6|NHQD)fvJUXg-72P+EZ(EJR(n|u!6Qk~KR#dG|g=6(2)@(TP5 z>ICr9G+%@NPF{z9S)C^Q3eC6RH_IdVSJa8&k7+)EKO^tMUy=_Tp26FC+%fz$btZ8C zR&!5g@NcOzhkruz-h;*S+Cc8ZzpYLMUe|mTp6csb4IZgehwrIQ13plv1>awt2>xw# zI`FYN3H(5Hy6_*W(}N!`AHaX0P6q#-=11_0dfHRYd(kHEB78Mp5x)_`0(d6 zUxw%Q`w;k%>eS$++su6n;mgPy@c*Yy3%NMbAl{ex0$vf~b)XcAM#PEYO zpTbYkI(zUVG~b6GEg!;vsm=&~yynLapTbX9Cx>tKqIq6(_{r**{!u)yU#nAwpRM@{ z-2bHc{j~sot~w$7T6OC1TjWjn@6~C;@6vqa@ECrvIw}0Xz49h}MR^Opsyv3T zEl(WYh2N;>)rW7T&H%ohJcHk%&KSO{<|pvIpfcBxAoO2IlK&CT%9U>6LkXk*76X(lsXOgH#FabNAfm&IdwYl6Eq*gPnDAEed^5NfxiBh9xI;jr_?FKr|MMT zP5X5mzL`29{3&(n@E7Gx_~+DV!{7U7vkwt`1$hkLQJob2Db07`v7T2SzK1$P_{QpF z@Xh68_&(}P;d^R+2H!_Mhezs^9xtBbk(&46$IC15{nZKJXKB6$zeHY#AE-_feuL&) z@Z01O{0Hj9@V{$5f&W|Hg&(F)AHKN$yI=rcUY@}db;j_uG(Um&bbn^>lhm2RH&(}c zqIixYb$s~g>QvyLQ>O~wRbGRitxg?&pynI!W8^LPK%EGFrsg{iPvBRp(}nld={bA= zzge9Oeyutqhfm;Rb!PCJ)X5$0{j+$E531wCZ&#=6@GAVD>eS%(suMcA0e?=N7W@fy z+79o)UsorAKd(;e@E-h~+net<2Jko389ICfUs|0Be5t?a>!!nVcv&6qU&V7=Nu83z z%kYn?Q-!apPT=qmzNR`2_yTpB4sXLhrA`OFt~#;9Q~0Lp^x&JR(}#afK7@Z(oe_Lb z&5z;x$*1t`)XCw;YJLv?xx73pp6{>a75I7b0RCk?ZV10l^L6;2*X)qDhh zP9DSC>ZI_+^>aa8__Fdo{JZK5;j3#tgRd(e!w*(x3g1riGx#R*Is9mKN>3KgYj}qF z^%LLW75Goo3E*GSR;D3{M z;7`aC`1N|+F8n3U_u#L~2k;xz$>7WU)qGtW!IzUy;5Vx?gRiLh9A1%oPZ!VYHg$aX z0?n5lUWJdXgoeuz3vhqvJ?s?&iV zt4{3j6uzoDJ@_f=^c_BgudU7qevUe0hfm=fsguKp>dYNpdbW7JTd7lq->Oc<;Q_p% zP6*HS-=}qlH{rXf(}w%&ncokJ93I2>RwsoA>U166hksX{A-thZ=I}B65Ot>TziXdo z4xht+q)zGI#dCa49pB*<_(|#n@Wt;kUoUD7ufxw!rwLzKotDERcwe0uzOFin!@KZH z)#<}OtIojT8T>kR#_&DWnK*m~zg3+%{9tvw{}j*Bx(~lcoeKOH>Qo(GgFmcJ9e#;A z4TrbjPpK2Z?@*`X@C5#nI$iioou0!7aPN!e>rO_$Ti^dUd;)*BIy3kO)X5$0<;8Pc zUL7C4zB*-xSK%K~rv~3iozUS8_?qgp;NMiI?eGqK19cMkG3ula@4-K#&H#RvIzxw# z;9pQ@0+00mJau>uZ>r-xS3Ji*Xgwu|m*IPD3MGk19D`QrKhPMtD*tmjp6 zcmN-$6T;`})E(Z0U!zVNUjCH6|8aN>|C2f?{PWtMuEYEAzp68Y*R-C@;bZti>P+F^ z(0XPLpTqy9PU(f>IkvPO-{BSb^XdffM4g($>+mFOj7 z@4}Z=rw_kaoq@wM_{!>x;kTB2vv_4FJ*fVb4i;2Wwla`*(kuR1gM=hev_ z?!8nz#{<>z;d`l5c6b$jggQ0&K(Eu#;SKol>a^fvb=nT^z<;4m0?*Y+9o~cgMx6n? zyuNv#A3A&lzd)S{yr$07;W>P$j`wo$99!y?9A1Xss7@8W^;YI}5y0R0tnmckFD;eS`B2R})jzQc#`$JH6ZyXuS`K7~K6P7d#>Gk19Dzs2)? zUY#=h9Ca!V58$t;6T&Z4r|$43+}qK79c;rdQzvqG3|~T>6n>35U5EGKORF=4-=I$B z@G*Qjb*AuJ)tNbb4*!rkrFroj$LjbFufSJPCxAbpPR-$U_yTpB@K@DoIXr@YQk@vS z%zftlA#r#YzOg!e_{Y>4I6Q-Iq0Sf{s55c+48DyzbNI&Uc&`-C(Yg=cQJo6>i|SMz zUW4zZP946FIt_=n;CrbP!M~?Y$KeTlUv;|h9zi|2ThIzId!b;=H}!jDs@2A`@EI=lh@nK~`_GwQS*-hrQ{P6B^TofN+0{pR)C zgP*0&0AAMo(BUKa`RYvItEw}FuP@Kxm#X8vRy@aTHD7Xg8GemARroII1P%}3H>%Ts ze_frX!`txN)#<>$uTJdn6n?ilJ@|3z^c_BgKd8;sguKhug)BPjlA@F z@qC|FrwqSC^A-3+9>8BvCxkzt`8xbR@+SN>b=vTney%%$hw>P{*iPpAs1*LH9=Gf8 zKK$M44B<;p%dYKIhp(tk>5bw!Zmf>)@CtktbprSf>eL)whc8g4 z32&>@a(D#)q&hMD2kIma@4`1$rw>0-oq@wM_!jDn;peI|arg|rjXHDqP3n06E1siu zAHJhH75D?{R2^P}@1{;2o~zSvcniLlIuZOGe>1O(j>8l9zUp-0A5^F3@Bw^(buxIM z&dA{t_z%^Y!M9Q;cewXv@f?p*$A>r6DLcFhKTe$*d^dGMhd1CqQ>O+0KXuv;@4!!2 zCxIvGqz>=Ff2Ym>ezrP8hmYVFsxyJ#sLs^kIs8g>JnyB2f4^*L;s1KxM4ghu%kWH{ zD*Od?0*8n2Th(d6OAnaWSJUBb_+9FB;47*VJ3NIys7?=FRj2RpA^b6QM(~Z*89RIm ze_EX!zN0#GhnL<~Jl_}8DZ{_2PQ~E?{55q#`2OnD9o~d5_9gxL5ByklB8SKDcdL`a zf2B^>;eGhB>I~rnbux#K;VY^$h2N~s%;9tRD(aNpUOdN%I=;gz@CE7w@PDgQb9f#8 zNp+g=#U3=T)0V>{_{QqQ@D@jNzd=6Nk^>JE}8>Z?BHG zSn(XK`|#b=slfM8r|R$;d@ptC@b9V9aCi&8uR0O@2z5FRPvHBj(}ka`PS4>3_z%^| z;OD9{a`*&(lsYr`b?W2}_uf%F$K%xT;rFOhc6b&3Gj(e4=hO)u-hiK`P7D5yhs^7= z?eGr#EOiq23hJZ|@4?SkX98cK&eY*K{8DwicNX`5Lv>0HFT<}Jl` z;Qyyi)8TFS?do*k$Ep)MJcZw_P7nTTb@~n;!XH#;1iw<9vBRhE$JEK;H>opscxmzC z`97^q89r90;_v|ef;u7mU+UBy-h{uVP8m@k z`_;+dACr&a%c?Vle@gQ+_^0J__=@V3mMEU%R+{(WJIgEZRn!UKk>+dgedTrdn(8#+ z2Wq|rKSUnEKc!9#Pc)yvPnUP$o2t`?U!nN{{5p9C-&&n9{0_}e;19@W@P;~b_`fyp zm5S&1hTMnmqD}?=zJHjnJ5~5cU81jYyCZVC?CKNRVRZVakAN;k;5nOW7L_!x6$L~4)>NUp5u?z@!?H%%J4)#A5w*% zs7?*Ok2)cIsTa-uH{d^2rv=|voi_acLl>}HJ`#e@*e!>>I~rCMP~g&_@U~I z;J;930{@Zbr|{$DIs7zrymuAP@fVse!G9|+!+)(#6@Jlm=6MD1i`5C?XQ|VGKle3r zpPTTj)M>-dQKtjHLGv;ER(T3PU!5L&togpfhwzKk8Nu&WXAFNtK80VZP7Z%o^K(vS2)uB1}I{e~Cb&l}c)oH{3cB$z^4o~6BJZ$=1 zcvaqq->b(R!Z+7^=I}9mqRte)g??^r2H!^gIs74YO7AJ2S4;E0!z=J7)Cu7GsZ)a= zCa=SvRi_F6rRG}>kKiw<6T@%U_qU0|yYM&E>BG;}dIs=o48F-JW<4W^Pv9R?X9i#A*QS#@+*`VMUaP3%!#7cglfjTw# z*6M@~Z@@pPP7A(+I&Ft{;2W!x!1qumb$AcHg*pRxTb-f9NAPXbnZWl~XX@}AzN0$c zGR1TJknVrU;br)4>Qv!tsuRH1lZWuHsndXm-!k{G3Ex_sHhf=oI`AJXW;!u^M|D#8 z0qXSNU(tLYzOQ@;KUAF&{2=GQ~OhapQBCy z@2V5RFV%b<{zrKeex^EY`0bjH;P=a8_<8E2@ZWsfoO>7km^yv?6m3V%)WGx!pZn03zKH>*>6Z}A+L)w~b?u)G4lQ=I_5K=U>DX7W1xes!AgFKNC7 z|E4^GKcG$wKV0(({Fm}B{1J8f@GCSwfZrj{;4^i`@Fj0G-&ao@K7;>9ojLq*J+Ak@ z;yGIP;kh~$_)MLu!)x#t)TzUtQK#YX7W`FpBKY&_bR3?*-`+IePj=z&)PL{v96o@* zTb&I4L3KtBpTOU*&J4alo!sHxvc+>;NgW@)fjVV}SK%L5rv~3zozUS8c%V)T{#A9_ z4)4G>QYV3bTbd}@FUcjIy{Fr)$!h6JjWlaQ*wA2{xx-~ z@DtSu93H~=Q>Ov{l{!s_x8Vn>(}7>8PVDd$9;?%X57p^AdWm#eh5uZg z9R83xbBC9fE1vJK)G5QCQ>Wta0Nz(8gfI4(`8r;AcoTk!I&JuS)QKD(!>?5*h1dRW zzJ7J#74`e@KdCc>ucrCT;bZt%ohiKbbF-cqd=vHO@CVc>Enhsx@0_5&Z_wct_~YsX z@XzaUYYwl&|E*3FzKc37hezLd>D!r#7&`Fh%ie@mT#!!!7M)EUE1JyPHQ zID7_QUY$Ao0IkPcp?HqgeRxHk3j7dtst&Kg*HEVpKT(~A!&~t6)QR9{sMB$H0^d}f zF8l&@dJZ4J>*{3iYtf{dh%EfcsT^%3(h&pA5SK)iBQ-eRPPU!Fk zJW{6xpR3b$cn5y4IthHq$IbhE>hK=?XmtkgvN}VDkKiY$Gl5ssnL0d&ch&JeP&~&^ zsZ(-z8U7n}s_;$K2^=26f3HphzO_0{hqvKZsMCReQJvV~DLhlB2j4@TzQc#`+teAs zzoX9B;ZyiM>g4c4)tNiI^ugl!{zIKI{6uvs4iDfnbwceLuN8-A%ek;7y7 z>*}QNkvd(6_u-50YQCNh;SZ^kIeZNF)tSPdQ)lMzIsAj_lvXUB=WjFtmg1K{1fUl;T3gS4v*k9bz=AebrOen;h#~b58qUsfx|QSw(5-GJE=2q_zeDK zb>{GIspEa9c#hV6cuSoM{1A1j4zIz#rA{55sMBzG3x0q)5&Q&oIu1|ZhpE$r_tfb* zd;m|>$>5i(GlE|ypTLh(X9mAT^SQ&lm5S$>s^i1|rcT-6Rro3D)ZkC56FR&BKV6*` z-210_y|*3SfuE&L0{?(Isl$8l^VAu@*HUNb@Dcn1b-a~}`@exYCHP3ck6eZi)TzQZ zRVRSgJn+9o~eG)M>--QYUhF48KF26h2j_ z>+n8&tj-YrtU8&)$MAd9nZjRIXXfxZe4xjJ+BVd{7vDW0Qs zAO4~`75GopsXDv{pQ}@c|4N;P!&~sz)rsI2sMB$H0{3>)?_0sIQ>W+f0sI~6WbnVJ zGjjL@UQ%ZUe?*s%J@|*z8NffI&d}i_ctxEFd}np04$t8qQ^#ASc#iw3Q*wA2UR9?G|Dig8 z!$bI*>NMcTtJ8FN8y=|BfuEsH?C=!6o;p4FK%KtBhwz#@Blu0~j2%9OZ=y~PpQtl; zcZI_GsMB?LAKp-B z2wz{F%;97BPU=kI+o&^h_#EC;r?hJE9QRbmcX$Q9yE*|pQm5wdI=rP$6P~Eka(D#) zx;inur%vMVF1)QyAAW^81BYku{nQ!5?@(vr@EJT(XAYmL<9)n%j@Etn0qRuX&#O~) zcn#iBrw(82Df9YjIJ^ZvM4bq}oH`wcC-7LEE_`)$dJZ4Jk5(sx*VGv~d;(9@nZdVH zCwI8FTJao@SI38cO`WpCtMF8v8hn3sLWeitKUb#(KU$r(!#nVQv$XR3~tF2>-o04fw0-G#%cC z57gsT&fMXp)r;qwsZ)mU zq)x@*0sJO)Lij%F)E(Z0kJM?yf2dC6@ECrFIw|}Fb-E7k!^i3j;lEZVbNCp3k2+KM z#p=u)K8H`#DXmdF#~ak~9bSPyq)q_8Po0{>>+q>MP59I5v>YD6A6F-azo|~*@Gg9& zP9MJfGv@U^aCioPMx8NyEp;XipTTo==J3tc@fH-%(YgQv!}sS`Lngs-Vi1Ad}9O^3JP zfjS-dnd-z2PvPsS(}Q2BPT%1}cuk!V{8n|w4xhp|Q74B_)R{ZH^oip6hU%2zPpVUK zcmUs0oe=)^f1B57-Qi7mU7a@kgX%;MkKx;@lfu_hr|a-OyrIqzzL`3i!^iNQ)S1F} zRcGe#IlQS(Y3<@U?xT(mKS*AI@2*Y&|GDOC4zI&o>NMeJtJ8vCE|1_}S0{$ws`

zU3goaKKuc7{y$y!9ra9EZ~uQpkf`Vx1vDaP6wn~WAT3H85F{XA=#d(lgz9hviBcQ{ zH6q9$EmTJ-MnDKf2#PpVgA_+ZiHbTPa)P3aC=#T7*N@pB)|&VF&H3Z3Fj$A@u=$MD0|lfdWdnU(k){=e#3!2hcrr%~L|+=U;bo*LYJI{$ptOS}m`PCYI7 z%IayCco%-6dVKhL>gkvG0RC0=4B;*HMDRW3WB51J6T^Mwr|=WxGx!nr|4{iV{Acny{37)<;eS-#gJ<$K{1Wwa;gu|ZetIR|hksu^1Nghu6P9=c z|ABhO@HNykDe)=%TJ_A}pHNR);u-vh>Ty0&+;M01R7$)GzgazX`2Ol?l(+}~iF(@b zzIr+(-hA^puybo{61NdX=3E{gcKZNfupTK>23_o6; z!cURU;pfUL>lW|%Qn?GiR^EW$DsRH?m3QDXc^Cc{c>sS#K7hX{2BEO z;YX_{f}bQG!=G1A4F8VuQ}_+?8T@s7<-gxDhu@}r2LF}ZS--f~8`a~&|D=2s{dD|gRnMZtE6w7*J@r)Kcc`aU;tlvA>ha+Bs;5=r9r)qu>A_z-BYzJ45)a@rvu+fJ>3%b;Uo0~@V(VDDDfeDtez44VD*ekJcduylfaKs&#c7f@K`+y`2Xp1 z;e4#Pqqz&8s;35jMm_ZsZ^9GxwBTRT`)ZeX7d}&u4?jgc{SqI*Q}qntXR9YF@iBa^ zo){jfXIkPjc&45?{3YFUR^rZv#T^&wap9GJ=b!g#iPz!I-ud@qP5701UtWo~;T83C z;Ww(MSK@uRtDXV;HuZ!h9>J^X8N=^W&!ohs@S1vN@Q2ltmUsrQtH;@>xZ}+i=X+I3 zyb5oqrw+eEJ&h9g;OnWU4ZmMKof7ZCKc=2O{9*M3B_6^zRZj$eR6U~-pTM_J&lLVQ z^&};p!namW27gvPixRJFT-VZ3=OQlg1b&cuQuqgTo_UEc;D@QFvPp5r6V&6DcnyA} zdK&P7dYUEPf`48;9r(A@(=Blyeyn-|_=V~jl=u)HsAmMfTs`9wkKtcYPXZsSXIA2K z_$lgHz<;D3XVc=2<}Unn_0-^Zs;6G!P54>rX~7>*PrJms@blE;!)NO0m-qmFv3iE^ z$J7&*_!xePdSduf>Y0}K3_eoN9R45mWF_u+#T|d39vAND*X3%7*WuTxrwME z!*5bg7rvZ&dL`b6-=dxY{5|RkOFV+#rk*jpuAWJWPvLi~X9j;vJ!y$&@O#zcY*yUy zdG%CEyb4d$Q-{0X$$!0Ql(+}~wR+m{52~kA;yw89)YFG=qMo3{L-?Q66Tx>;&F z@W<6Ng&(G#q{LJBlj_OfC#h#q;+4&d`~E{cRrvYpsg-yG{x9`-@axpmD)A0{p`IT6 z=j!oGJb=HXo)EsQe*PN5SCWt5OYD>X{564pSos*fi9CV7K|Lva7v<;h&&U_>H>#)d z@#2n;>3&>zpq>UiQ@#oRn(`g^=3D2V-!A-{%KPwTbj|?&ZRH2>v3v-Br+P;4ophct z{5JK(@b{}Hfj_AH4E|^N9KNP{7Vv*7?`%=rWvOTLdvoFIsiy{ihw^p!2jxxp#_DOo zo65K0Tg$ufE!E@0cUQg-KU6+|Z>OFi{5a(!_-XPnd>8e^@Qan7!pHJC{C0T;e@O0b zS={Rhc@@5pH{oykPrk1QUs2wLuO;unJ^28>qdbHkARog&FQ34_CZEFh(ml`ML*-NW z_vIP`1tEUcsMEM4MuD@r-gCDJ)HhkHY@_$~_fx8>r_v*h9 zya(^Arw@Nl=Lz6%d@g?;Lin-jnZj3AK7rTeDf|TWcv}_kc}?Y8@b%>#_{r+&!8cXj zhqvSb{8aUX@Ew#N!grUC;Ag340^eWx7=D;MfnTVe6uz93?>>jWPriU(s-8-#c+dMO z@4^q>D&M^Zze+s~_|ZCl6aIC13x0!oI`9jX@4~N;`|zKrCxG9f{Gh~#@VnJBg8xQ6 zWB8xtG5i7bB=CPKKP&M${5R@Zz?WI%yEt1HcQkk5kE*8zUqwCj5^utvQcnxMj(XZ9 z-i1G_9v{BBdio_kfWNGsA$$k*L?u3kFWt?*ACBQY^-N2A27jx1=J4ayla;u$O>xKN z)#JiXS5LLX>+qG-(}aIVJzj~o;j5~r3%_1Hy%O)k*Hq5{ewTW}5|7~Psb>uTrFteM zK80_ro*6t#`d`tB>pDgb9w0bHfUWIR`o;v(x^)yP{gYTlAHvBEm=U+EECEkPY zrJg=~CG`X)9>PDPo(R6CdPXHafqzy#Q}_nzNlHA0AFZAYzL|O!C0^OKxbGL$Q-wG6 zeOC?Ml{esDQjZ58DBpsQ;AiiV zKZh~=RP`kAbJdf=&(}Y%nZqwu&jNmtdMeu$cf4A87yd(e4StDw8t^-mZ>LuQUPu0_c@1*m$OS}vJg?fB=P3P&C_yGPJ^$g*@&J&gR z82$(K#PF8RGcEBM{BiZn;l6sZ5_djT-0^Sfap6PtR7<=Le?~n`c&r|;#M|)a)zgK~ z)zd5SKHS+i|9)}+uYNRt4#N_U;7hA#3~#AtQsPtiTh%jzpQh(0E%6M#yn38=amT*S zQz`K(d`0!t;gNb8CGNpjR!R z;huVu5>Mfqswac@)Uzn@$_~YSKcSu~JXBAu#2fJK)Z@X&>S>jD2fnjY11L0`9A)vSV?_EA)AHOS}d@Mm-Jq zE$V5Ocnf}ldOGl5s;67xKKv`{3E+QL&!EJI@Ke<@f-lrFh7W8<_?hZS;BS5*|2}F4 zUrs)UU!a}^d=2HDor*hdAa~)HsHX-W>-=^2Ox}cFsh$?RvQhrIXu})wF8q4+`0yQc z{yw~?o&o&F>KVc}`b2)u5&UfZzGV#mtX_8tKUO}2-=Wu?!_QPcgMVA@>|EUIKJ~cp z>y@v<@0Zu%zfeyT{+RM!iMQbotEUV9hkAM?-iQBAJp=eF>Iq9cg8xB1WB6NM%=ev? z_!OS1X9j<_deRcl;D1(+vrBQu_0>};@hbcY_0-{?P*0=8J@}LAX~TC^Pp8Ct@Tb(% zhaaY%pu|J?)9Q)fUs2Df#3%4))H8*jp`N6~Q~0y$$>85t&!WUDyB7CdsHY15v3hDH z-hjWL9uJ}n&j`MQdd4Ll!{4Qz1ip`YW+gs{uc)2{{228(o#Kw>F8safslm@sPrbyO z@b{~y1;1Q9?Go?8S5=P>|B-t7B|d=H)iZ=Ypq{A2$MDtF6T?&WOiO$QUrRl6_}|r& zmALci;*M*p$A!P7o@$BL;p?iW34hzm`PU7v#M|%<)YFBps-9kn_u(H?&j7xrdcqQq z;2W!F41e(T`9I&Ez#o^-;2Z08Q}`$48N8*}b#^cAwVU!4_<`~&d*xAGKzf_gIe zUzA_K|0S>NQQY@y>Z!tCQNC8<4S1*?5B~O7^3P)n{vLS`{vo*!Utb==HY11L0)DA_Dti|1|I6xeOS}fZMm-JqIqGS`uaLLkKT=Nzew*@L z_}y|Jeye%{_`S*x;J=X%;XhN)2%aiGhCd~b;fZ<@_(J&^d`bIrQ}CI37I0U2XRqRp zE6Ho{56SCrPu_y>DsRINll$;5%lq)5dH&xkKw8AA%@?f{1krMH}lW?4F05g z=J1#G=d`mDclIvcbEX~_{?_m2uUjqgI{bO{G~w@2k5}St_!9f)KTmbxYpJJK;(hp= z)iZ!^qMoqCBlvRa8N;_%&!ohs@RigvgYT`Lw8S&`2h`*2Q{3?g^;Al{3jeTr>hKfP z(%;6>i36RCEkG_qn;i-R*zre0X$Go2%oEG zSmGo2SJg9tyBp`fuO63p0zXwfDZHtkd5JIJ-%?Ly-{Ov4^|&QogAdiyfDhEuEb$in zJL>7cC+g{zxDWrHdII=dJ%bV-!hfKi5xlxdzSp?KWBB#zN#Jum|FaTb!2hMZvtMy9 z^BVm1ugiaati#_eZ^74*x8a|Z`|$ncefY8RA^c2v1iwT+hTp7vi{U>~ehR-&K7-$( zo;mz4%4hKBfFGou5Pq~gf;C-C#+Gx%ll6n?GTIiR@9PvjN&J@PvI*YXDZ3AqPf z;(+}7y*B(g@$7w~V%D+d;L{E&L8@Jp1h!EcZ^;2%+s2fs`C7W`p(2mUej^x#h_@5BEq z58#`tCxpLw$^7fa5Wb>(1m8wI6Zo3S$M8+$34BNOr0`EEKZoxtU%+=)PvtYk9Y3eM z3;(>l2H#IT4fyfOH%q()KU6&(_*c}^E%5;E?~(spIVkZF{7m(aOFV}AI%fjETs^ZA zpTj?|o(243o!>dAxJyr8|6Ta;>Z!r6R8Jiqs;3G6ih5e`>-E0cCEkUfrXC-DyL$TY ziOw^CpQWB5JW)>sPt`MqU#OlK{zvspOMC{uR6TQeWz+oo$qe3 zefSaT2}(SK|3^I$yr=gyD)9;YCG||R{3#>C-B{rpOyFmzMt~WVZ~j{Yw*LBua|fWew^~{ z68GWXP`+Q{L-_g1Mie{;#NES-_jdom`sXI_iC(t}zxwbz@4

8&Jc@O?ac^|%qo~r@;k)!k14dH)P&lvtM`2_x=d)l<&ZglK0>zsizPBlJWuk6nO~$ zhI%6SIm(aVBl!e=rh2CETa-`Wcgs`w1?tJ*zf^t!e@tHaY;oU9)Ki5&t9-4*8}KXD zcs?!GEfr348_h#3i1O_jaSa2LFk?4!=X*g8y9JhW}FT!+$I9!ylCo;eVA!@PEr=_$%@$ zeCappxq`d$3|^JH|69Db_sOg9HRMfrQ|`e(F7LuWCGWv^lMmosc?dsP?tHFz{?ExP z@NXQF-%k~Omp-R;_zCK1z`ri{;E8(L@UxWflz0#Rd-e3;*QqBc@euxmdLnqDo>7TU z;Qv(56#g51KbF96)AubI{C9fY1^h92^{C>Gf0ftZncRavCvU-Dk@w)0H|5WR4_{s$ zz+coogz)z(KZJi+K7udxnf&MU349~vWBAtc1pZd_r0`vppTqZ;FW~P~Pvz+1{rk$h z@Z;q*cuhSG_$kUa;pfU*@YU4QfnTb87k;hWhkryp0sL0w2k?94L->a38Np}DkKuoj z$M7xGlfa)*eg=OAe?a*wfczK?oZa8LO* zd^>p;-cyec-(C4W{2=)Nex!Pa@S~KE;3vw*@Gq(-hJRD}Df~kD9Dap7gHPn{F~z;^ zmRI2q$(!)UAB(}(Y&d;s5D z9>UL8PXs?q`B8~a;1{W93O`Cc3EbH{|9$2Z9;qjThs);A;R5~@^;f=7-1jo|RN}@?H3|av%O5^#t(O>plnYcglzGSJX3tud4hQ{!w`huN;(rpP9h7Qho;CT|S4u zO+5?vA<8?)7I!>e?!s44PYr&W@^$!sS@8ht9%>2s-E*Md=2&Z@SD`rhqu); zfUl>XA^bk|MDSzPGlp-Xo*4e9dZzG&d%(}eG*ya)e+yba$^JzaQ1zrOU~r>Un8KU6&fct<@U{CxF9@T1f-h6n1Iz`v)S zDg1c#%-|#YbrXJrdNTM|)#Dsr-0?Q$EAR*8Rru-Zsl$J-d;>n0d+?!p+VE$U@4%P% zpZwRa9{fA%>BIj=`2fD6JcM7No(R60@+0{A@(KKU^-SR(S3ZHaw;rhKi$J@`7xw@SPR-&lFS#6$R2$`4C?0^dvb6T|nn`-I=6drRTRD?f*y zBwxVqS5M_j#rwQic^7`UyaxZZdK&Pnm2bjtl(*o2P)`SbtMXm=-EtrP7xe`22b3Sc zAC?c{e^<{4o+>|v&*d@vS@k6FO!*o7dHEdvl6n^KCEWb$h7%Nbe3RUTFV)MxFRH=c zu6!N-E_oCF7WK5?Ybf6?@h;p|j}QNddio_kfG@9}A$%wObEOErk^0B*cdI9cd&*DY zE%^+-vU=w5y_C=3hsm82i#x8a9v41Pz6w8AUWc!%o+doeud`l>x8WPBrwd=Grw6Zn zJpcK(58qNf19(gM5bnz(`1a}`vpOiFwT-%UL;_>a_+mUsr=T|Lg1i#y(?o(lY4 zc@@5|dg}00`3C$Sau0r}dfM=1-j;ve=)hNy_u!vbPaob?K7el}58)@OCxRcS{0M%G zd;&jJJyZDU$|vyS;D;#Rgnv%ng1_P5{MY3UJW#$1|GM0VFRPvaeva}3 z__yUlcuhSc_*Ke};Xjne@HN$wz;9Q6R^oH`$JDcc|4cp3SBg6h_3MQT-$*?*_+IDa zKX=sO530Wj-%LF%c&yiLmv|SxwR(K`Z}htT5+A^KP|pzlgnFV9AH#Q7PYj>yeN9Vz z2H#&jbNCB7Pgdg2SBpCysvZ~a>i4&+C0>UgrJg4IL+bHLybV8IJzaQpi~ReXUWxbN zC#z=w-$3UHOFV*~s-7|2(|IN(K82sHo*8@#ohL2v41SS%oUavk?CCrecqp&JFIP_; zK2g2_Pvsu`8uhf{l`ZqVI`D?P2cM{?58qYy6O?!a?Pg`f z<>&BNzJUKmJ(aH)?|-Vi3s2-V_@C6%fX|d~!XKA+;F-J&e^DO5EBZNl0AEf%g4g6@ z_~#DEzt4=}bG@$wzPfs5@b%?$_%rHRz&BIgIi-02EqM+8DR~{fo4g5sQLo#AAE102 zeu&(M`|>`#FCW6aee(N^;DPcnd?25~FP6{XOCFMcUpj}6mCxX-Z=avX`9|@cmr;)k z|EYSa@J-cIhcBm|Cj372c<_DI(}u68o-TX`{r+vQ#0T&_ln+aM3_no$Nr}(kpH)6B zap%w5wp$W!>b>dD}1>2(+IiFzuh6?fT4JyrO+>Zz4@1HOfNJox77X_a^fzO8zC z@Q!-?5)a@zt0#mXp`KxhkKlW%X97Q3Ju&>>8Jt_P=<>&A#+Kgh%RW!5>jiyTrTj zE7jw}pHol2#0T&j)iZ>@q@Jk6$MB!1Cx*N4%%9t7iO=A7sb>zat0yaQCoJxGzj|DF zQ$5uZufu<>o+f;v`|;r0sJ{*WgL=B~j`BVDf$~25FX|bMKS3VB|E``f{1oLU zB|e2etDYJB67{4dp23|%^Y15}Gm1Ojtey(|HhC5PX7$wJ_bA_h|5EP3msL+2{#)fc zCEkO-TRnaFKT>z1iqGfrtnHN|C}Zzp2F8xPX=E>J&O{re5<(c zChDoeS65H1#2fHWsKPg`P^~_6r0r%BYIkULqMe1=&yaxY*dK&N_tEXAwE%?dm>A+{|>6W+;KV3Zm z{8{x3N_+?(s%HdW_FehsYh2SMef2nJ6?Zgu;onwI4Zfv% z>LuQUU#gxKyrZ6WiFe^wsKbFg#0T)PdWP_$b)Kli$M74~6T`oyo@t5C;J2t} z4!=x2S&2Jm7k9i}Judu8^;Ao|4*!{Yn($lHKVhASRwzMPD*?Ve@s0ycvU@ViD&R9)#IE~+_A2nN{LtDnR@E*E!EQ~aS#52 zdfMEG zc=g=kj!!CID{&A0vhpqXQt!^6&mR1r|H<8lSCtRpYsrW3_2nb@%DT4+d=uqk`1bMy zzPfr+_#w*A;a`+5;OnWUGA!QnIm)|m_Y?U)*Q>!dQBMPYsd}35hI(4?t<=+j-=Ll@ zysI7`-d0Zl|D}2c@PT@U@SW5%g3r}6hR5oO;k&9QfxoDp89YDr=pRRl#Ufn8x4hQf9)iZ=& zq@D=wsb>uDsV9ctpq?q*SI-Q7sCwq`d)1S{egmGQP)`rOvAhpIMm+;~NBI!GpFDzpQ9Wb$QOZx?iQd-~e!O~S z@Z;2z!WZhv;DLIa3yM1q)l-3AA+N$uQcoRzr}7PWt(AW+Jos1D(}w>}Jsr5Go*w+` z>gmIuQBMGWMIORWRZj$e_j~f^a0Kt_JQMio>Y2jVQBML7)sw=%rJfAFwR#rtiFzs* z7WX|{JyrM~>Z!qxlsDi*^?2~pm2Z`J2Y!KidhjdM@VnJBg16K&h7Z&e!+)-x1irg^X7Gu6 z=I{s9vw-ib9_QlXj&t?6@CVgXgZI=^hr8S4do|&|QBMp0S@pExo_f0Q->JukAFG}| zyr-T4{EzAx!UyV!;GufP@W<2>!_QRD6h2nZ4E}_A=I})M44%oIZx?s`n|fUM#X3(F zev7;g|EGGI@ZT!$!KRaEAV^dRrvDisly*tz5!pzJ@~uT z(}utCefcgO_`BpicuhTh_{z!$@b%;&d{y;C@XeGT!FQKW;2%=Y6n?n!3H=Mm{G;me;LiK=y;|^h$vf~()YF5nsk{&0R35-T zuAUISt@1kWKTbU@c%*zA9?QG% zlhotGkJUZ&;dAv2;HRi(2tQXn5&R1I7=DI&V)###pTZxM&*0~%XAXC_%lFFQO}TSv zamNePZ!vc^)%s2u9APgJox+NZTMyC>B0lOZVx_|_u*HlX8@lmAHtpO z^LvZn*QjRLY zD}_I-o(z7j?sEbEfxL2gaoaSIs}~ zK72KK0RNkMLinc258->rNAOHN6L>?r#PGH}fj_IB6z(fOhY#fo_zUW(e6P6URCyPk z$!qY}9g+Y1)qvM_$e*_+{Ij~B7JO;-bl@%Zbm1qf$A`Z~JptTT&j3D@58=zIX9Pb_ zuRDfcB9Gx!^(63{m7l@ykNIqw>zExZ^+MF8uxKsli|Wf&6n(hc7R0!at~< z7JRDvY{OSpPZz$XdVIL2d>{U4`2fC-dWP`NDj&fI@-h6Q>WSfVy{{?U-7$a8XYftc zGlyTS*UjM9$(`>Pcid7vF8mJVtMK2+>+nyirwM;rc@Mti2lMBl4c|dMUHChc@4;7* z_u;##X8_+w`4HaFJxB1p)H8;+m7l=-@+o|O^~~Ur@+mx#XYhm7<6KeP(b*|~PAc$* zybAxUdg}15@(uVv?!k{%PaD3C?zsaWtEUG)Ry}=qM?C@jVA~OXqIx2Dp!^6v(|IQF zuc>DW|E78p_$BfbewunR_zlW0;8S_!%HqCfs;3H{DPMy-JLk_s1Ad-*Jb0#_7JSKi z{yFWyzpb7gysp>v;VpRpzf3(Le0ja@5Wbdt1iw-}6Zp2u$8cZgN#NJ2CxsuRo;f^J z&jNmvdMZ~HcN{D4!c%z-{uA{y;3w)lO}M*DzIzLPhk82ji`CPGH`U|A?@>0ef9h$$S5;3N-cU~${&)5G@Q(6*cpx9Z|D~QGd~=;Ag6|<8!=G1A4F7`i zQ}|ftnZaLC&m4ZHdNO#T9_Q-fj!S(u|M|j&kJVF!FVs_qzgay^_`T}!;MLvo-P`cD zsizBnLOnfrQ$2n7JJmCQca#s|eR%|5Nj+nDr2GUvl~3U-t7isZD4)V>o&0_>_-g8L zt|{*5d?^2XRp4#)RN)_1PaXaq^)%pp^?300)YFC!mG8h~c@Ms!diwCD&J)15lZWul z)DyuEQho&gqI?41N*NU?x*wJE92t6JE^A%zeBHEgE!UF zfPY#&9=xY~3m(cl@O{+NgU8DI@VPvIAE=%XUfn&vhatQzAHfe*&jdbDK88=^3H(U) zr0@rI&vSUDo(23E^;E7c?)a#BT=;)%NBHsTX~5mp^Ur$|{(gB2ev*1R@HLh1!Z(%s z@KeBe*S&k0KTm9A$&D?1b>ry#_)}lpTKvNPvOg|X9hn|`4oPv zJcF;O9_NPQj^9+i0>4ULg|DifI{atKH{gGfd+@c?(}p{1<-2s??~(W58>pubukDfl zIvBt|s-6(;sV9QB)H8zbrk)9WYxPXwzIqb)(dtRzJE|vx57o1PpQ)b84~zTmp`I!{ zR!W=ykqBw$0z0e z3A|_LhkGaM{FCDO`*wbKq*UXE6TalSxwqhNmv`Xr zmG|Huk@w*n%Lnis@)6vZPv9rXr|`=f`TLr|Pu?&8`=94T!QkJpaFxZ@{~!=I3d{pW7h6PamEt zAHvsa>i*%i)AIfV?kS(ayUM$_7SG>Sz5#c>sq@44Qr?Gml@H;E+WFzZ={i6BOLl&^ z6Xtp6r^WL}asHoO_!;)T;E{UT@bB9Dg4fQ-UpItbYwrslC?CU9oil;oWA7IpeJg)m zSO0&NSHEj;oZl*V{gPsjYJl;AcFU_YB~-|52YW_?r*RdnWL!H_G!d z-1|j-4^w#hVD1S#ejxW5+?S{DKt6|Ozsh?ucvtxaJo|Z`cWy6!j#7CAp2$6TP2PgL zcjx_Wc$DVegU5f&-KYOH_deYFQ|RJmp=~+ z_+5HVoI8s5Y|lvrZqJDex96k^x96k=x96k|x96k*x96k@x97xz+jG)_+jA1Y?Kv61 z?KuhI_M8mip7xF44fzOe&()Z&=V}7C=PHKVbG3lmbLHHLef3;b;I5u47jDm$e^;@m ztFPMuJi0s2h47i4w=q0bPfS-&0-viVg9qp2zg{@{|Ff^>wV&nZapBLYrvXpZ6|0@Yws8S+WUpS_I}~^ ze&=v|UkkXMzj9CU{OQU0`J3?Ugxp(jH_$!6U%LnRYxe+u+`dnMe{!?@xt+rAe)?yx zJ|{Ey&dSf>zfgVw-%I(*&x`kPh`a{h@ZehjZ~t@FcUbAp@53)r zK7>2p(fz~E*+%Dwca_iJW98lZi|3D(Z@{~s%=_E$=)1aq_?MIq;lU+3Km5naC-7ML z41TZj?gPd1*Dllj!%x^w=ZANd_u&^RAHuzn&JX|Pr*!{t=ZZX^!NYc*cYjele^>bi z{2b-m@c7ES--n;EW8NRay{mM7c%*znS3ZN^puC$D&ma6C?{C2GQoap$uGabCzq9kh z1Leo?r|tZ3=bF4fgDZ6@bNcEz8^12*PgYL^Uqb)hG50sc{9EK9+_^qK{}}#)`V)9p`4k>JoWEWM zf6pHI`*QVvx1T#w_0-_8{&NHVLG`rZ&JXo<48ET7L%4Uv*Yod7;r`{hyR+iG1?q3W z53uvVV>=K0DCI-AccbnfK2$z|r@z%b!>>@jrvJO$bD;m+fPZt>e3v#n(0P3LPt-qv zd)Mdp7Q&P3a*yESALl-TcPF`z;pw%xFW{S}`RBs<^YY<<R=xqhQ~4JB zpZn+Uw*#Ll@55hGK7ik+{1E;|oo573l#k)>P(Fb_q`doB@g8ii8hkU|XA^FFwcz&W znY(b?%ZG2R{sG+f8p3Bf{}^t2#qdXzpTTXfIXqR~`E&7pY%dr7xbk(l?bU?Om2bms zuP*#4BSQSLpr1?^iy9N6J^8D866Ys|vqV`3Btf^57%oJ8;{p2fs@B z0B(DQ@UikEc&hh1f#0C~0&e?O{;zlsC+hQ3gWJ9h__F%v1ueMk+kso&hugjZyz%{f z-yz)g9l_UCK8D-A3B0NN9B%tA;2SCL&WrbE`_|x|@=dty+k$VQd>3x}`tX+WW4P@b z!w1@T2Dg3Z@D;SL^JMY;+P*H_@^!fF+k|hc^S9x)Zx`NHz7My32k@PhkKnfN7~WBS z3b%b{@I95!;I^;x*W$f(m9N5W-#YvN<$G}3w+}x}`-X7aH-dle+xge$3EcLb!jIGM zTc&W^H-isG`RBLtx8gn6zE$|C$~WM)uLlp6@4#)}9{eoj1Gw!Q!iUO_;I{7sexdRS z-1bf3k@B^tiuYstHsIg8INz%Uw|zTs+t-KNz5(3!9l~wj5#08T;kIuAw|(bu+jjxC zecito@5lD7!EN6r-1cq3ZQm~3_VwYm?*MN54&k=%3~u|*;g1}Y@9X@d_=pwI6FZu?H*wr>izeKWZ2Tgi&|VEb0#wr>M&`+9KOw*$9* zdvMz~fZM(y-1Z&8ZQlvp_D$fnZwj}47jWCR^3USE*}g5f?c0HWP`{q~aN9S4TYd<) zeMfNHH-_832|U*GIfvW63;3PNyU!Hw!S=1er^+|swr>l5pYmO}?d!u6C+qVz5eFt#cH-g)~W4P@*h1Z2X6cJ;IZ-n-1ZIOcPc-E+rAU{RQUvM`=;>wlwZJY z-^zcA_m(JMgWJ9h_(RJ3aN9S4ucm#6aNBnTw|oq@eG|CtJBQo83;0aucb_ZXgY8>` zKcajSZu_?2sq$U8?d!uISAGDueTVS5@?*H|8^fPceg?OF=kQGV`l5J0wr>-Du73V) z!)@O#+aQpjVBDn24hPU+lZ&SGKJA-eldSAGGveJd{%@9ik%YjE4Q0q-l{g4@0w_;Jb) z;kNGx{v+)h!)@OLZu`#Rw(kONf1i^3V(}hq-x@s7`I~Utw*^00`7Ye{_2C2M2XNbW z2tQT%G2Hfz;i2*~xa~WKpQXI>Qt^IlUl%@9z74m1yYRign}7e)hugjbxaA|b?K_6s z_s>(f?K^`%qVNARxb5q_jC)YN3b%dh@W+++;I?lYK3Bd6w|)EYr<4!jwr>Q_l%K$D z-zogx%2!?~-h=I1gw|#qX+c$vQz9D>}^N--R?*#s$@(JAbP2rBd zKUu(S--`2}SO5G@<<q_+r9zZRX&E>z6tzR?K_9t zz6-eJ-PaZG!S=1eZQmx`_HDtJ)A_q_+t-Izl^?)u-ywWO<;QT_H-^`gpTTY4IeZo6 zoh6F*WBa=By7G0n?c0Q}p?n{1`wrkgT_OMH`w`ss9mDt5^Du?mzB9P(o55{g=k>*V z=;{1bxb0hqAEvwqw|(1iU-=%~_U*%uQa*&+z7f2y`~+_MPT|KXpTce53?3-oSh9F; zwyy_&Nc(o+wr>w^U!McG?Hj^v-x1vQoxo4l`4hP9o5BamFW|OsWvSvloT_{cZu>Ui zq4F)b?c0H$rMwTfeFOMV`61l)9lHGt@?K^~Ds{9yk`^NB*@-w*YJBMGTyt8!ier#VCK32XCw|$%N8*u2>-1eQp?dPKmZu>fw;yqY@6>j_1;S+uT@4;=~HvAUldvM#g508}( z;kIuCzf<`M-1eQqr^=^r+c$&Xr+np&#e1`TtMEkm4&3(b!LQZ60o?Wt;kNGxZu?H) z)}O#_-xU6k&cA@$zLhr>?_s8V4Q~53;EyQZg4@0wc&fY)w|xWn|+QN9nieFyMml%K+F-x+*k?VG`EU*|2w zd$4>JZu{2Zwyy`becNzX=kLL7-#&af6awr}NsiubUF&R>Juz72Ro z`4-&v?ZDSn-iO=10lcaF5N`X9;2S9)!)@OL?kPWq+rA6<7Roo>;=S3vE%@d7^BY~b z?d!uWKY-i5L%4naJcir8F}$WfpFe}!zH|60$~$i>zF*tdh1Zp@!)@Otd=2H>aND;F zZz$h~+r9(%y2?jz+jk6aDxbk^U+3+`dpKTyexnMved}=B*Mr->ZMf~*gWJA+xa}Ll zZQlrP`%d7t?-XwPrf}OggWJB9WsCP_`&Qw$Zv$@odT`sf1Gjy9aNBnTw|yt@u|B^E z-1bf3mS4ba-^y~ud$4_LaND;5-$>7A3vT;%;GXh6-1ZINTPQz-+rA@sOZgaX`zG*h zm7l|H-vzv_yt{ny-fZ6*d}rmmaNF01|491|;I{7&Zuv3X_Ko4T?+kAH&fy)M-+4#z z{o1}Rd{5=;aND;D?<(Jh+rC}+0m}E`w(kJmQ$B**zGL`d%1`08?+orMUwvoser(@5 zd>?&%_TaW}8*cmd;I?lcZu^FC+c$#Sz7x3ZJB8c6DcttW;I?n2TD%|Iw+gp?8*tm# zgWJ9xxb546+r9zZ_6^~-ZvwY{Q}~_w{4U_OZ{=OZd$4>BZu>Uiwr>k=`*z?*>G|~G zwr>FMD?fzWz9aZ?%ExfqH-QJr&*8T30)Ddc?h3_wvwdsuf$~kb?c0K%s{88Qk`r!!7UBitpFX&e9d7$J;Zr@I zZMf~*h2N)qA8z{&;ED1P-1Z&AA5wk_w|!^unerLj_I2J@ythY`uflELI$il5-1hCm z57g&3gxkIm-1eQoZQm)}_D$ioZw9x0EAKDfgY8>|+rACB?d!p9-wxdN?ZIu|0B-w+ zaNBnTw|ysY+c$yRzA0V%)>bawkL}xlKc&xa3vT;%;FkB{wr>EpeTQ({cLaZ2=a1pG zZvvkyKZo1C3;0vYyQ>uM$M&tkGv%9b+qVV(xAI-M?d!uA$`9bS?-2f?@-w*YJBR;A z`#P%@->>cK!YyBi+rCY>?c0XizFoNUz5Ktw--p}21Nf54M{wJB46i6Zh1%Uc@J*;w&7LfL%8i5!N0Dr&l9-qJB8c6DcttW;I?n&gT;HW zeXDTWw*j|(J-F@Lf!n@4xa}LjZQl@X`;OqY?*wl9CUDy~h1qi7;gI}@D=rZ&f&K20$x+zU9ETzwr>r-itw^`}X0s?*wl9 zPT_CT=QoAhz8T!|l{Jd@VEb0#wr>M&`+D%Ep3e^4_U*wpQa*s&z9HOGegwCDC-5zl zPvEw13U4XDfZM*6HH-JQt@1Uv?c0F2mG|McZva1CpWh+e_8q}(-xzNDCUDz#4!3<5 zaNE~it9TE#Zw+qyHsQ8!3vT;%;kK_2w|xh2+jj`JeaCRyH-_83Gq~+LhugmO4;Syp z_HDwK)91Ghw|%>C%lF~7?*MN5MsVAA4BuJL=M--H&fp#8Gq~;RG>Z4Ir}9;}?OTU; zmG|JbZySDq@;$ii+lTj*58<|N1V2pq6mI)w@FVs4t*l+V2ivy_w|yIM+t-8Jz8$#j z+k@M_0o?Wt;kNGxZu?H)wr>KreN(vYyMWuim5&tf&GxOqZQlmm_HDs!-wxdN9l~wj z5&S*+{KjzGH-TGz4!3<5aNE~ir+5#xZw>D2`E0^%-xmBR<-2g(*N69&AHZ$jA^bSy z$8g&>h6l>e;I{7^ezNk;y2bmkeO>rK`8M44?ZVI2=eG~HeFt#cH-g)~W4P@*h1?-1bf2$LRArhugjjxb5pUi}zsr*5I~p6K?yq z;I?lUZu|Oh+jjuBeTQ({cMP|EW4P@*gWJAyxb5qFw0J+ZuM4+*>u}q*3AcUwaNBnP ze^mQMaNBnbxBL`t`_ACDZw9x0osSjoVWj7?3b%dh@T-*f;I?lYK32X5w|)EY8> zwr>QVC_jPQzEk)u%BOJKH-pE@H#RKZo9*kt|E7IAaND;Bw|oG%eM7kIJA&K36ZoAv ze*(9CQ}|T*1>E+nY*f65`;@Q2ZQlkwQN9JYeLL`nl=tDbZvdYuKZM)9BlsiA&*8T3 z0{*P_bvG{FgY8>`TfPameOqwbw+pv@eR!(#58$@%5dOIGW4P@b!{^G+;I{7^{*?00 zCdK=)eO-8_d>wB4HsSwPz74m1yYPkb5#072!(Y+9Q@HIrgIhj>+rG}G#e1-Qt8m-5 z4u4VS_u#f~8}8`uSLwlR-#&avir>ii41?OWNbcn>QoUxVAe4R}rY z7Tosjz*kY;hugjZysrEZZu^elYbYPXZQlgmP~P3Vcn`L34Ss<>zfHL9+k)G^UAXP* z!)@OI-1Z&9ZQn86_Ko4T?+kAH&f&JN^YP;S*uE~@_N~Kh-zMDlZNqKfF5LF*!)@OI z-1eQqZQmLEjru%haNF0}qIeIMuflELI^6d4;I?lYzOJ6n9^Cfr!<)*7aN9S6Z>0PL zZu?H*p7JT&_RZj1C|}vKcyG3E72Z<50k?fU__oRiaN9S8FKb_K;kNGtZutam`=)T) zcLBG3E1xLdLtE#s!EN6Ld}rlbaND;7?+i=^r3%7myaNBnPw|yhH?K_6sz8T#1b+#_v!%F)6R^hg99d3CKZu_?3 zwr>w^`}X07>G=%dwr>RYm7lR(=X^d?^3-iq7EU)pMW28}-~5@L2iECyVblSS8O_;Wc?3p1nWMH{nq&_ZHk+ zG4~FW`27dxp1`lr&ld~$0~_c0+P1}e`0ysVx8R$~efZx#oB#Q@ zA^c@|48LCIsclz$U3*_GxSijJ+xdrZJ7)~HbI##*PIvp_dF-4$`0BsZ-)9TI_@aC^Vb&c)Y#*j^XDfxHcG$@_3$9>MK?rtp#S8T>|h zb(i8f?H&Snch~&$HGns^&3y>>_s>0oyZhxnf_uu3;lVa}egaST%{_)EN9I0-2M6Uo zgQs80J%#%xmdlo*vwKLb&ZRf!o)U6mDNnDxWUCuAQd~k8jq!!N-%_Be8fd{wc{WZ9~FCU)H z@_YafAI?35C-Mkx`511${?6g{>#w_K@qNAa>nYsMKY-iwIfmQo&fxaC&R)gWwdb}D zx97GEw{u2tJLeQ`=dA5rd|k`8;Fgcz!4ludzmI}9UZ48}o+=;1?RDpHdtG;*;yLa4 zZ^G^Q@51f-lNg?^ke_n~kKEi-cun~^+)+M*r|-&pD&6Aw@7B*jRrt^4b@&DP{?mhp zN95lJMsT~&DctTega5C3ynT!3d_d<3;p6r5_Z7i?c?`Gbe-5|j-`%hHx|VOkEg!*O zd%y74-Y?wxtNRyU*ZMuUeI9#o`#gs5hQ3co;n(~j|9-?m^bgj ze{oTsAHc0Ygc~Zs)Aib7V6f_I7V6f_ML-`-Iy$r*J#x z3~uL~!|j}vLyGqq+w%#xbJpS6ow`rBowEhEb9Uf%&K}&(Ie-WDe8TOVBf34Ga69J| zZs(lA?VNMCowIUi@jktu>-mJ+IqUF9=WN35oGrMWvjewt_TYBT0bS=D!tI z=N!>f-6!16IfdIfXK*{`9B${V9A3On&+Zd$=d8mcowEtIbGG1i&JNtp*@N3T2Xwnn zxSewZ@7jIB?VMA%opT1abI##*&dL$R`;6_^Bek(X? zEnQ!a;Ogw*>Kx$e9O3GW9u>|r`xAe^!_}F<_uiSp)tST9S-{mWm&8&NI_I;p$A_$G`S@!mrE|uFe9k&JwQ9nyz`m)!D+g-r2#` z*~8U2z|}dz)fs(HIM3so{rQBeGl4JO?em1IGl#3QfUC2FtFwl0y|aO zbA+oi`o3_U$3ODt6Ryq#zR=I_aCPQzbrx`SmT+~}@U3??aCNqHe;;*lb@p&|4sdmj zaCJtH3Fnz;p!aV>Kx(fjJ`jd z=TSdz!qu6;7y5o4uFf2;&H}E^60Xh~zSYm0aCNqH{k#cRXAf8B09WS-S7-DC;XD_= z>F;;AIurQbJ5#tibGSMSxH?O?I&1o0SZ4!QXA95#dFxH=2CI!m}ZYxveX8@M`K_{=*yy1xH}t8;*>bA+oi`oVCXndS*s zX97R^-c$IMdBW9Mz|~p8)mhUuPq;c;_*Ty+T%A2!odaB*BV3))4~6qQ{<%M&aCIi| z>>d7m!qu6>)mgyRS;Ey>!?)hqz}4BpXWrS-^?btBIl$F9!qpi)Hk@asdBWA1z>k0F z^MqfSCtRHcT%9Fcoi$zagsZcKZ@sgFtFwozbAYRJgsU@pTsY6;pZW6%S7!px{Pzqg zT%9>wodsN-C0v~~eCwSJT%9d^rr(Fq^?btBIl$F9!qpi)KAdN!dBWA1z>j*r!>`N} zuFe9k&JwQ9nyz`m)!D+gdcVWf*~8U2z|}dz)fqh@oagba{(QpKnZUEb-|ujB=5Tcu zaCMe&b=L5$cQ$Z!w(yyEc62?TaCHuFb&ha#Mo$dqnQ5MIbtdrR+kKw!EAxb_vw*9! zgsZcrYo2g*w(zZYc5rp}aCHuFb&ha#M%!?n$G7OuFf2;&H}E^60Xh~ zzV*%quFe)d)7K-qo=>*%xH_XB4(FL^o^W+0@T1=E@GJ9#tFwTsvxKX&rfZ&X zb++)W-tTaA_HcC$aCMGwbw)oD&hvQk=M%2Z1fKnkzu)2N%;D-R;OZ>l>a5{g?`+`e zY~eHS?C5$v;p!aV>Kx(fjD9qnXQp|=)tSJL`tuI>m3hL|S-{mdBW9Mz|~p8)mhUuPq;c;_*U3%EKCfZndOqRm z9N_94;p&W@6wWjIeSg2h)tSJL`gs$6Wu9<#7I1ZzaCO#n?`+`eY~gGD{ZF_$d$>9W zxH?C;I-@$A=kfP^o^W+0@P+<916-XsT%846oh4kIHGJ!x4P2cq-Tz)?2UlkgSLXm% z=LlD4^b_GcGtCpO&IG>K-~WWGGl#3QfUC2FtFwl0^?btB*}~VFCtRI9T%7}4og-YG zv!4v-slQ)c!jJ!G&us-?y4Uc<`|W%~_wy}$<=(?jAGYfZ@Z$%}M|kEw!IvMj^U;&T z*Sq%f3H{_jpC%9*IB@C={+g@ zj-PJNc?Mtj`2w!zyI(xV}&v4B---UBK{L20ipWy0`emb1bqPF|S@Z}@T6ZpydQ~2)h?R*JWe+5^64Of2ySAPpX?d-lC zeD3`{{OJ7yeDjZXo#+|ieAFMq)t|uCpTgCj)1PejE#L?5FX`T2!FQuwr-iG(gR8%X ztABv2e}u37`I+E5?~k4t&U5emF?@ZUpStSvA&0BKfUCcRtG|M)zlNWF(&pB{FZcC% z!q@jU@8FA1vi=dS{t2%B=vkPLfBwX9^(XM%Pg_q4zxd}*4&V9w3;5>m>^e1E{S93G zEnM^O;Og(;XMcVM_{sZ6_}2R;_~o`X|K!=>Jk_7V)t|%FU%=I0!uQ@^!8iW=*L3f1 z;MujUzlW=TfUAFmtAB#4KYC6$pYxM!?=k%J2=fGUj%{S93G zEqwQEyKe{I`18}lm)<|Xvro49M@={%^~Z4aCvf$raP{Z#)zhrMfS>&Hr-ZNk^QWS} z+pg2X)!)I@-^0~Ez|}v}pJDf%;1_>>qUVP5T>1QC`0`zLogA+I0_8FTB5_-_k$-;p(5@>W`im&PV+*T>T0BhIv{@9EF6y$|r6KL6n}@1Nl7PqF^wXTo`^KZUD5hpWGUtG|RFUSa(eeCho) z{G!i)`1W?z-_!5s^MR{>gsXpot3UeLa6YS_v+Kw3gMa=c@XSAdQuygM)?dOk{|c`D z8m|5ZuKpH2d$IL*@Xgk|haVp1^MucS-#mJLI3M-LaP=o}^`~(4=kT5X{8GS|kGGx@ zzVrSHzI>ydZ{h0i;Og(;>L1|hAL;)2Gr>SEPjK}|KZp5T%dQi{)t|sGt({NdrziS6;n~B@ z3;5dmYqSJ39kO==fnBzy+4Lue$aXn_{#fJ_~py( zd71y_F!SAPRne+%F0^B;cj=f8(9{P`c?Gw+XH7|uuiFU9r{W*N$^Dp2V zfBsAO$)EoUKL0YCe+yTC2UmX&SN{N4{|I0C?@uQ9@^RJ^{Q~Cc&rb|r|E8VK;p#8o z>M!Byui)yh;XD2Q2fp(D7QXZT4!-^!>mT9jpWy0`UKGwp{V`nq3Elfs_{pE29DeZU zr-09X!}@Eu`Wv|VTe$i=xcYne!M}bD@Rh%RNBCKv|M2x6Sby^3aGvT<;p)%f>M!8x zFX5T@SMa$%KQ(;g&rbute5v*KaP<#x^^b7%PjK}|FA3*!*5^O`^jM!KeCzX1;ir4r z`4X=F3awdzhWC;D>vdw{Z1$aP{|a^$&3MkMPXD{!Z|@&p&FzdCt8*hA+H7hpWGUtG|S+ zzk;j3hM#`Q=GMSB-rvF(-rvD@cem?|aP?1c^+zuY=cE1@uKonRdaB(wh3~vShcCUq zfbZ{S*Qw#^Z{X^0;p*?;>hIw@pZ`Gj&(9H_dH)1I+|~M%mxuFIe+pNB4p)BxSAR+O z`B(7M#^(v2JTlrcZ{aI_{=-k+-^0(|KfvdIYOgDLWjG)8$8hy0aP_Bf_2=;2Gwr?weD3`v zeC^Lq1wXynuG7NR-@(=2!__~))j!gG{u6xT{n4w!d9J)ahA&UnpTpH(z|~*E)nCEY zU&D_tvimmh-FzlYDge}JFfY5mbJh4WE=3|D^wSAPmue-2;y_df-EC%F2fUkT@<{ur+Q1b%wD?LCEG9%Y`x=iXnykCFA) zaP>EE^|x^KcX0Lh@Rk4h`v5Yw21kA5|r&*o)z-x$8~&z}T7_s^dcp8cg=r-ZA& zf~&uVtG|J(zoq;0)4>n^{PggZ&wqd~2J4T0Eu4?~W4QVgxcXDL`g8c{X?EWNzVrSP zo_T)-KmNI0r-iG(gR8%XtABv2e}u1gcHaqp@#jB!Z8*=1fBlW&tGC#7a=7{nxcW=D z`YX8lYxv2(|7qyn-@?z{-@#Wm_RoK~`X{*hqb{6}`eV5I6S~hog)ja2$>AsOFW}1? z`R6}e{S93GEnNK_T>U-#Jk_7V)t|%FU%=I0!dLqK zAHMPLA8Po)`y2S|hSuN1)jz=1Kf={N!POtVKAg|a=O4pY-k-qt-k-t`|6=do60ZIV zuKpUX{syl8mhSWK;2VE_dic)!2l)1{tUvnoa6amf;p$J|>QCY7&*8hL*?kN6%KJU*<{R3S6BYft6{x!ju-{bTAjc}f;hnUCk*}s|RaP=2(^_Ot< zS8(;$@Xf32z72ft{VhE6{tmwKpT9=9`X{*hqu&hYqy8AK{sew`rS+%qgZJm~rS})` z<6Z1JHC+7-T>ULv{T*EWJ$&i&AK*LhAK?q{pWsXHPu>vDQ~fDi{W)Cy1zi0leDPYF zTLs^Ee+{2|e*@p&*{;*W)jz=1Kf={N!POuARyd#C&)9uq_{#ee_{{rL_}u$TxcV!& z`fIrQ8@T#g_`#q54!-J`SAPdre-Bsx09XGAKmVHDcY<%cKl<%(o+s~*;cM^D;p#8o>M!Byui)yh z;m4O+e*<56e+xf)e+NH(pQCS&e|}Q<#rt#k!TSq%=KVEX z{S93GEnNK_T>U+K>F?hGe)9eizW4qKeop=S|KAPgss0qM{v59U0E7)jz=1Kf={N!POtVF`UotXYIZ*-TM>x*85ZV!uw0O`YX8lYqW|^-PvGiL;p)%fEAKDhC+{!eYwxe%Gw*NV z>hIv{@8Rkn;OZag{`oV(50CMA{(d;mmG{T+%^mH04p)BxSAPjte+5^64bNU^_if-i z?{DFm_jmL!vd^CpuKo$G{^$?F`KUjJt3QEny+4I7zu)Ic_x=KY{(L)M!`0uw)!)L^ z-@(=2!;juSz~|mS!WZ5@!B^g&^x-_!pTgCj!_{BF)nCF-{?D&f@QXh`HGJ;<4gC0d zHvb;3{sFH35w89TuKwsv;e2MlVE2vTC+|<-Gw)B~OYbk?>aXDHui@%%;OcMTTkr4S z2Y-Hg_~k+NdI$LabIqea4CkZ%7_R;VuKpCR{v5vkWxH4`SAPdr ze-Bsx09XGApS{TXC;0OFe4c+4&hzB`F?@RmJD=0<=ktN9zl5v5f~&uV?|;><-@xbI z-@=dH-@z}RW7iqsn*Rh>fAr>XKI)I*>QCUyU$XnA@Pqf~bnh?VYwxe&>TlrcZ{h0i z;Og(;3xECx_|E%B_}=>``0NY(`~N=<=c)b_uKpaZ{sON4629^IS9I^M;XChd;K$GR z-~YqaKfu*L!qq>))gS#yIG@blzcGB}{Rw>Q{VDu>N9!-)>aXDHui@%%;OcMT*)Q1h z-@!Nf{HJ^W0N;O}^+$gi&PV+*T>S}L{V81iIehUN>o4Fdeg4DO-e18_pKaG^;p*?; z>hIy|AK>aA;cM@o;JY8Np6C?LbN`*@G5qi`^Bk`J0R{P}O; zJMZt{$1k?$XN0SNf~!Az3+8hzpATI93H|wYK82sWKZkFTYX{WV^$v+O$2pNI2Ne+*ZD0#|SJ39kO=FT(k#KZdJ6fggSTDg5N$KjiSU_ZRS$_t$XsH*od0aP@a^ z_4joD`ws(rsn37-$@?ex@zZVo$=kzusy~IRKZmQofUCcRpPp*>t>DZ1n%D4||NY4Z zzV!YcuKoe8{t>SJ39kNV2Uj%{S93GE&TXm zyKe`-e2sYzKYRZG-(2i`^q1j$)E~pupTO0h!quO{5B~i_0Y7L1|hAK{yq+kGea!TY0kg!4Rle+-|0+^&u{dxPvPp%;p#8o>M!Ao&hA^mPu^d{XWrky z=icAL)jz=1Kf={N!POtVGn~)f`(yYTlrcZ{g?H z+Wb5C%KLlx+4~3h;tT!vKYtU>NBuEe{Rv$CDO~+IeC_iu;B)`^v4k)D`KjRBJ6nHC z_urp)aP{|a^$&3MkMPCoZT=H{qtE}p4d=Pn=Rf=qTYnDM{0q4HOSt+gxcY1O`P;3( zfv@!W58vtYAHKPj^^b7%PjK}|XUs>x|ADJNfv+^5mnfLc_^$&3Mk8t%*aP>$37|v(y&wmVG>id8A#`{zF;dcJ{4_AK$SAPvxe*;&4 z3qSb$JNR7R|HIebKfo8>AC2LB)E~pupTO0h!quO{cQ3HH74+}&dBTtW?=Pz0mp``i zEnNK_T>U*<{R3S6BYdwv|Aj9fVLj1%!+9ROKZeh4;h+C-^%rpUmvHr0aP`;l)zhuB zf$#kDr-d*5^QVLFKE))gS#+I3M-LaP=qfmHzw}e(>ifho5}@1$^fHHC+7- zT>ULv{T*EWJ$&)2Hn#yj_vdGX@4bJ5Z$8A}8 zo7eE2_c!p(WaoRh`UklBN4WYYxcZ~_hx0l7qTM%!@4P>OZ@oW-FaO=HQ^M6>!PQ^G z)!)F?-@?!S`P0EK{`~jwjrR}m)kmy9`an1z^~Z4aCvf$raP{Z#wLkv_eB+-#C4AwZ zKNWoUht}W1)!)I@-^0~Ez|}v(_rGHIo!}dN{(mr>=gQ|F!?&O7pZ|3K`K5rXzl5v5 zf~&uV?|;qu8~DomTX^RE9sKfH)<42E{|T=C=%2&+s6U3QKY^cp{we(6{W*N;{RRB^ zsn%b^)!)F?-@?`3!PVcx*FOINzSHMFeBu2Qe7~^%T|{`D%N9 zD)`Aie`@&J`y2S;v#q~}tABv2e}t=lf~!CJP&l9cldM06ul(~TfzS2(ANcwX)?dQa zU%}O1!`0uw)!)K5{`ISaA0BJ3tEWHIe1K=4WBt*;g!55<3|D^wSAPmue-1zR_YVbp z=lvyo@BI}#`%LR^;p*?;>hIy|AK>aA;TP|p=s#q0jy@dDb5;8DMxR-K4p)BxSAPjt ze+5^64bNU;Jq`Ti{VhE6{tmu=hh1lctAB#4Kl;~jKI)I*>QCUi=h=Nz_|l)BobLSv ze1AQge+^fE16O|wSAPdre-B^#=g$D2d;bVuc>e@H-^}`xe+%cS{uHkM9IpNXuKp6f z@cs(Ee1h$zhF>1$&l`MuUF+}Z{{7DYSN{lC{{&Zm^zWF@ORXn{&%Hl^pS?eYAOFFw zQ^Ga>3apTO0h!quO{Gw(0p zC+{!eNAIuT^AA{m3)lQRxcYmz`UklBNBGJ=e#3h0nadgJ0ff*BRmJpWy0`J{rzP{V`nq3H+qr|G;EE^|x^KcX0Lh^q;f)4)BflkMNm3|KXc!_~-w}!g;Dcg{wb@tG|G& zzl86+zk+Z4>sJlGJk*~z`0~Had${@sxcW!9`X{*hqyGr!Gxz5|hOfLofuFrUg)ctC z=3m0qU%}O1!`0uw)!)Kb-rv#v`RU;&?;qfcYgvEvpW%GeAH&t3z}26^)t|$6{{AiC zOYbk?NAIuT=eODO)56u?!PVcx)jz=1Kf)LO^ZNup{IK;z{}s-2sn37->67ey4p)Bx zSAPjte+5^64L`if?%TkZ-rvH{`uvA4y?=zOe}bz&`gk}W^~Z4aC-B{Ktv`jYyg!Fu zyuX0&KGCjI!`0uw)!)L^-@(=2!T|{`}20+3V!ka8h-Nr2EOwC9TYX{WbjJ{SAEM{Vjav{T+Pi{Ucoc6I}h#wZi$RKZdJ6fiL~{pDBFj z{W*N@{RMn?E1Q1}SAPRne+yTC2UmX&KluCy_{yLEk*?2w_~jFa5?!ugyU>x|(W?@!<}eg6+X z+|sU7!qs2F)nCKa-@w)1!gt=^!B5`b!!v#V51)H~blq@1>W|^-PvGiL;p)%f7k~d2 z@Pqf4@P)qrhi`6S^Kaqm@8Ih1;p!jY>L1}3|NeP`ue?9HUO3OC_s8(l&8je}t=lf~!A@!uhB_hO0lJdw)vz&(9pb@z2i! zzPp*dt{SfX2Cn`VuKo_L{vJN_?|%mPQlJ0uwLbshm$~&P*AM5Z{uHkM9IpNXuKtq# zT)S@tKj`~^_}=>)_{#fxxcUdU`bW6>C%F2f8-(*&d4CMQJkjR~-+F%vKit&LmvHr0 zaP`-4^*3L1~|=h%HG_{D#J7~LqG=i2*Y`0`D5ogA+I0(8(DvHlW?BuPvPp%;p#8o z>M!AY|M%Bb@PmK-s^M$zZ{SPs@8Rkn;OZaY>Yw21k8T>y=j{D4eCPcMeB=EoJiDRI zzl5v5f~&uVtG|J(zlE>;^Rt7`?_)hZ{PNZ21AKQhkLKZg)E~pupTO0h!quO{PyXje z1^jYvpC|n6{T2N920P!v)!)I@-^0~Ez|}v}{p;@pKY4$2vv8g#?~mb&8(4o1SAPLl ze+gHA1y_F!pXuNK3_o~(3qN{)2j5-a`bW6>C%F2fn}_pJe+*ZDLVup!H-+!KKZhT@ zzktuYzlN*7fvdlTtG|P*zlWdv`_BQs@%|CM_x=e!k8BRfEy8)u?rMLom%``nIed0^ zJ72)pe!he+-79!@H@i*?S5F65PY=($XQ2CiNBF{hg73X&am#Q%{X^e$)t>~^@XbA} zzkz4&Eqv#l9enX+cAXwRb06UQ%+5!j5O%63hN~xmZ@njl&%8f}tG|Gs{W>N5;9kLJ zU+Q~@tG|P*zlW=5fM-6p5k7aH;F-@exz+zX&lIkn9Ikm5aLu!XtG|LDeQq^;@7}<- z_w?rnzSi>t-)KJYhR^)I3H-da`{r=nw}9)uB|P){ zR`APL`8?rg-){@oeLJ}B+rt-r-vNI1`;PGS1Fa{z4d&+ijp4d)0-yPPQ~1g6o5RoF zZuc$Wdc74~_pRaQ+^*BWkAB}4etxi>@8P=d0M~s-_|fk>!4H1l=(d=rKhFtW?_ap? zo5S~h--7PoVD-{=!DH@|NT-}-$M z_}ZVVobLB6;JR-K&-}g>eB<}6;p^|P*W1$lz8zfm?coc*?*L!>eMkDk?Y_|`VQzlk z7_R#!@R{E?g|Ga+Ieh(4yKhPN`&Mw>w}zj;+OE^UGrw;OKR?XQ_wXz4U%2i&!jFF6 z3BL6EMxTs%`qzbo?yon6>%KXB@Aoa>3%_p(UzhfJYr5aJf$P34eCzk^;B&uk4?jP| z?mNOY=LxR+MxTPY`F&&f%w}zkn`=|zf z^!v8(v){LeYrg|r_Z{I!zwZP;_g96Yxw!0_Img5{Xf0+s^2HH@Wp4^>pZ}>e!hdxKHbhA;cGwN!_WSo^}qG^8$VLo z?>ExhhjYHh!_0T^8@iA1Te)YS7G7uJ-oQV{U9W>Zg}*+zdJb^)M4u7fS3PUEdOEmz zvPbH9f9~1B)ic1=lO^GG_V2UjCx;(D*UqirTR&gG_n%|u*YLHU zFX3DN&-&k*gXS~fyymm~?66<+so~n!8Ls)H-mCj+K6)K;^-MUgp4A=x=bjd>o|(^8 z*H=#gSC96f^XiE|FTAgMHgL_ShpT7t`QdfcQ^D0U!spMl&!q`|e42T5$ME_aKOe)- zPq*_4e7Q4E;d}QSetxo@FW{S>G%w*h_X@ss-@{++pEEuDweA;q`~3c@&*yj<&f!fw4_16;3bgzI%paJ{bRi^DnV zb;aveVTjlZrQzVz31hX48- z?ECyE4d>RoFX8)l+W7+h@RNP6ZsEV^=UaIH47<(={vrQ*H^Fc7a67;FlCbk`?m7G` z+&A!Vb8p~xx~E-#cGvLw>M!BTx7u@3!S}E7_Y3`%<_-Mf*JaXDHZ{eG_Sbqmse^1wZ;OZaY>Yw21 zkM17MQ~fbq{TckjJA2;BtMMIN&u0hkpKI4S!=Lh1_I`}+5q4gAU&C+mIy+y$pY-}a zxa#}iE&Td#GC#xheu?fG-uIDDwaztMbEvPz4{*(4fNKu3FAeXjIV5n+VFg#`9>s0XUo#r)s_HOeAzVh=eeET+^?a*D_Dw^Xr(W@Rgs>;cNFbT=T5qdS4vi`u=}_>-{*pclf&WeR~4`=!^aP zf4Js9!dJ1q-U+_Dxp{P-@V*N_AH&yvK7ns;;q!#=Zfu^z_ct*w=zhM0AKbTa&A)|f z{wKKRKfyKs#n*&$(EM}wL;ub`5BG4*Ke}&to%yHQ>y6=C_XNIjPvMt^&l7&UtQ zGhBPg?-$;8cNd=<{Ni51H>uB!{%Phle0vA;2EM+%c?(~Cx_JlR`1u~bbw9&3hv@6V z*Q@u%60Xnr0)DgKvgdpY|IPc`{9E{o{p;^+72a3#FX8jM+Uu?0TlX5ia&O?hwXZW=uPge7@H%>3OSoQF0oUu=!u7gZxL(%@uGcle^}5#oJM7f!>fzga z`QG8n`AimQuyA_=kU{4*mX*{URMRz>#E^; zT@75XtA+3GYxnKo8}IDlndS`N`uP#Q^Vc=|rf{BmeiFEzpA}rsZ3WlQHD|bwvpUT*+5+a z;oEPseO2(o1AV{nt$PDMtnGXYpOxkv{P+O#9=?2#`2aup`4N70kN!{idQabD*NNeq z$CxMdCz_}5m7mYy=gQ6(@Xe#mOZaMIUcncSGOyuFzi$JdyC31J7ua(+!XK;W>RZG4 z?0?+*;iso~KYac)?}sn_d=20Hgq?5TncugCAAiKockqLI4`29o2KdZ<_JHv9YCZ{E z^VwdFkMKofofCZR9(`MQ-`%t9dD=rkzjWTlXBkdWM}Z;F)_3U%trBH}J&^ z%v<>0y`w+h&iC-+bIk|%-hG7c+$Z?fJ^A)M@X>!>KEXfazW9!?M{~&Gn!^UJIW%z1;Rt`G z_a_eu?|bO{`Gj9y=g%kn{3?Gw;j`D9m+;Fg%q#f*m&|MU`ZvrQ`23aTEqv$K>EPRg zo$ujC?-}4rKR?1xetv@Qf6=ZVJvf}_^3~=s{OqqQfp2`CDSYqu&EY$rLjhmC!R}ka zXMVnduQX4(KMx)J?9W3__x=ICc&+t}@U!L#*YglPB%G(7hZwHsA%W|8Na1=Oa=4y{ z0v`zldLDYXo`(Uh=V7G#-e=z#&iTInxwMAA|MB*D zTf@K3e;-w{YhWFL)8x#27`Jel*;P>#~Pgd|>ew2N^*u(X?-NW_ye1Yrp zIWEIaeconpeg2nleQxjI?T`7_FZh4@Kc|p9EWGd0|ML=_`F~E~+wZdf|N6g2SN-YP zY^M8`IIsKaK02?xj5x2o%pV?hYA-2Vds)M^mm03U9N^l^0M}k-{<_t#y(Dn$MRU@5 zz1|k*^?FZmz1|6~*SmN`I3K;<9In^9f$Q})aJ}9mT(5V8>-EmRJM7W(kizvmwD7~5 z?en~YZ~oZ4hp&Fie1LD>vbpZq#GeC5|E z;Mwol^-K89uT#MfZ?y9@Jo{bq27dZ&^A^5#@8BE1Z%;qkbq4t2Ps~U7-1jwmR5)jS z{wHvK{;%Nr{IB4T@UIJdxZZa?T<^OJT<^Ph6L#u-m%;VEE8%+I?cjRfb#T4E&Tzet zqDO~4dLNbW+4tDF4SehWxuSoso!9?epI5qXi}Sj#?xXXXbK>`xYtAdU=3K!w=RI6= z?%|sA1+F>A{`%CTId9?WZ{g}c!PP&()xY?@a9`@r;p*SO)!)F?e}rpaBV7B+9uxN5 z?d|sc%nJUsk1$`uzuCQpAK&GlGw{`W{c{FB|6B78zIcau4?p_n$pF9j^+)*9eS&B2 zv+GCSAI{-g{E``7i8z4nO$$0>1oHJ72=rZ!@pp+p~EMKmM6{13zj0 z@a%(jzJnk9d=Fo1{_u-?^aJ5M55C_Res)jbdLB}^o`)Q+=b?a~y{Ckq+$;Fv{XS2) zo}UJO@aLz6>-p*63x9rk_*V0T>-ibsJ3l|c_54ItIDb7qG5qAuNdiB(r*J(#Ib6?A z0oU_W!ZXbuuIFbDzw&e34~DPn%J&KIE8i!;uY8{Xzw&(oT=(taKk?K3RX^ce;QGEd z{-Ln*0Y7e^pBemdew`BjbM8C1z8-b(=bi2PXZVZUqsN9l&-`mUzl8t1|NgUpzuK>} zh3o513)g&3aD6?R;1Be#Ba6p{oey@;;rjZuhrecLd+FiY;{~q0#E%c}tG#4!z1|Z3 za_tNLI`=bN&u#RC@Vz?3S_W?e0 zKf}+@w(E@W{WHxk@byodPw?5!eDhjP&);|GY-mS;BRl zD_=ixzP+0B`Z~<}CjRwVuGh7K>-ntUdOr7XJ)b>X&*ufM=QDnK_`K=)%;0)HOSqoT z9bC_62iNm?hU@u^{QDZsN6)9e-;rx>HO^~p2e{@oz%{qoGr~D&ZV6m-TfsHA3a+{B z;hI|y*W4~}%`Nut2Q>%HErV-r`u**L(`N=Cg%sJ}q4HIl(ob z39k7pny^#z$>EyM2Cn%uaLq?Q$7nu!t`h(GMy~g%ejcKqabDNSo)`A(=lT+^zdv?_ z>-P*JT)$_S|4ev&{hlF(Ki2;oZ4KA!s^RKC!_^c0YNPdY=w( zeLiO|3}2V7U&3|1haW!CpJ)0_%t!d_Hs%xj7@J4G5cX_tX&%F8H#Se;+0D&UxX$PB z`OWNn0nct~UcwiC-wM9=*Hy!hK8FUr^|`h1wV&_cTfc7)U%C(Q%sqNhIM2P`H-;bl zz6o6WO5xg94%fa4xaMENwXX`U^EF)iYT(*e3t#vgI=J@L!?mvgzVW$@aP4b?Z#DlH zhx6CIVtD4~bNJ5hTfnui60Ut!aP6yxYhMjq^KaqWR|nVm9FY% z63$cmis5UYTSE8eJcVyHf4KHlz%xI;h3ost7Ow9nPjG!dIl=Y)?c$~3>(%%FIb7d2 zZs7VnKf-mL5w7dZ+wi`+&Kj=QTf_Bw4{*KS0j{3e%l_y4CUAAG;QBs%e>FaPd3b%@ zH-YQ*uHd@w9(y=&l$``To%`0CE)J$#v)5Ae00AK{0)+W85-aF1RUzTWNK?R*TMr{)QKn-5gR|(g?D!BGl!?mvlu6?y|?W==dnLk|n8sIt~{bD%Jo$o7#YhMXm`%2;3 zR}R;{3b^)F!nLmou6@<;nVwI$_SM3*uMVz#^>FQLfNNhPT>F~fSLS~R=dXRmaGlTL zYu{G^*S<=)_Eo{PuNtm>HE`{#g==3OT>I+br@Ptb&j8oHM!5Dh!L_gGm%@2!Uol+! zO5oa83coUcxb{`Rb$$zf+}GRhqiXm|{qMIN;5y&IU*+e|@LN2jzv}mF^Is0ChuQl)BXTMGlU$5-^0MEQw{5!ta~HkM24A>~kpK8$Vydx9&Ck^1rR8fv9$9xb~RCwZ{UkJ(h6ov4U%lHC%ga z;M&&_uJ6M~xV{gc|7tiNeIK5}_4gRpaQ%DaYWV2?ez^nuL+%6om+xVpSF>LWd*0xl zz<=yZ?feS9bFbh(>%ND#?mhf9?icvcJ$`N2`DXVFKDd|gkNCfLeFwjepYPzea6iL8 z)jjIMo;$iP;rDPa;9ujuh2P)3g+J8&1b>YC1phXFy^Gg{osV+Q;am3&{ORru{DtmE z_^-H+@He{8Umy0o%{_&`$9)gi^WVeu{9oXD{^MT{@2lrOgX{S(;d=gea6SJWT+ja* zuIE4cjj%`0{}Qg}zkuuc-@^6$w{SiGC%B&f39jdV@ta|%p8p)K=YIp&^WVVr{2$?Z z{ztf;|M?rj9zFjlT+ja+uIImo>-j&x_52TTJ^!=c3VZbYCvZLgE4ZHj3a;mWgfAZC z-{-)0-)bHm!=BaJJcchGXr92&em;fo+;jNWzYi(k^KZB7mvEi0;Isc@=WF=p0p<;S zukU~0TYp_0{Ot4T;am3se)txE+>!fh)D~D@e1zh_o;o4UP z*S>1F=HI}zuNJQJ9bEhB;o8>#*Sl+V@q# zwXYhkeKm0HtA%S{9bEhB;hO&d*SDDk+E)tKzH<25_f^2P zuM)0(RdDUAg)e+x9bEhB;o8>#*SHE`{#g==3OT>I+bn*RXTzDBstPjKxk>ce?zUol+!O5oa83fI1J z_`>&9z_qUuu6{}`@)C2*Zj;o4UY z*S-q4_Eo~QuL`bx)$qCRtAT4@EnNHR;M&&+KkMf&xb_wOVK`6iD~4-d30(V1;o4UY z*Zd2(_Eo}lzJhCCHC+2@;M!LU*S^a*9h1739fxbe;m$J`-b&(|unBT>C2F+E)eFzFPQMKYzisuO6;_4RGyiglk_DT>FatB%G(_AH%h;1g`Tb zT>Hx5+E)SBzDl_ERl&8d8h-M9HE`{#g==3OT>Co1_2*O3pN4bOpHD5}`tzv*uD>6+ zh3oHqwQ&8tuh}WQug)iMov-1ChxqRU;cNF6etwvp@8GA>yoYCgodLf0>x^{&{%wLE z-J`dJulL{{!!O@yJsJGUfBynp|9ulX_=o+!*QSH(zb|I-*04v{&*5+Jf4-@O@4w6E z2|qvF=ShE*`2asX(tLy;{W=qT@7Ia`EbL$S`52zLC-9|v3cozUde(5wr-p0(2e{@x zz%~Eu&%=IQzl3Z4J$(0QpC>%~9-k-tcS^P!#x-{n;t~s~xfnnXGVkH%A2c7}Yxfa8 z_v=sa*<E{#p(a)#w?PKlwYqQ(|NV`3g#Gt&Pv8%Jr2YM|E&R$};Cf%o{wln_`seU(^!u*iI=_KG z#LxHe?5Q@l0ls;P`3PS>*?fYZYV+u?!~WHen#b_9djjA5u$@ogrynuT;afjnz!y)l z^EG_wueX8g^|o-m-VUzU+r#yG2e@AE2-oYK;Cj8$JHvVE^~P|$-VFYa{@;I5!u1^P z;ClW$xW3LV{wD0vb#l0#=N7(un(d{7XYM_G;Xc4uJG;&ZzxaM9_`yB;+pvG-9>X{8 z34HCI!cX40hHIWjxaPJvhduHf{uqD%Zs1RFKf-mL5w4#3yTbeG`YBx3Z{f>l_#WZ= zpEmE|TlWDzdzSAJzVh=E{PJ`=AH6&5pS#EKrF#NjxTo-qzpgc0UoUF7zK$Q@=eye1 zivd2l&;Bm#kM3sY*Kl3GhU@wV_~Yh(c-7xq9^fzZ|6YmN--q}87k_>d__N(t@E5yR z@YlHS;lJ(P!~ewn0)K~l{10L0liV}-d)!O7=DdSz&S$vx8@(sIufCox;kteS*Y%HZ zU1x-AU-N$q@2h>KaP4ai*S>1F_H}@3UjtnGnvG$P_Lac3uN7STs^Hq!9Bm1+VAZBVUPBkz%}O;Tyx&THMbtFxn1CzTmFHtXF1x}ivm7(FX6|( zxAPVJAr(&UmaZgI>SG%&-s52JM}!2a9w`~*Y!KNuAf}O>mS~4b57yg51Qxj zrJpb03*Tc2zxeqIp85G2zWspRw}H>Rr-kqQ`W<}Z-ow{^-vPe)C%f+mU-|h7zITs4 z6wY(w9>cfp8T`)owx5@C`0f0;TEXw&Ucm37=M%pAn7ytFp51hJ)%Qy^eC^)A7w#?m zcrCk52R~iYyob;HIs^T>c7BAPZ(=^db>HY;F#l`Y`51njYz_%L`>1&eUw*_qhhN+a z_{_b8&#z-W6+H9vEnNHR;76ZB57)j1_}Bc~+SdfvzLI|n=dXRG z@T1QmhihL2eDCK=xaL;DwXYh!_3Jcny}w%c+Rt}z-M5EpUjtnG8sXa41lPWze-G!W zeZ_F?D}ifYDO~$1;o4UP&wLIwT>EO^3qRk&HMb6~ef98}UuU4}>lgfdBmepZ*L|ap zg!9zCVz~B|z_qUwu6^Zj?W=%mUnN}os^HpJ3)j9n_|fOk!?mvgzW4JZTyvY?+E+A% z^W6G%Vz}O434HD6Q@HM%!?mvhu6>no?W=-oUo~9&YT(*e3)j9nxb`)|wXX@D`5dB; zhV#_EV)(+(CveRzg==3qeCF3F==%BvKi|;5e!+F$8m@gcaP6ywYhN8)`|9D^*8tbP zM!5Dh!L_gCW8wU@uM~dtIplEdtAOwQdxW^q=88wXYbieI;=1D}`%cIb8cH;M!LS*S;#a_O*xK=~edgcn|*y z|M@!kukiJ1zbX72+1zrt_FKT0e!hh3eN@5s*Rbo?@U#EC-@teNb9xIu_;otCuG7P_ z>)Cw=xb{23wciP@Jw_i7=czr$@Xg2l=Xdzre~wP!C-)qFa4+CH_Y%Hw-@^5K?-s7# z&!6CWeir{7zTOwU$bbI>*Y!7WU8jeiKf(6`-+#9G2tT?{@U43^o38r(P;}M*oXrlO zV%Le`3%_pyU;A}Z_+??&$>EuwFW|aw3E$kxu2aFYo1540lY0YS&+U8*U);>RgJ*ue zhtFfqW}57)j1xb`)|wXX@TeMQ#{=dXRmaP2FHYhML??{g^O+E)eN`uQ5JxixU@tA(%q zIvrg5>fxE6AK<$02-m(Qxb_uYE1akH6~nc!1g?FhaP2FHYhMLi`>NsER|8-899p>c z)xl?ezK3gW16=zW;pfk?_tym1zM^Y~^E~?b7_R#!aP2FFYhO8B`zqktR|(g?D!BGl z!?mvlu6^}z?Q4MVeGVgB`+aP2FHXMVnb>%Jvi`>NpD zR}I&`8o2h=!nLmsu6^}z?Q4K*U(t2Ld1_xVeBpCQ;M!LTpZWP5uDKO(?W=^JKhwW{ z!L_d%e)RJVT=#9^+E)kHzIwR!HNds65w3kraP2F)UO0d4D~4-dIb8cH;Cr7#3D>?V z_}0(YaLuiOYhNvV?bqqx+E)+H{QLmdeMh+VHNmy7C<^DPeZ_F?D}ifYDO~%?;o4UL z*S>1F_SL`_K8F^reRc4epYP$C+W^(N{OcE7`--j~&hzN!W4P{{z_qUwu6^Zj z?W=%mUnN}os^HpJ4cERJxc1e{d{zTaGsi54A;IA_}Z_N!nLm) zp85F#uKSj7?W=-oUo~9&YT(*e3)j9nxc1eG0L9?fMCP@8?svu9L$r{&Q{t z*WOFG_FloY-x{v{Ht^XeT7L^)`OnuKeC9vD_wciyAK*v#5x#ey-8h`Hem|4IU!dm) zuHOq)a6NB(xSqG!O~U)?_eBZ(%6;MLtl+x-2%p`}<~+gIcQ%i18urZYZXUyRK7sG< zY3Eb;$*-To7nz+e;JeQ;FX5LvnO9fiHC&y|)p!fvex6;wgYUhkho9UB_|ZL@W1c?$ z7{2w_mB2Os6t43*T=OsBntut`{42QTU&H6#*<6jc@Z}fzJmFh^PI|b`5AeA^CnH?@ zn&8@3bhB{&+E)xe_`Y(u-Y*6G_=Wy_!e^;JpKzV8;j_Eg`3A0iwQ%jLgKzbG!gb%_ zYJ7xWnJ4_pJa2wA^Mvd9&*4|*314ZR@GJ9#>wFErGEexGdBU&E6Mki$@beegzJ{yu z5w6bZYCO6{I8V(#hHL%_T+ed~*ZfPk=3l`#{<>(}s^ zpKsuM@9g0V_W`c{5w89TuKwuO;e6B|!_}X_)t|yM?=0bS_X@858m|5ZuKpIT{tmAG z9}ZaX6ou_s4McCvf$raP{YK^%rpUmvHr0bnoB8^}gug*L;-y^CK7dP2J<$ zgmd_O_YD5Y?k#+GKYRW=_~L8Ld-&kpTL*@ zkFI+UpQoPtIKGHRl9kG!w2~SXt(=xyEIJH@rPDg3RSr{8UzHlA+jgE3R>x84-b5`W ztAkdG)p4Y>owxIr4olR=Ez=!!U)S&Rde3#u^XmIgE|2SVz2-CT&+Os5r=Ql}fH!sA z5Z>9^^|#@}t(-^js`4>>tb78W$Om+t?+9MZ8D7j8Ud$O@%o%QT^Io&eoZ*%a;KiKb z#hl^AoZ-cs;l-Tc#hl^AoZ$<*p5euu;l-Tc#hl^AoZ&V%Z|!B~47Yp$FXjv{<_s_9 z3@_#kFXjv{<_xzv_u=-q@2yil-=*z!`-P9c;yi#?k9FRF_mvOfi|tWAt2Jc4)C z6T_>AxqJfeD4)XZxC40Y5Z5z;H`SBDYX`dg6uxwT^EtdH&*9yVxxDw<^8BkiIQQW# zJBN>z&*64lZ{6}d zZGU}uUp)cb_O}7I{SD!^ziqhfZv?mfjp4Sx3EcKKh1>pSaNFN0ysPt_!)xM8)24Y&P`;I_Xp-1awt z+y16-+us4)_ICug{hh;Ye{*_ z9XEm7{-$*G4Cs1(HGu+usb{SAGh&zd5|C9&dy4Jni=D!)<>9xb1HP zZu=X;ZGYQv+usOo`y0b;e+O{e-x0i{^UdJ4zf*Wq`8nL?oWpH@y$vx>_4shx-vC}w zz5%!6hH%^8HhdPj>obDe{>E_I-vn;^o5F2>2XNcp5#07SgWLX2;kLhCP@bpluMZ#V zd;_@cZv);}K7`wx+i=_82;Nms47dGF;2q^txE*%@xBVT#o9fBnw!c%j?e83J`9xb1HPZu=X-ZGU5UP3N1yZGTgEMfm~T<~)Mi{$}vmm)z}W3b*~8 z!^g_!a67KIQF)%WzdpRLo&aw9+ko5thH%^8Hr)0%g4_PaaNFMmZu^_UZGSVk?e7%c z)%nihw!b;NqrA6qd7d_BA8z{_z?+>RT;ZGU5UMLh}J_BVyw{tn=_ zzazNqZw9yhox*K@=WyHK9B%s?Y*L=T?Qa7<`=YxYg>c*7Hhipn1h+ZIaNFMm-d9fw zxBVT!yULH?cH9hZ`#Xhq)H8?M{^oGoU++!jdD{N^aNFMiZu{GS+x~`d+ut_a_BVms z{-*Gz&UXN}{T;z;%4cw!^Av9TJBL@)lf!L)y*HQVIs1aU4t=;CH-OvzHsE9RgmBy6 zHr)0%g4_PaaNFMmZu^_UZGQ)F+usq~_IC!i&)s-iA+h)$>fa?RyMgD4)P> zzf$^cu73cxIgj8A#k=CpM1;t9Nzl8^Bmq(k5@0xr+TQ%`*7R$0B-x|;Qg+r_O|jm?eh{L+&(X{1h>yiq;UIOg!lGx zfAcuqUf``0oComkcbzxj9pyuKOZ{#5lJXIJ(sTVW-0}&0p`H|8(>e$6)`_lv1n+;_ zc?Pe`r|_114sXi6ca-N`w>iU$Im3%N!;3k?i#fw>ZV|kgGu-kCyqGh*m@~YXGrX8H zyqGh*m@~YX^E=CPuGyU7#hl^AoZ-cs;l-TcHn#{~%o%R^1YXP;Ud$O@%o$$H8D7j8 zUd$O@%(+pXbJgYyFXjv{<_s_93@_#kx4A{|V$N{OC-7p<@M6yJV$Sel&hTQ+@M6yJ zV$Sa>&$(iAh8J^&7juRebA}glhTGgCcrj3rp1c3|;l-Tc#hl^AoZ-cs;WoDjUd$P8`2=3f8D7j8Ud$O@%o$$H8D7j8Ud;Kw z%5$FCoZ-cs;l-Tc#hl^AoZ&XN2wuz?ZutaW%o$$H8D7j8Ud$O@%o$$H8D7l!J>@x1 zY|ijv&hTQ+@M6yJV$N`zTLdrW47Yp&FXjv{<_s_93@_#kFXjv{<_s_9ym@)fW1BO) zm@~YXGrX8HyqGiG<`%(=Im0cVz>7JC+g4OGv%l7&Ka&}4!52hZav-y%5$jy(DnH6-YL!lc=c504fyc; z&O>)b>#uB4?zh)V1Nf8gcdrw+;H%#0Jcd7cqFd)2-v5c4e-58r=-k_~yq?~Vo%`_m zSvc@^H&{}17vOWYjV@TNS1*G4WM!)ubPe5icBef>Mv0X)+02DjjKc^}?+(5-&}um9Qk5I+2q^D*3d zGWhTjm!HtpGlN^t96ov2^(^4Fp2~;I^RXTe-hasTRN>ZBgIiAk?`dvzxb-yQ))T^K znnMe2JstScU)}m6c=J)`OYo|^3-2ymK7rT&;k*yG`~cqiyUUN^me1g|CtQ98xBQ&` zxXV|zDete{UIMt?URrRwJ;v}2_jJ!I`tXKsFO_Y}$F-gSZnx_e+}0VxZJm9%-L5mZ z-L4mKTfhI2@;Yt(OYqLKZk{o``MmF~c%<*aEuX@x6_+2vEkABXV+V_u^ z&#Ra}yqG`yow~p6!>uQSTh9XiUiCCTR$k9yH8+P4K76tBHhlat=N)*r>U;@aUEO&M zpS|394_;l%c?w^=()kc>`4N1qo(bIYQ+QMP1>EvEeNDHXTB|(ITlIXh3%9=)25|fP zVgmp5Z}fQscqXrXynNj3c$e?Oi+?v9-e2FXKZjS};N1H}xqqg-5AVtYc=dIzrvb08 z>pX-vYtGwrum_P`>QE@4c-4#b|{}$An(BK z{w;yqdWP_A^>{ah+v8DX$MSLQ@hE`Xd9~nn{~5#AwfkrIy}BQ$>{LFkJ)Q<|dpvEy zZ4Q07^<;4CS-|(zeEfEKJ%{h*p655=J$WB)>&f7@FAKP>)Bj}oxVFwF-1aMi+kW-n z`)Hjr_&<8?`t&|kKCZ354!8BU;rIN&&8G{$=u+1|fcu*B1fGVjCx@@6eC^Zab#5Z> z!fibRxZN*J;E!niIs8$1=QHKwZu3=lUQ6)r9i}skf^W8zmt-?1{z6SrK@^$#WpK$AG!at{c3;q>( z2fm+t3I6O6nm_!%mG8k1llS4D(RmHw^V{4$j^XKs?z|@OqtrixFSO1Dd<~obXUp^d zmU^o218x5B3)NqTAFrM!d{6bX;3q5Jf&W;(1iwJug{6cp&GIVzZg~y*WcXFc?~}Pn)4=o zf9D-|_W@>BTi zO6Sclme03)x$_Xd^gHJXyn4$??z)1v!D}}=&*0q~oloIadE-mv z^BUgZ@=dzd6T*AfxO@U%C3g3RJ$QP@bt~pDfbXIF5I#{pgCB8^dmNv@t9QEoIsBa` zyZi#)P~MBmbJ#%nDm+m>fN!dN9X?e)gx~uE_k5!TuioX>pTO_ed{THWAHY}px|{z9 zzP>zzzehfWZzrF_zben+$I89k%k%8(yaIT#vfGyi{EXe*`XjjIWBAekr|*B{!c4{?k2NK89Z=PvFe9_&TkhZtI6HDWAZvv-QJk%4hJYtsg%5 zgXX_fUjM)JJysvy)qaKW7ux#a3*{5|D{TGno^D?m{B^c|cvE@rtL61?s=N z_}{p?#_f7h-bcwPAr{ypW}@Y(NOe**uZtsmY~K7(Ii>xVa$_x3BV|4Lgw ze35DX@LO&D@V@d1{9#)^yrp~wKl(WL_&$XX&vWa^;r4mU%Gb*0b;((-$AeqGv5Yt2 z))T_5Ct1c*ct`8c;OFUi-4tFqU)S6I<@35&c^|&GK>G!^d>d~01U}H~^Vu@)9Z+7+ z1rdfJzjf=+;Fh1lEuX`m{F-|__2Tk6f2`+wK79BaciaX%QBO$M`Nr^{ zD4)RX`Q!+Gy7C#^@>BX{Zk;*2INz@?bH4E6eBs6U((QcV#reW*{VBYv^^f4i9N^Y7 zrQ00f*Vw)vSf0ZPT7Lt6kL@q~J9fVCmG^hoRRX`}8FxQ3g1Y2f4%=MAJ_KRhui)(;I_XZ-1fH(xBZRbw!aD7_BVyw{$_C7-znVoH;3E)dfzP1 z!S>6C+kS;`%eUc{PvEw{*)r~R%j>cI4VUp4{%*YvlfdnFY*P5rMeh2~;Fh1lEuX_r z*Kxf=%Imk^!|~xZpAc^OHr(d~0F8oP5U-&aPRqOHKmT$n1-O^nTP5A%XeBhRk;d^~Y^MOBY^MPA_1h>zRWy|;! zUcJ!WzvXcId}-yV@_b%!kXxq*w|oO`b8EtF&LQ0LF}&)#@989Pd%brAf8(C+b>M6n z&*AnuwwIL8*SrD0Qm+q$%Xk7m>l}BTr0|LU?jFJIIvK<5y2{}8K7x02dHr^~@Zom5 z2;r7*!!4h{?fF{@ubrmHO}OQ!aLaqgl-F;0A8z>&ZuvIc@*TM4m*6Mdt=l8~BYJuG-bxqCvm^(?`yr~939&u~Mx z&ICSL-+2!{ey#H%yt1zI@pp6g{@9BDRw~w?!L5G+xBdm(`YYdi?tTw${Z+X2*WuRR zJg(ed@!ffabj_iA{B!3My7J=_o;#n>m9L!i+<6baSjVlidGd4TL-*-1!VXQND8ObLTzySo!8@&z%q9{nxnlbbnCJ_ul3_fj8goe0+L2-`~M`2Jda} zdbyE|J8lHG z<1WGNxZNK<_i+>WY)3brKHQEwfZK70a69h!%;!FC24863XK*|29B#*5!0ot|A3yhT zJ$SLda64`Qx8v5~cHHJ!<>L-_bMp-0wJ$pFp8eeU1U~+}%Z~@oozLJC9k=q+=gxcZ z{;saSc}_Xs{gU$#UfIKW_uS{sC-BzpEpU6k>yMODR7iIAKl%K+%l+WSSZ@8WuzJc63zkI&$ zk^AtLJb-^m-hdx058hpf^ z>y%I7_sIwFrg}#3rzghg!uL?V4L?L4!GH37_xEfJKVJC+ezrV?UnU>GZt_k3_nMnz<(o8;Wx_%@Q39i z_!IIB{`{`4fA|{mIecSz4)5r8=l!xg|DN23Z>gRD{%LswzPCJte_P&$pCXUoW34BK z|3di$eyu!(|4BZ8KPexQ`Pc4q_p3g97v%%^ z!SV+D1bGNQSKfy2{*-$iFM_{ptotAMuhf&kua&3pyW|7-U*scrw>Wya7K=9>SM&+&27N{POY4DG< z)jx*Y{4==Ce*(Ap&*5k5^Fj;w`Z{jDj913x>*}Pn-1Sq1|I9v*1b?IYYs+{5fA6kt z9~cib8L;BW-?tb52 zC-5K2Q}~7Q0sKn&2!4}1gKwty{ipDUm7l|(ug_KH@U`ULmF4+tDEHy-mIv^byaE5B zJcJ)4Z^KWNNAMrXV|Xyr>o)MQ@+o{OAHW}%kKijE>26;cd|mkz{wDbx9?En0r{&&N z<@xU`_u+@h1NislA>3ZSZNnGJNATw#<<2*TzfzvS-zZPv@0SnYq4sM8|D^I6d=L2) zeu#VyKUto`e=hf~F3iK=XaK8@CT1^>zTqIm(Sr(%e`yM{V)5L>+#`%Jb=Gd-hh8l z9>TYix8YxrNARQMG5id90{@LXh2JP2!2cp2!Jn3A@RuB`Il~9Ke&+C3DWAh9%6oq( z&wmr;eRyXlciaHpRK5Y6z}H{e^!L-C=_bP8F z&-1_JK0K5M@Ezq1_~+yye1CZxexy8tpDK^x-+7CB9V3CCr+f;(Mm~Vg{ay=f9TnKD;gu;9JWZaC@E;!tMD< z8*a}}BKVH#kKr?|KY{P5dSeL-`1P+SHwI2EShUDct_Po5Sx^ z-n+Rxx5wl@+&jS?H-N7rZ@}Lw58+$L+wh&`5quwc3_n_)z`Z}Y$FCHAy7B|~FXbb6 zpq>nVgYr}OTt0_CE6?F;o~Y~gmh$}HEcfB|xuyXAe&rkRQ=f6qD?<2ZlyAd#xzw#E zg72$*3_ntyz*{zd_{qu-;OEOn@Fkl+yeFT+|Dc{Z{4RM8U&y^%%kzKXNp224{5A3b z{tkHq{y}*N-(KE^?=6qui9Cj%E>GYin?L*l&A%(@MGlxe6H)N0k_X3hVZ_6+VEe>BluPF7;c|e zOyIXFpTg~Pj05QDdZ^KVjK7tSBG2A})n82@8K84>QAHW}%kKivj zRrf#eSIMXF*WcsvbNIIM93I>J?<~*%t?KvTEqMUnUEY8nArIkwn?L+i+ss?&KvOV+0L8rYTtPXAD`vC1@E2aybYhJrvvYt;_?x^`9tSR@am79$MF7{&b#m{ zHg@|mhM%{I^UB@jdHzNDCj55gyYQFG$MDLV)PGO8zoC2+{zm1y@LjAQ{t4^9x7>fE z^}`QRegdz2Qon}}pX)lT-B<4K-0kKZ!0Yllyeju*<(`dx;O67Q?e+fv{(j{f@XAB( zaUq0%M)@{;Pk992PaeaMlqc|$HoNWX}|+{2)ExAZ^P|%*9abKzhd}S>QCUi$W!=U@&Vjl zza7C3Q9gs)>!?%siOSF6zmVtf>*d~`%Ja9^SAF1Yh|KxBeLJ%MH4b=m)Ad2z7Ajg zM=sxjk9T(cOYn7-@4;)z58<0BKY{l@qxHi-tbFy6^7?Da*WsVG^~0N=)%xMj{#f&e zcaz6I|qzXac0`2^nDajtt)UqVoy7 zw!QP}pUdm)cC=3T@6OhB0-q_r1iwl79=!88*FS{cqx=Nk+|}h5@V_Zv{Y!cMW994c zD+XFWy!m<8zXad&C%R7HedUMnuPZ-+SH7V2!;e+I`e=Fm9p&rrRQVRX`9-ZCe#%d^ zet7LmE!!E1YH{qR31--CCRAHpB8^}}a-y8Z<`)8j(*vGV$BUv~LA zeA{z%{lhC?(fZ-vRK5qF?d9@A__^n4{_y7BT0i`k%KML(*I!w3`2gOLx8QyCFTsDO z{vN!qo*{f<>xWmqs`bO~QogzS@D!m$}E61b*eu+#GuFm;J(dAHJ4+0AJ~CETYhx8b|WBly1Z7=D;MfuABz;pfQ*@ZZTt@H^!h{4x0yUb)E4 zZ4O^cp2O{Pp58yp^MAeaKKvc>0KTQX0pCF$!uOW9;r}g<;Hf-@UnEc9SIJZOP4WT! z0r?32ggk@4@Rx4RQ}`?7bNE~2Is5~1@5%D~caZz=J>>!XV0i;R)$JvOAE$g9-j_%4 z3*<5U_woc@nY-I*3co}70lcRC2p-Ba_+QmCg>QRl=6)9f{;cvjeC>WfZOkigz#@D--aI}kKjL$$MEyyLwMy7_josfKdSr!-c`Pum*?=L@^yIS zFxTILzi8y zzRA#CZwdS=KWzta6bd}DbIf3Mv8zw+EZA@|{X$OHHx z@&^1n@(_N8ybZrV9>K4Y$MC5d-4$e19=;Mjy!^2A&=pA$rE^L_y6#*d;quKYZ}4r zdvFO2g?^m?p+q~bcCxTzC ze^)eyZ?Aj;zd`vFzO(WJ_#W~R{ClhEI)|SwpTf_P&*3Av_n-1S|FEUIpY-9M`h)WT zzVQ{#8}O~w6T&}qt;@IJC#pY!Z?}d!-vn--hfm@5dDj8lJ|{nd+xH(c_>Q_xrtl~2 zc?*0uyKdoY+uu#kl;?b)@;>}Hc>s_9?4GAJ;P!dc5dI_ewBf0GBKTMy!|ikU3EVz! zpTh0)_5=8ZTF(e>pTp1K*DF7T-zT5L|MClWUFGm6mG_=4&wsT`-F4-|*O3SCx62#w zx#kwaw^P0ix9eW-Tod>Q zluzOI`>q4HeO`J5|EPL0_^0GkxP2~q4&O`p9Db18dtQ0|#pj&iP0cNUe@{IPxP6{E zgxl}Lw&7=}CxVaUG5l(I0>4R~!XJ_k;LpfMaQhr}2Dk6sOyTyswR5*cx__vz>3(E6SEdH%O5@54KG{{v6t4Y+-O zEri?m*V=IVoMZ&I@3qBn`y74(x9^>$@W1K225|d6+6Zo+&&=RMo$nO>!m+y^=J40Z zbNFU*?}g?0zhCad?e~HL_(zm)!0r2MA^h{ox8VoNBls!u7;fK#OW;H0Q~2-X1NdF? z5!`;KH-kT}{1k4#(>sT+^gFjNIsDaf??vVL+vi|?_yL>*O)qeh(ypZ>oF>pXz>K055(&1pZ$2WboqOn}B~@`8oV6 z@*IAs-1ExwKS}Pxe=ZN;*U203JLMt#F?kzqzrz#3Uvh=pml(djJb}MWp2D}558yk> zNAO+c8TO^4nJLv@&P=SkKiZDGx&M(DSVx4-1_J6OO?;z_P(mON_qa*EAPV}k_Yf- z>e=7IkPsjuKYMN&Q{wjG0-_+(0x8G5T z;4eDL%{hkK_fHe}*6L5;2R-F_2Jmwayk^BKStIykC%b$G-&y@r_&)ME{0Mmt4|PA_ zy|g@k`<|>1KS4bKJW@{sex~vv{9<_kaAHn~n z<7V(xGq*2O_`32r{1V-MtE-jg+&#j*e_eMR-Vrq_qy});eF);_+|12+pX?8qx=BAnS2EQkUWF$CZEC&lF#9(Jcpkv_g-F}=kMh{ z{C;@=&*csH%dT^C2;u9?+wg`wf^RF2;k(Hb_<`~i{yq5sex`f`zet|Jua{5Z56kE9 zXXQD3t?S)=$ z{NM5s{KbEC=as=%mrvpA%IEM+0v=jpwoJpcD7@58s32kU}!+ zLF$R%$H-&2{r*e>|AF!;{1W*9{?|L)oJa8MmCxw*J_Y&C{-{AUl_$%e!8s+)d zH6I_|k_YgO)zg4C8V6 zr|_|S4o`L59DbMb-kRn4KPC6!E9-tHfFFFcJ`V$5SNRaWvAhl6N*=*?mdEh@$0y_lPA5lJrKP6A#UG=2! zfqVd8^JaHmBltx53~ry3nZoUN1?TYM_xoiii@V+neV(cTUqc?k19=<1xjcfm>>Gp2Clj58$WCNAUCIbNB~u zcejfizJuI*ZFxSQm;3OU_A!9R$~WLA%0u`--iH519>MRF$M8SP6S()dyB($QCzKz+ zA2?O-%;c8}KvbA^akF z8~y2srpQ-!^-qZRs_+`pZ;djdC z@JHl1{8_p8y7K(j(C3hS_*>)wd>eTKZm$o7aC<+o4Ik-zBls88pTO;Vpeg)dq%LDjk@&^2a@({k0yba$?9>EWi z$MBQn3EaNlH-H~}w>z&9yeH4#r^}~s`+n9OexC9<{7Sj^`tqFbl>6|<p z;Oodk_@?qUd@Febx7Q6~cw6}dzPCJuA0Z#WQ~3yfwmgGxW`9@0FI0XG|D8OC+waMH z>zC(0QQn6?AP?a7x%USADdj`>DtceL4Y$9uBKYf-kKy)rQv%;q`2oDR&fye;dY(m@Ez3ey`enk-Q_-f`-|M;asapYWg2kX#}ID&(uUjoBe?zUeGIqXy-(oY zLeCH3->`YYzb&7_eUTaC^M-HY(5G9`AhkpVS|~pOQD= z2mHxhCn4NEx7LQ+<7Nc6=QA;Ul?PmZ0$)#_!tHs^0Nzl31pl}^gST}ZPT{*LKZnQ4 z=Wu%-i1Na{D2K*>_2tQulhMysi;1|nd_)YQz{#SVl zf1bYYFo3TuAHm-t&){3gr*Qim?;QSV<#V__4tj4Y&wp>_eRx+Mz>k+V;AhE0_^v0o z>#7aENcjjpmdEg0V9Jmf1C1^H<#x;`^Zn+ z?^D35A9G%X4?pVMhcDDqgLjk<;1zj&8E?R=$~Tws5I#}91z(c4;Wc@Ot~o^Tnet2U zx;%#W)^-oCf5xK3v9?14L{T=u(m5<;Ht)~mWO!)-fRDU0SmGT34NBJ>) zqI?GLDL;cxm7l|h%2(>;`9Gk%2cIfmgFm5s0QWVYCj4pTLwHm94t%AD-Thkxk8J+% zHI+}`tM2diw+~-i`2jpoehh!T@)>-n^~~U#C_jg9q5j%i%JZKb=I%cO_#$!Mg!knk zeCcSHUxJU1cOJu=$2sr9J4ZNA;Jw40_ux(CQ+QS0hu7s}`0Om#pTQSrJ72(SS2@q& z)yth%HY?A$^>gQS`0PgK4ftf@yalf--=-_S1fSgQdSZC*7Uw;9{Z8j8d^mMJgfEmI z!N>9myndhSnZk#6JD@k-g!CQ|zpTO(?az2H(o^U>ccjXIs^&c*u z!#fM-)wh-RrS<%uu6P{x=@sX7cvs$l*W@ku?0K%I4PQL#*0Tg}zR=}k`0xeJd+_l} z&QthI`5}D#d6yr-hw?Fe@@bdP;0yU2zR7X!_5U3Hlg01d^Xj*k=YN6RhhHWS;Md3- z@EheJ{4RMLzNCNmIf6f^d<=g~p1_}#r*P{Zz+e1?o5Kih`3!#QN$ztgQ@G{l%Xq$w zd+#XEe>JVghp#OU;7>g1*580{_^9&`e#ZCR`rGgg)E~h&mB;WM*K^lZ7d~74oE6tq z0w*0I$pI z@bPC{z5(yZoAir+xuQRWFUcc#Z{Ef@GW2Lu9NC} z%Imkh54U^^{(W0N{Hz~o{cy{taLZ5Nme1jbs=vBTYdt!d=CGdt-ra<`r(#u!L2`rpQhj2@4+pf!Yx06e?jZa;dXymeSdlV zmiOV7Z^8Fde+)ld-h*2{gZ2fS{ zSHtr9E$_oE--2&$>xWyu2e*6*xBLXYjjbPU`Ra$t>$kiQw|oo!!k4)Hjp3H>!7ZP{ zEkA*O=ONvG;g+v{xV(PL`*6#*;BA{f-10rRbB+eTi%CTz6C$!P^}+s`5xTzDctfC_|A6y z!!2L^NO}F1_u-ar!M|wN58U!SxaCv0nYMno<$G|;r*O+p;9I^@k6&=hS6k)vTi%CTz6C$_ z2zUPw!!6%~TRw$beggl6%^z<0>c`9Lx4aLxd<*^?n?Kz0J-FplxaBAC8*Tk?%U3^9 zUccpixaC{$+wFHF;g;{gEuX?IKY>4F>xWyux?OqwmiOV7Z^8d&^M_l$2e*6*xBLYD zPg_6S^40Cj>$kiQw|oo!s-1NG!!6%~TRw$begc1s`mT` zZuuVE@+sW%6Zi(Uez@hUJC@gPc^_{17QA8Whg-e}w|ok>`~?0%TR+_L)t$=gx4aLx zdqW;}p_)RbOJrGuyfR2)_RI&d2cdJ-}{L3Dm?h1J6|8ZzV^Ka z|FB($@K5Ntb@*qWcj=19*#>+I@DMc_?pU(;Ook>Wqb-> zU-h^GxBdl@H*X$Vc#9 zA@UHR+{1|x;ZvB0@_q6LDz`w1Y5&X~hyZjh_*|(fe z;Ol7qQ~37!{NWs)>bMK|3-)w#u6(w<@24v7!B3a_@R#a3uffkyz7Btfu9F6QpnM1) zZ{+6Cf}gK^2R^>momT`u<_706{MQG${x00E{~r9eI&KQTTt0w5xx0IPUDD3cp@GGq`;oW&xilpTmDF_jV~?4>!xJ@Z01y_#N^9ez&{O9k{P`Ms)k!KKwPxcj2#<_u#LSr||XV1Nd(L?e=j9|K~rQ zkKr4tCxdS!pTf`3?PUgkyYdV8^?z~ulEdGmytixldU%Vx3V)lt2LGXcr!;`udK&Pj zu5@!~!r!U>7Cd{EyMEg6+xBoh5&YfiS%Pma@4}n%1pYyJ3cvIp?)vP*w^n`#KkN-I zKZ5u5{4Il@zn$B!3B0ZT8T@k1XAb|g@;Uss%2#$P@4Mxz@DH5h*6+i=p#A`U&$BLH zhbMb$-{F_)emI1GRsC)FSI%_(9k}I};FrI^%`=9drsod{{OdYy5B?4L7=E^VvW(B* z&wSXee*xd*3b)S6=ga$Yp8Bis!{jyix8?O^ya`W~Z!O~;_zB7{E#qDIDa!Yj@jm>= z$`61LV5l-$?MB_6aIVUTg!L{K2v^a8SlbxSH8E5SH4(Y z|KF9bF5@-$pOvpK;~n_ByY6aQ;2HTXKp*O&1od_(11%XkOAnet1^co+Uo<$KF`AO1e&hs*dF9x6Xs z#%J(5bw9jV#w$^I{vTI=bs4Y0+sfCM@g}_YE%*9u2=5;E+ZC_7w&2b0x_ldMJsrAw zBDk$%32r?xe5jr-+}4r6t)~Z{s3(P6PakeQ1NcHcL%8*f;MOySSC4md&fqqm3EX<7 z@Va_ta9hV5ZaoWlOFcQ<_Pw%u`TDdT58hQz72ZBp?-#&liSrtKteyaF`8vFEf}3Xp zZuus>qkIUre2cDpo38bD;LQ`=dLp>>FTp#?$8gJc;p1<+{seCM9(=6+6mI!Gympf7 zAHXd?gm;x6!7V?AS59{Q8Qk&{cvtx;-10McC3XFCxaAk{uJSqD@|8Wx*Z)j;kFM*X z3h#X1t;dI3e+^za#pMII%ft@Rs`9aO;oY z_IS1g@1En<8N>VXF5LQi@QQB#DZF->>+i!mzjZ!<&yIcNirdSOuKR}({c?AE9K-E# zHiOr%aQO*5y2@Q2Gr0B6;j=aT73*BUD{DH>;WoF*m&@0ot*j))T>->RE!@dSbZUpLgLM^(1i1 z_uyUSQ@G{(@V@c`xSiJ!K2d%IxAl$T)|0_2SLr&3+q$Q4>zTo8>Y2l>X92gK9Ntk+ zsllx$fX~!Zhg(krZaqzS?P@pY5Ny+S@2V$)Th9{QdSdukJzco(}!En0Nzy3 z5NzTluJF&e{l0@!>y+Sx1I>zRL>IJdSbZsbm3j~Byj8L z!L28SkJZzMTh9P)JwtfqIyavY+&f6v^-SQ_Glg5v4Bl1G9Bw@exb@`lv3e>? zy+Qx1J`vtDX>UJuSHPwBcj*bl}z#!L4Tr zUiqV&PYk!7F5G$&cvC$+xb>uP>*>S0>KVYTX9%~R5qzwkG2D7Gxb;lnl^fiArf}<- z!L4TwZ>nbjx1JnsJ(aJPum8S!Jh=5#;nw5BD-(BKHMsQzaOy;XPx<<{9uGcKPZe%GKHPe0@a8RU zJ^|c%>Tv66z`N>c!mTHSTTcrlwnw>KVbUXAHNV3|_g-&1V9)o+;dVX7HwZ=5Xs-z^x~Tchytbw|xCu zj|aD&DtxRSA8tK0xb+0^%I$7Gb-493;MUWGH`Nout)~UIo;JLzo(|l4BDnP|!N=-} z;nvfITTcS7Ox=8XaO+9o*3*YK)iZ!w&k$}sBY0OmW4QHXaO;`C$Lg8Ft!DwlkuifS5GlW~u2yQ)Ncvn3c z+zTr5>Y2fzTu6>RG_8Cx=^4W&iT^ z-+aKW$Aeps5BIgs8oaLeqXT$LUWa#IX)j`~yh_ziCTefVNS=L2|8^BKX%>L0^C)&$RwJ+|&C_4R~GsP59!aZk-`r{cU(n z{T+B^W!DqIXX;;qcRuOzU3f$N3B0S@Q4ct_nD?c(w?cxyN3b9lGod;yFZQ25&y%Jb-uhabAbFzTvz9@80IT37_rf))UeB48fb)LZcx}WdCJ8hRw;X~#7@UeUVuWjY-&xi2-+ntZ#&3C#vkKv0q zxqJp6H(Y)KuWjP;Q+W4X&S&t&#?I&PiOy?5U&HN74(~tg+&k#075B%UyVtJ-+OI0S z<+(Zg@QEJ3YVi7hT)qx(s=op6sJ{sxt3QOdR&w)f!`tfbz?ak?!H4Q!g7;T<{ayHy z`V)9p{XKYJ{VBZppqtMCo~VBa@2P)8SO1v)JY8>a`~1q3uKpRktNuB>vWn}^;r6+f z$~T@WT~|Z(d+;UoSK+mlU4IR3pLYr1WA)eJ9rZWh#G zpF#kE4hTe9tbh;}p(0=um%1@gL5he(QR$#yM8JxG5fv*U1Z1gz8kMDpXmEiFqEY-g z#WgxE9dL=_FjWVg@xIRMa~ypi$IU;_(HGZq_LEP}q`iCh`+Y2b0r&60^w(zO)8~CXF9XwTj3m;<3@8GHO5Af(F*5@9c zDE|nrKGF8)1YanBgvZK1)0H1R{a-HnmVKP9KZX~|pTZO6C-6x5GkE&3mOqEDm7l>= zE5Ct9H?{m0?%(g( z!wcni@J{&$_;?e`Kf?WcKLfl}{t4bHe}oSc%a5LM*{kq!eggOJd(GgZ@>6)B{5iatS^fg<-vi6x(M@fB3D1>Z zz}vgodAy_q#UR(=J~l)r&DspW6sQ{^}CMEN^-s{9sSe2(RJ@KpH+y7GH?qWmMg zytCz>;0xuC@J#tZ5FbX7Gu=SCqm# z<7#uatj=$I6eMb=kK}`4f1o{3(2^`~;pTe+EyLpTft{ z`k%r5`+N&{s{9K`;`u$f5FO)xrCy%oH1>C;}nZsk{FX6fJ3wZr`mS4gneQ#_H zPn2K5Gv#mKac=or_*D4~JXQXVuKX6>f05;P@KpH+c&7Xwo+$qaufN3dPjLU9-v}?1 ze}<3xd_MZ#%f6*QZ}}5=r0<)>@LKr^+`lh4gXfx0;feC+^q;l$Wbjn;3wWjcCEUMv zSinooui&xrOZf0cThAKa>F>9x;O(t!egoe=*Z!WF8eV-wcXj{S!u@-a4ZPF*4({Lc zY2lu;hesc2&)dO0=K%Md9$qNt2=|-;?l~uTqnr`$IcK=%M4QV#56YRqJtu~H&J><~ zl{(a*$y!fzvU$=t$ z{081>zJ~k!7M}dOR?_pF}b(ZuE_-*?%6zwg9wf8HrPQ%(Z+`58RAi9K%$_xU-z(R>E?`31cAXv@#x zKEH%VH?{c!?(-{nr}+}@^J{oD+%{O%YJ#RaB^RbrS!aaWv zFFwxZJGjpu;7M%rJ>2Jybj=TRt^WjXKHl<2xaXhY#V6Q&^xVsy`}_o++}!44xX(}N znosCj{|w&T!tztNzrN4m&1cwr2KW2(0zPU!hx^|jwS;?40gpb@o_7WJb(C<=S;G_M zRB+GPz&)piXUf^a{kR&q=j`Bxa$30O?BSl%!7Jq);GWaNJ?98-lrz9R=LGkh5#A~1 z4ELPq`!D z@8LdwghzL^`2p_pCwQaz5$@;f49~Pb(eo~Q?tlJk0{5I4-Y91Z_nZXoIWu^4FFW3p z{*AW24DR>C1-yUqr$qnxzZ_ny&6jZh^9U>WQhh7oacMbgc>3+;72NY{c&Yp?yjOk$ zuav)od;T8ouM-_SSN;LsJ=NCV!##h1`|HOE9)GjtjPPFTKf^tL^8Cvl`s+#zkH5w8 zr|?qiPvD-P!u|DT4zIO889Y(`0`B=sxW5h+@Lv0~g4f?>$5p~Tzk)})o;UDZ`87ON z|F>|@-@&KKZ{fN2e-F>J{toW>Jv>$Z5ngNk1AI{a3GVr4_(J*73od)ry~Fl-0x#cX z9>c4zGf&`4<x<7_^KSMoXB|2*je?)UK=oqB;Hl;h@KN5w3;7WqX*~nH()U%vY(zafmh0j;hr;vdrkuHlrw{S&K&N?nZeuW)>p5y3wSTj;eMSKa6kSP zy!i;rFX5eh4fp&F-1nu1N6Oj46L|yo{2ko)xrHZRYUh0q_kHf*o^ya#%IV?$`QZ`n z?+*w1BW*n=x_<5+;n}0qTe#;(^<{4x%}?MyAH$2|9=M#9L`5D~jQ+TKOIo#(n zc#_+C7I2@>;g#l>aGx*enqR^FyqECk%WOSsxc8@m`}1z#m2zsh&u`(`(w=t*_ujYg zLf88qUdcPSUq?OMe;;#%cb{$dfdQU8$ovG4^}Hjz{urA-!@G|)kACp7&!g@G6L@y( zB>K<)#q^2!6h6NBHdpfrJbRn@jIQ~V{#Kix!}E``=gr{lFWUS9p7{M4-agy*VF@qq zZ8QsviSzyMz%jYc=zXaTrIr0 zt@$1v-^ILx*S9f0!27LvPye9#5k4w^fJZuCCwSi2@s9NSSbp@v%RYzK8+dwO%ZcIf z{mrNFUgs-;r+2gY8N9x$c?$18$$SnkAEZ9Rvxk~5;KL)#b9j@RFX8>e%?o(-h2|^z zqs>crQJAmcgZ8I_HybX!9p{{c@Wh;msXv{tT~Prk?-MWzWlBF`vL=&ByRg`!j|2 zFS6%N;N9HvQ@Fpb&*8aVM>2RRU%>tKF^Bv6geAOqP8a>>{|dN&?z)0|P6=*H}LZd{?Ap;4(>TEyim>_?*Cn$4({tez@w+x zdV07&?-AbUc?Y=f*9q=9BRtvI^Pb`UIqKwxFMFPAE`~SXlw9rG6yD1dxPShT!dJ?l z!&BvC@ItkCijRN6nw%Nonhke&n)m<2Rd6;C>y&@Zu$1^q>Em z!oB|qyx80P4DSCu(-iJGb9kkk4DR=V1>D!4!@FOw^(^5&U%;a`*!&9a^Cdjf{2K1_ z6}-~?2JZ7Uywm&^?(+@2)4sLvsUGj)<#*rq|DD|^>fqh8%nxwCPLJ?Zj|X_6oD)2H zw&joTOzSzr{dl7nUG_P7mz}Q(+~;F>_eVBAh5LL0Pd;e#GkB)&C8qG`livRSb~B3R zaPN5rPc*-Pd(U&Y=PcocatgTTtl*wg!W-qR;r=>U!F}I0^qbp0)O3ALwS_0Qu=xh= z>)gQyt+R#u>ud-2^KpPzv)f&ruO8mWkMLA}g6DcX!uz{e{uv&Bx_R`Y=(F-;_)7o& z6y9q-fe-Q-Jd&sIQl7#6d@SIx=5u%O4p1*-7FSGd??(Q%O$81DH~c=T$UPvAa3gC}pZ z`4sN+b9nU@o6q1rzkoOIvH2YC^Gmws3%I|&ui)ulTYd@m_cImTb2sql3-56C`d-5m z`4&FrHs8Q~ojZ7R>tDUfZ|UD?J=nv2e>!-joCDnFd%EV2bX{KqJiCpp^91+&5#DJ2 z4EOoy#g{#g63d^!eLjX4nxDdbK7n_dpTT`Tg(tVQ^~~WupTR55FW^3((>1@OYyS&) zb~{_o3hwzOywUs`?(-Eqy1nIZ;67i&3(argKHtDQ&F|no-@=nmv-Navzpf7O^bL2i z_u=qNeuSsLWb*_4P39+f^A__F?(09pyZhPtqaVNQx!(sS@bL*YAH)5AmB2lB2G76v z&R6@8!b|xap2{=0AJ+mN{iv-ohx>6Y;fdx8xc`0KE4b&B@Ju;tx*k_>&)L8$<}7+CJ>yKHtI{&F|s9&JI3kod>wDvxj@m5uR;rodeu|&OO0BXM_*RIm10C zddW3ChZjGlp2Iz73itIS@aV;sGo$Nua}Lk7?hM}k+g-1&mjyikuz3#mzaM@H_rD*$ zfcxJMU&8%7uHot5+VfWM;qS~h@ci%W&%M_0^48{Ccyr^sT^(;j|A6JUaKEnh@LcQZ z;Kkot&H?W0?BTx7Biz?H(6!DH?$^~B-f2IhpSbMjpmk2*zRno#>zu-Uoe5p*OyPdr z&f(o(+rDM+>~G8$@JOD+>yynd;qCj)S8%`nOL(jGtl_Diw}OwF-@qHq*YHB?Y2cC0 z-wxg=zlA5t-@^yZckovGbAabMUq|><>lxs+@=x$s>lxv_=Fjj(>xo`^+0RJpnZW&i z7{i;_+5K<|_s{JUxW7)#=*pSHeH|IReA(^o{t0h>+B}Dse{c7nCEWiUUjeUv+m34m z_xTe3y_#Rcecvj&<~Q)=TkZGGHQe`Y3(wwV>ulgYzk@fLZ{ex_Jk%b((>f1u-^U*A zIY)S_oD+Pd$I;6!d)U3j_F;C7=kW38ZGL@?H}KWZ+5F)eAFuIgbGc4GU)eRjy2dMb z^m5DHz!N>L;XAEw3t!3`cqQM#C-N4a%J=ZIyo2}h1H9DZ9=_1yBRtnW4Dgx!3}1VH z^#A*RuOa!&^y>aFfv5K}PvD;8&l7SqpJG0jFW{cz$Kg4#=9ic+TOyQoBz@vqoml?c%yPcPmuKV*G-Y6%7drl7beJkO`9q)0~ zlQlegqj?SYGlMsGGGD;scbVt#`fR?0`+cK;M_+IKT*1B1 zC4A8Q8ty%;;QqWDc=RLoyfxgPcMDH6-@x-ogoWe`_9A3#Y_$Xh% znqGmhj^3c3cJA`?i9YnlIu0x%V3GITgH7&Ia!Hy&9hCylmmouiE-|a6kS% z+>f(^CwIU1RsRq0Oy0x$yW0E!_v1gohr8MQ2rp9eGrZCBMz6l?r}uvXFOKRb+~=q8 z=6yDwz_*QoC03_spYKT-m4Ps>tDl* zKeL<)?$5h{Cxgw`aKArn;hxjLlRvkd9sH=rE!=bV@aQirr=#oh_58ekE!_Lrz@vY+p6}q^&lc|YjXk_mP6zkrJ;0M2{QA}FZx8qT z=K%NI6Fj=>{3>UJ$MQ2gd8ExpKXuu+_%qBW@a{9sW4OP+n!*pQ9ajSP`5FATHJ`$L zehwcrpTT{80Z)F;)|12iye#2`<_oz0dG;0Db4qxloHg8YD!Aut;GJ@6xaVx)p3}ew zx3 z2rrcr!)y5z?)fvgU!N&F-rM@;@Km0`JwJ#0b-RQY$|>NLd9caR2*%W^m6*;l&$u--CNj2KSr=yi-mN_nal%a|(FzCR@)6?l~pg-yg2w z(W@<|g8TdiUTD6C`}`K(Xug5_{0<(y#@5rqeSQzmG~dB}{s6Bu-@|?W2#uvol+~@c3 zO!FPw=MV5k^F2J(=ekFDqx;MWp6dP5NLS7oJ}77MGnYNjloP{!JyZCt-fa6bgU8BA z;hr;x-&Q#bc%qyo-2Xcz1-!Y>eXrhsuHeZd%}aQ!_jPM{`AIfk!87>=K7Opt*YNno z54c*-7M^^kc>_<2`(MrP;QeFGTX^yq^F6%ydh-sRKHr}A0Iwcr^F2J#{v6@izu55( z@Z?$dyITJU_x~Qu8D4$10v^A~=5u&{viT*v|9X4g0-k@Qov#%<()lXkpar` zm+k)mZ{A|_C;0FeHb26XPqF7c!~2i6{g2wqo_9K~3A{+`yu|SC`Q{1S|2s)D`a^8} zDZKtX^Eo`%dNO$SUOQh4c%k_mUdos7{^QkW`0(%QGyP|5oh3Z_fbIVpp8kT(SMY4G z`3=0hp{=Ke*Eh8FY~huTtAURnvgh5whv(V(+QVbL@9gOJvz!CG`9kv^o@zZuc=r2t zz6N;n9`h5ty@BPA@c3VBAI|XhFVyFsyXH*et4XPNKd!(Gi=cyurGJ-mFEc?YlWX?}nw+MgcYeT1EtBRqeR z`3ause2wt&{kIbj=?A^A`1RmYP?msj5aBuSz zUfth(4o@Fqp273`nJ?hO=bGpAhnX+o?b5t}M>=0Cc=FM9yd}K)0rLtT>;1+CUSxJ& zYWl)_3vYG48hF;*dD+1`&A0GgzK7>;R-fVRE7WIr-r72Qc=WG!{T|`vPuu(eFWzVK zC;0eUd)^V=+|>LGA9P&NFI@IK`Jg@Tg#P_@zNYY{uCD}MJZc%t?6@a(tkd>!Gb<_CBtKf%*y z*l~^U=t<^hc=F4(&gc#3^ZRW-C-CmAHXp;s-?I5By#Ho<-UJ@${rn7`>$p;Q{HbP!b&b)wE|74%Hoal|0J@@CGz?0kAd<^&Jox(jQffve|!96F1d(Ir*C?|t^&I0Z^Iebvg67D$# z+;dj&{MEGw|qF%tv@~$A?|bpW($X+I;lp%kx$J zhWP~E%VT(UU&~M6-pd(0(>halEuX{t-?Q~(@cNI<7w}TYmBaH_**cf-?y2T0xc9z< z7r$oDyN1Wwp9)_8uFY@YT_Z-3QtT6p^F=6kyKvx9p-57&4PFI(HU zBfS1X^8wy`jrj=o<2u7*?Q^uh>}R9nn!s!2$MEru_PkSg|9JBm-0ur1yiu>_@K&C| zJ6)#>c>h*AFFAaWFX8^WP{4~jzvJqL3B*~43T2hUz)_tgVD*8R4J*Ke`>BfOIj@c1T{Kf+5rKEu;DS$?F~XaDm8 zxqJd|jy50Df5m(XFWzRJz$4|);O#wa{d2gV*9;yjX8}*-IlR<)S;8yL7w|#8f_tw@ zc=rhF`5Nx;-zvE0Y~ax&EvJTi&KBmF72a zpReKBS6KcQ?(+@2`%0T{;l96nc=T?&9y@pW%(y~FCXEJ{0#Tw zjoy0Mx8a#~yc4*e_ZS|1kIhfv{(K4Cb7t^VIVs$8=5Wu+;JI=ZaL>u%z7I=y@iaTG z0`BuGc=8=KU&4KUP1k%y*Z0#l@a8)$zlMAM7G6Bv<{P-r@8HQZY`%s2{2tzE{s8xV z9^n%`KEum9JpAf?PV`HcJ#6I@_;5R$pTZO6Ch$=?Gk9@F%Sqw+XP9U3x$+nAUO72@ zoLSBiUVo{1LD$cxEBfozE4crC1#7rp7Zto!&Iax|HQaNy@Lo9$+;ett&uQVKa`te~ z>ENDofX6>=`_scc=Lq+l0iG)71oxZ~?m1_8uAJy?mwon}7+&b#pTg_u=U-hH3A~lh z;N@4^dv&6ersG<}3wZ?}pJ4eL zc=a>pHN4evZQ;$kZT$`0k9T*CxA0+N&)dQMx;nteTiSdNk9E99c(=Cs0p9+Ht^Wk? zbzCF7)A>5X{X9m$eA!#~3B37Dd)_JBum1$z=y+%FPM*U1Z?fZ^!~J+Oc&X!Bz%zLc zAAaBRm+(>hR=|gw*nA21`_vkq>$oa-Dc`{3$60<2_v78d{rYX-nbyC9`+01y@jbl$ zXj{(#?)TFk9_hG_@I*eq)57vka6jG=Uh25c@LJb(^efl&_8O1j^-XO(3Ec1dGkE(+ zHlM=i=-#);zU(ovsxaS|?(Hm@jfcyLjUTA)V`}~=% z`KY_>xzA7F*&A&=G2G{;@J90q+~;TT>`j)R!hL=Y?=+vmeSQIN-fa0xxc9Gs_fPn2 z`@9sMe1mxjAAj}{SMzIlv^B5b)sL9haPRRJUi`StH}GD*gSS6y^DVrS@8Q0l4(`{{ z0Y3bO?Pm}7KNoz2`*|7Q={qcEg!?+q@cIw!xFY@j-#;&HAJ3G@FKB!|6I`5{<&}uFSS2Mc>5^Zp8@Xq zBiz4Va)!rRfAp)D_s{HiZT%Ct=TG5No!e72Lm{QNkLzQE>>aNm~! zJ}Tz~&mLqsBYb?MdGza-{aN`OJW_rP&*W2h|6t2c;NG7Zyu8PIuI~RS+|SpXuK5h^ z&zHkJcL|Sw+SXaXlh>QC;MMEQOL#3`!^=0?d<8GwWWIsN@*3W}&E~gsJ#PcAf7#}D zbRAa0`}ykN`91A?9pHYxdU&PzBix_w1ozw#Ui`Hk*BM^Pqu;pfb19#|eVs8p zy^kH&6z;#TOyHg~gXhXg;l7SJ+;cK`rJM!ab8@)nEa9zk3b?Oh1^1j1-YaJf_nZpu zIUD$>oEq*qTe#;m@c6!VzIJfWY2lu;ho{Qv;GT1Udrl9}m2-qo^!Rj*kMQyyUvhOn zJi{A#^p4A(ho5WUiT?d5yi-mBALKK*=g;B3KN&o_r>%bhPvkk=^9y*c^{?QWa!Pn1 zU&B3r179h>hF8kj!W(%5_xu*_*TWv(DW`)E@&nxSkMOP5Kft4V+5Vs4iF|~6e)P`E zUWMxio+&4W7xF3G^Jj3su2OiVoH@LaXK>G7!2LSS;f1c#CEUMnR=_=H1@Dwo!aZjV z_nZoz+|TY~8@T6e;eMPAe9->v;KjYI4=ucs@8PAqgU9k7?)!Cw_gc>Y&y;h5r}wdS zj__RbXLux^94>pi)?5tlw4NzEQBDFMw4NC})qDycl|P63dC%aZ&iexH_w6Oza|?L- zhhKX2da;7%@)GXX%Np+Mtl(*;e!>&I58J{$w}Gd3{IaWc?%<=mg%`KC`8~Xpckpp; z^9Oh-@8Rud+x!t8-`9MA5BD%X!N>cVk95tS;r@O*`pwJU`upt(Jb94qe+>8eDLmJF z0{8PfgIAhQ;r@JcxaVZ>PB{y>=j3qDS;C_S+wm4~&so7er-Wz9S;IZ2f_u&eUMZ)B zd(IZ_ISqVJ&JON5E!=bV@Z=%3KOH>M;{)7tdU&CnBiwTaxaXYUjdDh~=bYi56TJ)l zSIz|PIn!%=4X^Zk6};}2SNER{yn4HN4KFpnh3A@Y;IVuMAC$9)dv7{;_OrI01HAYd z^PYYK^CP^J5AaHUg4gmhe6IbB-hFw#e4P`xuQR4Ae+nOOX6sMjt8ot!dvAO@ai@;zk-jqH!tDwE6msM_T$Yfcp95;;7Mz~h37h7 z4cyPy4({ixrR#j{;oZmB^LFs)(dGyG()D%@a|SNe})&z ziGJ&{xA7w^X990>^B7*g+J4Sf72%h}PNV&1}Iov%Ip z<~HBKN1d+&e9-lEgeN**1KiKo3GU}>r0aa0=^tn7jNWtE+eY&fczQ#dpThlhH-Sf^ z?f(p3>%Nr2`*+&>9G)pBgO4||^(^4SN1NyH=8Nxqb-gd)oz7PQ&vl<%!7H7w3Vx#( z+wYq;@bn|U=IVKCcr`WO!V}Fm@aq2*S2;U)EpOrZEiHc!Z*OGY!TXz-AK;DhdwBmB zHh+efU;edM>&$-pvR8d>^96i-qIm_6KKX-JIa_%0DdsJ_yOa4I-rU^0gLii@Kfn{^ z^zdS4^GA56^$hUg&Nd(Y&gDLII<5)4xV6p4@FX!`z?Ne&DJo{Z+e+@5gW%IrtzMgx&)~?4L z=Er}tezx%b=3jr+w>`Z0-;cZU13Z?W;F)})_4<0UM_YajA0A~sgXjNX&+EtSIgkGn z{azdMeq1@+k826{<66P}xHfPHJkt6j-Ov2G=ufr}6L|ZN zy6)iBhs>w&M)L{0)p5<>!^v_|xbMRp?)#9zeIFKZ--jIDYdwBn_kH-=*VsO+FyFlX z@mJSZ39lRTH9UWr`3By<)O-i86%cq+U@Z!1V zC;0H)<|Dk6NB{q258KM-C-Cz5<}v;I&8P5I`=7zz@R5hBeOSQXCSSpa^AoS;*YHkW z!AtoDURTe&%4y)gujk#tH#fF*9^k&7Biz?Bz&f80o&~)6Zd*?Qf5Ye6KCIv$&^kA8-=8hq*VDj#Jv+Fsr-T2Cp7#Lv{W-yX zJ!iPDC;IP~`|sCM({v_}leU$Cb41PQL0`BWs!hJmj+}E>$H;=LPRPec;cLVqJ z?BKqhJ>1vR!F@dkxUXk`KUnKI!5<}`9546Z*E5CtdJ?#=X9n-S+V&@dKi1C+{K@ha z+}E>)`+6$4uV(`vzRK3qz@M$>-NC)j2e_~22>102a9_^}Ui?>EPxQXa{jdGJz-R4j54{LsekGHk;M}KsAzHah~cDxgK z@fORU!aq^-34FYj<)`r5YJLuHZf)}m_+2%h!?VLTUhU@!eoxJp@brF`U%@lYZ{X2q z+3{}S@6h|G1|Hwr=3Dq9l)s1Pnm@pon(yJ|3oU z=4bHa{`Ps@9G=P-aGzgZ;|07^&I;bjOL!w+!)tj3FXbC}C9mN=zlGt1 zx=z>dPV*ahtMzQ*k$eaD{5?F=`~lwo`+Kg|bA%6iJi?qvj9r?3I>( zgm*t-eu6h2w)1s{_YbrA$)8>Jzg?J5;jw%MALVm+E?>ZXehIHNzk=sKY3nKB!^8A? z1n-qo!#mAy;k~?pw|d;c-z>iYJLm%_3z;67ijy(bHJuK6Wh>tDfZ&9C8p-EH8#=C^QP&ki0x+Sb2^ z`|%#&x#o{>zm7(Dsrl%yF84Y50(-wPf!B{TkLk*p!BaiX;JF?z;gRn%{fliM)^LA6 zS;HGW-oX>)?BRah9pK5A+VdXay>d=)e;;y&$6sPOlfS;)=kyikQ+TEM8Qfow=5)<3 z;J!agc%cR`5adYr5WtZ{W#S+4{F|&u`(K9v|TTx^slr%0I#VK5&NjnxDM?a-V&E z3Xi|q)<1*$`seUU^9#7=FX5f$SMW^N%NicZH*lZd()IJg4qhmyg%6Li_YZscDDU8v za*ps$KVP5V_0INlgpUuk=Z!vq^P<fuB=AQ4oWWap3Qv`@fctg5g!_G91s|2OhWqEM8+iJ4w*D>L zKgZj_OU>`$x$Zv)c&qs%+DR!##SKiBT(J9wm=7Tzl70FS=j)_;VD^8(K_e}*@D zJ(~Q(<$0+zKZTd-$qe3Ueh&BZxPV8GxAiaK!>e^(;HBo*@OWkO8+fbvEj-oy4nAsr z5BJX>I(Vh~zyV&%dwBLDTjv1p)rS*&kdN?O>z|C5=ga#!h1Z&&!3X&q?$_x8o;*S4 z3+{be(KWw@SMmzp>3C~+qxpvJ*AcvxxA0Cm2YCDqc3elezfPUtx#rJs?|J->m**wX z;~Ct4uR4c!%3r`o{XTaIPoHS(U%~yntl_2RH}G1%g%9!_Jkjqr_w;YH^&H^-bL$aa zX#NEE`^*{MXnyifm-`$&$)0x#A2dIM`}s=Ywfd96{XV&bmwMh6y!Zrt9{}EJegpUO zwS|wG-@$!<_VDzZwEu8l|BFgPF89CE{1o2HXYlb=mOqEb->UtG`+N?sbiXR#xpLOl!aPPwgp8Ypl=N9hgcL%RDzlZno13Z12 zQTjbiO<b@a~6fegm(RvxSHA1@AP!hsWP(`3HC*Kf)XN2_8M&a?aQIj)n-f4auc@bsCsp2@#m?!WKz6kck62KVEg!&}WS;QqdP2_H4T zg8TdGH9Y+uTmJ^`*W(smYJLZAbv^Fsx*iYk`p4{fkJtDK9zVww&z{J{rX+Q3(arfp1*}Rn%}`ae^1x^0bc)z_8*=+ z$NU8MzMbKP<|iM%+<))e6y9il29IB1&pU??nqR>EdSAkm@74apOZgh^J=wrp&2QoU zy0C+fn%~p)KI8yTH`;%=pT`rt)chIluOpLxzuf;;^HX@F^~~U-=I3xf?+bYPeYXB3 z+|T<8?)Td@yi?8w?#H!-N6)qA-NF5SyN73*KfwKSk0ZR&{0Z)TJHtE8PyXX_|9$;a zc>4Xe{u$iYGl!R&U%)+o2_H1Sg8Td$o;=Uivw{2k7G7w62lw^w;f>}GaGyWI2hE@0 zK7WQM&$snYBKthjfA8t@Q+T2I8QkaR@J90sxX&-)gXULo@6Q^Zyg>U85Bm?VHNS%= zx~})|Uh@aI|6DV`N6nA$Og|r;;ki8eh|B#?f546_fyeR`p6chjIXshR@Ig6CcqL!K z{d|@1Sov#sBCp_`^0)9tk6U=B#|L<+&ry50ujdH&^$hUTKSzQ`A8F?!)<6BaNc4CH z_wQNF;id8yaPMahA0J}B|6Rf(?Pmcm^|*q!dfpn|D}M_gzu*;bLlxe*ZczR zpIa{BwdPlFe?M8mJN3DO_sZGANBIu!*XbUf>iz!#?&tRiFExLH`{$Nt_@Md84KMfK z=cn-GMcRLOt)DCB@Lux^c&5({mhkvTEq?{C)Xz0M*Zc20>G1*XzaKxs zTRray?&t9gA2mO@(d9n-d7Q%2t@ar{yxM#YFEzh_d;SvMYJLUx>v|0zHNS!T@0Yjm zRG%B{;GVySmzqDo{r%VxK4|^~_s>Jl@Z`m|Z<8Be?!WhB3NJK2gEx8|nbS4Dfcwu` zIlR(+qks>}S;M0rxAkw}-iIwb)BFzZ^DVs4diL;A-oY#7AK{Js1o!7X(=|VtT%H%7 zpTd)u=(>P=Pv-DK^9y*V>u3pYG{1uTb-jiUn%}^?m)SbE@I?2$9o%2A3drPLF%|pvNb;zn?$DOXW{K`trPZ52x@}^D}s)9?s#7UQZYB zO!G^)-}hGVO7m;D-}g3f|9q*2SHA!7+V>yc>Tw4jw4NSbDCY<-@JQ=S;C{bN;lqRMzBGrA@(iBp^YA6S(Bl$b>G1~c$Ge3$ z%HP5LcssftAK`-@pWvPHM|k3&Bf(Su9O+{&_kU2%6z=Ox;GOm}h5Mf`$l#6U7w}e| z!wWqw;iZ1xRKY9dY~Z!LrYmO$Z}qr?_j-JU`*}anb$>p?eLtg*z1;s?f8HvF`}?8< zp6j?$cz9iak6ORf?ys8zJl6aO&-C+G6kql~^ZvvAeMJl@&KmB=yM_Dl zws4={!+pMo7dl@9y!ZWw`|(ElzG2vBJ&xg-@~7}zp1}R}G=t}Qyo8ThX94$pSiyZC zD!9*Y;6C5LeIHtQr^h{f(Bl)l(ChUGkF`IM{@zF5=T`GEyvywMaSHF{30*lUJks-K zaK8^P;q`0nzFoivJ?{!W%1gNCSMaE{{0+R;WB;e`PxG@jzr}o`$No>B@8k!}-~5vf zSNGK(UfjLB@+1Az%m?~i%}?<1&gLV$Q_dOQ+`;CfPrQ8H&ow`RcQc!h;pv^s6L@x; z4__VE44&N1d;$OPLv4Qw_=EM|J1F7ZuiAN8!_#Md>(%o%@Sk|vZ(jKx{`ty3z_VYo z{2uQ4XZYX#ko`U6@h4sG=cARMz!T-q;GUnuJ%0u7K48zghI@Vm_xv6FtF+D*{zQ2P zf4c7*Jh_b>e^1x(9^uh<+4+iYdAV=3@+a`0kjL=X%BS#OlF#6NepC8)+j^Gp-%(Bp z{~z)-e9%5ra6c~_xbH&?KP&$LznQL=9$x8rk8pq90q)N`nO^SyZIz$E?;g45N4`0iV@J#t9xaY^8e7SGWSI!Ln zN_h(Rp3LEqu9plx=z3Yf->Uoy{(JHb+&)Sgl`r8>kr(jl`0lIMmlfQPx1{U* zR`CA$w$2T_dVzTjk6&oMh5J4w$MEt_Ha~^;4>M2T-2=>L@ZxjK7x1_1xN`XK$(QgylNa!v_XqC# zQ^I|JDtM)R*uZ^%YPj#u7Vi7gzH;J!aIxL;oj`0aJPIsCrzCH$fC0)F=X!2Ld7 z!hL@#y7p%S_x-8izCT;I?@t5w{n^2Ne-7|3(((52$IAzJ@;duh*YO_UFY)sWf3u%oc;V+49_|D1$nOJse>%B(m5!ortMd*|eoW^bUc5x- z9iI8?2;BE!1~2@)!+*`sJN!@myu)ih@9=QF!~K0-1^0c}zLcbG_g6^PWG}<}-No_2xM|SI!dd`CEAUB+F^w(G$&E zc>8$sJ>2upaL{zuK{@Km0`;}=@~0-oIcX;=9< zyt|#vFX4l{fM>V0`4zl;sjai5CpN!^H}VSJy<&Z}&JDbOndR5?#(WEpe!{$gcOPTx z*};ojnz!&uzK2I|uyuCu^gpbh2Y4p$;knNH5ng?P^=yEr@3wWG;E9fFgvXy~^YQJj z=`B3;mhQcUhu*?NZ{eZ0@X%X$=q)_-79M&F550wl-qO9d@X%X$=q)_-mhQcUhu*?N zZ{eZ0@X%X$=q)_-79M&F550wl-ois~;i0$i(A(Qz(_47xE!}$y550wl-ois~;i0$i z&|7%uEj;uV9(oH8y`_6^;i0$i&|7%uE!}$y550wl-ois~;i0$i&|7%uEj;uV9(oIp za@}v?p||kRTX^X0r(M%qc<3$NdkYV}g@@k4LvP`sxA4$gc<3!W^cEg^3lF`edvD>P zxA4$gc<3$NdkYV}g@@k4LvP`sxA4$gc<3!W^cEg^3lF`8hu*?NZ{eZ0cetjv@X%Yj z_ZA*{3lF`8hu*?Vy^meMz0Wy3^cEg^3lF`8hu+e?xA4$gc<3!W^p@_ug@@k4LvP`s zxA4$gc<3!W^cEg^3lF`8hxZBa&|7%uEj;x0j@R@S9(qgn-ois~;i0$i&|7%uEj;uV z9(oH8y@iL~!b5NA-dlL+Ej;uV9(qgn-ois~;i0$i&|7%uEj;uV9(oH8y@iL~!b5N2 zp||kRTX^X0ov!IEJoJ|Cy@iL~!b5N2p||kRTX^U#JoFYGdJ7M|g@@kKy|?htTX^U# zJoJ|Cy@iL~!b5N2p||kRTX^U#JoFYGdJ7M|g@^YE@X%X$=q)_-c6Lo~;i0#5?=3v^ z79M&F550wl-onHCTX^U#JoFYGdJ7M|rF(DTp||kRTX^U#-FpiUy@iL~!b5N2p||kR zTX^U#JoFYGdJ7NtTX^U#JoFYGdi!6l=`B3;mhQcUhu*?NZ{eZ0@X%X$=q)_-79M&F z550wl-qO9d@X%X$=q)_-mhQcUhu*?NZ{eZ0@X%X$=q)_-79M&F550wl-ois~;i0$i z(Azs-(_47xE!}$y550wl-ois~;i0$i&|7%uEj;uV9(oH8y`_6^;i0$i&|7%uE!}$y z550wl-ois~;i0$i&|7%uEj;uV9(oH8?-SsmxA4$gc-S~w`c39FyuX({?-rii zz`TKn-ois~;i0$i&|7%uEj;uV9_u_F;mrfAX9GO+79M&F554{LYkCV0y`_6^;i0$i z&|7%uEj;uV9(oH8y@iL~!b5NA-dlL+Ej;uV9(oH8y@iL~!b5N2p||kRTX^U#JoFYG zdJ7M|g@@k46YniN^cEg^3lF`$>ovWFhu+e?xA4$gc<3!W^cEg^3lF`8hu*?NZ{eZ0 zbnh)Z^cEg^3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F550wl-ois~;m!SZzlDe1!b5N2 zp|^Ltrnm6WTe|la9(oH8y@iL~!b5N2p||kRTX^U#JoJ|Cy@iL~!b5N2p||kRTX^U# zJoFYGdJ7M|g@@k4LvP`sxA4$gc<3!W@!rBiZ{eZ0@X*`4U(;K7=q=rQ3lF`8hu*?N zZ{eZ0@X%X$=q)_-79M&__uj%oZ{eZ0@X%X$=q)_-79M&F550wl-ois~;i0$i&|7%u zEj;uV9`3jB&|7%uEj;u#y{5PD&|A9q79M&F550wl-onG@32^Uo4iCMBhu*?NZ|UA! zc<3!W^cEg^3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F550wl-onGrW$@5jc<3!W^!78Z z=`B3;mhQcUhu*?NZ{eZ0@X%X$=q)_-79M&F551*(Z{eZ0@X%X$=q)_-79M&F550wl z-ois~;i0$i&|7%uEj;uV9(oH8KbOHnZ{eZ0@X*`Oyr#GC&|A9q79M&F550wl-ois~ z;i0$i&|7%uEj;v=?!ASF-ois~;i0$i&|7%uEj;uV9(oH8y@iL~!b5N2p||kRTX^U# zJiJeUhu*?NZ{eZ0_qe9F@X%Yj_ZA*{3lF`8hu*?NZ{eZ0@X%X$=q)_-mhQcUhu*?N zZ{eZ0@X%X$=q)_-79M&F550wl-ois~;i0$i&|7%uEj;{O1`oZ3hu*?NZ|`|cZ{eZ0 zbnh)Z^cEg^3lF`8hu*?NZ{eZ0@X%X$=q=rQ3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F z550wl-ois~;i0$i&|7%8-@-$0;i0$i(A#@m(_47xE!}$y550wl-ois~;i0$i@Oc6} z^cEg^3lF`edvD>PxA4$gc<3!W^cEg^3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F4?mZ| zLvP`sxA4&0dtcLAc<3$NdkYV}g@@k4LvP`sxA4$gc<3!W^cEg^OZVQwLvP`sxA4$g zc<3!W^cEg^3lF`8hu*?NZ{eZ0@X%X$=q)_-79M^sgNNS2LvP`sx1V)QZ{eZ0bnh)Z z^cEg^3lF`8hu*?NZ{eZ0@X%X$=q=rQ3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F550wl z-ois~;i0$i&|7$Tp8yZNg@@k4LvQET^cEg^OZVQwLvP`sxA4$gc<3!W^cEg^3lF`8 zhu+e?xA4$gc<3!W^cEg^3lF`8hu*?NZ{eZ0@X%X$=q)_-79M&F550wlpUdE(xA4$g zcuq@kaWdi~oI1zo+>L-1Gf)-skf_u---jOV`!K-c-?5w%ynVO%8SeWK>3gPretjP%aNh_2Ios!bA7+^MeMsTH z4|BNhLk9PK$l<;Z1>E;x1^0be!+jqrxbMRT?)%WdeIIsk--j0N`>==mK6G&3haT?x zFu;8uPH^9cGu-zf$}aDpz7G?)??Vb7Z}-%z>u3(o?qr_9n>(8?=yx$+!fVYJaR2?t z3O>#)r-T;|GGD_>c?HkqTX^ws%h|)rdzyFj&oe*Jm2-qwdfowEYyJdpNxe4uMSz)$A`{B%CRPv-;tbUwgO z=fiLRkof>Voeyx$hb4UVi;u9^3wZwq`#ZH19>39i4G-K)`j`Fh;Nbb-e+LKeFZJ*< z?Ky%E?lnBPwAJt6nR^2-uWI$j@aRhBE&Yn-9X$V0^PcX%cMo4(*6Ig%;q?=|azBCB z?!oWO=fm_x>(3D0dVK_6UB=q8fEV7*7#@555+3+CuHeztEdKzW{g`` z{bcpQ@6N})^8N|o>YoU%{#n4)KQUbWvxHCAv39QD)%DE};PEe-C-BP0C54y2ZS@&E z_`3DO8lJgt;Prv!9KzMlC49Px)z|R$hUPnXa|QDTu6^tnu6?Y9Yai?2+Q)jh_OU%& z``AGDeQbnlADigff8g53f?Lezq4u#5UTXh=Ykyn7wZARl+Q(LK?PCXU?PCdC`&b6o zezk^czslj-uS$4+Ynu;ScySB!BY5Lp!?o^qaP3PST>DZF*S@reYhN1S+Lun?S}(yZ z=i{z@DTb>bmT>jM0lf5nNZ{&+6t4XzgZG}ZhHL-H;q{GdTsCm^X8~7#9>Ue1Te$Y0 zBf7>NuKlNnNB>~$+{2rfm=Ey$d*&m2dWiW1kN?0tNay1l-PJs#|D$;XuU}@qfEV94 zkKyIR%$M-^4(18G{it~gZ(eGi!Sf%Oui@39c@D2{Z(hJ7_d|H~F{`iO*~86u@cwq@ z4ZQid`7wNWxp@mu{>Qw7XOA%N;lrKHM|k{S<|pv{BJtCgEy1qH1PNh=Ev~nGSt zFus0W3wZE7%UQvv7n`T>#M_y{%Y|L<8eYHCa&q`|b*taN+jm%f36K5vZsF~j_f`1{|TjZ5O~iQtjX+rm&DZeZs@DFTURnJHo@C~S@W{P{*MDgB+aKZ;y!H37M?b`Sc>P?W2kf{Sd>|4=cF(A%UwOQn>nI4Oc(paP`9mu6`)t!+DQ7vtMuF%}dQIc=2NMBe?Fn z)^OeT*uiz*qk-$b>oHvSJzBW#dvtK!_vqoe@4AQUzUv6reUB6P>3yWz&gYNrd&Kbg z4c2ci-(9{@=s(ecAyYpJU&54gP38jxTt;?ZY8Ff2!3l;V<+06w06ju6GI7^``KD*Z9J9y=!=TB_Ch7u6F|uyuP4c*>VoyePmw3Gq2y$ zy}p8%?j8IJ=h}Sk;kw>Ey!~-&&j8o;j_`J2^%H!!w)qLXxTbk8_$p6 z(N9=?3cre9ZwA-(uHm}g9Ior#z;(R^T-SRD*Y%cgUGElN`}J0EU2g~f%!T%M89iLr zyN4Iov3?uiy513-jkT{5NgB#qi?%XPp_B1b#EGPvNTH!2kLA_CBP5tGFLm;WzX8 zEnM|G_+uV#?P=ipJE=YVUH@w1JHVA6+-cqqYJUh<{Sy8YmG%D$uKEnF>s`ZDe+XB5 zO1SE4xZ1OWtGD2A*3OStk=xZ0n=RbRl>{zJIxkKmdQ zHC**A{O5dpJ9z2)!wLMff9~=_`UigYGi?4W;rg7I!q54<&HoIZf8W-70e`CJAHucX zkKoVq`Wmj!Wi9-bUf;npZ_fySo7Ye9-aof4?mF)`^=Ay%d`sXT@%$9N^7d@tSN{|1 z&jOx!eFeYlmu){ef@{2v;V=EN<+pI%XBgnG8?1hWtA6pP^Zt2(*T-#fqzo-2Yxz#;494^cv!#Y6|zViAlJb0(oAHmmNzk?62wfbZD*6TZX z{dTM0!yB(3;lo?4{)F!J;XTp+-u?wVx}x>Z5*~a00X+5k6rLYjd)9E(Z{X!yto{&Q zc>We%zuD@K;Huxj+c#PLFy!7?h!~07<_RM2Fa>ce}_`@cNj>KE`J@^t~P&$Idixaw1Q`*N#a!w)=v18-ku^@s4> z>$mXsrB;6gSN)D&Tm3P-^85}Syu|AFaMh3SMya zKY&l3pTdXdTm2fY`VBnx`a`(txA5}F@{iz)E7|egsXlFAD&_LNAP2B z&kjCRR)0+Q`VL+{-Rk%7@X9v6BRqea)t|suULS7e{U1El>KE{}*DvAaxmJHb_xcnb zYyXF9Ki|Oh`gI7`>(Lgj^>_r=dfCDCzV{fe_q`ol?*sR6y$>AWxxeo^f$!A+e=+a> z_BG!B@Lv57*XzXrJkb6F*XzX^uKEq#+j9s%@%C)t!&9vNNAUP6zW=}zuRn$#d3^`3 zpKSSicyv{t|M0-;PvEK#|8m~{`IEf=;fd!j;qBJy58$d#;epq$;f3dK;K>s${}8VF zEj)jM)gQrY&)>nzbFBUtuKEsMKHlp0@Lv57FCS<1Cveq=_nG&9a<r;4smesG}E3e(dB&HFib58=gotv-UcUcZFLpRoEBJaIpOm+lF?bx+~tCoMmNH&?UI zL2Gz@74sZEoMky1cytY`FW~K!%n#wKE1Q?_%6$u8d3!2&eND?bg6B6dui>rNxA6Sh zR^P$fpEmE|&F`A;;eq$V08egY^&@=pc24l}*R1{o-rmG~abL{C&)B%c@Ya0=&pvJS z2k^k_6L|btt54z8-aLb6cQIeX%e$Iy;Mv-|fETwnKZG~Ft}A%C%<*FLa@Yahtr)mLqNOStxdEnNFR1=l`s1lK-L z!?h3W;MxZoxb}f#xb}e--h9ojtAm$;&qH|O^?P{k`@jI#elo$epPaz8p9J@x&rj_q zAzb@O1lNADfNMWl!nL2Q;Mz|P;Mz|zxb~AZT>D85*M72rYdh3y<&T{SR0E4zB$A;q(0Kyu#k+AH&N_{r8!5(ZX~0 z4&Gkc>U;QgJ%5BJo-@H?_Y=7CgGbEU|0|xqfCruv!-q>+JC|_fui!WH{1o1LP6n^t z*Kp0m;5h?)xPp2m;c3;4iAH(nO`3LaQa}s#&p2C%%!5{AV8+hV51w3{?ge$*fTKyqB`A_pLy!QGdc;mi<5C35~$MD|0 zgSYN`c;i08%lF&&a!=sBd-!OK*9WYA0bjW<;febJJabRsx%(Pkz0ZE{241@#!VAyY z!ju1g^qFxyg4f=k6msb3cJ6?%`wR{l9Wwz+?9%JaRvP4|lhD zmBLr|G+)EZdzf$F!M)56;kDOq;hEPT!TbAK&JJEa!2B5AyLa&7fmXkV*Iqxur=PNQ zeFBfaZ8_mt^ZuW{X}*95>VNq3?^b^RAKX)TaVcxh8lGLj_aAuTIfwA+-|Y9+@L&J* zhtACZ1|D7F$!ENSFMPk+eQ8_zhoc{nNwsebfnlea}CE>wmXparV65PQPyqSN|mN z8+d*S5Bz#J@Za$I0n4A*$=;d=cV;A(%co%fGkze0H6 zucu46UcXjwy?(9XZ@H1JuN+?e<`d4me>j47USGrG*y{Ii{r$oK*Ln<|G{4@x=ZEmf z^Ox{T+}PT`g6kZN9DX^k-@yBu*nU{Sui^Da@XG(5tA+op*LQIB&ji1L*Pp-xzux%C z^Zrr&60Z6TuKG1x^@s4AdV5N^{{G?^uJLW*8s7o_RUh9GK3&J=PyCem^`_pQC0y$< zgWu8X*KpOB@Oyjx7OwZh$8hye3s?V~z+ZY7o1ekC^Xt<5iQ(PfTKy8P_NVafSyrFH zwf`K#_4-o6RbRs&?ESNYYk%nA>layjdidbudjcO1ynmiL?;ow-1GwrFxatq#r`HQV zy{Qr6Td${^%f_I*O0@wI1o;L3XwLgZd{R#Zyzh?cJ!YAL)58=u$ z;mWV!m-)Em@8E;ycks)5dwRI~bAqe=CvdfY@$`BB{GZp@>u(I#>s)pWB{v)`qw}z{}g=@SzxW;RQ-_F}J!8KmdGw1#DNUvYOwcj4V@8DiC*M7ByYri^%pT2*CYrh)cH`Dt!cQCVMJRUrE-Vdse;WzjEC0z9>{ExjpgKPa3@IUqXL%8aX z;9vYN|9gD+>G>Y;(d#?7>QCU>e}d=D`$vC29>dk1C0zL_TpW$1zhz9@P~MP0#|(we}va>;N_S6ehB=! z_wo4ySAGju|95cJkMNI7_WpK)tA6qPdH>uowf!N6>we%GelM@j;f>!vEa8RMZ|QzN zY6pL`*Eeu|zV6{q^!hzK@^d~<;Lq{;;05!3(EG6%e!kZ);jLe93jZ6g&){l*0q?#3 z5U%$#NAS;keGS*^?;idiUO&JG@1Nj>^L{w-`Vg-A1Ng;nW%DP2t35gV8GiqK1J~=< z5nS`LhU;7)d5eFs<009Vck zSI)tU=Iwv2`X8R(&3-lvJSO2th&+p)RJssh%d#LYMaMgz|o?ma_ z*OkJP``Gnn@c!=RTexyMxN>^9a>AGVkoLfplfjj^1B`;Xzddk>G@_wbc_ zSkK$9=Y}hI<@E>f#yx`la+}CxL&$ z>r;5;?Ky-GUSGnspVaVkYirLAuKj!u|JC2K`T?%?2QQmn@9%ki2v_|QKKXU!@Zd4l zpBs4Pehhzx|K1j^=g0&6gcW~8r@DF=^57+#h;Dgtnz?C0fFu&gX_SQcMyt=?Vg+~{f7jWh5;L2&>d1E58%PutUX)!Y5U>9+r9np*u94jp1+4r?u%E=+po{*FrNe{cBe`E>>FwtkD?nb)WA0y!HG6Ui@5lX8n%v$?L<{&im(Oo*%&jZ_f(8^ZEmL%BgQt3Qw6 zZ}R#Yu65eNKj8HpT=^sXQ(iy8_5GH`>*oFQ6|axsYEJ_Hw%4a{wPypr)a`737I3xa z2!2(sui?sX;Xmc|9bEY%eC72MeDL|Rc>TP8Zszqdy!ZM9ervB!;i})j*Ir-1RbRpH z?e#}+)gQwjw{+A5Bfe~41c!QFX8I{HGJpwIlS@yEa7kR`Yl}b$MDwc zTe#-W0Dr&Nk8st8Zc!l>OdgchrNz>@XY&vgx~2F*8dY+-)mjG zY2I)89%&3$e z>H9#q?lZ68dOy5|f5iJIhu3d-(wX&g1XufOxZ2;s)t(OC>Ny4c7XF+!c+0#WH2-7x zx4ivJxYpwuejWdwVGf@@ZP#1EwH~){wPy#{{?NeHo;_Uq!vNQO4vy#jp!pxdRlkI* zJuA4{pTV`Ctl`>E4&m4J{w(3i1@?EEHT(*nwfVe*E5C!E`@6QE^zi0`HeM6_2Hu_% zxcYzb)_H&4)azro>Jzy3hZL^$eh44lYS&xBwO(rY?Y%uaxavFjZO^jLzdc;9*Ax63 zAGP+Jz_mXt-Zt-_kNA3w;hH}Q{Mx^6`6>Jreq9+n`Pz5R>@y|&voEph+QQ%DITd{P zos%;;E&S)d>#s-fll$BGem%T=(znm#3~=R)@Poo~CV1dE!Q1Ekshkjg=s6L*+**5L zxN?^8W6xQ^1J6m|ce%j&CxxqjGI;q!YflbW&IYdjDd2(UlyK#2;p(3XUOvIvQ^S?B zgOA=n4LtCi7Jk-Ueg44JKRvuW$L9}RIU`*CGrYoT+KHldKTscel z;sG|FSMb1d61Z|wcZ+DDdCU$BinzraE(_5FCS+)HC#D6 zxW=o22cFZymD9mBUOl`#+uAe0l{3PRyniNm;5orN=kw<&&$0O&!qq4f3N2(;cCwc z9(Yay|98(x;mXP2<)P0X_?52j^9Qb+0v>oy2|w_hEnGPjynL9?ANbuoX9rhK0}ni> zg+IY_I=FIrc==GDKk(Oi&Inh|1P?qXXy^0i)1DK;l@r0shxq(~2ftwBwS+5Y1rIzY zfnV2iQn+$5c==$TKkz^DoDEz#1w8Pa68=cf*}|1m!OI8v{DEtI?ciFc4LtCi7OwTx z!Ijg)%Ln@Wfopw@aOF(!z;l9k&*zWUS4h|AFL?O?pFeP|uO(dVS-}I(N#I&vDO@=j zyexeFz_q?MaOD*6z;jBt*4GxUoC;pv-{%ip>uU#BP6H1-r-f^Mb#UeM@bZ2>f8bhQ zBV0KXJn)?0J@fga^%cUE6T!>-`uu@kF7f>bu64SC2cDC_m6O8No(x|8mCqlzayIZo zUl#>D@SGB^oGrZc_hS|PK0bfom-qL*JGlC%fd`(`!j;p()jvJF{7auda9!63*XP^` z-agrKg7?nnk8(n|KIca8;c1o=!yl^mKXCQW3Lf}%CGe-}{SRC@89ew4Yflb8&vQ0# zV>u;U*R_RfyefF@cJiw{=jvAV+&VK1y4MuhU#G2|h5NKl=P0!j%)zJtu~1e_q0s zvw{b{P7}EH=M=7-3|{}3^-m7h{=9)Jr+_D(Q^NImXA4(O1s|?u?Wy7Vyt9KVr-7HA z)57(6r-Lh}r+dx-*XNxPuAB+pUem5C_~3m0=<`kpS55@aJtv0i^Ue~koE1FqoCL1V zJ1JZ_8N6Qj{DJH9&IYcW0-ktI3D@VHEnGPje7J_sAGkj6?BL32;HBrZaDCqC;L7Rg zo-@Gpd1r(xXM(p^_xaP!=a1eegz&=m)d*f**>YmIa+dJMb5`*7ik6eW^*$kmt34Tf zxRT}MaOG^^YEJ==y*(vd?>Dw^wWoq7-kut+oE=>4Y2f9L*mbpVy>IE@YEKWZBg+}! z${FEm&jcT?U^&5u=JV(D{SRF2iQs{^Cx$C$30Heo@Wk7bz?GB2C!dEIyuOOhAGqFc zY~YIr`}=2j{-c&t!j-dyuRW)N*O#}P8m^oj{K#_}c^83V41g zYflMR&K9oms^IMEIf#9$x;B-}i$nXM}6KCi)NjzTZdY^G7)$T;mnN z6VHj^%2~oSUMqP0eQQqw*Xv*k*LY>{z;klAayD>{R{_tzXYDEB%Gts-UKPCkuI1El z=7 zf5Y07z?GB2ThGbhf#>9KlyKe0-omxMDtPj5mQ%x(vx94WHSl`0 zoEEN}4zBgp!voJ5;K~`{T3-`9|Ejeo=;!lCIU!u@D}uLQv78vLoF!cAYXy%zCxPpE zK?>J+W$^M}tvxwhIUBgftAG#xVmT#TIa|2KtAZzT}To))g04zBU) z;eqE2aOI3}jn@RvzijOZK02R2dTtfMHC_?Coh&DYD`yGUc&*^E=Ol3Dq;QQ_1~321 z+LObTvw>^83i|)GoD!~_EnMSO!4uD^;mX;;HC_$8{+hL?h3olX2iJJ@@W68hxN=6g z#%qG-|7q6!6${O1Re77OtEM zUVhHnQ^U2sc5vl1@WIzb3)lMU;L7RY<)?i9z_q?cxN;`=@JY)FJ~^L1T3;btIT1YZ zoEWb4wS+5Y1+PD0?MdKTUnyKU89eZu9Io}Xfh(tg=O4HBlyI%DEnKhf6}f&ZQ#l&=znK9C0y%k3s+7B4?L%aYklqD%4y*Fhpas< zT=`(Deb;aXoixN;i0=d^IGuMV!99$vo3+B3klzDBrm zCiw7f%LzU+pFdh(AzV2TJn@_uuJyHqD`y3-TWe1O*ZNA~%E{o#J1i%MYkh6t$|>OW z+bySrYkh6u%BkRi=hSemuN_=D4LpCFwWoz^eRXi<^zinrmNUS$zDBrmCc5VYpPkPi zt*;QS`y>%O@NtjfT3<`Ja#ryCU3OgwT>1tvxkd>uU#BP6Ho&U9@nmuMV!99$p^%{DEtIjd0~m@Zl|%6MSwyf3&_rxN;(R z;yE!~>uU*D&I(??+1iu9wZ2lgax!?}IXPVGYXetK0nguL?J42+@g(k3FY_E2o3k{(iNGmls-l2DoxYxL)5U`Wr1L`22kSC?|w#AB*6L=frU3 zEa4ij6}*0fwI_k=?@Cg*#w&vdo|D6svw>^83V7aFdrG)+ws4JC1#e$(IW=54JGjQH zfybWH!j;p(HC{cue4VvtfGcN&YrH1-@LJ0W{%JmcPXB%fuJMZCiRZ*{GY`z%^b4ygst_lyLn$+7_&;(Y!nCxmOfB6#^6YflVU&JwQiTET~BTTTMk-~FX4^?8Q{tp;To?A z9(ztO%;%53habWd|DII@FJEcxiQ&pw!VAw?!H1pYByfFiIfd7rlfe_u$>GY`zyz1%H4c({EJt)hEQ4Oh+%zVe&~o?l??Y2o_bX$Q|er-!%aTh0Ji&IqqO zXM)F`6MShtfAqb|5Z-!D1TW9C_QY`IEa8*qtl-1TEhm9%9;Wd45jI{KJb0PqvV7#>~Byo1MHzo&cs2v4qHIl({A$3^Enhwy*(bDkr(&UsGY znkOlIP=CU8&T|e|&ITSo%D)c`*E!E6Tsd3##&asT&Uvok%Gtqpp3}f}&T|XbyzSuX zpB}Dro(H&cM!5QCg6o{;;9utBuAC6A{)yl^=Q)NeX9*v?e^zjv^PIr-I+eoJKN(!- zJm+xbY~bpj0zwD{U+42j`(6mwctvoX^BlvKvxIBBR&bs3oWPZn!ZltQT<1LJaOG^^wf9c} z*E!E6T;Jo}!qq<&T<1L3aOLda>YoO#bDmqczNgy3)t(-%bDjsdaz?n?Gr@JvbMTe< z{L$YLgmASdg6o{;7_OWpTzwBSuKVXBT>U)3bDd0NixrFPv?G~<_3a)dWYq*}L4;5z5I zh3mO(2UkuH*E!DvT+eMsxN;`A&Up_0Z9aeW+%|+OCxYvo=NPW_wS;S(uHZW7Ie}|^ zrEukBaGmpX>5bT5H)jthf=RCJ?J$LKi%IV=c=XrpiJ`aK`XM*dT=ioo*^GDYe z!ZltIT<1K;a9!6DuACKI=R7BHT~`WMP6pRG&pBM@DsSM*Dd0Nixr8fc3*Y;7R1S55}kInOy<_q8@~zwBje)_o^uAB-U`#H}wT>JA5uABz0bDmqc_U8_+oF1-oo(H)0 z=Mk=)39fUVga4Y(AAQ~l;mV2NI_Eit>+{YMuACKI=R7BHecnmo%E{odpYxo<^?7Fl zS55)fInO0rpLe!! z30wG)@2eGD=RDVNEf*oaYv<_X!4 zY2Z5NxrOVxI{1NKR}a@Y&jVaJBfRvS39fUVgKy5~k8(nI=Q$Bv=RC)7Rbe&T|RZ=hH1*<5j_R&T|b{&JM2eYT!EOxrHmI zgKNBcxXyVV;K~`{8m|eibDo3$p3fiUgm8^l1lKvwFX=ipoO`J?wOAzb4X!FA4a3|Gz)uJKyIb^83i$8? zyB}V{m9vFwyehcPd9LBg*}*kl4P56uw{X1x=R7BHy$?^}d%gdG>zwBtuAB`#ew@F5hU=W?60V#reB(J4Jo%>WV>MhkJNV9X z8o17RZsE%5;G^gCaGmo!z?Cz?SC6;;nczC-IXIcmAKk|e;aXo2T<1K;aOEuFT3;); z&UsGY%1Pl`Um0BIJm+xbY~Wg71zhJmmvH55;aXo6T<1L3aOLdaT3-!Z=RCJ?Jum3s z8m}I%bDjsdaz?nuYl7>X=ivX%=Z|tixW+4j>zwBpuAC)Y%IjaLKLInOOzIUQW%)x&kp^8i=Q2-kQ`aGmoUd~ZH~PM?3l zHC_>1=RC*o)8}7s<*eX3=Q)9&KL3I%Cxh#p=Nx|e{EM#phj5+qT*6PEf5FwB3a)dW zYxwE&FSv3VxXyWQ;iu2P;L7RYI_G(SpFaPBD`$f1oaf;C^Z9f7{0pv}2(ELUWBBRw zFSv46aGmpuU?wK32hX&T|dd`r5&j)4+Ak za|_q{>fp-h;X3DefNOn?aOF&Jo%0<0U_O7ezCyTiBKqh2dxLPTuO(bLE4a>iPT*Q! zDO@=j{ZrPS9Io}Xfh(tg>zwBjuJyHrE2o0%oaY*@^|ga5r-AF7=N7K@)xnk1!*$N{ z0N45&;mVodI_Ej~pZWaJ`U>HCeUIQe=Q)OJeJ$b2S;2MAa{|};O5w`M;5z3yhiiRp z;L0iBI_J5BYkh6u%BkQw=edS!eeK}NY2dY=_uIm?zB;&adbrMc9^hJEBV0KXy!P{% zgNx?#N9!wuD<^{MoaY#>^|gd6X9d?e&k0=XD}^g3gX^5<9Io}Xfh(tg>zwBjuJyHr z>wRwp*E!EMT#Kt+r-$pD=K-$uHNurM!FA4a5Pa{uU#BP6O9D&n;Z*tAi`2hwGf@0j~8m!j&_@bHx-|vm!I_Eit zYke)@%2~l6mXsMT*9@!ws7TCaGmp9!?nJ4aOE^` zo%7tnwZ1yIa(cMVc^=?eUn5*O6I|y!2N$2uAFZztuAB(2bDm?k*4GlQoE2Q>JST9i zuN1DF46bvYbGX*m2Ckd}9{V}ZC0y%k3)g*;3a)dWYq-|e4z8RAu5+GSxYkz(S56Pt zInM)J>uZE7XM*dT=in0a`J?p}!j%)jbESx(d4Owujd0~maGmoU zTyj2t^!E-ST=z*LxXyWw;mTRU^?7gw*E!D#TsbLR?|U=2&Uwz^%Gtm}eSZu-`1#Bw zTsd2~_OS}CbDnFsa&~ZyR|D5M&n;YkSJJ^XUOimrJP&Z?jBt(D1lKvw!KLQ&M>!!} z;}yYm&T|Y`&JwQiTETVBa{^aR3fFjLaGmpc#*LYQMo%3A7 zm9v9uyc)R9d2Zp#>EIf#9zwD{()0PFoDi<@ir_luIfg4|3DEIf#9zwD{GV}SPzefw<8m|bhbDm?ka+Yw7*9xw4o)froQn2KYIfpA}0}r0y-#>)w zoaYj*?^SN$2cA>GbOJg0~2oaX_qoDn{F&IH#v&%x#9 z^GAO-5yI7;2(ELUW4LmbaJ6Rz*E!D#T=Qp22m_a}HO|2A+9N0oOUtC0y%b z3*UN91=l&xHC#D6cK{=$_xhlkUe_gY$g;!yC^De#EX%&wuysAzb+p zynUI^e|Yp1>xUTr`G<68_Nyga`73z&BA@^81J6(3zkOAkKPg=K8C>~m_{Q^d_>ubt zuKWV7{6l!_IVJq2x3K=%!j)galiK>{2tIj!4Zo%4KV10@T=~cF_*~!r;UB!6wZDTa zzlSS-56?V*fPd@~HohZV`4c>TiO+v{>G{DG=kxy$y?;Wu@*{Zp5TF0>Klhv%-uUlb z!j->*2bc8sC-BL068Pe&HZCdM_n!=2U)@8|IR&xD?ftQAF=!eeDs_czIeLz+mi0{a|O@;-trINYtKpG zTlW;M_Gj?;!cb=2OPuw?fg6}-1hVR{X zaOF4f{DYQ%3|~CM=RZ7k@8HVs;qeD7e-E!bXMnfvBV73te0ZPbpTNUsTKj`&KK~Eg zL%8xIc>P|>U%*SxiQ$d=60ZCeJb#bnAHYvMCxI`YW!II$m7l@mcU%4%UU*IpuiZCr z+{_XuKWfb-_7z{xIPDUaOL#yo{CFJzO~h{Pgu2uAB+3ocPM~`J;8QgezwSZ*OgX$C1GG z{x*dxCxf?tU^yjR_j$H(f##j|a{+R=R;Ht_s=meayB&*|W5PY;j3YdHhF z@|+Q_oC!XhSWa-2`FQo76T+1f!Rzl>P7Gf@$NFapSI!Eaf7@~rc+ zUCYVgt>N%(sD|8c;xd3uAB;9|E%TI@XT{|aOE`c;Cpg_t(=29z>QCTzx)&KI82P;c8C=AFgOQG5pbY_4~bW<*eZKEZ1qEoXoqd(H?~&IB(nVL8Dy=JO}|YrC!xuAB&-T-ZG0UmpdLFcct33_8zLw>*a6R|w z;L7RY^))SLfb03i2v^Po4;GdaEavk^&nZH^V7H_g6P?;CGNu6YFye$(=g;JUwB!*zdk2Uq(Wc<>vR ze+<|C)fTS%s~ueVJv_LXQcr_g9y2<*(quO)UQauKTMAT=!Q~xbic2@N1U8hU@-n4%hwF4P5yJJUFoY zL%8m*mT=u)-NKb$!GjxH{t;aFS8KTLukPT=Z{Wd=EdLm;`>QQn_g6c(@_TsjtCqiq z>;7tRt@-?HZ(`3+L%8xIcz!8+eiy^PvmlmxN-`3@-@FN0soe_XA4(O1&=4cp8(hQxH`CUdid~*mNUR#@bwR#xz8}dl{3MI zU$&g!$L8a%`{5y6IT1YnRm+Ltx*xuTD`y37Z)iCQT;Bsp;mXP2$&D>1hwJ+u8@O@` zczq+wDdGBF#}=-f3Lbb)3s-wOxN>@Un^=1W_@y`YxoLzeXM)F`6a4sm-0$VzhY8`z ziQxI|tvxaPBCU70a#rx+R+f{%pLb*behyqY89ca+<>c_+^6T2bl~cg;%yLTjkKfJO zvxO_Cg15J`oEonAyn`#JfhTvgoECnsTlx2A;L7RY!|z(o0RI)gt`V-B37+4{a)N8m z=a1G!2v<%7PyWDiV)%pnx|VR|tl;e}Ehm9r{gu`~DO@=jyuG>Q?*G8c-g1JUn2&qnIU!u_iQx6e zEGLHRbJG&8oE5zNgykggyLtD zso_VSvx6(Afw!NtoEEOn%N<-fJ-q&s?HN3sO z>EZbwSv>KKS561d6U!OkSN5C{uAB*;{HEoEKkfhezS{bHj^N5!z~e*v`4xQhoCCOW5_tSD z%UQ$Ud2#!^lf#v>f!7<$DdFdP&K9nm3SR%Y1&Ix?-oZ$c1f3NkZoDg3BrPamo&_ACp;mTRT>%Xv^6#h7W9n9d$S;OOdSxy0e zmggM8l~cmwds@yB{N+D79Q31`qjae)5G&WwVV;2dd>t_&Ivrfo8?45 zWB+P=OV3%rl@r7B+w_nc$6a$0!uP|MlFYtI?r${FGLLo6rwS^HPxHF!=4S55>kA8bFrgvT$l{#n75 za{#X&WH}kU@SHVVIXS$2pyeFWJ*R{#XA2)5U^z8BdbwTK4z8RA9(Ya%Kk%F$uADtQ zF04HhJolUvxN?G@v!B!WCil1hynt^#Cx$C$3D56mISKsOb5gi+GI)7k%h|w3&ne)_ zIfU1LWjPi6dV#%u9l@1T!-qeyoCe&a){c^t#f_WTyE{0=^>Eq@PJ{s6zk#jO1! zT=^4xxTEC+KZM`Y^CP(O7x3W@mb-*2e+6HhXY1vF?w{)uc<^k?U&9m6 z$>D0}1|B@ea!UBdbGC5hRPf-rma~Ibp3}gUa|~}Q%jx0Ap0kH5XMnfQu$&Y4;5or{ z?Z4N!D<_1v&$PN29-eRGwS+5Y1+SlOIVpVQIT>6zYk2uI%PHV%&pCuEr-bKEwVWe( z={YrAIXifAuI04w#&bHja(a0D6w4XmgXc_e<($BSCtFT*J^NSV8(m=IwSX%ph7Vi& z`2+ZY=Ol3Dr118MmXpJC&)LA0Q^4ydSk4x{^_&W>oFjO7j^#A)#&eG0%4y;G<1J?o z-+RsgSI!7e9%nhh&)dHmuMb?__VW;~oCqGDZ9l(+zshq~aOE7pgU4D<27j98tl`SZ z;lo*$a|plI73{i7xN^4e_A!=I!~e(Ivx6(Af!B|=oDTkF&*|aH*~806SmU9g6Jg0>#r-SFOx10e!dCmw|&IC_hXF1_7_`iOCX}Pm;kKoE#z~k52 z&#&OQ=N!P5lfZ-5Sk4;0^ZV^NTsa$fd}KK#{6fFau!Sq9g2&IdoE`j~-kt`ooMU+W z0?X;)N8UesxN-*g@G8qWfw!I${G$E$T93*J;q5D}E{2buvxF;W1+RCOlft7{+Wg7j z%2~t9S6EH~Kk%GGxN=H(eu3p2(LJYzD`y8!&bOQv{*`ZaXI4lDS56O)&$FBn{vOYn z;L16H2QRmr=$GtYjqi&+X8~7E3?E)Re*Ib5S56DhUt~FZc;h((Tsb2=`D@DwR`##PYwtNBTsaXu zexd#R65jbfwt_3?0G_|wax!@EDqE*(xN>rM-dfHf{F@)J`BTD`vxVm$vYZEX)R!}AYV&IG@@=bXTm6a2FMoW?i*p#A3seChoh!EQW$ zEN6g!!PmtISIz`4-)TAF4g6n!|MPv%iQvjvz>{~_&#&Ou%I$N(0bDr=Jbt_7tl^c9 zdk$C51|Gc4a!UAtw`U7iP6Z#1EoTQm_VzSz%Xy_8h+r{wSz0C zf#;vJoR;o69b7p*y#0jb4DiZxM!0e&ckKW;xC!$;3q!j-dv z4xSDv$hE2n_BpS7G4{xI$HaOG6+*mG)l=KZ{bE2n|S zpSJe2@WOLCxN>@U{3**B;Fadj|3}xIN6W93bs8@m(@qqM8pWX$)Nnz&xFQ5)C>eu@ zmLQl$&{9y-h*}^@RD|*}1ukd-0s)6E6)*}yaRH4?hCvCV1tLZS4ImgNDACr)*sITe zzk97b>mL4L@mZTUJ5N%nbI$J!aLqZx7k};M%x>ZSTlY_+`v+xu17Cz>n6sf@{tZKAXEa4g3an z-CMZkoZ$TzxH&!ivu(}**PJuF`FuBL7EiC&msp;wx0y9IcxalbKIN_{6;n>gKN$fzOvsNzk`3ieLh&gHD?b$e1+T30Uo^2J;y4z z<{aU>8@M?Q{A90-7Opub_~y&qoF4x7HfMlq&KbV?GB;;-tLgnS+MEEcIdk~p3O8o~ zUs|6Ct~pEi?8a`+3ZB}W1g<%2c>fJ<&IZ1-IT>7Yw(#cb-JBi#8+8A`HD?cRzShk- zz$@K9aLqZwo3C+m8u;~p$bH`3!ZqgvZx(J&5C3|bGr%?H3~#>L&6$1c^!~Y}^$Fmb zGlw@{>EPy_5 z*|$yapQ~+70N0#3eD%d{&H{e0IT2iQmhjc}-JBKtBl;4iiFs^FS)gs*Pm<}~m(+ng4zIVX5;b9(rDY|a4JoHP7zQ@5Ylt*6&r zANK}u&6&gXac@HZ7We&cDSYkA5-1mh0o^an2?t8))>IwHf z;l3x__q^RDJ>kA5{QPZpf5LrFxbF#HuifjbfN#u8`0UPZeFfj&#d!_)J?ZKRU*6ST z@9@LjoR4tz5A5f$^!fYg)^2?W@699l*3LJEAIvlO&idzY^)KM+U&7VDf~$WGSN|5S z{vBNX2Y7!odw#;zKd_&ZQ&05|;p!j3)jx)-e*#zk6t4aSd~W?qxcXOc^{?UT-@?_u zgRB1lSN{>N{=x01ulL6Khj8_e;OZa4)jxr&e+pOs46gn;T>UHhx4QeMhO2)ISN{&K z{sUb7N4WY2cbN86{}8VJ5&U>txBnQf{s~R-Uszl5uQ3tw9Q4zB(K zT>VG5`UiKM_Ei56uKp2R{bRWLC-A-XPvPpH!PP&9tA7Dk{}QhL6(`!PP&8tA7Gl{}itN8GLK~bGZ5!aP=?Y>R-XtzlN)S3s?URuKvND zr#;u!KZL7)1XuqUuKo#J{ZqL5XK?k;;p$(&*Vey;tA7Po{~E6TEnNLOxcU!p^&jEt zAKhizzmM(t30MCFuKp=p{WG}w=Wz8e;Obw()xUx-t$$6ok3Uag!qq=W zrajd^gsXo7-&+3^uKpQZ{d2hb7jX42;p$(()xUm{|>JEe}Jq12v`5$uG60C zAHvl?f~$WFSN{w?v;H~VzTQ&6)xU(Re+5_n8m|5=T>U$^`Va7fKK_TRfAAgCp6VaM z)jxu(e+*au1g`!mT>T69{+90fQ^IxsS8(;O;p*SQ)xU$Q{{UD25w8Bh-KITv`uHEN z{t;aLW4QV!aP?2&>Yu^YKZmP-1z%YI8m|5=T>U$^`VVmRAK~gBeCM>M`iF4!kKh~Y zAH&r@fvbNCSN{yI{yALz3%L51aP@ED$6L7jzk{p)09XGJuKvN@r#;m_gsXo9SN|BU z{t0|#{ZqL5XK?k;;p$(&)xU(Re+5_n8m|5Wd}aMdxcUd*HSMYXAzb|}@*&Uphbo=+F3f|kl`_yptZ{h0S!PS3&tN#dB|KNM3J=H&itA7Ib|2_p* z{|v7FIb8hL0_^KZEPv zvvRol7jX42;p$(()xUYu^Y zKZmP-MYqp$YqOa8Me}t=laIa}k^$+RRKZ0+pe+*au1g`!mT>Ue+`sZ-< zFW~B5!qvZpFRgzESN{R7{v%xdgL_YVs(%Pq{|K)BG2Qwn@RjvX;p(5k)jx-;e*stj z60ZIgT>Wdf`Va7JVqgD;9K(;u8%(xxIX?&;rjS9gKzHY zj+?`MPx$N(ZhZ;w?dy0I{H(99!cXP{e5amp-!r|WC*1dh`<`&$6YhJ$eNXsOJ>kA5 z-1mh0o^an2?t8-Lwx1Ec)&2RurajMgzlCt$6YhJ$eNVXW3HLqWE89;F_dVghC*1dh z`<`&$6TVPSxbJzNOM1e6Pq^<1_dVghC*1dh&(#y|d%}HBxbF$~J>kA5e0i07o{aFh zdGP(yo{LAh^&$N91Mc&_2;Q5=@UwXW-~XVSlfrxRjBdYgFo*BH-#$Nu@9%A&pTgC@ zg3s*p+!{VNAK+W-Gs1n(`%ZiAtaAwWJ>kA5-1mh0o^an2epFAm?+N!k;l3x__k{bN z@J2o1zUTd>Jx}UM*Zm3iJ>kA5-1mh0p75P|(sh5reNVXW3HLqW>e<5&k9M!$0p4uf z>*5UW%}4m+hur$v4@~Q`vxI9uF&+r`VxNlDc7fhAI)ocW8T7dTQ{eJpUemN-u5}dn;&&^f*+h-ulWO=hj86*5nT6M z4A=dZ!Pn{u_dVe!^@RJLaNiT|d%}HBxbF$ys3+X_++5NV?t8+0Pq^<1U#TbD_k=GW z=kD7Q?t8+0Pq^<1_dVghCw!@%aNqNXF6jyPJ>kA5Ts=4N^^;tm4BkJ{`4+yv+IbE? zJi+-6ezf@oeDe_Zx;wyEHm8E`9_-d1;Ro{?zPIb%z~^?{7OwsD@Z;m%c@1#w=M2|= zM!5Df`(a#P+fM-3ej@ntX>LC;eEoFi3A}l#^AvtG&)^619KN^X7VwpM3D^1xzI=+? zPYu`o+`{*Jx4wh#e%bi|KbVj3y?J#1Y5#?K!hKJ;?+N!k;l3waeG0hm3D^1xK2uM) z?+N!k;l3x__k{bN5196xt0&y|g!`Uw-xKb8!qumM`<`&Eui&RA+x-dmJ>kA5-1mh0 zo^bV?|H!oG{<&^HA$;~Q_dHy{H_vwKm+-xL4A-0$TyxfN?I(qw9_se9foskduKnb2 z&Dp^E=z%^$H*M4HS=B(hFvxaL<3SZlPHgL_^!Zjy{_x66egKJI&-(6#`clhB& z&Rh82yn`QY&H$hPk()Ep|Im5xW7F$+wD}=?{zA7tg7+_Q9>cfKbDqGP%6Uq+pKs0J z2V0-RkLDG8ubyz<6YhJ$eNVXWNmtM8lAdti6Mj@rxbF$~J?ZKR_dVghC;XtEblsnD z-xKb8!hKJ=dj9WAdcu8A_)b0Hz9-!Gq^l>~_k{bN@Ok6z{~Ese6Xz}b^eX2ad}VV6 zcyB(!=QbyJ;IwD+N;f}*YkdSi*!mc*k53Z#@)d4=3g5rXc?Lh4=kSAh1z)Hq-1mh0 zo^an2?t8-3C;0J8dcw6ng74H5?t8+0Pq^<1_dVghCw!@%aNiT|d%}HBxbF#9pWs24 z^n`1D1mCD9-1mh0o^an2?t8+0Px$OD?mn;KyEi*;;al?#-oM_hAK;Do2tU5jtq*=; z+H-Cm!uK{Of~#8$*Y_tR@ZIa&aZ|Xy4Re;g{xZ!SGNJK zZX6#u5J-r-D0@9C2)1i;G4I({yALT3b?wJaCNKT>Q=+mt%a*w2Vd%b zhO65MSGVA&rk&L-gsWQwSGNRyKG}T>SGNqVZaG}t3b?wJaCNKT>Q=*N)~$uBTL)LS z0j_Q%T-}20w6nTJ@SSyw>Gt|f;Odsb)h&aoTMk#Z0XyLOErqLF23NNnez0x@T-{2zx>ay>tKsU_!qshn zAO753#}Te>!9%8<)h&dpTLf3P7_M#!T-{Rm*1BbIb<5%ER>0M*gsWQxSGN|vv~C?- z-3GY2jc|1fetOzj-9os!MR0YC;VbKwz|}2sG?mt%9pt4Oh1ou5KM% z-3GY2jqv_Y-SrKAX4+ZZLb$p`aCM8}>XyLOEu-7#3prff3b?wJaCNKT>Q=+mt%a*w z2R~Z30j_Q%T-}17opx5Y5Uy?!T-_4**1Dx|b<5!Dmc!MpfU8>xSGNkTZZ&*o-CDT1 zb#QeX;OaKQ)h&3~w6nTJ@aDaC z=fAQ06RvJ8TtDB@!PS3&tMdrg>pl4SX;1xpM+n!?cO>xs@7?}WxcX;s_0QqzU%=JB zgsXoASN|HW{w@4u{X4k&4{-G#;p!j!!nCLQhj8_e;OZa4)jxx8t$z+z{{pW5C0zY0 zxcb*{^>5+o-@(;?fFG^@2v`5$;nSY#AHvl?f~$WFSN{aA{wZAj3;4|XmvHs3;Obw) z)xU+Se+O6p0j~ZdT>XPzoc3(&`5D61KZ2`&3|IdIuKp=p{WG}w=Wz9};QRO6^AoQA zEnNLOxcU!p^&jEtA3S2(Q~g7@`bY512i^W-xcVn>^-tmIpTX5XhpT@9SN{^O{w;iA z{X4k&4{-G#;p!hea@teYu<@)<1=-e+F0o9IpNaT>VS9`d4uEui@%H zz>n{9_s5+o z-@(;C_@!yjmGuwd>L06tA7qx{{pW5C0zY0cyI5&HC+8$xcYZ+^&jBs zKf={Nc+|9~`iF4!Pv9HtpVIB~tPHOHIb8hxcZN9 z^$#9B?Wz7DT>T@s`p0nf&*10Y-TygU_kRIb{}QhL6VqH`WNt>^)KP-U%}PChO2)ISN{&K{sUb7N4WY2kDc~h zTmKNQ{t;aLW4QV!aP?2&>Yu^YKZmP-1)p308m|5=T>U$^`VVmRAK~gBJZ{=k{X@9= zNARWfkLmXNbrZPyr*QSp;Od{l)xUtNe+gIrmi~9{`P0GGe}Jq12v`50nD$iv5U&0a zT>WFX`X}(Y^-tlt|1-Gy=Wz8e;Obw()xUzPe+^gv0lu{UBV7H1$4`5ze+XCq2(JDy zT>TTc`loR9&*0~G+4~n<{R_DImvHs3;Obw))xU+Se+O6p;0e>7&ELBHhj8_e;OZa4 z)jxr&e+u8)?>ETc`u##V-G0AN0Y6#)60ZIgT>Wdf`nPcP@8Ie`z}0_*tABL$wEx=r z$8hye;Od{k)jxx)e-2mw0fgfEzk{p)09XGJuKvLjr#;m_gsXo7 z@89F@{}itN8C?ByxcV1x^)KP-U%}PCrd$6OzPJ7zT>S^Q`j2q+51usbss15c{Uf;g z$8hz};9KjT!_~iltA7bs{|c`DHC+8$xcYZ=>p#G^)_;VnfAGuGp6VaM)jxu(e+*au z1g`!mT>T69%=(vb^{?RSU&GbEg{yxDSN{R7{v+M`2fs4yxwifxT>T@s`p0nfPvGjG z!qq>6tA7qx{|dghuQSwe^>5+o-@(;?fUExqSN~u??Wz7D-TFuHrS*^L`uHEN{wZAj zGr0QaaP=?X>R-auzlAUV!9Mz;YH;gExPG5f@Z@Pv{r;a2uHXL?!S(xpVz_?) zPeRw%|Ka+5N*P?g|0jp5e*xF;yC~sv`}nPeAFNLY_dVe|>pa4J&!=3{6YhJ$eNVXW z2|ubQ-1mh0o^an2?t8+0Pxx6q;l3xlS5LU_`P55#!hKJ;?+N!k;XC!D+wZ5!;Jzo^ z_k{bNaP@59tABK_-xjXl2X%t$_d#`V{XVE3uHOeW!1eo}f?q}de{%Z?;Vbh9e*9;* zK8EY}?IdvhzMT}V-?x*&_4{^mxPISG0iWAGOSpdDP6gNRgR0@WzAaqWw}b2Y22Y#z zJgFz#_k^EqeGKeNVXW3HLqWz9-!GeEPI!qn>cz6Ta5{ z3HLqWz9-!Gg!`Uw-xGdRPq^<1_dVghC*1dhtLN+))1LExaj)M1uHSbwhwJwph499H zFVh09-@g>W_4|&N@T2V~hWp=J0^iyCHC%I2_$IgCTLRadEnIVQxPDK_4z4+SxaO4b z#`bf7Yt9j_IW>IobMCwvxaOSTn$yAcdpUZz=A7Z0Gr|w{`)X$8^nTNv2tNCFdmn)x z&dw9~+B}7CKjzkF@TGYU*Wc+b;QBk=C0t*3t>6pWPYw64Q~2zoZl4|eU_QW`e|76e z_{lta=Ct$vBW`^R_dVghC*1dh`<`&$6YhJ$eNXuL!*2gI-1mh0o^an2?t8+0&u2}0 zF4Pn5d%}HBxbF$~J>kA5-1mh0p72IJ>GpdzTDb2C_dVghC*1e^wQ0}&f7|^D_dVgh zCtZIh9j>o4W#p^)KM+U&7VDf~$WGSN|5iv;G}i{Rg=Ek8t%5eq-8G{X@9= zM{xCz;p(5kkE6T)bGZ5!aP=?g*1v+Qe+^gv7OwssT>S_5#`=$N^$&h?+Ee{QxcWzM z^^f7|pTN~Wg{yx7Us?YWuKpEV{cF1QZ{h0S!PS3&tN#dB|KR^jd#*q3?w=5@{t;aL zW4QV!aP?2&>Yu^YKZmP-1wZ|ly}!fNzlE!RN4Ne1T>VG5`Ui(;PxTMs>L0&$TVPvPpH!PP&9tA7Dk{}QhLEqr7BJGlA}aP=SQ)<5{IX;1YJ;p!j3)jx)-e*&Ld z{}itKKZC1(4p;vIuKp!l{VTZo*KqY8;OGBz_x}i2|KPW$J=H&?TmJ~I{xMwr6S(@P zaP`mNXMH{eSN{U8{v}-fE4cdCaP@ED>fgcDKlq(#&ztDjvzk;iO4OjmbuKpcd{Rg=Ek8t&mo;~e9xBf9){S&zQr*!L|!PP&9 ztA7Dk{}QhL75r%ZYqOa8Me}t=l@SJH+^$+3dpV0rWd;X+!`#MGjSO1)D z{R_DImvHs3;Obw))xU+W^zlDj{Rg=Ek8t%5o;&TS{vllbBe?pL0JH16=(_xcUdxwCBwFhj8_e;OZa4)jxr&e+pOs46gn;T>UHf z+WOaU^>5+o-_foA09XGJuKvOAO?#?;2v`3I-has5KQUbW6S(@PaP`mN>Yu~azksWM z30MCX-rLvPJGlA}aP=SQ)<5|DX;1YJ;p!j3)jx)-e*$l;e+pOs46gn;T>T5U`j>F^ zui)xm!_|L)Z>|3bSO4I7)1K-d(yf04SN|BU{s~U$^`UlUS_SC;Og>dzc;OZaKt$zYn{}jH}_nE`>eQr73zR#_I@2r0bSN{sG z{xw|vTe$jnaP=SH>OaEOKl+1dfB)}OaP?2%>YviBe+F0o9IpNaT>VSB^{?O?>tDmw zzlE!R2Uq_AuKpuj{eu@wd#Zm3SN{aAf6q$c>Yu^YKc`#&0UG!`qy;p-@;ec zzk{p)09XGJuKvLbr#;m_gsXo9SN|BU{u%uEZ|>uV9IpNaT>VSB^{?RSU&GbEg{yx@ zxBdfsVf{zA`Ul5pPxTMs>L0EoqD+DoZ*@?!u5StvlmaVm*&jjniInHca#=z%~`@VCx-jKe+1W@HC%I2 zxW3P71J|4_Tyt``|NBR9&8gt4PrG{d|Nm3No9j4l;Y;%lzW!yM@TGYR*ZPkB4Q|c=Uz(5bt$FZ=)1Gtl5U%wR{QULqxG~)KgrC38txw@6^9;VX z4mrHH^%eal))RidvGW#wFz?`Nz0bn^`z+kQ-v4OY)4$%~{`C&u*?wZUf1ic>_gT35 zXK?>M3-|A{aQ{9F*ZXV>-&+3;uKojD{YSX^2lce4`iF4!kKpPb!__~5Z>)a`SN{yI z{yALz3%L51aP_a?>R-dve}K=d{|HzA;E$(0)jx!*e*{VqH`e*Q!_0Qqz zU%=JBgsXoASN|HW{w-YnJGlA>FPrwtDjvzk;iO4OjmbuKpcd{Rg=Ek8t&mUOw%=u>LVz{S&zQr*QSp;Od{l)xUtN ze@VCg72W#RaP@ED>fgcDe}Jq12v`5$71N&TAHvl?fzQ6t-rwQr-SR~PY1YuUUG!1fAGp_PyM`72-nAlEBM-aCUE_{ z-z5CoB!#07I6JM+8(Z-M=Rlt^*_M% z^Jo=ZKaX~VpKU)i+<%<}-q`vRTyr|OzAn+jHRlZ1oDr_COU(XcdVMu#4%eIzuCGfh z;F`09YfcQ;*CkeP&B@^NTe$0F z@bk^woDr_qMbJ#Ii{8INxZaN<_;Kmx$MA!BM*lXqK8GL83%K5|OSs;zE4sa3*YJ64 zJ>k812VdUGtsmfrTRI=%zUQl_J-0R|gm26v`nS6IF??;F!4K*Q_dVghC*1d>t0(+q z9a^~W3HLqWJN1P7p0By2C*1d>t0&y|g!`Uw-xIFSr%Jd!pQ_;ce5!`8ZfEx=-1mh0 zp78Z;?EZxNp0B;6C*1d>t0&y|g!`Uw-xKb8!hKJ;?@3orxbF$~J>d)Wg!`VayQC-F z_oS;Q-1mh0o^an2?qBb4|9YqE^$u@tYxgJI_k{bN@bj(h{)GFUe|kwzxbI0m9yQPq^<1_dVe|dv1(y-}CjC^o0AKboGS$o^an2?t8+0Pq^<1_dV(A3HLqWz9;;s zo^ao@y`(4H_oS;Q-1mh0o^an2?t8+0Pq^<1Z|r@fh5Mdx-xGe&>mBZUzTuLdaNm=z zo^an2?t8+0Pq^<1_dVghCw!-#aNiT|d&0Ns3HLqUcu7yV?@3orxO%SP`#ZV!#}vN# zcIO-T>CVo#aLvi#y}cfHaLw7nHK&Ac?&6MnfNRbXt~oXQn7BC&TyswFnH{%-ukPsP z^l;5N!!>7wpKtHx%-%G;?wT`)YfcE)^;*C+X9?Gw7_RHJf@{tit~n`uW8F4z&Dp{= zCx`2L?ckbI!S#Ho;d(x_@Y7vg{|>%4AK-djkMRB7-JIaf)9a|`ObFlF{0Oe+Kn&m7 z=kEzz&w&)Kum5LoeLa5%SN{U8>$r#e{_y^uu0st!Jj8iRznArg_xEr-|akuFTU4#0Y6x`l76*YU%_YU z3~#JY3qM(h4!*VX8sK|dKf={5cvVezv(d- z-``ch7q-5H&&(_M{#V@m8a^{`;al?#u5JU})(3w+?YVubJ8lTy{NOV$&MSh?ZG8+s zKgG>S;7juqezFc3d~WM=xX!D9H&1rQE$QkE-`M&VzWY@-r-L8udub zUYLjQ#^yxuxvh`k>X5)sHYbCxp6<>ohi`uFIqtrN_fNB)@W#A?ugzZbDd}K)ia&vbbJ06 z@W%F2!FP6EHC*S_!cWh3$L-)7JFfxW|GHa0!jHe}Jb3%G^TN(6gzI`m@V%W^0$wc3vI4f4RF}1AJ+7MtEcA75wG2=Sj~2y1lkA5-1mh0o^an2?t8+0Pq^>-*VCR$^@RJLaNiT|d%}HB zxbF#Hswdp{g!`Uw-xKb8!hKJ;?+N!k-!<)ddV$@abUi=ez9-!Gg!`Uw-xJ>0^RR^b zp76ar$6ENt&Z~p#J{jOTuMw{E3jSu=P3INDbzTvC_7dw1*LfvyomU1w|ACv6!*yN- zT<2B7bzT)*=T*aXUM;-0>(#+^UISd`6})@eb7$uj!gXE|T;~%4Nf&Z~gyyh^yvtAd~HIZ(rOUM*bbHNcm4UL#!R75weAr_L*c>%1bk&MSuN zyb}1y&MSrMyfV1XtAIDpbJwec>%1zs&Z~y&yjr- zZ_QKq{wv+rNj7lpGlOfNTe$j^@cAp3d4y}9HC+2_;GeUy=L3BAVtYQowZ4bX z-v8K(=f(ivm#qAbf@_~kxb_*t zwa*p&7rxQm&k4N$6MG)Q)pG}5H|}*;!dH*|-HZGD0KbPFw}LNT?T&kd>$p8!#~tB1 z?(758o;SAR2Jp>m?0E>+aaV91H-+oC8~DtQo52sSb;sSpb=(77$F1QyZUg_To7(dV z-oMTr_XO8*XCIt)({V$%j=O-rz>XWi7k}oCyM*hw8@P^}!*$#p{IPc20={{@JMJE? z<2G;|w}b1rJ^TmlxC8vqy5pYVI_}~huDx>%*Kt?y*p8dP`!~4buHicF4zBYp;X3XC zemy&G1z)_$9rp;=aeKIqJHmC`*)Z+%PxkR!0N=dX9d{1baaV91H-+oC8~AJOxEcKL z7I)k&T*p1Yb=(@R<2G$tOjp7zo6Ie_c=JcsMJF!wdj=O;C`5eLZd|twJ+!U_k=5QT% z2iNntfb03ZhwHdCT*vL;I&Kfw^Lc>l`Fw`!xZ%G}d+N9`T*qC(@29UPz*pD#_{GP) zYq;JQcJQ4&pG)}qy3cYSKf`t03a;ZG;X2HQae_%wIEbGXiT z1=snea2-y&K<)6Bb z>k7E`xrb|?C0zSFz<=Ut?(07ly!n{xb%d*D4`14I<_zzx^9XOg_8AxV&+Ol(*G2mY z;M&g|u5K}WX2)H_wVxEO{cPY@Tb~TR_=MZ{7Op-A_`>=e;o46P*M1tf_S3?(pA%et zM!5RSKQisF{e*DsX953m`}%VP-+j`}UBcC81D{);EnNG_;o8p*uKg5n?Pm{HpBk<{ zC%E?0!L^?r{zcYjfcKwrea>+8S^WF-y3DN460ZHkaP4OW*M1VX_OphoPYzd~JzV=K z;o8pu{_uOauU}R0*~8rX(Gjlo4P5U>C%E2^2DqM|XSkl9BV5nV*?&y0i}n-1wf{L> z@23m6_P>N{{|Q|CU&FQk6t4Ym;M#u%*Z#L~-Jb{e$)2|rT>Cu2wa*%^eKv6IvxRG) zC%Eq05q^G!dp`~S@ANuqpL4kO8N#*C1^kisbML1S{O}X*eqO@0eg*Ge{m6^=#{|B3 zjq^2pX1;~*?7HOe{p;PF9bEe<;M)HluKW1_*ZwQG_J4$H|0lTi-@DwWb>HUj-tOByT>B~E+Rp)g-{;!Ld+^2m?Rg8= z`V;*81b6?O;is3o*VhO?|AAXS`>$zd%@5$(&m6A%Z3)+Y)^P17g=;?>xb~C5wVy3q z_s;>owfpA?*M4fa_S3+%pBAqDoZz~DM)=9@pZPfLulF+x_H=u_mvHT;hTr*C?&G}%-rM(y_3$^m$E_dWv&a78 z#rtFU(P_7PzRj&)z-Nzh>l3*C9e)kizvHL$^Rq9GyScFKqn|{s3EFz~{F90DrKpuiz_N-@-TJ zFJJ7lgX{CD0Y3YbTR+14PdX1iK7G9$%tLr%9?}2J&57aLPdHEDtB*TR;j@oA&*(Nk zhp+zItuNp+TVKMDvtPOBR>7P9aC2(-*5-r|}o$V)u>-rY({=eLPQo?n8EBM-8M>YIp*S&?G%{%zf ze1Plvj&NPyz`ib}o(sFaAzartqFbLBuIrn?b$wI#$*yk(*Y(Zey1oT`W7oH&+v~A{ z>-yGkUEcwIu-8JhztgI}u5Sp}^-bvC;@(G6xc+@D zgU`Ost%R9bD@J`*(ZYpEJ9@Azartg6sOma9!VoZr3-3>-uJJUEdt8>s!EeeM|WIO82^` z;JUswT-UdS>-u(ZUEcw&>pRlz`Uamly`L9$eM7jeR|MDfjp16K!Ot&s9dh{Aynyds z>DHIr?pL_F2GpZ*}WS_`!1d60ZI&eD^zh z_w^b0(tLo=&4d4eZssBUWFEoy&vwU6;HS4aPvOgx^9OB8)GhpI`|05NdpZOBcz5^vPDi+Y-)!(X)6V+)J0bjhYx}*aaQ!{F7_Pqu zm%-Q8KZmP-0ayPLuKpEV{cE`Tw{Z3E;OalXch-M|tAFsh)1K-d!qq>5tA7kv{{*i7 zDO~*v_{r|`60ZIgT>Wdf`nPcP@8J4-76V*=&tjz8-?IolZ`yNh{X@9=M{xCz;p(5j z)jx%+e+F0o9IpNqd}IA`O}{2AJVOV1YcVJn0{CHd`RHx zpTgBYgR6fISN{U8{v}-fTll$i&+`tx{wwDLd}cnv)j#-xX;1YJ;d=f=a6NxwxSl@= z{Pa$@&lIlvKZC1(4p;vIuKp!l{VTZo*KqY8;9KiI!qq>x?zE@+hj8_e;OZa4)jxr& ze+pOs41TcwIb8h?RmEQIfSc!1XuqUuKo#J{ZqJ} zKN(!lpPX*bp8~$M{v}-fE4cdCaP@ED>fgcDe}Jq12v`5;dei

mS3_KY^=%3RnLO zuKqb({R_DImvrl2!PnNmhO2)ISN{&K{sUb7N4WY2UpVcl{vllb6MApYPx#(EgX_7G z!}Z)K;CgP9@bi1!aVz-z@0{20{%@VP@U_k9;Jzn(_Z~NAgm2A*FPiq$-;W64`uheNVXW3HLp}_>!J*-xKb8!hKJ;?+N!k;l3w)^?rN3!+lS< z?+M?jC*1dh`<`&$^Gh!23HLqWz9-!GgzJ4LhwFW(pxgUS2|vBhdcu8AxbF#Ht0&y| zg!`Uw-}6f^=?V8e;l3x__k{bNaNm=zp74!&!hKJ;?+IV3C*1dh`<`&$GrXiH-1mg* z??9z+{T-+b?t8+0Pr7=-Pyb~1CtQCgu7>OHNwx5U&FSF2CtQCYYJ}_WLj_+p?YXhP z6Bokucj6NG{KM|NQuzK~ooDcsc@Edds|8#iuaUG!`qyyvZ{h0S!PS4D zTmKQR{=p5VJ=H&itA7Mn{}`_R30(bCxcV3HrS&i2>R-XtzlN)S3s?URuIJJK*K=v4 z+jA-SifPaOU)=o@!qq>5tA7kv{{*i7DO~+CxccXC^{?QK-9I&4{ad*DcX0I|;OalZ z)j#;kX;1YJ>DE7jH`YIftA7Gl{}itN8C?ByxcV1x^)KP--@?zIaQ9CK-+#>c0AHDp z@R@n=Rnwl$NA3QE>-iJG_56w9dj2Hvz0FVIz9)SBpKd=n{Pge63;5Z*g!kqxe5amp z-xKb8!hO%LzN9DI_k{bNaNiTYQBSz<3HLqWz9-!Gg!`WGxq8BVPq^<1_dOSv^n~mA z5W)3)i0SrxNZ?ELg!`Uw-xKb8!hKJ;?+HJA+}-CL-1mh0o^apuYcA;t_dVghCtW@1 z>IwHf;l3x__k{bNaNiTYR8P3?3HLqWzUS9o(i84`!hKJ=dcsft?e2#Z?t8+0Pq^<1 z_dVghC%m!ukq+*A!hKJ;?|H*Zdcu8AxbI02(*S|;PaP=?X`uM7Z zpY8i$TKK{Gba3Aj?t8+0&#$|rC*1dh`<`&$6TVSTxbF$~J>kA5-1mh0p78bmu=^9f zFdyLOz^xzQNAuw8r#-jkA>8+b`<`&$6TY_jDO`Or`0BIW{&V=oUZ(|oXI{d$<}Lhi zed`Hde3A13e)>Y|3D@US!8c5M>hq}(?t8+0Pq^<1KiYm$xW7N)`g|&ftA7Fa_a|JR zPgQVzK2^i@`P2YkTK^HQ{y{YDss15c{Uf;g$8hye;Od{k)jxypt$z+z{{pW5C0zY0 zxcb*{^>5+o-@(;Cxbd`SWA}LoSN{mE{xMwr6S(@PaDDuc!S(S&PEXwHu7Gc?e+gIr z3aV?P`gd^kAK>ag!qq>zV%mRg{bRWLCvf#o;p(5k)jx-;e*stjl5YJg_}co{ zaP@ED>fgcDe}Jq12v`5$CexnkAHvl?f$zWE-B&5RxyfTLzE3-YFKvAepPLu(!06%_(y-&ec4{(2PD7fjg+uGKL@TGYKKi|O3kKw&_NZ==1 zpTc!s8GL8!bNES};WJxb!Ovgi_F2RCH>@tMZ%eoJ9eis(z?-kM&hWE!3%+sM`C#iq zxH?4et*uYt%ZIq%Q?Y^XAMJi`$qxR6-Rm#Tw}2m>;nvr1t#9C3-@~<^0j~Xo-!$#e zzRG^@3S8?~@b;Ehs-@)~JdwRIm&z9)(OgBG( z>wbvgPrjR5zk;u?OAAj1--^15`;pUg{(<`2O@%lQzcecKQ&tC1; zAK|^tso}d$?DsO2@QtlMz}MyjT>Cu3kG6h3_RFL0j0*DrLwg&%BA4&NNz`W?J~^_07Q|Ej5_Fm3Q`0?(} zkMQ*io!9WqE1Wm*op}pCzRs=h;Y;fG^j~zohHuSN_|m+9zw#;e{sq^0mGHw8+;I$8JvegW6~!?$1i{0e@u*L4He z{1&eH{q3)P{s7;~wsO?Z>U)P2!H*z;)afuH&BII&OcLYd`J)-`vF=ca}`o z>$m}2$DPA<+{Im|bNuy%_jcV^a2+>+>$q#Uj=TAeYd>xVUwo%sU$~B2z;)a`T*p1! zZF=0(ce(i${QNz3ec?K83)gY`@4WW;1AM=>>v;F+dd&~un!ot2>3Z!yf*-!y&0oPa zKY?rh=DV+beg<#u?&j~{nqR;*|FE9UZyx04RPgnWJ8$5c-@-Nj^gYx0>%E)P!8cbs zKf|?tglql$9@joUglqj0e$F3t@$py;?;q-X4L{t_{k`lIKK~l$Te#NeaIN3NwZ4RF z{SmJ9HC*dYaINp)T7QOX{Rr3k`8}uo8|xOr`v<$8OSrCY4A=Eq!?iwzYyB3k^*LPY zPxqScU-v@?*ZpvYHy?9f&mZZd^Wfgo`Nwy;zxi`XHUoKiK*;T(|}ql?(Xx zIqvIrW5t=;`_hOf-y`%KSw{=qk2 zJeOARg{{xwPyb`L{~dg8>ksgI|F&CS!B5{?TpYK9-{02vaNSq4@1O2JxAg&h_gjy@ zm>wtfXypAG!!wmySveF6VH zTfc`dbiVMH+xiARv-LgvYFj_RwV&XA(++R3^>g@Kece^2mdVl`wKl>eP%x}?Q=`}dk_I!|1K86ue`6@{}QhK zr0`GK{x|T|o!sl9gn#FA-TVVw&z}ZvOo~@8Ft$fZy5X zSMc8c9!d+px2-?H^*lMlf5g_0@QwaH%MVYx<+gqS*Xt{VKf%^-;M&g){_D2BfNyO- z75s&^{s>o}4*n56-{9K+?Ecg9)$=fb>;8%0dLAy}dLAb5&Cj^k#SXrExbp&jwDm3g z{0O%`c))c3%U^Xqhp&Ihc?v)Lo_l?5;JwY?!JljE3%J(Ta6La8xSpRq{P%7C06*D& z!XKHQ@5^lc0#m^n5q>DlXoa7x347#C@Mw1mEd? zfZx~lvw`d5!ySBU>kGKv-z)h1H{JbthA+(H|2;jgN80=qd}ix&_`liu9lZYyciaQ~ zGoSD7w+eo?^&R}KSF{(Md-%!L&mK74|7&x%K7ea|4A=Y>+~1#Y&ClRkU&5dBKzF_e zxYjrDC;yqde_FV{FK&SAIWYh6X@~xgii>U`{A`}VX9st^j_|qp2ygB4`q_h~$6dbL z&57ZE`g(U>EBLIi?>B>g>mRyq8C=id0)F}}ciaZvo44@oZ@cy3PfYh;ysCAtYq<8g zf@_}%T>H%6{qMNr9^hy53ch=`TR*_{dSCqHbpQQx-1-Q62?#~x~rL7;}E89=_;OY7PjrCc;7q&iu&&~Jn+4J1>E#dp;J8$8KS2zzIGTl$} zR_Am0@iop<_{;6%$yyEnO>d-|)LcW`y?;p#kl=yd;IX!{J{>K4IoVC$D~tzW~x z#nz{AUB?`*>$rn!{sDfn`4wF2JNRGQ`X0Wu{^8F|`>5vvuAVEn?%M>e^UdIa?Pm+u z=i?>(!?yka*ZKzj30vR74=;NB#rxU--&u#~XQv&S*Sq^|2|qN>Gq~ndaLqZwHD`e9 z>&)|qP4}~VyF2a*-rM_W4*%4P-1B4y*Yml8pKX2*U%bN|cYx3S!g-WW_o+D>xaMSV z&8gs;)5A4qfNM_lbC>J~t~nW8b1JyLzHy>kpALRDKf{maBfK$>etx=t|Nai2{iW-l z!nfvox_$gs!!@UYYtBfwIg4MI?o)FjxaMr=Hm87V&K|BgE#2mv;hHnTHD~$o>HgI@ zg=@|Rt~n*$_S3*Mr-f@y_KVZ~=ziP7H}7=Uy@c<~kMNaw4PToNaDDtdf5db@tA8xq z=a2Bc&B@@JvxjTW0N48YBd`6q8C=i*BmAKAg3tfTo>#l+{MlbSU&0r*K8OF&HTGPE zFYUS?;Ek=X;9B3pUvKMAaIHVX-(l-VxYmciG(F$j)NcO^`1akd!wP=;?|17HxV|2h z!EgQt?Zthvh3n&i68;1FdI((iLj%8~?Z1U@Z2tpXU;jVD*S3Cs)wGZHAHsFMG5qas zbmzN*YyTVg-E99EeDz%S{T>Be-~X_O>-!CkaQ*vi4OjmTexEzL>)XR;_VH5qsA-@3 z=>CE0>pUy?gKT{Q*XuWfKg`x|;S0N7CA_fp2e{@p@F&~)7QV9iBm7yme)i~T2hCr= zpJVGIxcVe;o$ng1^UdM9pLcNW=K$BgXH{_BpDkSX&k3&kXN2qfC1#JA_R+_&3%I^t zB7*DV(gd#WmsrE~dfdYG{SrA``!C`8x#I&|Uq@}=uWa4xrG=~K0DqIMKf|?t{@7`^ zzqIusTq~=)!mK{-e$d_~u^MT&&;2 z`!_f*;q&i*$;J8ud|_U}m*z+K;hpY$>r1?W_h0JPxA5IRyM3PEo7sym_SwOAU+uhy zFYf4kfG_Xs&i4#on~#_H>~Yik=UBS^2bcI9zI~s&z9D?`r|vo~=%4L6MDYIp&X@2_ z=RAfl?7UX+m3eZBXY`ZX=N7*EVt3peK6{w+1AOr}ZcYV1+|aE*!WU0)Uc+bq>AZnA zpYN`F3*RiApWyum+Dxx;5~Ht#9Ev-U+VbcJP(W>ET*Gz&EyjglpfkCrqD{+rPB?7Je|F!++I0 zgsZ~>-q?O3xL&_8Tyt0O>%Z6SGlAdKd=1y@H@(C+@U^|(Gk9XYh3kBC`0Vf8eX@h= zeD`q8E#ddE^E$vEVqU>@zDM{yY<&&a`8IH!ZwuG?p5QAx-wv+(e1L228UBj*x$8T^ z-(fzp=e=I1h3;Fp_Bn^|Y@Z=~_D1)5h~R=-5!GG91f$Ms$;UDT> zb@7gp!gXCX@XF?AaCO+i8|#q6^}boaHCOZ0|B4T|4kgy#-uwX9`Bss8m{-# z2L3M}bo*)HPqh7<;E%BN9bC`V9zIyN0j@4*_^fx&ff26G!4s#?Va=VxclP`V;idTk zuFlaVzJ#lD3|HqBJpC8fIf1`!aK47W&-R(Z)p-LyShpQ~_3%$#yenr%o<{#l3TVKN?Ti?J_Ti?PLw*CZP+xiY3+WH=z*!lrJxAkZE!q$)Q z(fZ7uG`&A(zv#LJ@WSTL;hoJ7;agk3fDg7lg70ko627qg#PG`2ui&ArPv8ezzlP6k zeF`sa{RSS``V79e^;>vv=bOWGTfc|vIa9(Ld%Yassa@9!t~p0|Z*yw6<}`54Y2mZ? zy64^rt~niCb9(r~<_vJn8R6a~G?)YfP4`&`fMXA4*79KN#7JGeR*aLw7nH#Vn)Yfc5%{*LgM-PrBF zhJW4_&Re+lcY@#6)_3r|c~7_Z=>fj8^E$)z@$Cp#pV_ZW@58;#3EnEnNFM!L`p0K6}6GKftxm5w3j(PoCaC+UFd;wD}=i`&+=Z z&j`M;IWb)OOyJsQ3fDe2beo^SwZ9y$`{4vv|L7^xK04nWT=P44X5D-E%kJc!djtIK z=4bdrZ|ByJ@axCUXHT8(^LuaaJb-_#t)IiMFc0C2JGeOu_?5PP3D&Yq-ugg@4|+yY(Bm&NqYWe7EpxY)%f>`4(`!e)jMy?7T|&?adE#>s-Na zZJm$s?vCz!Yq-w0f$Mx*_=9ZD39j?);d&ho@T+v+!k=Y6!f$(9xBuC%PVZZtZva1A zpE+FT8^U$I3;2KB#vM0;>wK4RJ;!2rW6zlc9=^n#PYPc>!+8eR{4IR;_wISVqrb$> z*~2ye0N4BqzOwl>eE$+RzlCdl2iN=_zO(sfc>5AJfA%!@-}-u^=Fj1pAHo}(AHlaT zb#r65<|lB?U&Ch~aQ!#%;-zl>7Owd_xaJq|mCY~VvzNO06wYyJVgv-wAO@>;k52Cn%hxaN28#^w+3jm;n7njhG6NY_jA=kVD-xc&=xXML7% z&DVcxzCK?_xZdWk;o46M-`M&LUcc1!%;DNc0oVLJ{9yAB@V)gs!Zp8z>;2;d|M)%J z`%VYH;-1cX`0|ObxcIu(0Kc{Q8UFkyx%DIb9p$Nib- z-Y(zw`NR7>uif8|=W`zO+?jdp>-FB(Ug}#Hexf{q|LIigPY)hw-adRA^=AOzO`gJE zBOk&~lxOg5H17!Bkmv9nLxmk=Ni0c?0h5mmz$QeQh5$;YY|@@QdTGY`h-Z@XO^9{OK>XdYs2NnXKE(Yy=z+43d)e0gy0`u@D@`2&8d`cQ{ITG!78 z{3)s*!k;T|!f%nc;0LKcZTKnj2!4sY1K(TqWB9A&UHBpL1n!;(dhogKzkRs7Uk~65 z<%e+hKGF#8)-i^=_mL*>mFiF7Q$63#;BMU|{QOhw^)-iA@(O<2X_jBWw|bxX68>oU z3jP#%@Zt6S+25*R-!FqVPPO`V_~equY~&m8K=~%z^|uAzP5CzbaCro;zu(r=fuA9d z;X!8kF8sOj1b*V_mhZvOmiOUT%LnkE$W!=6na92NqC#oOAUHt^^>i6NU{s5k;{t)i!kKk?}jNz`&6L_xrQ@E>Nz+Igg zJot^BuM+O+%;D~MD|n(hOSr2Oe02T1yM5e%yZsu%FF(^>UrqQ;@)msgVavDS$DeB+ z!EZamJche{--UM{^q(8o+XQ}vya)fj>i6L{$_Ma6&bIZW@J;d|d>8o$9_W1J@KmS1(a!>OK{9LVb3cpZZz^~JKX7EepCH!3Z9G=Q6_$8Wm0l!kdgkLQW&RgGyZMR?H zo79Ioe4Co}rvX1e^+Wh!@+SN#AF}h?f^RQx!?%z}@P+!*fgh!LWB3X3F8p+P5ALpa zefXAV+4=|YP4X0e%b%{?c)bkaANjL+20vP!!`=094BuAQfeCyk`4oQjk8J$~eDq`U z67H_2bNF7WU%?NOFW|dhWc8Qu#b?ZekFD>|yRN_R6I8zrKUdy>|4!#OgdcE?&D(_U za;J4un0Me?{f~JJclVbrywLrn2Y1)!0o=XsH-x+T89cbv?t?kp z-H#@4SHFO}`ZIW>`g6FuJ}=;|{tE8u2l@IwNmaiNch|QN?&`PTu6`R{sD20TUXNY4 ztKWmW`h9rtTkC%ccdxq)?&{}oSAPtTRDTM0*S8tm)t|#%{R*C{{u184(9Tov@%8iN z>eu0}egj^peiI(5ejD!Uci^sm3=e*1{ZHVr>i6NUehPQ>hww=CNARKQkKwNV6z=L5 z@Kp6n_)7IFxU0W}yZS45q58G+*Uw}9Q#NkEUHvB9)o;Or-&_A9c%k*ja92NpyZSwN zr1}H6yT1(KuKoz_>gVuO^(XMA`cuGN{W;v-4=VVNes1@R1-y2n`4aA)w^wlYyd8XE z{XDwo?Hb%YZ`a}Oc{_x=`%M#m*iE+17Ce);;fMUx@)7*BJIyu~q{-+;U4|0dks|61@J zeqsBt4Ugp!{FW`Q*m(Z$z;Bbs@PoFtd>4MmR^|zOle`bF>pDJw@BT|$e+oZHK7`-$ zGs|c2(Lc;b@U8D|>m0+~{dfZJt8Y{IMe+iEvHCWHUoJ1<$Eu%m_@!IfIxG0O@&!DV zFX3O;yes(i^5BzqZk&JjeeIvg>u~ox(STpAK7{Zsb^UC@k5K&<{CIgAezy7(!MC`l z^``@0s1GsxI(ZjLJkax(+qZ3b4c?H~ z;VbpI0pDwP+gI-S-PL#R(>9S${@0!lTX6UOs{1;Fbt2?b)#<=p9rwKN>bU2y4tMn@a96*8yZSSDrTTNYtG|G|_vcq|cfJBW=eoWHf3Wja zgS-0f>jsy1bsESgsuRLp9rt|g>bT==Az!FY1b1~}xO;!T2Y36m55G?Ly8(P9PvHk| zWBY#ypXfTA!QK5dhqrY;#_)MV?_YfK`aQ8e=hr8#22HZUlgmCvf(1g3^ffn37547R#c_4zj z=YbgRt`}YSe(FO4-~IvCpB~&j5A@;gd0+r{&jTskJr4}w?s*`CyXS!s+&vHEaQ8eg zhP&s1DcoJ}3V35ndwtE|iM)iLcALEqKZkGs7xN14-k)E>-Su$=f3&Ux!Kc^v?X%=H z_||*b`s?tyJcQTvI%&cWQ2iGCD0v$mZ)fvH@H-!39>bfOw+la8^%MBT@*aE(c^`iM z({0`X{HSM`r|?$Wd+#yJ=`4aA4CoA}abXX99P1rtn;K3V84HR;Q$EJcpmLzxA_%UoT(4 z-9B8x-9B8w-98LHYwL37%k9G&-0j0U-0i~#-0i~cl)pl zcl#mTjE~@@#yLFEcnlwEJb|w?p26MzcJsLYpLT$q_c`(x$(L|f$IavFd_?)+bL;2# zYIy_h_Ib1!@4_>U6L_d`58lzZ51(p0guB`%>rVvl$UE>IbZS&*ZSbw9AU>*hwm;A;ck6R_>RiA;IEOpb-DGpeI6k{)^Wvf z_xHQ-#Zk8Y1pchIo4fmoo7cVX-ADdX&6~pA{VRjJ``u(SKHrQ7dLDP{arK*<@d%!2 z+=16Lj^QnhyYQ*jm%uA|kFIeaKGk>#cYSy3b;mn=hn<%Z@)P+4?&`RCT%D2f1@fhQ z4tI4b`1#5&;6IWt;cow|;9t`IseR$jweMfM*Ifhd>V)t@by{%ux{KgFjXUs6;{@)G z$NfFm&#RBI^U_CtAs@nB9XF4w^CRU)$k&dw`Z;{2@fg0;xPZGpy1(b6PSEmP$RHqMjbp~)(Cxs`fGlaW38Qi_@M(|X1#&B0> z3U}W}p21!H5}vDm1$XtAa97`bevI2Eu6|9Qli_@-`VF|N--NsGQ@7!+egv;nKZd*d z3EX|ooBRA8w@!B+2h88y_EidZ_1)+6uug`2qBu!nc%<;BBpE33tyKEBL;8-39vom3w~J#@#pJ z&4<|avkqU%L%4e`X~GZJye;_SHE$b!MxpBr{3%~GkKyh)rVIawyUxH5*StOWt5kmg zch5H|{8H5)!cSBE4BpY}aRk5h-qz56S6OCJNcRcRzx&6HD!>kVx@-2B6?&`RC zT%8@1?;#({`|y#*19+iv26ugQf6vX^-qF^dBfp<~3U_tfJg!bh`5E#@$xFC<{-4uz z9xJ%}JoY8r)m_1_XxVWE|FOOwzALZ6-RHE|;clG`c&>GZaQE{AO}MMmf>)~3hP(H9 zBDky5frt07*HH|2`>hLibrN`@Iz701A8!D6{Yv4%Zr1-Hd?$GZcR!akg75J(?N9i{ zPc|RJZ+(jS1b(r8&SVPTQFRLVVS8Ae8TA%?fyNzpP2(8e(zpxHHSWWkTDP0WT?gLt3_GqA`3vMDxU1vladpmCevJH8 z@&fMm`C>C3d}Y0#Z8vs*Zajqs!AyU@4qwU}@I>PlJW_w|nn(3de5Um$LjHVt7w+n~ zd0cn2cc^~W;8)4(aQE}T4Y>Px;1FI@ohJMN@)kUlx8d&Rfg|{hy8m|IE!BzPZFv{| z8hHZ0N#27WEO$Q-;`-_K{{Z>=R(gLP?w;R<@Rsr!+&w>!;I2*%PgG|Nck7XdM|zB&A`7wb6+e!P4E-&sGGvV^;zhgrb`{hUnjU+ecZ_wz6{_|sLt4tGBf z6T)4;n(*_q{ucZac^ke-^G0xYTphSOt{C1`oi2P2&6~j8arNP;`Z$13)z1`O%7<|G z^;ZUWuZt19)V|8$ZXb@}!IYw4 z--oV#4esjH;g#w%;I2*xcXgWZ@Ltxp7Tnco!(E*So~TX-?&`#FSEmb4RVRVFI_~#c zxa+DrkA38G)k)#5P6l`17tZ0X{uo}U{uJ)&&)}~99Pa8@@VV+Q;jVu0we|huzVBRz zyZQ}yu(h4{CfwC;!`=6t-S6da{deavMm|#gF5K03zkh>uddR1$(}%k{1GrmH3NKV= z2zPbd@2_y{aeW&hAKcseHio-8Q@H!Sa|w6XwK;sJeRUs%?<-%xzyDs#FX7eDd6fAe!^$X+whCEo(?=zKV$gi>Sq^znEIK(pZgA*w+C;_`|#DVmLI@( zd#8B{U#8|m_#fVFp21h|F(1LVIL>?wclX5!d z_r)dr>c`mYbOq0v=E2w3_ow@P6E(Q|eG_%K`+XA)xchw*A>94Gi6-3rzKIsx{l1Ad z-2J|Z2=0F0L=1QL^)CFY>O%q_>h~}A;O_TL^x^LJO$^}f_f4d5_xmP>aQFKrGPwJF z6C=3$eG@s{{l1AY-2J|ZDcn7m6!2TL-)8V#UT&|85`Lw=zdVQU{a1aC0o;8*dI_)V zzSa50`u^<5WBB#%`U5}3U4P&=-eBwR!7r-J2XObilEQD%yhHd!nm2=Qd!*GL!N2%6 z^D%s&X_n0i~%ecpQ3pu@SWsSxa)HP zPqwk=?HSy?FHyo>ojE*LoeJ*G*Anj5w}KB|W5*kObA8{QFR#JV{Vd;ryLE@~)0J<+ z&yu&`{a0K4HavZwc?7>nbvkg@w-_GY$NJxew}*E8J-DmehadEQTh9P~v^<6HCm+Jy zIx~2BKU?PrK0npgIflEs6L`C_bxz^2ynsi)zI@}lG24un@L2gdyd$sRZhZ^*SjW4B zyZj3N8Rdhk*Y~aKXANE`Ux&MXhHzK62|rDJYr!v&x8bgz(Pq2@|DEc@aM#Z+{3zuU zxa&g?9&Bs-s!!K=0C#m#cuRGLaQAa}8Qj$w!DH3Q;qG~90(a}1!Z+P)`@Dc(B%i?# zUs%3`pCF&ZUA}^6$}iyO%9rr-@50?aN#KR{Ne}Msn|*lwTI=HgUa3wBck39!-TP)4 zJh;ETj&gYKRGW7Scl8VSS9h|{mz%+VBroB2$Sb%z{slaJoYh~#FOsj|u6|AZbMGIz zzPP?R-%@oN$ZszX;qH3a+>E#2_x+aLkJ|9vRX>8?`YX$K;HUl4Jcb{yd>4M#K7@xl z?>)Ht{@?)a>Zb6YJ>JgC5dH^w2ESH5g1hgdtDjZC||)(khi|Qe*LT;@aB!@nGQVo+8;N(fEOozV8d7R zk6Ax!*RR))7yXTV6CP^b1YSSp?;H6(JbdxV8$N`m|9Z!UkKn->f7|dJ9v^G4cteVD+*qpkiF z9{l>Sjd?40evqv*`0o1ghRGot`5HXA(Q*xVev-`_!jt{&xVrG@fwukuy!xHhAHy^4 zpDDaJ@Pv(Zmhj*>n|B4T2M^!KH@>%iysM)v--ZYG{L@B0f|m!{>!=H_zG|Brtn1jb_oyNalw;s+j(q$fBm>3)oH;Kw|;o4{gA-RS01|2hatRD{S4la7x1~f zgcpCabyx6~UUy4)_`YIeJuCR~9}nB`;0Nn{2=qFN;Q9V`9%Fdm`U8)je8fim5j;r_ z-tY;$uJcvGqeq>&k+03y*BPB}*VPc7->0#WZ^D!J*z0ZpU#QPHyd$5&i|y?7Six&r zfAB-BQ`d(&yrq6N;Hh3;Av|;E1zz6E`qP0gU#8a`Jkq=ge4_nO!UvkS_M`RVN-nkQ zd>0wrK56>|KGeKpc-OsN;Gyz!cwMiH3SRumuJcQHUv*aS{5;*q zf4ts@^xk$ItHUR|+5T_9EA9U#e4_Ib!85JD1CR6^lEKT@*m=+4^#{GhUU%^HblnHx zh3XXWMEMF{YX3BTvfiIk>uOWgQ zuJF-kY^*bcM|xgr!c#pzbm3v$UcWtftn1zY9^TLL6L@}&yl%mfmc|i0);NLpHBRA~#yNa-x~*djfAtS9-`Jl8ywtdYFEtMI-0j|HsB0X; zTN+33SY!A6?&|dADe{@dIeemV0WUSK;7g5zU*NoG9Ku@~NAT>8w{P^V1222#F}(KQ z=3V$e-`7as?(6Xso@t!JCmI*;G) zywtdYFEtK+iS=t7!dn_g@aFsN?RAuG%nz!#ua?2aqw%bU*izo(l~<08Yl3+#wk40IEPO(F5soc6@002 zu)z8?4&g10BY3QF0`F^_!ZVF?_(bCZUTR#yml_8*WBnS3@Rr6AJk~gY_cc!8nZ`MM zqHzH)HLl=Gje}dTevLzTOXCPW)%7-k_cc!8nZ`MMqHzH)HLl=Gjf3A{{TfH`SmOlV z*Eoe|8t3qd#s$38xPmV=4sOLdH4fn|jU#xhaRTpaoWe7WbNEE#0$yrd!Iv5bzs34B z4&g10BY3QF0`F^_!ZVF?_(bCZUTR#yml_9pAJIMM)HM#_EsYa+U*i;>X`I6+8W-?V z;|l)p?Y_G4JhOoBAYZ~CEnmTRk_W$EKVLh`Yw%sUU?sWs(b){zdVJXDIdZ=B+uX+IN-h;nI z-iNbik{AT$A{ulWYe!F}H|EoOs-}U|fH+c>I4|yGaw})!~!|yH+;rEa?;rEoc;P;ZZ z;rEtD@cYO+@PCoV@cYTT@cYXX_ygrV_=Dwr_;&IE{Gsv`{xJCv{;%>3{s{R9{z!Qa zf0TR-e~f$rf2@28f1JF4KU+S7?U)*Z^AzyZ^6%$x8Wa@NAM5HJMeSlG5o{wF8rhN1pYC35B_m^ zAN~pX0RBmN3jdUR2>-M^gMUUof`3k)!~b1AhJRi@fqy|hgpze}cRVf3`elD<8wRewX!a0)Mo83V*7+fbS`v!S|P!@VCh4@RQ^f{2ciL{_pZ7 z{Oj@+{KxX(_VxXLtGouk+X>d6I{d-%2D~W`;ZK)0;d{zk@cra%cuyX|-y!e7Pm{;+ zkIB36&&d<`74jaul=tB`$p`S;mj`jWjWO)s~NnVE^C~v@z zkcaTo+=l0gS>?ADxbsmlvnTrT;NOtf;Xjr);0t*O|C784zsGy6Z!P%!*omW`kB+!&oSKfa{^DEZ|AF^f7jl}E;r*9d~lhrhw!+i{RS_dW*!9g?;Bn?5yxhsW2@f7?-h#ItW8Q{Wo6IA4>rv(%xa)sRf28FT z_!+uB^x>|b19;xCIw{=sa|n0+%;2t{Be?5l4tM<=!(BfoaM#Z%-1W17yME5#uAe1c z{jA{cRX>+-*UuF^dA{{CxZ7s-C*1Y34tM=*z+FE>xa(&V?)uq+yMDIeuAdRy^|J$a z{fy~uf5P9b*LxrC`Z<8#wLjr^?N7MtX9jou9Kl^bbGYm081DKxfxCWA;jW(r-1Tz? zcl|8sZhyj`_a$9F;lY`<4_EN~gW8{4tnbgh@-=v&`+gnn?q3afsQYXPZ^@hR*xj$; zsk{w$_pb<^E8l^;e#Y=h`2@bF`q_s^=ULwd@ahxhDLnhA`4HavfO!UY{T#uAvn-#( zCz<&e?)o`_yM9jLuAc?m^>YSy{Vd_(hiyF-{Lr6jf5QJ`Y98EueIFL`I{dF+uzU!= z?$!2vx)%K8zgs?n@9m!V;7|Lq|*2@=fLyymheo0$#q)d%%LpGlhTghu_%fTLzDBv*R7XCoA(DURQn$cm15em&#A!eR%>L+s&JB*UuLGuKfw` z|Jmw9aM#Zc-1RetyMA`zuAd3q^|J?e{p`bCKT~+)7VS@X^&IQx2p;Tgp2N#Wnvdb( zBg`jo*Uu?DdAQ{Z_)_P01`nTW`4S#9%;)g%@#YmgtD7(2uAfVI|2~!vwp!nx`>3CF zc($kI8}Q`|%tLrAZ^G}|pKy0SYQrPV8^Q0|pYXf(Cp^{ltqZ?vf5LO+dvMp!KD=_* zPx#;5{)A6nYJD5Qt9{LLcxxZ?F+6>d`2_CzIfW-LwtNAP_BNlvT|Y~>>*pL^?qzi< zxa;Qv?)tf;zr^YU_gvqf+x|w^Px$uo5FTA*b(-*GV%~y>@(3P1(LP6|1Fv>AkKw7j z3lHT9ym*4u>A@3uA0EgD@ci*sCxsU+^C3KZl6eNtb-W{ZpgK7`Rh=!t+C-7W8-HaFTO!*l+m6x0GIlQlYwHaT)6Xlni@fAE)KDhV#{&!x3N6ObX z<1P4Pm%}%%R~@)}j_tye?d|jN68gWtaidNT9{kek_u<9Q%m?u759TSnulx``Q9grL zf3i9wc<^WQ9Gzfc2q(r=jI%@X4dhOL$SUI&*kq^*3+VY%e)Ow+Sa!S zUVof<2VPun>xto^_Cpswxz6TI;F0n@c=11$@56&{n-AfA)yd$wd;~A#1w1*&j%x;w zK5AaV)7*Rx&(1fm;Q82m0Z%V5U(zo$U%{7Jf9*c&`>mQ;f9mj9-hfZ8wtNUr^!jSj zue0^E;MK!ypS0nvCum>6Cx5Yd6L|eL^B%k<@55vH81AmOQ+WPRTTcN`biJL?TXwu9 z{cmsDxIWL}%Zn{v!J{2*-UYn=2=gU;xtGp6JUiYzxbOOTFO{#s`^wkh?t0sRyX$QT zch}n{++A;5aCg0J!`<~Zg1hT&3{RhK{prH1`&%CpxVzr=;O=_chu3$pIs>@7-lp*U z_m&^R-Ssx3>v}tayX$QZch}o7++A-c@JiRiDcoIe3wWUGzzpuLwpBVC7w@I=?)3?Aw_Jc7IHa1M9Z;exL7HG|hbV8>g+OLu*v?K zPuqa!Tif+Bg!lhw^_%eYF_v$^%U#VQcyhbd>A*|XkKx5%EZ>Fa@&vxT)ABudDeuFV zT2Bg}C_jW3@(kYktIa!tPgdqRJXU@J5C3NQDLj|Y;KALk&m}zlyVa@S(Lc-=@I)Tm zZ+)LcT4x=es{ajmUv)xwsy;N~h4xz;o~lj+pU6A#N*=@Of%T^gZ^;vQBJaUd`2fDW zo7GR@!4~F2cwL^sL-`2alIQS9K8DBg2|STc;eB}luWxDVpTVPhn3wRhW*u%M zv3w2QlGou&)oH*d@(>>GW%ZlzYESbPJeRlOi9CYWUu1PU@KW=}@J!x?r>c{{>o2kT zJ$Sjdc^{t12k=;)!t3%OynM0M&)}JS1dru8ye=QZ%S~2)0?*`Acq}jAb@>cls-G1+ zjcwiqJd!WrfqVrobY6l7tncSk9>QBMvw55FOs*t5buQUGq9TlZWv1RhDnUBY6uR$lLHj>xtm0Jb?$V zv3Yy&;??GTcq$*jBY6rBNMf8yai7WvV0q!$Rl|6I?H$9u{?&?o4Kup*HUvp2;hC zEMLN-H&~q&Jdg(uUf*v;&+>J6>oD^Me5pDiJXgL2kN?f;wBdDm1TPM^dWtxq=AFP(c>$lO&J5m{SMczyHtzx+ z$d~Zq2+ObFsXS<`@1ICsg9q|Dym*V%Z@^P|2#@4Vcpz`Vi@w!w!&7+#kK_rw{x+-A zgO^8|_u-j*0B^}t_)`5J!YA?^9vx-#j^Uwv0fNW zUL9@gOyK2l<~?|Hta%?ED?fm@W|^6d;*W; zQ+OaR;Khkne+EzGB|MT>@c3k_vw+v-OL#f7{0g4QgNLr~!&qK}*X4D1`ChBvfM@a$ z9?K(meu~xUz!P~459M8Wr9LF^T;7Mrr`o&&cwL^tgVQZPgct8K&)}(i1drr7Jdls! z#c5W50#D^rcqA|2fqVupbi5^eqT`*z`|=7NXSU7-Jb%CW5}wF|?XeG4rv|UfLwI_o z)o;Qhc?%xM+wkHHs}sRfc?TZJV|XC%!ix`B{RE!Md+-)AZufbdLI((`AG~g3?2oFAL>uJJ^kC?aMsk{x3d&u`Sc7M8jb&=%<$fp{+KV81~jOB;O zXBxXdT|T+k@;UOZ&KRCP;Es(WnZQR!U$f!SzpkI(<|Aht-hsC@j^W|1_WR|#@JOD( zTk;;fFYm*{XIT9KeDW0Y6dpd;dH|{;n(kA=dl*9pWmHzena^4A?H)7(S7A;h{W%FEwuuULIoe_TiCy0B^~2 zc=CC>PLAOdjoqI$cb!*$ihQB5`_tufiga2|r)? z_M_L=`JeI#K2slt@NcPp245;ahhMMy6}<5i>vQul>+3mF^;__ttIh!acXuY3lt{ml9^hyP0X3f@${`S|tqZ1Gat4=s4C`~dzj9d8OBC_jUL zS@{y4E8lp+`g*>jd_74ehUB1-rE20f2dAv=k@h$q1SI6-gvRq>B65R zPvCpV$MDz6C-5WXEBJO=Pw>R`b$&qk4*W^V$MDZ8KZ1Wlp2HhA*nV5Uf2{lx-d4Wd zT3_d*)c*+nd(|1jU#@%xugT}|N60JqZt~`n*4Oh=c?*89d;tFsc?y4*dY*JztWC@UHsVgI}Y3A3jii3co@50zOi{w(I(OZd1MvFO=`X@1y;ez!%Do;g3>& z0)c1)f%oMxysi2p_9S}^C|1=*+Jfd&sBc_-&Oe(9{k+S*9^Xg@+G{feB-I>>p4LA z5FRVvgAbJN)0LmXGvy0-u6(V%zMir2b@)vAF8u4tC-9~6WB5(VPvDIkwg2IP?$^Q7 z*4G&+-+{-<$MEgtBlu-6v-?^O-$nTae3N_$-(TK-`uaLAlSlBk%ZKo5Gp_>lwn&Q9gsm%Fp39DPO_g zq4hMQ_4VALd<%ZI@&ou6$kkJ++q%WUHCD|C-BBEZNH7-A5wk-kCb1*zo2~Zy!CZ{UEYD;D39Si z)gQrE%IEN*@(cKV_qF@Q5BRy`2ybjwe_d= z!u9o(%Gcqs@?H4#$|vwt`7!)fWL{@Y+JxfB3(s zehzOdzkokN^_TFT^6fp>*ZFGYBlt-9A-t`81}~JK!(XO+1z#%P+-rS3Z&JPm4{x^q z4B($sK8451&)_F1U&059o{uUY!gJ+&@XM6%!)MA*;WsN^z?aI`Hm$Gco66VW zja#fgUHH$HPvDXAWBBdLPvAY}SMW?8ym)<`6L|+-$zyo%8|{Diecb+sx0GMNcT#={ zPn2)(y}r&DDIdWz<%jS?l+WN(<>&BsDqq1X<(n^AU(ZLBZ^7%gYX8GusC){Kl%K&5 zQoe-um2brB>p4#O5S}aFgP*H>A6_ayg?~x;0v`O!6Yz6+1#V|ZV_ zg8xGGJNvB9J5qiGUn#$U-|rQA4uL;h-hSD7{gdSpye%KX_mOAt*U9JbcgZXG8S>`7 z>+6}wTkx;Q2k>9WQ+T1i&ES7mzJxE7Z@hecJwH}Hgx7y(`=kfIP5C~&t^5>zzpkCH z0-h*e+i!h6k5|49A1dF4Pvi-_l#k)T@9nrI@KC;jzfkK5Ua`K;Sosb-Q9gzblpn#r zpnML`m0!TGQ+^4bDc|m{uk#(sNARWcL-<+BXYj_-`ZkCEyYdx0Qoi}h_4Qn>d<))F zehB}iThC^E4!_f_XEWY>)%tqw|4Q4RtAr}7-0ss0#Ve9+c2f#=Fk;dQOQfL9v3KWiJ^3WATn-(HUk zkGe$%`SkH2^JMm~a9s^5W^ zsvpCH1Kzb!rwgyk6L=`^!CUeao*s1KM*Sf?muK+mwU*D}^~>I~QD+PRc8v% z+AFX5NU=kP*a!LOAs;B)y39=zW6nftT0@#+bJzbaqTe|J84fnAsC@a1#N z8}M+Gc?j>nRM$87J;!wb!PCxRHuZyI&*mNIy>G9o~X_O9%eS%}KTgl#a91aUC#sXcQ;pr9Za*&*`+8@HeDuVj?FV?;GSA`3Zsue7a%b}j zUVp94yMUL^w)_&F$!mwM_p|>z%h%y8#_L?zTDe9=&koBdDER6=dlJ4k2bHvhw=t|DG%Y@W2{aS zUdUVU=CPJ?G@Je;Y@bnc{X96!SvOY}V`A5tPcy*Z7nZeVAT|Y~Bq2ryy z^J{dyg{NJszkufh^Cdjp&wK?BR3~`D`u-1;ufandZ=HU=^`Sw3wapvCgWU2>y87RO zr^>hC!N;sl1P?DT@4&0$Y&|i&_>|?l@aV0UPvE87|M2Qj zZ}|$I>iV{Thi?DF^S4|56+F>?2o77{|Ecy>4PNMaRfi|)e*>P5ZT%rU*Y&DNf2FOn zMemun;i>u@!6WVGKD>2`-R}nQS8lTVZwfzHp2L&Vtj-wzcI7AV6Xhkm|30fThrdbr z3Vwn-__y`*HPLa^;HN8JhksPwf>%1OHvCh{NANGp6L@;M9aj(jb>;i;>*X0dIoaxr z;6GA6hc~XW^H{(Gt!D=RmFkr6KggHx@>E;T3jR0cgTvR)OZ~;Ro)8|XP80rcuUT8dr*FR#xsyslLD5g71Hl=7k@7hUSHTOg@G`?vM8RD&gl=R(}P*Kpwtj zeVu3h#PS{Z$8}tN_{FL-g1dee@Iw!``r#4l^SZuu;O;#3;jT^rzq4oc7w`)|WBrWY zx<0SV=WsXg4DOCAJaWB`^AzstRPbAmv-Q;8wqEB$k*&W4U;N(AZx`w`3&yn9l_na zV|aX_9oGcjmlyEt2lhVL3|{}1eLt&&N7tFp;pMgF3wU~s`4S#oZC-oF`uXbX`)GA| zOWuIz|M{Jb^BdBCVBUdGj`+|R-tgG%{zp-i`#uf1#C$Dg(9XB~dIJcM5(Z^C~pZ^LhqNASPOWB9#4XY1+0A13d?cb50z zyUSDfp7J65Rq_%15P1%NvwQ-7yL<{iNj`(0Aur*%ymrF+`Ep+$)#3YZvc83I_w`W| zzMZ~~Xv3fSZ9Ol+f2pr)V)*B@&My24c@KV-ybtehdG*HiA%$O~{1AS|C0l)V1*5@UBwX1pX?)CHaAmwZDEqAkg1O6!GL-?-p7Ce%-;s2`o9r(V= z$M9X0Pv8eC--AC>`2qYWtbq z{7QKb{zG{RpB(jrjs6Vbg*=0o@)7(dc@F=Rd<@_6?KbZO{s8$DzN5T=?g9W$w>|HB z{xr`Ao@Y+=FL5Sz2{@kH+f!oexT!e@I3PT`JQ(?e}U(*=P&d;@x0@C&-2r~b@n~q z)61ux@8$W>^VF*|@_awfbIP$R;iRXpq@AB%*JdeHn-1C=uUU|Nc=L^qY=K0d| zeLWAp_s{FVtG)HqJb$^DuY10q=b`7X@Vx2y%e?w6&%0i}6VG$cU+ekU^MgDuJb#_%GtUq9 zy!8C_p3gn+dHt+BKg7!~JU`U);QRmFKX357=J{cs*FFC?&l|pb>j^!7qnB@c{$|f3 z&yVoD_0X91KJ)wq-ug?=Kj-D=o?qR1 z@ci4J4?TaSS3mdsdM`im{5zgcJ^!xfGta;0dFlE0JzseKQg1!MkN>$Je&FRp&*z@E zJpZBRZO?z?dF=U5Jnwq`Q_mC6U+%55=lOd)?|c3;ug<{p8$3@vuRI@mzK>Ub5Ay2IJ-^AzFFgOH=S$Cj<@w6&-jxeEI*e_vLXmSO5Rl zG7QO`=+;QfR96jB$Y?|wYj-kglq8`^xRPox6GbIvV(6Ai`!-4?$(AH!8(SNqB-_Z6 zbKMxJ5b6G&&(}G3?zzkK`S$tu_nF6I-tTi>@7I1_XL+A<-sf8QV=VZu7JPyQ&$jp! zTkzj3{OK0Fz`|c>!GE{#ms{{ZEciMLe$0YzwBW}r_*M&k(t_`{;HND35eu#(Lq9xW z!5_EuckB$u&uNQKk_G?Uf~Qz;hs7t=f_Jds=@z`51@~C+$`(A^f$i&$Zx>S@7vrzhJ{hTKLs0{0SDkx&@zX!FyZqmo0b=3qIX~*R~3@G~v=RtsL+g73EAP78j-f+tz<6Bhg|3+~tzj-NUfJkf&pvf#-Uysiar zYQbF=Jl%qyW5F{l__-F`W5JUxc(w(vXTftU_<0sQ*Mir#;A1TK`4)VF1;4<8PqyF} zTJV=G_(c}H*n(ed!KYjBODy<83+}ey8!dPP3%=EYkFemoE%*crUi`N%UxIA7vhXjn z@Q+yV%PqL0G#n?X7CgynwN!3%=2U zr&{o>7QCSa-)+HLT5x5-TU+oW7QBrGKViY!S#Zb3u%EkF@I(uqWx*od z_e-K@Qo?P|9CmmLm**v@aS0BGdh)G_4$rt%i6xF;+AWVCb_5%A$JKdJ22#w+o1}kt z5PuDBWS%2#|3=BTXP#37{^gRtmia2o&yjp{=BqMaEcr&vL#V?)S@M@KUyb=Ol0T1m zz6SVnC0~d6MCNlOUz7P7%x6o!GV`2*@MlQ=O>Kacqg$v?t;ede1= zei-xTGoK>)!OUO4e3In*F@GWR4$1doo>MOVBd2NqEzGAdzgzM*GJi4i8ztYKc}~Un zmrMRy<}YP_j^vv&&nX#yvE&;u&yWXWH`{AJ9Kk^FhgU(S54_XB{88qwU_M3i2bpiee3In%Fn=ZU4#{t0{wn5={3Yej zd{gFkOa4pdn=!w!JU?)9umRz`Z;HB8KH;>3clw+6MdrRI>dyMW=Y!3CSk!&2$$weF zeVc)Y?t4tW(HJA&+#fObvF84^xo z=04Bdzl^$XGxVoH}_l2eXzODGxu%gzQ^2;ntRO&n$IQX{+zk@D$?-KeV)mO?ki1xlewRa zd-oKTdDV~a&2)Gsb#y28_cX5BfOR9khNr01t5*Fv#}RZXl!V~TtqAXR!CMZz+dM_5 zm$7xbHQTGsdK7$%4tt79JVk50s`~{WIXu3EhOPi^X*25Z=4yRgb>NhG#w}0s6jkx6 zl!>HSE#l7=dMMZi`~gwBh7cKm#v0&N7m7-4+^U3+V1p;Z2B11bH#c-!##Y~}MvbO~ zwb105W-v?ySNBXRaGyXg6Q;fo%C&z8sJ^tDf#+3E7R_{sXA=zNyoRBAckb!?H@MvXq2&_2H z5ZI5(YuwjnRLFhVP5;4tXTyC}^2JR_!1JmbCeY5;!;F4=B*dute`2&KPZ+Jeij4jq zPe!+f8OQK) zvRVb51)1#mTsctD-&T_E0;ApSnMKFj^~fyR((bl)-LriC8r*iPr|6XDiC?&3qQ7U7 zu6mxL0ykH7nl1PEI=a)>;2*Iz#GUFX>gY~wwmfs(Urr{QI-h(M^12r8d)UqLCrx?u zItzC(EPU5}c-dUItmEctj)m-#Gh407a<<={Rdm1;T>J0^ap>LyomcJnwi>QKAT%>k z7ZEB=J(KdFg3nV4!@T^&iPhRo=;R30^b~%PmylId_S-a6OLhId)g4(yP&KQlzdJpv z=xA2a--LV5PjoC|7me`mB6X%r-3nMB!Ji~*G8BdXr0$SZR22rbant;fo=Qh)>>%}9 zPt+GcJmAqpc&z1vMD<87R{R$3EZ;QubQXYbrql)4nB|-APK8_4Lj#DD=T3yG>Rx7i z1#U&;qMz9mfKig_X5i}tj#f+H$*PqQ&lh~pq$%zr^87eiaAsLFEYNapqLP#5yUBv6 z8;R;$;!Y$U@VWAxRTN|q3#igjRUaf-MaQJQ+xwohAxCNl9BVry4HG$kdwipkP3x#; z^(S&rZSnGcT&>65GikItS*(j1pySvccl&N`! zd^wfHtI3&*_!5BvM_>du(tDE)#+WFKzhPK#qbPL0pukV+9)rxqlk#eokM0iL~%j-hn)eXFYuKZDl9Y)_{P&K$QFca+wfV6C@ z1Z07Gxe?cp#*`S~7!AyO%HtdD&J`Y>qQ8S1{UgyNc@T?pfDi@;O4xAavU_P&>MWc~ zYQ#mh8ZP`sqvlC{0W3$A-L1o|r`m5)8J?nm@)*=~wcda|YL!WiRE>;c6Y!2!?l@=R zcuJ!4a(<=r=wgTAlj14bsm6hUe-w2N>&I^I3_1bmsp?x)2LBzv2)d3vf|j;pEb=bE ziJYRQ5&1^JKv!w2lIIG#su{dQHJW%Mi05;i+$6ou9qgSr=qP-G9ac#I9V2p&NI!s1 z7A=_SM6y~SW67r@qb*vR18_qFo{LQ2fY$w8KH&Qz%QI2%(YTkpC%>DxrqdA)d&<;- zd&n>Xp^+fC4Z@mO1eYLG0YcDq`NJGsTrAi6xJsvo@!8PfVqkVH8IjyQzXqgnTy)Jj zp+QV_uX<+8OmqeBz*=-Q`L!9+w?mv)eSrsPxnQwZ^?PupBk&8H<5k<~7Z&XJvN?{) zh`mnDzCb>+NM#lMgwfafYOtlXf9f|i(M&CIvp%V6)ZdqI(OtT@Ji1jXkXZGqYhOn< zH0cKAnUv*D$tuE-M`Guxs=Sl=Q_}E*!N6?IAS1+p4X#&Rf6gq(EO1Y^2YAjx?xPh= zcOHhHeF??AKy6eCF<@;`-hL>p-VziEqYTnWc3d*hD{wbu2($+hnB3V{tB`6x^s0wm zi?Shw^hHlqB zx|)qRjPfZ{9{^tpY0HnZ9D&8OkjkrPS;M4C0I5KjMueCd{~eO?-$PIe1MogXWAx7e zxZUmLqnXU2uk~Q2Xai=1l08M}YlE)GhaPsU>5wF22ww+|@4NVhq=~s5(DR&jKGfe+ zcXF>G?5`C{;V^CIWm=`u$S19y!cYLEp&UgRdWI zC^(3%OMjW;=u~hl|034G^l&BB?LaY}23@Q2wS(HQlK%jM#&n}WhmHn=lf7!dP^l`y z?V2I1Pfy|1#;dvlPNs<%m*q9_OzMpe8Iq6RMaf$c04XxY32c{|^#rk5vzkDpS7plE z5o+AMA-D_h$t>~(WB|U|?D(AQ2{OV>K5Os`1|dm$Mo-a!@MFf)1Vnv=oe;V(Ks$Hj z(ati}T~sVpS;DA5Mm*&d&cb5UJBhimpd!B8;GZ8KpIb@={}zM)dtbVsRQc;^T7*s^A|zsP&m*@Slp0zaHWQf2G0iZ}5A^$A2z_|D3^ZVeqer zho7G#yay5&Rl29AFql`(GpTwoo5i1nF7=N-G`~z8;vFa`$Z{0Ws_z9J<(7Qi<~zYQ z=(>71+x#YSyVVHN7Uh{SN+JYf>+x>w^e*NxqGip)=-*$TRj;#z@QZZ{%su0X^! z{~%(hhr}T>J+sbZ+d}4o^g2JaR3TVXt;NI)&7BGc5R4R zuMo3pV8%$8E^mkteVZO%?f+RLjE_aoL_|vC#8Iq8U6q zBm_wY#H&sXw&i{`fD%(g?oXj7!hfzjsTSgUS%}MGU{P*&MY%t7{)^>UE%bcee!2{h z-+?HciUIMeaUt#?_NrTj`*cvFcM3fb>{jyR=@8#pAub(|pxoC*?gOFcPlujg8J}xEPiLs5GD=*`iJ1Flr%**59eJr#(59pq3qCvAp-#Y5o3R>WPqLr$9 zh(_Dy0BoiKn+<{wd})qxoCef$oVkg3g@;yQZq7n}|AyCz57!8GO}8-gs+O1?GmI3f z0~iIzh8bNR!zjNGm<-fRR$pqE{HjOWYSry*Eh6Q9M7dP?^j;QPsd_60DIZfwK-d9< zpvwVMgRYZ9w86tFLY^(in}HT|eIv+vw)S2c5urkU3*r8lh}TFRW06pkpu!!(^- zY&Q5s5aU(vf`v2}t3&XifYa)_q{qPS1V(1z3ff#^p{wAm_V{|cQ@m=TCRIHx6lwCL7~b^5(j>-3`F*g6#$ zs(-7~+u#^%v`nN^V#P473R5{ zz+HyJxy%#>UC(%pf2sixTk1NFG9Z~fxYIv5axYUojoiqn*PZ5b!h1oONn^w0QrrWd8P4YzU~K~Vx=wMG4Diov z0G|PnbBn()Gg17uUzH*QxcUKix?}_MsD7N~2^ZKo!1k&dP{fc;Cy%v3SBln(Y+oRV zgFt2`7&g06?yQe>Jia9|bsw%_oMFA&pcUI_laUFdaZH+|(gjgQr#n?+Wc)XyIiUwH zD z)C%UOGW97_FB$7*AVc;Us@SAFjMC?k2{mcd;+376M!=aeh0RJDz2>8d|jE;KA} zH;mF%SEL}34BL|-XP&4*8Z&iFCY6~4WDI#S^3*6@)kZ3?j09DQ^Apy}49@R2A^$V; zg18@9mUJ#ky1a;Tkg)+1rrBh8a^^#Gro zAJ)9DJPE52xu&31z1H9u9h?P^le2G$J5mq-Ab>HibyxX#{kT`v@1^^z$N=c>Refaq zz1piD1wQ5Ea=D9WYVgilj?8htV9o7-TH3WbmiTt$mo4N5sCG+u&1|@;>wWuDGb`=j zlS@(7{-?Py#dX#_haC&y8%JQvLhh#uY*@$@lEAtUw_I=weA#SyV8KF6n4?Ww2zidc zbTsLB0r;&iHNpHhU~2Vt9W~3D8M%X*0-0qDhw%uMtXB>D5!F(Pw0Rk^=>Jqn8BVQ* z7?UFE*DRDdYPp|xnygBZ3cT!9=ZI*m0W8gcJg?fTT?z9h0NhNXk<0 zYmvkWCI1RZhV7n5V$1Hp)c>nj`j;2?mLGN5^;B6#hYYo6a!8wWqfN0fv)l@nVSRZR zT~_BwaNPET4i1my$w@)ipd6MQA|(v=pQKd`!y&y{_9l4RieAyouM*}bwuB<@Yb}TW2dEizy$^Ol z*Hp8*(Nmo(3NZA~jH^lmQRN0v*eX=+-!m~cK=|GqS}#$nGqv;6)nZ9S8>c(xKfb4N z^~&{l#!z`wAyWR<*ncK=BpW-HqX5+m*g*$#PFVFZ6shV~lS)^aNJ-1^IhzMhL)kBo z?z!(!s;81kw_2T_9qVyD36le*bThOJ8-36KnrRcPD%H9zp2yGjs+(l24*_T%zVxtB z2)f#1;U(y5gR58laR&*@OKw&7czo`}vG}e>^BcC$>=9YwCaKd-{Pjp^C97 zL;cyAhL);dabL)gYPW>fh#j)B@D>F*~6_wk)(1z$G6m0<;6 zR^HfKhGQkL6a{?&!bK{8zb-o!g!J)F+5=_0M{DaWEQ5`raRaTfS{n~9zm@o*lHio7 zcMQ+h;^8_Ar-pcb`z_<(53X`*UraoJB7#z5c}F|I5wA-7PM2Gv>H>cL-`O(h!E|fp z_emaaOZ!!0Bll9mZ@|DqyYg&in=s3ls{N`e@>n<1nd2Ri`I0ek2&otka4i9PO-3yaD5P5gRXn!dXHTDh@3L@eS7VxZ+MUPRQ}b%qq@i| zRUgE~L8}LQ_pYb*)rEH@xt_XRTQWQzm9ua&qF*RK03{=(OFTSdaiQRpsmqB2+>7Gk zYK!j>O&*2xplgKDp{!jz>$K(#1;4A{-yqj^LRF?dBNe30jZ3Yi1-_x(PNAzI*D69? zsz$}7(lS30svmpMLuIO)VA@m8X4}fsKqEraSOmSSa=$@v#v}&Ij+9GM5zseTu3YI1@>1=0kkdsZFSwn?dhI=+_2Bk0i=&m(7RI zq>>or^#L!=-Xr}U`9_kQFHyeuYx=A7HhI$qw(1V%fvP~WOvA`^0hKv^#ajYzuG}E-v2`IM?>&B!l!&e?WZoB zt8E!V>k&e$EadxXS5Y9V)!w|DLL#5LUW@!fY_96&7KD19N$0zPpz`ssT=S~#09M(y zaiGmn?nz+=)!zGxs;vyc_k`dhgd=uZMI3gaUGs#O}f;-zin-)_w0=LqTfRL)X?U>Km!(HE+5qQ_%P0}dfb|9 z-cr?fUts$Y8ji@hK0Op#|0N6i=&iaT7~GbwKyZ8tFnx1J2=JZ;aQ37VU>E~xD=~sKlg_R3>*3Ap(A#!KXbt?*!b(6#9n_y?#d%41~oESK?QSjJ9| z?{)WN=`E_d{hfr#arG{klq(@S_h?wKjSc;Rg9vzQx3CRw;M5x;!_-rON9q-z_D_r= z+G+yB?4taF8YihOYim0E_eP<96p8vKc4dt_Qi)B&ioFS*N&!!+04D~p1vxxozfyXd zH`eKLyPo>PMc3f2Uuiz~j0>bNXxsn=KF9v|Zrljou17m;`Sqn8cy_LXo54dH!5NL7 zqVA|U*zi3v%ZFXm;D`6_@YZlnx;9&`&i)zAmoM$g*Q5nh3CL!Fre{ZPy2cjQGIh!G zXa*7d$=o)Fs)#*Lhq?9Vku_`{%hYqNW$I2%#GOiy<=vNX&*6iPjzzT~GB{sqOyOG87+=DX1a3tR8b+#KJc_)`Ks&ga ziv7vRt0I{L>X%T4DwaxgM$6KbwjN{!UDt5~;1jE~AMHI|(s9tCIoP>*BU|e{cZvtw z%{gzzJ?^Lxxn{F@j_R?O-tFzqX!Zpg9_$4#Rh64^=hzp6j z@h!PLJTMq!f%?)0?mLK^%lSqPAfqX1eLGep*_9cMROWxuXw{xrjdHKDHL7ehT7;@a zjpnh6p%g4 zDZ2oor(+}5{;{+O<&Lh>)96zmh~@Z{2#i3)TP_g40AkQ}s4Jav09R97L);mdi}yqp zR~bEq0LnH2l%ZTdaX5mwDN{2_Yl@^xVkKpuFz(qSg)#qCZoY+wQ!wX2m;ZK+Rxp4& z4YYPKbQA?sMZqkn20P$owSt&rvHJ$iVooe+Js~|*Lfg&)DNE-oUN!_>l|ojs@j*4Y z6>L~dGFVNpW>6MWO|II_F%(*rGPPqD7pBQm3+&RNK#^kD2mKZ&txUonw)wM@*I zeTusDP`{ufx8FxGF=Ogx&>aYBb4#&SeZ_kWZacWps}jV{hFP?;R82J$uf(OOArzZe z(lf`qu+)FWm7d4myHMv-v2yzLX$r_Gs-+Qi0W?#HICxg))B`Yrr`(H9yqu`7M2u_v@7)XW05?@FTc8V@iV?C5!!3oDLJ|LUoPxlZ zYz%#YuZ3g~+wgC!2sjItVNl{#b)lSpK9ZBM*24W99Ao};H9g9~>xhS|JDdyVYX&|u zHM{AlGsK{En13iQxO|Ks;G@5mF)mjJK3SFmnO-%r!%RoJF^^}|K$RX*znycz4QPs7 zXTOT=NB2}6>#4LcxEt~BVGK1Wcm+@&hGC$DU)JapKr>~Hlsnfq;*s9Nz7?sUYkUXR zxH69~_k2}<1{wxm+Nrj{y-pLvYoX^su6baK;|Y2w%5`oh`OY z1$lR@Y2muvjJDX4d`DUG6uIka6<)n z6ZjzT)PU_QRCFUtx6sdU;ql+~RG>%I5cQ*-c@dm>{De4DZDKcrR;KDrY-LLP$M~VV zt6OJ8nFI=I;i{(MDa`1w>U__Y3{-A2e~T^fj1lt|h_9)H%&V^YxS_-Uv0n8Etbl7? zelZ;D%{JN5ZlXLK^;6tB)?dpA#d>pf)%z0bx#R*9cdsgs_3xPs$9fq)v|>Eov@IO# z_k&7ZCxG%;=ZSXej`0=7`m9BvSYLHjC2*ipR56is!S(O{@>d(gmNbS_%y!(yQ+JV3xz*NtQ&4JBB0tN(2cb zyq1>22(N^sM0m>eD8lMyWkn+V=PN_j?LZ3FF~Yw^GERh>D14!9I11Nco`_Lcx-~iq zhvNGTU3JEZh-fn}g_%Q-iJAR2nkak>6B>w>eqbuI%h>;m_Rc>PRnq@4E_il5pP6)H zrr+1_WTs!!>1s?rqtm}VkMua5-ox|=r0M)aO=x(;9F|@+YqW?hI@;4%1U<)A!ytSZ zi=gM&)Mp&()Pm~x{!Ok9Hro(Z;f~t73@>Xes$afd7hcQTOI1+mL}Rz}0LmZ>UpTkg znTN0GN2RC>^%G2aWvDtx2lfF^B?=VvyRDQNK8NwND#^Pz3w1Uz)Av}4I&cLG{dnfM zH5qEVr12#1Jq8YZljEG!KvGg<9cbcJSBBs>h2VRyqbisMK=1lVI!6pOLpY5>I9nFN zr@?jkb-ijn(9sOz2^MI2jB_HhF#PmId3qFAJnde z0=`#cXgJ5kp5m72a~;k2v7q|m6||B?Yv`?AAl5I`*lT|8M@W~dlZ~W?tE0HbS}+u> z`n`g}_NA>jc8Dh9BP+hBfO2_U3cT_F=i~yhnHs+WH2+rgZ9!LUX}n>_Z>la-p>}8B z?w^OJ>}Yt0{ns`|1$;6U8pORS32N@yC^uyJxl;aM2HWfl zSdub0EkGSHUuMV*YqA=@N%sRkamI3APE`kQZgVE$VYRJ$2S1b2L zR57|cxE#@U<13vReYZBkG6Lm3h#@^F*qOa`K3)a(O3ci^F7z(vRoP&V4lb}G=z0TS zhK?()dRS`!!<(UA$nDNTz6Hpu0KWg$#B5T&XOw=1N2-2Z$b6Rphc^mkd|fbwj~sj? z6QYp#4y#&>LE3dYKc+(}fjwR~O4Ue%`ZBnM+b6vnbskVz^xXx=-i_McP>m5C%x8!H zhh5D%BD|mt9D~`)s9nwX6Raa_&t)`#v$tDKDn+fARMZAgXW@rbf%T;-cq73t>fSM^ zyN!CG*#^$h@Hl`2>Is8UY%rdSoDUvH;T~3n(0(Sp9r#9&k-WW-W1l6fM{|ZlQ$pa+ zg_mg8CnLPV{ybL-yWc#z{>j!bQ+AUD)&yN6nS+myuyTN2nMx8#p97&@bqVNW6%&i+ z3OT<26>_auNOvmyY+SZ__Uo&;RIBV-HvfKHsm|TP59=Rs>9jh}gRTN~E)nuFb)JyN zs#7^GTdhtTQ71>#=|y!=6seI#p;o%ZTP*8OnPmIIqPT*#z6%J87x+5p`mGhM7Do;eV)TC^0sEf~`y_RfNhPa}l8TB%^S#kyU>tP4Eev75!Hf<> zJ?#Z$<@?V|b;yKbak9s!cV3ux7#`^GF#0Bz%NNAeo$*p6Y32gl`UhBSA3jgBCR-dklmA8b%Zv-L@jcws&l?dUz<97M%;%`Xhif8zZO=U=4jLQYZvRvuJ> zy|}aC*^WED=*KGpJJ)kq%#c;0N(2wh=bq-&e2IG@sg%MaTzT zuUr$d%?T)NtByP61~2C5Ld*OOX+8_!&S!gZ$017P95)kpm&RR-w5-qR>_R@;%yhb% z$@E&CE+(2jC4-NiW!6=$@>pgcD9;W@mh$hzo$~MDV;*IpWpqb2P^!}&KH4E&5c#%Z z_9K%`QjM6su{>L!*~aC_wV2&ho;{s{?B??9VPxIU zPdJQ~{Ljc^|7(koZ9@R1ZT)el+}^l{LNwbv^O$EDxO>&Ka*j=jYeLkGv=yQ*J~AO% zgJ~0@r!NFrI7AOK8xGM@X2T)65n1ZC4tMIdoR7mHIv3ech`!55;Shb1*>H%CXEq$7 zBbg0{=)KH_L$nvO;SlY@EJKtUv_*#9O-tNart8oKg?s!5p)7U?%rIrj&Z#q-$wWit z5Gss;_@o8-9ln74j;o(>kL9Nqb5ew#bX(ZkEqAQ9wYZvH_m#Mk&wIiL^Ph3~XuY<6 z5Up3Pp&BGqwq8Bt$#Hd4Tz*=w3BoU0uh*MXuNQHp<+Tlc8 zk>7#d-D)7YT$qv5EvyY5bWDdB#UCo5G0n0|T;oayg}L_&e| zA>r~HDvaiU5o-*`ANd~Y<1thjLS=>>Ey?dn14%OHAw_f4p6V+c#Zz-k39byfhBc$7 zhTuw0`|9h++JSp)B3X0CUF!@{h9OFiCAve1@(x+8>iv>{tQH>S9MF( zons;vPt+x6VWfabg5`JwYq;*kWYH99<<;Jae4KE$!dO9;k*eaXivoAghyialMib4k zri1}+-5yS|!6F1^a9G1DLsOW=nec0d>B*FIN1jZkq^q)TMv~>I7X~ zny|z${Qy=6E*Bd(LD0PqqlcUSYt|wur(KHA)vy%xqNJj=tF7&ShlR8{umJCWSAaK2 z`PS`0S!zv4ojf|l*%cTU!4$a90tRp9X79sj}!L(d{7+b!Ma6NCAsxCdRoG(OBhMB#RQ`crbaH_%fjZ@AdRf8Tvc-2)*!84u$WUN*gM}28i$~{|E6esICG7;N^cne_SK*-6^Yi{^eL{BHw)*)lo$ ztWX@czb_8MyxFM#M&gRIQPscX3>GUuU2tGmeumtkk1)d6AVQyJV%5%KX`M+nnoKqT zs)iv&10o3W8U$TGHsqGy=NeG^!?^lue@yKu6+0E?eB=m>mogg%>#ClL)P*g_6m;K=W53 zO}G(t7V!IGwe55b{c7M)-g;#(`ed0rx0G`RQ{vXHkdLTZpDVo=0&FhsA zuTPq=_0>E7XM&l(HUk7re>L0#`vbpQ!GgL0e@!p<*R)gGgPbFW%bW##K~zrz>R-oH z0aAgl>4`?6sup2ifj$@7*WfUZ7IYSJn%@63E|yGZL1(gRMW$_{9u zuzh5$8kvZQ^5wa!LOjn8tco4QwN(4DgeHvQ1{E2g$eHJnbNrm`sH1nVY|i5N1d;m8 z73?PW;nX7TBdLXpfJccu%#Z%2rVl`jqKPZUkc3TLx>LWaG{v1QXoL*w*v)~jm5;Oz zTxhgkqa+@*_Q%887d#=_f4(iGJtpC!IOq3<_T7o#bGbx&?8pDMae2IqqYN;{>*ZQl zq1qw7W0_!stjqhj1YK?t_br3TQ(%Jg&n{(aw;fl+h`>jHqrb8V8h()+*dKH)5`;3< zTxxEdrTYh8W&2vD&Wc6QZ!b3)gd=Cte-RvXFVF-C*#9E>bdhT(Tqobgv-eXGkUV?; zg$rnHei8%Rm*87Yw_r!NoLE|AQm4eC4mgHw;SwSxT(gT$)5h$?4gHwUwPxX2W>8ke zGpSRGr>H{;zVgHfS<6Toud95TBY_rOxjf2T;Nib6**5g{=y6Ft6eFFFqd zN%*UY8xR`*-B%NB7a`_+Y5P9V+pn%dHtMU1U+%6A5WF1q)kMb^07s?mT~iy=9d}mZ zEZ%(yO+K&atBJc7*yJ_dq8l;xFQ~<@%gFe=<|!t|BpToyz}4se9tH1^9jc1n~{z3`Q$$FeXb^JpF3axi|a&! z_W2RseF@(#tjOn87T7-Dey;ZUXC$k@=f6EGK7S7^t+r)+JPOO_UOX@Nd9P50KJOZj z()fHlWa0bw^D6K;Yj$lR|A%v`4^^ct_;n8X2IK#9PIZ!qZ9%b+Fp~`M8Z=E9VUi@G zqv{WLWBy|@oMK+8LmMPO!*%%kg}h7+5c1f=guBLL>nwZ|{iXDQo{ztq#pTqN8| z)w%Jg%%RZTDpqE9721OKk9X*s%~vc0QvGgRb4C=zrf%e4-bLYQ## ze)tQQ&?mH%lM8;FLK$S_rK-Lr)l_wlD+}-47-u{US;l`&44Uk&$H7XRm{!RKILAnS zCvv0Bl%VUO|9dva+7r&>7Pe5|`3XT!l zAl%i4US|j}+W`5-t`-4^y6#mwU;jv+%BFabc=+^nJmo9rIvA{+f74ZFxHqg4{0Fq< zL>s^io@^zWI<3)IermhPY-DD=$?%O&tuPr5YE}0Iv~!6Zf~G$0thupB)T;)~!Hmyj zikTU6lr)HE{Musq;L=%mJ)VgpFY~eVx?`;7J9BjLTxgs21a2Hse!m8-Z_&x6&ud)u z{s}!T@rHwvIf#BC7dx#4Z8Na4wExpNivJRQ@54#>B5wm z{PQNU1>Wr$ccQYh@L9gGlYY=y453xH`g_i;17vcL4lk zrM;6uR+2I1boCK;LX-LP=Q=W+3l8B!iy8854jq|D53OfxQ$|&Am!Opf?YlGdOG%XD z#()Gnq{W=GVUEYwB=s%04RMyjFT=|4ZO8q<;BlzlPoLDwfvZM`qIp2~cv zCSCo|$J*5=pQi)l)8aBD>Gd?@^>>ZeFAr70$bzgyG|cZ$7k>}CnG;p8Fzq+n`w_uXENPJr>|j}Kha4Iz7#9URO>%zX(zQB5h7Ja0xL z5?IFf$wfiet3W`~tLCAYd2(x!0I-4+RDU!4O)#BDqKSe1g#Rqn0?I7Kd@V2nYF`gX zycYtOrf5^qNY#kaoSU7O5U-thC8*wtVeC#D*q?-c zgl$-U-+1!d5RXUDh4EU(!#kUJudvD>wx`1+ERMXS-_Li`z)4TJkBFfKZk#iRe1_t$ zr!cL50p?{weF^;!VkQ7AP7qlpJ3x>~cGqHN-rpV)U=S+Ivll!3hffBB<&9{ltS8cG83$^LE@&h6|XHSOC`TENctmwS> z5q=e?qg!PaO>qEn0*RsRI# zu|+%!DSsEZL}q`710jHe1cK_%<4A30sthT=Q!q>?(N*iFIte>+X^oVB)t|=AZ#)B) z59*wnL-G=nsVfvX6ysGuwE~k(DqB5C%wY!eVPpc|QGO|7vXue%BhD4bh`OurD6kOQ z*wOTP)$O&=*!Fn84w452(3yZz0LbQL;KuYT`7NdK+4Llrral3Y83UmD7FLVXoaWq~2` zH=E6W*YrL~kYPS)T8WA?=FtpHjFdRXKM!<#gqOkoVHO(k(07|O@Y6rz8K^%S|Fuxh z%3OLZ1nkUs2A$oFa00`0k3tE|~ zWYAK}$2w-Yv@Q8#thrn2M@Q?o1@iTT@dXB9SuBE{F#c;C1*Jd`=Kth+wh{VNEJ-)1 z2w#ANLhp-3&_bshgd1ZKw9xS)^ut;d`mS8x!c~5s=o~{MquzMGPxOaG=f8UIOBQi$ zAcAVoCjC6=<^lkNpa}O z9(h79zbB9A<<|{Tl;`++PvYrKV`$g08aj#|AU2|CU_7e)G_j%>6tuX#!M!#vu0D1x zPq@Df?m^dca(&83s%ohJ`Yle$=Hs_iF;CGF)FyuPS1bOi(o)5I8BiTy{QjE=Xe`$T zM$8*z4})HaD@KRHhR0zLf@`_&9p9)$SE64dv(vm}XMvs=DZLven>Qk_awSFB?SJ83bz z3%84xOIckeigJ(Bf^X|Mc=kn^T5^<{;e4LefM9dj6;O!qX@JZAd@A=6G7?=YhiK)V zBmg6`VwkM_pOhe^YPOo&0#LyUuoER{y*tM(OIkA2`y$ep}o%!m2Gg zm99>1LJAt)@;MluOAg2+fSzjaTm|4T13dC8m<}+Ry~uE28PhwYJf`DsWh{bjsDqDc zgsHIzx}n|$1nmE*&I0_a8n4H26&GKZqxS+OqTZLg-&werp~33;j_{U>x#^HNRb2}) zG^M$S!vPyMX?~8d9imm@1kM?Pen7p0qf$89J#~a^O4V<;FGPEW5z~=M*ngL~Vqbze z!@M&CB9bl6j)FEWF{Blg$D!piW0b;iSW}N2u8~*lRg3=mP!0=r0SCu~rm&ETASv)Z zzJH_9DBNy4<>Su1V;)ZI%O0xo z&-+k<0d=2{77Dc>7CQ8FY#KrYpDYecElp+R<52wSR9L>Te%Y>Y5n;k?i{%$ z;mR1CU`Q$)jd~5jzlgz!B5hVBsb(CvTq->DPue1s^LtQoZNoU1!5#Z00|qs}1avtk zt38s6UfI`8$N3^H1JWYK2Z7hv3UJsdl}Db*DS-ImmCk$yEZKgH#~trbfILbes4V$$ zg$*d|GSy(!euWNDSB8s$e z`wtH7hJ)auj`by|u#O^rI!gTWvkU02ew%M zJ5BFQ#g%5&6Mi`VIxd}7r%T1^Y&)R6vx)aO-dQB@_}=+_IMqoOb?O*(CJMh&obOJ4 z(R5m!C7`3MII+`3$jej*A&>P=tF#hAc0(F{Htj-c4 z$NRsK$Ex#sT((-BEyL)Ylc%W8@3>N(A;J&eKgOlghX%eWbSp^_-$$1VlQMOYFo{*F zdVJ<^1I~XI<^zm6JNIkve9wD~cl2_$6BHFF@}`iNsTYMjR*}czv9*@7=MSZaww~0! z$rf&<$`g;uv^rx!QGqgN33-{SA>^^j{JJlmjk>5Biy}8e5$yk$YX>R?k8F(U?J$LX zK=*c2!H=WkSA;Kz)z6qzQ&kvOmR`k|z20z&P#%aiD8isfv?ii?_-FK0J{RPy>j zC_2WzIsW`04a4{+7D4xos||uT7D4xo(|~~QzfQ0iR?2k=uJQ};ld5M-$aB|{g_?1r ztK;h{{Hz)Pay$^;kvij?>@%E80l`=88M8pND!D|?LGI=$JeH4>$g6v*^i-3Er@*vc zclH^#zLNip=RV>IC4YE|j^p>7exrGq2r2axt?YScAAf7GoNT7xUpo3BC)3|Mcb+~` z6O-^;o-?!f_CD%DbxL#K!H$oW3OQTwHtSoz>2{^bhE&3W*ym z32a~5+fWY&B(S@jS$|}?B?&K;IAIRsPkekfb2cIdry%n^IR#@u{IG9FJts2kCk7nd z0fp#-y8&Pgqw?j0i@X~H?ZsDRsvVw(b_wC52T)0lq?g4bHQR)8?q>`Y;4mj_L8Pbd z?9;AyyFTsuwz~_fC-N43V7DHpmg3IID7|&!TY=&{DMBIS*=d|&iTYs2S@;60CCldc z5Y~cY={Gp~ifyILP@`S-wF@SXFb60itS-r?R^mD``hQ zglo{UU<>j{d}UFhTqKQ_ExQRCe=O@4(>x|ZGERR8b6@y_RpuF7;bQThk@8omn2KL0 zHGvlZK&Cg3b^(VN5oY`k(}J!HVOORW3dujbQ@>;o{>6B6(S4M#>vxnAUe7Y(zfW8* zd^tUqCaLHR`op9g9t114ZHdZh^bmIP`ePmc!y-MrOS_c?Q2vVak7o*Lf7B9fn?l>5 zs}XeuueZqy1B+LMXGizQ@)+i6J&5xk2J7bVVS_j{7E!;j%mO0Le?a?#-+#cBm9L`^ z;r!!RqUK`t!yqDJeS=sx7EyO?*}^?fxIbXH|IF}$4Z9@FB6Sz;^$qvW4dN%Uh+aWF zdM|Z9$MAjHpg$Rlt{qt-=m(B!7v3f4rRw%rbnTfbg1!*wLDxK7`HtJj$hahyC|y*# z3q*Vitpo%T{JDdL9$}1wR!FI!4n(XE(UE;V>mUwq_NvN%N_QmNK(dNXtDP5b=UlHc001NfOvYsTeu zB7gtj#pP_!l3r+d+(aJO&orqMOD>q5w>XD2EpQ6$97=mX+&% zF_hJ7t!NCAWUAGx>K))bYo0p;n3x{!3LF2t8hCo5980?2`)grFJm|Xh2xAi8hd{_R zBBURLpvbv9rt=Tg?CH}_!GRI>);2~Ket*OEalZZr^v8iLTnZS2U)(@(IGhD+s!U&e+I+vb4C{UO&tUH$-=Z#DOaVBV$$NTi4D`Ve+q=t=xiwf@Gs zawG{5>H;CKQ}qTxV5J$g{NzXJ`qKI~WJ&0)11S^rs3lTmj#_&25PDzQ;&IN`uC z5JR+aFpW4wY^3*l2TfL=i-d`cbtURTe9azCJM`H?jb zIPFiMaNmG;g07{9>6UUGvY)wPf5CY>I^4IxOKml!%_Z}wt5a{NDFey z`%LpBxt<{&v&kb~^EBl*LwI^Ho2OLj-|vCCl&oq(d3HmUacAWpyoGrTjC=i|>q%-v z>v4rB^;$pOyI%)n60P2(Tbc;myzlV+S3t%K*b_MPBd}HSQ+Zz15#@%xhb+AggS8%S zcjg`S>kO0uLp4RO1s%Nd3`gKK{49Ww+)t7g_j727jb*wu!Z+UuE4W<-1-f6IT#goI z0WE6^rSuY}IWHo3FRgC`7J)Yd2idg!O1F+g4QJOFv()&>-MTNmJr+T~FitTDEn^Y% z3*&Gg1YIv-lU~sEtXwDIig&udJ8>Hyx+rXC`w72Vsua1uX~a;H+Gv(Y?JmqOOBmILmhT}WlvJu$eb?X$Nj60TG7C|Q=NVr0W}k~BiN*18 z4_$^N&b6?uaTZPnbNCG%XAa5+REg>{1};*`YBp0G4h)nMy~-r++Irv{7|BBjx0`Fx{6>cfEFDiv18LTVrQp=A_Q3 zCG*|IK+h^#Qk8d5+`CLum|twd&QnAly5~YO0KxZ9P#R}NRQUe(I7G9d7(o>fYY`uA z=jV$0bNlF)?pDDutm3`9H8)JgP`l9b*8kz%EgJ%YuDqY=>j!X+(KkG!w_zjoWSFg% zRJ6wWeKXiQR5RWZtbfq?F~jEH>ijxH1YOJbQ|Cpva& zOpZm+&D}RZ!0*4{l%k+3AlLo4qL(h^=wVP&VA#1!C^=(ga4!a$Pu)z12>x2bnON}iNw45Bk zDD4$Zr&@olb&JGti@o{4p}ptE!^y+(;E(;pvK24?_LcGSg_zc6BbTan3n-`tK_A9{ z4)wy56rA>Ot@g5t@N!#`8oXVHbTdQNFhr&<5(xAW2-ewc>O7o5cz8GcS*8X;GGk39 zJL2uv?PL%tm`&kKYq%fOZO0&}6?8oX2@KY%WPoD&>r3rEe!mf>xLb;*B`Aoyl{moN z7#Fv-;MO#_GY#&vxVSFC-PMDdRE8$l|5nQDFs=HqD*G;bF@Ia99h z;Yywt8j_yjqSe80r;t1jG}e#f5|16-F&isd2wG1;D^nXQLOWB?ngLDrUraZ66-MuP zgYYk+cSd(QbNx;oC!wXk&<_yuC-L;6;}u}dy6P4Y43qOrDoG_t3VR6A8HOFA7aeZ| ze-tQp4*A#WDzQ zLcYHhLJ{`H9DO?NU=YxPenLnKdYTl@Kab1oLde07R)87Ge}~Z`wC^bJA>RK0G{)63 z9E@>!(Z}-R_h#2<^Npp>f)O0=qo%MvglmfaR1lt7M0^-J&(K90>XjP*%Pl&_pEkHp z#>Lh9yO$f>eg?N!TwL9fOaU&=|Ju&TpC{LuxU#(5pkipBFVr2~F*V>0A$`=K|FK%5 zAHtmtZwz`PLBGkMuQBLLD?&fjjf!0abXLQchz1uu6N`R<_~&b&6Y&8c=z}`@Glk!~ z-my)y=x#`3<_MlOq*uox=;kBWAk>LP(9OrqqU&%73cBu>s~6YsQ6z01n)hEG|MJ!r z%#--iG9R8NWic7qs3RH<)vpZ@-aw;{f9Z5H;BZKvRk~bm!<`Q5%DXS&$`KVE(U6m8 zAJLG$g;SS7bq>j5Ua<0|5@u(^_r8S1Pe&Zlu&*Qzg>^*3Dm)J@zN$|{8LBuV9;I1( z-FGv?BJoF*=|3O;vM%qx^?Bp(#pged&^|vO*(jesd|rH?f7@02+`m%${3qOL(2u

7gwOw49*4s6`3^iU_xXlUhAOU%M`?WiR~CJq zvZVr_W0Vk7bv~!b$JKdqjLhc~Flnx8%ZV&ys+zz$tqb?k`1Ka_anr!pGwJh}iQAsa z;MZFgeGe%bhP@NrFUHS)o9{O(Z{!T+YPRReW72Cl$In2=?_2=Kx<1Os>~`oAPJ_OA zLlnCHzvR*9AAS^{qMV5ubS*NgJ zjm5C9y1=laQdvvFP!-=le5Kb*vW1@-0NTLEIREuK80$-TAI$?N_1PtReBJS$fMW-< zxwOFt2*Gz5f<=3l$}bS2u-?X!0^w(hhS56kyU&1E=4g96Wg^?F5_iBVOq+pWUI!ufWy4ph(Bc(!3ey+q^fIJvJF$R`Q7W|kgb!fP=bIw!T!fFM32#7+qJ3>J~a&KY@Jq2%daZ<61Fg%uHhw2 zU!l|Um`*`D=sK|pg^E)tX34ryiGMck?EG?2L;mqC@Op#rFUHaRGFda9q(oGUruNI0 z9?H^t2bM^As#4zJ#mLvm`z2!Qff1X5#e(w>KVtZ*LC7til*|c_B4fGh_mjmQs#*Fi z$0R@E%7&tw(JIq092u@eddtU?AfigCf>^3jViEPWi9W*p3E@5pSGpo2>R2@Vn zH=+FaRbBLIdU?X*)2Df;7In_hMQZP#%)d(<(HAOF{mGx82c2trzwg(9voCEvdph}5 zWcAaR7}XRVlwS~LtFY-9TXX#W)j|_3HftQni)C7T|H)+(AWZ|(=R7=xyZs!P`&-D2 zr=jxPr_!@3{gKt`4}FZWXWSvQ>Nu|(1OdO0u**wU{RsGik9v|9SUBVd9_48SYjq7# z-((Yd5RJ4ks45I)XjC; z{qT!y-1MbQMGpwQc*w#LXocl??Wij;yr84f_*2W!yBPFRts)#%_=9vLX^-ScClsj8 zU-QKR{7V=&qkmaB1l{WkNX%p6DI}&cF-`=Qs4Val^!wIgwXiLfL*IG`Xv2Y~e}l}I za5vLkwdOZ5%@b*Wa~;!ZOk?$$n{KjN;Y$yk*ZR#NN3%n;IzvtT3s#?kjsPHe?ifIH>ejpw5CG=+5vn9xI`57wO^3 z)RQ3adyx4Ycn-gh9?`XN2!qaqL=LbJC;4Sk+UAFd&~8M!lLJQ_YmX5yJmJL;o+ANG zWP$^l?1ljIJ?i^ddxsW*x3K;Xtjt!s^4g+f!Oz^LtKV4{4he^eh6Bsi6J^U3*@dxh z1#BroGEV9KDHs<(9TSJ6S4+jZ6a78PkiTp(VQpl^dKkMS7%#cdAr8m)Dae=BOE02# zx&2HX0NzbBl=IZ-U)05_uHG~P3sfrjNQ~S7(f%Lg0b<1H^UEsMi#Bd_v^dt}Iurov z27Wxl=h{dWLxW-}^x#|+(^Sn^ZJJ*VG2;J;+Td9Uvi?zcX4t+EVvB7hBQ6NBd^eJ1 ziT`?X^CkRzZ@4<$GBmO}y|Uk~PG#!nBU)lNNQ~p23Bob0aUZEpuU9ocj_vuDZci8v zK~kOWVS?4^7A8=gO8hf;vW_p|S^z@TJeBEtbY*cdeY;LqWm?WGVI669NcH~{=6|n^ zsfUN5wxKXL(wMLJg^giXQO_7t@NryY7K=)0eS3>B8>(=CiuayTe3S7dyf?@g^CA;8 zrjQA?4UDU%49FQGdy{XMib?U|OZ=Y0uRF)i(gFX2L_dA7WN z9x9c84Qs0}?WadXnN8hI;2p;Mj{GDX_S55y3pL6!>DDBzgU81mo<;mEmf?x`okI0X z3GL^P6RWd&)npA&Me}HS#7y`gi1||6T;FBd!}0Puw%`2tTuX}arS-o@+)HcQuPW#K z=^glWjA`J_=x+q}68)K?nJtIk4@_Fi4yq)uLmYc++c|iN3>=<3uC*h-8kD-C2xkRx zJ__pvyAl3@M82yIoTc|S=5QD=Nc~Xr42i^4d;KO~|6)faa3sk)B*p|@eaJIaWEh$a z*wF0(zPq3jtyYi0J3KyZ3!55NxH69XArHh1^+*91fw=p_QW%|XJ(C1pbvNI*5X*g0 zYC+emH5}N1b&0x^EQrybjn(ydN~@(wXVa*v;5kP+r{*A#`5t1F{ARf*ITQVzQS!%? zkxG7fSSz^#lrc&=3vWWV0!2R@DT=;+i&1ppH{c7Cq-T)r3HSmc*^G%W=z%b=s(p1s zhyQu>21c0-C^M!tTB_9}8aNkJK^wIk&a>Jm>A#azOB!(}dl6s4qyf4%ra+0~=jTaAxg`f7U?gD+bb+H<*W z&(PhG_ACu(dm6^KXANV_msUmFlNU_X;gO>4=}7Hq&(%!Oo>pQH66Yb|e}n<#OGso| zei6`@@CWLmzn_NhW7@PX2bq3EE5DWL2XuP1;Oq2!!ABbVA4C`1ynj>g(bR7#8%Ik$^Z))wyj{XXa1{p$bz-_Psw zap(Mgzh^(sdCq>0ZHi{2EHU9*348}b`^P1e13u+d4=vUaa~EF8>=)KQ7)q=0ipC3< zQeVY$@ep&p+D#%DWOa-8cK*Fsf`^Lc27Re$oJep>{VE_gonzX6C!cM)=7 zriAZ8(>%kkX*57T0eDcDbZSWv?dy0VN2vwij3NZTT27&`^l|c&`FVbj#q)}CyLA$1$Xy;U+L7yOh z%asl9F^DZ!*$fp@p@ZL8^9%tNfl}Ut5{t729CCZN5{tS;W9PV+S>PHf6^}gNbnzyr zg18qlq~#w$p~zK9A0k%}K9OtqMhdTv;6<2&@c96#$W@1;zW@{@=fP(K1AHNiC0HtE z;aj9Ne`rs$T--?KIgek`8Yu+~>un6v5L@c7ybwtA{qE5KSiJ?8ED)-Th3eb*4Tv7S zP@s}9JF5Oc(TWyQTOkePDA;!bV+vu`;Pms8;aKiorlnQ8~%n6j&lCY7!}lvF`-_ziE^rvxpI zw`F$ewMJ@P)p{1sSd`bb@u2U0Z7oh(vv^vfb~ACbl^%t4`7w;OgT zs;fz1RNd!Fr!WaB)x1kmEhH5wOhXZ7fCwXJurD>4T!Qt#V25iFSBJJUL!`cshl1~-@z%D)vlf43lT6S16a~ZB67$utXXyz$;G98@=9R&Y+WFi&(_IE_tAVU-7@}-!qa|# z^g1nBk%4_SCaex*>j8`TkG16SJ${+_cN)Cx|G>)^yhVcd7JivIoV}r!+(|Dsb zYP6#K^{Kz#rpP7_g9pE`jFri!2wS^w_5tkN5S)zr;4HT^BjYk+u>8lw)TOiRTIyEG zV2n2yV`5=g3s+Hi@_QFppv;Dt-GIiGKpYm;dDIhC51n?x*(%B+P@I}Q`&O)M z(-nBCh}(OO&V^51nh1BD^q(2#;mvk zuj!B_1}~&}*9^9Cs0(fA8w^QrM1eRD;{GE|k+UyP0{}VZP;(e;Aj~kqh&pO@FGm$f z`z_K2To=M2BRa-{H;1gc3~U1fYmEhKEu6rCa8|3`0PnL6{RksA>F^w!khvT};r`f{ zIO@dxyz1Wps56QV(aEkFRZ}GRcrMKsA+elAFj3E`@9$Yt>$qwIw?0(`ea~}G)YB(r zl&NvRhbhDguYjTF3<6)E6MHYIL(g6@x#%Tx?$+8S*BRKBF=6!_-9v!I{MRaa%S8Or zmgnBoJj(euN)yeamErN3fh~;*t9e|wnmn!&9+w*)FB`n#n0R`wY@6VDKBQDq(jPFu z{&FlUb~$_%EXr)r+-Sh&;S1upwNOGH957xrmQP)AlB0|{xpC}GD6UWa^F}P4AA$f5 zl?Hie4?{Qhslve_E&p1|LqO&%o>!`QyF2Q<*oRQD;rvVJ>9hT_lD}v0%bXZNX^`84 zVo7B!WGN6&ozqI^cqap&77Jbuvcdf)goPFsUCUVuIA&vvphcB+3qIRO2w@z5$5L_p zFJBBeex(Njj=z(w3kX^AdMwlNN{{VXGH5n&x+&Zo2l3ewKp!KwL@flW-{;=~FHZ%T zR^R|7tl%JFKs^|ff%eDOQnUp+8rXI*VYNSw04(-Den3kXqsI~hJ3l6@*5hCc>QO=T_-DEJ zKYqpk1uLv7($(a9Kp{X*G=Q!qiwr>PXaHSJ4iTm=2)jx0n}=Tp^7F58bid1xn|!Y3 ztTq*%!3!1L5;TX$GN>>boLIS#@y>}| z0P?Pr#&Kt0YsBWPFEQT^x|pVRkGHErve4MBPrT3fu@_zFm6e+}G1KCTe*kAI4U+E< z;(XKt&b#f!J>u=|Uh#HmA)^#q>=|2JjqJshWDl>ss7E}Hf+(&8Zob?^tQyGg^fH2r z#Z|qyOT0VRSJW%sds+TY-%0D1ioTsLY#%LCl?pJR?VfU z1Sq?2yXWr(ODxlH*i&iVM7zaVzn8t}AG>>EqB=62(V*;J&yPry%BugBv#8R4>{*#^ z0R9)Nl~!SCJkCE`2Inv>RMi7iOpt227zV}!t)8fzA&Y8YXD<137bIu0#Zci-&eEqM zsVa`Hhg`tMS_J^br@F+1j5Hkn`u`WB$#FDet{*2 zUPw9Anh!MxjPiW4R14F=fL z`|hxO?P|V$jVJG6QryYZYBuoV)dqpZ_!s^%G=wXqufRCb)c8MOJX_hg5!7c5>bRJw zy07&iP|^Qi!Wgj!zhs|HGPj}*shBJ+}TP0(2*vlDS6lG%u5s@B{(7Vmt3@$FSwY}F8d znJ0g|Sw~ao4pKQ8nB=twWvO*lZqc_Nh+i zF}+FB;|HjZbRttN)rllEp9v71BY&aQ9)wbR#(l{N&Y>K1!Yp~Yy1gGcs-Wc(ygxeB zVd)BCc*O!ivx}dB{hkQ1-{860ektk}%>&EAKzoxN5+OU#jRAl|%?1^mr{|}&3Lx3V zw`3PTnc!cTLwCk7KA`1P3^B8tQZf={T&c}`;3c|9j2b!70IZ7!&^7Y8D>Q}Iq5*V` z{4D@5|F?*R`?UO?z%T5AlYMZVwABVb-H<$s!Moez9IAUdd93P)lvir7EYbylOlR-= zj$%HIc0L9O*1V`6>}r`wrKtHPm8kw238MovUfq#jj}tDc8Z2imCqfAugafR8 zpQbI9^fFKaXA6hHT-C^Rh?-77q|()7bPf>+#0;OEZzOLs)%3xXJBr&)Qs_W`aueFS zOtnI*TMnf<#&AW>ecy(AaO@4AAvcQPz9gp6Q~x>%EA((E)m8lJPP;Ih6Xrm9;Pf&u z!;(2UW(u!xUWG%QJcGByOKTo_K2<+Ix%WX1NA%(7j4=riXs1Knd*}2!(Uz zh#P1JEUd$vguL;pFPJlxi*qM1nEEwlwR14MI*lF*PCR!;JuK{Tw$IY!-9S`#88H+e zy6KOq6JLT5?q7TdgdC_5HYY#?hdT16wot0La({H;S~zj4JD zj~YFvD&m>6aVN@-f0HbpQ~2aitr=Y18{i>dF?Eq$;b3;gL2JPwE(XIy2XU(a((9pS^kn+wVg7A(5&J%%i+64e zZT7*~sFKda349*QU*TCN+AoL}uQo?Hg)I}*(icP)&P2V1q{qbtZAIHWz3>OmUayJv~ zUPUB~GiXxz6k~CUtiJjcR>gVk^DsdBy@Z3t6gJ?b%ZS@Nm5Q(u=b>f$6wJ$IXZY1Vu`W8DMzM*xn zo)dm*&e8otd|EC_l=e%Rw5pHa9>XQc(M+b_pK$+Jv01i+p8 z&&Jari4R_eLP)2}+KHLEfbxl@`VwVr#*CN-dl+&!)YrlmT>b=UPkV4N+BAa<>rmr^ za@vULDSkHI2XqD{OD#=g&Wsf%*cWMwYawP<6%QA=B4OkZc z1Z64k2ZFQtOhv#ucVqwG>zd?9)dj^V8f2OG02`>85}Hx&aY5&m(u7iV0ngBXxrov= z*&dV`3s<3xopM0ar@)&Bw>}J3=)9qxgKTD*Rx)7FnM=UX5bnQy2{nbn4MA#+6w$Cg zg0{oyI; z#%t5gxxm7ReJ!ujL6EUjJg@lM^OYkz09D}qd<@a@)-XE50v~e zwH*MgRc|-MICK$^%v$&YgAye_ZDrYZ;%Zhlm7o-EFK)bCQ1gXExf+kZlqZaYt0%Z% zScK?ccXT3(IFBwG@O$-%~HYJ{i+sd z|AZ|?;DA_*~Jbg{inI6MXpKHCuda&jU&;{q?E z^iOS^c!AQ5kIzPTo39gfSNl7@#^&i<@dq?HxAO(6y<9aH#Y5!3wTG@KLssp`d!#zEzbXV5F-p}lAkIx9 zXJ;5)c-DCW>SCc#_v4z7ExYJ2TQ-yz{IC)M{Om>blGn8|f};bwdS!LI<`Lo;q=n!B zQB95DA$2ANf91Fce)={M!l9D1;DXw~2p)o>1#csQ=R!(hUaGdibt!l!yrT1eL{}#F zih!iAEs##@l#uVzC2=NW zuMe`a%JXor-^JG^(jbgYt$}KsZ7B<>J8k}qV1r`%HM&6&VmCG@e)bAvR<_s(;T+9m zpRJncF z!i^1vuMQBdn-*??-3XTm;dC>!=o2ki2!f6gErjQ7!m}%W{X!m&F0eY>or$NA4TIv1 zl+cj2ND8EGjP)_%g*3+6LA-Z<7u&`8#T&5xd;(J4=B1wu$3 zmRuYJdBl46Nuo?$98XB}l+H*Z@&q*4`Fc^IZ-Dma0?0P~U1N#M{g-x=va6d-DpM^G zxX7h*?L>@2mq84Xv=-pKD|EV|!A(-795+7FK^yzyr2$yh{nuSit7Tt157?xt1x4f8 zXl zYqVfTMogH0xuEWS(X>5pBOJ43Cf-YHG~NHE(Ep}||IHZvrd-vFNgbBHQ>eR0ot||} zyf93Pt3=jTfMZ|pvnT*(8PUl20tF+5{hh+6_J&IABqDoP$yZ+v0? zH11f35QhO^ab~c6TY?muzrEloulBNdd$0&L4<)oClu&iD{3w9vHaECoFk7ZjWIR}= zP<+o(3yw?de}X1px9v+w6(px)b|%up<3+hPMw11A^lM@*V2>kIh)MYBxkg&uSpOde z+8_e-UV>IMptpyCa(*k6YAQkX+yM4Jgn@2}0F5K)QW?R;`X5)Qtgl9Z{(!HZXAS5V zVW6Ba3Kd{ELHiOE`NGx7=6Ho2#msRuSZck}%i>uC0OOtm3dM`crUgf@<~C_?QzS%Sgfrd&vYt#44(mT4RCE+W$?sQz-Z&ok)7V^3 zQYW1{j;&3vBO=&il+LDtbgLk-w|)O4?d!SrV(dZ~8Pet6oQE%Hem!cag_m_~jDVo~ zEDS1T>Q=%)h5;&PqUIeA-Cm&I0W{8kF5*swd|EClf}p2AAKe3=Z6Sbswz>G7CZ3Bp zl3lbPR~_UOeT!8eWpNmgM~A@E9+Az#51S&NCS zL_!$FoWl!A!AStiL@;#UhbcYWV3+G%pe<-;<@GHpJ%?&~NE>mar(14)^r3sz_XpkQ z89CgxU3QGL$^{WEUs;J4wEk*JzzXDJol15rd;Hodfw7Ioesu@yoX-2Xml4QA`e0(`m_pi=+4+IdDcNGSxAi zNK(HrfgCs>e~|;V^O*zOEyUS0vt!@>>^3fXQH5z(c=8)vElG_-S_jU##FfCAwuX%K zZXT6l9!ls0m*au@bun1buPgB}A`}{At%W}TTD)xT1plT* z-4oKWK3!JDB&tLtq&+#FE`pMOtAzFl4z8yiYVeEHG4jJB4%JgWP`f$p{n~9A>DR8` zYOxhIIz0>gd@R$za*8)8FXon}$zQy~AuUhGAa5;L00``VE0jRyP}M{_>7z%4l;MKU zD(RJ{W7(+SN5njyW{zQktq_I7*D_o!6Ykamo=4|UEkMO{n&A>rn%<|ngPNTBfsP(E zKrY*n%g&}KQ1<+7N$T6jQE48B;n=+*^}zu1v=$l>*0L71MP=nsm-|6G8k^W*le&W8 z#J#@o3M@;QGah6va&lbHiPrkaZ_>(4={dN&!JP0(>dJOpXysE4@HZ-(FvAN%o)1OG zfegM_*92$vVz35;@=KGHRn&;aEPO_ghDbjSj6%vfW`r|-U_poPVv^9a_0IZ6m`@&~ zSdrQn+)W{_1wBtk7eKXMH>~pO)rKjp*Rw)^dUd}Z_HQO;Tf(pd!3X4I{1OzrFF<*~ zP$aj)ZQzD>kMvZKEng8E{x<-&o-MPL&g8MQ-2faW{}gOph?mkwS@0p`K!MC-{?9ZI zB`nInk6>vzs-C+GMg~MKUw(s^q?4{blN1EO1bDCs^q`chxW75J>r>rH56~&<4MHO- z4{D3}FqPk08(#Cs_zDAb*QvOIFOkBI~mzm8hOHsU$VRq;l2X;b`0ZMgKM3 z!{7adIaE?>;a$3^=1}L)(-=q9Y#@5qGc}E=MKs4Gq{uW;Jx#Keh~w6+T6NnkumE5l zzJZ<=VF_yiPY3ZF#}6sqOl-Mk6$pDi@N3eNT(1TRmTmqYP@#S)OOgn9Pf&04-XV_6u>)jfu7qFMoLN;MX2V@UozOGY~o zC5HmCeGw06drerOY$Sr(W&{N&?zTVZS(r6I!*R(uR@z_X9k*}^wrM>!Mf89lWb4fq zKc8oxXa|{cu!=j^?(TFe_H(Fpuz@_K^J%I7>1sZ`>3?d> zr)hk`Nb^Xas~MYT3F?bSDPLHZ8$+sat7;ICx*wZe-{Rei_76lw`=5o!=fX++Sg-O1 zn_A`Sc_9%wGLkA(ww_%cz;epTJFkNtGb&!)C3qOC$wtDv4CDV3*{b%bab)?%G;m{8 z>?0oT4-Vip0yrhBLwtPs{cXzOQx%0?LGlnFg_v99xCwIjYy-d`oN6slWeadSiUnzI zlHw6bLyx+jKf`LeT+NONqemY00S5cOpJ8M5PyAByM-AEzx&-MD6QdqIt)cp>$%cbnO)#k;YK%$cDwa1=%Ty1V)Bx4Tq>@w* zNkwwd<}3sUpY7cVH0ON$ij<86A#6lLkK*(b=<$T3+J8Ejr|=K_jcDC#ySEhh9zs_< za$0MxY`6z$W&b{YfZ!DXgL(MxU}_tACxSN64FY+O&_1WW7Dxtx zolLn+V|KkEurn;jF~)+RaOO`K0pmZcR|DfefXAVJ&P_bm=i+!N!hRU_7(RNx!8pIa zv-Z2cx3OvuhRQB4M@Qq#A3Bb7CL!co2N9ra$ZWo~a5)0CNH7KxNVMd6+?lvQ&c)hq zU~b8`)lJ8cfS;_EKQHyo|?T6I92Kp93uZ2?Dv%GDQxqMRcW zmF)Sw5{dh@rBMO?8Je&Je-q%8@zrQFJ;S^A4`$-IrKm<&)qQd9$45^ zF`J#>&(r15)qm(iR4~QpCuYiMTCO%B?1nC4+A%anz+F$NT_tQVtm{yXXVHkbLJZ3E zov(+~rVP|Z3`*u|3?qKXeMs)KI;mm?%mcEGJ@^!R)8CF0|AxY<2atu@qZBQ?D$=Hz zcrsv`H-YD~&B8C5v{|xlukF|37z4jzqqdo6;~AD1PPV8&9x+D083dSnhxH3J_Fc-0 zmI%}e1~nb!>j*R4PZJYHm(&yG%;Y*EOAY+OpsIkhA$`akp)Q_V1o+6<;}&iT+nflV zy>cDuhA}h_w{Df9hJUO*d&uKW4POcCvFmU+Kkiz2pEj`_8vtdn=`~iY(3NID7fchj zN3TGShp6Nm8vwz~AIw}#Vb~ly{;*CZs9%tx5ys;iU3jm2iLAD@%r!;kwvbsO?qUHA zi@4at2M0KEf;m&BCPxEcbqN5r8i2vk0N64BfcXHx{Qr~8)-Ce89>1uhIc_Oczfb*@ z+yD~ee(O0pP&71>r-w!;YvCvcB}S<`K$GybZy=se6h4(ry$+*IMZH%drD+xc}=3adq_wph@|kcVb|fJCt8Bly8hp z*(1de8jcXk(}gn5e~ATcD0VRv3uAG$7H0Tmy<8~X0*cxWeE?5szi1!J4!Rf|Jx2Yi z0|)cJMwST!Vri#~)*8T*=G`%2^cIqrh2|O;X?`HTOYlpqW-+le<@*g~--EHZKO(u7 zA(Zb2HFy`}riDQ&?(sR9{?cc{O-c?k51`K8%G3&`R?;C7)B>j8B&FF% z#f)^{{Z7mLGlpb*Y7D?&X*)bjTl(Sa7{oCC!cI(q45vHP)X!zZ1jm#tuB%|Mn+SzC zmWLO9P^KMfAV8@>e^3l+pc^Llfi;bP@qX<9VKBTHZjh1cAOOOo`uy4$QvLOrNVNid zE!M*O&})KZZxa%d&BIH-WK28MrvR5VMka1qT%Q1bw1F=PfoB?X{ebt*1u!2it}6iC zjleLpsR8sAQVumz3o=jOXr3m9C=KQpf@y3L11Qb2A5TH^TqnNY$piC{x;l8DHc&W} zF3$}B<+F7h&9awbMuk~4>Zq+1tT)rhEg*bR|XO+0v~!<3oy zGi%{^omXeuXmcj{ZIp`74t3S1$f*7R@!5Khf`WdMJMIl3xsDIQeua5WKTz}W;7KNV z`8ETOV4I}bR@Q7MlC2@?=X?^#SfGzH7CnbJRv_4CTQ{(HOZ;Ms2Xu)o@?$^*++k3G zdasj$JFWp>LxL79fg`ZVVmts?NYC5@X`c{&1rethwo;wWT@;|*0HVw61b#4cwKG&Z z%FMlXB%L7zzag{Se{DmXve!2aLavnh3KIfZeX@G@;b?~a*<);3vFV5YT^|dGZ9xx{rM~2!|$p=3!+w} zHhBK~Zf0iOff|>(YNuwk(5rTi3}e$CTa$2MKT?{e727sgM0q>l~VU&P0;_F&F$;N=A%J?hReWXgw0#5 zh;8#{e$h66wZFFci}*{0Ch;#O;1avaN(eP+0k1*wsmJ7fgi)PCy*Q2qd$zD&7T*IV z%JM&uV*8kwN#2%%=~MM$eK#h&{a2cB^oUECa2gu57+W4QlmQcUvdna5qEgy!FRGlZ zCtj?DhfuOvSQ9oSO7SXrL|-A3%9n|1!HQ};XZ^-(_r!QL6AuwpqCRkD>o3c35(TzZ zYox~k=iOy41wy174w}Nly4qz?cj7DZ!=auWq9sc~z_b?r%mNeEv#=I0sQ7Hth93{5 zUK{vwla1zxB9pEzEXTXvh)y9=c=#?e{M|Qv$D3?I;I{BxV)(mdd?(S4CQLPgCdD$L zOLkBt8S_2+aG3c9kb7k8WsOuVllF^CBamdEvGVlvkM4wWVc%BH#zR3$U}LU0DW-@fN}djISfu~JxR2*roj$~LQ=R^T>H0_;1HO+U=+B($Z~$$ZFR6YYbKG&e zfaN#M6X1df6)?@GUxu6J9`4pc1RLEehH0#Yg)k?~lk=~bXXAWho}nYfJa>dJsu{uP zp$JBQa0?+ARgJ~S^$?gG{YRMe_cN)inf#2g9q(&CTkep6#kbvMEdHyMs@N7kdnjOW z8&I+TB}kyaPhdU?AogUc+OuL9z?`bT#E|-&If#8w+Z>XE{UDj=UlufXTo0xv>vTHP zY}?>{GSiRibZw>|(PWEbb(GUVS21i z??IY_nTI9QW=L7^XFPoz?3n{QJ`Z<+e%x?}_mAHh%L0!Av*+KEGn?Kr+5hNcfpF0n zLCNot`QWQPav&We;E@}0JkO{32hZRgSH~{`z zrbltk*&SEH^l8R=cib|jl}^9L^p83{iRs-s?PPkZPCvx-TAj{idMVN-__Te{pVj{8 zy^yB$XS0E-9)Pk=cA~$e9b$ecB;;xV+WrjM%XCNvwfrq6yf=F^h# zsjnMr|HyVNYW*q3+8;APYcFO3-d_KqfR&0tRlNPZd)bjZr=E(*Mz>j3eW$JUCDwlo z5U;-j6y=H@Q18Ta3!ObzGu=q18!}x>r>|r>PKT44O!Iy!;Ql=s>Ek->VVd_k;r?fG z(SiTx+l-(UAgE`9G#+q1BfDviX&5bd4{!Kz4<9m^#v1g@LQRn$4-z6lU07;zxDRPv0F&c>hRkYCAl;Mb#`qmDK@Nxhj zcA1AvK5XCvn6AQK@APP~;*{@K1PWU{pT9@7C@*k3#0}xJJAD zBhW^Br7lH4*ZdQ7NPM;ikcd>#oO%Ix@?{C72A*hLix`N3 z(t+(Y&vanNG5nzeo9?`g9oTe#2ljdZgmjA2ahe&B<8`mq#*gqJ4!i|@2ovELWw?a` zM18~3YP%?N8-()Ny5bl5IMjU-eK`rt1>j~_gjSR_mo9XwQL;>@e(7DhQ=co6sA+R! z32~DMvF>X<5@J9H#RMHMpi=?rP`RM(eaUBAa<37r)EGD^reM~>xpZ)v(+Fn`7w*)X zv4qowrU^LuY+2xh@o)UH>u&_z5EBifnK0~cd%|>@9t+f3xKR8N;A;Sm^Y8Bo^?^t4 zqy=5tzkx5V^kiS{N*w>uP7oLwxrRLCNIN|;;{7hf2@+xn;JQ|1?EvO6ouCW)Yd%{y zNQL{4E#Eg<-8bMPzIakJE3K_o=lDqUH6Bi+A?STobem`Q}=uVV!6ixBO8#EWQ5&9;iL zy;|5NhfBLKCK|RiAfLbbD$FQP#RTmjpbG%%v+e93W{K>W2wJ2uf-n;ZIR7{ttxild z?Ws3~i*$5O45h7wU-&(>rf_K+PJd-AB$rnC$1lTFo*WZYFQobapg!9-{lZl48xuj( z&I{1Ss6v0$Iwt7lV$ME7dmL!v{ww(%A-@j%PRr_-)i1<`QmBu92LL``Ht6+&8seYo)z1 z_Bi9R5%&{vGR}{_7U%n7-8SD>_`;t5PX$+vtmpe!ru4CRc8PJzWdYLnrxtiV#&gd? z-#9XuiDnD#QF$*8XeB+1E@o%!7*jiYla|CcF(>0>j`h})obmDP7>B_9%)5Z(^om0{ zF7Bbo>L2W^l%e52M)vV`xTBh{0 z5F#K`PR1urY;5a_o!O!g7oYMBa14PHembyPid;eL8QioQ{V4*2H`7 z&$=(`{;UVG9>fCXGUdwD)oJ+FQZ#v1e$MRww$E`rdxSZotfJGXVSj)L{If?>7|Dz3 z;R@5?$r$hdJ#5}+EIzg^v7=(gY~c<{o1oUTN<=M_l6dgOA?uMFI5m-vK&;@3+g|Y+Z?zRz(ePbP(O;+DV^reDZr0T$Br*+Y zyL$r;3*+CkVQ|Wi=~P+1FHbFyeE4`E9&;*bwEPQ?Jds+Y+J22GnR9UO)GYU{zO4Kc6|-=#p!YL(x77R#Ri9{3;kdh7g#c70l5I4CrR~=_ zZtizL*4QbbNL2~Vg&g>W9h06|KHM9XXa0FB4zlQPu}+d3KMG66VLBjUm8Z4v9o&^e z|IZiyUleCqcLn6KXY96HyX=M`vWvc>%fnDnsKbB?_kSl2zmLm6gVfo*A%y$(3OJv| zRpt3=C;~O@g}a zN5szh(hd)WB&%2SwX7*|OVKRt zY^_~6-PO*x&f(kr~8&calZ||DU);6RH6;j%4>_)db zpOaBCCYGQ#XhClgK_7?hgkDegP3a;~1=^#Ng{`;LbZLr{AAM-$5TL)OR|gNU5*PQrZanRfPTYH9|QV z<<7P&T#QhEj)NOBRbkq3${f}(ov5G^oY?B+awww z!$#T>5_0X8vf{DfnY2+yoVQvhZDr~bv0yeSyE_$6qz(4Y|KG%d2Z8EPL*8aAc;p_I z$sN3rbx;e2a2(&g?{@(fix32|w93K&uX6@nGSr_ufwg^&YJKglxOpr8~`*Uv`U z5MPu$g0$FYc*$T z)(vHi+1Hr}D03a~*L$yjSJthlnD{lBUn6Tl*H9qfZ0Az+?c8y-!HR7}O}9y z?BvBQ7hJ%w7oTP61s@aC)>zPHJ;lXO*k)^*gO2#KTXo}Xe@J`Jx>}AxRnw=kR}tg% z#7%<#Gmci$phwl~IBtkX=puQ?Yv#B94*VO|doRl%NIuLl1f$Hm1>N2;+f46l-CkIF{|# z`JnGk>r#P3K|oo9@ddm4ngxj2b~o$ZqH;O9OV~4>B%zs#KPf)j8hvyL0&HW|V`){A zaS{y-2cP0d=?Nz9mE;ebI75$~t2!XF1UUaYN+wrFP^flDMo`BWni15NTApek>=}lx z17tggSn7FDr+;CZXW;p4{pGBvGPoGl_nP6WV_l(5nBuN+;#?KBEVp>~fY=-osH2G$ zFus5$Ru<=3p(-dzjv@(0TO%Zr~CMKOgK~=kmN`-y+PLgGxd#I*y^~YXqYbBw&8P^(I(^Y&ICd1 z^1lOcW`}D$`I65y;G_3MvjzJlf~G??dqd71^NhYZ5yy((!VYB@j3}G+z9`52)x5Uw z0(acU*ML5Qre=aKXtQyC4hBhK3m>6Cn(CPfeWA*^K&f@5MoHHwI1XDP(9vI!#|Y{> z;Yxvk@~q01A^u_(vkQD z;j4_Sgr|^q);YZb@fX?=P>3UL_%(;b3cG!z8xZ(;j%QX~N9in;S zGJR~oa9+laD=avrlY|2t4$>V7FIx4c)6iJ&=2iI z`O_9-jSdg0yd(p|Lb0*`O~0S^K^^z>XCp`oUdCX@@dW=xB& z4PdeUtFshep6$O47;I*4l)p#SN7KLng=c{3A##+-CXO^`+(57{NlBkV;W~&AZHi zXkE>lB-Mdz^1#NS9+Qw6VRLKYm-6v>yB2`zQUK|D>o77K^&SMkCDe@F&kGf^r6vfd z=MPiVr&m*Loo5~k)q_^Mp%rh!`^{(jA@exy8`B!xJW^|rq-Fq1?lI8{*;?ukc2ylI z&&OP#pzkmLr-dKo52V0UzwQluFA7AlggY5bzl$ z*wtRmqLFqBW29hwYA`;z6h_%c#HbX+m}M}gT?%8FVC=a$z+Nm}zS`8SrAct51(R$$Oc z64gRMC|7R?f*vLaMoHaBTO&9JZt~-dGB`v311Dc_76ouR8Jx8Lz_~+k#sEh=I6=)i z%{z<2T2@-3<^q{b$Us?n7pr}OPjmRB%fTmC(k9p+$iK0ueYcV}`KKeOu`2nq6W|bm zQ&C|k*9KO^5r@CH^bz!Pfk%Wa#)mJWpsWQiA;RJOFAyd@xhyFSLU1L>HS+hIDg-ku zSHWj)a8&i>b-Ai#J`cucSue!H%KqUGYF{4?E!$yjRgj#3(N~A6j1wQFC$Lvy_v^_g zA-L4hzrV~#iv}p#A*#o34fQ^-wmvn1tKcKGvKHnd+~Dhj`06=QzY&Pw{F3Gfi#+BA z1&XyoZ#;XKv$w!zRz2Ak<*FITg=WC@*_L-;1zoOM0+@U_cogYX4}CzZ7YbyVx{ypr z^h3hH5Y{>f*aHB=4RJS69}cVI+P}^msF$Nzwg5M$2(ug|DKZ?2t0!1}^Z;~r@y)oi zF61!4m8nv{In>W9*fze17C5zh#;{KhV~@tkdVpd5O9u1-JG(m5n9jlW#jqxHr`UUK z3kB<4Vj&D&CB!2kF;orp?gb><7Z;lO>!Fvy!TY+;HVoLNVev^j3eASatKWt-EYeZi z1tqSBhQ)->rD5@+lMRb6(00c77YLEXY+(^825K#+xoYu7pa-oLx3Uc z_oQ6hgD33hb(i-W@#F!^Z>buKfyKjiHjK~L>D#bfb zS2mWA6|V}V7;d)redhDI8=*TcrbYj532p9f^2){&mUU@R3bxIXt2@|Q#Hs9s%OFxW zdO5GWMH^D)oQih%Y|mX!1D*gF&a=S2X+UD&0pvXQ4Q_K#NmF$)JtKu<90L%Xp zQ?4m{u(=P5V%V4fshT+AHh4dHbjg4SNg{Grh%AM@N8|S5nTeS3XOiAEvnCQzr3$#e zE~C2fRF`m%6OL*PHVDD)vJ2-vp}F&+AbfWNVk9<+r&ye|FlLydk=h!hkPa}yG{`<9 zNN)qlIPzao`sm1C>9x&ORw1_Uc7S@z4$WNOJo>&d3)|qPt?rWH3pB~cp0Ih zUbj<0m^TM=L<2AKr4i*w(f%oNpi_>?)op5vPggS`g&#}k(ysqggvdXHujDil4QHQi zUm7;z(*nTM7Z^WG02nz0U8_HFDUCcI2-C#4RcmNbLwYqg2<5&21O$3O$eHTQ3Y>p@ zr4EQ8z`B3eN~rb(RiCXpX)}Ip0(9uj(eD$RLjMX)gEvR7YRw*IRVad+qbuX)=+@{? zN8TJg6+r0zFoedpW#EBdtXcvIi^pS%x=VlFH@6el0AMbii!25Zd@cIrD4!119O5(B zxQFAmND5F-UX-h)ms6jJ9hI(jLcK4jQ;YLg#DdXP{`G+I+4{Dp+PV0Z%wJQD+|YI& zo}A)P@8U_sP*!l4D)Cjmq8M)ZhPy+M|G|if`U1**_nxrFL>z#;%Ms%i1Ipt1kct%i zSli;oEyq8$(^2&FC_}oVPz-ZjE$I}H^VyaHp82>{;8B7uzbq>E(WM9b{(zUnUsa+Y zv?JXh6z?$`Y)q(eUI1Yr$R^ zu_uPpdKSAZJu}W(3z{HwvqpB`jtbE>tsdz#rZYIwScAD=Oq;9FOgtS1k z0i3yE%Cij=Tlp)1u9XizCzKupX4K@$nOx)FsYXc9d!k)VzIvUU)` zB5yIC*Mdwr+&WH>;RMSO(t@`bPXqv3NN8N+9!_auL&*~W5xr->gVJm6woWGjbGmOE zoz8>ZRFb4|x^DxW$W*m-B1y$D0l_acG}dlM!mt4pE9F9~;ZzLmnS-o_LM~`d^nQ;O z(tdFs1XjeQ%I_hcvF^J$+8uLs^Ih4cgrfxrZw~}PYhgzeOUdak7fK{z({BXjapSsJ zd;_0r85;DH-Hp>`h=CEAwfq|_ZGFIuNFxQQQEQFFCnC)=NL&aPf|O*CQU$47ibmoS zkp>tfwse7{PfKE4ejRZ+r+6T&0}~}=XB@K@Bnm7pigR~Ee9?u$#FzESkPPna{TBCM zGloG*&2V4K|77UDS@kIrU~50jbOi^Iy*^tBCRQA3oS{p7ao*U~|5w^KjYZeB6sN|d z(chYrW=5N2XG42&W452wz6Nv@rjPmXz4s z-xrF5$&|rl61g0dQ8_$~<m(C!L*Mmwg(`ZPN&qm`VvkMu=uXH+s=TCyXuQR zGdcXGBsIX3#Y@OgmY)1HyD(Rt~u_7{YB+ZP+>`X_+v?& zQ&&BLP7rS<$lJId5Ek7m{cJA^NRbzvAiu4OG@(0-h>*-6v0vd?HLPiiVF0#sTAB}I3JRI>9+dEwUq}!GhbxGdQMYGX=vh#PubMfK$_@Z9AZ{2cn60#4CY!tnq z_v=RFA9LxTHq0?kGh{+ZX6J7eT91%c%)Kl)Xv$UYw z)_o6aE!3!8&z0Zb1&9Zn8h9ZdnoL0VVrdU3ii6y9GLAZH zqZyXKr9`-Z)bqd<@C!c3(J<}mu+Yco7mR$988D+IN;C#5MyhH8h#6#u^8F39!K~#D zy%~**=z{r|B$abppfidmsXevHK@hhC(YwuOyO~twIFZKQwx%W>5^@re3mVxC@teuDX(a;?Qc$6QrS!ZH05#O zueXE@)&qhJ+8G9|{S00L1MeFKubaW^M7$|fwi4iEkWKQZ5WPyzy@r&s(hld;VfhQa zx`UMGj+ojvQCC97XxXc)%S86ZkRAJfMT1h$UkLC#EWva`F~V z&a4Xg)&m>&pCr@gB0Zv}nh97M27&>@bg1q=+Ej)==yXGTe!Sk;HcM`LQ zxn5nd$&GXL@WA)AP-*Y-+3qA`ko>C@HpGds!obO?L%1-M!R9ndG zogGcRnL38rE&6?&+Q(YY=4nQg{eoXij(hq-XZ-cfVrHF=)19uP+fh6_qhM6DUEOlzcj);`UizQh|vts zOa7;?_$0@W@Y@zX`Ok0*w?Pv%?h@P{5^hfjw~>Y$Zw#kChihnk_T#T`zvV6HJz;6EQhlh)Be0goAw9%^$ZapG5M$F{U>~%Ymf_xJX(;;)mT6MvU%DoXAUimhX%*94Vx;lR8#?Ac?ue=jdMYE``GVtDKyB#!rHamX9ys|sw3VLM*p8{Ul znopPV%3t1z&FwS1_t_3(6&%(R;FsL|UU^$JZSBtZE7pGZ%@D6_gSVkxnZx!dOxqQI zL%p)SdEbWb{a#sK)p(_@(+}g(j{em=Gz-*T^H9Qv?R-GJpQ<)UD&lfvS3ks7hdTQT z>uzkopJ~9;mD{9JRK7{!z96K$9|*>j%h}ce#vu6`L6-fX&#w@7cc|A$oW;Jk;BZno z)gX{P%4jlT#V-f8#kv-YJmEkDL7de6)w5CldPe1qS_%K5OWg} z-c7XczKqKSYbVGCn^1aW_QVXjW*sfLdhb6Vw+f_cqKkCie?ZO_$W4TVtozXm2y(rL zQpT%%F2$AU zW0yuc|1!013~+{mdNQ+oRMo&=lEoGuCB-2#_bu+d8>ewa@KNq`zc>it+4AxC_^i0^^(E1&g0P()`eOtbp# z{ps|iuNxtsVjb}*@+^1_bsa_uQT!~<&dkpQH8SgRW+LthuWmBL=@~*`E4|&sZI)P0z8ZPsL7hDE3T0MhojVLZ`SJqpf)Jvdc zcX!1?TFzB??nbl7XS>kQIG?Jdu}Y|b#+rwey3b&ZlyPM()vQd>3!TtvTA2o-7F2eq z&nIcUc82M7tq3@HniepgG{r(MT67MFK|3)(H3)J+7l+E3C3Y==EEZQDv69rzN~k7A zL40bv9uM9Rn92e%?;@ra;T=Ka)itpgSPMn~+4~wnYY=I#1|BJBn7$8{-CDq-H$6iQ zyB~d6N_vZaaH#iZ>gS0Vh7s87LT3b1~nU}<&1B~mG z0pxJdLz|*Nb)Nk50&Q@maA=5Mt}0K$a22<~Pl2lNgB%2gJS5e}Wgy~?`}1ptQr;Aj z;~triQ>0h3t2y|(ITdiYi#vgF8SA~_o@1r{eTJy8=vsUaxDd`~!y{@C3n*eVtko6( zF+im?AwDcT6C*Ux)S+HPA_hJ;jtNG4sPSK-U*sAD9J3Z%^uzdSM1K+mFI@CeS-ub< zIVg-@tg$jI zjq{Q57^SHT8`@!)+l!_2hmwo)lRGOmKdMe|2qSj{kJ03G-us>x{})M{;Fr}~T=ja8 z>YQ?4jH6T`Ye^7e*UFSW0zp_@(f5k21Tm2f@bs@RXg&a8`uZ?e``$NMtm zF_PE)rV)9_7^kV#Ag=mxJ?yeU?m4qK@1Tl*qt%Kxx+a#27p+W}sEE24ZH+ur^MM;VFKyBH0>Zz6vF;Kx*iBe(3xKm2tHd22GAb6Ba)c?34W0-B*Y$|`u(eODqi(O%Hslx$@wd5 zNX`?PIsarLb6u0NeS`DLgSqZ*(i#nAySpGc6>%MYkDcaBAz6@|gapT{F{*}?&vru% z8XFS~$8*9WyOeg5{WlZY}2b!G4K@OhtFdxLkZ4pzGV#uj0u?~}&j}L&WDKGJA z&);F?X{jMKS!`4jiD#xzL{j){kIpTm6Z!a;?{!ISVT`jUlPu}|2ag�&vhWS2nT3%4d8Rw*^jN0b z>hvI{Wt)k+NpGeb;=RxIQ>{Rj_oX<@a#vMk`3Pj7_lTJ-%FZu|FB+4aBNHubvtR8O z9jlZBb)xA>48+Z7KHC%!-qBf4Y3V;*i+W%WJ+Y6SH;RX0>*&-1*x^%Q<_Y^BKvD=` z-G|$FRVam!-EoHtBB|KjcgACu&5n7XaPszI1a+2Hv_@q88wNne8%~I%D_$Tc?ge~x zBj-=qMr{*Q{svOd4UXdMvproCw$Sr>+5p~ju?(Q+`I^Td&@o5=_g~;aSfxo<8D1Z5 zgxOSv`FJ2d(g@+2hl+$%uf;Jb0}9D4pxUb*9x2Y86)tvDmKXmRsP*_WT`~v4uBcdNq>j?&^`8jM5FO_?15M#s91{ z*K2K=}%ZAVU-O{(t@0PE`B&kE~ zSI^Rzrx34vw&VB>SKH*q)}k0L;JQv^dJzP)9`Qcg zV|Wki%+#yd4;RupK50U+qhIg@fH|}CfKW@D{F!*m?gHAQT(^iZzeSdFeYOk& z!N23x@>5{Q;(Y_~(FWX)vWNCJo8zlKVEY`T$Cv3YM_8BLl_^HsHc|I_9i-#NTsE3C z-uFRK_2C=g(L+)N;e!Ou5}$2lJjKurPcE9lDfxFmGVe}1JMUQ17>luUvUeMsDpNw+ ze%8XD7^mEwx{t)rV)10GaKX{9Y#8xita9~7nY>AmHyDE9F?$E`j;R6ay+5_6>1rNQ ztjDJut1H6WQPhr$Vb}$p}ua@gr<>=%BEl1mX1-?`k6QM>JQVKVgAy14&fTCnP{|3e)@DMT<0mC`K-!n|4IjZv5+lP87%#vR` zI#W~B_e|lng3;9Hcwk)E%9jweaeO`OBWG(64^vNS*XXYbk@6hIYZ3M`2uoFvJ^%hL zsz|0-3-$maB8xiy2*qT-Xr$Y^xUqA_2)OM20%Vr$tvHkWZ$A)f#5X|Gvl=S-Y!d-N zxz3eCn!6|@%1=DB<64tCG!Ny_>v~WA^B!b0yKcCDmtO+A@V}O?ro+j7BoXVrIqZqo zR+h!w;06<(Z^Kvds=&@ZRoCRq09a}ejBeBmGE$-;;2UVxSqr(>O=3sUc=$eYf#Rzc z0(>c9n@0#+hj{EX1v!#a9!VFLSPMFd>tpRq@Vlp#yh|IgT_WS^iHwZjpg)51KPs~# zOUAEs`-*=+-Xb95mJ=-H!UL6Qfsh(8sX`{kd%vX$LPkZAK?2R*@+w0GE z{~LSpWxiZ{c>RTG05qvOJjGCgM33i^KqRz+!@_hBZwLUNty3i`dn10KmP5Vyy(pUp zfgEZ&9^~=?e9Cbz(=VSUgzPy{SMj74_r??f9!TJyq)3PILfwG~vY*cSbwy!!M&SyiUIZ z2#Pbe8u-`6uQ>mne{nZxzKtjLvnDia0wFf&Q^pgn=b$Rv^MjA+K#`*ElT?A+3mI^G zSGI`y>R34^oPKI?)y;Nu4;Po5;XFBq8pLzmunzMzhw2AAWM>@51)K0{duaJ_*2DZ) z1+gnv;j0r~R2=(oLKeDNFf<&NCPb-E0v12!`N&F>F zootvy9%S=Eb>I=K#DMJ%1(OE>`}M2_n`gjQXfP}Khnh{E`7lj?&}`&E@D~U^8Vj_{ zNXTI=d_gdg_k5L0(ZpD}TAwinA&Fm8@zpivl7{V0*k`vgK6@;TQdm34WrIPRHU|V& zrH}QrN64oEM1GO18a~OTBFw3gn)%r?csBr)_zpwQ+uV&ec4iJ=V10hPcOkbee=#`Q`WP-}<=*8j+`CLHoaEk*gT$!c8?Y+C zF)6u1q%B_)+2e$2KHGyhEy~<}=aR^&xkj8Gfv~&xGz$V-e@m6Frou+#T?3`O{^sPu z_FaDgmOGW4_jCUUi>!KrSpL=ZC@fxmrd9dRXbhlrX_i)xe22;&RYNvtD~Ivm;)vxJ zj)4rq%hTW_CWN-Zd^m(-ZQ>Z+AZez0I}M=+IT|&aeG8C%wnspVaggVu^z+rVc+qV{ zUMcV4xIseTFdsr~sX_<0glU|{nFcaKirLh5QYvE!LL+81N z2$uHVN|DE*zO57q>guB^=YT z4rS}oU9`7HQ|RsgkY_0;1lBVd4jHG4pF=RU&BG_3ZMMMbnf)86U^pLGRn zx_zS#p1=ppCTNJk00XjS0bGHk%@?HNw>7Bn!2vnN`08z9Di1~m^$eD{_p1X9Hj>%G{Ez!>Al z98N?vmlW92zp+9@i0tG>+k7k>DtJ|HM5^d;p0x%+sjnN0-F5B+IxLJvAa4JtpUNAH z{S*-494&bJe-w-TVIj~71paTa*w413&?FYSVuSgA9E&}^SZniI7o_`=KBi(-GIqi&|-(v7*$4e0z8H|W6rx!7N2n;ib zVKguR&L=!n!@~#<)Np^o&uh3F;cgm!jBqkwQ_Ck#ma1jjs1c|-UDJz@w4|nu9QW>L zEtJS{U(+Dsu;nles-^J9SH#s;sUm(JJR>hDwdUgwO11^@`yDQ=OpS5Wt;nZXG8qxY zN{-6(*grbyQ>=SCi(-Ak%&3dsa1)zRtc|>s%}=pv(@%mVVN#?3gJ{dG4=$ishjK-+ zIyLlHtjid*gTuNYB0Kj0hZAm(8Up+eme%D6yHSWQ_5~=!$`eos923$Cv60Cc8W>p@ z6J}67z|#p|*22gq%zKyo~z+Z+-;PWs_HBxeyDAlM3 zgDTatKxU1@A0yQih((vtsd8W#UeH9vN*02zL8}$KcDQ##J@1m%@ayMzmB5h*U+X zG=JV(t&_a`E!ka`ztzO+iXAsX_^SZ^ppRIzmN{1FZm#p6hnyjiZvngC12Bs~Hh_uz zIvBMHcobnBpgYt>c(u+ogzzE__an>`8c5%j@FyA`KzNLXyAj503=^YKr=a?-&<=ca zlcv&)aHxiP*@bmUQubs@Ru-iB+b$Us;3fhEMJ`nv=IXl87xcUb;ku)c(POX0>^PG) z7|5*kSv6%wjSK77>)BnHZ#br>r0bt&KdoThx?tWRtGje$#*?74a`%4}XT|N^d+e8|O-`Lw|&ZJhLvbGdXc&7iV=+18Z__KP(jvH#)5(>c>Fs z588H^B0+!>R~vC&F(62a{`4(A!4DVX$mVFT{*;J z4MDM5d5Z-=J{{DbBJnB42m^q8+902>-$_drvS4Ivy3OYB%RB4DH+Pz9$;c}3+*dQ< zRnnAG@&=8p@(MoTo`Yfuih^UmR!_XhH~y&|MN|>A;+FtOL%YEZYa-F5({B}LA)+vZ z!Q&c(#T{Gj7>EMvynkuCaA615^M>@p`X;T!22NGr%{z#60?Tsycd-dd1<<4NM7)>=a|VWGx*t(*Ms4)gYMxYKFMy z5J6@sPi7*jN8M%k2*)QKp9Rdl3MM%5t0*i4=F{64rl*qd28-rpHU6@bhjDsvsvY1( z)qq5o6UoA@xnz z`0sUn+j2TUkr4N|5IWV0I)$Omc#H{gEd{k6v#+*`; zyRkM2@wuJ(ycsQ=o8sw-e?3c@^8jow)FJ+)Bjdsm@a#f^_9+&e^%oTfqLk3XVS4GI z34U1xKTX%p@zPlk^%Ax6LQjrxxc7M`LZC4BRVX5xZN@R_sNLTf8E?qfO(9h+XA1!f zrI$1w6?}H$lQt2GnsEOL3Bh`q8V9`<`>2C~dEqO(hB=tW9-@ij3N#pUcYZ$I)&ALs zne5oo=PSJ}_Y(*M5ki52a>zqKBJz-`8hcxq`u1lweX~R|TOQzIR?`5n&L^5Tnd-o2 zruJtZ)n3<9q|9ZNje1a*YmhbZd{dH#_Ue+AN^i?ah9Y|G(MLs?2ndNh?hnabxEi;G z#8Gd1A*noRKXve#Y_1V=y;!3C`&Oxjg9I~F7S_0XI|kh985drU10JiXZpCaBH;<`l zE(6!Cq&%N*<#&+JROhCH>;fS>`yk1}$KlA!#A;(gKg87MINa}Yg7Q*4_6GZY)mAh? zZ8K1qnzb8JqZF>elIS%YL5Ij$7|<6wve$v$9?(ttSz2N#f7U8Uzo>nL5X z>+m{F9mbqSORP83#^U8tNHt_7|2XPzaP&VJom}o}a{N_W&O=0$eG){A)w&;{$rjAJ z@HBv{g!_H5!R^4MjcNeV4np)55XC|_vig+Qvzv^R;j*S=Xn_nYTzgP$yhcnpm+Q>7 zUv}-SX=Oh$BZ;<$Zw|4>$%98Zf64oSXv7ZCmkP;Tsef|o5~ zExw{J6zkK^^h7jP>yse`>f5&a@*Np>*_x@xq2wh)h(f(UZ8-XOkXI(K9foeCC)V-KmZsg$|4K|p6 zNz52=xb29V9;uc=zY%txb)F&#{QKMtgvmkccR2T}u}G(@do&=?7Dx7JTz81$C=}q@ z!=a9H2995sk;u3Gd~E4^C_e8>i(BeL4IHh&{HXC7_kuL__#d!1?|JMq4}$S*cO>in zb~5-S7ICha?Xt)Bpd7APYl_TF?~SQ}GIGaox2bp`-A+XW=5aoT^`jer*>|RQ5iU;p zcS{3ZE}}ni#cl)wb~|nH7~qjUTqn3k`r$6xBi#utjal2_m+ie3e_f$b!>;F%e(EiM zkM#ZMOKAJ6G8@!7*kVB2r`SD)em%aFM|#f>L2~$cq?h4)iAQ>__kk=<3zpM(r1x~8 zHeNe`YWgpW0T5YBT=(?vWd7%Qk@=HAK*4RpV}Q(GjSMdHO`WvN3pZ<-|AAi?XbXQ` zp$mpwPv(a*{bjxkdM5oXvt38#+d7NPTa;b}=o>7DpUm6hdx^~By$@uuL9m=g=6ylS zV}E!5|CISpL;tPJUp^-?F9rh2ycZq=WPYVqaG7^{Qp>#mCN1+m_+^2f;jb&S`CHeM z`749{W!@IO7G)mAY}b)_uTCQKn}8BYb&D^r&Tb5n!%ybh@V!Ll8@vx>acQueM&@Uq zq|9%nEf1#W`3U=M%bkv144yIAF0LQMtitgClwDtVbEOhvsbG%tWLP8xBKEQkM1mb8 z2rKqoRisUF9$Z#m@>GCZF*Z{ckI~tEdZ{vm(3!z1InqX?JnR6eKUd^h9!v|~m!1Y* zNNO{BL7%yO4eV~k8?H)QncxhgQyYG`yyImM-9%4tC;i3ZiU3@(4N2OS8Ln!`C?6@R z06jK@Giwgxes0l*DX7%3;0}zuhLl5D-12T|er0kNEJ&|GTIARu)0whx^Pqaww z4(37KR~0S-(%5n~A`8y{vM8>V95o}vZ&W2d(C?()2t`p`!*KuQde%a*%D^vKh*13n z!i?`Z0w{PScB&xjs)NWVMd`a0qlq47WLt{oyf#eXNeR=ZtPo=xPCc_Vk-I}1ASat; zMK6tO>%3&o5dtI4=hIJnQ60gvRTo6S9WthH5)UQZNcR-S2Y1LD%-a^@4IDWbQ}c>$ z@gI$@UW?mBaLARXnbmiAbnnZT{3E^j_jJ^5fuTYd>k0G1fZzhhHbpo`w3FK!;02Q# zw`SoRg_EftKbFUUJw3ME(O&rs2WlUx!#k)>IeKy-Taf7oGdp=NGA4BZrVKr|iw71_ zL$&*REu=8D1(2x+h~U7lTBQ^Sm#CC2Clf-0|92P2LWfSw)jAZAs1wq-cl#A;J=0OJ zvwsIxxb4D2(z}@bVF!q)N!wbe0&FHgn&*R2&UhF5{5kkX;x~XzF0zBfY&mN%WLJIg zO16EIXp9=)6_v7F7x3@3R6>tBjbH0k&M7#U%O)d%!m;IW{X~vgGRG?9AiW!boFdOd zeS7Q!x1R`cXKCD%nF18U)L3F)h1^iDG&Kwglcp30jL5&S0L)b(<^NVwxhHKo?S;K! z6;-uXh$R~?jAGbw7$8XvC8lV>L`PQ567n~be9Mv#74XOs{IdqH<53%kE<-L);2Coa z1jh?e37`R(GbZ)mV*)P#oTlpKvX@&vP#!LqN5$lQo{ER$ygA+Q(?0%0dNN~O``Ht0 zuMKk_(=a+5+1|% zdCMJy%j-0ega=C+`b~aWt*g_a))^>4eJp+ca-OGD?KSO8m|gY68De8C+{P4u5dX>K zic#5svW9dENw2oCRX6ZHBu)+FM;GpPoA4UAW-##g2|y)E?$97*{R z6e9q#0#~lm@`zT809m);HT)l-)TZ&tB+oC2*+-+>mpj@%{-`bI3(WtLFj;4|TbHw1 z6IY_9<@OEk#~efiTt`P%3M)dLUZ#CItl7~)>d{0?bO3eh=YJf+!yXo{~vhF!$KH0E5nV5hjBO3b!u^nq6lBXb1KAJ(fBPj^70oou64*CxFJ94mClzs_S9=RI*G*mVAdxZ+cgqZh`@WsjUKwLiGw)A1u{S zjJgPQzPU0*o2J}d3IG%*d49I#UP4A(^a_?_0}{St7X#YKwc0vu z_CEk{ZRfHW6b&Qi<+6VTx_hMNToeZ#eX$z;5Q54%H!K4xCVAym46fM>)*?w#UIiR_ zFOpk;&^R`#X^e?OEW!m*XS@WS3`QX1bI2v;IdSAnuf(A%B-nl@ zSa*fSgPI!+AhbS#Q2=fNAe+Pidmv1?;>?tH2U@CEF4(#ua>*{WbpzXlxg-LW0~I=q zF`N2BEH0*EQY|MAfno%~0M}2Jt7WS>v-om~hZ8K1s zxpJTpK9Sy1;00l(z&&DS>uqLGF}9;KXnP{R zaEO9U?I0WDYFa``;67x{n8{I4=!#d6J(Ahs3?W<(_6YE&gzqHmyuwLDqs7z-eGE@M zso;=bFUxCsGy0i!kr0L(JEx&)?kyN}y8i-_*NJ2&WhhjwI$*&1b%qdL@-7SBFC2O{ zO-4z&N+8jtkZ4);%Y6rUl3h9#_4R@qc*R7NXW@2{9oA8r>H}`9ZFu(T|3;twpT-^W zqW_1z`oGhw{{~;GTC^mX{^y2>GQ>`KTWak6ZdJ92Wt6W8AAcN#BeYyl4c4n(o=mko z4a8C{Z-OK134V2j#^RW#HMK-YxNp>@Pd=;uRWLBFj8tjr*-ac{FUMaicM9aE>AgjI z7M49tLHBLouBGNt*y6Nda9gN8eF*ji*aK6tzX{vgfyC?ppov;b*_6@FTcn@6oAU-& z=!OAh)xYLje(JyQSfKh3k*Ho=%If<~P+=~T!pA1`5@CKFYSh2yHeFRQiA!N^pi^s; zow#u09gK#pk8*UkcMC9D5!xxCCVY*DD(2xbl+?P$olLr4I*w;*{~#nXy;2&}o1Jg$ zy1L?-(hOJa0bw2o&01u1T}Dxybu9%-b@Ee8N9JkdDxZ|bn}?f;a}*w|`#vX&#p<;{ z3|Yi5hZx=fM4k>{{vkI08wV<}B~b|*8}S~q3v}-d1PNuhkZJKXLd}v~ZmR{;Bh5rS zTQ``8D~s@OUUvD>3Sj&TeXxEoUtea@KTEz^o6Y+T$mO1HB({_&jx&)fblyu)%nI5Z zOeXxdG=NYi;S(CpB+SFlNbe%NUBhn@-mKxFgjW!z)~b63U~7}hsQBh1Y=${HkP#2) zzC^CDVx5KB_s#6BPMTTPm>n!qfK%p!)qJpOl)vzMUn{Q+U9ik+rkf8c#R7q6|`Y1SU$9~=KXh(AW0FM@`}Q`$6G zKflC)ra4jJoQHVb7pcBN^4m0w59FZo9?Bi=jR@a&bF30FH0GoVu!}jl7e%-JzCg09 zc&>dFs$ug$Tg0v#!^?brN)i=Pjy?H&9?Nrv*Y=0_b(eyARCzqGz*pvK3Dap-;#Dz+pJuk;U=qHuok-;V&0}A+QSV~dD4FnK zWzgI@lv)~cQwJ)s%_8$V$ZS@GcCC?$dgNW&Cw=hiw%&6@@TK`^iR$fC8(&LmVJTUOE+c2Tv&H_N+=t{8y=m`?BSu}@eh>c9p(c{Uj4 z)^jkd#^8YzE}B+Tss#fRw6UEL|9Ewt!PLX%;3*qCd2u6 zNp?|n#xLJQrm61|S*d7H&n2-&jiEVgMmTZmhqF2<;?aCXC?{$&JPSTs(pYM+dpvCk zNn8?LMN>SiPKml+RLA5MrxKvW0Ki?o*V@<<7~Ofe{8-AJioe#Xb-ErN6=L8}#RVm+ z7bgzI(FoiKbV<(LiA%$w6}(Fa9?@;LLAsm1$LQ_%^}xuos$SA;$Yf_ga!deLqgd^n zMM)N`U*#9~(MmehJhZH&ptjs8z%GRy0IJMaeY*5a>PGyMLWKHfnufyE2;dTx_s3yY zu^I^1DQmS5x&=+CL)zV7D9!LTpi%H9e4!Z z|J&5;kHSLSh=+>(s9)+*I}Fd^>JbC=Qx6I>EWADMlZ^Bo`1X z*)<8Cz%q=rF9+k|^JF{9macW$+BVZ7mWh1~-$3F;u42RuF)`s(LY(08BGaA&|W zePmSA`p3++478WcOLPygt5m^t0NJBW>9X9s+-$RvZg^s_^tPPuQFVxayP6V*GyVee zMB|;b$>))W)xYHnt?A(^7OcWO*rxFj33bSi3rjxE0;JD9z%o~nHji#A5hz*Tr5S~8 z_eOz<;59*Dnpe&wtQRa`E&M3jjkrB1pe!%&2#>1pA%>~10bFn^6Syln4dlXp6wzo&>N2o}U+3W`M^#fTTWJ(1UA?~1?B(In?C2HpHssMM`N z4&IY>MXTX)nC(F)bVu+Nv>(#K-nB?QCK;$cl2_>Lt~QG#|5&Mev_L8?QdKrg#Q6_3 zjNM)Q01LlsImKV34t|CL{0Qo*E%3QNF|n2Jqy?dCcvTF$-tke>C_7i|VIA%h4Nh$w zb6d!=Y9i9z$g`%EZCc_5u{2cir*>}FdXCBYOTl( zO}K-qW306CdytnB>IyTX+9JOmwF$q7Y===O-+qTDY)ItBnlDh}*ggSq zm_0Rc!lQrr@|bbLUprEpS>TyZ%DXYywOG96-d#2LAiHXkF@{PcWHTq;NfZs2o>eu3nKVUf0 z)Yo|7=ra7-anAniUD5f2QV9IUVsIH zwu-*=NXAG{p>jFLSCC(LX~t;9GB<$Wq|6Nz)F`Z9=p%Iy@}Q~cgr~BPPJ{b|qWvp5 z{z`fWkj1{wX!SjsiTh-ly0j4A5bY`~{Z8&zkexGMun9>#n`k=9!`uixQyJ9*okIX{ zAQRqqHso;VEC$i?zJ~Yxm&9m}1-vom^`|Z;}55|(|kp$dktT` zg|8^TJ8Z*CxQcW*d510H340NL=tZ0;hTSIt_akf zufgc{@JR!m#i*qQ zidORs6scwyC_+s#P`JuBP?#D?h$UMA8Lf%{>!Q@}&8AlU$$>c=O1T9DpnG%g)poa| z&c=g6tpTz;&E-`g&VOlTnH5A?UO{&3e^{iIr80A32Rx=!IR8Se(?^fa8U+kf%0SWT zfPo^_b^}GI9|6&{eUD#peL&wq{~sD3{}aj_!k~m_>XWz~#PfXtYRo?Pyy0Pv`hPp( zx4w>**%lq&r%|m`kp(I5aJ36ha&C3F8r-(bM(Y>9IlY1-uM_5M#|PDFj{$alN4dX~ zvi~lZapSwUBDM~~?xKR%{xt(pcARH8PQx&#Q2n!6qjUC0ke%od?4w`j)z8Cd!;ZWH z3BigJgj^`aq

fKgoEXg=dGWXICs*4$If@IG4 zZnoE9S&zKCVpqB7vPO$X>*||K)MXGBXoOQW!mDZi2%~^db94R^aN{y-`q&dhrBGRa zl0~kk@8lam;p%6B;4xb#Kma-H+NZu=t%vSBcdu=9E{HF!MGoTEIc$&a!dHi9lg&Am zugdV6a{#a5|69QJ${-$6^fZ+<2fbS_@mZ~t`!F@niQvE*=oPnmAQ(fo9Ikk}XO|(C zF##HPkHjvTP(c?CHMUNjm;_^&O* z>WB%&U5nUI6E+WwOXYq6ER0bn8JuPjek(?7=0vfmC1y&C7Dw%b#mN1u#EfD@o z<+A1eN^!9aC+}eyZbBKbzCwZ{K?crg;a0s#nnXGAE35<0!*@Cx``|5Bm{>&BRZ{RZ zW_LI1NmZmu6eJ)&t6^kW@@57#NJxLgn$dWx5SL&r2gqtJY9$oUOa#RvCfN}_`~i)y zx*Emnj$e;G8JN;kD64=hPkEp=Ha?!@>2n6Xkg__vUo)P7U6X`Xh~ZL<&q6d!hz5=J zK0#-^Hnrq|=ihEI5bl3AP(Kx7pj35Xg2vrVoiI>}I%FV++69Px%~ta!Ol_1mDF3kP z?0tvQEjtKca^4KRuLI=?Pl{h}%lQ^Mz^y6_QkE)XK}7tr<-7>1hX9@Uzw)4fbLoa^ zA76}hQvlpNndIsMWB*4wW2hPj-GI;0BoU|7y_3V4M8npL$-GJUqH1i|Llos>)ihfP zxM{$EG3tZ!n4~|AN@V5=CUXOm`5c>t3)5IoC!2BZ#cYvW&Gn5umqQIW~6esR`Q7{D`bXE$yKPN7A(Wdo`N1$gw!6X??9brQz4 z-I7r1e*BZl0I69>L-%N9t-)ufL-Ls7z*xOom2;+|RilvtYK#v$WM&vBIr3Jr9vyk# zDE`PULT)@?B&k+z#>?cqxohx)s!p$Soodb{F;?>);A3Lo+gb`=AB}H;#>X+wf5R8+ zkB^Cgk6PyA?`Dl}qQ;m1Kkyx$=Eols1K+e#_|_o>cpu1@uG}pC=nxkg7H{*XI!%gB zlj@g7U`aDXr-{*N68zE#ENQyxG~qf;q+c3=B~25ZhC>|i6XKUfU`b=sX}0P#2l>Om zv^Ozv-pU2~^&0(}_AXao$-4bhDQ><_Gfk(l<(?prH>{t-)h5&HYDau9y4pXJ9K#UF zw+$~gvE_0@4bmbOPMbJy+Blu|4V^X&pCTmfN~T2|433*I)4^p7NMX(3<`CLstCGy_ zMM1HsgL}{T`4B&&4aBHQo<*zP%=IQA9~Nm#_5r&>Izc5?{r@CZVH;aYI*%mre4WTu`4|Jt-FhZ;gjUGK*ntHkx z=iV)tNDdohbmVypM6&iHdUc9M;qY0&rxAC9(>j=f7?t0hRZ&LvhJ6rq1}nm=MmGUD-eDIQhZ4_HGh z7XZc_zCi(s?os9bVmANi;fR)kqP_}l(o_rz<6cdZ{MAm$yoQLb<^yJ;CZACfD<|S* zNcszC(4zr?1Yh7v!pW3`AL{fJk3CV?EmGeB2jq@zQI$C~(HzboVlRgk@4;>zjvv_b zyi|;Cs3sopS$OD<7w+lMTevzzI+ef+QzkIwU!p=SR;)hB)t2;7HE9S=inv;Qf){CO z40mbD(lh)js2EAkLax{GSGcHoJ6LhghZmxdqzW%YS&ppfhZzGu4nIrDtNJs0T%aCkG4QqVO}j?X6|f6Fg~$?b zQy;NrXhTvnO+MkodbBuglK3UQK~(yEj=Z}N*r`zc{XUg|`QfUdUr4yfW;&8=gd*VJ zfMjmNc^7$n!Ta?l`MPZqWdDrRFz!DCKGwlB`7Krxi64zRT)l4~+;<=l1bXLo5-(J* z~0tM=+OMWG|2whQ#2b323n>=SSX*O@eQe7$2X~#IXJ^b_v6~_>Ji} z(q17dD^oMH3GpbwPr$-NY#Eerc`^F|Y}s;N zz;|;2tVi_%TV;^jauS*5PW(Y=c^Pj5-C}aeA|(mS&>$29gY=Js2-=o=0Id#=V{hic zlBWzWH2UH%rTKXP?SZnn*R+cm6=|RZwFE6$h_@famXrfn(hSK)8W*B-iN~;2gnB#= z>go#ZGm6@K(fc}0<*a8A0N`)${3qGjA6}qVVeUyB`zn&{gZfS{k2yYENUqSinJhl~ zSwmY7W=B0zz?PM!nrqsg5klK6zqE=}zrHK9i$?0)k%oF?nz|9xlky%Dk2|c$fb!@| z+bHmXv$xzH?(AGXLU<|RgzltP;j)JSW)J>$;?8cdX9k8%Cd5O%DWp@XgGJP8$Y%1b zGy#R`5#-^PsU@ExAxy&k#DNU`d^SVRS^6V`$36zh&<{KYYv_pYL63bhWkBsc=cLX0 z8wH9hv7z0*WRpSL$wUSjtCz%5;-LP(ho(hUCpU;}4dE85E9pGQ2pRruXZ_LY#ck1C zMF(=V14%sgOH+lbuW9mtrJ0UAiMuC&AEEWX;NS852$mz#A4=xPhz%`KSU&A#If6m+ zJ?g$GH4xE9*j~#_Z1OLIPMUH8lf3^5?*odn3#Xew&JBknPk$VnT!YG`4vZ$}=-Hr} zKIT!sKpG)PuNeDLIz?670}X0e|3cZY{)N9nxe+;n$g%2|j_GjyYjApu2ib9@UZS7R zIA#s?8p6v$f(9Ya7h?pQv}H_W%hk7Uf}w#-hrwK=3}6^04V0joO+le-Av&O1XHvBB zf_^|$Bm>5DMb*KD5vPuff@YmWG0ayC%J-*7dcSUzuf(6AmVE+&f9kP6Aj%pjt9#Pa zO(0X2J{0J0xJ5kn%abXhbNDMFswxBn(Ikn;XWmS;he8%j3>YC$A3KZ-M#tzsazg)w zloNju>p2;o@eOO2n=%mJTTo84eg;5nd^jT^)7Ax;k0E-3s*_2 zF^rZ#eYTuDCWUUmI-2>%m4_>ZxDhmzACPl1ae!CG#w9F>y@wyB$@BQAHT-oAtfZ=6 zQggshC|!a4y{EzN$9% zKV0DE)OFT4zIUM`0^;60Aj5K0h*Bevi0?D8<{0J3!^KXzfDtNee^vS7pg%Vb(Y+kP z#X)ij34m%bf8%HoMa|7RYkA{_aHWS=ODdlcrds}k;(-4w=7f|okyE{e)n~AB2xpY| zc&C&Ripd53HB7h^4T=ATb9q)oRsZkg_%*m|5h++FB`w>QyN?4fSBqvq1_N}8ngGVQ zMjzeY)jY0bDI%0BtB&OO%X%&1{?cm^wNZ4v7GbAEU{Spb5}Q>21vDvmnd}6UV7xoy ztvZVz-B=+G_nDm4HFS`SgWX;HMX5e~=th{VE}Wprb1t6UccD6Hui{kG93rKwv;iQC zilHBC$_ETHZvnunu8Q9_ITsKx$g5zNvvuBZb()YX^tYy1eS>r{{T>pod${soI12q~ znhJRk(Ka5GE*vIjmvjNeYKh<|BhB~02Z-X(qK$n)Xv@yPOasI1Amu3qXcW5D!NIbp zcXvO*0&>>z6cUug-blhj=^{KN`PtPp42H?Lg44i+$DTJ4hUT61dZfLJ;Vlw-CiY5v z7K&f2PL0r&zka2Gku#Wd2=bny$;)RAL*V=2<5WJO)Hb*|EDSUZFy% zdJqzqlGwX7tfr%+v;m{c+JX#&@mT)4LT`<|UX&Ea1O8D`BGDxf-L-dEUbs0?Q6;Uw zTn!LcUFvMnlcl1h{2VNYUzC&rd@re~PrVOhacr=hM!Y{)W7VAgNUCO*=)8Z`_+)8~)KM}tDlNMoLd4GKow#D}n5ypET zh_OM?oIVj&4ffrLux%BJ@TKwpQ-sgrq|m=>`8~fC5hj?JR2}dbAi}*nf{U<8B`v~s zZ)*{@!Y>Q_Fn?X4cQw9V%kOZ%f6M<*zHa#ynC&_(|B=el^1pJFYWdrS1vO^U*JZ!VDXXIRJ$`f(46V5=gwgmfRq?3Tkc#j7wC*e0TtU{ zebWAw9dD{kC1jX|uIr`-T)_;j+(CD5VOSsGRPb&*^NOIO%70xkv9;nJttv}`QOaCTD!0O@NR;xTYK3&z$u|L!sT!}rzITjIDSu36&nF@NO4@by_}3o= zp=D!_{~quo3(9}VgoC>@;cr>X@l>QaK6lNt6(K`8tkRy>$i5hULL9$Au!y7PLX;a^ zKZU`1a{UV?;1K2yn?|x)Fvu;@QKRZbH&WHBPHWM&i-JB$VQ?NzqpY@0b{jJA7o;&_r!!p&J)r??}k&1_Ys`~Gh_%=?Rf0G(8 z3|!zEQ5^wml`IW$CVxKnIwi78iQ<789xDvf<^v0_PD!B@yQu}+D#>F4A028s9%*ve zczm+aJ4SvAWnq9~fI(sMyc{f#E%#%upL87ZdFQ3_7oMF&*%VGNs%K92M zm)9z60bv@FCcFoP#5|cSbs$Hzk@sYRG|dBdRAmmhvt*#V&T7bsg#9sT2_S+EBg=_{ zN%oA!gi;oKyxWw|plunp2IHe)>t4gwlU}xD=?iR4mQ#{ot6xyIumL&tDb3bnUbgyx zEsuREbjZ-L`RII)7<>-r2E_9)#rM*A&`T%ex^zypl;RiOZi;^e<3*1>MRZjcd;MTj z?89hK)~zThgyBBvE_6ew>mgyj;b<5e9NJJM*niBh)Kyr**3^2zouD#l^N3Vb(+E>hPr|F=v41JJ zD07tM4sTiFVVJD#KAk))&05#YhjKE11I{y2S^T-Uv4o4?GJ(&`a{W!zDng8C_=ph$QlHG(Eq>dC;S)r{&Bnp_tvBMr5)>xm%iJgpZb+D z#U$EiN3@?OyBzJomp}&_dqFQ&Tw0C0CP6xckHzZVH8s(3yz&89bcNP@`2?zv6LHK* zlU&17k`VS)6{Vx3SQU8_lcQ&YqBriNi#&HA!JnCs};=nZ3 zGP(e7a{^Xx{v%cXX9#_f>qYilH`$Q>Jh+nEcL##3+rs@HPSW+LRI&l)Qq@xi%2dh1 zn%Uk{_?=FWfF0kbf{}!41zPqBNS0*>9<#v^g2(Wem@f~j zxAfNKRfyuA3XD8kjK*vD8WEc@oWc+vF~+Pl-qsu~4YJ>`G?iX?9voQLxeI#yFt~HG*GIcirLv9A z@3P`dd}BHL1vd4GS6R*#T#9iG>8F~IEnL;29)FK=8WOG&36RlOkdk~xsHrcrvhk|* z2<*E?52F)IK&45_QwupCLr^A(rs|44CG#|-@3_4HA7<#r;Cj8EInE7rymFpPkmJRL z;%zQM;x1(2G{O}-8CcZKPL5lPRNg=%zOxmm3N^t90;oU#vGA&{4qz!*-G2dMdN+P8 zhqi(==k2`t7fSy9@92u_=07RQ&_{jV_y^@iue_Ciw&VUk?8u?YhXIBAV8k@ZS9nQ|yxzP^(r*LG$^HaF`QSnjm^tNGTKqndwVrhgZEwO(C=z`@e{Or@!YV8j~s32VDh4aYZqZI$z?^(my!QoO{MPF979 zDe=P%%y0&E}0 zNQbW2v+>jhh@n;e(V`5Afc#An%+x&keV16v%@a8p+7;^V*z zdx&ZBEhj44->BGf8+ixsJ1@=i?q}~Rt(@x~tXZ?C8J*FTvy?Vof$tS)MvX71%X<;_ zN0t{ha%52j#88@B#3^-A?JI}X04Kf_sWUH<4MxSmA*DHTyjN8-?5YX^P>U`b#Vf^S z29Nu_v{wZ6L!1%RdCV<5_Q$a(Y{b7>vSP%(L;zwdAY^8in}?#xEK_BFXq|lz-;o(wkG@aI?Y>ko3j!vuRV6I;-{hQmBUgYy7H$4V6!exk@kkb zxff}mE~+;cBdi0W@wGsfRsEzJfRa=be9w9DjUYZM&;7_nx%~~r_4Pt+pdpUkNP_c9 zO<6f;HNRuJKqJ`{g&>pBYBJTcOkRCNa_0s1V3&u>`X~O{AIS)2 za;Q2cX+Es1&qpMi5 z`3h`mhBrYKNzfDtD5V&6=L=dj!qiQGs4Q&c0VD0Ki_PY@TK(_Q#tkA2%Ptsx6_ z6}4hh`$r!FR!zOsrXZ@Mdm3BMqUTYCACcGFy&tpqF%<$TJ+WWp8OTE^~$s_H*@q&@ejI{X3|!rh8}6ZO8JMfa$OOPX3^ z1Er{X0tHI61Bg!8Ar1Dz1CS+AVG=K6ak{I{^ws!NR1qNJ9^OvhP!p{(4CGKh;hW_` zv@VZ5VW`a0wC<{fCVRA+Rf1@$fk0>Z8|VzWFv=I`Mynj6X@N&37&{quGD7uvHUvkAYqw%U!-QEEjxPn&o$Sc{EwBR+i;8$FIZkpPMzy4K>TDUY390 zdrtcNo8+=#n&k()EPs57EU%}$`$53g3jX(nGY0CWiU3*D z=q1!AW;uc&;C(%IB&_0}qe`WIq>d3C?A>NNX> zL)aXL`JZf&E9n>7Bb(_LUP>?N7ry-4^b0E_90e|TQ5?MBi$r|HX=uaMqFIpK@6wHC0ynmZVNxkx_kcZ?PB)wWk>^;5 zQeON=MUU&C6abRg7vVVCH>g=GHhFOcdhoe=vVcN8(jHNazVO(qnvIA3)N`%KyV>Q_ zMIM6e3q{MBe~NPZeX!_~+S^Mr7VWDRU5TS5SpkxfUX<{Uql=}+EIJHPFQjCe5Tv1K z|Atnw%%?>=1M;N*jQA%8=@vT%gCLI4eB$9b5Mz{Iwk7BIWk0Q}LzK=>E1G==21UY?5*6sh~A1VNl4 z{XqcxzoZQq&xFpo?~=}WF7-NMZ7eyvAg9cCdh|p($gVSfDZ^wwgVEAEDD?dM!U>kD zNKFK8?$Qh=dE^<*=kE!jwNE!T&h$W8KP7Nk)3{<`GRx zJaDn89zL33!P~E*)=li|1=A-1=G+XOo{fv70rt<9`x!@NxXJhzQKF9b$y6&%4Z%>{ zAZ$p4JRZns+P(ZR_6?4)uf{l4W4wB=!C2piaqlZW7OVLovbj6jPKw$e8RQI>ztO9-4&jfG+9~h>2t6=|MFA-uqIrNtK zrD;AX4yqnRiW$#Bg7bbB30Cv67Z3J~1l=#or@P1b&W_(36i-dy>8tT0bTyhK;rm?31yNGNYuv2F*bca^)+7igZ zT8dx3{1Y$Bs0?DeQPz1DRe$~<;?dEi+kp>j7_xfo6Y<(`S3JO1*L6Pi39^1r7Ab%i z7)RX|yFS_o;51wy9(x<0^w?YCuiPnOH4{RO#cC}3R?&f$!Vk<{%{ z4sP2U9#x0sc6ffV<#vUkMmQI6o$Oi>#p#xb8xPe$=5wDiGk|te%>XN2502`Jpdo^S z)dgVGTyz8fh~GGzn+rPqsRvuTm<{uvtzN5<&U_djVmiozRug>4(1Y&-23u+B)EY_c z$ip|pD#|M;&&eRq^yVHD%rJ#$ASEWumiq%|-AEqWWoHJ;9FUsx@NFid5V!Kv~w-&K3@#IUSCb_~Ev^ zDBAuj;u}dr0=v~2IU+DOO7F5izF87~b#jh~(_AuU-C8Il@xkuT$e|v>(?kD-X0e>M2!QPwLZvj5vHo@Kql=Sa5XjMHQ z$yP8>Vy7WK$~laMvBcD^wezfO5HD8E7F5M6IG_TyW{y3K_bB!3ZY+I~`V7C|Pfg-4 z&aqUv-Uh9?h^*j_@!GB2mncNs*TigP1{V4V&y*w3T~}xdP$H>AZ>Lp_gXQqspmh_z zmu%3g;C&#A7p1iY4r27MvN32O{wdZ;gKp5GB1fyXomu9K>Jj|%O_-XITvZ_0KLcf| ze1Q5J=Y^SlRTU+V*1Lu>^u5dlArl)I@f zJVPIEYy~`mtETEMcc5C_iBex73pVPZn{)#h9pz`4OwaMeVQmM1GS#4rQOBNwsHS`D zy#}DzaB6hp06)P|eL?$5Ldh%Py z!XO50(;ZRqf)SuxnxKUo3mENP~eI_h$_S_Kb8)NvQ)lc|BwOfZ@ z_e`P=Q}w{UHP$4-eZ}}>Wt+!vRS92pI!_mT_?91BD_U%c5;Ev`cT+d5ful? zYx#ZDGAyJqcF3*q)e;zn6Dt|a>8i5c#vGW;$EaB;MkCLuDFCUFAIh(2|b?vz?Fq&0~z(z{gQYY>GXB&C?ZoBO)MqYhh>k>lM4k_mcRv+NKt! z3}ms-BG!?W;jL%5^NY|;{BlKVLMrP`(vSi@M(PXs36EiFb4RHDQB;`wpkXIWvs38D z&STeQCwr~1bEuYKr@FAyhoP5@Qan`kV*~{MN-S} zGLkC7P~DLKIMpYqpT=HaQkg>jt9$(=^=t|G*MgIOnu{hb`~FozzPXotKals>SH6Ho z^E(=itTP&RDuW#gFG4h@*?(cENCIF*lw#=D_r3Bsh;~&*9{j{tZ@16^AOoIaqNsh|oj&F8_3~kHb#y#G$Woiy@dH>Io788wVZ2|SYT5c1m#$>LK4@$gEkATxLyAe-MrdUK2;DOy2jHy zD4yEDld19ial64&9(dRh&cY2K*#Fsw%7&eqnEwLwTE{;@!5YasLd4!Yy-F{dU^uoG zh33m>bFunCAU+R1dAfR1KuzOPoh!>MX)x>b)fWHd>v9 zrgGqY+~{tE8i;RJ7Zg_~?ajxw^5~<3^?0E(9&~WEo#2c!I3v{k%u=BQLv1{=qe2aB zJf;`Eg6Y@2D-|2kGF^=jdF*pOVm5p3W<@chCZerB2g#mRX!k%L%m>LFGcy( z#N6BDHMF-UjVpB40dD+!9?K~P`lbHbDUNN#^DWei^!Xwnw(K3v8 z)Ptw#y}77T1z8#4j?`G1ofRFM^KnIFF2WjoU(vCjaHvvrY&iky*lYsSu}J{DIyRcH z(XkA|M#r8fY@8D!_jda!nm3*%a^scRjHYKD5iM)ci#;HUg4x3kH6I^@mtVUfxhU6U zrY%n~K;gdC>D-u6F+DcK03}5UZm=8KIL zAMV>o!Ca~S-{O4VxJ9It_T4m{M%aYB<272Zv0VV6|MgKl@s4%zd~4kl!qwA&tYhfO zOh+}jD@7Cg33g{mv7l{xFS7Un6(<9S+ZYqk75m1`>^`8ppJ9IEv0v=LM|4hjGmhxW z*N7;g&LWr za^IhXjZ7dTMM9!~o{!-W>6T={3l$OwA<7*XAe5zphDkNRb>Slmx~rUX;o;Ce zM&lXd#j_kU8LWRj!=Zg;{Pjxp3m~;-V4>Z8(vayuG9_HUJ<7-R$e`3Xt7j;6C{-OV zwe~{ov!|tLrYYxcjI3fwWHjY8DbcIR`G71NKJfKlRO{Q#!!bT^)XKWu1$y(gK zo`T8XceI5)fDWUQG_`IQ7OlU4;M~)NUjNQWg8e;+PvOL<;yv7d-JOChRNp1AfH(?Q z!ZZ({q2Bv~7|yLR8Lq^fM#aoN8HrEoLMu>-znW2;xB@;JT>j8-3?A5E;i~DTjqr%uE56-_veW#Kp3}IB(8v5 z8T!QVFbF)jr^1y4TP?jg{DEt3rf}6`kyyKcb>!NGYQ_g>NQ6p24npCoDImEv7@@JD znmFf(6$c1Nf?(-wZGp_?bOCvfeP}l=4cQK6I8kVKB8{cwcqRWh>TmFgz?O4~2Hm7y zhSY8gfv!7o66glWbSw=pMt-E4%5O5vh*4vSl{I;3C-kuyEZ9;J#HXU}>UHnuJ@`x! zFJOa*h$BSa<{{iFhm5+GWSs~o& zDd`)74sqgx{?Vu?&fX!f^&}c2_Q`X1t5}L;7q*(=o`omO)oK!{1Q$8L)V;Wo+G0O| z$#!FQnQHTV3=rYOPtgWY?P>wNam5aKgFRNU`f3LrX&MXFDt>Vyju2YQ#^IQ};N2cJ zCnVd3Ut;@FnpciQUnyx}w%qSfI8bdmAKbfQKdef#04!@c{Q{OKV8#p8U?hKGFZyic zye9w+H|`_T%AeWK6{=@Jig$GU(M7lH!sd8hD=JTe5orv@gSaCp8P=cyEWP_P#;iu( zy29=iavotg50_uyZv^v#+qSLOp%=6$dp!1ZA+b(E9=rD#W4yK_(&arY9mrqrOv19q z^IR{h5cOSEls3tE5GAbwF)ynnHH%U;e7!P?cT7r=`?lV5{c%eO5cbILuvT+Os_HzE zRR1b08>IWDJU};!xnJH3F;4xsWsQO1$gRT|z{Q>V$ zGjBg4+MwRRFKg-*{<=ato>3vKov&j`h5#%akm;FVdEA9_3iVvFR@5CO;*|9vW@F1m z$xBqygzq4tjm=>SmF7zlx>S=ROFQ!%ya~v2~Y+QnO&sdmqSR z*I+r#sqzm%%VYoMiR;K53*9ibVtS>D>O0sb-o~5`gwoVhT4Gk+<3M8!YeCc+=n+O7 zXg<`wRI`b?SnNQRwdj*^5-J=WPO|ej^gjxfDJCsu0xH8a{?Tvg#y|OCZTH6Gm*UDq zHlKbdthP7kE06}87E3N0%I{;YcokPJ3(f&&AL7;vYoX4q*sH0s%q^znShf~p;O z=gvhsJ%5xz$69YXTMj3ZthIS~Wx{|Ksp)*^Kojf5((%JOAqaon-WE@hfI$+CsUgRIT4lZPeYjRXR#xCO0g2r`5&qW`E zOqe0@Ixp40^PW}rLrAQi-8?6c_q*e?)Juya78t($Q(O?yPUC{u&dG}_G#SL)9{^}h zU^IaH07yVT?0@Y*J?C6kWf3u_eBgRA+NIUT^v5TyEk_|pc*1ZYvEI|_3f)6EP8Vkr zVL3MI3SB}t67P+essfYL1{m!AF^jlCEqRyq(m5);xHSIiXD%2iruUfEua&AX62p z+xUyDDfHP*8mpCXG!>-UFw|KanRrp3?mulqLcI_TPlzq&Q+|RrrJ^23m3fdK!mk{(byJIPJ4g~z?M*+TH!*}9+V8IIb z&aa)m+1I~dMXIsBo?JC1{jzl!k0~q)S49_YXs&6i(e%W)LZ^Ya+XWz>Kn{T6WG7Rz zGk~z{H+F?SO_*n5#JH@73R^KHYS*SQ9wW>n9O&pnp`oh!Nv*$IX?99Hd6=u9t!Pi) zH2uYvV>g;j3+1QT*XU@AMAJYl=GI2lHrPhd&#qNt!bgyfSJlA&7C2V}CHN9p_m|*9 zFO-&G@j0y;#b_l#B={3~bjAL#Pt>DXj*;Loj;CFr{elp@#CLvGw()s?f*AK2f^P>W zI71WcrV0L#ZS*k%1Tl?;7}C^H{Iy=)CQ5b;iwIU2YsRdlcWlLusaVwpdh0owaaZVc zCOS-;ny^H^!ovXztnJd!d10;JVrP~!w(Xct$MD|x`dTZ zA4hnfhNB2?)o?w+j6Vu~{^r6)s3<^a5#dX8`nf`@Ax&UiPUSm4(f04-FWMqi_b+5c zGq3wVkh;uZGcUJS>T%ErQ;)Sc*>Z(`1j5qHM-pJKl|fd9qIlT<^;k(W{}HMVyKRlZ zLZDTnP->txQ<`S}0AbV2YZErjygFevbED4BCxWc>AFkN1Wz*>fRXvA&fqy@>nZ;o$ z^(5&AHg9euh)uX81#0}e>8ILbi0-j})|TGJ$@pt^feKjt`KK@bc<)>=)0^=2e=aQj4$n0oz~(Z6f0c5w(xyj=*0xf`D+3&{V8^dn6~e)1C`y% z0!y6N3?A-NUmvyM?O);x!TOt=YJ0ai<{HPL#|VV|D3Wt%K7WNdYn>t#@Ma`1t>U_Y znCYjGCr!;E9@h}ue<-QIM5;HHV2P1bxia1d_L*dc*qZ@hTY3E$d3?e9J=goaF{lea zuY2FrCO;C^L9J>;=>f|*6xeWt!OZ%H$c+26+1{#ul-*W#M_|Gt5|86$e z;V{)Nw#o!O8*GWkfZ5=CBZJQdkHCe&V*hi$o`IjkFO0G}!CzPCFBsVcpAC8r_|FFa zK)E>^{D#@6oUqHj>5ZNZ7B7?8;AcSTHv`WOmcws0*azQBW`j?8AIM@-u$*Q#mNUrDSdkmcvix#SQ#qe#rYk7I)yGEG<2mw1Jk# z-sI8i$lMG%GtmCDFNZ#^O*q%Rmkl;_$W<6Gy@QtoW?{5aqCmD>&YGa$AN?a1&bAA~ z2{yO>GRIkA?%G-$wT8(`7Ma?_kra3SU8{<{7-0WDY$t)(jmo4fCHv`#U>|v~-9lWi zr-4y*jH9bv-wMc$hjo+l>d7tixTNwSRzi_mQTxBV=jIUiPLI8jAi)j}vV(jJ@auo& z)E4ZO!Jb9)g{xYU61^E`wCvT^NKQn1;YB}kkAuE{Y7od-55RV_G>-dF8H^TqP@)D+ zMjl*02Kk!nbBpC{(D$#vkIrH2?Jvu@3lxkOk%;-7Q%0&axo|C?MmSExS%e!1JS0`s z)j)!}Lj#fOCIaYpD#~}t!I}kESRLBI2>1d4ww1 zJW66#>~c6QhOIc2roJ2R7q(*Uqj#ea=)jIN3k>3K%P9n&G}Sg165B3dPb6~Sa5+l2 z-GnmSf(9!3iIp7Ga;=yZ38HNG40vml3~X{5k&_0LfsXc-A=D_%+2 zv5#b>uKQB*vE@D_8j1#bWiflrLe6i0+|M}x-7WdZZ~ed_9}7@d7XTNtW%Tp4UFU;Z zzBnIhsHLiI!b-7?7lKu_BXEarfyWWvX``#CcatxdX+UI~^o(Z(AE?D)TW% zKEBM4kHj^OO&}!+^+y4HUk|0(!z8ZIowcc%E26?NW7WnC z%>v^SH3fGVgOeZw9&eNBd*FCp4$1TDOh?`f^Xl0^)lZ<@82oZi4Sr@h{>32p77i2K;azJhe)&n^%Rn;b0GNSDA2Jyo8>gGxFPKN%*uL>JpNk?s4 z6^Tb=Oj9=iBbzZDqO{lfOn30@N~nSdF)lSj_~~eAFq_?{aY_QKi}-=NIl?;7^Qq4`bRbT%b%8@?-hW)fkEFi75+3^qZnKj(a_V9f*;(wN?z-$V@rjJofgA7I}yn-)L=o`Iw}%ZW@fH=ML|WudeNZR5IZUw>`@cJN+5|E z!%6Jev4db2MFa(rV($tnU;(n{3BiI<5%2eW&Au;pm&DKa^Z7siJRan|_kPdL%+Aiv zw%J)f^tj_W_1!c2TS>SXYKFH;8Bz{aFQD?wZk!9aIJ-Na@^V$D)KMBZL|RAC+;Yz2 zHhP* z$3QQ&Yx5Y0R+*cV_7P*bGMMO>IM8^&)?%9mGS@<6hA_t;BK`?sm=fXl6pSg@c^}vV z8-7o0Uql=`ia)kpZ*Bm#vmPq}`3w+yC37u!pWZe*tgiUU1#J2J%X(U!B;T~sW=HnwGgjKX+3%II?M-G57Wz>Q?$4{dUtEPWHrY>%Ol2sN z$L1?qKUZJ;S8sxLhc0D&5Lt-F8wgaDN*5;63JRCqH(0kB$Xktc^YyH4AKH1A8CgRy zM89(Rpz-P7%R95h?q-e>Ix>`+MA#F@q7A*z2C8~(qKIa%=%H~DzTp8BtMYj1=tnd= zr_g`X28N`UYfTa7%iwoj@OyIb+c#DEuHbjS;JfOSoxVt7cSaH0E-aD=fh z$d?GO+RK_-(v-NNL8EEih~8j?=fw)uhHDy{a^?TWTo(nov^<9hia4)0W*e*pyN?ZS zWj8g$0}60QbW1kVv_0WI=BeTia|?r1t{Y4ay&LEbvD9l?0C78grs*a@Ik!_HdJX>n z`ikM|W>8hl=1eDB`4>*{z0;pt!^Jptx@Q^}92`_zsZNBd6tT+PlY^7DBE}2kfaV~l z2;@qE$cr}-zON6W)Od9>AXSWeL2m&yRn84v&A#if<~Y6gtH=#1_V;|UjOU9V3pqK^ zz0b>-+^HmIfClT^k~+;pt8`8XH+@W0QO%0r^gWlhupaj1!{(9Hi~SyXB-OipYIBgw zMFsf^hY@ACDqaSG%ceXw0z3W~T_*@M_b05^1=s8uEO$|o3o0Fyc+=p{_{v{>J(Q!cI1ig5v;^^8@&U`6G#SxxY1trWhRXp9JnqoC~?0Z*+gvD6tVnFVm~kO#9}=0 z{}S>KM+LL7q5P7%Yednsd{KU&S;&SbtzKz2?v`zRcX|U)+|OIhw6$fLNv1@2nFUaq z7k^ic?t?`51AcSS!){Ec2f{u@#_xM4&ETeK%|jEI_xNZSMvJU1xB9~%xGJ_!4AI^? ztvvYFE6abwW5B-Ea;eNMy`yD1l1!3EewsW|_ZgH06w|^J z9c-M%9`|BYHnt(e>>Im4sPceJS$E=J?jfqGB%J_1R5Qtb zS!G|ckj;?2#67)_YMttTADpp^X`qygv1-#uplQ!dginW;qT0jLJ~uhnOQUuYU))5v zh385+ekl-@BcPJZ_p=QMlKJA_8}(z9Q}F8_Q~mrq2Ve6&iK&@;zc_e&=N2548VBg#PjV$WXJOo+wdqi(L&`@tA;H?iL76in=-PRecfnqGi=cUYCP)>!%Pd7tJu-Ne@6I5lK6_~JCf60)n?`qN@FvggrxZN( zu_ru@c~~NAaNG03;xb5nTkwOqlQhlcC|zrP`-Uw@-YycqDs`WR6fvb9MIS0hYUo)# zr+fhENRU2M*v`9i2P5%1ppn|ww0i9>zfvoQegJfy(vr`wlaNpJwP9oqCJ_Eondoc% zGG9%pdwn!-K+fTld^L~(A7S#cbl&5-ibS~UPCpsNVxyQIT+eqXH$%vlx9KLZk#pGp zB&bsAP3jt-Pj*z3Nyp%O+z1#_JhXua#Vk8bp_xo-pKPwOa+ovOSNe7Sl;Z7Ziq}RZ zwhRwjhGtM@_!{Q~ICI_g0bxxap$_j0^FHUUbDvll?Xpd zz@|j_0sdM=Z_yevK|geKfOce>iP2z2n-eJ(&NCKvFyx0P(azpXLTL`_ySK&%J-7cE zt*^d!flY)PsAS?l*xw)ciw?5l=!S`i^^@DGEI~IIQk8aO`BZwL;R5Y;bc$WyZm%%B zcYEdGkjjKl_}4Vzk9C31?sP>_ik12~YN#CsT$<3JLLil0TvOf^BK8O<|6 z;m-$DqHZdM8r2=-5r?T>J=lKw*x=Tolyx)|jchG2$Q*{R*J|DWMZJoZuH8W!Dbc{( z<*-mz)?X|8<}10ckk{w}_k5eP#lPk{2LlB_D3&!k>1g`PB7n4+stUuUd zxN1GdA)5V2j^sJZF9Zb3;)?-t=jT$IX!f{G_zuS4_GuRG`3S>B-r^IIMppcOw(8IVXY5E$nG8j0?l5-D^C}Tu zQlM7xDS!RWkVP{$#eC>$$#!s7th+?2`40@332hkf$!Dsw+%8JonUIf7BS#{9eJeNm>^5fA|H5PeOZMqj>HDtS+LW;?@pM>FSei?C#5D8GUTk*tQ4xJxr24{8Zn#4YJs zH<98$2C(XCxGJGEwZxWV+33fM`!nGlps6n8hWIs^_|}ZBcjjFpe2BJmgy>Wd#d>H2 z#SYPS@ES?-KO1PS7D-a^pd8&Z~)cBNSP}KoMK#cCTy&H0O?1xwUyCb z_On844Zf7|jdSj3ZMUHCR#9{E#7kN?TXtX~+(qOiv>g!?9s8V&9_PAACi78{CFgTq z%V#Z@H*RSMq*ZSlD+>5KU6GcHcV5MOKU6iB1y+;e=V4S-u97p@J0S?! z3^w^V$~i-(FMfA%?Q}6@Y&fw2YF7h5#l+^Z-Y$`T`fv`VD!Jf?t`h>yt zSArSpe`P&X^|`iKQtWRh_J8n^pHgOcEy;1oj=Hu zVrNTe;V|ymk=Mgx?cv}t<}GluG&<+y6%EG2j#7MFccTgx}<^F+Anoz!#;tKR|T5Bu+_yX}Rura@e6Nz!jLY(fe?M7@n(Rw!qQa?&E}-RVlblR7xqKwQrMZ*Gf^ z7WY8!`~HQeu@K+E&t`6mXPveQujB$p&?n{<$vf9yoxppz`nauFt#@zp6qj@Ll#^KU z{_i23`Of{{_P&`fX!`x%d~fpXyf*nm7%y!#XX^6mV_N$j%%f`Rq#92va#L%bpELn3 zU&BMA6F97@esWntr>gC~1yu|EhK*Gs<~bIE#;rZgZ-izx^~1?clxxVDy9oJMqzlPi zq^uy{a=TYr%NVL2h2~i1r4z(!8+Z*INrutT2UT2`MJt{+Dm;kjw$xR43wed79c`Ga zpsXB0W-uz(JT{Lx8{c?Vw-}!bzw82e1QWYI|+fe^@SeN;zz4k7O zGZo&*R&o>Wv{N>L^q~c%#Wmk?yb-5tdWlF)+XDAORBDiL@0i9mx;MS%x|uws2Bkri zq~&hL(Vcn2Le@hDQt$3Ikmfe?GMF!)YMX(pj#Ho{YT`DHk#RqU%h;xLBBW<3o$b?4 zC~d(9qS?1Ay;}WUG<&Sl-}&?I{`#LIj-6vL$AH1$DGV;E|Tp$tdZY+X9XYg(dklbufV5a0+ zV5J$*%L*>s?ozoE`Rm>9fBQv)OhP?PPLsw(TZV1#Bet)uj=YY5s zcmiUuPk#xEc@`$RbE_rT-YV$`A-3|?C%D*F#W?y7wxoxlK8f)0AZ3Ln2!S*o2s~L? z3?ZvJ+Lw5tuiBw@PyW`)Fy*-iEESyotg|01@P}URrTfy|gkj#mSj&+ zA2r~&+i1|k1ypJXb62%3aa(kv9Mo$wKK6$X&XcN;(cKT#y-*N~AXFqS=Mb3mM-%-D z%4l9C!t1;E78Wh->L!$Ho44<&530{exL!xo1=X-6oNzr!7T5gs`rh6ox@|a@qW2g7 z<14-yC(p~Ph5M7AG%cHPbZ~NK#J0xTLmXta$pUPSS6X9c(w8Y6@#!;_KES7omG0%! z`zu{Q8l$^UQS|y8_t#Lo0oRx<*?k{QgQKlvE^4ZRplulV~i+VoCBUeDo+u{^o#l& z@jzOAV(Dgck~R{(!#j>g`|2kQsUp+{+|!%V3DkTagc56TZ{p;j1{H?!R%-ZUstma{ zJZ^FwJ~%jNwu#n2qS;4)+(L=%tAs*{byuQ-L@tSBD6vhI9%({?k@;>4jn`VwqD?@< z=z5YkyWN_!A8PyV_0|)m{lR^0WtJ)7{{BOZDrDj>q+%0Q4>aXx_2)$h*XgwHdV9Ov z9Ta3*D9~D7N`-c@m-X%iy&M+2e4HQgm}*SJ?&Of)b%6zc)={m!&t0lEE%q<~8m48o zs)8)FMO$PX-`3Ks&E@D_H2eIR=C0b_t6A!2y?`?j!m<2WX0Pk1a6p7fBh2iWj<%l6N9;erq0)*Hs`?qPF}Gx8`>a?!DRn77z2HkA zdy>BSC`*+Q>n&8ZIVgio=T$Jqv1fDI^Y-_x-x1w4AXNqCr!W!C2FVIj#>OfGEh_xP z8#E5PDkRHJMdrFSfBLfadL4GMT*Gl4td-YEXOPfDz5I%of@W&^-JQOj_j89qUKP;~ zLzSA89c0#^q1yK=xo$!edcd+=U4SXyUCD>|oowTGlr%da)7UFl_kg#hSW_eBjt}zY zyCv&9lez9gQmUeT?OSu|el_T_fI8<|Hoqt;R#`w1?TAnNZm3B9R56OKW zW({9NqJCt?I^J1mD7W&d1;q}mz(CXvn6l$)9(`ZXqiLz*){S6KW1bPzxI~nb(UOia zZXO?}mNeidkKon}1VfStdbEmQ3qvph1hMySQOm0Ar1`*g{FQ}T3xCD^8+C2mHWEot z0Jl_aEs7GkPp#E8<119HF}LyE0o1L_;LqV{GC2EM8A;|>k(z&!;6K|4{1d<=)6&=@ zV^c82v`;P-VDYY4!wE*}PZ=B-VoPoWj%jN@Sc?^59a~3yRTJWe$EgB@NJRoG5%g&t z_MJ_T*lSgi%=sMynV@?t$V(>eOiT>q@ozoHuePXrN?AI;{FFUb_H^eUzqPBB(~4EV zd)d@(aJ*)tWFuhe$+?FtU+7Xd`O*;F5ckwnpNUQmUcO)_Q(hXIs3>wq2> zsFie9*%1QQMpdM?ZFj5(Oz+XrbuXy6v`c>ALw$kGRf86P<=i|`YdI@ zNhegb0V_iU5ypo$P~6C5HQ;TGUslrpP3!E4Z(vra`XymOUxv7Tv_Yj{`Tg>pvSL5V z@@de25IVtV1Qug^lMIYgTrH(`e8Ej1D2BU9LKN+Lq{YfLKTlSGHjUH{$}#Jt-N;f* z;mo}Shb_`LA3vxaJAy^qP_h&E^!BzW&rSVB%rq}^)!B%d{U7#xi^(*Q9)c@Gj_u(k zX}>1?1Dm_TW|74(#2=T4LXHRri29d17Tg zIpQnqoaLz?v~FQ7axJbcD9}lgE=h*(7vO5|yLsQE%ew?9 zPtv5XlazO4t9QXywD2+Scl|2=O@~ zq-g{VI8=3|y^!-^>_SUweEue&MgxbzYQ}@sJE$f~+*bO18JiwSaoq)M+{RdZ{jroqN#J?p8bUr44YLNBif6ZzB%XCEOA4$j&Ps)d&5_SOL=^pSG`$#048M#o^8w`^Z;Xc_!hIq+F%vNb+LQ&LZfYr6h_BPc#Vi)Lej3!j+OI{DvNW5IXwL|DkIu=i^C$0c6E!Et=i0SqEdIX-ewqRzCpn|7)XR;!yt5@O4b$R3T)pe!M;D`d$oXQ8{U&Au}6| zql92v%bIXKJiOaB$iF#HjF;=A-zfMDYAS_ccf&9p3?$r1#$$6B&*rU0WqOv_e3rj; zMK!TOGxES@4p=6VQA25lFTSZIw?@|yWLi7eNJwGeuOEGZe$f*~CTB3Bj;zG(29a2s z`5GM_B^rZ8BR8PI4gq4=92$}6X$8Y$j{_${4}(yI?g)!OlJd_nAr--TQvSqCp}SC^WI0DD-c`9y4&aZqC-F%1<@<=W)9A=V<;p@IUb<`qr#lPv`InV^UM&GSQnsJ6oP;rTCR8 zYQmi@eqI+p!;GJ1`E)nrp8v$H#-Efb3>E-+bvpwd{BUBLc}SIuDD>-Z`^>N3+wsd5dySzr^}Yi0+zL{V;wqVI>6-Q=w+NpBCC~Z^qB$Nk z2DG@6MS`FAomPP1}wAN>H@eb!qQ@Pzj@~AW}!yN7ex)OKUWwtI6IF0h+f*}I#6x-~(9<15& zhw~4Asok>wE!BLKY?{e_uOD~SQ42*ipRqv*BN|FpVD-~nWjNfoc8YB<4`hT^RJkT! z2I>j&4(`9EyuqA>2H9d=(asBs=B~;f+-415Le+gW;ED!2W9QA)wnf5SLO!)G?mr|| zlz1;xZIcicJl%tui$W##-pU?CpC2m_gNYg@kCe)BvqCRIg(M9{Zg?Ob+ar ze>A)sq3IEC?K5UIO57`2hS=Zr(5_`s%@@+QG1fA^#Unh7EKd@Uq89W}u~ zSCV2)y-5*!37rMN$z+;!^eB^^M0mpQp6P)q=n%j78miJI1|=?cHPvIWAC1n6Z(9?b z-zeYiVpEo;^e#$gH;kJ^Y0^Z+!n~NBz<}}KB_7}^V+eCxl60DJFuH`0y zh(7k!uf5cpZphO|W?VQn7u6nt;V3T1zvo-tXY70u-j2)5m*fctee_5`;K2>$bnCwOI&;J!@+&j|>g`X311d49_h)+Px)nGM0Z zcAntlOMD5V0m1Af!3Cz{D&a5RsL+k>N1p#y!h%)^Zj&TStBLa`&_8BTBXt zrm(pGuUf=yEUH{k=t_)AOHMJYd$W|12=8Mkuv$lxHG`AbmFTs1nchn&>q&KjBm>I& zP+Q2M6uGYPOR4}pY~X2KN^x{}Vo+tW_@S!V7_wx6-T^2cLj+%uh5eaJX-v1v2I^N- zzpCxma+jMd;Nck3Vxs2Q*|3X|BWvr0ZmbY-eGtXCF|f2v0>zbo75&5MVjRX6pue#_ zJP}t!J{u8shwo^H(+d>lbaa4y#&S&Pqt5 zTg4BX0WY-#_BF#?9cSoo6na>A6`1(j)_!mCp$6R($`;KondxUcfhA_Eo4B!$`RPsu z#$GBCX=M?leL^ z_WPWVQmiJ35;qjTMV9zngqzHV6xWos9U><#JoJIDTc9dnlro1gbCvlv5pr|%Vj;}ZAP5Z`xZ0{JDYWh1mOw$U0@_us44_!?XpNM+dn{>&rK z97-Qz75oJG1iKg*=5-832V(Zwwk;p2ti25d_1_wbve}7nooKN>{}Hm$@?-}(H8?Hx zPB_0vbHV;&purg&gsQHlGUBh{-v*L+U0sQ56Pto--Rkl^iV$s15tb(}WqIEGL3SV` z@I;>y*Pi(U<9+p|RSzrT2wpA?@sF*^Z2`K>9NKtY`jf;%+j;674OXKVG(T7^KZY=Q zP%Td*IyfLYGS!ivf=jI?1Q_04?%N0D(0OP<+V}ko`CyQ1%uRL6TgKH-GUiCerxe$wAIe`k>aP9N{5?v`KNv_JZXo*0_W}b&Xz{3 zjjMiI&w1Iupaz251~N;Ru$h22X9(St?~ml`9VOD5eOJ=^9fY6aAMS}`6N0y?qslI1o0~dw_1Kg$vFrQwwGM^}R<|M-JJ73(-_3??|<05_tb)kD*uP5-@y_9?>R9lf- z01@|eo@T~=uHot)U<`bQbuuW%n}aCkd&b2pO3t;B#~|~B3)XIm+AJ&TzbsV!ysDyR zeSEqNXX0N6OB}bUjYj*<#od=}K$`%5JWj@-Sy@BXGo@Y8aEC#BjrJY+pk$6IzN2AE za=GMikE?SUuIkosm9xtgo;%R+RtIN&K*l@3ln5VQufA&_e-##$b(9<-83=&1y`MenT`!(0 z{GyDC=HwdE{_1qE?VA9jI1|6EG8wpwU?Bdcao&5Nluau+Cy6_&;MIgPVael*`+8)> z@6apmKjov;ypMpOz6(z)LbIw{Q`CP=*}S9_+5+VV&OzmU{;gGBTlpIYmAA!MUwIk) zhpKP(l~;U_34QAQLFIkP&^N$U<8itBQ@9pYw1TUAS1PZq$8$~+&jj$8>e=$HuZ-RZ zo%ql4&dB1pt0SY^D*>o}gr4epKiyqxmfbW*WWKI`>l{dghxaUVb8U)41oqOwqa zEY*F!d`i~LE>^U@4FscRmC7{u>wk1yFT3mW8anpf@SJ%{-*8C*GOyPNsi;Pa=3i;R zTr=PLppU^+Tzlb=;@Wc}gK-(yL1+D2iIKygJ3mf;JD(9i?KL=IUNJ1yQ;z)_gta(I z)*er<{70n!n`N6=D4*@Q$SJ-GVER^0 zGKqXPUJ=+(?2p8p-S1T57LwJd{dCGRSt*Pcq+$H~;S|PFkFmGMI4+6t`ZSE?NsRY` zQEh8t#>79{-71%k@Do1-N*O)d8|+N5-p1)kLJ_Fbxvj!oVQ_l_NB<|3tq{yNf_+>4 zy1UzmuSpthTBYH=bl*K$G#b8_S&7--ANX5`H7Rj#pVe9!d?$aq<$Fc2Et%`#D$j>} zD@Kn{C8S;lI82)FN3bR81d9O9fj2(Cn z_N9TXN9`gEYOZJr=2sh^0}Qn8`$cNp>3GVP)Bh_wW1(EvM+nqgI-(Jl_hof*DcUzf zmV^%5*Ov3WNY2;SW+yz&Y8Hoib#(!vF_<4k}tIn{Hu6_xx%BOz$}$?2pIJ}wI7{5j(_15qc7U|XkWMzFkJ#V)4RwQiY`&Ll z+6nw~tQ^N0-Ue6M65s#>=oJ8*4S*zhN=v}*2G~9ToPu&|Lu7knU|ahe<}Vg%_G0Xg zi`-q1RWW-w9dcF~afPafkXle}zjbh9Q{qevk>GPv&L~T5x!wF~9o_irqPx?6InctPdxo%xMzn zY=vodt0xwyCW>qRkh`LBCYF98^n+LrSRx|6797*(v6zbXC#XhuU8=fd4ebz!WWsn= z{`Po*wG|k4Ylyqr+J?>wRR3PyTVuL>Eo1gdjRnd-^)_pCs#P%MKzHv1mD%mPzUY$` ze~b7B-$>gN&bK#B9=irDFRk7^8z&nZ*O7r1wcPzCRh0@2;=R53KKW)(dqX@J3Fy7+ z5m4>ykxI`o^`2d!^h-WHOzFu!JzVMgNmtBPMP-$3PR3GRE~vKOu0n9o?sunfrc0BR znFR5Fj+3OmV*5n(c$|mk{()1dFO1d4Z;~AY_eVx+AuCI`*I3qvwDSUiCxNu7jq41O z_?z^r+S}ftIq<%T?4%z#%HG4cO)sQh4Ey+46>{BIvI7nL1%55wmF^r*`VXZ~@@c2^QN~7A zlR(AWf@=Hi_VjSZdJod-1}=P#(^ePI1ES7rf%Pq{cDnMV)EUUbdarf|0xa3Ay0*a9 zwqEzgBx|b+g@A_eC2cd0bgFH}-%3R%n9G2_Xx9ujplBWv$WlXn{(FwLahWC_nueZE; z+LfoYqlcbp(FuCw`sf6m?0vItF)cbld>$IIu?>jo-Lm8l(>H#!=upd2(FvY!?R(JA z#g9=fhn7&ktwkpYy}Y-a6x{V)J6K`PGgKL45&Z0F@sEPXyeR}d1nBrj+<&mx+mf7` zW$pmpWsZ`{xxXC6JfEBjQIM;mYzdSuCX$8d*fNLL%zcgLQ$bGrd;S_-)iQ^-z+Vnl$=83O zlCR@$sytQir%CRDmiax!yDk4~mVY{bgIE%qV1c-il{r}GMjfyA*JQ5MLQ3*nc%a=` zbEVL%`8V||q3TuSuu}I2dw_A~M@fvMHuu=x$-oxPzK4s5)n0okG6jv3@inda44~9B z7yY~`&mZk#Vd>Ai_#Ntf#hm8!aSV+1J?ng%QGSx!UJaScx;13y4q<4-?>)M6wo=g#VZLHQ_JFTYdLkM{hLCmb#9<{sTAM;5HuELej(;RF`iYL=;(MI_UU z{~kW)dO!A!7R6UcjI&2t@jnd(#UJyHir;NF@usNnq_ zrOiPu+Uqw)yTduLc)L%48S>1Q^t_3AOuWEaw2L6O<*Pl-k;q8Qq zt8>BZZzqUmN0jbrb&=gyX}ba@nq8oDTiyraIt+4wSVK3SpAuJst1p^27$ zOtP0NA<3Sl1dfjKGJH5_0|O7?4| z1IfOlv?Lpty$-FCDK-5*L}a4bZ|f{v3rXD~lYpfD_$gCT>+Pb`IgBc4{V;5c~TU>t7igHzp zP4+#Sy#XIca{;0>>*WvGqBL{V&M3_;%}TR-P@3;&wpN-WDb1X7tTerorCIu=)%(qm zv3kE+3Dx`gN>r0LnMCX=uk=SKtqAqR|NpS5&}E_)6uRym3VmF%&}8qdEc^NFf8S18 z*XZ(oQ@uYyfD~BoBR>73(g*nTVx?`O7|ouqbb){WlG1sk{h&xAOPAQJgEeg?uxUo* za?^~R57!3lCWyYoja034Tv!0`tbz!u`{uu)sVSh8!u|f3DmvjFeANb}I6s~^H5dyS zp}SfoZq!+*jmkPD`|Q5M&;@l#w~Q#rVtAdeQ^)w%{9>EcOsDRs z&>ne%`+$u*&BcaJ_qs77-P(q#og@$qe{?KoiK2ZwysnOg+_OIy-TATG`JB$r+3;f& zjGxP|Nb$1>V<>w6oYOi#Z3BJ|4ER~_Qmg#PX%>D4dVa2&9q_X|{D2*n+=f?tfvQ1Z zyl!DKVuV?_pXP4O&)ghqgNts5s)`1^FYN_%7@(5Mz}C26gI>MIkB7*9&5z%!Cbt!U zaY8sH-cu$2$zOe9B(sKOz}YzgXYm&U&aPf4DMBvRzctQqON2~iK!(dm9|_Rn@H;;~ zTSduBce4sr-;H$KS;uRISnH?bj8MW|&C_o-l?JICcb-zBeX-G2TRGmqxjqHgXG^mG zW$?}Dl>6P?9}iafiO1L~45L+;UjR%Z{NuY4$_o3tjK4;A{*ivsE29YIn_9LBhI)QX zfhZHipSDs`=~2wNv!NZdKDO_0JVsa-#>QEQ&} zxmNKMZ<8~268E;k)D1Gpa-_J(%P14bi}pf*@c(^BqOIXCIo)?73tBc$#A=rCPmEAlk6{-NrcKUc_-W56S zVrxQjLf>n&-(V0xYCyM*75RaI6d9C-Za@-6d~zcE+}qmcIv=C>uYNhZ%#p=x?w4~Z zQ>MT#=cntM`Q=<|Ha0zn{BjP1l(DiItdzKQZ-Evk@|-OBl5Fs#JFP z!$jW)20N&+GhZNzJMTO;8iV<=9n|ReJLS`t3BMuS=*ZleUCl7}-S}}c^EelV(+_GG z>kHi2lhIFQTXcAL^oEH*DUl!!kYAZrOJ%?b9PakJxD#@ek2sq+_=ijacZsY26FAtL z_DtI>DyEe13it3K$rmdr>jI|!`ZbfO+Q$N>PK7BW1wZb$J+@&>c!Av-Mc>wzFjTe+ zx=#Sa>Xs+1NQf$c9qxz_`otE&@H!J;^b+_LkwqxSs!^N34(pygrlPv308qzO=!p9b z<^sIrQkfiIAl=dUW0`xrp^QX3PaL>4h{hYGS3d-b>s?lK$YMk#imQyqTPxVDtyx3W zxgxgi7F3EYk$2l%s+fKC?ia=tc$mG((5CJOle~NOd>@?2C?~^9O(I`v`A?&#KzZhN zkn&V{*C9pQctOjJQSWS$$`QMd*F+mms$r~5V5+;v3QB40Zplbb)y{!KHF-af5Beku zABlBQI1>{i3Ue}evY2m@aCeFn3w!i0tE8-naQS?-=XtS@ zSz^`m6P|HKN{D%$Tl_3&6tKE37=TZqK@(DVxL;&hg06>NHowbo=SS0r;h3m??p|`s z^54c&Mho*k??|;RntqAv+Q)b1BQyKv9}t8!vlsUF&Flo_R!JqHt$y6nBrS1iRW>9w z%}uT5N=9;xZ2FdV-6Ln;^*KNIY!SnB+C5c4X;r$Wy|2=&bY%U~2Bv;}F{36q4g}!Yq$>?Z~N%Xy&O@H^ui3PYxz_&Gu6*yT%YFT$(9;iMsDG<38b^4=`{V-uTHVgbMyuy z@{yiR?v$t5*FZdK4?AdXlc@`_E_!E&_)Hv%G+aMC!|La>>yQ=eZCNS1G=9}I6@TpO z5Rdv8vF=k|wNi?cXiw%%g^0%8&`d<{nw1Ph^vO=mMKrHIDWWS|MP9vBJVpBsInm~u zbN_CK!08Vzep`HHcY++wZcGD}xWU(p71Zr0d;&de7Bv{m{W-?SrS2^Fa1&_|A~JiX z(h;*F*-t8cfKT70bT6NtrL;msn*2{FhBVW@k}R!ZshLN}g3Rr%G@ z{PdWd0pxi=Cg<>51atTdI2b(y!d5CGI;yWLmG!To>5TeDVPvIkdq6~%ZuF|8D0iLO zT0K?YCMy{2n|*?b^2+t?l&!?=py^ID`@?6q@3szN!D(zQQvky_)Hj|>+eU$EsTmz9D58jkjt)f{a zt?Hi5mYa7WXGUB6O^q|z7Ek^?xO{K)A4f)l@qIyd}IbpIG?-mQzS)2 zl3*K}<{S8>A-K=+VT)!TC|Y;uV16`v7o~6V>8+H$)~B|e%_9^v1Ar}QwN z-k|hiYd_gPD*c{Mf1&gopPr}mOFsR)(vyAqL8b5a>8R3mq#00h{{3;|-VNp1Y<<}K z!ED^uJC-~?xdnh<-S~cQO+?zUG{E{7f7$NrWs?2j08Jelxj9fd?M4W9=9P$;&1M7m zWfyu%ZOyNDUwW8uSGD&>aPJd31AjHrI)DFSQUWD2_Ra$_12cO+bw;IZso85z~%RYWD%`m*Q z$3nv<(crHGzrW|gB9GDNU!Y{^yX9qXB*F*HYN|kYh;k3M(z0lR=u?xnY`6Yknn5@h za2czco-i$9z_6d6v1WqrYn;8>M*|<+(rAYZTZLov#h~L>;1c2KuSubv^xeYQ_)-0Ze| z!Mq*agPNi+|B9pv>*w3+dyA{?@I&@2W=8IInoJNSx6$38A9m}aTGgWG{l&7S)~Ym9 zK93ZWDhx`f>R||SEc#{1f12%YkS7u)94sGX^uFcUdyek2uOJnPmNSLNhU`K5p{I+0T1?piGO8JjI^TcsgVAw{sd`{``) zns9HSt%RR|tJrQ`dXoHQIt&sjzSb_+Wo#pG^Tj#&vnFI8RAwN12xNrP7n1tUJyqRU zSSzTO2hM_oZ1qj)L*w!;R_pimGVPF?(acwE1&7RTPh;vGIEl}S$0N0scBHQ56heAa zyXUm5*HxK@Z?oql-ez6_sQYQe$S9tvE92~3zfnr+U??8Rf9(rX9i6n=z*+wfX9Mof-g~#5 z*LlHlBIw2XY}pgZEPVwoq!#N`!b~25_P>XRId|DN0J zn{%6M+@1=~SX<+E&m^};wB*5a`@`;OORpR#NTdMx9X1M%6NF*fi^!oe9 zhDSx5<3vqj6OCCJ8J&fBUfU7$(QnSI%JIG1??q|U*gXw+apEz1)BeEYV zReS8a_);FM_C~`;AH~v!wa1MKqK5QzcJr!AKG$9ogy12LfW{TN8Qkf8{z>S^0s1k7 zA2uGj>j8~;*k9vmtiAONMHq<$tO$32gMA&FSHWn9B-!aZYc!4jpJi`!$5?i>pgNh| zkez7RpGszbhmZFDOcE;R3@hi&$?s1kzYhhypPLXNW1rrugA*f-_>n$KL*RNAjkhF;BydZ*{wAX(HdoWhHLXWZq|D z#fNq!Q=&w;{&~prjdR*uUcIEm&6yF@;To_7@u2Y72mQ2O-4(U6pVm6d)Cn=>u^;ZZ zV`*#jbKBhShY=A#QYApEkmCu3(yGyrlbqJx#~}MX@duL`9=N|gN!Qsbm7j({g}wEG z@!S6=i+){j&C6-4x_ZYiGMo$mILqd%_;h zao7D{_T8d^O?VuWkgtlDW{%~TD!S{)0xOjU;3@4bwBp|-{F)K};QAh0V6m^so>jI` zt+q0^XI7j;@Ap09gN*dwa&8n{)BWdNxFWgkm&o{(qkGEM9}sci(T zwRksYqFPplVdx1E=N}5>Ybuvk?w|TOrx{S$*W6nVS9Pi_O|w$vIu^^h5xxj+%iaIM zF~hD3aOVqyWHJFe7pc97u@J3ubc72a%|88k@Af95OjMh4aSjuKMEKsP<5{tNr)e}Y zfn*D!q3ZVV0O1;(4%`E~{lg5pIbWOb`2u~up3jN!5#V9X&fgdF&RVfYsn?oCBbszb2^X9O1GaEmWoPw}j4Uz=VGBL@#vjtv&1E);MM~ z`wydGaF=HlMX$?oHzhGYV}>@k!j08_o4zL1d^1DcOwyrht!!{{h)3wl4>~!MJp5EE zr3_X7Dwd-I?!$M=J-c&j+Rmxm0oUqtU8|qjG8%BYK3~S?M0h{QB*J^QPHemJU4}F&6bwmn5l#>QG>(3nPn_uD?tF_vDdkPKu!O4Cmo}tg>TY&6M9J|~X zg@!D4g5|iI9EtF)_P3V5vP-wuu>^CdmeClqOPe*XD2?APgBBqqYmP4I0xtejQ0<#y zk$qb$+c&+PU42)2JA3<~)~cMq5S@Z4MFgavcDI99%NOe>i??z-!$N<6<^PZMQqzE<7R zLCfId3OJBl;UNg{b)J+n;CfEJhm4q5u$hc`+py-F&7T3zYH@aQF(Yo#0NVfLD zA~kIbbge=<^72RNA+@}s0__2cA|kgcTuN?o0rODGfOUV6;wWgD^$QMcP5nZs`Zlje ztGMgX{k7vifRB?w^W(~qGIO>s$KGlLG-}%!2+EHsaW^xnCY(i6wouihLYkJfb8A^c z)ir>Tk?bn(YqgtkmmK$hKdE{fz~2bRSt@9Sp-wI6dM7KCa(6L42*z80lyAf5F8aKK zK1)a)#AD6U`1Y)SJfT`?bSK`5wko7bBK!`os9AR~uQWNhd7Ns#d4aoHjg@uN2mzEY z%XWhDgIK@U)!|fwBoQ9NU&3eVo0sI3>Eih3{?;UUI@cO%_K}ZNP(0#)0+zgf2L&25 z8&PDrmtr|tgQMpvYE@oQn{SHxe-o-Y8up1=m51fqyP!x*KrRYc{Ckz~oSTtr9r2$Y z4_d;;p5>U18)-QEyoku*kdo=F(s5*kf#RBVZeQAR{8sH-GxfT*V3^y7Z)w}uJK4{s z+-|Gj_arMrlI1qMjX&K|lWB(qNIA;|MfxF`V_1?vGI6wSu z%vJY_%q+WZnT3(^Q!EVF;kUGH(m6?FG6>zEdb$V;I|K&>Kv)wv1uI$9e`%;n?zwcg z`K|+>P*kps80a=c(VTDb|9}fO@G3^O4_SM;!OXZ9wMu8kE!c0l0fBIN$vv2>Ye-UF z`SILPDpdWlDkM7KHLfmn9}>Nl^2A=%JEYQu;+j+r#nct`UtWeI1Wj_` za96es(*+;=?PfEBMtU?>bzq_03x?Tm@;g{DQI(>2GS?kJsfV`4NI7NN%F490%G6v5 z%GF@D5|6to!fg1rhoxX^_^aORyz$a`Zsa&!2@S4qe?TNJDi~f|JECBuZsIr`nbJSy zN1})HO5Q=x%PY;H#$Wn{1B%Wme~Ml7rE=at>`^&URxaW;uoj%b_!-PJ65-)+W7;s< zh-ZcrEd%VEyTw%VmgPsQe{Q9LZ8xE40im*Qi~ES%Ghb!@W0Eg*X{{C3jpP|SkuMU( zss5!DXo!-M$`Jzi+jat*6Bp_^xLvr0eMizHbflG2rTZJE!ntqMYlgU&hWZ`MuBV5sq#5WWT~O@ z?tx{@=WP} z&dQlYHZhLc%Ne8*JYGg4cUv-eOm1++iL#hF_gQFKGNA+EjCTkbaiIg z7c5!d$}aQ=z8H6t*Bo$Y2Kgu;4MRFv!Gz2MJh5MhSnhPAD4A7*`3GGG3Kgq~sX$`G z#M(-zGYpQ&M@~RZV%6AqA3?n8nRi=@x+cQxf8M2Tf3>>(hN7|a9Yb}u3$*cz!_<6q zuGs{wC(b9S7CKw6Xf_(%HCdbT9ZZ?G+SHFLGr!R-<}-4~{@ok&p8G*m#bntoxte4` zdWd7(bqMKhHSrTp@;2q7Dlg_4=JEH=GInmtmEluFjlzkW5+(nq ze!|<89NZvR-r&Osi;rB_o|m&}JXys%uZ&!{x?gdd)x}8VM&h!|uiVvFGAiWi6>Q)W zZ3t~Wj`JV)cyoMO1;V_PhEU?N?$lCoQ<&g>?m5dQaVct9N?ct^TxV8-2F>EFMFwmS zGS5$}98`m5(ai)w8f^XtVliq9i3(@TCe;LJ;UFCWg++@UG)vR0*gj-GtoMp0_r#_w z8Y?e~zZZaItL4%LD#+vhymm9{w!s}+=6jN186Bbzy0m{tb$dVorpgsf5S+YXKOBc# z2s3iTMBhDT=Imf&la=2?E8No{bkSX>t%xO9hvE-Wr_55G1~4};SQg40H*H@uUnh4g z-WUOtZ&TT)Ph=N>4K^IG+PuR{QjDrN0Z_GZ1p&0x>juvf!5(bF;;QyY3LZPbbdA8Vk8?7vv-F zB=q1r-_kMaZzsaL!2#P?Dn;e&2xM59-GZv3O#N)6WiRnPkE)z$h(pyw0VBRUqi?mm zbH=J6%GA1Rbm=4Mm z^L~_zVt(uS&`W|9jc-v*nMhi`eA;2QM5dp*1@-(0*+ zHlSGr>iL2TTo1G2EoT-%te|g{ps2^+pGr zBRZp?(~@hkai~5IGd8w2HdNeEHEN~x?oz1vYam&u`(-nRC?ZHfB^Xwqa7{Y3X$1;l zUVF0y)?7FTm$=vO(^}pbvT4tnNW+trb@CS=I@9LlN0jyd4)%7g@qf+BE)0VNB^1sTY*}ZoP2RR>q14U zeW>kqD&|Y=Y`OWmJqMW8GHdqY0hS)E_dE4hI>Wf=be1XASsi4#xW5j{dLG3yb{>Vs z)|R+iZc$API1wub-EEbra`s2!qU=C+VnWpi010`$!Z#^8-TR{+ksql!hX2+2=|&DR zR_s&CmxbI_o08kRolX$UY_7Jm+4b}xRpslOHk z&djyd12a%0UFMKyY$_sGAJ^?#KYI%EgzY(!P6yHz7PTsYzT>UKq^&v-a zOyVBc?Wjp4#XZ_lc4NNuEQv|?6!`qTMA2|z3~r)*Z#-PqmlD3n^eOh+1F|okMTQqw z18*Piq3y@2C!!?c1} zO!%VN;WHdG|C5guTEk%F23WY1EL@CTTvII($E1$bo3uHLaH+L6kl?;U25A=g{k%}a zoI#;YSs&At%Y6WIjU#ANh05ol-Oz}%*?0|AZ7pF%EfShrx)yC#d@yNU=~Z6qM)D-5 zDeGyZV@CtPy*0%(QRCqFIl+5m{<6@ngr@tpurvd~>K+tky2I!cwIHP7>Tqg0-IgLf83NNlVN*zm%IlLg{r z5;#lT;*}m9`+$$vcwW+o&EzZGEY^&KWlba7)`&2 z-N#o34Rzv5UrUrG`o%#e3-&X2kmfP+J{~ht8P7|)ayFn)obV z(vDXMZKK=Bp7rkDT|?s9y%6#U^SK`g%X^Dwq8KLZP8I6E){liU8P9|9H_BoB(~h3; zT$e|l_%!@~$RNy%`};x95VtIXxY%|xp=RRIV&5p?sWm*!<srMULUuSCw4X9`2XLe;pWf$wLxVE3MS%{K&Um7=kyFVpC-_& zu>UiL23sXeX z5WLiKiBLy;m*hqT?=x!QY+eee)J3odv0()MN`ztb!oq?Gy;>i^r}t`H z=g^|<^g*$1y$r>Ayj{{-Tz!co(o7Bt+Dls~_~&UVDgx0&_+b7@mPY~~dsfvR5Hs>R z>yQ_F(PC&dxE+B}kG-8e8{L-nB;Mm?qBX5RA}i(qkLT86mA6jdgzIPtm2VZsnUP#Dj!sXlB1*-nRaDL|bWdDNjm1o-R4LLa zJ)1W4HU|M!m2BIwJy0##71<<@SIZ_-37c znYWZ|xu&MoAwP#zMV|V%zw~GKo}sKk6>l_c6D3?c2V*9X?XjQo^U2o`rfLP!WGlsQ zC-L3Q6WrK&i+K}YO7C5&`8Hmw@n$$m)ny9KEsZgZt5bdFmxR_9%rC1;eiY74~4{eM+n@EIS=IcYlJ z-b<*RXEy%F3_{O>re&CBh{|6pamQ%tqUr9X*ilW?^O?cy0(won*`78H{qSmRQdlnq zd6*!vVWkMb)_wI1Y?wTlO58(&_pU!LW4pAu%MF=J%MM~l$VJKh+*vdde^K&{J>br7 zh&Eb#Yo)k7QM0C~JQ=0sc2d5TxEiKLs+$hr70~a&vOV>*R%|y>^+Zg}IhOY>9twFJr8VGfGlTgc^Wa;!O7SkURft`eSYY^+lHu~Yi z6}4U+FVin`|0lU)Q=5h#D=J!5sy8qUu_u5M<~bLEc}so-Yqz)2C}(y+VVUwBdS~hi zjfwXWdxR-P+9^r)=KB^M=zt90yNfLjap=%S#4PB8i-(N_D ze~<{DBE+amp&L#J06)Xg@vMbe*nYTqgg3DPpYKSr7$%LU(-YoqPDHa( zqVXOLI{}mk9|cMsL%`V$S}wY;=ipQLxuJ|8hOv@fusK;e}e-DxTzV z%k-q+i>Sa9UVrJpP2d$1Gh2o6RX75eMEEAq3tN^I6Uhv=to5$J8V2WXQn-D--7Mb_ z%SZW*LI#*W{htk(1oUJ}JbwJN;y2OH0T1{8@mKtI1e?lu`RScZt#jPSAQf?ECL$#`ZPWgrAihCc3N+zo*{ z)x$3+p}pw2QER|dpvn(0`+qP({6qc{#%z=<+tAV2l5IQfwR4Aa2kH zK_@xK8D}>y8s33p4ZVMt?05Z~74YaqEO68;5F7iSqo!8B79MKmv(Hkd;>7HrdVOs# zH|ty-CU0r+PpYf&qh=1F3QM1m-2X2AMtD0->uxTQ)KM5ZanZ@9uO}5P`J!XGvFVKk@ z(l0e(fAX`m7fZ_GXhMs!#Fw_nUy<5@)bIGV6;f;7=Id>tPWEd^nEa}A6K~BqlT!Gt zdt15X{L9rr)Mi!CKzfq0pn}X>rje3z0P07pps~2*`wF`GQ&rFi^-WX|>oaGmf)1B- zYuJ%`^l}OGJ7O}UAXLz{BUC~4?on$Ujqbkx$Jx8U_gww)<15BQJ`@+0`=LlyTy%#@fTbh^jB5bSfcV;zF(@?*prxj0cF8pC-wmKz$K zk+(#r9!9DeuKxKV%LXkIc8!j}pklLH%dG5PL-cnST9Li?%2>6__fRmc=_E(umiSUu zWZlydjfT$|Kiq0#{A8z~(m-VFCmE)VJQop^^R1nBMW#nU%TCj@556qeum$!CFN|*& zr`QJFk&)M;Tv}c?_}Py<+Jz9N$sZdbZO*D1s9X?bKi&fUQF~$jG2o^CBZMi8VKc=S zBFwnqzifiSb7^kkg_{g;gZmGKCp4M9%jBuO%E)UZ47nq_g)qZTIa4($gPEUBX=aA? zmtg=748utzG=l?ZWiT@hxqQyG*<|Kh{F<*{(L-4*i8CgB0~DA^Y{lGF?xw|dJPdmM zig38bz^2d!jsHd)pc^nkXI;qdt1TfUSGHbhy%n^jl~-#YClu%4?E#N`@K1?C zk}WRx5X0D(mC@w#vyr>ytjVXqY$ZMyr;gpA~~ zS;NwjPddgf1R7SkLbWiD92r=AV4)P^V6$LL`FPNH@1wcItiI@{i!4e_mrz)7J6vxF zXU@uUJZe|HBt4JC2l~QT<>DB;g8!E~h69e6;F?5HMih2Yg6+W&>i925DtL;?<_! znm?SMVF;u6p(@s-Cb654g|w=S8Qbz@G+&sBE@dfz$22t#c{r+k&0tAUSrWqNDRf%@ zDYLMsYAOG??NBkQJmZSMKMsnaJWNe_<5?-+MKD0CIa*g)x3D32vmZoL^zU{7?o*!% zfi}itNG($Oxj)(qYU?4T+H&$Fwk8#x$H#YW4AbgZOMu#V^*KM2;-tCHW#m2?&1y9_ z1y`O6I>K{C%%5<7Jqox2^BS-I=JtJz3M8-G6lN=?f zmRQkd2>Xi|2Ni!;JZfK0 zV-J6f*KJ#pdJ1U0Q+&?(FR(Af{t>XIQArl`CTFk`_p>nir+Oa0EY!L01YMlH{LtzHa84TCEg`@W_0Kh^g=h`LB^z!S){hL++* zZarqo9K3jsrMdO;9|uZ{zC3q+mcIgX*^07n zM6TK8OlChSU#9w_@;pHjt5QrIV^Ac6E*tWIiMY(T08ZJ?gZr&mVN4aJmGi}9C(wl-2zYc)QL9= z3qZDD9Nk32waOWFdl<66YS3BRbI5|P9fCjm4uGn;>+AVr9In9~>l1<61ZYn`ym)#u zhLwO_m|I(XI`laPL2Q@rku+84%7G|8haOl0CKsA!6o&^YEy*u&ac}%}#s)9`9 z?us|8|2_-bqCP%636qt}KPbm>HTPku7iyL~i`0jBlAk0M!AglxthxYW79KO)1LAR2 z08IO4;fYRW)P9;ewn4zui_{MUCFXxHt|h6zq=&?Pt4Q!3m!#wBbH3!U9H%ifo}qWF zoX3UA7`%FxN}m~KI@wt$l1BAq#_?)(GvgTX60L?3N$y`BLhj+rO_^tV=(WB`-9-|V zh=Si{USiZu2!R91Seay7AtPp%*y_vNu>)Za>wM0kgIPVpnQ0GZKCUOs6y0E$@nY&! z0@C5CLyLDO4%N`>E$G#N_Bn4Lb_#GhQ(aI{+_&h*$@EC6MwVpQpUT9KOk!Uoda=3y zHrFGb!o{cuq{5*HSvvRQ#WMilG>-y-eDEm<^ovwEcKfL4kTAH)RlqSVD1=q+=FR>p zH&d}_ZL8epU4yIK@bUgC_wU`IRc=Q$tIEylrtR)SJmLS0I&STAR(XbtW}CsFRJ`dE z$OweHcdWnSeHn&wBv9f*DY97ALsz_}Yl17@xpZCeIsw{q3@@Gn{6HtTFcVp2AGeq| zt6>Gr8>`_J7(3Qp2qJqIpv2?Uq}L*b7l$iX*9LVu8rX+RRq(0kRv4>~Z5fZ2Q%`YT zAJ0L_ivy*gRnrG)jJnwlX`a#Y9y-GHk~ITxkUR+`gtEZi)5o+?v6=V7J95v%sC>>} zGI1U&|215a>dUGNqKk)sDEx6%ak%JhJpX|z?ykPPURCh1Lw_ZZ*-5;b--V@6EN;{~ z=bgz^@u0%SyIV^vYvCpw-!6yieqg{d9k?+IUD)O+F8JMvhv8cq;Ix!;8;qqCs+Ugy zt6O}lY?1y{WGGgzftqQVAy8Kr&EzV&8nrh8jxXEt4qBi3(m|6cR&SsRlarBnW~v&X z?ontfGu0PQLMF2D!Jq8#@F*U4Kzh`)jgwMQ>Z^{R!yc z{~5?abMQ`(regO6I0JBq#{GML76*pDawo`W5XDtM{v8*e^wB5=`7J#gcY>V241#l) z1a_1*P9FW0p^VoZAwjjmMehnmMGIJ3)RR-H@ET6J+vJtduB-^O@?=34~b9 z&4l>&qhI9QeE!m*u#W(f|2|5sWclFa0{ zjL$uVi6WgSZ%Jll!~}0Gzaz3e><9IWyk8?9U|aZMKeS-qvtXNAur9))N2!l$h|IXd zge_3J;O|-0g-l-qcokMnCoC`HBY|k=Pvu1vHl9N_`oiKQ7}6HG2pY%fr4kgWt$1Qf za2k@}bPlRy2sRGedqh;roxa#_J(CZAk$t}WnTH1WMNEnCa_~Fqe?|8{OPv`xBRBjc1i`7l!H#mpG^)YI^==UMy0_q&X7iCB# zS2uw>?-s=FrT{q7!-tmRD2%i>y0ZO9-NtAszY3Z^iVc^rE~HPu4(9|YQB4ERG`A~S z>aL4yA*XQnhEOY3BMA}0UG9@GL!a{~zYN$hQ6W=30VdGGWVn}}1xht=CN>!IbU_GP zFEkUiku3uV!vc4nLULn-zT62mumrl`m)z5MfrX@D!C)pdq<+JD8d|n>eA&$xnAbKw zIS|1lefvDJk*R(y^64D5l8`nEBs6)#hC2BvokESX#M+!P~IBy7E zz<~Pu0p`^k5dGW~AX2W*BSahFh8XK z#OP0040Dl~iLn`2HxqDt+tSY^aoY&&j*LmtM0;CAQ@|ZukN<}m`TtO2ejH6v*}}1` z5-i>oP-+G~w8l{m2oR%cA%wBa2u#i*vYZr}u0}9Nk{FfW0oQlOBTbr}>w3aET8XWO zXm)QpG^xG!oE{9pIuat+qokwgPZ3v`G2>-0Cs3#Zj77+_+Y zEEn;rw35KOk8uIm=ll{l@m&(XtgJVe1lCP$ZvlSQfa&@c!hi;ADquYf*iNaIHelBY z*j)x}wt(qxuK|f#iRedvrZx)I!((n_iRYle{hmwZMGdmcowdQ|+|}2MGmYC*SHvO@ zjxT;wF2eKJd{d~KRsVmhiW1`kB(;e?2a@oAX62@hVcnRyznE8_rK+Z)4zNRFog;YC8>eL7v6x44$5^EY7+LXW>Ppciq(b ziR){{YL(#Gf@2>A^2xPB@PG5DuHRFT%*i>Lq|5s)d2;SZucg!WNPxL8((DJr2p(u4PLX`JTJVjm_#DmDf$eSdwP&wpTtSaKEZTP>oAoiec z|L{5A&)RXs2UzXjGgKEjGfB2+3Bo6n zCsuIF+*FOgE4R?}ZpRy2`N>;N`LHsg(EYK`#kcK~@}V2&-*+M&>5;+{&O5YWu@$L* z_{1gY{}e{F%Nqs`MqyaXPkffM@~HC|)Nu(<%J67+8GWfW%*IL#z5ee?%pA zM-I52-{EnBj<*n$Q*YVj$V&I7Vq!(ch9KQrPF@PtZK3UIub-2H1oQ(8mAW$r)h5!vK5ENX3 zy1P;-Gr$=@&KV$zReX*nYvw@M%mBBOE|?2o<_z%No~(*KbzLgUZ1o2?KhYUBeI-R+ zhvHQg>2QD9-M@4wA2RB9xVZhD{q%p6>)@2h(nWyA0DlEDeLt1QK98*#hRU-V~s%K7MHxMXN8C0nbB*a(l2*V>eUm7=K;c) z$?UvesHW8=c+PH3emjF-FnW}^zIOQ~&2TS>#i0RpwY6#g@RY+=@I9M>5g=z0aAVvF zHvg^9*}I1nE5uPd0cg0eaCB^eywX17R5n#6Gors6Rur~C7@pm^JC2?{$kOFvHWscF zNYCTX)iA%JP(hdK<-5-EM1v%PcZ35Tz>$-uY^_9|5e+07mORg)^~AAX>y^vpx;WJI z=3urtI5Br_6^P}FR?(nr!D=Lyg;o7Lv@iKn$kv))P06<-Ripo{t*C|IGNMb4Z!5?m z@!X@=(Nu7qFP1ob&YR_%k5J%AH>0>v=s@|Y22clUZSk7hs;WVERm0?o+1;}Ik3#3NY2*_&P zGS?p-lPcgl^9{I}AaGN><}Fh_MWhi#$_1|R+ALUYV4i3Bpck4`X>6Qaw8#v%oKH@} zA!(#~K4VUIWR7SPH_yn)aXGS#iaiPNsIRUZ$+uJg1V9-!>Cw`eDtBWdUoL{-fHW0McbBxj)TJ?%o;;*oT!N}tjTZDl)@*r?V?}bHLqNR1yrDt zs{Qw)(WrpR;B$7yw_UJw(JqR4BI}!HO3aY4A{;G<- zC5%7^p__T>hsOR010RMWg@NNFpy@rHW9%Wu2G=Er>ut1pIO9JHvuF0_9RL;rkzEOf zfJOT)9Csu=i2fnC@h^_6$cX;Hi%NqJcb;uL2#6(Dp=4Dk^b`&kce8lDkcwBJo=*;M zWgJ3h{$|qIjjeU?|BGMEzIx#d$G6yTA=$Xt(b$Bd)Ey=it!g5~Ubh-!@buKwGp{)z zQ+0ircv}z>iM#Kl#NjuN>*?ez!c80DrUkjd6?$79h{c-7I_z}IyZW&7NWR^VY}^=n zOU{dIKl1#yeb$+VTBwG!)55L@8KLSSMNdFxs^oMMs12{yzhw;}1KjC*yqPq)Yzs$8x^^nZ z?H@9{^e^fjXiFLB^zIvEVW_v7V?rtFwKg`?v*1oF66{gEp=_bkJcf}q+&8WRyQ;t7 z`EDk|-E}yJ;WP=4>aSYsK#ICg2VzxY1~A#HD^JwJK2&-Bog!T|lz!S9nEW0jO_Q2No59}6&?4s5r|de?xU_&#bEN*gsu44cyx>lZ0QPKlc(GYx5fd$|ri(f~JE>TAq z|8>%Z70Qg5i=8h4>Yf`Z|!c zRdE>a;%Phq>LGdV&r$}Dq_TC9;7|ePm!V{DRvpWdZ%)*M2euB6L%*JB04}Lol+ZJn zHsJmT(ncBJKtk8;Yy+6_*Lzt+SkJ=gzQ8ktE_2?OX5%JP=&swhj@-{sY?44Oq+U+Pe^Pag4y}%8D_g;MG zr&8g&0esA{-Tdhu-HvX@({l`3WczaX9GI5QM175Tw(tt&TuQeZ>53SHcAl%X95VUc8-UZ;|IBrnG$b1U{18>+s0xpPu()THY>| zt=RIz>%NlF&3f+_(ZXNkzhRIxxc1u1<0%UK}6NVVC{6 zw6W{s)AF`?k9E$wl)3`Cm+#>)gY(m|+;~$LZj;^0HWO~8Kqn&WE>IM)HUAiUf=p;% z>%=ZG4yLO-8c$uT5rtK8JlE%3Cm$(RlM^7x;ltl9rwSHb80sGy*6OYru73zzO$|p5 z)b779cnlo}h)+8?7q z-H5Gy2AEQBLd`1XXeWQQ@cAmU8a^Y)&*S7ri`ZYM3EZ(e!Njj7VIN zpvV0e{Qjv6LtWBULXl%xyGg2v;XBM(x5BKEo+rhCiqvVZG$m3?*VA0zp)g#!gnst7 zgjzTkf&WwDH__`EolKj@7FV>y(D`O*b=u(#SE=CA9u7LO{?Cf0M8g$YWA6((&7gZy z=y3koWR?J$qiNrA=(-$}EcQ%W8_;?~_BrSSxmf@sz>Oke2!I}SbOV6TIlK)=9WUaS zz#~h;R?lGD2J;Rhr?Eu}wg!Uj4#8HSrV*Iq0JVZ1OmKXOOVaWDMEMj3lzAED@zWjm zW#Rj8EfxxTXsbitjqU^Ae+%@M`Ou?G`eWLAKnjJZL}3YA`LR*%%Xz?Je}?HPfaCmg zD#aaxU+atMC9-o>G~4Na<->wrf^fCX8^8EwK2&h8ueeNyW@+Yx{&D+{J z?{Az8X#s@1L*+44O&BxQg_2YX81Q=FVi02^vZk7lBeJ5F*4Lp}HIE^BV@pcA#*2|5 zykQ86KRwJMu2;Nkbf&Wn0SW^c!7Oaa{!|ORalojzA*kI<$IDr&0i#h)B+QJ!^_|3V z*F(pa^c-RQmrrk{UviMF)*Y^d6nrDbXx2Z94BCpaWtq0ti}gr{Vw_b;b`2 z?=B|WadvuRy)ecC6-A8a7y%_GsY@-YVP!xuc?3osHAP;i%9BzbvLxs7}kRMKZ0I2++BIv(0asR=ChJQJ@?)L z_3T2Is^#?1TM!|wSr>||Zg&)6ebHc%zvBLTJMz8Gu2|5W=4(avo$}g+*bz!nH45W! z63m$6Zqe|PxTAcU>G3$=fXfbYI|@N~#c!W`+Tqf^xM|SW`f**X#kK@9)FKI>5esfg z4l+`=hBHpYNh$Mlr`jL}Yu6gGr(WNbWq?}_)cr}i&=)b2JDD#$oVIvhC4_J;1;!uL z7-H`(sbQ7>Ycz%;>LaFRB&b`3J{#Hgf=9Q@#**$P7&PBdz#+|T2!t8ZV3dMoNY^6S z>(QCL`(Ohq=Qb$2Ko^J_h8KO`0J(raLcE&`z5I5$7_b()#i$Ro%f56{fFaHwY~7Oa z=5P&RMJyp|?$e0I{ZGm8g})jtvo!Pxh*N`DwYVy+<2`Mqm|5ywboE!i#scDXKZMna zs=JU*NtQ94zNIt%SIzkHOj0?9P!vI?N*go(9=1zQ#YM|t zCmv>DHBQ|CO#8xpGPeM{Gq_=ULdtwrH)K1pB{JjbSh77rSaV}MO~aVZywxZPmier~ z!URUvB#UfS^TC3RoSxm13jl-d0i<_(Or+S<7>ej7GTRr=AJRBYC)hr=b);JzXw&ms zrRTqFT5}f9D*%z^;5{^CCF_uDmC#!#(}@yo$qY$+SLRQV_teBN)s35#SkugcnY3e< zVe%NpRIZ}`}GNcc$Zo5EfX;82&VVZjB8S_Cxaa67INLZz?vIX*k(PmWu`dXC{F z0R+k2M1s%G?!;T)#W%H3_FBjvQUr0bpkq+qV`J2-k2XomCMBB z0Af8LwQsX`DlFPNgFl}Nm8ZcFEg_Wi+ztTuIWOJg1^*j8qwvCE-!rI=)Z4jQhR1E+ zuoyvOA{whe>HhwNDcz+0zz!RCmWF?k1=gbzNpxc@t@*C}TL276cgN9c>pZWGY+)bmX0 zxfFbFalJu^j-;31Vkl>P^zYmPr+QcBgRa_bW7x23Z-DItCBP?T!~r@ zP|^70ugT6@w8-9xfZ^V|jXa&W(d%(6YZupl0uGWgkbppc+4@4t@#>Uitm8pRW;Eb{ z%8c?n!on4xtrpOeWT!jh(d5FcT3O_sAW#0qpgus2V#X*n(S%Y|E<&Cdw9T~A;jF$T z&7(e&J%{n5Gyfh#*>FFLJPTDSkwIa=pcQ)*`Vf zb?1+jB0OVFme&c&=^*pJWbzNjQ>-rkD;kAy*LA!|aOJe&8 zaxf6>2IFD{%BP{$aimvOVlL2p2*z~)6x_8ravxyt z2`>6Gb{!x+iHI?xV7GXGe0~U)$I}Rv5YSyJ?!e7ElbUzRvljI1#cPz_@iuEFM@d~OQ0sLCL3h&ty@%{`%xX-#PD#s0WzSTrsk)}X zg)ELUVuO~g4&Xx~Hf!AhMg52%seKP{FShFxsRYAXZ_OHJ8P8wKS%$s|z_S~Amy%Hx z0mz=##|rZ`qKRh=@SwMf#_^LjgoG6wWvv~IcuxXjP+B{YLxyzVDKMmK4z@E@#ZP4| zmi@qN0ppBq$F@qyz;&z*62fROI4jrIE;aJf60pl}77&2!Iy9tefK;+#)yJTUR}V?Z zEIV5MWza*{T!@|uC32MQZ$2AYxW~r*M%rX7yo; z$wT5|vQpEquQrB)9)nq?yv_xDfDKiFU zcOyHMUAGU>(4KCVP#D6jg+AvW4XGAXi_7&`jb6oKIqFt_^NDi4c43qS} z`n&%?a3VWc?f@lvPcBq7ZCTdGYr1_RLq_!v43Q?}e$?V?tc1e(`X2XRldlp|V$|7y zQxr0sZpZMK7`(@V;ZImZ(Xv%Ywh8Gow)`|7@_7QEoM915W^%oLe*;qdKhO1=xl>Qn8oT+Npg-qC5AH4>33ev&M2ltW@JC=BQaE#MCdSz86xf2h`uoGP0 zK75|=i{8EV+Asm)iGFA}C;oDH~?;7KS%#5B7h=LjeZ* zL9oWd7aQRyBA}Bf<_^_K&z?pn3%ZBuWb1nz-asc+U$)W7rEE(lUjZ&ty(2n;PJW&3 z*GWs%TBUbx zMkmd|nxzv8)(G!X9fF%R9!r22|4gU()|A`;jVJ^3djUODkAwia_w&-R&JfUz0G(8# zR`9e!DRpU(WP6JgT!-`pq!0dE0p!4F2GS-$LR{!Vb;ChG$hG5Ki8xzZ^n){?fQ1`N z!lBKK%DS0tCQV%}iVkTrqr6A3{|Pe?HbRq`W>>|3`iMIEr!J(?98@M|>R=|?H)d^6 zs{si9fYCQws7^uz$!{QeO@CB9z`)H9*n=Mw0O9OPtg0(ekgSgdB`6ySR1TpiEGbBQ z2AGhTiB`=0h+$uG*g=_q=!OtPf80!R&4j07wOoL}(`mGD0iIS1&g#AH@EWcE&mobO6hgjn&=#MvFisYXiyFb33cg!#RsT++ zk-oyWbL{c>v}-o&x8%*)tyoP%MjzQviC($@i_RaVUcC<8<`P+t_AX~6R56HgM{zkU zRPFb>`$)f z!S>IrzF-%lf_OC@IrXTZoCTIlFqQ{*a|J5%Fw&O+kkE+o9>Vue(2!Ee6x5(meJYEz zi535#6@RP_EfZhVQFE5X&`2sq7o_vc{Pf~NNj zaF2+ih$u<8sz^~wq}b@)=W`}Yeyqn-)J2z!XUS|@zkj|@xhm=$Xs=O`y^~72oEM|vel+Lm~SiuIp!?l`&FcwLUkofFYEmUnwsIecHUL>87Me^RrhNVQo>mrF4 zycxg?|F`&+49BI?hFb~>8f92K#Vp50*^<`xP3!@X&w1e%Vm&LrC-GZ?2ZyUJY!O=d z74FtyfQfre5kA4OHpT8!1}D#fhycAd-A)NwzBJ)vi16?RljejwE zcmxA=-2V)C+!dQ{|s5}`gR+I76o#kaQS2XkIjNOjFdP$(x<8lRP zw-MN39ob13@^|k?T9YBbSwH)njX)3(>K-8gFB8GovcBPpTag-PM+#REucJCba%3-n zu)k-Euqi(qVTb7Pd!~B&CauAcEDqPpnCJ?ZS?@vBKpDDqiu#+VQJ&soP}T%WpY!?J z)K;N7!8%4UJ5SSMh9b}n*-b=F zCdNNt5(a_ZPY zO=1*mj`ka?l17SP6O3S&rfR{u8o??B30BPrR@d-UM+6I-#b?t1+`~k+Q@f39F_6vY zd``CD={-r@*<=pfu%-;f*8A*wy3TAd#vQr}!ZQb$dQcX+HJ3iHTC#3}K8e_p?tS_ngc; zF6kHq2lpWk*G8+?o#Ljz%%wK{hi$7Mw+c@=-pI&>e8CMfMal_GID5m*CG=R9SK#+5 zbR|Obc^*%eT?j2Wre@`A3_8TJ3o!OrmKrPzY_Xs`fu-;*mS^P~S)P&di?99UUk$g- z)l_nQja-E3W(VVNEf!q>d!ArlpfOVT+mXRr-KgjDf2p1#HC>)queZZydZE^WX#g?F z{Of>E=Y2x*7OO|_)Xm^pq%ahtmLlYxO+M#IT(}G@`<&&B?4xYqVS$io zDd35-mo1Fe>N|pSX$|6Bh~G)L&H?SM|3F)7V8V5JvFZl0cp9s-NbQ{bFSq^a8&57> z)d|nY3J<2IdCbpNC=s~z(h8d6O!cA_UqJ};#ZE6wdZV-0xyhRILSpjrQ&4;aVxgIM z_)Wwz?sc32RwWZhk{~`Y^yVO3&Ir`-VCl?_O{T&suIJ$inI_d~&qb;kXkx~T$pnX`Oc?wB>P05? zKYT%UCk_V!YWHruhIx^BuzO`h!M?-1$b2y#@epIJ)QCd0oHYXP3;A?MPFh%+7nu$7 z(aR@buqK7}VO6c55mKST2PxU2vu`Fq^X|yTI}52xz0;_i>R^{6$cs!pr2DlJgRms5^e{{M3YE)KrA1@pl-5XP!xk zRIx1Bc=+WU+WDs-OR=M2gy-q%A`pAg!xzCcnol`2z6&q<0t%W(a69R6 z?EoSS^zO@q(%p$uHkpCmdd_QPL>q-bc(!akB<8i)tnm2s@pydCdQwKYt!V!H(i8`JsL1b=@w;v!pnB z1?`iuRKw;fnyZ$AHC%8#dKeB!Qm$zW(^=&0CfJ5+Y%K3n7c4Pf0u!E(xgP{L_DXa( z9X$eZ2#$Mz!{=;l_)08I0u*Gy7)nr%kfP5@B$TKLk%dct;e7;`Fj%yG3`4Q%cf^;} zN-wa7O}oHJJ{#Nesr8X*U@FmiN==3_lerI?nH^#^nd(kFBR#GUr&@Z?1B7h)n*35toLkwt z5{6HZf>B)4*orTq07=iU08+lL(d+?<`I$!%iEqjI;4DRT&R+tN4wFzKJF_!>B4hB_S<&ch>!Dt^=DCe79rGS%{vg-hPPs zf>!E1$N*HkU;;P9?T^Inj=D7(M!>~Uhqm6`E11)}-5O_15TL~XD#;Yat|;~l@C9h^ z0et_iLP6or=7&FYpmJxjG-07PGW^S*>77zVT z8)hbxXI|$p`ls?7?Vyhay-x-5BF4cu=z(~y7RmP)?4L3pQwOjH1va4I`->nwO=u#W zDoDAsBJq2z_%{$gNt$--5|*YN4gsONVUNLj7rodJJQS)5Qa#(n_4fDg*?NVbtzfMu zEVYPR6AKjhjD2;Ea2N#L1V?Ku(vJd~NWYdU0sr<=PlJ(Xed=>okAkbKnZwYrz?7m? zt^i@0gc=iIg2&TvEx>Eq;G)!HfP%sHr@=aMm*JPWM_}?QQK}WvJu?uE`#&ne=G-{6 z0a#T}uw%=Nntw*OXG!V?6PKW(OsK!Q3WN533u()4h z>99MTUo500j@eiU!p@%;;$$$(b{9sqK4%=+5lKG;4akrnmF@<{4F&)5s$Me{DpU)) zp(mc;4xbdMu0cZX*85L-hC}~PFIguI{c5t6V6&Av^(HLjAI3$NjO?9xDVeF zBiOcLATsGRw09o(;7@~OO5kV*xduLtyzazJtJtA~+S%(ht~9m&tX3^{`qguBvQ@0U zLX_tdY03e*Edrx`<#Waea+dl;lO;{n0Bz=|LLMH{Ih5}dUfRjaM(;wcJ}=_DKVWM> zQCi+_Y1pdIrcSjNMDY2{mn8GDjxUW2i3D|<3E{qwS(PB2Ig)xmU(kZzEs@w865^eE zpa|LG)ddl$ddrWL;nQG(?2J%5igYMiJ#6BJsuhS!UB&9wxxwF^pFI|D-PTzFt+~ar zY;R9p17i4kjP>70d=g%_TAKR?@{KwnZT;~`^Sv4kATHggg1QkTw5JJ5N(OA+t*O$} za5VUXmkgPmsdpLfqSO+JCdak>9M3dyLsbP6*IT`1vU47*icnc>1j(-J7h(bmN7MGc zX`XLuMR~{LBp9aqg(?{+z1wuG^y4Nt9PDcac!mUm&jy0rpO?CZdK51ZCiqw&mcEHr zP-UKt;6tw7N*{X$M2pp?ljU$g`bc&Wo6;Dal_iGBCf?QONR|}F_11CMNYj8+u`NQq zE>LLhsZco_^P8Ayp&BSDdR<&?lgJ{~)kN~_lv*s2KGobr<}mVOS~rh_S-Po%R1cjZ z@Q_oy6CD&e>-_Izim zj%du+q@XbpL$7s7w;iFCpi`#miyo;PzqlYaz@{U&2TI5S zZ%Xbw*#C4bn=sr%(-fpL@;{S%yu5RpuJf)E*8Mll1`uU;^vlab+*-nk*ntrXu2mXp zo@Xzlk;))jBXqS`xHX9gU=!pv!Mh|V8Se-LZw~}-L$K*O z&p?=*CfEldw5w>;xj|M&QblOe*vbNm$+ zh27_#ouJL<0X(VZ7l8>4C8xDdBgiT2WAa35pe6o}C|(reMXD9jy+5!KxR;?FG{Bkc zkeyet?!gy`q*i>y7MK@4sRamndVm2tZEPM#Z5;HW=P9A9>(v)X#oa`*Yv1QwVha_i z-Vhw`GZp-Qe8Ml%dk9jbe0loRSSGQt4;JnSOxQCS8^KZ5R&h9Tz%OAIqbPs>)N|K^y%?%+dYagR^1MS6si`VoJm%b0M?|{7vc^Dt1M8qfAHyWjRPSW zfVRxy05mK;doV^ry?SEwgZ)2?X-^oc{vjh}l3PrJpDB^Mi*=(Kfz(#*4XiaZqhiTyMG07QD6SKwreM*o*M*DF#_9^~d%Y2JDv$gxxr z4bp9ztlLwn2-f`+)8)(lfPU)q7*Z3dW=llRhOfNH!TYUX9sM7E+0vXlz)uDS`!T*C z@-UweC0^rPy<3=4n@zNla5kdR4$HX7@cgMkFbz*nYnTj4Of-?>DG&jaF?zLe7uTIywj_+H=W_WgqAuU=DhI5+fqE(B76*4U?lv897c2 zrBbdlIsVu$o^91ZJZUR-G?qw?{g+NgtW5z6*lGATtjtY+cJT7NQJrBKA}`A*z~9rT zs5N@DWks!X0+TZuZ!+di)KHaU+WyPqSu*FH0^15ysf857VUpo6gi6*gK4)t%M(w4c z@Eor1!36ah8G_>Ta;= zxl;_wrxKBs?RaB(7OT4QEK+swq#9$;NCyf)8%Xi zNUCIxvdO$sG9M|{nR}YdUGVg-XF0ZGXgQ{ZSZFywz2uKh!3}H1Ln${2w?0Fqvbm2T zNg6>H0Gf6OP4O)Nz`6*QVkp`9-wjX$>Ja&21QOf$P7^ExU_xvNlZosa3r?k7qF2L) z7DPu2;=*%mX~1BT6>|(R@PA0XWAA*l-fkNWbSg{8`_CIibMO3qjF+{9#mG@o!hYaP zOL(8f{15;JBy1t9zr$fSB&=ouJrDr$WvZJ5Xb6CW*SE+V<+}|57TDm6lrFrrQ#6#_$CUn4{4DDbPh{)|T^l$BFrxdb)u4(@Yde|U|F8xhR4 zG#G`yk=gKVr&uRowE!P0Ir0|Q^giEiZ=TP2-+6Wf6@6W!Jn?8T;r~w}N>j>s;Aho( zU7om~k350*7lf{jjF6z4El;|obw5Lc`}!|3Q5wzahSt2>x!=d9%9TP3Y^cL_;%id3 z3#ktFq}a)H0)8lpv&NagLsB0IzLH{r)aWuuEgwXw4xKAYs+K~cSna({OVyb8fVXZb zBxwtk>}Wb%W5gg)!mUBqPLEKn7}bHJFe)wKnnPN`Wq6Y8SA5zTDagJnY2iktwYnpd zqMkj=;o?dB29eM1O|AYD>(^>WL$ggOw7^FFGDtlkq+CLT28%gMsWww^P>Z)up2g|~ z;i5=w#d~RD&XcroF^5atIsUZ0X0J@;ifXFj})ptjL*qMv8rUK zKOuNaRvX%I`!Z<0E;K(qqb26TZg2yqbscGF{~^4p7F61115nGp1;CWerWG&O7q=|buvgkCZv zN@SL{?3aC7rXldPLYX$<-Oe;n`@WQfo z0}saw+0Mib8tsJ!`?EvhI|U|jw}}ex&SF%I9Ybcr&ck2^_urnPl855AbS3ZkRacVl z@gxfy__S#{Xr8krEldT5;?XBbvAodA)mMS?-iVd#Q}vBjJizMj)g7+d_?k;yfMyp% zQ*`psA$IQ6!9I91l5YMkNyk@@V2hW|zJ?t~L5+S#aAjMbQu@E{)v^lR0|#|ktd~el zOg6Hpg(fCSwHJ=U6@UvE(7zNsK~mE8;yC@4lT`gb__b4}uC2X*mldiAZck`@4;XfG zsIuT87yMsKQW#-}%SA{Me^Q8-Y*aL%Grt7MdG3Ig^MFJeIk_oUZ8czvfK|OoM>Hh{ zOIp;w`};+OYp-Y#^|~@e9Rk!8wY(9vuOVe4YN{lKiE0*kMhW54#zGYhks0Ey$ft^b z)?(WbIV?%xM05jnBZ&B%_j_1?_u@CG0J9&3w$%)S_bBwl0nv>8TBGAb>18KUsV_++ zD$g^vBx_3Xefqdxp&uI?+URpGS0o;&O!*;|$uklo(jeaaLuXy)m-py;{g*t8)tlO_ zi~0lab~S?adyAxnYel!F5nvJDzl9~gon?=vKHfIj%z)igdTSsZ+}>sHv+DL0;}xZw zb-0PmbIdpqWrz^%GqMfn<~bxxY+x&ijdK^EY4F zovA|_fcoTDjvKN6PayoN@fQab`<%%gG7G}J@vORE^b4PPeCBhmIzj7OhTqalI3|>( zFtz*nyTrq;Sl$Swdk3;BeI=Bze8kmr5@wqc*wmhJvW&&AnDp zM~hz;pBiMygtX#NTvvOd^Q@8{R|?TUX_OE>H!jEOaz7>-O6l39TS%&x_+nKRD7qFa#7IC2MAMrYLx&qG#>+@Ay= z4<3534zBc}J*ZSBzZE)Mn?y-i{|pw+A;LlA3qOVwdB7wo?g2>*6AE8LSkBcU5X;$q zTs8wa*S!^e#(z zV+^HNN}&MakJ}M2!^o*X?sKmG9hJz9?~0JlG~`!NH+JIA2A^{#5`E4O<##dxZ5pIN zuj$Aghg_k(L6+d@Ef^j)7)(hyazDc7Ra1fAk)VwM;QoUjw1t^I&*9n|rO zNfzS(q;Vj0C7NT~FeRJXH>Fou3cgJ+_*6|prG$wUsjJ`Hn*W_(QA+S$5WxLcN6YG6 zRsshZeHe_gGRzg6C51wHzHLX&jM%pKEUO~)Fhownvk8f`FP%t3TO#`Eu%|!-_dgVr zCC|z2TAne6)Iq%4Sq|h`D@oyHqX&nVdXQ8dkiz`m5VWsyBZVO5|AvrFIrX$7hxP6X4gywtUGLih3a zLGD>Qa;DCI-kmz#{+sCrtl6S>CiZ_I66gQ%YbS>uc0>a;Ep-lDpHTV;!$uB83GJ4S zZ?&lyhKpN;i?(Hw`W*R5s-+<{Mex~HidV`YwWJ#b-hVhuV9)PGW%YaULfoff4N)6| ziY$X@Yau!SL`zrY<}F&jesH8hRW8Llp>0(8p(KUbk5T0>4^zI2hsq-LsF3ohhYiV- zaI!+1ZBq(KHlFotX}%=(+8;Da7wpt#Em*F&a)JLRyq6-_dPxcs%;8#ZjU9>~qHs5t zDcoow=Tq5+?Da;tfu)e8aOrk*jY7u?%`XmWh1$0_l?=^m3=JFeII*d8^N$r0?F@+$ znPuEE8~JJet;G9^PrWbrY|Ud-8KjnWr3MciD9b#W3n`zvr*slA0#d9N(WMbJSkac| zC^3&z&@A0Nc5b9#KDAk%#j2lpNh(rH@or}(q&8L3O4!Ng5vBp%^&q7@zn@YDCZ4XU z)yxv0e-s}+dqE{=DY!x8Uc=B1alFD|`|YK%#ED9z9Z%W1=V~qd3;;hqXFvJvCA|5E zH;&v)8F$O~uD%H3{!2iF4$W)AJt;L}Z-cH4T92dgwAD8D=SWhR+Kich)P*Wrx34V2 zxZUJ81FxWoc+6Hk5sxs$!y7FY+OhR^H9$|qZvpiXVZE4$-(#2wN2a@;DT2-{1bnc? znp55aKQJ_>mnAuDMv3Bm%$AF%#ZW)!Y~SZ=f^o=6%(V^S67>}RA8e6p8)y#XKIh}` zz2tK~LJv$>L=ByeOFdt+PHV|OxIIauMyVggQ!UzT$$6i4<)S7EJWfEMyLGsB2{4pB z+(hHN+)I_D7KF{m77>zzjZZVm61oZSQ0V%GEi<>zhL<@|sV1?0D-NCH{=K&J^1cg1 z&Ai|YdLDf8g?rYvd&4che4tYBh_=re3y{*Qz!*>@v#$u{l5C;)AGc%5;o5F#V@jv8 zv~imuP_uLbSJ&7{t^3KOGLdgWPIA9WFxYbci7j%yNO-6t?aK#;M)3bDzvuB=y0Us0 za-B;jcXt`&h6uSyhQv(jFL=Bsw*1#>EzwDN7OOGk^f==n-pN=_ST35>8c7Rpqx8ql zi|J)`B)Nv#tA<()Lv2VI)VkVI(-rnjP{a5a?wsKN3cuudW*Eo1Ww_2zoCVh_EWyjG z?B#>~6-ZLJc)A2@3c;3TlA3}1B=J9k$EuGVIX1XmT)#kxBTHdGzT5P*WkI*Y@27>= zOZZ@zM`ul?3V*$DRIEOSKNNEI-*=@7|A!>m3ttOgJpf&V0&|Hg0W#e0|Gu%0}N z)fyi-g(2T4*zN1^RR-H4L9=vf2Uln*_sFwY4K~y^;GMa0 z$a+v~%64(Kq=nZV?K0_TJ4#p{68fB1ck=h5{QiaC(z)nvsGs^zbJMOA>L~06w#0R- zH9&}uFO!sKxhCQK_;xA8 zXfnfWscDm0(1sfP=?5Ci5AwSizoiRV-_XC+&|iK<(|43X|NK%aB3xQsf{lg#LuJxj zjQo^&o;-{F{#m98Ui)hDdKomoO{L7geXr#VuVlJGZY}w_17h!GU5rOz1uAK!{>&U} zl0ygXMJlQcvTcOy(}wI!pbM_sn(VeET2*G$A2hRC^R1p)%@zoIq2DyQL)%;3qi;&F znOPm!URGH*2Re4MKVQ=NZ&V6tm|wIlWxb=RBPf?{P`@tLM4mT9w&2~aDFBgqk`!)7 zdXRR!HOsUTUBYIoP zqM_v2atO%j6X76-5EBGqL#M_s;dzdp(F+VSNh_%T(KZ_QjV`fO#o8=t;JuU*yY#g- z=TL*xHUFMgEU{i&%d)_4k)Ia$kRkP@;45kTD7J}ZklNOgq)u%qi&PgOCpMRXGgjf0x7Xy=YU?@<*kqQQT5Z7yExb ziK)7KRFv{Cz}@W?fQ>pHisH{mF4RZBa`2+vfPS@VjwJJ}2n#G$?V460!dsJBOk7}F z_!1;B;R6!ywFu|f#vmkH(oe1@9^CCDcNC;xXFWFd)1SP)`PJ2v!)6RB`>%;XmP~aY z=sK1yi91RJGT6MSQk2BV+LRl2`5;AIYa-VX43XOyc{(Zzd30H--3!Y}Hd&1TgedF> zfxz)4$Qd{RqrGYT(|iH$ulf+@Xd0kah&K(XXPTJ7OtaUeX-Dz z{F}qY>z-LBLJ;F}QAET45B)1amMGO6OqnXi*E{hlLSi-y{u=`~We<@w29cg>;0VkZ zro6bMMwkj#)7gqq;!6_80A3c>O4oS@VK$)i4J*TS{wa`oUL;XrJNGNF&4bz@Jmiu;=FRR8Om2c<3;fNo0O$pdlgnbkHIf99Q!ARb8jAJc7kjwB= zbv2a4!tZ(9WE^QMlr16T=TXoI;B4EARA6v6zRA+QZm2PYh4uHlH_5DTwVE@|&m?B?qOV#IDL)+07V)5Rj z1@KRFI;%2iM*Ezd*PaZ#$NQWu@lLgbZ zEwRseW)0ENT(PzE3)3B$wYv{g@Bg9M*%Qpp8o=Y-k*nQK;Q)}&ITvxBL1LjrYVSNM zDU3fyZY+{Ly$sCPfB~y2{hTfA-zXJ1eFs04a1XFiy&>>LYNX(95r=CyQ3ctYXCj8R zb|y^_#Wsh<_GI}eu)vl5G|9k2o2^VJR{e5T6Wt%&3^`n0)=SPS2RFkp!dT$RR9@7d zoKYz#!}4*%Yde;;O~HA|D&yemWfZc`|tmmziLwjiSYb*C{6_GvhZ5Mq^;yh_NkEvhXjD|ekIT747YIm8O*uG59#T{?U}!}L;u`1%Yl*I_5a^APqquYpCwy#u*@ zS?J3r`DZ0qtOnre`Hi(yJYnOLwIun~vJ(ZR_qZBreQT*Ag*yM{K7i`!U8u62&w%7#&vRv@3}Z|t`* z)?8-=tXZ#h!lE}1k8v!+2QxC@c0NuAPJW|y*ICm-8#85Tx*ckQIWr7qbLE=Ew7j>( zDFF8KuAmV^Nkg4#+bv?mv*7iUGrW;4j1l|P^9W|-U0#zweqlcc=J6c1PtE?q#*Q4W zYYFB~%;^eSay!ozM>12Un!brD)}@h_T%gBI+o(1o*iurW#chncx70#%9dmMr0T7Y6UO?_4Mp`0BEJqyQQy0K?8L>HJmqD3LKh_R!3_~ zIq%NV<^0b5T(z96;ZQF~jfGq35yZO_7aFj^0HY1OXKdh!IobwheNM>=)Q5bAOOl-m z3?F2R9X!&ts#d|24OIC_X+AMj_zC_4J-jTFin~^x(9^J`(9_@WlY#G_8WL@YO4P(K zBU23#^n3|@ixYj1>@Cf^;=gMk8iN>A>2I}jnWNGd!T(e_@N*Fot!yLJp*(t{C-!q> zbwWz#+ds9Bx1FC+n2?KeTEKWx=$Rguv<)@J^8<(I#%tLIX+T1<=qteYC2;s~W_CSytfeLSg4D{6slimYMt60JI@icp8*GJ7wBmp?5X*S&(Px8Y|uAj!`*$w7YBNKB#XVEBoG4FrZ!XphkIo^3)gYY2`a z^_Ju)R=Ie3(e@XqS0o1e|M8@q*x{HjICi5=LJoQYX-{BcpHJVCz_MwE&Ycw8|7Q;X zc&2(DxdBY$m`*T⪻L-xFs8AiMbMhOmKgrhU3NiB|FCF4BSA*plS({-h>5`u8dso zb0&c=G{Nls(V=ZIj{3;xdG+tjw}Mci`dgm(9?iTmin+mrqE!?^9(`vCNMI_m?>;!n z{z-N^qyD_%h`^VGhzQhSA%U~V-_X3Uys(wmUylR;6&ETxUGo(VlM0s9-vJ)rX!2f6 z5q!?PoDU%A zViI*h$4zt$B`@k-q(J4=?|5SBAWb5abXpD9nK+>G`{k4u)lV`NsvdZjlo1Ao;(fa@ z4Q%l;zf>!xYpE9DDbo%z_9gdq zm1mbRbGDoTGf#!m;N2Zu$mjtwq`59+%R!-!C1as_U!Fzk7IJ2qUO$4YwK)H@l&t`Y zC?k13>II2ENn(_5`@P^NhpawmM?WjayO!`9K7uuVNU#+QO`3#`AU*FFSq^)7GFKB= zW2*=XjGa05{R+1z6;oTA{`_$ z)gyn_(XfktN5I`bIxW9FLE*5KC;wwsvRDDh4r!7ExmZCrrfcJF5OwHST+<7^e#?XL z834_}3vBoYHp;h1HWLT^adyiw%LIWj7v0hYlc1fmCkkjMixVHqQ4cf($NtyO6E)!e zXCrEqx&rGJNuRuju2kKN(7inIaw7y|5zJJtGU7-!BODU(EF)TvXk--n9cuwp8?C;U z5a%&B0hMQt*qWDv`r>zumk@Ya^A|gI*^5eHJ3AZFb};K=FEH33j8gqA*0$bF*8W;J zC&If3Al{>prgPrExa%uvE|@Qd!WCWA$EW4ZkHa=(=;l8R<;pXtXC^q)1Z74p%SR;v zp(LyHwL^IfQX@I882`}R2v1e;y*333#tp9v(9Xb4xX>dWXEv)>#ggFgBv=%=P-j-v zwk)rCGtlPHa<*cc<#X0vEQ9>?yk8}9DKMh_?wt3xX8?z`=-F#Yi?Y!}S3Fq%`^NaX z;p~P>P?@T^0sKR7ICbSQ&7UCmGR$r`!%G6;*FM*6>&3L5J>=*U_j)xU!a~uAqs_zt2AgzG;wwEZQT|{t7z~m)dVu@eEu>Y!_4Pc zgODd$EkmB%Lyp`|h#|+b@$P9Q6>OuFQ@?qDU!)@hrO!EQku<4(ZNJ1kZb2EueI_E* z1LKe+o~?r=&tTG>OlpHrK(G_DL8}R(aQ`RBilU_p*olOaioOF42j*yj{Ii+A2-dH5 zxqzIjRX*q63)y`^QB~&KA<+f?c|8{Rc7U>1dNHFC$h8Z;v`a5sy#_FvQS65-*AM`0 zr~L%a8^GgpjuOH}>OAZ_Y-Fp6CJ1nM1Gb|yuu1~f7+{l3>oN9AnpYu*C_UK#!gEu3 z6du>A%jb2EtD<77R77m#ZEv8(7`qPs_~AN5x9$E$5*LcP6IZ+r0Jrl+cRbBCQ$2^d zj-hcYXb^F4v<`5FsD2aU3PG<@n;saZwpM?OpNs6QS>)m=bT-c|7b4Xi%rZTLlz3boa!gWd&r^Ug7x z4@8V;XR}ITlcKeQ9@Wy$wUtFX<#H_Tpi?bD8h1-2PJK<(^YB~b8IpCu3<;G&-Dbl*~8qR(|-81F(G1`NYR#;jF2jvr@`!gl252Rc=c$6EZY1 zK_j@_>P}>%U{`LPqNLoqPc`K>6iBfCKR-}zPY{TGQjLJ@7pn}%k9(KS#DeBkd*f5(P=b}d0W9XT-gK}lq)!)#5Hvl}DxB0({iC|1Cn`S*H{ zZfSW}9!kr99iPW8!ma)=coJ^;Rz7cApK%*){MOM=Rfmeesiwr9<20%EA{R=GF&b# z@7bs=<>MoI_vyDWwI{|ssXbD=r|13MBQ0+yc}e5?{I8##jM&mPnu+j-Oy!!c zgLAvah-(|AdZ6-W>AjPozvg|m_@4g>6KjWTXtxl{6~2L2;8nEO{5SdE$Y6dxZYj;@ zSB-UD#>9m|2o&o39$*5FujUZ*7Ww}v{>!SKOVxjaqe=Q`eP1ZLr%Uz8#)=%QS5!^R zAv8Y6jDQrV7(np@d&3X#91HD91bA|iiF+&Fp9g86Ln$D8D6P=s8Ch$6tOzSCFd1`DMrx(8aFI3$sTb)H`t9UNXp5EIwtv8N;*7_{>((d~6ZEHE6 zyLkf-(@I|?1WuoVm9Db3TgmyH%+2HKgFcWtEM>R;}#m z-E7mk*$9D!DQuX6r`*g>-B_^Y1tn4;OhaW`V$Uno8D^JMPw7^IBMx&wu3HIKSvqX0 zo`J!CLsHO2XAfFWOOv2{PidQzs4gpI3ajv#^Myxe?Q)klptA3KOZ)kKQHIuN-*jCT z=lvk%U&g#!WRu87fO2;IYM(3Xa*cE{AY735Y^={Zf-rks@jz0+~l`VHt$iLhcN1Y>% zX5>~mkehkH?(HDVm@~YqPiZa45ufln`V$&51Jw&Pq}`hzT--U9MDbU8pN03hS_X5D zTwk+gGxWwBGdpv`<`9-m`GC6UyHa>)F6ZRA*s#gKZ@}^FEKv_g<670c8bs$yUjsC3 zk*i=y^zA6B5t2?{ampB$#q@d+ejx+uGwzKzs19Sti5~L`{1R>k-mlPWeFBuIZu|8@ z;GQLSr?NAp%5k3h0ys7Ma%?w?j6a4x6Vdyk*9@$9gDGE)ska`oZSI^PIDM0_B)C|k zf^Y0B=?obxxS5+L0w+zO^VNNV5bLvoFjjuWPP7KjKY*3e7ZKqQA?U$ch)fksZf*P* zaSB@eM!l+pU=T)He>p3YCc5rYwO@j>zg*P!@mbUm#Pjg|nDB1!;TLs&mR1v0&i>&; zUzKb=pkmtOHII&tPN?#pnm)D1bl#}aaiHYKWEcY)Vc+A1#CsVbwN1G~T|H0VzI5x2Ga@r4jV;0&4gh)DQc3FlpRzaFyuz4=onj zhM}No?E#Q_)r~ci2CUV;$ZZVX;qozJ9f~A+*AQHgZq3%|u1xc|3glnMv=?b~og-}z zp&=dj?5ct80ddsAv;qn6goRgHdiYw~rm=l1)E4{-(4m(>#0fD1*&a9_R)$8rmFZSG zy@Kgvou15eT@vK709HnRR*w=F6fISaMao|W{M^qR$_Jdc4kZGtx{Hrn!xwNieg6OT z0?t)0!fyG}e|?UY;*eT^$CwK^`|Y_9Sup-Fm!5ufslI^o^Cz^He~Mo!@(TX?YP@>W zzg>E|^~T6cPv3z~a7m?z#h^o6$`6@Mk4NNMei=-ruf`alq% zKjvQ)tx5?#ki~lON}7*rdVm(@zc8l%ZG&;Gy4-1&mY{1G4^YngbAk@>;h@nwE% z9WC>_9@jFz1HY_L7ykNcT$TB6GVgmsq|DnwPbqUdi~Wntb1oH`FGg%egv@t)Fq;J=`3IiD(qSkLp-&Qu3l8sQ8Y6N2JZUwlI?$5>| zr6^>cd@0%#oI~()(UeY%$YnyH7a~?Mo=yO%rF;Q~zF$Ho`>;e?vCXrk(^n{8v6g2P zZhNs8!w1g4orwjV)2pd0HM#?le-?)p(onro&KdwqL!>*9L!|i}JwPlY9#YRju+uzO zYF@87)QblFI+-Zd*!~RcP&;Dt)mY1PnociZIz^{vF`cB-A2NNhPS0jqZcsDzH?8%; za}XfrOLz^QXv2Cui<(5R`=OCRNl`o*c+X+<0D^J13(G^A8I}uWc2q$tn1_N=jnSZG z-9dOPs2&XoDGl$f8zo!dlY{W>34cZSMa|%g=6FFj`tPYI`47eEOW)s87`XG>8k$%D zo-FBnMZm2D{Ep-8W96(}g<6Rm_NnL=^}Z>Zk^atYg6zcd3nrgQo)#0FOG_N6u1?|K`ZGtx9b<7?kZ3OwV2l;D5{-%h+)i?lm#Ci*9 zw(rqGaLB8SOpSKqS!#4{hSroX3D8vos5N*98jzm=V6_pH>94mC5hZ?ORA|*~twKwz zzFR1c8u_HjEhMCKnQp4nnM~Ky>8qIL<`7`&G93{8tWk~WKbeNPLi{^;U>xBwm_FuA zP?@`GiazM`SIXC~`6JNaY-odv zsg$@E8S^dm2}*@{KqjW_Hucd?umOe&)XRc229J9gzJGrn$_J1v>>>c;T|#^wni)q>_Tp~=V_h!#YXtRFBr=~7R0;8I4-vdS5=J;G zAi%W%p!0b5&DktyFzOPn2j_A)w$kyS@78vy zGm6XXi{+$a>(8GdDQx;bCIlkBzIdiYF9(FxcUQ8I8sH`Z zWW?fcTM7TAgiEZ1|Lph3KFy&D!c0&pYn_Ke)&D5e#I}zC#zj1m88CA3KOBt@{3v0KTpE2m5#)k zK&IbbJQJ->yE$c}jeYgO_I8{9AaRUQ>FdOfs@C$0`w8$1P&LWm&R641Dnku7sU$VX zq|#JxlS);$BPIH8a*|c1N#&_)Ov(z*F(V`ttJUY7vhYkQS?cSDrTw|}vVlniXUl`eh%Waa%k>T+?pXli56h zCK}y;n0-Md=!=-W)!dqwppKwlMVCALO>i9BhIdd}#-DCvQe<+w2h6*q$3gtJ;7M2N zI_7srKHLZOqArPW9+)9nnG2}xFUK1+ya&OGOYO#VB)p~OF4o2|3j;C7fV;_(W~3c3 zw_TgP3WsIca2KhQOVA^JOUtPu$(+m$c6@rmZiz9w7S^}obOSdlPhIR%eL)1#K#?Yz zXo&9LCaXf~bLcOZTW$UlEbCw^8C7zY-m-pyA3y0|@+Z^>RssiC9rJ(#*P^DuIpbZ5 zKCX8>Yw$e+d}JHEdVX5QdRW}=j0GQTz;9>w;de)2@$@#F?nMwA#gkz`lgTLuG`RyT z4ByPaFEa262L2fIa2fm=n>8ODG7NuWs+L5df#3PHmc&}ZW7XGYJsJhS(ZJ^!_<06? z2H}xsvlv;L~WLMK3oD!?Jm@n zE;CYE3wXhx=CA~BQMc|3mw=ewHzo27TTcmvLux$Xxl3nV8wEenz^4J;u;1S>)saOm zVv&HEvOSoixA~_*MiE zhcI}i|IqHY=phj=5nE*jY`v-xfNJ{}KBum_+ zE5l34&jlUHc`BIG6bW|V`G2(r+B_hP0FdxT+jw@i&F|P5`-~zmpTpsrpoW+w$4OV2NpNr7$Ib0iyM&NVVwdFCqM#@ z4k3Q#4G;2jMTnoeNLjqw9xdWOBHX%Epc*1^?$Vi7C;o!ZEwG+pJ`uLCNAVVLN~Svz z(`uhk8zwTjN3Fsyh1F>^B^+D&dQrv4&T9Qh)Wy{_Q?`)5v1QyX0*;4@wLO_Ug37!{ zJ;dKwK+g}Nv`v5p&)F~BVR!RH#;mdt4ib5ydYzm(7-wvKj28MB4%J$q6)N zxzqW-Hu70G0nVzsHP5O*c+4L4$w)F0x=}L0|C89EY(mpgA3Zf6g=+zY14Tt0Yv<|2)RODY_5%jqXR2(IRSw5&=gJAaVYu!0%E0Kh*N<6Yj&Xg^=5+lk6z? zeoYW|N{Ik~Ed;<10KoT;g~WCO{}IK6(^N)CAH?f79>su?#Hlq!_~1j7|E6gaA@;un z2-%MYfiPa6kp$=s0mAUM4lkwfT}YrWy43I0vCNJuiW6+^I=D|35h@<+X>Ux+&8+Y= zbtzY{fU6a+~s>$C1)H3Tn?$6JJ*l=JHbwc>sXp!1ec{q5f3p!4(B-}5o= zIf_V&3iYhOd08eWbE8MeOe1?0TGJ*IH+h4(TdK_63>>C!D`kWaxYTVj(TXC(H49;` z6olHY^)J#Bvv+0I<9{>Ym_3<;hGw)*EThKn>W;EOn4`mED&xr%{4D_!aCDErw!RXi z){!a#*$j}l{~P}?k3;njspkMMdrP?gg=ip=LZZN{34$uBOSTiLOHNRQ*86`naei$d zMqBvO7nRV^X&-D%ea87kmUZ78aLj@Eiq7>MhJ~E@gOopYz0h~56fV(FFRlP$eK?EH z;nqOk3J_E*;3W`^jZ6KF&Ju(7SwV0SI5pc5PXY!(4_Vl@++=^^2`DfFAF@Dx5e;m42!PTCa26TTH%}|X(2z-(iDb2M z2wN}~M0+1Y|6(haOk&#<;lC(2DWESTswTtQ*jA^Bwt=usxN@iD5pA0KZLJOiMm=!f zxB4EBe35|w>}u~xGF4^8BW?F$^Bt5Ry@P*3NStjloM5u~~(yge$Jzm-IR zKxW5`??Kj=UU98agoapPK_7)Ylz*5=Uj2$+F)CGsMcCADki+#0Ba!WEuyAByVjghZ z_$cs@G`Bt969gDEao?H7oT&&BWgc!ZnQ_P*LFn!zu z(QA6Gw-3-*lUbTHTx!`Fy;&Ho1v)f->K`0xhJt6HRcROh0l;H)KDb3 z(6bh00MSIwJ2|AzAK)4nm`2_u$Wup|VOn1XI}>ZRNBz1A`-$MAj2KFAqRD=srro^h z*oEX(^&HC9WcxaiqDje;oad#;YQ~pQC2syuhfFl(OB8P?K(YOS-_4w^&3T*S{w3U{ z!{~82J$_%~yQg7)O}2M@jXN>@1tsrmoW}GTqyvtwkI2_2ywol z3}`NXkbzE-tlYv~(r)7LZW>ui{rmC1=sGDp32Jb-&ff zsd6MI^N8Ks5N*Z|8bt<|Uoo2oV@bvQhy}?9fLZz<4P~!YG(a&$J^I+O$54!bh!o`{ zsRcm!n8wEp)xbO!P`PTGhdko6#l@*j=heU&aNIG0rqhZ4ZP&_KMy%oT`>2?)tfWW{ z{`{R|YlMKLo9{q#ii!ms=N{rD@Q`v730>_!Rjql@9)5Lzp);B9`rw_`LQwl@U_5Gi z?Dc|L!0|3n1{`nUKeiXfp@HG>Bw|zVSN1FzPXu-$1B2cJY^}Q+GBJhzTzBkV%elsgaVzP6CLz)IjsZ238+#&;^|8p)9>p z6-C*TB@$bAlm_hWv6}BBn;VDM(F;jUPubuF58KG%S9riKXfo`v%~P9orF~Q<+ZiX^rP^(&hr`m) zH~~ktSVr`MzQE`I!1yRW>5CX6=`%pOX~4xV;W@f&!g5d|N2wU0%lFggz9LuY{acB> z#c~GLBXItghmY`eFeFa^YHI~e0gT8;_%$ROnC`Ewx=>wj%f%SV_Q$5hq zP^|;1RO$_EB8iPs%aN493-g}`=h?bt9(3nIEpw)9km=>&x)AK)AdlanlbTOv*2^a| z0Y`mQ0V~8{vRmc>cUnZzg}P`Zalp|047WSr_{9yj;=L}H*K6c8mDHPu<%U!(8rUEs z5C%+h>p6?RCaks^fDoiQ0&JdwMTAAdmjebq(ZDB%;CTYn-}kd&W=Z(9rBIxGvikK} zg(pDJ5uGP=7vq`3dTKYzbk`_2<9P2dFcad!XuIZm%J0|3odiQc45%F0(Q>x^J8CZZ= zkpUQIiT7osUFzCr5>VhEtg1^*pQVq>dF>t=tEef9d_U-OT=J!8`=}PI5=zW`gu})m zsF50uLXtNPtZ+zAC`$pB@(hqsxsIl2jB+Q_wo}cRpT_j{La98J=`C-@Q1vvGdJnVa z+X~C`(1JGiN)X%Rm9#oF0OQ?zxc_-9rF=jBvx7a1KF;4q@Jp$0uiF%**Mb0S=_tM| z^)ABGCa>fL|Rn zQZ>a*UI9~9AF%fwSb!gSjGkYFJWW95lZslh=t2GUg)AHtVb3kA_^92xkbJ^L7tpY(uzc=ORka7ExeLbN~D7?eT{PC(Hs}gM#@)^xFTJ|l0E>3 znn|HitKS0%gZ;_5Oh+(8?xo?!5m?-Rfg!AAs|NK!kmCs}y*S+GY8MjaDABa;Lv zrp_QHBGicWB!`?dzJu@33V(fk5zbVFTI2kS(e#dawv$V398Nk1{5~*YO=LUb3;9AR zF|{v9Phb{sE7o05rQxG37?SZQP z2z{s|UC#{|>ug%GwfY;$Hp8qex?;CR5Vtyt!`%M?=$RRSPokINOTD0=Ok4?SzZ4xu zHjUX7nZ0#8<*REMBd2aBshl%jo`Al(jV7Jh{8Qlz6#U}XM*IE6n{RlU^_ zHTO#ngS@`$1*0&N>X4_L`B&p~4ELM*>1tr}zo-@E3WM3DijOo#=%p#1plXC(jNpso zsH!e;>dL*D15L$ZV zmmc=W&Z~kwvVN#all`F{d2kq;5?}b?*oX~k$BG10qf(EQ(cI@awMlwpNuNXPYFN&N zQ*HF`b)td(Im7zbuUj1bvwP_fLkAz-EV#VDB^~4qI_UDr0^ATFbE%7{WDu84P-11v zx?Li8R>>hfx|6(^%}#=D1Cgpsm;kE*VBr6bfU>eVfKoLPd%>oB;&5Y|{~hG{xy0dr zHQ?|9(4}sbW}KL)4(@KPmwrU>HSPlkvC{A~bE&SHSz+8ilt55{2Dd>mUuBpMP#+CA zt|kv$Iqd-Ka_Wh|FS-{V7r&J;k{inMkX(Y699tn-lqseaq~~_gEztdL(*h&t!AFff z#fXgwb}Zp0;CSPHVZ>STy_H8b4s7B2!WdXFyVoZA3l>3b*xN!iz`_I-1HU%?-7BBB>B#uEqnYB zY*`W!5iX5rFT|5^d4Ht7W6=sDIZYMKQ~hdsnil2!c&BmN14Rhi&hu!-s2VbA#mi3}6U3wXr-3O`$ zS6%6~?NZp24pQi7=4~hiNpf9}Nqjp>Z%xhJz}Gdl;qxJVL>dGXUjt zJl7>vH&X?d4R~Y@Ea5?PYB2&1xOLY1BH)ewqj=W|F@jky!7vS`WuTq-15SfA4ItS6 zc7;8Jw}65k0%qG#ZU&U?IzSJlIvOe1U#%Vnvodu}HFR0mGZYx=S{AM=lc*~l7PWG{ z)b%&k0x=oe801)XH8vCnwh;s^yarXt-=oWvHmrT8mC3@Yd$9b zKjM3NGCQlWcswV;G3s30<_yje#J6XhvSR@CY(a@rLtEH=rZ)m#gn{?ITU4cllN_-(+ zF-8R7{D+_B>#bv%F?8c(Ydo65>1&rJN3$SmtID*R!75q(q@7+L!rA4g&XTVX*Ki-r zg67}4nMJig*Q?WafGoRk(jeFc9?>6g3`I%*Zk+#%Ujf8$hGPzybgBvaIAMyd-c7=1 zRq#p2y3_N5RLv6PsK9n0ca{#oN>=7>_uYWSQfnXOyX$Hj2h9_M4V_@bFbeDkJz(kw z4Els^xCzj68`=?jkU$6DF$k`{dk$XTw#r;#{o)z;`T@stpkVCX;d`VJcEL}g>}l^> zoF{iNRkC}tX~%tzdsj(;we{I*sLK(qh){g0yc?Sg15S`L^*(aw)OG|zvTB6&v^216 zb353T5|M<)OOEQkjeptyV{Dtq(CsWz)nzX}p^sr>^*1Zvl_ zpv05FaWWXPx_SSGe4&}iV8f*{`o^i5r-Ms95~{?s99;>gdQ4I=V!&C&taR!wy_B1# zw&X(A@8kYEQ7MXHDRL@_&vUbsJGPT^^SU9}>2YE80`3NZRrk`A?8kr8HE^QEhP~9j zSk82*r=-+JfQLK6w#9;P7UmE8rTdx?ezX;bU=c$K+ED$^9C zG}QXIPR6LP0yk&$GVI(-*# zjYB5BkT+miHk=3r{YoPU#JU3@m1uadz@CKqa5ehn+n_HnCW*ofKC(d{PU*-6n@v+XX0#>wT0_8V~42M1a=3x0%j=myJZ z7x$k9=NTp_#{wls%SEUdkLA6K#3U6e%9*~nG?F!XO=*heCd5JWBa3w7qciRvQP#er83HC{Q69PeqcCH%akAm%o=ECAtwc7v+TMx$*!tVOK2!rI%W z?-9`({Fy=xDo;R~2gXu|BG;AmZo>83 zsXrR&)c!0S38S?q5iqzf0&L;@AN)7k;}yeDPFjc?bYiU>21V3I&>02t^3W@6A;^y} zz0K&5`usW}0pymjJ|fH;i>W}Gm9FeEQLIpXR(^~I@TE7X!!ZGs>LT;6_42_n{GWal ztCXh#RR;=zY6?EBL%I>EKv&bm7Rh2ghJ#|7jnXB_-~xrY1A$3T9v~vA+r6)d8l}n% zOgGKqTxb%v^QlQUPA9+mMQJA}lm37hZq5~A(B%yVn9b0qRzZEL##hv*m%*X1dTbN* z$*|_-aqV(T4Q{6rSMxPvzVd0v zR@t+-&%Z=y`^p)1q1xyyE*OKWdCY34vr?&drRB*in4q%~byg~~PP|%^de9h6rt85YYq2vu=~|>o^1sUB(!nlsE@E1j%9y1*)tyG26#j( z*u4j6K7Hx8)|I*E(>5N|4;3q=_iHXE%Pl#J?_cg<>_Zg6o@BA)M_!W#tX@jc8TcZn z+Rl8Meb+)&4N{GJ}xOJO_v(}E;_%joT5}mkG974)F~tb?u!AP z$IJ|F5c$ftFgKW`Pg z;v~$4w~FGCf3^BrY~#{Y0>L(tnAErqPVNdgzP}3!T;#_rPp&jmy z{F3l(ppWSZU%p(aCoD%6!ow{W8l`jF-`f7*>?x&tcbpAYS2Z$h^U&n4q4C!8|bM|ifQc6TaDQE@DqGsL}Uz`!1pyHiu=lw zXrBIhsgg0-r?bcqv`%fuqqTuHBT~qp-6lc~0xclW>)Np7gOs!emR|%)4%;dPg!uL! zcBig+Gbo_Uo$g%w(lkVnI(-uoluKX`AHOycm)b-&x#wauciwO`_ix)l#j)T{4#RWt zAMsvos8rS+LOsIjQ12$ewZ{3+fHTAOj&7u0&VG$Tdc@|aCqX28By^?5?*xfNgaa}i zo#!JL=f2dU3Zu(x>nDiBaI`PNSoamNIO>8F5KHG+1X|W7$VLVx3aEF~Nr9d=;Zp)1 z$q{izv^exbluGq{5u$A%-J%zu{GWsAwm(@vqj?ZGm4r`Cug9sBoSml<@_OVO%Vsa; zsduQYFxk(7RoJC7#7vy-grWQXh>V@>=>O>qZC!5o2J6Z7bPk(U9XMXwW_=iu@f|bD z!e!KRKoTuI{bk-aTf`KK`5PSzWA~QQx>K(20xsa#+Jlq)$Jtm#GI-hC-C!wBvDx%! z2EWG{+fE%_>PeY)8RPiSHn8vD;Kc@bh1f{q8m$A%fXt(AFn_HR|-l?V<&=;Q%%cAP7dN(^7vdn^!`e{(cv^zVzNTjXF&c zuX?@pJaFUu*SxS6-47Dhq%g_4LUM@$N!h(8QJ+^#LW!6)8H01zazxVEdaTfqEKP-N!wBN#;WGE9YdP~xgy`?{@ zS5H@O`~}8GeG_v5BeGtw38=YuNO)E5xx=v{Y789T&J)0EfkT>YAaLORo9=9#1^Ca3 zvpI8!FtWeuuQG+}{+mKthxSo_XNDO+jI_Yis$9MzsErN7<<)X299dHxihOcoh=%Ii z9RvG8S^U2cEJ9bc-gKJ>O)TT;zUhq@^LUo0Wx0R7H5_dY_AEx*;QsSFLIqXlN;Dsn zIu}LIY*GnEPih85!2FXD+bi{#2R8#I;1~po0Y_i_$Mj63p$g(- z*w+ehD`7HH7_hc>xbgu<9lYYeQ6e;pC8#7kVds*tORe9a!^Vcjh0*-KZeX(G!srmM ziGT?>UhhW!XUhNQ@SoH#id#zCqRs~Mp&Kcc7!K{;H-oL;Sg>acM$ExezvKl#@CQ+h zB-F*6TE-MNkxKj5ld%hQQS}%~=-ZCm@kp9Dxeri=c=O`I=n794ydUJU!t><+oA}RC z4Wv}mx0resbGHlJkn0{DO!&oz6P+N=fRZ&zCy)~%z09JmdAeaZ$cme5i6I7S&C`gM_}^0dj}np+Omc#^b}%vjCa zvURg1g&6i+iyOk=R?F&hg}OM<2uZdvzTjd@&H_5NV7P?)0_^{gKHKqHu;Wq=0?De9 zr#rrt0aZ@~(eJY~iii()`(YkLUmnenrV7q+-q$H5;=SH1`hG2y1S>%jOt43dJ}=CH zFpfqcbAisWnBALj{&|iVLOo8uzXi*ozVf{o_SC|3=tgPzBBrZdC~0wVHn);=1_n#S zkKiantVICErSe7*e|bI&?7!5Gbfo2KFA_)tlT?T8*HKN%NmX@umNhCbVEV@w8`(^d zWI!zdkbjr@`wGMy%nUfL(45*xxO7P(fAoCDJ60wy8beYS9&1}ZMp$ZDAJN7wY52AC z9#zi$C8GP@LT;WuEs~pfH1V-h{j$`BxG^zO5AhokpXmJP#mPJCf+3 z(j}>Q0xUDJSbHl49wB7A_ZjqSY^lKu#xgz-g?he$}5)nfLemeUzfE(vUccr0gL8u}vtU^}hY-gMb{1`IWrN63)x(q;F0*=7| z2iujKuDG+1+j|QT-9$uW?k>$-88MbrM+**zyKN-IP3xg(BeOa(5?yo!&&Gfin}f}0 zaaXW9GB!VEb%Ze<=_%D@WyDi)30o20fnp&f;MmlK-BZsd?gKvS7|Fp?_(LW;<%_h! ztqm(&F;DEm>K?h?YJkY>ImG8%Uu$M7m6H2^+sO=s{mZnX+F63BrB6>S2aK1+p>2Ha zE8j_!%kkQp0vecQjuA$5k9>p0jLv#N7HR1~eXHS|tXVO@?g-#M0G97fo#a!ZBwuKV zhyaG?pHEW&9UTudZ%9fRqtL^!cEB~T}+hB;J zrjmB=3e+w`?ibw%P)Ct9#}A zMG<(OZ-bU${AEnpCjsOCF(&gNBIMoISfsrPwY(WkVVW2p<>!>_#OdoiCH{8|Uo#oV zAq`@kc@w>N#;4bCjV7{+sYI6q9C~>b3sv|UVx4ouIvj0(3MS5KEK-bEDyjymF-U%W z97uSgAsAsdH!B&tG6^Xw8S4N6yX8`c#K?KaP$RICfknH}O2%w~iMx^kLv;u~HunEI z9F8SEjKq1tH$J;+hl<4T)|IapRPJ%tF#mJs)F3=&QZ8(g%(8sb~6s^NPVj2X))Z@BJn1yQ7jL>SAv0UUO z4W8!Vz7n{xk`C-CM(M9ndzadWw|=bnN9w{>DIvPB1dyT&KZ(@|=|ZOnU3f-dLb{-h zaEef zpV0T>^3dt42_FlkH@B_>RKRgJFk}5gNVCM|xFs~+Dj~dY#D#+Qya%z(ONlqoo|Pd+ zPh41yaUozeRQI@0dID4;7#9E|*8k-H8zJoV;+Esl9xl}*g#CDnIC8LiYtd69)9nio?ndzW%sV35hB)a0EzggTZ57q5f9#@CGk!OYv;I7F2OiiFuoPS zSU)bT#&~B49q8FZthD{Jk`xhF1-d$rOUCe!5O0nHEL2JvL5+ z3{1=|C_?2;QqNvR0aMveUB@-JVmWOxNDAy1mK`6s3M=JlD52J{G(tl@!EbKKA2Mp_&zyH4GF5-+ z{4))+_ggW{f9Cq$*^7df2V|tm_yOKXY`?P?@|RY^m3eV@C)Uo&0w`JL#4Xm znx>-PGjLZ(vdSJoshm_Z@e;R)c<^c9Vevrp{Uf*)EZn$74r^@b}%-x=w(R$)W zvE_%L5aH3vHvd;(Z!eqc4$1ukC!MSmOok)>0vxgeqLY%)h!f|}m*DUi_P7a+?q9@In=y5`S_Km>=?eF1_Sw0@tOc);H%-GH`7+R21tf{*HkhAm!Evpwo#Y&}Exn!5>?93nBw^M3u!$m2zy-;!%(@MTYD>*ap z@zFJ;9o6Q)iB~gBR&t_Vd?7Eoac05pKT6Y&ZR+i9q$NfA@BT+i{|)^9cdNh^0xJCj zbONzNfOf=s^6RjU*lnlDx+5SfBS~!F4oqW)!u%T~-@6m}sH)Su*5gKb=^|gaL=O^K zWDU~Vxrd*2lr+5LkT+>zjjL@22aWP9o z-i+QZ68UjXGn@I473Sx&)-VmSgYR=(Jp)4VrI#G#ybkY{Xcsq{6ML}-eH`Z%+=xBP z`;mm@r0xy|5uBkS1t+%!Qn_7G2ESbNawcc}f`3NZs$K+N?{5cRdgQVp+qzG$7{o#IKB>^*$0X=QK;h>dB1K_s znAz7_fotnR;)u$2F#AtV$jrc3JzP_qy!jM&VSo!8q<+*Ft!*!Un#x)FjzR_|ID7Ha zE%PsX@pRGIV>zX_$`O}PB}Yza7p!1Fd~sy>tt}`)Ken1YbtCsQ~A`gx#iG*=4AY-bm*vQ8i4W zb47>inX8$?7Q{43;PzJa%VI|F#G-&_3Z|W)jrh;&=$WqfZ_?Pd*3;OW_(UDcR8HTj zlr5;G>fL}Hg&&{Zj+Fn6h+_!3rWfxR!cVtDqCAFhITHP?kMVzo)3$&PfZ8N=7+E5aWT0Hl9SF3%OC3vPn|8*+h)cD{ ze`&ZuqD&CpMeovAM^kn~4t5Aqu|tqy2I_e(^)U*=y0V-9s!D`Fg_?q2x;lMHnk}#p zwlg;qO{Y5_cLtJJ$$6;+;PHZT&rlLhjxut)m8g6x%P6gDb-Ix0=8}fw8tX)ss;d*p z>LMn9w3_^)Vws>y$tAjnGeR6a8Taq*xdcEeEV)Gg&(IM1@~mV5isY0&hDwePRuU8R zudc#t=tQcT+t7$`cPog{368{!F93i`RRtE8I!8B}xEcwr$crM2E5y5+3~1{&8fSu? z045RA8*?e?3N>z9ZJ6h6!Jw@Q^{}>u;EtfOtHMhOcqmF#yjBmwnRu<;_&DVh00M3~ z;51p?a#@VV_OHPZj^{hQQ6pT!UgU^n&+12RN7usH-Y8A%b) zcd&NkAyL(@O9#T7;&AQ^PQu9+>o9;{EZ|C325NHC3?TYDO?(vyF;)(~?1Nx9gni?k z>ir2iuWyzGSzcqx$dezeWWQZ zGSAyko~-R8o(D|k7VtoPS!#gES!8kwa2_&OYiZrK;mi}Fq`ylPYw@dEngV6a1X!M` zj~ufXu^0U^7^(RRg65p&^Us4_x#}C!j5x1KcuSvgn1h95-wt*X^q}4JgR%K=S~L*K z>V&OkokssmZQztQ9jsZq_&ydZzAnPa1Wy2iFMZH&X0R(2j5r7G?G(yGjVCE0!qINP z&19mw7nxRf;0!pHwx-s=g!D#*Dw9CSJlJ%-&A~;GjGwQU9io+xUXDv;0W9Dc0*nF2 zK#f<{dSSYtq)oXfh!0CC3G~9PV`}oiBYCa#qisVh5y-atjBz4@R64fE3tQkbEznYKDF~2u+BnhQs}Q#U5HOCly3@UL>D^qA6@7e#$22>A>phzI7&`l zD^5u^g--D~bF;m;Ce4+c5xNcq<$xhwtIpuj1etMYe*AJ&h?f``&!`W&G6AvYD4$v^ zqDve;G^%t{yjJDb$>w76m)?9o6$fqH6ALm}|7$5mL~=UPN(_S6h-o|xbVzLb#Vh!q zVms9&9&PvoH-VH(oh9eEL-{J&lmbzpO>s*kx>N#6aKKvz--_Qu*}!JNapr1c@h+vo z?abLuUzhrzq_`&-aq3|)Z z4+f7)V3*6AwJX&O3}SaR)xSRAc$CzM935Bg*MzT_0iS@I|*e2ExZ19i}LLQ7C=M$sHFB0#` zZ!Rv~h#1R$s4J018g$BU%}DGIB;utt=B6}#a-V3#;w!b)2yw4#)%GXlV~O@CAu_Rg zA{7(7If4+akS|6^&<^zJ6 zQgviCEzH&ioFisr)@@@>SFdQJY30#xzfp~*lE?!9j=gR zE%oepha_x&;t|~ol)Ggfad!nD!E@mDZ$hu;IdB;NgfV>^GkeaqsGc*YvF(2m9GR&JNLZ3F_IZ;I6Lh6!8Y6%Ye&2k8Xbpz=@s= zqVbN0d9q-h0?aUD#=`4iVfTOjPs75mjpAWO&`&d9#uq}gfM{$zYFJlloc}Uv1jnbR zp#6?SI`|bf!xK%ZOH=S4wY6-hC(eJD zyixdXJ(QIur1#d2UhUY1XAp3y3t%t9s~tP)Q|DWP`hQvN_z*y5wc~1(7G`$Lq(d5a zwPPy5ml^NqCK?Q1`Wq5^qL*#1EJqdu%OgHQ>S*^Si0AJZK2Zv6*=gMAcJP@EuHV_q zeCe$taDBTf4ld7NNVti{Qy9eaSBnsyIU3J{5qL`D;jw$E5y18RBSQH4Ag+196>$91 zL~nF?E~cnMN7jxXioqR|^wqhXj>c5eGgKPk<(Jt*Hi{$%lAB6!=y_q_|M)gAHpYxUBa!GYjZO1LKx${(yJN+6n6HIdLDTqlg~j!R+WU&S5U;I zD(IMV;Iy+%Sb>!<&JVAS^ zmv=J#Zp6xmZfX2tt2w?E2Zaa>&&Qe22QCx5d2yL=pT~Jdonc7!JE&(|=-5z;Ex_p3 z6_Ua-<1=;5I@X~EgyP~Au62V8>COTx*r=VLesLhw&2p*pbgfz8D?%hAowOju4-x+a zivD+TabB>*5g@O;_SDlbma1Bk9g6T?0Im+LJ+%b@KAZs-9Ajqq6Imo{-jdZH3D7Pb zM>dTB#Bu!#+yKlI+NW`tr$k?^-zg?Ghu}%KE$at1+(rbHf)f> z2z0@idSX$(+RgD4=PV>dnJs=kEc7-!uSdOM;01b)CYk!4;K)#SN)`w-Fa%On+95RE zbKphYK;FNyzW1bo-2&K7rw{WM4AoefY@odlK5?R($hj(fGFkcw&yMTVWt+uvR(ui$ z74a{PIlLwsyu79kpRhI3c=<%Uc?NGPf5Ui}Uls?iXFev5;-7P^;NhB7?^P)p@9qCL zy!R4ssFKa=|DP(^$`mRS$@I8f7vd9ZAl^45Gmp=h&3o%&aG^sv?{lmpW4bv_)|@cL z}W(0PF^qlogIWe3Xln;X{l- zwyE2b!LDd(mk23F)Naj^xV3wUczuqO+6P` z#syEX{x1y~bg4ydgQAPj8lBabfF&g!&vmHNhQaHp+3dV-G= zR#@(J!F5O^FaRqOu(hF6^`8Y6*q2Ipy!wO|Kj#cIi5IbM|I!f4WF2Q{B1kPjQcT2q zBVx~P(EB+PW^nMGg)^IaOUv*AyJOM;(l@)p-F-7UPVJv^rU-;U%Otpj^?OoVt z8`uKPfRZ|Nw@vR{beG?g>XT}EFi21fBn5{BIv4WGg6rQ;lc9^iP;TiEtSyu_!kwQ# zyOW)z@Mb#(I}j4SF;!fc9DpcpBfwosSjXY&Qet=<8qAkm@`$1V4vU1w+$1^-ht$GA zRs06jT=_kz>>)mD{1-mk3y#6SVKjOL`*I1#PJEn`EZ+u8~Xt_B%pxyxeRi%tNr8q+@+r3w!Ye4oZ#sI%Bz5Z9mk@wNP&ZBMN z81&F3x~}1Z9=)FX(o?t{Aq19J8EmjL+8BVr|EqU60=Q{qV2!h56wY(CE0NX#VvPi6 z7H|d}H_89&qwuevM3f!l6ek> zO#G@Gyl18j5774PHP9-Xy3XXv+6$=cWabqnQ;FhXfdap(RuGbzrSx1*_Nl4w?2wf! zvWaZ0O=I~?{e*XvN1BDH{a+$ z#9ODlh1{e93_7Af*5Wt29RF}$IjE6|GQ(S{>ThypkSzeqW>OcE$v8yvnyjujIdU!$ za+1`4kfS5?N^`MN9oSN|H!0WGt_t}o$l-W@z2P!1C$Kew$vk1Q6JC;&K88f1`sQxU zTb9X5P-~DQByx=;@*~jX3mU`(37S+E0{dzBRZkd994hH>JVZPIwB@$gn>%7JWU22< z(GCc=)##2pOKXa5;tM_S7GYNMICQf&c&5+CEq=dQpUL2=&pzzjmTQ@8OQ(D*@AN$> zCp)!W8+?{c6+*?GC6x-whL^80U2fI3T^lU=uXN`F z%xpcu?;TxBFCrT?M1-F@YN>lYvLfT26KMoG5U!B_?)$7tk zl4og;BBy)-?c-+8^IljpWYGrS%-SVUvgd{T8!}x*(^)|Epa2BGoeat4M z=W@j#)eoz94*+Q?QpqE@r+TAVq|cyH@~jI%AdIQG!4%TuYzYliRe+D> zsL#NQlKIwkQ6-Pm)I11gbgER{OU)j|x?U`=1*S!~5H--qt#whAdCMRo$~fR!V~1XJmYg=t-B zp|9(GGYW%{+I^;~%__k|z|k`iAMmbXueA6!#z=3t6D03OJ{`unD3yb`FDRSQBYgRx z-8+?*i`{d_<7VN|rS85EBQ1Q2WE5X+7l4iQEs|HX=7X=s;u3gya5Au1NZF<$fU>nL z7K%6;Wa`DjHXtlkH)IVwBj8}XQ6LV6ZIkh;TfyPS`k$R%QiUq{yHX+C;0XS10DlD7 zdAbq+#!dh@0D$xFfCuN_%Kz3HCkN{{S&ClcI`qF7ru1OF!y6i`Ph5g^lBhVX&MXy< z)9N*kRjASJbBWV>{Ig)3*15M4X%LLlq7TkhZPgsKBjdC#5Au==Ifqv+8`a=)z=J6D zUsf(pRu8RQj*(VE#T$|H_$!xN2)=9vbu3Li^@m;pNL1BpaE^jilC#y3@G*d-$*tOJ z&WEIb(d+Das@>C!G$~G*7P&%zGtepSImM*Pt^cJ&A{6E$2f$+4G5LkY(n3;WF(+BC`Ces{T4ffFhR2N2MxuyS5>KdComzy`X8rd zvi?WHZ0eO$rE9pTXTLrF_60Uemh=050_ru;Bi$Ag(U~67@}+-Nch!1L$x(**oNtLh3PP zzAYkpaT<<)i5IGgc5O5FLBK`;qNl z3+U)KzD(Latk1C8uc9n73z-QVeN>5HV{eO$Jz(F%-Jw8%^HQxp(T6db=9cKyH#PJ_ zZ7{Kk*9-MvFYIZu5|r~T*ig{h!jA7R!2{@Is4Ckp;d&U4I>*=fze!M%#DneBwhd$I zzju=AI^VII2?+4a+k7zQ~eHOWIc+vtXvsLkb`;T;N9DGjX7Ck zW-5Nw^Ct6U`szTCrly*lLh;@!)L1-+L#r48p$U0Os4*XDF?Lteog__FU1~(4w&27Q zD;*HH#*eQ-Q!kVaszTHIKNresoSz$~?%wKxu_KpWNO^D+>gxX;wY~Umax4pnbN-oY zZURY!cwDN#>F3yQh!_0y)Elf}Y)g2@i_v#%8pH~wH!C-;O%JYYKpT3^i3tzYhd7}F z$;TIr?c|&$tOD9 z|FW!q)48}!nJE}2$eKtcz8lHQ{+dl}Xh`!wX&@;s3@_rZSr{H5px>WGgGyK}WaLyi zINe2jX05L`$zd-XXJ=Hh7HLqD-@|FX3(5M;vw zX=)VkUj$+=mOEfX0h^Iuu4!v#-lZdR0hnQ|BGxkEwPgZ%_1$v$y@;cDq=>PK@YRE- zo+2Dp&Np@9&IzaKZU|D|bj1q&%zUjpK(8sKv$gmPNaQ{s~668cH^F*VU|6`x*8BmF1LuP<8 zm--rWAz?^lG)x$h&+P92FEYJ#lO{sJu;sAZ9UTgGHDGioSWFXKFSy6aJPY>xIhtjJ z;X=17nS@#tknffo2SWWP7M=r3`%VISn~A(vRZu*uMe z7x4CE?0#*PFM?oTTK zB;jK|_eGo_;S;ng&drr9|x&&`ziV_8ck#lP_%6QA2!3nDIR;9)&?~d3iw=_aC6x3@H~Xgyq9(|5hkZ z3b|28pYY|bB|Yjo%;TL!?NWlGZ_4L*L=&a_^(_`Y83O{6x~Q|{kTkN~Daz~&Vzg6L zQlno)>=bFai~^>B(MRvb*Zn$ORxO5mYXs| zTJG5lw%j)mQ^3(2q$;%>^x}&B|I%__j3b8?=Ykv#4f695UI>SG26M`SIoZLSr;r0` zxGS1Uf|ISOf*~XmhZq#~zru*;8L+8iv+~SYAtN5O%WOl|1~% zT$_JKz|mPaExnF0=Nj#0e;Wzfq=XKB1M46N%;I0<1wy8LHjDS@MM z1`^osH5mRKisY_OgeJiOP#u^e$I5w7`%ww%R!IRNdVd&Ust9ZX5IW%@uEj(PUzd}Q z_NLNFss;-BKMXh~f!+*6NTR86{9*hv_5MCi<&5fXI8*OtsI|5?YG(rnk`ZU>IjK2O zB}lyvXbES(Jjke;bu>+(GxZ+BnR??R&eWS?h(_o#!fQ9tcZW36)Z!`*ArbLjpQ$$m z$myLW!eat8b+cv;Npq&&`=krz!X8$^pKi=%=H{STf>c zRZl0W&m}_gmWLbq%^jvEzpkNcdaN;p1e5iYHUeA1jbxY2`)|jv)lZ(r*UgUO_)m_9 z3k#8CFxF!nnG0C3wg`@^Y0uGMI+`J;Ff>W9zCnCZ@a?6NVCR8R3@eHxp_Q^BwgI}znsM@!(TokFsTp%=4MUm{d)lJQ z5G@GH5}GleQ#8?xGw;#i#GYRGz19pDkVk4p70p~&GoBGE99fc@(NCCT=of^!7PG@{ zB??Cf+#z~mLt|L2oo#{mmgV)O{q5c&&X9b!%)~P;3q^#REclhmJgS@_9P07?QkvM* zBTW0s6EUAv59l;cvs3pWjd4Kr!!PcgMi#z9hazn9N?LUwY*ZUwBaL79zQtiftonXcZH&A_pv~?SbbM5oHeVDP` zA`ei-MxX&{Y$5;sc-{)G4fEwrAKc0m91wxgt~jM2$~e04KJ_=eua5VXUw6RkT6}%Lb1s0@l0GO*$J||;9ym99 z)FW=TPAp!;F~E80HHzJD!r$zEW99!y{3mnx zrAOTcjohZFt#V(P#H|A)oYq0eKU z4BL4Re)VGNoy_@Pi>VF$bHj_N*>^`Srpm&EOSM9#v$RzT{W=_9i;?-ghi|p5rtZqq zuETH$E|F*HNT1;ul30fK_zFS%xS5;vsJ{hSAyL@9a?2u>@ay{_4quJW7X_WHVm-l! z7S1CXczJc4KEx&Bq3!lHbo2w;1oLOP*amnT_2wE%PkR#kvbqhw43|oPZ(=Xz-SaYf zJg3rZ*0p#loRlPORH3G!BZAjyMCiFB7t5?2@IK%u_*2ZWD!_Zr9HY)o=I10Y?S*EI z8%xby@<7ls;C||YlBgorvj>llbu*w=I5~4 zcj@z;u`7_w4TaLt_#lr5OqY5Ra!ZGT`-ASL|N4pIRM3`B6n6&q0dPO6hY@f%dm8T~ zv=rNZ6EWckz}S_JOOQI>4%_N3v9LFVc%K-$l@5vr$ z?Eu6Ky$6v!aQ%85YVE|PNN%Z%yCHHc;Y`TGS-LVw)+&oqb$l(9x>Tnn-fsr56s}Cf zI~$M7Sa}S3?(`Mekg=&qvKdv(GR)H6_y8%R*dyGRZ7PxkwpVPhIqJ?CSO8vm3ay13 z>oyfh!s)|Z*Eit>3_}L6SOeL*4B$n*fZ7wRmsK@bC!8^h;c*xykC8JB>3uz9Mu|_N zqD-9Dh_l0}5mz3Axkv9)FaC&e5$35)GM7N2jtd7ojLH`RUr{0Rp9*!?cl7mRM6 zOh7kiP^d7jR}q|f(H}aR7P++c*HDTf3B0m4l|Ds030FMe zIQ=`#4Qf)y)FcZZAmH19EX-pfKS01GYG_TelBI@%TM9C(aDvlZ+QSofrJe;a!OqfI zDP`c58;0s#uFmM zX_@eSHmkCGTlQbp!E*8>7Aq}thd_XH{f^$2*~55ctw@C5Wi>_b**eI zU_e8kj&Mu9L)(eNkO2-wPhs9T-~0mSl>(K2J$#i-a}LU_1NoR_=la~Zv7n?P*O!mQ zlp5dS)vz4<1K;KPCL}WP^)UO$&+RQP#)XMb7ju#(pFd)u@+22O5m+%KB|o? z!}~Q!!oZM}2{>u)07kQtzjI4)WYPo(8fy?Z;-5&+^^SF|zYcS&F+$2gx6f=|B~5Kzj$gfS~K|!@TJg(S*pbVLM`m(94dEIUD>^ z;y&rI>Br$H!Nv%uZ4WS9P_OvSqPKS`RB5Wj)``DtRbM+DcX>l-+W$*cXJ~L;|IdkUBnVB;)TgaKSRKwPz z*<@>ipKST?mrJc=w?0XzkO3XiO_HI%WaxvxW|il2v=o`XT#u4fa2}VO;ig8a&>8Rv zY9H3Mu=EwK?VzlMoIXMAqsk9aX}p|V^i@e#i;2c^{lWA_Ef6(#lDb2k$WW8<4b6xv zOwCbtq&CE>u|ff6hN%dK45J9SAmQwXq6iHqkpw70bC6&Z4|A-+)F{p|w z4OjTw=crB5Q-b%Q&JSluop*iD)cFXR_pmqBSHZ?QImPc$lmS&p(UYGkUm% zM6H4Fn9-kb!B`zKd(%^JFJf4?lSDDL+^exid~+9E3&#ca?~p;a>LK1Tb5h;32|efx zQWKFr?|b$7W2Z6J0sC(vDVu1#s)5A*5io;hU88(Im<&f1o_&TgcAb15YNCS(E8uk( zj%3GRrCr2CmZxY8jSAGc409RaJ@5t)EX21gu$n?!={w}L`p^_^z}tad_mBX>!SRem zdi2#ai3dgVF{upGc+lNw0By-|l7c}<5=@lWH(Am$+=Yu$IM)YX#WL~}OTi2jSdm%R z6x%o=faQ9@f&rV?htr$TCX9@9n}7+TS+dglPZM6;!VB8P72pMy4ts1eeQQC0yU_wu z@Kv{vEQk`DKYu2}nliourEojW+|2o{GF1=jeypA=@XMBRG1mdlu#1L_j-adCeLlN7 zE^Vl4z36_NRG$}6mWY?-uhiERK>qd8u`f@KhpuF$m*v9!EJ#kogZBCd?vo*Q9A3mR zJPPj3h5S~Kd=$U%Qss;6;CmlQSTx?p5e?c2zRRO(hN9R5gDmqc0pH|N)iwmd-0xe2 zadR3dLRwv<$7|^#?V{A8E#hUxEN0URYawrdP=5H5rdSr)g3AMTU6w6CgMU(QOl3D@fsA4@XmYR{YSD>SKfHJ^cEQLDKB<}^ohTz3x$<wfY36!LSvLNv|sRP8nek+N8 z5LR+@s_L_rk}h45fg}(fG9b(~u$^0QyE7MZd-2Q7gI93#Am$^&Ja~zp`v&)uET$t- zefS$%$VnzjU~6TtrJfVpKRcb+n2y*;SQy(*rU6X}d<%}62}eC|Z^roym}(rKDvYJ% zx>OCkC2NV&Q5?uoZzFR6D+|$jpr-{E+{wOrcd0n8YHJoXr(C6={;6?Eq6Y=Z8-Og_ zr`8Al95o6#iFGc%aRza{2@t>`+-1S7fyF`3bHFMQ_jH@NIrO_d;KRZs6Pmx}+BAk; z1eT?{%WjqloBQ-Zv27=6qTUecS?>T1X;J6I$O##07lgz`EsrT0EvCkX$ zr6gbGGo-{#WMK*F+3r@zO3Shr;thO_{U1O=%lkE-SMf^;AinV$Y#IN7?YCYUTOFh! zwkihO7{LZU)DwV+^`AAGCQcwk&;)iq1Qw>#k47~SChLrG79+(JCKc`{~EO@y)xp$sIRFP1909MA&Bt zvPg(|cMThB5aApcz63qhkysu-6{}IaAHhzl+GxQu27y;nJ8{?cf3;j0Q+|tN9!zqby~GbC!+9i{&qI#Q*Bp5C|2QI zvg0eHSp8ERIg;?%{}x|ZuER^ka-w6?Rsn^H(LVn&IJ@$Dhaj6~i%J-_3CBnm@gH20XCtrC^4W;KT4y3aoS^YWcbkpq%@)aCi-k|Uyc(p!U@ALuEx{SN1|KLA%}ZcI%9{-|bmv$nUd%)vn z+G_Lynfy!*rGK`czaek0rF<<&&gWYcY{yD!SBUgvgy=~P(-S6JT~-Z~sxUA?Pq&@a zrYJ)5o<^IjH&5DR;}xdw+3+K1LRFAt^#&CE+yuV<4axs60<*>P-9Z}%L3R%lDIpm+ z|91yF#<}<_;<_D8y&{$9-k=KiXPK&dqF0QnGjl|3kO=dig7t2}ik$6ChPNU#u&!8#6q8K`i21*;}viCaf?wIa2gkTEfSkih46eTr9Yk~tmjW+seRcvi{x zUsK*r9J$Mbo~u#9K>QoPU>9fs)M4>lRH0IDH`l^i)^`^2MQhXLBGDzcG8AeuOl^)j zMT2|f-R2OWe2hst8w2L5}PNcJ@66pvaC9A1nRy%IS3?g(bL)}Eg zAhM6MZJHgamjIvG!yHwwhpyAeWZDw{p2<3`g;H?-3;x2+tlsh;ab*P?wLBdm!iX!| zh?x-+SJruaR9x9kpkob$bI`_4l|zzXh6aNGtdKF$IpfOQrqFO)U>@SidZAFtAys`) zpySFM-*sG>8_1n;WwV}-y2Ur$5EstLxU%y|oYQM`7k(j1l6G{Cn3AhsX|7dzEC)?R z5+MfPCW-xJ$hh_0K%wBiqy$pubQD^P8a?49amDe>1i#Xy?wR321}Jkc3B{|JBLg{b z{x?owlJ-;?#wjob!%dtF-W^`$s{+1$Ex_?~L!S$JS|Kg`f1cR{e{NluByDC9sN6#= zRd8$H)vK`%UZ!4NR-v(J4cXL%g_duXaavFfL^_Ue`^zuZ!X)swLtf@jdql&1_EGlpD$^S@2~K?h0Uu8#A1!kP}MY$q|Rbg(Ahgl~K){NrjmM~4%MT=6fLCy> zxGw^0eZNUl531H3@g8F7xkL|-yW*E(>cC&il|1-RTu0yVc$unz2}{)eY~>tv=fen& zO0`!vnm`zAC)Jq^)+}vDg&9NUy;O%?xGDbI%KE1GMe;sMsE(WB-^BNjh+ecGh;eMpoZ*||?~V1{ zi15u06ydUO{!bBhI{CkfaL8N{VGAIj2=BvVln5Kwj4i@~c3Om!(VCfc5?eJGeu=-9 z>#>K=rM6QCI@NZBG4H*_C`3s(Ppj>O_M)~IjE_*;da-gi)V4Cdhw9|Sm*#;a`teY) ziD+X{4`pHh^QQk(=G*@IUuFK!XCm{zfPjJv;4w<(i>t?$c|wMkc?uX{)+90lh7yIL zUdf3aeJ5bmRAMEqwsEl-b*SzB1r8B?XFm|*Iy_VmQEpgmKaTa?sO|m?im=Mp=MdpU z4jP8R#a3S|gCNkgv{H8oIyxa!%)3O@!Mq!$HqK+Z=y@tV4Llr$5n*>COpq*kC$D-9 zxdJPKp3xgYND+pNO!2g!Vv)bj2k+ZMwA2sLvI!qZRjR5LqGgeJH%#69AGDOUb z@|!-_;>uPf3$&u7L=52cXy;ZUhNA3;bD_$k!WXH{H3V-8RBH4vuIp!Mb*9pkC`9Ul*&=?uC`Hu(JNV`ElL{t32nev^BUtdTX@mc78P!uw3tS zh;z4)2QEaZ<273VL)0mQS4ncjTB#~RnW^vfNJ}gtmZf~aSuMd0m@JMruI42xq2mnx zq(5b0sscqY9xF-0YVJDD!3UZ%3aydGo6v3V9 zvl+T(k?}x>EA1eoTfd^VuNBmX|7{D@sBtW`hCF6>;q@fUO1n+RWHkH)XxPYlieHXU zsvQ_%LYx%9G?EIv;5wwbGM@6{Vn^lASv5v->uP8L><7_m2_F2vbHD^wXBe{zw^ohi zS3#Z4UEr4U|4fmH*6 zIRBbM0?W1Y2pO^Ynk0u1a;Y{|M3puxXfs%|l&2Vx!6t^m%gq++B(-Xe*6oOz$?Fcb zrrjxQ)jC^6lT^&<5zBQaN)5aKpfdnH&1|`v5!UyGy6O@hAd_6K8ie}*4tn;i(oHi7 zhA-3iy~R1H3*J*?RsRqv7O1ozp74C)LvoZ4WjIn{qMn!G>Zp|96#YC`|de*unR4q!8K? zk6-lzKKj>Cx*}*nuzwl5yq>5EHYl+^c^9A!jm4Ts&~y39aP>*U{JSmY4!+g-1MBrK z?%+Z83e9y0NkJn@P`y5NG}lXsy(_R+An#^m2<$Sg=RDBGDr8rf-|GAkI2Qqoc-Edl zRoE64O0psFpRHjoKTC7Z9%eUpX(;HYs!LQGoBM)KVzmaut){)&Q^Z~Qv}p~WV{!qd zdVyA2Ln{DwYe*){))4d@_$n+mYJE=@zy|&V8aQ;*e~+v(u|*iI-|F}BmF3#?8?amp zB!JfvzDmRM2%FaMK4G>7BdeWg0qqb+s7+06m>N!tTb`1CBZc)Y^8JVD7;@Sc)^~fO zO-Jc*P=v}%QsZaE5SF{}J>699?;!%){)0pPK$sk}mY~>#KlFAlsQ!!unA?HLa z*Y|#~*Gsd#iZCOp@O~5FHX2?;m~*Fizm#yYhKmR%YIqc328m<+e+6f`vhY`!fZnoN zmd})G5o5G_F^tgy8psG^gaZpedj;KowNf>WNsR;!e@mT@#_Mr{ma75E{tr=C{>K*e zpgqx|&OHVUBI-J%uY#cZK`;O5E@FNZdpbm~E?p~S1@{3Wlv)0PTD}Mur_*1yTs2wb z9+;`1XTfq78LbdciOC((oSxRJOhKcCHq-<}f9P8YYTES zR0kStD*vmGDO~pPUjp*qtIZ-?q(A7n2IU1kb@10TvEd7C|87J_?c2SE>9CL`!tw&B z3aj`t^E7o{=e)bSa4;UR4odgma@8o)DniH2YfVbd=Px>P`L)vaW8Q+MD^YdYrBx&` z)REtDRI5m$`uiiTA`u1Q_F5}#*KMLA-V;Ve{#+%cT8E5^j0Iq;$isxS=}K9qt3T0* z=`9vuI_w3~aL3@Vp&nJ}XO?<{o85&|(6xYn2*}yF3RevJmk#Zi(w!@|0CXvf>|E~> zKGC_(jAvho!6}Guf_X1>W;;DY>>?^v&Ei>JsDJGsdC*bjN}j0x)iXu+uiS5t9_nAe zSOETwV|x4#-M>ai&0Q#aeXO)8Z;0f!V0lFpZs|(qq{Z+bqi3nTl;Z>~S3~S@^`~jL zCgDaJK7)m0RX~?_17P75;I4-G{7Hy%R;1ornJNd1U4l zLm6HRo4M4*NE-N&7^p0CdMcUB*=iBeLuCGJWC?ouF40XZg3~;N3)W6$ySfRVooJT+ z3&e|{|IB9E+X^&f#9$bs9f{~FX>VWOpxYaD1}Ed*7;10u{}$`=F`Mcf0&w>Rs(JM! zk@TMBke;nL8TaWOB)`I{WX!(MJy7eR= zuIX3PdM?vkPQy%_V0gq8!_&O*9o-U;F{5RuCCvCD+!8kau3JJB|M!tFtc$-U{P%j& zw1m0vpiTWe03>?AJ0eBb&)p;?<+| zljjvdA1XJ3eib7a6qk(R0uumSOyGF{mjUpP(o#N&{WJa{8a_z)0S#{_e5Zz266WC* z$h)4fhn=(K`jRkX$^m~s_zZ>-fL|f}hlW2Ud{D!$65gZX*9mVXtj#w=>H%i@b{Hvs z{vM<_5O0H?WXcl7ckp-%YVeR~rCInzA`yVK4*|^t2%9k}|NzXyrJ3ZI}N{h8W-^Xk{0a z<5^F|32CKB&p(1vuv~3_qLK-3#uQs=H_2p+(X(f^IQ5KHxTwtLusP|gz*!nNtJPUJ zd+v8*ftzrK1X}SURBIxDMFhqJ_!xlM%Jk5@$eP$O+0?|JXfyuxp^C?arrMDZ4Yw!Eqb;xEE<=BTNu=r%t=34xDr{|P>c^0Xt#>yTVzi+hxi%D)iLrk?hF3Nwt^Rk0pz zIPe2Co-Rin^%V5C z)KDU=P=}*Yhuu0#9j*u%JMjozcgs~7n57Pn?=yAy3jnD@eEpX2P%Zyr!VhZr6T)|C z_-(>n35)XjQ#8n7*z*{X1U>i64NI7eZ+a)pc-be^f5%vcjqVd95F3kzo}-q0%x2_V zc6VF=u%m(P^hjt%Nuet$t(hvGgLFa9w$JHkh3<&#aU8m0Eu;UZT9POzq6H~&KYs~2 zeJp|B&@e3;p$=c}CkXR_Fz6YDztjvlNy4RCS0tj7=wAwnHRY{)Tgp56Qe=5wHK8?mzKcF}DsIYa)d75JIx>S7K^KCi^}l?a^hT zA))3^C6nhmNH0k%U$5h%BFRJ3M$*klD&e(~^7)uZb;V5=r0eHUCM)r`V&OP2DXkn4 z3~0hzU!~6l*RTy8Zvri6#-djT26};767_8yi&8}AfzoYw2 z-Q!!J7PL>-9`mkFgO8>~bxT#xz6p`jh1~@emXbdMc}^FWez9y47Z>)Ay&)I2^B&{E z)?rrcLpe+f0|saOGYc~5ga^lDsdqxig&lic$c6otEz=C)5vaTcE~Pa&Y6Y?H`m^mFkKp!!+=3uA#0R{S*~y zxmxY2XhLgarXyrRN3M_V1`GDvCbThVp;APa(hXc(&{nt@eaLi!4ao(j$afkOI+p;u z!Knaj6Z#TiV?sv}rU?ytI+If49H2hc!q-jrxI-GV8Lec$rsdvBy@Bs?sh$=aE`<>B z1-?gdb%@ru_7@V4lOhg$5-uV~t!;!pYnC4V`rpT|QMtM(4+|u5(szyG0ttBXoa7{> z?PEwEcr()Zd2+hx1M`S456=eAF0?tpgTbPE{7h(-jq5unF6fJRd!;yQfo}7PN#LuM z_FkUdfkG`knB!cGp1lZ-B_GH92*}K4_|1`;%RLj{#wY>T?%bTFtxMELFJp zSJMj);3&*;Rr~LfPhv_zm)tg@Wcp8CLnR-7Ek?Gw+s@6+%s!rv0E&`9cm*SrGZ_2V2f zn{Xa@1?G2|t1fZ{JzZx?BYlo zD&H;elh=;86?#_{DqO$@Z>z9rYLL(Os_Uy198B>e$i)w>R4cp+OYK@b`QOni0l#kt z7HOxpLRx8xa~`%gcKvFnF0yk@|H0P9pI?ck5i})V-6T><{e>C;hNui^WXQO9O*+LGR^umddIY*&5AqKg_WBR3&SxM36I2|V>V{ilk9ekctntiI` zwiW2vpaH-buG2e-{|nS<(6i|y(Z8R!LNZrF=MYzWQnh4ihzMuCF~)bqXcrzMLQ(IP z?r}gJ=XQ()e^n$~){49=uxP87#StF)GGOEB{V+l3!1{b0?b|uRTeVA)0Gt^AfvljX z35f$8UeJL%>g{Q(hCoO%ClDDvXRE)YOOD=T1Fsc*_*YlV20f!N^MEj%A3hAZ7tN=} zaxa?Gu7(>o_u|#04||I59olrNgv>_MPmYkC>ODXn^xRjTUyGAiHcUJL_jR0zc3a4$NP zu`;7I4gDg>e(Z8q(JRwMbM|bhXcHF6OoAC@U-ng`O(;|Kck33Ipf+2q!RS(G6FzMz zT-5o=^b#8|HiClRkJ7{@+(CdgAp?MI6Rsz0Y{C_UwM{7dK&uh-ffJ?bE|e-w^T{bd zYHL*`G;K<#mAVn%&*feeZw6Z9Ud$$}-HQT85eOK+I<|XJg63P%y$}m(tDMmF6Vf@| zi`vBJa4$yMcq+ITYiNyaT+!~ueJ_D8<6c|?+P&5}-HVGjcOve^-Nnx8qkA!Mr(JzD zfv`gLh1?6(LaMOu55~Q4&5|md`OSZp{1#?5bjkadJ4=RO)Zra=$&-p=lq~MW6S~v` z`;6jDfZH7O{K}(_P`OkAc6+^;-?$gU-mTzXd>KwI?nQkhmzwFww6WcbN=Re77fUxm zGmLvNgRpTgUL$PWi+sYyy%<8+xEJ>jHtxl(gwO3>)RqnJo-X&9PH8$uOlT5qm`l4C zho?#-t;aXUy>JsY?!~`f12*o(Z-k9|v5T;AFV+!uxEH72uIOImVG=s5c3RHjUL5ur z_u|5L*wom%_TW{xE82u7+r4-hCFx2EyB9u8sD#{$ElVRc0=H*;XKTcBL>tqTWH%E% z+_uNK7h^e6vs|@NnwXNR1ZYbB+6YcZYggeH!p4+rC9F-!9Fem;a|We6iIo&>j^SQR;W*54Jud-fG29CmsS7#UC&vT7b}t4`5&e5^ z10+NDg5CCRd?qW#ySQw41@Gbx>1L{4;`A=&?XQYl%Oq;30lP@!QM^cO?3G@*}Ft)`D>ZxGv~cnk>D15ap=;zhye*!ScvyjG9% zy(6VgXB9Nl1iEA*CL-l?dLZ_(PI<#~?6SeiJ`7?QcvV!q8?%9`ns@}o3NSrQ@5Xu{ zt#~Fsy>!bp;tKV;6-SccJaN8@AICNq#0ItM_4#rGUmkysLU$CNcZn4B{zVkd+IOke zs{;#yo}~a`>bm(3s~es-qJ86XXdG8a=+R$hzfHP0EpCOLbp_3JUxgmZ3JE!=^9G*U zuqnu;g19GyFM2p$v?tDXc!BtT7HNHnlkAWj2LR5QiLo9vI=E9zgNitqR3(yxNjx0b^9?$WaevZGH=OIi3^*F2dK^eCd=i^D;yb(M0 zd+*|5* z&E+fo=dZQ$JJGM4jd8|PpfBPP5kfa#QrzzM5k3>Gz5v>(CKp2;@Xf959&hH zeAKx??m{_h*-C4XD|KCCtkH;c&_+9&sP^8(1(sryMR|QNayf-)T7BC>$rli0n`|GWU&s1X03k^k5Y6M|ZC>gVFlBt-rH}4i?mH z*#Q&<{=4(g%%$kUGzX8!O~bu#i95DKc9)c|Fl9LO$%|PPnNfcV??d-gbMVcW{L4pD zWy;kr&#=9KD#Tk|5*pTiHUaf)hPB-{>tStVmDPu(S@^hU--JOxz^dGL zj{VQCLk=9@LF(@TYv5;qc%dE+7rf|e3 zHbcM|q&g2p@Ul>13sgD?{N^Pi#e`|bsYoxl?Qk2_r{FTUN2mn6TC z5WrC)`*7nHk9b8(+;jXxN#VOUeqEKsI38=hS~g?$&h-7&u2-h-Ulal&#vM)!dfX@~ z6SuBmyXtE1Vwx=cd&b*HdDC|fCxLSjwSZL@5SUq1Jvr0rg~Ly+HrFkd8@|)G<+?JV zn+aLzr=Dm6D%=G>u@Wn77W%5Ud$D)i3%%p~8+Fa1hZOSsOYq#SF!&S;7$61gD1!tl zm2~r+%}ifEyjJE4O-b;Xtn{Nf=1d$!_ zxmS9}ovD;pt=)_9FT@A5KoB_Rhd>~pE4`Hv3lFiK-|oe3aW8a>V+w?hXqvLpSLN2> zN)2xDAw$0P`e1}*K9qqs@tM9V$=!TBUlPT2!#Fj*{kkfxP!Veq?UUo%-M1R{%-(@q zC09bwu)F(w>~F&DbT6ZG&k#vK^&d2N`-B(zu=9B)y=Q9K1|%jb>?Ov|=Ox2wETRux z2bMd58v61ocUCti{%L|A`dAtGaYLGoZzS=t$t5EP$F=Lmux&UGd@u_+l?7!(poG3+ z%lQ;0u$K}Wa4_wl?>T@y^FDi+D<`5~G)yD2tN~c5^_9yB7%B2eZ((g(qOPY)Xr!K#-+FIN3 zI_K8fI$)fA@;{@SyRe`imfq;Vh1dmZNzjs?~nSWVD%Y5@wTIOrF^(3#cf&7FuH9!uI>q6H)>GQ5dI*5bFhT0+?My3BCs2 z1szZYoM`$S8R!Or0+aAgDx+pKj9(DZusU#XBAuaOh}x|R*=6|{wHq%`nR=@gV{U&5 zZ6f}g{W986wv=A3`rx;sM&Aw$ceWQHUV0^GAV5P;jH8ef1*&=`^A*g@Z#RLyidZ;9 zlB2TDehOV3Hl8E&NAQ;vugZk)m7(@MeqJ1*B()8Yq^^caUXc%pqbM9wFg6z6TXNK8 zGE{_9nvV3DMPn1x&^%oYDJsoJDB9_2E9k#ELt+j*PckqrF+mb(&$1XiL%s5EI)6kb z_rmiKC-DtpK|%`|LC+lKLPe*j?_P*if*y$9D>uQaP@U(vy~TyqihfwSH%*E9kz1DHUd5a%Bh z`t3>^LaeMTD?gS{Yl?iP{tF#M1>qXn703!qw=kfj6$X8P8xwA<;mZgoX*iznC4l9q z1sSnoW%bNZvcWZmUeq&1>TBdEu;z(OkNWu`0 zk}V=X+QHb8td}S~3^F+<4=yhyBW7N^x!4qPTnTUWc39Jb6Y&XmPlJE1t7mf>Y>*19 zuL@jI8%2zX_nfAtrWdf;G{;{l6IFvbGq4eu&szG-=x^TB5@nq&(eT$Z${GcDBtR#$ z{0ovtB67lAa1ZK9ebE=!cj_}AU1PREe$akQ>Tn$fB8MqSdk1gT*UdTkg@$3~zROf6 zvnx4XT{VQ&8%4akuv~AxI8|H3_AqKxYTg6y92!&=SOc#DIDr_GfDgby0{JW?L6>jU z$RI&hS|?JlJ46AGL13%7-E8%ZSCb#J73@`kqG*w05GLxdwmTBp)L>C8c&u3edCmyb zRAP@1CNCjVodd-B7&AW=dZZ7gWsUBTYa3#)e;c*dO7ry>_LsIZOL(Wu>$Q0j`wR17NR|CJ;6&rDwxI7f(DJR$w-o zUmivC_zhF+&|(U+lpr>gGe=WxQx*_r)E44@BsUervUg_HsDpw)D4N}Nz?9Qk*iEU;?hfW?0NiX-W3@+1qz1RIuO4v)gEuqK zynbTQc8%fwqkOPvxo!a2(Uq~2@9fIBbh2KyjyJnyOV!>#Z8|qSLOP?V>j3IzT`bgH zzKn(+)UAUi1L&c;tPZy6CsvABliVS+T9>+e}g@+i8^&8Y1 zI1VD(!~cq|>+cB{I^d@WOLeVz*twJJyF94+$HXNzBCG%0s{q#7IIHu8jvnRu73$t03GH1&qK(h2wbQeLI~;;;`feE@q08T?e*OG z7tQymr)j=(@fRopB}Sjtjf4e2TrtEsqP+<#UQ8}*Z8GWJLps^ygv{>obzxbcdtIn| z^PH(si%Ss3;SEWqLVWNReijk^eZ4IrIvq5p{X!z@HYk?%b%$WCt0jV&yx5omFNOyw zH63W>n2AP=A{nEV#Y5O^=Rc3I+0K8Ikt2PO2Iqf{4Qq?qm1BH^6C+|tQ@bVce}>$W zhRkfa-XXg4b&)>8XXr*-uFnDgha|d!A}xstZ$-Dm9wB?t0W?*tSngWCIzc4y>LSw) zZ{ZjLtsbR{Bs{+lj4>RW{^pl~R3R+W9l%wvPeQ-TFr?Mr0%iU;c1Y7) z!w8a8&Eh=5<_w9cgcJ4qmkCP*nzI89;w^}+0s_>VjqMt@?PATtc0;>|WI(|a)Expv zt8zmixn0NJCDdZ0pgYe|7#jr+O#P=$od0@Ij`+XPA?soesi2|x>XhATr4?B<25Q0{ z#g96Bos%%CVqU-|Y}G?HVT%QdCX6f23)7@3MkJV?eLlt@K~Fpg;n1uE-kZAOZuHqV z2@$##v+o%g!rc#$e*hgCu~GrQX?G1^aj!bD1 zL|z;AIspqn0fB;`=g3o$I@lkm{k1TCWvqHtnb>mOhU7Xn1F>U-J8MmAMEE8RClYQ# zSey6S?7=KobG-hKw&&$%u|mTeIXPz4ANkF$`t=XS;+Spkqw2zO?FGgPt>R$9a?M|7 z+TL^kcH5gkShu|sBMsl0STzs@h$~n)91U+IJt1BF+-}*O`C4f%Zy{_(JJ%65qn*nM zbF_o>ciah@co8Xg#dU&*pLmqR4dD#rUnBgg*^bjV&0(H zj>!|Q{Rs*00B42od32JF>8OP7`f)G>vf5%)WA1K%Z{0|UL|b*@H5=|a3!RUQAe{s7 zSF%@rf?P#sXd<1f2N@(ppuq`_{GtCr=Kwm(v#R&Nvnh6(|%J*id9)NQY<3U#ZUf_9SnA*2p^?!#ZK zJzfeC$l9auS~i%Y#5Net;qtr3%@-+XvwXP*U!bkn|3%4B0=tlk3tGeWapmni`};Na z_brE|=gd(F_M7E+gYCQ#*oO-CH5~5{`>*&Y*e5bKWFN7y&|O%phsNLBAq_HbFN8x{ znj*ijG5k+mWf9p-RL9LgDT!{2nMeyN5s7gBdmaVV5`P8#j-j9`niyh*D^?aTSWn4u zfl(CaX>D1BL_yD?Tw>iXfA`|=MA*N8p1WBlna5)3+Hp$`CXyob`4TZ6PXd=v)fQCo z(xnMNMd~6Gv!fcT4@3H_xoI>PVC;taKWF+MseKq<>^5TEi{VGma|6?(qG$AHD zdcSD~=i6_-#G8r0nWTE#sEUY6PY?suJ%Va3P_ccbN+P@-XeSaiXeo(`?D9UP1tcFC z0LaHI;p0*JJ4=x7H`??(WxuIUdYENhKT5D%_5VOu^q`Q5HfnrYeHu4zEUqKo)6tg~ z7x)h6zds?fe5wdp1Bi5xDb=S9o+t4eQ(@)@OuYE{P7okj|9TX{K?{l`%hXyK6Ya4UUih_cNg)~Bt?JWD0QL~+zVz&Dv<0yxM5b;~pLx3{JovhT& z_iyIPUtO7%4ritx&Yhh(?r;(*?~KoZ-`jl!ap1T=Jvw7td31ZkD776v<3q<1Tm|DS$n5+?piy>)k;1E2zU&dXgyWn!%q3QcDm3 zb3iRYQh!bO8kxRrLVZ)5548}o?nc|a@1TIl_9K+5ntKiyCH5SjdW+!LOFz`W3od}wPEOai25~R zsPuC;uyS!fMm$0+>tUi*qvNn3NB!{-aYdOx(g5RGb|bA8wbT_6(wb*AIyC! z6YM4NijoUuK>{Jn$3DTrLK_5DF(ob(b4&P8*5d3$mhdU zd59KT!L7)O_&0v%^N+VtEP)n$PBk*6#Un!iho`oMVPt)GkITKDQW-{mx+C~-0Nz)u>r7F(4j6dR z!1U||Bzq{MOhEqq60!{9vw)PuPru7nf%FjZ6U@6LRVWaM9|Pe3B%U|I+QfH71`?nB zAVvnG?RB}L1E0nxI=1BFcg zF}EgKdgi#bNh<$b1(t$e+%Au?4{^Le}i_V+NspSc>;(xyzya%(?)-Y{|3vYq_jy=L%Z*s){pc;Y~68 zL6ygBlm+TK+)d^IBaTJf55EZUKJ_Dl7COYr;Rsr&hwq`Fh1&K5NjyJRPP3xb0c8a} zIS<1${dW*r%@zNv%rDq1GM`~&e)^jVGT;44Y?-J2q-E~ysb!vlUozB+znHhN?{{c5 zl>sF|+>aYWsE+#`;_y8rqT@Zx12G=LLxt*a-0!e1)^{VqxKfHRdC)mTIMK_M@P{Y_ zI-%Qt(xbPyu0u_ws|YsLgob zs;v6XJV<232?&{AN`O}R&4Y|Tevk)7Mr;=dV|XVehNs4$GU(}^ zmG%0zN^H+ls4%#3pe)P60i7%G2x8ZwkQ{aTHOde`k5|o-dM3YGX}1y?RcQYFS>UJ> z*$bbnpSza9ImlKdr#+_iU#d;qOB*?8o55-ub9EY%fbLA+E_M8Nsk|)TnH=Q?^{miR zyrYrZCErSUD0mhs5yq&&at2MvVWMbvT#rr!Ns};h4~%M8z8k;CLTAmc#j+9KPaqyt zTIA^AR^)u_)Isp%`Dv;Nns||3GLCy%hIvBQ5_D{f#8NfG8HrE1(0gULnYV~ z^fW+5Bc2*pp&%XaQ&XUmY9u;} zgm2Su5#jb4eu{8Q4G$!2PCf5U_$s`|{jU$$3c;)DEIfe^(fJ)Jm}Mxr%BDp5AEkRF z#}Wp9m9hR_Cc;o8d>@-JikI%8j1>~t30E+oZ>@b> zz$8Hcx9wxR&xA36pv4%Jb4yRoN}7L0Mb!W^LP75SI`;OW=d6M+>d2ZQ!3{6orUVhNQ`l>8qJCn`E8fq%H5Rk2f#mt)(m+nWV;abGdeGLDTNXlH`-&Q0xr}f( z4bLE~x7>bzpQZ?_+?hmA9T;@C7I=zk8MDL;=^lpd{4_3rFsRiyb<4W48&Mu%k7r%k zn(!|gP9eNg!cro|aq)q(C z?4T}JhV0yK039wgNL4dv5X8`P8Z8Blq ztL?m(wJ){Mm8G%vAL$c=C$q(N%=fx0LWtnI$O^<>A~TC)HxlOJSkN;TUoroq)7*jg z!akJR?$TGT*34IVie36Y=J@NA>KNAK$o%pyF_?E3{!GbQX>S}A0aU|T-J&`3F|%L< zT%#&W^Dz$g#C%=G#2Q+_cE5Nz!X8*gAC$pr?jEbGJRaM1+ojScx{sMYu`6dVEZ37j zEL}p-GpQfM0KfAf2c4FyH@_r60)X2QuA<>4gilE`aMdAvSi|QN-ml?P^8kM@&Epmu ztBhI8@A{8BZRhr2b-`YRT1X!5wEYXS60-HZTs@J%&fzQ=4QvCYSSXSTcIM!nwwqv@ zL}%aH&NZUZC{xeb<=viH^fdP0J((Du0c>pwbV=x8sx37*e2ADA)OXsp z26EnMi?A9@2b8KEZXpg}V3rW~5X+6o$vbVE2yvLtI*W?HSM06CecQP9G<;(Ld#y!O)xl_qPQW$aI<)G)YyNz(V^(Kpja!I|f z@CBLjD9*=^y08-!wwSL}up?lrI0-tTj%zrMa2esE;tZzQr{NMo@#R+jgf8Fxp*w|8 z+@0wTymbX&=Wh3Q=nF7$x|Mvv3Z5rtRys-Y52DiX+o9cv(hy&VPZN8=JEx-^^!(Tt z3^}p73n$a88rH0|F*?E`t$j0=;?=xjvPwod4r<(mvUF^vb=4WF=nOe32`Sr+YmEuhA!wy!8g8|hF4GzQM*Zmnf~-nHaTZW z=Zk6FxP@NO?|1p*iVFLH@OHi`$^Ic?Wp?8w)(#|lE{cPj2toRJ5Y%&ZaL-ST}qz6U)g?qoM~6n`Vr+>A{+McbwZq9wgS z8fiG$R?4FaWLd@2h&Hzk=6GfVJ^gx%->C-hd7mNBPINn8KIdNe;o$_oAgWg1Ax9b`A?AiFd2izJG!fyQ`n^qkbUVTa1M4toLH4IM$rn-%j%iP^L;& ztJDSw5KCGu9qIVM858Z+zaOm%_8dX)dov5Yqfwhajq5E(dpI#3y)A@MTcMxF@|X`# z{v5ow(n>Edjru|d%JXcL4S_OWuk6I-4aYqU)T!mrYqs?TqTf+IFi)QbEl0i%XUs(m zQaOeo`LLq6ZxY}kUzz$IzfpN~ z7)UK4FAX#x{y~U@WKpj4&RC(8I*cewuRQMANVd-Oknshy<-j)D#e&E%7BkYGBWbrW zEt7DJ7Iq)f)RaB{1W_@&@j94$0Siu01%qhNqtj}V`wP0~fG+46gTHJde?pbX=VAFh zseZ+8{^+D&E|jlS?ewNG+Asr;6v}!$QW)Ma6<`WAT}3db04E+*YP!amg5N0FT`bzV zV}tlSY6RQS4&4862Qo%hOat}R?tfrmaUMsj3RpBd6M!Y?nUB2K|AfDgQA7xcON{{v z#DCh~99~uEj1*48=#7R8Ax3o1P?${LuWWfmahyrOl^>X6fQZn6s|H;l06hqPbSs#| z#{DE-vxMIa3g}7D%2n_qA{FmYm>6q#!saq8W_>GYSY;jOi>9>s5t zs*2*>3-x-4f5}T)+(NgTXfncX9|XyHNsf|n&?O9>5P^eQth7^{4q&zHdc;A%tx*JA z4gzx2>&O`@(8~hpfMUiXx#q^K63oI@bF#q?cHr9wwc`*VpNW zO1jr19eC;EoZ8Ws5vY4mH2IA9mgiC(grq*m3X;88X~)dy7Tr!GlctM4yHJu&wSEWe zDrR%y)paeYw$Z42YyDqK4*tH4ntL37Lxt?z9IFt`&NRui2FY{O*jw43?!jOGg|E`7 znS%1qq3)OLg;E++V=zi$yZ;j!%Y0+fNG=I~0uU4vr*@fRZu&$lj3A~B z{msR|CMuQ;Jg`C2sMEO{fC)BU9@IO)99W3>PhF(C9C!nhB~$q~@Y=9}v|PFHQpd%L zzT&;eETY8vQsVn-Oo{W5$j>!-LG>Y@TH`6`nbnO}1QL(PO)qEH`7%sSIbnLvWVrf4 z&r?j#VzY@u;`RB|gHNY98!qAZ-Q)vs4m6^>-V(=F#beTLGpW29TnY2pNVl;*f8B9YIiM=P!X!5APTPvegj|g%|A=@n`8K; z+C}8fT^00PhVs}w18XoB&Ox~<%T9|6EFmjn*%-$0-EcyS>~uJnPiwYXhO~02ll>|8 zRVM$CC|;m?vCD@dB&w$E=(xNN7n_*t0UsOm74loIF2=7;zK=afsyYwJaSg3x{dfdV zXr$k8*maF?5g#D~t)mTAQb%vSRabtZDro_dElOh`^0S7HEY>m@6eG7PVeab8?K{Zj z0Zx~amCJ_1z+0|SSl>r433OT3ummbfJ`$*VzG)oy6`>r}tRq>&**G$!g|Sq(;G;D4 zi1OIfFCjL|^VlLJprJ~KSc0CHGxf-H;gjSSCL$tVrliaOogf0}Lo{3pw!{>wPP$L3 zYWR%IuicS2;WHT};r=W9MfVH3y_{meT%8Dcd~;fXg5;El*4Z5MATpssk6-$lF>SHT zyAzaPqKr7Avv59anU?|uhcA4kHS}5_XPsOlh0v1sfe=_b83Um2Mbk;Nx9HRE_+e0C zb1cY1RGlP%tR{wr@O6ga(6HbwqqU@x&1>%1^CTcM$?$1 zt{)4>eDpB97~MVBrE1<#j zqb>s3xCcOnhd7o4wUw2AlIm3*1mmzZ91usrU>>=HZb55O(e{(PMG=WhJ4@oczSOo4 zkb)ZX4u7e+5fbSjviyEIbT$&lyBbHRz@o-FWlLq$!z47`K=|DB)%JRRhHk;agA z6wTn|R5O1^V~Y^9;qWF1Bs@W&lygQ!xDB&kasnLtCDoBN)3<+VBm8=Oh`+Sb8e-C4 z=M{YU^V%$aRV!~026I%3999@Tbh`@?L|f51c)pvzB`CvAip?1$ zuwRHkw2&)=Isi{R4(f1T$tsj>;n=nw=~8b{snuM0{}WmLJML%+CECzYYkmd^K(!>Q zmIg{tjSUp9jveC$CMcy!ly{hRj;I0MNO~+|BYYAVX_6T%BPZI{@404~(@ z>rj~G>USQauy{V1ZKdUFA-dnz>XU$y;cokR%k$E`LxncR}w3I@(mdi;gq|s zY$QroIp#EIfdzl>CermIv>jrgedvf6tc0w0aH6OOedR>o>aP+-3oZl!92B9gyuA1} z@HHSAlMYpq?V3ipCsxjjTbhK_K$W9L-d{~dYdHQ8T5|U)wwp5b%k`u>VkFrHS?4DU z$+>B)9VFim=Nr~T;9b)ABeLLBK-BHwC6sc4@FfGtLLd)dNOYz>e3204cObu^{QE_* zDQ^=&`HE~O<@RKSDuUVWI7uJ?r?{mt$v7vl{=y*icQ(FqJ5t(;xZe%{s8>V~~N{&hS zibbUSmy8hQcL6!ISng124u>*>S2J-uZLU z^o>L=1`1muuP=xxk+a37CUexddz}(_XhmDxC4hmO!L>d;j4LxXu6bcxH%8#PjJT!% zmt3eIJ2m|^K&Av_8T>iGZ(90q^J9|QGmL(BKPRb6i*@C-Zxmg5?!svti&|;!B&ooy zsip$kbjCRm+BhC*g@Vg7xdQcg?1J|qv`w@fo`PL4e7YRf&J?^6ZY%XRTvhJEhPvQ4 z>Po?Pp8k_Bw3t!`#Zg=h80%9%2v&GJ2(o#h)-IU5~j z2M;=Hc=LLwJdVyc?#bDQr2&RE!2Rkp(a9jzzk^_~sMB2MTO1{SGPrfcelRBu?ZM`3 zEzUI{xYX#$3g(lZnQd64Ko~B)@5yXeuxhF zJb6)*FRmOwFXxxm~AT-*w}u>Og^n7q)Ml4*@~Xv&IIaV${i@}}CTjU-t!kv=WrfC1V!? zebD1#Aw}a#ASq}4>E@lkjFU}qxfib+Cjgo_=L%H!WOM?uRWvR)VcocVCPE|4RyXp) z=nPD>t?Q7%S4k=vufUk9gI_T_J<_6@jJt3X^26`uprL4-BC?Q<>6nz@)_l2N7ATj^ z#dmkX{ZKeHvoWS*?gGTla2a-rv4@(EUluqzLu?<8H7FWK*EJSS4r+;WH-2&_t3JRN zc%7M>978RREs4uL-)oJXrrP2I8XwQsP#5D^!V(R&vot3w-b3NSG@hYm3f^cssCtfkX?<}Qah`X&QUwT*{q#< zpas1fXH(vil$>3B+2k1%E%{B7r#y}Hm#S)V(u*_li#88-=R--h0m)X-BFMYDNEyvr z#OSuQaGs0wNTh0`+!ZsCLriM@m^rkOYA+=WLY84=A z(^9@c8S~7SgcY0@_X&?9<^FBpQBxLEWTxy(o4f|}Mc-8g*hwRa;Xx*IkDW$r( z1-bN|8K=WXLTP=`wWhS5jR)0Wi8r3DZIMtH+qI4Ll$EVArtk0Ou44QzPIS3-CA$ zOT{c$k40M;hms`Gq|fQzgY^i!9!b<+$ayZ6Hg1gkeG-3}dJqN}Au8O3akMg4fnI~O zJa=n~mK-8Q^gOoZk>23V&X6p>*rV+D1PS zGmSP7vymp~*@|V!pyzA+Mf1YV7@u+JI;hf-F~Dg7n}{6s20n!d2c71CH>{S~X;fwA zDght%8=Q-DEhTQ)Wv(W0yn>#-C=fjv)^FMSUd!Is5q}Z5EcB^DXa-q-_cE=@ukp9& zQEcX~*A1KbpVT|`9S$cW@z0O{i(!T6m^cn zW%bFk*{@%Yh<>QWoJH1nXf@8)SiU@`|d__d?n@zoaMg z)4YEzd191PP0G2B>9jxHt`Qq*vFti1`jTEeR?4tNsh+$(iV?ILE69o&iLV|9ycjJG zc`Uw=;~`e=$GH$2GJWgJ%Pil~%=9gJ4+F0vf;JU6@Jg#1HJuzxk&t~(_Vi3)p@s6C zzB;_=|KHkR?@*)+;Pl^%p#b*(HHKLjr;-o_1q#$wq?HM`3QTc(e&IEs+0Tosu}5Au z7HR?B)DJR7kU3dw0$4(oM3VWyM^%cPRzZywf5001`8ugF_`0<9sd|+$??KjEXUWOI zZ5x52Ba}Le44woU+z``{oJXgC>3WJzI>$Y4OA1Fh2xYhi{mJXt{{tF;i{s-b?javY zr3L}Q{@PnjYN_Aud{OjeHh)43q@T#Q!Z^OH(DrfIOE=$N zz*Q2OT(OGh0uE0aAl7skKU&2jS^KvQsS4iov>(uMrem;;z*=SsnX^Hl)}~Qcz;kjZIIq zljhTOyZZQx>ZtjZuxhedC10hklorDk}+=aG3Cm3LSDP(+$7dQ#~P|1I^f7s-*2IP zQV(O37${bgF4AqCY*o>04T5)*x9lKl$8+cbvq9=rPF3ItZj=#t2fztz#S6Rw;2?n) z0PNuAyfM<)T&oEW)9_rv<{a0lglXup{?m{?!yEVuPW4rUIFe`s#2S*XyYMqShGe|E z3^KMRCr`smBbX&@wYc>`4u|CWvGNRuR{8d>*oj2!lh+EHrD{$svRPmkU^wLn>fJrU z=}YtDbz7VX9~J9=4MNwu_d+4<@^69g*Sg2w2uBPRP!+PdC_?^Vq8r+0(6bsjz2fhx zRxH#PkS0znRDf)C4c<~8GWh6K^&X)WAbNPpp&|@+E^mQv0VHJ(Iqf5|>pV-x%c6;f0oK5aFA(o;^yK=YI>^;u7t+ z!ol{kL1f#_DTYWT%B?F(l&e{$5^Wt8sYLmP>p`HUVpUOz?%X9@clp97(Ss-^=vjtxp!uxL&yX4x&-Fm!R5qzu#Y*XTpe8Czm`NkQMl3#LEOa9lIv1p*=S1l46n%`$6 zU&`T#|SQOF5Zz>2Ur?0XmC zwtNp==phtE&jxb*{|?RT{Ce;fct2uRQ4R)zr@ci(QyAXl@iM&0vFbB!Z9E;G(J6dc z!&@}Gm3U(HL8HNaRskh=c+lkx+kKhEKjynKzSYj$iR&BE4qDOQ=G{>tVQ%N3C;u8Y37mgdT{EziGzakDzaBF8 zcRIojHoAZf2uR7T1Qjn$1O+ps=c@6xnG^hsn2;rM;Lt-7skbDqPQ7SKo5s?{numNo z43h`iu7L*1Q2hu+W6b>w_U~%(4!b0*>yo4=*}cBAJ*2x1Ztc4oCk#Soj-+roB0^_c zN|U_@#9+B3Nn&$@SWixhQ-55@q9PLO(PtAR4tjc^KyP?X-%Y19-na)?_R3#&3Oqm& z^sK&1(eaC1K0tsL75iMv8;_&>n9Cb|_#tw6qZ4TbgEz}B8o`Ti%7{h%+YlVnusy{W zp=AwoXK*CR{(IT9LaQ+Df(PU|f-O09LS(u90u~9(b%7b$zbplcr_mL7TBIX}Ovf z(*R)xNYB|6b&B_OKqzp(|3})n2UtCIaeq>iaXljrqg)1qa?PWd5+=kMCUOrwgfvkU z%~WnlGn2+R2_csfxg@t-O45x~au+G2a!LC*GAWenc|YH^e`n4))9HDi_j%`!Is3Qw z+H0-7_F8M-_kPX%(3f8*wSAF(=(tD2S54iXL5xlB(CaFU)Sg&~jnL&EZ8Qq0G|^Zh zBX6Mb{>-d2HcdAgk9Zo()TA4Y`(A0F!9B?IvPYk7F6frlk?4uVNq52p(b`z~WEdpb zJx!gG@HE|E>Y~`qd@+>7fs{m;=p9C+E8E6u!JWjfhU%U4jeP_?nxMnskGBoFS1O>+ z;I7%AD!$QGb1tHwxE>)mk6OT={Ymet0!7gF63E?aBuer@KnFHv7v z*#jSpqN|=bna!1zTqT!;pKtKmDp|nVpiHw+C1>Hkco!`+N0WG-Q+Ba&0c_Yh#_eP^ ze)Nk{ZZ;Y{97x9hds>;-xC;aLv$h1ExJU5&30`;)3gB9&oU~W=r3Uxs9s%C10=$1T z-2xv^8T`8*zTPbcPCb`*1P`4zEnoIE<$jiRG+8xL7#rjm9pspVVx^cn+Qie59D-fm zT-<1+eOXp$ziP6*;cp9m-v8#m`1u#`(vgFLe3@;+l}6>$yS*+9f{I}O*(yox7Yj9R zNe;DKapL`+v zMP!T55u!5*+0FCJaxFE{=U%I+uKi+N`y6eo-U|xliR#zHSrvrx-Srk~ zJVD{7abCVl0s`SRyZKRGSytdy;O15x1FmzUYfZ3AowV7Zv~-3H^V|)YL%fy`xw{1jP`B*eS(4TD1PHF zCc@3(C=t#PUnYz}#EaN99!}FOmkV^9%@esCFT7Ry#n)xi*V~O!y*or04ChN(p}oHL z0Gj^K?&9JG{-+j>zGt-czbzbXwcXKbcwXb#6BT8{w$yy)gT6TzvAV%vtntE8_ff=C zUjxiy*0?^h7`}Vzm-{)UZE4kG;LGW&9;eYaj5}g^V0&s~$_G^G)%?D%d;853x8aTh3tBW8)|Zxm5v`{GV5vu46R$zsY>mHYW2f zTm>iAZsD_$%Of|s`>^SUYNGQYfqmwD&Fz_ZEx z#yd>r-@lU~^H8=NX)>=*G)c+)o8*UNzJ!m>>(@6HA!zabU3d#Jrw5oHiOww2=t?(i zjpWzW2p{i8|H)Ds-YR$i!zXfLp_txO-6&T6{tc+j6!Y;JAHD=V1n`Nj zqO-+OUB<11ww$-QZ?eJUxk-X1*93kwLaulBDRhrQdyqf&JM$nWqkLZ_s*ex2Aeo7J zRR;anyx2cusgKkMNlL>~l#c%olwKI= z%XDTyY0*wu%XD`-rS4Bc>Bgs$ly*%~S`z4THtBT4V8o`uUk8 zrDv16Uhew-2TEVw=F4<*K%=+$pwD0xZF(@VDHzdD3$12*QJ52 zcYh^iZzgVH%@#H!Pa{$?cwUKHX;#N)+mURs^_m#=-~Z55#6OFTUf^>|D(ZU=CR>AG zf|&;sQ0!;&CS~O2lDh#+@{4^RpiUwhpAM0MXr>^kk^;A_5o2qH%Rdh3ABv* zLlP-%@*|4Js=YQ9xD&)X%i_Xul)^K(|7-^|0u0lBQTcg=-7v|2o)qBB9EEzPdAe|t zYZ$rWAJP5=SXQCs@wwp9{P{4LaUV$U8OQwgv^bPb{}_Y{-8UAhE{dereW+$W`9RKv zp;e-?(zS#N5lNifqvM_;Yy2rOe6=UQ4e@#0rPpDNF3V4pRxNDpzTE&bb~}FtfyzD{ zYh{WyiqNSw7UrmJ1fD7)w=08+qD6^F?UX)%@=nc=|B*KcDpx@0nJkvP*;#SY$ zO~5n|%~wCC+H>G8v)t;xbkB;%`^dY&lL$|*ms=`r()$l4N)I25`M&8+I(LVim-tgZaM^K3N`vV=VM z5|N!}>AP#2UrXOu-MqBPT7W9fY{))3$Uc(aI9Ka%en5E)NBk|y?yljx%MriDpX)pu zMs&qFd}}+cJBJkIDAOXX$JqJJN7t2VSmplkeTOijOq6z4fte@`C7v|SBDrby6sp-CFf!)L ztXauo-LjMiD(9~^=5StJYk#C2Gpknv`z#5iiT6S^KS?pf#pg|fth3pmO%&%94?Ien z6T!y)itU1H6}8QT>60sXvDixY)wxd_dzzJUgZl4=qk3zGXsC6sDK5B%HPQL~Rp=qxa&4onl&s&N;6BEUk7%V1GR_6`BG8Zr3w@TtOe3kH&LQo0CzE<2m zVjtw)+-`(oJxmwp$Ig&VQ15M}RC$p$3$zgl1w@?NgHv3}rOLhaUn>kVize{xF0&6j zd%AuE(+=C;EX{4(+zTdR_IAWv_WFFOZbMtUp51QNDR=KQQOENDc#U0ERh-;raDGn1 z)|2^zOOjhp2BtqD%t9|&%j)_S<#+AyQ>c0tOqo~efN|kGcv7as9V$ax#Sl4PkLXf( z)?CK;7i6OCudx>ccr!u8IdUHszS(C}+_33jWALJBv0Gyn)u^cckOwOI#@N+pYWQCz z!2ZB0EcqS&`;yhpFRlI5bu^Ud@dakUGQtCiN-BIL8ffEnO^Ur`&Y}3G z_{YPPVKU1^jaVo7SA%7K(^h+ExuZR>%y;csV(6TBmMrrX(+cpo&-?Bsi%J^jJ=z~6 zZE6La_Jd)Twim2k#`)}q2IKsBxQufTNrp=G>R8%ler(rU%|a=ge@)nGQEp?dWrzpG z#Hwzr2 zPU%7=gK`rcja%vkdiPs5*ynkG;v|p7j8-OgDjWoVIl7KTlH7Bgqk4 zcw^uaT$Py<3hK7`0S50?DO6JTM~u%A!Dqz9`Sb~2Td0v%2Jh|FT~Av;_J})JL_bJ` zCvMfSv*?e%q*i7N?K@fR%%?q7JxKjO^y<*?pl`8b5|p&Yg!V=CKVuFP)n7O&&uh@P z+sKDX4P$@vh(jWLXo_le8r5}d$O6u-u-X?@ZJd%ouWB&5>#|V|)qIK1hjI^3xoeX0 zm7;9&ddFqqHdIq>1d#rKAsTA0ahILwTi&^9_i%Z(fM1+iz*mV&g0vQq=uA6fliL1A zJEzEXpym99_A^%7Lo=WXRc}k-m8w2MDobp`Cdp*?CJi+)00Xxjb_gT2!|oL{EOq^R zH1`}H@3peYcZne7f64Zq`I$a9c#m}7B{3eeCaFFL$FquKoEKy67^}cR#{@UB@2ft% ziE)FpaAEMieZ*C);j;%?&$f6~r1`2)O*a6#Mu!&04i_YZXOhhfVm(DG>+oyEZlxr< zCoFnnY1M)f_W)tKdz&c~e(g6#4vFxlg!4>rXX;RHR_koIG58!)p4V!wLpep#*&i`K(+)%q$w#um83R8y6nnf964{=+e7hN9+bCe)k0mD|n-6)`M zrDF?uBLT;ACfl!K1=gMTQHEGiQn-=vr0%P7x=Q%O|VfZ1#3PQNi}amVgBM z(tZTns$^*u>yq}Jw6q+T=Dv zD0U_PD9k-vmlHe9d&^rCKHi6iD}1C6pRMqLK74}0t195eo$m4nCwwgP`~34hvw0I9DWqq3jH4npzle1 zE9V4$Gjnv3z;r64#y~c@n*-2Fe(lP0L4^X$TjYt@0&2o@Jr?D&a>1x7t?W~ zI?vQs*ET=$_Gt62O@09+RIPp3DGYA`CZn6lZ=3-a`#lMDIbv;mhCZGUmRU%O4JCe1 z7=KI*MJ-;~;}5EOBJRshnmXhDJ5pk&LSu7oyOSiO-S{JajW_YGc z+q{)xy5h+c!#f8^w&OSct;^bO3Y{TJNlO0%NeF{08%{F=*d9C{3&;B7-WACuGwe1K@%rFw{h6L;UXg=2Wh2or8-Al5rg!@Zc zjV1h|Un(Te{SbtPxRPVMX=&m>wKA7q_b$muMZ<%h2Pxh57X^^R1o5;Xm}x^?%|5A7 zbz>AsHQ!;`>)kFvNk*p67>8=SanWke7~rk1hvi%8&M~Co?^Hsui)2W!eQZPYkwK`~ zwFyFPHGk2K8gsi@OyDrC>WIy>3iAJrDRrI&59qx92RdKcWI2;!l?Np;#C}CFy^X{_ z^S>%M5?<|QfFk}zO3^0sEwy<}@vnL1_mUm0n1!&NL@OcQNMuq#;A0?OqvKEy@@6aQZE1#zemKM@X&6Yens5fLnt;lNi3+C zp^CiBN4jeR*zQRl+$Bmrkz}DfOUV#9m0!JioNy&UOtHX9;UtYe_wouN+ck zX*}J@r%>LeUt)?ID6jq+z~5-@tRe!w>)i)Ld$$&AV_es}*NMXI#dgvkb=xSiC6PjQ$5C7MC`Vl_k84S>lwbX?;lqaiHsBNCXMT13 zf5QGh!hhpw*UfqIG|v97X)vWm8TK<8rx^CmhW)=PkiQ=5`@Vk8cjQVh$utRh?DGNO z{)=B!f^Ydh`o9yG$38N^e$bY%mmBsv!(QXD)0@HRX*9$vM7yHQ!O+_zHLmOZ496+T zEuENchLc4n{=7;h**i%^Er_V>W)#Y5{e4C{-@nn_U^p7@>hh}filJGyLGFe7e*=rzmLm^ zEH3w|TagW}(l1awAHa0W3e(r{+zt%upX)r;1F}LguAx{x18iFXqmupcMK%$zqZHV7 zhP69bIsah)%k2Nj0iD_86^>~Ej?A?F~Za(Y|XiFZ`0YgGdr2O3=Qk%wG?A* zrQ7&9=s!>!>=;=%$ERo&x)nhv-+f9*eUA4O17JjebAsfyZfX!JaL*XvMmN!7u!RRL zg7u~Vq_rDqG3WTkV7Ou*)<1vp0_mC!pb`&2sc*Z&cbVG&W2Yab zuY2f2mn5p1Li!kDANW5$;={T{qQ`b){>eRl^`E>)!`BJ=x=&y7lFOGFXWS6{!;?97 zF2QJh^PQ$Nk7oMCcITf@qQy6b2U;y^S*NRV-P(CfG|@yBq&%#*l`nd!p8r|vK=W$M zPcY@>R4@x{D-$5#bps=^_}75QdVx`jHm`9@jtYDqI8-NnqTAaw*{pvWw=31Py8K_; zLYY3lJgo6@sqr$?F3j;UXIPB-@zGa`Mg!=&zGnwf-I8nfHO(nbU+f5;^%Ln_a3cTY0I zW2a%dY)*-ZOw^k9&$YjhhD&a7Oxz742!KZ|NP-_86C0T62I*0=TLSM%t_ zyY3BN{!J9u$E>gzW?&iwzWkkLIkDTVFFh}Qe^8>*?^)Z54r)fZ?^vw)EECSxJvwAz zxu&ZcsAzcNDJGyE%P61tj%+}>h;j-gQrBa9ZL(W2hp!y^o1Fd(syw!XCvKT`g|KRD zdIQIY1Z9}i5YsYnovqL$zFBF{-jl7Qyrm z8V*Hs-+er(7)>|Rz)+1dp!VNC3J~-EEubPZ`1WUbG4(HZUtq1`i&r1GS|#?ja&V^j z$AGg9__Rp0x~Q#bJEH)bJdv=ao+ilQ66EW1;7Ruaq{%Q;eHszz6&UMiqONzHKDDyT z;;%idaoPKWE6o5y)jiM^cY!FJC9MfKfAh?zc^lODHA{A%;3o5ZiGP`3C0??VN{m|l zypHMERSceVe@lU1r(Lp=xQmQsH5yNoU4qDSk0|85!x)0rejMtV%xkEQcev6(OfC!b zspe88k-xQ@vrtK@si{Gzr>k?`GmWLA+2B+|HAg9vs^r#7)ZB7&cESy7$aL~Z&Zikm z+Xu`Qy2~t;4qZB;%A2f|oj?-U;;e51?=sJgHW% z9=3o?6{PB8A3ntE0Qc(0o>$(N4&wT_iNsZZ$pPaiTfeCsw|>3{&v%Yc>~iDE``AY? zX<__O3EPnn$iB0WmuKGvKdHvT%ig@Zf=SsgIniXl;De;>yTDrGx|Q&65=1oI#aTbN z-0ky|gg=NYooqTgw!>W=k~KeXRYT3+hoeYQ->l}(_tcB8Pf|A*vU$xXh8XdJeZi)h ze|)Wo|GpaHGM}K-r$JZ5x~JO3=Gk;LR8xZlC`{|lo%v~h#A2|^M!SMGz?Hx>k*p-R z*c{LnC3w#Xxgg1{em_};E?Fd~mlit;A{uVGCXnFDA61BU6oU6(Rg?o%vv@jn;y9B` zht+o7vm##KQF_99JSnlBw{+0XuVTXfY^d(GBFb9>=;#JDxpUNyO(lUv$j zaaPAjDR=MMo$a!fr7?J+x+58;OfLW#Uwe`7Bn`B_P9vC#<(z-|&aYam|G+CT>%T#$ zt=kcfq?psZy0{(Nc)QbdxNp(7=LL2Zso2L<9w3V-ak)Xys=&QI0DpxIXbQj$?Xz!A zEFvfT>1*_aKC#YYOVzieqS|C%fjPp48*_$yA#L#a%iZ5&^K)Wbg8)ZI)us}fjgwpS zU=se^0GYMY?5o)}^2YZ~Irl4UjU^ST`Hd>$J@Gh+-qF->nTn8$CzsbBuGMWuyWG3) zBPI#9iRg0>HBSduF)c2gJwv%bP&KdEO$07j^-QmoP>nh}ca`iAx^% zVEYcmK%DHe(wPqhxihEs4!g?@)%baa_mK^+FXdw>%peS#1o3)zJ3SHg{D<2wn>zn) zQZbPKTZO70RwdQ`8N14`Ya>_u{lHK@TP|s0bOyq8Pj9)q0*%RdwgID)Sw_PzA8ST8 z6Mtgel4aIAU+JECPcC{*{5Ea~_zL4=4t^Ly#1(KzZg&9=W83ljz3Xkep?lmkn%T#duER4tIfK@+2Zu#7&d>4PdBJVc9Mz`# zy;kFmdiOKG8WJsfS0$`>Z}7{pa|#(7HKBu&P-pi{kUG&)k0Uh#?mV~O+cGa0qX(MS zZtrIDMx7R&ugj{Qr5ch2l*k~Qe1I@bQap6a>iDhQ^1Xy`FEct>(Km|11PeY}G(o7N z?H^+ok|ZKMQt~7#7HtFH?7#abQ$p)0sh7}jb8sf1-6m&BXluUHCA4&;m(Yj&O7)(4 z$8^9w!Y^A%DO0T9Ba+ZYNvJkRy_Mf!H_p5`bmqn0ljr1Do5SNjCBiFLYB7uDZl+)u zuFDm$^4_)f@7T`t&Jy7{q^n3)w9WeXRC$3_3cs|UuvFJeG4&vD%|v{O_q7FYf(Bz> zCkX`(cM@T@auIbF&D}KL8(d*}BIj_lfe<3yq+#r$)N^OvToqfQSPt=0A0~L8s_xh|L1ux;F_*_%jsMSkWr| zPWN8ts&lx~H3x^DLL0h5C9l9fq`aK4!Yqflk;mk!*lk{z0yzreSMAw{<#oNgcD!UT(+XJ11EImGOWe?fGg4{YctM% z>S&zZsci~devgv zakR(D=Z>EE!fV(vDy6|{gH4aSSxJro9S^*_mlsOx8Os>{#KO9>C3DWBz10|?Vr3%y z*K$7_-At)K{WZ@!mtXjm(B*|vgnccJ2P5)!C6J82!~H+tP~Nusep4>&fPk4uDyyF< zzucHv>7KA_RZdF3uEWoyo&Z21+#8JCe{8(5-(Z1w!`5!nD1uF+DXcJrBzG`|HoD_s zTJN|XNJeIl5^j8ubHR;1XWGoYpUVE^ERoyzr58efz?E&|n^I3maxH}Wfe)4T8-C?V z^W8FqC~t|d6T|rr3o+StXb_d>Ubbu--7}UAx0g|aoz1w;g*pMvpI00B8g~=FM%sO2 z_O!-*5!h1(f{L9J;m*c&!j+OGW>=?37EOecUd#aK=};1n^Xb{pii>qBiS;&N^CWL2 z7`wn5vRJrQ-`35r24npjKt?srA31nyP&U>DKaY0znG)bM)HHR5eF6Ji|F^TfKRv+| zOSFBz5B%BQm3%fj+uLc6>}Pw&oZ-**K8qf!oR0_k{IF>~FKC@q`y1tU59j;KkE<=d zQJK;%Kkh|(!|~_ylOL+}Q?uqwU4DE}w)DVcdv%kvMt+XHWIK{>I(x(1q*e{E5L#6p zL>0S#2cfoZUSOaZ=T8GS`}Y~X&;5w(;pIelHwa3v_a=AzIj^!+q!Gte+UN5K>Z5A* zmSXF{mabDiktdw5eJ#U=lx#D&Ff^^CgkINk={ur5qK+b>&8!=A3Z7=Htu&9j`AX!` z$<+R9J08z^`7b*{@(eyE7}H9nBUOoi1CP7k`iYf0j}|(cGrs zj2$fdpwCNB&rZ1o@;MNk!*0&uvb?tjcL+`E0(|&2;M2EJPE($^x)jmeClp>uSmn)= z!S3f1W|NRrXrCwn;EO)?o*3okLVNTo6IzQF>W9-R&R#+Jqq(Po z(1g}e0ST>*0tju>XAMHT8myM-j*q?23fzTR3Hg>_OxHkYk{l=3o(B_%7ir`90bX*i zD%{eCA6IxsAHGLn9SuzRjZw_}|8IU{+N0(-+JRFtdWz3Rexq4IcE8c%H19VmFb!qx zCj&z_dzqo7{6??sz27La_$~X5Lr8D%8)k1!k&Q66$eJ_dH`dSb6`I-NdM|rOnfyk# zC0p9@pM^FHXu#E~rf%nlb7tUO_ zrE3-1^-sT6VS6-Aiv7dvY;veHQ!Y9UeT(MaS(`RF(bP@PmLu(2g|VbJ)Ja|PLo5!> zmNW3x(@&Mmmw&PanUlS&cC9<)ag8f@FVyZ#S?S)h%TF5bNWcqSY3QP$OTIarrzyTI zKCoQ(41{UAlYRbt?7!SyaqT4YWhLeApYD@Hx!i5~Kx<*=e2BqPq*hOL;HBuLs?6Lu zQ^A2pUZ){fa$rH&jhJOs^T=zWp^;}=B75IhR`Y`d*1{O}+S|{1LRqgYOd!7oGKG3u zSsx9>X^{gF4gUv1saTu2AR1)a&T^;D+!ugMq;&Ad)G2x}kvysa`q=S%#mn79DB*S?lM`0t&3wdT z3N?ifjh*@uquGU~cFT)q4lW2Xp2;sjQl)iGX!zLY$~Kc9{a!{Z%bZs|j#urM{)Tc= z?qSS1h?BzDYe5s3yIIO3Y7gmzbh2OWzhG7-7@gO<-iC0E`#_OJK+HH;d_>!?wPOkO z?$@iei}wev!BP>|@Wp{9tBzZKXOv9bJ!IXi+IRwhnN(eq9&LZw2nmiM@2I7yKX0^> z>0)KlFyLEm^m+#wkGFCE!VgWxKvcw!QsO zJEzYQ!$QtNUa4t;M*7J+KGAT?)>hE#YXa_{_)H#mjT-{S*h!#g{o@0r`Bi~haNCx9 zRm%qJcC2w_f}Kh^6!~Bfd9>NIwE0j%(xUzCH{tfQ-+C7gAX>Za6q2!C*A*#s*UZ9Z z%yKfLV4o&<--}`Ah`F&kPmingsRM~hr-C)e(afX2+&ol$20%<>y#>=J5iG6!%Ea*n zr6@~X?N{h-m%-UV_6!TwF}F2KG?gBEU)`ZK?ot4a)3?jubiN%5f$tgmdetVuH@aiZ zYSy^+f;AgB*ynODzk;A1^qJg~`WR$+*0^neh@Hy^TLYPZeFz5alqC~vY0qBeP>|UL zR5GlTyY=&)%!!9|oibY_*m~rT|f-Y0KM z3~MWP;cR22&MoF~UNPTA$rWECM5%l_YQOqD`P!wQ`Zndj0e<-G-_&BZA$7(`l~;gM zNKwVPLG=cxB0ODYQQ~gv@AKOO7XA^v)LVJ-gv^~iqH?9bDm9AxPv8O}SwnmI%Vvu1 zB7RjnXX_i{#p&~zmB(9!{e6P9BO1LQV&Dl^WZ7wfzAAT<&zh7If8Few{m*7&8{I&D zHFpqy*Q~=`&bPI{wO_e4FOhb8z{nSTNW~ZP8@mdlA>PgB1B7mtemCQ)!s@WxbLAe$ z3gfTPI}sR=>z1|wZq-yM91n#=_*}SZi0>}S2O4EuQ`}7lt|9RlCHe-l&=*DBl>HSu z3lqme7O1B)2LPQcNWC7=?+uwB)0wH-{ul3{ zFcz|RoY?t>B6%<~Ep^wY3dbei#+{+@@w3ALYC zk)SuY?LZ~{ZEC;ut`2nZ`xD`To-FS_vw19jHtGCOTI%p1l;`#jLM5)wvNqb{6RO!C8@4KW<9I)JAbS%j*_-f2 z+?heBr#ppE{BCQ4w2fr~dAVs77exjns|qiy8&K#HWm_Xce`t_aYR;#@UK%bFa81Y1 zdBajRsUhb|d`)j@0Vqemb$9qaYaOnh@6u?q{78qND|6?B4BC@qZ_2>E)rwNKoJfa% zLN%Wd@E~kksak57B&4NRs6Vhkql+Fds9674QG5OD4@D0__f^F1?ixZWbZ?7kETrl# zU`&J;EtG;_NmEXyCHR3_QkEkdkhCEb>qE3~7Q2=K&KCBYaN8>REX$Q}f4}Gn);v$- zu>q~RGFIuIL~Q_K2_1RjdHgqGCo&wI9SP78A&#QY;>)Z(Iy-* zbo@u`e5FOhQGXr%k2eLbpnx7d+-arU1>Eixj#l0N%2AZ-RSP-H}-|~SUCT#k2iyidCEAIkI~rmbRgVs&q{)` z)+u};pc&6)glayQX32kz=_D89;r`VauF~&!9RN%)YxMIR9lt%ES)&D^+a3f7jq%pV zVrOkKyS&-(NRy8etGNHi)_XkG<>+g`Rk=IVV=>*;6c|fQJWDD@x!aKp zdQ){*Fa|ta4G#g!nB<@tFVWXs;3`LJcyq4`9K0#9(LGr&QF=DcH8#?z4JG#}gKeg% z_{=;c&2>`eyvN3eLQIr*7c$6%0x~;hA#;2hnV(Sd3|Uud$HS^YM&{#g0hzV$s+)Gg zTUsON#DA1qBbFD*%=5 zt$iBP(9kRqUaVEh6txqR)GiaXn;NJ^XIih^?MJ%0=o;c^9fIeq{dk9)sJcw1v@>-? zHpl%bJ*IWu?=tlfak;>!x)*+x=FkqBia8SoMhEKk)J*1IkHL@@hHmQy8viWF`JbfO z5hF~+Pc7hVJYw-D4BW%H%@vuW1M_vWCUJ<_)e=Q`io#dJOlk-D0(I`W)&ueA2V z&Q;$Pls0)Q8Q0-0_8>4;=RvGxz(f-7R()F#Y4R@V_T6C~54w=f#c%PwMT+C$pr0=E zdwEiVb6Y`5$s0+I&dTG*C7QuWbXJMNRX>~I7YlUhGZyYf7|S|klGtslgxLf46pnR- z$+t;SuY-fARX)loYLd`qdGP16qb1fq=BlgzG6Lle9@XVb7t$+??pb6{rqpyp)Ad$U z?aBofmiCtr$9qi?ept_Z-Sk3bVmsoVuI0YD;?IA#@5FsNZlPUE=c8&@xaszJuONcU zPbkT1lak_`o2=rdJDW?fnRm`Mn>prB*$i@6Lk||N5OcYE!G^&#Jta%n6DxJI^yA*^ zv9-jJ@1bp?MJ3NJc4>J8;KPOcVtm$%?FgnBW3bE?|iGM9qvW7jn=l6O( zMECI^shGka`|uKl=lk$7g*D>=%+m@#=fh(ael#ttZGbR27w+W9Hf?1LP8Wmyp0uhj zP`IZL&sF%9JLRxvwN}_&v^CrIWVKw~FZXNn6?3LOsgfs=P6Vm?y-5h9HqJvh6qF8T zXQpeh#~Xp8y6d8{he+b>s#Pd=@Bo~mf(PKPDUF_jkj(^=AnX*e@3PuWr5!#DRTqn; zDpwE;Q88Axr?^<=Nx@t|+BGMd&yT#ROi-aVG->W1Y^*yRWbE--O7 zhiE?2jr-XonzHBs_%*XGt9wo)I#+#HoJYhE#d zrZE|aERv+*E8goqyX893%_>nfH(-gEyVf2{z%u;1ZOG0f$3+IjRiWsv4<^ zQ%_8X-I(WwDDqISlve#D>_>5>MxMJ&u^$T57tnn?5!H;Q8(na1j2Sw2`<_61JB%Jh z;IlVOOZx1=2YP?~Va95r1`e~vimNp!dTcMT8Ojr4|pjl0tHwo43y^XcPL(VXd68kOiDEBggfnrOLTPwVc0&0ar1d^@r z+7}21t?-u$U!r=8=Dw})**^TL!lw|HGN!rzy%V`KuXb-oAfu~0q9ecg#%QjJ{^Jv7 z?_!ld z-68}Hs);{{zUUP$jpM695TBfSVe6{RK~5EoKqY2BYTV0DxR(gWS~X^J46ZITdG+-}{#D%iY$H$n^dl*;4zaB4>3I;tMoJ8a5k!sz)oVlVa55xXXq za(A9V9x2a#91u2E?}cl{v8aC$>98zRJp$agL0q6s*6nI=o3!i_Uzx1jppWp!UnzO6 zz{Fx<6s8f(XTZe2Kp<*@b#|dx65=il z+GA&WYnBSRfgaLJ^0T8EH)sP)Q1d9qfEOBjqo7B1s`!o8n47Bo+823{ zVO5fQ@Uj1@OP2ukf9leWrUvP2#BQVhkN6TM(#vqNXQ8bx)<+_HBvJGvnd@cXs#i(L zlID_2Yqhw8)o$mrraR?Le^9&Q8oBmwiHxoD6_zH@&!4kuNvHf8kOM~V{)x1+HGZl;jJ3ipiJ&lANK0#uoo_-rl zSWgLJcY}|m0nsyp+htrb75Ix$X&8t5uoTA)A{<+(Hw~k?4^OrRU?lN@Cer}?{ERgK za;`4+u{8ivlG*K$DQfGUTG;^Y9zfMI97?E$>q5}k_Jd1T^AmRKnko~+dAh#=4Y%Q8 zSZM2;(m>y+GY&laYCGq5g!2H_XP>ch*kQ+O}>;urxVz)39zJD zd2>2{gSz6Lf!HCUnAT{|1KwYP^1#6nU%J7< z^@d&p4C?9ohY-dYkSD@7XjAj$2W{vas=k$)DtD)YEjGwk+QLb+fG4Nii*KR!7TyaQ zH#@s}F47yy-!=khI_)KpEOpYc^zgh*SCvzhIc~l(JfFsKJ2*~OWovQ2NQu$hW5w8W zk{$bB6xOXdgbz{pULXGZDZ-lJBK}8(Z%zweOE?jp|FR7hyr){|jv5cOU@UUh^N67W z>EP_+Kg1rohS5|Y#yLCgm=n8{T5#vqa?+;E_XMmI)`BN~qZG@3-pt*vAwG%Y-87S# zBQu!}Jc%hRuL;Y2jbkSg8=rx$rOSDgK>K=7DT@310WglGq7&h7Udk8`oeeD;4_zlu za+a*o&n~335oJ$%otJ6^4bUN?&|6)SbQX72Y&XTGtK$xvXJ7j(fCo0tzErV0Wn`y+ z{=p=pp58y7WiECS;E`xl3FZ_dwSI83*%(Iy7_}LG08@D(pH{=sE{XIpj^dJIobeJw z`ru@t>j1g*fyF-yJy&`qZb1HZ#eVjrZ z3qVZn?(utpXlhT3!C9?*O3o^yUpp65;7HYLc;95N{n=l{OtQI(&gzEs`5R>1T_lm< z&1_aN3EYQlQGvg6{rzWp%i4n|6Jo}A%9{Ai##5DNl+{`aRFWgTJv_>^N^;vIQO2&q zVI;y=P6>K!x8g!$2hpr}hU3#S#v_Y2P_nx-6zIJtO#_kxT3`EC4u^hRk z2Hj&=c9+kZR#Yz4ao6~~icMy2fRHI|KL9W)!6?}xwc93SOSK)hYdrDSZhy^Cb${OFoO?D0ht6aov2yoF-PVz&-Cfz6G@+FrUc(i`^D9iRLVi~q`&GuTVl4_b-v=x3>yZ2O*5`xh&@dh~zW0{Tf`lX7D`33uUnO}RKm-&_asz8_OH=26_oy3!B2l6m4Pk!2ZxNVke%AWd@ z(`8;p^COubq-CZh{f62at1Qrw2x%I z$1_`yxfaRt-JV~nj5BS|wi{sB{Ka{$sZxL$;4ZLeEsvfNgi73!yL@V~d)GqMMME<7 z^9?nts)PmM$a7Bw@P+Ong|rp$D>s^;zco=n-Q2n{9w|FW;_h2MW^r;61Z+4cYqL7S zmHv??Y)R)c(qVq6`Xm~Q2u}k<^fbUL%8Nvvq|o&yQatr>KaM8(-=YMvGYpwJobjc;T*q3BI&bk0gFB37`Cm&{ zDO&;(pQKOTe-hp(yAHwO?hRi&j8H8Ly+X5O8ay z+RM6S<9!|aJq&D5Q;J_+MB0a0YU;`}zY)s!?(9ksRq6okg$fz7sZCaCoG$-CLjfK7 z)4k6+$}HHD6l|=W#m%%`S#F%!o!>=a8|E@NdU%{yj44#E=(dQUn-lm*p;G~MblS+- z6g42jGC*Q7#6# zl?TzDFuc#PXVK4uEKB%}VNNJ_ zKkgx4m@f`F)FO1h6oIamj^7^Qb+bIX7|>L*X_WJMBK+wST5bM-|Ma;FB&_54~fC#&caTvyZR6Ye&ON5rHznRij~bc@E5@y={YdC&RO9G8m;ZD zfQK7O=6H_6G7$qy zN_jy? z(Wmp7<0?w+gZPLU(@JYoN5 zEzwjSy7S`9Odj@W#KX_0?f}hx=`@!-Ai2uIHeM!w*Y- z0-ytOAiH8;QswqBgJ|qVx*NBT?Iv#5xRbKN9A+>>Jq!<~?7d}gasfcR|2{$9WU&3e zmj8_5>YX!VYJ6mNf5W{m*-TrG3~*0M_cRss{0;(5YCtiD*5!X2P^_;;HqrL;*|ewv zf6Hg10mYQBvkxfp!hS$eNI8_XK;@xa`TFH7VW%w(DE8}^KA`yXAwQt_yjI#pb!6za zAH?mvtDIu)&~q{d6i;Q#kv5>Xjr4{A#SO_1v3Pa1oWX$Nsj&Hfsq%jw#1(%0zsh`Q zmC5|%PrS_g^Vvw|omOR+`Iw!(%%4TiCG*GmRe>JXFZOfHmSp|{Z{5b*Cc6c*-LoEJ&TQ*w0%{&rXiSYU3wjlFhz2$-X zSy`t=CC~i9LqpTl=tRTgzQ(a8cNOBcw42@C0Woza0l0~)7NV=voumWFS(be#SxQf( z^l<}&lwudF5`Sv{R$2(d+&ayTg=%zZr5R_<&L)|g)T1+J#rq+{&|1uB`6_vNjso%N zZhJ^8Hs=)O&pra{l?>(tefWGAAK|6Zh{h(-z#0!<%}@BKrCPDUXMQu0rlf9WQEM%K zY8~`1`ke5h3n_lS)=mw(2MnVQ!M{imWB+In!(eXxael*DTgw`}$C4lPd^hS&=<4JH zE7_jlz-J0#0_Cy3@id{0`sOp$eYU-GY4YB4B@>$rSZD~PhAF9cMHdMnU^pa;ySCi+U z1M|E*@1{4L2(NjtVY}#YkZVw_a8k9d65$5bitc&)M|?-CizG{REmqQTP#B~LQ3B?o zFx62ae9^thmMl78cenOWtyTNdT(N?!Vh``mR&9%5qu8DHAw6jy=*j&o$a!(itIk@k z=3Y<69cty;aY>*jGaf~1?A~f3J&W!+<3mJ$nhlipoTTvaK76mjNBVHJ!gk6cy5|E5 zxAO5d3Wt1nq{2-Vwkllt%MO0wh|4vRTe+LV|JY?ubtHI}MLe&FO+@$%A9UizEjvwR z`EMCUBYhV;W_;=@%q@0C!{?<+EeSpg+?z5+jWafl32sKk@6o#7DgD^YMn9)A*gJ?A z3~J$?>T7hNtMROjQq|S4(P}ie#kK1AZcKJGw12bQtvw3Mc?;q5{?}MF*>CaRxK#J) zXk!I)JntTgH%@hAYeUi8p4{XRTc?OJpxaSYkWz6*$QXpf}g#x0h#Va_(GhN%9JL4wvfC zx9*Oa>;;tj_`b$oYa1r$eO0y3YP_eCb?1Xk6RQ^on%Gv8G11)9p%cH4z;Oyh2^^|G zHGw?|_HW`8&5g3EXn3o2VyXS5U9IBT2c-V5zXs>ux2i+D3DL=A-g&DOm!3!_ zns%Tut*Y4p|F2f&0J2#XMZ;_)(NW5BH=Lk^Rou1qCzlIAY2t&*r!C)6hM$XF=-wZ& z4>`G>lctp|aKq^U)b-5cL2?~bkcMhOgacvzphzh<^t6c7ZqHS7)kS%v;&%kZdmkI+ zSlE8dS7bWU(BCbQHtWhdtO`{pUx`<7s= z6eNm28B>hgQPWO@tjP07SMdtmYl3XyA7|9+Nqo*l+yK(D56;5e^}I zjQTLRk!DflN8SPxs^tuYxcaTWs*dgb0F*-2 zOZZ|Q=`OVP9fDnJz6(_!BK*-#ZA+{EY7(kB*@V)pC^Ri+f&$08;$(xIl3-II=PFQeKVGCp(=oQtY zYASal=4o<^^KUR<8+(G!evL>r~FD%Kw zqE~3AoY;@l2R8sP;V{INHCah3F zZ!JtR*3CRgSZ=* zq|W*i>dv$XEjNz(U*3FNi6krthU z$H{TM-oSxI=k!syJ7FR3IWIng>DX2mW4wfo=%P3o@<{Zc~an3-_)L;a;QG+Q64>g#}Pf88G zuQ4@{inxEzfpxnaJ0Sn;^+f1JwsN<-kZn#$2Nt>i^S*%69-fjE2}(m6D21w@#u;$w zA=K>J1EzCsPb+YzWRW^T(M}nayN>{XYbhwWMyKIA({LRkTqhIE!q(Zw))mIqN{*84 zfFt=gam=C4(zZvp7~Tjwg%=PWcJmaQV!L_tf-Y0IU*0FHJ#tT3IBjc8WjWlkEY?iM zTSa4|Fk?8Bo2&aZ?!)~6QQ_+eU{z(hEOg9*gE!{je1ppdX+IAnXq+{R4+VI=J z&ig-6S$daYA~1#hI$|g_sEm)3PUW+9yuD6tZghVIsy~L!DoQGLKPUz##QAT9sK7qy zU5DVo)s|KiP`QJuLv*nc;eSCGAA(|*nsZg!c*CHf??){rN&DR)Mu!!y44kK#8(Pp* z;-MWaCtLhut}m6tG}?ca7t`Rs1xu5tuCYf}gd9o&*py}$0l z$i4X#_0HQ*JDDFWC>lOwUtCXefxnvJ%xE4u2XEJ!MCFcZm`;OUxtm3bJtG9!^M>Ql zxn>^k{UMO1X*X_#N0$ut09*x+q^5oS%+3WsF#lQSOWfLRsPvUR359G7d#J&l8Ne31 zg#oM-IR)#Fnl}N~`>jW_K~;Q_2G=cnqL@S8L7K55~L06ZfSE#BRh|__HDa` z?Bmk2a|wl^9#~@8Z&g~|Mfs7MFDu%Bo${P`3^9F~ce$~(t89asu!+lGeSD69Yvb|7 zT82cwR=WAG`D#z=-%#x|^_S(S+DFu?+FzfJSX|4U!5e+umR)u{ggt)3U(-dAoW)s= zwp+$FWHgQXp&!@9porh-H|p>0cT4z?cD}DjTt)=no2pQ+acdHjA-x@>aT6`y9ZmV@nyPC7B68)FxYl~L4vw_hUf68?=aL;M(QNtmDPgdL2a%Z301cM z8vLpymrq%ES5z+QNdtOatxlv)t~;%LYc4|3*Hr}f)pZm<`g&`HKlb4~h3EUQ48XnS z!*476oDWMG-J?D{QsH|od}AMXhYysvDjz6xHz+`rU1z^YVv>2ceno|m4qt_;j~8zg z;MVz0gl{^Kw_d(4cU1uK+;FJ+U}ACPrME7y^d6RO&frbM!)7x&%_VrdZd^fNL)(;0 zTMd=0bh*MW*BTex5N^5Xj?V55$ffB zf}X&FB>Z&-KMFkMZlwJmZvSsG%xJ>HhU#_o8PJ4OpFt}3*G*83nwHr8EV0$xUh@!LFXglH1KZwVs@9Xb7+N1cTNN$9FnJ`Y?iaO$Tt)B2+g~c~oYD&8m^>or3 z7BM;}KZLPe)|{!^4R^_w9xN=K{-+$%b=CO9Yzs@Yz+oJ$sjlhZ>Kmj@9W?}y_Fib- z7u{?Qp02fTnC^G2-J})oSBmChMP;?0G0tNOw2Tq^Dk5Tc4MB!;05c%8edu)F=EgIVaYCr~s7j7|YiPw1SI-9;2Tr*|!8B{)DiNWo;sbtqa)P z1-6O3-tE7ice*|0bO)py#2?p7SJf??SMEJ~1`-W{J3_}8+Zx6Qpq_5m8?pdbenLgD zX#YE4^5f~x0#x==CBDmkU#Y>?$5T4KFS$A?o!!`m!yR=I?KP{QN!J9weYn1Km}u0k zYqod5V_N0>m_XL>tKlz`X^tyf22s0VD+C9;m zYtx3PQwZD3h_}g5Uqc;^Q)!S(+>PpI159np+mBegGA~G)C}eeh{YDc!6DVm=5ASnX z*$_;cy@#st*s+Gsbp~0i88qRpk0xO%&k;a@D}Wh=3f+ztqPIL%awx&{y1&2p@2(`WNEmJ#vW3 z7^uX@&dGyK>f%P34o4AlkKK!ouXT5AIxPcH<*PWIqFvNM4!7q;ZU!wjfzr4*qdlcj zYy*#-d%wws`@c4NEsH?0q2!^OHzZZ`{!;B3=l4)!uim}NE^&lVEdHclx&zTZ+;#@> zZJY^V-lyy26?4H=s>hA)`CeWC8B{8-05c8eZFWTGOi zejgwdE2mzOTi4pZ-0#Qu`aTn)u``HIgcsZ*&78}BvdZ9|7?h0>lo`$M4zS8#T=z%u_{sVz402^yHKcr(|S?B?FO(wk;PE1j>}2Pl)5 z!A&!`YHj*RNk6z8deyJyCfsl`>A}rEhWgRkb$q9-R1O41Xxc0Ob*$)28;bGZuRl|9 zUXa*L?H-BU`E{Lx$YMTJSw{rlh4#JC?ZfwU!n9O<7|Jon-??fvC-7}cZ+~@B#wH$G;gyEa> z+!(OLzcV>g&7A+JkepxfFMQCQtNoF482tf?MDMUjBdHxz`LRx@fXTsDUxKONbhYK^@m%Xk&5O2Q2aLg4MY12PKb)&L2D^I3oHM_fe--m9fc{y}e5Il&ya=hcYCz z{7K4uQe+YST;W2J%ytotn`}wit}ua6e2&2H3lMoL8Y@gwg^A-2oY7EEi0%FDCv*3) z*)y{f?wQ7ovX?He$=$eH|+|lba7H? zh$(k-gN&I~*`SPLKw!$~^*X7KlaPwld8QTi23j!<$lN)^SXMv#7|ng3(=xG_#7{JL zrovBdriF#%i7i4c65*N~%yqJC;Qr`ERmxB99=N)mn%L>%*3SYu#8Lu4@fG^_P0A52 zQR4bwlS7_6&z5Xd=g)#I{+eX+m_jcoG*%QEl;uIPHYm%@0-7(p0o9v?_yHj%<0Yh~ zXkMzMp(LfEPA4kYw4z z@U*!=95FYzX~MK2Cx8hE1M~Ik0qJ-ZPVRivh>^jc6x*gUs05C2U4xitS<3LQ7`u@oI6{2zSLPiSX|5R_+#G*O0A)vgHTa#%9Y_xrp~PnVV2Eexc|- z$@^Zx_o@1R-#A?b&R7u@QSJt&(DzB_szWyPp&Gp(#145-mb>Cq)}zz2{?ji@fkHJ0 zTh>kiH#2*}&3N(ol0qMryBLno%O_NAB21IfGucrt!&Q;dCPgMUWY+=Qlz0bZ&pOVs zK2KJY$Oq((wIEI$6^f&G71~vy5eD(BAl6|bT;~^%%Fn5|D$-n2Am)EGF{2UtBxQ$Kp&R!& zhJmhs6zDo$IJhp9C4pjnObJpV8l2^u`|`cC)F64uRUTcjJ6-5N+|!*Dgd*<2>%7f1 zVyW8ju4HB82hC``jJvTCRkRMQL&dq#Ff+JMFHU6{^Z$&XRI$fijIg*;xWcm203nfB zz@%3`;#jQDh9^}09*~Lf^#iRfbU*)_Mtlx+Y%;1A#{W(-v9FVv&iH@G^lUOyOJzc6 z%on!sW1wz4rKLeG5=U{lzfBWRR$puTbA!Qk%E@Opcigq=B%$BO?OkA~bli6& zbnmjbmP8k-o*=tI*sQgcyJNDW%rlhNgOX{yWY&sH8f4Hcd$y0ETJBCra&nn+{+CrX zhkm4$|3VmkT$Wo!DAp5Iv{LQ*|A1>CxR~>xh%^srH@2VN+skh(5=b@fS|{gh+xMCd z*(^(|tXb%WkfoR`iExkp>V}-hf7Dj6%-+0MP+|kc} z@Wic-$%*f*1h=fa_v$5X{2m-j0>ViMZzvjL7$9SK#b9O@%>}R9n!HLW(Z>xPe4f$z zSldK`Lrm#A*ej1UF)jQ~4Z&m_P}iE@*m~L8~V}XTgbVqEtdzA?fhWJ*>BS9V82eO@KQXkw} z282HeZMV?}u|H$^Iu5LTtXa^$NVLXEC#>?nE6xLHfscf1cg-KPqR)kB7V)r>|fztLjf0MLj9#a|E*p@N8pj zD;QG)Fx-gqP73Opa@S{>E10djsf%ga8WV8Z0^h} z>+tPBn(mdH;+|FBZ6$&Tb_|A(-swhKi~#HR+{Ej!*y@M?-N zc<7j457(H@=H>1QdiII%Q+)yga{|1ZaRklo#?eiH`f>9;!0OS7?xpy^KB?aQ#*4io z6}XSi^6HZB77$W%KK|eG?P@#5Tzmu}?ukbd?&vc-U>~jd9AqCeZQajB4MnhNVwgUt zWi@&3>`nWaXfqqhZYG}RR;U4054o$dNMb|Yn`BXKT}MMH9wvO>Dz!VhX*PTnQ@Ag= zT*L7b&AInraNpspibFwZn+Jb7(^TC^eM8tt?a)B#LnYs1nw`pmtBp!-Y2%-Xl~_kK3lhBwpFBa^;|RULsqHPRHe%%`QG{V>ep*}<`CF# z^xRIUdb+vRXb<@gg1)XV#iy&}u`+&PwXL*idGHZ&iwVUh8Z`d0oq(<0W)RQand-Q_ z5CHQYnL=7Bin|AuuWvHnUxBJ}{woDCm>^q(4c(Uj=!G88^;Gm?AN`i1?J3XvE=MdX zt>q-o&)UfT8$+nqaY&d#38IO0JfjnQk@+4n#QNi>V-v3+3I}-UulwZ0{=z3(`aq?B zZC@1C_2Sc0`r$NV-F?Gv>;VBFlP-c*d*We^(x)C^GE~zAX*aEkaLWcv69P&aXaAUHoj(AISo*vmfkMde;);l?8Q?X@(K9Diyj zUQF*LxyYn%kdd?&57FFFCvQP(&lL~R+-pwFs zQ{iJ5HN?i&Kt2^J;N?Ic%5#?}6xG0cepkgc+I3KQ3RP3?5}*6V%Pjk01JwNCz5$xn zt{_0u)(y>urs6f;f4a;IVwJo@_W(zMTTzOLr$nNGH$9#Nn~3#{&LuT=mP!DCQQu1d zB@JbIMA@f$;Clq@UoAGX*dXO*BP|}L``4`~93r{pZePk}IOtRp?dM&0r}~B5vz|nLZiiVzUvo*BNFSNUsPKJ0}M+WQ{xDJZ3^D3>~INLbtHqEtHuX$8qq z#;0Z8yIv{lXuNBcs+qV{qFUHfjW&u{=sJNyEpalGZbj=&W?-53 zT}cH;XRT7Z`}*qr6jl|x7YLWtJ@FA?O$zO1!0#*!<;>2fBJ3-69)3JJYr4QoUT8@q z^Dlim**msvVN2hWq~DTdNOUM#zPmS2{ybM_q3TAPs{aM=^zJ_sg9%rjMDW3iRyG$3 zu6YV-si)jcSMVv;|1Uvf6!DZI?o-Y~lR3u_C4I}>yY2g87ej{oUzD5rR14;QZBW@8 zPnI@uRYZC#R@v9+>`zAPxRqa|KGDpg35Z3b5A}LdJ*N5RN=DBXtqZI*mXEdvyB&yYc(o~{Yt?D-f*`EzuW66(aR5BqehQAHInU-dlw(oO;_ zAx!`Kderv_wh*&hgosF8v%Tp$<=LUB){4;X?)T22aK2QDJF^ zw<>>_j%b3r$4cAqw1?vDIm;ld%ch4LdbEe-G^V(|V{MM(BZ>FeRh|?|s7CKQlX!s_ zO@46ri+;wuxH{9k zVRGppr^bEC(72D78ZjF8GmLUA<&vmIk;F_CLNgP`IFeM7TjUZd-+;5?hSuL99B)y=j54=ZV0n~eHoH5xvR<9|SX&gfT%8!Co%U=_>S9=5 zTKh$~N6cC1gt^h6x@vpg#=|VX z@1U%502xk*2OxG>K?B_1rIw6S{&5ErDpVZgb%LHvz&2 zlN93x<}{k=MGh&hqv&2d!A|Lg^c1KPI;Am2DA4P=INR)rAQFh)xLM4EA+6X^(X z8}{sYL1|YgJueE^VTnpeq6MGPnrO%UTSEX15s9+MDALMR4U_MlBBlWSa%#?GO^8xc zdjP&!Q9DjDBCbkEB&KbYMF>_LMaxs>UK-{uuf|)=kqT?l_Ya{i@G~Zg|vBaL=3Q1SSy=ctqqMhiQ}~ z@ijzxKlV9(f0Eeb;~{+XP7*vl>eg5CUIc`>a{rv$vHcq?aCqTNQ-4IwO^E1@2plg| zQYQF$oD&@3Kd>$IXG^LW`{f+k@;FO5Xp*++{o#XV*o4nF4-!Rm$lORs>0-3c$-=`Ns4 zA7P*ZrAUYYgkanRP#ywWNhC7+Th>CDt~Ae6O7?I%*XLN)oudnj@edM!&Zr7p-o+Rb zapnxQ43t-YjJY9;ZotI-Bk-tQQghOxoHR z4~cDJMygFA+Zmae*WJ6)+%Qsx8%DS_13;cd_&izmX<)a*bz>yvFDbBeJi}W&Fs$YK zSXuu7O_LVSaHAblLUFnq$pRBHqm6K%vJJ2#%(g7BVLSAsQvT$or3xzNbkf{G3 zC%0G?#qZ^+(<(DCMuGZg9=?+-^o1j1ji&`oWUTkQU`0uf#rQvP1jnM;;ZEwAhaJ|r zKiu+pXXEWrzAeU^ER*k+;=0@+9QKQ&e4V(Ez=`{^qi~*QCnI88%hJ58&gpK6 zny*NEcbys>m1-H~%>3K_>a*BT^cqjxY6#oJEYaeKh}9JARM9oC$e9znhat~9vn)hnnyZO^_;#k;o!ng+*MB$&^!v^McqbMVA zhbZbd;}&+9Ak2M&#+`q#9bnyBh=wT&v<~3VoP}SMRq)bFI?J>b*Grc7He8 zH3s+|0R&E&kdx_+<=0v0U&lyQyegg#HNP`9&6omYQc1HBJDv$q$WG%80by~2)#QL4 z^^?3Y)B*tq9IsNLF}tZkpL7*3U{nHkJU{E9(D_kONzTP-R?F+>Wh*IvBgb{%?2|`6BqfmM>-TMTGGHz$AV|b zfSKS_Txf#y6F$d?F0_-bf?Au}`3Am!HqpzMPPR|oZfwVqCLt+y1gDsf6rOwDPeLmW ze<5jS1gCusX-Q4Lm{4l$SskDg9v%VVIrv_Z+6~q-NtIt$d7d*yRwG#J!vR64vYW6| zFsKm6 zA=k@EtHybG;r`goa{P+!rJwQ8bhw0VCljzK1Puw0sg(1~4k~9c=8&TS7ZVW3W#rgh zAnbjjHVXmE9FmP;_dy6Skio_yAVKyW;M@v$sPJsbTmKEJoME=flVht)*yu@}=XZAB=_+x;shcd2o`F0#NG0Js=MPEsI(7Lk9lHvA7j8~ehFy<+Ot>C_0x6!zKyQ)dvRDjWw_m*wL^WB zA@`3iSB~JOL^yylCJ759^^DjtOqcz$Jv&0Wg50DGat1fJ=5&|aif>S*k?nayy+qST z4(BWhy&|7u_;hE*&f&B1_VnV;Z+42&`1R+PVOM zjPG!tu5>;^c+bEZ!79>8087n;ll8KKodX5&WW+zKKSL1f36O(#L8{?otw9~-c~*f)f^2L``ZH#ELzeyJ*sk{v6w3 z$zvquTUFjsRNJ_+N#a1qY(k`)TgS16CdzY4x3oJAqrK_>1bD~X_Ym< zp^VkEcU`2VBB`;1m9kMbz)nspe1&GkP|XV2;82f&6;X~Wz$}@B%P70G@cSvoM_?SxPd_=)q$H5iKKAj}`%uco9$Z zO({~mK9=6aXv8mgSw`p|e`ei;(J*TBVmrUEb+xRjN;8IOzSL>DGtFB}1H2+dmn6eY z>8kiuA?j|%U0}=ScnD62$%WKi7b^??IAd9oA~l})=M`2ch19#Pq`L4$cCtVrlBp2( zEbek5jd8}riufoa_`ol)M@k}w6aK;k)|>26raBFv#$I+o1n@iPzKm_%+8wV}gOI{_ zCK~@C+keu&M96cKCT3zEN$h!uLH}b!R7&H&GRlhTU82)upYh&rSUt3iRwu3>Nk#+r zIC^GnOUpR~u423*gx!FCXi(rhqM3BXf2X*v$juRBkO}%O z=(YG9Ly=Vq28J-6muEeI%)otsLrLHjvB#c(D*>Qjb{Ou zPHlW+NIlj;S;wJP&O&>IzLL!+bc-0@4ZA5A$t{o=0|0p*%yaRm3h4+fk_;MC0);kp zaQ&0gc>B;YDH~97{NXII-%{;uenT5TcZ?3>R5a;r;VLEe{j)E2$ce-vx2~~K>6OvP zDwI*I`GlzPQC=kw>}B^rAUGePGOuNz*zRaPp+m&@4iQY@rzy4ZW}MwXkF6NuzQe%~ zeAP#g2Uv_saYf=RGc2X!;qM(>aF;l|EKzz>maY`a81vJ6A3$fxJtEoIq5?6-Is~L@ z`Mo77nmHm5IMr{OvGTc_4S4scKE`{Jd{nYAQ3Ya*Oa|zFzY4%Xy<7r!Tt+BRwfuvR zl4|h=-A=h+#*CFfzENvTGvu_?gxF(#9}bAX?(be-|8LL4jbSzCPN)n7Y)F;CYsl{J zI9C+LkT6vkvip0?dc>m^nASk`rzYdcT46k&ZrclI&)()@laEfps?fz}=M_u)!4Z9Z zCxW3e39iF3#42;XhIrhzYk!_)QZ+eW#Hd;jgyl?JA3T$PJNaevg(Uuhm+-t)8@?*? z@;O>^sg}Fz zK}8?t?kJA0BF77F*~AdoUJZ0+9~8s98yH!;!1R-iFY97+6KYrCe04!k>AQT6Qpi5A z4{JtawGYc67%F6pULC{rkHZO#ZLe(7sja0k>Sk8Vk~Pmfi!HmBS+Z?N#GPfic+yKW z0Wja^$oqApHs0A2hlk?juGpk(K@o?^ICW`()C5l7?#wgTvx}puQ75BHBZ=yu_ICOWYV)Fn>)+M4k=nkO1N1zjxiB<<4crGk%wG= zIs|~S9z*X~ZPb)~;ij>7;E2XyBe7)_b)(xcJ{L1{=1Iupn?j>3OGA+Lfx;A|$bgx} zc)mdMMhKCHj#XIyWNBf^tb7PTr|{9UO&Uv;_pd>=sw;G28+t^L?z7m!h+4J}ORQCF zf6QlK)d?&Kwe15&_FNo?>P}byFdA1IId($slW4*q=!HNMa#v%Zf)^k%f;(cjF=Dnv zq$2{cF>B$|BaBsVMggDhA;ZMt*!hh~b)AA@8d0n9k!g)8wP}Iu0F~3BIHS0P)WP6# z{x_>A+ps%O5+{s{wUi0}f@cB6-bB@f8^1e?+#T^FcyTwB;3x(kcqi{lSs2a?qm3^! zl^*)b!}u66QVO1WK*QA0EPuD^?n!|dqcoIg2EIg*+5U+1e$0_)P}p85{KI>%a0v4= z8Vjvtvc3+*8Pz4#Iis>XL1I%FwOZ+f(VqhV^B&^g&)%Cjj{3$b%pP$r)_S4q ze{dqrU3otflvG8(o?*Ga&|UdchOdZFDu2MR90lmEoW<}_6+e{W{RnG49tfi!(*#xN z)?rVoYuGEHl>ItPjR8~7))XI#gN~4#<3wE;0Z%j7@C;IAa~PWLcS-i_jbcn_FN|I# z^-^8ro9b3KSvb+U#bQjTC6sCsKR^i^Jas3Yr~Nyu*I+D-@g7X$v+P6gKP`ookas}J zzR_!XhxJL0D7bB_R$@Oj+K9IK42(m-jeuta{e;^#f?+OxLwF>^{ltIddCKGISvjd0 zQ9Q;E?GRy7tp(Nu9)-4m&mx?yh8tHHlJZ01O!5r}EQFNTFPQTAOy*7~|C>-=d{;53 zI#i+tC=bbM|3i{hk3d%Ug=DoD2W2p;E%z57q>N_BW<|baGjFHPW+TQlJnP-f6!Ce+ z1VS?ZPm!6YCY1_`c+q>(i>nhaHbwR1wmtGCMDYkdX%QqMso(cCH+|n2km=VOo$^kk z95W&rPI8NVTrBD%`|6g-uvJRkpB6dTk%QCSGZx=aK~pu*qf1!JSQy!>7cer4&G%?D zfo$C3ApYuvv9AWXh8p?xF5>}!eetcZ95YP&18IZ1Z6B|IDojUa?=bKV0%I5$?{gF- zn_MKZ)ffSk;zXuD2-yeY=6rsd@x=;OFy~{Vz`Ei z|7$tI6%f{fxtajMj1s{NYhhpAnnHP-?3$H_PPZ57Oi|m%je-PCal5Rko$k7OkOS+0 zj-V2CKu5e#o)BfSsso%T&jiJ&dvHZ#?2JexNyr9~lyM7Q`zvcjjE5ws&N`rS)uqQr zlBcVOiMt6fB<~Y(FVCR3F$2@91+}*B1LT-D=^~qv)CQ7wRy+s+h zF!7i$UP32~CFhlDy{lDgFAR*l4viJFFb41$tOl(BKJbe0T=&}&N?Tq)108k)LXF0N z5Dn_yD2x-emFuNtZ2G;^nA77q_=j;Y%LoBH<6R4t@_oU=47CRzxMIqp-MG&(P(Ty) z@i`l+g%L}8n7v1cEuF3lXgx+rXa>l`OM4j00PSHg1F(l{iT-jK0F+WL6B>qbqwHb! z7k+!#KsdMUKA5tR*F^6=7woEj*F;wfCzeGMxm$?M7@}R(y(AVL=g)M}{K^1}<_88)G(A8R zD59&{f3H@w{03p%sG^AkH`-NIBb?hd4rt7x$zWI)O<#s}(R5>|4meexW3zUxD42&#z{kS@Rly7Zpe~pR0-c3_t!p={ zGFvZYb4h~G;Zk9U?AcAt238e8mWM+a+_tZmpzKZrYU)Y#E%oosK>fRt(h4oS{uHbf z-is1CVYq(@Ho-40slw|9!bo5TK7$G`gWmB`)#kA8(goKZeI+f_iY7}BgR&6Nswh9q ze=MO;hc@yH#Hb+9O!(5bz z@I;2cM%d@D1Dmd9pRX6Be+rp2F&!e8$4AraWq`bMlR7tqSdsXE0y(>?~!! z_iq#4roE({TPt*WkWbkH^Vz5KnQ%JTewS%Eo~{&{PeyOZFK?^l(Ma-urJO|n1bHl2 zLq54+7Au3Fo%O zg9zPFvBfexOByG(G7L{q;lCFm{FVxzXZUp$KFshS72e8lUlm@-@Z&1HmSKIy(jta= z#uD;_Jj&Uom8Qs5UVh6>)EJx+DL+H4WwuNk17 zXTL-dej@Ub%WUEpaD$?I7~o|##Skzrv$?te;epZwu-O@ZPFd(p#D|mlIZ}|!7f%N3 z$7HV}b904IGBX~Be$2}g63dw+ZjvsQZ4&bEvh6d90amatF)$v1-UxU`Dze%$JXnPv zW|%YIKF6W@e!XaqY^9QTvDR(0YV-oXR?Q+DMt$c4(@<3=XBkdY;o{7pFf!1#d!VM( z^t<{^{{xno>5+U6XMmco%WGEaIO3f;Rb1>_xe4HGoaTeK!LXw~Ya`XL9QKMHwN)qorT(`P|^Xp6d*>%LX0bqS+zFbIASOsW**Mm5k1{}f0$+@fI=*so444k z4Z9fYVIh`c6l((Q^J?D1d?VT{)7 zjOu((7%%#n4|6cKp9FHJXXB8>!H7PB_NXe)S{H+ujHX#G1 z-5g~29Pig+?K&C%y6bBLG)^uhPq)hH!YWS+LzpX$#)~Lxt&dq?Z1du8zi$3)6i{h7 zuaqFF>+fR2v25TZ0#*L(921M?`}vPezWYn)gpqT^%J(_rvonhCF35o7w&yd5$_zH^ zQZbndCsAsn7Jmx*P(H`3n*K)ZU5ZvuAB^bgA~{`+?tZiy6KV->-!>U2`qExP+)}Qj zRHH}saeu;JAhir3^|YpjHLRHp=&&|j#W`g2gt4KO)w$yqUtOrL-WA3ao%<8(g{d9= zbakK3UWwb5f;_|q8#6!~jAH;c_&FBsX{)FPl;VF{fc6l9DI0uZw%-O{C7j!KWLCfi zcQCANa23Pa20vj~jU(80Aso(ZJ&FL#9%vcN>=U{qU1h_VWykQvb;9h=eKoV=*k5tm zdLj>DHkAP~+k^o&9BLzAHXO<`%!Y%{@#+13Ayh$DB7~&-ZX<-wv;0DMg>Y`$j+p^w z*D$OcP8WRjIXb8`2M{05cmil3y#!o+^8J~o%5$NcL zJcRKs43O~_44{;E)bKNy1eC&H&wFpf;43rz3{EDT+jeCJn9^1FX@+$vf6K5g2mX?uy+;|wYzjfrIa)%hqI7vS$g5%J`+R#{#vtyh!rS6}OD6Kh|4v=6ywP!zt zst}S(GnUK%7jD~=AS+z%kt_|!che8n8t$C&D;+H2qme~0W~t$B=8D{3?4#v=H+pW! z!~2nz`_Bwe?z8-DAl^%ck?5^+@lGHi4UML<^CeWTJBFUtmWQ@ zVP##aj%rPgT^vAdD(B4l<72D(!_& zB3G5kX$hS$Y$2K4uQSw04xp!JB$ zNC~3e_^3*#Ann=R%z9(#bP5x-k4Df&A@Hn|4ouw^)2SB!J`&%riYh#g#9z;fMEZn3 z2?qVh{fQF1#NcNVGzWU~jd_~yA4-Ms+XWa_uNHpq?@bj#ZTlVjB<{M4K13<%k@)u* zej0uW@nae0NIb%W7;dY=JsEC-u+MSfKEIAVgFFT6n9AXz!3IZ`HV)4I5To=MiK<(g zNUr@O$bHig<{{R)rn^yz=Bv%oVVYwM?Zz}*DCnPn3b(EWz)R8eJs50V@y{;_U{3 zpAhv^bz0^$ztznpoZEKu17OnDWiYJVc^1Cn{!jTIj(;tvZ!1~^<@Z}HX?lJ^srG$d z2~EpsfCSTG&%VYsA^p}LABf-j7#W94HWf@#f>8&8C0p;Pl58Jj5sX?DSjWNXP+<-NusW93*}62kd7`-7Wz`^4a|e+vZAdky3>*s(N+HobI2c&@if zd#@MKVbVobse+}*Dw?{TOWQV!4EH})6_U$?%?}#dg!*_(y z#CrRy6+%<3r*t_SV)w&sTk@VRhmROwIZR*x^+#zEq53S@Kq;F1N%>Ii(enQ4J-@sU z({;IRO@YR&KWZ_o>yLXF*5y!?VU~k-l$9kPucw_jN>wPAkwz8D6$xQ~oMzmtQW_st zjGw^r3l_?cdiF&_}Z- z9cL5eGRiUYRr&JQcjpvgaYF1?Z%HV%P+CDiZ597po)wTM_}DN!A`+`lzq<>_>5Yr% z=y-`7Amdb0?OU;(~L)(bdq}I2P%}-qG`&(gBIFf(bHQze>=Q;OUL%_Z%P^M8D5I zikJ_sp=C~FWW_s+BV2hJUxNU?sTO<#i4W!(;4 z1h$jB2T9)j@t@*4yD?e{U5ps>mKdLap8yfYv=5s3FRjLfanodJkFlbNa*`Gkh9nX) zZ;6v^RiY6D01B3|o1;s4Asqoq!2+VT_^*GPk~kTy77$`NST`W|?)EwMBC3ni^`|bB z7$R^dJf!lNT%39ETA>AuPzssTGRoB!=Pe{6U;5M>e8L4VrI9VRQ`d>bauqBrY-QbA zTLR33RpVW0kv=6<)WKdUo=*Txh6ph|79Ww0trhvky26TW%hp^qS9@a;d8h3t9KxlNbZ4RR)q7YoQ@agsAtmX4h?)!yRFtnNH|j7*H;3Ph?+6>qsVY4gpb= zPjgeg>+kmKNCQpzr-+&y(BHpG1A#l?yMX>4hBG%WUB?~b?K&{vmy1orHFY77Z(w=YBSi@=xdw}=)Sb_IPdQb9v zksDmJ^jx|A^b34t?l^ByLZM7#sXktmgW8to{T@|;NDvvi6K0=*@WHl6C zV>`X*16RpjE9=d?-GRTy%gI#NXR)h7Y+(%WYXlDjBl!JF7y^tzI4URv)l3AB0D;dj z^)A*blkjhnSBuz!^0#MWCIR139W5n7(tnE1my2wg`lO0;AtT6IaWg3G5(8xlfQPv7Q|mhJt6sN zSAe(b;fYi`;9(^4IeH7$?(*MR{!=A~eB=95unsKy{TpnsqQNoBgON4Xzo$2%^5lxa zEtSnm>U7>;a0Cr+5YF6++UUb5?6&dJRtDqPXS@(*2VeqC7*&g`7*9G5l_6h{)~Fja z${?jt=_W|J&yg<8NCQ*Y*~gw0n+$8oBY__iW=G&upbE&5YczR~NlTpBJxJ1`X){UE zqFf&aJ~8Dq<0mM#2yQ})2C}G@cA1{vAlGJbpxWKw#z(Jln)1H8qk?)|yje$1}!l+i0K} z23g~z7IvnxWa9mLm zF3o7fU})LsyV?i6MR835DE9xz|6BN%p|!tn5-h0FQ~LCDHnJ^>%32s8w(LhZ~#R^xtU zR;WWX9T-Q4)Sp?pP(&TXNXEm$L=iB7^Kb4nH4`~d6NB@hW>SH?GklJ}O0iU~;a}ii zx&azd9>J|isQltABHm}FYkj+PZ%VoaNGHv($~v8;QBzwodJ5?z#h=$zj+PY4111Wl zoGRbIVQF7FSvkI zR}{{iVi>3JdU26_Olr=6xDGf1u&Ug4i;9HTqDum}wXN8yelY`%rC%J<0?`1RGkGE1 zVIDzdNr>aIb4J7nG>#kro!!BM+UV>ljMvHN=HT+2Ggm3cb4DjNdxAAm9Y+%lMkM7| zNEw_W{kYFjt%RK61^;g)6o{e&#qfZL@!PQw5uD7wBvd870117LUBy}ATktQYX^{eu zM+K0pb8LnPa+vUK(<$8T1ot4p-Ty!0zWYDozBGn(mlfP4|0C|Tfcba+6x~w<_v$++ zlV$k-cNis;-v1HzgM#}h!QJ&g;x7CjaUU5?nfwz)x^LkBztFt^F#n!T$>a^eJy&qg z_>Z{T{YTvQ3GOz6JNZB2zVya_qk9|Dq2E@FGC7NXlUJLu2OGd?VixsQ)%3dpvIdF8 zS>7+5W!?o)nISY3x}(Z6?r-@BzWZ=;nT}0lY#GGT?$G|bhKp~R`Y>FQakB;D@q(%& z!nmMcxK+_OAumbO>UvNOu9L1Cw#ng3!0I}PN&htp90SF0LVc=Hbmz+q{W6U&Z{tN= z@5{iYPCJqKVTkv1l?KnzqMEf-$&ReGU)3qO8yLSvY(UCNuh)&BDV&I8cV=1y zT6T>u+OxJJAx_g^br)lFVRoqSQ1ppJq{UiY$W8jCCd)CJFQMg#6>G%9vh0Rjq`J;$ zx5u3{TcA!$Hz~CA6}hJFzm z%kxF@^IcT=2_J=Z8lm=TGoATSokSQqfmmiBWFM3yXkV=;S%8#v6YG!HSdRA;fxOss z8i%BwUR37LQsGR8QJg*b^M%F#;U7KuG~=&-=7U=Bh#K6U#S5O&j2<`VM|gN4lAqhw znlj-eQNIYywXHN)(J#Vd8U1pYqKwopBBAR^kmdz#NP^`T#P~BitY0Lv-FOjAUs3`} z{{^n`{ChssH9x(kUsV?&iT8q(r5ihqOjUgy0G_v)OF_c;`3?%>SRpNjtH8=pKhh_X z=m5YpqtgvYqJ^e(IJK#XmAHEPMR>SZzeF=xIsGD%D25l$0$>3n#=}2iSc%s0(D22a zM8q-S%vxNz)!z>~Dm&uZ84ltAVJ644!aJh~Gx>P}uN zOH7(k%?m+6U>pKSGfwnhgtx5NL~EBd<8I;eF=nTLE}ciOY$xV_A|Q5biaf3QqU`sX zeLxPXn!k>I#=5YZ_P{(2Krb}2b+&=*a`1v9ukW*LA~=QK4oIn{-0X%N8=e9Djj zlVXrSZ&%kKsxvA@?zY|fQT2QMqYU(>=$Fq_z0qLK>I-KXuc8FeY!E`?IOH9SpB+(R z_s+roC*Qe<$$B7eC+rI&d6AqYcWs9b7*~fUBgWCJ2OkO7y*y?IrhlfWl%0f+5Quu? z%RmoDV;U`(GJxrb#AGSW)xL$2ROCD+#Go?79g1`xQse%Qzo~4toRfTL4Uut&R-nruP>cdBx1mkBU$n(2lIPM+q!b(qibXzUFJ9zBh~dRk1OmNIoLABG%&W0BvZSz z`vdK)ABZ+E&~Xw};*P1077EP<6fHpW{UY277d4Xs|!(vNojEL#RctsO@t|C_a6H7@>qG6ZOFbY5+$z<0f%#HXozC)N!XXr+cC? zl+39Erlp;ac+w;e*U;N6>WwgCaxsh8WQ{N z7d@`B3jsVV-v!uedl~xS%&~~ck15kVFm5dmmp{?d5 zL&2v;&H&&ApsEe=-#7%1-kV8H5LG)YVvO5%l&c&p791lcvHF9B;bmo4 z+nuRcT?hLsh=MYJp-zou2Do$xy&Gn(OK{;IQi88Ox{BiCbXJiMlg$#m4AF`%y**oW zsUymQsYFoJrEzCKbAc*!tP8NX6CUI)4wlcGX#4vdKi&XyG*paBwI?h;y+{(mkf;(H z`Le&nJ_0P(zu`ZGg|jp>cS6G)Dxt8DJT07sJ+p)bHwX%1VT+hfK~_*fu`rYyGDzn} z-+T~Ijx1=6!8yc$R(pjtR~bad)Kf{-ZMa%6$j{Z*>*Q)9{!ItgkSOEW5_IKx~We1aJtct1SjUMk}%p&54z z%Qzz$#{PLH$+Vc6X%}RQ`7hIfHoQn-29wLJ!tk+3=yQz5#0~C0=k$$Un{)A`jC!US zB}r}&s9u+N?@KHMIgKUR0=Uuy{O-5g02%}RfW;)=LIRFvZh0wlE1A0Yv zz+Dk&ZuPqW+_c*nf(wDq{%S5{b`R9g=%_-EUKh|=0L{G%95-Lje_S=BZW%g?Yc4Wy za5@rpLc#`{yvK9lFe1D^<`(8-zrOVbM_Zl1>w(-Nj9r_H?FhK;F4xA#(%hna%+D>x z$E@5aK0e63gOAC%#rbgMmf+*fT&{1Cje@zQ7#f6{KKD+BxV=?Q9F<$aa!WI=Bjd_2 z)P|w53^he)GM0rwxG9Yjc4BqI*)~N{0N^b-3)Xm6TDhlg9fnT zpMXxpWJ&m`B*ZuZ$Aod#0n-uA@YP2zkMeCQ5LG3&{B(PYdo*&}9%CFw0xx{KY)t_zv$VL>HNR42xMG=d|6PC$ycWHfyG zT57n9Z9l`KUx?N?$`?|Q_Y(+2VyCf_Q4k(3U=!(QVWX2(@1GZ~@;nE<7vvvWkx#TD zA8jIk-j95sA9?!%kt=(d2;`XmxDCf~f|RJWxn}P_RBl zd_wrinii>ir(Z-W+xS9spFx~TDqk{6q%x(zEEM72Jxju`UnJqb8fQQmm zL+F|XAgBtxrpud(3-TOJv)V`knD>IuF`u}?*g|EOg$|F;@fJU`Mwvw<*%#AQqoK&z zf`p-5M@^x$0|2!L?8|dlAQwy@t@#mLqp=j+3V|!R(vR##3uE+P^#U-Nw}aUvFlR52 zLZIj>J#J_>=1iQ@-Ay{|;NX4}c^Qd&u$*x$m!yv?Gdh;h;Cxz;4k{j=6VxuE2H}Z@ zT>$CCZ3uanONM74AslmHX-Hf+O_V`1*9xCQ@HGZL$V9F2(5>BZuqRX}IPi~pfz2q! zyGxScAkfQd4osz;?)2Ef(q-?;@tN}+$uas&kyiJ;KS|HQDXs3-V5vLKc?_#g?AbHH zA!@G0ZIYQ)5o|`0%I7iZnvuQu`zaT)S3`%o8SI-got~kJ zW}N>Y8_ID=_%99RFTY1i_y0yi8D%KuQeJw<*fWxbi@nG`+PJt)lB$}Fx{K%Mvel1f zk#v}FV*W2_hW9&s!J?K-p0@&Tq-ydX7>l zUbPYQloB0TFH0$n_*m-1dUc3yHT^sTtp!wLoq*!RQiKgRr*_Ak-A-?()c?(|1}jxwAq;)TX0>=?j$^cCcR*cTa_Dy;Y9 z$+Y0$2}I#yKhy{@a#F4LJ@AhvwbaD@4d=b8jK)*7o8W~bT>w$exdMT~tJ7+q_UsSv zU66O|saM+l{;h^M1<_(+U)sOI)(BGur+;Q?$+~4oaT5TjC~ScnUF19#vG7F_&!*w$ zF;MPTr9=;&L>B;!ezbe!j#@NekPIEdP|~hFotJ3q5sDY2Y^K4KUPs?K@%``=@%)b zul0+Fe3^d9WwLqtu=Z9amzv`^EksJ}T7;60|*=w~66|a1MV^MzaCw%mF%)Vyt}p=P;yL9zivSNk6s; zG^ZI~!U168S6nT$wlWfB7!x*vvG+v2*jxQ(Y>ee-9F`kX$8#Q!%4vg!-pbf90Ov@$ zdLF_0VzJa$#!GYW0UjO=uKlCqFYG~b=pqH`*Tsy{XzFcjnha?rMC1Hp&Zx{0yil#4 zY{iqGZ{hhc2x6LRL$E^jV5{Yb6y+!4UnmE$h^_vGYu{vS@ec&Tb5*^Ry4~4E>U4^| zl=k>XUfx|8!WGk4D;Ri`l1UTBiQy-_i{$fE$XdqN(S9E4| zN1JyS9ph5DR}AeRM9SBrxZ#OvG2W_-_UuI%5SHRC`?#(>mIA`N#xM1i6_ktE0Ub3K z>X$`)na&qdFqWvLV8%%lYDSek4(5VR1D_O-Xy>COk{V_bn8eRdDc&S~9p-;jtY9Y! zhXcW;8Pxf~Zw93#PkoQ$9Gs%!L^}AstDw_U0_f6=nt)DieAS-i1+~yz4m0BdMll_gH_e^1IDk#SLQebW>$*k zEK7fU4U;)8Kz2JeBqlS%dv@tbmVW3;A*+B`YM}HzE#bU;j=9IgP({>t17lFSx^!|C zk&dSPi)Z)-k{wdYT$0rd8TcGe02PL@eqy5%QwJC9SvZ=-(x}{y#qDTR!a&KFNsX}# z(Vo>Zs1zi&7k?D7z0=SywzsM>|B@K~UMWHn@pX#IX za3-f?+(RO^lRLE7>adky z4%qV%MCi$-91D`(J;=xBD9_1aHl`X-U2tES>Uh3?5N{a2(JJk6|D~K3MQCJL>F6&n z^kBVp^mwM}L|R%9iZnI^jbgFYP-I@tgMp|m!TF|<5)Lek6T`UGP}WtMc8s2pxR`<3PA zDhF%f07#Va^jIm`7%3+)2;Pc^_OH%0FztvYw%f5RL5lT`?V<=!R`vuyvD4VbQ&bR& zOEW5l!>UTHdna-^@(3%nSMd)^5c!?o0v0LJJH%}SP0Oj-042rvO`bnQT2kI)n35Y_$c9X6tzX;`{@GBcnv)Rh=cYW_O+<*D}`k zkYpJ}a7pb9q@Ed`T1o%NVA|xt#`6`eT0`#utx>~f zQAW+tBG~g=L;S(LI`4pB?b&EmA}hzM4Mn)$ZqW{*B048l3U<&o-3jjx^iy!Q!oN_k z@C~7$e<%gBH!E2MD9A`b?l&Hg+&h`M&tywoiuO(n4d|ln7Y2({kV!o|gfkwc^fqBd zpwF?E>jL$1qIv)l%kMx23M%BjfYgON)&h9-8x2_Qc7QblfKJJ`1^(q608{@Dh|FFc z;9u0Uth`lbep^2+wmx)}$gGhPLckmbfRYf(Y3o}Wh|DVUbY+^u3v9WI%#Pql6D_m) z;jqda26mvlCLhw~0M{7Fq(oGWWEUi{Ty$I{uhqN zjuehtfCXx>zrFw3rZA4Vn(AP@a9n${=6D#|_CCiRuGG_O){+TG9-by41@Vxc>gVBI z0*b&swfP11&&IHx#A{Fkxkhmgnr6@f`hr|q}kpN?wzP6oj0`4rAs?cK*w zjtCh=LW3wqlHuzr#6zy^6R3IQBG{LXUh5&Y#nGOK(LN@AFXfsrx+a&Mq3N*I82Ej2 zG(>SZoUUany(UI>Y-30*Tkh*HO>YAud#P?R5pLU{jjSv(6e~$HPOk?9$`Eac9~lg` zFXdOtfh(UQ@u0RzaI%z1$?)zyWGSdxSJmuEr{5-hShR=o$G;kDAgjq(qAH;2v5DXj zoeEqETbGLyWgGK(xDKq$7bb6jYTcAj%d_eWotQK%WEnZor}5#u@jAlmk*{^7R|D;N zaV6;VY;0V>8eV($YHB{qu{Mduy4x@5M8Xy)ufZBHeBHkWUttA$g}N+Gz9dsa*kH~D z?lIU4Eo&YJM)sj>qBxp~mC0NNnHo7*mMwF|a-URJI#O_zFwhqI6cvrxV#+eSICsI0(xOq&nZLrgtaVnN4 zXT$#i7^{x%VR;bonQf{};{wAnZ~<)A`y35bMp&Sz3*M4Q`P6f5NLC5rU`#)NIt0gPs5QAp`&Xr`OR7=})EX@@u1^_Svf(258MUO4uktWqs$p~X zwRGzsk$DA!R2IHE2xYH)i{jH&%P(KSqu{>IUWmimGa&j7F)UnIfXvp-q%`lYyhwy5ZjMf{|sy!o#2~_J{UXcVeL5$c>^1|03I+eD{>k;u>K?rC91@P zhiIm5N8El$h|N1&B_s|ul^#HrGF62$o&qW~*)dLLaL`*uJ^ z?u3JV{dIbA z=BVieUGpSwA^oU{uv4+%+|k5@^q|d8w1Q2CQpZkz9}t56-(K2JPyAD^F#2gA+Zgya zX&3z8tU)jo-kzOF?uh?-OB6q@1x?nX9lBjaJM0<1XusG;RZoxYuXX+Dp(4i^NF+#( zPZX3Bk3DyMc7HN3BozHh%dj&#H$KP0JtX~O{G0mhAO@;TBM{-KXPQ3xf0jhQ!QDwB zJ1Gt2jO#za)xF>fSh*NH8L-y&;1s$Ccft$R#Fjbfzi%%Y%%CkpA&t>Pgui4^tNBZj z{_QYk48u(H`xSkvircZLhA>ziXElpTzkr*?taQA;Rd|0XK*zy7CLR56N5{*G zj`&bI(h8vCJw?a*>LL$^Nyjsw1Lr^N3KVrM0<)-7NK%kgbrJS%?yrGI0nq0dW&!uE zq^SYC`0c=}2EdyMc%lV-DZnN7-8964azB~u&)v5(#QSs<>rEKc+ixNhcT!)9LOAGI+aLIs4Dwi(4bh6=60S{OrGyTX9!b8g1&#`HZJHGdNkPoyMv|k6aK9($a4co75-)g!9iUK$h7gVd7$S{+q5>tlXqEYK}#0u$E z6zRZ}W=tUlCRh$4tS;#0_JHzY$U`NXNJ>M?P3!NhW_}m+HDJ+03$&2$R2(+%sYfoh z8tr!K7Ba?5=!9;)Bpj|&^rA|^+I|kQU0zNOtw&4wolx8RgvB{uYA1Ao&2>wrAATcZ z-P6NAT=K&X)A$EW6pgJ!A){Hq_A9Hfa>fuvM?)(e++*-!0d#05wDDe{BUBCNzl9b}w7<^HsE>s+(s&<*an9>&Rl!-ld$_x!OVq^2b zOXrtE1gn97TeMnZHn?>Rb-;oOt*l$!9ppo zNYvu#vQ)VvOA0lTJ{c)+{^2*&$ItQ4PFfs$12*I34~j%|!0KMKKl6>gcnQfuDcECz zV5k*zm(d8Ry?JOycGMk+9wva%h~54FJhGAv~RuSZ|QzbG!-kX#dOqtN0h-Ns?O-JxWNm1m;nsYu>e67iZ>$($6|` zBCE%-Rm$;fSh~9{rnHJVv#({`8_beD`%P)QCfOaUz>(`N!u*-yy5i)y<-680??_B% zY;|Vdu)kU!Q${>94qA~`52G?X`<-7g$nfZ62jHu=l%EgW^?S<$s;<#7Wb~$F( z^T#KWFy}6goo$OO93oh*tD3YCd1Y=X&Evk*cFL<1n=DiJA|!J2d`eBrid|$KMan?k z4T`EgNOgo>evQ=LZ9Yd+V3dfZenfKx5q!mmj6%oJ2}H)cTRbo`Y2INnio+^c7Sl#U+zc z#9SGe1;i&EJdR*oMMxam1xA^~Ub`v8zMW6{4&saFQG#+U4>HmH*?GByNVxBAcD-Z+ zsoVssCpVF{57NS9IEBq%s_Qal{`-^4SD%$O$3silfUsRp*!)~BAh{K5s3*gY64VsXX-Djs^8Wr1JcXH=m>J`hYms5t!mUvKpz^$#gsh6++TF>_?DvIQib;987LSt78rA+E2-b6_F!TDl z05g~7Yi3H3nM))QLme0Wigx^(MVJTliwJWIUPMewh{6=+=S-3#);N(l&LBp!Fy6sO zPY((VF>#drXpONyUxw%xLH`_H1bwO%{bNj`(AOpUB%l{(RF!XKX|gPtGJGqAj08gw zzI8^*$^Rcbd22g8`NmQdm*0~|tfC|f^yD++{(tu5TfRbG0Z)#0X>F7;C5hf~gpqJn z6_hO3jFXq7XHLm@uR>AXd#UdPU zZ!xc>^J!j|?%$cY1xX|;8)Wz#ZE_)fQye@0P~s55uh%bv`wP5y-c%meGnG%k$d~u{ z<%W(Cp=L6sDPlw(&mniYfmv$gj_B8V=E*k{bPK$BC&A8_l}!c}33q%#(($1i`yP4~ zODqsbWXlzRjb=`?WQi&y-DF)dC9DMhbD69pJtWF;ggp{>()nZ5l)ZQnP1*f%L2}9d z9dZ#(DO68)OFmNI0|eJ3dC^%;Q@NCpFX{4y$}kqu!hC;bVJgD_NvsaOGJ5M5QHD-< z@${gP2P;E~<)RF;+lw-AO64v(nLw>Y4bb&+h%y}SDatT=o?jW%0;EH83Sc_@X{`TQ z&E^ke|CtPWX$IH)ZZVL>c=0UMMJB%99-?ixGbUWjv*pAQ`Sb2Bj&L>1)zMW~& zmm)X@m;A8l93*(oC-y3l3{DeP&!EPjzgrl6B&>g4MrBafr0t##Sl!+^x0vLxR7XcI z!j5`T@~z&JAkBMDRA(kqdy1IK|2E#F5`9Yo81$MC43e-J5Rjz9;4|PQ)!4)5HK>sX zzuS|OyMyN>1 z+~f1K88#c%xYy@tG;KCc$oSUh*&p#SVs`pG3&%5nPgN4;s+i}J=YoR4zyelSi`Yv* zPJQr>$CG&Y99`tUgZ#I}{~z`$w>oZ&LR#bCLa4LbQEV6xis_Vl4{ANx&~p9>LX%|eG5Pq9SPbAdB)&291nNoN6N5xpT4B^gmDqQ z4T_EJ3cQ{YkbI+9wZ#%`MK?MK-3tocE`_e9MmH-M-4u;3_r7p+@j>Vk6}op7y35Zf z>K+S5*HWWts`j~Q#TYI?WIdcql{q7Gxj~QFhW-b zn~lx5A=n2yStj+UAZBBd(;()FLr7i6LkuM~!xb~_%BN5Uf$N!SVF zD0&~!rD z2-E@{vahAz!&5!VwsWV~MQ$T2!zWxej^JBL<6}byv~#zri#3nsDgk&1jj3WJ~G4T@|xkj2AgSVU=_T*A|hq)3v*K=MY(cwnEFAmugVfpHT*^ zjxQVU@LS_w?3ok54T7<7uj*H%nYt_kH{p?P85IrdO}H z`R32UsfEs@*7QQmGb6mUoiz`u)@+Jv8ww%in|}&3%PN2b1n5XI%i{8r7R=|DFN#1- zqRP1W(|MD-IXU5gi3mOEfMq^Oy^pPoFlus;&Gk1X7o09Ej5R*I?J3K>n@IH zG9YT`1s9se=X_h8-<*s@i9|;m8&Kc8+@p(?^)NTgaAyK33R{TVbM& z_4lF>T?3-L2AdOLYp@CPiSzv5-3f(&#pn3Q8nqAbXop#% z>Q?M$K%xm#<;x)T#dwe-_EaVKn!2y`Ws0OB;7FAezs?mtjL(+^a3qbDPrK#QFVEZacM5s73F%1V zE;SDid(BdNo;D($ciIQ-hsu`*N`}8d{P&1=x@;Hm$fa#OZy&fBQ0`X|blGn4tJ_vJ zwGbW&<28=M_L{bdI{Xc2+w_*eu)`Dk5kN1SmdqRPknen&?*is>qrS zJOYicK+r=F*m|M>$0SB%6xkM0D5M)+0)8q{c{=n|ewOmd^R%o<1IY9A)Xxk&?N!YS zl-!lipe-A+On&q+@9tXXKEL;W5eBlJL)`_XZM;J{%Gs3i7N&5?HzV8Xlj=yQB# zjp`#&(=AbTFFwnHPN*SY#;7lb^AqNv^{~C9NXp7$dGyvP7RaQLPVuQ&Oj2Pfv8Fmj zq}3jZsTA>&qJmD**J=;XDVypoDQ?UbK8#JD1?Wh6S3Vt+PZn*T@d|}fY84*PYS!DP zwx8BvYWo73o6EKl539D%9xB?thhN>c1}UQLPiZ)6dv6`4wjYGgft66(M<7~jJN00j zErSs%UTJ#*>BJOPrqcE{Ud(38Ex1OL(_|{&XqjWt_P5@obR!f{o&rL#ehVWqiYr)) zw%@?rcE(+21rM~nCR1HOO%LM%&;;Nzwf*y1LdGaf#>Vr7McdZ^K|tFN5Z#tDimau; zW6}1P1c7Z(@7rp7EsM5;{-*swT|BIzCZ^s{`jSitl_NaHT`B_7c9kuWvW z0ygO|Eg*{~=(5ek!)gIDG3o$+znEX$w)$DI~= z{?yNMZD0!Z>{a1!@tZ96&=x~7Vh^*RXwv?9UAKQ!%i*^6@VPW{wLP3#DD59>)aMd4 z-V#-}Nq@12bonw!eK8)y$h=?MKg13=Xv2^5AK)59x6!YwyG_~9u{dFw1=fSEbSk7C2bEo5zQY`EJkpfEsqh~ zZH!m;P>poD6RuuRikMZu7$r)3sCkr3q5XpzF3|oNN$FmiCP3R25bU8c?8|T}Sc^Se zy9@;`bX@Sj9)_VdHBLi=lsyawF41VVe-y^CnvB2e6%%_{1q4Elw12)Ox)sM1S<`{X zVh=q8fvpmnSf)L6?EYWvp||KO?cv!|+8#;(AMN2wDaSm|gO4kFcuYV0;~8WRvg*NI z`Aaqj;)#J)Pp!nU)k!suTw$MM<6F``MNhIEsrF z79l!t{(Yg+c))5u4wa&wq}ZWTO#UQ*BdMQ!`b<7qtYct*q;Z$Z#sk*zDCA&T#}plA zZSgqG(`9=Z536;&{GwRLNPfk>#$>UM;u?XM~R=vRK9T+YCpeqJpLM`yZ=1_>ZgEU9iw4w#`*06 zYq5?SWtT)4XO0LSSjSJap)IJ*m32G?T&8urp)gj~WXuLC))s4kAYdH_h;FJvXN&|M zi*+Ok0$a0gx3!3n_EM^}hG`La|Z{k(H%P0I~r1(*+C~3c>D6Uf+oMB0kpi=xJDU5fO1;krz z*0`w19~^ObeI;&!a8iQ$%YEz&W9?fO(8S{}>Nt0YykBn!X9j z0=CpEL5b1=?l?rI@{Ka6@B9|f6&-c7r>6*zO98<5+d*lHwM|ky+fEKeVKnuWag)q&Tco#8@q+h)PjO^kTJ6@y4tGj-*cV z=_C1Mv4^MnB8|J$P&{A{O(8MU9>(Y}YlK!bNSCb_9#(tk^{m*#i~Q=gRLJfy#i`B=z< z3R;FdOc4}wWij3aDPBK}G|R(5#Ps=soPICpnB<{76R@2e8kUFH!PhDeotR%q9=h(8 z^05CgArD^e$KpViSI9$-44*?-NgkxWU#2hgpm@LP5pPt&u`HC zFEU|lW(iAqg)zRB7)zfvVcaM%wll^$i4kFzpL`)$4M4V+m4%AIv$iXNIK@qXPPat zF-A%0cmZH39dqzlV*4inN^Jxe(y{v=XefWuv`IobN_vK*V?S=CCFwYdsWT)UZLrJ_ z#w-?@oFWsX<1|E1e^lf(Nk_qbVCVOCFdaxo24O7{5ly z+L=HdAD+hYP!72c#yCWZz88k?kUZ>XiAiZNje6|Vqx=T_X0ya33ydcj;|qzg=qVG8 zXM{`@GREr?<0tc+^sA7VQ;bm{F@`)FrjgQ$S>`@YGRZ@SF~IRhrz3zobOISe9# z%@#KkdCfK$fmt3p-^ubYhEn~uqMKMAPDnW-c{pXr6#Dn?Flb46sD%tt9!>(I#AYKv zsg2-59yV%dD1XwDIf4h%pZ9KapvM$OP8D3-YCJd8_Jd zv0@$=_|>gU2lB8MbCjMZ$RSe`F!h>(X9GK?&^OdiIQVsa&U=#p7k z9@aqx7V_W_RvENGf3PI(lc2&I^TRHz>w_^J&7j2$YGML)jJ}`cVIrqkM2fx%rty$G zOrA~{DN$3X#|}KgZ_wMAC1x|rRLXqD_*i1(Ei}=%QDD5o7|%cpWORkcS*$hd~>x2Ezqo5~%QoJoH0iFs7Lq z z`ONbJk;nXiz$_2SC^0`!s^7MxqnID09P#{M$Q1gwFIvEuACSSAAAo`RfdH5v5D0mA z%}acgbpOkO5%+u?l81knP_<9%?_nwo$wT=>YP;zoQ$LXj-Fg?~3kQ0u&JP8zfPuZ6 zm=4Vkn3ePif*dkk04?Td%n!{4#ou2N#TX1ytS^B;czzg1PO?Ts^6=B=QXZZI4Ji-3x%-L(HlLUun#-^oLPH+VY{xOl@OrvA*pt@NtD)G^ zdb(3sx4BGI$iCExifc$fE~o@!t~Z0oegyvq6R6`yZkB)soN_8xNWf~C#zPX&m@!g* zf&mBjNlJ_c1tyGHGl)jYbneC1B}Vq6CXC%IXDKf+MuEiG{tpwzO#jED+G{$DG)|O0^X6&Bmr&`QqAT^V3vUB;Vc2qQ>x$gayyoQ zwlW_{KnM9uGeRm_KuSOcz)}L*;jzTlj{v1Mf(r?FHHRf2^%KE*OTP+9z{+B(_Gx_; zCclsb{4$;@H&tY+Co+MV{{@NCCyBhK8KHbVST=Vf(}4t>gAnR}7vzws0<;VXXeuav z@v$gIGms()_y7dL5>QIyx(IT56VNe9!2WqG0pIqyssvQ_zZlK^Qydujv|#ZmGR#Lv z{ul7KBe%w1u`GaNw3>Tzl0qKrAy1;bxKRMm@%X&BLMw`b?pyO9 z3p$O9Zs3E=(?O~PXAf#wELL17NLQ)BVp=NWV{7nm_5I4TKid^MmMH977lqLB*r5XVZ0ovAUnpa4BIVqfZ06te9QjQ9T+Y z{t-9bhBW%}7c<8;Ty==^M_rpC5hq6M`h%DV$Sjl1VNk!9!*lxI5fTPm_732?z&F>i z;OxKh7;a9$i3fZG)Sr~O{omlR+aUc~aeg5#Wsed$Vzcz*AC_BO<9|=wPv?-wqR?e^ z#CW6zi(SdTxd&$YCVU51)9;nMc1UGq!}%aNVB)%J_ziBr1^izwxA@b8$$RI1z7CB; zt+5q+#f_MXG(pL;nr=o#Y2J^U#BGCa7}ym~bvUqtW_6(P_3k#*RXeCJ#B9hlBF-mU z)Cna%Pqch#UfetM68sTtF`o5t4r;$Kabt=2sOrcI8cortF5eVf42hqy1bOfm+$E6= zZ~ka{zrUk%8~On4hHF<&>6z6CoqQjC@5d2C5F4JM&hjnd!>)8Xg*uOFZT(5VkBUl- zp13)2CqDG9pM0!3rIEMnj!F1zB^?`R1UgRdC)kljY4dQ^I7)`voYN0-Zhjj`=!JEt z0`KUIa8MtIvY^0JoJmxLvHGV!qv*Z~#2M&+kT-7p)9HEk5wtl(_qI zND0Z%5b;_Rj2hu?qugkBUFG*C9r=`b7aqBKeqo)a#QFP)!|{e)2+P{zn`(K?Pq>aucUjN*X2p89TC?=S4VPO2Tj}T^ zwY#OIwYgyCs9-6entb2w%6n$OD_P~XJ4?{zv|gc_(A>sW^Lwp60080@BtSHX+8UQ;`frFVMFj{KuOzQGch zi*)SQJ_KE6<&!kD!@qrZ;2tUfyzGiZNTBwdikpLW2V%uz;4O)z$+%C3{xE7;n82R! zU3&&M-zhiSZNa6QZ#j6#+)`JBWr| zVhr(`Fm4hUF2*=Ei|e7kl^@1P$>pTaImx7dJ(>X=_)W-hs71kNrQmABzKje_Y7GLo!^1@Jjm`@6AQiT(;;y zr_Jxdr3u+@YeJ=F#!bp0>m1u}XM#at|MPHMB=$oh#uGEh!JvChppGffr_qahV5x%r zSET5T9-uc!-hV-?t%t!Ng!PZanE8+i{Xv( zLx+?_f9u;p%aD$_f?}jBMhlQ)`z;8Br6ZTfEfVDPIkZ2-Cx=)EO<@AI-Zx!UBC70X zxQLBuSWUU@V<{2mvoWU$f7zF~`-%g-vOrE@hWD*E6PK5x61Nx;Z`m& zTHC?aF!CR-2BwAm4=)bQ7Q0}&3&z}M1}$dL?IuviFNM8oBB!K^6x}|P-XP_ny;w)5 z!5{?ths2mT+k~O9Or>mKjEnbk8T8K{G+`VUGIfSAzL6L=m?b7%U^HjGSR*lV=7njb z4CSPMags?M#`FV@KiY!;d_Nb0XUM|_`AqV#fCN&ry^O#t4|k=rJiI}ve%sk5EDt?o zK9Yyt@|on}Nm|+qc^D0_l!sn;EU^tIK&g%3LLPpdj%M&DZ96ICVPc<Px!I(M{V}yswATibo>(#rAadIk`L4SXi z3FA6}afmUtNQ`=BiSY=jtHXS;LSjsx6Q+^UkCXhIWReG0DscSKcOwAbzXj1WH?nio7Oy zP$S>4=*90c9mqp6VJ#Kpkg10lFL`*sQ&1c)ixKQE+|yRSI#{8K_{dQXO5Agm$}UQ#r!BoDVGR+fiv zpaQc5CQTQ1S+qeLFfIH)@Zyj>+=N8M;QN;#c!oS2k(Ai=Zgq{^F;`RJnX)g-<9+Fh$WG-Z96z508_*#Z%5LS_g8KjC_Ngn!ksw@v%R4y^N_z;<%dHwJ-7!LYBUL2B#SCI(+4>M>w zgW8!u9q*XU@^CY!I7EuRmo3~<9zL4I+}`Pfo%fkAVg$xk#+W2A5=;`463Hx+$w?-8sNEG*;QxUD@-PVP zV93Lh@|ome1c@xp7a=gqgK{&=!yHP5{|A4zH>sw~NAgfdK9fB3MGHuIXb-TIhuV0= z`62?~d=Ua653h|wQ~HzgMhJPB*ft~&|K>we@c+P67?OwbG-|u)BGU+w3Eesm@};*E zc}?;#YBU(wyNsF-o!@aErZ;_pAcssBK+BMa=7OR_7USTJLLROIf$;p0NaVT-a(YwH zG0DSsW2qjt1F_dC4<}h8B6;{}rId%~K*R9=;O;99>{=`2;iL?0>kN6o%M#D5viNY( zjm-_F(CFcM19GPUu^AB6fSf)pb9G?|F7wEtE<)~>ro!W5FGL6i9b;#o?9h;f0g*SD)C#Z#GU3i?bZ~44e5rID~%={9{vvfdlVaM zs&5E3@etX;+t;mbX@KVgV(Xgr?dpvtz(Ktx=^MCPic}3ulKzyyj{0EM;yal$K5HIs ztSa|XR2OHHaVNz9;v#KhU|s7ACfhpuM^kw#Fd1I0ymRJUrMxo+Rw?h_Zx`jg9djzm z8yPBZGMBf$C~t#nDep6PiSpim<7MRqKkR50t}^jQQI+_rD)A*%;nJJtdcv5E;Lwc`TO2zLMf*5*fDA2{v#?+Gg%&d&;E0| zV@WQ^mWelY1S}ZSmf&=4+Zfxny-B#Kj4~S+2%x-YL5z-S6Mw%ZQN#Q_d?;AE*$==0PvZtGJFZxNlFAh` zarj%Z5C_+aKF{8-)a~cc?F;KudLk<>=;4&*&o@TBoT4;+O{Jl`o@k(Xvg9|ZUlISC z5Q?eue#VtM6$`)&e%mKh4%t}UNZ{L>5Q zc@>_iv1t8B;8I&p1oVf;R6LT`0UlGr&>rE(KHhn* zIs7+# zPw2N4;~~=@ekJ+^lD(HWYQORQ|S-{yQ`+;{JSte&Gmi zpCXxuuBN(D`^fQ&>HqUe^f$dM+GmKMzY^Wd?K4);ANMEpt)}uPU5WlAN&gJ107)4J zeT$%P{S*3!W5WGU^6*3xj&prB%Ct)DRe_tcX#UO-@7srN;rc8b&h7PtiQbjkE73&% z?kmxMR??3Z^dB(jCkXn9e?q^mDV^$5B;;T8e}gOazkJ_2tbZicFy2=_Ft&fV@*EKX>p^-;du;QetuBDXAH{f zkLiI-qWJnFiPfp}(Ed82;8^l5SY8|cM?Fy&(s+J=HV0qKU`%JOMpB`Ta-P;>1)+#) z6eVR(?+rpOf6^l!n$2SlH0C#uQ(|lJ2K`MkHzVt72$^a*6W((xx%wVKRZjATT1x9g zTCjf#Fe+SFw{o+M3i(=kqM;v+HC!yQ5=ucA;}i9PTLxYV8X-021jG&f{eBZUYR_?1 z;4uBEO?@^WhiN2y`gE1MqLem^mdFi0B?2YSS0eBs2$Vf#lz)|pPBE$I6>h08L79Wb*Bwe-tXkSH zf%K0o{Xu-t1Ye2VyqvlulhhT`;!}!mN`{s-wGbwkLZxmb{I*3YhnlF5E++nSnjq9( ziB;k)tHkS8iCe40&nn?|AbN+Z#CMqE&S7AR`QNXy;677-vpRFYSZJR%XZxTwNp8UD zl2^T)O69S|Pke5Jn1t?8`77}Q(0QDB2e-UVu0Cg9j*9nv^KsO7}hXm>(K&1!X zL!DDTBTx$gdaEbCbL&spL7>_KWEU9^5a|4AuJ*;F)L1EH1Ue=_Z3L)-K)VF!TLH2? z3{W9Jqro(dpy%9?SdPq(fx2@{uyvb3#6}wf3Ae-77vU{393aE~G8`tup)$NphPTRa zhztkIaF7f$W$2QjLEnHI_y!$={&0yu5}~yJ694o)BKc;hCwpi|LY&L{!ePN4k)bm=hn=63|zBtR7cbb>%12vE5Iog>gI0`$nY%s)}o z!SZ_~QJn&xk2Uhq#>879cEvkX&Y*h_{zW!OW8$ujIN!)`LXS%yh6H0T>}1K*%y z(C;nrQxTf@Cj#~4@#nF*f`2AH&i(SsLCHTOMEn%TQ&s>=?CD6IY7;BdR1qk%L0DU1q z0<>O$ni1$(0a_zK?Ff`FK&u6)8-cO~XuvnjKWPNIOMp6cWB$37KrR6)5E<_zP_h8c z5unKgY9m0)4lw`BCQy9=8Y8gs31k6?_~%)u9fE(h0z&+=EJN^59KPH`<4+$M_LN}{ z8LBcIAjAGLOqXGr4ExEjuMC|sOqHQQ-+&wV1|5TbhQz-Gp|Stu@=tTHPbaQ3ucVXV z8?7#_GcMRx@*X2gLHNUc1JbE_p%3@X>VprlT0LEy%Ic`N0{ZSMlCJhyRS%FqYa~=6 zd<~?~y`!vY_zSMDtyvDrk#tIjh|ot|=@aNUWCcBC?*OP7uRvd+*2SlCOdr8Lfx-Xq zOYcLG82v9)&2*n7jju<(mGbcxx{b!yYS^U#{42nDd@ZNLA2_eY@sv%}xkm-4Sb%mD zXuALv3eXV(eIh`g2+(N)@%a*e(i#D}NT6o~)_Vd}BL|>70eVA#8WKo+qbzBK0JS2} zodRpA03{IU76E!mfKmw5U4Whypn(M9PwU|OnF2JDK=lOHjbF2TOdwE@PT}BtrCnJ* z?kCVm0h%N-&Lz-30U9GfJ_3~p(1QIeAI}o#Jpt+^uwEn3Vt`0KW)C*xqfSpJ%g2KL zLOzn*LOzDb&?UoLWcW83j*{U>8IF+Qa2XDh;ZPagCc|50XwWy{2EIYZpnto>AA`^| zzP|E}n6`&~sG=e}+2+$+}8c(3P0yIH@rW0tg0Np7-4-;sV0F4r$ zM+uZJKqCa`83J_?pyOXL|12d?3jx}i$o#X4K(z&^t;o2ZK<5W=wVMji=L9+?K>hbI z|Li5uE&;kIKtB+u5Fp~8?wJPvJO=7CzV=TO{Nv6M{PVaBeKLGhhR@3I85uq;!+*+f zkqn=b;gd3aLWT=vXwWy{2EIYZp#Pl2e*xkDVthT`hpP8~F}|k1xBTkk>+HT_e0`Q$ z_ww;I_ck;^l|XWQwTq0+2_(nYi+flY+7n2QuWbaV zJAvf*`mF$^6UZ1}8x1t%qj|27k2X#rAJrs3wv=Hr88($+I~m5yu&oSllwlhgww7U> z3|q&o z{x^~bdESTBW&)k;PI?hQ1ph*yuUKCZObnK$#i_LCfn%6htltD&I_R$A+bcM_%5i#b zi{}KSsX%uqq|V?W)-E5q*I)%FEiM(`CsXp?KrE+c4^Je{Y$ujG*D$`eZykm3yVF0w z2X68UJu`A1QaX9=Q@zCwmw!rJY=-YX>|?9iK4>c$nhwoH_X*5uIlbDaxE&eZjVM!E z+>ngqjqaqfZTJwKe@=Vsg5e0X{?kVlIG2i(uaDNsIp0HBv>|V^vZ%Bs#>$V*6S8`XPU@9_?W2J`U3!6%u_27AN7b%MXbz0U&*tzrUVT~!{IC7{S4b{ z^gSK^oC=vcrWK0mFnKg@EFKs`TaX86$j*nV7TAj-@)vCUwmXi5Z7S%A{*|MhD@Pku zj#?{6PaLj9YFFjx$EGNk6K@+^nLtM6XqU>->nlg2Do1}bkq6~Hm7|}UqVnZ250R?l zPg;78f|M8VU8g|dx=_2^SDE~X%F#ZRqirik>sF3lFufc_J6bvVMdj!Rm7_0Jjy_U3 zI^7fv_5G`~2Jk0+Urr+p?RGC%wcMiru$K7g8?Zpq21@a#cmNJ;1z2)I&&s8as~nB4 z94$A!8Fe~XIl8HGbXDc(^Od9C%F!v6qr)pldz+$0N4`t5qCe?+(UEhWSZN8ITO+FF zO-Vz!ElU2IwK0bA4Ij>7>51PDPty(!RbIg&N5PZwic{O~)Qa*8=hZ8h5Km>`DInI?*+1+@>FNbtXEC|$~_Z>$h z6_NTx{cA*WfQ5JiikBizQvtp)iP*m=W?}4TO7&Abm9ZOA{9%zEg+@or$FV4))tp{5 z5pjBJ6G9!qi4m6c{O{cf05&Dy-w23eBVL>0859@vn^C-(%;fZ*ucj=x7oEMPI?9u{ zS$RdP)_{W{_ih}57AFHvGDB_8GeSw?Y`xxCm)c^~2^)%j>W%A1XNDk9vtcTyr66LGH2 zZ4~Dh9lefd_old=)48iVQ56TWNV%Tq z9zz6wl*rAzN5t)jb2&;lZbiAkkHB9|ag^&^X)L`6kQD^ULB-2Ub142S#qCtmBXuZ8 z0mZpyM;sKNO>tyKdOeCypg1TX&TKV|``u< z-5k%MI5X--juQhM<;MI##L?_WV<`V?B5n`M+fxD(b*?lOZ=&YP6_6Y>a4qVcQHVz& zQkx>TAVOWi4LMR^rj~DD?o~z9QNDqBJ6S|y(FEMf4#X+K+!H0U(mgylx+X+pE!xGR zyrQf-#BFxyTkeO56*)8v>_aRS+4;?zDGAss_)Q;C43bE4anlMGouwi#-gk%xo)-|e zmv3+orN{A%LdMdw7tgWi6nCu^9G^^S)U%Y6o9s?Na;RKX4K&Xn#EIV^*0q%dPf^*U z@gvK=f(5G$;eh2<@F3TSV^OlOH96*>{Pky!T+;xEG$iC4<+7dLPot~%e50AY)-*XX=XUG}}5P&ZtF+QJU>dL~vRh{QHVligRRmSP(=& z6ANJFMXA)wF93^#qP&PpJe{H#Dzr`BL(JuO5ddwftUIhwwFff-3GiWVmVOk6Xy+Fa zwM3B$zB|l)-4=21%wd+;1|ktHdW7Rq9M7Q^WhVdS8_~k~J0C!g1Q1K5o&^437LhZ6 z(5n#VVek;eV=?swF`nkkY_rBLDkcWCWik~k&tu^86bG;PViw}bnfu#-yQg}T{HCaz z_xEZ_ehoaNX*+71^`37rE75ztjM;ixf-XW&D=fnjlJ=d4Mjf z!JUDKV*0lPayfc&pN`u$5Y6W3kBHWx=y;AEK$PZ?z%Y((L$nq}(-BqIMB)(-ma>}S z85M(WW_uG@fmR_5!Yb1R2EDlKTk%Y9E!!@h>6K-liDz2$l&ulZ*VFSGrM3|uNpwh{ zlv3L*h+cDMy*zAYl|pN1cB%ABZ$?d=oqrZa>?scG&2;fN73G) zXhVv24n;{c0P|%4J;z+2)kQ_FaBU3yL0xDYVq~1Tg6>Wi?A} z`FN)0;$9E*#3LBR>2Xg|^4~zqd4E@w{O9oCEYcgK)!37+tS)p0H#xl(HIzrF<8TCF z+4N$|!0}PJs&x~0dl!(zFk$gtigGveUb4Gy^j?Zk@`nN&_s+}2KpIDDB5I?O(bY_) zw%@*%EYgOfI-+24O8^I)k-862LIybx8hWdddY)m*EWp+Q#(eSv!#dC0>%BCX5|$u= z*1CZ&IAI>a{D9HoJ3xbm&$a+jjD0n;e<-(&$FtqD2~kJ(Kk*x#{c7MzV3pd28HBtX zO_wyhms-rK4L?Rii80G2c`tRI2|zpmG}4uggwWT{K84@t?61o*!2d`g7V}{jSNGghYI_Wsi7TBoesX#*^z_8$w)O%< zcAMNIK?GXNXv9lx_u#iF5SfR{n)A|{Ui3irJ#I&0Asz?ZJ#B)rsL-F58s&>=19Z)n zfL~PfI>lRntUg-@#4$U~&5f>xC;a6gwa$#GE27-@;;G?`320*bj5~eNr*WNqDX^qw~3)xko^(ny0n4#k(a^i+KlGfNoiA z@zm4KK>mhUJT<(Vt839uBPFrW@JGQ{aq)5c#2+)X9e2{VYrWr}83I2NTlX*MBd^g# zpHq-B@afTN?_JE8KPP?cydk;~BF+MDP#{Fe$YdBPj|2E>wVI z#9va2F2U+>Mtx2_cOI7k5;A2Bm@?AhFehbe;JC%&X**o?rtEKrN4WOD`r%9I*6p}O zzknj>4?xCCY|DUWy`z{{W@WzcZV&EPRQ)3zYUuzH7i6zcwd+a-I4tDMokt)`{!VvO z?**&7I%-bedsVe=cnXt1rMT44ee9(7x9;v@zvaUcA9iO_0`-*V(~x`g8_@#KW-nDUwni;Fkr!JmW5$% z97mT~mdm+MePW4p-!(yRJd` zeG#g6LzL=|%E7-=>w8fx(qpwEHF=NcdyE_PRR1iinmofnA99}oqe#>m*d-mK!~X2B z_1e*R!J&=lgV3PPI~kO%>`Txes`^@}zJ9oXb2`(RQO7`*Fxt1CVK0e_NFbuHNQ)1m z`!KAai?r?P4qdfwSCh|VH&WMOG@`#gD4weHJ)tIVP?dhV%hyWv_j;3tn_$e*4j3L$ z0Te!!Ud(6yw!t79i_ePZ3!GqE!kK&q``w2!lD~D=8x1QnOs&DzpR_-2VZS ztb@eD2L4(zuE)E3jd~Q)lTnYo3EV%<5gFPZ=WuqwNZyk@NsZrw&$wr5xNNgjP2P%c zq^rt+eZiQoTA>KZINY*<`atQwONC_Dy~W?FtBcO@e1c@y?_<^YvNd4Hd1@{`Y9E|8 zPr(CtCP%CXGmDL>AB4_gvM%4YlCK(0+Kq=S{jRmkOl`~c8t(7kLZyp}ac%~lAQ>e^ z6*lW5i$PO6=RMxS?RNPx?V0(T-IKL5F5ht6Us_brqDa5qrR^N-k9yQv;yuc)PO2YZ^CKpPI6p%gl3 z{F$a|Jyq=%=nVZ>XDF$TE)vJEgg+}XU(cE8!cS9Pt!LTikS{g)JNGTr7Y$TWE7}HQ zy0k`5(ffT5gRzEVD7g;to@M_M^t9448dsFn)e5Q`Xk}lA#&=aa7nqK&GsgSjLuq2T z+%pHN5!xq9-$Z*Z{*+HP>FB@uw;WyGb8lm&R9k0VnDCn+{SeP@XS@2)&4a724kc7MdZ+4=peXRydXSc#VznxKJ?GCUn z<>X6f(=o}cmjuc_$VVX%<468Q{9WXJ^cU#A`WNwMTp8av+~nVe*@^tw+-~qG`V6uLes0eF3Eos@bs;}C zcT@1Irx$gYvm_1cO~IhN@i}p!@Nr9blpU1)UJbbm*q0+sk`q9joE zi7_5!>Ek@qk^7)|^gwk`t5ut663NnAzPF5``w-32vr$Q2U-_ytyCrYO2M!Hy-NBpG(~Kh_iKysoR*Dvfa;c|F9ZdYe{|rN2hI91^51$* z`G*+!->;g#?lt9i82Q~*^Y3`~TFSruI$8d%Rr5b_P5BoZ`G0w!YW=S%{}3bp`&IMT zy{7yQBfq<9{vE5Xss0VQ{#klg|RqFf1HPKsW(EDYUxxNovPS1Hw^imCa zODoZfzntFLx38tXXBx=*4y#0OAHe4N@_NuWFBK9##E|eILc-r73ch{`q(0OkH?nnP za#poV!R-J;dnrq!V>n3PuOk z!4V7-UdTIyz1-T;H(6*IqU60oW+Uk0FpQ+a-HMDDv^6?Og0Tde-YAn?@$nHuOQvrw zBFIG>$r--88nM^fDoBURHy$b9yC-JggBbw-=$@El!-pDv zcOW0q+N)YwARA;+Un++{fk`7v|3;#eWN0S?qkw|a8_B^e{Uc7^@9vVJ9SpQW@E*JtsXL_e;ESE&Ovxp;BLwe>RGwEAh|kltdrF zc(4aQte?)a<_){jH)TBuc7LjX~$yR-9{`AkU*BNSx}u})=U*^X|BWyP{_W{d!z!+mfBfwiG8rIfqjtHpav`=rzlt8LG-Y1xZpTH? zyT~bHkrH0vEHu2Y#Ci-1XO}O_?&_K4aJMGU&1hh&-gJ8`t#eI1i9!i|JY>_0bon}G zYR5Cq>^K2bko87y(r5Tl7Hf&cJlx z`BDuQdC%Jw?>?lL7ey)gv<(Q0iz7{0ZBHvu%8RhAh`Yei>4PfDR>*$!7Fw102Z*F$ zrHn)kRIE^y{^SR016!$D*SIYGiI&8fUE?yel7aqS-BoL`eheF68UFsURMq_M)WSyl zOw3TGme3Vzn{Y6&t<=urn!`hfn(?fz?2VN_Dd59bSc1V z(KVbc7(}zcYCC|N2F()qo`wQkkNXKNWBz@QnY*a}uDqU!K%K9+p5gVK_8SBvo$Nyi zu-ChXdM`wJI*>({3|{C5eHqHKK9)L*Wc`C$uUX1!%YJ0j{^-O)x{^qYH>+Ng_Ihbc z$U#0@D1I7?FiYP+W?|R}{i%)&`ihpSVNHteg#n1x`L~>=klFw!V7# zYhAJY5k~pDR4qTgb|-8cEnL1%uH-Uzd)^-D2j;vF6U}Ezo(8sMg+5u&j)nf0pKFC)^*~{kUN8aU3H$m7?uR@I z`GWz^S%9(4*D5d(Knx7$*U-T5ArXy_1WwPnUh+tEm-gb$qUfiy(5b?BmKM_O!^-+v8j`x9IYl(1MjJ z{Fcy|| zoVYM&p7kyzuOT`jXI=!Jn(|X5o^GJd@s>tG=Th=o6S&V^p6zP#HYINxHKl(5_BD!> zyv;}ox0kZok#pfLx2@nha4{MWD4sIacM}8xBliWDKdKr`%r459Lsa77LGh((-9lSd z@jKkjN#Aa)TKA~l^H%o^RofDb`CEOm`XM{0!04{^gW{NI!?LT&vo7IWK8L?Il)A*$Qdl=GT=yz7Vwt z*wL+qzJL~q2%N+4YUqpJ(n$8dlJN{}Q$~D6hPNUDA5fx_cq<~6yaMrL^*l?B_*7su z2*~&Pg}&Jh`lgcqH1KF?lRJ;UKFz(?sw@pNiC&3fkyd!P*$%0hZ4*e!wBP3WBs zS?|PdRtd3Ncd%PB)Xvk;n(cwJcrQwz{-~5C-WR zsX(mO55L5YHOO}M{gt>+AL-f#9w?z@#7ZyQfWP#%1#QnuhtzG zOHM%64SL}*8c>tE;C;A*hd(*0yL3n}J{+2ZQoHcw(y?d)NVLoUMjUxo=&qkh!f`7P ze(9T$y%1OZd2xk=FLy+u;`G)MyqLk4c%uZvn;uPYd?=96>BE7WI2|`B=EXUvCRS`= z<%u`lK&1ofVu`wfQ2p=69pdogxHA+|u|H-I=~kzIpu*8jLs|NVxKS62!_QtkYANmK zm;hv$urQdGL+^J$y7c)!5q*)_l^ld?gAbqDWIQq|ykmV0J_K^ zCCH(B8R@wGkc2eS@24kZyN4lF-sNtB?{(D>!~F|1=D$QgOX%lw?jJ2@CHY7E6@I3j zGjKB>TM%^7V!mGCx&YaOEDjjMNFZBqLKQaiT#=>0|7S%be2I`=>ClV$2`s&i9?(wjGRisqb&1kQUo4-% z|NK-XuDH-I~tAbKf{-z5G1-M{T;t3K5qWnH{X1xtWLM&!2q-vLnE$nTN#c1@^^e5 zm_|x}nwxx+?HN5W+9-JnMg2)*2@%!c%jI;pw3MEA*DULQhM)b@?0E^#$lj ztU0OF?NFF7KYfB2ii&!LhJ7W!9_i5D&n96HS`7VRe_Ry~ddm8?{`76V?8<`G$eo(= z`$Tt~@;F^uhv$y&M)*x|*TZjn_u^p8-!C4slv?OKfLgi{H4m1NU7al)FnZGQs@VYY zN^kk4W7I8qSpWn+C)H~qYM!Ofg|&v@D*-}{8r7x-kJ(X^f|P-WfgSTb3k`64M!#kW zUhIe$`TiCtS_pApziCHPmke-Z6{0&JJy0}{M}z&HTU zt3OsMq4Fp0#P>*lJBGL6@*U{&k+6&bMak<3>|o4^3&(IRKUq1OAuf6#K$jkfcJcn0 z5QySf4!VYV&aLE~LZR{WfRgvUc$%u@_W)ABi67k87y&oH%KkpIlxBN@_hbTWS}tF= zO3UcBxRZ$4Tl*mxGYTbgX&vIEnd>z1Kqn_RHv(QLbM8gEO(Up8!jOq{y z8r3r@mRfv~(62jzb)3q})x=$iVBoeP2KKZ8^CoE94Yh^FfyABOiW@v%XJCn*)4iip z`DeAVjlq~Z(dFg}rP6pTb^3%VYBJt{fyGx1rT~&nScqYs=nw}*0_udmbU$(2=|VQ@ zVHxJ&YL)h*lPK%0*F zAQF>k0h)zOxJ5|mwL?5_0^EUTyh^24kad@EwyW5fE$KG|P1-M&2S2)!o3FpdEsjj3 zv>(~vQQ9dRCTjgAARSdTx1W+vYBseYPd;+rVXoC1w9n{ZNwk?Y3`CUG!z~3v z8~|cH*be|LsC+mq1!jX`$(I6M2%|*~mCvbF1h)Yp;z~X)#35FQLyun}4h`ZG3-fn+YUk~rHwoT)zXg>S z!KlOHx*C%h*D6KJroY!TVz(E4Q1!I=JEsDHp z)4&&WL2k=stO!ZL!%XCh;$l-NROmtQokSbPQcon1%KG&_>Smb8g}odq#RL=?14#3| zLJAE5G^8;sg~RNqp+9E#m!RN}K1BgqyDbUEP9i5ngk@B-lZd#Ik0^QNXfYL^HUR=U z1|mwb>5m>9pzJhGk(eH)mxIn``q@izXcA5=Vx&Bh%+YL*t(IFzFgKI5N|^*GmgNyT zG2E}H4|voY`twp;Yd;Nw>W^vC3}x|0*WXH4vP>IK1)ht$?Qw|+z4rwD6cJezexnkP zh6xchIbg6S_;z}O8X7j0e4Ifdu^%NTeYW4i-daWOgAhS@wKa$f3I30Oz^V$#?blH(qJ{+x}<#_mxGA ze#(|CSsO^Y3#AFXz9=U;d2$ zKD^i9Ki7+YjsL#hb`}0>4yB0JI)9-5v1u`8cQNzdDGL5`{u_LdvK#z&m|p&WH2wv` z{CE3e=D%H^{eS1bmRJ_E{zq1WN}4Q>e`o)U|GG^A|Eakr3!wF8;CP+32M5rx_7F=Z za$cOBxKmku9c)aipTn9_nPwY=g$i9rNOv`8JIINOpemq+fk-3_0o>v}+=A3`9H5(K zOF@!LSs$e!F@eq+Q%WkO+=P_Ats3MZVLIh1I#Q3)#yYjS%4*eGbhvizVau6FON}i8 z5>a%xTJGUOT)j|Zhf}MetR52SJ<_MfmcBk$Uh1wBiJSZO9F`ZWJXr*EplWrTzPd#x zYI!eQkE;ol)jNxhRIjlG+w={cxyivGZY+AHsOYf0#ty9xNLdR{M0k%R)Yzgm#LFWJ z15w_?wPjjl;fZR=#jM;=>tHXO3Zk2YbrM zHCSM>#|3TOe}&cdue8of^S!hfm}$QL_B7w)^oRq->Db|;$8_snI1R?x)2!Rlv<+$4 zpwIj|&AP`w(3*T!`3Kpm*|L%RlkzydE`X46*TAWjxL`$3k3Wz(zI=eM>xVAu+2j&R+`x&M zknQyNJ)y*()Z{JM>1zBIIxIG87#tyrj>O_k4=+*Aebczmus6ZeG-SWlid?>%(&9I( z<2O5fwI3^+FYO_fJZ0E5RoJ^nb8|@#8ouxW9)cm08=Yr*b^DAMZIch(RtRUi8bxq0 z(dxhxfi83C^vp`jkSx7)GM#nHLG8r`wQt@KEJw%EP79sL87mf*m%gW+oGW^`9u-Mg zeF_ihy4aJ;g9MZ>3=2;mS62WHW;pS?s5Evq6;n5dkW9h`{ounX|#&p z?r)H)KC2cTic%wOF+U!&_&?tHE!K7}AFTSL?AoUI3ireJoE`jB6PjjSt1&gKO+`)a z2@+@9eX>9T(9(#|-QCkvhN5N+)G zqbt#5Y1-yY>)woAham8~PN=(d0MuQ_T{SkSH7Y>BL`n72&%LZkiTY8%!G%3#Gl4PMvAz$b0Fs z@A72nwWq?=g2jFf_Ho@puP1v0-+5Ys2aI;uBoGduFsY=9FZo~~vQtsB^oCQ&+7wdn z4w89=&m4XVb&ZTLL$D>mOXyPz$f$$U9I&gsdwU(^a$}2?Y(AWP*EC`CnT3@pCBa*a zlYpIuB-&CP!%3C5A3B4d(B@-9+YjtZCfbu;pUMSt`P#Cx6?ud9hYgnPaI@}W8(cm$ zI&2i_7TCCeeu#`(-BE4P2C%_#olyXkUS9!q786O9-fS`#DodZ}5d~k5^*lS3gyflg zj0zS}lhVPs(%C6ly0si~1p29Du7YI(W=0&!!y)gzx-gn0l;u>)pEnsJA{or~5U0W1 zWpG+@mi}JKWlhx>-APTAh{BKsEuO+k8Qo)oF;ATkFEf#&OZI=my_Z?~qTZBsV_5_G zSxi4S(a%cyfsy&3em8VyddD4PQm#bXvy^+c<~Xql@&=ZkWxK;8lFQcuzB?J(c{s3W zdnYdR*M3s_T3HXLq(Y2I{$T^*QB~ru118V^ex46iQ4ab(k4?CRM~$rJl^xHSQ{2^Y*D_DkNYc;BIJ zwie*k6pfjj-+ zQcI?GV}^DYoYE^DaG?D%J|a3POTQO3Ls2#*Z$22#C|^EYq+z#bH^~4$Q@c|E8fE&2 zb+n*h4OHt^K3Y3Yw4b&i7*mVOh1LuZ_lD{R_V0;kXoeE|weMp*2MW{0u!5>6t1WbX z9`&MD%)1dZ%6u62*c~m<2vIQWpvQ?>j}t*-{3QF_NPO?Yi6h(6J`Araa<;1O^4|v! zz3pxn_IzO&Kat%ktpN8%```#S*-%tBN*xRaPlNPK6WB<%rs_!~zGdVp*yIWK3!kbV^{zB?kZ)g)G4^cQ7 z&WQ-#;rWU;k&>Q8eK8K`)oB?*`?oVY^L{Mag{yH~S>Ls-ejbf!ik_foaHHCDvORkr zYJj~;WE)?$YknAc-4G|toRD;9p=ptyq9-F#MKPmOY>$9 z&b>sJ_RSoHFle#N8tY$RX&Vft5Kvh#2Nl855dP3&M&8R7x1(LMW?EHkM;3x z%>l@Jw2yu>Ln)3P0c`XgfmQ(FeZWNaLdsqSP0d2@aad;3?bY;c-GxP`u#vp=45jO5 z&{agcwxC`1_@^}YX6Y9o53u6i`WOvS5cAS3{op({Go<#WAnD~7sLrTv{U2$l-pJ22 z579G5{mx@ZxtH^`%K8`Bah}Q32o8S2Qv%-4sWdwHs=54Au?v7n_-CiKClj7URu?qR zynGZ0`-Wi03)y4(7j%t9$Ek4cb!iXT`qavO8IN%YY%5sPuV!vv|L5&$V;Z%xe z98RP>uB`6oC`fnEiIlvDfU9kW5kD;&M*OszIR4YmE(Zv@XEL}s!{@fa57QkXj(ZG5 z256v(g<1*r?lOFHl2Ex`U6MCtpAs&W{T$k6)XeEii@pegIR68JQq;HngGBveH}lvb z)(asG=S!A%c%N^Yl3P+M9Bl(h&cRvIA zk&lcrpZthv9B-~#&xv9DQBH;fDG%e{9-?8qCwbA}7KOegQ8!XQnh&Fie!|5=iR2oi z5K_u&&(sdlKzdO3R1lK{y1GG^j>#c1I#(a2?{~I{og{4P#Opb2c^YdqokP&VKqu)a9V3Xwt z{DyWYmFq3s31qaBRi7EKLm}{p)AWj>wUc3BUR4erw?S6WI^9dLmGb!2hDIT#{&idKmUG+oza?++|@ zkLzC8TOpjkE+Ccdv_5PtMf2iF$-@jx-oo3di+?VlUA{rY;L1J{1b&SWm+HrFBIIVa4iE4}P4K&3-E!Kf+ z=u%)>RFby~gSi+FRhmbp@zrxhf$2~XC7+w;k@gH0_(CuG84xkv zi20;jC=Sbn86XS&W*Rye#}TdiP9Ss__!3k`TK@Hl!#S0ob3qfL2&THAzb_Gc0gVW^QRQZiw}Ll`zI+>Y4lYcxHLy;RLc`o}CTGc?d z!`vMjIdgs=qvX@^8D;gbMpo~UgE<#Qi38ut`eBVCl=W$iBK-rm`=@V*?L&EFDlok# z5>j(6Au^L9dIH%3;o5~76-W8IM(|3f+F9^Y@^9HsFQhLFVKAi@ho$P$r__P?UNK5l z<|Mr^)WC^5`Q#AVc>zvJ=)>OTd06GF!r>C`-?(*@9dF?(_4B*LLT0Z8C4VwX!~-E_ zW$6UOOAg2@I`j9?yYP~DyhI#8LT4U-i2MWM?AU09G)Eq?Xoobxti0iD33 z@~XZm)c(}HG())j9<1RvxlvBIR=JXY%5I)k0G*x@g2i;ks4Xw$170SG{AjB1;s-pa zpZ`(RB)06ci3`R0eRW7*@(=tJ@1NDyd!ie+`hrJ6$FCAldE_pzzE|%Scz|i=5U=?M za(N#i=SJ^|ks{9sq?c{saml2fzJ#%1AsbTCErHF5RBldBEUtfxCn)ieyQpfqJ0Tcz zV9znIz{o(w<+<_k@|50lNshm6hpZ6c)jYuGW|~**#(8fTgH_)xb{rwZ@kkiX>SVdS`PFNYmF%f2N9vEIdR?q=Wlq_p_>qK5D{i8 zoEVI1h!&3=eFr8?ah{zPG$qx|!0 z1QKmX8V1i{7#{=|Km`F5g(AU)o!=j4Rf{g*u;)(YF^$xfbNHH3$0@UPh%ICV3zbEI z+Tb>+PihBY_MZ}WDNE1fGxJ#Rzj&TB8yweV&4$kFMs4TCQcWB2e=3S`u)p$Qhe;Ei zxj$IYjWv*AB7X_wM|zb*1LKON z{FX9Yp_Jo8Bp(4rhWBJtX7WWP?{*#(MJoie^Z^F_r>{o;?^mH8Y0%GK4tXG>dVB6I z^a~Du$DvGkB4sQV#m?~8j^b|Ov77p;CJ)FRi#<9t+FpyevjSvI6?Q_|sBl7kjh#@@ zsK1=MN(;nZrG(>7!a^xF)QORh%pJY~2-qt*$Lyum~g?%Y9gBqj6FpsSHw zcomW#VZT6>@-%9n;r$H#bX>_hg3(9tK?hv&vDhGgns*njz^t~bkRM}^&!@$>>T4Ru z(;XibI;ekv#RzT+!TgOW5b6qzyKOVHYUDk8=nWcZ4(gje7P6ktD+XH~dDrQ+QMvPt0Hs?8M*}8jp!d>_S?57#Q&jfIc1k$Y`A>aTB z+y)?vSxBdb zFgNg0bbprKB<}K*M>)lES&GMt>EmcHV0Xj5F5a5)q= z%FY2a7!$u88WJocUu4p^<2D99li|lQd|!qi$Z)9)UzFh@8U6!d=sZ80^jpyFydA`N z5WWzp7i?NE0MZ5AbOK~R2}XtF?>ygS5NqQ`mJXzsK(#rj#vdTw473IY_Na5vZcq%7 z57=*)ZD+cKP-#DPB>4*&TCTiVK=_2voe-+8Q>EGJf0%n0=%}jek3R_kqM|n_saUN> zO)F8fp@I?xO(2mOoM=?2SnaP^)X-`x#fe}Q4a@{`I}T8_Ra>pJwLYpAt%?s636BI& z5o}e|Di0OzILgCU0Fn8BzUSO|B*8w`fBo0`vlf~AIQKmE-e>Q9_St9esrHz3717eQ zNE}b(_+rzHM3jm5^(DMX`jY#}qJe5(I_0+2Kup^n?}~Aidsrr+d5%!5=Dx0yNvN-# zv)GM92nI>5;mlfUc*EX&r};tgNPjESP9L3P$KqsWg`g(_-e>aFxZu z(q-U4=i5^4A=yV;MxJ2_PTm6p);b@_XxpF6Y8pzGE;8rBusdh;R3*pQNKUPV*! z3yRwc2T+xbb%e2DOj$kW73r?l>d{`nhplk+Dn`s__Yi#9ZJ@Q88WCfy*SY{Su#YM{ z;K8n5Grnz@>IH2;JKlEsAUYi~M$Z$3spgB!Eab2QxbioH&goz}x$kPdKjBGTvOi>2 z;x%P*b15{Z5}f8&r6*Wd7sOyEerRYw`XabU9)HDL-M0)gW9~5PlEVRdgZJ_q5^YA~ zX_t~)W&4mSxp$RTe!NJ0gG%79TH&lWh_dS%(HHSw{qGq*C>J3r?fMs~lFR(>>l42- z-?Itci#_o13AxWhm%&&cjw^U#*8mO+ZJW9=AY)hQ5p3PtGg#T?MDJCjl`j+j1W=zH zFFYkwx!sx3M=;e2Z<7J8)=|KFfZ}Ie`gkfu>?ideUxQk#Dfr1OE>0e0^!~4lqlK)9 zcZ;F5HXTFjd2niCXw}kO^Gc`rCDD=MWF! zI*aprMEimlcTHqBC#hxt6?YBG zJnhen$=``Dlp?bnMr0NpB<{39GpKTng0zJ-rw%YNz##>x`s_3T%=*N9JJ_h^Q>O_t z0No+6%E&T`YElyWolHrSe@tNJ)pfC&;X6Za&1pM%$vdAvvy3jTX*tyFK$StFCi2%96ZvZlk-z#Hs;%LC;umei5BWahR8Yln zRAAjZYB_t>#_rWwJaNC!?$r?OUVU$GVlYpxm>Z0=0Ph4*XSOrVmmKX|Xg@RmG)4+%uyt@xi zx7YQ(cC;G2X_@!V0w#M4E=(d9*lH)tzQ!J5O^jW+8u!o|Vz&`-Z443D9;V&uSZ$vB zh25?*wX{)<3shn$6PWKlEpv{v4z~&mspJyrMHO z_d_YA-UC#Ue8%G6o*xWMZ?9lBUf`XmdgIUc=f*qQw9#y=&uoeG<@rvql^YfJ?lnA+ z4pWLFUVm>6dZ_f6*u^%kC8XftBdX%b(W*Dh8)ZX}z-;z4VvA-u< zOhKt*4c?um;^uLBGF3`8}X;oeYjd3x+JC2Ks(fp*0UXLM8{2P!_h10AF zZ_l<{JvXwn$~>O{4@usG{S}t%Z-Hm`8Luu1oPKq2_P90P0&9zx|Ghw?uNM3$FI ze^@3z{+H1o(h)PS=@c<@*)v9gC@ibX|BSQU*u)rJC%9+k8`~2r5g?1rK}gkgjAPTiUK zJ(T6Hbe_5a>B&Q63=e1Wu#@x)I%l<|nRr4kZ~YE9Gh%^`rf?cNAN|CQd83K#*c;?7aHVrrmdSq#6iVATT?wodk6gk zv2$^*SfSXdGw>w>2fgduW$Z&Le?qYQpTYc>$WdqG?J-S@IqU4qP~2ANvLc9^ZvN}R z@*RXLx#p)zU~_y3`{uVvL-h?j)_Mhrr{+Xwt2D|NN*NpHd2Kj}z~ot&Ek+5%(CI|w zvPAYOUVFoe`n%)9x5M2ZC-UtucQm$7aG8K^vl6-wONn) zF$#@v`P>H{IVVmyl67Z#$>~JJ4g*i8^$)o$&iGW1B(1J`QK(_;l5$0#I0y3=9k(|g zkMbJ~#Ts#Gx||}*3F6XNo{)4fdOArf!atSBIR{flsmWTwq)^(5y2{Pr@j><@a3q?> zVr&|5tp}DV*TuCf`M_xrY{e9Di^c!&Luke@Dy>lb3@A9oE`44QpO%-LXXag6i%b66 z@{6Dguj@RQET2rfC!4UbS^G?bH~nBy=->DYMEf*|l>c&wPs+o^nk!iclM~2v98`Vy zpkA)C>&&BCuX9kt|f<%Zc9x!{n_N3kd)Sjp0+TLyd{dj~~`^6~R z-+jxOU#Y6BoFN65F>UyA<|@6#Sg|B++9>a8+89jyh2TW|+3jQJko3K3T&YY3|#!mAN+OUA@oDe~`+Vy#>uu(JGRS}pgE{fiLT^QAjg z^9|I26I}ejN3iRuBWv`BW9?ThQ8^~np9eo7sMeGX^&6~!5c4Ql8gNGr)eEE7 z9Lk!H)jrQERL+cq1)S7@dTrQtikgA%PC&J^`@u3ew6E9;EEBYBG+M!;$8D>oa>s09 z(L(C#4Mq)J6W;Qnlv8xOVGKo-PbVu!#Hw9L1Q%$HOI0i zcUjq_Pn^DHd+y;feU@VbV^bQ@3+d#4f;bSmCwce(R$4V1_3ORoSTIPzj^4*mBGZcU z2qr>#BRQf+-=Fs^rVzBHA#Yx8{3w`Q`(+=lN;%Zg(#I%J$EkG4mJx^ zpt9Bb^pEEJp8}7fm2vFCSTvXrCA+8%3&r)2>47IQtDJJZW&VKE9_e}1=9CjoP?Je& zxuM?rEcMbkZ^^xuD4*rmpzr!)1j>@r#t4B2J{~u%TwF?16DsIMg%I|Jt)Ip|yu$g$ zm@G|I;Y?w!Ujcy*dbvv(0NYr%2`H zaXzp8`C-jvrHM0nBszZX&zh25I~M(tj-^VDdPcIW%S+>?vGKEp52!rv%s3HwJPLjV!H~II2wEU|wjMYDL z=5y*_{p~+?dU&@rrm^3rMdC?aMdGPcIX9ypYn?@T)>#Z9xoWPFi&=HS_y@we{G}Dz zks>QE4wkRMf1hpfQWM0LREvWT4dlOKh>1J|P|_0%g=+cM9+C;tTOf5&Uh2@D(eqRw z`A)Vq$>R&NdkfZe^i#O)(_#XgIcRA56OZrK-a{t2rHxcD=bk0~RCM&kwdQTc(l%+}}I z77zNu=^HI;F)daTj3!!khpt`w`JlyDaAnL0@tzI8WAN?KvZ)l%zff!{PCe)1sE$n~ zl#d4d3adAi21Eu~DH^=l=b5I?uG@tO;qzAzc_WOBHfxQ5H~iQPJhNXcz@u` zy@oB(3@4nbeHL-dp&H>_-Jo|2RSiwLp?fX->R+3E&jjqHexTlC^=Dxlu1)99i;pWc zl4V?v`}2~=TfY?Lifn}IL+{l*hLB1Hzt{yHp3)l{yn#Ga!0y+iNsc$+nS*JiRPmWf zKF85IDZUVotEFIjPk^tXe7s@y2Y|EL|NmU z=z+QwP~O6?_3uF6ZB2gD+R=9e-F0GN;!b;4X`iPLOIft?w!*`J$z5elfFMZg+9(Mh zAR$bJ#t%H+^Z7drMy`??gc*l(7v&~6l|{O@`Va7fc+V}uH%ge+>|hm`qx-ItW;_aD zNG}lLg&43_Siu*=Z{`bA*w)nG9e%z*A1xC}=WulSJi^h<1n9JNlHRZL2f*my1K#jQ z;5^@DDl@;T*~chO^wo{d#}E#kk)w+7?zF0*H{_?ts=-Sm{{@`Bm^DpoYu$Wyx0oup zJ*vMrYS(^6m(j0O$vKZ};9XuChF;7;nFcS^Tl{iq@^bxo6dq}C0h^==d@qY=@Mo|| z-Wtk{KUdbI@d|?8WO@LkV$Aujuzm3LZm{*?`oXVM8aQ`??F;q$gzYd!Z}hz%mW!J& zP4wVFE;rKpT!-S2k>=zQ!QQKe?g$@{vmq*Yo)#)d5!9R|yVa#Yc9FLf+p+r1EICyR zv;cv<6F0uyxq)-m`9&Kzd2p#DII3sYx{#GZ-~f1kc%^ z9`aB8GoTjXN8|ioqf>MLv?(7gn`r8cmQB(xmBJqGD%-BYRJzzNWH=ii)l(>5olg_; zBhf#V7K_vGJf87NJDmE*C7}Y$KZY}v&!gO^PdlkzzYp+%e#cP^Zya|{2^GO4CTTub z4Ei__1E1e}N8EDf|KOXWSb-?~PR&D>IF)&Zp)*MGvaVM)S$ z*bF`g9^N&AoEnJb0o+n)*Ncy)56>g0o~Cn6i?kAm>^SNl9%r`V48TAQfyH*D_as{g zYrh0W{jn?-(VuwFCC$v>*u7>%W}O#%*#8CyQDaLK; zKrZO%JRT5pP+)vi?|_!tPD4x1){A)_s(jCh*i$p>OcL>>dNN~&YDxRGpBY%h6w|*m zNsSS7MlK6fzDuI*buvPse-FExir_1=A7vT~1n-=*lM~c_*l!x1r-pzMT-Nr6?R4@3 z$Z25me3Sr)TEhu2EcF#jG4?ysA>3uwB{gm`$b2F#tTwRRcT=~OiRRK@g({Z@&3wAc z01ehq%h{*dPamNo&=3{YB1smb==rb|UmsT;zoaZx=eF0y&n^ss0yc?v*17+vb63^5 z8$~X-*8g~q&#z;Hs>Pn1`5iI{M_xq=v=++uELgd!anlI5=5B=e+4&ICn$n;$#kYP~ z8mdG_4g~!{_r>JjI+*Ysh8UdZR{*wghB|!0@E-m$O9&OvE`;z*#1ew0Tbd9w$%V)? zAxsp+>Up>?LRfTw3R^`U@ zTLyNg1G`G>Mh8EAL?Yk`Gn%ib1L)4u0ScJt!0@2-SCvXij)~wdGcaV9p@aLsxlcL} z3`7T|h7N*mM}`hEL}27^7doh{j$sdS=I;&U@BHuSAk1dsi^;!-+`S=axU|lFsg@L0 z@n<1T)4b+UY2AeP1F@USM6Ut&ooXwTB1;R@%K?E2Uxng7DWspFc%NrfP6McOlQK;d zSGyncNyvShzMm5>`{Uf&%6D%ZT7%J6OY%B*MNRpeq5L;P@qV+(k;}MZjH9C4Fg!!f z?~Qb8?%ESfA*G;Z6hxnz-)Vf%41lg`6a%W`RjC?xY4hS6%apUg(B>VD^(JBh)BP%%4VDroaTDiCBQdL7elrrAYu8a_=;@CjnDx=lJrgYR!!=)bJkK)j}@>S`le==gh-(B zlkmxb_{cqI5k!E#<6At6j|TobgXJqwi1c*UCRj|YaDTUu*cq*4Sb4E=2Q6*@G1MZV z1rrJSo6VgMzFVT@IAyl#tRXgZobglrs6HKek>|=9a zzJz*z5{(l2BjkQw6FVj3z8P}Ygdl>WUkp0)R|h*blY=GXZVVlb9h0CT+XH=x6|$wy zdHAE!pu0NQ`Y}K2iKF7&y99x^p){DkEZFz8p!4v%rNP#1rEK)n_5CUkX=5;-H{LDh zSKO;uo6vA*Kd&*6_&rcG^8#3B=AW+<_#F!4ag4sBYV|;v6icDpTZyX@HGvfb_}JZN}cB4qb}N+ayMJ` z%#>0iNWG_Tx7vLx82`mrb>%D4OhGYp zhz$hZL{R6%`!pr{nf}H{q=-b~MB|1kmc(jqFRYE#{JOBFa&dSQA)%beQK)fgL0ju? zsV?6ZaJSbI>3X&zbf8Y25vttk3t23%Ig?z92K$oL@pGuaGAEpjJjUaM{Hfl&a_ZP> zcU|hm$d?7zmITW`8XYhEOCRbrY;@K`Qf=sZe6Z0N>n5xwXo-T@b)L(ZN+ z`Hm5>F)grOuh6ZB6}2{Q3r=_g9j0|tL8_#F-ev-7S;(T$8Tx0Yi~iFHs4Zu&*s_)y z*Or8-0_2+V4+6-e`cPT~On40)c@P@6U82+x*hO$8Fz++9v9g7s@=c-gCDm@fXVB=* zWXrFbTuPO%G_J=#?SS~012vlJRG(Xz-bZD`${oqxY56xy@-JoNpZm5XAf>dp?5@0| zM^O1`m$8K=tdv?)omwuU5%FZ?;)H*&H{8#eb1W#Wjx4W8)Ulp0^0y0JwP)z+Iq0G~ ze*WHI{Fl%ZA=U^DX1NmK!88~mZ9r8LJfZTAbK^(+D(NsULh;e~KfDqyWyiTT7McrD zjk&duh-BdpYAaVcGw#J4kYG}<*}u|#qqclkb^h8Ka*W005SAEkUaW9muPt9woxi3A z0*zlsR};4)y2Ygp?(lnzjzJF=ZHcM}(HV~P?1Wn8#AKr>48;p?IXm93$gSTa1zWEx z+Zx{^dngPe+F#$|su!qy!mK8)DK-4eUoB*s@oP6F-6u@b?h$hmZ<&Sv%|>8BB1~xG z*JQ~&w<`v||4o1XPZjw6(bTd1$x2PXEq^AZ`y+v={*>}HhZ#cg^eW0%-Bru^hHh5NG!jjY~ckrxubxPnpL8!l@CRR8L{5FdPP=Jge=CZ~uoMk*6dP=ve zITcU6L-!Z^th=oC=NXGE$5gTf^-*uS5=4TvgPRf10AEn*8lm zEhot-;0)M5l5NiU2Z;+w5q*#or@6hnQ}=vSVFgC>2Z&`~>WxqHtu&E__9-}O<3||5 zK~O&@`cqh`GiMQlGs4QKjamO0%^|6=n#vT+-zlNed?fR*#{ERAN0`91sq)=UU`^%g z#2RF^4>K-*Juu;o^z;wLgI`t0&)ORz4j{4RV5de`z5{gqHhE}Te~H&4Z&o?c3&q*p zHF!xj6_Vw4CIV+wcw=?t+QZ(CyjPHVG4U{rpxQM&z=H==Bdq6)pg(CoUF&{OUCt^7 zQS`Ah`(1iCDn5+NoO?5(Xm2-B)UH9KMG+PQ?#B$F8&)4}{$lDyNY1a5Mav~2C%ozs z6fq-(pz4AylF06-?L!{DmX=40F?Dv~*=anc@mgDst8-U*l=}Zx9v$6X9-(JyK~{?d zl%DzjS|YVeB8e)~5(xzd;tU_=%qa!6weFtk_|+AOtBq7@GF0cjmDP7vxspq-){w#a z=2u{Qp$SQp9cClhv7uX@1UZgdSDui{fbL&R`Eux!jh0~f?hvaWUv!-tANZ?evE;Vo zRIqYa__!Q7^_v_yb+g3LW%SLbU#Xk@eVMg3EwglPDRCx+vIbe`D7W9z(`>&_$39~8 zv^7Rgd*?ssX@#BiwC!Zq78`VZJq@{9!zp3(v{q-vd%Q`M@~5V}!)?oNAN01O^IQ%O zbnZzrsl@lWt#Q|iOD76g%%!<=$D8NGYwq73Xx&(7_?lRCpL|$#A2czs>ej}O2sHiEM&tmy4}VXp8mnS#_UT_}9w~y;v`NR^4ZxteW#X(GLx)b}XxYha0hK&3duJ zQtx*^XE@W0kk#RLRO+FFKF;oE{u@K=wI?r>I?i?{n@jIxt&OF3A@pQzeD5%w(6SUG zN|dmLfMgr4Vb+%kYvTw?c$~3Hr4?NtFK^!2MbjB~x;Sjp;JRTUc@WcPsEUe9C-z@AI zv+!~~nemv|eAXPeh)JS5m0rdytd&Y zhN-pVSXPxeR^j-SDN;B2ury}jL9g!9CLEO6tP0VNJun#mLDnvems@A6a3GR2=7IM| zx`lV`!zlcYH3mMJW9bOFFV<8htv!$)89PT{!rPJ}^B8E*{m5Da)xq&N?-m3neD-hb z9Ej3D{D?b)?!T->@c&PCj$HZk|Ju%R%Iz6D$JEB}cd`afHM+m|&z~`Mvs*X=yrn#_ zmJ2wLZ@FMY8iE}ch_g{4rNF6WWsf?UIB}Z_XO<`7yK|Q&Yy72&;jeZ*fckn%D1S+2 zdBRe$t?_+-dGbT%NqRY;N&3GnPma>p1YFH52mC3q&*i{bw@R(|>fAkvfiPmMo{696 z7ye;P{{E2CN=uEuiiT|jC#yuu}`!*&3{2jWi4qm#`xK(T5J@nk$vlGD&KT& z9=Jcin(!$$v-n8aC<5^tzfx9cUw9L_xEzNGYlA~;{B@Eo0;!9;o#<|{QD0q@wurL$ zUgcURaysp$H6@G_wOQlDJ4}R;7$;ycnH_+x#tHVns^i~hoOad5vKiw9cm&OWj$4qgHPMVi*cMES1w0c-6mar_;zLXpzMhKOGJ1|l- zHLtw!c(Z>jxiMj7vTs(u38Yr?woj~Ii~Ay`jqUM^&<=x@Z+Ej-h<8f9rU%3(SDK8f z<3G=@bzgzw^2j%MrD-XxAO?=sYd&YiBqKC+}1jnY9D{~gt-@AtQl ztcQVtpgRuuYJ-|pZaHIk&gd?pgSm8DN1oYR`RLb(XkHoq$E%j?tF`_sqis-8z5mHG z{&VT+Y5;#UQ`H~MbGGEcagIk!FTo!lM{u6w<5-R{!y z#n)h6Ob^Y6e@O7}8*qP7)N({YVgPV*<}3$*TAMRkE!v@`hIy8nt}%25H9bE_!-7|1 z1*oWM8CZ~kLK-T>;BVe4^?#cb*EH6P6Ava*seNmqzteD~-@SHVcP%vWJg5!8vkhF0 z(a(5E0AaH)#WWE^PBegX>X@x)v7N{8=X>m%5`PR8mIO*Ov2-Sh_={3SUgZRoP!UC~ zp_Z+wREDgZ9|sqps7iHDijPZ`AIU1Kf)<YRcVdenvJAUJr(XEAo zUE)VvNe^=0O^bw@^40DuPGGxI7MxCHtnuR`2*vixuVfsrE{8Mj5Z^Lnf7iFHXS7Bk&P+k)tmUH#g(VeU{wjgW z*TW|V;{oIfl4$1NrHaHyu;sWbSl&u{Ks}xHU7p7J6u5nQ%OD!8T%6f*OplP-wo{w; zCd*vAJ(H%g$b0H}+E=AGwHT}QAoUj@<4;H|4uvE4pv*&Dgcd#lBbiAo$@TVi4s|WTS>GT~~ z9E^O(*^0G|AGtds8~2p2Dqj)VO#dszpeL+~4W1L2up+g>9XtnPf(*;0=EX|01fGKc ziPJk!W0%6lRH<|7yCEBEs^qNk$D$!X9Wcf@^|VDa47U?kuW?tg)l>dz{t|L~%{rBU|Xg`&=bnrzX?1TK$IL zB_O1TM!zBle`N(u-!jn<_*>k#RgO<-)a5S$bCHceZyf9~8 zs2lI6`u=#d^ARTl2A9e2Q_4dT5+M&S+rD|;lb6aEzs38OZ_gf(5vYwnz<1Cjt~Dq4 z1m>KV!__CUERb-72I{?cusy45uNj;4^~K&xy3kp#XZ|Wr_1upYJ^e&i2wI*b-vqv- zpQj$^ls>p7FOMUSYjH0BQ~|4V9139>gBCjrddKV6H0|E++Utg0?==6P;bRw~Nqyo( z&*R2fFbe;~mxoOY7Y#!P)2?H6tg0Ak)ba^KCwh0a8z|P0Yo1^Ov%f91TY9Xoy|wX+DsX7@|;B#?C9|w`*!g$=dd7KnwzCz0}jW5@?9oH zLzPL<@Erj|wHNxeU#;5X!;kr%9An4!RK^*`&sof-IMv1bTo;%w(!Oa*TV!`Cth7Yv zf#2teDHjj0OvGMESi~t846AUOlOm${r~;ij5WZ!~S8zx&M>Or47T&LE7a%!-2gFaW zj@2u%^_))xq4ug`%^d@fgo5Ln`XgTTuu5mvZb}57;q^%V_Xs+FZgnD?O-tt@Ubb7& zfP2aT$pLoVz_^_SgW_ksFsyRQ=_-FsZ2&h+x}){#Y8?tx0C%^sDBZB>=`GaxteA6R zqUk|iWNU?D2VT1*vmfva=xPPuCl^UiFw!Lu$95~5P?7hnG7go%lsI3$IZ2kmS5He!S$rpx@kj5bD^xY(05wGkI@t9!;llp~t3O zrO)QEihagd8EaO>4r$u_DS56AiC~(Y^LxI0_IN(2asIr?iQLRBQ^J?BAcJ*tI8N(~ zO+%V_B2;Aip`++cscBAB0m}l$1kU6?&D}*JFlWKae3A-%$dn{yl_K*@Sb0vokg#XC zG`)IbIUin+D=6LO^lNd`?lYX|@40E({eMn0VlIYZ73BhaU2>xE^+kRfU%BIE;gA`( zf!}&gekB${IN9}UZh1hQDR1C3U*)wCz3=j(=T30FN#!s8+VY}?PUX*4`4c*okL)dU z=9CF$@w1_=t%QIos*{UTyVr_Tj2&R;s&!MZrj2Kkfb4)on9sqt0yZgn{356aq@EFm zbNt^TiMD4Y=bYM4SaymFw(s|enbqzgiAu^_O6xRo^mXsa+vu5uwtWw*vpuip{crUC z7n%1xbKV!zZn5B8XW#;6T?H!~v_@=X==r>EN6h)%~wURf%#dw*xvddf%k7C3WW=K?gB zTH4i65@)I|L;uO%|Ll8igyp)PdSBY z+j4fHa_P04-BVfZTFzFKfWpS!&Vs5QEhBqFD?ze-8EV;YH}tX}O|dO8vM~9+TE0;& zi^7so-CnI*QD0&*J$(Hz(Me z7tGC(++^pTn#3E{)Oa& z+(a3=7<#QqNq7xm=dma*b>QfCk&iV?0rX0$X2%$fNQnY($czR09(TaZHs`Ue4FChk?P0%z7Z zI38E5$L`}wnFyL|nc21!->C;E0~&r?P0zh~-Ya8ikJav}rrq28{XD-Rtg)vv@0xa< z7(Tga*U{IWqyPn}#Y8B}}n3v^qkVFgG23y4$!$P`g^DPjp1>d7qaUK zFWjI<+q|pv3#>$7Im+}7fiON~W*>s;$4j>S0A(hZ=bjFPt^x$GDH1u z{*W$p8QZqT|7C33L|<5{48e9s=K=oMCQ2=}=$xiKY{n*gn8tRue!GwD7(2G7_BUhe z-+gSqnxR;fcKTX5t!c|d#zjopLw3~F%n-A-v9#_)_ec6b@6OD8X;EpK4@=%P?dn4B zPE?sSMe|V@BNwGp6o@|jVqzJ$qNjjPZD1&6u@1c%V)IsrW-&UkRXS(2&|JYBGk|jfu=4c*w}%%QL#1&=+kl>>3;{*w88yTj z+sFLoKz!;@Zx@12rs27?=nXT*x^3P+iQ1wC|4WMA(jDNBp@*27u%wH_r&ka}3$P6` zOglo5cE=N_D2GdkOP&xZ_zf(?CiIf>t1cRw-?SSJx=2PE!$GIQ7MpG;J2TIT8h+(7 zQ=(nvV@`k$(ku^$VHloec4E6q{wIrUs^sNsC3`VI3W|okTXxr0(?11R(_Hyo?_c$_ zs`*5hGJJ>+g$u=g{92ax6z%Pc1bmiv17qkiSC&rxU*^iv^VOkl^S|s^fBvJO`g7zW z99lGl=`#0d{u{GTH@-D*-4?pV|KZ{8Ldoo zZ1UP{f!RSqwj!T#PGF|VkZd5Y&>6lY1TPiBPoxdc>603T9*f4rRlAJxYV{crRRs!T zL5r1&iF0@h-7uA$S#hqF*C8q+UgEF(A>J$TTD5TLxbB4A#JF6c>q~@n^2!J6^nEoaRo2 z%uvI`HH=nH2VfmpNP2(xoL(p=F=fXk6a|!!sEJ>j-))Hj;+)vP>J{FJBLN(REzZWj z(>DP!5cec3oD&_-d@w#M98PkgCQVHo7clvk-6eU-R%)}N->FQ1$RQY-G*F@!jUDc2HV3ULh%vA-wF^Bj%~I{+vN(e z7VkJkQoUxXjdOY;Et4vD8PQRTYFmm-h|}h=Jk)o|M>IY_Mu@6)fp|^63H-3abVVT` z6eg*aowx82Vm=kpMd@BU`}Gap{#Rx~-*TpEywOt=#c;Z6cSmaVIm86rd9KSJ`ZD=e z!nbnVs@7dq9cd+i>k-B0xJ_iV~hF~dfsi&6}f%Ym3-YP#Qlgw+)zY%5?c=ce(yllubznSFiHL)W1 zi&__N+)-745mILt^|mMsfS>Cf6& z!*nnGHuJty=@U)qwtbh@9$Kf;sl(KhJ1E_Sd91wZe_(DH=JD*eEb|yKn_`xE zTqu1V=5f;{`rYPD(Qnt`g-ghP@{7%R>~my$og@V3kfwaYO*Wp}jhkF>9jb@BEO`#K zhuoyMh#*MPUIuvm)I?fo=wodG*1R_hJ+<76O%dDRB-3l52GRNOEPeBXNfA z7bmxN!N{qm{{MFQaq8I`?APk8^A5KddGuy=&am_4Mi*c2eGI(|BOmCuD@L5=pPG3# zB$e6(IhTy5^JY-lWyq^6lT30?lTtD98RZj4F{H#d_2*LEZqT0#^+#cqOxU2r9K`MS$V;bhWUBdz?c_&N=}GZ87^w8sa$3{LXkV#b~P+s9i z>L{QvNGwZG5|A0&+9Cl}&+-D2Y>xypZAJQ<$|3{o&rp6OgX+CL0ZYSdvuIb+>+=WQ z-%vJ5@$%VSOnGm8Yyz2$9Dw%GRORu&SGlm1C+Rv2`<7eFZ!a1lY= zIlfb|SAMtV*2sEWVw-B4$5&K!c9p8pjsAgToWZ#8rR4?nPTwo56v+vhZLLDyJ9mi{ z;We^QbnPvwp=!9C$k}VgMNHlT zUB?%S61-K9n+nEpovt9jb#4Wbp*b%Hbbc9MSi}Rb%yv(+&+E@M)7&ez&kS%1?RAUZ z?=;K&>nC}0ey?s^Kh8VrVRKRCO|wr1cwt>2vR+czuJ8&UnJPK?N{OoFiFL5D?!RmX zdmGsdp3A0Upd80)q7LsWK#SC_nVdf6B+Jb@|Flt9Z069b(?(&Nn{6&|D9}+)GH~Kd z@Je{t(B6&fwC{C(fJn%J${mddO0Bf(5o$vz%PINYm-jOwX&JQ!4FGy8bjFSG3{ z70;m@E$tB=rD#90c^3@r9e$A49Ff?co#>&h;otMMyCS*8*a1@QCLcSUC`iu8+N8Pf zk-)_sgEtj8QI_Kz)@mml71r1qEQr=0K{0CUa~L&h9g0!prlP#Y6JH~tWI*8TtkSbrJhlmX^m+z0v-H}>c9dp{u*n}B^3ACKyHjyyYl)wfaE z-)THN@i||Z{`iNF`^zamx?WEgoSrKA#*a612(57wKI{ZGKGbBjw(p17O*8NsIkRSe6ATEz+cu;8uK>k-4IlPZq#MFskMoNzF-57F(rsQDMbiO??=U2{9ybWb z8+d_G!BygZ$<(<|1Ij!T`dQR_5O0l6otf%#T9`V~kgig`L#I{ibDb81zo7e?AI9E& z!hbZPJ5x6%#c^`)DvBZ#5z%?GM;Q`A9le~|D3n|#(E#x6}6P%e6w=Cq7 zYz)x*`bx!VCpM@pG?I{HjO{8)sU87^dZLtS@UEeVw$>3Nbx1;)a712pd2O}DCR%6I z+D+l%+A2YJ2y_nk8A8ER$TqnfU=ebroNpX_4x|-@Y{TJ1A>00BYJJ6QSG;?fAl{lh zb1r}xiFev&p-{owGtJkU=94F7){E)5Zc@*_GfjA`qYmwu90rg8<(I1&9IWXMgB#xu z19z9{%Y-fBL}aFu>{?)8lB)XhM<|>x%$TLSe6Kz4tkAjDZB>QrP-2VFVV(Y#x6w||JZ}K)(5P3h^#&sinBOg4 z6~Bq`%pB3P-OY5O4^gMt(Ok{>sY^wkKsK?;ltLaUWAGMl^L6UEUOZ-Au&-*8%*we) z<3oKPcOY3(7f~+3hKd#^R=!q0EoSAHWAo*}a~;b4R^`}dOwG0V=!0BZ`C;=3-__-Q zZ{{9Zjg&TB`3?^?%%9ZRVP0x}w|E!vE5KwaFx-f|Lu7S+YCL`BAE0@+Y=P_rxt}O) z!th(n%lo^(e3X}s=H;*S63@cx&9q#axYGZ{496NJ#eBl0OPGQ?R552b-dY(*p-&ey2hw@*M=TLZbs^p~~X*j0`+$)QiQle=l zw~d|J1i<0#YXa^L7=;kH+WYoY!Q}MhD|Y|+;S7HKIIE{8pJYgX+F!tU>Phx|Gx+o4 zt!B2Mr+@SS)#oq=yJd?kOM6I?)}oCpZlT!E=~{1Vz3&NmhhoQ`9WNZnA;HPNOx#W- z(vvt)qCG+Fdq(rkrbEeE#49;Q!)5uQ*zo5hYvmhzD<80#K$uSaI=7D3Q+(m4@ic^f z7}=Q$hXb*aM~rCR<}4aq;Pr*bNG}Ko$*Z<3(4z+}Uvms~3@*c!?2>R)Jo_U9n4A)m z*)O$2$}mo+dADrq)0b)FIdTC8Hn+NIZ=(~Psm?X+9qUA|VxU~%fY*y}Y+&RT+OSRz z1pSviyS>nC+&x>Jnl$8R`l_chdoz;`?W?A&z&c1eGv9}4#_L~b+6{M8mb`fVJ-VKy zty^c&ib&!@Q@%&*#y;s!`kG?)6M65aKj~vWSwl8oU|CaE*t)4Ge--|RvP92(N$uEl zz%zfLk8Hl;S#JOhGV|iotvvEQIhRj{=dzyUayFke>$A}>nAFC~kCsjwJde+uMV~~r zUYOseK|ca0>)ekL$1zr;pEBP4oY~_6UFY%k%Ng$;b8W|~l+2ySdqjG?f6VzrE^F!W zYVqVOvf~A$jJJPwyqBn;KVCV~r^nl)b2~j-H}=b4n{G)irfFnM&S_~iWR%lVuhvb) z*_K|V#&k<^bxpT)r}-#@!QNmn+frN~`Di~f)6$fQ=~lpBQ78O`a$3<*dE3gAQ?t0+ z&&S<Zrl3yF-S0t2eYrP|iE4}o zXq-Qu`4^)H+%`%KtJ_5U1!KUH_Ouc^;2&_vX3d@kFLs&C@LE3^+*TyMi}qvgX(?55 z&kr`|C06W_e(%3erQhe~wZ{)CWr(dJF+2U9gyIS9uB7>)0bE6*ApPFN(Jd@X{8Zn< z&&>CyCeBH}*Y`^eu|@%nkKT&Jj5PiF{i)90LueI=J?ZcL_fKW-&HKr^7fqg;m|}Vj z{!dL*a;f~=zcr7{+n_0E-X6fE-VRn<_S@fN-;T}J_(AsRgIw+d4e}Id@GktDwX)yA zT@JH+nOlU(+5!oapU&TleTdL*h7yQEJ=S3E`Q9d7L9MgsGSW}8_$%_xKNTuK%dWuv z&8Cy_UTCEC_>lRgaEDj%mP$zRGtp?gsuAj($*(QL(;s*0HKb-+s zqL#2uuNASSypzpuhj)Vc-R2$3Z}*W#4~O7VB{LzQTvA>)rVA-A?`27uFd-RIzCB`F zTqz?{S5j{LD`dRm?*@p zK5+b|E}|(dGWOG-KFQCLWE$Dx9Z4N#LJe2T@Y@6TU8&r{%_3yvng=@d_=O9)04Cbg z0tOW<1DG3USin4LR{GuMcwCjoOaA$o0nW+k9^XG6Q+bflm_O3FF6Jf|DfoG0pZt|}R7xUI@4j6gbm~V3; zQWin%EQ>DT4cJeW%(>84ua@eGSrk!hfo|eGui=n}luC@f+nzzf=Xpa<0uZDiLSHGu zDs)VcCCpw*o=-^$^E~exl)$*jdc~IFWrbcMwyDeOIMKX5M6>EUvWMtdyy+@OR_1Rib?t6(E$YEMv(i1dj zcFq0usg8IcPq1%=7bo5oBFw!eMF1*{a-{JO`i8janDT}l6Fz>}apA&Y;~Q^bty*+i z&Hkyz7H83Rm0$N~(6Kt;nI!){9QSDYa7!5L$TX) z4nKTWs^rgVK=}>U5K^x3+oW|fB?;J{m(ou1dRfu_Hb}Xl?&OVei|<@Z8C~W6#vmJMskeF_rp!!x%C#XJXW}=1c;`DExoim^0Z&A(D&nTVfG^73R#h z&E$|n7|GiQ;2~qonZf)$!q$-qVT+R%e9`7LO%wXN6g|!_`l**&bgYUhcO=Tso#I38 zHe=5G;TgzwW6u05Pg&=WbQavevy#g~ve=U9R-GklsO_is2=$meYk>)Cj&E;$YK zn0J&umeO9KU*z6>7U`iP8_Nw>IJSEYTVMj-67$+=egPFFh&Ob5Emiq$OrnA`b>P5+ zN7B7#TS{QFfbv=1IKvfpOd7#AV!By31)n2C(iejPP4%d^j`Za+-$#_!mmKZ16T+ zBDq*^(wDI4u2OHP5yMe^Xg z-hyLjB3Br;P!G4Q@V-wx66s=AmFnZ zj+?G$7C<>Xdxqb7f0pLjsO-!v{Q2uc!;IfpxX)Q|pXsyNmfUIj{G9P-W3e8C^;Q*m zvu5kCFQ?i)UiWaC=L3B?GixU}5VzZfgI*=C$$YA!fhL~g-mM=bqyZZlD%9j8;c%*+ zr|r@j+OsFAp!-FGH}h(cNt^&+J1!7A-}|DbxU?rHZB2tpLtgn#c%Ttq*hDAE5;)bc z;vRBhZ(yo3^ zI(lUswXUX2I!$T1*_>>pG~JV?gW3|(+B#I8493S|hujd#e=j{wJdgshU+$C(QfxF% zS!Jk`%3-`M`RI+Ys#EqAIHaFwTYDF(O5(rmKO7v*$?u_R2O?5(0(nAWyX`PK&S z_FCWNs#K(8Q`{&ZnR^*o+3z-7sfKp&$`f)ET7}Bq*~!TXdB$GH9Tyo>OQIj3%J6%e zEcKc*O`=ro))F4&5M@O16QvFo*v!DD;)3Q6LopBcrhTe`JVMTlI3i^8&-07>m`$P-m778z}!OMTi-p4drD*b~B@uqvt*t;B)~j zK793O4&me3%dX$Y2pS20#ZK1V)(7&EOOwy3 zjE2y9GFaaIW^S4+rsfK@Wl8T?Uw7!n- z<%DZ1L-7L#;p{ip-QjjbJ}e0QWaUxK$@7R<@a>u_TQ93AZ?AQpD!k5p;|jto40Su) z)zeeXad~-DUWst%_?vuzUYEay?(lHajTO{KKWP6->Qk`P{KfQQwCx4_jtGj8R&s|eZYj#y{l7x_ z**?do*X%U}Vn-oQ*ia?7-D}!VAYu1SMX8de8nc^U8YoXb4JdC4i1br8-Q;iqg*82` zKTh3Je}^HfJc+LoHYNHv^F3$V+w6IpltmYE7L|acb;ZH*f00?ZAC46YA!Jh(v3V?R zQ@&OVCsetScup@Ne!^8m?Le-jdO0)C5+mX4LwqDgqSx`bm3p-IpefZL*UEQi!QdWM zEyOm_Nu}1t?WW57Re}5$Iaf}SGBVd;U8Fi~j%eS}U)NFpLC#Dg&^d)3*}UE7KX?@t z_Hg^^DX&`Vn zd$l0&Ll9`liY?*Mz-boLGg4>C3U_%a7gIn#$XOR63%eR|o0W{pJurNaC- z0{Qi$0hnk&RvP*URFsRrp&Qen87wQf0>jDNd92 zy7tIWtlt5IV23>WRR-PGAhJ1>zw!Yba?~6iL46Jga>Bf(r4Lx0KZDweeG97P+buB?7Za&lK$lc~;x&04P@q2?Za@<~(fIeU+(=O|XmNBt-i zd#o5+>d+~ZxFD}(8Q8pc}gJy%Xf!QU|}Er zPAH#DI?QyP@cS-BQYC|e03}_4U#<2MIozT;+Y0m1u@nbWF25=7)e{7+8_KHGm7dZR z(8dFjb5&$lIsN-ZPISJy64%O&y<11RO+B9Y`3>HCr%8OOF5XVcyk*&>x6932J9s5U zHs{04G&Ha8^FKzD+h%L6u;0ba3yY47#CMuys$fMV%%zQ4#@WKRUp-+**E=t9Borqx zFjDJL7U`#?N*d1)Ot@on^|qxZQ=sJ^{ywn}ZXa5SmoxNehHcWz$r+Why}S7 zT4ADAEjG5PZQj#n#o49bVNP`&HBySi^5iQ0dCe%Ro1V$tK1Ux*`^Q0@k!4L?NRzde zPnRezzNrIuWwsxrEswV;x4po(ggli1gLqC{#=+1S*|9GHT{@Jy-usS9fa?ly?;VEG zOC^U2Fcre}*rM>&8gNwxlgKP#^2?zX&-Fs1ApkJ07n}SSh!h`(rS`>2^P-dIK^RL} zQxJG(j>?~3&(L{0d_<{I0r}TAjKmU^`AvooAFS81?WPGb|H@c*(N;u+)=w!-a z>YoXrJrJR54Wr(rhV2+H_1C|lr`xOW4(o)AekDJ@>~6dLk-e~oz`zQ@;-WV<$Yg32$+yhv3T27)Kl=9vR{ z*x3X>s3z?Z{Blo9OkXBY=UBzi_kd?}i!oswS?BHzmLuA>2V3|0WBtOJ-AvuV*3Hny z)kUM@g=o2jWy!3*&3@~EK;>(MEHP-hAYg7b+nb5z=1y}{NRHKHld%#GPSejgHCwA0Dr{{zJIaHZUdFms~mj)z<+$O zeVoSkNdK|jX!7wrTSGKveD7HEK73E&V98Y6SpE5}m^YZgrdWun@sM`30ouw_G_|?I zu8;B1N5w-1`WF~fG&fQgkQ<*Y$*FI=>O0P_kM5=GyDnRw(|kH#+V$*!1{PmJl}`thb4x6BQ>UnVMY>+@&&zR?K-@P1b!oN24WJ2#8b zLe@Webk(;lMla*{zrg5;T`_u*Pfs~`JBSze!>78F0@mpP<5=9GY5nj5F_T?0sgJ_ z3ib7MU9~T!!BirTS`A;&IR*Qjn3Vl^W9H+9y7FJ-H~!e~o5Q}LHZt#Fy`QXeq-Q9V zU$=VJX6li*Xto|~vj#5>U)Xvsq@FZ9-A@w%ZjI4#z4s&{L@f>Hk?g$w<3IS1ztj2g zEc-ahTt1*iL=mLtremB&7W2w!FfBIka=jkNJ;}_8%u77i;?cF?<~B zDq{zwN(zQb#x9qPCEvh$Z!!fX!+yH24;t~C$JBf8-#ROBtnV`V^QVFMlqyusde*d4 zj^WSMN9XKD+3i{P18Hn%!P~sYuT{SeVIf~ENs0PGTo|3V(H2uF)4kYiXmPwe>DPu{#8gH~2!9&lfYw6Z%?IJC|!99C1ga`M)? zcwrxsUncJnHr_9#i*)q{6PzWttJ06of;H-#?Q5vH4c%YLq?ZU^Hqbxo?x~yxLg^vY;bMbNGLpk`!`a6H_n?N2ArFGR;rCG7aiKGgy zy(AR-PIc3;GGd$#Wn8EPTCs$ zxbe|I>>Ev|o!Iz5AZ~XUtDi@>FfA}7I1A3tZ?W4fi34~ldu;axHcQFyotNW}o6)%2 zvh5@e!$Tnh-$3kfsglNJPwFfzU(Fn2+(Gwa5~mXZ>kqueE-4p63|sA~k|!&q6H!Fjil>TsO@bOmv=t7Z z3!A;!NK2+5?RzJyy`N>;^N#iVak!q3%{K%jsE_19=2dqY{(~=%E|;3TlgZ^-0& zmNJflA=L0vmdVKyOaCZI*oqLD#Ol%O*y+K@O~rYQTXe=xPAeHhE7i>ds*FZ84mCt+ zJwUu4K{vtJ{0+RWi#@fDpJ3(I#*&22%ldj+Fw#~Os{A~BQ!sXhX2>3t5so5yftU># zCGs3vN8COvn#4azwabN*hH$_C6xDch7~Qd)8-NcPAi0aP49M8? zgRxnH7u~ebv#p0n-WN47KA;`33*5zZwyM9>XItVU_Pn?^ z?u=Ln+wI!zJps*BZT3StEWj443d1f8z}A)J0IYcdAPPp(U2o3rD+gkBfT5Tjcqe?w zG827^`ISb9p%zp9byV-?eOy6-xn$r ze6Zo5I<7X6{MZ8eq@;5N=S!9p*Xe-ou__fiO+eO=Es!?~$WN<6sB(jT6<1^o!~y^$ zJ!D*0gWfU>P9R%llxMG~&ceE`@^fpq)1fvLxovgsYf9We{q03{isY}AOpsH|fW6)> zt=_MT^{#{e1AIFcH3hj6VcT{>&_`)iHw=u%9&1s+(HXTY!e#YnIPs)33N!S+SUt?5 z?=b({qV5RUF9#txUu-$XM|i^912>a#o!&Bb;UQ?bMJR4#WrgB5+qjbcSb`?-z0V#7 z)Jdr(xe7i9T6LQtjgzY)dO|SL3jOoY@{79IAYn%E2u5y%Z>$$)M9rGeSX)`UvhMdW zvYqGN_j(Bv!fbUZW7FEogd67Z;c!Ko_XzG(2EwgMvO`Xq@YEMG0Md_s1A;G}PZk!fi-?Nn84 zWR;ql0M)&gW5XW$^c0ky+(3!C*z*=Z9gPPk8!~*z&}sNWL!(>0ht2X*@c6EPA;zEC z;xW}55G*Z~*9A^lKV%FE5ONF=u>4E zpr1#J2IyJBX>O-zC(x@lk&w`*DaCozFD=v2_;Y4d4^tbHDa{_5H&W2pC14;^g|y7f zA_Aflt<##9>A6_78xSl$3a4VFHA*%dfOQ9d!4|jluU}xJN(_S*gbUe&a#O z6AWKos@?wZQO2sc)f;R~8Tt3I3q_63yDc zyS{9X5BY5K=6wluT!VX|(nf!V-IbW_ZInts`lfyKP|hQr|8o|MJf`XTvb+{}6W2%c zdqlWTFamo(Xa0*81@V!Sv;kQv<_YUU)07n)vMwxh7F>U{YLNO@1gks(ElFI?>tmM6 ziu+P)v~wnHl~e|kf_RBnSAc^OT!PufCn&2O-Vcmpafc3K9d)o(x1PFCL@q`pm*_j; zKX_10{;tkKZ1Fxs@00dEgU9AH<3{=#Z2h1YWqz(QJ+2%_I@MCI)fSAl@;+7a&`~V8 zVnmH}JLDEjl|%uE6_$;T_d7!atnwaX2V5OmX%tJ-q1k=uZVwoA^D;ojSwGh3qwN(P zTg(+bAj&6RCeA&$Ozk`~IB1!(U>A$%M6i4rryw|zxE=&>HNlzrXMH+-&lWo5%zRLf zr|*$_jWhEhA<5ljH_jHFebLOJuP^`E;3WP!(|4UY!lA@fe9O&EzrBN-$rM*Qh7F%{ zoHDE%9k*Q@=-BM8i!3d~qvrr#-!n5mZ^|O~t@t=R0BVclgHL<*L^E)vaBZG<^nGG# zSNDA8ZS+in6UfUEU)`&FKFxzrenR(?()SJ#Y?}Bu_kM1d`*~gNAMJ9#xXZn!c5Zze zy4-K?axWH=TVBj8b5FFM!^AFMNV__v`PC-h*(uPmg?$z4?DPrFPTNs8gYG}ww*q)m z;Im$Mss47o+F-sLRHwNU!%0;|Uq>vVL> zsX~Yk2raAPQ%k)QOb4CjXT{0mj|*m+)&1<_=<7hUY2i77sS{nvPsW}r=TYxvRul;W z;4y|MMf5GfL{7p!BrR%ysdGr+1Qy}(lIh3mXZUWJLrZQuT0o*~lCp^ud_GMdJllGL z{r)vch4_#G_WNr4{j=u#G3oCg()SN{{{CoctpHYik5rDi07 zGj^i1UR$xnLb-V5l|->cb(g4sXOV!&z-sxx} ztIN}cQ;WcJHopmixiqCk*L}^{YtrtGCS+HqPjGaXb)oyOae7^udS3YcxCsx~2$e%) zlilY|ld+_Ig?Cb+$$I1jvX1D)zdVfd>D2S7HG%wMY|69W?I-$2I_THI43S5wOqaJ4 zOv<*2*3cTc!Re;J4XR&d+NtiBbfY4KK%KNur5%*_T5y17KG0VENpK7s5B0H1Q@}PY zV&!b>*G(sfPz7+{YX?Vl{^0LWLL#Pk3D#q=Y5df|Hf2kMp%Df~6bC z^JyHM)9y!z^~gnoD;!R6Vbe;xqc$6k8o`MzFEDS-yHH~nrCytI(iCd> zWhS>i4;jKT?)+G%o}Vb5M{*Q3NY+Vq+FRV;p$-Ht!k|Fi^a~0J#cf_EP$3z)|F`D)YYul@Gg*(wv&jF%{}?fclv9wRmas4Qkp5nETv4 zEh3&^&gy2v`yTNblBNV5^A%UShgN-|C| z?i+}@9k9^gCwQv$Y!;swIQ>*YK{7XMHPahlq0qEQ_Sa6(ON{V4WuSM1-zmY{U4nj1 zWi>Ana*fFwRnFBS)#pR!RQ&iVnHs5(`hoxn!DO7~DsKN1Z_>Fun zfLI3v|8Xb_%o2`mAA@z1-BlBP$b|U<*5khZ9UydkMh+4E2RjxHP9HvK5RXpXPq3-? z;hVua-PUO#<*og0#w_6pbF)o`$tsd94W9J?RI6c0k0XC^u|8vbh-0@Ex1X z$fep;7(CwTzHxy+Jq7NWq!|6BFYUCW;gjf|s$K#Bj!MI?zkehRdn=&Qu~!6j3m+er zHiPkQrBH1l>{|V|r?1o}s70YcEp!d2BY@4gUoM>dy@Ix z=@#*= zmXR;v6NBLs^x=E-A%jOgp8z69n^QvXQZ9l)`}{)j=>GUP~U_f%@ zkF_17_N2e5q0jT+zdr5`mXpIY|0vvbuqY6=&-+4H)d9nSWX~TB_OpGPV27r!*_aNLaa0BJ*4IU7MDzx~YV(ah;m_t951 zbCSfc&1D+OlWEm>jjK}Mo;EFC20oo2WI?zBOM0{{RXunvW=WmoG{(FlICdm4h6^@U zz=P$cd=4K)$~M=$4r3$~E#n+7Q?ZU+7Lg?^XHbEMdTnc4Im*Lw)i29(PN7ovZqLSu zwL#Os(r2?TQ*PCXx&#(D(iP=Q7}KwDaCS2OIv*Wn!ZbxPjhj(9Och^i2FuM{c@65o z{{pg(SD}ONp9R|QzM}pGP-UM9$K6ga9ne!S;DVE1RjBCEfzi@~LfCR0BYnqi%?s7^ z3;^<|0mxDSGEN&+Z|hb5GSrm8H~cN-Cj+|uiNibofSCr$2;HV=<(oOs{c51*Pt)}q zJ{Ev`LbtkBrxS2@Re0NtVcInY|&JyGhMZSI5H-{^b5x6=Kz7D`uDe@3htJAMJHZ<4 z9}8?ebT}R7sx6O7#PSw0eeHCNlLBflLdOuJ=%8bO%G$0lIZ&jM}iM#a{;;6CA6xW>u2>r>AD`rxF{x(aR&63trXfK?OCdQXW z&lx|MxVZ)=Ad=dd_)f$df{)-iQwDpww-no)s=m!1G$O%`_Vub+`~ z_Rx0W0AdT7veVIfy?;^gXrrGsJWIph9ZAm0MPc6Jy=dl0)RugpzP01{x+NFt)!mLV z%(xM(Qbu9cl>u9MGB6NdF(cI&YRoYRhI0zT5q}GCddgr8+H9pNw4a@P$dGx4BeK_M z^aWp4dAIr|?fCkb?&F+P4z_d*smuWBg>beI?(Z8lMl&lH(WA=v?&_z#yUloPv`HIn zQqexzI(WSLRzu+J+&x;1VM;~bxVG_k>=*OEp3HcJ_w6TMK;pB@m-pFB6To@fpcl9D-Z9bUQGSEZMO`0x>df*?3Bgc zwWG;FyPxl{O@_t-$K`lzu^?#tU6QA7E!2)sCHts9!LgFiJ&Eh{%)+jiOJ`LM-7g`8 zy}epymN+(xhH)l@Jt?+n(uKS8fdrIMrz;C0>87f(_ilPao{pgR_tA5?R#~!&DFi*>#9rZN1c|rG=W(j(Qf_95!Z_?vpcR^v!AJ^RsI##Eb~=nP9Wa? zl=Dj^r~W=|)HLc8qMbgzg@VzHH^{YWA!w2Cr?6dE%;yS>?l2vS% zeNru}lKWeBI_=oZJo&~{Z=j;g)mGz*3oy-hmASKuRqj&JKE{Hda-t4jRCn8Hd?%|E z@yLegakG0eDzA{Y{jhX(5?@{Nps`WD2}SV5WVipZ{ykK!{U3PL{?7%KJNEx}Qabd1 z;|Cr4Kd?jpZ^-T3|DSOvP5nQmt^X~Y3os{|i6^bj{hrQ=ZT=dz8BLx(lQ5C2i|_Z* z&&X?B4dK0XVW_*Y?@xBG=*!Zmh>G%2E(q+zALSs%5%QeY0F4^_H$le(fbB<=5y0N# z!j_HztL~aEyrPF18vg!o49NbT0ojf|)X1Ja1h+f|rf+s1<8#yDHEOH_;A@~wOu=v0 zLa3`{Cr+(0a1^~ooL9}UZTwT8$Lq|MaSHZdWGA03*4isIwOrW9ou|SmX%=QFaqc!2 z!wTfJ4~1#GAK&OGDaE?<+AJCMRufn&Vf?c2hXhv zZBJd5v0p^QvP7mI>0Jh}Z@HV=(*GW$1M9ckrETedR(fn%QI-78o(Eut zekk*YOquEj3~$JYm~_|hkf7J()%5l124fSKyyEkbn|aODb9ts7v5+?d0{Af0p7IFY zJ$xnS`eYBEkC1D8C|cax2J4agkBn{{FBbN+XTR@_-eMlL zntM)KD54!O^pC83Pk3O?yXdG6PobLQGu?cAPb&!V6E#buoPT6xYfxb)r#LG0|KCNt z2pJfE9@@9@E1IH`m;;1##>q0S<~x*=pO0`(?v)J6cbnMf_TOr{Ew`H+-ZaA%5I)#@ zSX%(qJ;SxPpPFUWO)SusO}SEe z8LJnqP$HZH{qmk=K1K26d3;cCik&ccjyG~JJ~d*#;7d~-1*XYetCUz?X2%<$>0<7Y ziX`sb!kDWP3$((h)xi-$T>RL(nrX}Eb$%VSBXGWc(g5RrUkt^>Yw11q+~a(lReh3H zP|-0NyRg0;_Plk6WwMhjL&4< z(3bUlkmdaz#@{B=xF=GV(eaPrp(+9IVU%vJUJy(2<(IDM{teS@$AEK@rUShqZsymj z`=Jrun4#K*Vmc^2x&4HrxJmLFP5jx%hhV(6lT$46Ec7-XFWoMnbJqDmUl9;t9-*64 zYT^nw(<0+O0++N^GygCPjn!j@*}5kU(CP{uIE45ub$j!w{h@LS3Euin{F4$W>iSA> zcF%rGW7Y}+j{gk(N!X8WXJLX6Ypc8Lt8^XCco^;1`#!V>vcydfG;sx{c0>~D6UIgo z|CMMgQ^1>V`U@oHD)BpYg&8(o(Jvnq>0?U%hmzd2eG0JE?U6j{?0^~xJ{?x<-l%0) z6-k=YZD{WGMjc0Fov5CTXDS1}Qae+p2YMdUM)Y>_a=4FA;j(s_xFH{Nh~Q7XG&sDc zF}0@)r)RNvLr?;YvA%n?k__m%hNhn9@tff!0IIc42Jzk-bWq-9Y4>gTgYZb3XqRsk z2T8k&-&OV=_euRyqZzxVKxRCOJ{9%)~;htalr zL;Ko#_+MgBsfQ;8)fRqn53}29deU(JpoiB4Rjrd!&R|#b>z!_T$W2^QCa0OHzpuE# zOz!Lk16OCIe#DYQ>jrz7=0ho}ZL}7-zC|&G#c})*pdq z%%|VjEYA1`h)*ZnZ=>X@0<5|9Tc{F->T^bVitlaBc<*f1591Fal;9J?7Zq1Gl4#?L zI-}o_FHXtu#b?pAGkozU`YhIjzdfv(qb+9K&lXQ1xf5SBE&U();@?*p`W0Wi2va~R zZsf)s)`3y&EnVl#h0h|ac%YXF$QJ7hs$5_dw6z^MTTKd_bRkNj@+$!v~%~wLTgj7(r~tfDhDDZ6`ipTKYfofj=#8?~ja`wR*B0&D4_1SXRX&1RcLpzvnWvRvw+30fVn~?l{fT1l7~p5&ex9+ zNjt~^MbdfHN`?je8+j|k0>i?6$*p3XNRcdJC7X2u%snQGz85)>jvzz{Oi^K-z zNExQ;-S0k}rwb1J{m$tsjaLIV^IR#RkkWBif)$oBcs*bF~X`7gN+S75^v@Nmi2Fi65E9{*U~~dajipJ%PIaW%$uU;6q1#>WU$Y$j{oNdT zUx6VXXKFtBlKetVW%$MW2z^K67vqUi74Qp}nmh3e)7bx!Uuu zVUKq^@(ZQ<{Ne#=+Ts`Y@(}O~e-r@@WcYe2VA$$m zm!R5fgK8(}_w!|rv28Wwnwo+hE?njhwkW8opL*!?izj?ygXlzDZtrI?uRF`UYhde< z*s=nz?j3)F!JFkj^z)|kApJ3xBmcM@_UW-pSao+R_ywfSp49v;R240o@A6~ChkzF* zqbm86yvT}55``%9w(zr9#{J3e-u)7A?lyb4nB`6B&U>hAaz3v820IbvJkQbhQBMi> zPuDOF>E=w0?4wR)ANA4JJyh*9h)~K+b^awN?KYBnaQv?wg=I_v;wyByUBRuIAr0B= z$yO?OfwpDEZ%5y%1IE?G2dZ%4I-qT|T29wVH}JbxKAl9B)$niHJD&n^%I{ci3%STl=ip zhzAWGM3{@1PRE{84J3^eW+eZs>Y0(WsRGQmlg|lr+-&Zjlga)XspWvFC?bH`vsVaH zC%Y5g5Qqf@_Pqacik7w`wrA67VfHdP5U`ou&T&N-&DP(fT!iiM61I`6tBcDgyAwb3 zg>9BBlhY<_m$w(TFWRLBkD?_FbN!x76L<{kTCo`|=_HdZ6YPWNg`Uo~3Tvh?8~$Sb zGRPDx)@A>O#hu{!4b-Y0jZAZ_2nCzWDvlF&YXg(iz%nIkRI@~g%vP;FU|@`{y4dCl z(GUEAjaR$vw1`Z}HbvrAZ6Y$Zc|p@z`c>28wLeW?9&c?r^%hM>Omek+!-K(eYa7gU zk29D{-}P&32x`pB)cAY(#!zE(YmL8Ajdv+ogGtC|mch(3g9)S*#xs7h`^-L#CrDyQ zj59M=neuHz`imR@RN--^!r<%)?juH|paIDNL_`YeY*uqno1fcOtxC1Wg3WSGqFk*> z4+h@@ZX3+xjgK+-D+M04eKTmN!Hm#g5xGju?-ORZ?b}v- zU9~;0ecSgeZEgF5f6?}h2EGB4&&p$o0NYwG?EC z_fEa}#LKjYed>+)q|4~tibd+l9`{i+YVQ&Xf40ZKN<#98;X!akcI+v6G|=1wTj|<@ z&HB(^RSX3rYPXp2vgR09h`$JiNuLcyogwBt`CxxPi=`>ShS!u*+*t=#;m+FM8*{LA z|CH4Yf0OORVK6GZ!S*t|5t?0~(q1|U~J^%6$0|fs5D$;K2Svz>a7968;pIL0C zL|kew&<-oE{~q^#%mbD2{Z`$(kgK>{oCI^Te6U5QBmPgZ5v_lgiP+6}`v!#eOvY(Wa2cM+Ad7y42&kP#<{XyjT&iW1-BaTPjQy&dQ*`( z(<9s~0AMzlBJ|=GmKt=GyNjhXC~>fUxuz^Ryz9^XqlSM@YkK3pM*;piHEf~!B7^nh z*bvv!v}Pv9g6P5gL%+g@ri$z|^#xyNkOb{YODNN?@>*nVStNh^oLYkXL^ivba|*~E zPx^ibmke!33RS*OhIh+b0;&I1>za5GCp^5Jz&!FSfAsGT6U;ze>EyZejsTV%ng25# z!g}PJ{$M+mY4FRqj>^nFdYOm(GRsVvtBzKtW)|)I+2oHrYW_I#<0k(k@&oUz%ieex zc_HK1DetTc=K9`QJq{a!l02pJXriVuQss{rW4sHqnlAkN*6n64It9_nsFYR0H1(7C%b>ZG@?P>#8;EuS9byjBC8swN-m&+ez&`qlV!@@Om~D5qm89t zv-|u0!ipm%yG0gI)m?o#?v~m7_xsdC^3q&N)hciB@<@Tx*`E5FXC!4yC0FPo~P|tU7 zG!wBAdNja24$$Lu)fxC3jUk7T9ef>8B+-{eTLiT6Zl8d@hDS+2Kxsqjd)&!KueeARCx`#$dIP0X%;1=p zA8u##Pm|{@&HORw_h*?7{=gse_A9;pJo7eJZ@MiWpn8M+-&8m6f0Tb}yYXA`Q%LFn zEw6wSZ;bw2pg$Ms&pB%OoBRpJd)I46jQ9T~{PBKE&Llp=u!8Z&$r11e*+=-hgeyJ3|fX46ylQM)-u19^;RB`dYh3c zd9&Vb&Ad%eEm!iE@sA5%9&uV)gKiT>%&c%II(CAvSLXcA&3QxD89i$vEA*hjS}+jR zM^KPuzixj`KkwAuZFa;o@(p*J6ewvy{_(X3Z?*2@I^uzu>xem`6?o(&AMzc{l6}v% zQ5&@tQX$Po*En)mslfDPsr#)30F!TgVg9Y4+!gMbgnAX;+HRiF_#ZMGvT|VbOKy>k{35fq;th%@E9MVqZ#+a3s7krw-cLj2QJZp9W;`-Ay20x>$Lbp%Rb<+E)Q+~ zmsQCdaG~YK%4f`e-^Aj)R4>V^zH4o&`r?RAiEuFFJ&uh4s9lS01z?Ci>B84HvXH@~ z`>12F1L^8KoHd7}Z>#R6=*$Cap2h1Q>Q+r`br7HI|A8RfpZU*df@Vx#dZ<(APerC1 zPD?6BrCsT3pppvXtcMt<30kOj1p@hwn9yqJYL2OX zBYhjH1zQ`Mqxd)YY-mHul08fJ>$z5Z07>9ox7vfJTlXyq@NRW%Td3wuwNy8*JKu7= zwE*k38YNu3=_Cx!cDm?LoSZ$I{v(_lRVoF+j1y7efW~WX*BDi#GJjH;*wS+0>H~rq zDz{%&V35AOL2bbX1RUurzCB^aORV{26N#K#_omi$hJA&*Jx|KoX_3T@d7N=9=AyPs z4--N<+q&}@v`)rn;Fd1De*-v#_$;6up#Ra5tz2YBthtn4l2f^2Y_}DAOE1o@R1fyk zMUT6tf0e^c!}^xj9v%WU@^f3cbzf8Wz4JrWZ$(4fYQF_kexCMl&8Dp`)F}3!LRQ_e`Y3TTl<;l$>yL91kJ(4Rfs<1~#xLHdo~%0w7V%d& zM=R-wQtDncEWZLggrM_8(A>%7(?Q(<@^nVQ5IzJ~(5k_Oq9wZtGDqi#uiR|K`{Lmw zd@YES9kSw!Wq(WRB2Wdb0paIdAcUs@$p}zr+NwLOT}Z4uBm85hp^t16Ps{`u%n;l6 z84<}xA-3C!2z_RVZ8&8j$?IvD3p4MccJFP_6}pHxU2V?jMV!$E{2quTFDtd3>-XBp zsv>(uYj7C zP(MNV*=UF}K0C2@JkBTwO8vZHst}6sGW>0Fo2AMat}6SQxW4$iSy$otMQ|nk6aSZY zxBj_FiMV`1T*o~P%GmpvT69f*nk8%k7R(@dW+7_d3CE3>`% ziY4N4GBFS90JTYSRz&>Ox!gToq7eMV#?~SGhG8xAVyw!u+)Z;I!D*OLA>FwrEK(Q! z#>6Xpi#3*U)=upjKnL!AByN?_7*)nnvK@$hhgB+A=ggX@2oFCIj|TZ?vLvO-+xUs_ zH8^jhFnp09lGG#r)SPfd9qVtooty#=DUPf#<3!a|_Gob}P^-bvPxp1-`C_J4|GHF_ z5a8i}A$Z+I9Ys(g>bVvFD~vuvPGbBs*EkW&#g`Q(%o$S@q$rXcT@YmysZ)(z%|Q=@i(jBP6_D3YWj$iS1~(6?mE=v((M*i; zXQGFJ(~X3K^RFDYV$F=qTsuCUy4fpu&RQ@cbM5$3rn#91Qu8zX2zpBXk8U`zVn5brs-yg+@?;wv-|(Yhf{ zDsJF8GdTCh#G=e!!BX4p$Yk^hxFEMTFPC6fvYZWQxd?U-7xF0NNOyg5f-HVy3?r}gobWSn;?gYai?#!B4BYTzv zgRr^UxUBJ}_q;sv%=rQ5{j^S)=keVvz60y#l<2!1`Yv-dPSnZyR@wVE?eTh(<>M*h z%qm#Vhyw)S{`LWdBbhTCfr1NlhGOeINL2R=6Mge8bo%x#dx`e?lV{qC#Fn84Oq3LN z{70uwNZGdG|K15L9^|3xs; zH(D~$gdQ#KN0u%JVQA)7HTTiKNYrm}`#$A!IA1FDO_=5xNp&yNX-t?#Lywku)|{@a z!?Jv9KQqrn{|NT&Pab`3Viikg)}|i>WrDTIiu=1O9o8qWTxVc+;gYWHdLEEI50=BpCs}YrPLnZy{JJgsMLeQBT&H^a`tf2m_t$R$$1;pd+6O z$4p0piw+A{ON{krz`L1heBrHeM^}(;vfGuf#TB8lr57Y~O7~gE7_A+l-e;*9Zsctw z2%WjZvsDg-)V%@?i&fUJ*5%%>AwwddIt@;bL^StPM4(T-Z$e0kJZC>i6{VlD7<_6L zgYilp;jD7L4UvE9%<4_@%md!p>`Y`N<35iV$+&Jbn}t%KGwBEa?5Cd*q)M7IHDN2L?D`0#p+T*|X ziJ8Tq1Z*KwqHF7f{z)Yq+zfVjhl0J-P8@eJtQ*G!6ys}e;;4Ue57xes0SG#CJ`6E?$4+6ec?ziXJB?Pb$y3wd_pfek7m zvxmO}#q9&n%@>6v|M>tv$o)0PrbZ%nPj)8_`ZFC{LBzo7A<=Ut7qRt3L@J^lvJg+; z)|ExRr3fxi*8t9?lMAW9)_n%gZSZSc8PwPxfh)6qaQj$Wy(NCVv|9ZYQ*X}>^`7q6 zYiK^E@cM5NZv$xiLpo=kmNhH>Y`}=eM$efy)1*9Q~*@h*3HM?f7F3cdepJyir6sXK{*0Ql|eihj(%8g29I|bS>!qsG_dta#4)nw;nchq80 zt&8Tt#s++FteuFNo+rmbiof&+am4|Jw`PFp^3qy-Q@REcd^E@)42xY%DO0Dq=ABQB%!i_m}Imfx^~6 z#=r21;Pm;}5R$zWFIQe+%PtcmG=qxq_PMpL(N^S?Zs9r{1>` z@EwZyPJpsGeIfFyAJSDW?2B^G9#+sj?yVSY42Lw8J!>juK^!QjWtDwUwVV~OZT|d zOt4h5xvGf>!{Ed;^Vs6vf=XdblWF$`9yI_~zWwgi!Am*Ir+bNc2~;~FmZ{!fkMsr7 zjK8YLbT*kjXhI%WAqp=H3TJ6;r7-$QU|*#0w@hIp53vW4wFj<~%9fpw^hyXsNDlX>o*$NVST-5xxZyPJZiY3^D*C3AOOwv7=>rZHbXL#hO< z|M>;4Mc^5~9tW2~51pBsGaa>zc%RO^7#?Bo&VvCrWnz&C5M1$V=+K zM=V=4wJZk}*J(Mpt)F_^xsP?@;F*{USz=aBH%rVN*ZND$o=c9p#1M@Kz>SJ6saN5hsUb%JZAb+&{3to{s%r$Y{m1#^EO#&bQr#T>jV7vI#Q+p=Oa7A55>@jD50u%98>^ z=Zfm)PzN7k7C}-=u~EAs^`x#-Siq7(7)7g(%q6)Cd9(VI=X(z#O;2_YJma@hL_6`I zol*n#$?nq&j@r%%rkw*AL%dJb_nrprhu?dVzyJM*U<7dru$KE|qbqnW{JV1Sx8>J6 z*8mI@^8V(7YO?zhUGY3B7orTLP->2MrO90Gb@gd)vio-x`Yux;{X26%(xKprfjnli z!@i{?7e(;R{n{-%vSdjAo)I_TvGBixtcHPfoabKsAf00~rj(2plYtjQ)FkF!`djjV zKZ2O<0ZaB8_KU-5X+~T0r_rpW0v*=7nf$Uyopff?cA$OTG8RDA7p>Csl!3B2P0QU} zUZg$M^6QTNjhfE=kwIyKF{qUhwDS9(`>nX96%lbW#id3ssR^18p9uW<*b3fmc9 zxsqU~f>X7)w+1C@sY0ALW1bCE;Nut{} zVEQy2wn{QSr~X73cj~!RFE|NaSh)W)7*w)KGO}}Q1g}Jcbv)8z(^+U&ldm*nu%`ChNr z>stzuG=(JhDg*x;k%U3Vy)`>1gz1BLi?VKJkXz=IS_a@=^sp++-C%4}t!N8WR+3A< zCk6%_`HSS9c`t=S`!YB$;5AT8Q&VJ>_&AtBS9kcY1cmp(#=zC0Vlz0IEd1TwRD#}O zYAo<*1@hadKQ0vTrd5KS)paopd_K}P1tn_fK+~wNO0(#UxPsgtOffBY_wyvY#JBb$ zgCs)O>T6`Xt(g`{)Ttuj7k1mqX-a)gDMGUb^Xp>~AgT7PQCmo*?0;+drEx2+n6pP< zNdt+gYmEI-mDA2VQcU$iRw}5|uYoj1F26BlCsoc^z0B|322Cmch|)FKBQdfe$jQ*J zPvHWiIB2U$xaM-q!NRi+w%~Fxyj9bME-YM+p>%3YztSp7{NF}|;pFgB%+JVMe+?fm zEL=#M%3ik=W_~>d(}n+jN@+{I8%MTij$3~p4c=D=zeB(6=MM~iR|M&q->(Mg4W!TI z#ytPt?AW`AC71L6m;xJA1k_5IwnIa!Oz%Bmc?dtnj30Og6;{H&+W<2gJVVyVq)JlwaWrDX5a^hLisA*X|8f!t_ZzUa?oDyMx3Q%lKENA!3t+(Y-MQumMC_j#w|s*_1MZt%R6>5 zrw`Ixr19M%8o#e0GWF$?>r&sf7BrYbjk^ZnNpS7PiB5C4}uZ^@)fxbzh}- zGJ?cU;vpQ{g6wxJPs#Js;b*0Dh1aGkoNdYc+KQ5m70L5+!&94oOVBu{*?Mwaqg#;7 zv1fn$8xm$js;tLWKV8lzLpMED#81h`5$oZ#)-RTn9E_~@>epA&%6&`fU#oDI#P;Rl zr!it9d)u^?UYW|r&IH_QPQ(LxzumZ_=*sXoyZhcSa6cnsFA6`y{+R$7dG^NA-~aK) z-x$6n@^t1&5fQQ$tq1((G@Rsv%CfhuIU_Y8$*Jk5%E=$PDcsUbH+Rt^>){uzUo_S~ z?{}3R?ykg)dS7GxOBGHd@yFvY)$~>v+)u$-fB#_Wx{`7CE=e~HUDdccd-lhJ$K9Jf zqd)z2nq!UGQ=669nsM{4sm(9G@WKmgR|B2j{HFhm+}Ns+(;Ql(aWTB`Rg5mUpB^;n z{-^F)PoGQ=Pk(?vtW3_Coc-Sr?9NNP{cQWS^SM`32Y95o7y@QMbS4s&Ms)>Eg^Pte#PaJpfw z>nsV!4!&7EkhWv%09$Ic&)*tT7~Hx6=h&9R=C1Cc4f~gb7O~fow(f5zSwhKqz&U&w z%-)dTC70UIk*WNwi|v8S?M8Q)9ecM3>923#wRTByzoF{}=~R&w*ZuEMM@4e}ixRZu zPSjw?^G!pZJZ9O})4PW?e$uyObH&u1qQh|Gr^i-?-VHw|Lai*>?!MO~>1o{QSrv{K z!LwgxEC9aIj;#(spcT4W=~Y>;)?;+t%MF7j#wSh&Ya=b&?1eA2T>SPU53-RIT2ke# zjKrET(qhXprZBKB(z1om?Z)@IO>J^P<#0(uB|c8E18N+Vcm0F!n3PD% z+fnPW&5O%vprVA%Hbz40hk{V6p^Q&N8LJsmWc}L?q_L~2`+B&FxI=wYcm&qYavG&| z?<#7`&?kVc`8xEw381j8&rM%n&o{GS)t#r}v1LV>mBZY#9e(a?8V^64$KS=HspuKp zLj8zR1}`e{vY5kZG1SuUAQ3c&7pb$Az^!F#Wyub^N61D`IOISYK z;r4p&pKP|D7ES9PkP5T$^*m$}JD6?9Hm7Zn*B~bMhUP5beoUD%=a|qcM}P~A_hDt# z%~RV8Gw_Pyb1+Ai2H4LDKGShteOmLLVJtIFm^7Cw5SC<5Z8ChcmESP#sZH%zj&+wS zMzK2NGH;A2i%-nbraJlBsu1hn$;trRIN`$+PBspGE>d zha>0DJI8dv$DgOzc!jhY0PT8LPKs+?<|`aK^q=Y>K+TT5nq`6yX6A2F<1YB#iPT`n zb{zJOF;!JaWyMn>^bE%k#s2J)Ca(oOI9+)6<7kzwtKlTwr*6mY7@Spoxt%y3;U;hB zs%SDlJDR*C8)wqh)hAhtvXe8@<+bVCW(*58maVz{4BU*u=>{@u-VT#%68nc%l2+Dq zdx5p+d&09&vw6Rx1+LO=`8D_XMVvJffnG%AHqzDXcSkUdNTInmOp?vT(slw86V50d zRPn)-qPxl5v*QTp>NOh0Sl|VQ$`d##rJSe1`co>83ZjvX64!7(=JI0*apCvsdHH`9+2Gr{+w93wWx1#eBJHet^N3p>S!drEVQ2ekx}8YH*f#8JiBbd9F|+L zpHfuYvzO2!x(DK)bZtyZXx&_dkvty)qh)8nuO$)HO`~{4atZ?7e(RoUHIp38tzOl| z11g6ireeNGjR9W9UJ{`dEJy>+lX!_A<(~AV{i0lataRgF=G5e!Uq65?k}0 zZYjh$3Vt~tGT{|{LFs~ul7nUWzpi;iB}w>EQ8Y1Qa3u7W_Pyfw0s`T6mGfbw&;(5#5oH*KKmkv31(Ts zKO-7iGryY^zm=9aT%Laxtmy2~_4x=&qnX-fD|WS?6Fnx9PqerQ?x@WquzcMgn`uBuxJ!O771~k_Mez%=h{L9dFmAW)i+xf&!{4_7( zd|2gd=T@GANXeU#Wc5fCtbG~6P%_w=XP0~oQn3lW3k-cl7z(vs0EQj}LoWq^BFSOc z!i1skw_)gc0X~5VUF!3@2I%>|sW9{!sA)U5ftqhvbw6b~eGM*vjFCI29{da${9J3t zwzk919Pl%ztB;?J!q2tU^9_Ek?0}y=tCIP7RZicsR{UHg{8V$o&z_NFPF_WQK3xy{ zcsM@5tDd1KqPL&pW8!FXz^k4hXc%gn6(Hz0f@JVBoTFqPKl=sAIV9&Q8PW%B*#b6| zfSloW$hpfPXS8QmpyaWs7_4IuI$-tin*=qQhpzIzrq3l`hZW7g%K6B@=e`2<+4(?n zHQICgyU{9;tsAu9P!Hb1`=+0OIxQDB{Jb!`1*z7YhlLJh`%tG5j6v&27tZ6xB?rVl znQ-$+Oz|bSF0}F+XQT6aY{|C6-Pjd1LQY*wq z)tNFn^N=znFVK%<<>5>}5b4{xl7A82yE(m`4bJ+JrX%arkwM8y(~)sGy=6u29hvZv z%vYhq?gnht!Oa;`{mW>As;wvc{$P@GNlx!nbGq=q^S9G1>%PwCyylv#Zc#dsV?6Z@X6n{ldkOhK^&0&P}>eyBb z$uS2s?_J-cCMC8Kzn>~26Pjvg95~_(JSnqF`!_@|ZN;_}#Tu?)b@xB& z#|?#TA6}lPTdXaUo3QjD*V%tjBj%3x@a%W8iljTw$gAi%u{f_jKf5NcZj*J-`{K=C z&`o_+nViImTwUCUQPA5nA`_N@Q#l3h*m0m`*;}(tPeoekzuib92((F?L0)7ASyA?O z^}uj?S^Fk$YNk*I3gNOhYBu|Dp;!Gk#ojIw1Y=9beijf77zSi#YLxA#KE+IYUik{= zJqGc0wEJYQ8!^l4g61^VIP19v`vJR4Uo^g5MecuP`h*z{Ze5|Ry0@sAg=j>(^?B0V z@N?M^=W~D(ED(i?Jny=iw5RfUGp-_G=dF_08@Fbo$l3#!$fWaeH|x%kl!-Ow7$+~~ zAdt0evw=li%)dKInnSNIeTg{|vP-<;ICbm2Guq`gXi{n6qR zabD5#|KX+1ix)?n_nZ{-Wy^ftVZOLS%b3#x%lY(ywsUT=F{-nnl(Dr3nbw>+;7BlU zc9-K38Et0_eD+$B7ft|TIAbjorwaI=)hOQi?BpzMF*mwa6>sOqdS5pdtu?f#%%p*-+zrUHl4k8NwakG&=O#c zB$e|E;}Mfni>j?@=7?5*l+K35THqsK1gg|RW8~ov))}mR4cN|J$6$sfih{8%U5uHJ zv!$|0jU4L7e@oN{aj|`;vJaj%Nff5k8p1fbnCHKJ`>K^7~3G7GscH=MPDSr z&w#T8ZNa&Mt^E8Wfc^D0VCP^h71NnO{H>3LJT9~ zas{d9(}nN;zC8mB;4PtUpOC0*uR-zfWp@8&QJK6JATcYpPm}PBFr0rDmW7COunpDW zqnJLU&a1F6v^Wrb!@2grw|qpm2W~Kk4wLw@h`teg_YqxMH*_BD&(&?PQ|yGD&>S5V z@M9N-6K$moXpSv0f>|_KI88hIp%plm*w&+mvw_3;cH^FG7-<-O$38?9+;=_+W^*4% z&r)gd{R~D4?gul_9MJs+I7qVa1ZssnMMJJ2`#nOpv_tN>u()t+Nh!G%_3bk<_O#6X ze_~$d3c%RABNA2zwk6S7lPlv*o=iORk+-4~uFpeltIChEz@3z+$$=eI$~Zio%t~6l zAQ>I<6^vnx?#UIQhP4O6u>&FN9*Jsp%LXoW(~dmXk8z?aEc`$|h_Tj8ZSd6?cTmN( z%@Zq1UWFx8*JMp{&cu#b_HoSyv)NX+gXyoRzbNf5=CM85vBMK?KF6;o%nkC%^U*|1 zRZ%9W(XXwtWbGKv5>{nR8tvqF3p;zUGtNF>6p~f3!`)}hV*4)EREklen;{5x@z)>@ zxDh{Dx*tkMZMAS(i)spGs=zyJB3t{zUS%qXG`=^e^4k5(*Is#F z=TW%2k|lkJ-wYeq?l1}}`E(?P&3@wpN$M=~nQ%MEe6Y)QSPVUzF`4_5%*QQny4!BV zC+a+rzt&;7)-+#~PJbtve+JlsMOv>lXnVBNU2%c88)zg`4aT7S6QTD0GNL?4fk zeCKzy4Ppu9D6&$pUM5d}YN{~xC2uyx(}m~!wmri#_T3C;;yynKY8Z4o(2tgV=!f$+ z0sX9m2&yLhq@dDTa8+N?4@WU-fa|tO?VbL@5Ky!@eIXwWWb4{c&`MFzK`7|?fPx&H zs0Mw7u0@{C2t|?Px#h^2cFfCe72LM@Qri-wwk8lPryBF(BnJb_!J3!Dx-q(L9ZOCkJ!( zUKufcnXrlKx?6W$LMzE|hQ5&6n6RkRB>M7*o)-p5v6picg%X>Y=(-`7vpij z{e!p{$Rvz!flpE&RpTHyDh~kOHo)YRO)t3V1&Z-ayHYh9ncf<6(HZ7-?(BB|c|CE- z64w_Q)kSU@K3<6;`~r7Vpu%z4fS7Q*UD=5y&aDV%Xrh>JsXU+w^KFJEB7PO^h$1TV zqV%weC^TUT!`a;;3UV)M^$>VJh^qeeB1tsS{svE1okM0pg6%*PwWQOBC;L_>!;+Z0 zc{trb4JN(aCK;@Q<)?9%Hlu!_&8U(lpLsAzUUp`EHk{#fYBU`QLtwM7gA(=z`J%5a z(RBKvWX@MXyhqr$Xr-@yN77fNzLh1`jM@O%AK(yW_#23(pVakz>WXq+NPf5J!cD*G zL|@=qYy1_6E0$&-x2u^s&Y+)0q%zAn&v00)zoiIoRlVe$2+#B!6fNGJyl7%X<60lc}Clrj6Tj>@H)uxbwynsECO&`;n^?l4Z{QFi92WEF%n_}C+RWXhHh%ZkHqa9NxH5$FQBt3=Tm~s z7F3nIiyD5e?7G;HTEc0?pC$AGrLosnCAzlTPztx1>u&}H`Hu73l^xCXS7hdg?XhTX zR-471TP>+F^jakJO2oOHt3VG~bu$5t&Gx%uKaA6JY+1jkv$q|yXqyk;jxW?V_vf|Q z@$rJI$i+vp&65AFEctV#QSWCCn@R6v&6jJDDc{MQ|4t^olRf_hKRw#Z81#va;L8hvJ^y#SGeYe7#k7Zn z_TwBP`3MU*h04jv(ScF%=7Q)shE9FK0pi*g~ z(A}IVRPzt-DoVBDQ~2@>d#iMoU5W(p2^c!qcKYIs4pT#r=-cL;<1CrIO{V5@B5TG+ zF^6^QqRHV`pO1W!ES$to)T-PE#|SM#Zo`koj=h!Ui`p((vp>$NMU;!5#zz%GU)y;z zN*DxVH~`m6K5E=!#dda!SdSfQbOCwciBwmG+h~{R7)rZgmbVt=XGhlWsVX^Sv)WGY ziX6xCT%{-`@6(#0)NJ7u+p$&2NnKK36`xQ<<`BLU8`+3W7ruIrxG3ATA*@CeEBml| zwDz^CILc7c0wr`iV#ABVD1z+D+9gY)WuLI;wV7jlTt8~Q-%+#7=PTx#%fHbNF)nRs z?N86a)H(Kj@WUg#aeoEs3H5;h<_nID1BTgVh)T+YF`%rzeAL1|7&aO6Ly zsP=_gCx`pYNVKIi6P9bU`v9IFwzFIih#w|rp^-J&(?DgquDIa*Wd3$Kk+uGSwP>&V z^`Ec+ZZO~PF;%?hUg3Z5h3QVZaO1u07eJ}PX>(=t&~=811w!V`4kG3hL@pl=gfn8J zq2@r949ydEMG|)a+)oLYl#NPCoCYf{)j6EF5U{SXOK|L59R^3O^P8}-4JpSyySCB8 zGP69kwHq_erhJpV9zSfh5bVC!_8%4)2Ynk51U@78@% z8=9y;X}4m+;q;=JG|h`ibWfRxsH2EpJBbet0L5%|fW{8Fpx>$I`TI(?^SYg0Y4%T2 z+CS-0pI4JIlikJOT1D}cc9Z=brNx%b?5sY3hh;0K_dwtbx*APhr+%3p?s5P6rrn!#6x&(miBYxV z$3Urw@t#GMIGlZhi45fJqtuIv;6Ag}RAVHVLsSL)XoK^IZl9omooXPnerRh=^{*fK zrRx9Kz0$AW`y;ia3#*dtk&*S+*&DtDneC}_=V$HTa4A;XLsWdRo%n{`p9pA6b74)k zlo_jXiyeB)E~_z|LXI7gWZwap9_*m{ZtEl#g4ZvygC9kkcya9pVFN0qX)DlJ{u;43RNx-M<1! z_}(WOQ)mgj9U{4S6@}~-qkssT14q z+1XcODl+7JqiSL+IK8s=+)!!t+jbd?%=b(pEJe$Qe9425D7G^y58ge*yQDq-UX(08 zX7--5as>3J-yLo)p3u6ItSCqejAkkPD?J5`*6p)?(2T=*v(w}(X?*Yaa%;i<+1uoB zZq+>mHfVXcgV;JeCKt)+{3(t$;T!pm*^UW6BfoO*9C9UR2Ho=!%KHzxiEn3^uoJ`k z^j_xZ%uDI-nh_As&rX~JbuKEU!iY6`nFf$@_o1UVH9E0e)&lse2s$r6b=t3-Nrpd zCb$d6uy5mdY~52&Mc5~kg^#k|9hq>zzINS2neQ9DB5Y~{E|^r|4ET09IWxQL!0kW6 zp?X64vr6PpVx%b%jx}^WYAIU0{RC%S)r5w2*`wHVx&ZUSxm{z2x?A`DS<@8X0SFce zQA?Wa30lEw%I0UC)Avp*Hj(^j_pSH>HtssPqWj9aW~*+j(k2kevVt*#cx@AzK}9Aq zYK_?vbtdF_*-ZJQxta=_RwBmcD!AiI`xNMyRTWxV)#n9&iiYmb@LSPKcww5BBF^hp zAj?=$k$ANSY)|wt zb=`yeUOOS0%>C_UtSpVY`gTAa%o69$3@-!>7&O0a0{NV0u)tyNJS+Z*KE8|SosX_U~ zu^6G*VGb=hoXw=;yn;hj@nJ-s6u!k-P1&UyCc%_cKnq6vf2iG9n)-%nzg2d}o3&+S z);%wS1x=x$KKt}2krs@tB>y@My!QN@qB*8$ZKi0i&neoK zqPKrv?4!Z*i#4=0p0CF9yTuM2Q}fN(!CdRE8coi#T8FGU2b9E*oPa(gdRJ!Vsiz6g zlcAr*hMHUK37h~O`Pbo4bIy}i>cH47%WSQZT`T=f93vEnZlyU8%xdc-PpU+ zItWNpWN9A5{Xt{Q;%z(Kn#L3$6x>)u(qOut)ikCwE55F#L@l%zoa=>wv1QXTHUoVA z;wsNDjk5TOP7q3a@S`Nndqe;jIeE}dpmEh-)pa~)D`bQ#$2?MurhzV0rb>145G81` z`|@8n+??R7fdcI8Nsld(DX1Szd=<_389UQs`*B&AOycE6`fhOtTq6MOsi>b=Jl;Le zSVzsNZh0fCYAiqI7%$?ITZ(2aubuH|>e1x%qI$fd-Jbf!k+|Vz&BC^8EkH!7jK5QT z8cRSC660V+=phK(t)*FPNdR|+a#;yN-$=w>J1x*>Y=NK>J+N` zGme1jo0RdB1cl5&?F{@TyE`~~ZaAGPx{ZqTQ4PQ9()If7Fc;O=qzhl04We{Nt=UE3 zpH~j?&GJ+~w20LI)1Q<1qYkAWHE&cLak$`OI_LkXH}xx8G;iJXRv)}gb}wM|yqj4_ z&>bebz+aqXCF43TYqRMxBftj3-;^WU#v-fthr_($>^5_SjLTY zu{!GP)|uYh=qYyLO3p>@MOoxHD^9tHoXs4$_K7AdYmm$|1OU=|DRxGUb!LCOSZ5W_ zvO_Jno^k5Dmdsc~rE_<2Jx}%V;yQlJ0PP9ovLfGmyW|L+9Wi%|qsmQ_qGu!1GT}CX z7!;$qol0|1k)LNT(~Nagfmw|FuCYbuEH$!>WpA)2EVV;`TtK-u1$r-%K)f)AKLSCW z^h-v@_PV#vu*>o+fH8ueU38%ACF>r=Rl~+zDie2%u4?nIAUfgw5|@c~mNEg=Cq>dt znxblr=bqk(xHK9K?XBAswJyRZ%s)>q%lhEPUARfY=qX|oz`{M3r(n~jmT5Bl8KhVPaaS=7@xF9vvYR!}@aGjL zi$6d2$5#G4P5c>$j%EcPj<02vuoK2vu{1?XhEcBIG@%SAq!<=EKC@3pbO2FJ=4?oilZE-k$k)3#rp{NkoCY zXD%=>#D9@s2*oPefX>eJ^C9-xIQRPls@BK3bo)+h`&r};2m)!i23ZJBN%7{Jnk#%goP}RgfgkS2AZUyBH7i>;JFnMt0sV& zFNoBMbFmNdGi>WYLS!+6r57fj81?~)EgRMnMfY-2od#Sm-@QPC*oO}WyTzl~QcQI- zax9sF?}gkj@R+!E@5FG)hq#msyP72#2MoDiybe)QH@n8&sj#B{4kX{@&Mr{jZoz$Q zgm1C`$=H->+te z!~+}gprv)HAxACx?%*9vwoW=A#~Ws4nEJ)iDH>YYy7;(v9%(-Pac#!qta>}C%$M7f zxmzz%jDj0;yAPdjsF+bGE=jYfieL+_p)NnZ8!M~UrXr;)-|E9HApE+A7#~Z1E43GD zk!5bDxffq3I$u`F#az(&eSWpvtfSAw?$EwO&u^>>csb!UO^#YoaQrQFJzaR)Pl+P* z3Xd6h2@`x@1|}J-oK-x9-B)#PO5obCnufg#%_kLV${3X}v+i;CzN5+DGw;+^xcK?> zr&Na6HI#sVd9l`f_HU8IXGx}#1a>`@%a>V!w{WatHVMBgeL6FbDt3iCnqukYMB{V4 zA*Lr;PO4>-aeLpyC^4n?3oQaU@s<_4!9@KYi>YkQhWWswP{^30JjZ!V0%j8p4pBQWe|)&eWi$dN@}M|h9;OH| z?Y++W5!e-uq?c&{z5IXdy?K09)%EwEK!BjZ6O|}btwFC!Dr%yjMnQ8W(UbO~IF>q6 zvBb6xsEI?Jf;Z7zueYXFTWi&#R?(KW)&W5Q6J}?Isvs)hfae$)6eXyY`@BDUpF0x! zwa@Q){r>vq^@4lOKKtym_S$Q&z1G@mtvxB2FZcY2`7-sZ6kjm&h4}#F8@)Lg!|jau zm8nou!1~fUiss275Zk<8K{*oK)cA$0)GH2|V^)1yN9I@NG#2D8X z`bywr=Z%?P0THq#<&^*pQJA^>5h@KQu_1}Cbo~$|hMBdFhHGh4V#ifrw;oq89jmlm zzk|l-(#f$h9w(V{N zk?&S>37ZqbhoBatMO-P|;C-JcD*t63rRiVKku0>LNG5@P5hcAX71Wys$Y7!YjZ0FB z)iKgqa${q%>`#r-oAMqsq2nxHB-LqN7jt!9SAYGtXV;fqB+^e&oTVm(3+~uN8;L;f zdiC6t+%?}^gT1?$H%QKoo+fVvH) zU$qJrSd7;67i5m;R{0fN)z03cb4mKKmrrL9RJ(GEg0oHj4GtPjX~N8xqwMR z-^Wv>cyNbG8pq(p6N$YpRSLZbSAfv&Xhn4XI?T`4g&OOLjATw-FV~tZ6$>>=i_L}j zZCs5C)y#Kcwzq2^$f!)(hS5Sh$ShypsVfX|i|+`*SxQskj#MNXRD+b`HDTzrF;rZ! z*~A+;?z`@U&Gm&_gpJ#!5ScD#bS9^@H@rxr#sDq!w@{;9oP-7Qq|MczIteTV2uy*( zbh!X6e6^xWYSJF4Nh75sv9{`@@$af6tv6@cT7h*)C#@$&H0%RYXXFQwIfPAM6r>0! zs@z^k4afV{_dfMIt4Ra}AfU4Ik6@2)9(y<9g##s>GTd?lxLV?NN=?Gcj^v#Ra%1Sx zITXDOtJ?DwGYo;uG+r7K>0oL=n30>tdwbMnxux$?gcf-PUwDv!*P){FiPwi`U}4x<IWZO06KO3k*{{|kJ9e03r_jD8=8RA&2~ zBj;7B%D6zz8yAQ@PXLgAP6{6>6iO;v@urn~9?`=&jZAr7A1iTdtalx6i7DT}Gbub5 zgxqX>huwITQtu3E1^UkE<_k5ZIt~wc)G`7$aj!iNIZKPTN|foERq2I{?dX|P!0DS{ zMywADzM%@SoM*c|ZRRCOMw&-~9!1Ad;c0fI(~Mf3g>?Ot>Iad}5+# zU;iV690W$zJC>Hz3KVU<}A`zeU0vAHw?SI41w z@ic@z%kWexgXk;F9dm+8mp}$8G_jGekttpx>%*JhSLf%MJq`J9)oC!5 z-6eU$u>sw%gQsv3Jg!plBu?~k_!kIIE=Y%=>?)UEH})gEd58B3S8r@ujQ&pin=xQw z-SEa74Y}c8TMk{ft>OJ3*a{nl9v{qO6K{3nRXFaBXM0=m5y#s|si4jdYJye|CMqN- z>OM-CWTK|EU8i*i`oWIv(A=nycAOfLNTCXpF%WX37}MH6-EdIkfUq_QN30tT|N8m- zOtja?<5Phxe+1|;KmS0Ffa0nQ5SIf)%p+SSpb1tNq0t~D%K}Q!NQ7GcVDq(R zT2lX*b3A5vyhC^xnQ&dGgn?q={h>14ct=5|436Kx63Tu993>sE#f!n>{Y5p+?n(`V zuQX!bRq=^@Em`)^1s|e>7sy989Q%CE*?ijsG~zsrLv_~PBxtk*gxww`l(xLz(+mxN zZrZ&lRP5bMf!6v#Gb$@Vb7H`q1If8UA2@qtjhTcE}bN}J%tr{jHi zc4#-raVC$?m{Oa^M~PBD&E)YJgYB296-)7y#4**!r}-7aL>iB96W8FppA7Q&XkVK& zJ_@C(Sjf&k8KSOo9h+aOmr3B$*|f>D#8yP%OK$9R4#9oDsWqQcl%v#AN+zKX1xfPx zYz0g51J21eW%BtD1%?-c%{TRF;~y|M$P!CtG{JY}_t--^%;k_VX_e8BFw$*PE}u&^ z?$e77Yb!d`HCd(>JJGXD*>yI_6#D$3cxO^v8x)%yQ!2&>hvI{RVz}q=O5t-TDSRm2 zEhxt4|9GYFIg}JW6zAucI+XlU6ldp`>a|^dDbVUqhmz@Aztgfnq#phE4o>U)LMiMC*#ahG=ul$%@FQ?B zzQmy0%#}gjK9p#vpv|8A@T`r_IARArs97O?>gMp%Bjdf|r=Ybg3KK!1{H#vxQnUFkU7y$yq zI_ChoUE9_eB1DhzMI`WEIm~eKZPDc3S6&xQ-q?6mG`Ww{GM!hV$+Ml7zi}}cDrV-M z16#j;e-3il5=n$lNy(S^2h|tiOIlm_YIAi#B=#a_^zs$?-uXgu?>^`M7#XK>zyV=RD%~s)+XG- z^d;SmzH|eHtg*_hNN-7rJj`qqF6jDxz^8~}mYRy?RQ-M(FGyZ!!grHzWO0b<_$K;hp5f!ltOz_=b}+A6n@*q2pX!q(|&Sr@YoIYLP>qnirJ zV?7*{UIlgc`6v1p%Tuo!XZk|ivaiALI^sUcI@c6ad%uDkrREGjwEw1x&TBgAbhN~q zv>k5Vc7W6JioOD(f?@1ScYL2R{Li=T^U3HybL|GTg*{xj^qaVz?2YK*JusEw_GR7aBhO6Po!WHI|*5x+%F z%bD=xk$s$&kEQ>~lnK@=pDhQk-g%rNS{Fa3q!y$@wevK%Tg7hTheL-kwU2H+)?uaF%MKk80 zq29(ysJnB0F?=Gf05M!2I1h#>CXM>ItWi(U2S)uum|oVXYaOYJtqVuj9f?C!xy_>G zwA53%mi@}xn$<%N`64;q2nMy5RKST5gGvw;D{yz%nz7u;2xBI%E-J9GHj?Wgwk|Mu zI{!e|Bsz!`cbd*xjnvJ7@~lA}F&!N!d1nu$kTE7&^V}QDCZMn|kldcxh^Pe-7{{7H zc>&S;!iS=ChI*?H*_ISKJ~o*e8xLjPrE*i5JWXdpP{{?UOkjEWvBK3GnvlQ@9;xSj zRtg@q=TU%F1KN^>2A9^>Xnpb}7Oc(<$el%n zZM`LT7KKEG5M8)?enC0YtmOvQ3?ilL26uqS0?NXZI|?et!2}Bjzc6{NDoyU>o(RHP zr?xZPe~Bs-6!h|IdDHMOPSH~w|DGHx&wR(~z<1a6BXXShz?!zhdj76dbs4Uqe{Tn62o^4SIn!6z3X2_TkR64132nqZ;|hK zQ9aAaM}^Tvl}o*F+fhv$oY@Nr>|TULRb#ei$i3Sj1Ai1c<`&MwI+I3A1K)wHK|rat zzVh*NoXWEBb6gMjQMNGfqbw4ljF&M+i86j=iffoiZ#dCsxTzzYsbWhlA@+}Pmn`c& z&gv&W4xcfG#3LlLAs=u(DGQC3JNc8lnSAm7vd3Pt6MKYgmuZcwiM@ykFN;~VgxaVG z$?SWOWEvJ2?V+Is!uCZltn$!V^H(skwTx#ona^6<%mu7W`TCFos@6Au)yHWaQb_y= z$|B?0w6*!e^2S}m)fboCI6w|o#{JCDGiGeGWw;w=C+Tpj?gP#8YWSmg3}y-3q2^}n zJ>=eEe2A^?sB}BKdbjmzTJL|%_fzzn!rPrXT^Ei?+vSn!q>Kl@(ZU+H9n9iod&7s> zo)doZ#bW4k%Yt%}C+e9qdmfM*U=tS>n$^mw^<;)mfnJ^heQ*}&0bdE=E)?8Ly8wGb z*WsrB+I~%xHSNa)HFyZNVh5tQ^!XE(a|ts?4_^{}rvPd=<_Tj_8U(j5^y{NHkY{>4 zA;5|w#*l`fk(X(jtn(~@Td%+BNMdA@bF06(;gICQOs45Y6nwvMYjR=rf*W>IqUnk4 zg#B}OrjRn4NXZR?X;>&p*%99r25s#T zpb>mw!AYyxpGfq%w72jx;K%wY+y=_Dh^6CM zd%#myi~42}qlE#)6cpHw$faSBuU!`nQ;J1ga%vAOLZKB8pTYwylgl8#2Jhk*nwbSnW>z0aZ*CIR#50{7u_;nAy_8A+AWmU!u>G2M}BUt3@}j z#SPI)dI9IH>m?FMnV{Ip}M zwr5Q7!jz9T3t`WksXqRr;`)kDQVc2I{oot$3p*+a*ij|+Pw^vmv*A5*R`?J6Xjs6H zDg`gY`mQK#koUtWhOfwNd5BrxFsX|bJ@Fi|co4~4En(ijHkn#_h4)1V)&l!-5hnlJif)%zMq z>tnC6UBkK9rF;cRn1!^#yXE1j1<(KKX2r#@q%0v#=)9uzdzT}xSw|1}bZn*f?lhRn zHI^B+XgXTI1Nj%3jS414%AUl$D0G*yZxc)l&e_YM{0Vmjly8FNpnS_;GxT4`K`Zr3 z*JqRed%pa{t^E2rF*OK%Og5)=5rFjY{UYXNQ5pa?d*u~7;cKJ8cYE*MX}rtt>NNB3 zAi6+q`XgI72y4X8woMg_xKuw*Edl&BG@qK6GA}VNC9d~PPs}y`1Or)tP3VDmoZ;}S z4culp{BhfH_!55(iiqh~^qZ_VSsS>`yt)0_P~sP9ZD3hYFsw~DVAh5Ij7PQ)z2hEW zEdM5kocnhEIP}Pex@HSe_N&B<4*$j=>==H2m>7O+#HsBF)~D_9hH$$S$-6UhVehf zY0!Mujb8<4y*JdMYYtAFVaiCok+6;##%>rK6quM6TBFAt6S>>|+el^uIV;yZtd79B zU>98BGcBf`_vwDN(pXg*r(-U4oT6%$@@tF`Ew+dj=TjM1uV$W*pXyFI+_acG1f)+S z08A*lDacXP3=l@J*ZIs^bsyEbfE9JD8-L371iz50^(~fGq(ggJr)&p9U_;oFlwD68sJzj1x=O(gzvublB56bFXZy0c?PL0}L!-p*!;RV>Vb?Rpw4Js!? z+U7A)RVZ^w?;qHe)Zcc!r~3cLi4-Y8?Rql)sSbc9aRVd@=v6c%yfD7 z9I6YgfLV%I(DpNl1p#0}^{4`Tdvb7C+cJsDVA-PHjzdC7#g$;O>A2ag8!NhbmuTu- zL{byf+7mm4NZP7Y&8MiDjmir!RrQ3*2;~wR!C?6q<>$RE4~eT-Sk3;78-BW-u%ym^2R>7y}KF1Q%}9Sg!2E`yKgqqyA`mGls+T) zHGo2-(rP$uZ;3HeD%w)Ff5obj+q?g22=@T=?>2;=4$1B00|)oq$(YK6abpNIl2#Ll zrzXPg>ZQy+?B`|;;N!6SoOy4o+uu5fksfEeoTu$n4+HOa9>IZmotuyOq^A*_rt<&T zxuZ7Fxw4d8q5c`x#XEK!CHpd%I8-s1>lwuC!Tj65|33}piRFSWn_d&&G9N?K%$|?` z8rE}vXQ@2S7{J3;0#ojGHh-v=B{C0aMo8OIIJ-?D2wEhu9=2f4u z4X31*jz;VXtGxsD6nJx(X5$Z%Wp~tmD8E+!4iojFn#- ziW!NNNra~IVd<#VvM-jeWhEYQeD?>u*Kic?NeO&SbrYkCO>1v7^w8S5VlDBYwUc#| zD0|Gd_0^USXuH81tkziR)Zii1fai>na$Ewtu#LzB2dA| z6FiC)FFK#fpmCGB7q459wR*i_VEow{l9z_h4xbY~_Z$-^AK=BjHUDZnQ`Z@Z7JU6} zJndWQyusUxOl;c$1T=6C0IIP7jm-wA!Fy~Adr9oY^l@JYX@KS?u7BLPGsIfewzt*H zh3nhBjpkyuBuejnu4%QwTQUJ?4yK4Z?>+=41LJP_J!{-AS$4>AALe$qsf*FM%eqTp z`k2@CoUuo|Gb3BxBc%tMQ{mIdYcCHK0_~xrF6cr$mtCf4g%6 zln_2q)LdV|OSr*zLdJuXpnjMF;b%&qj^kM;@8ad!GTdt=MRi-%5rAHWJ42Z3K_&Ss-gU&J=npfa|x zy2xv%-6$h>ZuBnLV%eIcfUMc{L2tS^dA=wmM-ZeJ+e{ZA`~D95u+J1tBCC8|flnZw z;k$|3tOB7uof10rij}WOGn0IU0bZgQdATh{J3+%3#=q0;zqDyhBt9}^Pyr6Y-JJ#H z+FjbH?Xd}+^|+1R_P(x-B`es^wYqHDoLAd_No_&`b)m_sNJ1MxtlJ2?O$<1QAFBYP zHnZPlF>jH+z``dwGyras9xVl3+g{i-uTrWVv!kV-*Yq+?>M z18;}g*sG}!5Mo2m>P#f#Y6K+&PokFYQM~h{4*AQo>Xg7+QV<2tcLXV49l+Ma4VW@8=R(dz!ptKEnQl`6!FH z*7itTn-GGaN-3x-u+GyE3F$-y4adotkdQ8o~eU%dm&!ryPHVy-R!WXZQucxN$m@HN*=n z+f$!218ei`Z12x7naZ`*E1lb>&|~AxAVRg?3_tABRQT=n=%u4nwWGRPA>yH zso5*UVzH(940`=ew&vRN{T#?gp?UlJ^nCoE{T<|`neMOkcc}Bl|H}^IC++XB#6_10 zs4nC0ATDnl9ysP(tiKl?53G|Vfycwj^w$25nKklmNN=(Auu-N%!Nu`AY7@=$xR<{V zF9qvN;%t3>*)N9+B^EI8>ueysm z7zKUJ_&QABbnm5R35_#n6MGiB-5JCj!{_k&`>xz74?_yW(!9;3o~9#UxFQ;HTTXGw zqMGLjc4FPBl@}3NldT^Fd~Jz;8F$IDuMf|5*dm{qJK=@M1Qd<0!l&c$WRKCpA)bA( zkC2NupnK%{-xgPOczuyT5U=Z6);H}h$dH*2EZT1<05zSA#9x}V$0AayWdUcM}^bw*DJ-OR<*3?D1V- z4Xn(l#^17$81IWs8b{QA2OISl2UOy=V+%~!b}@{re!`d%nvb7bg0l^xWi@RTCl_bu zy7~p&EV`9kS7W(|b#||G+{GrGQ|4Oby3z>=CFlv!FNxZg^gsDa8@An<$#wOD&2`n; zhtSD>h;%wY4AbS%*iyx8yM-I<#q=N*id#*J<6K{W7R;`>ixH&b1X^X9a7!gw|~tjnvR-a zZIgr2X&nGLY9x>(U-b*P5uF2dIsPvveN7&s#GiPL&GoihfJ{{87QRw3k3J1k$o4*N zazT*Uiuk~buzwaTu_Kt`8X~IjnWiWgCy^# zQM#NMG4+F~aG$VJ&lk!EV!LqKzV*FXzP0<7>(F8H3y}Z?yEUK7MAs^8yKy^5j?G#n z>C{j$#nz{_z{+x2KjBRxw@WbTDw6wAfG==bKhnrxYk3WqOmXS43i%XaDr^;Pk@;op z-jYJcdOzvMs%|si_~Gup1?6gDzCM6BsRS#EKbApf3XOZe+)I-Ytb0G@%4cd?exuKr z_-)}G6--`bCjm!;x9{l~84{yD{+DbhYtE4M!XzJrJhPHp_30;|l-!CS?tTWm25&62 zNq6r54ptQK{k;5E74J{A`K@ppO65Q|Ip`F`O)Ph$Vm-bM%5RnSi%}r=G1yAb+~8T> zt@jzq?K;rnMP9Aef5zV%25;iGcjH3&t=6X=8M4RL;(enuC&Ui56i~(3U$hJ19ZsA6 zdVp3wfnrjU{rYWCVJ7`wL8jeT;NXHhdW#gU&%|(J-TTm+{kMSR48}J$8;Tq8-v2KY z7e)|8SQ^Smw)WvVO<*{fGel0SGB8j9pKl1UR#|d4dT$fAhRu!x0&Kb=T)b{TkW9r% zqgY)){OuckARCi!C`DtR(^e$*wSOmg)Vwxq;C6VSnf!hcxHRL_;2n=;$K)0<_p-Qe zT_-#BtWXE^iLyGA#@%2u&8S+b^fc2<4ToG%i$%%*#Dulo?D&3K|A8IKUvI2eU6h$0 z;&0|Nn2*&+;Du1`&A*YMi3c#iQplX$@(;6s{-Tly%|EB$7eZ}5C^)C93Fu}b7@y}$ z{(e-==o^-%M5piFBO;vVcjxFMhE-_b6xIMizIn!B^I{Wy)z+B7tirt10 z2fE)ON{@4g`Lx0}f~-*sEqi3XkyD&7=tmgX9>bu|T#cYfpnY$qjmZjB-h{77+@&@V|HkZD$X*ZqN+ zdl-cQ9k4#hq*LJ3&#T2v(XThStx?!>+j!bBanRCN6bG$e(`y?H?e{f{RI}b^NgbwN z|8CPZto(vt;y|(edS@tB1bqNYiWqJOFEFpPiR5_WD`R4DtEj>&K~^8+n2D46v4{KE`mLt&Lzuc9t6Q29rV-06OT7czM1z8LrBvv z44+-$4E-hxpxq8%!4-316GhfK`zEHitHQ-;FdHaLk+j)B?Pl9Fbnv-O9x2|ewNx1RSWZAC|lz}?$ zfns{+;vmlC@0z3kOhTmHQ!%|Aj0&SCkKCkQhi;}&zL3!9nsR=O#rdp=?FpJ2LuJxzn;g`B>wv7x8*+?0wC|`*m#MElvo41YJ`^v?;{ zano{b;_5D>A@GOE-12LHdaqBdq;R1{M)D59jRuhvPP#O)KC z7xJAIW}9rGfvghU;01Q940wwST)@JP(Ixc6pqH5+iJOtEjp2!=JYNj3a6QpEpUeq& znL(OS%3DAn#7uPFXuXpqPoz`=e)hY3_6;cK5##8vbbLO=xf5qK=A?T$Rm)5^=pTg3wO2W^pPfklO zkQz3fw_I;nL@eLhX7IFX_E_L^> z5e#=iKLV*`vD{O|OfBQS2>134BtvxDL$PJdivbo{$h zZHi&^1@BFyx7JF1Fwy4|ZZHIFaw_Vk@bB?K#Psk zO43d>U)N^}#~ztH$}5tKu^&p1p0K3(x-yl&%pF;sJ<^9&})AFmF2Svey z9h{b*<~-=)P95CQX>oHNEDIi#uuYotU}o@ukl@{N9!%jb`t?pu3y#ew=FQhXQ0d$i z*drhgO>1kKf4ZGZ=cXkRiW9Rn>)qJw8vVrY(p_}6hPM!=sdwrICuKwrS_nr!Di22& zjmv833bv-q>&d(BHjQLHW!noH;w;Tx+r#ngZ*$!5lmMde6HN~UU7)82D+vB>H*s2@Dc+b>+mhAyiRIle#_Ctw;qlwy z&;3Ypp0H>eYnqWlv2AYfs_6UW!f15fShv~?&v$LMZO?3+YDsAur*Zx7TB(g~0lEX> zb%_DhpEAM8RcMBka1;}P;`Sccsklb(+H270%nC&fkw=8RWDbS<7dlKlpe2HC4Nu`H zLfn*%{7IVZCGl2qW8;ENq7tlH*)G`igq2`|Lhb#Gw9f2pjj32iSYJ+FRb}~a{aD(3 z+$_}gJhVvRSKh-bwzb0yoP?B5zQ%o#3~RZ}@O!hu{hp%cud!nN z8q76+4X-I<1}!p68*?hdtlI~}%=tmbXUpMA1t2e9Mb9hzO!&~%&dhvg1r9xWHkiMT@E!a@w6mYN z;6#<)^e&P}Xp)oeEfRazsaqT@r5HX@7!zUxV_8|0X|hkewd~uSCO}>KUCc=Jn_$9-d~Ec6^sWRZCUmOj>shoG(1Knvbj(E z<0ww!M~h9t3|cSnJHZe6+5ao<1YhZX7tV5!ue`bF*$7;hF@w_W2O^rye&Efy745YF zD7Swu(ayV{=aSj}ZHKT{a+MLA{k{OLek1p4?3!OIkKH#OfDwMJFzTWZ2J_Zwm3D{( zx(oky;#0K4Rna>uCbk=TIElvUC#()nPJS-Z`n5!1ckeqKy&mexFFsC~V2HVc+`@7`HMcc9FD1-OXvWm!u{l&H4wJ?x=q~ z_eWwQOCDllwZukq?Oy~|zNNl%>q)|YOBuEOS~CdTo&2d5hZ3{urk!Nx8i1x7o_2!At7=z|GE?i-VwA zo4RH4C*wpJN+;y9-JKiLZS0HA^*cmA>}{a@ zH=oYH)_MWMp}cDCG+(to2g=x)JX7k0fjkqmB^KJ(H+?&_e-M;F`v6c|N)pU_PzwK$ zdwr>!Ag8T4MU3O^U>y^kp`G47zZQv|7D8#fgs{0WF+b*WxyI#9KUvOM-T}@_+B`M zZEv;IL+`#eHFIZ31CDYyFxW%d2Jq}5i9%9aec|(rKLMxX{FGM`r^(SDhz+zBINFpk zu-sdLU8(Ca%cmwYd8B(kLyNVE8w#O>7EP$|>|)Y;Wa2|q{&01LmIesskzYT3Q0@HpW^Hd6fY!krt) zz%*db7a`^N6~%Z&^*av3X%*pMkwQ}Szfr$@L!@Gph9&nf;4Muj{wOi<5F-d~^#1f3 zCZp)cP(c%)<0t^&ty)&q(KRJAer|j?JJw(uzY6c97Eq7(?Xs-rdKRokmBHitpgRNr zq566lSUA@0Ck%#}F9GYlp8FK}G;#V!ONs)D3ycNpKYt*_FiGZgvpv;l9M z>w}j%H{BzO>WTGKL$x`RI>mYh9LviLTk-lHCNm+9Vm%RF)?}8yqNefieAe?slc5|j9`+%UWm|R==iySj zdd+mXAmBS7p=ZVu60OF!V?2H6NKcFhQafclZ-_Tqz5`<%YGDheN5FjkNElw%!)bZ! z|DONI+FYLIKjxU)V(}j@_Hml^&|`5OvhifN4sp;~`r3EoKi+8plm-72 z?*X`c=5ulu^V#PmF&_-~1x?=*_c3s{8~$@|7Jp$T6{i`{t2Z(RFre4c_%)9L{_}Y7 z^q&7G{O7Y>|0Dl-Yqt2$fZvG!Ts1Jwf9|+7pZ}!K8~Dcf&dk{d>O6f+A90o5Pbq6%YkiA!@TY!GC9^o@dW=SL--@AVDiX|}0%!HJ21hfjX512u6B zXL$7gD$Lm$(T|P|_kYuwzI5NyW{r>EHfst^b-uR?Pg!P`e|+?vvzjT1Ee$(&Eb8=k zq(QuR_FzL6=2`W^WW;I`*!{k&x-N0O3juB4`8ec#&tiwWqT-qC+C zZFkU8bkh%(ZtlW)-#-KwU+1=CC~sO>z4(UL z-QEL!w*!t%)7i?BoHDS!@P)2j1ODbFaqq_EqKG{P2CPc%!OCl<^koY<8vBwJz%Z6P zX7dyVQo~@2q+pSFVO%7-A}&zlO6GjiFV+{nTyGL&en|dmxLJh#(xHm+|DLue#C99Jg1?&`MUrSbt)HU{DSTj%nNQl6 zG;`%*GV)4Z@c3HRX2TG$1Y6g1%T2V_Zu4v|M*qsovl(_S4HY>%hti2Q*Mkfa)B{_7 zvd#MVNc__BlWme!o}{JIi-u=s+uX?(UN@YiHOKc-rc9e|lV}1}z7D)3jwgLnD&OY2 zsq(ypn+sFrc^NnVo+|G{JAKuTO}Y6yTi!UFvf;h_Te!9K2(uyI=EU^9-OqUzFq4T} z&n3h<`sIKtf10>;?}6VWCx7{UcCfBIHF4{EeRQ`Lqu-TviCdrkYJWdL2LyfCtbUrm?%2jWH$*nzt&v! zf`1yrJBAQ_y(q&JJY+3{r$53^xck7~(7;?$zX6ELKB%gMW1X!XjRQH`lB1tQZN>E1 zspU}O4bJr2Ocfa6UDJ`2bpIZ&IMb(@_F{`+YA~~U%fVVg@AEK{*mHLxb1vY8NQ@H{ z-mi;6sY_GxjU4|X&YU;cRAhs(9{W@0dnL@agO($orkH5Bb6sEjJK=6T8MRVRGK@4l zd7ZeOQ*lvf{JxQ=##3d)BUGvF|5-S;p>ypnPQ@(4W|8ub<>7$G?E5KxOFcG0orxxfzP$W+IR5KTbQ!X@|a-HRrMjp)Ll%H5m`Dr-%L55SVNps4l z)MT1d{!2IFlnNznsFhn|FOb$SEETN%cC>>67fq^5%J4|!_}WSx2>Z*%`YeR5x*Q3szse2=oY@#RRW)M ztG~oSh3_%oLwLvVN$fad$cqJ|>@bXyi)C=jxrST%ufWgk{4m)z?R0p;Oowd)t~@XC zgqaT8%yc+8@r0QU+st$rop{1bhizs$j7U6Tro%Qf9S#Vl!``B5v0OJvwzQ52M|Iw$ zTVV1}<;(om$6oZ$W)6~H@m0--OPHY2mx^Pr=~nr-N4= zhC%4X*9>Wg3L*aa9He^b`X8^o|zrH`$bGjb*GS3@7Ju7>9Hj1<%koO z24m?aE{SLhy9N{0LSn$aa9SGdKupv?;sjD#xQ4InUst`(xh=w@rZ=iPuV2Q}_s0?p ze}EZcr7}ZFGsq`qJYREKKBTTQy|A`=P17g-AwVw&n}Pl-M&DdowcczJgQArGhw#*$ z##0q|dJts^S}Ck?rZ1>zS+8`M2mP9|EHv@BqQ4hbKFk`OUH&JG#bROg)pMO_7^@)G zX#Yhdz!WC*)!g-e1z z=-jNMOCqtQ{@xVhPt)(if~;2`XJpf_UR8`d>fdLIQm@~#63lmfl+jjXZ>W!LWqn9n z9qMDB+pN6AkAmtPXb&P_T8N%MuQarllCsn%8!Y6jAl!*~#m zH_5rQ^-Yj%4ss+@Aqo?@F(RSyL_uI%C?b@#KE?sYO6XfyzkFk)g4AMP)W>z82r-=a z8gr(|&dR05?$x3&CySV~D~N1!TK@q&%_BpFP4!A00Y$N&by{eAgLn95NnQ(;!S*k`7xH9c;GV|Mq}!`v@MxiN4vpJ1ao)DeOuJow&FFK{uhC^D z(pfn%jy*AP-M-%-`xbKm5-JY|;o?&AzWQE{Z@+qhR)mXX*KvVd4A2pD?ozOLa|hyo zV$Pdyo4}JROD1nSZ)ef^n@9^fd<;SLbt7ODOMHmUi&TH@M2B&$I5{NF2+9{Lc!Zi- z$;ax~yKW%&^FbZg%$j|;~(g(qXO zz@>CDLTN8{8n=Ffu5jg0y~q|B_a_18m4ha4DEP9FW3PEg*v9V`!k6HsHx#;s-zZr9 z6u#xM)9&jcLvbc52p4AZUtvAe=^Y$Tko6pap^rODvI{W+W*~^~Yza-_BSIXi^{h}c zH}W|&zllCE5?bq4W_K4cVhY2mKL;o69cW zAp0!{hKyx^hnyf_VPYuB*64?n7R>JdCe~LJZfP0&wUAWVb)tnYo1WKz{2u68krp}h zyn=dciq$8I_O7pL~Qd=B^}WxJ0FKfr%| z@-@)+6PIITLJ)4d*`@FgO+UByr=-UVeBgf$BB3wCxyp^EsKKo?wsT+Gh?)6eCis&K z78Bf`%cGS39NTZv3Ur*XxO=%$Q@VH|dn+XSOqgz9c!9 zb02%RmGrIt>he!r^xm-=zjEV%DlMDM7kG#)@f!L;99GLb$bt_VEj32=*v9RK#v%SP zpy8(7)K7YB(tsEuUdck&Zt+(|tc$^vF@h4QA1PepRxh~xf!vp3FZqAwB_ofaBoPhF z3?znSLD!;`yq(4`Mq|xkz6h_?XNUTSGYVR6{NI>Hnn9av%X+1rY@w=ch9>Kk2YJQF zZK>Jk-^?gu2N2)U95N!mJ*g*QC!Ua2>_0)Z^gE*=2lC$+usY9aUW&MtJn2d$HJ`lg z*vf)}#-VO>K`&niujM>GH+X!8dAtk4YGO_SGkkLLp#x#KFLFod@?^=d1y{~W%-IAR zk0$$H`On0hCFb(jD<8?{?#|7J@)>_6-O6d~-kipMooOtZZp^twl8LV~I5NX;@}V!b z`6lfC1YRbI34`Q@3wGW3FUye_2AWL~@Z%rTV%zw=*rLvjeWM@sc3O7?tDRmk z+Pc%an#z$ly^5bz8T|&3@27y><{O{Eb80=Q$o>+fBC9E=-f-P8Nf_#AmV~i`g31!3 z=KNho*SoLX0tvNHBh|}3>g8tV(%s^hlsXm6C|D~;{4_qT4z~^hW)fm(SwFD1A~<44 zOEfNZD@GS5*?Nw|59@LQzJO1I)ibt-;pSpKH+jCsC0EKZ^9laYY@WQwTD)!|ep_HG zhzOA0!a_C)2)viO=B&+s#LBQ zb3D%mjSWbaO(7z#^AQ&pgvL7+O8s&{sI(YZm1 z#?r8vdsecK1f)bGK1S#O|8pRjQ!xw(8wKH*EC^$?9yD7N0w^kbDKIV_uq%)nFzG8F zjtI>2eF3+@oBAh!}N6<`;wV1 zcmO4$Q?M01CY;5d2p|Dr@zYwydRw38^DmjTMe!2jweG{yFA5FXvY@8%hmsMpfDlaM z0?Z(63Ny$!{7=lKQ!yM!c)P(nEDMrL1&JCat4r)-GyDI{s{-cV|6~CD9inUP0LCM? zu}`;fObpC@rwcT2X;$#TQ-n~!6jSVApkN?$+kRcp zKo+NkE?hM-dM6YrbXIgACq`Q$_}JC*r}e-T1MpnDH~-9TsW&|Dqh8@Ys`f z{AkdQ-x}8O?Z$7{RY0OsLNop1)Syu5qa>C-L48>YL%0x2^7p4sE!eWs9lXko&g-MR zlmot3=1y+hof9$b0l}MiF;fb_ZM#X<3_h85v*vB8oI_XrCLtWz7lL`PnHB zVX(z96qbWt`-H`o2>qF@X$}g@CJ&BXQ;3iFTZV;d6*0_wlbHDy@2x)o#`U~u)?wt$ z^}fQxgZb%9e@kGEMwc=?RUV<)D$0H^9#MN z#!GLX@1I2{O_AnNgZI$S_~4bY=jBTxjr%9dPX7Y0zx)RrLYx@&@PiVOnl6_+->8Eh z(anY2q~xXWk0XhZM<_!cITG-Radqa-Fb3Y%e`c~3^fDUM%G=4@YW}mjjJWCwMZXsE4 zE$~;Cdq?zDCg4j3DZwuvb+iz=&MhzOid>d5QFPoYSAdX@W^WCmbYojWwlO9_38~rCWv>j$`Yf=UH^>@Xa{ReJ9JfAc{#?8 zOljhW&o_U`>sx?|mtJ+Ea(#3Oqy-60t2hCNh)_OX&WRgCQ(bnQXWu&p#%G#$H@Q)- z+#FI?-|5O~DL1l>yUd0yp+bBo+#F?Sj2Axe#@)G*Z3U zX_?7e5zdlgJ|)XGf1+jTW2xGjd(|Fke{hUnk}O-xlh%%=1!Dk__I-&XkSE`XUxpov z?}&gdqA@d)8@LEYvQ8uUt{%ib@gJ6))gJx0QuJ|iF?bWbMiLhq0;_n(WcF^?G$2|o zW{JaJ!%OvvrV(43|dKN|Y`7OU|%v?2u3k zlEDuo;wJ7iRHy7*#$ahB1@6X~I>C2k){};%Ke=K|;TKKlYX1x*A?>70k}bupYyH4! zTgsj<0d8#TjGZGp;>IAvWx|sB3G?0FMQ@OQr`(;eiG{~I_XtfVo2&Bh0l++vt+;jG zn-4=`U&APCX%!d;^raIT3*Z%}eIt z*6=L&m-4!ba?7nL^H})VOr3M!R-*>^?%?)<_u$h>;wK z5Q1)V)By@kI6A+m!TV$F)PkAjtOMJy4iZ{I^Oyg|XkU_Bjp?c~!ttnd`ACp2kOo0u z#}YK+{e8H#GAS9b8Rhv%^g>(Tg#`z(JL^hY?hu2|LE%~I zEgofJe*t#GhW{G;MyfxBe;NEn>4?E^R@pR_*>fb)^{$}|SwJYB7n~!3ffM{L0KZ3% zo?0-b`g65O_J+YLWiO0vz5KVFB`@4!I1vALp579uN(*5T(G7P&k**;>d16al4oCeRqc5C@9!;quYvKik!1Z;LB_` zaFp^ysI+E0H-IArS2jFnTm?MZShh?ZG{X;0qvM6!yuUmIm-xGRR#(gpIlRusmNu1M z(HC}ss}{+$aH+5oNouL_PU-jtfzrGumdSK*y@SPZ&QF)6Fa!sik>WV&pzr_=N)atn z{cZ4iodt~h^DdMy7M^}6JC9m5YKYEVtYfJyOXonNT#kNJm0^C8;VeVHM|AKMxt#o4 z3?#TcYMnsM9vLEqgVVN(3R)QwBVNT|n6>;Y$Xb51SgfuDV?nETn3uG0c;nRefwV0< zr#e1X%i|;sl#t)9Uka|a1t5o^Z`KWeZtUKmLXRuvIDa{jI)-Y-ifU?dspj1&hK}>8 zWv?GE8;O@%_=2G$7zKt7k?#04F0;-i9W^{XI8AYOxdj}ORreA?3XID{FQ<= zPV^)SEMs+J!{D-OFoV=$iDu1Whvvhsk-@P%=xB@F!Ngha%$Nkn@}$d#sm!v|bXfsq z^|6;>rs-pO;4%vhH(fs}IGG1AmZ%vu%|fdMYxU)~kKv3qu~eOn6d%T-gDIze!WUl2 z!I~{F4YWe`YwV0U3ge%5xeKmN-z+@Kip0uaBQ!B>5CrNu^{#c48WZV{d++wxSv`$S5X`{l@hyYZZ)+6b6SqN@-v zH+#eRn9PXi$97B)#-8SV|BrC0$uukg8P0$UlEKVr8x1HDF*)E)t!2ruM$CK?`*@z1 zxdfYks`*c8-in;wOb!wD_cR|FZ5I$`X>c4D3y6*f%w0e%2lRG+mIdSkjC+!~c?Af) z#E-$m63PO5n95e(Py4edLs*3_%U1qzsu{@DIb!TMZ}LoIX1v3^Ev5{=p#Eu0e3rR+ z_tMIhJcESk^PO~$H6_4r3b{slpNTy_BfTdsFN`6zPuSuu{Xt5Ax8?webmU#}4~M^m z)#!*Du~gG2+D&>mJT^b{)CPGX*B1;s)rXUjWR`$=t1<4a8 z836Oy7O&%EkW<828;U~Vd;<}k_scMa8#mBk!Y2L$zf`k)*t5=D-Ds@HAB{uDQguJmHF&G{zIwWwZhIVu&@&MG^H@bR49^@v-H+E( zS04Cbw%cm_JLW%Qt-_8nKZ|&R3&%@ftbQpvBE>ZCY(O zTNO6SHKI*Gh}#Yj3>}U4rwH+T-QV`L37M~9)mn6;{9+B~upY~M`2#7p3$fn^WCxdnGnY#- zkOZaoGU=>(>0hi;qiOoT%3R)(x!mzH*y~DGf*#fZJc3~|mHwv&rvI>3L_2%?FaOV# zdihf`z1WH258&M+bGbItp?xq8o5izhiv5{mev+O3Pie(m`#?v4BNyr}$8n%18X)fmA8)GsD|Kx0J!}-?NrU!X`w|QQz=ZR6L+^)*Q zdaCTSUS_C#)FxA>xaT@!Y@J^1YUhKqdibc*T4C!v(xy5cJ=f`D>pWn-buV=?yN9Xx zP7~b88O_K=7r^lXcM06pC_q9;+#4#;q)i~#iziB;K;CcgzJJVK+?_3rlJNCSdzYI! zm8xs?6UkBtbKVG|#r66WUbl`mdM+6?wumzq$CSVkriI#g8i_v@d}LavM)$F&kS=I2 z`b|Hlr3Ilz;EXq)I9oz3AHFL1uo4?)K0IC@o<-odKHQP_;S0@&chZOT)VzpzY)Xf! zIhLixDwXfxzY*9p;%g{i4_*6scOn6|bXadrz1@6WY4Q-K!ye#ee1;BKx4akgcS`W?sC;5#xrjd zpc~Fiwq(+QHh3@mute{BRrADuF?@Lk=!%xtq>yU18?cxf^v(Q{Mw7R_aT*8n3v_Y|gUzsk}@cWLgY0nGFHu0-J z?S7X!IPqLf+up|?T6SU4Ju6f|VFY5=Lm;XfT{^AV=tG^R2h1Hg4jsC&IDF(f%nLf?T?H%>Epb;bj3^<4+=m&J|66m%sI*+G?<=3r0x@=+&F^)K7?m= zReev5fp`s%F60qc=22=GnyK5J8=FF}@-yr_Ka`qMa$Q#o&3s9&B<3Z6MTbJQiRc+20@*rm9T zofg?+C&FfXAsD+K*b!q!7y>b4H&RV&V67Z(=Yr@|^xDJNRjsgNSMGI_I!NPFaKZDXbh5$w?a~Lq&4^btYB6>*S*Vs> zMlR*ASgup#F^8E@DoI>l*x*fG3tXVyz^z87S#BZkfOW~TQ~vcqf&YpG zCW0qJ0W0xmQ6S#aG8Ms7nj+gAA1BN9Q{_7;N_Pl*wmS6WEw)2v-)1}X6SG-jS$-9f zUYG9m1gr%{eeGYtE5jK)B6T*;i>suKPRgtH-4&# znnA_Crj65 z=HU6)k|$&0u;*`)JCo~O-Pjeyq{?)o^NY{YL5Q@+JF(-3a@}m(H}{pP`>bzSKZy8l zC>+sW6&Ez_=tiGy*E`N#9clOocG#1X&KmnwsjP-UID)H8msrd`y_jPWW3~d{?;AMw zyESgL>o)d{E#XDVujWYmt4UZomx9du-=Y0~=C$ueXQaVw{|@g@&V8T$Hvei&K~pik zCd~qU%|J8pyR;Y1ZZGR)gU|oHKR89&|BH{>f`b30hySIAfYtx+d|*E0EFP>2`S;HB zTQ(!Lrq`c^<~DQB^G|u+cFQK!n%JKI{+vN9V1Bgo!Gd;!Z+m}5-u;BkeYxtNns+}s z@BXm7`+f87cg@_J@AuBT-`tT4&-;1z%k%E%W$uRxz9%#Hru~QV?thziAI-bJBJX}e z-uwn(;^1SdG}Z3 z-A~B7KQ(i2=xuc7-hA({y!(Cg?sv_*@11wQd3L)0roH#`?w9A?&&#`iGVlJOy!+qg z-AD88ugJTfkavG--u>vj`@{0?_szTCHSfN6-u>oTdHv73U!HeAFYo@zy!(gp?thzi zAI-bJBJX}e=6;Rgf0=tTKBM#Q4@=*Z53JNL`~E`(?KW-%`G4*CYYR|WzIGEgmfFm| zCv0l)E~XB4?z8;Q#S63ix%>RH$NHme{RVG7^?7lQCdyg<%;)rWT83!ndzbb2{x^FC z-)qVLUbx5luh{wx-f21CtKa^6D}SnE-%ZA1{{Yi?8XWT3aGn>>P5*5jJk5d@i%k!@ zpUvri{Dg9Fh;FpC5$$bJa=(w~e~@)PNxr;j8jjys?j}a_&LWwhXYzs&B$MD-{@tee zG(^(J&d9)<*n3nq`V72Fd7cCBg;Y|5H3r@ja^Ma0tEPTd@oSHM)vJkA8~9!H>)L4b z*^|d->3`p${?7kp{le_}bJhT)|Ip0Td^+7-w1EA#Snk4jmyt=E0)i|&#kt{oJHjU{Xx6%bT<1c5os&VHZB}NV-qhqqnySNoyGkCw5l;Zkv#>`6v zhVUve_oEiE=b~-C5f5_F;B7tS&xBA|d%)?jFKk!hd#6-=laz8N1_(!&beUECYubio zbpK{PR$zm&F&+TfXesXdGh&S=lcl0*A2)ui%YoqEGcr+wckYY_3%UlR?F1T9A*ZSX z++@$!A?KR$O+Pbol^Z|!d9jKH@BEh^B-{0+b>!PX%Rjx;jg94B@2)3JJ?t5;Kx~_p zT#uvLW!0#DDTTTnl4S#D8Jv|1XAI+w;7mU^Isx(wBEeu_;b=Dl6+eM5d>x4&pkY|b z0g4OA(cwDBv9Y!>SvFZij?+Q4EMJ&tWeASWD^Hd^I!CndOUPsNn)>Ru8+Sd6ZMJj5 zXD9cY*vT9;@M@&y1xd?$!3!|Wu765DBfh#+Z!3j)@g%NNlf8WLgIUv;^Ji0$N~emL zE_5s2fR4t*7RY_B`c+}VIbkLJevUSJBBkNRB>By=-wiE!gmnpf*M_j_%_jDr#{M}A zw$ylFUrN-0z|GdibgUac+}H@K>D%a1a@1CDYGj|ciP0QGH8Lp1Pt`A{QYRZPRWN$& zN2Ttf(PD1}3;TfzV|)tM176<4yevCw*1NHzyIvB%sD7vMN18SR$(&NyI5b(dizbz9 zl=17q@9D)0i+1P!=q|D{Ge5z2m+VK*^uBE*jJMUg-ptS^$(koYF{$}u=%a@tert~0 zR^!}B;;dvNM{L)Ivd-$gj1H8FBQWyhll(xzIs~-c@VSP4)D`2kP4U8eT9oFxaeS=I zsh~wmL}DYksdT2V;sxiKVdExtt{;F`b%S>#)w}LC%^ajRb3WKOBw1D`Qp_oB`nbWn zZfDu{=XEV)vNw3$!>1OsyGM`kp=XBdywXO_CcXiS;D{3lt#qrq8+U6y_6Lo7CChH_ z5VdHk5XDrtf+U4!fwN}abBL$zGWd$#IH;hh!i}G#L&lp*y9Ng17H3oMg~q-$M_yLk z1nXQJoGTQGpH%ERb4c4%-;En0Lr}7$K1N1|7aHGk<0MyVTLRA13sx z;jf(}eQ@S1;+^{HWlf~y?4VW?Z}eBw4Sb@h1MMPr?H~{LCJ?ZiybGmsB#1V6 zGoJ)~t*e^Ws7KI4{`wdM1;R9AUq7!`MAQx5D~p8nYX=di2+J(5$6Z>ROz?CzJ>tD} zFYbjG9NE$E03~cI**E}Z;r$g5R0~%7OHj1F@y7}wjJ;zTZ}1w`Frgjsx^mb^xe^L3 z8f}{9*EAXHT^~D#ZOLDXqH)YjHkRNc;hk!_HkxKRqmDh)X__{;2ih6T!oYUdhG!mR zJK!Tvc?5*KxH+)rbfQ$cue1<}*p4c_1X38!@~B~FL|D-E0Q==@HI zFBe{ZN~RY;t*ANti|ac->}|F+{<%Q3ak>dkhL2fD>sdo$i+&gzc%-`5YJGufiv0_P z3hdtXv)N73-;E@0*Gi>7hu*`^mqZf7UZ>{;U4MdM3}OGW!MkZEHWGOce&ogvI{tZ0 zVJ%rx7bCTatROa`{I0hm!@+L+xegvA%jU!Q^?3L6+I0Ri+47;`Qlgc=lp0>4y8TU4 zM^VpE$@7&;mYk9&PDL<}ev*A$1+|yJGwNJ7G^-2TFX-Oc5sPwJN3?Ucj*fTug9A{+{2KT z0X%}dOQDx;c?82p_l%umRlM(|v^RR9r#VIaQ-_JO<2a3c!#gH*lAg{9p`sLK-q#{L&%m;q|@9xF2U{ZDwmG5r02 zXt50F9bBhmBpZSbduPnZI|T>4lWohIVC*OE3swQz%R6;YkNUp#9-RrMk%7@~sU7bQ zmFM-vq6Y^tJ7|-s0#epHJO1KUSt5_y(Ntmfsn?MBm(1&z!bd@A69@BpjyPy<1nkh5 zQ@wM)DrtJ!!{1re`sf;L|3b)>*n#u!277{zW!j9Bkn@$R2` z6#x3*x40Ay6hPK2sr{DFK(MSFn#gZmqlvnb)^#5IVeww?bVYvXf?XX)HoP-yA?Jf+ z=^KB;Ejo67LF?VX+@>ax+DnRM-{H|a(OaB4kW!ptDdoLAo$B8;bb1K4HtrG3YnbL0 z>6sc^Q{zgQg)Y_XYjb!*YS0TAG=(p@6;)68ved(ukqG|U!IvYv4umhg_cHj>-OPQA z=ALz;>%3s%W4R?g?Vt4r5e>dZf|ZJ=y>AKlp0T1I`=^HZ*B3cx_aD(fYv%V-_>I(e z+Z9!5xzS|{t_>$sm*Uslfh`G-S)lQ_SLEOw{>=f zT^@fFFY@hC`+`*av+~=w4BnIRoA2=EU!LE-Tbu81Z~hH#`}zF=f-#AUc|suqT=V^V zG=8<)zTUa*>qfBVbyT5_G7u~I*L%n9$@gr&-Fy%7ww(T6pg6C;TlT2Gn^XN2=JaQ` z?eFI&-&60*-|q~5 zAMoE_?f1XaevcU0d*dlGMTs34aP=A3C=3{09NybADe&al$nF7xN-P_ar3d}WSa+;N zT3KyX4YtC)Ym`{FgGmA$SsK8~fPjPpZj_T}c%D^t<1~JZsrLVJvF*r8L2{Sa;9k_J z8eAD!*!s&1|0jL(V6F`B49FLHwI?JTbDb;ea)pE;o)8im{yZg*Qh(>}UfxxW1ZjA- zylc2C@9LnAQ{u2(QP_rGAPO6&D}Y2{FYqK$*zKMuEMMMh^hPKn&p>VTQuxP4)+U`S zKXy0v73#A6vCqvJ`*e3~9G~n~lLEO`_mY$`i!1N-#*Y^b7=+Y1b_NPUV+p9)j5o7+ zQKFyS&t}`TyPeG|yvnS^N>xOgw6~c}>VB0!AGY?Mfv<8Kl)@iPLC%EEG!qJZ`h@Ct zKcQgGgpTnhg!ux?h=m&MQX3IwULQs}gk}c?zr1{}Z<*q7KZLHNdwpI{p8Zm@^1a?qL3~Iu+JZR!u)^7) zg2Z8z@Qoz*`3?gmOFz8v4P!50;Lp1__dg4JEibl;0lQ@BC!D?y3L>N1#!nd-jPHHe zMAy#1USKbcU$o0Q=c$95T2GRhX3N`$)B*sEslG61(Sx(#@H zf&GY|1fcExMm?!vnyq-x?!afw1RKQwA|V3gSrwRvOa$N%5zZ9VKPA5h=xUT` zxVJ0tPs#7|#QG`ueV%A0CBM%T&A9UW%R1+YX4V;?^Thg38JI(HY2^8T7r>pAj5Y&s zJ-g5Dm^j?Pn*{vdHt^PXy$f%#A=TI|T>Tu2m<@WISY!LoKT-=Rq0JAdZN-QC&=zB$ zM?6T%_W5}hBY2f+lB{};feR8>8@*174!L?Aj4|DG)|N?UohC$_gO?aAclA1k5Dm^H zL^WysI0=Tq-M$^Xbk>_qmm)`x$@WrE+H?sY=pH++jy{Y2C>2v1`eXDwY4U@f=e_VB z^T^MB#1ZS0Pkq8s;Gj=7;Zd$BNJ*bO!Xvv073UEi0@In|Ji;RdjB4i*o_)RIJi@cn zz(BZl!8YC ziow{B9_{m6^X|QF&3B#7M|2Bpbuh-op7U6%Wyd$QwiPWLQGuD;d_OpQk>wAHlNvu| zIJ{Z6ALsIB4L}jQv3vgPMfkJlcgLS8(5P56!7nV@edA!!#_n>V7Y=P-ZwAxbhv1k8 z!7hFLe0uM%;J2+>{D0;A9BUXh;J0()z&!Y^yQV$(%_XDQZ(*|6VjX%Se}?w(&mKUu z6a(G{L_>h+1!HqTwE9!w^Rw>{>s{d7pKi_lp{I9$*cXML$IW@y2e`Q_J(x6j__&dYK9e-sO8~3lKkL$2U@US}kZ>@iY!4m)|v?uv1U4Fj3{kF`0k7@d9zyG!6 zw>kTrjnsMa0{8q`VjViwJAcM-{#f%?ZawAoKm7NiGHdI*7Oti)C$*QyNzL4)geQj1 z%#eKRiBoeAs4wSyno6%c}V+Z>m1HYl@7l7Y$4gF2sWOLv*b(5W!+%U!3UyOOx z5*Lb3fC>*iITwn#H-Fka_OtH^3arT`cGF3F0>{^U!F$60!u|ED`$D3v{I@-N+%xaK zFukFDc*e&b!@XeaZbd4@>+?qb$U5`~LA%wJr^ZrVe*VV48UVrNk0=GM`gGxxTS?yd za)eCFcdDrsyXj_u97@L_M#mRfsY{>hZh#M*yakQ(3*u@yzeH)DaW{zYBqzDfcrv#( zd-~{j(~b1Fq^X*J@nLRK|N76eo9@t1cGF+=XKB;C)QgPUmh^PYl^Ob*`oK0PMv8ZK zCt?RlVhJ z6cW06zBfY^9|huJ0_j(6o{VRDvL<QtRuzK;XGT>%`gNl@o+im+f<8((@N zgvs6*s(ABLX0#xHcW1?m!K$}|{vmIhpu)1twS4O^M#e2MjoMmzH_MPU!kw+Z$+8zP{u{q1-ic*z&-UM_!GWVwwgU&o1}n)Q%f!j4m*d0{ zigoE4w7Yc8cHrPQa^gJrZ zeS>n@$C~;^(W7#-i1c4UaqZt3lzTHvS>~AF(;ENL3WT4cy zFyLF{0@GXM|J9x-6I4~Oyx7$;2IKv{9lE`swZcFN{`0Rf`DE$bOI#?yil8H}gFYe~ z2w0YcPt;cck?#aap;+-DLjIoZ<3}1AUjrInQ$+FKpo{oB&D^e@j65hjQ{tF<=WlC- zu?n%U#K`veB+ktS7u}mRu)k>-DO~d~-q4p_X}=;e2{%&3NCyW4OZ%qWa z-0W|BlkGBbe%=qpPsX`u2RiRitUJ+Ex)a8LC4CLYqBUe|Cq_M0frajUXLm=XItl<> zkOrs`eF4l-9cpKfIL0ma0BS_v4ng0F+S%Q|$!kE67L(~lpxGDD990OY9pjcOFSaQM zG)EPw9MGix;8Z!FIjWP&0Zqyeb*B^VD;3oK37s%}KTrTzBtB6Ta(e;#CmExIcnB)u zS}=QI{(>&)$ka=F&XlHc(BhX>XUWp)Vt2`_i>=h}r9MH^MZtH_&y*y$XR zb!eEB<=U!xo-8X`jW5xkV1iDd_WTS-r|9RN>kF+c3z-I;f6+ywQ@%1LfJp}Pv+3&TwLWu47|ma$t9TZ9%m-yN~1SDa_Jpp-$XZB8?F0?EMd zQ0izE=h-bhXNo<$g%uvXaP1cEN4S2Z~^_;&>9 z8tt0|K0KU8hz5u$TvgB_5|B6<1M6D>_1w@MJ5p zD(~JPZ#QOdRYgktQNxGgcD&Fy9`lBKy)X{WS=Ag z@zQ$wjMrcj(X;S4j3LfMo!p+n?h01fHNX0z{?jvdN7p8PHa%-kUuV`a+v1OTjx3Pr zNhGz7{-&8AU*c|q?$sWh>%M1n&fgdYfcBi$x&Cl%ZM2S{7ty0UTDQMfoM)p|;1%c5 zI{8nxL+i}_yDhDI*wprD-CgG68`#E_otD_McukUYk22JP3p+T`fg)FTMX$Q6De5ckfr>SqYY;h)Kn28_-2j zcG+vd?-YL+j76%g*y;Vn%n~ODx|orht*0J-5Lt4w7=4SkC7tr={Ut5qf)bJ5*6{b| zYh=ZNir1y0z3{^v#V7NMp*X-!R|nxEQ{!jG3v<$Dk9CkY)+=hGvCd;G=CXKnYP?U* z)NDwobu@Iy{Y>opN4U2JBwAL*yCw`$&Dg|KRK@i|bbTRmwY8BeTkml1S?;sHsUv+P z4&fi~9A+HT_e+qw)>~7I@IqGh!7iU`KTc(6S{+QQFSs9@+w8~3KmN4+c!jsS*^dRh z`mFs3=kCWf-dO4VxGaA^&Nnsk_oJKH4|hB2od9)Qo@_W@BLi4=r!R#WzD)b$Z<@?j zv|gwCp8Na(`;aJYyFZURNvpN6SBPUp@?o>cu2#)f{$jy<2B=0P42j_NYr}-w$HR5;it~85j{xd+ zc(}D*ao&A(sVVk&I7#0m-*WG(H}T{$Yez9svh?u!?8Ak6ZalO$ADn$ z)@3Rn`Q+3gN<0}M1H6*&fe(tiYCpfE@JKiGp**UHd+4U>yZtPM>mYt{ZXH8+n>J3X z5)ts?pIrnp*#v+!Elu3Rj>*z`w|(pU43ZH4?7^K$ByP3rZoLzGXlGXRAI<+%PZ?t3 zK+b%}h|?5Cw|~R21>t?HnTu4CBHzd{2zk3|AhhpnPNq5}bZ)-AVAXnm<3(sqLiPqi z30n5zP;6>hfH>@?BY%@LD#R?NX&^SXS15i#4-+8b%|a!ZyU&%8*=vKiKurwUpD4B8 z&5TP|#8#vqZ$elT(txatrk@KDgb`~|o-4{1sYVSnQh#&IUor^u7Zro#4OYMnq_Ycp1gbXq8^;@M!)Z#1cwYH5xn)%b?T_+4sMNH*)3m*Swqt3_ zL>|_Sj~2s&cDQ_Wka&{g0z|3J9rr>i5%SQqo#2g&K4gE2)uO-JR)@HkXMg#N(13bp zAdL{Qvw~Au!4Le6vnW=hGN3AGpM`MHekaHm$E-3d*3D8v>(~$3OMlZ3sA<$w_0E$r|bqJZ3oW>3!RE(WVr^M?un;N;G-no;G zI0tEeHtpXEe>1>#IkzK_F(GkJCfp#h^+464$Xiq^Q_NhxX+{sH@h_%IC^p3n@mTL% z_+QcjZi%cXOBf7WB@kS<~?G=O^ zGzx)Rpb5ZKQ2P_K-_+>HDbFwivOEM^yPY7-xg%Yh8Y!@Js2L(;PwmD3D|t3Um}s0z zTM{55v_4Lfbq UsVm#Wc5F^P5(z_`X6HY57+>ueF6Os@%ksQ?ZQXZ|KdQ^Qh(Dq zW{*BrK6k=?304_wTqrhyO?o5z9Z7|HIBWl;UD^=h6I0u?Y_w6OlY$I5O`}%kjdgL`u}*6{Ry$v{k>npnKwMjH`J29B z##*OTvke7oB?5R60i_WaQP3Q$0AaC2c-gJ#J)G@-)NYa4J{@rNjZ*~|8zbveyM^kA zZ#*Mww}@#PG2<@LH%|07jrdqg^=8Q4q%WMP@oqHZdXy|@1FtYcW`p8IW=KFWwX7jF zkpJf{YX>tV%L;(}HyRNJguBmzjKkY>+#}Vog{I>Ut$>GK8MNY(Vb@hAO6V{v9-J6H zl)0CL5Xm=8>rxOtN^!gA(L+J7dQwLC_J}Ykfj|tRtHgR@Hm(Y1Ca5{tqg;aTyOa0VC;~4&V0j|g_Od*6n*BsrlDEwN_qNG|{NfEgU%#X+-JF=ipPrj7U-b6Rb;6f5QUW} zz#CbgFhkR@sN@b*+Ebiwz_&$8?sbs*%u?(XfCttVTfuLlm)AZ z)oq+X<3EKXAdbKNLn}6EGphK=6IQY%y1m#ReU%}%v=sWAe$8+=%$E?BjcA_!EiE1V zO*d%@u@knn-jR}LH2eIsjOTm$i}(U!_b9ixKx&gS%p5vAT<3k#*> zFY$pT&fw!T`=Mp}+NORih}-XkKz>dESSZ*SlUTssN^XIAXXHsXg9GyhTq3{ z==gZ;k!7uO-TrIr<;mqI#6}cvJHZ~|-=+_rCFd#vnb!4QUT3eUiMAA<5bIfTg57if znyTd&zmt%%FD(zi7Qh$XQ9OC*mIYcVmQ10`cfvgQ@1I;!GdohUWkHAV$@WwJ2f`)0 zCfMO_yR6vc;>3%pA$ifG5pnqp#JUvn7e;`=AYl4b)C5BTJn_OZ) zRbxj=V&RgV%w(tOV>o@JGa_ID(LzQ9v`(+GVNWE6(hOYN`}?{aU2x~lWcofN1(pS7 zRB~c*J`OC;B5~;4$rwRA{7bN7dFZzYP=YD>3CR?;pAPQlCsODOx)Nr*J`npyGoY0$ zz4S-mFFc5w9N}A|c#1CM?0h8*KRdwOc&g#>H;3=U1T4AAdl^B>(!eGFU`!%SQzMsx zcv(c0>;@yuYmY3PCbH1uZpZ>D-AK$8NR!0->^SpD za*gLS+EJr^f7zx{cxB=^;1c#WS^C=%WLTHt$FqMhw|X}*g_spf#PJZ#;HF!cY;F&>eTgqBjnq_ zraP4kW-2RwM?h51P`=j;UbP`?o}H#k4@C>1gU~|<+PFDT)f{;uSg|qadlLb0nTXCJ zB{F%!O=Pk${0&gWqdA*`74OzNuLe`eOne{uyOEe=JOnnEw!88^@|pN&%@stZznm`x z#}-&E>x;PjS$1D;lkm$gNB|h?i?P)rrEY4;6x?=nD%Yy+r`UwPup7W49ri{BWbSZz(JdhZ%bubXXcR@#UjZN60U*)|Bxl# zM<({u%lM#0hlr3I@<$!OOB1Cpf0QF_sb!UvCgLKFwKSnIf){d6AW~7P2!ErjWn|aW z!=MrddH%)+0RVwK@_#oQ1<{5mp6nHK#Eo`yPJAS=icO{?Qd9hbhfwSSrH$}6-fJGF z^S}f;37}&5t+L|B66Beox|BN#NGYj60Ihs1P!fnw1Aw?qI}aQqtRGlz*r9r7@i1+q zuGwU0YM~#hk^xX%#Hr~qO(5NGOhv>DoaK$zG68IFTdxc%aZ&uzGG|<>vf)!Mq_Q$u z^uq)>(VwZ9{bm02j!dCXs{)6lsJUb&9;C5u${5X_VLsdWb*2%Y{6{rn ztgjmr7jwSoDqSX3Q)Ph5r!j>#Ru;&PWo0py7T%I=sCUls+7}1xmlIobbj6rlypCgt zEGMoNH4mocpj1mJ6I@@arDea8xLYlwIms~=y7-df-%s&4_h9S?StnbXc}L>LaWKw` zp9aq?c@}#lE_#xQ5r(yz?|rA_*9;<|UDEkOHQY_?{t|>@K^-H(_*oKOUEf0mT#1n* zVPgEg^UAMiLN}G%7@bKpG6xk5L3t&azy^#cpXP781wgUVnmK`ewfnsxOB*6izG_1l zO$4d3+!uYl7n7`bV|jy`vrzkp+cpnou}Zu;!>y`KZeu=9>%VoVdTnz_F~%6~1x5LS zTun=s9T)M^4u*#jN`}iPB}-e67eCH3&~n|vgoz7d>=l{*!=RhyMCc^I$N)Es;*1;c z3CB;;8O>jo*BP$A`7;TZCQBa}uJd(qE4P!yt;A(o+`7tad1PreYwzu^WpDOGy0YC% z-_WdBSN7TE2Q1sYH}rl=`^4&wf}uVOBde;_{>C#!R6u(m?gY;3An#)*aQuR^EIs_e z!0*5xc-ERrs!L4e{^F3F%LNcD2Tfr=%zaBRGY*uMh;q0f;H4YfIwh)usuPe3#{c4O zlvz&S`x3iOa{fj|(pDPU@Xbre@enWi-mB~Rb4o8G_*SyXAfhB=``J;@FbP?NzfpF{ zBopJUxBWJ`GuRj)_;@Y`$(+sKQRjh_6&Ex~6`TcEFMDq*(*Va>31$}RMZD+?%_lW} z8weVwg>)gP9r$5QL-wl{X(!{C_@YVfBM&n1y57yzTb34b48}Uu*}j^2$h8Y7=@9NZ z1k0gL6NcK|eSv6mhv){MzfpGL#}P57)8N^CeTg55f3e~h`J%guCm&tCWeE(sBCF^q z_?UO7AnFQc5AMaA4*o`QJN6c=mumaLw$~oDD-#K-m)wxlLj+ixf1?o6CP9|m)r6Z;DxwmL)ktMkS=Mn6O@oW7OqTw4NVu`E^-=xvboy7bftwMFauK9DbL@JkVQG_Mqe!=y4#m}2-k(AQ7_m;S z$N+){z1k|m1}I}vi{-NATOcqU?sWd8C|bXx4l+!$MCZ;lL2mpJPYJO_olFAHg#snc zfWAU@6rZ3(xV)EhsGcEP(jZ9;0WNIjr(vf#@SV=y`@fp^bar2Pfx%FMyaTdMYagxE zeWpD`K_Y?YP1XC{en7d2*J*)0`()lrh8K+AR$f9snrE5*+OV9Dm?B|mKQ%Ezy?W=S zh(W$Ty|TW5RDu&Vx%j0$WUK3fzBAEG!EG!AGn7}Iy2x_b=$nY9X5SH$}hAgQ+&$tEP2Qwe{K6%!8>`p=$A>#Y|Q(H7SLjv}nj~VIEA2tu^lS z=IAxmZqBY|9!#}WsunLQZ*@DaOn0o~DMdG`-7mh)bP379ODBii#LNp9z%rM&Nl*#F zVHTBOJDi$+f|N~8mHzTAWzLb_uVUvA?^iG90Pk0c)6M(>Yl^+19?mWS4|GrfogPnE zbM`P7YiD*hZkinOa zD|u#{1PATm%S+ySffL6^8+>_K?>`-1`fG}x4_}JZ#KV{KFE;q{(-#fCTr6jOUpU&_2{TBlFJm+i+Gu>K{J-9vm?surO09IxY>^*4pD=E300 z&2GnMc^!W$zJyJe?cmE8^CAy^BfG(uZbu1HHaTC@UxP16!4AKQosYaSBdkg z`31f_?-liMo;1Jm#uL8u8~Q2u^6sNq_>yrgtR%4;LeH#|#RPdY{4wb*$;W-Hgv^JP z2jGMx-w4F|mM>*BSx*wdL51(WT2QsgsYit2N@5nLDGHpJIFYJBw-&@YB3b%@L_5E0 zu9@BGA&|YGX2}qr&x$c(8t)_Dhi68Z@2CTjfoH1Uh3w1>%B&JIcNZ-E)AL8WAV+`}jIw_-SK12@6gPuUpw)DWkggs-wjDosd%;jIq}Y+bg)Be|%e1)hO? zm)p+Uhz6chPOH7SA07_=Ra5-F2lE)eM*~>`Qcb73W zn8HX~B7e537}MXll13!EhK(IrCYr}xp*EUKa?_Z7pLs>Cnn%@hYq#{g&QiX}H^Qfc zV91#n=SQdtp5y)8sr7ZI))Qq0vkF!v@@I9wnblvV+LRVp3vj2IR``s(Dcv(mQ<`QO zjAN>n6~|6YmX?@E0?n_+5R~qhmS2jGGPqfBsGwWIjob?>do|4L)gkO?O1|dq7kPVA zBD$IHY%@!6rMMm|cFlVIsah8Pwl1nsviQcXFOwGG(Cd5XPs@&?fbUtBJF?D-{i0b# zvFI}Wk$7%&C&?(DJMv5)}ar(i;-kIcLk~zYBAs)&1(F{n+^_R*C2n;ZC83m zW6$y%EB-A0`Sw+kGyP@Ox%GXQxBLyVpnQNbFZL+zp=Y{8o{ldUb(d(%VJkz zUCQflb5yExF89IM&-H~?%+@>tz+FEhcl7Wy5bF>C2Cdi}cOG-xc^C~|z=vhe3eCev zCPzq8@G}Dmv1>IE^DWfyH-;dX+HbB=t3_B57oZU6E%drVqX%QxYxE##Hi?@JsZvwR z$hq4^HhZPjl4PlO&_d&@1q%V`xxyX)dUyO$*%gtB4GY<9HaO&aKSWMM9Ikueeo5;f4R~X~fV7U6tKl{5O6*Fv-NKTf z?|JDet6uOo{TI^1fW0OVovvz%!%!1zQoVre8+{HF3sya1Fv;Kaj4`$o}hoE3uvWZ%WY2rc~7(pvPo4*0C=lx92*JbtmHVGmFRqtY@ZhBrK6CDhm3}Jjc zs7TllOFm&`z+D-|m9B~1keZt1Bc0l_rh^lAXx}i`T#RT9ih_Td6}?p7)5;8D(ab8b zeQ;}^_?zxS!G(FWsTTg4SA~_Hxo8_raK=e7WJnjK<_r2>F}v^#b4@5PfbbyTTVtTY zxX?E>v*w9b?UL_Z?MY@ghN#i`0+;Awf=hg4*7G(do%pj?s0a*Nifv3QvDtJly+jnq z3Hd${@OWVJw7Obk1j&m_OcFO*ov2olCTUm}#eCsY+sqf5yw4d`w2=$JQRTV-LP`im z*0-E+8YKmgbsYdplf!#NH^Z*N2g9y zV#8r!Ox{yj=}X1&V_Ju>b>Z~x6_DC;M?gfjWk*NpjIQp&Oam<&iX<|w4EQ!BX43H- z;~ypj(4kT3Z(OhM(Vog0tdU;#rGQGTL#EOn@+<9PXex2SXDUs|uarK7nSA{bU0;?-gG)-R?yzq$Yg`@vsu=+!b2pKu5{0FU3{`f{ekI0EVh|{xQkn1 zG1q{)ZGq?_br4xWm*IZ+r%Ou{A{qrAIp4tJv)F%iBlE&%)CjIOB}=Cd;);jT6nc}1E)krE=<9YvTq3Fa1 zii=XKe}lCp((d}Zk$-{BMtRf{LEgN|imvDL3=F{zzg8IZy=KKwSK&504EuwX{>Hb3 z^RbhG@!`dZSIk(T#jeEJ{1pFE@wR|MaL~(&?&uJ{gnv7RN2LmU{wAXhRTOx{SUVsi z0tBPCos^+Bi;ELynx_VRpvYC?)bJ}_sY9@<*l>) zO-=lk0XpcIxJ2c7EB^;Zt#|6L; zQZw^wx`6s(JDJ{e7(|@1Dt4JD74^~|ZI3C~ z-<Gc@GaII?{SZQ21Gwo_W_TJo(losE+Hz~KhUxbP0NYgK?2x9 z2$HLZ#e<-u-K9Js1*7L*o1%KvFk9ySmPJ3u=mwKbKdCe-ztRx5Qigs~X?A|4-lmes z-QqGG7maslp+)^Y8d_~=XrH`xO+R8I4b2tz=F-p&;p3$+tG4XI@^IsJwQhGJ)9iu78rVn+zY?txGf2x%$jvMxkM-T@`;(_g#B1xbw9J7K(>&(}RE z6q~E_G!VN>=c(TO07nB(L0Ia@cLLG*(m)r437~a_l+z9U59KmWNd|x5ZK~dhV3g)U zZ(3aDZ3vm9kA;Q2~hs4dMTb*F~P=jJsew_zT1#uor|HRGQCmKaZ4U4M5X;2^OM9POg`eLknE#A`!RBbJHlbLyI0euF z5u-_Ju=hh+NOr?C(3OkZjqyyVgB95oJ{Ggsl3>NwfSn|VkyW+BKkNJEz(%!s^*_v5 z#CJjPu@pZgXbXlHkZ%p<+XiYE8^f3j44sylZpE&8=k#9Kf{md_`)uzFkozb_MWO@C zI|!fp`x|dp-{utG*3sX1nI@?h;4uf0POsDGMD^C~{ z!=%`#exe+FIJEkWOQ8J4{O>#@)rZMvJwnAPlP?7N;-nZSyX*?vN})WKcU zl_Yks>P-FXqC!nY8Um>8#gzC_SgyYbt-rr%JEy%2Bz3ZHr{1!8C_XSK6J zkrR18$Gk5w@7tv&xqOhnNn^Je%1vmHEdA5r0!ggY#M{-twbqnq#ai^wX1-`iqEUbZ zyJQSUv266l{8bJGYyFu4Kf$`FjUhiW@~U^PI!yE`g+JF#{^j1fORn6h%_BFfH~N|2W(jtjvN(IXF)aXK(nP9 zuriB3i(QP7@iuJ2X#!=hclsO+in5fexI0Qr_r8Ch1)>wxrlkv7FKZH%acb{C$eb*_ zzS1QV#?fLps+YHR%{4Xp)2A~LVV9v5a`$$A_Li7>W+9_S#^kOf=b@S;g6R!Ta6i{3 zs+PanAntWZMGobpF%%^59Y`2fg56vqg~5ZYRr{zE?9|$39pJL4@hougj?5CXWazaO zhyU^181QIVf+s`k?_m41$MSTv_e%8QcZT22A0={Xz0-TfPCjshb4{r``78PBPTrfh z$K7A`ZB0|ewc=f4`;Bt1-kD-b^CllA4QKQBMYR^TcGo|cdt__sZ`S8F==(UETRg~i zfuZ$p=9)127On|%NWlo9{8*;f(LrO6e50TsQXU)JB|bc9pVEaRGTdp5(F0mBh|0P# zvM-efPQg4rI=R{3GPq0Y3z^k|Up3)YwR2YNR5^!)52xcI)A7Esp~dR>)M9m9=k^;v zu{hP2uXV#5-B&cFr9&l`;PcS3X>x2e>;SqV>{}80XbugkU<~Op6VUg#sW!ihr=^&N zaFH`nn`N(?FZs=w#P`bFULPca=GxOI1*^PV0ffu9w22WPRG7 ziFjS%J+O*bGS8iZq#|(wuTuA(>73)3rkc^$LV3L=0QOB@Z!uY~hA|{%PCp zNmVvSlp!(f2DHjUV&paX1#8c8JQ*+ln-*7_#3A(oq)YIA4d|hsA5rG)--d?7{9Isq) zB6aVH7oD_ty@{t&GeTDw*O4rCcQZ|_m>Z50oiBbYN4ff&OUGI9XoKEj0NdCFpy#~! zE|hKN!UkagK|Wp#peLE9D!`rzd$~*xURaz_%tH3?qlGB`0H>p6=p6}IldBXG#@JQ& zlIzTrz`ILq9dtBw zw;;SYMYcLX%_4~VH>tJrXLxspiaeTvPo$-qwh7b1f$4_wx%B|e8JdQ07D9k|%&6>O zoF;0^4C>6utb*TYc@pmdk6LGQ_-w*S*U(8rB6S>XLR6lR`OR)4=StKld#B z0{hd`?9Z0+&#^!EfAfE9e{OL1=Z0RNvOk?sP<;0ObTD(y-kBYHu0 zR#Hg()HKxko0{3G#H)L-UkZ!l?N7}pd`TYw4HdY?!=0fVI^e-7f78`MhJcU>8rND*JT|L}=o0!I{(Ybh4EqX#^t1paK zCc{8<&^oYww|I5R);9arKaE#b3a^HlIdcjMuejeX4PPL<>LoNF$gRPv?^1^_LXE5* z>zB1(%d}szd;To^Iy|#qV-(WGv;I%6pJ3wUYLpbdk>lFolhZ3V{i4DSEK_vgZGz)x&D|2D`Vd0Sn9NWHUx0xlJ<@G0K! z2`bh*Pf-X5!xA2#PRoc+jhsmQmGm@%zf+j3x1U;v*uJ&+UUO{_g-s40Ee#x}9E`OY zb(RLBq?^g|s=!kq#?@W8<52gj*@yjdKxUpOxnZ8PJ zjir|o?h;y-6wO6n3qQLM@s}3|we`j#3Wu=KcdENp!W z_^x;6{*wEoAqyY6U`03{Gm9+K^gFz@Kd-#ZTplVLSz*K(#RPcKxfihCU;}2nXEwlj z7uHzzhLxu7Ls#h2kjWZ_rMz;~3zk_!Cy0U8Ev(+Tdni_N?&5j2%a!iJt?UlBK%V)| z$}>+V*n)#5NNmXRpbnP~?qQw{r9$>OxRS1M#3*x|vsJC=B-0Xq(o_wtC{XWr*VgRL zPuxLHP*txQd!dg`?}fifo@8u=*2dk6EFRx{?L*&Djd;K3HftH+L?@jqIAoGCu{0%P zDG?yjhia~(Cc)KIO$!)S3?ATj+*5xjNQlg{e$G96>2|ll`6GV?Avv=OAD1jG`IfeH zu`73AU)PbCzTPP$yM=pu;~q`KRHoA#TxruoWWHJdTYi!4vM?t_UwSDV6PKT%5XEMU zv^X&Izu|?k>Mqihln4vWLh_lSA&^AFuOUNo_elWEgW064I{&C<1?rvqK?^HB837as zz1WF}b&QwZHI+C)@qP~!vmKiF)5t$kAF)C$1TmYg&C*Au?^hYO#(IJa)R%<&N&CLx z!Ojf#f%RuYzXN8{(YoT%yJc~$hkhSB;vwhO*bydv!0W+^50tsQWm`wffA<3Gq0_^w z>-IlAOuXM!{>CWRHIo6s=#%5!IaC+;XZ6-44zKm5kbY4+-PgHY*M^v=0ou39JSCl!#r?X zid{L}0BJLFt@uBx`JLg!tXQ-%^U_#9=DaM)dRdUy^3SV)TX7{8X4XBNb@keC*Pj$y zOGAB;t<`xHeq+~kRvvJD9Z$Md65Qv_R@9%V^Y_m;dwtfLK4YKW=dW`5jCp+$^uY94 zt@WsSV10(zN6noO630S3Jc*IAzBT@^MS-(?sPN;@J-lfgu4!Pp>D;nFUw1xUxg7mY zNABwxx7vLen01IeO+!bx!kVNTPRwbz!k69fg}qY^;}18}&^cbC#~;2!zmIi)&(ZHb z=J%x&ouAPj^tGTLqjkKo!2B>Me3d%Wg*-mEf9h3#_ti!2E0`3Q6;vx=QXI=Dy+XV- zj?LA1m#Cv49Vath)O+6l6+B&|gMIh#bW@jXc)EtCJ%FbHpWbI6r_TXipE>Z9xbO49 z(kd^={zl+o~f8sRvK{>uTZMk`jfy%$Dn9BuaA->e!n_yD<>z z_kc6E00t!b4j0t#Tek9yBJ{T^79+kbs|M;Vo1n1!!KoX2xs5aYvE$$%J_ddFk8bNg z#IWe15)HaAeO_Ae5&dYk(wV%1@AFQmOtm>6a`a8K-YWyEcLw6*$ML=CSgMN8Z|h&l zH{|-)U_b9&yehM=V`Caew6ypsOfbO_;Oql&*9F%C zx00phT`|;(mv;t>TDf85P&-|Ofkx$IxgwGQVj4rBl*3hXmkChQO9e_hlr`p&oDx^0ic~3e4Q*`@z|ZEPoJ7y6xkq3@5$m#8>7P1RhMKRB`bXGn4oX zFRu#bsN_|7)g*&?yF>8Y7tab-Jg>Ye#}SxdAhL^PiakBZ|Gk2OsY@jxLY%Xm>?&`I z`o&AXM)&ctyDZ;AA|wnWa3c01oIKo0J{F9RE0waxGFPRa8{$5)*qw(Q`(h!F zzVByMy_ccygu;_3uVS?^AgX!~`u?N>=2UZEg1#rA?_)Isq#W_V99e6uiY17}qPx0I z{SlqJ{4&uQI+8c7gUPqp64`9J#81rkn1xcpWT?CZa-@Zxsra-Ic#>@$T4=)g`J1X- z0zWQI;29VCE=iAae`MvI^Iurlk#vY1jMxM(w0HTI@Xxh&6`Qd|Hxr)nO{{wQSoXl9 zbj{}uW%+j9LB=^)c<6LtpV@^+S)`6MW<#gA)VL41REBgQ<&hDJ}CeM5tY3B}Q1T9{zLgMBuo~ZQAlud$jn8eGI&<&d$H-!5dafu4y^&w&LHLpD_Emv!hF`yBG*m&M}EZa=3{HKQw#q z+;a!i`g-bV#c$&wcf|#tYm2ktbMECS_)G!kI709AnEy4vrXAv(p~cF)2coyM&7^fI zCOVkN;CT8Zkm(%A)PlHP#!IhzZZ0;JAGS#X^vD_+?mQhF&^MJ>tFH?OMo;SI)yj}a z=}sxb&L}4rRAlvorOY`H-8Fv7iNSbLJ%d^QNF4yCKkR;T@@iRIi9cfUY6Z_-0sUVl z`du}NNVT47~gW-SXV)L{I~QbY5v@(^#r97FH_X?}jU zc=u26^ViOGAtIaH4|Q=cL+(9(e((Po{QK9)7nshU#U?9()&XhMJJ2KkL;sM!ltQkV zklqaO4_u}o4Icf^!LN2v>$8(=tua`a2vlsrWS7vv@6X41yda%y%V1uhVu|yo%_*EK zd!~Jyvxv}}jdMj)`=|K_*FIs-@UBxn-fdFHUn<^hcJc0dI^Gk!YcF5t7r9S=nfSNQ z@AHAi-}Dv!ndpe%iFc1yzRrC7Gm>v}kxt>?Qbk2=!+iNVWaHoRX?u!)IQ&|Np3%cU zd!bmcOuTOUM8v#ra*SD9QPD^TWOB{oT*OJRXcc#iibXm@^7GmyRtmk=1P7F_(%pvx zF}k+E*>__yX-+AO>o-iX_80C}Eu6u9rDz^~WzfdGCw#&4O42B1IhLOH1aH2q^J)s{ z+HPF@f8xA)?RWW`jBIK*=T-Ip%z5=$@+U-6q{cQm+mSw1ayrl==;jPkxhs{D2Go>y z%*d;_m(XGt&~Dz~6q%A_nYg{tH&SnkGH*JWHYUFmbOqwrgynzqDQo;w{`&{+2ZFu6Snqei1@1tCm*N8VPJIDl zh4E94=ZJb({iIO{Z-*}=r||f`33U2@fj?iS{_!;hy3Jwo{Y@9|hX3pT@5uj2lZnK- zMiT4ZE?f8n+?hB`!~Drk5P7xu?S%A?DTb;o-*?0M-QZ1z?^C)~@qOoAm%r-5o4Frm zc>BjS_k|o1O9)%aBjOoMC6897bX-H+1SJR$;$EbGJ34r3IN6>YA z2tkn%{dme9Vn-TXPPTaOY9rnonbz0;|Dn&h=i2~Zz?`1=Ncw$@qX3pV}d-w57Lw?VpK>!?*UyVl; zUbKAv+yYJvWDEXBS9lTvYy5JnV_)=+1lD=C%z5i&E{3t=s|~|dYp>Y)VvXNgt#raQ z=r?CKXYlW{@(KTRf_+)vN_&mHVCF8P(0tM;G)-3Fb%1gFM<%N3J^6t=8Fd!qnXec}Ad2(MSKRwco0H>S(}{O-%+-VIzvZ1D zO_by)mCkxm1?!yCu12XO3s}Aye$6tdJ&oO?J6SuTG!8(D^)J-48({ZxBu|72-I zXcpi5u7`CJoFZxt#cM0)YN)Xt^HPf7%G24SvrTDyX~xsT^t2{Cy_d=OJ%M7D&x%DZZP9< zdP2*;5m<;H2iTOB(V;mCK26o*s`q8P)xc!XasB@yZ@%@a{4M%Q-rS{Ad-CSvk~P@0 znVputXTLPGALv2*sYDWP3++E|&4%_Lo$1P(v+^Cvxk*`X0H~~U0mpk$o7k1h^d~x~ zYV+&-m1hDr$ck2T30S0%B>NEnhvVX#-xsag)EtWarTlp{iI*l9qR+^$z1B^tijx1iXbmwbYRs0QPomi4nDOtLXm$a|c z&br&swu6VMlQ>Ir_VnnAzyWujsi9@nQ3u$ zkvME@F1v<5gyGgO<$a8+N`LRF?;1Y|y8RIV8)4K+DEMYNCrPt*Xf-vTE>B z7#pjG&^d3IYO#GSdomWUhjplV^ohqGe>~9g&LK7DE;@_eYAP0~!&(gG{C98kIy`uH z9pbi_>S>-j!hy-LS~eaWuf4Ot*X;G>QaCzo7&RJuud2LW!j>8}av6kXKcMW(dB zXYMz&&gXuX_&eNI!ko<$=Ge>ZmGO&r)xdt8(`>$v=wlqjwdXdURa>!?qj=yY6YPO~ z?I-OPk3)TY<}R(_)Wd4x$9*Cbta#tT*(?^$ed`ovs)yozMHiGhH~;37WMTxHEj>He zwEzuvs`dnXzt)w>(#?Qc+CR+O^NI2?&5aO?>7gOd>==O7P}Z7EHuQ%C37fZ@M46YL z*i0?mI<;5v#J2QjbdtCeD~s@v8bLcx(l*uGMLVv3lhhrVk!BmuvXlupczA~ z_=%o1y68XLK8re%+s=D|iq%NcWG{Z@UFrV)_t%yd%r4x|N)p5EIoT`HEm?mpAI5@2 z#$`@boBT}?p>+K4F@RRcen*knb~hueCkJ4F6s&qR=pXW`36L_HdFascEk(M9fq$~} zu2)5C1{Vi?>otT)dU+hR=-Z?m&h&p*>uAcH?v3=YOV@zCIzZCMpl^BfHOwK41;Y1_*t=kM z_pStj1Q&PsqY}tSH@G*Kgp*QQM#^7Gm3IjG7GV9(W#J@@A9>S~dV`)Z^@d&J&F!f- z=o!5?H3*cp`J-{(XlNs6nms~dCQ7GzrJX40lsJbcfWfsdWe3^OM4f6Ven)Egn$9*Q ztg8HdFQ^X2W>_g`lAB+uJ@Ayb!f7(-!4zgxG29|v1E8Lfdqa&CzZihxT)d(lwj~gk zi)aXd+O9a6%Uys9v3X`bz!U3O=qKqi27(?u`Bo`PehQXk|B78|xnkLsgr@*jo$lr4 z@ICK)vf<>4b$M{|5%3LkV#)ToU@&rryG|cP7N_|sAty0d*Q zR5|e80=(-r@J@?wr``bX-kTw=UVSfb0AmJBPj2aZDB(1Ku>z*l6{)`Dm3E?}Q{o9d z39bPw04;H+6nF7K2i^xIjvsb3G3@D@kdv*4d;6$xp@y!ILpNDE>XkfrftH?53q#Qm zYgTOMc;dU>C+vc8ZSpt(5A=zQNQCESpkSDm1gahn`iF{wMSkr<&nJX$kGc`9Qg(u7Is9y-e(2e9$2^s1F0_4`I!Q zl+`=;tiGinF&=kWapaZGL(}<{1Jn{le5jqkZ{{UFs@{2I1dO2B*L3K1!vT(K!vVq! zkSpo4nLusRR*dmT7qHsMuKDzq?U?1gHy_>wvpmBP1}YwBVVwP!(V zFB?}uH&mrm5AJy!5+q6g@Jb*aF2$s5_YNfqPShvd49sbVlG8=t+i@t#ve()pe-ev0 z6}^$n3K6q!4f3c4CQ`Qs_*>&o(pAOz{v=)SCxM+Y_Wl2c{v;69K5|5T`dOAp^CTQM6rp>^N^*ojo1BjF5?S>RUZKJYnfT(cCUS8k$VmZ zU>m09FJ9#A9FLULmK}Yye1}upb(q;CXaYXNp9JJz?Chrs-Yz*hCQA~X#APzb0c{4+ z4vl;+dxyStf@FsctQ9-bK7ASO(U<`Vc7ec7$QS_)Wy z2EzrR7@YywPopoVQ9A{>qA#Oe=uOd=)5G~|F6h1CnQZ7i1jm(@oVP#+XnSYo%N#0( z=xpUjpykcN<6}LKNWt~bgf1u(&3P$*9G#bgI4|?aN0x&Pgubw`17yj4AfilS}oGk~cJKxn27f?*RSd>}+bhFO^;0r0lE-=4$A2 zl2{H$JD01bN0zs_fc=Jda5i3)zxo2~`Kz-5`wA;%f0^ySX^lTQSAJ+HaM3C2UI4&) zu985sbjahcqg3G{FQ9#v!}gUw=!UgKQK+(5!4NX!Y*tVK(hMs|;M*Ddk+Hno71^Z# z+XR*Ht>svAMRWpa5AH%-!8cW^G(?RopCoi2!NJEl1#mBa1!S(YC;!XZqNnjcRYZ;_k#Ro z7eH@(G8>@(Ue{J1V&X^a0es*4au$3~Se^&ppWON%1K-2W&tK@T;M+Aw6y0tQzPHWV zQ}}+8n!ADTr_aq_cER`F5Za8-<+I1_8GM(t<)Dk6em)DnuUL`?-`~3VKL);k@{|09 z{tCW79lm=s>?wS|d--nR`-6%3%P#mn973Cc?~jiC=i&R5HCgaIcVQlU54!0;2EKo0 z0tn`cslMV*u(?d^j(+!_zNhfrm72Su-}_Ez557YX+6;V`4%#F59@QTGKING#_+HYS z2j8dv`acG~|8!RVLVrcSKb?M${^_2=_Yu_G4Se?*pTF#)-{(VUGw|K-hi&28mG^N) zeCP&ln}q03OJgVIa4AwurSu4yOJkNab^+4Zo$X3v&qc*+1S8k;qYjl_cVkyxGjqXt z7_rpaPZ{&_r(|A!9Qv7x?D)!`WnYf*&hG5Xuh^U6Tm$SiKhHJ5u4{T~H>Tz$*0#CU zX6ZNnd3nul@3p6XZh6hgKXx}bTX`OJ@3o%t+{C{&a+-tp@IE^yPfl}SFxI_uDBcAz zRrjiZoh_?TJjC{F$tACmKm}!LO_nWrAFnvimb|A|oF}X)@{03>HSbCg)7A^=IaBNj zYmB{lPW*o$#a-ZEzktBYzu& zXGIYF44yslgN!`gq)7Y3OIw4M?jW29!smZ_-j<`4c}4YV+RNKQSoG+i=%UxLL0vwN zxhrjz$=R};o4Q&$TXCYId8B2+9O>@RKZJ*xc|@@ZplYFR`Nr7ws18SAQ%b}vd+0Wk8b%?|S%oP` zp)Mn7CDytRp{XF_5oWBLngfT9?G->G)n7wlwF;NyZ+sPnwl#0FzLfCPR@L+VSy#(; zOp5i$@>@*yw)HqykhyJ90Ksg3w_zYQg+w`{NSd+6-#AhYD%M!OC9S1K51%BHp5|qY zrZ$HU#g6LhR%{r7U9pN;9wvMJpfYG@L3qvP+^Ugh1Bi($Yw)enjBEIwY94AeZG~9J z>aw=HHV_*!S|#I5i9W_3&CFm)qcOlXm6xd|{~65;b=)j#=Eofrn;)|5@l3t125eg`R++8Kjs@oVyU&U*>x{Sk>xpd_s_o|KW~Z7`Rg7FCR>qGIF5`xR~cYGQD=(9!^@MCf{|L{q69l&v6 z_i}`?n21gv$2zbUXP5C^EC975eE;;}w3_HVf04d3GMu^ui{GkGUu=r%#~K$*!%xmoKIn?BBpH#KM#h>PAUS$fSw zoDv=sRk6?h#$|2lMV`;D5qZd{9y0K4&1*1pP=pZFF|Hi;%6jL#MYnjkV=vZ@G%nGw z@uJi4-V$u}{jH3%Jg~xEneRTJeYhI$!PU81t;s;j>8gCGR%&nxu6ky{)!D$+`7T^V zs68FHsspZ0pXfAP*aoi7H$|Thu9PgTnXgHql~vWMTx%|Wmbb{GlcnE(Fl&)pBWu&m z2!WrKYt2oh`Sx8b8)k2126!#svVeV68h%A`5`8%LS+De_v_`IK{Y%z<5NWv;OSF2Y zXJ-W`SFl!}llTNku6MphksJfA?rIzZ+(ap$@APrXb%8_448Y%tMPw zB(3#jIlj15`aVOY-3+nePW8^RJ8xm-IoqJZk#7a$gyMW}tDu|n+lk+zFXJG7@o+A< zyv#~frz}4UTRG?#8Y6k{vm@u2`oH*_KG;Q<@CEL#*tBt0)M*g1FuyY&(C1P^MECzu z_a@*`71`T(2NH>j-l$}Zj!V>5lW|Mb2w~9#66n^BMg>JhWn5?+H-wI&vIM$=xwehC zucIhBRMn|d zr%s(Z=Tz)Nzj~+%pO`IvHJjhE(lZ4J9c0iLSmqssw))i#Qm!?2NId@#z=+++7SxbI z;kgJC4j-b3-6`v)Z(I!yn7CTjj79&V3;Wess37_PK1rMxSrh6 zMGRF;RTyGLZY{tivAiWjef9fx7#}Z!>7lYbA7IOfhBoQi{`LY$#*Vk`O=6&^JWu!-8p-d5t4HItxlILVvS9MN=K%-zMSMoA;MSnN?$Qlj zaZ2maiZbvwTJ;JYdB(Yy^Zv2VDBCDz@W)F2fC+wJ-;3Zi)Rk8@+zoiRvY|rcMaOkT zuwOGbSDJt54InNBp^F+U*iyEdIMCeqh2i;>8^Q|s8yAGwH$C6~+kO!K#&G6Hx*Nam zF;~*rc*-}RBE1=08cg2|I{l7E=JRkLm(gGzPC~yxw%X4gmigQVeueYkzB@a%PZ*9n z;@6&s$~!%_$F(>+^*8>yZVKmy)1>VLzyd|sCT*4+uFo;)8b*WA6#oE;lbm`!|KOkB z_dy%q>AvqAd7rWG>tnr6+xNMx*J=B{y{y-1`@RjBah>e@+U2#i?~{1=iGAOj$a3!P zv}nfN-QlPro`2CwJv^$V;4eV;#Ub zchc5z?U;Sxep4BT5waANE4Mt@Hq!~A_yb>g!sM}kGKThKj!`Ejk9D#yNspgx-QtB- zcqEFp7p{PGdPFP|^b4*Du$ovqW}jp;kj=N=Fir6TbeR0HEjzDwczlH|cSXmFm}RY> zVh{NT2|e8&GBdumzBvBR#n>JoLz^07`viG6ENsVb}Z0H^VUK zQyK<=s&2vRz{u?P3RSK1NM!b9iOB56Z9xnz`eo5omKE4wt-wShJXzQjf9NsmKvcMV zL;E?MC#E=)PbW;#3k=C^%P#3%2_k|Mi0Xy#{MaX`<})bxF-Ajp6arSH#d0t8q{MR9iYC%oDF{N_5FKh!4s=oj9Xcs0 zdYp7-#W#=}XdSWKOX1!DL!>V(g0ktXRC{r>Sg&n5vm)cXAg?u@Wr=S@lFrseCm`1` zq!`Y~d0Elvcv6C)4B8&^1lStiXPRT zebV^YLRB?x^ljgP{+Gdb;9Qa{{ekE=FchC8{7c`9=TCcnm5V%TelIlNudbR2 z87`7P*$ep+nki!+bZH?rXBE)C<#{5tMQ~{;K}jI47V;#l-75;AN??Z+n_N3$FKMX|3Qca0Oc&xJ7$_snK3<`PMPV@^Z0!PjFjS;0BMm8VgQ5|2+~k zzB8`|=Qu%2fzl68Bzu24rv{HW@%K1C{3-^QvH z2V%jbJPw_-yAe0ar4~=-pYJt4)hm6a$(0^5`=m7w0*HgOwUSK^IP!4SQ!bIzHYgAH zyCoMb6Pvkib$f@W(47FvsL(*+jE{}Q!xw&u8x+y!Xwj^t&_}2JSb_8_G634V7LUl2 z20t@f!UHOTn{Xq1W%;7I_BQMoUG=wCVbg&7TzRrm#OmwzhTsW@^XM{s1O8oqE8;0M z97O=ZTk~8-=nyP6@U(Rg_nGjB$KD0|21IY5cOHZf?8FMybX+CGozZZX8^wW>t4Bw1 za^V`w$3ZUS?l_RaR(}F2CTE2oZ>|x#4^6-oPe#bi>T%;0wph3y#2f-02`t=+gGLn5 zIxSJ;*$s!@L3U%l@{PJWP&lirXLZ5xxVJT=z?$s&$xcpM#4$?jdaQ+{ zLN|(wrnH3(V}k^D2=pK2`XzMc4fphh5z-5Ys9^V@xq!DNv#zN#g0wW@c z2z0SiR1QI#>F1~(y_>WNQ0i{{iq4g;V1B~TazeA<#e-KKT%jH)=DUSHa|7%;Wf~E~ zbPRweL0fh=pb4AXs9sOd4-&CeBKDPEeKZ63f^pt4QeirhLa(CL@%*X(<}^#otb;>% z{njmLW_rdlzw$FfXaz!RtV;~aMRun3E_Mqf1RTJj4%EY#6Tn8cTRkgzOtC=xojTHq zUw!Z-8@L$PvN9A0P9EU#*91>Mzd8U-kLD2#*l_p4i17J0d1BN4%E8VB$HQ*9Qwa39gnWc!Ha=jL_To4lb4P{5L49fu$Lt zhY5;o4N)hsllXN88`af^7zzC$6rs=}u*T19;aV)FDeq*#1w1nu+UXh{ ziML31YwQ+Z+e%~WjGQ(FKbjl1S`PQ$l{bP$G6o_7D`Mtm^H4p>O5S*2wt^;6`@3lk76`$J1%eb83Q@`3h*$!kn7-FLUofC5nm3@1#zN+ zV6y_3!8RPOMezS8Wt$6P&rJ(X$c`Hgb1+$g6LRtE6;7MTAWSF>DdJ(A1K*?d>VBB5 z$ENCaa-G-wtaODqyQ*JilU! z53lexaesFawwU4h$BQHrJnum39fqI5>gPMD&`ujr6dmKso1%-Z<+E6eD7Rd(XCpUqQ54jScHb zyJ#0&46IAJ>drwIig3I3{0czxt8tHz@1s$z>Z}lnHQ>gL++hDz?(SGKdZ+`0gmX5Z zWa9*QPJtM>C7%D$^i?hk;4w*n+k>3}T!JBwy=ax*_AoJ?xVU@XBz%zqz3xc}!9Q&2 zgHmO`+<`WbAoqBNjc$qk6)(MPSS0_luQ1BY`{m!j?02!1lK7tBQ{Wj5oJ@kU-~ZxV zx@5ln&+)(FvK0JRYW%PLulV0evex*2e{yH&dI;(r(*21$!~fvN6Zn5V32;KE06hSh z!2fueRSEp}#c@+&=3 zuSI)b#Gj|Nc;*&gpQDY?(pzBfwL0pk2PC}4 zmb@dLzvvO1K(so}iyE589s`{Q0yekC?i?{q_A^l)pC;Dx-6|WaU9r1#+OO_f3r-v9 z7eKIRznZ%)b;X9himQbcRY0`qvT=uq*3Dx=TyySbuvtgDv(*wzP+W%%1VM7n0^Rd? z8GEvZx6ttD-k?B7dl@F8(r7|(2FQu={2u?L_=9)A3%X#01ecT{J;mD~FhL}E1W-KX z-Q7k=!denx+u>AQ}Hl}Dq4D~D#`+^oIfBO!{h5@2*+2$G^DF9axK(2+4#PL zkVJ}GZA_+RR#hWQb@j$^@dCtXAwC77EN^+ z#}!7y@tn!hv_%5iYT5%x3~cOMY=fp~mw;XO6!~AOh?m4q-Id{MT*!I=N-=mM=0wM`V+z>aYY<f37!R92R9+fA02BbEhLDMD?zMdQQyBpILp-|xTj&qNP*xK%h?Eopbh4o>o=6Z@Z33CwqasvCx z|1d&l0|>iO;^VgJuSj9K9){_PZ7MwjG4r~+F)<2d)(&`*<8g(%h*KBS7KP|cpXEaU zr_f&n)L??b%zXwm3~W5PPzsS^2+RUhJ`6FCE)zhQzIy|xC!BBK0k9{@?8V`{0G0`S zFz8Aedu&Q$XX+Xb+ffa5kMVFFN-u~h->AmVxbn2C<&z$pnlwu@r#XyNm!_(&+>E!y8!6InLk+iX7P#l6ZPtL?Tt)p>OKb)k_j~Q zrTCCW%_kDu2|-qrF=?~vJ$#;+QP^S|q0jiH{NlXPfFOTvF>6Xs(`1CQhkwmw^~bLBHq5dWcQg`?T! zx9S>JN~UCEe-?6!ItMsX`~*1hh)!Y9I}j)%IoUVyRCky=m;JLotO>iV&0X^2T=bp5 zHlYj%t8RB}{*4A7(JzXDku)U5ULN z05q%OTxlKf^%S%gq#bna8^7|MKL9fl>xXgYTQ6&z64!aT>X+QdQzn!9tqO9xq5b)0hEt5gN&bBR>l7_>ZKc zq2G}$PvjM$9RGR46O}Iu3W&kSv)?&GZq#Gi%gQI53de7c{oA4%ffva+#QLn(&c&?1`;mK^F8XnpbA0zn$5+Lb86SdD0?gOsDaDZk;vQzeFk(}89K$6XV z$Zdb!XnidtJr)Ug&|9>#oCp`%S<6B~J2UZ1_)a4Fb;jeT@UPnhEDZU_12-jf@oR#Wo%?p-UzranFBIKXCWjG zY6lE+G3=ya^mB?|^<)~+Z5vSlG9YTMw~HR6i|TJ9W#=H?v|n|T^5_IhMTT|N+#Tb3 zs2J8I39s%v8>}aieW08&jG0b;5Ao&_SYCswgX3~=u+eZKxT?ORoejL^Gd8dz%D0?~ zq5HTznTw;llL8NQZ3wx0p*LOX`dYv)bpi{$h&*a2E=_iM@HeSlu3|boPj?iAFpXV- znrB~5IZ247r+JF-kr&3SlF}Ryn}u1#co(x$iEip>d6gtI8D|@~jJyE9Tm!$9Khpt9gT`yUt)zj|AA4!Sr-lcY~<#9m2`{OC_+$jCD= zV{*U6YD?5#ZpLaG$vKqhjm#7FivBv+lWdLo3S!Y(?y_z`X$!Gr%xXCdVE`q{`3$fg zXQp3egLS50eLB8jIl#1*!AMTo2z(2yKf}J_pACXlVs;!mLy*FSJfTlB8croqPFlo8 zKe;Wy>G#=KflFc7`6b56*re&bkq12*cVo};s|M;D(IqY@7T)yQgHgb->v>^zMQ2B^ zmH>%bjBwQ6j8FZl&q^LbWMs~e%tJcQoFbY2&NAz+g?zO3VzPu=I?b;^{;QI|snh)9 zkpGJ0_wF=*FXZ1K`LmE8&%c-ShGV5a_QNqUp5uGy=N#*Kg7v)JdiKfJ`F~jH53T3% z*87;1o@v4BiucsFQ?ueV9sWBfc+1xqp%p-d?6-VjY=U9o1Cy-4@v=_$@#s@>BeW8Q zh z5QO4GRVO6)IS#(9Q+ddaIG#_`Wmt86E!#%abq(r5e(O-XzMm(tHEWCp?rXdde0jJu zgpNWNai>)L3oJP+FRr{Kx(EfQS^G=CXAf?RJ)`%VI&Cy`1&rvC_+yuA9V)P?+K4@A z(|Jo(xSP4IGPo(rxPM`AJN`E3E;K(06jT~7FKNQOaG8d=)?qL@iKc1w#=XszlSuRU zwJ>*i-B;D-!b)>#<;E|n4bSq*jh|N;o~6~;<37F+{sE6S_gWY@WEP9hL4)GRuiUt~ zs(i_Whhk6K`!c+rSh_;LA3rH}dtzVL4SVGPPSI}y2VM+As%Tri*s|*Vx}#~*|4P^w z#4Yt3Em?~^X}T4t;u)@v_FEBO4qR9^ccJGC^k*xKsh^<7h|B%>F+boDxQ|OHLbvf4 zDlxuhJ9ffX;;3Q5d@38UB)<0|c>vC_r~(KhQfWAQx$N)HAe$v0H`@mqZ@Gt$sGRh@ ztI)4{FP!v;*)xZSW1xgC4!%ZekpOFtz0$2m0c4 zmj+aESl+02m+O68^8CYSkTbyGq9R12wdokgyqeCvwt^Ll*J7Ffv=$NFRhPig&|q~9 z>?PI!C(yGeHx@FQ#=F$aW!vM})-@mmSKJeCm3L2C?=E$`n`*r~1Mgx@a{d=@jrKz- zK_9`^E>T7VwCOk4+BJF*K~U64u>HEf`_)TdkmaYmJ;Sda%zRtsSCcc}9tvTMZz({i z=g6s?paT2E^If-*C%4AFiv1DW82eoW%!S*Yc4(8Sk4&2{jvX)H!7d%=qLjly*4C^! zHcKWA=gVvk@5V1MWv(guLC+SMH_=>V7&;!fQ1^D&_|LOw@t0mKmrhzGJn@Mk?= zB4Bz}gX4a6h7JDV9f9A<@|{fZZZ>$CFF4Hh9)pWp2jxOblQ!^1!+z3$^yy*KBQ~R8 ze#%q47So_h;0VW*fWC)Q;t7y{T&;p5Z>YKMUQhfzNq?x*Av@h7X`=zrz2dPpOg_zz zksZA_Wj_4e#nO$}qi@7a*J$5P=fh*4b~YaxGv8u9jLv+E`7n%cIUhd55n|?iSR2fo z53g-bn-7C=zRyEMS{i4@KgCTOkt?{v>+qRwx0?DNIWto;^(IJ7poj;<7c|0S!jz{L zvj+Z0>(!Mk819REAb!$TSk`&<=9_gxaa?R4?{YN;7^+%Oqk1q3vrC^0SEa~ z?9C;dvkrpx@wwP{zvnu@=G9n+!l z{2L~TXiPC?cV98+YD>JX7fz>WOY!1JE^iuIf&`2P4>=(&D?%MUbERXEo92T-c?}Lo zCW~-^bBf_L#cef=kUP#@>kV#)t<3J2|M&%!*fM+%wMMr_F-r`%hvERCrwKc}TwfghFB|iTp`FSBN`XC5H#1R;@_klPD*%V}f#X$`|1d%YXu?OY0@%UsE z4rSH*e)JCC)XM2Ahy=l{VZhc4+fgyT>QE6}T8fokef$raVQf_I;mJK+u4<8Au<)CM zAKtDP=SHVNQouoz5h|6j()`8;fN}~38!XCjYNb#$T8R%n+3@pbp zikPq%TgQuQ^$W3<7Jlse9b8uA0sH)ulgl!d8R?i;Rq$&J4@B(f=D@|C@YE7)$vx#$ zd*Knt4dKA}O^hVsB#LdD*Q7f&EFU60c2n7HWU%<$F^g8}g>Yo*6w>YBB?Vx9WwXzv zZnr~K;Sk;vzNdf{hF{O)18*n%l~gU)8MIM*xBUtdblGHl#N_*1kp#VCB7ct@zUs~i zjb}EVQG}>V$bks4AdomoDo~QtQ6f$EwzBFVJLw5OB3%Xr<>KVC?z-q2beMd#ExYJg z_74_wP$bJ~I9F?&rt=+-ad4&|;ypSv8Lfa*%IE;TDYU*pb;-}b*~2HPClACrz}+d1 z{Bn&@P(DqA7^tj{L7%U~ zC&JpN0+-NVsrxS1=zMzw(CEg*XJxM-kvhhKlj8XYjuZAu*NK)GOsTL$b*Kgw#dCbd zCw1=$Mb26OwBXgIN4?}vS|gknLxTl5VZQ~HCipRDz@R)aLx(L#_;}ltLxTOwl#l3R z3vU)ebYcK~7(7OG_kOlM&ZNh%_g z_n_r*_>8YHLbL%hW{!tv$F_L>=K57G9P!VTS?^bGt;8(Eq;BT8MQ}|Xp5XO}`dX_z zS)0TC3xX@(3@*tv`xp4t(05T7Tp0xShV~~0IpDpk@ zN7?ImfIZ{h5v2|XnaWvVTz&cnMkk@$h5&&Tl!wTpJYkUSI@ov0{PFPxjVH8i=#*)ymD$2n}p~@-X#CyX%o!>qv=*qZl~+q%I6PsCI9%r*V&wwqw1u@T6)SYEi@9-_HDR&ytH-~E4J}nadx6%8 zr6SV8%VW06w)8B}0qkzj2Y3s^P@3zRoV|j?6BeX!b z0`lZ{#$@qtgB5n9ZwY+hc)&ba>=m+Qn*NVQqcJvMrQjX;)-kPLrXo zj;;lYlHKaf6G-V6&~A09p&C8W>&UEZK-PA*?NFSrsQH&m#oyc-Zz?7VfV}?GkT~2* zlrcfW7UmQo!KuSp?N}}zMG``n<^ZV%`S-MfR zR&BwxTiV!6y|RJRcgYmW+jJo+waC@#1d<<2T#w#|oWe4^M_pUU1%_o+T?h@F$%6O}L?-A8z_g;u9 z)Ft>M3~|8PVcO86cV%ObUQyUN#c}=+Us!|PrMWJ+8So|uybTKpuG|KNMe0~BUtHDDgSPzbvn83qd;~Nax5_y$IpMk=7{t0!Q60NZ>@Q2j$xIpVr58z2#WGExQi#;txf-Hx) z_^q)i{1fJ%sr++KQg#*mp6`oHguSA~j5;%dOC}}coX>2v_Q24xou(C}wq^~ESl-|^ zPc|gInkD$ zY3v!^)3*1#46~r_IB)s;+TN2^`=W)P_Yd-BFeg|ag8g@m(mCFuYv^=w{XJ==x>kMz5vio^NgVrmw| zpzm*-!bJ5=tDnaGKhU39{%kr1h4xa$h5(T$;`ue#q29#0p6IRXNqz$6JA5`LZrFur4+S9lnqW0^k-$8r2OQC6Efna?5FFpdQAIt%*3uZ!(AB{Q$N z2)dCXZ!w4!c5y&it}%7Ae1zQ{?`e~RUhHMDt*y!WQEzs$*4LE(sFzg4L))Zxw_M7J zY``f|ch24(2QuJC12?vBIIs6Yn>aIbhHa%CRG_;^OgMkvXQU=o=l+Dgd>G%6o$0YX zumHeiWOSjT)K)eIp{xn`4(y$^sD$ZOLJ{{17>zOK?u#;T)xc{MXm!SZRhNEGlg2!f zB7YRxKMLdrJn_QMHxehTkzEvDL^`W-Ga@34bo$|0+v8OrpKObU*5Sxmv>9<+0qjEm zWqnSS)9d>B)g3J`m7Zoa$jyM^-Ml7tj*Vy^+y2RgAW6ufd8+%5Y_W467dR@Oe+tNv z;{Asr-fvX@V)>L(b!Y<=f2)3c+7gq}d5zQc{sLwKSw_WVM{4xvMQ`h$t+A=9j?a7F6&*^oyt8aZ3Rfz+ax zq_PHFQ-KNg1IUeIzwBP}qIL!W7=YIBRrmsp8fPUU6`~Rp)w-nU)3k0;lVWP9Ptve#fXbt#T?h!+~s zM^|Bl$tUziePFz*_Z2-+5u)gzqU~8X9t9CkDnbRS9ePwpUDF?663)(Ja_0x9+-lL+q%Ow`!|4$=XVFgO}5TqZ!OE&q&1X@ec|gM37q2KKOxx^l{QnkTd40Z zwEbtjk-9y=BRtsNSf;-RyicM9lNKlci}u`RS!_hg=Arvk1`NVGl4l`z{MZBGws3Bb zutNY@JO{LLP7P-X%wTSSWOhZ$Zo~K742_h15Gv$6EkcWFl>^sClr2yXzeRNcy~qG} zq$d(`*Q0>Xk~Ku{G*4bx&b7Hpy`jIT;1?c#Fh))!#Hm{bSd z5A&2rPVx43JkBjzJRt7|6oQ7hz=O-&sif3q7I5LRKvNgNdElWN7hY;HcR*3XuE#v1 z(1(y)X*NA;E+dnqMF$X4o7gd7d&QmDGAnmDcVbS@={K1X$+@8WO8L6bZjD_G{FC2L zmGAXXJ7(aOZ0Dp!Bn*e5H8Ec!r^AK{X38q~2LxK`pGr4uSdbmqJzQe;Y$7Arut3YR zxIVDT%8BTp8pM{HZE$9*xVB$K^l}R-zuS z4DV)cFb5Rjx|vh9qUNp9fdC?MlsbM0XC&3LZtGWpI*`A>yN9bh{37}1N?sbG1JHzD z{}0W^myf-*b}_IN>kBILOu%eh>rJ*%apDMI{X<8t7*fC(lUW z7n`O!obiiC+;B{!Q7(1ff9xqtmib^eOkpxxtg``x)aSbp-dVAJ;HrohWAK9Atf0aV zlWjW7LyH2f-U~wC>f2f(Cx8;1W>lLXo2NfM^93ocE`1ey!-!?W7b$zULe_NA6RVcj zF-b8sR{}2NKOam%o281yxWSh*zev@i0?l;@q!!&(*RoE01s?!@W$8}})DY`E4PZsK z6cD{sKWXC|Ql9(JBINV?3O=c_enbDHUp_STFTwql+f;&K@m#C;a6E~GQTAu^+q;^^^3T2s(+nD)M+3Iz97J5=6-Yu`NBMD zT0%v`{_BbTrj;$NL}&7tV;l8JLeGCG2ECW^pF^zW6mD)OyS5u~Hs7L_mC{0OmYe{y zR$h{g3?i6FPD{6y(x!T5=xz{^)0njDmW0&3uhqoq$WADdgeyw+uNP~^m#f}A8;t0>>sNsa;1}oSHP)LVb(!_6Pz|+y zm8dhVU%9G^ztE1p@+Pev6v<}L(hf5i8_kwjBmdlyn9FUkKluAi8UEU@ohPgb&p0}> ze+cQf-mwf8S=D=Ll$ zDzL$#BM~AG%$4_eC7}l{MhXX5g)mA*Q4!AJeXf09xcJu+M+G&+{XMt% zsgfkoYE{m9A4^ep3TA_)-Ewvwu4(eHVPc!0wt}*KP}GCoVM;_RXroOy|XacgMAuJ=1Z(kFH3L>S`;k$`wZaC0LN+GHj#BRBK z0{cy^ud0w0$ccu)Ir03$D?oxN$f;Y69ee+B?nDoRbuEc-3IF8qPceIS&6#ZbMs)?A zY2Usij|vJ*sHh(SlJ8d555FTTQELOo6BTr`LzKinIs z&JwEOT9qUuwy6LDxpCu-)IK+kVWl`pC^*=d66Cj7bfGIW8ruXA=89jPdo7F{v7%wz zhEkJ&=_3jX++fG<(<_v?@s+RBb76Qw+p}z_o(tlJSIfGG%h`qAK19mSJc*PVS4-m@ zVyaY_nu`>+z!3OZVKQ2YR9(ClfhUZ~Ls>H;!C+?AM1b}3#YTw!vDgDc-M!M7+<-(4 zwt@$HvNn2YEg71NH#{CSmqD(%l!VBm|0@}Uv}8Gu|QyXTM=(oS^k@@FY8~zBrwIp>lLNf1!6vb4$SWXgk7{ zZWYqrez`G)HliMVl7*p!v1I zHq2^KO?0Rp?Q~kbK4&Y`Zg4^M&T9F+XxN<#f5#T94xT{SZf15MGvbLG#h^Dk%|sy+ z>ygO>0{vf~^ShO(U}BS%sAXb{l^D*%b}KQ8i7f0^jbgq1bfziwIMdV9(iD3V`G|-= zThg|E1j?iZt0As0_ikJX#it~$-h73)$hJddFyHZHHN#-mmz{;pRd!Wl_rFBWjt8S% z8lnB46z53XGzUKz)XlhU8(<@-TjL>~upUpxI(mT&(n*wZ{%3s&FYkkZtAh@L3#wSJ2t7uI2 zTrN=`9t~nW$7y4y9VPT0E|05Tq3kGR=a#1qdun^UvVNS~hltv6;ln`HKw5WkoExf{ z;zs%kK<3_w7^^AYXoSMZuW5kFXf=v(V}(V$8;+{TBp?9N&RA{aL0cq$8Ws>SJXRat zU;!?5TGrcb0J<9~yH!@0jq1}c$f(-b!cq42PqtpzGJV;fm^m3nNjMzoP`=Lyx@<4` zvP1pT{xalVjuy&CSn$=;^3}bn++MPlp_ep zWk6|QhK*krG>$w`qwy#8^6NyS-X04kJJh^NR23;3f->w)<4U z0kkl++pQJS?H7^FZr}e|r`=YMu~;X4-hh*2O$IN35((icdH$pxC1sKv?!k{J>@uu% z+tN-REC#9hORy5^n4|~2mN6v7x^4eJ8A9ItX{T+2%%6)5dC-k{WLS;sW;HGj zs*d7{EGIqc8Q4H~s!0BiC)n)@y_(UkyI5PKYz4}&U6<{sU2V$0gLcjT9rZ6K_mU|z z3e6#;k@EaW4dAIU@gp4kwZx*a8L|6f561q<#GwvYoG)zi z8}j@~y@02T_YA@~)q)|mPs?!36@$hWw!-qd-Dj>+<5+Beq^*4d_bXLK(|YTIpdmbx zi(G=M3C|iNB4;RweK+aRv9y!#d#O+ z#_SF$=YWb{t#vr_f$JE?!oi`1H3t`Y(zkdOR-e)Gcg3sTT&?>Gi{yji`R#*c=mM|! zidXVH#bZ>f*bM93E`RpL$Ync?rjxsWaN>eNl=Ijv%pV=y;j?$3cXH^D8QcNHK55NZE^E}>I!HVu{^pcCv<6o1Z0UJ^ z*6bOKeKOux9l|*O&T&Q{fqGI~r0n2h^b)e$V_YG$mRCG*MfFj*cj^Msfd%4uHx`KV z??UqdrI}EuQ)yRE^~OAh>bgSjW_4`6_dg;4G#64A**5|7N0sU?03aK4SyeZx5%`+3 ztWYYKU^98@139PKsBZ43S;!4|vZ|}_WM>*YLbO`Z>Af6>`cx~gWt*UzEu=<SHG_5--%)SXjE5L}A2b)p)%f}@2nrQsnk zRkg5$!mpqlJB4!7@v{b^iq`);3_))e8V4w!>$RKQ^Xz$C$vf1ueRV&cz*8*kk?B`g zKQTj7tp}Z;AmdkucY`Vi^a6b}Kfy1UJeJL2IX&KvnIus~K~cF7ct{`)?W&#^_Hf}S z{%Aeu`#cIYl?ODbCj5Et+Mo&%Hp<4 z*?)hVjk;-RIONr(g|J=n5I7S3fwan3GAOW=Q!k2jF|L3 z2NeZ;T~UDBvrA9bJLHup;Y@RmIn$h9&U9W{x-c!xh1OZVA}w8;mL8s#9+j4^PfJg8 zrm>xT2~-U|e__T|2HQ%i8|xr{Tn$FV%^8iChLI70 zvGEullLt!3647ZD0j^6;eVJ2>NAYBA%VQAkC3r%l?2x1MB%OzM9_-VGW6c+%2{{NW z;0EK>cR=-#viOmdFE^?#?~$jz;I0Q?vQ@B=X8hz7>sH2*#kwz&|3-Bazr^&-MaAku zcv|z=h=pwym^c9Dx+16~H06>99eT5&QQdnlO;)Do>k#iZGb@RI4ZX=#^5b9R1STDmYTU6Pir zNK4nIrH7}bN2R6f)6&zNX=@H(-&>uvp%T|x476$6rgg@uiuS`+!;-U`$TO-U6F^cp zRnp=F`9{_!f}<$xG*n~V=RN}O8&x>xmcbZYKWDSr_PXHX_OFv z(D>Tvh}eW3#!LM3D*rU`&&#pju=T}l2MefC{ol79T)z7M_4VL}rRnRzS;wcY2X@+? z3wGLN(Vm^oOG_7~rAyM%6=~_(wDj<_^r*CSeOh{&GyQ+fdH`Ra^z~rEv(ELP`3I~A zdtxGh%a<1F_231(+rfJ9#3ES_Iu6zA!TPoxt_K^BPhAiE{L)?zPTxkGiT}DDxK903 z>%r&!cD5eOwQ@7ogD0(wjP<~jjQ_eGTqecRxPZ~{KU)t*0+VF@wb;63tOsY{rE@*7 z(=LnV>~vmQx-czWl9sMWOV_5Qho_}SrKRiB($k#j|MTlXkJkSW>w*6Sy&hbOC+QaU zLi~mGphlh@s@lrQQ+?!D8YOT&c(Q7z>%j$cWj)~e+Weuw9n3u#7aeX6XP@pAfJ=P{>|CUDCR!g(zMI9%Y^g8L7} zJ{aEglNLY2Qn+8tWA?zd2>W4&fKV@D~I`#2jX#@pmdjW=mMjmV=rvGA(j+0 z9f`^)B6PoKYrF!ki2UusZ)ZCxxrj-sAX_~#WjpOTblY?wh$zjIt@@s4*XFaRUtLl| zE@t;yo8qyfx#)GSS(^vnes_0^3NHK+11bZVr%pu``W)J>$G$4FZGbN;${0ts{X@iA z-!HyDU*CZ$S>G~@FJl)}U`(Cx9@_$(u2WN?%MsO*p zeDm}_PdHf)RNhB$Lo-@8A{E?DTesrdWa|bJ+!GM=Q#-r=E`U46(YlS30S?xUX`Nl- z{qvoDlx)_TmCq2|)fwP!-UV>^tCHX@Ah?4v!2N9(!0qk8-5vz@o}B19r|_4aKi}EM z9i5#!3E(CWoJDt>1n_1dINnOY@DbIy+#{RS;A6OSV}x0a)hy?hAJ|3Ow^pP4HIJga zCV5Z+P5aGi@M8j>Eyt$Lww|yHZJpuhx6GbkOz8ax?m7oJJx<>b{BLcYETm|+b&*_W!nK(SYL<9rWq|9o3*dSH-o-1-DZyJ?KFCV1nD9txH*tj@$)sAAFT; zT@Jx5Ve3*z;uBHx{khJb=0w?VpSGxd&pQ9fhw{q|-6@9O}r1=QCn{yZknb4v-MF)+X35IuSpz~Ka?JFL2#+mYtebdTFAdf=E#*1UGHm(sh@#I9uK zVE0~)wfWfYZvQgb(1!szeRswkBrGhj>p5<7!`mg``av0r2BVjZ!cW3gD2ue-#)%LK&rrXgPU@7VLPMOe@MYSD8r+L;}D z4X!B}wd%rwdiD8zXxWWC_`EX~iG|?;a9gY)7K}}bj>F67P5d*Oe;Ck6E_sxz^>h>o zH02E*oSwowAOFz4I@+Iq`teUM{yB+%c==JZn16V2Mf6bqIgo$y`6rKm7{e^O8~!Bp z3A?7RIKFNN`ih?q{QpT`amS6B)IeNwAnd(67k#ID@{aWtFHQMx`ifz&6}R;jwX_?T zX|DAuX%x2($oP}`ikJGP@PM7NzlDyuFcnq&PxKY1?m~T&9o+j6)YqB5;+H;uzWu!& zTw+-Oq+bboOJxv`r-Iul`~KB|j)?@9sjsL_1-Db|u5qlSB?OnLugJ2&?Yh3=5(ocY zF$T~x^%XOE|9Mo-cl13-aGCmw!KvVOipo*%CAsX$1ed9=$VmmaQ&e8%K;;jk0WMQt zF|$0Svpa2F?@yDhyPx1P^%eeI0C$h0b%O{lQ(sY#3T~&J&2~r;n{EKOOnt@DUMZd3 zDY)?taL*E4roQ5qT>y9ECrR90NN}0@ieq*GTsKGS_9VDWeZ^Nj(b=7fOml<}*do)s zFv@3$OfQWB@JxNhL<0Yl``|qtWOFIOW$G(R0IpMgh4Bv1wiA7YL3o+^ij}AQx4z#m z#Ao-t`S#BG{@V2bo~f@05_q!jJDoD;f0P{d%Ly)1Uvct(ZR0ZuKX)Ont?Eso-{s zr9&M^|MePx%hXpaI_b}2$>ku4rwA@nUs1OU;L06iIE3Ib^%X~@g4=0l=Qzf2H-gL5 zSA2S6N@sTp?w+>fP`q_Dz-8(y8g~KQu@2mgB)Cj{Mfom(d%88*x&sI;ky=9i6?E;4<|UzFh!U;sAFn!DZ?zww0uGcBh=}21jSVz6#(n z^%XCrg4-#feCLqTrV?DHzT#4X+v(!lDs#bR>ZgTJK4bAMCGbprMVDP@YtskGasF;3 zz-8(yUOyoPlRNGD#)V06_YhpBzT%2q0JqXHFjWM%1ARrF1@q7AD^7Lvd$m7vhnmz^ zbcgwNr}~Q5@!6l$S2P~+ze``y3GCt)d*zL;bMSGd*f`N?;yp3;pckZ=?aQQDn^j_5VJGud9ks~VWw;qYB zGjM&=G14kdW}K->$njb1vby3*O@#<`p45vVw!fE`AD5zi#>{iG>Ng)}gy;}bzqy|g z+8sZjS}I*?%L4BXYj1bLrW>p~NDPQO@1}zVM+tp*ti53}6xB(be z&uZ~k@p$Ra0LkwfsH9?7CrGBFtCJQ>S8u+;(bWq?f^wJ`pTYLymKk206kJr4v@njZ zq#Xta7^92(tYAqHJpg~Keg#{)=m2@;zZwmzX=Xst%mG;9H>5!1ee?5sp}4FX=y(K-*!m)|4&2afP)5YBrCC(76z&tP1YP3S~@ zY#&4&&C@PSh?W7HuNGhU@*-dO*20?dAK=R4H3ohR2e|+&=@G07&U_~yAW_W^^qC6I zXb56Iem~D_^mHQr<7)UN6JBa)9{U`i7QpcwmRB!&&2M-Sz#Bv(E&rodoR8OsZI7D^ zOPd>3jQuMK$(Xskv$(01^Env4a`-D54UgmJd=LKk)x07udyLb=iwwqM(E<2e z`_myD`k(@Y#zo6H2RvaQd}MBH2Amvpcf|n|o;H5u~ zPc`Mg86hv72qJv}8oUF(2Z&h527rq<{2N)NKYOz}JmC|)h%f*rVlRX1_s5;;YjItO zIu5)87uq9I3L(QK0s&Fs!fvjP9cJ+X^F&^;V~%Dlu|kh|I=IL}+=$m4sKd1~d5Ay1 z$z%4072aA;`P#YD;PD1F`&!hV6KV*sGgjF95%@ zR_gvw)MU?CY;XketME}AXZ!>UiS|YryFH)#iof+_MLod^reS zZ6BA?ZlBd|M577RNMq4p+Aj0t*O4SY5iJ+WfFG z$g6ky7MyAJJ+!iX{pdB({g6Ekw*ZMgr-9!gqIY%a(rWXaN`&b? zGkj3qndU+JS0YOHnkW=n$@344a+Wc91o?e}GC_J+rWSY>tVBEu@+QoV^H4(U_{^vP zs0!zrORG$yD(uVMT4nl-t+JR@h4X+ibN{OF;1awZ+;eLbTgNzx8gKb>iAJBXJ=#Hr zJ4@3+9O3X6!aC-Y>tLt+ZK8B;pXOg64*4E=~X<_jYy|C_^B z@j?nXW0^5={N5&6{hIJzEMw83c>aS0NRbzcLSU#M8Jsxl- z?plck(~bObPB0kQVm`EEMae~+Y=9E1032_BF+w8O6n_T!2d*vJt9&Fslkl&X@RvA9 z;gE5-!>Jq)$pP_axLSuW4D2N%z@c9wa}jbp%ed`Pq@rKpk7cKt6D^d^f(P|tkO(lT zc>de_v*F}dY|CRrn%%X#nBd!DUOgVM9nD+dal_iMwzDoDea$DL1l-f$bbSaQ)6@J) z!Udn^_kU+@F^rQqlDYEY&1AC7vg6O+WYRd18 zW{a)0%f2p=!eDF1bI4deW_R@P!AVb-cJ-CZ7MoU+jhl7Fq7`0o3yAn3;MUfN_m>-I5}3JYGrT zOI$1UA%+oww{a^~c{ZH=jG5IgTuU`dRxCyp8!4@UC=LHUjO=jniv}~Ir{3b(SOwt6 z^s-BL1W{2_Wb-YdDurd^OfrO~rNXVa>BUyk5l zS7V6q3vpbp(!aVIp^+T6U<>XAY&SxifP<>uN7OBc6|<-OKSt2FQCvXg7$3QZpgZaU-sL6-ihzD~?ZF%XnZ^hz8{l7b==)aY|GP0iT~S~?1at@gl{Aoi-wKWd}fBuA!0 zV^2YmZbuhs#}5bTc2r<7l-PxKJLVa>9Tm|9Y)4|fh11FG16N^M*Wo@0>}Gs!?)Lh2 zkv5Br`yFNw1;0@eQ1k_)1~&Eu^3W&YW158lt(<1jt6@DWatQI9SQ**4kA_yStO$0;)7&=uz*LO z6c8a!eY6~Kl9zO0OjP;N_1?W2MHe7xq8ofGPeF0(p|s37*MxqKCGiN4M!HrvtX^pCF5C!}x&qrX5J z{GZFV0L!{ERV!DW-s6ogwqJd$U)AH4Sop{^;qd8p3>45ZbGq2?PVBt8)#xE^CPj4= zOMl>$!?O4;At6h*n z=q0A&4`|GXKMVIy}?ittR#mNxs<#sd$75aM%J&~pny81EDF@kfazcT?ttLg3CJ z+aVqyP$juo38T{01A@mS!6N*!1f-LALU5ddAk$i7Zmlsl#=4_jHQ|W}%N7ru;|*Vr z4gN%IS=V4n+lbC(=lK;)=ksf`p6{B1; zO3gnPOAc6m5p5So;Z_Y#7*Z495tI(ufRM~k^3lJSx=j$GQ-4eUxec5l11cnhjyE#T zJsrh-=5+4M)Z+tjD~wGgB$))6G{A%(-(_^>HdYepw-7P3A&^|UB|e`!I2vgplia}QFoMP67kz~^t6m# zz4VZ=u77c1q9YVw{yCSUUy-s`o4-Tb(ZH{vnc$R10FV|2z-fs=V(3s$06N?IkUUXt zRyF`MBxzUlnegw#+=m0VBuDHE!WuvAMZiufLuI5i`K^qM?avOR2ZUF$MNl2+ZA~t^xtZ8d%-qziJHh%z!{eaMtb| zu2x=$(;E9s^bLz9I&}*+e!vy6%m&wt^n_2ZNI?*9WeWH2#e1}h+Yku1Fwhe+P~GY1 z@_yev?NWQogh62CMAMB1PDeC$U)>JU#!pF zac_w&<)3#oKB;L*_!O_92H1hpkoy_D09Jz&a`uNRz2Fm7?i@XUaZ<0FI~C;OI%l@VTt5eiySy)uhJEe^sk(O24kZR z!}2^A4pM*in%{E?;O8lizW&oEXI+ z3IzshA908DFnSE$5X8AQh#mmJ^50p3P&i--><04V`OoI)dG$eJa?Ka=@sE3&_rb~d zpXWS0nX(dC@P+^N86FsG$`FW8s>j}AV)M-q=aeNccKVIMmdb6Q8d)4t2 z?S?j-k0HW$!#TeoDh-!Uwrtl;>MbKGb>>32@}gl4e!SBgODsq@a<~;7{aAER(-?0q z2Z`ZUG?3W4kcErLT>{0Uh3e zX|GlrVfhx_H~H=U*0<}hi@^bHp#{=qz5XKgbrc(T?ByPgF2s9~f`C=Q6S3-15Gc9X zxhC;G#f^zeLO!Ov&H#vd&Oiu<-#}sk2j@c;ndzKDj4s635me$u9@Y2NL|)6|n49hK zyNmF{7k*0LUt=rsCrSwZEu!}e$tR5Y#cvoAL0_m{Bf0m=TN38vN1^k>4xbl@Sl^g0 z2v~iGh$j+qmq@Kb_95${9jhp>#oCh)AxyHxgrVDLeT8%`n~H0AT#|p^CiIKV(#7AT z$WBcQ0I*M<#rgxhb+u_&$cA;ITJ#vvD}j}pk%>L0a2nNDkyGOSix>sKb~TL8@T2-t zXt0xpW2ovwFQh_D-a>rpSKbjbq#5j+6?0LRXz1G3p9Ovis>9A0YdeFVP6r6aOJ2!} zBV~u+iDIK*KM79a6gA=jy(0T0SH#M=_fsGYvtLY8joWXScetUipV^ueBl4IgsOZ1Vz= zEMVZCy+EygwePD>$*Usn*`eFDE`%@M?pLGN6IWFBAZu^#GDdm${zB`0FZJOu9A)g~ z!0|4h;ff8>^u`@LHY8)9Ab!E!44l!7!Mk{n5Ue(5zzyAbF12bH7e77VadqTF$vE^M zqiZ}M-UshkLxEcS1G}V`w}j|vG%1 zr*7DiCrelt!ZCg1fyn$Sn5TY^6=gcQ!GG1IKg*@DI%*`}M4?8o<>%M97J0&VbA|Fm z(&C+gBXhURY2A!`b!C6pDBg?OwO;vE@C&;g){+N32DW_bJrNHsta=}I0b}~w7cHKc ze$nEIsTVE22WYrF{KUo{J);Zw8**YL?mNUi;kfVcliLahkEycnHL)Fsr=|{!LkTy5f!V-DHT1`T| z)HB5iQd#_WO@#&O*HxX3Y3P1n2lo&E@f}c^HsGQ({pdez$zW+oOKcGLnTXOWZF&L_ zIkg^=XNM}4YAvFc1zDPBgw6nGmL}fOG2JCzs2ndYwCdtcTh+jOv{8#+-E%#r9SmH4 z7vT@#PIkC1;+$cX6l#ebi7$3uatPqVs}@&@w2B)nHwdLtXN%2Xqx6hrcV+jWkE$^H z)brm#)!SF+Menz2JeFIrgIL;2P#5BL^j$0OVCEfd;p|#`O`KgxC_`BS#E_@XV-jbW z>@$g{n&F<~%4YN_@pfJ~ce!BK>4 z9UobXh9bC)^sBaEASK8)qeK*ejH9NB_q3>=MYBg4j5o{<;TjvsnS}m9y|G6bd{QCs ztl*!OSV}%JilHvF&Uv^FNsumzlCOZN@E?DZ!SEl~o$2_EbuNp$cNhR-sv_vkNSI<_ z6+#xQhNv9}O-;iG*fYv%jtq}kVtR6o;yVPgDrwd-2dm+{(UO%ct{sr9g5z=hcj#(n z;#ye&1r{ypBp|86TfWQ))gad!=Iuru&EDc=Y-lrEiGOS;mp@N z%RYgMk!bkr%t+r7Oo{czzBpviyvzPx-wxg?&5%>{^|*g}Gv;R`mUImtz0Qxd3bGp{!B3I$-cynIpK{y+0c>PlGFmACn2X-f#z|?;aAm;T8JOAsYJkHNk}aVR!^74a1^*EdJ<(; z@{N@Fr&Z<^zQMY_AD8@QCb+auL@v*tEloc1Z)N^;<_B8~rOZWyP=1ZO$Y}TwD<6AQ z%N}v>aIsi(3F;3F$#vJH6DEyf2?)W8ds{!hU(Fik;QB2uDLk=p|D4zVF0(?sJg-v2F-Mf zoW;72 zYRo`2S$@s4ctT>}<-EqPCMk@pJ=}sEYq1pXch0>_Q9@3{HQd?P~Iuz(k~M-=)M@ySf~TboeZQ5%6k4 z9A?N42D^`Vyw+2V7VxSC*l0jq+M#~DkV?_GI+ZZ+@hNzEaaBhiuG%^fl3HvkAK0fkColiS?I^B}ln4-%(e<)V!?3?7IP200T>LE--J z6cG{mU%b70d{xEO|DVV$T66~`NbBubg9WePBT5uBkwnhXgV1U%tv6b&vDH=ziDDH* zP7Y-Ec$7Yh)_Q5G#agYFR>3P6t^u!jN3DW*VQ-HLiVBL7-}|#>pOce-eV*_0$B);G z?0sft&#YOq)|xeI)~t!v+2*E%tVV3*_B?-9UgCX9x2*FO$BD9-`tXuQfdpKbERi`% z#=Q+kOZX0bI6}erV}t^zgndjV(V!eNoluFPvcdMx>Vviss*NIv)bdzx{A;fJNAnaq z4%ZXc@tBuxkOq<+`n$yu7)LP%8h=BvkHHzM#G3Z8u`WYZulhmgyRU9-ucg(!-6)Vm z2>syM%?_ZVu{|pK-~m2R-4rW)(mocOOD9Swi%QQ>>Fm&zOWoMlS4Cr=X^a&Odv(`g z-!RS%`zU=5l*jXotUavFb~a3ljvuf^XMM2f4krL1*Eb zOt~!AGZ)HFK($^TwK9okt|ezi`jCA4@mf5vt?uV!w{wRst>k^nqk%A2(V3+ z0jat~ealXbkW9~q=~PMkSYBOLq1#@1JXv&G)o`$(sOPhcuIn(&L8eUe$eE9R0OnFt z%1astu!b&!6G|HPif0#;izl)}S1+H#edVlin_9i5`&1$OWYuJRV5G-=Zz6KeM~Jaw zSZkqdV*}0W-QaDtI3bQrgwam=Z<%ETow9pnP=aAyC>r;Q24Hr9+^PAeS1 zq*A<~oTG1MNIdg>^L@DPNSPtXNcvo<=~`W1)Bc<&GA@Ryj%S8f)(4HpU`L_yO)tAq z&hey?!Y2`r6n}C;^hqm4q4z@3;2#Ps)usN5sPJ{J-s18Re-u}}zRE%KbCu2xy^Emr z0`A14ghQO*CUs!2omj*4AdfYP#^4ziF2dM>e6H{IwQ3qFWoF_KM@xYno;{UC_1iY_ z&hD>)%*=D+iT++6{1hl8u$^i1pY$L+keOXIKlPHUc`46`8B)}>9gS&&H#yBphIzof7oh!ukGMwlAI#w6)l+&LuyKoB5uJ6ekyK&nL4z-XMcfH9ii`9b|4N={v-_Oj(={NF zfD`s%@{Gn`<#&q8_qXz1T5`7N8*j!dBw)VXDO_F9f4%uzxbOHw!>64mNM=VIfJ6i+ zG3tP~oWOhoRPLbTfNXtH$DgRziC@&7%|ZM5o%`%PTioZL?u7SiFU|Yl{=7?eu504H z(Z?8nYk(PF$fX+dRtPpFq#&Sm8iI+muI-nEN~D;BSR%ZHu8mbm>q7$c=*q}bfZDmyAliRP8?j=#zq8w`p&7WMF{n09M-OPLWkl#Dyf6sBhyXAl1#&0y&?&kKp7nOMPhq=5< z9&*!R)e5iaYiKf_aS|YeMEC*Q11_HaFf*H{r zy``=OSITXFz>d*4^-8W-y>Taw$z{I}`i}BuexZ&(+}>>}qcaub{2kAEDfzb+P)wPV z7$cWcZA={@n;eN>=$PHTT9Ea`ZY#fKZrn5Cl?N6YR zL&!zo4baD(!JGbliiEw%D6D@Cf645CaDnCmuI<|#RB1D<2xpgF)fs1BJR!DZQMMs? z=4awml({v*+oXmeW^~OET}W{11&;g<=e%&9{(4vhSBFe5f?Zw*s=vYfJeRe9+a%2%J+CcHs8HWTEL)XnfGB&}zR3`t`c-8_wGv^zv);V0jVTrzf1**sf)SanyF5|h zV&j|&*E1L7Hs^4q+7LYWB(&8?rMdTVFXYzcp6#$OYyFbUURo}Hqk;~X`d~8mQbmR! zmaDbQJBKu%R(Wu3bFy*}&in@zc~R6>UNEFpTU7DPiA*VnCH@@X_P%ii(Q%?bIRH0?{- zBwNJ(Gwl_}yeX;gE?KuZr)Hpe_BpqaNSHTo_+@@CnVmW46G#+-{f~=6p6#yX%9w z74o|qG3TmxFi&jl#t*=xO;&eB$A`YKYIQOkGdSOIO0;tD`89kLs}IiQtMlv9?|4lj zdM(HBhD%l+mg@RS$8WY}%p_FIT9!%=L1FrLrKXhX_@Ip(0`{g)w3dz? zovkFK!#1x;vn8uMI<(eH{S{KsJ?v|ODW8cij}7OOVtVlQ`+ruVTe#IqY>z2TwgpV- z5?y(><@N)A2ym`GZ+>Yr&RNHX@4zSMBir>SV?M>*LTw?D@4i2tAJO*<8!Fii zS`&<9?zC8~`Sg#2!6XT+m?#+h(rw`b8=2QP&yF1Qcd4HFFBmh*+&i$a`@#T+vM_*@$ zp8KFJmg|Gt57P1#QT1#Xi*Shd`TF|6w+#8M{9(SZw*Mh2NLA;b(XZrji4r1jh?(Wv z(qsRtL5oG!V8B`KIgq;>$PFXgD{B5N0nsI+k!JpO7+98}bX8=Z?OKN8)UW*jDa!3Y zt7t7d^tQ}8nx6f(nvQJEY}x|Y(?97)#+Fx6C9atnQ7wY`Zm;29`DUJKjd5RwdNrZMhq@Q?A)M_5m6)O~fEj#pkzgH49oVu>XYB(^eVPDmt zj1dC0Q&o|N^bfx8W;5s<4{6yOsPE?hTk!GwT7LO)Vo7vDrEVP0o+%UAn&7TV6f#-0 z82i72UG=XwJgj4fz}2R!k>9Tmo_jy0U-g0Q>-uXrw>LPrVkv#ay^>bHctD`634UBu z-C0yuR1@_K`Mf@unXhR#MvQ7=`s@g5qLyW@mb%h+>URDVJ90r;L6PhI_$tIz=ay{qtqe1-3H z%R^Y7uMh|?++l^Y^A+x01YxzS@ay{xZ$Q{R4`IK2g^b8|c35F;zC!M;ETC`$#0Ux> z%2)VKk359Sz8iHJ2v>fvqb_fP0j$fvbrl*8Z=*t%;f+4cgnG>m;kYml2ZsaNrwuDxW=yT&2N(m(!3I8N9vXC$EUPmW9{+^i!yOSvVtl zPK};Fik>G$&-0__+0pa#=y?jy^Jj%OgPxA=+5P7Uf7|1-NZ0%@;eV#%FfTI0T=juW zdPO|vj_gT-&3&t-WXR`_m=W7HqT4GPp9jcnkKHa=Q?h(ye_k&f7|%zOVBeblAqJc< z?-tLF>>u|>mIWITcpN;(0zLG*l+c{_21)VkqC}vbz$+;a6^f}dQL$C| zS3k8{#r)J-J?Kf4u{^a&Z=`IAQg|(gK%xq?hZDUH?sy&6@jABSbzJnC@b?9*5`OWB zxkojR`rsgKYGo&5jAj;E7UWn-vYx|56Dd9(I@oWnNP`3Xi<@6O-%oD@q+tHOildHa zN(?w;#~Lhn6s{&#AM{z%q9bl{TE9T6jTo_-Vmh(MWWgdq9(m*SHW-gD%(q(!Y4EbD zNo~ag{_nied@rt{){)FJE6dJLc)xyXdH4NFyna}<1j!)mtii~I$SRaurrY!?eJ*|8 zJYC8+c^9W!l>}fho}f6ZB3_?ttNmg%(+TM6V7Ba~)DR0YdPA!bofZkKHhs8Vs4d}ZF~(W6ZR#f0 z=as9XFMF7bYT_f>>RW z^}&xQJ!*1v96DLZWEK)fuzS0@K?wYoxTts+W>tcCu^2}Q>GEIH_7 zIfFD^0%ygueOZDp_NKq8VU*wfB@}%q#;+kCvN|k%ZnRmVI520YyW-i)V34Xr)yJHZ z;QIuaEl>LB!I$xylQpBS@Fzo_;4~Omuj}e{bBK{ zj}>=oLa6~RZgGG=Ko_F8TRRgmtSIhQa)oJ=g^ zii%L@uRM_uX1i@W|3m_WZ3}-&8hoS3U|W-77=WYBM`4)R!WF4N(LzLVbO~=N191&M&tJV0AuWz|68*2&AT@ zZDGA2W6@R)c(wlb%TLUlw%b<%@T;&Dg{xr1;pBtGb<`0D@DYv@fY}j;-X;lj&HhY% zoNsz1TKhOrQnlG@()y26$_f8-KFS?Pe>Gb*{*sxhuB5rHJds)okDX-(Filk)U1%7W zu=C2lSN9dTE4HE1K+uF8HGXgbGR;QEOD;|47B6>q`xWQf57R@9tHleoyYZpQ&j~| zYmwRhWMe>i6XgkS%yLP&j9{!<*6;^g)nx0+Wh6a_T|DG3o`Ys5y>}6q#)lB2tPUU+ zm27<3N-3{#VfcMA3;bM&Z)g~v4XDT?*4Zyn7sqfJR>@fuw-X@c(V3rvxw9DYe>WWB z=ms2W{|{(tWuLU8@dnL)Kts94H}rwXt6U)mMVNyKcv-64(vOHPD6Ex-9BKikYUm!M z^3`8+E3AIXgmcKN50d*B$%Ru-)3nZ3{8Dr4`ufylroI~eYM2Q3aObC&A4jjr)`^{H_gpF#*sn5{!;E$r%Xpr4 z>&S(<2pwIA!+o313-Mw^e-K&Na3YHNcfanI@D6&>tRKEffmsnh zcxfHtup7rBc^&b|pK`4>e)lJO61JKDG2-nZYMUNhW$fePp|mHyIxK9;-1~l6wX+ zy?sAWmK}P}T&PvcfVL<3NaiWt7aB(vlTX9r7I53_UyfnLL5~ZEbM-D?fG;;~jBU$-1(GQHEEZkzH z2Dvq_7!@hTlnI^ zd00zD3~}qk`QJO%toSj(I8H127s=QDMt10tw_p!qf0g9*!M(kSu$AfQ7{_Z`4sB6= zeNb^I6_9`H-!(0)31;)Ts>$k!9>>0>F_&dg*2@0uHFUJ)qaJs7&7GF~72u2QV4gCJJqLW-!{1A{(- zg)VmN5qfF!{oiia)AZ?24nCGO_^-KqU@9WuYl(m0}nEQ#2H=-suFD{#=sI<)zpFfmjocmBtWEOO;a}eRPav+ zT|Z{r^nay+sjBOLq=E5xm~_y<&jxg*fmP%?8W_r)U!Z}%?NUSojSa+kLa1(!xqtmQ z!rWOFi*QGnySA(==6*Z>e}y@txj)lvI1??{^aqHonieA(3tOV}jnIzKSNI4ikyL&4 zVKEryPmFTQf}2`^qBET|$xz&0G3IeBbXYdw^&G6xmK#pwK%^*?Zo-nq19ER3l}Q=^ z!EJx*N^*@4n?!wmsFbFKCRY*6GiDbH=Co?)1VCd7dRQ(Xm*sW?g4}m_s}I)leW$ih zwzeU50T9Hj^~(pt2wszd+y`fKb%Dj8NVxuZG6mWQIG6BO+abcMo8%MYhMs0I?6 zrMJO?yVSvm+0wzb>!|;7S^~UqEro$KyiC8d71O_M^F!utP1Vc^e_dwuZSS`|#91CK zd(^joxJZF46qbdzQ-@u!Du8zhOQRw{JzZm_N$n*1)G?_gb^!HAgZg*0M`MPW^HG$- zw@5ZL-b7X51N?&Tb-%*KqV<3O_@dxgFa10$8P5wb zC9@vTGB~m5o$laCMI~s7Ti6B&`SpswoYhFf(F9f%y<)Te*79x;&gBa8)Hbq=2BmCc zT4AcfEjzg>PB$d5@vMi=Fz3^>GhQ)%Qp)x(FgS-)yZOo z_CEd(BE;*<6|h3*6j$H&}@4RMgp5qjpz8U6~x`ziFrly@TzUu?8ms@HW-Q!k|rkm-5&Z2Zu1!CfUL^jYXaw zCRH2Zj;t2t#?|d2%GCmI6W4}F^RANn_*U!K`o`t|nS9@7gpGW2@K%9*r}pHB$af#I zA>T=KO7u(Hj)1NCj(xTc0*25aV3TWSeKY;;8(eLQ;WyDQoxZ4ThxGf?`~D~T{n+T2 zSg4KrhDR0fCsi;W^i!+c0MN7#8JlXni?7@`m0O4hvXS-Ys4YUR!bpzu z2C=soyoFK3o6o^16M0Q1Kqv`+4L?d7|0ZQ9j?UAcotTKIg2v;+mkaU<X8qcva}<#2@-WF2b5b~RtJ@Ps%DQpYx)OzOl6Qs;aI$o5hGQao(0u-vd= zQ=}(a{<#fXOikOcxz=|yY%zi!J30d?d)6%Ph1z%}awi`#C7O-;1uvot7Ua&>##JVB zV)5*cW7(oIg6egbpDW z1Ug>U3F!4XH^QGhPQPHdgtDjpGs^mb6C3*)VJR=DwcBqf>qomZiir?P z2W4!B=<2Dc(L8|{aEGYIxYx0TC}P^JW_eHWHr4QUB23H}A)nz=FSrFyXinq_k2-|k z1#h`XUVCXd-EWUKAp%RWT?{i;Znk#0m+8C0Bji|&N}mHDPdEHl?vpXyAp$8QpqxFC z>&2vNh$4eB!Gr}dC%3nl_U-!{oyYxOmWA>UVjLWSa~o1b-^3X%53Ur^0TK(Ca_eQ= zi39EFpV&X)pBS@RLbg~Xyqc6#l32JA9gcrjni>0D3odHkO(kC-Fjzg7 zhX*5W^7hy0e!SlBPjKa}U6%)5`e@Uhm=G4mK$mouCm5a%{5{@*KUDEK3xvPr;BOdk z{i@oeOELaT&REDf;5wwIWJS6^1Cq#|ZTMR*{B4z(?TWwlyLODfV>O&z@Tb#}JH#Kb zg>1!i!0+&V6a?@M<=eso$Sm|Xz98SJ%^LaE2KNNcaSZ?1L?i*p*^N6A0sDk=H@ z=-!pubT(|n5X=1U?K9_++1D}~x6ize%+1qO?Kov}U_#3`J;&J-HDvRF>?gYD&ZovR z7o*|q7iYz@ugx#T@~3$dU$U^t>2gS^Qr}51ouXE1E6PlulkPMvu{Ls#?jS$>YuX#m z&07x3a@fp~V%TBd0Sa$E`ilJ9G8dKy2dEK@xbf`B!O^AwFSaRALG6)R2H3s$C1cGz z)K3}Po!*t-Cp7<`htS;Do|nfSzOrQuASXT}+VSPF(GEjoQ-p+UpgygqLjj;{&MBg9Ye z7UI;uc@t+@SrD=b(%$%j*2B1v#Qo9bNs6bcN*cPK?Dt-p@K^C|<5Kb=4>p*pUTN4R zzUaMf^})C2&|7w{F-A+@ro&Rr)oCZMli42)#^^d&JsIu47>Z(TzFCwT(a^I#IK6Z$ zVrU=!(k(ph|3*JEOj3T`!1`cMS#PxWa5S>m490I)76%ix^>)G0_%k^@mf5^_=R-87 z#V0Y0p!lNIXOl&~kxdwtpV=5~dfM^gZ)P?vW=I~aH%y&mYMLHKjc8n)^yFzTCKHu1 zE}t^qUi|KrY{Pj?TI}chZgOi!Qs}lZm z$?TCYsj|^%%}vLnVI(q~$%~iO3|rA~s3L32GwFMzd-G4h#<`fW^cF6nl0?;WoHHat z_J?6B8eibUcy=613zpQMN0Wk?iw7F8+8ggC>c{x1`yyq9(GCaq`Kc018=a}N9;GF3 zA79IN^}!_BBAwEupinFmiXzI%glfx>>bUPBOq)j?t34uN>FDW%I8)qY0)`f=DS$ygH(V^-x`n8#G)vjFJ?@ru-#(_cj|MV| z=0uP>3XZ3h9StIYtRg3vv*~MwNJnmeXkoABsyMB;@|4ksa!RVrOo_`q;1U(j_PF#U zY*Fl{;BjG_vbsgITk!CIL(JKcnD{wEh{=LT40kFCXTu3X&}3V3W}>bUD5+UB`|_hD z&JOS3?0Un?@{vh^HkVhD^{Qxsc-JBasRqj<+Y_&s|^3})tqII>sIiH zSAF3%brUIo=89r8N1ub`>`1pFhIb}#!4%tGVrS{9y<5$Z{n|ByHq7ys~*D zlmLH=lz{Bvn%lwn%Q$z(HHw21rO#;70xbU0Bx^VQ^0_X!mbC~PaX8DVtjsqB79x~? zyPaRm%Vjzl&vMOBeGu4MD3DkdOKk>-ycj67)Odpcg00Tf9~if_p}a7Hg4*Ak!J|aN z?`clz>d^Rj_G%lFi1n_wAK?vBOrX0K^*fspu@va(I{gY?m(r1aR3wIo-z4ncqw!8= z9u+;=g}ra{5_iAj8H|VgG7nBKz1DKLrfm&m4`T$c*mFbhtg*pGa!XtvEY=S<@(+F@ zyZxGl_MxA{{phw9>IOr&7b&mn*f$Jsuh{7>?NTnw-Oqqx&(2(qsB5=Pd}pmwJ(oxI zNU$~B=Y(4BPMb+?+INmVch+TUn5DWN)_|9rnK&vu*Ih`Rmy)cz%0it8Cqx@{hZP9M z4sD`%K9Y~hz$ACJXc|KH$xTzzHD}1_r#rG2@H)*7r{%ugF<*=-cd-N*;c+Z&MrY~= zj?P>h!zC=~zs$v&>0kM)k*PNW_M_ zlK$7#9IH5ARgI8L!sdYA*oO9sCabDp;E?Cjt#L2$e$tC`mD5s$*%sV)-!B;6^7@P- z&J6lEeibC#e&v6TZn9t%}X*w}D3%%;)xVcA^7tQ)` z`;gYAl?{W2th88s!MkaBX_H(KCylyQn88|o23@NzG2t-e`QR}M=3Z*AxPHz$=Je`< zTiZ)T*N(UlV9CsJ8q}gzn=D!qQB|roHw@*bE7N-l{zYEXMk9q{pQ&4VT!1hWH?b*tR#^=p zw;D)*nAuVO;@gPaL0j?i!ZGb3+Ke@EL38|EG7T9T=u;MBtDg8>b|ATAHS(>?_U#o+ zd5v$G7Yq8FRKTQM8tgtC}{Iir-SLZ>_w@M%5$d{>=EN=W%$O}?Y)Sa1Y>8)o6tn7FErUe>sdt$pSe_H*JD&Lw-=DZK zUiD?;U*+CyHGdnLjb~3U9i8c!`V!ahe|6_d=EYpeTu%m9U=8r*do|6weZCmS;K#}H zpXW;6J-3}-<0t!ZHNmA@99;k#OP108B^#5#24La&bgn>;(EpC{EF{BoiG$Aq&!6`L z1yH}g_&?FVH+^6F z7{3m9gl!$raAp*dhp9)AI;WAyxYsT=4M0=CcyHmVV88tpZURB9uzf@duvqx=a}DOy zD!g`kkj*e@;t!GLDZUIZq9v0nS(YpE#KaF|EF&sSK;({c7N;>wiRo9F{nEzN{?Y{b z4BY|wthZ{L{&JPj?0HuOUZtK$0L zkIsruv!0m?C`65v+Tu-5YiLsK49Qv$#Uuvfgo`ygnOmHE z$rSUb_m5jl*yNd2L=->w_4;65U&p8egtK$GlFMDZ&b7)(ukH(`(T4yCWv`ByC$+1E z`Q)HhIo86*@v7In>GL7_MAa)^%I-7NrKvLy6oaI;RD0>Ym1&pK4;_0so)I)#7M$=0 zCz0CW4L93ohfbZ0h#VAUyhnx|y4n!Tcs?S|voIMN-*B(|mi%StTL8_R^Lqk4m!a>4 zUw+Tp=3e;ZH+cTaI`6?+-0^GuT5sDddsIO6`&^OZ%3_)qf%{D*?*@yP^FzR zF4fs6rOdKMe}bMYT90QRuvLv|T9=2? z=bM8$$`gHGjbb4l7il{fsxF?juL&y8Fmm>-;2Oxzo$Yu1Re{P|k*wYCdm8Qr^yZ)u9JsPT#YpxgUvamH%BNh{RQc zoO9WNbwvYQvV5nK@Kma%w#K7}w1&s?p7H$WYy90Z$vwZ4SpCX?{2Li{nVvf6(5o&} zyBmikYB}d%5w6=8>D)ys0SHDhzj74gAA1B&mbQOH+tt=KK{PflXWj4u zqouJgQJHpbkN70ve_}NbqDC%&r_zSrwV7S&{9S4uhJI;fr-t1)5VH3!=a2ID-X)P* zwA1H{N*i`Qfg^;wjBM_^i*PdOG=~pGgc3eH@X^ef5~1Ulq*)6atTwRrZBn4}J-qqX z8z2$fgGshYEmOTE{4Rjj`ipDYR}kZH&n-zmS-wRrO?;q+nDRuH_n%AYe*3;2bWx_T z@*n1Iav!ir+Y@R?_OlPX=@TFq_mO>rQ@zJgtKBRVy`7cupJ-o+8K@r^YuM9KphC)U zq-DtS7QKFSru$Jzzx&Y~XW0<`S=ISZD7gK)EqO5ZF5|clRX#uQ!(_H^KO)| zCdf!X;IEBp{*I$<^A4!NUs+Ssz@4PjWcmrVngVdYC($mEtM#|$g;9M_e?3Q-R}$q$%wKp&R#7kIIj+@+AD4PHcyT zhD1`tQnBBgE z1LhIE5O0itwEKL(W5q(dUDF{S&y7AVPq2!S1tI~>Dv|YR5}N?${D{@*EM>umYh<^( z#mdR7^sDJ$L`Q06L62wuZq~hgOyO<~LMPo{^e*^kMjFqMe>_RJ*&H0OPA0uuDyQi~ zG?thys{psh{y@sUGtw!bwB@UkE2xB#$j5e{dx0v)gnb_AI|4ldS?0r|4eIL+>Ozoe z)j81WRDhMU0joY}ea}VoxZx@{(AV%3urqijGT^{)i1{b4e@PLmrU!HhhXVaE8}>LQ znQ^y!Pvw0cEPYjzFxJV%ZocB=C)f62LV3Os$GIeSEdNW>D=odhr6)kN^c>}n6%kaM z^Fc67bqVQNWoCQ)wH{RDaZeH-NO#m)fj-v0BPAh1G``Xkg|{`q9cwj`;N4jwf+buZ zDZGmUI^n3hz6bK`&=Is#AKaqBWauYOb2$1X&jN0Qg&pFi3tFC`6`|$c)p@k!*GXLI z(1DjTi}13FbcYvp58azP6%09s7r8wXh5xdowv6ajnWGaSVC37kN3x5clpBJd5Z7>r z9o>s^aSgIzK(q{{wwE!Sh<+~sKCwwHu7*&nkHRIa8E^}&au zTo7Zn*LK+|0&{}9$a2$7CIOaO{W0H2+;>=0mT(@-{Mmwc4`YfdlDnT>0Sp)uf6p(HzU74!1=#-rIX<|B3A-)BgkJlw1DVz@W(v)=fwTl0l)n0L;S7v(~t2;`q?c! znzO?xP}`iACr**Mx^9KEw{$hQGSQ!&OX2$9uunzV>A?_AJi8#q!+93@X}x>%N54Bt+KbEJ5@KzH1;ZFLOT_A-YikJ;WNEPPyt)>=gikD)zSRA~L$fJ$}RmXm=z5#mQ!@cjX z_g?P(yL#U`$J*3E=wO3;KS1wq^G^FM;K*zfs7mQp0sJL9m}C&xNWJ;z`5IVd5?`D6 zN1Ey~?yJnba@@sul|1&|xx9Hz75!xAX#`>OHd0kmmpL6=f!!$k^RiWh*;sqy2Y-eV)d0h7V%^&IFLi zYG=~onFabeOGRw+s+Cz~V0cZ#P|1@ZnKZieCexktd*$e5Fuw{Sv?eO;wxW{WC-DrQ57o;>`u zd{JxHT^wE?42Y7mLoY^#a!(^)H{U8KOpe7fx2hvlB=UCW0~WL@r7b-eI3PO!;I>aU zKyB$^QDX4l1fc{j71Hi$*K0p#udr&KRwL=sHP+K%I(LPL`EIZ zoIZ{Wnea0zmr1KH2?D*uj~&M`{Zo4DW9h2?d06f)X7M?kAVue>EDBNjbY9q&t4GAf>BJJ|I zQH`g*z>=H})F}VF8`C2w;>{m>pf`WQfaU3p%4nWF&Lw`gx_RDMy)K-vzc;`7facl9 z@QdBY=JfHr2q?;XvpMLQZgvT(fM1jK%XiuHt&JELY-K#u*>s#dYka=dqJa6OtrQ-p zZ!!k6Z9=R>4jJ9ZnYig*5l7}$K~5pH3tg(2WcJ)r+h1wKTu+)H;097}de(gLXA9Ei zzBrZIM@fHMyqcA)6pPft+shnQ3J?@H|w``Kot5#oIMr0qRnxAG_eu&r$tPHn-6KTq`MrH7L{ZGlLqq+!=-3q*0)5>7!VFe zN|-lZ4@dXlLG|ZHaAD45lxArFSkVC|{KnZvBY9w5o)*xE>^ZGg<5(=w!>CcIRAv3h z9=QiPfW|3^2ddTH>4il5j7oYT+Za9tn}f;v74dfFR<%RTs$|;aK_v~`gMc4?ZxbLU z>fMKm0FxKs{b>dM<~EpRFPW^j0da^%+|&nSUD45W-PQ@3=o~I1%43DwCr@rh z1P@=)B4ROdHF&<2cYRKm32&6g6(YULFsfM6KIKiHO_`i#Z@`}Ms`Ow1nI0|+c_xzb zmPYN$l7|TUS@TxuUH8WwubSrD%o9(UT#xZ#wY6%KvzK0h`U_$&P+F%6sYLJ;O-LQn zP&6?Gf8nNRsuR&`HlouuUq@H0ZPpcQk>HqDrpo;ESheEMn`QuTi`L#~>eCiTy+h7$ zt)`6_zJbK>WH$ubPI%7pv-vcA>CR7Wgf`5CNm2(f~G=n7P(iFkjb5i zDU)_2iO6Dlf<+pi(hl7=zI2^8;u`J0MYJzdD{T`SHL9llB@kpXh^b$lb?O2h@z0A0 z+Ji@bEzVgV+|QHU8@3p4RTsVVF$gNaXrcEs<9hGJST;S^p~PgPTOTDd`>~Y9#RmV& zk74*}3nU2Wpg9O+oRd@Ly0vuLybb>VnuaL89>K!c$%$uX#Fzx*2hZrQ2N7ZvpRE*& z)#iBB6Zm6f?r9|>nYn)j4@Fk_9%CWyGRK0{vhrZ@@r-DxNygDt9}x{NWBHlXJF2mC zqu2Cf;9^(>HPh<^9*mCq7thFta5J~#92o5Nq=e7xB2gh$nM1t#X^*a!K3;a7x%?+(QKnK?ccdjx9 zzRRGRXTXVpuT^r#eUu-aE!FI|B}YP3zk7I$c;#}+i1*9YYw`Zb$t}t=egBbvT0>f51iQILZQ-($ zd)LY*`20dEGnaP)rG!FZ-HIZU%BDSRBAASn8=pXQ?=&2gNPWfpd;|=i zM4bZ0q+&;}V?6O|n!3)4lgf4P=#oNcNyCZx(LgpG(=o?y6A$?5)heB*V1h?wRcU)g z+|}iBEsa_xQ(vO0wc)$Gn-<~pj-<3=3GCIE2-Qv(L3dIAsZRqztWX}|W~+TH#*}>Z5L_eCibC%)Yg5xnL1s9$X5sUg~uM>k3ZqBNQ~_3ky7du9O&YaTahTvNM}`*m<>UUGiL)e3ziUf~35?K}#FAbGn0NK|woP z?IfE-K&1Z{FA(NWU}Z3tm4Rt(9m+c^1Ia>G4c~PugR%S~QCePMb>L{Gcrg(6;v`ER zKDe};;z7d&*iL$N*vp#GpjUkif~zgf7C(&gH_=Un+>j%>*-E*BJ&RRdNporVw&0W< zjY5TS3HPC7WS?WfrI{nV@e?IgQ_ggjC^>j{uF3qhRFlly8}MMEHW{oXs%B2(;WZ7i zjhV16^H;g@)np!&1E1o7`YRI5MK$8#Ys)FbmF=~V&(%X&J2KIDO;u7Z9a|89E6&1Rcq>R9}O zZ3#GU;fVty!pAx#Y+(Zua{OqWX8IF;t|;o9F>y=Jg48-Wd*mU(9sic+%jaN5;{_@; z?Y{{}A!y^4qPyIJr$}K$tc+X2WM&)Ew1W94xRpRT9rJ87!!q6G^)f1>0)Z#6cQDr$ z^s`%6XvSX??Dcau-|k`04Z*i~hAHrrms3VL!2@`6H{U7-cl-_UmV{?>_O4*RHeDA1 zv{97ev|Z36oiLGkFd(r&!q(Q3t+4Jb1EGN65_S9u=G^@13XPwr1 z!x_n%X^)@Gqoi?HnD?BRQOvhg{m>&%&l~H6zhK%kwk{GMgwGqnRp){57Mc_)nps}> zJ^j{=?5^JktQdi`O7j{pR!=~rRji&;>Pwo_x6+1K#jSdyz6OuJMx)tz%Cs#(>A;)& zbu7o)gHXTR@{=yZJi(}k-753ibHyaZE41)}NZv&HU4ga?fu=Y$FJ>S$J-pq4W9e5S zPiKy!ScGx(zvUYCCDfzEB*Uc@Y^3B-NUK7IYeh#@24Z2j&T3^{kY5zclWt^}4z~wQ zpBt`=R|2w&A8pO$uQs09=GFv%{h6b=Kk*EgLPSOLf1YyMg72RrY!ur`!-^c8a^w)* zYunMYgQz_G3ixHX``+-*+vm{H3#F`+y~ob8q;wU{*ko}dnx z7MyHB%&LgL7&TF!Jj7tiCA_B-+ieCqX(hE)ty3x@Y~fJ>$APh?mwDQ}vq&qCXx_;V zxUTALdTwGR*HAQG+){icDki+UXo4IX8|xizpvdOqz{*WEn;+MVCqIJ)g3H}f9o+vU z1XGx#GG>>v{C0ycNR5}yzdU9!tj|V^0?_W2xQ-ucR_ewXHJO>_KwdRd)i?M~$4sxB zd}Busa|?_`rebMA(dBsd(7gc`51^N>&|}0Yr-Jc{V3In3f$#V$2^QW$pI140fY>!G z_J|f;k%?B}YcgZCu@#)e30{q_%4VGHC(>C~y(*qCOL?BdGP74$)!ExLmYI2~+9|nU z5ScTuIjx*vK#8(v-{mT8;UNtFV|SYzjFjB)08iH3^4_tvWHi7Ol62r#9AMC0KyY5pM;gC7+hOw*#nu9#PC~NE z-5{JC^;lP&2&z`D|IFh9CCPOP{i2MvM2~b>?M2ovVS}9y!urjYRBBHD$xs%}tu_k> zr@W*Z&4d`us8{zvxMt_6Y$SfMe<^J{Wd4j)X4)cTuKh_TWd8Ghhs-a|a>)Fc=jhB| z3{&7-_zT`3t$;nw^#bU2EAih;KzC~1M9JQZoG8f-oi;|o;sGn>wA~q08#W^t{YBw2 z*X{m1^0RAX+k|2kjytf{TR46|$B0D|ylBGBk2T|R8$r!{f})!3las`Dw<(UfYk~)~ z>&5X0XZj50{macrUqfFaq|c1b%n%#Slz_^MX+^884<>()wRUQOO_7be>Vu_x5dMlD zM6*RQd#{K@v*F)Vmm@se6e}ho$V0?x(R(su2aA^8*W{SzA~T30*9KA-!`y1P$;S8< zjrMv?+|6eH3)V6lzR$ikYXeispmDB|?_1>?f&|ZSC-mmXLc+eG*L0z=F#1ndyW!np zJije?`w%dJ&~CcPIMIoRf*ElvpEHsoLdHztie*NObtXGT$=R9;X2ioDDmEjI_QN47 z;RJOxFZ5Wea6G-(8dT6ww=Qtn!ey|f#K&*Mp2Zf#!5AgNhour^Y~wpmB?h3>d3W1+ z5uDmtGaV?H!EM)j&(49&$+xE=ND}mhRE0^18vl*(XYXlPI+?bf5n*LV>`NH=`rw4cUv|yE zk9@*KEZQz(w)=1~QK9RDo%t|w*|i;>`rz0LS{Q2E^@#vL0CniN$dLS?|7o5?iJ7#W z1J<_SvC*0vqP=tqt_e=+CdjQ&ucTD`ZmVNAEoR$5D^qKg!i7Z7?{<-sfpi54>YZ;`ozs{!i?4wA{*bTS_dFgDSgU ze+byHBzG`>!mBhGJk#!(t5wcF^al%|!`h0|U8}7;kq~RuQlxK0)8?JZiFSRCxy9a1 z7GC4homkIr)yS_kEw`EQVIh~6{+(~>?5L&Fd4}I}y$|<@uwdbh*?4h4{g;YN*8D6@ z@6v0bnNFaEN7?6sf&Zwc%;REDU*Jb|5Ya#sQiO}BGU-TFg^s@Xg{6Il9B zLR)s|qwQ}3X~lU+QXBPI6I)E%KG(i*TF8&wyP@!8-a}j4FkVnB_vY-aeodyH+ z1!vX#P27`a-80=g*wxikB0~A|AF1hQV2rc7IU=7f&pj>V_Xu-FI<5o%KFxLDMV_={ zv)BsW!z_gija*TAp73{Sc?5?J6?R%PYkNd9eTkHgJTGnzN||Bvpf_6eMLgi8uN38G zhu*)<@YCJoIjYEn-_RKJc~mhK6d2-A2QGX0z!Zj55q=jy&C(Yht$y+8x0vjry;XI1 zccsOsE{CSHB-GhU@zgrj4ls~<9M^T6yW7|BWf}OyEIWC~%I2&%4%#2jgQN+0HYiY@ zv7a-G#j8yIdQF?vd{=o^w0-^RT%pNk(Fi?m`#Kn0EmFJn0=1dnQF}9Z|1P7P;0WLq z+y@!lW`&}Z(afH#!`Y>gt4p(6Wu~lOF~<}U2YvhlI$VHU{SGa!<{c? zkO@ZHLx1*(#vyp-`wqv;d0Me|e=5>9w&2`*a3AU6Ee7n;+{^lujU7i);U9qkgPtg+ zX=UT-d?jThGBA4!$L#Md{Luk5W4(o=4jhZlKV|@bMipf418Mck#(rF3klKDC>r%+M z8-NLh(0G0D)XyONsx{U*UDjoOJOGiyfnNy7dBSQBb--<9Lfj4CCIl?s6)HCB2oE!? zLGk^_s}ClsX3Tj-)*SBlI9+ax3I9_Z8x<7k1!DNDg20RX5}PVlkf5+eI)ccpaxoyN z4{mr;j<4M-5q`Z|)w$~Na^Xq;R)yU$e^Zq=1Y7D@$z{*u%mS;@f#!~DLjIBarBdPC z8_z20555GR+)<;bPRF0wQj$r@3O|CO-p{+Xl-CrVF4t7wzS8I!lX!O5d(divo_B~| z*tFXaoRQ=;J9N!4+GgDl9P*Ck07Fnij*@;cMoDXepYjYZ0CwA7qG85o=&?2Wju~#j zziuzd+%Nm7{rqU2>5-4fa7wX$6yr)D%7nX9Lx(-bF$Gs%R!8+Tstl`}$S!Id|F+e@ zx@SZCsxW0!>W1KjQ;hT>l&5xppa$-*8*<({N#@*h`omqyAmr26oC>p34$sI9!L`&8 zXPGpUXYOArwo~r;+=m^|>AsjeMZNe#|7i$nlY&CYu%T#Bov3OHeRb!RgoTdtN+VX@ zUQ&(EOH&C2v>w9sdaU7zEmH*mXdiolC0AU#08x);eYw8R)8F`B1^UbT*Nps%^^7p< z7HXRr)-$;IAUOQb;Dp)-Tl+S=SDd8bZE`N{T9hiPxL}I~Z>`aaazo&cc0%MTo*MVd z^c0vC&$4v|921NWZUPQDgv(Yd2XPs4Rdcm<#X=fcyG*Qqy<6WkRca?He(uR!p8gm2 z&gW*X$3UDJ8^ryq`_8hijt!Ef>oOWx0x4fm78>s#%($Wzz2Bn$m>Vs|lw98LutNLKCCmi8v+>1E#VfNIV7dlad zAe5rx4v`;Cw&Ko5iv-O6`T+PP28`UJV~93PgM;7IFw2rHeA_NNBx6VGu?lx9kWS^7 zMML#3@Us0-Jytl^P(oKkzcE{@IN}IC(Y%dK4RJ;M^hXXcL=pfb`05ojs(d zoc|iC5Pk9N*!G2oQeS(;Ipg7E3ViqctA-tVsEZe%6(OpFOP?#)3w1gL)8x4~ zGp1a1MSG2ZEU;X5K(0!H<~V&xxdY16O#bJ>ebp>LrJ4@(yX`o0ZHdwQq+ zn+l%vPJhB1xG0{vx)R6Hl1;VVAzRd9(_0#g=p+Xsxa#OR;*8+8@(Hv6W5Mz0{7Rac z5UwI7^O|%_oDcRuq=1g#+i$98*IVs1RWEwi9%y#mq2+^J{ugg_X)u z7!SmM6wJ7ZsfasY<&Od3QMua54lO&9J|bCYrbDt6H&ZNR@Fc(Yb{%Dli)u3l^z|la zYO305yi?k02DO>P=N1FL$qqbe;hH`I1l16H;hN*390vjLJz%nz9=Xmw*3sE6J-p@~ zC)c*9?S{+q0*29G{j$4c3?ajGK=rZ8acV?34c3f;9;*l#?M*+D#3*KjRx@zs4#nR- zHa1!caV)Bia21k)>w}{R;H`nbMiug$Vb9IMRGx<9pl{Z_^$&K5-i8Gw(OWDyJt|in zB%`;%!3n&Dtx_t@ufboDdaKm@>7NVYJFgJgL+ayNLz#Dq_&+;z#i6334fyptI+Cal z?__W_{vE>j5n+tJ5-~d5Rhl3qsUv7ML^>EQ>^mWe{rC&&>Y(TmWQT+JGpnP%Us9jK zJUvEBCU%{0h2qgNZe=4Z^3TZAm$an2h6q%fyf`x4L1LXIFD`+lY9r4)afKhETIb=z zi(B3Ofk31W*4_IAKf%@9?CA7@B=XsR7GTa-Z^_gIk)Ht+*v)@-$V#%;;HHy@T^iDK zZQ==^b#8F~##4hD3;qMLP#mOk~d$}0x!Qm z<{6H*(o1r83nuJPZru=0wB!xJ(<-B@L<%`$qn!Jdb8?h3jqLE(?nn5gBncJWti;#d z4||Hwac{fRIwmGrUR}T%g)F_VVyaR=?>~@MKLcA4*IiZHSV<55~bP{zBNSnhC+g8uh4>B z^S(~@Ztk=^X&KW3dr_YzT6dVm)Xnwso}{)omv-86hF;3gZbDSPeEG-&i8#VHXwuOV zcbmRxG6WSVMt=}Kw!5-Anzt3VB;mi$v2lgXm=q_rhhlUnxW~?g;2w(5;i7x&T!`*L zs0^GD?_n4T@zHzdk8*yR-ev|ZVYes$sp$t-VN ze4P6|_hgSVx(^qu+a&5R42BQ!qeE6X`p{Xq&n@h(;=`H*@`p2$=LzE|MxE9KQkI9b z)(5j&N;Re4a)ikbVk(U(2}*Bg5K*4S7@loxA*IFaZcT$gQRAlxVcDUdb2pFp;W-A@ zCT^yP#*boT-BQ`&io8h?tY%uJ-D9QPmd?d0l_5GJ=IPTa$Km4GMtN0gu_*CI7QyUH zC<_fS3r`)=EehdEbr;n*W;{<%3PB!)5U-j^l1k4GecA~|*Rg>Dfu%9Kvs$!qwa%65 z(9lDpj+7O33c<-#DrW zaEm3G51I?$UmouO|3uFn0REcNvqP^i?Q6Zm_o7k2KLY;diA8`9?t<}}2=Jw4#0aB0se$?zY*|@>LQGby*pS>6#m=@2lh%mcL4T3m7X2?+rx_R zoxQaSzI!{c{~I2a?A><^ zwpImFbb?QYqYj|9_D6i-t#D2Z-RCb(;9{cNN(8}V#{4MwcBCKuo(62_;V_c8kj?%A>e%})PJ}vtFf&E?<-WC0Bi1H7KelLykCr7`V zqWs~}?-%(k9eEPJi=UTU{<5$wdOjQFKi}o~T9p5J^xPKZm;PJuFAI;2o+Ejx{u8@A z--^nI(et>d{Q1%I*HQW3ba`GD<=-4V?}+m6EPAT_b<#M9Va8TFjj>_nw9Cg}BtzF? zz+nU8SF;%4ZY*^8Si>pYn{vhk{@gQ-Tr$qX{Fq7U2hrE z8ReZD8`nX}y$E}J0oPQtSNt=%j{7z`?pwDj{5X$Cu-_Rpp76gyk&gQ>Ot(PDfVaH+e)xyJNAPMc0%A?1DB0Y7R1r*-^C9X^Y)3V+&O@l^u#o|wJ4zeIlK1eduf_tq@h2VnH>!X8S^IRtFR z();RdIEG|LJhr!XF1oewLk_t&k80tJ#fFC6;BAhG#fVSJ-oc0U!K0_DUGA26of3dk zAIx0*7xiCDn$zPtctM1Xg#Wd>;d+yCAZGuvbm9qM&g4!Y-o~HkWrl9Jt@)S_F!4 z0SOleGm3H>@|hb`UwfA<;F1xq`B`PN&}?t>n&ioe>o)W>w@{2(94}p#ta`U$BC$~TxQR@`wniTZFCNK+KwyduiS2KhH+Ztc zdqSW{;#@#r0Hk2y2~`Kv~A49|7q)uFKqCxzga{8BkriN}^-2{;U8Vd#sRQG%c)q!f(^dIo~aw9JMc-giE>&#>;)Dp{{V{r%3 zZh{S`?1|-xK^qgLtDJo}JCX65OnXNZ+H37He!u+>?X7BD&~ZKLiqZbpVs)tty5jpF zowp>;e73G!k=L6Zr(zC5Gd(VgIV;@z$AYM;4;xo=7eqs!{N+GNojHD%EE)9rGv#BS zEM1KN>06lqm&^>E22t`d-o8AwjYz$>K=sw1FDLBn31E$g!hJaERaVzrMk3L14V9VS zY;lB{hzPP~iwIJLx}4A@gr`mqK#+##nlybnksV$p%4e3d`_>gt6+3EWa!@Y5=KX|5koKFUvg~@JHM~mIWXDv8v&sxW9)M%#wWmByfA`H}J;l z#zjtND1Fi@*RI$##{v{`+6nvueWzKB?vTD6UM7uoIH_X*@;Iq&_8d+w({N`df)fGs zg9s-lL^zogEZ+&75NT!zgdSeVAK|;k|G52`Q!=NO5l8woZwt7e_-N|AEtoy){3mM6 zQn`0IL3;PBsrq>02f07x>AT~*F+Lw_%%)$+v?K#x|4tR0l<8Z3vfsDdDlmo#PoQ8P zucbtN;@sPA0zR*chBrd<(L9|UQ)*$5bBwHoZr1*i3xhnYy7@Q@-K^ktj`)8YiK^{I zRFcDyX~nM(`-1sdB3%g9OCKN#IgvlEy+k&l_9l(7w&W2bbn;v+;#-VvTxg6G@Va3KV8av04dk8{D5zr&CzxH4g2 zzDTvngiQ40zPnXD8-+kNIbfEGa1l*M{qbAF1Aw76Q|5nO=awk{PahMG`3h2Lj5x|P_Fp7+h;jl4 z#Jv-j=;{Jy&TyP&=v8#$LAZg<7sy@DdJe#njZ}aKV%tkNFUoc)B#Uz$snPg z_1|o-`1E+`A^NSe3HvH{&=1^__lv@BC&8n9kD29~P zGB{HaGkGq(*q{sc2BHL)4<<66^zEh_1e9z4eND;FwpaW{Ah6Jy>fmtAH?ZxZrn7wq zp^FsS*8KhdarZ9ZQB_y}e*yubq9-cRpjJ|ins^0;HbKw?5;&t1r3!^sskA6TMMX&v z1tgF}GmfKFX=|%iRBUP0iq%rQ6v8C|?}#_N0^Zo;$VHTkh|KTv-RI0?aslQ2_kEuC z`SWPzoPG9X?X}lld+oK>Ub|)6-ySDxAn{KMq-TKrBiC|o+gcWUbAKUe2+(la9CIfb zkU;4;+6v+(4#gC((btC@2DHtRZLI3YQAyvLPt{Y>aZs<`x0i(FWX)xan>>%PiNn z-D455#%9B3jrz@wo5_icdykc`p_k-tlQ=?RLDQFswD1N=kxUYNPJkSfZ1gcz#ibA+ zfvZH!)fI_*=B>qHy}9ariy;gb=9rtL=vi;B$j6Y_&9d!M+VU@FZ^A|$*KBErAm-L- zP`HZAvwxT!obpvoxvDA748_F}(Xr|(+=7ezfZ$d%UH8-35Ch}WXkB|elSh+T%^*K> zjqM@(|OH^?UFS77^ zPs-hJy_rhJ*qdI-XVg|#O2n2aVG;7R99GW;6?%sX`pkdHda3^PNMQ|~*yvJq0Bi1c zbB2IZ+68oqKu(1?Yu?}&-Q2u?DjDu9d?CR37wfl)ieUN|o@Ok`9|(g?*X3AuuXnj0 z8&{&~rd${o@xrp|oH~?^1n+qoHF?PZgd)Rm)Y~`FJ!mQ0EFaa%GMi8p!|lK;vaKBB2Ru-N@uZ zb2oX3i@BqI;1Ha;@unL)qC6xPQZ^4SsI`I6)OlvvahkEj350hcD+8nln}t;|*{%e< zK+z>xPMsZvrymNR`iyxnd}?|41UK|Su;5cvAY|_7I*%`QpB`EHTG`29e4M=QiU5r6|(D-}m==?=fzOGRyWfy-`H;_6_Q68J)dwMKE#| zoh$>!2@T>qxC>62k?8EyLh}fD)mh!oomGxWJioqdNHFxN0I0iH2z{NvaLN4>yJM;< z%jcKs%TRcfvo`DKMAG=qdZ zve0zD0%*fia{@)XD-TlLo~S*o=WIHT3ZZcXV&e$2<#I3{DH{Q2B2qbnjH-~_Z0$C! z3PgtKUS`&s<=$1Qw3y`_Jd8ZU1gvJxuYpOw2TH=04JmE&tL?{huP>EeqFLvG1)>b3~eADSRT zqhM6xLE2#zp>=d>JCI-Hr3EoZ;dBt@8f{&Y%N>MS!SuCmcqDWSb<^4cw_tU!XsuK8 zS2dDURw-%5$p8%_K~cs=T)Q_4$oR4ZwqWkykYF%;)jS;j<|^jZDcawh6<|kYw>QO9 z$Bj&EZ<;`doD2>)ZMM#aD_#~Qki6LoR4stL1PitiwdmB@gc-~@R%()^`olh;Yylgn zCR8PJ1L0|zo|h~t+vQmgr?HxZhqIP%DmI-bYKXyvRWw=*)O7%ZY{hyja;tL~tidVh zooC+X)?-aN`~<1kMxtvLkoh<-~EisV4_DgXj*lA4|?B#4E9-cHKf)B2NN-AkRU?HP!^nBj>Xdd*&I?3>oqV2wI^xHkyevoLHReFtv29R?RGLQZ z_+*2VEn9K(OHLIsuJfWCYLvfUP%-l@laqs#bf&OA?$_GUDSDUC-9t!E3rugE$B*Jw zp=Y~3D_DRl0%ZC0$s$X!Co0O;CD2q{Xj#giMxJKf&K&q%jsP8vndLrPVArjd5GJnr8L)#KM%(gv6g<@m8 zaZao+msWzq{YdQ`Q%x#Swv!mJ(Js))zN4d7innM`C>;@uY!v4OG;r znofe@%uBMW>fC)eOQs2DxHPsNH;?$d?tsXyI)rg>ps={kd#xK}oe z5L?f;%G34zsg}2_Xy2s&;{u-aZ(1T->}`aSxj`Z&1VYm?&5KuPda_mr7Jt^MJhU|- zcLSJRTe6&*zc81yuFO2lTUWS(G_kH2DQLWfMH*FaWiea^C>*A&EXJAfPdgJ3C`$xo ztCgis6t0BCFt3~sbD12Kr+WQvUSBGwv|L*r&z}tiZms&j7SZBj^=HY|oREd=qSMgBY_5WA^#+r`k69{MeNq)+>i1)m{^c1c!OauK(x@nHE}5E3ag&MopH` z6)rqsHe%eys`sS~n-}d5Ac|LcGQMk-@MlMAwXk zI+W;^F+?9UB#1uyv1wJ|KxCvX{4SNb3z`XivE{cZgwE|rd;1*yN)@zScmHGOb^qU{ zS^Y;1`j57GY%qgJ|Ep}CSK7-{bpuAlQ#+`Rf3*Ak7kod+e!rdX+O?VPmQ|SQOZ3Zb z?XZ(T91BcVnp4}u3VPC#BQs^D|O_l zZYy=%sMcCnP=zlr4L|kp9N_-LiOd`wp&;e*_L{5uU4?n%Cr}#)4$ETL1D1OyC2y}V z7um8}`zy>T?Ui-vhCx9U=1RXjZl(5`gH2{#j_tnAHBS5$JL!Qo-#99a7V=2GC;_;V zzcJR>rO~797yPj}R~h|;&lToiYSyPGyv#A%0(JtWuNW9xsSkF7uU7>#iLkt(8h+om>|KVQb=V8coJQ%bjT(y0kqO7mCx zv&3c?+q8e+tPtL(Z7e${)_%68N9^}-%Oj2XeDlDHWicFHQnI#-*QuLu5-*3+gjK?}{W>$RNL+Cv63e^2JaDyPXi!W8jM0&l=q z+ev;356tCTX3vKH$fzf?_WszJRhl;q<=pvm*EqEs*^7{P4DVC;J(k~_@MMH_!C|uK z2Y~3j73(u_-&*bUu=P=Y!&QGx^_|lFZW3{H{AmkXZ~4%(3dJuMS|%$f|ZJItK!99adwG>37!k(SY+%!*w4A7XBC<@UNHv{jp-r7d|vi=wIwC zeA}Mwqhspgpw}g7c+CYl7$x9Qo1_QWtJfUZJuPjZGrwO!WblXm&cg2@jKrgG;~uoT zo!V){ax5-KP2Z9AqLFm$Pre*f(>4v#_#1W6WW)_3`Y91y=Rz)LnKEHN=tMVl!f(7^^ z;FZBnKLd>qk$~y$PDnu0Xsxg;Uy9v6Z|Lbpa8W1K{A&WV%G}P?i5EkP4#Rl)x0?lX zEKS341My|%?+~11j4g+FD^gEmc=GC1`ia;$)LMhBWmaCMcq6r?SsrP5Qn9(rR=M3g zW-H;)sxq_XW?dCwKW9su^Dpyef~o>Hbv+crlci7Z7FxOS0d+n>TWK0DHOf3Ea@r zDHeSL68GmriL-EM-au^z(C2ca%V!nlpMz!tWhx8m0z0+3=D2w-SGbCsq1cTCMke*4 zjO;M6yD*=|n>>^Q$6Ld_{DBy02RJ(QAjCs-Lk~^c4_U8Ax1MkOSs%*G704(&;@4dqEh!ZC$EHLExsuV zn&Ch#3eBZ^c8$SYHfbl@fKt69_*}SpHkEP0`oL4#bl|ttX*)npuLILaN$wZ0h@P5& z1)kH4s;!46P*pkEBO1#J?PO%#np&zRQ;A0B@Szl&%N+M-$OWf|4S#AyHg~xlJ@KKm zaBRG)j=#G-8K0-`oVshN#9}xh!}yJ&X}S9C`f2H`w4N9Mz2U zxn$QI8l2RiQkkZb&_8bYB2S4TeTg*)+)7i??!b{8&xF3_DMF&Sd)C(sW34CaT5Aue zx8YNKd%z&t9%1+JIRE_0YcHX+dHg^Tu2tHOXbH*9v!G3}xSYV5zrrR^rSVA7-<)b~ z0vq;h6WGF%O*m(#cC;0Ldb8d&IWzvu3-P4cOv6#cl+r)-I!L^RU#o)&ea%!P*@O+i z-jPyyMWTb~Nd8J*>1)a{=Sum%!M60off}bv^AOOAp=GOSf8Nv6Op~5w7&PE%+Ziff z1#-Cb?B@FYga@mpd7g~c7?7f-8SblT%2Cq{mzw61LeuR6&HeReTvv+vYML~gwgb0M zPfarkq=_>#0=}ANLnWbEz_)mQxHyXt8Hg&b;7b> zXcWeQv}m#AR63}+?EoR`IV(|^s9wao%o_ix(7fqC<(n7%tYYnEp-pbXkeV0@j}E3U z(^euyZ7jJ(Ys7%+DaXf0I(21irbQQ;bN_;(wc=(TVHJ;+0)qMGGtea&t>%z*wZ)Cn zP9yWx8m}U7!5tu?Qg80`G=cfv+<9V%REql2=h$zW=T1*kScN>`!xgC6(7`SR0s0V# zzvM=|i5%3558UNNUkJrL%bmeku)OJxRN^?dyezAgqMX3Ir6r;-3 zi4rl50ZkdXU`VV6GkN|fXi@xse<6MrzEeWqQ#R6tTeCBL;(paOOWfcQqkAZn92`XG&SpWwGb^d@sZ`%ivDOU+g&R;b^=uR_l+ zVv+OUi}X2~+=-e^*Gi3oqQ(e%^B9|V#FViLv)_ev1MAJCMBc>GW_LQ#Bl+sv)2;9; z2W+SAFe{EUv?mO{d$cbMA^?^bylh3nasIB#qChfi8_AE%X2Ip^^&W)3XIXK4CD*)P zi7>XC9hu&`yP03GbKwq;3}2IA7rX%n#+bv4>U%<@#-^-eTiZStzzR-tNzEwh5exUo(EH zh>C{<&}+)4F`j+J;ksp_>Mz#RVG$CmB4)E}@{wsd=HtV}bSiJr1QcB5Av~Y3;Q6rg zD*zQA!S;dKT&ycMYDWQ7Ify6@h8hAVp}Ww^J8`UV!<^SP%sSNMP8=4fnPiYFF91o+ z`hQuRkrXw4Xl6&*n@zTxw_cYFfOu?noeO0#{6jl?H2MYa(KQ_G@a1(bqN{w?uXzWc z#Qx#wq5VN}y_+e5c~c65dFMNI9kGb^%eBE)n3hx1owAkuP8rFc^Dn9&ns-z5^Nrp- zl|pqu8-;4{%FQDn{0ejO@w3yQyy&e~abIki#Sn9zpzCFio%sI7frSZuer&lNpHLL1 z)8;K@^c`a8TE4J1L^DcO54>=N=!2|#3HdQl^paD1HjFC}naGX@$7}2e=B;j4>Z`@C3RC!^pP#G|${ka;YEoBP zo>YX_-vNAbBR@k4{YIruRJ#RlxpmuaDzy~^BfYtFmCAHr2JQFQ5NAB9p&uB@HMu$z z*c<3r=cViZBDC68FQgP}nHU@3QcTqYP07EE>EX>+b6$o76;f27Sd2G%HOgb3ua4rz zEZuK87aL$eU5@lwyj8qRJ%3XBMTjz2u4y-0E1}Np>zhY|kv_A?Y{IGefp@e)X7NV=u?XsNAcmbW#=>oz5hJW;}1Vp9uM(+A@7jE|dIN zds`yyr>PS_Fb{&;?^xX2H_GdnM4&{X z?O5%eK3vHMyJOyn_Ghm<<}b0=ihHSUtDvyI+8Pi4S2H?UM=K9kwg!Fn2W@|+_F$G) zi@>Ur+aPZ_<&hJFyeY#yDjGC{xV-Lzy{#TT%@^3&V0F2}#RuThb4@NytquJjyPRq_wml z*6dE7=!hiGFDWcz#uT9VrSlx~id@IamY z+qN?nTUWo(Qck!%O?c9f)bbR4Ugv0;lbj$Bj?%Xf@^CD3o7P0|iB?vO6t=vLW2vMb z3B6{94bZ*a^tE*&76PbB>6mi3r+Hl@s-`q*O~brVJBYk0`xga9`3yEVlNKxUZenny z=^s)W+G4I?^X&0g#X+iD3ErN`wRLmIquzY;-6@+&{sS;+X2eL3(2b~!pj^;>+zdh(dRg%CP&oP|U3&9=88vOcU+cBbFNLzHOwaVz+r_cdi@ zuF_lm5h#*@&z?_vT*8lESA{vT$LzGk5RqRzT04H$5#oe>6wlsyFuGNEm^Sy(-Cv0i zq!`tzfU0xXp-Jg>8ER20-QJEjM}p%V5+lAjYA&`&nE<3IiI5{IiJY_B!-o%v&)_EL z=44OGUyYJSvn+SFY=-_mf?4SbPtRCLHmbq|r7?C4zfaN^T}0s|u`ydW5)EA=bO>>5 zqGvHOmOc~ujbNnDJH-%1JK7K(qGKQA=@ZqOI0zfl-<%HOTBwFJM@00 zO4VwoO=O-{F@F~Ovs&|bN=ad9pLiuvMa!rg?Ov*?sy`!;(CUqYozok($BhUu;f%bb zI#1qlwHNZarGBV{6cJ9Yd2QFuIMZcZ-c3?zPD!j!30_(l>=3W~g+C`!7qe)}RA`>Q zOLu_IO)X5vs5Ao3peD&!(LZQkjQ;7RTgZ02RZMB{wMyc-=TU&hcna4oHV;>VH;hr` zclu`KR@y>ccF|15qc0VWz$PqB`4epWFJKSv;NsBy#eA%^_gDwj?^ad5rHPuH7!z%k zZ8AMUCPG!&kMe|zl%e;(79&UizYUB6GZ7TaLxv4p^d8Lp9w<1v_ z)=<^q{z6atPqum97YulI5!N=7-ze9Gc)^-yX+&f{yy$?W!*^qa*}NZ9PnMu%nkaFi zM~Sj^U(YAM@VI`i@A$pmtXilP5&r6L+UcoNw^of2ng4$4lYQt$jRbgk^)TA77+kLQ94n=Y#}O26!|}HBf64b zZ6v?;rmg5&zQ&s5KHeBR2dlRx(2)Wj3VD*^Pz;wC?LXVP>ABvbgm*(H0o(@B=|EUv zhWfopnwF9+F4;kglOzFCdmq%uF&x+%V?Cn-7$M*FyfKC{zz$Ejhmr_|EJ_~D6iRvk zi$3&B);dnL=1^;e>EM5)lb&hSoUf1BYBIi}3S5D{$)MJTA*Y=RlSQp0w==#O2)*JS z&Evb}xy1PH*xv5=rm3Mwp8@yT@jc@UAM#DzW^*%|9loVeolpkRQIsnDHYOmAD#~*d zwVD!RD0ds9rziPnh~_K~=5`SdjucHIzcpBrWpg2f3^5%!As6i~7Df#sZV|aP5apV| z)DZoS)lzs+)HpmO=y@|9GO4+JJnZ^;Ydn1ZYyuCPx3!Ch9cn1jr|-QM4=2TZ=+_^r zHX37SUFo-{CW_+7;Z2%Ein)hMqePolKh(`!xJm*|Qqtu|aozH8=hu<9x6Fm!#~4BE zSWQ1Q{RA_N@C_oPL|?Jg*chGVAwK$h|Mw64LKCqUfgx)5e*B{0SF_Qp_-)olB;pTq zGJ;@q)9v91H~jO|15}=`4;+g-1tU7tAGuUKWD4=OBID3^aIi+9o-gHm&p}8c7@mUc zgW0ZM%pXphG>6jJggB|dXKcI`RLM-Uj7A!N^m2vn_GZd%fqP0xEQB7pfRYuhuUfYKLf^+ZXHZeN;L_=xcNopU!@s zt-SX0cWL@8@FKEiC*o@tdtcZ1Uxx$}k?KZyADa9R_{^^NLfGYcA0D$GTDwZsSzbNe z`d5cmr>M$3`yEQGV=Zzudoi&;l%e$qWV82~<@p=geVE_NXCW_jmA$9f(9xQ89;F=e zwFozE_UBg=(?P$g$n5yaqx%IEO_~)I(fX48XbTcwtv}v8JEnS$^gc1-BR0ksG{>tJ zy(sqi-Rz`XU(qKrmu#-B#i)dPI({^_*>+M6+~}aT=S<{T!eDKlqLZ+HS3hXAr`%-& zk+{^@KxwDIQ-e$FwANkH^ehdS{+Uil!y0RS)Q&e1U~o4W@PY+4kj-fic(q2J^=1{7LPc4oXF5BC0C3eM93NyCS*&S7Yb6;TgIxue$K0%Cj_VqtLUe z(=^Vci4Vcp(Y#+x-Nmqvs4yQ7dr&#eueUHPHMGTkPAwLp?y0=Qdq#Y&xslt^!!vY2 zUuDP*|5Gh$Jkh`3W@AyyWyyjV>DJlQo)1DWv}+H*uJ%6lLz&IDU+#>@?rw%3P>s3_ z^498C`Y_w|zp1xWtxXQGx?g|lyEyf!TR7%g6*QKW1D_j+a~8IykM!p?tW9q?1w(On zsqKJ8!Ro2Jp0j^Qqe2eJW8Z zGs!OQu43FTBH!Q3n)#<2KGcp(rE3R1G2%V7OK36D>Lih$FQi{^Ai*L&VmXi{TFZeZ z@46pNnl6?9pz2{+Icr}6c=avIw!S@3vni#IM?{hVwQ3u3`>;KF{ z3>z&`f%!`x?Vz9Yh+8N5Ri%s-_V(~jYL&0ENvJS$t`dm&$3b67?%+9+DA{VmiL%Mp z6yB9WPZ#H9Cj1L3%*#QI;~IPL*YY>p0T1qVEJW7H($gz3SOO?~uSbU{JSS;BN;i-5 zvQGFc-qA%go(n2${0BJV?L(w{T)ei6Mrgb~q?45sGMaIt^sXgWFKbl~FFm6R+2S|D z%cqroz9Gpa!=~5~iSG1p;|gS=2tdTB`>5Dn@#WrmGP>cl{Z$Pgjhwq$)ge{+Y*ujLc7f(o8MQ#K|l=x>C9Qy zVa=)iHFy?a?V{Fqojv`}>PtK5JebcGoiAC1 zPZFI>;|ay>ptJWmKLDMM&6Gsv?9Jbc&JO~r)<@@fn}1K}RJ@Emp>yDly`nSE|GYPJ zruljHfX;Q>y%zR>&Lw=d=zKG?1$B-Eouk@8=cn!uKxc!^ltkyxHhnKTZ?`3UbpF`p z-xE5IwJT!#)cN(cy`u9?|MT9^`GlWm59qwp|GWouPT;de=fh%e%%Snb)u|?4f>S#L zxKpM-Gxv;k@LM;yEq<+BtqBFI6x%_aB;Q{#f&^nQd@< zB-J&PIJ26fW#moB%lHGgEikTC0ZwRn_6s#|f7|^1TIH8A&8|zze}|tRWG4XX>gPu+ zw*WoCVwRB0R|2wsZ3xH{`*Z64f+j8cB!7~6Zoc@L^!)=m6#d-xs+!iknV#~!t2Wr$ z9`B*N!p=mYeuTcIxmu^~x~Z@*ozN_5NY~n0v1a;%T4-joxr}~>GF-xnx6Zz67gh)) zmpn!jTM!?1o^bRv5Zw0n{VhcvIZfF_Bu zeR;6%D$C?pova8^VIrjxMtBd%&d)s_9Bep8BE|BAAo!E+s+|PgiNgx1nq2)E1I2$XhMX6+5U$@AGvA>ElxSyq{I0~2(1e7cUfm?!JqO)Q+XjUX8F2qM>ARSmL2%?6>b4c!D~gK zf7pKP%ye#7tO0c-Kc4+$Tr;v%lD7`)Q>R4(HK{KgcUAFx`?$sKCl0ZFE8ZdPb!)Z( z5UzP`fX90pRES-a`!pr&IC@+t%bacP1PST6jWE{#OV~E2Y4X!_j+@=H1+~H)r*r^Y zVJl8pEg>rIKX*l6zU0DCYU(Cs4f|q6w~se_3xe%pTUle7{hTE1MS$IEqf_^khDwC& z6ZTUj4_Z64qc5xkFAlwtkG3iZJQ6$<~{%3kKU67S)Y)7G~`# z)G2>+b~@jE!JR&}8hA)Em*(`Z1(}jAYOv`nIJidV>sjPoPr5cvP{AmB%`LYWFcu;b z5s`(SZK-P>m*t5qX@kBnz~qK-GfJtrTLJFPVDreFT_rrx_qj#yuF<~KlSru_<5q3ATZ^Q zwBtwUcTLbxx{cH%wCq@??nvO@=s#tePyDBR^RAwfjmW`?tdq=AKLhtQDe{Wuj38zB z{e7ej!(Cw(Z1ATFUn|V;OEkY)cfD4NO+0(^HdqL!?+(_VYxU=7{TUfU?-w^0w1xu# z2uQL_Q4*5V?32IS2NKN|W{MBVse&Z6Bhh}WNUD1KT67@K>1hK`Zsa*jvL<)j?s)8B zwyPE9q^o45u-tvEx~ZXverMIIOQiu`8hz4x@qX!{rB?YTC~A*?FvHV6MK}0a@AS*> zukv^B5}Pj=E(eBaAFH@q8a>wkt>8Cye{4nUb!EOmnV*l{5{ty<$EL>?#p+`JP`Zr} z1emGq??=g&?R#fvYv~7a$#6F^IMa`Y;~fC})`p7Jn)YvXHQXT>z5-wGuY=)8UZpt;s+gfgTCVDx@vGb7(csk~!l{;m-`2XwZ`SuGS z41_%-;>i$D1)28O2JC zeQQ*%%i-&Qf;GNi2WuFhb7kXSUdC<=SqrcAU}#{zoWC-g z1{TAZ2fF$|h!ZUVp+*jO%X@MF+D!S3WiDbO;Tv;VCO4n(6b!9n5G_~w0Hxb`$Pn?e zd~Zlv#>8U6O?DxJhkv4_V=~@yyA{wJS4P=FmqYE$CQME-F932L3Yv3cIZ6%wmzCc z5}*Rc>jLn=ly5S5y}pYtHa&<=dm0~`*Mrb&@fTQ1GC6!-*>wl4U-wweM>~2wA9|%` zaVA6GllaBGi(b2KQ|x!)@~qIR(78Rs!K_^ZU{`ELiauOX=+KAfhy{gDPl-QQYX!$% z9F4oh@`#Amc$U9H`o~D1xIqJvhHoEQV(k_fq+&Nb023I}f}6e>&6k;9gliL_tL1h* zfEGAHSNE}%#dN#f6!asB_0$bpN4W89OCTq!St`+kmJ!rP(<&Mp(&y2lsmftg)EBU&F%IV95^=7`H_}XfK@v{dC?DF?vg_)`K zkOgK2-8@hS&~GTB#0k-`Lfv9VuJ+sjq=wrN3|+S6-g#zx40oWwMQtX|13BFq!Cj zrx(4tI;$?qp7)L<74F6)8wMkw5(Ip4J91}LqGJ*gRUwdV<|oZ`5oFqlC^N-p>zC@z zRx6~dI;B_p=_lCqq1b$^P)W{gQUv_Vs{sJxS)+cTrY!TUo+6#@4S$I17YQZIyv$zE zpDJ9d3TqmO8LW8x%j#!?(N2FPt3dYC%?;F`VZYYa*_xzl-=&Ut_8S`{8#Ttt7v=I^ zVU9QVr&Ui)+Yi&r^_PeYG0F6M8$ubKF@td4@wIlnD*PN$micdkIn;0E0Q-xd#);Hm zXPMdYwmQ?Q;i@nNaoV_4RDH_uSLSfToEC@3DJSXq5bwFWo;!Qbo%Os^b1HJm)|YwS z=soYRbQ|3`a>_QPD>^HE@$ONJ^tra$8rDBLL0)SBcKg@&V`6NK_Jx@Q6v{z=PSkjv z51<-_!2(#po4+&tEbJPH(|L*>K^f&Z&gR%;e#Ect;%$t*8@o;Vn3CZ-4rQZ-dV%rn zjcY7-)4veO1Vdf(TJY-CBtmIuKB~HmnSS=tW+ut{ z=Z5a}lqi@eX7lbG>=_eHbenEP^R9PHVO_Iys(Jb{L6O3`h(|Q(q?&b3)vwwq@@sn# zx_|GNr1ID@mdL0xiGlxC3D>@TRpMNe1JmKshKt31F3u~g`Np~a8OWsnn{(1a&qFA; zT0-&G+cD(Nl4@pDY?bgN^#YVU1ZXUkj7s*H)MA?&kuMr=u6?7o)jVjAr|G1B(n1Iu zab{9-%A znVT#bZu)Ms-$SUabRnD7w|Uz@?QOrzVaIispB}d*vVTO);?HZ`g^70c#ay;%SA4`T zqnDA}!^ajU%kDmLyQD`Az^D9fxYK*EH&T|7kjwvLoFOYPb>4)}g9k5>XSc7SSN7lL zX*C~nSsh1XPQ*%6VQ%WZE3TX77G&``#NFsV?}_-lCy5v9k6BG7^K2Ep$C~ZK#l5V+ z#prMmV5jpvSQ^MD=u*JavR{}0=_}`#1RJ}l0a%|8y|{fPF#U}J)12A{GC5 zl*@uD0Yw5TZ9To!$ZV(pnzMl>Auq62vkJ=X1A)#hnvgZ2GuVw^m_vw!tu8AMaM`J# zU-RMd9KVoXQBwYO=39Eucs-$U{utiijicRZLc32x;X<% z?5D)GljgAdyH3r?AZyXI-`@?B*on7@P3ueAGx4=siJP=LShU6M{-INIF=<>w!J3SZ zUly%u#V2-KV+QVzXU*XV;n(Uw2$h5?zV2Q(EJ)~KuJ#u0=&b@J$4oy}d_s;ebq%cl zPVLeY(CnPX{(gxq5T`TJ5jd&tqu_O6U^tj zp>L9C&u@+P;v{{f;Qv1iYHR$zMA||R|J*CE&F#L%sd<64J>mbIetW<_s2?h=$VyK} z^u{39U2M*F)0@n;lf9wHO$>#!&B9vernFP=4R2>u8byL(a7(?J8lI-;#c-Ek_)^W4 z?B4%kUDYJjxx1Pb=@JN>|1`LFbE#wpgoL3_!K7ado|(6z<~&ISdi*hf#|NgHfLTto z-T)WZoO$xfl9~^&uWvj%FuGBQ`QW5xs3sxwAnNZ0CrH{{0d8snVWQ0)AvNy2FZsU;i{ zqj7w5R`6Er%-?2S{~6r!UmPGIYwUIH=jyvrfrnnk(kwZf4713P3==+2DKm8D1S3{` zCyTyj0=0W1cE)|V|2=v!xk^JqBEoKd5h>@8LQkD~8NY0YBxfC{Pp7F$*o9?6W@j-; zN7phL+gMFea zP4vx6eJ=U*>%{lyj}uSiJPKgFoQLpa<>{gw&E}zpf-j4yy4_?Ot=^s+gseO!^Jo+Nyr`X^;RbB7_X3=bfzb_{oDUylL z@=)o2PgJ6tk=TpRrbw`e$XEj?9qEPn%xT|4F84np0ar zUP&yuoHu)SU0O+axI#?hd3-58mvuiy(pK60aB*qT%4z%6?Cea)1xwDg*7}(;!M7=y zOChN+W&1f33?1a$oTFwCX=^^)?Bj#5CS5x^2P3PX#&C<8sx~qVoYxmW4Ph29xW(Zd;Ze;AA2T89X6JygBBvvW%q zp^i|s2~voZ_DQoD-SA9fM{mJee&X33R4hCVrR(sXeg=}@Mfom^|=)ckBOPsb!~nL>5de11mTPMGTxkLw&%h56f+575V! zl<(Bd5w2?bKf7ifbYJ&?-q^**U#`5Y!c2Tisw}onkv^4=a)$kIu7h&%64uPfH>WZUKP*2o~a}1gyGH>EQq?HjJ((btS>xAuVHKn<4GfdnrVeLXE?3uOnZEa zMv&$5_b=N?!M+p}@kD=v@WQd4kM6f8uOgn95)5bLJ^hmHPJ+V^E_{Yi zFSdV?K3ksk771~@lJMYS)`12dE6frrVrggyAycO&5!Fti#n^3H7iCn89VE@yX}0r+ zT3G(^rZClJn7^@f%sM)7h+1324T$EWrBwB@YT+!_(%62smX}FHVbESJl?$W)q?&m4 zJMTaaD7QhnWlX;od7RI?C+{Haguc7b4e3V8u_?5Qx;3lV&JA{la zy^GSz21YXGq&KLRfte|lTz8V6znJ`DmnpZ4j>o9Q?V=?YmyAT(LB43&UaS$BqS%ed zM1$8sr2mj-gulwXoQ3R>d4fF9kY(zHUuk`;i6@uw<|~|8pL;6VtU^T^tBz-X{x<7w zGDa)LzgMjN54}>&7n$K3GtJW$wO)tYLxSHM$f7BuL*MY!48s{x2IGg9V4^L{ylA8y zv}RCj|J#A5U>|GG@N54PmNS#9N<*)d41*}tOSx+dP$Ie&+-2b%T)@C?JH z*!$L1bHxuU5sVyn>tE-jfzn*0xhXu>OwH^XbjsHXv*nR-c^zVTV`eH6NILeI&jaaS z#Iv`(E_My#(+!ucWR`Qg=-F7mvN%#*Uhr8U(z~-;v}Hn)h%d~tnPERjNd<(x+={fa zQIh&xzj#$+VsT#TtyAH?;4k_YIE(*&V}=rbM}i$Mi_uq-7>!O*&Lnk>ndyCxUe=ty ztI7{+en)P8CB8g5(59rd9{&MZkxA+1cs~AM0y`}{wX6BsE*C$LBu{fkmv!@~Mr``> zh8}gXpM`cV-r0jQ0LK-qyt3iag5}W%sB)$|dO$Y--b*%nXF?JKCBt-3<#|?)&*#p; z*v`EO+KXwj6}Dhb`A7tKO4!M*Nt;{Dq*d7l_79mHbXn|1#_rHfiZ(hugLYH5@Zc=T z)8c0ehf*t_sS=nXwR~te-Sfk7)>p+c%))&v!=e{f5!3+#CumWE<0Mz#s;fb%f4L^i ztrp;5N|B*`;7qTYuR_?y59Ct z)yMq+l>w1EuWjA6s;gQcvbqHxAE{5pQzpBItl(aBMgE|M@N+o+_eD@(DAP)A&CeLt12x*~b!T zX0w%GzLy8Ud9xl&a^4O+w&lEBh3xgeCI6lHN*n$=$m748$_v)Rf4l6R|9%Jm-PDr* zCj6Da+vaQ9P{Ik#lu+_3ZuS;>pKp^*W7u(J>KMK_D51Yd&WB&lNz4Z~^s@Ya_MIP% z*zL^+JTC}*eN$I&KJ;TgR1PnZyo@8om2wA%(vEEdW!06DR89KmkMnuU*#UQWfw-%DHLqc0P|&^sn^{7M4UDtid^4^!xK+PSL5wGdxDgL8kJhB(KhnYNgUEv7#W)~QM~=6KTm3__W>WnmAM%oW81VN zd_4GkDn70#ZjFzf_a*SLC-LBaD(|j90I@RH8c8C^oZ#g{Z4=_e*89Oi#c7h(x|$dF zsZ3yWX&&`{QTGUn^yIZo}Z#37J=e{xT}s7K??IWQk8^;EJ|2VppkU&sju zE=rX@3y!KNl5mv%jD7z_m`l1v7K}q(X)Zfh9)*6cq6s}G?6u@6l(hTg-5ku=r(*1s zU)ZNIxg|*Q=J)$lh7At#`)~HSOqw~hRT*c-)$*+I(FV>=wY?2-*DR;falJ+?=xpwu z%<&f$JuV^ZIY#eUmQOqSb$<>C0qU@f&~p| z0Sx{Q1$4^}G*l8Fo@4LPRATt>Kc#` z4=G2hq7uJnDB=mk$dSIkqj2~k#an~t%m|=1Y=%Ae`znz&Jxd++0kQck&x+b6tc|Q67RRo5Mi_w4Fu4Ul+fEU~gPa|Ur3P<3Rg3xl$n6j*{3ocQorT%jVQhOnC_0YHK0-YezLu&O%;)`Y}dVrJh?xb4J5-E zW*G^?10j3v(hQ-8Dl$&(6q)rRr@VfrmVfj&F6M?{czkX!JSv~-0LqI#b87Db?!^9? zJ6g*NwqjTP5q8xrDN_l^6=qq0cP2a?2We)^0=~;tIg>Yw#A!2%m6Eqen)V{ zE~Xtytj6-7w-)pa>4iTQJU8ZA-dn&^PlLJ zu_t_d?cw7y|Go_D=DIyOr}8?YZg%*sMwB(|xNm0S(^l{`+Ki*x8?qGCtOYe|ZIIL9 z0A^7`?9XlVIG%kjGs=!|q|;m~)R+X*KlU)xRCTxKD~QgZ!rXTR1g9GYGCWslpX@DZ zTbq6jx72cctZt?>>yZfF@$mQ|aMKaU9QhZG zgr~|JBlSx3yeb>dHf70<>h>DSr1Jmhg{Y@nr|G0-tF%}R{s}cZv3FmjSrgjg)_jz- zcjF(O(b+jTql;(oURkxlEqcqT?WaDnY89*uMn-Ui?Ay*0k0=ixbyzTT6pPN%K+VTH zg6Z$d;4P1Nyk#;DtIs&xjSK)Qo1L2PfQ-Z`houM8R~Faz%i#;wZNA76)_&>G`~?Fy zU!G5V>Dc@Q8#iAbPkhN}{(_mCFR};N+B!9V!S=wH$=;Wthjr#leYc$G6(pp{zZM@Q z);`%9A6+wW0Qh0}=(cBkbZd=|GH3q9-EaXI(4n;c;=Uf9J_RV7E^L(6A0<3_*m%+V zGPE0R00S~g>xcIB(w|JE;|DOHlhQpL-IGYi5nw=PrF$6SW-(!CsO3&koUT3j9{vSC zYqp2@@5E0^$p4mwICU415`9WmjYy}8i2Nn>fm!fNs*3)HKafeURnbj$N1&%St>Fnz z9(f79j~Xs|DXOQ|tiGuusL1ZLRBK@We{ci?dvoGsQlvPO(@2?i0xln9qnGMZ@DSl&J&0fTin&Unl|M&o1*^uU*G#K}i70w;E^HLhFK>R@~i=9eq zlXd!QR4&5a&zPeJ&%DY?dQxfV(?y~B`g+!4*LnOqck!m-QV&N_7W>HBl~5 z4Hr?~;Qe^ql^gmMY{KFieQllWIrbyp>+JzPT$u!U2zYV4@w}QkxdktThT|dnMzoY| znXi9pVHb`D68p81?eB{nB~5O=_Bq+3Q=$c`quD<6bB*VUw^2lC9!McSqe4WF3}ff+oA(vVP#FbgsSewrI23-4q{2Ve(PNC5n}fa~3QK7IDUI z@S8eg$9j4l>nYA7H~)jtIVZv#!!;=odOlFFSnKIGtf#Y>ky=lEzoc}#o-V1{Y-^?V z08(E1>%kDGDCLt>MW+G`CR08EgL2v=xweI>GP#o`EjKV)u^l3v-fTixO!P4SDJyKr z169J-JNDQZz}xz&vJw^2DK9NN%vOjp1QKCGqHTO{dwb7570NWPt`Q~u&_ceQ=1==t z2qW2;eJW$S)a{~?)0b;6=q*Ok3p?(HBVaklaqQ-F%P-$*=`WRmzf?wgioetrr?%X3 z18!s99E`XuAK!GIcyut_?Tlcg>rc-Qbz{&M2e|t^nEr~EkRIfg&y<_KeLy`ck@cHe z>C|+w>CJvqy%=YbH^X4$AeHc&>cz@J@;iQV!f&b<43p#~elq;h`c3tMUy}TYpPcZU z>IJ(bd6rG~{HD@vOi^#%^=c5kM!Dz^M#T+;(Nm$k=IJyO`&15d!{4{dIAu-$?Hf3J z`&9jd+~hHuSrh5>!Lk&(ocRQF8LndH<@q!3W2aUujB(%`aG+pKAkyupfI9(?%(t3n zp32ynS59<*Ey1Fl%shL#b$}bW(K@X;H!qU}kp876p3KS3K7mi#iPvV{**ckbZu(oy zJI$v+(fhvK7WuixqXiHEpj12chOsm4$;=z3iFfhRlny>Vuf@b0HBS5HA37bR3FvB` zOOv8vU&>^f5%2Hgv_pH4x6fs;|Gl>_?WjN9yeVfsi|ef5c5}783nJC1?bQ85b1$jN zvsO4eFXhL$L#r30(87`a$dIaW9%i3S+gG_)85X>`ix%5`MAlSGYLONe?G4E zlUvP4E>&sa$LP<;zx&Cp=Hp#{a;y1>j~kUF<|C%76hFoCP*WBs5P5LPNusBfS0s8^cP5sn(5E^W@W!9*?K?T{ zNo6_iGf6=M)^W$vWG)R~)>g>atXd>wysC%flH9J4@zgt_o$s|Bw-Pd1$QWikbqLn{ z3tHpCse4c$)GX_nbhq*CzRmK8BJJekTg}Bo2qg5fr9d*w3M3ps^yEZ|3eA3M)=$F9 z2j4G`NQEpYXZ)-+lu{u-Cmxs_vD$q0%SqWq@&Q9d90Ec)<2*(z`g3O^Mt1tcpE+?G z9+uBZfn2*D;i@m*6&jcoF3SlGgwvu83Jqk4s&b{n8uSfEUaU_TItLpy@L$sVO!8ePOpaQv(b6z4ggm^l$BHDcYZ*H*)^e%C%7(@1%1O3vW+mDjdiTUYdMB@Bf?sG0sTwn^ zJ8I>P@$4`>)vxfQzb7mF7t^Y3h0~}o_L!ghDdmn#VC6sy*=Q#G)*FuI+EF7!vhO{h zFS+_WtYG27&f1h9_Cx@qu5&ea*wKS0j*Pa9@Jxz{pj%t zJ&5nvA%B4JEl@nRR|A;k`lq_r!~k(gEF=a&Zu%lQ9oo5>bTuTs5PV+#j6iGjqaIvj z-eYi7y&aRloa5!?~~7y;0Q8Z?rf zH&O^0m*=`QOBk6f^VNDSOr<1*3oFdlLxna%Sqah+AjFI|sOCbSg0_chy`yR)GzHh`IVgOYDpE@eV)@7ZOFxW`MM@6N5>fOXrG*GJ{nA&nhBBpX4AE$ zZ1$?Sc~VE5>fWnl>P*ft3w6{Rk_}&!wNN|~=jJ(Jic84XsuShdD)IwzCR0;-Lu*{9 zghC}mdsyyKt!E!O3YtNbvcX`Be%_(2<)LJ%FmL{iNGje?lVPA*vHd;j?$-USydj=_ z^J5;@!1;1%jF=wFW8H}%>xw>$! zBV)P%%GO9_!iDfDpEmb?F!~;Udo{$2Cx80$Bf<4Jc{}}!*%V8eE5z+0L7wK)m(Pck^zeYHDZ%Z6{h&x2RumXPRvely)PanhLC^dWTSZ>1wor$ zSDqNp9{q?&LqpB+Nd##dnBk+09ad&FjBTm*ucoAFAb6~ z)A7>E>t&TqAl9j!3Y3dn{s798VjGaN$eKZqOv-g5+@`LD1ySR@&ASj}%W@TTC<09A zC0ZI*M4|BwffI??J=I(IQ#8?5KnD$2T?$qcxp#^Z4H6um^7@|YZ6=z_4@tr z#~&0=?Je}Pbq)I-Pc2}9YA2provrY{7Ef)rN2FgPo#IF7hWiHbdk=*Fe77B49X5?u zo)ie*JezkY-Yww2)v}+>I0{Z|)7q z*!8JJ|2n!MG_^bVuu=wx8HIwsDC*3Js$3-F5KlOjQfaMH7OeHkP4m*V&@V!jtv9Te z<7DGRPUMF6N_$VRXm*ngiAp3-^BwN)e5x`Q9rlK3Q}Fig_*c$N6qmdH3dzrJD- z%S^ho@`z-AsLWMGpWKcpL`A%@oTOD|2g37b3q00=92>^&{j$4!j_2*(@W-LPk|BPItl)R&ZnEkVwfUrFp6F_Oo_94o0q) zUQYq=>=X{X9l{TZU;vCWuYFKe z8T}E&7mWehi!XQT*V0ObdErsDYs0i4 zg^eTRyCvs%_5=4LnL&7?qDN6Wp_h=sdq3V8;@F$bgJOFw{g<}Z>L_Et9gLsagCv~o zmP9hp?o~cw;nl{*)!skMseR5Hr-W+U{D;)l@FU)UyOI?=7HhEe*V?#GDU$OXx`E^@ z0%%@S5ZbCq0G?L)t0q;Lx4wd^+kChLs{hNiqWTSAe-mF+OlgSt9$`l}otbagm1yrS z*;fe5gUC)cOTSZUr#WXJl%fn!1hzI+JGC>YjPswfCF?GY{Y-yK{T7yKen%!iiJ-sH z)J)fYrZv!rFWG<{&|}({=R~gD;3)Vw|cv&VspzTp|*9|TkNcQ?J&vEeLgvz z(vj@%-rJ=StS})fDld(V_o4ZTpz->tf={jJLqy+!&=^jfZfiG<}! zU_c^$-kKzexMZW=sbbELvdLK^b`lIfGS~h*G*8uq*;rd8QBHZ$7nMDH`a>`d)GW_6 zyXtqw%ZqkZ!3`eqvOlYn`);JBfuCUH5gXnC;xS8Dnya7HHW5x#qfGiovDL~LdSop> ze9)P0{i{5nBz#ca^G;}gY(#Dt1sh=_&g^vYed|!sTuPcV-|l8mev$JqBC^~u53E?4 zSpK&jQi@@Yh_8r?Oz=S&Jdu>LSlJkj!~R)8REKWNhYO*h4~?t-`2Gwoz4wMi0*v zr)APNRGvkkUiJ@x@bw&4pO(v(1lD9$K}-ra5n7@>6|s@tP?!aWr=)MZk^MEYDugIYETzP!wwk+0#=CN$FLdOnyfN63j5H=IqTOAVaH-qKrP zoW|J{+GLKs+2cN$rn`M=gO6gNDUw}Bv_-1?(u=m95TPsF%uCUqx) zh?_e(Ll+D`G}ppibGLqAJ~;$=shSjBWveSqZc0a@bx_oNB(0`|U@jd(u6!wUJA;+Z z&*q4>R{314szfv)`fX?@l!|;(Kd2aW&sINlb&*q(h_70N zSJJ@%4z6sMUkwz!>C9T<3F(IseF)Qy7WCgzA>V6q^cnafvONuADQM*CZ?Oh3R& zx^w&cC`p(jL+C7Bcl|yqfVETGpl=aBqA*lc>s{SNHTkier2I?t5VBuN7i;gZmD!Gd zhQxW|e!3EUd|KuEmJBP*wNX4~pn4`KkSUmp_-t)nHRGT=+O(fJaNVwWX?-G=gFSIp zVOCft`~<<68=zjO0gh~regi5|BTlWjsUi%#uza(jzBn`Vhv{K2C_w*BT~##R#IY%2FD5xuqtsd;OLCQ}PW zT1O-%)!dU>0QY7Oru`pv*}Aa@OW$V`OSnmX&xIT?u23iZfdXrWRpU*Q$YNI3N!LSNEek*WE2!n z3K|($kspC(%`D<3E`3+-pQ!QV8oT3ntg*HC z>O3k|x0TU9@|4)mN(^hYF#J^5@>YcAr9jF^v!+@=Y!Ad=cIrMMSB2?%CsHu0stAu5 zwH4;#W&UdMQ<-R5J}diknVNI(lkJ(0cz&a3F{L7#g& z*o$!`MOPT*>UV}~X>^W7?$YQw+p$)IRdu_WcjQ;`*jtL}yEOW+pV#|c9=(+^v1b5V z6%L|8?!TT1VT*0r{cnXS2I;piF#93Jx-CfDB=O0u z%D2fZx=Cc$rn1j$R~Wyo`xEWhZtux_r_BsIHT68%1>&!_w*yS&vgpH9;5!N~ja9VM z0@s?y&0miAN*GIoXVDn``4FkuHGCecw)Z7xdK~nSc7e4w-bWP$slxo0a%~%{?FDY4 zhp11H>EMI!}C$!j~tPmxI#5f4U+FJjxfRDWz zOc}n0a4#!DgtnnRINaR%vKX2b9FT`+ldn8{yqIGqq48PkmRj?|GDxsQJDcj*fztDo zho=!gKVCT`5Wa?0aT+%sY(P3jvu?F<2g1|h*=ug|;^(`HpC1hteYwZ@`QK8X>=;$^ z6hA++l4+iZpU;xA-i%ynRmD~aIUF6i8_vjUsxAgtYs*saQg-A^d4Q4fp z4}BHo1$GiFwM#)lFRz^j&83_h>HQ8Yo5oP0Gz~0t|+ z5iG5;lcaUm-Xo#=XAqtU1NtH1{uC*Ydfa%mfQ@P*S~~`v8#S2`2(9roS6^XVxP^`A z!B#NPWe(^ntp0vNz8UtaClf7?JpldRrs=sx3LqKk4zf^PK%$MD=$q=ep&DuTzJ+jY zJ$(Hfd^|EDX{wubowz1*c-V*VM*%3i6_&N|c>E}el?Y>Y#DKUOoK@Dc@O+yYWO=zW zL0E-3)X(hm<11{Hzkg>&Ad;q6 zb#X&K?bWb!rUJ`vcezPD)Frmb=>au$3iTIr?ZQ{ye?b1r_c4E9AM+RNWB&d2F@Ln* zzV-L{ea!#2eas))$NV?!WBy<4WB#Ij%-?+<^M7;ZzTy9!ea!##KIZ@9KIXq+AM^iW zAM+ROWB%@``7w**+f1-KIUv6vFFd#CMQl5-Yi-;g`+uz!1#;bjm<(pqTtXelvhVO9 zsbYctIOR*RzY&pbDfaee?Q79n*Nt@x-U@_vm4_OGEC)res3m9cJMrvF>cz4pH)kC7 z{Pqq$H`1@S^nHbpP}2Qiw)YyaFC7OuYL8t?fUaYbbY82@>@-e6}0@ ziPdTqp|%~Eh5tj*X5^Sq70XaL+2q`!L#ntxBR5^+hR$?D=)sl`v};6D7gSNU>!>J7 zSu-enq%ExFFQshg%-Dp^E=lQ;nK?%dpoHQL1>K1UYguL7HrWukgvUd5>B^fr*r)|>ZQ1PcfT67G?JEj zSUF4e3#6DeBuoCQ6G*Kx`TeU~H3<*nw=o1wI<~?zvfrdC(1Yg*T2lI4#;)tneg(1r zpeB2ljZ~z6#+@3G7vK9PMHr^tl3~HfHCfgc$7lpYr|HCICH}kgfxdC>Z-zt(=xc{_ z{ptKouRID-t*W0oP|{S&Ce_2g4S~r0R-8Y`t1runt?}U{t^<0KB$#Dryr4xCK>|3Q z{WXLZ3}2NNdmI?2ab}S%ApkeymodLE;WU~V>xexG!7j&qBFSRU)~Mf1Mc65}G6XW( zP~H@!X^BrN^gC&1`k&R`LG`&~bYHY55)_=YJ0WIq9z32LguY^u0K_5i* zJt!DH!3|%-HDcTcb1)moK@i_y=Hs9&(7`P>ojEaEc@kbj-uxj_A+uV>Lhk?L0lTwj zKwUH$q+z7}fCHzMr)p&2JcyMS1q(j2ZcWkq$T2fv4^%mFKThtdRjDMG;?B(cPk3Dvi=wH zKw@_6FSlsot9E(Fnw&jR)$RZ`HEDly`9XY;7Fd}Nkg}fKvy@xfsffazfDJIGCu>W7 z>PAo@HED7$R)yJpS)s&fIE3X7+H2guGc&DeV19aRRXqFS8kAVf--Lgg@6l*CX=1Fn zlG?2>ek0bOq%onk4ur~brJ}olPKE~-Yc>T!S7ip4bL$`fN`~SQ5rMxm03aNp-jrsYc}r>H2cj4pozJPV6Vk$E#i4=4I_$E1pap_o+kB{FAoMHR!mBzwm3vMAgk7QOqRB63 zLtKAg_sPlrc=j#T;;WN$V|7XX0y#}8g6hOiPpU9g z=qM}~q2pD_Dv(NJjHlGu$Fo8EMUw!m8@vI4NS#4GHu0y1ASss1Z|6R&&v+_aX zWzScB3%4)R`qLBnzVgDFogJ%AcEdenY{v?XCv78x)9gt*YZ9811N`}Aqp4JeS1>X27|D&`I_J4@&}@Y(FZ@ ztJe#)P;q?wL#goylfH0UWiL*l-)LB~VvqEDCtBP8HvLYd$^XOjyL{^R(eJ;k<`F5N zCH>y+7fhw!&?-y6UxY-zPrmNa@3~nX{f_j1C-eJ1q2GrYoM!rcqiS#T`-`4?rQeT2 z|2OE@p%a#V&))jO=(j+C@0osAZT;`l@BI4z)AZYA^7qm2Ry26+@ZVQ%^a`fZ@AI!( z`aSs;(eGC)J^JlGzbq((f_q(*FVed;WDk{Z8AG zLcbUOKla`PuIgg_9|sjfQxBSyEh@LDpt7*S#55s6KgSM+Wo9l#Wnw!Slr1U-c=>cZ zrDc1Ww%V<0R^C!GMG??ewN0&D+MFXQYPn{f|NH&SoQ*9tyWj5r^*b+RKFfS&o|$>( zd7gRC%rj1VFNSX8H?N2653!l_@q?X9d2UcNCLQ6Mz*l&nnPY+nH-&qycTAXuiyk~k z%`xEvUh*I>$AqbjySN8WNpnn?%jY~N%`xFSUh;4>#{|yY^x&CXjtQJ2vM@Kd?CA>~ z6F4h5FA8KV;E8U9i=!g1>?BDb5%^OTshW z#PQ}lbCblIF6JhcH!0?Zr<_D}H#gb5>1l3qd2_Mc6vY>yFroLHLkPW}R^xtF=snJa zL^Z@23oQn zPhIxXA!U>7ZGzG7UWGIq|NI4J0^6wbE*^M>YEJQ920#~fj%*2)lN-Frgn0eq;{{l} z6ETmLhqpbRPnyTJhCc(ZLw}d1 z2`6>TNLfB=oA+3}c6$29q*bKFbOIr*JHtg#>vMdgC$wX4d%#f#ybm)v zGZnA?-Zb3$VlN&IKmC)^1^|Tjk6$rQh;dU>bhEay-;mQy*xA8Za#&@0^1!ljLQ~3$ zyb-cY94B>h=LI^!k6dm7yKNmd$8o4{%(n^{c4@xlGPFpk4jWZmeH-;p=%~U`{49us zp$y|(&S)wF-aZyi1TfdKqRx0-$}XGcTn}QN!{PkID zha;io)`24{93&H`FpvxJ{0tjf`yL z8J%9(Sr8J{0JmbJFXPN#%N3;G<|xt5g@gCov22G6$iK%R*`2_C<1XxqKvq!j#Ui|T zhEK4#e2FwO=nxwm6R6Jr=q;T(y1WLXOX88kQ7VBm#A zhL@o)FvP6E2Lo@>d=T7e=O})PDPc>a88W0|USVt`Ip6{|fBs1BjB16}b-a5;L`hZ* zIlxi$J9?DubG$d8$AJ;{fRqE?&3ItO8)^0)-)H^YuBbizY**wSB;|O3ld{+dkItK6 zFJi~l+|agJTFRbLo4qsf8m>TWA8OV;pvQhBx!?N~9{OUN<_i;e-)I`HzOi3f7L^h> z^mK~o`aN(to^p-~0yHPXWddoam6qp?-9RYv!v4`%zpN2LS!d7j{%t^SPHv^|^NL^~tRaoZJYj_uxmbNcM{gQ!cp1b&eR|)d-Y;hntU*CU&|pz&QxFT1RXpge z=oJ^KA&KC8oHU1fTq?fH@~;Tb(x}GT8%aMFi=(Pr5i<#8Q{;ljEzP5~$j-NXNEybP z?txfrR$c1k+}}3L1Zq6Mz7m7|fdHJ2V*4IL9Lln=Vo#*~mINofuRrTs5`!nA%m)q5CkIQylvkFZXUx7badFQM2ex&k5wFq@+&o)?}wpCpTnM4xoJpOcS0y+-mJ zH>Tt1^`uo;m0%?*06@;f3N)zWp^KRe);UndN;{ykS32!=JUTkgazC02N+B3WGJy8$ zN-k)RmU(B86an%T5;e(@2>1;0kI^7mA!89`NGZi466v+v(YjYDB{lRI1)+WCatyRq zQA2jM?8{d%(OQx*grFZCiNl%DCt zx`rxVEVYewC{74@t#bUNi@pj}+X+0w8(Ya(1!QTkGd`dii3`+OK2!;jNO*5>w9bg& z(h7f_VQ$4uJjca4gY%p^>WsinSNgG#^aD*AO2#tuE0BRa0wn0xL$!@DbVyZ#gd~s< zXR0kT3ArO%@%_*M=YxHj7e3MKe5)Zqs7k1MrEbP4=>3WE*CeHWZd9%7Ylo% z9smsiSBGEKuox^xtK9THt1puiP~EXoo1z-0a1lJ?!S##tx)Y&&D}< z)DxNqoBT(re{d@qWNY!|NHNep+;|v0!s)o}0L(%h-@GNSgVX&Vk;aMW&Sb_!tOcKZ z-SZLBt8q5e2YibP5w`a>!)PN&AN5J5+todwPcVi;*+O*6rS7&Z`q^zor^xCc8IkMZ{|D?gn9?=+FFR>(J{Y6?$(?j56iwp)`WLE(!l!9n_-HC z^~3y4Ftmo|6Se8FFnl(W`V;d>Wa5D4uk6zfl^r3PA9mgqGSKw9Z-5=>TMkK+3F(G8 z;X;`R7Cnr)2cF|rhn0aE#iqp#FQSVezaSoe1M&DQ;_-J9k9Qe7J{_s#QaI9t$qqGP z?n_l;F%2nreLgN06p$;3aQ~Ry5MGpWVm){fzsLUfJeW;~iWlXJIeFN4ePu=}@)m}N z#Le$onIYpN%%fZQp9M!QgZQ3v1HDu;lb5!4!siDNz3%yeL6|(q!u&ugbj=`2_H0P@ zYy}585U6K*zmd?KN_d=w@gL3;gye58Ym3LksJ6VYtEG;yNbMTb5o`LLb+yKaGD3j8 zcR20YPh3-C{v`t!!S-y6?CPI?v5#-CJx>OBc*yN}DjCcE@3g0^-$q_XqCI@&kC1JK|5(1k{J)T`Uv_{8iGZ>IvGrIjaA=GVe0`R%-D3-s8MWx%Z}> ztUb}ESZ)Rpj1?kOpjd6yQ}ZDw!h(w@ihq6q`39DTqzL(iWx%<(xBjK8KtC3HeCTwa z6wgWVL~d`BR!>|@>)rbRVqj@VD(rcDgs9I~eJ=S3pcOQS;BQ4d6q^0Wg_|wQ(?0bd z-=jLQ(;^J5=}Nyajkztq0EGv<)LBfIVL)tez_R_bom|xqTF!#uO*1WC^6TYTkkNFBRvT}!rM?{Htwkm3vR~p_r8cLc+*0v@Ep~CDC!dD z9%Bi*prYlf*&2EQz+$m#)!_=+X`+sG)ufJ4Yyp`As`zAl-*)?V%=?p6W(otU8NnSJ z=htmit1mIn_L$BVEMq=4+kmi^t7vVpT6ta@hpJp6hg7e|sn2qt-KV&jjvMK3<2Ogc zzzu1fA{lW2rtn0yK%|O3S#q-Ga@@Oc58Dv7^M}M(S6QAuFjx8rOyg*MX$&7?)nlJj z#P~?StI9svs3PA8yU%bAol;2^FU%)>FzCM&Ls$+AVMUyVJi01YfAWUUDU1eSFpM-S z_-8$lTt>xcNfuV1qx(l=OwwRidF*i7y{Dz~5qDVRRl` zXMyp(76zaNenqfjkHaCWQcpGM;uOh_JYJK$+#L0+fQe@HhnYK|iEIo_K-i!$d=*HV z>ZmCU(gAb2fps7h6p(D8>p_B# zR1lrXbtG!{0DxM=h!=DqH)wyG8owoeG-`FUztjc1z|XSZFnbc;(A!v1J5TCYV#UIf z!&a3lHb|(w>M!3!^YP+Ib=3bvvoTiJw`EU}ltrMrJTJx1KsLyY0e@_lS^5$k@TK@U zxMo;t-YqrnXj>0Kw^aXHCv4%z4Z`f}iSoH-nj-lyls;4c6Ls3yY0G>`AM>D~K_}qI zswjTFe5;95=Gu{16u+DiG1fWN1?LQM14}(ghFshgxEv$UTQEt1reO-gdj26siyg;mHnlD23SFy!Wg+IRI84PvTXtVO zH?<3*nFxOXV-xsXm#(39Muw%#aH6zEAWt@dZ2=h8HHAD=k!Ke2?CQ@`YAmGXkmsNV zTMzWGB6S6)=8{gNn~m>Alh*aPr#NYk8l0_b*3QahY*stpI$Gyuqf6XJr1SD5c5o0B z>$%4C9NZ*lVLv$fipZZe1l6+8Ib_bM0)nWF5NkG`KyH;{K_#681iK68b~O>D@C@OZ zFEoU{o-g&O-uRMv%#SYsN$PpIfzVKo$rU(=X-ij33jJ6M=n&QRB|gW-AIfp9*!=6# zV7f#H#7`DvPF7r?E&Z}6uo7*hqd1iVLr+%M%G9n%5)77_DxsZa;aiCM)Iq%W-X{Y{ zz1GZS|CFP$`;+xdAN3O*B0r?Rst>+kado-0%SKvhr*GubnuRJ_hm3r&Hw~s%EU#+L zKhN>c0;Z51sF}m4^)5n&sr5wMbS|YKLQFi+7;W&_IKwRGp3d^BA43~ePo>BbT69!? z#r~nbh0+qWS2Yu#c$*?cxE%1pmj|j9+a!an=xxN@sN1hzPIay2OSrzCnHZw4x8vpy z_4QCh3f0%o;{LDJ*Cmwep5)FSvph#mUr%Nu{kQb>X5f9oqOgdA9p(@8b+uS1A$Egb zpVhr?vtE6D_l+U?x-Yu^-$P%=_=|^ehSq!j^8H`x>w8xUTm1h*U*84blH#}kqT~PR z>%-93pWG6nuX72M{>A$G*-w~<(HsxU4fS=gT>UTU>sii&(bxZ`Os}P{@21IIpT3@V z*^$@RrwJLR){n(vmmoNT`g;0jOs`H{`^b;4y9@Slh~sPDr$XVM9$&wASBSnIy!`K?ujBm1 z`&;Sj2PqmkOGPDD%Z}jK@PB7~T`}nY>FYl=zHWJMh`yf0Nt(Y{Ur#J&9!7JFlN;*m ze7X8x($^6eAB?_!jxxQLzOMY5PwN?9H^1b_>+55M3{&gPa8rlA-cfcq`g%UTI@Iy? zfzqJ9<^nxT>vXi-7li=cm=gfnoia8$565TX(3bp!ie62$YkLmbFmZ8aTUkM#e+K3b zVz6-1KVGM!+a}n@`0G9&mc_0_)i=SXk#^&%MZVZKomi3{vYHxl`7g?U=Pz~Tf6jj` z|NJ`hf77a?{8(k`ZvsqQgsjYu=E{7j&C_*i%~AEbHZk=W)(TB^EBInFYvh3SC7qY) zHOu;(5qbjf#5UdiUHJUQlK{#%zmfdXx8GsX=06vqa}@&t`k3kn12Vo3G`gesO+FqU zICP{;{1`34L%1jHg-^|aBZB6v$u}P8I_DXGr6=%fPqJ@_@mKoDW7F{o=4loV=!o2~ zdd{7MiJcsEkG!Ba-C0+0wy*UQj0l8iU`5-Ah0&7V2WiX?aiUu9UXSUxQsB36lu+ZG zl)PTP*v@^>v1`7_M>bDAC7ZV_cq=+j5U&~^=fkhphaX`7v?tbmA#(F1KeTH-$SLl} zto0F+v0_yY|k-jes8Hy;@Uy8!oR)YU#?ch=WFO}=6K#P&c2hu1!v zq`vflFhX)SG>cqJ9u?j`*9Rb7}h?kbd$>-G3_(1f5)))Ns$+9 zpWK0LpB2(xXrE)t0_`(~U0vE|#V-xF&-PyRw9i$K|8e^~I8WN=lCH?jliUwk;jofR zm~~kDjDeTR95tgSUmRTf`27!DZh3HWxAx~x-w@b<&N=pWmN>(c+550{=X&Y~GS7zJ zIysJg5#)(+mDeVq&obS&!e8E&ysS+3m;_e}4!zh1Z#CjMuC)`c1u$aC71quE=#+adV$uu)oF6Yms zg(WM$F(->vsR7TdL*z|>NFG7LIu%FJzN~iQnq{lH0`)-nIh&N`2K-gzYGRs@7`#ml z9EyPHDK6Zox_~cap$eyh5hc#LFbl4VJBeJ!%+$&}%=v&Ve6jm3Bs=30{j|6x-q$Sh zR3=8RhK|P;dvh^iA>(8mVgbeM&Bi@~t2mDVonjhDi~0OScpG?ZFUssw-OuC-hiY{` z!demYuma|pRFHvAu!es=kt)9Wy%&S?A9d=7v#aZqKR}hX zw0_vSrM~t_|D5e}MKW?ey!JVx8;K6l5BH*3q*7Ph6y841UuN3p+=-@ruDPURSo<6= zFW5d=S!|!Tq`lBSyWnlow9hc<_|)HTZE3iDRzM2YsxsPFDU<%VeQug9?Q`b2D4i$y zY-E+APR(FZ!`f#U2+L6=!ZUxOeMW4quYF!x!S?wN+2rur=Rv8jkoH-FW|8*kd}Da~ z_)zDfzaM$fw9i6kc>CNVFW5e_`mudlOM9Vx2EHF?pAXp8wf&1d^bOj--Ro(eryly_ z_WAH_X&-kIa)bR_#r7!{Q$MVIKA6Y$QIwnqslV%#Z)-Qz*FH^_vwdR8CWqHPo20%% z+NT?uMcQX_PI&uVbE#>c9VMoHI$9mW`p;^4!S;#CWc%D9?S=MvcW$74+Ow-``*%~r z?K8PsJ?*o1;vcuq8E;AZRGp37VE=mAK0k?>8`eJUK^W{`se^;-KUV~Im<|d<2(WE0 zM}5*qP6B~TJeUSaP=*C_=pr80$1Ai9qUbfAV^wa0(ujehm@b~G{1NOb%9kEph%L}+LCkEx|2lnSuzqOnK_awh?iN1Y!4wS3jI|X;S)#~{;D%-8n zny+3r4WCr2_fOGF4BtUW(}0wi(b$Q;*!C%`^U$aHo79$tJCIB4(R3Q;Nxq3{NtBjY z=4OvTOA02eZ>?5?O$L`S1ID|9B?(XSajBmPZ@yUHcwE)T?`=OJzn^*zk~}B}qJU;yb)_ooNyxD_)%m#W~*r=}%WB`|Ap^+TIYxw;p zlfiY&fbqstCxoY|`Tg-X!0!W*N)Db#cclklUdw^fFQ}g5#lWLBzBC$M6tgq(eo~pn-{QCt*~pO{EI~v z%+@J&lUEzDfwQuJyhGVpQ9zc8ta@6;VZ_0%)RE^@E%C+PeExRzYLAW*-Lyv=Y!z~2d3%1EG{`ABDf;4r zeT8i&xfi+`^~v%j~mz?ubKAv*@Q`^J$}P|{q6C0z%Q!L zI#T$>zhC{E;ukAU2Yb>;88?v^aw=&9_{D!_fD!A&FP@P%e}rE=hI*(Uzep!Bb>SCh z%8P@;FS?ycQQiQ4vGz3Na=7?~qgdk?_n$x&Q?2q4Qdx%q9mFrbIR|7wL4@EJQ(t-U zkMN6i_1Qb|OK9&Tlk~&0cbZeehmMa24>R`8TSLO_opHU4z0+pQbj=$*yLUvD2Jwr% z_wxnqouo9{JCn!q9(v%be+TTH9*N8kaiS)#ZP?y<>ul6iO%sIUCc9v{PF#h z7k?ywl3%J%{+uXp{z(4Bq8{p(KTA#q4Ry(%*X6~*$)9!Fe#2jY`GI zkeN6}DN(t26!fKH&G*4h5Q69WVc`p~2w;yo7wzYnF=JDp+Qr_(DeySwxiU_UXU>S< z=X3f#<~h%o;JmC1{J#~RT`^A94pSsG_i&u0j#;H(QH-iwf8u z@I%a$K_`<^&vDS|`zT)_{+u|L%yB9FL&0?cOPo@k_oVoNr@wsoj6Q>YTv$GBf{%$~ z1K+}VfZGILhsSswGyf}z#drA&y#4Vqus$Gge74l^W5`lYA-n&4vgG+VuLq(K(;nrj z&CmnBO!p0_xnU8`lyCFSGW$3Kn>N$AZ!74}ySgJzGfeq5|2uxbjgm1=cRnVA*I9`; z$7coKMs9_tY1UedcM)u2!Sh+E`0CZFrIJ44$njP=>ZJQcDtmW)K=ZoN+h_sIIN8sc zc7@a42`SvEC(LN=}gZ9s-bHbqN}jbLvAeivf$c&2%)Oy;khY(|UBIGyvDB z6OTa)c#?m)NZZYurv+n0=vb_booyu6S0+|PNUSwfT0CH9y6b=shd9SEvvS1GDEcuk zDXm~kL`41_>^h_;FlKQSzfNcdTVxP~@Q*acab0AT3m+Ei7Bof+0dN~gB?|F%%K%(> zMiZ2|5ya%EF?UlZ;?3e$+QKJG2KMbf0y|kYBBAJ@rkdNcl8}t-%8JWTPbM>Y?2W=B zQKwIAVWH6j*C`jQa-O1|;6m?6%1rqQ2jwxEC;3ujM@!_x=dpzZJm|ui({gfks;*Tv zegu||;t1|YN%KW4e)fY=#WLpw7z+Mr;Eg7=qJSV(q~0urjZVk_giPX5iH_DK#H?gf zC7O!qMX2MhMD2q(2SV>#;GEzLen#gXU9=Nx@&`SNOdOQDiILY2orY^MB>za9t6H^Gsf8L~|a`{!<1jGmr%Jj;h= zKY2Xp3P=9EbCTb?3$y3TIPz^=W+Wy&afG6GigXvwFM8&%ez;NaS}-AwaT^5iV;x@0 z8EBK73$giyzj2a)_@kLdfB@3f8;9rqb&v34Giqh9%1}-4VY8Sg%H!zXDlCsIVr@uu zcrRdrv3HybNsoC!`87rj;?S~6{|A$6ad`eGLIHLv-+m4mHR~Ss=W_4E^Yn1}pC%k< z!q+6M^e#5{TTFP=`{?(9MN1q{@4;|AFrE(DPw;dW(*J?6p!0L_l%wctwz~TQdPOU^ zsFS0p68BI+oNjH$&e4t~^~LTv2`&RyqH4MO9uCs^6p#M_MO!!~;$)n7D&IB6yt7`5 z<#b;WZAI?m31&FJ^-jL^r-oEI!%Khr_8YKw9Gb#CCi)W3`4?FiVl%L#=rlB8=aoFS z?L_Z&_$;J+zQxXzuN_4$JcUD|7RMIsy$0R}3-+Gwn6jwZ(ixF?-)V1E&{Ij4y#ss8 z*~6hAF_nA^j-|Wr8T4hC1##@anEBm=E_B>A&dB{_nW+S9?eFD_3KfwFhOHU(bQAT? zu&7^hgSbJ4;D8d>)i?ytp}qsp(;=6mYBZ!hbj28yi*A3>%c8b*!7;=)Kjp)z&3PSb zzy5-+UGh2~KAbWlDj~e^Guut!;?%SZyW@9{PWZ>L?Y z{VVjsk-QgyiCrg(|Hk5ph=sQHL}h>qQOdi8Z1w(=LhmWH>Fq#zpOx1*G!DHWl;&up zhZ?Ry^UtUFS!V#^TR6^urby!qsBM%D?dkwJScmu9By%sa;gP0f>Rv|ZvTn$Ok=vtr&m zj&{SA{8c>Wfj;l>2O=08O#{eL)Eyd-4$}(v#5sz5U@cswI*N;#kSF<+^QpJ1)vw3z z<6*;~dK4TB#i=iVWHB8)lQd0wD*tjLw{aS5O0VcN$77Y0_K+NAR>D!-l^GT;8JTYH z2~o@exmgZ-1xHSthBgp)rrKAx-PsvCJ&ybXm{Z}!=up23VTJzWfVavO`teI?@R(3>L_X}cU&MGF#pCL3$ZRH z-Xq>l!ba2&Sl%JO#?Xi^R2%9s2F~W72jO1sdZ|7vIEGN9y=!@3O{6S#JW#Wgs{aWM?^Tui11{^DHr;;J}Z z5B9lqre}|WrC1s#+p7x~qXrk{4YNG=B0+19Wp~4=iQ8~;Q~nCew+KEk9fi*!HF419 z1U+yG7X!m8CcWf1@y&KKX>|A7Q}qV?g|a%t1}>&t^2Xy2?aIu=&6jx6xU!3^9avJ_Y`L+5Je$NJSTc_9J52ylQx3|pb-UZlsiRKG;S_Tb5+nt#&l z4}Ht}xrfHU-z~gVljsLVqT;!MB?jKsGCHr!us=^LTpZok-L_p{yKTp`ln+O3rm%|f z_C!QKKl$rg1ToH;Oh9k2OG!tZE3%a2CpbN+x*XoSNe!mwJ8P!*9c!50QFIR|^PVft z*a8Lsi(d6=nmFwC9*cMB?ihPnx_w-_dqDL5bbEkfzjroM3@CN*0#;Xv6(AMo9RW)5 z34H!W9~i+e4tO8%*T@F4E9eaN9UiB32{c~#Q*roq^Xaq!TzAk0D63j17VEA@F+ZW_ zd&EoWIdC@%A)@CgS&j_6of;;O#Vg+U%}Y_n+9DYz zdg8g3$`;I5KmUdM<=HYd)Gv>ts2tyzVV?2L6Wxa5`n!Bg_)_1CxQ3G+@LH-YZ#%Uq zzDQi9#r5VWwhG1dLre&6XwKHdtcB4nB=BhEG$a^*^c6)QO({|{6UlS1W8=p2IApQ zx+uN?AM&OhZzh?WB;HIjHyC&M-D;qHFwU~v{1&pY9n#Ac(K!-vC7J9Tb@bD$Pb?Sb z2}Y3RIqI89IBN>@OY(4q0TIY$c3k7l&nTtaHNgd5MF%`qOp{liIf{x!^bOE7!`VZ6 z{75(saTMLn=k|O<^b}1m{8und{&@HSGM$CXz_ION?k!B=<-{tMCP#h$EJ^nh$`h4C z)8nAD;6=umtyA@XX<_-Od})Y5{Rno^hG|V>{xW^I$N}XAn-9*F1m1~rU3deu$R~(E zMVgI%K(|Sm5$p<@I9lcC!&1FYv{D@FK0DO!O$Dji^a7|V%!p3F*a!^_`U;S!qxhfH zeSRuYzD7c&)V-J*m5!p_QoG~`jNvq$$CFNgXkg*i5PaS|h*7~UK!B-#DWUCWm`X1X z-wxnC*4$4%*1t^qXE;fAVY@-V_3L^12TKI4ehbVFdI2k%OhWyN`9q-^in#@?5)^+B z3KNmGFadvgo5fpJL>46#s%@(?5}E8)$M_&>%~oy*pwMaxsl|b)M0gFUXoQM;YdV_~ z_3BJn;wU-_bwMBiZeuIqzBH!le(8re{|jy+`(Q>m1Ncs#XoQdgeIO!znKnjP8F9|a zK5--pVKl=u9^jj!E=TH4_uX;g1kC9f7YF`9=Wx1vI#UkhwPKpO#$Rm-0aI9+0E_|u zD-tqO_UGN>^jv`~uR!#ph+7qN89%d!&lbAdb`-{gTrn6v!Ua9LUXK_UKs7CCM4O=~ z3%BCvG`$rZou)TGz^}#nHufM}T(@%~tY7a$=mxPPJYy0thDZI1f~F<$8AkDo=&lF& zWxf~^aX9)tM{R9ktO(R8+DX_?J6+Z+C^|tDbtoI&_Y`YAj16OuHPkphoj-kp8x*^J@?i{JlvDCfLyG@D%IcP8S!S`=oQ&?SjjkMXq zGdAcSw$!dDEJb5>?_$~A&rMl6s@i*kW@UV*IZoh&{(v3(dRla)qy*QEPM{a(LEgJ* ze0Y){NzfAh`c{4k?%jNjAf^qF2gVwS? zd;kuO@IHY*)F0@>Nw|dc-mNh^F51SZAzDzN9ChLz0&*P66Xr^@%Yob#&%_7N;h8DF z=5_avXTtpJo%#g1Dgf196|Te6n`8bS=+uYAqTo7^OVHuuFmYFoO2jMHObK!bQLn2( zvsv+;Zik++N4-k-&c*kk)~nu}gx=l`r1^9y)s{E;)RZ%q<#gE_SWZ{sb`}t<3Ka;y zP@O}0Ml@(56&~dsj`6c2R=v^$%l;RNsc4`jmm7{Eq~*qw@Vm+KpQj zMuJ^O!^XR%EDt6s_K$0U#dZ5=K`9x%Ry*L98P4c@Y9wzi*fqBeF#JAYzE25!{Ht4ICH-<*4XrCI;t; zsx^yKlU4HHmM(K*38qV-bx^Cyxo5Ahst~MoJlnCX>g6EXMLgeWe`byV)ObRf>>{DI zv+s}Z>hfseSu!BLBX~3fV*rmLQs+_*v2ohvwRzLg7q7>d4%i)_(;RjC;OR&M<_G5& za=qQ4NjYBpEz(ik6xX60D8Xz6QQZ+yhsIkKP0&sCxi5s3SngCQEs*!CksJz+O%K(M zqHoEakkMIMM(=2fvSAnsh+-w)VAhA-RfD@5uOb~~j15buY;3I4eI>-~1)vl=ftV_C z2HhK}!QnJ**v(LY*yaH64$jsxkxm?%1%x{s6MBF!)Vvc| zy)?r)m5uXb#3|)2U+m7V2yq*8nXp8)Dd`Jmt^X;6nJa|#5_ zHImJh%{EFgc86O6-UuI=w%@qH9ZO%^+f8nNkZFR=o8pT-VdcS7Eny-#wOHfzH`{xCpuV^yV|IKt8cxjtP~<-stQINxmta}nHG3iR zSOkpWA~6Ka^&W?WSm#OV-rcArTqUSnbA{gi)E#W{8l}p+*%$k93o?6+7oyqS$5XR^ zn9bB`z6Y&6emIuz|HbzZQic5Uh;&;Lv2z!v!@+CsLd+^b91|)-twK8vtd{ZX;A3>?IG4RWoiIP}M<3?Nx97NGMo*h~5}q?_Sz21!)S)_gY4 zf!9?ppd9Lsoze#CMuaTHwG6{D$cy#Ge$*Tc04;{CT;)~E`R8*{g7$bCae|jP?E?Zk z>AIf!57sqc6QgpnEzh$uqz%(Gb41suPaa^CLR|al%~9Rh%AVwFkqoK2&j$UNOZE6g zotyx{QIsm!PJ*Yzi`>Jwl%kN$w(3>R%J|fzUvvdVIXZ=C0AxN3iCZ?1E}khf7?tep3|L{3i{~mmE4TpV zC#1^@)$1cWOMA*0)G7yZI)3RW?H{(MMxj;efR8nO=#F<{cn7RV$=Jv0z?$MWMN zR|*i7U+@J5#WbcOWoX4R5SFo0Fy3(#ZIBq)j_5*((ErfW_V;HC$m)$$OkRvdIR?a! zX1B|g_X*5}T3xr|^`@Vp%^=faU!I0WOSCpzK&FLr&jmZ%Z_LIPXa6t`J5vYz_ z_iM+}@qPyG1Fge~K@HG2;xAZ96Llq_HsLliq>37|CdDHX7BRGaaT+EoN5r{2^QFFV zpkiqC<7hXS7fyHY1TtD2w)(?9e_`euzZW(#gydg?&xQ!LW>nyxym~FybgphsRMTK604LPh7 zJZr~FfodY_u^wn!@Fx-v65$-!vrCBFmN#=Pr zo>`JqVIoxEG$z9=5c0<)fW2xiV?Oeb*3qfxpyW0ZZ>*z{ysT{)HlVjHzuxwpNLj1t_5K8i0QL|V77RWB3|`!k z!))+&7OQIxQq55h&(ml{3@q21SZKdA697YtG!^}#)-G9UK{X}b{qZB!Y;%>pUCQ5LPs6(K>);L7ftHibz!NWqzyX>U%kOE0g0L z8CDxRCYm=qo{afqsSR6D{hyn69r(=hn?ZYN1_2W5a229FZP>_CZ}N-B%@=TC%Zieu z9>)t_#i)slQi3SP9yC@EBz2b%eGGZoxl|2RJCd87vG`;Hn{IzSu>LHbfZv zO8c9)zk+%7J5v0dehSc!LQ;fmzyq`i^e!wN$~KH1h8niZ~^D&ahMdeI>(6ZGLbnBn$l7SbMKhp z1$(5q`=#1T0f?5gL4WGJt8g#QaQ*_9a)$F2_}HA`oX1OjhVuYi2F`H4gU|ID&bhqQ zXE@)06wDdUH{mjHhV!ksk%My=@?2k>;p~@okvnjRGtSJ_KOEu={wU-*&UfMq9NXL^ zuX$xtzO)J}n(!FsJV)AG9loG8X0<3@lZ7-Oii? zt-mQ5$j_VbtCC}brjpiJ)`gcxJBoJT9V{ec;o(MGN85Hzr7@eMrsD7sEUd)9z)?Jw z18WaD1y6i}4N$G_|CLXMW6*(xxttlpGENM~5|5mw!} zdzBmxEvH#~p(OqM+Sw<|kh~U3Ao!g7ynyoxoB-2f?|PB8(Z%S3TPe}Fz@;)H8ug*Y zG(*glRAj_NsP}Lp9;LgQDHmGS1ca8Om`lJo4rdUCu8x?r5c8T5`W)@IK;EQEZN41Os`;4%oNjr%+^D|aNWGI0!xFNj{?dEfK#%yV(KE1- zUVab}LY`+}lZEK@5#B0j>R`;q3n8x}ykFs=K70FABqRk`2$31#ohh+*s~eFq8}wQU zH>#^7tXBO=5Np|ziyF0@mk2*G``4q@YP{eMkm>#LDb!IZ|2!kc`Bis{60CZ;YA+vR zM`c-+&Gl9?Dz8%?!OcQL`lvHIr8hB^> z>{@u|tV=*OjDqaH1>Sk?b!?e%ymQJE+UBkk?;Nmg>Yv0rpO`6l=YsE0wJ=d2$nF`|~O(}|Rth_rW0aE*~9mO&#zSCfmR)G*!|wE}4w zo0-uu`|ig@R@yHNo(UkKj$OcM@v)*!3e?}>N*63O@T_#UFI2HqKktVNqw zK+FX3&h*>r!8?;PQRMpZ&UZK9bpv>3yH%O%f(3;bEUGwNex5DMmn;Rjz&o`&q}#5<>^QJ^%4cXnVI8pby*x{Ox`q;u(rbyiGI(b`R*zE+ z|0%pP7bDg~!aJYb|01N#*S$1RmO94}1-x^sK-PkHW)5J%NVLOGw2G3en)DjHb7VCb z8aBh9z&kT5kpx9Sn7le2JR(@!+lz0k8}ID9f*rz-cg`OncxQy9v0JsoN3wGpc&5bLt@?h!_&#;9 zgd0^i39Hq)2oD?Y%=}SH+k?Y9D?ex71roPFE@}X7{|k8MPq$DS3EtWF8eVQxJ62Nf z{2$<*_aI|7=p7PnR5wakt*#|OtmQ#+QKOcLcP`w>R;%%Xc;{Q)@hLQTDgS7^^Tw-> zG~Stp+>adIIlgaAyz}8C4la%u-Z}AnZ3fkeceeSie!TN2p`#wWvkOxV$2-TsE$Csa z?_2{AuTH%4tsmIRe}H$s15?nli7KrbUIXv!1aPy)`p(iWpc+ug-vjTwIFl_Cj(5KF zfVR2o#5;$7^JnqSSGg+#>y?(GYGH&f2eeod?`%X9wn4nJ^ixWVsMR0pL`s>6v3E7z zx&0^^tXU=(0+`zURlVywW9K6+lLn*bs9(lU(|tS?8FUmMGTwRfhWhc&ulG{YfF0__ zJ68ecX%O%1(4_&q^XX5RI5~5YxdJvYw-1}V4!m>o_o#rHF9h#=y_$oaO~|VDd*Ypq zxA6Ukd|w;yj78RhcYXyi6T~|QTvZR=nb8$Rt{?BL_!zGnz&jK6zIZt6J1-r4ka*{W zak_WJu@l$AJ2O@h*r^BaOpup<9p34dMZ9(4os&P3Hc&P~B8%f0g6qQZ&MK^ls|)XZ z?aF%a&IP+!*9W=2bIYLm@Xql_hEzRsjkdTvW__eyjV9|mr&W?*KSB`1JNIBxnhJc@ zhq_=h7#h5@gCxtr_b6oT$2+I=stfO&nn+!8sCeh%@;|^kry~MnN+=k0+G2Th@ObBr zFTqltsI5$@_Mp63WB;K1&*Gi2?;(SF@y?C+Peb-~;+-?kqCja7?>w7jXc+H2DS<6e zKi+9EmjK>5g>LfM&othde4pT*-S_=GZLSmZSMu$2ZeX`M?~;b zDBgLIeiX`LW%|*d#XIK!?*Vc73y(D3dGzQY-uYH1GSA->?|gf8y?AHG9TZ51f_J_- zDS&r&NoAq^(kvA3Ou&pedwG^xTh0t9y}mG42JbBXMd$pd@Xi8^SPu#BoUxHAe&gAi zC`$SAdt1-ox^&vU?jSupJ){;SM}hXqdAN7=kU&9ZzBncf-t=E0`Qn0?|g@r zou>}G^WG&W4%UB`GKE@r=gi-y9x~oJHcN(jf3&_c5*hyi-g(hEh!Vs*KP$s+U3lla zctPT0)EkU)Wbn=nm@)Ya@XqEKO&==W*%aH$4{?3xFa0#$`OIQkJ=N+_grRun53l_{ zyi@SbdB49%T(jjUgrRsR?}>Nbi?BA{Ifbze-Z_%j4d9)l@xCs+GgqG1#ygo_U3h1< zJP+cX{(C>(>A&~mopS%bf_KJzZ1Bzn@FsXj>pL?yYk+fluHc>Cb^7+Q0d?b@gXgnD z`0>t<`wQL~BWdhb?eUSpJI|GPyVc-@jPFyINw`tzZWUJM9LA-O(>G(7p@4Uw;O7IA-@9dAdRLZ*c3t%?5(($~ zXxaKu!OqJpceG{0uK`y{17gDF1-NnH-h*g-gEf{b97S=mF4)Db%@*Ff5Ydi-=xc$) zIk4rND<&#ZW3Z_JYpkuxRQ)0Ln8RxL_k;UU@H}pf#(s0zkbsRaQ6pO6apVpTOtR$6 zz<`&67}d?GK?cG8QpHczIvlJBw~?yjPCu*=az;mzS^r(h(G&NU!P{fYc|fButMM~{ za~~GW(0WX%)(0A*lsrGA)Wnae=yh78d}G-g=kK+d!p~;P88Wy&?4R4EW>V|Hn$1BmVE`SdOO}bso97=_DRj1g^L<- zTY;nKbu8h&816Qk^(tx6-1}T`|4D{jhLj3_*pp^2D_j)a&)v3pKYKvSi1d`zqc*!d zxUwQwR86J^UonsTX==d10e)(*JTm?_5n>LaxZKJbu^j9Su6+T2xXdlPY8nuT9JQb~ zp^hRH2u1SyKeg;Hu{QY^Y}d9tE#FPYUM{S^Tm;^mfmoIehYei23V$uwB9B%H+xkNG zW#U68RW0hx6o#@J;01gSVmmci2R?>YTT}>J)z&4N z0pSY>2h#X7L&a%^y1iU8U}}&7WlM3jX27h#p)%Pzeou*C6kh<+LO;RrH|y|;C;3=K z39(MCCS8D^{rfmKs=FaANf{_L-}`sGN@Bv_>z7Q`4mZ(JqfXlePj^^`e+ia$T5;NGJsek40Ln4N79VD|i2FghC!z79h6!1-dgd_tBe_d15tiEf$q%o<@b?t|_% z%zkIH**Szix?FsC{IT{T6#QWMV>?(O$RC&UXce&Mvd@o-0S`Hb<2y2o`=T|GrOap2g?r|z;r==xRZUz)AA&L&WFwqlbK$_{IH<3 zetzh==+OAVA@LjFho67Y{P3ETGn60tT+Wg=%n!3lL-^s{?SGCRDwb(}h`Cnt!&x=L zF1Qcn2klS)927uK8yvrR4uzs$X+Xd}8($#OC|qw)LSY}~9BK_ua)iJ9qD0_mAFhW! z_^ty73lyVozeBSQ8^rGUszZbYFR{5Y*9BRhO`8VBO%go@$xd`^nS2L68YTJ(@YSS! z>w1!Be9y7NZkK0H5;}_NiJa(u7S;_Cz%j=3o=lCQV!7wklU$7W$eYd=jzL_fXJNG1 zQfpNn_)hE-pIUtb;)1}|nn2YIAt+Ii3_Z!0>on9#5Tw1m=cu2*WwEdieVzK|Z59hk z3>rR1J@;h?m?9cxmg3p#xWS2)$MT_E-ymbsh>ga}H(_iOwSyFSlJ{*OsbZnUs1eKy zDknxtV^RNYRnvEpfJ#uUP7{)=)z!$HL=0r;Nv@DI)~VA&NWBoGA_`5t7CAum z0?$!xa8ZB*cXY)|8-T_Of;2u`f=r`W+em*D(ns~{v=*}lLn?ecnucHCslPg)Q-|C{ZH zFV#%$IMz+@BG#*sx^HHY%5~~H^-QMC`9Y8>0ahUefe~Sr+BjocE6D~I4%ksWVo>K#Lm-}+>`R4r%=04tpCxwM&-a&F- z=FKwC_Yc()OFu97ZpMAkpz)vo0Vx$`@4nLY@buTGa4~q$`s*K5b^&|$QL>?@<^Br} z+n!l*Aya73-aRn4e*HCW-l6HQOC^2-`s=jswEpsvt}uJ|pDsRWSbuGs8=}8f?wokS~@lCwujvho>cS0K!T}m#SmTqxN@WBP2I6 zW#58ZaVRGe2Fhwbm6h(h+uvYMht!$#8HK#K8p|@HDj6lYF`C&(8NVCOo+Rdi@k||~ z8D#PLw?^8`Nu{(YWKIyDgD8Zvpu0rW07E)KtFE>ZT~yZ~L=910bHVHw3~@x+;TmQH z`L9V+{Sb1kT5sNalASsYbsP(c_Jb_MQ9Ow#l2KkABtG=kPoGj8ai|Avz*1;sbv6nj zdTnO`3zFP4ybKHZx~?(|#BN_i&-z9A6 znm`$rBWg$)ItR<}_%L0DPyeOMP=SzTSipNv^6xqg^)M0$mSH*+l_|sH5+5`U`=rix zctx_p`{&al%bI+!@k!mjqLmxGq5)okfb{I;37npm-=7nz`A`8&e-*yqkM6)|2Xlan^qKf_ zJf>a7Wot})Jwo-#`97Z@J=>+`HWE;ogcp`af8ewi0s5fYky7!g6 zzBwp|jMRtzC?5qd9Hf)|bqKf7APLZ*MFRn*CP?@97Sf(vW)~gwE80noWIU-8VTCpTQmI6a+ zOcP5o5=qn8jl=jt4h0f(@kL(cCAVsI8Si1_vU7$T zx$NL!^6Axilut{%$pa`i%Dr7pn1qne2blXf6Sgzqr6%knVVQTV3741zkzbmt^wx^IueQDrXs2OQD07Vg3FVogJ>LA&~5$IV_JfrFP%fHQhm;(9AZ zjXB1e>x1KE{P)56^4QmkOGm$qq}#y_9U{(f*}tb2?aCj)q;T#9B%?T&t@g#9FG=Mk zbY5A^V~RGp9M(SgU`l|W!6E3m!_tPP4Krsj6@Yw5fDLf{Y~jRJWH;^(T(D60-)=9h zny+#Eac~O*&r7b9(e3imAWV^Wj=HQZmr`MT>2hbImPW>*q%QX;L~V{zqO8OXqFGum z*dwHp@ih+U@!p7sBL6#Y@j_Qa=J>oT4C(_dMcRamK6endAf7Fzm@1q~8B!udc=Sa(xc)f_Ax|AAfy+j;9gCMZl z;92>M3nWrW0h~0V?9gfoaa~AsG@nw_zZ9P<5ilv^NPIRv8#h$PlG#OQM!NmsLrRP| z5yIf;@6{=*Wf3dHrBcspNh|5pp6!FCs4y{@Xhnve-)mz*8M0_WCG2=NgiDx#oQy0+~LxU%ICt%Q12imRox+p-jBA(;gu)(ffc$nyMk&auB$#d507Dr zY(rL2NJ$-UUTZTdd5C`94!LknQ0+TQf6Z)>Xu5vQu>t2W)i>sKu6})nemxR*_!pt( zo7eOriLYDh*OPFE>?723@)~xQex12cNSnh?@iXi>NZlpT2xQvdf%NrwPrpx;G3;NX zX_OW4fd`B5#TsJEllw8ZHWDvj+|kpvT&5nFO=Uq0ETIvMmi2dY=rw;&XWOQNjzA}z z3y~7&<)#NC)0H{QyX9~3kR6=b zvOxb9+naOy6n}C2%wC&mZ&$@|NsKgs_^Ex*G}zl&UiFoUlf)E4B*HF;Pnk=N{ZzM; zM>3ugzlsr2gQNHm1c*|8ZI^T6OPHj`t;ES5T@pNQe<3y?5r z09VsyJt3}=~lbNugcw#((}WSFQp3fbsqEGz)C zV&;Bj{;yoT!7(r1PSVfY>E~#y5_ztD2(o3T$^8^@pGuPS;3l>l7#lRI+QL%keTI#C z8150p1QTMPI!CbHb?ONqkBt4WAduR_qiO_R1RJTrPqTd}AhXPsrF!$qll;=hy4H_F z?_w0)I}(@BQkOZI4v2>)0{%0fJw|g$_#ooU)0qvToV-#j^?^tev;caLLBxE#X3}w@ zDWG0{R(rqYBN5+~>VB6defCrAyX({&3_Zyw8p_6)&oyNamjx&*LOjX#l6onde43%G zWQ9<+@}n@yPOeSaHKCM!e~G5-1VdRHP1$2eg@pcq)Pr@lb8&#OO-~xi-g>{jI@?V_ zz&hKyTqx@uPFaabI;76ZzX_=`%TQK;oJh)OhMwf33}wzR%7z3f%koopT)mVH^HVlJ zQ}+IcVRg0@uWPjCnW2=;?5peS=O;8}YZ-cy#~~FG`lk6@x8|ZnfjWB^@uWh=%&V`? z${^Py%Dm5{&dv*`Y(Q|y93i8A3+sk8ILDT}L3*|1Q`)?1pgc80PTP1)~Db)9VxE$5dSzbp(;rXDkty*a18 zI@@`&p=|RfLfI^&i8_NG6{#_SghNEI2Z+^xvFdJBx+DOyQE|lj;NfYwB(78cdE~KG5XyE8ZC$<)^A&zN$ zywiu5>r~Yg7T%|RkZ_~=1|jIGGIYJcF9M==!GZu?@9216As$6}k)#Z|KKfWH=wo>v zpzDU3biMR-C|$jSbe(AEy4=upiJ_~P&d)Y;t4>Ueh=x;FpA)Yo~Mu94w% zeS+6qmL--R>#Ic=T{GotsjtG`y1w3jRM*!#5^hwlBV+~Tsx;Gt?bH+UOdGU8=;k6o z%N+g%2e@S;CXN3f585BUI7jg(zrb28ra@DS)m%iB28SB9n<=zu| zX;>!Hgw0LZ$b>yjxZ`3SZ<7fxFwfVR`;{h4G|v~C`*|juW}Y7-VVQS^xqsTcf5e0j zn(*El;V5&zPr@?qdFJ~IO?a7kzQDx0*o56om~P_zY~uOpO_Aqi-qX$Vf1CK7O?;gX z;4BmGEfen@6K_qLu9sWQ`&{$gVtHQXb(?tpbgwb-t}~%O-53+^L6e>iE4>p;db3S< zLb}dxm${EM_wSkeuTA)!3AajE=ADi71M8c3F!p~7PdjJ&5yR7doW`X7ay+e3B!Qg2 zDW0ZIs2@-JSl%55o)#0q64%1h?$>$!Nj$BKq;lwZ+KqCgNF8|EPsg(g{sKI$1Ok6O@mRaMW0)+5}P+pjd zKNw~4nm>-XI4{8;VGAHSMS7kEVFV;KBXIQ0Uyj=S@2MDS6Pifw4_Yj{NxT|@d|hT% zMvh0N^J~iuV8lHLaO~)Ws#xM1IFclYgAaiTg{`WjG-@p@psiP#CU$X>-rP!NamUn{Opb6YSWKc^eKUc_4eG;_+Y-K^J zRL;Y^MLn;_9Th0x7Dlxza2naO07$U8nN`h!{3O(Z~(dMW}FCYk2$;}r0!jBo0B6a?l zY$_UqC=TD{s5a4YAUt^h(iF6(OihyH4oGstB{}pT%Z+Z)aKCA&%Pjd9T1lH_#13Cs)y`_0ApM)#Yp z{F=q6r{9d!{bs~tVg2Sxx#d`cohC;Oeu@n#?s}3ePpmrz^g;@ed-)@*h9isCQBsk8 zfxMO6qan9|H*0pH9MyUen8z*`jsW}fArp!r@8ZYncIKK~LA=;P_b<8xN172Iu}Dx? zIc9)c6sE+_U|#E(+{J_BPG@b| zkI8NEGVYXMW`wl4zXo7dV?Id||N1QJ488$lOIXQxhzO7fB6#$}C@A?ht)7u*=B%X9lsQBg%a`nZ z8ULdAnj&Qth*i*ADKSK*xCY|7C2nd`_q0WR{3v(f?=s4?&<)qkL+9{3$*ydwkhh;RRn$)cGxa3DfyZdzINa%~lIfiU z^@fQit>H<2NJp~}4LPGm5lv(0>H);0uqNt=lyVV;iBV+Z)YJL#RK9o~03~r13R~)eZ-*IH1 z){SYDWI1ZbF7zVc{Y+OM^81^k=ndge;^IaxW z<8I0Vf%@+>Q2&MM=+*}5FPE8GE&fmzsO^QO zGtPNMcg9r14Y}-$)X#oH0-W_L#VR|YI~y06lZfGatdABwja5}oa<)#5VUC`tissB1 z9&;oS_?(c)EiY}w!s2X03S~2)f-YEqw;^*XgK*0q7=cF{X7O8evRs6eyV(fU4J5}C z)#_564~b#9AavU?x~JvrmRiOuR_e%_3SklC3e%p1$TH7@^x0dP{xqf^h4j&bNUX*A z?`%x0uqGj24seM;G7r^dMV6_5J;0_g`l9w+i3rLGCJt#wx6-mTJ1ASV2=pW`n5`SG z3yX)sM1v-om|_U9>eD!qC zPD7;ScB^BsB!y<_(F~Chn7UdWD8P*;`Fm5%pMf8P{bK`llJpPt72}cHpL-F{%!UE> z1}-2BSlN322N+GFUVJ4kJjqie?cL}QNoWgoi?Do6V;Xxaj4|CES_)s?rYRynb>kN( ztG1)DB;m_{y%?dUw9qZz9h6b4oiAavI!i(P(an~~GWn=bB_0%PLat}GqJ7}g6?I6-XpINEUM`R@j8 zr_camV9><-J z!HMBG%zFFkeH<7l#f#FuQ5N2dNUyL}pM*sc5> z!){zg^#%X%80-yiV&dO%%U40e-I3T9=X*7)#el^NgaK71a;Fx^5E5Z(p?|lHuUCWB zKeBe&-Q(#KhTTI(=Q--CXQqM++rm`?5Gq(i36}d@H6JB#x!(}2m|48VQM46ui#7z_ zjts-6GQ#kw@=SL=EWBYbRk1|w*Gxz6jhV$i<{cvoamn?**wqjE$Ju^-YFqvIR7My+ zl?)<+$#FKz7yF=2?o^m5b0BaqY~O38EE?0^ngc^|9R6e2zR2;=M|h`)+lwfcq$D!l zF3&qr-$8C0pbD~`K-JD-RtyH$2C4dq_k~TcNEGS>(gsFp%5q2Xf3U_+Yj4EG2xEoK zUTxHHTXKRiI8>$F6v+@MFdyV7{+KQ2RYb4>~z6P4&e_ zKeQgQtjv4%nVbhH^Nx|Q()*~nKh}f`Ot{j7Uz>2X3EP~dzdzoDd(87P6JB7#R_6Wp z=DxYPA85i76M9V8zLQS(A`?Dfo)?*Lf(i9^mEMa@ygN)-Xx@)A_btr*^X9(Hgm0P9 zYr?NhIPMIc|I>!v?&khvbKlt9w=wtsFyFO0hu>9tZ#i3k|GEh~nRt~Z-T@QZ5>|S5 zntNZO{%)}O?phN@o9}m-d_ObKzc%q_neZ|^|38_ZdU%3Jk|VL+J6)3c%jc)A`f%0kU=-5sD-zO<-lJ*`2gtxbbV;t1xEQphdA_1Nw?f z**w$Ykg4SX&Wq@hpdf!(LR+rhWK<08314tcrHn{9bYs%0KNKwz&HDe?dl&F1s-%B7 zfdrzWJt)zrs8O>k@tP=6Vn7o}pa&-yQ4|+-QIX(nL6{L#l)y|Nz2hj}U0nq4_se=g zM1h1r0$x#3P*H*kbTeS^0t5u+`~B*4&-6^dW&htk|L=Xi=jGA#be~(DI(6#QsZ&)4 zm7+zqGzs|uh~p(>uHL;H@2kyeV=2X8vA~VamBtoxJl+A3IhxN%Q8!hCS*nUlKQ8(w z&2z*v$Sb7*u;&5v4QD>3Xq_E`v`;`BEe$*-dQs;0Vl1VDo6jTobnb{0z55I{cB-?e z-~2-I0K@1MG9!=EPp5luUNA?$Ej_&X-gpmZpq#shyLRC{ocSdbPUt8@;>b~t7h9uI z#8KoX^bq1fmKo$P^!YYEBSl4khp6$X3T1U3{;1PuU*x^K*I2$1xuq}N?pLly+(x1t zJ)#)r0NwoykStj)cpkzqHuundpdoZtOp~HA1}=xF6W`<@wVJ<-Awa9m4SWJYQf#&P zuH*+XL0tIc5eFc15pqc@;rsRm$fv-CWLff@05UEakgHMwdHf9p$h9^gS1LdT+kl+z z0;GQekP-)w4toVecK|t80n#NIkgqsqZfC3<8<3XK3L`7|j1;YXN)6HTl3!tD-aq0P zc^$c=ySLoCH;gRUX=7yevx1RlrN;QsRHXuP_v;EHm)L-uuK;=21|-)7NY4Zy`3@kP z@7Wt5XF7nKq5!!#8IWcqCt;+M4amoTQy6LDGg36b2IL{huQ2lB!*PtvLN38b^7|C-0IZ^@g zrqmc8nzxXggpuD~Q-Hj8qr%9Wd`60%wgIV?{0bxgcqopMyO2vTGGwoSJi5ci$h}Vs zMy^f<!v^Fe1;~XqAUjc=tOc09o&+G74j>=iy*G>;>;Q6j1-goxE8~E0 z&uBi$WUK|xfdR>;j^34ImgGq0H8xY3af3q71AIn`zIsB9(?}Oow>}t0&uHWl^qjd@ zs3vZ=(KGHTLC-x>V|<(rl!{&4&G)Jrr*1YNT@)ZU+kkwF>bTe~AsdYpRpT=@)Ae{m zAHg=2QtgnnRvs0PBz`VTe2bMf!A1tYCNn%(y>FthI$nLjG>Q96@TTh z;T+_FYs*l{habb89~0yUvi0DJmvi3rR`@Y2gK_^BFHVuX5F_IbE|<6Mj+J3|maEO4 zn1=%<`nQO;i}<+g#=RkWjP?P@(!(2E)Wci6$bXp+wQx~_^B$JJt<~n+TM`Z~Jkot& zVP)|a9C~Ffth*>LPpch`U7WM&^KwD;LWVNCZrv?sLFIXQRwZ&My*TRC%?Gc;@Z(ph zc_)8il?m|~Df)euB8VY?n;ATyRrvE_akCfu%aBh<;`j&``}@T_gvW|Jtt0rQ2*0X~Ag-mc2F^G~1*hQuLrSiGjt*?{~p95aZLK60Q3o_y5V zDmQakwmD?Es1I5jIWxg1w{f8GkDg~dJD?iWgDh0Jkvz<$zzY=m_x`3Dc}?=HMzbyO zJZJq2d9*PHpZT(kiew5{{D)>1vR)Bu=-h^LWV`B8R{b|t4V5P}5fCJKCW8@&sHfn7 zr;BAiD)-llIOe&QqoBNTc@fXBIz+tr0Aqa)rdVQO*@DB zZwg!(D6@pRa18L1s3?9U)5H}QAZEpbz@et5Fl1rP4iRWxDD^?%t7C9^rUx|e!GA=_ zboCSHRempg&a%^aFDPh#90@oujH~#J6ul*75#LFO*qNsbu!`C6yor3$ z*lot%;A!~PhG))W0?*iFcm}g-Hz@Z!uRyuf2IT?;${-t*nE)l<0cCR?SO#WC zFA^I+JDM80SV^4$nV|? z{6eMz3_fxVG1F?^izn&e?|epz>QzZ|7>mU@-;G%KY>ZwdWM&;~p48h$iM*F@ZWe!%r5Tm7p$F+T&wOL_Q7jRn5Qi2bR_EOzg zV{@kclDctZq8lEk8!M*n4WmanJ?QpN!RP~0XB?xi$e6k?y6G8((YLN9D3~97Mv6w+ zpxnXy@xh&rCCnb&TI3T<`SuFRz38oEF&-6Aj!p*U_*77?vOzh~2IV*f%CC=G>o%QT*=kwXma{Jj^o{d(S{R|wg;L5{0%k{GJb+eyc?C05b z`b7KvH?+%s71`x~wBg9K^W9yn_N%kTPOJ%tqw=2e9LV(PPE_qr#x@9-}z+u2krN{HXIH<-n7$~+v%I- zz0rC~+FNYp9IfzivGk`==!9^diSO=byl4jN>l)`^4b%=#b=cv&6TuA_*2t$ zMreT;)tLs6!?UKRrYAKf_FnG%x&LAQe_#Ln9LQYOxa%HEn=bIH_p4LhFHCvwm-2o_ z%KP#0ci?M{=Ua7IT?AbT)?;7>Y_2wM#C;Ipki)FsMx*V7yco&+*B9Q|qJbilLyPptvn6f7SDD?21*I4$A``D}_UZA3%ai?ad zm!7195OlvU^5!iqZI+8ytNEfEyDrpmjR+xOZ!t&0XUd%c#kiHTp9k@ziZ}a*>P7&2 zkxFD6f;<=UDSURcZj|W8dALNNUlzKa*3_9{l-PB9|BAv#N@41vTxE@t=m(tt#ozkR zBW}CtoUJ`#*;8%+CU|E7;*PLHQ^sLz|234fCC1O#w0e?kg)n;stKncoA@R-0dq0Jp3CCeSiW@sHH?!X7%|DKL5-kfoOl z@c56#G6z-*##7Y>cAJGR=P1;pmKx$ge+NzRpXLjnfU~E{{I%Fr>kT_P>c#JCq5ZH2 z^A)#fp*xv3l5sFCFs3Zv3-ndBpWsK~CdP#7M4d|WKqKR{fsyjlvXnXp${X0SOgbJW zBwFnwEM!~{w$m{Q45c5?@)-sT?%Nw;*~@Nf0VcpR**Q_=gx`-BM^x(b?Ks7Fp-4A9e}FMv?*f1dRDQGHvgyrSp-Qg#u7dm)}M=+yWe zRHBU^NGb)n1P-^!fRu7MDCHQDuSp*ysi;`NuZv}$#>pk*x+!`$dN8)zt1?0Ypx&22F`IePAofm4Z04e zl3<%Q{&f6J_jd=026_{=vZnH*((L)pKyvW^hO0tGUR{MJig-6-R~(rD3#9o02h0lW zAi?`UMnrHd`~Fcau;95dE&3{O+Khu!=nDWa3M={*$ii2E3xBRnd0I$)mrg+^|8w2A z9Yz1_O<%4Df7ixW;%|E37zY5t2f4sS6E&nZM#%foY5XuW?$AYYUgn!Hn5+VHaZ2b`aRGG!X`wDPUJaoq|am=jq()yo*gs!+K5+34VWjK9}+2h zmmX~PT&Rcp17hh@eAX$v4FG?-{uYD^2np3`8YWbV?sr(qH{Bm3*k=*}% zDqrA~zdyD7&u#nE|I3~Jr!%{7w5*evDIo`GVB`?^F2#r~Lh?<$vC` zPyN5l>3>pruJ7^XMdY7(Jks(-Ue3b9Tf8f9jKwgMiTQm6#vc<1=89Tq202RjuZY0- zLtrg3YoRc{Kw_)JAqFd`Hg()1BeDj{z*`DXcwt7#HuPA88X>lESoC6BZoo~bz^|$w z;Wgpt=__6vu#m}rMpC_-0Y;+UgOlnFJM|u;WHH5?g&)JGHc>yWZyKGe7e`h7pYxPN z^&8A8sromD0j98E#v8j+eC!CeCpxp*9F(Z<9^%4Vyh`fRLX)uCOPxe+6g~X!-pwM7O}L z#{?@nz&^@a3j5ar0YN26LKh6wulj`ludTmTwqjtIE*+^#k zmVKf3G}HioTWyYbPkzkr1s_C?;q{!yKRUe7-QP>w z*LUnb)tBqk7fPyczxMTg`pdrbciAO&e|sj?_YOY0^*^AO$@^4ajZ@!`@3{MWZTtEP z_o==Dr@s4>>N~uBeQSQ+m;Sa~Z1=Z3slKK7?Cx)}JuoA4aNDbep1=lA0>{+H>en zU4|e!njCpt(Z|dlqzQ#u1z_WU*kPO51`9f0u7(uur^b48Qq}EI2L+E@f*T1pQtnR zNN?J2nbfmv^X0?*sCKZePVvG(*aH(0 zm`|nPIX=jh`CHC|VkbBHjDj22n8ub43SX4fhHFfqn7e;ma(+l(8|NUtexeCPupo-jhgUHd)t?_dn7;Y zwyN+6sv+aEG9%y#`#n1W$xf-G3U$QwPtcY%Kv}dzZ&-r?`^Xx=Nge5P%|qNnKZfhR zL_en+TSUc&T{Lx=zVJv;dxxZr`Si6F_gb z<9FVvtTbiN8=b;<4=3OIko7CDBDAr{nvCLd6|)TI zx_18iM^4Ax#xKeK`NIbZa`a^?2jr@;zL*}M!D{pFWluo+=IA?cKk;TLpX0X{EfS%^v`I?EW8$8}1$$AsqGH6SfL zeTQ?OF)uh~J}gqZIS|?r!XC7K8LAr>=BRV6j#zLr3@}|habK*dg)SF3`$01+fw!S-z`8L83Td#EnEzl} ziwU6O2e)YuCpuPeAENk!)5FYAvF_z|0&r(vf?TQ+-ep20^Adt1Cu<1G0yk-NJ$@Lf z)&5Qx+TF6_@OhwJ?mhm*7wMZ`hP#ioU=<3u?myP{aR0HEbpNq`KiqxXTkBWX8_Qp- zRfI|4&Oc7qZ*U>*S6+krm5)P}1E9L|45-UE#Ofvu)Sv`VcFFT8JASIEiR)mluJpl%Kxax(Q#VLxuQpWw^Xgo6jgeyA3sIQ!d5#;JUWL zSeellYz8yyk~YZb(xG_U*q^LENRxH1!uSa0fMvn8Z577CGN{_Y^SYjIbnjMCY>wS{ zex!SbH+_kP$Tc=!<~`6cbr1A*skGaE_C3(6qy@C!7{zH$WAttQc~5o@ zvp;+S_ZdME`09PI6udZblrVYYO5iccRb~l3@T7*3nlVK{!T^VpwZOb|1EC|aV!8~M zQ-gIv_9S_HFHmN=_|=2IM(GiTx{4HyL#yCOoYc7tKhFb?v?k=o*K!Dw>2|%tX^#3D z(kP^jV(gk(eBt!^SlU4>3`wQ&FN{M~RZqCjqsLMFZ5`jYL_gNq24C@yidhhI2$6J}d>nlJr(>s3IYhhfF~N^fPws)!zBRcm4m6A1q*UBs19YkNLsFQp+Cr!6a4wKD+$h_`y&qpF&Dn?f>!v z`}{9yp8wzG2aok~_`&`5uN?F5{3ZOL#{QON-YmaT29f;W+sbzQVAi5!eh|Og3g_GQ z#omw^lY|4Vs}K%gqJ(y_v|q*r6aFp?0GN)m`ODy#@;UZxOwPs1e?L4EM1U}l`d-ax ztkw4Qfn1q+-%t(e#4RmaXdqV}EWh$0X)VS}t}X9F4&egNC0Ux|0^cKf9j4c*l1_A~ zFe}Wi?MM z_UTNHWcEo@v)7(|707Cf{`PYGRbN&^tQN0U^;fBNx?U1gyhYbK-CwQKbI_`LosRQ& zwZ202wemk+U+;9SuQmLgSYNT9YWa`W*VMZ9y1w@AiRD$6O`LJ`?&M2vi9Ral(n<3U zi{lwywQkBfJTVjE^hygMzOeGg%<)CeG`Z3Km1SOS|H?5hm0u~CYX6s? zabJ!C_}L8mSC)C7{VT^b2{e)bX*BAF{;uo72?xjBSZ`#+DLeRHbfXP@KQo}cw8-!nh^eeQpopPjKH z!O!|naroKoEB`0+voZgkpZTo4@Uxm?6474y*_BL=WWEL4oj>Gf#Zt{4_*t$bq|Czq z3tamZXd1Cu zsf`5=wr<_~LADVPhtxra_!SjU6oOyTlcgny1<;dg$I;DxyN^obB4uFQ$CSH$xo`yUgv@4 zcvKHxDKE}mdszn|Nu9gt!CmQE?U#`5q1NtPR!1=pN4q@1!(F+Ha+@+vgQeEDTId^Y z0K(U2VbS1i$0ZP-ak{rlyLldt(zUvY(pW2whqSs<{H`_^LgD8%GH5C-=Yi^OrdakP zeOhqZ-a9Y)kerLsw4AWn9hpKyGMuU2(bsEyyQC+oprPqbv z%cu*6jqdJPc5!bT!4I;1#(q2zyDDoU{v`BqVu?Z9hv*WKU%lbKK}&cv-4aez;}Spb z2Z_7S`)Bq9HSxTk*f+JWjrxnD96UQ$`uCqH#Fc_ zxKle+m!hYzCqS*!qR$CD)#l{sZj;+;^CrCG2(;fso+mzfkGzb&l`nUxm#gGuGG7o( z@9yE5^%43`*R(g3S}?15;^`5I-pu zy^`x#v>=y!PN;^RnIw>&YCbXIvC--cHs|6V+FWzUnWWPXNgox!?~O2g8Hc9eIwBQ$ zBi%du!g+b#;_tN3J7^@nU-im{3w|D;Hkn1?GZ-Xl?I)lT&QA~@3I$gBAg^RqcDx-| zTHoHDgiF(4`%wcxWlEi;8y~=)MQ`{$bGqmr))P)vWxmY4YU~5fC-m8#P)jkP5G?}% zBu~!sKPU#{beO>|FkcVj2Rf)Wf_I)yk~ldJhJHfDAY7F(utG8+DNt`+fV6o3^kiQ3 zK0!3VtL~&)&{zz_Ru=E}7sm4oJ`j1{!G|||7V+`L$FQieJAnd$m6hRZfal9Gr&f`W zL5y)a_9fHtib6~u*ldE1F9&*rkOOf-eg~<7Td#3rnm4@_==F&8je}1P0!q(})(gF= zyy>l$AJYss21OJVKEGB`;XFJOlMir_4284XCOI%}d3KIa*{k(idpx^q8ms!rfI2&Y^=sCgDj^jhbh?TF=?Pm z_`&Iu40NWme%9*n2kA1_rP?TY+d(9)xGqZgS8!ua(%KA;)R{g!vLM5eNxWQQRXK3f z6^4~qjgMs8au18k^|up3Io|~rpX`psvOZ(&|8Lj-CTIrhe=mFechiT`k7~20eFuc~ zhPLC|OgwS={2iW^;rYZP&?#6gZq-7U;WMUaTFr#>(r~v_GmI2C&f(ImCY7MXk^`D* zzDCKkrqY4X1-(Y`ggz-)scM~uu9JoWnPp8CmFeE_18k`>{IqImxfWUrr3{T8FmXQh zg7Jav@3B!P?5WY`*eHq3)#fJD*}Fp^gH7t;w}eQ-GuWhF{3~o9q)C_!iq6Xy?pUT; z^xEC5L^sROPo#K7`Z}=8?q?Z0>NTWIU-+a7yUns1T+X2fMG8;fSlRmBk;z zN}A#u)`T9)-UNEzUZNEKuXK7XDo=E}HmTFFpf2FpKq#y70f3>z_;4zoacn&zPY1We zy*Yw~gWC-@yyGZ>y2n~~P;dr@da5(jeH8>4>i(7KHc?e7qQbo;T4)*uKur5u5c?6H zK8TI!4If%_D@Y8j_O(fDa8^?X3@F?L=?i~A(-cKveMw<2sKGv3n|&V}t=R#zeU{V&mL7AWjCkYOw zDI5$S4&GKcxE!9G?-B{6H8Ege9@@b#8x82c8vhrG25^5ZR4?;)FKx-v%(i^E3xtOz zKAc(1VhKCwl8<*i$orjgmfX+#7O&B2hwY?n&54U#P=|Bk;tq*YR-;4os(l+BWDQt_N90i=0vITF zDnA>$SRfhg=@|B(yA(To4V z^_y~KnpS~JV4kHApoeEtB!JB7hd*hZtnkHk%9(cR3BiCUpZ-mxMkGoQ}H){a-EUg<7i3qLV4m!9K9jG`|7nB1PUG zp!hchWA5|ti)Rc2T5Vql6Ft|<-)|4t!?Zv&ek0E}`@Cd7QY6j+KMAEGnj@~kiX0_)Sm zFGzh;SReL*tPfi`9D8}mYTz)BB^6dJcX-lNeH3yxRbmGU`o)RPqUa+^9`h+it%o0z z`lhP-zV&~=RRa5#2JEE)Wu>vvXZ+6Xx9}#vP3tXzjhr#aw*(sm7_~?5v*lc0_-)ni z1yZjdkdiqDl!m`M1kzP7r1xU~L>A;fit`nhZK2x+WI&>Xo5APbH!68W8sT zeON+z4YSf%$1Z~v(9@aK9ZD=#7Pkd9p)TA=fK>-*obC(T^~_AbMG5uON@Kg%_`1^A z7H7bI*!z+3YzS;bFktZS=u5OOnyLErcA~tm_*egO(0nDeEYg+4oKqa<8~>XksUnf1 zDj)zu?kcvh3Y+!&kBX1CVjn;G-;O5&cjBE9`&NU0vHZb(zqxu z*E~(Q95Z8(s>p_Qva?bjh@h7b47tFEZ{TjoY~$HQ|Kh`sC;X40Uv zfEWE7&do+TL~sCt!{JFd2gZ-sg_tzR=sL5Lc#OeCx@%@6@KZDC{51bZ3T;hcArm+m z+|4Zw9S@TCGw4}2Ovxok)be>!i4QPkohE##S-yjJzz1QPY>7&ysm+US$PYanY(c6o z{9=6b0;fd(D4{Rx{~({fDUOET@Jn(Lp11fX|Cyi&^5Mk3MGG-jV)EXl@p$fCb_Kqw z#KM`2DalCCI-Y4DA#7iu6bsDI!)jSpi!}FUID_YiW2O8Pj_t8h27UtWx*@ax2BfLq zpmZ$g(i4wS`L1o@V5ydd9+@Av%IC-3zR1+d&Jt%7Z~78-=L zCwasIaGzeJi^P_2#-O`{wdy!J`>1`K?5lQZwLhS|V^UuZ=(wfi{^~Zc{pgHX_5mli zxXii0UqwG@o1w#mBe=0w%_KcdtNlB$rH3EloUqw0Tr>RRoqcl{yVct;efF=pjk!b2 zO^dtUjzk4j)eX|I}vDB zj0$NMDu=72Nw`_GZxW-)VWgls_oENhzsh{rFt)*wq@%9Ax=^2chWZBs|7BX}+D=?W zHff<(DBeeUV(6Q+DK}xyRZ;jO-Kc60L5}kx8NKyLRgMACrO+3d0B8^|5*UD7OeM&& zc@R@r3k$*`5jK<4!vK3MOwT7^Z5o?p><{M>sIs)4Cj3TL>;Qa5LUb zJWcq61^}iFeiz_jB`+XuFm3oj4*YkFe(W+B1=M7EBN>ZW4;@4nQPZ~)TE0fQUUjE5 z4|FIOkyc;+Pu1pCxweZ)W%_Tp%8JaFpOshBfZf{9v4GRGG%}4?DXc0Jdhg1o|slU2_Jme4S=?eLS7rQdSqxEvyd%_ z*_+>s#iY_r93SJ?1bpE`e4rmMNdq$l+B7hT9%aUmmFDRl9=yfOzZ?S9A0(nLe%0>4 zz9dr7BguW{XzpwUJ5ag9GF3ci`W++a;TM3%>70uTRfVR~5k2B%7DyrU_&WWvBE8iK zeElKm{lp#>N43x*%!oMY{SFif-Mxim(BQu#y&>ilV>ok8;zWVAf{{x3T6y{`MjShy0tw#W)wk0>%Sv9<*v< zsE)Aa=0cpY){)nWqm>wRtGF(}VU4$89b(0}O3tG`j>S-%B?}?<+j3B;$BP&tuyoOa z5n!FBE?Vt_zz!n!WJ8=zFR$xFYZtBdU3~O~OR)v12c`AJ6H~UhQLBB3JbM}bP#$vC z$Lo((ebsxa?}a_qhkH?*fdVUyTN9%#_Q{~IFCE@kzD{ifa>}hhOmIECvFa;~hBEO? zpszu0L80lCcY4FzBFWojU<+#2{YT(FD~POXQ9Lp97VF^rv-$!W2)Gtv zio(4s-H4SNox;B2ordq6okX~G9djbX@7SO1F~dKQ!HB^DtQgKAzH@dGarm8do!ki-oq zG4Vw^5e}m&ae^d%ghVpK%7eDYXfo9pJshmnJcqB(GNC7btzbeo)q?Em{Es{VdTESm z)#jUoi#xsH_=byvo!1NSnqbG%6M~2wkz#)@!Mu{BUL4PQjUVZ?;Q@0;7ktLadFH*d ziJEb_G%p2xqNUCA9tY68ixSP>(-|a6d>}s1ev*9UGw_-NcH$lqwmIC?!%5AM!{ah@ z;P?cy#(lNx%`f*;r;dp%{y)dEUpuZvdiyQ9%v`w#rsfKKEiGUE4_ar`a5lx|d-xT@ zKJZt^S&j#X(m`xTA;(v_5TG2m3hQIp+EHu}G3IfZ4E;)sCR{1@Fgq&OFr(w#HI{M( z^uLwB&Ts+Zv-gM{`1}Do&(#wZ%!;jtPGiKlkDQ>FK53}rG^r$HtDta@OCgS7L8hWE zL8Y;-+VoB*1aGQb!1e(&kjO&dsi&9;RqesG?cf~Z90s_F;aO;BFO6HyVn@U4;MN_c z@)j`f0kp9zs5bAnb7Kt23>*tbPy>s+Aw^O=-vVXWO69fpeB;n<>5QR+HQ^>aX;rwE z?@*ReDHbOSe@-m>y<9cs@7rU3@&T9!^pC7I$3LWq|81%`8l&+4Y>d9aO0$^fd9}VG zYSQ=uhI~X{_Q$RbU0M1h@pKz{(oZ_VAVhXQqKC5Ukaw}R95~FN zZJJ3(*c3hXaG_g};t;q$+RNyHsqU-}m4C6-1Hu*aSHonOne(TFJ!z9S1p3WMmt z4%3gAe?N2eFTdmDx^*dnG;qWHxw}7Jzu`GWgeVixYc;FkjVat6Qj!5?h4efd8No`TyuAaX@;dq+R^Z`)y9 z!oyE*$`nhr4xmZLJXMEt=Ge<2=$)?TH{fiwc!R&YFCBMB z_dHBb$5HO~bUy^@_hZ@p$@Os5i{0Vj%mCvf|By0mK8}+Rr*YI_k}8hs#!B&1cSaF$ zJy%0LShy78h1pU_rYg039-X7;bdCm4EI&G=-plJ?%TW@2!|@lk113tL$D%>VLj8wP z78*R-iATq1SXfCB4;s<`Hf1Y78E6x3>vrUeTiI8jW~>)6n8uyVZe&XzZj#j$Vg2Ui zsB*^jh{|}H6P2+%4ZF;Y!|1{qENp^B01B>(E*8Cm^vh?gDcovi+5Ji`974y^(xiXU zAUdbsad4NB$4HH{$5sd?rHu% zlepb#zlS*|p-5t^VFuKafEkQ;I3T_EDW%#2i+dldJyc(syVwt_4d`$YJz(A3vso>~ zyFWck+HF7_KgM)0(E z8%QD-d({!#Fhe}U#4~%oKnZ-kxv?W@f>Ti$XhkkaH|M;Mb0#s7C z$vQ+4i^8YWs_w*=3e4N2Fan5fGnYND+Nfh2k)pp`Dv|ZgrQmGV$Cy|gN8sa<^g5{f zWCE*r3Dp3M@e=w8sMY3c)Gmsjh5{5_=fcQ-cq8NCPWpW;d*Y!aAf_;n)d4V4K>s>< zR05q~{&hXg1^hyA?h(TQO{CAoLrC0#8%W$Wz+l|$X^U-dqC5xJykJ1imZXXl@m2gh) ze9A>A58=BWDTPjDH%RmfwC)Z5)Iq!=u&JRq0-71H8OflTuaQ|Y)-pDWGzKfiTuQJx zAegg2FxXuR9c)!JlZM+4IIeTj-82Js1)2%HBx6mGzCst6?H**aEPI@#vbpT55ZX5L zLMebfPpkQyU0^utjTD`J3F&0L`8apg&!fD=MH+z-rCpLe>K))^6U(03jib&}Mj3S; zmfGWnG86g%3ewRrHXV&RmUJY0G$D9JN70v(hmADS=UUQ8l}%HLSuHej`h}#C3$DXv zFloF4u)F+Obum(O1v1e;zt0-~^5I87T8D6X=f5 zH=JB2Vjti@X(R0Z-d1RMwgp{6f3?tbP?9Z&or-{c&ymz4FAK|s6qXb1>+pMoZI}4; zl%xG8C*&{~GC>~wW{g|j+qFOCy$JPzm0?!9QdKM3LsZ)}+G{$jCKL7LC+a&|>I)~; z_p++blzG^MoxpEetq=Y7h98oSgr8-%yd^Vv4y}dqnXhE#1bOL#zas0kC|NHPGE7I2 z^|~{qp(T^9cc)GFmRy#XrJOSP6EfcgQq5&~Ia+8ae+Fss2%?esUeY};@S(SOxyX81 zu7$}PK|K`q{zBt{d<3Q1bjSIX>1(6=+x!Yg*SOq0#pwaHI9{vlZJs5#ZWBX5o_{Jvj`FN)(Da)=$#^h-r-8n>q?xX=UQ!N{IYP1q{2$3d@@xt%KwZ(fwU#KJi?LERknDSVm`PPrE?4CU1`!|cK*RbcIEVn&F1(8|)>j2gZ%(#9kw z8`_AzD150L77>v*nPtoBWPsF8>i(?K+qps`nRWDs%!PdgQ?KD&ad9 zY&(|fQvGm5PW*niydo|xQ?z~vgw^+?!mWj$nNu9s6Hv6hRP{+2B~ zMp+HpC-ZOk$*sajdA6EYFbM-?4v`;FpPJ{%uav6&r^K>f-@gTfxESI)^t`-qg6A@g z(R%*DuABA4x^48VG+a`_X*P5+4nC&SnjIVmRx{&)sb)RD$?62)5oS}(H)CfG>Pk3-J)Xi_A4?!qcb|6Fz=a=EEgf`R`+!2X~i_+`~c*f{Qy|=&a0p z^k!xI3bz{Q0B|Le@Mf-{ykf#&w8=N19Hg_B7sHB9!ZHD2zIr1ZA6kN(3ik@N@&5sZ zDivT}1sI9oXxd6a-?br&f_j&{SawM!&=!UR9Cl1Obbj#`FZGvsBV}z=Uj^H8Zaxqj zI*5W_XtigdPH+Cw!ukY{1Ssh+uFB-3gBYv;P38B98P+ED-Br+AG zW;#=4QYJEXO;@F6I#x1q(-(vT4%O-p5zz!3T#*)IrP6~gDFthXQhua+z}0i*)DFZGxS?!z{{Ac=L%H<7WQ=Aq!)*Hkb z11h3DU>H0aKWloO1hWHt68g5ya{~z484{D^`1I9kz3ju}MMCmPS}d*~%^+dWIwQ9_ zRBXKr&5pf~6T_ECRXQdELQ{PDlisi{gLr~*TbFdLD~W)-slJG4kr!g7;9z>AtWOo&bJAC)AGsT zzyJ<&=^|pw0M&&E$$Al)97Ywo1))Lcm_ROO8x#?Gm}@u+s!N20!EcK#vu&txk_WR% zjOaRlPb}r0%KYE-{5I+>r(|M2V1{BqUm!Gc<8&-L0OrLKzI1r0!4`%(40AXY9@j&3 z0e)e_>doJn;2Ipy4xB-~5KIvj5grh7TSxzm@)bX!(*8+M7Cvq4D{7mEm8vO zfs{}qRgN)(SDiChOOQ#7k!TK_zJ?OdS2Wi0hT->yk;|o}mTWuT-7y1Zo1^iD;pu^c z5C~;Pg~#fSK?PZpz0{H*NRK@L41Or{D(hl&O{+Z)2=mh4sv^xg6u%LfS$em3*iL2# z1S|z1=ptrSag&3tSz$JKvsz43T+cJ#t z;mPmZZg?y-f@?>mmXw|aZict*B^2j6=|;NMxW_z$p_;m_jEcMKF5Uxn0;ePWmr|w= zlyuJCC$>=(Y&9RjvmHMD=f$$qJ7B2LA8EM^c7|mlJU0+4n*nf>^&;jS#TPlcex3*r zmh-%9jU1K}eMH0=V}d&Emb}xPyviRsDZf<>G{^T7v_Jhhu3wWQ4)KEXg`eTR-X+XH zH(po|6HGdoBW_s3ZaHeKyCkEX=z?puAiCg0T?^gL6*bar3q}ZJ$%#ifMwUUS9jeW> z2XKbsr&=31HIoFJp5SS_)8oQT0QP+EsMF8K37!^uN|?ene-|aw#Imp7xf;we0N%UP zKf-5h_vNnvW;cT$4!`+$tU?*QA}>`@=&SN8ULEAClx9WO_U`6P>$|DIxD5)5z3J7BfNA zfnl~5@=J+?r~_jTtPONxakpN3IcNoD-&%;$uo89F`TH{pTYd&Nj1AMv9Xtde` z(H@UfU^D63RwTyDC6{cC%O$H^>znZ@M3PS}1PE912+poE%6WqAD~I5K?Q7R^h#(@H zGz8W-vWcQ0=v}|hKmbrZjI9PH@IdGU5j>kQQ}ys*hSaq zysKL6h4@reS8YBL+za%f_V=##_vjR@widsIHDT?!h2K*!e;-<{HU|tQKXi{t;C9@$ z(Ee*RtW(;5y`b96Z2z^S+HAd`UA=z-lZ}*+$*!&9HnG*5G7q&l-%gZ zlyPwq&dpFJl7YhniIQa`dsrAFb|@4)xZahv-A&q|8F^Y)C@> z8D+_D%3uy`<`dgVGb*QBP?okJC__f3s~~3X7Tl5GhyTqrfV3Yj$i$_gV|+#~CLM+X zA4Fi2MFP8cXkE;+Y%*ndkR@^&__VMyHBhb{%iqWy&{CD{Xb9cHu*+c~?wr3{NBlg4 zd;L=6Isl%OUL}5dbmz_o4O~2a>1LW;8@4H%G0WL|=jP2BY@gB`kG@F`xkN}|43}=6 zo5dk+9%SU5%OAE`e@wgkSZ0-dnN}MR_)(0va3Vsz#ghrb_3tU*<>_ci%YP}a22T*d zEIeTB&aW;y;8)PWbMiG#2l9%GH$q1MqwQ(IY6Zu#_w%^dQ=2lI#ho)rh-tXnbHFkn zf!jy4Mgfg#Lk${IDpSB0a`p5!Nw9FOJ@vJq2)61Pq@^s|&Jon!kI*%an;}k$0_=Qa z*#mcBddsd3bECkQkLsHwRW?oFNUb7sBh7jNO*;FUIq}_s4W?W=3V6nO2q!LtY>Lj= zs|d<|Nj&Iw@CTqC>62YVwlv6xHnLA_6ZT_$^dv)8G+Jxy=Wlr8e#=YBj1_Tt3%m#Z zK6Mx6!HhEd@hP#Gi|at{rE-d@PN;}#uLgy{A2fh_Zr`l|#?ol6c-xrReRcQ^hndYt z_nq`2Pd{*d1HQSGPBcQgkLf&!Q>w%_m=W$&EXMJH4`kE8NCpHKT-%9ZQ>*bEQs{d) znPBF`vOoEKHR=!>k}W!2wTi{R8E!XhFG!hcV*v9T*Vf=pp81<*5HNLEC9Ym{P|QxK zBm9H8YH<=r(WzOe13?zT;?#wW^?ns0>~|)zW{gk5(iiT;ggKb%q(ljYU0iRry^bRk z)bO}A8<>@IF@By5YKTT`dPpRh*W$Z%jy>NXn|6sU1Iy@Q-G%SdMYi!5r-JaJfEASa z14#79yPR(c2_)9y&>hY)M=5v$bJ0usgTq}bAI(Kypxz{@Z-I-e%?z(QUJu4Q+^I4$ z5|UCULA}M$t}-AP2cCV`jwbxm@P}O6V_tW?S`-r) zQ>i|O>{yM_Tq?Z6IYmS)eny`1nA6_pDdMU?!^r7Q(3k8<(Iw@7%!w3zTQB z8K+Jics>)?jHl!FBB~)UB?`>frop45=DobHunu4h@4-;=b)|a!URDQW2K}24{L~hM z^~?!SxYSalrLwIT^6L>S-=cz|Qzg%LDQsYwBM0J*3ADXY>(S*|v`wjU!dGe2K1GrW zgT)Om3@19Hh2cJ9Gj|+An=*0)8XLH$&}bYuwFIKKL$zA*8{$v0%PMcl5LgSaK3I8c zp-$2V3NBQMiYmX&nt^!^Q-7_NS{s(iF`K?syWlw}=;GwpqCeVfbQ|oPSat?Oe81KF z#Oa|H(oqpuMD#?g2aL&emf%6fat!CnqRC+7hfjfpjY%pT!fI z~@a_y{Tf7CbqO zIPXNi6ZMp#9!Q(aIuhl9eaie{yL3OhbP1kT2)Z%F&-MX63q$B&)2_dXL+rqpZ?~}K zYX#p|c7gx!-?M+G}*kG6y3slCFbgePx@aD#pnaB|A} zkPB9Y3-6%U+o)Jnu2{y}>k{h;rJ+JdBGHuz{RCDUFk>A|w3|c~U?_n_Lz5V14O+)I zm%Bn~2XNK=^Q9z^q^}f3pE{94-&jUbbS9W#qA9iB%WcK*dbxYE3>q9?mcs8LtRe!O zq(TQ!5nsGy?C~`V$c@wdN7ykpwb`G+vL;YoSWf`4#Fm2H!~JQUZScF=u*`SKS&W4= zqakAsd{OhRssw6wk$z~KGvC-S48>51bvJjK}WROD}f8}DE0C*vBxI$@WZlbbFK#SVVbyPXlJq#c@&lu~ngaCtfbF0lUTS>c&i3=D&w!{W$ zcSkI{!cw$bV9~9D=T3CN;Cc;~5nTm)Y_#tU-tNJ)i^VY{fw7K~XedULS;J4|T1X=mL28Z%%C+K4)!#a*nihIBHP(Msy-EKs+?b@2wUnRQHov zfd$wZb1k$Er7Y?gE5opn2{vWPftN(ak+*|5Zpjr=xo7=}tShSz*V34I72rtD)doj(0Y@6mG)Fc({`& zg9aKIr$IGzny7}BP|b|>VUk$l$(@_ip81#mLV1ugt@2 z;=f7J&(ywToCi%s?`y9riJZ(eIW77EhB70ML)oXqy_Hq#5A0{WwtQjx+b$s=0Xxc} za8Zi&(Z%K=^&qvd9&Q=~X92>K$n1~>dfkFqU@)+M;S4s=e>m-pga1C$h9C`+4z7y3 z;}9<|Pz;|b-;JIu^DUQoJ={AFCSC`Q;%1q5(U28NrV+>_q6n)l5cjNUHm55Pow*3` zYgWyG;b&{$Zm4lrE+PLpW%B#bj)1Di$Z0`jh?vMNk=f5dU&vurH)pC_LvUokTSGW% zmE>mP)QR%LeK=hJIHPc@txxM$4+`>W<+v51QP2NXt9y~aAfB-{tp|Vcw$0`IfsItH zIvpmvF^0l%>TrNnQ3~?tmKWb47`8QDMA@3r$f0x(250H$AG)^d19t7byc!IwbPJ48 zuQqdFf{oPCOI7M?OpV*+u|hG*EALi{Onv7e6VFfVmT^^ma3->=p>^^652Y3YkKH4B z7|Xund$egZSexVNWIn(OqJXcYlLB_1 z=9(Rp>kqO%P?_xcHMrh|^W|d2 z6v4W(tb+xSvf6F)?#7-63nfje_>V-Gprh-F{$9rbu62Zgt+RjXwk2MV$hgM@^<{54N9h69HJ}fqV}xLPaZG!y|ISR;})7j38Wyj=K34 z;8&!M2)(W!t7@jp18p1KYyPC6Jz5l|TE zGw~R5s7KG_2rRaOcQLwkIix?^z7HmPz~2@~7vbZd6f0n8jz)nhxpO4F`)sO6Vpc)i-ms2H{bQZGdS*D|!H2W5`*?=;2BTk1#Zg6rG4BNN7M2^nA_gd?QkH z0zPvk%(@aqaO93sRSfs2a^{+zrR)hL46#qbnJbsW+pZu-V+b;I8;Z%7@|Fp-@8z}N z35L&$H2I0*#CLG?1>z9dsB12%>6$)#YWitrAwQ>&OdJk|1b z)0YJMk1b!9N}hkKA9}*h_L=||%B$zo>5RMx25ysVp*+%6oKptIgF%uMT{DpUtnhsf zO{Z2~t&G2?@FV3?5ZpMy7ycU@;Ldcho$|a<5|D*~S3FR1dSKBK=z@8y8{PAG7AXwg z>~QW*AP$Tji{E;){I>0}xm|?%iniI)z5w6!Fhr?{=ru0lR^|rok8sQ1a%Kg}as?0k zRX{a1=1oz}k>p;Gp~&>zO$sdL*Ef-wK(+JEb*KZ)3lG|14i-%+ccJ8K@wtYn$9X#W zb2ZmstzuuPjTy^?QA~(remI>>f#EkPSyW*1*w^oQRJJlH57fk`jaI8psTNIe&70{- z_wDWb{j=2MICt6p=9V9LUu@$3SwOgld+(zE;B=kRr1m+t%5%Lr7EdBK0kOY(EPK^A zYQ|}tSWECnP#JBCekfAHeS_WB4cy;2_UXDI!36h}UV68qm(O~;>80B(pqJi?UN#@4 zmWecc`#+$U+YbLjdU;{ALocI~=;b)5Wyz@aOfMIFx)*xs#e_)ah`axoUb;$oH}?Oh z(aYdPie3iFbG_LY&-V245?CooWiRw{byYihar@Uo+3O2G+rnWEzq%5SBsQZ%v|$YG zwM@sP?d~;T?v42j78BK!{Kfd5l`uxHl!HACS5`R4wJQ7uEzXd$d$rkP(9T$>-oO9k ziFw^IkB+$ms=)|Q9Fdx653alsyya`QTZn_@)1}lgS4QibE=J6fVUUxC7C;2Uyq6Sv`6N=5RONcEy)bmlF zk-m%miRA+v0kQGSYj1!p2j`9o;b1Pk2WQPCl3fZ`n-+-Oi?{{22L#NMg*coAatsC* zZ#G}YjSnyAF<+vsI~LmI4T{LWh{6aU~ujEAe0{EE@<>eT4FA016jl zES*K!0*?=&<&`BW9D}f6(4?%myyHLe)<&@asJ?-!2WX}=J+ zrIu@NXSnXPr%wuvBpo?`Mv~LZjAj3R{eL{6&nPpN@_zX>ZN`e=LXc?2IX)w!q`dh3 zvDkKH9D*q+W=9klEe z9U}UWfy!q5@o2QY{(#03|E>8mwWoXjEE$uSKjRK1-)c4A#5dRcd77{PU(BDMvI)=L z=T8?{&#U=!14t}o{!E(iNAu^EC3~Mg_cAS#*?;Q4nLpRrb+ns57f3=10sQa$anGNw z?>h6RGk?ka*)>#c3nF!?jv!i!sr`UAe2volnLHK38o%x$9u{W5@3YqpeY%xks>uNYS`%G9#HLCcI~cUu zzd}OAh+J_3HQq(@LL>*^!1{vyYm6SqKhRUQny=gSyuf;xWvM+SIpi+kj$!5mtt-kE z^TYg6?xd7w(g=jWX^tanr1IO1W7-Tq~@*RRNpAbi>c=X?6 zJ%TE1{iscyxN=j+g&4R9&PAT(a`UFKmIpELDuq!LgrK}U^X=ip0fGn!MQj{YgqAD5`;!)|>Yc#7NQWlDys=!FV;8VWm{mBD){S zbE)uv{*=P3{||Ss_ifYt zDxpYvEaRWf{PS@X=E|++nRXqgwksQ0By;WhGEzj6e*d>2K4Y=Xse8#3jmPy1<&iDf z)NwzE1mUx4Uq|0*Z1Mt;OPl#i!An&@p-I^0QH|chq=Z^OSK-(kwT{!L=zs$hk1F<( zR9;dFY(*{o|CLNF{U<4eUGb*kO&8YQBi4RTbyAai|hx(TG_> z4syYdLKeSesn_G$LwFkdp!mmc7LGeDvBszY9zbsPZjs{##=#+5R?UftV6S&szSh$Y##H0{(BlQ%%x`Y5oDJELqiy_5x__1a%t#D zsbQNr6opd;>>9S=8nDG9+6`D=1;*OW3LS?Kg5+6}iyZC&%Y^}pH4v!AelXD<@S#vc zLYap^_|A+3ihzxSfDQA@?*;Wo2xeN%XDbMeB1=-3O-bmV`O?emEl%%5YkqV;>F5@F zCXYtB6c{HWzZ=Gh+)`s~?iVs&Z{8`-ZDvTGt>#2|#!Nq+*5Rv4QL8yz@~$_B?kTT-c@#>Z#wfg>8)?Sc7@fsGJe0D;u5%xrU1bSs6gy3p%f@?j zesmFZO3R{OMps3@kDAt>@zJ85QvEp?zY^?K&bJV%g-#S7QO+eE~~`oxw#Qld@07Qf%wPk71*uKu*Zpd`b&T^jLsO~-cFGxro^;GmTB+aAM#J5*Q5 z2h)G)-lx0YyN2@qf%j-9@Drd|zjoO%wb2Xst6{^RuK#63{^oMrA#bcO-cPT`l_T2h zP(};-r_Fw^H=m_-y$?ZD-k{Z?F5{5NyE$Cr=So_jBJMPPfD-w2eT+kV} zaUrgpo<@T!xY6JDr_Sw8C*bouuiy9g$2YH6`rdo)xo54aQ&p!>T1>`aHR^3m5Km(pw$mH@USx z@7deyo)q)@oUe#%wkMv?W)G%1JUh(RY#%0lc!PYlnT67#mW|`98Vxk2A-6oy*Y)f0 zE!&Er{xRgSHw#}>7o?R2|HYcw>k+9Ut!2{kg92q;WN!26Y{-+I)Wc9l{I)o1LngQO z!0(a6&hGu0v|inD=0g&TaVGy<62fTyxxCwRMYrc6-JYwvJvVfF9@p)8Vz=k!ZqGA| zpF^BMxJ)ty#5yX?)uazD9i2ytJd#WwRhlV*eR*DlCOrnoVq8R6Y_v+T$cLQRAZ7lG z<*HEL)1qO|OQ*n&r>IhTQt5gI_O1f}e+_8nde%_$q5M3wQ)ryM=4suT=0lIIPm6SB zBGe)<{}G7Q^7eHIxdU7#yeoIhbq8|2W!Db(s=eu^!_Y46Z=PQcKf5sdkbWN3&rK9mWdu)xZiAV zBz;Z~PyAPm0Qy7jTJYLm4FID`4;JOt#Q+q%8aLqPt0d-3}a0&M^) zq-$UT;u|x^OmxlkgE?Jj(nu}jM(9zS9?d}0*U~o)D>{*+Xkh+RRg52*cTENWBRdJ- zbK$^t629Vry^rs0Kr_!kUu&KcV_>uo5Y0}l5hv-1AJ3UT!M)ZzWdyLAiUbWm+z~F> zZk637E4$3M4i6vwv^e}l(z~GY+i{2;9`^KJ`>p@K@xOyRd1>;_%4UYr z=1`3A-org3EZE!%Wrs1u(VjP1u3}zzLtYlI>-pyr5J(^ApUb;FS9E(G((Spr+jB#= z=W*ShCw6;o?)E&h_&LuD_YLJyal!wu@WaCgbmfO9AJNGVpZbGj$K~!+((>G_o{!); z)If!bDw|#R(#^)w&i?s!V|s5MR>+2;$}_)SuVw7CyII*oBf1V3*~iH6we~UcDj!RY z&<2=z&f1-&{CR$O*%|*Ue)zHn{#X3)NtW%G7~PuiZME3oh5{QLGPS#eZ`Wai`-XKl z(*-RD=CD^p)vpNI``611Z}ZEUd4f2vB&QLz+#X8 z6#{a(dojEl;M#NtZarn!4)=(?Sr2>M|MCBpJ$_|S$Q~EjFM9iXOc?DHC@0Sf%6Uws zqQPj6XsziPx$I?uX0T~<)$x(!qYQuZ+?dwCNb5iOZJ8sBFv$4)$czO%D_5%y3uk88 zOld!$`Qp;t36WMsP~&zjx94+fd!@o0+5a4Te9iMJRL7MMj^Zvq%7BO*RJbvtSwrQs z2)>R6^*DcE=1V@OACc!*Y1mFMJL^$p9D5^!>wz!sNXsi^i3VF7gyG2E2aTxRyEO5d z@}6_kn%vChIo5%u4a3Gy*~F|ObpG@zl{~v{S8YP1)gGA}=_qtF5Ak4d3dNn)@RFwATSrV`))C55 z($rA$xDX~9+{ZOSU6Ls#nx3iNYIoye>B@a_$KvLw@ZPM3O$X z-rQztgw!&Sxy-(XARX?3>-|{W$<^lF-Ke`Rol5l2_3QPZ2l;n>a89h*?eGYu1#P<& zn<(eu+dTIigH#ZiJ*Wtxla*1@hPGAfeEbI7#)H$bqT3tGHCi1>eodbimzHAV%z-%X z#%|b{Z;jV8ku!>QWP`fluD*+7&us}-Tc^khEo-WA@rGnz@Q6;dol$Wi5$eTO{3eax zD30!#Q~TJi@ZUAJtOPzsdXpG$Ikep#4JJDe(y3mYvwiuhK9Z!AlNF9qj#_zpiPz<> zo4MSV-8zAO;bbY2T#QU;P%JKU83?N$*>Q06!b{|eubo8w>E7kSzjgyL^sYr!|QSMKAD191kX7J8}$0)LJ)=?OZoVNE~7` zSPls|Y4TyCi#Kb05{$#40O1y>9rpU#v)aF3qK04|f~NQ=;!;ESL~q^c*P@4}XbixY zP%mV;=?cR-*D#Cinewnc;h~#Sp>=HE%4d$F&lI5f*n)Atrm{ryffY8Je~wdDpCM5-@v|96Ib2;ls@`c@AKysCEb;LhHwD98N$|E3IS z7T+erOCm}6Q6aEnqH>hU-svy()sjY6*`b@|Zog}UZ*;tIQ&q+p(ih+H%@kr3(UErr zJ5v`5g^K9N%>P?(dhVY84xHlf7Zb!}08S$7;DFq{_0Q}N$qV@nncf52gGtG|fHWsQ z7?E4=|SK(hsLXx&G5p8e{C=tQtY;igG@d}%*{}A^A9Bx|jdoBy zL{px8P8Hzhav7{y=HWv97+!tN3ZB%X$3zYC$J{eTS(>9>4njd0qU<{8h>zO%WrcSz^3LSFj*^GUF#NJ*@vK zD_R*`&;3-ScEuarlh1DNp>pk;B!;4kX8>fkUhb}oTm5bZDCAG@iw}#Yw|}ie;`+Z` zlYB-AjK{tMYn58W{q)+(1Kj87H97e{N&edq))bik=B2BZbEtIEPMQiCX^IHdKtqm` z=7S5o90KxaBfVM(C~W1bKVf=>-lbM7_PIdF>R%}y_i7TN%H}3fbrf`$o)b0Al~z?` z(3M6;ML9Ra&E4+zXTgrNOg&dh$3r9j0S0=?*U7K4{_tFd>=%s z?7TIQE8Kpn`Zkt1-ws=|QXkNOYpmit9iNVxPD4U3XQ`&tlE|QIJcSXxcHdm-TW@p_ zO)|aLt^T^M4av$jYq4JupkLT~?&a%O1)7Lt_d5^v!c2lX+~!k-V(4$bf5UFTuLE70nv}hjtD1eFgv%q~ z<*Xppz<$`Jdc@#Z<%?JU*(&n;xh?HI_v`8X5GC7iUd1?R(3&qU&+Ua*9SvF9pErjW z*hL1R32klQ;G*z-69MWET7{&yEG%pneHFl2Gcf`TgVc~-btyzpJRj=NZJ7(Ct4F3Q%k3K#PFGH*EFXzIqEJNd#* ze_w-yeHA#BG2g@g%f@}a(`_Z06B*4_;6kPY8f!Q524~}-LNCG&GBd(P{D0Tu8&v3q ze4wxQ*VjL>m=XWzHaKcVwadCM8|u%#Sq;oA%FL_x4ryO4x&pVEAHwIlF?zo!b6j{G z%T@0ugwNZA*T(Q#8(!bxd*$=rFUTBi=?gN)hS!V3YrhLK%?T@UvQVe{77$n|ayFDD}!r-1s zg1@spunKWACKEvWPu)j3;giuPRb%7l4nIi=vgFH~OCt(Ps678V>0E!o`u8|{S=`8` z9bvWUzmB4vOaDi9wm0+eSByli-}GBe9hV8hiF-7>KCN+Gcl<8USNZW<&--JG-`ALb z^{!`1Cq#pXo)U(kFc_lKZe7M5KW@?BvNpY6;-gw`uPqb)*Md9P_f!HHAP z`rxyJSAaZ)jY58WJqqWdp6z7p?4ZTUh4_fQiD5`4!O=Qe!I0>jh0OODdA*_(k#}^bhYW>bu0Y9N@G)RLWl+AM_pW zr>_YeUsC}r0MTap#7Y;{r_M&Pz;Lw60XI@b1yV&sHIiva^We_&=MLAnw?E3LA|s-QUgBJEW9=OsUI5cCE;L*lhvvrk_3Tj zn_1M7t!w$w;kRa}!UU zg64IRTuAOlIo+qfuS%~vsl4g_;(6Z-w{7)b1@ zOq1P%lq%SJx5}=fzVtI~w8zq;HSG9gHeXUe;5B<`;wj^Ja>deh<#9!5PJI+jU&wGo znQxpdPGmyy9Pfv2Zyl#8%=aP8_oP*Ld?IS}plrPKebu}ILjKbCZ4K2U=1ZsdXYnrL*f0`Aoh<5HtsPdwebd3uTJ_R`>5Abw&8z% zv4HLwRy+OSqjDMk=jU4pOfkL2JYRhmXPa5Yj%KwLeQq%ikNJ`{N})<+5+8{@gq}p?VFtZ$qXn35Kk^HZr@cc6!9d%$!mNW7ZVLD|LvA zRbjOMyvA%LKLz{{@K1l7G9WzWVM$Ql=x&zg5qc;$j;9Fj;E8hm&c9wP!N|-%Z7hub zz?{~VOees=Tt7aXKh3=`COD#O&6v~?5luv7+Bpm^lY!%@y3~s`E%TU;?aIfbw(DP0 z^}?i;ncwre&>osxI_Z%AVNP;)XigSpp6_f&^r?2uA%5jdzeK|`)j;9Qb<-u{SJT~W z=4;!N`Z)6;#RMZ#FQg6#F7EqN>f%j*62N7yuo>>V}ewcdZ9KIsSV;~ zKh>ta%zSEBDF-$Ll$H=Xmq#op}_UJCE|>Wb2q(i z(#k%Yim_RcJy!hlvZ*Ls?$h^tI`Y<=Y57l3A3!oOD6AQFb4j{;%}tjzx{b-Mf&xh~ zm#d(6aRf*8o-XyH0hvh}Me=;sslJ}+LnU}`X6DeOEAud5Xj}88EMXfoUwTLh{=@Gx>BjZO(Nw*e2%_mRZ*zVxI``$5>~G`fpx2`FGCiY_hhL4(-R9?|s1DS{CZ)b};P{rJeAsweZ)&=BoCi}r9k zZuJ_tiwilP_#WQ*=yj$wuCH~a% zjBX3*(u(p;;;5EqbX!OtpdlpvfR<-;TS%`C(|^_SjBX3*;x&|i)0Ss+Tbk%iOXYta zfp-zenXRyOi9V~hkz_z+nqq6P1&U|W5v8wW&DE|S&R4zm%l&7VZdW_`X9?Y4q(D9s zKH(4+xtI*_(sH}XrdEf0Q9kC$zjNrnr4tOBh#@+x=V%@M?f1#B zuS8nq@G!F#%~c8Q3rsF;bSH(#-dR)v-*~rOEo5Ji{g-~;GITz*Z}HxW;43sjHPqrj zX3;KtSJXQCe?YChei5~1dej=|Q7eBA(V@1;WnY{b4Z7pgb^jmm=?hJ+$EU%W|3~o& z$jW1fM<(s;trwZ3XjvDTZnxC+AX7>&){ji*`~or^`wPfa^9#r{=ogS_vph1DflOYX zFML`CfO4KT#DkEth|>WV;NJNXGH>|67SRg#%rt{^xA{E326X~)6EP8neSMJO z!@mY87t71U8$x^#b!vS;OLIt_TBX#_9AU`O;TE)+W{cD6RNs}{3ptWAJ#xG!(bPub*|GIqLr$phf(o|! z;);r8VRjHF5aw6Gj}oVF3Gz(U*NOA2z@oxZ$5YPo>}m|5ReGXxrj5-R3IBKfT5M1A#xex#18cPik`??2)%00FNYSuquJV)9<_>CiHW1N zxkfCaA4(IP&fPT4XsUJW9IJzid-dn}qE(q)!SZmQhkB~Ak=BcZ_cs;zM|nku!}gzM z7tF+2sd;V9k6+Ev40x0ntxt!m>q~{;N%CjPoWz6O3#g4oY+T_ zw@Z+b<5zg|uZadpj3VM4iztOcz6K*Qtwkh6LVle52>D3&o8VLmTkO5#>B(gZ?-%8O z0t^byd;)Ytr409cchgS5W!1-#qcUw)1gX@8Whn&CiEOC85=p))Afs6%6xt9^9Yp<;G`U^s zAM}1SUH7(yZ(eWxunY4F+6lNAxw8I*XpX^oXeaO`m|B6m=tlQwOX!kXs926}Vgu7n z{0RVD7%mO|@C@D)?9b!^8qS7*3`N+8W}oSLp)y<2vK=uY4~+NlUZ0MZ)|jLv1Fsqr zwQ3mZI%KWQkhQj|o!*~R1k_QQ00y28Q#pf=nyK|=kq(&|X|!S-?rO=EojVG~2qE`F z3ItR8QA%m;^bJU)l+7r`Mz%{yh`Bvg(td?@A-zPeCg)Fbm;0NFF0j8}&z@??i)D$jKbi?$wM;0_lWMb2`BxoSuBI2Al_2v?VGFi~*-V*K7?n+B(R|W@vi1GBLpjnFL?(o zasa>NVD{SCwny<8#Zh=PIIS$!^DIc;fdo~L5(VV^18^Qr(Qb=@T@8wR+Vc@?{a}Bf z3}5YrrDe-PK(*L4vg|gd-$fM^hF@fq#6^CQSD>+rrldU_Mb!nj$frPCoAzASE%I5- z>OR@=4llRn2>iCYAU@8ktWx+zRv7Ckmx~^w#L7M$>NDnax<$TR`OPo#ewEYv$SrcM zgydu2W8l;KpSoDH(DJ@!fgqNgA32wG*<5{;6s?an<{#PHXL6L}1&N8vImIO2rHGA8 zTChKuS9GParQ}#6(8r1+qhxGv25R(`RIR6|p2*cu;(ARpbDV}}!87(xbsUvV{|YhC zF&ZTHw?++P!{c6Z6p3<=qs_z6K@I6L6M6aVS_MAg92H7E_OhNMTQq1jrRLyP)67}) z6Kq6o<5lyVcgc;$RQj7}FsDq53B+KFT2y+Cx6H=TZGeEx&LQ#fh_sP7G}>Ynmn3SQ zMFU-ukv=b+*V@#FUeT&x6zm7x04lK+JG`?o&!UMgvAEU-C&9UZ+Q?^gI>^1cCAqN= zZ5lfC-OOziIv)Geu)}+T`oQJC=#u=9h`&6p`bX-PvjTTme>G!; z>!UWLlVJxw@b*_)X@pw~+|q5;0UFlfw&x2W!5b3=YiwOaX7^}uL-yM+a``msVT1B+ zE_;v~#a2T1(eyP2k<>FiS$U78=pZUEKF;m14aEgFx04zV=Df{C=A~O^Pb=_4-stJa z#L=6UR{7=R^I?kD7w@%Bp*5_h4W0Jl+Hdvss8&j}`rOu?HBBkh zR7p+Q2Z#p(O|a#rGr!zCfLdzV^M`z&%f6!KesiW;#u8LmNfJBc`klzkYqSAY0yGrA zZG!wW0b6`;MnmoiKF_|N{iibb(a-(a?^W#0O8=m9ie{QGDx7b{gI}%J`4&F8QXlJ} zGr+ez%J6G@^S89v2onA@Q8qacTz&rU<|#6-$6+gN7bOw=pLj^U4Rxb?flj%L0|b8Z<&bG-G)KX*24 zcG=Tq8uG+swWc`hxt&?>4YRh9RZu~|FA9z%@1PE04x-nv(J#vI2K#rGSaB{51(!KQ zIC^0XhcI+89q_i~cgyEMD$)utAmjO*+pCcZ(!(~ivU08@$D?-`)lYA*$+>|- z%(PTjX85e_)u9fzD@{`g+uId=rCmGRCU#xn?h0M1Xx<9&5^4Qb#Z0x>on6`30=RYW_MLN<|L@^z8-7zB8H?9-*L$OddG%1A=XS?t4Gfq=o$ z;CO~$wYeE?bVn??6-tgHqf-G!?MmNFcdb5cxA4_IWT1OQ>z#tR|Ir9B_yV*yfBc6& zC6F8eIY$~IMUqETArHcdNb(S220&V2oE9;Zjy*vnDIY^AEM^Ca(XGr`BFW$A8!&|p z??tQ`Qa8;y7jo4tu56UvMp-)dX0+42hn~!gpD7A(h&A_s5fX2bb}P0Jf#A zT$|)eqaj_862e>euS*~XsoHe!T~FcrbjXeHy{3zrt7qBU>&N-GR{Af+H9GwWGTEoz zzkiU1u=KRg_!P%)CVr4kKpjq=z9 z^vX#0(p-bdX&Sdq(ak`uZy%H4`a=r6O=1Xl=T>0mQz53a>M|WdNwB{xyWJIT!0}oX zFo+bxd;dmskXdin%rc>?bj{DBJs9|$e9q016#H|!58UgkI4N2ZUDI{3;6W81OM4$J z!b5agk>tZ+fmSf7oLochjRq%F$cKG!_1nCU2QBtGc-Gs)+~MtEIGz@#<6eGZ6H(`B zd{+irS7b>ND_tC`TGq5&=3D)B-Xu6sytF&xH#$Ru8~dM{R{CJ#8~PCENACdSea-k^H`W&(|-=g1Ze#zY1xOGn9$ z&5De-)tT(p)@}{c$BJX3VPZJEK-+y%n%?r0vf8znlL!ui!1~oZNAF8qIefUt>(WzT zYS>4#@e`>STRA`q{^joZA2{KxnS@($A7Z4a#J`Z;rpo+lg?zloLhBm>=9Y_yf!3-6 za+z=Wp`(-OTz&e930&f6y2nb^w4Q9iex3TUt?!3ojUPw^@(PjWp2Mh`J>Cz(t%jT% z8a9R=tPOiGh9+sD*g!q_yT+w1i80X##H0G)2|+-8aF?nY`fh#IHykPD*7q|O>=_NN zHSi5u1z*uBpkjuh$((z+#M5(b&{GbgO32JL+2KDjOMIsX7?DoiyQ*bHF`sMmQR&}}RkF?5EtbS||4(Hoe!P?UZapU}=NX{Si_U+IcCi8dvpq=C!j1TYx*SU{=i zVJjb3z7_8Av1;*RLHU|vg~``cI{U=0f?Bs4%Dr$@MyK>iiCuGU&DetF#;a!v+OE}T z*JB}?dIwR^;9T0lMeg}~d{4R&qnmt_ODYY2PGgxg&|QR9qZ^W*P5EAZ<`rrnDikjq z8l8XfL+a{!clKIL5?H*#jr&>%t&E|g<%&u@FXuW#9x323zMN(moiIaGaX9S6^ zvLIXe%afskGsgqCNah`GId>kI-w5xj-9L2Ky_zN*aI!A^Kiq z_t`b)Zs%$=)4lyEBu1NUGJr4Z1`#LF5QFQ&>a@u9X;|+NtvalDh&KOAIRC0g^KbR7 zvrv}!_Fue{4bjW_AsXb{pV^JqnYZ<$1x7vEi686YhP2F1rf7QpC7hy`m$=#(XHN^` zYIo~TnsoK2S|VYTPksTyPw7GPL%c3P_>@(}5I!-4@MFTd8w{`4>ER9{hj?PIupXYl zE-p~q=ng*5&k&+`rl8LpP)oe325G6Y&#TdoR(7C{unR$quUCzVb0Nv>#o_VGF<&(7vR$boAC9l}iQ5(#a zC30-gcd>bEd&?d8Rku+ZOHW+uohETI!jun3EDWap)8VoOC!+w$>hnksrJ9D zo}&Rvy_WA@&b7}2I>Wv_l<%7wG(??7M#We}{WAfz?GMZ$J*Cb_aytkSme(7wlY`n9 zwK*HStwHOVR8VEn%M#P6_ks%<&FobsR~bazolu%CR0OgR7|dKwOLw02*&mwwI|iPHaL~>9rTS=lj}G1jOEx*Da@ze7M{9z zh^TAJ%IYE*<-CD;@2P6h%Sdt|MT)Le-lqeIEm0CVPk7Y~lC^myyHea90?l=0iHiyO z5i%nzH!OoPFcBkGPOcKS)1sf~?Ag64BbzUS9hov;Kg-~{6C*{7&^OnicV^ZLpL`uT zyS^l{S)J*Mv0E}V(2&TWIQx7Q?Jm91ll1esYHEiU;IrvZsMf35>Jn^fUm}`*Tzp<; zhz)6V{#jjn0qcxn?~GMK=de$a)`5JGy5sjX=*3D(K!Dv5#X+*OIRzG~8##7%S@5`c0_HWo|C3mypc#yH2VcBsfiGqCI^maV; zxIUy9HIHJ{JaYkZE(1(ad3m?Eyt!(;prTM--=gwVLS|Uk`OK8(pM$GaU(`37>2QC5 zkN2a+HptZ);`~_gvUUDXxy-$y_`#E6OFOrvqQUjziva*H+pyge{ZTx&6#A{fR~a-<=EW$zfIBUuSsH=q9V(U7#uPVD<`qgtdFP zzff2ojS6An6UV^VChFFuTqIW|Mc3zLBq*c50v6km5o4JA*T61EoNctd# z9Pze2dmU7LfIn-M=Y=@-tz^g1DZl?W{&vvpy~P;N0X$wTNo;FypH~ys?kqLoBJc0y zKRV3OPov10BSh^3v`^&boyGYWbO-FuBOH$84V(FDTiolcFhgwGF+1AR`{ubKr5hxA zyMH4Z{O6~!pAbB4;)^se+8l&S8AR^R$oCn=&W-sw*Qvq2y6 z@;}hQ0eMu$Ow?TjsAWGa%ttgZCsgs^o_4hqOUc&Bv^={s#79OV(z;qO%BC9dto@K2 zL@o7Zr+eh7D+kgPI*{FB^<;u7R-*d}TOoaEIr#FJ@Qayty|S{IYZ#nV1x&O8 zzy!Tcy&ChgSDX5~cjG|O6;h-OaCcI!{4BJSu{O9f^zy*jeuM*A4b-OYG{+R~j6=X) zn)nDAkzt`k&+00D=w}3NGebUX}-_542i$$+1(gdpQJ^b8Qu+W zmxXr&-NoF+s%~v(LLn8{SYka;#N@!-mi9FADXYK z-m2P^N{xiDRZs?udTzBYxN9Px*9A|Ah~{JN&aa6){4W8CczCp&W99=W7(aWaUH~}} zJ{|8EZ;Ek@Bv)@nKw$q;5Vb#;UmBTlCiPNjOx0`hg-qBX5sS-gi?-1XUkQ&)4LO-+ zfhq%FU0c7qbkEF}JkN}CuDYcS)A!5Yrhzu@b)_p^KYGXb45+DE7)dWBBg=kSbYvYB z+*r;Nbfqny-}V*lq4T)@6JHCBie_e^z&eD~xQuB+XwA<;+zySJK8K=Q6W_MA`fZ~F0(lhtl;>@@P&`TNp&DL@^Z~uq_vVU!pWO}9Gzx#(3jFB43uF!2vJ=) z;2eq0`vJ4}7ulxWlAJniJSn37_1fJ0UZvGpIOYZhk%*}_j$3R&8pfj4(%})YuQq~? zZQqEVmYOn;!_QaUv?-ZF*=HZ7PyKAbjQes!yOgyulU`epC!lFqNpHTDzIx8g6iUzA zOw%`$ip~CUH%bYWF~=S&hGf_cf0Jr*k>Ez1Hqj0){4X~4GP3P(4{oDX53H(10#b1r z(E?Y_sb%#gJ#2%`Oc!vHh`pGLk>h23EI zj%U9p&mRe{=w0Y9P@pi7M-M448VQ3Nenyj2k{H!|A)6`t2Ukc=jXZpXFu(PMDZ595 zXGDaegP$=%lllRm>s|HQ6l{021f9ak!0&Yi63FCuLmS(#i<|DroJr}}W2HXMdgS+#KFdW$YtW^wy(6JHkGMOu^fEiURW*i&H3$MWb>tC zz?G-exPU_}gE?}8cEkM^W7f>=q{3e`&;6L2l0;vFP}?kxwak@WZ%JY!bIOP(E~CSK znT@l!J;F;1nReMGDmwRK3^B6QJnZNC?LF3bUfT)LIkfm7&hGz+!sA0d7-`UYu5;lX zuKugFIXSWPo=brv7n~{a1T7hR8{xc87`-N^bKc2YDZ!`?*g3%uE-PPyrFPP$tpS5B zv}qvU$Bl_dAFiOrkbn2hDi5|mXl$i{jVmRakn<^eR~or&A7OaFo#Z z0vr@dD4R$6Ot$D70J%?TUo0>%l6g3b=UjO#_C`jQml_F08G7e9ZK5_aPPU#@LoE1f9q`(*V01Y$RE1X`Y^WVinEytAtq6^p+dgG7)m7$*gF* zablpKP=DvUx-eYmN!+CeX|&c)+!tg(GG;-wFbjL!fPLVt`(X`7L&e2hIYRB4sWr?V zcY&xDQ~lP>Nl@n2wkZ?PXVwI1v21@?{XGZwj;RZ>OQsmL$yt{ zslVz#gRZa_rGg;d+cWQvAw4hZv7k&;0Rza$5-gTLO`j~us+oLlAqF{G1USyL+}*RE z&4Q}G$!?dsg$Jq`c}dM=vgMsTjz&K zw5>Y82Nd?L*F}_?4vCRy(wGwG`@~^NG*L{65BkK4Sj)nSdL8VXZ*b@3x0&`baVI>7 z5vOu%vm~(%!n*GZ*eU2Jml&VN9Su%d0fB<^+BGIXz=fZQgOG$| z?z>^8JKS~m8oNLa_w2`>T{u_BAZ~;TST-BsS@n9XlC4qDCsl}SWTgS z3G^i|X1yY90eIK2u9$`T^33`2F?H1?>JnmBHAB#@LWJqv#jH(fWPYOL?t2@}p&fbB zG#%c+EaNPe)Q1Liu-sRz!q|kxzXF@>8J}>^hEt^VP+K*$h=&qQk=8(UGvtUk zP`6c@7BA{(jy3dd7Dpg8M^BB^9Fdua|3TzA>yvdvp1!wq6M0Uv;?}<^KKPM`mOAp# zuS`HkjD86E27064)QezeuHVi_;^i(d!IWJg=(yZ^rxyD>!uUqF-NB*ntm@3}AeYyB zg5t{4z@%F_|HjU@xSWR0avDSnsvF(EMu+7zWDYeExSC0v&pF&3G@VZ&y6Z2(Ruq1L z=I77gQUf}XkB^hydx^%cAzHddRJb8?)+*L^>levhX|V-6ej$VQ>qDdm=lqT6$s)J) zO135S2xXK0-Ywk=JAq*1N~R9-jsHOq(;Y5z8D()Bj?i==LzmbM3nidiBtr*cveLu; zu-p{%Kkh+p$p7oWtb~I8Skz*`rx3n}VRzQauj;g6%6PLDEfa|5HDQn*z_9?U-+b7eANw=b z1$jP?oZ`nihHGTj4|)5vugzsHh3oY5lGk==NGI0HL+$Sum~yN2#*SPr-%G8CT3a(C zJ~z7JDHbw!bkO3=Bse0;;ervSxmsxoCL#9Jj5({}$uoc8Cx6)SWyVa09T#!c>NLFQ zuwy#pTx85b5Q)#+ANmL^%vtIO?nrikw32V49+Fj++ObF5VA3bt9vkd;vKwj_TJ?J*X2H9 zi`zl1yb9EI*T}3qwPYwK*-dG1%?ceb&0K5ivasUo->720`ep7GU-^s&OyON7ecTspZ+?`{0-RjZS_uGe$=|B!= z?Xp>2t~Po$gaWdE*7HW$k1S<}u<$asWq3Eh^$qU^x{~m&+I_dLZ+V4V$zA49WU@QjTvt&9uF5R&}g@(fLW=$iFvNY59lits@!11txv#XDQ=-G(#L`k-n} z5BtP#zwU1sNXxeV9Th&xyUcIYA^oT)8}sA)zfXQ{UkvYv|HNIUpT0Rqh0>+V{k6DD ze=TJHFJ*rZ0vGn_lu_2FZPwqXJKpc?)04fzK8>-$*6EWBx9C&3@6&0JdK>M!2Kw|S z?=pW-BkgBb-RUQ6#0Lj;SxaC3r0ZH5+#)gAod>6*71fr~6y{ExH@He|60kqLw}wXW zZ#V0-ZlrJB=5O12djGFNShxq>gT_quUzvINpX|!a6UxsEMH%9n-Hjj7>>(nF(4T5O zl!p%$+0B%9nDVL(yIUL7tev^bY_D>M7Jg3Ru2Ywxb<^*c`@?4i!=05*_#bR_YP8@F z%V8*rLJ-1w`9izm@w#Q*qoeLjX})8f-`l`Y!B~u}A%LF@Lqf z{dq3~e0PPe=+B3^iY6d*7jV(`WBq*Qt5X=Buy|@RSDJ`;1or84y|fcm(splM{FyuG zNK@^KBi>C_XujpdPf=$SZimQaop9R{1rxG2%D!ln!4E?U7|4V^YSX`>vmqfQpx|>4 zy;XpNeYP!t*He3j;H4vGp+3`~EYkWj)ihrcV=F9%@=@sH8A90YE$xgp#wWZ$S1Xh& zvZKiFkA$caN&YByf?+*lbHa_$Oii0R(e$o2d#@z#Mv`wD&(X3<#9<^oqB+Th0+W>A zh)LP2%)?|0+ezbk`om_?bgMC4a(rR7NagM%Ox5DS6K#|eANj;mEABc#vt|iWFF>81 z!cO65Zu+ko?v_?%8_2irUWp9BsNwP8lrBnFTo6JL4qfAyFHSZ;&6UGtCDM@q#R)-o!>V%|hJPgBg=>(xNtshgW1c8j>F;Uv7HXIbosKbQU@&U+5zD z(C#55T%l1Cp^3C|3XRc_EB3=H)&Z*0uQE6BzLO>ki7|7*7{z-+5y2l|{$RxLA0zg! zAuqLXMozSa?Xtv@-u|;`*K8KGZzS|<)$mp0~Mh7ho71pMEO7*J506_-JytJ_Ksm1C+ zqkHcOtbS2HrWR`(tD-vjotgbMRAB}>LuY82h245Zl7Hh?-_kwN#zuDsABxsRA%`ej z8Zdy~pIfq5f;2aZ0JFMRsP}Ew`(BmxRcj-ykMWS%PjhDSIdHbr7Py@IU6rV3)BYqq z6DA$tlYULoy@6MD5B5p@N!oa-Pa0rJkz`8s1cXT9K(^r^G89QVo2_ZHITQ2`{AYJs zyZFj3w3dUX6>ev+QY}tig)oCVz;UtQgk$W4ZWMIP!wbIFE99Z)6W^?6Af{R#a#w`> z(eqWtq^M>-7kYu6mC^o(y$nfvKtP*Dj!r6Ff}^^cSfg1k8O^uz6TWg>gj< zkhM%SuZ)|d2h`536L`&L`Ow@U>Du3xnHuhp)b7767&_=U?p5znhZzMbaCSU(hY``u zT2yuBe2HCJx}xcGheuPW&)v!rVYgW|YY7G@k?ofPw0#U&sCLF$K4k&xBY(NP2c&u* zhTx?8^6VeN{%gv#N~1%|xju-kK+^GkV}?ZDR_nzV$d9}@sQAS=y?DdF7!+;Uy|P(v zWk^`T0!ZD3%*yzHWlXT{(NdWq#$)NDE0A#6aj%hnrG@ANRY!0b3xyIj%Z*NI%SxBv zyJ)awCAH%+^G%8}Ap7}c(4XY&5@8F}AHQ$fRY3K`L(V<;0JHZShAmGF{61NMb#P;s zdD?I%T{f6s;7;3(s#<=;yXTQM-k#EcFCVvR)KFGn^^L+}DrBV#=_%hbPkP)RX_p{# zdzA&r>Blv5AjNr~go2t}zmghwpx|-Mu-T2{Pv5_WJe2NZD!pEnnKdloruEvRo7U^# zW$|?78z3lPg2=WF%0ybnA;rv1d=h&uuh`1AM?q>o^%`|4fkoh8IjUKn1syGG#zm4R z`MZip>mv{=pk&ZxA(Z%;5IHjW)P`ELD@{H-Oveb6nvIqvvYEp`}59gFlA8b z1Fa1;LH}6#5+dC$8d#G$@@>G1`y~Sb76YJD$ZitK6}uW;YI^nS$P~dK`6WkLh>JJ8 zB0QJa8oppD?J*1gL2Db`IYYzw5a0q|)Ee2$Gc4&|&EPs`4qNp_=0X9DpH(I_^J^QK zQ5Bg=R2|k(M)uXDMwU%DQ(EKy9Jx{=i#*5@@BpKg3Y#yVKaZ9Z!(6FuxiVu_N4oa~ zR;sWX7?ti-zVWw;&CL%tN4?{b(QD!&9MLoQK?VH@f;}RN4K7uG7B6);zR~#Zy^{tE zdJ#{ZDO6pIeG>I#7V#4j=biQy$gw+WHdD4(d-QO3?WMqoU3VZkw)q7u6k3cwPW{v( z?E)Eiko5H~;r#r^Zn_imtx8j@bLYuU#$^1X2y|zhe)52QP)XMuu?0x3S)%ou4434h9`*-iT-tOOJ z3+@Wncx<8j8)tv8?tha9&#c{aYhR?bRDBB9vQ;nHTJ~kDnXU^^U_{tu8jRt(;56}-Sv6daYm53?IZ)@2Bv9B1#L_?;)>+^M@c+?t>IZ&4j`QUptG-KCA}J(2wcoyiTpz z*X<_jTBthPqHH~~keU_l1YTvvYV1iX<0pfFZ=*4)%ST;RcKrp|^T!vLn)!A3-A4Lu zojxToAIfF6vyLq=r<&RCs7+1ycLhyAwxo%}rZ^klC9i1ZVbDI_-w$o)8N(6&>dG}ndFT3|UPam%M>^u9v=1MNQ3>v; zASLckn@47!%?`q+k>qvAe(oLADHw9A2s=HM7rksJtod1)8gl689}>LsXMCQOToLtW zd=h;)-=jk{(X@8IF*_a7=-$|3Z7wx0iCxF`aLQWlqoP%?j(EkW5Z~8=d2S@xk6T(& zgU4(cxNTWk6S>wKcmj*lKMZBuAGLk%%y5$GK zO+;UdP}}q_+i|A#;9F>=-_0<_fc2!&J^0!!v~`mgH(p*c^gH!VaN%rMs3%g1i2r zJ(YcQ`ukES^&xH0K)p}bR7cYovg=-HT5cs%;ieULNDuK53k`MK*5sKKVxDDr zEPDpQ1V#MmbCl;AUmEG<(ViDtAf@Cj%=1t}Kppq;EhWK~l_j)WTfTqS<}`&eNY7O^ zt6nAwmd-ddltW0^y-(>QCWuN{t%};(`%tB~v@zjfQ`5bp8(0zrc%z3#H~vX=)yL0k zbSn=P&i8&{YjylN9|1FTkfcsGx)0exw6UZ<7=vyDn(p0yBfiuQG`aOh{Z^TViX+#{ z)cps{$UGX{RN1T?IK3VuxnEu{5PNfOaMrd%B-Y0Oz_%GX1PTd0P4B9o17ZM{JM!)y zb7pM1o55JCI5x!@ZI#9dgBT%8J#`5WVq6WmP$pJ=o*2cxNwK_~_op67nlF?tKGmNc zNQbHUo%)n70~t}?A-(0$};^1C8@wWteGjV`Nb(9jLjxw#QzHnX1LF*FmkT7pr)J)S zF~49gU}g|)bo6&zG=)9(){CKqX85<-eg3-me{k~%HBkh@)*Yh}k4D5Bhq(<;!3bUlPFK^Q<*)qiy|IsvLcNqIMhG^Z;(Q+tnMW5fNSU zpwj@B46336G<0=$FHpC^)`GFgYIo2jlEL%F{L;YB+D6oWxx0aDIPsZ3A;}6}6|g+H zLnkYD)USm*!|u9BjA?;FhD5E*UCUmOCl9P_W!kg>Fv{Eks)1#t8(?gxDMlgxp{-V1 zFCn`Yp<%xj`n9Z5CnXQh(oGtwZF4)fLXVJC$orTPYE6T=RckA_T@>TrP$6;o#l_X- zz>XY!98o#Z04TE-6q0?wf(j`aT0dr0%k);7w_`ZJqUDud#I{{}OG)NfgMw!2qdmc7 zL0whQ%mv>ve340*Rfm7i+**V4vy-2W;BlwVQ?G`A2^F>ZQGcu#u5^AlW~LJ(e)mO! zTMb*Etu<`&Xv-XqV~T7P{8xW%^Wdo3wRrfCO;|vL0sv?7LVuA#9#P!irh8_ynZ#(7 z_k0^^{kQtBNL3-P-Xv1s4_$^6Z23ceeb=*S5lGNZ8Q~es!rXtqi8i0M^qWuOY>r`n zGMxVb>Imrf#tG{Kgyh9gh+MxL_5e}fpNvZxP3>8tx$w~}eN-4PP}R+_VnlM1>npJ7 zabx>-xD&R|2J-p-#U*5)^QreL6}X?J>$=c+hIlE#XhSCa zoke1!8?l>wbsWJ2%OxtuCpA)!SbndXw1pDISOjMs@;}n1cPx}fGeGZIdQNY0?qnes z`%CrV`04b(vu5Vd;l74vR-?1JuG?sFYM@l_4?rn(oujMsAOel9T|!9O?~%BN*86dC zvDmzXN}%;DvM5J%&UI0C8+sf~y`e3>4tI`ye5RyZGFV^DfG6vY?tDYljqdDU36Ws{ zI1~FQiz1fKxuucTgRBPl^6?Cq7^&o^A_VJpMiChs-32B-S^bZN?btWGo^PeQ<^41k z)|d`!wR_o8kVkD08r`S6`j!lEJ#S^_H068{FO5*6`{-dU*xHN=OhI-y{84_*Csy7A zTedSVTDTX9$QnAQdvDPbVC8#lSQIV(E1?@IehUL0xU6Z6L8JSCUoY`$)@02w?!6X1{9``h3N71D#!`5@aO+8Hyy8K4qgE-pMTJm>w0 z<>_iDalX|9McVXOs~G!-U6|KNst3Boj^={Q1r(S4hn-|sT952p-|CTkoBaoM)BDU4 zj(YGGl*$~h>J=N}S-sfJRvh>a?RUrUXejy0_PeL9yn%6v2W({p_q20M`x%fYz5by{ z;+oDoqRq{HADm*LI)mjZUiBGgb9s!W(bL3I-x~DEw;tllB0Tb~^gcz>JP-BS7p#59RYJ#L4ZBBhf?!J94nHq+tx+}TN@W|*!-+n z^d_!vp3|-dB~DWt8Z4l7daPdps|gX?E;<5f{02w)$}uj6V^^u|7*rq_-mLE>GI@~k z$a1&mQvv}d;?vET2)0F{%;)k*>kPy=_Zn%|lh$_;E!6$)XT+Pf%DhBl#NWUN=-gH$ z`K-FCljAzOm9ev1A4`q+os-c>mvdXYk(IbHs7iigXL_2~U~(rNr1pnVI1fG`ir$1l zPqf-tufldp-lPAg^J?+n3gh1gUm;#N&Y8*ld9_nsh?jDh^0inhe;ysG!y)*EP4~pA z)yI(= zM_z>?Hx!Sn)#}1@gCaHpLEZae_1RRQ|M|*P(^0C4Ja}c9X`cjFtfEaaAx~~?wM{b{ zfWgz8o-Crt(#W*6HfYWKQfWxX4pr~+@kIzjM!*rmW07O%cc}2&wG14?G1@LOuw)bF z=aNInD5fTJ7_h&gl_J z<>aAlnHQB-h?x+!(D?i1R2Vk>KL*W$+kzRtIoNe$&n|-04mW3G7Ub>WN%H;gGETSh zHWWV-P|(b1s?P6sGRZCfslctr4|u)p^3`d4rSLwnky|yro!Q*J0lSvFaxBR+Um9?h zxKXP4Is_%(Kliv+Z%BIJ50MtB2_9v7;IcwfE7}vecK$k^wK5;GT_rjjJh)6C13UFR z-O^tqJ_!2(5lv2(e-%kK^4@k?=L02Czsovlv~8@KSSkJl@(}SmtkKkVU5=7pdaUiP z-bbc*>MkQfo3*OS42NcmcFoYEMpFv?dWP?zXdW@O*Xw7tx!N zOpKheaj6IyvT55W45$wO!LGT3PJ%wC6$TWlx56#hOnU4Nw`~K--6t(^Nv_{VJE{lW zX4Arf+%r}iCGF?>a3F&vNc6~lBM8~E?|KT8)P}S*;W|letnC}FHxtTvB!^LC%Ki6OgCy@ibE3fMBaDO<=c5Kn=frINbH1SBLis8T?7#lxuTHoT?{zrU) z4JooIL|Y3(HKK&LywEpKN=4mnxE5QD`7LWUh+J-@Grh@W=igqEYKxxIYbu-Qxpv@G zVgtvGm%G7i>#-@ln^*xj&!TnUa%K$gG()5t!^u=qbHrk0bT84nW^u2|Qd>r{iz!dk zLm&4Ft+LF^Js+F7P{cGABt{cbu+1Lpf|Jw9P$m5`6*`6!4( zH4{#=Y6R3jKdO?=@Mm*xBkAtzqfQr(raVJ!xjoP<&b-5=*2xfhhFW|z5Sktqd(}Yi z_9sHnn__ZJ9`sbg!IWT7IbIFwTGtcVKX>*{{)9pgX`RNcVi(vbS;}G(5AJE>(gh|b z4{gg`{t2nD?}-80cnDos|Bm5Ad5oLtPn%rxC1a>rVcWLAc1+XNa>=-o#dTv!6FbVz zFs;3k^d$}6d-%g;h0|1+mn?wH+XCZXEB%;sK)u)_`$AIx&eB;%v}!kjk~ z@Yb^4r=td9ExHt4H+Z7cV)oY=KiNCa65Zk)K@n7EbntTOLX7~Q8O9C$u2XlMIT44d zp1um1hR_sLejH)YA~=MPO7NJsQDn28wvS{%T97&9FVBulRk$&@SrKs?&9!^oXLZ)y zS>potSDKr-3pT>n+kDAkuoZw3blF%Dw%XXXze`*DRBk)|)Q$ktQ5!s{X^L!SW)ION zw*9Y0Y8`MNF9&>rCxqdemD-eL_G>}#U>nIq;8LlqLb}waz6lKqUGw zU=e`_I`O|T235NV1N|BDx5{;f{L`j4Fds@pWC6n|xIt{jJkN`8GBaiL!sG|qZO*)s z9B|gmZTyL!!F^Bn*Zm&%*Hdqi;V6;K`+7zzNyQ9ZRpB<=Mk##)G4f z<6es-|H;6{Q=3ME3oubCji#m`BwxT{_PcVjAUntu^mmZHi4|6qFJXwvZ+V2ZnMIki zr17Ade}bK+^*?b{tg5Z)=6G--A{SL0pysD9DvbxfD{sk_MOq|U`xu-XA*R5B4`R=z zq;`4_i@}-fP5u}SHao8g<(U-mM98n4+|T*O+IYsf*OQ1z8Szb+gR98b=zX~F>*}9 zTlGlj$;SjE=|fBF6_{-=y5z&7?!!p>Ze9XQs&4j=%pP7neT2G61R!3=gAG(-JaueY zbjj#4p;XOc_?sOC8lk;zaIG-Qf+PP?)0-aiHu|^d+(O)1{J+f0C}!=e(YcJg{)k!o zeKdUm0R_I_Fml=D5}SPC@Z ziES07M_X3>gbCDKb`oA3`hp@%Upvo-xgV(T>&chMv)+*1kL$&S4TT8blP&wrmb#vh z-ATPzf5`4Le<5V|5x)?!d%s@@*X z#$Ijwq*mb$m%wXF$kyH;fv%a2u{eM;j)2s6vb8JC{c?w^HGeN%SD!O9aU0sbL(Z^#@C8TH1}=d=I%H;&5Lg z4wuzdEt&MW`MUQ?c6N6kR(aj7$7FbHep~QC33rdMjnst5{i{uc;T`Vb`$X`Ik+|tTA_SVe*Cq8D;x`f70mHRiE(RU~;al*S0=3U?M86!q*RNpy=>eH%$$pc-H=qtR5a%4{wEU8js=wH=KdF+g$z zig}LN6vT*VP41`(MMaARst;xtiu*OsHO*5BvPt4$o(?NXyhOR^zZTU8&x_fu3my{d zEJhk`HMZn3CF<|%n#kFgi2~qrVur5T+ploffoPt4w5G)Z%M%M`SYatEtUh>1Ty=fQ zYS2zdIR=l?<_)JX`v9#`Vo%SX#ez#=tB;1OMz1G7XV^%>RAVn}ex0ZF7E|5P^tLu3 zq>+@=s=cb=wnO!Q$a@p;s;aC1Kgv(FyaUVH7e)?RyUPbR@h-SHI2AO1%fVaL##WGGJXVU9Gbtb#f=zq&wFGQSK*qY6ZjkgTY+3MiaSxZ1dZ7_-*Ch5eTmRG@NR_Hy$c;Y8 zexcg_c=VJXMCW|73*Wh&%gDt*@11HU$ui)1wk}9Om)HTW8i+-3%Maky`3vI^`LZu!NkgF z)mjV#1}`5W1?FyN2A15CR7chMCorF34+o!OHtJ|Ko$wsEo7;15s!R-+`3q;k{$-5S zPty#|y^y{8e>54ufAmY3TNwVYR@4fAN1hwCMHk;T^fsMT>NLuI<5^QW?Weh>&Tg@- zI-7GxzQ%v;OIO?rJS|DGyVkl+eT+^sJ?gp!nYPSHS=%NnN9Tl*x9$3cjn+SoOC+}M zji!|4JiGfmtYRD&egdXyJ44AO&2&uX0(jR>{tkDGL~vr~1f&|-EP;*at-Acj$ahwI zqrgChpveDd3Ku|yQ4s7fQ`@8KYOCxP7;)p!$uCqV^r0Y6APfww(C|8xQ|R8@0wv z?WPCI=0zvtO#2dz)iiI`XmU`p&qF_gYh8#t9aMs1%Jp=?ewkpbv63Jk2L>_iR-R(U zX3}(+g+h?CA&8}wi--#KiI)(=W0XA_$MdN3*Y(kan)mdUB!O`YUDD%yr>*v454Vuw z3-4hsF}#x7I6&b`dc0vDhgZ)kQ}bfOJw|HLgN#&1V=Jlk)1Z-@4a4D-f0xQnl=js; zj*TnPG{EHRxm%zG@x68!BaOA?tVzr$adPN&#}V21JnF18O`lAuBEYk2Vq-x9(7p#U z7c!9`By zK?|m-1vI(X0f;2Fgp2-CA<${Y`unP&61+g`;hsBywTpKYweD_$(!=HVpn97xH zvqsStzUihi9@RWJKZ26M#xOR|p+HCDXg=r{>chWkhQ7L-|J4l5^?zK9lDFUfHjO)m z4^%2)g*ScA)5h#Oc#pjq#^L4DwN#%M8@?}iR!s!^91rUlL7ji5G#ZK0vS%FbcyVjB zAp-b8zg4wxLOun4t+4aUAbz7vw}hXe^UIkFky#vS5)ZGU4K<1TD*>E;=!a5=KdySBNGGtotB3Dwox0Q|7(=ChYeb(SSl-4mZ#b&4yj z?sbMjtL7J<=Bi8lLsUnrNMS!-S)qsFGi$QBWD(bUqYzr4I=FZ#=iSe}+f z{pIBid(mHB#7wm`>Mt*E*^B=2qFb1Pj%)d!)jP27sXf8>vH=*go!Z&JI6OUqC%^yGC_u%|6U!-NjA&@qILw9s*+o!`=^2}5W> zOXE8(^g>IcEKG=7*kW^t(5o$t@4L8HS{mgxg}8sVG`>s7KXRSMtkE?Gx%5#Xo!>t{ zb_KD9gWKg|*@hGi#df5{9@FeCc8k4LZHVj-o=5t4je|#W;DZfD#i!ZULVdkL0XxY>Y+b~kt_V3s#V?= zKK0inVQUKzIjGhR_r7mZD(Al#V1N7p(m194`D7tZB;9{@3&0hhp=HEh*ebxz7_V^m zKUlb0;pr~EMBz^mUdijzFuyH=6N zGzE^yJbm>P6HHT!l6|4=_Q%6z^BI=CBl(#cH+TdoE&6Lky*MJz)lu13RAhkx~ zD&O!+j94a7NvA|B{ZmL(6Tg2M)ZCp?#IP?|VNIT@PXQEK3Fkw> zcwc?o#!E~XpZ{g(*i7`O{AdX=zp-*dn`dh;?N)I46pTD358IQp5;%zo7k z8wQ7(c4wv%I*eNkTxde5X=@iMzk-Y#w!kY9XrkSps^_B}v&QmKQsvk>{}ko`SF+0Z z%(N*O$m{ABcc;);|Ir&@rjfa;oz}^3{XkR%-!B+6ndChYZ;HmP)%jbgC|R)g_`ON& zsJXDx3p1xtllEBtsFS(EF?Q;I`;i1|&Oh+9+u)-+fq+`nH0`$W-(i9x<)?ZX-F4c# zH@8Ne5g(6kufudCON99Ghv%zf@%TbLVt1>EaMeqZ%fzeCJ1-o%_zmU+qQQk+Skq_t zD!5aZ!HrE;QVlhquL7J2X`#T{3giup-vY)4O`!B_UANMFJmDmYHwbkDgDWB?`Y5OL z&oHC(>s8Z#t%;V{gmkx!_hM2=sn+zQ@FB51qxpR3T3v~5H{SeQ0Vuj@h~}a0<~S>z z_=FD3g9%1QxC*U6f7DJNclXlIoBmD;+^u?r)9s-?TlEw_9Ikrn(vNHVZx9UHk4ap( zza8p&k22=e^#7W@9Nnml>6~fEUasSEhbb_3m+)Gns%v7#;(PInR>35xOUZu<@iXfY zRmdM~wvY4+ZJB)Bc2{3vo;jXDQ~<_DM?$~ncJ_#f3=K?l^rNaql(N$X&`bj_M240NZDX6pCs&!s6_ zeQeg}!Tz*196t`9Nk7pPO$_+rXoU0FcUmusUjjdw*OTkpA+K1lX zR9EfBQL29UI;;Aew}Psterr_^c@{`e=ojhFk!@y*LRm8}{FJnv;S_4t z6^Et(YQ9KI!#E8n1q?CQ^PQj?YL?TKI{zwHZR&VcwPsn+m>uY!R^>zf`As@?8<5De zcYH0A=yPVs#;s_&61+-QTHzioz#cZ>H0Gr4Qc3sF`M)+>=j(tw)!>9`gLOWymJ6vx zP&mh)Y5#P4l61nRYn5xYcs^Vq+c{R70#kx?aUrg*Wa=4TppP+~1$ddTua@!AK zZqOiY)|BO~S*CXMclDaU9Qi7&% zLpW6JFJl*p+joC8z_5ok!=H&lKG+$1oM7&{$v>jiw+7AePqoRl(m&pwvZnQ~SGL4} z_%Npnt}n>?YqGj#b!yK3)|{3878GQ9hjL*#Vq_&x3k+1iRHk+Qu}ldgQ=%*7OqQ;V zTzS6(L=4hvu4~HlRxDKnE7}7mKop@( zfu6?6Ssm4h#~V~HYCjFYC@*%4-o5xOA{L@sqY+3fEc<@V(STur;;c-+YRc0Tm53g0 zrmFkJp-H4wZBrF?3dT%@ilV}4dIS|(81o?KuGz?2rZTZ`P6wP}nhFJvXsDBjhW+Ok zkT@ty+XEmCPruXiw|{i;dgc(=-3Pg5$jPA1w4?4T8gdd?Ba`Cwa9RPIMRQ>(1V?}e zg>m_;P6#eE$0uH;;D4OLEN#DdRMeSv|Ku!vHm_tUjczD zp%CMAmij#ia(pEyZp~`VKaLN~^&he9EByy}2HD=Z&agE#M1*>&|AJn#S)a{hZTEi@ zWIaEZ^&Xe?A3@d|3$jkiWS!?95@g+%lJZDj-} z4WU8x>nBPxcj230y5V%F&nV&evR^nHUlrd!grFlP}PXb!xWT?oh(Bw@~dDoVGkW{|x;9CO2Fr&;deYal3ykT`4~m5OHg zkIhx;?gtI^=dKRbJC$Onx5;@l1W~MN#j5HI+FR9ycyHSOOtm?@H=@bj$M*xf8{#>Q zL_H{IgK5%2Bua0++EJ5Yz<-V?@m}}sptW}J;;Z-&z+E|(>d*R6A3TyIdR#L~tk;E0 zLpz*PVZ0GTae76i8w#R=@xX>Qa`EOMGltPPT~J9hIG_wSvI9VFQ$#p^Ti!@CTUcc0 zR=#awATAF0-ObhZ#|+_7g;ObmQD|12W~=H{9CaGhnaT`7vxT#i=X*!s5`;&CFmMos zWjD&TCe$CtN@o9HxG35?Dsr+=^70HuUe{pN->z=bvZDb7xQPve;zbGTYh!=XM_ci( zB9IRQw!Xms)XA)~Ds`sotD)R=Q%M_+->$w;@O7@su55kEl$cY$Ahi5PEmJxN!B&2Xqr-14f6M0}8hVEw=@flBtZCxEZq;Y-M(2ak? z^3;oWuQ(=tVI?Bb%5>?x6>2gwyu(XCHZ%(O)H+~6Ph!~+4SYakd_OdQ72QGu()=a6 zNje3+Ob|8i9I(Wh{Qjc|Xu+$MtU~=1Laf-BZf^&qvRHXH;n$(g_twCjB}SR})y5%7 z=()(@H(sKtqQPpsu;L7lMrm-FgSJWNc`;5)aSn^K`=S0WVV&QMw%=W&7faf*uoS-% zn6Q{KkQhu8XyQ`j=C8%Qbmz>$gc4v>>BYtuM|w6)IJPLVb^Fj`vD3r8=#GWorfbjs zP$Q$3KT7rbg&RNVrYffru8A%4X8VWXcNh@qGQs9O0DAlunf+*&VSTTnqK-zrF;oDV z_3shO=zq+?Brb)QM143ek>`-`@UGGB?L%w3pp`eFHh#b4<{EjJ`jXF_c`0hBr-j?$ zDx`$sh^a2dVptzIEy}^)E!!LH>|}fOF;q6$gnP7C++JSd|B}-x;i^xf(=#%&E1t~} zCDSk0(Aui?k)3N|8|+$m9Vy124INN9D#CA=Ak(KNN713|Q^_4_b|--*a6V%wc2$Si z`HYOymkDNDmFd05+S5j!^a-5SC@cnSJ`;4fu;R>AzZ{2RXIM3K{9WBNbAO#dh7jO9G@;?xQ9kv(e?fny8cm^%U> z!@mqPKC)X6gK<@j6VX|Te}2&XpFCU9pv34{xR~_`w=&AEs^>DvH|SmciI%RZLen`X zv^K9|%0PEC^Yhg3j{o#uq<6^+or(Z{v^KWdpL`NBN*#g*V|$ABIBrXQ@VGUM=i6%X8%v2#Em(3HZ}E+bL6 z>djEoZ3KoP_*94G)s!XKi?Mnd-&`9%7el%y3LHL#IS6t{P3#M&@5j&63GNuGFD+ex zpQqQC05a$2=?U`{m38nWC9DkD{@}%_Y8bl))n1&M3vpd&-6wMueE+{NyQ|hmSJDFq zl}iq<5ayx+mVSI#%%BtMV`}o6KuL`*OP!8=%ISm1fjRw5{A##kHpz8*s9R;5|YExUf;-M33tZ%nV+`IMzwQfdYDOuEB({SE_ygJ z9;w9gcul(W_wO&Eo6IB>OWVx(o&{nr_1G;8{=HYyNZ~aLjsIbbEdxW+|Mb6Jh@IT$ z&|Qr)bTd0?xuH%AQMFYy2-Se7zh48EXw|NXJA%my3+?|@%tpD64h=vjnWGQTb!n9v zamakuOQzV+$yL9ZI!PQ|mMI1JN4D2A6fw^Xz*X;RCei}<8EkdD%#0TbdtLlcf9D}b zL`&IHhw6S^X<}#%lquwg2jK1n|gs1UqjQ&Oc>=2I{0S`Ib4!0a(XIn0n9*K;pNAku8&8YbD zaN@%K$~C?jMI*=fg3%E3Nb}Pe+2MrlU&hW8qk$>UF1PBM17;FiESU6Qzb^KQ&i-JF zk3s$sYt_VSe+xW=(?5YZela%l+i>{jv*Z}QIlZRzst0BDs+bjK^s0uFKm?D=404@+ z-)Br=xV+m{ZL2qZnZ_&7hbmd>ZQ(V`H*U*OjbzT9vK+KcgH*4PQ)KKr6BjuFMKJgj zgr1jNPl4D#&S6mT=e3Tho7@GH!(e=Jt8;B^MQ(vkUP&GCf!MVlA+z?_RoifCqke7j z5DH^IN%RYfZ1lS7tx!{iu<*y99WqCC?(EQ@s?DAqQosK%pC6jN`SU}cYRcrDAL39& z#Ln1ovf}@{hlkDveUe>om%z{f0K|LXtN4>pb0Z_0{1bm-3pjqj*$GgAQ*)rgKH6WK z?yxisVmr^l0kIXvCih?+;?xMoubghzAozKxc_v7d{9HUUgsXrliJs{tw1OlC&)G$5 zkckiFKiAsPIpu&XeNPM?L*hFB_D!bSn!+x(?~4wmlcmQmLqyrAH0}urmVw0JDf_E5 zPo<&Z8grYZua4yP_MDS^DtKQKyblka1A^zN08#0G&wFwkdzq6g<0VI|n5S8NuJm6r zAo1dOyS=JnV1fK4_FMw$uGfI;c`N9TP zV@tS&r+&0lQ1XlV3}H0_FG<*) zd6`-p1lV!`Z>%9ftNh)K_Ns{F%)W{imw@Ql3(0+4by5zc-1=^eD12}|hS%&e*7;ri z^=N52l5*#eoAMGCswL2=jIJK8ZpK9-l3l8ki=dL!#|BKDUv#LfSE$O|P=Igk!;8Ai zs-yv!F)W_~sy|WDpyr^?=%_9x1a=0pdTbON@u+g%`zgMioU-P--NMXLnPX+z;cw zHrJsP#ADqNF0!9Lrnem13Scy-g>w8EyR7HRNth+r@H+pU-!U%If+R}+j_@-R$JVM} z=i(!`lOnw^G;>wzcV28^^531%Z!}5sXi2=Epsg_;mydlJ#IuX_Hlp#-t?K+0`)c*N z^_?YbG~dAAm(dRkI`)+m5is1cb-p^}IPGCvTp=Ya? zI0Y}*BOFF?b3*JV?ITKxIDsATzo93umr+vI|H^@Fb8@-CQ@#((|1Qpu>{UloF7KCoFDP$vL3vkvS9urx zkn+5O^7igr-m5p|+PBLl%6p9O1?^jVP=0^keyLM`A4}xQdx@l-`+F1L3(A{TP~P>O z%6qDe@%1KktO7HVuSBL6mItxv%ME@}DAu8jeQV=4U88F-tzuwOE4A(>dfcn_#Cq|S z_aBgc(vM7k){jg-@p09u9>8Bq%+PW# zKB}50YPUDN&{0E6LO0Kb6^*{%v>;mRRegNffFb6bDO|NQI%bxZ-8%nYsN%Tx19o5- z2R~z68p5&>JG#8iA8`P-bg(G6#8sn)oF1H>l6)H;bUFsIkJ+NNHR)I+tn$>Ift*fb zhQmh>8dp6ebaQ*^=70{?a42ff#Ii_F?mDPWwYlclJ+?V&r}Ba4h7(XALg~IRXvUB9 z;w(LD?Zu+hT}!f)fDMY$7(V+@Ennir26?e#StEwp8qp5t>_8i| zkIXw;W(gE|d1Hf8FKYf)y%#xFEXDS^s58}IsHuMf30EJ(IxbI_-Uw1VaO5Ql;M@FK{eil~ zE8S7QXPqCFMlm*q;ZaT#;)cT)j0_4$Jy+-dq5?iAG5ACiW+!xNaK8N6_zB!Kd#?<0 zsOdiAxy;0`8g_CRHW{)J8W%((p^Qhaz$Nv;#MH*G;e}kicnQbP$q@ePl9Df43uoYfAO#0<7RZg)xA!?912qp6wJ zXFg+9_M&}W962Z)Tbjg|miY6NGBuzpr%iSKpOi?=(-{6izKVCWg3ibk#ChCngMyHm zmLy*?&JfE<>z^o3oqtx4ptl6M;^gEYNl=$1IU+wviFvI6d;1j-5LwhDE-A6gW|#*x zu>*8K2?1jRIO$2~*UI%X{6)nR{ zQcDB;OVlGDeESYvHnbZqT_wYnLESh0i8eoeQBmq|;9PniDNxSeLp!qp|7b14w~JTx zdg$6ef*~}`OH3^B`oE5gJ+l5Z+~o~!iwj**t2^VOyJZ@n_MySfn==jMW_%Y%KFm~p z3YBN(O<`K4t3Mv9s(G_doj*9+fr5FnLr_5KxlDWO{64!ge`4>j9r7!>L$99x*?lMt z>ik#FxJR$+c&+pIpp&AnqLCN4y>#H8LbOeJ@d;%OUhFvT zujUGeCY19TS;Bp)be~4>Y35jURf0^F5;6@Q45)5CYb|@RoFi-vfBbK)WUPrlIGKkw zk`6*U5Xh^Udrsu17PoNH3tLm(o}(?KccbrdJYCtm_;V7+{5{lzMDt`(fR1KG>E~hI zp_o|nqwa$-{CsF{u2vG)OhYDDFZI|xkP|2_HKo;vjm+n>Ff(xB%mHv+Hr7A|2KQ#4 zd@Oi(+*Eyo>kIgV1o%{Pe_`~*TXZ+z#9+h8!L$}O!IK)|k0Eg}t!f3Ky zY~$O*=-XvuSsz$$vg5n!$f=tf-)f57%=qrbr`-7dgg1et2Owm89mZy`@dqGo9&EHO zXYh4PFwzWw3)lOGSrS)s_;CVgFxJpuUHkR+N!>&DJWH{4{^7Oi;S9r+M5Ay^)1B)P zA6NJ^q0Y%|hB{3ewxZJt2B=yG9VHrg+(0K?C|uOw`j?l`fS&2e14g#zXCvu!5Gk3@5DXr0K9M~3)eJBeHHUd|8v3)E2CFuG!AKcwE=;SKZN9p7S zDxjUYa!v%juA%0Fl!OOQj$N}CW4-v;xwF!-T^3s!&UH32m`RlOdq5n{)}KEvLQM&> zv0;Z*HjQk4eE`Vf{e*pSI|6UpfBG<+4|hqo#O~N!HBaquOf7s5jj7S$(OZ;N7Ives zOHgWoo!xe<(gM>l+!{zLSg8-8PW}$e`$E;q0@n+d&s6t-PoC=jPJy9u?szTcu!Gf% z#x-3U)5Vt^Dvx2&abm8v0Rux%o>bH_szmO|I3&v?;;6?{Ck|#e=ESORlb;X~$OD1H z!I~cOqSeKD(GuTL;6*F&9JYn?qJ{sqbo?Y7u>slmiG$ZutN!x2?XH} z7(ZbAW*BV2ULcd&IM4Ax>c5Ng{28gmd4>}c%l)ccI#^K8(k+e*6AQ`-h|GR%91)Jj z3gNj=Jd?%dXsNOQdl*t+ekEl~Jt*I5y8o1;CP76D|Ca416)l~i<}YTcH4apVQRUx1 zoBFb;;9m$ zPM{y!Z)Tx@1-0Y`Y@Hw1#flh`YrV<{@ntpfixDGt_2TS~jlfapXW{-!y#BsC_gs6n zAsJK?fE5!dPfm)zMOf4P)ip=e%5~GH+~nEQ=AXNC{V&gnv|Ls$ykIG)_NNem#Yc5G z{YrE(`KDKsKO8!)-Rm+>+Wg|k?rz-{B3<%VQ6wD0&#U&&R#*w5AJ%sJc||QYADBSQ zEh}K(tEE_NphQubYmUW7QMb&yJ14hG&>kClwP!0uP<~BSO7|$zw{U#lidq22T}zljE2sp7Hn!T$qVKz2hw#4s9wnk>I%st8H{5+P8K z;Pi${p5zc1Un&v!X@IbNcscB#HQ@L9M&;@?WksLU6r;D;kWunp8F*=au6#uBUupyQ zZ-q&|+){$)*k^&T`rLmBPS2-91-g7&8#~;FrX+b9C@p!h%V)_^1X;ODRoI;5sS~9~ z??{dbp3U~0m%Nmx^mR`a=izTHn4ViM#Sa?BnwDZC*Xdr>8gSe2^bZU5FsZF3@zi|1 zG%t>BoxF&U!!PNjP5>xF%feM}5tw~o?^+D*`2N-~Gd`gANt4%ynktwQo zQg%Lpe#@29Nw;O?+$~f@zcp~kb0nw8FG9^*5DIi$U%eu>-X zPB&HfBorO$m$%T#R~cwuiFlU5d|U6b@6wVOE;_NWCdszu9{Mz zRfXeAycr#SYiDV1*Z`y54auQIi;o+0w~|Hfuq2NMNs^0fl;P{KZ>kw|b~!q=dBK*O zqeqwv)(m<|S)=n@8}0hqEdI(aYS0BP-wIK|*cY``Pq-#E>u6$-=|#T54F%`!20jC= zO$@kWS3B^;r0|D<)uzE&=~uzWutfL={uzWT0H2&(-8IXW;?y4GKj1j{wM40Zk1&&^ zszM46cgyXU4+U}Csj7lfqR$u%`W*r%m-DA?O4i0t&}TgKdSBMt@hmUnSdsS}@*D%- zpe;(?CKPL_+%750OT$CfR$Fjuq5S<=bP?{Bf)3%((Qiu~!u~S)tXWT+l6J?M{_CJq zSC{7Ym%XYv*E}0r#P;%&HL=xNs8#;sj*{a4s zxvyCb%$gudM8u&lx`@AKBF-blMzCq1{esVSeFdMkK3VvjN45<8K4Io|O2yp{iP9(u z(WaIZaP`GfeY78pou(F>?99f1zi}?qbkS)XSXlGy60pPSLl3$d)Y4TEK>20lKGFeH znOL7TCi&^t8`1^)NhiU`_gk)+y!r%V(p_G7diB%R@YV0-0`z z1Pf%7EMC6MIt=0>C|tsQPcZc5k&$ryr;X>&!)hNN&JmB=s23P~d%)~hhWeb%nG)^Q z!_{9UuD-I6-7ilaFFt{S9+!?qvcy!L6etyHyiqR{XpjCf6X|33g}D=g84;p$nX1f0 zh?2s}M(5z%3;2y=CnH^Y(~KoWtxnG@u?`h;)tbnj((9YV**N&jHVKBiq>m0RITIF|PMho5X5{;%1LSV(h+=WeL z^uS}HF%!nS91ASR1kXzh#d5;wvKTRWotxkiA}XoqsV*VYOdBg595;wxR%Rmb38X`N z0$uJb|K{0C73ixcZ&KbEWU-y)jb%`C!``gCp#ufA@j7NU<}^F-k`poFbiBq%6``h= zd4bTuSfk&T-y5yJ8nnZ(gflgHnQ!Y5bAwwF>-*K{=4m>6h5tA<1WxRb%*&y@M@?}Z z9S`70hD2d7`5nn-Duqq#V_l+9(?8S-^i?;?Ln(9zs~ea8^h|maDt0ElczCYFd)FLD z21Q}`P~?h=wBo`bInri|(`Q``!C?czYTqVPnk^?U{mqpY_hRoQ*J=TaPo!=($}I8w zHv~<^mKLEsNPMK^6rBff{3S~QSgKUM-+bM9EdOhhLw| zMMZ}O7472Wl!KKg=3rLm@8mv4|6PL#bbj5t`v>Jpq3+NB>iX8#XCs;9=w%iaD^Ka%`zNk~c4{H9JU61WsVaFz|UdM1# zLj=C^=41?f64;Am!kWf0a1YL$t+;-m2RK!y%<?GEYf^###U0F#MCrcUVvCR% zFyMaLE0*RLQkKB$D(oqT+Hn!>2-cT|Sp%8lIsIF^nV(30EGIshSl(HF1a(i#HoB93 zK7nSLndqNFuFdEgpMC^g<7o9z4SAS1Y2%*`fO=kQfw~U*_21E_PFL{-`qZJ+Bc1#v z^r<6hXqmqkGsvU)(x>{9<#=V&n+e}MoZtLUc$2EMw^XH1*?9h6)0gf;i;}gXGg)P% z)=Pn28_&oDc?)_ONOA-`e7Q&}8$vitxSzSC-6;;_E%jx!*0{E7^fwKINcFUX&p;7g zZaxN9;C5TJr)%XThMYtUSkGZQM9B+205T%Sxh3>+O-K_!<{+;PxUW>xWu-qGjw1C& zfLC??&AS|0lzKQo*6THwB_hxBPc=l1RBUB+4XfCSUBK=Aiazpv)#IfLG|yy5Bw04U z*2vlE(xs0tDGCZaLP8>#@>zK~xpF+;{YI)-z#q-?ZCUINu$d*cwI@qE-#sQ-M()t?g!qq{L817S&2 z;$e5I2vi^8;VHz=-OlW+_>DXhj4c|Y{eZ8TXn5k!+Dhb`NnCEuVPkyd|2grN?f1( z3DN1&>c^HqugosSv6XmaU;y)GYEwlE5rJMG9>kcAyLb3Q(K9<6Oa^?st>zMV^x3!f z6g{I=43zvl6ZsC2L%6TjQf$=~>x)>2d3Gexney^TxjfD4QS ztPM!O-~tkGJIK2c97D}_ZJGpBtHAvE|BxI1M>aSA&E%c&k7wprF#qeBd!6S09~lx+ zcANjVXU2cKO*GHOUzoS)@!#6UKmFBr&3~{XZ~XVjjsN%X(;IMyme0^ZRDS?J-9fI+ zFr??-o6nG*#@qh}KmEnjAHz=HeLJdD$|4x4h--MHPveY&F%+zRg2hdvo)19=UI~} z(I9FqFaA$xTmIr#1+KZvoG_MUs9cZ?+2go~O^4Od3zFNBQ~NQE>p95vSIU>Jrun}J z)Wps%KB~R@-cXHm`ym8Xda()RHSw#eCcTTnC#ydc$ z=`Vbt7@;G56;=hBiSW9qvH+DAU@B_P&SZ%zGh}@aAWk}uyhn}0VT=(?NR|W3}@{Z>c{-PV+zq^RXN$c!ocYE>l zB5%XAE48ICW%5m0TVNwzzK`Blv8fo@UCvOr<%XT#EBT5pCEVCv?O$wRzN8j$cE)5V z^a)>+&F%qb!#(GW_sqKmN>Jm*U5vR8_>s(guZGx@J`oi!U`Tx;;ysZRksjgGkrE>p z9N6}B8PSwxw-5|1YwyNw3cAnfZR;YcxgH}y@_mS=1E>-d%2d~Km1D(Iu}Pp9(@up`vJak17;;f z3z!3o{3s~}y>dZLL#!-y3x^~AgR^|h#|<)Du)rej7g{LXiXNQH-R_8B|%L;Ap~V; zlnbyzX*6E?g>a`PG}W_(r(fxs9vsJIWLxbku5^M9eEXnvK$Tbr{>NiKbD;riinT+y zw9sn2%eqRXTLZP2Dvh2QRH(WO(Bia7d3Z3=USxbTOd-^{h7nEMqwe%~um2{U>`h2| zDcQ{>@nOMB14Y_<>K@a0 z*^G%9ZDjQ4E~Iux75wQgENt{2Q8*3(N(`;fY&O-q&8BTsMtqa|P+#0`pP0DJs$h$X z>7^~IMeb9v&0#+6V4G5y<;gU21|NfcDQBn$?)6yk%YyT70P|7llLBB)C7cJ$cw=ks zZ0i7Pd})ahv+@ zJoO*LEGI1cT>l7#*_cg=G@5+k^j;Kg>Ms9o^i6ipRYsnn|6NXT_=Mipp_K4xzUmD? zFr9Ed2%4wmK0c?7yMm7`c^~TwJ~lizj&~@g-5-*8u8cSTu4W;|b5cP1H!2*zTedFR z1v`|{2SwlY@31bOW++tgp0ytm7!sv_VU56|ydlE})2vwenaM;h$ppH?me9 zzxz}G!$iHFa!6PX&UCikk5OpOHhBVXViUHI!uznf7o#O2W!g~(8 zBRO~vpa#J^!CcD0dkV#E0^ThgG0(uW6TClt;d|iy#tHvx@P0oJze^YA;Jt;yId~VM z;rh?=!LsCE4lI{5ZF690Zq0xths?*>ESag?nOSl#nKwg4hEhUK%`kvBjoOyBZ5PNP z3`Y4{N^_V#a~jwjpnFe}vuJG+-}j@=WUoB>0*Zfq{Qu^9#(BW+Ue89{2=wCLy`Jsq z(unZ;SH7STaqHQ|3d5N7#AO1{nQekfa`g=ckOsy(*w@LfZ_ix|Jbs-AkB#R4sQ)MU zI1g7RnN{v-g>(4m=xxRO^e}@-cUX!|&;LX|CZ%}$q4OU@gLwODj4207+VF*K?5K1*S{<~Gk<>mQ2hR?3_DXq zO*wWZL~)LtVK91pnNEwb=P(cyI`Mc+(E6t@V@rzx$k|5pJb93c=I?c+{X2mEKT2_;IprU~$ z_^)4M7+mN7*H2~37^xU7O&mh}sHcnhUX$qY_B!1@vCmk2e!?VTbBV)88?L{f<33DR zL{`6nR_vkPe4xF*u^LwR=ewHxvv^7-#13-#z0P2Bq&QU#B)BLgMFux4OJX+?)1v3 zH1r9}FI><7WkA>oGK*(ve02CD3!2H4N?@(>TxHo|fv6gi?YK=(K2`U}3|1{?6*iO8*xWoRmF>R}b0Y^Qg+ zU#N*HE8Z>Mplj`slJ_^5mY=vG?e&l(t10h+@!RjEkqR|m$6wac(T<~=GpmGm&SJ)( z@Me_6uIZG%=!QSUs~Mw=>HdIeUkTe+_=`<281O*w`nzr1RY%|&cUiCWr(7=VLea^* zcwjo$IuSRW`KYXq@)XLbCEcwhvc%cxMuU24Z>Ch4PYV(Qt{AQ;O1K~0gG#tODB)t2 zpbrT(wVBNlUkrZ|KvOZ2H*oG(O8MOiA(wgpk%i6=(0boSFpL%KB5eO zn7<}3!S#VL9qhthp6O7NINPI);X#K}njz9qCk8aHy0^$*Z10WBfS}ai+-*xpKlIA0 z++(dCiN+EQ1lRw%M?W41wqen&xy@s;drWf=(M=-fyNC2;rep8OzplB6SpB#JI0)hN zTIw|Y2v!Ecb~|QDvcrkM2VT*6SRxT5$>?m7u?lrG3V-x#Zycu(Ng9F}f-OOihK(mE zW-LG;us*d5TY|bma&8JaVnr>Qyf&tmpg3Jm`KA9G0i$p0yJhUHWPcrweLG8hL|`l8 z?#Xkvx8$L_Qz8rF7ov!JMj7NU9UHHXbm?s^!kDGL(UU}vnZp=bE^;PKhuzPtKGB1z zna*gaxu-bu%8nabIO-GE9J|M(dL+z=IFXB7+zVvT(`wkx|}jEmBdUYb4q2)5*c zt-0W|T=3CcaDFbhFbFc7^#p~3?lGMQ{sxbQdL-)Q1ps1*4;u}l>BZRKt@F?AKb6*- z&dlCV;$3XfSSH4SU)aj%&zd1h+xd8Qa4Neua~o{YA;NK5rOLfTHRmE(Zf(t~iQC=` zhFRelws6>8gs$g$Sg>v#jxDvXs;$c*IQ%RL>wu}eH2+iHJVHBQiwDur@r)Ax@F>kT z>%6$#ODt`BHHRwr#u1C=HiICQ^91w>@fM*zi8gy!AA+@hm@9(2Dm;k17uASYQ+8q^A~{%wfd-Dw1tf1$QkPm{}!m zBeiO}D+k{pCmy)K*`QAPMASgXToR?%9l5wrst_%ZR54)tT^7@`y;PB`1^;RA!TT)V z^VU<{g9itx0yXS&y!8A~^IRfvh0v(pel)YYl4$8V=E*)TaPJd+m@#8Zp4myC5(D;G zX)7!<@2z`CL2nPZtuGBU(_$v5=@w^{?nd?tWTdRy<5+1_*8qFKdO-+tGsKGO7< zbST3WtZ9rx(Bxp@9hxq^7yJ?Q3O+>PKDYRn#?g+t9c*LJISqU6hkZMhHkYoOKy8Tu z-yX7r=131BaI>>NC&fW-8nzz|tMf-243=(GS>!1*T+EG*)@v5(@8oxW_WRB9JIsS0 z!|!yRr1mZa4eR{5w{+rnrhn@AofMnp^zez`bxMyye&>;ka{SH{e{=lK{PB+8Y2#VQ zq=?`7*>!*&0Eyq}x75IQ{o#rNT#uv7(3HF`DB&8FpbvL9kZ@D{&fnKF4wRPVcR0qL z&KzTx1<#w3pT$+k7gxVoS``WwRq-8KR41U-dq!V#i)!^GHEXFqT^_X)J?=bWalqmT zWAwuofmLG_Z^E;1Y&Y+_7g7>y3FJ}2Bj9MVYjSQq-t?kcxC+Ax1+9R;p)@Uo)q1+% z=v;7YE?A!nHspenbHS-WkS2Nrd5m+979Mr}%yoi<*2Lwo>;6&%)^K8`0RV%AM`v;D zC^55`q_SVQmsc-=%>?MP+b3d6nNln@{xmLjkx@|1KIh6Bs5*50<#Z*xQ0n@#fpz}J z%qAU3T%=>}ER9-3XIy$}z>7$-@d8}H$ZuW>`@@)y`w4YmA9BPM!DXF9K@^6m7pGGX6q?{*?>>|B*vh>##nnTBepo$Vw-)NI1^N{ZSH9pD=KGmA zwlKTJSgYzNU6OTO8(j>?uhUAbE?4t}Pq@KW=Kd_nf)kcqH?$G2evh?I`(ON>@+hRz zPbwt3MuNPh*|wV|&DB)MTbfaj)t{cLXZHG@Xv@6W{|2NivPOl^Vp6Ib%WNRJBOEyR z&rMFUg|~qc9Ujq{K@qA)~jl(r*po) z8MPU+b}DVq%Bw}1g{VdmNX9yV&ggS8{Av-Z4f2q%jc`PFmb`kwnsn;ZjT#whK80L$ z{?AOhoRgH}qa4K#`2P&q|AM*lh0k1R`~)d6??lf)t;WV1qVV$1drULvY-Z4&ZUSup z^g@hq;Pz@(IT2tK#S=$j7(P+1*w=*%$Q@m~%U3j<6B5YH#S0*Q0uA%x4-7#)o#!PkVu1+hng zT_+fJO?X`xTn`rj2N~wLea`-FY1AQdu|NI%&OM2oo-W;EqQFp4?#Gn7ezCB2Ps#;s z4r}2VpSY45s+LB!VUE~kIp72qKY+h@;ik>;7p+JB zIDN`}nqmQeaZbDwe=+(a$6v(X&?+(D#NhSt4e$=ykPh@IHRtB|ixVGo{KdF&j=vbo zQxGWBr@U~rR&B>$%zuj(BnIsLb49^MoB*pW{^F-W3A?HUeHd)iAit0_qTmob|Hz-dY??GxcC|@3T`oCaXd^N=C%VqcC{{4H_|A!c`IF{=B$os^) z^3e6`fpPxQm&X9yi}aac0Pw(4?mXy-2?)>r2nOJc90RZ+_0!CG%nTpUy<#K2ITiV( z_~w8QI7e(^J|D1m9v>h*Rh_?^$-f1B0A6gqj%xS{`R3NV_5W9zCHd>ycdh?<_j*+= zy7ahwP%t!(6Pk=_fIC(M3*!Jd88Lf94`TxUW1Q?T#<8GATmN6!^!jfoHsM8E|1YE9 zjn@B*gbd$d(=Zo&G0DK!+V|CG7O^p1YWhJjrpJ!1_FQTtJB zKsJsjE~m;SMHD3;eqbHMg8S9Nx}ff|1Gt&RhYV8eBY4Rg7kus@e_;J9v9;eMO?ZQm&b8Hb9n59A&s+(N5=b>92e_b zGNfw3rJpEYavbDo8117-S~x6TQ@mbAR&0{z}H8(|66dkS^3e}AW)%ug|U5uPZuho2yJY`AJ?4dGOZ&VzecPMcNxwg ztkNYy3?PSj8$VbU?*Ar(ea#-Z9)%k}EB6{x<>nA+sXzE$1PNT-`MPAOyy%*p>@dU% z$ur@`x#j)_NvUZXA6*XL&5`mv=lTsAzuZM-^23YOL4OWE{l}aAO%L57TQxs)?arc( zYM@uu9=i5f5r;(gZPj&R7KR!hSIkDY@*)2lCdO?bQ)-JAdVm_|rOVFRB$K9Gi;lfk2p{bJ}rU}kW4f*iK}{mTK~d(qTvpO9$w)${1KlB8t1IM)W;bo7z(&5x(RSXmHI6~1b|K5WA_s!OAM}hRV1&n z-AG>YjDzxhqMCn7*$=si=nl!DAC=DwAgjYCN8q1QqZ?yd$W3McZ!V}Z-H%(-~E&^!$1 zcbo|DX0Pj!r3W^?E8O@pKgUQ&PQS?b;ySZjJquiCGSv9^ofnzx$r+XIr@(0aXP&&A zGPj*&``|h-R4z0_O&3G<@=`|MvQ*|?b7t}ve9X}6+)Oopl#XY^|Igu4gySO!`Bb!r z4YSwoTbJgNG$1!dzE&dpx_JW(IX9`+4D$+nvi$8lEphlsR|eQlH=9_?%%9{<`r75Q zaO@&Mtt2_yP`8?CWd6^i52of@`SLHRE0gQ|cb|9obb6%>4?2<;@|1tR|E7NUG$im& z`r+~moDes8nb5e$`+ILJ#QpzFKm6B&GSZpphiBCl=!Yw@$q@pIpyBz}%;MdEZdZ>w`k;ti=i#P@tBpPR2A zKIi-O!-tx}qcfk|M`+TC&&|$fHGY%&;l@*#*$_^7-qsTTKh+PvK85m3KfFL8q?8|| zAI{Bp7S2nL&uWT|#~g8mgZ{EWpC;Zz#4%v2=O&w(iH^Q*RH5|4J!grv?$i>~53hbf zOTsQ~;;wfZC(@G954ZeEZ_*E+b+Y>Lefr@U2K~-;NJzYSzUp{dZPCeKF;eX`U_b0SWgtid}$XgPY8{p;6qq4k<42rTsoQma) zX6q=8aHT?HZN*JED?b}|lOwbWvkl=+NG48a*1#c*W)9LjWw(`ysQOIQc+O-t=*7R0 z+j5!sz;-wWN_f$QKWw_4FPPTA%5oMsbi?($($LCqm-kXfAdlcdw`q;BnB7>bmwB@7KQyH<5*z2J@|c-{#2KTKH2Vzi9qdMy^mt5}qxf3pJr z6cm_WSnM>)S(gU@Q>DDu`*C-$0&QRIJHW@zJAakLHo?@S>o;Hwh&K2Y5T4YYm_ znrPdppZAa=ub(%%##%p*&J8(Cf2iqDH7_=FAtrCG&)?f$If)MX3yiG4Luv0T>MsQ@ z7o!vum;v68?eEGn+%#o`PsI2)@rc=j%Ghx#T|w(Ad)8L<%D{j*X!-400GcY=6v+f1 zublP5l5!c3K5A)$CugSJ_F+Fid7C^ZcCV-Fj zN9cR;7H)(vReSNkrWPy&y*eV>ivCyw>>%IItP`%CHC1W%d8#!;hZ(oD&sV9X#$N^b z8?p&L9^fF(palThKp9_kq0puN(Ar=`*?lVa2iP0y zVz}z%P}3-(FLqP|YZB*|BxB^x!ADNmhYPt8&W5jrs#Fq&ZLdXKx@c9nW4MupA>0Q` zq7%wVRAP<4UyZ^WBoSbve7Dz@;bZ|CaBROCZlKu4iZ!HJeHy%JVt6%v>?=h>abX;4 zl0OTkGZmJ2vF%8_s;0^hHD3s*c%dPl^c|tbeP}y6k5KbV1ezCwntn+Nw_n{A%}1zd zIPb9qL-1F&wfvR!{#MzOGE9;FL(K|&o`o78q)vRCQP}TQ7#862aS?j5xTUvqeRinI z%5#0D+{vF3nS)niXticwwF$rjMaoM;JMb=Ix}&*x5A~Ob)BoE#fBmVf|Dj7v;$irZ zYf)2>xg>cjNMihW=AJpj$K(ah^ccK}eV%@q!i(2pmeG5y`C0ot%!2%^Er`8>-litr zozd@pSk>p3u1rqM@)5E5=>@c+2l*mhrS1t`t2={H`5f35Yj12r+_-8019$jARrB%B zR?N4c9o_7{JJ*gKXos;f)f#2`SY78|d9kK0ca*661HCS@(XkMgqXEGqd7<*;yss#Z z6Xw$ao_=`cd)^C;Sm@3Epr_P6)v0Miz`NLjSl1!(VO!%CuJ?KyxGO0pUW~`M-WhLk zHVF5sbQ;pn|1VG*YH3u`zgl<>_D2ohwYz_Tg*w^rM-Juu`>!RuM(OLMU zV|4BexSe^aKc(>A&gR#}?@Z0}S7!D{ji1ybRp+q}!*uG=)nzF zmlZ{$Gr%=!CPh}sA8M{364ACk+ zUDU!+5zYz9t0j|MTPAA3C}qPIvX;H3;bf324_r@}cWa?~wLh9VC{(2s`Nyil-!S5i z56I=A$V_Q8u8U<&%~xYao7a|=ehFC%;S|A)ai_B+Mb1gVI?qWhWU%Y}_ukfkSbv6` zf^<~00Q+rs(E;8(GdJ6#aLXH@CcC{}RsvHhij1-^^^d#>qzX;~9USdTHJC1(+PE&= z-XNPDytFKBZ!n*d{`SYx$N0IB=>dO%FBZ;q$;-Vu66Ef?$=q)plg%9ua?ingc&6^B zGP##ryQ#V-=W?GHVcON@D1807)0ECZ zAY63{5RibQFTK4y2#h)0LEt2wf$~4u6qi61%+m@1;c<+t{<`*_`C0+n?_- zygcW>te>F#zs55)ZRL1$kV=wh{MJxS2uBj~i1O8f_oTUt7~4Ev($+_L_k}kWu*4#6 zlP*2=6kWOyYF^ID1YE?>H}KW}_-Mfg!B}vOtrRj#fa48%(qz6rH4j>e(2s9sE@`LX zf;b`gG{XKpx5;DL3jcPVI`1W8Lp4JmxpHFWA?REmsF=>1YE*0uWP4O*Qv@uQhoEBr zkeeOk>&b7LT=iWbv6=NyiR&>9QCDPh%ddd$lE35ObdYgF;%*IKD0gKbaKlMOC zQBo`)iW&woSXYXXB~h`z_3E!--*64#J{+C>@a31f()Ox~(XNnEE5vjbobMAMrTvT4 zew@Q4p0%r7fQ-OxP_u~$)D{vEKuxKO{L__J?UVGhmMB;LKPtZwS<)<}WuP2^P}6gL zS1DGdYKmCnZVhs^g__u?TAgD%8R1|x(O+VU!kTy^CeyS z)rp2!w z+u&7VH{5KZ}376bnnD;kV&&3btMZ_*@HmY9&%Ei1zKvShRJJ3c}~1%=^=!`;~MCiTbWJvf(H zy)w(vB68`cSt?Sdr(i+wVev<@>sMv60zVdb`%QMme~1tW zsca$>nhiw6W&gfL+W8PTU>kY*EC*hm*VqC7hP2^NQi4mA04Umeb=*ym69!H5#0Uq2 zq|0Sa7BOHYVhxHO9**^r7^BLV&r?)BzSV3-xOs`+|7S!L+^iB!Q+dg^sYxqj@6Sa7 zx>uy0v>~#LPgKZ)N$n6{C2wxZqO3ko|G~Rx=^f-xm+lKy-83}tijhtZ-H}s~UzNG@ zm%G_>PV(o$`|iPW?N8kIErRz;gXgXGoRi!;c;A!4q)|N5o}AkZ-undUP7UHO;+fgs zNyc$AkXx==Jx&9P^>kjWbugj57Zi&V63nE@C zNt1e5pFYZsw)DxI5@p0=t8>RtiZ?&vN2BA+jrc%mJpbDx{*CF@N*{iCvm;*AwQ$52 z{w{CCU!R^C@p;36_ipHri-w^+(jwpSJl`6#r!~z^pCzjL)34L1w*>?Ch6(_Fa{mC} z8wLjeKjsAie81xYfX|zp5AZL;0(clTkP6EH{8uE-0em58HUhZL`{ZU8TiW$`5}gyd z8I8@uza4F24*t`nTl{as|4P$LmhSV?X5rsZR0#i3aR+}+2*cF_=i{C_q=*|OH1R@B zpMz=~{0b-jVKk96CV3h9E}z}QmdkQyZsOmrb0shtHykc-FZHjZWbuX9=*bit%{3y- z-{D}p5%J&XN^}n&p=jia!@8iNY6DY{c_@#DPo2C3kA~-_(a_uUY1#W5%s(!ojT4{_ ztReU_cmmdv3a$=#mz?S%dal$KSy#Oox}esTq(u3x5y72y@^dKLfU8wg^!C)9v#RNJ zoj;A)WeXXn7U%in0uDb{Nn|AU-@m}nlGDQAp7|?8KmL&*>Bj~EgzVgZ@uyA#xy)ko zl`PR22Fph~rn2&s4t)e1PYkFTtTBQJj2mspVBs@0*&h*(TV;dm{BvpoL|%_FJ-Dcb zO#6&bbbckyda=&Gkm7J9OE$pfu$s2wd-iB7BTr{#qZev@a+C|ZnqDHY^{Eu*DY-YF z^7h|3oQ1FBNuP+Hv%^_=xZi}r#+^X?F7)p}|46S`Wd5jQ9{}e>gS*}I&C;}wa~#lHNRdGUlT;qhi+Z(zTL zPW!cSt6~zNgR_Sn0Q+et~f@ylP0ZIpQ^LOnp z)txz;Uorc#Viw*}l(%9&quh!87O(#oJ=^*mR7Qq@qf6YP32)Lyo$gW>oEpU7v=FU` z@>{OgJa){U)JN7{!EEB=%xA~E(lEybhMJc3c7659OIH^1$n|xR>d501NuYG}btl@F z>+4qkqrSc@xPp7TkYT#}2RX--H7K z6Kx>Ei7U!OO`iZDFRmG})IUY8v4H$>%?|P-^aP;=c`rU?laODCZQnI-VK#+db?IUn z2RDo|$p4wQA?c%`8r>vAWFhb3P#yqsW)=s778F2z4|?FB?4W)hM1?W}p#P@MpZyQ1 zGtTZ;lenO)w(66}&boW9+<)dn!5o(C+&|d_30KXFendJz{$XSfr783O^jpDvMJ*O* z8OE|(}^^@=d|};=ed}XLfJ> zF)%`s{#T>+6@)+KDX0jS1=;yj>^iM=nN2LjpqlueUVOARVi2;-R(4%vFAcPOH1B+z z0l69A2$cT@ zx#4a6KkR)6c$CHVcY=vTFZu=~8WnNXsEHybN-$B-L0ke|rhn(Mf3HsL-?=xn=-=AY zV*Oj@KpoM)Rp{R;^lw=apTSVak4Wy{vN%6I*j4xMRHg*tMq-`A%hY)|SGGj?p%QVP zY_{b;nO1mkFor^K(mrs|Ydr5EcN(MErw)h_#YcKspQ(H-dzt=bIU%fg2NWyQ$vr)S z>G?oWtdsX(g}AxCjJnXm&~GDiy=uox8qJw%6Eq$6B}e`Qr{#x&Xb#n5=M2MrlIId_ z6H}f;Hx4E(QP+-u6}?iTe$^UFFtTvkIL3#o5Iyto^FPJ;;p1mu5l0@GT{jRwR^6Cj z7EE>9Zvv!buhbSll+W;?@KlNYkU&08h&?BP4Yj_4ObfpwX*7J2j4i;Mr#n&fjF^;Z z{uyR#bo@pjFcF53Dqn`-i9|eqV{4*HUWU)I^S2;%$#0>S+>e{^!KvX2G@A4K8&&QK z4ujI&;VeNV^<1)y=((2vc)Ser$EdsSl~DD-mTyKGy9w0QvL}5j$n;sq^waS!OaZS? z;IGvOBah4}w)`^3hhl$u4v7YP@GrCV67>;ZVSuP(=RxVNVvWzwr^-z6@QQpE#g4=& zv%^LjGrF2Zuzixc1)*IonkxpsOQ+XT4+}%+3RH*chn%vFC0}2I*Qy8JM0|Gp!K=;? zPN`;{xH4*vs%)K@cLHfp)*r3~(T55r6>ys#twqQX@D|l_?7WqaBEnx2%kpzQ9!*gL zS{Aie>qXtKi#k>p)dNLI5#_pwNN1WI-4Z(Bii6kzy>e12_mUy^96VFdUUUT7K8e7v zV__y*=v8T_Q2jDU4!H*zui&y=4D$Ih@^@s3Xb^wM;wv=aSUKog(D2UjNJDIOrbl$U|X6yqZo{unc#trWam?1P> zcaG6+)sfv1#U-Dz0-x~bJ64W|B(h%p4YpVGMqY*mrB7>xwptXO$4sHD*-uKJessJ+ z^x89MG3@{%q!ka0OKipL;#OO+96ab%DQE?1Bi;2ZY7p*VDvAZ54RO2|=(ph_gV8;` zstKjpEw?!a$B&?}@M3@tWAP;Q2(R)G?9+H<8oITYj7%q-(Vjx0lFD34!nL(zn;zQ#Ta8<&eyN;PBduyc{vrcH1+(zI}gCCq2V9ynvyYk7)EF-e{QYHT>)^fXvx{b@6W`r@1q zjJt44Z#FhJ#$%4A1t;Dc)hp-2EIoc#u3J-)9U~Z zuR9oggZ7<^H^3QqMZ6N~mYawR?V=EL-24pmMvQLNl25w{XhN)%gs}wJ*xc|1t*cJ7z;yGNUb%}#&BdIN0&A{|HB;mfu#((}; z52VGG^D9IlnOEjOCBSmfvIAGsd2VaN!b~$+F36$8^j@awjXJWfg}92ffF4FTkgJqN z!ul2~JZ^#?uMLHpP8apHG}%y?ih`+q-a3C_~R!y+d5dGE8yw;{MJCQa6U$=1R< zqQu6mg;DTFJ+G|=XNWNncq)L7%0skC`21JKWH{$hEtZS%oo8?-84dSA5rMEwGzzad z(KZ^;N`z=Mx@;Vw({u5kkF_eBhhlC4>LjK(ROf*u>;#ZRqq?XQWxhFX2rU|heQzeQ zafWiFeT9-plQQPXTD~6AKuyS)2kFqWUB5j6{z9xGf|;?J!rV49o~yFWjEQo`O?aBw z7&}loXRvJ>e$J3=<=8 zNir`=kS)=^$X6Y_7)Ia6h-=#i#ttBLP~;Ths_lUM@K+e_wBD>u%@_soWq2ySV!aq} zWiT`g`g-^q0gZCe6SZH)_3%x2pQM9ZR1x`O2p)4G!V{6m4CqG{f^VRbJB1Y*)c4eV zTQpqubBks@a&qh(Aj#S&5UJGSmyL<*k~Jp?={*WVphwk1|93&u~g0mv;UZ z6Xg}UXVNoKJVdTjy_`c;y+see0M^1`6KllW%Gg{jF@V6xlNkNNPp~VY=ewm!W;?gT z5}3QlLIuQB(NiEC>L|RSnCXtHg*oBJAfY~jxlW0EvvXhKLGz8O#T>_Y;o+_(AL2*# zkx5>4IJ(fF0W}?eJkT%)y%Na-dTJY&-JQ~(jf@OzIa;<)#$e+$59zNj>ZfMdz9>OrMkf{Au!ptp4D`ss%m_1&F4&ceg+2hp>c|21e4YN8I$noaY3Cy1;R$+kJ;9qCp5iq) zNzrBX!^-l{E8$k*L`uwSDBw(x+vV2-9_6oP_dC$QZe_({EhojTDvtgq>e zSD0nF)g1_9F?LzO7In9Lv&PAMMco)=Ng8CTQ3@P~w_;)YE>H2=$|Yg}4H?gt5MUTf zigGcKIWUz~_0naP6ERfaaJWV--nu+s##rE}0_c32gNQegM4f=GiLS6yA(l#^Prz7aM=ZtMcJhGF_0Y(%s?HN>^sWw2$IYaM{Cqi-7VbgF5SlQUvtxquV z1doymdqD4yal!)MAa&^s{8UHw9`6Z`$>x-ya#!$r;A)J+Rs17}&!+c%K&jEJ)4B_> zcX0;7?!2(Oc+;pXSqLx&tp~Bs8vR~iJ_zD1EH(waOu(ypU7za-%@gMp2G4Z>gTSbp zr(OO1QX!v#)Q-SU5Z}v{zZ*x%G)6b(R+pfKsp^J9_zLPbo)T1-pNXBZbBLrkB~S?X z&?3evm(Lu6E9^?~Wz=KGRR=bI%ZFS9Wx#epHDvN9Ce>g~7U|U)>QfH%&FxbTEUseM zfK_2#5nFH)sxvfoJPO2I7-)6qE{cLeXE8M4N-nCeyl5uAf>LBBVyxU+j&NIuq=Obk ztJ#9^zCac1J!vo4SUu21CfTbXt}})(H%Gb0bD4{59vWE^UJA+qH3NZ=i@|g7KO$Eh zb4fHhOC1VHd@qj4rZ5yuuMgk=1QzjIe!Y8Rh5b$TqJ@(H`&nLwy=a*nTl`##+t8rQ zX%3bhDm>zGjyb48vi=o z=ymPRM(vi~xuS7n8{Hq+?;*kuDl z(B{f9*#8HzLA~+|7OT`%Dyn@HLnPiJ57*7itL#Gj*W#rc1;K{oio*AfgoQ)`_i%gV ze!!TiKa#S&37mceBc5aD2$BMI3Cyn7f!#@bp_&?d_7cX8mTxa&71sSYl#H0J8`I7} zo)dGb@f^W!CcLJ=4$0JZh;J4iJ=`?{Mj)G@5x1PPJb{mi^ooc@dvA88r}#_D z$90{aAPzs%sMAPZNhH%uB$2P+mB@uyC28gx(qw_^`-_O?cbUe`kflfC8likl(BMmI zd7eO%C$Kn@3A`ifm*FYiW%=I6M;A8z2HI1$3EIE+1lxN8>xd!U78V6^J&M@TuwWbM zY6y;E~n=yB6daskp#OQs&;imgqcF0}{9g@)DkJvfF3WJRs zvIp&2hqkjHTv}Q%UcHe8h3+;%f87>Y9Xi^G2TVxV#9eKNxC5&>p+}lDDAHNvg{x*3 z=pQe&yzm4I*fYcj)FQ|YNOQe3m#u{J3A42@xS1Yd(dwKF*&{Pnuo@irJi+Hk{hs1q zE#EzS2@*ifv9A8(NT-0qGmdMErC!!~CMkl*w5H*S+p1{Vi0~DbmKTD*;#rnYLNUNo zi8Jt#^aGpfBsrJkTwNnqGc1u4ZhC0mmO8p>M%j4U#^yr4I>4PUMmYy3leZ8L-7qM4 zg72(A3QzC@ji9Ynog&_t91=RMyg+DlgBrZq#>eqQ2En+YBg?XF0WT;R;gA_9mJb3u z>Y;Lq>JukU&K@ieza4Ksqlp`oAHh?|LU^E)Q#MDqlFYI z?upmYtogaiA!?hAud)+5mS_hSV&cG86Mv0z=aE?nd!u7Syeg;M9jcTiGf4yQqi6#M z-t-xd{R%w*$C{rJ>bxfb%qLZs)A2PDY2f(cK3Uq|txDktabo);Y=(raHE$zpXrFF& z$;`con{&#t$s&_`HrMw!Kq%la2K4hO$1)}8tu_q=ywG4^0=e=bnB8O$hyz#P8@(5| zT6oAd6p0+r?znv4BIX?$$7gk=o+ERPa9WT#tQ7`SQ*;rtlykt5^h0#|LnXbLS^}Nc z&H|&q5m`codpsx_qWb0nwv#&-u$!PqsFN~rHcVI=r7$430=_u-Z4dHd+ChB70J=Y~)Gx5y*jf^eJ z_(`ttA27sTJ&ni)PjgWcc=L4>fkqsAk8H!6)ZN+0R#1ny$!RI-S9Br_Rx!}w&A1Me z3hXE^a;HB)6f~)?*wwUXkmMg9bOGRrT5g33BlkZAtpmQqM)kU+S`L-ptFB@yt_?(I zk)oX1*)WfEfb4zgxRfP*4z>LN?JUL{D%LR0_SBoN5a!@as`*XoJYE3|e(czHF|c%K zS@kgAvoU4$gf2j?5v)|x^F>pKaJu(g_9SM-cvTN43Qhls2n-c^&;}GD<*RihLv(jG zQ(|BS?QKx$=VId{JH>oUOb=DZD~?XMTTra=`3iC;;Pa3H#Ao3L_h~%dL)Zq7hfgOS z3%`|5jkNtM0 zf$z}^`Fk@RSjL;r$~>4;jW#(U04qes8vBu+4N63{2~(>j`LI>LVrAO9aZX5I|?!+()1*@J{V7Kt;3L zWCS}DQmCjq4}d_26_5V8p#!mL)ZzqrtVKA*Zn;TO3y}0eXo)SvuvH}X*${R_{L!FY z>T{$gTYUj$)fgxz{N5gbL~aYY;tS<>r->rJ{YHHIhmp)jU8b>-OPa2b2M9DZ_iwuA z9csGV`En>z<1)SzWITm)l7R1cm=^F8zt@6&60T@?_#^YY&^)KD(eFFp%7I{s`F^up z8^Y0M*HlT@5at!*Vz?mb+7R~0b3=Hbq^l1PYIYrB(vM5JHiVy*=Z5erCjD#8uJ7X6 zj7Rnwm!u&{k3p@BJeZ%%xozBuAyn8iB+N?{qXgZhL$K;<@&X1m#4<5CS(q9!2CVe! zW9Sf!J|s2d&O&bXXdwwqj#u=m1yjgv+F(GEPDmmHRFkqK6+{3AsCuS?2tXlJ_sNmD ziDQ(yqN5f7@G&kn17QDYiwsvnRdjVK^oT_7UI^1hy1dY7Y(=L14eB_4`pEh`_!)qjiCx z6*S`cG?Sv1#L${RD7j0B%a%v1aq}@o{o=u%5#0a*R7O%dsED0Q>BKJV_BaB1c>860 zE%QId74=hdcs05mMGeQd>f72yyFs}`*rkwo2o=@^GzP$)>bvdlrfdfLm8CQ31x)wIPGHm*ayJc%~X#}BzSms89z{eH{d%8*tKde zTxVQDnD{uo2iswrzS<(Ds63Bj48DJdqrj)4F)7$V$&pGqdB!7JPP#~)ApXqKx&AY8 zp9-B<3(XghhhMVQlXBNwI+nGRjxl>#SdfSxfnDOm@5Sso^M0+mnD5hJ21T>#4yHh@ z5071`1!ax7F2$9CwV$~TXm)+bqLl!HZM!pKor$y?2^5Nea3vB~^I7{p`ALm&Le2!=T7i3a@EMNibw=Z{P@1L{6`>&cAmT)}QRH%+RVoKytkHSr=UpRny+5o-WN zkI?I^2K5)I&O3AiQj&9Y;2TKPC;Mx}u;xSz%L7#t$O@f=`I%acK+@__hhvRH-X3V? zsKapAUpl7|u^csLKk|Je&h+&kh_`Ca9wOS+bX>#wfWm43PpKZ63Iv^>&GPh=DvUNl zys7}3b;;%xu%}iLE%>5tLw=ygrvqdqKwNt9BxDX{y?HvF(^vpwP)f1;p~zmSSU6;= z1NKuwX3Xocd}p#eG>TmZv2)fPvB9E99R#_JnAvM1C&@=DCSx{zBiyituLkP#Sr<@-db5MFeV z)fyYEj(@MP^0=2@-rG4G-K>s}F$-k3<$C}(#%ebvILasd$kp&~0{VNkNP+89q&onG zTMDJ%Mz00gX63;RB|zU`3M&t?$;z8<3L4AY93;9^G!>bWcF{SYi{ZCU zDR-9t?fn4>j(WUJQ1G<>O}(6|3F8@D!`)-@i`4Ks0Fw$SPvgCHw^UHZc7!;DJ@3{X ziLt&&ecPAacem7v!!r8sQ)d93e@;64jyp0psq_1g&M~3>IFu^wQ%A@hA+ot59Zc~{ zc*SNsLM&jCn^}TDCWIYn8Z5~p3KA5sU!~Y!KgOLnYGZQYPqu51w%T>Fj2XIJqwNX| zj8))Fx!av~JtasX#@Mda23>5|(Kgs!QEpHR{HTB9F1#lk(6M%DyNJ?65Cw-RtnqPm z=yYeWTAZrr1&R*}I=zov@vN%yaM$k=1-_a9!0e(2aMkRL#X8E*g2B1|%#IYXC6I=2 ziJ$;FMzFOIVOqYmomm)Oxr6!IO}&I2nIz3Xu$hLQ9MZ7>Z)7~*s|jeUaa+ed)>RjT zet@)sDV(=1{5vS zpRRf%2|iEIiNmj~W%-a=b;z)L@_b<_3G!{-VsH|RLWdGab9w#@l4^Ngj6{^@7^fI* z*`T&*`<$};5-_ORwo`Y}MqI-gXcFqB<@lwYP&0UF(J@%iWOR@wiUhi~!!lua z%p!i1xz_@oj!(A2@055%BphP&TIzwA#HF^Bpj~LhMz#kbJvaIt2Ff1YNz z(TS1-jyhHD2u#4D%tZnL`kY|Kj0j-z2U&tZCNu1Bz>*mbvcYcp)dss3ci~SdyT-RB zx7uz;_qN(`wTvyg9WS10+EE=V@J6}YopyXIwIznwj(rVSw&NTdY+elPp=?Jg+tJxnlfV^o#DI zd|Z?HcQ`f@jPY+pjDMd8nHv6ee`E9SPur3EPx5b1RJK+AeN`A#)BPK#7`h*ak~II0 zmb=}d`!9PX^6&WuEb0Ct8*D)g?9pt>pW)x(ETuI$^sim4vTjHCWYdniZFW1R$ldO= zqeD_V1{knx$Ne_gzsJB{hP&V8-+byaj~f0xEd1@?^tgucG++1;fjR90wR|)2jXaqj z<4GA$g$HeF1W%TeDuIO|LY>@+OYhbjNY@XOh<{Wp9sn1etBTrTh zR4;7N{8)=?GCxkn))Fy(92Vop1He9pAEz(2`SA+oj`5>9*8n{eS6dDbQ@!yd7ElLB z`S5G)6Xaq`y|nFd>&D_X=>A9HDNXmAPBL^qf3r>ZGvsb}=)N;ch>gDw8?dDN$8E6x zjDhvyPUd9&(eZb3Q5)@81JP*nZ=PLXO02*w|K{i2X@{F7#M*I>0n2uL-qWDxMF9pi zM@_9~f8|P!zY&ZCLh|R~fFvnZr<8KTfj|8|R zK90fv%X~aVaR0V$99qtz+Ti0Ll4?HYc`GgW_&aJ;hL2bLr1^LWuE~5n1yj{xd^{w^ z$7g}a4IlseGn2Ij}?5>Znx()L58hQ&oF2rKew~NezVa=)2Fx#FZd09W+WQb z9WUf}x1N{>ZXU#!wer%Zw-hN}Sam=+171MPkHWmHwY$*7Ny=rRb9b5SAbqHe{TuN2BnsB0pgJGU4R*vKeR`gXE*p7wbu6T zd>HG=e9a^4$}QjV0L>BlQ2^AOU5f)Z5=(fEb%|Njl~0Z*ChCyfy<7gvzhFIhOfIU zwE24Wh9tgz3y-vuUdGkt>p|)XJjM8$yf#GL#aFl+rmD?dg}Tw)ja1k2E|gw)7aAJM zntUAFuU*P?eJZwOJTQkNew?XC+{5F8Wc| z9PwYd+nw>U3rmQ}izf|O^4HTg*qdWuug6{ZOh~{)di_K4FTaiUtPxSC+mmM(oDwT| z%X+qKciQ7-39wUPl*_vn&L)nI^BEZ!_wF5g_a4AK=%@x+i6MO1Sg|cc}~T^Dgi9;N9N*>C7L2CuJPZuRZu9U!UXMiv%i39^ute_SLlaHm>>7Y;%g^1l67M7 zwbbhp;%i?E$LQcF#juPpNGpz52=o*Y%UVpCafzSQp%7pHv+=d#U`QknP5;&S+Ns^h z^UXZct@~s5LsC6HwHjZ$ibg-jClg=0O6lQbFs{29U;7LK$i&xfoJUQZVx|l7H;b?3 z`DARl{pjC~uQjekrCN!v9eqtgd~F)XsAloCxp@?Bjp|7Pu;Xj@$;1DTuWbak{Q3CW znLlX4=!5HjBEEJjyQ;PL+D~EjogOl`fL25W_+$-VGGJ6Mce@jZI*=t`c+zpG2?i|5 zW0DQ_#u(UZarYmIuRW97M!S|mz}oHVYFA+M_jbD?a<@C}DrE_=cGVcLY}Y&-IVI?N zG6wcR-2G?bYnRt2%6xSS08{Tg5m#I99I1|E!T*)`+9^O`Ykq{k5MLYgvn|yB+wryG z(s8>LU+aUUTAsIvuRThwP|I?OuRXd-chPuUlkJT*;%n`}F-DDfWtOeRoWtBPKf)i= z^B>0gC0bq-Jz9G;cJw}wA~rL0%rR7}R%&L*RLkXVcl6>OECI{_7ZogpDZPVCUiPPx0@*ZMCCpb-Wz`yTBo_0?(Jb-D$@wf&`m?HyHGge^0Q%?iB+Y{SNK; zz4-WlIldM_HCyuJUy83~z+2GVNc!)@*SZSMcPqa3b61-mTaB;vg%K}2iTIlHTg{Ip zxF+*s8}YSwL3V~82T!y4u^n^A_)-15R0ECRYRlnaY8jry0{Uz5wf)h!R>!T7FsL54 z?(1S0Z5Rs`{=HQ0c8BiYJ|vNUdm6B$`;%?39b;hIhvjZ}+Ob}0Yx8e+1D5T$!3OJ!fjtd(|6Aj0N1!3C@$p}Xua&@-Z-tNl3-Pt{ z1^3OO5v}_x$8E&dnm~Srj~|?3^YO{d9p~d? z4bUTSwfT66%EFWI@n4It4LGO``fm{4(tNxi+ou1oV)QR}yG8#8GP6zp!wp!We;ceT z2KF@E{iXQYXqMI*KfeRvXzLZ*4>0XnwAgOXT)Eqw_H;>V&!q+|+w-&y_U0Ja>v8wL zF}`*IQ2##|UmFHvJ(;inTk*Bg(g(X0Upp5`HD9+HUwf7coZ;&izR-MKi)%7pw-H}E z0BmLWdit9-Uthu8alSs+06h~|o396{-gt`fHOI^$>KMMl-7w`acNJ=Xb2n0T;9V&F z$m_)iI=T}{yUqAoUg*_Icl z8nEQA(`~Q^#=u%^&tHkJ4Pa@l$qPT_t8o@C*w3`5_A|RZ56j)|v}e84+HOyG16Jgp z4b~L{dm8TAim%m7K`olc*Y;(GR^n@07U6wDeC-E5|6Y9Ua&Uy9+8ZbDMtrRxp1vQ` z|6Y78HJ*Ndf&V-4wa-9oranKsx*PGer{d|SF#T@C*M`Tl)F2D|7MS?`-lss8j<0oJ zvn(Zi!E1<2&F9Z`{8`SQwfy-vf1><(X)=C_dAH~5_&J0>MX|e+dDoXezc5KZ-g)?Q z0e=LZlyL*U-ozjI8s^=4LY7A_6A$N)6F)YWtG}_f$1v69{Fcp3LwR=$f3D$A6@SL? zCxD;k@ilj-_XX#~&MTMw6!lKtNz2waPXjx6gevux6xd<(+MFR?)%yc7P~oxt^m_i$ zpmy0&5&j<%J`!O^5G+vV+&Q7bH?IWMOOdsXTkDOBZRtigoT#CzmU+)bfL zE%}7Sskyz;xEi*ko#aYF0$?mGd=8ZkY4I96KL=cnO$> z9)dOI89QWfnHQg|j?d`&>0@ZEj&JcTS9Tq3B2LLg@Qb~R0Cp%{t5<`Oc4N!Es<@V) z0pEnBUeT3)Nsaz1aH zpT>Qt`KcP$*7#|de8>7{UTuCl;>X|Pr#JR7{4~F|IX~^s7PiSxQ!!LFrh`p?? zd&l|d3t+1aetK6JD8^6o|Ax%yg4zX|pB|O#S~VWm|IbgY^3&#JnxAIAr1{CderukW zwQ+4U?|C)+q)#(mjFUugUWDEb>&|n2jh@TC!*(ZJbj{TQNuw3ztZaa_a{#wqxr&3< zc{xLvWj?pUa)vOS5b^KCPvd6LUq!SxYy14Bt6}V7(bS4QA zgtX7%v$2B;^5im))Pz5aFbuc@Jt&>AXUesdV_2M^6sV4@JA{L#L+_N}!dA3_buzu7 zQO*0D%Q4A`^|4*8P#*W2bbT=awcvJS38Iag-b5F12-{~KS&sYo#Xid}0o)`jRM^o) zZf{a+F^q8gmtoYbhfw(}6XlarsDX(J?o=mRX;tV#zx;ec%2bRbl+NiVU+`4Q+#1;t zU}&I*y98JrYlh(~e1i+vH>u)Flq(#z_jS!=)5)ZltiQ^VEGuGiz)tbY>u zdFmN4`6*|PN+-{kqJn8h66c)xJr!4~B-k*{tKNQ_lgygx^tjt>_mQ>iPBb*1q~H*3 zq%YUxp_6d7&STe5igMqbP~k&PEqLES4%vtX-(5FgPo>!IqlIq@Qn{J*T6K@m3HMk2 z0Y@|=oxKAIur520#1)d&A8P@MooA915`yX9K*6$AorWs~s3SpPq3l1}yw6l0aH$k1 z#j94Y)$+3%SU_HuR$=cz1ZAWW>mGH&51$07rgN=?!42p=Q+)}V_C?DM;5!eFzbaLG!5{>H28xs1{$W zj<$AG`M9Vp%_|cg9>rf)-4e|vZ$78l#J(;x&x_i)ey-mOg;SbkhUeprUS}cvba?+5 zKmDG54=LpgV9zhFy5f!I9953Ch=43n4}rY2F`!j^NF=?Vk9!zD#8mvUGOdza&Gbek z**-i;CAlBAgYX0h{ZvD>qLOT4y5CSq?px5dO0sztsco(L5!cpKlEw0UtGbSRsMspW z`0syDCFz>ZWIR~!cf++sd1I01lS_qwW8wZ z0DOXqU$_S!o76tijwY3E+Pkx^6|T}Mv{CUdZBD^|LdCxmbkM5)9U=7pjr#r7i#2O) zdPcLBeck6-{cN>y?Wo@isfRn@LC;6U{#cNGuR=>NJRRSs)t zA=tNr!=N!n8lVZl+EebCAW)iog)93!m7;pv1lAclcS7^P`|t1`cBB^IQFSm=KyYJI zbr@qpX|NC0D)<30#uZLMYQQJxO_0|p*0u`UfzUYgFP%Zwh^o(VB?*1VYp8I%&Zyqk zxiIW_)obctvlk>0P_8EFtX?&A<-r&f#20{3Bj4VSF^VDp+N+-10@e%_UL-JK1y7Yd z28Re}Y%ll=D4BI0lPbF~s{T@yMs=F5%E_iG`3Y6Yk*YMX0S8JgP`wo^TCK&+QVV={ z=vwSTH=&`}fJ~H8c;UnHFssr#MqRkeFc>Rz@Zn@ z#DBROU-988&T$g^9(o<$S$TtV@~w_n!n)@Mq);qxbB5q0Nu1*e*JtD$&d&y*wa(i^eu!}E&R(aoN%Tf~e ztFsvqi4$<;q#ZCg$Kh4~6gLyQ+^Y(qeXCN){|U87HMLkUP1F4> zUPFZknp(Um=`~KC!^p^O^S~VwkV^1$MeDWLfg@Ap`_n&3EmohF#C?P9TG-ro>rRb} zKBg9@@Cxp8#=D%{Hzjl4XBS~ zu(8p;6T3`6G7;@>8c;7KfLgd#Ks6dr_i9k}^4O%RlR&*02jw-OE=vG)<a%w=t`_i` zEOD|Sao``~SYgYZ@ZsyQTq~M&@JCwooHP;4D>#_jVb4ZiyYuF`fB<+zc6uP?XYEXP<}GnSD=pq;-M$VZICss?`^;JVsa5kb zj2)cs?GAN0g;>qo=?s-+#w*|sosQAJVw*eA4ku1+#2jJEC;K!6_I3s{aDFL9&l{98 zaK=Vwpk$-7rq1Dx(Rz{w(Cg5H)t*Hz|Zc$j}Wl2_R6NXIMo5IZg8k&ub~~qjchMd128tf-QWuRD9*)s>XI?o z^cGzjgwLsXN#$X5?1*1gJdaH{Wkkby2yooRJE|1?=BN@C%TGLk&Ge5_w{#c(T#1us z`qCAIOepxlH<=+))r3PVDm%K+m`n&}citASdU(bJ*zM3d$yC`;2s`IEQyW~=L$ReO z9+FXR_JfGdt=W}f)%*%dDA_VM&GKIhdL`_>)|8C0aT$2o&hifzzvDk~&KwEJWeBKr z8wrSnH8+(sWcjNw6o=o!k1O!0E7U7P1VD!z*Q{u|J8!2PQlLsl12?nQq`UUG7KB#3 zoHfiK@l_t{32cqg%`VDA8&B6wHh5w>p8AXX? z=tnK+E|$JDuqnE5K=8bbZD$4WL!8_WCnGu$zPXwowv`nuC<}a8S~Dx-j9}-?GXtG7 zON%$%uwGnu;Z+zF?R}&|eY1cO5~iH&R&@p~f=uEm97HW+^dE&CRz>t&?TZbNad!VRnU z`oj^?wJ0H#du_gs8~)+lR9;mPvT@J$C3tYbU?5O<0QA8}>L39M&n!p>q(_tv!&wAF z01K{q0>IPA6!jAfK#dC+UD$dX_^_rRoj#GWlFoiOQs?%`P8rvyL@+h4kMuuRw9e~O zL`X!Q)v*s`#z zrZ-F%sC76B7wais02H`d)hPxl8&6Ab2+~c(y?Xj-o}LStPv>{Ks(B2LAsq%*cbIQH zO~Uebf}9K=CbW)<^~15=4)m4fTTJeQVI}<2ctPKc5ze<#ap@65$6NKYVGK( z>D@GM8H`>3Sjebo2>`%Q2TAJQ3E`Mfn3E$7VG#4Em8J+MeWzyMNt>cup_|10xk(U! z!Cp?H3Z`Ly@z27MCG*Xdou-Ctl&U|d_whi_j+d|J4znB3T{nPTBMtb=eZQeeN}Oo6 z`M}n^QsM*>0}I^;SbJ|SuvT@MDX;}u1ZxxMV6;3SK%y#n2Wks3 zj&|Hezzg3Tmxp|gJUMi{yvI@*&;^ci>~7!XZ+dg<7s|LtZeN1`JlffkvbJG3hLC^dqtx)+xD>hwW4QKLD#qr z+c$5hq?TdBqi+Msg?jrXDi+JCu=he$Mrn1g91u439h=%Ha>@$6Po%e)%gO2p0Y&N~ z)geO%RutUOftf&{wS2}OGtbwz;;$0v#C#+VUMUw!i|1IrbE$HLCdvYR3gQZiVA7JP zF^%duxYxzQisMT+t4AdQg>!D3)J`Wp1npe_u&XHYFxn~SQOScuSy#UUl?8~e0H)M| zPwHOaGA`AD4IjLA^*cbFEWq;5BT%~#?2`35o-LoODoa8}06I#e zdTJZIuNXy$7wv9V_n^!5Frt3$zgw92=IsPbSWhHiq8%Sl%r6rSCj79`a6U8oIs+<3 zhg8cylrPYpkxo=rI?;bF8bf~M>rApdRIz#Lxd#9XZ$R7T&4YMDI!%eR7d;bRyP)1# z8Dn>6iTPUo0^)7#91?hf&MxVBUTqPOVYA zLJdG8waAHDcObpY7brwH9ylEI4Ki+u&)I<3L8$Ci2ue-RXQ0=NPKbUA%nJV_K|f67 z_e5q-;_=<#Yg-O9=Wl)LXp4G==jbA( ztm(@tmuiU;fKHCu#GqQfHvuS`b=9HfWnlS$|@A@!qz2yb(`j~tKZ0M4Or zK_FuT6yJk*aE3nMVs#`^-}H+{=lV>g6|Mz(*V4 zX$E*2fOk+%4gBUdfmc0`?t1)sp@&;nu)BcxM`lj2OK`D-&y*)D6l?qlw`tNyRKk4L~W8 z@rF)nUMJE;7M(A~8IQ8lvdDq_xyrGh5aHPsHW7*}N~oc44Qd$8TJlaz79Fo)BDSR3 z*)VAgL%wO4A9oU%J4paguUr95*m0mDi@IJXv=?1wVowKu5Tm_MULys?#L(STu|Xek z0G9tBn)cqv6xy5i6lo9C^3`-rWbkU#NA^OkylO6hk#2g*eh|q}a%?L5^jJ_?sPL|9 zrB2tNP6F;u0oM>|7uhb)x3FT(Y6AM!%o79*J>Fd4417id5lnDU^I1H@(2EH)>B|{} z7mRmeXAEFk;)W2uz@^Oh0*nXt=X{)~@}G!#FVHs}B;HU36ov8(9ygT9XZUH0H(~$( zblgKR)$g1TICq;9V)Lo0lr!nPk_;3C-)$zq_;gatjL+sb&Sj@LkMeW9s>ahZrCR60 zFnQaZ2+U!p#c@c8$DMcJ(YPmsb)ep4SRMCJPe8{Mo8={c6Q(^J;S65GJherf=_jP zk3Xpro|S$Sh(){yK&j)%=!BuMY zdUiddTQp%7BftZ12?#}<+zTI|<49_NDq$|$gpmg^VH_`#&e4M?QJqh{2oH;pR=N0q z5=wE^8XwR5!CkwmvDBApSvQz4FYszim=j|5ASYn8WyeI>I?18!m@Xs{f)foBDh@9TQNZm|6!u5?1`~+C=impn!3YJ zMpwwO=;ITZ?rZ?%RK1Y`{FR(K^opLNryl-+CI*Zi`GmuEZHr}Rm*xi3>+IXW1{fzo z<7hl=CbNONacZgl0&cptOT=jT7Y=leajEYa7892Ss|gY|T#2boMzh*LaR#bjzTJC> z)P}QJ&f{?=)3tpswP9rBYR>$AGVJ^L8;9Fs0~_MhffezU6@Tta`r7A zrsAP*@LAnLq{t5+EN&*)1GJagCE80G56BRc8qCeV<^qONB|*qg7O_~@v=!hzqR82D);_a`Kc%p6kO;;A-L) zm`W!Tp5R?QA1>6M3Bwyug0YSM@+P$gsU8z~0;k+{Kc4z9M=!!T5cW`LN%v|L5su!nBz;@4QHxcPMwj27X0SZPYVOt$B zU2hPMFVsU8dEs49CIr$PrMkyv?qfovGG6J)~DY zMOYO2=WA@%EV_fT-Kk+`zyncTmR%|VfN^CGXW$w{s^?@R_?Y7OLA=2k=-=k|pRCM& z89!MaTcoh#iEMAawD&#Q_NIL<2%}8&BVaceoyr-gK+vHAq*p@x(7$2%E(fnEr?h3Z z*6*Pm2G4ymz73bMX-kD}$~vFgJVl z>dkNkk4H$Airb-fd{QE5>H~&sY8o;TV9mtmD&+}`f+E^B{)Ju~Qs0}!{BLlEN;26( z_MvnJHYp+Po(6)Xt9j5@{o+wwE21S%hw z8vY1(UX}XmX8d-?Z#{S;0_&h5a=ro zk+&c?BZu;5Z~h#~pO0sdLZl}S1cb=+8Kf|I-iyz#%y8Jx+n>bq#5n(%sUl{X&CGJ7 z^QuGXZU~NWID_ZnwA?xF&@jlDl6)7W8jra`ZB^&5FoK3gl(--rv5yO!bwOBrw`RLA zVhP9chD}Q;4W#9m2zj(}A38wB0pEm`4Q(Ej`(x#;5OA)*GE@3f;*Y3zhd}!%6>G3S zynRraDhIS=?!O=(Zx!(-64dO@{aM==sa z6dfiA#+s3kKZEelor4$T$Y9P)8ICA8#Lg7<<8=H~M=UKjJi#&9_*7Z$3Eq_JDc+2U z=q{_@V#piL1{vdkN*K+WSGz35Usu(|9oXc~3%iR~j>?j`17pw$QA`^c{a#?czi#AP z?!XSf%LKft*PX>ExYFqiBCZ-d7d;ODd6U|VDQ-wvfU&QHfr0s`OKcp|TMR?}U4pj> zLYw2Q&=lq?dDR(YAZco72&#wOgx9Jk?`EHFV3d~U>(1-r1O%i)ToLnmU)upPZEyyp zRXv@>&VT%6lJ)eH}hi{qj^!yXQdEOrLU(K+SuVL{^ldxVb>L*)!0 z?N$dy1pJT4RWmT#ZI(I|b1xW&afS{lhyCvi(Lf0_B8Hp~1K=v~G#Q)92Q2ky)?SmA zkt8u1ouHmbr~9R_ySNS3ssTsvG!e`RUaR`xRyF~^=Wct_tAO*^zl3=P!?u=~!L9(3 zYpWWD*9^y0H7*Fj0|Q^kfe>(Yj}(ca zy8`r(?F_xUQMxIbAVpx<@!aBXJEo@;Y$1KvlAbpra@Dda4t8WEjIID2OzSKk+0_}yWQuZ-H24VmPw?CfStcu!@14ODL04SDP{L9)cc%V? zGRlz?#c~mfL*29}hxhWWvI712F9OEEYoB=1)1n9uAWw!ZR zd0zF1h+zy6AAJvsNBRz0jnzu%qp#^%NP*sy!U`W=eQ-;wiIkBg_~SYL0C-lrX-1OV79SKGo16%asUTIb>Ex}-Vpr7fH*fT z$0nI)HHT@XPM-+ug1p}*EWZZ|mht}+_IDM!V<3p-D)#GOc0Od*U`R8bB!kXY6-Wks zgF<96Y$|ICgky{-Kxwz-WR{GhnW2dSMA@i&+@X8?lZPgxXus2xRU`@+ps~KMIsq0! zi9OiJU@|rkLsXOUwa8Iu|0f+F@tzH|!%bsW9$qKGfl%$1;* z;QK^A2a9%MD7b?!YvI6E$eUpjVm+4_+{l~Z&RdsYG&lo`oU^v2Ln*Mv9Uw+ay>r&; zG#NsJ(xYhToAl5jFXRBHZUM`fQyFmvp!5yGQ9|8vs!j>dMQya+;o=+&|8sbTp51<2 zAJFxcFve&>G&oM5n|3r{U9yeYEZ=yfnp*}kgZm#Cm{(?qZmarX1&FRJ#U{GrQ3X%% z>I`QHN>Bi5V3CKqdd;uajrVe9*|^>!4c9uO^Q>_U?IR*gKOBOsfqAF^DzOSPQKDI2 zJ&yhjo)Y$gvyGm_iD`rI82MJ)YtlZDX$fgLFbv{|eSnTA=UoZg4>HXsrIOTN`A~~C z2bEWgYy2S4ia^S9Q`!*%lYxFq#$l%&_HmK=khO4C~Reg+fhG(Rq@=%$B zYa3GAmjelbCE}{O@m~~!t7oI78(VyrMHJnvRYOUq}7Wmfu-;tUDY=?Vr6VvZa z{LvZs%6P^GC2=v9w;|%FM@-1)-ak_0t&%`OlRAF$c056W2Aj>Rr*cYHmZ*R5AIXjn z_V_m*^$w*ExD)ygd?(Oa9UPXyS(wtqIR#}YB3Aok2Fo&NK85;#-1<0Fhxwq-Q2N!9 z5mW9>dXynPDk%d)Kew_Nt;=|mNo}wg)8iubSdgY!=LLb?LQ>J0C1|x*T?cODe1u$xLa?bi_GTRdiHjV|zgF;*RsM4!k``X*TgoPa zZKcfsL9d6sp2O6o2vQqMq%|Ug?RD)S76p6uhy?@=wY~ygiRy9`f{8)IrR94MOa;6Z ze5=m95wk+cV_F_N>CII=@e-KC=PuXb^K^VRa*)W*#@EOSyMCxDBoX+SpLSMBKUC!s z`Qyj7P@^!$|MIY=OORwkWE%Tx>a~EkP+0q|eGFF z^&h6|fcoEaO^f<_e!KqNTCabWsedvuDi4WfUHp*lo;BzmVsshtwm-_H_<->3*deip z%=%(`R?yh>*<>M%vn4cVhL9YK(0vIa#FvwSB$#e{!$QdV5a!kgM<5A|7z`C8RgC== zG+WA5{ztp8rSef=W!B+vJUK(YoNACU27A2_Ni?#2CeKx=NKlI935bVR{_Um`tcmAO(M!w7re*=qx`l>US{@$bL5qMSn+xCXK&SQD*|IL;L1Nt zpMRZKNxvQG*@{eL!iPG%hvfC?P#-caVoLns8%<3Kz$skGIhdrd4Yx#i9j~Yy80>S) z_cJQy3}W#b#j`C=y6?%C7}nLV{}cgW`M;2KdagE2MG5D!ZhEl0FF^6DW6AVRXM&Z1 z;gK0ur$rn_z5wl6zTp{!oBr-21_R6gAO;%$0~lbVPi?KyZ?I#f(e3R-UWTyGjZL(;X`=r`|R zqruQFSiv}#)$%Qr44UAH0avhdj;8ndF}2CRFT`@3Q_dKQO_3(7FEzmMwVG^_MzHOlCnp+2H^x6B@kPX{i{UM+rvxEFBo;U?ga$b{IZAT`-FG#Z!pbw;0?pC zL__0NJ=<|D-r0CY2j{wrYsaCFFkuozqgPP^-F&6zW88Tq(RUlBbjb9Kk@5%&B#2{m zkgJG=;&Zv`ezZSYo+vOZ|SU+?luQa@SpM%Ein~R2AI``hlZH zcd7q4p&Zdj^lAzQ1nPiMRXUsKt)#nuD>a%7bygv@^sU(cj%M-OdeRLVKOV`w>b?zy z(7CrdgA-+_R%1npNmOetLsi%{)kqJE946*QSQzMQ7$eY!XQRixYUwY9DdCA=j+p-? z5kDV`ZAqRvoZ#6n7jm#izoZIu(=QJS9)Alzlg4WNd^p75XD#{#sBwfdq;MZ9$!aFk zg*#UM(|IibQ^xNL;#G}ofl0*+VR4PS(j|9_~R zhy7>U*%9E3*Wz6=OL9AJ0AaM!&I+o$p{xg2iSOcx5li8iFm?&e#W(I@)41N0=J=F z>D)L1tm+p1^x@D}5x_-(K*0JHf`H9fJCuZgy+XR5Vgj{(O!I-(^1UkDsZC66#(j7z z!t)Q~b<96y3*?$s@^e-q`zCG%-o^Y2Tg%_4mHZ7{#+sD>4CbGV*DcE5qm}#!);G() zH}h{u%FjpxMqzjiMqwb(gQr^a8YW9&OAl!0+K*tXa9uWt3g_>OZye|C_i5&RH}gK6 zfg8|`8!HV5h4;sU;WW=^^dQQEtASn<#+VwQ7p1Ag!@9C`m&Eld-#&PZX1!AbI|(k` zE?g{8A4!Y+6v5sK*vrfoDZK z%AJhIehFU<9CV)`a)51Qe!zAcTu#K@W#Ax;EoJiM{>W@L0)AqquNhdq1QzEa6?T;r9;A>`?oa8MGBZ=glkK2tBs=Bu2u$?bN|ocYz_1jf83OGNlMO zP{uzFwTFwxgEfHlfv}s}aoP503Ds`SjgN!*5{ng_p}uA-WpUd%)u#>Og;w*Bhl2zB z{xa2phAy@X^7W?B#`_A1&N677&yf@1M49}4RZbE5Q1`3Zs~KT!PpD!AcB`>&k4ObJ zx&v!GfjQW)$yu{vg)46^7V}3sA(Phkuo;r&dkrbEV8(?-tFGWcIx+GV1s7_T|Jlv;}N~u zjh&Hvp8|)jz@Dz)04&rxAH9wVnSp+oXBVw>y8@%I_ml@w)wvGWyfQl9aNppV+n&=V zE#J3%4xZ`?A#bp3BkH*g`%P6n55SeT5NHBGPvANKGmy+cX@{;a!W0x+9 z8wWpKcACb2eMH zxgqQ_?{72L$IW$>xh}=k=!=wRm|+5Le|b%9iapKbT!^yy(uE09Y${%GPJ?zoLpZ9Y zS7C87P=)~U>iZ66BfM(wTEzaVIfVhXP1D?JS%|CVctIGU!pl(zbH0<1^F=%7qDigg zyi4coV{+b_kn?&w=fkb$EY~>?H#tiaau(V-JM{cDI!|i5Zwh-MR^k1SSsHfC!Cff~ zXM}alcf1mBbJKX!=JPk>B_yH4UX-VLe8FT*>f@noH$43^7mR~MAPv47EmRN1ab{Rm&r35|@L(0#+gVpEwYK1c31a44o`OprG0vl}8%;Ds-m9L#O{zaEjFTB4{I6 zWdN^XW7SFqo2TXKmSMH~5eI^)5Dfw3)t(-JjL;?|WQl^xy2nN&!){~PCk=g9T8%aK zmhU*;S~G6!Bje3!^%z}3hd9bfei&rZhYrA9=v)~5b+V_~Y`7wa&)2(}LC$*Bt92Y* zKpuVXJs7dX>+#abDW;+6vgn8U(F-WT%F{|6(nS$Q$5$@g@~RIX;6RlA{N*ghtlg5( zwq7&jRWmW&8VQDmaCg*YSW$HZBc2cnGJks!+A;{kUs-|x7I5Kf5HWc~R3w&)DFmWkwi z8I#@ulf&y0U1ZJJ7=8{9(X8BCP^A{70BM6i;Z%_Nu=_0X`MmHjxz>k!nQISoz1&={ zFxPv`^&)e<*j&@h_16J0{zqICkEX;@_AsunU zZ0r(z&CSaarbuD^+i$}IIKaE3ZFmPK!{bty!pxNrP4>isofJ-w0{!mb(FxV|<%~1( zX<|+_3Zc|J>qRmlr_*8rOUUr*KPDtdnMPYa6>GS145k$@btL5kOfN4JE^??uT$fln!$a zoTpS|SiXAUEyB;8N;S4}e}*8q$R~q4H`Z(`vFrvCJT3`3 zuttDMScwv*ni-6Qk^4~kSPDfWVVK;jF6+#uZ&im5)Yen~OSD)a#l|?w^502VIN_y* zXrX>kv(0hi^%_m?&=n3i|y-@E0RV!A?g?l~$3-N&}v;s3jO&#^biK?}nZ^+mckkK0+eZZbg z%jQVrp`<$IOg)QXfL!N=E6wvLxz>j}ndei?^+a=Bdxp+;gn2#~SJ8XJJ0L3J;hXq; z0kfipQ!o>x0OsSE2}bIi$wbUGnEN*Zvq{#f!(Rraga%V{{lOEwPO!4Cyu;K>uz`mY zUTHpqsTfG!-zMDH)QJ zJ=`LhULWZRJ||(4&>g(6-sR%-ljsq{I!1qU1@93u4|ai80gV;;gbXnW0Fy;AuK+q} zE0YQ)g2p(RKqe7TAz30tg!dQ7LT|mRM=qne(O~(%#@i@%+zW8SV0GY9E@#EOOq?g; zqAwm>2ChL0YRpa#(3J^-19Pj_MSwC733Jsq978a@33)XFQ|#PNAvZIkYh{!E(-do=89$d=*Td{VcB}|>)A>R}khWzB8(GtP>P>b#8iC4qJI)4N(~cW$fb&oj zSnfG4JM8L}7noO~W3Jd3H%}ZqrVKxpUVse+FR2<{ID%IR`aqK_dd~|e*sGp~HY3w4 zMYa2M7z7KYr-XNVAQMbTnH(5s^k8KwY{yJ>os0uE3MVaTj=~T63nnV3L>|`qAot(n zMBftj#vq^`l~Z2iLrbyOrsd0m#vzC7{zm$f;_3Zas8~1~KR`+5(D8=eu-3@(t%NZ` zVT+lFvnsJ`(1*@EgoM6w&f0~UE1!=(wBYkd8U-PH2s1RC(b=*;Ek~5%pKe&=tO>)6 z74i)~3{W({K(i(u${D-}hE`t6Kxicv(VKeU_((H)9>Q|JQPdZ5GpBd%zzTCYQ1l||QIh!Q?F5wqyqDU~!Jl0e# z3dQcA>6(^J$T)t^1oMQ#|9Vj>HgQ!IPCR1fSdK=1FqAz97fy)pI`1{TBoAEjHSE={ z%E}cm@3b;dSiBW;v;*Vs@jEd@2xx+1A3!NB*@apg1qt3hJBou}^1xG$7INzl_xoy| zk|$(2tQU2NTgD;o#ZbTA<=!?_SeMXbG*2@Qt2anBtUAjK#iQlp>3*o+b%v){2Y=xh ztW^xO!RsYUmSm5Qk5?nlme170bYPA)I^a$^Y#3RCI+UGw zQ`_7{CRu<#JJRjZW!^`}J0?rZD9h|XEtqncEG@Xro!^9Ps{~pw2RWeyeq6p#5a2wB zo$BWpJ4gazy|S&2SZQm$*wcdgEm;k_Re=~V6y00np;ED%*|0M9Vn*9b&k}ez+ESsH zhC$G_`mGoR*84hs9jeFz!cZkPUKEYP;tbVCR2d5?t&~}X06Es4C7g)ZYhr3TJM)n zZ5}p_lm7)=o;WiPE~{Y_1ecZk3NG*RYjBAXs?kEqXo(iJ3@OJz=5QqeASH$Ay{*OA z(TvXL&Zmsr%VgJx8(9;~V!GstbHFL)xERJEcD7)=AAMwGquq%lW4iWdy845Y=$f-#e=5|UX92x`9-U%-3F9)J8kvw? z_z^CS~7Km|{#hMN&%$V9aYOcv3lSx%Q0KRww^&Ez`vuqxtoS{d}_K`Wl?&A3c&4tm#mGmhSz zO{8UdvrI*;H#7A_ZwPdIUshHmdg`5I)n&QzXO6VsGRhzCUY`1e<7+xm_5WLvUW+YF zGwu0hCX$RwnHMqM!@g?}Q{WBBkdV~1wm{tBRe>j&Fw|Re;8|P;IU5;CEPoB67e;0Gi z{Mwi}k7WQV=R3T>9gs+9fE{vHno(P@Ek~4{$GIdFy9z`K!$nuf!om0zpj`A|URum@iVEN+!@?&_KGAfi~JsdX|7r$^3Lg64aR`s z!?Vz97^bogH?rv3cucxv52!xJqZIz{3xbe{Cj^cEbKdFaIzVbitFBuO>>)>C~4&9SI;7YH*^~==X|Dk`( zv#xAs|CswNS^hEG{=$}}f52czEBE9#kDmPx`u|`?igRK+>#V~eQLETV zyvEt=PgpDzMv}#@KS8^mCw%8$gSYpSSPSU>Jg!M_c)gP_1uw8*pDyDflrWsW)=(Gk z!m_dE!OsmlOWkVt5?MCBxH?#b(c3Hl-QbB3znm)Zi!6J~7?sj0RN){ozvixpL{9FC zubn6^iVoV4GUzf8%A`RekqU?tUG=UWX_ULc8==NK7iz2|{?|Bxb-wdfd|6Kfby?>< z3ZK#?f73E>wZXgQ-FtMuVO>0(b?FZLxhu_HAP>&9<>M10%rPhwy5^5S)0~6e?;$dW z>5V)ZM@z=|{BOi4Ay2GJoG}h+RhJksUX}y%+P6aaGx0b{W&q9`jXR>}eX~8B*qZ;E zt015k>sV6t0(x-q=!w4guLOs__Xjcr2zz-A#?*W)@OS4(k)gkWZvgHml+n7QD?;yu z)2mC|ATGEraZfWpYP`}i$VF&SxZis^=q=Qr4R!HXkrLL;+tibD-y-_kU2?8!sROUy zheqq>ecIDTc}%kVA$B&x)5t%bi3<1(uhto^)(5h+-rb?r%3Q7IhH4dPG0_cI)pl9qUP(b8%ZH3)Px7d&m%cy1=Ku!&nLSUT%G4zDJH ztW)Dq%R&B8)@!X%TNL0Hu-RcA3pl<<g5tZt#c+jb}#UW$?O1_*RS;ASJ%wX(z9s z1;q#OKKcqo*#Xuf{EVwSO;?uDl^5vBLUqNTz=t6BEakqQ;Ro;c>AgY|(xIKEJJGwn z_Eak!D}Kz315!JtzLV|d{8YXE?34N$Th7%TO9kQIwdvio?4E)YAq|6F<6Up^L9iZd zi1W4KED?VGQYp_fTaAAe%|2PLTu4xI(O-;b~X zmLonuZ_63#L!gPlMS55f|FL?6Lp*W8H>?fuQ@j!;Sc4`hphYeHwXGtNd4aMT_Cyk7 zQ3q{mmg&o`v>>JJ{$E;fD(^VH2i*yLOy7m71?5|gIL`G%iy+K(R_PH%nM>3byTEz6 zDNJ{P@r9^)gS>9)0F#?z4@0q^+Nc&(bSBiYe`h#}p85&Kr$;kjHM*7SdE*u8b>kQ9 zY14iLbX4Xji9vYlv4}VY6&e01;8AA(&b#$3ufv3Ft5jC?n8HIV{5o*f?CqvVJQ|4~ zTJ9CKVB9R@wod8muOfY^OJC2h1(z)VMW5A%PvBrTQNXt8aN#6};&tBJ2iR=6mnC}X z0J+AU(Ov<#Me_mU|!*AVsYSMu=>p?TZ$t!7VB7srg(>0h*mC-%%^`Ov?R zCf1s3i{NLZ!c1Q+X^3RP4c=G3+6}Zpv#{%BS^W!nGWH~IRZ9nQ8n`gxB>kO&RJ;KJxhM!n{^TuATLsj{37K(%v0uf}ap5J3 z6?SJJC=*VwBO>!K*0#N(tl*UUGrc2T8ncGpF5JvsU}1oZr&k0Y{lWNZ+9&{h4UaOa*%5 zU!FdwB1n=}r-*0$mh!7CtS3n~RMG^ZJnG~nF^~Q!>!6@WnBUqeBmtr5&gw10Rb%3) zIYrz@3@Ba#9^M-#gGJPTGY$P9W#w#pA9}l({4)qScIqfQ86i|S>&OjeI30V$Dt@GV zz3O7^{OMqp3y1UH^Q$5MmS3k;_<`O@V&1VlC2+sd^PolqE9L{7(TrU!Z%cS=V4!&g+wrTM$PF;1aQ^|MNO|qRIkx6R(TQ!r*1T*RxaqR^|2YTGu~IqJN{U@ek?WeI{F=A2v|>*KYm0 z&vk`#A-nhRH5z(g?rEoht?3`RRZ3z;Nyjz90B?LareHP{yx2eA?H`z(zWy$1&;8W1 z`dq`QbEz0tu_nfK0Nfi94l%R3i^<@Ap(kkmX~ zpzWuU=kHJZov4AS$rBh3y+}6-S@xhp;ONJ0-#p_6PGhvrkK` z%IRRi9GI~xa)eq$w7A5VV17Kt2%nOhmG{8Wtgtn}{U6sqON0dJ+Zst%~eSJn`f=#we|7~9z)a5clCABSQ) zscopwOVbsm#VHZhWGU&9#C@)B!amdqwM1WO!8pw*U!6>uD9|d9Cfy_8_UHum8(25Jq-L`o}kAEbwkqvCNu? z-)Z`D*T>KIrjG&G16v@k-)|jc4CcQMXJLNPpaACLhYaRVFNRdIQP;|YV(KD!^ zQ`#Jmnp2u+;e?x9H^!=%jq?IEPsfDfF6C5Gz^NvtWcSf;6zq7y8`A?gnq zR96SP0~H7iJ16Kl+2g@-vIcc1J9f;UIVIR2ul6VMLDTVd&t}cF#!VEr$uP45YsKhlAhQ1~BDk(1vi?@8_>EEVHMs)t#EzN35b_9HbhO!CW6bQv9QY46@x(-FfP_i{6Df(yfivn(a+ z!*{&BvcgZ^pOIQz`Gi@*J!s3Mhy7)(jwL`~(~nY)mYCj z@DkI6H~EYAPao$NZ)WYjUG`l%`nSd>!fKN>8ct7tkWn`fkt=IchmrT>8WbBaGiC;; zODYr7G@wZ0_kxim5#Q(|G=g45`YC?Q1zfD%Zc@)szkuklOs+qKJ;9WRxL$(Pw$G#oD!+ScB@ZT|~g1TKA+Soa31s%5Mr*xSr{+x___oqp^HJByE2W|+VxKP!4#2UlB?xNs zS5oNg`G7^1m4RnbUXNqf+v3PVH%avbjJ%&zC*cdG%puU74w5bsScNyvGE3qVm6@i$ z_&+MOf3DOfmEu0?|Ddg}{~qXpLa_ps+KN966CAfi;sSulXfg(WQj`&miD61_1_2wO7@OG+^n9KUlKJnQNty(#$LQ7Ayj%vI9@Z!U&1q^D)aXd z^NcW#-i!_RTS7%;jYsmgvT+yw7B~J(yX9D403?5<$z{V}O;`k=8i}OW0-oXLI?N=L zW~JQW^V#n6zS+;J(3O@QpjT_(iStAYshjMhKZEBiDXH+z#6liidCvfp#h{Tcb#Dl9 zD2Bp|5=bWEy^Iyh@S^@cCgHoqWB^&!dDCu?imdf}x4*VR1I+q;zjyBhf@cnw?2ld& zZr%S#eZght?meWoobNvn{I0j(h5p9i{op;__x*$4n}YPS%3S(`_Pfx3IY@siD8Dp# zza;pr2+D_f_a_H0AM)a?RKEyPUJQOKg6jVnygwUMv%8A_Hd8$KeJA)Wq)_TAM1KB} z%GP8!FyH@-&slwd|2tT(JpHWnYwuQRUm`^|ip{3^b4U1F!Y_=)K5e%@Y}JQfF-6N~DS zm*8c0$@;p*!`sPMP_w8fIgc$P4~O3`v8skdl2Hj|{G$K6q_i$^B_6q#5{>EG#_h}< zOlr^Nhogs24zJ`!7B2Vrc{ST*D0-z8e~0F^!z{`Rbia7JN!~n!Y{hC?2@`G#%eiL(ub|ZX)RUfK4jz%i7uLE9#za zKeb(18P(sD1pI8}5R+ok2my=e#KAf0D zIP7YJ=_2MQJ`vx>d5xEozl{9kJBEGc@(N`7)o@+_>&;NK(6*8AdMcyRN$uhI!4g2+ z8crTt*zz4vswTsRc>w7Qz!h8C7~^zQP&>WnhIs%;SQ}r;m0h((o6fC`*Y?@upG8vU zJReDpD6A1#Z21GV)NE*@n&DJ445lD@5gDm#>)Ppk_(EmlNEjAxJ391qq;S)@N5>;Q zH~A6LatlC+5O%TZLoM<@P`vt3Okv_AJYFs`^wh!C(+9Ya^$U2nkoAXa_Nth@OG^DG zl_?XyHU6fZPrnLX*}Yus%S-aM)Rk1b=)eMa0MNQ&Y8s`1rDV zn-SlXehpvxgWHLMFu;Y?Uk?E$tUf{PFAg~ z?KXE;EdgRREvxA2(qZuzYg*>fYTxZg#ry6wtZE)zeUR^g{;s91Fm%K`jFkQHcI#6LwVqq+#Xp!>gyz>cOp@$dF zBLoFkp*~YaZbUe5JkSd1K^eDq61x6FsC+r`zA;9q>+CMeex}~0*g?fwz8x-EveHckB;h&&#$%r`fm$BzGcr$u|l>|}29%h-2KexV6 zU$q)jx}q1RbM<;C+Dd`B&pV~e?<1ORbVyZ1C6!ApxA&SwS8F>d z>_j!m9HnmPCMu$TT^4D2Tp|D?fP3RGLZqkjYyTRjb4zZ=lkCXYDaw_3`_&v43H zFkx}3H?S@9{;)4gP*7Wb7)RQ1)NRxL;K6qP)lh?EgSR_aV7Zq5SxPOg0?c=e-i$LO zm#G(_*%+fa;n?BgkXxJ=0k2aK^#(6GT=laU=6my`qe;r7OL7EwckjYR$s%gSha-Z# zs1+3o5((g{(=>v!m3&ulb%K=L#o|t2>?9^)38>=qlY^R%Drk)0#mcJxqeUO|9^OZ! z1mOku^O%xqwb#CjWDy)b@+c7pR%9Eb#p1(23)m1HS`jl*CGNod!CYO3={*v)&(VbFk_q!4IX8?II?Lf3Fzlf}qJ!QImQOQ)L}r1v4AD0qHZbbeAYX;?MIl%yJ5|C0c8E$pj*)*`#gJToZx;I_ds%EtiBqetsQV@Va z1OYg=N?nIQ_Z=u~oy1=*vV24MnQ`0y##H^&*0fShUI*~1BEemEJPuuZ)&>a83GI*X6ZBH<@S?|uTC)I4W2EO$ZEG2w(x8#fnISOU|jZvTMDQej%Y%fKyqk)Bu^P4iqx6Eb=hvCe0Iw zPK^H<;p9a+is{~|Mctduj%{eXA@tOs>gjXTv^#=L2GsEl955<3og1$`c$5Dt5;FS* zkw0FAKOQ9h_;APPj^wlWTujW`TNaByE`UGoR`o6X@j-k`41)Ev`V+O48>w-lUrKVi zWc|Pu`ShO^*0N@NB(Z-aaTYtg5$-YC>mwL)a?S0&kvK;)HmE^mw$mV5srC-wnbWnx zRk2A&bYTt;)saM5I5D<}a7m4aDza5#3MT~44hpJtSCO{8VRj>sBM;pV03EARP0L-` z_&6#d06({y@K6uz7SRhyip=gbE+tLX5=<9a04hPOB~G6zWU{jHx;Qcv4bzOakZ`KE z&XI}pOJxm$3GHbW+Oz32zPa3qX#4{Kwl<_oo<7I|Z0!;0_mA1kSu-q7#~TsnU-@)= zIrkhsQ5XL_b3^50kt!PDD2Aq_E>X{BenJ_fNA!2)8M+lQH2pdIG_0D>@IhnCG>$UP zoT2Cnj^VbytM5oNG>A%WybLo)E+Sno85tw5M7&WHp7&i(<6Ynw#O_?bt8_E-J^||7 zL!3Czt3&2d_)V(D__4%^;W~u~n*;r-OWYJbjXV_5k6Uk6U_VzNb}HCPpFO^Y2Om$oEup;FzP(`NDZPph>5}gb zw0@c8K$gFfUXrVA>#H`Rx?qf9MGtD>PG+vLEDJ?XqubD2e~O)4Yi&Rku8n6%QX3`K z#?-QO$?ZV{ILD#QD=$y|N&f7&J(%qlu|%xMwXR#O)~&N#spv#E2pc>T6GN|!KVC3s zgY_=t5Uf+j<>*~~@=~m*KW{wNO~TY_156*Qv+?l!iM@l?@+LO6AN2>dgZ}}HvvH`I zxbbeBE45?ll7ZLNd%<(+xOB-Q2e?Vr)hl_e#O9G(_!Z2r^}C%s>Yg9sA&}=#OSSrL z;tbb*GaU?y`D?5i;4L_U?~#7(Cc!h~*AmT>DVzR55uOU0P|bqhprvya%E8C2o2qfe zq3Gd4VeRwms-Eq`5dXx*t907v!AjJ(O1>eDBUD5k%L*EGh`z2qT@+pCnwIJ!Xczin zgl|Q4gRi98hlC=@5o{0K2?Vi`8B%ch*W)S&Le;elFDlDnRqP=foyJS=DMC9c$T5by zd5&R&VmBC{VJcul2{E!t3f0#TkP#SX$GW60X#LQeYww}hk<=%7Uk(VqmYXDi0!i#! z%^XMqlrCPyO69`SbjfG?yXEk{W-zGmf}q0BB^EG>Bgn-TseHz87ZCykhPy8yP0EA( z@$7QoAfx<9@&KsB!q8RiybSnUvmfh6K7|$&XuZ0^b;NK-X$c<>xE%1OV_r>a0Q;$c zenbZH=krO8Ahui^;dZx3)v8b|2HgB(m`&l0!WH(aBT6Ag%*a!Jjy!cZGk7fW)Tzi* zN3;vd#Ed5=qEoyqY-_lmL^3LY=8%>qjzykIf%quK!#&4T>h_-L`$(QTf;n7_Omz`E z9hBk~E54ZK=ooQ|Mj%gZ&&X3pAWscDNukcWCMnz&c*K^5qNlKg?c%#`CM*o|8efVu zwKk-^d>iC(Bij9+lA77i)vYRA?K1MI`tSSgSRWyAY#~KV&eR-^bO8ey` zX6aoR^FnxRuoOLsH){Lu)Be^V91`PdN&C!OV~IgrazD0_Mk3F^8<{b*4sMA$Fdb(O z4*30$dHk|GXB|y8OtTM!VAgPpd!Y%oB<4fz(hIp;V>c(}J05I8%r~GdzDTC#i6MuD z<3kQc#JA3_MU8)vewE%8Nm93s<6u}u#=$I&zlNwSlIT7NN!{U9Ur=JGP_r{?hbkOu ze-9dGQ2XZnw01(V9dMiV6m2qQyq7sNv$^^Q z+frMI7&#S+4a1MYAI~4o-BS`ecke%q8}FXWh`>_|5=HP<+hq7En~qzlp(fEQzBJyg zCQ%!L;o1gqAJcvmA98{^mJ!Uy!k`c46iF+G#??wwvpe0})oD91D=DsMUhTCJw%1N+ z+ueH)J8If{Jd^Ore)OK(vUe6VyxU3E=NbGqc-=UfI0r2*DgQCtw1SN^cso`Y^WEUJ z{UP6BEF%w)aWyw9izOTjOBha^Y%F1=u`YF0D;oRP$Jbgs(BY1&`)f~aKP*$?hq{&3 zpT;6pUp2jl(8PGNvPc5fOI%^9@sLi%$9E~dZm*!Yu(ix_mJxYIL8mPm3#cOAXpnfN zQuV6>mOd1HgQe4Pe<(Hn&~c_>6uHRy<51M^;pkRHYCGsmy{vO220uZj1u${iBC5E2 zN-UyXFzr~xTlNv^HNhgLw!|XNG8XZ%`uJ*9s>Yfgj#P1OM|nV-IAl;gDv{2$f157( zsjGc&94D-HkYvgpxhuIvdaoZq9kp>!eC57Y-H0$%Rtpr&#Bw-3s7Q_;z(Yr{laT(G z97Xm0*09~y&D+?E{kDeyas5I7RyXhCUXjG)qWZ!$sX>l^JfWhl>fI1Gkf_Tiz)bQY zddJIndl{f|EuS}YHV8!&p;NADeSf5_*7Hl8Qjq$r;n%T_H zCn2+OvJs7flxfE-wH>omJLaHJY>;R`OXli9^F7LrkNB@;YVi!$BHLzf_z8IV)ZT~` z61~fza6_~Ge?I7cr0V-nteVCsKF<0>Z^vcJKAtTbitS1++lOB9MT{u`KbCG>=h_48 zL3@+4Wg6eifGD)Z!?L^z$^Ea7yTuwr496_k>%wIHx{vW<~W?=K$;*8NK#}qVV7;Z0LX1 zs~8&z#I;)+FZ8z)u%Cqkf}a^yR92rjx2S&bx#YfIpc8kM<*b;dJYMKt${8PI_g&{2 zJ%H{gj5~3GHhhV`W9gY5JM@tKJOQoNc^7S?0B!U% z+sT-=-|+MN(Z8+g5h!_mprr}96uCs_M*^LJ^3Z2h9i`#@CO}R*&)z;0lSVvo;im5;Z)B@)KlN@Yq1NFc>Bm?@BOm)iNLuyk~o^p zjxCu$!d35tqW22$ao8d5C~gDJFF@U@y>NU2pXW8b5^kASc5Yqhg2mdfmpb*_hQ@t# zB*t}6hihHBatYhC)E)hhg)ZnDY|_=?gm_*uawm0Ydyb34i~T#<472kQyAt#=m|t!@ zV6aSq`7@Z)=~#Cb>Hb(l!U-m-IKr_cJs2e7iV_r<-D;=zGKM6!Er}W)1JGThHW)Er zvF=QOi7-*xrC-fll-*1F{lr(1R&-DT>XPhGP1X8|Yf^uh zwcEr~vOBuGW91peZi+-45J;Bz9df-oOemz4WNUr4MO*7HFJl8|i(+F6#U|rP=ikMj zYznq%X+OfB@a$EFMCqwPeW}M#RQThlt2WUszM(eW3zgl$QrR80Q4sLI;}hxddIy_g z_^?fuY=#`OU0rGVxhNbOj)QErpZ@WDqkD%>Q}-rM(#}uke^kbm$GEWc0r$MceYBBR zAtebXbmKHHukXrE&MnOB%` zn2eCEI|a1IbdQfI@9jO=tkfPmzW*MrwZTRUG3iB~Tx@G=e&BE4BTKU6;E0@KXmZp3 z!sbkTbQFSMI!;)`Pq;wGt-uKx+Jnl^M1ISEi<5?h74e=s)}1FtbM`2cxyERXL^t8G zrW?1m`cL{JL!vA2GV7FOkg_b&pwf@H!LHs~4AeUB*zKI0J=f``u{yeC-K*vY&?ws7 zpuu!_f9nxILz2xbH13~j(CDe5ngyuE8p+k$cqsNi5IxD>k>{-Q-s)|%;bap8-gGhtZbd&jOhePv!2ah` zOk{gTpTIS5a!||uVu{VqQ+1D5T^$M&>sjkuJE?(oI63@bN}9wU?RWo1Uh74)n4P?y zDJ01HTulxHEF-LE+P0O}qYQiC`}A!Kg4ed4UFULw*QPxKp)2t6cI5sG^*#5waocps zTA2`|5M^L2&8qC(K%!r`%+&?ouM&Dt!1QBcl6SGd}{69S<1Gi!Fc=?oE< znX>WgxU_4ly(_ey9IberlcH)``h(2%YxC5FO=E!Gq}(cp?+O~{q4iTop%0(e$EckT3Gp5HR|{O0;?|A@n0+Lp4eTZVXtb~MM}H* z{FTx!*Bh;Wr@za(>TXeg=L}66JY58rPY*`T5kh9|OzOJ}o4e@hbS~+vtJBo(L3X&M zjGf6>WNYwdmj=2za+t2JXk=hwazdVvBJ+6eDuI}*}tqv5o zk>2q}l~|CoyDf@=^;BsAex)Jm= zUs=bx`8V{x!Fw2%ZZib+PQc}@8}Z)|)Mx1~9q#23*f$iFooj6&3$*W-$xDN+99z)(`_ z8dlRgH-BA|@U9YIm68KTPHm-%3&#iRaTYguhZ8N+sUVVg_(A)**P2gY;lZ6Pz_Pw- zP2<3<{^KYK!cE~h9X=NcAfmCw%!}+HeiVYcuUak9*u$=pIZipnv4#ec4=)jGOr1sE zNI1LZIU;~?3oXi}f0SFUobiWODd*xAFZj^7ZB3#NlIZwg_=<1>HnX}oJh;cLrC?iZ z32C9HOtFocft)*`_qpNk0L$j0ZSUO*NZ;OGE-u$gD0VM|%MxC&1l^;}P@8ABtHDq- zA}vA79CO#x=*%k}b|WV-tkFjjsXxio7%M z%Loca689QJ5*3legdz@wO~rm(SGpq4`8lFB4jwZ1bAq}^{0meb-h`uNC&caL@fRc| zERMZ#8Fpby8}~*OoO;#COEd=(@yAZ#xMpfUmtG&=$n=Gx`_M+tCJRZDwvnBwSn{?a zbwC1?I*vQlUS;uyH0|`&0+swLoR}kAO}Ct_7lb_9h?6byipU3ynjOTB3q0!N%9rI1 zLa{&Ee6zeObTZ#7HMBN2DTz#heiZ9!wzye$5AO?k!@r772uJZPdY*V;qa@gCp^JBW-<}H0Ad`wI{c- z^I&q5K$R>x?gH9Q4t(OJ58U`xaLUM>4}t4U;T|^dI_3-Bz^zr&aPw^b`Y@J~-*G8K z-1^FWZia<0hTet4=&(AK_8RZj<2vtDdJ>qAgrfD#i5N)IYn3$_#P&@*X|^DVn+Iu{ z;_fDUeO)G~6n<%YFTB5ZJ&5BZZaUi(i{WIvW!{EOg2p;;Ex&#s0!GBhv7 z!7Q;Qj!KHnW;A^j=n_X}tRu_(J;@c&Ti{k0y8d`;u4Od905stMCY91EMi0|B=BOFi z2vT5aZTUDo4N8?};Vq%lvu zx6qc) zI`29Cs%`H{p6HlxcPwEHGF?LrF3^!h7tglRR)II2tQq++>);=kwE}?rbcUAo+!|ZW_|rE}Vi5$ZWDivotOKK+2pw1l8s3}j z`NQkshUb0$Pf}^_6Mm`DTcP)kTSv@p-E5gNgJxLfPxF^Sw8c6P;t!x_Fo*~4aD%u( zwPj2GN{Z^-Aeyb)oDFOcx90|Nk}BHVAQpIiIt+p+CvnuIXYJ>G3-tL9+0y9*}Cfj??#)%3wv zlHT#0cv_G%u>L&lsvzeu(sPSDCd5f_tB%1qSdiMl_Xvsy$IMSJ35xf$hB_ZJZ$B+y z09IU=J!aMxv(`9h@4BuBNsU*U_n>Q0#I@90u_^Lr}U}ba-^Q-zKqK zJrU!n5SVkL>g-zh2+>D1wYcRT`uY&V-b}8Yt%*g5K-6E93iDGXT)=9X^O{V-kiEFM z8XE8P-PJDhQ~TN-xxTI$*-S%^!Jl{|Z(d8EIH}Ypbj;{H;HZ4Ue1blZX|ew za4k9r1}%8}DUWKd(AGTZ8N$-0AA-^p9Ol$*H`_KV6^5r& zn@=k*viTp^;Y2HCzpiS5XF@Vt!5bxDOadA|yh!^%r6uza&U}YvVy1)6KPD@5f2K%n z%MZjM8pV^DAvtzO@-K-0q($8;z%G;`k@zgTmBARJs(?a^ztq=pBG;|K%h(9=%n2s_ zjNqF*THnNv!odJ6b!#kd3N=2Cndwic?Qp`~)+@JF4vRrmcI@q2pxLIrNYy&`t(mg! zNL&zn8y%@y?c)57b01cx?*JIA*IUoFsSUh=UQ$Vu-&|bn4M!7;aU5sG$z|H+CIAL} zde&*C<?ZaNU$Hnb}k1jc>jTV0K|i9({AM(@Msuw1%txFEs zDN^{6-;c~Qx^$)K%m(lJliiZ((8nfBosYQkZ0!g4pThT~LksWwFSPd4-k1Dp0zKaz zG}be!HZ#`7l61-NUd#w9c)9+(s6R6eCAwdkFFmqxk|!xjl2|L!6(dZ#s~;@ABn)qbjfvvW(Tw@ zOH1D8Ro;{znFZxkfRsi9e&u7Bi9bZGrG&i0VCd0E{DyYzd#nvMm04e7ccK31&mv@gQXj?IgnfU*gkF{k!7R zMVs!(!>5|>gipWW$N+VM2j!#T(^;4O|AS9!HvS*r)1C+ZZ}I7~Q~z6h3g~y%L%n4U z0Yc3{T|NqrYJNLtMs!)w-0}P&&X-p67O|=uizA|YWYCPT&3i|C+JEo{Sv<6Y;&3zs z`xsiSM~KxpK9!kD{IhHC#_$bF233QQIVDOt<1#!mjDu*!ijE?#IK6b$VL*EuxDIZK`W+SROAO-vA^^x=mopikZ@v0l8{uv z-T27d3H#>twr`>7wvW2aiXwT#4)_c>vC{hZCX^hF%Z<}tb+RT$75d@C1O$4M5DP8x zp55Qt8m*1bwuzOwy8MHPx;(?}&*7G-6()BhHVY2wGSW-zKwJbA07MqK#A>)xx6kOazLGt)^RR#VcFk#m zlabDo&L7`fK_E?Bxo`;>ijv`}8PGHQ?lZ#ON}2pIUlHW(3i~q3qyeYGGr}rX)Fn?S z!)t^n2X=5m-s&W*p{TAoRp^rrQr01%=uxBxJN(OFl-Z&pt`Ag-&VQnbTCTw14q;Lo zbB_yjm@R&3-A_)dCYsG?Ef+$AjH*V~OJ~EgGq&+MmRz-?8rHUYz9FocPhzb7WoQQ; zDK;Fz@>g~7(~&q23#0$6OPpVA697ZibiRKOHD>f_EgL!4JZ7yu)oc1t7_naz5nNzD z1ij}}gp-qj4ZOdsWm`6K$%gQgC(2x#EmKJuLcuJomgl0C4Ld93#bluxGzXJm7Y9>| z%7!kvR>p|a$I%P0#SRYxd2aCT;Q$(rZ)FST7xd&KJ2qVQ{RhnFU<(J6#RoHI91Ru) z@c`_-QCJm}8fv8`m)Yf`_=Bm%f2u!0i#q2JN7-8BqyqL@L59MuB|S5Re@u;rZvZna zA^i|Bsvs_?0U|Vb7oVu*Cy8OYSu~4d40oE7i1)F8&q5nzS=3xj4uvcjn_Guaqto_y>dH*qh*Gk!WJ&WMuS2BJlRji0Fs-cTuV65_;!R-;kfM&-~&fe;JaR zIFAk9XeA_%IDTv^xh_A}UWVJt`FcS})L(x_HF)E`y#pk@vYK_DC+}JG1vHQa5ccl* zMEOCtx1o~8>zP#Uelj$8-*?lbt@JLpOpJaenZh`TLK{~!_OloYr@L>$5$`yjkqwB- zddjJKZ@EY!qnTFIe7jLkwswCalP{T+^!fh>^!LGiwoHGovv2C|8gFd;0s=G3wN2=IZ<>kIqgl?X%7-$G;JEs z4c<#2weyMO?8{IpB~|4tiYH;~Hy^OYWegoxAT{@;S-%!-)$_J~vvCPetb>n!^w`(LK04f;iOdT9U?b zw-h}Ku0VI;sU`j@P@&}TA|mtZyZ$Lk?|S!W&qZX>Kka4seR`qjg~;`VH_cP3Im@5J zcDbsU?}=fPr1*@!#b?3@v$KAzx@NfI1gE{Pyjk@mK(nIhk@AD0I2kT4A_VKC;;lll zS9y{uG8|5xHlDp-a+S*u(CLYKlfmVjirHM*;C*zK#R+evI<_m$>MoKo11i>l0ufVa z`)4gKxCUBb1C3gwZ78q?Q|;|$)jw;&xm}3g%U)_vYgq6T+fq@>lu0}u%3~02mR_y) zK8AP5^s3E`$a?kbUv!9G?)KVGt#?bvMlWoR=-qN&)jOxvjaqnoFKN|ggUqgdv0b6T zRU~npjZ1L?#tKS6L%Ix1V`59hl`y-~P)Q~`-MAF-i#+{RJKlh{ zufh)_09S^>Nc`DyDgSw&;NE9hSYX~~y7x!QXY(N0jP|A?@7|*X05`RY$2(m>0HS!J zb9d9NW-{Zt#L`DMO;B?ye_eUIrXl_#b6q)25y#;?sK`FYhr+Oo#PtG?b^<|LrO01b zsVLfz9^MB_T`X#KCM$Jy0FWWlwV!DZ1W~dD9u5jjvI2n!*C1pf+>?R=7MXt@oh!?Js% zd^B|uE=dAyP)Bdxsvpbb?^)Me)k-%~>x+%_NFP{|x0#l_S;{LOYh}I2yYX1Kvi)E- z1rQt|#c10PPL^hvT17v+9|pVLx6@vpw6|CbYuW zxjk`C!tLk-P1&!x^Za>|!%5VhRK*pdx)o%S2Em*)^Q(WXJbksO`nS7mTAVx;hMt0O za$&6|M_8)fKgfsa*BPKPYuQ`1$0oXSA$MZ)QGjaYH?;kA7Oq`#;qB?wLu`ha()FcB zX?&i@rK+@Em+))S&kRYSIjy2_{7ZTwVq-0jR}zG#DCkMMiZhS`-t~{X`ebPpxwM(C z{45L1r~GY7;2-(bvgP-3X%kdbdLDJbZW{@Hmy^f)W_KH^yW3TRSVd}R`&-IyCs$^u zbHBEs{vOl4V5q8Jn4>;ti6q0Etx~9^HHJt$AboKQMH7N1+$L4zZFohAT)ab0&(}-lX4{t+ zxa_tkP zmMYJl9~GQb_udeyw56>q5+^nv{(_1+R3AwmjcZ}JJrc)aOumDQnqIGS;e|piy5GwA z?cGo7IFhvCw^uZ5dXe+nTS}tJrv8*&_v%7FLQ&N2Ql=HyZk^K8Jl5c1%}QVO-CPjq z=wb^GaLmVdh^nfrNSG#uxo~K3pS$?IU(^9mp?}61tqb-pBbr-Nfv+3wf_}`c2G|YW z)2r_QkygV3de<(~jLiyowUwIqYTAJWg!#cf);z1Ocp%TS%CT^H^mCbJAUW{1`^;Ht zQVB%Xj&on%k&!RdC@cYCK#SUsjq}F>i2jW&$y^olf`Aqu8}D0OLui6EzCxqSz{NJO z6w8p^c;9mQ;l^Z4qm@-gFZ0RaXx{PCZ9mg_VAS{m^6s5AqPfymXv^T`gj+ z1YxAL3ZJ@5JiMU7;Fg_iXLhKk`z#6=vj5H z-mo4uj`2YnghTW!pPY^5-RqB`g4X%%_VLpUNy5%t!-az-UGz#oQoh;tev8r0&Cw|V z5J^{tPQ6y@s$Z^R#ADersMg3Ju=kYCj2PZ$XpiHEvMlK-ZH^SFYVUMoX|5@ye%j}Lob^>3xKV-{Wq^fKU zs^n5O=S}0kgRm8OPKJOCCW9j&^LSD$lACGFVwFuUHX;(eiiG@VB>qLpv1hZot?2sx zhg6w;XWI^T@c*Qb8shCVtQ*5MgOhmH4%5tU=&9>$kj#`{VNK|MJK8%bM@Q^^aqhjY zWkKP@lk-S+)q+T91c4Af?g73N|3k+{QDv=?qPoI>#1}61c3sA?(+D);v`TU7u?Txc zI9XHXzsZ9;57#}L@fS`#F1~CtaoUbYnzE%hZHMu#R>4Gu#fxeak;;w3;`|YX_PeVw zaJiI``_Eulm`fgg}_@9zOQU+B%@h?XPgW+f|B z&ye#gbPav7$D_Z`kn>?oOXr?Me?w2wC6%iYxFR`WVUK1ddez*ssw07S+w1Dvv*pcX z^Y*vYfWN*A#c-GUnr7K<@bqcUw{D^81Qm+BV^xnTyxNYL`G=F}vkZeRZxDUgab<=# zk(5-NU1q@Dw#)1gqA8sK`mdEl!Uf^o2SE52z%i2`tFnazZ>1@O0O_WOQ;(_BB6Qxs zg?Ut?B6ty*@pcPdgaKZk;6+Q`E7VJd&-r)`FY%NwiI1B3v(6y$NKWBLyNXVHpgDG( zn4iKcHE8%*@kbi+lg22yJ>C1GrDkc`HL9EvM$@hhzB!uq4%K=MeD`Hp5KOELe#~L_`6SAqfMfN{HK%z=U*fQVh&q6_!srY zah@Cv-m)D6d(i#1@91Cj_}Et0_vr^?XPNaIWH6U1ft?~bFm;SpQRq;;gCt{qA;Z7G zzv#&aX%_?bnQC7CMcie8*qvAtoAoaeo$Tme^yK?@;9s;+{zc|W5KeUTFPgJNV2XDy z_pU?5yE*rwIm^iriO*T3jCt-wy=w@_DF33pbtKFB7rpZL4>%NEPa6J3gMbcxL}Wdo zKn_OQ*=>cLv2NF#gV7G_^%9D`A)%bOu`?|}r)tKUbuemC;TChODZpi^Wl8}aMvGvQ zfR@0N^)R|rEN#+27fwE`wPNCdWTH*Ll-bT_oadj06xNpa*6DMaWicP_oXZZ#wAOof z0!ugRu(Q;-b_g&%$>S+^F0wvMS1r>P7^l(IVP{6x-)N-p@qGTc{eqixiv#TO`*if< zx#s7s4ttKN?1Vj4N-*qsG;B0@K!80vovFHj;c4F(P@_+P_&GnGdwC1|cuxO7`&`D4 z2N?vYcpwdiwtJU$DC#o!-*1wto^+Y z3NdnyImCNf;FOX6fmUUd?qo&x^(kq#@Xt}e&#X&lOI z4ks_G_DsTnXJG>;T%@f*Bm#naF8%!%+WwE)__vQYFv^DP=uUYA5ah(Nf(tNBF6 z#P7%_=QSNy6;A!zjHaP96#E%lNnK(T8g~WSLlNy9a=zatBhWAP`;*1tx8Q|C&feBa zXNMeV;6u?~`j*5RRm2T+(~+8|iqsCk8-9&kXRD_xJNV(eGyy*xOh^kFKZvLk(r+S7 z!vb7raQ^o2ut>7QRr0V(9)YjGB~HswC#dlj=CC(9;j_ts{}*Y2?gGUC1BPr zq|JyV4?KQ&N4$pqXo?+Pf1tn!w6cH_88R4({vMhs>3Suo9CoOjzImKIU}?ww3Wg(F z9wOI~=Z1Jo{typ%KW~WZ6#s)D%H8k_y5oFlm~wAsvA(SKW)r)`{|)8AW9Zg) zKkEpGPrYx2fo^^`o}@~?nqU=g^5|`-B0n{ zZhX}H?8ecF9F5o%@RGSOcMcQZRmzz1HZmED2Rm4va5;N6~c>W%=!KDH#ZVaRYJ~H#yeI}TTCPb z(mT4!9312fTxAZ<lDKRLlw5r+o7NHkm(4`s8qt~l3SU9E}(9L0lelQCJLr?-@t9-R_%JuZQ~Ih zgH%(7kZT{{+29SL3tcO_!$#HZXA0vGkffuwaZ`2M0*BWJ>AmxWH!k*16*a=I!^r4| zd35P?3Z`D^QX~|EtfsyD$H_}2@P=^?8F!79L&g#SlX`+p{>Vd(cYkdmCm{9w5RN-4 zfd}^dR6qXsGN{t;ZGb7Y!Izs&ibLSy6iQhN^mvP!9#fAhyw~RoR4_HYsNOnhCLLIJ zyr%aqcziYR!?m6!Q+K6)m%34H_VRBaHTAU?&|c2Z&bv(Yb}1fxl#Kq-{GlV>37gM@ zcR<=lOOW;lX&hZDiFMoNX>C-(5MPbU5vZ(bpXgYV-NNbSgP#h6s3k zgO}a`{BbNyBr!)Kqx!_Y+M~Hm=>jT2yyfgOm1kogd__1%8aVlYO{764PQuRg0DC_gC{A+v(28d5a;{T*RB50R20>H`nGc32N!--R2BM{S-il!z|-Nsi2rcB;G76l0ckQq<`f;jT^zG6w{D?@Z>Qq!Y@KT_f)< zrgzGNTP=1Q?CoBm(elZ{RPutj^`A@=QmU{6+Tx2ANl=UubO+S-CW*|Pu8p6riS%YScA z_RIcSF`Ka`JHaaa|H7VZkhAV}_hBq>=z(whcjOG$}9rsJV%MZn*JHX`Qtf}b__D4z=~`ZI*@vgfjt;bj>hUN z9F9+^4$GdN{h`d=skI;IZRTdTP;JKC>?)q&_EfvB1G77|M(Y3vO~=^mkk`aHO{w<3 zXFbUt7UkGF9#bEf93byLE$&mvUF4MUZ_34P5!nml59pn=@rU$0Z%?|_*Ds?-R!@WX z-rf+;DW1LGp_K4=}Ba6G88svjX!d+Y|mPkYb@BF zweit%{dxqpXRg0Ff3#_9N}1kp!uf}mW%zvU$V4T~vp*{$oAUP{hW#xjf189@Cz|Ga9{*#J1)GFg^?#{)YlgfzO% zL4r0jz-aIW*WYcmgra>_Lt?1RA?5zdb3rpPYGZ?;Sb-Yh53Dv(5sU(wW=aFmSbnHB zR#5X$7SRyzD8l!mJZf#pD38ogPRl8e=PT6QCI{}VOv#dGzX78)y6N!h&|z~ImO6z~ zfLX_mG`K&?0asJkCa`EVIOH>%UOa}ce9;~E^x6N7c=8NA3cRp22ma-JkPg#f(K_O= zhm*&+izFW9d*CuGEt_11&oSp+RTg@xd*$@*hxym?Y}lrLny`=sRc5lgbC5O&2SilJ&bRMVQ-HQ9EuiGZY>*~5c<5oL z{p!%8Q0!O03VeKdA3H(6{S?)zzAs?M)y>O-%I@x1*=d`vY27l7hyWa&f3qXId)V zAE>uK@RnTF>%7&_|T%0tmbkbvRC3YLbIU{O{0L?~X5jK<#j>1@#-Dn0|1tOB?r8 zOusj{KcOXv=XV}R<|p~%_#S!c={mCush`qRcp{Y2Lb=sZA?l6kWB3$F-YZQZSCm&9 z!mK>|?g9??+!_&z{?http$&8TLAFeSfz2jcPC?jes!&XF0Ce6)v({!P6VXqTf@nos zOO61iS7Czvb%4cx|1!YhAfDgU!vicnddE&!JU|JC#rGc|EM8g?U~wM;UD{-x@Sf8H z_Jfw^^nknZW_m#Jo}zV--ZtnI&iG>%ApUG_xc*`=mLw61fpcgbVZF%4qQE1NfWE(C0P}cit-XrT22@AdAm;v#Ihrs{>tpzDxBjsS^%+IS#M^d(O)i;`0r+ z%5clICHdqSi_g~#MDyeG@sab)@OxFai_e&^7HlJPMlJZ!`W$BDpLuqDNPA$w>&r(x z$GGP+_*urDtXf)juz4&Zc0GepW%Du^b@iVdMy-8Bs|Kddu*JKOWCkNVj{X4tK3m>Q zO(1YjZO_zG3g>J|-YiP`Iu~d7tK`6K_wTqaHh4D(HMCIE-YwW42#2qr9BXa5TNoEX z{4=zG)i8-v7IFIK8TMeYPe)RR*!fj-#QNqU=6PFb2wAYK{;$At+sJ&dY>mq>hXU`S z2OUt}W4VcxoT>MY+f3}Y;&WT(Jm|WgQlF9oXASH;@ZWrq9r*8#aRYDBBF>ha#I}|< z@Y^#3pP3u@an!N7fp25o%HYF)H}JcU$sc&}TsQDZ54eG^0@F0`@p|tx@CQ)g1)CKU zRqFcN;2m^26>?HZ4%`k-*}EGYmIQjUu><`V#Y%RK9oPmGbu3w;-*_%|;On0Yi21Pt zkJ3`e#SUz9Os}_O9k_^1m*^t&dG|v*PRFNSFD)&HhrR3EbbRelT`1|V<7*w)Uc3ao<%txF-LmO?`0!tde3ip3f6~Ebu3WJ+n)oyr}l{4 zj%=2H#oj0PyZY;`{ws*L`6emsSr#8~E3gN8`b;OKXWN=!C*L`xoqR%8NWR~Gt*jl6 z+LLZJH2N|Xhv_Q<#fqi&$)wehmRH+ z@QA+RR6g5YsNHEQni%kfI|$&iMC6Z-N`AAUB^J5f5?db(WfL6{87wvl97elO>b+BK zp;!~tvccQN1VtvB=@1_9HWdQxCfX|NQrHxGg4=_K$k#hzO?Jt(gEjGh4rBD!ME-;a zYzz8VZT2WcBvg37hc{OP;Q?P0pF`nJ;Q>29E)@F{xx5302H^qw6L_-gfL(aNsaD@& z|80H)HF72VUhsGrk3d!70jumW=rzI~>4CoagTwdJSXUxJpXj{{WTXF}Zht5Ju>+%p z2Gl$iG``Vsdn%!x)bm|yh`mSS-l<0uB(RtEG{&D(fFc%O=l#vr+3L)CYw*7O%jPCI zsvD~OME+2mMta-BrE~e|)0{icPz%YDKb6>#`fz!Oi0Qoh1vhJZ`2SQLf}4Lif7F?s zeN6w-{t9~~ZxfAYLIf6jul?RZ=`c};uFzRXZ~S&~bNw^N-Oa_bUt$Dqi~88ve9q`& z`~OBG$m(M+RA|C~tB-w;`S|bjv2$nrPxY}AI_hI1(oXsWf3>K5O>(aZK;t<7cVX!NGz(8z{` ztVGeMR!zhu zD9pW^?{O>ZpRBAvfdG-&=5Zl6-?{v@+J8}KXQ9j{mt{4#cl^V#(~DDau1A`s{93U? zsSo;Yn%muhRfD(6sG+E%NIjfArstosBPlzY%8z->Q{3KMrhR}eeB48OQYXdjYf8*1 zZm+S-km;u_b0@{^AA-!m4)d!_=4(l>qh_&`XoMvoT7;*5FSTcVI^(B~oOQ&2=!0}W zU2@#h642#zw<5W8Nvc+J?f09{u~M@Xhu}1mc5S+_tPL0Q+})gv^xXZ zxWy=6xkN)>U z*AF!;SUjtR+EJ1 z?)tyt&f0XP;8!89qPWl)1zMTr3A@Y+Of@)c@`;z?LK#5-yDQ4W8pA9^!MOD`UZ|Jb zZbpQg@wB-7+V(W`8U|WT1C1RmkObFQtJ!?z{(Sgj^Jh{Itc@-8LAm_Npw|p+w2I+r zO?=OEhmn(^52eGVBB>}*k(^d^Z= zS6-yhpLS+xniGD5cjQ^hP;6h0<1UxtvL1~i2*up@t^y0?Yx22&)n`Gmn)Z<8yaNRN zc`T=Z@22)MI@jK<@ed8&bGw-Vhd1)3JXrRR6TboT{HkFE7#{l%3fc zfHA#MYo0-2MgnKE&^z`PCt43CqPtB*<1xfZdeY$KV-`fLiT&1pP@pl-vE8Z*a9e9RF7LC`-xfaKntS zWcMQM4fX3@uC2027&Q(&4MX?}ntm!2mFMqAUJ-nSdGC0AtbgEbZqeypqKde8%4uL% zZN^^?vm{kW*_UBQC>#4*)7;cT`V&<77fVh(DZI#4*0_7RZe5_p>f;hNV1WDOpdayYwm=vZhF78`UqT zZ>FnWh(?_eI(K0whnLnY>RDKqypkxaTv5MxcrhK|YWm@&1>Ps0(rX%8LJA?e2FYct zhj)}UhLRXaWM?!KJANCr1&x4u2%dRt&O){LAHr;;y)NDTy7cqKlxStn} za~W!T&v4usu!rY;-(95R`Y=Lo-8nAVAzR6vPKS-}6@c`5WZOiuX zJ2mNHtkP`PH^KU95dTHjUs930u8+iLtwY!CNKe;)oZs~xbX45|#QzU%?*boHk+csd zkU&)61UIA6^%68nqM}5}N(7b+N#u-9BnSu!DvA;mRD>Bo5eUqTv~&7*Iik4 z)%Aji3JH(|P!ZHs6fxk9a~ST|fWmywQ+?)^MBVrQy>EWXoJ*hT>gww1>Z|^gwAiE3b#&GW| zxRckAb4_Pb!zbxU0>#-H(Zm#+q?pf9Rf*I3A-Vk@(fWFRl0fSwI0F)C9r!aOEsQB- zI83^a0Ta@dEG6^qU()#i|H-^-Pq|wv&A=aI%3XkRm~!!8!4S@Qel?OXK~2C@HS2DU z&AKbEvFAm+v1eV)e*_>Z&FA3X5l6$=nm9JEx5TE~uV8cV_!iil{9j;mNn6-3eg;k& zVzU$z2K9hAKFeC-^Bl%QThys+iO+pk{yueHy1EVOTrc?eKy7PlD|`Yi@#*#}_04{~F?Tm_tn(I*l1%{RuHEV%6 zvYi}Lre3t5TG)IuEVp{_+6-$GFEAy%j>l@*VfNU>F$bM4$g{z#zvXxi*>PitwC)mp ztw&or2I7#wgB7D;lpUw%uf;TkW3THppW1J!=0H7o%4#(m^db5T^1G3>Lz-_SPbF;?ht7kZNP=m@naMvtvCe~%t3Q6ff<&a!{0KGWI)Ws~_S=e1Sey;9#X zc71<=Iuxt#?r42B-6`6rOAG0*O{@wh*1`o*pJUyVsMns>R(=1F`ku&C?dzUc-?6_{ z-}V^^O?yGr*VDpAqNqL|RYD3jfc#ou2Hr@7+#Fn3Unbbgq2?$(06O1D8imOP?E4&u zDCU#f){K6!6$m}_mdYl@CEI(W#3zB1zBHo$Ej#r;^Tb+V9(j;9A2+w{y`|t|-IPuF z8ggbgq~hl;sIATn2A2;0kK(;N*{281u(7x|I1{V1;ym*$0J3W1_*R;8GS&LvX0$r4 zf>)Zi+Rr!eIr{e+`+g*bBkwPZzL)#xAR=}AjYyneK6wxL@dL?hA^Ib?8|H5ng$k{o z4}^5z;dY)Y@o#(2d8CN_EY+KZq-4UN2as)nd!2n4;!HU@!n%uB(Xw`_wRWj%`Oa4l zbx3`^$bMdCKMya6u7Bop{bTDzTmr2y-Cz~YJjp7h|1FR>;)wkH8=zAk&cZ*|FQhv4 z?Tz4h0)qH>oUIWh?S<)tlc*U#Bu`?Ru}^MWROEdM5V5!G6k1m5!-H)+t!G=|`9j;O z?+BVyLXL(MS!8|kpq&FCOw$7Z{#j<%IpU*eGBa)cZDkDW9d|3NMd-2z8G&rB$5ZQ6 zT*=qZ1DS9wDex*NU}HG2HHPil#&97=OvsU|FzgY<&;<;w4Ac?+yA5*RGp!o^nrOZ) zXr=|Qi;y7KMS7t(3%{q6^Y6C z8@uI){|e|()QV|>!kUg@Z!lD(A(aElwuFK96GX@{ST*!mG*nu1;3Xot+!~J8;iqNY zFUb4gxU&8XQd@<16Cdg-&OqX=UB(fW-+@a~XwqEZ4qk)> zXhwO_Po8t!#{QW7PSOGo3{iURl(fKoaP2DE=7pDja)z^N23GRamlhkT|#&vPaUrL%h8q)O(0?p_v`JVG20X%XYm=oA!qZKJyu%e$<5OI zbuuZxBB^m0)Ln&FddOBjxjvI-42!*CKP5nfdqA4 z{&5A*&|@sMI31uv?D%C=&=U`M&l!Q=a7kDZ^1HzTeZX?iZ(|h&dsmt>Izdozg&8=NZ$Xq|6jxiA zjUDK?o+OA^bTJCBaE^X$F5jWbZ^C;H7EF66fs7n?gDoX#C8o_CAMv_xu33K{M;}+m ze}sBuTZn@<Bmg+)C(iG`kuf}9isNqtBUPS4Q$Ovk2? zld-H2?GM$SHkZ{3MBUezV|L-auOy?GkO}~R%g=lQ0^g;8$mTtSg_~MbXWqdZl&Hk7 z({~gR`2nt(eMh0u{sjfB8##c5?ntkng$*!Y;3*{phlpNevwL6@o<{8pL29}P79_pQ zqaI>TOG)&}HaroQwy+@&^Sdla&{3Zs8|^3$>ijP{$~YBW7TXgy72 zG3F?lenoDzlvjyzj6$s6+@n3s$mnsW-+PzAoPk>{{^m z0BBd}HtbS)L6%eE(Ai>s`zd2t%b-W!gXj!9F|+F!q(Da88_YIt1a6?C=7TC2#1O4j zxFUm3V;WK!6nMx!F&_q~1Z5T0#NBtI6J_kEM5a8d z=O1t+bILJh~H-M9CIb|L6CJwVAXXCLE$APf#ufeyg_q>H@T=~(5Xb8vWSY!A?MHbDZ<=^ zU+X&&M2dH{fQ37Y{}A3A_(flZhuSSX7d5tM;ZG!Zs~!qmM+hNDhH7C~w$N6e1og9| zNlo!*-&dMhYz_|2Wz8rQeOGDz{19GHFvevI`8MkP^li&YaHc?SE2ypdGXzhBed?Rp zJH^vaZ^WJk{KFyj_SDc$9oMtNjb=KOLxO^yX_YM3rbdO zxoXD+A6OT=Bb&&&ouI8{nIG-r)AaN!6125MsG9!`pb&BRk@v}lx#2;9$>H!SIfXFM z^f=6E3g*Ad6Nf`0Q15GCn9KM&@23H6gFktCpJ=foYQ)tOX<) zMaVH?k8tD-+!4lQ(hOO-6=u6 zVl8|_XRd|u`8h0|RLb+M#ek|rk6~N!LZ*zYUyT=9kJr(7Aq;iUGZrt@e;p@&IU62( zFXp4*|4#mf2WtLbf~UtBY_%ADmZ+S@;ChT5y}RJ(J?Cick(wxcCyFOj#l{_kna^7WbT|v z5#1Ee-|+K&81!xCZ+HZ3&Y*h9-!P(vfZ;(iI|C9RB;;@S`hm9cH=MN$5vk z@?+S12`XwMf5TGbbwtc6>X3G}0Q{@@8{WkS@%#lTt(k%{z1@!WA+Me$<<{`?qVv(YD?~L zTuO)$xOe~ zxU%=QN;&Lh#i23NkUkCRE>Xt|=tiqt>Bc$&kKH*l}Dsg11PED{~FwY9C4`hS{&%Uj9Vn16k(9{Bdpc2C! zHfnt=E|SMk8AfrqM_ebs^JzFOmV1?_ornBMVqHOwIk6^SMMdoGV~xKGVK89=nrICc zq+2022ZeD?>UaAvX`s85mWAO&Ju!yA_j+v0eQ_IOqt|@DFRJ2pgx~@eoYTKX-Z9SU z;2W9KbHLLwwBR_2&#r}#`=d1&$1hyg@&JEz8sc@|4{l{4G%~#H0)U{6>?m> z(AH`8J}S^F&F$5g7={QgDg*8}d~gXLgbIt*SEEow+a=XDN$5J z<`7}!I7~IavG*#@3w15X3+PG*Jz?)+8Y-WNgTuLnrqJoOgU zL2%e`EPU)5j$sX<$x79%H&@;UWRqH~1#XnCa%|rsYB=c-S}DY1X37)C!jNcB{K0Y#hhU(mNzl_>dV1n0wp+N-pj_kcZ_Sa!1DkK&e4_BRPky1IDD2 z10zO|;1Zy6qU{;0Ri->vWUfJm*Km#9izNXWY@oGy!tII)b2#sV36%`nv_+WkBxjTf z5myVRXdFZTkH9eEc>KPY{MzXkm<&!_My^nDQ5ajnykEhrFIF&JHe<(u%39|-z79D~ ze}i%X{mUhkjdh4M#Fshl@rw820Ai^H?)*`B@9rt&J#foovV5^u$SNrEUH~KY7p_)g zHW7+ibK|=QQov=Q!qxo+ryx=d3%G#-u0Gr@d`RvuA!5n&32frT!v5L zC*%O&(H+mG*az22s5F8JLCyr8RFI^ZMv0SkNS z?Y!1`o%8zT5y#-_Z&IvrQtfifpYR)>{m0g*B<2{qdhjZQUi>M6TwqPE3lYe<(`1@S zdpvrF%3#E>?38PdKWTAv4BKN=>#7D3h6w?@u(Bmbh`j(~F***1=4JwBmYrrW3Jrr% z#6CO~&gYZ_rOX_Q)twWjUzwEHod3YJPep>J+*nU&f%Ysvdj;!=s%G@8n*Ts? zY~W)Gp~y-t(2U-~{=YdKc?jIM_gmDXB%&YQ!UBsI$G@4p3Gj@aUX~h5)N%bSkaFBg1WvAe8T*GI%BD}C)L{|Vl2hFtO)IRWS<9|C}|3~~JAz#0=)?R7%oG!~@ zGXt4&*|hn9_5YBs{}zPwv`*0R+xhzc8vLF79Pxbp_r1yvO3v5+<&mWP94cS`+6|z1 zLVk{?agQGU)qMS3t71D8|0ntSFMBzjp5uJ1N|b%!|14ks9}y<65>2<2uOC@1(7npt zoS2xS3UW1?n8Ssnl~h@ZMNLs=7V7fFD1$WvoTgCtls~X#zW%!iUHMbZ@0Kbqxf_Fm zbGewqYpm|<~;_EC9J^R>_ z-_O_olkmu|H}N;}^-oJ^Vq}SG;$OzGO|UQXzC|L5U(eTn+n$!~8bq{IyM}gD?dl8A zXxB8z45-b&m#@Fe@3iYT^7T(mXjgZ;T}vxfyH0GQUB8;I|JFrZG*JkeO~|zc_ed>p zr=cPRvW1VC&nO{Z|Fyf5>F|>|kQ}J!;O?O4&`VJHN%-J}-1&fLc4Mi8eEpx{Hfj_M zOi`lhS_G0(RP7(jo{8qIm^zDjEB=@A^&fs;Rq>Ei!9?mNo^utpO}_r>Lm*}=U@Bk#mAk1a+agKw%YqxW+@$!rzvP>3fyHm- z>wgs={(8Rt{V80sCaK@sn>FYm*;AleGhhF_&qZNBeXOmp`$RAQALQ#F<^d!-Uw=0s zBTCJ1&|OpvJ751<6M5gFSz18l>mP$E+s@Zt0CCK*vT6*+O3Tla^Y#C|?J@=yEWe+x z|KrGZ(y7u+5yV^7R6;0FJcO9>bQQ-Fx$~rAs#9m+N;EC!6O7!E*ePxD_n?F4&Wbw}DGVwS<5qF^Ri6V!gcg4)YMX=gWd z#B10E7QOtE`g3cqnFu$?fEa{iHL>NR9YZGPy@bK*4s|Ma)oo69rS7Kwh&`HkrE8@P zRlN}g=*L7-=Ks{2>8!%;>MZ<=C_$Rh^rBu!^e9)D`!knj<5@#WJM0L@vRfG}ip5-m zWvbRy&Q7h)PKT?=*oHQxVHbwx|5765Fp778$uNNRU|Mc@>b7#@{BmO~97S;BZ}3=unq*a2fUkM#yF0T8`?tyMv2CfU2Ln?wS1S3lHM=G zrs6Vg$FgxEmg|aI30n13Py&_(QZ{N?fRRTU8Mt4ZZllo1o@``$__O+PAciV38jFRd z*!m!C)r|_cTq|;LpU3*Kg&&onCyMGl_hGw$n4OAgO+ic+j_mA!>TLfJRQ}%Pj#X$O z$%)Nq_^=eQUI=In-hl6}hV|M_hwdnbqQV@5?K>InU~e}Z!MxAG>9pQISnmTO?K0)T zn{Xgq;0MoD<-rN|QBdVYm}Pm?gE+@=)B`f(-ZW7{1x_!|-%p2Em8+gIXyo;bGpO3x2R? z$=95t1#ZHSuet;Q;v+fsgO}#%p{ueL^DYDPdX7U<2dh8c319Pj{KxNC@_X;*s-pJZ zZ}ia2p5_-A{4^r|nav5!)XjSP0pkP_8b8`qv|9^o;{J}s(J>|cn!ii7Pie0$lPF_5 zl>zrSpCMm1r=pEeJH0l%e**-DlNNXdXmLovYK|*wQ8a#SX$5)#=TOYV)EJru$pc3n zMg|F6Fp_eXJn)dMnd#XNRcuj zo+3pzzJ}@r$HW?nL6I01w|bCXNSETmUR|b3aRt43=J;pqKnd$^bUSIo{O=_GqxcN8 z28$mw%j{>f?Pu0Ir~;D9lKDL}Dz`#Xs3C*OZ5)V&GqzCsXhP_Rc?nkM3|6fnHlf0` zpYzR5^GrOZ*LoZL`$*b?~58Ej*gR*c0;>mOeRTI zu8nJHmq>3=N)G0>#;}tr!wvC~WnTR@MiqsIwh%_|X&I(24_egaeO%`LN#-*EI1E(7 z!7rZv2odli5Ozj}3&PXX@L?F@_;J1t{@E3rf+NR|G6blNL+>BM4vEQtg%O<2JHfP9N0C;*l z`08xokoqw4tt=0l@IHpq31<-rI|L=etI)v%(!|g4=F0~G7hhgKO;ufpkH;qEKj2}a z3)uYz0z3L+O=RQHu4#}Z{@g0w;m`c}4QcJP_V$1)5#(_GLNV;) zh#Wxn8v78$CT9Emv2QenUk}6b=fXt$Z6J%VfY_GHpSJ*7-!x9%s32iI0E&%EA89Syv;C6 za4WQW33bTC!pG=!dwxMnM*|N#P{+(t%^V$V<|r^QXHF1z8fXNQckgl}?osDvya-`L z1*V$G=fTvHh<*?1sWfwi@u9Lcg3uuLRs^d>97o4c-U&wYf(~TsC!kagT9{&0V!n#=S8EWxD$UbsDX*H$ z*M{5mEn%mdufN_#m70GBh!}CSzfH+Nj=^LZ`1LBvz|=axY=ti}h4*{+L%!cEbuMT8 zM}7D{*yXH;0ExfpfRBe3Gr04<8}9krWWPUM-Y*YB#yvX@BTvR#5R^vFN|n9@Dg&jU z7TADKNWg3ngq3FaMi^a1$01eCPhM-Qfq~OSjA3B!|Bffb#5i_7it`pdOOAaIxAfd# z;va|Y7(ZsaQa8GaHfn*V@l1vSo>l#Uvah{$)q|ONFKS@~2mAR3!q+ z<=dVen@G&ncxRQe$t`otXb!@vy4ZeS{&F@c**7pLwPpa>RHp_2c zCe>7$9=pCgG;E}GOY8cI&ccMqEkQ;{>kfRCfDZOsh0;&LduhKagB{iiw2*3~b)rCw z(L0oG<;nzmZ1l6xES0t_bXg{Hc->yde^!pwCe&$N2R2ulztq`;{P1Sf4I{=n`9A%5 z{Mcmw&F|k=-y^jrzDJM9P#bAgUC)-(K$ktmcR1DCQ8z9ws9j!zt=-;(dJy??I)-;r z&kv>l^Gx7tbi`ZSD}E^9$HHIwyWmXz0l+RjzZtV}nqIWh(_Jq*x|oAc*UHxFscR=4 zuP;^wQL<>Qt_@sFpMh=}meG)QBJTTy|El@}8ZG;Y%1Os)vGA9u z%pUK?L2Dvf9J}vR_)koxXyWymp0oAfOia*)9eK!R(Lv7%RsGLlj;ukqkZ2?gyK_AQ zH&90v?~zx5cX9andlSa2pMlL(aMvItV3(QX!;{fqx6QPn}`GoTl zPrr#xY*EOjSK!lW*pTog<3G=YCVwY+`dW~iUy5#7hRN^SF$zHV159L#Fwuqjy#X%< ztfh++nFaMahq*&{WJ;7Ec*Sj8zypXp=ybFJ3#W)9I2xd!ulSm^sTbnkRL?0`5_|iy z0nUcfQ$Z5xPae&+9QrX1ugR@0%M3T!_=cvzym1{Zf`Vol?J`tx?56etrtaWvDPe5! z=+|9vTm^(<^i=Txagvgv+mJhKY9_1~Y3@E#b2NV*##wM+cDb>~`5uRy>jU--{;HsC zrD6@I8;8q`X5W_n$l}nUJk;q6oR_WVBL=D0T4TMQ-+(sHgtv+L!pfL$ZyvsNMpn9w zuS<mM?>y5JR&} zKlmwL>)lqv=8faEw9;kV=;ic9`g^W+MjA_v8L&L9$V*U-$Oi}4>fLOZo>z!XU=rBX z<=cwR+3H->A81f>xKkclcxW!=Gtoxyz022-(O8y^Cr~6|D@d*Y@?~)?zQo@wP|SVq z#W3%|qP14@=W?|Gr7VNP6g`Y?UV<%c{L}PM7P=3?D|*o~&BrtfqG6QN&+NM$>k}ld zK)06XA9VR58QMG~_TZWYswy}#;%Kh?CS_2lv|ox7E9xLjcOYkt6p54@n+py&MNo!M zq@5{rMQXY6S?H>EuFt+V*65K>JeL`}%Y!GmjZe`LJT3j0b*^rok+$8Il?O2TRw~M- z_>T7X*12Gharx@-e=T%##STggM?Uc`bQ_H<)xm{TZNaB*<15#}&-7#JT-}zdE;3dI z%d=ftDfS1Tg=LxU{7vAIIz)<-LvD6T2lKe{;*`j8-{JnAfstkCd(Y*@I=4|<4jnzH zq5lE*vn^x-7nxvf`7x{A-B!B09l|*#F0E`C<`Un}{k8!RpkVe*odvBQS`dOrpkl&aP+(la_ol#oNOfc=aW^5?MuuTvh z(RgK{altB3gHcP!|2LI7fBWzXH9qBN(eN8F`5>l6GU=>13hvN)(OJEE(C!I^7d%QdK6Pa3uMI0|)N3n_d3h1R>5S^&(c?HIxwYssJ ze6z+(7pz*BOj)F%!tir3{A_{WvLvWD?t4=4?za*J^4J>DXl6IQj^SaBu0hj7?>)}z zq3XJd*SfkM$Lqel4!)=38auy6@CQxost47zY?gTRDl*P{6q6{^JE4qROPM=WnG97X zJE4qROPO+2#w1BYzGnW|b;-4q$ya4osxs>n%E+~pNmpeSs4|NZ%E+~p**r{eu2y9p zPbeeTQsx6GGkqL7cN#J2f{vZ0A}^m;uZr=iBKoRAy_&0D<=|D%=&PRU)dYDZ%)&Q+ zpgUls9%-e@9)5n>| z3c%@Yk-n3_ncf!ZcMl=jG19{uVUvD6UL>qIsE=pI)*~aWZ{L=lNf8wc^S6wj;4JiY z$>?bBj#j-pPpX(MyVpAdzK&Hmj%ctehOUbK?x?q)W(KF`_+yckA2FcYGYn<8sR zE{A-D1zh4kMMs;xEf2X&l8lVmc)%Fx2~4nwh4(e>vjHys!>b|7LWSEmKbukjCd-UC zW{;#Mhp1Id$v4X~Knsj0+96i#Ow9h_p42}OQR)~dZ_zN`ymUR4Uqn!&sM6Y`nqe(j+P+9r3_vjxBj&%XOcXkNAqvS7R@GH$u5V_?_;X3&xC!Y(n6%gl zN^gs<1Jn#0oQ6cpJIOz5(kjjK_C1wi(QF)Xc&`C)9CFdKH%>yBjDMo*4|RU~nie6i z&1e|(tt@olIhbsd`rE`i8kLnzRaAc1@yv^`$(s*r@fwa182h01f(Yixhnty$0!>s= z2vjhz$eEaGrr-@@>qqM$+604dsUu<9za=}s;5s=TeM%rb`OoN>aTBJ_i7Dps?{R-a zVGs;P`}^|slueJQPXTH^@Zj@42Pp>&7ks0NJ#80TiPx;L)-IOCV%YsRpU=^Z*ZGTD zb2!4}gi?7Ta=(5MWrcm}DO$CbFp46o7;n^ge@2Q)V+(g~LZoCS@~Xioj7@-4Wnm-B zHa~t6=gpw?kQcOS0)H9F(1Ufr5e)J$3Wy2b(m>wcFH`~zDo~Fnv$L!HlngU{F?R_( zq6la=3*VXj4oK_oLG`?S=1aCReq75=^AuI!;SH+53@Kn^0XS8H1mcHq5`+s;@Gh*q z0p}+k3=yS3>z2Q8g(A>=ehsYKB4U-4k+dZ2U{G1|%TH9ayU~Uq^`kp#2r1eVaw9#_ zhs8qak$vl7+`bu?(=nvzqLp#f9#v<-V~xG=d!*d6$3t?1?!|6nOpaZK_~pf_M?X-A zeqYW8sBFLSk;jb_92qS)PGk>f@cIhtu0n}_v7WJi7udfq*}vb|zdzZ(PvTdNzb|B4 zl(8R-y7=2?)g&@|TQww3($JfR2g;u_+8$n+=6kz1?ywCB-?ndJFOskRj;HW)!=(R5 z$j5?Jz9YwI_io0euX&7iZ(4dvO1~|T^xC}_@lxx~%j0lqEbI5|y{WkA&FB3GXt=zX zm$M#w9G6G&O~1&d{ps4h?Ey@meK@k}$v)b>9dNT4#Zs0c-E)d>SvvDdu$xj>Gk0_q zu8dDz@Ak0&3@m^x-{B0+*8?9GA81I!RtR+k2l7{_#P<`tJhiycgZInldg?j|vr^K7DtE-)}iIA9lk7iCQ9=Dbv#eJuWfO^w#EIX7Wcba+^3)q65)w^ zDxtod7VmquxG!#TUy*o^L~B>$9-TC$LajH9I;`PtS#^3KJXo)@&MF;WcT;KpGN(~% ztTYa#ep-qHeYJ&;WX%Vbmud^|@5GDJ>P}rkgCf42?W(i7IA=Grd#5f8RrK^VA2TI8 zwKh}{`=Juo7izyUE28mQ^@5#-xg)oeq4S!IAC@P*r2${y)B8b0$2%*5e zu+UPQa?M4%;bA`?Rq*w>5(So5DYA#^Ki~=7zT@Dd9a5LBRlklUQ?1DDp{1Jh7!6 zZ<}|qfSl}vrD{uA@s5kKLIz->AnU0ReZqb_TFr4S#N2eg4W;+rwLS%FJ|(_vp{pbq z4yxKNMLKn%=&~EE-nBuNbEP(aE(JwQZfOBOZXvrUPm*QV*qVdu#ia?((B^J}DK)0h3Sxf|6_XNPr0fr`z`hXvl=g*_Ur4quU_F7H z2`*zlO{$+qcN3yr9i#Dl4Lli{mfr_2BLdueQIC8+xS=8*@}#GjC;po$@5DUw(Hhp= z3x%4@T3pkj6XS7sII!GkG7lEv1=l8}4bLRR60kka;cmFioL9LF(Sci_}aLhG=L zJ5&w|Ry%MV#mWTPA_H!7}73$65?Q60m2&;62ifIlx#Gam+=R*!CgVHpN7Q)~))uMYCl;XVgCcJ64nV z1w&=wqo}V@{qt0UeJGWFiaz0&&}GBcvCAr-r?*=Ox1_gCxZ~>K>7{*|>%m>HXGbme z+QN&qg&D=YcOx$-*u1#vXnL{sNMqHZj!vAYQtXd-dXypeljKz1qig33$QY}gU!1WB zWmB}LmeZSo9#cu6V8E{d0Df6lbtoM_%gX$#Jm(?6%+s9#C;iwbDr#y&(8ilf~B^H(!V+DL&btprdtLDGLVs20v(`lZbx;wJjrf1cm z4#19R72^K_iNC{&u{tXOcx_7zt!w~S@{MgdlX#sbpB4?s&>j(ISBwOz94VLPU*-L( zV1xBF4WYLGNE{5-N<2~J_Gp;{TanaS5PR7o)1OXh@5$^|+ zdSDvL(-_$$;y81lEVlzU^HH!Y8y^;78K(t?;(@lXEXUa!3O$gvXql~1ft>>ZUk?t3 z9$k^A1V0p6jG+PO|z%U&c$TOw>tNB_T=KTxiDq`*Z zR(5`&{S3dQ2bum#I-#t)%JwJJZ`AoVrAqYFIt#XbvT)rVZA7hwJz3&!UFZ8Q{(fEj z{WbDF>VJw#%4rD;0xSoSGoe2;507DG|)ydzg zh~ttnu2vetU4(As!CA2MM?B@O;Pue-_PC4oM!mINnfpfkff_!6Mdc)S{$6)cc!nb$ zPpcO#pZUJA9^tfbZX5wHMAJsRZ|X&;+%pha{#(L?TT#J$uT)_E5ip}Z;1g{huy+4r zK$TFndH3T=QVH4NT3t%XBG^8KHF3SU}LLjLI?7=)-S z;+rfL4?h}xKP|zF28HfKVxaJXCxuFLHt;&$#;X~AE#k$c0aw#&&Qo|*w8lq-YQ(Wn zM}~(SkfPG;T=HZ}RUd3_uAe$oTL?dgo!%)5l=uw)$bX;`&lv7pUtCd3~0;KAqPmsq4vp)hg15;$7Jag(h7$bM{%4~@V$7{zgb;Plr(Kj-`Srz(7WZ2l_` zP7zO6MS1Y{yz(N%Z!(}<&L9W>&&+`rDB{ot2(O-kRjWJykh^Ge@(`$-`L3}NL*Pxf zu@~?#1P*v#mwaI!B8mKT@RHA8zBWs?EE$c_JNf~B(Re7IqjG^$lT~8 z#hS3@>Pf-K^@P~Pa$pevgXQT#7%w5&uqS|ZeSnx-Dgn>clig9sj-?Cs zG*9~hk3xkH*d=DMgz>cu7S5 zR)iCoGE{hfiVUdo4bs__ra=o!D82a;6l44JP{Kg5_!RXbS(4zxfkJ%#p0Duf-U=Tr zFhO{(FzqYhGjA26XW<_1rs}yD5W|n5!YRB6;o>@8_;E1^7uGbzv3+^1Af=zn>(P9U z06YhhjD*q)_y{}1BVmL})Q8U}j!#`6@xPVk;{IxA#fhW^nEFP>-G}czlM-G}eZB=h zqjW4&Jj|%H?hMRSDlpWt+k5KWtOemPy6 zn+j?z;W7-W{6|>75~gFxnr9gdd$IV~FThJgY0QUI!TX4#dw=SclQDhN#`RR_2BM=r z0L=yt(DD%3ZJAz_hV2ehWUO5dr|M>H4$pt3zIwTy+CY6(^B=@(r>{9foAWPRl^V4O zW`Nq!u^C414$XWsDckAW)Lc+&EQ6yQ8u5v`(dmTJqS{G2=({zR!9bURnBFt6LAhQg z6D@Euy8=v@m?P#2;^teS$25Lh^)1G1NH_3t8s!x*}4;=?zfE(RxU>)LRy_Ied6VXTXy&#$ddGis`MDycKMPw!JjzN3z z4jVs8GOdBAu;yL*cP&4VRQewlZJAo&JiJQt*8>QeJ*p!Tj?le^VB9h*HL z*@xrxI-)A0LrJi-<1ei07gmK7JT2J3dpVXaPp;t7<&{SnRGY{n-t+PDPs{l70lq{) zj^zP9?9xr=DzO&lD9>>b$7S^6V07akLaZ2%UzE6FRPpy%QA22+`O&jj4Y`bu8G!IP zmPLqD!~B#qo-sF>Cqv|kLj>^@ci6IyqTAP?v7holY6?3J>>Ls9X7A@kZm<1NNkI!c z#$CZ-7$yjD1KPKI3v*hI`pW!9LZBJ9PHP*dn9J~*#oDoTp~5Ew%1-ldfPw=Hb9~KR zJSY2_Gqt%-(7YOQtoboPQrEbQEw0cMXl)Tw89|WVNU##j%YzH){&qvR%!wh7#c;&& z!zC0UTv^B>y`5%GdX6h$MBp@;-MJM$RM^#SP&RMXDnu>Z|FRJ+f{{|1cWkQ`O~Y$~ z>3I%CZN$vOZy7dioy)FH6t^AOBFvyV&!?m=Zi@r1TYbNv$!}0zAg>jgeA$?^=S5>u zLK zT5)SW^c-qe!vF5jS(8z3g1p(1T%jNW`$s~p*{ar!_>N>*g+1P~I%zVO;F6S$b32Q{z12w&Ddzn*D{=pe6p{O-yx$hTe~$MOl!NsV${*pq0x9>K zc>jL9{1V>362E_m_fN;~ALIS~@%uTvzbAfwJMSmN@3VM6Dt^DO5%=!c{aL5*eZ2m& z&f$H${hY3Gk|5-bDAFux`!6%{qO?=-sj?YHkpBKL$ z%KKB|_k(%gDSkhQ_t+-}#t?eW;r#^MgGZ+D8i$|BILPa+#z78#IpVYVn=}+W`4PvQ zzT%Y%e+;cGtcCZ54~F-L_l9@#7D-u~%*|(G$I(*JlFoY*>dBX~?lle>dtjf?)O-PD zq*z;c6jl}Zj~Y2beGxd23|<+B_0XMN^+n6~=Z$(`+5SAeZfiUJ;CJ#fe%VcWK4!D@ z2|MA2c_(e!Si{+M5zVN+Wj(8}0=h+d@I-8;gSzTi^flU5w9Y$H7Sf#j8!Z1<`-KjS zTn#x+U7n!vs0GLzuPq!p4GxX8Y1Ph5*{>6R*tXk< zPSFtAET3VS1k4vJoS8)Qhp=bA>uxqQ4iVeN28ZwVlsuC!GKo2fvpc zT4p=<_}ZIbxpU{&x@f=q!ii2P(d%}!8*8MxY!}gTT(B)uFAH7Tv)YMCCsnqnN+ZO*1=+K^Qit1n$}T{V~^e|EP%vFn&^sCqWueA z1V&we(b6;`9OqJSC|G zELI=VE*45J-_7+n{b#|8#ioMYXdF*oJm&fBh%#{Fbq-Tl7E3x-^T~CTNO0VXfkWB4 zp(VqhVJ;nsR8ZmyL;iEbuZu_Dc7)ZOua-v@@h~H?<7|nnTM5s8fhHr8?ZFbRaRmul~va?IF)#wo`llD8e!DOl0Gj7}V&E$nnmNp+`Vt-p%U7Pw`` za{8mbhd*i+{83i~vyQ#O7}^;=sjnt&bBAyR&vp30h-2@cs5UIOhMgjwAEBUCftxsQ zkbl7+A;-ccP%R}9A$Vl=w%*de5=ip29FFA6de{3Tm-m`d`Q% z5Gwp-ISpe6%?Hx^g(ZZ84qrWf5>p>^<9&&*zxG{(TC|p-yuEF;hvs_0apdRjdsW z0`e;-2K8iA1bke3%x~7PMfm7@9Hrt8EhrX^kkLhFYNb#32eY^+ zpDd7yMd*Y&?^$p-!mUDU^^iI;n(pQWsG5JWkqGtCqd1@U+61_8?DL20s06bDxKM4! zxMfHV;#gioWueaFY{;?`X+4@s$R@$!50G%S!29Y9B2~dAJmyi#CSJ%{<)5KUZfHl0Qas}JJCQUafMM%JlVQq%AgEM#Hc`7~+CXW!pFL5?ydLFr!RT#w|2$Ene`C&0Mc8r5&}nt~8(j zGXw{F1^zw;njJowNaG4HV6E(Y?CF>?aB?=GL`e+c7pFDJ!b7(Io+tEqLFnWP1~{x- z!AFbng9HNszUB9sd~Aed(1kjG(}m3TNN%=z%EF{W_oO&U(om|8MNq~>8~FLIQ^}_X z%{%eCh{!N}Gx{{N(XO9k-kF&|1TB(?$CT8&|_pB zK!$t;fuO(aFeh`1I$F4a!wCfXOoo-3JYtio_jrVja3my5s1ZjV;%Kpou(mmNMB=~o zR>lduueUPs%c<@Ua=*eFob+37-Da2b+U52F4$J>+|BkbtM}H^U_s`hBi}1^OO#A+I z`}ceMcbEP9JbqcO>xt_3B>cj%W1Z6C_cZ&SzX(t0Jg$MH**`=Hal~8uwv9`eV3msW zJq(H>{wdm-O1RQ22G3C_RZSlxCb(?MGU{to#$Yyu4wd7%0x3>LPQ*ViSK%#Q+{7?o zA8(LATz>MArcUcRW}=qQm{7=+dl=x?Gg0}0ttfmY^#uTWvj*Sd_o?~#1wanYppQ2I zC_nj7{Iss`v783r0yD2In5+!DtpFjpGFY|FO;g^kjxex{!_G|D{}D%B>zusmP;E3VS-0Nh;NB znn#lqwkbB5kG#UMhOE1EgfD`o&%asNL7Eb^d+%feqOloA)wxp9_>4PY@AkT5`*hTn zqcUybZ4lKrr_u!MR~vu0@94Wp*OXSD2a^o;9gVT~9YwwU&g*PCL}NYfh{k$6bq{kj8V3GFW`_*?6?_ud^oVu6&OIBs$y@G+pqlIs z-H{7j=cACckKP1eEX2Q7gPt1hTE0?*Sump-`!*Kb2(WW}_Z>sl9 ztXEZP5u7fM_$oLey2m7--Cs;a)dW!rRVd=vDU8-9yoUaOAEp&hQB?b5e=I~psl1@? zSr#0Coe)dy^3U)*@VXf^OZR)|WfDF-b00-2^TMnMcZphgp?r5^Vg@hfuAOGl?RLh5 zDbGEV=uh%(gCRS?nFNkQZh<}sijdQdux4*&7O9(QxRn9NR?_w#9GJr~j|;dQm~;oT zZbM3%F3C1zg#q6j&J$oaLp6@c0xj?@Mmgj`?}X{W9>4D3Xaqh{{p$Exl5Z+ydQ*iJas77~-BkNPa_J;Op)ku5Rn3juD=z+bq+&zxT z*^r-tIat)@a%QSPinbKr&Oi#V8INl#_Gn=7?Gp#e7TK*cyO zml5x5d!RSI7dI#N0+BEA_g~ zX((eMYibLo;ORL1u3JE6KJ6YjacqC5%Bb(#1|~xkW1l-RJv}gGoj?jTbii=ERIC!n z^R-JiP{6k$JBg=K;@Jz_!HbbuDY09zdsq_J6z$ji^zX}dVTAR}(a4?%{R$k93siSL zszR5RVMCH#VYFAZuIEa}vQ*psiq6o?CAV_u9mb(anMe%4eq9#t*RQ!WYr9_`KDAZ9 zwrZl9-=sgTiMJN^PV~oGB&275#QHo`p!!_a?`gT{blQf1{Mkv=gkB4x1iG=mhvFZ^ zU&Z(*{6=E0Vq5M&SjE0Wul^_eIJ)quf3OQj-klh}shzI2Ciu1$2QTjpt6lP$KJZ=# zR(X0vRz)X&{mhPDgvcfO>=YzJVhV#|a25v@{;G7}&J29u`Go9?9kek?z>CZ@ z6j8#eBopZy{|s{-RykU_q^t9ZYLkNHVGh#T5`GBzuNhJRWrhnDH6ukHejHE`7+5Ux zc%KqTM&TDoM$TOMk|FXi$^{Q*xmhKvX=V|y{WiCVB^BG5;mD_}?}E+vquA;N~ue;|e3xp~znT7i^D zYy`wH7hQQ~Z&rC1sziSP9^hcYZxb8|^1I-yN`N~@z@a~a(=aKbb>;v>x>-CGETOSW zP&BzeJ>4RX>|ClUiM{OIia@2is6#9d*FxQ@G*>^#08-kwVumf`QIO3lnC>Fxlg^F!V~R&3&O@4yc<<#<(iB_(&fPbB%zyBtcUK_ zNqW#fE!TW^EQ!Fo;N4)Y%dmOr6V1Pz-5eayQ+e>8KZ)ZL)1PD(tTZ3uy{~U>iuW;w zTtmT7SqKH-H~?f$L04!PV14PqIdkQJb0;DJuEBTl`h6wQAg8sQF}mup=pJv13O zLdgn_&ooaK^ri!|R{M)PMWIQT9L)w!B#5Z0M%Y@4nMNu@eveACiLZ&W7HGl?Nr5!B z#zqE_HB0fH^bAl@1`(C~V#5%nse|^J<^wmAQZs-`85$-Vq;Vl34zuyKT^DrA$pQ;K z#n!luA1x8r&sQ$U>ppfiEu zb(!XT0XY*8FICNT1#uzGeJBnKMg~9bh$7UVZ>WZM5NdJv+* zv8)B4SCD|eJ-{O1PB$HAb#>-;gOAe~7 zG#ye;q2oo{Qe1ELfu^Y&=h|RvTILs8nkV0F%uj|NO>f#zoP$+5=7(rqkeFC)iQ->r zj&h(?l%DrLinW)}B^n3ZmZ~4rF^#SAl<4OU#4Wl{50{FR93TR8ve zR5gdiS2vu)T$7U3pO*8N8v)O14l^)-Wl9Q@D!~)n!6_LZ;Ic~QjqpiPe&ZpbvR$1D z2b>>`<-y9l^57jg`Mc?x#*$;!ow#75?zzKQ8GfJhxtrP3z!*?8r)7Jbkie+B z(!6mhI2Usy=IzmW?%=o_wA3A(grdhX?Q#)pFz(z8w$>T}$Z-LIuc4k>x8R}l6VmQp zi)%g`hF|1A)k6a_Y#qY3{lGq#t<&c64sg?h$YXjdBCJEdGJ`vw2+vjz#pfI>-*yJ(YVxFEI4OQiY& zDa2LkKnkh6AqFzcVFFMycp|wQiY5{kVLYak8wbmcp8)TWeXo`6mI&_H{LXq|liQ5> zS@2HK=R((G-NxKlcR(LSY%Jy$jsiL<(ezP_!iKW(6o_X>VRwT5S{}TOTA2s&*E7)8 z85jrXki)U0-jqy^2eP8^*N4E2zZHKS20VGzok1)BFjTT zDvo!CZP$@})w;0}5%IYczQ1>l_x&IwaBL_k2BFzvSl1g%bg~HYM4*3RfCig#b4@P} zO)Q|HIeG`H_z+TFx#cjUd)QZ0e3`R)q;w~MLahT^mW>jdkJ+O=t#M8EY#XjLi!>y- zJRgObm|9F~NK6geS)MuR^@x3f_v07=m1aR!Oyk0gl8Wyt&8u>xjg{sUyp);dj0ckA z9gW=(qzFhp9v%A9kni)fS#|oFA(KALhHNpW(v41Q)JItppSw z(R^?6HrNq9eF%t^IHLepUHluoua+7AK``}t=V#l?x^3j{qAFAsMWUY|@u@;$ks#rc zaZqN|$G>U=4Qj_g!!xj?y0qIhP-HV)%-vP*ir~N{pp)A+I-VDk8yJ2e8dnpXT?J}! zjdTbaQK%w<{C?Wpk5P>aSGw;%*hP+3Pr$+%vXCVGH2f&&XaMe+}_e8vR zhK4o!4xg;e?F@);l8t*G+Rd1($KZL3eAl`@BcZ-mAqQNjC|2K1@%n15heb+M0Vmdm zdUvQR)H?y{U98>;uJS^KquQv~?yfmJ$&>R9MocI4(g@mLr@IhE7INJ67H|mNd5r{! zVe&`Xt-$AO2?o=&1Y>0X97PuC)c@8CP zI+&-v!PsKd7VMM@4oTo^>t@sp>`&M3-HQ(t+eTJ|=i%8*ZSK1uMD&?;G33vZQIisz zXN|$LcSnttXKsb9%u$O+-7e4U7Fc~yZqX=@Jmbg!X{_$@?8c;Lv_vc&b%s2<$1cb7 zAzMCUq>giuAT!X20^~vF&i4eHfyOBN>)Cj?$i^mG@HTmtjLp^ZEE$_JJQF^Mb=9X< zFRXNG6ifbuX{I&ly_ZDE{*Tua$(|_Ftm}U!WawSrMb=t3pcQv*6)N z!ZwPasA!|6cjAKqTdYyjanYJagLLYDI`q9w3q!o;x6H zVu~L#X0yDTAL9yVe>Ez1wCn5#iP+@_@s8 z&TACCbxC6HJ*gyF>^X>`u}6_}jg<=ia?S=7A;|9i3BP5A37&{`_E_iaO_BC>*0D=E z16V;V?#-0kv9qmWC`^u5pFz}@Z$&5VomFMpJGBwVD5z!$eA5HaTegzNwJT>}sPIF4 zZ(V^}6}=*BNhp4AtwGb1zw}-wh1OW7p*)$K{S#SBa#(KE)8f(VG@aa^z2JJ=Y?2usbVT;xH1}`9(#l3B4SQ#uW)mgb>nvfPS zP6)e-NbeAvU=|T`b(-9~SDF08LtN2T=1j>(gsHU3*N_7=xGWnXEt58y|GGmq zIIMVDHRrmJY7SQJsEKHQ^D>@<3Ty2G{p= zc-LE(fDOWkOcHZM^H%?sssi+PUxqWLZ~vn8eS$VIf6T;u0{R+Jiyi}>>`!FH-?#_k@ zmOEGa;us#XO<&kOM&XzKM>1Z5j&m4~v9{4$?tsA<#Y)Vs5=%;~I%Z&Xb~m<)!$%C$ z>cnM@MsUHGGk>%6woD?1n|(co$2Jp@A`jjLWz*!@Lu3-9SUshp2tT9lSjQoA#jo zOwG`P*8op+W;-4YU^HP32e&wka5&XKTR5?pwzAMP-OK|Ztdnv~?ZCdM0UKYP)(!(X zQV!F`%^B8bU<~R}Tzi;X_H&F)&sf4v#W4n{?0)6KwRjc==3mLg4e1Rl4f3^31+Op}o_xAE@l);<(L+5>eiVZRV+!MGCq_{YZeWxpMKYDR$c{MPPKUXPVJQ={ zP1gh&V?&=*W9+cO0YgiHxl=B`UUCt7l_Kr$wyfwb_$T$e{~VNwI6}t)E~~Tcl}%|t zH&nRf34smjnZXx0LcU(I8u%0DEnzhtKij9|PMn*th7J|3cwc1B+mHQ>zJi%!UDqe0 zwoo5nV<$R*s*D+G<8S+aqpH|$^aNCfDpZEYQAx2;;i848e7kjfM3SRZ9;EXhSP`1c zB48$Uk068=41bJ>4~+9$s<^{8ydsQUCq15f7}Ja-wFvd_^V;s3iM-~-j+TA>w6yH< zQhTTH!AzV06bSq9`2ZFxQoY%e*FDtrsl4u{t~>FX=Rd@S3;8Q*pT(3gJb96;OE`vY zqwSAFzZ%kVQ(IYR;lP~8F^s_n{`pQ1wX)CNMjL8y2t+u{y~u12%fe-m)dn6EU7mrg z*})OJX!iR9x}z@^F;odCV|@4}@$>{~!C+KKixF`PE^ZU-xBp7A-*&^^r>OllM>#Gr zd&B6^!hQ?9qn^rnm&4}jMi<5GM=CutGmqC}rFKN~kHg!lS;fbACLvkS7<4cOTG0{j zHd!p7{vc1BgF%TP*UEN>KkW<3v8>b9SzgLXnAqTPk#IQ&Vt>Hlakw)#l&v7`(Xq^ql zqOncH674wAWM43Lb64-YBTq;VTeF(t;S%y1UIC<+kvuTIX9$2 zs?DB}G@mhZ(E(-;g!~rUl?22gcPz#!5kUeccW69xoAEiwdYvNywFt*XXf48OKY5nO z^GF}suQ0~qrmGd1JZr7qh0QTwup%*6m2*a*M7$LSGUh^nR%v`&Y0kZYf)>t!gjpw+ zhf3M60xlZzr{j?8Ba$W=2$3Yye*sBsa~6V=aR(!`X&a#>J^*REJrB0BZ9Vo^+qPv* z0yBVRtVf_$$T~UhUsm==94H9s9kWL=F$%Pv*1akAN%)sHV4Nh`Cu8{`W+Fet5oMbc zdtN+0#LmbD=0`z(2-qjB-?C4(%nvcq9UOxRYHNAXHkdp=C#{jr7itM@^^YsCe=LV} zI)Agvw!vrO0ytLP)2mdnPV~_2ToU9OwJyX$6MPl7ZX_7wC<36p!BWYgf(Qvzg*nQ6 zOLX$(;Wr`L~G4Ja3D z{!^h7QROXWxCc0DXl{31MS6gvl|Ku#9N|PoeI}37{A(fZ5ZQq!Rh$EgYeuS;j8?%a z!4nXyTIPp=S`Ioq2D>#*5FC8dno~5tY|sn6xd4rU(rTVCT!}&+l7YhqAfXUJVqoZF zu?LR_GS9;S1tj4<0HvW+dAnMipb;M!J;(BDx}jn5hKx!+#}Zk@d*wzPPaP4S>V|Q) z2(?zx3sY3rGV#dk!U8YBc@aCx@eT#;@s%q$2vrQn&_6)QWM20q&EHz}Gk7O1 zUT)VfE)NX?3E>&1CZ{7TuNcc7ok1tXE@QavHZGsY#^~nKG91FB@&?H{5o>j$1O4hf z#am_0$7}M}h?s!W0w16zFylCTib>d%?Z+Fu3tgJgSTb5Jjgk@8?GVdKPosYjU5e3x z%qbw1*%fO(xRI<94Efl65*#vGTey-|a&z$0Sj?ga1Eqi=C-o&Gz{A+u@^!e68n>>Dz(mI z{Ju6(T_{B_Ky(jNA89GB)nDz%O7jkXwgT zSHnLhhsi|ztI~ZEm*iQZE&LN#jY!i4Z!m;EJ3P}IxhWDU_HXw(qazRlPHQ8~>T%W- zi?E{C>(_&<8d0wd5zHMb$f`5Gjt5y?nAA2q$f_S~{o6@j;s%4f=`4XVL48m7qh)*E z?axXWXSI~?|C>0go)R1N>v2|V2ap}cPZZv@?j+m6TXzSi_TP%P`aQ6B3D|QL)IUbs z1D7?tbmm82*s_lj1=57M+oYuK+n#A{=jgCWVipLT#8m);v4Tf8h?YUFNK+yE&}OQJ zK(he*L9Brz`*|6RRRw}2#$$_>gR4qZlWY*1>#qJ6{VcMb{6mi5ZNXiUX zA8;umzJLZ87a>M~=}j?T*+?rGT$<<;%os7MfpR`GGyG&5nBIj?(1E{;=(=Je+TI$H zC~xG@L}PnAU={}r;|v_kP&{@^^9RA$jJb@*7sH|gO&3ANb);YbLB?eqI(9fQsGZA! zm}(YR_k?sq{3m9PM%r!GeFGa8L2bl`jP-(dwuT$oG(=CQATSD%$I*CdMx^q{Q$(qj z?=nvW=58p_pGwRz!pbDi{ioci!uZrv?6_kvGJW4X+hxz5CpcvI1Qk&RW=Qlg0!t+n z*?J}_mly2&$;`!{OiA*wtdu0Iqa%VIiC#E0S zs!;|ECIfS-W@Uq%pQrDDxf7?z<#_bcU}aVV_PHPr2^^hgk^_QB0oJEP_Y5zs&e0;r3(kEw;=cEsnkx`BrPS!#^QB3SP~ z5u(uBoe#aFK1QcmIJjH$9~DM~s`DByTua!S_?rc3-UGb5$E{tuM=wHT$bem757l6OxAH!kp$sAzYF~2(r&Nk!tCSn)cstL zVF99RT}yh)#(rD=U^fem!*)5AGN%vaKV7=IUFAEW;M9^U20fPXD z2gC#u#NY5#a={{+&lA8S6%|fng|9i{dB`ZqlH5sPgZJ(t8%!@x4)9U)vE7&1NwP2l zYHn~~1&nATKmeQ#j~=|EH$vHqJz(WKiuFd{<9JTIOM~nKG}fOE|4eE1&G7J{et!udYDh@kCGGrcQ~Kuso2pN;sMi?j!@+YBDa z3$zEuLrnUb&({L9nrKU8^B2GPEE8V@OGe-;9Wh^qAY)#O?_su|SN&{?&t(2@?cT4+ zd&ZE7&iA{pG7k0FRG|g-D=C~^YHYF<2<3nB9qr-m0yV2*Ne_Hw-pty7-w_zyea&dp zTY%~_3#2)Z1BSL_GER?Rlf=IP1TVBdVQ?awG_MCJ<=fQcYcA9LlKU=LlBY*5&uJ*h z=7xVQuuJg6VBpPO-1w$vAuB4%fY@p`hRgUd;z-%WDQ|iO)eN?+9)m0$k-|U;R~fO zp-I=fUPM(>g6s4zLOFbl7w3l6!XNjx#Q+`nRSZSa`jK8JtO7g6(cqn=f z`?92`q*Ua+uld~nkG3~~kD|!p#}h~(2-pFM##J$Dlw?;SD6C`!NhHA@nMf2+5YZ?| zKs*oyX$!M>eZ|FUcI^u|1?(9N-LfT7gTqJzb^9yx4iYTZ}|)RYC#PY<`%HKKk+r( zVEFssHFS?^*Q49%pU5dViTrlql`-^2aFg!0Zur%@9PFkiqgBp`Fxt$AkUDankn~&J zm6G80Sg%!rR^G!a#&8o327cLlOYL#HU9kb#@wO+tAx>pu_?v+U>%ZeaV5z1GhhjQ1N|mD2k{Y3yX?UBggJVt zO7}fl24p}hCZ7+VWt=QX?_TY}hGCdM$^#zwqA=Li-jBIC2yv@;i|^+)lX_xS?ftJ3 zKg5lANm%e!{wcc}M2c&bc_G^~q6by!^+#6m8sE>!lTv;5`OIIRxFNKfcv{6z=TySuPFD|G8|s=h24hL9MCDZD8ZP$2Xw!n zD9M<;lg$zwIn;}MMHr4J7I8ry*L&b3zLKKc%@ca!=!^`4rRz_F%}9j8Ib8l6NRcI0 zZmg~o{wspPRK+zsMTvWgavkFy+35pMl^m_tj}d{*Uf5W?j}bS0lpT2l%d~gGF^PdV zLUE*_QTdRdF{rpG_xrMQfIKKd}@9`bNd3@}>f|MM}m%pkr5K#Di04HE9a^B#eVs1J#{Jh>- ztblYvsl2rxf`0A>srx*2xRY8UB)L=WIc$0vW;AQ*cE&$blncctzaa0%SyP z(81o8BB}wc&LEcVoG4NQ<4>%Wy949m-i7pb;#=(x{SG}z$p;u=E%K`poVCaRt+{&w zwPs;|>{7I*COSTZ#tvxnqQ{oVXdiz$^01mAsltS2xVp~)RCdy> z=E#Cd94ldm>ji%;XXr{)mO+zprltKGawaqn{Yc5hDaH1bHaI0K0Alz_X3#nL#Cufx zFH`z+uyktWcrz+5<~ksBY?mz}FIkyQ{bP*&Ej=-*@y&QT8-0xSs6^&0 zRSAdQbbP-Z)Zb^R?=tuMc=dg<``t=?U+vTBe-g$>{#6sj*`>bcyWbzF@80frExw8L z51c;{;1DTGae0)lp?unDzJ}q(y{%w0`RuEG4JX2L3wh0nvFMCa@*C52(Nh%Y#kxN_ zI0QG&J`+b4>z5{BM{aNmyWc#|`C#m>1pwlG2t@+!veyzQ)E44Qvt1JZV=5IyETJ4BFNWp zI<5zY*BK|V|&6|ddf5uE$HjD+&P+vmZr(T7{v z4F4^Z;f)3({=YF$jPb0PVaTe4vvA23bGiBfw6*t%9q4PwFlN30MmLTBY?!dCVoEAt zKMoR`kzCTWt&pE>v>4fZ4SzNKJnkAXWtiekUay5qlgj2MNPJo$uCmF$w#L`czHB%b zZNPB-Ao-SipGc^fg7qUE`ehh?Ub0w#Q78sj=srhe1&s1Rz$3x9s+_1FLXDp)tEasP zHn|JSptb%Jx>O@w1_48^Wf}f8AV4f&9F_raAfY%o3_rOr4tUvQ5NA>=AR*vE)mxUHO)ic%1X4i z>~nY@jp$i}r2I`AU=`4}(g5E|gE!ZXW`dUm8%*H%TJggpFuZ$Ysk5Ptd;g34m>87f zO5kAWv4>+^j+EA3YTcBT=*?DJXdtUf?H*syPX)V<-j=oWyKMF*9+ElmzyRfP;fc{= zzoPzFSIET;if(cB=DN#f>-{<8KFoX68-t(5^hMqf=HMgEM|8RiSt~WiMSR0|Lem3xq z+HXhsn_5ke_cw*BeGL~FGe@9wuwE9-lnCp=K2!cDZDqCw+T?3#lk7x{I}AT1m-0dF za;}Q<9XodAmDHSI)VdIPy&LWsr2=duX(~@*s-SXv_|y2IeE@AFLiw|5MR;@L2Yp)i z0gkfQ%4*YsL)iM@ha#Z=N15kzcE@~SqjVANL6(d z6|1L3sOW5y6uJWMQby|v)Zm?XTMx5*E4ctBhSQm;L6ld%g#hMi6~mt|v`=n;#o>gR zAAAuVIUAK_jECGqwy5ZH@^{0oXbS+&2grlT=;O*Jl;6OWG{{a1(N1vAd&Nmt( zⅈHy4RrGK9tTzTg zj{)O-Q!!$6ZdmOhxz>%)LpDr&&mp(-L}u(G!=5ldP+{*!@T2iE~CU^ z23l>`{5u?m@*8QQ{|x2qWFGe~T1(^@!qqn*Aw1fgcPKWpM`h;ON2&;FrMh0!nuz(V zuzl6zSYEVh0n#Sa7UDGF%4H`oTi9N_8B+#U^8FQDH?~qv5#)&~JTs7 zcvoLx-9tAX3dJRc-4g9lE0uirVtv7QHe!i~?ap6C0p`ypdve#8-QfXApl%upu|uvL zhf|mqT&@8u;S6B=dj-&9yn0X5DOiL#%pp1OBKaRd7mJ%{cI99pR+}MR1famz=DcBku_8#TPJ6p5gbTp`s8dvf`v05y&&GRTYQ_vxOhvBmmAB z!M$UEdqH6e2n!B%r@zmaGe)Y8!}NDy3D&+0r8qtf3BntfvpWXXn(u#Qo^gBtu{^ov z+K#j>^YlUkEYj#@=VoFaypvQhFTuBB8FF>yhGAbd-mtqOTn0;K(B#^r7F~7iO>5$0 zQeYA2*oXpeL#Ie($ghjAlSq*>nw(lOO**L?N%BCptAY5|F~>vcFG(&-J9#ZDvlsAkAPH?K}Jfd2|Sh^qc++`&_j(ZeawMrW^iW zK|eE)swSXOxHToqUJWfH(f~rBBZkxz0(Y)a-lY4Xx1F=_N86E6e8C{i;5!EdQrxDH z*7)LNv_>dDy(oe*>|COplW1@GtP#rV1*I62Da|k;4aZuNi+NkaKL^Qt4Y=0iQ}F+9 zcNq2*RxpCE6L4{a|Ep{ZI)u=KlJU+g?aCTP1SNg(B7RCq{p5 zz>FicK^q*J6{Ei>yhM3D`IsoL;eSMV;SfhV|DL0j*S{W$S6+UsnL(F6O1r!W{)4Kr zPke*RCJRs>-xL#0==NZO7?e9|l&`ow4&_*tizC9Qa>xA=pM=m4Q73dOL4lKZ%D+lhu2&|DD_(s4*8B{N4($c+=9C(VP`CaiIBM0_E zuf)rNASXM$U{YSI_HNXZKe~pOqCOkngLO-Wf42|{ z@%$gOKA>5~Uj^4&I^_{P86Q4GxR>%bKKvTAuae90;U;*psqk~-*)nsYBHiQT<4Fh6 z=#uWd2mT!C-k)+bTXwsz5$QgJ)@`Dx`=X1JbkDyMk=sV|01McG*G|7`?; zu7W^=TGL}%8i5~xjg(7?)e744604w__~Gg<4P;SDIfj%Cgc!@}nwL6zfU3%{Ro z##fW&vOevzE+Rbk*_&ra8<4Bq^N5kvv*b3m?>F||XBe%9w-`{Zeh>wa(9gv3IjxJZ z?E|FPVXZ$K#!3@yH5aH6#K^^LE692^&_NS_awY+}7M6`?a(6AiMV#HOT%8ayXB}dmk;MXXbd#|2Id;M^`~C|D)SO zbPN*FK`#NWvuWB<055|5;ba1e7j)el)xrPqz9uP}4=+6_`r@KJ@TaMj2UsQdE$4N^ zX~RFJN3%Q~(|H(&!-<=YHFP=^?ls_9Z$maI_pVCg)Eq`J3d~*w2n`i*Dg&-U zt=%lgGQh*amMf`2iHZZ9FJ6ZQ?=7{b)!B*1{d?*(sLZ1x-oaXWMl_9x66;Q;)W0#zxute-Q!wH0MXgHh^O}wvhDb~o zT<_&{!RZ-IAZ1kcRB0}hG)GQtGEE05qH)*oNi1P zoTPwYiucY7x$(Wmc|t^VO$gUH)Hv(ahr{Fla16!Kxyy9!`e$X^e5w6izS`Y|{pvNL zb7{f^UPJj-@{MDfGhx0zLz^&n)F~6@O5|`3iCV7-9pj7!0pvDb+HDsoXeT!d?I{=9 z-!!x~$U!_33H3jKDu+9V{u6pqZu@jP0et8IqANIeiebea?a^7rBTG!!n+&^}U^r{& z?BJjE{PO|-)bURt5#~L2jw~_b70PmFyxg`aS8zH{Fzghj<9YTDCjgxw{*dTl+7d)Z zmVLw&J&;9ziCAY4(~snzEBI#^|9J6-*^Y5~@{>i!0V?8ED3f3~N*Dj(X{vh3>JOgaT9=v(9!P7Uv zs(K4aJX$$tz!ahzKOrF_A^a}v=fKa_G+d=ueaBDP?|I&@4;`cZRz-NL71{3z!yVx!d15V2+Dm-bn*9%epI*lkAq>yE z1%Zz}Yrn+?h+ozoYH5z##AEr*mhGRdrKlK&6|ik^;SbMV#0I598MFwuc?0hi2Hy0n zbux;KMW1*A?|O_yKjZc1g23l^-Q-yt$}rdN%P9*CSti4VBuew z-$(&qyHV~QVzm}ys0ENoim z!AuOS-~C0W9m=2H5Z?4C)&x*g@T2|!4rlomW+Fc10?(Nbs1Lc?b7nd`a){w|oyN?8 z>Vx-y8Z&3854kXO<{EzZ8b%s352?2{#>^ay6=0Rw}ucB^kd5{yus?*DBU)Vxc< z*e#64wd@7iTk@An(m*wsvnX++IkKy7T}}Zmu_*|gm0#7Vs}~1j`+8!Ki?zdc{_@Rl zzx_7Sc#TC{JvcWC>DXJgc@qzS56_TUKHkaReJisc{60Uh#aEpJ&{c()`6~>g1OZ5A zoOdiTo?02Y4Dd881?EgniE)WvU`sP_j<|5RPDkLpUvzzHV>o9KPBkdmES$qG9OBG{ z65H~Z4?E%Y_W%iOxP{dNW(uow2#t3OGz;xl7aEWB<63JQaPrrFi@RKeQh-x5cI$?# z?m7mv>Oyeb2G%UNUtDmDkxDA*+i_CA#+5vmaJHkJn}zeU3umkgXWX0Dj%y6(?}Rh= zx8NLd;hZLLK&-$8yM_RckZG-*Tf|O7YGIeJ*G0#-C#4Y|xL`gfbB=61S{Dn`Z8(6V5iX zzc5FBU#y@YF@iW0XV#RP0n5V7<=_U?HTtG&ong~E@GEj;QkMMaLJc(bjvNLwO+WFY@5;kOzN*ZfrworJ+|6))o$?ZS|x5f~=`2I{knU7dhZA)E78L4sqY<|3SFKuLKHFgXX9 zFwTV{=^{|xgYB%TVuMffSK$va5m6*C&H{|H0E5vk3`rV+VMSq#Lhk%kkQ{^|Vo_k6 z2^eP$N7CT(A!;`pfAY7l&LLMtG(V2P(gS3e6UfEYT zz&CQWTo=3obg~>v8rRerwqw*{) z-QW&ddHjS+eZVA8Vbu294If5AZZpKyam!Ov-F3a>y#xl;} z7)-s(Jw;aX%ZTZ|hgJt>-;+B6w(>b~u^qPkn88;|06bF1jTn!~5(MrD#z8K2@zvbt zBl~HC@RipamHl$)jO4{AudOUWRHqF`&c5Oy2;78HVb5V2hZJ6>DtoCl^5`%KlPFTH1G z5QBdlS#Y1rISG)zF6Z(;FoV7h3Hf8jdqLjHO@=wv^TW z;4oI?BG~t*GYE%i*|=wHz2OhA<2TPPr^25~ioyA8pK%q|@3TR1atF3A>XxM6BbURm z2#_dSY{JBw+I}8s%pJpcEGw>Y^Q{Masxuda@!)<{Fr*1l9vJ_!s~ARdF^iO~N`_-z zBo;0}hu<{SUIzIb-wJ(QK9*Ln-*37V;`2O;nd+nZzuTH4K&a4I^mFuaptcmg3+Y#? z(K$mM<*YkUoF|Cm2m@cB5k9oD|I-+~D_nWB3wHmf7#(pueP__rChWovO193zbrq#~?M|yF%qfzx9cL`XN zI>^#s({~7az?UKtZzGKvD`mCAfaBEF_R|l^&}4>}@zW2C>#31mdWcR!=VAzaa zA8oj~N9b%d4>|pa40>kFVyfv!WH>ZqruvY9(u^A=KlJKHo`YWo8hmw0(64QgmII_o z+FtTnY17ByY+0*EFVBja16fzkUUwji?SlzH-`*C;(E^IubMSwnp6wRiaT`bB+;(sF zMo(Z9Z*@0bPrArl>$DghXoUl^i20-%yr}xrDpk|z66l+POA@QDI8xOg4URKFz_yp9 zfcQOzqEU0-Gb)3^=Q8t_J-BS3(V`BIw(F>|3(^SJ(3Eu#A}tOr$*BEMF+0E z5~WFJRtifL%XD-CDvZj@bw_h$MdB5ZzZ0(z`3t!tDyF}6c6SjITZH$T)eyg;$GBZS z2Oyg2*c)^j**lGizNdzuyp~9;l;A~>WPin_BUq{dW6Q)*^JeachX1elK-akki9_eY z==GyZsj>ubN|H50lz8AF?vxIa^;^Tg56sQmEYYLv;1u1P;*?O?>n@P+7G5;Meu`GQ z_X;MC>i8pQ*T#MPBC;EGysQ(jdUgbY&Iy>sX`JvxCX8z)oYyW5;B7>Z#e+eR?*|`E zkc(molC21G;}E#lA~unTk+VS<``mXaMxf~$5##$9V%&YFCWh8OW!0;h7c|)5$_tt{ z1?--U7;k{^0jdow&`3Ss?T2b-9HbMS>$|*GDGo&}uEynB3lqz5p+o99qm3(*Kc`Bx zZtcKpi$@z>*Bbw93u$3gCgblLIbFE~xN;;Q%hf2@SX1VuzG&y_3)$gp z&%v}xyv#qI?Lm`3C|-?a*GfBWh-fFkDqkY7pz?4OI=4!b4GRFvSs}f;0;m8AnfviU z8qXhv(ExY~@hsJVf#+tWeD5D|JBI~}@o+K++VhzE=V6rG-o_$+9#A1Uj!bKoW z*kJu#1s<`4(p(Y<=A!_D=2m))xs|H0caajlYp zHCpB}$W+kFS%Pk?_}M)833(WLt>YRP#nfidYxFBAy>3VI zg|*GywHPfttvFU-A7%Yl={z*&2FE3h8f}-kaJ4l2bRgh%L1K1{ipd%1>%)ZX7r4g6 zdA;3c1D@yuj}{Z-2pin00j?HPY{Z*liplsS-YS;POKo<(&R{Pdh;5$7Xg|D5k6GnS z(_@Ba*yRvw6lb}_`&K}R;yo6m5>m!<%tdULU~5+nwns)KNMeV1+*?us$cUm|Svt_I zqu5+_6O0(56-CjKgIvzvM4}ZiDBe_9FUj&!+)i-OrBIJPbD3+>(4L_!bMPx|$e82y z(~yA)13d1=51p#;ke=@2Tp{Z;8C;pAz^4Yb> z99XMd(Z(W>(>cSsEXQgWNZeF)1=h-6k(9j=PQa@EEuBly0aZQR`X|FmKC1pVVfBTs zzdKf(gBeN?;uwi3!0j!RdI(sFMFuKAMy2?krN7!8g?eY5mM3JJN4D$1|83r+Gpg?!HQLi`HU_Naa{3jkJgROb8u z2@#t=zJ8-v5RA5nDIRsnhQD+T*#BK3jn@Y{UGX#5nYbr-ev5w&=q3s97_^k}uy7r+cgKAKpkAi0dte043IMOd_8O-^>6m1isG2|QgO zHDFk=wg7Mrq*!A_S^}VwaXjrw>>Z>$79z)m@C#Nzxa{=LnD;$MgHoaQx}iLA8lumh zQ;^-jbH`@qftk=PYx~onydsI+6IVkZM=RtwjDBWKZjHAe@Mc;)K&}<^b}i2U>Fl4b z0{0wGgJI;BqLJZCth>%pNMilmv)dDvcZG}^eNkX=A1uEd+^2;D%jjqziwWevKo}4J z=TfVzPl9>yQvmCe$OD7Zwk+D7kYC#h(r55Fi4G37IvU2!gwYL)MDh=QfhNbv-$f%2 z0zyv+ultaF8Et0MD#9?d5c9ay)&%{N-zvG;(V3BazMgPTCPu$-Yp00dSqouYj2>Q)4-|TY<&*q7Q;K?GbF$)GwkgQ`|o=iaJ5h? z_c-ZkOZ4JI9-xUcr9A=Q1kT3YV=V=byZ`bW^z&#dw%eNL;lQ1*&by-hj_7fxC7k?b@;jpG4w|j{M z?|&l%!~gVjm6 z|MdxMZ0D(*86*jNDu&`58YIH%%uh8s;NG?PbH1@&k&TDnG=hQ00&6!*UIlpqIF<81 zVtiE)Z*22QoUk?8hS6n8Np8kabdNu7EyTLe{+LjRIkxi|EcPkS;$okkzy1K@wkFg8 ze^zGzI;dZ2KmPb5^zZ+j`(86_UkJBc@bKB^{sU}yuQc~8ma#@Wd-FCVF12eH#bzl+ zmJs?FPU8HCnV9>Y&aubelRvrt*lYFq7-Muf_bjW?9^|3mteE)+f2N55?m_meZNJ!t z{4yzmAHx>lDLawV)K27?-b76JN7|c*AM8)VUH zbU*TVu3W`B;tb%h`Ub!up9cJy6?o;AK~z9|`06lQRp-eg%MTbxphV(aomF!6Zf@sw(l?Z@nfuyZUxDf#j{TaH*7w;ako z;mNTj23ex43Ow-+a73^$g9|517&L}LU`*uBWcPgK5+X3h{tFSXq^B2vzk<(`6wzvN ziV%W^=Qtbw)g&E2$5*SBH!wy{Y}QrUi2ZX856glb33vH#(X+m_&b&^|h#U z{0wFB>;yk!VrqERUsa>{{}Eq98PTYH$S0~A{u!mU6e~bY%P%$LT@x_FU~kw)Tpe)* zOj9j$lCI;uccG4RWM8LJ4bDbn5SiW@;#6($z#1 z!1r7X{Nbxnu73-pW7t}RMa~eB7Lj)^mjIwzAt1a;qzk4R1bVx4PXG!31#$W9>s5R55Zl$ zhd^#u;h6&VwA* zin9|#@R$}Fm#bRn>%Ygh5Kx5uE5`)fiB5%4#{kpaqx;*uT?v(6FLjkJ{D{~Ot{8r6v92?12&3x#5 z*{V5$Io++Ahse_LNN_~-21EpZ(G8<*T`aoCVlf{dk^Pz{i|v9#ak))1e#dAOjU#A2 zL4yl_B4~7EPt=<>>(MyS-{ZUJO`GXgL>rHrHqk)bwE6t`Y};?_T-lVU#sJYxn_>IA zn0^FEt`L25NNg)5Va(q|D}4mq*h()>(*4LCkf7XQy4?GE{AtG7(KnVm)z-~_#Nt!w z2MH-s=J{nPQht7OLy^NcD-9Ty+WrUCP~>{#a3<5IgWX#3254i0@#4>6jxBF8X;UV6 zI+F*ROy1bE-SZiemM|&o>hbnVXd0qJQ+vGNaS?N$E(P_16Asg8_Dwku z3B1lKi1|{;u{;{(sXXim%T3|C=w${vWwIOiuLWPBjqQIT2UF&FtudIUL@Z3x921U# zUS{H(wJlJ?i)4_&6P)^a0-Hr1K?3xu1psTZ_#MU_$ez0W&k0gu?^A<+UIj z6(W22Ca-lP>^xYH1a>~`Z~8xaXDaar5{L@~egkkc#Nh^bG1k+IF?GLPvz=ayQJJab zZ^MrPbOl2>YY;tlEvi)3g_vj2n}H!{of-7t@s1h5rcd6SIs0Di%s?ec7kD!O0~y4r zkhqTCjKF27&Q|!U%pmD$pW2PDOd^yF;@~U~#3uSXhS4S#?1Q!-C<-7!f8l5m1kGhG z5>&>*LBzt%fcx>4&1nj};e6FHi2V^^THNT6etnm1#&;j$<5JoJzO^MhNFGT1rmHZXc$G+TFRjM zSkf9dtrvOu3`SngM*PuKY18%3P-QDr2dQ!fb2UvBbhj9)T;futf=odEDF1D`oNGV( z$?wpmAC#z{-~`^;zeOLkHRv-$ARa}ZZSnMho`61?fpMC%5Px{fbpSi7zp?w{fJFya z=!Kzo+5TFDMxkJ-I%LigSr zOJg`IccObMMm*(LP_-}q#275gbCki-1MKDMoJKqCH(max_0DLR)~2P#6YRYrPskUm zE{#=&2Y0iXjxvWgVJtvp*n2$p=!vw@8|C*t)M&FkeAcp29s)&gl(z%)d?6DYuW*_A zVF8+fb`Ych0l^kd`>%#!f*Hg)t6(q*Kp+WrWCDhYVj`ywFk(n{e>X+4#B<_F7Gs4! zc$ttadaemfh~MxVQgxjNo4!es{j(^RWb4s0njo1vI3jGfxz)A9aX7$)3BSLuKj~j2 z7rTWV3-A)mSsn{8;ga9Ts`Nvty-}QuAhpgMq6HeW-yYo`H#9vK8>F&^YPfYhhw@|@ z#`Ou#*}ouXd{9e{sNJCaRg&y+8f}rVl@eA0epXw9_d4HGh{!hKz0O}LaF>fsNvrbc zUA(;MN4&6)OcB%P&aTawJ}ti0rcc09rcWty&}h0=aw_kTeZ7zH8ZmR(WZeHc2Ta>E zCW;)+Vo?r_vX@vljIIpeVr6N3(Uo+(dw+dH+E&8-gH zE)v0K=)=(N5Cmti0nH{QSubzkP~d1D1RNO!t#5f ze7TTJ?Ma1T#uOg^@Ge2MCpCzL&h?xid!5IxI zGWeyoATbY0t;dr;&QLH+CbPjpjI$+%s}o=p_JiEwQoemyWA%bMqZCO}Py~On_mz24 zbpYSfW8OJ0!5@ONVuCrdpo4owUyEZ>a7sX@DABLA{>;gIWf^mH81b)1n2|jzGY?@# zhgq3p_{^S{S;6NC{>+E*1TD%=hX|c7wPkAN8f|=}KpENdC3!*BlFS82&XZd*@Dvv- zd^g^}%FJa1gQ6GWiP4d~zl?YO?80kRXRF5t%W5yP_7L)*X?P34B`~|(Id7*w$BRXG z8kO%Nb;U|mE2EP8@+zh#w84E$P&mfYB~la}eJsYle+yN#!k0yKWDM)576{{fd zVF%f^Ks&z-+^GxBrW%<;!r;sw__`>aVf3Wf#b+sn^)ln2B2(nD810t$-zB25l;eon z3e3qvDP@*Rj3&-+<=DaQEC9}2IgIiPI|Sz_Xu1GsC1{Sp6xU2p6x6r-+%6KKbGDKQ z^(@(^X1Lf5O6_NMQV`(&_j9V0L}<$#L5yVVM7%{>%YQjKD#kmq>^O+iUJDa&gE)>; zHR}pxw8JT!9OBu_-u4b`Avge2!6Xik3qU@-^oe5@C=fX{g)ezP0-RNlX}T}Pe1SJ$ zC{6}1%N)iFfJ^Kv{--&}LS)5Vh@@2mlMo0Pl~<`|H2nQh7FlK^0%BUuV!UcOFh{iU zoZizPNtmPO5mcRIAl+-V;&)XsQFsy{mHTBg?Aw7*blR`Q_J2Iu{;0Cpani;9|z>$f(?w1!# zM_W8Sw#8w3V6xrb^)~w|xj~wRM|AO^2j?|04jMHG?_@tS0G$lAg}KFj_6Ik>XlmU3 z=L*Q*?*5kQ{&wjUYN~^<2iX2AMO>S~B_f&8;_TwBa3lRRH78pS9y(fPHNiJ<)Ie)P zjTeRg1#CBjCJ<#W71#@eA;jH2L5REZ%;tz&@TDg1TYg2{=a7TMeUdq&L=!v>Kjw(w zzDP=9cOJ#jL5-6wSk}m)(fmG*-{^#z%!Z#wN5rWI zXO1e(S&nB^`HLSRf3WCZ`V)sh@&8~>-<=`7IY!}AQu8v&)M2p`l=O@Ml&b(p|eC-CeTB$|;dbj2g-) zJWKe5n3~LwyHMpYgl`1{sWTtu3x0g3B;b#;=*#eASTzIiLmuJ|=JIE5;}dc948V`? z#F{_8?^mxl$!6Isj>5>phCd0kFjk{#H)W0ikOn5kY^X9>5}A)(ocP(@ABC6#<|f9> zyc4p_pa9Q^QKtZsX9hl{2WOu>fkP2=CgSYP{RaA%l9lkS@rk8&!d*(kE?|hOLs?g~ z{$UDZIn|D+>H#f%IB#!b5N%ICtQyeIjJP@KG>@l#*HTI z9ae-(OBO)Zn}H?LM$qn1;wPW}L_%Oz=1L@pGd#19vwO9DM`v~j_)Q={H$QwLGGooq zkIyuOYb*8xL#&O%8?R(Yv|qY~!lLJ{)2J1~cHfwsbKRWFo6K3Ja&FW)lVfrYb8}WS zne!hi=d(KJTd+`EY)^G_{)APQb%HIXzloZQ@Zp;&t5C5{r=N_5SH>ln>S0ueKQh;rgdwP zc6o-ze!lcQ`;BXXJxEu~WSEz*(t1h?=xsGh0FT=9bRww;Coj|zB;lp>jwhhDv&3#@ z?NyptO1<&Y?7F`YGIXVo5l?ZALoqY81O`D#*{p{xq9kPD`K`nfr(I+CKe&w@g_MHe zsKUID31(%^cz@;?9Y0j!h5ic(|C>_cJwLNb$4^GQmt&XtBFXJK z#IxLoCH$cMy8d1!-}~%m_4hXU{>FYtzQLv#GEm2$zhP*(GP42k!51=ziiD}7=3xk; zmSmGPC%E9a(I2cq+4(e2k(toS$jR2ad%vI?T%)Dj!&)Ph^M# zpHeoSES29s9wSi}LE*tbi_65$F4)-&)r z1d=O65ke?z0qTrQfd^h#0#CcQ&bd)XW-+jf4ixo+93}%h>cE9Mwl2Cz?*4Xf=7F!Q_dyk|z*lGqAbz zQ<|F1rN5L9o`@;6Py7*c zYKT%|Z(e2=u0 zM2HRwwIBq0#Ynb17Jt4XFRbfAymiM_h0qT#{`Xj>Dz%fII~JgZrn*}dX8uKlI_&E= zX}mzw2tk%8!ce4o84;+VVFts~8+@VVe$bRN-~OGaV1-1*)Ls;|=uN()qAkU5( z=7G6TzVVBJ#+*p&hV+CjX|EWnv|qh& zNm{W);(%W1K zpU^^Eh&NO~HJ>m;=oR1(@`+K{E56q7y}%b}#|$)BW)^dzACk35&+NC%+!~xdxy|(b zVi+^4Peicq7mTfEq_Y*57A90F)-;3UOXOTAe0qf4fyLG)r6<^Tbfm?F`;SHuPnZeB z2OLCvz{{a-vYyULzON^Sklg*-iCp+jE}HelIW(B%@3iDd$e0Kua7__kQ$)5?GL!6nnp7dOw1%v z6?m4~^*3^%eh+M-(1}$eGLk0D+D{Lne)AMO)KgLM3;WRVJh22a6ZdH~(3}k0&q&{6 zH4yj+KLB#P2p{b^w9>ppRk_ZJ?eZ(YMSw}XS%%=s9yDv85=JWXyv)e};z}KMbM9Wq zLIY6f%wYlabxk@hPUu`BGALf?kb8rTgiU>iSYdN6fdK71DG?m4M@Q_=yLKsIbLq8g zatND3d7-emK))5+-Q^9kQ0k4cq)g~?exr@@NPW^4@22}6Bl|9kpo6-VAQQF>bo>`; z5ApT#EAtb=)p+flXv};WneCcuAmsMYWSavyz;>6O1=z(JHrm`9ctH&z5j<1}m)duA zdMu&)Ph!Z3q_&dp9f-Xx5CJq-y8mQAsZfECZ~0!O8A=Mm{k{XXOC81_Sn3lrF>hOp zS>1e_CLGzE&w50nzb45y5O^5j{Hb*FWB`m~&O~0`+z~>{O0KZ+E;zYlS6lj!}G5Iu^gc&6PE8crS{(vCsekVlM&)sx5)$$%4nt_9 zLwu~sK!Mm-FdoBWtLGP;WE=?BlHDT}UYM=daiCg(BnkF$Ba~DVP7h*q&n8=dHsN?l zx&ef44tJghJ~Jz~mv?Z+Vj<`pj-4;WAqAKY_n@(|JP;^gtnaXp#a_^!^YGa0j)U32 z+o?D|a|_NoDQk;eO`&&`KhqP~O}-za_7wmkGXM%~5}HJSmSYbE9QYopc+Qb$B}UxS z@E?2*Btxn0{(#`{bf8r>0N2-Nd9kn{I3=Sf_khRf{~;Eq!Sw*-a>8i`Zrz?xSzVr1 zlzm_sYF&Y;{Hg7**Xiyx1p&LVdh#+e@VOZr4wu08^4A10o^a&<5Qv$Rzs_Xw%Q-=$oafUo|d>`%RYlDPf;{?%f;D z=mOQoP)sSt%&7JZ zXG1>rI}2wT$fA$5lb`%aZCr#=lf%v3{?(j~r>}FV66TQ zS;3>@(H!>T+QV=wyLJmqSahE5KS8aPP_jXDQ^3-2pm3$3pq>~e@*9#}o=OAd*e#dR z#l1vqOSS>{4&U=KtF)CQ+78NQ1y0{03MJGy3lLdqr?l5n`N4mS_d)vmFs1Ya2dihc z)DhzXK?Sx^0h*@zMi7Yy%FmMU+Jsi>SS<9xJWSlN8tQO)RX4ljHK6Fj^cl>U1f%jj zwlzMtsL!jU>5-TF>@*QbV7#+b{_M_xpb6ye4IFtFL=}$L!nQzGUyTE_`HMZ8Xc(&4 z4aZQ$_MCv}aMA^8gfQ)ENDNStwG@L~*X+SJ4=+mH(mol%yAs6&ue$i|#dm%-mT27_EGFnS91W{Oy7=)2pet{Wp z)N-~FU(3m?JeoWYRkT-Z`a@M)JNo~iYSUi*y{e5sibhrY_nmQ7OJS8zwbb~kC8??{ zgMU1>0sqBdw`%np|4`L--Oyas;ux2$rkm{9R~k47C6_>R1~6WY_B7SQq_Z_L$1Y1z zxW)9e$J^+hcJ>pIp7!Pt_OzaL?WJa#IISWqx!zLsz)%fvX+j=e0htC7l0f-_j>n;r9P=jb*rz|I*?Te; z_`n7p{T>Y$82=b7qkerr#?v_lMdwfAU>w8Fp0u$5jdAL*IA(kVWtbTEund|}JT#aV z;)O`%N!*Q{y5(nu?hrXGwlo0rbpe#x!6sK*^>9sLtqkZlntf52B)LQ0cIZihBm~(<0;`DC z8jl2gR^aJrc~_<#8EH|WAdo+Apd02vT9Ky7yZWL^peamqR$td*3q!SC<?fbzeV< zbXYusnTGq3)!i^AYJcvJh^!PWU5+sPPjL`~IVBqA{v;L#v|vksSNKqdHp`TT26`MA z@`rhWatcH^YLW*++Oci7n}L2fT`hU`Sr~z%z0Cl|xL653- zwbyd3EH7NL{e*5MKaHetk0wFTk&E(3N8uvWN3l-JJ>U)iD}gTn%Q=u|EPE|NGE`Hl zUI#A0k~a=}B23F}5_lm=Gvx|!y_-Qnenj!`^+cNWBzas$Rz2lnM#T&;r9Qxnj+s(> z>~3I&mPwJJ=e`O-?(IvF1%cZLL0HXAzB7mfkmB_$hm#}n#H0bA04l>`O7*}jHAJ?O zTN7urUSy=K0dnA7YJZaan5?Se_?|^_KyUYc9{0y%tp#;015E~YK_IKR5k=XLUaH!) zf**%3V>2jcmc5@JNG9I5i8`6M@J8#8EzC1ae>}_&ul}(4@kJ{zj}KMBM~3ueyqLLa zZr7;1^lS3gN?D0!pYf0?EQ<)X(ohokai|Du7MEc~5dCK0a%!uq*+I^TC<3^fO6?U? zC0Z>ym4{S%CRu$oiY>LTd+D)+(2mG>72&5#xCi0R>#^xocz1?l^^cTG9YS)+={6!@ zPj8D^I_(~yxq^0aJu7^7-6Wl6ns7`1OD|)1ToqZjGh$ z#O!ykmug}VF#I{yz;UhKAA#r@59NYig6tVqF};j=R5LPgA=6^=3yWmFCDSGsF(nFR zV60A(#9AlGyOn$odowDy5CVw52v-Zd!7WuI)C2n_Js23R8kI}oMG^l2+z)Gt0i@Ip z-AXBqamP?#i=ihdq+ra`-VbcJhK-N0hv&!Sys8(ZUu$~_pjB!LedZFOuZFpHk7u!rZEq-2 zui&H%-#baAcCW*aCOF6MiG;hQc!T3%bqq~`zr*mul?aM{PKS32;aca6MNRTKsb9gP zI|~p7C&WnC`(RunK8_^F$6u9S|uodN$v68s`mHsZ0^K0^6L zm94@BsaFQ|ealy0a+C)$s?wBqZ5xmCt z0M19j>?>l!X=6|%!vBSE?u&VWST;nrh;xfCrvNhlMVl>V*^x>|4z~ z;SGQb0z)wy&^fk!*qu3R0x!?A*P;So3-(Os=5MgLY7XG#p-!C>aQxqXnJm+@uGnK! ze_})o)!M$-=q-6c>gFtntZziyZVf-~`YpE%r;g$j1^xLo_T9f6!7}gxarmF}Z=~1$ zE#P$A)zXfbu;|xd_JN;p+O${9KB!KGF>3h7?NS`#9_{S-#_O@aJ)_u#9~WP~kL( zVhYRJto~|w+oSTNesfxy4BFtl<>9l=nTT)aJPE9EuF#SFE>Vfj!xOyd4Aswzboh<> zd6hhCoZIyCb{+21&xv?4|HJxukAA+apO4~6JU8m+EBe_E@`6ZqlxK}I7RWmHqLrDr zn~pSeWH0?ZK|gm&&1#%$_4gC{d5?bHrk{iKGhaUk=;xFAxlli=^s_=g$Lr@@{ajtB z@LH|E=jrc9^s`)t&(_Zg`gu9xokun0-N^Ya5dHM)=QDUZe%+k^(Sa_8uj?BBP@pim zQO6I}PfMP4&K&*yU;TZbe!7@H0X0JkuP#DihA!+lU5z6e_@z2giVnX+f49?Zw@IfP zp=)-Lj(GQtqx>PlU(grp%ME%=Tnu9MsV{@hw)Zk2?|l0Isry|K#K_^q|+G(1Scp;eu7n zF+oNtPx9rLVR>{A3)!un{{%8DW!;au6!;vDM=3D{06hH)d2Ru9cXND%0z84*#*iC| z33)0ZkJ6CGxRAq`f4`+E$h&UTsNd{DM$0sY{5T-93T9+2VtJ;qv@LYN@eomnD#PJ# zzV$<+ixQpAAL7WRa9Z6kN{9`PA+58)X|11OutaR@h+gaKm|2|@=4ksC!>E&eKPQ&A zHUsB(udbQn`?*CqoJR1Mwhrx+eH_V8rol@RXpLC5?+~wbYij;rEY_Bj!n&C->dMT( zVNc*kGjK4R)^@N+D_s3&Ogfhr0G6Fh>$|rZ7?F}+TW|{Woo{`LssqlCN9~d_s|r&? zUueBzO~zv59V4mUM!+#s$iuejrlZNTE1b6E3V6hVHz$M^K~_F1vwNNwexOoP`eF!d z^DHX2z=1*HVZbFhz;sL;^9{v_M~+R6`uTWobjiCv1JJ%r6;lSn)WdR0F(7b&N~*Xd zWNmrbrC5kJoU8GkuRt^+MqvSS(8mv>gL-6Rw!-Kzay-}F<)QLrcbOLSUhMK_m;j;6 zIi3ZeS2;Y!^RWeZ0wB&y0YH8Lpgget4**v%odMtiGpi|>EX-g;YJ6csqb$`L#uj>9tPcf-Q&z!$@v+?ToECUUmFI!9_ZmVWh(YmSue)yooL2w{1B68cqHX3}W% z>11PR@{yd+5~>O^Tf@)!^J&sKQld3UMorRDk4<2SNc;Z*e(rEw3_t&Y^xt*-bLrpv zlj#@#$@KUA$@I6yrpE@$&9ct|>xyu+Hu|uc`xcI>e4h@LW5Q{J;gm*q%`me=%V@lt z(|_S(nt>NIY{fGEF^XzbC$5iz&feFwC^m3BQI5w^E3QQl%DVuBZEz>&Gu=+N5j2pQ3ifRj#3?QSnpprK1Fg^-Ww z66}9prp5pcG~b^}ZRZ7nvN{WmFx-$#-vNw>eM19YO<2@BS866Bl|VjL?Ks*?~|F2Mu~KL zS8ko$ool(nb(7C1wLcu;Dgugs8L1V*Bz{Q)2(7~1#!tiEw)*tIs>xD&`!mo5F#_lJ z_0{mA{|EZ2YXr^4`YH*SM)cJpENa9o91xE38~VT??tG$_N(r^Z<}ZxJ((RqpVHlL5 z3&9sQE|s(8$708zwr?hwk0h+l0~?^NO6^hjf|sLwo$i(3Z+E|fKn|{{XHv_zpASI7 z!7%Dww)ULK?7=W7D$UC=Y2;nfe15UOsZeR=#H5jTN%MxJnNos! zjd$xbUfFDqtEfCg6}wTzD$1v#6gA>0F-D?r(F&#(5CbD{V(M@+T+ZrDawCxsK=g%Kop#Q*eSAyaEZXgdT=~z~m5)wgsU@{_x<(x_%j` z9BhPsDSX!UOGUF@|LJ^?xTqH}>4jsjFsvK|9X%yW;9~j~O*Mpph#b-j&m2rb;k%!S zZ@d?p04E+yJaZJ-UI5#cV1uAYps*NEY_P&YrrV({gM&&aeK0B2Z+u`#2_Q~3Iuwop zw8^-Dro;^JMl~@=0B{Tent{p+KCFI-1^EO0CL_he462iacmp2+zL)Ulvek4EVuu7C zkai5tU5E&yL#5;lJiss0QgN!#ml;wgknzTHCB*Vc{(t~813t-E=VKgyre$rEhff{> zi@L@P&XtF?LLO`*hrxjdbgWAIfINb8t0m5-;(XgwN)=csUv(0Qa3aVE{%WN}5c~0M zB(|>}VL)0QVysdbx%!~61g_>WNP#4UAY!3Q#Ja)~39S<<%PZZAOW>U{tx8tCPX6jd z5A%s13R<8Z3;C!MTH!6h2{EUxlQkg`3&{-MPDJGWRErZFr=pGMoNZ5SwRA2CP$gtQ zbPL=Wh0KAUjMoR}92p%L-2I3KB=?b1Zw9YOFFc=Hkg6CbiVTe zRq_QsMmBFS31?*`cbN|PRwM&2dOb4+8cu+ijpg({bv3_PW~>T=~> z(%+%dv#Z#}Dt(1YFS13amv>3;k@Qna>@y_slsrhk#@gjf!Bgv6K#0+^#~^C8Y}GHCvH^(~zqxYb<^kImrgf;ooB7#*$Qd#Qcr=K1gzt9`b%)q_`V zVd)C%z^x-+@+CMm9d>}%8kcA91{+1%`{q>p*E`TQmTK?)>RsA4dllc$ zMgeZSW~g_Q;hofb7T;6UdkWuwJx|h0d+b;5Y!7>OG6^C#v@pzPDHJQr$%LF4aA7u7b}hI_h1j`?ls_!`T&Jg0^(;n^=$h7FYoOCEUGQfHM@pat2o_z;xhF>j#mSYq>KbSu#rP0%Re_ zVdc8=oj8}HK%Soyv%eW8WV$zaD`Z?3tf?+fcPP%4eDnC$7t5Uzyg>p^7r2>vhszv) zb=lWN;^{(@q;tP905tOt*n>m3z%3Kw&B`rzf3jL>jrxS>p5y*<`C~4 zeNbmq10%m}Cb60r(Z9ce*8;Sbcmw0{>fbb8|7Pk&c%2l*3+18>Fh)hET!ZdT0s=Q# zq4Z;Kpoa@ODqlrJQShvy%AJP)F2M6(mJ_YSeF?)yc(ZrqU!6Z9|C;=3ops0oms=j% z5ai0;W>n(v(y((Y{%Cr;;d24fVS#S=h71tEsBFpfxE`T#_#mc#4S!f})HN89N3r*j zkU3JJ7&9@gi)!5iqqHb^LyjjnA`8pa(ITH2{*TdIXm4*2HhOT}P!b;-mOv>C*DQ?L z$pEL)n&naaY*K0auP%UBH1}}%Hs)b_!gQVpTU0gYdc>N+DKw2u*RnTN#^+R2I-)=q zC3`hL-78V!z|9u`G;WjzX1K3)O+5*1J+zjcS+MhrKwP9=ES6tcS*D7A1rHHiO0HfJTW;IcdhBM%vKeJ`MfKkZB&#V z%e=0lG7y#HM!|Gq51KDgeG#MR`~*wKNmrifbqVZ6l4ueVksg)bq0vE`B3cYj6G4o8 zz+RKFe!!c31WP*z4NY2^gM@5j&e5lOg5?;2sBU<+ls7fSV-=+wgxXTo3#(X;1>u1-{==20wxn ztcI3`Um0B@@-*9rdXdCC%0FQiAul8f*j|x$Gh)$i!1fs2Dc;75koOf81!5PvY>yX- zdI(X#LZmNX8-5x~5qiUpt0w`P8_N$9z`s3~@2S!(P;i9jfgODU@~+xnrKwP9gy(e{ zd6zVM&X;aog1!s>SAEyze-)L7DDc0Ea`|6Hfr4;m=qQ)}B`PwOA188rdQFt~ryv1o zQptlqw1&a#gOcKtUAhmERJ)DJZo>FC)3PcX0mg@8KqL~*5TRUl&PKvHTF$Cs1o*?q z7J5>L`)V>I@CWj18e-h=4d7#x@fN~EkGi~9qxG>ly>2Lkq)dTvrX=0=Aze}Mr5ws! z3f~9t0{-H*8&?L4?h+yVm1=U~$KQ~{U5=8&SN)v?+SiO6?v2=HivNN2F2wsM19FafGC4QXXHVqMDM& zT~t(4^0=Ku{f<2TrpF(Y$E%T0L|>M$WfSsvK2rpnl*iK%6_v*m@n!m;#}mqmx%@rF zSpxQ=eet{Ucm?@*2mZKxtmW}shEpDUf-^Ya@dQ_A-~qvGb=m-`Ll9?Rk)RItSTlIM ziKlKxFFg3eD|uU9gG*(|gwA~o1#>9|2dJ1OOPK9PyI{y;t*!?b+4u52`1s3shx9Br zL!;m^GriWxEPEoNU`E;F@C467AM{f?3PM5asG9+G^<0mcmgM z{2MIX^}Q3zl6to&f0nK|yEA}=7Ndp>Q6j3c0FT%jUf4@&$Y9iP3m7VAsn5h~sLLC`|lHhoudxJ`|0F-OCe?O_xRH!sf_wT8arpf+&y^6|f!rr=AMP)Q$Z=Eesztg{4BheqS zw|3+(;s1=vX8ZR_q-dspKZ&Sl|DKDl{~!DJ7;q`LzXXq1?$7Hf+|OWe{~SK*W=zI| zKfIE;BUOpU=BW_e1@TvWM2GDAeyLd+1pemC=MfG)tl)_Rz$m?4fZ;ASy-ML(0U5JrptV zJpy7X#HTj#F>cf1vn!%j)1f0{+6;U_a+P5L;U0$vYwGrM?IMgLc#K~NL-rY2XkID; zJYxAL;XI`u7z_Qt2lN9T{Na`CA+OL6c#Ywo|D1z*KtC{xbxnqS`)Nu?Ahms?davgD zHR^o<-!E716@2e6?{}Bjebh^y-2*RB3-;N3%82}+YolQ}z7%H#srn77SVQLzXz)u} z+u=?htO?ZX9+N#ke2h+5jRAA>bF#yoA+u>iMy4DwFM2@F=A}1YhHo#ZB5UIj6x`&M0hZs@}|)>f$4a!lv0AMX;Z271_wWiUDR zk^%se&xcu)1hC<+1UbWL=K%>b@HLlc_3~m(2S#-`Gh?5<9ZUog zbRr431MC}KfA%;~~k7z6)W1;0wculNo4wA|w1pAiF}IbHa3V&I3f1^@91{+Qo@zfF?F(x(P_ zK}(?zbGq;+^Ig;D38XQDt4s!2l|%VZQdqr^u2 zIE(9m9giiFB{fOR6djsl+Hoq|{aLwWI|#@0jZRaj)3nSny(4K_?2s}Y)l!EPFv`M( zg$7(qia|fm?UKy+N0K>00=OvjfzOMJh8#Iw3)f}Ycue@CywL9#k92ewB*W1NtTdv| z*gIZ^(-9Z*ZD$+g%4@gyJ=nZzqk?{Q(p4JnEW{J8&fT9ft&& z+Y7#7W}e`Wki_@%m2bIYk_Lt8t*3C=`?FvY3x*t8r2eV4{CNptQ7a?A&mSh^{ZRE; zFAZ;2pGR0qv-(gQ(Tk}PTc7KYlF~2KhvH&fID=^XD>4qinlJ4DYUWoLNTMH{C36rV zc;;%<(Tj^FrGRNKM@dMQ4g2cDVOixBDT~KZIIj~agq+tIK6>-X7;A2z2h5CLt%Q$T zwZ1|5D$S#`|6GQTZyx>*!(j;t z@4NRwW#xoXeLskDm6!uYXK^nii8*BKj%tQPQ5qwy6S#O$g*4*#0T92>{jqg8`K;6b zrRW{cku?&%uRIAf3*2L>O|-LbjG<9Duk=`#B=OAhdkW%ri6*%yzjdm<1a=664}#m{ z86uhPR-T$m5`jYuvb`g&hm9bYSSQH*bHqtDe=Elm7n}e?b**t=$wP;@`52P-f?6gCUf2v?HP zyK>ayjVW{f;+z>rJ=UJ}LFho%?&RH}6{Gn&O%?N<)e-r4fHSImvgU6+f$g3k%)Ivq zUHt9NaG9JRH|+>d(b!HM@4w3vypgLzK2i@$q-5&&16So<0Q z3{#===Wud|6j34mp*gArzmjhw-V;u86#QDh+VH=#+ur7^@dWP^Ffa}UVV*BIII;UO zT*2_}+35+k_62*@dIPm2GxvIfXM2JcdA|BIvp({y8WnoX6AUt(8F6V~0rGk?YYW`4 zE9T8yQm~S54SITBom8+ys2j}u)TO7Y!R;dT&37+K(j!Dd(gYUw62o-uKX@e{nK-ef& zqLo;t?|`v14A~!HFKq?u1e(mF=5X@-lLeW_bHt>r$B-C(bR&adW8T9B+qVG~s`uDA zqSk?R#mfqyQ;l1L`(z8%pxQ2j^47fW70|6s=hSr0tZgy}BKud=h`TJD6c5KzvZLVc zFMv#pVcb1&mOP2iFWva@t6mfuBCRzNe32XT1sambhaD^gHJC!U)3|gwcWgMybK@4H zHksyE!H2mhkcx)gsj~=MP7EXC8KV|0z->oKhqta^p^dXXRVtR|3qFt-a!IEYBQMd5 zLcPJuQ@z1!P=n!^ygk&hOT)a#h#N6@{tOcUcZLRo-bEE=&+Rf1Z78K|Bv6S9MTkVf zel=in>}HYkyb(VItlY0}M6bXW3W96lqA==ivYDsr#Rkl06{wl|0GEOhUe50~^mie@ zC(1W;NXqnwTRo~jT&gRxW8U6}fGGP?asrk$EgezUAqG6kBZfse3rl}jh+?7~O7M@} zB3?&Hu+Zw2oa9f`$;b&PWJ1(2{0=@+h;IF|0=tuQeu7Ep?)@Ag2%r$nqB@_@U*Uh06s$nuF5`muPu-Gq>drE`t;g z48drOs6XU<)f_(0waX0eZn}2hk?! zi+I)plbqTl+w$*-@<)fNkML_eevj1WkK>!x=XD^dX??y!DpQ}R7D;|PTYF?CqoGa0 zcPtb0yF`?6z{bH|K9=)HWWh%E4te*IJGsn1h#x~QStnq|S#lXhc<3(t*nETur@SOo zg&#v7C6t!xj29Q&}9 z13f1_Tww0{iQEe&hsJFs<~SK4Zfa-5O^*wA)g2F?VQuf2!Ursd!z=~1CwIbj9&TI7 zLhs33>MVA#kD&vQ2P-*<_aR|`SFrf_2SyX^?NqyIvzSt)8n@0E9GD2P(qtDWP<{$d ze=Y!Pz=3@&g4^uIl1NjkeNu9Lh<52WLDE1w)Ze~=ObdEhc@$N-^+GMA^G9`Jzg?hU za^W7g_4YKnVYt1MCSKEq?z?i|(?4rGiBRsI(lfAnlKEGX`;9-3Ohi89^y`BRh zt@4lT=ltY^!-iZEIN@+_(cD(qMeAX6ja=i!&YA5F>s{LJ@X%!V4BWnfNwxHmhn zuxKunY&#x)d0;?$N1x19_ipuB`0~Upt$zylpHHHEuQD{Pqg*o;42Gt{6P;T&{qzP7 zXtl_lP>US5HC4KoShZf%TPsUHloTT`Etba)x>w*v$8EnuKMNMcTm)fb9U() zovtMqtyV@)@iwRq^ag1E5{cD_l>O+|Zf7oJ?`CX*gZ0!KsNoB|+xSzBD?3$UdnoY; zO?4e`r6}wYZn2BScPWaM8-TnVNP7?~fm{C399At{XPP$SlU+FW3Y>e~t)DfpUtwb` z>2v;IJ}&CVwZPZK551*A1bc(ev*You=p`8E#j;7w34FpRnrq=EHYe~tlVH?1iyy~_ zx!i|WBkDNXbm8@biEkD$r3`6eN-2jIrA&o`QXXY&OhJr|DTt{!JWp@2kmz6+ix7Ol z#mcDIq5Gq3vGhj*xKK)Z$I+{24?9I|I9Yn+?HcKkq6M)%vdiv~YUz>t4%0m{46fZH z^Pc_B^unz~(cS2UcmEf?Kp9z!2&%tT(_0quiSCCPb~(iK|Cjyn%x}6M5*JB7JTKk` zwePFO{qT>I*biC1!NLG$#%JgbnHg2=f-rx+$IMvaOzW7Uerc=wVKZFKOTwPv5HlgM ztr+%CnULq1f}dnbRz{{NxRBn!Dv0O*zK6sD^RhuLqFDvF&lu1_y4Y5OOlBM@8O2Oq zCv~v^1Hnc_Tr|S1(j+2b^KLN}hNT30ns}rPX0d=@{&8^2Ah%$QP5b~EpTaj2%i&e3 zHHt*0T&^UTLp=MyDxs$yrY}rpBN^XBRN@*7Dti6!5c+n6a{$$f3o{~c)dr5}AfI*x zS%!(}232?y3U2V9jAYQ6@XD&Us9iPtIJ5>fsi$y1tr%fk!S!ZoJP? zr@EMGu)>_n^2j1v8+%`vCUABpu)f&RsGs>|WjwQ87yJ%eX=S`C-UjueOk?)()w?2n zyahII*vIcPD@vWBK};|phH4N~42XY3Ky-+Lc-?@w(}1`|gD5c|PHkL^VSG^Z_^t?)(^)s*PtUCdt%C1-P%=3*PDg* z07qmM$Vm+?X{M`}IP)+do`Y-C_1|wY^aLwy*K35vBk94q=r1m{Ly5N=LF6@uamc0i z{lXYjVdahV%~J?(xo_~2oRPnROJNa`DAXgJqEkWt`7Pj*~ z95qe%ew%!a51D z6Q-Z@m{T7{2%n&JsagB6djh8Pq_o;EK()!-RuHPey^v0Vw7FeQj4#OQ$IwR#=0LHy z&nLyw8v(uSf+jK+LtxT&iZ3}gL@6FIIQD@(V49tPiUH7uh3EY64vfFB zIt#7{nb`HtyAy_iQyr2WDDh$pVnj~-0d{JtOD=;ijuje;veZ{|f$P9Juk{64Z`Jur zdh68214IIs##QF4llq2VRz}&+dOnshr!9)b1uHKKnrWmsD6-BPg4?mkkElqi@44h26Vi7;9m zz$(gSEXv;cS);O<_}^NlFu^(jU}W2jhnB|8=fG;TWp+EV$IZz{DE?f|k}0qcmmL&+ zHi++?Pdrb5;1MPW1@RqDE;tIROt5v;y^Xecbv0b>c9$Ce9^;x->cZCTbWYXVY z+~FqtCZvOGK2&1TPZ57*=x!75Ow{nsHSQ_KZ7}JtG=93Q&*LWl?Z$oHq$@G--bm2! z?lEqm$-lwWYrKK;z6l>^^5w!W{D%m+WADlr9f0`Ydv;S`J56lPRlS@g>)`iUXYL|? zh3rrPCB$25>CIo z=B}v6KKqCfg5X9VB_O9kUYLVL5{DreI^I3d$d3O`@E71&etZf~MnpLJJVJ45!N%)T zkowa+!E=EKeLKw)e1a(C(MX*fjqK*p$gQ-3kQplYP9Kf*Pbydxdo-dCcEZU$504&= zyzNCVSrjHjt-QQ?ZsECLB)$+^KE>r+M|5heZn zJbUB5mPb{vM0sUEw-#@plQ)d) zIM4~rjmq6T7T^lQ^%#zdJpqMY+JgdO`?~(6A-Mm>^A1!G!@e~sfXZ%{tF@qK~0EPM-J+$=nVSwO>xSrX`fT%H+WkkK}lOP#(Mo+in+QuNI!n_D#5q5A%~I~+Pz=Lvlx-K5W( z^!=XQa*0fAd~ymehSukIa?QuDPC1q*m%J>Qo6Z6Qxh@Zsu)I{l;#kKMMmW>z3-W^t z=K&Z^sOg)CV6+Za?i#8=8M%-PMFDRTp8~~MBOAxH2$v5wLvTze6BLH%-F|i1G2m`# z?f84nMmL6p9AJut;~eYOL}&3tfxbbXz~LO!uiE%A0mFS$F- z#a^&bTU7n79bqv?E6DjhZla@p0p$K{g4goI741tLamVhtut#B`7+A9JiudD_+@&MZ zch7B=-yyKj`AL2vEKuYhen8d}Z(v2}Frld$pT5i=VPBYt4$>0mRcIlKtyTj*)<<|X zp=?A(Wy&~Z#}UFq>3gBpbk1Ise($uR4ST!q4(uta#=%m%9=<@k3p1D8+fe^3?t29m z21cX@^Aq=@{QdPsJg+bm7q_H6!4qLU>fgLyx6dKty#kE|S*fLC#=CsWc>hkPDc5PH z$E1;ON%MhDQ>fFF#iWsMNn=Tx7PX`9)lulGX6kAKbreQjGj+9I5*1Nb>y7yZg}|8m zb)A_2IFb8x-XPu^!Qtk4`XaQbX>zE9c7-QI><8M}nTh(l^$Ybn&V*1*2EJ)2%2QNx zZGYv?v4%X zM1S3Z1hM-6AkgS9;jyCK)eniFi>T+D=`YFDVt)#A<=V2SK2p& zmV*oYQ4$aU1*fiUF#c!3L_Yt<@?QJH?=XmT^!TGldAZVIq3lF*>-x=k6#a2KoE;E# zZ4wkxF=wH6ie$LkfrsAKVk?GKGLhmR?-?+rRnt$3!Up0<=zAsr_H23Gp2dN>!XD7W zX;LE2_@{QzB?2xK2c)nhWNhP!TjUMg04)J%2n802o+}DW#OrI$;&>D&qstnF6}2ZR zx3_3N9{(J{niTB^v-=jlDVWHw>r-{1ji=>%jtq9zsaX&?#6fiGaYw+*DL2EvRs5XIauenTf9^BzFf!>56`B0z4i zfxtdLYa$uPRSCnV>79V|WI`#xS+XBT*gA4Wy z`2pVSUvPlXwzPLrXn+J6cv&&-e8yKwL~q|C442E>|Z^s8U{H}-F6;$PatKevhhgC_o0 zn)sh;8!KrTg?0}K1e)4Zo+EDG*Ym=c!M5ehh-+!ul1{bB1@^PPiAYz z`_=izpGN;jesu*PQRAFuancTmZL<_`3V!wBt_hCNaq@WvOCxtcnGRnak%W}BSV)A5 z`PCN%lkrg*S}%LF%22IwSBhH|`a1xj#1P&u5cg=~#|s_ePljvL$EZCVXdRE+D(^Bi zvPTRld0Px$NTu*}Hwj(vA|JRO59Ah;J7f7NPRRuu5`Enga-V@eAAtb8F2xNy6l1N7 z`QPDjg2H!*v)$5!*N-}mGs~|Qx%ghsgdH*k_ml$5s#NU=7vohNKguo4>aY@1(pggM z1fGEwkG7A3P0s+xB@#x0^Z+1x08B)+Uv+kqm#R}o%Moa`)K9udj?vTA5$HI)wKz2% zqrh2mB_9Q~5{AvTucY970Bz(s#Y(ymD0J|k;q*jJ*kTYs(s9Q%+}XmAHO7NqEq@u0 zztc1~zCU@1-?%1-@6YfJ?SlB?wFrXt)ogHQC^DnZ_i~isR~;{V0bm4E1JDM1d4M~= zx_E6D)C13?OMMLP33214F8~yb2`4;MO+h=xp0s>~4}dURyjwWrGkyljr3Ql#PBc`o zV11r}bsDI5bd)FY@bdG`It3*XG3QA<%b1EFsN`anaYUNK4Q!<#sM3c^h?k?3S^U5s z6S}E8Sv>eyWCiZZ5E@x83bQW!9{uUHptwd&>nsk6}5svO8Wx@e<2bS~uCsND_EJ+N}s@{YdIrKVhXP zf$>QObW6tO7*k%2uxf!Wzsy6=N#~`+DWXovzTov z%8G{v_cs>etzwQ6bb6g%jp1N`F6T&2i{Z5Xe2UO27*73(U>F-hk^(Rz z4XPFFS&M*e+tnhQ_GF^toF%+(!FhJAr)Xp<@Hs7+#l5I6=^)&N7jY)B;&^4nGo;LZ zut-Ll7nR1viBzb+yCk581V~+hx@FRQRC*Lji@s2qUj z$g`bu@>6;XhOW^||IbK-A}i+gdT7 zUv0SV1&2JKCcepNIO-rd_%;q@svd|TVcr|lH79O!PUg;nuJyypXx@Fe9%O{l7A1MZ z^K&vc=U+u4#O8rDg5beCD$2X6@~G$tpiCYWy&WGQn;6z^#TP}H!ci~ z1@A@c_ouYzb$M!z_$4}JFuXTS;GKgzj4nb_@5y zd4gPv!((|RG)c|-QBLwmoy;qtWhg{Q2wWbEvT^DKBm*g%u?|42&-L@ta>=|{Dg}Wp zw3U(I6co$^(+Nc8sT(t3ydELuMK(QJ61IP%D`c1DS4*DdyY~(1EEML&-9n)1(%_3d zI6~xN!tZ$Lyl@o9haXeTIcG#~Y4b%FnbGlmFOpV{GJJS~coz;iRqh*{DzX;5d?Uj# z#Qf?pDv;qbT{*DSu*y@hj;Uv6Xb;~I^{WExpBRRHC{HA5rTsH*Ev2K+_X@oIS0Heft!?Gow6OFcfM|@<}J6>1C zukJ>8i>ux>l7XwtoF%g2-aCz+Nmt$7HT)p}gKZe)25x94m~cqMg$HMGzRvaO94sVO)>n(@WPA)5(fQ0p{8OSKFn>1Y9t)X=bh0Sz z1MSB72kCiD!d7V2kh3 z;W*!BX)Q4#C2t;t5$PlK#%=>3*Go4vsF`cPRUDx~pJ9XUEDgC1oEBILsL?`gX*JHT zYBk&6jKSF|913UJu7zw{v4QPA zs7p)y%3Uq-E5oAv%0QI!7x)#teI7)A3*+)IKI9Sf>r#fbz^^R2z=jyhujH_PRtx;f zf!7bhuY8fkj_cCy+h+WVcUhEQSpa@zE_E@J)KRPd62DTvO8Axd?+?bWoN%=8D@( zeEWaE68zeaC~je$A2U-n+s_jNJv0U?s|>E*?TP8#T?hVR2R_e0GkJ~xt^}D6BT$-2 z$`C9|gs8w1obG_*S3?+wZHAQ)o{48Z!5zb^ci@7!evG#=7NY?1uM@n#m{Ng++z2dy zD|SJ8A!}$NC_a_~$d&EMiYLdHK_fML7D>F!*uw|JM~1%8%Ps1HZ^`>L;GJ7Hpv0+x zD|V3Cq6fW{1&7$lg$qzf#V>XzK8Nd1C@-pCm+uIhkT`Csu4O({QRl;;dhkN@{pUK1 zv-nksCSR+*k1g(Igp(0nrb8=4uu`02QYWg%CgHh(tV^obIO6ip)CjXOKACTDeGJOL z69Gw*{3C9Yqzl5C493P$d=rV(ag_?O#9i^tkfG3yKq7LMvS~qiGfY;#HJ*4@T34gw z4Xo!8x&e4lGuYx+rOv+tn<%CF}4 zhDCx(fYbYVE8!lbL-k{8iWfSVszg#1qOPcHg4e*q)n*#&(v^m0D^vEh;lVhKumrv|~T!+EB0@8moP!fkR@o&!Pr#8X-o2_R(SrPL6%?kL2pk0Hb&}PVZR}># zLF5)=QibmLB^;JV))C@Y`_fVRG$1}2w3$K65M-lZ)?9;vHl_{DP_Ru?WK{m48AL$| z3JmSU0JEpm2pdEtQnrYOw-FJV!7&+W@#vM!TfA`PL0Wu!qzVp{g-g(;%`2Co&G({L zq2It^vzG_Z*d%X<8*2}5`$nvDCfs9#djPyBBh5G&-ejCC*r3M=!#PgM>0)e@!WEeKW|zyi4X5!qxf%80I2jIRQzeeqLq1h4 zN!E7|c%l*2|10T6US!;So_)j|Kg%ppYjBj!iGA)1pju~PV*6F<&8;1wYACMzRTq9% zg^VQihzV5WjK^i}TBUEJN^H_?4t&!uy zW;sUcu0W`woHGk#{A7KJ0t9@!p&0A36BP^XLC zot}!+*r^Rds>puSKAxzyvlg7gta=YBHNFJ__8wGrli(6_b)xA0^CAhbiQ_6TkwOX| z=Bp=A&cpWexzL+Hn}mM*9C6SW9hN+;LsPWfCCer*))H-b^q{>M>ZXw*q%Ia^MJwaV z_kok^d85F}D7Ck;*P-yxA`Cqaxkepd0BK$5e2UweMx5wZwJk>A6z&s5gp+R}EyEGq z)YV(jh|17g#@%n+PmSwXqvM}y;!iN)i%s}^svA|@h?+S^C$ayfX(p!i2{IYoQ%GS_W#z||KVor_ZX2O)*rjl?|Zanme9o~&>= zn9ba0fZT}a=(Q`ck(G@z_9AW>N5elv5C%q}Ogk}pFo4G?m;kq4%WW^Vz?idv;4#QK zBc+_lpi+H06R9X z&V~Pj0pqst4(CA$y1P1bCzjC1=e9M~Lb0LyTv2w7;%=}ZsTq{)RJF z!&WyT4c7?lDK`7nj%SX9fSs=tL8UrXeS0#GwRxHc?@(1G-WgIb5%IC{7Xlm*px%JL z{v7g8_W0$feewvP6S&kYK}Pwf%>D-eet8<9S{Yv>iC^tG;YbIKv-E?ei&SOv z;jC!JjWz()4K#NkL~Lb^+ViC(i?k&^d7FpJUvI zjQgQ+*BRFb7Xu-bDgMgPM&ll}Qo{)w|E}*fG2Ud{X(rr_bR^i zu4dEsWM`~`e9wOWJd8Epue_*mgrE5@v&YkA{BDdap_;Mwc)mXt!3VL&^D{&R2eHTV zm?)If%^NfUcKc}syZVU?i6@)zv!q?`2rSM2drVqkC%pZ=Z9Wo?h=WO$;o4ADCj z{ihhii{rLZlf@J01>;J$K$8WmKzhs_G_4UsM&_!Md0!mYgF!&=J4Wt{_kv0rBs7qb zob6CQk~muU)ECT+lyDxXYpcD8pp_>p<2dBt3jeWCQX1+({P5#IObr`jW&94YB+LhJ z*?dw`X)fjzrV2tnhGPRM^=jMpu!>8eBeS@}aq! z|Jvjr4l`DZn6*LO$9SG40ewFYBlia@DM25 zHkXxbP@CvBN16KCgacJi|8y`^b)p`qM%CB*Vo-I3ZjG+;Ym%rjs?L^XHYmT;qE6jN z7pOwM$gsP^xsc!R1h?-EqiOB6WCYSm3*F~Wezxr{w1ng%a0 zXreKM#^Y*1g+@{Z-4>8E72(1z-o5c)DB6WP6q4lKy;A#|bnSgsIfs>c1iG_k2!R>p zDC#H`Z%{rdyH53yvh{3@Fh>PPfidFrj$w?#$rHa@%?)kn#-*BD&~8;IDE_L@2)Lvq zk2GS&c}?72@R#Ae?)Fo8 z41;GCsBkmaBgSI!-as(kSeBzWz)Bc_L@^pU@hLdH#qi-plEP6}$Y!7uFo7|FlW^Rr zr#&Dabn-6W;kS5yQlij_zWOCDEOb%e( zan)Z%`U*fIn}_>CxdsRgS^%U60L4iIjdgaAYZ(MCFtOig4wHpB8Ijbuuxq1T+!mJz zh7P*EGMqdYCvatdf&W@s7Q=rPX3d?6vo5f41?@s7jy*AAZ#2?W{;OoAPV%sA+^> zyI<-!5f%P#<=5_dtVMq9v?KpV{94`4gY#?CYyLfctU{Mv4K zqLZx&zotzEjN7^wlq81y6^lx4_X@I56^Qby3H^jSn_0^huuko>xhR~?bipEx9VUV) zKdD7>b@*&qz5HtFTCPsEK%_2)*J-O+7HFgX#?|Rvks)A>y;Vwp_+yr=Qe~2Cl`rtS zEVx_LiD2cq22c5vMbxR6=~@}treb+96)VQlXt=%ykV*ts?t@l{{Es9V{VOY!d_EZk zt9@86id+Vk@&~sfGJg7o(%w6ln@WmVyDh?3#8mV?-9U9ylJ{BYsg$SVcarwep(h-Xc&Snt|j~2Soa1QZ~i;2jE-|>=I zbirxAp_#!ImQoMn`TBf;;9nFrl?9d~jv=~EDC4{o{JbvVw0%{9c$$JS1!@rp%Cv=n zEtz>_d0cn&9tu5)9;7j%K9o*oyt_oiP~_PME&AO#U#vegcq}c$;^m7x9_tn&6-9Z& z)&0T_QT7DbsTf&U|2Mu(V^8f*Fr}#2PLmP^s_p|6f{P1!q}v^-P6kl!e)4FMBoKmm zr<}mAB17GxhRg#YqwI0MS~C+v1gLKrP2l0wr;-BojABtP_wNytP7|PFNQu~B(}p~@ zVJI6b5XqkFe&xNwC%_hpJky1?wzLQ3+ysCL3Npct^*RGe~YD^rW1 zh*ei09U1OUs;_)$QW)ip$o`UAHe!D#eWeNSX$v*mnQYv;1)4ow43~uYC*x+C`2KJ) z076&CxD?jiWBgW(%kab5D&8l@HRKLC3kxQu>F1+9~DbP%N^BijgcGk-V8H$mdqR^ zF?j<$>laZTiyIY}(UL?*ToZ}QPYowuOl>WB9fAV3s}rl^kA=Ty{bYT99H^=1J%8uS zHhhNw)r1_beno`8@dWmShC|kA)35&s|M-42tMpgoea|j$k$)U!0FZyYr!xkTm6eZm z!mnO9;os#SJ6FnZ`rpa>epzN$vnl^Lzbb}*Yyge*tAC)v5&ZIfw_5WO0d_{gv5<89 z4kY6U*5kZno+opU^I_^Hc>_yw0`ukR0E(W+deZiJ;^z7?=Qcq-zQe1|^qXrH0*?+BlfjfAYJ{7}ir9DM4mI{#=hWPgR zGJn1Id2e88PGG%{lo*FLTKY5#omBrNP|ybo57Edvb>a_*cvHf&<)MtwMKbh|PR1us z>>mx?hOj997}3`u7Sug~eV)K?7z@XOMw6k+Si6)6PaQZ3jnCR7o(GPP9U*1e1CZ?2 z6)!np+FEv|ZeQ^9814+hDMG@YlZ?=N!VrPr;n3mf)twwz{MyO63?`#BAf;V*==9G1 zo-7&;nX@l}{YFqX!Nmpm=t$u^@h7X$Ed;9GiutWpd)4`HiT}ViWm)T>*sQm zg_GaJC?ZX(u4>L-{#V&=-!{6%{l}Ye2j73ZOBo$`~>B+iz3{-+!ckrjeH$ zyA(m#b};+x4wL@>%KqaSO}9d8Dr5E^-^0^<|6%O6bA<1}X*W3n516t<3QWr}S9!Kp z0XJsYYc_#ms$LS*_;OTEnUFD@QKKnS^v)w&SuRJ7gEJh z-(Wgh;|;@QOOg%KZ{NQ(nw51T1w^J}iuM5u)6>E2aOUTs)wkgOtT1*snpQ)|JX#4S zo=YBkMoJ+NgTkq{$CG4mW1BKGvkku8m%fv>0Ec_)lu$*H@x{m2rwBSK;8CMZmV;Z)(d zH)xZpYVS7oIw3wo3YSkF(e-2*90jJ2U!iwOR)p)2*Lu-z5-_N(+oASJ?cv!Ojy=@eEPUlcw9 zUlUM8d)sH@S(tHVCpyULlKXyBGvE3er8g*V=fZ-fVZG(qUFIjsVL?;oq1FEO7dRrv ziy2IW9-QxHBVsXRh#dxV+qd@G;coT43O2;ffx1VL{|lpyG0=?r~% zwwbYqK+q%a9~43L9Ly+eqYUci1aA?-rQuJ1bw~uk;k)LlK#d@1&L)AqU(2P`MFeUPRM}q;G#|NvZH^$k42@Wg zpen{50ztRFcTfaj*u^K_I zGxiV&>ND-22zu)cJ%V1L+X8|D2$zQEE;uBD=0d?<(v|)ojd(`)vOR*X65Y^7Bj`k# z0uAbPsR8x}bhX$rGJ-k_+9M8>vlG!57YEXZoOdCPtt`r2F6Gc#+#A*Z;G4>D$ zI_2GiBB=Crji5rhEg)zp!lmJ7*BlZ-<Hb-JLf^KH)ArRE+ zor5CC&uXEXH_~kZL6;z08h&K;ArUmT!62yL8fnA=-ODzDl33fu2>OA83MFrPl{KhS zKhU)@CJ6ItXwWKw8bQ#78wB=9xwhM&K0~aHpo^{$1Pzs91woS~RwF2zv4=p=hPMui zpng)l2K9HkEg_E2$zO0t~?}yvY?eN>Ec%jft`YPC zT`S|sSOh&UP$LN1s~3RtadDPo?*Rfe2>R0}2pVJ~s7eizSRe?ypfF<(fuNVCL=eeK*On9ERoQUzH#4**!vh~{(X4i>LeGl7Dl{1`#UABN+^HsgZ6miT zv|s#Hp`T3pJx$!ONq_oGU7mKULcPUb6}k*A%e%aZdyPpyrjc6}nk4?J(5oi>8%^A4 zCjEDf+^W#8;;#zrGwJs?apOPH^*b}hWxehYe^n^gguh_?>2Mi8$hbXCe36NN!?@E- z{PZU7EQy~VTHVO43hfquRp@|8A2&!L<(RcMg-t3rd}vK|{u`jBz2G2ym*gYiFZ z+&AG;ANG6Wf62H@jl0gcWpJ4;Y}~e=>hcb6;y0PLA$PuZJlBzf%|AjuvM%;Me zUf5WOp+3rV&@yfb|9_&7(*1*|K1zPu?iCR2xf=&u|q{lwVx zF`z{5+td&E4`2~P*~QrYavPg)TO&W-_w-l;A!A*En@dUI_0g&cyg4pp3bmdhCUnvN zbCT%a@uuwuyzZ<|5wD)M4Xt<=5m6Ah4{9ZJK)eXiD(i z&N;h7AMKfh;4=C{6Zi5vramJg<>S2QYmv*r{&(OjIWfvy}P6_?t;pF1CS3B@5 zu*;KHKTG8P;~*{#j~(mujQayEDV)V0XHF%A#);4(g1*&%+xgSHcX?Q#50?G|=TAK| zZu@98`op?0DIaeaj7h{|Tr*)Z4!-u}A5(bI-}1XaFCQbfq=Kk%c$|G#@_52Nm*~Dm ze9`_&`;8{WcWS%^mg5k_dHx%{Url{=UJ~{+lJGMR$``QQC#e4#b421zig@;75wkcv${sA z+4MYXC&pt`9||m>Uj+=C|y-ov;zkYoEhklSwovbjw?f=8u)I z@bd+6bBCk2W8f4B}vj_UDv3T!3`i=?&=s5E61v6H?wYrJO9N|1Hg0R#VxCmCZ z6FCB_VQian#l-qE(0{(T22ar+IA{9^%n&ttAuw4TPM!={I1a|H?N%PCorG(dcs&7| z;n}?RGY4sLD3|oU?pm51>_Zu1@bBT|BR=F<&+j0fVMxK)^FPAENP5A)vFNWvUz!s0 z-cL%gqaBx9DQ9>yVWp@NpmCWIuhp-27Ttj4G(lJT1e3*8u-~$k^s+Z;o+D{8IN->wR)xLhX2e01GJzVY5mJg zBSw`miBPM2#PI~3iDaRivfhbNuecjakmfvwFT)%UP6G>e~Aw2;(YWH zG>J+oNnH{2<%#$S^+ix*J%MmK6)28ys?HL=&{&@aqsjf@M;=0~&K?J@uRiP4-ze_{ih zJVi!F(=xUo)mqU=d;_V-cRlk#Gk&Lk;g^(hq9uPgi@)JV@XPgdtRfMdn@imCEvR1k zGEx0mg-#sk)ZfHiikf!PN@NSYlGLdn1Y9?|=M40s!^Y|>gucP{Sc-25wbi8y;(RlC}zTVU5 z_^TA*oXgO-0s1@vT$1LQ?&P+uX|1CWv9Mh73-Ij;o`?I>S{aH5Y2}g}V!wA&DmV?G zrnerhI`>I#z&3j<`^@e06bGQF&Io5Q6@Vz$GoaRi(W+XT@bwaob9V{vr0Sp0=^X?f zIAFqEYPkt_0jS4%TpI@Af!H^4sKLpUsNR!A#t0#5;+D#S^B92ZlTS(H<0-JX1PGs} zY!2gI{S-6L*;x=mbM&%pH*bsap^ z3R3+g6z{QNi_H1yK|37~3g==)nX(SoPK!n+I`aP}x#nCbNZ6v*%|UMNPD_|VpT2=A zseXNUP;9QiV6P3UTXjr;Yn^NakJ#YOJ{w?Q-{P@W=oTo7K^w>$_)fOwII}wnZx>*Y z;dBw#ovM42XxGtXyHUjP#clSqhTSXmSX*wwsq=+|c}S!dz+gN>JPbj5>eNhxh`t0B z9@JFv?NNDUrtJge18w(#)`>SC+}>AOGXL|P~z4Jm1)8MfSaTy zty^UQDrwnp>RhDb6y5{PS5P6G$^VMS`$&Cme_jt7WW%LPLvLJ~-qb&@2Z1|w&=`GA zkc(Vj<`zw#@^t#oB)!%`jm?5f@zjn*7H-E*fw?!;pY26WKu-tI=sj-X6E8TO4}hhK zIh{u{*xJY%z=owmlkXi=EpjNIsCGO`5{8DGL;2Q^^HAPLdW-Zp@W!ub7vtv6Yx*4i z78)5Yl4JJZGo+Qczrz@w!56Z~K+v4QZyp8UID;?Mp+Z+Njr6Oe5#?W${#Oz8^m9Zk z`l>0$A3%ht#`1{%L>$o%2^8^o9#!M9Jn1|KTVuM)@ff3HCv^3hm-Tp`_O_<0^NqXJ zxT}r(C*0^1DANxY{wWMl(fq=gQH}xIy}~&&;rHMgUY0MW-9vac8t)m}yV!WY1$T_t zkBxV<_P%7ir;&H&h{R|(USFu>G)YofJZ`WuI_t}l>MVHq+*uDY_ZZwzX@K5%a?N|I z9qKCT^jp_q%)pE(ZVVcIpg;7Hl+jyI?L_(v`(*)gAY0V_Qubtn+DX^SSg5Zksw(EP zGQQO}57kU~TPXK#mb+T+B4m6bIdO(jtx6jgIZ=)@>eN#xf<*>Qk>hlckD4N{(nUUD zitOETk)KaPk$zp|uxOFVjf*_Z6xr1jd5kIYNL^%4Q)C^H0rRJ($Xa+?Xk{9U+^N;J z{uV9rRl%gK5TZVkbBQ|j4(1(O`5NaR5c>jMD`T&|gQ@P+MYb|ExL$zp&0H(624j}A zL=82X5_tlmPF<@()EN+I8pLLOWl|j<0pX2;IMsmYVn8Ho5ETZ*X5z=nSYkk|5OlE? z2~iL@_ml>#euNFFQ_JXD8Sj`puW1lT26yA(Z2@;*Lr|k;L3@j%(Mmp&W|6Ff`;X)N zl*g%;2{i}Y#efS6zFaP>^pgrja66EvD5$~}r~+|&iBzdUT?ChuktEJMT`yD(R1WtX zp$o%Lr2FkxlNSfQ0d)qo^x=%$*o= zm%~Ax5_$>s3kIN|6x@D+@pxw%lrs_k27?LHeyn*N+CZw`h%BPdF-&yu$OsvO0b5eH zqOFDY0g$cFZgsisQH~VcN`Dlx8%i(c$#h8G@CqoLmGMtVoUqgFKN1T{^0o=2dPv*n zbv!^a5ebdF26jZ#Y&aJgICp3`6E&Q3G@K&@&KowI{YV?SxDlMtBc?ukAP-`FI%+sq zYB+~$IF(@G!pYa$aApurJtkGOK5yA@-ZyZ*kU5upn1=HsH}zK5D1nn~!?}}i#y5g< zn+?Zr;0)1l=9UY7W@|X-3Y@Q>H~2{*oGc^^C*L+vbKn=ep-*pYH`n;-f;R{iG;yCW ze%t-X_}3bDv2p(}?hD4<2$vTv4u3Q1M&yG=lBVV9Ev`SaOZ5744pv3sYA^XVpbjcJi2Lf;U(d8(hG*pN}gQ2MoS0k6!h1n9_dYCbu}e5Xpx?Dd!UH(7eayy zAO_<6ok^{IEBUF(GGE?)Q|4W3aNdx+j^}(q+MP%nvXC&GocSEXDnnym)APc1%Z=J zMH%28ao8DQAd~cn>G8N8Z7zmPrwU zuWE{5yRRGnD7c1hVkd^hk5jb9)>n67<%mINv1rNyqrtY$G(6Z0ncD&mBFcE-%10;p zz^fTzBfxn~aa+Y0E8#KNB9)Qv?0|cVoD@!6i6h(ZgdT4>69G;-T=|fF7(;UHkZbvY zfIPYoSB<3SAq*Sa>=*KB7M=142iMs7;aiGnQ`kGfIoh- z*8Jo5cKwNSfT5ExK=E!2$96oB%)u@WJ}24O0$;u~*9D zf}PIEC6)k>YC&y!qKo-S1a~~+uAl6oowCMqk3?npx4n5v{H;*5s z9l+%niG?LGs_Pz&uJW`Vsc;fgw7C^5!Jjz^S%K8(43Fg+k|fv@*b-ASY>PUf9nryd z9K=l=7#wUk+jGo13pESvf;c^#Jo0Hhfv+yt6Sx$vEd%EJ&*r+-xUJV>Pt#~*;dwOf z^NX^!VvTkOMhUUy5>jDzdCjKL)gU5OgI)f?BsZe+=VBL;$wn@K-~`c4m!MK zBMS5cGbBEkoPcD!#!;&a;#JWKeqx(_G7KMj5Vf9uD`Us#HIRLFJ5vk>`qkgIzrdPt z_LKp0Y5>?G(8#}2smscE6N$l?a^$-Zi}BYdC6yoQ#bLA}gzHa5mA{9%O&rw*f>l~1 zr*NB1{R+<)d4Mw)8`w=YaS2OqA?{E>ZYP|yyE)wD*)l6({$OHcysotz!sEUJYV235 zJ7Sce{ykdLm&a2+DVE>&woXq}h+`Sr}#x*g4-#OXm&h_)ajRSCsWPa9;)DSgUl5E!DAVw$;iWX# z=S%T8XQo#>@OWs0R$TWax=TgS43V?y4EUa4){7Je1T!o+epPb^MA?-6qdSgh--;&| z90jOum97ni!L3DWnaXeU{1hZX=nzf$w zZ^!yyCJ#EG{;0K@{^(z*{~(kasffm+jXrfzlNvornWbH$jJYqd$;tERG^$T(v_A9Y zsA6a-g!G4=Y$u1OA!&`+u#DnVuVF192Roc&5(!!%0eIs@GA)X-jsj5?02R(-V0))S-y~iATEIjqAX~#eK$SR^xd>Pmp*hPD$cq#?$kyoijU>;vx*a!6eXO z=vW}E)u(P0Fx!QnISzm+ElxR(NifTghughkhvyw4BOOx_$T%L3i(p-xng|zHAD<;D zU%Q=Q%y~EB_{W9rxSMx}`*vTLJqtoB%yZiyDalvY#oAE0{nA+;{c#P)D z7P%Cr*)Ln%u^ce1F1|dDa@;aac=`r>S^Y^!rPE|9vmQP5E%z>(p{wfj9u@Wzcg0b%L$ zH$z_8NSnO!+e$c|l-*EtCiknosUWEkcc^=~>P1YC{$t4=EGc_bHU#mMe{vCyEcDKj zZS2Nz^SM?>Im;M*Fig%^60IIQ>HwgF;zo+Vxf~{2DK}?9tYZ=4nX1X`s;~EiLmgzx z?M98W(!pL~Wxri;kkv|j#gDxhLpaH8HKp^<07@*KUo)Nv%lLkIRttX9N{VIm4aqZ;MBq=o=}%Z4Dkt#jxH4F>I4lkJ()|B{Cg_7#qhwcfHT^p zkH`ImCgq`Kp_c^lhh9T3XwrnS&j@a}%Bu((w?D0p;C3E@Be*^DuT*5kBEfAB3Soj7 zX>h>b-0~N2`w}w7cKYV0S+RrSHV@Rx;P!JFU4q*VCKSkDA=4^~TP$wPar+P0tN)j5 zK|E>}x)_2=TOP$tVdo1cqJgCM$vg49L}PpJPf*H5$sG-sQ_E2F6>YXuc!>rgIj|0l z?5U75sj?OH)~Qm==fbQ>BghaJ@8_*7nlfUm$big_WU9r+S`9Lp`Xf{58zjb7x{t%L zJ=`}j^hvb7He)Kh<`x(Sw%10oA`^t#=sUSWc+dYcj*ZF}xrv;)?;H$1)yn$eUUJsS zH_KL=w-D4CL?EQtB1d`vOPjMmt9*+7JMe3_3 zL+Lc>FxZ9+*Emb57v)z!mJvXW%npznPA*Ij%ob2&tQ4KX$6<{QtChQcE399U)Fu+!!dQ=6(cuoTHjS1PEG4lFbrlISF~8kBGd$rc_T8M z{0TdwS>AT&9EqNQ=y3AWEhah|4Xt^V$S6ZY!^w9_+9C3rchDTqN{$;O$AyyiJp4lX z;ti|_O@4+_3s~;{E;RS)$mhKX4ZV#YvR9U2n?mmKdduk&w<>gralLR2xElnVKz`Ar zF4tf}M?_PVJsZjJtb9uOGoOou9*s~+sXh=GHKEFAhVu>B3#Fu*&?w^{BjCRay)UlR z;7JoTA{w+zKBc(b^7&op+i0vSTA5za&w?K5U!?FABfMnQpLz0xy2A>j4a|c$p@G$O~w?C$L-rdQFSjnB`p#~u`Pu}=D2%iI) z9Qpl(*xfZlcEu%caJ|Ho{&T42X;@+mPjd=*J*pTYv=L9^mUYmDsleP^!&CX~SIK}Y z+GU`s+;#teQ2h=VcN^G_4>-;2y2)VS+68D_!nIdp8iAXePEs7R^)BZMj(3*^AR1fa za#?>WK&T+B!na@Dm@r8)k`sq*2JX77gn51qO*KotS~HIe2@k~y_?Q=O1-{`8<=aX) za4nL84W9~F{ImH)X2nw$U0Hla+*2gcXq#tcI1EE>F7P2)#H-b9NLw)0ICNvO{_?H> zDEFu($c7CnfpV3QvPjZzW&}4u2&*ACyC$oi)r; z0}J9+!nAVvDMv2#!tHpzP+nam=8^KBrvvjKR*d}=|KZ71zPO)#!F!?0adDCk8me+i zcKe*3U!5=aNjcFMx2^Dt4UBOXPePNhH~eBfe6okbE2m0U?TenTp ziB>|{RS4uEhP?$?iT&Mjph1o@^!9FT)(SBDSaO5|Z?wfrC_@~QE|$2WtR5hBdS{2g zD!sEKo}29Kh)(Y0I7H}H7`N8A>*0oz7nj8D?6xB;W@opT-q@WTwsQLr_Ng0f?_L4y zuiyas9YM0w!^W;pDg3Amg!YzWD%q6LSt=T@S0}GBfemu~d;8x(>(#l>bQVI`2LY`zOQ_9wb(@s8%kp zA>57QAnnpiUQ?eFpfHR8J6-83lqANdNnS zBa?jAa6F80UP?|$xF9(+L0Zs<_Nfkl2S8x{>E7V|sovmNmo_e_E?$g7!r;IpZ}5sv zPy+wn8~BEKS9u`G^JZ4#>2|>I1}{gRYmkQrTk1&Ufe^Pe6HziIDc{MZW>*gn=#>`&lPbr2e0~pOy9>xKH)ueXP=7xldJ4#Fety zeX7e10N$s%auEgu>;rznO~I(exReodA>d!X2lB_gs0$uYE!yk6{~Y|r<)47_se)c8 zH8H!oPZ!LMB3`}AzApRv${}Uh&YAHkrG1m?>-0R}Ef2q1@~~aaruV5{F97kf-vepD zZ1$^6R5-%l%6Vu-#~v&GSRbz{ z9ridTH8QTPuhR9Z^9GbJu%Y{IwGyhKEVOavVpQ1S5VL~yMN#}pf8jn) zj=s$8&LU@!cv)8PDxoi}|0F_xft+8vFQd2+_Y;CxA%myXe^92)EDlC9bU$MJw#hDy zPhspcfUC!I+t8gzfmw93%%Y5iTpJ2eUq|*pZy+2wQq?w3eZfyO8m6pA3bfZ*T!z(F z_>G+4S>E6mQ^Edtt!csqW&T;vss3X7d=@sUYdKeHJegJbXL@?x=qlLi$^0|_B&_&2 z%)+BDa?wxLl5TgvTyEH!1z;NNN)2`a9ONnJP~W4>ujYIcftMDA*Hz$shO(mYez;xW zDMBf#PRz;N=`8V~-QM8y(vaX3Hp|=dWnRm07N;>?&zDDvryYK=iulsjbN!cxH+h z8%fSrNHicP#<)0!%t&{bxSNwpzKgu$`rBxGc&#t9zM#@H*c1HcUXOM0UT@&h^~mfg zS>r4&;5fht+DWr*Abj{17}0HP=s7?t^VU2{#PBcl6GU9_6KmqHf!`BY>Py=oNGjUb zD!-ejXkY96wS>M^>^{SlhS`Tb;794 z4&23P5J0Dd#;0$=sooB)u?ilo8=e4G@pq6W@vI_H6Va-NS^<@IZXG(o(YH99e+Ebd zIb%GIaYjdZ9K%`Ck^NINe`FNDnC+y4!4alHqFC?diGLxB#7WZ!!V{{euv10#aPHw$ zn+Ea71z!*zluSS`w3Oei^Pg?4f(w_=$LnX-Fj*)>P8|oxA5XwI&JI629gZ9xvVuFW5e?xmSOz8+sQCU;|GDDAR8^$FB?LCQ-r^ zsP?P7zMSOHf_UajhA*^>qF!Jr9hr%xRJfD}b6mvBxXAAY`RCnOB?6W8IMA)EudUw+ zhnfEPN7KXw%Tc0V_1OHPqkehz0KV9+RjZ1&#=aDVeNp!cwRkLcUC<3Wj!jFA48#Jw z*g8L*gRt_;9_W5=;3vPj7ZcqRJeO;sEL0)~O_E6gt%l!tkBK@V4Qh*nCwyDfjizpd zYZ`JyICaaeCv&5-_)`gPT{!YG2b8U9+`{5PVwG6U${jt5dE?bbExQOm!Uyq))uVOJW7K)=R%A`44&C|9D?qT zpKJfm3EF?uuMobD_BaX-FC2X-x^W(*Lrt-oev5yqBoW@)DS z4eIQCDmU#>17Fm0D5`6Up2q?Tku3*P3@t8W3tZKGFN*^%67l4QlGoxt!3@^)Oa;-5 z&}&>~Wd0Y9hW>5wkEAR`qdEGxWI0P_nFz#vt&D*{I;HeeE5RX$gX&T&#s5#&HT<;Of%knAXsogq#@RP`Iu6w z)Qb;~=RjzRlQwJ-UQBV(&0Lp{SO0bif6eWf1snC5E1bo9FrMUweo>W6FJZ6Vgn0_q zVPy`UWnkWHXkvKaf}g1ZC)H5!k|&6=`hZrtJV7u%jR$E?=6;$mxX4>nlce^Z{d*Vz z1r2C|M$6qNq;TY6wp5^reFB0WXNSp*^#wmBCOiQ_mM>7BlesBBJu=>SZ&X{Ka2D6e z5P>O*qDrXeaq0az4|zI6F@OTotMJmV78$W6!nMhqiirvMvO0(cwc$GUmA>yU`ql18wCdqN`^ zx|I(dnp=Nvl%)Y-B&Z~HG;dzXbhEB?vGYA3aL(c$5Ye!^p@IcXtWlzXi3FKwg0!MyMWt#9UO`P1@6kzwaU2^hUaJ-rTkDmo zMMTAL3!rU9>jm)w^~M;`<#9Db?vp+UVE*z z*A7S7AeOyfzcBn9tFnGwtHrp~^!vU#>KSWWwiqK7U|ChR!m+a9ICleM^VK&jhxC&) z98F6?O@HZ@l{KtnLrwPAIwcQSy*Sz$%XO1RuHUy+xbCTruB*=eGFZQ*J&zsRF0Eycmqs8GEd+(OQaAY2B4}0P0L_W>0e~pnWlzhBZVE+2)!FL=x{-;m+~joo!*FtLWK-@XrCxm`j{j~nifXz$RRvSp#Bqk)udlWaa!L>DKQjEGftX&(`-?%qzB7cYtYtDO^? z!=iI5A}zTiV!LcVBDzbcqHOWStJ@w=;cKKN`{Ey=S9YN%NS3;Yt=a=Z+lhv);bJ_6_eBYby3vj{^0X4i(@x-dK!@50?0*)XDPUu zeA$WPGxBN5m$ikO+CUt$?+)mizhczLY;-Ftj=z$8hZr{T7s{)Qb%`#ijCQY#Rc`-X zWwbK%UE(1s49DV{;!rKk?Qq#^p~eHq5{_3alNX&ifB&2PdTI%p>vau>o@@=zaXn7% z6W$j||HN#>8a+Zc><0A-N7pgx11573F|>6F|KBB77la56+LDE{myQcvID8U+P97P) zZqYR870$r3w|C!R`_&#AJ-##1{I+t(X!Yq8dI9%3V z|2QA!jtgC~jFYNwECw&HSzNxMwQHzxEv1JepMDpPA38G}-(`K|i0rzq;bcoVelhdr zi%_GSmPCeRmxpd^#dv8$7zB-9JkmZrtxs2sOl6p4vG-c+q#*XGX}sG9;c{fQUQ^!| zyKi;pM`r6W6~E4p9~GIc$5ecQf&YuTL6O;dOvRV`@kNo@dQ8QS_v3p;X6w<97gJpe z3$ImptZ|NLDVU2G|Jm6<1gYL#3nl2UhlfPcvC6yRBrU>gz1OF1xd;mkV&iu;F0uK0 zd0B_~CL2nI^O@q!BJC{fp{A}vw4d?^EDJqbx&7vGLbA1~GXH-Is zf27ID{7+;I*WpY3XKww74_?ZC3xW2t5@=`UtzR6ZxX;WJUQ!C?bAuRuS~R~6kgfF| z;bb`gu(WtXI$b)Djowsqo03^vsL9{EV+FokR^ap7-ctUTYtW7yL!UV?gFriW5i@ebG z+ZhtJ>sEUjQIrWO)!Z-PBXTy#|^<(RtoE*y~rt z4nqy2ZHjDmo0X3Z^YMHsS>j}lER^1FPZ3M%zUw569eJ3h#KtVVI9kJtKAo8AvbFAW zOX~A|d{)1UwJo;o>qhThYLHbhOLgh5EH3Fqi3`V`mN-OSsdiS)e_c)7qDUYj5wT=@ zuha90^3IdsfD+&|JD)B)(3+&{2aFX6&YqRK2KThD4tlB0QmE-9)qLFo!8wDKb#&Ym zQYvh$?c;RoEo-FL#}2=#Z?tS+2FwMm3H1a{H3X=1uh496Rb8mI##ldgXnkNR1; zvX_zm-$j`|S%c!eXa|!L!G3>*H>OdZRMzSk_P)o7Ah?3;h=`pJVT;MAy8EY+$7(7# z$jS_NE$`u&LAAAf|E$C_%yFUo+e!!11H+ywj+I@F6E>~!W?Du{c-%_lzIzK9Pkms# zazo6NijIxQeFfrc3%A(vvDNh~xpQs?_2i;e-X>X`!aDO{;=>K%z`#V1}{c}Gu z{arsY{pFeI|NHc5ub(PXk9xoqSpOCGZ5?Gz`XTzXK8PZ>qEFlWEA@$s!O*8=&oF)3 zYN%PQ_p`tJSbf@~Cu@TIFZF5H)CSG$tWO(7OH=x^T5DtC$xMCP$!6c5(pX%#x-WcU zhqXpqVC*lhFQ=gxvgSc8%|b!sAs)sfUCy&dI|Ce+4%}AANJHC?rQ!JSJUGRd9kUP@ z4VBg#5r+;c)_T)k!Aez5rN$H$)Ov5VV~iWm`SRy7H(A?KvLVDt#CDNp%F}bM+JOaz z8c@E386cFqSG@0W&z^VX`F`R+X)(`$a(HLo-jKsF5CK61&Wx{DyD$ z3pHvV+VJg|P-8?b!#r%tn4&Ynkr%otMOOX8p*de#J3hWYD6sy?bnC|bnOJb7fwWq# z=>gPykB}Tp`rc%askpEBPOJ6a{*yA~TsmkuH%(uobC~cG_T5LKWr9+3mi8gsyTf|E zr<$(_ZD-xGUOF<^%*J!xV;EG%I!ocNW{hiUJYxbj=;1B8G#f^MPKMPjcREBkP@=_IfR~Qe>Z|>q#mmwnwVl!U;-p&iKsUE zp{bQ(MC!54T%g&P!Gts;Gdz8aw4?ZTJ+vr2p!fX|z)^x4kQl`}?vMMNF+|u19X$;- zz9nt9(uS$)h;-0z2r{cxv?G9BUs@?zGJuVpJbOC5|I7LIYXV)QCFggEE6 zrIgSuF^p~pN+HB``kq4X1ly$wiQ^pL1e?KoL_#mD=bR-V87aGYVhb!b|9)b@7soNB z-3LDASX0N+HmE7ESDV-^i?77TlRIwfWJjEdRN26smLjd%uNCS7woOwu_{tv0K5D(c zEmsxAm+=b}FW>GMRt#Np`Tj z4J+aWZ6UEsOZ7BIPur`wHRaqW_@v7Ch$kVC#w^{G7mVxL`|tOloo?U@z`zz@2r}{9 ztoof7DJDyT2sxl`&q@UdQj^Etl6~`BRY!{A>gYU;(ZK4mmg>;Z7I@$4P$hoG z=7;-z8JhBzl107^U3PZ3>>r`VFZ9~sw?{lrOa-63Wgf_vJ{GFyz!ZF{si9Q~pyx-5{TD$71lKIcbn3&<}(BOKxJiYG;F&bktpU(gquDra)BR1HBhB z83)qI{6gB~=tatTTn$Pt-+5eB^bM)izJ-W59Lq2`zY{pKeU0Z_tv0baf*fE0X=-y@3TzAN^D5VcO1B<;s#O* z$N=)0_|toU0yV`UD=F7;RnfU7g+Xh#LewhP>tD5P?{SY;iFgQwxAfG(h+pK-S~XuO z@QMPh$~e>?FMSx*IL-4~)fP~kVF}c&0s_QO$BD$VI730YQBG;5qzmgoUBOx0g{Mgt zDhoqMp|`k9*u|!pjL7^Zki>b4!`~ilQXcs^d9W{NFFzy`!#~VRT+0BFqbxDqOSp{$G4@dTv`X8h8ngh3r(t+c{wA-Byb36@@I8)7-fD~f;fV^c`hM&WeXZ_C+`q2>% zp-aa*gs%3(IzK$q51R;OuNIvD5i=(`9F9tc1(IZtd*`6Vyd9oh>AX$#>;Wy9nN#Dx z{*~oeS}zFFUK|X9Mu<$KhTCi_%650?vij1JhaLIq?z|d)G()c&y>^sbCc!dWyGMTU zEpHfm%=gY1Xv>~=vW1&{XP)dBX5KDIwu$jdAM#6iic+ZK&-R-19#o-FmeD#C=r;sQ zi9u&@rt)y}vQc`2C0zycPa=$@>*F-N~qjsoMQ{esHjaW#S84DFg98oLw%I|{eMpU9~n z@MJUbDT)>~%B%2OQ{}g&!f#EPT0>7+#3qneq+>J#sa-M~Zdhp}>3EPDRWvqmXEjj- zer(;ZA?xR{=jhd^mq@#Wkj}MLIQF|K6d?~++8H(Q9F9Kff)qhl3TAbvaWW4bgllTf z*tTDTUsUk4W1oA?#p{V2yM zm%j3l!?OG3M3Boa;D3Z=x9T;9&uiI6ZSaO3pF*?tQio>i2ptL1uJo5v4QFrhv!CEf zyUkvcUZnCnr&4C%G->RPKf>0M^wMsI=|5HB6* zSGf{QvBMon?@c{!S=|U#caE#>&~$ZwvW%(fa{cP^{OWT3>atvQ*?x5wImA51UX$LL zTd#5vvN7+c&Rf59VQSCoZ zPayJcdjgHwVgii?zpvB;>c$Z_87=(JCeRI__*N#+*m2F2|ENEI?%!EexH)Q5nxrBn43Sp)N6YFbgfR!pIL{v`7?tsJ%4t(#pcg19(CDwcBNfwr6s*? z8jDUt4SBxz6tL?~hh5z*r_G{mI>Axgcy zCcRr!$d*-ZK>p#KXxvrluhUhY)v?M2=xkKw-%1=pFR*Z<_W~hRmiko=a8;h~S1JF5 znN^~&)q?1KYFyWV70%1ts8p>o&G;m=YoA?zB|I)F^4_ zu4xGX{WGX@gDCEnD?s(OSAeFiBk&o*U-NHB4Ri7K6-<$B z<1dxi$Nwq)XAi91V`@hlYI=ZIHtkwQGK43Ad?LPg^vHKv+@Sb}k8u7Ee=e;y{tv6+ zq~fKYAv0kc-Bk>D{=GxRU6J-H(k8YiVY-~Epq#ZxakMPHXLy`n<;t1o%6Wc2m2+ZH z&SNU)X%Z&$zo>Ma+u>(8&h1n`jQZh2e)y^%p5uqN`Qg`oxc69>zrqhc3DW!T=lEe$ zr*Ms*?jb)u2z*_IN5BJa_UF%&nu@}&W;~@2H5Jn^L;-6)p4R)I>aKSBi6RE{NR(h$C@?_$e4+WPHtY{9Zhw!K`_cH{nMYc$k4Mt{q}CkC>?SF`C|D8E6{8`2AQN?ea+Y&;Yg}j z!hPgr5P~6OGq}ed>o!qXDdWZ%&2uN+!|J6Dy2nfs*&|9Fl2uQ;0g)tYLbxx~7ph)P zs7=5?y_2@kJIG={>)+fBT0@}w+M#+7e$;zAs-d?bD5MD58ZnU}VkgdgJ_|ZZO;5&~ z%vPW@uCKBDYb-g)TyII-=IR_JiDPB}TWh_YQX5jYy%`6%_B?FiMsJ#htGqi2)oo3{ z3_9rxH#KRuq?70w4&R!d;jo(l24!i$5`)avvw_E~w4P!4Wa$|Kd7_CM=dYL4U>GBp zf6oP-&t2axp<;0EI?78j|83K{bZW3pweIjclWHbD9V1e-FtN{RjwG%3-*@xhU$po6 zsqc&Z@7?VC!o;oq_mH1H>A&Aen6?#^{3o#|ySQ|ePw%Rt9Db>e_bFoQflw0{8ir9v zb=X5obt9?ipWoku-VdHVa(UM5TtLMkJTT!wUTeJE!A6WwmzWZZ&ZCu@}JwD9AJ_bS@CR=lTep<+W(-@vufOuh4kl|J`tTmG?4X;sKJI z-DYI)Ur9tGe`h_;`It9N?go_?8}N6%(oQ&x(cg4v=hI4W?M(1w?PMY<7fqcc8QDbU zfSL(1csW}(bEi}@!)oUKelvfj$w!X;)m)aH8CE1TV83&F`7S`>r8nW)$EMe1F{ZGA z1wU45Yb<{c0$T)nKLgS5+hJ&%C32yI=_|ZXuo?}B&w)W<2T~XX%#{QtR{9iy)v>ch z`zmB}sDWM(l7i)JsId>7uV}bd4mGlg!==MzLoL^XjT9y-bHw#}z@8rlL?gDG(589s z+k2sOWEN=w?Y7M5$uN`ttgVa<-qW6t^Hs)5ceWPg7ZOk&i=FF5t#{DJ@TYWoSleASe2DH9qQDqLwQqPbvR*8sA4?VdZ$T_3F0wPB%haDT;86O`G=Ni zFqvqH=TxlA|JzjlNypcE`&!<6iW)2yS$RhGuLrv96BBYLAwEHe0(a+GB(V}W=>2qx zZi~D3&fu!wha;JB$lMkmwuAoN%tj)A))A_t6PQi+b6K-=KQl?C6Zof%c2x3C3+E?Z zB$P$;5f&~;46|@y;sXDDFd-aO;yDYkrXo!2ur>N;*P~pw+ASyxnaT&Uxac$v-$=4;?MIzl^m*lh{BTE1i6tI+{O;z~;z=8IH}< z91oPdr~XT;VJ9|EcHOn!$Ui$aZ>1#%t$W>@$>uGYayO-CvUx*{>-23W(tO@gX+H08 z8kouFspw2TPs^h5c?;;H@p;b+1kP^aDDinpkzqISSVK@?CZBhZnucb5Y#3m8WH18o z;Tp%}VLBCDuVyp%0r(QwyD#=yu5_8~-9dX2)sek}t@bvuCov&mUclfP^9F-AlBp^N zZzXklOGyU*miP&UK;jEEjeQ1BZNk)Pj4^nlRWS*aIg`OtApwJD&+(i-d9FcX@RYpP z>-v^s@cMecrn-Q}w%+SAc&$x$Q%3%2Rpg76j<5hNLaq0>eQ308hU4Y#={8ntp(fZG zif*`4MD`w@LUS%LUXEOYucbKQ#v50GY4C1W{X|rDqgO|$c@LWYxmmp1;LpNV<+}v7 zY_p>+rB$_FkDgky>v}~7Kf*U?`b6vZ;OAThMb6$%U61$bWf_+3&?PIj_cg@pbgA9% z87qzP3N4KdIMcG_UqVpp#UBS{!O2472Vrop`T+Efm0s->UNCiwaxNe42=??-I9#s>=f?qe^;1h2)PM4;bGbmtUo`%UNrILTQtaB7!L6$z5C#I333jhKa`f; zGWXLiJz9ylj`$v}iARPPW;LhQxBotS&Mt?j-TzhgoD*bz^@IGQoH#^$z!v^d4)P0- zJ!i=cX3sf>g{sy&b@Gq3=X`g7(C>d~&-ujQpqZWRIU9>1BY}UEx9?LMcc;NAeBu6$ z=;D+$C)O8e+PKwKa~-Ql_A5Q@z;rlzlil)wwSSDI7{6#@gz=|_3B>LZpYkLR7W?V*743baohJj7WWlR(N-u! z4;`qIC@stZ1-plaUdnJQ%TND*%YBc(12M$R-Fv)|GlO8_PF~WqE3j`c4=%x$wj7sbng!wj>tjud;-AXJ70X(sfdrl-w0dMET3)KaYCvA>0p>b0X^ z!%AgdsZ7}J&LY%LGBaEOfwk^<4fQQhp-cIuN?HiL{#H1f1WYZ!TH9y4DXpnErv>;ODg_|N_&jz zr&klzE4JCLS?n*>(izH+#l>V6unda}2PeK+cOTXRKXpm)s`U!Xm|F4t3e!XO+FpSf z76~A)=@0j6BKiUzZ%PTB=bfCd9$6GB>UghO+Mj7ytzw}cu6)%a5bCD2XNvlYf5W+; zVya*_2J&8ojI*4OJJ;9*@u0q}R061&5>UmVJ*X0$f*4grE4_eufNx**0l*fde2d@= z8iCmXyG`xDm+q!!bIN5A4`$^^BXf>@M98~hTm_X@2}k3l6^Ma!p!+ypv#nL*sm@zz zkl|Ll7wo1i&-{llLnc|d!t4wEr$X;+J;h7EM^r=M-a}5un@_=N3FZ&D)zOIPR{36bDMfCHFI3ZO;zjP5FSmavo(_LK=A1I8MW?7yU-af8~ z9sMHmQbpV_M8w5f@q53B2mK;`Pbgfn`}s1bi_D~FFdD(ss5@IDeD%aCcRin8;thNU z#$#lI$E7Xq%^fHczN{ny@LoNUKle89!ia_%<2-CJsx3seosB+~VjpVeg=(e|=>4yH zpcR9GcqnPRT%cY?7NJC1j(`rUtI`c2-CYtK8G>Hquq<=3SS`S0SFkFqBHpz}{Z4;WFln=JRp z&4zJ&ceB(;!TF~vjbE;2_2K-}RX>(resn+MTR5Zszwjq`%^^WEJM+ts08CHb*eV`eim`)OH;Uxc4KDxy8JjtE^}p%`S}P51CHtRz{Jy;;Xcs^Z7w!6a5i>4|GaF`T}YakvVL!~9pv zOxCkL|S&>^gC1frukoyDB5$$b{jccy%_fd8Bw}IP?MdD^&kzN$1Msf)B zxxk5VW>s~+PdPlYryVp>jC8LCYJspO8&@E5t7A}Xnk@kZ(mb)gQ;qe-Kb&~lxRf6J zp+Az|u$b3+?{c4le_x9^Bg8FNuV0hNd+JOnIHvnL?pDd+lRMaQeWYnaS9wbxiQ`G)<3Ij84%w#v zjyx02y>yV-_9k8jAp(1g4t_h7tnF>{Tu_}W&jr>Ib7X8-BmY zNcm;!5yrjy6L~&BC@8K(j5qj8^aHy?BsxR8Zd@{dHRKMNeB18oQ48#U+a9d(h7@i4 z>vd-au_pSJH?)@<$0IFVQ62rrE9Ox{x%ih`8Or(xAwc+czv~{_cFlp#>%r!p=<@2+ z5qTZBg|@nyo=eRyyRY*TAm(HS!F`?V*Z0XFLhRbL6Ew_S>-9B8RC+Q_Y9Cg>)JMSD zwU`Hxu~`1~!c1mngMagDxg(r67u?BhfbZ@5uG~`|TJ${RD=XBrvrWdFH;Ar|4PHn< zTSesZaEHS|lknVcyBpO(&U^E6C7%0HccyIH`G#M3oDS=FTv1JILiwBnDG~@J^RM2Q zRYgpGbw8{v+95{BWZ?-<<^o2Scx*cH$So3Ypq{G8+6HsYbBbOh&LHPLHL-)!8F%wD z_V8<|FC$%F(uEq&2Iw@YFrE0bLTgf}v5HSYvl<8SBCkg!?^c(=pN4LmADQ39>`h8X z`)v4g_zN-Sn+^7W(gIzu*}t2xjnx>lfc_=;K{~O-ABOiTHzsk<1BF@ZWkC{#HEyUs zm9w-ndmUZ)F5MVZ1nG(AoW@J7cl_mrh^Xs#X&79@JI>U&RUm{AeK#8u20epNJgB(H zg7F;)L^(s=G1*lHtU*Oo_}V`2;@1QRtK_1}&v;E|PW*F{#GK8_&q%l15J%ea2A5+M zwc{Y>6*?((3iY<5?@lnyH4ZLHK$fpCoMah{ZQV@fH}2y@`FXxM;z;3!0e2CTn9d)q z?GI-b)`sXPnX5b?4uW z3s4`2_OFWV<2QLlGXI~_wg^+h{m}3WQ@_x%2uf@{{exSipX;YjNOg+*TM5`Uymyq_ zCcHtIZhyr|P{^;b08RVwjjS#Y#}4i=+Xo`NKx+;yTjd>r@tW~|ELKk(;-gu9e>dA% zv5^4kR$gDa<}*nT>;kFL*}l?SeQQ&l*P<8`6V*}-SllxMi+c)-e|wN${H{mq0xISM zsLUbCj%6Q1*`U`Ud%uh4vkzg^K&~2IBI~o~2wQq2LNi~Fe36~~dT8d7TJPLT4as+x zU&~z(vwyRfPB_&4bWnDEPI=>+x*84Xz!av}U(7g`2V*y%vF$Pa-K^O^b>rI0JkRWV zo-Q89nL$Z|vg>BIy_kk?r^5P>f60S}8|49RL|Hz9@)K{J3uYwq-!xs`{KSYMZqSY) zbnCC~QyxYRU#4q^DsOe*pQPBfqK&@ZoaG$f#(+c*BVKGL2&z zpEbx~#Hoa)KZD=h2uqweWK(;xnf=3mmArqA$)>iD_n)+eJ>oXx{ReVbXdCkW34^yS z?=R20(+2T}%KP*D^#8ZW`yVWD_;KD5h96GeKaQ6)Epqbyls|HJAose}1zKRW%amw; zrSLu;xQ>D{XdBDf7KDl!D?+0QG;xQk^G|#pVlv0I-cR%zea8+G+WDvp#8yYw*03sM zdktO0!Xp1v7%$!37gQ#6_981&d%i|jJAnO^=EE1u$v|rq9~X`s9!klMD6S&=YjPQQQdXkpfA)1YvzQ|p?r?{uXrHE^HNYyEUeuy@xyYaftYMsdTs17^ zgV#!%O(BuMxtEkRac6S_?XHP_;hoGSu++!)!~0tnKw{&fY{r)Ol2&b2rHsJ8nx8__ z(sT+_3~fY?&T&s$@i%(e>K10n4N8azU*J(UtZ1hT9B4$L(OU2AE5IVp@Q*e-7?P)h z5=LE`MqI6TV`1ij(2u{9P-iWkv7Ju{1ib;2Z*qE%c(0YToI4PEe1W%qidji95)V)J zl@8M_ri#vQQ|YjdEGeY}+6GB|Yp8MC+J+gV$b^z)=uX>kW^z3Y(MIoAgl1;J?HSnI z(D0n$=;Y8hOm+GO_#kf9*h1T|ANh<&@(T2XpL4hbO}D6jG>WPGL9O>~-vXM9qH2Y= zZnJ7!RqH*wFRL0M9^PF#x}#)c-<0l++9^Cl;nneJ#V}Z2UsVv# zxu{8>gPB3Um>h$rI2E0z^|d&2Qe+})13&>;7j7{VequKKIlE1BC1=irrzg9T*Qw;b z|6v-mZZQKE=->{0DFELyXM%TXj}-1N{az6LcewwApJJQ1znryPgPNm`bjE!kMB6*y zyv`H-2%NuNvPaXsX`UB&pKnr&bwEyMLF2tP(>^Iq4C^2Q(+tq5nUKc$7kr%OVsDcC z|AzAyGD`)XafGSk{hJ(3HwD|V;1}5T7={gKVj7D~uQm&desDRZS1XV~JosElDTbfk zuKrV=*VCTddX?tsrH|3Fmu9nYU-9eSZUv3~+;LbB5i(N}Z~E`|`r%SPeBTd$YvIDg zX8(PULtXv@GD4MeuK#|4A5Qke-h}%82S1$ehpYVXw|=<74<98=%WFou%i;R}A3Z$u_-oF=3koOGAbSK;>bIT*mg@7P2xOkrj6i8w`aW{ATYy#VT;QDgD3=6TK7+egjlljYg>Ya2Y$lDUfnoQu*9UZE_ ze27EU*@R9%Q6YTlj`_Fq+9G_~)ayUs(=0fnPWY5_avE>?9hZJStup;wa<1Xi^%TDi zd^%n-0w15&-YrVH%6nc$`SH?rdtK{2#;e1pLnz0`r;XVC31ha7PrG*eclgxa_5U?| z%I<0Sw37niNnw5me7a)i0H0QZ7RZyf_wngpr1SCVr{I`2r++SX_;eRx20pFAMmvp9 z=j`ww@Tp)+eERFCG~T=uPCxhJ3h#h!{j1jSY5n%w#3%To6h8goPKQtDePsCbu)VJJ z#_;O!=>f{|@##;-^KT2ETC@KhK0VCW|F`JVMLi6ku2&#@T8cDBb{al?I*?`7(WmAQ zg-=K1`uKFq&MAC)tB1p>4v<)^*m*q8y>m{?Ns3}I#hU!G%)ef z34VolsD^lH#IJBlhYF8KS2(|OC%@esbh1-*AKHfw>(U-tm)rzNvr&njY+8|D<};|` zX+H1oc$*;XKES|Nu4u~E{CcdUec1GAZMS^Z_$?aApb6XkFBJy z@9qk2mhW!FXWLR!N6LsaMzFEJ+v4dW6#mwt{r@W{P!%+$d z$BiOU8R2$qd)m_di1r-x1KYD?q4jrL?K$M%wUK`>8s5Q*epHL0<(YhpZezSS@yO&7$>5n|f zL*Rky`*3_vsA5!2ysk7W-0vSJo*)N*k<|s&Wvv&F3CEAg4oBat?$?4x^SYYQoMBlN zFTkK?)%07yZ0^anv}ZE^vK(RMBC%Wtx*5pkDVbkxZLAwo6~8uHHVr6^aA@dUnz1;^9ZcF+bgH!ZDY8^Ah2ESvMjjWE-Ovr1O&_tb zNl1I{;#29bB&zM6VD8Qj_=VU6#9RBQiT!>08^EzHu1 ziooBgU+=cO=Q7ri56rQJ3A5)sa34+njI1 z43PnM=pG!7U7H&|@=7k?JO?@&F54JtG(|!DvX$QCRT8vtSd6+h*FUB*_pvfNl^6XbMZMNG)AVdq37yCLIJN?f zf2ABQ^Sxx^!(3a`$_tIZvg55yotE^2}g$J7Ph^Yn$OX>$;EJ%+f(jUJGRq=i*JvtY%f_7eHA;Y z>R8Vss-r!Jl+C?(Z9<63d}g@D57?M&yUXD*H!DrSKu+)j3Yl&DI|v`!KH3u9kz16$ zi&k%kArnbRnixn?(U#D&)m=7^;{)b){)c^3N6L0V?PkX)tWD&h{?9XAj~p$t`&u?%wl%BIL0cBs*={f+gxMB9c~X&phy zl8RZ|X~^RF1Jgy#L-?edFju8M?q?r2u;P0Ur9bAloITw~nsz5mqh9Yi(s12SMd-E# zB`p zn8inQGu{xT75c5t9mWM~(e;(FxN1Pu`*Tj#%G*>7vX!$PFCz&@(1wvQYthJBZ{2lp zl#vC6l!UXz>#0Gq*BkMx!dZy?L%xT?#Qv%_Wz={Nslu_Qku<<>8W1bwt)ztqi;gy_ z?IcqFm|Hkjy5aW}fg!Vgc+2!r+3INI1z-W%*|oc=U*n9g^G&T+MZk_{O`R_8+-_#n&- zr|fWT#3kKCJQ=20S>AO+HYQQ=ly@}+SFN`sM@$12CgsfH(g!7?i{(!@NnKYuD^<+R z=3&aIEbpej=H|mMInOH;0!#WYYu&Lr+CIKvAcj&s)h|qixd#D2jb?qoP}#0$l4FC4 z7x~lB7IO5ahKqc4t=`C**frQ?V8{9m-29v3=DC0H8gPdTxI5SH*k# z4L0zO48VJ?#C|F8R`xRR4pV>~v;*(KX*qQ1rSN8I&}eUo#RgziKO9(Z*x!LQ92=ho ztGN=A=+_5pUTQyh0j$!RIyk!ezBOeVnY&e?W8OBn!umS6KKo?=*DtF3ZQ3?m!_l^I z?1DVX6I|;gtoSkwSEzuNsxuLG8C*H)#CSWne*a+xTt^yQk9g#J19jgyU1>IiA#tQL z%_+19Z@Im8b_~A@8Wio_0#?H=-A?B3-R`E^lSOW-eXySk_Z6%V5Wa8k3lrDa_l1de z-ZK{@c6jZ^_A4X( z^#8ZmuY3ee5r%x}Fvi)heBKE?%1s);c~$3#d&T1Q$qMI*+wK{7(`Q}o1Ti}hWSJU> zAm_ZeltnL&QpS|rrnHcPVWj+VGihX@;VkzdBjd?rN~DS2uI9ivXDzeGzbTCsw}jeN z!ZN9y3iZT=RG*m91vDwo@PnJZ&`ztMaI_n$A(Y_W5~Rt#ffM#MrhNWajSe((lGJ$m zWuL(=$RJ|ASF=L%RjP9sR(HxQWmy3sLJsNMLbBTH6!Q^n4vQvcruR=Y#!KfSY%3dVJu7$8C2kKQf3mk1J)NwKPQHxG5ZqlRsX8yoldMq zV}UV6pgi-tvvsJ5@)dlb%xk?o!jx8dXABm$bQ&+jNUMeOA!@>%)CS6@YQv$+!p-Lj z9Ti*ZhhO>O$9|ZzZ%1lY>7-Y}*LdkCH#;L+XHxI3-OfN#t|AnM+9?sUt6$-bG>fCs zJjC4UPK92hfynxj4KUvE!Tn`ep!?qz%9rS((5^YlyUUj-Ui^B$FEFg|dJv-O)Mc-n zZZ^w+%mFRu8v||UHB-m#N7kJ#Qd<$gcyQ`DIL=G?Y0S6jgmsN!bb)fv?{Y$W2`QOxapn3lMQ19Hj@#+t5&wwR^Cm1Wj**aCl+<|i+xK|a8 zq-UJ`k#Fp=X4TN3CJACU^RN|&jZZI;D%SbAq$>0qK~>X>rjb`wDeO?yJ$7Wsrk>TN z1V0#dAa4Jggnui%xn|~5dc#zQj!)?|g^nM8X}6Cf^Y`vr?(pswLiR!5d}V%b8fDW& zX(zbs(xT*f7rG+PBu-lq9E+O!T%2M;K#b6Fqe=dqUGev(`%1z0p3a4{CbZwB$lJr(g>}e_I&te=DMKGx-<8&=j&I*_#NkUqK)8$D_$4SBD(Jj=xf^!9 zk2FBmohzuD3xQs%4^2%5H>x^}E{q4XM!{@8yin92AUkAro;7TIh0a?U-*Y>T?=$X< z7xdFAvx(uxRh9Wu!`TY@^UvA1rRw`pvqQ_vq%;jY85*em1QIhdk1l5uI-ulBUs0Fb zaEB@M4$`(So_~r~ZuVc2!Nv!_@(TP<{%=;l!-yhrdI8y3B8P;@{Or#i6)oJ$QBj4` z5tU_xqOzy^?>A(GDS1k8e?dS+HO6AlI$mGUP6s5bc;Y})2~Sw>Q1G(uLP2-nn-=HZ zYM3mOO%05)<_(nI^OLA*cpm5AtRCyD$;qv`qkZ~t6+LaV5 zyxYiI6@SUK?gMWFmKYVGnZpmLXdT|C|DdjQ`4W0eM2JDe0bMIvCx_YS+blcvZQ}n!JivW%^0SChxOllV7kMWa-dGo71{B^fS#U zSyRoi`Es0mR5M@x`Ic+PD}-vte1+%~t>P8lGj1N<;=&;UV|&*9KVQ*ToIT1MVy%`Vn(`%HS*I<5ySi<1``5}3rQge9%j%G(;P27+RCwA zV4;_9Wn+lLhtc){Q$H`MzQuFyTtL9KX{7j{(GdJU&4u8bL6n#{0I7vy5(5I-mu^p8 z;tRuZHwWR9w2-p&rTtYjkktDED=bPscd0AASfvjHlvJ>*U%@gJvXaEIR}^09=Qu}| z>Kr5(2T9>e`4;Nu8^cbu!UI1dGTX&}h%9$;Cd)#U>pi$qOd}Mr)hl?Dp#pJRC}7Lg zP{_=~?thKlMg44cA;>vEXXU4#NwB?3KQz2Js|H48Vrbg&*W0(Kl=trTkAgaPc;}Qd5`cki)u7` zSx|Dehh4*V@dn>F$VX!PT5kwkgWNElq~4`6VoJ}sw=p^)jdG|I?_|I50T=#$o>1qy@L*l&<;F$5};v%d5tBS6M}+N!Lb{2c1aQGf-*rzi#;rYd*xfC!3Kp~CLhazoSZqU)tdSOR)?o~q-rkeUT3RAlTL&iMii)5U_lF^ zetKT)9t|UA@LcP?vi=U#-dy>Z2NAh}HC$}hSD?ZbSTAUbEYTkv>QBn^*|O02X4b3h zOqsu21lrn~a^7Y}b{xivm8@ zF<1fb71Zg@0R(k|8Pm-SH`LGhPokFTK4bP+Q29+o)2K~Tc1u)j@xS~BS0kaOKM3Ww za5oajGb^rpsZPk6l~ovOx)$Fp6|u?kUnnif%@gz53wfqKq0;B(yYN8&{V@Msp?IyCe)={)Y$r@Jql~}oL4TmW@{12L^pzj;!T>sb zTC`pW)irS+ACJa1KBdGbGu*3^2_f3=0aGV}f_F)TW{gkp@sG#rI? z?@xG%49AA&YGQ_GBjM2RC{ZCBK?YH*KC9a#gro(!p@1Tbb^I0$kLE?FZ|(DaSC zjbPPo5RV|zMhmp-47SvID@=n>i&`;`utSaYC^007?aR@P7Uh$}GCedLi7v@8r8=!5 zdZxw&6@{Aiwhml^q-L!*=sn0W;(gc;FO0*HN!9+T0(7s&2p|d$HEsql)$vOT@k%!3!fa6e-@t;%j(Mxi}i-vl%CaZPm zI%ahpi$7qv`872#(pJo35$@O87IymcvSsLbqZl8=hN7!Hu8?Q+yhzrE7D~vCo>x^Y zRmF>jwyRXiC~PGt`ykYKwccYiAWqI0Z7h3PdfsJq`=rXuC@*>{MJ@C7yvL#EU4o%O zWD|Pc$60jblOnQ!2Xpw*he2yv5^5SsPo)=*9*25&dny~cE^#r45YQEIB5S>x_aFfb z8JOTkqAGrVH%=TU-TlBVYNiR=UVGaK+VVm(b>?^Ape}W1hm+Fpz$ul7%jCnb@p>8| zHg&W19T6-G@o#X2o*B5j#8*(qX7LaOwO;WcDJMWEy<_M=WT1)-8nko(U_=K4SnFMddjkJgZNqK`ooQCg1v`w(! zA|PYUSmwDm?|>VbpjSq2w1$t{%^@!}I&B4+zH|)o#x+M@ zP2zl~a$q}E9U*rs7w?9vl@gZe_Q)a}t@O+q5$)Nf?A41u-)bSv6r$lAsRMTnj9w<$ zqn(`jLN?2lzQcjn$y}?xTR85-O98KLRFj5JBeQkHI|UCi^3yocJ6=z)nVQZi2|}o8 zoiJ+)4#epf1OGAGXZVkqig=*2|CpUR`j0u7Bt~Z(JVpr~qlP&5GyejbK1{wbn3(gK zT^&p&wI?poL<#txaQvj~$Tv8xxq&uE-%h?90Jp5&qQ61R6#1|-fV0cqzWDRR2YfTM zJVI#M7^6+mC81|4@jw%;+#c0C^ox;9+=&`Hs1Jt)+xvco8Xuz}&S%Z`fzO(|h}tj@ zo0RuJK)JBE!s@tkUHOzN^ES+*mUD=Rz8-q^syu4Bs`uu^Xd*J!mq<%?==y#l@w2tA z@$1Sw=YEKJf%lp)8?9RJyj2p~w-`d&UOyxD z?1WBV_lEzfxR%}W0xh}&9SICpz6QOP>rY|gV|qOZaI~qAV682|b|h7Fd1xy8Ntti+ zSz9_wS77xJV-+T_R>92=I|}`-b)Cnz&-)BzY+!Ur^oUqx?sp6(wlw}=z5=gjpN7P) z;?q?-@(HB+31eJ|`Hk;NXnW^MD6xcS;lx$KiK})5CrG=VN@G{9i@qMcGB0*j9)*D08nSr|+0yXb4|62NW60i4glUvP)!cl+nfTtmiW8Ta+-dL0SFrhu`vQgN z3iyhZ>aCnaEhQAWQHWZBr)4}P^K0G|Ka5ay;-22Wj@+Ov;O*CG&$!CF^%N$v2^74^ zJaV*a&Ot&eR>e`-t9f7Ly|h4^-p=#AaTn%$GJn72aD1IV+N}huCde|Q@Sv?^uE%A~ z;K`Tkdluh)ONg>JBpq-l)iAI}X!0+i#^4jodE*rNd{(#X!`VxdE5*U?!6!dk7qT@R z*@K*+$vasNxfH)5Ge-a$KgZI4Qc=A0$p!-f8NK)3WamuR?bdL*{>kPx>-qvF#umK` zHRUjX6!1h06Y6;WAAY0LC~*4EuT%%-CyoYyiIYgvsov49-ow-Np53Y53;cSkw_a~t zxDhWs%+-59Q14LH8_&6L68#CVAU@TJlfi*zVMV<3wQJpo7n;M4^e|tLbbVbmZ>xRX z-PQTjdDhqccmCS-_0<*D*S0Vfe}370S#3=`yw5}8asy7z%X+@RJ^w||zjn{l^n9Y8 zxuSe0{VCTUu0vVnov0=Vl69}EM_Oz??7r2-CISVuo#Lbs_ZS}U5o9eRt61h9M6^Ah z80~5uqgv$|r)^3&zV`^d%@TEs=j^5Dh1SWv`{;R^#qZr)&r{v==1X~=;+|LQ`6~DP zj-DsF=f!$H+n!6C&NX0d=1fdv^jAp1+fa_N0idlo?A->PJgrgl6rUbg^)@+6xLYS7palW$`W?(Xe!GnOf9QGb76@nlZ z%CJ*MlYbtLb}N!rxZrgm@C!v*#xY(}h=~^dy)MGWOCyxuTgG0`xUUq8@lX#-anUK3 z!i%=7Pd0oD?QY=GS)3CB_SP}DOyLGiVdE!PnQwLd-FwtOA#sSQT9zjyJ62NinJn ztEcq(th>8us@M1$CUdxPT9Dg1w2;lf{DF&xt|pW1triXaN{{2BSGKnf{RUI)??e|H zi}FN&dJh6Tn5C(sLK^&;~&{c zR>fGTX~;mgk;#39?v%Y(|5lJ&_TRhMfLdRi4fuyIi><{M1MoL=3ZR;X%l4uJav0AB zsIP5fBn^GxdH@Ynv2B=Y&_PV6_ZC4vo%xe*deV&-Tu-qx?;SM2tjf=(_KJq@+238w zt(821OuD~Pcs)gs<5$vxcOXUF>E5$KQ(FY1iHd6`PH3R#215PBCXBTQ)HzBJE~`c< z+jtQLQ&gx?_f%-tHj!Q1qHf4b(DY1Gx_Ig7R|;p>dQZG6;$ub|zw7^@)6OA|TNSTElwXtVUHd3@YlXros!kD80Y}xMbEzrQ;WB+LuPj{v4i3 z!sh|p=Ppo9A%SMzdB*4KMN{*rDg;st_ayoE$g(-kCGYYU6pr6MP4AIbE=9?GdIyr( zo^OM9+~2hTZuI{7f-Bi3eQHkMd!n1uf7RD`=`SvJaQNdtQgDd8_tBo_IlcQE>G5o(R%Qo z#|F<7t}I^i6+rm9Ko~FmB|L{(-5ez1`A+|N9CLpf&lLQCqP$DhBsXtP?e`sz^Zx?~ z&un!doI!54BVD1&T6on@uq(&wsR)XaaKDf2V&XYp{pLQF9GN9Km90q9U0NjQLYBdU zodF$qj%fvgxM8dYG!qS|r6bO}FKf1%tPCzLXA=vTUME?SoV%wp9JD2uSN_?jcCtu) zwT|}?ZYw9etpVbt(Ybt*`0kA|b)`JveF-YO3CzPj=7W&Ju#? z{=CL2J?Y~|CKt4P6hgk%ZUq$RcLJH!cFFufuZkej0?OgtgwrWsDl$OzCqjsnk7v z2-eH^mGqd6VRh^=pE|ih!qFqEBHwYh;bkx!JV~y{lbC^~C=b=qvvRBB=M6l%wc7z% zkk?|B9ltWn;7+xftP0Dm!YcE)C@F(`e! zmHvDK4M$XRPtj=JRly`z!Nlm$@v1-*BGmg-W^k&4(Wwe<4=T9ORRBSL^kE7N>d@o} z$Df}?6rGrElSp;UusWVs(r9EFe^rx)D%AVIX1%5>8=9(2ATn5uvC1BtLBgOSjTyao z%tlr;e_%AS%*tgPa8$8a&#ZGcOd4aWgd<{y?&|mqMW|&;ux&o$r(jtpK4aQExbd=& z9G{`wDL!MRit^8zT|p{YH%Ho#(FG^6LXF!Q_99$1+3Sf`)R*<|7_QqVO@h>FYbdVO zw!ccFljYt+-->1;Z}(G=PNg=|YsGG!tJIj%_-RI_(#%MexZF?E;HL?v(ln>i%p?tB zAE*CwY|CMQ-LWm-DH(F&iHRmQ@#C7SV8j4Rs7d}w6D#@Sz}dMZOugl60ahnNb{#yrF1Kr;%HIgNfTT( zFdl|LQ*O$!k+FLd?C{5_fUr zz>T-icWkYf2V1MjJ~oWiN7+rrzeq51x_d5Ee*llMTyY^?cA z*<%N7h)n7$W?v^m?k7I+6@OCupFP!H5eW98^4jL!@8r0}!9t84{*aMCy zc|e;#c)`)fa#a)T(KL;1kJjRU?VI`4TBOs$57^ho3==d348u5l%Q4e5mABc|q5bD6 zo7$hs!}(j=oEEAY^vgQFJWQ#;Kjp7PZPmAj!>mqI;a`F?Y6*x?2jj*D;x!*pyK%nw zQPEbNtxVBjcfL5H`1exrsq{xmL~|N{g83TE3>+bZbHU5RSxBf-_y{m-y{9kJeYb27 zN@w13OWO`hSr9TFd?~zE`{mT#?q9P;Gu9{B@g{^Tg$TxW@+f{)GSD zh4;kNO)j#q-Gxu_o;br#+20Re-RM64!cTOhU&6P3*i$9#s6Qv`PhUUV>3%rK&pw(k zQR}}vK**Lm@fSrN??+as_R;!tzE!m-vDyDp;8%Ts-`=14;i3NbKltB!d_(4PK1!dM zh{+TwMb4Y;=G(|>c*@L^v1}qrTI{}%*luFY8-M;S-Hlz}!rTtWZWSBN*w*isc!sh= z?jDd9IF@;##s`QAA6ZcUc`WiV$smM#g=3VLqT%%eu#qSh#s+?hGrslKH;!@T>3d=Y zo1SF;UGv_7y8HKKbdVPgWHIO!}na8i7ePNAESP*Y9fKby1AFAEV8 zBIGk6VI{m5JBRfzYR3gPu9+#@1865?5aHaiF~5h1AT#{zJ4yK1VJnFf749JHMg~zi z%+h#c9y4KK-N_~I)_Sl0VkgoqDp@jTKh2PM{>|qK&!>EWuqe(kyyyqspd2^X)sqy+ zLQSur0wS0xW3+g4dcli!b4?baw%3 zW^Y_bdDYRg*(`Gv$yliYl3wBk5ga=+_>bLFn{hYF8ry)cHj-9a(mu#uQ8(x?5O<%`abQOMdwj2asr zt~E&JUj;Iwm#IEd2PPWtq4}|$Bj4;7YAohaoo(P4et51|7;5?nF`Pr{+w^+ITH6yz8K^lUvLb#WT6ZCHii1i~IE?xy3+z#C!>f_A+ug zsf8NNa9LZyI7AoM%@LAJb$d{1Rt*5?#!6jP@x%_2{3A9ZiSz}}bY15f!0qZ3U(s-l z)CF^hLtSvG(-xHH_}YTEyNC|C3+8kWsNu#+D_+~iub;jn3uNs(M?#@we(x5w!zxESN{na2(2UTs){9IKzpxzy@U+i5vkUt{@=4^_h5R%hV*A2NC3$32 z4Rq2O+~dYg`$8-ehtF2e+~>pv;%%DQw{6FEa3u{*j3%ZUTeI@2$b!7YZ0#GHHH(XD zy#{H-B(pt>d(0YV|Sd`oui`xCEda*ylC~Z7bB* zP78&2Mv!XHv$e4NE##(Tr!+UZG^hMRX`bJ4D{1N{)_N1R`r#-FpL3QnzNYW=ceL76 z(gJ9vwn%siim(I|jPy1CoIui3abg}^STg@*1}T}p!2Zs$zq9P`Q~ZT8CJy=1aaNUn zIDs(nj{ow#UUvG*rMSQkYZQ4SFNsI}m?ITa0wN}U>GNX>jsJOGVu}p@=VWahgUf;5 zaPF;e{}ow1DI6P;!CbH!(yiDrxjGkBsy`Tj9oiiL6lXOs=_`v4E#zeoT}AubNa*0*MkwjQYI-gn1C0^0bynFug6+C4t?0uoD zvxGW_MR7ic6-ve2wYe2@*WfG;qr8g9s;(80Pr6n@i+8DsZmeYO$!q3#e;)pw3Tm>y zt||LG)Od$*C%RWrRqXg&!LvHrR)w)ub-yox1~!=ME(x&)S0qm^P_$Bo`z@}Cb;tG* zgAlZ-Ep_Ecy|9zryj|!9*>z&4!`*2oWmiQvRrPzVml>;% z3c#S2pKA8u%#|{?%Lpz82ElqsO1PJU3Bb#>&-OxwnXmmbZ=f}&szvr%@ z-t;&!4_@W1`ZWMG)Ywg(q`ZFXtK-L_ciptzrQFrD%V9O~p2aoMU3yeSU#r4stUCK2 zkq^6v8sDVRfNqz=sMM#HQG!B3ARB*ypPxxZBqs&_`tW#o+>+4hR;&k=eZCxiH1!ty5CH<>?ghL|wwHwTC(=C*d*jrSwl?sahFT0qSM zv1OR)x@2u{r}p=Fic3t?5*pl(*>Al(%YEf9%>Sg8&}GM}mzs6E`^9u}wWxN{q&ALo zo4(*+Yf`5~e*cvpg`BD}3MF-1@p^jUv=!1E|`~72LbN-k97%!BB(3;uVA7dpgP5EOy z*xL9G4Nk+CF+U+SY1+_y;CkoE0#B~6c_;*BK;83fm&`Tnyx8{gsXyJ3HK!c>Ff*m; zyvbS5JReo)H)GnsXWzT2c`x}|Q}d;AB88Z>-gD2BD~2E?c0ykn8N?@1TX}I+bXcL} zEjx*Jv&Fr#pel;P#YMyNtUT`H<+(u5!q=X%n>dWk<4~)V&?<~}t%{yb z4chEqjk_bN$D)+`NVFj7IEeFE)F?1iN4*SgAO&w}1@AGv(&l9TpB@()U0K+6W9Iis zWlrZa>1VaxZL>A7%(S6!Zz!W+NzN)x=3nS%OpLIpQ7LkEP8Y+&{_OZ`z29}^1;gJX zSBo`Xl&JB;^9VcPr{QLbX4Y!f9xETf_3{CHjGRa!lODp*my2235rtfF=0Vyk^SoKG zF>|!(F=V*oA;{{EsrlN$PjJPUiqpUh+RLWC_XJx<0=jD7;4gprl8VBV=sCK+sN4pi z%B}Jf5dtk-RBnWcBM=4rFCJn>ImNWla_>ShZ9BMr8vIW!X$j=3_-DSrU%!Q) z!8{e*l^^@RSw9-?M|t_;jDUOd{|MfK*C`HO?0>BvT-N!Z4~uK}=a5{0Pj-zXSjE~GTmr2L5jB`Aa79Lj+l zv>dl7Cc4OOP?CVp>TR${nsot{0zwhRdAGWQ&7a){DUe)^sx!7M64f9<77US9Ia0;_ z8vrjf#TkchG`;hqq^tFMNu1~}Fh-BuLi*{$B4Fu-wC(n-)O5Y|Y4B5~gc-B>3m9%p z?$xU33iEte>wPL{3ao?Qc3`Dv*Pkkgj`&91<;(AIiD=fDwca8qKnz?|CYYw^& zf(|zt#{zL}hXvwPi0`{~!y!`z-p%x%hr;mIxr#BgB+XTf;7v<8YiwIvOVGX)GujBa zL?1<`;RbbK!b_9ZCr^-O1EpVSm&=7=-$ExeqccVZvo9i`6FNu$c!rN;?wwtftBgyD z{4j0U74OT6UlxuJ9~tW1J=FX7JXb+XjT99e#GG_i*_B z9{ll*K+8qO|a9b;S-r9(w5GReB*GzjL;1i~87zmD`3 zF@-H?<%~^wJ!7lEa)!W1avT;kRyP63=)2_M4XTEpo~O7+s}&(%pQnJilOQ)zFX-Td!haL4>L)-CYCUXCJ-X^ zoxCTFiI?V?SaYZWc)k2NO$_z!DrmPlW9r9>8fa~(Hx>YJB5dkW5@fO>NCLe6Agy4p z5uHSPCaHg!oz~Ta>a@aU|J$8P#7|H^Nl-sVrHE7T;VN*5fAC!$Z808>k8p7JjjfSN zH>gO=K~%ivo6k}>6Hel+U9LDtI5*%J5ARP#I7+{5Gbp)msj$-}+3fvBCW}K%_!- z%JEO-<<_3wDF34)!)lD04mj)frZ<7f%t!-`K%lRo-s6kGlPk4(K7o7^p2qX9-^=88 z_EI{`hIweQIJAd3jLbw7SB<{y6pp`Okk0)4LQw;) z@cxA+&xTD2x#AtUY-TD)Mlx2^XUW8hx|G1@tc=ecnY+m6Sbb&|Krm}Z76zPa;$xVn z)clVB=k#|=0LDV*Ac@>^k!nVR(?D1q{lq&UqvfF{ZE;v{yzFA#GT6DPvD9~ht-oy69)G5{9F=9K4St5UTHBx|D36jARj# z>3jsyldK?kwG~h9#)2Gb{4LW?^5DHnI8nWd*}7LZ8q59PPwi$)Pa!N-t=D^hZptAQ zTTr+K#M2pfROXnSjDPT?5VZ=LNE{PL-oDDqRjalb*rwB=-r%F)lCcvF)=|mN-wHKe zMbi+e@7A#CVDvQ&o5qBu&h4@I;p{r@O#MI1oq2qeMe_dxf<{Frs9h9{Q>gww1YJz6kigUTYzxnKFP*|5uSEKRT z6GpVj-OF}zZD0Q(G;6)8)vJ2B7fhjcp_#4o3Hr&xzFhaDrULRs&^1{WW=8Xv2NC1@ zPuBLkgotS^b^h+PYcfqU+= z0e@Td2+=l4EubaVQ#PkK|0K2HNVOqL(*eP~`FV~q%LU)&Q+W_T8veup z&Q2=7S%>mMfbF1r(`rFaTazJ4Xy{?%lbWH|#H-mwr`R-LGb(ryO-+!B87+Sci=Q{m zerMVUet!eH;P7b5WfH4E^`kBqeFc(RFz`!DP7qwZwazj>NE+#%CR-sw|JpiJ^`^aD z)vj%wnx3(dWifP!q1^=q3&z_9F!Y%zvNpTp`nWhm5&rNp<(yM~+7YFIpL3)3X9vh_dNrUx2q&8*Pt-8p7MK7LNc{0w{ z7}4a15<1uEKu93lRzD)y1}E*<$`NSverZERZA~lS6iB3!q?2~~35%9p1uVU8qNuut zsswIO9=NF?xL0O7a2tm7cL!;g0Md6BpwbE56GKQXzX(#?L0Wl?YtytmNZW*v+J|aZ zV`<~MwZJ;5ey-cW%(Zlo{^l*7HjRdhr%aE5H@lG>`K}wuWROk`(mAB)gL&(ngQMO) z{s~5s{pn^QWy>(jPMu*gY?{Jwd$v(y-hJmGd^5*+o)p z+-K$=g8b(Q+qgEJn#51!#Xm;T46uM!NHm|gB4tuTuUYKSw&OaL=Qb<-3D5ot#1)@N0n|DXE< zqB>)T?~#qzzM@`H)EkMVbG3gO60llID}6_#F)Ojz%ofWxLIR3YG==&rPb(w23b*h) z`x=di?ccF3&3-L$Uk|sh_!DGRti~P=p|#<3mi`;!n%mf}xXCR}Yu?e=%}px}?a#j) zKNbW+_*vX9#Lrbj9e(z=*o_W9H&N&M@blb#0e)h?`6+&?kniik&-<%4grDjO8^X^% z_I1PfIp;ycPaPB_yqxs5!;jJsKY7$szai?m0CyLoo|y*-J9|Ps=L~V!DYlTBGHTZO zmr&#SuyfbF0e1G>Uf4M+k9wB-d*Fj=F+0$iCU+{?0LlFE>J6b}2m9IqC4_zw57s-9 zNe_R(5OVY+hmbLg9YT~ky&#{PHi(Z?@c1!&+_!&-j}wn__$X9BNpTbSIExC`hmY&0 z2Ke~Rc0a|(wg}%4ADd@3L?XZY+lKJb&Ax63AJs553^aJ6!^h2UIeaMnDLz=fb#w}r zV-G`#L&zTfQMP!8dv59*;^VL*9X{4rR+XzT;k$6A-~@M}yHm7nY=V^?!H0`}|++nLK}Cvkl1m`b29ja;4ABIt?fejUF#(DgZ330OKt$R7rU(+U^%@EVrLwdUhl@?sy7{# z<)`8Lx6=HtVz^!_R7^IK4XeG>emw)`P7DX;wpmZli;>dW_IyIL?eIs`mCm2gd4+1L%cqST{%2Nlj$O_}DuC!&|A2O-fCj;l?pX z$ey`|zH7($F2`50iQd@$!G!37O=yQA@C7}JCAqn+xx;OYTIbdI$xkp(rw)t{l&$)) zW%lrLl^*WVgU7j8vQX~@q1{Dy-SSdA6d691hsX!FU%E_@j{X@3_7 zZ?eX`gJa%BMI3gK8Ej+DQzX?B?9@MQdNZ|bKUt4++h5HsUZf9Y@7L1BTH`(|mKx*j zl&7TXR;{SFHIeTssCJNdpN;`kwR|3Fxy$k%RXE7I!@cAD>s9Vu(ID?8_YQyBJKekD zLEg3Q9c!K&+&grIfhB{yE8Tm-;MKVIy#{%gxp$aH`nh*y-b}q}$wXR?aBuo~kJ_76 zk(MfZGrA(4y1N3!-cKmDvp2fJ_iW`ox}wgXQ--FNYB6&~s_zwti2E7imW8Nt{x(as zvRsbGBdiQ>Z+56Y-%tdsNbi;cfn2#nYs#~m_}Q%u-r*r?dGQ$f?1 zR@T?kODGa)d0a%A>btY$U+q6cetnO$+`|VTaOsrJRQg>nG24ah3L!wEsgy#%_z!7Y zj-6YMh*$M#<=$s5K~r0F>yg*5CNo`akGdJvCF6L4H2q;*HBRKl@#lFnYt{!8gyT@?58xNR@XQd6l-3vUZ;j)2F(&o%gmZGLo{i zcZnLtuOdD1+t+LUsaNWanLz-p?0R0RlxKTSW24OPb^dX`yAe89(1+pdfIueuVSgF1 zSR1JGb>N_hQnB*)8_!EiriAl@x!SjU!Z&2KJh_!Eo?I%7j}oa_hyca$#(2Vp#_8{Y z66Uv&EC-#xyjJ!q>v=9r%S< z!i|>&o34mLluOkuhCieT$mp`bSSe+(x zu@Pi8KV_i9uB6T>u-nmPf@({3{mp5vhOkoh!5MHBb3>!DMSkjca*JK!SE*Lsl==OG zUq!RV0LcGL(I*%&@oiT0L|Uv8+EIkEp;Nr!ijdnJg86mbGL5x${@^8=DAxhjcnGuV zs{WcthcDHvDr5EZ3r%p<`Ab)*(Q4QoymF16WWUw{)36ig#5HzIkx%PStNv`QKS-9j z?WfuJJl6SUA?Lz#$e7BGOqOf>0|h$O_2~2YPOa-;|GbAgV1NCHzvC2FwpTcE5-xTA z=-xLLr2inAgg_;s+Wke1NaUS(m7CEg1n!dvn*HjT!%Q>sx95_2e7 zsS_?u?OQRZ34U<|N51E}XRg{C^V9qLf_vr;Zwc?2bNBbH{Mq~a+JVRYecUq_EvMdw z8l2He2yzKkvtC)|5C2ytW0NY^{p}Zd0{!ZpsONS8!9lm(deI~ zl8^LbZnPYDk;J#OC-kY4dcb);bEcb5KO`mS8@s_sG0Ci2S>x}wFWE-U27t5!1pr%~ zsX;Px4Z;`x-EXQqktWnqjxq#>9413~5ZpCZbjg-L*5)?Seempq*i_Xj3#SfhaH8K^ zgourceL(iYAPdg*Haog4mM8ueeAlTqGZzJ2K1jecN|76$OOHey5nfEK?ekh%zUCyy znwH{dDMzHeIlc04`JH6^_s)0Q0dJ5>6lndI&DcCt~Y&46FIhiY#7ONQbff1I41 zfAZ+t){i!)dB2b9dU&?YgE2&2fVlNgcbvYm@fU3Mx^!jeYKm^+zt>22nkGz7poe8MQjrdk@m~+uO0K&uzacC%&Z*OBLq+ZFhsp5tQ#YifL#zn$OR&&7h6F8xevpFL1 ztpKRe9%w|uFVS>scNUn6qOU!^4i_bYL31lebJS)|JGfHG;AYnb8ZGz-jd1N! zY<^w)av1HjE4TESyUot!uZk*2^_@E0YF;nV^aNJ8I!q#(*V9>es zV?``klVwSyg2ed0FIMZ9`!mkYyI1gq!`64Npa>=_HnGc(n*FQxEBY?Pxg?K#p~OC|{&kBO zx^VyN@7tQtsR^V7T6{2-1U>y~l=W1U#k7mG#=r1twzWCphTw(-%53kQalh&nfyIn= zXXP5>3?yCOPo0E_rlx8y)i0)MMU?I**C?a_66F&ixX?dG>+e+Gx5!VMwN?C?;JK^Y zXsuuUGLLR7I=1i#wa^JxaY-nWOWa898+; zbKS_ebbk6Ao^#@a^~E~!r23nzJi)uxZH8?~o*eiku>ipG`s4{{MGQRWSnMbAWNR>! z6C_=~?#4D>YKvRGMn0jsI8h?`wAievk|!E$$rIQ%D^H{V1@h#%mxOtEEtDtGWSTps zkS9*UJB}T2p@crO?OgwnGk;d7h)b@u&jC}KpNxh)&qFBH_mP9$OtM;4XT=JIvtng$ zzWjHwa`ly%=j2<$YViv{eY$2K;Q*l$$`F|@Qd@jMoBqf0@vWyrk8O<-}@6Z4HNCp!og=YE1*>2IMNxrYp!z-`xtK0c!9k^a^7 z8VP3Xqdf#pDsZ&Zl*Htj=e={t3#Jm(`6nLegh&l7T5zBdM?4TBuPYCzWpzxLR!5ag zPcn=6duS6P)pr6V5F_$cx|`g{kMI)6jfrZgDe=cTYmmQ+R5bd9(p2Ap{j8vWuC+sp zFvUTqLoZ*W4x#EVQ2>-FV8nmgK(G&d*_9HAIg?gSCXS*unpWrUbC8&_@21u+Y?0H< zU3pXMNnq~ra6U*f#dCdwvpxCuX(Lf{U*YrT%t;g z{NcGA1X>!a9NJG2if9k3K$b1e)m}b!pl643cj&}F$vAhle~ePe1HlpF?z| zR_hMYmp=twsE}4J@GsT_t$g(~wRVaByi&l-pnthJUDL)w-gKBYb{x$WNVxlz&T^57 zTsl8pVbAl^QBp)_`p|XK>fl+a%8$uQnHAGF2H&T5Nav?NRyCe3PzMHx2h{RRkH-yEL$5xswX{Yfu}|ZX=y#Y0qb>+duOuK_BBIL68A2JiNx&Dq%LUIE#Q9P{7eRf4TyT%@QNdCNV#-% za#>ZMRsLu5^er;pHa?o?p@XlXNFEV}c;Prfb@In*yqFf72wwWSdMtoxoKmngjH1tN z{>=t&R1G*r>S^7OgOz?5kC+L;`mK3vuY!iRE%2dgV(SjBlCs z*m6M!sM%|`a3WaRGC`Fg_I&ZHs`SpxQ>>l*RK3Kpo4OD@9pr7sUta6va4-Mk;Md5! z9Afbz|ApX7i9bibwAdcs>}GQ*EKzXf_A#z^sfVa%Dn#h?)4K4ixe~Gz+Vud5=t&aH zYULueGS#;-hmfcc@;TWhve#HA5D9XX($uIr{HwL;2z^+9aP#yM#o9nr_>ZclRNtO` zGzVt! z2M6y~xOaEx-JZd_B1nnmT&H*22Jgz;J55vmM)nSxa_<^>7j&i8U-leWV16AH$A$*7 zMv^C70BA$031sM>yoR-O$$Ag-O)F3RGd^lRa&`s3IroNb5NZm~5G!sQXUW5(#0$8d zmhGao@#IQdouLfOeFbrm3Fl5(%$kXa!ZlilMNwHy@Z)Z=rRgef%1rrAy!^|`$o2D( z?~xl?D+}i}CJ+>`vn%hr_xDw(WPw_#`|Q5L1qg8VvPdtJ18bDQHnNig^|OT6 zbXH&a%rz9s2AY*c|B?f1BTu@HpPn4ZgdhpyjW-N5|lyFu{o(R1-`II&MKBS^VU)l3GjramRlzfeO&L+Q7VyqCpvrn zVV}*@e+2i@P8pMpZ7$SA=lr-)H1hE3=(!6Q6U38R-kMx9{tf0Dhy#P?ityxpG+7TP z7|-W8An7NSULV&e-kLkT=X3$f(jpM;>jq-4{J|P>S`bqf%TDnmL+sM=^J#El?H05 zT8e$Vop4oy@G#JoCRN4UR*X^uMv8=yn9adn73;g!hHbIU4z*Y6^YfhVFCzCgr4;MO z4_2WnXB}5L_@12r$`w;?c!^%An<`gEcB7KUiWo-63V)}e2(7AG7GpS?)mM@gMy*=Q z?b?G%PxX)Z4lXKhkEIwVH9FEF9Ww^OE~1TrN^QND9jqBu<^mIuWiKtJ4Kh>oU07r4p9&%vS;@uM}Om$J^hs?zUDTr-qR_Y=!_h}wGp;! zMRdd$hz-=WZ!UsqF(9P6u*z@~fu@tiRweC@rUut*7H~q56#%n^QjsU=Y@d8|V{ zlGW*vXyn;>(?lGyq4~$c3l6DR)y6e~n^Be%>WR*@PlZJj3z}OC<6f6P#*m_meJ2*fziLQ>8|{(P94#yviZBh3o0w^a>MHxk1Dwyo<{TW=)Bq5RW-ht&ot<8 z-c}`F8ytD^gspx&{gC7dML(wdTSqGhz4m%#=AVaor*6OI@Z_n*YXmlO*>Cw;$u*5t z$v3L{yrw+O&lUdJ+iUwnlY13cmH+4b&(hx^eCU`EkX)7Q{k8hApI^#FAvfWgY?EAogAc}2kDpVq=nN2Z$Wy1rE}9C1>Zjj z(&K{nX9ekqAU!)s&kfS^g7ktQZ6uZbAf|v{eLvk3i=e#UgJ=@zsF=p;pER)| zeoLU1ka7jj;o@_M#NNVk!B=&!;WPy-Ic=m)6YDwVwKM@F=G`!!1W8rJr&DEDb8}jC z@@{>Xpp7bf@J9W{thb+y+sUlSb^N%+VFGDUhn?&d{r+!Ph| zu9>R0-N(6)4Y}Il$=fYQP+K&4uX0Mf{q*K|)ub=sz8?$8jP2_6`<+u-Mm=FeRl}o( z>I^-GO8$@jW@p)0y^bZGu{!I~{IU8xJ608p)qFQrHy-K6>Kf8;tiG9+9jh}}WXEcs zkN=Ox>I{w5a0cY&pC79ea@gCzSls{`*FRQg1#*|21jg%LHzW^iXuR0R4XNY*WV~kd za^tn|aT~8!dDM74rN`WO9e1XU*99zN7_ZN!<&W1l+3|`pUR;+X-b&ZF@%n%?9IvW} z*ZkjbQJqF+3?t)AD(mGU6yXq08yK1Ees*NeH2sh`w2aH`Zd@GCM%T|rd0hK{GA@Vj z;l`y5IflqPmPd`t&U(y^%X9TMF89m&pV}?)aQ?XbkR6vA#)XRqH7-XS?#3lX8jj0@ z4`%sjm(K;kWZtq5{xA8ck7KpIiz$fTTG@QwtHCUQhk^2;u_)jRg)y~Z@y*Y17Irv zMwUuzHWtCPVHwRVMzpwj{v^wdNt@j7DK}d<*atK}VO!RW zrJo#{5B6||4RJ<8y%prVO&nI2W&Vt~2y(eU&C-lNl~hK?Cgsn|%+fBUPpWswxzx)V zzEycr?lLyjJWbEC875ybvtgX&G-kuTu!ohqBq8+d-Eu3@$5T_jWGa)6@TPhz%AvRQ z^fuD7$A_a{p|RHKVHV!Pza1D%??k$wvD#!Pkv6cHQu-(B$s*E7@9RsKlKPKE)eCnx zUCRk(YT2A*Ps49#zwN&fFjqzvADdN~`+I42r8N~wwVq8FDNAh#l??-b%0r?qL>&fr zb0B=*pTeOjM8UoHtFKGIw4}g$k)FTZiM}Xoajjo{s2b(8g`dvh8iYNA9IB+PYWGgc zg&!H?l)oi?N$D`ZljW2>d-b0c;qctE6o&xPUcZv$y4W=8aO^*S2Y6Of(rS%sh~z9U zBuJf-f&??yA~~GoWjQ-C>NPb{4lPE6BxR{26!WmLeS!d_ED^mV!CqZ7X*uaZBq2Y` z_0Gd-%>%jpm|xJUX(E(Rq_K`Oy78K(@HW)2dW$#$MQ&XYlj`9ww6;xEfnTR#{R^&U zX#JiK0wLA+>ww8O(PK{1O#O?^47YO?MLf0lKkp4@hIyiNo3ZDIlXZzdASQ-d;_pXl z(x2V+N&g5Yg=PL095n?cgGx(mls2gSUQ)Ut_4jm5+-%c}Kb-!XC~X1*F?d4bX-oT& zA=0XoVVsIk{l&6LE*s0M`LW)yY(IU1YH|JrIpxjmVW=XZoQeWo4+cneO%d#XdSX2P z+OYx8=kIJ|gPp*?NF12zo8pZof_gpXhG)~0zsqn~pzAdaujn&;8tK{fDjqaVm-zP` z;zr~SQb~VFyN&O$5#hH#I@EFh>xgW4IdTsCb=pWc<(+@9v=7)U>J}whjD#b>p{gb! z$E{OB98vDb`SdDjs}~y;MJ8YWsGNg7MQsl{0o&sYt1#4bTc%Jh(!Iq#1|tGky`@eY zha6YG&(0MefA#3tMx?(KPGDF%mAGMfRgbx0IsYU#ESp_m!@}y3Vd=_)8@Xrg)@!O=XGCtEW!3)L5fR?OB3CNjiN@m|{ML6R zYzUkdr7BG#nhx|&zcnxFpQZ;*gmr%URBU0X9-m6gC2M=(j&<(c06@rb*lq}h!DPL2 zTa#T>YH&1Jqz@ICQ!AoLZSM@xF&+nx7xuM=ADHM&M`ZWd>r$#ItKnP-cxn^%lh(i- zM2ItIuy9rm=PoDQBHe4{%zLRH>AhE}_E0pNN4j4o0LCoR0@gn|=eqrIs+lr{Or$d$ zYlJmoViK_IZ0W`=(tVP$snVs9-lbG}S6HdQ{Hl2d{a_eX8~KK{{cb^xwhrF#-J5!Lv)} zrgxd*tYd$%bbh)v_&y>?FAUy?>9XMc7}C=xu=A*R@(nIrd<8u`fWYxN{Q%W1y=@TB zRdKX)@nVOnScT%no}zfMcw1N`-~DXIC2KA*H64@9tP+iiHnQKU{%QQ)4)J5p2NF&Xs0#!q?ab6yoL9HVcq*za8A7ZU+33Hd;bg9h-x_+EQpq41pf+C=i9`y zJOXMd=TtaHdgezXgLQPgq5Gk!0sApw64!^%IP(10uxueAWLfhtdSE7O%k049u^+Ci2gcud2M{A~ztI})ZjWL| z#%_;NFxBBrWBL^|>>iobIMA4<+hpa$ljqJ!9LL2+@#zD<%=E#yztIr&8>Qltu3UxW zRcUIQ-PG}RriN$S)sEue$gCCZ4{EuU7hh>(I&;rTfPJFzV5i`Ko7eT{gCD*L&KwdF z{-MdtzRmMkEwlE=Q!D+_)obw3%0^4$e9H#o?GlW@+`(yhJXu=mpTj;1zx~T!;J6R2 zkB2QJP3g9SCDA0oLYyxT>+vdO)U>}&RIGnS8Muy}>P7?0f|CxMao9ZQTJvOi9B2I# z*=VF|y4vA;!u-c|6*&Hi=8Q<(_c()|>>p3<*DE|Kg?gu>QZeFFU$k%5C*41Lxz@+- zoDz-phwB-4!Hy>_s25j-$oRXFu|W9kg%`{2ky#KKpE8kr_z5CE30~bj<82;I)z~`v zFK|=i-m&sm8oJx8fQ`bo;@>#L^_sQ(1>kC<^awVE0{&wc5l1@hp_Ou4?qhL>_8Z%u za^EXw?MRigcj8a`<+LmEg!x|-jqlt(QF6z(*B{0ys6o!@?!jylaxpLJ8#Upvpg zlh99I7dvCA)<^raMw`bL7BsTM$x~Jxh_qb9CKy;57D{7I0f+NJjH`{3Ux_G)8Ny~K z?Lt1xioT6KD29kJsWIOfhB7^NjJ;jZJf>imhFj=0CrmkR+E{pas>=cK{*;Z=h{A@; zP}%Etm&(4}|2?Zz-OKwjQ2V^b7|4UWu(49mwqJW zAFp3ZwE0UFxOm;HdjwgJ>X3E1vZ~bk{*~oA?*0u75L&Bc|KTQ;ZD$X6&BI=%FPU zXF3Xn_&%@MH*%jsZhw$vX&Cj>&xh#qZtnB0KmEL?K5y(QX!#H}7VrgC)PzQ>+BL%A z$a9lu=S+`&^g*NC_{a346tHp1FIBkSQk{cRk;E92DjTI6Ni#h*g9UODO2WECs+M<9 z2jVS;c=33%iwgz6z~4$xL@v?h(YlKi_H5W)c_f`ON1^jWa#sOw?v|R|pj{M$>wtI*A@Ne$W!eM<{75zr>N|#ZMHLYn>HOHH$ z+GF!Vuo?I6Y!$Q7+Fr5pd69(0YZ2j!AS$dE3Xc;nKdY4E$VXF>rKW%!8p1^Cq5$Gw zbTWWACg>jS7F5HAH_6-xy(V35DuDPF&uu(2eJucVpL~X}{ZD`fg?|{NDt(=%!R-Rf z!IRwG(XYUN^O$dq70N2n?ud7e#ZY-ox~=-t!~)&2-SQStES68RT>nnMX(F*VmBhW# zOh9LG3mYbEZ(nMRTsU67<>My_16W8TS^FCru&ZImK6)arbvW9Lur!(Z0iWmZGCk%a zP*inU`WR@%MNmP01O z)R9D?YQo|y0I41)Yi3Iw_)af2pL;Z@Bg*>%dDBavU8+P*u!W*=ig~x#sL^+MQAQHI z)J5o`!Ud_)Ft%?+|DBs9UNGE+pe&+fX`i4~=r-KpyvdaOs2%QwN)a)B*Q3A9Xeed( zb!Ea6alxSOTxo;aa9dhnbNtw`4OUilDQ=-EsiqR(OP1d;^xI5wF5mna`ythdJ2W|< zkB~?c$BShI^{QbNEX_e2CZ?BMmoJ65Zt^#$8!VJcw(J@@6Akw^bc@f z_*HVy2*wAJbFnA)NX zjh3OSM)s3Mis`iMg~&3EZllt2ywCH&dQMLJImR%{Xbc| z`@Ue0%YCQ{@*--3#qH?(XSN)Y%jI(aX&J?6>VHUKE!tG)ufnnis2c)Mt06WXZwWy8 zdd!0QG?oQ*TL5Ze9;ib<|3PKX3P3&JKqUfD;=LTG+j5|W1fb%1pcXa$C{RZPpiXw6 z>H|Tfe9H?3QI9xqJ8scjEa}3nx z0jL)sb0GH%Ks}?!EUsonv$(n-05u{H)K2}aY+(SZ$${EA0CipmP#1)tY|FF6-!*{u zn-CsHN_N`D;4Q;|&S3jx0?_=7+X7{NL8LJqcHZb6Y2F)4v1#j~8S=nQ(34_bgacc+ z{5@{@CWlDNaRN%$yGl9`k<~D<#NU0nA*MM%%(i;W4$^jq zWC!Wxz1?`Xk%r^_&IF^vqXJOxeqo?43_!i8$1JFq2WLSw2cSmhfjakV1NEl>)C33W zkO0)U4xpNIp!x-%%JM+X``X+Hx(A@D9H<=wQ2TcP)h`F?^SxXLKPC-37;8|R{l4j) z+}`DP_?hsy+8&qt?Rw01@bju{2k!_#-INDvM=q77?1BK)!w%GE0jMb*K;4l8bxZ*2 z@H|i#upR)^GXbbm9jI3VP(wR_Iwl9ITL7v{9;j%EwQ6DjYC8vNN&srJ4xqZ_K+W0H z;p!<;lakZMDdb(>^8z@p5~Y{N%L6#i=`o9_d4sZux;TJyp22YfcekUg>|p_@%N?k? z0MtbtKwX>z_4@$S?n;^bdRe>1R~xWC0kCoh>`wu(-*o`CR}R<*zjOV2o7CV`+-G>| z9>7`IZeULd;QU9A+5UYrFx$Ty131?h9M`|0``9epHHDU98@j%&XPmB-r-$Dq)*d%c zDpf~4eB6u%r>m4iS|+iY^(Ktu854fwZ%H1L6ZAQKDLy-5N-grYJ*2=JI zpts*U*!1>YeyeVB%ivQ7%G$)NR}@2a??PVm@brgQnZ3$PrkoXAAi3n9^H*#D50E3% zW8Z^;ViL?7Z!59X#Y^qbXf!!P&;GT&zs$_;M}N3_uw0tZRZfti>PWwkijwO4AX*-# zE&pR16f}$F{-+kUWQl(y<)u2veKm`r#%(Dy{^d8-kM7n1mK@cCCt|a?;~;nT_`2~< zgZ(?H&JxcK()Qr}6~Xhwr2GEZI_bN?bC|9Up7*)Z<)0kD8N#znQ!fPX4-WF}?YVqI zN%j4_;Q7hm`w~*U-!v%qeegX@w*~QPcbFa=JcsG=!E=~42hU-8TkssFHw5*}4&I09 z2f=fgZj^HMhUrehvr?@H?g`p|(G@P=*}x^ zm;e92{gy||#aRDe?YHb?P5LGKEnQ*7_3XF2-%ovV_u+n!GW%`=Q*)ia`PIL6zs2hd zsQ>ML%ZXKCGuPd3*^icH_gj`D>g)XHX>itlV*97wiG=2RJ#v}s>6{MGU6Rq>uPfv; zQ4C|9-cW$gi>eihv~ed?%PysajJ9ymNng>TZM!PejEo z>cDn0v|mOiwZ3P?_|%VYYMWRcYi=!KhhmecH?|_0sAriN@kSug#WHoLNN7gXQcZr(fZ+2fH7mX>3VB<7Rp1^Sn)>sbh-T4k^)b$Ohsn z5u%!HzD!&C_f#^WNgD!r`5nL8#n*JJpWHR~ds+t9U_9#l0<}&1S>AwR`cX)h##uZyH5h-%^n5Rqw4%GF(=Y@{wLy@+em%$Gw^ZrHWDpcla8BdR9V`8%jk zYTz(lR*)yk^}A7EMrFJz;2S2}+VKPH^Y(*W`)4qRXMXabg4K`eqJq5o05zRCzx^>y zTXPZrN*}XyLAs4JXHE;^%Ndv0eimyW5Ew}`n7(6MOzLzO9Yu>IcDJ`YczUTtmEOL? zjBk@e+#5wOPfAbVP0Tw90wLP1&Ui(=5j+qCR9gnnAlW{0PLtK%S*3~q8cA$NKK5sL zSk+KTn1*8iB>Ienv)a3`h?C$IW<2))iW;HX*s>T^;=fm-Cl0XAss}PoJ9|yk-dW>p zAQi<@V=DYVlb;_Y{v%|yT#QSd-|8|(5_eFpn)8DR+DJ8o2rpI*snNtg3`@WVb^hNd zp)J<~iM5*ElctIP=nI9Y=EEISh-&_vD?~NNS~VXv+`&C=>5V7^T4>i;&i_+@A0_^F zVF8RHb$&0a<=uvwSaL~v28O`UF3VWhDMN>a1}iP+*mriyb_?Ii^q!(#VMil#qAEl; znP}!ktTNf!_X}&EoGVW4_o+DCznO(?wbNs9*(kjesr=Fk|A*hG(L~TVHCGHx^n#!m z!3}pPO}~SAGtQV4Wov7*V)S+ITrp}JkY8I-`YM@Y*$mCoq$e0L4uR~l8y~Q$Hf3Nd z1W-trX```q3G*I%D>4n}71{u1>D0CU8v*deoL;v46;pAyGytfHV2;M3kAvm#V!>g~t+qH*(C< zM^ohYbKf}{EBMJD+b74pQ$&j;{#t0!G8zePZW%vn=tIz#^r2MEe#2=%gzR1qGa7aO zr?+bSsNu1m29b5q=>y|{5!cx4{q@BmNh7?S_r=@)-WUEGgs@o$Q<@oz%k1t^Q@h6%viUy}KB zS!nEzTW(@ySdiEyCUujXFXa1R_q`^N=twFZ@NXoA=agYLmtiCsxJK`kS~Ap1>79x%v7oAi zHWVb$(!GJ2=^8-BP!xg-if1!bKr`DA0pmZ~i9q=0maKa_az&$J3#iI`N7~;rGT7+~ zp*=ry!{)>Z+uPPI$^&i}+%>r%7mqjGPh%$P&~6+9@Yh;^rky%);-_!2ZFWdejA^Qp z@B=aY7Odq82!J#;4Vrxz=A&4Ope8X?>7WDh59U>o4!(AM>r3Ex!^><-4E;N{=kQ)w zrL<=cWI>hPrXZ5|QJtRw;p*SmX!+vEWt(cUQ}eg)M;A?JQ$N3qY{g#(gxhptNdA7P zOl)+|Wk^{be9>@rta(9E`(94}z{?8!uCIB`^HF@~T8aM7&x`)IO)OBjg#`nxyZA;W z88+A7#>WYcAU`R?e^9h>ek{f2AUm^!RXTyKdS;vo#agIG`3qn_s^y4;jbO11*&fkCRN@k=3m?&asEJO!uU>B};cxBGyUE~R{ z%>Vlv7;`&R-ax?TrF%N^=qjPK3;A%iMZxWuvGFa#{*`w~lrap`l|z%$$$Rj_g4F-U z?fiNhBPTQXJgkg1&*^2hIRZj=sI~FRI{z>G3c%d-sq~Muoyl!(x||1T(q%FJFczrO zd+EE@MermD&lI)$qWe_WF45*@Q@ z88@vSODV1;+Mt5qnoN@qg^?!)N87l^T^?F*m<~74?GmR_O_aoyowz#-2G;VMURTv3 z9EgKKc%5W56Yo(Yd56fg2^Gg=R>#jp67!5bhVuQoCMy=4MmcI!G&vNR?9MPG2aW`= z7KjCHR=%LEMH8ylf6X?tHm~UvNxZ_VnV_9%aScIQrtrwkJ(@(v-(a~?&3cF?Q;U^& z*IA-w*!4%r8-mv^lbe0L>y}!P8(r@uU>>FIRmf(B1 zaSq35Nt~NNqp8Ov&;3)2h&21eb!uf=?kw5)EBHQ$@19dRQjg6b7pPH^Y?;R2I{%{Px8>}d|H&*OL-E(cZ@ z;dTUTK+^jwyPFS_Us~LfqBW?9-4(Yaa#?r2Dt5_h$+w!HQ|yv{7Q5uj^OvVr(MgA& z^^{XMlAkHJAD_JI)I4H}SOdKwWozjvaYbB{?z9|DK@O$49N=O#CW#!efamNoOlz=y z(r-&_iFO0?u4Gk7$EDW5;5EYESq$8|bfs}GvN-OIbSo>}9mev+Eys4}ffW#6FcRqc zH%T=sPN?y=Y+_i*dI?n;A=0QEE&DV6jW7(0$oLBPnG0ZHYD36^SGsp-I)rAJXSjH@;@f z+oLO@sSzdqNR?!Vq+eLhCV)d+xrRXm%S-n(RzuEJ(-;PKjR-lv$W7!}{s1-@ut@fK~WorEfF}q`tWNM?bm1ti)e1 zZ!J3da{ubEpo!&MDQMQ5T6K&LCMc|_s<=a~S5jELJTfe7OWKozT5XM2oQ?!V!zC!a zq``*=MYruxboczCMd@duP}N$}q1J7JqVH{)#SeXL-`@4rYAs2RpePcrU#NWIN8Eva zu^~)sHCp?JJTXPh3_6Nfy!=0r#86C{G4E#Q45l{RYz-o`eQ}RX2H7;B}^UHStD$wRYKqb!K@n7K)f@P%D-aqQ~$-7PFJhm<(iF-sCsl7iaCcwfI&5LLUgFI@$ z#*r%y<9GCwK`lf)@H0zoTp2#bE7Ed5y-L5uA4ew|R(mfkAT6~ae9thJ{}5>zq4%1(6W}Zk)YP$s5eQ?6$Tv!gh8)Kz5yWy457R9)EF8D zW=GY()NP`4KXth5ud~Z|<&Sf}i}kzF{T8$NXSm;*_5I=Qw^-AEWK%0URlo1%w=j8& z-g)cTO?=(|9NJAb-h=pAP5c_lh zaZm>k1h&T7#?s}pNMbcjT(9Be<=cPOaI#_>agQ^c9BYP?IcaqxG@M}D59DK+W82vj z$XQk&;Ys(9?HEtCbJbhwX(>p9ZM~yIw=L{>AC#5?BY(gbRN1~91b^#iyWoBCzFrGhu6CNG+ zm16}x0k?;&IC@rzFb%-LRL*t-J|`ld8--&2s~BdC1lb6w>+MF>eQUmELXGOQM9XN|b1( zPVYvJ&~9Yq$o+8pB7>_8dCQjxd3Qi?D;=_PkLvP|8~y@wI{jUmBbGU8qG`Q#-(q$m z#B7t_u7|f;r(Hv=U9skQC9Z9C{#9fFeB*u9-tH1&>{4A=bUbJuj@WYdTRua*v97w-rKwLeI$%<5KbLHHV6gOy%%DHkLBg z!hhTSZc?6??Dyzef1dkM;Xmzul=w%gI0JqkFLIOqQ{BZj&)RA-)gqZ8HbIYZziooM zfjp5XY{82_-A2pr7G-pcWB*vy;Eh2eX;B-(^7h-x52ZZ&WbAnkXAXX^VNcc(nu0S6 z2mEkRcXX6uzpr?U#_g@trt0?^-&1~J07`7kPBn!9YZ?7j6@-Ad9GnFI>&B1p>s42%hxv=2(NA{Sd69z;z#v+#)EdTrleztyIueMWhUIs zzNm|O_;c*h{zS-Go!akFC!)*0<}xogKQNLyrX5Ky$HxgaKZ8&)1miWr!b8{yR( zS^g#i;dTk!Ai3X$B&67 ziRdB62lsIP%znwRj{7D1|C?c4?|wJNy3eHgHxKH9#1JfQ`41bUAV{5%qFSFn%r9@+R^E5Z=asz-YU_4&-cW@ zrRX$!`?M0@S6JMZh5W>*)3S&(*{mn8WhsBL2)S2!V3x^0^(n*Z82$+GwH6pgPWb{3`Pw4Sx0Vrv|@D{9A)x#s2lduOk1-;8&sF zq+js!C`&Qf&cDj`qtcCQ9UQT&X`->(BN1=9TcgPT-GG zA6h)c0cDCE-A}@|coPtXdO*+-%wyiLVbMK@!AUx%wxFUGQUj+IHIYQ?#^jr^h|2lr zb_0eN@=A-nZYAP#ouT8jFMlX(TLRG73o0E=4y&~~INpcg)(ZtQk*)GaRZme^GA?_M zVo~YHaNFRh?4gxWte+S|w@3NS8bdAW^3GAOl9MsJN0Wn#B2OFKt|e^({()I%mRl9xNl!Dwms)|=)YCoCn8fONA4>gbP#0REV{Mqd^* z6~Ybr7xnJb&q(h9rIGG4?4wk~Q)%vOeW>#*chpdF{bEHC!jbKOjgXDJSt9rkyd`mQ zXbF2aC5l;>6f@o2eXs2n_&3krrLTlM`nen0WG1mNk_n(LE~(H!!%D8==7*EPhFJ^phd# zZw+OL9+b)MXNO3pa2CvNj1&Y)FLWin#Ot3K2hO<#M%}ik<0GW zTM@^;LCt{YuCx4cdObHzujh@^!RZ>c=a6;En61n<`YcbG3YyNW5qV<5A|BH>(s}K_1oWK#3vG1I zKTFR~34d@ag23kJ$};G=gxOc~%o+V6-OmBKreO{;H-Lh0|2)M1OqEBFg^N_o=iEeg zlH8M-_|vBfYb~u86RuzdFYEj#?e9nYU9oQK-wpD1uk+^#0L8l1`A=DaCj0x>;5E?h zuoq|OMe4w{D{p~ruB)|&<@Rv79@zQbTly0|wHG5*g)10o6n(AeWu`Qo0kiw~uXk1s z*-2j1(K+R98WqjRF>HZ9XSh6M(TQH!M7d6I>QIRtdzZss(`c7-i9b;}S)I0X@FnA) zyozRsO8jGjU%mY5;8%%1F!)vMm)kG=n-=wV`eZL&W3HW}MS!_ebc1%XB}|I=_<)G)abfXTyhD=5gVgS5ey$uh-)L*NNC- zS6Z!Rr^>X|`E`0X1;n=F7qOuIk9Gb<8w;=aC2;#i12vmJg1-x`Z~wf`f10nfEG9>A z8?Z>ri4s2=kpr!yne&%K5^F($)EdG&_z6zEwWo2dl+^}Wcv9`ZOah!basHW`{tpjz z{_vkdpXB~k|y ze{Px#+|yk}=|ThWL+0-V=|2QxcfSFe?}R3}xGs1e6;v>tv>mxPKYca7@PAfe`32`> zsrbKXb`i)gXxJouK6pSLr@CG|FZANrB>nKo9jsgCKbc&U7viTMR<&({5I@X6eO?Oz zjtC|4;*dDu0EpL%0TdFTM%#zzCY&Y-MIz^!J$Ng!!}ltT ziIKQV9Ma>^LLDn3Y!v&H3RD2@*~g&BqqkEe)%E_lZvs!J{GzFPj2>N|d z_5xIdhEK2R{O#_Lk$hjXh6xdd15P3gFjv*dmh}RPzh0@Bnuq%P3XzUMvx2%}x$Z9q3V_Biwb+55L%JO!Ikg?A?GYi%Gw|&0S z1lvSXAvPCKqoFL*z)tbqF#Nf5rpua5Z}7g^V~86yEVxBWgVXZX(d z7d;o(NY(&~eh}(E@xWZ<-O|MPW3^-h=lzBQvi9Vh{Wy|XBQfP{$e>eOfg{%bUHKc| z#w@AwGD~jFV#qWFGjP}Whma@gjj4^M2!iXjB=}*)_i~>q_*4^qs&SvX^C>R>JUhGd zek8FKl{5cGJ?EX9YF?rG0<)m~Hjr|!# zJnP-3KGJ=nXFu7eKEQn{;Zs@AX?|Gacd$=cGvx(_>~V8#J!cud|IY9oY0;gULQQBc z?#J-kM%vrt0NaJ26nld66~~cLgEfm!v!<#;qWoz*$)+Abqs2xg@g2& z_8Z+gLX(iaY)6`=w~!ol_I0w;HwtDFNp_>ONKQ3C?q$0?wQ0 zK9%qZ`gETH`n<$G9n2Sp=-X7g5It@|jz*6mlkFU@gh<%iHSdp0v+{@?Ec|Na#L^IJUxaIl55F!wuqPt_hifIpaGp4b{f#R z;WtnvgvjZ+fb@nr())Uq5iEqrHI2b1s8>w>w);dAaTE_eLA}QRe3~^%4b&_+O&MA8 zUM{l)EIgnzpM|G?FBZn8q7LbY^~`plI`+o`M0ze#QmR?Ed8t70?3( zCb$Ci9Sh8M1xl>I$6B9xPwRmKHLgIfjs^av0^ZXFE>BVL(Yj`@gwr!!o~6R-&rL~# zU7l8#XOWVi`}W(qzl+N=#pRinm&blvo*%!q9yjsPf0GzL5JxS;3~M{N6tMPNmW^Y5 zkGMP~$1j`x zl6Y;-=e^nnZq`4RY+dz1bj}wgZ3DOLfBGE1TXSpK_<@@YO14(4T1Kq!IiGHxTvVBS zHMPyPyy)~=W$#6mWFhC~gT@csG}Wc4Q)^}KIhCnOiWMgPRUc%5Y@YNxz*rQ*=>2B% ze1lLie#905gx1Pb1rWL=7gntLBnu*vTmcAGmYkDZW3V?*ZPS!onCh17^jfFqdoSw! zGOcNDqiE5Lv*}~S_zSvob?H`!2HcUgbIt~M6CddPTbUlmtX^DTnBfrgi=W^nThqi;3>h6p3%BfOZySCRE?d12I_7jT2Jjgzx}!n`n@8rMd6dU zclAO^1@E?2k~|$hQQD8seOjZbghgPN`LI4-{sQMi$QYN&Mb}*R$m$==IjHcFM7chD z{bS8zi7>Q``GPrUXkk2cm-)^K+!Puo90F11b%ms#z$^di*_vn`Udr!k@5#j^ocn0( zk=6wm!FVVhUA+98hOs!pL9c$iO%YNXeBGZT7*e}AII^)AY6pe?n8HhQkaI%z*p<>- zk*L)=c{v!G4%$yv^*6VEYuOc~5=+I9lxz(ZB#Zr8L6@!&zG$Yyo~u2M2UkKTTy*2` zcxr}5;+VTJr`}xJ#B+PkY(M4`Uy;>ZBr%8g7T%6fQR;$2949Tp%X4maaobW{6U=AW zT{rQW+`-s){4AZc_{UPdfI8tUjz>t3w5$r6x0>e{nh(=sJfnD0{C$7h`G&L)g*3s5 zb1WlXzS<$IRWBGBrYK3!UMunt*75{>i#3N$Z>h?|pX;Dq$!t~SjR%9lsca_fa6p*F ziyZjpmm}$nZdw+SVM0}(*7QiGmvH?<*BqW5m8M*9xv=5Ys$8m6$Q?r8L4dZ;MVgCv z?|=P_(EOzG#1);Z+Tpso{2Tl*_0qKBP&uo;dgt^$&YGp|LP>< z@QjP&s7kXam4@M}x8dmdxrReE%CB}`Mk1DcPeU_Is%UL>az&Z|S%F-!qf|hSd4F?s ze0eA@R;c0Sb0P^-V8Bxa`&Bj$sw*mTu4#`Lr|49clM``|;M9*>8!f79qv2_dKp|gE zT@tG9auOTyFn~SeYIjDq+PPKH`^K%>>#}laf#|g17^BfI{I#tiK5$=p#=t0M`BtJ$ z18*qU`ZMYCm@0xPR69ddv+_8_00|?wN(^B z0YvY_4hwGCX9>>Xf3ff0+>Js0p=yrfNAK9@v#<8@&cPJgo*aDD0$ z*cSU<1*S=piQf_i+}~$g0=6L#N}>cHfr%mx(MA#lpGq4C9(yU-t5QEF@$RJ)yn2Ig%?Dt z`aEB`sfRmsgNP4D7O=Y}o0CwYfWxh8oBw6r;$E1Q?H!6>imKM_8C(}Cl= zZI-^pCLn`bm7EKt<`36YCZBJfQ+SxS&1Q!ship+WsQm5oKZ~dM747tD`!$&!r+mpu z6T|t=gDl*YVKBV|5BchQo_*C8X5&7qN2o@V!8DwDptOwVcnZd|_io1Tpp;4Z(vjZR z>$%Dx)pzB8q>e84yF3Yjrut^=5u+ti+h6g7c5TwwefQr)A=-$fzQk9IVeJ;}`<`r@ zoPQ^cQlZ2V0Nw+D+B1Vk|6u@x;zsg#jpcDtA({#V0-2li-ic-{!~6x)HE}$)l-C6G zUaW)~7V)XN{OiUO(@U6(L;V`2QkkZHdq)zRh|mM4Ic^R(CEEP#Gzu3q>?qD|;uoG= zQ%U^O+E)@xSwWPyrVP9I2(eEhiOT`32%W{e>>TsJZ#0?qkK1Jf%~bd_BZ`?!;;AVTm#rV1Bmpn_?MxBd%cZ5p!+2M;RI@|(u& z%&)OFm`fM||A9}%kW1t?bp&ty2j0;zWc<5G)7_9J0baN*9fbUATQ%eH(k5EXV9lh0 z0{$h-qaukO-;tQZaezW?Z2VJy7O|`GPceRx zzdNPPEA+x*Dopi#f2pIYhvic?7WVDJPKaoCt zKe2bK{W?QmpvD?OZqfN2I>h$R691pbP^|U-9l#Orn@HvB&@}pR`W}c}0wfoMF8$N+ zh&v*^AwbfGQ*|_RB!3=j;8eP_ziNK|qU~3o-DGTA!N2Ytq{%kaE|w9wXh&lS`GZ zvAgC}S8oKT?(tQ{5|X{wTbsThSK;4|!ccslCk#L60T6Df;%%NVJbi+H{9;E?N0H{+ zjip@~QKMEVg>=;rFs@mkBp~53UipP09JK&U_`i{S(0GbUw|pe{;(z^?8?DbsO+*-M z012Q(g`OEH0R*=Hcvw;}7rYz5)cDVw7Q&noz??*yei7x<@czr&U6WtOuK&(!_tDup z1_X8N_Y2#-HY}-jzY^j??d}|Y@%?`}Tr4F`Zx*&Y;g)C!l`_M{39L?ds4zs0ZxvoR zYNUj_UjOnPj_U`Z4L(>4}XOV@|#e%X18;0vibPNgs#8d~Ye!p)emO+#t;R0nQ2YI~pGm zQ)c1oq-&lqpQ8GqYsqs%(FP=XDcw+?1xxg*N*GlYFHpUC68+nULW#Z)nIzF0&k#p2 zMoawnJ_7Ado}Y~9_1}9V9G=(smEO!G`XGv#Jb%BtYHL8l42Ah;v;)$TFkktYydst- z=lT;>^fSP*-~B82wuXjg`Suj}-}u%M%a8{{F8d?EsmA|_E%(K?jFG0QSINpL;GRM8ZZ>tdxG_^mbK`yH<#8L*<4jRMU;pftq?OW)kFEh z#$m<}V$QbLfO{U2VS@iOKV6p|;aAANBmFz|Ui`b|7VGja{u(fAh4@$XG_Z)`G0VH0 zOIe4G;Sp;Cn8Wedj2P;JzYiS_j2L;mdwlvW^?wDp`1gF$Uqt`;P^+_0MgLit zy9Y3LRKx$r{JSWGC;q+jO*Jpu{x_zE{5$g%*ZxOIe@XlQnh&))3zhaaWnrEkz&w%k zr~Eq-+W0B|){B3ejBKtdog~tSf7z$M*_oZ--_rD@n@i)N4dC}(44m1frX_6Gn*9W|)_V{yzIb-S@{18K75bMjWvpbB|cOoKf{NQkk#>F9huz9=oUjwFqck{%BH!r z@mT-*mqZ;>bFSpqx_mfzGu4J!Bg=<(fzE&r?|mTT!^2h7_%N#v`Teti`|vB2zxDpE zdar|VTOt0_W7!9DbF_!nEd+ir>^ zuI97&x2z-ocA#bW*WfX(wE_RCJ^qI@{=e|A^8Et-r6Z1iC$I{4Mve9I?lMo62s*<1K-<5a%7yex&6n9*5d)iwwB))^t<6LbM?89*NzM4s_OPn zq@grmEpegTYnpE^7;Ksle3vqF*2|QkJpaMGr30b-%hKD;+XK~8_^wGzVbGd29)n(K}Ub5+6pGos1=Z2fxaY~c=jnG~>17-`-d2*-@!j8($no6` zR&U662S{~b^Y+{R0hs!`dAk>HJI>pms6!ytxXipg3ebl0w&LORynTIHdfv|PU*O#p zxQMs(d>!ZiN#UBWT5c)91b=pz^fLd>@NJTRTlhA`pQg8D;S2YXJXun9zj8UHu#}l^ z>Srw)&9@@us89T{qz;C3W%Q~44IL=@rE8JXe;K{W%%eV5ujKGgmpYvPi06Mo2mYAO zsd==_CSN#@=Bi7Q8~ppjGVkOWTmWGHyY+u=4FjK7voh@-o2lcZu#O@BVY|Ob=d}5s zfd=#|FCF8rTN2KHp259e=gZcuFvQ7bGv@Nr?5i({8B#=<@mkc zp3=|V`gv17vz6H1p4DfEerD?DA^l9%&xQP`e(T$L<%E39PnH;iCmH+WsNAgIyuQfj z94h>7n_^o*25pKhsK?LcoWh3S4ZCOdhyut}QhSog<1cwm8(SRDtYMz5`US|C_E!rtmO9cn+OMQ@@CGZ?O z$KmuePDz9w*ER;qTgsYA?LT_2sxEB#o~vcBY8gr`1H)R@hP6E8#7;Io=8{c+)Uedp z1MgN722^fOMap6csoD~0=joqDCE9_UggFiVTQQW9#l{mv)q9_#Yvh_||Bo+-TvNsp z|A=w+IQ>-+nAey`mqs zeX-X7KCeZ}$3py%n5X_&5cg#!`<_VaBjsr=of{CgaIzyK;AsO%7N(JMXNM9^#D@ym z68}@B(0*}6suye3P%mMT$-cdxpe}EwQ3&XE@P|+o{#$@vkkrPp!C&cz+)Ns3S7YGlzfOUHbgczlZ{K zG{!&IUc^yi_J(%Xw$!GwrIv$H=`l&;P25jXf5iH**a@Fmu)jqk(_Na@uTE{6(7jx71QgB{&uHb zd$vf%BgvAd-(Y7S3|Dk=__KhXdzV*ex1Y2JyoDLa-6TwvT)fg*0PJyuqiSV4Jii@& zr=OF;@4h@U{7>|*&fC6FyWFzX<%Zn&`e=i{o_gvMC)f$%PU4U=F+WXCVVkvz^W(M7 zG)jj*Prp;b?rFG=zS0Jr6COjAp2P9AgP83`0i)$zkrvyVo0~`5`E&j_k_f}ruN$dP zDLLHXigMY!MVeR6HOF9AqMUV)M{kBxfEOdJ9*_;5f}>^Zi?(52@eZMKp6agp#@;#R zT~fqN+DCw6JKkWx@QX&o?-20+g;R*{xdafJuW<@dO@dhB;WqltL=hfEJ#z{XOU!TP zt1iAR61nAUn#LQA{CT>OSMY+h_!Fdwc@HXo*wi~TDwH)_kDWY43{tcIGSe!3&LtEM zQ)xR7D%HbPdP1gBUAt$HIn!0TNlCp2l|Nmn&pVZ_)nlhi4=$j-@JtY{@I^eL?rQX7 zdF8km_+cIl@UQCXp8mt_C@vU_wkMSW1|Tm6KjNKphHq-nbsC=WDaPW*?)SU+MesUt za!^A?I}bmYmXK~yu*qb(W_@xFycgEg!NDYw{ejg`n^S}AG#dQFtcJL_pNne}MuT{1@L)%j4~tas~NxxwBoliVPjjVuz#l zns3x#I^|~OUmZu?@S--9iB<((4`ym{2xoj1W6QkoH{8zT;-DN!@auUlM|XZSfg}^) zsT%#4Ec)6hbxFf8|0jJpVI5x@jd^}Ia#KZ<_d5hThZzliKRq-0Nsc@3@Lk3?bf@J{ zI(Pn!L+#@_``FWc{Lnt~_2G2}(+4{0^SnKm1s8|UOTy=<@OiL3mjwR|zxNBDUmmLb z%YqWVbyd&u@cFdGcLq1{%r-aoz9#R>g;BvDKg?XEvzh@34}GTTRY#Uy;haMN#T|=( z@6W;D#K=K09G{9w)peehrdOvyua=4Ml*Y<2Hk>5llyjSaADQ-=6iGTR_m|fly&8}% z%4K6UF~EFI41idjHYk?=Rh@U`P#IohsJ;y{#fhcR{RlU?b@M;65_9_KGa2eA`i9e! zm{o{hi}S+@i{;Uv*y~V|7k7yqvXSC9fw3Jby=A+F^4;QDkL~>z*5uyjw z8Y4ujc{7^uBrjP?nY;l#<@W&zE+iARrP-WwGdVv_=X}BC9OP(7@`f?wjFPj8oK@MJ zFJ*ET09&i`U6(VO&N+#kmyvUVc|Xgv_WexG{nI(OxSW@TIgR3`>8~Tap+J$XU+r~$ zJ}O;GFJ8J)-Q=@RXf7uneniOSIpXVtg=|5E4G-#PK$h%;5o{S^Dz5L59hI&g04;%6mtbxu8>x{)10nY zquZ@-K~w9JqhR0Xl$u|dszm1ePI4JH`-g?SN^M{kIyvJR6Rj#}- znetj)c~yacBEDtijS0RmE1q=wewI_7^Y=5}l{dkahYS+7?|4_?S;Qemp zP0T6pS>ua+t-KxhVy4tJ%d(7dt(9c$bZqfu0K(o>-r@V(JA7B~@LjRP_aWWBS_&8Y-7_&xaz<>m(eC_do(w>>JpKX8Bh zxXC{DaUYZIV|VxQef!wv@{F^OjqYQ#eZ1#BPO*>I+((6dJf{zoFuNb%82<&~bA0$5 z6h4QB&oX;1310bzOFt?Mch4ok&-s?%@N0|j41OA>|3jF5PrkGGT76!f_i1fKs`}h| zucP{$dhf;b`NS8^LYl=-WX`#JS=sb`M7Yz7Tb4)6Yh!iukNaV1dkC11 z={Gu$de7Gi*+VzLGVS=)n^Uzd+<_7$f#@E7LDG&oTEDL_q_kbpKkxNw{d?7075}iH zI{sk~49vY(sy45w_EuC4NL8=simxkDj~IsR^Gd|lz@HwG-PZVz^= zZr|D~{%)`8_`5||n(rQeFTb&`Q1Bb)SNv!$>Zv?h)p&rN6YQud*lP&aeiRgQ4q!Jt z;}mq%^e_HV6$uVfNcBRUCRD0pC)Rj3k}RF1w&ibOl4sLNjw8thIY~I=$?+~q$rN(@ z#BOD@EO+q>(j}ZY(QC{ppf;W0>2!jj6TMis1noH;soc3E=J_MWSqF3-^rDz^rrUA! z>+I6&lKp#CFJ#{5`Tu)8im62Zvh?`2=rm1TVuVh;WD^+x{#~;~H*%*_Oqq;MFPUh)aeqUY)bql+kZ;kkXd=oD?}&_<3*+ zzOy+-80RptIUCgA|6;;^$dG@UEu2k0J4Fu*H8(I8o|XQ@ld+lo_^mij;XnZWf}nRH*ecEOZh zXpoBgh=!uZsh0l6lwO3EhM`gKv9bZYOTRQoPvWY=xJWsN9291b-Ht0$DxX8D!y>J( z>SNIdtUnRxzkERy|052)d!6W8=0V}TMQX}mnxd^jHNo7a{4edmMRJkfj&>bSb;xwH z394AB4MsJo&YP})=$0p4tLTR^>lCG;qh6mCQFm=zQ~$0DotzQ|G@e=Ky<~-7Vukml z5!v#+n|D=iG+D4B8Xub1^lZ#Sa$y0$G5)z}eF!~}3QxeCqMiVnF*HJd0w%ixVd#V~ zRKOf(&1@;VOiyw&hjCJjO;VTv^UXumnGl$d59A{a%-=nq3(P6cW`H><1g4r`z??-p ztiJ<6R|95km4l?(gRod7Fj{-)fwBpVipvFKN!NK{f75Urm4+i09+d0gXhnO~I~AFM z;G!G|E^y}tf!`w`{N{#`nnCUQ6a416EoERdI}8>4?zZ+}{;Vf)?O~jhnolUq;P>Pq z>O}~@y9LZN{H|P{3%{YyWZ-ApuOOqE&>eoqPEW&cl7pYx1N;H{Y@#75%2H%x zhton~7Cr0Oh7N6}{TE5ElZb5$2Sf-VyumL&yeIE#{Hy1S#1b&~sJ^V8So>=)Wek<+ z2pS8JEqG-az*3+td0i>@UVqXDN;tNxiUXA~?_IG88ZsN6sRcDlQWoKRzEXlD{8(8Y zYxSxYR0Q#ou2gmN)Y802OTe;4t9w>Sc~8@P&LDxX6pt>q2&x~2IE}NY@Dp!tPE~Bs`K7$w38b$>l^9qklh#ovGm+hvu$I;>G{sQXHV)=;!lz}? zzH*GGV^CNqjnD?LQcZS%dFIvvI$D`y)fsLFtSa7^VELc_$~AF-aGE9#QsDbx;LtEIs=${CWY*3lOqwe8 zFdi~e$jdCF&AdETHic-$(X_^tZ&eRud26^xW) zCr3W83Kszd{@CwqOW}bj(()<@$&4}C8G#|Tp)d~t>M`q1t)`glBMk-2hU>|)K>$|M zA&nAfDVUivUn(>hfvw|6XjE53LVuBRpdhs2DO#iFCOunLX?BaRki~X}N_kPVo&{Y@T1*q8)oz(1Ndm@n7bDl{# zDnaShY;{ZIQt6^z0UjkgAedY!{d;Drke2XjZ<&81o7hAs=?dEBUtyct^K5(4zsN=U z=ka7R6bn@Ti%@FBkL~&C4U?DJk_Dff>5s_d2SRWZ8a+f}LBqGx+&A%Zc z&Hfut(Eg7j=XE|w+N}5$k`AqvJ0z^~onG$D(-!rWMA)*j4SM>C6 z4{LZNTLayJ%BUb}5VcVTJ@rQHNWA4ei+)tf^0gH9Qg4 zpr`(9rp?}=24$9*tW232=(y7PFTI~b&?P$(Nb8;>EwD36t!-~U2^AR*{_=&>6RKyL zyZTeHZvH~32-Ejqp20}dMKvpp(u+jtFWZdl0E$cn7cd9Xa&xlq=U3s!iK#fioaE8~ z5kW4=>w}xtXRYKK%1svjAWSEH094FT9VoEUot#c5E-oG~=X(7m#kBK*;9@tu)%+Im zmYRzEz8F`esl5qQzb>W@MX^z}wzrX>!O#iNsllJIe_md80$&zE9|Y9I2gDua8$|R^ z6IaNS?lq%ScmkTI>bY;z1o5A+$;v#4%uf$jc(UY>-`tl6)P&;tQ?koc9?8@MSnN!q zN)>L@l<=W0Tv-sMm{L}M;fZZb20n{4V-qzxq{^O+LN+m`QSUgeGQdoLcU)#}(DZ!F z#XxwCzwoVC<jp%ub(AzkuCn!Jj0==$R%VsI*rAiq)GW1hV-VFr8sTD5ax6hz$%W(J`ns- zVrtAIzaPKTVPCO-J#XfjfCK%QGd1`Fq{B)c-~OOfYa9LR@nI*wOf;AdZ1g?L`D8em zXS!xdIkn<2^3gLkD&N6%pY&O@et#t=+q8R}jp>i=xyA26coyf@^=1TUn=SAh?$MIp z@tI9CiRqxBL@C31(=G#b?d^#d58z{~K``Bw0fh1POp~xQ({$@G7Ka;T#4ZAG99~cB z$m8bDOq(-1z;e;#G%S-WbAcJ_)jD!9pI1mU5}wy^s6=w-vI7jd{*5i~VapWV;NL?7>%7j|iVb+VYrH?zv~_u7 zbu@WBO>2TMq1K`RHW9Mi}x#?i6*W8WwgB%4_e)jCmnhghi-xFy5g&@=e?-=1tamPw*sQ zp1I@@7)ym4iNd!a0#?CbOURnmB}8fLUN~OOi)hdc>Z9ztd2szq4&&!tixB-0Uq!(T zEV6=XAa1Y%6vq3|@LdBkgFI%;Y0r6OYwbx_?Fktpt(L*XJRAIf*w~OuC9(U!9MSZg z0?}&lr;DZYHL2R`6k$#a0|UT-YSdcbI>M}EKe9FWosa_|o`;uW(s&MrwbF^ubImKBkt=>xHV)~iTIQ34979OT;^4f5~ijW&e5 z?gOe6_OAxV{Hc}*1(7}`P1Bs8SJtMFOY}jqs6NgO%NpRv!nY#72a3!LZEpR8@R$iV z#=kXb;Zo_rS)Z`*S*G*lxJYB>M%P!3z?{z>&eSS@^Ur9tZb=;fKU`&TnP8=HxZ;<9 z)?EsVRz;(eTf)zCOL#I}!Z*5=@Xd4ylAp(^1jb+tPt%obR1*(w=9ln4x9NeCvacGJI&m&4-5QC8Nu|GKQP-O4&GEUQmgRx@qKx6CpjkhHwk!rK1M9sQZaB%c;b83dPm9ObW3VT&w1OopawGSb5L3yEH5G zph7tbws)(u)}@*4(mauq#@;Q>Audg`OEV)UjlEl%uV$(3H%wpzC#6Skl56{WE~<*C zG3lr=E~?W-*|?{p2D_-oEGq49KD_W<4*qj}&fh|!Smoa%t#bVYN(sr4NY%nb6Kld{ zkTB0ygvZkWM|Rl!No%FiP$ zpKsL;%5!z)TkDfijH5p9dF{cmM4wY)UZ0cd*<*~qyET^oaxDKjHtf0I>Uol*0WCPC zrVXiiRDK3=(ssrx2FE8BQ2v@(P(E?Q*)z}*1soM^b)uLomZ_(yd@Z~yHQux({PA4%qt!v0L zo-=5h8}=gqt)$z3Xld4s+n|u@_d5s;uC$QCXH*0j8F$JGEi1X8hMN_FLoGqLnn&hq z)_Mh|uJwxH22GrxFmEGrfY$uveEQIP^MZr^;nqY&3n&d7Eg5;IZR=x2^XDFrrj-9a zX8T%y&^xsVxhu1tE}!CBDtW=}i8T1N@1s@)G71ywi+@Kew>Az;ApWm2QFX{Fl4yN0 zLJLI17m`Hg$crN?76^<}L34!=Tl052@C@)DXIBuWR|OBnT4jw0ShDbu@eZg(X_yH9 zc=+Q62-Cp~E9go!_&tX7gZ!b1L8!ImT|Z4F+Njhv!n8(vTzLXf^7ymvV5E-|H~^c4 zk=FLE_%0$1{AuCWF7YpY#g_O8r~dA@N%~Lu`8+cWyX8z_9nBfcQ22tn->;fW$%h&; z7XEpnx`fQ7wuB=++u8-z=?M`wpQ83*)N|oo0usBCicZWbd(M)*-FT4W8}7Q}_>w^> zIOE$*I2j9*!1PR^wOZ*zN_UOQN*4Y`fX%HP2+UP7j^S|6^ON>bvC{lxJ2#b=HTVy4 zyFKqpx7S9U{(B1e4yc2-9gG>tUD=-LmJhq>MN z1HG)N^}g~Cy_Gro(Pv>BPoj-MoG+|!&dF<=X}t`NMmfgHr2jp6&*T-O4%*D@e1h8a z7xlPj5tvF7DhO%MP}8khTmspdhsKdIi0kH^$V=9 zAl|dg$0CzRA~pb3sS?&z8*X>aj_Z-)Nb3WHSFC)bhhQW_?8wZ;)pON03QH|+@V7D3 z!cr*pAuEpE=2Ba;xUyQb;^V@Fy;+rlvT8Sr-7(Si7K)d27Qin(=dUCGTcN*P~Y z>fm8(EG=-9Su3ANzvEb%0L=#fV&;Dh6J{{gs*a(mKEAXxm;yCS``>EtSMPO?^+zNx z?4DLzYBOEjMZcQ%|N1`Vj(0-4P_9dXZV0pS1lHyNP zBzIeq1C@mH9k*+rj1n7%?3A?W>IQ%F$0Xaanx=6L{%>~~aVqw&yNo!z!Jo3r2v#@4 z_q|y*sh<#^D!Kd@t9R~K(}xZIFO?m;(hL6a5zl48_4Zs6)cw z;NkFNUyEE4d=h@Y6(&yFcV`gaLgI3Mz-oJ{gbVprO@IB|g%8-K@6KTE7y4etosuqo zj(gU5uLTRfbd^6w77RRsS`RzmE?81 z;HoN5M#c^P2^GhPRxGUM;f+MX+WY(^;rfDd5;sYKf|LH5K(paYq}(i<5ZN74=r1uj@!$eYAH4 z$KY-Coj7X%JZ;P%e#TYtGhv*xP$IvHVm>c}q_#;Bp?%&JJc%^^lNeJJOEfnVB0Ea4 z59MpucVY~Dj|D*(_M8jjYSvld0q2=zYhBnFaZoMQly@s(k*CI7rmb8tmWhOk8zyFG z*6<lE zC7PJ=h-F&UQZ~#*%rYO=kvYwRfcT{89o580KQ!rlIWr5ZNK7WuFQ#+RR3RAfa+|Yc zrLHk9!1cBb(;7>D0Y1|EU5aOye)wpQ^lL=V(QjT^vsLGOP(}K&y+BE>1np$cCbq#p z=NK^S=A~*`Yx|V_!MNV{a0DR?Lchx(GsacI!UM}5%saV9(bM{B-}9>c-G+O z5%Fv(1$K|kP7WtGdz-4qdaI(|^XJ4+L(7RmLia^JSxJK~S>5`#DF-(A|GURMSk1p5 z^=c8){HQk`sq_1G(+sz`7bXYNo&k*|)!tjxTi)W5RZ@4o8m_Zp6MRMP3DH*ew$x^olVd6)_SEZ6zIy}JR{ zK~&|Hym+X=v&6x3KJhv5#u||T2gfi^<}9Yh!_2aA;~!sN6;0^!?Xm2%B+vv)W+~O@ zWnv1@o3Zjha8(U~63-(23J@v|8b-en$-HGWj_)e!GC-rl@m70B)^H43@v;+V_(ebH zzuG%U@Lq55?oFnyzhwK*X9(}7xbIXa#AzRGU`3*?fleJd;H#-EXidbH4!u5!3CHBdz{x<&67m`3nP zRD*1+*d=qK1<~X=P_!j=`Ac0p5~l;xA+|`1FCNfYwT(24&!kOWlxxt}Hm#}AUFx`* zV9#yBV4#^ir$2U1fOEW)H%IodD>BIJjquTfHEkd*%T})W_3mVvyx7ae8z>ibS2(%) zDc(%XSzlASm3%#x##%mbYJ?IPklRWh#NYp*#=RT-n<%Y%%iADO*W*?{_^+V(rPl1l zU3X_iV#6Ho)wH$z5+&BWp8clFH zp?5a_;gX|hhnmYlzs^wtA^M(C#xXla71Pb5zo9KdNMnY7Mw92}$G<4L=9tZk3!28` zGV7_xoa>6YzpHrj;vS77y_J!t8jH7cbfw?+XyUq}uDe5g_Ex2y$8uosR&H3S*h&eL z_l>`|wPJI%w<+GiwGMsq&hYx=SFW1OGV;2j;6fCPTmbE@^sXsRG!}28yloayE6rnu zLP-ATB@nT&kQU-Py44v?bp)SIzcX)!=tnnkgZZ8DXJfPW=JjG?vBX3e^3gfy;C8`u z%!&IySyML)GTvvhlb+;G3yYE3f!EgGf)OmzYxCMzYZJ03g)9OQ>H8N0YMl~EuTm z(fHP$ky~~nq@q1DKfr`oVO|@0?^L_!Cxn0o0Kh}@g8o!HI$1*%!iAiAOlJ&IxXP{v zW_6XA4-MI=vIS_7D%fs z^|m}8M$2DDb&h&}E9(9F1Sw)(!5&fEAEDom=06wBe<}VU>ixIW*f~XQY+^*w?Q?(C z^A-}#Z#yHOFckZQ&2eEp+Ro@f7+QV89&lm1wVkmWVW{*8`>6{nXgi~TF!cF^{n&*S zww+N(7|MLYCR$ik&$iQg2H)j9GkylI4;g{j1hG3M_N8(4o2o=fM;Rvy0A3xEbK)W zmIj`MJ?+BMz_YN2U051;7WQ)sbHHoWJNW1z;AbEZn3Cs19B;^G?c?B!uSs__;ea}m z&N=L_+L_!n9m93^Z#(5mH#*p~569mX;e8bRj9T^F$TL+k=2oW~>7M*U6eu50lLF<8 zcXm}=C-E;?{LzYsEBou+Lb>M#_bdvZ19_ek+Rl8&weoTm#CgP=TiORjy|r+p2D^W2 z;Vc>zO&nKNvAkkss^s}rwQg7$jx8W4aIns9*QpO~zB4cWWvU5Vw>!7r4e}u-e0cNf zJk-qI_#&^CywIr(H)!gGR56r=YO3Tr9%mfW^FI_7Z7C36Kc-vjqL{1bLU_o4Y3;^9 zZ)6!Q+%AnKF0dQcWN3+3{5?d-<6mNE`Mg+1{LB1Et5#Fcp`LAUwV1b7_g_^-y$fVx zx224bsdWQzzXatT89^&DWeq^6!M^4^rXZr&}9 z$>zj;MDFpHdp^0P1rx-AEgv~G&VVlW{A>9AAaw*gm%2`zv7Lp1{fc&*VqNb#dERc%^}ec#EWm7Y)!EpYQ5j!f zAw_Au&Se8BwQFs`?j~V(s}R0P*Er&CZoP`CHh1PX)y9%#r$oKiqTWAZmFuoK%*kz$ z1=r+97MzTuK9)B<8hkd8QuMj4HpLoe^1HcxvUYXmGeNZ0Om{WJH`Z&(QN`8g#*)Vk zQQcpKbwe3?iJ=sy1@5}aRgsp*V9t>_Y^J>0HmW$9905b@h_sBg&=L7B4%ojoOzmt}!Lu2p3k=HVPAw5xZO1@h)ulwo&^Lh7>@$K`v~Mwo&^N z*4x4ka$&vOMty^@J{GnJutnyK=+ice)A1wvT3DV-*SBrdfrRa8VVhmLJsFM2e0H?* z+D08zSNTq)Wuyp`cAl!|DcIs9S{3Pw>t5&V7ELXwOD!GIR#0gf?j^hqJB!pZc?m9R zdRw;QZL_TyvdK|U{MV$3=C^Br2v`=0+pZAyU~L6Tm%Q;?h@Nl!6dNGZ6JoP~1GCn4 zazPLC+VW$SjBMjK*aAU3j+G6Fu_`f<^t(S6DI-Mmp{UNQrzy3?!MR|U!@Fp*pxogF z35x5y9~E^NT0v7Yv2mOR5Q}~%WSKsO7??$umqER!lK~2 z)Ht8MunX9%Jkfj%gu8Ba^6_vYWz~edhmakOclMYPku+Qs^d!c@>Hce*IG7oXRY7e2 zs+4I??&Fi{H=NI$3ZZ}iBuEczH(DJ%k z=4<;p;8ag)K9-?*!6d!))!Rh9?Wwm5c!N$RZxEHeW0hmfeb?oZLPsC)-=}%@n!8Az zMcdiYbAr3Zs9HFZPk6F zt^crh{Q62S@^ME2tQpqS;9qjwh4zMKq<8dg!&$ANsyhqHc;r>H^ee?qDAkS0kDelQ z$mQeQKO+o-m%~xh%*&Q^Ehx&^08k`d%mRxfN8_Z>O_y~P^1-EYx9GP$r|DZc+mBL?0}M| z=f^)Uy5@+@iwc^~z#jHxF#g2qHkz<`QIEz&-tuZR01;^bine$CafbI&YB|#XiCXs$ z-b&8N_ap!3towVh?$4`yY4XQit<{MW$*$g1CmJKH`HQv%KcWe#lIhKkYv#y1;+i*I zFKoEwuMQgyegC5r{?Ht~t@Ac#uUdc27W;5YBf0=8N68Vj5&jII*XOU}Dy_-i44&X4 z&F@ZY8+J%Eu^Z}~q7fZ6XswNId=`l(NEAzczXzsCQ*{N`HG5Pyrw)s>{0->G5~vju zz3W-#(DS2WzH``0`nyMVSDW=NX6Xbmy#1ZIvF9L%&x+YtL;wl(lHApTO%vHa>yY zK#DXd=ObMHNV>X3VYo#dFJEy7ZgKa z4-0%J3_M&AoKwsCt>ok>9CAtYj{AcnbJ_vcupUJ8h$b#CazU}=kMfaFdLS-9YWtK$ zTFxPJ2(QSrrF>K``mms859~N<+J=>fgV9mL9!bpmAh<&7AlauaWppz-HAf<$U)Avs z)>OwgR982DRT!CmHgK>33#0-tc}Mfgwig z7d2ka?@t=P&)OZO_&OH0%o`C6IW5?qJVrbk%MkxwZXDz0!luYH@srcqjuJf*6+&~1 zUI~3-$Nq zDK7^F%$?lpQ?e+~HNyC<@uP2L=^10&5t(*IE*l#|(k@Ji346gx9|`GG}br5hL2h zmO`s~i~Iv}z+LXZ-THHzz*Q92oqJ1;7WdwP7G*Tn;rw@+N9I4v-$>ipmPz#IE7~B{ zJti?mfFz7hYv59UGxTgrs^peyLwXjR!dMwTMUj`X6fB=$H-C)I6{^EK0KAKq;U?jQ zaNRBBZ5kay`E;^_px6Is$4t`aHu(FJ37Qy*OfRH);e^;N)A8yBMizHqiLW+gSz-{7 z&$Z9!Y|r3Y0wkFQ&+uaWrCW2Gsv7)XJr5!39{NqF)So7akUdM2c8HY0&+jbqej-yP zKfVs2LH>frMc%p>^e$MCE0wA@72&RTx3lBg|3%3G%`1_XH&T#nD*dWGg# zO*eOT^8E@>qB+_p6X2UwPMuk?-j8xm?I?9iJ<07KEto=ds^t8uov2$JNLZQiOaY;l zE4)`D^{eAu1<_0XR+k)hP^zS6lF*QzMDhP!`SxUF!>()kOJy~nx})GA*B6%{sYPh5 zT8|kjrG-6cVw7u{vaB#`OY;MXh{#dQhy11=Yq2sVco`hdQqrjRMyx!DajMSa;4en_ z(|iDGrQRgS0>{$2$~BYgqw!&7`JGNe^?PgToObYM zSjqoHM3FAoj{1hifr3_~WiAQDDvOYur7)76?PCn1l5#Kl+6>Z!N`P~IsL==4bRfAj z_zSf`L|lKhPg<9_@r12W??%>3^kAAqRog4|w!}M%>JtT=lqeX=R_Vsz6qgmanQnIa z&3sFvc_p6*Ms7Kkt{Pc$QekkCLD~A%(JL>wlGLJ<)one4a|D&d32eAFvHFVbvkgFO z3*Lccq)LjeaU+K~6#8sz((3EzD@3t&2jJa@qUl8<-@jau>pMhfmIUioL1>H~q}OLo zABHk4)J!A1|Nqj5-Oi*#kJL>cHr<7F(}(@og>}=1UFpKQ>BGL~!n*0h&UInk^kEWi zsJELwtlEWj(}(HACh5BA!;W!b-SlCL!O|V|VSBlB-SlC*yD-y-In7sEAvS^%Y=I{Q zSheZHT!O9xwNAp`JFOF2meGmbNm31BQN8yex-#VRrd75R_Be&rSAGzQKST=lF24oQ zAQ?N3k+`%FV#{qZ@fwXX1#x}Z4VocrPUsvDbE9>h8j-6L!?F$vkgF3rtszI;W35wf zgvsf|*x#?si{-C%bW&326dBW(TDARrJJIr54%wK zKaY-t@_H;ECeoEOd8tJ!=^gc9;xa|q`?(tY<%>6^ur0E7SI;wz*fZ7h7^b|gKUU8( zo!D=y=N0Jtf$DjMdfr<-&(vbKSI;xO*say`WPB0aWQ~z>ERbHkdY&oBeqgVrA#<%Y z6`8n1^*qy&MR{fF{opc9y{nhFsW6N<7-J6is!G-v;G^AcTH{ENtS zr_0JS{zZP%e%kH9Ze%J$G`^^#;9$GO1IsDNp^aU(S+Xuw^7|`9Fm_1f#)ouMDx%)& z+HV<%ChvO2$>Zpx7&2B6Y`WCK1g{JW=$cPQ7SWO9@UyK~zW~VD#!{~d2H|pz)%hu@ zl9R5Yat@fF*T0-{u_=bT9+@iH>j$e5qL(5RM6RpmH^1=`k~N(!2k*!)BRgsigUydD z&4+iVmG~295Flo*6iyo_)vM^Ooe^J!{!UjlM7`Zi@L(7IJ$5@&C68Y&+&U7KnG8cl z7I)o~-VbnhikL(+rJEFUnWPv*0^5&cMK(BkW<7nlf}5s6wNi=|LJtxGp$vdSAIB(v zC4x_y;$|!;faFnSO;a;?Pnu%J0UopCj@ql2Xv)>FOx3m;bGXF}LX$EJ+8gVWNK1>T z2x>wCPIl%(T6@%#U6$bik=9c{ZeTMd2R6)CIpAQi%d$0wlLUPBw``@!Q4b?);>O5O zk4enoZc}X3keY5qNrG$LvN8IZ!1@34I1Hm{mobH5J4EMG_oL;hl2r|>^TcIW2~X7g zg)2V7^yIQq33@Gm)JIYmXfPfFQzZjPEkw}BA8aH4ZH+uE=W;Zi1xIo6jhb@Db-NU5 zSa8|6E|q4^w7sd4*Fd6JVqi9>&j3i?fX0dHU0%dsyE0!QM)r?a;~T(_dW+-Rkw#vm zn-!_5%^e6Nf8d?B9+B3c@fKf)B=NpVo74ZK>bCwvgV*^Wf7JA?C!%kSv_3_`*5#2F zaqh^RraZH|T7(QUo&aw-n%BP4yza3*h$#i_lKvGr0fClxF4uDsQIJ-!$#s)C_Z zG#qoi=O^Q|v8Dt)=VGIhcueq?n>ANoG`o9j3w~>|#bZCz*l;T0<&PX3|FiiKoja~( z)I}rDJmWNRhQT&XG(+Z|RyXP*n>d;?hGsQoC)&gjI|R8-W!7}+3`e(F6KRbTZF9xV z8gNpi7>)gh_(rohO-aMSJ0|Qnd!{Qv&g_x?^WV*$wMx@%_GrAq`4iLpVOm5hpJ%U6 zrFWY@)==ZbYG3OFMk2UMI6Py=`4U_t3dM#ywgkv&jTDJEt4aE_SWYZiIPOxhi?ZX@ z;vCznNXuvnR3pZ?UjLGkf)fb~>E9$gJTZ%(H@tq0vMQE4yw>q5hu3%UW_TS_wmJRx zXzL#_w8@h2UE{Yz;%@jY{XsYU{;eVEF7f+kM1388Ut`%Eei!V4kM4pQz6 zdw7wj4ZSN3o;%@pWfs5L2NZrkZ+2um;dc>LrAq!{v}Z~1>|dPi$X|J8dyoC_e}nHe z)nlu7z;{A+!uR71-z6P@%flJm*X>YfE|>VP@crTO|M%eg=NE+d{y*w-H++{!2EI?a zSonV2kbj5ozo(@ChwladCBDCPsiWunJ@4>+2+#jxd>>go5^~jzzK@L7?u73r7`{sc z2H#I%a6i;3+W&L<{^#%h--GYF(^jMJ530}I@Lfv_@I85<@V(}ke~0f2zw`g_9r@+I z#P{_VJAA*c)8YFJp55@B{pgZ@G^MWlL=!6Q$O|WBtSo3&gFpXE%3|q{s6{BI zsGLrJS&a+t@HFdgr(r$xeO2qql-+?DYqzV3CRA0Or%8v?28GBb(98rk{M}`0CNyK= zZSs(1>TQ`Y%FmdQPr@UG<-wJapg198S2@Aev-;dSS>kVOk9LkMrqH~FgZPQ`?~f+Nv3dJyWZL~mfkN%U(ehV?+NH5XX+rk;_rYPMVbjuAKXC1mx}=f=~l!+ z@`J(boz1stS<S8z;rlj#r%Y5s!jb1W?YGoS%4gIf-~_O>E2@ahhqBD`v- z{U{9Bi~ne|Y7++G1oQFd(fnPbaj`X&ppI$@lW2Bb&7**Z>+34L|8FYkYbyPp^|_Tc zBCtc|kF$1wwlBUJqU}B(&@)g<^^=){A`d%YdZha`Y7YcKAu&i@R zLxcbBrK$@Dw-(w`)*Qx$DxyE0r=GgtK*lzXPMIRmc`>k)NNM;DWNHa?Ct}&U5Z;yR9~x-%N0u65n-y#a`qNKE0qrOoxfQoWXTn03&7XCq`8i(7E6XpRJ!1(i0|GaaaaYzM4n@ss2pZ`wpj;Qydh$yZEw}9 zZ6q9qOJi-hW+9dalla9}`2h>@vcZ4z6U}u8ora{n16hB!Q$9&y#f0{9PZgAuo(Nu4 zs*I8iRs8^eJge}S_qndg!C7+lniDuB(qqG9iX(Idz{U-M9(?1qN>AJ%wCjocm*;ru z4{y7x)K%y7#vBGbL9jevDOxw8Yg9y{`$gl!Pd$G%fvt6L=oBOnCfQCZ7#)rA|McY+ zR7@_X^uY&cH<_`!EtD%|-ayK~7 z0y#moT1mp{8OUPNIJ`^y7qbwX7#EEe^35|}74ouC%SmJg<;qIzbhar`B7v6o4HWb1B?KiE2) zmNP%J=|GS4XVbyG(rb>JYbx+?p0Xb%f zSZ7s#+x2!etKk)+Wc`IWi_(5bitSwGC7|iO%iA8!P zW!&B=P_ue8`<^tlb?gAfFyQHvNC<&qy)eSOy=c7zGpeG;*koVI3tcvSALua_FE8bE zw81|~$E>i~XEJJ$Xyu!cmb)nf&rM1)+g@?5Ad~Z`H!GBQjn65W z{x@tA60W^hx}mc6RIb6vt^q9(A|kk?Cov#&MQugJ$4%otp( z1y9crZE(*l`3>zAH#3`b2#Kvxj48HNr0l%mPUhsPlD=nBnrvuni_u#$U+`+&-=w4K z$AO*Rn{tmLBZgE-Rh=vtkn!I{uT()>UOt1}XkGfiQ6%P7F?Xn1B%}n7EiBj$G854q z2+2Jw1o5vSBu|f|Qta>hNMI9jdPD)>*Psl}H60_$(*1bS4k=`_IyowI|7f;UKpio& zy$I&~*%4Mf-8v_|9&jcn&hXxO<>*Fagl;;qb%nfBAYg)B%><3X*O9~@y@P!c+5w0h zS<19-k`w^Cf`)ed^=&s5@e>iM9-9ewGi%w1JM*xg7@XHsZJZ1XzBNb(As(vgOy47| zy)@r;<#Z#xZiCbb8d4^hsHwvM9!1I@(Uph>Zn*)0W8U<^IGffoc8%D=lH7l62s|); zBL)#oHAMKc__nq=RM+%H1%mL&npvj*5Fj)Ap~h0zn=RGMmg5ZTYXl9B$*%G5P_d+$ zK$(Kd1Udy%W|d?w#*rb##*Qmr6FW}%hh(e%C$IphR;Pdc@rVoW*<|?mbjMN-deSk< zJ;UcIJj3x7{)7FAi{u`?sdv{oA)LG}Ty-N!jf5v3DO!hfcHWJE9!D35MLHZgARPs3 zH~SBw7K~BjHAcOakjsGnP)FamX>u5rrgib9Xi)nvQ^!&zeNK}i^Tx6jbugFJb@bSl z3VKCwR8X+{OSmkx|(-g-Ux4L7U+*-lDdnG+9N?sJBUs-fqqn z--soT`pinyteHPn7W-I^q%Y8Isb#7r&w=Bi!s@IoR!VW0^J|6H^~GKLrtxRrsD~Xt z#iJ!Xs^#w?xG-T z69<+s!Adk0#qbfnhq}mO?o}9`D3!p z5QiBzuWzHSoY~Ht)egw0!I6HWN@W(BDEzgEZrCbd(>ba%FR$5-;xHRA$QeXZzzPl& zsol~xeo71zDpj)oNW)Q4*_NNO6vPx~!EohLV6H-$mdhQhv93g(h6!>tsI)qiZIpp9eyA$j@G^z~I^*h3f;tK5V>LMgqG ztofPEP7Caf99jqI7=Ri0*&rI;pO{0gv4Ccrsx?hpX`wG5*{Wc#bl1q|;&Zs#oV>Q?)wEdI0)&!=Sp# zuX3mK#lyh}no|e))l3hk^XUXt5@OrhNVxkR3p zdT$+!t}YQp%za5`DvzkMxg2SENMn$=E?Tj&9)}>1mMx#uRib2;`LIVO2p18yET;0LLa!>eSj#%ql33~FtU(R31#B_@%) z>E^ol$pB_uJ&p2SinNXdgpLl0pbm<9pV~H7%zMdhYtqs_+Tid1Hu^+wmDFTYt=Vw9 zy2No1fy<-OWFIa7cLyZfERv+VDa0F4mqH~Jd_#5Q&>PkjhLut8?e4f5ESB_Q=OwN@ zQ}hH2MCp|-2kJU|ohlhMLijMUD0ode z#0RXZL#QAwKRA~>u>asLdv39QRwsI5yf<=ab)sg^ZECpWoKIxoahZa4WNKN z;Z$j!5d!OZ;e__9Qb2c0D+#fAm?EFMQxf7s8cLSh*F5GVi6(n43(D--8Qj8C>zV4Y z<_#n?Pu5wWF~%HE?Z!axY?CSpV7c1ZP*&9*&Cmrqv^2(8e5&M|)SNTkp+86SH`%>V zjv1v&Vl}SN?O~zF^yxDZItf)37Wx|KWcVlzK)C+rJoo(&Va{ak0c?!18R|>q(_h|m&n8?ib<}WV5x%=)b z%q*A5rWU}3r{*2qG(x>GEglYt+&r9kGZ9BOuA>F-g9Fxrg|4A&iLWbhbxE%5T5S8v z$-Yl3osyUhzX4={rE8i^DfeeeVKUPhN|_y&qD_|IdXsoz#;G3H0bgXdX42)I9a(^L zmwz<)XLE%Vpg*|zhWwLc$9{H?^Q}kpi{8;M=l`>!9cPCb>csp`o(Tg!jhV4YD`e+O zF@`V2!)=j_>WL*{)VmI{i-b}o@guLJ3bSi(D+N|DMmzXRd^Iv>68;jeOZ!WdyM&Cx z#7RY6e{%5CIUpE~{3XV^`%C;J-cfXBqJPmDUjN;lzr;yJ<~$=YN&XVAlkd>!edo$X2%i_~nYtyqc3TNekl zTK-NUoJLMH4##grtbAom`|ihr0$4U^30uAbzmCz!sKv4Jj+p!;tS5trfBE8Q<)=if zl2^TyK460C9Ku6EyG!0$xw1=+`g$FITf^n7#qbcQXsk4!V_ zT(XLGfNJ9B^&;Elo9gr7=s|t?q8KOjTE(kG^bfKAApyh2IM+4Q(Xx3t<=ZR{9wWrH zr-mU_7q1ArSd`nv{$;6>zYkLn`y=fb`UpIYr^d?P1vCuVs}$1Q9vP)*O4B+Q@>>%>hqGjrSr~O*Hu594J{7 zqpi)cB)5!#C#9Snv8Yo^6?N?C$zg+7Ja%B9#4-rUNr=}u(rAlBFdWEol3~0KXnDu6 zPJoqOWqqCcGaY_xrZa?D7wVD8u~t~lVU3<@{pYMUzHR;ZoPNlH069YkRHEK!Y?mT= z!Lh(n@~@Oqw^|-#U{0q0NfnuYl5#>D{J)`;6EQ&iiX~X3YV^~dF-9q^Oe7}qQeU|} z(o(OG&~M2g=^#w3k5y<Wg2k8E z!M2v!Ecc>v6L6zEFyPEe{xc20YMyho<&1k&gX5X!JD?;Xf3XPRy;kF?#a=EDk7osqULkvWWt+5(OBujvt~#E7Yc zP)XpUi@RD)PLhM_Hld3f{`_|NXP9=16~X}PbTN|+7i^souT9f0ls_8$Zzpn(gjqSd zKslN$*vnAL3M#RJm~=u6!(w;qBo|cy;?TiPq;;Je+e7LrJF<=`s5d;gJL_oSE2ao) zB+zp1O*RWa>Go(r-S#N8rTzQJeb}_$Y23B%dN;8@qjFXUelWg1=jq7PU#_saIRRXh`ck zonT2fR(nN7!3yntI{k#4oUx8Oj=ELtbYe{YB7?cNaYffyu`yA!)k@zgl3=VNXaqu%b z{yIY{)CVo_8!?qOE<@4&i*Q?J(`%-;Q zy|>JHtwU4@_00piEo+OA@shE!p6E+@{-mZYR)*78siSqy3O-&~aDrt76-u%p3U57WQqlYY|F8zE?U#qzFcd}kMU*3h?+uGGki<t(o!Q{_|)mdYQCXETHD*1M*|8TijD zgv5Vp6{b7#m>nxL-1$VLO8O7c)XRlT(?=D{P4|F|c`-$3EQU8(S-)9~INCk&d?FWH zq)&O-A^&&yvmT`1p{QIBD^n|W#Gi8k?kC!%U^Ak|`&27BXpYcF6nPg;T)2^pb+agy zwOZN&x_4RWA=hE*ud$Hd43uvmF7EAcKIhy3FH)$+`aEg9n<^Qkg0i++Qn};c z-J2e1eZkT%B|hW5(yg-@&&KDiS{5QV9W7u_&Fh*w!^T-E@^GB5;{f={BCtE15kEbh zDK0vX`;*BHtO07hPtHsh9A(!uut39_;Rk~;hE&h;|H-21d{C@=Bb(%X=YwN}6TggS z<>&6h30?%#gcsY6*%>cv>48U#<4>m)LyHToxEbv`p~V8~dekCcvAXj99X=c&V%5DZ zx%luSnv+E?@Y!L4j_=pK+2GxW#ShfwX1mLjc*(rPwgJLvmJlN+46=!F<_)$Ti!>zEH9gXFNX48<$nOO zW=-uj-r4*QNLWx;y2VS^BcH>jhSYU2iGy_AMP2?|T;be!@eF(%;~%%dpUJ#v8;Qf_ zx8%=XX7=YG<<{aQRL6{P$x-=AI9TiSUTkZs1&(cm^E}C& z9F)bh@J`-K?GPDSC56{5yvh&(r!=R8$}zUA<@{nj`_HdW{w`OHa4Eu@jo*$e z$ah{sYj4&Sr99 zEdRZxPU?r#=CIirK&_e9AcuY~@)iF)1x9tfk&)NVsn~F~##|PZheG5d3%Dnd8%o#U z_nrDxdGBcXY6MW6V&8$!gQZ0xy&3$j8oC8tZq|P=-Mgl@sY}!H;>yNjqvfw#9r+vq zcqvt~_Ap^)V;^g+rbd{++zuk3__HXcckr+K7mb`&=-iM}FmO%xFQ5uOWC64o8Hz6wAww#=O>U?inaCOWaNM=h*+m@qut~#8FBsY$ zjev0LZfxqyjmaq$?OiYJP*tl{)pUp5&D6aWKeVVw>$S9Y7ty0#BOeX@?`S%_$jxOY zU!aw6gap-@Ol;O^ARGJvuCk+|oh)6&xmRbkRcv2-mkj7e6`GYOR=HvFKxxm|aj(x` zFMXhQr0sThPi`7O@zzHNg}ik!FPues`Yv<;RnVLDL+d3a66!uSLDRyH+55(Z|js z?A*p$9fw{>)4qy(6f#}Bo3AhDduOM}1Ysk`#T@BRyTF90*5k}Ci=Q*xK!{w?$| zhaXVB4d!+R(+!Lkyn`y1*LiCr3lJI6U1NA6PbvSdW;QM1rS zhW||lH(L4s@%AS0RTfv=e*yu5A~y&TRIF(mHBnlFYl5PQB=}st(X@cts+FoGv{FsO zjiR7Nl4zb@uVU3M)>c~gYL&X+h8T9RE~r(kR-rESUM~u6hzjQa{hfL4l7Q5{@B8`3 zk39F8XXebznKLtI&Yb1N=3;nWj~fjCmGmsvdIdd$+1K~6f&P?-2D)*+h}rPZX6!(u z9&>UHq7MWKe{uA_gke-2sT+*%)sFHCc8scNv{l~>eu{OV<7FiQ+ z7c;KUaHLXa`y1Gc8QTRxN`XsZeE$-pe9PML9_@HfeALB^@zwSz zdOwsZB6ee3%owwF=FS@U4L8oWeapuA2_72f1Mkm{vjdSGH%=wNxZP{>#*Lk7-JM!( zL1Ko-yzZx6TlxmgY_f3+65s#G^-rfp#_dcVX*vn!>4(WKR-)_#f!FP(FG-EG)f=Dm zrQY{Wzo+N(hsi4_$a)&g=MR(8(w;;4t;+9wV6wku`?bH(BbPt)>mW3=XwY*MvEi%> z8u}d)($jl%khXhH+Fq6xU6d?Pq8P`IQh(02KI4x@t1xl+dPtwj#CvgBrdv~7x=JhY z=+w{*KDd&7Db!x7ai2eEc)aG*BMUepwSHC-x^B_Hm3QQ~{EN7R-Kjx=oUbjB-@G?^ z{{;hT;@+pNpOJY{P3z{88W_9Ufs5Myo!|1Rme1}k!JVyH;Bk2vKbw1i_^9Dt6|Oc* zOEGF)F8w_nlB7MV)&+HnuR5KYwOaiOn`kIk0tw0i=da2h?ptkTa9y$gzgvY*87AU>iCt> zzTz_}d?K|^4M_AC472Y;xU3)$sJzSK9=Qk?wSGN)4 z@q6Ys^$)zm{STEx?4r&gcl2pFUsqp}zoQ@K`76;jMBc~IPxZboy6Pbx2z5wnh4@Fm z*x1bsk4-G#^fobnGsK_+-2FPhJ@3ENwpjYRhj{*xqpKEkfcw3yu!=hK1JbGc0 zOe~OAe?FllT?3rn1MTvM)BqSg{d~ei2H$=JHhVUuMOsJsdyAR*ezT-G-~EQr)kB(r zruGNqy5$>MV%VzN`GH@{d$iK94zXG@38Ix;x2JYy+``>m5iX0$ju|tkPxM26T9n_0 zLI@KDzryDIq`Hs~1KI>dHH=_$ui!~wYIH>S<2gc4(m!Z$SKW*EvNi5i9$Mp$r{*j< z82|Nk_@jwI=Xc2I9tgE$VO#&)dCFu4F8mYPOLpAc6g?Xub-llWsL9-shmx~tHs>eS z!K&Ay9x4lmjFc9K4_cocX(!Q^H~0h>T#w66@UaY$L0VIw)s@QxT^(g@;k zWzKg*yg~2yCxziaN1CT0NL)0GbDEXP`nyf(KA;zkSUV*HHTnpy{WJ8Frjr}?G&;GThv?+ncdLWezpqapJLb%b=$P&&J`H@= z*!crS4$g>@*tbsvCh&*H|u>u2gh*L`{0nbzLBa=nuie5>G*g7>j!}9 z%Rf#1)lQBdLJ7avBcOy%el#XN|4Z(K`rpnojvjW5wO4k}omA4@#TAM6&1)D>R*gWP zTKgKR1+z#TGiX4RSQrR(;wN>dLIL0ZZhmOC{L73PG(7r&*Ub{LD2&E<*nciuzKF6~ zVZuIkOn(xcKv&ZBS%p5$DlZ=ny?R90_fRxo?^LW>_D(PV^N8OS_M4FVHGmd8o1@m~8>B+8T>)#@dU}}i>=|5d7UJCm~&3kJvkXqK0sAWBuX+cBL z=;xEIT8l*Zi)}^y_Xs20&v`T^rrr4)YkYj{yS&@?MfPC=9~?b<+o9*m-si#VwwbBY z%*;92dz|c#9+5jVIqRKCH#pg(%{&~5YtU{H28^yl?Yhsc8@)**#JZ10+DC4;>`Cv> z4|j{-YkWs)gMg@;p<;oDFqM;7d@v1~8F#px(Tl@v8<-i|h?oN`p)8l5!VcFVCHW{c z+j*aOnQM(hQ0$8Mb;u-5D^6DRbw#Nej=wz}9+`q^+miIUxuUK_qer@pqMKW}>!?*5 zM!;#s{pYOS@(-xG8O-;rb|r($_;cTb64!Vf;~>T4YNiA!a9QdW*N#br;K(?G4?MRx zryjN&63xZ8cMIm?7Jgt4b-PAiE-Rkmoo&~BC$1GAf_bps+rE6Ka(nWfuR{e@yXCCc z>HY3SlaX7j!!qehSsW)sFIwU*M3)4H`|mb{9-7-z;bbjHuD>{X8arQJjj62G%l-R? zaQU+gHF3bX>4!-g)?M6HoDmFvSLZDs^q}0~7nxXfY%l-^T~U5& zp5~RYZV^9~dP_$u+cLf4U~tahM`Pl-Taot&KSBHjJDN(HSt}J-sO9j$)PN=M%Kxr~ z|Jw>p0A-sZ#$6%n-LYPLoLtgJcjHDV^*ZukCTl*W+6YN_(_?&l!x~aauaY3o(H)=FsurMQFVKh|qC3mU` z@5T|)gJq6$YTnCw@TN-QDEQ};KAknb&LE@~WLkgs(4h58_|cd+)LPGa?AmYojBV>r z@q839wb0mx(+=Vyu|~i>8|J7N3rXvw8R1Q6;r_?P4?)Xmee+p%rHuqc8+sXc9?aQXt})P%;u>V}H`i2m&MH0J z%18Vsp=_dbFlzOo@o$H#{u!DblipQUv9Ek3p5r4`o7qSCaD04JRek(~qDVQJnc(Vp zIEH4obn()5EZDh^+&xnMMtywP-o&_Q9;i*at^5)$4mlx>uCLk>9fel$x6#8o!X20H zmv{!12g4B>zo)Xg$l7Lvsr{lTQ_H`FEt|I5`Ger$5tYuijI;paJ z$tsAJp(du$JA8;nZ@HZ{8?9J&1i#qXM({vIud!w0a|?=8t!xZp*EWMF>H^(v96yp} z5f%N%oaF`Mhy?+qlUun@&dAeku|r>|rn4{IOYQ<|;OIemb032!kZ-DUkcHM;hG*II z!_{J6CQ^-wGtY8*8>UYdR_jz!gQ=3DQ&i2+S2e#@&d(h}V@}G33i#wbIXQQ{YVC!k zitZc6w$e;0KC{808$zUrIss9q^I?N~cwV%^iyyhW8_HLM+{j7XPzvQEx{L1*3Xv{Q z?CB--v2PeXF=@t!AUeaX z`l^3M_c7unGt+0S<@Zef@M%BEAC%10vIam1gTwy#qU24CM6ixY?`lBubT>4tBTf%% zH-;mS&xjc@y{=@XOLXJy%FW8=>#1KSqaja9W>@heqI(&g#cm4)i`NPKK!5BecK!HR zr4lybOjG7M$6gljl06HGpXlD_hXZq&HoLpXx#G?_pJ@F~A5pYzpN(P>Oeg%*JEqBB zHCWUyThC~?o<#sV&uDdh6@h~{*AntiOR|N*VNPXM-U)woKAsq=i zqIh;Ls-wd4hT7ktB)hF*tA6@Ay&z&Fq8mupCs;1cmGW74$3}Xh>#JYBmJJ!rkMn>*zvrFF$!k zx<07(LX?os{*h)2^XpJPFozBqq+#he!>KC%^@3bgMF;2qkMX(Swf~pnb6N4%8K0v3 z|80B@0P5EopAQH8UyRSv(|V52bvtQ%-ag%p&o|EB&iG^;0|$qX*cpZ15oZ)Rc|jz0 zE>Zzh6Jq;fkb(k3%Rrd8P{&(v9a7=G;6PIvH#=M>@a|!gQdg)IDfhsj(w)_&V*ZZF zlf}O@CUzONjokqJ!aM!pZ(~QK{+eBry+E+4iEdU)t7$)kIPL_KZOKq|EIVi zor`tLUfXB0+9sdipB#L(bs@)5?6EB7z?7Y%7p?2(mPN_$ZE3oE zR}nQo;jxpi0%%U8{G&+zo7@tWObLh1T{^`Jo%{DGp=rbe-Xsf!mLl_#B!?&En;H{e zY`va+g)0m?rTu$IU0yh!mGvj|z6Tj(_%Z)k#RrL0Zie?b%2_O%UmRjI;}_>M^NVA) zhJ&k~4gBKRx?+s&CA2Ut&r_={%=mo;;$|<-OHNk%m9KNYExt0|m~l1%5Qj)_w+_-+ z!SEHOmfAk|@H$6Vm!?sYRA?2JxEls&3KjfNY+p5N_?t;d;ZRfj!N%@bC1%M-#by<5 zK_3js|BTHR!K5*9Bp($oVpbs+2D1ipVZf));dU~|1MYyFOVpfENVd2t_p>Um0S5Iy z*I)I2i~4yv+P)OKmxt_2p?Kpf13V)`K0ZFOgkxlA_B_yP27S!@{*?n--ZUG@ zB&oU%!icfzjb^MOU_+4#a<9u-4$2U#SlC^>bDowtewsawa}6+}Wo|fLL#v>jdIF*I92&vo#|^-c*Aemi<}0ko?=-|QyY!r*fg2crHT_L1wqB7PN9LD$f>lNs7H zok;mbK%38}Y#N{ySHsE@sURq1j$!q9h}VNlOS2~k|K)7K-wkk0=+Tg7uM!+}F)@_g;Bn~^KeiGtTnCfd6FR*N@_r+o_fX4wJ9)c{ z|J_%9u3xo(U*un@z%FU(x#By&@7vMJPAm~|YSLzv(Q_4j+W~0&7M@cAJAquK+Fa+T zz>rUwXlZ{e1o}I-Jp*XjkT2IjR>^u~l2Z6Zj>p-3ZQ{6XE7oekik#s}5S*qe8 zJHxahGJO%IRi5HHZH`;}Gak3G_?M|KI*jeLI+{za))$OuA8sH&Yn7(Khg!p`jr2bj zp~VXZpAbJgA5z-r`30p;-ilzW>gCYvr^sMR=i#{5T~k?G*4^@!=1TeN7R+{oEaGAa zK_rTPrFoU8zl6p11v{l37oT+~D~OIoq%#w0OUt^8^S7>{jNWx%ax)FJ57G!POIE(skZ-w!ilr#z034^Sf8YaYAe4aI#KPIFNKkuIKzqGW6PB6E7}Rh_aqRu20t^=6YAr~ z6B_mTmPzVHs7)avX$jRS3Za@e+!&ZKYF?=0OA;e7PPW4MkGMFfDHMWhP-{qj!kPf2 zP>0BCbho+DjNoZxGFy1qUmfvZ(D=4ps&0>dReg#Ms*8=N_4llKj4>em(~3UP!Q(;? zjT%AcN9`F#>xH!Z=)hyq_!u$1vtUoxMp30}V_wk4Apt&IBjwoLs)~|P^@KY9WVnkC zB_DoSYCEm6P{&;B4wMXZ+4w~Y5|2^06FhO$yaIfkJu zRH2yMGX6C#=k7P?RG5>6KAo-sMTM=Khy&1 z|B`=r?vw5E5AU)&^{RL~{KL}EUH#_27mq1bt7RxF*D1(b4!K^7Msn;2*XPX%_(Wt` zt)yl34mb*W1Z>ADu4UU?Cy)zhKfb{!72O z5q5Cm3CJ0Pa}4L<^QfLONR0zKaXT63lgd3&fus>~b#{Pm;Bg%yX;Z+@utF-@H{!ME z-f7!hy0pPbVt>Bz`k<&W8)xAp>p%oaVf>+e$Tb?q8LdWI*6NeNdy9cmv8 z&%}*GX!f%pLSpXrhxvi@c(f!-q4v2-*15~%N@G418-wl>G=M}h?bBd741P%@#(jNm zZRShi9Pwnaow*pnodWe$3q`Up){TfREp%VBm+CViLwl3fJ?XDwXZ_SBSDQ4-{g-uP z(d&en=J#oi4(rZez!~%}jd#w4L)5|ix9W*uIQxFp=Wmg%gw))msERc@dcb`-68m=E zgwCmn&Zh#F!|wWu^&eZylHcU3ts4tu`R#YZZ-9d%0f#ARBF=p;)S<0Y4SJquMm_HMGRU zL6?7^1zOkPTTQwk8`DgF>fh^vgA9I+HJM(@iu?Yc9}=G0ZPJgZyfHD1HrqN7_OU@( zmXx7Cpo447t*J$gi4VNoI|)rMN6*Q}$t(MG#I%?4)6uniKLFtF-L*gIsrT`*<3H8v z^AeTvGGmioR=JmR>}7#_`H{Wcs~1`5u1)k`r=QrcPMEfhKTUTSdpZy~%)l~Eo{J)- zNA?gtOdZ`N8B>MkLbZ-5o?Q{|!t{`qgYF$ToJmd-w*y_7gpk zPuCgIcf;{f1Hx77L$mjSiIT2Ed4gLC&&PkFrkCn!O!QY9IM&$#obg21p!rM#Zp4ed zj%X<26A~jqhO({?!{N4O#U$Tvg@=g~^*rkbx zx1*dW5&Cfo)jti`RfamuV=XXQ<0iV3W&`%l@C?Z*fXJQy>HcM12i=RC9d66eu+A5` zu=li_30>1h_MT2|(A9^CKUKZNOHDBVMO>W+s(*odn)w`=u>(OqsU6=s?N@>s%5iJ^ zdUx^6jauG>j96uI71KrjPHr`A5b0(+JSh>2mm+I`7Gj9q4vcF(hl=<@5T+8?y)-?L41?`o&sONdGxG& z1G+a&!Y!kN_Y>`0Pynk96SA`a4>5fr{+Zm7jgXh2Gka|4>&)NKc$0qgIX-^cJ}8xf(j!J~4+(6>aPL+2(ovDX1Xqo|lYhF-(hU4A;p=NRk<4 z=L8dhD2!aF^A)!7h%f)v%fArEQ^1tT8RMx%$wO^C;k2$$t#a}$ ziKMu;a^Y125s4vBI2nekJNiTEUkq0*?kcUj_&A#E>O${vS+vaPK+7BWU7>;X$n19X zxBhcI3Kf@vYrY1=r%@>y5|_hZF&dqui$IOtN4NZ+&Zd`eKP=F;nWH~)~L&u z&FA=kdXGVIoK2dj#BF;1pi&VAYdM59PMKm#adXFmHGom7ru*t_1_$fjcI-h;|C@@E zzp(LSVCpo6@gt4jY}=xR6LV#4uYNZ#y}KnZ9=Ln1VONr0jufvH-W6xb(@_Gmg4^Pv z!wUUbh5WT)=f(qE{OYp+7jNa_VhdcJTsRvpaNBj-`V21cpCdl$n~sjxSSIy9P0OQ| z^=n|{nYXsV$c&=om&RPIh4*w9$N3NekqkFmBMr$=rFz{ljGDu^whK@nP@Y45n(}_O z-8%N6j)z0wHF>Ed=5ob!`67Oand*F5V09R|VikkqIGKfFzLHCM=L`4FZ;{!)z!BuM ztmX;lp2%in`pWs0?&7^unW>UIPOGzgS6{vnAU1rXgu=EjsPQc5+tp~0l46p2kNrZL zf}4<|8AeDh?F%fJAYFnfZJG}G<8(920@?S3(lU#%m_V@Em8_v(j&B;0`871CIq(%! zu}wo-2-J$+qLdWotk;aO^Lw{2tqWoX5wzt>o4KDLEyt<>S@%aosi_LmlR!CO-CD%kac@GJH>? zg&SSV@W<_BsBjtHvkVt21M-4fe~^(E^w1APEwDRUZysVk!iWm1&F((uk2U@eYvy+m z8~0hLC@sXLfMQq=qa{-=s~mG+v`Pz{bp6UdN9(Q61utg!?MQr|JETsprK{$x71l zowNOX2q;r`I(U68-?*>J*Ee+;`JDCdYoU$_41kVv<<1EAZ2H(2e94rmLpPY#5!kkm z9&@II{~)a6e zI=M+YAnok{5!C?cgE?OL4MBe#7T75}{HJC`e__35#rVEI8FS!)xJ%2C$RI`;C+1m_ zrbuiWT$y-vsryS7vzPH|JD(`_)fmM;YQ;9~CwSNI@gmKTiL#2Z*dXxXVlLI#beyOX z#|gdin)bZhJJV3dzr^yiE%kCP{J3Ea@Z$s$k~S^YCLbtwKl2vLLu}_%ocn#4!MQ-7 z)5y+wC7jq-P#6ky*eiarS_zKvQBYUriZA{_P`GTe+W7E7Ya5bI=26 zs*2`f>0o)*BqRby$Nr(qlK`nFB+LK|fI{sD38dj95o)`Vsbo$2mcWYPTo!6;P~)I# ze~e}AiG`O)L(QNnkO$aS+{YExd6OHirPEPnH|fEJ$M?qJZ>{w5N|kP9$hCvB#s0*a z!KVH&-)&RB{4RA^?qAP_JA)g1urVcXU5`75!Fr zRPI)vU{tvg^6x#UNE%{{}1r zrEXf(mfE~ftC4{tk+T+;(_)P{&RlR}nOwCcIETYp45;bsZ%tG%H{9kb zn^D5$WXjCEWP-2R1x)3fJn0#7$v#i=s?+&xe2@8eM1c}L(d1|r+c~{NRwvR$5kczJ z9QkHe0U^|JH611V%VBBxvNdx+a#vuP{J_<3S0VQ9_?~ZN@T*RQR}jcaUEsh8*7_#* z(Q@2l)|%6@`nSc=bRO>5m24}F-KnsT=kSp@2}TsT^lt3VCvT~|w{3Z`J1^VE1GfDr zcIPqsI9MM+)S2HCyU|W#wVq~{(#zIQHR42bRtslC&T86TdJ3OLcS%@P-IzG`eqmKH z&tT$D(b=X!bB+}W`O}ymsPm_l&WOZsDP2buiqYN_@o%wsp~lKy@ueoJv7JxxD+sl# zyb>r8xJ(jHFHSU9o{Guh#oJ3aDh-hsqAa_mN0_u>7Ll>|Bo~N^L%Z6tAZ)_6+S)Sq z05ZhxE-h5Y;x}64$VlvgApGWyE)u0lN*=Hlv2J!8$1g^`e`k9TMD%3^53aN-7GR>H z7L{4GTAvWEdp@TJ^?Y*BoC>RLLFrs;j^a;Bn-q>4y}bq@=YdkUv~9v$1;%BFu+LUI z7r6FRt#xtlstt@J-2YgkDf)m_E`8&HEV_+}%Ys6#_YYVwPX%JJpm{}^!oLX$8~!NF zjRV_YW8#>gxR55Bcq_I}O}2?V*@{JJJlRVc6El4^(cwCSQijSpOXn(_tokV})w=EI zxQ|~tsc(ZuXHWN~u-KRwHB(@bOPMpn_fOa^dVX`(4$<@AOWQ-wnL!~#&$ei3wL)yU zGDTn*iBB-+L|41UTJiL(YxEultHUnMzH)#L4qDI~$FHC18vYeH-twYrc$cuw&{(8a zy)-6n2};h@CrEsrp6Q5)_F|*-XaHf3KBos40qjq#xI(H?wl%k`y;#{xtSl2K_E+>av|QvI9-O>Y0dy1(g^ecV!T;Gc81Kjsk{8pE^g1j# z#1sKD!9u80_8HhoCBcaALJQkGzp->3AI!o71=s%n(mv!qCtu*~YVbRcGB{4Sb-%%R zFX7-g>SVpRJDALNVGyp5WhM z*W=H{vvp)yq#D6An^_FpThjb&@}LzL)FxSj_*QHGo#Vb_Kb1(|3IR-9iOO&1prcAUW_@FB@X$$t$IzP z_trLub+n0poG_!;y8*SIY|5^pR@8{m{h{QY`d z(d(obWtOAe#!g4RL zE#Ll}i|o4GKwP7(!VUmj^q5SGwW1>?%+pymgqfVN2)zX8b2wyG^c3|+$`gvKHv3I$ zh%Uvyi#;tdN&=`>;g6VcV$*dgUhb}<4<^4EZF4F9&>D>&T=iF7icgt?zV4xJLu>!H zK{li?$!Ag)lqA{EZG3dDI$Lewo>^DSdn&itvxPc0@EZkaWQ*>+2(!H`qnYqCWt8sL zPt{lq7K;rj>0{N)&jzKOyS?OZVLhP-iF!NpgZkKhN0{6Eg@BO0uWXr3)UVBr$lz~Y z&jsExVZ!;X*PVPb^%^1wt9Q>!AVSz5OMY`+z9pBbpGX@tCfq$B|KSLtTk15O({Ekx z2c|skeqDa~d)%+TVeL4XcG$TH&-kVrpmXZye3Gb2=Zt;T4tH?ketnb4je~L1`AQ;k z=mHBd2JdrGS>}Fy%Qk8OH$9_9OD#ZCCw+J-GW1nbqdPY}0F>nAme>mAap;o+@@5uo z!?4T-NJ7+lD|rGD?wmJ(%stBp9jzQ*i7{d@q~CpLKri`fF`EK)LodgxqCb%-jW;eY z)==J#s4(UK3ib(iVh#D@$%@!w|KNP<#Lri`t*Vw{b`C1hMkUHK^9`NouaGCnUWcGI zYg)kX+X1J}?hQJwrMPbJ=UUyZw8nq;_XtkdA^#iJ-e`Hy-p0gF z`6>Ai$TxkGq;#Zo|6bEsmzTWCszlCb_whSZ*x@BOZQA+m^2OphPU{+HC%e*8-EsbT2XU{lUSFN zl{IT%s6%0@yoaZqch--mO+9aVxT=4sU6;7t!=n#6{JRzWM^ekEn$}ASN4M;g{D3(( zXEZZymoNTtqjUUEEjUuY-NBXfI!A}k@!ja~9*#xHca;uBXR$(L!T8vV$kdqFvX2&o zYNO|z44gm1e8K*P%A@Rkv^m)XAnu8u*O>JHCyr+G3j%liz7`{*ue*wwQghi+V+edF zl?VX!86E04Re%&RA~mW7}BDbkWTl~QZY*etoh%RUxl^RSVr$ zsHvg30siLDNBJ^#$AXI93wL8;&|MjmDe)9WqI5~ zrSBuW_^~B9H&&VP+wPg}S&;$jVvO->Vy9h;`L>Xob1=%au_~eF^!%>K%n zHAhcoj`E?Ne3Qz830|QIe(qOJa5o==`CVeth@7I#@251s<8>uo)spD3**Q*-y&wHO zGsh`7sVqImZeU|eo2x2@GML|OmwLVMS^a;E&#NOl#OLa#viNMs;bPB{;{EofiqV_o4Gfz;B7(~m1^#<=2-PZdJQd=|-gcS5{b+Tb!MhiD zUJP%PF&LWXqvL8GhLOlysPz)Ce}CF6CKl`AzSE z58qTD*YDH8JY!0j)6M8@NHT>mpYy=!lj~!3O~mG)Pm|#BCN&vDsxX%3zD*w5Rs=yg z^1)#yCI>?@dXw=_@p^#$GrF6AIY4s8S=Zzd<0G}04Z-c#GAV9zC!AR=QytDFQBh|MvU#4H7jyqIBT~U{4zkCT# zd+`gJB5|JdRiW)q2JM_82O0K1Oa6?fzcGFZq&l%LI}yHPWl%L26C^gF#cV=e><|d^ zkRezQZb)@nGcb*(^7_fCEATI8-jri~{rp(%5nKG)sx9fJS#;dcL$ya>C0<*FiJoh5 zlCe&otK(P9r7sYZTR%CEq~%>~6>i$YJ~;c=9_$0AuUrILA2vi{Q<=0S@a38TTqcI-b`kx=bceqqaMN(~jEa+G*O<&n|` zICWXiP7V!`gWn<;j*{-;{T5@X(}BYE%t~f1r&6{&+^bac;xVx|u43Ya>Jprnx{JS9 z#8T*=a|+)8cFts!2q3ClyULd0z^1aNfJp-uv;1<99fx<^bQ| zF*&{1uT_s1o8QT!2cipw=z<-CssohUf8$ql0U$G;w0~Xhds=|W*`lWoOZp7|s;-oD z9mu-#EXNgN;<)f&Hm&Hv$(^^LHzx-N+9n^+m{UKvOjuu%oOEa}`|$r-|EW3kOMm+- z>)-I$4(dNNsDJArz3Q)-2qQRpmH$$@A=T((gNtE-J`Wn{Y&2ZjgEV^ZhAcXuagK}k ztscwpj4H){$Il{Xoefioro!Yp_Nl8LQeH^D1mOTjQu!^p!1Mw-I{2dNS{GD@V=Y-W zo69*5#SS*Fwdk^xP|^E|=kMCsEj=EnHnHg*&IHtzDlm;9?JZ~Bcm&n=IObgdx~l`C zP4kat*?}Zm5=g|H)IgNfvvtFZ@SX$m$C?g+DKc&pep_>7%|a1+rv3Q8!{7J?4@i`P zDvM3=&8F6(c#^3Hx{F_0;)Zwm!9CGiooD=l5%u<$4C5a@1uNno&h+A_F$XH`o< zk#DgoO-L1LKZJctp2LwTAuoD~=mGj%nHOq5oG0e`HOjip7? zh3EmL4V-!s1OK7}okHropV)fm#_ps752V-EpdB>fTWSI$VR0qQ4=c6Q?yp*^b>69i z^s4vgPqkIpAy2<5tr^SvfPjh8&QeR@Z9DHaWZ~01y&#X>N zu-_XJGw(KC?!fSXD>Y8mRPk*z z(tP=2Ct%C*lH?borBeaDr1dF!Y&I(?OdW&Nt%c5g`nV+kXPcyh7^gz>afV`$lod>@MiGLTi#yUswW^}_yO6+fvgsox%OcK zl+cBYrU?L^=U?fduW*1=78uG2jB%4g?Z1GqAFbBwigCJ{&YR*-XtK4zi(jh?av{FN z^RI2yZ!s>7R%M&sm5*<2c9CvW>^naS@iyJIn=3N4eFQJUMG2L*&E+>|o>0fWROhW| zyF(xOTW)0xCK$bTHAz(4#Y4s`?Uk-ty{p=4AXyfESMpoL;-Ssr>V*(gZzYMUw;Jy7 zy%H1FNT@~7W%E-_=(i##khtSP?u)A469)~2Ngu?-6(#QLFZrqu6`3dq)$YSD?1Q7i zoK>hcI;0JqjPg(Rd=y|w!}GA@w^Mw> zRdTh7?5fcZ^vu^osO^X3ajpgwd>MBLq4xLS#zOOlbWUh6Zv*YG@$QVmCk~WB_Qb(o zCZ&YWxjYiDHpc*=HpOzNG3Nls8yp0|g=siNe87vtX3ka4u$x!~g46%YM*}?q7aaL=^)Oos3j|YJRUfN8jP0}bAm_=c z6L|7vv8l%2*tWdz#Mq$zC-{T<;Jf$zS2A5mtp`(prCO6Q=Cgf?RVNYeALN|+l;#6 zhg8UCcm7M^@(<9wo&Ga9liEMRUDiJMKiv)Ix&8491n+bSMIl$6PM+i*<_TSa^QK6U z#Ta1bM_H;D!zb7^%?}$c{k?)6A8}gNij~F3qtKngf-~_|EDpr;zO}IM|d%;+u~=# zl(L7SaEw;;dqR-bco`ev_2=Bhp2eo=8`=#X)rpt6()yPH<3^FKGDpPB-2!(#nWZ^c z(OOd_E!d=bO#KZp-_4)F-rjwZKRn|OV&T9FQPD6F?JiQju45Zse>$RITAzfD(7DX(0;mCM@q6ZasfCoQq>%`j0@$u@`1 zMh|cu^|PIn8LnXlz{!#1%Z!J3yWtWaMR9hH3e6$&MI@k4__V+@!wm_^|Ja7WVFSGB zCd6e)z1m$|F1q5ze7|z^g$TZNVi8B6Nc^f+x*Hrnk)*Nfxf5u7;T++?9YnpV&!e1} zV1MRVsu&>rm^wk)=;tK^dsuLgdVTzuF1eodRPg73^ez#(~Dx!;Vc|O5Rtgg-{G^cGrCMG2J-PVeom2~A$=R{_89zu=+COP(SmG&S_X03VJHx?j7W$>K|g749G-Im--RfQ_Z6^ zS9gu%KWo$Vr0(L*`!rXXrsr_&bYLK+Ow+$iL{}3vWc2e9|Cz+wPwK$mJ1g{PdMv5A zne`yJpzNKK#r%Ab-S&abcfIMjY0E3+57LPha4-~1z%p+}`!z$F}Zqn+K|!P^EFyc&NgSA0I)VLRgc zhslL)8(ISJw$fL1cX4g<2&`~kSzXwqv~X-)FfF3LsY)5OtUxb}-<;q~J4^n!xVK-Z z^*!cV9BT!xJod-Ghu;f$8~Jr%AVM87wyUptMqWhil%fe~SCI4X%$E0WR{4nkn7Wtz zCX*qUUqtnGTT{#AArK?>t z&XH0P_9)49hDCvObdyxDTL%w4q5PW8f>JON>W~J)=vB!tdDxHy%lRtNB6}F%6BAd8 zAYBu!zGJPYpnp+8E`>d6{Du$*3GV%XNN`QRTz=wYBagsn!jucriUo&hhi**lbcLx) zw9Cm#B>uQsALbUIDiZsvetYo_BRwpo%=N_q8sC^pt6JBLz0g}NQzHJY!anMc`|y>m zCQDkud^3BeX3r@_8@P7q+d~iT`3xPf zen_7z-D2nF5BlW{WtU^4Ap)3}#Kh$q|81-Mp;cKO&=*4K96pnnF|mU!t}?W+<$lq^ z-)RRg{qNJuLS81BIbyKze5DZjHhSTtYK@6UjZ=g7+nx~m-CH%6bJ@5YA#}yPpba7P zl_gr+-F_qaq#WsIDRG-dG$sZl(yN-}Ni!hQpIk$O;G(C(=Q&kq>xc1spB=7gB*|h- z9$b(sU7@{B3UJ$N^5NSw*|{NvY>}M@7nqRgG^uV9JO=u9dRD&_No4e_TE;p(ZP-1z z>O>a`{%V66|D@Z}m{61g$0BbCbxb zYw5Ln1>$-@XyCs1hC9s-+gk%@?biL&vu-u9)5(jq{$iu1%hvyv+1*1z*h#|ZE*Uz_ zrN=f_akS^ch}II8@Skw(%2G{ybZLG{%JXsTVUB0YqN#h zrE6#+SGcQ*4>n>|+rN*NFcbr@;{5@jyn?6er^R@uJ>P#97yr*n0E$-lTn+}+?d*kX zkP_Ju;=BK1eE5i@qu=^@zy;P+mV5+P7mN|=BN9vo#>AKaN*OXZ&%_p#di2ku2kMNV zlt>g=^d7WBsT&qwc1&WnHY>bXy_psg`?o~ zR3E$1<=`V^hKN%kJtlMCr=u170{{gmf2$@!r8bpnFaK%nRGz%fHV_9~L1QM{2Il2| z;^l9R@SW8B*TQ}8c3sGSTAxcOsOKqq<}ZsJaWDol1?p*_Jzp*v8$db0M}@7miDMD& zip;U=zfEK%_0p~^C+Eq`m_z=#2Bj~|8+_|x+n%=?EY#JilA+m?dEuxW>bSxLj?yn! zyS!k30AA;ig-O|5@RV4i>j4rCX(V@LkJ^M* z)wYgjVuIBc%-49klXa&JzHGNPW^dI^EM%+zLRD%c_gxvm@4fb=Zk^&a@6HJ@M@HQ8 zPcLpCTHf(kEIKKhBjf$1+I!hjEjKxm0+aDpQ&+#HyKa=^`2xP;frHtck%MbSC(Ahv zXXU$6R%_`T#5U8{1qKwEi zjQK`DK-T6nGz;If-CtBw^SjK>;swL2v;C4-6$fzlX|1>ALzfsw!0Va|xppH{vC>yi zroQS+Vz0~mByDMN?jI?~5?;aqhkc@NvQg)JW-c)iMcqRdiiPN#T*(HWIlp+ZL8b2V zhmY&?(B!o}@S*aBF2i${p>I;o9dgQZO=Yg_$}F{&RtJ2_ zQV!p4pL6``w;f2zJ!1Up)?VM`Hpu0de!ZDh@U-7n$ zN?@8@;KhR9iRKAj)o0TW({=dnGG&dPBduq|^~G1*tVQQB2jnh+-}W3FFCf__ABl#mV*AR)<^jAuk7FLf42TPAa@D;SNHF=-L~Jq)T7ov!Prs% z?7RARgbnWW%|ZXt_5(y%avBY$tW#(cd-V$@eI&dg<4;_Q7wuJ39cJXsmojf9*z5i? z$O_gu!vE*h%*Tqx#Adh@xF;2?#y{U{nK!tD8+l{0NBjk5gdXu9DBVm~louW$uwloV zRCLprIEogLn-+0gZA^q{k-gHo3ABgbDu}ZCCoczY;;mRU`A(C9?`81gG%0w)dUt?p zgvxpt`tgPukvyq6n{qYk4mFCs!!>S6sN7wr;$H#Ns>_xNVWYcpPL1jFf=H4j4+Uy$ zk{0eKAS@ia-p0Fa zsaNl>8@=h&aQyMPq@Nz~KX4Z+HD@H>cpEZNeDBo#^ov^hr2fKt+mgPiKWDp)-Avl` z5n=RjeKbVF8<}* zTz&A2K7<`=M0YQteCNS3^*Gv8k|5=W`kB-rmet7%?lzYPawMVo#&y5wI^x9xmk$by zvB5(jK(k;V|I2jTR`iGKe%=(luw>Hrm7;y0vIBYKC62g)^~n@m5*7Sh-lne<+>rR* zFP)}Hn5XtZwJB#H&Zh!sN=@(0Cb92Iw_q~n1xzjJ3UI{!5SK=3X)EBdqtRrBt6-J5 zaFrw)Ht%b_YN7n_vqM zl+NRq0IpODe^jzsKb4lqrfsw|5kkf^)a$2l^>gqFJdwJB$DLX$BlR8rZop2iO7q84Fe7#v*ZiZdL7%+sKNH6Mt?u z!Zc9(qIs#bg-~ctyZQkN@=Jrt>jCP|Sx_hEfI48c12w$|sAU4h|7lPUqLQ7ds9jz^ zk({apgK|Kve%^s9>jCOl0yQ8D>SsMbP0NB>)RyhwgYP;}uU($o!7~I3e5G5pcMnkC z%Yr&zpxFJ~F6hR?5ceHPs%Umc54d{^ZebSg3ijq~Z_BfAzwuRYJ2pAE-{}E&EhQh- zx)Y4;c|G9DAANwu<+Hz{z4bcL(%u(4a`AAJm6Q@*;!2f_KM>5*nI(ytLBeE++LNN& z&w_+jmoPc;gCJq9B|O$j`x13QT9-?ktF+NU+B%okrL?_*G@B`L<3bYyNps62E(JFv zKKHf=dF#lYhcbZ*v(I<&CRM0*)e8T0%hbn@k%LH}Yi9X-o{gd2C*0SUNVD5+Nq;#EYV2u*(V1 zIY#2hP!c~5eibH8=2vncMc@}QEwSmp3x#`6wB%q8iWz>Ay+a5}uag>!xx8x$POpy1 zeXKw$$@KJV-=`s_vB4hL#^7s&Aza` z{44&){2S8w*{-!Kso9CH%NUj}h6R2uH~kK!f3$7-9HsBe3fL?E45j~I+w{pwzjE7j zPw7u>n_i*xLDte9?HQr;h)c)l1wQ!PKS=}=P=X9&oi?MG`w@2X7<-*gnr<#7*CH{= z&{K?w<=jSztddLkhX2=JD|;`N%-p7pyg|p~UneD@C{ivY1iKD|o#Htmcs><8Hw4dJ zcXHnk4W6e2&zZsVkHK>)`IC7&yUc3~-Sav=X4&aEcFB}ukQbVN>_CL)AKb_icUHk{ zw6nU`>3o1PmWTcIue#Fr<+1)=)qv2onfJJC%|3@gOx{V439Hgh_n|je?z&7q_l|*O zK)()74qU4z3)eq)tE=De=|L7C|4_AS`0nSJ9ACR-Y2FOzYnd$FW!s}!3Mtu!Q>eqrttWnx`7qVLM_1Dypo?>g2gPJq}LFgl^Ir@zCkjy!d4WH_D0Hg&sU6VT(3_t1f zF3vBUnL4tB?OTo@G5tAh)%ov4mV2(~&3s?7@vC`9>kviOzq((2L+vl{i@W|l84<*7 zuocKI5J9|{ie%-73$_0CKg%8D#e;+vUW_c_<1<}KnM*+)@kWrc4=G;TvI6t!s=Evd z(DfzuIeIVLsaL*M%8j!t;BD}=sjEaToL&!?!(W`Ha3AkX;}$ZX53pI>RE5z^H|9;)*?kB2Q&r*B2b?f+e!4skRtLQ)tR;1x7LNqB@1=sm;Z*}Wu14CEr zOu@Dd1sA^f=uVU)oT&Hd8==;|q*BhyUlf}CDd}PV+0;VzyVi^P>>uio;iH4!nAnvK z36eKY0OvznCPGsH_M&{^5o*Tg8qI||Dhg=mWlP0h_VndpJd-`?cq4L&}_<Zt)SW z__KaQ~`I2galfw8lLQJRHOTAG~W zV1#kIZ@F~@U$nCIBTcX?*nqbYMHV9MaO zAhsSawufy)y*EG5dyCDB9$QfCdL5J3HYV=);|_PB-DyH*{WsAy+f|M+EK`R)^OC~6 zXusCVi(8l*wWR~#*i0CCnr*P`dD$Iqoc2$wUY{sprl5M&p&REz=UHo(SfNw*u|=d} zuYa&}5Wbe3pJ|>25#71Fxb2rh%(B#Q%3i$(8RNwzr*oK8hX(yh#{+)!7wjHITi69z zruNLg0hZACi?;Pm;Em#wXhDyTO0nm>MEPd=ODyvHh_MeZBcWDfhRv%*1qZ1;6S z=^G1K>UpBj#E#VNlTHFdA|MqgbP*&VXF z`|5zcZWHY7(br>N{?5^tM{viS;~Vmie!uY2)ecCW+7VJ7Yt)p}X_q{dGFjzbqPzGR zwp0O-f2QaD$vt{}>;d4rG4aJK8*-K_=1Fh=D2<8Z=8Mm1Il<;-Juh^XOI4Jvx&O4+ zH8KL(ja387{y^g|+kf;shO%wJf8f{rGP`elQS(u=eC=aBpslu7L2J1WzBo!RT)B`d z_XjJq$HH`>#>8O#W;+TW!GYDqunETwfwfCtN50sq8bcjtK+RTq(BqognGS4Sz$*=h zAE|j*THvR$%W|#u2B_1`T5=XDp*aLFrw4wx&a!YFhd-EZl3}HcSD|F+=q|y|qJ;HK z{Q@6rj2~N!{fjlQG5=-#>AHk z_?^Z_OKVSDtg=$hTbh;3Q|HB0tT3JC#>9C)HBd||R7x8Y>mJj0Na&^D`-%2_jr49h zWn<#I_8VRo zPdtnNX6fmr=zd-`=?3?hbaQ&ox23ae>JYA;_2-O-dVpxKSDSB-_Ly&1yF#~Gp&K4a z7lPl`@63Gb=sDkhV5LGG=cL!3IS==mZ{0sJQ0e*BWZ!e<+gkhX=G#8K=iBEq-O?tv zJ8r%y+7PH?LTi3S6p3zop0j=2LW(l8{F==oCYT8|o@6YuJDy~c7C&>N=Z_Jg#XZz0 zp9!^}Wen>f=A zh5zUZD?tTdESsi!@ki%S6R%S#KiP|S+UttysjmJeK0Vs19I<)!Q15RwK=%d7uKa!W zaAtS_ou?xHeUc}5ey8P(cUs~$7KKzWSJ`6o?4b$@B$e^+3z7p^_K3~9LwVzmuGYi9 z&))n_%Ng&q#1*Yd)Zg+}l_A@sU3$dk*+UhKb3F)>16cNm&0DFw@kcl6;ooO(ey8P( zcUt0#Rwe3hd8^8(f^~Yt=Gj9P{Ml6yBnPnU5u2A#UjIJ#=(H>=TJ@&C<*mvkmp|Gs zp{q!Id|=Ce4=|lA|HJU==#GZkui(IjiD6N7XLDf9FJ(t(%^x*u{^$VgEt?9Hr;|R1 zUu9W%uV>&z((tA`c$n-pW7RM304ldFys~6Cfahi5J(GdgkcOwDyfcApZgB99+b%rM zBF?LQQ?u|EX5hUAJy?GggdTWJw6Dp*JAAwFrY5Bn3Oq9l&4N_yM~A}q(;($GBx9{= zRO|p(Y!9%rG99TRZhVYdvhd|(!NLD&8vfrLd=W}@=Y4uO_`ljNd{I{Nk{tMVkdTFc zP6qyrH2jU6=xP5v2VV~d|2fjP8919g$;VYBXus$#3t#@{UHd92kv-Z{+a6&%rYYNosJ*DNhd{v@OTW!LWhY`G#t%mN&nb zoX6Mn`Cj^chU~)ul%p|U$DdSx@u)>lXr+_oJK_D9U?ja3vKLkhapO0%d~HJj>!2| zm|TF!iH*s=HTC#5HF=ViYp`#)mdGjRCAX^Fyw!}hKHCJ#&0!NXIZBN(vvJEbhul3f zb&`;*Q!VpE;g(51dSo)%9%qo|r8DVAk4$%vkUR}Y;+F&K+_3+YI9IMjU{F4SFX<(s zlJ4J+WdY%@#WCbTjfq-&kyJCWJFf71@hjhjC5O>W1V<~*lgRri!7&a6N$w_`pcH%F zP1uNW%=+X4U1`-ab9?(|x@`R^@as8prFN5#s6neg(#ba(p;5>Kc`?1g6^YFs(6{+7 z=fA2eX$ey@va?Xfvn)#7A_!*I>q3sd90}d4ozOi|UH2s|Mt5WQz1j|!>HNU`31>k& z?r`!!sTST$N+jPET#HooF3XZ>m~^bvQfR@mDCf=^l6?cx#a3z}5{kj!?`grO6Yqbh zPGBduDRD5JK)bf%R9dRCB^->X7h$1HqU_+p5hkQx6`q-%obYoAQ=IOK7jJjrk;r3# zqbdZ(vh5dS(^*Q{q{|=;`(9zG@`7yZpcZ8lt!1*=i$VM6G*@X?kZpBtHhZyb_XXMJ z2HEE2X0sQ|77Mbq2HEE1X0sQ|cD7}kg+eDI(xg@y9-3SS#|K}l^|gVoq4tkh?uDra zm!p&%$+e_$z7yJq3+?#viXfN3K{MAG9L=x{BT~6t`*FPyUbMBJ?gFopPwEJI5@o63 znm_f5x_WU*efhfj_*8{RyyVCA{(53ygrD1@fVzwR-4FXOns<6(r0OFtH2wuCV&lsJ zX%SHHf&ct@cH;R)r2He!7kRCm-YzaSW3o#*FLd*I$qoL?n44Tz@4rn2m)>Lt*hm#I zM4Zc|NDQi~k6ln$Uqw*A55uAHFRGGZ;n?9uE#1ZEQis2UD~WyT%RhMxvdjx#M)$if z4PAXVPhM1W=*ru9)cUDUUs#mdyL(A;4&TypOo+U_l6OHYN{vdd-y(j;JahMeE$4}i zi6NIt#--VzBkDBWMaufVyEyR+okIDL{FohZo5|V2;bm;3RbsCxD-e4H`+u(~+`?tn zJpt4qfB4kzEt2I1e_F3_D&}4xeL#%edkHeMpyU2tbw1%It7jY z>1eF5xEyF_%9D68MP2T2e#q{nOQq|Uu*jA5T13FIEgsy&J&)Nb0NhFT!88C4$#?Qw z?o=bO7FsQ;<5aa#m*hn{7MAl90*sWuOR$RO-$$?%Y#+=Vjc^CsVQaT$ZW{mL*irSd zpMY|AcG>zBDnQZDr0*SGZ}g3xC`^_1H{0L8dm7w)bqkZMXPFlJnnK+6vAX zZ$7X}Xs0{y=r3u`hOQ?x0^t?oi^#wVKcqjmFtZ@~_#O7aN0kH0e?9D*N_3IXr20bTxP`<&r<#Rm!#e<~ zS?xW#4)pLBf@p}=B#0C!QvqU0ixBqMRkZ*K4@{KRTE~5@CTb;qHbU)%L%Exuqou8+ zKDJ+d>^#VTz3OOK_?cCCn4-jJu=XZ^isYfEm_Kn&DnTD-Y>cxrAKvr9{F$?Zg znm>sx5E(LU5Q zNf+04XTKilN3cW3<8$~dD}!~)hH`ZwR3zA{Lw~>)1YpNo^?(rCe~z_+=U;E{K()Qz zzh02kx)MR~AMDgCdXtVvcyOLxQ4X{wz)zF~^{77mQIFuLU))dfBo(13wLu+ceWf?1?=*zkvIVfcr9TW&T+#4KbxIa5&wT*}B%UIHm1O zh)}D|AVh=DT~r*Y+Td6<>$H8I6nvg*pMyxD^LTOG&SAmlR;6IZ)ZQRV8JakUgmBx6 z5|=GRHZc{Ip^i`4;ywyp!aC!YC3vAS`4A&a{WKIDJmma&$%DxR|HQgAIqbg|F6Uxc ztmwmT9l8|GXnlCm7R7lc)-_hI9=k#XaxEz3_&7=<%-7k!7TIWV`jIjyK7^m7&&wfx3jswn?n}AtznaUR9p%oK}H&xW*_ANfOfvZU7;j-j=@Hd(Jt+0P` z36gJ=Zcb(;m@pz7Ke2(|Ww`H_qk|EGEuY6>|HNv(<9G92Jf|QWB9juTc{0-{R8p{s z-&B7NlGAaOu3)taf+bv`s^D&}pb{pNnVCG{l(2Jy#Ou^Z{N`55VZV(0blfQp#AHA` zWXZAPn<7>J4z-UkC|A3*m>yyOY7G{?>Sl!FPiSaK=~PNQHkSuJ{1B>+o54?-vgmJl+gv^B z{VTe7c(KO}MAfIxUn}(n$$Ic6kYsw82Ni5|i7R zY>8!OZhdkSp}%FA*%T($pT4vTuuZh}XQ=2*#@fZyp6rrebjcMc;aOC^YL~pwB^QNh zh9Yp6xa8Yi@^Cl zNV6y`hk6qL`bwx>^82tq9Gh{7u88E#X0JLdK$$ijhkx6UD?BYC}xcYhOZJSQH8-@J*V zk>gGnk^CL0nfVtCOnv;gz-{a}pw^g|&qXXl=B8Dl^lC%%hu3u?A(nekT_ZWav9m&7 zP8sLVD*Q5llib4?z-o=O$R!d#F@SqO2JXT%+_;V}^N*_OtaM> z1LQC}6PUv68Wz*g{7HFf%r4>G-vnY0%^>!aYe~((+nj-SZwB710G^ZH4^4Md!guTD!I^GuPD)8g*)4nU+9yZ^}{~v8{0v~mC z_5UXj5D}T6MrpO0KE^h<1{EcWnn4oV%ZM?m6e4 z%jPofS(FN4f#(n+a%@E2He&%rRiJ_b$|m+8EBmky%l5`pWfO4E=tewimc{fOq3T0} z7I9dPe+r*mkNiLJN#+^61Mj?n6UvAU+IO=jBbsY@f5YfiiIpvjI6;l$I5952{1dzb z%lEh&=d6FEMAz0A)T)pu6Igevp)x>-YQwaLtG0P zpPjJm-)V697mUwqB)ReNE9#tyQdt61X!8+`RL4! zr}!w(J^dAFQur8xP05FNhsEp0K4CLB#YGp_9ZhjDpWB5XzHO`8hdWGiClL;l#7#98 z$I*3?xSj)UkH71;4c4QdEM;Qcg49nO6_S?6;mWnJMZ%SDFut(`?y%E7z)_+V z#2H$u5+$lSa8vl8a_Rjq59gvRoGuUN9TO9~gJJ~dS>ha=gFT$blq(0$?O8ZCdN}{~ zaKxoVa8_sG2u-Rz&-8)wdvd#e{T#}retn6+Nd4-_f%86bLXcHLkOK3fGBG|iElct* z{6P=54V%q-ldEeMzRR&4>C^S++>&=(YFIH3D?sW+h_#*4T8mU$Bb zLqTtni}05{>-o_<&&YS*r}}u;^BMf$j@3Rq%;BlyXo~hlrqYnb@yTI%<_dslB? z+rF<(ct7IX=K2xWw+Y{kJd2g5uE_PQ+2i$e-|>mKJ(v`m{Z|dZ8amOXUbb{eO*j-a z;kK2yNBS`Q1<4)#jN3T~O|~>{Cy-|2)=#Y>yOrxUV9}y!}C3gY4R434tT@l+To(g(xym)v4tE)1( zbK>*Fh|1)oLpH)7qGN+WY-0A3qou(*G*>a@IHpTL0LdePS(z9M%;64Z#HT$>YGrbA zo(DQ60y;K4MN0~{K9NZbuS~|rZG_pOI+&fORNam$k(_i;RA57xrgbUP4Z&Q$DIXKP4dcOSwYpOx354OlDaZr8(?Bt zlekibs#5VGqo@)@{9(qfJqe7hw})eGy8i{g4JK+oCa6iAR-xYL%1@bli@3UwUQEuD zC_FZz8gWK-A+v+SnEh)K#~|D@EeTpZgvcjOK~NrR)e;fCf$61BALEboF3k@Pcf{cq zdrQx=juHY?8w2iwVmQdF#b44YIX^6}l_=sM;0U*uN#Z*D2&h(V`3o0rUy~}>L86zI z_f|3~IOW{kI?`G|27F*eFU}q@MP6osB{Tcd%B2P0G+@O+>J-V*PvMmuiA66~B_MRgZ;MaRQsK&_n;Ta3h~<-MZEN|99plm9Nx4$ zQ?&j>U~?tN_eH`8x-LJLBi*V0_7jk~G0d#eD*08l0Vhk3e`+shw63N(7=?vwV_P4F zi;jUNUVW@qSNNO&KS)9ri$UoB zc0Q@J0}6c60Qe#uP(BVXCaw2?HJ$+uJ@E`-+L`|?4}G~yx$zkMU!yP76o=O^86Y!G z^NHh2lkvkmg`+7Yn?v0msHYgzhk%OzTb5!U*z$k-2N=LU(faTM@|n8eE16MC_0rN45|=ph>uTHmH!?kDYel<%@n&ngX4 zmzMv0sIr$J$QGLJ)S5f61FxO9r{O&XqELb z=<(*LJl1>5^MK3KsUEoJ$|Ouup7|D~cgk;cd753G4h24s{Wdt~S)R+M1w6T?R0)+4 zW2yM}T~Y-}by1S1riZ$uVv^Wju2h*TwVNg3fK{H3F_T#%x+%BES1*?6*=1TE&_hP0 z?O($BHbxLUS`RlOz4)kq)~BCW9!9h-`<@w%RJ$Iw1UdDgAl1LoaG;|zdVC_Pc1l!j z%VJuoVlA#t{Nh@%sJ@D*zB3GL@B86P5HtAu;W>043XR%-s=%Yi@8@?2gV zN{KeT!{Eko;DttpP`1j|)AMdhEGnYEP{Ws;nm;9qXytkwd$$isj$hCTb1nQ&GsbXXwcei#$ z_n%$+T4!wz{|X1clWgH>ec>OS1OIS?f2xD8>`V*rwKKEu&)6LPZVrAI*}|&6@OR6B z|CO;rU!QF4S9S+~CBFn;_EW9L^}vsAYdspb@mWo_aGxCbW^c78ZHBS7^(s54`t%(R zzOvi0`u#e78GP+2?cOA7H2eeq`Tgg&xROQ!Km0GZAH)~cCV0;BrL2}=1)z{|7+*Q8 zDml11kzXm4>siZVVh`>#c>1xtf>hJe8792cyeEY%q;n7Yg>*iNHNK+20 zKHP@?jm{o>uioh&TlXH@V)`?#`&&$x#t=6p+4Rc9ml$>Pgj4i@)35<2z?@a_g40K} zsbPok$XXc7@s{ZMHv?Lli@Cx;W^9jt9~+E?J~@_49%NmXgrhpZ??oUHcvd3#EUKQ zn9URSCh=rT9JP6(c-3*1IE+N>1ShceksAzL9AY~4WcW=&eLu9wg|GT>t`8RyBFWI% z6v76b4?p!OfA!(5KAcV1BTKjQ$t1h6*L+n|UaU34C|hLI)%P2I(6z4odeOf@`k*pijR?sp(pya>cCT8ei(tf%O!CPtj2cAh%!Gc z=~|MEZRNFkd*Xg1)v{@${o`1mE%75Y26*fI$X+f_r_HwGK|HS^tmj&h@A5ReJRJ&r z9{X)AeWODC>z5x{Y&?0vwGCYF+e0*w9SdhJ6Zt!-!D+iupa_@UO z`ISEmO$^R?wF%J`GhVPeZxt3YY^-Gb69o@}1od~7PDsVE5qJSZEORYtTb+1Y#|%-% z&uZj{7K7wxCQ7_u7Bz@{INQYYyW@${Ab}egKDptEx!q4FT$4P8L-qa~j#ktxbFw8w zNwvT-X)W(fcE?(m%YD*Akv!ijZnOHR=}c;}gT05O-{}1~6LDVRf1n|@%16?qOY4iJ zI2aOZ%>!$@qt)fa9T<3)mbR(926^=wK>sgz>Z~Z{@e1{FRTcb-f6`(sad5piCH6SlIM>! z>Dje;;>1lndKzHe1EAI9IhA|Id$43rtDp?h(ph6EGuD+EmM!y(o-*V))g-E}+%eN? zlhdcL&S%g@TeXs#R!@nK!t_()nze#um$6hag#Dwy%E##yk||1n#X2`IexdW~kyiVV zp4HAfFx}KUL)ZGpgnwtMMJHT;{uoVsHOrpvw=5~f36XUxUVf%_#$0mLaA+BVqeb$} zLiEeUbA*Z)KW3t>!>mD0w6%^V5@6KbVX5mlE!sk45NL2VmW5J{L54GBC z+@kZ$Fw5!AGu=7-x{cTObLxA<7~YR+)t@b$XVw{jUhwa&u=7ACQx!hf7k)i@ z|B~~cgRkuFyfL3&+4GD^IKBSgpU2`3eizxyIw;~dto=)K;O}bi_jB---N6^~$-2fdP)&_$a%BFSKOm8Od|`ofn=YTNutyP|sfh z5V0`o68sub5w9bR46i5Br|bRc@NS#`n#6h2xUSH|b%iF~AsETCk&tF1vf&coMslC1 zp3$#`Zn#=itE1Lu_4F^Qc9^SnY*=GyPLBP)7)F>u$rTQ;etkB<)^SaudZa4}rYXh$ z%aOj(FQb6Bb!^Du=Q-s+qjGMtP@cjF!I6cZ8gw@Y%aE3;7SzxmMJz`=F=GgqQ~dL8 zm8a$0Z>vxL=42-MFuH+Qfy}e>iwwNcoIl$OiBYl6V z{Ggk}I6cq(Zf0Gk=i2Yfr=@RpKPuAKxgW*ptM!A{UTPukKJF`5R}jKB9ItCAh1@?$ z@B86bjLW)UkO?G3z`I{p10IJ1QTT*~Pj)X*)8^^F8};8U{1=|<8+jM_u-bZGpZ)`N z=on)Uz2jyRqiw~{A*dC<1u$2G|V z@GVo({cX;#?vQj7rVtg%&0VbJ6sRd*iKDN?8#U$6G`{AZmYH1G_#DdF+2roE`RN``P7Y<^90k8nKW8qo*Psjcgtmk}qNx<*oAW_FEtHTj|MKRfMS5nG2Uc#CU#+tt9=dQ3KR3IR% zpwL*FP&dW3r)~xb6+ozP5CZgq&VL9knGqXt(v;YUiYa)sU`GVS@l>U^*}H-4m7mM_ zSBG%3*~8_Fw1cJXCzO(c30oZz-exZ{h!c=-oQ8$*4=Q@#U73YKOu1N zj_4O&ML5dfRhc~%{tBdVj1Twz3jWTrm6CNdRx>Lwiu8s^!>h^{;;-P3q(<}AfB6+# zE&qE`8Cw~yoDlR_uBdC_&6ZYbX{Ei=ibHt@`u}17{OaKU_x-bI{C}{2X7<@XGym`Q z&%X}(U+kaW<8}3i^H?Vm8RXu_Z!NivSY2AY;e5S#t?R_(s88?lV|QoV>{GBUsgsgs zqEhOZ3j;)`;{_Xdus-qqco>Qd|&yC&e((p&0;N2+=t{2bm)yuj$$)8V-sW9 zR*rGTCNpJN(&Q>LlN0p*BDB87h2n;2WRvb@&s5hCRB#!Y``dTyp4W^a2_n3GEt3R+ zsPwE)O;j9fzgjFNg=f}s*P-=Rbw7EJwOdl4mRdGfZT`FQ@^>)nv(7lKZ2`Y)60hUu z`)Dy`*>CsL(!Pp-X)Xfs`?d~|o(5?*N8=;?y$yvlYn)0i19VP5>Wi(gM%XM^CDNsg@j z4x-`z6215@Mi3pMYo$bdJ48)(a=`BmbmSsL zN^k}b5fQF8fYyF?+8_@>F7(11Iayf0a>g;%wj01hA?+IWS$_Cetz=e%wr+M#?1_X2 zD#f>sU|KRoAA#G%)(m9`!h`s)H$K2|*#A5Kary!O@Az}}kz2!`7aISO>ZGgPqr6fB*aH4<8bm zV%Y6&GI$jBV!f_s%m*Ln%6*w^Nx0O94V&CpgY&o#jqOT-w>oSsnFj>n+M}=L!c+=_ zKsmI=nHs=AJ&*oTs@Z!>It?B|SGivB27^cKLLFu>I9RuQO(GU&YnV$A3^;|766qi@ zw`89E#4X53a4Pqp4VYMSVc;*ca)feC|3LC;#?`QN_o*Ftg*QCQqn6;=Bmd~|?6=S=RCfET__jtPj5$KoLs}<(SMS_FYNgKO%jI4TH1U+EEcouVd=e!XLLb9k(M=$AKYR1m){f!w*S^a((5sJ129qnrysKyV zVTX^liTWP48XBqS1#~x?O!c=1^3_1r=?pny6WDIbkRUmO%dS+##BeaaS3PECm!^M# zk)DoAWp-VE?v(>62X1U)h4L!uKkCz?hA;?v`3_X}t;;m9Jekc`x&zA>oo9<7xj|~K z-LJwIx{jg0zEAL-GU6Mv2f{1(d*^aT58<`C>Z~rm9?RWDbiGg1#hOstE6`Eb#HyF7 zl~p_*rKMf0b%k?zSzj{5oaY#4@cTb#BMWD_6`kuqU@gKiCQ=8Xbr>YLn5Sg*Z&Sac z6XP4wzxguEFkag7S z3(%IF$-+&vo}>A($07!3otsHrAHB~lAW74$wEm4>+QV(XXOwVJJUPxwJEL zk=5Ef&0@;kiSEaOZWtxpK*^5Jnz5GyqJj~SVm|j_Vr#>~1S^;gc2uAVK0_1f< zNQ!ziLZN(LM1E1O(kVh3p<_o9E0pu&Lvr>&E-;-OoIc2@m%Kixw*23Xqr)ULJ*y9D z{al*Ww$~a@G)<6(#cP9V%irg|JyfxOT@Y7RA&R%V?P%iRWko{kFPY@6S`EzA1>ILF zDugcRHtf^AJQlt9K2}` z-pmNzi4NYh@G~o3l~oIe@8-fw<@EOv+~{EcX7KvxZz8C92G!|;bRpa9Ok_=L>SAZO z7s+F6e(qpRM&tRu!GHQ7^}nT4TR3Zv7QyfQjz-mFt)VW8_rRY|{%}4~S^mS|7v;47 zZi9cGgWr{fFH{r3uW;~{7|!bjU%A3tHiLg8n<8reEQ7zjgTI>muKmJ75&YGAS^Je3 zrh36wuJE#4_*-GVPTQR^f*UR4eJkt^^fNGqIk3k|O-#``42mV9R%(L4ha)Qt5zeF> zd3b`E9A7GzA~OTrm~eY$b|>3Wrgm+&$wC84iB+xcy|pejc$|mp)0KC_BYBR@@mJ~L zpM4EM_w3bZE^ztw?M(aP$r+H_V(^eP1qU=5P5p5Qh>dDBo(EO7YMa7)Qet-gx*E=y zzpDbA(#-x29-0_U1m&Z3`Z@=2dnz|4)l>8z<&1O(5wsxx6*=-zJ+d5ei^9Fx3{LHy5{ z?s{$yt}@ z$KkR~^yl8QC$cbo@~+McnGUu&hpX9aT-zx~ZggtyCApEI%3y9ogwI3&v-qaEFdT>$ z*2bO5Th6dce-q$LOoWjI$%cG0TJG~TE&OS9b`oZf3Zt0kD-i2F)U(<*#AcsXsD-g) ze=Q92L{!asKhHkUE?g<^k5te!y(V_$Mwb1uc>d}vWL8kKY9Lgt(xVkN9uWSN-`Vrw z*ZB<>YVC9HkR>knbu}$ZZ1G$6hO^9fU5XK_NsKG0O*X=wu=bbQ+7}j$*8cf0C%lM5 zLu(QRaz7z}+13APu~fU`O1f{0=-=SVcB!%2?OUhDd1@+C15WwFotok&Z64xst$p~+ z!dUBetZ3(tiQb<7Zw-T7P3Wp%fP+1q9Ge9LoDvN(ela-l-Q!LP4ESo&Z6X+wb^Oi_ zmEEP7ZYQ=G*q_$b zFzZ~{SyJpokW&Cq19E#QDh!K2N|{HM%m_QZWPpvaL)|{VNUF&N=@q zXy;Ox^{ATr?4Zd9i^A3N$u2aLgC@FHs7#*hJR7)z#H0oF=@&^9DrtU}uC_octO8Fl z+MZ#Dgx5#>N&tQeXjoo&tnS)(*i366i)EP*@wH-*#n<9%OKw9O#%A$d?s<5f&42EB zcmge{DgS5VKLkhsYTG_{gVAntxIRAW^ZN8&>I_q2EO)BaIYf)jt1|@Dre@R5F56g( zX?6bDqHIAn+sl^CtrE4nSSh+ivlBn)-`4F_e{QAih!bq-{!F9G)5)ljM=M;X>)!Wg zzUT5ZyF48^dF;1!Xg|x-uf6IHE~z5e{Qk`PA`QdkEQIl8s#mve{Qa5NEh!Qpp#SZA zcjgggkcqvsSM`Gf32eQb*_WKcc7m5vq{toq$SgoOhPf*^D*O;Kqwa-%okE~l5wSM8 zv<-_$@$nJeLlAm6m*nk$@ojkDh|b=GV5>geI7(Pb#wW5Lf#-4b_4wVq{mBoYt&XHt zmxi@`7@VDj9z4CSH#qw_IAa|gu)Kpa*ufbao}i);Nnp3x$t@Q@W${ns0QFf(Fbd`$w*$fjYh`Q9@RnkF`gN~Lzpp?dPfDa0H>x>Xuf|nVA=gtlQ)p)T$4H;tHTr7jnyO3iUpi3i&cB& z(&|!1?s!ZVC!=;=6l=X7kmOW^N(6)zl;XR32IN`>MnqS9S{b|G)YWEkw5(bbQlM7q z#47v*gk$W;x|(zAX{*zZ?YJ>RgVj#Fohdz8T}Ycs1!l5J6{?h3<x7w+y@L|vI1&|6cEf8AH|LKU+ zNR(^x@s~V}%b{i1DDY3LI10Y7&Dgcgs()1+dFqHgOQkhAf>t>EEmW<|auBN|l>d0Q zFGMJyC$YA_2o1l&f10Q4$m|Y5aU=6FG*(_%$|a!f8JX|!*=J-bY}-YA#OC8uZoKf& z^DzY4JaIQm{OG)X6Em~{x90sPzHNzrQR2gEKX!jXl2!eg`Ce=_uc*=|EZi_Kh~2j+ zczA^S%g5nEtX2J!AirIUO5Omk6{SxNV`E=$E@$kB5o4uj>)DPA+E$=(vvS1rFd#4V zFdvP;oIE6VW5@H;<|UGy0_lQYnlu zMEc~L#CF4XIL&Z++n;Np8S5dx(xPO7T%Vq{SbX@}^afsc-wMm|okA@?v>53-Yha+6 zBaMkTo2GbbMrlp?Qase^V1zar#=G0|2NUVnTJEgZmSR|P#w$o_(-b4uWR~tZ9JFdj zqDEX78*vsWA1{2z8HmY5?ns_e(l$^SRH==5QE?wvd=wu`7Wfa;)28W8fM;-ANq&xY z7mcJ+sXQlE1c}B9=+Qb7uv}@J(%X+JYUFT^8C zwnGhlw+|za!zrq4T|V2&+-&w?+13nq**bl;`MKHb!?JbxY|TDfM{YLzuxvlKY%ud; z@@A@4F)?_kq;(JUa-UkER2V9siZWWMr}|WoHg*WvR8O3bveX{#vEdZQIbOXak0Ego zb|sHSsqGE6xeb|~abvjC5##v9_AC}L1)%dFHL=N7#|!F9;`wlm9fs6HoHiO*1J(hs z?mubLvfi*3vKX9~fp?Ly%FDpMhG!0}Xdmr-4Cskk^MoZI19-wx@o*a=J$<+3$3!yV1iUtp?owjL~lMy0>DOdEK)E8%RScM4N0C3i4M5iAs3^2rug&w{IG# zraU{NU+nmnCH}ep#3hz^OaF;dMc~MWMDO0Ma04e!?{04}G9l;Q?d6t5a190M#tO>K z-Cj2$gnmlwu;mi#Dbddmf5#fIzMg(ba#G2NN2kYom`mgzd$7zvk%k&2KDpO(hR$>T zA9s)l#U+Pli?`yV&-s0k0lL}hcrBJlGI9Onnrpc_$>lqi>*-Fy(_uJqR)|9uWGbk-x`LWqI ztHrrQ%e)kwBU@MJ6b(YbJ0FtzK_8mTKcKM@+2fOrOX0T0{o#e*ranDZ)xBe5~ zu*5(1pV(!IH};=+w7@% zy_Z4wZ`!}vz&0*|FFe9PE$(j+8t)=~KTE&6-}Kp}|K(e1N=tv~xT)5xr;hlk<@sU% zc@A@V&apg)_n&8Hm*-f^vrGSZJ^}Yr*PfQ=(=%BsCM1JRsR*A#61iFTV{K*(ojPLO zZt6mpifq32Zg*h+Zh7YRpGR*hQ1v{^)6{<+^P!YFqSf+D>OYU!3#N{kW_il`&$FGY zYnWAHL_qeSGN3Eeq!+R349L*Eu)``0!g6E(^!|?-PBwo9z3^ zaEjIcWO$Abzvsj6`*2?i7l&^e{Kes~eYvQ7Gv77%dOOd1fsow;`HofRPESLHgsN1Hop5 z*3_qOKAQ(&FehScG2r?Td3H_m0`4<>)cB2X9-5u(`4*A8d{oSlwuh^p_34^+cB7P% z@2AgiQ&)K7hGwr+E0TKkgCm4~(*g^*ND$Au{*pYn1?g?N$e$jZbog(%)j7x2LngIBPh_kkq%Q7y^Lh&Q zO7N%)?PT?g!1LyWGHC3r(WDWj3dBwnnoDVhb%XLyV=)P8scLo|_`=M?eD~Ras5)ScKopirx8TYQD}dUrXNW%W?Z7YyX5yIzDO>_D|%}!c;S}uS-~W?HIMQ zk~uane8?S}-J~%PI#HRKu8orA{UA@ESXwq%*O!bXv_WR@_1u}3p^VE_U4^BwuQY&K z4bHj^L9)IX4*}7iwp1s-H6)f-FK~?4Off^2nb|O?Op~jHMJ*t-WQxKRFRfTOr(}jg zO!B%c{r-|^3aPAuP=8E}K$7SBl2?=_;Yh+1lC3TRpbiKYZOXuRyF$9Qn$XZg6F-17 zlAVUf>eiAW{D_UXzhndPJ-e#WbSP`m_4r%Yh}7ANYIPen8r*3Xsq<_qGsgz5%x9}m z7-h2$n}g|OqYl;0&Xs+GX26?|xdXjAs?l6ch~98VoFrhzn#_)%BG;#XwQmAF_zgPS3Lc`0gC3y--FZ18Ug!WkbbP8J}iK<_~DtzXd z{(GAL?(pBu{=3+Xu$qXF6zSnIHFLQ>`P25&PsOEs;?zQm8v+#lk*`ocm9LZ^jHvw% zdeYS}i<{9Cr^Z=adrzDiNDuJ7P@f(Qx-wJ%rdmxjudb3A!-_69b;et6{VLo#t7oWhuZky;) zoh~w}ZtXq&#YQ;6nxf@w?<1Bd$0npy*ZZv%K(8LPoe)n{6vtB)h^*$g?FU@9Biv39 zT*22L1HD{7DT8?YVdybxH$5;;%YuQnAZRZ3D9+6IPSu0i7 zB%#CGL1H4}0tOb)is)K8eq=uQc}Z-<6s~LyR135x^+yX-m-SY_^_M^x6=B)j-1Teh4c0c;d^-$=tq@U=Srg ziy?j8bB-2iZMo)M0IF4y#CRHi_zzAN2ARj|$x*i7D${WE?8I_WGe_0Vk+OXxbRH@h zt)lw$GA@wQEN_@lpMLrDydEsmHr)idS(0f>YOi0iUQ&a0?|NS2VyJ>cc435?Y?m_x zvDO-d`Ref!9+&7rDyI_rKOmKTBrQbcB4NeH=|RhN*ymJkBY={6iKYuF3))FPf=t^o zrY!yAg}F6JAtBb_?cme~29M01|r_Hv!3JZgY$SEw9J z(iEFZV^I&e<2I|3rh}h`{z_l>I~+sD=23t(6U~sAy9X}wFs<*#aG~w7W_~c2(p0yu zXZ%Yo;a4*?BKJG;f}VOi3FT3z*R-j+EzV2*2F(k_)a|y4BA7VtxOesg8=}bOx1; z^tmj>o(}!M$|Ga-YftUcV6Rlj#IszU3#8CkQRosnNdVtEUe!(&4@=t$P^E~+5h&r@ z+w{$Yjjg{^21tOxhJqVx5O2ioF3q>{tCgR5nXgcInNWXPcNWK74Hq)d2k#fM%2lmF zWpJ^@%K@nz`PaGWF~Ct_eblGFV})VzR2OsIAFF!Q=Wb{V{W`I_^--T*GuutdIu&QM z@{}c7E%srng9(B*z(eUQnZ1!Q<2KkC~o)jWe z>%wJHxKl=*cK1i^Ua7~ZBcuDJ_36XT>N!g894>>x86B1Bmd=q2MDpw$nJ=8-&XM1_ zOpm?fjN$kJrQm<}$|XNga0=9FL11fi5b96ARi*yY z|J>Jkjd;nzO!^rM#Z{)?0m~zgFQFMyl`?ZH{d9II)R~uk6CIY-dpk^*KV|x#;#ljm zb`$Disao+pD^FnoEHU&&PP<{Fz z-@m2j#M*YZa_aKen>gmY+4$6pDFx>UQ)fJ?<5v>{V{LCas3(%nJZGU&Ju-TrhMlM; zBo9Ff!sG5(T!;UX*>!5rp8YkXPDL){S&!_V^hdDIZnAcKe1C6#*}6bwAudW-N#jaD zcj-$nbn?4z7_W11Ne(uQzf<(L1AK-hqa*ibk7zHE9BXNVY>t^ ziw3m3kAZ3{AYk|l+Y5Fv$HCF8yff3y${DmlebT@NGFWHWI2FV1|EVmeZ~iPOxmzhY zjf^o}Ddi!}O_yVNd14`+_@X}j)-!!3X+GCWxs(@8QjMO^*nEz#r5%Z+x5iE`$H{=z zPA=czmtBVGH9<%LX+N_gQ(N$vIzeaVD~#y5eHfuW7_{@H_nqmIjc_8@$Rh)6CBw$-$cUNL$VgID_ZN7V(J~Mxl08ppYLmWMU6AZd5Ua+dXlZ)+!JcF*Q8T`hko5`4#_+`G^m1Zk zEjTS&@kW@NEgyt=h!N&?_c^RYbh(QUoq#$BWid6A(D0^!>n~lomxxp*T}~+E<&S3R zQ5J7R`p5?LELs6>BSjl1BGI&u$4bkeiuRn*=OMASF(G9k<9)YNitV%I?Np%cRM4}X zh5{wpPMgwhGTILk{-DM3TQ&xMiFuW1(&KYxPxc8qDci##dMmRgu>zrYWy=O$r958^ z$jEI@!_eemm5K5vw4vqP#@->>>9Fb5iJi)WmSqDfTUO^c4B-)y!z$Z}Njw#fS)&0t zF~8+Y0Lhiz=Zmm%*_AfvB=Qp$R&2r8iuOm1)_uMRg~6%BC04~AtSZEArfS5dFiu2| zPAW*9mH&6D=t{^oi~Ly6XES|A~<)l&<)OG9?o?v z;n)JfNn3kq(^(3D`F2` zF?7@PNOt|v72+dhxWcut2;82s{o#@d>7e87C@xIhZK*3+pw zvCx)Fil6gn#X%a+6%Bjp2|N*!_5o@{+Zv)pV>*b6wt%fbKx6nv?9zgAQm(gZ!!xL- zGO;rA6mPZV>#erI8?=^Yzxz3cDsT6Gt8H7}Y8zfzzLK}vo+LT(a^@*SdiKO!xhn;U zRH;5s+~t?Q%u_D!vQ$NZTRVlr;H;1l91Mf-Wu_?evBO@PgYX)Xm9Kpk9Tis6u6mKF zLrx(H{{oc;#rL^!Rj4A7w@_gCyE?{Br(_eoByAKitGP+Z1kiA614Lu!)+JJu!& zcIK754{B252FLSX3FQL6pNzxEIILI3RV@9SbrAj*eLc_b>5-wgHYA)0*G`yl?NnY> z{?vsZ$5X>b1o?}?bAVr|$3%}$PV!F?+GC=H>~MR-k2ma|s%(+#tljJ(!b;xj`7rz! z3jmi|nH3a^@zUy~Ase)JW*?|xZ}on^aF-JjyA@WJub|kDB&(5qq`sXtJ?knlni-N< zg#t=DJQ*+KC7+!)hA&{a)L;DY_xc5=V^t=2rW^JAxI=@) zxC6_VU-&^dne={o{oi2!`t+9-G;}C7K|yKsc2Rhefpr=-rAEtc1S^oQ4|9yr39`3_Ce0Y}+ zt9|$3*2> z`G4ZWNrXM*zhkoW5Q_pl5hmV~m-cw-yw1u=HHnRSwW^bq86^LlW=CM1GHkC>eJ5&P z8NDJ%9akJroY$e!d|X#dmLT!K4B2gQEl(x_u~BUmd@2V_*mcv$q2JH(Tl+jnq^4Q! z>sBizcJ3ky)F~#Bn&}eq*m*(nrXdOvwatMZF}$ToX*`N$+Bv-@K@wKGdI_txQ-^s@ z_*05`#So2DU68CRv-hu3cmm?oy{uvgU7)rLeyz6gU8ZlG8x|)kDzF^CjkjPE)hn?b z?~08&b$)C_2_mQq{-AcnMw~0#hYKE1vtuKWTk`Rkni?A+savwTS;w1-`t-=xH44=a z;ahTk(6XdX>RE2m?c?Np+3QdZb{(n(e+70}xo_o%QGs4IcIjG?e6-w1 zDlQK2XS7x~tg1U*q020~`js5jStlI|QfIB^a^lp=Qz}obJnhsRhy1$0%ni6$n+!C& z=XA&dyH--D=Ro%tRd8U{nWuqwO!E7A$0Wb^J0>agQM_X^EPs6au>9~UzVv*^3io`7 z9-r3RZYLz-gLonPxeGt6O~tp*3-X`q{&}XTBz#FCu5fpjYtM3=-Z3%pRYmM^WR3?t zvI*0f#&5KA@7?^>m`g?k4{=E3uQ)G=-Tg}Jrmm^DqpZjaMt!2?`fq!s0-N)o<;&P5 z#|6ow_>tfE4c*G!My?^`j$AIBq)IeDmiX}O=2{kKvrN$i?+c^#$>jOCwpC4|15U0u zM%uGJJ^52{B;E@-M-({yYF_l*-vj7=tUf*Za0v~LSp1pL=G1Yy&g8H!@a_Cm`qHJx zR&WM;{Fn;yd<|zk%#2{_yCKs!w0$yhONLB6bunH((Ga2~Oip0+<4Sc$Q%XFbNlxH| zlp7K96`Gt#D(c?xw!T@uoEj!pEFE8%*8|Gk(@32FykjVZpE|n^)%v>#)`_{Wk~phK zU4&DbStx^ag~fpu$V7E99w@d&BE(HUlb-wFxUs<$Hqo*1CXm;YD3@pGYgHm zmRYee*Pru4%yqdo(t?SF?I#r)b3KZ0G1rrX?I#WA5p5!LAM`|owPxq9cL` zE7bBY45yyXEzzDG5_bP?K<;;xKOW`Zu225t3;F%!-#f~Gc%S^g@14K-s*1eWoG!LZ zE5i(`^zGYlrf=V_eai3pQm^_Z2IiBmP;eTK4w6^#U~EQit4?AXPZ$zBln+q-V_7%+ zC}=t2OAtd|)!0)};qhG@xU@n;D}(_r}}g;Rsf(s1T$bLeNs zLrh|3U2I`~_fmiFjr3gnm)s^uZ9lycQPr|z-DXLOm^{_E7sP@KDck=*X?*XOsT

    *mdrtKJqr2j% zGocbG%#$D5Tm4O)Qc^6R8fBm$_!;!ENJ+-pP}td!FYFv3R=B||=1JI@pp(bpaAFJi zJucJmEtbdk)Q++Mn!s)r2%{uty((Iv16TBLEL&LyRB=n7H96$sLA9mj!)8N_+Fw_c z%-G5lfme?>DXl5mtJf5rBv}! zFpdB3(w9HRsVB+i37 z%>X2!_j>FZpX+kQXX%BTkI#O7d>&RB<1>cw38&gfS0PQC$eZzDe2eRJQO`R4uFK^1 zbLB|^KA~XzDkrgQ*{fH7ID<79By*pZo@1u7$SGOEEG6g!j<(~8Gqu97ks8Yh+ud;T z50a7xUomn1gQDPJH%M?sN1?GcV!+&E(kszpU{f<+;}UDVOw~$s`HyGwW#7^IC!!^8 z>{6ImU!U&O`6>1|`e?E)9S`v|S;+`h57~^httCDJxO9Agr2l)wxN%Jl8a+2hNx~k& zwZGiFX`{&Tg6$x`QTBXUI|!u^#dG$`t*YtjB{o34tder(Up=VzJ&=D%>(b!oumMOR zJ`GalLvs=)Hpx0j_OmsKXXTDCfd~_|?fyEvr%*J=g9>}?ik+4G)mb@>QGV96Je(jP}Oo87HnlXt(F-fJm?Fup=< z&SwPDZAgE@V_4|XpwYkFLHq1UI~jMVD#lW+P-}#Kk>ZvS)Tg_#lO^)` zvCfZ;a++xgyDO|fF+#Fv)MKsRNARMN|8fIOspiAhVRpZVa(Mv0HU9f8$A8CKeoY1c zJ04ulb$fX57QU<=KQ7z5ey#?raGCm=<~Pr3@&vW-pT>;dxvRyJiD-MNck9R*8!@4* zeY{wbwtHxB_=$C-f_JO?Vf2R;h|kwMfx}?)!-E)qXe9Xl^YgZh&puFEgwMigoYF0S z2A{pkXGC|$m5A()gX(Y7y(ByT(A@{flqq`tbw>(*dx|?>zv#o?`S2|t-bmQ9T|0W{ z1zt-oi?|!n1?={ZiRi~cTUh0uejM}GX7r=sL`OS@n?ZP%XBpdzc09UDv?Ib#H~y>- z-B^SZwLWyCBYn=Fpoe#7`=4D;-MRGRnsYZFZxq|c-(03N#`_zhAL3d1($$w4)1!#D zc2%DA@M#Q&C`8U94_ANK#2Sewr^5s7pTp=lUn%f9oV62pGI)Hx4si`XkZz|?C5#HB zFIfZ%GK7=9LPA_D6y~r{g>i*jb1?24STr!MYU^zW8twmwOK4qG-P-%?%y=VCA$XhN zJP;HHEe|ULgsV~~{OUE8tW zk&ZxANZ7T93LO+tq1J?vL+S6+H>mu0hnCdKkI}}1Aq3Cx^4j$4#0X){Sz;Ft)7KM* zpR&`k{DRa0R~okD9F6mW0V1`$LD3bv;aO?4AC1n%bgnHgyov7WK6~u$7h)}toxIrN zv+^rt#q$>L6u>j?QkgpGrRFck#IBuBs>}Bqmv4aO`)l8P^RoGV>GBPg@XJAvNw{0LV$CJ%|jRdQa% z^hcTjKQm5?47tppZ#7D}Hhr+8cp}))o1F==bIx=R?$Vju=qiWy9J+emq+WCH*T>rQ zi%;~Lcv~lx>-@xvVdDhAqLs=9qF?kH+oz~)e)T<%WtDo zXJGm(I-I3b6{KkVv;jq@?7N3f-AoPPt^6mx#(I&kk7lL~u8v0qg46_9*aoiON9;yo zI}@weK1=Xigfmf-d3}($h$(?x)bXrIAV3Js7LIq|wS9|FtAzzjhZJpvvvoR2 z*9N~dQgMlcj9My4bLM4kNhkGcUbH{5 z+|h`s5&!UlR_kg%Fa3{`>v%Lywi^15)A%7YvhO;HhRzzg`RLWB_wa2sF3zo=>^8`c z_G0_Qir9#WWzmufr_okhyY=19cBYE)t;sC4xQRZl%*BPf5N6$3jA4A8?Z`*`k`nX| z%^z-aCKeLUUXke3+*yp3c&mI+MXQdSGvuCOuJQh?vy06Z&n&*N( z4E0)Grs#}SFXds4l>tR0CyljB{2{LVP`;Tn?Mr^d)@>J z*HawK8vaISHKZEHec!RN=lk$PAKvA|W*;sfjQAn=+4YFQxg~)qLck+i_IMe%<<%g6 zQ+iWErZ%r25X^-jr$nYQMxHqcV!uy>h;BVk%YXi98v$4 z^a3 z>FVBTddFyqhLJx7IMzbP^;?wRHl5Nu*|}`ePDR8*v6K{1?JamA|JZlG%77-gh%IEHVo9G6^M_= zKSY=E!ZVZ|<|R~CMkFusRek#Jown>73K*s`n5!uM{+#%cIq`FS{Jr#IUXbWEFK|50 z(%s_LuNS0gXS8GnqPt-B zlB$#%-L*MCKg*t^*O~4YbEekj%?^;HT;*QUF)<)-=){wk7-Prn-`T%l_~OX}APsD1 zJQna;#?Sn5t5|J5j{Y{rX8)dnfN8GfL@C1{r>x=tq}u_=#Sqlt z!&8R*@^XW`?8%0Gg5*IpEt{x*nGK?4=6V$3ny3>_KniR zAM*unIO7fBhPBki$C38&wpEOr=A@cs%gB}SGG3+*5+T%GYUlodfx5UDnJz_V<=e8$ z6kWwsVXhdA`hYbTk5zwJJ^J%*ju^JjUv+D9)R#d%YjYA=?U@+PG0luy%_uxOZpt*w z!qOVH=V_v96JOVk`V0?@ncBp^YZIGlIE3SK@6*`qMWmX7dp+_#ycul>@(~)WxD8<` zwZn9e3>JP++`6jqWF8ix$eR~71ER()ODMODjRvU^>q0zWM71!<%p zjZvFHY0i~8vr(V^K8zxd^cCburVe^y_B423WMa2;so6G~r$`!d5nEy1g$DUn9^9}#FosPxzDB?cD%xb6cK-yM374kF5Vs~<}ysX4>|AZwwhaxIF=3IW% zr>m#jidn-|%U;T51^jqlnwLA?_33ZyyE{`<)Ive*47*xaG!B(*Mk_Oy9bbkwNF6*B zg@;M7+Pbd1B3S=QkSe&QdtaN!<|{WHfSijG%~yWSFO_b-awEU46O@*Zq4G-`Dx0s& z9MJf@c-&DhtKk?SZXb-o-v|@x@za-d;eH8;bSsn53?LT16>Z53XgEQ; zuc1O5RO90Y_@VA!2bpKG(7QXeUq43R=Bw-eslM^w@4rnitdh6^$vmSUu~u0J@zMQ9 znirnJM=rtrAJ&%+V8Vx5Uw2n;b$z}2q+4I&$7zq28F=LeP7v%!TTip~+mOdUp?y`^jf!p*6plKYL)G^`gHN$yXPeqiyzcR2(lZ2 zS8#03mQ0D1WIgANn#8+O#ptqU{`NEiDGahqtT)WNohUM-NO%gAmwBrBio$%x^tT4? z%S>Y!l)rT0MFARdB!lNlDER);f^#NjAuC z@5b8Zi@9o9QWw!Q!(TOt_iTI_gmESN6m1LYa^*%rNW}&LL&K~OZ^a-yB+ya<7sL3+T9dZ9s`#>jgn0AvDMoTZ zVmBbLBq!;;1$?Q1GRc~UMMEKDcJ2s2Raq6RpQX;f`dNx>82se46Yu!>57{jpS$JyP zQn-BaU7EMlaVP@OLcD^x(M6|^{v2bY#sje|YW!BqTXfZo_!bnDwHF*3PO>5nUsc5( zEGW|z5DpMa2XsH5EK3~vcmeIzDY8oPEi)+7sE&Fylv`AXcPkUiM|VkT5SjfR#H$Dg zSCw~N_%SBj1-vv8HfbRh?X$=&i_^#J(C**Q*6Hnq9MiBFz1GDP?Yx0s^{k7ZeafJr z0BDt7RfZRZ&}Ki{2xigpf3BN{42w$epu}1B&ewNTZvQJ?mi#zS6uzORdw6@g@4yCJ()G0M9z!8Z>Lz_^u8orA?1#4h?^+->g%;_wtS! z<^9J??pL4w=qdV+8Bm3Wov}w2;!G8nJf_V_TzE9C0-u<3584(UVBcE52llMr_T+(n z{f3Rc7d^w-I0GEE=kr2cMLYU)iq^t^H zGy_wa^Gx&)HIltDmY8_=ho57DW!*{p>TvVmY?RGxIe{N+-b%Nkv#sa^MJPX;6i-B{ zi|Y8zQLe6-!gUlbRV$Wa*Ngzaf;<&g_#;A2IB9`wkm|6{Wfe0Vh?#)+OEbTNWV;3Geau>s^wF+- z(t>1ryYi;~GG9R=XsDe$d0w3*RS&(d>r@X1Z ztW=Ol+E=2(a;7>ga#^z?^|!QHWmG|zf@He|s^BxBltj`;dsr4E+m|bE>M!dQB$D=( z=&+or4vSpYtVsPWZB`jouv$T~-2zoG$MwKRdsr4E+tbRMNV)(nvBPpLYgUB*mNqLR z`{`wtrPw>!58Dd6)ae9%3UH@Wm;*_ZTH-1sMzIA!VOA;dV2Rt<*PjEA>sa_5r<9YWPLp)68svE@p`A0x2~9nk-kb|inw#H@16_@+CX>m11g z7Z7=Fb0linQSK_okJ&$Hy7-o{0a`NYvQo8W{DvnONOpLFp34^$EhY;*!6I)Lqz5yS z6O|vHU}7m{>-Y^%pc@>#%&=^HDc!&lLOelj1$p2JD14)^f>Ji+*BLCW6q1jGynLOemLSvit7 zTOdHcH$b=eXjlFg3z9dtD{m^LNeL3S*jJ*(a{6bkmbEBSe@m?{RnW;76mU!Y9uERH-J3LR+oP#!|?;ULo-GE08$v_x*&0G;Y`LzKzC@)7y9mdhvrnn z-OYjzJXV~CcOMOPcXHN>*Z}VW2d}JsGW>wx{a|x=rD07kcq~E>?`K(fA9#4DICypK zlcC#!ckJfyf>8SPEZx?+^zg3A!n@kTvzix7oY_7Z`YLz_Zw_xpxMwarzbZXQy%*uS z%q#HL=Y_;+CR2qki~=4e%~ z{;*vKL=0jigwx%FW{OBc2a%j4nJKg#^WBkXY;vR*I1Hnh1Q^Jy-S7t7LQ=7!JL%w4K9spCHs-3gGKRh|;nC)0shp^EfSUud-} zG>t-#Zxi<6IzW*WvM^gyUHZk>)jWT|_ZZQ8r^84$PlKRF-Rm6IsbE#fjP#wZRHxPC ziEpbJFzULxvxP)Ue~&YRbftpy*{)QVE45q!Et0TOR@*Y>qb|7`GWn$|H6?wJE4A8{ z(%kC)$28?;TIy69*DwkhjPxMOZ8N0VK`X(?n;dl?lPCESNGQ$Wy8|O-!WVZ*NcpI1 z4w6qxwlhpsQRMUU5l*|D*cF0UZ++@rSr1MbA7Uneo|8s`M9!fSq)7-&wT5s zgoRO_F2z`$Nj^`9&r_Ps)0xdvW_eh8@$!u8MRPW7Crfjyv`1dX?PFy5Sm$fa=Llpo zMaTZ$$+(B?=j3_c@!>ch9_hmhA0F+)DF0BtJ^jwrM{#-LG}PO<7Y^7E?q($-=@K5_ z{wj9KED3rW>k*7r=jkC~d!+Xi8k$%u`EHPShdciD=}$LEo{q=5=<1i34wn8(6WU8K zc&QWqt#y42FU9hcYZ%3=_X38=vyUn({Lje(?uJ9=j5h2a!ScfPkuyl)33k&*iI&6_U#$3lOl@% zCwy$mK{ISYX%hbWHa*Ykp1E%a0*Ea*x1^w=^`piE)OXYmS&91^t;UumO_4cK}CN51!L=W~STI!|cBk4AaM9oq6^vSh8WKI)C)jRb zFDO;o=cv+B2jhBmQ;@&REE6xz6m{LJxwDZ_t)t#j(*YXuK(PS;U4Ks()>x!1UkgxiKx=5Pv9~f5i^F$D zxgD<1>hDVrMc5((Q3xBGOb?Uyp?E*^`N_YJ^TC?cA3%O+9qbP#DQ0MF*Ri z^&sWL9O`ta3K8#cVh z5m76PO5N{hs4Equ%Ur3|u9PEzj*tZ@ysW7-Zjls<1P-v=4U9f}1b^WNu2KJdzKA9m zy|fLJH`ft9YeYmBttFQH*6l7iPqjpp&AMmFi+r-$mZgg}W|n-nPwrCFqtYw&VacsN zxl=Lh;o>Ykvl+KM(|n$0pJ!w?PlsYG&p6AY1ql!%ie@J;)1~fiseP`bHSXZAqxB+q zC*H$OXFC*p*KLRY^5Iw?9_GVwK0MNgQU1YvJNhxP`3Vgf(7uYbzGZEHVzo8c$qOLv zdf2;XsG5GEfqtaR<*_mHdF;3KbCS!`>GI51;PcpT%Twm^G`l<#BFG2rsk%j?XMuRehAvLMJU?wjd5$+)yd^Ceq@bf?SbYdVx@c#iaS z#0K11QPXnh>T)Gty+l38x!H)m#$N(TUGMy+gNd}oy$VO{mZIUq82td{43P<3^;r6#3Lsx8z}n`1L7rSBy*1H+lLOR%%~t zZbRG|n9=Kc*LOY?z0;TW?2a3yHW_CKdsx%xgEdLEHm_NNkFZs>;Ad9Q4O`m^U(n$S z-+2|(EAJXd>!cN&MZplWa3~LLy^>pL%OO_sV^lMGh4d(m_I7OJt}Sbzq}|H6V5gzD z(i{@vnby%O=;^SbINThDmCV8c-7hAGRwjo06=|3md|(g7J!kacZ*Rc4U&|snG|EF6 zkzZB5>cWrWsRH@V>%J*dG`Z6ufS>K@5P-s5dYE_-!)mk6Hx~<<$zJPr2oZPzlo8Hx zgLt&qZK&3Y`c^QOuVW`zYE9rikKUe#u$b|1=)Oqw)?XTD&ktvjuKYMZ24oM zc+uRi|ZV(!{MX^Afxd zyhLaXhpp4?t#+E4E9S0GU)Z6p`nR`P{hi7|n`gi-X|W4+#Yyk7kf+KlbdS;589Q=5 zFBcTW-u@^q4s%io{tr))`)7X}Z;>N3T9KevK7Uve64}PlZuaKh#od{uqbSTu* zHsG!T=macQ922OqLH;_qZmd9y1z(!K+{mYrr^NG@$MaX>y)mBO736P##|j2wR7}#0 zIEC^tcV1qT|71=6i^1Xh#9F`2e1jUXgGpa>rn#>Cr(*KEu;7W!?gmkSl*|bJSmLph zo;Sw^-V8&xLxZ1hHbLK1A+4LUg<8s=7QK+h;aP`rgE=x;KPaAH25COsyWb5y_-yU7 zY_aTxfOzs8t^lNEzZ<(o-baHZmcfk!?UHV*fFs$7^B(u?1B7St(+?lUt%d1{!le}h zI7SQ?9!y`OivACC=K>#9b?yH|Ug3Fy5JklrH7c>F#789x znm{5mI?<@0v_48(HMCl#kSMJWBAKZ<9ml5CR$Fc5R&Bkv-fCM*5fy`Ccvux_eOHMu zJY#soH!3jy?{DohlSx2(@BRNjpZn*dIdk^jXFt|nd+oK>UTbZ#OO<{%>7u6J3lqmF zQ8qM6JU2|7phRPQNSqcXA`sqR_Ci?l$S|==iDGv2`MzOdM2TCDq>gdl1v|f$GB4dZYl5znUaY&i?5%Dvq5_gw6MmFRknKz%?Vb3k|d6Q z#Koe;2B_D)g@?1u>#$8`3yhV`@)nP59nHY{$tQv~Fw82Ytb}T_{hkn2q*RsYt^YOx{fK)-qhF-EdIL| zcP}!!;+{TaG<`p+z&3{QALKxMW$;#PT?$8I1UP*IrL_M*foJ0_V%GBe-J;1NA$`@h zaG$}x`x%@t9)gxr*1_kco#0bVWm!E!`}}W3h(!)TN*7x#&1!s!$Dj0zIXEu0_sc2Z zzH{%o-#q{Q{oFf#Bh{?~({ECN?N8cKdrfuINe#iJ8__H5lyMj|`;-6B4wBfktNh{Q zzi9avlAkH5{)M2>xB9m!^=B(};x22!iE#3jmcO6!8GX}OV~*W?bFB>$MwS0A-nmqFy_z_$+aRp)?5hBYLBM!phe0xrirRkoJ^hkY^rW~>%{EWcdAye#`($A#^3Z3 zfoQbaZdmDRVy7C=VvQ3imEI#-GRWKcb(|V1s!KH#M_0NqXm$P})g9AhlKV-kTILD{ zAg2x3(Nbw2*>#00k>qVeD*`BF&bBA}x6+Z_>tkfzZHW%jS{PNL0+3+cuEEATGe+7h z?ZfQl=PDovp7jmEwbm|X!dY?=?xbI1H@?vFR1>>!#CBtAi+Uo3?`j{OtxXt8vq}qZ zrZ(gG?76ld$Jn>?x8iUUMq2~B9WQGZWD1|qcWmZQD7EnKb;lp##qVXJb)hH*&hsud zOU;#F1v@hCciuwBm#w!Q`E5!z1jkBu$Y3V&5w??FZcqL#*|sMSSB;f_YaD6SX7}X# zub1$o@BFW0%1gk?>FgftNSZXofjDcGh%%ayOv!_7c;eukx692ZunrH3rp8VHM$yU@ zUR<%^qe)B+D_`^$j)3TDFQ|+Zc0>)$$C6#+uLQ3f_ZHBKD%$-=Gwos<_qT-UdOHJ)0`YV31_#Xh%~y22C5>mSJ6lwhj1_*hC>h$j`>^j)FW=65ve?FwPbCF zE#g%ZUxKUfmO-*jW^wd%h-40g7xe}bFYRsNZ4*$W5il?jOB(OP)Oa_ zpeMLA%=6RxX^&()aPML>GQr4y-$GoOLFqGk^AEJ{^54B%`Fp1K>dSu=yZmP?e~`tw z?cqN(t}9rc`QwU@GtF+OL&R7anotb!Z3un^1hf}*(@#jT9=W|mr3gSd`PsX)UAPOh zbkUI3nEny(8})glACTuwQV@4{#NL0;E=RV7*u%OT+djzq;m6EIG*t^jX+jL^TQk?m z_qdLEaks*_!%7Jh!ANNa2LPQOR(z-w}+>cJTX)9)(wtxs!BhX<*OW4TmXpE zZ!t3=9RhvkH;F$2lhX^WY%@)V{Lv{nKFaY!f9=B`9hZB5!1$;2x$^KwqP`PbcS}zv zu}!f-glMe;g&I3Z+6ncWj%Wxby|I}qU5Y4ZEUiG58b0h!SlU~E%Sayrk(^o%fE@iDtps0SiP25lUSqqmfwGu`M)E7rscm$cXqD-I`Su4{)%1Ydx>uu z+D_iKNPSqOG61)9lv8P(89q-5pHsu<@!@kw`22GCtl^1&LUJ;Dj4b>^Je2D0z z)>A}xmE9A*)#n8C-^H_4^eksoCsncXRnn!@J9+#+`7WBO#urhfh3bl}TdM&^UEz~U zj%Y*@uO>Eibw}-Pc<%GPdSoT5oG*!#3GiW&j{Zdj!G{1|c+X0gmIJlr8>D1Io?Y=e zi%$V%E^2ECPJWhKov;&AedC}l+ojy$7;Ag9@>v{>-oa%ReZd_XZf*b(`d%2wPi*1y z0v!DOEWg1u6%P8$vOTs=b}xn&e%vlPD;GE9cKYk;A^;4bF1+`3JpM<&3Qn#f@`(+QC2(x+fF_Z_VrR z9`s++$hQQAS12tLd~&&{R_{~Mp&A~iqIug~D7WRtY-qrwx%t}8a5P`l%S3Q;vsPsI zHVL924OD$@B3q7Cl6u)AC_e>B`Qz=kGSV$U?Qftehf>CQHEp&!EuWmpN0KqDYZVWYE0(rNPpRAEefU=V*YE~u(Aw< zqv=er=HF=gk2G!jx~nG)$;rNK@?~ugf+zN)9;yBaEx+3E*R20@`=>+!gaVvyIU5H- zGbOLLiZ~9>A^~Kd`F*&dt9Rs zTpNPcS41Rx&+9MA&ib7H4Ol}u9{rU&@;q`Me9qf!DAAShBpMb|mX_>I9E2aaUiTcG zHzB@H{xx+FHU8DUEj7j5n#70nR+Ttv=MuQxY1O7Cf7eQkq(r7<6EmABdE5TJW`AF{ zzt8el#CqrmN33rTpYIQM??2|ewDZ18NKtbfgKaurR1&n6y&c(LsKy_+Fv+GoneXjrgB(ULL7j z%i%d^=pTCvV_HWK2$=j^AW!bXAVpwXUn`uo75U zqqo%P3J&F?l}~ytV|iD2phFPR8V`u3j^3?@Oth&iWC99tP|8sVe2*axk1%{;0=nS~ zkO5QYmd(oP;uzn#+6cDdG@`npFt$vgqMT3btBDqAP_HwkW=9C7o(4IrX`q+_M}z{nu;6PU|BMS?&ChNtbcZuPuYQ$YZi2>cX`vTB$dW^hp0?-i;j^ zh&%$;!)-@c>7l@ah;d2f4c>a zirnFk4HzYZRHx40aoz>y>&V@NSyw|G9_=QxJV!)xe^dattotKdIu%xh_6kfw{7O>N z$I#jA`eyTj3%kGyat_nSFY(xjKj|pBC34a+X;iH6)0~5254@UAjQGR`P<(@JzX)Bm zKeE+y;hu_baQhG4qRsna6lK#^WYJ4na92?zQ*yZqvb>?roGlOP&N zmHhKIA|QMe`BwiwMGDtuIvxEo~1(1qdyw{S?Xl5Dugo3UOdGnJh`LW6) zmKStAh5V8#xm52~ZkfGbsQ1Cj|AW1c(>r%i7j*sF-Ve@|zro)3&6ZDJY0zz^sJiQA z%Z1LY+4Onpm-W{Lo5`kMp!8vUYk_^Ue0ZwltGV=jdewib($%jt(N37vY(4e4bUTvE zrH@zoP?bkJn^*quT>68((htbha|h`RsPJay_@ngCc+2fCsQy|O6csiO3jKtM9^5?f zP=Z&5@7c>k6cvYXP$<6@z+|u%m-*8ZONe7_WFL+N(&toL=zj9jD_-7L2h^;)jYUf> zo_`W2&jrxPH8WZ0En#ehw!ZOSs;5VA5N+uwo)Jl1hLp%@6Ow|W8k5& zjlHpN1(S*siqr5jE?UpI8@5=%I*2Vjz;`I zCr7GKXR`%ecbR?x{)eKjJ+H6y(?MD$_+ za8SW%ocY#OE^nL>^H+5ZBHmMFd*fKHBaCF)n?LzxQ#S!sRIV^qzB=kJkCv|yi#sUh zkBIqPT%xG%XaS!8rY`4%#mQ0jrXirrn8#`>)^!nZYS6-*Ke4xX(4^AF+DVfezcOi3 zDtqb2HJkKZf(w-l4iLfd)Il-eLIAr%+JQWVLpEM>y zxAPJ_h`48!PY{~y@7jciWJ*R~^&HoVH@o68+9YefsvMLiV7MBG*7&` z33GBVGK-4iH}BRL6{iXnaTa={*hs=ZzYj@B9rF;>mMzfmcb~n@IHY@IHefktz4_SO zI5bl-b%D0b$Gi6A+tWM>rb;?bmzzK56My?5Rq98F>hASv6P5=U?Y#WqEqK%;PyBbs zv+W)&^Zd><^%>7FH=d2Z>>1A~d)atOzi2%3G_B!y7L2g*6lce?EfJ3AqxqjRo}vHK z@l0a5eDQcDsm;_eWAnyyHe)hgcT@W<{!D`91?LT3?F#3c5|;Cp&+~8Z|8npZNGny z__Uk&sA&KMs|ibQEXY9^EXYe-XpLHS;JmBDWbaLg6&H--*e~k$%k!(jaTG-Jk7@|^ z-F3e|`o0@@*TeMqC%34_NxBbTftl4)HjZyv>L|wH_(@`@Ud=AyW%;QLg4FYkvuZFB z-lN8EZ|U4VsTB66x_Rs5YMwQN+wU%He9;l+=jYamv=%KAjDQpVOq}vf{$w6d50k%?RZ};;PFzepxR7 zdhwsBzXIZ10gM~P@m)G!(R(%=I*YYip7{G z3dazHLX8JqAt4=FOP7!1DxU~~*!WL;0V#u9g_N0}#co%pV$H|Eb4x4S->XaApJ z0cdK{v(zW#vl$shf>I<-n>Pa2h6PTffbw@J|Kx0bYQ}cGAD4Y^d8}Vp-gCoc;18oS z6DoL#5BZh;gL@_8biPuCRJB}jt}A5HlSz|e#eY9NgMvLv&*+8L-@6zp9$9@>Zt*Du zDDgjgF$qK^lwhLkCA1`eDOGoB(n1vVX(<(Q{Q9K*z4%(8(c+T|3Ds5zv5C(I<0nn_ z;&b$z#Ho^=ClF&x&uhH+Og;HdW=edw8{h4w@2Zr}cN6vGy9s*ooy=DCooq+--58#H z*N^Y|>ANDO^Id_Se76;()OX!_@?D@O-*xgt)FN`=$DHL`B?DomN5f1P!%Po@nJ%Jy zLZ@|HZW1luB-O~m2l*^F-dH(Gk*6Di6>we>iTuAWoy_nt_cbq~?O)3#D+nqv>j0j$ z&(VLTYZ&;UD;455Q>C}~=uuV8J4Bp$Os?{7>L9}1;BIemJ6|<&AoA&U@4AaAi8Dyo z$dK+fPwK}=_oeeX@XxH`$hoPDpmfYUH`s9)0@^JeJ*xTB6TBbml(_lRQtyWE^CPm1 zZWioTTrh|GGN1EWE+h?&+1$O;5(jhRi)PxTM{9YX!c%Chv~iQYzhQ%XJ?8f>J2^F} z-?CEPrJ%(k@X<`k`AyolbY7_OKA^nv7m(KQrs_oOnOl;+gEXy&Xl8kx^v??KNBu*s z{#7`d^B(h8n-R_1Iy0lPAfvcu5+sn3&V<2@_@m1V)eX2D1il@dt0Ku^_)-yBVQtq` zG?{!6PF>5XfJNLA3~MudFJ$OqWSM`M1v(gvI+HX{0El&0^27Z?Mt7mOjD1Wa(h3LBfE z{*y7TP*rrEXsUikT2@{t>$?UToOjpNVwu$hBW~i}HB(q;p&k5OcansFeuw*3T!xWH zvx)3q7QD}m7BLKB-837t)0i|#&$2uFy!66*qHJf~5X*HS6M(PXo@Wi9(~Wv5D>`8lT;sgnmrrPtRDAza|tHH?fUWMper0sH6^?p?8xGsBXtQt1&V zA3ruUbT7@#4^k@ouupO>a4j2WoAqxmQb#yFqX-sID>bc@z@lEfi4|rFcRP@XroN31 z_Y%&x{EcoN4Xf%nquCns--@EH?T#d8BI98h!twq~M8ADb67TM4C9&50(s#8!CY7;R z4+LMdK$=-0oVj1UkEy#C&~pH_%Ddq}8fFo-?xPOSBx>Ndz4+s>L{Zk0OsXbGwOmz) zYPvg`I=8Gdie_1+tf+MQm+{R73VXdxhY=H1L2*}|^iq|tdT~3?ks&VOLi~BfgsEn? zBq~T^6Bkkljr@yISd2oWWAzbzytJd1yT z*}Qe9M^c&Xi*?e}zpc|>8){Kh!OeB{+fNl$Zt>!C`JtH59WvF5`X_pOFbmb*$Y~Hg z;P;-$f#%C*r$^gw@H1_G?=XmKFTz@z=_`)5tYF~hmX&&m-?3nYSTpm;)dP$RGLs+S zP;|5hHD04%y-~w(Gn$%S)@?{&tr)s+C}KF0!x(1GyFmxJQFH*EI&&+3zH6HCEnL4k zk4SDD^wa~mIUo$Tg>mkq$Z`ZpKPBQvn2t4QAJImJ=n|B6pgqvhh;eobYM0{6WIdY9 zcbo2k5H#+Wt23`Yzly3J?OH56Rde7{EyRs!+yGQj0(uJ@`Qbm8S;I-s1kkUhqDMFQ zQMKRQ(vfq;fDe>z&~=a;Yn!-RAS{ip=TYN&iM?rA zs@(t7_!dv7YOSymvR#ojx3L&)ZsSO4a}Db0L=POQ{b%H7`r1~1c}qJskVg-T`i;Yh zAGylnM-C{!3nKlg)mz+(XM&K8TJq=ZHU{Is#wnJ$#YQStOTDaKWZhx~-^KllTKkVk zZ`W}DI;rR)f6@p!@tEJN4vM>8Nf!Ad%$bKT8#}sQ^qbb^Q#5 zt4-vt0X>>(0oyviBH38H4IoJS?10crmogwJJ+&S$cZnIRe#(<0rb6&g0QON{+=V%`_ zn*}fxQ}ycySg-!u4;&$yoITV{p9ce=y+%Ss*Wye`E11?}Uw^asMY_o%#V~aWd*1GO z{@?0fLvS#Cp(RdEymQh~Ug&z9HaMkBFV$<$_>$LIee1;Y(raQ^Rz^xWW@u?IT}nfA zucj<$g_en`1?p3GUUL?I3wV?^FW|n$BO8K$E@4FSmZ&UlJkSQah7?rrW6X|={Vr=C z14I-6UDQ{RyXIN9kk77rfmTs-xc;Kg=1@Fkx(1z%OQD^7L`AY*%;7aRr|O!U??wE- zxV)5<6A}MsomNtjyj)BsMaFBjtm8S!DbBe#4HA)8NT#eFhuY}Bt4*TBPBn1>&T`3{ z32Ttk?dh75Jw-PgD>5a&Z_pMjwe!PPz*UgF4ClJ+_`t8O`@yTwPdrzivqqy=9G8S4 zlpQ=GscCca;h%+1vyCZ)&w@F9fKx5tT);IxWVnc5{*3wbPGGSw*|?iN_a-2>9~nHT zK)~qzF_7Te?QH*Q=jf<^nt^u@u$s-tUO6-buOeB7^{4}sU&?aNUwdc<+1u`$V_SXZsY+u2S8Kl(-waTcJEV41MX zid!a?7086SQMRwhGaA5Dl}Mg6ChFJDW*v^p6X>$*(Lu>eNC}n3v?aU0X4Ch9exy(5 zk6Ska9v$-%@2mXzJEu=o`X5wJPB&`vm%45%6n*MPT4tcdkpZ1l___Op)=@Hlzw#C@G%)tS zJls!9L2y5<@)A^5YUZYMsZZS;5A>r-3#6ru3v_BMe?QYLeA?T60w`DV#~Y;n&u|G{mfYMcY#nRdAlZ@#oTzbTW3993t4)He?!Y-rH+fg_^Toh zZ#2CaN&ZS7>a>R2JpaOfa!kt=gwUK#ss zWim6Zo@Q63Ew7CIwlXhUnX6|-Qre{~Dgr+&tKDTGe$>`yv+7;eU1Y_Q7mkZ0&nwl? z4>i2bMM=2C{CRU@zJk(}YI;S*HDE|}(}e73zv`M{s2pU`gA>Ut7P&*^lAc6n)*gxa zjAk0Wdsy@C=tnmFN>HWiwl;m5M`PN;^`_moKMJ|WIIs0TdqgarSSKvmFS)7Uf&B%X zLdOX;4x(mPe8WN;OH-O5MF4I_ZKcwU8#j@ z_3#(A>p?8{eG!p@b`WftYlI`@ub-{?u~x)t(K)!Sww`!Wj=9qv)`XP8U++U{nN~Dj zWarNZX<@)1r(o`JiHN_Yc9J+*U#X1{Zu1lIFML@=-{(MpVeNd%N@|jPjaDV*gP9oL z@Kzu(+1S1CTD&S$7jHIh1P*4Z_j;ye+p>d;Lv>${iySGMis(MdTYPL)tM0z=5^MCC-HSwjxyLQS#U4XQ8D(2p z7SG_*Y>>}4Hb?x4y$0xyN*@Ajdg!zNY1cb^=PmYHhq%j>io1T-(_aoBk~N+r_#sRoJcaT|zvcr|NTiBqtl`#b9IdSpUdmqld^f1d(fA zkvl;7YojMobO@p-N$Qx*FF5Ep(-yagdYB7aj(E4$!RpX&C{|(1y<<{T1z1QEz)4{XKuBKNsrJND2z*r2y_3d+F z{y=*g7kN2oVM9KWS6eUqoln8Mae~05l+5uL z;BQxN{H?D7fwTGlFYx!i|1SQT4BCf29raQEe9I5cIJi_~4)Z{--;9 z_}YJS{ePfduzyv#^`GK3+`r_4;r1`-G5YLZ3TEN2v9S3rBe7m8|{V_JHp}A+jP@%I6#{U+6XjDcr>>i8%B7V4!c7nkd0yT#by6adygT6Gjcj#q;*0O@R)!zmVbqH|9uIfy=wgpt-Wco6 zA3q<>B=A(_@jI>Pt z&e_iu6f{jFpa&K`{@F!D&1n1qa^N^T-Y%DfnUf1RK$(XwVug!f*cp**-Dd8f-QM!T zt`2M8^796PKMEF@&%KU6Z5W6JwoWpq?tA0}HZ^>W@_0WiQ(x5e!{FjW-pkm%qaoZe z5Gcyvai_4U$w5+#H`Duq%sNx!w^dJ%rGD2&0)YB3JNSsV>?Af`g}J$&ELp60i?7pa zZkmSP`6$fr68(&ycu=24QZ0-0Kz*RtllI$m7~7`Z6kb~)QO48?K)IQ+Xnnl+eLSho zLl_F;?0N2n3$|g2N5R~QTf!7gko))|{OzqDljxe{=`nC@0*Sc z@%<&`@$Ej~d-_RwXX~b?KPS;2u=~&IcOT?`I6IIG-derZMee>v^i$6C(H)f9QHgGd z{#X_c7`Ys4=FJOGfDBRMOstR~fmOw^N__auPt<3fom?F8kCehEmOKM-9|aaZNOk@* zF%&8h|71$^wn6*Q8Fg^ZkX;|K)Ynum{@+h{afJ$vvZ5l%uN9jPX9J*gJa89Tu*u`F z@2)DYX`K>Y%OZ0<-Oh2}2gD5d=6F@NufKa&PdE=tetHGk@G9e1Nv z>sr())!d5eA}G|tXdg%;BWqggi%H(C@vK^Z^%O3G8@%J9DSphOg)@qfuQl$8=^`q% zd8zf+RA*M7nml*p)>Hj+_uQ&_=k6l|0DPHh{l8bYJX&;ea_n9w`(yX3tz0vAqpIn8 z03?eu1hyo&qZs}BPsjYns%0zFvKu|Bt$4W_8;QnYHOc*I{C%n)>@OI|>|q3xIDS9o zaQr?|jFU+#Y#cnrTRdsMDXo*3DS=5#*MaFS)diS}y~QU~wN?rfyu@!wbWI8G)DA|o zR&bpP4l8RNp{BgVPe{~^XgQ>}pp)|7txMJnmHN-issB0|3;0nH+VPy09M%uz-=NB; zQ2wnTuf~5Glp>th34C{Mr9-LaTK|a(Vv4?j%J7qxmBOhJ;MCxn%Jp+MrN0hJ)cDU< zWLn-L|1+1r*D3zkF_l~9V%#y7z+O+MiJeNR)f@$SxFj?V-CCWBI*sHf%u8OEi1Iq< zG=Eq>>(VQlGyj?D%o@!8inc;nM$jeMDD7^oK0~Xog}pj0IlAanfAomTrx}Y*0y8DY zpZr___NukmjEtC?j1_Cb_t=)k8(fZzV1j?L5Mji&seWuIs!hYiwC+u+Nlr0bsKV2+ z0#C-$icn1btgM1>Z@kylU{aG^WTIw`Q*zp*4He;2r6cEN$IX)-gsqYI;Zd zoLspW<|Ax)!rC!4t;UWzSzB!8mTbkueXT8~7L1*UNP-u>S5EkVhG6+G(V7zj zp;Tc8X4NPP1)9fqEZD$|U+h-Ud}V}BKH{XqOB_Q(LFoiIim8!8YbNQgN4dzm?9Q@g zDpw?>HlF&25#$8&7xB{@`K?laYL!`G7l<+38>A+B%1_HD$$F z*adTFGXyIjO-tkJYGs>8qDz?7Fm0{K9dvnS)}5Gozz`?U}~1F)O6MwSr=jhhck;hm4#Mq>G!oROO|`YEKzb(zI2h4l!U82SXWMoRV)Hqyv`H6sDHVYu(d9fom3rN8DW?UrsbvrmM56=TQ zh+~OP?=6>}c>EWgo9<7pxg>_?rqlAzP5YDC``onY=B_0UZm9vQGwYBC+CiK_CPsY5 zO$j~raDC2C&3?#4($Ze&ZvHrLnIQZ9SLjM=z+6=;l-L|g%|b>gBR==}WXUC82Bvy% zvr8HBdC#U{@Wi0d3wCguDY<8AZ~H{W=gK+;i8UV6yr2}qy@SPtNfQU0hX_ZCbR$dR z3=+}*dhuuY5lvn$VRwuPOB{9Sq)T_hq~ES|z;XOt9rmfEQT}ib{ieomwQ)#IrB`!G z|5+o4d{lE{G{pj|d>%g_s}^JHmX*kK3W0Jfby2eEMSWtnK8p9Esfe^ZUYa#WYN8ai zH{D_F=s1}0nnZa5ERpif*gui4eqcB?zD(V?xQFI^QGeP%Co}%e^=If#{n32_ef#s) z#9V)9kDhQYys2>*eX;rMP19o+&eQpj*B2bY4j;tcuhlk>~(VX8*UYfK+A;Th%CRJr!VQV9`8NY*>3FGj^QWd z=C8*?^4DV{nV+*B4efiTyA6A)W~Quqy0$Q4aNMg&Hob^HZH`6vusvnvwD}+!!OX2R zM*sEvKg*_X?{h_|TLQ8>>@SP+L1(koq7w^L4xK#o5BiP?H3>3G&_C;qyvX;FFInnA zl@ES#=$1_LgKPPjX*wA15|m4Gf;fr%s1Lrp^;2yX!Ht?C>9yJy-vS-j5*&%Lt7~mf ze7z4$6~+>dxP^P75qG>W2+psdM5WQ?I|i78fHs^Pen)#K2V==g+1akrVzF3GTI)W3 z1e@yo9XKO(?Ge)1y7IS){azZ)Jk2t^W>%fQ4OT4r%(gnbVV1u&psw)4C@Y(D@NDdd zQBgT`%jzyu;vFX?Z_}3)1cGLpv0%8m#G=D303ng~AG_vu{{RPP(gK&t5}Xjb;)C+$+2a&qHzOL6h0$S3Sab6aPh;L3^fsq1&=b!=Z$Gx zA514%rYDJY^T&eB5pVI@zEMh>=^8y|$BjXd%L$MQa}znfmZrtWV!8xT>I{}1B+iPE znkg7D-O-z>?~Jqr6YvRD5&WwY>lU1neuK)S@A$v-tN4|53B!{Asb9r3=U4Fq24J${ z!~8J$ZH0ta;$-F7_dh}2uI&4rULnu3;tVD�$}2e?!pp5YyQ2Je(BQn+7#~3+QU! znc{}>TpK_@EH556>6plFk;)i8F!9mhJ7%o?2g_k-;m-wZ*%VAtmO@Q!i2ASAR%{AR zEZdTKuv8#{A2RMss6izM^NFy9nd=Jan|&7y=L;^oDwDs9m;k~|Q8cQVH9wr36;;Ji z>=BjQh_OkMLpWy-2A5wdKaj^Sig=ekagle)aw&|lTyKK$Ri&~geg8=5=PuJLRb%&C zTL9j|F&3K%owwM09hgS^9V~Vpopk$S9=^pYuLjkg0B|8t>kkCe0BWSAwg{KR^`;^C z?>oh$t86a9_5_z5!G{M3E^NKi98?kjRa>1#FH=^x{f=c=28*UuCe22db!xp5zr}>G z$VuWpn4HH?PW7KqmulRX>0ccV9s;e# z)ALCZU?##w70&e|OI@qNAwtKUv^I*z=kg5$(lR>v9B+&m5^JCD2syXBa{$yZeFEPR zub^$F;uS2pAwNV!mYzxu!Slv;hHE#@25Iuxvv}D|5_IvP?n{)(amn?_st`Mj+;!gp z=~d=rK|F`?J2GgqF!7wlMlt;LvxwP)P%W_$3N=&Z=U040B|)KyoVu1m)Fr!a79i{u zWG|O^ReD4)k{0eMYi7>eH4}KN-DAth39(|<7D_pVR?JHr$`3gS&L^bZAzDuj!FAW| zX3mbKlElNi`|co3FvpxEylX{0(;b=GAtRN6^W#=vVoN~KQwQ|ENY7LlV4ts*kGQV5 z%(qyPO9UkX=C9PcollZtA2;i%T?a-Pj+OZDzIu?Y<|ezXD3-djtX<^@HLtBhb?Qh* zLo~l|wQpxUd}sxEx2uDA@8kn*^Q_G4-3_Mx^%lThx3b-5wXN6p6R42l-AX~&RXh4e zP8B&CnOdOB?MjyCF{L{7n|~xn$R!T37Jm8l7TsJT@mew)g7KU}ar=~t6nKeO`70HJ z@`CEw+DIO({5s2mt64AN$e%_NJsvch%k;+7ta6mxG*_5L*DVH^B)3p?vc;0oW=TR9vC-=|Dcf@=$OX?9 zynDZ>04`Gm?=3mYwyAMI8SRg54S;}amm)im z&h0z2p?1?ZdU_Yrlf8eW=ROkpfPd^7Oa_3Ingm)s-wbBzbM#xAs?al!Y7A$J*d84! zo8HW6wm7=2V*RVXjF?{~51vYYf3(5i9ao54i)|g!i5kiQBbECt<;4An&k4Tf$|t7m zdVEiGakBA{aU6{ANueJwMc!}>ibf;$!^ThT@HRKujB0O!|8R_nT7tifTbu}{!z;H{ zN2CGn!{qi8O~<;oKkYRg%`~O>);dtZ#c#DK9k;);t^go1)6!>hQQ61A8l*ol9WPrVhBweaqqDf3RI<=h2~}YH9qF$haZPM-9PA_!BAh2I|a5k@KlO z%O9wI$$Zq`sQP!e`Wx>|pO?QMyZH&>{uPyb=S|m-4LEaUkvWS`aPChK1_XG&K+VAW zYEU8DS+_QVH6`M49P_8X0U={3@8bg&iIxnUhT}RZCWH^?;(EuKGF$B=>L10#9BG>L zc|$P#Z8z}0Z0v=snx6wW7D}II97v5ut?@HO8vjKxjo*#S?ixw&%i!|x4nm=FQz)0i zyVu`va2QBY2lb3?$!?&NeIOBdV^{+zzpFZ z!nIKC{q*$^p1W-CrhD_-J9sVa?W6XxaJchvgTt@h?ga-y@H+uzFU-y-Z`M!pw@LOp1&o8i; zafw+pw)Crc8MR@?G-WjAWsDCq4p+wXyo^y{Mu{>i@-jZf>r$&hoQjR5Mq`OHmh@%x zC)9&UN)%1zeke9OR6h=GQ4ET>Tl+bpl15WkltxmHvIjH1&7hLgswkSJAt7zxKGOM; zgVCi%a1*+lg^|9(R!1|{hjSw~})skcA?9K4hXW{c^PGw`TW+9a=YU7(!?DW%>AVv+5fX*0*$*^$qJ)-&5-} z{$BMd;{DsfTfY}-ZeThZf`4ZLDy>j!u98Jd{P%Is6;v-{hy`hxib(8w^gbtky&gV? z{n5S84xfJtpWDJ`>z`cy!Qrzbe9j8L|1^BY!ty^4pB>@zlCb=c@VS-p$WMxQwbdDA zMEwtDM9a6G2^${J(_S37wzrD*=(XDlKyUAcJw@65=M}t5=G&FL8h0K|RiSA7h$|Mg ztjERt_UQ6$oDt0CNOgW`G?rl}p6C_v%Z$}vGI0un_sF|;?`m?Tv;F~TE zL1IaXx?>qvR~;zY*|l%Tw>$v;EZ~1mIaXQZtK*_ATWIVuP9u-y7==iq(v5`?vxg%q zXG+$Wy2d!Rwa{U@g|~7Q?djN~a>u;AL)PR7h)^_jYH4~P^}F`yzQ}TE3<3?**GK(- zLe*L2!8#!>1iYAwG;)YA0_fVgsv>`)Wls-dgAHe3|2z9mPa~Px$?Hch=PaYO6M!8| zP_XDZRu{o!D1!v((ymvUJm$0XfF9sc+NDkwMV7xYz(R1@O^kN2u5EAM@d5P$yorYDcN-vAT7f#2t z?AUj8T7`y1uq=EZbxSa2sME>;=gq-Td!b!0)=OQ?{R$vmo&RCi>5gCIuBA8^12FkV zmtm_{R$M3XT^_Mf7A_~(m=<4WXW%5DzlSY~UXK#fO=`anyJQivh(!H!JH z^ZPozBOj2MXxE)iMCYfd!s@k$Yopu3%-t%u^jd7YC)y)k^y4}&+8y-qm3QLnnf zjrk`+Qhm8ds)|{&ufPDy+y@B$@pgxW9OUB-Nb8Eyg*Bx-uRRmq-^-L7%A(caRJnbQ zzWrFx>G-tnI^iq<{9vkL{OQpH5JEvsIa+jVIu|sjXz4J>G0+i6IF}T zXg7B;|HBSttc@%WPUOB4P&8I~H0}JxCX=qdpHsRg?|15QLotwTO)U8>bWdOACR1kc zADY2EMd6`^P_O;RYSwsj(GNSh{L5y_rpfFjdQB6>JU^k?;~Y6wxSg|AR{0FhrPj)m z*GudLabuW`(AMuVObrh_ljFZWl@WVv6+c9{-vmJ*hGlj^;eg z?;I`XX7xD(Y0%kZIecE0J~`U-OFl{Ou4I30+GC)MApKn^IeoP7npls_C}_pf+l?pv zWtE}wdXcBuF|awP9D-<;BsQxbL{itg7MZLGP{86?c{@rRLdFh2;LsvOC#x7_y0Uhw>fr7rz){qJ_A11z-atS-2KOT+t0wWh7oZ8$3E)wDMl5onvVV+3CEksk?fiL?J@aH zykE!p_Qh6fGks@vwhM!2icdh3#q8Wo!0=IUuh?4bhWb9T{GE_&+1ccsYBN>KjIW3; z-^h%wfO_h?s;qIVpf*kzBxy#g>=vG!2IFhu#Xn)vbhha7ZZf=69!c#g&Sy*PGodgF z9et~R$`B@ljEn4Dp+PGoiD?H@=O+rk*i=6h;r>Qjr1DcQp0O`w9u-OV||05o2AeDVkdi3Y)2iAudCdcCD*$l?Ne6xVx?E2{sc;Y&hjv6Eewhqd&Lg{WH47Vk0O6%5L6jFx)XA=U}BfvQtw8CWX zH{c<`;$Yk@48O`Yy^qsLz(5z%RV*t1>56p<-}#()z$(ta>XO5btn-IWsQh#;IIp=+ zO)5s=l9pFnF4(8!!wHRC6Ggjrm^winX_OK;w{z_i%gU zmPl$e%S1f8OnW{G-q}LC*KT)${9`|R1JNF62lo?4_XNL|?(QPk#@E}nEpzk)7; z6gG=6BDS)xa`K|9(vxSG62TUt*gyonX{9D(+)kmWzo0wn?-^Y=ZL7YKpezxCePCQ& zvSDi-H>PPNAU$6JJMxxjg?G}k+B1hU@=VF^Mw;AwaMTY(t()u0J9V>c>SJvVOPGZP z{LjwTe$7y>N)g}r3OUi_0yty30SlIxP66xdm6kIozn^t<5B4T~E@gV)!e#ynhsjd-pNdzeYAlizLSLTj`Gme{ zS%I%Y*1+2NF$>!@5wg0%{9`6h$$vn=L}uZ!aOU{+vA}*oK(#916M^UCosI+0!mT za7~D~HmJnctP9!RI^i&P&FBKA&NMNeW?#Jy@zadnWD1oD{{6;W+l_PTl7~oFBWlLJ z8cU90T51HdM+%|h9toMS?<{>LePz=f7F2%kY~`AXT)h^oeSO#Od%)qM zzgx({Uh>{>K|Y`jr+?5|sC!Z=r3N};uuf#AmI97q z#b+<_;&<8}^jP|1ZFr83d9Bi;8H#hk>Q{yf#kunbbWNA=r;QruH(@vAo+Qddyj2A@ zPt#bYWT366GwCrVse+#iOnYT_`Z7K?J0T0qn;lj9AR{$3q5)xizTY#n^pic~>(%4L zsgzG|A~#2$GVPJ1R2na}Y%Q8sU9tLZ$1%9D*mv*NsO^17Hgq;uMGeA~Y6xMqMeEZC zF>nV*G~?+Dc=K1fE%=#Oa#&f*qaeW1LqWB{?7&^zwmULX_)^y|v%!UJoN0L;D6H`H z04`JVWoUE{IclpuE0(&vJ{#qRR>02pW2xVpc-ZpT9PHVvdhL!A75n-kuB&8HF<)#5 zJ|%LFG_S;~tc}0fbbh>@GrgbvRFh;0g}H;EO=XP8$}LZa6}NxQ0rN z2=xqNFoP(SKha~ElCSq_%ht2=CwfZ8LC?xK=x-PYJC$ULplHH_l-eC@ovt?umFt;qNB12G~V@JyvHa7%^eHnxbA=~g3yQT1# zt2lnwJQ-01p=SeMD#yh2w=yB5dt`=5_i{}GeKFvE!?Mx&e7R%lIKV}zhTsMYXr8LX z)7BfEU!nf766+nEpVkjBOU^fVx@D0X-ekRZ<6QEZJMkUE^knmA=H=fK=5K{&FFnBq z={|fq%vi6CNqHHI!;B6x!hMM-{FAMOGp)FDJCLW7#hkA8FJnN3_;FeoZ`9Kx`{6|j zgV$#vG>u{r7@>6TeOd&p%GE<1@)X@1U@v&~0|(GdE?_x}(n7R?Aw_Vwy)a}eG)(m1 zp|Vclh71O@iKXr_J7_|+G8w^xm{l2rh&?ynNY``c9_BNx6vTYyEFQe+0xu>+3&=|x z$4iiSkIM|{{~QX)IKj@4XQ_6SDN58oDN`h$P_CII9gOo~q&?Hu0m_S2dVF8wQJKdy)aY9=`6-Obsgg7hCK8}7e+sSXLM^sShm zMl{a71AI>nnDMx5SY*o=WBtVZd#rFnFqv%1DZzb`&A8~BqD22a(gaI2@$P0B#)>x*B;tj!GloD<{vT2*mvJfecTtCam!BpW}Z*tB4 zCv4Mf21iwL&;f*#?{Cg@{VBKK2K%(K*Mbp346p`Ue17PwW;#m3P4>AIK69uhADui9 z#T8Q2fl#-MLQge$<@gcU(;-uE|3ou@5Z;COmcA8!rHx597=7?>2IozW3C>Kx{NhN; zS$v>`vQ0DQKV&d&2v$+nzH?@ZJsI!LA_HmGiRw_)Ut38O$y;)eci%Jwk5je>)1-{z zZ7mPZ*C4hgZteuhqvU@&MwuqwaY;y{{!~j(SIUD!DaHdJX09qot zY!oBM!Pp$C-&AjD6fD(?gm&y-4q}f8fors09Gv2Ye`SJg7~g&r6ikI49V}WhpO%)K zfD-dT{VV{dEOI*ma+`Cw=*jx_*Og`5A)EAGVL`q4^o6a`6qv>d-Vq3d&wr*!Pxa`jGEX1jhX^gac ztd!KaDs7yJSI$|lYMSYlv%xk5|F+yn?*v+f7Tw5$594JMm<27EDxNe$nj;rlBBh!Z zD6XFR_jyZ1-+JZuTU{zsO!?s79>x=uie>R=)2u0rsAS@Ys#nvk@wHR2O*?ZvI~K+xUoZqHPJ;EDt$Q)8A`)nOo+#Avor6Lx9=s;>DOghLxwoO;e8C5 zMYucUL3C&iufbh^>{(W6MN!Gh3j#>D^^J`L)^r$GHX7ZY@MZcP6YV4qhAl&!f~^b5X|YCay&Eh*lvx7OY?{?!L~&4fojUA@ zmW@BVu4XSFgQ9Gq50%wxIKQPvH|%YjHuHHV_`sgxYJ>C(+d^0shGDh(^k+6-=Lm`; zB$1ROEKEWh4KxJBhX}9_mEFx748dwcT}Mc+&k+_86jKiVmSOKYwpj(S%~lRlh`zCC zO>-goyu>^{!CvUyO#0jWlJYNOVfbEm5y&pMapwqRAdE#Xo2fh^knN=iWM5X(z%M_@ z-t)vC8%27ih}=QH8+v+%h+eEuF+?rbh7BtRE3Pm`HKo=sP1mLVoI)0SC$=lq12V`-ckxOZX-GeZU+tckpZP z9coMHpog`Mi;wOJCNc6>rbOEvr-=$CG4*Q>_e&!wDM_-T4ib>mB?Vh=v{lxlth7)Q z_2CW=A6i}LV6&&jm|YLq`*S4mb5J8*^zZPZvqeCFGJHPg$ay1o-cS|o$pHz5#ug;f zH69WslB9oN)c*WT4)_N7qE!XFk$4b~EXmWRD5@pMT!H1y0-fu1@^NdtEa1Vd30}N^ z5kFuEm0nwc6+HWN*RW`&!<1qhki@5Y%Z@&JDP|f1Q>Nsr1&oilWB&hi3c)x7;2Fa8^lonPa;ga`~=La}z@ zg3ROoJ!Sp*<=+$6&{$Y{2cZOhRJ0Cy?bP37sM>9h1#g3%-+9R?`> z!%Rs<=2^kXOBAzWqM4OZLUz5dtD0*8)6SK1zb~A$=%{+48B&vS;naI}qdB}(h`&X9 z&A#+3;@@VYZwMY78%mrE>%&8SMn#8>Tk3dNs%oYzaSd2W_gi(ru(aqM2av3lf?Gs+ z-4?92+R?v55rXgs|7LqfDFK-qr+q6F$$jk@C5V?&vyC13nBjkdXr;PsFRe-;oHf_8~(sZ)B`&W-$3$w-L_N{y0kv1`Gw{wITTP3z8p0 zq?BbBv;FlgQlDIJBt%D)SZA0BF9Q;IRH~RLlB8uR+bRJhcw(b?`Al%*24gV085)yQ z4wNK?E#fW2n?eQ<;R|wK<-uDxt@39QR8%$}Du4Web}oZP2eCqWJ|ET_8b|Nb1C5(L z7aH)L!4b`$2aUU)-x(UGZzin=3VjA1LPM<8Dtt!hT{!fE=$(e%&Tzsd!5MW1eGvUB5;O7#mDqsr6RcuPE_8kURdB1)o>EjzJ00uTwX?EG` zcSGQ-RH+UxQUi_;OD)PPrQZ#~w^V8)Y)qCQ8PTFUMQ+rEe3?a>a3XS}Z>I)gY0wc$ ztmTC_d5nvoC{HSo#&Ty~pj8^&d4*8N7J%*Y&tSq6Y~f5lT_FaG$*_369aVB|EJ#ly z(e_8MMmxPu+dlb(?PDe3>iFoD{BBb%LLMDI?WZ+{w&p|qxR{;be7V#6UA++i1;s>q z_2S?ngbfg#oVfHnAp^W>RWF|I1yXg6bHd-mXEdF$wA_k8%n^&qUE<`Z|0V8ZFRJc1 zsThlqovgB-h3sWFDz)Ehqap{>Dc8v$wRvTr8rX{)y{k@z*Cy4&O_R!}8>S z{rMq=rcK$)UoVXeUMm!7)~kZRt`mBvs%@jOL=g|n&ALWt|BJS4Ec6!sfihFzfDT*= zGpW^!-N#49{WQLM65Qc|Xu^rqN@!f6g<7huSY1JU5Udd2mHFk&rPP4LjH=Rc3$;D~ zg&b9DCyh@emtW;xd^%K%U+lTx;g>lE@_LAuSix(JfBpTcoSGA>A$TkOhwR-@UD8%Q zB$9=^wNqOEBFXRJM0&kb%N#W*a?}+?k-^w!K2F0(YpS${C3%xQ5$)`GeVMF!6Bso` z{!yD)=N{qPYQ34Q+D{?Iny+!u%u_Ys%4iU&MtfKIo}GdM`ht!EynVN4;*M;j`^R|Any{|SM@;fOPT$$8vcwnxFuwPUGY&S3y1VZ z-kn}Gx)QNkbOoUh?~so+KBN}S&dQk z5kNqpwV05ZILgGKuNjis#5lW=vO2QN^)t{T^#`+&?vT9>4RO`AuZ|Rl{i?CdH;y&& zvf~fgZlq?iG2d3!OxwYIuW$&UQvpczo!3j)%tCWZ2xv=ES@fTHn=!maya16F`X;!F zA2wF&0=|#>`&D;LE6ywStoD-BfN5$T*xNu_q~3;loB z3XE$Sleyj&V@1ZwrAL_-sHUL6L{OKO5m49yUcSbW{3#3E^L^f-=c0@`n=})~@1`kX z>o}3}-jN)%;3bTIFbZLu|0W@nLBz?%79a=>dO%fYms*Cy#I=P0=0d~F5Okh@RQ zj}!QD5?B9t@ztUTsgu~hE-1sxss`(9zoW~ueLvcwb^p8Ex|h_Ax+7EaMt4Y0X?e6- zCY1H~qjr+I$sMCyo@4bqId$|=CXK{@%HqLnKOi8D zXi}ylEN`ca60BJ=;oBH{J6CU3a;$#EDuJ4}@I;LcgeY2i9PP)-LHc_iH?g`*WU$`J z)YJv}ZKj0&33hW_Jb1a7oS=fGczB_XS`N8b1-6EC(NV(?Dk6qYY%#caiQPrcfZn`+ z{;OuhL#qrF=3JeQa#9;f>6gF7R({lf!L9bxF;5q3m!z`|!FPt6Ky6g%&M~&!pZ^cm zn{@n&OL{?BebDhDQZ!=+gJe?G5d5W;@M``cd}J%uN;L$paeqYDrH;?Lq-?awl_ zm)nAC7DSk-0~&w8Ut~(%WJ4eyX-g7g=zAu1Q@yc9*p|xRQ~NOTBpB%>6qvwd$bA5Z z201R~PY`!0PC{6mC-P!iZEzTe$Sd5><0RU$BE%RIA!6}*lOIrzNPZX#M8-Jz;a>qe zq4|!@R8iJo*BhvA= zP0tEu;~Kp@rtx69Ym7yDH7$r1%(+tDqtJ9o8%dp>Syl@8G9@R{32DX-xaiu6r3YdV4^I+F%}0!TH&XtI zoCL0TCd6abw06Uk+X>K*uR??t&vP!@C$8j>(dZn_pkp?4A|gu2Eg!2@5PMx0WDG|< z>Fm}A4<5QDv+RV)P{2w2d2lj+s+Z`HK9b_3**bE4*({#b59oxxdg=Zu@gRQ+9=?XZ zVpb#74@niKZ9^sXWt*O22gCRK!uLht``YlmJ$&yD-#f$iVgoALG(D+rLik=9zSoEE zRpI-b@O@VJ-punNR3URz9~5IYPY8^}(z*O3k1Zn1YwgMVsq$3*Zho+23NQ?Zu?!2Y z>Ph6hg%@&fD(en&1D;_M@A1>BDWV%;HMf;%Oc$|-+gNC09Pi-=NAi=rd&+EFfgfRIcb2&UOYa19+Es#yVCS}T9I0KBGqJ*y*N-x7y^c{OOZ3LSNa|t>!*QQ} zw`}*kAExySe~i{c_P3;cy&s;6Bf7AQ5KLa{u-zfc7&+(*?7Tj_5oTbe$b+iPc~D>> zFXd(YTl!Rs^ixhWv!ChE`z2IODs5lcIRh(qfest3E5M*#^_62D`oi4vW2y09!bttt z7%_PlK7wJAc}&AWbY)6{w^=r{ukX4plMiY6i!2rtVb8Ixa|O1A>I+_+4V6&fzGu_R zNOyc}q~)`Wu28{51gPc_x1$g2=`M`MhC!kgD<%{3(AK|H--+Msr+(aKL+I!~hEBNO zc|izeiGG)gh5Zv98e=dgWM}+24TP0wBdvVLTlg3JOsT0fR~#@1Kpiy@?pX3pdx^2A z4Mr%h+5{t-y@jVD9>;HF7`n~)5^RI^CEni~bH5o$O)ZR-cR~{y&#Ly84HA(kb#D3W zK6T|AK$aqL%$bsh13*c%d592p4LEf>my#d8@vQ7MT|C#+l@O$NBITdPDqo+A=iz%z z7V;H&1mntGj9Mng5d`0>^j1or{vidN4po2Yw{A!eukj$=hT8f4#+3;~@KMd&&p%a58pxgE${ z_9~rAP3^?9rg1NgJ!g5|^ir7K6rg{gKf%H4oH&2IJ!`-Ebs zis9d$V2j#?5xLD2kU~PKG=(y;(3eSA(;G{0wW?EPI=!jkN3FN)1Z+^p;!^J|d>;v% zVHZ9?!{QucGgznTFg6?bev=J|n{Y%Sd_^|Vq_c^;&bL6D(OzjPf3j(AE``x@dp8_K&F~? za;HpFapzNJ`Mk-(Q;$Kw*IGna@ zQ;ODfwDMyweyR?tSVx9k(3zHp8C)*R>w;MM>oJ_V7M?1As~QF1s?!$mbqo+9h3!Je zaWX4!IoHSwZU^TWp3E;ypF;}zxxO>#2Nf@;zX6gvddbOPdaQ|tadp;^ueDhd!M2iF z>9syVeb_l}VKHXWo?Qjtsm~C5xSdSN4lqh^6Fky`sE*2AedZVDmQOkjb9XFMXT(APRF)z_L}nR|%V*+R@>1e2(8`o|=y3g)hp``;*h5mA$uM^etC(%U2M zYErcSmjn-*Ttsn!N>zFzL%qWPVeU)7qbjnt69^F28-yS#YLw_Cikc{42x|fXZbJv7 zvd!Rv8;yz>A+14K0^QNHZEGCSLC1x0oRLvfbOZqv!j^z5FoLKcC?fX~SpzQ1 zs%|eyP-p({fBrAelfL)Xa;oaosZ*y;ojOIY1M8Up)2LL^sRKm`faRYY4N}n&_o-4( zVTa>B<^HrbzR2C7yhBLyBsyP5Zvdc(=hU72kmz%Dw6HDVOq>)3QkjZwldhC>KW%+9 z!S04&r~k-|Dqc=S7bPA6Pn_bW?kkW-;Q<-z@!Pp)kX90MWHQ$DM2>Nvn>|lRHv$q^ z2Sk^@v4@1lN*93%d#rcAC5f#Bd6%3+`GYNS`Hq7el97}%8+q599b{=p1$8{RKzfu5 zdwwY7{@W@Z^EgCUnnja zPyb5wr%p60LbS&UZ^#2i4hTC?{wQ&P&1tAK59^#--Fxs-z}`EEiYexN4_bvDk>1;~ zJ+}9bNbl(}KQP*RYT&=fv`+64oC&Iz-cy4*(tCrqIK2nL#vG`41&1IL+BHxYMK%6H zg=^Y7v}aVbM!6oVa4)N`z1=DPwv$ED{SVM}mefT@CvLUR!3a_OGcs#XVQ9HRBhnLn zJbn5tiq8HGr%&a4N;1x;M5h_-o~RkcnQ@$%CU$ngY}4JDVK7QC#-4+APG%NyMa=1? z{I49}%OJk<&zVhpZ|PBAZ}5J8D%z{57lL)@*pp~JNEfG5F5MsH2bzh>I(IA9$f~(H zZ0-?t*f!$dn3uJ|A8b;Lf8(#zft;+0q~E&(JMF2cv%({!h!#9Qtj3=`pib|XHJHCd z0_nk#MhV0}1N%o1mhCGMLVd|Ho?o#SBD?M%y&%e8f*L~GLd9;V?fw%_f+qJ^AHq1- z&iGyu`w>t5=~k@jsRi9*EkrB_)?$7RUFoAY z-Vm;`}K0xeRWUyBFkEO-4>EtddNS+jG*m zG8704?B*4T_VIdnvriWlFzTUXw*W33f%xXm+$}d)4B2j9ibK0m{A8dd8b1N3U;BTy zIl->`j`ewmmW?a0Fai=0P${qZ%Nc5q5ejR-^`9e8cL+yO%<+qz^%&}#U(v$;3t(wJ zT4PtE&m!|nxEiby-Xi1}IAtg0g1Dqr)KhTu9C*~@u(&4Pg$hMn6K^3-Dis~s>l}P2 zdMAk?YD`Z2CFIKAo}`cY==Y- zWou8zxe-stflY-b%pdlqD$l|TeICe9=DhowIP<|2Sz`{d*&q@_zJO{-LgYdb~x z+rYAE=Y^r7Ivs)}d}W!Tu{d4%wb$bI0Vjy1=Mu$LJp7&nu}N-6XP{BjW)@s)!w|KgXt66(hp}Ed*kL zi!Wt6Os>sRXoFl#!{9PxuohlTUZNDdYbNo1hxPvI{Y0QLK4(WTCB?i7KDk}=$BUY9 z$EK8(Hp-_`fTzB1JFMZexhJ$*X1>Zx>L(L&(9_`qZA)1}XNftT}R!gm6gGv=4EL^H{ea)tYa@j(Oi3%l;2n>2LS#TyoF;Q_yx^urYW}il^I>S*fFWs46M#=w!!o=N>!JsVY zONioB=3Ifj-YEZf3W87pl5=&^g#eB7IK(Q5O(CqJ_nC6yhN6lyY8k6i&C#P2yAnHa zmI?ST0u6#7-e)@^<3ta|P`9RVCe(y_nuo5!`rjQ&_toYZ%kNFJt{)3tCMQ8NQ;n+F_Nxbj;~K z47Q`z1bpjm!&u*8nahsFG^2^8>q`;pkcpL31=0nH5DNJuIOa{!V- z9Z{0OYoWxZi3=(}>I!EhxQl0Pnp#@?^F63xao_t^E^!SAlW4Fz6Jx3p5h%VGZ~(T9 zr3a^Bq44v=9Jy2#<1-JI&{#xZjFQ-_S5G+*GvHlhH9Cs)Sjk#LF(_ctw+kucu{B2d zK|Ca!C&UgS$YUj56dh;3|0tF@9UyZm*rbOTPX}>>d-d(Y1Kxf(I>fjpzSz8!v&Bto#b;l^i8Jnl4owr+yU%sMJq=jCYR}+NRcT3RPHhJF`BN zp3qNSD+I1sp#3nwIB<8Oy7sw(%w}IcsH`jmhMut$%3HEJk2pGv@2IrX+Oc(b_gIE5 zeS=_5v||J;t4EZ+^Ft0VHO1JA&?=$|{6Te}$C_}f;N*IguW1K!gqkr~zRfFKKrZ=+ zr|MouQJV8ulW>sD-i-!0BU=d**4J05@tfw1->cjNf(;4|9~`oI^m5YpQLA9f zhxN`a(%w6RY>#tNP&F&&oY0&MnwlTAF2KFhc2_R*6b$Tq07iEO<08Ot&N&HQn%b*? zVDevT&C9O^X6#w)U+~&`3*NoZo?xN-{JtlF$Jt#e3q7bt3>Gt#@y*a1N;QT-!4bqU zNUVz)iJAMI;cv!i4v;ynsd!MaI(Oh8GpHe`<`q5pJg6q0h(<@vDta>D9K^!-2Wk+G z5j!sdBqs|*ajJtQ7j}Q~Vq~g0$zCk<)u6G*g=isqa;N>f15a}AJJk0CUOZODb&sK~ z&Atpc>`2(ACNni4TgdSya8tVlhXL=-C}x{|by$NxfmzQCS8-{?EfDy6fuKUk?n)Nw z(E0|{(mJz{T|uUP)pzv2E>yuDwQl@d^b}T(xC(2~hFpRR^-bH$N^d-e9o8UCs!DUr zW}26gM)4-jE>PuQ9_x);xhNwM#YXv2{5o`+kEL)Noo*N}bh^Ic7wPod^Aw%VMt1vN z=o0A9DswAP)5P#UE3#x=Q^P--Z{c;gWe>8_5e!D2q26$WMJjC>SubTkGZdvlMOPn* zR?`f;2Tx!QNm*B6{5p%37xZG~>^H&t|0FOQAm$i5Db#nDMa0ZWBVwMHvuh;EY@Sz} zqiGciNFawypGRhetl8CA9?la&!-4@%UhD}9RzX_lnftxre~@`GcXp?ZZ4q6>X5i*EB*M-BZ;7>WE@wAxLWR?*+p%$V%$0k*WLzPL1}@9G;C984QDa_sNT_E?p z@~&U*2JGiDgr(n)XnQBVlzsy3;x<=#3#jbuO#DF5Rv~xeqjoOa1ZeOJ#m@ziPwrdV zX$YeRs59wUT&9kplH3qN!It4SI2@1SIBsa1FC&o@HJ@X1<$ZwAguK$Xx`~}bN)E5K z;W32)P3?5PP^yZ^S-3qF8~2+j5JA);Z|63Cc$Ia(iKHSmg_^DWhKc~lP1LD5B9;9% zi4TC>Pv2`EU(V*XLmnK({hZQ!U`KcpveeE&z!3_7 zPQdE=z4+slMtbqGyLMlf1$ z(0Sh`-DL?q^!>|Mk?nyYr(<{oYH;Bp7YxaVpSP1lDZj&>E-a%S7^D17aw+Txg>oif z*kTCXjR&oTpB)831j_u?jpxfnKDY_!+~{cRf=GaBih6mhH%?=fR@?roEP3wG&%^q; zO+QZ`A@PxRQBz@`tpe@QDp7po4?1U>PIQ6(F3{gM>hGKNbH9F`r=N$eR5dbn`p)|K zhK|1u-%)Mu_)7XfS_(lGXv#SkoCf~G(#+tgF{e+!BV`F&VK0zB3a-tZqcDeIJ;^mq zOlkh~bqF9~i@j7q`l;Lz-@u>sX0QqF6Nc+&6kuewL9ri>#>&0agh=>2n*gt$)M@v` zsRBwrl?W@F&e+8i#RKXfq)Z5?y>j#wYWCAWcC3n1-iyG7Dv*_@#6vl4zh;N0vVB|^ zEVu@#v2ZEx3PDix`GTK0Qf;)l<9Qj#YLxe2iL=(Qb&gfgYVMB+kW+QFQDNx0$!H1s zuVc^{scsEBG1PV(YgVt;6qi^zr^ZVk4)G)BYg7|Vl1;{R0G)|t2J?~;vQR=ON#R8w z2f>7l4|r?XBCWz1sNd~kEBpTeM!*^~gaAa8At7XUZvltY{y6*yBH>SNvqnuKBnH8M z=>RwxlF8a*O%$CsI^|#6 zFBhw>97O4sNa_N3PGpt_xs{?PBX^9AM*6T@P9mVsWQqJrYBD0dCwy`AL{I^2(fn*4 z1)3~ZFbTpS*b4+Os4&v3UB}QG$__vXaB(=_xFa1N1DXxwAc46%=dFaL6sUpZGCa;j zFZ=^&T-F&M1Ycx^3aF1)y@ISMTsz>Cx;E|}B2?r}x9?G2tetg!BmWSj*5NGlZiKvc(H67bYuAURk{N>5Bm@ z0$L$K0%Za2lg@t;%Z?E!UHl)ha|Zt$tqKV>)a?82C&Kl6BN4osd)*uq%V2U8LB7Wt z#9HoVgQT;MK#HKeAN z5Cf+>;m}oQTz$TfdHSU7I=B;XFfwD{e7g~FCfA4aIN+cssTaCUx;O%oYrHZ*5WgJ2 zbR|JzgFyOT`h@+q|;irQC9#QzMe;NLngbzdJUx1%U_%j{&98@oW zUkhhj{s}K^+d;nalGLRwiJ5aw(nQ6mXqN$|);(QwG)SiQI8bY3Kbr7HJMbm}9!FU% z`pZ6uAK079h*0=(1`U}8aj_Jma2%{@NF(BEV{;E7&CtFY{POK01u? za7Q<%)!0(AZ$-&GzpI+EbgQm9`djjba=LL)VGeK@XxAvbe*m8SS4fqxFXVV1bu_-d zp=KRXnFiw>03+j_b{bKDBMERTsafDPZvd}*ZbNv_ZjXV-;Twh5jPPzFJoaY`C@r0a z#?PDEdjc%00tU(4>_wSGJ=UrZ1!sL6oS~gI={TI)^DXFeAzf}AYK85$@Dml-D&YXe z_btS$82{&Bo)EHTBqx+KE9*HOItNTUEpUI1(B*@ot--t&mBg}jDhLL!0nDRi{ihq{vK9{XHUnHJU6Ww;z9tz4 z`WUP{^@`aaH>?FO<7H@YKXV=rrC(U+O+`w<<)7s@sJN9@ zGC~>98vz0ddl3*{(_m9*EqLu9KFL5un8E>50CE5V5Y(qCxyWluv*1|I5F_ggI){DkrU00`%a)TFak<;g+?X10zbMk zH)d~g`F?C?+|PNR@UfB6rc*_mQ=xXj1w-D00x@x=E3w)YxIDQq_Xn6HaAgGD7<4yn z+`-;-{mt{=FDp(6XZPID;*Gj#+|PL%!Jvqal%Y4Wc%YtYAL#Soj`5In??NK1Mi7>Z zPAdNilbgr{5ehcSk0T?fxYQA?;cSH>MOOZjwt}BetvT6rkBHrToKQiRU)TZ1i0z+a z<|W^WcE)|@5e|Fjn}d4Zqny{KlhGb6_N~HVe2aIVT)30OMgm^uA$Mpf^#4CLzw7M6 zV8Ei+JHbs5+!Vo0B=tj1a8m>)A-Jgo-|Yk^A-EZWlO%YY6O6ljMm0xpGYNJ% z!3=6Hii0RaL4TwF#TU0C-4i6;EcV2AuSieW;=C`g3WnURfdmW}Y#PE`0xN85`Pm1< zpxq4Jp?+GZB0)zcfLit|>uHssn-q{;WlzPl3MoLBvDXSo8P)bSghw(>a5BBBL#3=M zby;==Lc^&Y7DL+r|6g#Nd>jYr=pE~0ddE5p{)huDYZj6{eMYj!s?ZoL*QTY&{y3Zr zyn!tr+=FBWhQ;24qzYNB3J^PZ-`u%v5H(%G`OC45$? z)-h+jT&h$rk#1l4rc!wD_yl{arCs5n5b8vL0(xi&%h3u;q;P6G4WRZNkJY{fW|dIi zpD{mScO@Bye-}HImHv0*mJ9I$hkkblF5!vG5>y^-K-;*n{>(MbLwnE%?oj&uPLm3Z<(Cz3AL(!9 z$Kl=SXkab&&+rPbA00!3p6%1@E?jvTxnzBHCDKnxlnj%j8IJlT!x<_=M=0T9^9Wrq zUlYmKE1GYG%GW{VYePMK?IQ{gOpmDv6>Aku|8Rr!P3>!in82Nt`A+$T_?Ba1pMip! zlw5}2rX}r2O(lbA^a@A8VF2%1K=;J203cr2=*t8)22yx+B%0;v=#DD7TWoZ3H2Mdh zVBi^PiE%$SFKQprA7sI`HTvUKe*7g$QfA;MGc*(Qc+d<$k2p6c{JS8jngN^` zEJoyHSkKUd@;zC2$KC{9P2BRBm&_q0_VR&y(vT$0WY#*#I++1XpM@ucw0ny1jtwj` z4VjCOIYVX6LNWK`Ax)lSZY|1<$qZnIB0NnzO#pzcgo$u-95S0Ka}hI7K$;0=VCXmi zE)^i16c9!r(`1NN#ZYrB<9Z68#eg%x3{}WyO!i+`zPN4hG z0U1mkTqq9~P^Vg;gEcI91Q#mx{IW5s0Qm}3NXk%!1lE)>kbl+C6XSsbOjUq9Siq^O z1v*&6l1Ff1kwC6caTOEf>vajNDdU5G)n!Z!$PY+-u%ejBYRdS)r7*38Ic=WrZ;N9fFJB(C- zmj!rP3Lf({gqLBj*YM0JJbGwU@aXeW>bFzDD+0VC1@GNP;h8p-cC24<6y7u7)C%4% zr+%9iyi&j`Rq%K_T*LYm+b=12GRUJK--}f7HV5QSg2IIy-+2|U^PvELj_?M61)Ll* zQu3T)gHiamMXLObg94iW8H*P#QSkXt@Rv4Lap9`=-ydivo zv;EAX;itsapDs34{YN|SKUetatm@B)g3kjM4dD}@Fw+Ns_MU0 z)t?Uq|H(${PjL2qG4(HstN(i{JBDDP1D^u~_!*<>&xeA)ywUm-oc*Fx|AIhues~@H zY-ZQF1Mc6~$vB^VG4x^izMXa1M{tV7SDidO*sw|eqEI|Z>3`Lxme7yM%W zHouxb3DrA!`QHd~{6GYiF^$!K542n-&%YkAzg+*WDBH#EFXq1iu@Jj4G$D8o2$m!s<0NxT zCYe%f>0 zdp|G{xrij9Bb&IdH;ygLO~m70Oyo55CCc0eokqq;wV z3!mUnMDo1xP^^P`$tGIOfMmc&Rx`PJ0L~7Fin7c=Q5v2_c;?}WMMx1=9F=n=S!D$u zp-Q&J3{=inUt87JI`zdSAkI=>L^8gT`3O}KQTWPGUt{=)T}{Ov zDyQMHq~SWEq2S+8Qg0|)HzEsHno_;`qHIkWMKS)>h=j&BqrC{Uc$6Xzj|%mes~+>! zV~Kifz(e#f=m&QwU*xwb132+{OJq{I4hvml@z9hCQ;Tgw5rDpj*ye6QvCZ8aV%yMU zn`d96>*0dHN;O{?oG;qx`GST-!mrln2iE&ma=tJyU$iU8U5WYPAk?fN3lf%z@#=1Y z@#=1l@e1hh0v!V#`$HtGqqGXy;tmz1KriyH8M+*bI%vHOS1$C(tz;J%9z7UpBKUHs zc}z4CEAD_{hK56zHYAw{qi;h*rD5a1%>Bslry-{qD8=`T6g;!=1dWzr?Oh5&Y-ff* zx;2AU*1@3DKxv*Cyi^vN!wcT48|L;(Vnw_5c4GXHfcx|l1Y<8kIBCKi8Vz_nRnPml@^Q*sZBxKEE9^&*$XGlB>zz`pPgVgO zvkeMt!oZo<{>1o!0sIn*{a1uzJuw;Zh6CP@fY(pptpq$S!QfmLrbq&BGT;?seX`8&F%`|IGUavzVma%w^Dhsbw1Vlm{=M;P=2g45a` zV<&7bQelNP-+q|MW?s*5PZ$xjc0nXa^YDH}VZoPOP5&<7Jf z7=<0<{s{c(uB^k>(80;4IQwVZd{Ff_ae1$~a#u6fEORp~`Cou>Ijyn_gD{W^tS!u4 zSu&+C6OMR`Qe25E=~$~wSY80^^-}Ob@;V&Vu1drH8P3S8E5SA`F8>M+1)uK;bREPs z^4^U&2#1cGa?25p4en{ejYtZH?I{Hnr*cq~KZ!6iIJ~ZQp`yD&cqVHvrtj`RjXSd% z&O!h5D4b{Q8}BkEt#u`Cr-k4Lh1@fRjo`|BqkU3(q;=dU2)r9!KRkH#DRsjGSD#vk z!Y9weejPkZ4+{J{-?s`=SohP12fCk~j|-0ckF7&xV`2pQ(U$Wq#~ReBaEiGyMOQxO-FJw8xnJAU@zpGRqZMZzgVWWqt?m zOjB+x1o|#;=dLO_5B|J54a3dq*rBBN3Z}*e9cIjT!>Qw*FEkHb}+Vy%4uOtpc7` zh<^;6Z`AxkHjG%_-)kSz`^ke6@++FxJ_dxXSV4q&_5)!K{i*vQq*1v z@a2X-)R>uGc6YYV*^C|Je%Rv+r*6c%J2b5s%nQTxnZZ5UQ4d%8t#Q~L+o*6ev+_pWj4rA5LO28c+n${MjybpgvD2}DLLB+m%ATq)>2K2&%8XuBq7QRAD>(=%JL^=I0 z@r-;_B756!NhB}>`x*FW`u;I=crK2NvIZ+*XA2AonlvBEUlDaMvIkkkS&i++o5HvF z3ri#t%6~>6uJx8yI!z4yiM`e)og@=jYiC99uMTu#2eq}mR86Lu-wQy5_qN*uUZpCp zRV1Isx?*n}Kd4t=7hRyMW+n2VK5M1L7#wfC&A~CWb{|>GEAr7-eH>iIm@IzK;(3R4 z|128bU~2ie?cA8zfaA?r+=&Y}NZGtoo_}S|Ai^_+0cPOx&Slt+>sE&SIRl}c(TfSr zespjY#9Ey&cqM3gbc$MMARKHH-^r_^6HukziOq_e{4#Kdn6Wqo@D2Z80Q>H2_)qt* zVcD=O$7GR5v$u>G^Xcap=UbCS7LMZb@G-MsO+XU|(STiO9c zSl@j~{C&0@mzyEzfVKTRSWyEzx0I30V5JZ-An^`H&Oe%%e?zIm&iz5*&hs?RoIrjX zl49ey8*{n?yR2@^OwbkkSWvw)O{4nQUx25;YG!_ahQ@F0L-q0N zT_$KrXdO$_zf50Xqogkon)iJ6WIR%g@&&AVl&lK)MY{SV+U_%1k5JCTOC#-mVuxzC zLs#WpRlD1>>Twj|y&O54sKy>XTQ}A~W1|B^@D75vuLltj0q;%M2-q8{kAM>J+-~5x zMB_I8;Gibx>=E2+ zUP-RmyY|EJl?X4*Cc!BCj*%!#tU4f8FfBl$QJ3iS&y271SlPRU!=161Ol|Bp`($V| znqJ3`~kQP#%&l7V&$FzoEtV4X93+jVA{C?ghUs4W^!Id3DV%W zVNdH^RKpd54H2Nx0Xv|1G_>ExLksKxtVUsc@iAd6nlH^fU?~TGIT;$vK`gdnDs^%& zeT;f_X{GFs)|s`4+2%|G$CSlxbW8~*)owqejXQ{= zUar6-6Au70Pl0*&O~I#i{ci=6tiWVwFn>^B?$=;W|E*wFZjxRt0~l9sxhVhqa&ALx zUP&^vt%my+eD}m>g7vWEc$!IJ51Z9wMQ4N-1{Vh#kH4wz;p!uBh}42*%S zSCXFW?PGtB%_`2@a@5CP&3e%J!$Wuxw&Gih)6>;y!dk#%>eM9=e)^{cuL2nlN@mLj ziFQsn2wbaJuWXECWk=_dR|^L#435VD?dgIkbN~xgFtfY-EV1JAB$h~iWK^sGblelg zT)eclhTO;mD+F%}r#=}JCT3p)#vH@1f!qvKI>Bzo@JdX+;EZ-pe28A7F!*;)ovz^1 z3-N%@K_1N~7VBC#n0$y1=gpNnTfo&`ZH}{E_T8HSJJEO8GroIGyndx#C*ZXOmmR*^ zGLv4!FJNxpy~!=S@0r04vwGchbMejIs#*6my?N42Q@n4wa<>`&VMqf8FqbvSEuV*n z;U7#OCnc4eHjucUX3|(vd6^(+U+l4-{4>$?nP~LU+b}WS8%HlnivY@~;EfKPM1)dc zB4Dy)Qp#unLHYot;0I7*DWzMwx0gMZl|5V23H4<{C%n2MN+)Y$>7-k_;Xj|%J{g%L zTlaqyM^s+?bt_q=%k5OTVM#kbDe#(#~N+ZrZcXvRM$PLo_Wh4uj zRHmsuQTNp8elg0==ac}eZ+k3WV(2M&Ri2ZcTJ%VIxunjwI?ukae)@4vdK;`Yl71Wc ztZ(%=F=~#&*HgE>0#HP57Nm!AI{!;g3KK9)I59j5NBPl-p^3~28#*WYz0sO(W2d+Xn@4u!df-$pW=M%~aL|mg2{PFkEoErer zk=Nqu@4W}|59*Ez-2vdJ2Wn%D(N#Hz69-60*DzlPx6;8jPCC`=8>|RIeK#(kB-MU9 z#Hu)CdD|lL4T=07WrkBXRFIrw@#)=*BtO2&{2P&^UgQUe+$NE4Ffy>(zU>Syu~yr2 z_48K!{6;?y>E{XkY}G}j@1&pS>gVtAgv4iGk92%KuEXne_-LK~{nJ&vU&pV}PY2F6 zd`HIu>0jN{l?XJOzyhiABTTDwTr zd2#(a8q?pWl((Brd^DqfORJ7 z6HgU}e=k}QPW9c_C@~ehNMiDTMtWIWn=h`7^g`rcW#F;mh$XP5tY6E!T|gxkpMHy-;lWP)aYe)-7+y9)zD&2Hm+ zihBP8-{mkzsM%n?*MU(YO%C6A%^lu{@|_-2@ZN*(AM+iWnY#o=PReProSjv5P}S=q zuG|m}k2zskgFuDr`!+9b8PYSms;qxgEX%p>mIJTI0@ci5e1RQDxT|**IBTA$Jy3cL zMk|lcV%Km4eV{;hi%uGiI?#xFM z%6hj%o$u~oEXLWA)RCATB1lKUV^DByIm$}!ENy*k#X*JM{0G_A)wS3lSX~PZ;U23M zOC#np0kvLZ0_ZN>TKThd$>EoPN}78E{tmvMtKQ$|dnfh2jPKN*A}YlujYR{{(X34jG(X2&ayp0WBm>RnKG?-VG&3G-auDq)B zrNY2&!c75O?|j}1mQn~lbOty;2HZgc;9e`~gK)UKs?fL`J4L7|@lL0K&|`vJ^0?fE zbdOVTY^$8Cz=;Kl>~ZZMV`So{QL6%z0Z*Y;N(*r<*=;G-oc(b6F%XQgCm16%%4^M< zfg+0=<+lrO#2p*CyGky0L9$cK@n*2Jl8!Ej)1uEVuK+ZqM^J34;3D9UE^rVYA$QzT z%R+H|ij*BW*bpO6lu#zwVRiTrhWaZig>TX(4uza~*Ptlh9OAJo&?dI1nA$s$JOfwR zkZ)4JNAtlq7tbW$M03bDSt0vYU_O>-E^San?;kx@=wsoTJ^#t!Si;lnBY3i($NNJ$ zyBP{ca`}6x(K)pIQs9TfapNf*jz=x~HV()74flV8WZ_izEV9jdk={L+X7KYsqEZip#zcn~Aj? zh*|ptG&X`d$%@*o${{)V$Ra>u7FjJSx9WaWF4S!P3n+B5Ds&Ft$MYR5_)@8z zeukRe$v7IiAkTGt@2B3~eCMfog!ksVNb7Kc$oF1^5o&e`-!;4H!gtNCTJv4At2DlA zcGZUOnq3)uC%Yn-qy}{RDbkMN%QPGP={_QPH4*dfEt-GYQ`uAXNWmzl#f-4gqd)u^ z`r*+^qWS;K|EB)O{0}+)NBQ5n|3Us|HNyYw|84wl{jc*s3=W6?c^c<`_xzXq?<2+k zo^O=@F%*QH|5ZKy*H-ht@r{jehyP7GDgT=%{BJ{)|J~I%|7#D~;1~EGay7#Ll43In z|3fVm|6>O7zecMa!~YH`{`WTwb%+1qy#fDY82R6y>+!!fn*ZJ07OKR|MA`7e|&fNAKxAR$9ISS@!jEne0TUC-yQzP z_c;F7R`I_@4*#1)BrE=Rlf(Z?*i*m2{|^4|;eRFnG59T{)fTg@V^@x=YQV+lK;J{_}{aQ@;`=>|81|w|B^NTyS}jz?(o06Ps;!PBK+^Y zDF3^&asGGxKYo$_Ay*^(Z$xY+;eV*5;(yFQ{?}-=WBA`b#s8kcPOKR|M5MJ|0OH_2j>>DR{m)kk*xUNbq@c#gFW>N{I8b$ zZ-iqH)j+?18;<%kFS)c&3-1MojFh$)VB*1ZNb-2&l9iMDA2P09dB`-nRhh=Jy8gU1 zybMc-gg(uvlYoZ|++z{1h`2tYh@c)#veteI++X$-JGs72A{8yBHlSN&yiVf%X6)uY z*4Njvn`1Q{)JY5nSJdm2zfY>KllT-3h|cSwW@|7YL?=<%LXesGsz&C^vSNwVN!-QK zxL$Y+P~(gBDxJiOdxd(h15Brrhw)u^a!gtI^{>y#-C#J48``nEY+VIz zN2S3_t*Kx@JJd~{?$AX8=02HVjrfAA7SyfUn!1;-urb_YEoZf`G~RMGnT?I2m?o5Z zG1_FEvtQD=#6!~@opM3Fl!qB}l-NyBDc_a`S$%clk@XTU zdIE{(;kb`0^tL!h1b8b&C(c0Pvfk*hKDd2=x_*q7FRbT2185lBxr0+vOju0w%D^dN zE`VF--GQH7fsbn2RPz2;Y@*^~UhI@k!T-H!Xm1MZmj4z3z%@v43i}9}rNc<)-Adoe zFn0VAx}=hGUBSE1S2NI8+?xrdZec}RASV)(SGkPgABuiRF!`Xgk)6{Vb&Kt+f1vZ& zS^MQC3*5)FjcAeH`5cQMHhP_O{eJ5V)(Ty8*;w6ZH@u}q8vXL84I6z|y+(tz(e5*t(?HH&Y1rs>op(&ZfmM&YxV)C6kX_o3ocqo808b;&nwcZ zPqsE}RcXCewO9LFjkjy|pHaIzm;M*+GMV%D$oY%y>LC@>?fMjrblUX}yaDQVZIyPt zj1B!ryTYIRU$*Ow$5Fc{mi!m(x`#QhL(Z6Xi4qEA;n3+(>snM$a}<~cg{>24C+U<& zneojl!m2=*nx0`MS8ZMR4Qk_4^x6`;=156CkjITPRo+qS?lGEnJCegX1NOTu4Y4t~ zUN5(9ynkPR46S+O&Hn-qrOY`RIb-lZ4@w*!4j#hRD5;w6-yeY#r+*7EU|pCI#>H~W zY{EQ@aVtFV?R1MMx{gYxuoR&!Pj$USQm5X8jEyi4Fg|G%!&L zR=I8J0hUEex3G1ID#rf_h%`cmw>lX!N#rpZuRw&PA}N0lw4#RoHp$59DDr8G!7?+U zcs9x}PlzGzj*^+P(VcoWI#S}*-SB}sa2Rt^Z9Cdo;OfSP^U+@l3EfKFMJaw%XuH zpO1OG>Z!AJ{0aJI5-dirs?k}5DI(HIYd5jN@#vI!E)7RGfKD@XQS~EF0-GP!ridL- z6t9rO^gJKfYhClNIGjQblR17H=lC?Bj5s~Z!KqmPW1q_U9dbqRNmhZ>U>S;EaBX(J zR1f?Dk=A-3m!$Fs&{I4<=hw$40~9`K=Og&!3o%k3l?uNRe6E?R@cA~L#8PMLb$Rm( zLy6hX*$$3{{T#}@@x~#3_6tTE?D!8@gFrF%W5>`;&ElaWKT)YxikYOt2!}5}>nEbn zFB5^etk1~d!-iizdbpXJVw5*UR9R_qKZw#ls?k{fCX07?Q+NXkz~ZJbSPCh+iXxif zFGnpnFo6D_3uyy}o$pwAdPgJU<91ZddgQgZzKqCgJW_+hJb+l3TUU}9m9E7SCN*1{ zMT`{z028N)z+K}AZf@-tl;m2-8plZ4`!*^@^)h({;*YcAd2iVP>)ah=(?Z$%t(#dl z1WvyKK3{*Vi`tuLd~bdo_)a~7L{T~Y$p?vV>5j#W9`ilUFDP^sYjhncnJPS_kUq~2 z7Y3?pPtohfD80wZwbwihJl(ReaXg7!TNpS9Kb8&oNLS{(1UZpF(1QXC(kIxwb(r$x z|6RSdJOt<)U;QuY^%Qgd898I>rIzS+9;$}(Pho2UItlxvxQz$f%?ubV0XW6h0!a{1 z2&7NIfUx=@LJmS;-nbm>PfkQE!z7e_#J?%ec&r*q4l5U18-()Ohp2DNsSQ;u2NBnZ z*c1gb#0M-Ad4Cf(Z}~?y!Ohe@C}I<+rVp&2Uc(a{VosH@SjM?*wqoW1)q>GlTd|ky ztS6PKhy%#F^CLxj12DE^NLgp2Muouz>T0Gybz$!XavzuBA0_kX0qY^GD_lOmBuL}B zyR#ftF#M@Vz?vKfhRm>XfnNKl*r%|l^Fj8iWVB=+n}I(pJ#p{%0mi66`;;G8(K3mY zD(*)W8_)MK7n4rDCZO!Ne2Y(-@2S{)voT=e^36a##Vo-$PDExK;gue%YUa=3!^RBE zOZe3{`4!t!Km$?vM+G&Dq3N8rYe z;Z;gMct16LVAu&I9f&5Bgb%q(sW;;Jm14c|2pDY&CNN&JKu;v*g<0mx6G?V2gn_O# zW4-2OFxFAR!_@p|Vsvn*rr3W%LiVT9Ou)wAZK?z{Z}AEg*fSOacVr!u|CeE7ntx5* zncta#b?OxF2E&g#5?E`n6gEH95nyndioRcGJNDvf#D<&!JcA>UlaQH+yC^>*>awgY zSa%788bV-fVdf8T3XCJZYqecx2q$}k_tcqzES%-VA>NH86UF^UE9a~)+$UHt9zBQy zsossejTE*ZZ%B`T1msH6e-?bg zDMTVq4EL84!xtb1M_ysP)GM!|pBxybrJAZ>DNmH8spH3%Q9b||vnS$54PkETL1IPJAQd&s>62RQ#LiWUpNzbO{<{50sC^K(nXg zN7$#LE};ujjXfBo6cdprvp@x7x=D=CVuoA-$*Z5!>?tTb*bArD`vs@9Iu@AL4oBAS zsk65rIItcb3~+`Sfp9p0z?BG$oav;eVff9BYpLCyHBX&;t=h9l2FfFnpIDc^=;wHT z{Bqqu9~BN&6g z+mZ(y%EL%Nvrgmw8ubPZUkGn7p-xJNkw2}3cy70Lgv?X4)a$f#ds{BZcR~m^105mi zk4j@dB5E*5VFV4aM&&UCQ|n+8LhWr8P+}t@O<%nNl7U_Py}0!QqbzEmXO#a_>aYfx z#Rv+A9?K9ScB1~tz`86Bz6qxu{v(m!+rdYqlCSD@^DSBQ(Tg00P_6M^W&wXWO`aDG}C8@i1Lk*9++-4@5vHTxTL3&U9y5ayjLFwUV)^J9Kjsa6ZF- znS5V`gD_tivmXMWWxaY~g%#)sIT@yOHzrpXz*h;@h`6myRV_SIuOQqYAARtAtU%ER z&yO4aeE<}x-WXLqFSO05NC@0oj+b*@;*D*%@{a6X&U8O;g<`p}ls1a6eQV4ZlVq6N zX?o*%baj~hIh2CK?6}u!47<9ujeAnz-^4lV{-}yJ10Ul2pQPxmxvccN*y$Z&trM$E z;AQqMqEhJ6sPG`%W98ng&YQb&aaqiX^;4LMr_}?Iz62l<;>75b4d)z*G zyNlfa`YQ0udx^6!lEG_R`?#RR9@+Ehbp7P9T!%pYp=J|cLt z_VtHY9;$;|vr%2JI`ZSaOZy5)Tt9(!u-1pom&}kqJr8N5ZhqZ5fBHDK5&|Ocq+pUV z{aviT=j-pe`g@)Jmg8ASZ|U!?IzAbSOTw$u-x>P5GoMIr>hGTVdxHKRqrWTkcPXCm z*L}}%ED-jyeJs%W_dG5b4p_B$`{kZOtR#TFr738xZekV(dMM}4S(qUqvu6;w_*3&K z>Xjz(XfL0bQqgn3iw-Q-fg%8n(}7eMGG!W5mg&HG5?F)#aaf=fW%31o2qSO_n-C2n zYzeD~h7q*D?r8=7ikS)c*?=IbuTscQ(SUV2HJ3ZQL4-BQU>{=HLzK9Inlz+g<*N7u zjDvWh76c{fvlzLPkpO5h6n{v-bfAYf(&++(amje{4{+o*=K7TD$L;VR1zYhXpR(hztuR246-rreM-Jv+$`RZg`mypjz1UcO;DZ z!-C1-qaVql6}eZaMA0^ofy-eK%~cOu;>H%|DI7wAL*eaPhpK2nLJ+4|BbSg82b%xu zG*t{i8mx6{6r2r_G?hBdJPE;)kWVL>d`NM(=wxLYR9YlWWdzh9oo1X)14jzFxZ+5f zb99FxuyLOZsL9Vf07(_o=()(-7*h&U}eP1gZlmoc$oK(*g)`AT-Q+ z7Elmjz1c^U%^vF+B#0)7my&0VRKcDXXjq-~Dav5&bLq*zV#l% z!IA=2jH;syQ2&zw-r=>8qjD#r;r9mzjZulM_^=AG(Ict8e&Ixzf`qE*_}7pJGz-ZM z1owt8>SdS6u)DAa?`Rcp^J2lxCAbr^F?GQDrZ=_bz!e7r^0UDnYusY2O`yz$%$v6q zhY=7I8VD{5iq7LA(o_`_{*Ky~65;|t#C?3@pprRFDuatw=;dMzp%iN!3RSgn9iXrF&mU-gGus!#<$Upq<$?t&HBh>2! zO&vfbz<%^bgAO^Cf84=$OBBt6H9)F^biubYw}F-VmK-uR(S2FOg)W14w1Doi$HEZ%^nK@E!P+ zuo&o60eY;(6B|N5m(WKF^a9nv_2E|$W+*2yw)PSh17EGmM>PU}6_o$>57KJ_KYq06 zenN{-z2E&S(oYV_kH}L42c6?iFO8DBL)+FA&>A=t2SbAQ5A31X>N}Jvmr3TFJv9dZ zq_|je=FS%U;@!qe_7oW>v*x`uy#a6;iZG$Wq+SL!!RYEK!v|b_ACO3o16*W0@)wxc zicFbk*Va+38b9tu&eKEFdIK5txZksX4<`jp@5lrO;7%V?uOuHb)Xu4Fj8-%H3#~q{ zE2=mIXqBX`X!R!2>O|72cWixBaoF!<3qv`lmd4;o!eV;bLK@uvdTT>U+WHFc)fyVk z#wr)=CzX36y$tzd1`@cNt{euRkqMT0vdSG~MtC5(JEn3H7E`%8q``Qrv4-ZdizW3rX=>6Lfia-FV(qukP&L90s{2YKSZ5SUnu}5-5Um=g8L*VnkO8BW z){0l1lFB=x(8hX#^BNj^`&}`Om9Ut`CJV)WHKAc+hf_U03!1}l>gu6bZNB1c`AR&S}x13DhycJn`Ds*u_pV7UV15U8#@n}huncLIT28d$$TQeuu6Pq z=-WdTM3reK_h&N;J$n$-i=PmgkB6d(*@5SD|`#=3Gh8?zSxWrZ*j&BTq$ zumC(&%B9K{4r<^bvwJvo{s1brI;MlT{6$1?XmA+lHT-u9wySGDLVx9M#%W&c^>3zj zezs~Y&fEgwS1b@Mg#Qic(96^*9?4&l2W(o+(H@)kF1kXKT3~x|2>O^kxBn3WtTxIo zY6O4@*O(4~wLra{tP`bh2%zJVqot&=lzB|F3vx#|wb>w&ay>@m-Gg02Sc7A*JXriB z_$f4T&*KFuAn$-R=?(`a%S(%YJ#bGkVh3M4}hX22x!$8Xf+F+)GE z)Xz8Z)boLI-3B!>HwWOK_!FFO@O|6BNr%UpjOo`5P}jOzj^WcGg0C(~jaD-^oAKj!S)E^v0yy^BCp0NR<|eJb_xh`ygsH*OABBE*SO@&h@td&QluB z5e!oY&cZl2B8dRbY{0P#F_>i^A_mTBtj`Dy=LW#Z50iK+ipsS{MfdM8_ zpT8V525|vnC6sgV?N4I<`2higvX2yEmpZu`yT!ZFtVlV!AMRj%zdfSb-8Nd^k=S!{ z`e6e5VQUaI{hL4#I)LFM+xy`}T_kKX)LhCw!S7kRwgOc7bGMgFpL}W3r2n z3j#Xvf`F70vX2D$zaO^iFZdFC`C^BbX_r8G-4!^Y{c(dg&-zwL&@_pGBnDyoK-?Y; z$r&>ZcyWex-AkA@Lpj67Q%XH*rTw0Mv$E)lv4|VJV`SYbq6mzdDj>iWiGxkRgbR6W znvb#GnMKEr-ZW00?G>~p8rmjYmL7qIxt(j^Mqx}0#BFoQIhA-@&O{xb1XNgS~@{S}y=H!J)k{8lh8C@||Zm^#X7p`5ogn58*!h;fCk z$vZi6t|q96>(E6vNlhrwjj`?o9H)oG$ypItM{{~Wjp?8+DnF+gRNj64OyI@CR`87FVT6HyTR=L;s5YY?ygRuH2ALOUO=xiJO>lK|jXQE(bE zYnBZ6%-2Tj%3YyFWYcIGgT|no$j!Xy+Fm0*y87$rmAi&xh#q~j!{#ZaB{d~2p1O7#Y+v;k@n%C3DTd|Ww_2qPUX@ttlaZH>!2%0N;jsl|9~nsZnd4z;;`oX#gb^jZXFed_GmimPmg{;h&!=fM3&H^(`+)k@`wMj$uD; zi}Yif!}x;N#`a^&OMX*7zCM-QPZ=r4=-Nuzm9y4RC~*XOT!!N7F^O_s*eV5Z-Hv)Bz-8zrYwA07ky4(x2Bl>D z_EJ75Z9uH$S3xXnSW1VOQhKwLRTr@>ziA@zpcH!{1YKoMPW)sLKd;iVLE~rZ)qpwa zx5Ml}n7k%cug^5zlL>cp4BTmi+v&H%ed}((9T0_!QJcYzwbw#|X8jvb#50KXSfAZT z3r6&Ry?F05kF^|e4I+&4ATmNl^2lhaNH*TzNE+{Di2wADjDHRBxc}iYrFWr37a>Y- z2c#M%P=b=mgH++v>*aT>{Ej?nzIz4wRg&)#Nt=mZ%!u}YElQbEuAk3-qQ1}5&uaa& z@nrf=TUGoY^>e;{w)j-VCulft>G0R_&3tF+_#t>wi8EY3r|aiD{ao+FYd8n>cZbhZ zzC8WBQ9mEl&sUwNE^iv@8<`Jc_6KHtTOnADu@Eigp1>{bpa{b9Aq9Kq9_tL8OrWWA zXE~Js>L`YU#a-@ZvnF8Wn~6!2AAbFj$B+B!PmzT;5FP`_b@u1?S53Z=vy zI`7sSo=gZ{=sTWb+&>3nJYczQO!B)EcZPQ=yY1u6yyq^%6i(84=K3e8dMzN+1Xmo= zjZS|`rjQPOtCL*0JG_e^@w7owaC;~MX|#ejERK|6%jdrXb45rpb=jpLsE+56*o|_z zk8su+A|!^(5$=+Sw@mVuUUb=|uH5EE`DP3^`V>!~E{sJb@{|qggT^8HE->U}NBM@H2|>8b3c>hxnYBcqrU@FkWff`fgx+YD_%xcZ$jXPsT^~vlafS1#;leX8dMH zPk?y9-@syoT*b5_k0Kz_5V>CKaGxsj@Muq_^3sND>I zxnP8fFOD<=C#c{@;W2zYgBqhoEYWV+=!y|)IYImsi z(hMB+?m?mX8q8ROFqPwID~ygm!zN-lH9eDN) zwC`IqD+XtsgUz{d`4$k3S!x`_s&}Z!WevN6BoHBRb|G|DeUC1cSU(b=MYum3^!w|yOO;p~@&F51s<*4rKu7w%=a z@W)Hy!iyr|T$)GWT)=SQU*RZUL%DThoN&5>i>BYtaN&<1#f85e34b##{P75!XXC<) zBjNYQg_-MJ8Fr4#@JvuHtC6c~>T=;j8*cF`L#f5*waDijQg}=sd&U5xr znWf9{nh>xdMNY}hJF+&(doM3wJRNyJquvBu_?sg4{c>A3-*~g6E%fgttxni=$Q_|? zSI~VCmHQXeg$l!R*X^f>re6QU?Dc_^~tYoVs0cGy5=PaFGk;npiXN19|e8IR1i*v!F?Wb;{M? zK7Eo(Ztm06D7Usjwig%nW3b7P297PJ5QBL{9Jk(ZQZQHUP4jOm89*@>?2b$zx&h}s z4^%rwGXahtOeBL6YN006_H4%DA<4DhhEuzvqIUOs!x!xe9E=fOAeUexr;87l7q-Z* z())P7$6Is{1N?8}Hybg|R05oj=~VVpE90@MvY*--kF7dqLrhxM=QqGx`JceU`JA3Z z5^?li-;W~Symc?~-N_98UZzW05wG0a-2C@WvunYYQq&Ri3GZ)gB;LP;)Fw$HLtpyy z|8@E}bfG_%Pe48Fu4g=&P$}hW{CQj8)=6|kC2n8w;sAJabU^I^kt3U3u{)+eena}~ z|1f>~|1f>xf0+Kuum7|9ul&{Y#+%K@NAdI5Urm45xTCqzzD4bR%EQfO{fFtV{}0pW z$EH{Pec5SFf9H?DtBv@>@Lm&c!qDw^R35FVc=AqM_ZB#*-X7rx@tM&U2NK;T!_hyQRx_>f11_d&qC z4%71I-j=lmaJ7>e9Ej5%drjl=fhoqbD=GQy#vX-#z4u&*tc)l;Xd3RV;A)VE zwwi%MkTfNK=sFzlxUO@_S<>QF;_nlVe2Ty`SJ!gzD|vz&9>|;hc_ep~cM(p++q+1B z*doRCA-#vBl$?rz4n1WUpT0FIj!s}T{I~pG?s1E`oR`Nh{WXGn?EX{qP*>Qzg9(9u zGtJP5i*OWYuiF^B7j1GI11!jOxXGH`L~}W4g}}G+E0Ve)_Y9o-$6?|luF%A8iO>y) zQl(;G_JPb2P?Qx$QP9b?7ABi zm@cKzo4_l&1l*b$>RTSG zF=oJvIu;_?NTh+2s{ov1;-vT;axi>YigVPMf++d}y@)OXe`Dq|Br^~C$K(z4u6>XE zEO&d!Fc=t?bj0;ii6t4~)GN;Z;*mZLPns{AG3GOtVfo+(a(YOKB#L=lvW&JNGIWwI4>^amkIifwN zaO+r%;^Yita?<(@G7_Y>zZyu#9g=;K94X7DUQ-Jq1hz_+fMtj^Z zV%EWM)%&$8@t8YsGjOXf6(@Z=^)y3Ct%cvO?ILWW{T~5$n1;LEyNd+XQXGYx(h(ag ziQl_Hb!aj+<6@favFG^8Ljl9ZZU7rVs+t9rhxv(p3Vkmgq*XHVSeXDVDZ+ppk zzTU09Fj;|RHdwuF(iH=u=<^5_IT7B#T%W7_aFW`dC-gB`cbJ)9c&ztNA$;mMfY?+< zrIC1mVf!^iLE=#Oh}N)D2wJlO@QHV}pb=eAO-iRhN~83U(PskzLigIjSbi8mX==!a zcY3UrXnhpi_11Tw2FA+C)*oVod~O0*Tzg ziN*z?E{GeUt_U-Vl_ijgW*kSU;?_-DyV%xmH!b1@2|KPpky33HSMD%uqGH5l{_oE@ zca{LQwBK+4ukY{m`to{_x%ZrN&+?q-Jm)#jc6O=n&(CPcmH$$G2Nu*fvL^mw&AB!4 z4=p4?NQi23@Dx++s#A9wd0!y83F%W}WIql$i^hHZ@S%p6jULv&t)}_i;>ZC_Q>%(1 z;Bm|SB05!z7=eYhh`_>G=?!`io>gOUa%rj5%Etys6 zHiZAz>;%tzyk!I=3i2mpabk*IprZ$iSV2-ai^zu{CJP27b9kA+wN2Q1f0LB@#57*` zss}Hj#KnDJ1c;hN55kGd%iJ|x#sjribzHDT+1!Pe4snHv!LdJF; zn(`pFwW&i)MPsrRv8507D}ohxJ@q07Scx7>&G^JPHF`(-1@@L9NoneSzlh(KDe}dB z8qRAGQ=}|4Mn!TGp_(x!289U{0yk;6ol<02B~Ptk5>9aD9B;;Z)*$MD9`Cf2r?XCi z&b*IUC|vJE$@t<%@2K6EJ%_UM2%B3(VK;FYg=YR$2`e8C|pyu zDZaqDr!82ufs>;_;v-j>y^JNf@=7g9__YH0lKNl!WT*ar&$hBh{l8-OWxu5VC;#{Q z-z(ey99uvbhI8CXd@nabu`^}P9fu=x!hIza-)hkRQ1kk}o4lylU5VzozPQerQS*n6nKrZE;+`wRs0>hlY05U(h+x&r(+AuimS(#m4^Mv`efJW4H2B zShT~Fy{!DT5O^*%KoLM>Qf}-V96b;MjYTc;xG!N@Uc+NA?GWl{kxs0;+-(aTW##v!5NMWL0S0rM`-9bZs$|K;hP1^nSk*!8^-@1Bj^ zC)~T}jO2yA1;AqdA6B(+T4{W3ef+QC_^Q_0eKBWRJ^e*#hgyC`MKtw=n;emGS9Z87 zxqfa%^w^ZBweYsJi<>+P84kDldPBH+(PcI$CJEMg`ZV57Yx`p0xC#$dbkT#abK^yfw4TTB&6m(hF@s$R%Y zTBvgtXQIyIU=R{aEpw~ickZ~wGz$bT)KEN{sjYf$%Ja3Y><;V&-xf*Ayp-?f?INj^ zJ_*8=i5yYex=;EI6=>bZBo-8u=rmOQLG;Zqr@&-wQ8BaBjdAQvQMfwWPq%u>l;-r4 zf;Xt1ZUyCmAaJxmT!)A*HAj6glk`an{)j2rU?xye#85wUac+(A6v9_zm7DCjzcZ({ zlD5ub4SrJ7TAY#wGru~}NBHYGdEEKOtSLH&9v=^J6&B(?xv$xMXB~Dor-x`A)au^H z*RU0cCZZEGIt#d_P6Y812x5(;v{JcIr#P~zE;ScH8EeKYq>{6!B%JuBspKcN5?#w( zKvlm+CH=RH{w8}rUrCekEv;P|y{&X4-PGu{T(OD0YdU{*3KxU$AUWn7{Xy$}Mfu^e zp7P_nl8SaazTY338{eGA^5g5i2jlxHb?!30>!~C=zILfbv>$!x_+G2Xjjzlf-_I}C z_#WhMc6{^cq8Z--JCDx_TOBda+#}cUh0BVwA?TQY&>_0lxEVJUAAgK9dqI4~!v4$8 z=6HBW2xm|hEen2dU!EJiU;JdYuYdFRCv#FhSe2s(>H=bUtpD?i2`Sh`e9|2g`5e{8 z%wlKW!1(9H5qs-10xm?$f=!k%D-nf`73RH zSLH`{lfPUez-)qhCY_cD>nj}h!jd09Bc8#3I-{R;L@DY{Zr;XkTww9}!K0Pin zuIbZ@BbU@yzuTGNbTbP%%ymx7eZS>RjWRMmlv$KZd(EUp_g04QT(6mhCV#FpoCkM>2^k&LZCi*3sFUG@S>A^4*{2Ghpx^WtF0(;?_~W9oF`^Jax^FX zGdYHj>TsL4z=GQ{@dkuVRy4+GS#q4j+0?@ZKCgqX_~x#4rmnlT^5gU>bNK?JUDL0G zsXuWRHKs}>!+O(A&Qx(dzp6iNB&Z&RtaN%ct6Mf%ECPv(NSr{T*%AvzguHexGp&Eq zMY-5$&4j)PQ+FmYcE-lFf}R96sXV(V&tzVh46{2<}+=VVx!9@u}o-Lq{qlXD$0MXy`Id^}H~GN-as-S^0Bb>#84 z)zFr<{@Wdb(?4S7oThl5t7jaU^=<#QaXdX_`ySsyr-^5AKd<_SXbPD`5d6N$Q!LR2 z{iz5uf#7RD6eo?|Z+~GLyN)jz5hp-3X`&7@u%JRU@R4oGP1=A1Mz+KGSEVL6m9g?B znrigsN)`e#xk=coDb0&jY%KtFo!VF3$`72@ zDJs+aDPc|J?BG_v*^zW;g-aqU93eWAK_!z{ChtR1XmOrG{ zlHIy&GRP)_jyEip8OI-sW!`}wNJ`l4IN`FjLQ;??aH-Kdy|0KnX1Wq8^iyt?avBT( z;^&qyC*7v5<2&Q1ZR>;d8kDjNJZCfFbNMqCi5!IETkR(6%LoxGU z^HIlpY8mYGta)0{X6k=>CX@DkGVHwg1C{9|cPP{qbsNR$dWHdZl zrLaRfw9OjI-eZ0jr_S-;ll*4sXMsJe;n{B03UmkFDq+79tHkU9uCO?q_+^{Y5u3#C zQwjL|ieb*2Sl?;`xJ~eG-!Q_(WbdceG0KD+ti>0Q#!2GmUa^f5dmjT zuh^>Sf$8%sT%9>*b>Wt)E@<{^t5!s}!UA6Ew7gEi=BHcviwCOsxuwms_}hp7ZsP70 z`bj>$jGvmh6Zo5YoG*!E3L@w-hy+l^XhUQxP%bx*%*oIB<~Z<2NM z1_hj3f5RikkdhG?3lkmxvXYr>%ChZ^(F|FghLNwH3Tuib12RF*`)|)Hx|Xria2q9^IqVH*_uF1Oq8P`H zk;*Xh{jjPECx5e+AJ-0*Imb!JG*u66DK_C~7s=fuakq+vtG78bujKu-)R&uGnuNvS(lbtjiFf`@zD}2Tnt$SJ;oPfyDq{GPH4gT zXbI&DCm+-7dow@Mx-VQk(UVdVcpG|H+d61ZU}0F(T`-fT=c!Y>gGCl4H51|LtxoJV z+kK0XZ%w+T#w>h!&KTOd#e1NS*i17msVktW79#uE&&)Jinxes{Zo&)17U;7}fttD3 zl9oD8jw0FhD5h6&{nlqzrn=+4W8?11kl5D30xL?*gqYh`LrH9Nsr{(OCo}3T`ZL>g z&pSBZVN7%VI`83Wx5b|6-(QdYH%{rm|J=4RFPELHEh5rZ`M6ZiIa8J4(ZQKaN!IHQyd@0Ws7peV??uTM+~b?Mt~~`$e}A` zzAmOopM?`Q!ri=D*q3dtq^KS0OS}7|uO)}LFC%IQT?%W|bs!cDE$mslGgI7xB&{va z^Dws`wbu0~!t*9|pr-)~e0;cggWj*sL1u5K*H9pHWx$CI=Q~sF!|G9dE4m~NE$wjL z6tJ2Xpo?7?87ZYW|3~Z3*g$g?@ZBFJHnm(Dh`s_-P~k2e8lVoKF%)QwV5(ezu{<)D zFE9W$kRK6Hg-)*#>!>vP8}})H8uoUR!&b=nq}|il7vwHYaNHwTxXF?rmnM{ilC|B& z$~a`Pj$&ImiY<|K$C2U?ORWKkPAnYFUjFbd9d>vYGs*E`s_XQst4ehfFP4)L(4}H$ zrjhFwf|@%T5LgcihT7Mc7=vVvI*s1aQgNnHSBxo|ERNHdhy?qk?xsp#pNjpZ!Hea2 zNQ^L$P=-L^#O_jQ;A5~SKEg+g)XnPGtk8VeEB%|EzSW(ZpZs?2QfLof)zL3?4VfH+ zyPhP5@T#-SVm!WzWhWC;ujNmL7Rx#Lt<{JvYH}0TRTP|u@v+7XDOpFP@!`yswGGM| z`&0lM>^4Zm87KfKacYt~co2pjYwt20D@f&T0tI;@rH}eeN?4~xITGT?&EfJwF2Fkp zO2U;WL4?^0BggVowojy1^CBmu65mN_lJK~BW0UMpj9v^8g3655a(Z5N-3@u_VCslI zqx^o#|D*B)`TX=xET8_QI>kQ+Eg;4i^_B47+l;|U^>a?_NQ3$8urUS$j_r?d)S3A^ zy(kFIdLHWfztcZ6B3XN~5#=l$5$}6)Rtm5P)SLblBVNCqlD>&j>IH)is*#0INmvl^ z{M$0U14ZKQCo%EYkyyFe=5Zb@-)2y?q&)o$OQU$=i8(2cFHOEe!$pMZBG9H`L}l~?vKXf&6}#HPGFA=%k=K?DW|EY z?DrX(G*^^_ljVywg7j^jKZw2z_O+ZS?AvsP5kaewH;`yt&;8QqUYj>EsRfP(<3TY{ zYiUWjFcifH6DDr~wU@;oS+*fA^ALUk8;f)ZShC2(9Q()wT6rGU!c2cZgG@(@&PRr* zkK=Q=6ulC{J=x$<)((sO2H^>m1U$^+i9RRB7ul@Nw#X9(PWkNgFBqF=?UUDgr^-&b zQ%jM3GQF350{G&K)2~={$b>=_gMo+J2_Cv*B>pzwkP2C`Ae6Q8U~{|*I&)y4b-o%Z26`3b zBw4qNyVGu2Dq@y9JTA`i)5(X$U9*n86RC?{{9Z_qJ-B_YmqenC9}~_a)DMf$?X}md z?l(zn-cZ%FgUY4-F5s5G^c)vXMq!#EZ%5hQSf-g{+@GkR6U%Onal1LbM6TI=emVvo?deNcV}*Yz4x1*dFJ^TC`#6@ zY3Fjy`Eu!*`Tf6;pMNsP_^yS5TgLZHs`MG($myFhvkLWJonB4(tJ(!rWn3#jQPdD>RB0bO09SX4A z?f#rPo^74g&biV*6FKv?8lPVuWHloSt*!LEuWyCmSm<$;k1u=LdFAgOc7wP-ZfDr( zY(6MZD1q`}_e+WQWwx~+C4k3;PHeRR)%1A<7GCzL(uu7FlB8W=<;XnU?8G8Ui(%?4 zfkWJDUiZeJi=5aJQp1Vg2--~bN9fxmBs-~^WF%xTG_#%c!`#Qd_S!9 zQBx)^ivgr!ebU2n_^gl%r6Hh-WbC-A1<}nl(EXCQk%OfVpXE<@S@haRx}YIP#MIo@ z&F1EKB)cCJp3DhqNV6XsWU+X%h%R$;WOBsXrca7**sE#B&0I7>npwS8{sZo&-ZZQW z^9i;nyXll6(OWnGL6ke!|0zRu<_s+!bcz$3%=4hZHh|%vBN4sPwrFi9nOYBOvj=LQ zO8EGgx`T}TdR%Ja>{6jG3N(IY`{&oSn|OLFnGjA-KgDl6QXFCDLKR)xC3i~8tjnF( zE&hsCEdUqe%j31*55Kb#3S3}_^+g2bz{Id-Ose%Ak(C&Apk)Z~2lAa+DOA_7*zN~Q zOhjNrIR;F26o4TsXw>o@6WQOkDmEx}s)4lM55oy$U<=IGsjr#5CW^7HWS52*CbI&C zQtTNTMDqKaT)VbL#JCpXb|$&`EZeW`1_#O_-~O@z*Wo$55Agkx37eprJBdP#Fnc~? zQy9QYPTQ23FY~rW?>rtk@5lOP1puEyR!DeQ2`cSvAYoR){@=W@PnHTK?whN2STC`- zkO$Tv=>p8h+LRo>Q>bX9lQ;AW#XGjwPGduCnL%Q(DbqJ{K_Q~2!5vaBbXrEJCNcRn zJTy%%p6axSQLuBvw?M3|iv^2@kj3!kMSZfUg4E3BA7V;(%JBCYn|~ z44;wTkU8CP=15LOnz<m~@K_S6eHAT7B{M`s$CJnV%S7p#ELj#c7rO zetrB4EV)0q3w6>WRKU@1hiykyWvIFh*M&s0;Vj}Dhvi3o?z?xfMtn}kw3IsBEa4z= zB(QvI*^jsRM^JUmysz>Xp^~jVTd~nEhs}{*_Vc)l;3iVNaW%xZw!AcbOP*dZ)l;LF z<>$0uK08ljmM`EQEF4kFQ$tLLX!`wN;U2|_ZgM)5oK~FLOMTSBYV^K43^W%3YxJ(; zk@L@kN15|aP2s0zAN27%{*er;EsG22Z`HiIOK;V@>K|b=4|kJK2xvxxZ}&of6)46g z$G+$(sRx-Azu&d3<&`=7*p*pJ+?`uMSMpJ96&qYJ4+SUzA7^@z6ZQvmK)`+{S}ZA2 zY$lxvC!SWHK!Ztl$hx`pWP=yn?fHc>)6|FOxxot|NxhtP9{ZRh0Y-S>#UK8ptSCG~KpDz9)fkIpqMb?$| zF*jO5`QvxNEUwa2$G*X8x#@7iT%RiQ1K#^9l>gzU1fAJSxzG!I2qmW%oA9V$$xUAu zmMk3tmSp#j9OF|!@@AWfg5F8}W@sI|P*F`&$8k>VL%Pe;1(7rO3kUK;`wn^gHBZ}b z6+1Y~iT#Kl(9APG^SA1!^VXPIVRxt^dG^q6Ll}>yVP{T5anlZT=nMF$X$RZ6rnYGZ z`}sOmlekYvTQm0p{`TL#^wwGY@|jf-xXJmNaBtk-QE&&vhu9vKMMpKwe}tT($Qilw zlIrnyS)WeY#j4u8a%w2(VUca9_qv-r)H)Ur@sTuFN#0seQwrU1y#M(-`>G zg3_wb45%07(=Vz2_FZud)AQz+@PpEo`C}tH^k_CRrFcjFG_PL;=yv6qB^>FOxV4oc z@ay}@%&ZsoWs$x*_olj{KAAOcR_ar>=2?0AW!pQ9Z|uCJbWL+X{z%E1R_PCSdoMl* zc_+rGpSp9$@Wny=huwSvT15Yr@S}3inNxMVhdZ~N{DW0SGQ7FR@N}Sk>*LNnmZ#qK zE#S@_!qfitt-_r2VLaLmzIv}d9B`DckaY}BYWMzbnw>T z&i&fgPB`Jh3opTWclc;`t}QnaE79R&Wx+E1YBcwK{BEvwmkuvGo>pXD6TeemMSDX`)C$;`yH6Q)zbtse$4KJFCjp|cy zMbe+szv}(1C5F7I|1#HVqqNU#IW4!CPrJSRKAZB==B>QxpTS1Mw>El@8$VNzWd}2j2WWTfw=e-3G5bU4}N#9)GMN(zg>jP%FqQPui8a^w|)2_1U{CU zz#VvFg=6V@f_oxIu~V&x9G=H#cVItz$TGrXcCpP%hW9J>#e7ML{@C7pxgH`fq+mu> zSs)KADfdNvM(8JbU&p)uRNi+!u>7Br_pvMHC5e$6GQGWbt$goDZJ7BiJb9BRUC1QN z*ATpYek^N@H@rjXEzzpI{q@`VgU1c9uesTfcNu-EJ(gQsuE{!Zpw1ok;s7Jx=h zXPCAJDJo0MzpHH4wzEPwD_6Q1wt7}Z27jUgE6(v`6ws@zJ?DMtxdxxaN*Ad63hYmJ z>VLrY6v5F&{nXQNcHg);K!?#kEtRP!52mx*Ue#!`Lvdn17wX&L1Z{~Gl+bYIslx1V zg4yBxk=#PKoRHn0^ObTxc_;Wx_bc2zd@eB^!i5s}oL&eYmyFKv3DPaQz)iFnf1J(6 z4%TV8nT4AEA&D`DWKXSo1iy8E-R=C7S~$*sH2)O+2lH>M-_87!(Rq9U)$*^;zwPOs z&i~qfI{%dT2lG#EAzb!k{{0HSWd2Qub~pcIbcWA=YyJ&Cu*B^1Vg}lGr)`1SUM@{D z+`ZAZKx8@hLg26C`kMIbt>N-MJYQ(FbSPFXl$P#t2<@Hd9&?J7ql{wZ)SW2Rck+bR z@{tfSqgTNxW>>um-XFf2V(!jrieHLWZf0u|t$a_vok*h86%%CSm(#0Mjbe|IKkY|3 z)d4Q-NBeZq_3hP<=%lZBehp|MeTBc(N}r7>J1H;wDLHwBo9uhTId1#AB|x;IlfDKi z1^iFf@u24LJ_Jvwh#bWs(rum^C~B=OKiG*~43Rb>i%9vwZgOf_W@H7@z@s#vJ~;&r zF_%=V^YMcI$TvQ5)p&Q{RTG@pOH30kqiDglw1toTI`03R8teLq5TLs*bz*~=NVm}C z+?jb7Nt}=I+1((V=E|!0QsKvX+~v19`!B+G98dE)S~(rB)qTR%%OmH7@n^!R4t^KC zJZA3lwUJk6xgfukpC|9Myc+POAAMjB36%$o*M^mf|-TaN|=s ze+?U-`oi!abZF988KP&^uwCmb@oHa(eM2{6(NL}4b2C!xVVDav?yT9R&P>_2V!$BL zM9b^qI!xcY)+-dmmbOq_c>J<4s3;9d@hkv^E75!nDtBUM>4O}M-0BWzW{jjpuQVm1 zAt>NBw!}y;%vrm5YY0&T`{T06g<--L#<%798Q{YnnT3zy089ox-`zcY+WrxI2H;~0 z>thDHM*TH*7H*=HCPr%x0?YyM!z_YW112^A@dgOK5q*uTp0Im3P6Dey&i27$`7f63 z!y%yj1glH_7z+rTZT{@5lk1B=$1G)RIDXXx6WWP$2*&tEp`B7#D81+G4P2mVtQ_84 zFEAZgj)6mp8<_la?)H?T5VC&SiYGpopH3!Gwi1|TbPExh;wgM=Oi<9y$RDld$b16K zW~U_)C7{4BlscC^Pyg_6S;}o$Oa0nX-%(=k@B9ri(D}m+9Oe4&QXO6ngEt`V2!cMwtsA#0 z|AH4qruX>QR~Lmci)&hsudn_*G8B=}HS(D4&UIHrRhXg>f+Lap)MbKIuS{iYm3*aD zdhB+DD`?i1kv?TJH3VBLsX;8Mny=m%8NjCw|3~rp@J!W8CwVRsN5Rc}zr_E(YUPwI zq0G|O{pzbfi@u;p2k<$q=gFH>2;l$;a&t zlCSCLg?7J5vyT^r9PQo81vq0*KRkV0b|0~z{QQHijKd5axcv6&l+A*M!% z&}w{*o|D5T;!FI2Hj9gs$yzYefT`i&AJ70oh))aaCV1%z<6yi7`4i|;Cs&&9F-)Wu z+~w>hZ#Qo;OTz*$RxS~g0Nv`l@~E-dJ+B zU2WDwC^4!cgy(i_A=UEYnwR5ggqQK=i)v7$Sca z#J$*2^;@|b37gXrq>ObQNy?=q8>@DLJpjJ^0zI$BOvq``ZAuKF0qakf1?a@SOF}L6 zm8TE%@io-ExWqJkF3ufgdX@ldTaW>3g;zqfVygWcQv1=mGqA^ zG@b%RIIs3h@@-+n_f>F^;mQuf#!1jQ&iK>fAA~>cJ_@HkzJB0!rIE|)<1eS=6>pQO z30f0tHPgH_paRiqA&lah;E7?nnw@6k`JtBuNwOu3)BUySf|{3xn*Az%n6F|Blat33 zs_34tqTZ?dB=wRe*8~gG(Qr`e517OG`zxW4vO9pp;xc?UI{CX zA3j6>KM*`&W&O3LYUz~u&YW`)(hbYhR_&NpAP?)b57bosWy{6;h zw~oK>B5MHe%sCN9iD1?H#0bu`wT^%cPlSlCWBZJ6i@#*^qVJH$Cs8ieTmwx+gEP)( z^-2^a$`HxDR59o{)(G2D=W7DuUu1U>#FN;WrT)a9Jm?-rGTsSv&%amz{d>tx4W}BX z=^8qyW~dSf@M5PJl;;9MVB{v4=aN>c(e8AhwRpk-XWp;s3~t>X5wMt2TDh<7vF_{ zSVp?+@ctUAe% zQwQI@mcc~1@t8>u&G=lKoD*}DrSUMX)a6sbzPz(S$^MU?snO?-&m}j8VzQmzA7ZI zO}TG|@tzzTPF!9-UTW9?_T`B&n8LpmN}Nhq1Z>Lx%ydNdg*s$<|8=>x)#%s4m2c0b zZ%tU|JITMPmpY$K=YNKzX?w@lLXhvY`}mj&Q`rrH&b4uBOW4}J4vYSzsjYIY4*bc4 zDx8nz9h5guA{X(WKoG6DOG(8 znn%q&z{Hyt-c@8)tRcEr@fNnsF+*q*M<73%4!FY|-K0djqYoS_$9Wh-7D71jGsByP zi#xKe%9ckQ3Q>vCCBqHEC$6@Jjh21lr91t#JPj|hwi9ZUiCHwrBC=6m-B=oJ2jG#z z*~jE09)F#Cjh9+h@ghq2`;6!U&MdvPvCN`H24WHH#|tdOSmngtCEY9?iXT=}_q~(l``}JYJ3m^I+btvET(=MKyRF^c93=R9tgS~(nc02K=VS7pJ1qegM-Coh zpj;BN@Sz6yzjq?*#gFAH_)5Nl$P0p$X>tv_vKmse>F=5G>rq%~^$b7ZT-5-Tg!6kl z(pUQqu{XOoPaFW#I{^kF;Pw5z7+-E8)&_6xp!r+$HTGEpom3dUAUhuM_p2iZY4E1; zFkh&*#x0&+z)GGONG-Qrf|Vp^5CfA%4@6`^xN?U<8Uj9Ji=ruR-od5F=h!E-p6zB% zECSUJY{URe$4%i?nFbV-(OkjI@3UAlLS{=FtY1f}E7_LyW31eJw_WTDmF3jzRT(VO)-A+cs{qB^LuT12L5ih=l?w4SC3HY zCK0IWwOF819Tb3b@}Hg9at4FVH#Y0y|H`xT?PeCVVi^Jjq?Gu;t>Ob!MttA`62%81 z%tVZg=tG4{6zkJ$hBY*hW&GIsot78i!&0B|C%Y20YMfXTZCif2u{3h9o{Xb=t`N|U zauIkI(sM1Zm|bH(S(t+~_9^70pQT5}PYGQ$jTv_Fi|t|;f2UcH`PDCY$kKpR4_@-s zf4g(_Lkp`POkV0N&_=7b6NQgn!Da4;S$?r|M!3A z-Xy*3+(_h{Oz&4ADPlYE_E?xcT7NKwSeSlNfA-DQHNey*-gn1F^_G3!3+%4S=zkeX zdj6?mUxRu6)hIb;Y!tEHo&4*5()+g=;n)74ZC3bo?XJx0m8@XPyxy)YP|Pcwt6d7| z<78Zy#s6NV1^7<@81Vmj#wy9zSHnoYPIcy7WCUMa++MYC$|5nXIFDtj{yg=NEWUO+ z?^OMH$`(FvJ%q2LFjR+QF1Rsr@-AF!xmYX9KYqE~Yq^_%EKP!VZ;Kw34*2DrIc{;) z2-wzR%;0>sbz5}3SPA@${M$*u-9Co?g}MC&4U0rMsYW7z!+`aylYQgKXitnZr=(os z96!h_M+5qX!TD0d2$0;vgncA9QsSF#{1dnGeFS2i_}`TwEmj4Qv;6N$iO`5&mn=D` z=9Hr8S^oFqNNH7f=Z?FWM|2X%Os+!h{h|~5DzjPJ3h6a=bHwA0NVG6%JV3if5fZO} zL}7L}PeS_S;G-Brqj$3u8{$|sWFw0j1nU7%((nVO_{SdF&_oX^gZs-c1pf)1hU82` zSzXnK&Yd>8;zv`yYY#QB73GrE^|Wkxe4DNo&5gERlW<;>m85?C>l z?#2oFG2;uvowp>+Lbrio3fQp@dImct^e$k>@K>x_G<`3ms!nn;s~o9JmvHsfXewJi zpdQZL;?(Axj=b)iv8epCN0oZ5zY=q1q;3H3_s&ZcgHtLKYnBz-)wcGtj9CJqE`ohuYKIO=f;$*^3GUvyLQ}C- z^O6$^<`}cyw7?A3nG-_;{Mif!j=t;h@+SQvn5Db7n^Z(xaAr!@p#f?AWH7=C?xw&F zqw_R}15>~-#_v7pvt;aDjow<`&;kkqU&4~TAO<$ik|~Z40-J_VvTu*{0Jd-TapZI# z(5Pi(c#%F%T!qPbKre>%LQeNGFL^oLk!AxX+~Vd;@7tcWtk-ubk=LhbS&mhhZ3e*; zUAD~Ek?Jv<}r^gamq6Dt}}vr`wFul|&%w%js}HWyZZ}zcgaH zS5pJYwfvss#4KrZP_WpU;B~M?&dr~)os6Mup5HS3T~4kCBACHE`;6u9;3l&C9bnNu zlLbpl3Ir8*mF0&le`i0RmjB#oX`^7iJ@`A1ge?6OUnKcnj=!@N49-_z`8&--%ME|0~>&?xdw0KW6r$ z*BAM6xW!=t^TjqG!h$AD&2CDDAIt7a62u)atJ8jz-g!WVkJfx8KS`j!9M|7;iOS{~3tNFgC4kFpZhqQ-Z z&?5`i-SAP1TK8ksM3;sV9MW!pk6P1-j~ZQEbv`V^`BdRzSZl<8hL4JsKO~^B6EQMn zvgPIIm(GEu2$jT8-9N@MROPj;Q_7ER^xj!#_QISX&4kS-2ocrfI4Yc8#8Ih3$fSs# z#Vrxys18kY9F?4>5`KzBxI2F8;6i?iYbl!OT~&|iQ;bsA#Ap!rVgl*0;hCF}LW$eW z%||s=V_;pTOE|IDGB?A}LUDU1lD4JN?Edgo!%oD?g&hTpsyicqdGSasU|xKD208{F zelR?cN!q%f@FRxpnTxFWQR`BYpJUnUAL6;uoAX9zX!&Bt{g@Al9aV^W+W9d|1pLD6 z@;1G|kyKHOc~06AnxkM|>}Uju^k*)`i_Irf%3WXBWAYd?DqLTrXe*M~@%BGx{XS#G zB$P2$n1=g!NVOR5V?1R-iPNe=iEI1G%Ar94+)?$zaGQ%`#>GkH^0u zfTkkA;=+`Q^PtogEDD5Z@BWQ8lY_$-ikteWY_Tn4ibDa25-*^2yb98odkpg!MhM?r?J$w0z$>47Yk~X$Y|uxm2PJ< zbhuDL*kBpaZ_{F<*VwtS94ERJocA@zJ}zbF7c>&*8$@z|6nm09JVBy3ba^dAMKy3(C1IDGsQ*U;7^ZSmbwW>)X<+HK5&wAuOBTOIrHS4e*qpm)b&BBv!;5RekH>U=3AHMX5TzfBN;o9;D zHK=+ulm3B!{?-fMqfh?1=|814KjJq>0-7C8%jJBgR-{j!iV(0akSeDGYR@DSUc^XI zpVJO1v9XZ=Ud5zXeluvFkN`fxfF+X>dem26fdug9W`25>8v$UA?H}`t>=&Tz#UME6 z0kp7FmoiSZ{<#+W&+61iuuj-t7VO+|aATbn?AzDEgK5*2sY!L(iM8|1&U^Q>S1w2qj&LcGmdZcPBR$B43R6>DnLrh7yvBZdjIS_YpnRpVx<0&B(_@Skcl9Tu)*RnoOvJf0;@rm-m4Zz zUN%NDrW*-rHe6&5dm~3?dOvKa5xiwCHM*hEd$ojVW=fHrfeRxm3Xd*kyg7SF8AEyy zC`3g)T}9!t@uWeqWdO4P_^cd?^=}Kpga_%6Y#mL&r7Du~Dle}A0S-B1HlXVEt@vsp z4D{bF?qeJ#VvkS+3ttRbQDf^FMY*b{=sB~o)F8P{$qC)wIst20_)^uXB8g#a?C^gG zvq~CtS{QAUZ}u>vBpP$rwAhzwlsR6R;pEQ@r=f%irZf+e&4Wy?;UOF3zjn7iu^gH$ zWG&qMa>0U5zJ9MQs#WJ_8&-6zSibv7QJtjVtGfjsbt-@KF1}y^^wUr%P|E*s{Ab39 zP>|3EW8G>5PTw~JH+=QuHXGo~Rh_K4Ds@5&-<5pJ2BQ?%kq|JXDl!E5plP3j@%a7j zF|8t@18>VtpUoTRS{K_8bKQE-^42|naaR%k7(GjESW`XX1^I1tgaEB*&4BCsV`dj^ zkCl70(9LN{k_VH`>hHF|jO>~0NOIu(wbGm03c>cv@0lb!*l_iFD`7NaGaPa3=LS|y z_Xt)H#vEW8y&ta0=K3{J~I8Wx8t ze&Sxiu_{ZygbaxWkZD6=XrJbFe`;P9j1TRT+Q#nXpD(H;vu)l{I_doF3vgGIr2)nb zUx5oNYQ5b|)%FElBKyWyIZuRpe=)8m9zO63_RCbzR1s58>DaFAl&YQEY&-FlndgAP zz8#_Xl%5?lHu2f!w>tc*m0{6Nt&f)ss$My9i_Z@|3th>iuHj>>Q{t=QwjYUb@6Xiy zXQ?ToJDSJpGNHuIP>KDCK4E_b6_h7TH~kr`P2v&Ky_fIPCeh8)0@-O^y07dsFHDN8 zolLCuZow9~I|ze8@9eA`CL}b%m77%SE`g=y&Qh!5dca7EU$- zg)QD`KbL$BADm6?nan)V5JS%$;NZKu>lBf=TO;M{+vc70V**U^g8U%9LLQP?eth@! zzxOl0{-o+JLtqQ+7<{lHG}H}bGGl7}gzs?joXb2r=O zt6?2f4=m>pc6h;d0zc8et?$H&K<;SyMGMSQv`C!G>DzmqcvvAq0@2=2gJD|5Kdl&1?4ABteNhp;eOig zo8=BItJBt3b@nW7)6yta!0VZ(s}q=nm?#$}Fsgh7h=209^eXbQP_T>Fq8HXzcOWKK zjYf;4i@gl79QBY^o-|10fgSSxp!R;LV%kK1fg3{0cIbhLe}ZV1dIhgi1-i&JVB+j+ zKQZFru!5l2c>4o>=?iP=M(QSN%<4hsZB+o|Ercb0pUU(2L?-1?x#0~!?2>er<6YhE-uAE?vZzQ=hM zu}Bf|GGq^G_!k)g#fJ)}J5_PNVMHUJZnWi^hFtEnysXY8yDeMrT#=$`6h(SHT?B%S z`#4xy#O3jqGF&*1)cI!EmR`>oa>>=#MiG$nsh%f{cVb_d=S#1?{o*O4#g4dm-2Q!9Tecg>$II&vuJYnK@`1V};9KH53Cw94co;2YS zo)sHx#*pZQ30GiI!9bg)mrQh87Vt#p_o4G{HxY+a75!`)a%1!Wl0R%G83Ji;QG`A< z4VmP$+-W~qYx4r;0s$(!#%bx#r+hWf=JZQ#G9@P$rzRP}L9l4tjOiuci8w8?vZ3HT zD%fqET2()&tnub0!$8V!jz~6+k#SQe5T9xD9yWPx%3Gm>btm+S4(9%^Pd4tH`VGk3yMKQG{`Z?$WoWkx zc!^os?F!LbZcQKkTpRAfG(IzIW6vHgQfu_Xid)^4IJK9a)WrLi zhng2**nL?;dWBg}Mk&gf=P3|G|Bk4y#86#^_?Qu`aAF71h!K2U#=>!v2Xf3?;LMcY z2<8Y#1bC>4pHSC-dwvi5E6sB&&OgS0TWel)LZQ*wK%OMS6t`iEoODg2)ZkGLQR!UuV|avgsz-8!zu znYM$cvsdwF#9RCsx{W{QFBn%FZ)h7=>n#3D7d&7cD!NsPt4+}l&7&#WVIED<&s6l) z+SY~zPUR2F+rR~OrrG6-^w8R{M2W&wlPK1i#OIXwP&V;VoA{FISw-$Dr*gKD513Cx zm`Plt#3ntf@>ZMps@w@DkaBMM6w17=8lL0fE&1z>HR+f0`kr1J%Tv9F?XWo2W;*!%B^S}aSXYAm5ZG8g~)WiJ6<_`TSpvEPCw=J zRn8UUoT!|Bts_n(XMl3B5qcqVIyv7`&Vbet-y)|tK)vxlz;CkNaE@QYHqUkuv+@?S*zPySG}|D=;&e}L#+-npX-U4)5GSV-Ud zFW-r{k8kvz`fzu|J>e7gBfgZl+x~tzeSh))===ZC_y42shLZjtegFTHzW@AtJJa`5 z{}uZF2T2G2SJ3w@D@ET&Z8r3M)LKK|BWHJ}@Af`p1(D%O&`xp=8TU3-5wCAaQE{+! zJmF_KmFEt<(JMIc=I^J@?w?_7iC~t-n&6Ge!S@VYQ&gXbvQ4ai7O2P&6m!{}IdKH} zg|0vT5YdK2xVXh^A%;v3Z{%bhVY;Scncm-hfTKs}rco69VWSSp!EF5O|5X29IVIcw zZ~ROBzXt`C-~XrX-(COTPy78J6+&~uByeZ>{C@1sHcl!RY;Hs@KJH;`1(Nj`<1C`K zFn5?{B}w5#zj{<30VB(ioB+GSz_bvm{XP+7MPXyt+~|G(1L30HswV`dUg|ba?t_P* z_;O>CC2eo~J#Lb0^kz~rl&PyoUv2i?siUYWV8a#UtE(=oZrTB%=IMQkA~$m|oC_~m zlpZ^77EZ#3B{tfCtVypWqhr+zb@$3|S`)9%fCk2Ae50|F;@hX45YswUR_0Vwq^ZJ2P`$QX3 zik1zt5mU9GAoegL1BlgNbJw^|xUP(^mmrqGjxnBASbDhQHPi3rCDp)}{y8^PeN7)6 zok=ibL$XAHM(-4EDzQUMfWZiX87N1Ya58dD3an(Pn?J+9fglLJ$jd4OJ38Jmt23(iN*k z$Q1EGS7qGxx4LsrApsl%?y;lF%zZe;E5g-ydi4TGM6;4^G%=k{9uSh7gKngvDvtbw zkH6HPg(r+?5c?M7o0KFtUmU6DIrL&E0gV=v3Z5o*=`ssv!chU|VL8wO=V1s5cDm@H zN$?FPukUhtvgBJPeA7}AP;cXfriaW@uBI8H{ zB(>73PY%FHcbJQ6h+roQ@lr|@7&Q?8IMe$ML#u|-cR~qfXX4)5q!}Kl>x@Ocle4mD z({^UE{CLhl6&e?WoHvUvAdEx^#<$+CDDe{ zrBLaq|`;JLz(u|lT(fdwmV9v^}@K6_&c_w ziMfO{kP^n#cz6e&8|93@ZtAPoO&Y-2=RIHE>a@tdgR-G`_tYn*gH5!7R$lk=_~}JF zHhL@G)qXnoYi4uZK?Qk@0{$AKq&t?4?RMhjTjkD-jpH>ZC<7{_q%Mpm8Y>@77I>bt z>Q>16WHsAKzpLqc)H;gtJIl|$U*-5`l+O0iJU^0s1_@oGxF*49s{P*;{_i?|xhEFZ zCX_slDSxzFIWC_CrS4L;ZXhyNi?{3pNls9zd9i3xhf7CHX$t<{|B8~H6Z;vSHKzOnK32?PX(2i^9!OF~^QEU!Wab}auV=~lIy1I+{X zaVj4z-%2A+<>TejZ8(+BluNVY#Ev)iz1PJ|cJnjlXCJ~k-EPY9@Te__o<3kce6&2s zZ_X+t=v3Zb-p^DuOo|VAAr3{B$CuC8^j=*}b^1zI)j7RR59sS)2-^*U0Jjl;U+|ir zy*f{nf!(hC(-Pvsy^*~-uQ%XxGtbjQbz>jmQ>(Sa=?Hc(l7gX?Zy!aY>sr-cT@YiNh}M($;)Nq z+57p|9x$_uax09F3t1!0K75n0Wm>{a3dnAJOa;LT6#D6w44rN6=;6e+^B^8~ zH16-RTdDUQDo;o>^j>>e*)z;AYL&<}cL}2G44h!bu}b5B&gN4!j^!E$;p&zZ zjH3g^l*Ym4f6OlUxJEsQM`>O7XpzbAEmmqG>zlJud~LRgM#TnO)Op?jDQLWy7n8!kdXkoc_< zx9eFLjuK(H`I8LyXGzrku`wd|>f@>Oepc=wF)?sL_lMa?Vlwb?1JgHz(51zyaGdw#KX4v7xu7f;s`??c*Rd}1d$`V`O>*nL_47?Q@l)R^9^-R=FBy_{a(rm{=S{gxAm zpHLH^DyE^P%#B|Uye9;pXXbuO5L=bSuMem~r`PrCsFeFHv$S$=c9(u8Xp*_#5`cik z`p5m29d5FO#k8)(P1bf9EB0Z5{%nPZSR(V>Q)Hy)8dHDt>rx8S5MujqRay`G|7#+~Kp!JT9hZONC1-fQItg482r;mAa19)B$pc0wFElgOC@o_X@9yI^k97 zG?H`kLHQ;=HVCW;DlGH#AqDGAH)!6f0t!sG0#eH&b`80m{0Za#Gn-?Co@5eP)IBXDrx{)|RYlQa4c4(j7y$ z_ZHoDk}EVoMzD_Wa=gS2!%%mo-Kk%gF$ZZEa(R*=GiQKN61b=L6FXFD>Q6l&`|{A0?Q`Io-Pu-r0U|j;x~}Bk4zdz8{;_kB`FDo3wS+2d17# zAI_!^wtcWLsP@_AWpl4yjG`AlmTlcaKjMo&aHiF`-K z_>lJTHgLhJ3_ubWl$o=0JpD5lZ044{7P2`?Hr|!9F&; zPxN7|_EmYPlco39=jeUvL)N~V7*mltNXopq>n!b=Zpa9&r>2si0g8Xfauv$N%>Iw! zn8?a4A}{z&(+T)C0tX$mNTLjM_lvq_2^-=1NBfNoo~t=z6{ug3_0~-n|e@#-g)|>XLGK3>RAa5 zzxJvQIhJ4bMzv6wm#uWk8=)85{+;ID$!E%&l)m7q-P9N%0^pOF*TEB~p-&6bzcEMt zDWBbIl778w7r0IJr0d!+P1CJl-jcz(L||e10WsBu5SJvzY{f=?1w~4YCq?J4_)1@* zmNf-8_iBxDttjAPXvHgTx?PXZS(%Hbn-2u1?cwx%J;@~d+ZYWaol8APhi4I1MEj~UY#~Rzt|f1_MCLiY==9!hHbw$0bSkPM ztH>p?E8&_r_H?pQw|ru9%CYtwevvtc*U}+sNhh=Nv-VAcBYQV`OV(JqXQLPAY1}NQ zrOvFg?ntjvr|{^PcUtQ6=_-;-KPj6&yhBj0MzeRU$(PtT&Ah4KkJJc$Er?LPHkgZ_+dURSK&322&ItqHjU7t+MH`km$fw)wvz zc8y-`8nn&}+BcN_Dr->$KebIuQe7Drx8f8yy4Km{KjjOrD>08XRBnPZ>b z>>=|1{(lkjmyi6nkpJ@^?HTg_+?-7Q1IYi`>@Pw7=9~T-$bZ!ndE{^IUx@r&wu1b* z(=q1-!~YcGzqTQ6+Wj)-XD!+PocR$=#Ocp+A$);3WM|IDZq0`C@dZ4F`7!%`mg(u` zGei6?ql!lFOfJ&ih3R?wA22=hnBLtoJ@-@J&P>lIGxrehy+8eTh&~+dGlrMp;dU#w zh&PNEgHy=#w8HcB8XvTb&SV&#)>d^n+3m2i|1Cs6@CSQ_=+Atha$kz*n*GjPj?iUK zxSVsoXA!+q4dQaXf5RRldZ%xr;&Ns_o=5cWA76;*hitl6dN^MN$x#Terf}kSV&=n% zUy9o+;OUWw3a4*1NS_@1#=&bq*osE)#P1{NksbH=yJ|FZm+$MjP$-OB6Piby<=;zHe+ulFtJCE76%Z$(noW1x?I-Z{n zmxJl(g1IQjIu8GbWk@tS&+P8&k-tN?sGSr!T{n6!rUlC;X8PZQ=bZ=n@Osu3D>!tmY0A^MNudBW0v0iJ)FR0z+h8*=a*$suZdLnb4~ zJA*31Q-&ctbrbVm!7f9{D0ZR8?-RT>#QN1fw=_6$!=x-@+g@4uiUd| zqhGw~KQ{WOCKQf-=K8|X>;AsvzH6Dao$&wP8vhCoKs(L;R}6dg&xyif6O?yPh{C^& z`cGZgY5ec)I{rsa|DPFuTWHtu-@S+PpZ?Fy|FxaQ|L#uX_t*bw_n)u-WVbsgWfr~J zb*Jf?b(=DZupOldC4_}W4_}Zkvl726~4DJPyZA6-hjYsCwxE7a9RHx z-(N}n$HxBL)rDif>+PN9oAHnK=i1Dl-S7Ub^*&)&dUxEcJ)8Siy#L(zuj(}Zw{{(W z#qj@``6rU?F7toq9*%$ge{TF&b{hYiJB{C8@3U+7WWArUC+ofXjy)UmoA3R{#=Pb` zg=7Bly8L?2D@r6`vf7n(C`!<-U_UDZ+~yTXX%B4T|16`fk)|c5MA-wqiqR$K^(vfn z(A%JcmnELGD;N3Nl|v&&yHbzomh8Z=&P6nz*RFK7(tTBHbA{VYiW#hQWtf|@(v^nA z7(gqoMoz!dObAw^Uzvq|rTZ_XUwQo*amXm53bas1aO&-DwNSwVE!6OS3VDIP#@;HL zmr&%anhkR}V`C{Ds!a?jIJ?D+8D9T38?~dRd3{-P$AOVEn>%_tu`?*$+_AqC>xNps z|MKRJo_PIZTp^Kp2dA0b*~w|yBonn6i&TGnnLR|+jLF;MhqD1b{BR#cj5U`2-UXNV z(KWhE6M8iVslcsn9?Opt>rtwEYu~BEVrSpxr|+jl`jS=$&?YzW%55LscOYZZd6)x! z;%Tw`?$RZw^)j{bH{FD_c=ZmrMWr5_K(6MdjiyY=i?)+8>v$V|3s(l!qYIl0w zZo~a_ZNMVW^oh$nvBgfFCzYTRyBDyagt3~Qtlq`WPJ?IlgXSxvA4@GkXa1#dIA^R6 zhaFRPg+m6LtzFP%*lQh?8&z&PKQz_?X0rW}n(4RmZp~&iNvxqEYxr?HN;eE{TRWX^an_LmA-Ccz#4T)|tWYLn9aOW^v6bLD9$pgD9N()JDuSp%{F%&;=~SMsYqfN zAVIOaDoEZpQnQ=EKFs=_55h3mv!2{(u*YS4z4|7Gv;EWDa1P97JWU4GpHS49XuLtC zT)0NauRctUT{P2%DpFY)pd>UM(Y4VED4`U4jvKU?HNjyJ$gB$45_UbUibc(g>Usf% zu8qzku1W<){vRF`?QK$73V;p0UGusxc*9(3Jegj2tf6X3{?BC`fsx`G|2@k(& zRq1X1uh2>F>-cT-Ha@S}@l8=oeFYvMPvm1#?W3WazQk=t> z=8i+1mgn;5Tb}*kc|OQuC-I^lS68^=!(bR#h|>x^3fL-BaViif7_lh2U~&~Eri92f zi&pb0TRBY$liY;}3ZF&J;Cc?@f*@4z3{)O71cV^XZuQHNdYpc6O4tIdZ8bRpS}Zfi zVCgt#2{voJf~qaL+zkKT|FXas>n~5QcL(rT^NGT^p4VZGNE3I7dB|0k>`VO2(??-B zo;afei^L8QFV9UrnKj6H(@p*%>shlDO14NyqhYDw5)u|?v4~#PWqVbpUU4CXW+!3B z1!ATGW@ebDO^+&l=gmH*01cVmfw6FHT0x@J50TX94MxMDAmLnOY&YWOm-yb_h7yk& z$a<}2r)QzLSG@x8{f$Ug?`${jRfOi~Cu(f4QAw$lW#(cI@lS zZsQgXOqAtRd!UH{8YL)cC#TvXXWj=-1x!b$7n)X z9=TM-)V$>nu%TE`*qEBduM{3gu$QyJdu&{X=8fZF(SvT{z9!l%U@~jVM@vC)-FSE4 zbwr3-26?Y1t|C~trLF#0?!AnUWOu0R!&h)!b*0=Iz?SIFWm)Loboolml#`=bK6j3n z&D*f?Y*#|=k3+?uvlA@|$MLb`l567S-m`RemK6&jqk|H55RrQ$q9zpynwQo z!a`5zaN`r9QSH&U-KTsv9Ac={cwwURw0p9~T2K&=QC3 z6S<1+06ImqE?v@myG8=`oeUTIJElCGJn51jJWw=?Z*WkWMnP?$uRw=_?y(KxJQlcf zb520*J4$GFg^ZRDmb|9SP24=W|MpDpYews$hwDtrun9o#HFxQ-2?lBuV+&Ufe;Uj* zI=wC!3oSN|CQrjcz*q6faEqMX<&0zT`UtFzN834tYKBS4xSxqf2Sneb!=5l^PuvcHy&4$C@BwJR&q{)7D7(WnSAqN#IG9CT~9j30Fjie|3M-yX~63pp10IB{St^$4PggRU>G+<`nJpv6sn}<5+3Fwilp(sVB@=O+o{~ z4!E&_Pt>3d1U;UZx5Pn zO>ACN#;G!I;t<^mC8Spa6%p>vq7J_AzZ~sBKy{m%RIOhdB%Ip*`diiJ%){GURaV*R zzOL%}FRiPByKrJ3u`#EP1Sq(>QvEs1yD`R7rJ{MkJ|D06RT=BwG6BxK69x4l@U-&e z)`dT*KT^dw)|dI7@t>{Tb2s1LMmcBRL~R7#7wov{XT>kc__ZtxR>St=FP|K=!ho9~ ze8bB=sBx}4fsw*DfX`66E2eqeK70Xli#Ked97 zPA1*aOPn_OZZ^^5N0K(X!E3%*lRqgOXYxknTfEqP$Z!DI1OQH3)j)>VtblvyFt5A= z2*jazSo&cSvib_wFRrZYPjdQwdhFWm1~qUJu%lHNA-ywQ3FI&LJ@jKoegAuFOt%J} zeFnT{@_Z)dtjQM7Z4bV|r#`?Vb>HoTL;hAR9I~dKfNtkcPOr5&Z=(CGis7p@^&F`s z=QVQT$;&3jsOV9kSAW}8>wq|J+wDDdL2uKs3_h-Fw{A8vFr12)5J67Ri8#bULkbZZ zntN%#Y9U1vWBJ6w<~epCxD?tL3uq&NCE_T@SwRdNcwUxBs=4X+Sk{X*{&_zUTY|Vo zF;U4akjN#`{B>4EEJ6GNMY&n)9!HZq(B$Q~ez!uG=NCs4(Uly2Z1?(2^?BA|p!#S}*Pak8yi*w;JWaf8qk>{vt4int+jMD#jo3m=?~4Q{}7C&I#U`sC`q=FE|Pkl*a?3jLKVjs#xrC_AUES!%0u zJn;^C=WuM!cSs^Cd}=tvSmL?rvW1-I6|CkkXDhww7edj|bJX}7&d?McOLcG(L?`n5dL9kADU!}A)#-V7 z9Q@$fHjuiN`kg2-sVaGgP!=&3;KW8kvQ8|SfX|A8{S~7ME)al4Gy~;YWY{v!hrA|a z+j663fsCoj@cMBq3x>^WN^1wAT~e{c8#Vq1r~f){T&h1^Ks}=Br%Jzt<2Qc{y=Y0K z-5!wuI9COj#FLXF@dGCh%QA$qdEXSoy+8Rv4XU9QHdx+aGprICN7QTJnLD#-45T0g z)UkAGKf>G+1x}pf;l~LuIR15=1#Yk0xUiwFptA8p&XGB7h?rnaN5ulS9r#^qH8>BX7!a!W6>bk4AZ z#fvO?XG?Z4ox|4ct>H{ts2vc-NDuVr>L$JOnvY^HrhmkLjt)VB%OENl_|^JP=5rMJ z8BnmhJOX;8z5{oCO32Uh_lYghvKONBwn(^GhhS9kqSs{Spl38(nO~Hrc=1O8wtMAA z^OuOh@cIqX<;2=26&t3!lm59#D`NRv8qN%-{)aXG-FFD7bj_#$N$mM6A&KOjlI7O= zqbpSb(`0OB*%qd&zGkZr>rU$!aMXCgO>imiLOtytO*!Se@0ns&ixaGiW*v~YbBtTG zVx?5zW{GP!|jt!>2nIYPyZur2j!Ee z(4SN2kDVivg{TXkdK~PCZD+iL#itYhGS%mY z3gPrYLEuZ-xMoI`K_30gG6%LUA$Xb969XHb0)fxy=jL5f+)F_{6PPUjI3}*J+Vzi#F!v z*Iyb<_DA#bCTHmRiWr}$kI*>m!nK&hspuFmaIEf%sXu^5bU#vDH#X^7PwqoU?aT@6NMFdFViXx<1maZ5#)Wp4tn&207KqjisU{9pfeefk<3 z@~ewQ#IW6(T&c@u0+pJMJ_@4(HuWl#9Kz+(sw)`8;%LQUuX&2Wjb^2_zBv6WzR7Yx zRy@W{G#$j>C@85rUwSwq$d$vpSho;sui{cR`Y10{Zr8KyO{5ItweD#PZIP|7^Qh}% zZeAERbCcB*r1yu%J!2KG)*TQ{ z?6PE(%(ir4JQI64JaIg+U*deb!zjdmjo;*7v$kUqsjk$Clt9(6zhE z=$jKV2-+^F5QI+{)^G(u-^K=L6Pn%_#;cM&T$ma87o5=Wq{FLR9k39S%PL{VzR;KG2!evP9CEa0>q!$Pb)ZQ7m;zzAlb!Tr}R4k(+PVe60nBHYVmd-NsmOuU(uN zs7i)Y6Zlyb1z#gp!S*g;1Hc||S0!}_77BA~$yY6d7gY_ra9}+$SxB2~CuR5#v;qzV z6%tE^nP^)c(!G;3BhZ@F07?!d-ijQQb01amx59g>Wqv{ZOi{GDy+Y~eYk_KN#7j5+ zE-z#MH4zFsRwzO>oz8RmWlFksh2y%JYoI2a>ywH$Mg|tueTRE`=t`h^s2i*0bKL|C zw5}UAX--aJO*H~$@IJvqpslO%Um?4y(9&`~g@3Sd2VVtf{%*?72>PdZmH)E7f7ZE% zoqDH*olZYLz)s($?|=Qg?)Cl8VOjxs|G%y8A4_+HVX~TiVVE<=I1H0b6^6O#cMiii zeSg0vb1{telj2RH&fX*WyO254m1!J{R|wFkFi$7SHVpnYb()+k#Han(__UbuX--2N z!mH>!bcc8n{UgpeauC`#A$McesKA)DD}Xl?{b-LH00$cs?HyPU+<_L6*`RS@Gz;;7 z6Lj2a8r}4%*YpZti^TkwfL`P`Khq5Gb1U5|niEFDXY2i;>^5+oYN9uAY)^Ja}=IEU7s`g@4= ze^3f!(yf8I$BrMXy6$Myp^vc7e$7h4{I#7 zZu%wu1HXOiJ_(uP3%fAIqnQruMq-LDbWHIAQlPh)d1WIkXRf-P*Waq_QQV)QL*u`WkU@jaL-I$Icc{h!a3@G<8K$#|fnZpCeTfH%8k& z4WNkWDyH`twR`pF9itKO;;Zs%^a*{hYo=ocJ$6{~+sR16Ai$f+K2;N@MPzW(QnH zg`_${QashO1WjN%$#VujR03|b5?@|n-6X#3sfwfgXO#SQ%7%4(&HSaWBcWr!n1~|` zS;Qy8aMW*z_kb5>H1R*-FCT*~`^Ws{OiFdnUrGnoNq;$;=|NuHN`EOG)i(5(M@Vzm zS%3KeNrr$98z)lvFX%6&yAJf1?W(_&aU~-CrEGi?(KEf_9RH=bd|TTvtGE=Elj5>m$vKM4(pXo;CG=uj zdP}v2wr8TLCp&8w8?-Q0nEr2u)BpYbU`M}l^?xI_s9t&sMP2~CRG;zZ^$S9;G>u2wXP&j-PZ95Dmio%* z(zfJi_|iWD zYe>H#5V@hyO}%DNe<6q%7LI9Bz`~-kf|CCn)jgEB{%;c>2k?=xf)SVyxSzn;h_4EO z^S3mhz6?iih888F{>!5D!QsLTxQRGBlq^d=U+A{q zb^OXz`$7Ezel7NXo4HZaNIS3kiyn2GHRv_1X3#JNu2bg}#t%FP*c>%bg)rj6P0iJ; zbf(x3v7HP`8%7HLp0L`ncx(ASi{*a-L@gqyI7_8~i~pH}rUO2op|$4&|1e!x>6~qCx;Pb&z}6P!7vEY1_v>aEs!c-dz){>j#=o62x% zGgB%T7%Afc1i#kg?{GV^{Dht=1IuH7VoEd={Sk?{-1|ijks*hAUW%4|Nv!2^@7jIU zC!+6xg!T^PMG(&LlI*6(DFHO}NscLxdjKed2c!3=rE|BVez^wKdi}nq0MQLag94k7INsALLvgvX`sbV?9%tHX@dDHJ81#PWGa(<5I z>U=1Wytxq*4ji3Vo%n8S$ymdi`1ya}VjSgonu1?ep zw2vLq98Fg``3H6J*Z80K2VI6w2*X^(e8dy$@Kl%)O%{Tlrb?!Q>dPN!+)7xscCX*% zKF{pU1@-U5%RY~#F4&u`>^y>uV!tM^lv~gsVX>7|X#!8^8dOvoxCRvqQgPsO^o?5= z1x+fRmQCX+?FiQG0ptz(kUwQ;c9mEgoPy@d8%N6v--dVNcEh3YiACv4K|*eoALOZX zJXn=%dsPu!$FbT1Y4N)m9;NVo^Tis?~{A&?)~cRp1?%1 zW-{>?8%X6mCU>Z$9C-r!Pa}=3G2)4rUITQ5izLr7tMvAX69m&losl|HQD=tWNlhr< zci|ztISz+W#eb*@g{PinzxP{$xR)7vPo%LqwK(oRnlhiEDm*zrklbMc(!H5NYOaI3 z2~v&&(#752L`<3mC4S@Xx?;S;zk=ULiwoX}s%*0BY-+@IbUf4Na6mPFp6fe%ngz+} z`-M4ukEigepo5t)Wcs`cXb=0mbvKjpvtR3{2U^eKDHmVUe~Wk0p}79<2k;0li@mMG zzrkM49~M38{7Bu!vbE~#E0mhLsu&ZPf3cn$M-7Xp_^(wBDp5aS^>6a0Spr5a|1x`W z%bS6EulYvEC7ijSAgf6W{co_A;Z+$cTciz8l-U~j)rmE+d0+R6Rp5qpy((ZgBQ`AZ znkrcij&9-3h&$RtqB{_v=4va0otVBq-TzbWDsg@PE6(A{+6yO zPbtNED32wpalpK4T@1`#8msuGj{BBaa=Y>Toc%y4snHQkjODS$g%M-k_OmJ$soA1f zqQ5$`&b)eci#IRr&_!Jb)i|AkvkScDH~AF>KAwEEpw)RTJIqG07~wV1TIoF!@401C zBHm+5T8<(NvPlEr-mHbh*ZvvbW~ZFFj(cNQ;^yDWXhzZZO-vF*j7#+aMKds_S467R zKXsSv93RhfW__Ax2`@6g;Puu>+o{34G7r(zCHak8=(HvTyN?H>a>= z2a@T$PS7ScdXH3<)Ay>wwrCd950jYfKUD z99gk`%G&e+t}9jGt9czu@!{Wz@9OxDg_*vXY8zMTyBGKl8`4m))bPV9=1uu5eXlji zcd;j0Sq}cLNz2hBS(w+J>uYDiLK8t%)5flkuKkvxr}(&m~nDTo^R7 z#p=1CXz5j;ZnnCRn*W)AURcWiP)}XUyG!gl28{ilBL|%;LmzO3Dhs)m{471>)}w{# zmECsX&mif$e*+#uwC}3bhCN4EH$f^xcmTQi2U&uU z%D&x>p!sOkF4u?&vu^7`c#V|5H()+^y%?F**X3b&;Q2s>GVDSKR};KFUz4%$Jyo;eX7#aw&>H{`)z|h zz4NpG7Jce{s_4_j>It-B+w{r%X-JbkE%3jm=VUH!Zlw*@FN9vK`EW|qTijV3KP3zL0tp2;I)hkyRk*$3dtX$EqjKmpwYxK>HSO6 zoY^;h8=cB)$u~y@=O5JE`g8P>xeeA&#VkC9JUx zr;l-vI8D8Qt=xXN4+OIg@k9fHCwOkE|bcXEJuVm23 zhRuy9hqGBZLlYOM1AE09rBr?<%f!q6=Cxc*me;h!Mo-%pN}-t=GSCUn905l~g4`u2 z-E3vJZWXsmoV*VQBshPn0s|G)LqQZeI>zJeHDV&2tMM=D+L^!?dkv1r(CdrOHy&WwnYU z(cv%Z&8oBhMSha&qC1zX8^!wX?bbi1L!U5XbBvH2aEXAjQ*WO~$xhV*j@kO}R)dg- zZ_M?r4DIFU-4{!2eWvncRGRYw%efuZ6Nbq_?M)@aNC-oBHMVmx=yfE-5?^2`mxgmC z6VcetjhlCo+}Pd^5CVNRwtHUFS3D5AqSm>|y+*+>SkGQw(;TQmV|!n(`SLzMTDB5ko-`V4-~1aIp0xex6AMTGbS@tN z1T$^Sz(N1Od`BIJlwrQ@PC{Y3^;ZT{s8J8=4)ntx3_%&Og#Hrg#R%k}K;-)Ehdf(ss zOLDlaOdqqR-_c~b<-HUq3|ml|^Z%>h(~k7{;M5u2^fzIGHRCI%pTXD7a(xK9Y<+cg zp?zIv%c@~~3`(f+_fY?_fJvbyGe^^LHiDT}Fr0*N-1#s0x1hBTLcd%?@72BDbPpL2 ze74%Z??f?XAN!wgN0;Y<^E<6e?mWL_awq+siYrdg@BOXc0Z<#vb_Qx7m?IgSBHAjf z{)1|2`_U5%Px@)PlIH-3yuaBb9?z{^AEt|G=y;|;o_mDKpu!Tjexe>FC-Db~3r-nFO!Xa&UtUb9u`HR}=kR9qd=E2qeN9`&&wSWZZS{wZneR=HYue@9bG`bP7&Lez=)5|CdG3?yvUGK- z^6OY*dzJT#L-Y}bqcWe{A$wwCJ;M9N9;Bg@Eg&7!D`~^LU-Tj^p5U_em90vim+n!W z@7q4c*mNZSreM^XACY2`m-ANg-%r;~s`Rej%+n8F=D%^P_;2I}{`>L#NmYsQt&^&} zMQ`-b<3f8Z(Bo2j?4`%&^!WxoEmz{}R`gYSw4!g?qZNHuML$u|l}dcgihgL1R@Apg zE4oHSSE=YaCBAA!zpzIu+HQ|l^lKG;yQ*b8w%XfDT3IX3P|Pe@sE3yE+$wsTxo247 zT$lKa5@&@a?{|qWsGgVk{cK)Wo5zg3PT=(uy+-Uc&g*CTYCaDe#N|LYsBH+Z8_ld|ucf?hHXg`c z_u`e?tLJ;o(!)j@KUfz{6{P5r0IL_4K5tI62lOYeWn7PFDyj+V5vzC(I0xI2*NePJ zLCZMqsE-zsIkLcV6E1fLclxdSctbB0BqzoXo} zTE=~c++yYKWx1l~k$rbC)rk`7CvXTfntoCaq=Udm^^;-LMm+qQ#MBE)U_DZp zGmP8r5Me^ahrNJowT-Vs56dw1$@L#E!1>aPU^4Jf-mMpYiIlV8WB`u7Ekf(0EN zWb%7>bJ!1!tp%V!?TN_OEzvD8++6;azFzxCvdsRI9G@tUdq+I>@|uUj9XJZ(|7J%~ zS*Xi!xEV*Y9uvWUyYTPvNrL0bOIPR?AQ{y!&eDUxh=8!tOnE0K{s{%6n5&VAy@uhO z5X;B4DpDXa(Y%Ryv3pR1hW82^UciBq&a`Tusm5L{`#VznQ;;nEUDNjakRx=2II-4y zs>tt^>Ecb4-kJ?;+Kb8U7;tDmxc<;cKL|>Nk$yh$@?!H=_edirZ2e{ZGd2F0vn396 z;VkR_T;m_yZN|+k_s^mdLf`K$-)pW@L#d&wjt`YlA1uZ2-n9X*S;qk>_r399xkptF z;bZ@93J-FH&!2m)1dxYj17%y^e9pyioguBuouY5RAELK0$(%j6e!V`e z|CAFcec{S>Rt;>Y%g1Z1oi3t`&{Xy!%BbGd-SQg$oZanCCXETVCA)c&FB2d0!UdH4 zk4IuOA(dEar)ei*wj4iE;ylv2b6Tdv{Ur^muOC`1T|?=@bMW9oJ`xXgvI@6anF&Q? zxSJ?>m}zss6_j8jcLn8AD>b>?T|pTk$PX@KEe)1BLlep^9Gi*eWz#Mh_|1Pj_bR6A zC>N+`#6KUQ3P}SQGKlkxvjivPaxSFAfUZcj1O1|3swf}fD6rLyl-U?3=IlIMol`u# zrP~QN1)kN(+e#XgQSn~g_&}~mUQrxPVMwl6qHrZVm`3B*_SI43;1ICHYx+oE>F8eL zf`}_os}i_BKJ~5PP{vt3xDZB@r{NDzyN87BhLzXV1r^f|DN}X}Dkjc~J|V~o15y~3 zAi~O~9vYB*^01Iwgq4Xl&=r=xRnX18Of2HIc9j#i-TBJE5tR5|Y{j;jbSjV!G{B1Y zyym~LaV@T>>u-jLYz8xT)-c$f@JIEO40i>JT*ZZ(n-4X?XCFoG$AqzW))koaCx291G|*Z7e$9FZ5S=O}}gk!i}HpB8GK zJ1K*1N2<;!gCd-5Zn#Q%1J#$g1J86p_{4CT^~Zq{h!C-vdwd%|$oDrA{c$dx;UgzR z`W
1. **Start from scratch**: + - Create a _new_, _isolated_ directory for debugging + + ```bash + now=$(date +'%Y-%m-%d %H:%M:%S') + debug_dir="debugging/${now}" + mkdir -p "${debug_dir}" && cd "${debug_dir}" + echo "Debugging in $(pwd)" + - Create a new clone of the repository - - Create a new virtual environment - - Re-install dependencies - - Try re-running - - ```bash - now=$(date +'%Y-%m-%d %H:%M:%S') - debug_dir="debugging/${now}" - mkdir -p "${debug_dir}" && cd "${debug_dir}" - echo "Debugging in $(pwd)" - ``` - - -```bash -now=$(date +'%Y-%m-%d %H:%M:%S') -debug_dir="debugging/${now}" -mkdir -p "${debug_dir}" && cd "${debug_dir}" -echo "Debugging in $(pwd)" - -git clone https://github.com/argonne-lcf/Megatron-DeepSpeed.git && cd Megatron-DeepSpeed + ```bash + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed + ``` + + - Create a new virtual environment -source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env -python3 -m pip install "git+https://github.com/saforem2/ezpz" --require-virtualenv + ```bash + source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env + ``` + + - Re-install dependencies -# run simple test to verify distributed setup -ezpz-test + ```bash + python3 -m pip install -e "git+https://github.com/saforem2/ezpz" + ``` -# Run simple example -DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt bash train_alcf.sh -``` + - Run simple test to verify python can launch distributed processes: + + ```bash + ezpz-test + ``` + + - Try re-running + + ```bash + DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt bash train_alcf.sh + ``` [^alcf]: Yes, _any_ of the ALCF systems! e.g.: Aurora, Polaris, ThetaGPU, Sunspot, Sophia, Sirius, ... From 386863b6ff7945a2aefc7502da9d884d2eafc802 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:34:53 -0500 Subject: [PATCH 451/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index ac879e20889..f8ccf32e38a 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -5,8 +5,9 @@ HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH -source <(curl -L https://bit.ly/ezpz-utils) || exit -NO_COLOR=1 ezpz_setup_env || exit + +# shellcheck disable=SC1090 +source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env if command -v "ezpz-test"; then log_message INFO "${GREEN}✓${RESET} ezpz is already installed." From 82b491f875621d455545203fa03f50d97ef950e4 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:35:14 -0500 Subject: [PATCH 452/590] chore: Update `megatron/training_log.py` --- megatron/training_log.py | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index b798ba9b39a..394f34a048c 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -7,8 +7,9 @@ import logging import os +import ezpz + from deepspeed import get_accelerator -import ezpz as ez import torch from megatron.core import mpu @@ -27,9 +28,9 @@ ) -RANK: int = ez.get_rank() -WORLD_SIZE: int = ez.get_world_size() -DEVICE_TYPE: str = ez.dist.get_torch_device_type() +RANK: int = ezpz.get_rank() +WORLD_SIZE: int = ezpz.get_world_size() +DEVICE_TYPE: str = ezpz.dist.get_torch_device_type() DEVICE: torch.device = torch.device(DEVICE_TYPE) log: logging.Logger = logging.getLogger(__name__) From 9ffebe1f3c3741773e501a73239abeec52313344 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:35:24 -0500 Subject: [PATCH 453/590] chore: Update `megatron/training.py` --- megatron/training.py | 110 +++++++++++++++++++++++-------------------- 1 file changed, 58 insertions(+), 52 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index cf64eaf80a2..c239ea5f183 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -564,7 +564,7 @@ def get_optimizer_param_scheduler(optimizer): if args.train_iters: if args.lr_decay_iters is None: args.lr_decay_iters = args.train_iters - + lr_decay_steps = args.lr_decay_iters * args.global_batch_size wd_incr_steps = args.train_iters * args.global_batch_size if args.lr_warmup_fraction is not None: @@ -703,7 +703,7 @@ def setup_model_and_optimizer( optimizer = get_megatron_optimizer( model, no_wd_decay_cond, scale_lr_cond, lr_mult - ) + ) # opt_param_scheduler is the old lr_scheduler plus weight decay scheduling opt_param_scheduler = get_optimizer_param_scheduler(optimizer) if args.deepspeed: @@ -1018,7 +1018,6 @@ def train( write_args_to_tensorboard() assert accelerator is not None setup_profiler(args, accelerator.device_name()) - if args.random_ltd: # random-ltd requires different randomness on each rank @@ -1061,44 +1060,46 @@ def train( args.train_range_to_skip[1::2], ) ) - + # Learning rate finder mode - if hasattr(args, 'lr_finder') and args.lr_finder: + if hasattr(args, "lr_finder") and args.lr_finder: # Calculate number of iterations to use (10% of train_iters) finder_iters = max(1, int(args.train_iters * 0.1)) - + # Initialize tracking variables for LR finder lr_finder_losses = [] lr_finder_lrs = [] - + # Initialize loss smoothing variables avg_loss = 0.0 - best_loss = float('inf') + best_loss = float("inf") batch_num = 0 beta = 0.98 # Smoothing factor - + # Set initial learning rate and calculate multiplier init_lr = 1e-6 max_lr = 1.0 mult = (max_lr / init_lr) ** (1 / finder_iters) - + # Set initial learning rate curr_lr = init_lr for param_group in optimizer.param_groups: - param_group['lr'] = curr_lr - + param_group["lr"] = curr_lr + # Turn on training mode which enables dropout for model_module in model: model_module.train() - + # Get configuration for training config = core_transformer_config_from_args(args) if not args.deepspeed: config.grad_scale_func = optimizer.scale_loss config.timers = timers - - log.info(f"Running learning rate finder for {finder_iters} iterations (10% of {args.train_iters})") - + + log.info( + f"Running learning rate finder for {finder_iters} iterations (10% of {args.train_iters})" + ) + # Main LR finder loop for i in range(finder_iters): # Execute training step @@ -1110,89 +1111,94 @@ def train( opt_param_scheduler, config, ) - + # Skip iterations that didn't update weights if skipped_iter: continue - + # Get loss value (use first available loss if multiple present) - if 'lm loss' in loss_dict: - loss_val = loss_dict['lm loss'] + if "lm loss" in loss_dict: + loss_val = loss_dict["lm loss"] else: loss_val = next(iter(loss_dict.values())) - + if isinstance(loss_val, torch.Tensor): loss_val = loss_val.item() - + # Update batch counter batch_num += 1 - + # Compute smoothed loss avg_loss = beta * avg_loss + (1 - beta) * loss_val smoothed_loss = avg_loss / (1 - beta**batch_num) - + # Stop if the loss is exploding if batch_num > 1 and smoothed_loss > 4 * best_loss: log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") break - + # Record the best loss if smoothed_loss < best_loss or batch_num == 1: best_loss = smoothed_loss - + # Record values for plotting lr_finder_losses.append(smoothed_loss) lr_finder_lrs.append(curr_lr) - - # Log to wandb -# if wandb is not None and wandb.run is not None: - # wandb.log({ - # "lr_finder/learning_rate": curr_lr, - # "lr_finder/raw_loss": loss_val, - # "lr_finder/smoothed_loss": smoothed_loss, - # "lr_finder/iteration": i, - # "lr_finder/log_lr": math.log10(curr_lr) - # }) + # Log to wandb + # if wandb is not None and wandb.run is not None: + # wandb.log({ + # "lr_finder/learning_rate": curr_lr, + # "lr_finder/raw_loss": loss_val, + # "lr_finder/smoothed_loss": smoothed_loss, + # "lr_finder/iteration": i, + # "lr_finder/log_lr": math.log10(curr_lr) + # }) # Print progress - if (i+1) % args.log_interval == 0: - log.info(f"LR Finder: iteration {i+1}/{finder_iters}, " - f"lr: {curr_lr:.8f}, loss: {smoothed_loss:.4f}") - + if (i + 1) % args.log_interval == 0: + log.info( + f"LR Finder: iteration {i+1}/{finder_iters}, " + f"lr: {curr_lr:.8f}, loss: {smoothed_loss:.4f}" + ) + # Update the learning rate for the next step (bypassing scheduler) curr_lr *= mult for param_group in optimizer.param_groups: - param_group['lr'] = curr_lr - + param_group["lr"] = curr_lr + # Save raw data (on rank 0 only) if mpu.get_data_parallel_rank() == 0: # Create the results directory if it doesn't exist - os.makedirs(f'{args.save}/lr_finder', exist_ok=True) - + os.makedirs(f"{args.save}/lr_finder", exist_ok=True) + # Save raw data to a simple CSV file import csv - with open(f'{args.save}/lr_finder/lr_finder_data.csv', 'w', newline='') as f: + + with open( + f"{args.save}/lr_finder/lr_finder_data.csv", "w", newline="" + ) as f: writer = csv.writer(f) - writer.writerow(['learning_rate', 'loss']) + writer.writerow(["learning_rate", "loss"]) for lr, loss in zip(lr_finder_lrs, lr_finder_losses): writer.writerow([lr, loss]) - + # Also save as numpy arrays for convenience try: import numpy as np + np.savez( - f'{args.save}/lr_finder/lr_finder_data.npz', + f"{args.save}/lr_finder/lr_finder_data.npz", learning_rates=np.array(lr_finder_lrs), - losses=np.array(lr_finder_losses) + losses=np.array(lr_finder_losses), ) except ImportError: pass - + log.info(f"LR finder completed. Results saved to {args.save}/lr_finder/") - + # Return after LR finder is done - return args.iteration + return args.iteration while iteration < args.train_iters and ( args.train_tokens is None or args.consumed_train_tokens < args.train_tokens From b4849ae836113aad708fb9806b74e23279b4e874 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:00 -0500 Subject: [PATCH 454/590] chore: Update `megatron/optimizer_param_scheduler.py` --- megatron/optimizer_param_scheduler.py | 86 ++++++++++++++++++--------- 1 file changed, 58 insertions(+), 28 deletions(-) diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 647e88b1457..00614483aad 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -10,15 +10,25 @@ class OptimizerParamScheduler(object): """Anneals learning rate and weight decay""" - def __init__(self, optimizer, max_lr, min_lr, - lr_warmup_steps, lr_decay_steps, lr_decay_style, - start_wd, end_wd, wd_incr_steps, wd_incr_style, - constant_lr = 0.0, - lr_constant_steps=None, - lr_cooldown_steps=None, - timescale=None, - use_checkpoint_opt_param_scheduler=True, - override_opt_param_scheduler=False): + def __init__( + self, + optimizer, + max_lr, + min_lr, + lr_warmup_steps, + lr_decay_steps, + lr_decay_style, + start_wd, + end_wd, + wd_incr_steps, + wd_incr_style, + constant_lr=0.0, + lr_constant_steps=None, + lr_cooldown_steps=None, + timescale=None, + use_checkpoint_opt_param_scheduler=True, + override_opt_param_scheduler=False, + ): args = get_args() # Class values. self.optimizer = optimizer @@ -37,7 +47,7 @@ def __init__(self, optimizer, max_lr, min_lr, self.end_steps = lr_decay_steps assert self.lr_decay_steps > 0 assert self.lr_warmup_steps < self.lr_decay_steps - assert self.lr_constant_steps < self.lr_decay_steps + assert self.lr_constant_steps < self.lr_decay_steps self.lr_decay_tokens = args.lr_decay_tokens self.num_tokens = 0 @@ -139,74 +149,94 @@ def get_lr(self): # --- Infinite Decay Options --- elif self.lr_decay_style in ["infinite-cosine", "infinite-inv-square-root"]: - # Compute iterations after warmup. (Using steps if lr_decay_tokens is None, - # otherwise tokens.) + # Compute iterations after warmup. (Using steps if lr_decay_tokens is None, + # otherwise tokens.) assert self.constant_lr > self.min_lr assert self.constant_lr <= self.max_lr - delta_lr = (self.max_lr - self.constant_lr) + delta_lr = self.max_lr - self.constant_lr if self.lr_decay_tokens is None: num_steps_ = self.num_steps - self.lr_warmup_steps cooldown_steps_ = self.lr_cooldown_steps - self.lr_warmup_steps if self.constant_steps is None: - raise Exception('Constant LR steps need to be provided for infinite schedulers') + raise Exception( + "Constant LR steps need to be provided for infinite schedulers" + ) if num_steps_ <= cooldown_steps_: cooldown_ratio = float(num_steps_) / float(cooldown_steps_) - if self.lr_decay_style == "infinite_cosine": coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) lr = self.constant_lr + delta_lr * coeff else: # infinite_inv_sqrt + def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 + coeff = inv_f(1) * inv_f(cooldown_ratio) lr = self.start_lr - delta_lr * coeff else: num_steps_ = num_steps_ - cooldown_steps_ if num_steps_ <= self.lr_constant_steps: - # Stay constant for constant_iters + # Stay constant for constant_iters lr = self.constant_lr else: - # Exponential decay from constant_lr to min_lr. - - end_steps_ = self.end_steps - self.warmup_steps - cooldown_steps_ - self.lr_constant_steps + # Exponential decay from constant_lr to min_lr. + + end_steps_ = ( + self.end_steps + - self.warmup_steps + - cooldown_steps_ + - self.lr_constant_steps + ) num_steps_ = num_steps_ - self.lr_constant_steps - exp_factor = -math.log(self.min_lr / self.constant_lr) / end_steps_ + exp_factor = ( + -math.log(self.min_lr / self.constant_lr) / end_steps_ + ) lr = self.constant_lr * math.exp(-exp_factor * num_steps_) - + # token based decay else: num_tokens_ = self.num_tokens_ - self.lr_warmup_tokens cooldown_tokens_ = self.lr_cooldown_tokens - self.lr_warmup_tokens if self.constant_tokens is None: - raise Exception('Constant LR tokens need to be provided for infinite schedulers') + raise Exception( + "Constant LR tokens need to be provided for infinite schedulers" + ) if num_tokens_ <= cooldown_tokens_: cooldown_ratio = float(num_tokens_) / float(cooldown_tokens_) - + if self.lr_decay_style == "infinite_cosine": coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) lr = self.constant_lr + delta_lr * coeff else: # infinite_inv_sqrt + def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 + coeff = inv_f(1) * inv_f(cooldown_ratio) lr = self.start_lr - delta_lr * coeff else: num_tokens_ = num_tokens_ - cooldown_tokens_ - + if num_tokens_ <= self.lr_constant_tokens: # Stay constant for constant_tokens period lr = self.constant_lr else: # Exponential decay from constant_lr to min_lr - - end_tokens_ = self.end_tokens - self.lr_warmup_tokens - cooldown_tokens_ - self.lr_constant_tokens + + end_tokens_ = ( + self.end_tokens + - self.lr_warmup_tokens + - cooldown_tokens_ + - self.lr_constant_tokens + ) num_tokens_ = num_tokens_ - constant_period - exp_factor = -math.log(self.min_lr / self.constant_lr) / end_tokens_ + exp_factor = ( + -math.log(self.min_lr / self.constant_lr) / end_tokens_ + ) lr = self.constant_lr * math.exp(-exp_factor * num_tokens_) - return max(lr, self.min_lr) if self.lr_decay_tokens is None: From 683386a5eb3da7e00fecd73a585a1e836746ef0c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:06 -0500 Subject: [PATCH 455/590] chore: Update `megatron/optimizer/muon.py` --- megatron/optimizer/muon.py | 114 ++++++++++++++++++------------------- 1 file changed, 56 insertions(+), 58 deletions(-) diff --git a/megatron/optimizer/muon.py b/megatron/optimizer/muon.py index c25865b6a66..c8e4bd7b4e4 100644 --- a/megatron/optimizer/muon.py +++ b/megatron/optimizer/muon.py @@ -4,9 +4,10 @@ import torch.distributed as dist from torch import Tensor + # This code snippet is a modified version adapted from the following GitHub repository: # https://github.com/KellerJordan/Muon/blob/master/muon.py -#and https://github.com/MoonshotAI/Moonlight/blob/master/examples/toy_train.py +# and https://github.com/MoonshotAI/Moonlight/blob/master/examples/toy_train.py @torch.compile def zeropower_via_newtonschulz5(G, steps): """ @@ -22,9 +23,7 @@ def zeropower_via_newtonschulz5(G, steps): # Perform the NS iterations for _ in range(steps): A = X @ X.T - B = ( - b * A + c * A @ A - ) + B = b * A + c * A @ A X = a * X + B @ X if G.size(0) > G.size(1): @@ -37,7 +36,6 @@ class Muon(torch.optim.Optimizer): Muon - MomentUm Orthogonalized by Newton-schulz """ - def __init__( self, params, @@ -61,45 +59,45 @@ def __init__( # Initialize the base optimizer with all parameter groups super(Muon, self).__init__(params, defaults) - + # Process parameter groups to determine which will use Muon and which will use AdamW for i, group in enumerate(self.param_groups): # Ensure each group has all required parameters with defaults for k, v in defaults.items(): group.setdefault(k, v) - + # Mark parameters as using Muon or AdamW - group['use_muon_list'] = [] - - for p_idx, p in enumerate(group['params']): + group["use_muon_list"] = [] + + for p_idx, p in enumerate(group["params"]): if p.grad is None: continue - + use_muon = p.ndim == 2 - + # If parameter is 2D but has a large dimension, it's likely an embedding or LM head, need to change this!!! if use_muon and max(p.shape) > 10000: use_muon = False - - group['use_muon_list'].append(use_muon) - + + group["use_muon_list"].append(use_muon) + # Initialize parameter state state = self.state[p] if len(state) == 0: - state['use_muon'] = use_muon + state["use_muon"] = use_muon if use_muon: - state['momentum_buffer'] = torch.zeros_like(p.grad) + state["momentum_buffer"] = torch.zeros_like(p.grad) else: - state['step'] = torch.tensor(0.0) - state['moment1'] = torch.zeros_like(p.grad) - state['moment2'] = torch.zeros_like(p.grad) + state["step"] = torch.tensor(0.0) + state["moment1"] = torch.zeros_like(p.grad) + state["moment2"] = torch.zeros_like(p.grad) def __setstate__(self, state): """ Handle state loading for the optimizer. """ super(Muon, self).__setstate__(state) - + # Ensure all parameter groups have the required defaults for group in self.param_groups: group.setdefault("nesterov", True) @@ -110,15 +108,15 @@ def __setstate__(self, state): group.setdefault("wd", 0.1) group.setdefault("lr", 1e-3) group.setdefault("use_muon_list", []) - + # Convert step from float to tensor if needed state_values = list(self.state.values()) - if state_values and 'step' in state_values[0]: - step_is_tensor = torch.is_tensor(state_values[0]['step']) + if state_values and "step" in state_values[0]: + step_is_tensor = torch.is_tensor(state_values[0]["step"]) if not step_is_tensor: for s in state_values: - if 'step' in s: - s['step'] = torch.tensor(float(s['step'])) + if "step" in s: + s["step"] = torch.tensor(float(s["step"])) def adjust_lr_for_muon(self, lr, param_shape): """ @@ -145,11 +143,11 @@ def step(self, closure=None): ns_steps = group["ns_steps"] adamw_betas = group["adamw_betas"] adamw_eps = group["adamw_eps"] - + # Get use_muon_list (initialize if not present) - if 'use_muon_list' not in group or not group['use_muon_list']: - group['use_muon_list'] = [] - for p in group['params']: + if "use_muon_list" not in group or not group["use_muon_list"]: + group["use_muon_list"] = [] + for p in group["params"]: if p.grad is None: continue # Default determination of whether to use Muon @@ -157,15 +155,15 @@ def step(self, closure=None): # Don't use Muon for embeddings/LM heads (approximated by large dimension check) if use_muon and max(p.shape) > 10000: use_muon = False - group['use_muon_list'].append(use_muon) - + group["use_muon_list"].append(use_muon) + # Apply optimization step to each parameter - for param_idx, p in enumerate(group['params']): + for param_idx, p in enumerate(group["params"]): if p.grad is None: continue - + grad = p.grad - + # Initialize state if needed state = self.state[p] if len(state) == 0: @@ -174,44 +172,44 @@ def step(self, closure=None): if p.ndim == 2: # Don't use Muon for embeddings/LM heads (approximated by large dimension check) use_muon = max(p.shape) <= 10000 - - state['use_muon'] = use_muon - + + state["use_muon"] = use_muon + if use_muon: - state['momentum_buffer'] = torch.zeros_like(grad) + state["momentum_buffer"] = torch.zeros_like(grad) else: - state['step'] = torch.tensor(0.0) - state['moment1'] = torch.zeros_like(grad) - state['moment2'] = torch.zeros_like(grad) - + state["step"] = torch.tensor(0.0) + state["moment1"] = torch.zeros_like(grad) + state["moment2"] = torch.zeros_like(grad) + # Check if we should use Muon for this parameter - use_muon = state.get('use_muon', False) - + use_muon = state.get("use_muon", False) + if use_muon: # Muon optimization if grad.ndim > 2: grad = grad.view(grad.size(0), -1) - + # Initialize momentum buffer if not present if "momentum_buffer" not in state: state["momentum_buffer"] = torch.zeros_like(grad) - + buf = state["momentum_buffer"] buf.mul_(momentum).add_(grad) - + if nesterov: g = grad.add(buf, alpha=momentum) else: g = buf - + u = zeropower_via_newtonschulz5(g, steps=ns_steps) - + # Scale update adjusted_lr = self.adjust_lr_for_muon(lr, p.shape) - + # Apply weight decay p.data.mul_(1 - lr * wd) - + # Apply update p.data.add_(u, alpha=-adjusted_lr) else: @@ -219,27 +217,27 @@ def step(self, closure=None): beta1, beta2 = adamw_betas eps = adamw_eps weight_decay = wd - + # Initialize AdamW state if not present if "step" not in state: state["step"] = torch.tensor(0.0) state["moment1"] = torch.zeros_like(grad) state["moment2"] = torch.zeros_like(grad) - + state["step"] += 1 step = state["step"] buf1 = state["moment1"] buf2 = state["moment2"] - + buf1.lerp_(grad, 1 - beta1) buf2.lerp_(grad.square(), 1 - beta2) - + g = buf1 / (eps + buf2.sqrt()) - + bias_correction1 = 1 - beta1**step bias_correction2 = 1 - beta2**step scale = bias_correction1 / bias_correction2**0.5 - + p.data.mul_(1 - lr * weight_decay) p.data.add_(g, alpha=-lr / scale) From 61655616d01c202699856bf299ce653d91984b1c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:09 -0500 Subject: [PATCH 456/590] chore: Update `megatron/optimizer/adopt.py` --- megatron/optimizer/adopt.py | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/megatron/optimizer/adopt.py b/megatron/optimizer/adopt.py index 4e5b3722dc0..e33fb83b1d6 100644 --- a/megatron/optimizer/adopt.py +++ b/megatron/optimizer/adopt.py @@ -29,8 +29,9 @@ __all__ = ["ADOPT", "adopt"] -#code from https://github.com/iShohei220/adopt/blob/main/src/adopt/adopt.py, copy pasted here because no pip and not sure about submodules +# code from https://github.com/iShohei220/adopt/blob/main/src/adopt/adopt.py, copy pasted here because no pip and not sure about submodules class ADOPT(Optimizer): + def __init__( self, params: ParamsT, @@ -132,9 +133,7 @@ def _init_group( has_complex |= torch.is_complex(p) params_with_grad.append(p) if p.grad.is_sparse: - raise RuntimeError( - "ADOPT does not support sparse gradients" - ) + raise RuntimeError("ADOPT does not support sparse gradients") grads.append(p.grad) state = self.state[p] @@ -305,7 +304,7 @@ def _single_tensor_adopt( continue if weight_decay != 0 and decouple: - param.add_(param, alpha=-lr*weight_decay) + param.add_(param, alpha=-lr * weight_decay) denom = torch.clamp(exp_avg_sq.sqrt(), eps) normed_grad = grad.div(denom) @@ -417,7 +416,7 @@ def _multi_tensor_adopt( continue if weight_decay != 0 and decouple: - torch._foreach_add_(device_params, device_params, alpha=-lr*weight_decay) + torch._foreach_add_(device_params, device_params, alpha=-lr * weight_decay) exp_avg_sq_sqrt = torch._foreach_sqrt(device_exp_avg_sqs) torch._foreach_maximum_(exp_avg_sq_sqrt, eps) @@ -498,9 +497,7 @@ def adopt( eps: float, maximize: bool, ): - r"""Functional API that performs ADOPT algorithm computation. - - """ + r"""Functional API that performs ADOPT algorithm computation.""" # Respect when the user inputs False/True for foreach or fused. We only want to change # the default when neither have been user-specified. Note that we default to foreach # and pass False to use_fused. This is not a mistake--we want to give the fused impl From 558322dc33083ddd71a0551ba1bcee946b88372b Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:18 -0500 Subject: [PATCH 457/590] chore: Update `megatron/optimizer/__init__.py` --- megatron/optimizer/__init__.py | 31 ++++++++++++++++--------------- 1 file changed, 16 insertions(+), 15 deletions(-) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index c961bc42ac7..47ca85f5145 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -383,7 +383,7 @@ def optimizer_hook(p): eps=args.adam_eps, ) # ---- Shampoo ---------------------------------------- - elif str(args.optimizer).lower() == 'dshampooadamw': + elif str(args.optimizer).lower() == "dshampooadamw": from distributed_shampoo.distributed_shampoo import DistributedShampoo from distributed_shampoo.shampoo_types import AdamGraftingConfig @@ -397,20 +397,20 @@ def optimizer_hook(p): precondition_frequency=args.dshampooadamw_precondition_freq, use_decoupled_weight_decay=args.dshampooadamw_use_decoupled_weight_decay, grafting_config=AdamGraftingConfig( - beta2=args.adam_beta2, - epsilon=args.adam_eps, + beta2=args.adam_beta2, + epsilon=args.adam_eps, ), ) - #--- sgd ------------------------------------------------- - elif args.optimizer == 'sgd': + # --- sgd ------------------------------------------------- + elif args.optimizer == "sgd": optimizer = torch.optim.SGD( param_groups, lr=args.lr, weight_decay=args.weight_decay, momentum=args.sgd_momentum, ) - #--- sophiag ----------------------------------------------- - elif str(args.optimizer).lower() == 'sophiag': + # --- sophiag ----------------------------------------------- + elif str(args.optimizer).lower() == "sophiag": from .sophia import SophiaG optimizer = SophiaG( @@ -420,18 +420,19 @@ def optimizer_hook(p): rho=args.sophiag_rho, weight_decay=args.weight_decay, ) - #--- adopt ----------------------------------------------- - elif str(args.optimizer).lower() == 'adopt': + # --- adopt ----------------------------------------------- + elif str(args.optimizer).lower() == "adopt": from .adopt import ADOPT + optimizer = ADOPT( param_groups, lr=args.lr, betas=(args.adam_beta1, args.adam_beta2), - eps = args.adam_eps, - weight_decay=args.weight_decay + eps=args.adam_eps, + weight_decay=args.weight_decay, ) - #--- muon ----------------------------------------------- - elif str(args.optimizer).lower() == 'muon': + # --- muon ----------------------------------------------- + elif str(args.optimizer).lower() == "muon": from .muon import Muon optimizer = Muon( @@ -441,8 +442,8 @@ def optimizer_hook(p): momentum=args.muon_momentum, nesterov=args.muon_nesterov, ns_steps=args.muon_ns_steps, - adamw_betas=(args.muonadamw_beta1,args.muonadamw_beta2), - adamw_eps=args.muonadamw_eps + adamw_betas=(args.muonadamw_beta1, args.muonadamw_beta2), + adamw_eps=args.muonadamw_eps, ) else: raise TypeError(f"{args.optimizer} optimizer is not supported.") From 1656e98bb92072054218eacd8d9afb129ff322da Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:26 -0500 Subject: [PATCH 458/590] chore: Update `megatron/core/transformer/transformer_config.py` --- .../core/transformer/transformer_config.py | 187 +++++++++--------- 1 file changed, 94 insertions(+), 93 deletions(-) diff --git a/megatron/core/transformer/transformer_config.py b/megatron/core/transformer/transformer_config.py index 359fa3a5a54..798c1b25264 100644 --- a/megatron/core/transformer/transformer_config.py +++ b/megatron/core/transformer/transformer_config.py @@ -14,95 +14,95 @@ class TransformerConfig(ModelParallelConfig): """Configuration object for megatron-core transformers. - Attributes: - - # model architecture - num_layers (int): Number of transformer layers in a transformer block. - hidden_size (int): Transformer hidden size. - ffn_hidden_size (int): Transformer Feed-Forward Network hidden size. - This is set to 4*hidden_size if not provided. Defaults to None.') - num_attention_heads (int): Number of transformer attention heads. - num_key_value_heads (int): This is the number of key_value heads that should be used to implement Grouped Query Attention. If - `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if - `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. - For more details checkout [this paper](https://arxiv.org/pdf/2305.13245.pdf). - If it is not specified, will default to `num_attention_heads`. - kv_channels (int): Projection weights dimension in multi-head attention. - This is set to hidden_size // num_attention_heads if not provided. - Defaults to None. - hidden_dropout (float): Dropout probability for transformer hidden state. Defaults to 0.1. - attention_dropout (float): Post attention dropout probability. Defaults to 0.1. - fp32_residual_connection (bool): If true, move residual connections to fp32. - apply_residual_connection_post_layernorm (bool): If true, uses the original BERT residule connection ordering. - Defaults to False. - layernorm_epsilon (float): Layernorm epsilon. Defaults to 1e-5. - - layernorm_zero_centered_gamma (bool): if set to 'True', the LayerNorm is adjusted to center the gamma values - around 0. This improves numerical stability. Defaults to False. - - add_bias_linear (bool): Include a bias term in all linear layers (QKV projections, after core attention, and two - in MLP layer). Default is True. - - gated_linear_unit (bool): Use a gated linear unit for the first linear layer in the MLP. Defaults to False. - - activation_func (Callable): Activation function to use for the non-linearity in the MLP. Defaults to F.gelu. - - # initialization - init_method (Callable): Method to initialize weights. Note that bias is always set to - zero. Should be a function that takes a single Tensor and - initializes it. Defaults to - megatron.core.utils.init_method_normal(init_method_std) which is - torch.nn.init.normal_ with mean=0.0 and std=init_method_Std. - - output_layer_init_method (Callable): Method to initialize weights of the output layer of - both attention and MLP blocks. Defaults to - megatron.core.utils.scaled_init_method_normal(init_method_std) - which is torch.nn.init.normal_ with mean=0.0 and - std=init_method_std / math.sqrt(2.0 * num_layers). - - - word_embedding_init_std (float): Standard deviation of the zero mean normal initialization for the word embeddings - - init_method_std (float): Standard deviation of the zero mean normal for the default - initialization method, not used if init_method and - output_layer_init_method are provided. Defaults to 0.02. - - # mixed-precision - apply_query_key_layer_scaling (bool): If true, scale Q * K^T by 1 / layer-number. Defaults to True. - attention_softmax_in_fp32 (bool): If true, run attention masking and softmax in fp32. - This should be true if apply_query_key_layer_scaling is true. - - # fusion - bias_gelu_fustion (bool): If true, fuses bias and gelu. Defaults to False. - masked_softmax_fusion (bool): If true, uses softmax fusion. - persist_layer_norm (bool): If true, uses the persistent fused layer norm kernel. - This kernel only supports a fixed set of hidden sizes. - Defaults to False. - bias_dropout_fusion (bool): If true, uses bias dropout fusion. - - # activation recomputation - - recompute_granularity (str): megatron-core supports 'selective' activation checkpointing where only the memory - intensive part of attention is checkpointed. These memory intensive activations - are also less compute intensive which makes activation checkpointing more efficient - for LLMs (20B+). See Reducing Activation Recomputation in Large Transformer - Models: https://arxiv.org/abs/2205.05198 for more details. 'full' will checkpoint - the entire transformer layer. Must be 'selective' or 'full'. Defaults to None. - - recompute_method (str): uniform will uniformly divide the total number of transformer layers in a transformer - block and recompute the input activation of each divided chunk at the specified - granularity. block will recompute the input activations for only a set number of - transformer layers per pipeline stage. The rest of the layers in the pipeline stage - will not have any activations recomputed. Must be 'uniform' or 'block'. Defaults to - None. - - recompute_num_layers (int): When recompute_method is uniform, recompute_num_layers is the number of transformer - layers in each uniformly divided recompute unit. When recompute_method is block, - recompute_num_layers is the number of transformer layers to recompute within each - pipeline stage. Defaults to None. - - distribute_saved_activations (bool): If true, distribute recomputed activations across the model parallel - group. Defaults to None. + Attributes: + + # model architecture + num_layers (int): Number of transformer layers in a transformer block. + hidden_size (int): Transformer hidden size. + ffn_hidden_size (int): Transformer Feed-Forward Network hidden size. + This is set to 4*hidden_size if not provided. Defaults to None.') + num_attention_heads (int): Number of transformer attention heads. + num_key_value_heads (int): This is the number of key_value heads that should be used to implement Grouped Query Attention. If + `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if + `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. + For more details checkout [this paper](https://arxiv.org/pdf/2305.13245.pdf). + If it is not specified, will default to `num_attention_heads`. + kv_channels (int): Projection weights dimension in multi-head attention. + This is set to hidden_size // num_attention_heads if not provided. + Defaults to None. + hidden_dropout (float): Dropout probability for transformer hidden state. Defaults to 0.1. + attention_dropout (float): Post attention dropout probability. Defaults to 0.1. + fp32_residual_connection (bool): If true, move residual connections to fp32. + apply_residual_connection_post_layernorm (bool): If true, uses the original BERT residule connection ordering. + Defaults to False. + layernorm_epsilon (float): Layernorm epsilon. Defaults to 1e-5. + + layernorm_zero_centered_gamma (bool): if set to 'True', the LayerNorm is adjusted to center the gamma values + around 0. This improves numerical stability. Defaults to False. + + add_bias_linear (bool): Include a bias term in all linear layers (QKV projections, after core attention, and two + in MLP layer). Default is True. + + gated_linear_unit (bool): Use a gated linear unit for the first linear layer in the MLP. Defaults to False. + + activation_func (Callable): Activation function to use for the non-linearity in the MLP. Defaults to F.gelu. + + # initialization + init_method (Callable): Method to initialize weights. Note that bias is always set to + zero. Should be a function that takes a single Tensor and + initializes it. Defaults to + megatron.core.utils.init_method_normal(init_method_std) which is + torch.nn.init.normal_ with mean=0.0 and std=init_method_Std. + + output_layer_init_method (Callable): Method to initialize weights of the output layer of + both attention and MLP blocks. Defaults to + megatron.core.utils.scaled_init_method_normal(init_method_std) + which is torch.nn.init.normal_ with mean=0.0 and + std=init_method_std / math.sqrt(2.0 * num_layers). + + + word_embedding_init_std (float): Standard deviation of the zero mean normal initialization for the word embeddings + + init_method_std (float): Standard deviation of the zero mean normal for the default + initialization method, not used if init_method and + output_layer_init_method are provided. Defaults to 0.02. + + # mixed-precision + apply_query_key_layer_scaling (bool): If true, scale Q * K^T by 1 / layer-number. Defaults to True. + attention_softmax_in_fp32 (bool): If true, run attention masking and softmax in fp32. + This should be true if apply_query_key_layer_scaling is true. + + # fusion + bias_gelu_fustion (bool): If true, fuses bias and gelu. Defaults to False. + masked_softmax_fusion (bool): If true, uses softmax fusion. + persist_layer_norm (bool): If true, uses the persistent fused layer norm kernel. + This kernel only supports a fixed set of hidden sizes. + Defaults to False. + bias_dropout_fusion (bool): If true, uses bias dropout fusion. + + # activation recomputation + + recompute_granularity (str): megatron-core supports 'selective' activation checkpointing where only the memory + intensive part of attention is checkpointed. These memory intensive activations + are also less compute intensive which makes activation checkpointing more efficient + for LLMs (20B+). See Reducing Activation Recomputation in Large Transformer + Models: https://arxiv.org/abs/2205.05198 for more details. 'full' will checkpoint + the entire transformer layer. Must be 'selective' or 'full'. Defaults to None. + + recompute_method (str): uniform will uniformly divide the total number of transformer layers in a transformer + block and recompute the input activation of each divided chunk at the specified + granularity. block will recompute the input activations for only a set number of + transformer layers per pipeline stage. The rest of the layers in the pipeline stage + will not have any activations recomputed. Must be 'uniform' or 'block'. Defaults to + None. + + recompute_num_layers (int): When recompute_method is uniform, recompute_num_layers is the number of transformer + layers in each uniformly divided recompute unit. When recompute_method is block, + recompute_num_layers is the number of transformer layers to recompute within each + pipeline stage. Defaults to None. + + distribute_saved_activations (bool): If true, distribute recomputed activations across the model parallel + group. Defaults to None. """ @@ -129,12 +129,11 @@ class TransformerConfig(ModelParallelConfig): init_method: Callable = None output_layer_init_method: Callable = None init_method_std: float = 0.02 - + adjust_word_embedding_init: bool = False world_embedding_init_method: Callable = None word_embedding_init_std: float = 0.02 - # mixed-precision apply_query_key_layer_scaling: bool = True attention_softmax_in_fp32: bool = True @@ -225,10 +224,12 @@ def __post_init__(self): if self.init_method is None: self.init_method = init_method_normal(self.init_method_std) - + if self.world_embedding_init_method is None: if self.adjust_word_embedding_init: - self.world_embedding_init_method = init_method_normal(self.word_embedding_init_std) + self.world_embedding_init_method = init_method_normal( + self.word_embedding_init_std + ) else: self.world_embedding_init_method = self.init_method From 102d7973f3ab28bca32b1afa3155862a56ced169 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:29 -0500 Subject: [PATCH 459/590] chore: Update `megatron/checkpointing.py` --- megatron/checkpointing.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/megatron/checkpointing.py b/megatron/checkpointing.py index 7202a7c924c..7a58d8b100a 100644 --- a/megatron/checkpointing.py +++ b/megatron/checkpointing.py @@ -356,9 +356,9 @@ def state_dict_for_save_checkpoint_deepspeed( # Trim off the filename and mp_rank_* directory. for _ in range(3): checkpoint_name = os.path.dirname(checkpoint_name) - #try: + # try: # model[0].save_checkpoint(checkpoint_name, client_state=state_dict) - #except Exception: + # except Exception: # import ezpz # ezpz.utils.breakpoint(0) model[0].save_checkpoint(checkpoint_name, client_state=state_dict) From f7f397bd079ba98d7e3dae35268de1317ffec7e3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:36:33 -0500 Subject: [PATCH 460/590] chore: Update `megatron/arguments.py` --- megatron/arguments.py | 104 +++++++++++++++++++++--------------------- 1 file changed, 52 insertions(+), 52 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 87e9f5678cd..709efaace9f 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -76,21 +76,21 @@ def validate_args(args, defaults={}): args.tensor_model_parallel_size = min( args.tensor_model_parallel_size, args.world_size ) - assert args.world_size % args.tensor_model_parallel_size == 0, ( - "world size ({}) is not divisible by tensor model parallel size ({})".format( - args.world_size, args.tensor_model_parallel_size - ) + assert ( + args.world_size % args.tensor_model_parallel_size == 0 + ), "world size ({}) is not divisible by tensor model parallel size ({})".format( + args.world_size, args.tensor_model_parallel_size ) # Zero bubble pipeline is defined on deepspeed's scheduler if args.enable_zbh1_pipeline: assert args.deepspeed, "Use DeepSpeed to use zero-bubble H1 pipeline" - assert args.sequence_parallel == False, ( - "Sequence Parallel not tested, proceed at own will by removing this line" - ) + assert ( + args.sequence_parallel == False + ), "Sequence Parallel not tested, proceed at own will by removing this line" if args.enable_zbh1_exact_semantics: - assert args.enable_zbh1_pipeline, ( - "Exact semantics require ZBH1 pipeline enabled" - ) + assert ( + args.enable_zbh1_pipeline + ), "Exact semantics require ZBH1 pipeline enabled" # Pipeline model parallel size. args.pipeline_model_parallel_size = min( args.pipeline_model_parallel_size, @@ -103,14 +103,14 @@ def validate_args(args, defaults={}): ) # Checks. if args.no_pipeline_parallel: - assert args.pipeline_model_parallel_size == 1, ( - "pipeline_model_parallel_size must be 1 if pipeline parallel is disabled" - ) + assert ( + args.pipeline_model_parallel_size == 1 + ), "pipeline_model_parallel_size must be 1 if pipeline parallel is disabled" if args.ds_sequence_parallel_size > 1: - assert version.parse(deepspeed.__version__) >= version.parse("0.10.2"), ( - "sequence parallelism requires DeepSpeed version 0.10.2+" - ) + assert version.parse(deepspeed.__version__) >= version.parse( + "0.10.2" + ), "sequence parallelism requires DeepSpeed version 0.10.2+" model_parallel_size = ( args.pipeline_model_parallel_size @@ -155,13 +155,13 @@ def validate_args(args, defaults={}): ) # Deprecated arguments - assert args.batch_size is None, ( - "--batch-size argument is no longer valid, use --micro-batch-size instead" - ) + assert ( + args.batch_size is None + ), "--batch-size argument is no longer valid, use --micro-batch-size instead" del args.batch_size - assert args.warmup is None, ( - "--warmup argument is no longer valid, use --lr-warmup-fraction instead" - ) + assert ( + args.warmup is None + ), "--warmup argument is no longer valid, use --lr-warmup-fraction instead" del args.warmup assert args.model_parallel_size is None, ( "--model-parallel-size is no " @@ -284,19 +284,19 @@ def validate_args(args, defaults={}): # If we use iteration-based training, make sure the # sample-based options are off. assert args.train_samples is None, "expected iteration-based training" - assert args.lr_decay_samples is None, ( - "expected iteration-based learning rate decay" - ) - assert args.lr_warmup_samples == 0, ( - "expected iteration-based learning rate warmup" - ) - assert args.rampup_batch_size is None, ( - "expected no batch-size rampup for iteration-based training" - ) + assert ( + args.lr_decay_samples is None + ), "expected iteration-based learning rate decay" + assert ( + args.lr_warmup_samples == 0 + ), "expected iteration-based learning rate warmup" + assert ( + args.rampup_batch_size is None + ), "expected no batch-size rampup for iteration-based training" if args.lr_warmup_fraction is not None: - assert args.lr_warmup_iters == 0, ( - "can only specify one of lr-warmup-fraction and lr-warmup-iters" - ) + assert ( + args.lr_warmup_iters == 0 + ), "can only specify one of lr-warmup-fraction and lr-warmup-iters" # Sample-based training. if args.train_samples: @@ -306,20 +306,20 @@ def validate_args(args, defaults={}): assert args.lr_decay_iters is None, "expected sample-based learning rate decay" assert args.lr_warmup_iters == 0, "expected sample-based learnig rate warmup" if args.lr_warmup_fraction is not None: - assert args.lr_warmup_samples == 0, ( - "can only specify one of lr-warmup-fraction and lr-warmup-samples" - ) + assert ( + args.lr_warmup_samples == 0 + ), "can only specify one of lr-warmup-fraction and lr-warmup-samples" if args.num_layers is not None: - assert args.encoder_num_layers is None, ( - "cannot have both num-layers and encoder-num-layers specified" - ) + assert ( + args.encoder_num_layers is None + ), "cannot have both num-layers and encoder-num-layers specified" args.encoder_num_layers = args.num_layers else: if not args.use_dataset_only: - assert args.encoder_num_layers is not None, ( - "either num-layers or encoder-num-layers should be specified" - ) + assert ( + args.encoder_num_layers is not None + ), "either num-layers or encoder-num-layers should be specified" args.num_layers = args.encoder_num_layers # Check required arguments. @@ -374,9 +374,9 @@ def validate_args(args, defaults={}): if args.fp16_lm_cross_entropy: assert args.fp16, "lm cross entropy in fp16 only support in fp16 mode." if args.fp32_residual_connection: - assert args.fp16 or args.bf16, ( - "residual connection in fp32 only supported when using fp16 or bf16." - ) + assert ( + args.fp16 or args.bf16 + ), "residual connection in fp32 only supported when using fp16 or bf16." if not args.use_dataset_only: if args.weight_decay_incr_style == "constant": @@ -430,13 +430,13 @@ def validate_args(args, defaults={}): # Tranformer-Engine/FP8 related checking if args.fp8_e4m3 or args.fp8_hybrid: - assert args.transformer_impl == "transformer_engine", ( - "transformer-engine required for fp8 training and inference" - ) + assert ( + args.transformer_impl == "transformer_engine" + ), "transformer-engine required for fp8 training and inference" - assert not (args.fp8_e4m3 and args.fp8_hybrid), ( - "cannot train with both fp8 e4m3 and hybrid formatting" - ) + assert not ( + args.fp8_e4m3 and args.fp8_hybrid + ), "cannot train with both fp8 e4m3 and hybrid formatting" if args.recompute_granularity == "selective": assert args.recompute_method is None, ( From c369de6431cd926db67fc88a38f4376b3865c898 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 06:37:15 -0500 Subject: [PATCH 461/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 33 +++++++++++++++++++++++++-------- 1 file changed, 25 insertions(+), 8 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index ad814bfd1d3..04283997a23 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -190,9 +190,12 @@ setup_run_cmd() { # min_lr=$(python3 -c 'print(f"{2 / (10 ** 5):.8f}")') # "--min-lr ${LR:-${min_lr}}" # 2e-5 # "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 - export LR="${LR:-0.0002}" - export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" - export LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" + LR="${LR:-0.0002}" + LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" + LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" + export LR + export LR_DECAY_STYLE + export LR_WARMUP_FRAC lr_flags=( "--lr ${LR}" "--lr-decay-style ${LR_DECAY_STYLE}" @@ -552,13 +555,13 @@ get_grad_acc_steps_on_aurora() { exit 1 fi nhosts=$(wc -l <"${hf}") - if [[ "${nhosts}" -ge 256 ]]; then + if [[ "${nhosts}" -ge 256 ]]; then # n >= 256 gas=1 - elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then + elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then # 128 <= n < 256 gas=2 - elif [[ 32 -lt "${nhosts}" && "${nhosts}" -lt 129 ]]; then + elif [[ 32 -lt "${nhosts}" && "${nhosts}" -lt 129 ]]; then # 32 < n < 128 gas=4 - elif [[ 16 -le "${nhosts}" && "${nhosts}" -le 32 ]]; then + elif [[ 16 -le "${nhosts}" && "${nhosts}" -le 32 ]]; then # 16 <= n < 32 gas=8 else gas=16 @@ -607,6 +610,15 @@ get_model_arch_7B() { export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 } +get_model_arch_llama3_3B() { + export HEADS=32 + export NLAYERS=28 + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 +} + # get_model_arch_70B() { # # 70B # export MODEL_KEY="AuroraGPT-70B" @@ -735,6 +747,8 @@ setParams() { # printf "Using model architecture: %s\n" "$(printGreen "${model_arch}")" if [[ "${MODEL_ARCH:-}" == "70B" ]]; then get_model_arch_70B + elif [[ "${MODEL_ARCH:-}" == "3B" ]]; then + get_model_arch_llama3_3B else get_model_arch_7B fi @@ -948,8 +962,11 @@ get_output_prefix() { pre="${pre}_seq${SEQ}_gb${GLOBAL_BATCH}" pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" - pre="${pre}_ntok${TRAIN_TOKENS}" + local num_tokens_in_billions + num_tokens_in_billions=$((TRAIN_TOKENS / 10 ** 9)) + pre="${pre}_ntok${num_tokens_in_billions}B" if [[ -n "${TOKENIZER_TYPE:-}" ]]; then + # _tok="${TOKENIZER_TYPE/Tokenizer//}" # Strip "Tokenizer" suffix if present _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa pre="${pre}_tok${_tok}" fi From 99dacb57e03c7afacf15e32f7015a3d048b469d5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 11:01:08 -0500 Subject: [PATCH 462/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 84 +++++++------------------------------------------ 1 file changed, 12 insertions(+), 72 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 04283997a23..18c022dbf03 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -187,23 +187,6 @@ setup_run_cmd() { "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" ) fi - # min_lr=$(python3 -c 'print(f"{2 / (10 ** 5):.8f}")') - # "--min-lr ${LR:-${min_lr}}" # 2e-5 - # "--min-lr ${MIN_LR:-"2e-6"}" # 2e-5 - LR="${LR:-0.0002}" - LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" - LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" - export LR - export LR_DECAY_STYLE - export LR_WARMUP_FRAC - lr_flags=( - "--lr ${LR}" - "--lr-decay-style ${LR_DECAY_STYLE}" - "--lr-warmup-fraction ${LR_WARMUP_FRAC}" - ) - if [[ -n "${LR_DECAY_ITERS:-}" ]]; then - lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") - fi tb_flags=() if [[ -z "${NO_TENSORBOARD:-}" ]]; then @@ -771,6 +754,18 @@ setParams() { # # Use best set of CCL env vars from Gordon Bell runs on Aurora # set_ccl_vars_on_aurora # fi + # + --[LR Settings]------------------------------------------------------+ + export LR="${LR:-0.0002}" + export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" + export LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" + lr_flags=( + "--lr ${LR}" + "--lr-decay-style ${LR_DECAY_STYLE}" + "--lr-warmup-fraction ${LR_WARMUP_FRAC}" + ) + if [[ -n "${LR_DECAY_ITERS:-}" ]]; then + lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") + fi # +---[Run Settings]------------------------------------------------------+ export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE @@ -835,21 +830,6 @@ set_args() { fi ds_args+=("--deepspeed_config=${DS_CONFIG}") ds_args+=("--zero-stage=$ZERO_STAGE") - - # if [[ "${ZERO_STAGE}" == 3 ]]; then - # ds_args+=("--use-mics") - # fi - - # ds_args=" " - # ds_args=" --deepspeed ${ds_args}" - # if [[ $PP == 1 ]]; then - # ds_args=" --no-pipeline-parallel ${ds_args}" - # fi - # ds_args=" --deepspeed_config=$DS_CONFIG ${ds_args}" - # ds_args="--zero-stage=$ZERO_STAGE ${ds_args}" - # if [[ "${ZERO_STAGE}" == 3 ]]; then - # ds_args="--use-mics ${ds_args}" - # fi if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args+=("--deepspeed-activation-checkpointing") @@ -884,46 +864,6 @@ make_ds_hostfile() { sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } -########################################### -# ezpz_setup -# -# 1. Clone [`saforem2/ezpz`](https://github.com/saforem2/ezpz) (if necessary) -# to `"${WORKING_DIR}/deps/ezpz/"` -# -# 2. Source [`ezpz/src/ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -# - This provides `{ezpz_setup_python, ezpz_setup_job}` (called below) -# -# 3. Call `ezpz_setup_python` (from `ezpz/bin/utils.sh`): -# - This will setup conda + virtual enviroment -# -# 4. Call `ezpz_setup_job` (from `ezpz/bin/utils.sh`): -# - This will parse `$PBS_*` variables and build launch cmd -# -# 3. Call `_ezpz_install` (from `Megatron-DeepSpeed/ALCF/helpers.sh`): -# - Install ezpz from `"${WORKING_DIR}/depz/ezpz/"` -########################################### -# ezpz_setup() { -# source <() -# ezdir="${WORKING_DIR}/deps/ezpz" -# if [[ -d "${ezdir}" ]]; then -# echo "Found ezpz in ${ezdir}" -# else -# mkdir -p "$(dirname "${ezdir}")" -# git clone https://github.com/saforem2/ezpz "${ezdir}" -# fi -# # shellcheck source=../deps/ezpz/src/ezpz/bin/utils.sh -# source "${ezdir}/src/ezpz/bin/utils.sh" || exit -# ezpz_setup_python -# ezpz_setup_job "$@" -# ezpz_pip_loc=$(python3 -m pip list | grep ezpz | awk '{print $NF}') -# if [[ -z "${ezpz_pip_loc:-}" ]]; then -# printf "[ezpz_install] Installing ezpz from %s\n" "${ezdir}" -# python3 -m pip install -e "${ezdir}" --require-virtualenv -# else -# printf "[ezpz_install] Found ezpz @ %s\n" "${ezpz_pip_loc}" -# fi -# } - ####################################################################### # ezpz_test: Run simple test to make sure all nodes in working order ####################################################################### From 6b08a2d327413899aa698da7794f225c6de48630 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 11:01:33 -0500 Subject: [PATCH 463/590] docs: Update `ALCF/notes/debugging.md` --- ALCF/notes/debugging.md | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/ALCF/notes/debugging.md b/ALCF/notes/debugging.md index 425785164bb..cdb49db69be 100644 --- a/ALCF/notes/debugging.md +++ b/ALCF/notes/debugging.md @@ -8,6 +8,15 @@ try that I've found useful. > and should work on _any_ of the ALCF systems[^alcf]. > (though it _should_ work anywhere, let me know if you run into issues!) +## 🔄 Turning it Off and On Again + +The first thing to try is to simply re-run your script, after resetting your +environment (see below): + +```bash +"${SHELL}" --login ; now=$(date +'%Y-%m-%d-%H%M%S'); debug_dir="${HOME}/debugging/${now}" && mkdir -p "${debug_dir}" && cd "${debug_dir}"&& git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed && source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env && python3 -m pip install --require-virtualenv "git+https://github.com/saforem2/ezpz" "numpy<2" deepspeed tensorboard && ezpz-test && DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt bash train_alcf.sh +``` + ## 🤔 Why Does this Happen? One of the most common issues encountered by users is a mangled environment. From 984cbb11eddf6327d83895c7f6d0c07db0f0630e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 12:41:52 -0500 Subject: [PATCH 464/590] chore: Update `megatron/training_log.py` --- megatron/training_log.py | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/megatron/training_log.py b/megatron/training_log.py index 394f34a048c..c571855da87 100644 --- a/megatron/training_log.py +++ b/megatron/training_log.py @@ -345,6 +345,7 @@ def training_log( ) if avg > 0.0: log_string += " {}={:.6f} |".format(key, avg) + wandb_metrics |= {f"loss/{key}_avg": avg} total_loss_dict[key] = accelerator.FloatTensor([0.0]) if loss_scale is not None: log_string += " loss_scale={:.1f} |".format(loss_scale) @@ -389,7 +390,11 @@ def training_log( "training/skiped_iterations": total_loss_dict[skipped_iters_key] } wandb_metrics |= {"training/nan_iterations": total_loss_dict[nan_iters_key]} - wandb.log(wandb_metrics) + if getattr(wandb, "log", None) is not None: + assert callable(wandb.log), ( + f"wandb.log is not callable, got {type(wandb.log)}" + ) + wandb.log(wandb_metrics) total_loss_dict[advanced_iters_key] = 0 total_loss_dict[skipped_iters_key] = 0 total_loss_dict[nan_iters_key] = 0 From ebb18999efe17a775772aa93963378e5f2e30f9f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 15 Jul 2025 12:42:03 -0500 Subject: [PATCH 465/590] chore: Update `megatron/timers.py` --- megatron/timers.py | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/megatron/timers.py b/megatron/timers.py index 5a9ba1f21d5..cd2c9c206d1 100644 --- a/megatron/timers.py +++ b/megatron/timers.py @@ -217,7 +217,16 @@ def _get_global_min_max_time(self, names, reset, barrier, normalizer): for i, name in enumerate(names): rank_to_time = rank_name_to_time[:, i] # filter out the ones we did not have any timings for - rank_to_time = rank_to_time[rank_to_time > 0.0] + try: + rank_to_time = rank_to_time[rank_to_time > 0.0] + except Exception: + # this can happen if rank_to_time is not a tensor + raise Exception( + "timer {} did not return a tensor, got {}".format( + name, type(rank_to_time) + ) + ) + # If the timer exists: if rank_to_time.numel() > 0: name_to_min_max_time[name] = ( From e3b0398d2f2d3f8ec543e99373ca14bd18a1e4f8 Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Tue, 22 Jul 2025 14:58:24 +0000 Subject: [PATCH 466/590] fixed infinite schedulers bugs and dshampoo name in arguments --- megatron/arguments.py | 2 +- megatron/optimizer_param_scheduler.py | 11 ++++++----- 2 files changed, 7 insertions(+), 6 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 709efaace9f..ce5582702b2 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1492,7 +1492,7 @@ def _add_training_args(parser): "galoreadamw8bitperlayer", "ipex.fusedlamb", "ipex.lamb", - "dshampooadamwsgd", + "dshampooadamw", "sgdschedulefree", "sophiag", "adopt", diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index 00614483aad..f62fb23697a 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -22,10 +22,10 @@ def __init__( end_wd, wd_incr_steps, wd_incr_style, + timescale=10.0, constant_lr=0.0, lr_constant_steps=None, lr_cooldown_steps=None, - timescale=None, use_checkpoint_opt_param_scheduler=True, override_opt_param_scheduler=False, ): @@ -48,7 +48,8 @@ def __init__( assert self.lr_decay_steps > 0 assert self.lr_warmup_steps < self.lr_decay_steps assert self.lr_constant_steps < self.lr_decay_steps - + + self.timescale=timescale self.lr_decay_tokens = args.lr_decay_tokens self.num_tokens = 0 self.lr_warmup_tokens = args.lr_warmup_tokens @@ -157,7 +158,7 @@ def get_lr(self): if self.lr_decay_tokens is None: num_steps_ = self.num_steps - self.lr_warmup_steps cooldown_steps_ = self.lr_cooldown_steps - self.lr_warmup_steps - if self.constant_steps is None: + if self.lr_constant_steps is None: raise Exception( "Constant LR steps need to be provided for infinite schedulers" ) @@ -173,7 +174,7 @@ def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 coeff = inv_f(1) * inv_f(cooldown_ratio) - lr = self.start_lr - delta_lr * coeff + lr = self.max_lr - delta_lr * coeff else: num_steps_ = num_steps_ - cooldown_steps_ @@ -215,7 +216,7 @@ def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 coeff = inv_f(1) * inv_f(cooldown_ratio) - lr = self.start_lr - delta_lr * coeff + lr = self.max_lr - delta_lr * coeff else: num_tokens_ = num_tokens_ - cooldown_tokens_ From 1508ff5e123dfb31fa6bf51268fd1c4cf1f9c4c2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Aug 2025 08:30:03 -0500 Subject: [PATCH 467/590] chore: Update `ALCF/README.md` --- ALCF/README.md | 127 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 127 insertions(+) diff --git a/ALCF/README.md b/ALCF/README.md index e782b27a939..7722c0a8328 100644 --- a/ALCF/README.md +++ b/ALCF/README.md @@ -1,5 +1,130 @@ # Megatron-DeepSpeed @ ALCF +> [!IMPORTANT] +> [`train_alcf.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_alcf.sh) +> is the main entry point for training @ ALCF + +## 🐣 Getting Started + +- 🏃‍♂️ Run Training: + + ```bash + git clone https://github.com/argonne-lcf/Megatron-DeepSpeed + cd Megatron-DeepSpeed + + # for PyTorch 2.5 environment (default): + source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env + # for PyTorch 2.8: ezpz_setup_env_pt28_aurora ⤴ + # for PyTorch 2.8: ezpz_setup_env_pt28_aurora + + # install dependencies + python3 -m pip install \ + --require-virtualenv \ + tensorboard deepspeed \ + "git+https://github.com/saforem2/ezpz" + + # test setup + ezpz-test + + # run training + bash train_alcf.sh + ``` + +> [!TIP] +> To enable logging with [Weights & Biases](https://wandb.ai/) (`wandb`), +> we need to install and login: +> +> ```bash +> python3 -m pip install wandb --upgrade +> wandb login +> ``` +> +> **NOTE**: W\&B can be disabled by setting `export WANDB_DISABLED=1` +> +> See [`wandb`: Quickstart](https://docs.wandb.ai/quickstart) for +> additional information + +This will default to using the default AuroraGPT-7B architecture with the +full [Dolma (v1.7)](https://huggingface.co/datasets/allenai/dolma) +dataset. + +## ⚙️ Configuration + +This is a simple subset of configurable options. + +The full list (as well as their default values) can be found in +[ALCF / `helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh) + +Any additional arguments passed to [`train_alcf.sh`](../train_alcf.sh) will be +forwarded to the argument parser in +[`pretrain_gpt_alcf.py`](../pretrain_gpt_alcf.py) + +### Environment Overrides + +- `DTYPE`: Data type +- `DATA_FILE_LIST`: Data file list +- `FFN_HIDDEN_SIZE`: Feedforward Neural Network projection size +- `GRAD_ACC_STEPS`: Gradient accumulation steps +- `HEADS`: Number of attention heads +- `HIDDEN`: Hidden size +- `MICRO_BATCH`: Micro batch size +- `NO_FLASH_ATTN`: No Flash Attention +- `NLAYERS`: Number of layers +- `NUM_KV_HEAD`: Number of key-value heads +- `OPT`: Optimizer + - `adam` + - `adam8bit` + - `adamw` + - `adamwschedulefree` + - `apex.adam` + - `apex.sgd` + - `ds.fusedlamb` + - `ds.onebitlamb` + - `galoreadamw` + - `galoreadamw8bit` + - `galoreadamw8bitperlayer` + - `ipex.fusedlamb` + - `ipex.lamb` + - `shampoo` + - `sgd` + - `sgdschedulefree` + - `sophiag` +- `PP`: Pipeline parallelism degree +- `SEQ`: Sequence length +- `SP`: Sequence parallelism (Ulysses) degree +- `TP`: Tensor parallelism degree +- `TRAIN_TOKENS`: Number of training tokens +- `TRAIN_ITERS`: Number of training iterations +- `USE_ACTIVATION_CHECKPOINTING`: Use activation checkpointing +- `WEIGHT_DECAY`: Weight decay +- `ZERO_STAGE`: Zero stage + +--- + +
Deprecated: + + + + + + > [!IMPORTANT] > [`train_aGPT_7B.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/train_aGPT_7B.sh) is the main entry point for launching > distributed training on {Polaris, Aurora, Sunspot} @ ALCF. @@ -207,3 +332,5 @@ cd /eagle/datasets/dolma/utils
+ +
From 1224815ff05d062ed8aaeb09a5a8cf97613ade55 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Aug 2025 08:30:24 -0500 Subject: [PATCH 468/590] feat: Create `train.sh` --- train.sh | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) create mode 100644 train.sh diff --git a/train.sh b/train.sh new file mode 100644 index 00000000000..e1f4e5e96cc --- /dev/null +++ b/train.sh @@ -0,0 +1,16 @@ +#!/bin/bash --login + +# 1. Source `ezpz/bin/uitils.sh` and setup {job, python} environment: +# NO_COLOR=1 source <(curl -sL https://bit.ly/ezpz-utils) && ezpz_setup_env +script -efq /dev/null -c "source <(curl -sL https://bit.ly/ezpz-utils) && ezpz_setup_env" +# script -q /dev/null source <(curl -sL https://bit.ly/ezpz-utils) && ezpz_setup_env + +# 2. Source `ALCF/helpers.sh` for Megatron-DeepSpeed setup +source "ALCF/helpers.sh" || exit + +# 3. Call `setup` from `./ALCF/helpers.sh` +setup "$@" || exit + +# 4. Run: +echo "${run_cmd[@]}" | tee -a "${OUTPUT_LOG}" +eval "${run_cmd[*]}" 2>&1 | tee -a "${OUTPUT_LOG}" From 6925291233bbc575b037c020ccb8fa4129cad68e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 7 Aug 2025 11:36:52 -0500 Subject: [PATCH 469/590] chore: Update `megatron/training_log_alcf.py` --- megatron/training_log_alcf.py | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/megatron/training_log_alcf.py b/megatron/training_log_alcf.py index dcd872971d7..a17e3141ca3 100644 --- a/megatron/training_log_alcf.py +++ b/megatron/training_log_alcf.py @@ -592,6 +592,8 @@ def training_log( tflops_lm = num_flop_per_sec_lm / (10**12) tflops_lm_per_gpu = tflops_lm / args.world_size wandb_metrics |= { + "iteration": iteration, + "train_iters": args.train_iters, "throughput/iteration-time": elapsed_time_per_iteration, # 1000 ms / s "throughput/samples_per_sec": samples_per_sec, "throughput/samples_per_sec_per_replica": samples_per_sec_per_replica, @@ -612,6 +614,12 @@ def training_log( **{f"loss/{k}": v for k, v in loss_dict.items()}, } if writer and args.log_timers_to_tensorboard: + wandb_metrics |= { + 'iteration-time/iteration': iteration, + 'iteration-time/iteration-time': elapsed_time_per_iteration, + 'iteration-time/consumed_train_samples': args.consumed_train_samples, + 'iteration-time/consumed_train_tokens': args.consumed_train_tokens, + } writer.add_scalar( "iteration-time/iteration-time", elapsed_time_per_iteration, iteration ) From dacd3d2ef39fbe125e1bb3ceca20b775106c4035 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 15 Aug 2025 07:44:21 -0500 Subject: [PATCH 470/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 187 +++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 154 insertions(+), 33 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 18c022dbf03..908fb0366da 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -124,7 +124,7 @@ setup() { check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` - tok="${TOKENIZER_TYPE:-Llama2Tokenizer}" + tok="${TOKENIZER_TYPE:-HFTokenizer}" setup_tokenizer_and_data "${tok}" "${dfl}" || exit make_data || exit # Print job info @@ -257,10 +257,12 @@ setup_run_cmd() { declare -A arch_map printf "==== ARCHITECTURE ====\n" arch_map=( + ["MODEL_ARCH"]="${MODEL_ARCH}" ["TP"]="${TP}" ["PP"]="${PP}" ["SP"]="${SP}" ["DP"]="${DP}" + ["ZERO"]="${ZERO_STAGE}" ["MBS"]="${MICRO_BATCH}" ["GAS"]="${GRAD_ACC_STEPS}" ["GBS"]="${GLOBAL_BATCH}" @@ -591,17 +593,104 @@ get_model_arch_7B() { export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export MODEL_ARCH="AuroraGPT-7B" } get_model_arch_llama3_3B() { - export HEADS=32 + export HEADS=24 export NLAYERS=28 export HIDDEN=3072 export NUM_KV_HEAD=8 export FFN_HIDDEN_SIZE=8192 export SEQ=8192 + export MODEL_ARCH="llama3-3B" + # [NOTE]----------------------------------- + # - OOM on Aurora with: + # - Config 1: + # - ZERO_STAGE=1 + # - USE_ACTIVATION_CHECKPOINTING=0 + # - Config 2: + # - ZERO_STAGE=0 + # - USE_ACTIVATION_CHECKPOINTING=1 + # + # - hmmmmm 🤔 seems odd??? + # - suspect due to SEQ_LEN=8192 + # (as opposed to 4096 for 7B model) + # ----------------------------------------- +} + + +get_model_arch_llama3_3B_customNlayers() { + # if [[ "$#" -ne 1 ]]; then + # log_message ERROR "Expected 1 argument (nLayers), received: $*" + # return 1 + # fi + export HEADS=24 + export NLAYERS="${NLAYERS:-28}" # default to 28 layers + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="llama3-3B-nLayers${NLAYERS}" } +get_model_arch_smollm3_3B() { + export HEADS=16 + export NLAYERS=36 + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="smollm3-3B" +} + +# get_model_arch_smollm3_3B_custom() { +# export HEADS=16 +# export NLAYERS=24 +# export HIDDEN=2048 +# export NUM_KV_HEAD=4 +# export FFN_HIDDEN_SIZE=11008 +# export SEQ=8192 +# export MODEL_ARCH="smollm3-3B-nLayers24" +# } + +get_model_arch_smollm3_3B_custom_nLayers() { + # if [[ "$#" -ne 1 ]]; then + # log_message ERROR "Expected 1 argument (nLayers), received: $*" + # return 1 + # fi + # local nlayers="$1" + export HEADS=16 + export NLAYERS="${NLAYERS:-24}" # default to 24 layers + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="smollm3-nLayers${NLAYERS}" +} + +get_model_arch_phi4_mini() { + export HEADS=32 + export NLAYERS=24 + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="phi4-mini" +} + +get_model_arch_phi4_mini_custom_nLayers() { + export HEADS=32 + export NLAYERS="${NLAYERS:-24}" # default to 24 layers + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" +} + + + # get_model_arch_70B() { # # 70B # export MODEL_KEY="AuroraGPT-70B" @@ -625,6 +714,7 @@ get_model_arch_70B() { FFN_HIDDEN_SIZE=28672 HIDDEN=8192 SEQ_LEN=8192 + export MODEL_ARCH="AuroraGPT-70B" } get_model_arch_33B() { @@ -641,6 +731,7 @@ get_model_arch_33B() { export FFN_HIDDEN_SIZE=11076 export SEQ=4096 export NUM_KV_HEAD=6 + export MODEL_ARCH="AuroraGPT-33B" } ############################################################################## @@ -726,15 +817,31 @@ setParams() { FLASH_ARG="--use-flash-attn-v2" fi fi - # model_arch="${MODEL_ARCH:-7B}" - # printf "Using model architecture: %s\n" "$(printGreen "${model_arch}")" - if [[ "${MODEL_ARCH:-}" == "70B" ]]; then - get_model_arch_70B - elif [[ "${MODEL_ARCH:-}" == "3B" ]]; then - get_model_arch_llama3_3B - else - get_model_arch_7B - fi + ma="${MODEL_ARCH:-7B}" + case "${ma}" in + # "70B" | "llama-3.1-70B" | "llama-3.1-70b" | "llama-3.2-70B" | "llama-3.2-70b") + "70B") + get_model_arch_70B + ;; + "33B" | "llama-3.2-33B" | "llama-3.2-33b") + get_model_arch_33B + ;; + "smollm3-3B" | "smollm3_3B") + get_model_arch_smollm3_3B_custom_nLayers + ;; + "phi4-mini" | "phi4_mini") + get_model_arch_phi4_mini_custom_nLayers + ;; + "llama3-3B" | "llama-3B") + get_model_arch_llama3_3B_customNlayers + ;; + "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") + get_model_arch_7B + ;; + *) + get_model_arch_7B + ;; + esac export TP="${TP}" export PP="${PP:-1}" export SP="${SP:-1}" @@ -783,14 +890,14 @@ setParams() { printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then - export TRAIN_TOKENS=${TRAIN_TOKENS:-2000000000000} + export TRAIN_TOKENS=${TRAIN_TOKENS:-4673780159710} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" else export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" fi - export MODEL_TYPE="llama-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL + export MODEL_TYPE="${MODEL_ARCH:-AuroraGPT}-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # NOTE: [2024-07-10] ##################################################### # - [sam]: For whatever reason, it seems that using # sequence-parallelism (SP) > 1 is INCOMPATIBLE with @@ -830,26 +937,31 @@ set_args() { fi ds_args+=("--deepspeed_config=${DS_CONFIG}") ds_args+=("--zero-stage=$ZERO_STAGE") - if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then + # if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then + if [[ "${USE_ACTIVATION_CHECKPOINTING:-}" == 1 || "${USE_ACTIVATION_CHECKPOINTING:-}" == "true" ]]; then echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args+=("--deepspeed-activation-checkpointing") + ds_args+=( + "--checkpoint-activations" + "--checkpoint-num-layers=${ACT_CKPT_NUM_LAYERS:-1}" + ) # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" # --checkpoint-activations \ # --deepspeed-activation-checkpointing fi export ds_args # --------------------------------------------------------------- - gpt_args=() - # we are now using activation checkpoint provided by megatron, see below. - # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" - if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then - echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - gpt_args+=( - "--checkpoint-activations" - "--checkpoint-num-layers ${ACT_CKPT_NUM_LAYERS}" - ) - fi - export gpt_args + # gpt_args=() + # # we are now using activation checkpoint provided by megatron, see below. + # # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" + # if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then + # echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" + # gpt_args+=( + # "--checkpoint-activations" + # "--checkpoint-num-layers ${ACT_CKPT_NUM_LAYERS}" + # ) + # fi + # export gpt_args } make_ds_hostfile() { @@ -897,11 +1009,12 @@ saveDSenv() { get_output_prefix() { # ---- Specify output location -------------------------------- - pre="ws${WORLD_SIZE}_ds_stage${ZERO_STAGE}_nl${NLAYERS}" - pre="${pre}_hs${HIDDEN}_mb${MICRO_BATCH}" - pre="${pre}_seq${SEQ}_gb${GLOBAL_BATCH}" - pre="${pre}_sp${SP}_pp${PP}_tp${TP}_${DTYPE}_opt${OPT}" - pre="${pre}_lr${LR}_lwf${LR_WARMUP_FRAC}" + pre="ws${WORLD_SIZE}-ds-stage${ZERO_STAGE}-nl${NLAYERS}" + pre="${pre}-hs${HIDDEN}-mb${MICRO_BATCH}" + pre="${pre}-seq${SEQ}-gb${GLOBAL_BATCH}" + pre="${pre}-sp${SP}-pp${PP}-tp${TP}-${DTYPE}-opt${OPT}" + pre="${pre}-lr${LR}-lwf${LR_WARMUP_FRAC}" + pre="${MODEL_ARCH:-AuroraGPT}-${pre}" local num_tokens_in_billions num_tokens_in_billions=$((TRAIN_TOKENS / 10 ** 9)) pre="${pre}_ntok${num_tokens_in_billions}B" @@ -910,6 +1023,12 @@ get_output_prefix() { _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa pre="${pre}_tok${_tok}" fi + if [[ -n "${TOKENIZER_MODEL:-}" ]]; then + # _tm=$(echo "${TOKENIZER_MODEL}" | sed 's/\/_/g') # noqa + # replace slashes with underscores + _tm="${TOKENIZER_MODEL//\//_}" # noqa + pre="${pre}_tm${_tm}" + fi if [[ -n "${LR_DECAY_ITERS}" ]]; then pre="${pre}_ldi${LR_DECAY_ITERS}" fi @@ -1125,15 +1244,17 @@ setup_tokenizer_and_data() { "--merge-file ${MERGE_FILE}" ) else - export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" - tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" + # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" + tm="${TOKENIZER_MODEL:-google/gemma-7B}" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ _tokenizer_flags+=( "--tokenizer-type ${TOKENIZER_TYPE}" "--tokenizer-model ${TOKENIZER_MODEL}" ) # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then - echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${DATA_FILE_LIST:-}" + echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${dfl}" setData "${dfl}" || exit fi export DATA_FLAGS="${_data_flags[*]:-}" From b69120132f73a329961582f1d5b3e551ef8654f0 Mon Sep 17 00:00:00 2001 From: Huihuo Zheng Date: Sat, 16 Aug 2025 01:26:13 -0500 Subject: [PATCH 471/590] cache indices support --- megatron/data/gpt_dataset.py | 66 ++++++++++++++++++++++++++++++++---- 1 file changed, 59 insertions(+), 7 deletions(-) diff --git a/megatron/data/gpt_dataset.py b/megatron/data/gpt_dataset.py index d09f08d63a4..608a6a916d6 100644 --- a/megatron/data/gpt_dataset.py +++ b/megatron/data/gpt_dataset.py @@ -158,7 +158,7 @@ def _build_indices_blended(): "{:.2f} (sec)".format(time.time() - start_time)) return dataset_index, dataset_sample_index - + @dlp.log def _build_indices_concat(): start_time = time.time() dataset_index = np.zeros(self.num_samples, dtype=np.int64) @@ -175,11 +175,63 @@ def _build_indices_concat(): "{:.2f} (sec)".format(time.time() - start_time) ) return dataset_index, dataset_sample_index - - if args.blend_sample_in_corpus: - self.dataset_index, self.dataset_sample_index = _build_indices_blended() - else: - self.dataset_index, self.dataset_sample_index = _build_indices_concat() + + def _build_indices(): + if args.blend_sample_in_corpus: + return _build_indices_blended() + else: + return _build_indices_concat() + + def _cache_indices(): + desc = self.dataset_builders[0].corpus + desc += f"\n {self.num_samples}" + self.dataset_index = np.zeros(self.num_samples, dtype=np.int64) + self.dataset_sample_index = np.zeros(self.num_samples, dtype=np.int64) + if data_cache_path: + desc_hash = hashlib.md5(desc.encode('utf-8')).hexdigest() + desc_path = os.path.join(data_cache_path, desc_hash + ".dsc") + index_path = os.path.join(data_cache_path, desc_hash + "_index.npy") + sample_index_path = os.path.join(data_cache_path, desc_hash + "_sample_index.npy") + cache_hit = os.path.isfile(index_path) and os.path.isfile(sample_index_path) + cache_success = True + if torch.distributed.get_rank() == 0 and not cache_hit: + print(' > WARNING: could not find index map files for blendable' + ' dataset, building indices on rank 0 ...', flush=True) + dataset_index, dataset_sample_index = _build_indices() + try: + log.debug(" > saving index map files") + start_time = time.perf_counter() + os.makedirs(os.path.dirname(index_path), exist_ok=True) + with open(desc_path, 'wt') as fd: + fd.write(desc) + np.save(index_path, dataset_index, allow_pickle=True) + np.save(sample_index_path, dataset_sample_index, + allow_pickle=True) + log.info(f" > finished saving {self.dataset_builders[0].corpus} corpus index map files in {time.perf_counter() - start_time} seconds") + except OSError: + print(f'There was an error trying to create the data cache directory ({data_cache_path})') + print('or a file in it. This is set with the --data-cache-path argument. Please') + print('ensure you have write access to this directory or specify one that you do have') + print('write access to.') + cache_success = False + self.dataset_index = dataset_index + self.dataset_sample_index = dataset_sample_index + torch.distributed.barrier(group=mpu.get_data_parallel_group()) + torch.distributed.barrier(group=mpu.get_pipeline_model_parallel_group()) + torch.distributed.barrier(group=mpu.get_data_parallel_group()) + + start_time = time.perf_counter() + log.info(f'> loading {self.dataset_builders[0].corpus} corpus dataset index: {index_path}') + self.dataset_index = np.load(index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_index.size == self.num_samples + log.info(f'> loading {self.dataset_builders[0].corpus} corpus dataset sample index: {sample_index_path}') + self.dataset_sample_index = np.load(sample_index_path, allow_pickle=True, mmap_mode='r') + assert self.dataset_sample_index.size == self.num_samples + log.info(f'> finished loading in {time.perf_counter() - start_time} seconds') + else: + self.dataset_index, self.dataset_sample_index = _build_indices() + + _cache_indices() np_rng = np.random.RandomState(seed=dataset_builders[0].seed) self.shuffle_index = np.arange(self.num_samples) @@ -189,7 +241,7 @@ def _build_indices_concat(): self.desc += dataset_builders[i].prefix + "," log.info( - f"[BuildConcatDataset] Caught {args.shuffle_sample_in_corpus=} across" + f"[BuildCorpusDataset] Caught {args.shuffle_sample_in_corpus=} across" f" {self.num_samples} samples" ) self.desc += ( From d64abca865a5e7e42958d467f0673297c1adeeff Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 08:42:25 -0500 Subject: [PATCH 472/590] feat: Add `ALCF/data-lists/aurora/olmo-mix-1124.txt` --- ALCF/data-lists/aurora/olmo-mix-1124.txt | 1438 ++++++++++++++++++++++ 1 file changed, 1438 insertions(+) create mode 100644 ALCF/data-lists/aurora/olmo-mix-1124.txt diff --git a/ALCF/data-lists/aurora/olmo-mix-1124.txt b/ALCF/data-lists/aurora/olmo-mix-1124.txt new file mode 100644 index 00000000000..69de81919aa --- /dev/null +++ b/ALCF/data-lists/aurora/olmo-mix-1124.txt @@ -0,0 +1,1438 @@ +0.0016916184 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/wiki/fused_0001_of_0002_text_document wiki +0.0008750710 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/wiki/fused_0002_of_0002_text_document wiki +0.0035602442 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/algebraic-stack/fused_0001_of_0003_text_document algebraic-stack +0.0040678996 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/algebraic-stack/fused_0002_of_0003_text_document algebraic-stack +0.0009361344 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/algebraic-stack/fused_0003_of_0003_text_document algebraic-stack +0.0020628192 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0022_of_0025_text_document pes2o +0.0007015969 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0001_of_0025_text_document pes2o +0.0006992163 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0006_of_0025_text_document pes2o +0.0019571719 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0009_of_0025_text_document pes2o +0.0019548375 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0014_of_0025_text_document pes2o +0.0007066820 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0007_of_0025_text_document pes2o +0.0020597481 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0016_of_0025_text_document pes2o +0.0019533494 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0018_of_0025_text_document pes2o +0.0019531326 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0017_of_0025_text_document pes2o +0.0019628659 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0021_of_0025_text_document pes2o +0.0019561544 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0015_of_0025_text_document pes2o +0.0019617991 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0019_of_0025_text_document pes2o +0.0019525277 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0010_of_0025_text_document pes2o +0.0019572305 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0023_of_0025_text_document pes2o +0.0020636509 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0011_of_0025_text_document pes2o +0.0008565472 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0008_of_0025_text_document pes2o +0.0019624300 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0024_of_0025_text_document pes2o +0.0007039007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0002_of_0025_text_document pes2o +0.0020620877 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0012_of_0025_text_document pes2o +0.0007031983 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0004_of_0025_text_document pes2o +0.0007040202 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0003_of_0025_text_document pes2o +0.0007034923 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0005_of_0025_text_document pes2o +0.0022880258 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0025_of_0025_text_document pes2o +0.0020648551 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0013_of_0025_text_document pes2o +0.0020587847 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/pes2o/fused_0020_of_0025_text_document pes2o +0.0032668430 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/open-web-math/fused_0001_of_0003_text_document open-web-math +0.0026537500 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/open-web-math/fused_0002_of_0003_text_document open-web-math +0.0028121815 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/open-web-math/fused_0003_of_0003_text_document open-web-math +0.0014233454 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/arxiv/fused_0005_of_0005_text_document arxiv +0.0026645337 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/arxiv/fused_0004_of_0005_text_document arxiv +0.0033332955 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/arxiv/fused_0002_of_0005_text_document arxiv +0.0038234059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/arxiv/fused_0001_of_0005_text_document arxiv +0.0032411595 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/arxiv/fused_0003_of_0005_text_document arxiv +0.0018294835 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0514_of_1380_text_document dclm +0.0018507614 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1245_of_1380_text_document dclm +0.0019057482 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0657_of_1380_text_document dclm +0.0018703259 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0402_of_1380_text_document dclm +0.0018761172 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0193_of_1380_text_document dclm +0.0018352288 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0081_of_1380_text_document dclm +0.0018542042 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0747_of_1380_text_document dclm +0.0019132349 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0893_of_1380_text_document dclm +0.0018986856 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1103_of_1380_text_document dclm +0.0018512106 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1214_of_1380_text_document dclm +0.0019219147 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0479_of_1380_text_document dclm +0.0018412034 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0427_of_1380_text_document dclm +0.0018822748 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0194_of_1380_text_document dclm +0.0018778341 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0379_of_1380_text_document dclm +0.0018892680 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0679_of_1380_text_document dclm +0.0017777129 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0247_of_1380_text_document dclm +0.0018257472 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0332_of_1380_text_document dclm +0.0018929278 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0164_of_1380_text_document dclm +0.0018973543 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0772_of_1380_text_document dclm +0.0018923846 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0007_of_1380_text_document dclm +0.0019048418 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1312_of_1380_text_document dclm +0.0018609074 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1167_of_1380_text_document dclm +0.0019044918 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0042_of_1380_text_document dclm +0.0018691585 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1185_of_1380_text_document dclm +0.0019162346 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0771_of_1380_text_document dclm +0.0018715739 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0586_of_1380_text_document dclm +0.0018754377 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0197_of_1380_text_document dclm +0.0018443191 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0053_of_1380_text_document dclm +0.0019074914 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0992_of_1380_text_document dclm +0.0018906962 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1235_of_1380_text_document dclm +0.0018074489 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0476_of_1380_text_document dclm +0.0018651080 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1054_of_1380_text_document dclm +0.0019007134 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0468_of_1380_text_document dclm +0.0018628169 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0792_of_1380_text_document dclm +0.0018280937 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0581_of_1380_text_document dclm +0.0018797597 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0235_of_1380_text_document dclm +0.0018177165 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0416_of_1380_text_document dclm +0.0018375578 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0575_of_1380_text_document dclm +0.0018127794 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0313_of_1380_text_document dclm +0.0018422362 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0856_of_1380_text_document dclm +0.0018829563 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0201_of_1380_text_document dclm +0.0018698975 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1206_of_1380_text_document dclm +0.0018489451 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0212_of_1380_text_document dclm +0.0018627014 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0060_of_1380_text_document dclm +0.0018608744 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0330_of_1380_text_document dclm +0.0018958050 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0038_of_1380_text_document dclm +0.0018401793 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0523_of_1380_text_document dclm +0.0018474051 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1213_of_1380_text_document dclm +0.0018500122 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1290_of_1380_text_document dclm +0.0018712428 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0128_of_1380_text_document dclm +0.0018795686 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1130_of_1380_text_document dclm +0.0019244501 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0371_of_1380_text_document dclm +0.0018485262 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0185_of_1380_text_document dclm +0.0018864832 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0547_of_1380_text_document dclm +0.0018297940 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0526_of_1380_text_document dclm +0.0018246380 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0444_of_1380_text_document dclm +0.0018548483 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0812_of_1380_text_document dclm +0.0018318399 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0321_of_1380_text_document dclm +0.0018494539 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1025_of_1380_text_document dclm +0.0018808494 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0275_of_1380_text_document dclm +0.0019095425 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1374_of_1380_text_document dclm +0.0018314888 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1176_of_1380_text_document dclm +0.0019122305 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0824_of_1380_text_document dclm +0.0018382881 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1226_of_1380_text_document dclm +0.0018257026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0425_of_1380_text_document dclm +0.0018329238 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0529_of_1380_text_document dclm +0.0019004427 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0998_of_1380_text_document dclm +0.0018316324 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0354_of_1380_text_document dclm +0.0018452801 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1236_of_1380_text_document dclm +0.0018289322 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1358_of_1380_text_document dclm +0.0018987828 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0066_of_1380_text_document dclm +0.0018341039 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0609_of_1380_text_document dclm +0.0018739570 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0565_of_1380_text_document dclm +0.0018423487 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0232_of_1380_text_document dclm +0.0018414757 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1180_of_1380_text_document dclm +0.0018673039 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0821_of_1380_text_document dclm +0.0018570268 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0478_of_1380_text_document dclm +0.0018741098 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1260_of_1380_text_document dclm +0.0018323693 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0787_of_1380_text_document dclm +0.0019105024 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0177_of_1380_text_document dclm +0.0018579480 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1244_of_1380_text_document dclm +0.0018897371 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1280_of_1380_text_document dclm +0.0018339781 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1341_of_1380_text_document dclm +0.0018675967 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0887_of_1380_text_document dclm +0.0018307571 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1173_of_1380_text_document dclm +0.0019018056 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0474_of_1380_text_document dclm +0.0018374219 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0045_of_1380_text_document dclm +0.0018938192 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0489_of_1380_text_document dclm +0.0019229224 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1267_of_1380_text_document dclm +0.0018038026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0895_of_1380_text_document dclm +0.0018717950 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0743_of_1380_text_document dclm +0.0018873804 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0828_of_1380_text_document dclm +0.0018665323 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0837_of_1380_text_document dclm +0.0018968838 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0372_of_1380_text_document dclm +0.0018963942 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0466_of_1380_text_document dclm +0.0018541771 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0151_of_1380_text_document dclm +0.0018698711 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0732_of_1380_text_document dclm +0.0018894950 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0127_of_1380_text_document dclm +0.0018452070 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1331_of_1380_text_document dclm +0.0019083166 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0257_of_1380_text_document dclm +0.0018727883 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0470_of_1380_text_document dclm +0.0017949268 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0495_of_1380_text_document dclm +0.0018674750 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0681_of_1380_text_document dclm +0.0018820689 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0475_of_1380_text_document dclm +0.0018341200 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0113_of_1380_text_document dclm +0.0018382941 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0764_of_1380_text_document dclm +0.0018817227 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0944_of_1380_text_document dclm +0.0018771059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0117_of_1380_text_document dclm +0.0018942741 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0079_of_1380_text_document dclm +0.0019245775 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0342_of_1380_text_document dclm +0.0018553394 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0760_of_1380_text_document dclm +0.0018901312 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0315_of_1380_text_document dclm +0.0018214595 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0757_of_1380_text_document dclm +0.0018886267 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0030_of_1380_text_document dclm +0.0018862606 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0373_of_1380_text_document dclm +0.0018750463 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1067_of_1380_text_document dclm +0.0019085327 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0825_of_1380_text_document dclm +0.0018545905 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0670_of_1380_text_document dclm +0.0019265063 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1195_of_1380_text_document dclm +0.0018365273 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0671_of_1380_text_document dclm +0.0018228662 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0460_of_1380_text_document dclm +0.0018484202 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0173_of_1380_text_document dclm +0.0018331162 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0929_of_1380_text_document dclm +0.0018968216 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0433_of_1380_text_document dclm +0.0018728489 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1107_of_1380_text_document dclm +0.0018190723 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1175_of_1380_text_document dclm +0.0018297789 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0170_of_1380_text_document dclm +0.0018905247 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1355_of_1380_text_document dclm +0.0018346736 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0126_of_1380_text_document dclm +0.0018700717 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0222_of_1380_text_document dclm +0.0018687740 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1187_of_1380_text_document dclm +0.0018427712 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0623_of_1380_text_document dclm +0.0018915909 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0286_of_1380_text_document dclm +0.0018636094 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0031_of_1380_text_document dclm +0.0018468998 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0997_of_1380_text_document dclm +0.0018560440 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1184_of_1380_text_document dclm +0.0018214870 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1293_of_1380_text_document dclm +0.0018172191 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0218_of_1380_text_document dclm +0.0018290169 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0237_of_1380_text_document dclm +0.0018288824 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0990_of_1380_text_document dclm +0.0018716118 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1281_of_1380_text_document dclm +0.0018794268 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0242_of_1380_text_document dclm +0.0017881114 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1006_of_1380_text_document dclm +0.0018695477 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0092_of_1380_text_document dclm +0.0018693027 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0889_of_1380_text_document dclm +0.0018972631 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0638_of_1380_text_document dclm +0.0018356020 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0509_of_1380_text_document dclm +0.0018897586 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0511_of_1380_text_document dclm +0.0018384992 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0410_of_1380_text_document dclm +0.0018880726 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0391_of_1380_text_document dclm +0.0018235586 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0299_of_1380_text_document dclm +0.0018914402 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0736_of_1380_text_document dclm +0.0018816617 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0213_of_1380_text_document dclm +0.0018569127 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1313_of_1380_text_document dclm +0.0018725271 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0602_of_1380_text_document dclm +0.0018843571 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0290_of_1380_text_document dclm +0.0019073376 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1322_of_1380_text_document dclm +0.0018417191 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0680_of_1380_text_document dclm +0.0018980609 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1254_of_1380_text_document dclm +0.0018677054 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1345_of_1380_text_document dclm +0.0018367428 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1026_of_1380_text_document dclm +0.0018378824 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0699_of_1380_text_document dclm +0.0018058036 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1076_of_1380_text_document dclm +0.0018855970 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1138_of_1380_text_document dclm +0.0018880120 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0241_of_1380_text_document dclm +0.0018245854 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0898_of_1380_text_document dclm +0.0018701538 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0224_of_1380_text_document dclm +0.0018432957 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0276_of_1380_text_document dclm +0.0018598787 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0037_of_1380_text_document dclm +0.0018639489 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0229_of_1380_text_document dclm +0.0019173725 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0124_of_1380_text_document dclm +0.0018594884 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0810_of_1380_text_document dclm +0.0018095439 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0574_of_1380_text_document dclm +0.0018622776 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1259_of_1380_text_document dclm +0.0018825616 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0603_of_1380_text_document dclm +0.0019246269 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0137_of_1380_text_document dclm +0.0018795127 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0147_of_1380_text_document dclm +0.0018866601 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1105_of_1380_text_document dclm +0.0018796094 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0262_of_1380_text_document dclm +0.0018433990 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0166_of_1380_text_document dclm +0.0018457154 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0921_of_1380_text_document dclm +0.0019166336 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0966_of_1380_text_document dclm +0.0018425678 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0651_of_1380_text_document dclm +0.0018472643 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0314_of_1380_text_document dclm +0.0019260864 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0924_of_1380_text_document dclm +0.0018877908 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0530_of_1380_text_document dclm +0.0018549938 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0308_of_1380_text_document dclm +0.0018447361 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0401_of_1380_text_document dclm +0.0018817288 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1328_of_1380_text_document dclm +0.0018576826 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1288_of_1380_text_document dclm +0.0018401439 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1163_of_1380_text_document dclm +0.0018631989 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0236_of_1380_text_document dclm +0.0018444213 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1093_of_1380_text_document dclm +0.0018741967 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0421_of_1380_text_document dclm +0.0018937940 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0942_of_1380_text_document dclm +0.0018935839 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0393_of_1380_text_document dclm +0.0019134968 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1338_of_1380_text_document dclm +0.0018675200 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0084_of_1380_text_document dclm +0.0018772070 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0762_of_1380_text_document dclm +0.0018705388 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0618_of_1380_text_document dclm +0.0019011909 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0492_of_1380_text_document dclm +0.0018746500 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0244_of_1380_text_document dclm +0.0019221765 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1366_of_1380_text_document dclm +0.0018651238 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1253_of_1380_text_document dclm +0.0018843114 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1072_of_1380_text_document dclm +0.0018259995 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0727_of_1380_text_document dclm +0.0018340976 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0210_of_1380_text_document dclm +0.0018469574 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0525_of_1380_text_document dclm +0.0018479453 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0563_of_1380_text_document dclm +0.0019242407 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0270_of_1380_text_document dclm +0.0018455479 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0426_of_1380_text_document dclm +0.0018325862 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1334_of_1380_text_document dclm +0.0018742951 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0065_of_1380_text_document dclm +0.0018697997 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0078_of_1380_text_document dclm +0.0018619267 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0111_of_1380_text_document dclm +0.0018554206 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0874_of_1380_text_document dclm +0.0019154249 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0412_of_1380_text_document dclm +0.0018443356 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0008_of_1380_text_document dclm +0.0018229357 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1129_of_1380_text_document dclm +0.0018899491 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1296_of_1380_text_document dclm +0.0018432649 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0841_of_1380_text_document dclm +0.0018257777 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0848_of_1380_text_document dclm +0.0019001873 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0076_of_1380_text_document dclm +0.0018478680 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0519_of_1380_text_document dclm +0.0018009501 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0822_of_1380_text_document dclm +0.0018640925 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0721_of_1380_text_document dclm +0.0018674438 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0600_of_1380_text_document dclm +0.0018682718 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1043_of_1380_text_document dclm +0.0018427549 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0264_of_1380_text_document dclm +0.0018558621 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0215_of_1380_text_document dclm +0.0018258300 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0388_of_1380_text_document dclm +0.0019065056 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1115_of_1380_text_document dclm +0.0019003209 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0773_of_1380_text_document dclm +0.0018347884 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0406_of_1380_text_document dclm +0.0018857147 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1090_of_1380_text_document dclm +0.0018655813 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0767_of_1380_text_document dclm +0.0018389277 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0793_of_1380_text_document dclm +0.0018738371 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0039_of_1380_text_document dclm +0.0018362008 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0570_of_1380_text_document dclm +0.0018686516 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0303_of_1380_text_document dclm +0.0018206997 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0776_of_1380_text_document dclm +0.0019071614 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0712_of_1380_text_document dclm +0.0018008154 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0945_of_1380_text_document dclm +0.0018590650 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0477_of_1380_text_document dclm +0.0019155993 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0658_of_1380_text_document dclm +0.0017769544 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0731_of_1380_text_document dclm +0.0018799115 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0195_of_1380_text_document dclm +0.0018888472 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1292_of_1380_text_document dclm +0.0018334075 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0029_of_1380_text_document dclm +0.0019020180 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0488_of_1380_text_document dclm +0.0018281964 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1102_of_1380_text_document dclm +0.0019314419 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0963_of_1380_text_document dclm +0.0018519837 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0755_of_1380_text_document dclm +0.0018099277 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0876_of_1380_text_document dclm +0.0018738874 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0860_of_1380_text_document dclm +0.0018872350 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0490_of_1380_text_document dclm +0.0018479014 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0621_of_1380_text_document dclm +0.0019091327 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0304_of_1380_text_document dclm +0.0018965531 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0389_of_1380_text_document dclm +0.0018653304 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1223_of_1380_text_document dclm +0.0018987721 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1309_of_1380_text_document dclm +0.0018112236 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0935_of_1380_text_document dclm +0.0018703577 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0077_of_1380_text_document dclm +0.0018987462 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0939_of_1380_text_document dclm +0.0018412652 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0109_of_1380_text_document dclm +0.0018774230 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0016_of_1380_text_document dclm +0.0018200938 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1155_of_1380_text_document dclm +0.0019167423 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0014_of_1380_text_document dclm +0.0018941669 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0327_of_1380_text_document dclm +0.0019016732 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1192_of_1380_text_document dclm +0.0018013874 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1104_of_1380_text_document dclm +0.0018199958 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0011_of_1380_text_document dclm +0.0018604353 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0959_of_1380_text_document dclm +0.0018784838 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1215_of_1380_text_document dclm +0.0018696938 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0748_of_1380_text_document dclm +0.0019072044 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0405_of_1380_text_document dclm +0.0018818575 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0099_of_1380_text_document dclm +0.0018652260 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1327_of_1380_text_document dclm +0.0018861993 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0885_of_1380_text_document dclm +0.0018843313 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0559_of_1380_text_document dclm +0.0019018207 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0217_of_1380_text_document dclm +0.0018654141 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0720_of_1380_text_document dclm +0.0018124821 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0551_of_1380_text_document dclm +0.0018984222 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0894_of_1380_text_document dclm +0.0018407002 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1199_of_1380_text_document dclm +0.0018946082 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0331_of_1380_text_document dclm +0.0019001650 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0859_of_1380_text_document dclm +0.0018670612 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1057_of_1380_text_document dclm +0.0018990572 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0300_of_1380_text_document dclm +0.0018995237 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1377_of_1380_text_document dclm +0.0019236742 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0380_of_1380_text_document dclm +0.0018612864 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1121_of_1380_text_document dclm +0.0018673395 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0116_of_1380_text_document dclm +0.0018792485 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1337_of_1380_text_document dclm +0.0018394065 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1091_of_1380_text_document dclm +0.0018532436 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0447_of_1380_text_document dclm +0.0018405922 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0415_of_1380_text_document dclm +0.0018444330 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0006_of_1380_text_document dclm +0.0018940195 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0063_of_1380_text_document dclm +0.0018463930 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0231_of_1380_text_document dclm +0.0018866914 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0360_of_1380_text_document dclm +0.0018457370 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0735_of_1380_text_document dclm +0.0018782653 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0711_of_1380_text_document dclm +0.0018847284 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0438_of_1380_text_document dclm +0.0018423815 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0900_of_1380_text_document dclm +0.0019027074 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0035_of_1380_text_document dclm +0.0018361075 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0064_of_1380_text_document dclm +0.0018940670 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0413_of_1380_text_document dclm +0.0018215708 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0130_of_1380_text_document dclm +0.0019162600 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0204_of_1380_text_document dclm +0.0019061173 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1153_of_1380_text_document dclm +0.0018285422 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1108_of_1380_text_document dclm +0.0018497699 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1301_of_1380_text_document dclm +0.0018720661 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0046_of_1380_text_document dclm +0.0018240812 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0834_of_1380_text_document dclm +0.0018709973 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0953_of_1380_text_document dclm +0.0018484622 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0734_of_1380_text_document dclm +0.0018749492 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0205_of_1380_text_document dclm +0.0018585468 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0872_of_1380_text_document dclm +0.0018993125 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0174_of_1380_text_document dclm +0.0019160923 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1212_of_1380_text_document dclm +0.0018417181 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0808_of_1380_text_document dclm +0.0018621153 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0814_of_1380_text_document dclm +0.0019240997 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0923_of_1380_text_document dclm +0.0018483985 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0869_of_1380_text_document dclm +0.0019309032 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0642_of_1380_text_document dclm +0.0019061166 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0260_of_1380_text_document dclm +0.0019049171 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0971_of_1380_text_document dclm +0.0018080877 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1242_of_1380_text_document dclm +0.0018669705 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0816_of_1380_text_document dclm +0.0018270375 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0504_of_1380_text_document dclm +0.0018689676 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0975_of_1380_text_document dclm +0.0018854989 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1335_of_1380_text_document dclm +0.0019021632 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1266_of_1380_text_document dclm +0.0019086154 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1348_of_1380_text_document dclm +0.0019105257 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0149_of_1380_text_document dclm +0.0018688622 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0467_of_1380_text_document dclm +0.0018549201 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1342_of_1380_text_document dclm +0.0018687808 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0362_of_1380_text_document dclm +0.0018876964 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0073_of_1380_text_document dclm +0.0018852231 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0967_of_1380_text_document dclm +0.0018651567 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0381_of_1380_text_document dclm +0.0019156936 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1036_of_1380_text_document dclm +0.0018134282 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0683_of_1380_text_document dclm +0.0018715365 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0930_of_1380_text_document dclm +0.0018856602 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1203_of_1380_text_document dclm +0.0019018166 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0102_of_1380_text_document dclm +0.0018596122 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0349_of_1380_text_document dclm +0.0018970324 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0880_of_1380_text_document dclm +0.0018596011 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0363_of_1380_text_document dclm +0.0018404441 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0206_of_1380_text_document dclm +0.0018731259 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0846_of_1380_text_document dclm +0.0018680015 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1111_of_1380_text_document dclm +0.0018484725 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0995_of_1380_text_document dclm +0.0018462672 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0418_of_1380_text_document dclm +0.0018043945 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1032_of_1380_text_document dclm +0.0018768238 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0424_of_1380_text_document dclm +0.0018691867 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1196_of_1380_text_document dclm +0.0018549773 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0448_of_1380_text_document dclm +0.0018428781 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1238_of_1380_text_document dclm +0.0018526826 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1332_of_1380_text_document dclm +0.0018763549 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0142_of_1380_text_document dclm +0.0018312940 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0180_of_1380_text_document dclm +0.0019130115 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1228_of_1380_text_document dclm +0.0018356139 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1169_of_1380_text_document dclm +0.0018757080 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0566_of_1380_text_document dclm +0.0018922094 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1252_of_1380_text_document dclm +0.0018696378 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1216_of_1380_text_document dclm +0.0018564618 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0358_of_1380_text_document dclm +0.0018747817 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0596_of_1380_text_document dclm +0.0018037933 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1056_of_1380_text_document dclm +0.0018842334 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1152_of_1380_text_document dclm +0.0018880962 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0908_of_1380_text_document dclm +0.0018852454 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0863_of_1380_text_document dclm +0.0019258935 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0826_of_1380_text_document dclm +0.0018664390 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0703_of_1380_text_document dclm +0.0018539604 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0754_of_1380_text_document dclm +0.0018820039 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0374_of_1380_text_document dclm +0.0018706501 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0156_of_1380_text_document dclm +0.0018738398 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1239_of_1380_text_document dclm +0.0018520938 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0344_of_1380_text_document dclm +0.0018583268 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1246_of_1380_text_document dclm +0.0018461179 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0862_of_1380_text_document dclm +0.0018591204 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1321_of_1380_text_document dclm +0.0018448338 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1037_of_1380_text_document dclm +0.0018482994 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0697_of_1380_text_document dclm +0.0018396398 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0541_of_1380_text_document dclm +0.0018373273 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0104_of_1380_text_document dclm +0.0018566847 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0234_of_1380_text_document dclm +0.0018398595 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0430_of_1380_text_document dclm +0.0019139834 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0082_of_1380_text_document dclm +0.0018918316 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0840_of_1380_text_document dclm +0.0018297887 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0831_of_1380_text_document dclm +0.0018751209 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1066_of_1380_text_document dclm +0.0019008854 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0796_of_1380_text_document dclm +0.0018463936 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0051_of_1380_text_document dclm +0.0018536237 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0274_of_1380_text_document dclm +0.0018667058 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0211_of_1380_text_document dclm +0.0018575307 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1171_of_1380_text_document dclm +0.0018484386 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1272_of_1380_text_document dclm +0.0018953383 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0622_of_1380_text_document dclm +0.0018708542 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0763_of_1380_text_document dclm +0.0018657670 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0782_of_1380_text_document dclm +0.0018778205 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1160_of_1380_text_document dclm +0.0019147164 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0369_of_1380_text_document dclm +0.0018866640 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0129_of_1380_text_document dclm +0.0018386811 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0982_of_1380_text_document dclm +0.0018644587 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1357_of_1380_text_document dclm +0.0018384335 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0909_of_1380_text_document dclm +0.0018256657 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0115_of_1380_text_document dclm +0.0018910288 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0112_of_1380_text_document dclm +0.0018587420 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1211_of_1380_text_document dclm +0.0019021554 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0738_of_1380_text_document dclm +0.0019206735 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1336_of_1380_text_document dclm +0.0018502001 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1095_of_1380_text_document dclm +0.0018063843 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0392_of_1380_text_document dclm +0.0018237877 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1282_of_1380_text_document dclm +0.0019197506 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1365_of_1380_text_document dclm +0.0018727967 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1136_of_1380_text_document dclm +0.0018451986 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0616_of_1380_text_document dclm +0.0019028778 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0301_of_1380_text_document dclm +0.0018476846 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0382_of_1380_text_document dclm +0.0018481131 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0590_of_1380_text_document dclm +0.0018638395 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0100_of_1380_text_document dclm +0.0018998986 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0568_of_1380_text_document dclm +0.0019052566 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1333_of_1380_text_document dclm +0.0018851817 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0238_of_1380_text_document dclm +0.0018500819 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0390_of_1380_text_document dclm +0.0018419942 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0663_of_1380_text_document dclm +0.0019299515 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0949_of_1380_text_document dclm +0.0019178310 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0162_of_1380_text_document dclm +0.0018356134 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1209_of_1380_text_document dclm +0.0019312048 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0798_of_1380_text_document dclm +0.0018258568 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0710_of_1380_text_document dclm +0.0018399533 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0779_of_1380_text_document dclm +0.0018744345 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0669_of_1380_text_document dclm +0.0018714698 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1339_of_1380_text_document dclm +0.0018402046 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0273_of_1380_text_document dclm +0.0018690259 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1002_of_1380_text_document dclm +0.0018847487 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1170_of_1380_text_document dclm +0.0018862909 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0594_of_1380_text_document dclm +0.0019025441 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1362_of_1380_text_document dclm +0.0019048226 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0899_of_1380_text_document dclm +0.0018855320 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1118_of_1380_text_document dclm +0.0018571632 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0655_of_1380_text_document dclm +0.0018777523 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0740_of_1380_text_document dclm +0.0018563269 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0759_of_1380_text_document dclm +0.0018749821 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0161_of_1380_text_document dclm +0.0018737137 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0972_of_1380_text_document dclm +0.0018223511 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1277_of_1380_text_document dclm +0.0018959707 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0271_of_1380_text_document dclm +0.0018631523 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0664_of_1380_text_document dclm +0.0018903911 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1190_of_1380_text_document dclm +0.0018795364 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0800_of_1380_text_document dclm +0.0018495524 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0093_of_1380_text_document dclm +0.0018536896 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0399_of_1380_text_document dclm +0.0018852574 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0774_of_1380_text_document dclm +0.0018923357 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1074_of_1380_text_document dclm +0.0018677450 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0400_of_1380_text_document dclm +0.0018851258 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0346_of_1380_text_document dclm +0.0018657540 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0297_of_1380_text_document dclm +0.0018592106 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0095_of_1380_text_document dclm +0.0018371471 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0431_of_1380_text_document dclm +0.0018825750 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0561_of_1380_text_document dclm +0.0018942577 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0907_of_1380_text_document dclm +0.0018909400 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0620_of_1380_text_document dclm +0.0018524695 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0564_of_1380_text_document dclm +0.0018366115 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0181_of_1380_text_document dclm +0.0018658956 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0445_of_1380_text_document dclm +0.0018637280 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0943_of_1380_text_document dclm +0.0018970477 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0560_of_1380_text_document dclm +0.0018422840 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0106_of_1380_text_document dclm +0.0018856931 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0484_of_1380_text_document dclm +0.0019042016 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0165_of_1380_text_document dclm +0.0018824662 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0849_of_1380_text_document dclm +0.0018718026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0615_of_1380_text_document dclm +0.0018404168 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0751_of_1380_text_document dclm +0.0018955843 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0357_of_1380_text_document dclm +0.0018471907 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0897_of_1380_text_document dclm +0.0018880475 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0608_of_1380_text_document dclm +0.0018064113 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0171_of_1380_text_document dclm +0.0018639737 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0198_of_1380_text_document dclm +0.0018501609 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0105_of_1380_text_document dclm +0.0018406059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0090_of_1380_text_document dclm +0.0018670782 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0553_of_1380_text_document dclm +0.0018515311 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0601_of_1380_text_document dclm +0.0018740600 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0148_of_1380_text_document dclm +0.0018436897 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0913_of_1380_text_document dclm +0.0018810607 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1087_of_1380_text_document dclm +0.0018601461 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1207_of_1380_text_document dclm +0.0019126128 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1222_of_1380_text_document dclm +0.0018865197 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0347_of_1380_text_document dclm +0.0018545476 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0141_of_1380_text_document dclm +0.0018278502 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0802_of_1380_text_document dclm +0.0018790521 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0648_of_1380_text_document dclm +0.0019087638 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1306_of_1380_text_document dclm +0.0019086703 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1289_of_1380_text_document dclm +0.0018534333 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0279_of_1380_text_document dclm +0.0018818039 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0994_of_1380_text_document dclm +0.0018907756 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0203_of_1380_text_document dclm +0.0018608936 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0682_of_1380_text_document dclm +0.0018718580 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0458_of_1380_text_document dclm +0.0019088074 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1055_of_1380_text_document dclm +0.0018609412 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0854_of_1380_text_document dclm +0.0018038430 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0364_of_1380_text_document dclm +0.0018779187 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0852_of_1380_text_document dclm +0.0018505518 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1049_of_1380_text_document dclm +0.0018825681 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0292_of_1380_text_document dclm +0.0018881821 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0883_of_1380_text_document dclm +0.0018472746 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0597_of_1380_text_document dclm +0.0018767929 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1039_of_1380_text_document dclm +0.0019068838 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0550_of_1380_text_document dclm +0.0018862796 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1315_of_1380_text_document dclm +0.0019214525 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0687_of_1380_text_document dclm +0.0019226715 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0925_of_1380_text_document dclm +0.0018606431 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0158_of_1380_text_document dclm +0.0018327872 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0503_of_1380_text_document dclm +0.0018977414 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0062_of_1380_text_document dclm +0.0018820707 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1127_of_1380_text_document dclm +0.0018641175 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1088_of_1380_text_document dclm +0.0018767167 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0486_of_1380_text_document dclm +0.0018701187 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0805_of_1380_text_document dclm +0.0018505233 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0922_of_1380_text_document dclm +0.0018337531 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1197_of_1380_text_document dclm +0.0018126747 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1034_of_1380_text_document dclm +0.0018500296 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0962_of_1380_text_document dclm +0.0018505906 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0708_of_1380_text_document dclm +0.0018617534 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0698_of_1380_text_document dclm +0.0018345821 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1264_of_1380_text_document dclm +0.0019188646 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0980_of_1380_text_document dclm +0.0018932106 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0809_of_1380_text_document dclm +0.0018837969 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1148_of_1380_text_document dclm +0.0018539522 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1161_of_1380_text_document dclm +0.0018197167 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0320_of_1380_text_document dclm +0.0018690040 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0309_of_1380_text_document dclm +0.0018256024 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1178_of_1380_text_document dclm +0.0018480117 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1349_of_1380_text_document dclm +0.0018681399 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0520_of_1380_text_document dclm +0.0019021710 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0376_of_1380_text_document dclm +0.0018895428 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1023_of_1380_text_document dclm +0.0019047181 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0613_of_1380_text_document dclm +0.0018855437 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0282_of_1380_text_document dclm +0.0018964219 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1356_of_1380_text_document dclm +0.0018921764 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0435_of_1380_text_document dclm +0.0018316352 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0146_of_1380_text_document dclm +0.0018516763 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0277_of_1380_text_document dclm +0.0018652002 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0463_of_1380_text_document dclm +0.0018523233 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0163_of_1380_text_document dclm +0.0018743603 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0659_of_1380_text_document dclm +0.0018567925 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0864_of_1380_text_document dclm +0.0019080064 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0690_of_1380_text_document dclm +0.0018301721 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0208_of_1380_text_document dclm +0.0018249580 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1013_of_1380_text_document dclm +0.0018813171 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1016_of_1380_text_document dclm +0.0018551290 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0202_of_1380_text_document dclm +0.0018646581 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1230_of_1380_text_document dclm +0.0019080001 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1249_of_1380_text_document dclm +0.0018805160 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1344_of_1380_text_document dclm +0.0018719004 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0527_of_1380_text_document dclm +0.0018422914 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0143_of_1380_text_document dclm +0.0018576947 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0248_of_1380_text_document dclm +0.0018826346 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0813_of_1380_text_document dclm +0.0018202996 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0098_of_1380_text_document dclm +0.0018687034 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0928_of_1380_text_document dclm +0.0018283517 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0455_of_1380_text_document dclm +0.0018666737 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0187_of_1380_text_document dclm +0.0018458003 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0101_of_1380_text_document dclm +0.0018784618 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0160_of_1380_text_document dclm +0.0019165496 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0591_of_1380_text_document dclm +0.0019150472 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0580_of_1380_text_document dclm +0.0019096007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0338_of_1380_text_document dclm +0.0018951413 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0619_of_1380_text_document dclm +0.0018636033 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0532_of_1380_text_document dclm +0.0018801762 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0091_of_1380_text_document dclm +0.0018595926 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0341_of_1380_text_document dclm +0.0018285462 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1361_of_1380_text_document dclm +0.0018661147 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0643_of_1380_text_document dclm +0.0018321209 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0871_of_1380_text_document dclm +0.0018238375 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0507_of_1380_text_document dclm +0.0018473142 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0579_of_1380_text_document dclm +0.0018929779 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1124_of_1380_text_document dclm +0.0018922866 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1256_of_1380_text_document dclm +0.0019020674 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0351_of_1380_text_document dclm +0.0018574034 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1123_of_1380_text_document dclm +0.0019023583 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1274_of_1380_text_document dclm +0.0018658007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0582_of_1380_text_document dclm +0.0019263296 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1363_of_1380_text_document dclm +0.0018838421 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0253_of_1380_text_document dclm +0.0019077058 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0794_of_1380_text_document dclm +0.0018677016 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0937_of_1380_text_document dclm +0.0019069591 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1325_of_1380_text_document dclm +0.0018757994 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0976_of_1380_text_document dclm +0.0018876071 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0836_of_1380_text_document dclm +0.0019093391 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0387_of_1380_text_document dclm +0.0018224155 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0319_of_1380_text_document dclm +0.0018011020 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0337_of_1380_text_document dclm +0.0018788858 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0249_of_1380_text_document dclm +0.0019211780 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0284_of_1380_text_document dclm +0.0018378218 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1316_of_1380_text_document dclm +0.0018793777 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0868_of_1380_text_document dclm +0.0018629749 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0481_of_1380_text_document dclm +0.0018727372 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1237_of_1380_text_document dclm +0.0018654836 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1128_of_1380_text_document dclm +0.0018548777 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0159_of_1380_text_document dclm +0.0018802148 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0577_of_1380_text_document dclm +0.0018415794 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0305_of_1380_text_document dclm +0.0018377748 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0407_of_1380_text_document dclm +0.0018932316 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0521_of_1380_text_document dclm +0.0018911868 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1134_of_1380_text_document dclm +0.0018601946 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0153_of_1380_text_document dclm +0.0018897654 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0052_of_1380_text_document dclm +0.0019276232 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1113_of_1380_text_document dclm +0.0018543013 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0957_of_1380_text_document dclm +0.0018555126 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0350_of_1380_text_document dclm +0.0018464948 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0877_of_1380_text_document dclm +0.0018784567 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0979_of_1380_text_document dclm +0.0019212272 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0890_of_1380_text_document dclm +0.0018642543 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0915_of_1380_text_document dclm +0.0018558482 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0567_of_1380_text_document dclm +0.0018940979 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0167_of_1380_text_document dclm +0.0019153170 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0696_of_1380_text_document dclm +0.0018768687 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0556_of_1380_text_document dclm +0.0018919257 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0186_of_1380_text_document dclm +0.0018648516 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1372_of_1380_text_document dclm +0.0018452263 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1069_of_1380_text_document dclm +0.0018684198 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1174_of_1380_text_document dclm +0.0018715785 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0089_of_1380_text_document dclm +0.0018522881 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0628_of_1380_text_document dclm +0.0019094008 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1255_of_1380_text_document dclm +0.0018534555 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0986_of_1380_text_document dclm +0.0018895471 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0272_of_1380_text_document dclm +0.0018354636 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0745_of_1380_text_document dclm +0.0018805944 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0817_of_1380_text_document dclm +0.0018905448 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0318_of_1380_text_document dclm +0.0018725921 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0169_of_1380_text_document dclm +0.0018584122 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1369_of_1380_text_document dclm +0.0018651832 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0891_of_1380_text_document dclm +0.0018413498 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1323_of_1380_text_document dclm +0.0018522595 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0500_of_1380_text_document dclm +0.0018844116 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0298_of_1380_text_document dclm +0.0018423878 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0343_of_1380_text_document dclm +0.0018416312 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0036_of_1380_text_document dclm +0.0018998049 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0437_of_1380_text_document dclm +0.0019015746 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0617_of_1380_text_document dclm +0.0018470154 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1077_of_1380_text_document dclm +0.0018599366 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1046_of_1380_text_document dclm +0.0018664504 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0633_of_1380_text_document dclm +0.0018885638 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0624_of_1380_text_document dclm +0.0018454824 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0984_of_1380_text_document dclm +0.0018818185 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0395_of_1380_text_document dclm +0.0018313168 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0219_of_1380_text_document dclm +0.0018218889 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1291_of_1380_text_document dclm +0.0018999432 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0022_of_1380_text_document dclm +0.0019185357 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1319_of_1380_text_document dclm +0.0018569415 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0544_of_1380_text_document dclm +0.0019293328 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0408_of_1380_text_document dclm +0.0018673488 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0108_of_1380_text_document dclm +0.0018792552 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1261_of_1380_text_document dclm +0.0017608926 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0261_of_1380_text_document dclm +0.0019094092 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0677_of_1380_text_document dclm +0.0018957909 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0904_of_1380_text_document dclm +0.0018309363 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1070_of_1380_text_document dclm +0.0018999699 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0280_of_1380_text_document dclm +0.0018319594 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1116_of_1380_text_document dclm +0.0018659322 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0044_of_1380_text_document dclm +0.0018272376 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0709_of_1380_text_document dclm +0.0019090018 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0780_of_1380_text_document dclm +0.0018782011 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1015_of_1380_text_document dclm +0.0018343919 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1022_of_1380_text_document dclm +0.0018620814 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0356_of_1380_text_document dclm +0.0018239137 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0312_of_1380_text_document dclm +0.0018059976 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0423_of_1380_text_document dclm +0.0018659400 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0480_of_1380_text_document dclm +0.0018811025 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1320_of_1380_text_document dclm +0.0018734525 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0291_of_1380_text_document dclm +0.0018705430 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1092_of_1380_text_document dclm +0.0019021394 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0685_of_1380_text_document dclm +0.0018662783 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0879_of_1380_text_document dclm +0.0018731764 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0768_of_1380_text_document dclm +0.0018381929 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0459_of_1380_text_document dclm +0.0018418093 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1156_of_1380_text_document dclm +0.0018819648 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0323_of_1380_text_document dclm +0.0018709593 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0853_of_1380_text_document dclm +0.0019046035 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0138_of_1380_text_document dclm +0.0018207018 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0001_of_1380_text_document dclm +0.0018316259 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0958_of_1380_text_document dclm +0.0018685142 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1109_of_1380_text_document dclm +0.0019281788 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1168_of_1380_text_document dclm +0.0018718903 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0316_of_1380_text_document dclm +0.0018933502 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1200_of_1380_text_document dclm +0.0018703207 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0646_of_1380_text_document dclm +0.0018869175 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0906_of_1380_text_document dclm +0.0019297890 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0797_of_1380_text_document dclm +0.0019086814 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0639_of_1380_text_document dclm +0.0019054407 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0200_of_1380_text_document dclm +0.0018253523 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0293_of_1380_text_document dclm +0.0019325438 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1047_of_1380_text_document dclm +0.0018535922 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0827_of_1380_text_document dclm +0.0018795320 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1194_of_1380_text_document dclm +0.0018664075 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0464_of_1380_text_document dclm +0.0018638728 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0353_of_1380_text_document dclm +0.0018938379 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1308_of_1380_text_document dclm +0.0018581932 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1009_of_1380_text_document dclm +0.0019003641 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0422_of_1380_text_document dclm +0.0018672961 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0061_of_1380_text_document dclm +0.0018721213 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0355_of_1380_text_document dclm +0.0018897893 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0191_of_1380_text_document dclm +0.0019072332 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0572_of_1380_text_document dclm +0.0018717260 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0641_of_1380_text_document dclm +0.0018893794 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1303_of_1380_text_document dclm +0.0018788473 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1305_of_1380_text_document dclm +0.0018263065 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0226_of_1380_text_document dclm +0.0018719328 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1343_of_1380_text_document dclm +0.0018706147 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0571_of_1380_text_document dclm +0.0018575036 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1218_of_1380_text_document dclm +0.0018117550 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0713_of_1380_text_document dclm +0.0018461351 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0881_of_1380_text_document dclm +0.0019151428 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0419_of_1380_text_document dclm +0.0018963694 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0845_of_1380_text_document dclm +0.0018953954 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0838_of_1380_text_document dclm +0.0018710534 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0742_of_1380_text_document dclm +0.0018381884 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0790_of_1380_text_document dclm +0.0018460766 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0605_of_1380_text_document dclm +0.0018723902 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0487_of_1380_text_document dclm +0.0018503678 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0396_of_1380_text_document dclm +0.0018922701 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0770_of_1380_text_document dclm +0.0018646071 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0361_of_1380_text_document dclm +0.0018700677 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0094_of_1380_text_document dclm +0.0018446412 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0629_of_1380_text_document dclm +0.0018318657 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1204_of_1380_text_document dclm +0.0018352471 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0728_of_1380_text_document dclm +0.0018594037 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0548_of_1380_text_document dclm +0.0018919110 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0988_of_1380_text_document dclm +0.0018139825 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0555_of_1380_text_document dclm +0.0018434120 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0221_of_1380_text_document dclm +0.0019119555 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0675_of_1380_text_document dclm +0.0018293943 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1028_of_1380_text_document dclm +0.0018296420 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0513_of_1380_text_document dclm +0.0018766019 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0652_of_1380_text_document dclm +0.0018414808 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1001_of_1380_text_document dclm +0.0018544155 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0240_of_1380_text_document dclm +0.0018448547 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0088_of_1380_text_document dclm +0.0019223865 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1283_of_1380_text_document dclm +0.0019081765 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1165_of_1380_text_document dclm +0.0018333250 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0269_of_1380_text_document dclm +0.0018740564 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0121_of_1380_text_document dclm +0.0018446218 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1234_of_1380_text_document dclm +0.0018794959 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0182_of_1380_text_document dclm +0.0018745029 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1232_of_1380_text_document dclm +0.0018699782 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0517_of_1380_text_document dclm +0.0019191539 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1279_of_1380_text_document dclm +0.0018386079 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0496_of_1380_text_document dclm +0.0018387638 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1020_of_1380_text_document dclm +0.0018373025 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0225_of_1380_text_document dclm +0.0019115299 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0296_of_1380_text_document dclm +0.0019236015 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0950_of_1380_text_document dclm +0.0018362277 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0183_of_1380_text_document dclm +0.0018382665 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1151_of_1380_text_document dclm +0.0018467086 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1314_of_1380_text_document dclm +0.0019108435 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0781_of_1380_text_document dclm +0.0018309001 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0645_of_1380_text_document dclm +0.0018464568 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0019_of_1380_text_document dclm +0.0018659468 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1373_of_1380_text_document dclm +0.0018564427 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0348_of_1380_text_document dclm +0.0018290265 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0888_of_1380_text_document dclm +0.0018887476 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0733_of_1380_text_document dclm +0.0018908862 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0074_of_1380_text_document dclm +0.0018989668 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0970_of_1380_text_document dclm +0.0018731163 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1221_of_1380_text_document dclm +0.0018726913 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0494_of_1380_text_document dclm +0.0018762932 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1141_of_1380_text_document dclm +0.0018616122 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0482_of_1380_text_document dclm +0.0018384335 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0546_of_1380_text_document dclm +0.0018721451 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0806_of_1380_text_document dclm +0.0018336444 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0326_of_1380_text_document dclm +0.0018022324 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1041_of_1380_text_document dclm +0.0018892364 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0386_of_1380_text_document dclm +0.0018068131 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0027_of_1380_text_document dclm +0.0018059579 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0666_of_1380_text_document dclm +0.0018788448 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0612_of_1380_text_document dclm +0.0018875379 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0746_of_1380_text_document dclm +0.0018070240 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0136_of_1380_text_document dclm +0.0019029995 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0409_of_1380_text_document dclm +0.0018992707 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1278_of_1380_text_document dclm +0.0018725762 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1018_of_1380_text_document dclm +0.0018195643 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0179_of_1380_text_document dclm +0.0018575007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0429_of_1380_text_document dclm +0.0018636686 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1112_of_1380_text_document dclm +0.0018861857 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0636_of_1380_text_document dclm +0.0018586092 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0540_of_1380_text_document dclm +0.0019173146 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1164_of_1380_text_document dclm +0.0018523085 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1024_of_1380_text_document dclm +0.0018651616 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0598_of_1380_text_document dclm +0.0018472634 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1188_of_1380_text_document dclm +0.0018293119 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0365_of_1380_text_document dclm +0.0018731080 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1219_of_1380_text_document dclm +0.0018717051 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0758_of_1380_text_document dclm +0.0019042224 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0428_of_1380_text_document dclm +0.0018791471 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0968_of_1380_text_document dclm +0.0018706125 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1295_of_1380_text_document dclm +0.0019271516 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0552_of_1380_text_document dclm +0.0018768360 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0454_of_1380_text_document dclm +0.0019057967 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1370_of_1380_text_document dclm +0.0018982943 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0875_of_1380_text_document dclm +0.0018702588 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0456_of_1380_text_document dclm +0.0018730837 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0339_of_1380_text_document dclm +0.0019032527 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0132_of_1380_text_document dclm +0.0018687659 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0896_of_1380_text_document dclm +0.0018468720 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0420_of_1380_text_document dclm +0.0018363290 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0015_of_1380_text_document dclm +0.0018724510 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0383_of_1380_text_document dclm +0.0018765133 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1350_of_1380_text_document dclm +0.0018115212 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1068_of_1380_text_document dclm +0.0018322939 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0587_of_1380_text_document dclm +0.0018320414 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0145_of_1380_text_document dclm +0.0018701311 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0804_of_1380_text_document dclm +0.0018346909 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0446_of_1380_text_document dclm +0.0018959318 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1149_of_1380_text_document dclm +0.0018327166 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0956_of_1380_text_document dclm +0.0018820050 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1119_of_1380_text_document dclm +0.0019257100 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0135_of_1380_text_document dclm +0.0018291171 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0914_of_1380_text_document dclm +0.0018545792 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0788_of_1380_text_document dclm +0.0018306372 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0989_of_1380_text_document dclm +0.0018728107 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1251_of_1380_text_document dclm +0.0018453362 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0335_of_1380_text_document dclm +0.0018835295 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0144_of_1380_text_document dclm +0.0018326795 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0287_of_1380_text_document dclm +0.0018952474 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0278_of_1380_text_document dclm +0.0018521983 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0522_of_1380_text_document dclm +0.0018435895 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0531_of_1380_text_document dclm +0.0018293725 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0668_of_1380_text_document dclm +0.0019058004 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0152_of_1380_text_document dclm +0.0018597526 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0516_of_1380_text_document dclm +0.0018145097 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0569_of_1380_text_document dclm +0.0018574845 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1202_of_1380_text_document dclm +0.0018855389 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0919_of_1380_text_document dclm +0.0018196574 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0985_of_1380_text_document dclm +0.0018754655 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0125_of_1380_text_document dclm +0.0018526184 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1154_of_1380_text_document dclm +0.0018227247 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0947_of_1380_text_document dclm +0.0019110234 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0216_of_1380_text_document dclm +0.0019028642 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1019_of_1380_text_document dclm +0.0018307843 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0010_of_1380_text_document dclm +0.0018761834 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0196_of_1380_text_document dclm +0.0019017444 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0799_of_1380_text_document dclm +0.0018921847 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0266_of_1380_text_document dclm +0.0018524751 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0604_of_1380_text_document dclm +0.0018873638 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0403_of_1380_text_document dclm +0.0019336533 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0878_of_1380_text_document dclm +0.0018969305 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1027_of_1380_text_document dclm +0.0018454210 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1075_of_1380_text_document dclm +0.0018277482 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1353_of_1380_text_document dclm +0.0019110484 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1276_of_1380_text_document dclm +0.0019130451 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0307_of_1380_text_document dclm +0.0018646888 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0873_of_1380_text_document dclm +0.0018567673 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0545_of_1380_text_document dclm +0.0018662453 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1031_of_1380_text_document dclm +0.0018656394 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1172_of_1380_text_document dclm +0.0018649889 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0228_of_1380_text_document dclm +0.0018662391 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0650_of_1380_text_document dclm +0.0018269771 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1227_of_1380_text_document dclm +0.0018400189 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0705_of_1380_text_document dclm +0.0018380056 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0584_of_1380_text_document dclm +0.0018586877 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1040_of_1380_text_document dclm +0.0018307351 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0440_of_1380_text_document dclm +0.0018436218 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0469_of_1380_text_document dclm +0.0018494088 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0634_of_1380_text_document dclm +0.0018663997 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0692_of_1380_text_document dclm +0.0018553709 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0578_of_1380_text_document dclm +0.0018538059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1243_of_1380_text_document dclm +0.0018697949 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1120_of_1380_text_document dclm +0.0019140026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0537_of_1380_text_document dclm +0.0018495141 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1285_of_1380_text_document dclm +0.0019233718 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0718_of_1380_text_document dclm +0.0018355907 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0903_of_1380_text_document dclm +0.0018316432 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1347_of_1380_text_document dclm +0.0019007437 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1084_of_1380_text_document dclm +0.0018480700 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0918_of_1380_text_document dclm +0.0018536385 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1241_of_1380_text_document dclm +0.0018639914 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0258_of_1380_text_document dclm +0.0018918419 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0933_of_1380_text_document dclm +0.0018128165 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0439_of_1380_text_document dclm +0.0019036799 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0562_of_1380_text_document dclm +0.0018682514 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1284_of_1380_text_document dclm +0.0018649201 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0192_of_1380_text_document dclm +0.0018341436 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0209_of_1380_text_document dclm +0.0018441648 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0119_of_1380_text_document dclm +0.0019171286 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0965_of_1380_text_document dclm +0.0018700420 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0245_of_1380_text_document dclm +0.0019162796 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1044_of_1380_text_document dclm +0.0018376284 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1089_of_1380_text_document dclm +0.0018205568 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0072_of_1380_text_document dclm +0.0018682892 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0750_of_1380_text_document dclm +0.0018344781 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0704_of_1380_text_document dclm +0.0018464707 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0867_of_1380_text_document dclm +0.0018128792 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0067_of_1380_text_document dclm +0.0018341206 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0637_of_1380_text_document dclm +0.0019213774 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0835_of_1380_text_document dclm +0.0019163265 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0589_of_1380_text_document dclm +0.0018789254 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0255_of_1380_text_document dclm +0.0018295082 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0457_of_1380_text_document dclm +0.0019007284 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0684_of_1380_text_document dclm +0.0018145506 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0676_of_1380_text_document dclm +0.0018787394 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1364_of_1380_text_document dclm +0.0018961785 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1058_of_1380_text_document dclm +0.0018739282 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0978_of_1380_text_document dclm +0.0018843646 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0259_of_1380_text_document dclm +0.0018706314 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0083_of_1380_text_document dclm +0.0019030468 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1298_of_1380_text_document dclm +0.0018728270 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1231_of_1380_text_document dclm +0.0018217090 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0691_of_1380_text_document dclm +0.0018916742 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0188_of_1380_text_document dclm +0.0018652824 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1376_of_1380_text_document dclm +0.0018627009 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0695_of_1380_text_document dclm +0.0018687061 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0345_of_1380_text_document dclm +0.0018533328 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0730_of_1380_text_document dclm +0.0019004980 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0725_of_1380_text_document dclm +0.0018622517 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0025_of_1380_text_document dclm +0.0018693479 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1177_of_1380_text_document dclm +0.0018304992 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0719_of_1380_text_document dclm +0.0018413407 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0801_of_1380_text_document dclm +0.0018956016 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1179_of_1380_text_document dclm +0.0018875657 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1082_of_1380_text_document dclm +0.0018760859 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0370_of_1380_text_document dclm +0.0018528936 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0483_of_1380_text_document dclm +0.0018902015 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1038_of_1380_text_document dclm +0.0018992719 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1210_of_1380_text_document dclm +0.0018759828 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0268_of_1380_text_document dclm +0.0018957802 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1191_of_1380_text_document dclm +0.0019038385 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0505_of_1380_text_document dclm +0.0019127512 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1061_of_1380_text_document dclm +0.0018618835 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0717_of_1380_text_document dclm +0.0018380410 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0003_of_1380_text_document dclm +0.0019240799 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0741_of_1380_text_document dclm +0.0018973885 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1106_of_1380_text_document dclm +0.0018298568 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0791_of_1380_text_document dclm +0.0018280632 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0674_of_1380_text_document dclm +0.0018523376 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1162_of_1380_text_document dclm +0.0018935350 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1367_of_1380_text_document dclm +0.0018699494 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0783_of_1380_text_document dclm +0.0018406304 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1053_of_1380_text_document dclm +0.0018462301 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0855_of_1380_text_document dclm +0.0018628542 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0557_of_1380_text_document dclm +0.0018558770 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0071_of_1380_text_document dclm +0.0019083828 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0285_of_1380_text_document dclm +0.0018779468 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1351_of_1380_text_document dclm +0.0018862027 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1258_of_1380_text_document dclm +0.0018473214 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0815_of_1380_text_document dclm +0.0018340737 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0333_of_1380_text_document dclm +0.0018951646 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0649_of_1380_text_document dclm +0.0018489962 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0538_of_1380_text_document dclm +0.0018601429 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0214_of_1380_text_document dclm +0.0018791392 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0524_of_1380_text_document dclm +0.0018882402 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0223_of_1380_text_document dclm +0.0019133154 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1240_of_1380_text_document dclm +0.0019106616 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0311_of_1380_text_document dclm +0.0018907078 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0917_of_1380_text_document dclm +0.0018709038 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1063_of_1380_text_document dclm +0.0018672721 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1268_of_1380_text_document dclm +0.0018452050 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0176_of_1380_text_document dclm +0.0019023331 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0626_of_1380_text_document dclm +0.0018317003 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1125_of_1380_text_document dclm +0.0018189946 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0103_of_1380_text_document dclm +0.0018143876 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1158_of_1380_text_document dclm +0.0018802984 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0434_of_1380_text_document dclm +0.0019031458 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0441_of_1380_text_document dclm +0.0018101783 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1083_of_1380_text_document dclm +0.0018730260 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1273_of_1380_text_document dclm +0.0018800628 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1007_of_1380_text_document dclm +0.0018043363 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1265_of_1380_text_document dclm +0.0018559722 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0938_of_1380_text_document dclm +0.0018863892 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0220_of_1380_text_document dclm +0.0019165899 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0635_of_1380_text_document dclm +0.0018666864 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0175_of_1380_text_document dclm +0.0018543807 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0693_of_1380_text_document dclm +0.0018573839 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1052_of_1380_text_document dclm +0.0018465772 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0769_of_1380_text_document dclm +0.0018344368 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1114_of_1380_text_document dclm +0.0018954131 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1029_of_1380_text_document dclm +0.0019024576 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1004_of_1380_text_document dclm +0.0019065884 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0640_of_1380_text_document dclm +0.0018558021 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0352_of_1380_text_document dclm +0.0018752061 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0485_of_1380_text_document dclm +0.0018435877 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0155_of_1380_text_document dclm +0.0018677608 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0983_of_1380_text_document dclm +0.0018477022 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0789_of_1380_text_document dclm +0.0019131420 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0359_of_1380_text_document dclm +0.0018529167 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0034_of_1380_text_document dclm +0.0018418107 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0134_of_1380_text_document dclm +0.0018396155 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0905_of_1380_text_document dclm +0.0018380837 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0414_of_1380_text_document dclm +0.0018640920 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0040_of_1380_text_document dclm +0.0018261607 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0941_of_1380_text_document dclm +0.0018857729 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1233_of_1380_text_document dclm +0.0018316924 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0964_of_1380_text_document dclm +0.0018420481 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0041_of_1380_text_document dclm +0.0018601815 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1135_of_1380_text_document dclm +0.0018700967 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0996_of_1380_text_document dclm +0.0018627433 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1012_of_1380_text_document dclm +0.0018009095 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1142_of_1380_text_document dclm +0.0018334964 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0795_of_1380_text_document dclm +0.0018562517 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1271_of_1380_text_document dclm +0.0018952945 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0325_of_1380_text_document dclm +0.0018957471 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1146_of_1380_text_document dclm +0.0018903193 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0294_of_1380_text_document dclm +0.0018986124 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0832_of_1380_text_document dclm +0.0019126951 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0068_of_1380_text_document dclm +0.0019127986 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1250_of_1380_text_document dclm +0.0018696966 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1205_of_1380_text_document dclm +0.0018881237 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0020_of_1380_text_document dclm +0.0017916565 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1062_of_1380_text_document dclm +0.0018345644 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0535_of_1380_text_document dclm +0.0018888820 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0775_of_1380_text_document dclm +0.0018799015 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0473_of_1380_text_document dclm +0.0018454800 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0934_of_1380_text_document dclm +0.0018375825 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0157_of_1380_text_document dclm +0.0018717515 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0884_of_1380_text_document dclm +0.0018815580 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0368_of_1380_text_document dclm +0.0019132873 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0417_of_1380_text_document dclm +0.0018934204 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0585_of_1380_text_document dclm +0.0019136916 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0450_of_1380_text_document dclm +0.0018699063 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0955_of_1380_text_document dclm +0.0018369088 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0239_of_1380_text_document dclm +0.0018287061 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1297_of_1380_text_document dclm +0.0018853511 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1317_of_1380_text_document dclm +0.0018578264 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1140_of_1380_text_document dclm +0.0018424915 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0614_of_1380_text_document dclm +0.0018394723 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1144_of_1380_text_document dclm +0.0018183789 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0184_of_1380_text_document dclm +0.0018618538 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0981_of_1380_text_document dclm +0.0018667756 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1150_of_1380_text_document dclm +0.0019030219 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0017_of_1380_text_document dclm +0.0018259681 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0654_of_1380_text_document dclm +0.0018739132 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0595_of_1380_text_document dclm +0.0018222264 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0004_of_1380_text_document dclm +0.0018736658 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1359_of_1380_text_document dclm +0.0018530615 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0974_of_1380_text_document dclm +0.0018521730 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0714_of_1380_text_document dclm +0.0018611794 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0283_of_1380_text_document dclm +0.0018944609 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0973_of_1380_text_document dclm +0.0018578283 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1378_of_1380_text_document dclm +0.0018029391 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0295_of_1380_text_document dclm +0.0018312668 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1368_of_1380_text_document dclm +0.0018807775 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0678_of_1380_text_document dclm +0.0018636517 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0807_of_1380_text_document dclm +0.0019068081 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1310_of_1380_text_document dclm +0.0019258320 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0936_of_1380_text_document dclm +0.0018522632 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1220_of_1380_text_document dclm +0.0018690306 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0702_of_1380_text_document dclm +0.0018727699 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0910_of_1380_text_document dclm +0.0018761633 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0765_of_1380_text_document dclm +0.0018557141 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0865_of_1380_text_document dclm +0.0018426367 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0661_of_1380_text_document dclm +0.0018212726 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0716_of_1380_text_document dclm +0.0018648054 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0172_of_1380_text_document dclm +0.0018635705 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0499_of_1380_text_document dclm +0.0018975485 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0662_of_1380_text_document dclm +0.0018864033 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1014_of_1380_text_document dclm +0.0018194497 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0722_of_1380_text_document dclm +0.0018362078 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1182_of_1380_text_document dclm +0.0018441557 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0830_of_1380_text_document dclm +0.0018994381 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0055_of_1380_text_document dclm +0.0018714778 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0366_of_1380_text_document dclm +0.0018945506 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0432_of_1380_text_document dclm +0.0019243890 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0012_of_1380_text_document dclm +0.0018800951 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1045_of_1380_text_document dclm +0.0019435971 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1311_of_1380_text_document dclm +0.0019086007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1201_of_1380_text_document dclm +0.0019204008 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1270_of_1380_text_document dclm +0.0018784517 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1101_of_1380_text_document dclm +0.0018350019 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0097_of_1380_text_document dclm +0.0018868562 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0660_of_1380_text_document dclm +0.0018664000 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0281_of_1380_text_document dclm +0.0018480388 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0843_of_1380_text_document dclm +0.0018700870 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1010_of_1380_text_document dclm +0.0019040717 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0398_of_1380_text_document dclm +0.0018281358 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0573_of_1380_text_document dclm +0.0018726366 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0926_of_1380_text_document dclm +0.0018866492 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0302_of_1380_text_document dclm +0.0018704301 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0987_of_1380_text_document dclm +0.0018360013 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1340_of_1380_text_document dclm +0.0019166229 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0230_of_1380_text_document dclm +0.0018513733 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0576_of_1380_text_document dclm +0.0018881498 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0140_of_1380_text_document dclm +0.0018998363 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0715_of_1380_text_document dclm +0.0018288808 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1287_of_1380_text_document dclm +0.0018881156 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0267_of_1380_text_document dclm +0.0018536323 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0009_of_1380_text_document dclm +0.0018629059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0023_of_1380_text_document dclm +0.0018432812 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1286_of_1380_text_document dclm +0.0018400956 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0018_of_1380_text_document dclm +0.0018551258 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1099_of_1380_text_document dclm +0.0018292019 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1122_of_1380_text_document dclm +0.0019052066 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0549_of_1380_text_document dclm +0.0018124314 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1183_of_1380_text_document dclm +0.0019203051 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1059_of_1380_text_document dclm +0.0018611505 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0139_of_1380_text_document dclm +0.0018228522 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1275_of_1380_text_document dclm +0.0018927100 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0189_of_1380_text_document dclm +0.0018809545 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0607_of_1380_text_document dclm +0.0019163395 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0656_of_1380_text_document dclm +0.0018001775 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0882_of_1380_text_document dclm +0.0018499106 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1048_of_1380_text_document dclm +0.0018298680 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0252_of_1380_text_document dclm +0.0018626866 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0724_of_1380_text_document dclm +0.0019173997 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0993_of_1380_text_document dclm +0.0019082887 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1110_of_1380_text_document dclm +0.0018842003 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0057_of_1380_text_document dclm +0.0018275759 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0317_of_1380_text_document dclm +0.0019105271 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0858_of_1380_text_document dclm +0.0018689440 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0902_of_1380_text_document dclm +0.0019097096 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0384_of_1380_text_document dclm +0.0018894948 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0766_of_1380_text_document dclm +0.0018722010 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1181_of_1380_text_document dclm +0.0018393599 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1217_of_1380_text_document dclm +0.0019215106 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1008_of_1380_text_document dclm +0.0018856141 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1021_of_1380_text_document dclm +0.0018344503 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1081_of_1380_text_document dclm +0.0018452596 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1064_of_1380_text_document dclm +0.0018991159 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1117_of_1380_text_document dclm +0.0018409474 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0107_of_1380_text_document dclm +0.0018528107 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0940_of_1380_text_document dclm +0.0019104344 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0911_of_1380_text_document dclm +0.0018508138 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0829_of_1380_text_document dclm +0.0018938719 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0870_of_1380_text_document dclm +0.0019055884 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1003_of_1380_text_document dclm +0.0018262329 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0778_of_1380_text_document dclm +0.0019044814 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0630_of_1380_text_document dclm +0.0018780267 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0442_of_1380_text_document dclm +0.0018556998 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0508_of_1380_text_document dclm +0.0018416994 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0599_of_1380_text_document dclm +0.0018357545 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0927_of_1380_text_document dclm +0.0018289552 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0803_of_1380_text_document dclm +0.0018643323 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1159_of_1380_text_document dclm +0.0019093068 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0013_of_1380_text_document dclm +0.0018680102 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0647_of_1380_text_document dclm +0.0018321037 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0472_of_1380_text_document dclm +0.0018372995 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0002_of_1380_text_document dclm +0.0018486256 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0122_of_1380_text_document dclm +0.0018753150 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0026_of_1380_text_document dclm +0.0018463011 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0058_of_1380_text_document dclm +0.0018159565 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0606_of_1380_text_document dclm +0.0018920931 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0991_of_1380_text_document dclm +0.0018238402 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1193_of_1380_text_document dclm +0.0019041197 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0839_of_1380_text_document dclm +0.0018366312 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0322_of_1380_text_document dclm +0.0018964137 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0047_of_1380_text_document dclm +0.0018767660 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0451_of_1380_text_document dclm +0.0018774793 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0462_of_1380_text_document dclm +0.0019210375 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0672_of_1380_text_document dclm +0.0018335542 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1302_of_1380_text_document dclm +0.0018389962 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1133_of_1380_text_document dclm +0.0018160760 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0706_of_1380_text_document dclm +0.0019033018 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0960_of_1380_text_document dclm +0.0018342059 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0461_of_1380_text_document dclm +0.0018564667 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0554_of_1380_text_document dclm +0.0018739383 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0744_of_1380_text_document dclm +0.0018696068 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0542_of_1380_text_document dclm +0.0018329030 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0506_of_1380_text_document dclm +0.0018842180 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0847_of_1380_text_document dclm +0.0018560068 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0818_of_1380_text_document dclm +0.0018828525 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0375_of_1380_text_document dclm +0.0018164048 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0627_of_1380_text_document dclm +0.0018812121 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1011_of_1380_text_document dclm +0.0018451559 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0246_of_1380_text_document dclm +0.0018439416 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0588_of_1380_text_document dclm +0.0018702442 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0543_of_1380_text_document dclm +0.0018854307 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0631_of_1380_text_document dclm +0.0018731889 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0593_of_1380_text_document dclm +0.0018913032 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0329_of_1380_text_document dclm +0.0019165026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1139_of_1380_text_document dclm +0.0018767340 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1126_of_1380_text_document dclm +0.0018513167 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0749_of_1380_text_document dclm +0.0013288917 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1380_of_1380_text_document dclm +0.0019268002 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1030_of_1380_text_document dclm +0.0018655835 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0288_of_1380_text_document dclm +0.0019331632 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0069_of_1380_text_document dclm +0.0018665087 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1147_of_1380_text_document dclm +0.0018897743 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1371_of_1380_text_document dclm +0.0018568955 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0823_of_1380_text_document dclm +0.0018608965 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1060_of_1380_text_document dclm +0.0018052057 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1100_of_1380_text_document dclm +0.0019168952 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1224_of_1380_text_document dclm +0.0018969237 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0861_of_1380_text_document dclm +0.0018507702 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0753_of_1380_text_document dclm +0.0018363308 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1080_of_1380_text_document dclm +0.0018710553 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0150_of_1380_text_document dclm +0.0019180508 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0449_of_1380_text_document dclm +0.0018590105 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0916_of_1380_text_document dclm +0.0018829422 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1262_of_1380_text_document dclm +0.0018299199 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0951_of_1380_text_document dclm +0.0018592116 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0336_of_1380_text_document dclm +0.0018753699 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0227_of_1380_text_document dclm +0.0018461572 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0168_of_1380_text_document dclm +0.0018626886 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1098_of_1380_text_document dclm +0.0018880948 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0075_of_1380_text_document dclm +0.0018513032 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1071_of_1380_text_document dclm +0.0019101075 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1035_of_1380_text_document dclm +0.0018488416 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1300_of_1380_text_document dclm +0.0018387096 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0844_of_1380_text_document dclm +0.0018411846 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0251_of_1380_text_document dclm +0.0018560705 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0688_of_1380_text_document dclm +0.0018966691 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1304_of_1380_text_document dclm +0.0018517051 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1225_of_1380_text_document dclm +0.0019122017 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0178_of_1380_text_document dclm +0.0018400110 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0043_of_1380_text_document dclm +0.0018445836 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1330_of_1380_text_document dclm +0.0018305015 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0443_of_1380_text_document dclm +0.0018504420 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0518_of_1380_text_document dclm +0.0018718299 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0501_of_1380_text_document dclm +0.0018689653 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0493_of_1380_text_document dclm +0.0018512911 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0033_of_1380_text_document dclm +0.0018523347 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0819_of_1380_text_document dclm +0.0018361083 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0999_of_1380_text_document dclm +0.0019189143 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1073_of_1380_text_document dclm +0.0018254051 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1299_of_1380_text_document dclm +0.0018493121 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1145_of_1380_text_document dclm +0.0018933666 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0786_of_1380_text_document dclm +0.0018562733 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0701_of_1380_text_document dclm +0.0019099767 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0465_of_1380_text_document dclm +0.0018343736 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1097_of_1380_text_document dclm +0.0018478744 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0842_of_1380_text_document dclm +0.0018712990 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0866_of_1380_text_document dclm +0.0019012623 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0096_of_1380_text_document dclm +0.0018564848 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0080_of_1380_text_document dclm +0.0018480257 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0644_of_1380_text_document dclm +0.0018974671 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1086_of_1380_text_document dclm +0.0019069569 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1166_of_1380_text_document dclm +0.0018763692 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1137_of_1380_text_document dclm +0.0018294833 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0377_of_1380_text_document dclm +0.0018053846 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0689_of_1380_text_document dclm +0.0018741190 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1157_of_1380_text_document dclm +0.0018972744 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0700_of_1380_text_document dclm +0.0018822438 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0583_of_1380_text_document dclm +0.0018986565 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0324_of_1380_text_document dclm +0.0019118250 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0784_of_1380_text_document dclm +0.0018623439 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0539_of_1380_text_document dclm +0.0018409868 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1078_of_1380_text_document dclm +0.0018239198 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1131_of_1380_text_document dclm +0.0018390830 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0114_of_1380_text_document dclm +0.0019021054 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1360_of_1380_text_document dclm +0.0019275959 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0110_of_1380_text_document dclm +0.0018478533 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0048_of_1380_text_document dclm +0.0018989246 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0059_of_1380_text_document dclm +0.0018626284 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0049_of_1380_text_document dclm +0.0019245024 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0367_of_1380_text_document dclm +0.0018410363 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0471_of_1380_text_document dclm +0.0018887022 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0021_of_1380_text_document dclm +0.0018917279 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0070_of_1380_text_document dclm +0.0018892320 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0397_of_1380_text_document dclm +0.0018450127 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0820_of_1380_text_document dclm +0.0018273267 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0056_of_1380_text_document dclm +0.0018991843 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0673_of_1380_text_document dclm +0.0018330377 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0961_of_1380_text_document dclm +0.0018961392 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0289_of_1380_text_document dclm +0.0019015425 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0686_of_1380_text_document dclm +0.0018246617 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0892_of_1380_text_document dclm +0.0018598920 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0005_of_1380_text_document dclm +0.0018133762 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0385_of_1380_text_document dclm +0.0018510191 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0912_of_1380_text_document dclm +0.0018743293 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1247_of_1380_text_document dclm +0.0018841601 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0123_of_1380_text_document dclm +0.0018805927 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1257_of_1380_text_document dclm +0.0018531046 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0739_of_1380_text_document dclm +0.0018023724 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1352_of_1380_text_document dclm +0.0019157325 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0340_of_1380_text_document dclm +0.0018709676 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0610_of_1380_text_document dclm +0.0018973913 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1307_of_1380_text_document dclm +0.0018997579 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0667_of_1380_text_document dclm +0.0019311211 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0024_of_1380_text_document dclm +0.0018997666 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0653_of_1380_text_document dclm +0.0018220302 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0254_of_1380_text_document dclm +0.0018152469 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0625_of_1380_text_document dclm +0.0018506938 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0498_of_1380_text_document dclm +0.0019001601 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0977_of_1380_text_document dclm +0.0018195061 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0404_of_1380_text_document dclm +0.0018705565 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1269_of_1380_text_document dclm +0.0018666372 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0932_of_1380_text_document dclm +0.0018169995 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1143_of_1380_text_document dclm +0.0018995026 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0190_of_1380_text_document dclm +0.0018519940 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1375_of_1380_text_document dclm +0.0018299149 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1263_of_1380_text_document dclm +0.0018292637 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1248_of_1380_text_document dclm +0.0018635921 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0334_of_1380_text_document dclm +0.0018578501 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0054_of_1380_text_document dclm +0.0018712484 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1294_of_1380_text_document dclm +0.0018945984 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1065_of_1380_text_document dclm +0.0018837219 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0761_of_1380_text_document dclm +0.0018220231 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1051_of_1380_text_document dclm +0.0018621332 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0952_of_1380_text_document dclm +0.0019129078 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1033_of_1380_text_document dclm +0.0019037357 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0510_of_1380_text_document dclm +0.0018756068 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1346_of_1380_text_document dclm +0.0019094743 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0533_of_1380_text_document dclm +0.0018721656 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0086_of_1380_text_document dclm +0.0018816698 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0931_of_1380_text_document dclm +0.0018757105 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0199_of_1380_text_document dclm +0.0018952671 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0726_of_1380_text_document dclm +0.0019038353 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0263_of_1380_text_document dclm +0.0018808379 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1000_of_1380_text_document dclm +0.0018386710 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0886_of_1380_text_document dclm +0.0018970044 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0378_of_1380_text_document dclm +0.0019222774 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0811_of_1380_text_document dclm +0.0018347427 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0694_of_1380_text_document dclm +0.0018388299 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1329_of_1380_text_document dclm +0.0018682032 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0452_of_1380_text_document dclm +0.0018959927 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1189_of_1380_text_document dclm +0.0019111375 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0785_of_1380_text_document dclm +0.0018677360 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0028_of_1380_text_document dclm +0.0018080180 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0946_of_1380_text_document dclm +0.0018979817 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0954_of_1380_text_document dclm +0.0019163596 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0394_of_1380_text_document dclm +0.0018537028 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1094_of_1380_text_document dclm +0.0018283748 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1042_of_1380_text_document dclm +0.0018865309 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0534_of_1380_text_document dclm +0.0018478335 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1017_of_1380_text_document dclm +0.0018401747 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1354_of_1380_text_document dclm +0.0018439807 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1379_of_1380_text_document dclm +0.0019266766 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0233_of_1380_text_document dclm +0.0018271878 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0528_of_1380_text_document dclm +0.0019024668 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1324_of_1380_text_document dclm +0.0018617232 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0032_of_1380_text_document dclm +0.0018705730 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0723_of_1380_text_document dclm +0.0018917349 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0133_of_1380_text_document dclm +0.0018421526 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1005_of_1380_text_document dclm +0.0018307939 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0920_of_1380_text_document dclm +0.0018145374 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0250_of_1380_text_document dclm +0.0018993194 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0756_of_1380_text_document dclm +0.0018558510 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0154_of_1380_text_document dclm +0.0018245510 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0243_of_1380_text_document dclm +0.0018541129 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0207_of_1380_text_document dclm +0.0018381114 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0632_of_1380_text_document dclm +0.0018369944 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0085_of_1380_text_document dclm +0.0018167475 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0752_of_1380_text_document dclm +0.0019214007 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0436_of_1380_text_document dclm +0.0018599216 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0850_of_1380_text_document dclm +0.0018746494 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0087_of_1380_text_document dclm +0.0018742702 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1208_of_1380_text_document dclm +0.0018381546 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0851_of_1380_text_document dclm +0.0018768773 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0265_of_1380_text_document dclm +0.0018459132 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0611_of_1380_text_document dclm +0.0017998809 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0502_of_1380_text_document dclm +0.0018642597 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0729_of_1380_text_document dclm +0.0019009944 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0833_of_1380_text_document dclm +0.0018496478 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0256_of_1380_text_document dclm +0.0018449845 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0497_of_1380_text_document dclm +0.0018484815 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0901_of_1380_text_document dclm +0.0018590916 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0737_of_1380_text_document dclm +0.0018728336 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1079_of_1380_text_document dclm +0.0018133155 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0592_of_1380_text_document dclm +0.0018928901 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1096_of_1380_text_document dclm +0.0018508042 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0948_of_1380_text_document dclm +0.0018770165 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0491_of_1380_text_document dclm +0.0018890555 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1132_of_1380_text_document dclm +0.0018427916 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1186_of_1380_text_document dclm +0.0018575144 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1229_of_1380_text_document dclm +0.0018888786 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0120_of_1380_text_document dclm +0.0018826666 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0453_of_1380_text_document dclm +0.0019024099 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0536_of_1380_text_document dclm +0.0018794673 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0118_of_1380_text_document dclm +0.0018493040 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0969_of_1380_text_document dclm +0.0018469335 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1050_of_1380_text_document dclm +0.0019069849 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0050_of_1380_text_document dclm +0.0018294520 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1326_of_1380_text_document dclm +0.0018946885 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0857_of_1380_text_document dclm +0.0018363991 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0777_of_1380_text_document dclm +0.0018198899 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0411_of_1380_text_document dclm +0.0018488040 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0328_of_1380_text_document dclm +0.0018904332 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1318_of_1380_text_document dclm +0.0018323413 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0131_of_1380_text_document dclm +0.0018800347 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0665_of_1380_text_document dclm +0.0018705095 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0558_of_1380_text_document dclm +0.0018775255 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0515_of_1380_text_document dclm +0.0018695225 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0512_of_1380_text_document dclm +0.0018233862 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0306_of_1380_text_document dclm +0.0018852382 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0310_of_1380_text_document dclm +0.0019036863 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1198_of_1380_text_document dclm +0.0018753493 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_0707_of_1380_text_document dclm +0.0018671430 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/dclm/fused_1085_of_1380_text_document dclm +0.0033006672 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0012_of_0020_text_document starcoder +0.0037221571 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0003_of_0020_text_document starcoder +0.0042851441 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0002_of_0020_text_document starcoder +0.0030614735 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0017_of_0020_text_document starcoder +0.0024361940 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0007_of_0020_text_document starcoder +0.0024365761 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0006_of_0020_text_document starcoder +0.0042101924 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0011_of_0020_text_document starcoder +0.0031415370 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0015_of_0020_text_document starcoder +0.0024398634 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0005_of_0020_text_document starcoder +0.0051185163 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0010_of_0020_text_document starcoder +0.0042972443 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0001_of_0020_text_document starcoder +0.0025242806 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0004_of_0020_text_document starcoder +0.0018291745 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0020_of_0020_text_document starcoder +0.0034469649 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0013_of_0020_text_document starcoder +0.0035435419 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0019_of_0020_text_document starcoder +0.0026357893 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0008_of_0020_text_document starcoder +0.0025316778 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0016_of_0020_text_document starcoder +0.0036148148 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0018_of_0020_text_document starcoder +0.0051109371 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0009_of_0020_text_document starcoder +0.0032581175 /flare/AuroraGPT/datasets/olmo-mix-1124/data_fused_gemma_eod/starcoder/fused_0014_of_0020_text_document starcoder From 7012ebcc475dff3c31eb0779d66fd78a1687a4db Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 14:12:08 +0000 Subject: [PATCH 473/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index f8ccf32e38a..9bc98c4c3ac 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -7,7 +7,8 @@ HERE=$(python3 -c 'import os; print(os.getcwd())') && export HERE GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH # shellcheck disable=SC1090 -source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env +source <(curl -L https://bit.ly/ezpz-utils) +ezpz_setup_env if command -v "ezpz-test"; then log_message INFO "${GREEN}✓${RESET} ezpz is already installed." @@ -33,7 +34,11 @@ train_aGPT() { GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH # 2. source `ALCF/helpers.sh` for Megatron-DeepSpeed setup - source "${HERE}/ALCF/helpers.sh" || exit + source "${HERE}/ALCF/helpers.sh" || { + log_message ERROR "Unable to source ALCF/helpers.sh." + log_message ERROR "Please ensure you are in the correct directory." + return 1 + } # 3. call `setup` from `./ALCF/helpers.sh` setup "$@" || exit From 90aeb8221a093ca1d7dcb67c46e919392222fb1d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 12:39:48 -0500 Subject: [PATCH 474/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 71 +++++++++++++++++++++---------------------------- 1 file changed, 30 insertions(+), 41 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 908fb0366da..f2a4788c9ef 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -2,28 +2,13 @@ ############################################################################### # [`ALCF/helpers.sh`](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/helpers.sh) # -# Contains helper functions for launching `../train_llama_alcf.sh` -# -# To use, on any of {Polaris, Aurora, Sunspot} @ ALCF: -# -# ```bash -# $ git clone https://github.com/argonne-lcf/Megatron-DeepSpeed -# $ cd Megatron-DeepSpeed -# $ export PBS_O_WORKDIR=$(pwd) && source ALCF/helpers.sh && setup -# ``` -# -# and this will, automatically: -# -# 1. Setup python (conda + virtual environment) -# -# 2. Parse `$PBS_*` env vars to build appropriate `alias launch='mpiexec ...'` -# command for launching across all GPUs in our active PBS job. +# Contains helper functions for launching `../train_alcf.sh` ############################################################################### ############################################################################### # Source: # [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -source <(curl -L https://bit.ly/ezpz-utils) > /dev/null || exit +source <(curl -L https://bit.ly/ezpz-utils) > /dev/null || return 1 ezpz_setup_job >/dev/null || exit ############################################################################### @@ -218,41 +203,45 @@ setup_run_cmd() { "${ds_args[@]}" "${gpt_args[@]}" "--${DTYPE}" - "--shuffle-sample-in-corpus" - "--blend-sample-in-corpus" "--accumulate-allreduce-grads-in-fp32" + "--adjust-word-embedding-init" + "--adam-beta1=${ADAM_BETA1:-0.9}" + "--adam-beta2=${ADAM_BETA2:-0.95}" + "--adam-eps=${ADAM_EPS:-0.00001}" + "--blend-sample-in-corpus" + "--clip-grad=${CLIP_GRAD:-1.0}" + "--data-cache-path=${data_cache_path}" + "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" + "--distributed-backend=${BE}" + "--ds-sequence-parallel-size=${SP}" + "--eval-interval=${EVAL_INTERVAL:-100}" + "--eval-iters=${EVAL_ITERS:-20}" + "--global-batch-size=${GLOBAL_BATCH}" + "--hidden-size=${HIDDEN}" + "--init-method-std=$(echo "scale=6; sqrt(2 / (5 * ${HIDDEN}))" | bc -l)" + "--log-interval=${LOG_INTERVAL:-1}" + "--load=${LOAD:-${CKPT_DIR}}" + "--max-position-embeddings=${SEQ}" + "--micro-batch-size=${MICRO_BATCH}" "--no-bias-gelu-fusion" "--no-bias-dropout-fusion" "--no-masked-softmax-fusion" "--no-gradient-accumulation-fusion" + "--num-layers=${NLAYERS}" + "--num-attention-heads=${HEADS}" "--optimizer=${OPT}" - "--tensor-model-parallel-size=${TP}" "--pipeline-model-parallel-size=${PP}" - "--max-position-embeddings=${SEQ}" - "--micro-batch-size=${MICRO_BATCH}" - "--ds-sequence-parallel-size=${SP}" - "--global-batch-size=${GLOBAL_BATCH}" - "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" - "--timing-log-level=${TIMING_LOG_LEVEL:-1}" - "--eval-interval=${EVAL_INTERVAL:-100}" - "--eval-iters=${EVAL_ITERS:-20}" - "--save-interval=${SAVE_INTERVAL:-50}" - "--log-interval=${LOG_INTERVAL:-1}" + "--rotary-position-embeddings-theta=${ROPE_THETA:-50000}" "--save=${SAVE:-${CKPT_DIR}}" - "--load=${LOAD:-${CKPT_DIR}}" "--seq-length=${SEQ}" - "--num-layers=${NLAYERS}" - "--hidden-size=${HIDDEN}" + "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" + "--shuffle-sample-in-corpus" + "--save-interval=${SAVE_INTERVAL:-50}" "--train-iters=${TRAIN_ITERS}" - "--distributed-backend=${BE}" + "--tensor-model-parallel-size=${TP}" + "--timing-log-level=${TIMING_LOG_LEVEL:-1}" "--weight-decay=${WEIGHT_DECAY:-0.1}" - "--adam-beta1=${ADAM_BETA1:-0.9}" - "--adam-beta2=${ADAM_BETA2:-0.95}" - "--adam-eps=${ADAM_EPS:-0.00001}" - "--clip-grad=${CLIP_GRAD:-1.0}" - "--num-attention-heads=${HEADS}" - "--data-cache-path=${data_cache_path}" - "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" + "--word-embedding-init-std=0.632455532" ) declare -A arch_map printf "==== ARCHITECTURE ====\n" From f41b3ab18469f4a2baefd73a36decb8ca2b1aaca Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 12:44:35 -0500 Subject: [PATCH 475/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 142 +++++++----------------------------------------- 1 file changed, 19 insertions(+), 123 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index f2a4788c9ef..97a343fd66f 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -299,61 +299,6 @@ save_dotenv() { fi } -####################################################################### -## get_machine_name: -## -## Return current machine name, as lowercase string -## -## Example: -## ```bash -## $ machine_name=$(get_machine_name) -## $ echo "machine_name: ${machine_name}" -## ``` -####################################################################### -#get_machine_name() { -# if [[ $(hostname) == x4* || $(hostname) == aurora* ]]; then -# machine="aurora" -# elif [[ $(hostname) == x1* || $(hostname) == uan* ]]; then -# machine="sunspot" -# elif [[ $(hostname) == x3* || $(hostname) == polaris* ]]; then -# if [[ "${PBS_O_HOST}" == sirius* ]]; then -# machine="sirius" -# else -# machine="polaris" -# fi -# elif [[ $(hostname) == sophia* ]]; then -# machine="sophia" -# elif [[ $(hostname) == nid* ]]; then -# machine="perlmutter" -# else -# machine=$(hostname) -# fi -# echo "${machine}" -#} -# -#get_machine() { -# machine=$(hostname) -# if [[ $(hostname) == x4* ]]; then -# machine="aurora" -# elif [[ $(hostname) == x1* ]]; then -# machine="sunspot" -# elif [[ $(hostname) == x3* ]]; then -# if [[ "${PBS_O_HOST}" == sirius* ]]; then -# machine="sirius" -# else -# machine="polaris" -# fi -# elif [[ $(hostname) == sophia* ]]; then -# machine="sophia" -# elif [[ $(hostname) == nid* ]]; then -# machine="perlmutter" -# else -# echo "Unknown MACHINE. Setting MACHINE to $(hostname) and continuing..." -# fi -# export MACHINE="${machine}" -# printf "Running on: %s\n" "$(printBlue "${MACHINE}")" -#} - check_and_kill_if_running() { RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') if [[ -n "${RUNNING_PIDS}" ]]; then @@ -441,21 +386,6 @@ setupLauncher() { printf " %s" "$(printMagenta "${LAUNCHER}")" } -# set_lr_args() { -# export LR=${LR:-0.0002} # LEARNING_RATE -# export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP -# export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS -# LR_ARGS="--lr ${LR} --lr-decay-style cosine" -# if [[ -n "${LR_DECAY_ITERS:-}" ]]; then -# LR_ARGS="${LR_ARGS} --lr-decay-iters ${LR_DECAY_ITERS}" -# fi -# if [[ -n "${LR_WARMUP_FRAC}" ]]; then -# LR_ARGS="${LR_ARGS} --lr-warmup-fraction ${LR_WARMUP_FRAC}" -# fi -# echo "LR_ARGS: ${LR_ARGS}" -# export LR_ARGS="${LR_ARGS}" -# } - ######################################################################### # `get_batch_size_on_polaris`: Identify MICRO_BATCH to use on Polaris. # @@ -593,27 +523,29 @@ get_model_arch_llama3_3B() { export FFN_HIDDEN_SIZE=8192 export SEQ=8192 export MODEL_ARCH="llama3-3B" - # [NOTE]----------------------------------- - # - OOM on Aurora with: - # - Config 1: - # - ZERO_STAGE=1 - # - USE_ACTIVATION_CHECKPOINTING=0 - # - Config 2: - # - ZERO_STAGE=0 - # - USE_ACTIVATION_CHECKPOINTING=1 - # - # - hmmmmm 🤔 seems odd??? - # - suspect due to SEQ_LEN=8192 - # (as opposed to 4096 for 7B model) - # ----------------------------------------- } +get_model_arch_smollm3_3B() { + export HEADS=16 + export NLAYERS=36 + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="smollm3-3B" +} + +get_model_arch_phi4_mini() { + export HEADS=32 + export NLAYERS=24 + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="phi4-mini" +} get_model_arch_llama3_3B_customNlayers() { - # if [[ "$#" -ne 1 ]]; then - # log_message ERROR "Expected 1 argument (nLayers), received: $*" - # return 1 - # fi export HEADS=24 export NLAYERS="${NLAYERS:-28}" # default to 28 layers export HIDDEN=3072 @@ -623,32 +555,8 @@ get_model_arch_llama3_3B_customNlayers() { export MODEL_ARCH="llama3-3B-nLayers${NLAYERS}" } -get_model_arch_smollm3_3B() { - export HEADS=16 - export NLAYERS=36 - export HIDDEN=2048 - export NUM_KV_HEAD=4 - export FFN_HIDDEN_SIZE=11008 - export SEQ=8192 - export MODEL_ARCH="smollm3-3B" -} - -# get_model_arch_smollm3_3B_custom() { -# export HEADS=16 -# export NLAYERS=24 -# export HIDDEN=2048 -# export NUM_KV_HEAD=4 -# export FFN_HIDDEN_SIZE=11008 -# export SEQ=8192 -# export MODEL_ARCH="smollm3-3B-nLayers24" -# } get_model_arch_smollm3_3B_custom_nLayers() { - # if [[ "$#" -ne 1 ]]; then - # log_message ERROR "Expected 1 argument (nLayers), received: $*" - # return 1 - # fi - # local nlayers="$1" export HEADS=16 export NLAYERS="${NLAYERS:-24}" # default to 24 layers export HIDDEN=2048 @@ -658,16 +566,6 @@ get_model_arch_smollm3_3B_custom_nLayers() { export MODEL_ARCH="smollm3-nLayers${NLAYERS}" } -get_model_arch_phi4_mini() { - export HEADS=32 - export NLAYERS=24 - export HIDDEN=3072 - export NUM_KV_HEAD=8 - export FFN_HIDDEN_SIZE=8192 - export SEQ=8192 - export MODEL_ARCH="phi4-mini" -} - get_model_arch_phi4_mini_custom_nLayers() { export HEADS=32 export NLAYERS="${NLAYERS:-24}" # default to 24 layers @@ -678,8 +576,6 @@ get_model_arch_phi4_mini_custom_nLayers() { export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" } - - # get_model_arch_70B() { # # 70B # export MODEL_KEY="AuroraGPT-70B" From df0c30a01f3103e2e139da4149b5a1b3f393accd Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 12:45:23 -0500 Subject: [PATCH 476/590] chore: Update `megatron/training_log_alcf.py` --- megatron/training_log_alcf.py | 1 + 1 file changed, 1 insertion(+) diff --git a/megatron/training_log_alcf.py b/megatron/training_log_alcf.py index a17e3141ca3..9982f868fd8 100644 --- a/megatron/training_log_alcf.py +++ b/megatron/training_log_alcf.py @@ -674,6 +674,7 @@ def training_log( ) if avg > 0.0: log_string += " {}={:.6f} |".format(key, avg) + wandb_metrics[f"loss/{key}"] = avg total_loss_dict[key] = torch.tensor([0.0]).to(DEVICE) if loss_scale is not None: log_string += " loss_scale={:.1f} |".format(loss_scale) From eb10947422886dc901787d75401bdf68005ef1f2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 15:28:23 -0500 Subject: [PATCH 477/590] docs: Add `ALCF/notes/AuroraGPT-small.md` --- ALCF/notes/AuroraGPT-small.md | 151 ++++++++++++++++++++++++++++++++++ 1 file changed, 151 insertions(+) create mode 100644 ALCF/notes/AuroraGPT-small.md diff --git a/ALCF/notes/AuroraGPT-small.md b/ALCF/notes/AuroraGPT-small.md new file mode 100644 index 00000000000..91f42a6be56 --- /dev/null +++ b/ALCF/notes/AuroraGPT-small.md @@ -0,0 +1,151 @@ +# AuroraGPT-v1 (Small) + +- [WandB Report: AuroraGPT-v1 (Small)](https://api.wandb.ai/links/aurora_gpt/5qxfdak3) + +## 📊 Performance Results + +| model-arch | `nlayers` | model size (B) | `tpgps` | `TFLOPs` | +| :--------: | :-------: | :------------: | :-----: | :------: | +| Llama3 | 4 | 1.98 | 8018 | 66.75 | +| Llama3 | 6 | 2.18 | 6786 | 68.76 | +| Llama3 | 8 | 2.38 | 5874 | 70.37 | +| Llama3 | 10 | 2.58 | 5179 | 71.36 | +| Llama3 | 12 | 2.78 | 4646 | 72.34 | +| Llama3 | 14 | 2.98 | 4202 | 73.15 | +| SmolLM3 | 8 | 1.67 | 7316 | 61.59 | +| SmolLM3 | 10 | 1.83 | 6448 | 62.72 | +| SmolLM3 | 12 | 1.99 | 5780 | 64.09 | +| SmolLM3 | 14 | 2.14 | 5238 | 64.82 | +| SmolLM3 | 16 | 2.30 | 4734 | 64.77 | +| SmolLM3 | 18 | 2.45 | 4363 | 65.38 | + +- Note: OOM for LLama3 @ 16 layers and SmolLM3 @ 20 layers + +## ⚙️ Configs + +- Explicit command: + + ```bash + MODEL_ARCH={smollm3-3B,llama3-3B} \ + NLAYERS= \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + OPT=adamw \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/$(ezpz_get_machine_name)/books.txt \ + bash train_alcf.sh + ``` + + +- Llama3 Architecture: + + ```llama3-architecture.yaml + HEADS: 24 + HIDDEN: 3072 + FFN_HIDDEN_SIZE: 8192 + NLAYERS: XX + NUM_KV_HEAD: 8 + SEQ: 8192 + USE_ACTIVATION_CHECKPOINTING: 0 + ZERO: 0 + ``` + +- SmolLM3 Architecture: + + ```smollm3-architecture.yaml + HEADS: 16 + HIDDEN: 2048 + FFN_HIDDEN_SIZE: 11008 + NLAYERS: XX + NUM_KV_HEAD: 4 + SEQ: 8192 + USE_ACTIVATION_CHECKPOINTING: 0 + ZERO: 0 + ``` + +[^lmax]: This is (~ roughly) at memory capacity + +## Raw Data + +- TFLOPs data: + + ```bash + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-124405][id:Aurora, 2, llama3-3B-nLayers14, 48, 2025-08-21-124405][TFLOPS-lm:73.04185] 73.15212991672594 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-133138][id:Aurora, 2, llama3-3B-nLayers12, 48, 2025-08-21-133138][TFLOPS-lm:72.30713] 72.34283930368927 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-134052][id:Aurora, 2, llama3-3B-nLayers10, 48, 2025-08-21-134052][TFLOPS-lm:71.56954] 71.56953772155975 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-122810][id:Aurora, 2, llama3-3B-nLayers10, 48, 2025-08-21-122810][TFLOPS-lm:71.26275] 71.36638292006606 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-134526][id:Aurora, 2, llama3-3B-nLayers8, 48, 2025-08-21-134526][TFLOPS-lm:70.29466] 70.37302585667582 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-142250][id:Aurora, 2, llama3-3B-nLayers6, 48, 2025-08-21-142250][TFLOPS-lm:68.81743] 68.75536285086024 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-143234][id:Aurora, 2, llama3-3B-nLayers4, 48, 2025-08-21-143234][TFLOPS-lm:66.88754] 66.7547450957777 + + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-131910][id:Aurora, 2, smollm3-nLayers18, 48, 2025-08-21-131910][TFLOPS-lm:66.34897] 66.319195539024 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-125937][id:Aurora, 2, smollm3-nLayers16, 48, 2025-08-21-125937][TFLOPS-lm:65.7598] 65.87611405627682 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-140507][id:Aurora, 2, smollm3-nLayers14, 48, 2025-08-21-140507][TFLOPS-lm:12.39282] 65.57458612915367 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-135339][id:Aurora, 2, smollm3-nLayers12, 48, 2025-08-21-135339][TFLOPS-lm:64.87912] 64.89387696083435 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-134749][id:Aurora, 2, smollm3-nLayers10, 48, 2025-08-21-134749][TFLOPS-lm:63.40927] 63.266400862713915 + [Aurora] [NHOST:2][MB:1][GAS:2][GB:48] [@ 2025-08-21-135130][id:Aurora, 2, smollm3-nLayers8, 48, 2025-08-21-135130][TFLOPS-lm:62.34549] 62.28304152993416 + ``` + +- LLama3 tokens per gpu per second: + + ```bash + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers14, args.global_batch_size: 48, created_at: 2025-08-21-124405 4202.008365924828 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers12, args.global_batch_size: 48, created_at: 2025-08-21-133138 4646.389722691857 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers12, args.global_batch_size: 96, created_at: 2025-08-21-121138 4986.43082836443 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-122810 5179.048714554391 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers8, args.global_batch_size: 48, created_at: 2025-08-21-134526 5874.068043287903 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers6, args.global_batch_size: 48, created_at: 2025-08-21-142250 6786.922466326357 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers4, args.global_batch_size: 48, created_at: 2025-08-21-143234 8018.191437878496 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-134749 6445.155290877832 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers8, args.global_batch_size: 48, created_at: 2025-08-21-135130 7327.6685718892195 + ``` + +- SmolLM3 (global) tokens per second: + + + ```bash + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers18, args.global_batch_size: 48, created_at: 2025-08-21-131910 104714.98091017039 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers16, args.global_batch_size: 48, created_at: 2025-08-21-125937 113606.28598021298 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers14, args.global_batch_size: 48, created_at: 2025-08-21-140507 125730.55726532736 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers12, args.global_batch_size: 48, created_at: 2025-08-21-135339 138737.32674433687 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-134749 154746.97733244646 + machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers8, args.global_batch_size: 48, created_at: 2025-08-21-135130 175552.08689111052 + ``` + + - Corrections: + + ```bash + 175552 / 24 = 7315.5 + 154746 / 24 = 6447.75 + 138737 / 24 = 5,780.71 + 125730 / 24 = 5,238.75 + 113606 / 24 = 4733.58 + 104714 / 24 = 4363.08 + ``` + +- SmolLM3 Model size: + + ```bash + [smollm3-nLayers18][nlayers: 18] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers18, args.global_batch_size: 48, created_at: 2025-08-21-131910 2.454792192 + [smollm3-nLayers16][nlayers: 16] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers16, args.global_batch_size: 48, created_at: 2025-08-21-125937 2.298546176 + [smollm3-nLayers14][nlayers: 14] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers14, args.global_batch_size: 48, created_at: 2025-08-21-140507 2.14230016 + [smollm3-nLayers12][nlayers: 12] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers12, args.global_batch_size: 48, created_at: 2025-08-21-135339 1.986054144 + [smollm3-nLayers10][nlayers: 10] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-134749 1.829808128 + [smollm3-nLayers8][nlayers: 8] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: smollm3-nLayers8, args.global_batch_size: 48, created_at: 2025-08-21-135130 1.673562112 + ``` + +- Explicitly: + + ```bash + [llama3-3B-nLayers14][nlayers: 14] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers14, args.global_batch_size: 48, created_at: 2025-08-21-124405 2.982239232 + [llama3-3B-nLayers12][nlayers: 12] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers12, args.global_batch_size: 48, created_at: 2025-08-21-133138 2.780900352 + [llama3-3B-nLayers12][nlayers: 12] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers12, args.global_batch_size: 96, created_at: 2025-08-21-121138 2.780900352 + [llama3-3B-nLayers10][nlayers: 10] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-134052 2.579561472 + [llama3-3B-nLayers10][nlayers: 10] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers10, args.global_batch_size: 48, created_at: 2025-08-21-122810 2.579561472 + [llama3-3B-nLayers8][nlayers: 8] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers8, args.global_batch_size: 48, created_at: 2025-08-21-134526 2.378222592 + [llama3-3B-nLayers6][nlayers: 6] machine: Aurora, env.NHOSTS: 2, env.MODEL_ARCH: llama3-3B-nLayers6, args.global_batch_size: 48, created_at: 2025-08-21-142250 2.176883712 + ``` From 50050fdf7607d1d46c7dfe4366daa667d58f2e5c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 15:41:09 -0500 Subject: [PATCH 478/590] docs: Update `ALCF/notes/AuroraGPT-small.md` --- ALCF/notes/AuroraGPT-small.md | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/ALCF/notes/AuroraGPT-small.md b/ALCF/notes/AuroraGPT-small.md index 91f42a6be56..13d3379e294 100644 --- a/ALCF/notes/AuroraGPT-small.md +++ b/ALCF/notes/AuroraGPT-small.md @@ -40,34 +40,35 @@ bash train_alcf.sh ``` - - Llama3 Architecture: ```llama3-architecture.yaml HEADS: 24 HIDDEN: 3072 FFN_HIDDEN_SIZE: 8192 - NLAYERS: XX + NLAYERS: 28 # default NUM_KV_HEAD: 8 SEQ: 8192 USE_ACTIVATION_CHECKPOINTING: 0 ZERO: 0 ``` + - [meta-llama/Llama-3.2-3B/`config.json`](https://huggingface.co/meta-llama/Llama-3.2-3B/blob/main/config.json) + - SmolLM3 Architecture: ```smollm3-architecture.yaml HEADS: 16 HIDDEN: 2048 FFN_HIDDEN_SIZE: 11008 - NLAYERS: XX + NLAYERS: 36 NUM_KV_HEAD: 4 SEQ: 8192 USE_ACTIVATION_CHECKPOINTING: 0 ZERO: 0 ``` -[^lmax]: This is (~ roughly) at memory capacity + - [HuggingFaceTB/SmolLM3-3B-Base/`config.json`](https://huggingface.co/HuggingFaceTB/SmolLM3-3B-Base/blob/main/config.json) ## Raw Data From f12d9703e9e6aec920d67545532e49b0505465b5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 21 Aug 2025 22:14:25 +0000 Subject: [PATCH 479/590] feat: Update `ALCF/data-lists/sunspot/books.txt` --- ALCF/data-lists/sunspot/books.txt | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/ALCF/data-lists/sunspot/books.txt b/ALCF/data-lists/sunspot/books.txt index 9502fba1f5f..a97584853a1 100644 --- a/ALCF/data-lists/sunspot/books.txt +++ b/ALCF/data-lists/sunspot/books.txt @@ -1,3 +1,5 @@ -0.0031025147279277244 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0000_text_document books -0.003102019887362634 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0001_text_document books -0.0009996745994661548 /gila/Aurora_deployment/AuroraGPT/datasets/dolma/data_v1.7_Llama2Tokenizer/books-0002_text_document books +0.0031025147279277244 /tegu/datascience/foremans/books-dataset/books-0000_text_document books +0.003102019887362634 /tegu/datascience/foremans/books-dataset/books-0001_text_document books +0.0009996745994661548 /tegu/datascience/foremans/books-dataset/books-0002_text_document books + + From 7ab7e357dd2bcadf8757d2ea883495d2f19b1a81 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Fri, 22 Aug 2025 13:31:00 -0500 Subject: [PATCH 480/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 97a343fd66f..b74d66172b0 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -221,7 +221,7 @@ setup_run_cmd() { "--init-method-std=$(echo "scale=6; sqrt(2 / (5 * ${HIDDEN}))" | bc -l)" "--log-interval=${LOG_INTERVAL:-1}" "--load=${LOAD:-${CKPT_DIR}}" - "--max-position-embeddings=${SEQ}" + "--max-position-embeddings=$(( 16 * SEQ ))" "--micro-batch-size=${MICRO_BATCH}" "--no-bias-gelu-fusion" "--no-bias-dropout-fusion" From 456abc61ddb63a941553aca447981c26d9b2234b Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Sun, 24 Aug 2025 23:51:46 +0000 Subject: [PATCH 481/590] Added muonclip and fixed lr_finder logic --- megatron/arguments.py | 9 ++ megatron/optimizer/__init__.py | 17 +++ megatron/optimizer/muon.py | 213 ++++++++++++++++++++++++++++++++- megatron/training.py | 97 +++++++++++---- 4 files changed, 312 insertions(+), 24 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index ce5582702b2..0de06d0976a 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1203,6 +1203,14 @@ def _add_regularization_args(parser): "--muonadamw-eps", type=float, default=1e-12, help="epsilon for internal adamw" ) + group.add_argument( + "--muonclip-t", type=float, default=100.0, help="qk threshold" + ) + + group.add_argument( + "--muonclip-alpha", type=float, default=0.5, help="split exponent for (Wq, Wk)" + ) + return parser @@ -1497,6 +1505,7 @@ def _add_training_args(parser): "sophiag", "adopt", "muon", + "muonclip", ], help="Optimizer function", ) diff --git a/megatron/optimizer/__init__.py b/megatron/optimizer/__init__.py index 47ca85f5145..68c2d33e8cc 100644 --- a/megatron/optimizer/__init__.py +++ b/megatron/optimizer/__init__.py @@ -445,6 +445,23 @@ def optimizer_hook(p): adamw_betas=(args.muonadamw_beta1, args.muonadamw_beta2), adamw_eps=args.muonadamw_eps, ) + # --- muonclip ----------------------------------------------- + elif str(args.optimizer).lower() == "muonclip": + from .muon import MuonClip + + optimizer = MuonClip( + param_groups, + lr=args.lr, + wd=args.weight_decay, + momentum=args.muon_momentum, + nesterov=args.muon_nesterov, + ns_steps=args.muon_ns_steps, + adamw_betas=(args.muonadamw_beta1, args.muonadamw_beta2), + adamw_eps=args.muonadamw_eps, + qk_clip=True, # turn on MuonClip + clip_t=args.muonclip_t, + alpha=args.muonclip_alpha, + ) else: raise TypeError(f"{args.optimizer} optimizer is not supported.") assert optimizer is not None diff --git a/megatron/optimizer/muon.py b/megatron/optimizer/muon.py index c8e4bd7b4e4..05b239f5dbc 100644 --- a/megatron/optimizer/muon.py +++ b/megatron/optimizer/muon.py @@ -3,7 +3,7 @@ import math import torch.distributed as dist from torch import Tensor - +from typing import Iterable, Optional, Callable, List, Dict, Any # This code snippet is a modified version adapted from the following GitHub repository: # https://github.com/KellerJordan/Muon/blob/master/muon.py @@ -242,3 +242,214 @@ def step(self, closure=None): p.data.add_(g, alpha=-lr / scale) return loss + + +# ========================================================== +# MuonClip: Muon + post-step QK-clip +# ========================================================== + +class QKInputRecorder: + """ + Minimal helper to capture the input 'x' that feeds an attention block's + q/k projections on each forward pass. + + Usage: + rec = QKInputRecorder() + optimizer.attach_to_attention(attn, recorder=rec, q_attr='q_proj', k_attr='k_proj', + d_head=head_dim, t=100.0, alpha=0.5) + """ + def __init__(self): + self._buffers: Dict[int, Tensor] = {} + self._handles: List[Any] = [] + + def _make_hook(self, key: int): + def _capture(module, inputs): + # forward_pre_hook → inputs[0] is the module input (x) + x: Tensor = inputs[0] + self._buffers[key] = x.detach() + return _capture + + def attach(self, module) -> Callable[[], Optional[Tensor]]: + key = id(module) + handle = module.register_forward_pre_hook(self._make_hook(key)) + self._handles.append(handle) + def getter() -> Optional[Tensor]: + return self._buffers.get(key, None) + return getter + + def remove(self): + for h in self._handles: + h.remove() + self._handles.clear() + self._buffers.clear() + + +class MuonClip(Muon): + """ + Muon optimizer with qk-clip: + After the normal Muon/AdamW updates, for each registered (W_q, W_k) pair: + η = min(t / max_ij(q_i^T k_j), 1), q = x W_q^T, k = x W_k^T + W_q ← η^α W_q, W_k ← η^(1-α) W_k + (optionally divide logits by sqrt(d_head) to match attention) + + Register pairs via: + - attach_to_attention(attn_module, recorder, q_attr='q_proj', k_attr='k_proj', d_head=..., t=..., alpha=...) + - register_qk_pair(W_q, W_k, x_getter, d_head=None, t=None, alpha=None) + + Notes: + * Clipping runs under no_grad and does not backprop. + * If x_getter() returns None (no forward this step), clipping is skipped. + * In DDP, we take the MAX of max-logit across ranks before computing η. + """ + + def __init__( + self, + params: Iterable[Tensor], + *, + lr: float = 1e-3, + wd: float = 0.1, + momentum: float = 0.95, + nesterov: bool = True, + ns_steps: int = 5, + adamw_betas=(0.95, 0.95), + adamw_eps: float = 1e-8, + # MuonClip extras: + qk_clip: bool = True, + clip_t: float = 100.0, + alpha: float = 0.5, + use_sqrt_d: bool = True, + ): + super().__init__( + params, + lr=lr, + wd=wd, + momentum=momentum, + nesterov=nesterov, + ns_steps=ns_steps, + adamw_betas=adamw_betas, + adamw_eps=adamw_eps, + ) + self._clip_enabled = bool(qk_clip) + self._clip_t_default = float(clip_t) + self._alpha_default = float(alpha) + self._use_sqrt_d = bool(use_sqrt_d) + self._pairs: List[Dict[str, Any]] = [] # W_q, W_k, x_getter, d_head, t, alpha + + # --------- Registration APIs --------- + + def register_qk_pair( + self, + W_q: Tensor, + W_k: Tensor, + x_getter: Callable[[], Optional[Tensor]], + *, + d_head: Optional[int] = None, + t: Optional[float] = None, + alpha: Optional[float] = None, + ): + self._pairs.append( + dict( + W_q=W_q, + W_k=W_k, + x_getter=x_getter, + d_head=d_head, + t=float(t) if t is not None else None, + alpha=float(alpha) if alpha is not None else None, + ) + ) + + def attach_to_attention( + self, + attn_module: torch.nn.Module, + *, + recorder: Optional[QKInputRecorder] = None, + q_attr: str = "q_proj", + k_attr: str = "k_proj", + d_head: Optional[int] = None, + t: Optional[float] = None, + alpha: Optional[float] = None, + ): + """ + Convenience hook for the common case with separate q/k Linear modules. + """ + assert hasattr(attn_module, q_attr) and hasattr(attn_module, k_attr), \ + f"Module must have {q_attr} and {k_attr}" + q_lin = getattr(attn_module, q_attr) + k_lin = getattr(attn_module, k_attr) + assert hasattr(q_lin, "weight") and hasattr(k_lin, "weight") + + if recorder is None: + recorder = QKInputRecorder() + x_getter = recorder.attach(attn_module) + self.register_qk_pair(q_lin.weight, k_lin.weight, x_getter, + d_head=d_head, t=t, alpha=alpha) + return recorder # keep this object alive somewhere! + + # ------------- Core step -------------- + + @torch.no_grad() + def _apply_qk_clip_once( + self, + W_q: Tensor, + W_k: Tensor, + x: Tensor, + *, + d_head: Optional[int], + t: float, + alpha: float, + eps: float = 1e-12, + ): + """ + Compute η from current batch x and rescale W_q/W_k in-place. + """ + # x: (B, T, d_model); W_q/W_k: (out_features, d_model) + device = W_q.device + x = x.to(device, non_blocking=True) + + q = x @ W_q.T + k = x @ W_k.T + + scores = torch.einsum("bid,bjd->bij", q, k) + if self._use_sqrt_d: + denom = float(d_head if d_head is not None else W_q.size(0)) + scores = scores / (denom ** 0.5) + + max_logit = scores.max() + + # Global max across DDP ranks, if any + if dist.is_available() and dist.is_initialized(): + dist.all_reduce(max_logit, op=dist.ReduceOp.MAX) + + # η = min(t / max_logit, 1) + if max_logit <= 0: + eta = torch.tensor(1.0, device=device) + else: + eta = torch.clamp(t / (max_logit + eps), max=1.0) + + # Always apply (η==1 → no change) + W_q.mul_(eta ** alpha) + W_k.mul_(eta ** (1.0 - alpha)) + + return float(max_logit.item()), float(eta.item()) + + @torch.no_grad() + def step(self, closure=None): + # 1) normal Muon/AdamW step + loss = super().step(closure=closure) + + # 2) qk-clip (post-step) + if self._clip_enabled and len(self._pairs) > 0: + for pair in self._pairs: + x = pair["x_getter"]() + if x is None: + continue # no forward for this module this step + self._apply_qk_clip_once( + pair["W_q"], + pair["W_k"], + x, + d_head=pair.get("d_head", None), + t=pair.get("t", self._clip_t_default) or self._clip_t_default, + alpha=pair.get("alpha", self._alpha_default) + if pair.get("alpha", None) is not None else self._alpha_default, + ) + return loss diff --git a/megatron/training.py b/megatron/training.py index c239ea5f183..c0938b6c4b7 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1100,6 +1100,10 @@ def train( f"Running learning rate finder for {finder_iters} iterations (10% of {args.train_iters})" ) + # convenience handles for DP sync + dp_group = mpu.get_data_parallel_group() + dev = torch.device(DEVICE_TYPE) + # Main LR finder loop for i in range(finder_iters): # Execute training step @@ -1112,18 +1116,41 @@ def train( config, ) - # Skip iterations that didn't update weights + # --- MAKE skipped_iter GLOBAL (any rank skip => all skip) --- + if tdist.is_available() and tdist.is_initialized(): + _skip = torch.tensor([1 if skipped_iter else 0], device=dev) + tdist.all_reduce(_skip, op=tdist.ReduceOp.MAX, group=dp_group) + skipped_iter = bool(_skip.item()) + + # ---- PRESERVE ORIGINAL SEMANTICS: do NOT advance LR when skipped ---- if skipped_iter: + # Keep control flow identical across ranks continue - # Get loss value (use first available loss if multiple present) + # Get loss value (use first available loss if multiple present). + # Non-pipeline-last stages return {}, so contribute 0. if "lm loss" in loss_dict: - loss_val = loss_dict["lm loss"] + local_loss = loss_dict["lm loss"] + elif loss_dict: + local_loss = next(iter(loss_dict.values())) else: - loss_val = next(iter(loss_dict.values())) + local_loss = 0.0 + + if isinstance(local_loss, torch.Tensor): + local_loss = local_loss.item() + + # --- REDUCE loss across DP (average of contributing ranks) --- + loss_t = torch.tensor([float(local_loss)], device=dev) + # Only pipeline-last ranks contribute 1.0 to the denominator + contrib = 1.0 if mpu.is_pipeline_last_stage(ignore_virtual=True) else 0.0 + denom_t = torch.tensor([contrib], device=dev) + + if tdist.is_available() and tdist.is_initialized(): + tdist.all_reduce(loss_t, op=tdist.ReduceOp.SUM, group=dp_group) + tdist.all_reduce(denom_t, op=tdist.ReduceOp.SUM, group=dp_group) - if isinstance(loss_val, torch.Tensor): - loss_val = loss_val.item() + dp_count = max(1.0, float(denom_t.item())) + loss_val = float(loss_t.item()) / dp_count # Update batch counter batch_num += 1 @@ -1132,31 +1159,51 @@ def train( avg_loss = beta * avg_loss + (1 - beta) * loss_val smoothed_loss = avg_loss / (1 - beta**batch_num) - # Stop if the loss is exploding - if batch_num > 1 and smoothed_loss > 4 * best_loss: - log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") - break - - # Record the best loss + # Update best_loss with the SAME smoothed_loss across all ranks if smoothed_loss < best_loss or batch_num == 1: best_loss = smoothed_loss + # explode = (batch_num > 1 and smoothed_loss > 4 * best_loss) + + + # if explode: + # if mpu.get_data_parallel_rank() == 0: + # log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") + # break + + ## Record the best loss (use the same global smoothed_loss) + #if smoothed_loss < best_loss or batch_num == 1: + # best_loss = smoothed_loss + + # --- GLOBALIZE the "loss exploding" decision (any rank => all ranks) --- + #explode_local = (batch_num > 1 and smoothed_loss > 4 * best_loss) + #print(f"Rank {mpu.get_data_parallel_rank()}: explode_local={explode_local}") + #if mpu.get_data_parallel_rank() == 0: + + # print(f"Iter {i}: batch_num={batch_num}, smoothed_loss={smoothed_loss:.8f}, best_loss={best_loss:.8f}, ratio={smoothed_loss/best_loss:.2f}") + #if tdist.is_available() and tdist.is_initialized(): + # _exp = torch.tensor([1 if explode_local else 0], device=dev) + # tdist.all_reduce(_exp, op=tdist.ReduceOp.MAX, group=dp_group) + # explode = bool(_exp.item()) + #else: + # explode = explode_local + + + + #if explode: + # if mpu.get_data_parallel_rank() == 0: + # log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") + # Keep everyone in lockstep before breaking + # if tdist.is_available() and tdist.is_initialized(): + # tdist.barrier(group=dp_group) + # break + # Record values for plotting lr_finder_losses.append(smoothed_loss) lr_finder_lrs.append(curr_lr) - # Log to wandb - # if wandb is not None and wandb.run is not None: - # wandb.log({ - # "lr_finder/learning_rate": curr_lr, - # "lr_finder/raw_loss": loss_val, - # "lr_finder/smoothed_loss": smoothed_loss, - # "lr_finder/iteration": i, - # "lr_finder/log_lr": math.log10(curr_lr) - # }) - # Print progress - if (i + 1) % args.log_interval == 0: + if (i + 1) % args.log_interval == 0 and mpu.get_data_parallel_rank() == 0: log.info( f"LR Finder: iteration {i+1}/{finder_iters}, " f"lr: {curr_lr:.8f}, loss: {smoothed_loss:.4f}" @@ -1197,6 +1244,10 @@ def train( log.info(f"LR finder completed. Results saved to {args.save}/lr_finder/") + # Ensure all ranks exit LR-finder together (prevents stragglers) + if tdist.is_available() and tdist.is_initialized(): + tdist.barrier(group=dp_group) + # Return after LR finder is done return args.iteration From e9467fa966166915f1075ed1765e2daae51f9daf Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Mon, 25 Aug 2025 23:26:06 +0000 Subject: [PATCH 482/590] Updated muonclip lr adjuster --- megatron/optimizer/muon.py | 147 ++++++++++++++++++++++++++++--------- 1 file changed, 114 insertions(+), 33 deletions(-) diff --git a/megatron/optimizer/muon.py b/megatron/optimizer/muon.py index 05b239f5dbc..0f917b36ca5 100644 --- a/megatron/optimizer/muon.py +++ b/megatron/optimizer/muon.py @@ -46,6 +46,7 @@ def __init__( ns_steps=5, adamw_betas=(0.95, 0.95), adamw_eps=1e-8, + adjuster_lr_ref=True #original moonlight lr adjustment ): defaults = dict( lr=lr, @@ -73,7 +74,7 @@ def __init__( if p.grad is None: continue - use_muon = p.ndim == 2 + use_muon = (p.ndim == 2) # If parameter is 2D but has a large dimension, it's likely an embedding or LM head, need to change this!!! if use_muon and max(p.shape) > 10000: @@ -124,10 +125,22 @@ def adjust_lr_for_muon(self, lr, param_shape): """ A, B = param_shape[:2] # We adjust the learning rate based on the size of the parameter matrix + #adjusted_ratio = max(1.0, float(A) / float(B)) ** 0.5 adjusted_ratio = 0.2 * math.sqrt(max(A, B)) adjusted_lr = lr * adjusted_ratio return adjusted_lr + def adjust_lr_for_muonclip(self, lr, param_shape): + """ + Adjust learning rate based on parameter shape for Muon. + """ + A, B = param_shape[:2] + # We adjust the learning rate based on the size of the parameter matrix + adjusted_ratio = max(1.0, float(A) / float(B)) ** 0.5 + #adjusted_ratio = 0.2 * math.sqrt(max(A, B)) + adjusted_lr = lr * adjusted_ratio + return adjusted_lr + def step(self, closure=None): """Perform a single optimization step.""" loss = None @@ -205,7 +218,10 @@ def step(self, closure=None): u = zeropower_via_newtonschulz5(g, steps=ns_steps) # Scale update - adjusted_lr = self.adjust_lr_for_muon(lr, p.shape) + if adjuster_lr_ref==True: + adjusted_lr = self.adjust_lr_for_muon(lr, p.shape) + else: + adjusted_lr = self.adjust_lr_for_muonclip(lr, p.shape) # Apply weight decay p.data.mul_(1 - lr * wd) @@ -258,31 +274,49 @@ class QKInputRecorder: optimizer.attach_to_attention(attn, recorder=rec, q_attr='q_proj', k_attr='k_proj', d_head=head_dim, t=100.0, alpha=0.5) """ - def __init__(self): + def __init__(self, auto_clear: bool = True): self._buffers: Dict[int, Tensor] = {} self._handles: List[Any] = [] + self._auto_clear = auto_clear + self._retrieved: set = set() # Track which buffers were retrieved def _make_hook(self, key: int): def _capture(module, inputs): # forward_pre_hook → inputs[0] is the module input (x) x: Tensor = inputs[0] self._buffers[key] = x.detach() + # Mark as not yet retrieved + if key in self._retrieved: + self._retrieved.remove(key) return _capture def attach(self, module) -> Callable[[], Optional[Tensor]]: key = id(module) handle = module.register_forward_pre_hook(self._make_hook(key)) self._handles.append(handle) + def getter() -> Optional[Tensor]: - return self._buffers.get(key, None) + tensor = self._buffers.get(key, None) + if tensor is not None: + self._retrieved.add(key) + return tensor return getter + + def clear_buffers(self): + """Clear only retrieved buffers to free memory.""" + if self._auto_clear: + for key in self._retrieved: + if key in self._buffers: + del self._buffers[key] + self._retrieved.clear() def remove(self): + """Remove all hooks and clear buffers.""" for h in self._handles: h.remove() self._handles.clear() self._buffers.clear() - + self._retrieved.clear() class MuonClip(Muon): """ @@ -313,6 +347,7 @@ def __init__( ns_steps: int = 5, adamw_betas=(0.95, 0.95), adamw_eps: float = 1e-8, + adjuster_lr_ref=False, # MuonClip extras: qk_clip: bool = True, clip_t: float = 100.0, @@ -328,6 +363,7 @@ def __init__( ns_steps=ns_steps, adamw_betas=adamw_betas, adamw_eps=adamw_eps, + adjuster_lr_ref=adjuster_lr_ref, ) self._clip_enabled = bool(qk_clip) self._clip_t_default = float(clip_t) @@ -347,6 +383,12 @@ def register_qk_pair( t: Optional[float] = None, alpha: Optional[float] = None, ): + + """Register a Q/K weight pair for clipping.""" + # Validate tensor shapes + assert W_q.ndim == 2 and W_k.ndim == 2, "W_q and W_k must be 2D tensors" + assert W_q.size(1) == W_k.size(1), "W_q and W_k must have same input dimension" + self._pairs.append( dict( W_q=W_q, @@ -402,6 +444,14 @@ def _apply_qk_clip_once( """ Compute η from current batch x and rescale W_q/W_k in-place. """ + # Validate and prepare input + if x.ndim == 2: + x = x.unsqueeze(0) # Add batch dimension if missing + + assert x.ndim == 3, f"Expected 3D input (batch, seq, dim), got shape {x.shape}" + assert x.size(-1) == W_q.size(1), \ + f"Input dim {x.size(-1)} doesn't match weight dim {W_q.size(1)}" + # x: (B, T, d_model); W_q/W_k: (out_features, d_model) device = W_q.device x = x.to(device, non_blocking=True) @@ -411,26 +461,36 @@ def _apply_qk_clip_once( scores = torch.einsum("bid,bjd->bij", q, k) if self._use_sqrt_d: - denom = float(d_head if d_head is not None else W_q.size(0)) - scores = scores / (denom ** 0.5) - - max_logit = scores.max() - - # Global max across DDP ranks, if any + scores = scores / (W_q.size(0) ** 0.5) + #denom = float(d_head if d_head is not None else W_q.size(0)) + #scores = scores / (denom ** 0.5) + + # Find maximum score + max_score = scores.max() + + # Check for numerical issues + if not torch.isfinite(max_score): + return False, float('nan') + + # Global max across DDP ranks if in distributed training if dist.is_available() and dist.is_initialized(): - dist.all_reduce(max_logit, op=dist.ReduceOp.MAX) - - # η = min(t / max_logit, 1) - if max_logit <= 0: - eta = torch.tensor(1.0, device=device) - else: - eta = torch.clamp(t / (max_logit + eps), max=1.0) - - # Always apply (η==1 → no change) - W_q.mul_(eta ** alpha) - W_k.mul_(eta ** (1.0 - alpha)) - - return float(max_logit.item()), float(eta.item()) + dist.all_reduce(max_score, op=dist.ReduceOp.MAX) + + max_score_val = max_score.item() if torch.is_tensor(max_score) else max_score + + # Apply clipping if max score exceeds threshold + if max_score_val > t: + eta = t / (max_score_val + eps) + scale_q = eta ** alpha + scale_k = eta ** (1 - alpha) + + # Scale the weights + W_q.mul_(scale_q) + W_k.mul_(scale_k) + + return True, max_score_val + + return False, max_score_val @torch.no_grad() def step(self, closure=None): @@ -443,13 +503,34 @@ def step(self, closure=None): x = pair["x_getter"]() if x is None: continue # no forward for this module this step - self._apply_qk_clip_once( - pair["W_q"], - pair["W_k"], - x, - d_head=pair.get("d_head", None), - t=pair.get("t", self._clip_t_default) or self._clip_t_default, - alpha=pair.get("alpha", self._alpha_default) - if pair.get("alpha", None) is not None else self._alpha_default, - ) + try: + clipped, max_score = self._apply_qk_clip_once( + pair["W_q"], + pair["W_k"], + x, + d_head=pair.get("d_head", None), + t=pair.get("t", self._clip_t_default) or self._clip_t_default, + alpha=pair.get("alpha", self._alpha_default) + if pair.get("alpha", None) is not None else self._alpha_default, + ) + + # Optional: log clipping events for monitoring + # if clipped: + # print(f"QK-clip triggered! Max score: {max_score:.2f} > {pair.get('t', self._clip_t_default):.2f}") + + except Exception as e: + # Log but don't crash training + print(f"Warning: QK-clip failed with error: {e}") + continue + + # Clear recorder buffers after each step to prevent memory accumulation + for recorder in self._recorders: + recorder.clear_buffers() return loss + def __del__(self): + """Cleanup hooks when optimizer is destroyed.""" + for recorder in self._recorders: + try: + recorder.remove() + except: + pass # Ignore errors during cleanup From 99b2592f980dd798cd212e43e6ff0e3dea3dc959 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Aug 2025 10:08:52 -0500 Subject: [PATCH 483/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2059 +++++++++++++++++++++++------------------------ 1 file changed, 1023 insertions(+), 1036 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index b74d66172b0..b7053dc37d4 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -8,7 +8,7 @@ ############################################################################### # Source: # [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) -source <(curl -L https://bit.ly/ezpz-utils) > /dev/null || return 1 +source <(curl -L https://bit.ly/ezpz-utils) >/dev/null || return 1 ezpz_setup_job >/dev/null || exit ############################################################################### @@ -33,21 +33,21 @@ ezpz_setup_job >/dev/null || exit # (e.g. virtual environment, location of executables, etc.) ################## helpers_main() { - # NOTE: for debug mode, run with `DEBUG=1` - if [[ -n "${DEBUG:-}" ]]; then - set -euxo - fi - if [[ -n "${PBS_O_WORKDIR}" ]]; then - WORKING_DIR="${PBS_O_WORKDIR}" - elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then - WORKING_DIR="${SLURM_SUBMIT_DIR}" - else - echo "Unable to detect PBS or SLURM working directory info..." - WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') - echo "Using ${WORKING_DIR} as working directory..." - fi - export WORKING_DIR="${WORKING_DIR}" - printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" + # NOTE: for debug mode, run with `DEBUG=1` + if [[ -n "${DEBUG:-}" ]]; then + set -euxo + fi + if [[ -n "${PBS_O_WORKDIR}" ]]; then + WORKING_DIR="${PBS_O_WORKDIR}" + elif [[ -n "${SLURM_SUBMIT_DIR}" ]]; then + WORKING_DIR="${SLURM_SUBMIT_DIR}" + else + echo "Unable to detect PBS or SLURM working directory info..." + WORKING_DIR=$(python3 -c 'import os; print(os.getcwd())') + echo "Using ${WORKING_DIR} as working directory..." + fi + export WORKING_DIR="${WORKING_DIR}" + printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" } ############################################################################## @@ -77,49 +77,49 @@ helpers_main() { # - Setup run command to be executed. ############################################################################## setup() { - # Identify machine we're on - mn=$(ezpz_get_machine_name) - export MACHINE="${mn}" - # get_machine || exit - ########################################################################## - # # ezpz_setup will: - # # 1. Setup python - # # - load base conda - # # - (if necessary) create virtual environment on top of base conda - # # - activate virtual environment from ^ - # # 2. Install ezpz (if needed) - # # 3. Parse PBS_* environment variables to determine: - # # - NHOSTS (by counting number of lines in $PBS_NODEFILE) - # # - NGPU_PER_HOST (by magic) - # # - NGPUS (= NHOSTS * NGPU_PER_HOST) - # # 4. Use these (^) to build our launch command - # ezpz_setup || exit - ########################################################################## - install_dependencies - # Set command line arguments to pass to `"${EXEC}"` - setParams || exit - # Create `deepspeed_config.json` from runtime params from ^ - buildDSconfig || exit - # Specify output directory for {logs, checkpoints, etc.} - setup_checkpoint || exit - setOutput || exit - # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) - set_args || exit - # Ensure executable exists in expected path - check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" - dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" - # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` - tok="${TOKENIZER_TYPE:-HFTokenizer}" - setup_tokenizer_and_data "${tok}" "${dfl}" || exit - make_data || exit - # Print job info - printJobInfo || exit - # Save `.env` to `CKPT_DIR` for safe keeping - save_dotenv "${CKPT_DIR}" || exit - # Check that were not already running, if so, exit. - check_and_kill_if_running || exit - # Setup run command to be executed - setup_run_cmd "$@" || exit + # Identify machine we're on + mn=$(ezpz_get_machine_name) + export MACHINE="${mn}" + # get_machine || exit + ########################################################################## + # # ezpz_setup will: + # # 1. Setup python + # # - load base conda + # # - (if necessary) create virtual environment on top of base conda + # # - activate virtual environment from ^ + # # 2. Install ezpz (if needed) + # # 3. Parse PBS_* environment variables to determine: + # # - NHOSTS (by counting number of lines in $PBS_NODEFILE) + # # - NGPU_PER_HOST (by magic) + # # - NGPUS (= NHOSTS * NGPU_PER_HOST) + # # 4. Use these (^) to build our launch command + # ezpz_setup || exit + ########################################################################## + install_dependencies + # Set command line arguments to pass to `"${EXEC}"` + setParams || exit + # Create `deepspeed_config.json` from runtime params from ^ + buildDSconfig || exit + # Specify output directory for {logs, checkpoints, etc.} + setup_checkpoint || exit + setOutput || exit + # Specify additional `deepspeed` arguments (dependent on _newly created_ variables) + set_args || exit + # Ensure executable exists in expected path + check_executable "${EXEC:-${WORKING_DIR}/pretrain_gpt_alcf.py}" + dfl="${DATA_FILE_LIST:-"${PBS_O_WORKDIR:-${HERE}}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt"}" + # Setup data + tokenizer via `DATA_FILE_LIST` and `TOKENIZER_TYPE` + tok="${TOKENIZER_TYPE:-HFTokenizer}" + setup_tokenizer_and_data "${tok}" "${dfl}" || exit + make_data || exit + # Print job info + printJobInfo || exit + # Save `.env` to `CKPT_DIR` for safe keeping + save_dotenv "${CKPT_DIR}" || exit + # Check that were not already running, if so, exit. + check_and_kill_if_running || exit + # Setup run command to be executed + setup_run_cmd "$@" || exit } ##################################################### @@ -128,206 +128,205 @@ setup() { # Build run command to be executed. ##################################################### setup_run_cmd() { - ############################## - # take in additional arguments - # and append them directly to - # the end of the `run_cmd` - # custom_args="$@" - custom_args=("$@") - ############################## - #### Make it easy to track experiments by date ################### - year="$(date "+%Y")" - month="$(date "+%m")" - day="$(date "+%Y-%m-%d")" - today="$(date "+%Y-%m-%d")" # kept for backwards compatibility - started_at="$(date "+%Y-%m-%d-%H%M%S")" - export YEAR="${year}" - export MONTH="${month}" - export DAY="${day}" - export TODAY="${today}" - export STARTED_AT="${started_at}" - ################################################################## - # NOTE: to launch with DeepSpeed instead of mpiexec: - # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` - ################################################################## - setupLauncher "${LAUNCH_WITH:-MPICH}" || exit - export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" - printf "\n" - echo "Using data_cache_path: ${data_cache_path}" - ################################################################## - # WARN: to disable Llama-type architectures, toggle via: - # `NO_LLAMA=1 bash train_llama_alcf.sh` - ################################################################## - if [[ -z "${NO_LLAMA:-}" ]]; then - llama_flags=( - "--swiglu" - "--hidden-dropout 0" - "--attention-dropout 0" - "--normalization rmsnorm" - "--disable-bias-linear" - "--no-query-key-layer-scaling" - "--use-rotary-position-embeddings" - "--untie-embeddings-and-output-weights" - "--num-key-value-heads ${NUM_KV_HEAD}" - "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" - ) - fi - - tb_flags=() - if [[ -z "${NO_TENSORBOARD:-}" ]]; then - TBDIR="${CKPT_DIR}/tensorboard" - mkdir -p "${TBDIR}" - tb_flags+=( - "--log-timers-to-tensorboard" - "--log-optimizer-states-to-tensorboard" - "--tensorboard-dir ${TBDIR}" - ) - fi - dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt}" - - train_args=() - if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then - train_args+=("--use-checkpoint-opt_param-scheduler") - fi - # "--init-method-std ${INIT_METHOD_STD:-0.0006}" - # "--shuffle-sample" - train_args+=( - "${lr_flags[@]}" - "${custom_args[@]}" - "${llama_flags[@]}" - "${FLASH_ARG}" - "${TIMING_STR:-}" - "${DATA_FLAGS}" - "${TOKENIZER_FLAGS}" - "${tb_flags[@]}" - "${ds_args[@]}" - "${gpt_args[@]}" - "--${DTYPE}" - "--accumulate-allreduce-grads-in-fp32" - "--adjust-word-embedding-init" - "--adam-beta1=${ADAM_BETA1:-0.9}" - "--adam-beta2=${ADAM_BETA2:-0.95}" - "--adam-eps=${ADAM_EPS:-0.00001}" - "--blend-sample-in-corpus" - "--clip-grad=${CLIP_GRAD:-1.0}" - "--data-cache-path=${data_cache_path}" - "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" - "--distributed-backend=${BE}" - "--ds-sequence-parallel-size=${SP}" - "--eval-interval=${EVAL_INTERVAL:-100}" - "--eval-iters=${EVAL_ITERS:-20}" - "--global-batch-size=${GLOBAL_BATCH}" - "--hidden-size=${HIDDEN}" - "--init-method-std=$(echo "scale=6; sqrt(2 / (5 * ${HIDDEN}))" | bc -l)" - "--log-interval=${LOG_INTERVAL:-1}" - "--load=${LOAD:-${CKPT_DIR}}" - "--max-position-embeddings=$(( 16 * SEQ ))" - "--micro-batch-size=${MICRO_BATCH}" - "--no-bias-gelu-fusion" - "--no-bias-dropout-fusion" - "--no-masked-softmax-fusion" - "--no-gradient-accumulation-fusion" - "--num-layers=${NLAYERS}" - "--num-attention-heads=${HEADS}" - "--optimizer=${OPT}" - "--pipeline-model-parallel-size=${PP}" - "--rotary-position-embeddings-theta=${ROPE_THETA:-50000}" - "--save=${SAVE:-${CKPT_DIR}}" - "--seq-length=${SEQ}" - "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" - "--shuffle-sample-in-corpus" - "--save-interval=${SAVE_INTERVAL:-50}" - "--train-iters=${TRAIN_ITERS}" - "--tensor-model-parallel-size=${TP}" - "--timing-log-level=${TIMING_LOG_LEVEL:-1}" - "--weight-decay=${WEIGHT_DECAY:-0.1}" - "--word-embedding-init-std=0.632455532" - ) - declare -A arch_map - printf "==== ARCHITECTURE ====\n" - arch_map=( - ["MODEL_ARCH"]="${MODEL_ARCH}" - ["TP"]="${TP}" - ["PP"]="${PP}" - ["SP"]="${SP}" - ["DP"]="${DP}" - ["ZERO"]="${ZERO_STAGE}" - ["MBS"]="${MICRO_BATCH}" - ["GAS"]="${GRAD_ACC_STEPS}" - ["GBS"]="${GLOBAL_BATCH}" - ["NLAYERS"]="${NLAYERS}" - ["HIDDEN"]="${HIDDEN}" - ["HEADS"]="${HEADS}" - ["SEQ"]="${SEQ}" - ["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE}" - ["NUM_KV_HEAD"]="${NUM_KV_HEAD}" - ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" - # ["DTYPE"]="${DTYPE}" - # ["OPT"]="${OPT}" - # ["WEIGHT_DECAY"]="${WEIGHT_DECAY}" - # ["ADAM_BETA1"]="${ADAM_BETA1}" - # ["ADAM_BETA2"]="${ADAM_BETA2}" - # ["ADAM_EPS"]="${ADAM_EPS}" - # ["CLIP_GRAD"]="${CLIP_GRAD:-1.0}" - ) - for v in "${!arch_map[@]}"; do - printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" - done - printf "======================\n" - # "--adam-eps ${ADAM_EPS:-0.00001}" - cache_dir="${PBS_O_WORKDIR}/.cache/" - mkdir -p "${cache_dir}" - targs_cache="${cache_dir}/train_args.txt" - for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done - export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") - printf "Training Arguments: %s\n" "$(printBlue "${TRAIN_ARGS[@]}")" - export run_cmd=("${LAUNCHER}" "${train_args[@]}") + ############################## + # take in additional arguments + # and append them directly to + # the end of the `run_cmd` + # custom_args="$@" + custom_args=("$@") + ############################## + #### Make it easy to track experiments by date ################### + year="$(date "+%Y")" + month="$(date "+%m")" + day="$(date "+%Y-%m-%d")" + today="$(date "+%Y-%m-%d")" # kept for backwards compatibility + started_at="$(date "+%Y-%m-%d-%H%M%S")" + export YEAR="${year}" + export MONTH="${month}" + export DAY="${day}" + export TODAY="${today}" + export STARTED_AT="${started_at}" + ################################################################## + # NOTE: to launch with DeepSpeed instead of mpiexec: + # `export LAUNCH_WITH=deepspeeed && bash train_llama_alcf.sh` + ################################################################## + setupLauncher "${LAUNCH_WITH:-MPICH}" || exit + export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" + printf "\n" + echo "Using data_cache_path: ${data_cache_path}" + ################################################################## + # WARN: to disable Llama-type architectures, toggle via: + # `NO_LLAMA=1 bash train_llama_alcf.sh` + ################################################################## + if [[ -z "${NO_LLAMA:-}" ]]; then + llama_flags=( + "--swiglu" + "--hidden-dropout 0" + "--attention-dropout 0" + "--normalization rmsnorm" + "--disable-bias-linear" + "--no-query-key-layer-scaling" + "--use-rotary-position-embeddings" + "--untie-embeddings-and-output-weights" + "--num-key-value-heads ${NUM_KV_HEAD}" + "--ffn-hidden-size ${FFN_HIDDEN_SIZE}" + ) + fi + + tb_flags=() + if [[ -z "${NO_TENSORBOARD:-}" ]]; then + TBDIR="${CKPT_DIR}/tensorboard" + mkdir -p "${TBDIR}" + tb_flags+=( + "--log-timers-to-tensorboard" + "--log-optimizer-states-to-tensorboard" + "--tensorboard-dir ${TBDIR}" + ) + fi + dfl_fallback="${DATA_FILE_LIST:-${PBS_O_WORKDIR}/ALCF/data-lists/$(ezpz_get_machine_name)/dolma.txt}" + + train_args=() + if [[ -z "${OVERRIDE_CKPT_OPT_PARAM:-}" ]]; then + train_args+=("--use-checkpoint-opt_param-scheduler") + fi + # "--init-method-std ${INIT_METHOD_STD:-0.0006}" + # "--shuffle-sample" + train_args+=( + "${lr_flags[@]}" + "${custom_args[@]}" + "${llama_flags[@]}" + "${FLASH_ARG}" + "${TIMING_STR:-}" + "${DATA_FLAGS}" + "${TOKENIZER_FLAGS}" + "${tb_flags[@]}" + "${ds_args[@]}" + "--${DTYPE}" + "--accumulate-allreduce-grads-in-fp32" + "--adjust-word-embedding-init" + "--adam-beta1=${ADAM_BETA1:-0.9}" + "--adam-beta2=${ADAM_BETA2:-0.95}" + "--adam-eps=${ADAM_EPS:-0.00001}" + "--blend-sample-in-corpus" + "--clip-grad=${CLIP_GRAD:-1.0}" + "--data-cache-path=${data_cache_path}" + "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" + "--distributed-backend=${BE}" + "--ds-sequence-parallel-size=${SP}" + "--eval-interval=${EVAL_INTERVAL:-100}" + "--eval-iters=${EVAL_ITERS:-20}" + "--global-batch-size=${GLOBAL_BATCH}" + "--hidden-size=${HIDDEN}" + "--init-method-std=$(echo "scale=6; sqrt(2 / (5 * ${HIDDEN}))" | bc -l)" + "--log-interval=${LOG_INTERVAL:-1}" + "--load=${LOAD:-${CKPT_DIR}}" + "--max-position-embeddings=$((16 * SEQ))" + "--micro-batch-size=${MICRO_BATCH}" + "--no-bias-gelu-fusion" + "--no-bias-dropout-fusion" + "--no-masked-softmax-fusion" + "--no-gradient-accumulation-fusion" + "--num-layers=${NLAYERS}" + "--num-attention-heads=${HEADS}" + "--optimizer=${OPT}" + "--pipeline-model-parallel-size=${PP}" + "--rotary-position-embeddings-theta=${ROPE_THETA:-50000}" + "--save=${SAVE:-${CKPT_DIR}}" + "--seq-length=${SEQ}" + "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" + "--shuffle-sample-in-corpus" + "--save-interval=${SAVE_INTERVAL:-50}" + "--train-iters=${TRAIN_ITERS}" + "--tensor-model-parallel-size=${TP}" + "--timing-log-level=${TIMING_LOG_LEVEL:-1}" + "--weight-decay=${WEIGHT_DECAY:-0.1}" + "--word-embedding-init-std=0.632455532" + ) + declare -A arch_map + printf "==== ARCHITECTURE ====\n" + arch_map=( + ["MODEL_ARCH"]="${MODEL_ARCH}" + ["TP"]="${TP}" + ["PP"]="${PP}" + ["SP"]="${SP}" + ["DP"]="${DP}" + ["ZERO"]="${ZERO_STAGE}" + ["MBS"]="${MICRO_BATCH}" + ["GAS"]="${GRAD_ACC_STEPS}" + ["GBS"]="${GLOBAL_BATCH}" + ["NLAYERS"]="${NLAYERS}" + ["HIDDEN"]="${HIDDEN}" + ["HEADS"]="${HEADS}" + ["SEQ"]="${SEQ}" + ["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE}" + ["NUM_KV_HEAD"]="${NUM_KV_HEAD}" + ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" + ["DTYPE"]="${DTYPE}" + ["OPT"]="${OPT}" + # ["WEIGHT_DECAY"]="${WEIGHT_DECAY}" + # ["ADAM_BETA1"]="${ADAM_BETA1}" + # ["ADAM_BETA2"]="${ADAM_BETA2}" + # ["ADAM_EPS"]="${ADAM_EPS}" + # ["CLIP_GRAD"]="${CLIP_GRAD:-1.0}" + ) + for v in "${!arch_map[@]}"; do + printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" + done + printf "======================\n" + # "--adam-eps ${ADAM_EPS:-0.00001}" + cache_dir="${PBS_O_WORKDIR}/.cache/" + mkdir -p "${cache_dir}" + targs_cache="${cache_dir}/train_args.txt" + for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done + export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") + printf "Training Arguments: %s\n" "$(printBlue "${TRAIN_ARGS[@]}")" + export run_cmd=("${LAUNCHER}" "${train_args[@]}") } save_dotenv() { - if [[ "$#" -ne 1 ]]; then - estr="[error]" - printf "%s Expected one argument (outdir). Received: %s" "$(printRed "${estr}")" "$#" - else - outdir="$1" - mkdir -p "${outdir}" - module list - dotenv_file="${outdir}/.env" - echo "Saving environment to ${dotenv_file}" - printenv | grep -v "LS_COLORS" >"${dotenv_file}" - export DOTENV_FILE="${dotenv_file}" - fi + if [[ "$#" -ne 1 ]]; then + estr="[error]" + printf "%s Expected one argument (outdir). Received: %s" "$(printRed "${estr}")" "$#" + else + outdir="$1" + mkdir -p "${outdir}" + module list + dotenv_file="${outdir}/.env" + echo "Saving environment to ${dotenv_file}" + printenv | grep -v "LS_COLORS" >"${dotenv_file}" + export DOTENV_FILE="${dotenv_file}" + fi } check_and_kill_if_running() { - RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') - if [[ -n "${RUNNING_PIDS}" ]]; then - echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}" - else - echo "Not currently running. Continuing!" - fi + RUNNING_PIDS=$(lsof -i:29500 -Fp | head -n 1 | sed 's/^p//') + if [[ -n "${RUNNING_PIDS}" ]]; then + echo "Caught ${RUNNING_PIDS}" && kill "${RUNNING_PIDS}" + else + echo "Not currently running. Continuing!" + fi } setupSrun() { - if [[ $(hostname) == login* || $(hostname) == nid* ]]; then - export NHOSTS="${SLURM_NNODES:-1}" - export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" - export NGPUS="$((NHOSTS * NGPU_PER_HOST))" - export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" - else - echo "Skipping setupSrun() on $(hostname)" - fi + if [[ $(hostname) == login* || $(hostname) == nid* ]]; then + export NHOSTS="${SLURM_NNODES:-1}" + export NGPU_PER_HOST="${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}" + export NGPUS="$((NHOSTS * NGPU_PER_HOST))" + export SRUN_EXEC="srun --gpus ${NGPUS} --gpus-per-node ${NGPU_PER_HOST} -N ${NHOSTS} -n ${NGPUS} -l -u --verbose" + else + echo "Skipping setupSrun() on $(hostname)" + fi } printJobInfo() { - echo "++++++++++++++++++++++++++++++++++++++++++++++++++" - echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" - echo "- Using $(which python3)" - echo "- WORLD_SIZE:${WORLD_SIZE-}" - echo "- BACKEND: ${BE:-}" - echo "- MODEL_TYPE: ${MODEL_TYPE:-}" - echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" - echo "++++++++++++++++++++++++++++++++++++++++++++++++++" + echo "++++++++++++++++++++++++++++++++++++++++++++++++++" + echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" + echo "- Using $(which python3)" + echo "- WORLD_SIZE:${WORLD_SIZE-}" + echo "- BACKEND: ${BE:-}" + echo "- MODEL_TYPE: ${MODEL_TYPE:-}" + echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" + echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } ############################################################################# @@ -344,46 +343,46 @@ printJobInfo() { # will launch with `deepspeed` instead of `mpiexec`. ############################################################################# setupLauncher() { - shell_type=$(basename "${SHELL}") - if [[ "${shell_type}" == "bash" ]]; then - shopt -s expand_aliases - fi - if [[ "$#" == 1 ]]; then - local dist_launcher="$1" - else - local dist_launcher="${LAUNCH_WITH:-${LAUNCH_CMD:-"MPICH"}}" - fi - if [[ "${dist_launcher}" == "deepspeed" ]]; then - # Save {PATH, LD_LIBRARY_PATH, ...} to .deepspeed_env - saveDSenv || exit - # Assert `./hostfile_deepspeed` exists - export hfds="${WORKING_DIR}/hostfile_deepspeed" - make_ds_hostfile || exit - export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" - else - LAUNCHER="ezpz-launch ${EXEC}" - # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job - # echo "ezpz_launch: $(which ezpz_launch)" - # export -f ezpz_launch - # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" - # if [[ -n "${DIST_LAUNCH}" ]]; then - # mn=$(ezpz_get_machine_name) - # if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then - # LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" - # elif [[ "${mn}" == "sophia" ]]; then - # LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" - # else - # LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" - # fi - # export LAUNCHER="${LAUNCHER}" - # else - # echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" - # printenv | tee ".env-${PBS_JOBID}" - # echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" - # fi - fi - printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" - printf " %s" "$(printMagenta "${LAUNCHER}")" + shell_type=$(basename "${SHELL}") + if [[ "${shell_type}" == "bash" ]]; then + shopt -s expand_aliases + fi + if [[ "$#" == 1 ]]; then + local dist_launcher="$1" + else + local dist_launcher="${LAUNCH_WITH:-${LAUNCH_CMD:-"MPICH"}}" + fi + if [[ "${dist_launcher}" == "deepspeed" ]]; then + # Save {PATH, LD_LIBRARY_PATH, ...} to .deepspeed_env + saveDSenv || exit + # Assert `./hostfile_deepspeed` exists + export hfds="${WORKING_DIR}/hostfile_deepspeed" + make_ds_hostfile || exit + export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" + else + LAUNCHER="ezpz-launch ${EXEC}" + # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job + # echo "ezpz_launch: $(which ezpz_launch)" + # export -f ezpz_launch + # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" + # if [[ -n "${DIST_LAUNCH}" ]]; then + # mn=$(ezpz_get_machine_name) + # if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then + # LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" + # elif [[ "${mn}" == "sophia" ]]; then + # LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" + # else + # LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" + # fi + # export LAUNCHER="${LAUNCHER}" + # else + # echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" + # printenv | tee ".env-${PBS_JOBID}" + # echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" + # fi + fi + printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" + printf " %s" "$(printMagenta "${LAUNCHER}")" } ######################################################################### @@ -401,30 +400,30 @@ setupLauncher() { # are the largest batch sizes that fit in memory at various node counts. ######################################################################### get_batch_size_on_polaris() { - if [[ $(hostname) == x3* ]]; then - nhosts=$(wc -l <"${HOSTFILE:-${PBS_NODEFILE}}") - if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then - mbs=1 - elif [[ "${nhosts}" -ge 3 && "${nhosts}" -le 7 ]]; then - mbs=2 - elif [[ "${nhosts}" -ge 8 ]]; then - mbs=4 - fi - fi - echo "${mbs}" + if [[ $(hostname) == x3* ]]; then + nhosts=$(wc -l <"${HOSTFILE:-${PBS_NODEFILE}}") + if [[ "${nhosts}" == 1 || "${nhosts}" == 2 ]]; then + mbs=1 + elif [[ "${nhosts}" -ge 3 && "${nhosts}" -le 7 ]]; then + mbs=2 + elif [[ "${nhosts}" -ge 8 ]]; then + mbs=4 + fi + fi + echo "${mbs}" } _get_num_hosts_from_hostfile() { - if [[ "$#" == 1 ]]; then - if [[ -f "$1" ]]; then - nhosts=$(wc -l <"$1") - echo "${nhosts}" - else - exit 1 - fi - else - exit 1 - fi + if [[ "$#" == 1 ]]; then + if [[ -f "$1" ]]; then + nhosts=$(wc -l <"$1") + echo "${nhosts}" + else + exit 1 + fi + else + exit 1 + fi } ########################################### @@ -449,131 +448,129 @@ _get_num_hosts_from_hostfile() { # ########################################### get_grad_acc_steps_on_aurora() { - if [[ "$#" == 0 ]]; then - hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" - elif [[ "$#" == 1 ]]; then - hf="$1" - else - echo "Usage: get_grad_acc_steps_on_aurora" - echo "Expected exactly 0 or 1 arguments, received: $#" - exit 1 - fi - nhosts=$(wc -l <"${hf}") - if [[ "${nhosts}" -ge 256 ]]; then # n >= 256 - gas=1 - elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then # 128 <= n < 256 - gas=2 - elif [[ 32 -lt "${nhosts}" && "${nhosts}" -lt 129 ]]; then # 32 < n < 128 - gas=4 - elif [[ 16 -le "${nhosts}" && "${nhosts}" -le 32 ]]; then # 16 <= n < 32 - gas=8 - else - gas=16 - fi - echo "${gas}" + if [[ "$#" == 0 ]]; then + hf="${HOSTFILE:-${PBS_NODEFILE:-$(ezpz_get_pbs_nodefile_from_hostname)}}" + elif [[ "$#" == 1 ]]; then + hf="$1" + else + echo "Usage: get_grad_acc_steps_on_aurora" + echo "Expected exactly 0 or 1 arguments, received: $#" + exit 1 + fi + nhosts=$(wc -l <"${hf}") + if [[ "${nhosts}" -ge 256 ]]; then # n >= 256 + gas=1 + elif [[ 128 -le "${nhosts}" && "${nhosts}" -lt 256 ]]; then # 128 <= n < 256 + gas=2 + elif [[ 32 -lt "${nhosts}" && "${nhosts}" -lt 129 ]]; then # 32 < n < 128 + gas=4 + elif [[ 16 -le "${nhosts}" && "${nhosts}" -le 32 ]]; then # 16 <= n < 32 + gas=8 + else + gas=16 + fi + echo "${gas}" } set_ccl_vars_on_aurora() { - export CCL_KVS_MODE=mpi - export CCL_CONFIGURATION_PATH="" - export CCL_CONFIGURATION=cpu_gpu_dpcpp - # export CCL_ROOT=/tmp/oneccl/ - # export LD_LIBRARY_PATH=${CCL_ROOT}/lib:$LD_LIBRARY_PATH - # export CPATH=${CCL_ROOT}/include:$CPATH - # export LIBRARY_PATH=${CCL_ROOT}/lib:$LIBRARY_PATH - export CCL_KVS_CONNECTION_TIMEOUT=3600 - export FI_CXI_RX_MATCH_MODE=hybrid - export CCL_BCAST=double_tree - - export ZE_ENABLE_PCI_ID_DEVICE_ORDER=1 - export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich - export FI_PROVIDER=cxi # Required by Aurora mpich - export PALS_PMI=pmix # Required by Aurora mpich - # export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich - export CCL_ATL_TRANSPORT=ofi # [SF]: Changed 04/30/2025 - export TORCH_LLM_ALLREDUCE=1 - export CCL_SYCL_ESIMD=1 - export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) - export CCL_ENABLE_SYCL_KERNELS=1 - export CCL_WORKER_AFFINITY=5,13,21,29,37,45,57,65,73,81,89,97 - export CCL_ZE_CACHE_OPEN_IPC_HANDLES_THRESHOLD=32768 - export FI_CXI_DEFAULT_CQ_SIZE=1048576 - export FI_CXI_RX_MATCH_MODE=hybrid - export CCL_BCAST=double_tree + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export CCL_CONFIGURATION=cpu_gpu_dpcpp + # export CCL_ROOT=/tmp/oneccl/ + # export LD_LIBRARY_PATH=${CCL_ROOT}/lib:$LD_LIBRARY_PATH + # export CPATH=${CCL_ROOT}/include:$CPATH + # export LIBRARY_PATH=${CCL_ROOT}/lib:$LIBRARY_PATH + export CCL_KVS_CONNECTION_TIMEOUT=3600 + export FI_CXI_RX_MATCH_MODE=hybrid + export CCL_BCAST=double_tree + + export ZE_ENABLE_PCI_ID_DEVICE_ORDER=1 + export CCL_PROCESS_LAUNCHER=pmix # Required by Aurora mpich + export FI_PROVIDER=cxi # Required by Aurora mpich + export PALS_PMI=pmix # Required by Aurora mpich + # export CCL_ATL_TRANSPORT=mpi # Required by Aurora mpich + export CCL_ATL_TRANSPORT=ofi # [SF]: Changed 04/30/2025 + export TORCH_LLM_ALLREDUCE=1 + export CCL_SYCL_ESIMD=1 + export CCL_ALLGATHERV_MEDIUM_SIZE_THRESHOLD=0 # Required by current oneCCL (MLSL-2881) + export CCL_ENABLE_SYCL_KERNELS=1 + export CCL_WORKER_AFFINITY=5,13,21,29,37,45,57,65,73,81,89,97 + export CCL_ZE_CACHE_OPEN_IPC_HANDLES_THRESHOLD=32768 + export FI_CXI_DEFAULT_CQ_SIZE=1048576 + export FI_CXI_RX_MATCH_MODE=hybrid + export CCL_BCAST=double_tree } - get_model_arch_7B() { - # 7B - # export MODEL_KEY="AuroraGPT-7B" - export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS - export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS - export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE - export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION - export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export MODEL_ARCH="AuroraGPT-7B" + # 7B + # export MODEL_KEY="AuroraGPT-7B" + export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export MODEL_ARCH="AuroraGPT-7B" } get_model_arch_llama3_3B() { - export HEADS=24 - export NLAYERS=28 - export HIDDEN=3072 - export NUM_KV_HEAD=8 - export FFN_HIDDEN_SIZE=8192 - export SEQ=8192 - export MODEL_ARCH="llama3-3B" + export HEADS=24 + export NLAYERS=28 + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="llama3-3B" } get_model_arch_smollm3_3B() { - export HEADS=16 - export NLAYERS=36 - export HIDDEN=2048 - export NUM_KV_HEAD=4 - export FFN_HIDDEN_SIZE=11008 - export SEQ=8192 - export MODEL_ARCH="smollm3-3B" + export HEADS=16 + export NLAYERS=36 + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="smollm3-3B" } get_model_arch_phi4_mini() { - export HEADS=32 - export NLAYERS=24 - export HIDDEN=3072 - export NUM_KV_HEAD=8 - export FFN_HIDDEN_SIZE=8192 - export SEQ=8192 - export MODEL_ARCH="phi4-mini" + export HEADS=32 + export NLAYERS=24 + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="phi4-mini" } get_model_arch_llama3_3B_customNlayers() { - export HEADS=24 - export NLAYERS="${NLAYERS:-28}" # default to 28 layers - export HIDDEN=3072 - export NUM_KV_HEAD=8 - export FFN_HIDDEN_SIZE=8192 - export SEQ=8192 - export MODEL_ARCH="llama3-3B-nLayers${NLAYERS}" + export HEADS=24 + export NLAYERS="${NLAYERS:-28}" # default to 28 layers + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="llama3-3B-nLayers${NLAYERS}" } - get_model_arch_smollm3_3B_custom_nLayers() { - export HEADS=16 - export NLAYERS="${NLAYERS:-24}" # default to 24 layers - export HIDDEN=2048 - export NUM_KV_HEAD=4 - export FFN_HIDDEN_SIZE=11008 - export SEQ=8192 - export MODEL_ARCH="smollm3-nLayers${NLAYERS}" + export HEADS=16 + export NLAYERS="${NLAYERS:-24}" # default to 24 layers + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="smollm3-nLayers${NLAYERS}" } get_model_arch_phi4_mini_custom_nLayers() { - export HEADS=32 - export NLAYERS="${NLAYERS:-24}" # default to 24 layers - export HIDDEN=3072 - export NUM_KV_HEAD=8 - export FFN_HIDDEN_SIZE=8192 - export SEQ=8192 - export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" + export HEADS=32 + export NLAYERS="${NLAYERS:-24}" # default to 24 layers + export HIDDEN=3072 + export NUM_KV_HEAD=8 + export FFN_HIDDEN_SIZE=8192 + export SEQ=8192 + export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" } # get_model_arch_70B() { @@ -593,30 +590,30 @@ get_model_arch_phi4_mini_custom_nLayers() { # } # get_model_arch_70B() { - HEADS=64 - NLAYERS=80 - NUM_KV_HEAD=8 - FFN_HIDDEN_SIZE=28672 - HIDDEN=8192 - SEQ_LEN=8192 - export MODEL_ARCH="AuroraGPT-70B" + HEADS=64 + NLAYERS=80 + NUM_KV_HEAD=8 + FFN_HIDDEN_SIZE=28672 + HIDDEN=8192 + SEQ=8192 + export MODEL_ARCH="AuroraGPT-70B" } get_model_arch_33B() { - # 33B - export MODEL_KEY="AuroraGPT-33B" - export TP=6 - export ZERO_STAGE=2 - export GRAD_ACC_STEPS=1 - # export USE_ACTIVATION_CHECKPOINTING=1 - export MICRO_BATCH=1 - export NLAYERS=78 - export HIDDEN=8268 - export HEADS=78 - export FFN_HIDDEN_SIZE=11076 - export SEQ=4096 - export NUM_KV_HEAD=6 - export MODEL_ARCH="AuroraGPT-33B" + # 33B + export MODEL_KEY="AuroraGPT-33B" + export TP=6 + export ZERO_STAGE=2 + export GRAD_ACC_STEPS=1 + # export USE_ACTIVATION_CHECKPOINTING=1 + export MICRO_BATCH=1 + export NLAYERS=78 + export HIDDEN=8268 + export HEADS=78 + export FFN_HIDDEN_SIZE=11076 + export SEQ=4096 + export NUM_KV_HEAD=6 + export MODEL_ARCH="AuroraGPT-33B" } ############################################################################## @@ -631,179 +628,181 @@ get_model_arch_33B() { # ``` ############################################################################## setParams() { - FLASH_ARG="" - # ---- [Parallelism Settings] -------------------------------------------+ - # ------ [Aurora] -------||------ [SunSpot] ------------- - # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then - GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH - mn=$(ezpz_get_machine_name) - if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then - TP=${TP:-1} # TP = 1 - export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" - export CCL=${CCL:-ccl} # CCL - export BE="${CCL}" # COMMUNICATION BACKEND = CCL - export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 - # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS - gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") - export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" - # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" - echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - MICRO_BATCH=${MICRO_BATCH:-1} - if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" - else - FLASH_ARG="--use-flash-attn-builder" - fi - # [Polaris] - elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then - # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL - # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? - export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS - # NOTE: MICRO_BATCH is exported below - # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 - export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" - if [[ -n "${NO_FLASH_ATTN:-}" ]]; then - echo "Not using flash-attn!!" - else - FLASH_ARG="--use-flash-attn-v2" - fi - echo "Setting up AWS NCCL OFI Plugin on Polaris..." - source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit - # ---- [Sophia] ---------------------- - elif [[ "${mn}" == sophia* ]]; then - # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" - TP=${TP:-1} # TP = 2 - export NCCL=${NCCL:-nccl} # NCCL - export BE="${NCCL}" # BE = NCCL - export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS - export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" - if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" - else - FLASH_ARG="--use-flash-attn-v2" - fi - # echo "Setting up AWS NCCL OFI Plugin on Polaris..." - # source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit - # [Perlmutter] - elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then - TP="${TP:-2}" - export NCCL="${NCCL:-nccl}" - export BE="${NCCL}" - export DTYPE="${DTYPE:-bf16}" - MICRO_BATCH="${MICRO_BATCH:-1}" - if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" - else - FLASH_ARG="--use-flash-attn-v2" - fi - fi - ma="${MODEL_ARCH:-7B}" - case "${ma}" in - # "70B" | "llama-3.1-70B" | "llama-3.1-70b" | "llama-3.2-70B" | "llama-3.2-70b") - "70B") - get_model_arch_70B - ;; - "33B" | "llama-3.2-33B" | "llama-3.2-33b") - get_model_arch_33B - ;; - "smollm3-3B" | "smollm3_3B") - get_model_arch_smollm3_3B_custom_nLayers - ;; - "phi4-mini" | "phi4_mini") - get_model_arch_phi4_mini_custom_nLayers - ;; - "llama3-3B" | "llama-3B") - get_model_arch_llama3_3B_customNlayers - ;; - "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") - get_model_arch_7B - ;; - *) - get_model_arch_7B - ;; - esac - export TP="${TP}" - export PP="${PP:-1}" - export SP="${SP:-1}" - export FLASH_ARG="${FLASH_ARG}" - export DTYPE="${DTYPE:-bf16}" - export OPT="${OPT:-adamw}" - export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" - export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" - NHOSTS=$(wc -l <"${HOSTFILE}") - if [[ -z "${NGPU_PER_HOST:-}" ]]; then - NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') - fi - export NGPU_PER_HOST="${NGPU_PER_HOST}" - export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" - # if [[ "${WORLD_SIZE}" -gt 1 && "${mn}" == "aurora" ]]; then - # #### [sam: 08/17/2024] ########################################## - # # Use best set of CCL env vars from Gordon Bell runs on Aurora - # set_ccl_vars_on_aurora - # fi - # + --[LR Settings]------------------------------------------------------+ - export LR="${LR:-0.0002}" - export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" - export LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" - lr_flags=( - "--lr ${LR}" - "--lr-decay-style ${LR_DECAY_STYLE}" - "--lr-warmup-fraction ${LR_WARMUP_FRAC}" - ) - if [[ -n "${LR_DECAY_ITERS:-}" ]]; then - lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") - fi - # +---[Run Settings]------------------------------------------------------+ - export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS - export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS - export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? - export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE - export DP=$((WORLD_SIZE / TP / PP / SP)) # DATA PARALLELISM - export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" - if [[ -n "${TRAIN_TOKENS:-}" ]]; then - export TRAIN_TOKENS="${TRAIN_TOKENS}" - export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) - printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" - elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then - export TRAIN_TOKENS=${TRAIN_TOKENS:-4673780159710} - export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) - printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" - else - export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" - fi - export MODEL_TYPE="${MODEL_ARCH:-AuroraGPT}-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL - # NOTE: [2024-07-10] ##################################################### - # - [sam]: For whatever reason, it seems that using - # sequence-parallelism (SP) > 1 is INCOMPATIBLE with - # rotary-position-embeddings (ROPE). - # - # For this reason, we only use the default LLAMA_ARGS when SP=0. - ########################################################################## - # # -----[Learning Rate Settings]-------------------------------------------- - # export LR=${LR:-0.0002} # LEARNING_RATE - # export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP - # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS - # set_lr_args - # -----[Learning Rate Settings]-------------------------------------------- - # # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then - # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then - # TIMING_STR="\ - # --timing-log-level ${TIMING_LOG_LEVEL}" - # # " - # else - # TIMING_STR="" - # fi + FLASH_ARG="" + # ---- [Parallelism Settings] -------------------------------------------+ + # ------ [Aurora] -------||------ [SunSpot] ------------- + # if [[ $(hostname) == x4* || $(hostname) == x1* ]]; then + GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH + mn=$(ezpz_get_machine_name) + if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then + TP=${TP:-1} # TP = 1 + export SAVE_INTERVAL="${SAVE_INTERVAL:-50}" + export CCL=${CCL:-ccl} # CCL + export BE="${CCL}" # COMMUNICATION BACKEND = CCL + export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 + # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS + gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") + export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" + # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" + echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" + MICRO_BATCH=${MICRO_BATCH:-1} + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-builder" + fi + # [Polaris] + elif [[ "${mn}" == "polaris" || "${mn}" == "sirius" ]]; then + # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL + # export DTYPE=${DTYPE:-bf16} # DTYPE: BF16 ?? + export DTYPE=${DTYPE:-fp16} # DTYPE: FP16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + # NOTE: MICRO_BATCH is exported below + # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 + export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" + if [[ -n "${NO_FLASH_ATTN:-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-v2" + fi + echo "Setting up AWS NCCL OFI Plugin on Polaris..." + source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit + # ---- [Sophia] ---------------------- + elif [[ "${mn}" == sophia* ]]; then + # export LAUNCH_CMD="${LAUNCH_CMD:-deepspeed}" + TP=${TP:-1} # TP = 2 + export NCCL=${NCCL:-nccl} # NCCL + export BE="${NCCL}" # BE = NCCL + export DTYPE=${DTYPE:-bf16} # DTYPE: FP16 + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS + export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-v2" + fi + # echo "Setting up AWS NCCL OFI Plugin on Polaris..." + # source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit + # [Perlmutter] + elif [[ "${mn}" == login* || "${mn}" == nid* ]]; then + TP="${TP:-2}" + export NCCL="${NCCL:-nccl}" + export BE="${NCCL}" + export DTYPE="${DTYPE:-bf16}" + MICRO_BATCH="${MICRO_BATCH:-1}" + if [[ -n "${NO_FLASH_ATTN-}" ]]; then + echo "Not using flash-attn!!" + else + FLASH_ARG="--use-flash-attn-v2" + fi + fi + ma="${MODEL_ARCH:-7B}" + case "${ma}" in + # "70B" | "llama-3.1-70B" | "llama-3.1-70b" | "llama-3.2-70B" | "llama-3.2-70b") + "70B") + get_model_arch_70B + ;; + "33B" | "llama-3.2-33B" | "llama-3.2-33b") + get_model_arch_33B + ;; + "smollm3-3B" | "smollm3_3B") + get_model_arch_smollm3_3B_custom_nLayers + ;; + "phi4-mini" | "phi4_mini") + get_model_arch_phi4_mini_custom_nLayers + ;; + "llama3-3B" | "llama-3B") + get_model_arch_llama3_3B_customNlayers + ;; + "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") + get_model_arch_7B + ;; + *) + get_model_arch_7B + ;; + esac + export TP="${TP}" + export PP="${PP:-1}" + export SP="${SP:-1}" + export FLASH_ARG="${FLASH_ARG}" + export DTYPE="${DTYPE:-bf16}" + OPT="${OPT:-${OPTIMIZER:-${OPTIM:-adamw}}}" + export OPT + log_message INFO "❗ Using OPT: ${OPT}" + export WEIGHT_DECAY="${WEIGHT_DECAY:-0.1}" + export HOSTFILE="${HOSTFILE:-${PBS_NODEFILE}}" + NHOSTS=$(wc -l <"${HOSTFILE}") + if [[ -z "${NGPU_PER_HOST:-}" ]]; then + NGPU_PER_HOST=$(python3 -c 'import ezpz as ez; print(ez.get_gpus_per_node())') + fi + export NGPU_PER_HOST="${NGPU_PER_HOST}" + export WORLD_SIZE="${WORLD_SIZE:-$((NHOSTS * NGPU_PER_HOST))}" + # if [[ "${WORLD_SIZE}" -gt 1 && "${mn}" == "aurora" ]]; then + # #### [sam: 08/17/2024] ########################################## + # # Use best set of CCL env vars from Gordon Bell runs on Aurora + # set_ccl_vars_on_aurora + # fi + # + --[LR Settings]------------------------------------------------------+ + export LR="${LR:-0.0002}" + export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" + export LR_WARMUP_FRAC="${LR_WARMUP_FRAC:-0.05}" + lr_flags=( + "--lr ${LR}" + "--lr-decay-style ${LR_DECAY_STYLE}" + "--lr-warmup-fraction ${LR_WARMUP_FRAC}" + ) + if [[ -n "${LR_DECAY_ITERS:-}" ]]; then + lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") + fi + # +---[Run Settings]------------------------------------------------------+ + export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE + export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS + export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS + export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? + export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE + export DP=$((WORLD_SIZE / TP / PP / SP)) # DATA PARALLELISM + export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT + echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" + if [[ -n "${TRAIN_TOKENS:-}" ]]; then + export TRAIN_TOKENS="${TRAIN_TOKENS}" + export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then + export TRAIN_TOKENS=${TRAIN_TOKENS:-4673780159710} + export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + else + export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" + fi + export MODEL_TYPE="${MODEL_ARCH:-AuroraGPT}-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL + # NOTE: [2024-07-10] ##################################################### + # - [sam]: For whatever reason, it seems that using + # sequence-parallelism (SP) > 1 is INCOMPATIBLE with + # rotary-position-embeddings (ROPE). + # + # For this reason, we only use the default LLAMA_ARGS when SP=0. + ########################################################################## + # # -----[Learning Rate Settings]-------------------------------------------- + # export LR=${LR:-0.0002} # LEARNING_RATE + # export LR_WARMUP_FRAC=${LR_WARMUP_FRAC:-0.05} # LEARNING RATE WARMUP + # export LR_DECAY_ITERS=${LR_DECAY_ITERS:-} # LR DECAY ITERS + # set_lr_args + # -----[Learning Rate Settings]-------------------------------------------- + # # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then + # if [[ "${TIMING_LOG_LEVEL:-1}" -gt 1 ]]; then + # TIMING_STR="\ + # --timing-log-level ${TIMING_LOG_LEVEL}" + # # " + # else + # TIMING_STR="" + # fi } ############################################## @@ -813,65 +812,53 @@ setParams() { # arguments to pass to pretrain_gpt_alcf.py ############################################## set_args() { - # ---- Set DeepSpeed arguments -------------------------------- - ds_args=( - "--deepspeed" - ) - if [[ "${PP:-1}" == 1 ]]; then - ds_args+=("--no-pipeline-parallel") - fi - ds_args+=("--deepspeed_config=${DS_CONFIG}") - ds_args+=("--zero-stage=$ZERO_STAGE") - # if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then - if [[ "${USE_ACTIVATION_CHECKPOINTING:-}" == 1 || "${USE_ACTIVATION_CHECKPOINTING:-}" == "true" ]]; then - echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - ds_args+=("--deepspeed-activation-checkpointing") - ds_args+=( - "--checkpoint-activations" - "--checkpoint-num-layers=${ACT_CKPT_NUM_LAYERS:-1}" - ) - # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" - # --checkpoint-activations \ - # --deepspeed-activation-checkpointing - fi - export ds_args - # --------------------------------------------------------------- - # gpt_args=() - # # we are now using activation checkpoint provided by megatron, see below. - # # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" - # if [[ "$USE_ACTIVATION_CHECKPOINTING" == 1 ]]; then - # echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" - # gpt_args+=( - # "--checkpoint-activations" - # "--checkpoint-num-layers ${ACT_CKPT_NUM_LAYERS}" - # ) - # fi - # export gpt_args + # ---- Set DeepSpeed arguments -------------------------------- + ds_args=( + "--deepspeed" + ) + if [[ "${PP:-1}" == 1 ]]; then + ds_args+=("--no-pipeline-parallel") + fi + ds_args+=("--deepspeed_config=${DS_CONFIG}") + ds_args+=("--zero-stage=$ZERO_STAGE") + # if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then + if [[ "${USE_ACTIVATION_CHECKPOINTING:-}" == 1 || "${USE_ACTIVATION_CHECKPOINTING:-}" == "true" ]]; then + echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" + ds_args+=("--deepspeed-activation-checkpointing") + ds_args+=( + "--checkpoint-activations" + "--checkpoint-num-layers=${ACT_CKPT_NUM_LAYERS:-1}" + ) + # ds_args=" --deepspeed-activation-checkpointing ${ds_args}" + # --checkpoint-activations \ + # --deepspeed-activation-checkpointing + fi + export ds_args } make_ds_hostfile() { - export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" - # ---- Make MPICH hostfile ---------------- - hf="${HOSTFILE:-${PBS_NODEFILE}}" - export hostfile_mpich=hostfile_mpich - cat "${hf}" >"${hostfile_mpich}" - # ---- Make DeepSpeed hostfile ------------------- - export hostfile_deepspeed=hostfile_deepspeed - cat "${hf}" >"${hostfile_deepspeed}" - sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" + export GPUS_PER_NODE="${GPUS_PER_NODE:-${NGPU_PER_HOST:-${SLURM_GPUS_ON_NODE:-$(nvidia-smi -L | wc -l)}}}" + # ---- Make MPICH hostfile ---------------- + hf="${HOSTFILE:-${PBS_NODEFILE}}" + export hostfile_mpich=hostfile_mpich + cat "${hf}" >"${hostfile_mpich}" + # ---- Make DeepSpeed hostfile ------------------- + export hostfile_deepspeed=hostfile_deepspeed + cat "${hf}" >"${hostfile_deepspeed}" + sed -e "s/$/ slots=${GPUS_PER_NODE}/" -i "${hostfile_deepspeed}" } ####################################################################### # ezpz_test: Run simple test to make sure all nodes in working order ####################################################################### ezpz_test() { - printf "%s" "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" - # [ -n "${PBS_O_WORKIR}" ] && ezpz_savejobenv || ezpz_getjobenv - # python3 -Wignore -m ezpz.jobs && source "${PBS_O_WORKDIR}/.jobenv" - printf "%s" "[$(printBlue 'ezpz:test_dist')] Running test: ${eztest}\n" - eztest="TRAIN_ITERS=50 ${LAUNCH_CMD} python3 -Wignore -m ezpz.test_dist" - eval "${eztest}" - printf "%s" "[$(printBlue 'ezpz:test_dist')] Done with test!\n" + printf "%s" "[$(printBlue 'ezpz:test_dist')][INFO] Running ezpz.test_dist...\n" + # [ -n "${PBS_O_WORKIR}" ] && ezpz_savejobenv || ezpz_getjobenv + # python3 -Wignore -m ezpz.jobs && source "${PBS_O_WORKDIR}/.jobenv" + printf "%s" "[$(printBlue 'ezpz:test_dist')] Running test: ${eztest}\n" + eztest="TRAIN_ITERS=50 ${LAUNCH_CMD} python3 -Wignore -m ezpz.test_dist" + eval "${eztest}" + printf "%s" "[$(printBlue 'ezpz:test_dist')] Done with test!\n" } ############################################################################ @@ -881,84 +868,84 @@ ezpz_test() { # forwarded to ALL ranks with DeepSpeed ############################################################################ saveDSenv() { - echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" - { - echo "PATH=${PATH:-}" - echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH:-}" - echo "http_proxy=${http_proxy:-}" - echo "https_proxy=${https_proxy:-}" - echo "CFLAGS=${CFLAGS:-}" - echo "PYTHONUSERBASE=${PYTHONUSERBASE:-}" - } >.deepspeed_env + echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" + { + echo "PATH=${PATH:-}" + echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH:-}" + echo "http_proxy=${http_proxy:-}" + echo "https_proxy=${https_proxy:-}" + echo "CFLAGS=${CFLAGS:-}" + echo "PYTHONUSERBASE=${PYTHONUSERBASE:-}" + } >.deepspeed_env } get_output_prefix() { - # ---- Specify output location -------------------------------- - pre="ws${WORLD_SIZE}-ds-stage${ZERO_STAGE}-nl${NLAYERS}" - pre="${pre}-hs${HIDDEN}-mb${MICRO_BATCH}" - pre="${pre}-seq${SEQ}-gb${GLOBAL_BATCH}" - pre="${pre}-sp${SP}-pp${PP}-tp${TP}-${DTYPE}-opt${OPT}" - pre="${pre}-lr${LR}-lwf${LR_WARMUP_FRAC}" - pre="${MODEL_ARCH:-AuroraGPT}-${pre}" - local num_tokens_in_billions - num_tokens_in_billions=$((TRAIN_TOKENS / 10 ** 9)) - pre="${pre}_ntok${num_tokens_in_billions}B" - if [[ -n "${TOKENIZER_TYPE:-}" ]]; then - # _tok="${TOKENIZER_TYPE/Tokenizer//}" # Strip "Tokenizer" suffix if present - _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa - pre="${pre}_tok${_tok}" - fi - if [[ -n "${TOKENIZER_MODEL:-}" ]]; then - # _tm=$(echo "${TOKENIZER_MODEL}" | sed 's/\/_/g') # noqa - # replace slashes with underscores - _tm="${TOKENIZER_MODEL//\//_}" # noqa - pre="${pre}_tm${_tm}" - fi - if [[ -n "${LR_DECAY_ITERS}" ]]; then - pre="${pre}_ldi${LR_DECAY_ITERS}" - fi - if [[ -z "${NO_FLASH_ATTN:-}" ]]; then - pre="${pre}_flash" - fi - export OUTPUT_PREFIX="${pre}" - echo "${pre}" + # ---- Specify output location -------------------------------- + pre="ws${WORLD_SIZE}-ds-stage${ZERO_STAGE}-nl${NLAYERS}" + pre="${pre}-hs${HIDDEN}-mb${MICRO_BATCH}" + pre="${pre}-seq${SEQ}-gb${GLOBAL_BATCH}" + pre="${pre}-sp${SP}-pp${PP}-tp${TP}-${DTYPE}-opt${OPT}" + pre="${pre}-lr${LR}-lwf${LR_WARMUP_FRAC}" + pre="${MODEL_ARCH:-AuroraGPT}-${pre}" + local num_tokens_in_billions + num_tokens_in_billions=$((TRAIN_TOKENS / 10 ** 9)) + pre="${pre}_ntok${num_tokens_in_billions}B" + if [[ -n "${TOKENIZER_TYPE:-}" ]]; then + # _tok="${TOKENIZER_TYPE/Tokenizer//}" # Strip "Tokenizer" suffix if present + _tok=$(echo "${TOKENIZER_TYPE}" | sed 's/Tokenizer//g') # noqa + pre="${pre}_tok${_tok}" + fi + if [[ -n "${TOKENIZER_MODEL:-}" ]]; then + # _tm=$(echo "${TOKENIZER_MODEL}" | sed 's/\/_/g') # noqa + # replace slashes with underscores + _tm="${TOKENIZER_MODEL//\//_}" # noqa + pre="${pre}_tm${_tm}" + fi + if [[ -n "${LR_DECAY_ITERS}" ]]; then + pre="${pre}_ldi${LR_DECAY_ITERS}" + fi + if [[ -z "${NO_FLASH_ATTN:-}" ]]; then + pre="${pre}_flash" + fi + export OUTPUT_PREFIX="${pre}" + echo "${pre}" } setOutput() { - # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" - OUTPUT_PREFIX=$(get_output_prefix) - OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" - export OUTPUT_DIR="${OUTPUT_DIR}" && mkdir -p "${OUTPUT_DIR}" - export OUTPUT_LOG="${OUTPUT_DIR}/output.log" - echo "${OUTPUT_LOG}" >>"logs/latest" - printf "\n Please see logs at: %s\n" "$(printGreen "${OUTPUT_DIR}")" + # OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%m%d%H%M%S)_${HOSTNAME}" + OUTPUT_PREFIX=$(get_output_prefix) + OUTPUT_DIR="logs/${OUTPUT_PREFIX}/$(date +%Y%m%d-%H%M%S)_${WORLD_SIZE}_${HOSTNAME}" + export OUTPUT_DIR="${OUTPUT_DIR}" && mkdir -p "${OUTPUT_DIR}" + export OUTPUT_LOG="${OUTPUT_DIR}/output.log" + echo "${OUTPUT_LOG}" >>"logs/latest" + printf "\n Please see logs at: %s\n" "$(printGreen "${OUTPUT_DIR}")" } get_checkpoint_dir() { - if [[ -n "${CKPT_DIR:-}" ]]; then - echo "${CKPT_DIR}" - else - echo "checkpoints/$(get_output_prefix)" - fi + if [[ -n "${CKPT_DIR:-}" ]]; then + echo "${CKPT_DIR}" + else + echo "checkpoints/$(get_output_prefix)" + fi } setup_checkpoint() { - ckpt_dir=$(get_checkpoint_dir) - export CKPT_DIR="${ckpt_dir}" - printf "Checkpoints will be saved to: %s\n" "$(printYellow "${CKPT_DIR}")" + ckpt_dir=$(get_checkpoint_dir) + export CKPT_DIR="${ckpt_dir}" + printf "Checkpoints will be saved to: %s\n" "$(printYellow "${CKPT_DIR}")" } ############################################# # Build DeepSpeed config and write to .json ############################################# buildDSconfig() { - # export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" - export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" - mkdir -p "$(dirname "${DS_CONFIG}")" - printf "DS_CONFIG: %s\n" "${DS_CONFIG}" - printf "ZS=%s, MB=%s, GB=%s, PP=%s, DTYPE=%s\n" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" - generateDSconfig "${DS_CONFIG}" - cat "${DS_CONFIG}" | jq . + # export CPU_OPTIMIZER="${CPU_OPTIMIZER:-0}" + export DS_CONFIG="${WORKING_DIR}/ds-configs/ds_stage${ZERO_STAGE}_mb${MICRO_BATCH}_gb${GLOBAL_BATCH}_pp${PP}_${DTYPE}.json" + mkdir -p "$(dirname "${DS_CONFIG}")" + printf "DS_CONFIG: %s\n" "${DS_CONFIG}" + printf "ZS=%s, MB=%s, GB=%s, PP=%s, DTYPE=%s\n" "${ZERO_STAGE}" "${MICRO_BATCH}" "${GLOBAL_BATCH}" "${PP}" "${DTYPE}" + generateDSconfig "${DS_CONFIG}" + cat "${DS_CONFIG}" | jq . } ############################################################################### @@ -968,17 +955,17 @@ buildDSconfig() { # `file_list`. ############################################################################### sumWeights() { - local file_list=$1 - weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") - python3 -c "import numpy as np; print(np.sum(${weights}))" + local file_list=$1 + weights=$(cat "${file_list}" | awk '{print $1}' | tr '\n' '\ ,\ ' | sed 's/^/[/g' | sed 's/$/]/g' | tr '\ ' "\,\ ") + python3 -c "import numpy as np; print(np.sum(${weights}))" } sumFiles() { - local rd=$1 - for f in $("${rd}/*.txt"); do - ws=$(sumWeights "${rd}/${f}") - echo "sum($f.weights)=${ws}" - done + local rd=$1 + for f in $("${rd}/*.txt"); do + ws=$(sumWeights "${rd}/${f}") + echo "sum($f.weights)=${ws}" + done } ########################################### @@ -990,9 +977,9 @@ sumFiles() { # is built appropriately. ########################################### make_data() { - python3 -m pip install pybind11 - mdir="${WORKING_DIR}/megatron/data" - cd "${mdir}" && make && cd - + python3 -m pip install pybind11 + mdir="${WORKING_DIR}/megatron/data" + cd "${mdir}" && make && cd - } ############################################################################## @@ -1001,79 +988,79 @@ make_data() { # Ensure all dependencies installed from `ALCF/requirements/requirements.txt` ############################################################################## install_dependencies() { - depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" - echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." - python3 -m pip install -r "${depsfile}" --require-virtualenv - if [[ ! -x "$(command -v deepspeed)" ]]; then - printf "[install_dependencies] No 'deepspeed' command found on %s in %s\n" "$$(ezpz_get_machine_name)" "$(which python3)" - printf "[install_dependencies] Attempting to install deepspeed via pip...\n" - python3 -m pip install deepspeed --require-virtualenv || { - printf "[install_dependencies] Failed to install deepspeed via pip on %s\n" "$(ezpz_get_machine_name)" - # printf "[install_dependencies] !! No deepsepeed in %s\n" "$(which python3)" - return 1 - } - # mn=$(ezpz_get_machine_name) - # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then - # install_deepspeed_for_xpu || exit - # fi - # printf "[install_dependencies] !! No deepsepeed in %s" "$(which python3)" - fi + depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" + echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." + python3 -m pip install -r "${depsfile}" --require-virtualenv + if [[ ! -x "$(command -v deepspeed)" ]]; then + printf "[install_dependencies] No 'deepspeed' command found on %s in %s\n" "$$(ezpz_get_machine_name)" "$(which python3)" + printf "[install_dependencies] Attempting to install deepspeed via pip...\n" + python3 -m pip install deepspeed --require-virtualenv || { + printf "[install_dependencies] Failed to install deepspeed via pip on %s\n" "$(ezpz_get_machine_name)" + # printf "[install_dependencies] !! No deepsepeed in %s\n" "$(which python3)" + return 1 + } + # mn=$(ezpz_get_machine_name) + # if [[ "${mn}" == aurora* || "${mn}" == sunspot* ]]; then + # install_deepspeed_for_xpu || exit + # fi + # printf "[install_dependencies] !! No deepsepeed in %s" "$(which python3)" + fi } ################################################# # Fix for distributed key value store on Aurora ################################################# use_kvs_fix_on_aurora() { - export CCL_KVS_MODE=mpi - export CCL_CONFIGURATION_PATH="" - export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH - ######################################################### - # if not set, CCL will complain... ? - export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" - ######################################################### + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + ######################################################### + # if not set, CCL will complain... ? + export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" + ######################################################### } update_ccl_env_vars_aurora() { - # export CCL_KVS_MODE=mpi - # # export CCL_CONFIGURATION_PATH="" - # # unset CCL_CONFIGURATION_PATH - # # export CCL_CONFIGURATION=cpu_gpu_dpcpp - # # export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" - # export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - # export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - # export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH - # # export CCL_ALLREDUCE_SCALEOUT=direct - # printenv | grep -E -v "^__" | grep -E "CCL|LD|CPATH|LIBRARY_PATH" - ######################################################### - # if not set, CCL will complain... ? - export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" - ######################################################### - # Sam: [2024-06-29] - export CCL_KVS_MODE=mpi - export CCL_CONFIGURATION_PATH="" - export CCL_CONFIGURATION=cpu_gpu_dpcpp - export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" - export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH - export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH - export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + # export CCL_KVS_MODE=mpi + # # export CCL_CONFIGURATION_PATH="" + # # unset CCL_CONFIGURATION_PATH + # # export CCL_CONFIGURATION=cpu_gpu_dpcpp + # # export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + # export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + # export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + # export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH + # # export CCL_ALLREDUCE_SCALEOUT=direct + # printenv | grep -E -v "^__" | grep -E "CCL|LD|CPATH|LIBRARY_PATH" + ######################################################### + # if not set, CCL will complain... ? + export NUMEXPR_MAX_THREADS="${NUMEXPR_MAX_THREADS:-16}" + ######################################################### + # Sam: [2024-06-29] + export CCL_KVS_MODE=mpi + export CCL_CONFIGURATION_PATH="" + export CCL_CONFIGURATION=cpu_gpu_dpcpp + export CCL_ROOT="/flare/Aurora_deployment/intel/ccl/_install_release_2021_13" + export LD_LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LD_LIBRARY_PATH + export CPATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/include:$CPATH + export LIBRARY_PATH=/flare/Aurora_deployment/intel/ccl/_install_release_2021_13/lib:$LIBRARY_PATH } ########################################################## # Check that we can find the `.py` file we wish to launch ########################################################## check_executable() { - fp=$1 - if [[ -f "${fp}" ]]; then - export EXEC="${fp}" - # ----[1.5 Keep track of stem from file path]------------------------- - exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") - export EXEC_STEM="${exec_stem}" - else - estr="Unable to locate executable ${fp}" - printf "[ALCF.helpers:check_executable] %s\n" "$(printRed "${estr}")" - fi + fp=$1 + if [[ -f "${fp}" ]]; then + export EXEC="${fp}" + # ----[1.5 Keep track of stem from file path]------------------------- + exec_stem=$(echo "${EXEC}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.py//g") + export EXEC_STEM="${exec_stem}" + else + estr="Unable to locate executable ${fp}" + printf "[ALCF.helpers:check_executable] %s\n" "$(printRed "${estr}")" + fi } ###################################################################### @@ -1083,69 +1070,69 @@ check_executable() { # - Otherwise, make default HOSTFILEs from "${PBS_NODEFILE}" ###################################################################### makeHostfiles() { - if [[ -n "${HOSTFILE}" ]]; then - printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" - else - make_ds_hostfile - fi + if [[ -n "${HOSTFILE}" ]]; then + printf "!! USING CUSTOM HOSTFILE FROM: %s" "${HOSTFILE}" + else + make_ds_hostfile + fi } ################################################## # Setup tokenizer as either Llama2 or GPT2 style ################################################## setup_tokenizer_and_data() { - if [[ "$#" == 1 ]]; then - tok="$1" - dfl="${DATA_FILE_LIST:-}" - elif [[ "$#" == 2 ]]; then - tok="$1" - dfl="$2" - else - echo "Incorrect number of arguments passed. Received: $#, expected 2" - fi - echo "Setting up tokenizer with ${tok}" - echo "Using data_file_list: ${dfl}" - _data_flags=() - _tokenizer_flags=() - if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then - export TOKENIZER_TYPE="GPT2" - _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") - machine=$(ezpz_get_machine_name) - if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then - export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" - elif [[ ${machine} == "sunspot" ]]; then - export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" - elif [[ ${machine} == "aurora" ]]; then - export DATA_PARENT="${DATA_PARENT:-/gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" - else - export DATA_PARENT="${DATA_PARENT:-${WORKING_DIR}/dataset}" - fi - export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" - export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" - export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" - _data_flags+=( - "--data-path ${DATA_PATH}" - "--vocab-file ${VOCAB_FILE}" - "--merge-file ${MERGE_FILE}" - ) - else - # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" - # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" - tm="${TOKENIZER_MODEL:-google/gemma-7B}" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model - export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ - _tokenizer_flags+=( - "--tokenizer-type ${TOKENIZER_TYPE}" - "--tokenizer-model ${TOKENIZER_MODEL}" - ) - # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then - echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${dfl}" - setData "${dfl}" || exit - fi - export DATA_FLAGS="${_data_flags[*]:-}" - export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" - printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" - printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" + if [[ "$#" == 1 ]]; then + tok="$1" + dfl="${DATA_FILE_LIST:-}" + elif [[ "$#" == 2 ]]; then + tok="$1" + dfl="$2" + else + echo "Incorrect number of arguments passed. Received: $#, expected 2" + fi + echo "Setting up tokenizer with ${tok}" + echo "Using data_file_list: ${dfl}" + _data_flags=() + _tokenizer_flags=() + if [[ ${tok} == gpt* || ${tok} == GPT* ]]; then + export TOKENIZER_TYPE="GPT2" + _tokenizer_flags+=("--tokenizer-type GPT2BPETokenizer") + machine=$(ezpz_get_machine_name) + if [[ ${machine} == "polaris" || ${machine} == "sophia" ]]; then + export DATA_PARENT="${DATA_PARENT:-/eagle/argonne_tpc/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" + elif [[ ${machine} == "sunspot" ]]; then + export DATA_PARENT="${DATA_PARENT:-/gila/Aurora_deployment/foremans/anl_24_q2_release/Megatron-DeepSpeed/dataset}" + elif [[ ${machine} == "aurora" ]]; then + export DATA_PARENT="${DATA_PARENT:-/gecko/Aurora_deployment/foremans/projects/argonne-lcf/Megatron-DeepSpeed/dataset}" + else + export DATA_PARENT="${DATA_PARENT:-${WORKING_DIR}/dataset}" + fi + export VOCAB_FILE="${DATA_PARENT}/gpt2-vocab.json" + export MERGE_FILE="${DATA_PARENT}/gpt2-merges.txt" + export DATA_PATH="${DATA_PARENT}/BookCorpusDataset_text_document" + _data_flags+=( + "--data-path ${DATA_PATH}" + "--vocab-file ${VOCAB_FILE}" + "--merge-file ${MERGE_FILE}" + ) + else + # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-Llama2Tokenizer}" + # tm="${WORKING_DIR}/ALCF/tokenizer.model" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" + tm="${TOKENIZER_MODEL:-google/gemma-7B}" # fallback: Megatron-DeepSpeed/ALCF/tokenizer.model + export TOKENIZER_MODEL="${TOKENIZER_MODEL:-${tm}}" # USE TOKENIZER_MODEL from env, else fallback from ^ + _tokenizer_flags+=( + "--tokenizer-type ${TOKENIZER_TYPE}" + "--tokenizer-model ${TOKENIZER_MODEL}" + ) + # if [[ "${TOKENIZER_TYPE}" != "GPT2" ]]; then + echo "Using tokenizer: ${TOKENIZER_TYPE}. Setting up data with ${dfl}" + setData "${dfl}" || exit + fi + export DATA_FLAGS="${_data_flags[*]:-}" + export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" + printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" + printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" } ############################################### @@ -1155,37 +1142,37 @@ setup_tokenizer_and_data() { # fallback to default values if necessary. ############################################### setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] - ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# - mn=$(ezpz_get_machine_name) - dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" - ############################################################################ - # set `dfl` to `dfl_fallback` if not passed as an argument, - # use this data file list to call `setData` - dfl="${1:-${dfl_fallback}}" - printf "Calling: setData() with %s\n" "${dfl}" - ndocs=$(wc -l <"${dfl}") - ws=$(sumWeights "${dfl}") - dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") - dcp=".cache/${dfl_stem}/index-cache" - export DATA_FILE_LIST="${dfl}" - export NUM_DOCS="${ndocs}" - export WEIGHT_SUM="${ws}" - export DFL_STEM="${dfl_stem}" - export DATA_CACHE_PATH="${dcp}" - # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" - echo "--------------------" - echo "Updated environment:" - printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" - printf "NUM_DOCS: %s\n " "${NUM_DOCS}" - printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" - printf "DFL_STEM: %s\n" "${DFL_STEM}" - printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" - printf "DATA_FLAGS: %s\n" "${DATA_FLAGS:-}" - echo "--------------------" + ####### [Set DATA_FILE_LIST_FALLBACK based on current machine] ############# + mn=$(ezpz_get_machine_name) + dfl_fallback="${WORKING_DIR}/ALCF/data-lists/${mn}/dolma.txt" + ############################################################################ + # set `dfl` to `dfl_fallback` if not passed as an argument, + # use this data file list to call `setData` + dfl="${1:-${dfl_fallback}}" + printf "Calling: setData() with %s\n" "${dfl}" + ndocs=$(wc -l <"${dfl}") + ws=$(sumWeights "${dfl}") + dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") + dcp=".cache/${dfl_stem}/index-cache" + export DATA_FILE_LIST="${dfl}" + export NUM_DOCS="${ndocs}" + export WEIGHT_SUM="${ws}" + export DFL_STEM="${dfl_stem}" + export DATA_CACHE_PATH="${dcp}" + # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" + echo "--------------------" + echo "Updated environment:" + printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" + printf "NUM_DOCS: %s\n " "${NUM_DOCS}" + printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" + printf "DFL_STEM: %s\n" "${DFL_STEM}" + printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" + printf "DATA_FLAGS: %s\n" "${DATA_FLAGS:-}" + echo "--------------------" } generateDSconfig_new() { - cat <"${CONFIG_JSON}" + cat <"${CONFIG_JSON}" { "train_batch_size" : $GLOBAL_BATCH, "train_micro_batch_size_per_gpu": $MICRO_BATCH, @@ -1216,26 +1203,26 @@ EOT # This will contain the appropriate variables as set in the current environment. ################################################################################ generateDSconfig() { - if [ $# -ne 1 ]; then - echo "Usage: $0 config_file" - exit 1 - fi - for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE"; do - if [ -z "$v" ]; then - echo "Please export required envs before execute $0" - exit 1 - fi - done - # \"scheduler\": { - # \"type\": \"WarmupLR\", - # \"params\": { - # \"warmup_min_lr\": 0.00003, - # \"warmup_max_lr\": 0.0003, - # \"warmup_num_steps\": 5000 - # } - # }, - extra="" - common="\ + if [ $# -ne 1 ]; then + echo "Usage: $0 config_file" + exit 1 + fi + for v in "$GLOBAL_BATCH" "$MICRO_BATCH" "$GRAD_ACC_STEPS" "$ZERO_STAGE" "$PP" "$DTYPE"; do + if [ -z "$v" ]; then + echo "Please export required envs before execute $0" + exit 1 + fi + done + # \"scheduler\": { + # \"type\": \"WarmupLR\", + # \"params\": { + # \"warmup_min_lr\": 0.00003, + # \"warmup_max_lr\": 0.0003, + # \"warmup_num_steps\": 5000 + # } + # }, + extra="" + common="\ \"train_batch_size\": $GLOBAL_BATCH, \"train_micro_batch_size_per_gpu\": $MICRO_BATCH, \"gradient_clipping\": 1.0, @@ -1244,16 +1231,16 @@ generateDSconfig() { \"zero_force_ds_cpu_optimizer\": false, \"zero_allow_untested_optimizer\": true, \"wall_clock_breakdown\": false," - # if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then - # activation_checkpointing="\ - # \"activation_checkpointing\": { - # \"partition_activations\": true, - # \"contiguous_memory_optimization\": true - # }," - # fi - if [[ $DTYPE == "bf16" ]]; then - # \"communication_data_type\": \"bf16\", - dtype="\ + # if [[ "${USE_ACTIVATION_CHECKPOINTING}" == 1 ]]; then + # activation_checkpointing="\ + # \"activation_checkpointing\": { + # \"partition_activations\": true, + # \"contiguous_memory_optimization\": true + # }," + # fi + if [[ $DTYPE == "bf16" ]]; then + # \"communication_data_type\": \"bf16\", + dtype="\ \"fp16\": { \"enabled\": false, \"loss_scale\": 0, @@ -1264,8 +1251,8 @@ generateDSconfig() { \"bfloat16\": { \"enabled\": true }," - elif [[ $DTYPE == "fp16" ]]; then - dtype="\ + elif [[ $DTYPE == "fp16" ]]; then + dtype="\ \"communication_data_type\": \"fp16\", \"fp16\": { \"enabled\": true, @@ -1278,11 +1265,11 @@ generateDSconfig() { \"enabled\": false, \"loss_scale\": 1.0 }," - else - dtype="\"communication_data_type\": \"fp32\"," - fi - if [[ "${OPT:-}" == "ds.adamw" ]]; then - optimizer="\ + else + dtype="\"communication_data_type\": \"fp32\"," + fi + if [[ "${OPT:-}" == "ds.adamw" ]]; then + optimizer="\ \"optimizer\": { \"type\": \"AdamW\", \"params\": { @@ -1293,8 +1280,8 @@ generateDSconfig() { \"weight_decay\": 1e-1 }, }," - elif [[ "${OPT:-}" == "ds.onebitlamb" ]]; then - optimizer="\ + elif [[ "${OPT:-}" == "ds.onebitlamb" ]]; then + optimizer="\ \"optimizer\": { \"type\": \"OneBitLamb\", \"params\": { @@ -1310,35 +1297,35 @@ generateDSconfig() { \"factor_threshold\": 0.1 } }," - else - optimizer="" - fi - if [[ "${ZERO_STAGE}" == 3 ]]; then - # \"mics_shard_size\": 2, - # hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-${NGPU_PER_HOST:-1}} - # zero="\ - # \"zero_optimization\": { - # \"stage\": 3, - # \"reduce_scatter\": true, - # \"reduce_bucket_size\": 10000000, - # \"reduce_scatter\": true, - # \"zero_quantized_weights\": true, - # \"zero_hpz_partition_size\": $hpz_partition_size, - # \"zero_quantized_gradients\": true, - # \"contiguous_gradients\": true, - # \"overlap_comm\": true, - # } - # }," - # mics_shard_size="${MICS_SHARD_SIZE:-${NGPU_PER_HOST}}" - # \"mics_shard_size\": $mics_shard_size, - hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-1} - - # \"zero_quantized_weights\": false, - # \"zero_hpz_partition_size\": $hpz_partition_size, - # \"zero_quantized_gradients\": false, - # \"mics_shard_size\": 1, - # \"mics_hierarchical_params_gather\": false, - zero="\ + else + optimizer="" + fi + if [[ "${ZERO_STAGE}" == 3 ]]; then + # \"mics_shard_size\": 2, + # hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-${NGPU_PER_HOST:-1}} + # zero="\ + # \"zero_optimization\": { + # \"stage\": 3, + # \"reduce_scatter\": true, + # \"reduce_bucket_size\": 10000000, + # \"reduce_scatter\": true, + # \"zero_quantized_weights\": true, + # \"zero_hpz_partition_size\": $hpz_partition_size, + # \"zero_quantized_gradients\": true, + # \"contiguous_gradients\": true, + # \"overlap_comm\": true, + # } + # }," + # mics_shard_size="${MICS_SHARD_SIZE:-${NGPU_PER_HOST}}" + # \"mics_shard_size\": $mics_shard_size, + hpz_partition_size=${ZERO_HPZ_PARTITION_SIZE:-1} + + # \"zero_quantized_weights\": false, + # \"zero_hpz_partition_size\": $hpz_partition_size, + # \"zero_quantized_gradients\": false, + # \"mics_shard_size\": 1, + # \"mics_hierarchical_params_gather\": false, + zero="\ \"zero_optimization\": { \"stage\": 3, \"zero_hpz_partition_size\": $hpz_partition_size, @@ -1359,25 +1346,25 @@ generateDSconfig() { \"pin_memory\": true } }," - # elif [[ $ZERO_STAGE == 2 ]]; then - elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then - if [[ -n "${CPU_OPTIMIZER:-}" ]]; then - echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" - zero="\ + # elif [[ $ZERO_STAGE == 2 ]]; then + elif [[ "${ZERO_STAGE}" == 2 || "${ZERO_STAGE}" == 1 ]]; then + if [[ -n "${CPU_OPTIMIZER:-}" ]]; then + echo "!!!! CAUGHT CPU_OPTIMIZER !!!!" + zero="\ \"zero_optimization\": { \"stage\": $ZERO_STAGE, \"offload_optimizer\": { \"device\": \"cpu\" } }," - else - zero="\ + else + zero="\ \"zero_optimization\": { \"stage\": $ZERO_STAGE }," - fi - if [[ "${PP}" -gt 1 ]]; then - extra="\ + fi + if [[ "${PP}" -gt 1 ]]; then + extra="\ \"data_types\": { \"grad_accum_dtype\": \"fp32\" }, @@ -1387,18 +1374,18 @@ generateDSconfig() { \"prof_all\": true, \"debug\": false }," - else - extra="\ + else + extra="\ \"comms_logger\": { \"enabled\": ${COMMS_LOGGER:-false}, \"verbose\": false, \"debug\": false }," - fi - else - echo 'Please add the correct config set!!!' - fi - flops_profiler="\ + fi + else + echo 'Please add the correct config set!!!' + fi + flops_profiler="\ \"flops_profiler\": { \"enabled\": true, \"profile_step\": 2, @@ -1407,7 +1394,7 @@ generateDSconfig() { \"detailed\": true, \"output_file\": null }" - cat <"$1" + cat <"$1" { $common $optimizer @@ -1432,116 +1419,116 @@ CYAN="\e[1;35m" # WHITE="\e[1;36m" printBlack() { - printf "\e[1;30m%s\e[0m\n" "$@" + printf "\e[1;30m%s\e[0m\n" "$@" } printRed() { - printf "\e[1;31m%s\e[0m\n" "$@" + printf "\e[1;31m%s\e[0m\n" "$@" } printGreen() { - printf "\e[1;32m%s\e[0m\n" "$@" + printf "\e[1;32m%s\e[0m\n" "$@" } printYellow() { - printf "\e[1;33m%s\e[0m\n" "$@" + printf "\e[1;33m%s\e[0m\n" "$@" } printBlue() { - printf "\e[1;34m%s\e[0m\n" "$@" + printf "\e[1;34m%s\e[0m\n" "$@" } printMagenta() { - printf "\e[1;35m%s\e[0m\n" "$@" + printf "\e[1;35m%s\e[0m\n" "$@" } printCyan() { - printf "\e[1;36m%s\e[0m\n" "$@" + printf "\e[1;36m%s\e[0m\n" "$@" } printWhite() { - printf "\e[1;37m%s\e[0m\n" "$@" + printf "\e[1;37m%s\e[0m\n" "$@" } reset_env() { - custom_vars=( - NO_FLASH_ATTN - USE_FLASH_ATTN - TP - PP - SP - FLASH_ARG - OPT - ADAM_BETA1 - ADAM_BETA2 - ADAM_EPS - WEIGHT_DECAY - HEADS - NLAYERS - HIDDEN - NUM_KV_HEAD - FFN_HIDDEN_SIZE - SEQ - ZERO_STAGE - MICRO_BATCH - EVAL_ITERS - EVAL_INTERVAL - TIMING_LOG_LEVEL - ACT_CKPT_NUM_LAYERS - USE_ACTIVATION_CHECKPOINTING - GLOBAL_BATCH_MAX - GLOBAL_BATCH - TRAIN_TOKENS - TRAIN_ITERS - MODEL_TYPE - LR - LR_WARMUP_FRAC - LR_DECAY_ITERS - LR_ARGS - CPU_OPTIMIZER - DS_CONFIG - OUTPUT_DIR - OUTPUT_LOG - CKPT_DIR - ds_args - EXEC - EXEC_STEM - DATA_FLAGS - TOKENIZER_TYPE - TOKENIZER_MODEL - TOKENIZER_FLAGS - DATA_FILE_LIST - NUM_DOCS - WEIGHT_SUM - DFL_STEM - DATA_CACHE_PATH - DOTENV_FILE - YEAR - MONTH - DAY - TODAY - STARTED_AT - LAUNCHER - data_cache_path - DEFAULTS - ) - # LLAMA_ARGS - printf "Unsetting custom vars: %s\n" "${custom_vars[*]}" - unset "${custom_vars[@]}" + custom_vars=( + NO_FLASH_ATTN + USE_FLASH_ATTN + TP + PP + SP + FLASH_ARG + OPT + ADAM_BETA1 + ADAM_BETA2 + ADAM_EPS + WEIGHT_DECAY + HEADS + NLAYERS + HIDDEN + NUM_KV_HEAD + FFN_HIDDEN_SIZE + SEQ + ZERO_STAGE + MICRO_BATCH + EVAL_ITERS + EVAL_INTERVAL + TIMING_LOG_LEVEL + ACT_CKPT_NUM_LAYERS + USE_ACTIVATION_CHECKPOINTING + GLOBAL_BATCH_MAX + GLOBAL_BATCH + TRAIN_TOKENS + TRAIN_ITERS + MODEL_TYPE + LR + LR_WARMUP_FRAC + LR_DECAY_ITERS + LR_ARGS + CPU_OPTIMIZER + DS_CONFIG + OUTPUT_DIR + OUTPUT_LOG + CKPT_DIR + ds_args + EXEC + EXEC_STEM + DATA_FLAGS + TOKENIZER_TYPE + TOKENIZER_MODEL + TOKENIZER_FLAGS + DATA_FILE_LIST + NUM_DOCS + WEIGHT_SUM + DFL_STEM + DATA_CACHE_PATH + DOTENV_FILE + YEAR + MONTH + DAY + TODAY + STARTED_AT + LAUNCHER + data_cache_path + DEFAULTS + ) + # LLAMA_ARGS + printf "Unsetting custom vars: %s\n" "${custom_vars[*]}" + unset "${custom_vars[@]}" } convert_ckpt_to_universal() { - if [[ "$#" -ne 1 ]]; then - echo "Usage: convert_ckpt_to_universal ckpt_dir" - echo "Expected one argument (ckpt_dir), received: $#" - exit 1 - fi - ckptdir=$1 - gs=$(cat "${ckptdir}/latest_checkpointed_iteration.txt") - src="${ckptdir}/global_step${gs}" - dst="${ckptdir}/global_step${gs}_universal" - convert_script="${PBS_O_WORKDIR}/deps/DeepSpeed/checkpoint/ds_to_universal.py" - python3 "${convert_script}" --input_folder "${src}" --output_folder "${dst}" + if [[ "$#" -ne 1 ]]; then + echo "Usage: convert_ckpt_to_universal ckpt_dir" + echo "Expected one argument (ckpt_dir), received: $#" + exit 1 + fi + ckptdir=$1 + gs=$(cat "${ckptdir}/latest_checkpointed_iteration.txt") + src="${ckptdir}/global_step${gs}" + dst="${ckptdir}/global_step${gs}_universal" + convert_script="${PBS_O_WORKDIR}/deps/DeepSpeed/checkpoint/ds_to_universal.py" + python3 "${convert_script}" --input_folder "${src}" --output_folder "${dst}" } ########################### From 4eab242dc10fcb5c09f524ee9369557d3335abf9 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Aug 2025 10:09:15 -0500 Subject: [PATCH 484/590] feat: Add `train_aGPT_2B_large_batch.sh` --- train_aGPT_2B_large_batch.sh | 54 ++++++++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) create mode 100755 train_aGPT_2B_large_batch.sh diff --git a/train_aGPT_2B_large_batch.sh b/train_aGPT_2B_large_batch.sh new file mode 100755 index 00000000000..b4053262e6e --- /dev/null +++ b/train_aGPT_2B_large_batch.sh @@ -0,0 +1,54 @@ +#!/bin/bash --login +#PBS -q prod +#PBS -A AuroraGPT +#PBS -j oe +#PBS -l walltime=06:00:00,filesystems=flare:home +#PBS -l select=256 + +cd /flare/datascience/foremans/projects/argonne-lcf/Megatron-DeepSpeed || { + echo "Failed to change directory to /flare/datascience/foremans/projects/argonne-lcf/Megatron-DeepSpeed" + exit 1 +} + +PBS_O_WORKDIR="$(pwd)" +export PBS_O_WORKDIR + +export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" +export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" +export http_proxy="http://proxy.alcf.anl.gov:3128" +export https_proxy="http://proxy.alcf.anl.gov:3128" +export ftp_proxy="http://proxy.alcf.anl.gov:3128" + + +# shellcheck disable=SC1090 +source <(curl -L https://bit.ly/ezpz-utils) +ezpz_setup_env + +# eval "$(micromamba shell hook --shell "$(basename "${SHELL}")")" +# ezpz_load_new_pt_modules_aurora +# micromamba activate /flare/datascience_collab/foremans/micromamba/envs/pt29-2025-07 +# +# NODES=$(cat "${PBS_NODEFILE}" | uniq | wc -l) +# mpiexec -np "${NODES}" --ppn 1 python3 \ +# /flare/datascience/foremans/projects/argonne-lcf/scalable_conda_env/cache_soft.py \ +# --src=/flare/datascience/foremans/pt29-ezpz-2025-07.tar.gz \ +# --dst=/tmp/pt29-ezpz-2025-07.tar.gz \ +# --d +# +# micromamba deactivate +# micromamba activate /tmp/pt29-ezpz-2025-07 +# which python3 + +MODEL_ARCH=smollm3-3B \ + OPT=ipex.fusedlamb \ + NLAYERS=12 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + OPT=adamw \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" From 3848f6fbda623e57973051a070bcfb501bc95aa2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Aug 2025 10:09:59 -0500 Subject: [PATCH 485/590] docs: Update `ALCF/notes/*.md` --- ALCF/notes/AuroraGPT-3B.md | 94 +++++++++++++++++++++++++++++++++++++ ALCF/notes/AuroraGPT-70B.md | 18 ++++++- 2 files changed, 111 insertions(+), 1 deletion(-) create mode 100644 ALCF/notes/AuroraGPT-3B.md diff --git a/ALCF/notes/AuroraGPT-3B.md b/ALCF/notes/AuroraGPT-3B.md new file mode 100644 index 00000000000..bbee38caba1 --- /dev/null +++ b/ALCF/notes/AuroraGPT-3B.md @@ -0,0 +1,94 @@ +# AuroraGPT-3B-v1 + +## Organization + +``` +/flare/AuroraGPT/AuroraGPT-v1/3B/Megatron-DeepSpeed/ +``` + +## Llama-3.2-3B Base Config + + +- Model Architecture: + -
model architecture (yaml) + + ```yaml + DP: 1536 + FFN_HIDDEN_SIZE: 8192 + GAS: + - 2 + - 4 + GBS: 12288 + HIDDEN: 3072 + HEADS: 32 + MBS: 4 + NLAYERS: 28 + NUM_KV_HEAD: 8 + OPT: ipex.fusedlamb + PP: 1 + SEQ: 8192 + SP: 1 + TP: 1 + USE_ACTIVATION_CHECKPOINTING: 1 + ZERO_STAGE: 1 + ``` + +
+ +- 128 Nodes: + - [pleasant-snowflake-3180](https://wandb.ai/aurora_gpt/AuroraGPT/runs/y9r7r3mh) + + ```bash + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=4 \ + USE_ACTIVATION_CHECKPOINTING=1 \ + ZERO_STAGE=0 \ + MODEL_ARCH=3B \ + OPT=ipex.fusedlamb \ + bash train_alcf.sh + ``` + +- 64 Nodes: + - [jolly-silence-3185](https://wandb.ai/aurora_gpt/AuroraGPT/runs/28qxlycg) + + ```bash + GRAD_ACC_STEPS=4 \ + MICRO_BATCH=4 \ + USE_ACTIVATION_CHECKPOINTING=1 \ + ZERO_STAGE=0 \ + MODEL_ARCH=3B \ + OPT=ipex.fusedlamb \ + bash train_alcf.sh + ``` + + +## Reference Configs + +- [Llama-3.2-3B](https://huggingface.co/meta-llama/Llama-3.2-3B): + - [meta-llama/Llama-3.2-3B/config.json](https://huggingface.co/meta-llama/Llama-3.2-3B/blob/main/config.json) +- [Gemma-3](https://huggingface.co/google/gemma-3-4b-it): + - [google/gemma-3-4b-it/config.json](https://huggingface.co/google/gemma-3-4b-it/blob/main/config.json) +- [SmolLM3-3B-Base](https://huggingface.co/HuggingFaceTB/SmolLM3-3B-Base): + - [HuggingFaceTB/SmolLM3-3B-Base/config.json](https://huggingface.co/HuggingFaceTB/SmolLM3-3B-Base/blob/main/config.json) + + +## Experiments + +Would like to run: + +- Large Batch Training + - Identify largest number of tokens / batch that preserves stability +- Optimizer(s): + - Muon + - Lamb + - AdamW + - schedulefree (?) + - Others? +- Tokenizer(s): + - Implementation: + - Llama Tokenizer + - Gemma Tokenizer + - GitHub Repos: + - [M4THYOU/TokenDagger](https://huggingface.co/M4THYOU/TokenDagger) + - [openai/TikToken](https://github.com/openai/tiktoken) + - [google/sentencepiece](https://github.com/google/sentencepiece) diff --git a/ALCF/notes/AuroraGPT-70B.md b/ALCF/notes/AuroraGPT-70B.md index 530c2145780..b722862a5fc 100644 --- a/ALCF/notes/AuroraGPT-70B.md +++ b/ALCF/notes/AuroraGPT-70B.md @@ -28,6 +28,22 @@ PP="${NHOSTS}" \ - `sps`: samples per second - `tpgps`: tokens per GPU per second +#### PyTorch 2.8 Results + + +| Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | +| :------: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :-------------------------------------------------------------------------- | +| 8[^2273] | 96 | 8 | 12 | 1 | 1 | 1 | 12 | True | OOM | OOM | OOM | [northern-field-2273](https://wandb.ai/aurora_gpt/AuroraGPT/runs/odz3euea) | +| 8[^2274] | 96 | 8 | 6 | 2 | 1 | 1 | 6 | True | 54.66 | 2.882 | 122.967 | [chocolate-cloud-2274](https://wandb.ai/aurora_gpt/AuroraGPT/runs/smov9k5d) | +| 8 | 96 | 8 | 6 | 2 | 1 | 2 | 12 | True | | | | | +| 8 | 96 | 8 | 6 | 2 | 1 | 4 | 24 | True | | | | | + +[^2273]: [northern-field-2273](https://wandb.ai/aurora_gpt/AuroraGPT/runs/odz3euea) +[^2274]: [chocolate-cloud-2274](https://wandb.ai/aurora_gpt/AuroraGPT/runs/smov9k5d) + +#### PyTorch 2.5 Results + +Using the default PyTorch (2.5) provided by the `frameworks` module on Aurora | Nodes | NGPU | PP | DP | TP | MBS | GAS | GBS | ACKPT | TFLOPS | `sps` | `tpgps` | wandb | | :-------: | :--: | :-: | :-: | :-: | :-: | :-: | --: | :---: | -----: | ----: | ------: | :----------------------------------------------------------------------------- | @@ -73,7 +89,7 @@ PP="${NHOSTS}" \ ZERO_STAGE=3 \ HEADS=64 \ NLAYERS=80 \ - NUM_KV_HEAD=8 \ + [NUM_KV_HEAD](2025-07-07_num_kv_head.md)=8 \ FFN_HIDDEN_SIZE=28672 \ HIDDEN=8192 \ MICRO_BATCH=1 \ From 96c5a10accf12b1acf3a75b3433f2d9fc95b7b4f Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 26 Aug 2025 10:10:47 -0500 Subject: [PATCH 486/590] chore: Add `train_aGPT_7B_chain.sh` --- train_aGPT_7B_chain.sh | 38 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 38 insertions(+) create mode 100644 train_aGPT_7B_chain.sh diff --git a/train_aGPT_7B_chain.sh b/train_aGPT_7B_chain.sh new file mode 100644 index 00000000000..0a3c69173d3 --- /dev/null +++ b/train_aGPT_7B_chain.sh @@ -0,0 +1,38 @@ +#!/bin/bash --login +# Example usage: +# WALLTIME="06:00:00" NJOBS=4 NHOSTS=64 bash train_aGPT_7B_chain.sh + +NHOSTS="${NHOSTS:-24}" +#!/bin/bash --login +# Example usage: +# WALLTIME="06:00:00" NJOBS=4 NHOSTS=64 bash train_aGPT_7B_chain.sh + +NHOSTS="${NHOSTS:-24}" +WALLTIME="${WALLTIME:-"02:00:00"}" +NJOBS="${NJOBS:-10}" + +JOBIDS=() + +printf "Caught NHOSTS: %s, WALLTIME: %s, NJOBS: %s\n" "${NHOSTS}" "${WALLTIME}" "${NJOBS}" +printf "Submitting job 0/%s\n" "${NJOBS}" +job_cmd="qsub -A datascience -q prod -l select=${NHOSTS} -l walltime=${WALLTIME},filesystems=flare:home ~/test.sh" +printf "JOB0: %s\n" "${job_cmd}" +# JOBIDS+=("$(bash -c "${job_cmd}")") +# JOBIDS+=("$(eval "${job_cmd}")") +# echo "${JOBIDS[@]}" +jobid=($(eval "${job_cmd}")) +echo "First jobid: ${jobid[*]}" + +for (( idx=1; idx<="${NJOBS}"; idx+=1 )); do + echo "Submitting job ${idx}/${NJOBS}" + printf "JOBIDS[idx]: %s\n" "${JOBIDS[idx]}" + echo "${JOBIDS[idx]}" + job_cmd="qsub -W depend=afterany:${JOBIDS[idx]} -A datascience -q prod -l select=${NHOSTS} -l walltime=${WALLTIME},filesystems=flare:home ~/test.sh" + echo "Job cmd for JOB ${idx}:" + echo "${job_cmd}" + JOBIDS+=("$(bash -c "${job_cmd}")") + # printf "%s\n" "${job_cmd}" + # eval "${job_cmd}" + # JOBIDS+=("$(eval "${job_cmd}")") + # printf "Submitted %s/%s: %s\n" "${idx}" "${NJOBS}" "${JOBIDS[((idx + 1))}" +done From fc4d16790c1a2744edb20dd516ad2cef46c6b8aa Mon Sep 17 00:00:00 2001 From: Marieme Ngom Date: Wed, 27 Aug 2025 03:17:18 +0000 Subject: [PATCH 487/590] added cooldown phase option to constant LR decay --- megatron/arguments.py | 19 ++++++-- megatron/optimizer_param_scheduler.py | 63 +++++++++++++++++++++------ megatron/training.py | 9 ++++ 3 files changed, 75 insertions(+), 16 deletions(-) diff --git a/megatron/arguments.py b/megatron/arguments.py index 0de06d0976a..3ca98641ff0 100644 --- a/megatron/arguments.py +++ b/megatron/arguments.py @@ -1750,14 +1750,14 @@ def _add_learning_rate_args(parser): type=float, default=0.001, help="fraction of lr-constant-(iters/samples) to use " - "for constant phase (as a float)", + "for infinite schedulers constant phase (as a float)", ) group.add_argument( "--lr-cooldown-tokens", type=int, default=None, - help="number of tokens to cooldown learning rate over," + help="number of tokens to cooldown learning rate over for infinite schedulers," " If not None will override iter/sample-based decay", ) @@ -1765,7 +1765,20 @@ def _add_learning_rate_args(parser): "--lr-cooldown-fraction", type=float, default=0.65, - help="fraction -(iters/samples) to use for cooldownt phase (as a float)", + help="fraction -(iters/samples) to use for cooldown phase for infinite schedulers (as a float)", + ) + + group.add_argument( + "--lr_constant_plus_cooldown", + action="store_true", + help="Perform cooldown for constant decay style", + ) + + group.add_argument( + "--lr_constant_plus_cooldown_frac", + type=float, + default=0.05, + help="Cooldown frction for constant decay style", ) group.add_argument( diff --git a/megatron/optimizer_param_scheduler.py b/megatron/optimizer_param_scheduler.py index f62fb23697a..120d540f6b8 100644 --- a/megatron/optimizer_param_scheduler.py +++ b/megatron/optimizer_param_scheduler.py @@ -15,6 +15,7 @@ def __init__( optimizer, max_lr, min_lr, + lr_constant_plus_cooldown_steps, lr_warmup_steps, lr_decay_steps, lr_decay_style, @@ -47,8 +48,14 @@ def __init__( self.end_steps = lr_decay_steps assert self.lr_decay_steps > 0 assert self.lr_warmup_steps < self.lr_decay_steps - assert self.lr_constant_steps < self.lr_decay_steps - + if self.lr_constant_steps is not None: + assert self.lr_constant_steps < self.lr_decay_steps + + # for constant+cooldown + self.lr_constant_plus_cooldown = args.lr_constant_plus_cooldown + self.lr_constant_plus_cooldown_steps = lr_constant_plus_cooldown_steps + + #for infinite schedulers self.timescale=timescale self.lr_decay_tokens = args.lr_decay_tokens self.num_tokens = 0 @@ -102,6 +109,7 @@ def get_wd(self): ) ) + return self.start_wd + coeff * delta_wd def get_lr(self): @@ -126,8 +134,37 @@ def get_lr(self): # If the learning rate is constant, just return the initial value. if self.lr_decay_style == "constant": - return self.max_lr + if not self.lr_constant_plus_cooldown: #* + return self.max_lr + + if self.lr_decay_tokens is None: + #steps based + num_steps_ = self.num_steps - self.lr_warmup_steps + total_steps_ = self.end_steps - self.lr_warmup_steps + if num_steps_ >= total_steps_ : return self.min_lr + cooldown_steps_ = int(self.lr_constant_plus_cooldown_steps) + if num_steps_ < total_steps_ - cooldown_steps_: return self.max_lr + print(total_steps_) + print(cooldown_steps_) + ratio = (num_steps_ - (total_steps_ - cooldown_steps_)) / float(cooldown_steps_) + if ratio < 0.0: ratio = 0.0 + if ratio > 1.0: ratio = 1.0 + return self.min_lr + (self.max_lr - self.min_lr) * (1.0 - math.sqrt(ratio)) + else: + #token based + num_tokens_ = self.num_tokens - self.lr_warmup_tokens + total_tokens_ = self.end_tokens - self.lr_warmup_tokens + if num_tokens_ >= total_tokens_ : return self.min_lr + cooldown_tokens_ = int(self.lr_constant_plus_cooldown_tokens) + if num_tokens_ < total_tokens_ - cooldown_tokens_: return self.max_lr + print(total_tokens_) + print(cooldown_tokens_) + ratio = (num_tokens_ - (total_tokens_ - cooldown_tokens_)) / float(cooldown_tokens_) + if ratio < 0.0: ratio = 0.0 + if ratio > 1.0: ratio = 1.0 + return self.min_lr + (self.max_lr - self.min_lr) * (1.0 - math.sqrt(ratio)) +#################################################### # For any steps larger than `self.lr_decay_steps`, use `self.min_lr`. if self.lr_decay_tokens is None: if self.num_steps > self.lr_decay_steps: @@ -157,7 +194,7 @@ def get_lr(self): delta_lr = self.max_lr - self.constant_lr if self.lr_decay_tokens is None: num_steps_ = self.num_steps - self.lr_warmup_steps - cooldown_steps_ = self.lr_cooldown_steps - self.lr_warmup_steps + cooldown_steps_ = self.lr_cooldown_steps #- self.lr_warmup_steps if self.lr_constant_steps is None: raise Exception( "Constant LR steps need to be provided for infinite schedulers" @@ -165,15 +202,15 @@ def get_lr(self): if num_steps_ <= cooldown_steps_: cooldown_ratio = float(num_steps_) / float(cooldown_steps_) - if self.lr_decay_style == "infinite_cosine": + if self.lr_decay_style == "infinite-cosine": coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) lr = self.constant_lr + delta_lr * coeff - else: # infinite_inv_sqrt + else: # infinite-inv-sqrt def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 - coeff = inv_f(1) * inv_f(cooldown_ratio) + coeff = inv_f(cooldown_ratio)/inv_f(1) lr = self.max_lr - delta_lr * coeff else: @@ -186,7 +223,7 @@ def inv_f(t): end_steps_ = ( self.end_steps - - self.warmup_steps + - self.lr_warmup_steps - cooldown_steps_ - self.lr_constant_steps ) @@ -198,8 +235,8 @@ def inv_f(t): # token based decay else: - num_tokens_ = self.num_tokens_ - self.lr_warmup_tokens - cooldown_tokens_ = self.lr_cooldown_tokens - self.lr_warmup_tokens + num_tokens_ = self.num_tokens - self.lr_warmup_tokens + cooldown_tokens_ = self.lr_cooldown_tokens # - self.lr_warmup_tokens if self.constant_tokens is None: raise Exception( "Constant LR tokens need to be provided for infinite schedulers" @@ -207,7 +244,7 @@ def inv_f(t): if num_tokens_ <= cooldown_tokens_: cooldown_ratio = float(num_tokens_) / float(cooldown_tokens_) - if self.lr_decay_style == "infinite_cosine": + if self.lr_decay_style == "infinite-cosine": coeff = 0.5 * (math.cos(math.pi * cooldown_ratio) + 1.0) lr = self.constant_lr + delta_lr * coeff else: # infinite_inv_sqrt @@ -215,7 +252,7 @@ def inv_f(t): def inv_f(t): return (1 / math.sqrt(1 + (self.timescale * t))) - 1 - coeff = inv_f(1) * inv_f(cooldown_ratio) + coeff = inv_f(cooldown_ratio)/inv_f(1) lr = self.max_lr - delta_lr * coeff else: num_tokens_ = num_tokens_ - cooldown_tokens_ @@ -232,7 +269,7 @@ def inv_f(t): - cooldown_tokens_ - self.lr_constant_tokens ) - num_tokens_ = num_tokens_ - constant_period + num_tokens_ = num_tokens_ - self.lr_constant_tokens exp_factor = ( -math.log(self.min_lr / self.constant_lr) / end_tokens_ ) diff --git a/megatron/training.py b/megatron/training.py index c0938b6c4b7..ab1942bfcc7 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -576,6 +576,10 @@ def get_optimizer_param_scheduler(optimizer): lr_constant_steps = args.lr_constant_fraction * lr_decay_steps else: lr_constant_steps = args.lr_constant_iters * args.global_batch_size + if args.lr_constant_plus_cooldown: + lr_constant_plus_cooldown_steps = args.lr_constant_plus_cooldown_frac * lr_decay_steps + else: + lr_constant_plus_cooldown_steps = 0 lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps # Sample-based training. elif args.train_samples: @@ -595,6 +599,10 @@ def get_optimizer_param_scheduler(optimizer): lr_constant_steps = args.lr_constant_fraction * lr_decay_steps else: lr_constant_steps = args.lr_constant_samples + if args.lr_constant_plus_cooldown: + lr_constant_plus_cooldown_steps = args.lr_constant_plus_cooldown_frac * lr_decay_steps + else: + lr_constant_plus_cooldown_steps = 0 lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps else: raise Exception("either train-iters or train-samples should be provided.") @@ -603,6 +611,7 @@ def get_optimizer_param_scheduler(optimizer): optimizer, max_lr=args.lr, min_lr=args.min_lr, + lr_constant_plus_cooldown_steps=lr_constant_plus_cooldown_steps, lr_warmup_steps=lr_warmup_steps, lr_decay_steps=lr_decay_steps, lr_decay_style=args.lr_decay_style, From 9b46590492d53e1cc5a6b7c8de9d013969ed9a17 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 27 Aug 2025 08:35:09 -0500 Subject: [PATCH 488/590] feat: Add `train_aGPT_2B_large_batch.sh` --- train_aGPT_2B_large_batch_pt28.sh | 66 +++++++++++++++++++++++++++++++ 1 file changed, 66 insertions(+) create mode 100755 train_aGPT_2B_large_batch_pt28.sh diff --git a/train_aGPT_2B_large_batch_pt28.sh b/train_aGPT_2B_large_batch_pt28.sh new file mode 100755 index 00000000000..5a31b628c31 --- /dev/null +++ b/train_aGPT_2B_large_batch_pt28.sh @@ -0,0 +1,66 @@ +#!/bin/bash --login +#PBS -q prod +#PBS -A AuroraGPT +#PBS -j oe +#PBS -l walltime=06:00:00,filesystems=flare:home +#PBS -l select=256 + +# Get the directory of this script +whereami="$(dirname "$(realpath "$0")")" +cd "${whereami}" || { + echo "Failed to change directory to ${whereami}" + exit 1 +} + +setup_env() { + PBS_O_WORKDIR="$(pwd)" + export PBS_O_WORKDIR + export http_proxy="http://proxy.alcf.anl.gov:3128" + export https_proxy="http://proxy.alcf.anl.gov:3128" + export ftp_proxy="http://proxy.alcf.anl.gov:3128" + export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" + export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" + + # shellcheck disable=SC1090 + source /home/foremans/utils.sh + + conda_env=/flare/datascience/foremans/micromamba/envs/2025-07-pt28 + conda_name="$(basename "$conda_env")" + dst_env="/tmp/${conda_name}" + + if [[ ! -d "${dst_env}" ]]; then + echo "No cached conda environment found, creating a new one" + ezpz_setup_env + ezpz-yeet-env --src "${conda_env}.tar.gz" + deactivate + conda deactivate + fi + + ezpz_load_new_pt_modules_aurora + conda activate "${dst_env}" + ezpz_setup_env + log_message INFO "Using: $(which python3)" +} + +train_model() { + MODEL_ARCH=smollm3-3B \ + OPT=ipex.fusedlamb \ + NLAYERS=12 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" \ + "$@" +} + +main() { + setup_env + train_model "$@" +} + +main "$@" From 6b32c95f5ecefd4bd2ae2027acafbf992f3a8a75 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 11 Sep 2025 16:13:26 -0500 Subject: [PATCH 489/590] feat: Initial logic to prevent `NaN`s from crashing training --- megatron/core/pipeline_parallel/schedules.py | 98 ++++++++++++++------ 1 file changed, 69 insertions(+), 29 deletions(-) diff --git a/megatron/core/pipeline_parallel/schedules.py b/megatron/core/pipeline_parallel/schedules.py index cfb3bab87b5..5351ed58ab6 100644 --- a/megatron/core/pipeline_parallel/schedules.py +++ b/megatron/core/pipeline_parallel/schedules.py @@ -3,6 +3,7 @@ import contextlib from typing import Callable, Iterator, List, Optional, Union +import ezpz import torch from torch.autograd.variable import Variable from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP @@ -14,7 +15,7 @@ from megatron.core.enums import ModelType from megatron.core.utils import get_attr_wrapped_model, get_model_type, get_model_config -from megatron.utils import print_rank_0, unwrap_model +from megatron.utils import unwrap_model from megatron.model import DistributedDataParallel as LocalDDP from megatron.model import Float16Module from megatron.utils import Profile @@ -23,6 +24,8 @@ # Types Shape = Union[List[int], torch.Size] +logger = ezpz.get_logger(__name__) + def get_forward_backward_func(): """Retrieves the appropriate forward_backward function given the @@ -141,9 +144,9 @@ def custom_backward(output, grad_output): grad have the same shape, while C++'s 'backward' does not. """ - assert ( - output.numel() == 1 - ), "output should be pseudo-'freed' in schedule, to optimize memory" + assert output.numel() == 1, ( + "output should be pseudo-'freed' in schedule, to optimize memory" + ) assert isinstance(output, torch.Tensor), "output == '%s'." % type(output).__name__ assert isinstance(grad_output, (torch.Tensor, type(None))), ( "grad_output == '%s'." % type(grad_output).__name__ @@ -264,10 +267,13 @@ def backward_step( if config.timers is not None: config.timers("backward-compute").stop() if len(to_skip) > 0 and args.iteration in [int(i) for i in to_skip]: - print_rank_0(f"Caught {args.iteration=} in `iters_to_skip`! Skipping!") + logger.info(f"Caught {args.iteration=} in `iters_to_skip`! Skipping!") return [None] if args.deepspeed: + from deepspeed import DeepSpeedEngine + assert model is not None + assert isinstance(model, DeepSpeedEngine) # Retain the grad on the input_tensor. unwrap_input_tensor_grad = False if not isinstance(input_tensor, list): @@ -312,6 +318,37 @@ def backward_step( input_tensor_grad[-1].add_(output_tensor_grad[1]) if unwrap_input_tensor_grad: input_tensor_grad = input_tensor_grad[0] + + # Add check to see if any NanNs in gradients + # if so: + # 1. Identify rank, print info to std out + # 2. Set NaN values to zero + # 3. Proceed with backprop + if input_tensor_grad is not None: + ezpz.breakpoint(0) + if not unwrap_input_tensor_grad: + for idx, x in enumerate(input_tensor_grad): + if torch.isnan(x).any(): + logger.critical( + " ".join( + [ + f"[{ezpz.get_hostname()}][{ezpz.get_rank()}] ", + f"NaN detected in input_tensor_grad[{idx}]!! Setting to zero.", + ] + ) + ) + input_tensor_grad[idx][torch.isnan(x)] = 0.0 + else: + if torch.isnan(input_tensor_grad).any(): + logger.critical( + " ".join( + [ + f"[{ezpz.get_hostname()}][{ezpz.get_rank()}] ", + f"NaN detected in input_tensor_grad!! Setting to zero.", + ] + ) + ) + input_tensor_grad[torch.isnan(input_tensor_grad)] = 0.0 if config.timers is not None: config.timers("backward-compute").stop() return input_tensor_grad @@ -340,14 +377,14 @@ def forward_backward_no_pipelining( """ if isinstance(model, list): - assert ( - len(model) == 1 - ), "non-pipeline-parallel schedule does not support model chunking" + assert len(model) == 1, ( + "non-pipeline-parallel schedule does not support model chunking" + ) model = model[0] if isinstance(data_iterator, list): - assert ( - len(data_iterator) == 1 - ), "non-pipeline-parallel schedule does not support model chunking" + assert len(data_iterator) == 1, ( + "non-pipeline-parallel schedule does not support model chunking" + ) data_iterator = data_iterator[0] config = get_model_config(model) @@ -359,13 +396,18 @@ def forward_backward_no_pipelining( no_sync_func = contextlib.nullcontext args = get_args() + assert args is not None if args.deepspeed: + from deepspeed import DeepSpeedEngine + + assert isinstance(model, DeepSpeedEngine) model.set_gradient_accumulation_boundary(False) model_type = get_model_type(model) forward_data_store = [] - input_tensor, output_tensor_grad = None, None + input_tensor = None + output_tensor_grad = None with no_sync_func(): for i in dlp.iter(range(num_microbatches - 1)): output_tensor = forward_step( @@ -428,15 +470,15 @@ def forward_backward_pipelining_with_interleaving( communication between pipeline stages as needed. Returns dictionary with losses if the last stage, empty dict otherwise.""" - assert isinstance( - model, list - ), "interleaved pipeline parallelism expected model chunking" - assert all( - isinstance(chunk, torch.nn.Module) for chunk in model - ), "invalid model chunking" - assert isinstance( - data_iterator, list - ), "interleaved pipeline parallelism expected each model chunk to have a data iterator" + assert isinstance(model, list), ( + "interleaved pipeline parallelism expected model chunking" + ) + assert all(isinstance(chunk, torch.nn.Module) for chunk in model), ( + "invalid model chunking" + ) + assert isinstance(data_iterator, list), ( + "interleaved pipeline parallelism expected each model chunk to have a data iterator" + ) config = get_model_config(model[0]) if config.overlap_p2p_comm and config.batch_p2p_comm: @@ -689,7 +731,6 @@ def backward_step_helper(microbatch_id): bwd_wait_handles = None for k in range(num_warmup_microbatches): - if fwd_wait_handles is not None: for req in fwd_wait_handles: req.wait() @@ -1159,14 +1200,14 @@ def forward_backward_pipelining_without_interleaving( Returns dictionary with losses if the last stage, empty dict otherwise.""" if isinstance(model, list): - assert ( - len(model) == 1 - ), "non-interleaved pipeline parallelism does not support model chunking" + assert len(model) == 1, ( + "non-interleaved pipeline parallelism does not support model chunking" + ) model = model[0] if isinstance(data_iterator, list): - assert ( - len(data_iterator) == 1 - ), "non-pipeline-parallel schedule does not support model chunking" + assert len(data_iterator) == 1, ( + "non-pipeline-parallel schedule does not support model chunking" + ) data_iterator = data_iterator[0] config = get_model_config(model) @@ -1353,7 +1394,6 @@ def enable_grad_sync(): # Run cooldown backward passes. if not forward_only: for i in range(num_warmup_microbatches): - # Enable async grad reduction in the last backward pass # Note: If grad sync function is provided, only enable # async grad reduction in first pipeline stage. Other From e966d29008dd64c020073f043d26f251f6076dec Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 11 Sep 2025 16:14:34 -0500 Subject: [PATCH 490/590] chore: Prevent checkpoint timers from bringing down training --- megatron/training.py | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index ab1942bfcc7..385ed9064aa 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -1,3 +1,4 @@ +# noqa: E402 # Copyright (C) 2024 Habana Labs, Ltd. an Intel Company. # Copyright (c) 2023, NVIDIA CORPORATION. All rights reserved. """Pretrain utilities.""" @@ -7,6 +8,8 @@ # The earliest we can measure the start time. _TRAIN_START_TIME = time.time() +import ezpz + from collections import OrderedDict from datetime import datetime import json @@ -813,10 +816,17 @@ def setup_model_and_optimizer( if args.load is not None: timers = get_timers() assert timers is not None - timers("load-checkpoint", log_level=0).start(barrier=True) + # timers("load-checkpoint", log_level=0).start(barrier=True) + t0 = time.perf_counter() args.iteration = load_checkpoint(model, optimizer, opt_param_scheduler) - timers("load-checkpoint").stop(barrier=True) - timers.log(["load-checkpoint"]) + ezpz.dist.synchronize() + dtl = time.perf_counter() - t0 + try: + wandb.log({"timers/load-checkpoint": dtl}, step=args.iteration) + except Exception: + log.info(f"timers/load-checkpoint took {dtl:.3f} seconds") + # timers("load-checkpoint").stop(barrier=True) + # timers.log(["load-checkpoint"]) else: args.iteration = 0 else: From b4a0e7a9789c7bb7eecc84adc4a1193d7ed8ba67 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 11 Sep 2025 16:14:57 -0500 Subject: [PATCH 491/590] chore: Prevent checkpoint timers from bringing down training --- megatron/timers.py | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/megatron/timers.py b/megatron/timers.py index cd2c9c206d1..89b2da3795e 100644 --- a/megatron/timers.py +++ b/megatron/timers.py @@ -216,9 +216,8 @@ def _get_global_min_max_time(self, names, reset, barrier, normalizer): name_to_min_max_time = {} for i, name in enumerate(names): rank_to_time = rank_name_to_time[:, i] - # filter out the ones we did not have any timings for try: - rank_to_time = rank_to_time[rank_to_time > 0.0] + times = rank_to_time[rank_to_time > 0.0] except Exception: # this can happen if rank_to_time is not a tensor raise Exception( @@ -226,12 +225,22 @@ def _get_global_min_max_time(self, names, reset, barrier, normalizer): name, type(rank_to_time) ) ) + # filter out the ones we did not have any timings for + # try: + # rank_to_time = rank_to_time[rank_to_time > 0.0] + # except Exception: + # # this can happen if rank_to_time is not a tensor + # raise Exception( + # "timer {} did not return a tensor, got {}".format( + # name, type(rank_to_time) + # ) + # ) # If the timer exists: - if rank_to_time.numel() > 0: + if times.numel() > 0: name_to_min_max_time[name] = ( - rank_to_time.min().item() / normalizer, - rank_to_time.max().item() / normalizer, + times.min().item() / normalizer, + times.max().item() / normalizer, ) return name_to_min_max_time From c72e2ebd25a37fb18967cd056fbb9f1b4e76a1d1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 11 Sep 2025 16:15:24 -0500 Subject: [PATCH 492/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 304 ++++++++++++++++++++++++++++++------------------ 1 file changed, 190 insertions(+), 114 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index b7053dc37d4..fc8e273d900 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -8,6 +8,7 @@ ############################################################################### # Source: # [`ezpz/bin/utils.sh`](https://github.com/saforem2/ezpz/blob/main/src/ezpz/bin/utils.sh) +# shellcheck disable=SC1090 source <(curl -L https://bit.ly/ezpz-utils) >/dev/null || return 1 ezpz_setup_job >/dev/null || exit ############################################################################### @@ -47,7 +48,7 @@ helpers_main() { echo "Using ${WORKING_DIR} as working directory..." fi export WORKING_DIR="${WORKING_DIR}" - printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" + log_message INFO "Using WORKING_DIR: ${WORKING_DIR}" # \n" "${WORKING_DIR}" } ############################################################################## @@ -153,7 +154,7 @@ setup_run_cmd() { setupLauncher "${LAUNCH_WITH:-MPICH}" || exit export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" printf "\n" - echo "Using data_cache_path: ${data_cache_path}" + log_message INFO "Using data_cache_path: ${data_cache_path}" ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` @@ -207,7 +208,7 @@ setup_run_cmd() { "--adam-beta1=${ADAM_BETA1:-0.9}" "--adam-beta2=${ADAM_BETA2:-0.95}" "--adam-eps=${ADAM_EPS:-0.00001}" - "--blend-sample-in-corpus" + # "--blend-sample-in-corpus" "--clip-grad=${CLIP_GRAD:-1.0}" "--data-cache-path=${data_cache_path}" "--data-file-list=${DATA_FILE_LIST:-${dfl_fallback}}" @@ -260,7 +261,7 @@ setup_run_cmd() { ["SEQ"]="${SEQ}" ["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE}" ["NUM_KV_HEAD"]="${NUM_KV_HEAD}" - ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" + ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING}" ["DTYPE"]="${DTYPE}" ["OPT"]="${OPT}" # ["WEIGHT_DECAY"]="${WEIGHT_DECAY}" @@ -270,7 +271,8 @@ setup_run_cmd() { # ["CLIP_GRAD"]="${CLIP_GRAD:-1.0}" ) for v in "${!arch_map[@]}"; do - printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" + log_message INFO "${GREEN}${v}${RESET}: ${BLUE}${arch_map[${v}]}${RESET}" + # printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" done printf "======================\n" # "--adam-eps ${ADAM_EPS:-0.00001}" @@ -279,7 +281,7 @@ setup_run_cmd() { targs_cache="${cache_dir}/train_args.txt" for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") - printf "Training Arguments: %s\n" "$(printBlue "${TRAIN_ARGS[@]}")" + log_message INFO "Training Arguments: ${BLUE}${TRAIN_ARGS[@]}${RESET}" # \n" "$(printBlue "${TRAIN_ARGS[@]}")" export run_cmd=("${LAUNCHER}" "${train_args[@]}") } @@ -319,14 +321,14 @@ setupSrun() { } printJobInfo() { - echo "++++++++++++++++++++++++++++++++++++++++++++++++++" - echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" - echo "- Using $(which python3)" - echo "- WORLD_SIZE:${WORLD_SIZE-}" - echo "- BACKEND: ${BE:-}" - echo "- MODEL_TYPE: ${MODEL_TYPE:-}" - echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" - echo "++++++++++++++++++++++++++++++++++++++++++++++++++" + log_message INFO "++++++++++++++++++++++++++++++++++++++++++++++++++" + log_message INFO "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" + log_message INFO "- Using $(which python3)" + log_message INFO "- WORLD_SIZE:${WORLD_SIZE-}" + log_message INFO "- BACKEND: ${BE:-}" + log_message INFO "- MODEL_TYPE: ${MODEL_TYPE:-}" + log_message INFO "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" + log_message INFO "++++++++++++++++++++++++++++++++++++++++++++++++++" } ############################################################################# @@ -453,9 +455,9 @@ get_grad_acc_steps_on_aurora() { elif [[ "$#" == 1 ]]; then hf="$1" else - echo "Usage: get_grad_acc_steps_on_aurora" - echo "Expected exactly 0 or 1 arguments, received: $#" - exit 1 + log_message INFO "Usage: get_grad_acc_steps_on_aurora" + log_message INFO "Expected exactly 0 or 1 arguments, received: $#" + return 1 fi nhosts=$(wc -l <"${hf}") if [[ "${nhosts}" -ge 256 ]]; then # n >= 256 @@ -501,18 +503,115 @@ set_ccl_vars_on_aurora() { export CCL_BCAST=double_tree } + +# ================================================ +# AuroraGPT Model Variants +# ================================================ +get_model_arch_70B() { + HEADS=64 + NLAYERS=80 + NUM_KV_HEAD=8 + FFN_HIDDEN_SIZE=28672 + HIDDEN=8192 + SEQ=8192 + export MODEL_ARCH="AuroraGPT-70B" +} + +get_model_arch_33B() { + # 33B + export MODEL_KEY="AuroraGPT-33B" + export TP=6 + export ZERO_STAGE=2 + export GRAD_ACC_STEPS=1 + # export USE_ACTIVATION_CHECKPOINTING=1 + export MICRO_BATCH=1 + export NLAYERS=78 + export HIDDEN=8268 + export HEADS=78 + export FFN_HIDDEN_SIZE=11076 + export SEQ=4096 + export NUM_KV_HEAD=6 + export MODEL_ARCH="AuroraGPT-33B" +} + get_model_arch_7B() { # 7B # export MODEL_KEY="AuroraGPT-7B" - export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS - export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS - export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE - export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION - export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE - export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - export MODEL_ARCH="AuroraGPT-7B" + declare -A model_config + model_config["HEADS"]="${HEADS:-32}" + model_config["NLAYERS"]="${NLAYERS:-32}" + model_config["HIDDEN"]="${HIDDEN:-4096}" + model_config["NUM_KV_HEAD"]="${NUM_KV_HEAD:-8}" + model_config["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE:-11008}" + model_config["SEQ"]="${SEQ:-4096}" + model_config["MODEL_ARCH"]="AuroraGPT-7B" + model_config["ZERO_STAGE"]="${ZERO_STAGE:-1}" + model_config["MICRO_BATCH"]="${MICRO_BATCH:-1}" + model_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS:-${GAS:-4}}" + model_config["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" + model_config["LR_DECAY_STYLE"]="${LR_DECAY_STYLE:-cosine}" + model_config["TOKENIZER_TYPE"]="${TOKENIZER_TYPE:-HFTokenizer}" + model_config["TOKENIZER_MODEL"]="${TOKENIZER_MODEL:-"meta-llama/llama-2-7b-hf"}" + for key in "${!model_config[@]}"; do + log_message INFO "[AuroraGPT-7B] Setting ${GREEN}${key}${RESET}=${BLUE}${model_config[${key}]}${RESET}" + export "${key}"="${model_config[${key}]}" + done + # export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + # export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + # export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + # export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + # export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE + # export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + # export model_config="AuroraGPT-7B" + # export ZERO_STAGE="${ZERO_STAGE:-1}" + # export MICRO_BATCH="${MICRO_BATCH:-1}" + # export GRAD_ACC_STEPS=:"${GRAD_ACC_STEPS:-4}" + # export USE_ACTIVATION_CHECKPOINTING="${USE_ACTIVATION_CHECKPOINTING:-0}" + # export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" + # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" + # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-meta-llama/llama-2-7b-hf}" } +get_model_arch_2B() { + # 2B + declare -A model_config + model_config["HEADS"]="${HEADS:-16}" + model_config["NLAYERS"]="${NLAYERS:-12}" + model_config["HIDDEN"]="${HIDDEN:-2048}" + model_config["NUM_KV_HEAD"]="${NUM_KV_HEAD:-4}" + model_config["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE:-11008}" + model_config["SEQ"]="${SEQ:-8192}" + model_config["MODEL_ARCH"]="AuroraGPT-2B" + model_config["ZERO_STAGE"]="${ZERO_STAGE:-0}" + model_config["MICRO_BATCH"]="${MICRO_BATCH:-1}" + model_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS:-2}" + model_config["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" + model_config["LR_DECAY_STYLE"]="${LR_DECAY_STYLE:-constant}" + model_config["TOKENIZER_TYPE"]="${TOKENIZER_TYPE:-HFTokenizer}" + model_config["TOKENIZER_MODEL"]="${TOKENIZER_MODEL:-google/gemma-7b}" + for key in "${!model_config[@]}"; do + log_message INFO "[AuroraGPT-2B] Setting ${GREEN}${key}${RESET}=${BLUE}${model_config[${key}]}${RESET}" + export "${key}"="${model_config[${key}]}" + done + # export HEADS=16 + # export NLAYERS=12 + # export HIDDEN=2048 + # export NUM_KV_HEAD=4 + # export FFN_HIDDEN_SIZE=11008 + # export SEQ=8192 + # export model_config="AuroraGPT-2B" + # export ZERO_STAGE="${ZERO_STAGE:-0}" + # export MICRO_BATCH="${MICRO_BATCH:-1}" + # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-2}" + # export USE_ACTIVATION_CHECKPOINTING="${USE_ACTIVATION_CHECKPOINTING:-0}" + # export LR_DECAY_STYLE="${LR_DECAY_STYLE:-constant}" + # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" + # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-google/gemma-7b}" +} + +# ================================================ +# Additional Model Variants +# ================================================ get_model_arch_llama3_3B() { export HEADS=24 export NLAYERS=28 @@ -573,49 +672,6 @@ get_model_arch_phi4_mini_custom_nLayers() { export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" } -# get_model_arch_70B() { -# # 70B -# export MODEL_KEY="AuroraGPT-70B" -# export TP=6 -# export ZERO_STAGE=2 -# export GRAD_ACC_STEPS=1 -# export USE_ACTIVATION_CHECKPOINTING=1 -# export MICRO_BATCH=1 -# export NLAYERS=80 -# export HIDDEN=8192 -# export HEADS=96 -# export FFN_HIDDEN_SIZE=11040 -# export SEQ=8192 -# export NUM_KV_HEAD=6 -# } -# -get_model_arch_70B() { - HEADS=64 - NLAYERS=80 - NUM_KV_HEAD=8 - FFN_HIDDEN_SIZE=28672 - HIDDEN=8192 - SEQ=8192 - export MODEL_ARCH="AuroraGPT-70B" -} - -get_model_arch_33B() { - # 33B - export MODEL_KEY="AuroraGPT-33B" - export TP=6 - export ZERO_STAGE=2 - export GRAD_ACC_STEPS=1 - # export USE_ACTIVATION_CHECKPOINTING=1 - export MICRO_BATCH=1 - export NLAYERS=78 - export HIDDEN=8268 - export HEADS=78 - export FFN_HIDDEN_SIZE=11076 - export SEQ=4096 - export NUM_KV_HEAD=6 - export MODEL_ARCH="AuroraGPT-33B" -} - ############################################################################## # setParams # @@ -641,13 +697,13 @@ setParams() { export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS - gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") - export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" + # gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") + # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas:-1}}" # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" - echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - MICRO_BATCH=${MICRO_BATCH:-1} + # echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" + # MICRO_BATCH=${MICRO_BATCH:-1} if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" + log_message WARN "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-builder" fi @@ -664,11 +720,11 @@ setParams() { # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN:-}" ]]; then - echo "Not using flash-attn!!" + log_message WARN "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi - echo "Setting up AWS NCCL OFI Plugin on Polaris..." + log_message INFO "Setting up AWS NCCL OFI Plugin on Polaris..." source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # ---- [Sophia] ---------------------- elif [[ "${mn}" == sophia* ]]; then @@ -680,7 +736,7 @@ setParams() { export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" + log_message WARN "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi @@ -694,17 +750,23 @@ setParams() { export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-1}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then - echo "Not using flash-attn!!" + log_message WARN "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi fi - ma="${MODEL_ARCH:-7B}" + ma="${MODEL_ARCH:-"AuroraGPT-2B"}" + case "${ma}" in - # "70B" | "llama-3.1-70B" | "llama-3.1-70b" | "llama-3.2-70B" | "llama-3.2-70b") - "70B") + "70B" | "AuroraGPT-70B" | "aurora-gpt-70b") get_model_arch_70B ;; + "7B" | "AuroraGPT-7B" | "aurora-gpt-7b") + get_model_arch_7B + ;; + "2B" | "AuroraGPT-2B" | "aurora-gpt-2b") + get_model_arch_2B + ;; "33B" | "llama-3.2-33B" | "llama-3.2-33b") get_model_arch_33B ;; @@ -717,11 +779,8 @@ setParams() { "llama3-3B" | "llama-3B") get_model_arch_llama3_3B_customNlayers ;; - "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") - get_model_arch_7B - ;; *) - get_model_arch_7B + get_model_arch_3B ;; esac export TP="${TP}" @@ -758,29 +817,45 @@ setParams() { lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") fi # +---[Run Settings]------------------------------------------------------+ - export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE - export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS + # export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE + # export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE + export GRAD_ACC_STEPS="${GRAD_ACC_STEPS}" # GRADIENT ACCUMULATION STEPS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE export DP=$((WORLD_SIZE / TP / PP / SP)) # DATA PARALLELISM export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" + declare -A parallelism_config + parallelism_config["WORLD_SIZE"]="${WORLD_SIZE}" + parallelism_config["DP"]="${DP}" + parallelism_config["TP"]="${TP}" + parallelism_config["PP"]="${PP}" + parallelism_config["SP"]="${SP}" + parallelism_config["MICRO_BATCH"]="${MICRO_BATCH}" + parallelism_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS}" + parallelism_config["GLOBAL_BATCH"]="${GLOBAL_BATCH}" + parallelism_config["GLOBAL_BATCH_MAX"]="${GLOBAL_BATCH_MAX}" + for key in "${!parallelism_config[@]}"; do + log_message INFO "[Parallelism] Setting ${GREEN}${key}${RESET}=${BLUE}${parallelism_config[${key}]}${RESET}" + export "${key}"="${parallelism_config[${key}]}" + done + # echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" if [[ -n "${TRAIN_TOKENS:-}" ]]; then export TRAIN_TOKENS="${TRAIN_TOKENS}" export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) - printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then export TRAIN_TOKENS=${TRAIN_TOKENS:-4673780159710} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) - printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + # log_message INFO "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + # log_message INFO "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" else export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" + TRAIN_TOKENS=$((TRAIN_ITERS * SEQ * GLOBAL_BATCH)) + export TRAIN_TOKENS fi + log_message INFO "TRAIN_TOKENS=${TRAIN_TOKENS} (=$((TRAIN_TOKENS / 10 ** 9))B tokens)" # \n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + log_message INFO "TRAIN_ITERS=${TRAIN_ITERS}" # \n" "${TRAIN_ITERS}" export MODEL_TYPE="${MODEL_ARCH:-AuroraGPT}-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # NOTE: [2024-07-10] ##################################################### # - [sam]: For whatever reason, it seems that using @@ -823,7 +898,7 @@ set_args() { ds_args+=("--zero-stage=$ZERO_STAGE") # if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then if [[ "${USE_ACTIVATION_CHECKPOINTING:-}" == 1 || "${USE_ACTIVATION_CHECKPOINTING:-}" == "true" ]]; then - echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" + log_message WARN "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args+=("--deepspeed-activation-checkpointing") ds_args+=( "--checkpoint-activations" @@ -868,7 +943,7 @@ ezpz_test() { # forwarded to ALL ranks with DeepSpeed ############################################################################ saveDSenv() { - echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" + log_message INFO "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { echo "PATH=${PATH:-}" echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH:-}" @@ -989,14 +1064,15 @@ make_data() { ############################################################################## install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" - echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." - python3 -m pip install -r "${depsfile}" --require-virtualenv + log_message INFO "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." + python3 -m pip install -r "${depsfile}" > /dev/null || { # --require-virtualenv + log_message INFO "[install_dependencies] Failed to install dependencies from ${depsfile} on $(ezpz_get_machine_name) in $(which python3)\n" + } if [[ ! -x "$(command -v deepspeed)" ]]; then - printf "[install_dependencies] No 'deepspeed' command found on %s in %s\n" "$$(ezpz_get_machine_name)" "$(which python3)" - printf "[install_dependencies] Attempting to install deepspeed via pip...\n" - python3 -m pip install deepspeed --require-virtualenv || { - printf "[install_dependencies] Failed to install deepspeed via pip on %s\n" "$(ezpz_get_machine_name)" - # printf "[install_dependencies] !! No deepsepeed in %s\n" "$(which python3)" + log_message INFO "[install_dependencies] No 'deepspeed' command found on $(ezpz_get_machine_name) in $(which python3)\n" + log_message INFO "[install_dependencies] Attempting to install deepspeed via pip...\n" + python3 -m pip install deepspeed > /dev/null || { # --require-virtualenv || { + log_message INFO "[install_dependencies] Failed to install deepspeed via pip on $(ezpz_get_machine_name)\n" return 1 } # mn=$(ezpz_get_machine_name) @@ -1083,7 +1159,7 @@ makeHostfiles() { setup_tokenizer_and_data() { if [[ "$#" == 1 ]]; then tok="$1" - dfl="${DATA_FILE_LIST:-}" + dfl="${DATA_FILE_LIST:-ALCF/data-lists/$(ezpz_get_machine_name)/olmo-mix-11.txt}" elif [[ "$#" == 2 ]]; then tok="$1" dfl="$2" @@ -1131,8 +1207,8 @@ setup_tokenizer_and_data() { fi export DATA_FLAGS="${_data_flags[*]:-}" export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" - printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" - printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" + log_message INFO "[setData] DATA_FLAGS: ${GREEN}${DATA_FLAGS}${RESET}" # \n" "$(printGreen "${DATA_FLAGS}")" + log_message INFO "[setData] TOKENIZER_FLAGS: ${MAGENTA}${TOKENIZER_FLAGS}${RESET}" # \n" "$(printMagenta "${TOKENIZER_FLAGS}")" } ############################################### @@ -1149,7 +1225,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # set `dfl` to `dfl_fallback` if not passed as an argument, # use this data file list to call `setData` dfl="${1:-${dfl_fallback}}" - printf "Calling: setData() with %s\n" "${dfl}" + log_message INFO "Calling: setData() with ${dfl}" # \n" "${dfl}" ndocs=$(wc -l <"${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") @@ -1160,15 +1236,15 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" - echo "--------------------" - echo "Updated environment:" - printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" - printf "NUM_DOCS: %s\n " "${NUM_DOCS}" - printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" - printf "DFL_STEM: %s\n" "${DFL_STEM}" - printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" - printf "DATA_FLAGS: %s\n" "${DATA_FLAGS:-}" - echo "--------------------" + log_message INFO "--------------------" + log_message INFO "Updated environment:" + log_message INFO "DATA_FILE_LIST: ${DATA_FILE_LIST}" # \n" "${DATA_FILE_LIST}" + log_message INFO "NUM_DOCS: ${NUM_DOCS}" # %s\n " "${NUM_DOCS}" + log_message INFO "WEIGHT_SUM: ${WEIGHT_SUM}" # %s\n" "${WEIGHT_SUM}" + log_message INFO "DFL_STEM: ${DFL_STEM}" # %s\n" "${DFL_STEM}" + log_message INFO "DATA_CACHE_PATH: ${DATA_CACHE_PATH}" # %s\n" "${DATA_CACHE_PATH}" + log_message INFO "DATA_FLAGS: ${DATA_FLAGS}" # \n" "${DATA_FLAGS:-}" + log_message INFO "--------------------" } generateDSconfig_new() { From da80b206275b79547d5655160d2645ff68cb1b22 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Thu, 11 Sep 2025 16:16:01 -0500 Subject: [PATCH 493/590] chore: Update `train_aGPT_2B_large_batch.sh` --- train_aGPT_2B_large_batch.sh | 85 ++++++++++++++++++------------------ 1 file changed, 43 insertions(+), 42 deletions(-) diff --git a/train_aGPT_2B_large_batch.sh b/train_aGPT_2B_large_batch.sh index b4053262e6e..b2a8189345b 100755 --- a/train_aGPT_2B_large_batch.sh +++ b/train_aGPT_2B_large_batch.sh @@ -5,50 +5,51 @@ #PBS -l walltime=06:00:00,filesystems=flare:home #PBS -l select=256 -cd /flare/datascience/foremans/projects/argonne-lcf/Megatron-DeepSpeed || { - echo "Failed to change directory to /flare/datascience/foremans/projects/argonne-lcf/Megatron-DeepSpeed" - exit 1 -} - -PBS_O_WORKDIR="$(pwd)" -export PBS_O_WORKDIR +# Get the directory of this script +# whereami="$(dirname "$(realpath "$0")")" +# cd "${whereami}" || { +# echo "Failed to change directory to ${whereami}" +# exit 1 +# } -export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" -export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" -export http_proxy="http://proxy.alcf.anl.gov:3128" -export https_proxy="http://proxy.alcf.anl.gov:3128" -export ftp_proxy="http://proxy.alcf.anl.gov:3128" +setup_env() { + cd /flare/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/large-batch-training/tok50M-n512/Megatron-DeepSpeed || { + echo "Failed to change directory to /flare/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/large-batch-training/tok50M-n512/Megatron-DeepSpeed" + exit 1 + } + PBS_O_WORKDIR="$(pwd)" + export PBS_O_WORKDIR + export http_proxy="http://proxy.alcf.anl.gov:3128" + export https_proxy="http://proxy.alcf.anl.gov:3128" + export ftp_proxy="http://proxy.alcf.anl.gov:3128" + export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" + export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" + # shellcheck disable=SC1090 + source <(curl -L https://bit.ly/ezpz-utils) + ezpz_setup_env + log_message INFO "Using: $(which python3)" +} -# shellcheck disable=SC1090 -source <(curl -L https://bit.ly/ezpz-utils) -ezpz_setup_env +train_model() { + MODEL_ARCH=smollm3-3B \ + OPT=ipex.fusedlamb \ + NLAYERS=12 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" \ + "$@" +} -# eval "$(micromamba shell hook --shell "$(basename "${SHELL}")")" -# ezpz_load_new_pt_modules_aurora -# micromamba activate /flare/datascience_collab/foremans/micromamba/envs/pt29-2025-07 -# -# NODES=$(cat "${PBS_NODEFILE}" | uniq | wc -l) -# mpiexec -np "${NODES}" --ppn 1 python3 \ -# /flare/datascience/foremans/projects/argonne-lcf/scalable_conda_env/cache_soft.py \ -# --src=/flare/datascience/foremans/pt29-ezpz-2025-07.tar.gz \ -# --dst=/tmp/pt29-ezpz-2025-07.tar.gz \ -# --d -# -# micromamba deactivate -# micromamba activate /tmp/pt29-ezpz-2025-07 -# which python3 +main() { + setup_env + train_model "$@" +} -MODEL_ARCH=smollm3-3B \ - OPT=ipex.fusedlamb \ - NLAYERS=12 \ - GRAD_ACC_STEPS=2 \ - MICRO_BATCH=1 \ - USE_ACTIVATION_CHECKPOINTING=0 \ - ZERO_STAGE=0 \ - OPT=adamw \ - LR_DECAY_STYLE=constant \ - TOKENIZER_TYPE=HFTokenizer \ - TOKENIZER_MODEL=google/gemma-7b \ - DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ - bash "${PBS_O_WORKDIR}/train_alcf.sh" +main "$@" From f59df1d1985102a95b5861911079aebb95f22cbc Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 12:20:32 -0500 Subject: [PATCH 494/590] custom optimizers, schedulers, hp tuning and CPT --- ALCF/notes/lb_optimizers_settings_and_cpt.md | 78 ++++++++++++++++++++ 1 file changed, 78 insertions(+) create mode 100644 ALCF/notes/lb_optimizers_settings_and_cpt.md diff --git a/ALCF/notes/lb_optimizers_settings_and_cpt.md b/ALCF/notes/lb_optimizers_settings_and_cpt.md new file mode 100644 index 00000000000..961c39a4bd5 --- /dev/null +++ b/ALCF/notes/lb_optimizers_settings_and_cpt.md @@ -0,0 +1,78 @@ +# Megatron-DeepSpeed, optimizers, hyperparameters and CPT + +Single command to test and run Megatron-DeepSpeed: + +```bash +now=$(date +'%Y-%m-%d-%H%M%S') && debug_dir="${now}" && mkdir -p "${debug_dir}"&& cd "${debug_dir}"&& git clone https://github.com/argonne-lcf/Megatron-DeepSpeed && cd Megatron-DeepSpeed && source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_env && python3 -m pip install --require-virtualenv "git+https://github.com/saforem2/ezpz" "numpy<2" deepspeed tensorboard && ezpz-test && DATA_FILE_LIST=ALCF/data-lists/aurora/books.txt bash train_alcf.sh +``` +## Optimizers +The default optimizer is `adamw`. Go to this [list of optimizers](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/arguments.py#L1485) for a complete list of supported optimizers (note that dshampoo might throw checkpointing errors, we are working on fixing this). For example, to run with `muon`, you can do: +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh +``` +Here +```bash +DATA_FILE_LIST=path/to/your/tokenized/data +TRAIN_TOKENS= number of training tokens +GRAD_ACC_STEPS=number of grad accumulation steps +LR=learning rate +LR_WARMUP_FRACTION=warmup fraction +OPT=optimizer +``` +Your global batch size will be: `num_gpus*micro_batch_size*GRAD_ACC_STEPS`, micro batch size is 1 by default, you can change it by adding `MICRO_BATCH=new_micro_batch_size` to your options. To have the corresponding number if tokens per step, you need to multiply the global batch size by the sequence length (set with `SEQ_LEN`, default is 4096) + +### Adding custom optimizers +To add a custom optimizer, you have to modify the following files: +- `megatron/optimizer/__init__.py`: [muon example](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/optimizer/__init__.py#L434), note that you either heve to import the optimizer from a pre-installed package or add it in the `megatron/optimizer/` folder. +- `megatron/arguments.py`: [optimizer arguments](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/arguments.py#L1070), to add the optimizer arguments +- `megatron/arguments.py`: [list of valid optimizers](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/arguments.py#L1485), to add the new optimizer to the list of valid optimizers + +### Schedulers +Note that the default scheduler is `cosine`. We also support `infinite cosine, infinite inverse square root, constant, constant with cooldown, inverse square root, linear` schedulers. For example to change the scheduler to `constant`, you can do so with the `LR_DECAY_STYLE` option: +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh +``` +To add cooldown, you need to add the `--lr_constant_plus_cooldown` flag and set the cooldown fraction with `--lr_constant_plus_cooldown_frac`. The default cooldown fraction is 0.05 +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --lr_constant_plus_cooldown_frac 0.01 +``` +#### Adding custom schedulers +To add a custom scheduler, you have to modify the following files: +- `megatron/optimizer_param_scheduler.py`: [schedulers](megatron/optimizer_param_scheduler.py), to add the new scheduler +- `megatron/arguments.py`: [list of LR arguments](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/arguments.py#L1671), to add the new scheduler arguments. +- You might have to change [the function](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/994f2a129d465cc50e6c35af075eb3292874effe/megatron/training.py#L559) to incorporate your custom scheduler options. + +## Hyperparameter tuning +#### Init variance +Weight initialization is key to training LLMs,and to avoid spikes in losses. Here, we initialize the wights following this [paper](https://arxiv.org/pdf/2312.16903). The default variance value at initialization is 0.02. To add custom variances, one can use `--init-method-std, `--adjust-word-embedding-init`, and `--word-embedding-init-std`. For our runs, we do +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 +``` +where `d=hidden size`. + +### Learning rate +For the learning rate, we implemented the learning rate finder routine [here](https://sgugger.github.io/how-do-you-find-a-good-learning-rate.html) and [here](https://arxiv.org/pdf/1506.01186). This is activated with the `--lr-finder` and run for `TRAIN_ITERS` steps. For example, for a 1000 steps: +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 --lr-finder +``` +### Maximal Update Parametrization/Complete Parametrization +We have MuP and CompleteP incorporated in AuroraGPT. Will add details about activating them soon + + + + +## Doing CPT +Build your cpt dataset, here I am mixing the lucid papers with weight 0.9 and dolma with weight 0.1: +```bash +python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt +``` +here is the cpt command to run from the Megatron-deepspeed folder: +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune +``` +Here the following options options/flags should be: +```bash +DATA_FILE_LIST=path/to/your/tokenized/data +LOAD=path/to/your/universal/checkpoint +SAVE=path/to/where/you/want/to/save/checkpoints +``` From 74b4cf3ad292751d6ca1664153c955314c5dcd2d Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 18:23:02 -0500 Subject: [PATCH 495/590] Update lb_optimizers_settings_and_cpt.md --- ALCF/notes/lb_optimizers_settings_and_cpt.md | 22 ++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/lb_optimizers_settings_and_cpt.md b/ALCF/notes/lb_optimizers_settings_and_cpt.md index 961c39a4bd5..78d908fdd49 100644 --- a/ALCF/notes/lb_optimizers_settings_and_cpt.md +++ b/ALCF/notes/lb_optimizers_settings_and_cpt.md @@ -62,17 +62,35 @@ We have MuP and CompleteP incorporated in AuroraGPT. Will add details about acti ## Doing CPT -Build your cpt dataset, here I am mixing the lucid papers with weight 0.9 and dolma with weight 0.1: +### Legacy agpt-7b checkpoints +This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the replay+rewarm one where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: +1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here]($ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder +). +2. Build your cpt dataset, here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -here is the cpt command to run from the Megatron-deepspeed folder: +3. Then, we can run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune ``` +Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. Here the following options options/flags should be: ```bash DATA_FILE_LIST=path/to/your/tokenized/data LOAD=path/to/your/universal/checkpoint SAVE=path/to/where/you/want/to/save/checkpoints +--universal-checkpoint to load a universal checkpoint ``` +### New agpt runs (phase 1 -> phase 2) +For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. + +To do CPT here, +1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max BEFORE COOLING DOWN** +2. Mix the datasets as above +3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm +4. Run +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown +``` +5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint (i.e after cooling it down) and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. From d17ccadca8e7c1dd9261b3db7963f38bc262cacb Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 18:24:55 -0500 Subject: [PATCH 496/590] details about doing cpt --- ALCF/notes/cpt.md | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) create mode 100644 ALCF/notes/cpt.md diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md new file mode 100644 index 00000000000..4ce7e56814d --- /dev/null +++ b/ALCF/notes/cpt.md @@ -0,0 +1,33 @@ +# CPT +## Legacy agpt-7b checkpoints +This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the replay+rewarm one where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: +1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here]($ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder +). +2. Build your cpt dataset, here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): +```bash +python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt +``` +3. Then, we can run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune +``` +Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. For example, we tested rewarming to LR/2 and 2LR as well. +Here the following options options/flags should be: +```bash +DATA_FILE_LIST=path/to/your/tokenized/data +LOAD=path/to/your/universal/checkpoint +SAVE=path/to/where/you/want/to/save/checkpoints +--universal-checkpoint to load a universal checkpoint +``` +## New agpt runs (phase 1 -> phase 2) +For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. + +To do CPT here, +1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max BEFORE COOLING DOWN** +2. Mix the datasets as above +3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm +4. Run +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown +``` +5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint (i.e after cooling it down) and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. From 55f7963ccffee51c2b9100188fead42b47d5e62b Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 18:25:45 -0500 Subject: [PATCH 497/590] Update and rename lb_optimizers_settings_and_cpt.md to lb_optimizers_settings.md --- ...s_and_cpt.md => lb_optimizers_settings.md} | 37 ------------------- 1 file changed, 37 deletions(-) rename ALCF/notes/{lb_optimizers_settings_and_cpt.md => lb_optimizers_settings.md} (62%) diff --git a/ALCF/notes/lb_optimizers_settings_and_cpt.md b/ALCF/notes/lb_optimizers_settings.md similarity index 62% rename from ALCF/notes/lb_optimizers_settings_and_cpt.md rename to ALCF/notes/lb_optimizers_settings.md index 78d908fdd49..706b2aeafb3 100644 --- a/ALCF/notes/lb_optimizers_settings_and_cpt.md +++ b/ALCF/notes/lb_optimizers_settings.md @@ -57,40 +57,3 @@ DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEP ``` ### Maximal Update Parametrization/Complete Parametrization We have MuP and CompleteP incorporated in AuroraGPT. Will add details about activating them soon - - - - -## Doing CPT -### Legacy agpt-7b checkpoints -This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the replay+rewarm one where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: -1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here]($ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder -). -2. Build your cpt dataset, here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): -```bash -python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt -``` -3. Then, we can run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): -```bash -DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune -``` -Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. -Here the following options options/flags should be: -```bash -DATA_FILE_LIST=path/to/your/tokenized/data -LOAD=path/to/your/universal/checkpoint -SAVE=path/to/where/you/want/to/save/checkpoints ---universal-checkpoint to load a universal checkpoint -``` -### New agpt runs (phase 1 -> phase 2) -For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. - -To do CPT here, -1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max BEFORE COOLING DOWN** -2. Mix the datasets as above -3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm -4. Run -```bash -DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown -``` -5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint (i.e after cooling it down) and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. From f077ce62fc3660891ed49598d4d5fbda6881ab73 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 18:30:06 -0500 Subject: [PATCH 498/590] Update cpt.md --- ALCF/notes/cpt.md | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 4ce7e56814d..769986fd846 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,9 +1,8 @@ # CPT ## Legacy agpt-7b checkpoints -This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the replay+rewarm one where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: -1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here]($ cd deps/DeepSpeed && git status && cd - && ckpt_dir=checkpoints/ws48_ds_stage1_nl1_hs4096_mb1_seq4096_gb768_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash ; gs=$(cat "${ckpt_dir}/latest_checkpointed_iteration.txt") && echo "global step: ${gs}" && python3 deps/DeepSpeed/deepspeed/checkpoint/ds_to_universal.py --input_folder "${ckpt_dir}/global_step${gs}" --output_folder "${ckpt_dir}/global_step${gs}_universal" --keep_temp_folder -). -2. Build your cpt dataset, here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): +This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: +1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md). +2. Use [mix_datasets.py function](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) to build your cpt dataset. Here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` @@ -11,7 +10,7 @@ python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/ ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune ``` -Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. For example, we tested rewarming to LR/2 and 2LR as well. +Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. by just setting a different value for LR For example, we tested rewarming to LR/2 and 2LR as well. Here the following options options/flags should be: ```bash DATA_FILE_LIST=path/to/your/tokenized/data From 9a9e69c054ae8bff6407d92c88a059756f9a97b9 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 22 Sep 2025 18:30:51 -0500 Subject: [PATCH 499/590] Update cpt.md --- ALCF/notes/cpt.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 769986fd846..d5175a11417 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,8 +1,8 @@ # CPT ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: -1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions here](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md). -2. Use [mix_datasets.py function](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) to build your cpt dataset. Here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): +1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. +2. Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. Here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` From 56d0d80853bff500e0136c8f1b5c4d3c45369d58 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Sep 2025 13:41:04 -0500 Subject: [PATCH 500/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 312 +++++++++++++++++++----------------------------- 1 file changed, 126 insertions(+), 186 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index fc8e273d900..5d70ea98a45 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -48,7 +48,7 @@ helpers_main() { echo "Using ${WORKING_DIR} as working directory..." fi export WORKING_DIR="${WORKING_DIR}" - log_message INFO "Using WORKING_DIR: ${WORKING_DIR}" # \n" "${WORKING_DIR}" + printf "Using WORKING_DIR: %s\n" "${WORKING_DIR}" } ############################################################################## @@ -154,7 +154,7 @@ setup_run_cmd() { setupLauncher "${LAUNCH_WITH:-MPICH}" || exit export data_cache_path="${CKPT_DIR}/${DATA_CACHE_PATH}" && mkdir -p "${data_cache_path}" printf "\n" - log_message INFO "Using data_cache_path: ${data_cache_path}" + echo "Using data_cache_path: ${data_cache_path}" ################################################################## # WARN: to disable Llama-type architectures, toggle via: # `NO_LLAMA=1 bash train_llama_alcf.sh` @@ -231,7 +231,7 @@ setup_run_cmd() { "--num-attention-heads=${HEADS}" "--optimizer=${OPT}" "--pipeline-model-parallel-size=${PP}" - "--rotary-position-embeddings-theta=${ROPE_THETA:-50000}" + "--rotary-position-embeddings-theta=${ROPE_THETA:-5000000}" "--save=${SAVE:-${CKPT_DIR}}" "--seq-length=${SEQ}" "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" @@ -261,7 +261,7 @@ setup_run_cmd() { ["SEQ"]="${SEQ}" ["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE}" ["NUM_KV_HEAD"]="${NUM_KV_HEAD}" - ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING}" + ["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" ["DTYPE"]="${DTYPE}" ["OPT"]="${OPT}" # ["WEIGHT_DECAY"]="${WEIGHT_DECAY}" @@ -271,8 +271,7 @@ setup_run_cmd() { # ["CLIP_GRAD"]="${CLIP_GRAD:-1.0}" ) for v in "${!arch_map[@]}"; do - log_message INFO "${GREEN}${v}${RESET}: ${BLUE}${arch_map[${v}]}${RESET}" - # printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" + printf "%s: %s\n" "$(printGreen "${v}")" "$(printBlue "${arch_map[${v}]}")" done printf "======================\n" # "--adam-eps ${ADAM_EPS:-0.00001}" @@ -281,7 +280,7 @@ setup_run_cmd() { targs_cache="${cache_dir}/train_args.txt" for arg in "${train_args[@]}"; do echo "${arg}" >>"${targs_cache}"; done export TRAIN_ARGS=("$(printf '%s\n' "${train_args[@]}" | sort)") - log_message INFO "Training Arguments: ${BLUE}${TRAIN_ARGS[@]}${RESET}" # \n" "$(printBlue "${TRAIN_ARGS[@]}")" + printf "Training Arguments: %s\n" "$(printBlue "${TRAIN_ARGS[@]}")" export run_cmd=("${LAUNCHER}" "${train_args[@]}") } @@ -321,14 +320,14 @@ setupSrun() { } printJobInfo() { - log_message INFO "++++++++++++++++++++++++++++++++++++++++++++++++++" - log_message INFO "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" - log_message INFO "- Using $(which python3)" - log_message INFO "- WORLD_SIZE:${WORLD_SIZE-}" - log_message INFO "- BACKEND: ${BE:-}" - log_message INFO "- MODEL_TYPE: ${MODEL_TYPE:-}" - log_message INFO "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" - log_message INFO "++++++++++++++++++++++++++++++++++++++++++++++++++" + echo "++++++++++++++++++++++++++++++++++++++++++++++++++" + echo "- MPICH_DIR=${MPICH_DIR:-${MPI_ROOT:-}}" + echo "- Using $(which python3)" + echo "- WORLD_SIZE:${WORLD_SIZE-}" + echo "- BACKEND: ${BE:-}" + echo "- MODEL_TYPE: ${MODEL_TYPE:-}" + echo "- Using DATA_FILE_LIST: ${DATA_FILE_LIST:-}" + echo "++++++++++++++++++++++++++++++++++++++++++++++++++" } ############################################################################# @@ -455,9 +454,9 @@ get_grad_acc_steps_on_aurora() { elif [[ "$#" == 1 ]]; then hf="$1" else - log_message INFO "Usage: get_grad_acc_steps_on_aurora" - log_message INFO "Expected exactly 0 or 1 arguments, received: $#" - return 1 + echo "Usage: get_grad_acc_steps_on_aurora" + echo "Expected exactly 0 or 1 arguments, received: $#" + exit 1 fi nhosts=$(wc -l <"${hf}") if [[ "${nhosts}" -ge 256 ]]; then # n >= 256 @@ -504,114 +503,29 @@ set_ccl_vars_on_aurora() { } -# ================================================ -# AuroraGPT Model Variants -# ================================================ -get_model_arch_70B() { - HEADS=64 - NLAYERS=80 - NUM_KV_HEAD=8 - FFN_HIDDEN_SIZE=28672 - HIDDEN=8192 - SEQ=8192 - export MODEL_ARCH="AuroraGPT-70B" -} - -get_model_arch_33B() { - # 33B - export MODEL_KEY="AuroraGPT-33B" - export TP=6 - export ZERO_STAGE=2 - export GRAD_ACC_STEPS=1 - # export USE_ACTIVATION_CHECKPOINTING=1 - export MICRO_BATCH=1 - export NLAYERS=78 - export HIDDEN=8268 - export HEADS=78 - export FFN_HIDDEN_SIZE=11076 - export SEQ=4096 - export NUM_KV_HEAD=6 - export MODEL_ARCH="AuroraGPT-33B" +get_model_arch_AuroraGPT_2B() { + # AuroraGPT-2B + export HEADS=16 + export NLAYERS=12 + export HIDDEN=2048 + export NUM_KV_HEAD=4 + export FFN_HIDDEN_SIZE=11008 + export SEQ=8192 + export MODEL_ARCH="AuroraGPT-2B" } get_model_arch_7B() { # 7B # export MODEL_KEY="AuroraGPT-7B" - declare -A model_config - model_config["HEADS"]="${HEADS:-32}" - model_config["NLAYERS"]="${NLAYERS:-32}" - model_config["HIDDEN"]="${HIDDEN:-4096}" - model_config["NUM_KV_HEAD"]="${NUM_KV_HEAD:-8}" - model_config["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE:-11008}" - model_config["SEQ"]="${SEQ:-4096}" - model_config["MODEL_ARCH"]="AuroraGPT-7B" - model_config["ZERO_STAGE"]="${ZERO_STAGE:-1}" - model_config["MICRO_BATCH"]="${MICRO_BATCH:-1}" - model_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS:-${GAS:-4}}" - model_config["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" - model_config["LR_DECAY_STYLE"]="${LR_DECAY_STYLE:-cosine}" - model_config["TOKENIZER_TYPE"]="${TOKENIZER_TYPE:-HFTokenizer}" - model_config["TOKENIZER_MODEL"]="${TOKENIZER_MODEL:-"meta-llama/llama-2-7b-hf"}" - for key in "${!model_config[@]}"; do - log_message INFO "[AuroraGPT-7B] Setting ${GREEN}${key}${RESET}=${BLUE}${model_config[${key}]}${RESET}" - export "${key}"="${model_config[${key}]}" - done - # export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS - # export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS - # export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE - # export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION - # export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE - # export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 - # export model_config="AuroraGPT-7B" - # export ZERO_STAGE="${ZERO_STAGE:-1}" - # export MICRO_BATCH="${MICRO_BATCH:-1}" - # export GRAD_ACC_STEPS=:"${GRAD_ACC_STEPS:-4}" - # export USE_ACTIVATION_CHECKPOINTING="${USE_ACTIVATION_CHECKPOINTING:-0}" - # export LR_DECAY_STYLE="${LR_DECAY_STYLE:-cosine}" - # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-meta-llama/llama-2-7b-hf}" + export HEADS=${HEADS:-${NHEADS:-32}} # NUMBER OF ATEN HEADS + export NLAYERS=${NLAYERS:-${NUM_LAYERS:-32}} # NUMBER OF LAYERS + export HIDDEN=${HIDDEN:-4096} # HIDDEN SIZE + export NUM_KV_HEAD=${NUM_KV_HEAD:-8} # GROUP ATTENTION + export FFN_HIDDEN_SIZE=${FFN_HIDDEN_SIZE:-11008} # FFN HIDDEN SIZE + export SEQ=${SEQ:-4096} # SEQ_LEN: 4096 + export MODEL_ARCH="AuroraGPT-7B" } -get_model_arch_2B() { - # 2B - declare -A model_config - model_config["HEADS"]="${HEADS:-16}" - model_config["NLAYERS"]="${NLAYERS:-12}" - model_config["HIDDEN"]="${HIDDEN:-2048}" - model_config["NUM_KV_HEAD"]="${NUM_KV_HEAD:-4}" - model_config["FFN_HIDDEN_SIZE"]="${FFN_HIDDEN_SIZE:-11008}" - model_config["SEQ"]="${SEQ:-8192}" - model_config["MODEL_ARCH"]="AuroraGPT-2B" - model_config["ZERO_STAGE"]="${ZERO_STAGE:-0}" - model_config["MICRO_BATCH"]="${MICRO_BATCH:-1}" - model_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS:-2}" - model_config["USE_ACTIVATION_CHECKPOINTING"]="${USE_ACTIVATION_CHECKPOINTING:-0}" - model_config["LR_DECAY_STYLE"]="${LR_DECAY_STYLE:-constant}" - model_config["TOKENIZER_TYPE"]="${TOKENIZER_TYPE:-HFTokenizer}" - model_config["TOKENIZER_MODEL"]="${TOKENIZER_MODEL:-google/gemma-7b}" - for key in "${!model_config[@]}"; do - log_message INFO "[AuroraGPT-2B] Setting ${GREEN}${key}${RESET}=${BLUE}${model_config[${key}]}${RESET}" - export "${key}"="${model_config[${key}]}" - done - # export HEADS=16 - # export NLAYERS=12 - # export HIDDEN=2048 - # export NUM_KV_HEAD=4 - # export FFN_HIDDEN_SIZE=11008 - # export SEQ=8192 - # export model_config="AuroraGPT-2B" - # export ZERO_STAGE="${ZERO_STAGE:-0}" - # export MICRO_BATCH="${MICRO_BATCH:-1}" - # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-2}" - # export USE_ACTIVATION_CHECKPOINTING="${USE_ACTIVATION_CHECKPOINTING:-0}" - # export LR_DECAY_STYLE="${LR_DECAY_STYLE:-constant}" - # export TOKENIZER_TYPE="${TOKENIZER_TYPE:-HFTokenizer}" - # export TOKENIZER_MODEL="${TOKENIZER_MODEL:-google/gemma-7b}" -} - -# ================================================ -# Additional Model Variants -# ================================================ get_model_arch_llama3_3B() { export HEADS=24 export NLAYERS=28 @@ -672,6 +586,49 @@ get_model_arch_phi4_mini_custom_nLayers() { export MODEL_ARCH="phi4-mini-nLayers${NLAYERS}" } +# get_model_arch_70B() { +# # 70B +# export MODEL_KEY="AuroraGPT-70B" +# export TP=6 +# export ZERO_STAGE=2 +# export GRAD_ACC_STEPS=1 +# export USE_ACTIVATION_CHECKPOINTING=1 +# export MICRO_BATCH=1 +# export NLAYERS=80 +# export HIDDEN=8192 +# export HEADS=96 +# export FFN_HIDDEN_SIZE=11040 +# export SEQ=8192 +# export NUM_KV_HEAD=6 +# } +# +get_model_arch_70B() { + HEADS=64 + NLAYERS=80 + NUM_KV_HEAD=8 + FFN_HIDDEN_SIZE=28672 + HIDDEN=8192 + SEQ=8192 + export MODEL_ARCH="AuroraGPT-70B" +} + +get_model_arch_33B() { + # 33B + export MODEL_KEY="AuroraGPT-33B" + export TP=6 + export ZERO_STAGE=2 + export GRAD_ACC_STEPS=1 + # export USE_ACTIVATION_CHECKPOINTING=1 + export MICRO_BATCH=1 + export NLAYERS=78 + export HIDDEN=8268 + export HEADS=78 + export FFN_HIDDEN_SIZE=11076 + export SEQ=4096 + export NUM_KV_HEAD=6 + export MODEL_ARCH="AuroraGPT-33B" +} + ############################################################################## # setParams # @@ -697,13 +654,13 @@ setParams() { export BE="${CCL}" # COMMUNICATION BACKEND = CCL export DTYPE=${DTYPE:-bf16} # DTYPE: bf16 # export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT_ACC_STEPS - # gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") - # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas:-1}}" + gas=$(get_grad_acc_steps_on_aurora "${PBS_NODEFILE:-${HOSTFILE:-${hostfile}}}") + export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-${gas}}" # export GRAD_ACC_STEPS="${GRAD_ACC_STEPS:-$(get_grad_acc_steps_on_aurora "$@)}" - # echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" - # MICRO_BATCH=${MICRO_BATCH:-1} + echo "[setParams] Using GRAD_ACC_STEPS: ${GRAD_ACC_STEPS}" + MICRO_BATCH=${MICRO_BATCH:-1} if [[ -n "${NO_FLASH_ATTN-}" ]]; then - log_message WARN "Not using flash-attn!!" + echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-builder" fi @@ -720,11 +677,11 @@ setParams() { # MICRO_BATCH=${MICRO_BATCH:-2} # MICRO_BATCH = 8 export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN:-}" ]]; then - log_message WARN "Not using flash-attn!!" + echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi - log_message INFO "Setting up AWS NCCL OFI Plugin on Polaris..." + echo "Setting up AWS NCCL OFI Plugin on Polaris..." source "${WORKING_DIR}/ALCF/aws_ofi_nccl_plugin.sh" || exit # ---- [Sophia] ---------------------- elif [[ "${mn}" == sophia* ]]; then @@ -736,7 +693,7 @@ setParams() { export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-8} # GRADIENT_ACC_STEPS export MICRO_BATCH="${MICRO_BATCH:-$(get_batch_size_on_polaris)}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then - log_message WARN "Not using flash-attn!!" + echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi @@ -750,23 +707,17 @@ setParams() { export DTYPE="${DTYPE:-bf16}" MICRO_BATCH="${MICRO_BATCH:-1}" if [[ -n "${NO_FLASH_ATTN-}" ]]; then - log_message WARN "Not using flash-attn!!" + echo "Not using flash-attn!!" else FLASH_ARG="--use-flash-attn-v2" fi fi - ma="${MODEL_ARCH:-"AuroraGPT-2B"}" - + ma="${MODEL_ARCH:-7B}" case "${ma}" in - "70B" | "AuroraGPT-70B" | "aurora-gpt-70b") + # "70B" | "llama-3.1-70B" | "llama-3.1-70b" | "llama-3.2-70B" | "llama-3.2-70b") + "70B") get_model_arch_70B ;; - "7B" | "AuroraGPT-7B" | "aurora-gpt-7b") - get_model_arch_7B - ;; - "2B" | "AuroraGPT-2B" | "aurora-gpt-2b") - get_model_arch_2B - ;; "33B" | "llama-3.2-33B" | "llama-3.2-33b") get_model_arch_33B ;; @@ -779,8 +730,14 @@ setParams() { "llama3-3B" | "llama-3B") get_model_arch_llama3_3B_customNlayers ;; + "2B" | "AuroraGPT-2B" | "AuroraGPT_2B" | "Aurora-GPT-2B" | "AuroraGPT2B" | "Aurora_GPT_2B" | "aurora-gpt-2b" | "aurora_gpt_2b") + get_model_arch_AuroraGPT_2B + ;; + "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") + get_model_arch_7B + ;; *) - get_model_arch_3B + get_model_arch_7B ;; esac export TP="${TP}" @@ -817,45 +774,29 @@ setParams() { lr_flags+=("--lr-decay-iters ${LR_DECAY_ITERS:-}") fi # +---[Run Settings]------------------------------------------------------+ - # export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE - # export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE - export GRAD_ACC_STEPS="${GRAD_ACC_STEPS}" # GRADIENT ACCUMULATION STEPS + export ZERO_STAGE=${ZERO_STAGE:-1} # ZERO OFFLOADING STAGE + export MICRO_BATCH=${MICRO_BATCH:-1} # MICRO BATCH SIZE + export GRAD_ACC_STEPS=${GRAD_ACC_STEPS:-1} # GRADIENT ACCUMULATION STEPS export TIMING_LOG_LEVEL="${TIMING_LOG_LEVEL:-1}" # TIMING VERBOSITY IN LOGS export ACT_CKPT_NUM_LAYERS="${ACT_CKPT_NUM_LAYERS:-1}" # NUM LAYERS TO CHECKPOINT ACTIVATIONS - export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} # USE ACTIVATION CHECKPOINTING ? + export USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING:-} # USE ACTIVATION CHECKPOINTING ? export GLOBAL_BATCH_MAX=$((WORLD_SIZE * MICRO_BATCH * GRAD_ACC_STEPS / TP / PP / SP)) # MAX GLOBAL BATCH SIZE export DP=$((WORLD_SIZE / TP / PP / SP)) # DATA PARALLELISM export GLOBAL_BATCH="${GLOBAL_BATCH:-${GLOBAL_BATCH_MAX}}" # WILL USE MAX IF NOT SET IN ENVIRONMENT - declare -A parallelism_config - parallelism_config["WORLD_SIZE"]="${WORLD_SIZE}" - parallelism_config["DP"]="${DP}" - parallelism_config["TP"]="${TP}" - parallelism_config["PP"]="${PP}" - parallelism_config["SP"]="${SP}" - parallelism_config["MICRO_BATCH"]="${MICRO_BATCH}" - parallelism_config["GRAD_ACC_STEPS"]="${GRAD_ACC_STEPS}" - parallelism_config["GLOBAL_BATCH"]="${GLOBAL_BATCH}" - parallelism_config["GLOBAL_BATCH_MAX"]="${GLOBAL_BATCH_MAX}" - for key in "${!parallelism_config[@]}"; do - log_message INFO "[Parallelism] Setting ${GREEN}${key}${RESET}=${BLUE}${parallelism_config[${key}]}${RESET}" - export "${key}"="${parallelism_config[${key}]}" - done - # echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" + echo "WORLD_SIZE: ${WORLD_SIZE} MBS: ${MICRO_BATCH} GAS: ${GRAD_ACC_STEPS} DP: ${DP} GBS: ${GLOBAL_BATCH} GBS_MAX: ${GLOBAL_BATCH_MAX} TP: ${TP} PP: ${PP} SP: ${SP}" if [[ -n "${TRAIN_TOKENS:-}" ]]; then export TRAIN_TOKENS="${TRAIN_TOKENS}" export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" elif [[ -z "${TRAIN_ITERS:-${TRAIN_ITER:-}}" ]]; then export TRAIN_TOKENS=${TRAIN_TOKENS:-4673780159710} export TRAIN_ITERS=$((TRAIN_TOKENS / SEQ / GLOBAL_BATCH)) - # log_message INFO "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - # log_message INFO "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" + printf "TRAIN_TOKENS=%s (=%sB tokens)\n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" + printf "TRAIN_ITERS=%s\n" "${TRAIN_ITERS}" else export TRAIN_ITERS="${TRAIN_ITERS:-${TRAIN_ITER:-}}" - TRAIN_TOKENS=$((TRAIN_ITERS * SEQ * GLOBAL_BATCH)) - export TRAIN_TOKENS fi - log_message INFO "TRAIN_TOKENS=${TRAIN_TOKENS} (=$((TRAIN_TOKENS / 10 ** 9))B tokens)" # \n" "${TRAIN_TOKENS}" "$((TRAIN_TOKENS / 10 ** 9))" - log_message INFO "TRAIN_ITERS=${TRAIN_ITERS}" # \n" "${TRAIN_ITERS}" export MODEL_TYPE="${MODEL_ARCH:-AuroraGPT}-gb${GLOBAL_BATCH}-seq${SEQ}-pp${PP}-tp${TP}-${NLAYERS}layers-${HEADS}heads-${HIDDEN}hidden" # STRING FOR IDENTIFYING MODEL # NOTE: [2024-07-10] ##################################################### # - [sam]: For whatever reason, it seems that using @@ -898,7 +839,7 @@ set_args() { ds_args+=("--zero-stage=$ZERO_STAGE") # if [[ -n "${USE_ACTIVATION_CHECKPOINTING:-}" ]]; then if [[ "${USE_ACTIVATION_CHECKPOINTING:-}" == 1 || "${USE_ACTIVATION_CHECKPOINTING:-}" == "true" ]]; then - log_message WARN "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" + echo "!! Caught USE_ACTIVATION_CHECKPOINTING=${USE_ACTIVATION_CHECKPOINTING} !!" ds_args+=("--deepspeed-activation-checkpointing") ds_args+=( "--checkpoint-activations" @@ -943,7 +884,7 @@ ezpz_test() { # forwarded to ALL ranks with DeepSpeed ############################################################################ saveDSenv() { - log_message INFO "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" + echo "Saving {PATH, LD_LIBRARY_PATH, htt{p,ps}_proxy, CFLAGS, PYTHONUSERBASE} to .deepspeed_env" { echo "PATH=${PATH:-}" echo "LD_LIBRARY_PATH=${LD_LIBRARY_PATH:-}" @@ -1064,15 +1005,14 @@ make_data() { ############################################################################## install_dependencies() { depsfile="${WORKING_DIR}/ALCF/requirements/requirements.txt" - log_message INFO "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." - python3 -m pip install -r "${depsfile}" > /dev/null || { # --require-virtualenv - log_message INFO "[install_dependencies] Failed to install dependencies from ${depsfile} on $(ezpz_get_machine_name) in $(which python3)\n" - } + echo "[install_dependencies] Ensuring all dependencies from ${depsfile} installed..." + python3 -m pip install -r "${depsfile}" # --require-virtualenv if [[ ! -x "$(command -v deepspeed)" ]]; then - log_message INFO "[install_dependencies] No 'deepspeed' command found on $(ezpz_get_machine_name) in $(which python3)\n" - log_message INFO "[install_dependencies] Attempting to install deepspeed via pip...\n" - python3 -m pip install deepspeed > /dev/null || { # --require-virtualenv || { - log_message INFO "[install_dependencies] Failed to install deepspeed via pip on $(ezpz_get_machine_name)\n" + printf "[install_dependencies] No 'deepspeed' command found on %s in %s\n" "$$(ezpz_get_machine_name)" "$(which python3)" + printf "[install_dependencies] Attempting to install deepspeed via pip...\n" + python3 -m pip install deepspeed || { # --require-virtualenv || { + printf "[install_dependencies] Failed to install deepspeed via pip on %s\n" "$(ezpz_get_machine_name)" + # printf "[install_dependencies] !! No deepsepeed in %s\n" "$(which python3)" return 1 } # mn=$(ezpz_get_machine_name) @@ -1159,7 +1099,7 @@ makeHostfiles() { setup_tokenizer_and_data() { if [[ "$#" == 1 ]]; then tok="$1" - dfl="${DATA_FILE_LIST:-ALCF/data-lists/$(ezpz_get_machine_name)/olmo-mix-11.txt}" + dfl="${DATA_FILE_LIST:-}" elif [[ "$#" == 2 ]]; then tok="$1" dfl="$2" @@ -1207,8 +1147,8 @@ setup_tokenizer_and_data() { fi export DATA_FLAGS="${_data_flags[*]:-}" export TOKENIZER_FLAGS="${_tokenizer_flags[*]}" - log_message INFO "[setData] DATA_FLAGS: ${GREEN}${DATA_FLAGS}${RESET}" # \n" "$(printGreen "${DATA_FLAGS}")" - log_message INFO "[setData] TOKENIZER_FLAGS: ${MAGENTA}${TOKENIZER_FLAGS}${RESET}" # \n" "$(printMagenta "${TOKENIZER_FLAGS}")" + printf "[setData] DATA_FLAGS: %s\n" "$(printGreen "${DATA_FLAGS}")" + printf "[setData] TOKENIZER_FLAGS: %s\n" "$(printMagenta "${TOKENIZER_FLAGS}")" } ############################################### @@ -1225,7 +1165,7 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] # set `dfl` to `dfl_fallback` if not passed as an argument, # use this data file list to call `setData` dfl="${1:-${dfl_fallback}}" - log_message INFO "Calling: setData() with ${dfl}" # \n" "${dfl}" + printf "Calling: setData() with %s\n" "${dfl}" ndocs=$(wc -l <"${dfl}") ws=$(sumWeights "${dfl}") dfl_stem=$(echo "${dfl}" | tr "\/" "\t" | awk '{print $NF}' | sed "s/\.txt//g") @@ -1236,15 +1176,15 @@ setData() { # ------------------------[dfl: abbrv. for DATA_FILE_LIST] export DFL_STEM="${dfl_stem}" export DATA_CACHE_PATH="${dcp}" # export DATA_FLAGS="${DATA_FLAGS} --data-file-list ${DATA_FILE_LIST}" # --data-cache-path ${DATA_CACHE_PATH}" - log_message INFO "--------------------" - log_message INFO "Updated environment:" - log_message INFO "DATA_FILE_LIST: ${DATA_FILE_LIST}" # \n" "${DATA_FILE_LIST}" - log_message INFO "NUM_DOCS: ${NUM_DOCS}" # %s\n " "${NUM_DOCS}" - log_message INFO "WEIGHT_SUM: ${WEIGHT_SUM}" # %s\n" "${WEIGHT_SUM}" - log_message INFO "DFL_STEM: ${DFL_STEM}" # %s\n" "${DFL_STEM}" - log_message INFO "DATA_CACHE_PATH: ${DATA_CACHE_PATH}" # %s\n" "${DATA_CACHE_PATH}" - log_message INFO "DATA_FLAGS: ${DATA_FLAGS}" # \n" "${DATA_FLAGS:-}" - log_message INFO "--------------------" + echo "--------------------" + echo "Updated environment:" + printf "DATA_FILE_LIST: %s\n" "${DATA_FILE_LIST}" + printf "NUM_DOCS: %s\n " "${NUM_DOCS}" + printf "WEIGHT_SUM: %s\n" "${WEIGHT_SUM}" + printf "DFL_STEM: %s\n" "${DFL_STEM}" + printf "DATA_CACHE_PATH: %s\n" "${DATA_CACHE_PATH}" + printf "DATA_FLAGS: %s\n" "${DATA_FLAGS:-}" + echo "--------------------" } generateDSconfig_new() { From 7d7996543db1ac3a6f308bc59b8170cb1e86a00a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 23 Sep 2025 13:42:11 -0500 Subject: [PATCH 501/590] chore: Format `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index 95e8d15f9a8..a2d974418f7 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -48,6 +48,7 @@ from torch import nn import torch.nn.functional as F + dt_imports = time.time() - python_start_time t0_setup = time.time() @@ -342,9 +343,7 @@ def loss_func(loss_mask, moe_loss, mos_loss, output_tensor): "kd loss": mos_loss, } log.info( - f">>> total loss: {loss}, " - f"lm loss {averaged_loss[0]}, " - f"kd loss {mos_loss}" + f">>> total loss: {loss}, lm loss {averaged_loss[0]}, kd loss {mos_loss}" ) else: if max(args.num_experts) <= 1: @@ -486,8 +485,7 @@ def train_valid_test_datasets_provider(train_val_test_num_samples): except Exception as exc: log.exception(exc) raise Exception( - "Please provide the file list as " - "'weight, filename, corpus'" + "Please provide the file list as 'weight, filename, corpus'" ) if fname.find(".bin") != -1: fname = fname.split(".bin")[0] @@ -547,8 +545,7 @@ def git_ds_info(): git_hash = "unknown" git_branch = "unknown" print( - f"**** Git info for Megatron: " - f"git_hash={git_hash} git_branch={git_branch} ****" + f"**** Git info for Megatron: git_hash={git_hash} git_branch={git_branch} ****" ) From 9fed3b58d8a07c08af1051f2b947f266e063901a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Wed, 24 Sep 2025 07:11:26 -0500 Subject: [PATCH 502/590] chore: Update `train_aGPT_2B_large_batch.sh` --- train_aGPT_2B_large_batch.sh | 78 ++++++++++++++++--------------- train_aGPT_2B_large_batch_pt28.sh | 66 -------------------------- 2 files changed, 40 insertions(+), 104 deletions(-) mode change 100755 => 100644 train_aGPT_2B_large_batch.sh delete mode 100755 train_aGPT_2B_large_batch_pt28.sh diff --git a/train_aGPT_2B_large_batch.sh b/train_aGPT_2B_large_batch.sh old mode 100755 new mode 100644 index b2a8189345b..bc78bfd9f2c --- a/train_aGPT_2B_large_batch.sh +++ b/train_aGPT_2B_large_batch.sh @@ -5,51 +5,53 @@ #PBS -l walltime=06:00:00,filesystems=flare:home #PBS -l select=256 -# Get the directory of this script -# whereami="$(dirname "$(realpath "$0")")" -# cd "${whereami}" || { -# echo "Failed to change directory to ${whereami}" -# exit 1 -# } -setup_env() { - cd /flare/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/large-batch-training/tok50M-n512/Megatron-DeepSpeed || { - echo "Failed to change directory to /flare/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/large-batch-training/tok50M-n512/Megatron-DeepSpeed" - exit 1 - } - PBS_O_WORKDIR="$(pwd)" - export PBS_O_WORKDIR - export http_proxy="http://proxy.alcf.anl.gov:3128" - export https_proxy="http://proxy.alcf.anl.gov:3128" - export ftp_proxy="http://proxy.alcf.anl.gov:3128" - export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" - export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" +# Optimizer Plans +# +# 1. Ongoing: +# - Lamb +# - 256 Nodes (50M Tok / batch, LR=2e-4) +# - 512 Nodes (100M Tok / batch, LR=2e-4) +# 1. Up-next: +# 256 Nodes @ 50M Tok / batch +# - Sophia: +# - LR=2.17e-5 +# - Muon +# - LR=1.82e-5 +# - Muon-clip +# - LR=2.28e-5 + - # shellcheck disable=SC1090 - source <(curl -L https://bit.ly/ezpz-utils) - ezpz_setup_env - log_message INFO "Using: $(which python3)" +setup_env() { + cd "${PBS_O_WORKDIR}" || { + echo "Failed to change directory to ${PBS_O_WORKDIR}" + exit 1 + } + # shellcheck disable=SC1090 + source <(curl -L https://bit.ly/ezpz-utils) + ezpz_setup_env + log_message INFO "Using: $(which python3)" } train_model() { - MODEL_ARCH=smollm3-3B \ - OPT=ipex.fusedlamb \ - NLAYERS=12 \ - GRAD_ACC_STEPS=2 \ - MICRO_BATCH=1 \ - USE_ACTIVATION_CHECKPOINTING=0 \ - ZERO_STAGE=0 \ - LR_DECAY_STYLE=constant \ - TOKENIZER_TYPE=HFTokenizer \ - TOKENIZER_MODEL=google/gemma-7b \ - DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ - bash "${PBS_O_WORKDIR}/train_alcf.sh" \ - "$@" -} + MODEL_ARCH=AuroraGPT-2B \ + OPT=muonclip \ + LR=2.28e-5 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" \ + "$@" + } main() { - setup_env - train_model "$@" + setup_env + train_model "$@" } main "$@" diff --git a/train_aGPT_2B_large_batch_pt28.sh b/train_aGPT_2B_large_batch_pt28.sh deleted file mode 100755 index 5a31b628c31..00000000000 --- a/train_aGPT_2B_large_batch_pt28.sh +++ /dev/null @@ -1,66 +0,0 @@ -#!/bin/bash --login -#PBS -q prod -#PBS -A AuroraGPT -#PBS -j oe -#PBS -l walltime=06:00:00,filesystems=flare:home -#PBS -l select=256 - -# Get the directory of this script -whereami="$(dirname "$(realpath "$0")")" -cd "${whereami}" || { - echo "Failed to change directory to ${whereami}" - exit 1 -} - -setup_env() { - PBS_O_WORKDIR="$(pwd)" - export PBS_O_WORKDIR - export http_proxy="http://proxy.alcf.anl.gov:3128" - export https_proxy="http://proxy.alcf.anl.gov:3128" - export ftp_proxy="http://proxy.alcf.anl.gov:3128" - export HTTP_PROXY="http://proxy.alcf.anl.gov:3128" - export HTTPS_PROXY="http://proxy.alcf.anl.gov:3128" - - # shellcheck disable=SC1090 - source /home/foremans/utils.sh - - conda_env=/flare/datascience/foremans/micromamba/envs/2025-07-pt28 - conda_name="$(basename "$conda_env")" - dst_env="/tmp/${conda_name}" - - if [[ ! -d "${dst_env}" ]]; then - echo "No cached conda environment found, creating a new one" - ezpz_setup_env - ezpz-yeet-env --src "${conda_env}.tar.gz" - deactivate - conda deactivate - fi - - ezpz_load_new_pt_modules_aurora - conda activate "${dst_env}" - ezpz_setup_env - log_message INFO "Using: $(which python3)" -} - -train_model() { - MODEL_ARCH=smollm3-3B \ - OPT=ipex.fusedlamb \ - NLAYERS=12 \ - GRAD_ACC_STEPS=2 \ - MICRO_BATCH=1 \ - USE_ACTIVATION_CHECKPOINTING=0 \ - ZERO_STAGE=0 \ - LR_DECAY_STYLE=constant \ - TOKENIZER_TYPE=HFTokenizer \ - TOKENIZER_MODEL=google/gemma-7b \ - DATA_FILE_LIST=ALCF/data-lists/aurora/olmo-mix-1124.txt \ - bash "${PBS_O_WORKDIR}/train_alcf.sh" \ - "$@" -} - -main() { - setup_env - train_model "$@" -} - -main "$@" From 527fe81a5d35d6aa99bbf26bb5f0833d515f19a5 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Fri, 26 Sep 2025 11:29:18 -0500 Subject: [PATCH 503/590] Update cpt.md --- ALCF/notes/cpt.md | 93 +++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 89 insertions(+), 4 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index d5175a11417..7ddb8fda5cc 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,16 +1,96 @@ # CPT ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: -1. First, we need to train from an universal checkpoint. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. +1. First, we need to train from an **universal checkpoint**. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. 2. Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. Here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` +For convenience, here is a copy of the ***mix_datasets.py*** script +```bash +#!/usr/bin/env python3 +import argparse +import sys + +def parse_args(): + p = argparse.ArgumentParser( + description="Mix multiple file-lists, normalize internal weights, and apply global file weights." + ) + p.add_argument( + '--inputs', + nargs='+', + required=True, + help="Pairs of file_list and global_weight, e.g.: --inputs 0.3 f1.txt 0.7 f2.tx" + ) + return p.parse_args() + + +def main(): + args = parse_args() + inp = args.inputs + if len(inp) % 2 != 0: + sys.exit("Error: --inputs must be an even number of arguments (file weight pairs).") + + # Group into (file_path, global_weight) + pairs = [] + for i in range(0, len(inp), 2): + file_path = inp[i+1] + try: + gw = float(inp[i]) + except ValueError: + sys.exit(f"Error: global weight must be a number, got '{inp[i]}'") + if gw <= 0: + sys.exit(f"Error: global weight must be positive, got {gw}") + pairs.append((gw, file_path)) + + # Compute sum of all global weights (if normalization across files is desired) + sum_global = sum(gw for gw, _ in pairs) + + for gw, file_path in pairs: + # Normalized file-level fraction (optional across all files) + file_fraction = gw / sum_global + + # Read entries and sum file-local weights + entries = [] + file_sum = 0.0 + try: + with open(file_path, 'r') as f: + for line in f: + line = line.strip() + if not line or line.startswith('#'): + continue + parts = line.split() + if len(parts) < 3: + sys.exit(f"Error: each line must have prefix weight corpus, got: '{line}'") + prefix = parts[1] + try: + w = float(parts[0]) + except ValueError: + sys.exit(f"Error: weight must be numeric, got '{parts[1]}' in file {file_path}") + corpus = parts[2] + entries.append((prefix, w, corpus)) + file_sum += w + except FileNotFoundError: + sys.exit(f"Error: cannot open file '{file_path}'") + + if file_sum <= 0: + sys.exit(f"Error: sum of weights in file '{file_path}' is non-positive: {file_sum}") + + # Print header only once + # Compute and print normalized weights + for prefix, w, corpus in entries: + new_w = (w / file_sum) * file_fraction + print(f"{new_w:.6f} {prefix} {corpus}") + +if __name__ == '__main__': + main() +``` + 3. Then, we can run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune ``` -Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. by just setting a different value for LR For example, we tested rewarming to LR/2 and 2LR as well. +Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. by just setting a different value for LR For example, we tested rewarming to LR/2 i.e **LR=0.0001** and 2LR as well. Here the following options options/flags should be: ```bash DATA_FILE_LIST=path/to/your/tokenized/data @@ -22,11 +102,16 @@ SAVE=path/to/where/you/want/to/save/checkpoints For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. To do CPT here, -1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max BEFORE COOLING DOWN** +1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** 2. Mix the datasets as above 3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm 4. Run ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown ``` -5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint (i.e after cooling it down) and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. +5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. + +## To do list +Follow and implement the [recipe](https://arxiv.org/pdf/2407.07263v1) where the new dataset is incrementally introduced. This might be advantageous for example when the new dataset is QAs as opposed to pure text. Here: +1. If base model LR was decayed to 0, one might need to rewarm it before following the recipe +2. Constant/infinite LR schedule was used, one might experiment with the recipe as is. From 7d90dfe088200e8ff91077aee248c25b7040a9c7 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 30 Sep 2025 11:21:55 -0500 Subject: [PATCH 504/590] Update cpt.md --- ALCF/notes/cpt.md | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 7ddb8fda5cc..e3aa5c8eb78 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -98,18 +98,30 @@ LOAD=path/to/your/universal/checkpoint SAVE=path/to/where/you/want/to/save/checkpoints --universal-checkpoint to load a universal checkpoint ``` -## New agpt runs (phase 1 -> phase 2) +## New agpt runs (phase 1 -> phase 2: weak distribution shift) For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. To do CPT here, 1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** -2. Mix the datasets as above +2. Mix the datasets as above, I would try different mixing weights here to experiment with 3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm 4. Run ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown ``` -5. If loss curve not recovering after a while or loss is diverging, one can take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. +5. If loss curve not recovering after a while or loss is diverging, one can: + - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. + - If still no luck, I'd: + a. take an earlier not converged checkpoint + b. Continue training with the base dataset with a cosine scheduler decaying to **LR_max/100** or cooldown to **LR_max/100**. (I would experiment with both if resources allow) + c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + +My guess is that since the distribution shift is not too strong between phase I and phase II data, you will not need to experiment with step 5. + +## New agpt runs (phase 2 -> phase 3 or phase 3 -> phase 4: strong distribution shift) +I'd try steps 1-4 above and put more weight on phase 2 (then phase 3) data when mixing in the new datasets. If no luck, step 5 should work here. + + ## To do list Follow and implement the [recipe](https://arxiv.org/pdf/2407.07263v1) where the new dataset is incrementally introduced. This might be advantageous for example when the new dataset is QAs as opposed to pure text. Here: From a513cafa869702aab5da8d65bb3f78339ff60154 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 30 Sep 2025 13:14:01 -0500 Subject: [PATCH 505/590] Update cpt.md --- ALCF/notes/cpt.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index e3aa5c8eb78..2c66586ae89 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,7 +1,7 @@ # CPT ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: -1. First, we need to train from an **universal checkpoint**. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. +1. First, if running on resources different than in base pretraining i.e smaller num of gpus, we need to train from an **universal checkpoint**. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. 2. Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. Here we are mixing the lucid papers with weight 0.9 and dolma with weight 0.1 (you can play with the weights if needed): ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt @@ -96,13 +96,13 @@ Here the following options options/flags should be: DATA_FILE_LIST=path/to/your/tokenized/data LOAD=path/to/your/universal/checkpoint SAVE=path/to/where/you/want/to/save/checkpoints ---universal-checkpoint to load a universal checkpoint +--universal-checkpoint to load a universal checkpoint (not needed if checkpoint not universal) ``` ## New agpt runs (phase 1 -> phase 2: weak distribution shift) For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. To do CPT here, -1. Convert checkpoint to an universal checkpoint, **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** +1. Convert checkpoint to an universal checkpoint (if running for example on smaller num of gpus), **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** 2. Mix the datasets as above, I would try different mixing weights here to experiment with 3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm 4. Run From 07a7bbdf3a0f1cd5bec92da9764155b2a867ac89 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Sun, 12 Oct 2025 13:43:18 -0500 Subject: [PATCH 506/590] fix: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5d70ea98a45..5020bcd19da 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -361,7 +361,7 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - LAUNCHER="ezpz-launch ${EXEC}" + LAUNCHER="ezpz-launch $(which python3) ${EXEC}" # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job # echo "ezpz_launch: $(which ezpz_launch)" # export -f ezpz_launch From 889ff8df24419c3ee79af91c5f97bed02a89c76a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:51:26 -0600 Subject: [PATCH 507/590] docs: Add `ALCF/notes/cooldown.md` --- ALCF/notes/cooldown.md | 230 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 230 insertions(+) create mode 100644 ALCF/notes/cooldown.md diff --git a/ALCF/notes/cooldown.md b/ALCF/notes/cooldown.md new file mode 100644 index 00000000000..7fbb800cd2b --- /dev/null +++ b/ALCF/notes/cooldown.md @@ -0,0 +1,230 @@ +# 🧊 Cooling Down AuroraGPT-2B Checkpoints + +- Cooled down over last $10\%$: + - [volcanic-blaze-4312](https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu/overview?nw=nwuserforemans) + ![](./assets/cooldown.png) + +## ♻️ Convert to Universal + +```bash +TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1 python3 ALCF/ds_to_universal.py \ + --input_folder test_rollback/global_step136000 \ + --output_folder test_rollback/global_step136000_universal +``` + +## Cooling Down + +- 256 Nodes: + + ```bash + LR_DECAY_STYLE=constant \ + OPT=ipex.fusedlamb \ + OVERRIDE_CKPT_OPT_PARAM=1 \ + TRAIN_ITERS=137650 \ + GRAD_ACC_STEPS=2 \ + LOAD=test_rollback \ + DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + bash train_alcf.sh \ + --override-opt_param-scheduler \ + --min-lr=2e-5 \ + --lr_constant_plus_cooldown \ + --lr_constant_plus_cooldown_frac=0.9 + ``` + +Example: + +```bash +#[🐍 aurora_frameworks-2025.2.0](👻 Megatron-DeepSpeed-aurora_frameworks-2025.2.0) +#[/f/A/A/E/A/c/r/Megatron-DeepSpeed][🌱 main][✅] [⏱️ 26m13s] +#[11/10/25 @ 10:19:03][x4417c6s4b0n0] +; LR_DECAY_STYLE=constant \ + LR=0.0002 \ + OPT=ipex.fusedlamb \ + OVERRIDE_CKPT_OPT_PARAM=1 \ + TRAIN_ITERS=137650 \ + GRAD_ACC_STEPS=2 \ + LOAD=test_rollback \ + DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + bash train_alcf.sh \ + --lr_constant_plus_cooldown \ + --lr_constant_plus_cooldown_frac=0.9 \ + --min-lr=2e-5 \ + --override-opt_param-scheduler + +[2025-11-10-095114][I][] Detected PBS scheduler environment. +[2025-11-10-095114][I][] running [ezpz_setup_env]... +[2025-11-10-095114][I][] [PYTHON] +[2025-11-10-095114][I][] - Found both conda_prefix and virtual_env in environment. +[2025-11-10-095114][I][] - Using conda from: /opt/aurora/25.190.0/frameworks/aurora_frameworks-2025.2.0 +[2025-11-10-095114][I][] - Using venv from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlamb- +256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0 +[2025-11-10-095114][I][] - Using python from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlam +b-256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0/bin/python3 +[2025-11-10-095114][I][] [JOB] +[2025-11-10-095114][I][] - Setting up env for foremans +[2025-11-10-095114][I][] - Detected pbs scheduler +[2025-11-10-095114][I][] - Machine: aurora +[2025-11-10-095114][I][] - Hostname: x4417c6s4b0n0 +[2025-11-10-095116][I][] - PBS_JOBID=8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + to calculate: + - num_hosts: 256 + - num_cores_per_host: 208 + - num_cpus_per_host: 104 + - num_gpus_per_host: 12 + - depth: 8 + - num_gpus: 3072 +[2025-11-10-095116][I][] [HOSTS] - ezpz_print_hosts +[2025-11-10-095116][I][] - Detected PBS Scheduler +[2025-11-10-095116][I][] [HOSTS] +[2025-11-10-095116][I][] - HOSTFILE=/var/spool/pbs/aux/8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov +[2025-11-10-095116][I][] - NHOSTS=256 +[2025-11-10-095116][I][] - HOSTS: +[2025-11-10-095116][I][] - [host:0] - x4417c6s4b0n0.hsn.cm.aurora.alcf.anl.gov +# [...clipped...] +[2025-11-10 09:56:45,204] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 +[2025-11-10 09:56:45,302] [INFO] [utils.py:781:see_memory_usage] Before Building Model +[2025-11-10 09:56:45,303] [INFO] [utils.py:782:see_memory_usage] MA 0.0 GB Max_MA 0.0 GB CA 0.0 GB Max_CA 0 GB +[2025-11-10 09:56:45,303] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.83 GB, percent = 6.5% +[2025-11-10 09:56:45,304] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 +[2025-11-10 09:56:45,320815][I][Megatron-DeepSpeed/pretrain_gpt_alcf:151:model_provider] -------------------------------------------------------------------------------- +[2025-11-10 09:56:45,321819][I][Megatron-DeepSpeed/pretrain_gpt_alcf:152:model_provider] Number of parameters in model: 1986054144 +[2025-11-10 09:56:45,322546][I][Megatron-DeepSpeed/pretrain_gpt_alcf:153:model_provider] -------------------------------------------------------------------------------- +[2025-11-10 09:56:45,484] [INFO] [utils.py:781:see_memory_usage] After Building Model +[2025-11-10 09:56:45,485] [INFO] [utils.py:782:see_memory_usage] MA 3.71 GB Max_MA 3.71 GB CA 3.72 GB Max_CA 4 GB +[2025-11-10 09:56:45,485] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.84 GB, percent = 6.5% +[2025-11-10 09:56:45,485921][I][Megatron-DeepSpeed/pretrain_gpt_alcf:161:model_provider] Patching tensorboard from checkpoints/AuroraGPT-2B-ws3072-ds-stage1-nl12-hs2048-mb1-seq8192-gb6144-sp1-pp1-tp1-bf16-optipex.fusedlamb-lr0.0002-lwf0.05_ntok0B_flash/tensorboard +wandb: WARNING The get_url method is deprecated and will be removed in a future release. Please use `run.url` instead. +[2025-11-10 09:56:51,975268][I][Megatron-DeepSpeed/pretrain_gpt_alcf:168:model_provider] Updating WandB run.config: [sandy-darkness-4309](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2cmpsosr) + > number of parameters on (tensor, pipeline) model parallel rank (0, 0)=1986054144 +[2025-11-10 09:56:51,979967][I][megatron/optimizer_param_scheduler:89:__init__] > learning rate decay style: constant +[2025-11-10 09:56:51,980855][I][megatron/training:725:setup_model_and_optimizer] DeepSpeed is enabled. +[2025-11-10 09:56:51,981610][I][megatron/training:780:setup_model_and_optimizer] Did NOT catch: ('args.data_efficiency_curriculum_learning' and 'build_train_valid_test_datasets_provider is not None') +[2025-11-10 09:56:51,982375][I][megatron/training:789:setup_model_and_optimizer] Calling 'deepspeed.initialize'... +[2025-11-10 09:56:51,983047][I][megatron/training:790:setup_model_and_optimizer] Wrapped with: profiler= +[2025-11-10 09:56:51,983] [INFO] [logging.py:107:log_dist] [Rank 0] DeepSpeed info: version=0.17.5, git-hash=unknown, git-branch=unknown + +# [...clipped...] + +[2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 715.45 | bwd_microstep: 2402.91 | bwd_inner_microstep: 1223.71 | bwd_allreduce_microstep: 1178.40 | step_microstep: 160.78 +[2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 715.40 | bwd: 2402.86 | bwd_inner: 1223.70 | bwd_allreduce: 1178.40 | step: 160.79 +[2025-11-10 09:43:36,161992][I][megatron/training_log:402:training_log] iteration= 136004/ 136100 | consumed_samples= 835608576 | consumed_tokens=6845305454592 | elapsed_time_per_iteration_ms=3415.3 | learning_rate=2.0127e-05 | global_batch_size= 6144 | lm loss=12.316531 | loss_scale=1.0 | grad_norm=1020.352 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1798.971 | tokens_per_gpu_per_second_tgs=4797.255 | TFLOPs=53.66 | +(min, max) time across ranks (ms): + forward-backward ...............................: (3198.27, 3199.53) + optimizer ......................................: (158.53, 166.36) +[2025-11-10 09:43:39,572] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 167.02 | optimizer_gradients: 0.46 | optimizer_step: 1.13 +[2025-11-10 09:43:39,573] [INFO] [logging.py:107:log_dist] [Rank 0] step=136005, skipped=0, lr=[2.012568679065123e-05, 2.012568679065123e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:43:39,573] [INFO] [timer.py:264:stop] epoch=0/micro_step=5/global_step=5, RunningAvgSamplesPerSec=2443.9516881034883, CurrSamplesPerSec=2553.443452275532, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.96 | bwd_microstep: 2398.60 | bwd_inner_microstep: 1222.63 | bwd_allreduce_microstep: 1175.22 | step_microstep: 173.52 +[2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.90 | bwd: 2398.57 | bwd_inner: 1222.62 | bwd_allreduce: 1175.27 | step: 173.53 +[2025-11-10 09:43:39,582716][I][megatron/training_log:402:training_log] iteration= 136005/ 136100 | consumed_samples= 835614720 | consumed_tokens=6845355786240 | elapsed_time_per_iteration_ms=3420.2 | learning_rate=2.01257e-05 | global_batch_size= 6144 | lm loss=11.967899 | loss_scale=1.0 | grad_norm=735.534 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1796.410 | tokens_per_gpu_per_second_tgs=4790.426 | TFLOPs=53.59 | +(min, max) time across ranks (ms): + forward-backward ...............................: (3194.05, 3195.63) + optimizer ......................................: (170.97, 175.03) +[2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 157.82 | optimizer_gradients: 0.47 | optimizer_step: 1.12 +[2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] step=136006, skipped=0, lr=[2.0124363314442905e-05, 2.0124363314442905e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:43:46,437] [INFO] [timer.py:264:stop] epoch=0/micro_step=6/global_step=6, RunningAvgSamplesPerSec=2472.2740699287133, CurrSamplesPerSec=2561.3206693919196, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.84 | bwd_microstep: 2347.61 | bwd_inner_microstep: 1165.11 | bwd_allreduce_microstep: 1181.70 | step_microstep: 164.54 +[2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.77 | bwd: 2347.57 | bwd_inner: 1165.11 | bwd_allreduce: 1181.73 | step: 164.55 +[2025-11-10 09:43:46,446173][I][megatron/training_log:402:training_log] iteration= 136006/ 136100 | consumed_samples= 835620864 | consumed_tokens=6845406117888 | elapsed_time_per_iteration_ms=6863.2 | learning_rate=2.01244e-05 | global_batch_size= 6144 | lm loss=11.744588 | loss_scale=1.0 | grad_norm=527.276 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=895.207 | tokens_per_gpu_per_second_tgs=2387.220 | TFLOPs=26.70 | +(min, max) time across ranks (ms): + forward-backward ...............................: (6639.44, 6640.87) + optimizer ......................................: (161.40, 165.76) +[2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 150.38 | optimizer_gradients: 0.47 | optimizer_step: 1.12 +[2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] step=136007, skipped=0, lr=[2.012303984797229e-05, 2.012303984797229e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:43:49,840] [INFO] [timer.py:264:stop] epoch=0/micro_step=7/global_step=7, RunningAvgSamplesPerSec=2492.441497638933, CurrSamplesPerSec=2576.511402462105, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.13 | bwd_microstep: 2397.86 | bwd_inner_microstep: 1224.18 | bwd_allreduce_microstep: 1172.96 | step_microstep: 156.83 +[2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.07 | bwd: 2397.83 | bwd_inner: 1224.17 | bwd_allreduce: 1172.99 | step: 156.84 +[2025-11-10 09:43:49,849423][I][megatron/training_log:402:training_log] iteration= 136007/ 136100 | consumed_samples= 835627008 | consumed_tokens=6845456449536 | elapsed_time_per_iteration_ms=3402.7 | learning_rate=2.0123e-05 | global_batch_size= 6144 | lm loss=11.613136 | loss_scale=1.0 | grad_norm=579.721 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1805.618 | tokens_per_gpu_per_second_tgs=4814.980 | TFLOPs=53.86 | +(min, max) time across ranks (ms): + forward-backward ...............................: (3191.89, 3193.07) + optimizer ......................................: (154.73, 158.25) +[2025-11-10 09:43:55,173] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 144.15 | optimizer_gradients: 0.47 | optimizer_step: 1.13 +[2025-11-10 09:43:55,174] [INFO] [logging.py:107:log_dist] [Rank 0] step=136008, skipped=0, lr=[2.0121716391239166e-05, 2.0121716391239166e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:43:55,175] [INFO] [timer.py:264:stop] epoch=0/micro_step=8/global_step=8, RunningAvgSamplesPerSec=2214.823701527157, CurrSamplesPerSec=1422.5676763766571, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:43:55,175] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.49 | bwd_microstep: 2325.32 | bwd_inner_microstep: 1170.29 | bwd_allreduce_microstep: 1154.19 | step_microstep: 150.72 +[2025-11-10 09:43:55,176] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.43 | bwd: 2325.28 | bwd_inner: 1170.29 | bwd_allreduce: 1154.22 | step: 150.73 +[2025-11-10 09:43:55,183901][I][megatron/training_log:402:training_log] iteration= 136008/ 136100 | consumed_samples= 835633152 | consumed_tokens=6845506781184 | elapsed_time_per_iteration_ms=5334.1 | learning_rate=2.01217e-05 | global_batch_size= 6144 | lm loss=11.405020 | loss_scale=1.0 | grad_norm=230.690 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1151.839 | tokens_per_gpu_per_second_tgs=3071.570 | TFLOPs=34.36 | +(min, max) time across ranks (ms): + forward-backward ...............................: (5129.49, 5130.83) + optimizer ......................................: (147.99, 152.14) +[2025-11-10 09:43:58,569] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 138.00 | optimizer_gradients: 0.47 | optimizer_step: 1.12 +[2025-11-10 09:43:58,570] [INFO] [logging.py:107:log_dist] [Rank 0] step=136009, skipped=0, lr=[2.0120392944243313e-05, 2.0120392944243313e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:43:58,570] [INFO] [timer.py:264:stop] epoch=0/micro_step=9/global_step=9, RunningAvgSamplesPerSec=2261.5249756425674, CurrSamplesPerSec=2589.0806546406557, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 716.27 | bwd_microstep: 2402.82 | bwd_inner_microstep: 1226.62 | bwd_allreduce_microstep: 1175.49 | step_microstep: 144.53 +[2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 716.21 | bwd: 2402.79 | bwd_inner: 1226.63 | bwd_allreduce: 1175.53 | step: 144.54 +[2025-11-10 09:43:58,579283][I][megatron/training_log:402:training_log] iteration= 136009/ 136100 | consumed_samples= 835639296 | consumed_tokens=6845557112832 | elapsed_time_per_iteration_ms=3394.8 | learning_rate=2.01204e-05 | global_batch_size= 6144 | lm loss=11.290359 | loss_scale=1.0 | grad_norm=326.170 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1809.814 | tokens_per_gpu_per_second_tgs=4826.171 | TFLOPs=53.99 | +(min, max) time across ranks (ms): + forward-backward ...............................: (3194.54, 3195.97) + optimizer ......................................: (141.84, 145.75) +[2025-11-10 09:44:01,970] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 137.27 | optimizer_gradients: 0.48 | optimizer_step: 1.13 +[2025-11-10 09:44:01,971] [INFO] [logging.py:107:log_dist] [Rank 0] step=136010, skipped=0, lr=[2.011906950698453e-05, 2.011906950698453e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:44:01,972] [INFO] [timer.py:264:stop] epoch=0/micro_step=10/global_step=10, RunningAvgSamplesPerSec=2296.9740702454246, CurrSamplesPerSec=2580.0686696658236, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:44:01,972] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.33 | bwd_microstep: 2403.22 | bwd_inner_microstep: 1220.85 | bwd_allreduce_microstep: 1181.64 | step_microstep: 143.81 +[2025-11-10 09:44:01,973] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.27 | bwd: 2403.19 | bwd_inner: 1220.86 | bwd_allreduce: 1181.69 | step: 143.82 +[2025-11-10 09:44:01,981208][I][megatron/training_log:402:training_log] iteration= 136010/ 136100 | consumed_samples= 835645440 | consumed_tokens=6845607444480 | elapsed_time_per_iteration_ms=3401.5 | learning_rate=2.01191e-05 | global_batch_size= 6144 | lm loss=11.142616 | loss_scale=1.0 | grad_norm=184.646 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1806.288 | tokens_per_gpu_per_second_tgs=4816.769 | TFLOPs=53.88 | +(min, max) time across ranks (ms): + forward-backward ...............................: (3196.98, 3200.27) + optimizer ......................................: (141.42, 145.17) +[2025-11-10 09:44:05,471] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 131.87 | optimizer_gradients: 0.47 | optimizer_step: 1.14 +[2025-11-10 09:44:05,472] [INFO] [logging.py:107:log_dist] [Rank 0] step=136011, skipped=0, lr=[2.0117746079462598e-05, 2.0117746079462598e-05], mom=[(0.9, 0.95), (0.9, 0.95)] +[2025-11-10 09:44:05,473] [INFO] [timer.py:264:stop] epoch=0/micro_step=11/global_step=11, RunningAvgSamplesPerSec=2327.316936098493, CurrSamplesPerSec=2602.3285686919153, MemAllocated=3.71GB, MaxMemAllocated=39.0GB +[2025-11-10 09:44:05,473] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 719.44 | bwd_microstep: 2365.65 | bwd_inner_microstep: 1195.86 | bwd_allreduce_microstep: 1169.15 | step_microstep: 138.29 +[2025-11-10 09:44:05,474] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 719.38 | bwd: 2365.62 | bwd_inner: 1195.82 | bwd_allreduce: 1169.19 | step: 138.30 + +[2025-11-10 12:29:56,532794][I][megatron/training:1534:evaluate] Evaluating iter 19/20 +[2025-11-10 12:29:57,331870][I][megatron/training:1534:evaluate] Evaluating iter 20/20 +[2025-11-10 12:29:58,133928][I][megatron/training:1692:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- +[2025-11-10 12:29:58,134910][I][megatron/training:1693:evaluate_and_print_results] validation loss at iteration 137650 on 122880-sample draw from validation set | lm loss value=2.697780lm loss PPL=14.846734 +[2025-11-10 12:29:58,136014][I][megatron/training:1694:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- +Comm. Op Message Size Count Total Latency(ms) Avg Latency(ms) tput_avg (Gbps) busbw_avg (Gbps) +broadcast + 4.0 KB 25 980.95 0.33 0.11 0.11 + 8.0 MB 12 214.59 12.05 8.72 8.72 + 12.0 MB 12 354.42 12.34 10.63 10.63 + 43.0 MB 12 346.20 23.60 16.63 16.63 + 86.0 MB 12 492.66 37.23 20.63 20.63 + 1000.0 MB 2 62863.52 31431.76 11.92 11.92 +all_reduce + 4.0 B 3300 1828.19 0.53 0.00 0.00 + 20.0 B 68 238.60 3.36 0.00 0.00 + 100.0 KB 1650 108812.85 26.15 0.53 0.53 + 45.6 MB 1650 59143.33 31.96 24.15 24.14 + 45.62 MB 1650 72884.32 42.17 18.16 18.15 + 874.0 MB 1650 -259904.95 246.25 59.55 59.53 + 914.0 MB 1650 -296460.05 230.41 66.55 66.53 + 954.38 MB 1650 -223315.44 268.82 59.56 59.54 + 954.4 MB 1650 497077.13 285.12 56.16 56.14 +all_gather_into_tensor + 36.0 B 1650 61948.81 37.49 0.02 0.02 + 1.23 MB 1650 -527532.78 95.50 332.86 332.75 +barrier + 0B 68 6248.29 14.28 0.00 0.00 +log_summary_barrier + 0B 1 498.65 498.65 0.00 0.00 +wandb.run.name: volcanic-blaze-4312 +wandb.run.url: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu +wandb: updating run metadata +wandb: uploading config.yaml +wandb: +wandb: Run summary: +wandb: learning-rate/iteration 137650 +wandb: learning-rate/learning-rate 2e-05 +wandb: lm-loss-training/consumed_train_tokens 6928151347200 +wandb: lm-loss-training/iteration 137650 +wandb: lm-loss-training/lm loss 2.69441 +wandb: loss/grad_norm 3.39173 +wandb: loss/iteration 137650 +wandb: loss/lm loss 2.69441 +wandb: loss/lm loss_avg 2.69441 +wandb: loss/loss_scale 1 +wandb: +50 ... +wandb: +wandb: View run volcanic-blaze-4312 at: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu +wandb: View project at: https://wandb.ai/aurora_gpt/AuroraGPT +wandb: Synced 8 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s) +wandb: Find logs at: ./wandb/run-20251110_102334-7bjj8vgu/logs +Application 5d196584 resources: utime=21540179s stime=2039451s maxrss=46631076KB inblock=18009722136 oublock=2015065920 minflt=16872908673 majflt=453960025 nvcsw=11096789454 nivcsw=10717725 +[2025-11-10 12:30:19,247084][I][ezpz/launch:402:launch] Execution finished with 0. +[2025-11-10 12:30:19,247825][I][ezpz/launch:403:launch] Executing finished in 7795.27 seconds. +[2025-11-10 12:30:19,248203][I][ezpz/launch:404:launch] Took 7795.28 seconds to run. Exiting. +took: 2h 11m 1s From d16272e90ed8dbb9f5eae9b75836b3b6aefb7b95 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:53:29 -0600 Subject: [PATCH 508/590] docs: Add `ALCF/notes/assets/` --- ALCF/notes/assets/cooldown.png | Bin 0 -> 66403 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/cooldown.png diff --git a/ALCF/notes/assets/cooldown.png b/ALCF/notes/assets/cooldown.png new file mode 100644 index 0000000000000000000000000000000000000000..f32995abd8fff76511de23cb7a971a3c00925221 GIT binary patch literal 66403 zcmeEt`8$;1|F2z=6h#P?P}$194-wgwvNN*pJ7XUblAY{pp^|;yjgqYiS;xN1He;+~ zne7bfd!6$==X0*}2b>?c=6dFt`?>G;{a#+L*ZqDXwKbJ#E;3&vA|j$ud8+uFh=?3T z_^8WzmZ{=$R zF{Jbl>B+tGeD6Gx zcW)8j?GP{OuP{oRu8SS)gQ-qc{Cpa2W#UKJOzbNv`Lz4RYj)Wo%H-BSX6mR?vz7|jUQ;u|Px4z?&J~4wUg)>HDoAjHR z(0s6MeEVFydORsAI6(Wt<&gY5k{6+pzb_cJlm8}e`u5V5bej0E{azZ8^g|lc5b|v{ zUWF@f!g{V;Pkept*2||Xg0Iu!X_mj3g|cc3fW(83@TXQ)Gn{{=dX#g^oP=ZUvcFNxa=HJ z{v9;`ER7QKn(=Dzm8))wR|lV@y(L|YB;!c9z&H_k?UvP3zMmPZamE8keqfCB18TsJx%RFXn#FwCvz>&N5#0TUWO3;tBgABH!l2=&6`W096 zp}!B-MEAY+@A)xf*+Rd)wHSG1eM3+E3Kv5)tIivV5WCKO^YBMHX-{|=c)~`z3@v0T zUyTbp%A&+kH&GY8$)sYi!5oX%T$RLaj34q}O^xE|5LxhZU*Y^hUqaRPM*ln8V~G!8uG!mfwo{(oeQ>oea`wlKMQ3w~9qAn| zx5T3a#oKlXH=o`v@EnpKGI%bqes!HMU0s)BDj~GjZaH*WX;)-dXqWQ#*Z7<(M)!^Gsy7co;jB}&fLH}**vw=XYp-AhPaWmX^TGQ^-CFZ7c1>h?bC)q3uEpSwuBsN*^7bFIQ>Lxw}ML+L}@h2?4k(BGE4R^o!>f{d1I z(C<}-RkBr6R$9=@Kly+5LK}NHp{%LRTw($O{4U(=JirHY9DvVqFDstCeRl68UoeG@F zg$65?4@{Gbm`nZCz2v-9G)(y9oFIk~lO;_Bl14H{tp?ag3WE+^jwkuLOCelJNkvSb zeIM`TFqCxuemC}Yq^X$3A=AFwHupE%nEYni=G~2vjp)rpPlG0=roN^mzeT?-zeYb+ zzw(2Z`=JN4o3v}-M`kyEu#7ON-*|or#`K5Tn+fM!VD;nCS7ARhby088S*vL~Oqb%+ z-S5ieBkti_*d}UZ%$g*{KUyU)}p8F*!XOkm_51tW2#F zB^*V~05RAw>@pBF=r=4fdZIsSpkcsc5Mcn*-*a3O{UG|;(WmZR-K)CVx&g#{#COC7 z!dgG0M1M?T%)y`3zbr5ln;06_aXG?N`Nv>qWxHEXe@CBMaKaRvOaG4!(I6-pA~N0KE7>{IU;7|)KqKW z=YjbHHF{J3#;NhfyUR2eoGzMB?OiISF`E84Ygp?QveYio<$K+k(Ue)1@fGVcx<~YO ztOj%vELN;uckTHnMVom2WV>#Ey7Nb)#KF-E{yfWOdZJJ}S<4d}(_eIb zs=t`0Kh+}I>(U)p;X+f-rX;4&?$ zVOC*0PqIEv%`pcV;Sd|&dO{UCKXr1`zPilG#w_LN-|#`P=5Dj)#@p@$)N?;O`Ss>} zquAL8LidYhmBdqmqV={+XMK%Vjy^2$TpzripNvW_eOdLA@@tCbo7@``cA%<4!?LD| zli{K?9a;xt`$>mUd#I*`rcGARvsIv&Yq;;HO4Fvc3f$hrkLnhR+;E&ybypYH#HU5; z28mYRE#sa>|2c^u*u=-F=fe+X4Sm?Mv;g>B^lxN0e}lOQ(^#*WO%}}2R4%7JW;K^t$TX;Wts<0 z>zNG;uwepAdve6B&Xow>n%_{92Fd6RD0P$pVju%hzpL(>%FW76nd{;s`=~SB??7!sfM?iOphw`R^)+i|n;7fn z)vVR6*6Abnlb5^2E5{h%6L6_t>CQ>S96X4{_=0g-P)A_(iT|KcQN7JvddsHKzF`1L zb^*Kl{2OHvEe(4*JEbgnK!!oT@wc+~;4#efK_>#5H2P-rWeD{JW4U<}62>bH!R_Q^ z{I^_*|NEtF{3c0JLYEToCD9#;-Iw#LqoeiqL}qHlN5(BpO~%ztCyfk5mL$Qm`v9Vd zxtl594B=}8lqYVbuVSsCLG*y|`2rDfgdGtX;S(|8kD2gCL`0HD^6v^cDv$KvXQHFi zFT-+@=ZJ{ph*T8iU-%HiXDEY>L0)Tkp;qU)WZhP%{npHQ7X=H8n7YuB)&(lgD^I_0 zEprNly=f6nY^h02{jO$vxwGLFOLTN)Ow@F-A9C^|LI-4h2<|uOv-cL}K|XCB>i>0^ z8twfDg>DQ*W_t|C0Q5vNS+}; zfB*jUhquU3$8`UB3~eP6G8A;ptg`4oJcBUCmGM=n|I^Ut*W{R}2$RSU?fmqgCn0B+ zi~NtPAR@84D`y6YzRZyM?myoY64*xipQrdffhYtkD5>xG;6K@lh_KG%%m3dHgiZVZ zhWP7d|F4Ef=Gd;qA29CH;`axVniSZN58`^~=kv0DRcTqzs?0DkoTnAJbSm{u31N_P zuKGKHfOY8Z3?7HGFd5qgyRjHIL^bWO*Ylvu12dcV|8X z#XP(YV-MR5+~x)(YXzRvU!>e)A}0G35)5K?B+U$#8>PB*>C)RN8Q;TE)Yc5QmzUS# z-tn?!N}%AL?u@);#MRS$bEStq4enu!Ym{*j;l|dLmut~Q9;paKke~a$wgw4!4+Iuq zi&K`xPb3f>UQY$`hsc?IClfNX?%@-gyGuhuL!V>QLbX?Zl?#26KQ5_PYJ9sxL7`Jg z+~D?JXFXh%Pu$st$%#;9!6k(8(>cE#UbD|0a7eh$=Lw65L_#@t7ih8I zT`c=7r%Ut61K;^?xjjbw%`d($R7jshb>GeG`>7$;qVuOednw+CXnxBdnPiQ)+>1NSAVa)J9d7E98=XIcMBnqo+7?B(M@nL`!H?8Ho$ zre_R%#%sJY{UMB|u5Fmb%5D+2egXC-ENPY1rl{}wN-F2IOTFO0&8b}n$jPh7&ZR`w zpOSm(%OH?sc1lBE{_2`Zv-jrl{@`G%L8Ez?2V1?zx(R5GucPA*GH{@DNYdlewHyoZ zUQ+Ij!$Sz)_JP7I!>nn$c|nsmzZj`8@84}!>gA|^^D8VZ96wW9+-u=Ew`M2hCc{B=!8x7%LZFn z4(*pDNvVRyFCF2$BuU)cxpSBl%nC5j1QP_oq_*L!0eCf=g*p5}+%mt+D?Zbjr#b#O za;71v<<9Vl9|N05)B+9jX)z~_6ez%9nAE;9s(ANkNvp>tU}Jc<9FIa)IZcTk!&lqB zFkbTOe5jcjjbOGJyrgti+w(q>b1mJ>06XJ1%kiZpMu#O?zJHjH$Lb**t*wEvbM+99 z&(k|A~lw8zt;HVoE)~}HE!8;W2x~PcKq^ZOK1w;P-zEfYx3X|s%}~! zUFZ6}|>e_G-?7f;Z>PX*JXg|e|@xD|n1=0LnPq!}@7yV|SI%cani z_3c}u@D|U=rLORLK4crr(SMN=J&N3lv-9#<_+9Hy+}YXQd= zzDWN$gt0UR4Z%ePEO_L41x+Jo>RpU;lV-}?7=K+jH~$0UdCP$hVXG*_R6W#4By?7A zX7yJN=c+B2(bf*$q0F~AUs25>LdjfBHzw&l$79j{T5hWCIO6 z#QS*az(L2W0QA)EAP_I&yIb^m1+&+m+~{GwNrMtvSz^Z?46g6az29w~N<8G;-?e=e$d*rCu4wiNj%jGwzkPwji3>0oOXKPqOwlrOHA`*#nl` z_^xoJN?w`bj>B^NuGW(Xr%^asC`AC)y=&6n5{CC_<-iNYDc;f4kSj*8M@CC)x)cV! ztL+`<_a_PYTfJ+k)Vm~Vwx^u|W{4tLo~cw#TKmYhJ!ON{{_M?`dSx`9+JGv;7>t$5 zVjoHQkgHm_FZt0FN=c@?n%=`wKX$dwn$rjFx8Q@e8aD7Y(-5|IQO<;Ix|L};GMUZ| zZeI2O7<~Sy_I0|o^>r`Rsg)p}Dde{K>ZdO`FgPX^Is}+S{Yep-MLJVYuGSOMLQmsh z;3}wXKtc%vR*LvUUX~8o8ZV}D(Tyq#@ zzv_POw3p(RN11?}6lh9^x-(}NP^8nE84kMTiMb9mQ_wm0^s#(w(b#u~F{zPKLnCmMj^2Rgn%Jy+F?_6h0{;5D zJz|BnT59b0;J^ban~>etC&e^%V10nrvP7r-b@lZJ57SqFfA>KD=`&icJA}-wd)Q12 zacUm7J*F&%;4#z!KLZb^IoC%9*%ujsJ|8dys72{(!0uLFTAGZC?f-7Ty?^Wjqo6k= zqCZDp4>id8hlacfH3Gko#ljc4RSN^cK2HXDcy3;{!+Uz`pltv~oBL`uTD|18Z9$o1 zdAZTnP<%54uYA1A3Vd&?F>;*k!nB6pi6Qt|Sd&b@e4biVucZC*uw7`JaXtYnY_8%L*aKe%?Fm8n=0+sZ zVNSCxjhgfL(0cqk>lSQSO^vWr60dQCoHKr&vc_~9*~8f8*{y;lVna^mxf&gv)7N%S zh6=hzkdp=AV8(WCDsA8qhi2kTvvZ@JXA*mmPpPrT42|afjB-qWYSx0-tZS?E4wQRk za?`~#C!8hwh-@}MZS%kFjI!wbmQ2{M8KHVFRdZiIIYDPN9p+*&smmZR5-gdH zrZ@Q`Rn|JTK7aTL)H-#aJ1tXSZl>Je1iViP5x^e<=Mp!a8w`T57P(h8$ctC@|1{d3=~;+T%P*e|ZiBq02fUOr?dCgU>?OguSxkMfj~-s3k4x!* z3Ur9g9>_a0aib4yG8$}G!b-M+jHQ@y#pU>zEVnME-MZp(_ne>}AnrB!o)xOZ;5anv z&|CV~)KBlF?}syi<>2CX2UTICkvoc)bUD-5zv zz8fAKcX%TvImae6Uhx=vFB~le@@E|?)@}MgZEl_u{p1G^nO1+BucAtD!L5gFXsg;3 zr2j^dO;#+5mnl!5XF7aZ_!zL|r0?RA>p0$UUu^Q&;UI*YZd}0I>7liJz!_tb%ezEY z@~}l& z4(?54Jl7rl?*uwi9}vpK@-X@+mI;a)G$&;KAa>pp8as*N&k8vghF(;*cY$Z`r|Y8%$igGS ztza{hTJ}t$9>^5p39o@(fz<-BH-B01U+Jos@DMdF3kpJXm{kbm15SjvYd<>LKc%2l zN=dNSO06mB)d#EaSeleRsVkOh@hi&v+mQw_uetCLp)Z* z7j8X7kWWrZamtA6drDdCG~A1DAgqiFN>5JoMWmby~0ps znUQ_Ule>}vcPZHs?+sjBQc|c{y3{k#@oP_eW2Fvzf6rI^YA$-Yp{K}Q6O@Tx2^@gt z%KAjW0TnO*{4(VNyDq_2rW+amF#MbAPu;ww9n%Rs5V7Qpk2YelLN#Y}vAS@ea5WlD zz|ExXl04qVz_JEu{)t^+vsHx{M~n@ux1>1J3O{){`;*>lA(E>7j=iiEueK4ss9At= z=q}R)DqE(%aR%Yh3wN+t)4<9z{k)`x7HGEY+^O#u0UADNM#|*DXcPo1cy33@ zK2*MpYw2iAuY_~#Wjin$-G*IQsuCgz8RoBW+tZzBbcmY{nkuEcM5a`4ga0Oa+_|Ks zq7axZEIb0WV$2m7ki}QkczLRV@xjKv$DD#L{k)Cnw%n&4Ivr_N*Dg=B1*SUA)_Hgy ziy4Dzt8@H=Fcpr|HBW+&PzWLbH!+awienqgq_QcKR3LEdM@=#<@@G+&H%ZKQluQP)BXGGa%OZLiTiSXpmRi zkwZ&&K%HF^5SjEzsG9<1fleGM6}zY#=v?V3JIq&!RawQBvEQsPcvlyLuTWr^5mnX= zvQBWsf}6AFzbTHnP(#TN5Uu2g2qBt($lyZF(Yu@}k5#nU=8n)t>9Q$4B`xbf3VQc9 z9F{0fS>U3q4limP%>~}+V;qrEN3%Q9Q>O?tc#Yl5oGGZ!!<03npqFUI%8y1Ke{Rw0 z*kG>o2#m;C8fdX7@)0IB>zA?-he{nijkx5{J~uY?nbXIymBwZ)mcM#Dla#hzZ8N#A zYF?(=5*ZkgrSi^%ak+fh`Qdm#AER(jNxcZP3o^H5+(E`1Fm!A_ux8(&bfa*dhGbZC z{woiEGIhU<8Vb&Oe=~imJtnj6XP=lckC-XmTOv28q>oa1&&tf#OxNu+eXev}{HmGp=YX>mMDbQKx zDC)18_&}~p7D!6=NxNMSnQR66C9xo~k9@0!c=vXk=5+@7D-}?TG@TR>Z02(EIs1A2 zg9lNwQpy(rMqP6ivVV}fO|b@`rA9B!hqt7;zZjex^rHuP)?XL#5yfjfZd>+d~eTaVHM-na#lp-^0LV6^};$$xhDzL$sbzCp#|-0D0s$e*0wEE zoxQi~0Iqz=&<3hI+W&)fr1=HGjA{&hoj_L$_sCXo-@Y2S*q;IE6)5VkQxQ2)5?XLN ziiR?0I}pX?#7*Uh!{7f=+w|P2E7*5}gSw(j?~K!g;Y^gFHsdYFeLy6rj2NDA$L-$A z^o@JoCT4rfV|GZVwx*S5k?g9fw$4w#m&&g)-}y9^;}8ZT3(NHw-eDvr(~Og|(8+h4 zY|%&&Fp+K|ro7CS*fa^v9$5FBybnuKEzBAa8GM^Ex9&=UBOqqJ`siS5EaPP;A50Z} z6d;S%dOrBaEpGYF))YwF(5C;MWDL#Cv0XrGE7qdx^Kj_m5P5SPZ-` z>+#Wf(r>?ev0?wmwM%S)%v+X>8fJJ)@^to)``0cAqqEqD>Mj+@Oxac|kdv<-3;3uKSjMzTrbpM%4>3b*! z-<1JTwVA3fp3*SR>T+kd3mw%YL-`?on>yWppd!?+1`a?NXPaTJPTlH3zZ3KEQ09Kr zcSA#EQh4fH9e!@H0XIBbqxB7BQ~RI3QjcHmyX7LgbDN6B9MK7jq=b+#v$gH%Z5fz* zNF8s42!xO?iacD{@wr<_j5DqtFzp)~80{F~%O8K3*^O&Uh`qGozl^vXA){_#a0yLLf?t4B&L0^Y39UI zKn-h-3P5@WFLj6t-Gy3;tc>PT7dtv=33$n+P9pA)Kucg(@u3xb^u%dWc9VV;=zU953HVOttGAly~V%k{sG8w~+- zlAW~>hLZYM#C<{aL~UNL9heTPt?PO$w4@R67tTxseG}1*W;f<3*5glVtTyU?@^piq z7|^mx+_;_%iym8IcQJ3S)E*C@G3|9^PBa0}eA2pqT!8!dMcCMoBM9qK8TI`~m~p@v z=({IG3}e|8qb)bBZTTgd+4m%Gmt4@TkAHBoVXP#jYzoQTH(virban|8mon9hZ&?R| zBv}7Uc#f4U88zS6FA*N-xz>%#gakpy>os>uNau{9z+4(0Ojc$rgaB|3NC?m~A(i`&(FWZV|l4@!n0vOkdD!k_Ga`!M_@r(bC?e!q$F$L8Ko4H%GvuP%IElQJxK?+l$Du3<%huUP(?i6{o>u{C+jEBjJ{2b4zr~M?DRLG=vYT=?Qnzf-oR!?1wKh)*7OI$wq_{UQ z{XhFU{(`-qPuk0f35O}rL363|1DqSmW)Ff!;;J7|aQsbK!gPcM67iHJ{0%39DL}@m(RmnEGm%fhEPL>jlP(1icVHGCC?ARgZfN@3IU~ zR&B~zPs*{O{9&}}mSdcO4Y8)`ujj>8iphUAZ`S8J=Tk|y&0s|D7t=4Pq=fvfYyVDN z<;X8AQyw^h+~ouPUB3B)Ott624U~ci%q%=cK(xfQtq9||qJp?Cp&)f&%KUyTlcf-O z(BLUE!EMx5r)Lj)Q;0L~TLcK)+2YU=)>uJRIG&?wP2Kr%q3vG&gyZw&pHs4P#t~w7 z&#d-Lr>L26{YGsmBIQyZQ z|1-ybSx1|SB;^}&$eDb4lFD`1XtTy*#!zKT z%^?^9yo}qfxY@E&?Ef~Z>n@3M z)sYNy1Fb_{(9_T(p;`toe-s!^D7WDT$2s9q_CHuhnq-Cl+AxVgx`xRH?Eg0t60(ta z?Gbj;pkglH>!|i-9K|kQ6y&gz*beQMS|jtD&ui6|0pi`@AV$=)n48m`Py2L|)+ z$x%x?ZkNt*<7F29I^C_!N24yCwLCYSf2v^kQ@>XT16MGg!|EKfz-@h2)+mRu-}0J^*OvI*UPvS8Nj|bU)IqPFy=il!LTdfOB1B$ zOZllvs8dhr0#%sa(Wo6oOIAjUGoSnB=-qkut0My4O9;yy2kxe^ge~}?Xs@fcej=rC z6h1kvMrMSQ5-9=qteJA9GQh0(1WZ^H#>|^0(=;%HOq-3HoQ_ds~=wB z7rw<5d06-Uwf{pAlE#x0O+K72ak^Vz=stukHHVSH5 zOQQCKV5qM~!PRPB>|hkzIb0s*c*vWv;_au~XZUX`LVD?3?Nm7Rv>kRRpKR1#(?U$V zQejh7n=5wv%!aQkN6>?l^qBce6Z}fMwTg0j6Ky2UX7fLtPPgA@bgkMKY0&z+4TKGP z(!Mfmh>?wZ!${64%zcXwHll_y-^iMkLM0MhtO~Sv}*M@r-nQw`+>-}<%2xlI{y@MMZ!F_6eI2aj%emi zZhVrT?1=3Zv9kNOS5D-{8Sh;V4iGOvriven6%w7hE+0bnpH8 zb`KS6UeNxfGfRYS7Qp4p`Txd!M;0#f6`-gzd2PK)|DStwk)(7YM-4Hi(8b)9)Gi(e z&N`6=&j$Z(<%pH-r9e&ApyF;oVkaO56s~c$BjQe{Y6Xi7&)<{Acc&C?F?68pZvqeI zh)?YV`a8e)$kQ?8?o`$>6SS^Bg}77?V}coMssWVbo$QOFQm-IqJ42 z2Wg(NlRShJ`B@X(m+@Ao8f=K9Wr;XVd=IS%Lw=4x9j)phhn`W_}NmvwMAcYRDFts zk`5jFlq;rC6lyx`zzi9{X@;2LBipJ|hiNY$bSdQut1`U#o44t&-xzy~dnk{MbbrK3 zSkYOfoxh+@*|e4SwqCdTC^OJ7k^4cVlVKEC!T9owm%e1sBuA}Hcs4I?o)^;4GP1Z@ z9C{iqUkT2yDpLh#JRfCPzkS8MS39K#CoHSz_&0UMw#Uy#BG<%uX=uC_507J?t#52y zh0JjeKDC@)kK(g%b=7Wa@-HkbRGIcOG2_d5`czWp-ho$bzYV&(jA3!7LGk;-;H#Er zAuYT?E*c|6&46(cf&*4Ox^1F$sOIy^>~0iLYKAi^*X3-(eF?k6gG{SZK3~`S2%_xO zlx{b3p0k|v)o^|b9dxHpgQm&@#@2|fGvd5)A=*zGje!o04c^CMN!?dgQhQn?std;V z!#bw<>jMtaAn0Nk_3{oktk((Is50g<&YM|}xM_l;ftm0HHcdAp{hvy+#RVLp4r9ww z#Vp8`?05WY86bKN2y`LtJF^O0MFnELQ;%&>lVsaU$QYemC>vX(?W1}NRc65J%gdg6 z_{x1kADDC;i&4zkCdnC4UuqV{Ym%Z$R;JId(_x@Y>^)*Rfp|;>z0XM-tKB+mq)r6kJJj=4L z!$e?pAk})%R{WbkC)v9)wX2FhSPr5g54+ABNZp@M%Z71RD*q{xIl*cj#~lgpb97AKXw>v-4+J=J{1f|*P zc};7~EF1kg%;rKuGg<+2$2W(oq-6Pybv9z~(X(%EOoel1_PtBS^osD1uX*dKbXPG~ zgJ2Cm*bh_#6=FtgZvh$sh>=AXaD$N~+q&D&!;u`ugx{E8{V~8ucoy70a?`q`TINzamyKCF_bwPL1 zJ?q*@j#FB59T@Iv6%IGXX*On=4>Pj-8EX&P?QxHY2=DwKv4tHcrf9?1oDu@JLuTb1UdnYw(Nl_V3oy}zkRuJb~` zo)qAQ&VV<>8jfAu`J5HqOwm`;3}4(U)Hj)lBoSvqPx_Y4$lYoHJQMkJUSj7*F8&%C zl8lqtsi_|F?(4}F1J&)x7Rv|4S+`LNKXoR?=gi9TbqFcE(*QcAk)uG!iB(HH=;l_U znt;SyBUc0dTWyu!ieS1_b8P`@ZRVwe17)e{+I0HJ0{!iRva(MFRxdF>1`}kW>%uwx zO?4NWb?BxN<8!zvvh@1CR_J_<;KjYeb5(>A(l_yGG|%6vqCSsi^*Xs%GE@KLEU#J- zcUUW}L@`fEu>pyy7yTlHkf;v!T)%ph0cIE@e@zAV8#8i@;C6~ zuG>_ z3oqJ&v3X_1tjzHh!`GH}QYLJi9%d)xY_Yi=E0u9$aKW}9D+;$@Zw&n9KIhlTXXtT%PW%2_8u98cSej@he+d=B^wbk zV##%)pZ*@wgj8LtaFn9{gDw@4qEYk_t}Y{loR5_CUijm0gM;bO)^8f=7Tp>-Z(fe0 zKsibJ51yCUF99LDs)+>2)}y;q5bnVT*$Q&jU)UE7_}JkNoktci_E_zpMcUa-ue|?d zRZBOO5+Bk>HF22jeTw7=C4D4^g#j39LZ5TuWEpejUCQLKEaTwAK3MOEy1Kz4LO#8< zn$z0lJH3UA7<)&4!Z-XEp%IZJ^F7pq!fV!-c2O|N1|D5KkTh}g)%jr-66MU7<1pZB z!*|`16&~p0mp&umpmX0P6yo*kuD&LVqr5@>#~C6~U@roCAEq0=4D!!Grgp?RFyNU8 zd~W)!A_HabBA_)Qs+J>I?mG2Nu?pZ?6=;swpv5*ssNtDZ{}gCG&l))(koD_0t8dZ* zWf8S%sCAI8&gJ3n1&pnX$+_RN=D4^7xRAjpv1m?E2I>w}BOIFmyuFYf_kk=>`5 z({g}mYAU=bxhV?Lml~*r2i9k{2i^BpKa+K}aU>`ypC(h%_Yi*nhv_72e61D&z>cR3 z;Eqla4g-u4eH9)e^`kr;U29_$574YI#5r$d7I#q^w6mtjJK@8p}{SF5ZOQ@|fI5`a#zjM0}A$MCF_eQs)Vb+pRvr z5*VaT1V#2OFjDPNY4;rW!V~$@Zua+vd@zd%-1$6k0Ax&N8Nsd#jT= zw{knuLh`0q9WjC2Xq^TS62I0xG&;(GIwa_jw>_-ue{#gh-bNT&h+Vs=5)>hktlYu& z7ZQC)b_65k z-74XN>@E9<=7Vk;d5sfPr5piu<_(d@qJg8dEr)+BPkPkNWT+6V9~m#7Cp8XhPRZ?LALGTT6L60M1vntW zdIw))L=c3M<{crz#xFQtAzWcA98r*q5H7eNp@LPlyOE-oN_plF|)Cr4|`H5njduzF^+CIkL!MZu(S6j=E&g(#OLvVJnV zJ!r=FStLL0JMPDPxQ&zW3PMZ}+?e~?c=k9jH*}s_PNBd9 zwOfEFfH2^e*Qyd{rPPY$PW=p6U$TQOvh0$|jKnjlnv)-rYv<%d-8dLHL|<>N`WOAD z@e)(RX=xoKa&GU8u0K9z6*H*fJ@4*0-4|Vw}TPO0voc*iBb>_LQCa(L$LbNw=4CH2^HdqM%IazhvjcG)jor(-D;!v z?F6_=6et|$lm831FKgy_2I|`jF8WA2s}5>>y+_T%NDOM{!!ZPk9N!hzsy^};8agyI zPNWB6Xtmdd}PJlY(kUliZ~w$CpqenRs3!oewWG4@k=V9m2%pB zsrIu~zQO&!tl@oQFvY>l?fT(LL$1v0FxPY>ZRX#GWhxTq)^8fwSi1JKnq4yCq7I|{ z+9SkEcp*|wiMv!C%_oB!)h#@>9zNh)pzJjJ{VvmSCshx_>l<@^_4f6p}^V_#?flam-S_ToT0>a(!Szp_HuLWOo!ZdrqncX@}An3xUq46R)7`Aye z?z3NYoOFh7@*te|wh`oO+aaOd_n)t+w|z+-P8}EBZ-=NLhgKYkKH;*TosqB^Z1To?6q<~0iv%2E_)TLVY+HXXo-2p9q9ABsMAFrhuvJ#w9DE47%XgGloRaT zpmqvMzdp$;D2_DoNA)(XnkK4qaoFTG8e$CT&zeNete^EZSn8f&0Ta4Ix5*`~9T#~+ zkoJr)`Cc3!MtJ5J<#flE$_A1+L)0*R%*!x6+gRXGv5+QT(C3aB8aEo$XW?!3_so5Z zv<3`dZ3p!4QyX4nAQT9QE6$hzWnc0h%vK%#6JUbdBQN^gVJpqdxSl&FfKrJ0FP9{g zc||g5!=Xo(tXuY?U_|=W-hJ!4bwc;C<(&nGg=u7~$F-Niiax--qB=FYv z=>&%loT|4_gTJ3(t8V=3F%S932yDNAE;OUm+bzkk+gFJ;mpyfX>|WE*{J`X>nz;!>p@ zA!QgfS9O$c7~3s)R9bDqGi*T|M2&-$TqF>BIe0kHX%@Jre5#K6jEtWunWrgPuKhY) z?wdGp6BeC`l5_eLO4PpYZp+c$ZF5|YdiBi3NT0i(c@ zKB_#Z>$hdfjupG`{=qzRQYU-7r`}UK|L6V1hsiCxxY3MG$Xj0A^T2e1;7T|{COC{# zdJr<=u=m}H+RB15TT)!W8g|j!LuzmTrsUXPnbVj0=p#q!?XnuR7PdewM@S3pJ)^S} zt$me*Ma(5RyBXDksGoW8P0q|S7sc+F1eC_T7aoDx{cbbmZ$1j}MIEOWmgei6pedJs zLIo0h35xN|AyzRXlevpG-*p^?Xtvy|%g%>F7MhB)Z{a}@FoC!?$gzRlv!_uxnU>xC zGF4m0sg`5En0BYB%L(t<7)zaYe9SA7^ETeSMr9k(Z1`z4I=+C`Mky5IMC%z@-@kOJ zzTcZvNH}V)wZ6>8sOfm~vjIh)BAxHf+38 zD^wvp^ha%_uBb5*D7@RSZleL|=0Aqn@lRZ-EIIiMYrRz}Wc1^+Rrqdd<(a-x`cF>j zI+Ihjyiw=eL8)KDp>`Vo+mQP=gn-{0mp|?aj{|^hUaKbgQ{P$0xD?tE^#mn*)k@@W zI6E|P+Q9&0w+`kgJ2;}eD}MNt57#Nbqg2!hVA|p(ge;dG^mvybpmHsnh3n)UMbUZJqN&tB@ma49zEd%2<_3kW^twd(xYh7 zZ<3-Ol+%fkQzEZ9$y}=;8Etqt&*@F=``2w9?cZx;*2hr6*~NXHhMuHKv1E_Ywc7Dz z{_9JqsRi{A_ifZi2RSrakqK`?T=loYLtc_^+x46nO34Z z#MH;uohfBg$D0OFW*a8L%khdDDwn&4ZtQL>Pc#q)Z)~h|eJ=Y?8xgsXq0Q-^%`aGM({3$`e^L})cBMYYMe+4fS6Q3A^*Y&NfnC#Xw?xhDt(!E*u(?a0BBBi+(afSTHl zH0sB2e?X(vH5Tj7Si%{7@y;77L6CQ`P;05nVNPJb4hd^S^YV~de`+oORbYHD2F;4b z!C)?Qo33IvFG>tE;yfHhf@N)_Iu=|P3Idb}-I%5bn1pt*JjS;pQG$b~2qmuq*Hm!_ zsy4ABQrm2n7F~8!5E*aZq*8RMo>yU=`=@krW2O{)Og}Uu2lg- z66X$T)GSGwB|vf)$8p|Npm)NrK_dJaLG#t zM8J+R{YD5KdC8tJFcjt+#$d|;woxKL@WkIlU~`-2=2MC0A}9Gg3Unr4=S&a*XV50= z;@@rtS@-8TJr@yf68<&LEzQeIe4)>!mwF%GJjZ%T^!kujQ6`up%AFB@>#(p$zn2dF-ov`$mhFk{S3x8F5KVtkvQwA$7P3&LQ^V zyGW}}ChSL7C-Z{p1f`veJv<7o^LJ1%vZQ3n;$0Rzy_3G3z^Tb%opo^gORZ;T{%6PS z(PC5`SH)k2ERsk+09XjZ~;mqUo?^y)AGAF!rPB>7l za0pHmL^=6-?>w{D#%t7t{ykB>MQHqIAlrT-oPXcxz$Fl~UGyDEcn9FL?monl?S`Jz zdFgs{!+3p={nQ&%uE5!&Y0ofd#X@KJ%BKOmk$AR=z<5-6yC&gx=T8wySuyl@*J$+|9!`J#`A?w_j1nJXRo#9 znscswuXfqPbafMoG6sx}4`n5_Sg&pugQfKX<-g%CMKJ<0HxChS6}+Dm7MFWqaaNyrJ>$;mwm|sv}rBgS=j@5=}9J! z?AuyzWTp{C{=PP>7SwW7D;5+if8g1#eV;i++#g>PQCnxvHc?+lMsyM;|SF}TnGDpnqrr{J$_h6!z>1bh3sZ%Q2Q z51RZ1nX!iu6&ovwz09C|ycH1Y5vZm)0B|aac1A;0#mxvY&Ja!bluFtRj$zjR^l5c?0Ok}4F# z9QgL-wUG?TKme5@>nTzH%At3auDk98z4w0b{LLXfuFL8ZS9$}%^ygmzORZS=uV(4jpYQb0HW9|hlcrF* zeaEQ80J|XquwXr2W+7IxJm;516gns;w<(P)_M^DUM!=H+uPbW)YJU^ zlJaIGLW-q}O`~HAQo(rKr9-dR)L!6rN-j=j#YY@{Zl?NJzM`>#b|oWayC*jT)W_IJ zdRe-)2uzIi7@vb~VvIRq3ISk1gh##t7~h^f`CnkLbn2Go@PU2b0n#+%^pTV6WR-H9 zm$q0V!WLt#n`VXo5gQ17dUAE|3|Y-m{Tc+Cf3p+-S9tjHP%$FTjGMTiBV$D4?l?{I zu>pA`dnP?nBV8Hd78h5rK|v3JsTgdmWI&RpLcA!lsjrHBYG!T8#VEptKkQyB?$Nrt zvtyS727`&x2D4v2vF3Ck5EpiR)E7yCnOSGz!i|k{cdBQ|p?$fNu>RJ^x95mYqVQBt zRX_gM&*tm2^sLyNOAF~PcO-X!p&@HcpD8bLQhDyYz@3!Lvw&pXOiJo3;tTXa<1ZEL z^_qbRrXBMrgv~AkI@GDnkGK~U|IrGGzS2HHt`x*QXS?sc#ZRGIe%q6Q;0HLeonQJM zWdk!Q5Q%Y(KzlKg44}5DG%VdTfbAlJlsI=1Bz~1-S=f;y%)enmg;J(bu>af$wh{Uk zBQY*q+V&^30mH?GKjo;no{zbs`hSNv2-TvfwJg9~UWC%4o8kCqVBt?y(|ZGycBRIh z16!DG*C%>F@Z1EbC!uwe&xBLD6}IU>vU^;v$7CTB=zA%lCy3n_3RYo~m0(LoRrwz~ zlY$>idWmLh*hj390Y^mkFhMs3i=#s+v#xrx?J-Ml`657-<~hIWSpl~&LDdHcr9XCz z4Q_6KjlM{!6~5z05mwR?`TPv0f)lS)R6qZ_^sneaQNwB0Wxd)rT6d){$(TxU-Z{Q( zCdPg2Qk_y4;D!_hnt0y$u3E4l_?lPd;{ZRFV*GyGlkNFgQG@H<%@Ni;NGJTbqdnPX5;;k{kWi{GYd`pHGER=+F)V##Ptzh}b zMi&1{kNv`Aaf03o-C0s!SHpa8JizPbHh-+G-}?_|3tS@=23lUi;@ATDF?hDe0rQGk z$i4&*l}D`nS?K25sP_eeK>KZYobeI1^%X`U4A2U{_w6PB(?x$_^8*;OkxO%#5h8!` zwwK(HpqU(yYCh1v`-vh1@VYj(HwmoS#@e&O7Q~2;& zJdli;2>wcU*OBj)CHC(Pq)fYHab`ySPNse3pOp41|0OCx_pP8aAW0~cUs*&N9(TD= zrD@!0%b!ag_PwU*TZFlrIE>fK%_VzmBS=dGy`s=Ge{T&TW#p?0O8Fcnx;DU!u~SoC z{^wl)`!!dkCHvUD?`JhAenoQoeR;wU8=`}!mQq|yKIIcDy6%9u&k6xp7ohA*L3DHO zC#(;*KATd~$Jn~w>|<*FwKN!!F=tV^uq5qmRn=L{X*{sfB?+S)+y);+wH^T^DiF3S zaDbv*qviNa*A6j(=*<^Y_&MSJl}S{{l2-qUZg%C>WiD0IsRYi$=>5p8l+2q`Sh=M1 zCBo+4VsTE;V<3>+0kr&-0#l>cz5VLBO`HNV7ym^-(VrUOSBBMXT&@q+a+nM@#KxbJ zHxz|AOHXlFX8M&s*O6=H8DVbYUDT#2;CeM$4%lyY9-&DsJ!DM)?L-sj!5Fw7RiW~$ zwlMbYW-i)94$5v(IgH;#HA z*p+r7(E@IhG$gEC{?i5IN_{D6e2>0r>dwc+S=8N7HnntMYr zP(oIeH#uzQ`jaITEsO>$Rt{atOM8E-3;_`_`;Ri|wH6W>P7UqoNbEeZO;;wH$$APb zEVV4)Zn87rCjr+Q*UD@QT-EKW5gZUiqV6X7l9b|Ao$wl|H|?MDLCV z>5`vr{EV$XybWRFWDi#tTQ1q3b?Mo#;uQz^v^lBr3PT=+OZ5iZf5K%CpxFvG+(Mg7 zp}@l0pYnpal~D@;Za)&q)q-EnR6HMJM5`A8gGVIQzkSj=wjl- z=wHoK40^!NAgZo!70V_<9O1U&S=7Yhq*7xk_7U=**w4}WHT^-2Hg`B(ulsAhpy~gw zm;Ylk{#r#;WTDR78e^No-=b`iuv&J68kM<^;|;Jt_63{iCdvmE?h)7i*FXYyli_wj zRlC#MXE^-PEpzqFSA??oO!aRbdjGrmKt1G)4YtkiXE z`7_hI!!yg*7LUekpoJYAi*u|ONH8gSD}6#JFvqY8e{o7uRK1OeuyP^Z>R7YOb@YqJ z1yUv<-BQ7+RF#cHs_s=+q%qQ;7)XWj>&y|Gu5gBtt5%;v9wAlL!Nea4 zrin7tP6VTwZ76+Ek&(c_4bo@oFWhwp*W@wk(8nYfQd0hRJ^W_dvsex~3+_f+6_1uI$UR8II^m6XuJ@6J07hR;`-XP*bn97dZ#8U|U>F7o1YE zC0YYRBl>NY4e+af8H)W(RxtcZ5!!B?2S-K`xY8*w)#V`RWd8vYL$A}IS8>tAx*X$U z>5*JmZPs**i(`WEPT zo&fy_+R5%{EdaC=`TR;vb~x3Gopc;-Zx^W~8|gB1Ee>Uk4IW|bBSax(Ci81S&3JS& zH!`6~+{3@Kb3;CqM~eJ2o(Y^E127biwv%$vU6a8~;rm}x1QWm%fn)WDiRni8BMP*Pze2+#!}SH)THi)pI(Up65Wb%0la zf}lTnr?n~ngcl{x-Y*vBmYeS54L999wn4I2Xd_=d1}VK-xh2= z7#Kjm(&;fSKM|JQAs7q4ZocpG0f9Y_N6Y@G@)qi8?%WRJ_9l`l=nn*ocj>NPIAx1pIw)Y|M4OyvP6I%6vLrNu^ zLT17y*>^!d*ltRsRAWE}Ml`+wyfw)sl$U1W^(tYyedlLQ%%t}=?NndbuPDL*FM_o% zJ!I6gLCwqU@2q!B+=MVO-9@? zGr4WVC4R7FrU0g(qB?`xz8WMrULEUjANk91uSRW#HBZyXEzcX|=>Ni3z#o|ocHj8P zQkgrgK0_9{z$P!^krQt9drPNd2P*|ZO*~?cR>4WY)ZZ3{)+d|sP;JZ8cOY+(cGv2C zce45NVnOP7$_XLcHFTR9)$ctWb1EKj&&pzoJ_jR5f8Y_wB-1bxUyGZ23c%mhaC}YxbBex= z2H>xjeju>Ujnk%E#ZD%`(53j$-y?uP4{m_F(iL(eVyeRVwVHnU7)S%>(p_!*Bj+;B zY|WK&S^{Mk1%R3r_dnP8?#l}h82BwN=EoKx16R%rqnW^B18A|7sfcLLJ<_qNpb}`9 zHhJG?0}M&QSxfos?{CYP_#JunwL+QX>B$7kxrU5!# zIXxse>uEMsl}O*2sd#-O6S~AhXAN9MxQSZdRX|0XcG*?#7HP?p8Q>H3(ICd*c=G!Z zc5sf^U5z#XKkGizrF0G_-9j+Jh2#wQfC=8?ufPFpf^mQ+k$!^$Ge9?*5nqFFWsuxs zqkeZ$_SBhyoA$er`^O>yJHVlq-^?^#atZ*oprD}aYOv2(c2lwQ8NbcHwxZ`*Fg;ob z=T1y}`sY*EqCROu9~LiloWF@3cwy_WmWqxg_NOrcql9!m!#>>gYQ0$56(UDRhhQ5x z+gQ!^}h!0VfND1hi!u z)c+OK@Q66nZRn#9C%eoaZBu|B8IJw7lV=P-Rew{<G_`#M zV?ACki#jsw&5wf$e{K>s!vXS=fjrD8#&?XzbRJ@;n82V;dFa`t{JiXT58=-<$%uXd z`v@I09g-cp@|`TtF7Ez3**j&D))db@+pVEvY8Kt-PhyXG8_y=IRDMwpSX>}z8ockB zI<>9KE~H`p&d>i^!F)xplAV1=0)AMZhSW z7~A6A=$uNhnB>UQ6HDe{>>GAC4BPULetS#~Tw*Qz4%|OHP?jb$aEBMM0VR@wETj=| ztN9uod!+FIXcwTN_5t12&mOP|l-MM>lH69#cI?~txHP#`@A2A7Yz$a4)GgvvxNJMt z8Xs1QHpgv)2#kM(-vMVyT>qvQgxHZjpK=dbFXSq|++_ak2Y!7(RxaRf&v|9_&5|SY z;i)urT+ELjKZdV?!9zIQtE`K=;JC^RS&LR*8%O)UPhJJHZg0?R={Ic)gxV4SmKI9s zA>WOUbAdO$ZL7*Vzwf19e#J>cIi#-+j1u@b3*Gf6cgQ#Ec)IjT0GyrotM~*VJ<2u# z{D{b&Zl<}-Img=NQnfiZf^m|EKLk1ktqpYj%NVI#ahF!>t*FU8>Xw$4ZYN(^R!7Ry zg&MX|*42N;r(0gbT&&floY;5O zFh4GK;wxxX9UPxW_tK#hut%_OPs*;?+XO3s*3e|N(fN=wXjlcK8+;mq`^1DAdOd$^jXR3M^8fE6vG^ueSiM4ZH8RcejaxyZq zTCd6IR-F6;a*#-7TbBp6)^|L%nGkF9-7jN|p*%!}5u6Nd|LOLR^zJIME4g)^I)7|O z#=#iew|&Ww-Ga;UM1K~-B^55lRj1wYa8BQv(IcR z%iu367T`{}U$<7Czaw1u-+&kDfdLBi(2k4FFTWNrC{j)JICz)`Yj!+g!Qc(}Iy7;b zY{rY*FF1S1gVl~5jFhy*P4%NwDbEI%+y9rT+iZzyVeT+-X&+$*h)1E|HT3k?fvB-vL#MA zsjb$dC80}uCMA5&hmS1Cpb<;`*~mg}QR-mI&lb}^*-EugZ~_Fa~gP+tq`!D+vp@6_eW$$hqP> zM{f#L-}?Ie#9V%y<^wXWGP#~`gd)#V%eu0Go&>z1!X!QEG&MxEzwz>$CDD5YHz!;z zngCB#cz(g!2w;r!OGLm};?rn+dMLq_bgL4UK@Z6rGD*V5fp)?sC$a|0kDF$jcgUB%FQgCB_~zdS6ib-*Wtx$9JSkQzKzh#KPb zX87K-*->yh7up+ad=%6nbvBj8%yN~xQ&O(=XDc`RE`o^^K8a9SNw`W%jM((^WZx$b z6vPds`o2dj&-%pt`cJV7ryH7Bc=>Bk$81DfIIN>`pu*X6ac+v4C#RrZ{PySZZ%fPe zsO$e&TAbsW0hr*Z?(gqEGCohoz#zLa5$jc(Fy>S@$A~ObVv|1SdG|Qilq01FOe+GY zT#&U8a5|vv(NHz9z6`oN=eFzrGO<)m00q5I^3z29P8=APR!-@#FOX?D-5$KVAQ>Er z+fb@@U-~MZDvjHhR(xVzow)G`^9Xy;&wdtY|G&ihHgWWFKAN( z2Y;tdCI1%S8x2Pz1t;xsyN2tbMl0vU{qS=>C9)8oiv?DRvR%GWzIcK1S9E=jmExRr z$crU#ACP^a_+Lt7(Z)P8>Dmi7excFPN&AL?U}mx2=K`c}K*QmzIx`$D8iK`5`BssX z7K;)Jh7?KsMJqesP*a1Xjm6?-y<0(fS$Teem(wNpyn2$`>4L$6!$9L^!=lT=>1u|+ zN=~lH{AV4I<>F4d_XYgr4S`MlV#7&Y@EsY5;Ah<-WU0%20F=-%?6nucxo9UZc=?Uj z)3CJ>I9KD~q|X_z=fP$1Cf$NV^I%%HdF?{W>yz?_#$j8ZW7S2+cSW{SO;&x$1(gc4 zGhF`1f#}50QNH3!GUQOrcqAdUkUmB0rThwxsfgZ`=Izi!B`qVLapkVvVs-1(BEynbXsF*T<6)zf9nqhFuenWrpo-0Zc$vp1zBmnOEc(713q7yVqG&a?|M&s*@j5M@+~7CO3-Hw*=71b zdSC0Ii((15e4z{65|aZ7vF)_Qt@gRCm1LNY@nVD9F-UQ10 zVjby)NWjXBrT$v6@ad9xiaF%^V}ab*&jKGIoqLHh(a{Be!S=ppkKfM?tgxq;@A}31 zNe@Gx^dc?G!^_Q_^mHeROEmNK#hwlafz{2^gTE`l1k*!_f4h4*&CbHJa?y+395*HC zcoi&hUypjGQM6(19;sTOe49?R<+6$$cM#oh6BK0cOdsxF=1hXJ`gSI0+=I0Ap@pp4 zV{%|g0}tmZA%{C)$QObasqYqcM@ z#3mICs%fkw<{1rUEH8HPMP6Xat?~leo@q^?Bg5^tT@WdE2>Fflue}4TA&i(Ff80$= zq$P3sWttzLx10_y9L7QTvpkgHd|q4)ls9VSO?_pK)al}awqHd==UWpvKla!ITHXq< zwxwk8o9+8JzRSa=Q~q^>YaF^-PW>KPm@Zq-=Og zrtC8|C^}|OX16_fYp3i{YiSJp+!Ifp`f^*C)#n8SP+&_P^+j9w=$AHGXI!0F(I_Nu zj%lJAmi-&S&(3+$SPRL|KdvwGKc<3T%Ns+AF}%}=MnYmwa<>?y9Ut#izc#;8dmol{ z;r05@$ChNsj{0qUbP1+7=1DxSQK*ehHy^f{<}>cMN$M>wwbq-?(@3OLOuDsQ`kyE8 zxnDh8gCBl$th{!4FcCCE@bEqiho$hoEbSupkxDseG9tbLH(UhPOY2bfOmk1Bdcuc-HEKFEAnUOoS{_$E)FvpUtC=9sL@-Ls4d_7f_m56%om zRzR`He5MZv6%ALbIsJ>9+z?jf=bKIqUf_nC&Xx2Fk!zU*Lv!=aZ;K@#);`C*Y}CZ8 z)n&HIzq0*lbZPk0L$H!zo8~Mxyg2G{BRpom0Zup_uYFmUbLAiHm!c;SAnjG!X*0hT zjM#jb0-{ocDi+LJ#bHK$_Jz2^eIdEB-KItY49t=X4 zpWkwMe?}!CMaO~?fZVXZRj%8oML^dB>l)7uGq% z(sV*?e(Lh(^;|L&1bTT>+h*v?!aYW|-C7 zEtJtU=-7V>l8zkj1@ioQ2NMGX?PP@{v{nG~K|pH`DfLCKABvmC^e(ql;D2UTj9Pgc&bAKIMc@CnpseY67& z3g5jZIrX1=-!ZOx`1>8Bz)?;t(6Uj%F~MjMJWXf%&=`BOiQ+Ag@#`^{-iZ13I~)&@ zj9XkKKwDMNF?n0nXu1}yh1!6KJYsXIX=I~Frqk6f?%W{S-RI7Xz0Qj5z|jbhTIV`& z9oXyIW#Daklq9Q%Q|EDSvzEQX;QIVo<|!le4 z+oT*Uyzh%SlwFGQ*PuyV=8ie1csJoM@qU+NV!zRkLggbYh~SPuhm>a5oicDP)q4S~ ze=jkW8t!O}1Y>jB=TpO#3BlvucrczGlg0vfY3n;^-A3iWFnGHii(@VP*k%*<(GhiM zZHFuX*d4VZE9B+z`58DkXo~{GgYnA9fGGd=TcfU(t(zU7FI9>g< zdXG2!UWBxkd-)rlTb=}QO9O|d{o=FQLw_J!|JqMiNEKP)e2|vHuA%k2E7pf@F-Oj2tW`>K2t_r=J zwFGpL(9Ix3!&rBjDJS&zwxJ{}lhoQfCpB(IOnlmo*y#Pq@N^zlGgOsIR07|7nkN_d z-+sPd|GUlekwh~waYL>UNAgy>a}-6YpVoZiQ_aVAs&;j}vde#WyZbA`Tj0Z>0u>HY zLNFONo6qhj>TpxSG>WJ~rYP9EMB{3y+r2DFcdm+3mNwNOz9dJwdr#SC5zaZK$43qzg%gyooz?PycLA> zv>n7+Y0hifG;iYI;LNp;Rm|3WaD;a}O3)-;lW>`MLy2cs8--f4$BuWpl;M-Jl;fjQ z-I+E|@liKou29wFwY*8ukpS6$JBJ2|zO_9yu;kdG&AOp<>&c0V*|_G6RthC}18 zYDO%lJCfg3KY(epHbS#zZ^Q`_PNEFu%g2@Qa6!HGXRF*-z70jXl`cW-_FphvYn0c>j^xh1T*>l z7qzG`A)<3eEZd_vZl3aU{e00b9l5U3oLb>RW{F|UTVV1Z^K(7jW1rNjQHi#HK^J3d<_UwtqAZ*!56}-dJ#hLXVF0wtCpbgUhtA2cxT*X;$N}iTNMX- zFKqOIy19PXDnBKyx>%iow(5xUNTK8XP_u6r7W6#!lV8FWagT*68jrX2t@c^8o!oXe zb?NFym)!!pZoW}!CZeQ9UQLW#XhV z@4zgH|J>zecglPTNJitjF+atlcq#C(JDKAUuS;Yd0hZ&9ZWs=;ly%{Rsj~mgPS|;^ zn;W&*nYH_Eb`Iu?pT`MrNS1a8z+p8h49m-ce>+L|2q7gpNHpyxySwgye27F5j<+SbbW8kLdoSS`+3x5APmBAjrGtd^a1xNCEajD8d#@uk1t{*3Gct zYRJ_{Rw@~e{oM)wt|Em#aAJqOcx29N9-|jBoFA?4O>RF9klEv%bDQafW$|f$&5?~w z85t3n{HlzGG(B^c7Y{y@n_jk6B%0h9%*D)gh#2h2vI*qEET&;W;Kb5eV_et>Uj)Ik z*LQl3a!lwrdwyby_!MYXQjU2!{RFL_cPNy^M{gq5@IaUwXS}JKSL1}CQO@Iw^o;ac z%?eYas$B%kDM@y=I=L8pmdtQAxI}n=8&wL#;yi&1$8RX&q`z-A%_nO&-3!ZYGH$Y4 zeC0g7=KN|$9uZ!{ny3P9K} z_B+FFlRj=5tJUEV^#jkT-&)CVT*~MW#B|IZrKz7f@(ec!NlhBv`F`%i=(r5yp5oFl z*fy@+#ve(lUSGB?C9fxd?`P@YW_{g3F>A;MpMbe+j!Vy)0ou43VmM(29q@d83Er>O z-F$n+F`uwVP-#$b|AB?~KFm#W5w+%ws%Kn|->UgTyB_RRbVHY(VJGu*ED`kkbORGg zT8+_h5ZUU|qWnw%i2;mRY)}sK=H=HN-@ZMqrl6U;wJwP})iB2VTaAnSaEs6}FkxX* z=J=77I%uKk$we>g<+E^RhI8B*7*Bq>C+0Jm&}}3RH90rRdS~immluuYXcUR!Jh7xr zCH;jOjM}+~O32zHK-kmFlSLIf6D3@pK9u|G$ z>>l`Zk$Y>(7MI`qu%kKh6-vpAK-Dww4}u$iu|}{t?9aFTulun@VKpeu z*d_JcOfnX>QBk4X4#PsOaFZ|i84aMC&7gXqM#EuK4FY^p_RxELRwyE~UDCjV3mCR! z)f~DUe`-9J0n!#G{%({hg=}M(Wg|vsUitFgk@W%zaz{r+VKp3Rl$!QFgLFwb(8c2& zei>^NL(N*E3v)cBIhS_d|A&b1sYhc9~4qDi9v0n+g;>%FR6SicP&Vu8veBU=En2r7T{0o;k6 z1DBWB2CY-i@ZFK-vXfO=!96`fvBYfcr$#E!^w--E4muqMTZYbto{aw5BHel0xr+)7 zEePj-fXV{yg#(4^0BhWM=ru^HO2g*P4{!f%vh8--ijJBYBo`#)z9O6kx4So7U}8(x~Dpf zjn7x6YjWj@2b@la0u`dn?;NRNR_pPKHbfoRTrwrro^GYY(tq&${tfTie6=yI8&34q zsw9$$t=^p*Idqgci(L*gS4=}{f`id?OEe5lVUAHl1n>*ns3OlX?_azpJs*}0n(8`I z#5w+X&4xD9G1hjkJ^!8YuTfPPAAQxc)6ieC*}w>r}_3W7_JT@)_963b~%-gFgA~`W5h#$zzK0 ziO6H;sHYW#I;RZ__k!r^GgTi++YLU!z1*xgP}ZwlMd!q`b{(5d&ECaF#aTx;tik*t z=q^7He9{Zaw<7I7FMwBhU6R|{E>gCX@6LQ@OHU>u`oX4T@=bx)gNdB7Yq{a&0I3|F z{Mgf{cl!Z&K8&~J`^G=H=&!3HG6 zPqKA>eTld37O}udOOqVRt@Nd1um*|7)bOV|CQ&?g{GzSP{a6*rr$R+s)RO52)WfK& zSDQs=*sv$-Vty?c7+4RePmJqpa}BkONqEV=NmyW~HBvyo9()2tO@*p& z^%kmYZMwqpt<=d>?oBgDPTO0@f-TRRmnJxm0)~5xff;w)ETHNUcVU;zf()fne~yn#;m9k8D53kCn`mQ7TGZZ*XjS zPlx`l;O+9TW!(xQnrc#E?eHL49i%{27j1~H|LDcKN%_$Ch=K%A^PP8pe`B=Ey|UZM zf$^Nv*=2r%?`|(je}TF(M0#v6q&slCfQXD?p z&O=PqKH?}S?R~RIH>wPJH*4CP-K_I2B`M~Zw$6SQ2BgtQT}dLjNO0GcP(Yoq?wOE;1IYLigLZ{aSk76( z1aVh0#)m+fkBonGIG@!9prTlyy-4|)SlXJOAobQu4k4y)gY-pWOE__8OxAOoT6fV{ zjzV8Wl01unSV3|=QA6;vqdO)~(uljZk=hJ2$Ad61Sy&F*wKL&);^s|V?_BIV;5V8b zl$nYWHY8uX{tO#@uPC4vKT5}diqzpxe#9>BV18R01)R7)BC{7-J3+y+kR^K7r)GRg zG(Sr1{R&oSp!Vz_z~XxpCauBywTJdq*5n{~IvlAuBhV{wB-3SR{z>@qQcE>WAw#&Q zy_10un8lVpX{@oCUr=+K(C0tm#)JUcwR5i$kJDCmCPHp~?juJIld1B2L#LMOaryQ^ zbj6LEN%riqib(qYov!6}5g$waR5Tatx{*DhWU#NU5x1L58AHM%{xlexBmHeO1mEP0rI zkey&OgN0_h-b-!pzm?yrWA2Y%=di&f~bAz;4s#85RdESv6t$jKZS;+z+T5r}aZ`O5B zn##a_4n*oP_3XG|Y+$ml%dcT2hq_f3+*S_!Zm$z8mKi%o8|FH;>F(I!+U8k!A#kg| zz_Cks6U=rJ1cb`-=gwi~k_?zdr2%1PTF(Om&yH@BVd8f%nn^&^t34H)$!gcMt8XLv zaHR1ElqZPgA@i|)f)v-!Fu{4tS$zu=B*r!!mottf!Z!LMd`Dus$3GoTMHm(wOAW^w zzNMV#(X_8c#L7@<(@$9Aw({Z+zYfU~H)qM|_L$+@pDiFxWo!#Z%-h@X;yAVp<%(4K z#;6lD%T)u?p_k+A@EqR}p)klL2$^voi)Tx+l9OnA*w3f4DXv4~L}zG=wBsS(mvk$| zgfU5pd`wWz4xH8{owM(RZu0%`tW$v-93c>+pKBmy!E>JcX`0 zLC8<90#DKFc})j{o!NP_*g>3XqL^*DYnCpOr&XuXgwd14UaxT-Dk)A@xTp%vF_afIMxca+INsIJLWc=b4TgIRw11M zl4d1eq^%4$GEO(N-fHEgt|9QGOX``%@{=>P-s5Pn@G#)C$PnM!D9yDToRwJ=+NSw9 zzM_*#^dOQ?|9;ByF#WH-Z{kCz_k>$@3UIH9RX&N0aI*WpG#z~Ul!d=rpPDmlZ?gR4 zqdLEPv7g|NGu0&aG(8!}+2}{81z%*H{g5A0F=bdSNIrPbM;gc>5ALQ=Lv{(&lby%3 z!@-KHswsx1cNVa>7rcf7fkEwRKdCrPTVH_*If&RY1CV??Y<#MBpGE!G&!`O|q0X}` zMLRMhCqDgQPbv4*+hN4&Zm-1b?d3Lib%Jl#ZoA-MGMT4n*2Q*g>yN3Y+d80Of!b`@ zCKsdcM}Z9iy0nR<&EE%IyF}ULxODWEh~l**GrGAbg>IPQI9))Q(2;P;n9lU25ytB@ zbSzet%;((=DmYbhQzn5{6h&n#pZMOol=I5#p}(_IWM3K9Apa4wruNboI=#BeUC4m2 zlcPH!fKqSM>FXNWH9Ps)vo30hY9&0TJj;2%J0_0LF84S&DAwU2rnm`+0tKEwE*H+H zh-@}pl4;dC*3R3@OfJy3n)$}#g7+ID2jeL;&cE`N_2A*3S|B?NK2usqo;|z9bW`tO z3mh|tjySKI?b2A`dB2%>7Tyc5{*?P?D4Yt7P{V3`)ZP;(MfY4wY?)v^){^fQOLt|# zspy;AHIijW_!1fwE7OFW*&JbhVBXJy-=}iESA8g2G2hTEX4ZBVhuC%Ujhh`-#LrYn z+&vwN;e{n70nV^~Y@f0Y?yVuDyy!+lU*u=ik0k&fpG$I*z3d1kvYC`iNqc@Mt7xeBU zsDe@z#TaS`?z7}~w};vKrt&PDB85iPr%LRv^WowQ&=j^yWQO`PNqR`or0^Xu*J$Oq{qM3BC& zo^I19IU>;_9VXFaFEjG@>G1TI57O6MZ$1%28i2eyde$TI9J(LcsSQmfiD~Ns*WK`&PWPmDiO-2%d z6ae-=d$t`)_9ex1Pp;sW7mkv+QE!-tG&VvhP+|h*qdmk;@|&sFSnJgn6jB5{oA@;V zk)YNDLAPQkI*#9%JA`P7^F;?ncNPBheGy>0_q@fA!zwH1#MkaaB?iwY3{FutN-G(o z9HJ$=Pw6~<&Tex4JMjYe*hyBDoML1>>3a0v(zk-fU=i>*;yaJvL*yTX^JB~KhQaDj z_ggRCbbn7PaGwsZ{&2HD<64VY8r8`^Nsz>VF~v1eu|7^*6!FTcTc<>m{9+uUcHO#B zehwc|r0kR8ZP7ksEwUJs{aRnw=o^3h0{BI#yk@c411hV|aNvjUFH;WQp{0QoZf9DI zEAv%maEy%k25&WJ`o<`)InI=Kr3xR^*L zcAr?~wWS$nedS!)qtpLRGVv?griV3RYXZ~&o@5Hz)Sl)hLc1oxbt=w4BZaFR(^c!z zB3!a-mXkDi{^7TP3Sjz{8!@Qy=zvbG%*zdcrtvQ>w8I{BBouc&*G-E3z)vJ`;mQOl zS4>Ku5~O3!s25E@h0?HYRMdlttX2`h&q38XPA&*T{Xp&yBJo$69g`(WygEuP>!Dt2 zMz}VP8RdmFMF5=Qg4oX+!V{m-$Du0n3H)ISf4f^sNA5ncML!siSl#S@p!j-3`vZF} zn9V@bpHK)x7=P7`7)vzGjrE%p;$CdzF3%u$(je`dmi7K%DE(6jzN&q=~1);gmMa zkH2RDc#u}dfRK3Vs8#ih=rK zJS2IRgSjevisT^|Pc@p8m$-SkGtu=r?y&5F&i9OH2%&(6+(mdkP8Wt!+l5)5L|#j5jB0&MygE;hJ(q40&0WAgY=k)Nv5 z@q8Jcf!a-~vbu;H=wSCWHb~v|UGioHK}?h-eV-KCz~Q==F1K2_y9qM;erlEIb+CVn zlU{U_QUSbs`1p|`cu=BCOVo&oLdgXZc{yK#8zL>JX1w?HB~sM)nI}v$PuKXagMM}m z2ME05{$842G0csp=n1%t@QyEpKNj$tJxo)<;IG7c+r-k7XL?xc^<9s#_X}@}ux#ow zz3LYNFCsZDi!>eQ&|okj&{Z7Q$}{UYllQcFh$(jlA0rxD#iHb_k)73bIrwOKMc)vp z%#TLCRydM_kOkB-F%t)g=Jp!|u(bK!Sb=$}WVX@Gh6H;^j@geOhYPQN9u4=n(fut8KS!bWZQC`7K&A=^@IUA+x^uPUxBIO*-<@N4Y>| zi()RU-<2j4ON`sQ>&v7PNhZI8O@#$BoQ4i8R{B{_*?3UeB%^QS2h^F1$_Og}p37xO zc(B6Qqp$Id>tgtMRk|Ca!}i<$JdHN#+Ui<60PhW9bcYquu4r`lu|D&f-QexT&)78d zG-8)fA5d)f>hWnFZa6a|&}`1(e?Df-47vKR{GcF1r^?^3Lsr}bUgdS=Wei#<8_Y*E zvoPrE7T_9L9)kRz4|)*>dGwbvQt}NbtMirm6i4fdZ?kmEg}0xRzwa`?W2t4{S8yv` z>QMr@U^4NMB^kc1{c-!r)v~sPhZ9>v$4(YTU&M;Rhe0^94@$2NQuWgA%OQyz!UQIz z0Nr2lGM$2%R$ek)c~vmoDX59xh!SfJv;7^XUs(7&q#56(9%5)A4>Q5X`yt64qd6Dm zeIhz*$?B(K6EhclKFtU{7XLRXTmiONcy;;;Y=)o{npwKrqZgmfR+1V?VY}wax1_?R zo}2Za*2Qh@JO^VsTVZ+bY&!T&Hk7cDE15YVG(?aMhT4Q0;s~3jPrefKwoy0m<&I<9 zK5MVGfK4A0NOoZrB|BDp>c*ckS$PuHPP*3AV;0O8pk$EBa_Ojp;0Ip~jiwM|FGC!O zdju84+#x-=v|r=v+E1Y)TDC57&H9n07!_HIZ=J$*6N@c5@XmiM4SN_lWYE~a)m9#2 zIKUUo>8am+L5;+y2ASlm+Z~ylFsDoo-zdiilBJL00uXsPc14%gj8+l{cJB+}v{^Ns%hcCJ&${>I zl4AU;I&CVRYodMi)X@B-6cmOVIQHCbWFv`HA$QJ)`3*+%ykG#`xkw|yaIt6Ut0@(v zf;4BW#WNaKiLJS-?<%@^&C8abZ;yl&6g!Nv1n-M8i^))#2^rB1g@*CIE;ct2& z`6^g0%~wesGob7L$W}lmI^#3^S$bYTVA-G*TR~>;%LP++7Pa6q}d6rwhu?)8;>zm*nLUbbY_l_-T3c<8~>aTtE9{~$EkF^^|8X) z{PHt!PXnY)Ohfhprvjow`-bMNY&ql=uab&)c~-fS1~J64I#3i*KT66=g@u~D^4B@F zT*eU1S|AJJ8Qhh@ZAE$$WkSDp`musZ{greRjSq(n9V}@$**7kb31bG?-W#yS%4+#` z6-CF7ay4juv`nZ@D90tT#fU=xS!)N?cI53|H_uL15DB)&4FkiZA=Q@=Rl<2PQi$W{ zHPT_HP&}n)gLmcdbpMB}zY2>p+O~CJ+?^nW7aAZ0hv4o3g1ZHGm%`npa0nJW1b26L zhv4pBxF6>F_gwow&$+G3@2S!I7`?UjcGLo(rAXK9um}`A-gZz^aMK>C6Ol|t<{jlX z{j|9bWQ(A@(wH4+5C-hkgkT~!rpHL1ecuM%PfSI-6cRDzni1H_eGA?u{GV%q$Bc8N zra`Q2mE5az%$sgV{3U9LglcFQc$qaP@Na|}-#f(4{PB>?Hm4`c&ixGsJ)k>0%*6iw zlx9^l$r8(uVt`}Acfz;C35*}8qk;7{b!b0aYQ&{=(`0uMy3rc$gL365u=$|?REv(HZQ^4+XisW zfbTs;1A7d_SnTHwkEQR3IyEpX|{ebT`mPlu?8)=%$vr%5yu3h%4 zh`o5UNl}v@j_w2ZbD`Yn9;j&n)P!FyTGNi3_Oxg#(Z}T$&*X6T`qnuubl84|{FaM| z#0cz@KRPUGbB$&vR6hKG1Ak-E4Zi-1%vqH`UbM~BsgcpttK~4B|BdQ`a)x32Lrg)| z13$1>2Gx;=c{v#}1j1!Qnmp>LJ(y%xbDE2f{>`hfoWyPsCGzBUp`>5CXg*dY^^Af} zpW+!?pwQ6V}*hA+Spf=<7#Bl2mGj4RiP#*L%l_O59l(OK91Bw zW>{Hxq?JZ|ik5`PB^#$&C>ry$D`<2x1lZxLX!;w6;X6iA(XpVYC@k7bU0V6*dqN@) zpb~U9%CtG*34fagsA6Tm;t3;{=uMK}sLJ?5=n~O~^kHYCVuz|GoFNLog}`3fg15{h ziB7=IU{|bjj|C!2GRd~()^L6*J3KmSBk2XF2-`oUcfx{W;n0p>otp8GM@D(LLZ{QH z83M_1)>xDTfm1-I?(t=ky_xrkA5+Kh$9{%!LpHv(yaUgf?NQ+2Y1Tc=7t?d2hx$O4sNy5eQdr`g3GIX)`|8YANFE+vmvn+z;t81U0)qf0;gV!)bFer} z3yMwL^f=F1O-c7@Gh`jc4wky|ZP8I$dobyfuK{c-2+Ln?CA_xzIo8hfL!f0HgP@sx zg&;uy?z^Mh-+D-L>Y-_Gog;F;WB#>^Y_jiTz>@8)yzfa4<0Q`^>e%VK4#|>Sw`dTp zCfY2aq?JZsBR(3NqK6iTKZR};ZU7F#a)$Xbyk;xt*_)_M#ecs$M{`+fx3fL?0|U4W z&qC|wXf?N-R)4Gfkupi!&q=IHxNP0suw{dgl6t@X{eIGU**k>a09|1;@pEsEepRfQ zz=6E5d}{xXRIlcvXkH|QR$NunK*h!EVQYQ#RpP)f30f-JSDIiX3Aai0prKt7u6FPX*F4yN<&{7 zV|F;28r7hGrNJdOefXAgw*OC4B;bIFFv9);7IomZ-KM3sp|B#f*w#Y@z>A=X5ZU7D zLBTqrP1|OVQF=xX@u7!rK^tUP5CXJjoGaDTSNjxy*g5Y|18d7kv=RVz@Q5i>(AMlH6BR;cI81qQG?4~vqK<&OD;J}q%!!Ze< zXm{N*e``aAk~`F7YUwW`DE>vXg_wTgSj>mZ5|gx8kirA_VL>`3&f6IJDWT%6($_Kb z{Q?XW_aM%WOstES+9etSRaKHyw8UH)IeTpCvh*Xt=UQ7zYt!a6AuTA}VnYZ>SOz%u zi_6OmU7)E3%p_MCKY|k^(I)uQ3go%`jPMYXv~rigSoTok95^JUwi&)blLd86QmV#U5ZrDGK%uTr^R{m-n==SG4lYoER(3He+c1Zfm)UWWm()fW{eIr;Ka0z*e7T?)++R)eAZN$V^8;{*zZ>^EA6HD1UkV|Yx-0N zD$LyOL4`STNKg_sqtpHSZIk6YVkbRDGAw^NI;X?G>541$i_x^o7P^TyqT?cGqp?R&)4mIGnA=83Mi2Q%$OZY5$xpFKV(@XKdvk_KLwHfhRZ@f3jKaRxGr5AByp73L_4>K=^rqU)a zY!PkrGF_KFp7W*ZGK?t3v`1-jnV)K{|4t)bKsLuO*8dg8&_ zIjLj;k@lBVfNQ%7(uE79c&(O-njg;bA>oXZLwx#^TQ99N1g+G^`y3CUKZ2VYV&sx( z!i|Sy=%Pnd@HA99tRPSa+Odt|mrOAE;T7sNdy?nmE_k*q+6m`^2X2dAdjFJ#h9I{k z6^3ocH2vJA*QnhCxg_~$_>CPL&*HvOWwxL720jHqi?83A0=@-f8;8t`sx$=`e7V4K zpN*SromRwbflw32p@>QAF-hB~W>h3-WIlGK8A(E;0UZjS1O$tc!-9Ga^p0GjT->V@ zs%AXI3xJvm@1n$r_BdCbQk^o`YG%}G($C8}4uy@Zr{Slne&IM$UnvrRLcpNwN{hn$ zaILn2QEj~L5%{)T%i8^=Rm6o6RJitK?tr^>tH`QWEM+qBIngK~pOXc;4_OC$+ z6>d?H>@&(CCHzV3%6e8z_~d?vm34`pTDNTE(SvCo0G?Ktl!{bG%eUmf_};ycXj*fr zotkf|MV@-5DqePfv#`Kj^zzc8NX@iHv#RUg8W)nF#rBMI;P2R~ShHFFDqd@4;9P_^ zpXvdkg1F#A7o?L+V1}i5QYdu9Rzft3yrYnB0e~yooHS>NVd`o{RXI3!Q`yig2_CS` zylboLS}9uxSfC%CXJcCsM z^-t2=``pXlzhdWJ3TKqnJo_W|h?2oYWk~apkn%SX;_7FV!y~TMZNcsuimkcm68m&u z6nPMZbMzA|V7G(Pc>dd&OsQ>u!Y{Llcjfs_+?5t?(w5=d1!&ji;`XdomZIqii z3o|)WBPOz+i^dfhgSB()3*&duw9CWJiuxEY&Ic}LziV*9o+5X=bhJ#%kT3eoc<)By zTNc*yZf`^QWrZkPwGPS zP@~AvsT>MSUd}jEOcSnF5nVW>)I4HnT?L)>niuaZ5?kak&F}9fmn@9j9&mZ?!eimM<;pLO)rm-XJUg9N-%~AkE0B;u!uBA7iCXYFe3A2$ z&&_$}QO4qcMAwLpn-UzuB2#3O{--4*M?C^Zg7`OX6N1Ge8Ua@CQgx)?ORpa5Xs}r2 zF_uau*DOCNw~NGGa;us1rYO|4=R~YDnL);>8)~@}*sf&y{RKfHlzHAO``IG~64@W0 z#YcX=TmP7Xg%T?#=t09rGLP*PwKsf3f+wbLu0A+rM4`hNBCgLw;T#-hOK$belTNr| z0fSy(3z=Cqy{njpvQhD|B0R}hObsI6&bkmTr@#V)0sQ2w&b9Kw@)xS1w}=s26%qf5 zmDkQ4#gK#K`XO~!DNm1Jx7`?9>A=oVGwEu(naW&JY+&cwiGNBF7v|pGgyKOnvSMR= zirxzFi7aChYl*xF{!&*RTivJr<%8`B%SiXn&!0uSUukR|W=O{9p+_aFn*3isyY%@Gm?j!{fldn$mo`?yk=DAcU*&`ZPe%PC#jP%F z$$Qt-Ia^$Z0zZS=&E}w7ngQtjJ~cdc5|X9ANv-XXnAjUBY-~0@cgHVha!fu)AcG1| zv~#z#6B_9cO4YxHo6%Mup|e3;lNkHBvQwaCjCHRKH6YmYi+TDtK@y<;`(#x~{+^Xv zS?}HAmske$)%gK5s&lVnvW7evTN3`KvY;c5`h4;CY7;wlk(F7A76OiWNr*JWW&s6t zEBKyI5`7u*HY1?`;|%5qaGpvC)6G)!(kLFBsL}RdjC@O;%s%o@x3iebLI`;`$E1o=4_7S{ag+MV!*D>GA$#umbBOHG_fe;1h(yzz!2Z&;pqxms zphfX`!fljaB}}^RL6twaUZRDg%d+;%{Xgl68&Rl)xW#m|qw)YdWAS@W=|2fWj?rDY zpHQfsYAJd?-d{ySEms%gEm?kbOKn9bxRAO}S3PExlDkBf4C(OL>sUt8##j_Sm!Bn- zgAR}<6L#RGO|zMXpa4AOlLZ=+2-!7d9M%yJL7ECtX)6vF;Eb1}vMd9S@}D3SGKChb zIMwd(IwS<^y7TyWmJ4k(DA}lPcDQyytv6|MNUAy2u6{|SRYtfdR&6dG9JfcEY+mIJ zpy#WX+CZE2ryxm-hM$DpgTLlUtI3s%Gx1HMDu6AcixN3Wxl`0R(>Jg#Z}=I> z!X)29EZK*15TWNZ0tm>WnF*Q4xB zm?^y`!oy~2tmJ%RHARC!zKK-3$$OqvN$(Ro3VmtVE^|??Myw*xD3$DZoYfn zRCm-JD2SUnU72mybaa4X3m`9~`7S*wZO?kV39VkI7q}PkR&4SaH7kxF?fch=-A`d?_;XAHQaFPii;R^< z*|s!L21$x&t|mOJpXvbTR3w_PutzBCg@3p6QQl zpLlIJb-q3-Hm@2k`24@o1|2_K6a{ZHSK2Jc)N@=1s!wJ}2T59w|BkMaVc!OKJ}~>4 zH(vgK!3OWR_iLsI7fyxS9oKO}kiHJ!{`>zZ;Q^?K&;uyn?04-F#6kzM9#aEfE%viT z{>I06?DbcsepH|@L(WTK`A8l=-#_EO&4OZ~aABl9^<_U3UH7Uy~Zo{T%p9;_R zQD@_4K+JxrQj;vQ^pDy_4Q9F6T~Ws$l1t_QPX%xW9CA_0AVWSTP8&nY(MyE5CRqL? zp-6q~U_BOq`bZdMoK9n%+313lbD zF*8^P{h;QUA~UZYH4RXMrX=Yd4iE%rq2yW&!0yG|q}yDsh6xwO_?wSOy8skb6(tXy zs$tgGv94^eT0c8HW^qJ2gmumOnZN_02VnJa zEeLHup0+@j`lhLuU=Z%9C#RQg&dG>57u}18j`>o(g zg~}r%FfIMyWRSF;XF2ghj%ld6_-r}VU$^gdez?9gqQui*BnO`>+ab==1wsM49 zw(|4KQOmHdA3a7+sBm1Fe2)A!d4W^w=&173p&YG7;mHv${26LjHpy-fE2D|fU_6QO zcR3%4IsFp%5u5B=wyv>b{UB_d$%o_aUm2jtLf)sz$K67WsyaNnftm)9aL5z}i)tD) ztPH6hS!%A8u!lct4SU%Fl-F__K-pjR8=+Z2zbP-lnHLwjly&3hX8I8^omx-mJT4`& zRf3>zo~EuxXdFKcw{sle@{+lhq@SeUHh%rhb9}U7iy4fR9>6%|q zdnQYycH9knc{)uJoc{G(n^DGubDx)`9NT?2w{ccvAQcBDs6I-fNy zScI49z7CULF8ghQSK6)39m>&A5iv+p?%yz;eK)&|0O(xH2l{>%ov5f*qoaZ9JF1L! z2>CXAeLt5+#SvKkYcczHNli71B;@7{HJz@pHOMJcUpg(3K2Fy~S+4#hKG!dq|Mjtm zK`IWTdGJ;vA=$YvjXwSKtgjlP#`wM+@tRzH)u}6tltx=h+gu_Sd&d7=`tnCmwBfZ- zfzQ9GMhd<0W^-7lU>T0}*Y5~6J+wOkM~|bk;Ndj+B+g+IC&NCWM0m;K|D7iBNI)r@ zUK2>ALZG%=vT1tpkIPI8@$PY7@sXw$%WxMqgz-fW$twFKpM}Ggt-7O!8L*lDNkE<$Y1vD*u580wj~FD);DcQ2g!~q=5sngKLtm-8b0;=CnF? zWtSJRGea2Ak}!Q7+FSv1+ge}Rfuo-vkzw~jJ~dp6!zk-#S0>^0q}4dS4NPJZi|}3t z-;7C+;81LM6d8Z^=-l! zzG_%#RSiJbDq6q8L%ca;qn0l<)IaB7_wJHYYE{)HV8QU%5T}?|4ED^#BEs;xJA>ua zQ{aN>4-Vmy(IFlsv93eD8UmV`;`x)7AdFeEV|zK3U=P>Ek#Js%^w~9_j19ZMjF{WdZ>_DzvPV>tHs7F+&5_A`D(qtLWI7db`wNO zHpQ2%5HYV1>Y|)0S^LEi+{luEKr=Z)lnFx7ne>2ZkNg zZ|d|0itD`Ci5Mt#%Jc4`fYABqu}n8-nI@4l+Kyr*H*^IcWnx?|rSoJ#P``)2Gg^YG z`qwRFJHv_4-i3w`zJbW1;ibdcCmYoj3d4e*o$$wGu&@f>Sakq-o;XEwe6KFwN3Tk( z&Q0r-9CBlVNN*)u8ny$uDqS>CcR}L39fXW#rfr6JB82vaU0q1{)oR*x$t`!cl(y#q zgh%)^s6VFl5x{TA^+wBEMH^NpBaD`6XwEJ6ze%OlB)~*ibtbhY-6T#)@IXP8`h8p&Z(Wck;#Sak~Y3I;2dIPxQ zciAO&u0q}!R_SJ2=)>d_dV&=4b6DVRclNyV`S@b_Uf*v>krwx~8T(ED%W(&;T+`cI zzJh^BhnmsMW(NQ^-`N{;{F-J9QlQ{5S(KSFv@pdn`?cLhx*ckA{)jRKCVhy z`J}XO)ou8bzDUq1`NnW8ao^;>O?B?o%Ik7g8;WlKmlTmn3>R&PXK5^ALJNn{rzD%9 zjYEO+Z>TmtE)79|T3$L&G+SPP*M7iY=rvql4w_ARLkk#s_YE!$r0=|OfdbA1Yf7IA zx6tgGVj6{IVrB@4J{gHiLVBpVmX#>7pi5+VNhs~XqCtz+RJz!<^0sD3(_H3U-1r(Jr<)=0`Netw;f+l&4) zu#@c_GTC9KHOXI&McC}sC$0OfnZ=Mqm;qBc87kgdoK!Q?oOLGviRET5Ua}kNCpCv_ zXf+%qW{33N4zG@_aQZEvCry*SF}@NwE=fP+wtGO zZ@OxxeeH;4cxgyn&xfRH7wlx8OC?pqu0k#(6B2OuK8i~x&Yaq)ya4}!SLy=)#O7J-|t@1dG1j-P5Gys5oxEIfjkEaJx^=0CK< z)7G4OU~=(v1Ihw1X!zCGK)k)sCTxUjc0Z z2{V})+d*9jEA##v`C~ic4)rSuc@wSe0GCe4tnnA3{jB=}>X5IG?is6X1^yUsy#?{d zgunhNLdh>T7uUncl)_uwW4+W>H$7$6lX7C$9Rv4e2r!*c4fEV2S2iA{bUt26FEyvk zlIih$H4V^ictPOp-?o9|O_-5LmBYX5DD-?z_)vRPuS9=0N~1XGxLV$;lJ-_b1c0j~ z1~k|+m$+Y*R`M>U99Tn@xEDH_$mt2IncwiKMyvv?YZgsy9tAD!Uu#yOR)8!hQ>-vO zS-qKx2Dxk7fFaqS5u;nqwNh5!_~I&ZO)_={;sHDWd{5Vf@iH?!RbRk6uhCUHwuly| z*lal))T@`v~Z{YtkuE?Z^UaJCU_CIo7itTU2MsyBh3lUm#oTbHbV$<< zAPnd=Yaat=;^N`>O!@^_&Vo-}k%(9UJZ zivKRc_J;vP5pE_e;9zR+-x1yxpZrH=1q{s~wSg#szwL-Nu@>OHRMgKq89wNrgBFec z4dF9}N%ChYF;^Q`364({jD!(?dL=~}$k++b!0dtiV@nzSE%|Mz)UJ@JsskaIN8N5SO}BdnDTez_t|NNbS#{pzG|RH=^t z`7x!+*c&cF)-jUun*alP3DIni}h`nX5y34JL)ow{#5~PL1}OZ)6v$RB!yh4%t~4fj+O#| zt+h^DVxWrXOhid^p*FpaDQxhtsA-&{#~kK0=#3)UL(2dmv;y$?jF00N9KkWO(r5H# z+`XN3`HWnqjMP?NbCUKH;WeA*pz8eMhLsu`O6e3B)0tN!J->6Qo4Ag~xsA%jV}#q* zy@2Wts2}M&32QuDf5vVY0i~MtSfSuu?w4<0_!o>H_8U7del6R)VRpR0Yb**pPrlC@ zv7+xr}ZCCme6BpWm|7p8mPC;eVA8@6s=b6yL|*>>(HzFpp22;TkMQRLw| zVoDh~PhMPia9;zJuFwX)n_~*?)0*&07uWw4V}$Cs8owAa?kwAh)A8)26L_CM_gSR! z6;%AMr}zomu+6{QR~SR;mflvDxkM^ckq~j^uN2QmAc13$oa8IVT@#_`V%71_&o#{X zQk!$whJoM-i=F~xL-(#(Oc>}%Dq35Q=UNfNpqOEf?ud_@GSc+KI+y`W=OMpUXrcew zq6OHBlggrn`Tn~%!pB|vt@XdfhXRIZL*#pUSnb!nx0dUZ42i!4W7nGQy4;V6TDIW~ zJV@bG!}K;!FMoolImg}OIWC2Z$fzI$Y@Z%c!$re~j|TU*)X(^WKkdfEFYyxeGB3YL z)dA?lEq8aM5Hmgcs4QpxL>;QVU^xafw#`_@$$;G8`m89VV2$7c%*FFO3a3;Em6NUM zi9tub`Awu@@iGYM{I4|q&{`4h0tX0vCSdpN;NMj2XnAL~dW|n}da%v3PN-#}_LINR zkWGG8LJcn^%waWYQ5cNuOnnS7|00ˋN!x)IZWl)(BvN;qGADBqZv_7ZEBYIKrk z1M{|)(U5HlwRgxli7!QHN7|NM0R!mNj#F-f9zR7M3%;iq zJUesvB3Y?hiOxXM->hD%e9+CceXkzO6V+n&e!*{}ZU^&A9;04(BKZ`Q9>!T?6!glS ztw-~v7*a&B(?xuK5_w`fnjd`#3W#d^+14X1CHyO)LqxbX2M;+*|KLunQb;z@(7vD& zGFv2XCbdVwOo*QXbD)_XBq%GyNtIY3wqRD@)jTI9mZ4}5Mtm8|Z()`m7(`R}D~3jy z75r%B==v`C2D=mJl>1iw=}RX=;xfko?r})W^PkFdsJ9!cK!8Flj|ix5D9>wLne>R{ zXP)g;M|Z3m`6+kxI?OIl_Jh$w&fCu7*-mJ@(Y?@1nBUni^kG4n*s!XDW5!jy)C(@H zZ0XiY?>F}Qba6%V56wf+?fxuQ!~LPx@4DV}@omNLcK%%De<6NN;n;@yJ>D2}sCWC_ z-6n63dE;AVz(Dqqq)wx!lob@u$h;?}&tyFfdSom8n{Ch~(e_&`V_jKh@Ts%Ekf%18$7+Y0sy3FelQ+kW9!2dLO>xJ2s{ z{Z%Ys({Zrt!Q-9-o88~r0^VBJHIO`NiJ|{G1FI-{HIJqk3D+rI$EkEc2YfkKHOR;! zJ0UlSKbv@gC44eIn?9LINmw-`hA`bRsVv>$JP9gm@dQyB-i*Ez=*MyfeBnM^O{S&@ z%KV0Mt(&#+2aq{aKUket{{Yf$6*9@EsZn*8tlKrmTbb1gRJ=Am(LjWV(sG+8>V)!oUSIKnA3Q3J?lk5L$)XMwh4EWzj)XSLLj#yesb&`fz@GipFCK!g60VWw=PZ)}U{mrhJ~m6k zLy0f=BZ($P`1j2>!-KY*-_4vg+9rVw6e`5d0n}J*WN<^q=R1zr!ew_oqn(1K=%Dyq z@SG&!B_&N78VMuJ4O)Od*Q@yu&Fdg6A*Q`=>4$CcI|IdXlilb=EQ0=l07AWFUp#4Y zjuVxDlDEMmP#_1!DUmQ19lV{Cco2)-GSpj!6;6;JwngU+1Y}9Chaa0nSg`NQrE3gX zv1BJ8)?;^2R-Rxjj!;QDL~p^h4FTz_r$b$C3a^=TA(bLzQHM$9Jkwg93UuQD9BW9@PuiEz>eg#y#n-D+Y0zyrwyO~7qF==-GhNj742msZFf(x zwxKB_O4w{Y^^u5Vrh6P*=DO`giJze8Wwg!q!fe4rRGng4n~0N9B>y=#q_#(8!{~9h^BQw?zsXHEezLl&FaGOx7&VPQ zgCrCUP97VwiyX*D5*Dv29p0Fz=SjtrrgB}ifP|e+#=k3f5zb+gcqm8b?%id*h~#O? zGiE+GIV$+eyp+qIbDQwAFG>W*O)w)G%a*-~7JC+3>~+U zQNX{!W3Fc(|6TGRR9m#X zPoXGvm&TFS$6KIThru9{An8uXbecUN+o#6JAJr$gWpFEiS9k zdvSeY+yYM(&2=y;^+x%8n3D2x$$-m)G;h>02hLI2Q{cfP?JjI#%I=QlH-|^OPTWl1Mc`~l! zH*R`G_dl%9|A4y8z$v+(!shekB9iL^BK0q_H-!FCj=x87@|9{j?FrTBLQB( zFJU*y#pq$w44RU#4o`Sdawdfx1FR-SbtqI!mrq**nPLi?cv03BNb-oZNTP|2C~~m=?k4*iA&mVY=l5CqoC4XQ`2ZeA zXQV>=&l6~Y&anvK*aOC6`NZfzng?2zw&Ihc#rrurMS9mT=oPDnl zt%35#8AC1?U~>4G3^rL`ra*QMqeP`5sBe(v=Rv9b08N%SA4#+AlsYZh+j`QOlo56F zNOUdqttInq95jGjQV@S`i}ew%Vh7r9OoEXdDZ+>SbBSWboKT>NnxFK(3;YvnWwNLR zR$rQlW`knExEKSbRpPpyUyUj9Me?@R(F0}n-kBF%=-{f<(^~KyBJwlvR>WglW3ojc zqY~{>g>?*RF*fOOi7M`O#0S#Rx-7MDN!#|bg^pJ3&fxX>+HTZT#|GA(9l=#sA!qd} zEQ2=|j@%OWwi&2n=`a?@NCG8pzpvBtOMO zD5!5cGU6ht&v&FQ+#4=yj;NEhnK2`XPq5~JJNi?uJ=e?~HeR`CH{_&rZdpzW9ljs7 zkQV0k_al{T%u23oqv(${6!*Kj5;K3Um$_}z2|7uzaEO8^3&RKf%pVrq$sG({+n@O` z%SR`rq?oM7){A0@SIfCK4#|98WZYj}HCkN;A?FjHLqk(A$HYYadPCzUW?#%dj`gN0 z*3j{G1y2O_8)fvATE!?|r2OoZTSc3a^P_>xiF%@?i$imFdP*lKt^dp{g%CM~lwHRB z=R|P4Ol7>{;Dn_j&B2>?To)*$>;2x~eO=FMJlU+k_b7RBP0xLGyW{cfqSfyznK@fy z?=oEv08Lfs;wSD$juWx>%jS<|0@vASwD@d0Rsv?FS|D~U}0*xxm?ti*D~tvTBi7Ok{hshN26TEjwDZe zhwDES9}2nZ??ElFmFGxYYeL&xr1swBHIV3h1q#@1xDvvYou5C9T;bRI^6&QwY}*@< z6{M#SpV3_otRCI3|1BbAJZ+VrmFTys4=bsWY~Zo zUBo-El7V$D&sRfXF6XmpaFthoxlju2rzGiBM0fq(!)A(7W0TNL^;m<~Ec$U5-eC7? z-d00bqwU(i+7}~7_~t=FQsh%@X~d=GXAl!~<`wr$j2^Nr`gq@S!^quJTSo)0mS%26 zyjPt+>onEuB%JB!u@$L_x*1IK$O@r-(>hzPr+V>Rb*EZz#eaMgp3rZehHhQ_F=&lm z7Z;P0hNlCthQcV%Oz@0D}7zRr#` zVu$nl0HlZ5%x&}fO6JAul6BDef4WRpNJv))Y*_SI=gEYQt~&K|3KQ|7s&MvR$s7dF z%(|t@XQsP%=e(DjY{TDy(lI;M+5BMt#^T~N`mFER!?N6zzxM?LB&+`)3QPOzl&||` zvtz0+o2{{H0WBcspOsdOJwR@J!jfhegtZ`uqGBYw7dblAn+?5*+6kJMA9?l-BTD<9 zi`bKh|FHU>gPf#dAvti5eeZ9{; z-9v9&(~7*`andOuONen@A(2pGG2Xd{f=<`yUukiIFPGOFqvsy(?#Ud;JMZ(q<}f!I zbY9lTS+F-x`X1|p?grN;$zKZrFh#pDjw?*u&6sD)mWxk9QrLeMi7S*#$OIfPgyu8Q zsqFFDDSbJ?AInKNn2tSF@5fNSUQim%&XlJVf7@*>kDP(R-o{|^p&E5*nogK#7#e>P z%`cb7j4ztK+SF(Z?%&NLZl*i=ML!&;8e3tv%RSxpmBFbNVA;G|=s0^}uUh{uE5Z9V zymv4lL`@%&LE{{6@hgdVaZx)#0(((0+{9IJgSt5h`sg?JF6XKs*^B4tM((8LtWH1Z zftY=}#jNEpw3AQ&i4sHeoQh5qAo|-0=ze7VCJK%RF#6H z4xh$Xo2fGLP8jA`mN4B~{sS-38b_%=tKH3)owW#_!Otz$)$O$A*^Ms7RoD>{lQ>Dp zP)l$}c+cx2FRhq-T4Cswr3GUsGrIdYvB+J>G`;8p=t9La&t(alvzkOL=l*^L$*&KGP9k=Q8`ZO>55JTv5Q`3Mk|}$H#ROp!w}Zm z-8YO%5zpJ7-PolCU!n{c;>hd0)vjH+AD_xrrU;#{81v4hp2?iJ_wSlUnO970w;hc{ zJqb6N->1Gxd4YBb6k9~A3!bItZBvZ3E8SpI2tRkcuK782UhCwzk+=CxF%{siz8wpu zX0}Y5B0k-cEDAClJNIF#H*P!`d97dYIZW9S5j(rD`#MiTwq7Iwkt4koa_&thce?Xy zZ5--TlKXO@W5&LBi2t=`^NaYuO6P#iC0Z+MQDP4eCu#RXFFHCW^%!KXS2)RE9;xE^ z&?cgt??3ix@j03FWOpm_Kb8?saaOi9EUrPei2t(`fv;@CgUgPD*!6<(jVXhxrSa1u zDsy{e){5=&H>=LXuOkdE6SW!Ucx0!YAlJ%gNb0$Jw3o*^Zdzr>?Sz+|iN}!76YhPJ z{v$1Ob2s$;$ccT+$?fTKzPWX-ri&2hkbu(fsf;AvqEB94q4lJqec7k#{r5h-nhthV z%JHw{7E}xrLzS=h*@$e4hXYm{Z(^Y|lSY@{sEy(Ot-elw+%xYRH$WR_3;$1ZUm4YA z6RnFAthih77B7Vs4^oO1FYZ>{-GV#CtvJQKxD(uqTZ>z9_XN55zH`p~bJjZN{=E5< zl{d5U&O4czy`SgVv-ee#=5yRgS_pSe`a=qvU3m*-b;p8!YR4S){cecR`8a;&k^b|1 zw!M+ZpcG{W4D09v-zVtHGOT)oznXmvD{n(YhF@^XCqY5w$7?xy^EX_4<~m>JRx5rV ze2-XL&5cZ7%6(e;_H7BWnjI0jw3?n1nUnfSRW0^1xqxaQ;iIYny;Rg%uwzIw>ajp` z{J_V$a%I$Hrc)-Yl$3|d-LWg4T!FhTA1n3&C+XK6kr!}-P=pqC{K43FcfjUvOq+ZZ z-X!%~cwLH9As~oanDOBO5W~j}jO5f4c9^^5LBU4U;to=N&01W1r)tNwN_?Gt8orFC zOmIHA9FEYn3J8;_@V^#5WDZEJY#9CFM!->hpU+g-c88cy)DQAfN_Y2Arg z_B!_{D1~3uQm{Y74L_XNS))!OG|y+U6`iaL8nF zR>8=cDWA&?o^+DV4ghKGpTHp&d%Doz}hP$ z%%948IDJPSNrV7K`i-^N_Bx}Dv`!_zE_yC0pQJg_9tpifM@jO8!SQ7lVxvJ3lU!p- zIV<1^sZxPy+^YR6O#C$F(LJtqJU3axGY-qo;7 z@SF5k5*`%3>v3Bdn*4xWfzx_kXYEgifn3kKeKkIXHtkMX!?yvrjKYUQX}JuI%}TT2 zuA>rz1EDo8Caj-d1p56f=IKO62Wt`Z-=}b4`I+b9V&WTVwf#{3Dy%^*sFL34#iSfR zCq}M}PhP@Z68HP$YE!%WYPS#erM3LaXc~PnQVfNJtieT|Xzs%`5)+I zOViV0r7l~^B=0B`qFWVJUh=%B;k8wZUQ|~v@d~CWuYxnY52br}9(t;SS9edit<>v0 zCQVewD4J4&(*=F%mwxX6Hw7$n-f?CKM~x>EC5j*YZ#FtFSFVRop8dgUkrtIYwY(I-EVmBp3HA!}>vYrkj2t6@XX zZT#`WW#&fXo$pQ6558nvgHghlWV@c7!in?EA3o>DVeO8fU?8hTcMY~t5;RggSN}rG)F2Ny0GrN^4GOZybg~Aye#GOxYi#A zNS)7+o5!YB=zc$+qF?qG;BKu1=+=37__E)vyP%&7mn;ciZurSu<@3Jhvw8A9tEsL= z!wd{GTFYOfW`O7Laci2y<3_;_9G3`i>A7?FrB`}}YARnhjWsPcjXBWx1N>HRyFBt4($QdE<-kl1N^+3p5LRK`Z%FufyYLf}Op~Y+KJ1AJ$*az~33P~6I){5pw3 zG88}m>FG9g-3Q0xdC{<0;}xv{XGxvna4|XM4mbNl+~cfOncf%XS)t;w1KNo(4@2sB zj<{7C5(f$b;m7XP%i{&R^5b>f-6%?~vzzTFpJQyNtd$2~6=IpV-A8^>;cYYHa2!NQ zpaT9zVN+qH%%CSU#C;|%zz^GV(8$4h6F1tQGwm!8hR!)(0`k{+AnjIqD+|`=K)54r zfZiFmx{{=3xU{ymSp&<0gy%=s8;OJ!=xR;Kk;GBlcQ{U?2n`wd02 z3Bv`s_p3Sa!=1xNY*1G^34=dQ0*UCoS>ifn=Y>rE@9)K-^m|?n?^SrHQy;-7zpC2u zKO@V|nCQuxD6J{pko1INg!XVxcFrO3kT3Qjc+hKEX=lY0+)o5HilDbYekCtxEruWa zN+J3z-^)8A5PCkQ;K9@X(^tKgD4_C2F5phupqoY2VC^bvJroc$R@clRMxt3>t0*g| z_Ey1zAG=U~Hx#VAX32}IrMfKI@^YztP>FtXO@JZ{*lu8U|F^H26wI-M6SS2-8$#uJ|_KaXjs zW^r1M8r3ek0HU&=MHMu);!>9Vg(N3}&ufH4sU-y~je1z4$$_B_j0icRwsJa0ADAUd7Txp0mCnP;0njPv^WdCi$aZw@$s?0CW( zq(#98c#l=YQ1i=mNHI&f!nd)wugl=NkwqyR$?7tI#sYM_GxE6F`$PoJ1+6pIQydoE z4hAeQO?cGyT3p{#F0Z4Lm|OS2%iqY`WD>kkXRpqL68o8(7p1Mx;vkk%twLvP|$(0_cnYWJupb5%z*y#Moznq19upGs_8WZPwXxQzp5imof}N=JKTF=6>qx zMXZQWD6AH0o0g=0p7}%#TlU)5g>M;ZR`ym&AJ*6Cj+0uVvYZDl2jmW2U_?X!xbyr9 z>t4y6kNsF14SHA-1k!62paFWF2Aiyagk$DC#&&L+*U||0XIpjre)P7g&}!T*XrQl5 z6O$NNJ5q2uu=4JUhs}f#Um^rm2qX|SVwH)NGh`!Gy7 zc^pq^Q=N3#%(k%PKaPlSW6PBuv)U>15CrVNXX?IC0fcOvTJF~sCoUBxkXX}Gpv}Ok>Zj=ZF^mive zLpBnnI2|4ddct4pA&vB4uv}kzdEQYs?}N{1iw_PDF*pOSX`0s$0ueacdoZk&)!JxH zVZ?Hukk<4W!!bX{n>QSm{Csl$baX8=;r?wBAp3~l;L|WyQkOKUiv{5Ej!c8w)$-^L?#H6^t)PXr7+eryvTC=(Cum;_qKSiQ~pNmI|&rc2A^r5c-L`g81Y4`F`}JM^$Qs0bGG*+}eZCswarr6&ok40`cm8rjoUHu#aOf;+EZs`<4`LKSQfDr(ZiZf%pvjJ{AogC&;C2Deub#*(usTHj89pZ zDny<^MdWQ| zhpb;Ur_I@Z%o5kz&RK}@+(8!2$kfvj?RH;F!$$a4j3-feRE%&*Y?z`Is8u$M=m*Q;Lg2d*i7f7e_q)*cx# z+Jt+ds36u;DU6>OO2n$zA_pGx^DrXbXcnaPsP-RecDay$wXvZ0pqcD~Jyl(%v&mYD z!1?Cr^gFY^V23(QCuYUvx-flL2|i0jp#-J{(a7OZs+FJpq@S5c&kS}!P_WVf%GJ-=Vj3jZ2VVaDG?RK~*|AZUgn{m;$1nCqjOzj1$_!QX z^PQ955jT0B7qY2&uuY8GVvEVd5b-q4K_`$t)y7Gc|$xsQ3s;E;zEM~PEF)r%#nxCe^2NalB$QxCF zNYVF!s|_g<5rnm8ltKWo)-US^wgX+V6WB(4zljGuvMwZ-LY}Y;-2-=|Gly>3(Ttz3 zk7}Z(oF)ohoS;QqtQjrlmNa^uypL|#1k=+kr^X3nmT7?!e11gDX}_i|P>?Mxl!Yfw zJ_&y=x~znd9i4_^QgSe8Xz@SE_kY$yZ$%2&Q&y1j648nAbl+x~X z5fl`hqM814Z{kBV%pT}JLYlNxLckE!7E{D%u1SOiDkEXtrV7=o8|?)%8dML-fL(5Q z(>B*}wwCZkbHOb!f}VeWULT)2xuZks$LfnLRGQ&MLj%MO&$U``6iX+wY`(Dd(}M41 zHEMKT`QPu#edcGIvH#;)MX;DcOnj)8+huMT4$CYbm18<^2%rVGQJl+x2Jg8#FH?2? zd^1`Zl;!q&=sLeVYC&V3$N>go53d-m)*yDRdaWXj^`sg2scDiEb?YCWG~26^N;-7b zOI}%7*(9~Pd~H?2u=lCWid{lFUX&f2m9AsB{w>sOOp>|;L2CAagp7qM@lhWo@A$o@ z`>h0m3|Ph;WDNJ?4Szv_;%hR~+M_A(AR{&{Q&rrHfsps}~Vsl;vL z*(1x=xP4S$1IvNfuL@G`+>hrxHI@y0bR)fQ11CGrS)ql@)E1sFM(ABGh|4kCu+o5T zvRfTnb2Y)Dpd#CG_ck?r;R0?;Q6ty6MUVvAqa5ThI|?E>eK&iL zgiSRv7+W|B^t%r}XXyc9Yl4s+4W~UYpPwdqesXD{UYe3cPuWRuvcqK3r~f2C4rqco zxk^7fMxZx>QDqA$jD2xU#QFe7Ztten;Hb_M=4jbV$X z_p(pOvxSFH9)Zg{i^8jLn(r8fTaqGRI5vABh^^Np|k6E3g}ItUH3C z=*99Ez5BmvwX2`69EC2qCo@|v`+Xpr;1BQLTj_^c$Mt5v`~7rc=r-~@O}CJ0(oN)a zmQfVY&}5(ty1n4x&m*RaL}<48(w4n-b&wgVMe20uVu>;2?oEs1@8gNg8@+2~jI)8nQI*`m2Y`sD>mTi$$)cMCq6EgpeYqZ8&Jyek z=tRR?v5r%r4Bv?yk(zC6(IMtqhVGPiRX@-9Ua`0ztGa!X#~Li;{A z6WwFuV)XXReW151f{ig-H0k7vpI>*;>PPu;;8{MDQD;$2i>_cY6X~tHJYtNRJf3uwiJdI)ma&DcninTj zv(T5tJH`}2ZCimzMCcz-Ya$^wayR`h`DY2PbsTnIl=H@8-Zj;k+Dz`dK04^d5GI16 zmd|v6Dn|DRwF0y3as71R4f_cA#tY}@rE-S-fSM+LVu5B=iB0=f&-x_J#nIY`a>1^l zT-}KX;nI516_Ea8!{$`UvOC1He>eoh5S1Ma6f^L8&mTXZvERM+X*lS=rq}ooMNV<|K<^ zKh1Rme76L?d%-`p9*_-D?PY`2P2P^v0>bs;XvW4o=_MpI7)}#BImih8K(kh@p2((E zneSeG26CDT^FcHMORqs4Kl)qbD7<^29EplYu3O)#C@lh*tE}WMo!?Mf=-6p?z|_j5 z1}}wbt`)U^>q@aFY^K|Q_ZRnZ=^9m6_Qh7dS1Y@Qyd3@)GDLrBCk+(OTGt2?7X6m9 z3+89xII~Zs&Y@$ox0Fc#;s7&Fq_yW@7xCK&d(~VIQVw!zEP_oaW3`q`ggXn!LC2vr z+jY&sK~(s(lhlJ^h$$`!UiF1DKw$%b0d9Ne9IeLe@T^F2TeBG}V70QoIf$?(+_BVh z1+WihW7SpxrYx$<+*EC3-QLLvX?&dxh+uP<7r#uSe^0zB@OLW-yER!`eo*~{;@`f6 zhy^hE^{}_x_ON74>7(gCV!QDmm~=bbh;0Lr9-!it4@J2>ez>pX{9yst*t}@IqVL|} zz0D%xDLn%`)_)}h##04tPM_e8P*-Db6s8xV`5EmI1t$-@e=5FM2{^MvA#zPdG0Wk1 zq25JMC+BX9GbFlE8*j=&A@JBx>P|uS7mMebRFgvXXDM0TOZl7_&&^pFFZ`EVcU)j~G zzW@5ik3+jd?C-D4f(na8!+H&lBc2nQtjmQ(#R-a;k)bX)LDBs${bPyexiyXcLpF7a zS3ZsXv*q)}a8}3*d&?WSLPj%e1pA@Q>f&`(I(^{%f%nb$)h~+hDqDrxoYz%3#T{w87ki;R@|p_1PFp%dN`=%>U?pdIh@Xe z#NcwTpwo^jQ^YimQmG`aYofBLRTGKfi6ve)TrwF?ihZ;8`t?YTg+OcY4!wuZPfMyh zEE9HV?MGn~08c~>54o@qDv4nS-7zGEvH`v4aH+Ge%(Yi?8{)feZOu?jh^7fkCN>%d$$1{ zl;+&ljeug)c{+i@rcTG*`V#p^i?DgEM(E(B?E0t_v+azw)C7=)g%Ua67?1hWvGk|W z*jT9m!aa5$`@tX+$e=jF=?b?BMZ&4*XNiN;6t?Dz5sdzCJ-8@m1M9v&cON~?3u)O9 z|GbTtqM{r`Q@*5mOroto3Tu~vnu2Dymog!Q!6o*E;(s>lmri@o_jpJG3{d(x@7{P# zVzssc2_R&iR+R{zuA8HQ#ybUa$cb~WeMlnwbCT2|sXMq(O3k`D{sKmMk(fTZlMlOA zb9OhM4&CsI)J>Txi#kFuM{`17g=PfWqC8SL1<7pbFfdY0vpG8UTpl_lgQCsPcpg%LD1x#|BN** z6gHswoe#>Odg>#wS3Os`+HYk9iZ zGc{?RxJSv*`>_!l+~DzTk-cBNzmRcFHb^+G_Pg-M>7Mpf#F(q7VObAvRao^B*pOjp zH{kG%OixyfG(!@SAfpp0Tby`(<_ihndEHm2mW>ghU-ZwS!E4KKcpbKWl|1F)@0s7+ z$uW&y>xZ@dVDa&_`FTt#+892H{+2WHxnm?b$IYsL-3;$@iHj_!j!&8*A5U#ve@*so zqQ4J>MFVR(M=HkZqmk)1OMJ@AiSKHS4GrH!j36^4h}ov*bY18w9E@X=L3*zhUOFXpJR0Ta=Zg*nOn3mE zuOmmyh#&M=04lPPt&|Zl$<%HPVih*8`bY8KF8}K)_lScksf3EkSv$BEoUSt7+DFA> z%x%#Y`YU3P?+2i0A{y~^D}T_$K>-*O^v93`fs+TpnHQ+uO0P=U@Y+yK#sdOLDs>Q@ zs&@D%ST-x|Bl66ER;R)8`IVw57FL5%AqB&f4?F6+qcXD**z{+%m5O9xs(fO-lG@^6 zivDN35y$vHqc%c2@1=v;JdM4p1XCfL3LyTro_u z{;2Tz9)J>+$FeW5^efX;kX1Up`m#;6OiKX1P;EPquZy(45J9?2E91&II{|+`Z*kAU z#`7^dXp1^k4d5QqEazZtgyz#<&izk9g6vH5p|!2ja?YPk(AcxySItB~Jk@%_a0{9OLq+E}v7M8c?&z${6AM9UZWf z6|Lo;B+$6Dy2>$d8*BbK6bJjdVgM{IPE>IsevT#Dhub~;R3UosjEVA6-J^OYm$pvBo#!6VqAyXd{u&mq<1JOiP&Mft_<7q zsDHahq~(OCOrks@rmDm7cQYpf@(Q~QuKg?q0ef9y6R;&C)K(-~&+CKu97*Dy07??uJp~dw}@S_ zaL#=kPX-Z*5NxHYtkG`P1~ekvcP)$j%#O~%rT{1T{==ubr!(r;di>dt^`BRI>G0Br zmh_mZM4Q&}BG8cz21yxznLr9dz^)2$cyZyDth*{&TCX1-9@Gv}PYkvL#O&>3zkJcs z3*#3P*sUuqXVv8JZF5^ttEN8=V+jek`A!(bVW;%N!P2EKscD{2t%pplf9F!H`_J#O z0-6)fACdu_`ccswG^(zaMbhP4TS#sq88k(;svm;`I!&kZl0~idb#pJIXJivU9qUeI zxR&3JzrkW@cwhD_3o2KVNJ2DK+ZjoRM~UR+b-U6ZmR_XDf>Js8x#Ul}ul5hjQND*U ztyi*Q-G9Yp5d>t@e!9Px6&vwbu}8Gp^gBYNQ3(9pn09c^xo z87wAviTYK7wz$r=`Cs+m-~adDG5<$P_-~)$V*zBfxv`BV=Kll{{Kwe-4{yCdqkb@% z(f=2}eV?da+LKR5ev2y)1{>6IS}l$kp?Jffj9ecyot>SpIm4ts6_%Arv9gW1`J=68 z5Q`uJT-`k6Wu2V3_j-E7t?X>l4iBvovr~#-?LOXx%{7L_n(An?zr*0u^V9i3&y|Ya zm%@M!=wQ3^&|gX>&?J-%wg0&g70Fj|@&jK%6&t2##c2>r=vgKGBV^|3 zsa4L(wzG3kKQp^ZD4ZHC>i*mz(6io@pohXgVRg2joN;*210H=NZlT%ybF#dm%-ub* z9qe@)Nmvp*)S{vZ=S$3sOmBb3?J#VB_f|CbV9QHOg=ZGx;$005O|mX76+)5Yx$Ajn z=jqwE3}t|i))t*^!$CO4;AQuG$?_6SD}kzP1#7q816u)NE$^eX+hNpus;6BQRVP7> zrmwhVp){AUqOw-0y?Uyl)==M(*1@ZNi_xOepJKQ-xY#oC@|1%Vv;E<3w|a8ph5<7I zsu~8YH;sG~N-#wrQo>=Zp4Gd-3%=QsH!tRXvPf(uS(OamkUMVrE4{sa%#Dn=`CsJJ zS9c$uL60Rbqog7U&O1vOUhYS1MoOQ|CXT;Hb|mlmkps^Ma%y&4Nexm$vHr?hSw*f* z%h%XzZk;44fzsfjmer6e`AYe!1Qp)WEmY|8@9o`g*irTL%F<8K`Wy2L$%K}@F$eR3 zx&ddXZGo--oh0=Vuc_lZpM@BHz29{$RCxRQ$B7`wc@!1Drxx5%)^j+B7BSpx)lO77 z!6wdqRf{Lz8LuUGO$JQ8)cSJ{5Pq3Cnwg$v0v!vzzWO^sBZ&aMAAyXFs9EfkAcuz# zhU?($oSP6_=?Ly7)>=;J2PT<;r9gxLc<8JdpaVMc*NXa_4|2fL6rd>(U`8ve+ z{DSk~=p^mz?uk*Jgagmv$$4hYvr9IX2E;TZcCKk;GI61*Vr)zwZRZsi9B`LiQTZA- z=VFG<;b7-%TR(R!2CBp7Zr$!E?(YA9Gh-&pxi)<`jp29BxGW=xl<-H}f%^6a7l4D< z=HnfDl{8eEof zWU4D-6tP)+I34>$ zdIc}yi^tQxe2y2GvIPs`+i~?yyVW<+nWSuAe#^a&JWHj;Nvza+tK%rTbA9-?%iiu@ z-;mjKuh$!LbVuxXqUfgwm%U*H%9??YhIhdgOHn!ODAldr-?I8b=JBluEkixGaTZ5; zwWxFWy>v*psFo81tWJCm>3;66EgNS~k#$@ol}rw2%(Ly1rW9OcrvI+;c--&Y&n_l z?Eqe0iO@Juos_o^Z089Vq|S4HKD~XJIYisW_*6Z4XHDrfg1DT(iz0(0dV9{r#N|-g zNR8FqrUyBd+gy4mvoM_F@jCLDA2rtr*ZaMMJ?3!h#asr*PhE~@RB<@n)`{CoMe-yP z2ThtZ9WRPy6;-KfC6gnG16XFULEo{wb5WzHCZ2wr z0M_3760r?Uu~y+`xs58eP z7;X}|M+NXIDUaj{G+FYUV|iZz)Rijo*mV)5uy<)+?qj?WW#;+2eUHP-RhH8<0FHUb zS|aKS6}@8HwmK=3m`MfzN1zVhp?&x%i(3E-oI5yEMfLAXf- zbyFl=yfw;1mJJa|@2~^uNV(ejYSithCKV*wsV;hNnO3RWu9G%N+S53f>2Rj}15yK7 z_lbH^q(ibbboc9$HW;$5XO0s}Wn0vvbAa6CciiSF?k+E@kyoLGg%UZbpGF}AQKZ84o zJAd}^u>D7CanW}wB)ve{F}+w++#bOt4p$K&o4?pz80!5D{{E@A}6O&@nFsJbmEp9>Fo}KC}0Q50OIGk z*w~V05~FJ=>5H08!^G(E$tb91?s*^`7WLJZHX${B&BsqrpVqwzUB1}qe1@SAr%z-b zwkF_7_;hm)E>ApQDr{i*F z0KKSgPrKCSb@H0Cwrk8LmNkGQ=a{G9-HKQA@uO|?l-xSL&UQ2X-=44=nX&Fy&r1C_cRNOMDU8R>cbc_~Jly+~UK3&3w2wIoTS5NaC zkvU%;)ej+5zPps}J9Yd4;IHOTG(wH|cOqI(iGr!9+|+LHOFSBA+4qP! z|LAS;W=Ndfsw;|{D^yuz3g8p%j`SN9{ZZIqy0vZG-tq3MYNrwip5@YiSPr+SWFy(yhomPNn-zP4ZTD$4Ewk zym0C)pM?d$Tdu{D@wK~gUpzYAw0-ZQHubDN&g82QH9dPbi$!Biwb5ake;9eTX5-Km zzT_Kiwh+^nyarbNoC}h`i$Dg^4x*ULu;f!p2kRAIPn>L7MR#9uLGrWJC>th@7OMtS zuTdQb)aWMZwXlPugfab7nZpq9^RVExV{Hc#8>;~9Udz8e41eY=Kl4AwNEZ908X*=$ z^;oG_`=zPtK0{PmlikpEK;<1^i7O3#1sol^gx zDmCCp0wV!aT<~Dsvso=Xz0-$%$vX^=t2^*bw*9;F-;~Q9`yl}#jy zw+g=Y=+;?c^uud0pS`&(u`r0=+A?WTgoYB_zZ}s%1798-;OD_Ls*?8%y^$9n4**L| z&Xjx4qYr+k{c>ACyu)l4NWnh}HL+c`@(y=(v@Dd4+d$~3s?U@~afGdOaX`8MttC_B z_lix-UvL!^1-=C8C2oY)y4VO$ubf%8ZzvZ=ur_3i`}p3j{&G&iTfLo5Ot0NFSh@`x zY$z1M*THpp^|%(RG`CcMKbW9tXnvwro)a$PHq;_p*>2J@H9j`>k1AVKp04)XmOqFn zv@sfWrd8?<_1xR5d+DB}Xb!5y*P6ru=w)M$KsSew8Q5a>1 z1xm*xR$~l}>Cd3+_G*`s%{`1w-;c#m7`oi9TdEd)Knq-g|S@-^jwk9uIX!-nf)P-C}>kRr?LU zy_K`lyC;QQRbAN?falR0!o#USgj0%!6Rm%z@S#VCpCu?N6vMU5Z$9U6z;sIvLWe zbfy=wAbsg=(L@p`6qV0JHpcPSP%FLTxGBoX zU~-H@UOCeeTugdc%FUQ8Uv2!~VU}1JJsdWLX=D-q6Tv_ ze=s2bLgRRCz?}G75Webqt2`w#zs%DAh0ySCmdJqhhhuU!|AnIau@Eh=&tDKv=jz(! T!-}F1;4fJz#gEkzMnV4vLTRt? literal 0 HcmV?d00001 From 710246a581ab85cc6435a22c90c1454690e4333c Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:54:44 -0600 Subject: [PATCH 509/590] Add files via upload --- .../assets/ScreenShot-2025-11-10-125411@2x.png | Bin 0 -> 327887 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/ScreenShot-2025-11-10-125411@2x.png diff --git a/ALCF/notes/assets/ScreenShot-2025-11-10-125411@2x.png b/ALCF/notes/assets/ScreenShot-2025-11-10-125411@2x.png new file mode 100644 index 0000000000000000000000000000000000000000..cdd2f3337818b5a24301e1135fb55143fe3c975a GIT binary patch literal 327887 zcmeEvXIN9))-Ga0K}3+ERFR@m0wTRC3L;HKdheakdlBhfKnO^+P^1^>kVprq(jkyY z4K0Mw0wjbR-1|FcA3ghf?tbq5=^vhC&9deibIx~;@s4+eh^H#@mo8qvNJK<*N#XIs z=R`#06+}d5#z@W)t_+Wj^Aiyfb6LO8a@A5+5;Fzb^BS80P0V>c?Hvi_L_`wOo{q+* zw&t!ZCgzsb4w9^>1_&#QwV5QVrl2yPvZJiImGxt9XLB`gl^3Sowx*(HtkO~!B|OCl z9oU<@8nbxX+c~(1c}lYW=vR#J`Q$QymE}hjS6fL|E#;>yvOs5Z79n0fUOrZ-i!2h( zW)@=4AIkmNnD9-K)ymb?Q49d^@bKXE5a0zmTLSn+MMVL8cL8_r@(^n9xOh3Z8hi3M zxZFJH;*UNan!A`fTRXa1107gS`ZP8Hy17cSvJ%Q!{_Nb<+T!26JGlJ$_yj`$PD%j$ zynKMaRyX&w{+sG2CI7B|(pXIExw#9_&h6y(lC1pq?g99JboHZ5`+q6>cage-HDSDW ze>6L}f!I@PPjfr%ht~Gy4laLmdsjf@r(&(Y6blFo{iU_Ki@CEb(4Nqk(5srMvpJy` zE1)YY;Kb;FKdnxfKES``kMQMRE5O`=FzwES`3)N>yG}$zrf4lI`&24h`uYt~Hb9lEw?CTqiZb!f;^^Ja!;8pkLHu2_{Z72t9aiOT$1EYMNN&va%VN$4!we=%omJ3HFSZ;7lGL=p;EfWlTZ;Af2~6 zf&^7_TZ#7kZF?j9;9Vys0J(`|c;UmXFBAT0R2AL)r?+8BASclyoC~;U;Cv2V_>sGW zd`_ITWu)iPVT$Q}3q97h7vZP3+Amz8u$5uBF%Wxctw!N?*Mc0`MAudRuUgzJ$q{VL z_&a_@_w;G54g(nUUvFKep`!KAQH-mi$5yJXH$oC@=|FnVZW z7%m* zL|*D&CN^(3ZORgh+6&GL7HYZKa_NEiY(JaRJG*d@>3#8$%c}#bQ*j5)#o4)B8*x6e zZu!vyPN%=WZ200OQVU3*q^_9uha_nmU;2jEJtHmg&R;x|>TC|#hy7=Ep|K-n!;XwE zo-G}A;_`i!>3fAe$}E?0)65lsVEc)uigs7riNwP)+IWqWwIq#P zfJf4&@Z=6;+2sqk_W1@>d{z3IVUsSF0nR_%dN{oGCM{lETTGv$zpK~O={v1#+nmCg z*qd7YD*Yuyw;<953< z?>OVY2g#*#j*B5ga0b^}>`ov*-+3WECd=>V8ecM#oO#DW*%rvgcf~vF!gw8-(piCK ziKl0td~q&2_a;c$h%2ACq1pfG<+ooH9BK53HNRYPq%FxL-z0Vo^7t zp@>E~sdsb*v8&l~`P5>R<{udDk;TYzW?x&r`k6@oCGVr2!6&Y?I$@a)0q=wd-FMDT z1)Wm5WNMO|a=N3VR*Nk;I>+>`osdR!l1Y*l&lpM8wqOzCT8q@&!}P#P$MdzXWH&wa zUaB@*&MR&@H(o4lRh=K-4D%Gm$H?5Iqg=m`d4|cD;YJW&($j~x*zGtOSknW2J_T*Y zK4!ZcG4{}-mNqe5yX{&Vd)uWp$u_Dsp*d5N%MlNUI42Vu72?^XZpXcNs3Sl1NKB>v z$tsH3I0t-Ts^ zu}`Q@a?h@Q>Jzx9fmEA{yGGZV-fT`nw|xytXGzB>_K?w*F;gX3C8FOULnFgM-(H_~ z=FE(BJ$1dj8#I+}II`Pxi4Vk&5dhwmz9ewH;~K~HYELa>Sz#+|?b7)6Fx3k68uDAi zJB&hB4s(81G2{Wn;23<$!N`)ro^RCXL-Vzul3B6VrclyQ+RJchoUrW3U_W1nW9Y0# zD@`lYP16rXH$Qz7`=BPL-=SnGJDPqvLwsx-& zIx={#?S*lD`{&V}%G4_DM(uZ#OqJNB7cSLa^#?ZF7TXS_R8i$L0z5X{{*@oKDtH8V z#&t3)eSzloNmf;Z1+dk+s!YLw`Y-hzhnbi9I>s#`m2>*sbl>wpg`n<1NNVws{!!lH zl9r7MNgI>}pc-xj7kb6lVj67ra4}q$e3hIb%s1?MSXiu5r(@@XPVY|8O&&RpI6eNY z=`Ns3iHfKSG<_p>h>2RP!12X={&>H8KQZDWV(o+VhakCuYcqU@Nl{HvUFIKp>@`Ez zw8le9`brFErf19?svI&lXf{4=tjz$+-6o!`v#)P6ow*zkC(5ME#4I@@u`Pk|!7na; zPeN7=uN}~>@xSMXc1Y|AZ7=x%(GD2ngO7)Hd)64i-9@a%q12WDhU(BBTXy)YZ zWN;{c(1W{t06};=TRN$pF*?57G~^Jd;IoBRDvu zh@`GvVD5tFXye9H%%JwWyw;Iclh&`TOd-&ao)-Gf_olL)CD(lbEq5PKK4ZSX>J?we zijvosSB?;)RE~~~ED86&E^x=fqPI>p3!HcjaQx0XOCZ~eSF5oRbMq|JDdAxpB4Rnto;LRsV=TqOUt*s?bFB4 z3q6$02~pBTHqboL;)zyqpanNYC zVC36={o4&C>*NzNPuHF{m|u2L?KUXWEXTz8eZo3yT!zAZU;1V5WiK}GhS7yZuDz~b zahusq-9=tfxMUwDds$f6TaZa?&eKdj(ORwj+02KY*k^G(B4a*{-#rH|tVEO~UgKmI zRgz)8-9N7b8q+v1`g21m)q zP(*_Z>dFQ+e7w={@XO}?MN~BeW$j@reaQFSxaY%!yazKW6U37>U@OP8S#4@~$dY5A zAWK{txJtI#xYfBYEoSi8(0Aj`q1xKOF2LS^3D<)vP2O;tVJtN(^)>`}eK@x0)CJY_ zNj{Uj3|jPn%=vT%^&_H95K;5vOYFhqPnLs|Sw>@DoVxPq)zwq???3c@ zw(|H&&h1xMSMEGkJA-%q$mUA~Z(i)%**nWfyEKL{) z>-DvCa0>8(c=<`;_vck*ZBj*DG*a%m>7?+_XS^t8Jw<$mgn~upCDFe=1TImec)+|` z9*~lgxb#n!imK#y4?Jir?FPOczWLXRCoOiIPocpbOhkk~jh{JwE~Qx33$#IlrB)t} z)#MG3P;n@&P7Ao)`{kCWi0vt~T)zbx;}4eXTQi$1Y0+SqOR4_te+}x-57`Z*;h0fK z(460H%;*0Y^^XcxU`Q%MS_+z{MdysdIYIU->CG? z+IXpGX@~ZeUq-lN6{uZl5rTnXSsIV_b~f?nnUnQwqpkaduyz&}JkPG6xaP~A{Ccy8 z#B+WzaI$&qb>YhMKO5zx41(md5NsTAgpuCg$pIM_iWIif=(Ca0@fkj9y7z#G|9junyX)F5DKzl zfsI46h{sib9^SG6!OWv5kf)&=z3xxDf-teob5aj^BPZTA%|b3k>Af(}u0kAEErUry3cJ#!|L;H^Qp2 z-QaR`|G7P3M0n1pOg-MHCX7#$a^3y9XQk(ff7ya8gP(~P%Zs+OG>0=i&u40#)H~0l zGyY$?n_?$)cSpf2!?pyET%GR!FIB`8vn7%RUB-&~#axG(@}{$=^Q~2VeWv%OgcCC_{{)n<<8$u_s;}%b$LMKtQjgKoO8?jCp|AoGx+`RhuhN%< zrvBQ|Ke{6GM26M&qHO#>YV=EG7U@$FMF-le5xYp_Fwd3C~^LO_94Ksg-#lKb zKP6Ru2Ti|&rhkwoJptw4F!LK`e#4B+Z{5t_xFHn(zjQNvKK}YEf2f;wqynK1#}Gb? z^@&JQ=6hE0Wi!?APx+eP4v1+D4-tI6`7cCnRY=PwT~vi zvBy?fgEgg?fUD0WFpn-ke$vnczJKX?|Ji%5fEpo{u}00o?Sb`Ad0FHb=}n{dkiC^* zPMjyik84%^mvDIXf>OE&|FnB&#%zo*jO6rrUIw8l$JVlzt)MmJ7_uhda3ybc!)N2t zKr{oj#gV0kmjCAWvdP@EV1Jk1Z^+W(2E8v3X5U+S{_ROTdC*cr$R-q?{*0d-IQ7|j z_G@9Oy)}r}ZWqbZ@?4QOwdDE6ePX% zGy3J$qv{&*T{`cE1gVAL&Wb5~#wI<@VVCHqo~zIr&`EOWIW4NM6eMp+`Jku|70m3> zo6|KDyNDiNJ6Gh=ztW)*CP5ktL;D&q9po1XW#YX7fB3^*53^m zG241s+D^2$9i&IIr)aPWS>qM28QJ~ZQgn>tWZ0?93hT`BtXLQ-Y1~?M#CutWh0y4g zeS|&fB$;-@NX~ZWWjd_*BCE!V9-G9_=c#7JU!#hGhL$uXKO7_ZcA7RIPjWN2p6$n} zJ$UhyT2vLW0@WX9c(%{QXnKsC0;O$Zq_J!zm39MyTFmypobS75+qzfjt$$IOE-MfV zSoD%`X8Jxk=>)7Qb-;$hUk)Hc=NwDA)$LSBYqFJh*!m%N$6K#iW{#=>8dged+djl4 zr>j+6xb@_^_KM)i6!jImzx!i~9ukCT7sCD&mu7QbAOnzw_DN#&jFXZyR~10ZI{zj7 z8=>gNC`TmC|{@80C*Mgz*+KuyVlNA+&6-Gnk1vzM9Ys6c?(e)#4}UY5p; zgRB&BY2%?z$~|jm1E1Q zQ_9kZW3iC5R{1xv&@_TByfu=G8;(Bj9_(=QJL)rZ`VQwGdSz$o6v=UztK6ULi4|oa zovZIUn)6A&MTT4BXz`i0i$9yO9HE-#gNP7xL1*b=s+Ce8w4# z`$Z8&`K21=U!QQ%P7S1s3REhjQ)K7y+7{PYc+Q4=O9j<*tja!sgO>uowsn^~w}s0q zdh7ZW3)X?cwa^}`*Os{ho#tUAP-K}}`>emqlC|-X+jw!l=8yHm$o2g=!lS4Z_6Mp2 z@=|(k9MHCiqX_ci9b?(BZN7SrV$kcUTnW>xAgqACLJ6$6(sWL##w_`1y&(y-Rc;vcW6*2!n6r6VzD&P?vQ zT;tYgDA%cSN?e9FiI}3%bCqL#WYM-817R=hxCVX&a!MXW>NEdg)j_T^auPa+8a0p3 zk(2dsH@zsWb*EMMmCDM##iF9|W;5;Q$sfcKCvvKGBiU*u?1sy}e&q^-IzS|&nWd$; zKM%Id_R!l?XfWSpt1-Q2Du0BK+WA>|fnxhPFN-5f}FA59gyTW`IA)%V{X!xNiPzH`;Dvo^k}Zafw$GR6wPABnGA6d6ij;Efea}H zpR`Ehv8!ue3dwT!$PEfT6B`eR!M!ftOihcSm*%EfZm?Yo6u;?l19^@RYL>^#!&#N8e7^~0agATX1| zXsI^SUET-28%>AS>VA~=@_leCVLN1S^I3mgP?37ydLk>8;UP#Mt6c5=e959d=S!lh z6g9!N$Dh-EvZVZ4_yj7dJ%0gs!eSslQgus$d zzR$-=lP7X0lM`OyAgi?KTYY02)p-rmL2rDxb8U85RT}x)R3+n3BSoa#w8Z)cly1{P z_Cv}D%m2yhTjRx6v!#h;T=RM6&nvETO%cs&1SH8*#tLt%WMVDA;{7rIxgK3(2#5Ln z1|v16&swoHBMsCRsop_n5e@L)>gFt-$@!@7djsc)iLoqa%o;=B-DM@adg)Z#$~qD;gPzj-(S_E{Q^YI)K%QaBz1d z%z0o+!Y@~`S`D8hK9 zChnMgFjY}Az};i1N{oiEd8SzTe01F_F8>JtPQ?5#H#awDXTexzlFQONk3C(6J|@VF z;I{i{Lpds=96ry6&h{9>_;!w`5mSD5OCeIcdVWa57Ov99=lfd{qF$$;@$U;_*DJ$v zKstBrQW#KchDT4dj6)Xbr3@N#(;N=uJ@;qjFCIbS&rC=?P7sOek=X8~8Z)r%+gywh zA~QVt3Tj12A0N!+3LP;?y}uXlHZd3*PzX)ZG)y~F_vAFgS}``~1Y1CQtvwOw(QzC* z4%>X5^mKJ3JtD&cRyI{o7@XlTRc`dG2yn~@$%ngR8E5O)fFyt>oO6n(i^aR{?E672 z-xtAj8U|iDWy9pquz|%*y*JDaH^(J0$bb#;E&TFN*yb8SSWDRXvtI4p8(<80^ULXD z@>1}j&epb#Ct}7!mnq_Y6gTCVl8H|gpxR1Nzp7>0rH|2lde(+dK8Kp@^Zi+pU!k0Gwl&TaIp(7 z4(K&*oZw?b4S-r`azS_Ye8a}`p-z1dxLWu@E*Hqad!-C549+aML-wk1;}AkC(W#m; zn9$2;+JY8^Z8Shbf%dpQ3RCQ8@|8F#nvy3bQ%<38=AtMOX+YS*0W(9g;zsu#)K8eZ zemUEwq3{ncadVo=OO>U0^fPRFNqJ{C;!}7<8g2?m%FE%&?M<9isbja8z z`(kwcx?aE47|vBD{m>cB({>w{idJbCDc|V_;SaW2FkfBu%Nsi{yZ1~d@9GO|RCav{ zoQ3K-Je!u>_80@tEc)0~p*GMDgq1gXD5C(GUVY_TnQY`jd+oG^4nVV31_d1Ae1kt^ zw4p6{R2)^$Iqf$fIlE#h7b>Fgz$pU04Z_aRk`9dJrrD*p&Czf<*o>n3!y)Zdx8vl= zRZQ&pmj9yZ} z{H(${`r86bk_zVU=9iG;?LI<3s&juqN{D zd&MTmunpFsp-+gKDzh=6?Q`id_R=hvO-1;2>OD4m%}3QNsSwD!#RPP6A&04u$28tY z(A2uPP7}T87jOV|n{8j%o=;*6mGZ|~IyKOd#?Fgr>e^SaZA_OscsSW@UhKJC0xkJK z@7jTogE?X8g8VoRaOv@lBrF}0^8vc^5L+{z8aO^+c5(rJsUle!l7y=?=NZyR}!w{F`+MrUu$`8ttd}VtCQTTD1DU3bp9mG(}jjMqyTp zmG>9bJ)j%kI*g5I6T9%%L?&BMkzslJgCB+*4l~)NfOm~8QSclTr9(zgLQf=bR8H?97UVflh3eXj=xET6zmI?R{DFWED`1H zNR>0hsETZWZk!VsR>TeYk9_N-c{R3jxaG`Xx3i|EvRW1>mx*;of8Fi(-{2aMx%2~u zmnDfmWHRbImva@@>UDbqb-gwY%Gp8OFwGX@kdre`?5Rv+wv>ZqC+XC6PDR6Rft&In zF8z|Fe2%>O@Ug8gfuvz+cz^x&4ejzeI(pylum&BIV6QNnSi_{!abX|(soLbHPs$DlfX9iS*@t& zkRb*j-MIN`0AD=IT!ZJr9ex&gkUp6`Mww_na4@5N1K*beqHC498^JGqFfzQd8tKen z!h4?I03%?I zhpCX;y&GDnSUX%FFgPe+Zce7> znB2RXjko)RilNmL*&pVB?f58&$CP;u1*i8d-Uj~w(B}M+R1`J!+Asv_0vS5G+QW}W zhmjs%bx<)VqDiD#ya0bmqa{77{t z!DtOhThn<#zoc<>)p0RN6Qr~;^wzJiN_4fWwN=B-K@2M1Uz&djYf$ITmI zRbml>1_2CYREdCq&&R5F5^k*>G#*ab8e=ouc^mETaA;S%M*uUR)(HK?V5l^Qz-#}f zE;VYZ(l;6W7rwo#QYb2~qx zT}wpHw{rOYdFG)4wy|2-ds6G+9~6nLoyMY$0-TLz2-l96&Ez6xFo(vY_n((53vg#* z2R0t~WMK!qpa1mk*o>}BEI!9?~ zEBQT7y6|n=b!buT+(opeaRjzQ`xzJl-3VD2mrKwxJ@3RaVjw*&eBwXhp~gYS*sjeh z-Q^f7bxPay!@z)J@%5b}{npLdI3drEOSV!@b7>)ok>&Fu-*=1Pnfk&Jf*X9Ish4ha z)R{7T{`SKu?RQ$$1yUE7Qf^C0PeCkvtE*1mJ^V#x+Y;9}$~zLn(Z|c8 z=3n0=tHliLX8NF@1%%(KSrKSa4cOb>#Lzp~Od1CM)tA=4F~|)+n8Ot950&jZXYW(> z6ldVX>@hb`8!F@i@U5H3)~l+{L}@}z^l!>p;A z7a0LDXB6o07$VG?vL%FVjS!)diL;r~0~^t~3FRKz2;HKn2p|@>+&No1u5VaWfb}5+ z86oAF4X8Q~0;hHUNe9;$^dQ3neyNkqsgg=XEN3Pnj`2#|YUI9wWaXGcmIxX$FF0@4 z=!4w;put|=0nYM1Txszf-PeY9QCe~z!S=*MqVTQ?1m5IkC-Q4G@?d&-wbn1wLH}Sf z1L3wDMR;ue8!YivTqBH$KTPoP3@qUIsD-6Sm*BbA>@>-ljBnck%_LY!Pd)Rx_B>JV-Qy6>P)QNMzs`Xvgs51e^LoEiZW&1!59ALw*VShu>Mto27{R*lj`Go^bUmRxi_{4FAF)G!*MKittaS%j}I;3cPtI!!YlbrIFLwh^jwv;VI57%+hLx)<6+ z)7RleL;ika2jU)hWHSamUns~mQd-+$88KdUy+dlg(rI~ComWRYkcTj>)He7Efvzw8#w7rrj(mPN6>xUo~)cCho2N|glddtIA}@!8uh zF}sPWnDtCsD6L!^xet4D$G9Q$c&nf-NXyF8iGvXg{~?*?<&u*?|!T zbICe(fzJx}o%&CfLgqYaIPC&6z{)oLgg3u{I076Ly5+Y#wB%2cYw)IZw3ih+y%~1k=I3I6>}GRKz8`AnnHK9fdgNe0XVfnkGSqAh{Q&4 zJp#H)-eBB{Fr?E@vF7LqhgP%+Pr?aAaq(1Dp_Fhz#`MT1AT_1@eDvk2WG_UrTlKh=GT`Qxz1yCs|dlTeCX`FR7QsW%^rf-Ti!9E%$}}a19V%wr>&@ zl0uQpMcJrM4(^vE3l z2soi?s%1=5Z)Etn|GDJ6YriaN8mN1oeo@sEZD=&Ek5M%4vMHptp!Qozjh~I+%;F|n zh%gN80nzP9pcm9yyE&N*!Jo_}!F~8xaeS${r2d4_(QeDTFb0NmJQa@>cXffvTxTqK zvD?%<3h_NWn0Cjm=JQpbZv)x@JK8|+D1i@5Ui-!h#Ovd_#1EW0MpIt+p!_h@uv^kD zOT7;~!kGQ7ZoI2*MCq)d59dJ(I~Dz9#w(*LJeJprUXv9-9fK~l_-^;no4m#bo@Xk$ z{yI7^mvD-FT1dxYP==3LVlhT1d6Q3hnHS=u(GL~UcD|~2+m&!;IO^3G^z>w6pQc!g zlp?@A!2F-z?-wN@EMfYSCCuIbz`GaVpCo#7X;n6q%_-L~B!osMe=!!gRSXP8)?giG z`?d!pk9`&WxZ7w)G90G?t7QOBbuA7VO9@y#$6J1J?jP^pe zKS=m~N%&32Y~)cpZM)U?4f?6AF{7tU{hEo6(R8jCuoX<*%6=1&=i(CPQ(I(i&=4$hA)S?6l%XL-p z>ad?jJAo?NUtAPV)Rvy(&R@wPgoHO>n8jpW-W`RFgS74<3iLGxv(&q8HVWGzv<4!+ z+w_J2vc~A{;W(*7t9rNB4vVUh#cL~YwRz7bQ8i3{ z^7zPYwM4p3h(KCf1lu0mBdgcGe5-a|wIKZFRr<~|gGU&sa8XKIxe28$4v(L*E#>SU z91mM7Irb^n?m=-S@Q0r;APNj6XxjcNLHwBW=8GF0w}yJk*iuUck_l8KCxhEgxq713 zJiNbO+#xIakzPHY1%Iq619na^q?^5hgXN(0EN!UE0N#F7EiTc+cYe)~)C1<`p^!$M z%*>#nQcZ0V*WILR)qJ+Khd`&%`FO3EgDhk7qO)j0hEC=po1D_TD}_~DCk{{VzJYCN zWfii*K!n1y#kQnQPK&ROc|K8VDp6G z)EHG_aHnh8y$uL~`@I+xO~|DNZ4NBdg4pjTrQFh8Qw9!zF6nS(cKS{%ZG^!KWbv{e?<=;6jZg@G@RxnUq3`nBFr}Y;YXLRzG7ZM+^RI@RQKLECe1*& zz!hrxjek1#!Y+i<;y^+Jp7(X<%`QvsIG!YK+M8klno$a$R@lZ&KAgB&^ZssjcWoCo z*n`?lOf1ryzQ1Hf8QM*4qPEz^2&uL^>rIHv?(7d|ZT6_CEHIOZu-RZLRbH~qY1d)p z7J?6~@fgPi=g$m?p7gWMaBKP2iF?n-SJtS^V5I?w7wnkLyr%6}9g5;`n#p3`8BV6f zwVM`IEd{{aPV%lHY5{)Uo?>3{#;Q7Vc1>Y8oTYJM3`E;`27Mi+dd`c znjEEw^LPbkQ-Euq){e4|gqTw9mv=HyC+j$Oo2buv_-fA?%QGa2>OR6L(It^}#0FXM z{53JNOcrQ?>j!~1kTq~#hS$+6Z%t!GbqX}K2?VWj!t28fvvd7HUsmQ|%oj8$%t=(avonJFqYz{LM zii&9OWO8vG#qJUUEp;lRfX3sv7G80r7NufDw^I*Wgnbx?Q?dvRuZnBD!Q8g{sNi_8 zpMR}peZBf3!EO919tYR~s5%>1S|Qjf4Ez^Ijpx#U$yv{h85T-aB}LmN`3WSCcw|5& zB4fM3an;%~ScKHp+iR!1Qx_Vk-qT70r%dnCeSwoZUsm`fpb%8>$dRj0D2z?Xzs4 zK;|2lKO*6#`z&D5&7jNDy^SN13HQ~VUTRh?wDu(gYC@zIKlb{WE;0ie^H-}8nE^Lk zbjc!hDJ6tABDwSIeOX@02UMtq?b8Cvrc|Fbdh3wOO5b;^Q66U}6sG?E1c=yUp zCZ0)#pojdg$43=?#bMMSeVzEW>jSo5aNA*7<0&&j)e+5ohHODmu_~}J^~|>GHK^KF zOiMvw5oHv0Pe6k9y}6?Lz6#F`F4RLp0RNR~^smPZ!Qb9sXEAu{1)2JB%CCdZS6IHD z*b~-l2iz3D8$>RVO(9$aYSUSFsZ<72`(WWhr_g%?)(0EyUoG~#>_!IhWt1zc>fx~)pXz;=!fI8WQePna zuR;%$|C)f3IP-Q_+>oKtWwo0#G$_E^h#+e?d~4eBBE7qODy6ir1T571q+`n!(~cRT zC{ayFFeTO*tALg?T%}~MtZuK5N?@xmVppNK0>N6gKaWJ#Oq*=?p6MwpJy7&KKEg!x zP3<&%+t)p0Swwe8{gXboSA;!i3HNrYSG6?e31UII?v+%`x=m;SCx5FaY*C*Ge$Z8UrBcj>yA}=iinQa+aOQZCAUofene$XND(77Fl4)!1{7@p)1)RJe3(Sf88cZ zc>0Z(2LE7<&)qKtFiK0J{JpW#lgO-*&s)iYHp>^MU~#m}KZz8eAOD0~NQ@KEjb_q6 zI4f0YOn+xy7s2dwo%1yqOm=jV#CX6oC~b(^5>@EM^>`^MN$YIzn8QAti6 z5jK_O9c|FDHlKN-0ue=-vFG-iwi-~2d|g7mF7ac}_8Q0K&Op#;a(80h3+Ri-u^R>Z z-lK_6?86n_Ih~u=2~tt(Tceq`V<1eQ%F>sHxqc^_>^@=W$66hbqlq=US&=8oXDaWeun?O#R2jWif;cf2M)IA~VKC;}(!$5DBjq>CwJ z%a67|cb#(e4SpKvh|ws}J&!DEKh%Qw%N>x;K3{_1VD! zRa~j)C&hL5s6OGIbxAzWN$MsQYGPjHQ#J8EDs*3BREI{#>K)&)@3#8AYvOzDk@Xu- zza-SN(lC|>YK2cp~w@EB?|jI4Xw_ggz{m}h#d z)*Nm77Euc*^wL_mtl$Wu%-A-p4pG9G9nQ#G**lba)B)IV5(@Y3sp#QWtLjbz3dyq; zBEYLW75~8i67&H&r_ir(KtG>3GhmY-&vr14p9&$-?S3F_G4T$Dnt-E1Lmr{RLvaqj&ftln zoUx-Wo6^|7GL`n!R!b<~V#WdLQopF{0N{0LDeNN?d%?B!BX2BbGMqdFh+k}NfzS<;gUsp*cV6Hpw zNo<7*CG%?Z;=8a4NnyC}um-b~t;Awj_fhe#NrQ$*rVU+IoK59Rm6VCSfsoj6c{f*I zKzo7@I?eD{r_g{;lPg)s+Id%78}7YU-0NasSN=#~t%r+tz!N>!1Ash-6|Iazv31|N z&W=UKj736)ouj?&)6kU-Fi|K$trBQ)qSh+%=p)CzqO!WfNsgvyGD1yRtf=-2G3n<9l#SC;x8ONc6*l`M>+r7yA@d|V6|$Y9Qt99n zv^!onm0OwMP>?CZX}|mlu2x#Ry83|oc9at#Z(M-ATVP{)cwGU9n?^{1L<{F@g<)M{ z#r1B;7)T7mu3h^_rZqA~9ZW6Ko`RDy!l`MCy<;Ktl72&D{i=5r8Z-~1Gjz)ua|ow$ zIXM{!bYerRImj?o)W>eUP-DFuR$hOXiMsP$H9J@n^;XOj`|{N30;XIIyNam^=!$^F z3nZY)*t?^4G7fr8Ruu!+UflC{7!e%##i7*wwFa7*{S>#s{7jfNB&@yCwqJgTVNNYers}8B9x7LaHGEGWWIY zTPS@n0om^r)og$Y$SDflqqh#Xv@eGWQAFQC6>^OsvLN29Hn#niGBHP%ed(ibdBqxI z5K_koa4Ycb>7N`havBOim-G*Enl%KpB&p^!Zs=pLM?(MN1?WW^Xu_O_v|6B!w}W$A zFOwji*qoSG4UWGlPwk(plrdGX!3h>Ew`nK`fK%%12}%Vrf|P7vI(G{~AWp|Vsc~8I zrwA5$D^+qy3$M=9C<`BHG;uIiH#PAud7AL|OISZ&jHasEkPp*;9s3u`Z#jv$50_CI zPfuw%7A;WQpDtQyQgeZ-u#@i2b&vVT%za6>^$EhixS-Wf_tMm0uF#n<&tnv*$>B20 zzncr_$9qAM#~=rtN`^#jS?vdUx>VXw`T%;g)yl_fPFoR=Xzh>ay57?2n9dV2`&Z}e z$`%#VpU%b0%e~Rm)Ia{-diicm481UKQTgHPUi;BIrVnhCx&kZ~y7=xWZ2l0L=4JWo z6niNiwYmH9x|&PeF_6)n^WmX|g%H^LtkTtQmi+~|bzUK+d$)Y%Qf<)aIe#=pp4sne zhiZPhpGWgq=<%!i9Vf95)3AYCFj))3UsvY#MO(_2p-0hW+W7wyJ-=hgQN3JJA5c_F# z4naga!4g?mjSAcU z(TM)J-{0yGydwCIj2wE=u?GH3)M%q$97QIM+-xc+Zx$g!5VW$NJjV|Ako4y+8mFmC#oMZO0Knr~T_;OVn$j7|>28RV3zu=UIhZTGn zoYxWDJ+C5iTfnIzAmN4gI={SnOa?VdGd;saLvGG|+P@>!k~*#G0JAUD-!%o>hi=a1 zL=?4(84Hbt-?P`*qwC!t`fSML;hV-7C0^|=Hzrcvsa!KnxzV^?LdcyXA{&Hv8YBT% z*1meaXzbd!u4=R91t{O2yV2p-!$B%E_cq2m`ta35fgh9vZLg1jD6H0kQqbhRLJYc)TFAv(##F23r zkSRZ%JUKZ7FA3hYfu&Z1)%wGTh)Gyp%ABXXwL4t?NmWB*LUR%zvsDP$URLn;?nF9I zd8l6UmO5%bhyp;~=sGmbu;m&6#47Y|nNO6x+ZseMmHAs=)-aU!3|N{(Ec8|mB+?q< zqXf5Jl$lCk0vfE!k#tBF;o`);u2nB-eYZ(Cw#x_JhyfRZIj3n)?)_r`e~bb{$zos> z07EizisNfd*>Buv_J3w!-~LtouI-+*Uv+~J4wxqK-ijT=1f6Of&RTp2nK@ z1968nm>w^LAzQ!^6wW9fSG|+)IWZN|jKjtB!-rhlG4tXr12G$Tvf6+%6f#7o{^w72 z)tUMslrP?5V)z>U?F`lLsR)DDRcm9S2K$!-Je#bXb<5^ZN968$`Pd8#&P%%!OJjxk zMfb{OeYTc%R)*g~^|U&xYAjs<%VP@yD&jpoJ&R~g{PJhTSgF;=ok)E*Q{6V?P3v)(v@NR_;ByGU7_Mhd5xmkC zweQz|47NT)Mp?OdUQ*nR(xqs*oPUsibhN+n-Rf+4dAY{jwfQ(n);dv9CZ0n>J+MiL zo_D+Y`@p)VU*B{mE}#PZ6#B)sqqA@JTl-9{C|U+KEmUx^2W&4rRWTk-6#!-yZ#X7C zOnaMP0_}K$CEbLHO9w51G5y8m)c>;5ANC6DXP0UGa^NK>Y%n+^lG@MYG@1kwHyA-N z`@dlFU%&i5V_5HzID8uOU3P0hEx_I0IiRDht?P2YX28a7wE=;E*Q?tf;#wg1>Y$De zIY=Jr+lMeR20ocsi8;r*{r99fnxzK*Yhp-}k^5yHD40rES#bz4<19js6d2fg#Qkch?o2YTuEV8-CD%~_rJ<$Nynq2~&_MNDBEJLQJHn>2K5g%5 zcyVJwtsglKlm-3HLEpd7)T~*xCiYrYO3>8D28=B*W}=_xk$&y%qVQAg7sYa4O9vrd zgxF^&eb?LOcGJ*6RoG+Qxj33Op}8p1jK>!J(_UkH@8?WVvJ3e)<6d_sava*B3h1Zq zI4~7jMLGN@JMQ6X~`+-5QcXC~GoH!sn z*$S4bZx{?fZoNT&UjOJv7Okc%qQDADcusfC9vNZfyKoU31+5 zP><-y$jTU<_X*N+2USlB zzj^YcRZdQcS%+qFG=y?(bA-yap}+vyV)=?wP_Rr#AuTm^ZrCIQUO!l%$5uLVkmdD# zT8`YZzr>#r`s5uM<6V=M4;LT-gWi}i$Sj%9%^`i446p4rgxMf$q{Mp18qu=y0fsmn zsN1YLx4xw(=X#BjGJy3rkLcyTOpk(+u?MH!uuQlK*N<}TW`8xh$CkI-_{7YaupsJ3 zca<|8n(#iJg=5|=DrW2Ao^k@vskm<1JH8sL2`>*F9FaN}E5`9LGc)8>Esc~e>7Lls z3oz=0Bd7=-#(zVu6IZ~v;sHU}_EhG&vI`U|DbAe{MWcdSCvS1x{)?sm?c}&tO5Pzd ztF-t)$x%sxHLRS!Zd&DVLt(_4yeFZZt7429dSEJoaU6>_KDd0qfs)uNQ}7mVom~G2 z_m6u@r)qLbGfsGjK`KGC9CLtKMqSEERd|O*;Ng#AwVx{mHj^i|qp^ zp`$|$$@fL)I$In16N$1KvrAHStBa~yMp!3ujAxDAxJNRkfW*PUVKd&-$BZO=2|XAk zHOgei$c{ResH#wzAFxmYa$ok1w7aYVOn$nMo2I@n&c)v}t(NUtjI2)yuGw$bf7o{J zIX&>Q$655IptQ=s(|m8Y?_0NoQkg3wx4qPqO>gnho2D34CiE4zqn;gzN84^S6^%sc#xy zsd9V3JZnnZ(7~J5K
  • +i8G{vCHyCBZ{Vk~EQbZZ|`@#xIBO_U%#{^{ft8L@AYch}4RskFN_=bRZf{d8U`0o2rxr?wo!dgh=CCsr_cToZiqAJH5 zKf{_!r&(1X$`kz~mls4j9jCDBEG!wC-Az_Uk?awPLjJW}?Cc5!)iaLn(<2t9UtSSI zTJp7vdmTKp>9Vf-OG#{e(zim+7M>RK$HA^pG>1}v%rl(D0Z|Z8%;Z>Ti)Xd|)nBF2 zMRC)xXp)P3MyqD_d@x^`P<>r4B-GnmedQspP{qOx{^{7-8S zaEMlimK$Jjg@$cIRHzhjc*Azl=SW;+BQh&+~C$J@eN^)JXe^2D)r8^9J08DH%+cB z)V{9DU^X?ZS`K{yBpp%d<2DOQmEzDhn|pti&uZp$@&!xhsaC&tWHAvcqR)uCG1T(~ zVm`NtXYF2=4-7~G+o|x~g35KG()A-fjG|Nx!LR5XFbPXKAl4aejxGIRmY!355;3U; zk2}o|_|z^`(Z1Xd)+s$iPI-L)%72Q1ZbcbnOc+tYzJ%4At!j?)%!EWe9Cj;@-PaJS z5s5P6(7X?Z*l#HxCJN3n+zIocW;JcWSiAmZafzs2qX={*8Wiz=oE-7h2X8^9RTRjR#Dz07xkg_yY?IrQj zb$Zr-0i%;0^8ZQ-eQ=Q(5`q|oSQv-&;{=MG7yoXDl3~YX)a_tc1O0U2~Ox!FfDg?PN$sk zoXqM;z)e(8!vEpZ+|RtH-jl3ZXuNcrk}WTR6zYa*RH=#a!2&(`m%aKbQhJWkTOn;pv@tdKBqIeGeq*(+`cG6h zvI%GFD0H%v!T1YzLPR&ux6}+*T1AL&WU&Dj!ZZyb8MbU5J@H&NV*lH6DGf|ww<7t= znom7b&Bcqh)Gr&a`YWiahL`69Q8k})>+{%q-U-{w5d#>D6>wo?o#Y43t)sCdiA1%` zLjnT5`xWeB>3E|*wJ_?`6E0N&5<@XPMwX4i6>S2+D;i`v{KPd zFGYy89jzb75_5C!?r*59U&Cn6s{?Dy$jrtjhADv)2m?cNYg^nOhb9wPXm4K73=9l> zL}5{v((Dt61DklzkvJ z;ruRj2CHMYozQYRsm<(3 zQNWVxBk<002yY^Vj?~`wv{x-9{Cp=x8)i-+JX?#4>!Am&`ND)&2cv8{!La>)HDw5_ z-gy45`xvv)yWV8fz?VVrsD2RmpvsQmk$ivVgSoPF*r!iZIJDo+Xx}4EAAbFM;z9i8 zvX9i(C>%!$V&%T`GU@G*#?T~gseQoLBve$Vh0Ja*j8mt&bKqKXqW09NY>-!hq#(7EqT z4ObhfH48$$SDM%nc<shs!G_cX`(fuXkC-+ZqTWHGo<@rAw+yXz9xWwX0L68XyI z$Z4@F<`Pf6dJ>fQ_7#13bfGr;(ZQ5V3;b~??R=$yFfvgX1-A#%TyHs*5)EhV4IF2A zE^f(z)ZaA^M1;9vM;Gd_nmlyEL4TK4x7F$Q)Xu^8bu=`}QDRzJT5R@&EpRf1-oX!f zeYJ2Fot)&Q3t30cJB^?pU;sD7J!Ie`)BULc^_79Hpnm0Is33j@akP~)=Dd}ky1|c1 z<-c#KuYT@~ekmk>kvrP&e3_P)M5|db-7kjzkI6O+L^QhMW2sue{KkLpRBq&7kVq`E~Nqn0IxJ*X1w%V)iV z=Meqc!g1sVLW<+X_xHyU+ge>{$;Qr-Wp$vHjcjh!eqnE|*QOgI`f8#ghzh@EA+vZb zyI?ms&FA?{Es-^IIl+~tD$fp8SvMgI0Wz`pp5$p2!ahy1B$`M>peel}Vh@Xsy`rUw z?~;%t;}2bvPEoqBN0%dKuX-wm-vasZQm4F!%vt=|d)qOkH9i~sRVynt8R!>qHIgKK zdvZ+|xL>i!V?F*0oPtb_%jb!{?TM~q+@fl^YCW;~KHfNq-tO+HiTr?kpt2pTph(Xm zcf;Y+twF+*)Yyr90Ds^c5Ph8*+MOo@THd{P~SK986WQQ>^4|aaZ z<&{8f7SSxNOR>@bjwd!eEQwY0R3U_Xi?fHfCp;{gd)eG4E`aB;Tm>rmRyluc-`XAY z*X=Q^(VZ4XpM=Z&i7VDNHfGb`?9Dd9GX_f*>jW&^dmAtC4O}kE;xD{U8Em$)F9;E{ zo|CtM00&eEi)xtZoj-}WYD6|&lKuMcTj$CN-;hJGWY&t>jRnqC4t}PFl_7Oi;5dq* ze4N6Yjr8vO$65I_g~|pjt%FuYWctmlY{X2hQeD8%;T*CCw%72I-~IKn2P^P)yW#_v zx(zhP#DI?dyDy{IdNw1k{isPMMvuL#2>cm|Z8 zW{nT(MiWve84n0$c`08Ozq^7E-;61S=!ETH`rgYdl6aQ}UO7eIvBnyUe@$Y^y7% zck%Xw;v0_NpU_BoCp5M+#aGUfa}b9a0-7bMk?-CEr(*~spL{buV(3n9-9jRshDfYm zrxCxSqN=Kv9wYP31Mt(7GAMb}JqO59=~cPy*J9EpfPo&;lbWB4IAje+9L6BpVvL|* z!UqKf<*-CC*Y~U|NBcQaZLdxcz>Wsf78RIx+7AF@jaEp35?_xg3wK!k0MFi+eW@%q zAqeF>2zCX%REl_-NUpQ<<#VRmfQpX?z)Pb>Y>$SpCVL$9t`=^dSxjbYIU`@@1<=XN zRtLe7{;RD<9a^xWdN0Vq$h~pq;3TKPSa{t*$DWCg+c$VudFt0B+{97L&ODlGN$r$6 zqJGgXZ#F@#&a6Z0J+4*L1JuPeLO4)}rWq(ID(cRqlia0UwE_XJmzqH(y`#Ov(GzP-=L>sMA_C(u?P3f)lAO?i=>$Ou?t7u-Vw zC56@R6Tz6Y_-;8vbm-N2p7z)b;lav1BaDYECDrJ@XZr|VzlCSI+}(p4JJCkGT)za5 zZ|(v~)?(j-(pn4i9akS}1GoC6o7$fvJGiUlB?SnPhh68hiqB#98g}W$nAya{A{TOI z<)raT64nQUQ*jv@ z*F;*W#ZTI+(DgnVFgPNH=7S7+?vHC3`7qfG7Hq}0=I*Rl5`-f9>)I3*2VAhJD6P?+ z!sqiUDiU1&^d9d`X7-C&t#(EDbNszgB2mv5Efv;!mc49k7=76z*xH>?_$0lE=t-|F zu1R`t_)aNQV0&r;Iqc7`OUo0 zu&&niZMk<_6LlMjRHft;ua+Ui>Jd4w;Ph6K5%C+FcAu2+5G|LooV<-y)xrvg*=D+e zcl?I(I%ZO;LiX!@P=dtPW&D!s1e?*~PDd_|xL$(vv_E2V)~m;j&(qIJ{sIJxCIhhL zE&yJMDnq&kv|?#a&FF5ETc0Zhk{WlXK|RO)sG(Mt7=z{euk?eIXBMYkG`tIJR*Y9u zzE@f!N>4JRMg&%xgIsd6PiXYOFN-~w%bJ93>3vA-j+xgFsX*^~D8to0yPk{-MkASHeqz(6Hk1ZdRd zZg>#NwdOWExWg8h7p-g1BdfhP6T23=O3wvlDksi3Nle?_CwOKlq@(~@t6q9>G4Ggp zjVW{jpV@+r#zOmabRPZy2u({d6zxc!pz9HLUlZ*`R+T68Mc*#>o_!^bqeElPzp>}Y^ z*fj`j+QN05UDQuMyU7#M`7|`m5PxxLEkVo3y-NJ$q>bQy9}@?G~HKukgDQ z3Q2KZ<5?=2LCR7u4)4c`nmK+E8nhq2{t|&S-=6|ba9Ay<)18@}JvHcp&OymUP#Ad- zbwu12E@#~x6skko6jt4xOFqMEJ_w>0|GsaL6#ZGoA84yTbqMbYLEcK)IP&z_QL zkT)o1^_L>HNVTo`zztMA(hDT|A!vR(pFx_ej$q+K>D1KJht*?ELDcyCAsrtGzU~&} z*j9o$0ce}f&q?&y68ILd-Rok0D!Jh=CrhE2s8~d!W6)x0ylm9@UeERDkef!*_LEKN zB+46#*F6!YV}B@9@AS>VdsPI}{RQR3de0HZAmnP=ZP7`SxtQj2Dy& z-KerSWZhp`a@mg;9z~8CXt`mb$zC=HT)lDM~iExF;@*^@dnQ)szZy@l{eQZR--es`I{R)g;Ts}Nga?MbkG$DT*Z>T|}w4?`M2 z$$=uI@Zpq4-{J~;&)fqaS(6i@X)BXQ2jTk0tn!x1J)e4cKnQMnPLAP6C*va3w6um2 z`d&UcclG|v_l7x_KGZr~&yLjU_GQd!7L>@~z4X9+W8Y$~m5G#Wc!;~VLd1vYr56F5 z!BCzr>^eU^d8Dk&8)3P}@p-#JJs$9D+7ljIf)Q@s^`_=$L$|DYt5~~^cIK(`zpgtk z^go~9RN@JvP9l_kxNliGnd`w7cO{Q-weTL{w!sDGQ}2)V1y`SlDJm-JnN2mHWAxAl zY8XSYcOKRc?uLvSJvZJEo&LC6N2-~cmYM;F3q3DK?v{#8brn$@NBlCr3?0#hkRwWu z&ZViN@74`Av+6&}!p?VnOG+GdN^Em1zc5Srlny*#=jWGURf)UiTy`&uZoJASW@*W6 zv;zBaCY;=cwc9y@GcPR$l*q@Do)V>2qde-&`bK1Hp<^&#FC#m9_~1-U;l3Go>J^di ziFJ0(yHaGjT$*P=`q=n!6Ou!_6ISNl`{WJXr>Qe4XyOq?41=Nyy5kEq3b& zdaGh`Lk-7Da^ST`6a$J&a}$H%Z$%@Ds!apjqs@g3ph-ZbC_FCC*xckglm88Ie%Q!0 zT}S%FY3&KmBNg^uY~2xm@iE{DY?*bgBCjS|x-Z#bU(dZQrtfaj1a$c5S&*9%wda>H z(&MyS0mxDJe4d7d9DBCRp{%w{Y%M+s!|KfX`dQjzg%>YA2Vf}`0w@c|Rtx&Nhi_8O zQ>v=!-W7GMMT93La555B4SZ_!%se$TqF}mpx)XD ztHc7uu)mpCXciGbW24{i;+^lxpyo`*D_M*=+AfLv7tE+6(1)A>eRS9nGRwb_XYPR3M;(a*cx&dSF zIB++jwc-fMGit!f&dzRrrBZyW&gGM-X}Uj@7Ig6t(qp`eFB@iN6w-yBP?JT3Hiyt~ znY4sN4gcoXb4K*DD6n45fns(7#d09K{<)E{1mGmSMj{`8+R!8GmG2gLbm>+BXIjV} zQO(>HwVVcc@^|Mmc?vi}>RY_9LQzmv4Gr*aO#FQZg+6otjClL%p18}+FSKGt`jZ?y z31Y1wiDG1jJ}!Q?BPC6TJ_0>nSo8P0@kzEoDx9E`DYl7P=iZum$pk2Y)fw?wC_oNE zx3s=x`W;u_^1q==orcI{d_rIrDctanqi?0lkloGAT>bN4X{u?gc@K;5$D?XqVR`cip=* zv|0n$yydi@7kQl-$%TGTD4cKkE^#~aj9vU`?VV5gCe}a{h~}Oa6*PTWE`DN*RKEnX zKSp9n4g@f3$@U12;NV~o+dUW62F5K6`QFfyS>ht4?~=2Xi;Ta#t~(K6$A#x#ElF3$y^xawV<3 z+BXJi%`MD5 zs(B zay}WSUg0t-H0q*~EHnT|UOZ`_^Bge#z1>q{Vy;fJOG?q4E_1Zzol;GP0|HZs z%k8?2`pLt%M}zWdy$r{4dX*-v(volcXAx91`%CTq7FbKj;hS4Pa6mEGD9J1mQH}=Sd@F7MoXJ3~S61FXX6iuUr6>NeZ+@SFX#JO3_uUJ3 z0Pe@lo}M21qi+j*f1A9AE?K*%6BS2L{}+ z(ZHE`0kO}Z>cc;bUDWBlArmc+xUrvsn}Pmw!U2h@s;YHO$MSI%$hYVhfa=F}^JjFK zlp&i6Y;BW7Vs7X-kf_hS4&n~WQDg1kZMnha?srzf!y_)8IZIyF?Cr6ILMDpMNBA7c z)yMJ=*;^!z+SQHR5bMr@3JDlZeY=E=88&^n^?ptst<$?zQ z{1=U*x2^hYS7A512hIQ`a|{qDk-LCE$L1z1*G*c^Kp<^BD13Z~JU`Nt;z9~*e;W|4Qp*_{0R(w%XW zlNZn35pr?&r2|NF$GZz|H@e_d7?*Mo>Qie6-Cw(0@}hI8XA{U>%Auho*Sn3l_lWKf zLQI8!+&2+0AcM%A+DhG@YFqxL$|9NNK=xnEGmz}vw|{L7e+=TR2=s(GJ?$A+R8=MpG=NtwF~jilUx=Bzn6fbFv? zY-_s_)zsYB3iN)k48=WQH8C;K8)G2(v4q_{XXH&WODgYI|Kc3~Vh8@*6e+={kz$Ls zZEbDQLx!F*RGf#J&O(DPJN6t*045iLZBk^~ZF!{BqS9ZD@z0C?ypbif@6spBwyrh@ zaaVo)gjtBt?;wOTk8kIP%kG<@p&dayN3xQ8-_hW$jTRnd#TpZ&lURnE|t=s#Dc64#fQ>rnnBcDB98@K4I;K$lS-+ z1^!V3|1*EtB;T&yoUj{Ibo@g<|E{yn zx+gHto%S z&%bWY|9q6jU0nkBk?9#3F3EHY%0FD|d6`$!^xwXu-uz5~BjjbYEFSxbL zcv!S3sq=Wk_#7?Le&xUa!-6l7o*ZY|`Ya}{u>ONY=&MPk?KpWi2d$5h?D*=F#JK_W zs5&0-mv*WAnAfO4*gICDM_GEPQn6t5MliU9KuBB%Gud6gTh+4bQ0O{;W&%-g=v*cI z>!YfprCmd&%LXFGh{+Fi2rWB<%V!fujoumFOUYLr=R%gQO_!@y!F!>FYTm-is_)pU zhjjk0`uV$Bf1`Z9H1@J7JM<4%SN8f#eH}jC^s8Oo{L&js1-gi_ix!_1ejUy47mFgK zrxVC=Ak(fVPCTW55AXi5?8gu2<~J^B1~Alvtkk^k!xf6p%p0SAm(MUDI!xhSQ) zRL;@6?R2(OzLHn_UsvgiOp05N-H2Yr5IS5$V%&r*=wAbmzdw{-0|*>K4ooY#f36o0 ztTt^a!tj`@<1!_zeJ2X_B)sgX{$*bP#Px3?_*qL2=>_MrZ!Xb<1pc>=lNP%Oq?;jo z3ABIQ@2PKlmxgL--rzTR`uN+=HV!F*h5yJ!|8B$>eWABmEF4JK)pQ~rIZul%H^$H;Mm zow1EQW7q$=`TwCT^!k9vNQtj1#s3f)`eZqai)vLtZ*^Pat$)xjGYv1}QX{Fd- z_TLxg|lj|^P+Tv$M_g?TkI8OV0`3?_dP=>OBZLrI6z z2AYmtHI-}?Va%io_l)X5qZh($%_T^(EcMQqXZq)Y2Qs^Ktzgz4-}u(CJQ(E%XTXd4 z^rwAwn788&I_1e0dmrZNZDZ`L?hV0M*m)k1JI;=`U`A_HAW#F&Ge;aksI+oh)W-WM37~S_8|(i2{&mXRJka8TFMt8bw+jTsZ|-O zR!KIA?neRv_k{#C$G$Ta@>#d>y# zeP;s0ge}a>h9kpt!!D<@Fg&+wK=TG?b0bN_yV$+J0Hw2m=U7Pjie)MAo1EoME~?eJ zW9amj-Ne}GDm}%RsE$E0h~hxB^$3v1GRIZwtZzVBz_um%Nj4e9kU|%zEU-x(x~fY= zNAyayW}IeNZWUHA#}&eIthkgXxQvY0#Uwex+si#+z^R;$ADGFMBGacYgiCVE&Xn%9 z#CkNl!X9bHyQ<7Wi^gj{=RSi&OOzC2B9*`|^4IeDA?IeImRToagqlaj=NZ>ZGqVQ+ z)1tYBFw_GIP}#ah<|cXD#~&>5(9WI8wb$kIXhh-HM)O61x;`2TRWPc;S}WK#sZQqZ zAU5o*lVq@h#V6OBxIM`N*2bFP?G1Kvn}QAv0S3fQLt?a2mvY2uyKmtRI;?QZB-|IZ zQqWp$PQmJRLO-O!N7r&@j|Csm(zW{0(zxS*VOK;z#se3;1w-=v^aRc(C+G{^QArXd zS-HC86rGc)o})+OV+;bGxb>*bK*byY%k7^t4G4B23adC3=5wuAWx_SvJ2+Ul3u2gA zSmi?*J_Airk@#GD%a0|S7niNjAO+>{?Ck8Pq?dA#hFf@*w;P3p_8+&)t^p{J3Bm4a zJ4l`l&8 zbPiopuN2~Rb;Q{6`7xAUhI2_kd+B>tC>6cDbZ}s3_=Cmy_7v>G>9y|*qJvv%V%bq# zH>_eAt>U=laVu&Rls))Ia z;6o|KxnNum`dv->*1#oZgOCxMXo(<0?7FPkAS)rw3_)?w+I{JqJZ#xb{vQCNFOo<4 zM#DZ0RTsMpc4;>1{>sF3`BObrHXEZ(;Zp#oa8IB)mA3$c^`xIG7hO6ehG!RZhn9GP z-@^9B(FDCG$(xe)on^*;w0%0n@4BvQ-7fE0QRYt+T=J`>R@+z1X5&XqTl{OhjHPvR zGE+1asWI>sd%lf^V%Hn}FyG&Hsx^sLb=N*IYwGcyjbd%nBC_=I!?WgMD<@L;9hpVB z9qC8Fi#7{^U87}T`{>I7d_#F=Tb^R34IS~&L-E~2_90sZ9VWun#n@m=IBq=h+ukF!}^7`;9kpc9f_dknVm%R!7eGyhTpAl=%eUG zJ9fe17f}BlqIgh!!O;1jlee7~eKl;IX9ERVfOV~Q;F4aQNovyxs#A{M<8N3Mh%NrZY3P@@2UuY|1r!eb(|6(08&8nvOA|2PT%rDm*Io1)714 ziHz1-%w9=TvvB-ohgJb3d9}5HEV^_1`&n_vM5D>%0h_^&+A~3qlAR0;)RL({$n}UQ zc7jvs{QCy;xJJ#hE)G>20l)fkr_J#WiAqFN)j>->?4J?(mhuHO3a zAyIkj!)AW$_1B56sCN>4*aa58xn5ts?~!ZI2l5G`9jn0A-xb^-qS|8e+ysoZ0)Jv} z^SIEM$cxeI+y?`S9G{1CNKjzWG=GS5KQ?}PXkp)3d<)j`hXwwR?Q`}zF>@V>$_BD6 z=EQZz6d3M{TL8AyXW{808ed`EX+8qY%OBA?HDI$LNzu{sMxFkOnSrGLQ|x2viDLc7V^i^7OV7G2o>8R5(E8`*4@lz0g_k%p z={mo7*VJ>azYKsS4dryuB7&8$W-N>AeEj}rIT!YqehZAdCceASJsUNUNeq-5QWyy& zoVt2e)SorM?r@cgUC2))q{_@q9-HOrDcX@2t%0(Q#9%So6X1Nd&?k;#>*#E5-NE?A z(dQeE^y2mdCad|eW(F6^#1MDFqZX1}9&7q!^mKaHtPuxTFF>;A7={GIV2A_TKbpu3 z*uL-L89Q!7yj)hdwpk@*R^pDD4Uv2zP<5}wDc~SU!@`Kr5^vv;>9t5aX=X;bxx8r9 zA-oWWk|abHMX=M+(#;ddj{0(O;wWa{dVHbE6V)Q<74ka0#21!ew8_9LuEZnIz;D$T4}ICdB)P~pO=THZPmtFn|dA6KghCWC_)(31P8#Z5%vLKTIg z1jlkY^))IzfGu3WtAJ#h>yHR2HOKD2s8>pmhc)KoXZL6Fs!E(<2)I}55zWFJVJ`7i zxFL(W*@&^#YF3FX){2#(il7*sp~5!wwSUB>mPtT<)#WsM=lDOqb~m5GZ9}hltKSXw z^g7b+HmqPY3%^(zq7o755SCn3D!880WI=>}~#?RDfz3HcUESA2^nX+z7B#(Bnd}rDvUtx~u*P*^Uh$%*X<_`t}C)eag zv%USrx3N_Z3=Cu~T!)>MkBTgWT<95dP?Y9WQ;EsLNAD@k%p>J# zV=e3aMANKB0x#zIeYSy|ghsC>jWEwI(1a46O(DT(5!)sv^WES~y_m)q0pXs;M$3ux zY@VPj9;jA(u*$}vsw-O&!!wD52<@(58)NvYfyTK+5++=v_n^~C6a$CGs10n_S#$7i z-_>VIX|Z2#)ib6cbgv)&;5Ge`8A_*goym3WZ$nlDr5B`F4TcLL&SNR!F0k~OXUZq1 zb*{a^v)OGI^h^&1nC$s5zukJG-*kB?gT=|?`~5i9XzR|$CaeXi3a>gT+)tQu&7S3Yj_Bv`xr?_5mcMe;OOgmNQvaJfQei%(~$`^~yDTYTn zw6QW!+poVURJY9ySEPv8`Y=N2Wj7PG_>QU&CUEM2Yp;zDZ=1=wej}N}qUjn1=!uJp zcNC=-3~e{#bk5jsKJTvSdT@~0Cc%Wct^Y#Sl32~otO1-^GdPWnR0#9Q3CoL%q-74% zI7h92Fh$FLnCo_w)8&w3q6Vvk7{mrd#9z-KEnB=x8)}4Kd)($_9Bvy22Kp#TRU%Ed zxP|%!u6Ic6UPn|-2D844b#jgtq$tjl_@m^N;(2~g$#|#MlTd+gIoeNFS6>gq#x*t# z-i#P2U(U_7pWQ;F!{L8gG7!Ju(5BtuYS$d!|zBEiJ=uR=HBP$1aL`G_!+wW zrOD)THWO%sZ4<3cbDs@2;%TZ(f{2DiHgr8hEC0>#bG8wUr_3r6csn?_ zNBewXAu^m^?9yw&)d__2zO_z`x0*brhP3z5JP@8}N$l6jzD0dyY^G?xl=fmX!d(j4%r+0)Y2o#~+ z_oBr}$KL$HOP6cCqD(#lPaH$qOEj=z{0+K1hN`SKNfS3WGG+Qwu=B$!m70N6YjBpU zE0@Xo7;f-9a{x&&sF69F+uTz~6s;zNghJXrjgp0bYP^>eP=z3O;^sm>;D*}veYjj; z6R^FwZe~{EX%#zQn-E)W+_)pdVv-6S98EN$YRi>^xw_i*A)j=3S0L0ram%qi>L|+F zVqTOtu}#X#i~IzMiB(o5p(Mcz$Kpo?Y~nvuz4(-;81G{pUg}D2lc{i>!D;nn0AJhZ zBr~9$8|6PF-!lYiYdVumrO-326D~61?6#w}c~sSq*a_1(ppRhzdbrj{8dy!fd7FUF z8iLK$v-#+!o5$wXdbR6@;%vLQS#SnWQypc5V~0|#11xt`W z*iU~tbj+=y>FIpch{?zs0}Mk`&d7FI4f4c(wcf09@3E%k!)aTlkgM8xmda|KnW)aG z09Xg3o<-2eTv9X_6E~`~$+q{2xH1A5qQM~IqrJf^QfHyPH6m*pf9?GaIN8NV6fEV; zmshyX8SrcOcHVqR3~ZLGi2sz@-0OU$?khFvM{j?NxCgUqyc`ns{g;sybHaIJr#hFV zo%ALrm#>?YC!!idG~?5?KGr6v~!1of<#l3VCBaZpkXmekB_)2 zS*#A0GeIVu`ptTCJ68%EX+`#-yw3hBGRKdX3N@0fo*&RG!^2SK)2tSH-=s)LDBcam zp_wGgE&Etsc$iA7A8KQSjNxJpLu=$#479$kiLS9o!bsdSd`x_qogPx;L$-u#dq*u? zL(%SXubuvgV>6(s20i~rHL$u~`pJ{B&IBQy$2UyPtx?MsEVz9AI^_lA(coeHiz9*$ zzw75YQ(P1~MT_i>M;3Um`JEBK$@TY}#`3y^f4DBG9v)6^hN%uNaa(rOyS+kJIvE4& zyu3Wh#=j&8-CRDhZFziY70@W&T2o9Wq=k2LdjEV|C-@%vei zz@agn-Z&W>hz@EtA-bcTV-8!`1!u7)rTxee!5UKjfws#Qz!qUK?w6Ss>TxH> zo6d&!W==-b3t@LGw?4d^BPpUs{X>(_pQip!u(AAZ@)H%7j-kFI*p`a#d=f}syyUXH z)N);tdRm`r=y%sl!eif?UGT#p<$A=EPBmFP^axXAWC`wOJV&H1%6It7oNztmid=(eV$kkZ@tcX`qBuHwih%f&6V zN+{UjtfKKiAg$I7>u{cPMgr2L-qzt!MN5&5xi01TJ%)a%VBnqCl!bi+Vmn;rVxtp8 zgZ>_PE%BTPplEb)=S7!e(oEkd824AxSASYe=O1Qp9>2mdxSH6QJCr0MIG}iCw>I=O z3RKG{KeCHG*%PZ_z#=4}$r_)ojy-Y0057;8-Xy+d7+YKbO5IRTG!RTs{jL)x==ZXW zhSOowzC{th-45$Zy+;MPbpV!iJ5zO2y@ItYW;up0? z(?02NI1Jq)!F6H*zNBBeR-rb>gT3K2(ug3W zloHS>G8S`?KT#7)DWJVS%xtsHTl&7b{S|SFyr2+;0kf2qTW!6}ibOlFAOe6)& zsX2ae`YxRiIBb8C+a({qzgSJO54WH*nV1X;-Y}z1v@s~HUXr-HmTUrR?jRX5u&Yx_ zqRwrrBq2F>c7J#5>p#=W)7O@!eBNGm!>s)T)do)Q4hq3_N5d>D(~OcQW^>tuheg%Q zlDoA86!@O@km{S*F_?#6j3)-9wtW=Mty9Yxu5B62;e7GDk5MXNWmO0yw4GouP-EEi zF0)a4JjOI!*e|Nc+*qPp+qZ?2T*o}R8-CY{)e16Wz?2`WK}ak`xND-Xc9vR_ucmv- zDb>cd4HzxEzKIq5CN85jVk1g3xE^xI$YOvwcQE$7Su;l{kqSrx$gexd8crdN7NW`c z=+EyM8f2Jj<=c9VA2FWb5EiQA5$yn{xONPfr{FS-p=zGGMi!?@HBXjPuW9f8pjZE= zfYz;g0yV;*ht{;Yfhb&;HrKKdVE8N)U~>8$C$9pm0LN+Kp*yt~VwLM<(D%gz?TBn+ z9i#+@y0|Ahq{J2vXb5DgPxHDZ*?1PAK9yaMb;GrOoRL9mt`|dLBSr+4B99~<8{KNI zXC>aQb?Vf6$!Ds*c5>EC+s8}9e>`Ck_o#9C>+Hv>mK|Q+oXovX`C+6bsT;FB6=X3) zwp|i8S&^3QVwU+8+nk*!qFDTpMt{cv^%fXRTI2YU`OJpAKyk|p^Di=}lIi&_RV__y z75iR^JE$do7;Jnr#rwb~cs_On>Te)^ZwjYSu%Gmu8yeCX=}^Hicb}su&(4QGJINx4 zYx{`QW6*@=X9825^u;wqtXC24TwN;D1xru^Z6@a}T~0PSBOmBIaA{F3Liz~5ueC0d zn(ER1c2Y`>ueS?($xzmZ;(iOWzd4To)Y}_(=!EQ|gZK6H^For7y>~Ofzs<-`ZhdR& zNG~hPVSK~tmvl()NnW)2P2lrpCpK|Tw_NfFY#v0S9vG-(*3%ocem94`&9MiE3I|B* zIH94}$V`s_g}s3}9j*PDRd-dj)kAFE7E%(VA5hqr`J{*(Oz4`(#gNd8%ZG>^ywvI& z%tAWSD)x9HrjNLB`@N@E7NTFsoNe1^Swkc>zL{{`F^vxiU5UBup#RUdyVn;Zxp`tr%>%@GF>5&ey1}+yS)YNi;cKuD=-DV+$$I{zMLdZ%kLA=z z72LH65@yKZhvapUXdm=h|Jba%sG}p1nCm(y54P%97cNL95IFRDHeJ!Pn%e)MYx$D0 zO5b4GJH?*RClF+QC*YHr4JH6~;nZt3Gmjy5BMU}b&OMF5IiT)s9C4}z_0X!NI3&Qg z+9Fuu>1XuiFAIgXg8YrvC3$&|*ktjE0RtLY{7`d-aBipt+Txr`HF0V0&hq#q@z5ra zPy=@N$r4Z(V5x3iY*awVoFqdoiP}%n&@AYKE@YFizJLH2u6~WoL8s__8C5w!mk_e$ zo1u!L9BUt)4m@og44=?Xdxkt4A|s)dDi7Iu-!ep1!AZ5^!ku#AgMYk^ z%BiRb<@>tE#&P7Vh**UDZIQ~EcPi!KIfJ{6tgk-lwrabJM}E} zaQZp91z}AL3Mu&xsL4*Ss1MneJq#N+V(Yo|I+?F$Y&PWt$fspc2XrYg&Rc6!(supO z9vo2F!T|kb*_(;zW#ey%OB^USqWCC5@_#6M>bNG?wvX7Nf=3AhP!SQ466v%+rIa2u z1Yv}fG-Dug009A!ZUGt6ve7k^Zeiqz2}q8P4H(--dXW~S;3Ncr z<-(lQkfw$ogvNk*hC+;#u32;Rlg=132_66Y`J4#+QY-Uu+k62oCC}I07GC1M`_15Y z%Ws5YgcSUQ-?5Y+=;(54VtV^xyvMhMIl^aK5oxUc13C>Fstq?t`AXRGG(z5C(JXJA zsArP-By8#3$mV^or<7MR&o@_%Bz_(|NYH;SD)*YxY>cOFmC$e|bu8tQ@eBED^&QDQ zgO*RB4xft*>&n|%PoikkYfaYQkQ<2Y&rU0!{W$9eHg@zPNlS(Bfx5dhz{LcLb&;7W zB#MhnBoK#;vSYaVkll>HkuPr6eG^?iPQ`n~MUCH2U0=`MVhxIPx3uXnj_OFc%yU!# znNf6iCSrC;>O?avR$PF!33JB#*o_)1-67hkZ=(xdMNHA+=I)<8FV?Nka*Wr%asca@ zOt1nWeLA!7S*66reP4 z;nt%+$B^Gb57)v6C5jz@qs*uKv=Fr;H#uGJ7-|2pY^vEE9S())-|`c)2;hyt-+=ojhakbY&lpBnedLC z!k=7Qd?;h{1uKIFi3>vAv@h3Z_i2E8>1UHqWN9J#6YnM5xc10x?PwZjua^S9#yWF{ z>Rb;TNb>AJK;LI3se2-e`)eE&E#)NH3CVtPaa zF?h*A?b|HFkAbmiGa7*=(}!7baF7IM;=Y4hgYpQFTyI1SR2F@bLYiS}PtBZQk*r6z zaNFg(L6_cjU(FM*w2vF}_R6!RjC>Wie?edOij_G_kfr8R4EvxAPxv~}Oj7Vf$sIvG z4dAj!I3tt)Ac!*}577{+bG;=!F-tJ2U!7PDb!2Gw&knlGh5TeG3{Hh1&SVJpn3%*F0^5vhN3lddNfWFF0>KSoBDClt#xKzPxP|-im@{+&DM=N2=g6 zix+cvpqrl=vwmS}L%PdZP7W!N&d_G{prAF@nqnM8yo_5!6ST( zhg8QzrQ{a;G2+8}<*R!BS(n+$pLhVl82|g?H3?ybXTdY<;_L!He-`D1XZ-vQ{*}vL zRXPt*$yYbEEtTSSZ}L%w^q@r7fiY&UB)L%$CBd~P?T2`Rd!cWo7(w|N4_?%LDF}w# z$DjGyY7e&LSsc)`ItCmQ`a<*R`%?+?*B=h(E=s-x>w_xIB+)?qw_iz;;J7}KD3Hv7 z?2GK)Jo<_mD^d`RX!%gw+qX3BC7+6P={ePR--r@Cg&h*LexJyD@<-(pclZf}aBK#? za$o=w7t;MIo59pci}WnQYwt6wCZF7SaRQh0zTi1b$(M`;C-vf8Y`A-$HJceHtq9ScfniY*nx zf2_+%Xxj*BZ=MCYt(j4(_9yn8dc(%$ezE5Dp$}O>Cu@&e{H+Ae(MTI2d-joy~AAd~0+pA@R2Jryhgf8r*LJLle zEJt(e(x2J(Fj%++X!4g*=-~}!Z}3}Ft0v>%;(L=HXWlp~Q;rXu;QdU>|7nivE$;A9 zl-2X_R_-Xwi^jI@F4Z*F%g#JjI*7IASLWtwIc^ostzM$O3FR&elUf#+rp+IT1xxTNsDVM_dz&Jvl266dq>>ui+2ZH|D?76JhVzbY@ zUu*XdYeKPFwY~V@iVubl2+*B-$MU+U6~ikbef4$4DzT+H3`qznVtfAxnP-@|S%cY= z*gBK!SVh})sl0OEnXLK62TvbtuxM6dL5sqb%3LvFsvP6~_uo+PP6Omo-i(2s@Xwi! zSw)&Hm~@fmB4_7j+f}lOqT;gkry|y2i3f5968){@IF2}S{yb%%oKCd7eM@=6GYEc< zz3n~0LIE)ZJyux^9LkBj_Q$DP{eQra_Zgpt%~h4hJ$+$E)1q*Q>yM|0SZLs!hM9*A z&yHKSXhzBDpK&z5k=oe`P2_G%faXn(WONlQ4zNsiPf2zV^rlk92;1#4vB-ezg!D2> z`2!6h{I)H4jzpN$URrU@UujcdXEFZiuJf9UuZQDeT#d&+#Jd$F6|5zC6Pbs#);4la zmO`(ZD~;D3Gx2^uEFYhkW%b@a%|`oFoXl{zZs@{z&gPPD$sC<#0)yflaH|$Xx=2i? zD5+5dxz>C4k%rpl-6xv~2L;fGbRkwxEJn0!bk3q)4H@u3FxTu=eUDn|@~4c_s8Opb zDD>&}R8D2$vig0FK1@9sINWpnlCiBH{tc5(?~OG|qLWcZ)52ZG10aP#7HyNd9j!t-Fe_~-6%#=2LR@$B11(DI1 z)+d+U?cDB9o?At5tM*%`uK&n2^KExjtJ_ZPZg@waiowY}xerUZ>L?{{4M-l-fPQM`x+T^5TBSH?wml&sPx^h(Lqud~n{Fp~5mOinAk=TO3}{W!5Q2>*|Gh z^Too_ewp-fJGip7K$r;|c+Tv6Q53pg@f%vq3^W-=CR(oxkx0Q{78@~chXJhDIWg63 zlBZ{;Q7my6C=rcrC8l{74!jq?_CEW^$=SVZ2 z7N~-NnNaDVnHog7KCqw0vu86Oi zL}l~p6#dPmzwlb)J>Q5a50(1!g0SDbJdV68-e&SW7_umR;uw?Sj23ss#hatyBTuk> z=0;W;*IHS(9>~ZSv4$umnTwM?V`7-ni`c9~-hL)nVKQ?N}Xsf6;pN-KI^67$IqceDG zsY79; zzt^K3iIS#mGQBwflaXOQi|lZCTo_!_an3)TYys0F`vI4GsS|&`ley-~7il@yZP)-%*cY*Pb@YsM}(yc ziwpxNGK#rZk*BIpyNO)`Y$el0`n|SbQ8x3MmbjnQx0|oLIxjuc)!kLGsKXSYzX!}-JY-|+;${&6J!I4q_hi{VV{)>YY;kmRy-oA249n zerIXy{!k#DTLE*NgDuHQZdMg`+O0T5e=s5XU!!+9KPdxsLdE7}#?UJm2*@y2*;-3R zO4&Mn3H9C_kE=qY8oe$Q8!YB-6(!b_W!$4=LSq7G)DRuu(Qefh(A6PL>ne_{fJdS| zbW5vItm@@gvgj)|CRWEnDP(mrDo{P&MOSl_@CIA@Vd+L?2ZT0m=njleqbMH8&SErHR0-FEej?|W$2!PT%=V+RMzbF~LU!|XO@&=HYcu$ZmQurlHG zO~S}QygV)uO_!_%d0dOX7T%8zH+btjn#UV-moFB&a621x5HIDxA8aL-xY*nsp*CJ7 zQ+e&Oa%OYh*=|KAidk2D*_8@TU@#@?yd~fJFa#5wa{LVeWTNoR27?5BGSAo-SXgA;3QHn8IN<#}$Kfq{|o&9@#GV(mNAStx0x?+2Ph zy6avQuaRuvDc-xBk>S3>F8S@7<;-=?!tGes1pjYH*U$*&C2F2pdP*Mns12BY80)f^ zkRLm6G=*hsw1!fG$D`YQW6wQ0N|Zb0V`8wxOT!<>U06sQx1dyb{Utg5Ht*+|BB( z>AAUAO|+bKj?M;&`%g(c<11HqT$Aa4(x@4+W!C5c=gsKQ;FkhN!4^Bp7Ei$KdEJ+X zl451ySvz7JUN@#-(6qO za;5_Z$;!fde79(F0o4P^Es0=WbP_r8fQUjzr>?FKF{&sc*or%2L#M>k#wvY2jToi* znnwy^Zd*`nAHD#_7>HLb#ix;Ga!6ey1t~X)DhHx}Z|}ye2r2=ruc5(^=Q2RM+LHQs zl9M3CJm@^v?z|q$%@@QG$oMs?VO6oEL&v^`Et4NtdVDN+%F<-+$0pi$NS%@XNzR;L zxGoWYgYq!(k?voAtBz)h4wKR=S{~Q6rwd5?$m%w0d#=&LR1edio^P`?RFpz8H4u^N z^1kDpuoHP`69}N!>uZ9xc27w8H-*HK-EOs(J(inmy=plL&mOJ;#UY_bq)Z26J0$l- zyeLKHMhm~Umnf-_9pA5(YVBlIupq(0ube#11IC`_9w~XZeRIEVY*oG>ll#1mhD53` z+$2KafJE96?)8BVAP6ai2a6Qmfk3C;vYN)NX2;yU*~^X#K8!eyza^HqOAQXi%kLow zH17i~;*LYxt1>+K{!rM{?@0L|=4tb6ASX~oi@nWSi5E7pz3So!mG1GEN@rV}$B2&y zR1FKxV^Dl3HHJa&5SHe}GM;jgqBS8}qgy>+7V{_1vUm&5R5?9=++=8Dq zIOgyTd5+a0*%Godc*(`!QL?_3`j&o~=S;MMcgj46Tm(jW@Bv{l17qR%k@*1;txE({ z1dl2nS19&eKZ^~%?ND(Pk;}EuQe2+q^n5c`hgD^yRM_*yJa~|y#UMe- zK@$vIUhd&1D44Dzo-)H-M;D*7oin$pHTf*X7SUa*{Hp&ubdO@*%yNl`C!F?x zfVs8)$2Xu|J`o5pYiUbD?0PvK@R7xrre}!WGvuZ6FsZR=EtIKFUm?prw0o2b+G;(t9LoR4>QCF1Z=6x(@8QcIYbpL>^1R{Cy)Co zi2=Pg@DEMwNu0&?utoPvPTPhyomEmi1rWOHLX>So4|MIdOCa9HvMKVy)!&g@pt~u! zeHub_6bA*m^`NK>R1ymC0s%379gVB+{2D#T@+xv$t>JN9km-*y9v3ko4xY*4?6fQ$ zo|+wW8)*mkuo01vF(6{Oq54~S@ywEMY9|7B$D`Pn0pj_wa533?IjmUb6iy~zNL(53 z0dvL<#@q0<-N;?>?Yc-$yK4i<8Zz(Og3>hru}ENdeC5}LZhv${oTN6V=p$oe$I4jJrlxb+!Z=kXoin%& zN5GF|jG2yhQ}Em&rB0vcoe47QNKQ1ejVp_vmK}DK5)E%%k%{G~t znE@ihTbSq@<_;CnrGm5+4{{(lPzqX(3fYNHTo#bl%^~KO+TWd@en<`v&R;-`QKXL) zvu%xb_EQFGyv02qLno>3-UKx4>5)_E7wRXOg95%qBQ#!^7Y{8g7QP5Zctd-_IO{X9 zItC7tQlQssQ*C`Aum@r?kKGtfs1%c-y)I4qxW4?E z;(nFs;k7x-zGY6t)dL~J>xu;o`^=`cZ}l3)FvSSV*orrbKOXxdT9^LAoJAfV3TMjb z$z|)JKVkVfN?5){nnc=8v&$jR(_YkAn~zoH@HsfLu02(H9?*fCp8bHYY%q{{}Lb37)*I@ownHz1Sj#1j4@S7j(d z<0Vfcyd8$m&TT8j&LD~16fqY3$X4Av=pHdt4|Rlyajeeh4b~>A=o(SH50+`xef0@T zc^q@zlC}2eUv-D?aT4=oFR!5*JGNA!(L4!6pdFnYQg^4kgXXQ+t0AEQLg=!^)ErWr zQfX9g8j*|ru|>t(4YCqC%Bi*pT9%dWw7zVnrjSkP?seG+K}I1>2p5SG`nI??_Xrou z;p+Np`IN%TMedIB1#{-}IxHC&%Q7&&9~f&?TLg<(-_F)QkJeNQ^5ILu7BbK5!5?EA z{N0jlYW&Vp)x~phx+v0RYf-pv3xy3I%)`Ci&lhtRtre2}pjscc8u!h?+e+_0&L*bg zP`1?+77%$5@emB&Hi~AK5A1l9(No^9PoYwtx(5xL{q?(-$05b`kuk7nPTR;A6nB=6Thn6h6?4+hy;?Ij zv)4hd`AP2{e8|9Z>%x|MtLQ}Ns8|i z>D3U=4QraYN{gq9=cG^gAz%`iGL%b9d|}~MF`{^#z|VrUyQqUmx}j8FAds033^y63 z^%?d#6HMoe>c{$8L)1%+uc@RNFSbnaX{_icXgqY8%oCAoh&jlh!YX=NWh7?*1LR^< zmyi;AnWrI;^@akjYiczGTgg8=#o#H|RdgfGb~X70zq8w09#}iNqxM>@)cPl`(j+yW zc{}$(f6;h6&!PNw*Y<4fjcu{?vl6Q)t~m_op--X7l)sB)*7MM!4$Xd_#i2RcBJBxu*XGcDK>_vnvjloI4NV@M0!V56bkCGtGE;oiVK2H zI--sGobmW|bz^H(1gh*tunFO4dG4b#U)pWv+ezvwIl){5kgwY4<&>!FP8Sr*-w9|| zIfD99u-|-_mmSY*njn2OV`IENt`%uh3aW{JHuGZU;YYC}ngmhkm(R&5>uFYL zK}bZk(Wttp@2B;NNm>G*$fVc+b8Hc$j-oH4!Dh)vChga%F+E zym!;Ox2LCGcVulVT3tjuvuRSogVz*+N`jD`ZqIEk5>t_iz7?AX2Ly!&u!Sun`Cs(b zWaD;uWB%s-2O?9LC%k)Mf5m}sDKxa;>zqZNq_rrnndwtwiZ?rI&IU&wL%>Oz5U);9 zpHhern+_Y-oqnOTar@D#T!#IqDQO!gz3pJy)Uv9LWW)>o6zME!*Hynf|2FUlt1iU^ zD*Zkhe}C$J#8uiX4JVekk=Lb5J*&pc3R8|b%@k$2D{XC=nYkol+g+GKa$V|xC|X4Y z{ch9IZj*({gznS&_C!}=cFN7Z8#rGtDfg)0B>Ul4dani83|m=c73kEIK$-r0Le z+dM7{@zF(@7j|MQ(X!kxTP98!w=7lA-PictHraBUVfAj9#9mK0EYSxn5bW=@fQetqDM_ob5nzL47A(u~RIz{PrHgO}ef3*g-dGILO+n4LmFoQ3z zl&ujS+cb+UeJIqc%G|Pxr?Q68`wqD6L^#=QkDMquIM%UOS@63WO8D@(-{3tqU{W=}v)!$Yph1Lvp@Mt$=f|WF|~W0Oxz!Ze}ruc}ZdBK}RUK-8b9-hKoIS#JZZF!!s`47`vbVCP1R0_3p=i9=8ASG`hT&{;JlM@PlD zUV6XdCVW@h;Utn7D$h7EAt;%+4kNhmedTIKHZ}x{2eb?Qm?a z+Zh{MC=X6--r88INEKPA6Aj@&p+)0j9tvHUqyW=H_p7&~LP3f5=gdK=3+ z%_XALoViagQX`^bn@&^XXucCm66=%lrDK8Plc=i8lO`I!tG>t z6DH1hId%HshO8@!EP2TF|3Y9qAD*> z%0Qe-hqF1q$2i|Aq3y_O!-dgT1R^tZ0q0@nhF7Yidrz3U4kum?Lqc!uO%Rt5H^!2P zDOxD|yLv>3M=H&-VadFUJ-Zxz(vW4K-p&lwEpKo?e}W7A-< zT5y%ssrDu|k4st6;@!rHTAw|{AAuV_V=#>6gt<0?)k6Hhqlh<=9BPj#XgJ)1*~%Of zX~QEVWn+!4@{=AIVNKd@WioDPAKn~?Jh?Eq6U9qw#kH}CDhjIP4by|SS40BZ$p|JH z8O^-14`vBIa~+9uV!kmpHZL=xlz3b63AzW3a1Blo!`6Fug_EyTybh~KZYq)o`&xJ_ z&D^gt3;p%?&k(N~Tu41oNgS_DEP9_PE-*PFi8OkNEErYeVaZ?Y+}nvD-6>!dKYbFH znZgTwXtBj{tn)*KJ99xnLSm$ZievppA^EY!-tp1^S{AE>P*!pr!t7Z$U+gKaOa0i4 z!cv~tB+*k=85eK$Vlnlz**Q(6!s*xI^#r2h+Iu2dR>#2;73uZ_oo^;c zPD!y%8Y;*w+1jV|W^F)XO%XrYnLQr{a`e3qDsgO1*?JF_Ik}|4;qdocB zLRel!4s$-Uf**XmnuMCO*Uqr8tMj$G^rzsB?|o@X?_~^{d=sd}$P-fRBczhYcLqU( z-97hL(Wu&$fq70Ue%Xa>yb24w6b$K+<$R)|D1#gCo46IRx6}+LUBwZ*FcK=^pyRYS zw=$?%w>;DvDC5t$hde?HZv#nZSb-7VPQoY**= zC5~~B6jyfmY7p65Vqu|(x$BNE#pX;G4DRAQun8WT2u|_%`TDTP3+L?_8I_;z&84^jfHAYq95iPgDcWb0%VQ z;+bGa_8>iFdxmdPL}qb7q)q|6R=)yQWW>V9}3jAxn2+ z&UiZY`K7QEm8FyvnD|~fSIqO65Ee0?6MeLSZi#lE%}`h2)iZEqe=NC8q@nfn)TEfN zL)9tC`PdQiS&D$_dY_lFWuMI}6%V|E-K@HNmH(79uKwQ5-et-8q0;oRrKyg@lQ>EW z*KwI=*>fKD2Ey#>raUdiyY^ANkL)O^YSvD`$onrP+2^bH)^4Kvp3L3^RgPv?@AnUM zW-4ZKLIinjT1o1dJek3++xKXbfC^L(`CJ@+nE1s@@1;;19X#NUI8e8b*SMp-_9ofY z*h(udb%J0h_g6t-P|hjM6(x^uXw!#xG$~ zPvp%~))#RnX70MrKnz-EaToTE@Ks*Q5fHpD11BGT7bU>!Ho8&4IVi&g{Z=41*8h_1 z$F?ZcJrKH|3-F6gl6S4Fl=MEh-}i9+Ajp)X13LADv{@5U7nn$y3us#y=!>VzhqFZS z>2?{v6tYra5Jg3rwkzJ2nsfzm^vs~oj(_=BQz|IA+%32qnCD_QKnB{iqu1`}iRM(j zpNbNPGwsbWaJ?zHHa%l$`{he=e@Q79@7%;0$*ql8cc5hmS-;DDM5S-sotUAv+|)6w zf68cQI$*nA7wgh~D_hFx2|ah8Re!vX_9E|*{1KO7jdAH(3{%%(p_4`vRtiSL(Bk^g z29<;93Ly#+AoQhHovX$%)CKBswYL`T#|PhkPZX!*4HsdH=eip|edS)06 zbiuOS*$p2gxUx!$x&OEy&Xk^0NzcLj3Ix z4dV|z#dL-HTrAc+H*c(x?awY>#(l%U4upRe+k7cCy-97S2GnTj9 zMWegNQA-teUSeT7e_?gqVJ^gT3z^c>r~dwG+|_7CO@btoan)pV%L>WdQsOkl;9;dO zD%tR|&!aGA4AyB;G$Ls_FEc`nKP`m+N<8VB4yFcq!Ym2MXfYA>9L-2@V}9v8l-MwcCptg&G3m zEA<}ED0uP)4(%UyRmK*(c=A=~-o(xR$!W8BkACQTUkwcr0Zo+E=0l-$mhW|GFZ=UL zF&&;_9#iu+^Z8?woyHC0VAOi$Hc0LK%4SC?2Wkl2I<6wkl0@n1Lt4CLDQ#7ot@3Dde2Vl1fMeaH}E~BQ<$1~2%c+*Km4vLUE zT4T`AW9C&~Tss0NME{P7kkhjQGPuaIV+}x~h^sahzDQc)%r_)gl-!`5iB#t5hZ6e8 zIq!|8SCgO$@0znF;0#RHf=m<`>muRvrTW0d^k*pe`NalqF#Z!OT`56cCg{}Vh1lHo z7e!p3b*&$K?CWqJZCP$D3)+{ z1q}u(`ITtjEE|o&lSXy(3Pz*^B|8VD4nA+pYTI6cWcXD^l0`>enwBz`PLZvSA%Q@UgE|fSx@9Kn z@HC2dzsInD+f+yMtCu5Bf__^9j~&Y+)@c{mdgQLJsVu^+B*qlLo2Hs9g%XWUK&ne% zcHx=p>Z@{9k*e9+F-l+0U)Zp>G#d@{*@R=1*!dThVYpUk=n+~EFf z4;Cm>jpz|)JXL{`KA2p0?{Tv~mJ{#k|=rFigPgVY69cDyXl3X)h)|_tOlp~VtYcOXySTa`b_`w zLWee{3f7Rz+98yaFJRgcYo~cKqj_P5-jaa5VQ<>;YRV)Trv$2Aul4WMqo5)0TNbnB z+OzL@jXEgR6K~i()5xw!eHQIV(2BKJ{FFUTLK|&U<*TnCdjOB?z8;kDmkkh2r-8Cf zdl=8epM@l%?akwNnT9{rr`n96pSZer^Fb!k=FA4Hz&tCxzDV#87m!+Zw0XAANh1pu zYyyLlFtJz}4N;wKgezcrsFsXpyw@Wkv(o7qTwkA&@38zyZ=-;}Iu@@}U~5p3qPYJv zZ^Nm3ijSMS%oj&0)3O30ahxgvB9)_2W5#Thr_%d~wQ82^Bw&V21TG|8>zjSV{D?ZE za>g6nL{u)Xo`#dO&jZTVqBN`D@g$F#_J={GxIa2z__ll#T#>s5y1P#Wwy^L%D=P@t zsx4RF2{UQjS87PJD6sJu$2Dta&09#245}RLD%S3H0Mm{#t$9^bP?_diSrk^LYr1c; zmSV3KMII3HoG8p|_VZdj$k_UB_JfL4Wv*e{%)qDB_`>|wfK#zNOiwB~X*U@)-FeV= ztk*?oGk;@Rh{`4G3OFa2)jNOMQKoBm>-Pt(W+S6{~{amcI%93!fOo+ zgSE?}*eds`x!Y7Jt5^Qz^Nr%Yx$l?1?N#>M=b;l%SH07-L==H}3sZd|t0!v;eBZY@ z71v+!d%*|{74+8hH+xTO-H=mkJm3uf9f$TTypYMM)i8%kCy0-E1A|3EYM z{q04Jn@u{@xgN;Xz)0&wo!lfzF|9oXIXR-X;xM;Gim%GeHXv>9S+2`I@^o}r!g=0* zO-)=Mx;{~Pz>+p3MitD4!cU6Sa>@|XF0k15iFL`qY}&SxGUIY3)Ha2zEUm0?0$0Qx zAHXVRdi(G?qeWSyHDYN;8JCWZ&e#Z>%xmy)XJ-uU=A>49Oa{<>(W>0OYd65b6k&Z; zXQT)Q>0B|>?JqAbonF0;*^=oZ#?c~2sxI}C2O>qQ7Vd-$4?7a7qF+muRxBO+lr2RA zx=$?wfop&aeED*dPyCfG*IaoG!>z@8mU%EseCeiQ`FDFN=T{w3Ovt&YYLl7)jD0eV zz!YrC*Gphqv%V!0Hn}r)aS;$&#dvi``~%A$lB=(%mcTAsF53{c&$GJDHDzPWiJ);M z24LQaJnPcAn#6JvdN?2M&8Zko&j(#z9k;|*aAJFh#E@f{EZ!DAt=RiL8o*`2EEyN4 zwvNRiW>X-G37?8PJb^lcBj|M@3SE&spUi6inC)^;&%`IhPLi=4Mr3`(E+4{R-QJe| z-26xs{Iz*aHy6Ze(dzDsev!ldq18&Lao<8B_b{(6;)+3{2BT7sqtdAQGl=w zN24NLSY-AxO0NQ8=32p%?dk^Wd_BC#J@(=6=d;@yc|B%)e`a!3 zLh8Sx^j%EfviS2pAgz(2OP6^vxg|2^lSWRKS7G#=%8ya2$$5UKefvNS?S(gf|NoP| zvkv>Mw@%7RjHdeM9BD0E@DL&=p*EOrfv0@4f^+ zJfWz6l#Z+c;GHni+E0@^zf_06d0+jKF}mOb?@st{VH!b)w5HUuKJGmMV{io&p-T6ll7k#;g8*a znJ*_W!g=j?Q__2WoLN=<*7xDZo%g(M0c^Q-Y<1Tb`0uQVxA)qb2vhb0LDb*3#`=Kb zJZ-1r33Z2e8H6uGM-Cwpb-_8`XX6*4CUXam&)%}Ut!{&FqWd3`LnqtV22QjQ zjv(>X5Qgtk*?B`QNOlvoCD%sp5dJfO|BSy^=aq-3i&?E(3u%uuz~p>j1x8brpgzqc zYR&u1@sP|{l`FM<`9DLWM(I!bs>Obfc3b|c?I_P^nxIPVG=1D_|3}`gX{Y#uUMrsS z34UP`|C@f$GU<0GC}SG*a^@(t2BFaQq8N zzMk{-BWm3+JOwCB@TWldPkpU_SCzkHQ=chY6H)005~=Xiy+I)+=RYviqsE$9ZO7!! zK0EW%V8wshH~?g?PQK3!rV0FZw+22wC4Ua?r-ap*6c6s!2lm^*zI>v0u=2kE`&m!k z$^zbqufpol$5(|Fa!#w-qvWxkG4{MwRPQ2iQ^W#1q3~|Wg zm*4d!yBxrnTD>DI|16FDM^?S%^qT1ts%<8y=eCxO=33cJkETkoShMUx^Jwbx^Em!7 ziEpio-6EkT^QWa?C(SA>EV4^-fV`BXuVnW%-xy~3*djNpb_&MVC>^vdi3nt+U@x|iCs-O*F()ld&57sLrGUSeBGkA4cWigRCq^H= zK4i{Nlg+{@J{CjC=b%Lv#d{+yccZ^`5%>y+iG#T7-CYDy^YI4`{1^AT(IVv_wPg~; zz^R~(S?Ow2B)>Wcc*|wmh@jB_4(SI-+2_^FjMF@=7`qB~;cXMOb66Lnquk89A~ZQT zpwEpJyl!{@H*U+y?05t$NE`nWUz2V`(g#I5hVfg1E}f;O&i43@M(@eIOzq&P@cTDn zL@6{#1Ad;!DwxN>$)q7HD~{}Lc}{;&)roBvd|5-Y{rfYJfa1B#f9zTxX;9@$J$2E*U40r@2t#!%&2D!dW_6U-9v(44S&vjN?{^>>_+-zV- z=bv~3jPa{~#QncPKlJLnS1Di8=|HS_@zW4hlU_$>>!9&CmGI$)T`?y;a|IyQFUJc1 zAH1eERMm%0MH2#%Mol*1d97~GJfYbr!K_xyyX&|^74BuKsAdL&P8N2iE`Dlp z|36j})iqq;Hm&a}?vR1j#0YWiSX#K*@POt&IP`_4XRl0msIijaRJ#;|UxDGrp@6|} zebL=qGu%EDKB6+T431s(25)=>J<`<*0}E3ZkpL53A5fhQ60C22_{#fFt$QWmCB_$X zB7svPbtcJ)Ad2rD<}@BA#v1j^E2IE{aR^bHx&*R^X75RW;^?^4UlaT`UKVmkcd+T1 zZqCo61>aw&xyVn55KJ9chdia+>8JziI2Rm79&d@TQ>F5;^<{wIv{9excl)$&1baQE zf0%5_{5fH_$@#v*J~xJF$}Ff9cGVXIbP;*CQqvwul*9Y*bGs!|3HSoM^%FX(b$@_3 zL_h5(3z3I(_qvs&vpsS8b=2>*dgb+}?1cG*{Bc1jRMHYxRMfx3G-x!zqL54iT?d;L)E_k13r8`;^4Ga7RW+{#V)n{aO9fH@SOIN?|=!>)*bD?Q%VqZeOvjn zX_v(U+ln~(gzsM{%aVnFoeS7@dGx={1+%o5k4?0iKr053A=HcQIW;e@d<@K)u&Wzq zEDLz2qk4C4QzQ7pj#a<>9ng|TQMVz#Cgp7fkYQ|bGh9Vli=Mz?`%a|?JBQoUGVa3n z{gv~LI-w16yMFw!+;m>`zMH+|Hv>sOR^9dB`^a!K$KGg!ukGrfx3Q%zkSpo_0cyE8 zK&=B_M*#QDVEgo^Z@Is#8JX7M2m;U6c4GHN`|cw6vuOgclEp^R>(jC0`DEn?kl*@6 zF)=E(rfLFQ>NXH(`Q>J)KE&|)2}?4QT=Bnf>CFz8Vpk$SYSE6vz^UE%+w&2z`9Ztr zd;o}ALG~K;M+o2X1MJiJ{pwFxF1@Y!wc{}?*~f@7v2KM;yh5bQ6y`&Ppby7qj#CMbnjYrofIn-8X>2xU%lDlNS$7t`D{ndVn~w7#>KE!m z017rrd*yyff!FHA&jT}F39B^om#AU?Q0eow?EtGV+tGwrkV!`vQ1mDi5?3h|F<@+> zc5C&Oz$zt_L%;p~B9%^8)6=?WLSEh!`laRELsP@`VS~rsc!9e`%T#K`MKfj{rqnqw zdSKJmfOCr4^MBpR1&1tCA-}|xIe{h-eLA5}DespIm3^CC{RW)GW^bS=-RiaxYMf7Q zhaVy|ww%FG*ACQg5)#*^89 zC%anCoIXh{pC1Y&hXhz?R^_&<6hrl&3_vx`5g!PT8Xo`L5mM(%u;KnG&wnMq6N$R? z1nshlKTw!bm3_UgH*4`P#V%iv;eZ5VZT6HB_h6tmfzAII8@(wH?Gx$=mvLplb`0r1 z+e;;szc)s8DZ})sCsM`LI=?h6zj-}o(k8e2MAT#x0Pjk?+eJm)KDk!_`7cXPiT^UU z@DjNASwOD-G4l^tlQ5#P~$(?X9JtPY+L7);{PiM{dGvohr3NWSiR_)EY`!flqA7FPq9L=i^ zkkLMew&7(nm0Aj`$=}R#K5E06@U02@*OsaQJdgOR4dw5jp6%LbUIJX71un~*Jp3hz zKp8$8SRwnIsTFxNxqn0JDo{x*psFN#3Ge{G;e2@?zcrC|gZZ*e>yX*#M(#}bmlVfW z|7|fp&WnvD_!wKi@Eoao7kTkEfLgBdEV%E6`i=iUZzaJP_WajD6#&PLvC#))^K{zP zxCs9RCU3Or=N$f14Z1!BP#x`x%zCOQ_-`ZaHF^rz(8@4sL-M!I0!qoKy_{Ru>CzUC zs5Gl0JbX&FeB6tWrA|!wCjp1nFvKBrp2{`+*tmy|B4wxkc)ai&OVo- zWCfolFg{t2fC>S1Z*1nzW$*f(ofcghe&E+T?F5h{i{Ik?!}{;&i(1#JZE5v3`~p*! zF-IamDCa9h9#8vhoza`r*3woPNc}pv^6II7{9$AH$@|15I6w_p4H2VX@{|4w=ZBRd zj!qZHO@7>n4GUm~{21Bj7&=Dj!+Imo7;j03G-k94RSpHzOv8v&VbGYb=+ynoTlka7^_qbQX`C5=KH?Bqa z4Ny=HDQ~terrK-3r^Fh-5(Nfgq+c7+dkoXe87%03iIdOszoHok8w9F&m*57f7h1P$ z4yMeEf6A_LiHMK=!Ghm_Ii>TBZpR-A?M{@w5#qJX!~fWw4mB4kHT1{&8HeB1Ws#TWoh9AtKeqipTNk~XwFY^@t7ni@!SIuzj>qt#@ zIjZYgt9U$9Qz;cP%sY*l3lrVURJ()<{GEe)sYJF*#&6%qf@`t|Wst?Hf<@HU%uxLh zfN(&|yD%z1e=7y^U790^;FQ3$tY7~5Jii_dkPwd*)jdL4s1ESXl|jMTq%fYJO21y6 z0NcgSzQ>kJ{z>%)zMIue7tx__4;54aN-+ji22xy8)m< zwcv{^zpNNmU)f2AO_&s&sW?itQw&A-XP(V6?qN~j_60e6n1i2Ex)Vp|;P$)ZTMYy& zd*Sxye=UKvOP@PCJJA-nUCgJpdDm%#1l#`rMMdPbmf^VD0GG3jq*l<-9hZa13mKsn z(hmTDdNt<%yLz4HpZilB!-2+KDL@FX9^=@+gg*`iJN|&M^FZ2ZzlB7?e$9usbTlWo zK7{E1D+#JzWu=Arvr-C1U0p{W_woS^xH1scrZs#5Bl;?3E` zRIJ@CvFR|``Rv$|ZEtR)t~Ws1=IrifGq5!G-&3FYac`4>XudWQSj1v|R_s-Y4kl z7oCUzwRH9W@Kr#%dWar?zK7}N|M7LzaZR>g{}Dk%R4hP61SJ&^1f^ppBB@A;grqP+ zq;t>Yrzoi)or031J4Q&ih>Qm57$G@eY);;`Jn06Pp+@VC&C13BEG-{83f3#N8QHNzf4u{mDd=DI-5Ct8 zpppX%I1>}MyA=Nmp&<(h{=kR!PYl(NZ0cF|E7qTvpmY-MxkyUNgVDg~fA1C(*p&=5 zA0WmatiZ5k{LQdcUQ#M^x&ITqhnBV{`tN}|{PzRE(mZk3 zrvnjCQiw@^jB*`R3C)dozd}Oc2pyvHh@Leg`5TGW#sV#+PdLqRZuc7>O!-+gPK*-F zG?U8P>`&{0%@1Yuns2cmT$Hba#E^u}mzLksl0QWrzCiScw9|5|c2(*Bmh&C>sv~?s zwFD~jpd%Q@>eB9!Ut2n#)FkR1T%Sb3vpVtyLdSED-{j-hLJr=H(PyRzJ28+z$#|2Iu#h|OPT z5sO{d{@u}j+5%2f>#5Y2By{^lAyE00q!1N*XTTV(Hn4@bCtV$QiZI3LOX0`32D62e z*#luWU6wq6yS;Q;L}*Js;0qU(@agXsF6#B|{Ozw@@Cok`K~=1a{7s`CREk;fOaj=F zpqi?UMZee7*@usGJ|?->LG_qT9nQTLU*IHvB%uXg6on{-M(xs=kR9m>-TuZ&Bfttp zU)p>gU#i^W4Xdo6EQh)AyQ}e@sP*p3oBxH`Hz~N!d| zFLvLJQj=86^W6kz>jrt!PAt(DH-vfkuH8~)qWH^0fPwZ=Q}FQ`$j$-Kq)N)XYB-jL zW~YIMXA*sYy!7<8UH1U$qBeGSL1onDua!}W{Z}4|eCz~4ImVLSXttt-ROh`UCMzbn zb9apj_mIgzeE)tLyidYiLUKO^j9h4PntUI}65mRimniQ#miT?5uIqGS+cwv>C<89a z9f>MM@w39NDh0^m20O^nb6M)#x0*O)Z|=q7$Cy&!=R+zlCtm@D4f%TJFLxND5aoC` zXhgjw*S}rxag(*Dg`Vab$Ss&D6ax*Cv(BJ)8AMB(LxsaZI_>a-7Aw-{Zg1mrHV8Z= ziq&O*oeUDINYou5VYK0M$Z80t^`MkK;XLX`(O4-oSKOZz9B#-RGP+E{CwtXB;~_`Z-VH17pWY_`UOdBk?ur z3p!j6jLhe~S74dMamh5hRuzo%h@Kf%HrR@RW5vON5@G}Keq#2 z%ehtnELkfr^yoPCL@WAX|NP*DQxgw|LKSEVp_78GWp))?!tN31E``u{>Nu-B0x0Es z$`Tp1GGbY87UKNUUq?$VKQS99O69Z>H-C8Jk#onG)yPUG7#Z|6)*55@J~tIKF3cn$ z=KX5Y`mvs$^4k-HYWljf^*S1;N5*GPe`u{!k|R!_qJ-g}e6UWK%92&Bt)@>s=^3AU zO}C0lt}N+X5Mkg`*XDahiPhA&k1QypB^2R;8Gce=bN9E(9K7_v!I_pDI-2=^tLtn_ z+b@!WWEDL;+VEOkIst8;l`+{Jc)i`LJxU}0GRAuSBYib**?8dF=N~4XnYLaGa$C|I z^ulKAnkyl|)3-%St0(nP-+bj51w zqnb`3XQO)EuB3!u^O>+WKV+C%D$6#$UaK-LJ!$Dw$y*KD9nuP%lbx`{CYQ7ZXf-3X z;9OrWha|DJf(WlKyjVyP-q2!aq*+R+WU75BuNiC7>x<}t&MO!)WW&oj=>dd;yD3b5 z3{)>U#E1&3=GMH%#U{~)^xV1a?VeNs9^ueedy8EykM}M3@`;e$&Lk80{aek>rD|vN zkkg*B!$i4mpboAr9mi<|HmwvkB!Xs{(T9Dd`t~mC=)ne+^_j{G8^}icQbJtp5vaVp zZJ$eJZstZTZRV4q%`X2JH#^0%CC@s1AFAU|FT9|#Bu9HKy?Db@yA{Y>#KjD^jU^q> zR7gaorxk82sv}Kr(e;Q2rigE3M@t`Iupgkw>HalG^O2~8a@d{J_Ivb(4~PEtYHa%12-cBAQ3$5-W6%y2eg z+CH|Z$h2I5wIXwHt%c^+!b;Nb%bpqCcYh%MquPl4e z4IH!u=Q7+v!Stf6v-^Vpr(@omz0qsx5zea{Hc_lfD^bznZ%|9rGPUKAp8;DqMOkGu@Z%TV~Yr?NlDQRu_->NiG#2fLxmZaw`(N6L9LQ%={M zAe5^`C{1y8H>FOsUN7c-EjzHVn!{CZy0OL#kI{D>uZbUbz<=!gBGJ(jO&`6P>UZ1F z(69%kc85P0ufX1Qh3e#7vD-pp!pH+nO`VmTPSzusA!#PX-T?0;`%M(jy|S1~=4h^W zh6nPEp8IFJ%jX+rrCJPje->h$Sm{aO^~)B}ungd$D#bM=z{Ke2Lh|S1DQt5WGW}dQ z8~k6mxk_M$Q>+jlbJHI#=a!U+<0rGroO@aW%NEjNF_X+n2gtkz)aap-`6ZM0oE288 zf_Q~)qS2N{b(chy{+k*8)5I;3lV^>?6aNxT?x)z0nh^T*x{nl){nPv4i=>EhnC@P7 zkl`Da59LlU?7sKfWIp@_b*q$+YPvX!d4FEqxx<2jA6MSeDMX0V@)}UU*}LziHyNB_ zoGco`&A(1g#xu8ZL^(?Cb8z`RAfO#L+Ky|LC}ycIz`pLa@I|K~vxGFw zjy)0U%C%XDz2nB4mG16wIy>c7V?AJi{l&|=7_&}Gxa%)9%z7V>NCdPt|{aS=+6Bd zNS>{2J!W`Mh1uQg{enzyLtr61hw|`Yo;SH&E3tc;3TxLNd=a^6|kh zg)?U#xCM)nZsfp${RTXBv45$vk%ni#{H2pz@R@%b4SZJ6sKX{nha5+=nV2V^-8pV_ zs3DpQmWp^UQoU)lUJ>Q$Uc1rcBj4O+Q&4DGI3SmFgKB?jq!ce5s`DepyuVaHR8;hf zlifM&L7AADWZLR%!PWj?^1YVo`SIze&CO>stvly(B)@!$eL?*!UigZqhLbhB>NzDn zGU~K2`V^_`oXI3HKfCInV*Uz*70ViK0TnM= zo^sBI@|>#K>~7@_Ez5j$?~70$`sq?zgywj2c(}6KIf$+SnO7Rabv|@3!|>=_-dmng zQuir77Z^_yoCD*I-kU9@KEeyI<{Gs+{hnKY`tJgs#P)`fQ^b1QJ25Hk?BX79<|e$G zX)>Co&gijG>CMi~!E{p6>;{g^yI8BQMuiE$1GTiY68fc3#w|35QyFVcWnrStN-`Ig zy%7nm(bzKzVoz0Xm=6dv(o8p>7M&T+Gw-#Y(*xhY5en`1f$qq0K^)$-KnJHfF?k`pyqf3FDmq^`h5u7v$Va9GCujuRjofOQ zgHUgc4a`lZ1{Je1*Oy+>*M+&hjEag9b{jTuD*eEB)>smE@J6egQ;|uU^s|>iK|!Nz z$gI`~LAnb~4dh&Es;cqG+S@~+YyoVi9Ch5xtwk-fT6#LTz7&yxA|OY`U)G6KTle+bbJsq+qlU=o`R&Ts$X7h6m?L3!q_ zb@Fs_a>Ryi;B?7Mstn)AopfK{KJMo#DwS9VLtERqQ;V1@XN@gd#VVgZqU?nY_ku3f z%12jcCRAx7BQ0CnE+1&+?rzdo6p!pyBP_uts_Ab!I!=u_)YIs-I4bEYotj-4Z;6!H zSA1nmeLpucsnotlLPt#f=|&!M7FmH{D9zwE4LT&zZdXx%g~Z(PrINqIC^Y7X7FBK? zmmL@7A^TaU*miVl{HHlQ_m5IKsvZ`2az!lLIP0bzv;JE)O7oV%_NVSwQnIpq*uBOm z1IJsX=!{#m;}G27eX@F=tucq{24QZxmh+)K(sVQ7$lK1U-%SS3zQp$U*H2@Ck0)mt zE?{i;$aT!n1o`ej_bK1nk-ib+fI~$6=^OMRuFc-+U8)Q|gOc~Z$Fqjmf*G1Mo>CfT zRJ>dtAesHQpnz~0mTz9RHS_n@(%l!b^N#Jxv%5DS%uo|?-1Gd|Y>KlDdI+Szp;)SYKir$?j|#RQ&J1XDwR> zqt~w_O$-J|8M{e1yS&Fu_%R{eFcO$C^2<8*vnuxjf>Z4?M3xl~?}29qjEEdGrtb(= ztLhsOh#S*Upesq$hZ)aoz%bG{rY5h3TQ_Kud4)r_wzr2&pOYEDE3KB|>FUBLV|Pi} z5FJOBqYsm*>6W||1mExSwl6nvC&cV^oWY)P93vMOvs%Un+lACJJ7*SXX<4vq)9tOLN7BuE{4eFY zbo&>qeaio6`vRyUgLYea3AjjplW%eZi`CkkK84VrZH8DFAy35!^<4Dyc2@aaRC(P1 z6{iF*2T47T)bXt#ugW~k{(&UB{*#|dU&uN-R757`XJ(p`XjJQ&bR-UF2ro_4>F8%j z*p5b?gOx4_Td#)o_e-=zu74_A%*v|B8|anl!UtSi?^cl3dk}yT7euO&;nx0r#+Fc2Oi^Yc!zOz#Q+jkROW3xAJUICt}LCuLMMLHKh$;} zX6b!MNBz{*#$;;WwmaLahM4RJ-h4t>KWt?Bq#p zW~j8ZbW75b`%P<65A~$e3sKKn-;^aR2ro@RZ6fGg9(uD7PL0k>6{aU&#IPi+V@`w@ zr%PZ9Rgl#X)Ch+{O=syse_a$BH;)trPHCKFsJ{#2EaYFPolUXf*=G)Zfa?h*LnwGG=pB*{Uu!~o8 z;`kkYLTF?^i$Cdm&)Hd;RmxSZ)LsV*Ne5PIp{x{#6Mi1<7i-g}D@P2%q8J_L?kWRPFS`(0KGDjRUJ{^JFC zAGZbcmqLvHUwP~X#gt^gjGtEj4Vdy&U0d7~Qs(hv)FTnb%EYgoonzjT&vK;H`4i&q zusLs*I9l^}%PO{t3Zsm*L+oG1RSMr&#RssU==qxM&$etfv$z;ZNJuQq&i5Y&sW64; z@;-R1TU+cyzwXcNAX(N)UZN1OnQu+R*`em% zc|$bnBfefE`N4lg7ziu>^a2M*gr2N^V$pI)f$<3n1avvYv9aQ&P~fQNKTJ$Uc&{%_ z#OE_fy0G%ls0uqyKZp>v;%5wE%m@=SIsihz7u4wNV6XVc5$}_xp0*&@ucgRL#3GD? zzIq732m|{K`B|2q9@vq)o#oU1YgpK5E^ctR*x2OCe(W#x^YFm zX3lJ~)h6ueC5;k3la@l$iB>O=p2#szY~?OIH~2*6Xq?R=;*%?l_!&jq^oyL2Sy|^L zt$Lf^&eY!ew|w;Yo#z_CgF#qhm(@#(Mj*QA>o^jbs@2`xK;p5-2U2oi7^Jlw42Z9aSB&=tXb0n39;L$kov#v8qHg_ByHTvAp&6!3Wo;mXH zf8g~T88iCS`t{9)8nM~-j(X^%u8oP}IP?dYQk~tsZKV!^=7xl&p@^LT^)0ruLm<~t zRFM4thD-elsz8dwl#09F;C^cAY-{5BVoG>yD$@1y4ROq;MiKp%o&w1PN&f~7$BNum zNz9j+c3b0E(`Yce&gPM=tqMXbo1Yj*ZZyfnOlfGqky)(GwKplsv+)3aT zE6*c*3ilvgNBlc|7@i?vD}7$w{S=Lq4B>jM1A9#d!#cR);)|rRp*?Ta*1B2zs$0+E z412;1JaLoFOrF`4T{@!bfDBah>I<){Ge;|{-8D3I#ExptnH^~o9a3!Ax3$In-hM!` z7p(Gt15F|gza)G6dnA8(!}heTw^bH8JIlD>Rf`p{jAzt%!OFMP(J7R`vqC+s2PXfS zAeDyJUI#3sOz8L;3+&0@P>j5Ih1EPIB{Di-${O49QjU2tEvHOrs5)53D=28hq;q{D zk};5$)->qY2nlsbWbG|x~>vBDj2I9zXo7U}4ZbK#6LO12X z%K@jATg91DYq71-Qj$qhtBncPVZ24(%tcnN2YSHpjSk<2YM&KZY<$CxeA=@bB%uD{ zAGpl9jA7pjaM%X|>vQ_GH(Qfrtd4|}DMV3FjWknK(r03mEiN+O}iFUjEKe7bHp z@6GCNGPv@Y#czTme_Y9V!~TQXq5k(h@NGT4gxzB;6Bv zaDz)$hkPW#cq7>A4db`IH0;F>SR`-qDiFOEV_Jz(!xZgdF`8AxKL` z9{&Np6=3iEOc@m_r51veK zSBuHD{bqS9T+DuTS)wH1_sWIxeK3tWKEjpvVF7c3wrOAuf`)<37~`gsgxlvm@3=Im zXtrKoZ?ku7sZp;UiyVx2&iOm#ZyfvXrgKU8Ldl?-ibglXt0G%$%kua1GVq@7f$>l3 zN^0R$2VDi!q}zwa3$RhqRq@TeU#w(A#!wyePFIm0qf-aV7qLM6KuR-hA55J(8kUtH3TR*T7 z5G@(ePRO%v^`z!c!_u7bbj8wC%{2-1gI6NXLl<4VaIAX@J5+G+FRlUX6-~~ovwp0V zmiX2dIl2J}?BV4@K0Q5-$Wu?2YT4YDCYt4I<@2xOv1RKMOO9V$j;>UYk8AOXiE0)N zR|S0v&FrPIsaoLS(61O<+P7)#nqj+4;&Sm7l4lLhJ1;T5uFsV!Ch^X;9nW&jV|OGW zcAIMp12m7DLnXw(V6gR|@l%uW%I4-srl!06G)<8Ly_6RZ=&metRC9h#*{wiRR<5E} z0(_F5$bCt6HZ}%iH_HYY#0xBDr{(fK$47X-uaB&EwzhZKWHE_>63hxeg8ZU21Ap#b zik?Nve!E7Wg}^d=t*{f~V?I~C`i2^8sUt+lQUTFGo-@}~DCRHsp}X@`V{p}>PMz@n z2RY(F#tL*=*`~KF_qAisI)(98Cp&Z1 zC-SQGZD52vM&|Jg;cfl6@o+dVm^Q&G)fjYX?$GzXuzK=v@x~dh`q41*k95UR&E#Ml zBV@HJSf~G8sIziruV-F*6FG{`$#3nWnzj^=dR`Z6t0dtWtJJ#MF>h>bG?;a7WJ^ZA z?}++}vn&fgZ01Z+df~T3{Z=MF;Ez>p@TPfZ$-RV|=SJoWoCANPkjfJ8uzh;b?`rCqTM`%H=;h-;bH0TcSE zXXDu4{O@A_YFqel_+X)!pk@EtmR?XOaV)d6emKw76e_`GQ8qh1;GODfUacMA9YINR zbsS++HG?0yw6fgcgUbhP84~goA^>+sWS%I794u@REZ^_ zVjy~xVbE%*ci*Aq{($GKd>Cu+pT@|Vf_O^vT>?JN{DAS1)};!%7>N_WEt8|xs`u!Ra$K!m>#*jn z#twjy)FA0_xo`fv+eSD2a1d8Ur&=~ z1@#=a*Z$pRmA@i4`ucTwpD%uMAz^rdBkFY!0&@h*uJD{E0Mf|%?HS0Nx;F1E)pl;C z%`}(xKegPP}~8Vy*QOW8MY zrCP<9n)XocBuxRP6Jky|Y|c;<5h??m9|02`zcO^fi9gzNqerN;m*J-9SL^^T&Av)r zsKhd{a^$U!_)(11+T@$s=u8)x&5HSfqk)0#1Tk`8tNO4UxKN&p!75c0uqoX5Il}Ns zk{JAlG%O!*DC5P@Kh8o91{W7+rm*Rp%{T0Ad159SQU7nx7Y1%o7=E)W!@?FEf}DR) zGyQBjS(xF*^c6v^r})Cl@4OM+7>+TZJ2M>{SDyg*i$a9(DQkXsj?BN~FIS*(rY8od z&&&!;%u6MK0m@RVb7F5C(^sFc&Ns5hr>I`?1Fw#wJ1serS7p@e`%*Y4vw=G|Hh%45ht#pGCP>Gy`1k%8%dDcx3=BA8Lh z$qfz|vB9Ip=8x!&xphv1S$n>ng66Li_$zW;L;oR`x716`WX5|5txPuELi|i}vi!6L zI`k1R1$K0$Mkc=LDw64gvxIdGE5pn}lSijN6nM)E?#^>wL@hnjdpAA8%~fxTvH$g- zA?R)PHwN1MAyh}?-ENOeH}7{EGh_+7SRdttcQp2U;tRJpJ1ao{5PEA|qJ7jDk;V#H z8>XJ{>sTIVgXo?2JuPV?z15z}X#!vLmqb>Qn9N_!A~PQw#c3KL zRfimTupR3Wji<k$&%lixfTq1t9Kg3FCMMoVaJY25q?qOSP0a>1 zf1b`=v)(`ax0836sf<#~iX4u(= zB$H}|^@=3SVqy5%&e*uPrJU8FE1NTKYx%Fm<0||+%jTyC{%OB%{RlrNJN)f^jl+01 zy>qc*q^VbJw1en?^~%_bmzX>W?5D=IK9?)AQqaoKg$lIwUOlX+TB`cle3!MCYb}_$ zDX1r&ee>2f zOPmTz;`IOWSooUB$D5)U!Ju%_*#rc_>?VhV)#8Ja?_Z!(XKZbkM!=9_dGw0|LO`S4 zQ|TZe-4w-_T$Vj5qTFhaD@ue5viBk{1a#ZRxav6I3p1U@PEASl+FhxtsEH?Lgnk)> zKQ5Hibu*d}(oq&Zsjy5i@cAK$jD{Qo@km0IewNJd%8hut4RS%t?x+Bt6=LPm{Y6bn zV$$oSRF?>drzR$qO3(G^*#d=#7C4Dv%DhQ; zF}jK$K3qBTx02P7lQ(rg>!LUwl4l>*OWf`QC^w)xQYRW3?Th@~u_uO;Ps7U5mS*)P zy%`qfAOXS#>x(aO+?aEg;QhxIrx_syi*{WKu0lhmEM&>b&ynnS2VdYr2lg8=riGA9 zXN>m+VQbP7VrI8dUU2X70ask;gOrpMz3wcw8_zVdoeWzF^iFAHPbgZ{mGK(;BO1W; zN&SlFJvoNNdST($0}}oHAMFe#b}uUfD&M{aj)>ON>djaI#Ae zS8xBpR?siy3sGZ-EIwU&n`8bM_necLOBTJCcDy2Yo5BO zOxQ8iZQk;U665UJHtyZ_=IsuVZCr>@8E<9KW=3y1;syu$+uM)VDn>Mg^1d!JGoRIA zadv4bKOg9*vBYuX;X^|Mp6efX?`|gEpD(6dw*n#W2Sn+J631593YTG35_!F8H~IN!;|HEYql56u^m4u z-h&>KDz(os(?6nLg!i=w?^Prf%N*u zM$tlVS>Vn)L5}Xx$$3D2`&Z|O?m-GCca3?|S1heLbdNqxy^p)6t#LL5-|^)+j-*0z zB*X#Ur)Vy}>BeqAnb6ZS+?9z6xg5tn-V(|PfH%jpXuNhY7R-DZ5NB!QmVYyV`hzx29rSO!i$94|E3z}PWT=OB}L*CpdbwnjkR-&wbf~k`Aormkh&Gq_{$CO(h-`Z>c9v3b+hIozrGjy@P&Z-dgpD zVl{rR$Fr#FYI>yWO*leP3VXe8X@oL1duf@b(AM5w`j~L|X31=yk5!|k8=t9ZvNZ~n z{&zh62T67H$dgm72UE~9P0gQ0yV67#YbIKQnYcyz)|YB0T0yO#K!V4i{0;wyyj?;O>N2-xL+eq?U*#?Fp+o+gx z+d~R@0+*q!1CjbwH+)w>rg?2z`*SsE>gg@NJKlM{LnwB=u5HbI5XrJAaQUDKnfq>> zIjI5%FSLwU7ZrVZsxC}OQmQM;$;~hSct>#%?A0{Q z`0(_OB+!iQ%hY~~7qJ=;F|OVa0I2KsAqSif&%$VX>n;DIf`zLy5=W#8CQ8@3nwd^# zyM~7B;w!-4Q-QY=3K;9kLZJ$-8Fv>Aow_>_p#oVc_f%Esh-tngWpm_gWn>wfC4Lh( z&Mg?pCDFeAv}HHT^piIqAR4yp0%&@*pYR-B+QUW?HagbxbQIYNqXAih+nUl1zK9=Q zz66A+mI#@2%94bvFa&F#$?t!r%>Kq!|LfNWwZtggIlb__q$5^3E;jZ&h2z`{HND!j zeiN-f(3e2^;IEI?wJti2PG&wfXZ=y2IxkfT@x z4-IuirqzpUajp0{lZ3K~0ZKiPnr`p$X#YM#%`0tfP4K8|`H4@2QY*Zj83CoRBZKP$ z5+($9DXgQ*hH`#_?XFEOH99fLS^PlE7^MC3mM^)jeSdq-a(FaJ*=o3|-p0kQ+_Wd; z0{5c==F_V)X!EntB4(2d5Qz8!c3~YakHa2nu$U z+o-s_mTfA zKkp3Om??h9e@Zz!8?l;8@6_Ls!I9uN7T3WQwPJ|qA``JR}l5=yH zzQltx;|cTGSrerCP$9i`%akZ|WFx|JZaro|$-8f)`)Z|0olwd+{ zn_|w_5EK%-Iw3xFHH8}F{*0C4`H9Hlj2J=lf#P|{LlP%%+r|LX0x)1iDDs!;{CTri z6*>8J zbi6Z6;V!zfKY(^Cy9G0RhsovvSU*CgbYYgMYG|BA#N0Iyz#52eT=}Zzv$&X50A7{z z#p(U0v0$WtKO)t&^M;>z)bYfg&ZrXhZtXHcT=Y#fHTxcyW|d%Ywlp5{R$o`lzk|B( zGS~7qv0=YY3_CFEw3ng%-m$%1nEo2j+*yIR1-IFDMNaO8H(U;}{vz*~m?ZVacD(7- zJ!LJ4Q3>{L#;o)kp)p;K))NLo7q(68masD`L3GipkrWRF;L##`d?oIg^?hOU(pyD* z8Vo_=n>sSx`op@iY^CQ^xRM-x9m1Nl!b|Ns?$QpFALk zg`h?&O@R$j&IRw?BjP;9$9ybJ0^izlTGVXQZi^Ia*mO6j(*NZ6+c^d2z=F?AL}q7B zo{>?SQp0aBjZ9_+4P34d{*3Dwbh!%hWye!gx&0W$>?g=G*xg&yOIa-n@IA5CM)r?v ztBds~()u$^mhuM1eSLk`Hu6iEu3qiP1jP#BuU+rfrwGT82A{o8UA3yMmGZCOco-$2 zA`=u~jNVEcZ7u58h1~xjDq4nSYs1N1&x~(Gbh#yeqXRfv0}_Oft{i#1D>HO)C~GFA zDEVTs=#xTGUT3JY^PG#j0^iHMHO`>zLfdxbdsOww`(b}Ju<9N-z}o##%C@`Ud=Z2jFadand## zg=uuSVzGJ4jE0pn`{7dNQ3LDVC8A_xj8H=Or8MCVW&!j& zhbMwt-Xxzn1#x3;-wh2rzWYAX2o;l*;Kk(@rlfwRnWFyu4Ex@Es*9 ztus0QHC5v-5ryI1ibE?_E4pnw6(onX@X(K4OdNZTZ}Jn z?XJ^K^Hg3%1%e-kVY+;Np!nAm2}4UGS?;V|%i|2`el(Bl<%t0HFuyZC#{tqRLQcuV zpV78*Blj&qa`363DsT_gwA@S{S#^tnfFZK2%SPCj5&D`HVu#;c2T{&+h=ArlXp{K< zgxn{;r0&mC;#~}B_Mh(h{vt4Qz45t5E#u(xZI{Z8hi`VT#*#oXC~F)-4*`UsMa|2{ zJgR!`h953uQqKJ`k`;cL?he_*R)Dd;tjShiQgr$FiSW&Wee_*r1l@P7mzZ2)ej8+{ zdw^d-NuKicpd(~-W#ddF6LcpKy<3I2V+@A$rU@XwvTl9J$*A ze4 zg^hLGvUeB_MRgJ4;&-N;PDYjI&=|-I({=FPQF*B8C#`00|Bwmsbab*U;v9AMc*B$u ziO)lFc;F9W5@cDSp6~y5wD!wgguf@xo?Gv!5O$a!lF79U-@X%{3dP;?Z)}G?;UTD~ zCgX&#aBFvEN+$=<*z3nJREr||z-~kbtzhag9F790znhiXb8F%59$)5&9CZX1AM{s|>~=e} zI!`vN?p6MT#4X?XHFht^v8E#YQcA38#Xl7Y7>xNaE=Nk_i$Qm>>-H%Pxl6`&w64gi zGE}710PCwIvZ5ErSbX}tm`V8$JXXh_eM3K%(@oajO>4MQeUhxReDp*}VHP_IW;OON zF_JPW(?!*- zN`LVRlgMdc7(`~1b(H9Vt^PaXAH_&BL#@t8>@F7o#QA}Upc8U=tn#3RQ8*-bp<-Mn z-=2Zw$ahoRWq-4>*f?+sz<;Nf946n^T9c`?`^vo+TcS#tCrom*b#z1u>R5Lef-pUZ z2&N>Zz!n+*0T{f!-zK;3Ou_`>m)PUV5w-$){n2UOiyGBZMrBLs80mHvX^xTA*=^87 z`3Ks>2syFdB3I*z*)vjB=M!982XLXx?P(EtANXMd#CA;kht6O4tZTf99_uikl1Uz0 zhC@0JgXTx?Y{7_Aw6vCP)(S+t%skvfADxFsa%*LAUJ@ttElchi?}FDLsmDm)MZw|A z+fl-Adkna*+tMwdUkNe|0AqOz^BP}ur#rlVArN^q^o-1kT661;xHRTf`G*4rIN+6~ z)CpRL&m-sG_r28VyyV;|HeACnpq9aFQZnsnU)!QptlvIvj?@P|O2g>E4L4#Mq9`|I zL`h$HeBx;fj+hl;WFp*o7fer3jf`M;usijcW4}*`*6qreA7_yrS^nm^20@xVq{LLT!H!q>G1PDN$#z3@ci| zk9kc*Njw5wllq{*PcfmGl^z8!919-N>;38g*ivzs#2q2sbxZ-o-s28xSoAmd<`HO)(JUs z8PnHI3=Z^`f={0>I^YFd%6b)8$L6QKIol%1X+=!Ui4L@$2kwRyu4f>6LA%#!SB`|V zkV$~j%S0`D_whnt7QqMdaipt)*(N4s*oSxaG|&y=$D5p9=%V~L#-6JG(x-3`+2E6B z{&uS4xIXi|v9-C@{Pz@HVtHSF>)atDAJ5_5u7$zXnztO9o;07gr+>uUqyPc&q`SS@ zuZM1#7UG5XM_|zPlc1z+T{{S^d!GCyurDiRbH3voe)3hTh%N6MqWk5*1BIT*?!S!N zlhc4#diYk=yK*GV;LU$~2jhuj>Um~X?~Tm2D9pzCnZ9YGCfM$l4Xxx_9-)Y%Wz+ts zZEy>hID~6!(ePu(-@MJOI$JQZ%{mfoqAO1BBwDLa3nh_}EkXEs0TOWFv>-|llfyE6 zDNTr5j@$9_yAVJQw$*aR;BA@Sdh{N$k#l^Os}brs2n+F`tbHUOY;EGlg>D~(zb*8@RCun@341R-t|Oi>d6L9H-Mt36jk@p9&_WWQZa)`pMF)T`rzU85{x%B{ zs8$_#kJm&^C}GZf4}JeLwHs%g-`kd77Twkra`r^@HjYYc_YQb=NKEk!W;A(*=8oa+ zr29E(3@?(MICQ2dY@H*Z8Q}XHy=rS6jf1u;6?~4MH(4Rd_%R3R@3y!SB1i^KOStqD z^R6F*@0Bx?xr2`WTO|u0wX9~?LNzGgf8A=et~i5*%Oq+{PMJuilEzKDp2S5l@YY!r~eXs3miA`Lg57L*|@ts>9~E;bwP9?@C{6ppGsL-&SXf-cp6BHu?wW=qGE4vLm%R! zAY)&97wha@hDYm(CDK~}wVi1^2!8^F&PV1Txb;>VXsz(QEw82Ejc#0 zzt>k6{l?X`xqi%TD0-niN?CE-T1Cumb`G~*%^U3*Zn+6TNLjC#j9ej+RuU6eUQX7} zpLO`)huuvWSQP z{ruLug7RTBeN}@Z&xPdA3{4wjT2{{#HQO$4Gitva1>C>BuE?<0*I_v`;fwKoRXQso zbho!fOf2Y<#+h2b=vOcMoksxv)K^Wrw@)mMBOrD~a}am0DeNj&%cE?5q}p!fMX=9( zbO7q5%Y^)mTiXW}ep7eBgYKq1sGdcEI@10xZixbU106>gN((NU2C2}WiQ68yP;Nj^ zj5d8+T0kQKw%z!iipr_{!HpqdXk}53Iqx3;TE*8MxS@CPx zW%Muj^W%?m?#@rh#W&LFjusRpI9^?yU9%k`PBa0D=_~F~ePwH6 zje59Y&_NBa1mNYnaha3(bALCZd?-d$Yv<>>qCYZ{qKvf~Tfbv}TOLz^0KJ=yATGm2 zr@jii#hHUclSQLB61vWz_}u&BRrvGD*TbGgm0Y*6=~63hhpY zo*=FTM@jQ>6+ru6ow|Azl*k^VnnJCn*yyWUnPAEGcdVD%x-zrRq0+Q9X6R;jlHk`t z_zs0fwY=J(usvGJ_<8=%`6IvNO}CCrR_YN`e`J5=yrvJuyhvd-Pu}LlOLfA!93T38 z1%l%3c00VBD11>SKknXu($d>F;vAutq71id56>c=4)wM+p30i?NfogL?K!{g{>SO& zy!5%fO}RvwX6vl~aRD^e7?aisVtXnGrsY761D15IXLSXpP&%ngLR@w@F`Fmx)j`;! zFzr&n9^F@+vA0(4X%8so7|)lU8!*}371?iNf52cxlSFI%BEo+}bKoLtv;-NIm9KMP z@4*z=rL>xut@Ghs27$dBUrU#`#MKf8APV9HJ?9;q4a?0s%*R+4h1JE53?^jk7=4y)Poe-cTKm4G7H6<*lcA1 zu+|BZ>W1C^x2E1GC-e+gy$;yD3Jz}WNo`}lW?KILMA3W5+^Hk1u-}Mw*~Q-Giv3~9 zP?B^v`yJwgTOuyA5Id0<=LWKQ|Lc35>4_98O-(h==q2sMAt5v5HppnD4DQ`sFuPel771V-OMP-kd%m%bCnC+FV1%?=8!w%Kn)q1am-@l( z+hK3UZi~LH5a}m}Yi^g*i}?7YXiyyqIZf8j2!bvj4#VJGbpVmf502i+%CfVysf9CL zq3v6I&6A+l2)gP4#_V>r+)i}^iaRpoS?tL|15dR!I|euPLHZhVK3cScPHL@9(Wx;o zFz}wXHeGkODuAj~&$|rkBjKstaD;$qvh(}`Z1*LAyjB7Pb?tvPe|6-^U-D!OANOM6 zCmgi?HutFlk(wU)ipTzd5u+et`>laBaxn&E8Uf{B2{>!QMYlCxOf58gLjb{^1*$wE zqbQl;!YAb#<`L*)j=?t%D1#VSO$8pLzkw;H8Esc!&X&LaX9rv`gRG)g2G$@Un zQ@RgFxW-IGAs_|*o9QDpo6+m*U?E0!Z5=KW`&F})zwq4$`hJRe#R)g@NC!NB!de?QAL8MEUIn!XE=eiKs`DFko zn{#ijjw$^j$&Ooq9vq2CG7ZXG{J)F+_SgBqbVNQVE_cbRg;QWh)rC0iC3pW^06@#~ z@YJ>s5aXwYFJZgWZT~@bW)`&W^^ILa@E?qeBb>wDz2+&RRI?Jnn?WS<2UbX=pD+T@}*7Uct18<{OUn+bL_=9gS+LA}+ zr$vx@@~pXTduBXZ2XAZEQ!;Dw0ue;S_%Pw8daV{mlX6yypy{54)<%B;{()@WqnKi4 z@Dk6vw|5YHhalJ7*5xGH{l>XWjXuOS=d;K`{W6z`Y@O~bJ1*^)Tct+ml&U38b^y)) zStrjQ7GiqM5$!1Xqgnj$)w2p%)zn94gOO;)zj`7s*?}V{kpsWIYpDH```)jK*m&tE z)7Po6b0M~J-{rTm`EL$9i2rEv$YEiu^L%|ykG;f3m-5?9A_z5zTl*yCVZoY(;~dzc z_D(}t`dXq*AXd2YF6_Sd^3NchU|_%)_U_*Hrhn&6Cb8TWkTUf)j)Gofu_~OH)tmL) zfj{?t5W`BQkzP97Pn_IXim7Ch*3@)14I)ps=xuw0%BD6p7Z#9QeaoN^fJi$BU63q{ z?=6&bTpvui61s9^P!R& z1EH5O|2BHzIH-~B=txQw_ z!m-#ihh5J1p>99O=l*8Y71m#mXDB=L!}}c6wa8jFtBI)a6Bi#79xAgl{`!VsG5eKqNe`4>Te&H0|oQyf4lk#{uQUbk$$*$UV%g^*~JVt zq+p{!{R7|yHvpS(GiQC%vSd5zUAuBeL9(r*y~X&2zs~C0`2P79AM5c7sJp9c^@GO# z(;BZPck@@OiL3at3=KtyFN+-ZDt~-}?>f0G`S?w_2-~#V9T_77MgK%xnOyA&*t}=| z#3)_ab)9aVE>cTwCI?ow-$`}!7HE!&4z?p)|3bbw0(AAgi#xwJ4&D>N)*bB)&o$uA z0=cfviiSmG-(`dNFu6X{&70q2haZmMs9bb%EbM+)!85`lj+1pLHL1-BvxOOn;I%J4 zxDKID8oKyt0piB*IpkCy5c@$u#a^8na``6~DA9jw6CE2(i0S$?H zMO#cV2NG~LAlqL5F7f+S`To`5AZcgd%aR7y^R0(zMub*t8&gGZ!1#FP@1%Dkb2M@ zuz~QdTyI`r`lKB49RUKy)ax^$96u8_@_f>yK){zO5ys^YU@#BdeUU^_bk?D_N5K1W(o zL{qqkSQ?;q$=@wZ)~i}I`KVog6D2}30N2Vv?G#t0X8`86qlDI%B>O^|}dJ!JxZ9h3b55Zb=btyh4 z1=b>S;@ncN>dh&<$8CrG13;G*HO#%$X{oq3rbb}UbWuyYB_1ZNAK&W zh@rL?sNiSHKGh{77~N=nHnQs6_KLs&4`^7ol8>z8;h@CFR=Epks+^ zO~q>_*{W8Se%R2n1x&_S@>Lji-Ty3SK6WlK=p4$SbgMH@ih=LyYrQ^iEkFhX&*M-s zMRf55O9+WCmO6eaju+Xnug?Sg391O2A>ySKKx}@5xyg}Jq+q+1Vl*(X^nAQRgnDZ$ z!ul=h+Lc{*s+~&j&a8%!!h4P7;t&?SOVJqS$@19*kKIZ|I zB_>Ak3^Dbn*{u&xTA3jEIp+i6-9-~|ym(eGWG(NVjpFo8D(sZvt{kLh+ex7$L7d}F z`O@fk+4+PZ-ObfEV*kvPG=1-d+|v=zwH;2cxn?#D6O+ z@1c)>!tGB_%v23lh0!ykvJQ9aGz=YSb8+Is5Tp=)4`DOz4~StSGE{xVWof z`GN_1cS4lwU%W+QAB-OwPvv28BMqWT7jqsx&|0EXSa-ZzxGTUt2dzRe83`;@bd_Qv z9*t!r_`^P8ZmuY#zS*9o75f!bkuWu-4PU3oGGKvcvLr$${tzs;vrArjmxnks@ zx)eaWU)=HtvEKgV-{SDTWX@2r*p92Pw@|N&W>5%1@fKfq=aVHp5{2%N|ChM$d&Tvj zN=F_!GG=Py2WNAfZV6~t?mD8+V}F3LZ11C9s>t5n(EMSVuF5BLDc8f(iLTv`* z7y;8s)U8b0(&A(pX4gx!5cT6-+V)PL@?d~kY#%HS9j`k<77F~%~ zjG|?{Y$5A*N9S44<%dmZ%!HW4Gbh_HXCn2LfYTfjnCWKEwA9r|0N~!Te$|q_n7_R2 z37GOBUV_Ogl@)NRtjUw7iJ#$#QlJZY%FtW4!zc!h$Hw(@>P07sGCT+}Mzj@J*jZ>Ayu) zu*o_M({-BHKpp^)>r#`(;7O3#Z|I7<_Sxs$pZOBoOFOci{oawATk!Qens&pX_k|JK zrBSDV{gWp{CY>+1Z>wtiaC^*KxHQeh5R4sj!pJ=fg5NaktX&j+;IpUi_xzTO!~H`- z#tL7;%*=i?1W>ml)NnnGLaEbgR5Sx70O2T1&2{0<=GCRjorIX8`8iqI{!1K5@ke>( zDtlCuzhhXxoiMxq<4M;~j<6I1eFfr;*3otJAVa$Ic#a?0E}*rRX$$WWtMF5bmw!$f zmOCbd^XwxZcL7d77p-%*BOh&IZW0^7e5yP)H+F`0pqk+%bhpEL=}kUQe;{2Mc~K8} z4w`p9A>APoLJGuCs_g7aL0OfyU2k{U_Y+-Zm~<&Eu7XKBrl_su0WdX3C%(OXC2w#k zOtz{!{$`2c*@}FCr}P|`ooE(NG7XIo5Ca?|$CFigeAM`js8iFk7fV>Sg>zB;5$P7% zg`9_3OTphI1qjc{d z*2Wt3g(*`BQ8}{$_{`6qLse;cqouR?E5TK_Op228wRKJdf~sSEtw*D%B$n@&YCiWD zAYa`*#N0%s3^E)ng*yXc71@|Nov62p$rdoQsBX)L(iAJ4`V=V7EgMacEw`}zV%(eF z3WI}++34bMEau$XYp*PeFq9aZ;o8hN&(Q&ti`fI^?WnUF4-wTBfghJvZD*IYTBEzD zhIS((S=(I3(-m7ECb3|?z+i3~us-JPPkPpy@;biemadza-Lub}*5^XRl)<*uZ7d!d z^baW!oX>V1Uw+g!t25A6hW0tc?D7^Zuwo1BSE!V5JTpajUe0RZs(Inl*C_$Ogq|FAN_z*Qm*EPul4n4B=JitBJ3_2DO zRFRSfq^5OLU4T3Ps+anI&!SxT)JcG{j!9P0`%$$kI^ zMzzwBI2*hI4rXRtNn6C4nyLn48D^r3t^>QAQTi}Vic)rX>|NWE{I%m#H|GlY!XENI z&vmWt8$466dUiCQt7mk1;vA;4QNyuv*PRDrGhG|&BcV+FY^$mZE3Fvc+X^ok$JmCA zRL!PO^Q`L|@d(rio#o&+lJ^=8w zoV!2}kJi;M$4wUJK|*DDrzzaT)4b1dqW%d7B#@*oa3Su!<*<`US&A54nnVe_k@XTX ze&<#-M9Zj`L+n`Y8SlUwPjz_DBI0{Y&CE=f$5Z@)BCYI?@4-wcXyi)4V6ak`G!9U0 zIL|O5DRJ6B4y{s%jsc!R+)n1gHD}oyuCBXvpgOv$|E5t@Km2WvQCso zxOgXq0-$lG9J_b%S}6dauhi8PV+p1Nq5mEkyEOQ8)-=e^{@>l}-T1l(UWWLFNoaZ0 z`4Yua?+VtK^m4EjhP5jjrl_;tU0B$gxEe`Ot%$yB+EJ9( z+B_}W`B}u)a6H)&HC@jcnl69aLlk~RdRQG`xAm~dzz=Vlr>{2FsXtnE96EyLJBeo1 zl2Jk$g#P7VCJnV64sFiP`Bdwlu&hi$*U%g#3alpJp8>T0=9)d|bAQG$U&;!6vCkyH zo*=o@hO9M#cKv9Y@$0ZkBN8;R@XYdA@J`F(?%0os_kZvbMGCNuL%mP8B}g$m?uMN) z7HNk$-;O$TjMgU$I@o{BM+jk|ba%+zptHTFg9eNct3`+P32<fU>|70A16g)~!-yr}2AY2A5|0nHk*3T*TF(5=!a zF=}&S{pg(zPsZi0w;^glyxPH#Z_A~->8_@Voo)@zd<(I)fA<{v)rfstA>jH2sS31!KbJA-S3G=6}27$0j*)hc!2R^~fxpj#jiP7Hq|!*)(l z*(s*eWeH47OiN!r`b=yMJ=ziUBg^`V6pB`iw$L_DDTx({ztFUciiLKHLyY!b9$AZ{ zpT9I{4YYC%4h8NA5 z+L<>yUwgLH1$Cf z$=bc|MRvP}TrPK7_P6)Tv8p{oX64>`YROHh0nW>H+4SNkeU5YzPa z_BK1|fL3fNy>>&A8?^RZoNvu;@teiVz(Urt^YUm`y%STJdS?TO=L>n{vy4xW zJ}G=PmpU%rcWa&RwFsFyqT*i|rI#o@rd%PFCR>e5Oxd%`vm#N&152|G} zVTg!}mmA5qFpwc8GZh{mMja7lrl_))}|JL)B3cce?NZ(R#B@iHt=YLnlr(@WTS zbzHvu9q6bo5eyW9)S?;XIG*&tW%RUS#fFUMYR{06-@5} zWXzW2*=mL-KpX^!@1^^K++lD7L5s9D22T%LLpXPKU=KV~HE@ z_LkWduWo)2k?Z3Byb-q}Cq`CSn!U3c!oM`YvdyUzVaV?~+h|!6R3GuWNv$y6ri;2W zU>&l`IoAan1fy?!u5Rjs8FJ})NmDB|etLj=TiZ!N(`7i+rAO9)l%z_fv(1>K>VJ}Y zm2*R3v^>~H7p|Zk1GHKzihTarl%i=H{Cp=@3r|@L6~IA*J(lZj$@LY^w7UEu<9%k{ z$ydO&x4orJXuq=aL)Mu6nItb;!3|XyCg>Dsq)VR94Ca+qxEH$MSw-0EOj@Y804v0pmI=afu~up;mOX)v{88V)bO1i{o5( zK|(@dDg0Ldx<=mFaV-#9Hv@Q5!jYnCVk$9)r~c)g8biB*#t#!Arb@z&NRz<MGgxHHi{?AYRN-0$+I1)ji9N?$)b~~isw(q#_D?`**P`aUKjRe6$Yg( zX;vv7f-t40={LRc)&YXG-6{g`18j}L>H&=XSq0alS~FUg?|4^7IrBz3S!nIPGErh2 zrA#k5uhVqp{D9JQWLDa2D$5*mD{CiW^F!B4$o&NN*{yENG6~!ORXVNrSG1#2)_Nv& z044U~oikVL-a2kPT6jXIH?&I(#s&;Bx5ay>zC9msjD(5vhH2cS|DWfvW)uG_uRIkT zNs_hebmQU1*t&35Ldo3jG5F|B0r`lND?`0lI_B4a2@k(1<@qOS9_1MKryy!56v_hT zDxShm1KNa};*WByeA^P5D|4BOoZ*o!6PQPRjh4?qAJyCUp0RPDnWoTKv0KEC+u-Bgtvi8?W&QSX_S0KILp!@A7%)mGX`_5= zkjCVdAJQ8nkxNz(D}6)FhccRiFYeZfLlZ4ul#$QPC6m-1cNeO20_8z0fVly4j ziS9b-fK=H$DtiHvXcNUBjlc{=4>uP_l)yQiiL7?M# zxvTq*&OPSaZ!$76rZPK6KiN_X22!^|)Va4abKbqAJSX<*q4hcY8>c0h9!85zzG_G` z;WOHd)O9{;zukGVt#vJ4?{Z|$(^D4>n!aQXA2Zyx^WOCa)JLC!%68|c=H}`gVYo1q zJCLU_XiN0`K--;{WQB*O`@?b{deU`S)ax^6q4VTqQDUB7m47ZWVX5nIKDBNj-!K9Bj z^isWL-xIZRVOkucj?dNgdv}k|&Z+T_M&uQrbEf#?kMO@8e_!`?Y_ z#=6P=coC@qAKC_-Q)aa>)e_SSZg^EIM31}wM$4zuPd@Cb!6Jj^9#p4^iLt#s4G}FW zbap$(?{OhX{L(YA26L_^-5`d>2qjHb0pn|LtaeQ7m7Y7|6f=(71ZQqxC!ei4;Hobp zn`WTQdu&650sW`QC!^X+BG~B-RHT~|Oj08Ob+XJb3r<4FI6x;Bu9`%?_ ze>vS2Z>Saj6;ODKgh92vI<$x+EATxhzOcdmQnEw8$7EAMLn0qty6upl7{!2CJ;cIu zL|1{)%+wT@Gk7ks(fu@r3v)Sj^=28=zvwG4kT6<2; za1%4nX4h%PyZKcD6tlS8#pt={go?1mgSbYg)YWL|D#qgyN1U0uZ$|u+mC5mR3x!MZ zD$dOmN2(M#80MI95kDZ|AM3h|80=518($tD^#8DyU@z(c%{H%Y8hr3KlXJy(0mcTw zF)u4Q8C1HN+6~D1OZ(=|)~pkwU|rE!jS~?0+egrMreVHh0ttU6mdJ#f*mG9TD09Ck zL_zX7q$s~?r_bqlOM!o@lIa<=4wY$j=u={TBw;#$Yk9QZ_ zZ$)3uG??{{frE5B7Z+8YuUrgwSDLn8**0*;+>Z$i6x(a;77Bz-Ck50mkALu|=uG3i zARUKxsp!@9Qe4D4oKpje!VZoa}*_NR2(#Lgd=VB_9sbuHSipVQ;bZSY7@nMtbxjWvi*!*HZhW0QkMvE`S?L6|9$m}0VX4P32YYbMcXpUf#?wt6 zwiN&uU+FIdV@6=!j>Dv%{w#^xX|s_rGIE)o9rAt|w&nIeMuwSDM}OAN)q(n9W;Q1( zTApCKoE$++-29pARlTb*A5!v7oLWBA=`-H0wgIb!UOFOQWc&r%nTo)1fa!0Kzc>13 z51`kN$cIXHdTefPQ(m4sI7-_gm#krlEGyW2$&T zMda%DRyzELN0p)VxlF`{oVY4rOK)$``0}zGkxm_?ItlHpWL4tJAgnkDIW9 zoOT>tDvxsls%5Njw07(94inYRNJh*=a(Zz}qUQZzF*C!I`?&%-f+u06n=j;6M)Lq2 zhdY%-C?rT!*8Q$h*dSZukwP#4d2mE6N>$6r>F$6c%cDe{E--{{1JswAb(E9?JOTXt ztPZbZYY0gO9E`0TcyBFNDnhBgJN4AUQ{X;Y)7!83kg~FJ1{iF+z1*2PL=9kj6ZOOq z5BuBmhO$whXilsZ6M@r7WXI|4_-1!5WgU}zhv%u~n;-?+&8#iF)Niq}XEA~s#AAsb z46>9X{#yqWH0 z6^mGr)7c7|nwqNVqKSMsSjfhV$b^{9hL4GTK)MO&V0*)kJbiWxg^6d>IbrhI2xUFg z%*=b_-h8T}L|gfwC3co7HY=-n%s#!~;-*|3pu^6*uYDX^0I2IXhJE1b!Vg<8#1dGk zw-uEo)ylHb<_z+o_HU8tIwf0?k4!#JxCDWs*2X+lF-K_7D(UkH#;4YoUZKCZxuq1D zTz{QDuoH0;E%4dTu-SFtNzn2@ykZ!`%H-wo`meVZxlcujUdtR@S<7V9E_413{1f z9=c<-oG!6pefP^ur@83p^kgT^s_@ooJRo#3kv;wZm7zBonZMOHuE6~(&z4TKuA z!`A%QYp+NzGjCTVr}-#zhps3klocVcA&@E%$1K}F;h2t8DzkK*JcOAM@ZE4lIm!50 zm$u;2_2dAj#Sy&bJyzD-(Gz5eT^$tOXj|4v^m7fM#fLk{jKA;jtI%F*oxT5BU#Vof z)zg{zcARY-Rcq@ZE*2N8rIf;24`Nh-MPF=VF&1MGH#8Esmp%M`eE2&80(Aup-5*7w zZK|gaG&|un(J3j`#z9P8)b4cv*XIf3(K=(YX!76>v%+RjZLid@gx@IB5Vu~|e_Qc7 zt{O>v=vMNJx|aYsg6ta%(YM@D7bjoG=%o6r8nt}DT(qrP?S8HQmpMe%&aagCQO`L# z+Lr;!kA{jYpO;~tHdnXkrHW~BF)T{BYuruq50 zSE1TLhOs+$mP_>*_^y1;(Cuk`RL`h$3cWjKgh`VqMNKXt77%FkQ^@*A%T7LBl4}(?ScnvFi|&~PGp4foZVtE$_+r+_(39!dVO#6D?xjeK zBdUiEyP>G#4U&@V6(3ggc8d$d$|Mf?Z}}RoDY(C&@M(AGcwz8Fxd|uz(L_XTkT3?i zZ499JeLjD}b!h&Mz;bR`-zg5E2ycTjj%&&%yztO)=o#{(+yCgDd8qqd&IRN>=bb)( z?!(4WhK(Y28jbXJs(4D2G0m`2N6WbZ^fGbAbX$V-CXdB|j;#HQ27cg#_|}u_&znNv zr_;7OOnglf ziiy>n=-~Q9{^|(Dt*7U?P@veCc=gp2U`M9AS+F&SM%)CsJ#&4@3?!DP{FiX4pWS&G ziI>%pF1eh~1S#3!8YiQ7W-2?scDb{$@$I~Lm$BX)#-^&eOL4DFjcR+!D*Ni^py#go zD!TXf@6IiXd)Q%?om#{+lihA{{9xN*j9P@gcFEthb%m~z3&JjA<7LH>07w?;w4j>d zTHR!Pli5qKF@%W^wYAN88`E(QHl6&CNLan46o^ zgmG&*>gnAmTlwDNfln76J!|zvQtob$lWp}Ne|QY3x79GTwqD`Xj?)miEd%@yp^i|21yW_sUmFb^QRS2trAwFtl|m>xqMX#7>$T z!LF;E_Wc?#wi=!@R-Y@K2b=-1CAygIFJKyZE(|8NO zo3>VNI{L9cv&K1NZ%N`WBzl_A303-4tH+9tQYiUc?PXY#i~r}w%4>@0y1894BINViM^b9q(O|`!x-Qt@p)yVRHM$6)7GoT%Rz$+3pZJ)2O_UV7XUz^ z(HbKK4Qzdmg`%E*o#9p)rDZp9fiGTdep94g50LTf6&h?#FzuVA^JR0ONibt&W?Cr>_G;Q$;^R?84muoH5!isH$+5vwdQ-*b0;8{KkXT>kZ zS|__j+SwEyW;o3PeEX^~u;bNUr}=e<-JE27bzilZC%Okr>r9IU-S#6Rv+OVC0VUTL zdeyJW!_RU)nrH}-t?C@VXl`miMCQ0K$XJA)lNjd({r0WfYsett5i~R<6=l)E>Wx1P) zLu~ZUh}c}v*-28><~MY|zv6kOnq$25n#CE*>el_E#+xeuUNS{$g4}AF_goMKyI^qR zPbI*rcal4=&s$^S=#(F>ar;BtRwp&Rd$J6xa#S0y2K}{JlG_U7jkrMhs+jdK$!&!6 z%>7V(Lj^?D?8R0zYvbR=z)_BjS~Ari0E#!*80h>+1~b${?=?h*SWMRT!n$k1x%jkq zK0WE-cT3&(kr~K~Y`#mTEx~F!QgpB2dY$-ewNd}7M!m&Ws>ZA3v`2H#jYj@ip0|2} z%T|ZFnDuAmugzON2LPrI?2|Odef{Ezo_XMR>_A+>Crx4_IBu(9m*@D>k4F91WId0| zADs+__q6AKX%70dsi=P_%VHN*q}zcw0+9F9RqA`hBz)(;dgBb0n~FLc&f=$L+uyxh z${2zpa846%A=JKjH%I6bXx*fjaQ*T|*lz0PZV--7*$X#;uUf(xCLeZ1P)CCHHWw3l zZ>vrIkh!vRDTw{xlWm0H|K%GefD7Nt%Qm}F;Hk3Ta&s^!|9VO2HHg~MH|+V$LPG>c zqasMMyi;L@f_sJ#X+z`>*yqBAcH!dB?5n&?mNI<$nO8z5#uQAGl1D?iAumA)lg}YK z>xz~LZvI?|m6UUtu+@1fvAyF{b^l1|nfV;Abr($7D4|RpG6RB33b2j7bjv7X*0GI*!%^v_W7v9sicbVkmgYuaf0g-=5 z*Wm2k#EWkOzo#DEZzdHGIa=)9)e6w`f3$;mIs3|My|pA7sR1-f;?6ySG~XZr6GUnw zSddAr#RPgAJ|zB5Xfh=k8VWZRgA8?oNU|(8+5zfvLEnLeXMgizx%h^O)%tq-JS8=J zqByyz?r~DOh+VFnJ;b0xxZ{HzIQG)zrb~LghqJU8|BI|VaBRX7L|JK3%-qw6OMOq7an5CxBxS}{QHMKT*p}&9|As+1SdhYanI-YQe;QCd9-wp^P zao2iKaQebSLPH$(3?EdVbfRB?O=Wv&q(Sz|ZM&+?vTo+o=3_jkRVE&P@8Iy@!EHct zQj+fUh?G;b{jAU@5H4VrB{>*{%+Jqcrig55XuP<>ML-;Z)9?xhWOB$e(l?FJ7tj42 z6@nTvzN=wQC|D3gzq_a0)VPx^@Fl$R&!p@wtY)`tDteWjLU2{7u7%y(b`F%`e}2K} zlu(0Y!oRz{Ap2dlU7+yG%fWd(@w_lZczy<8kIIMlK??*TGdCg-xgvriLIkhGf|#(| zwwhZw`TKBmkH7iO6R9CamAk5wXd0b@=0zQ=hY%B^ZWwjQ2%={vy%~Rm}qE6za>BWf!OC^Gf&ubGSHNwRg!j4;OGOJQ3RokZOQq)f;xh zTm-3NV@pdP_d^=GE1gU&Hd9tD>O;#_U1uxa*T0zgIXCw8N<=HN1}#+<6oD+Q_&KBf zTk`V{Y=7~g8nf7TMmo$9c?oliT1JJsA9WU<iRWvKd1FDD_I*t^5t-hbo9CB2x(2D!3x^9+cxMQBAV30HPt?@a z^q%wWK@N=^5-qU{RXH)FOH@G zCaq@&F8+7xv3I^+FC^Q{4A&dl!1TNLvRhzu$fQ9_zIsY|Vn4P2-o^6QW*yCmc|M;B z0I(aQ5*rc1_M3KRgdcHYOr8e@-e~#!*_WiK={Z2o>o(1W0A?V;jKAi*e_lP#Go%LW zhHk+gKc|lO-;6$25~TH)wURZkxd4Uk-PzQ=`0gzKY!sMf!VPAyG4?*$TQKt45&1gF zPVF#(wiy8Wg~6R?aiG8N+r9q^<%=YlL~2ESMNfsIlYh4zN0p+d?fGMqOSp3p&XzqV zt{ii$s#08dCg>MZF* ziIc;Ht(_+!O4%$kpdutzux4k-^8HLFd0Bne`YB%$S{7K}J4uDK6nC!_1GRgEm;QAR zyxf4X>D|sC4dFj**!Z>u|7D56fXD5{T?WDpg8mx7ro*wIb~Mh|{+n+2(MRZE?x@LVuLbz}lxBgS-KU#c^mS?w+6boT*#izfr7wR{`er zqpdB?e)Os0Qc&eLt;9$OQ%mBLzTL;bzIEdc6Xt&0eJZ zLh*f-=O)kc$AE#lk0dKHNL26NZ(;}in?k=>MZhobp(SA*z;}|UAawb(+UzSymD6U@ z*ZYGBf9~r(d^`cfZ{E9mEpEpEBEbeYL_P5t<8mv~BoEk|*)8eer9>E=~#J9|r`{G)bpUdsla{j9+YN!)=hFUv#+02|y|IUe> z)POFTuYjXKg+bwaNAUr)fkrImb$_QuKDb$6-e1URHBo@NYGUG}AS{JAZ3}v;$xbyC zE-8t|nj^NXJRB+UE>kP+`&Y||{y6xx_kXzftULZ6Y;7=kwEx;R-0e_m}vv-*LPyhZyv;-3Arlx2v z3VQSkwD`;r&wBCiYXG)B@Y>y9(9GYUt2wG~j$+t^ru|NSte8p2jv5ZwNb}7)|Bo~gw;g)PemFP6Bl zfh2QNf$++aFUZFLaQRv6H3Hwge|8?_uc#V4Nl_2$G#hFg6ZA$C9Hb}Tr8S6u_ox2` z&*1sRGwR6O$LhS>27v6V(FYuOdHe(L&(_n1PCuPMGd4Ku?!4>XsAU4PFt#JICBI#R zFD^&O^~It#CkUdE%SlewCg5@LG05%IGj$ zeXn(#Zxi&5%MA77=OF%gnSz6f1W)(9@)IPM7X0xq2Nz!c2&d~xa?aZp;LKH7?}?U> z9x3Aey9O2!k)1bKMNrAyQ3H1^ZSJua2=NMk8{+8z#=EN$6J_QO6EPZ5qNf>mZD4ke zc1GVdUa%Kf49BQe9$}5-^%#II*^&{7EPp<{@f!;qPKTTrQlnq7XaZg8>#NVZ7oX3! zI2Fzor!>FaJ4<8+4k&(Mkj9Ll?^XT?cxLuj2L;Utp4_(^LLXC}>4Roz2n9q-8_$_q zoNn>W9yhVy3!B*!!7V)OtY=PGd&holy(xiZ2kD9v^qTf0jg1RHJDOZaQyN6P)8*XM zBtl5C>KZ~#-tN&r24BT(S#wG5k63T~@p>iVw#*DneXQXMwEt(G@)wgIaN7bzsU8cx zD^Ox#I8nJL2x3_6z#h4MQ5eCg-3@g2F@mDN=S(Cxo-h*_K&T_@yOu;+1D-D`I&@e_ z&RPL*WfK6K3h&T1P%YVb)iMbfZ`Z;%G{5uH{$QM*5P?x9MXf0*zfEYIy2IMoa0NBd z6Cx^#lLBx%PwduSXuG+$9KH-3)%4t!nIpjq4u(c*nq#I5`-RbGw*k%YJEIv^bOKA!S`;9tL4kg8&wo4Q&R5;A*rvvR3>b;!Xm*9?EM z^7nvqaB8iomyY1YZh8u;nzF#h|Npc+l0;>@V#L_l91tM*$!O@cVUIUlL0S}~?g^kf zj;xCOtk%9V`r!wt6mq&Oyatoj{J{U? z$_Gaiaqm8|Zh&`qH4HGx0%lBXv(^Dx4Nx2&4TyQ3S@LWY%vvguC@9lqR4>qV0xUmq zG>pxX1O+Sj4iQL-^-V8KacQLQ2j4#k__zO8{txZ_HI~52_%u+MQmJQp$|6py@v6o< z08#u_GwaIU+RFcTJa`eA0t#=q38JcVh8^wg@yH)eYVZ3&!XCI@eSk{N7~Kme z^oczIBjb3#DZ%?v0jpst9W#mdJ z9|c}Ke=wBg^v7#D0uUa(uF|Zt^!Z|e-(94imTeVD(hBWV6OwNZ-}&6^vrbZE=lAMmV>z(j z{zm9{R=I(~wb->R8d;2$CF{IY?I)uSb5m=B<8@8>KPPef%^$DmiQPiOVfWE)Wg=ksal|k2hbNJt%OR>o!1r!01prA&|J{qnyyTEc)&(vu7 zdnb-~$=39e(9m*uOZTRbn;Jh7&xbt6~{CN;VX=>fU`lz;rdUq z^;L|si^X^EHV&#I^mb0mC?UqVj@tVRpF>1n=gjJ4JRiiYnNJ0Al z@iEzp%4a0B82HmdLTjxSIb?LVjL-+njh8(zH>8eByi+=? z0EYl5S;3mOg4%vv)MDie`6gtV0Za|fUj{$Asd218o|Sx!GLY81cXnUKSeL#V=Ea6g zc73W~XRvf2^E*#4sE0fvh!Yo$F*NaF{T{!hMVJqTFYI+k3VCZ!#zARStUb2y1yDyQ~dX@br^_q^2 z8(&OJ1Fdd?f{n}_Yhi>KF_CwV*F2i(@8}df;ve=4Cbw_0y+K>kF{Zo5)R?h!N_1qS z5TLgF$+9L)ZCkeeNDcaw@v5mRRI-X&O3yrmtt=7{iTo`gYi_Qf30oV-EZ-M1F*9T7 zhy#_)TZgK=xjbrrv?k@Oq%MAKtZ(c<=f(%}v6(u{xS~{-lKp}(q0J$AL7Ft1tlOw1 z%QwrB1=_U{s5@gxwi&B8Ry-jpcvs!)8j!_W1rhkjKUV=Hb~$gd4p3PD45~{{Z4LRo zD|`<6Dkl4YmxIso_}e!*nX9j47NF$~9{f)89Oq=aGr(uXH(QFzUe!DwoO3jKj|dDZ z4`{{|EUIzU--rlT$XzYFiqMPp{_Hcl`fo!L9Xm^Z+y|U0WzUjSk?u^(Ej4=Y-%~wE z_^N(WS4ULeKb*~Ykst|+`W5>(~a$~2&dDM#ECtg*d0If#mCL;YP6>c#w)6kp{ z&<1F8==Ncd$TOfDfG6>jIg_caU;j@e3q5-H-sVoo#cV;j14aPqh4ZNfTh7d4bv45K z9~I7V85r`^$Lx73^f!QOU2`-urcC8O_0bFum6 z6s3yx_V#*;a^wAnpEnBphA&!jtv%_!KduZ|xh|YDF5k23rdFSck1eafJTSmQ{%Fby z%rv`%z|94CE#deJWq*S@v3@G5@_n7U|AW|{XPthx5aoXmIse8YP*Cf*N(0>nQI@mc>0Rj@IegNsfk)i$4N`nko zhoL?|*j}-dS3S7V-;eCu^Eui0f}x!Qjx5hstMTJrgolXyulTNQES-|tRZ6%_8}O?^ zds9ROh0Dm40rMLy^)$cT#@BRUU1p&H%6w){5;@;KT0V7Henzqr*Yfq2k&*F5-7osa z`0Tzmoq#}zsqME<**pQ7`n&pj&QR9J>#v_|2$hZ93-jsEA!V6roK>V%kSVdNW<~@9 z8}>|`mTAgudR}Y}`?-+{FA>LCt+YvICfe6YvN|Et5@0O07|*1ED?2_Jh+9*uR|=>A z?)-wfw!Gj-yF(Dun^}3xW@f~`u^=ewe8$K^C!o%mk>Mueq^5;^K$H{&*rWi3dAX0Mtk!pj19q-7fgMBO^N*oSDem zIoQ*W;G3lbpC3#7ioPd5o>(l2$$QQv(HwjxV{PMkjJ;r8jr!Od{H6T9PvAGmVe+x~ z*JU#s9!g8C*t}Gi*xwtlCN=2$ij=KV2>+~z9^>Bo(`T2D9n&}G0cdf1)6wW*JHki^ zN2Ac61`+1(y2Huv z8v6gqF@TC*7dCcsU2nW>$*KEP;eWu5|Jj&KqL;aN#x!8G!GqZDL?u3}=zY8-LJBUHkO7zs)zJswxTr^#=~>|^Om+CysKXml{QM2?1Uo)@amV)o+rMS9Gwy1pUq5+^kb6F z99`Mc#g+iThrFfmw?hY2{9Q)!U(vkt9YrpL?m_@POb&uI_50iZ{3vNj=uGcs{uitk zbB=8$_L^n25&9F7g@qZ}L%A0r${PZo&L6YTM;OZ^Eg5Zk3kzyS-u7PQDISFH&z>2A zvLdSj0P({F;QPr#IHyB9k*qm;HH#Z3K{o3wZv^3s1^#DJ;Vz3<9?Se0WAgdmEs5OA zS90`)&Y`TPrly8aFfgPaVfQ~3=pOqoH1qH(_Hv7U+j||t*9WxT_wGIg)qe!b?{`l9 zYcs?|Vq_$fO6HoSqf9-qo0`m~HvP8=n1HEnuFP8AW1x@ifDu0M;`cOJh3>GLC3~gN z$Oy3#{>)biLjH9`Ani6lJoIEHnl>Ec>DlaZQ+ryIw09nNq(L_!qrsJA{MGtV>~Ks; zV>3284ki+vEtW;ngjc<%S*2y5y;77J>uCT#W9sk-P}j5+&>4PnK~JfH=P;ad>RSKB z7)&;H-{g4--?fIc+!JAaf5JrqZTd$bM%{CBO8O#B24_<3O;9P0U91*g=}7EGHU5dEP$x zrn+i?)Pwm}m})Lg)^Ac84s1(+ z+_?^zkTD>tclqm{^X&rd;EjbDAHqt&&`jrg)L6*3=F?{Q+8)?=w~~DF6SzgLPjdYk zCtnRYh5P80l$YpmO9@EDoWxep%LnFB&9g_f{g2krBcFI3eRlRR^<7Cdf7_YJ$dHA| zrh%an!K9g_84afRWLRNBB&_KUni5UfoN6o3Ur@G**H1s(f2j1~dmY|8Qw6AdSTFUg zPd^{hf!`dz-bOP@Zn$Gxky5%hCaS!WqzWEptUEgWpZO>Fr_k{70jDQ80sH!O3eB!y zhU>3I4|$r)m(noxmU4jRa1$6e^dLk{Zl5dd2q?(XT5%qDn*;uUqzFmU%q?uJ!&P{h zSIRi_@^28vU~K$}XYOE5wEkM69%{RuczELmxUy2MdwC&!@Sx4VfAO3i(tqt^CEyGG zemQd!WWQ5$cWxt{RE0b!u*MyUHx@RNiH9rM=401e@YyH&9fiNee{nm$Tp&p@lWDH$ zb<9LvkhcHYc18Gwfc&ZS(gk)?Ip z?CWiNwwPZTz+mJ{elcSnsdfQO$+%GJ6!@D`on-X?B+oQB53{6lwK&H+S0|vK`YYOd z_2LP|@2G(woqHDEM>xM3FBM$~E{zHD^6!owX+nE{F?8Z0u%g)M($BjqO(xnQ)8AHz z4B-Ox)A+&Dvjpuy`32yGh^=fC6!Ky&+96=r3xg4Hc)h#7M_d|E&R#g^Ocu6DIcFES z5PE{%-Q9h-vQ}x>Sk7!}ex7?PG+wWHKul~HFnL=gm9QP9{KOxKo|;8b@l^MHSjs*1 z01Ayxai@}=6>~}w%g)YyvY{>i=^_)FwP>MDYaj&WxE;Zjr|fVn>Z@2lb1A$4jx<1G zfIHYecgT45n-67h@@A09nUBA?0myL4%WSi5jG@&=2(U9VIceiD@XbeEI!60E(W!-FSC>x{PTK<9fKOyimz`d4Nn{CkVix$*v_V; zr5y{qoI=&5WviS&7W1-T^LeD5z2j4k$Ze;M9wiB?YIScq%~G3BnO5^1svdxB_iFoF zj@>IJXYx1y(Z{Yf>yVps+@9DT&B4)wUr;ONWPr)#KU(o{bK+1Bo`LylHG4uVsU#2! zpB9&MncgHM2x&xMyi(KpZA|jT3t2#~IH~!!U^n^|X%k77$NQCLs<{E2S=!ljF4Xvrf80uI|I#h}*!ha% zhtgjzIBI-y>1mr3MsiwEXJ!R!5LPOgW!<6X= z^Wd2%h-X&d;MyAO14yyUX#5+@pCDe05<`Vs@fjdqXRP)OVJcoPV3s(^R|mHieB9h* zn!TN!RxKvy0qe>$Vh;|RIWRew>+-q#GH0TVZrDL+_y8t4&cd3tC_C|=)> zY%i>j*J8yw>8xcpTU`HfPw-|=7zJ!k$6sO)nhBjY>ej~Eo-><@FaM9Rs|<^3`}$V| zQCbmD+CY$2x&%Z)B~?<=OG|e%jtU48D%~X@UD6HGB?5!AbPU}&^X?(M2Kax!+~-_8 zbIv|%uf2Nx7UIQ#osZVb-=!*4=_;!{fkCl>WPJ+(7U%AP>;DF=AHYh3^_w0V&F$%oic)w4)eE`T^7ovDajNF4EP5? zAeJ1=!4c$3Tg+aTt`&Ga;Sr8lqbg%fqlV=}#0``;idvm7@l$FSSREYkDzs%2JuEoV zbj0*EgI-rzIu5o&q$up@=qJ2z7VZdI?>uKEiA)K)g~rP~|+QTK+Ap z0Dej}du@knZ*!Mdf;D_jV*0!?8Q2b&ucKoK0_7GQH&4 zCgf+d7N=nSsur?ldF?gFTzh{%Vd+X|QPn(hE!_nGkk4piM4Cmx1)UJd7!{eg789?! z`b~|#eOj!{(5u)nElg%07l4tfb?06Ig$(Tl@R7x`%~T`~a;!`6Z8F)cR?@sM|9a_3 zx5^3Rm_c-UI!m4(_sjXaq0Bg78Xguv?Z#1UD_o!HC&``vMlp<&w|G|fi__GY(NIsS z^6fB}ZwGO|u(OJm?Zr*?b7i-J?%lMtXEod&9mtp(F%SU|Oy{-_1IR7v&<-lcjk4YJ znKe}KSYTBNibg>Z9y?cEAyA{5wssaT(|F(kmaHC7k@njAhViox$Y3gLKRAs^l z#O88AQ-YfhIW`%y3dwjY-||WGf)gH6bGvKnl>@s~C$6IA6{6jn2RoIDV`Y`|RLaT8 zvh=cwb9Xgph*NJt&#V#oES#M=h|yMjzc#Sc+yr=k3chu%8RylLWqzB83Fcd^QZ@DnxYO6= z8P{GsV23u)vD&CQeN>hwlj7&xiM_H}g};jHk}D?d690zmCek}-a6D|^dE-W+FGa4) zMM8SBuN)&eyU?cMW%Y_Dn^>+V#CoRO7em&wv1HKXB>w3(2HByRI4zUKq(p}jl2~i{ z`7ie*r`P3FU*v{t)r}OG^GFR-f1wgsXz!Crl(pLU9-gQBLf^taJ>s+7&H4OY)u+`< zaRu?=MQg6y<<_4@+#Z(C$#CWvc4YSL!GGUY?yc`TH>~l+1|)?G%n|p)&aSMfJ32Va zZ8-ZeS$~R+%VFZ*oq1EhZ1x52Hn$Te@m4==sneG`bUJ;vVf$u9YIm+1r8B?fIu2&H z4t%6l-TOrl$sK6;q8UIbcg4S35f+z~SL4K~lEGXIz_M!-3PzxAl=b0(yus5B1g+b# zm{fRRN^>34)AkzdN}%=zzOdRD100pBc-)Bo!7YuA&bpNNYdi)YhjU+x%3{nPF;M)xN?j4-A9u2eBiK@hMc*+ z5*BhJ{bv6+(v%Lvi7$lul#LNC67fC5&fQC2tKU=_q?hq3H)-iMbJcO&pv8SD}={B@8WhaS>qmn1=PESFp)r{fI@qjLbWrcS&#^a!z;H< z%9uVR)jG#kf6Ribz|1 z&O=r`#NrYI-_je536`-OP{2dn=kvnb&IO)UK%aj8cBX~1CV5i`)ker*fTbD*hFhBV z3&Cu$u$eW*#zjN3&l9qPFKsJzBOGpE#Zw^Om4uqutDX~Oewu1YmTZ@b6S-rNe$|!=0N~`B=`IpfXc4qaq?5_DW!ZFEZhf=u-(!K; z8i8{~#OW{cW*g}18)2u`_GVQ}{*bUd~nk&RV>KqD5RW?8D0Tr?XW!nK=0RI%_ zXK8SvDmI&KT6>O+c(P)nKMV0?%fbsgSI3KO97*#>xcDX-gFGWy;O8wX1FS87hARdR ze*GK>UjbN}5>Zl8O~YA-`1L}ww5jjWLo=DcFNp4QXrU5RQ_7UHHmQ!3^um2QRJmC& z{nE{Qv}IYk#d^oqb0_C>_X8hFgJvrGM-yi9Ffd zpF=CtT?fU&52mUG^@LJO>tA|%Nu+XYJSXMCrYxt>Jp;~P^oN&LzcFmI@M>&smRziT zt;stlY^Jg5id`G1Z~OUQxdX-|Ja<8w6}~1tkDFj)DXph+5RHa*UjaV#S|0<$VJ7b2 z{(VL)PdR^&i$+&#R`*n>%xT#b^fhuAGkLxod?J@y_EX$ zxTDxhZ{ht*K~CO)o1@`uG3UrnJJN08Y&sdnvl^Y2Yt1jByEW=VrEnf|t;Xt#S52AQ z@S^gQ2H$ra#Qm8DCdI`~@1zncLDyn-^^4Z2e9X;@>Lc~ysc{-l5AvQuO6WBp7O6*t z(~DpPGtQWTau~m#+K=n9NleeY` z<=LduGccgU+E+ZjhzYGB@y?mpxS=RkyFgB;1mSl!BW_=U2J*QIQU1iCq1 zNMOAiejOZo;l7`27IMM*T#>Svxz6TBh_NM&^J+G-kA&0NQ8K8g^j4jL%)A1KFGI|>IuZhXvhpxdcn`@^z z_+%vAHR=o<7Ill`4XF*(EA6+o?6H{>6NA6O{WWI`($l|0$+GHJTwcP;x!bMmE8*~F zjrAQ*%@fGLqidF1RhZ6?z92cejF{*qNaD9BKU-K?TaU$imWfzd>1GX6@|v?d&dxbl zocLBCl^!9hpx_2@$#3;Jw!?1=^O@Nky*Y0 zmY~(X!D|frT%b)7^vyRQOWNe1Z{jy!i{K+VH018om##+}Sx<}SDbWf?+k!Dc~^lZ?x~H4X#4d<8=7I~=RT82&W|ba+jndQpamjo?L(7Z2XlN=viK zaJJBeLe68>=!`dEe@XRlR%G@kMOx$yuaaQ;kGBd8EC;G}pS7Or357ei#bb+vz02u% zXec`eES=u|cHhv_QiKEn92G0EaoSbBRL|aPXKKp2Gg%>4X~9xVK365D1VxiurAf_f3;UG^*s!$$w5R@u`}-~`9@#nXDcmG0I=KU!K$mO`Y)CL z-WiV~M{`z8Aomy{i(RO%*+^?Ux}yLff83*%3&e7IsznVKBCnhN{e+z+;7&&uJ*DdH z$jsh`OM-ckxb+6ic60Yltw;JSQv*=6kd(}kwP8f{MXfgzjCkY}2HxO*Myk6eT9t|; zGHY1o)MswMR;y>t$|nT|5yok}EqYsPV0@W^6z}zjY+K3uU87w@>#gf`Yw)Ryc(2pa zldT$bsRiVikMg$tP8t& zCOcvq3-i08AH8E;D74Sg*=QIl5o;5mOz$9jDBVMc<@$o54lXI$79z~-xvfFW${1u1 z`Par;UiTVrc?I;kTj7y82V`F*ll`r%M z#?bTf6e1u?=@cF8(Sn7snVT3sc-(wX-1@GIW0AcKLeH_(Ss!Bg)bx%s^7Dx|^l!FF z<_oy%vdZc{>gAuU9!A{nmT$Lnpi^so4gS?*n=!=?8UreT9b=#$!|kwo{S!& zl5&xxQYM_1mq5(>yZasn4&qgSE8<5O`k%o0MU^Ez0P_MSFTVN{J-{>qP=^RJ%NvJ0 zd@{h`8J4;YV}xeP;?|by$>nF~(3HTRy0{k?#Re^AKLAw(Den3pRQb+g#bx-P?~kOa zwr`EX{{hHw1gf$tN;NUDJS7gLoADl7N&eK z;=|PYMNQ1@Fx8uaRa#YjJa+|d)1`&hZ7ga4^#!2Dse>$HM)F94iPO2%s`_G_a%?PX zI!AbG*bs^*tLN1ajjDsM2S9LevCyVu1eW1EJDthg{E1{-7_G|taC)xV1nxmb=#-;^}Yhd2&=|vW;F)9WHMr;M+0`{$()Z}xJ62`Lh<3QV) zWK`AtztLSWxcsBh_y>EHrx3qBNt!KYVWCOa@+sITv~B`G9%C0qHigDUU5frLtb$N8o;Nk+NI#PkQ+`sVE4c2z5SKm>enyK44}^&YZ71*L&ld^e zK$d+XvE6*GBJb)wdSiOy$TTAT9h4|HTaobrr4CC(8Ir%nYNBqV7Zw&(P0@Y4;qbX4 zbc{QIk(oRJlz*Aw_x3yq;$hR5!6&CM(0N$>iW z0LUhvrl^+j*1edF7ouiSkz+(#13_Xj%N}TFQ6fdN?ct1J?`oQD%h{ zYS{U7eI<7eCiTLrHdYRPH*)$M)B?Di`P*#Q323Le%`Lev)_$dVBVr{dXPnyYyR%wr zwS7(FlkOm5hV>GqefQU|6;}$p+wQCTb%H5>oA%et?&UfQpV!)ku9xZdb?anzRYz_* z1bH}QJR9@anTokeD~8<=OC)83i>X(;OG-v63QzOeKM_H%${Y9sJVAGy6E&93E?Y&4 zFSYeP_-&9ibd<|2-sxgH*U--qWv(gU%M#sHFz(5?)iwLs{qYC@$1ZS?o-5=IodPpEoBfxiIZb-rs^fq-$+)Bo zaZ!zap~HOXWPkNgoY8=7h3Tbv!!Zx-=;)`$Y4ICsqof{Eg%0+W(N$F^!cj3-i)r{9 zdoRYdGSbh5Z1WLZ9GrLNH4x_KM|q09-zfy7fv{YLb#ns&1T(7^U`1soXgU+?t}zi2 z2#gEy9ypjI=Rmm?n=>a04SNR zd;sHXGO$Jz;nBS^S7~^pQ5NKcCbwGODvxOon$(V&~wVZ+kv{b?Pt?AmKyTykFPXvJ>0?-I+D3i>jy6vVH>VXq2%hMLj}Tqj3k zVuaew=k1jI?puM(N}nHqEi#{+nr-$M59eQPvcQleF=Do?ewMw#Z3VN93uQDH-o@U#}GrbYcWfgZs;enxQk9dP6TH{`EF2 z*jP?t$U85)y8w;!oaSZazCh``;@-#6*(o>Fz!W&%aG&-hm^l#`OAP~4^~OR*qQ=yb z-zhHP!NQwB%Kv*vs=|538_s!_DX{=Cjelw5ae?uaB|W3kBi@hKy7s|^fyu8A?JUb; z|JV)5DIyAGF=S;N}C?@)Gu&mjJ+p;^UtHI}<|R8IW7ero~3w!FLen zczLDWwgEzwiX1t?W=&TaCm@X7+Gb#-`ffiG7iI2MmED29L#}M5-w75LM=mJs3=E>8 z$y~DhV1j_X&E&i^-jmna?`|!Qn!;c%*b_b%KiDmk88>-RMS`dmSLs?I0j86V4@fm7 zU_g_)w>H@DW_D}x5NtR&IHARh@m;Rm%g+)ut1t|z3OSO^oB3irW09NvRO1cRwD@;~ z!^~LC*7E7GK6$;oFHT|h9j;Ki)(q+b!a)*2W+L7Bd*JPzG;JR5#8S*qF0~8HuWv5v z_;AZ?4An+75fciQt?y#q-l2T9md~?g>_hY+d=`FCG7(fn>FjgjU+iP2Kc?blwPlao6rI3!>N-FKR*2V@})5$MwFDeS9@XJR1 z2TixMZ0P1f2j>nlIfp{M=_-_nA)w{Mm?Ab*pM<-=lcIEC41#S_yq*hm&r;Kl8D1v< za&#?1?)E;k?r1OIP=z$hJq1`@1?FGt^GbJ1UX*wCJFK~Zq0%=%X3Z4#br^2!IJ-Io z(EM>JD0FXT&huG z!>np6!>M+zDF0D*S zQp5ZC&pZti%_j1SH*tqRy8OWfdyXn_`0Oq>h4axCj~Aw1iV(cTD{bQW5Eg>q3Rz%*I$4M7O9wP1d*WWAqC>Q1+biBmgVgjZnanUVU6YT@?a zJhoXb2nYkH(1k7a{7VNzK0t`VSBR4V9-W?ZwIvADFRm=+MC(*6Kw1g;k`KIsc-5$X z5%8VQzOAFCj}xa&CEYQpkBoSs{EH|6g)H;We&_bf1iJG(y{(Ry*txh| zQ3{BAgI(}BhuUY5gRkrz+=_qL)o!fx__X8MRg-gb)NTPkuh^g*9sXB>n?Cy7#GA=^a5Q>D3iLIREDrMt> z>sYUL&C(iC#T4g1ZOL3C)DIi7YArokAk<};0Fdj#3pBnPsv>m5C^%cxo;Z%SKJJrG zBFAhFz|!(cM`?M>`j{qi5QmDW8)~Abvfo%-XcvP7Y*MCrmWrFPE&^)^8(hQ>trvq# zFA>JvK`SG4ilh!bTeK|DY;vv*)GZ=gb1wC(+!(sq4!P2_UrpF@JL)o`kzDX9AF#;1 zb(d?f&CBQmk}Z#Ey3!`3YmsSpj$!-c%Ep2ilYT__?6w&wQaA#+lTC!WCC$Eg_Kxw5 z1Su=67h)DfAk`WCM9^|#_R{V%owngML%X>!o?x7)k`k8_3Cj~c5!Dk3-r0D7z&ZB>MI!P60{ZD_^L3^43EgR_$;`&hS1wJvu-y^ z*KzjhwYRq?Jb9OlY}^&U7%8Bsli%!31sSj$F}SUupkVu*)@LjU6VvUcaA|>ug7mOq z(!66Y(@>T;CkYpk-#QhytV~2hB`N_mtIS-)jpa&-HCVhZXfYVe4WEI&bYp{ftmlC| zj%n@eJXW%Lv#3g6+5wb>gby%X17B4@qfJK#l?iD**|Lq19f*u6LIaU!X;hK-?i@6C zw?Kua@o()6A{49#Hb~-IYBjqeoIurwdY#cMm@#w)*nV~XV2BME$w2NOb9~`~b;<%& zd;p^wcfWYh!{wYi!30Y_%d2+k2+HtCKB)UGX>t&+XIjTEn+#I5E4)q=&{s(P#hk7C_J~H=@RgPe_9M7mu6zRyz4_-PK_%5)B~f@4L$CF? zt4E}M={7~4YsUrQj71FbKt{YkQD~rZ zf#gx7V7QbUL!JuhvuDpL+hUmtgc^2knLXDJyUU4(qeeVy6H{A*-6hs-7U&jR`$SZ$ zx!F6%xH)N|ao%Aa)C-v*fv}^$@y6IX#oe)HeZwmqGW#)5VpZTZtvV2y-|WZ6hGZSY zuAPL-a0(a=j8D=}j=hB3fIpPaC=1VSf-1LDEAizq*{NftWJKw7CWFEi;_ zz+mR?l)4*z<(V?5`}vp_{B7_gA*Sp>2Sb>@5MsGBX(9q*N5|Ko9q5fPc+%~hJ-h6N zzrzm7a@@=t5#fV7{!J`dWAtHK=y(*Qo8p$6>{*vb4gP56(0Nh5%u@U)!&97jKv5>` zQ{@{LBbIMVn_I={8opQ)5*jhBghlvkuO^U@LxJbAvD4M%NSL=9v9s^)S33UZ2nu0f zql!(Ib*Gn0=OQ99jJDIh0E?E}sgQx4lT!;#l*_1?y5mkoFrmQk9im>F87xUj$>8#~ z9=Z&R7_s620I}fmm{!_#ptST$Utic68I>|)NjkA_w+q%)ip#iJZMRzlqT^O=xvi1} zuohCWm&cp2L2V+o7w*`7p-jPsoAwJ|-kILK>GAUQ$YNKTlAZjCtZI0q_Y&&(;G2o| z8MWEmK-{vjj=+0E%+#c%u9|CU_WGp$jWHxjoQinlY~@FzgO@8LvKPNsx*z+uwZ65 zs~IAE-jEa-+;mpAzIZ$-vC&UyPJm>d1MhtwIR;uSBz{_vLR-fR*#y%Bl}6PKMGp4nv9PYlGym>cKxyd^hLzRTu^tcUQ<#`l zK9r7E`PVBOhg$fYZ=BaPXPOi2AM^H}!qp0&+l=pRcl^~w)SyS-dD|}gk<|AeXgA&4 zByq6lNy;#<-b|g}!ae^T0MyELUYAd+s}r9B!-6#j^~F`%5)=4<<&UK!`S}1KApSmz zx(@dC7GbMaRzuK8e$vwIT`ITeNjHRnfkCC5GBd6LECma=+f|tXFjP4Mo$YEIE4S@6 zLLpR*;eUCHv?XkEo~HW|r#-{3Mm7H8y=usx2I)Scy!y*$2!c*QlR}bY{L`>R+61X5 zsLXf}gn2waM@|0=uOEI5C_*qu@*>t+}eaKKAwIy%Gazmh7bfkH8fUfeEq3T?v@T2yi2b*fwDjQl&3t(5>i zCG`4sWIZ+z3lh!*SANL;pDtJp(39Xvl|m1x4T?9Q5fyKx5_FnxG5!U^7nNL~CBV$a z!^Lq4*p|M^%-qtn)et#zhD@-O^TTv9;g7Pscu~|2BqzHeU!2u_>gA(eJexYSWst`I z_}<2#)k9Ypem@AVhC((R(^j_y?depCRy|&ncd3lYaRkM z9x^S8d=oNJg=xGvmU!NZ6&g0OT_Dwt?uqqRuDNzKW6nn5RrF-7l`_Qg=3=2YNnxe8 z^gMQ`U?wv<@=mbcK_*1PMd&Vi7Obnw#k-}Qrwrsloja&uhPD2Ufi8{A-RC19X3FtN zpe}@*2X+pYrWQPFE+yDVahg}2+Mr;9FTwWOpQ&z;Q|ZA-65mN$)bM-PdDz)}=)(N8 zC#blT`RNqoOSUF>iKxVisy^Nq&EH43OeM=;SIoY!~5?u>q?Ldb=}( zo;`l3aUV0toQ0-KpJ6skr;+T9ojfvw_|x2A!w9N&XC2@=hCdZ7-Je zT-sGtF`B(_tGZRONsqxm)sThCs?ujV$%w-4knAUk4<*F!GDAX5qklg2^qT?J^iiSU zrj*DY<*Y;e9uTSbIIX2xq@#T=&NZ)4b7Ib{F()^dXhk&_=*{DqU@uW&meJKpM*CM# zci^8fqQM1=%{w$;zJhQZKE%HerSY4m_bJ2rNQ1l3LaW#AFpR@H$4;dag ze;9+0wK*8__F|l^fjx2?=U}%(?8M zb&yWjt9QhDR6>h`8{XY0f-gwvyakjMs6wi0n3RSu=4Q;74}IzC3hXtsvXZlHEgB$x zWG9?aAbK1+5Hs}AMI}r?vPNmzze@iG+KeHjDuwku^oc3KW3=L~!oqgdmH>7>ptB-a#VybQQI}561u+ z5)yn&OO~1u=vba+!M}D?hV*9{Q{Hjflc=d@Ys5F$9V1)HtL))^@)n(*`7K9j4oEj% zO#yY(eIPdDN}|?7eCpIG(Y{4XUZS*a`@6Q=sU|oeKzya=upEux!c&h!EP&rw7<6Ag zTCInsoa}!|NOsTUuk=#wVXxeVwejzg9;V@@~c@J>>r}W&BsYB6J4yWG8(N5Qe6>Gapgl`_tV)lu&>@ z0<2L&ADgREz&bk2q_P@AJ#e*}u|ScjYa>eu-JkyLfp&a+RYvvl+Z+qsuJg@$2ff_C zOOnH#@RZi55`(=mf@E=p*%z`cqE)20Gwv_2JRki7IR%NY)NkZGI5MnC(CHqaK6Ys_ zD_ejI)jgzoOmE}C@1gh4818I@d~uW$slPD{&`uy08CpdLc)8R{)SIc%qovP&u{8f% zb-5C@Y*(Q3)&8}*MAJzFCS8GC>|<{IZ+)Q6mpaHQ%Oh>yFu8*9pp1h(EB&~hM^hM@ zg?y*WIDP`Nw;u<+NyIV(~ox{UJ3PzNyx z4LX|V60dRjI4E4GB6a^;saF7dF=X-(z3GtrzuAu#C==ayYa}On%AAtR(Mmsru?4K&UP2o?OQ;&Kc#Pfo&RwW6zkka z;|mw-NKMMhETUeVH~bG#|G%4lP!ZaF4aI2?k|y5Xx#U$;v37X}s(|!~dh8$FwfmHP zAyjzDUxwi=$Jx7oXJWPo)MQOPrfp&G;k=iFaz5C_aKztO1~9xo}F;psEjdm8J;mhZ#aI0U#V*`b~R}E_;?BxW^R$fbIne z0}fN6N04TZMi(AacPdhFI$Bxx4-n9#lqpXD`zk$pJK%Djay5}j;D}gVd8FX?u%8h$o&U5Y_mtS37&!sDH zocJMMC z-6Qj3`6FjRV?8B})82z00q^U4k}U4_F=h1nX93;8s(DzMmP3!h5=)d#Jpp|0St>w$B-SN#`7sbqJGS;wBGm9ryV_ZgLjwGF8h`vnNc%LZZF_<1 zNhQb}$)0+xy4R)vy+^$%d5QhU*Zq2-frataYwB|wrk0kZ^MHS>%q*Guo zQtek2e?&*|y=ma=yC19g`2tk4%h6wfPgPvZebjiQDEV=B`Dct8!X{E(ZK5V#Gvz%f zv+@~2fI_o97ff{Iq1Aw4`vHC*y!6%zHmE{LWB98+Hbh!){RrKKP40X%@oP$rsQ zjy(|!T^0uP-!vteJ=&MgmWZ-9mK*li=G*??);cGul$`|OBNuI~ZB#!-wX>9Ix}lu< z<6YfgDhueR7e45x|8*mtmouKH^M7Xr#w94(axopwHv8`ze%f(p8~Str%oP-0vONR* z#6k&(MjEq+Y!cLohNh&uI-6~K^g)n+Yxe_Gq zdF^=>1u+twNfXN@0c;UXQL4OV20r-zgthz?3-!>LGVgUrm z_QaK;Z#F&gnmh+3Qjbwv-{3BJf6 zQ>NVJCUCoK-Myi0&;}wPy9?F_4vLk3As6TE2+)8{lvdCts6S!3@J#y|S; ze*PHfP3)$Us9T0Rbv`umjq{j~-WNUo^j|UxZW$sz7s_N)tG?G56r?&+TiPXm7)J_9M4im(@J!|-PiEKz zTB7TySjWromlK;UMt>QXf)WH*Fvg8v0~D&o5Si=@OMoQVN>1Fw`(A1<$!0a^eD837+`9RUV@64Z{AdjekL z)im*PIdXr#?+^n?mQ$9E+!@I4Kk}9zFVQ%Uaz<-muU3j>>(xNyZ9UKbFv6kp9)P+A zTXLlw;6wc;;)$-l4GG=xNxPXP|0S^eXp`lT74Dr6au9aNj+@VEdzya8E1=U6m>NbWOJPzOzLC)?0|)D$UW0 zI^;*8k!lnklmWnlx{i|Ntl7Y5AqeR<$mKnf*9f@r(UX#rD=XFYr25>2V0VcB?jg!X z2{8xdL|b0?VqB3zw9%Y_!V76aog9o%0f&r(!BRi=^xz^W4h`Ex0#8ezB$dx{b~-q& zDC=PE!0&e+ZmgdIs3Py*c-I-7l$$hK4s^q!o74|>Xz>T^U_oP|cJw!$mJH;4wZ&oO zelAszCs^^XBLe`h`_mC1x2NNMQqNqHvtKVlObh8bfU!^dmzYo#G-U?%WCXUA#W5Qs zr7;A5pPjXqyEeAP{rAoJ2TBl<%l@hvQjFq+aL7fB>|2+EI@kA@TE`p3;ub6^M+-}$ z>#XaQTc4fqKjB}69hUy=vB;mNRfqxfubO$;=@ZM@@%<()wYVv6)uC5X<(?%S~vabQ;x zJ2;LJ_?`etD;ZDDO76-Q?kc_uwEg9k>BA`Dai&yxMtLzD6s`N*tw_0o?B?jwpHQbf zZR7(6s6zJ`2&#{gsi}R}jy3?GSD_*AmHrjS}z+)uUvhraW9THhsEWlVqnlz+@5ZWH~@Y$ulK(Qty8V&<%7D!owCUeAdwd1c7^Puh`-C_&o-GBoBwz zG07}l8X0=vY&$Ka)L8sJHT_p6K&O|}K)~WNhOQ)w^&>#a5}!75`a+J+0i+r;aBjxl zv!ie@j`$Ri6_B*xUF<1+yn7_1!Y@CG`m~3r<^o~2*MZGO_K@^I29)dx_GFDvlmxI0 z8qxJH9QhssZU?CBWBSEm#2WN}hVrOc3(H@0vHMUqk?vZ|FhB|wSmBeNPf%wk3J3L; z=X$4)ehWyoNLm10;CnMmN(%OOmq0eFJpZLqWMJ?Q8Iy<_de&K4bEIeg5A+7cSIMc* zw?I<#2R`{PDQPtCgbgb zZ&)K&D?VqQlE!zP1=(oBjmhaRmiI*t8(Y5{qSMpA<-q&bX=!SrrRF+79C|>{7<{aPQqF8(*-pU#J8IbW3_mIR2_w1xWsq$V2q>6}JLh z3heh2Wcz!tYN;j)D#yoNkXyZF_MAFtM{9oQbr6e7AUvsr*%J#9P-Vv;Bt1G-|6@0Z z7&As;7{UF4I6AYGSeT`bj7_S1wCrm92`}Od_u#m%F~|sXpU`PvRqgLD7~|yR9BnzS zbD@CbqdThn0FdE*Qs>{z@8{)LF@Tu0Dl3Z_rF)rxS(Ph2hfBd?5&H*dhwDBn_zgpcQhjUf6GG|N+@td_*1q^#%M4b zPi0#h`+_LL`@KU>cO8tWPDxnK%B&%k#I+EQIqX&ZvC{nuQ2MJ^`stF{*R?(85yF2`_9p`?tdDQpk53G3*uG)hZAyrS^QVw-;$%MgnlqCxS&ffSFA{Js@+EQ zh!8eCQL~N(Jomkr_+y_COo1&cVs0918;=}o4<@l;Z7;RS%Fb@7HfR}oq^(mfL8MoF zm50G#8#|oMsvdTXnmZig;9{@maRcjPQ zVkwtF4NoUSAgJRhxSpb3WPhjVn6joIQfRz5X=U4Xk0{On;_K8|EBZf3{WAf8#WRhv zRiaCyGv+PC&e6KF0T9@Z(u_qygQ~DjD|e^;2`W&7pX+xed?|pf1zGva-JJ z1`sXXT6dNMtB3;C$p@_A%}+rt*g5c`*+6eES1O!t-4Wo$v==)29(9sgmoZ3LBmp*& zMHJUc?utuKbp%7pPLbZ>wCiI`7Zx<}!Aw>hJRFz3cJ8==07QlWfFdEs$c8bcOU@h& zJ4YQz1%>-g__!6+yxRG&T1k}6;{q_ zg7^3Y1F{*CmD}T9mu_6WYI07b5_=Q_hxx*~t3$f(w&CDr(wwYxNqP_Q{rR2sZw2A~ z!T$0+sMUZh}KaWLspo8s+b7~Gg%9nPTb zBKhFzTA6+B0x{Ld?UV_I>(@ONu`c4BKqnCT=fD3zM+v{*m8!RqRy>Q=NEBpNr0wW< z`zBlX`@qmEoS{YqBIZtSg+!fQwYanzWJ4FnUL`etj?lk;i?1M}{S;I9<6ocl-*bOl zX~lr#|N_mu!6mtF!W(w8* z_kn>|CLkN+Jo@m=_W=Pq)Z{jjbuWGRnL*on2i#3ebS#rL!e3f#%r2E4C!SwK065gU7rdvlhP}-h69txMZ70uQdNtpp|+DofRmZXRa@9>fnwfRf4i>WN@tgoNeC?vN6!2em=oB zuhCWbuF|_a8nG=ZvNJ^u{S#-BW6=&D{WGil`ZwwuX!JpHC9;LvT{VM^m^$m2+0N-A zoQ~O|S#Y+KosO6H2t;jg?X6tDL1^=}P%cj9>0HRdif>Lz3g)Pcf!hEwm13Soj@lsQ zq_6;GdZ|LTuVSmhqpV`a3pxfLerHtF+cjoSdlp?AtqxQV6fvTBT9KUj$QBXTHR_(A zuGnD;5O)IKwcXepfv5X>dUC8zT0OkgW@#=TjCs<*^|D}dAp^HngnpzNM2p0 zGUxqv*HahCNdro@l}A@+`R-`HlB*q>yH$W3svYxaWPY;sS>^iLxn;4gh!eAU(d7S= z9sfaNpqH_;OKfUw5pV5(84q8;+4?YsN%7V3aX2@_86A$hG3Nk<6*bZH4Gha38s)_l zOU$grss}w%mk13H_j-oId*w)~2~D4AYHqH7>2vGSGFT3xqFp*5nYru^wpKr#RDwuS zGOXTiB7&w%by@$uv~(BWob9@kK78_PW7r**x9@zJ^?8PT(kGmjhbb-vao9iHr|$7w z2e|)V1{o{uU|F?9ZQUI1z7jQ_C~Y09bz4Sz886h4%m_gD7S&ISQ)Mwv=S)@Czn{6X zLRl}%%f-cV@(C-Pm#N(U{nQhXtye*n{QcQQ=>LwX$RL>V>u{c-qJoYm)o9|o(w!llZ zOp--CJ{(u9-{sOT0vNRx>I!g{I(72yug{*0XDccaKxn|m$)43WJ)^%vgZd_G^h$O8 zGV8U{kKsDHx)--3B_u93od4rFdtaiVjZThNgp9M8uC}NUq*tCc_vsq2Ri>(pFIk*> zk1a^gY4xv}As^yGVtOh}89&GB0rLA>wehR=C&b+CGLo>g#{NqF{g>XugEllih}$YO zZ>&atp#)ZX|ISI&O0S`VI?(T&>+ATW_M_ukWXG=3>I^u_9}@D^2W3aL?61=^erc4a zO$aOQ{PM{JDJ&9v+Y+ti=JIo``3#mq=;Yqk9?5_|#4v#Z)G9FGUZ-vYY^gZAv~(Leo}v*abBmX;nR7H4&sBZ8UpKJ7$3=D&iT{p4~R z>O)fgg>sy&n_k}3az4YgSfc8Ht01S4>N}IdC_;g8G{-2aqV)@uxZ-kjbdfg5yA=14 zUuS0S2rY8hTCtFdAu;vqkz`rIdH=5Syz7IbpYYS7HC!-~(iq=dXyWZp!^j@^dVS!m z^L$Zdp2rmUH0Uer6cFfepmLNjF50!ZDZrMk_q{FC44_?3;OO^) zk>%gNXFPp8rcL8&{^7RWR9h*fk3x+q;k&|iV&QUq@XZv%Bq3+N>7i8rf4@8GN?ul@ zLIWnB2dsvd8Ze>~aS3lgxvOd#otRmr1)2%PMHW6NwQjsy&oXM7cO&%0#eyyKZq;>u zy6HDM(AdJH+b!I?qeAbbkA9J29NI|{v6D*KSvRYvOd0JH?z&*OmY%#MoarixbrxfW z=3@NpoUo6+fdBtPz}8=idZE?gT-xaW}Ed8q$s8R z%=M|M*)I4=S~lZ6L&&zkrl8U3rUZY%~I{lP>a)kIG8|{&AO}PuTGioO! z^xcACFD5TqXk@Pt# z5m-7#o5BfAQcX&0(R;hAv3feF?5p$YjC5Mr+_TNq@bw1<4hay2?Ck8N9x{_{h`D(d zjUbCc$Hc=xgz?=QD=(jR?7z$0x-)T(sId&K8rc*rkxD+^GL*tU5R z!AsrM#2LpQD*+W+%8IofrEoKuIPyA|;Xm|0FZhsW0%)St*f@E3uKPZ2e;?|5g{NDo zk(7m)q!lQme1qHkJ#3|wJC0`DF%G$TUr5ho>iT9^s7jT%+_zJl&kjlNqOHD*@t)Z2^CkCu%y(<|$bOi6LMv#UU4HdDeZ?yRtXlSq z9TgXO7GC1%mnKSm&AxX_*JEF1KR}mvqE5DtT8?W~XG%93yd*8K9h2WrxeJq&lG4X- zk6$sLysxC6sbb$&t0f|$sz<1O441l~MOOq~T$`kf*3r7K?RKgr?P~XzFZXUI_*6UQ zIvUkqD{Bp#2@d%5^_#OhOEkBH>kpZ@Co$hUV zwg(P5(9zz4*4bgs)9#OQ(k{zTtVadB+h4D>QxiPbDDds6g(@{#p!?4Ve|4rqMR_hg8%rm= zkB8Qc26A{H75DE zB3&1bt=9OTyFxrBS<(W8EQ9xWQ0(B^E%@Ch<$dq1=O{F6KO@78MiVG9^O*4}Rt z8Yl*YaDQ;h!9JJ~e@x+D8tdhju39*KHLe^OOV`s4P6X<>JzzDvs%V}Nh8)Ts7152n zwCIc*+)!%YC4m1x*&(p|y_R7UO>5~wlHi#dqCnO2R;+O**7B52JHh#Lc16VKTML@f z`ob02PaR|~>a0_6k0N>Ig1pQ@Ke-3i1@Cdad45Kl-8o?hv3xeX&|zz|LM-CT+D0xw zhSz>1|Hu%=lwvHe-21q-n3|2@l3*HbavIbd|5jSn&hkvycN&WT)cIRXJ@b^g< zl(M@Fd0)*mh+5rqO*9drKYvLeb|RS?_Win?Xa{Yi z-9r^ZUvqcj{0r%t|C&EQz$fxHU&@h$XxiJ08W)XF7K}GYH<;BLKKWQwraC*aRy4Pa z`HtDL;?erfj(B=Kp>$csmj%R{YSH>OTLfZpwd=ton?&*nS=XLO?*^m>YBD9 z2-2W{gmg%Ew@3>jNE{mJmJ(?W2#6rvjnW*t8-3KV(jJm3eC)Bj`!Ed(v;D&bxiWu>7-|i(Gn65auLNl)ytz-N8dJ85OUzz#&*6 zrNY=V><(7@z;3bN?UnNz@bi0|)3LMB4KI7Fe^hs842&Ysi_P#JxGfZJ9 zCixkY(mhF9u$gCBi4JdQ5tk8Qgb;|I)4;FqhbZHO;GSYV?w!u3*tcirE8*j2NT`W&|vBel1Eo*3|wZFe#wIL0_jKJd; zN@>nZHp=)czFA4Vw~G_UH!781Y?yByDE0tbKAtIT&-b{A)eWuc$sr9n0|_^ZOu>UuF}<7h7InxNC;c;iw~C5@eNVV#EpY-2k>G`hHvn4| z-2;Wz3D$kL9K+i;kKSP4!R-)%0}C8{pgj&+S~9j?Gi2a|Ex3tiKGqf+^iX`4{bu68;M2DG#A@Pl zsf3}zCNZ9yK4Nq6@lLx8UT{a&lZFE32+vU1vKAu~FIKUIyAHh(7kZ-o?0y7 zoZfTH}$kP;lwdUYMfe0`sJ_J+oc;{&lB|$_s~+D z+-!Y)=v*sgGh#@(E;s@5h6=ywr>)AaE z2feS4(iNb=s6e3OlZnsknl;rEp)MzC$RcspeM%yuklym-DRha7y41n#c{}*x(it;I z!PE{A?E~I*Pe02)wLv(+QKYqvT9wIYv%vDAVdmXuXG^)*{w1afG?A914}iKN(fGdN zf#=N99-65ip5Z9kDK2Kq%*-8#9bg>>H|W$5IAU%%q4Q`Z@$s?H1Pbn{C*8-#Gnp^f zs~?($b2v-@gT12{3q$MCxg+l76i`sBTgGmTG)i-vSm1t*2Yloak> zkxqfRZn1IZY7Mj|pOM0eGMLmS$tDZO*s&`L=Pabt9F|rT2wOhw*&5@m+_m0pn+g&d z?X7`5KoLM&Gb~x0*~?yujWEKi?5K($hocrI|OKMuc(A%df!_{6+ z-w?6h)T-1AaD#i}r8eca%*>{=?^c!Me3UDtDYi4J<3c*F8EsB$aw+IB;ad(w zPyv=ht&<0Z3zGd3;>VQ`DI5(gEjGJ`h)Yhg7KEj+vUoLjndZX1F zrIC!-UQj#HfphClXx6S@G0N#Ne=vqS%KDL64Lo3`4G;) zzO%D4AU+cO8Zfw}q=<%n-}J=c?Cflsxrw5l-VPI)OT4x(M#jlDurP8%MzVu`3@PoF zflMcg4mO|~#LY0k%HzAX>?q2oqH8)&y!QgqR1Nj=fM(xg?~cmb^gky3wVrH6&QNx@xYw%skn01Y)R?Qd$AK0$9VK!OFVROjt`^81T$Ay1+$D)Nt_9k>sBH2O~D%m`IMvfo(K zUqAHY>ed~u9e)KB5Ltr5IhIJ0$)@~5u_epoLGMNhkIpYC)PLv|pl z=}UY4ITrnnSHRbGSb%URzW*f+5j!p}Mswdook915*9T&{Y47G5`P2dNO9s31_7zJ>rxu->8XBFVN$ICXecj32JM0VuM!p!MSl7_(&?l(P(VD0s z%)F3F)hEKc6hZu=T3?3?J=Usw1adwyw>D{K>H<7lzdzctpHV7>sCO;+fs#R5oYOw; zc4I#yY}Q=afdHMdhHHVD$rSO@lDOtM2OA$L;E3XY+Gy^9n(`CM@c==f)KKUQUfqjk zu`{AHBUV&?IW}A8kHc~8ij3_Vj1|UnE`RVF!oWTI9_%V$_B(~#szjOelK6KOa@rPo zLmc`|3kmavb;}zq0@qi@&D)Y zQuH2|k7Ja+(^-E9GKJS;}&^drN z;xtzXUyR$kzGi-^am)u$#B~)b0#-ct`pRoH0Y0Lb=*v!F=c9Ds6ryx;dipMFb-Z^t zty4kN+qQ>xp*|6bZDh_*O|F{b2wH8lcGs+CM3j8>_Cvy+Z5SEP;kjG4X!i1$$;;Wt zghpq!y7me14kI!2lgyrrf!JM*w9UQ}nsI#6YE2)*=6Ntog!i3}FqBGt84Uh>5Zn_@t5MNs3P#k9$j0CbU%qGjTE0!vRJSXr z7XIf7fIuYU*gCq?aZzuVB64ZA+5S;Gn8;S*EHN~w>&|N+6;%nl#yAAJqwt6a!~E{; z<`QO!u0`6$Cl-<27wbh6`71TlqX+RE#)thTZIf~Oqk%WC@WFs(Xb8FJFJ(Q}BDJ47 zb0^>$Nn;#d+izX3f#0=u61<`CV=sUia!x9>+=^uuY6cqQr(3vW;5d|p9=vqbk444J zMpE9tf~sFuqv9>}vUfiCQAWpI7P{xGwgFzvsOy$^u*>YZ%;gskbhGz+&y#WWL>7@z z2sDfWfPcPZu6QmYYxF$$;Ls16i_wYyI6a!c@hi<7&2#&P7QuIsU9t~=&xDS$ zH`pbgdGdRl)SQmR_I(=7QP#-F$p+4`XS4}<308q`rgVkxeEmO{v4z=M+eczirjS-v zC6j*FO##W^Y!Q{ZR-mUi9-7x{lYraXx4kNR34x;kK)SFMO5W5L4-2@4xl`7}Cui|v z*(Xb@rfE;kfn$P?OK7OMBYk+&B)+jYe>$g>S3+t1)LP~HQVv_;oO4)f9WuGTZTWWD zjN13A^xvxk^=5KLPAW+Jn|Y)|*1|N?FR%DQp7z|ItRcK$?U?WRfbi!pLcy&7Gb)~4 zptVgx8O=VzKE+4R0VQ5H`IF~Kdb*FBU0;HZy6*D}uV$sS>OZjf)4vKBa<(t+!b?Ua;8=;ikuv>r*r(Pa@9)F&q16eQ!SBQ76)IHI4LY zCYAYMM;yfe8g7tMb3EvKs3U;|e#3~>rhOdssG7sPaxU7QkL9@I7WIWxurcpRzR(`@KZ)XR>-0)7lD*sM-SHCFe zB>Ex3yl8YTj`eY5Q8$;ODr!kb(!OZ=mQ|#l0Z(dDl28o(3I(UnwV_`8*{zo%*%6A} zs(Cf~WfZ6MRk=45p3y2D*>L}9>aXjrfjPD~4ZH_QX7+H4VtEJx^esDaS-ctb0=y^@ zG(-D!&{ha5o9F%WZlc{|+}#K?YC&5g;&J^NXJfO&!YRmTJhs!-M9?6RtV_ReJR@Ma zK0>{*KJi|8enGE8T);L>vy|u7q{7#PoLr;<#$q3bt}jz3L|BO%5=nMS07l_&AbGK>AR#a1RZ{5u6l<>dx6DMl1S#xPKUk1tymv=8O)cJWJO|j=L+Xpx zP}Kh*Hh&w4mvT?}CEBnuAP}I^Omr{eGfp%8L9b0u5_aeEU}2Hc8+u(Uv_?Ak_g5DA zz*u+WqLn#0atXQMrRknJB}U0&|<|PqsK5KHz8kY1PE2FyZ0zR(Hcm zFjQP%G%byYO<&F!6<3vxztyrw2J~^Odg?u+W>G9`;n-R;yO^Ns96xUa|M{oA?3x%j zOz)o0;l0+y2kPLubWy-x zqyn8Ea$N0zDG7xj+)Z31Z+oXA_6EVoDs$zcSe1n~uFvJbeg2(~7G7vHVHxB?EL=~C zoBrfmVfR-s@$)Jh*xE_Uwg|J;$BR2NqN8P_qq(>JX~(uS zNYav56ir)EMjuppg+E8`e~mOwe5O5rVVsqtVI*xIHi zdwyo3IwMK3$WvR16$?7D;5l?46XsqQk;;3w-7aqhDUu$~8H;US=qh8edoH%`lCCj; z&x9!3&c5-G{tEsQz+DI)Sfwsl(8;tEc@q2FU^1E`u+cGB(dPCD!rx!We*i<1#Qc12 zanS_q)@+szE;SO6mVSdW+2XOl#1?f&thpJvBw07Jx!3SfgAXwir(L>8V+eQ~jm<-r9Mi3VQ*$@^T3J2GAXn4$#iq7Zz30 zlvL@v`~9+`lBoLnw0yGV>LyfjUzvz_8(V5vv6sK26|FR zCddRdFIln0L77%@$)7)oqVck^J^N}}O6er#edo^H(wX=ByKV(e3@BqCt0UN+={4Tf z&VFoR&2qQcXLpy4nvM=(MG-FqmyPGW&c~Mq4Yj~9!Iia;)gA6xQoMXigR}zBR73-& zpZgz9czP)dqx}pYz)#VE<_s?)uH%h^UUqnvBJUxEZ#``)+}BwObK}mkV!7>gBh1b% zJc~`Dp_xnZgw`efq1rhM!tqlKo zfZjxs?fv91(rc(8T9I5m;1M&vWczx2Z42$wC;d38zVc+#tbCRJ%54^0e48U5Yhzpm z-I_&2^V1oTH4-*Iljlw|>cF9AZp3n7+lBIyMWegcGY@kF55EpHYnh1+>~}Mq$+O#v zI%H}F*v{@9%?;a6kWKQtIF^ypx9$T#rDj_SI(x82u6d9I6hymuz6{J{V%yc)yZMm_ zDeFhu!Rp^-P-%3#?HBV}_Q%X!gL#5PN6EU9e8EMBU;E_V-iZ3Q$wha9MjV!vY2ux& zJwXeW!;&pMPWm%eU({ATbbyY&;AqXjeKZd|$sUxRcjyz5Ha}Kyx>J_J;-Kxg2Am;!^|+}N z3#}XMFus;qpU3_34rA#@#Q1C45a^!Ftli_qskWB&D|I$pJaH8C`N*(>Z?d;-{g{J; z6W9BtX^WG0At>4Hxdf^`Q5W`@oEELD&qqkMSs#!_1gdOQJm^SKRe8yly_P$9(zwPQ zqQmiWs(0t!#yZPW^D!e^1hy>gX&GVRT!6vAB+$eAigOID;InYrv&p~cZ*J~o(9xe> zg`6Jeo}|q8ej@%1fjp|vM0!RKMB^)UBX3?#dp6~Iy8eJUz=T{3NFN)y^4dT&3^c)M_$CzmEysb;RpVq%b^Vkd83Xeq$vFPetmTIT7$^lt zI{m80XeM{`$YUncyU150tU}0es^h#RNv#2|$!Y-vEmse9O#Bi!Nr$_%f)^4|F1FI?kiin7w%C@g!Vk<~}t;NG0Binf!PuA<@_63uhp+C9CZ2?rO`>|gPC{FI}Qb6Oy#q@=w3 zc>{45n%mdT84~g#Cjo%-5ruQP8DN;;)`Xscm!X_UlYda_OD%_!$Yd>OT5t=ZiOTQrxk*_R?lydYxoT1lB4r8B)0vff_S1~Oa)-VuJ-w8U zMWB2%prpKP*TD@$ZYtel<=8)wiP!-SI4-QAjXMHq*U;dMIT}O{{zl2n(g9>~92i%V zJiD1%t+q<@O?PKt>!+P@;l2{_JMVyg?;qjaALGF&kB?PB%VnLF%XWqd zRxxu6glE={hRT5xF4G_1u|(cP#Tv=-(6g-iVh^UdSZwJz2}%(~fKiN(;TVus(DkgV zI~FXMvQaWHP;uX|+;ABo>daz^zZsDH8KXi0=`R1v41PI^6`FRvahOqg#l^t*3lfq1 zm5#g`>K^0k+5N1n$12cZq;rR{L!qYzGMssBJn$9l_VU{?UqqJLoppG)gHye8*Fg^e)TSp^yE!%y zIK6)c9{HmA_uXdwwiv^c5v4(qxl)B&{8B7+Mge$*M*#(s%K_rVq!Vt5!7c5 zDQYN^;5eU>VMKCqvs2JSvFrkeDI(;sgcTLB%*W6QX6jqkD4=;vmN?aAB(~el%LgS= z!I#jbXyy$+k8huIoM4Vcwo~!~T9Z{=cSE(>N6I^!XHN(;qog{(X zdXHJFOYI3G9O%E8Zdx_z2pK>BT#Jq6WjAa z*%ESwdr*`e5cv4eQ~aZ+%5a-q;!&|)O1ikZ+zX|M4DIIL_xViL*vf-ia3-2(OVS^5 zH9|L}&W z;*8m5aBL`jV%0*w&dROz+m5c?(dixBh6x5439+#U@JFS_!o>~35SrL!3LMv`0h5KF zdIgJn?YOL>vu9n<9vq0_{xoW!n{cGh?41XQOyA(%OXr6?u2WTx99^=|6yy%N#KqWf z-IhJrE+R4sMB!&_)(i6gSh&}+>rh#{r`1s{Jh;aT-MgsZgrPaQZ>7uSMou-H8S?)RN2ruEcuH%U&jQa(rjOT zh1qphg8{j-LVOmAbl*i^1A~D-3usdNwF^mN11(FcnQAL)G2Rii$(Z@1vcdiG+}|I06!2ls%EnXc+~oKYHWxK$>YwTd*Q~Qf zN_&+~8wIAJps2o>upb%agJv`mIf1P^J8h^zvdeXOkx=rjvygZE~8XUi?mK z$Z87?iD1tM{&1LQe9SM8`0MV4txu!CDjxx7*3$=h=P7y(DuDlTu$d(FI|5%^q=pt< zfKFWkBja#jbxCwgOvJSdLfYQ5VGq{#{6MghI(kWo=O4$GyHHtnQ@g~1U8C7RYY z&qCebWmhQ!&3bx|Smm{rmA?!RD$!F5aoC>u-fuY}B!7mM$<`tza3MoMHT33xBDXOz zZNx_2xacLx)flp(r6B5a?K!n`?TI_J55*#N7otIg72NPU&wN6&9>q?T>i&L&0l!Xx zaqXf^{AMDH8(D(t%V`mjF?^**&9v9?|-& z5N`RRTs|w$FV4!&_T<&-nQZK%vdTz)q8mS*%5RS%WP&uup&xCboz_7wFbpxq|4uyq z_L;wQ!Spe4S0>OVuR>d)R!Gm8=K&hO$Zjc&|Ayk=c_j4KN@o$V*Wmg}X=%NWT1tL0 zC-5GF%>ADu_@@EAQEthpl>625@|$-w6(W95ED|?}+8-}l991VeT+I4kB+Aw#=BJg3 zvB(Oq{sQp?!6)K|?HNYMd02D1TjKmq{3xZk=WhsYC}!b zvlX4C8aIi&FR;I3+mFC=D8XG(X(%%yy2Pg;;8*^1RZ#WqHB!VB!?d>My39y_58Dn` z&&L5$KpILvZv2hUSE&?g?qi&nf_=sqq_`0#!#NDFL6a;%b=KPBW7&aURS(&oo4oO; z&o7_&n^*uf5ZHo0+!l}WB6#s*dK4EE{Z<06v(lfXn4vfBACu2Ep8Or&NyuEM_uhMz z$R^!EAJ1~$ow7uDlaOd@c-X}u{s;3c4K#1q8C$4hiC^D~@N2Hi4bIg%ke>D&idKeC zB*e1M6UVF7qzI6Al|*k*`|!Og|EJPCq-SEYor`pOqE+6CR81k!pAbr~fVoj2LzMVer}|>}M4|A2*Szo?Uoph~kt|>2$B{o}`(MuBB~;Biz!kH` z62wf2iP^)$$;@aRY8>(;rA8|9KP(uy29~=XE3s3yJXbIH;`JM8X5Obv+4M3g`;II{ zF-LMffaN@dc5)Oy+kKoNG%D^#ZWAwpI_aq5OjeX;h^E58c;IiXu41RTbvN;$<7 zebl#|Iz2%rMExEh*1i;=MKkG}FQ!uva^k!O57R~;WbYgRtuk1b>4~cwoji6-UD-c> z?PBx~$p*=SKr2j>own)Y#R>oaG8;ff7?>rHai6WOdE?9QS|9I}70$*i*v>=f??=!8 zd(Fttmop^8bX)tF9Z&DcTU!V>I-IN%g@{}@sx0vDV4Qrqf>eM`W+r10=nLCq7{{3>ERp(9LC!=mQRfo;aI_6KW9&({&M7_b7x?970Q+mr5Z^Hk> zkU*l`Ywqg;OTH4EZly^D_ivLTXJon>NoxR0&GKOB(g|{QJ|M=YR1EmDh zZHH=n{vt;&ucvsKbkZK}(2b7wo6SP3fGHb3Byu0DbWw|u;W!O6$*g=wH;XNWo7R?%n=3-nY@p?QfB`L0B)m zlI;f~tC0-#|AX34lV~%x2QA>QkXJSEu+laF?Y*=hascw&fmUBy_wF!Xtoz*QLL6r< zo?OGD^i=yYol+fwFQhg+MIv0jk!@GF=Q-?FkmY&DgN-Q83Y_krwbZLk@Z0z09{`P2 z7KbY)R!ZD-!`U7nr}cZqX6ewHn2H~?Al+0|d-EW!id&jvdV=^!&4zUSd^7JP^{md3 zbU*p;H1FGwhA04`^G92q+AZ#Xw3z)V`)j3o&^j<0CS<@qH`Cn2HE6+R^r~vdWbXCL zSb#I!n^Jp$O`25OICRQsCx2Y~xa#dD^W&*R6p3^YO*a7f(iP3`~^uum!O-+7XB9!QhAkA#JB z@~B#n-pPWG{I*jX>HhnF4TO086Kzfne|!JPq392CS2c6~;X?EEpxum*TD*WY+c$xw z(Hfqd@$`OST2JZTPD6rs%~md8`OXAJFAnapY|bHE-Dyu81OtN$x6pIqtXq3Wgd811@6-~e4v)1d9OD_aTCRo=re4C?{M+IDZobT1KZwM#)-Y?=R@iIid? zO-)&;yu_xz+|4Kx2T?SU{M}vR0Ordj;GBF1q3Jq2Gb1S$4 zn^{u~=~n*L-7T*Q+$gNpg^uvI(E~KC{FdGzSL%|ts&PfCLiNXm21tC}*V~^+E7}Fc zm`o=uU~iDQ3-uuaCp8q8d3#Yv3BjdV9Qv6oW#vmxpNT~D2h^}rJDsP=;MJ;9kK(iZ zyJvy4qQbXFjx~{=v4#b+SrAAQxIv@O`ll>9{$@)4h9Mz7Sb+UXU22B&0J){X6`5=p zFQ4gQf`EQ}{{|+TfU;2LF{YdX^{I5qM@^bMd6Ch?Gu3}#MIf@kf^Q>AwYhi zLZlRm6BrN?jPITM{CN@_=Nj=N(vZ5|1)DA@AY1KoS#-Z3LiJTAviXzC`)*l6dSHK; z9Yd$7Ct7nz#-f0>{E0n(!Upk87_hdIQ}EG7-ryaLy5-0zAMa;O_##lG$Rq&<$Yg>_ zDOw0zt!u5f(}faZZ}y@Q3Y%d=MD~7`q=KTo7P0GGNOCd>%j}{s_na z{`W5IcjYJNALg-L*)%`vJ3wfF+~h(8;K2kAa>1cqqGq6rKu0%ozoeu2LXk0b5ld{9 zZ@aI&*tt$24XujopV)*+7Ni+u!4quhkJIegLDU@s_!33lNENp z+9^!0wAc_=2P@yz5z%F+7RnG>Z>BS@l|(Y2XK|x0TAto2(!_lk0`v#*0{ubxmCzJM z;~;dhQCNFYpQ7+HV^-5bJ}6)vceB00Z6ze(1peD{CL)x-jRZOk=vvpEz>~HxEQK+J z{{^>!n%^o)^f{n6R^IcE?g>&v4Pruh6QJ4%IMFxbt(3F$a<*c%W)%)KdPCef>D#`x z)6B$xdfo2HAFd>#|A!cyU}Qgk2y4xHc~@Kg)t~uYiymUOTC|3x;hT~f8c;;BW0IHy zLFg2E~&zTh%RQH@6pN_ds(6dfDGc%ChxjJ)c6t36QKEJ6p zoJ+k`Veo~COyr9C&|jkE=L5mI`MFZO-dw8Mi+w`_7n8>Cyd1G2;MZgZd0Q%#A40h@ zygCpqR%F=2xaqyAFoM(1I=GXcLqG@WjkRo7-$-Z`7BHg4w8gm@LHo99dx00B%$Wbb z$^fkz2zs0b$6w_83Ufp%dCo2KkZYK*n=)*ns1Tx6e{5MZr^QSAP`e4%L_nFr&aIeX z_LMc=!nmqL&AS*mrzFjhl(MZ@V7)d;D$Amo@m?yY-l`)|`ft zJ8$hUsiOvo0v|A~VA@sy>|cl?;A!+~SRu4aaP0z0$w=UQ*tk@wY~gn^(=5SjEv#$&=z6?SZa0U z4JK{U%>VTJm%s3Hw1Vz~fn&xTOYL6bM1yYIaYn&khn=O@KI7D=HZ*XCRq}Oz|A8U) zqE=W+;3>M5`1^6Hl(TexmI^mj=iJfPF8_kw`CDd(J0@+d(r82voq8q?$8Ue$8@N({ z-wg`7ZWK6+?Jx(gVAMQ<6=;pu!Yle7AxpqPbI#8U18-ZEL+RGIkiAAhh#*X)!GTa25`s-@S$L8r>o?=qMT% ztVkv2Z1A1BQAF*i>8e^x-2js;4Z2d6(B%N7&yL#fEi8YDRNn`SNC%v8p<6UN#$Qn^ zT!Hxi`|fLCE(AK7s6DMJy{}#cT%FxPfW$xn6>zwDjPLWFfa4bEMz<&TLfoFoJ_H9x zHm5F@-C!AHp-&*O{I_k}LYcylx~@~)$q`f-(EI)r&MU3%^60n=Rl@dMfXX?Aaqqv_~DQo$8I}|3KmTS=j9W=APE&^#eMKYmQe(Uq6ld zn@t62gFQm6Q@3K-F$~Sm4nDGM<267qf&oewoB~ju?8X*wNT4nlESw05{q6KPQ) zGd%*?o#G_->D6QgmWQS#Dn{uhPOL%+|* z{y4^>kEKY}=Ur>E+b0i@7F85G?mCVs5ZJfE;0DJIYp!%39JHTDhcbU*a)*^Y4)XBSh|JXc(4n)k) zD`P;l_#yW|rbkN5;??w_mxl0FI|uwekicmA0_ef4%rW(p>g1x}YIj+o^YH``INJ`vOwabwy2A{!!%4h`H&Na&&VZi+Yy?A+hdd{c(l;4mJtWYx zt&(;mygR^B)ZIFej2ms^h5Nth|7D3Kse|kv z1PDeAMpzW7Cn6e9ux3+07m^3T!szLpeU$0N)G-l;1il$wk?|P+fT>jEPXCc|&c?du1jp%pEK z(9;Q`Ij!khU{yclf<-G6zp*-7Ica&Bg+6MK#9VN_bRwl{n6~BsRjN%K5ePM93$x;9 z)mv~@=o+qMZzS-*zP{v$dUhGgxZXqxWc{Jv|I=5}^vzvG37TZS3r5SBJ zgsNesy6HFJ(@KGY(8Sq-R#a?!Oz63fPZQ^QC}@ocvj{lrD?1`yQUAr&Obf7Gm9oQm;F^M|g{r4%kpUw2YAIG5>*e;RM-GK|34}LA9mD;-~D1`!oR|&IVXXqCKBfOh=8YgdRl^aO|5VN* zeyH7`Jp>9qyBHB>GJTCN3WWP0FgsD}F|Y}jG_WI=D~^EGc}3r3e|99dFfj(q{yXH+ z!?L};^L|2HXX^p$_APs=Cf@Xt6c56?Gy&MLUquolNIvVo+$)ltun5#_xdmk6a&=tU=jslP)am0B zeKylcj|ih8GCf&YCYK@{_WU^iF7$o_Y9_4KR$0!R0~zDGiZ7R;8>~ptn|&rQSz;=s zUKM!y2-t3-&85HZqIENCj4JlZF3hCOUB1yxoWUVlgp_ln32&>g@RRm}AXX|g3o(=UA zRntX(;XYs^00Zwb!NPQFZ6GJl?&Z#oPdhyA)=;rYJUbqqVC;8Fqbox81RV}nxs^Je z@vg@P@lM%4ZMKlAd@q;2G_e?=cl4?FPF@(fJq=^3SD}+*092qq?hX}1+Yj@#KX)Bs zo?i68bH4;kHvC^7(`LQ>JUTC}J)JnVV%Qk}H_3lIVjpapChHQl0XtG!+-be2q+sl1 zUz`yi@MyKLadlH5t=OBIC#SO^&g*_!<^wj`7n%-}*&CLt_>@J@IkfW&EMS`h$md^B z7))-p-+GK6LG-BOG%_+gQc0wcYA2*lefLq1@VCMITXr*cnBaofx}jaw9XRU&nZB=n z8(R(-aMtJ3IISG67Ry&0I%HQ#5Bw$%#@G7>UDV7g8#K%-p|3%_1#?`$$lrxglWkSa zQJVj9v1v>lY0A1L{vP8fdE3V00z%Ll--!36XQ($Io>_1h zq+|B3#XXMXxOxK1Z+|5>$cf0Zo;|}sFRCP)Lja@~`ldJeANClI)h@PQ$JITr&}ojH z+Qf^VTBFA_BR9AKv+{6eCQzm$IB|RtRnD>vg4{q>M z6mT*a3fhXc`*kOf(%_2ZUVQX;RAV7esaYKv zW5KDE(%W8)%jS3uEF)U~e5=|7y9%dJ^K9G!+>d(n{`)#@Sv2UtjlHy!AX}0i)%}iE*>d#M0R8ABQOe;vK5c2 z+zzQSpPInbzvQD7-0d!tzW)?o`zz;&PMH1P@+wo7@xZ5zV*`UUeEp)&f%T1kWo>I4 zyrhc)&Y~|Lw7U&13#8gohfh zAY!N_E0kX~V=n6=m=b0%q)2=4evHhL$7-jGt7N^|_8t>ZEH&7Hn4-(><=SxloJwLDK*WM!cxnx!DbBuTB z5yzy{t;dWV)*elUxpxNLQXy5ssWD-^J^Hn+Ki)&FT0wR89|Oo_2BVOtP)SWe!hMj; z&aW7XvRpFUhNd2*yTR`tRdNfN7u~&{7{Tlo$;!lsch2Y1TBe&-XCqzP+fM5yt3}tr zZzZ%_s>&qw>?QgVtq(d?YqcdP<0A)XO>d+q?&)aSYCU|X5d3yT&+IZ3T#{c%8J3iR z5JU^&D|`VL_UMv|%fJ#|WG*(fmS$6-oy}=RyN>6k6|W*kxJM2TCL~QX7CMqL?Zw307UE>ij%LLu+J6|;iSF*M1rw{;nq^pbdPpTeA1<}#nevo|dio-u z?08!zr>tXF7UA0b5X({vxjSTM7jht=5hN_!q6E_jEB;Q6#NN88SU@ z-4az78+YVN3KkX{Ppsrf=srHMxWr**VdF(sE6MEPhN4ud=>AIC2h`(p; zt<`e32kRqFPrF?Y-`ruqj^7CvT@Wau3*TVT-(I(BOQe63d~>rnik&CGUdhm?@xgCa z^JC`BC}BRCV^BxZn9HzO-+VB{LGw`PEVcC6kG%kXeF>HX9kIffKx5YPj_jBaY&Bou zvd1o$gC5=CLk>qwyB6)rMqWxv%zs$vin#XjUg$HWM-`h?!vt@8pA9JxRo^c=%M&NJ zBmW%bEdj-}8uXEY2SXJ{qDLPh>!xOI2e!-&MB5DYy|x`Gw>>gZSWO3Zr@-{fNp>8x zpQP%qN{{KwzX9GOXL?~Hzy2~2v*>&}Qg1#Ol1(i=ts34gC7PyLB_cK4{*1=*cH0dr z$yMK~h0m9?`L|1s8{p?sE|A)RDaY$;QoBpD{DAYohHjD{q#RE^BD?&Pt&DLh#=p8w~aS^o&Y#1Mm!_ zKtumyx$Cf{gxA^O_<_+Ic?64EYnVrHyi}7QL!(L8R#K-G^kFcB;#pTbjf4{Y8Da4b zn}_G60lR#8mlXV*W==(|q&{*nK>PKj;OcAqwZNDfR%De2>CBd7WgNJULGLu0+Hb}a zydpEc+UJXTx?>j+o{zfZHV-sGDs13}*$ALRf*w|%w%=OmW{_9Cd>#p}PulW0F6787 zRjl@8y+&=eKt&hWG1*Bs76rkd6ueZn8q!!>9(S^d^(l8U{((d4yGF=P;Ry<_-e~|c zwWiX$p=y#g&s)A5s=U+j+n85}+}FTw#fabW74mpV4IQ0D|AF)~50jSfWPa)8stm?e zss7YPW|~_id&_sHk&C3C+f^GSMf=x;{Wd4NzQ7YIbtqOG3g+@gsNfV$9Z!j{ykJ}U zpJ53jn3<5QeEJA4$|(G?-K0{o^|*PsPO9|nTR3{}U33*?m{9w`j(Z!tG|M)%WhJ!L6&d#_%q7ZU&JJ*5hvJzYp1Nvwoh`cdo3somUlLbuQ;*I}7 zXDK;2$SQsz~JS0JpX#6xgQ3 z;{8;K1_-a6C>Ep8VB(w0-$KfI2XhcP8%)v#UeE@<{KpLWBZ6I~c`4vspN$R1RYtMf z=g!1<-+-$8DFAweptEblD6ky-;GkBSI%lB}r`y)^mJ8~!vHa1I+ftwxbyn$(_Khje zGgEHLgc}asu=34xf%%6c3LCu2ff{_$vLd@!n=SvL|u zoSM@*%&;7o*Iz;KGt2U^VEECa9EL4?H~iWv&@vn3A`7be%tWEF{tH1SYF%-8`Izk2 zgssx|AaC~9rG9|aFOTz0ck0C`&_@?$gV~iwGh{wk{FEX8^DR1ObJ6X^H}BvQ4MZh{ z2T7&wzTA!EhzRPRR7-cl(GOT8K+6?~xAUj!Bf2gkGX&^d+<{eDwMBw*j?EB#zxyms zT?NfnW|aSQFrFOB9i&UT{Od9w>lU%WBcPSWy_% z&^4Lri>q~dP#0Yklk;ksx6Z|Xd8YTMEv|hSHU}OP^9?Od9_&xKe>Wa8CYVpmhJlbt z*VfvGn81pk7^qEtEAHY;Hv5)|pWXN%*L?13 zC@_xIew>1rGoE@Mt-mTLwmOQ-2Qo^zu1sQx)LzkUv?AkQh1njlcK+I#6xK$|$h5a9x@8;On z{2yI!9T(-^bq_0|fT)y$qI60l0tyTr(%mK9Ak8p@2ofUQ-Q6*CN_P%1l+w&FAOkZ< zzvJ`#&biNdAMgL>5BSVA-`IQYwbtGj?Oxp5m*a^-$IgU*j1sr7@}#K+e(cA9Yfaw) zfq{4o(hoV%|1Y{x-wNQs((HZ{-@@hEvxMSN~F%%IEQUuD8SfXPRHEWh6JA`vgu`$?0Dzl!AK zFuO^^x^oM=fh(X#q)z$K_4Ct|SFgztVJO(ZPJvfD`lCsCYOJ7P^fHha)Gsi~(fYYH zyzn2p@b^<=c3$-TdfyU#11n7D+q6VkWnWRAADbK%SzDSTH4| z&!BFfO$(BcayLlZ=wKH`^yaG-&+VcvraqXXB2n2+E%C;c?iHzSaMJ^c@b^}&GyHi%| zW=Bx^)uaV??RcPvH=#nfu4kun$XBZ8BW=_2}%>70faa`Gh`a>CZ`V9Bn0CO_*1hpS}m`Qs30^AS?3* z*R+|xGd+BdSXC&LkAbft<1$oLbTw^TK+sYy%{(BERFrfzVdLg)rl!5E4*17r0 zS7vKDr1IUtS7bfx@$AC#yb4f^~?Nt0`Kv1veK)MUjpRq@^6doH} zJ-+n(bE!>q^q>{A;+;rL+@dQ!amxE|4ITo$e?MP9H=&M~OCR|AsE?BL5pa6wJ?BJe zpTQ8AqEFM)DK(rdsi0}VVtH;Tl%2BiL`QG5{y2ZE!g@HR)-J2Qu;Q>)5w{{L=}Y%W zS<(`sCB6!{Z|G&KE6(Xmw&dE`E&gF*>>B zfc%(>(DcQB_n%GwSMU%m(YR>xj(->A^hnkV^llV_j%wGY8X0cTW|SE+u>Vbd{%@2{ zOBKDYZ;h_2wH;&@Ru*u7rx_8~a#tM)ce9gWuxhf1n-KdL9>TDiz(C`bmL z3fm~^ukqs_(CXINx1gN?{+;vgoYi;TA2w$gUsc8-|Nf2EwB9fXPnl68=#6(scYh4zl(FIJ)|Zs4cTUQ8J<-5>jE?o z-lnq}PAJ2#WPe2V?fLd%IWg~MDVsn8gPOb3fB{hIfivm5{^6R3GkycX1aFb4HTkwD zOqP&u8zqLe)7!Tspj#?Zm0^{6fFQ+FVMpbatmL;EP&!wVQ-nhV~(U=W@c`9cYd#j3#cahC)Hm&`_J>R zkc+*3H1#4GdK{eLazr=3{pLEPpB ziE!YGxH}k@I>A1n70EAuz?RV@$wCxnk57Gl@~%{5`8j%x2#ro0PqaWb{-ZhEWh1O@%h0zQv0hSL7JRLJE5H1 zXFPLQUTQ3{jWW>lq1nPee%n7kevA;cPpT_gK+*o!J61@eD3DKkaUv7uk;N z7m-<<<|=5TWsntpSrZE=l1Qn{x^H6drw7hQtG$Uq#Wvv+@(z0JrC zw@-ptDTqGyzC8W2nmL;Gs)nAS=%8;jtQ0q5f4ozo@1wETThoohH)7nQXJ7uMumn^{ zQ@aW=CEb`dR4nGadv|cQK!E`b)`D~7CgtR|Y|bY;+kAjFcX9?RZbt?XQsjAAg0!<)iUaQm%uou_s~0>n-d{uVOS>Rg42q53bU+!4DzS#!46PW#4V zm0Q+u)O)`@grQ*un7+zQhhb^E3K_aG-p_cjKMWN0$9r++$rHOk=K117<$S(jSCX(% z;vSBNtdDY?Rk&>lOR&+jH+2p0HD|e0Wnvd^xnzajE_q@!aiiWRZNQs|esE+m@1Jh_ zN6{#3Qx0)0Gf2JA^UtPjUtvVti+_QW4>T%z%;7QGnbNL>$-o}HfkM+)e)_`<1N{+v z4?UetuGLgonZ5S!qP?9<%Tal&%z;h(K4e%pPfG$^enkWMT zYmrWWdp1?(J$p;$)qZS{I1Oqn!?bc9Qu^Yu#wM%`t{zedKpy0}46azg}5;WyRab*N0WWY{Fd-;WTB4z^gdW(G@8 zzX|X?OpXmAXA{2o!K%j2M-UB>FJmKjK^^X}Pn2&oeud-#<_pQ~dE|q)rog;8==IXR z3ip6A^seoilj4N%9~bTxb{j6{pxHIdBh@Z&4!8Qg^qxB2stMA*^Y$L5UV1NyoBk`~ zc&;rG>Y7fvn4D0$8^&(@3&bB4`Z@V_zz<^Pv4pJ>V*!X!N$L#B!XNbs{VRfUgXVH1 zbMVa8&<4zA_Ue4S&6bQ(HzFM;`?;H3xP@J=iE(wqhg`Wro)Imf&*4&2nFe$lCe1&s z=IQ-ARwV4e*2C;KN6`t88M-pL9RHnQ1_8bV!wQ;XA5t1S*2m~SFJZ8>sPg@yJEn6v zG)1(y_Bt4>#cw}v@&){FSgR^{1MDQ(0?LTnVeaqnjOp-rVVKt<|8W67-norPfXEK&>9Nq6 zyMLE@jQOj8Y-|?RC$b`Y%EhGh_R|vY^N1&N9S5>=QTbJRyX`6Bkb>sn9|c>0X=VM0 zdVLx{7Xr?Rq(N2u7a8IVOTy^I)XFw{W*R#?wu=hv7NT0%t5-~}ug=r}i3k_OV(WPQ zVZB~CCR+U254eAHr7?U;~DtOoDW~Kr1Gh?6zS!s4vEp6y0n z&zC+fVe}X{WXvpgH4Lv0L5+l5$qi%SjTkSVOv@5_|=MMriZSK^gy6yaw9!MTvxGHd{-fR zX)WO#xH~J-qBX;4#5=OOu>yHCAlUfL|Ne~FiwsS1t^Y=#%%x$l`N37kOKl#cRk#cC z4%rK)43Dwi--leCsEZC>2vZ`z6Yzs5l)0*?@p1xZYub2p7^Te7q?Hcj3?w-@0l5Z!{XaO6C;=7#zO9=Vq+cY4qO@ z40?wvtPq+jLwd6`lVnp6SlY@pwWVFM1~4tBsv}LD6_BptU>=jd(th}6;o6Yz^-l8#wXJXoc{V_qr&4uD}`{WfB6<5z`wRyV~{>uUS zg~@Y22V0InF{Bt&lc$gk`;k#yE>z2CDZS`zAK3K~*UK95Wu7A5=FTfKoj!8kpO_a8 zFFvKzEeK30utUTZ*!xd;4OC8&oDvI=Op?&Mk-#vqC>GxeR6}vz1pCeaex7!@>3n>h zs%oT5RC`YIj{J`}Hb9KzC!Hk#)gA7#&riT}wwEm!h@{~%#U`wc)Cqb22%C>K0N@0D zDa+dmPa!Z&9FFysoM`5S*lK=Bim%pWN2)+m-QNI3@$|s+TI`M#h1C$p+*j=!F6@F~ z2I?4$&rTHfwZs8ztz7xz*RA#uD8lToto|R~>URZ`&56AnR1S?J`hA};M4C#AQ3toG znWXp;DkFJHe$bYg26NT)V?-evvz%j29ahn$X>YhhceJ(m?2+4~#15>_ao1E-lVnk5vyV2E=0pqT`qRFW%AlSwE+Qmeo>qPfb23pv8R< zB?^())kuUl@{MvcMs|fm`oBet|v7N zMSGWiuD}E5$2akI2+p}L1`ZY@cu!*I=#t0oWP8tDd=L*a02gb9>7xYlPr;L#ct#RG zP#I7B35A$R{)F_t3Ir^0?3~K?*%her`;I^xkzUeTSwq8qVi|X)K!AY> zF_m)L=9o7-AU>Z|BepUof6}~x<=EDWzjwpa$mp|zzGs)F8v4+)S*rVlxAciMa?THv zslUj~{XU!L4Y4+q>TYryGW)Q2C%}dlHz^#ADj(;bG>>KQL z4)M9+TACui!`yveI7f=tP{XyhLYN`Rq_`IUp}HZNoHVXIEtX?$qKdVz(d*W*Pp^?* zP#r!b-9<0yWWJXe3%}(2wUS2P;#(j5ZR|^{{J{FMH6CX2=9Fg2&2X*S4zK6olrfZ3 zy(+^#(rt0ZwBJa=rlXO&-w1x1wkF?ps1+P@IBG?AU$?I4IFSofEUrbB*?~%S(TmV; z3|582bjjtAc9sbEm9(C3M*A{dJMuTxruBqeeIO^!}Hs=)EDXj;R((?&MsA zm2e>y8{KA8v?~>I`8hW~YphfXPd}jvTsyranptws zDCy zvhk#n1a0L|0lCb_+F2v@VLIR(@3x7~Xm>^IUHoVX+N%fA09_T5T}Fz?NF@y>rwpY; z*wMIBH?vpz>x@@bGe`R=ZBK=T^3gLwBB+8#3@^&q3|z@#v}^4y$G)7Nl}|uS?s7T} zVA!Fc@*Ws2>_2qY-{iyQllInQ;RUT_v+g3 zlzkFrbU|&u5G~V21j=Zw2xXw1@XQDeJ(D&knfK9UiFEwD=R;I=`rUJE|D`!SJW`VCpf&NPOv3F|t`NB2$;&}e&P`}N!FMw%4=;bOm zyg!{W&%(Z6H~#GfJ60pMmPF>Asn^R@^Xopca`+Z_CEBPT#i`ZKK7;*d=5nCE?^B8-B7LqsS&CvS z-U$(rRITx59PU#Y#YEdyW-gzhZ(L?eiaYu=6Po5QF=!TJ*N7zfZ$1VTq9f5u?%%fYD`rE76jUCdlnn3B+@ndsuoGL=CzqHm`Lv{z)AsEDbFV7GVRQZY;P zo-=8$zl)-XCT*VkXEE@GC$XNJ4(tEs)Br_AQIK-l)oASK^-OLaZXetE1+$;J1bIw< zLDryN;x8hfuc1SjDm#aSc`VTo{+e|-ZMk~ww%zRt`N5JQ_N?>I>JfRKzMFhuj!X21 zd(SIO&DwF}w{1Juq%D7TZU7m~BM`V;%LiOOW;+^WKcE}Z4z19fhaMmh#adZX0)YFR zlJUw5yP~nE15J0r>vH?!JigI@`+-7^D(HutZhr{1dBv}`Ij}YaLtVVODF_bh;3JP) za%(D>WwIU5G%)*K_KAb|nZ@a`c({RU5&kO=`P%T%`CxbeONZcPFGRdOtjU_n|}Xgja_|1 z1KKInz%0~&f8|Ux@DJ5-^sj^D|EXaV`ZmAGRDy||)7i?|pAZEyMt7>juY*%XCy0A@ z57Ex6Z>;wtkVzfwEQQ397-7AZXrOHzz- ztc#e^04{5p!b?lW?`#b^V+;^j<*?Sr?QJjuss7B{OmR3Dr$?CCk64 z%s0j}tlt%@bjU0AuLltnG=H(?I|9_K;y>5GGkgu^@Hu&9lLv^}9}IH6cH1;vO-^eA)^OQ<65Hc_Mnp$pp`YL3+I#)xQhdm!Z{gZIV&$m3!`NcS^eoLA z)y%%+XKAkLWoe3QfURTfs&d*?h}!eBN)w)}<$bP@n{5$sd}s{|7*nKpJkv4=F2B7b zZm%tgF+vb$X;I;1TY+Ot{s1-u#>A;xdd^m&NDqnTvbu&zQW)v#nHginhoION9|b?# zF}yTjL(!Z6b}21W8P^2aLV1cwHDMMzS4$S#I1oW&@d@m z{ti!9|GB~1;bgG)t9CxYP{h8Cpo5k2mWlSsY*D}RxFEpMNl}ep?$L^sX z={5OYie#x@&bDg;cwjryfNrsMQG%qsan5>MN)w}FIz`h!{-ttJD z(4ekhMhudb@TBUOc#lu4vi3x2xg@BV)_9|u)qYp*u z861o5^?ySS!aP2R9zjP6*Rq-P=qF4LQOl$sP z=njV@`_8h(<5E&TU7C*{YoOh9cB)}!G7P)AKPe=l(Jb7?o#JT8)zn)###^Y?bn5l1 zYaP9ePxyZT)KLaeScudBI`+$|K&skX)KD#QpK9(fRQWYEujL{3yf& zBkj0uN3oz#c{wY0erC{uuMP zxUgjWO*)A`hqC!R;f8lnt8-yXhjx8m&5Tl^DlddjA&_wmW-^ZF_7le ztN)YuyXI&Kl>|DMO4eiCSCp~yvR6@n#Y-CDt5?g8FOB9Yb;EN3-#*W_E;x!W^o6G{ z26o9RTYIf{8bt3ME;5Ap`BXhvT(stYdM~voj%+~bQ&Pawor)Tp(rMQqv zYGPM(=mAUi$zpX&$hse%E~-i0Xaz_AbNQJoZ;d9Qvp!RRXOZ0^)29x4S9l0D0

    a zDl{iWSi$^xen*!Lj@R>%Ope!#sfrXz*nAE;&+6Z-98hS2+^M8PmENq$e=DgDQm4TS zk;olz*#Ozbq52DR)r@{}x>cFm*#w3ru@{TC< z>t~CBc9OZ=hdi$^J27Hrr-iY}7jR2>5NMHAv6z(UVd_(z0&Ay=k~_{@U{r85@21e; zoJ2xijAq0P=`6Jyu!hYUQJ@91jwlv;|K}Sth)pZI6Sc2#@$fw)(+z+Tn2QRn3L}xDH|Ai=G{Ees8tE5MZYQ$w*y|S9&WWE%tuO2 zjqT~*{>rpjya1?2T&Y3_vuv)}$rEL?m>AE_~#z-)afPEuN!k{o6nES@|FZeQRRPp^00+~?Yi{g z#!|mI8~A$1O;qS#mYuF(hL6eaJsuPo`GxRy63sVJi|D7nX}3X4ixaB}ob@-eClH^+ z8VlrLDwA7IRi)Gi^9L9OQifJM?6)jWdB1^U`l>+$6N8;)$9qBSew#@6ZGFH@)AdtH zd_DDH`LdsOuUxKz#QuH@#fuiW@xfJ3zQhICTV{MKQRbS3DADSu&%)oa`)zpqssPQK$U!0qGOj@O+Y z?ABt|BA;^Fo@~zos(ELblpfOWzZS@XL-Y2Tadpe=8W#%-ahkK9Qs&kditBrvG^=W2 zh0c9AbJ{5n!`Iu8F%bzZf*S2Bieyme(%7NhDobq64PLTTm52ETQm{fO<1)mN(RI-p z__}oyYlOLi7WQFfBCgoYa}o6;cW2c+8e9D&*P(aZ216XnoH* zdoQUT!WvwV7$!XcFCE~o+Hw5T47q=#uBMth=>3j<-DM~IAX>ZavY^&~ZPojvYaT;% z_o`0}(M1W<>~I*^dM75Qr$>Vhyla*{59KW`dcEcw9<+=dhsz@~LMY<~vb?pz0_Xj- z!GtFZ>@r(hoUQssA1*2km$En?Z~lHygm?JoifhC9>`nh=SF-hl@XN(oNRjxWPzkdt&#;gxwB zdKazY;dMo)`-y=#mYcj)-UY-B9{fYvlMS($W`h47GT{?4Oj~mvU-sUgqV7Fn%NNWKZG9CzjHB8rUb<&)^o?rzu zz0kT)FEVlvPzDL&yJCdk%4rt^61l6NGyv#t3yG-Rig zsw&347!5w~4!M{^DYFf7@b>e2%imHBpb>0k%37yRR#c zZ!r??YzI2|We>}4jRy0!&--npKKNi`OM2wXjW-}pyh6GM^V_zGi?S>8h;m1l+jHtG z8AfYYjc#kc$m%DCH%EnB*J-zim* zKR8&gKZSH&w<-@eA1A%Q*U^<7N$Ny9096(*B@TT+@3}4dxHf09uBJtoE~AM9{9sv~ z*^=D|kvGtu{bQyHosn>3ue8f9!Bs{*t{3#_onE@v=T0IvKoj?2&n=ky*`|oE68K0J zc-rY#w4Z|Hf0mMCfAKA;gVFeAPU7KFX}gcZ#s2aqu+Z@`9)nAVVaHB|Q+?YhbjBS@ zX-;LJ<#3Z0bz_#+)aV$8<}2^3@CNRO6|rA{b<+M#@jQscoW|~&0(0E2SOMSZKIXmW z<@-(h?2_CDzz=52m@1^mN~~IxM9P9HCVr<*&(xugIE>Mst4bI-9scNtqz*9;x9oxP zJ2^aq4G`b;XidRgRCh3O9gLyGQZo{~{V?K<5^=bciT@!7R(8ZdvA?=&!|( zurXbS#E0KNYD##wFb#706v0shrw&7l`vD)(xmpBzgHVU!#1%~=3x<2}ZHf}(8Y%Y* zyp_~W8sw8I5P!X%^I@PEFIQ;sNIh_YA|4Pg=#8xqb*zRjlNZP5-Ys+0o zTI6wG`DXZ~GjUMo?oHbw@v@zm-%%;?=GZLu9e0A@76W!AFt!=&)yz^%{4w@3fr>16 zu0CgB%cb8vews4u$=JI@?NEnC?A{SOUmFpRvntE21*aDc7?ROraB;a~X?y1d3{)VD zKRe-AA!f|);%R4pgI;N?DKx+oH>n%y-TmBSJ`Se$sz6$H;gwfcZ5Teanz*;~gs!oX zanHtB8GHI=GVF|%_TV(n`+C1wuQTEsy>La`y;+A@yP$Fpa(jvQS9ncfI*3G@Ls-_$ z^gKhx(C{hCkM(SEX&}edhR8*+b&YLN^aR;v9@T{)@U*o<+2~T-|I~6Nvhxz&*(jNx zF?`rCW_;Bjh-~Kct583hJ1WWom;i;YJVj7u#HhWunxNvVx|^LU$#6Nr&+xiqRDfwZuM+0at{U?$tfTUQrR;2N?zwghSM5AEr}EiYA{esj{HbYd zUk^CNE~V=7Zl6jF0p- zLf6e!4^Q7%`$AltpxRc|PH{Oku9nyVO{{>*$;e#P;*>9?MjZ7ZXG`0_df(LfSb3y{AIp2X(|a(h3J5Pf#L-QoGz}jvoyv%Y6#Z1~j9S?j!xt1V zBwP718i|e^`k7m430o zntMoO*{ve4Ss&f-y3u2zVoApsj+gL1F_I{>5A%k?q&2$6HIMKY3!M|hr-0MTmuKTW zQ7;eEc|lZ{<9W5e8+!H+1TR)^v5#IZtoh{1_O_S}cmVs_ITMzmR%l5`He8Uiy1?}k ztJI)k|BEo?k<1nKT2F6JnY50dTUGp92+uTS<_qB}p;Ty(p=XDv(aEQ!t+}))OlNTJ z3lZmG4=Eq+WMNwqe2oOD`XJ)#;iSrQg01OE?S<;c~fG3;|*XTPXmShOhCr#*Q z>L{lRvELwnZlKJy&a!?DObkw3rg`$OxAEPB;ACB_Yj2gU{Yokd;oLy!)x>6g?fOR9ohUsNPK`vIeT3zGLRBpjsvd?){Fy3l{YjLlDa{AC~Jq-A8lpwvfwh{S6fU zBK&@9)k>mcljL$i`B0Vy6)Hm*K`F3N73YbMp>Bl^N!Bxd>|0`TI_^ZZk)IlisSyi< zHu!u|<d&~h8mpMaD$(q1#khM`Tcw=4md^u5hJvNjN3IH- z+rl5#Eu2weXrFSB8HDkdZv*VxaswN5+=^fSx>&w*&GfT;N`)ZP?BgAoGfPyM5a z4OZtwo%a*dBsrmRJ4GTF1=4M+el^5b48**L=!iQ2N1^jx9+!j50_m4#CHa>p%9kJF zm{i&C`-bUVBdbI>CMZQL3fPlIkR=m7Vt(5Ik?BErJ$F@aYK8cKUxO~bv&h-r1g~T# zH2BcFZ)6kU1k5&xB*leGg-hI=`1>f-|I`Q>&{#Ue;Mlj`zuwbT%=GUh=`GN+y_9Ph2B<^uAsk`u`|O)!}29@6;djPv6wL2HhjMFMZ<7G&o%Y4={ zBj@E++Sa5pB7veRIeJ&}kn)`(N=)ZKLyWYemQ7ojkBRL1vUJqfWC(d*UvKrz9k*a) zXm(#SX1K;w9s9*IwiWRrBE`@Dk?{Y*(C>S3o5D1YhN1d0Hd9fLr2C~FZm9W40VhX< z;IyGntF1W6;u_4|rWnm3w`zAXF7ioU^z^XWHj><(lVZq#9cW2HN|sdxR$&3tLwxFC z9l;czyqNZrqBTL=b)*dT0;ub7m-9fftz|01nB$wPQiET0cCT)v02%Zq8n|8=N#0*` z(o|BNC}pOMvIOsP9wn|wh@&0xwzrA}n5Yqzr5kw78@Fwv@75J4=UTaN&6Zz>uVDVdK2np_2o# zRVWlx9=6ggse!ssT#@eJo@JuJHwZKt4Z3)Dj$S?1T9@m_2oj;rD$9z z*iS`7wGEEbp>2s)oX^`FUBq`mU-tEP%=9O9o+~oEG6ZGrd8SeMuI~kZ(3r34%XR`C zKTlVv>5Uip>FeYmfR@@~pJN^6WE_={gg|bHAOj=6$s-g|0FiSEMbix@p9`9o zk2A7B{yJg8o(JjmrT%9l2eDNCrv!XqI{Ju~twK5^rs1;Xr5@-l;xQ(AH?I0fuAT&^ z0x)#Krj=I9QmaPN|>JIQJC^ z)uUS8Z;-VtWLYEu+D%zVGSS5@Q)wINlMt^}2{kwp@Y#K6^l(e7sT-`twdu9D5o#U@N+HN|WCR;}ezI6^1E2|gg zE)JiGf4s|nJzw5GoW(gje(@sCTiP3VSjOxWwu0T-G|JkU6gtPcCM}>P%YPadhj9G# zx~Tr@3qI|8cu0sCyPl!n!qsZdQbD`Wf!}+p?~Fzm3%7MuuU@EmqMcp4?+Mu{4oCO{ zJ}k=Ae7VA7paW}3jc%#~$E*qoo!kAw5bHTd88erATNUje7S96x43l)7wO^k%bIfgc zyPIxq`%YI;Pc(jqMV7BW!uJ!GWvd)VqnNQkK#_dMHDj%)kEvV~C zAb|)+6$s%e-|0t`DFT#EOq_2MTyx*S2eilaCk%`=w(X&iF;Um#xhfxh_O*<^W&yM_azU`n$G$%~U0q3(!WZyIGx=)( zn*lakHuB#IowJDB{5JI0Y)+Qer(-o(p2(O~j0^7ydtoN-Bx$fPG_H8Z?Jxhh7V~L; zS!1oS`m;rQXw4s)1y_9~$Y0c8Jo$UZ%0H38(-#3snYwUOEup$cl?2Ssu&!t)2j;h~ zUWK{*K6>^BSZ0m%QkmOD&`vz5cNu;R;dA z>~#=5#^2;lO%vmgcSd7u(}8K<4hGYs`{)}2v$BlF?;!u~5BN`xm^8+U$@eRoX7iik z{q`Bt|xIqqT&|g_X>XBBri1lx6UWE^{g6AR7(7zI7$`FlynY~n7S|a4TJa|UZd`9H`tHddO<`+?>_0iU(#I1JODiU#6gN4%h&!_v> zicL(JBR7ZTF6@h$=4*mQv%PT#rX}bXy$g6JOBk1{Ljqaa;ll*TwlC-kg^(M+G6#bm zb23MD=5iD)FY*!uqh9U9OF_|t_}nP<)3idPxVbk~5oNJ{q zhuFNfVxlP%F6=AU`vv0eT>S@=-PWLI4vBF4U33sC#tt@T^+3R zRy%3v#edss?#YXW>nhfV z&l@1x7a}7B$}z%poZq+J-+E|waW29hO1h@dX_1tE)x)}e5_8?454rdzxD3c%aDbRD z$Wygf3TL(xmCf5O9#?q`j}ja-Fd;4)(6$S<{?~1dUL#|;oKf%nBrTP;=Gyb#Qq&D- zhQmP*mmHqWHlW?+UwWa$irQ8vPd9+IMh$D%x0rx32B?g^$Uj|X1VP>szP2W6yk269 zHMf7uEQm4m6JkfMxWrYNy0^4MYg$d_4r0_VeW}zCf`dcrIKSFA9wVw}<8;b8HsO7- zm|0n}j>6te?1x9BIrg1P_Kn8ces;GCJJzj%y23;HV)qPC`w^wXr6UZ zwV6xnekSy#KBJfELi8%^0KmnXIHp*p2WP}Za%VK z(n4Tue_1?wKmXq#^j!4&w((o)BFF_FyPo@ff$Py5cU?@^;dYpH`tf2A5E<~U;-Qf?d zG2fzi6$ z)11H+S9w?c%T65q%l^{%?Inn$JVE6ngTc~?&N-DfE%Zthx-D2HD?uB&WHkYASQ#=i z1Y~Kipk4>+Xlu0AzftPEdExeRqum+MMwWV8k|m-10{G#eN87xS7@$WY?cH?(>HpsH z%$>VT&Jg+R?kOlCq(mP!d3575PAm=P6R2fh-^qFMpGlk?e=gK)eL z6Ru^F;p*~r2d6#W9kr>{=J7%0XOE1&BvFmeDms5d)^8sc3736;M86|jC2v*Wo!EmPbH}q z5I}t8whp5B zt!g4NLLsSZIjh?nYd9|drd^n-Ac5z=c*F_2A%Q`HrOeKFplaNbzLe1r{&<_J)c5ON zVrBBnq2dOARLa+UN5$EFSE7zB7hp+Y#zf;BpEk?*&PP>lobmTtKCQcwOZ@&mvi*_n zq@fE-YDa_@Pe@zg6Iz^M==1$0>8Yo0SV6uh#F=vOu}=HV%GEdpirE+rakBi~M_$rc z1(b)!*CacT9fG%CntzelQ5kh_D03{rH1!ujHgsi|ON7qGyo~uf+H_bvv&9anjMno` z43x&Sbe-z|XOn;A+2_IP*uXE`>>u}_>1ypK-uDq9azsvesaI^oAti#O#fBojd(>k|dba8$< zVY-W4h}IpFwW#HhuM#--EjNIzWd025v>HFL*J4=$RaF^3NX#Peqboo8ar)M`;VO0B z;l5&sK~=YO8WXvG2cMSZjM~d8!2J2JxqQ6Vj=fL0NY?h#h8;WoCk|=zo%PHvUhI0H z*O2w=vw@_1@2`#J@ARzBYJ4=8T28*M_No+}1Fu4?1@lyofg=#GO zACsm2M$_jC)Mm$#Q#-jAM(IRG7r0|aUVc(Q7Cqj13naTXr;8j(J<$TG5bdc( zh_zHJmUQPA9<79_NfG@G$h&a3Z<%!y|Gl2lyRriZtzy=aFpV=ma* zUVC7+82^o@VU5Y+{a*S7pfuh_IO>=TGxSmgIqPiW6y~^-f)RbYxQPCJ-c@Xz6eTY3b=N>?8^zf z_&)$WSRO-mYhH1?oFoA~s2nF$o@2M+o$NR?ds4jk9}E=0fg40Z^Sj4W&&>nT{rak$ zxS^Ec*xty+Dv;noW2$PK7HNCAu!qD#iwKA8Rtc{=R{?gVD(Xk{X$Wq`7OK`bU9$nh za^r1f@RqdINi&^=!9Ftq|8JWdT=Bh+&xTj=P$U@xCxUX0_AJ!kCW4_E)e{mNsSYZ2 z!{0Fbg7|5))Rgk$nX#Vz^`H<9o-t*7?Ig*;u1h^uybLv+Q$vd}Sh5W5o+q%4sEwJB#UjOMRdS^x8tGRBnKS4LjYg zAI-Gifp-EosiLfT`B@mZN>*)Vgoe z2=idB|IQ3k?p-&vwy4o(4#qmw^zQqir=t4f7}loN^U$o3Y3t<1LslT{HG=W;*KMjL zXgwb&P`ADlktsS5yFD z7gHY)jeajH+XEZ%AF0>G=l7GKHtA}N?!R3k?`^`2jxnDlu_+?tF@^g@&N`GQmZ~+c zAM-b7%2r4ue4lS(Zv{82xTCRKuNuN4cVx`{HLh&mfU=w_WDc3=WZe;sATD<~_~WFn z41&6>h_fv+M$z_k}8g^_Vhp(>ZKWB`I#RU&VWJ%er1CH zA6;J=7UjaNt%%anozfvK-Q6H9-Cfc#gn)EO3=JclLpKuA-8q!BbmtJ?c=mqZbKZT< z_n&8;=ek&V-}hR}0IJO0*axg_K|3A~1MeL`;~(zV$PXH-U~4uXuqViv4#ek5(`GAE;M!ev14BD&)(+^JC}7cQ+#qQiSnr9}@7OH2!M+=DjIe^>(Hdkg z^3*e}G-w2)wf?uKsC2>>DYoAOHgm2!PU^T|V{n+YYQT)=?!a|onHbnkPO6t0;(rvR zunqu*IZ`fl)ogCVlyBg&a*d>7F1j*Z7)fL}3b*=HZ84z*TLUc#8wsYb#QZbq*<*hA zjYz+Z;oTYO80)v-x%D(dKjNH&%J;BM#qzK|06#^x^iBeE{DI{E*xh4nj%ZJf&Q+Gk zr{jE^+UBC_z4%_kLeE$6GT*ElH76A+5bsQrgxVWsr!vvNI#%hIS$t7CzR^3+XY$*l zx$LX>ep97O5Rtt~Woad2oR81P+R`6RHYYx3E2}4c66ekBN_LW}OJsgNg=HpU)qY4}dHh zsVPH0-X-&Lc0xz(^C>h(75Kcp4yxr~2zP76H!1JFyAX6!YT8%tEP7eoLY9QIaGD3HkXc%hQ_f?q}pUYi} zk*+S;8SS)ZWgV|q1@Dw51W0=G2@|A502*~R$s8hnQ0nd^u_X)kG+wz39e1d^#*cL| z3+epy0crOtDs&$>v{sX;=lYh)T4+?BBz9P;`8ci5pN(PP_r_Cxsg|oj-T8SuAV%>mamtg67Omk&>_UDL$0)|E z;(;DJyLoxb`fEKM%@ohy@}DB{Ay5`vrqvt@92*g*&F0%xL7i} z?ZWwrhrMXU66*1z7f^;ht)@U6G&4Eq-^wi3(D6|3#7z^>J8T50b6O*GBVRZ;Z}mIw zqTvHPnNr=sNUaJo3*v}>FsY&9zmlK5WR;(eU+efSdsAi?IDv-0edDpFowaIVtu=;+#MOETY@vu+!A%N(@d#SC!2=e zS+N!~lFrhDNu`YQA4jHDEG*WCnH=Pb40#x20-OW%e1*hSc(u{*6hiJTp)GfS+3U>) zg6VCNiH$h1uqoS%u%B1=^7F4g{F0`v4Q{jGGBGmGXrw4L>@Xzu-gT9W0+sGVtkFWx z-}BnPv!gMnW%sPWko}x1)xuvcoOr=(e$=C#Qq;T9{Z`q}Wbs$IoyH@?f4po>=8|s% zH@f_SZ#_u{WrhmY#a#(IGalTFunsKa1L;16Xb+Zc_Xd1ckeO^%=gZ))Xww7lXaVQd z&YISU`BI|Es^nddZs@4giR~s#2X%YNUh|Ta;sr0|^lM!@0jF!Fiq?1?f=c`o=LAw) z0tACVfXMlJGgkfwFN&(KPtj(D!^nG=S0;V#Z3$ZJi)v-<-bD|C}bPr@dXH?Bn(k^D=xDt1aA zH% z*@|hOaQNs>J>}%advy>Gpx-zjm2DjU2NM^BW8Z$9A=7=Gd~N36s8CQ3NFT%2u(oY1 znW1Ey-L6NL2u8Lz+Sc2y50}>)j(dYMWE)p=J;4}Vzh8&sHcN7i6~U{5VG64dR%`>x zitDmj2JUYord@UpL!^^o+KKltWKGZ7V&+8nbJ;wz;WM1yeXWr_|HF0Tk7}C$4UY=L zEYebj#|kG)OP}Ivb;AD!b2@k4eTGd{4VIwmyW!XRyJ}_&XQkjjuY)s+ChOqgh?7Jw zGGfpmh$DY(p1yMNx6<(NyYaA_6ugun-rIFa@RtVIkIJ3tLS;zAns8@jY<+b6*_h9S zhSbZj>S7s8?C6|u#P+R~?DCZc3sBUwg66giw{bmmk*o#cL>lqB|Ie)^r_>m7tG8d; zX*6vbXtNnB32YN8lY?KaUf~pZ+qJy0PwUt~Nl6bWZ~tn!KP;-Syrmj48B#m|ZgM4W zhs{wt;Dh-Kj5Oi28^7^oJ7EdoRB#1o*o|9hJw&P0^}jEgaAJ3|Bi-JXR|g#S*LN~j z(oQIl$}Z#yH3~%V)XB0pi+uSuE9RAhR?NsAtxgbn+E~BZ@{KDXN8#PUG$9`aKsU;L z`G(5pUJ1&jB@yxlP%mo9CfrCA9eQ`G0seLF+-9~#rIC|oi`DHN>3O2`JZ@}j6yx%bX-}r0=_@qG z+r1@a_XB5Pp)+wQAy1zWm%>N=8~L?icAlgz2s-o4$uWu0fkfvcv9>{L2;-gno#)3Zg8#RGou!QAF0M+**1V}1Yg z=4bUe|0mK34Yof^(M6H!qF96+KL|71;5s(UO_mS|k z*;6TFaX2?KAw-Y*T@DB7-+lSNdE8D!Ohl#=1;d<;jjSmqmCAla^q-cq6Gj07QBDAG zBl%raEe*|v&6BtRV09mdAIZw?$??Z#^K1s+wh`hf?A!^haRTW!(0a=dpnVeMlh~q% zB`a>4_&)e`>WVQj@hi*I;Ogpr&P|>*z#5NP^bwC4)ca)|sN>8_@Or`PC&36e+$&22 z!k4DFkXdH;vYWQ+(vhJdA)^N#CJrj9%{XEn&a#$L=+`8_&AxM)o@VDw=@nwS*r*SQ zLQ~2v%tkwFt#`!>CHF2-S0yZxFCq$79<>%Dd{CD#;s!$6CABE8o2v3!G{tAQNAH)C z1Q^gT?d&CU9V4tQn^nhv;#q3$*P~_OEO%xhzsSy26RvF5ruK51F4JthUm;a*(s4tz ziW=RbbIqbvKb7P0N9ZAUR0(5+e3C#OLAnD~UJ#}vf$47=hg@gL({ehLsf|Ow7P95+ zhK~!0?kT7r#|Q$k-j(5$BXh+H-=xIT@X~BsQ_`eY;0bn&PF%b0=V?PiVwUP%b{}o$ znst#?B)m7?hHRxSFU{=i9}hbW*PNU9k8U5lD{?P!mNN%Rj z#Bve%(%EYO51|-glAS}>M{2I&U*{#Ch5RTd`K%OZmYxkKrkRK<^3qoAYvpRn2?(0) z)icvp3mejBAjWfaBkgIcE9Nfzk4bmI*m-aV?P9 zucMlfbD-SrcS{cAJPBNulUW5H;)L;^r#kBcMJ;7?C!B`k98rn%n(9RGx! zMA9&4T1xpX>TQ5>yhoaGjGUNtqYdKzgvPwN#_>KJrXJ#$gcB z<$H{zmZFaADs^Y*wbrq+XlX{H+yvzt!>Dq%P%l*x4L73g42y4JL|NaAvdxS7o;USd z1|(cQB~ubM{I?MY8K$c!~M$cWZsEed(vTbVn#^ z8gObf)Ra7!)_F~i{peaTb?#BMv%eb@$0#u0?s>~F0^Onq^J2`)cG0^Bt(MOm5N5+% z^)u1qlWUvR*4v8R)v{)|Y)0f(#!Xm(fV`Z$&DZR1#+`@`Sz{uZ*u>d(@90C9Io&|l z(+oA`aY`PJJA6X6^wy2sT#MK8a>XW3)Sply%Rdy(ZIk4E`h0_q-TGS2eC2*9M*R8h zNX$5HbII2-*6BV1Kj-?sRI=IR@dfNz+NoDBDXPv9lbDXiBi&kV+^Gq>rF-eA$~Aqm zIFdWcM)|KzLHni<&gbMf+ciZe2+wf&pmQVb?xZtc+_f{ma3M43!6zMo++Sbt{x;rn zV&L@K?PSW2V(&5U5$-gHo%51Y|N5Lyj1h59Cn_je%g`FYF`ijovUegQNq;uR7dSPQ z(|4SIne$&J)X4!S@CHC7uCoH{!NXdl_3cYSdSIz3v0n<|jN9%982$kd*!L1w( zTPq=hCnO;kQY90*cxgvA@vGU$9>vFF1THWCw&53W|LJL>jC_{MQFlX_vTBcA(^4;x z)T?|ejWPMvz3$`zl3izqTt2cnUv~TXJ!xdE;Nes1SPe8 z={v~vaTy*s%eB1K9tz0}q*2O>QJ6+GqIKgQXU;B`)KWjvPNUZBD23e~uMY5&25$2w zt5wEqxJ^-G=KF=0^WV)DmD5wdBM0acW$?2U#8@3hZ|-xbsxD6B?*2~4_70)? znM{RI@@XTHae;SELPO$eYxZ#enyc-w{FBJ15Krvm>f)ELwuP-n<8P#V z^n$CkP~ZK+jsW&?$NsQ@Xzi}7vgQxix+_*?H_XXM%@qY4Bn^<-1W2pu@H|}`vfm$? zv@LK4eCqHVRZV(l)J5oNl$gbv_p3Y(So(v>-&~_uv(5CJaC`rvL@5T=ESf+GN)9L3 z_gxdu1+N5(dX%0mItYXp_*7)vtbyVhk@QDqz^ep`ImMrN=n z9p+%!f~yj0CfhKTu|!y@XT-;1*)gpaH9d?n^DFKA(k*k%{is>clDA0KsgQHLHg5ZM zQKxNT!6?iAxK4YMWxgP*zMB(~@t`T%#t!0KkFIeefi>-~*Knf0kSRX0G@^f}ji}9F zEKoZdCLQJIvQHiccPy&m!IFgIK1++AXU4S9lG7|`-=@2+7PYy(kK)CF!_1>4z~TDp zGV}Rz@nQoO#7#N*$>O3K+wag!jVXK|C*+NSEelAGGK1I%?_z*jiKzqzr5e{5(geA%;69i*W%`lcgDN{R}Vg|dr9kWmhN zohaqE22#t6jXg7qCk?SA>Kx!edGJz!f#9s~cL}2Zky~E;`SHIJ zO{XXY`!B;&e<>UtQpsdT`~c!n!|Lgc<}s;bpgu34sIWxU1z0A4SfEeFl8Hxi-y-^2 zA%PjZ8?ttVaMjA@G_SV#nooT)NM%50Qks;EY4pe@6(7jKtY!QLV3{_N z42SZmLil%aA%*+-{T~g|QTy9JF!b+y;~4ev0F1?`vlK{Z;>G(yw`IkacEV?NC~HE6 z%6#{&g^3-mzO!94mlkazMpU#H>>4Qrq((ZpP)`dln7wmtgg%B8G-+DhfVZ2+2tZK*c;loTaO@x$_=}$hE%sy7o<>rC-Tg3F+3FW~ul^^JhTh)WKMPd-m!M zD8mSY{QfsxL4u84#p$NVwS0BJiKE_mEL*LTDm!xP8`9UhaP60~&|Ycx-$4k%QXV30?YYbZ?_zEoeu8yzz zG-y(se?3Y;1lJw*rpMiBh0*bDlenk$>C~tC!m1%W=W)eP<2(A|V*ka)c~MdC;*Am3 z^v$^3>rcj)DDKl_gkis`6NIdC2wQg=&;&{<&s^hI>TN32@<3bk>!bENemTSd<{1;dk88 zN|1x4^N^6Esu%6UnuL|6e z9o{$O0D0?{7D9Fv#F~p^2rseDZoAHbOYk>;NgR>Ua?ISC67W^?(rq}@AXV0mGOG-vboSq>NG98C$r^D z8RBral2&?xy(>E6jQW;+ZoX$zf4PSh5nqk>EDhXNEf=pn!v#PA0x1PlzFZYzGC8`P zjU6}NZtq{M*@c>O*qpgK`Q;G|)a{^;%6k1TldeVV*$SsQKeuSeC<|{*5@S3NRrRJfr=b_B-!6o;2Hq&SzX)Kq&GCxu@s)`hg zD|agEHZFYoA=@xB5W{teYj?T#qRBN z^t^2_#^FcFw>{|9=YNv!w@TfUvBlK(*3Id@#Ya*~Xw5lds7MKaT^I2)wY6MrjAb2!kxYCY(qyo5oaj$l8$JtsDl;>?_7(}d(QaA6^*jM^KFEf z#1(WE;&M|EKVJ#_{y`H@#{V(D>Cum;A+0o7*9Jm4f^CTnC!Z|4bhE|XFF(BnUcs4BY%2yJRivg%}Mhq(sLYTO7F#-L3I z2net$p4;e2)h9B5(IO=EqvJfEMw={#Et$hGzD@=n+|-9S(@jrvI;e8jIO)M1<- z+L2uj1a<%1BVC2idGNhN#c~DtNh&9Sz?ZGAP%X{ADix|D0+Ev%K zrfn5RjQKW)41)G*0W3zp7e>Cea*Do=TA5D=QKsiiwWeo13Cp-yM#X#f(Y`-bP>%4( z^pc_!P;ygNBC@tA{!w(n+~);6%&D77F}}avjyFiDWc3{`;oQO=>PW`bLBWsI>eWbo z@%`n7+Lx&WIZCsD&PrhJVhNEsALU1LkTpZ}Z}+VSlIC$m@OB<-o%|?$>Kx~h4^ChQM`XLM2GghcM zuPQ&|yVkrXBCE*o+G(RK{Cwe+S~JH=uRh_`{S%jxH5^on-E&`o3ZF;Ckt^Qm6N!qs z3rRWhde(z&eV~8;N3}#r*?HI4Q;J!+d4k-2Agxe%d5TutK@mXTCQPxI=eI$CH4AN^QLJVP@J!r=f4xK^~0;5rJVyFrsgM(Hr};y5v33V~JcW((9`jnqr3Ob#XH#7w$hxu?Y zY`USqX?nJhP%O3lPUMT4zZtXaHag^v$CH=UnjO?$Y?0L!WBXq{#D9LupZ73*73Gnf z6Roxm*_W`4*y`p}4vbXjNMGl>6Yk#`;`9Y*_^f%vCm4y46devOvM|?S@43iON6fa( zb#x~w)9!`Oz}YGsbMJrsigKeI*^EFe&z>I92{;24O4`t7{wg@XD<)FYr8NJ>XqS4S zy2XjqmBUR~AfdcBouY`VRpI2zHSd3PZiU>g^D0D0$XOXn!JvZGLC^3)@$1sLVqP4` ze~kt+9_CDWt6`IbK2SAdM&mfb!HgiH0w6Ye!@3@Q~LQ=k129WmsJp%t%F zpKS+$V@R_BDtaTHb;f)ybG3&WJzj6G0`}-Pq4^`M;8(9kUB}dQXsfd)D(-#{RNaO& zc1T5AU;R2?>e9X^O0^*K+xdFiALpkwW6S*I#CtMpL&pDBnUi7cFRT9VTIN$2BE0#W z=MgT|xa}j@ocooz7B*()t3aTbXuD?EtG^v)LI5s})Ta9JDx+7aqs?gfdKaNL8??qE z?X*b;sKDBzGB;!_RrdGR%Jc(K*ZZ4Dn}ZD9Saho)iED0A*cS1LrIjcu@83jKHW?&> z?=TVn;f#NKram?zJjw)B!%V?P+{q-a=BO8HmEZ?)54!?EI6h6m*qJ%YsYJ7UY-`B5 z3~$c8B+W}LG&PJ8ZE1J@v>spzPPM)=Lp7|fjZA~$OZFCL3<37kr-Njf+GFvqpe+pO znp&XI?V->qhfjAT@xhw+GfBU@>~E;r>n4=H(r2e+9C~B5j68{*aHHVg+Nm8$ENz^< zfl6x!6hJpGp)A*L!WwS1d6G!#62DrjOKjV387apDtHAiWC^0~^>)!za@U$a#T@WB9 zjWpNwau{vT9lFZU+SjJjNjc}UP4%GT75dnFp|&O3DT4S>MnxS}`wBN*{Kar?>d(FE zE5#?R$61h?QtQFs4^r2W2?VXWvFd2$GRsZ16}32&8rfCM0^IC~OBj^p*^$PX07lki zWd#}!l=UUg%{R&A@<$HgNgj1SbPWL)6vS^zH_SYv5!H^}nlB423hm5fJ!sH2y10K# zprzhS!b+*q@@1-mo_7m<2f2!`hgRmp$!_oF^E%<#0xLQ8_`|x6^BK1K&)T_>U{_czpo=JY zdAbS4?$b`ZNR=TAs3FWyF6z00-h3DbTrGS3HqMA3hmDIZH6NLq+UjO@OsJu(bU4Kg zidw)~-3MvH&&u`}Hyy1f6efRNsfh<)Dk)e>GHco{l-=9@%keHCaJ zn?^rgvfP^ztGb+9i@xe}tdSRCXd0_YIXTLLfuzY>B6auZRaGf>ldTsP;{{)&ZIXPp z3se*lbcXO~1`5=BzZh>QoKs3Dq(lV@fdX8@x+8uF^I{YBQ!h3R?Iyk5B)OwT9-P#y zvdSzOkY9u?Po3#s4DUcuIUQ;obM7h3Qf+#6kXAl zldk>`PkH^@+kD<9ncZ}E1Srvl( zIt;woef4(-<_s|9>)%j^9{I&)b+c495H8rziP;ni{II2ph()CR0)Ca)o?)DB<~r`Q zvbr%jw}Z{7?_il1%ed^ZRRLlK6XMxiAamAwII2I>glycAZ}o)4DZa^|wi{fB^{<_x zzqa;jbC3(-gq$VKYr!39POaOP@rn!mIu`yew$i>gKV^E>i(-)|yIPi+dC0NS6zv`! zvqw_&?15IpP*`uihWD_%0AAGG;CUW{`NbbMDYt>2BBm+1(oC4SCx-Tr1^^Nl zo=-d?&hVq%40a#nPV?W!~)seex=={@9p) zWUq20Oe>P$Bsz+EQ=AX)Je-L5c)*^MZlHP4Bar0AXBuLEB_F)*?|` zs@znV)Y?aqroZ^HpBgx?XP0wj`vIxvDDb*=dHRUj49Mu&Lsqd-d0DNsPQ)q~mjJLm zIPu9+z{XW|FBU4@rDnjd19b;0nPm_& z3~gb$oxP z(LhcN`N((wV^yo-sJ*&C;=g!++tdChxsj2EG60jHOO*H79qVe5@@q^=qkAe(o^gll zevq&H!=BLt`1aDOgIbt6B0k{ahCF%b`v^JiRA;BC^s?XXMS5ZJ2rPwQ=c?`bW~)-F zQTd(bB-f zH_OcW8O9_{{*?=jyDf!j*mUFJtiF3j>AQuy2Wz(t6dctqp0c^j@p61JJ@e=R51^ju ze&r8|)QL!E0Y9!{Ovl{?GECaMwC;&AdEKRbf03U69}Bt8XlWKecO~1 zq%5Xh+iKwciuoCW#YsoMmP{s6u{YFNIw;y5f*k2o(uW84#Z)^+8w z4JVsWrkW#Fx$p2 z^SNYN|H4iZ4c}aPMfw4WiouqgpP@Nqe|xzQ)ef@RXIU|xdS|eE8N(jERea90@qV;J z8YD#G>#E^wB7=W6klfe;x8G>VC)wqBWd+RMK`T_h7@bP$-+{vlJ3IZ?E&y;&_Xlhr z!^E`|=(;^|+;5NI-6XOXBCt?Qinx`6obckdqc_8U|7xSfKp@B~Jv&>-Q_Ocm9hA6V z%sd6n8t;&>hw!eNNTA=2?6xmU>kGNxXTm@;stxbPMh%6l%dLRP70<{&)cTf7Oh?*S zb}PN`(_jEzrT6x5I=muf1DW?d;aP-AKsz(E%;*UV6J1qY#H(A_`LxsjY~)N`Y}LQf z`?E0x#jXEdnUT}nks0XL>TCz(-$ub1p z=p-^q!ho$!P{IP)d@xKqd~YhJAB&KhGHd}Ob9o%nr)=~#wy|w@QL>jR$jYX`TU->t zl{XRfT&$25i$t#oY6DOb{JeM$!3^mBTbIi0uppL6Y6TL>KNH&l!b>SjznwO#?VVn5 zhshNlMjCL{@8mJ0Z>`OTE;WM9Z^J4MvC*zWI)e}b2R}*mVy8?^d#eDvWSpHRD z8$Z7L2$s~tG0=;AdZI*TWWoJjESXBbeVyPDAzS|TQ`bO8JAAD_J-fS^1l`D}$c!B0 z+0Tld#c^IP5zRLa$Aw9i8Uzd_G@*K18U@J(;jfb_KB&K)VR1j7EC@))b1s3GF{h0E zEEm~3>$}?vmgj0PFO||)b64$Wt4WeCGArw9qAJoVOY~Yg_c>*NOx<{ys*}AvMni9P zS9mU#g|m9{5f0@KyzbbvQIg>t{Tv-8SQLu9Ik$2kpg35PEtKO-}#1d^YR=^jsn z0m)lw?p+2MkI3@lz4ryr@tR@x;1dGTjSh^g2IG;(QTCYtm;MV#yJ z03Ccl$;3O6{dY&RM)|!v>S~Z@UdrDa1%e7kL=V< z8b4s0AFi!7B2M~>S?CZD6_5O|aYj?^RYP%3N1djWn zKk!nl-WP(B_QB*f3C18^UyTW;-B<_H?uBAr7ERfRk0kZME!QXmWC63RWc{}gi7rIqQ zr^Sg3+cINiX}`$UlEedR%6qt)1o?0=sz=+!=5H#Wt|b(e?taa;?^{sRoS#y>`A_Xa zxemOp`5b4$NQ*gmcMIP#*z*dohQv{ZO92lT$b#{!Bht+TRp(zFs$1|2zu(K3-3c>{ z0o1-&#dR#$dRJ2J?F~`*TjIiZ8IN;81j{LyR6J-A>%lR-a4f|6JD6L*D6^>y>NKD% z_`E0}+^psx4h(j(Fk`zNN8VO9xih(3q4(EDR zSAWu0$?&VG&a0v<%uF)xN~>i3%KeTuqP(ky(D^$e&k!r*#=(g%sFXGgO~%{;$sz}Q z%pyiJzbL{NRHY;C-Jp&ZH#GrOBaY{aj<$p}X3bl;yC&f5xU|o!Hi~&Tu+YwWa|ldt z=yZ6td0noN;7ZP1IctQ$(DiWF-Uzkt?-lbC=qpy+Lxn!rt%&v2S!iv3jVx>I@zePD zIcess!F3)<#@yY3`N5`c>e$~k|n@->o6K}0w{ zkZYj>@wio*XZ)mPnN8FSO9#o59G{yd6ti7!kPVv3?nsdG^G1i%v)HFqLvhrjWd<-> z_b9w8>z!{|ky*xE_mgpV#0QdOF3)7&39emEEI9+N3;cVAm}0pZ_vU*mD4Z_9qsKK z{;2-vNtoXAi~bF_|Gf2S z-W!DcIhc;%#{k!eXQuW{tEyn5j~Ax5-O*K*}?-CA>U-hPgnT)dYh^MaH59tyq_AXbP_+iwS?T zx%4LwJ#b90j}WVtl@EA$Cmc0;7R8;7#$C?FRPXD^_>psv5&;V1uFUPO?8EholYdHe z&LnXC=(Fmke;iNosu;SYye;h!A~D#qSg}_g<8r68Q(e;d&Z~nZ(TzuQXYp$JYWm34 zUmLV_zR=apy?{9krB_@Pp&JV@7I#}FFI4429a-+6z#}ujY4!IgTeuW&_U0tPXo|$V z1seH>-FBsP(2{u{%9VIP?ixuLLRYVX1bY6Iaq)=ns<9O7oQMCAS7&MCe$5e(25?62 zg=0s7_wo6+mi-0?y`J56FOb*Hty~EUBrScQk+p3%toMQ@o~v|;#o?xP)Q9v9N7|S7 z&=@B_72g8^^ho=P3nyyIRIKJC*}^<3a16Yg0E@_atJ4YD**3JDg~gzOY2|pyHOnhT z1CPdirG5w28kW~z=L-L`2`guTyHq|6HEL5Pd9K`*eDeCK%Nwt=_qdkb=+xLpg#k-X zF!hDe)#C*f&EK_TAg+}0!|OjeXKROij4J{RY_?_+{vAxs9GSb;xf)H-rdnFJN>DK+ zFYtRhP6Jyu&Ei)}o$Om;qe5+*mjwSGchGtjxMf^7zs+5vD{`<$Njf70RDkM`lk1s8 zmIHgoURSgjXlGoxwacsekWG&#M2~!h@Hx~pxaC;WxY-V78X8#)vjr7{F|trwZWRxx zm@t!vTyf)xR3Z-(V_sLwGqWA07rP|9LNdu?2wVTHpjBq+x2Ubg5dBWFsPFLPF!!6E z25joDMHYs&MCC|o$#vJ@WdVl9c#mb4ik&cY2Dg1rlvduz)W9UEK8w5`7tf9bhJj^r zHF`L^fN{o*0&la-&7r7|F<0LQN+BnHm-YcVSe<+_e$f>w-hG{!!&Q;%T%ivuXx_s& zY53e^2czi38Qu47N3b6$moN%Ex zN0?w{+T7q4EQdKeF;{;zWxul(R-ExnUgp9=tDY`#Eer;)?ohOii9kP7b6);HP1emF zVsym&tk%zR`b}OsBMQ;Q`ZXP%<1E#F2j@){R;CZn^Bn`aeCc`OFV# z;wAqtR?QPJpQo@VXKk4&_IZ*oUchwC*kc1^1>c6f{Mg`dP~lh1hVe-OQ4(NRmuwr` z^5W%x!%W-F;sIzP+T=5#} z)$c{2X)g1XW8t-(p5XNwj%VT|6wZuIyFIZbXCgZNB zZ*(4lslx}C%eb8KUGF5A(oVay0_*0SM5R>i=rbwq*~U~+;p_O1;ootJy@ob_yN34e zklVcXIteZRn4D3~kRkwi3IuA@l-`xn8#>N_9=3*>wa9?;f_(@Zl)}~>BDSwR{bvoQ z3%w0>PrHFswyLCNHUfmDWp)l)6Bo-;`#CY^+{^464LzN5e$)Uj)o9^XzjYUp@N(Ql zFPjn*9^}5Db=re`kYRXv-PD6>gfwmTvJgh0SBa(=qx_-{NL8-6 zWz-)^Gh4Yv`wC#uADZmWXWxXQr=2gJE!|ulN#)f#Qo1X49ljhc+K`q_JX2|{zwj`H zC&w5mZ?e#2U0RlYRSU%ZW@Mo)bWTk7y1t{&p5GWE&t}F`5GAmoDB$LQ+>QN{xAl&O zS>Jw7-vrwz6&P8Ub+cTE?}XS+m*yRnqL5qnV9&J~j##GOCALuDq@0)vTGxoF{D!c=CA8tML6t6B-^cOi9U_NK11?NBL9b>1_TgI?7v zM)HGLM{h@g_7+k|7b}WV+QzeTSW&|cJlo4mxAW*f11Ns38 zOVF(XC=<)95fx!AX^Ha}Dx$stB1ZLTgP!BDUS36tx%&Op1`NahR$8_){<>@E*fV4F zT-N(aS`NN&O}=@h)dTS!tlc;@Z``&KkQf-Esh4brJbj3W*X83*ii`|iWhg;Djc>zkXdGa< zQ;!NYo4>1pLR!J;l>gt z?tUaPlH2E6AH?|zVA@f_4HT<+qE+ZNj02$kMz!Q&rZC>UL5DqA`<>T@9NUdpyUU5n z`OS?N{(KfUD({ogIP(1)Q+|l3>CHHohku+4C^RNY6=_+6wME-{bMoQrdN)pIH`cY^ zHiUjZ**s;3@J?MJSC5j(fQmQ*Xup)rF5`ccKLFZVDk}?a;g)h|sG(iht=tS0!=1}K z5S~hSZRitA6G}|nQQ91oCY@D#!M}0hF*y%30PSKe$+QNP7jH;$bk#CAVrb`G>NKE@ zrW24oD>#}*d+ZA&&*Yo3UOnd=An^B@-pB5{9FqJ5+VHX1AMcq^%BqdE9PmOvaxbt^ z%w!*)T%tzL4o=G}-K?ac>ti82pRAaMT>Dt#z9Ii7O!e=WC%T65-6@u$_8Lu@01i$I zobgmotMcON7U}S$j(;c6555lkqx1Mr^ZP4s-%G-{ry~P#=1Q=JVIXa5(zP!V zi0LD9(IA`q0NXap+Uv*1Po2k<$gM;=*B>k0wu=tZjb~aoU}bFMMRydq(@<$54#QjV;w_L}j~9$q?k-O)+n;T7K}3%WyhCV>|0K>=dP6o51* zSE-pB*Za$X^pdEycu{li+Xfl1nubHu{47$T#qx~ht){i9C4Uv4>WHm#%{TE|_d?Bl z&hJa%wdFsq%(FEoZ5y!6eqnU3l>;tH1?~!rx@Ig99^gcn-HRIgiVE#U#-x|xK7z+_ovg^+v}AJJ+!UN#@Ipo^NXKKSuXrxn2VSt-*0#RFY(Z~!DJQ$I@PM>~iA(>vJAW_q;64Y!L$;A*imjB@UWlb5X5 zjl6X~9QFXr)lb*TdJFe*Sgn1Rk5_63ou7_#YGoSNNNtB!K9K_4cpV#E9a0l- zWK8y9-I-SAWUyQn#f8^kMjSuwW71bBZ)Z|?<&8HF$`UAP_GG#UfcI9b?1&IrIDT$? zo{B_nZi=z@vNLRMc^K0(f3ADv3xz8Y+yF-~?#oCI7ha038WS(<(L&l)&pPVHDHH=O zven)~O=+Bb^+6_HRTL3G8lMRm<3bPinDZ@-cLZvZA|yb*#sgQIb4I%82DQg?e$F+c z44-oxx6U*}+E`R|%2D*>>encN4W2JzFu63Q6DKOff4{*@p1<|1T%c{yXtOM^#{Yho z;9@qBA`(FhecUMvAh8J%yU$vaUufoVVc0n_D95)t64^bxh! z+MxeVDn&3`w4R_*&g04}v?k+Em+#IAKkOx*+>&q`sBnep zFO9)5s-(#3M9A-J0VQhS-|ujLQOx`^gsbcyxfJ#9y(bCkTxPlg0Pw7fWTQG~we6Kv zG~Kg+>!8m^vs3q$FRf4z|LZwDfjtB5gNTpehJ2SXFIjzmXVS>s`Cl>W(?x9-d9pI% zXoD#8Ts5Cwxv^E;nBH9}B^=K(vudLaRi-gog7Ga++HA|H{=;_);Qagy#I#+GE5WyQ z;TbaO3SB|)9Wki5s{OfKfiR7YX~Y_xT7+$9L)ikWEIEq;SR?N6bskI4KiSc;zQe=8 zz*72nY|8Vda7r_sHtMQ45s5tf&{f$_{M9&iS&u`ZVo8DnvKT<|o<}*TAS=wadZ4tf ztf7mdhb^cAQV?6bnS!EVFBjnf(}x!ft02(PY1OSc+v#gcwxjP|vg{tG4_c1D;wCsJ zdj(Jxw)=iw>PQ!GN}RBd$?J4v3}RBClY2KTsZg)56k)o>_rfAZ_B>T>_fCiB!dKmL z2N;%hc{EU}`{%mHUiwl~B_zbZ$E+Q&Bek6W=B}$uAe&^Ij41na$ZN`%=k|qEOVJUT z|8*P1;()R7vNbnt!ttWCDF#}BsBooehV1Xf>q zYI9E&2}cInpB@ZQ(H0|ETaE}09`B@0$92DGX@yS|b+Jsbe}311-u1k>e_6gkfX+(_ zOY8qgp`Ic9VbpnQgtGHOo~zT&%|&ldwxfOp*_Rp66EEzc%A-rVf_V*ufbg1`8X%4m z)t|D#+FSg^B+T|T2hEG5UIR0SCWpCn@fouQR0 zOTs#6xFu$8g)b0noDvl#mqd8}K1`_PL+X3o5VC(o?K1>BPokI%ru)$7L0|DA7DOwA zQj*t)IO!IG=^#)}e^HX`6YpBM5gs`VE-W1jbs@Bj3D086#{lDo4-rbdKPXK_qhaQF zHum0@wx-;rU*=@Sw$OOLeS?VhGZov-t+tVAU>o%`r0uP|c3x8|5EQtlQ5UC|xeUr<{RZ%BYzp#R2E^Z9zXngf!IKm9;%d zE)fS^=OC^gWeQOIYf*qL-ChAo^-SA6Wl@=XozW}7WzEedx<#jG-k-%v$l&r{zsf{4 zfa>?vy}b11J{LXFLTg3bC4DFUt|qer*9PUxw!s@twwEf5A3y+TJVz&ITnv}$X2j$~ z`1q|2l6*xx{;|`_N*ovBe*YZ}qqFHziwMI>@q42MU~`#)2}&HyLN9oNp0!$;D1WRT zt}`hc(8dahxbdX2%@%#+pAg7tIRfp3@m}&0C9mP0wJoUK8=4?X8K4sjZnhoIbF6>< zqho#cPip+{C{D8*sgTR-ixzwmVdYG~SesuByH!^kQcHJqrJbH4;U6sGfDE_vOg4M` zVT^W<^)H^+an|?NNq`=t|tC#E! zrDHI_IwMNbhW)(KJ5t>730E($wU{u2z(povT-S? zHru3v5~UL|V`OeS#_YLcHz(N_&IQeA2j&@m%9C@$pz0@$l(X0f8WU$*uI-kutyYME zgcb$N8y|wZx)P1>%SNqLhS~9kB)uu+<@(30@U|7p-E)UVT%PNRG+>GH&9ORM%SeM3 zC0v0+Tc@-sjPi0BaCt|wZkNWzz2Gt3O1KJP~^$~L&u6PP} zdqqCFr1_pO>9#zv4lIYcZhPD~oJyh?hm5!u*vafJOJHLhy54xY4C!oYMPcHa%Zv<4 z6<`2uY&;k}PQjlG^8MZ4B?$lc(G6ohbk|l*^*{=ZX9%q9#dumQ_iTphnw7s`)hx<#C`*sxMqAp3ZkR;2ENEPUv z@E(h>#p`}(%5P`uTlp;NE!j^`!-D7jFz^3BT(uFROTNBW`}%-}(HTNpNCNpQjuTfA z7Rz0|4{;ctMOYA;9Gx+HbzZXT)Bwy0Y9Vg&Ha*n|Ow@RyZ^M)2k&%)m6W1n|ZCItU zp>%3mU8ir)4L8o^Ip=%2&fi21(H50K^f`{=Fu_UWFQt>@Q=&ver;xQW?TL#vJM$_a<=+4M$rOQ7U(uW;DPl1?D>#_=e@fF1+^+qd z$&_u=Eu=JB!;0ums-|5v0`72>nj>j<4Ial%8p;-g(2O`eS@9`bnUS!+A1Dm8Hs)Z8 zC&B^@+h7%(I?p$O5KDGX!AY8x@tbsKCvEGGg0%?-EBk8ivHp{t0$lotj2J%V^+V6Q zujl=fF9zJMWK%g0>*(3@-_1Xca#>3Oe|XOBEx!L#X|BNm$#UP=0$$(M!k)Ou>CoBH z{)?02X52v16Jr6V*)RC8(KB;dR0T=fEoDBJ?DezG^fFXKU(E)KdfQw>g}mr2M1B1=nT%!8w*n@v#d z0g44%-Sn?LAI=q00oJcEVQa)qbi2N|A(?%I*X<{oGt?F-l;12PD?2f@JjLhpe1ibl z#`nh%V{&k$2{)Q{g~9NAUs-pHVT?+Eqq0ikWK5(8Q?W-IGu9~GfZD;HZq*FngInQr zbyOK*SxJmj+`--bNu2b^Dd?_0%FpzdUm`RH-Ps3iHTN@Za5gfr(U*Z|00SxzwtS;~ zh9xHxE>q$^w7f*PZzC9({6kQm)Q;C8 z_nCWy+$a&oUL0N@*g$y6sGW3iN`P-pN#oF1lKte3?m}uo`UcN`IK+W-lcqnsR9@2e zLCRe4KASzCX1!{e7_4R4+n!~=XdESh4@V|ez!wFv9;AQ5s;4pNqnC;F@z3G$ul5_$ znW=XYL34yJ1Af&7=8|Hng6!U7k#oO^+(!D{LYbs~k29G2s@IJ#s*Fbn#di=9y|+DJZr!|JWg_TkTg8&wacGUHr#k@IrYsUD8r^ zTQP#yhDic|Y;6c5@JfjVEFXT37YiW@)q|_kRVm@+%fZG3NnqE6xD9!}xow02r-JRu z?U`F7T&7BsawRD!A1fD% zD4Qm!mcL;K{*;QY{Ds;KR}n|E;Jmd(jE^0$eokdsho|g5%dozhE;WSB#necdOif(J zT1@+HrcB>|?tasKKd22@|#k}bi; z*Rv&$Mo{CPeU@dd_8mzlaqL*-og1ZYqW7cp+?jNu6w#{&UT=468Dsvur|q_hoh9j@ zp+hpc<>PY4^Wz8Rh!7Ayr@f8@lk6M+UHpDZXzxMmi4!Zdbb{nCgEjuO~vL}_kj z7$3aIYk{RD7ou)0I*chw!ehe-q>O~rJ~gj_X$3*+ZzFpHw%;rM=dfo8A9L-I3Y%rvj%rTejv+r7WXwB%#9cfzt}MC^D%hOrw6Y)~R#PE* zn2$r}q#IuT^D!TRi173g^J@7q_C&eJRY=itJm<=QQDl3r^~^MAeJbENtIsF6Sx6FU zX=>bl0}>Wp6(%!Q{Qs`JKSP8sn0<1D$kOJ6YF5^V$I6r)>1^DzTYpDLR4tb3*!KYs zHMbCKsFvB2dgl)$jatp?In5nLL44bgydR zg>&Y-joVei_0!+Z5z|+C){HtYA+bEPn)HiwkDFU`ji`|v-E4Al5^XOm3ttimP!G$L zU#uVSzCj%!GLKTZk+!EyR_O)D=Z=T0sWi#z1*_q%aF+anwp7lLblS=_jBipNU4ZD9n-L{Y(0MtI(;MddB8er(WaPrJTb@~quL@W zEo=fRSQXpEq4X_WQ!^ps70|Wr%b-PbdGVb}$af-z#QVfO{xgfD^^R;IkBqG>2+odV z`94=k9DCExCYCNO+u=F*rV@Xoh^c{041!e#ldtnZT!AkiyT}`GdSZC_Fd<2N=?b>7 z8wvEu`yE0J)6Z|wIBa2D`?(yA9fLuV+!C{-@UhZUvAZ%l`VjbHZ#;gTpE9?hap}aK zFEB1(QdYjQ%So7t9NM=U$6B#jFjUS-gB78c5m`KtVEijqX&kHL-?mOT?EX7rd4{0z z0wFEyK#}d}4nrQ1=;_plJW;DA)H;^y3hytbd}KtM{}c38awAI0nJgc1z3e{T5SOw_ zqN}={^!5)fzS)hFNm9Q{SU!45NWG69=T*TE!#75lpFtsQxtLglU$timRj%M_iZzWUYNeF>1fjw83GjYrorup7(uS|I7o(|?WkJf8e*T1(}2>k_v*q>|p2 zzn+ouA=#IK?MR&xp?ud0dgIJEa)6#;ow(y1M0_i|9bIwWC*HJs0sB_gU9nfsn2owM zV4IX3p2@Xif-^~i*G>j$d$%kH-KgCog;tOr<_5IU3c{jKvFE~yMQ#S!an@awpzZXR zY9d7K5+^FEWWKimfJIF;E(G({!>_BHSl)_M{A84~YThHK)AM$IsmO+Dk#T2(6~0US z@}U@md>(-Iz0l-q3UCr}iN{FUs?+tx=|jk%)8=-L{(kAgsehxK5YfrJ%chq}efWgS zu5LNIxCc@*tDMdR()M?Ap@O5cCpdFg8zORH_rCaxbp96t{|g(n*_b~*)a(D;FZ_k8 zir2o@;3~+Cw!lEW=Ty^%mmJU2D1V0zaVi&L4>f1?Ask*nc~QAtzV3Vu_1%b#=At9M$jWZAubxA4-3<`cYUSL z@bZF($cdlZUupG@-%}2%pIyFT?Aejt z>{ll-J#T1CJ^|-B68YgLtrs;?xCb)r+8y%sEltjs@2PkCD^*Xn!YhdbW+@<4lv<@Z zc8YdnlCk8<2iKcOKAED~`jZ}PRL0R{&4d}UV^_bnF~T*`p7@5L@cGAX$3Ck~6!gB| zFzeLYR_KlV4Hb&50TpNF=;rO;kP>#52Aq>rBMaa@upJEX@h+N~hYHmB7WGGW<`p_D z?}HsWjLS|3!oMBtpU^pPyJ>mW+X_iv@zGAs`>ogZeWp;`MW>>NmT6)~~RdM4N# zjUGrQLDkJu2>_p+=paU(*B+d)hemBV8SFRc&KQ8BYy%jR#G7H{+>QLXfNZHNhVdc$ z-~+j;J6TQuSWRB)IW~Acko}*@NqvNO(jaHorgvm5;rZsQ;%= z`M&`;&x`;x7yMZ*xVcE0nY_+=vX0E`=R^023|>jRb|`Bzd?@#?vD zOY=39sDIqQ)?nqT{h@ze!!|K0Ihr}~eN~av(aNp!=C{s-&EwZhQbh~AU}4ILy5wuB z!kIRge45?&Bati9lp>^`|3i5b4okAeuW=Ok5*r564vb9mq5_;l5>M zVTwQ`P9uvRA{1xkh;C3$hrCExf&u*ciU>=_q0%>9#xl#cnw}pz(6SbJT-PgUGHE5~ z2j$+T|1 zpoUT4n1UHctAKC89q-Z)y`4$btx|!WwZ%^(i3sIqi&bj^BGKGrUS_%=LqsL!!@DJl zq?3D z1W1MH{Z1(iF9!TeD{f+p>X#1ral`K`g&N4EDcwL)|L44I+T;NFTiRty_hH?IBz4_2$L=^xBGsFATa&E_sXMDh1I|K6plyInti+so~K+I2jusLc@X z%f<}vWuupWguY}wzasz(cF@p84ZLBzcwwu=WS;q3M1D*-9{Y~{<%agNLlv%>XV3rk zTb&^Nj7PMVd2dtVHmh$@e3+6}rf;Smnt`lwo=F)@;OfvhSgP@BS&dN@QO1cSys~~> z8)>Z_U+v(C7E=&A8#vE`GqGvL59X}G)KW3Im;T+wRn&{$7^9ZW&{C6T-<>Pkds>1J zCX)fE$Qo8^7K2UC>m!u6I*5W}LFZ0dzz_GY@TIN=SH;I`_F@jXM**&jMJnP-bqyKh zrKwhxs$qQ&{tr&@{F>-%GSRozA7)bsje<*1&A?j-n>l{jq_`q9LWC*4n;Fb_cpV!r z6rD-huWm4N(=5}dMhiaw-LNKNYS-<%eB35qK30rQr;spo2zTDOOCS95v?fa}SSZj1 zkBS9L4@%^?u2`!HJ_rUq;NKZz@N8Me0lkbypT4!`SH}5og{K|@2GeXWKY*791Z332 zf;L}_)sHv(7Mj{SR>vTIDhRaWvXL5L?Oe7l)~MThnkFyGArHMpbxxUK1k>MkI+rLF z%v-+0*=Wn;dpD8N7bs#{wcWe-t_rJYa*Hc(j~pC5n?XeMxvv%7D=k0h&hXm|-pOi3 zmDfEoZLCVWEmQI*d!&Wi-ggJBp5b)$@$6ll<>Bw+&Not9o_=@8A3wl-uZ3z2p#{FC z|DUuu18J(gIjFl;!_=*bYn{_t+H^kZmTOLHoCogSV2%Df`^LBGL%_`+<-`8-=l4I` zuW^Wi4B&k#@Z$n`y2=T65L*(d!zDH9*SPY;B`}uQ!>geF_XmtsQcPtHGc`2AP9

    7;-CVk= zo--mUhsAJQNu~n6zSuLjp}HrOFnwl^;Ih=Y*)3o)NK=HVm0(hF>lv2#9j}H*515~IEG{RLAwTudLEGh8@v&lBxm__?~#*X<6-@@_7 zeMaA33c=_9=(3}GXJ(kuDlcrR$QbvcO28Ou5wS+SWt7j3N>qY*<@=SW!`*bMmBYy! zjWbjb87{rjEMHu7?da#av^iB8CcdW4it+%RS8T|}J2RWj{9HZa%s=YA`KnJxec3i- zTYL~Zj%zj6&&5u-XSC45wrm*e_{i8pzTIl_Za7>2Hn|Ew=A@&nDExh(VOzp8C+BqR zT{Czy+Ei(dF{bh>cQWSWIEeo8#-avdInnto{4lY`pinv%e?-q&&BAs2REf}kwEo6T zL(;4R_5J%YpBtxf)vS^(0cs~uGBEnE_E*9EsprED88Ya{8E8rMvvooUqcGq5xqyeT zc96{h{ZeGDr8CK~BXEbFMGE4^J|HdPc^k$5cI$WSCvO)W&K{wCW1J_e3pz zyDuV@mue;IArXK2b3StMkaCNFqDgMrBM(eXvt! zc&|4Tq~51sI+;k_+Q6iCqzi9x(8Ma_YU1M2QG>%X5H*oHLef|_|MQTT5iSI^WJdB< zV^y@=iMCtm{-WX03&X}1%*Jy~ZvcsV>2{V|zSk}cct;sNB#lykLse&%ZZ$3x`MI(# z9MfDYk~q_sZ(YjNtgR(xZpSMuEt%+7%qklLytDQdFsGIdd&wIfzluO|8327OuE_N} zks818U(! zaPbudWIk-F6is==npexbHvRT3ZS6?XM8jvuP>c7_bOO@~Q+3d_ez&t^5NtpTaq1hX z+xx4y!jQhy#I)0nr%ICcfkCZ z$Qpg>)!&)FT_yL4#`hh6{B7;!83H^G(Dg<#bt~@AGg#iDsL%c#7V!-Ki&jnNQvh!B z+h^d$ke5Dxf`$92&%f5%4*E^L#-Ul{Xui9|V2tFIvZ2I*+7X|OW%@FZ{tDJ@=jo`N z<7?q2$GQk&IstSfGDa39h_{-!=wcZpIANNJ`bo+mE58c(b2{`4@qa-2*ITlDWzOIs z*paARHI-2z&Zw8 VpEf{YzPczJkK3#-m|6rzy3g zKs~nax0)>Vab^oX&sa*0lT;&#uKa2(c8I2O!e@bwnf!O&?zp9g=`-r8!G|0GvPig<#v8=1O6M+(^_haL@KYZJ zden#fWJ;)&ppZEGUc^++US2|Zv69fVrp-<6QF}%b=YH(VI$AjX5GFklu$G4M!tw>1 z{m9d=vSB&G#XHoJ4y*1E7+r4omTf>b^_|L7R$0gQ_ljk{Z6?d*l$)wzjTHL&c9k31 z^~Fl^U{^}6elke2j3u|?Vst4N%mxF!%!OcEaD0N`q2Zpgf9xRqyYxr~Q;BN3o~5d; zWj?2SsyEFt1q8OMLl<|lFNQ*W>4GW3JrCmta^nmuOfve`hc=B3{4k!^ zDstTrBU+s^PnD2-0VVFuR~sNF7BP-_h8zE9U;hUppFRKAw)~w?HdaJ!R#VkWz+cTC|9bJCY7x9V{~18n zs;#SK!zg6n!9YEby`{f+3Mm>4!A!7Hmiko1YtTP>k!f=7fjTj@3kJ}aoNA|)g=70W z>3^{@diJVxf*8^PL=4F#)NY-!aBSxZv61hgXGdVD zvDM7r+6q*^16O`vWPT1|j+bqU9sij9jf|^&?RbXDtKkWAa2!u zVyY>)wxDn+2r+Zs6$l{e{-pADLe0!OUAj3Lcr3w_+<=6D@Q3^U2TA?oW|eG;qP}Y{ z`j@`#IuLWDa+NGf0N%?>2GM8Kz&d&Mk52};hV*>Tj?P_=%~@oiyToY~>sGb)<)OyW zan^HY<~{!71yF%Fg&==F?#oZlwya|ObLtDe5DyuaScD-a>dCQew@(L|G7(%|89K+- zsz#^bz@J}3ud1Z=f57@zVe3)Mu=f(w7x>FtkZM5-JFk>&xz$ix13su!m?`|lZ^t0nIK`=TFEXK~IKMy~2omd?x*t*Wob1wt+bctdc1Px@JZ9fKI}0Y`G(>EXd} zd^kRoa;Kd*3n}qSqy_LNippV>b}VBjV9KGelTWO0<8S~IYD1x=mh;W*;hHO=l6a#| zY*>}xM!-C0NCr}zv8}50!^C*DQ`r=@r9v{2Lnf}c6jO=6>E*et$YF{GySnBNY+yCI_hGFYFO)rJxBIH5G!(^qkdbtZaGd*Q zQ#C9?gPP9aQ(R{o4gr@O*w?cr#w!ZT@a{RC6z@cm1#U;;m6VO7-++0+dQQeO!LIJ~ z-9zQC2}q|2x8spzHDiz{wK1R7%#`g{Z{TwdK95s|J-Cn$Su0HPJ!%!EfKE6Ek#Q6d zH7hPZv!l>XF_5PND!kFSrM7f+uK64gl?7wsgeRf{%(Gl{EPkKHmqNFn^>Re*GId?z zuG*yvv!*(f$IRZAe=Z!JqaPC*AJ>b`Flo71+>A#9Cp`%vIj;yBk7~pdkzAs?cNW5!E`Wd!9K{L$*J#>?}cxb?3ZxE&@L+fjh~31)+qooD&=d&Opu zsk(NQq1{}034BC{ZVb}Z!6+1pgbqVQ{I4Zo==xUDPBLHh#J&Z_HWSix@o`k{Q$ut=!5r%W_fDR zKnC7^8rwvPAJR_gLm|qm(muD>+yE$qbB^jBIA1eSavs7pCgK7chv|f84Ozqq!OStw zzV4iG>VlCrmBG;6@|O>;Td-lUoU3t*)|--Ycu`8ZtVlqY#QOu|)T6?-qx^{?wDGVk zTOt@+S>bs8B}z;Dui-5sxZu8DgjS4RD*j?7^Wl4N>fL}(b3f-LloMe8v1jg`yEBL) z4@jq+^K;8NK6vX3tp0}*FO;9o172CYr(<)iXTzN=_U3X4hHSUSCQc(H0w_UDmyHsK zW46I+f&~LoE_V8cdmcy0_eXoF&AxQvF(sX;GFB$6w2*m#fR64?u_Vs25Sn&w-J{=9=L%7PJCe*Is2<31Y@k?x-jX+#jq_qj|YP`w43r?0T}} z*BNY?IPECru+*H;k5TXY%%Kg>S=8sRTb+WqFey^x_C21L3lXABGR zr%%?7Hi^j$S#&7(a_hpRG!5~}N8eqmLWmmZeK*_IzDI=UY!z-m- zfo&o2yx2w9G7$Wq6|f&g0|h%vQAhs@r=RPqGR`^m;Kl>0nP74L&;yVkb+Fa zxq+N!I%T8fb)b+66WPdADv-9X(rpFDDzPhD?J2cRYeCEr_JCvbAU`ro*|fUe@*!mc z)ga;HS{M)Z7qn$G$qA!$8+ZQ`bmwuEqMs$FG3G#^LJSw9VJO7QmRB^^x0;oFWRa-O zYks={^Q$x;;Oq6FXx3WBEIVw$u}_)ZUzYAoxz|A0BeFwV z)@C6^uU0zfR$iC|eMGxUjW=ZO)ke<2lubQD(mWB>CTjng!O|}4pUv~vg7Mh^qol+X#WK~U9?%#^?$4{-c-z?-sh(w}SIOfP8 zBO!K1OjHkr2Q6J@#i){dkZ`{%0z&#v@q{=Il9;0v_sj|>BY4)Hpo1K9-l$0qmN(>R^C&k@)-ZDZ{?XOZPjpz8jh`$6N{@<0 zPVmY=NiuG72zay*tO%yW?N1v^i$kKS1Ms0;E=ZZmFV**&Okj|WXP}n)oj9!G;~VL> zXh`L<%}Nxc0HO93sKk)cR$0Zkkz?U^5Ku^)IA7OXMcZzBF$2NW+`5gL(@YKTvXoxZ zX{$1AP$1{7NzXUoIL#z5ZQ)k;`ayHXiCl*K9zSo{1#{ZBi<<&Lw9Sq+UD)4B%EnKs z#JJwF2-1VSMPoaT@7%gKa+t9RK02cq(VwVT*9@+DD(Xv4 zu}AY0t$vZ3Qm77Sn0Yg<1m*seve)|B%?ITRj)Bo;-c7X+p9)$@VL&r=50 zcZm=!MPdgGZ1MyqQCfl&pMVA3bjCnX;Lc~3{E88fo7V>j9l|ROQUG;hE0T$fj?fm6 zd6EJ$lL{iX}8(i~OIQ!-el1@2-9x+^cETF78 zUw-<$09VxIBwu@i!RXr2{3hJSTlfULmK==q>59qQ!60fVeTD2S>auC>$!Knn&RIZY zay=fb+-}zDI6pM)sBJQ590l~R=A2T`mrFy@x(A#Iv-#!a? zl=}FKM)0Y8`gjC*VigAmo#htx;c#hKg>pLOI z@9Ner^J<|cd6^u{y|rckSnTCKS>I4#pFC=WtXXJ3Cr3aSy`Z-#=c`M5w?yI0f<_Qm zj8?wH1@MaivpM!xmU^}EY~pSa#pp`i-k5+OQr zea)(K!UPnp)_R4cq1`9maUDR=j1K8@-&921dg}OgqPJjsdB1i^H8R4D8!QLSA8^Oc z-f!*J^%<4#seUK;ahF}1M@*rjXQZj?*f74#5cOcYWjr?`6Oj9^;o#2Jgd$EVp@w-H zRzM_ztN(-S$9D#tZV;Q|VcN|ke^Tk{?6`OvRLCsATMd?(Be%~%y z7CQT++l9^&k@GOFv?xBpu2!8MGL|3D7}$iPEZmictd4RZKI>Jdj6TOZk58@x7)9;maJ&;+Zr6 zj{II38CBF_ULB(VsNrSpF)J1SICc<9BH{JVMqTbulYcA__!{}w6_63%vopG2SMHWX z&GAyK-IZPe)$wRm&kW-?& zs9cmarP#sIq&?wLjW>US7JYx@3W-k!&cX<(YlXm5TkkoJOi1F4=mVG@<6j%6xAndRnj ziglHDR~KI@zPW9kzayen&V_oEJ9%OBNGRLW&=`M9mCYHNiq_7FgbWW;+?I{TsC8m0 zc9d|ejpCitsJ?4q_oPH!77fMz{Ro0>6>RV`&COgxM!x^3w{<(K?!z@OMx9F?x@mm2 z&^LmlJ9(I&W(H+@#ifi6OI<5G$iZh%wC!?-2It*&r!Uq{W*bimd!@~=yE#%?R8*%f zQ_67#ryA%}$1dx08&vc|&9vKXx|@qSX_jLkcU_tWY~2dBm7iXe=he1()M>$3gbmP# zI}3I-@cPe`Z|WS|nQ+bI+-k%jsksBYhG%qEJG&{kv%@#w`$S@TJ~V|mAcNBXPPX%D_N&>!SC zbVeu|HIt~0cDK=TWboN2XP;C0*3-4Uf7&w=ERt7vO^*DGvEHaO)wU6NhWPPsFT+K| z$dU_}!!Y~2dQt)EE+S5U1-B>e8mx#KUW^WG<9VaYu+B%3Tz9UXr&4UI=YJL@%d>!K zeiWBuevIi+ziTeOJ2B9w%_n73>thk)>pJwdd+^J)AN(32g9NO@wq*}59O6V+2Xg^_ z9QscygjK>HsCFiFX0A=YJ^!;n|80|4(L`hewI`fH(-)YKF~VS`w^g9~ewU<@3RMHP z-}wDEHo~f9X2ZUu3OZ~jcD5Oha!t5;?R2To2XWM594q$-u)=77FGLMOf|k)I%_5+5&tw$4m5ww%ay&tPQA4<8lNZzg$pL?tO=r zc&1<(4GB!5a&3pfK*QEIPqY!${J{p|RZVYW%$i8QC@^*CX*HTJP#5pCGWE$D%No`| zVZLW3;uz$?m6&a~^yL7io?Y?sy=&HqbUFc5@E0kJgEYGwg-`nC*9~Qkmp?PdTx|C} zU-2drtJTR7d!ugzEaIl0R8u!MxhQoQE!zMinqy{_Y+7axMcp*Cms?FmJf1xYOVIk} zE;G+K2O%#jW1n<2ywH=r78PW$4~s|l?nZ0(H-*(!9pBDBjLhWNZux3@UUZCr2EQ=-=c{kmAB@hNvl&0K z{TrM9i&p>fPq^IYy}O*dX%qQ9r9J->I=$Z&MT=%+_^IEP^En5l?16L z7$F%#QmBe0K(+d?`1*goAI|bE2S{V%BrUBl4p0M6Jubs;AvZ(wxB8w1zUduMImd0# z>dk_>wadZXOV&DJ5iX{bnFUKVcd5m?#rA|x6jg3sDyM7HJ0tmZwPUKV{Z89_KKcHF zyfpuChQHRKh)&K*T&^dqZ43yp~|8zO3SA?NX(bfvhVm7w!GXz2#NI0 z3uM{%-A>1zkI3HqF05y1m4;F}d)^%|*p5ATUPGYF-024mF~xiRYDmo~k`4EF*I-Jl zF0lt$_L^N9)VKFZsVW87s+mVOQR5cP2g3^CmS|+(o!ebY^YxkMDP9#$B38 zq_eH}l4*tG*|WDyC&vwr3z={v4yIOI_B2M(Z*DI~yzKLKg*I)hOR>`K@3X#7gw77X1&F9q^+N~dGtCF*u zhJ~!1F~qGnf0@~;66$s_gf6C-Y8Ca8OSJ10sEP0ej_4DjE#VGo-}w6ZEUyeq?AM5O8owM{#~UtnBZ# zR_%IMBm54d@w8&c*?QCJD$cs8?Pp(HG!Fk`GplnJ*DHL|tBZ>whH>Kb<935<7oou= z-6K!MVYL`vwrQ5AaVPTKhqV$}d1qTMrD!gqf$fOG9|CrmePskrqXH9?{-NExb&58y%?xMfjzvhv%C*pL+ps_QMa@@98Dn_TT(QBk< zm4wc4o9bw3dj1>|DSub)A{PxSX5oVp*5K{+CH&A$b5hYN*JF<6SpB{;IcsJmFy`qrww7IxT zl>Jiwj3UkFNU)sObM<~D-ptdMajSpD==?T;E0F<}`6|Qu@l0-On3ZlE$2(HT!%ci; zc(r?Y;E<9q2b(!dx+U@acZSSxsNR2EsQ7dNAII*u1`pD51VN2SPiN_^wa(pMlr+va zCkbMVL<1iUYmi&~mDY}6c|L!h5|H;+5*dTI{w%d{+3%2N^*S$s<5rHnFMqDo=)pJ* zSEx>O&;OB5?=k8n%wvM(>&HrIeayi5#zAh&M=91iOV+jRmQq(SuRqWGx3#-yiO`?$ z(gagy^UREMuD|ehC#*f8RD;wSxnc2zqU#dFvvS*1|6Jvhk;kl?aO%ZnD)o(N zy0F!aVWY1SHg;VKdw)bcLxwmV7ZICWqOBWSXbVthREC)ud^FLrASiWUX+nZ}Nrg$j zJVL^Di9SK13WHS{7t&GSD5NTPU|Z$-lGtVgmaJ{1J!CmTLJy^6C>90QnRUPWkT<8D ziDxXdwTXmxT#KezRk*@@=6^h@w>uI2Kfc~Ns?BxV9&VAMZE@OCXpusJ;Dus=7B9tJ zgBEvpiWRrw7K*z=u;6Y@ad#(BG{ND^`Oe)vcbvWN@4ttP5%P|VmATfOYp(UcoO%)< zNV_qKo?q_sJZF#b@*m+KAd*ua;`WaVpULcn!x*=T(A|oa=a*-Ak|+l|yk;4JW6b>0 zQFH<{YUMu3l9FV@6afx+&tQH3+C=Tne7>G+F|ofc6kYQ2#iPSX3(c(I$!^gZ+lk?6 zAMAW|;$6j|x(I~_c#E=Gc7Ut%)iF7j{KXUzFcznGZW*T#L&EkaI*P<@005O-0Gij? zc?F--$mM5E5fK590X3?mxT7=c5pdA z0m?I`0%^Pkus0hZ9&09gL7K}`yyUi@LYDM6jm)dC{T+y)kbfB(JI#2k}mm6ytRg zL*4Gz#N~|(E6#<~Wy8tebuSzJKbXWT>`J2|E@7YHD*^WD*q+j(;igyRWP9u~Xj^(G z-!##z{^XW@W;9G{ec+93jZ-~A?b4kuWkE{4JOQ3=&bPAdz~JyUjcjtCTTT`qhad3*re;olfrGbISw278(V?Sp-SZRF*@TUNYoNe_DBME=O11 zFyZp9)~cjxlOY>>@E{|nOfAp$eovBoXD}1+YxCX~Tz*qEbKj3v=4H^Zv|%MuY%KJ# z$H>G4b$M0X&&DyHhxj5Q4&#@^+T%&vZW! z#)4~^3mk%n z0*iY_0SlrKhY6n-Q`6QwR>$VYUJj8s5U<_~Nbd~2QmOh#hLOT*TGJ1I5St#zKt$~} zLyS$5sszS8?ie)QNJbpvl5KO@XpZYVl#PZcJKDyNHWGzX91I?NzQUU3PTF>GWTCe7 zkb$ykf)}l8VtE1pGe5}VQOhUZtXG848$YI~M>woeodeK1XcQ!477%nr{_fC;@qzOW zgYy)y13M)$el6tIE)t%D>EerxGLl-kqYKPA4)_x7C(@io}-`OncYQ#JG5am!q}S_EU7 z`ZWq;UY+<1ISJ9EfW?c9g|3r_#d zc4zc3E`sTl3jxTI*hDIw9O6UEM@v^~uMTh}FAvva$8VSDEaAEx=|%Y5<+En)A087= z0DMg%S#Q)_9qMa4vSX6&h3kj9{OTb%tNRDczTNvAwWaP=*0n89_Yz{!UXi_59lDEf z67U+1JK`EE0Wgc8i&Z_Rsn5d!=K<4_KxqnHbd?KUP0aaA1Rza@eCaW- z%0;hoPcUsNSu9zmcp=$K+F03_uSzeZ!v9R1tNns4)wb=Hwl-^<$&7ey2nC)_b^ROXu@L&X-UPfQ~ky zp~cV|R&|7lve&ss9s*zMW`b$)1T1F+a zABMm67b5#Tk8x1_1aKD79OG9@U4P;Z0CEkwHOjst>a6m(Z?L6E`qEli2AGbI>Wi8S zTbHmfEe^xPyYP{YPV-HwWJ&#zR3!A-YTs34$CP5)8U{{{j!}i;J!vj?;J}%#kw{Fj z^o*2G8uN>p+~(h#?s5_ww$JR#St}CUeCkU6h$Eq>Z<2Y{TM0(Itw9ObImek#TtGNj z;f3$)(oI)xSAp3EtN6)^FF6l_FeZ4HmU)}Shn|*UPD_DHPyifVkHB;j7n#2} z2b%Y9U2vc0;T_e<+0$1!DE=w=*t&{Glbo!^A=&#nH=>ga!$t#z#g;ZRTR+zR{4nfs zQ0sgpZi>oY=AvYrv1|f}ji7l-ya{65tQ3`JVmERR%;XfsW{;> zi~wp9!oZZ_$D@3@k4tA(cG|1?pv-G zx6Rn9wfS>#*P#$BPIo{(PLE|yM0iaC_<^kZY4EU!ftstS z*rgY6`4e@?UZWgKQqY8X_*>+dWcm1-C`!ulzV1G!%5U`<<1s!hH6yl2hEAa}%xn=^ zxEjN6IBB(w;P`b*^2jw6g?$E`WNc})C1`iMD=JS?O*xIXVvo83rl#&Y>&gTWu{JUp z+kePC{LOlJti?_O(Vtr7W-O}-efDjVYahz=%mN3Rv#hQ2?SN|4v2K&!d$+~ZeOfp< zSn=>?vxHRZa)F=nB9+k+oPm7TPgRD6^NDOg`|46(WNNBLl#&>$s329nh0Ti`f90nB z{>*;Ob9?sQ&d+JJr53jIQJGrAE``U6dEUkoa_kw(IMe$@S<5Ir=X0(E3;Ddh+3BVA z3(u6D{@Yb7 zcKECCH`WfA>yyi_vh$>fBmLzQ!}+$iGBNCH!ec(&R=pcHcPb5i(2z(DQ2_@r03E0N zP&u6_UjeQr;|5}i=gEZ!Y0}z8;BB4b3QrMCZBoN#>gHz^vRfjiv}wWn(KJufwU@)S zk~lZ%e#s%1Hv{1xLF8|8rKgzwSrl5v<)|HpC1*zNZpF*x9Ib2(f@e?sVk&Bbm9b0z z2;&R97;c8^eidL`AwkU;@uKidt=-c7!};$crYL37oP*HdOvW#)p7gCiEDk-=TCemM z(~TEuSe<^VxuV<4|6LI zUQcN{fUoU2Kl5IZUOIZo57Cf*T*;9&z1@!riKd8?X|l>4k@qQVgNeZUo?|2f!{Pj3 zb;c_V=r}7om-#(o-&Slwk3PU|W#As#ui-tO)Aj5{p~n&&TLpjDBU|j=kS%s3tTBg! zEzQml-pF|zEV^4xSmM_B8e=WLU4;4WEcBq^^2Yp}fh@knI|f+2XRzM5#5+k_1fHU2 znjAm2pYYVJrw;n8^s6}2)m=DVH9?ywIT1a`7qC6FvDy>;mCUB$nOn! zH%(*(Bj!4o<+110W)PMW3gGTcWo4YH~fY4{X2&k?A|9fdrP5#(09c!5>SRoN|kE&+5gx zEu>FdYlTaXyEUZGN9P>pAR0ku-^y9vPR#6h3$Kb(?@>g8&+nTy?+zGAu>K(Ge;-h# zhF5nf%=bd+2)J{W_S$4I;Bz2Nuvq(v^HEG`7ye&p-u_tjD+4CgU7}{wHo5$`wXl1H z$rAr+I=-<5;64O;;|h1$buAx9Mc=Qhacf(0L)LX<5NDZ}PXunN35@YqN@Fli(+D~!-~@Y*MX(cWR8r2@ zZ910gpT-x2k2+B^d%8j}k^N=@Y<*9)ml}>o#9W7sR=+H)(DH;EY9k}xx=++j3q!={ z4boG4H^tRt$z6p}&w7-z#ReW6-?%kWi%uH@9cwrphNZuMKkp*iFVaXI95~j}LXJ=1 z9AiXTFW}pK1HcaKFdU{N43{CZ9l-*vjT;?<3!sC|_y#cXl&JcFhrrI_PWwrlU2yhv zjo4OIi|IuDizU4Hca~+#>{e8i*8a+x!42A~uZ!naJnwF7byagj4h>Rx3FyLpTdgyX zi_hk(*uoRF-=tU}XvqvtYXo(raidEUyA^S_YmY(BrFoU==&iPE7-TholQ*y+B<$db^{P(?-bSUL+SIG!ub z6a%zc;;{kMiB)F@yP}^hFqh$jvIm)O<*&LX6*ympz%L@*kydY|5QYb7XFXmDI7(O78OeW)!ZM&4j}B3n&QCwd=cha7m|M%d{$-+Rlru}G~L6=u*Mb-(=$ zJwN*ar4F7J@ObMRm+g|07W>=^4^T;KxlnUnbYmIMaZ z<1%d@%SRYcC3Fqev*0R1>t2Jsx3&WZ@@08j5h6o)4UExL449>3TqQoBlc{wqAcVts z^|CISHEdfoL(Xe?^+)nIf~_lz*=eF1Deku3%jDfpw9D82QbX(cDf_KYx1N3p1y)}H zY(|x1HrriaRvIslnrRDDI6Aeb~W1U2aFKpT38SOf(%0kj`CI-v>tLA5BzW z<$w7d;;0craUA6I`!cftscNehd`+4!eZd8mL@}}{=24ufePI}GjL1rW!pq#Nvj{_NOpokV?E}with*gi9*)pu&sRuj16d+h0N>vbS z(pd6U!?#$^ukLukMY6x?Qs@3$=c=W2ZrR*EQ%mi&&%P~)Ar#L4iCUVW%DT@X?a29y zcqWPCZ1wz_wFrf3M3*-=*QjUPP*u0_dx6O`HQ6{KBUlthkndOy+x#A?zvEh)5lwu* zzy`t>oOo0k^fn|Y$?-5X>#QMzIEjl5TIFD72j=i1dqDKUPxLjt6zVr`ecEKKJ~)ON+9d z$TP7vtq5`^Y~l|{z|UW{B1C)hO})9T-QTc%Op-;8#iFJlO}CTr5|QheWHJWuP6*@N zGlop`y;CvSo%yGefOQUBV_QKXY!}zFn zE+(5BAa)hv>zLl4cj!Ka|3?e*0ftNX2g%_pE|Fm-GVXVnA)2x(1g^9MN7p7-5@tAi*njmLE8!S!{`(yj z8Bc`Z@-{dh;s4UxCa=b|_wSn=Br~P@nf$;oCySFyas5BjGI#1Q{ zKbGzxcP!}6QY~On-6naq|MsDHUTd;?wOq@bt zXl#suG1+yqu5vElFK1#x8>8O?sAM!F)@fa;LR(KIwL!{XG_R(nmUFRyNO04oRhwPt z`bFz%<)u#BFV(H-p>-qG8&m$aGsSoz_7H7Z-xIdt%DDWx4|Ts&n{d)O0uFoeqRU@D&eb!;{<4Fc z|Ju*dY3G&f(A#Y_5cxEhD0j?D!p(}tKA*Z<;tEzm&*se@$fLA9h;W#PUWmwvF z!D4wYJ))|H_}nzY;#p1qx#s!+Z)_=+8fbmezT5aG{af{l6voVmxWn44+aP$ZAng~Q zin3?-*3<&DK3q=NnwCRlBC@{9Wx|k86tJ_l>DfUM!T%f2$G;65z_m_>4>?9zXTp#M*;@>govM@IuhFo6fg zw={i^13?ETUI{0`&4w&-=dXd$J~}Cf=hC^j+IXGHDYCU4kQXBT(r>;zf`UHubA&(a z@+bK<69_6+myG)mc`y`uskBd^mJj!G;gN-e-jsV8xaU3-t0(N~dWMJYZ%Gqj2b8~S zXL;w4mTAV%AXvTUC*EwK5ae$GZSuWt6CiYRKrM#$o_AIm@ApQ&W5g8Zl(Rt>prQ#L zH1vICO5sl3;Eb8}S^`;*DBy{hRXyHo`gHUIyTTcL3_{gL-Zuc%#a?6Huf+jva+ler zRmf_1T8mJr>f7wW{B)_x@q6+gXA5dZcR#tsE%ta5;gC7xF;Hf`5Vsu+v33pZ^WETU z5?$E7O0GM4Ka`O@!@Qlh+_J>CG?{@!NcSG4vxuu1AXr{HMwf(^=TBHM-5HHrMH`W( z6(P|Gklkfm^c=JM^YmANrS!2{9x{xTx(X(oqR^FG<}-lFFJ$6xX*O8O9uw@+N_I`m zC;wQ3dYQ{>e`YBZ!g_(HS#di{!{TcbY4O=a4YD0Af7>~B1l-|XQiIP=c+Cs=?>j9-7%(NC$%b2RMtSrQbZHYxz zWNPsHI;PxsOtg76PlN290z@JKF+?K>%*A5f3&l_%M+HGg)WrU@za zDx-0YobSy)3VFU3JDEJnRozMoGG%;Fha<2zugh&TZYYO)_=vRnH}GDfZlr+ z`lfKSo~Isy=HQ+Ocuz|etY>v_bkFi=rLoz}^9jV0JHlP6Vw}YQM7?X#`!Xfv^vxbG zebuvrBM7>RsY_Xh+v(httSe{2D)P<-XC37P^gD877oLYO4)=SM&I~fXw8zsNC!^M= z8QZh{{Pf(jbi;odys#hNW`1rVN}U-09qlYz${W z5!f%GsA`38Jz~uK{_S(-rdV=UQHqf4S@h$HwBEpA?O6ZJgqoK3-#%ju^5`1jv3(uI z-sL&D?Lf-0!GTf&lT`|Har665ww_<03u=j$D)a1ZebZzk$DC}C|JE-42yvCmIPoBU zB36@IOllKv;qKDwTdiy#7PJ&HUD)%9v%-gIIrE4LHHnOvnL?{xIPh~@vX0=~tNkRumA$$P8mF{VcD z!JzVh1VwJ2q-)uzjU0IH^-aZk&Wl zQ1+vT%GvQ-i$S(^2dpMT`MYwFrxY-o@hb=Q{z9<)*5caZv_h5%C1; zkxQ!hVx=3@Cn~2~+gCvu7M^_8uSrz6%&!f$zc`Q;O)$0Gs)Lp5Wk-Q|!J6FQ^G0F1 z@1sE$DE(lYjMMH#TE9pb+mJdS`^_lkdp!bhV$T#^wl(kEfupe9q6;nEODaw87PYx< zy6=7@wGuYb>uS=GxV8b7j|VAn+j=b;%UdbLtqSHQ&JXY83qpc~{EmPLbmb-ZtV^7c z*7NV4dO|ji*Twk~HmA6Z;6Y_n)NmidDu;FBdznbFSFf@|3Qp=bN=4KbhYVhw?Wn@J zvLp*Z`sh^}+7o33a6ZTaJ(q3TK4En#xJ^gGt}V2Oj!$M$SZ?TP1v2?PWub@~qXk#q z__|lncMl;{gCuy+XsNm|T9hQ#YP`BZ(s>dVV6LFwFHFBgtqGNXI`6`?K`C3IVIt_I z^@-^lcE*q*uyJowf-h=}EoX*OVvaq_T;qA3XPDEGWx>IKNN;}8^#S;52Y%~@JY@5P00CO_`#66g3k)>(aB;EgYpDAOPPJwE>Dm5i_T z$q1~V1+eN`34D%0ok=dyL%INN4&$JR+DJIH4^iGxPxd;Wd(fs6O0?L)KWvjpOfm)B;8V zVp1iRD_9$X$oFF>HycVoJo3n?p(@I0V0Zl(R}#tR9mrlMI%BZo{Q+)5V8}R0?ptag z{|xMTih19@n_}hXBQpAUdmLR&V{-Z@D=zXaHW6D(+(sePl4FBd#{M&tKwDdh30ir-iXHH+6l8a~S_ z=~gdu6~EPn)zA0w7&t3Db$!31qKd1=Pk2Dds2|B3SdU{o>!lS!t}8OtIr`EXx1*5? z^^-d*{8{zs!8G4)W0#Qv2#{-vC$|KVIdDt+Dom%nF&^m=S8N?UP64xx;~yLL1>9xh z$1)QoW&vg3>*^XjLbjeilTXOH&j*+@wHba#&jyGP4$>4%-M>Th+AAl@=@kB|iyQNm zY-7V^a8S`EU5|7VQwvJ^Se{0u^!#*0(8JSDnvG>_jXx^w8S7t40jlo3e|l5&!nQ($sfHg8NDEWCTI;_|D)9x z%jfoj^oJXVL;(&HY=zKot>gI9(K3xQ}X(s2IBr*2BEuwL%C6qh!v z=6(4vSD9N~#WKm__G!quEiuBYl8ww{b;Uz(>l6gx&7y9YBS7Bu8Rfy2>Ax8U6b4?D z#opUk@Zh7Dvk_iel3?tU*i?~A5_tQ!%83_y( zk)cV>$f;DY)Vb^eeog4S|EHOI*Pm{u^hzPL?BUe$6`q zPSc`Xj0mfS(Zp1~#(W%_u4&hm%}aKCvV6wl7f5oEE7^rfQhv2VC5@;pA2>AtcUp;1 zvv#TXzmRNrGvA)hrM#=Z$`&;=a>gvu8Q9(mn)vSC`oxBid)iODl!q-3KxQ^T&HJU6 zfZkZ&Qt3mPcH%NEH{@|~o{J%J*z%~0&lV1%5=$g3Wi?MH;2w z8}<8zU-KVns3Ry;j}e&%qg= z#(|ylHn{{j76!;oN4flFxT*Z0uE+j0n8PDlHg^Q3`jeL#TI~{8+Bu%hnA<0Q(=%1P ze}nqQrzfEndgoN110Ti_sMb+HY>Tz~Xqm%df?e6D^{m32FNxX|hrRMN8-`(%9xP2i zM!4?=9oB8E!{2q+F)Zs^9hx=if{u zq>AKtb4Kbj_U3j% zg|+YaR-s~c`Kk3K!PCD~fIs>DuS2t}{{EL7l!ZBZtVY>$JRHVr+gUU4vy4^loi& zzVq;mr9Y&|DV>4I8NcPW05=d905<;aif}t7(pQuaam%-|(KfLU{xJ%yo(f9lvcQJc z@o{oPQbl4+Gh+^jPWgsinogkR4w=!*<}xEW9kmf zrCGFp&j-eKxF6I07(&->njzZ$qzyUvu@(*`xK}0DCcdbtxiySK_2{e2N;qx78;8t- zcGU_m`e?qSY49g}#sm{2Y%DbkiKxvZ^x3+huTz<=gct;U=4>Rd59oe#*Qk;eatFO1 zUc^`WU;|K|=wp}>b9hWmwMpQ4V;2weS0sxxWw^e!E%q27Psj4F8pL!wt~#p{VGcF8F15GJwcTFT)x`p!7{1edtgyZz)fl;u94wq!(KXMgK5a4wtYO1ol-bPu z&X-c*Not_peyidX8lHAd<_;mA( zz?x!rFQo2Tef&}9fd9V1BgTlA-6qMd3gge%@4^)4E>6fM6Gd95@%eyKEt60{ckFW2 ziEG7t(x8aGY_~}3EeX+q!JN8A#S}2*_?0N0q+`dk4FPb`W{haWVjOFy-+_-2v06nZ z^+<;IHAcogqQo!hqq6D@9zWX^?{R8kjIxlWtofI8ul!Scl2tp4_HBx!K17^|+HPdR zIu_xFsVU3uSIX>*)A>#wIw%to<#?5~B7w42Chw&UeLX}l<ALFhBP0Yv#Ahqa?D4>V3IzQRXo9l_dk7<^HP5Hx{~G>ix-Ezv&$3O< z>I|KdzO)~%c^?V)trCNUkt0vr+{O3t4lABkjBX{>7eOq*r9=fU{}DwAq@J!&*H-_! zt~YyGp=XdwA@Xv*a85N7AU)%eC6flT!sFN_D+5-4V!KIVHZT#`hHJ0J_2fzQ)3@qZ zWI#hmg9n{!_S8r(@HzQ}WF8bHj++~MGU!&yqb>k#Be_%-_XO3Yzs-Klow)#GMEj6t z=DH7RmMmwl0s|Oij%dcY1MPlpguWxt!u?1TEH=v85btuv;+gOmuX=fRS3{%uZX%zG zayR+9l(AB3ilmh0owA{_j+mIi?dT-TGpt^Ji43o=3VUp7iHhPIdK|U}{W^#0K_912 z%)yUy(j-dy;W3oIun|Aw0?c<}X#HC*;N5I;5^zU4x-?z?S|?~r7-B5boD{3{_REXt zUXa17wuYZd7;tT0LD=o2Cd)jZXkAG#o;enuX@t!< zmf#DWOT#+x5A1ceT&=N?{Y()hMnzWwev#27LMJahAr+c}yx$hj$G)}J0e#P$?4uA> z^ihwwqe-3xb8KLW@CCL=+NU;%fYP@0AjeZ!_)%un~l@%$E<1zuUG)emoW|8$Q4D{=(cxV%zL$?c0@X z_4=Q5_kV|>OIrIzN#D*{^pc@Yp!!r$PbsR>85FhpsINy5axZ(*S}4Ty$7EwyL;S2I zN9u;}#JWbMRm{@sjC6# z6viA5S+$xG!0uW0t4Zl&0U5f3TM%h9LIZC!qpI*ADpP1QP8^u*4iD7$kS#!Ub*ApD z7#uHopSW2B*7>3{QO2JbwJ=8+DB33|`B4y@RFT}$Q2->*`ZXeX5bWVIef-**dp5!O zvG#*3)MSI&t66LxVUAhFb)~2v;R&ZJpPA-(@RAE4@mFQW+^Jhd&tpK{nP#^rNPe?H$DDqz)!nB9JnyN9sNS~oaJ8t{5l!Ke2@rS3Av-vG_04Oht zi&nabWoIn$eAU`2sod+nAO&kq*nN54y3#RYl!-HeNtlnrKRMn6xVbQi&Zb1zzk!n< zl71Na53c_8>2wdtm>v@q;~}DA>L8=S&(o}HG!lCXWmgwDEefl6c~T|V-x}TyjAY?x z5?jkV@ErFb-_(2+vo6^)2m|sYPR4<+1sjE}3|mwIbc0g4aovVGac=UEQ8iLM>XH%S z_>pp|3XkgHpQ)ycjuZ}#noJZ8+tE%zl}*-&8U9f{SvsVqlr^q}fVXv#%SCf)<&VUy z{IcOba$YHc$4IEw9=X=hnt8r{NjPH}BWntg(T5SM`FJ5+yXepKS>XdwvJaY^%Vmm_ zd*Y`_w+96-=ZI!8I#|=n9Ti)gQmuj(yNVycFRq^`eggFOeWYxbaS7l({beglbuL!Z z8_wV)9**MRw8*)ZWq%TnJX~D^Q>Fb^_uGzHu#daCWA(B@>wFf1l((~cdtK$#vjG-0 z=B%(;^`^HvTePSiNmOxgpD~lzb7tus!qoGp0loYrK|3lf;9fDR*gdW$x9m;o_?pLi zEV1N2-(miCwHHZt9p~lw0{%G9bkJXByhrrLVQvga8=s<8&1(1h;F#y9W5#LG`rB81 zPgWOsA-jHDy>O+ERiHzmIpFX#qFo>#T71x>=iXw65YNPvEKR_DuFk1!eoU1{c*`LR z_idXr{YDN5rLa2hkXstJhlhFn1Ht__kNH9T%yaf_##L<2**;A7m%MWnwaLMc_Mbwp z+1u5(u74m4(f$&4+QSnc$V1Z_D#ZGh1ti%n6S$nVv25}hU11og(HVt7@_!YIU-<9g z>bqB&Y$y--wAt+d3(B$Su@ah3X5@TV==|aDLoE!1$+cv)o>^0$=F-8+ga zS57372$JULrBvXiOrqxiGwCk-Ry0vuUK* zT~lO;jxJ~YWFISE({sEfiNhV-3Fv+h-HWqU!_!^D^9yQVHPua%d8+&8^ShJdh1rES z(WMJdCY3W_G-&Cj^#S{?&|g3B8J6BlVp7-EyCV~KrM%Q%z9F$4P(F5=t|@R`vYCX| z*mZU?UK>6vzaG1a2<3SfI5oKqQtYu2+HL$rd+)|mYn6$ycbe^%?YX$&`7Qdf@(YBn z3^4PBGrpRIh_$d4{?vC#{#tj6Xk)QY=>#H@Ps;gEftn3(B+tEeRN$BJh7q*migl=x z({lMs$w}h%ikENh#Bf5}54_@EHYce?+1g8=5eDxeO_zjLSpafmooqTvfIXT3_}zI? zNo|g87a6Yp$HP%u;TcVADVUxfS6j~Ru1p`~oIA;E+8E3J{|vcGbqsotYIfYA@EQQz zQ;HH)`Nmi1u=gK5Jg=x|Rz#2glXoAWiN014d_(^#6t-#$&O=RLnW3f%YA=dNi*3cO zk3H5?Sk_=720;?t#`&@1oyzdg^rN4tdHYscH&Rj41=KpM$^TKw|E+@hQGh?ptt2Q~ zs~l^9nWm>E3&FQ1UekCQ){6jqBgLE zI}R^ea-6{BNY##MBQP}I`%!JkRpRAbG0nJ`I&sz}&#Nty12kRkU?zytx@E~0f0217 zGL6hs4|3fxb>1}?yR9t3t3-S6k=FfTjLcS*I_yQoVY6V2Dp|8p5}57Gt7d&XmDagY3`_vEZMVhp8~vdz z%ht+^oJJOdZ3HqA#4mCaN~~c;ikPb^SZZ>cN~$>X{N@r`Z-XL_Iu{2eBoRr5!!mR| z!fW^FdXuEGedYn+K4e7Pu7*@;{_(2!iY2}I{+rRjkZM71<`(9fm5g|Mx`bEz%~!<^ zdWB!|KOgTmmFHOq|7<-xot}n3zm;#t-^r|+ims$AD1;o$T1w54ClIX{;&a(%{Frr> zA<~26wK=kASgC;V=6TG0qpWofKVE%{=2dACZiJ%Sp1-2?T6#EnuA6kh$Hh|q25xAi zgL{d;a+TJ(>RKOo43|&6S1YpC0FBA?*79oKtIn5w>7f1#&Hn04|0&7#1<>Luo)a6Y z(U6tSb)qC4KU5-Sd3SfR>g8+8jCK5ag@)iz{aKRu|7yYb0qGT*KFf#^Wg0Czha?Ks zFkzIH)ZQr6sv6HYj?Cqxbp$gElF8#L!i&%%Lvb~@-LR}_No(Do7m98_2j%@ zt3t%}H*um)XRlt7tlNcOZntiRw!X6u6uOnTF-#?68}Mm#Wl$A4zHq&mH#JAqef*fs zy=G*xyENr;y%{sB2JCBgo#C7>@AA>lHuX4VsO<4e0Ca*#`KJeB%pO+KS!VLN4nKs4 zCdBFEjWcwSRD-v^8Xv>pc!nLHxbd&{OUqJ&Yz+?GEGSV;|5Hg>Vpm}LEyTF z`sx$G`@!t{n{VP9qTU|bCrk8xm zKO$?k+CG-x>5|GU(p2IF(rz`*Wqp`sR^%i;w4FjQ@Bjl#7q8}-Xn|IbIdODI!>+a@v>;T9`pNNL1~;(Tmt zjZg?G4Hf&mQn2NCaE+@u$*JPmMR3AhY?5VHW2I)*@A;ujGTxif|7;}g$+2AFl00i# z@$23M*1g8UgHQ`S`B(Hm*Utb1#Xr+}th|JPFYXH_z(Lq0gXo<5L^z7k(`osUl5^Ua~ z2^mfm2-`*3TyunI+SXcn)=W$vJ9dqXiPfV2=WX4&wa56B7kemP5XVg}Tu0<)$)Z;z z(W$&__G>RkWP(X3>3&-iumm280QM2HFw-QT;0GmdB}-#LvI6ez%6!YDPX zB&N=x<@&sEBtSta1TW96OCgXe3|P%0v0jur)q|#h)jDVnwpx42okSujhA7ONQ^1tA zN7+{0tEMSYF&+T#i`zZS%F-I=PG%TuyL#{6P-k;T&G~W$c+;_&iP?~x;Fp<1mqiy?%bA6}V zV=bgFAOsc7jjkiLf*{<)y@suSA&E71N>T2v0`e4R6z;rX3$i1f}*k zS#)W(WABr_>mm!fJhvv*^{W57dvTrdp)gq$(2C0e_h}1zZ+m;=gJ^S~mrsyRP@9^1 z8z(n@p%I#0{=kPg=i~VQ#NtZ+8wK9P9NNAozhZ3#K3IyxG`)q^vSJ%pItt9m1otKq z)cGieX!UFd>9xoZv$5eQ!k@`Po1Dh0MZ5khoKb&jVJK0Hs3E)g?gkZ`&OE0h`v*@G z^gFNWgth77V*D8*YQDZEq=|p`f;=|Tn!3E?0`E*kz)~K&;uCci>pm*)M1Qj1`mhv1 zA}ggsC~3n3FJ;P9Vc0o;L*`uR5sEw7bb%{oO}ulxw-%D@4bLqsr2TlUvim#sL${LN zAxKoO!#a89Q=$zqe+kXF(Xx9;ubA!&kpKzA6J^2#!#hNJGnQpcy0D9({eZSLU?rJ9spO&YL zTYs$=OEMEvK+G_SQJOT~Lk&&WffrrBWYc@Np9qjU&ur>7m%-5c>S96lv&y@}+6-oQ z6-muS!Q(wm$hG}Y*C*;ymgjz6t@9_j(u7!SX<05%FJweI;4E(g6Wt2beoX`?OuQBn zFh^b#nodW1LqEXXy`smd4>RNhvZE>QrhK9c()+lV;NcunA0b*~_hb(H%}MEG?7^ej*E9t?Kddp@`HTTVf$K=~av-$yJH z(;dY5vSeQrvVyG9m0NP)PkIyh_~Yz5;o4Jqh{nw@6N|S+f|^i0Rb!z^E^)`k{_;o6|0PC%z!7Vw5*| zD$?KWsyxeCZd1hqn}^i(@LTH@E8JD;KA4+vz^hitivB%1q*l%dB4v9cxgbKJ>=QLa zsbI3KP1K0tkM3!Sbd zL)Rw4i%9t$Cx_V+75$n|nv&1@qg&OY(DZFa$huC4lRS5~t4Mx3+UwnnVW%bLpYr3Q z90@<0vG8}R%HlezX~m5a$E$L@hoW`AQqy`Wm#FTqtFF3%oUCF;ztaKNOf24xy+mYR zi%&&#kiv0(o_{ zDI)IvB&1}=@25LZwC7s$bvMcR<^?nI(3YdalcVi(s^0U!5N>hT`_2InXsxk(O()Z* z2Txq4#U7wxRyQqmLC6LrB%26ww!-UbUZR9{AFNJd7pH)qjX7$X*B%JXAhyGm820}@ zLhwIDeLFk03=|^9(0a%YydO#$-s(P&IU2T*2$;#BWna*XbtwmnRzV@HAAOHvh4WBj zSr)i;S{dcKde}nIz&nyf`*}u>zWG-FC%6X8|B*pPP{U_LGHnKe7b-~>wRB=-L%#J> zC;bFfyZc263;`#>49&;6!^hVW=4%N@6E6tgeE;ulgzc?wVxBvO)8?_qFyslJwAM(l ziSZCJzvtI9V`P{O*IUp(Ioq?e)OGvfjH_AC!gc)Rp>$nKAU0}RCmpRG>OD4C~oo!6m zuYNr~g_;|0EMt{-y-mC_D!nZiHYXrv+Bs<+24z40Ou8^CVpapE&K>DdzHDEw9*NVl zaN_lUuDZ|$$G3?=)5XKJBffBd-IHk%_dR7*hjudGQut?`%$H{8jVi0(i(F2-_tu_P z%i-4|mio!bYOiv-%2hp}^PsXqjbH#hb9(>t(>*-8XT;(m8$n`Y96=2Nj2=c=lz1LQiFAHSlXnaiZa2FNh&OFyTuZaRTUqaTPzvu&vyqEA5oPBryIL7W~_U^{-^tKPnQ`{3f%HCzjB^SU&ld&>*?} z%J&3tJDPceX>DWUKD{zfkeTPW%{F#^S9AWW`y$y@i0(fnd=<_S#x%Mf7pWIhxT==B z8<3!`+{c{Oq0558IrVL0fZ;duC~BKj2tx$TA6ok7u+V_v#@Y5+Ats(mRh(F} zR!O&1+`_79=I+YptL-rb>8Y4wW>~3kqd1B_*V31Ti>K>+L`-e0e;PDl&N4ITgs!Ao z(arhmkNy^L*#~)%#EHYfJ5_RqI~0Za(gD>aESF>q1W)>&uZ9+f|z}YB4r3pZu%FzoGHcA z#H0DlG|99?D?TkC2dkOgBu)5ugY_YzJ~tKbXnu?ETJZki-ZZR}X5Ci;k+TjL5mft; znJ)@Dsc73?5FW-w{LJ!lbmct$ST~nBgVk661y{mvpLbACaHh~L?NKwG)^Vm`Ixoca z)U2j*M5DuC?fM*~BSMbh@y>7`$D_(m(ysdAVLZw+f6H`JXbH`J0$W0@rqoTpEtg16 z&AjKix{weV^X~AVkfcO`zMLi9G1y2UqQfyWns&6!M?LynxS~?9AZJ#{eOOAQSp@XJ zngDy+q?ejc^1uX(K93urmoFA~F3`Sn@gq!Tz8!HWxCOOnK#l8VcpkcnvLxljqJ`O@ zK%X_CG8f(}2^D8#LF)lDw8b&|2>iJ{184dFXM`t$`6b<51mdD+yat;N?YiCd08S7Q zC4KBI@n!JyIw4!T9riuwtt;db)+(qvcE~i=DVPX6M}L?uRB@Yjbaa?;dR-@-5_udy z{qFW46;v_far95fL=N?<<7fKsS)7$s&0F7Za-Zv#q)gPv^s5`&O?D4x1*5)B3ZL7O z+hXU{rHRcF;L~vY^WXm*bP1DfU|^MEPS@AtFMBNqCP_JFVyU#yghi0odix`e4D2oB zt^TAjR{@wdx<&xkJJt9mG7iEqG*6W?@N&`XG1q+XMm-CPcndM}kIq{3*Igq~UV0A5 zQl)@O=cI5i>s{E;o3uP2_Jwt73g;z~-$cGygzfdRU=@2c`AB!ap6~7r2&Mwv4hm8i zoVp_!FV@EA`$MF>;`fv<9xX-E4aP){7)>sE^d5Wu`m!p;kwU~;)tYg)vy(yqA4GGw z_VLCW6pq{_`|!Tj69@Y7dhek6DBtd3u3S`fM1srJ(4>|gT?XTFJpEMod;HWI^>(3~ za##<2Dk2$PtMcvDR;z=1V<44aKt=>!~aTVyH%`O@x?3G?n~N97rsUw~NHTffF1a4g5A zXi%^R(d2q59?F^X(f#3}XB*}M_ZSxNK88mN2bAmgiY{NA{7q#2`OyO{De6#(U5MbB z_Pz11B=-x!*Tl&ecJgs|i?-1)+$KO68*65xUr2ufLQQ2*qUJdX(I?lBDmfn}yiJig zIMJ)y(ZcG;@(<2r{VLGv0EuaDV`KATa9Ic9~mfip~{+cO7GpGUr~cY6D=wQskS3rq@RjeLG2@*TWFFr6To z3s)#D5kb0pzFG7nEjZUc9`DW;KK?vs?64QG=(%R)Vj9+7zLy`!#oShtYW+KQ%X>~O zBg9xh$@^SDi|x}TIQfo?P${!5aP(`B%Ie8hA>p`UamcxPLV|DQt-H+t&F7T8x_AE- zq5r|-7A_@_cgLMOo@}Umjrq613&dqN7&qQf;O-(3w(?M_ZX?7N4(6i;Fxp0-D9d zHj9klj*(DkX0?Incn+SNjPsL4KL2zrSN#xUqN_o$8^=qDqqJ?(IZ?RAWh!nC3LuF* z7ht|x$p{#M8T0Q^OuV}kKeD@6CZf@}Vx-qAZ>h=4pGtn`v)O4>3^?c zTVEJ(dG`CZ9eVGK-Y2P|hx7EgGsCOuESpxC5Mti=V9Y$-J)rF~ecTj6PwobM7=Qrg z{pB{^P)4116({yA$#vuR>oVoeD~q020w|?J5X00AXNe=iLgIv~qPum$XflUM=>hy2 z1IywS3WER-nfI!K`O*zz2mZKMwbwlA87%p^>$#A{zg`Qofh_~~tlbYkC80 zrfh3QI^@;IVwaGnc1f1zj+jH<1hw3he0cHDoV8N$+KqG|*cv z^8P7M0AFKq0F;#;d~{muO2p$rlbNXLGdpt}x2pPm)u54! zg{f+^f=LR=8n8?;R5f> z4G!UaW2t-HnuKa-n5lcTtMJrg^M_1X%LoB!d?b9WqQY|3$CcXhpZXpg<|Gp|aHQ?I z2jr~KGxc~8YC5yNazoMoZ}ZT_+dtiAo}S+61-jUQn%-fz)xjUF9Yy+&m+4_M+VTKJ zO076-hIa)6n3wjB#jpS2Fg+69sGMa4TzZ|AIdNGZ{KD1?u+GsE?j$M0Xqe+?_Vc5@ zvp_Ajd0NMScvZ;_Pe^685te=Y{~+1nsZB8X4uZcN*@tt?*bwk#lqAM5+qq}W5R<=t+F~(nongpIGShev%bRHji-Fsdl1zwT z*80BQ&L^dE04TeXIqA|(^OBm&uCxSzoBoy&-9JbpnC`z~VT|tQ^bme1&n#Pxs0-dX zagu|0Oh`4utRB=kw%k5#+2{q-x)bYY(7&7AvBO^W^U&u8uim1&oeQD!8%3H%4Agz< z2{cfZ8kb9IVp-0Qm~^>D7WnM94=xF*ySsn&DWlxN>2Qss{z2ix%|&1%gW_|#-{@ej z&!9M3N$5hfhiYYW(ZLS95~&tYlM?95m6Hkk-sZ)xk~ML4)Vg* z)&2vi+GCTxfQiY^ln~!{TY6tZnFbcq$Lh!aYZT&7x6-^5>&-pl(;xZ?IX;EA!7EZ& zkrX;WYfGe{z5828{a<;X2O_Gfcky~jW5owgT)rC*I3t1>bD94-73i0Ygb9j-0d{_s z8TG2C?R49*Rl&3hjnIOuj0+8Y{vT(OhAvB@8f&xUxcg(d?`laCV&sQdB|_rxBXf9; z!0>hC$41Y(EX^tFJIpA~Z4yR3Lc4I;e@oZjsHz#}EBX3%k>SnB^)zZ8dLUf&1Cs8F z5yy=ACL(d}lM!eiH*V+xZprAO5d8Bd>^*~V#q^-Vz3AZbDo0w1OQf(yQkIi{@v=zv z!GWf;NoiB{nfGOocUC$(=Z=kNO_57B1JI%9_TwP`c{#@U01q;MzHz&UZAMw}^Imxp zrxbVDLWf*p*ZO$zI@^4j2%#9&NKBk_sgt|ZQk+A5VoKPFTto@)^LnnA(yM#(&0dWE zwY${fUAecKxU3GNC%7gz`H>0P%-Q*yvlO;}rVPC4nqehVz(CU!4)mA^6D&RSfvMgMA^#nQv&8ujQ6dZk=z&*DR@uj+6TNqbN38c@j?^~|S(ZUEp| zEm*vl$0(GipmS37%*GeIGV}YLBGa!qTxQ)Hre*3`Mcm_4F<-4savY+@QY73RjE6(b zA>vazQ#o)X=CRwBX3s640Vviv~~n`I$7;W$k`^ z$V9gkQ)8)9dY{|0(dzM&F^eZatGe;UQB%q;l>)SI=E&`eV9R*espR zIbG@KM_QUd;t$_|1C5Pc?z3?^N#Y}XPM)l}WNBfPthT%*339vHe&oaZ@l?=GlDnjr zV=RH~xbG%xs6IEXhKUgjZraw&ap97lEOR#~fv+t&}dtG3>$JNcE*8D3JII@(T`j>MO@RyiG(Q?g}St72eo zd-6ghgYE5`=Rw>;n{w_cr}jizf5Y4j=j$ zdZAVxQVUtXZ|eDxYrGNZ1tQxIJKtaV+HU}EGJ;IR2uUg&a^k??jY1x)#h9+e^@Xif z6(2+Af2K74M@sk$SMoqZy%4)U4=H87g4SA~Y}*OZ>dSd+MaZBM@+#n%Hep9!qTqxT zAVjuX$*#G1;Z^uTJJ2`T%_{}sDmGP>?YWNe<&RI*eE%#taYoy#{hj5UDA zeVjr#!WCIkxGCT%9GOGLNy-AqWNCY?{AQZp*Vu?rT`rJcjku%YHqOZPW?DiXpYM z|H{#)Z!3ovlG*3YP5uu~dS28c%9hY|)*f2+65@+OJxhFXO0%Y%HAX#07U6)~h=smS zc1_z(&o4c0#}jF#x|kQ&;dit55S_{j3UUoYE=ij|l0{t|z2r584S;;R8I>xO%-32h zGTq{J+X~v(vi&7Z!SIs2Q*WhBdRfQq{mSj|4&vNlm2dOMLG@{^NLvcnuittE>TP{$ zZXkhj^HdF+*376G4JFG>;w*CbD3$Kb7Y|&wQT$-AWu*SV>r9VIZkN6HvuY&^Cou8a z$Uv`HzTxqNPl#JC!);KBiBo2=F|>Slk15Ofx(~kZNM4QaB?HlwdEZn%kh1?=NK+dJ zdTIr&F+C_#n3KJ&qklW^c$w<`X|kywEv@53TtYh<Zs4YPNrBXP z;ZBqP%z`?N>-N*edYPcdKgrCI4%(a*pZbO9v09>AqT+1&p$oApz?ss85VD-7{N%ma zlNLq$G7f)QQ`oeb5I#tj^j~s`aK)~Ho>C^YXj1>s%2sLudQU-shtU4e!^J!m$+x}2 z$SuSHR}G+*9vHMShj)+BE22%QJ1ek@$%;!+a1nLh=`U8jC*lM+v;FG-|tE1m|vm znFImS2t41k1*=T&d-(VvVkMip&4^XP)HH2tmu{x7TFjvc@8*6AKYs8n|Gqd7&-nf} zD1Uio)`in4cOMAbsK1)78c7d|Prr8}R#){(!vKHR`ca z>LkP}sljnea`BlFcVU0=b7d!5-=p8a`Wu02&jj92^_*3FP!|x%aQKARvE4}Oc!Yzv z7h4>G<7*SG_3`<8vxF-_D~dBk*4WVw*J<0s+t)Ld{im>zoJ-K<`fGALQM_=x2eOWb zc-Zjc)4E#L&c!HkLuZ+X)3@i1HTxN(6*_Y;8z|@6Q@SC%jZRJ+{yu9Mn&!~^e|6w( zckuR(4yp|FFfn|urC0)3i;niCQ@|6Pz`){?E;M6>odyAMz6HmFXP|&Sr;ZhL=j9`5KBIQ4 zxYJ8j5VtjtEo2m+=#Dsjf3?Pl{XWgjMwqd;7o#P4eK=Ae#cQkLYB zyn7Cz+Z|kMcZWDF?S-h6vkE(~P}u0W;*_T!xYgRjr<^vRlc=&v3bzqi?bb9+o8ewb zlAvgO5VA}4pBb5l0xI_>_jB?2Mh}Vmq68S?-0;)Ax*=Ns+XrSpv#>i(k5hPv?a)^U zqZ|E|*ybEy2f-O)=dpKCv;I+W*i-N7pM+wVYh}Z0#2oqp5)Qlx9VRM?oUT7I^-3Z{ z)N7L_0TVm{AJ6l&?j@JkNcs3;9NfY)Gdd*FZkIZGdBl2gN&T~54$T7@Ln|wKJ81*1 z1=huz^7w2VGA0^+ewtT?Sy=TjjvQGTg(ugWOBLnZ&ZiinAVkkX)!RZv3}x1+PEb4KLwdxV@lr* z$|g@=f90V~2w0&{8G2^tL%N^9&k{({1p!U&uEv{QRlZw8^*OPg>jA~}34azO8FEn46;$MZA(^zw~v^mvatDt}ePg z*@V^yF7o0g5Z3ooo>0k{2`t_hyYsgbe2+&D6T+4?^)C&0Q2H<%lDK9vpZa*_-hyfh zGXnWjhAJEBE)Zy8+VsbGA=^<+951$HeQ;PaUAhc@5(M5HGVXtk+56$9vv+foL=A{{ z^Gy{&q2+*%eM8v=&jyW40NriGd7Ej!T>}=dK&>zrdaoCJ-mbKqHuEf+J2$MN7WXDhr5E5M5Q5Itlx7T=l3} zCmUVY^E3{RaK7Kfet%iTAI(h1s9Kp zu zBm$7IgKmZIddcd{o`{Kax?6T5U_0qW5*z|qXZhH_ZhKJH=Q?3ch$rg4Mv&EOc|X!{ ze|_F-R3n5;1pF>rdo8-EzJ2oMcRxf%OftD`$ z&DAdae4*YCX!ZSXFMwM?cZbvbljC--Nk|i91U=Zp%*FSYyS6ibR%zk{j;Un`(M=SK_Z% z#GmI1sdcvBLh0;Did5|eYibzLUYNWoszJvp*{`khp!`VSt;H+UT6L%tgVn&FQ_CeN zkUi+DJh2q&ZU)2XLYT1Py8-Vp*FaB8Q9*LO*Z^Qp|K$){*D*mxQIUKHm?`b-c_3C| z-n%iv`fMXp6?2OZ^;c^7>!+U;S|(vm>ix-n-L`q^AVGp7BjOh%!Oe~QLbn&cydY>9 z^Y$*Phpp^za~_ zKFO3i>iIs&&Ms^rCSTYe3GA;+6eLe<)v@TwOB~Nxn{AYnv|XE)d#;*S6&2^Be7%i{ ztaf1aRDH}c)bP^mK>xx0(c@PKf%KP6p+9d^4G)TrD)6$q!tHL5eWD$v?)o`lVWw`I zcE6ag`iEeIJDy@|z*D{PWS7a{Ab5FBdwTRxe!{_7`rn6i7XF4J#hV<^eWxey6Ljgr z{Bg_AXg)d8VPeVX(4m%QPA|#OB1_q#w2~fw7TC};h)7|lh!Hdd)Imhk)l>tl%Qb?`siWTC z@i!ekLs5;eTSA=($qOSRbG{9Bavwj~!i(3{&T9FqPLfU0pcd{#%@38CCWM-|o7H?L z8+!Qrfbp+5Xp&mahW(WHNWoZi@~DPy3lK;1N272+dVSZb0N|JLw&{HWUFQ>!w))-GLVq@gM0|VVf|j<7B;_b#*j* ziL>Xitp`%BDPy`mpBz)J%-rd4H4THxk%`NBO-1 z5)T7N9*PSR2;*^4pN`ox&uP`^n?7`$0)2A)1Ygi2&o6K8pe)X*FA)OEiG=lJ2!%o7 z0-mRZFE3~azadw%f0a>xl>&EKTp;dg4%g6OaY~yBe~3fQJ$39XnQ1YBR-O0PeQk(~ zkvZum(y%bHhp)7?#OZzbss8LcCP~6u6|yA|Fq2i)Z#*^ZSGP}B)W&yYza^d_AgBz9 zWITQNwwqlb9R_1I%UEf7k(nu8E+xflHlMx>GkV9 zspQyeem8(>p-wW_2gK)ZXq^-HB*hEdIPs_LV6Ek58%AP5KMl#t>;v1jblI<2m}R^I zj4|@VF{P*9n@THTgADNSbr2{3{8Gz}43sm&r@5jn^=7a`aaT`YB7Za7J(vZGqOH%gubN}P z%Vr`WFsBx4P*mt&$+jCkla%JE%(9;hC~sYL>}arPPUwz_G}Q2)OvVgZsrbs2WV&D7 zRlYSjEm> zihlU}`)A}1E09=W!{{h-d8V*KsNU`;$mv97On7fS+N(7LjokZOOq)_@v+0kDFXS+7 zh6>*OzV*H9PmT_Ws#l?}I*oBvG_TWHQtAtTdF{otVr~U3tgAeB2FBXHxDh1lGNW%v z5<}rS)aCm7k!ouyQ$IC$x=kjlw2R7smGRyRYlCDb+V2R29#S6F5x+`WHwp>0|2*v?_u0Aj>?Po_XpY#gR7Q!0 z%=>qrm118C{|%$M;E@S~t^_?B{D3%6tdyXP(X`w=Bq#8gOWV=`ZLb2{MV}!N@j3jkmO@-%430`u53Ia()!><~lnn(f0boLo3BXrn;oo zx2)F$&eEDf5y=bag9_B58nFwEhDjuJ{RHWfTTPLhy=CJV4%TWyZUXz8--@<&jkGZM zUUq`n@yd@+fSkHEW~RXsaCOJm>Py~8$*=g=_vN1X-zm}vP^7c54)*5Eb406i-;%E; zSZCikIKb=J2DpPrCESLyhnX1Kn4EeG<_q##Z%-hGaGxmbVtX}szG^g67Pes?T3w+! z+&<+#qRthrZhyRNY2XXpUz4`bV%Oe;hnd>Gh6fq(+fe|bZ_9P4=6#)++$+2{(fZ{2 z3F)WZ1qq;pf4J)}yK+1U@OSP;)W=;DA6WD{64iCohg90q@^JzN^GpMcnD=(G$u=hA z^-xba|4v+gc49@s(8N7YIKqJ^_T3<_BprACZOhdtJU9wbdBRv-&WeE_n4OXvz3z|_ z5Y+Yg^q6vLVDo2EMv~Y_8VkouxcY<+OV;#5M;)>~k$>J#8Z7&v_)-2zK2f;W{CL|J zdjc_$fo^zS@(oA!Z4IC7+>!UN2frRqzLP%xcD5ymmUi8jC3aYEOvYhO=TVj!>-iJ> z{?+T9hl={-<_(`RaleeWuSxE$+ZR-u&!f9uIX4~eQ$1M<8LD7bvleZQTp~zMG5x2l zER5)CCG|72eLB=++|Qh2RI)jQ5)3qg;Hpm=+S4&LZE^x0$Av3S?McO&idyv&#(6T)#A!-9ehiomDCd58> z*tByUhVRAN`J+X0AO7|x;QSm!#UeR8I<$Ll&rm^p7wj?0;2iiqbUes{WDc$>x3mW@ zN|*Dcl%hP^g1YN`W|MLt_E13~+FIPk*pHu$%Zz-J5{-J(M0~zZ4an81RK^ZXc8IDyhi!Ckt7 zi)(H^zhTz~f2-P}+=?idDv|$9pXig!OC5uwYP@+s9(`@+zKfkpY17m?c~o@1zQ*2{ z7zy$QmUXNDm40O%V1R1g?TO2i&Zk^3H*_e|yYutbR`-dJQZJ1G)zOqdwOlx(B@cd- zhWo1OfNi10yIJ+Pgg+8+)etpU8OTae|Dq!HDd)sqz(f~{)`$(h_o3`K7>Xu6S;@|k z>rr*R&HeEmX8;M^zxgSa6o$q?ThjHlF4F`p+`7W($D?FI-Nxe0_4Cim1k|ai0p+O{ zPZ;u>NFMRZ1rJ`*zyA7#6PgVm{}NT9E$lL;EoqmM7GhViWaF*4nu_8kTgLC_(YpZkBO$LR#kMH-Jt4q zSvn9EnV5UZQ*iDbg;jPtG-7;pz8!xBq`sUp;-!5O8 zKz3v$zRfpJXB10ss;hVMqYJ2mS<2f-+65S>ZYyZ&l3XL3(dwiYYepA?ZHrq0n^YV= zMNjR~slQdfFJ`1zOF8aB;>%KKtMA)^~*kSFV>nLQh$o(KXb?U{B|c)$SKQSIDRRNzkRyDoL$&U@fN%&Bq914 zV2l^Dpd+~7@|;>{)D$HCc$={?%H-nH(Kr2i<#!2WN(E%6{ol~m#8A`rnZ&$>r}gHD z2Y7mU509>^3wiV=S7;l)kNc(~7|bQW4ks!$*}f0Mj!eZipmd9shK}L@{HxwxQvy%W=wfM>AY;c_{?_{%w8mM&PlGQmyb{n~qPuLkg+-%s^Fc#0+ym z%T~gCSTq@f%y_Kga1L?NMl2KJ_$ZEa+>`AA{>SmbKeCG`GW=`~uUvVRe&072KpksC zmAlKm{;sH_an;!ZA1Y$45g)|V3e(iyedgHju42BMVRkxc)8iPcCcnF z;a{nfD}gr+dNQq~J$6`(d%2rd@3|xDzYp=Bf`Ju6ETfF1cTuB`>z)3=g=f~D7Q^W$e@?_e&;K>r>kcYHIx}6eOxPO z{iG?GY&_N=9ls&lr5qdiZL5gzez4q+Vz*PXJxX?^N^-0;M~HKm!90#Yb}CI~)FECn zNDR;c8rmPFoQjZaKwo&mjK<-nV0~v==OT8BUpyopK6Lz!6TG&Q{!BesN1E4obadY@95Tv(sZ!!% zp`{SY>uk}%7OR@=3aBq76G_n1lQ1s*%xQnR|9#cub8LKM>)5|_C-#l%diMA81^-)^ z^1#-p@#xN9Ms^a^0&TE9u%-UmY3H<;UA2x|Rn%OY50ey9#h3qy7!xmeuBLQi)FLr~ z>^EXm@Uv@kYf6TOLT*>qIc4*}ZQ3vjeqdEWH*_Y)(t3SnCCv57YJ1+zv507#5;Np+ zV2;R_`b|q_8VE-fE9Hp)BC;}P*VTUQdUL$wD@vj^cV$DdY)>`((N~7n;n572SkNsr zGgE)x&zN*)Z&yuMmwqfyUxBl{ST~;9fc)RCh%!>Ni<*D}xV1p>r)fjo%Xk1k z>tLl82I~1lIf=@NwfWQZ9>KlWz7NbfUT9gHx0vn9lE#S1nHNKED^XF1h2?#3iHl%O?ZdAnRGe6q=;({(Fli@yqbx%MEUa2q2Q0$O zd=E!;ZSCz*N=)K3bP`F0=RNg&m#?W*2KHdV^rE{hM;;Ed?m38S+Q%oY>B|yLF0?WT zrWX$3*Zv@zH%17jfApv!TTOhfH`#X;Jo9nhEn?xitC$VGAA1Gc=P!E=V#^3)$4caX(=f;=*^V7YbM09Zlm`{#<@_4Hdxy9C7)Ca zf!#L5Zf=eg&Ry-&%k3t@;ye*{^)81N7xIc8zQWtykl|(&36Th2q=HX__r~j*CySc2 zd&JF7q(YD#EScl_pWof~(!UdLl1)){|AZhz^Z9IBQZ(mfZD{M6X25IUGi8JTpI@YH z_CLVj4`L5*0A4f!x*c)e{=M1Q0?qsveDS>Gv^2ZQsy+C%F^ls6{0bsoH;sL0ab?*v zLtG#M{p8+;kk{Uc|CYWhC;1`Cd~=`FJ1(mmrc8B#nf;K_8mC7|@3yy|2sxT*wa>!7 z)$89Z6Yx~Digovbkv}t3)fitp%>{`9!I0av*Ks<+*nTkEBazS!^E~M8_A`GPNoBz~ zouV%8D2I#Eb1etHojomDfqr?&R5(sv2`HNcDPq}~^BkO>mOR5Hz}w$U>*gC^{kdL?QH2mb=vfq8{%6||brc9&`@AWYz zq3oRMXT8@_cicxrtSv1hT<_eb|MW3`{=`zl0BEc}_GyO#WYF{`5IDdsO@uwLS?95Z zdq>{uEXU``jQ4nf2^&TrW&m8vfVCr=pTrBJTV%v^T8(jyC=ZohItb6 z3=SeOoQtx?8bn3!^6NGcNczeG55UKKJFC6`)?=3ST~aTM3z!Y|fYC9IouZoKuRFE4 za{P@wWvea*80IX!jN~h)ct}w{$l*;x+!iI>qJ}GJNv|or6xQLqXeMK*khgG;vE_k` zV>k#cr9VTaNlE&^2`uZ6qx= zq0q}bHlDtmM{Hl2@VB3j`}Vgssfs?L_TzLujK%2L?w=*RZs@DR>XCgcJADm&P774L zlZF!?b;=E#V>8iU6!70Anj5XuN{)R-G2OV^8~`8G$LYrcd|)f_Z;0XXKLjCPb+tEn ze*mtw?I1l^JQsgX@m%-7WY%L7@vA%~;`5G0q&JH#%WD!?_prq^lt)r3nK9@%>3Esv z&m>Hj2z#Cfwum}I2$ucAE9lO<LTi-zki8lMk)&8JwG|O zHud1`j5UFNFbCh2BF)G31GRr?PlRkk=#F45O$S^Q+jC)F2VJb)k z$DC#-zdnC@$>8IP`J`uAuNXr4mJw>{l4qim2|lExP&tIk|M#M^`H6CdVA4s4dE8Js zPx+VB@8c6?-EVGJi~|=3r)ZJIjs2gAmz%v?qtGIfn>#8aD;?3^k&{K zo#~&D0{g#QPxm41>z%j$`aZ3k8+h#6lgX97ujar{*YbM) zJBVqt1(n=@uzEHVSD#BhJt;;?+Mn%Jtsaw_Jj^`ErVZ-x5qHGKKG)1*;;S~p5iL5N z28r8YUyi(vH5m^sb`;>lzf+PsUY=@Cw2RWNRKEX#0>5qbbfESzQ6}&rqO*`TD?Zu9 zzngzd+>b?B_v7!T(*W7ifHRwBE@Yg}^TD^9<`u+?z&2L5Po4g+$)#Z8{+5-FMF*@b zx0x@ zl=&3!<7z@lF~Wd5ffz3@#gBE5fn;{vo|%V9OuXj%`rBZoc{ap(55FP0%>t z*uk`O&5~&3b%X7laz@8uOIkEc{`JT$$Wq~`wmq^kKg?Uk@A^HC87WC=sJ%dpc7<$2aVH~8>x z+9_*#ehJTZ(@%eUC$yC^r_F@vlg&sRlZP(3U6SiP^x3WWQz#}_Lgp}*Y9FVT#gO_+ zYxcoW@_d#$g1K4w_)CK|?mFka|m0$K>KwcOBf$JhiG)flU9BkOpF|t)^_@)b?r>> zFB`)KT^9KUS&Fubnf--e-6_A`iCb;&#d=$>#{aY$oKGpjE4OEa8#=$DQ2JSJaDU3q zdwk%p_Q^#0yFsNCI-VxO@umaLZzVk8;>@ncO*^M*O1zpD1VNDSd4&}M>B-q9Ln=Jy z$Lu^bRqp;^ND6?I$_D$FAADf*Hsc6dYe@!VM0Jn^xzJc!nnp3KkxY6v#O`f#om z)-E>ib;1}=nk+IP(Hj3uhOo@rrjWZn&J9qS_NeEp1&8gSkEvTvV~HoJmFOjYxeBCZ z?R@@le84L@GLx!fg;v)7Vl;dh(~-cJrq*DZ!5C*J-2TC)b+;UJ%ib!*wmmq%I%6jxU2+P> zSiwA>8*1}_n?AIphQiillGmY`&B!z@{|F;E=`uED4-Vd$k>J^y}gkr z*OG&X*K=2(G4JIC!Fz-x^FZoY98+E})#QEK`HgS4iQnxYx-Msq@zvISRZf4T>kv@b z0Tg1uzI#|b^mJPZ;J4rSA_=23c^Re@JAaLr~t^@1R9iap!vg;kj z70u?xFS!RlmRkV_r7u=}6S;TpI;6Imo(Y|_hxp%gv3*pp1<-WcrQgeBdF`4p(}1EX zAAZ^VFCeE4W8(VUuYXFQoZ1$PKCV_9&;oAMFZFrQV3ktz7EdA;h`k-jX{7{vs?+Se z>1`Q!xkLb-R_hxU7BcinTjHLMqq&dSETC5sZ`$)SGO8y^u}-V7T!-Xx3!;0P)0Lk2fTAVBDOquj);2)qbe!(c|ONtJGB?I+)kCWwyR=64i~0!O zMukL(iNAi7G!Z)R4kgPEIYWF#fnOf_W?|GhbmUn8R!;3;T@Q%I^(XC5*wdN`#?iY! zeQ2M-W*QYVG-J@z=f zU2{d@~(V|ZMxNgh$H=g}dykSyF_7PD2@myQhJ-kap>yqE;6J6Q}4^8SGuzIv;V_eD|So7}Pk z@p)M263Ut1PWB*Mt|yQM-nAyEIVR@GG32&Q{FQgv0PtSXgPFt*Kl4(VBP)sXt@4)5 zB_+ZJt9%@J9}n(>2N?#HnMoKBe8|jq$55w_I&)bfvx5yNj#Ob`GB88OSM^T&X;KZR zV%M()tI_qZq2(Y#iA$Qg|3lSTMa8wP+d8;g@Zg2JyF=j?+$FdLcMa|q+=Dv=cL?t8 z?(XhzE9;)U&pJ=_LMvvgHpiI%=>6+`jv1?*al&R*FCcDL#y=*wGv--v?wAudfKBTAySqxZ)uNy~&shT`$JAC#z~x~TLp14FzndM1q3 zj3NXQ zf$v$X4%fmK;I5lVn<}IXim9bxLHg|C5oh;vb2DVqDIhcyz*PzX_F~mY+s+#f418a@ ztIn&9_aPlrb>I(`m#ZrORTh{>QGW0WU?G`L$o6Vb#5p4fz176AmCsf~6X=(>2&-q# zrjB>{Z52A0mIX@P;m*|8?u(Yw`#$$7D~kJBa7<)&)pZpQ{wPL^E>Usc+T)B2Ew=oE zm}EdlJ>aG?>Wm}aErbFtDi0LkcE840+uF3(uTVa3talE**mV}qpS+Kr>z*gXjW5Tc5hR ze1TKWgdZ~GGh9ihQ1zI$E6#HU9O3>YX!L<1xxy4(@T<`UGlNO+ws?ai%_jOdl@jSa z`6a`qy$a*|{8%NoK(-7A?N}_WldepWjE82yB!Suo{U8%tCLzmGeo$a7NXeLf4I|n@ zzp8HBYmyNnK+ntjOaTKZ&Lq2e9xX#sT0yo)vTU5)$Jp5_ufH?p2io+ zqi3R0rkRvXHvHn))nJb`bF9T$@?wCraE$7<|0;m4=`8!0^1p+6zaY3qh<_;b=PVj` zcj0Q6Ow00bD#q`P zyuM=8uJ9*^qSD^^9D+3VvC1QYvisO*y&_fSW6LdUD-cd6#MgJb+<^M-gMyBwN~SuZ2w9m@~v~f@Z@MZ2s>zE z<;;4uEy_-@ge5{0z0B_NU+_kx5#%)k^5=KrcDaKn*org%lx64QbfK?j7q>Quvn>?? zm&uxhzYOF!pv*!$Q*%16&P%ND70?L(PV|vX;8wQHem<9Z*gPf@QWq@KX(K;>4EVmtKVl&Es_fk?r&kq4?9T;Hw4QlgPMP(GK&GM)ZJc+!VWnxP z(ZrjJs}@xIKGl%_jR1uR`uNoNpwoArDoKd8*^wKr1Rh5>$rEIwC*HBeuI3 z8MYqj;!t{ubUDZ2Xy!Rmlw`m1K%LJ!(zf$X8@g8$&1pzwn#jn{%N7D5 zc9cLZIZH~9D+-*Cgv)Wi5O+A_qrvZD)0Ti6%(jwSj&zpQ3tgnS>XCT^>BC=FuB38p z1@%hV*Q@tDT||x0-)2yJ@EKrwk_BJaszPFn)}HV;PJ=t{zGd2vL@flv`KmeVR&TQ` z6L+9=UqbdPEa*8GgtDkx;kF}->t0eZMuAm}Ezzpc)tB;|*9pV06e|~4Y(Gqfm0x#U zs8q@U9qQu^oH;V{C`5p2oe-&JlY&(9>FQtiDlKLoSJXz1pI^OIh^&82NB{ap!74Ng?-^;>2r6-sPY7NHi1Y5Ed9^9?7IjWeX5xfYaAbYP`v ze7cxHO3#xlXN7ul?i-aUdG8xcQ^CT%)or$F6$|W&N^eXbpUCDN(cmqzn!?nfv6ZE7imz*uXUYIyB*HLgr`eGIl&-(N`lQ#X0)s75|uJvp4Y-HNFB; z9C+El0M(zp@wBflNp9xPr6 zCIn1878j6UOwCtx@_V-j+n=Z5p?z#$=(9c8{%819C;fo-9$`gwX%`+Ba_y@BF;>&WBTqPNO? zIU(WeuyUc{g4X%L=yR9BWE`!NGFjUdyCS9MaVLi!c9-j3do-VJOU#uK3kZAld?0-u zF57ywt$O)k8Q9fyr^+ZJ(ZIx&%v}qQa2YNH~pGTj!(Fo`+%opN73WqW3mAe{UvU&AKNAz2e-|)5ytZBO|JhA@% z>9?MmR5`Kl8pZnC!TfwmTDJ#;{3J+|W=B(W&bF^SZzAS|H>5r#Ym+P^)*}t4&Nvqm zblBUIv6vg)^AA1xH8@SP&bmvCG~{S$W??oKqLJKrd7ZxH9$rhl2@*%dW$+M3$IQe& zkO@liUM@PuTx>e1=DldbFpVc*#9|q}1#brpQXn8WpF zKl@j=z<Y^`#N^O1I1G47Ix%IDKiSPX;fXgun@dxJ-f7s7FbJAD8!z-*TOKTo&fsGkAg4%Jk+I(+H zRvjq{*?TZ?MGq4gqE3C8|N1IZg@D30ftqlAofpFwL(dnx8NjYbtom2Zs{XAsWV+7EX~1X3IIg=Y-RAvWH1mGF;-{4P2Jq2-D8Jog$t+ z-H_Hjb@A6d_!1PfeDKJfy#U~k5WtiPN&gXzssTSFp*i(@omH5~cOi0QYVOF|K!xXLbs|__9Di3uFVrY8)F5!88W7-Fu@{2x zh3a>;=d51AriyU!^LT(@MoQqHGuz+m7XJjsh=x4ln7fqhGDSijM!)&lgn2&4&rZ9J zZbP{`qcrqeTM;&xavR7uCMXbLOQgw0nOf0ZV>7-=R1s2q@l9RR^tZO5RW(ZBL z|3b-c-g%;wm3xGk=QKsl06C*@N~qM^EmoCv?53i|3aE+1O#VSRD2P7WiTs}TArbTA z9C*NB1;n#Iu-JhNss`mRAaENiSZr#-oISfEn*%uRpF;*~T3U_~Ed2nt1I+9V=R5Kd zT=QJ02>w=}wNEb02Q<#mfa>qraN7mj;hd8yFShFEq%C&#pe3N?lOI5kpN6|y;AoDb zIS=FbBL0pJ_#FLQLC~3;7>9N3kKgQc1^&FV@sUI>{Mr38^%bfn?~LfP1&&rv3NG|7 zfh}Nf83x7y?w)~88Sj?>G-AXK)g_w2CqpgXVeRhXuTenA&!mX1XzZ)<`{OaAXA>MY z9Jm%ICLFeJ-59!<(Df*5t|4c(kp6&5Fd$TP-=w-NXOfWwfc8A-zVWw48Z%<5#PKum zcHhVT;uIZCdg28@rdr`6SDv>i*|v8F;YlEk5r(F9bszMN}W0jh@` z7ss)pSMBxS0Trjq?Fxh*ph8U~7_W!bYKrjQaClTHIQhqC-K>CPUK9Y7jPPtwdWTcg zo&o_C+eXZdC4dahcXI35nTvcywJ}W3ti<`>;O#$B>Yy}{SDU~pdt3aLXE|mVX6kRX z$N2@qT->?DEf!*qw)gp#OM5!rk-&=k3$LA#Br>&G2#L|qKWSNnim^XtVU>(aBU<)n zU*HeM%GC5sr3KVnr-?=UCnM@DL7R19@MovXe%^}+*c!ANJp7X%izeigSHxe+Fb=ha z3}>avula7hiQUxXGXe34Yf&r9TBiN@!n4|tTR{w{CkYz!4U>JRymr0C$k8{e8HmDF zkZ(T|K*lqg@I8slXWHxCgS}~Wq7L=Zw-7m_guUvuw(lWmnyQ7k3t7c8pdW~D4p2kE zBnGKL4dS3D3+YWMt*$dx?v=KTVegK})DmbJXxJ(kEO?+iacDC! zS+*WJ6%Rc>Zaq6?hFoGzk7v2~K80Zdt*Aq<5myLlLI3=R?CYAXAONt(5GU4u#*y{X zyVw1V_V zvkj(V6G{Ah!el~7c_;Ozm!1Mc1*-A96=U@WN1)ko&%=h)#@jq;Z#2S><;MgZ)jA^- zNV;E_KfXc{NrC!cHJ>*&K00q1L8XAVwUk-8f05belg=0H9UMI^l~fLAmNlwK{lAFY zKfCMy{}6%QRWVgOVi#k3TZy;W+Nmu$Mk)O2Vq(2M4*ddM;;lfL>8*-W{nm%JeA5~I zwWr1O3;t!P>8^gj-?OD30)n534AvJt)Cz&(&45YTs!q{MGK4DQW0>4HAviRVO+I*V z2oATwps9*@;?IzXx&_v&7~v^oF_{gIU%|wf04z;IG~aeLRWwk?p+D2|IoJGme8>y> zB#9P^wM+JMvS=S}UL3J@&Ixyq&B`(~uqX>}Tp+ZZ2^j`$N=j&PBnxZBW3r?pWun<@ zPZEL4h``ZPM+nvPl(}y;DZfI>N~U;!qHrqmLUr=?LY9KFyOxpm_B}( zPc$of8L%Amr5!uCZTgY(tGVymc)PjUA}Y4W7u1^y$mmJqI#_p~dZrYYR*qO#P=e<> zLGn<`=LToDn1~?r)A{VFHi3%gdyq{18PF^3*FIvIv*Ga52Cey-J&5Y-*VqG`IQ4mU zVJhZi<&)pb3XH!y_sg#x|BsJkL;%L+9ycY=oJ1bMND@nz z3AtgjRdjugFDYaN^3?8lZ$^)gw%C8tz4&cAX%-etW1Y~0gz`B~ zV3o>7R2&SCUzPV$M_?s)#WJDIa#xSU5$l$lwzQ2+EXWP@aL1|d^T6~g%GdE7qoF|q zaq7Frb7++Fljvu2r*CV+c?7XJMwsPO#&xUQwWd7Bj++?=Ge{%@UNxD#Htk`k_OnvI zY0~&Ca!s(fTDr9q6Vrocvcu4&=kkl+YNkK8i0LSowmsc@=8FrdO)Q;Nbr|BVF=RyY zF8%mpv35HQuZ%fHpB5mnSKe|j%Hm{VlY6YTdVT$54#~s`(hDqj?2GI)pa-7i!707$ zB;9f(=^o$=AP)4DfpvzQUF;pV(=xJ4(Q^cx1e80(xS$>EIrWf7a|7G{HI~jh+p;tC zf|^C-Bwt&`{JQ;qqD@EGl>gdSUU8yCZy-DK;wkzv@+qDpd$#|zsT<-l4CR-bkbz$S zwvnn`*QVIxA5doUwfE?5?>uJ&WK46@`|kxu1Pq@Wam^G@n2)u{!dmj81=^%psixrn+{S6eJcF@2u83_ zU*L|9#L%h+sz31s86TN50L9vdP|{?7B#G{4Yd_3X?Lpfer*^Et)l^@E5H8|{f9dIJ zyE^}__kT7|gI+Y)T^>A9Z86l+_j8QWcUxsdC4;+%-j2ugD+x>$2j`rrj|_{exDS>Q zhy+&F5vJ?$%3-z)jJ>g_Sks7P349R^u%Nq5^Kkkc^I~OzKGqNOdq=+dGa!BmR&*uS zTT2qd6NuLDoLRdB_Vs0>KN}}%$5&9qEOwBw8z%)xZe5A1j%6kzJD=ky=tVjF|M#TI(<*JXiuu)5i zO7U3)Rf+_xvXG<~rXR{;RigX{bs+T72Dm1aRf4&fP~^?3Hq!k|#?`~CJ@ov9!i|vL z$5Q?g@d4CCOz|umXp3Us($L?( zq{}{Y1LX!fj+x*GB%;YT4&l$-(17qDnHWMh_*;>_^RS6;S!AgIE#>WLC=YYOkK#!h z65Jtr`l*Fty3X_V7*OE`z3G_TLQo7V+qK?f;NsBuST2-wU3xsgp)%2WwJvHI%0gdS zUY>&<8;wj!E(1`ryk=!c&||Kj?cvLb5HKAh_`;XL=Pb>=W`7d3Wz?E^hu4_mGlN&v z8qUI+aZpe$GdIzNijM|p8|>?T^c+7;t<;=~@&Czp;5B?L+1~jFOsEyYqe%u_oY)06>bXz!Zjg=C4RRiDe z*kP(fEc43|*lC?Zm7W(y{OLPaY~;&@M&kCLH}X}x9xSEURA!V-h>S&M(q?jv{`~&^ zr#z-pFLq#)cv;F&*O6xqbAo%3itlCHsxX(BfBEPdmHc82MF8#~wl7UjSB*sL3Q_LiP-XpiH8?6=Zddtm z?lgDdc+aY@MUFhoALAueyl!r!c)%_;7`8zV=$cz7UNAROW+j-xn-{C2+bLS?Y$T|;P_-^!N>JMoV@`JK{rrG;mlb2% zyXU&%OL@+{B=ho7Ii5J0X6YF563+!``1ty>Nn9e2!j0-}{lYt7(+_gkhclJERvHIJ z&=43mgd`s6vO0X_{-4$RcLHKC{wK|GxwwJr2fTXfru#g&DSyV zi(vJ`IF`cO|3Gfl(VqlHSA8F}p3fVJNPxzFV1Tqs^=5)^o7LIVGZXFjd!T8?x6R@O z1)l{oQx{Vxi2pxb0IWSur%peLrTv6@56i1dmZ&{f53s_`WlPyOZ)(N*0{)XL7boRXEA?)Da;GwN?M0)LBoUex+}S+>fZifrGm*bCEw7Aj+U^Pi-Atv3 zq!-P;>fgDj)9yZih;4Hmvkr?26A#LL0VW>Tp@gks6tWR*mXYS2{fFrvWDl;pEiwC+^x1U+tKQkpV8oMmaiNu*%q zuUDvMvMOHsbbG2QusT~>QRO4It}weKwsfnajvVF;7+&{g(@{Cf5M-xrivYC{Rot?= zHEFha?%K>Kq=z#>^U_>#FIKfY%+k&nYEmEGJ(Q^ulR!(H?UMewpr9uVroW`C7+7-04`i{C1xo{`em~WfATf@kI0Tp$t(1O? z(ZZD+4oTc-)BzePUML;~-9Ol6hwkbqeI^OxTs+jgC*mvTu)w+YAm>kmr+E6q2`p~3 z?k9vEIduuk^Y$)%<6(gPVC=XS=PCg){)BB8$Y^uJ-oIZk-tY?@#v)7Y_-%N}$>RoT zAs`_9jn5Qq3w2rL&vr4maMmN%&d}`Q2v}YQ;DXF zz1;+2Vy;^DRl>=v9}+p&nYpd}2~p6@&meEW{fhHP(=^ZOc9WL6SCmYUj2h38%+y>e zTdp0{S*m~Xw;`N$Ws$6X&Yd{2O0pBLHYz`=m+aM(`en)g{4Q{|tkb+*L&YDgCr&7u z>GxWC=mB=8Gb=!IN!6UzQB^Uv9;TkyaPJAqh#uZ9ITC7bTUP@n8RuV0^S~22wYXx8 znMqMEyt{QG-r{^>6mDHZi<9D=-o)$U4dKc!NOMT#J317yCn}cteXrNhQ%2Ja7sP3t z2XuEbrP~Qi^)I)+2VE}v3$8w*r9u#-wkW%*sBmJr_yy=SrK;WNg8Gj9;Y%49=L(03kavbZR(1~s!+`RS*XBn8 zNgW}Z1uk!!&mz{ucT+=%nnsi%v6P_*k&y{`zKjQ< z*v-%y-7LbQ{5Uukg&vEFlG|$e+#Jt{_#yNd#7vmqo;SnPe`g4rGU#Kc`yX-j2k@7h zn7|ea3l-sFCX8qCjS#joK87z$3)1w@QM?fjGfS+yjL#o9lV4q-NIpBS zu#mzM^(5BFnNm*Go#HA*nw#kM>4mC&V-HZh8QG0mO0=Oa|25{D4!6^|*^Um_(<^Syv;2N$;kgisN#+Cw@)tV_EvLL-aldwLmHA-t z+FH`YU)XE@-ou3Kmca;ds=7P5{WMX(Evz(;?5z0Q4ns>EG2=&uupz(@J7I>|Xxx5f zSEJ6!64-J5@e(D&q$aK{C6i8#HfO>{2-rX4y~mN#uob*J9f3}DzDuM9`ce*+lXf_V z4{Ett5UPG`VdNiBXJiFG2A~U;yXDApM=lr|EoX`G=)U!o@cv z5%_=Z*~w~4HE@#m+sD!@Q07LzUoT7q`MwAd3!!ZI78yZWaO>bmmrYa;?V(eH$eb*f zwNgju)GM7?(w!!??VlvP6MloG{1I4sEE|)K(g6O-FyHruY*bK41~^NP6DGRISt5^PP5S@9tpK{Ag#OwS&=P+)3iCj<{At{;g2p82>YBs`JJ}%v8Bx zrg-We8fAJ?D!083iNQyrkq#282h-O!O06ftHTa_c>KLE}YqtgX)H5^mAkXyvyaH4F zCPe9ceu4ZQ1s+rzhh~JYo)G9gnVS41vh#FUmZk-eEv<$Tn3vy_bb$EJY6m&124CSB zv(({4r^Z%XreTPDO`i=yazvj#E|CejZe=y_kRa^#q2xmQF>h9+0{BAM3PPcw2y|pm**tMqQ(u(9xsq7Rxcq z0_uuAam6w|HRlFXATRMC6-#aRv}46fd=h%9hP(q>}(I^*-y!r}tWOB>XKD$MZvY)Qr?06;@=u-^IgY$d4%1bFVG~bJvzRgWfHjt<4 zJMUgCX<=@pKz8Q!@%KO9{OIH6hK72+ipd98T52=3sqN`_L$p7Qdv4{Aw$Kfeo*axx z;LR)V2<7udZX3*2{OlM0#qYZ=8_GycxXx?dj7l-u=Cp zwc|r<;~mOuk1o4zh8t+`Bgg@3y45Vov@r(fCV4NGvCltr-5 zt>eaML67Ua3!pv-K_ZS*Yi#JTs`lpSb-f1?T|Qe5daS_t8vkiKB1coxkO%bk0IKo- z63E zJBk}=EfGI-At2BQ@a2!ELgk?7ki+3`1<^rcT(F zPKfS8H@$@IR0HZn0)m=+fUIUJQw1W8Y()(eIr7M3?M-&kX=0zaRR83Fr6!VB!~R*q z=@33$uZ@WgJ)E(5pgJ^$4%8c@l)T8J2W5TzfjOC_ItApk3MdLF`ovWfIo!Y=TgrYQro2e!-gWC&_9V{(EoPTenkw1WcvQP5- zWgMaeQ|G;FLee-Jl;2v%g3Sn0&l}EomPK_XCZ>k;Gf07PDO9A<;X@Bsu!Q8ewnjEI!e=BU9a z-4?OJKF+xIaiN1AKvXtpPf}ADWy0`G+1+VbuOi*k;WGe4{{l`jIJo@oL3K1IU`gs( zFO8V450DxkzeBwk$AwQBHjsV5#F1=8gC#4@%ZOOx0)Kt%*}a4;8Smm>kCEHTEe89W z7Ei__dt$9bI}Pc5OcPC%H1pDEA*5k6l$0oSAlugF57DjS8-bYxC@k3BE3~@P2W9tG zUv`))#$E&aRI{s!CJ%psSSXf-;0!f@1Sr#m?1RD{!#tUF?U{lW#1yvxNjNp+-1i7d z^DsnW_~cJ-0%=@RvcX8Mz2D40)S3S*X17lzlb0j9H(^3jT3Gp3dZ1bydY5WdZ84S0 zv_YIH-|Es*$(_M2iv5|_U6Mj*-wHocTcn@qXNrPNSh zn>kq0A^Y-FzK}=6n<%$?QME1xImPaSmUnDHMD&2UPBw@U;2^;4k1GA8U)5PZ9#2@Mo6gE~tcKJ&=h1%Z zN+~6{aB)HD+xf&D_^!b{_2=!m3}sD}vHiILtpa$D7fr)V00?#QCaeqMm6 zloI=K=S}k^=gs)&C0Dyf-k!7;S_{JIG+Tbg7TZS>L2ZoHnkd!UHfqmJ$|UNuABgtL;?M#?{AR^i~w`5Mv5Qjef^;P0M zmFnwZBte0{9P17`7?s0jYK_Z0qQ|!S#Pk(hcj7l77g5vvN;cOuM!it4Jfb>$$}j?TgvWAOcAgsV)I1GEH+A6P&l4; z1%(TABjHh^X>Jj`sCT+|)e-*O$>u4@C@m|TnhmFBlR_m$K)62InGnDZ%yBhI%C~;; z-iHlMiUQtvMp{>-#vPW*JBn1yfCybh4DXK%+Edbc9~we7e#+M+j?B+dNRNRMfg~2g zHie_Zh%2Gx+)J8HbBVN(<{nhwq5}tFv(c?Ko!rBZHbNR@?Ib=iwFh5}@0{wX(itMp z4Kdw1la+Wq3s>t`4Wm{T%qZ|rrgZ5L!hvW8Yib@Io^l8N+||6h2>ZJ!$CD>)nfP4+ zpoSzZ$DR6M!tHhx!5m#r4{D6jk{*Y=u)`6!jDJgdtU+pmR!S(Ib<~l86M!G+}N)j2w=?}Z&;qHI4FhM_l%(?1&(;q{@ zv`4;=pA{fixzFBF4&OdV%;)KyYWv`-4 zwrrT4W4x8q@L8rBwU*J=f4+tc({I%v zmayr-B>6lj#_P=0RQ9a`=qoF=x5Tyh%>cGB)z87J4*Mn5(;VWrJrA__m0_{4*kH4-YuMd1~E55~sW#$9Ev!as&67LBgZi2e8 znCgxwF0Uh7(*0e&0jiTiTk4YLVc{uR*pR_}Gc;UA^|^H*zNrWOP4o6nXM)hT)0^3B zlZ{`}#$tHxGPNZa#qCpv$Fxa!Q}(W>%T`WsE39S{aQ-KIY$n5G0mUBMyB5!1-Kwg1 z?WJAP3)R&oX%sosz(`ys-bKZkQ7yjN9g;nN!FEqi)261>CdtzJ0J53KNAt+;^)d6^ z4$yd{TGuO7t0e`#?!^xuzPH+v0&f=S0OU5Sbz>M}X#IV}kwk*-;T`vTFWUfeP;#MD z2mboqK>xP&d+z#`_{+1`!CFjrtU?7<(aOA6*!!6vAqua7=T;&o zRO-zXRkv?b?BWzs32l))+QX=cWm3xh2&G&kb?3sU8qg}Q{LLsBmEF9+ZWw03D+HMc z{UNEspQ)448Xw-!Rb2mv&(%BYt-%w)zEz*qfX+2{J7+V--ps#m!$@k#-xR^5_=!3* z)8Xv))*C8{;)%wU`cE%a;kScljSclPGF9!EHcWDE4=pG6+V~6fTcd^&`aR{%M?5q% zkrVR7s5t#1`ezAa9g+RvGmfSUi@V`uCsxn<`$>aCTo1Qak_%|=6>5w{3n%qi#(U|U zW>{Mn*4_}_47bwhjc30Y`o^I+7#YI`E968TUJ6dsD9iA68uZq%ir5qHgpCf>-Ok3G zgU6N6STfA@TI~E%CfE}Qjc$$W3D;wYz~&#sjFRcR^_^tKsr3fSM^2I^2RiQlasETj zTUn`}1rGGCm*Rwo@2e6?r)X{dG3@{A{crjMi5(50ZNA=1SQ2ST7=D^n%O^o689N3xeC8i@tOav9hiar!)0Ij_1?D<=)&d6u*=M*2!Q}y zThh$EOaHs5D48Nz0Z@fz^?44a(W9lM70z2(>>;ZtYv+HBGdj{(R{98@dsj_lu#;T3 zu;3t>>p{?n91at2v}m*RjWKL9YLhW-!yrLk#`SmPgOcDun8+t%3toFttd&HWzFR-! zbSaamB{pG!TbOv(g^XO=eoy|Q0}DDzXIUdY0Ig3wC`8@`G|W_EPVIQz)KwIBIQ}hR zxRLZHkL}zEsouD8I`CdeniWP}MOxyb!89eT-0l^t?6KzT;QrTySLQA~CQqaLcy{(h z%NYfa`!S-F(Tw9=)Ga6qw&H@&Tj_bOK?=xP-tatyC*pVEv%cieMg{XGJRTex%1ufN zm3R#f7Ee9Oe47lQ~_JNk$Dpzhx|FS9HQ}uiYDF3NN{eDlxC415KLgJS} zw+B**Zo6fw@SrPM=%=)cKv7dPfwku?M~`=?(<<~vkGmV<^)ePJ*ri%43_uxIvu&4R zckMv(7WxBZHT%04TZPL{tyu>_M#cqVnEQv2Toh7E8#8rvjIHIGOLMaA29v!T7U9^U z;X4p(@sEy#Mq)NMJYv4~l8s}|$zDI%@*t3lSE>6wqC>z;h-%>=^7|GV-xMQ0okm#l z3vbpZ)<6=fcczXQfh7Z=*UFi{yobEc355$5de>Z;Im zqR}IZIR)Jw-VqF@$%DDTgfns8{?gNv#_DP6wqZjlF|#m>&^~P^rr!we`AV{!r!MT& zJuZLaO4_bh$l+8)BzbB@#SG24(}e0EQX}E?pC3;7QRVMGvqPlH)U(?JGz81DwowL_-PI|Vih(7O=_v$-u>2K0 zj7qG?uD8e8tw1p20iJ$gnR*2id$Y+4LqMphG(OK|7_3m~p?t}#qst4(<8mHj{-GTE{9f!|>gx+H_mMfoc6MU=7Nfc**<+h0(PCl2 zyg_?994+3WOSYEZk^^xl>a2lqn|hPZZ8l1pYL0{N9oDnOQwwqP6J`U384J?YnVHEsvgzsX_n zK5v~vQEH*_6C$?pPS;Xkd6;?g(Q0u6^S*fKv^Y$eZ4kgZF{TBJ`D{c97RaZ5! zv7Z!M>CP8rpFeQjHISW})^<@@FKft7!r}qbV#6@+N#RkMJayuPw-%Ip;n>r4&No=T z=bF5nS(tk+qQx*+{?6cgee2y&aT69N^kc4q)Q_dH>tD0+CsI^SaNKS?r|IwH{Y1I& z?B2`LQACZ?PZYGAr|an!OVjd?yw8i&G&)%lVlp>Wz8#n1i~z*yeEYw&uj@^{o z7)q8knOlfd^|kQ~MX?tYLO9ATqiAf~{4H?O-#2~HYlUhwMXGM$y+$;>F#j_MFq7IH z&n~5m*!gXdPO9Uys|^=R7x_!nl*9$bJ(O%U##yre7-U*B`;O?)@~#tF9cP(I*zprU>uXs*tNw9D^xw!V_gsH^KeW=iFY^5@Sm^@8bL!i3zJA_x zoZ#KvAa=9AzJRSQ=&Z1y+igBd+mYVrmE#g*s-{pB?BQxA&;s*(e=k44!OH`pm+N5j z*op)IzQG%7Qrdz@z_|ePu-N(XSbN3RP~gFA(r66JnP!|h6qn7ZMQ)w)zSZoA25Qjk z@c#Mh4|ks|Ze;}ma9&Vz&@E2;{d{X~rLb^5?Jp2T6gR0~PO81K44hTX-f}8#Dh>_r zbMIu*;N_2e2xD|PXKwNLxb|<3ZW)-LR67QDiNb>8eD$^r&+Xm5AQIaLprdVQ%z4p_51t8h^=L*Ece#~l%H!p z$_2$|=c`M##pjRIIlQ4bEjDWn@ufxT;)#BKF&-$~4!4kZP7mUHGw5F+?G?h&{GJV2Ej+Cz?20Sv?~?g?Z~0dpIh2+I1OxMF`7J=6Krp${(4WUc5)FhVJ0XR#9(l^oQ!Kafmb)IMo&qT zZ-41)!}j_HG`llN$u7VW zzogyyMx1doh^om3ii7Z`7-_@^w=D~*z!rI=L^#;oKcGzNZOx8ZxtHY_t#Rd*Dgi_I zQ3cx7b<*QA_dU7H%wZO4Xy#^$`d?d1_9PqxW3%KzkTc9%Hqk4Ll!Okb+6*lkEj6T6E4u!p~&ve0N9FA<=aT zSTL{TPzlRk$9+(~8wI=1v9MJFSTlZ$HS|TXT$LY>MkG@Y>fsFLygBo-LRh+`DY>Hg zO3MQjt&Sq&)6epKa(=#F=(GmK4hTS3h=v~^-Q_udnr7+FwcltHr%9fAXJ7$*;M|89 zgMcNC&8ULgI(85FU%2L<=cXh+cvqTr>Eku9<#EHx{9a#y6Hvde|2eoI=4jV)#Ld|J~+v=u0L?e~awcz(_tIp<_+t7paMvzss>vEyZBKS!hQ}F_cLG$ButnL?_;|unhSBh|_76b$pGSB< zF8I4c?XsYw{1534(m^38|-~RDi z1z6X`Rq*HQX+)-Jh|G1yoXbEZKZWS#>MDNPFPtuoM5+2ZT-6y_T^>;3m=hg}F;Lol zb;+Wm(4?h9)tN{hIKLG;F>PHq`D@~Zz2sr#c#$%JZYBtVfttzk*?NZpwDRafw@a-t zk^cnak%)r`?iZKG<0ITO5-kOV-d}4af*|Aq)TzxP)ALRvEE4ps8UyFZW}{k$Q2cut z4XGnEWllC84o>JWXBym_#rzv?{$o{6SPX{p!EGo66mTj@yg3Zt>gZx`X4jWGXXryi zWN(A9Fzln1DsIm4zh?YjK`|c@yuROcW3c<-GL8q_qPwi^rHit#ych)-X*}pjl(qt{ zI9WpfrtPM_FnCny!O{@XN&cm zUKupJp7-Ynt_3bGa&8+I1C0BgW06ONbf7pFaXAWWX6TlId`0tHjKE8VP~A1a4zfUk zY3}4kX~KNpFUxMcJhhMkgs=5GMMA{tZIQ{rhxTUb&dVPe6I#FYJ>$wB^w(~VS4s@N zYaCh*v_#Ig>&BRCtKv>=ff9(yB_C6rRh+yHQrq0hpyokB_|tQNmEYW6rwz3}J6Qe! zs?<+U&CJ{=8lEF}-C%y2MA0p#3!}f9>*G9^!s5*iXYTGj^xu|l=503=xV^J`sn#pd zQtyxt2FqDFRh1Rm8_o3E-ZmXbV}6VsB*k68HvZr7`44QbztH)qbvIFNy~whwtI|m{ zve8O|G7dwp#mi)KXZoe<^V3&b7D}_PE@BiWa$VVr*ejvacv?i?q3>VVb7o z@kI%WTEvE@rio_d`Slq#*wx3QQ*>v~S;}ddb-W=l@u^2FyZ3T7Nr1`2i8 zWR$eD=R9-r-K}+zJTM0==)=-fYoyH8Qb1}Ia^4ls>`hHQOw{+)a|k9wpss z-n?5V2V_3dpTuheFBk9o7(nTyg;h5Kxh!iUlm5SB#Mghuh)f@4VErIcS`~Kz&CGdo zF!YSQSRaN=8%0uLTJD5MRkcH(lCtKTK~LfLzzs=E;FT<1qz@sc0#&~MHEM|lkyWa> zHoKkDNso_$50P^iZw*6-Q}&+5t(P!TnHQSsg3b79!@kU|@5deGH*aV}4|TeikEK?W zNAb#&$q$$ygNyG&XYHk@!!${Bn(Z16RdECZaDcjFUXSHe-ONGuBY@%}jfLk44+ zTAg^6i{3<4$F-7sAEe|;stGjo=tkSR9=Zys`R^GSp9hc`jo8%k13I98B{`o~mPOR; z7vuhh@L0?C>OPRl;rEZJ>Z#BL^zWw3g4t6Str4>ObBBm5F1j0U(RdXPehu{&?#D~m zd$|ZMeXr-%dh=ACa>3AJg}GueQ(ga$u=kE<^ZowEYm{m!rK+gWYN@@cO;y!y&Dv_j zrbTQ)wM7-J9lQ33Q6px8+AH>m6{^IFEl7-SUhns7y+6M{e*fH$+ju1ReO>1|<9VLv z91hi$Dh+{}O#5G1AJ2j&ckw0 zE#xvDibd#Z5GFn}pY$!%;QXGsI~O|6j-ZCspRF| z(MZo#(=7_=?8%-8fC0aj12T)J!N0yV-M{Hc{4)R2T zzYo#P$7)@A7wXpJuXSx!d_IxMB|4v}<4?q%bZ16T?!)&QOLaH008_5BNm|%;sbfOP zW9^oL@53|s`b){!)WykY;aNOl)-Orcj;yqL7Juk7U;|JfH2ak-d~clkwFp*mL=sIa z(i<_e^SSwxl~Fz*@A=J(F$(0#RB2vO(yr=kC+4`g_N*~Tuvp>Gt!EEWT)3g{!cRx& z=`t2<$=?L8Q<1p9s5?q1mD@3E+^N~gE$qrWUMHfPp5a!#I8WF*-+JSBp^UdZ&SDkf zdeObo;mR`VD$+-2Xu{nk*0n#6AxiG%*e|y3ijxh+qd!5u=COyT=d_h z-WH0stY&;H;rg1_pGd^8*yD?TNrf|4Z^QhI4L^4d3wuUJ;?aYZq2N=+`epyE{DQ)b zYuSjXPm!y&mSc#ay1?@$j}q7I$7xtu|H}wR&EVZ{C~2-C{ssA&#)2Ld;KAX7sEg`s#B|(!dI!HHO7Ethy%|qU!(mz?ak!Q zz*5v}yoF#uID!M4eNY#>8)77y(kqS50+C=Q|TB4rL=v6pXRW3aw@V*pg?i`_>pDt#smZRW_SgFyo_2NRv zY+u+tqxsU2_@}mCeNH-y$Y+)e+>v~G3q36)KcM+6 zZ{5$AF#Bwl=g@Q^V{jZjrUOF>v0z0vJwp2XvJjEdC|>#FVr9sVYpj5R*TkKupr4Jz zoHu>k@oF9YM&=;3pKb3S05bh@(5<~9o=!pOL|+@%IvFw+0MQbFkOq4 zL3kOXFQLk>zF97uBbZf&lvhm;We1QJq7E~PULiZrFb}cN!@!p$h1JG4Q$xFIgFhwl zbI!U;2b7S6w)9d)tn&GsZ7t-X?fnsjL42ThDEIUzlh_Fyi=;WjJH+uOIa@r6vsb8y znj z>hE07p_x0kS}Xd4Hv<&Zr&+TP&h{_DpP}3vUMFod!QgO1FPf`O4uJLVi*s$%n5o+H zLsq)Vif(msOko};O;DiY4qb087Oo#-%YzBxm7jh%=w1=q)VS21>#@id+uvbzhRS`% zu=+9mc7$icBd$&L{6F{bkBxkxrYM7OLjRzbl}egRwJSnUO;*p%%y?51 zTa7IO{bJYoaCQyzW?RG0U0^A>(ei{Dy0O5A<1$)TAQCudp;B|P;X}Ed=3bCA8r0r6 zaf_;KhGNU81@t)Q*2Kc|!=Ig@A>vLYo{a#9iJQ$PS+A*nv@N~yJNd!2l67%(=M?Lz zpv}ED(QrY>HLF<&-FeO6IWy$-sqXuonKuX6CeZ!kbKD#BgQ|ruiVc3tF5OEJ41X#w zMOF$3K7OTYHP+<7A(nPWz4;2cJ5!rQWL<0N2r)V+xOA)0HEIiYaYt$=+_cbLVTMUA ztu!m(vu?|W)zU9#gPFZ7I=Am_Y30U@0Qc#;ku8+wBnyp>${n0L0gG=&P#8oGi;ONs zh^IjO)?Ss};H_%evuQm;E@3&Hw@HKnf1B?wgFy#|>lh8!+wSQ713%ey4c1n=9sIz8 zg{7M1jBzXp|5zv_G_~i|FAT82gH5n?GXA-w{2x)fSL(zp&&2T5+cH0BPF*h#Di+LUni z@}j!-#lTGjtS`_efe1$lA`VkTRkaUKmc2})MWsB}53|ez zN0{^MjP-om78XFK8PL@T@(Cnfq`XPHrXO&sS}r`O80fMTVOqr3m>_pyhxT2pj1ZsXrK$e%~= zTntMyc#?fsKV6C+YGy~BAgCXZy&3g{YYa$fHQHHO_28n9%S?m$nC{6eIf$-_w3q{F zmfcY}j(3^CNr_2$35gz_`lb^%1Tlo!WMmd&p40l(8~IJ{ZW)$g(;yGD+%5Ebbc$8| z8@kp)SHa@BUsT#9SrrHV! zs00cS#I`IJL6IhtI&SVWG_TEBDTxT-6~ zckXxBp*V2hspGhL?DAX;HuUsDXZ!OtsI2qsYJFK80~j)j$icWXZ8Q}tP306A6+;sokW*=*YQTl0H zH|kPfosxwERCmUn52H1oUxhjDrEWc$S(BAwZ2IxsGWarJfpf(!|K0uC-0gBtpJ!fP zfxq_(S3V|NCfW@aEay3nE4T`jj8E$KvVqQg&=7n4q}J+#B^ATB8=Cl*wX1QLXREH9 zccB-6ho@tkV^<)8 z;83^JG^R4o>zHev8(!gO3*;yps>OH+yciBg+Af&Y0rGo!p3H};4l&lmPlr(}xLtkX z2eaTEcJ6rx%16^0S-x`|a3L>tlT1A2l z&9R*n-7&`6svI+(0WFjaIukZX+Rk~&T(@(C%vKxR70&WlI+a;TSr^+Jcv!-Mu2GIl zA`{&JHBbypAl@!7f9uA!wsCM?sU6&~TAPCci~^@-%Oy$l&PV&|vn$r6fe%4)L$wIl zpcrO&ZOgXkz!6F(S6EgijQOHr7L&%VYfsZi26oI&zaw`%=~7C_!WAqVRybU#^zNyN zf>NI$0#I*egsp?g*U@{#W5uILrSg{G;f>p6uFM9Udc^?8cL36PKZLg*de*H{nhhGA z4o2#4bxqeRee=FbH`m>-c~5XyrmaNRUrnY1wqx!tzSAHB!P1+*Krzf1=1O8{njj4$ zxHAXubVT{~wPaMv4$K_ARXL+!?fMcGJ7sYMXgD^a{$x4n{RkseEP1}Q9?%R^M#U}} z4+s}kl2K^fc`Vb^@soA^SR0aqQW^PA;9ZqHJw_N4guXoLAGMyTJ;Xno{n8kK`@N%Z zQ0Qsqadc3Ol&kSPmV_TD2h{yWb1r%hd1HUMSV6cL7dIk&(Dt%D}sj_x0{-4kIt3(%o6tU^!3#LzQU0);JfD13R^Y7O_3DMy7# zNUA@gWU8We%`ZQ5dGi!z(LQ-aK(0GDm*V}H=cB%w*7n#73G1)?8Mv(jLDM>Y!tBTW zipyW|o`5V+#he4xqH9*#+PTioS%bLHH&$l3@R3Ehx939Xj--7+Z^b~X9~=Bs&U3P9 zn8N>#6Ln91D0Fi$WN|$y#8{i0@iLU6)eEZZw3xF{bDmq9kGzifa&d3nesQEbD`A4o z_c(pfB4EXC^bKMR1J%h`;-=Fqmy}a=k+5J&DI&@lvBPX6dYH#F*HN=c0ct-YF;rN4 zi=oh?lnS9X=!`&)QYbk!TeL&K*m0+6c?*yK@z0?th{>A_1*h3`bFak;!NB!R!$mJ+ z<46b7_;ZWVos2s8iphxLWU|jrHG|qZoWp7Hx%p#%q3&nAqY}w$IjC1mUsL;hir@@6 zQDn=`O{JcSN@!UbP5c&gKEQSej>PI=nsPDmCgVo4hq>e(CNPhNBlx`gjGSjN---|Z zXv*wOt?>*i&S(D;|I5g3;`55G4bH&PG~=eOL(7zx{JA{0Aa9g|Co~x?550m3deAVp zz<@6r)YfW(RbW1BIZX4UNlOltckDShpepa3%)LYIPVsgcG2xlQY`j@lD_hShUa#(I zzY_l|f+AQ3T8i=J@4AQ3RqBE~bTJX*wA{n{&=sXqYmy}r0Hnc!Q(`*{{- zqhu!boQ&uyY2EJ@?3%STP(He{O*ukb1N^3xJho_n?!1O>Jj`jALHW0*Ykx-g`F{Bm z=-Q?qzNz)&b2q7}E=@{>;@p>fVZ*yce4mBImE&n)v2>&joU{JTit{hi+up^M3F3g5 z2=9-*vb(LyvEi1ZSGX}WMUqYTL|DE_lv_G0nD4?s%nn>X1Q+(TcttHX)0e5Q( z-CwtKrSqBznxMkj0z`6bz;ZeT+g#1Pvl&%;FnR|tC>ZBV3j@>ctb7`Re3oigwgicl zDH>vI3s{Q9mP4x@pv3kP^muVOKp}=})j~AD>CP7)*f|+DcMPfvI`r*aPeydG(80dp zxCLL96I!VtisXIIITlgvuv@T@6Xm@c)U7pTnsF=Fatv+1GwlQ{!1uB!)QKRc4;;GX z?8(Ri8W)^jA_FwL=RP%>pnpFKfN%!8&2VqykJ|T}dj}TzW{`;O+gs`7vEnuxGH#6s zucvMiAnsqMgg&NLRciO*>A37=?rl)cV81KvA@)}#&F^koL5PM5>aveK@a5jf2LJGN zBDr}Ae;i1ir!Gl~L1**bTpNk?C8e74{2su~36iXH-lqEd0wwD-^C9zxwAbu`6JW#_ zsZzUK2;dqCSq|yAWERiBR4-iTsgt3a-JH5&Yfn?9jdGIL!JUq2|A?aRNmcLGYC_B~ z9yBmm+@Z3njRMsU(2 zhkkNvTw>yF6`S1bEeQIR%a&l7ihAEYa7EqZP*!N6p^B_2X!m?4p>ZbFwzSq_O0TZY z%qO3|dE?$!9^9El*y`hIV!YPsfXf}K)X8iWmIwqRdts<$#_+&! zWDyffsJ!o?0}QiK{RZAV%y%{VfjITbm`YWO71Qv5D1Enf<(eI0msOb;_0LITd4=guyCdN~UC&1Tcm zm693}YglcnL)h7KAvPZKT^?SrC)@5>!oqFCJ>)S#O=-s_rO^xn()E&0SO##6bOTPb z_g>9$E3%i$#4^_sy1@OZ0eeD13VT*1;7DxFR5~CZ2T5)GkV#VO}QdCFE{Ciop@KJYpDJG$T0T9O{O)Pu^#=#+FBgetl#Uxq1(RFQDQmJ-*$TNd{C3qz-KIl2~gu*F#3)jqdLs9V#9_fuD2@639bgpq!Mh8Ixq4gGa9tI->G{azol%hQWkHfi>U_ zzbK%rYWUu<8KeStdi^hmH*^q#l<0!+s_Lm6j(+Fx$^(R4 zf2Qr#YH87jA1r%@!_+K2=4=^_(RztFc4O|VA8lfI#gnT)$4o(-VORrw9S-W`q$vn^LW=NDshzH;mhImX$yvKvJuoR2Jh_ z$XTtzrCiQMJ#v-^0UWGkr~L-%ZTXR|bcRUoRN}?KG)bL`PtEwz2KqM#n}60(W*`B) z?wU&=>%u;}`b@@tU}6YuI=~K`F)Yl;sCA$@ZLr^=!H6-IHorj|=d@)I;@mCB9=u6x z_Mq#?`M1Y`-{+}G2IunwumnBVS8*}=)6Cg7(xT3vj=EWc`>yu@$!eav8;Lrtqiw1g zU&sCz8UMo?e)S8b+MMjkN>P-RV;Ir==pQX%SX@??RA&}oXEK#8Xa5GEXcPz-HJazPFcA<2iY#R`@ZT4 zu|J|1?J*>K87F(POZSf4b#!Je029rS&c%@K6O(8z$}B$Dv~#rUROhmA8zh@WKGXG` zXiBiTxto+q-sSyca_jQ0zB4Dpj+;;xhijdtehu(xm@e>bYtwdvJ7Q4_VP{lf;zosw zcX=s{7Ku8l2O3vTTpWz@254c?g>+D#gjDQJS7-R{Sq68I1n_0OPAW8J=ki2b$~#b6 zNZ!9?5@PY@sBI*41(Sm@U#~H0S^st8cm8}pdl-jPS{Qajt=7z8=2n(4V+`iN!u6tQ zyQE*wd@hh-TL|f!RY!KOnqlP@d`rE;cD9$a6^h`gaWf^9(gIP*dk@`=1W0mT3m#6SRgJUA(^ zm+Li)_nvRo>dPugz-D!BGZ(i(@Rviv$X%3SL<``C_!!5-tp=LB|V!1&ws#?6<+`?3M&vu7C+9vJsF~;sDo$C}n-ZkmJ zxnQ(H7t9E&bQRscrP%uA!}L4{hTwU`kV!vz;t<_eWUR%kNnZX5*O-_4l2de?34rNt{y$d(Ll@T2p8qYV~MjMcm_7m<*J5J-DoX&I5f&jrd#_n6uv9 zGA6)s?A4u~ztqQMd>+m(Y^fo!cNii0Q5G(n5mBfLJ7J=yS! zzFCx;0lw382z&CF{9sEH&hZ~%}uV(8k&U&1yE-DOP44#f<*wGGT zP9%%zb+pII*}2-3QQSOl4L|{a38Oo!?+)ZgAbH*+*`-TMZ`Oso-=Z$UR`CHnpx(x* zf$I!lg!71ssM(y8be>bg$NOM>vOR=#rF~y0DF}_0S`16WSJy_K{ z3^WWQpB+HkSXniS=4OrwIJ?D(7gR6lW{r7!d8v5mPZH&di9YihS!`5Rjfn)U9-dha z9W!reF@IbSEOxz}3bc~9dwjRXWn6WtH0zBke)!1$9XHIwbK2sfFs?gY zV7z}```k~CY;X%b4sE>>o|eDAb%HM?+QGp`wzP1_5(uP*cwJn`y!vjQuE*&C7_nnA z2q-bmKRhT#nfqh(QG$g3Z0U{yy66PfwKHvr&cKThg7&B?VSkSkzot`W&{>D$ri|k3 zwIjI-*D&0CPKE+Q>BZ9ZJ*!uw4m(k{1u((UF#BulmA=F^mjao!%+Ne^ko4Q}L#cPe zq_=QoC}gQ2v_L#U{}8{CR(COcrp*GqCXY|fio4#;t|A8uTCH2;K4^l^<5)^bhSn6j zi!Tad(IqWMO#`ii{`_X{(jhNvPhA_DfnWzAemlpjhi2N<`!~VfS6qimC0owRJnwqC ztj@x}B3Og=lH5(u@&4s|lEFvC31lnk!1I~s{)Z>MyKr%FaqzKu^@LAOOYqD0ZR26j|#2^;K2z0eB(k^Ry1ri;2BB}vsx|x=KY@YfZy)t$vb1%B(Sq`>e121Wbd@l!Y<9$ zU#m4!G9uDWREG3ou2yN1t}mMAP4N4&$G_Z4F*acoZj%yIUQtlLjaEnhF>eyr;4zRA zf4>KpTi#&J&Xd^~XDx~CxVRo{M*f-G6cj_prDN(kFDECr* zF%lDm^Re;kuiDL-ftLmwdB9+N6Ii$n${}8~Ma;GRojS6Sb~81#MWps3{GDO@Zx!0N z(BoU6<`-)&WPiDVimgHaK}TH#YTXbsG9+fyQ<-_Q)d1B+xl3M`JQd-CyE?Rs@y6Wa`& z-eEVY#otlz#wD&zwOt8Ka&-gwZQ(v#xZkL z!jlJOEH?zb3)x*uk&ad6fLw>J`!Y{tm^ZVAq?s#9>0$Y%V)cRFv&sTb1D#ZNTp2Iq zHIIE?A9SC>Zt*sid!94BP!tN)`|wYiV)WXYpCvTd=(&C1M5xiBH#Vk>ao+K~tRgIT z4FkcLiZr?zLf;Zm2@U9febZPwvgk>mgT1oKH}zbvTE-N-gXI&t;RspTcKm?Uwf=cB9<-Dl?Kb9>a zLyA8Uc;-t$;)#0o+fe?==isDUq4_Xt@Hca-M}MVa2woXqxgzAsRPWMP%4XSz=M|K_ zVDf>;G27Kmu|r(tS{df1=teV&!z8w=)H&b9CQ{#pk3N`QVMnwA!E{TJo#E+sfgKIov$z3}$mCoA`L!l7a!q@GE({s@RuH zdSTsdNQ#i+WqR~7zAJ5};Vl`fGgs>$5Jq!>b9pmg*%r6-i_uym^ES5(^EyPBzJslD`bU$?Dmw z&3*YZVzVlL$E<%ma@_T;0M}sOOc%RdO7{@*YrnyYm#x{dCp~n=$Wl)qm-vw0dm6?j zL^JrL$(FBmS5gD_fI*|=V<0pls(sSj(eRzAc2nBQd&@=HILEt*i(P>nN0athZ-FQtM4FLbkf z`^BnDb@~kHcBRP^CzZM2dzjb1$~xdp zc*{tmRg|i| z%fZ|whD9B_h1(ns!V%b%TjL|YG=a} zB}Dz_it^>Oex?@h9rt}9z5PgZ^FfupogQfw;A3Lqf1mt6A0cGaO1JxBWzu64&?3!9 z*}EJ0icw1bL6~Z|-P-R;DP-z(RG=P+QBGVULuJC^@r~nUIXaIGy#4P(mJHpw+m-Jd z#q=V)XEH>Byt%Zx=Z76$5Cc_o-9#7+T?9nIX7N6)15XW%9}ynut8nNDR+xe{-|?YM zHssoWEPd!Bx(Jp^CPBT;ID>eqChgX=r`bg<51LxIpHR7Rv(bjoZ7VaV?R7iM?p! z7vjy4#KsMMbu=?7mWE;az^~QCKTf`xpEC8B7@9mtwg?QB-9H&i!gq9h2*gja^=~Bh z+f5;=^CxP84Gi^P`1ye;sjlWX6&i^N9i1{*CJ`7W&+nBItd>3^_FHsqVV%yF3!}^s z^41SZ9Rf9M?<%p1N>s}$Xv7~-$lu~iphL)B{rvgNa5iqC!fZigsJ8Mh@kic8VGc$z zVx%MWC~!9H;hpl`(H$vFkVY_NQCm%GNvvQs>%;2+xs%i7s=aE!*NNL?{9VAktMwhQ zlU0{ne>;&^LTj$78{LjtNG;E@@H*KoqTHtL8zzeH1YSmbNjE!Fge>$ayxtI@B-$RE8YiJyb(mj##OjLMsE9;MYIxjLHU)lRjstO}0vy2y_u zeb@aC^4$}n5GY1Z@4@X<^-ohNH$?ySvVXsJcJ|eUb8Ox1!;9*J|1NdC8oRO3X6ttL z)@8R%)tWy>HwhJ6gCO>uMPRiMV#-fWmVcBZei+-`35++>&O=W1Nv_L z1XxFi($#l|f05d2X=$qMzWojLG9eHY3fzJ?him6cY+k`EOArFd8=kxm#qPzGbKz_k zSF(=?W2S3I{XOGPtVx?pM0W}5>2^(s#A9)Z@sCs`B|Po1XuQ>ct<8jp)Lq)3dk%ew zp05XGZ2kBDDRnpC zZT^^pkK%gCftL%77peVrm{f!P0Z+jm_!Kp`b@Bx1h-n{tX~~;x;qiS-HlXPwP(DE9 z_*fetY5QT#@R!>ZYDW8lX}}4`$@XV5j{-25Ph!6j zH($Cb1+CU6|9cY%T7-U=RMyZ?fAEV*5a?rVT6_2Nd~`90QC{Q$YnD+#v)<|g2kV%< zN8iKC6iV&B;mvjtsoN&EllICf*8MAo8dcmYmls&q{rnX-1j>`>TV`!xkfRMyvkvDc zE;}dRzP^(Zyb(NKQ$3kMXGqj(!wo0vuhqTqEF&PdE_u=uqX(|fawPEe00csxW%d)A zuYce=g)LU^A3fJvY%v$(--6rJ^Z{KM=aos@ zPZN~lsY^CF?oYb{xbi%l8z@rB)VUp8e~%Y!;b%EUeQSH~;uoHege>Xz|8TDXyC$k( zm`iodKb#VD!L6&$weC1KQL?cmrXJ`G?ay596Ot8-Nt4tMFJo%!+T{M$dLekyEu#3o zNe9D-%oLf$(jE_X=bSGP3X3RXV@`nv#nCU2i!FFo(!JJKp@dT}wKG?JPtW)Ns?EQa z>Ccm^xIxS`AWfedhrkEe2-uQWyPa+E(9U$+HaR0WLIQ9p(`WxOK!v0{nw9wr@55DX zA)a%#@>FUs+&O0*9Xf2^@rx+sa)yZdzU{bvjU|x(u?@>!bvU0FTI`B}d*o-bBJG>c z$Qm|k;Arx)PJh;&*e2i08_FL0(3^Wk_&l%L-Yq6~Jx&17u#I>oPbXtCfs~ z_{eDT^(N~kqhzFsoUXpiam~h>k+ue!1m>}tELZh?ns=h9p!#=Lnf^^!c3XRDL-Kn? z__f0pvBv~`i?(%TeMeTs%A)1$6Q-)!wP{bbN?HZSF@)PrsHWw6s5X4)Dtzc5Vol9* zPr@-*rEe-xy@4WRci_D5h4bvdU#V!%s|64D=h$UcmpZ~e`hdy7^t-7CQv7zC>c7gL zyfnCdT&vllMBZ2a$E{?-4^^LrR7R;O(6C;pMu@=}-+G>`3gvNv|O z=gE1Gp7F8upy-GsH;Ev*-ol*k@rQd2D>YtoD2@xBEXwBFHP1bGjE!CFEXw9AI1^|4 zexLYA=gpGBW>#7iSdz7Z>-w(eRTr-8pSN8ee@JlWOHA?+|Lq}O%uLLE;~wsp+DKQ) z%ngoej#%g*z!oE}R2->}A8N`w?ZWNywR7$Q31UFS%0xjfzJ~Lf`6ie}8{11M=l>Ie z(q?jnGl-`pFNlS}J8gZ-Y9wkMiFImuf;iDl0mQQX^~IL6qSdL1g&4VW;s-@`R;shpmBf`M%!-kG1V8;iMDLK z)6NWX5rMOw%uy6K)vI%MD~kmB>X!_ae$}PBBC{a*#;$?CZpT>Y{wI(m^uuQM~L9>Fj?q4n!rt!%3s?%V|EqApu?Zdr1 zi_>_LDval5htVB{gRcTIQnzUbz04Yal2UYVR|Gn>xSWR93X9(FHp#XN+V? zaT1RH1aY*n=hXb7e2cj$@#8n6zkLG^7MkQ=R5{gc z5g+|aRsEy1{<&jElTHSXl2)hE3Z@v~ms!QY##5Io>gyn=G_pfoeB%;Nm8MF zXZ5_Fb5rShZ;r$V8ZrjLuHWiOB`5!8{*U=lv??Y&n*bSO37M8vi1PW$M2=*WLr5xF z*b(fB!*;lBv-8su1b^~A&~6kT*h0xB%=O;S>}l@(^}tHW=5P+K_?-%X+xTLm@5;2~ zY^UW?eB^gEit&>{H7~H*O|eIzGY=@lPRc^2bEI3o0Ao$tyjfaOZq{J@N_51mCa;@_ zm9ouAx`=2^s&&KWbK1Tb3XkcWuYKudLb8Xg+M1Ij1aY?8Hm;mV&O-brVBDY(+r~}4 zj#HhB-7i2&@?)}mYiHvw3b7tz5}BzKD@QbXa^Wk$V)BYxQ1G~=Fup8Prd>Hp z@);U@Hs@`hEItIbJW?<9JD*TcaJZVh*})p$TT3XG5^;KOqjQ$c-(1cedrH%Jd17(^ zR2UU(n|f|LUug8AHPG;{KvyU@w4SK+T~|`dWzy@aVNx*;=SZ@u)X}$E(Wfj=IQ71B z?`eV$#61N8*>MB}*w{`j0XVU&b1BPTnrOG8b~P|Y(J+?ru>(wstw@3-?A2dR+`kmV z|LAn9H^gdT##t;04D#4iO)}+g#XOp*_o2z3Y~L66J61_n zM%66+@IKwNlr6DfduO_=e-w-}t$1B7-Tu)ln$qo!@oD`hW0<17d#*bgYT3tNC1*I1 z=TFzXYEX2rmVQ2V4r7B@)n%A-JBhD1H2dhS%+tURudm10+ycHtx1C)dF2Q% zULdt(AjppPp{5C=hK3P+|B_&4+-Ls$QQtwYD-d{WO;IFEPYKa419XCEznA>5^VK4^ zGz!}Zc2}yvT8&x4ASslmdF%jL;%vn_4Aeb9VJmLQPr_Mm7zN;&JpK-N0gVFKOtPam zP_lM}6YH+Y8#6)rWZGRnXlR4K#No=;5BIuWvN6Amzhvcr!6C>pS6oP=17&(pS9dh8 zbzsS@@xY4oFtB=@axu?tqPu|H>PIHuF9m-K+6rH~<~#0;1%hQEb-p-XveYxq2w;9f zSyn&{v_{-^^Z&|e!!ysQ7#z@d~B(@GHR#ujCHs`!uNbr9XSG<$xj_=~Ep>v7|!TUSJ>7AaYo z#fd({BiLUI8Wt-7ZRC~-v|-eBv+fh)U$d`H(& zMAUrYlm)B+2j*&wnQlHK;nkSBt>NIHzO@Bq`eGUk?)^G053no2_!U@^S!7a8PT=?@ zqp!Dd6d+nMZ`$6QV@MdS?Hs+R7GM%XZH)72(G=0MliJQB$S^jpjD6C zKml1!&Xcv^%%qXAWsi-*jH)vqCcCXK6in`XanRW&P0mqKg3Ygpyar&AWPBx~S^49R zd3if&fCQKlW|+iXp&Aoh%AXF?cl8d&W!GNP&2v^xyLk4D7jCx6_iLl5^%So4t}VUe zU8~>d(q)tpW;rYdQl&1giX1dd<|q{~b(Gj$qGL41j%p2QO68|oz2)Hknn#A#?~Cx6 zl35thv8>rPSor;E% zIccB=OY)As2Xj^~0nC5ZNh?|tu>^K7g*@DIY5uG@L)<@bstP%Xt1+qxx!SC<7i{lO z+YAlSs&cqF!<849BCD-;)s~aSJgKZIF8E~jTnQkOLpfi)2kw}Xi0Y&3~-rh}N*P+6W=P-sV z=js`TsnmRe`-qjg)Sq-*U%j`+(K^d!4{HK_5kY zgt`fLSBCU21d%_$W@6#jZf4rP4#9R%h7o!O+@QQkW@pu!eq-Qv#p&$a+?<2L3TtTk z?cp|m@5{Zm6yWUM$0J=AkR@7K3Rt2n>tE|zASNjpG35-BMH`PC$jJ#-idL^8w|~g^ zJsutzc@!UCjh?f!%hxqG2Zsmr3EW>4AOcM_ZW3r9hgS)l1F_uy`#CZ=C= zfm$qH7qGvaCJqWBp`}D%-|e0@kB*POoPN9|M=VP0>dIo>^@TclW`>EIyEoREV=3bo z5%B@d+KyzT_$L5e&>B!227exT`^j0So}ox|PjQUBe~`zj%yay4?jNKO9TEPQ~m*E(-bWj-|~^~sMKW=sQ4DSq(&f{+Y%qVH+-%!xVSFtYS&2>8lfl`U_OTq4`Jyj z35b8JtoMwjRJ^)ZT_!$5!sKa~5;=B2LO`XZd`P^0l(C~^Z0kscJf2WycSk0|fGu1z zX|~-m6(LbNl-r-sYbHA*%u<=5PE#tt2AkiNW>JDjYbkJ;IW5TgG<>mY9pb|?H zw~{)pcd`E*@{mVX=0ef-Z3b{(5hB*{@6n(>DUQjqVXvgaS(T$ajWxsqA*~i9=+9 z*PY0gHZDMkLEEn*5H|=vN4W^c%Za>uV0*Pg3p}kj_*}~FQv|bu^nEy!s98KWQ_2&h zBzNe93N2H5TqL$PqE*o{+Ne)}|4tSlUsjI0eJHxmHW)y>1-bqQ-umpCLwo#1axv|# z4yOuzW_XT-@Qg5^rSA_+>Z>I2S<+ydGPq8LkVFD|=f*V8D|quMkQNO1SvQX&=lK^R z&#ry(0@zHPySndushibQy%MT&dsvekF;oa3dj}vl?kK6`CV-6WtvfTMPmKMG^Yghi zoB_3eV$Z4}-r~piO3ljG!J_jGW^x;!TH`1*wLjH^e%Rt%o=jbBYzp9*;q!#S$Eia* zX|H`enY}gCxX$L3hYko6IU8_jHhxgpy*4Wu==z5UEl{5s;2lrCR`{BfX>a5_*x| zR0IS>dPk66LJvJCNbem2q4yFnbO?MK9(|tYcdqw**L%)?-#PDp`$~3p)}A#p>z9MtpPojhfJFv_jUJigy8aTpBy$&$4~A=lA+kJaj+t3A}1 z#LGu5WQ-!gPdm!6Y@CtNvzx!Xan(;NqC=R5I)x-6g})uU4R@3*xNRy3S6*;(q4b`` zf`ydRf|ArMo@wo%>`9{2?wqd0b zo88B53lS@jQI34QK@)bbN2e%nt@#wb?Fh#BJSSjdiCR2b(pH96t@b=j}$Q zgZi*xBy58>diqn>R)>OFjJL}7>Ca1IMc~oHdjpBZcj^j0~Fu>c?tuk|0=_%2?5?H z`OD@zGU2>g-!ld7@UWac48g4xmg|^cela>lLJrYZN8eFnN@SX|9zn7y9SL*a`waG1 zP1{aXjLBLn+l?8*0b9AAndYi zwx3@AXw;vpCQ)or@rbwe%vK{NLxY~)T+9sLJw*8$Y@|9wN4xL+x$WP7@xS?u$Iw3> zeY)A`!-qb7q4v;lx_#-IN!Z>j)qgR170iOKT@Y5So(t+>tLf}J*7gOEAM7I{9kj_R z&i`@%|K&RKe;JP?iT}rQ@9eZ0=fRM?wRhyy{Z=QZF-7-Pq)1w$4S*Un6SLWGg3-S< zW0}>6x||=Yzy;MKXm3eefJmG%;68Wse)_Lk^FLGne}CmKMQXo?M6E6)yd!tL5Z^k~ ztWB02m8ZQ4(&qabzp`OrY1xSqGdU!dWJtpn7O6Aw^5U(RV`5@@rK?Ns(+hCzBpKpx zT3cJc6%>rANJUgs@C&D+h=HNDgZ+UPo!8$#os-qj(0CywrG4byC_a9n)Mjdd+?kqA z;*F|M>9Ur#M8)G}KVd(6YvwT#OOhuNF;n2Qdzn|UB}=Cp2$Lx)Sr>iu=sp(Kv$>Q} zSYKDHLoYe*941Kafjz+IJ?;_J^2rn z>^DgZ8`*<%^lK^!zFWsXhYnm1`YO)hhR!8*ZIj`F2$Z9@hm+} zP^s)q6T_y{SL2gom|fj6sS;lj_0$fSEO~h;V>(Rs_Ks4Fnb$-&=k8HBI?N9KA`#-vl3WZ&-loz4ddFl3)mKc0|vt3d+R%;a{r=1&8tj zm^#IcpZI?{hxl8wb}V){*p8}xeP*`wlveNmLqod_n{HJKUBr~!2f4EnPYbbfyM0M{ zisAq1`dch|?b8m8TVcKt+haMeS$g5}GvWolY2lR_-e>&r0t{upp7qS%{ebY}s9$}Z z{uloGECY&U*%^59+Pxxk6PPbDKa3P5{8@(b&`a9Xo*m?=r6^In-*C}IiN-o4|5n6| ziD8%n>ZIGbp{HKMQ#;1ytSjb~TJyDL_k)K{q>f_*9dm% z1myw##vd2xPfPjnY>V34`*9n)(mt`)bN}|f&U?1;3U@y{fd-C09l;}=fun@_#tpnm`|lN@a7WEtu^ilh)uuDHU&OAv0xA!51&fgUzx`p8Fxb zP`x(!4{o7ex-zJdg6k(WFngLf_{W~z*3^|6=ZaCosfMg=jRVZ2-lUO1mdvCDuwHX% z3ti?VYq@)dD}VeAhri^apM{(3_WARA&_qN9A8K2?AG~;oZdZTIbW%VoY|oE4aU}CM;J9%UTl7m zspF>Z7O$7ZeMiip#r54pUx3nZu*4G!84n|*3a zL{F9v6qp^buQf(sHJbomG=V|cXf|cL01beQqniM?`U|sT6F*?UoD9QXH7YN7%+>UB zpH@Or?WRAKfBW|WqbC^tQ@3W_z}vJ0R)6w&0FVyHz-r9N{8Q_XF#Z(NsbVp>oGUIP$Za2iF4n-Y~mjR^r%Qa>(2ZK>?WyQf1 za;5CNEKM_=-P&c$*k!Cu<^oBzg*1PxLXyD~bmJ&hOm}qC6|)yL<%50InXKdpbhh=P zt-gYZ!Na<`LOA7Tybu;>vAPH2Ucr?~wKHK#lXhffKGFYePU$7sx1WD22^aDVy;4uV zQydE!$BTz1H?M9eV<$d_enFjYT9vzEFLM(gYo}azX#J}!=l_$OfpO!qEGl%}K6NEK zQ%avw(+@@$y3sQ_b-!tbNQkcJX)(Z3c<;40;Rd`n12@iXh5Ot$*+u^NC$3xRPr8}# zLahFx-YBPoDrm@8{mr3QHo<$J=V#jBjhAKA*c5C{K@ML7IR3$npJ_lQCOu&blAMkz zeyB!0QMO@LA#A_{(6ssMEM{Q%ia74(0^C_(7J3C<18(YX?R=#WK!?r19Aw<6PTudxvfjh z47j!P{34n&Ud&wP)(cI)n{oiWHGBlBP@(88HB%xb=-$d{n+b1uQ=iBVjAgkqGM)aP zvEXn^GRl32;b2HM?Sq_PZ59kp=qILlzh~%=C&brcIH?!sEvfW-J?`>bf&OumS65mf z|C;Zfa>~~QgC!O5`+d@)g5Jya++z6bLUf(<>Ty#XJU{Jtgzkv1hXeb}``cgb*kNK2 z+c#j$7#B;t+x)#^;4n#wym~@s>q>E{Z$OM3NvVGXf6V=y;wiw>@e(Xo?8g0%(fLbJ ze6)$uUJNULjO{4PP|H0y_2j@Qzg7*`pNZ@veZ1rKf*;wpUN8W-A4F^LmorJjV0o}D zVm1uyuAC=$F7%@%ElTrJlvBhz{q(UkcmX=UCreTmbmsapI5^?zF99(KKMIGX;GCHm zjXCK6Yn|d9dbXh7BmYaN-Ov1bDN&Nfk@e&s&qVukc>r2mX!ndVl+TMxc~BK#I@CpwlrfJZ5Pu%n@pR> zJo4AT7|gM&m%s1@1%xmV-vzP;KqBvN>$+C0q z|Jorgm-{^jIt;&{AuslvG?BD1e>LM_SO8UMnk7e*-}`*me|4sGppUPDN+a+erI3up zxxFLS)l`W13A)j;HXkT6QqCnQ@?tc2!jB#2cKl#WP!R*W0)zDvdRJHK-;)mrn-a$& zz3*q{(-h!Z%)dR9)eDl9(>C>6+9lPXGYxG$yQ-^P!K}3ZD#}Y}8_PwS;R5GKKk1?^ zUhF@LFi%Mtlp+1$k4}?nF}4`;bRV(+9_e%FrAL@o`KoyJR4P&<=g3bgW%vG=I0KIV zYAEYKpQ;v4`QHK#y)SuiBo$!BclDcM@{f(JKV}ri=w;bbR{#S=one%F{vM{&ZWNBP7QX}Y1oHXW%|bO?YM@*X_As{H$fV&9X&TLI3x4GPcy zO@{Q_z*#nU`{@k4d_rgP$1gGTVOLkjnd3!IcQ9fKXDkkUH__DNY!aZ*yd%zLs0=zQ z=FDyG>HYnPzo$aE3FLIjOwNN_cSXlmZbJSLR$byC00ItEcY;ECK=$palFy!bA87O@ z0w&_mL4XayoP%n`Mrs4OE)#X5iL%zh;*aZE+R0HZ}-` z7&9{mXCxjlG5vg{t-?hg!IH_aKuS6|Cc$7(j#$r^-97fA(wAWWcd+aKgX}g9P29Qv zlUq*ss_NQz{WCI{0{ahA5eV_!Q}n32dKDXQnzy)1t-isGFAuxAm96cUq9h>`)9J$D z29T<%D(@q*=yb~q4nA-i`I0G>o%IVD^ZtsT>*hIC*~0Amx9MogC)KU23x1|aC&}2q z)k^hQe(&tcwB(NU^QYtP`;{zF3BG&8N=D+oZu3RfcZXP*SuXdc0m9GHgQ*VcO)Myh zvo9={Cx$r=;H+qGm3)a*-aL2)3frzEZMs&L1{we3&Tp?dO%}P?2!?5 zOW)XIeceZC*AK{k3v}wPvISVuLyfuTc&%yHSvTt3jEkM~rALVcmaJ@a9N%K6d+Pl9 zX%$=#)-V2=DSMh{$n@@*%%i@py@B?=eVibfZZ~8)86If#^ljb=y^c3Wk4r~>NJs@7 z>sa@58nQm2fuWiL*e2$!bRAbR(wipf@dD`NV*;^NbY%s^~`q#2$94KAS-JIa@q zFMsvQ2d_i>y+==`FJ_H{1s@{ETvHpsPgcTF2p)pvWIv)e`@o^0o!erki#w4EB>3-t zxC~c)Al%UC>UDB$*IwpJw8hYQE5#~)8|bFimikr)I6s7q&RkeZq{d=E64m-a*wWHA zx(l%B{ZOWrkQw#f8GVV>xwk6Yhs5BqZcLXF5nVjAmSa{(z5NHAbC0 z;aIGX(Uixl>>eYk<@loSE2CMYcXpZE45m}NxLl)S)~1uxnBxpWth?kcR*#=g1N-DB zA>V^~spU^#3q5kv_E}6SA-#f-1*4=I7UDQ-1wvMrnW=I&*n~OE+OS%A)a={&jFecNhjy_9=;}n+Z z=>s7|Jx(Y%rBj9b^}tAhDY!S9tO5)ltqSwuhBV|0GgYU4DOzDX>#d_l|Im`zzg942 z%adf2p#dxYA>mXhMr*aPC_QlymT#EbnX#Z%ZX2kAZXGaY3D9SqNIg83@cOtmT!I~Pe92n1e{7%{vB$i@|?k1 zI*|j&15OYt3nB$J1QU-#W8YKrs|kKqPxt{#-VG#df03+jC{flK&@dYB*&22L8ZL%- zFwo%JU;&xSd+D8_J-Sd&!WY99U7U83$!#jG9g;u^C00+;)KJZol0L*qEX^>o`zd_D}m+CL$FTeC$1ll0Dlng2iiY` z2l?Rpj7c1#^2`spu*)dMXRvbq%G68V!^Fw(Mon{+w=j3V)R2``OqxJZz|-Sm3|^Hz z!8&2qt#~!0KE^kO+tQs_$8*iVMdNaouc&&+Nc`ql;OmsJ5ZT z8)B6EbTOJvE87hSSWVR~iZP1WBbfE8H;h(Em_YnVL zzar0Z&A(aDVbgpF_Hg~t18o)fGqmm-Qgbm>3uAVJf_l*_@>}>b3iwLo zi5n`cx%=I&aRa?^xi_s1w=d4TGcrA#$M>T5QMsf4UK(y3q**!j+Jm&N?mH_9U=de$ zDlv3-jk}{qGC)gRO!uRv%Z8Tcb**PXd?Jwr~3p|ZC^U~0js=oNgsPlkgK^`S@pGsEuDPl$=`GZlIyv} zi!oz}>!!kr70D6`H{)i|_K9%{dQZN&SdAT+n=b7KMo5%@=hJ-(f7NQfe=iHk-xi+6 zpTD9q5V(0WM-htY!(`BW)-TH{UFP1qzcIomCDrq4v&bfPz4dHiG}^2u@%H|=gnWM6 zsYfF=%jAB30X<}o5^YuH3olNFte4kV=mqtK-Q?XWD`ItxNORa6;&iRXicNLk*g7Y= z8hsr;gS294iH|+~4-40qR?-|bUkVp~+VJ(+FR})0=u}wPCXR1z$b{41L>Ft<`9#I_ z?*rY}yWaWNRc<}ZasmTIH><$qE*d#OtKpI#9H@0|{{Ts5&_q4bq-382EhySv6SsJh#ir0NDox2vg;ZcxF?5r6{ zjMDe7*0UC77O>GEiMM#^T$J)i<&A;o1{P{>F{u7I;eboVg4N9olQ`M&1jPaG_(gInADn0yy@TxY^&C3;_0wGbu=wZs zVl-sJNBjq}K6(mANOgQ zWQk=#PiOfLK{5__rbqtv_}3Jw$2RtYME+o)xI=WER@||>YgtP@sIdFxdPfuIGrU6~ ztkc8FgyFgg>A5!Wv&bTYn}a@Bq^8d=oKJ4>9Y^9#jgSz4gG=6FMKL=R-J=4r)ra`^7f*ZIaDUVdH0tgVG}>pA>S2Bqjy1Oj~;#Zqg?Q1lSKdZH~?KmOXKbYgm&Q$Bsas9Pn_zK@v%29H&UUofX{+y%A@;Hr3U?kn3sdWg2*iN;lTI~0?u z%>s+;9@z%gtE>5@TTa|CnfBdLUey}bGC9*Xw$b0*VPH7TFY=qN^3F*(nb2Gw&MU*! z)rBH?L*t5gGo^_XToVz_IPEZVEXOHawoIGYeNQC4F}0utabbbYZko@5%Z|v(CqlD} zse{>k+)o1;^!G0yzUZLUMeEj?izBD<$peOsxM0C@5tp_`T*w4*`+;{3ZyUeU!5JVf z^vQmA%kDzG0fX+iwR9S*tg>+50nr)@NNw;VtYNqKIA^AD6DNjc5=OG7uC}Mm(=!@) zVrq0DZaETs(uE3cex+&gXeMmk$n)xN@~NVXsxsh>6+6(;t?>ZwOlw|O za#pe$&W#+SaV$fPI4Lp3IAHa~?3nI>GdB+p!?WFovX`H~41WA3Er>(Au@jNxhLiNl zj!m}JeJllvH#CLWI95x2Y z1D+qgx(u7StXDvQ6NwhOvWvY(zvQd^MBVaP$s*w{bO%n+-65H(c}YhjXrp{C+CRz~ zJ?Z+vgCdvyJ-SL-)}Y2=h{9=M=le3Qy1iQ2D*aAy1M?Jv#)$Q*`?E2w6nU&u(b;rD zosskXO~~F#%iH{NuTym1G80XA;_8EHO0{2M9hoBwSF2RSdC8*~e4J}0e}ltcPiAwA zqE8&QGy3yHcUSMz@Qkv=XG@2l4Ibl&a0DTBVqpp9g;4lyt6G;eT8<=(Svb;nkGlE@6(Zus z1A9A{r)*aY4X1wn;l?`B^qDG!8#&HSuWohACmN?D7#rlDWgocmW7+vRHsjeqdqNA= zf$o8RkTI-Kpkr&$u{7YFE?dYUi$I0%+A&KxU?TEMwrGW+?64lB@9CQd9~Iivaun$@ zXT)VUQ8SSGHKtRQ9r&;%(2W{>B6V0oO4PQNOiIr{daDG4KYr%edU!X~!Pr({!Dcp? zH}o`1KWWrV5{59VCpuxbptnOc>Cpk#l%FJGF~&$X5czZ0s2XP0Bw-nzL<|@v4M1Q2JnrRoD-Jy zX*;K|TBdfc4Vm|RvNKO?8+y@7Y4ZtiyTGr;CYh2fOegp30?IW>5au>YYqHX7l+&ax z7o!paoqJI+8;_eS)(qjey|5SK*7+U+=Q~kr-^H}ys&5kNTEmG1=G7c2(W{7j$zb-{1Kv$1}e&5 ze!7I;%W`ffD&Wean-$nFPA3oG+cxO9i9U6836v z6{apU>9%b3OS(4Wn=|I($--js7Y$plPvW@6Nk7Tpg-;@gHHul0bl-B_`A}0s5kUc3 zXcuZBD+_~?n5>gR^$2D*CrXZoC-Sz+utwNzz@n5fKv8$RoAm0{3~m*vYS`UaU;Fs? zoZ7%kyAD8=;(iv0`aFeLl=FimZywZ&KmBy=`0QX@_k`1mymitejwI8u?S`+?$5G4( zCk5hyI#@W}q`Ql_&-7s<0yPRS?rd&bM$zP|8HMEOiomgKRDP}pTKP#wzqNs&IEHTaySoHuSQrhHYubBh4ZM z^I?MQ&TLNjE8H@THdLfZTj$;xrShwSg|;pdrFEPA8*@M?O32jniS5B_vr_OvuHrkk zb%jT@wq?7QV3BI32j8rKF{7gjbx-J{K)fa19_=ro1JuRL1*-1}aq*WeLi+Ks;ezxa zV>gi-l;M)e+jPUXD0*k;%ypQRQ$J8e29WKPTh~GqUYXQzy|}LkFfQ^H(qAsL8nhZ7 z^1W=fvQbOO_|#mG>{m>|nV2NPwrV=t=-D=LM{Vhto`Or?va{KU-HE>th=gnIrQ>Q1 z^f88+XSyxF4~@gSaq`h~=X{q{V&elL{p>#dBBa85g&qPW_Z2;#yysBS z^BpYKKrOUsk^uOP@VL$V1ed3;ZU3B|rkIF^nl0^P{)%@0K$)it1CiA!PJ2hUWh6nm z(5)Tn7p1y1Qo~9{`_R-t+JN3)z zFRR08#pvZV1E242C+6p})cJUKdRB?N-LcRAnGhrycD-e(G1Gw3b$gQL(O!udyV~%P zW}gF@(Ml{+Rrr?xNm`tF$D$51AY8>%c}DYASon#gRY_AHmPc#e3Z;+sw-wCu7mMm+ z+XSlg(2LX^kDQQh#Ye((sij}n*1snFOX+@lX?MXq**x&!O5DSPI(sRYQe`0sZ%Dg$ zh6_@|eHkVXI;R;{fQRybn6u_j(>{LlVn;gnTy>wXBc#roQ3*4!E`bOmeXn432WjQhf@S@Val=q$CV~Lxr9HltJ;3jxO|6KFfCoVVP1!t>C&o zF0SR;bE2@EvGo|3?c__h)3KBe93Glgry%`xC5k*onvjNi|Lt}hDT}Pl&N033UZSR0 z;OcZaTwo|lq2J9+=wUW|$fL|9{*sCF7qv)v|8hxo-p@(v{uN`5r8R;XLFFyS(~ALW zl$`Qe$y5AUG-KmLx1TxLdRJE>s!(ko@E)h89_PTBHFe){RSd5yI#0#JZFVf0jp~K* zGWvE-_P1h2x9wp2*~%=^>>3}^zJ9B`MbYW$Up+07%h-J?@mP@uAjVi@zf-vkUH#8U z$8TS6m1V(-AF?l?`dstFENeQDxI@qevt#MPXfqsr1qsvIraP~C8sT%;0Z-9VK6*&{( zL`f}{{_QuXN^+mzInB#3N9tb9i-G8^($|5%3VY4_QVC5>6-RH5^o)4P&Q}7aGzBZt$PwCGaxe1rGhEzIz52Huo8vVix6Lv9A6;u% z(^35)xX-^GH(E3(cb>j9ZLP47M`CE@ll@cS`B!66Fe%VH_V2Qvm0~*ahM%C_1pNuU z@YFe?P!np~({O$BaTVw?p140ECRma}bR^YKkgYn5$^s5#Wz=@qcu_Ea{GtraD|Ues z%VkKOY_c_3+1)-D7`VGm|A1^Z2&tE5TY8$)|dqCrW zWL1g%`SznBM@*1*7w7FOW@o#qKN_#o^Hq^1qLI)neuZ*$I z`Y>CP4)i6OB!R60p*mF@_-056`vCD)}1RLEG1OfU& z){`xw7u^e%ILmnK5C7ESv`NP+Om$0&09_w{CpsOln@DZ3bS7Tb(b3a8?a3}mCYCGi zR?M$e>|Iy&-QJlYe1%p_y7kt0m3Lf?A6rw4c1*^=tO;796Tf(y6f z!~UVA-6wm{BAni%m{Oq(G_OwU{}2r5fLrQ?ZMDm{b@f4oH;de)-ePmZ5u^DofroY6 zwVxymM*3vV&^tO>AoJc-K3mTg@rGI&A&c5tfhiACrUppT6J9K6_^UqRX;l6;%E)(@ z?Bh^FGg>!}!sXrWS<+l?8N0GUdCK%w+3&7aZ|>SJt*bjtS$_Lw)0$VNUV~TUcO9qk zWU0M$ugv%uh@}Bnai8Sl!SlgU!ACEl6=40wLJ$5&E+^dr1@yst&j+g%HmqY`3VsbM z_vUO~G2ms*297yY$EOwbYVkVip#09YK3cSZQ_7Z}_GgJ1;ZC2Smq`(h+@7wDpH$2| zUYE!y6C(=7-bC*#h&t^KW<>{;`z*v7HF}tB*?2by%qPU+9-lh}y=RTwe-x*(_Ga3q2%8yQAigsXW0>~ThJAmVfeBY+ z8g>Hr(d(0Pw(?_J$urmyb zq#6_@Ue}DU)jJg)>kN`HCq?H%d3GC2A<4oj@c3}HOBXd>o-vj}TUd?QBO(DF_pPm) z@*R0Y-z!#|0MCt^#-|@5lDMrOni=}86-nhrPvm=Y0Qf%fSDmef>-S4+@Vk!8OZK2bn88KQVW&S*vSUk| zJH8zQq`_bdL!AjNWk)_!P(l52%@2(&$<0Mlg~wjvw~B;t1}+0r8J zPu(RYcxvQ(|E|7*??J6gv=!7__({;64n({~S;In`(tTj~I1h-SoclUT8r5h74H+SbLf@95_Mh=d`3wt6fi*+H?kLyDyWl zZ0z(!S-UTtDxU7rq7-zEFLv8t37)OfuLEDm#Onw}2Q_o(m$q_fS9R)c zKTfwB|I_i$h1 zfeiTb7FfFBtupb)C_himjcThCgAGBFq1{cmThoouy?4cH!677atB*qrdkw2+Z~h9A z3?pdS(iuX_9^PkiRh_VgR65J89iz++5j(#qH}^-!*NW~>gQenhORc~bgBa_}Pw7AX z&GH6Nk53%uJK=Oj!L?fW#PbSsG4DGD4X&AwomWeW?uc9KR6RJUK@`MjB9}%ZM_NvQ ziWev~H3(LWB^Imh_USd`HSSGCE!)}ETL&ULCk?zsoQfwJouJ-64t5Xj&b<$%>`6lU z*hG079%vxJY2dx}fu0$)!LfXxpyeFY-R|j+C0JUdOUAPmtAC)~yvzyzL^3v#rwaER zOdFg~Gh|w7K$%bEYi}P< zb(!4pT2|j9;sG=+|JCeNJ4ts z3=`BEI&Y~Ge+-qJDi)}zupJr1{kDDhad=_4wFmxbTQe`@%cJ_!Iexf3|5VZoL2A={ z^Rdfdx=wRG6u_p~3pf5&%4=#c$-ASNDPp>*3@_&ue7BvS!c<}C*HsBnd+k+d&S`Nj zWupuj^qk2ta}g#<>$`Hi4&+<;f$~>~`_jH!PoAnN9eJ);%fNd+iBB#{DzWAyT@8Hb81psTm zuiINrGXKqKPJF3PsH;`7oz>ssbDWQn>8o{VA1x`Y7TyVpt47cRecyzsaK~iWJm76Vy-pSAqgG&Ao)WhN}qm&JY5vlMFQ?wsHtn z$Bc!I8E71#nfLHUX;YeSZ)#HSMOzI4nBpkAN@!9QQHT($OUR-3ZI3?0*?@d!2MKT! zLr=rkuh~^Lu(02H7vh0E;BwlM86{t4&aoaj*AS77)9~pipArnYM+VR%h@sXBT(TuQ z2d8P+&kem%fhv6qKmdMOEy4M;q2mNQJV8yHW1UyX8;(2}M&8xf$Y(*dhEbc=QpNtY zx94)LZO^z=IVr4gHMcAz3h8^6Fi~OZDXY))@py>cE`9(S&P&=Bp-{uY=VZtb4nygU zXjI8`%+#M19wT9muhm!zGQN3rX%z3D=Lx-LT2UK7O7723=DvPrCz^7R4}?I%r=|&K zWcH*c0v5`BA}^WBy)PE}0!<==d-bkE&3d#gMknt_DM{>>1RpFWTBJB8=l>3Gt7 zkBe076l7B~=`%eJ*DXVAvUe6(WV`2@W$J@Qe6Ze%-AD$CdfA~{rlq;=<%Ma%9`}_Y zDjt3*xyhp16fN2AKy~KFgPo!bl0&EB)uRN%+aTLal9bYbr!(aZQN{YO2`r1z!eFY0 z&Yvx3ii?DN+6|KM`B9$Sqb;0p*F-xTZi<`J5(r&p_yu^ftfx>L`f$!{`U7bF{eE$2 zCMj2()%;v$Tel5ONn&p9*N{6c`K{p*62JUZxN!g-dO+NE+B>}a)gBJ7~x)ZX6rDwejk(#uBZp-&yln#LYA-lSDMCurH#QB}Gp zTb1rv)(KFCx_!~d0`OU%W+l*DJ?Uee;Z>A!Op0nq z(-m1cfE=YlGh5pum!9p`ajqxew9GA-0{*oK5OhYnY1#}5q%QS!ij-T-7r{K8bO*b% z7RA4bS88?i6>q!JcZice-Zd{|h0zk3jDu2aDk`@F^6-@WCq0J}9%*^`x_1QM`mt8I z923&;m!v`4>OH;#R1VxHnfXIUt5MwDK>6xRUKZZR7>@j#?)~C*M_mPWc)E$Z(-CHe zsSSb;$2kq|@zAYgb;UKT&;t(aA>1ZFH)ie`Zv4o>#5+)%`nu^j<`*c*#Zr+ZEfPrL zE^npLu*5$1)A;(*DE?~u9V%7yEuxC#8@13O-*CAd!l@K>o7a9E3QlY=E&^!u{r%Ab&G~8 zyztLf--%lv1*uO~qt+15Hs*Wjx}ze*Rt1V9wk@Ypph-7vJ}lGK9lR}%wwIy$oL1h! zR;E3CZE+te!S8uMzav;+B(P5bHAjkGz>_lIS|&S9KJ>4&rGFu}+;+;|d56Xo^9T{$d4n}H zX-cS4(>vn1FrUTV0M75}#3DuRf*iR9(Qv=XIwEcgYB$30N6c2qm-odE=ta0%8s{-t^?Cxj@v(XP6UH*(UzH z+LL;qYW4sIsl;d5modx|Cw^-qnQ(#p1)#3qn7A8QM5A-7`}9rgVYU*4{=N{w z^Ivuh#W2+UhKt#|r)TqwPRrfeHw`*q(G?aW>y-w6DKpYPd^HhXhsI7owLC!tSuLOj z+31UC)EGfUuKL{Oro}ugKU^(ckbaWM1~3x^Z^~E0huzlc&c2m;5Pe0I$;m2ucI9hs z;)KQ~!0l=VXds)uvH_3RTVH=K#|7IpXM+(EghE!s{mG{_P~)(IDZT?&KC?Z}gM^Ng zvg-%FhvPt)M^-4Qcd5n{n`tIbQr{K@h+1D^Qpt29?DY`>iHdz1i!6e48!_8p+rE@G z9w4w%j+CreiIA6w%IIP_k$23=8wf5#X6YVv%V~xz0UHcyK=`S?D`=&jydm>|o)LhW z!e1^{_9rvPAP%au;f}B!P`ZKh34O9;PKsM=gU*=cWClZKn; zm`I^CHKcLZhqz_3+JOVOEq70pnN?`78AL4=l zpiMl+h%PuVR@au=_I%$B#t-w_wS>eyc1v3y+3kZgF6~oqC@S&S+brquxlG1A!IC*I&;y&Q6lyTz?Vsq85@s@4t4fzGy9^S zV=m)YI0b5rSZ9)G*)H2o0G9YnIQOY1>*?StFq3>VlW=%SXdic(JsJ$&ZyO4%N+ zLJYiG&(W^?>kluse9wD7-4PNd({tNjS4T~E*GRBXv2954PKap4Tox~)){ePn4U~MG znlHJ35n|4Ut9BzX%ITK;!CQyr%Rs4O$`s;u^ct>Px_&{*lf8%dY5Z%(nLw@yQKxtbUVqmLslOZ8`_16GH zg?!b|GGMv!qjs!eI-y$VmE$ioAe#EJ%C|P@{cg-*EP_ZTf@R$cB_^BDJe`sH zs$50H?-HO!!P>QN!l11C zM5o=P5%I4B@!kneXX;YYbf0{3-<@E2is`uida)8J&Vz5IpmDGzVIdCCQPYFkMQmx7 z_RDOPBpv|d*gf%_^CdUfP5?ZR%2l%A8Mk)hWJ(JUNPvsDA$7qKLM|WIOLs}z__C#b zY|CdMX*g7+zz!EN(FP8PsA5S5qM7mZ+Liiaq1C|lSf{L&22C`j7}N+C>)NoN& z;#eZd_uRRaawV~K@1QLcjrKom#CCz<`zqI29JDXb=9i5KJAp#9l zntsVKczCzWxut%&)$rQ$4JW$B=NrI-P4YsnDbod?XSapd^qi$#oajPUWMRcp$nY-? zK*)ZS#ioAjb$`SsA#MgVuwmJ~mH8=^Nli_-%FZ1307=J@?Qsr}K5!Q%ZttWXPGH9Z zeAtP-t?6bmasiZJ!AUz%SEZrbxTrv(&2QPFzfpWj)fEg61}_!^GPp~0NMsl#{M0aD z%!@KN3dz!be3hmqtMY@tP$9|bi#|Mx3I7Ct79Xn=J;IW*RdNGRF$PTa)YYJp8TKdH zX)kk93qr$$ZhUG~D7-65V#~wevcMx~ zi=ZexSv0;$jZ@6ThF`7D{drw|9Rg`ER!rmC2;OfWSXc%knok~fVX===o%U{1$)x)e zS~Y-q1a7T2ODwW!U&eF+~`hC}f@JoZ@VjT#hNgXWn;{ zg3dZ{VnoiWY%oXDZD{&{$=gg&h*P@Cj*_0MN7Xx>()A!`W3DWJ`##uBl~8P9a5Vjw*JNv2hlilJb9KE+F10;dq+GV?Q0$iU z;Y(OznD!z`p3sAT^sY>nWPy`fV^pUf?WKxaInSgi~wD%2ufgIvL39{eI6 zoA00&RqNl_C%IA9(wAV=tWS5fa^%a$$&X^E!^TfCGIC#Kejw}tcqF{j;CYxE(3!Bn zX@Hkq6C%_cM0D_lT52qBeT8uV{m|eM5T-*??|y=v_EsqQ^!xb$D4+BDuaiynfO;E> z5%sAJ+9+Y{JkyP*Buogg&EbUA+d8T;2|dtzD^T(_PZ_cp;UBrz7LJ)EBMOfX!X3r0 zUgEC5M%9~6P8K=EnoM${x~>AZ2#cE#a$CR4EeZY~{ovI)VYS8M9}L%GqnMb!O6js# z%lH=S{^jQDd7iT5j(svBCftHRz>rB#K*lBchS_gu1i@favEE zV7&Jf4)P6(q*^6UNkZ=(%V%3dSq0v?H$yjL{D=7SMoCM}|5o z0?K^VN{$j(L-;#g^+3ob6=~W44%`*jm+l;l z@z>n3NcHp!@vz2JRYF@O$qz8)cYQ#{cJJANIeKFS_T??{=Zi4xBxG?dm0+O6Rg;%Z$A74|nQ|W}`So%Boaq z%&g}Y_nr$r)I~>(z|1s|*Q}!=y{E;1`&(JjXy~Zk>DRj^$3fs8>y&g98_8z!IEOSK zDy-J#X631^1~QA*cl|t9Ozg|WWi~}g?s%ZEkp;z$%pwmJ9)5aE;8km3A!HsRTw>ao zASCP%)cv{t$jv0FbJ2P~@zKN~ZNf=PD2Iek^}Py;k_T{OULYDW*V8X^*m%HpnxU#Q z)HIasH)j~MhSqC|(dBKBAXDZu%-wa1lD8#^vPnUVPL<1MH43?e6@?RBE5lU*5+?n3 znf41VJv3kE35s@p$t~tCH^1=pY}8%xk9#Ft3ho01Svn1TV^7*2iJD{p4kwk5xl>gM zs*H6-8uFZtO%pmUi1w^BMxfIZbF6XAt61GjtEP=!QD>fc?M9)M!sZI+$!W;lkGhpv zb84$iIf_g(A*{lL{f9Pr3oM@Gg< zGdLJ9*S;P5Nh|6Eb={OTZ5Dk2;4MO^J^a=P2g9SR)?MoGW(uR*GpbB4ZL4`%O5Mpe z#`;=^ywW;>dpsgtZu5mLWv@6eKDW>z@f#QJeEevcLnv}7jGE_N**kGYG{o+5LnzxF zLPJKQmYn-JA|XaMt9ZsVsL;4J%6>2}UZGQjoZn?5s92Wj$Cn;YxU|_sCVWK9#N&lF0gDH9>S(OyL0KW=NQZ~QsQwgx{ zr?o^m<*AziVZsc386Zlfzm4rT?Wm#RyIgbi3i8>85hCXsJYUSxi6u+X+YXf3O7fD4RS@E@9&)+|NNU-o?F4tt2)QyOy~| zM0!6!*_6GlSsV%3oS)gd2sV?qCdMM4#r>9x*axbsu>iCGQ22?J0&oIm1_NyYeo1DJ z^!Ik|jk=&u4Lwod9&)E)WaK<5*+TN#O6JROk_%DDyE%UYhaQY*aa4_LQ9k*d;Bgmv zGvsp7LAYrn!X3THdH&9jAdmEfmXXnSV8b9QwJQ_1F!x0R{Hz+>) z;0huHz3<2O-WpPM-|#oFk$~|aAXVgB#`P=f277d2 zP;S?-lDE4zH?48w!{e=0dBWl_^3H3v0S4zF+j5}Uzcf5X8=?dpqya2~qZQ>$UF&`yZQj}|*6ld4=RGWGv8ZB$<5hQACKt9pP z9Ik$Qhme-_Rfo4Dij$$+#rz#sTbE@8@3-BR)JG6uw;!v11hesKIo*JmFq>xT-ax4l zFq9~OyS7t)i3`5SXB@srr4tTq9`_w;%RyUs0w;q~3touowIYvsX$>&0$-`L4p<*%P z<1au~3oe+MAJNov+iS~r0brgaybAHo@!YkeN0-w7cq3vY;gL@X3nP3nx;NWQpVAB6 zOQNQX8nb#BWDgq}U#LC6r$F(UAlF;c5_db#JGXzH)%K@e@$qd`<#OUa;>J8@l#4?D z+EXt?+qo^Fz87`33S^d?+2maX_s8Sa8`+nn?C@ieO3RVE=k9#42#(P|bb&5uM5GZ~h!4md3_Wj+#UT4Qsi zfGOWK!;R^Hui*odL{KNqg$s)SB@ftYQw|OYbS8gBM(k9W0#^tt7Ezj(!ehelU+}$hX|4wpPMNPvt&sd*7mP z-$vpdl#XVmtp;KD?y)8h2JmnOF)wzl-U+RBI^vQ663{BwS9nWqZ8xri`aHwc6W?*@ z&qA7Z-cWzm2Mip?{UvRBO;50EK(YxkE@1zUt>r^A3{LZsIC?I`7x&MV@J3#2vE%j& zf|>Rze_mVhI6W+sNiV*=A?8|E<=o*ieNkD(TXcsIo^jZUx561tYH> zhZ*CH5MVbz`dnE!Rh}r@Sg00WE7n2^@*7f1YD zru!`3^d4+`@IdlWcdER&b@YSIo@$H!0Sn0*hY5=Z^jwS)k|ZLKm9Yi1Wrn$(LA6QX zc3C&3L2;PDJ%p9xFs$(G9NjaCy6 zx*;k)FJ%2Mq}Gp`>d80yu6#K4yxMXNqU`~1IvXUx`|jn4v)pCRhU308kvqJ@?mc_W zw-te}-b+5>gppZ?+X$ua8AAm8RkzTmc=Qv+4wcY?1xmo+Tro|7r)ZMDv}#ADdzrda zqih?j-y_np6dR8xqUN_Hvj^8>E>3`K$!9$=I)366rm-eFi!RZ#jBgQ8?{VJ&B&uOP z&wD_qVG`v!&MrCz!kyZ@S==NTc|%NEMvg~pw?(g!!}va{3D}uzSW{Fu1hxPcW@i%s ze%IA!L%)y+JKt^suMIN$z!^nP@Y5n!;Ph zbQ%)v2y?|j&RI3_jJdrWvr3tZVsNdT?=s!w3#CZ*tUnD1KJxrMjaCG zscF>sVg7qfh6_W(SihSb#`7`73$kx2;~l#_(1qb6B?Y+wboO#n-Kmc1X=j4zB^+iX z(0c_gij5XQWT3*yPpkRH+!~x&MQ_~B0EN8dCph+wU;N7zFYDx78A{ex4rlOpe+Zsi zl3g2);J!D@+$T+lG7bgD-)~#Mco0@iiB5HQ6%5Fg5cr-K5+1T3kP+-$Y9dtj;iTN0 zUQW7xEHmC?0dgGeBAFXV?C!F9k}`@IxE85JJl91zCQNLIF7CVkOOI_!mOH7w}-$C0@O`fpH;=rdH{0)OtNy-&x_&t?E33=OMXv?Xc_m)})7 zV_!227Hy}7hFTm5xlR+&ylgXAZW(GJGS!>x=(BRh^~>?DO}~%X#2_NRcGf#1`OZe2 zT=o-+>4t>;wH9-HX*bCix4Vq@jfFI01{-b5(Ohj{dOm(syC1Pt^tbh+>B)>kRt^7q2Au-hIANoFa)?5O8$82Jg-zvkZC zHR?Qc%*X|JE#{EX7Y8%-bzHaJh9296nH_x?4?S5{6L^?DDR@6xfkV1WLw9|5mE_8K z@nSaeLdTT!n*}*%cs6Bf*@vCl6hU?TzPQ4E9I_0eS|hyn%~X8*b@MbSWa{G1{eU;y zp_5qUj5|kZG-t3gRFVNBd7wX@O8&jJ$9sw=iTM{cT$iWn(ucRKQ=J`w z&IC*r#42t@D>QwUzqPwLC=OUDUtP0Wb{5WGFWSl?E(LG2X+-*utLu8DuE=NlRprf< zsx*Gp!Ah2ulPHj1W(#@O`EH9H3K?tJAu8;P?_Z5~f2wtYe7mv{_wo|XGK|ZhZr^A0 zjp8=X`VAONGN*KS!dj|0;zYwc5^n*-K^3qb(B)TDaFlBHz;9=|$FZOFu0a;@8>wgL z;-6Xd+K*RV{)3>CAnd^{C-V$R zj5+FV<VZxf4eVydur~IF)v-%c$*Bh3-Hx5^&x(lS)k0lJ(imwRcuPt+T0I;t_V60a!{d zq{R5a8(Qa?g4oo{^Xh@-SwsrAW-6UDD64UJrT_8+9=TNz`+>BOl};|uFEg|2@YlTH z?CNq(9;HEe>Vuff7?XxO-V%!fDSgT|lsFpazg6-n@5Fg((*&)+@jEe1k)@p-I6ketrqZJ6d!^+^p$% zR{}-LAUQqdNwVQ^NcAN_c$sGXovp`)H8x0lODyYHjWg$VR4HyVICw446L1o7%rcQD z9bD?|?^m^d$t2gGXi=0mcd!p2Ux>+M)t8stUd%7L)Juj@gl2F!NPU<*<67iXCPXvl z2QcX_!}*O@^0!=NLMhi9Z4B`?ndxczC5EhB+8RA$(HOspN%~#DOvgRuRwa(_sD8~C zDi0cX-FOAE*Yp@%Ndxz`UIe#H=G_YLCw!5=U6Kmbo%B72M8Qk7=DQi7YfiByWHpN~ zwVnV*O|yVJOy}@SOULz>z)$lv^Ndmp3^j&(GtuBpJ+`y;9!~6@n%aH2Fj{NQ6=0XW zJX{m1klZig4P|p5-h2f}(Wo}c%u*v{&rUbr^N9e0pL3s1a3O)E-=l|LWV@4p%Xntp zuJFEM??vMCLT*1ows#SBOJ)7%3HPA+diE`u@p^6Ls@E552dmD<0|^MU?IBj=0sR)W zoV+4B1co6IJ*i8~wq=UtPO)+*%<{dLZH0gbe0h>CZ&Hq|gs02HlD;mB;tX_pTY{!- zcE)q+?H$koJ0X>DwB|EpS@TwFmvOn&hxXaG8Use?GB&0W@=yi7_ihSPG_oGGFvNb$ToZVyAkpi@gp$auH2m!cY&}69TCm# zOO^7un>#RxVt)usv$+z(DeEsfbOWxZ7d8ic>+FXb9zag8E!u?c;gA6lNdS^=O(jVd zgzC-vGvB~`@=h!f#fzlvhVd(`9za%XpJ@6?lZU+X_VcA3-8g6@`kB~H@IJcbzQ6M0 z(v){~?3oI7p@fORz_ZfcXXaHst2;AXeuEb!2W zL-WXGB2V0m?K=sf4ep63qi=(eH@=;au zJrt-$IpW)=yeoEgn=2JL-DYS;>-Qs^lk1X@yGlvJ^IgU>oHvX)+)5p>&xxGblp^pH zw;Sdh!0D`*ux=*bz=pp#zhRZ>NecQmMBI!-?mbbHOR27XBwuz6vY?zH2KROEHEcy_ zj8cled-kQS>pd;6rAF$+inhG|Y&Na}fO8QO&u zx~LCF#}QA(sPK@`l8bfkrhwDfX}zq1j!Qq_2%m=L$v=5py^b9mT(9u<{DZSG`4~x3#(eAPwPS!LJyiAL@feZ#RlF>qMUDA{!4$v4 zs5Le>TJD{3bt>jL2=*JKsmH|!)AM6GJxv%PPm+$*5t`kA$6BWOA;8ri_e89sJF zST?wNIVT}_J!_|F=WffHG7?LEkR(EL6~{QZY6xWP?*#eE%6z*V#oVoAZLtg3zl}u} zUg|A2g3@uJN~ca5>G7tWN67S!cIgEX(j8eW$Vs~#HQ6j@!0Ok--2vR{lE#STBZ%Aw zmRK1V>Nms?e#FKzTc?-J>=}8+L(`V__Jyb~?dHmnG5{D^j;1W@d``H#9C=#z@?8td zZ^vW@MorEAm`g;b_H`v!SDJ?~mi7+r<`#BB&zWCOVLVTv=#i4H_u?03rSB-XXP-8> z8abUqHzTe6$REJfq{~|6Bg!~K1*4(RTiHB%!Lq|^yQ^WgPU1yVqfUL+WS+_SSFN`b ziv=(BnL<>~7`m4AP43?Bxxl3g^2U5lA=j?=kwCwcaPcY$E7@;C&u0I<_EmR z(BT^S{+?a;bXYWnFi7=$%>dG-mbApZH+t|bkjkTE!X9Ujnf1chnuqyEX?=k|t2iy(5UvsA z_00Z5=B4b>)@oY9UWrorA7=0_6KMt`J>e<-i=zVFa@=>A;L}>Bn%M|xOc)OMXh&xr z`_7;^dPj8x855`c&iAcAK26eUr+-N|-)wmw=cLkFOd2)&1wbM>i)rG~Ei9eqEsP5B z+Y*zwZJ{8}xHGtDez^Cr4ytS>&LtY&d~eSuQb&UtTkDPlX#YJxGA3;*+uRm4p(7o;GkA1iPc6j<4H#+mQE=(^!EA;T~NO1;q z7pDI~mXSVL9yrI~$+74f`!okn>iNQaCW<+coTQY|3DY`9cY(DyXDg6ut_W;SVZ3gb zb+#(6E%^0}@_X~x5w%ot#|r0{D?yYmj7H>RN(p4$K*lQ4|EV4|nP}S+fK$ z)(aphIr_KyzY-APN)(y#>EgZTOAi=Wd#mfRk6Kh)LR9pS<-*uX{~%Spq3K3UK;8xX zL)0-x#So7bANc3giDdzHJ&Eic4p>MnGlj@$0L0KJLxTsA0X{M`BNrS3C4mt`R9-5^ zLYxTx9c#4{E#|H}hQmGFv&6C~gUdwSZci+g zP*0h#QPs<@G6J_@+T(?{veiscQY#dw_5@Fn_SwlxoR&SrMYor1GF-?Hq^jcMt`;P7KAY~YgtVz^5WQs zbgOu(oXW^VuM49W=2d&6XOF(Un)!M}tRVf6W4914 zXv^d=CH?D6&lm9}cYP`53&@&FHJJCubPis@ml`EyV^`}+rt!7rMn(=pJa4AL@A(Q* z&A_1$oLqc8iVJ~SDV^AH8h9ui zmOoU4>f_=?89U9RQLmUt>JsLrt#{!Z!;(j8&{`=amD}$VCUMQG&s?|BUgky9!1bj7 zTwfmnuixfLpi(Ac1YG)vjsE7&D1`T$-pQzppY8R^CO(o0t4L35(^Cwr00{SnRH5y& zhBqnFj&r8U@(s-Ss?CN5I2K!Dwvf7U5DZ(~)%do-o!Je$12UFj8(+Ks9vk=v_PyXo zJuiDpazkpXk}uH@zN_;6jslP!x+ww;G~X;a6(jx2P3gB_$zMI9ccCuSV4Bu=VEbf^ zzQcZhEZ}#K%zyS*hzUqCBCS6u6m{(H6!Y76-<}2Gq@GK3({8`>5&yG>{8qwu&p7zn ze;Ws`et&8B{yTF29l3u$0{>mDe`#@nHTmC>`~PO-3SvbLce?*MSB-5Z?R0Y1j81oM{zC7edLXk_Hw`bxcRkR*`i8$*@P_~iWM{Vgx4Crxqe^&4sn=}6 zo<{xy{RBgO6)4|DKK@3#{`&>SFoP3M)vLGQoc`}O+<*T_Ht4C6;)&k_=>Gjz{`ZRC zlF~yNT3Y>j$@sgSg@K;t{1owBCh>1gJ1a%tSSyyO{3S`>-(6lI1)!&}qi_C+QzzM> zN|LON-icAFe`aOrfSv+w?B5!$@9)m=XhCdzN%=oCYz9;(<$Q{@I@}3Mqt{DtrK&td zMXjXRnZK}u{OcwbGfloW-y7Y{HuSjrsTGX4GJF*LM{V_(S%(e;t$mON(DCgITlL#CnqYpXh3gKG;|0hSBfd{rzDB zz4&^x;8@0$Pb&ZW*7JYawuD8#d<%5>REvhGYLPyguAOQc7$y-mX&GD$yVXU?Xc|3lKq?}dXzr2P%WnmpLi zsu%U{EGPMXR8%W1sa8~x{uh2&XaTJCAlZ*zC8KeOovB`RsR-h*S=eC~P^s>x($Q}z zcfVKn+xOy`9*$UtksE5Hixv*OG5p4d{czwh}A^Y&lMd`m}ap<4ly~K!3?E3bHR3t&gLd}NXig>W zmgF>x*VB3aXf}QpiYaa`)5GeCmCAlLZ6pjxVx@(@5yk)fbb&pYwS)GAZuM#S^y4$y zg42&g3veQFz8SwL4gBkb_P7E1{CG+nvm*?-N&N%s6>^&@7mjQE)EO_LkR~c?O@H#| zJKg;dhYv477>`UOhl~^+)1T8W4gGOjJVKtZiYHgu-spZ8%Qt7TxaCMdWnD4W9%>qu z9;Edn#|MV(u6VBeU__cI3UKdti2!T&lpOijM*q^|0`oI@k8_{`gj56J2KfBK z2ihM@`vc^LjCSa}c>7_vA`2oeI3=@Le1B5Ahw4wqTGD;UmkhJ>tfg&egRj;Sq(u3{ z+y7<8LoblfZk`z3@MKX>;|m(0Io}k={?tH0D>DC^C7ToZ&1E-v6aMt!w|^goAHWAP zw=fKC=sLY-`k(cHy@~PR$Oc$h-Va&Nm`Eu*+11Bov*#H0l literal 0 HcmV?d00001 From 0eb4e65da6f41003d1cd6306cbb5d58669cb5ca0 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:55:43 -0600 Subject: [PATCH 510/590] Rename ScreenShot-2025-11-10-125411@2x.png to cooldownHD.png --- ...Shot-2025-11-10-125411@2x.png => cooldownHD.png} | Bin 1 file changed, 0 insertions(+), 0 deletions(-) rename ALCF/notes/assets/{ScreenShot-2025-11-10-125411@2x.png => cooldownHD.png} (100%) diff --git a/ALCF/notes/assets/ScreenShot-2025-11-10-125411@2x.png b/ALCF/notes/assets/cooldownHD.png similarity index 100% rename from ALCF/notes/assets/ScreenShot-2025-11-10-125411@2x.png rename to ALCF/notes/assets/cooldownHD.png From 41d5c07f5038429fc3f1fefa81c283177f538690 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:56:01 -0600 Subject: [PATCH 511/590] Replace cooldown image with high-definition version Updated image reference for cooldown documentation. --- ALCF/notes/cooldown.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cooldown.md b/ALCF/notes/cooldown.md index 7fbb800cd2b..966bbf7852d 100644 --- a/ALCF/notes/cooldown.md +++ b/ALCF/notes/cooldown.md @@ -2,7 +2,7 @@ - Cooled down over last $10\%$: - [volcanic-blaze-4312](https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu/overview?nw=nwuserforemans) - ![](./assets/cooldown.png) + ![](./assets/cooldownHD.png) ## ♻️ Convert to Universal From 8da9b0d078b48a0a39ed0f78d6c80f7deac79e4a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:57:40 -0600 Subject: [PATCH 512/590] Enhance cooldown.md with new title and example details Updated the section title and added details to the example. --- ALCF/notes/cooldown.md | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/cooldown.md b/ALCF/notes/cooldown.md index 966bbf7852d..80b8ec3275f 100644 --- a/ALCF/notes/cooldown.md +++ b/ALCF/notes/cooldown.md @@ -12,7 +12,7 @@ TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1 python3 ALCF/ds_to_universal.py \ --output_folder test_rollback/global_step136000_universal ``` -## Cooling Down +## ☃️ Cooling Down - 256 Nodes: @@ -31,7 +31,7 @@ TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1 python3 ALCF/ds_to_universal.py \ --lr_constant_plus_cooldown_frac=0.9 ``` -Example: +
    Example: ```bash #[🐍 aurora_frameworks-2025.2.0](👻 Megatron-DeepSpeed-aurora_frameworks-2025.2.0) @@ -228,3 +228,6 @@ Application 5d196584 resources: utime=21540179s stime=2039451s maxrss=46631076KB [2025-11-10 12:30:19,247825][I][ezpz/launch:403:launch] Executing finished in 7795.27 seconds. [2025-11-10 12:30:19,248203][I][ezpz/launch:404:launch] Took 7795.28 seconds to run. Exiting. took: 2h 11m 1s +``` + +
    From b654baacf5668d4be86d675404b27860b22d60e5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 10 Nov 2025 12:59:23 -0600 Subject: [PATCH 513/590] Fix markdown formatting in cooldown.md --- ALCF/notes/cooldown.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cooldown.md b/ALCF/notes/cooldown.md index 80b8ec3275f..4c9d0d774c0 100644 --- a/ALCF/notes/cooldown.md +++ b/ALCF/notes/cooldown.md @@ -1,6 +1,6 @@ # 🧊 Cooling Down AuroraGPT-2B Checkpoints -- Cooled down over last $10\%$: +- Cooled down over last 10\%: - [volcanic-blaze-4312](https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu/overview?nw=nwuserforemans) ![](./assets/cooldownHD.png) From 8db699ac2b870587c48afa39a33cf50168f710be Mon Sep 17 00:00:00 2001 From: "Nathan S. Nichols" <8723338+nscottnichols@users.noreply.github.com> Date: Mon, 10 Nov 2025 23:29:02 +0000 Subject: [PATCH 514/590] initial commit --- .gitignore | 2 + tools/cooldown_generator/README.md | 325 ++++++++++++++++++ .../build_checkpoints_from_tokens.py | 46 +++ .../cooldown_generator/gen_cooldown_sweep.sh | 122 +++++++ .../cooldown_generator/make_cooldown_cmds.py | 131 +++++++ 5 files changed, 626 insertions(+) create mode 100644 tools/cooldown_generator/README.md create mode 100755 tools/cooldown_generator/build_checkpoints_from_tokens.py create mode 100755 tools/cooldown_generator/gen_cooldown_sweep.sh create mode 100755 tools/cooldown_generator/make_cooldown_cmds.py diff --git a/.gitignore b/.gitignore index 5079fb8bf2c..b668a703f7c 100644 --- a/.gitignore +++ b/.gitignore @@ -18,6 +18,8 @@ ALCF/OUTPUTS/* *tmp* *core.* *old* +!tools/cooldown_generator/ +!tools/cooldown_generator/** *.bak **index-cache** **pbslogs** diff --git a/tools/cooldown_generator/README.md b/tools/cooldown_generator/README.md new file mode 100644 index 00000000000..e78ba361470 --- /dev/null +++ b/tools/cooldown_generator/README.md @@ -0,0 +1,325 @@ +# `make_cooldown_cmds.py` + +Generate **ready-to-run** Megatron-DeepSpeed commands to *cool down* a training run **starting exactly at a given checkpoint iteration**. + +Given: + +* a checkpoint iteration **S** (the `global_step` you resume from), and +* a cooldown length **R** (steps to spend decaying LR), + +the script emits commands that set: + +* `TRAIN_ITERS = T = S + R` +* `--lr_constant_plus_cooldown_frac = f = S / T` + +So the **constant LR** phase ends at the resume step, and the **cooldown** covers the remaining `R` steps. + +--- + +## What it prints + +For each `(S, R)` pair, the script prints a small annotated block: + +``` +# id= resume_step= cooldown_steps= total_iters= frac= +LR_DECAY_STYLE=constant \ +OPT=ipex.fusedlamb \ +OVERRIDE_CKPT_OPT_PARAM=1 \ +TRAIN_ITERS= \ +GRAD_ACC_STEPS=2 \ +LOAD=<...> \ +DATA_FILE_LIST=<...> \ +bash train_alcf.sh \ + --override-opt_param-scheduler \ + --min-lr=2e-5 \ + --lr_constant_plus_cooldown \ + --lr_constant_plus_cooldown_frac= \ + [any extra args...] +``` + +You can copy/paste the printed commands, or write all of them to a single `.sh` via `--emit-sh`. + +--- + +## Requirements + +* Python 3.7+ +* Your training wrapper (default: `train_alcf.sh`) accepts the same environment variables/flags as shown above. +* The checkpoint parent directory (`--load`) is the path you normally pass to Megatron/DeepSpeed for resuming. + +--- + +## Basic usage + +### Single checkpoint, single cooldown + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/olmo-mix-1124.txt \ + -S 72500 \ + -R 2000 +``` + +### Multiple checkpoints, one cooldown value + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/olmo-mix-1124.txt \ + -S 12900 32800 52650 72500 \ + -R 2000 +``` + +### Multiple checkpoints × multiple cooldowns (grid) + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + -S 92400 112250 132150 \ + -R 1000 2000 5000 \ + --emit-sh cooldown_grid.sh +# => prints to stdout and writes all commands to cooldown_grid.sh +``` + +### Using explicit ID tags (optional) + +IDs label the comment header above each command; they’re handy for grouping in dashboards or logs. + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/olmo-mix-1124.txt \ + --checkpoint-ids 1 2 3 4 \ + -S 12900 32800 52650 72500 \ + -R 2000 +``` + +### Using `--pairs` (compact) + +`--pairs` accepts `S:R` or `id:S:R` entries; when `id:` is omitted, IDs auto-increment. + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + --pairs 92400:2000 112250:5000 7:132150:2000 +``` + +--- + +## Common flags + +* `--load` *(required)*: Parent directory that contains `global_stepXXXXX/` checkpoints. +* `--data-file-list` *(required)*: The data list file your wrapper expects. +* `--train-script` (default: `train_alcf.sh`) +* `--grad-acc-steps` (default: `2`) +* `--opt` (default: `ipex.fusedlamb`) +* `--min-lr` (default: `2e-5`) +* `--no-override-ckpt-opt` (use if you **do not** want `OVERRIDE_CKPT_OPT_PARAM=1`) +* `--extra-args "..."` (anything you want appended to the train command; e.g., W&B tags) +* `--emit-sh ` (write all printed commands to a runnable shell script) + +--- + +## Why this formulation? + +Megatron-DeepSpeed’s constant-plus-cooldown helper treats `--lr_constant_plus_cooldown_frac=f` as the fraction of **total** training reserved for the constant phase. By setting: + +* `T = S + R` and `f = S/T`, + we guarantee the **cooldown starts exactly at resume** and lasts `R` steps—no ambiguity about “10% of total vs. remaining.” + +--- + +## Quick sanity checks + +* If you resume at **S** and choose **R=2000**, you should see: + + * `TRAIN_ITERS = S + 2000` + * `lr_constant_plus_cooldown_frac ≈ S / (S + 2000)` + (It will be a high fraction like `0.98…` if you resume late.) + +* If you accidentally swap the meanings (e.g., set `f=0.9` without adjusting `T`), you’ll be scheduling “last 10% of **total** job” rather than “R steps after resume.” This script avoids that pitfall. + +--- + +## Examples with real numbers + +Assume: + +* tokens/step = `8192 * 6144 = 50,331,648` +* 5% rollback from final ~7T run corresponds to ≈ *~7k steps* earlier. + +**ID 5** (≈ 5T tokens): `S=92400`, `R=2000` + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + --pairs 5:92400:2000 +``` + +**ID 5 rollback**: `S=~(92400 - ~7000) ≈ 85400` (rounded per your table) + +```bash +./make_cooldown_cmds.py \ + --load /proj/checkpoints_parent \ + --data-file-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + --pairs 5:85400:2000 +``` + +--- + +## Troubleshooting + +* **“Provide either --pairs OR both --checkpoint-iters and --cooldown-steps.”** + You must supply `(S, R)` pairs. Use `--pairs` *or* `-S ... -R ...`. + +* **Nothing about the data list changes across IDs.** + This script is intentionally **generic**. If your data list changes by phase/ID, handle that in a wrapper (e.g., `run_cooldown_per_id_split.sh`) and pass the correct `--data-file-list` for each call. + +* **Scheduler doesn’t seem to take effect.** + Ensure `--override-opt_param-scheduler` is present (it is) and `OVERRIDE_CKPT_OPT_PARAM=1` isn’t disabled unless you need to. + +--- + +# Cooldown Wrapper Scripts + +These helper scripts automate checkpoint enumeration and cooldown‐command generation using the generic [`make_cooldown_cmds.py`](./make_cooldown_cmds.py). + +--- + +## 1. `build_checkpoints_from_tokens.py` + +### Purpose + +Converts **token milestones** (0 T → 7 T) into **training iterations** and computes a “rollback” checkpoint offset by a percentage of the final total run. + +### How it works + +1. Each iteration processes `8192 × 6144 = 50 331 648 tokens`. +2. For every integer trillion token milestone (1 T → 7 T) it computes: + + * `steps_mod` = rounded step count at that token milestone + * `steps_rollback` = `steps_mod – (cooldown_percent × final_steps)` + (rounded to nearest multiple of `--round`) +3. Writes a tab-separated file: + +``` +id steps_mod steps_rollback +1 12900 12900 +2 32800 32800 +... +7 132150 132150 +``` + +### CLI + +```bash +python build_checkpoints_from_tokens.py \ + --ttokens 8 \ + --tokens-per-step $((8192*6144)) \ + --cooldown-percent 0.05 \ + --round 50 \ + --out checkpoints.tsv +``` + +**Arguments** + +| Flag | Default | Description | +| -------------------- | ----------------- | ------------------------------------------------- | +| `--ttokens` | `8` | Number of trillion-token milestones (0..N). | +| `--tokens-per-step` | `8192*6144` | Tokens processed per optimizer step. | +| `--cooldown-percent` | `0.05` | Fraction of final total used for rollback offset. | +| `--round` | `50` | Round step counts to nearest N. | +| `--out` | `checkpoints.tsv` | Output TSV file. | + +### Output + +Creates `checkpoints.tsv` for use by the generator script. + +--- + +## 2. `gen_cooldown_sweep.sh` + +### Purpose + +Automates cooldown job generation: + +* Reads `checkpoints.tsv` from the step builder. +* Creates **one `.sh` per checkpoint ID** for both *exact* and *rollback* resume points. +* Uses phase-based data lists: + + * IDs 1–4 → `olmo-mix-1124.txt` + * IDs 5–7 → `aurora/dolmino-mix-1124-fused-file-list.txt` + +### What it produces + +``` +cooldown_out/ + cooldown_id1_exact.sh + cooldown_id1_rollback.sh + cooldown_id2_exact.sh + ... + cooldown_id7_exact.sh + cooldown_id7_rollback.sh +``` + +Each file contains **one** fully-formed Megatron command block generated via `make_cooldown_cmds.py`. + +### CLI + +```bash +./gen_cooldown_sweep.sh \ + --load /proj/checkpoints_parent \ + --cool-R 2000 \ + --emit-dir ./cooldown_out \ + --phase1-list ALCF/data-lists/olmo-mix-1124.txt \ + --phase2-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + --train-script train_alcf.sh \ + --extra-args "--wandb-tag cooldown --wandb-note sweep_remaining" +``` + +### Key options + +| Flag | Meaning | +| ----------------------------------------------------------------- | -------------------------------------------------------------- | +| `--load` | Parent directory containing checkpoints (`global_stepXXXXX/`). | +| `--cool-R` | Cooldown length in steps (`R`). | +| `--emit-dir` | Output directory for generated scripts. | +| `--phase1-list`, `--phase2-list` | Data lists for phase 1 (IDs 1–4) and phase 2 (IDs 5–7). | +| `--train-script` | Training wrapper (`train_alcf.sh` by default). | +| `--extra-args` | Extra flags to append (e.g. W&B tags). | +| `--tokens-per-step`, `--ttokens`, `--cooldown-percent`, `--round` | Passed to `build_checkpoints_from_tokens.py`. | +| `--python` | Python interpreter (default: `python`). | + +### Output behavior + +* Calls `build_checkpoints_from_tokens.py` internally. +* For each ID: + + * Uses the proper data list by phase. + * Generates `cooldown_id_exact.sh` (always) and `cooldown_id_rollback.sh` (if rollback > 0). +* Each script is executable and self-contained. + +--- + +### Typical workflow + +1. **Enumerate checkpoints** + + ```bash + python build_checkpoints_from_tokens.py --out checkpoints.tsv + ``` +2. **Generate per-checkpoint cooldown scripts** + + ```bash + ./gen_cooldown_sweep.sh --load /proj/checkpoints_parent --cool-R 2000 + ``` +3. **Submit or batch-run** any of the emitted scripts on your training cluster. + +--- + +These two wrappers + [`make_cooldown_cmds.py`](./make_cooldown_cmds.py) together form a reproducible, parameterized pipeline for generating and managing cooldown experiments from multi-trillion-token Megatron-DeepSpeed runs. + diff --git a/tools/cooldown_generator/build_checkpoints_from_tokens.py b/tools/cooldown_generator/build_checkpoints_from_tokens.py new file mode 100755 index 00000000000..bc6e8c4611b --- /dev/null +++ b/tools/cooldown_generator/build_checkpoints_from_tokens.py @@ -0,0 +1,46 @@ +#!/usr/bin/env python3 +""" +Compute training iterations (steps) for checkpoints at 1..(ttokens-1) Trillion tokens +and their "rollback" checkpoints offset by a cooldown percentage of the FINAL total. + +Output TSV columns: + id\tsteps_mod\tsteps_rollback + +Where: + steps_mod = rounded steps at exactly i*T tokens + steps_rollback = rounded steps at (steps_mod - cooldown_iters) using cooldown_iters=percent*steps_at_(ttokens-1) +""" +import argparse + +def main(): + p = argparse.ArgumentParser() + p.add_argument("--ttokens", type=int, default=8, help="Total token milestones (default: 8 for 0..7T).") + p.add_argument("--tokens-per-step", type=int, default=8192*6144, help="Tokens per optimizer step (default: 8192*6144).") + p.add_argument("--cooldown-percent", type=float, default=0.05, help="Percent of final run used for rollback offset (default: 0.05).") + p.add_argument("--round", type=int, default=50, help="Round steps to nearest N (default: 50).") + p.add_argument("--out", type=str, default="checkpoints.tsv", help="Output TSV path (default: checkpoints.tsv).") + args = p.parse_args() + + ttokens = args.ttokens + tps = args.tokens_per_step + r = args.round + c = args.cooldown_percent + + # Steps at each i*T (i in 0..ttokens-1) + runs = {i: (i * 10**12) / tps for i in range(ttokens)} + runs_mod = {k: int(round(v / r) * r) for k, v in runs.items()} + + # % of the FINAL (ttokens-1) step count, then rounded rollback + cooldown_iters = int(c * runs[ttokens - 1]) + runs_rollback = {k: int(round((v - cooldown_iters) / r) * r) for k, v in runs_mod.items()} + + with open(args.out, "w") as f: + f.write("id\tsteps_mod\tsteps_rollback\n") + for k in range(1, ttokens): # emit 1..(ttokens-1) + f.write(f"{k}\t{runs_mod[k]}\t{runs_rollback[k]}\n") + + print(f"Wrote TSV to {args.out}") + +if __name__ == "__main__": + main() + diff --git a/tools/cooldown_generator/gen_cooldown_sweep.sh b/tools/cooldown_generator/gen_cooldown_sweep.sh new file mode 100755 index 00000000000..0aaf5075cdc --- /dev/null +++ b/tools/cooldown_generator/gen_cooldown_sweep.sh @@ -0,0 +1,122 @@ +#!/usr/bin/env bash +set -euo pipefail + +# Emit ONE command per file: +# cooldown_id_exact.sh +# cooldown_id_rollback.sh (only if rollback > 0) +# +# ID -> data list mapping: +# IDs 1..4 -> olmo-mix-1124.txt +# IDs 5..7 -> aurora/dolmino-mix-1124-fused-file-list.txt +# +# Example: +# ./run_cooldown_per_id_split.sh \ +# --load /path/to/checkpoints_parent \ +# --cool-R 2000 \ +# --emit-dir ./cooldown_out \ +# --phase1-list ALCF/data-lists/olmo-mix-1124.txt \ +# --phase2-list ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ +# --train-script train_alcf.sh \ +# --extra-args "--wandb-tag cooldown --wandb-note per_id_split" +# +# Optional knobs: +# --tokens-per-step 50331648 # 8192*6144 +# --ttokens 8 # produce 1..7T +# --cooldown-percent 0.05 +# --round 50 +# --python python3 + +# Defaults +EMIT_DIR="${PWD}/cooldown_out" +COOL_R="" +LOAD_PATH="" +PHASE1_LIST="" +PHASE2_LIST="" +TRAIN_SCRIPT="train_alcf.sh" +EXTRA_ARGS="" +TOKENS_PER_STEP=$((8192*6144)) +T_TOTAL=8 +COOLDOWN_PCT="0.05" +ROUND_TO="50" +PYTHON="${PYTHON:-python}" + +die() { echo "ERROR: $*" >&2; exit 1; } + +while [[ $# -gt 0 ]]; do + case "$1" in + --emit-dir) EMIT_DIR="$2"; shift 2 ;; + --cool-R) COOL_R="$2"; shift 2 ;; + --load) LOAD_PATH="$2"; shift 2 ;; + --phase1-list) PHASE1_LIST="$2"; shift 2 ;; + --phase2-list) PHASE2_LIST="$2"; shift 2 ;; + --train-script) TRAIN_SCRIPT="$2"; shift 2 ;; + --extra-args) EXTRA_ARGS="$2"; shift 2 ;; + --tokens-per-step) TOKENS_PER_STEP="$2"; shift 2 ;; + --ttokens) T_TOTAL="$2"; shift 2 ;; + --cooldown-percent) COOLDOWN_PCT="$2"; shift 2 ;; + --round) ROUND_TO="$2"; shift 2 ;; + --python) PYTHON="$2"; shift 2 ;; + *) die "Unknown arg: $1" ;; + esac +done + +[[ -n "${COOL_R}" ]] || die "--cool-R (cooldown steps) is required" +[[ -n "${LOAD_PATH}" ]] || die "--load path is required" +[[ -n "${PHASE1_LIST}" ]] || die "--phase1-list is required (IDs 1..4)" +[[ -n "${PHASE2_LIST}" ]] || die "--phase2-list is required (IDs 5..7)" + +mkdir -p "${EMIT_DIR}" + +# 1) Build the checkpoint table (1..7T + rollback) +${PYTHON} build_checkpoints_from_tokens.py \ + --ttokens "${T_TOTAL}" \ + --tokens-per-step "${TOKENS_PER_STEP}" \ + --cooldown-percent "${COOLDOWN_PCT}" \ + --round "${ROUND_TO}" \ + --out "${EMIT_DIR}/checkpoints.tsv" + +# 2) For each row, emit two single-command files (exact + rollback if > 0) +tail -n +2 "${EMIT_DIR}/checkpoints.tsv" | while IFS=$'\t' read -r id smod srb; do + # choose data list by ID range + if (( id >= 1 && id <= 4 )); then + DATA_LIST="${PHASE1_LIST}" + elif (( id >= 5 && id <= 7 )); then + DATA_LIST="${PHASE2_LIST}" + else + echo "Skipping unknown id ${id}" >&2 + continue + fi + + # exact-T + OUT_EX="${EMIT_DIR}/cooldown_id${id}_exact.sh" + ${PYTHON} make_cooldown_cmds.py \ + --load "${LOAD_PATH}" \ + --data-file-list "${DATA_LIST}" \ + --train-script "${TRAIN_SCRIPT}" \ + --checkpoint-ids "${id}" \ + --checkpoint-iters "${smod}" \ + --cooldown-steps "${COOL_R}" \ + --extra-args "${EXTRA_ARGS}" \ + --emit-sh "${OUT_EX}" + chmod +x "${OUT_EX}" + echo "Wrote ${OUT_EX}" + + # rollback (only if positive) + if (( srb > 0 )); then + OUT_RB="${EMIT_DIR}/cooldown_id${id}_rollback.sh" + ${PYTHON} make_cooldown_cmds.py \ + --load "${LOAD_PATH}" \ + --data-file-list "${DATA_LIST}" \ + --train-script "${TRAIN_SCRIPT}" \ + --checkpoint-ids "${id}" \ + --checkpoint-iters "${srb}" \ + --cooldown-steps "${COOL_R}" \ + --extra-args "${EXTRA_ARGS}" \ + --emit-sh "${OUT_RB}" + chmod +x "${OUT_RB}" + echo "Wrote ${OUT_RB}" + fi +done + +echo "Per-ID single-command scripts written to: ${EMIT_DIR}" + diff --git a/tools/cooldown_generator/make_cooldown_cmds.py b/tools/cooldown_generator/make_cooldown_cmds.py new file mode 100755 index 00000000000..79756595795 --- /dev/null +++ b/tools/cooldown_generator/make_cooldown_cmds.py @@ -0,0 +1,131 @@ +#!/usr/bin/env python3 +import argparse +from pathlib import Path +from textwrap import dedent + +def fmt_float(x: float) -> str: + return f"{x:.8f}".rstrip("0").rstrip(".") + +def build_command( + load_path: str, + data_file_list: str, + train_script: str, + train_iters: int, + lr_cooldown_frac: float, + grad_acc_steps: int, + opt: str, + min_lr: float, + override_ckpt_opt_param: bool, + extra_args: str, +) -> str: + env_lines = [ + "LR_DECAY_STYLE=constant", + f"OPT={opt}", + "OVERRIDE_CKPT_OPT_PARAM=1" if override_ckpt_opt_param else "", + f"TRAIN_ITERS={train_iters}", + f"GRAD_ACC_STEPS={grad_acc_steps}", + f"LOAD={load_path}", + f"DATA_FILE_LIST={data_file_list}", + ] + env_block = " \\\n".join([l for l in env_lines if l]) + + extra_line = "" + if extra_args: + extra_line = f" \\\n {extra_args}" + + cmd = dedent(f"""\ + {env_block} \\ + bash {train_script} \\ + --override-opt_param-scheduler \\ + --min-lr={min_lr} \\ + --lr_constant_plus_cooldown \\ + --lr_constant_plus_cooldown_frac={fmt_float(lr_cooldown_frac)}{extra_line} + """).strip() + return cmd + +def parse_pairs(pairs_args): + records = [] + next_id = 1 + for item in pairs_args: + parts = item.split(":") + if len(parts) == 2: + S = int(parts[0]); R = int(parts[1]); cid = next_id; next_id += 1 + elif len(parts) == 3: + cid = int(parts[0]); S = int(parts[1]); R = int(parts[2]) + else: + raise SystemExit(f"Bad --pairs entry: {item}") + if S <= 0 or R <= 0: + raise SystemExit(f"Non-positive S/R in --pairs entry: {item}") + records.append({"id": cid, "S": S, "R": R}) + return records + +def main(): + p = argparse.ArgumentParser( + description="Emit Megatron-DeepSpeed cooldown commands so LR cooldown starts at resume.\n" + "Provide checkpoint iteration(s) S and cooldown step(s) R.\n" + "For each pair, sets TRAIN_ITERS T=S+R and lr_constant_plus_cooldown_frac f=S/T." + ) + p.add_argument("--load", required=True) + p.add_argument("--data-file-list", required=True) + p.add_argument("--train-script", default="train_alcf.sh") + p.add_argument("--grad-acc-steps", type=int, default=2) + p.add_argument("--opt", default="ipex.fusedlamb") + p.add_argument("--min-lr", type=float, default=2e-5) + p.add_argument("--no-override-ckpt-opt", action="store_true") + p.add_argument("--extra-args", default="") + p.add_argument("--emit-sh", type=Path, default=None) + + p.add_argument("--checkpoint-iters", "-S", type=int, nargs="+") + p.add_argument("--cooldown-steps", "-R", type=int, nargs="+") + p.add_argument("--checkpoint-ids", type=int, nargs="+") + p.add_argument("--pairs", type=str, nargs="*") + + args = p.parse_args() + override_flag = not args.no_override_ckpt_opt + + if args.pairs: + records = parse_pairs(args.pairs) + else: + if not args.checkpoint_iters or not args.cooldown_steps: + raise SystemExit("Provide either --pairs OR both --checkpoint-iters and --cooldown-steps.") + ids = args.checkpoint_ids or list(range(1, len(args.checkpoint_iters) + 1)) + if len(ids) != len(args.checkpoint_iters): + raise SystemExit("--checkpoint-ids must match length of --checkpoint-iters.") + records = [{"id": cid, "S": int(S), "R": int(R)} + for cid, S in zip(ids, args.checkpoint_iters) + for R in args.cooldown_steps] + + lines = [] + header = "# Auto-generated cooldown commands\nset -euo pipefail\n\n" + if args.emit_sh: + lines.append(header) + + for rec in records: + cid, S, R = rec["id"], rec["S"], rec["R"] + T = S + R + f = S / T + tag = f"# id={cid} resume_step={S} cooldown_steps={R} total_iters={T} frac={fmt_float(f)}" + cmd = build_command( + load_path=args.load, + data_file_list=args.data_file_list, + train_script=args.train_script, + train_iters=T, + lr_cooldown_frac=f, + grad_acc_steps=args.grad_acc_steps, + opt=args.opt, + min_lr=args.min_lr, + override_ckpt_opt_param=override_flag, + extra_args=args.extra_args.strip(), + ) + block = f"{tag}\n{cmd}\n" + print(block) + if args.emit_sh: + lines.append(block + "\n") + + if args.emit_sh: + args.emit_sh.write_text("\n".join(lines)) + print(f"# Wrote script to: {args.emit_sh}") + +if __name__ == "__main__": + main() + From 76f5f9dec7e8c2dd3f4a3fcdcd482b4bfb40a1f3 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 11 Nov 2025 09:55:36 -0600 Subject: [PATCH 515/590] docs: Update `ALCF/notes/cooldown.md` --- ALCF/notes/assets/cooldown.png | Bin 66403 -> 474290 bytes ALCF/notes/assets/cooldownHD.png | Bin 327887 -> 0 bytes ALCF/notes/cooldown.md | 4 ++-- 3 files changed, 2 insertions(+), 2 deletions(-) delete mode 100644 ALCF/notes/assets/cooldownHD.png diff --git a/ALCF/notes/assets/cooldown.png b/ALCF/notes/assets/cooldown.png index f32995abd8fff76511de23cb7a971a3c00925221..0d34decfed645ab0260a710c4e0d780a1efdc64d 100644 GIT binary patch literal 474290 zcmeFZcUY6lw>FFoR1j?RE+Qfwq*p~iq$)};La0(g?*tL)0xG=)kS3cNIs}MPLMQYN z0YYz~2MCZaxPR}KeZI55>zwQT{@nj1nR#Z`%$iy2UiVsgLey0iZd|9iPC!6#L-F~u zR|Ev4r33_*ZeAtC{~~qK_aOnn#Ya}JwVkz9l*Ir*TW(`hpotl`yRALGoq#|>+TGq5 zU~T5iWMXDvWhZ$L4TarfvNDytrzNPut70!}W@+`@)6q=BQ}s2#(;6UZdQV#Fx`ewJ zegIoDXJaOJTN^tkF?Y#(zs42Ae?I$}=N{9qF3#4H_q0{inPh>EW=uldyxhF^q^>hb zIGUP^y?Q49pNH|^B=1=|JKKx#@VL3Tak~j{105}R_(VlTd3gDG`1!f;J-D1a?3|6= zx$K-+&W8ANjAv#}07ol(XDgr`)7hBDCO{Ww$$R(k?M(j}+}XPw_gr3^@=BzGj_p2FA)^7Eg zFn^(I4*J227i2UScg%DMEO(!P-#l#nAm&WPW&ibR^X=eEO#2;(RPK-N%uPZ)gx!i- zE2RSUeQifc=J4PLsi}e=-atqw96xJwccBoc(*0X7B@o!eozz-;u++@hp*`MsieX+g zQr*NQ82N(#+c)MRLW1Vzh9_RE=H?s?x8Ll450~i<5WS=v%+|Uy>mlo%{F(9Y1qgFf z!q?jGt;|WY$_~MOgbQQ^cPTe-GY$DyY2GJ`U@Q^kPGS(2K$k01#|bn^&vy{|sIT~H z&OLB^eH8Cn5iFb;DN13+c3)OGK;*l6^4%*C!lswo69>V}NoC(`>;{U%m2c&IYQAWa z@ccH-_r8`#{ASD!vis2O`9=t!rL?*BJF|B#lQ^?-49D7+_tHz!pRTOlQ4|Sj6uWnr z**c^B@mt#4H)|p~_Z2^BZB>EGn0QqMDJr8Dnn^}ce&1@K1Q@eT6qrLObRJ2WB*f@zS+EBjpR7q1M?oRx;t`3#O7MVIG?$ ztYX~oS7@*%$mbTJ#n42c!X#(-(xW!t+sE|CRf($TbJEVwKkK=&CmXmOwdDe`=lH%H zK{Q*hPX7p{JV78bZD-Fwxk+G_Jkeq@WS1sUKYW% z#}f*#qpMHFq#qdx$ZvPiFH=iOWAfIBjcNJDZeuCLgHwL0zFjQV%)eG#4A2%BilBcv z`-L21Y2SPUK_R@P#yvn?XI?*{WAo6U` zY_fI#Vv7SOB^#w1K{PgUWpDqHtG&^a_;sVIX4r*kFPpb-r*nAp5BLc7y&AL{W_Xfs z(Asg?6N!6ir+p^>Lqy+$`G+tOy_C<<532Y}v{aGAn*>r!1eg8oZ{D~qb3Oig9U^d5 zpV0h6_1jwyugt$?)h7Dm2SxFXfe(_?1eO;IrY|2*ClgWrd}B{r2)5g#+w~iszVw0V z{>MPxn?6mKApP6_GFh$LE6Q^d*dQV) zRxkIi=vd?hc(cMc9 zM7$E75&ZJp^maiAcvDB2dEw^b&A3~_w=lOO?}&$xg`C{Kt(a=6w#_ORb&z@S6}#0X zE8%6T<=eSQyRrH5Ld?+5vaPwM*Q(?7pG?U~y{LZ%dU`6;o2mV(mc92bvzd_D6*KXy z07t(*l|z}qZ@L;jxz`5hR_?uHE%);D)(6#ey*|<1+}(SgS5BBtF4k#9eY47H8=SN59?l<<$oDWPH400g&nGX4DJoaC zjvEp!$QiqD%d!5}Hr+L3UuZpQjEtJ9G5M$2RjM%Raq1lENwH`2*7RW22-WIth7{5i zDFa&r;mP}xIyHke3eM1Ys{YV6zyhBZFBxwUA2O1UrjeS3rqo%vF+tQy_{!=yfwFtwbZn!7+xzB9Za?bTh$NMcGV&pnoPbm>xX%U>BZhv zsxYgV_v{de1?J}$d@YF87Sxd}A1YnYq0|}HLqgI=55_x-b0O6xN@Zr_3uAUAb(TS6 z)#GO4Z^jLZ71kx!n)j2|#9MB`J;I|t#7WJg#*`)w_37ws2K7AY2`4c7E8a}RQB>01>t7ZU3y85q@!LPI8mpp*6% zRcg*gvr`MFZH@i13%%35D65%-jq2VVF})!AMnkyNs@&%FUES%ymA&rqM(7Xhdkmu# zsZ_NT)Elf?3}#9!_Gyqc5-_SqYa+R&V~uK%7?8nf_Nh0H21j&Z;X=cE`U`3z70Gnn zcHLnzIWjua=|6<^fg2JVTTGk}AAjfk&Y~?Y!dLa^z%W?nx#n}9=SJVXmXdo=#vhA= zi{BUzd&o;%D2f=S+@O}?Iez`o^O^UvH%}0>*Lb^7L&(=*RI8p^(s~iu<0gbYB_t zCC#PH4YnCX!POFlm%T%ugo{9 z3<>h7@Kjgcl+WPgdX5to=|rPOG-BG!rgQ0z{gu@#h=2Kz18AQMB86?_7|dd`gh2b%2=zvYV?QaAQGq8 z%j05SK+W9|+QJ!X6wKqy3UeBBU01|5Wnl)!83*zXu%4pgn$7jj%mg9}tFEt|?^11r zy_VP5;{LhcYb?ejV`@*om<8XJc|}-bK9<_O!Tc&BSi++yxx>A z&*>?AsHY%YP*ae6Jb3(irQ{U2?4*-yjd8`5YV9F%%PPmzLd#ZvJy#4-Q)3Va)f~v! z3q1l)T0wa>u8s9iO!pMK@HhbvrHa)c=+V>NlVQ?lbe7_+QX`ug^O3Y)?a$~8_RTe8sIWIJ;s zx1)au{23gEcp$wB;`TXBuiSMjKnm9vdu*W?FZh;WR903&>|UaqizQk|MXsmw`&xRw z%DB4W!y@TgcCS2K#XdcDG6BtE`-QbhqU5`SCj^^p_ujm|OH6V3$+I7%`?tvj-#vc!sji8^>oI)F zYfL(u3JOnb8UdJ~$$w4}hhBLt^D>nF1VPJ>+Ay4<@$#xf#*uC8dZ5L<;;>!TPMz_5 zD5(19`h)EbUbFTAUZ11InJ}Ju&Jn;Df(sWfU%kmB^Zx9^7*r;XJ4F_=VbCbCvwPJ* z%*`&UnS8jSt>r=g?_O7LTHf;8zd0;_6=ECaO|Sp){1~>c(j=QIX2*mM2cxXc4SlsN zSjNX;oh~3nhfVePq4KHei-Y}fV51w7e9yu5%crWV=LNa?y8Z{9Q|B!)=cSvT-NI_x z_CK#Yl5V^rdUu%Ot@iytrTrzgOw9GN`6?kPeu!-h$bmfRyewBP+KN_?4^v1y4?8dY z_g8*3vtzXat5~!zRkm^p*wr)zIK{BcDIw|EX^^EKs`6_y{@fM6mAmm`F1qpjX73K> z`g5&D2T)njI?>u%^4RjA}~_6kS}qNnz;zv!_n4(phx>eav6_Is1_2dAT0D0p$h2 zD~*pVQUl1OljFG&NT;dL)buy{dXmVOtgPqv$LFH0U}3n~?eQUu={n2#mc6TGgCm;% z?IU~5<@CQ!%V8Az@~3y=iW zRDhNjDzjrReE7HJ&Tr*5aanVf(4!FB4xE95!uc(D{fCU+C)?)om?PV-Jm(vtjne9? zDjK8NHTvOgJ6+=}DPPKEgGlEQ>EgHlb7=oO26ES$MvA>E?U>=A5LxR_^c;dA)*O3n zt(50Q)m2O5YA%1NXfy7KBzJpoelxPJmt~!8MlSN+;l1+>tJel7hF#;OPIhkQ*riy2 z8H~F{T}*B3T2S48J5-#0R4j09KKLg-*ZhI|g#`m_EJo)$SQ_|M+#Z%^h1jlwv?fx{ zpNyDCJ(?zZ4z^h!|GBX94Fw$PG{dgPvtc~_bbCvDZmSEOlh^*wW&L#+zFl632Sj{s z4%J!Heo1|7WL{Eh=26fj(SGOry07B5v7o8^rM!*aDLUwHJD0|FwJZj|@NtTImF?$$ z*H$#VJlli|Ct&|KU7er*T>rn#_qRSWgIAY2bp<1JpZrbs|IxGOcVnKiC!zd!VdK1% z|2*sYC(rBQJI#d@oxFvY91s8Z&z?JRX~!4EN@I}|^#6Bjdc%??U_B`D?#lT)`%j{u zoBqFSUbK;n#n5&Vl;!_!P46x{n-?T16r`Qsuz#NQZ(9B>wBI%{Uf&=mxAS)Tck};L z#Mj~PyYt`dd?%0o-#>d!GIbDM-yOh>i2u7a{c#I__}(#;fB2i(`Cm)c&V$$YVb|o} z{qElk{C`pHxuyJNmed9DhflLBLsR{{t^DtL{`JUz_d(|HinBQvtom;T;eRmwe;Ml! zdHmt~!(-t3i~s$z=T`m@hS&GJu-m`J=Kpo7{xvrLczORdHvfF>|1~!MXpsIjHvi~` z|1~!M8k_&(;m>*Le?)`-8k>KO%|Bk8e~ryQd+C3=T>m1Qf8Miyjm^Kt=3irTW;xRS z;hG0fqRo|p18z48*Y{b0XW=s93p(-iK^A2}<()g2TBeh8H{1RoduHsW&EE5ik1~aEb3?v=g z;t!AA!?+rbcQ7`4Yd%NI(e@n$m4S<3FRA>}?y;GmBpW)hEIoOSauM~IdTgT+0>;EOvYZ2^BjhSo zu9---u84i5*Xd-fQ_XXRd|t=mnXQ;TiMc>Lv!g;0FX7?n?mc(s@X+UIw=HfQ+i*wi z(~C!PEm_U_o-(-FD#*p+9#fg0^~Y(~^}a7GY8yvBb^Sf{>jD|U7FL($xWP@c;7d=* z;XU}hbP>F(M}3Zpw)DSW`ul?t^+x{lfbo&WG^B-rr(JvIhlHGXd+qiJiY090 zz`f10w_B+8o~vC#kX9}d70aHIcOEgM@n4Y-r%5rl$s*Uf>V#Lanfa^2U)v}newIl{ zk$&5Z&xgr)Q66kS(N8J#3v2sA(*X)m5tSH=LhtkLC7%A`Y{zK&O5*t+E#rTzyG@a6 z8d*>7BFTjFe1-bMg=tQ=Cl(&I4V@-Tupby&rMq#wRvk0f9n`2(Nbm!1j6je&v~$8NSiBeBGXX)S7ZW|DttAclDp z9?xftzYLiP)KdEPI@D@%+~>w&>e;{C6laT0XyZG#jq3_;XPZK9aCffc70_@)=NdMD z_>%>h5pmra(g*AjrizP`O}L@!8m@QnFoL~K1B&-`tpn!+NfS~ktz)diQuHF~@TY)F ztDznG3IIa9Hr$vG{!LW+EnNRZvu6ql87vd-=HmRN2S3NNJb;}8<}B;zM?vmB8`68Y zr%A5AKtXX{T$wZn{Pw~`4-?Bo#DzhpVf+a4|?Z4UG!GOsX5 z*~IX+&^f(Un~C`|z41E!eS!|hnqMm^UT5v>_%waA{jek^P|MImu=Z%h8lQDsB3WrRrJdcGky6)jwfi-j3rWC02NXPDV-gzVcPYx6MEAk*+ew$k+MRBV zMZvTCY~Rm?BSqI?4G*Rp$z;)T6p9G?+O@ReDZHr`@pSN6&MShEs!8wIc(xYbud;a; z>Jh&S>je*M`|RcZ;HB%}lSnue+(5%#!;UqPAgpcG&W2ZmZRI2b zJe-k{Vr`Ivvu?M~7@B%|pR<0%EF|P8Xur$NYG~&u zxGOa3{>(E)=VY;i2S!qf^w4tOA6;xv;(^goydVq=-NG&;*(SnEOT}QU`@Shn+i^x7 zZ$P=JHg#{7MoTT?`2h0n+qpT7OKhpv)gItxkKShyWun2~LA zDXox=cCeYY*?dgsBY3JJxDi$emGUSA2B@ZpXQA&1jaohnCSSV!*VXoOy!v<%xYP$* zuN-}s_ZWP5fNGnW!lKns5yM8llDo_OY$VkDA~u`;j03?*50vAn|+{SS!uF7uB z$7a4ok@Vt1oUr5VW#noe+oGFFM8~Uq40M4JN5WIPX>za~#`Be0?X&h>dX zqJfHegAOtin2!GH6B+@XE9t=3e)#B>6%@t zqpBVa^>Pb4KOcE}!g#ltA%OB>CjcJGr<14TiR8>%aLM$g-Cb=>a%;inX2 z7E4Vfgq?YiZarMMm%?CTV9VKW-R8yx*W@!lTI?PAxU_W!)kn zXiB5neKAynbu)?3g4aYemi*r1#Yy=Ym&BTh0kp9O*iFM1aW}RgR~i^FT&(=t6mStY z!J2uX$L$heu6(7u0)2?;c6(k?n_F|d=gHns*D$Jqh|&A=5OIVepqkJ47>f>)^TLq& zU|C(huNEoJ3vcX68m=Zf-%PFavg9d?r1JQFS~#l3L&aXYr4V zRz6!j>9SEm$kl;Esq~&aQCk?g4t=^gElZZO-X(pw+kX?O(&R?KQBY;_3SDP1W6OEx z%n^%Ie22xUi*r|f!@KUZu`3LqxnBGIB41h)i-CsdH+PLd%`r0vfJjAJU5X7J{dyrQ z`@`WK?yE-k$vp#>@%K_y3+K4uVT`WZzdHVa+!vBz6KnF;>nOn@Xxt9Qa?K9bA(RvT zULhRBZ(ziqJ0i8)x}(78R;eUeB|Nz30tS^=@K0E1R+fbYi`zyR~b zsJ{$0fnNoC55w%&z@|8dvMD5PCUR!MR3?*Zm?C>byQ1PQV5Xrv-{D8*h?QXWh%uk5 zd3cPA0;88VsLO52N7yFTnLFHSWgvGZ6<5Pe|LhTWT=rs={2-sS8`4RQrg2y}?e!-u z015Vp7C09d60QZf)F<+M5QObsOG`i}s}T2Y%_`LcqwAmc+1DB&SsNYLlIg%c8`a;S z@X#LIDW<}H`T{m6#ciVA(rEdG6D$Y_+ffAfZ7XC=P)khN5enj=1NF+OPo}QK}^Y?4*xl zW8HA6iz!yb7`@*$Fg;%{cAe1H>;kbO+IDMnUfrpuUqR*`;o?$XShh8kj(86{~Vx>1d(=Y15Ct}3>5(8InwqMdjh z@Ue=A)cf^V!d_p;Y3z_fELd;cb1hz1&tiz^t7@!8%1P2qS|4+ExSK=TR3^?=n}hn} zV&^e_w{%q-)hZ~}jt380*}YrC6pJFDxiE_oexb!hE@+_I?}Was0TX=Cfp;ti3=-pM z_gpQ|Sp3KXIfd{RO7F*KZ10$mU@Gs&LM20%XOTw%>_ZlCry}S^VDWe%^rH`EN}$QB z@>%F`*Q?}_Y`kr#5RTXz*D~0_*l#t*W&6J|n~y|`&%STuYkm?PDZQ~&rr}Tl+bqRf z2Y}vQMue1*>3T=$DjY5A4ZPcb1{pyPtv}O&uiQA#kh~0*!ZHEp21LrgP{3n5S2p|8 z0Lmbi2#Febg`tKI^|(`_nKW6@7oi5u2{4!R{z2NDOX%IZHf^U<(pehW5 z1M){%8RMVWnRU#sg)V(lvc8-b(9+zUrSmRVM`<<~hGlfOf4JV7ILStcXi0T<9J}d) zOl4Fb_#!C&00b{d{o*{rb)$kf(rUxe%Oj%8#a`+!zj)zd0b7sZuY2x$yi;v;X)&M0 zI(gJA!L%a&a3Q5qk!%_jDXmJo1jhSvIeJD^bF7NFQuy+Fe^M0=PwUmzM21ZmPGWbH zWkwi=^_j{3Hb0{1h>^q;_R5HeKP0`omy_vT@Wn)K&(3S@4Ss2{aZB{#LG{9R`W3lw z3ngzdVjdb=Oa!g4)bVyOOPE**91u(Y`1@v$w}<{$Sbw(&$MGT0;*dLd^a-WD`^#Xo z#;jLBUqt>WAEee-v7Htf_WU1m_vwI_&sE?#SXX(iU2*VQt7FM0lXxr==GyvRyC zDpaf_c47XD0wb~4Myb)kN?y?mH}V4E``la->C6zPnXNr8>D^YA84Irs4U1$NM7YCC zP6gXb&rkS5>QvQCipRuCD5iv6w@>DaA|Ac_u2%)K_vp)vY_Mu`t14!(o-op)i-+g6 z{XjtMs>`G#pmv`Y5y~K;SxA=(mjf&oRM|Vh-@HS+EtyP@fHHdk+fpk7ClGwUSN%xm zEh=1yv)yp$0>l!9)4YyxQ6~HnE&zh>ZnZ=;szNsyPNPW7^^C^HsBp*Y+Z+uy+NL$Q zV&uaUN2f$o%Stz&5|_7{jlb|hS^D%a2W3T3`t0ouyTnSmYEF3!r)J$b8gNzjP92D@ zKw3dp@2LLK`=xxBAu`hY-WeXvN-Fr%7v5bOA??Wgr4IGTAH+QM@!f>4+*`AyC+??= zJ{B7C-i5Tt(%Fh+kBsgf6LXVT7eaLTm0lp7_4ZT4xa;quKZ)Vd{>MxENS^%C{jqe7 zoa(;!C`MeB(4jXwo!eS(x*J-qq`zUYt+iasz(=LN{>Q}~w#8G4{KQYr*yS9JzFS{- z)XX^!?}=`F>Q`e=>=l$=_9f0xK+v)0;$EFRExJPSaB?zp&t^AEB(Xn~BmbtRqI{c= za02;aggCmrL)@w_vv_d|n>`$vu?deT!`D#W;9;E)&(OA5_9v4j!A!aNj#xf(xS6^R z?6&aojOcQ(=l=Uu0?&TcX^!L5JRWu#UL47Lkg*4LZEs`l{z${20cbWz_7G4aVsw5R zytP5shFf`oYJsH+%|!^1Ha~VBePVoQ>WUzX(ss-Ro%Gp4H%c^uQw#t!{O+UoA4QIw zl*2eno@;v*2sV~k^*3DS;Ec25I8mch5#dns23Z%?X&!9n@iZ1xo(fk+&Wo(Y!P?cr z5zc#qT53eJScLCsQk-yo(ZhJ?by0n7F5BJ7AIQvonr<=W25-!uN>IMwo`qd-{UhFt?9uo#)CJod$Nk0Pq78)xFK!h ziuV((1FOH*EQt7TRQpOwhi94O%x_DA(}-T41fHcgo|2aF`NpARHe_7g0hXeWg>(7Kqv~xz)L%m?V0%P@*2#9VA;H)a5ay zaw6H^Ck^B>o~Jpd7>&37d%ib87~jRYJsmPO|2OMpjr)QncL7)hhLd`F+kCAqLhD4| zQ?PMqm(UGiHa=aO4Cn%l^?kne&&AHKEJR(}%Q7?OY^~@mU3;m9k<)US%MK|uA>M%4q zlgPot;|Nq8M{5_+15P-(Qfpg+1<<658@O(9boTlXpjJE$_i83BQ?mNmS;(+{R@=Fc z`xNsT|D{^S=CbF|D6Ea+11iFZ-ogE#ZAfpv#pcUpP~7QysQrwb0Eac;%3F!W4?EpA zAg@HT-BX08sG?F!$Np$+EEGpa7qzrc-lENC1pCf3-}H}Y*nPoItE_p8z7mM(PYvu6 zJ~8%e{Xoo3MM$#5D1`LxFS}$_t7>4?9?jlHB2<(G4PyOirqu$6NK5o zKt7Egi$?K9Edkn3j>nBsbsp28+1o7ZD#<^==`Iz=Zgh@;!GK%e$$z7p%VVEpyh4>| zZW~2-&xRb2Hud_eNz7E0rN5h?;?Tt)fvLQXbFG;_*H)Ju?mSSvDvs7J-4=p9FWL0y z3lTik^WJK10Cr0WpnvGPD}Q2BkUo4(1(_jSW_lff9a5n&YJr!;8M=DZhze<*B7A{l zFp;^mhQ_b$EzmIYn2;13+;)!BQ&_#v4$5`Zs^umJ-&wG7SmDUMt78(fIgU&wpMBn{ zqYFBm@1u4o**jy5^*W{XR=nbkir+E|z1{fwdJb>@6g_0IA12GW;g|LO@5FFvF+f^jLZ@#r4H7Z@|;u;*9sc2&jR-Pq<6-Z-y70(H$KHz zwk6*-cHodcVL%cN%PKAN*mgrpVd-y{bQe0rc8Q2vf3h;En>G%jOJ?wFydgaA5&P9Fez}>g%p(g4Nxpb}hu6UtcC92!w6kdO# z>Zyu2T;LEZV&({9@P1nm)I2In+g+RL2+c3cg1ZDJr1+ddC44p zBzLEX2PtvaaY~{?ISQ|j+w7s7^+4$#0s@<|MrZIayA)utM=c%>zu;#Q%D2ae+r$?4 zg>ua0G++2|#68MJXWx^)Tyfyzfu|ipS4;s!Lq7Xc0=75l0eMt;7VJY?q>5r`49=}Q zB`VCMOpo)r^K}%?x^U!Try$>Z?Lg_e6tmL`ImymXAFkQ^=<-|J2u;6~n6F+;CFAN% z5H!ro{bdkLK#55-+mbfjJ9h2pgwd^?A$_*RGe7=k2xw{RH^MAfLEcS>4zUFQ^5b1z zcWqPR#irISy}K$(+%M&|{)N=8%&RYf4-N`o?>_aM{u%Uuo+njNxFy8GM4{E?>l4!? zycO%tpLAT)AOD)$H%=Q#D=nq(-PVb@44Ig$aUu3SMd7j>oEC~z_|r!CT2eksk5e%k zCD-bSOfO>1f-=|31sba$e0aF6Y!pwbI2BGp&wqdNF48*x^8hihZZOu0c27I<(Vt+O za;fJu?EA+voA&vdgw%Ej??P8)$)K}lTmv31Mo79M1)Bmhbv(JLC528qUfD3|#hY9K z<{xt_LoQfY;qmp!s3VDxk_O$L#dM!ioE^8W?{+TQzhm(0OFKRO6zrlC9tpy&fsn%m zxyq&J%gS*E-cMd;Eq9`XIK=E4K3F z0WxM9qTxT<^8e8BtdEU|8?A3ZjIsaCOb6kLo#uf8LWDn%hY)A`ZlH1*`Z7Pd*Ck}0#*%;}~w-r#ERLIOKK8Vwd zhK)==?!KoN5893@XeAz%6xtWhjBJ+T<#$~p9k)#)Q>R?;a3OLJ{uD8UD+iK0K=%h(L1?a$uHt;U3(gf zYq4qxc>Rm9k7vi{K+qN%;d-EB)U13_f(FUe$`|t0?8*U)aCoNdS8Cak-FN)5EBG{E za>V=BW4-u*5le@syT8E`a5Wo8_IUmlQF2dETpCL!`tEfN)y^xCSHKFGXlfW&jb(WM zVk*qyCCgYun-5C4B+KfH90m8-z$@~o07@=z3I{JeP(;A|pOXEFG-Sqa{pT*&2!Rhi zE<8Wg(o4)}#EcVjT(%Y$8n3`Jj@KMqzreG#MX6}sMR4<6c49Q-_PG($W@3yUQ~>2x-JH*YC8L!}8xUJmO6!W$+J9U~be8!ldf$O(;Gf@IpB%A)!Lo4Q%! z3S9(AYxSwd3-l0cVm`#?**~3JIvypBkwIa&xsv;5|&hQEsmM+vGh@| z`AmxURG}gSlNwCMQqS@ zeXP7*R?o%=gU`0IrGj+Pdo^yWhGn=_U_G}kFeGMwmrk#tQSe+Y?jh_ypsXN(-qT$^2V{J9<_qNiXxft|Lxcx|6-(E?oUUTzcRH(6w&- zJCD(g$>AQfajl_}B>HGu2nw27f#&%94o=2C$k^mT&r+>WbhU7o9i6xEkVW$oJwWKC zM|Fa(HZ^oOfp1_x04lAp66v6IfVQm+7&i3Hu-xxCzS!2l7A4KDZ@ar4q-b-vFOJS! zyum0{+q)>d_wj`da034Zn_^4;hZ+V}ZLHXjL5zp>0^6&KB&jzZ4<=V29fT%Sgr**z z6-z7^02D^gklil|g+?)GQAE$?IN!GTd(HN{Uv8ny)~_ZAZZkgr)$0thewRzL7mBTd zmd#deMhCnwr}tScsJ!kLltn~x)%Um&Et|?3D&7RY-Vhdb%L%mIPv6k2SbXc!U|-o_ ztlmdJgcN_Lq?R54U&Q*#9GP4 zrf|LI0nnNU0cj&pJ@>b=w81N5O8R|02aa8Wnh1-{!$%=bC<2wyY1J)beV;DrhAFK zR3;$4ra@U@L1uyv*%d(Onu={k?}yDJ3uOVc4vvV~al zCm9Rw$0HEnuueq7lR$i04a;wa@mQYMYB0pEQGzNKrDl+or#atQtZfMN#NsvR?&p^G ztsR^WXvL0g3@4(6RywR@F)=$Hl?wqQYXv*>#XcZxa`xflsdDk|~&(y%mv#s7^fxXZ)VzHq3 z*m33$4sIoR`5MXN)tc4D_zww20Myh-?%t#ulK3nD2p?|>9dQ}b<iKd8l>9`Nq3CuAJLQ>@t_8|KmznR1!H zn#0)4+ZYUUUS@c>0|3>JFZ>|!CnP2DIpVw}G_n*VM{i9GoTzq+YZw%Z>M}3Q)nI6J+8OkibN-M9`)sFEb4&8 z+&4?)TAH4GB|B_3HXbWmKHNXcl7h+(&5c{9e;Ui+vKo3@7#JCR*0#KT%X!L%py9Fh zJeVE3uGGrdO~)I*gVa)uw|l09U3h3hpRJk{JN1AR4n2C8m)#M0ze7jYx=ZkBa$7lX zb3}elii2jb|DicML#YG5_2O9D)+Hit`UfgU59zBOjw2@!FEbQ!;({^MVy5;^Jj2=D z=%E5VF>5;^_Ln}kaKO(O7({=DIdm(#;$#C47Dd)R`!~iW0U1s~>#zm`w@HJ65PTGA zbOTFouU@tE0hS%bRnBgiaP}`+KY4uF$(SWS8{XTdtzYMidHqi6aG>Ag!dV&GuG*vR zcS?7@`UF`0cz+s@cj!pMTXD1%y9^U?+D6crI2MQ`7d;udg=)=qu(s?rVuDX<2` zfQ=EU>l7PMoS!{@-heYoC4~lr!%Ba28P*ZsO$OF`>*=@3zX%n8k{VW2&6X_vaINXX zpEhwHiO}DFf=_#N`OIF$A2_zGR}u?PxNG!z?D_9fw>(K;@H)oILm(N;5AZco4*emxa@iezL2n~cL6jP+Ee59^n?Ib7NbiFzQoJ0RpRv39^v5%Qtq_j?HZ z9$YUQLiR9%wDr+drhR^pM2%#fy=V#jKCdA0kQy-MV(1l*_{m#cH_6?E!3;40Xv)!SJt*m~Oi+gz#}`l&i~kq;@&R_Osx)I3ow3K7cE z0EJeIR4*xnvW>8QqMb6~HrHBf%^6LVkdsP-2LfP58n#BO+`Fg{Hu zO|9Vs5h_hOI*-nHVU0d^jqFT+Mz`fx5=@Q@^yAT#*)?O~!sY93wXQa$AaDNo;dCPU zn|7y|f|*1ZjwO|38UNyot2t~bDLk;XuMIZZXBkjG*M~jn^s$Ox&Y(~lsb612UZnKa zGkS%<%qJV21Ono3WpHESrRG@FVkieEBM6gJ+GwoUIN7j^ujaX1IHVnMVOG8^pFhsX z(_zHke{bgma`CK+p>r7ZG(VaCWUh1CF(yWf6auj?&)VOl%Ue%i|5X|I3z+-uKRC-r z!{KpkDUP~hUj*UB4zWXB!sCzg#k{pgockFAQ_+=bRrhjVi!UDl`o9G0&3Z^5&h;OD z%HQ&AVem1{z(?;+p8IRm*Ev`Km$v&}w7To-sgz>9Ml=oPe!mh!#^7SXp0M-sTD%f> zwP`q~t1^wY`7%B*Q(nDXy0^dY(PrjLTvWgBX_&7qO|$x9RSDu`5NiUM_`;%KmRGu0 z3W4n8^Y<4aG4sJKQc@)wM=Ldj4ip0-0@hZZ+if}9DiDpeQHzw9 zcrpMv`ji-QXK}7Asz0wl%-fyVpxoBLT&C@#0!!-m6Q8G_a0-%qP88$$A6;C`yY_{% z*WFxI_eV1nB8xjz#RKz-4)k8qA6OT`zK|Kl(69a~0~kw{+1n#{=ZjDeIb&3=ck;E+ zl}6*M2*^Xle5X8DeH+ivGOCtL>X>)J1tWR5Pxbh$3I$&wGfVOqpKstAi!HFh)4*XH z8e$#RB!+-#jl|i9GD*OCtKNlccOFm_!1j+;XZqE0GRFhSXf?}<;4dyIzf10Kr~PAI zw1U5M{5KuNq#h|#-gH(4g)ieQw@1`rvff$e*XxepBmG8vpmpcexf_>7uYE=>8f(UQ z?0p`tS&bPs%JD_f}Qc=y&sjb`4}&=IMeE}>$Rx^mz*#-;;@+$(!;D{~#z(jIwbmxHEs z8#>O#BGIx;@I43DZq8=EPzDaFe9K9fq3igPoo?5~)Bs;(%cwSBe6DZz6pN0t`vW4C zHzFWzY})rzNU1TX=quOI(P)qLkf$Sut<7_=mabLSw~6Kib!%%F{b~asXO44Ejc_m# zIOS0~xOl$Ya{GonKD1qOjyYayK z*|Zrv5XnpLach$@#;$`Bfw(j3<^mDJZ8+s+Uts;7X3hF)<_8`nuQiC*%PbQT(+KIU z%;L!i=*RJmDgkj$TrH{)8ulT@?-J1h_N?M_!~%BGh}fTCYib|sb8ew`Be&HxpYD@e zb=b}3EU0;RlA;53UA`^mvJy=X?oJcHLoDyQ+m!g@r1`B*;#YO;pP+_UBTW}?Oq${{ zz~1~g&sy*&LR#A9?v1jZ4|Sn-4wK={DeCxY7SoY+3$LycfmJ)`#w+KPyuwqjgT?aU z`Nif78Qe8d#2U8HT&lscv}-lE(S&O2OZOoUoxbHej+*-(LQ_()!H@6wrPF*z}# zdtO!Ht22t-Tc@qSE?0;);%lF7l1k5NF?Z|gB+BCpgpzl1V9|DVeV%xy#J&5xd44Lao5PXZ+)ft(hDA zN-RPXyn38R`o(FdPG^PaUTCN^k`XYUhdV{s&$i_(jIZcE441UF=AlQ>OM04np$zLh z1G0|x@>U&}_X7<}r^2E=EM&Lo*w|?BJd}h5LJH$8GLxzC)JB>?!yEtJyGvw&{Rjzh z=?AV~Pu{fs(Cf0o(j&Rns&)!zMo?-bS@dsEUVdI1b;2+=8_Z-5;`U+9uLOc+>P7l- z3rKV*9`p~g6}L8o-6cPQv@nuE(GabJj@0gJlAgP*m~|+^@T@HUxwtDl9P-p7$+nX0 zEK@o5S+KiZH-2_W47yRPF_UdSIX@AU~OG5wHQeE4*cWS2kevijVxbQX# zwe(`O0IFKU;NRu$P4W~4tJ@#t;XF%93SEn?ab2a5XT7ikyOIy&murl z;$z(rFGJ?Po|B+d=`G$^p5SYhf>K;ZCBwr_=#AIxb>d9`!?L7e#r7KYwOjHmPmNpJ zMxJbKcx~K}4I&{0)*otCZEq{}uP(5a{S0tIK$)6I69bkzOF;%zx`4xaFMO>LvxPEt z9qQ&j{gY_Nb7~Rn%e+>D1shNIUHTdb_UThig=}C>FcvjrY9{^VL;4wTLE6%EK8RE8 z#=Wq+!#a#xc^M6O_KE5&gLPc7nqQ~cx8!k>30d)3uE2YtK%)ovQko$puJYG37Lh7q z%k?qNpCkCzCMKZ){$vtN(!R?VF7;7^Octl(;h#JNY)VEcn7gy7VNQUr=&;usX@XO> z1D$fH%Cfl~yH0e%(1C|_ZG^yYx~L2+bBsMQ8a7BZ$7e_wx>soz!B8*lg4iedvwmk4 z1_%kPK6rT~cN?haH5U*i)-lq?U$JHT{M~aOSGwebWeczh4<}c|!IPbaE6(6{VIUR{n5R&#qeTtln?O7Joa=vhDWmRyjGy z$Ptf2KO)8#;uK^wWy+q&-nO*9{%jxIgz)nz$|eb=Gx*U{f8{tEh(u^D;|Yv=epe{s zrAhS6=D*#B9gVHdglBFS@;0}3BXRq~ifp~J)M`jtVeV*GQv>6}`34}!M{{We;=bnJ z{57H|fF%tckqg;b+Y_G9#)Fh8OmL&`7flKHJn-wwfH*xnOWkU!VL)hU1N_GlKC7>R z@3XMe`Twx;)=_P?OS^bWDPG*&-MzR&DO!qq@nXe;m*P%w*CH)JgG(rG0gAg9C=dv) z1%9D>@3Y^1&ikEpzV-e1CyOL2&-2_f_uMnrT-QvhzkuryW+Bg#9{SJ*iL3kgfk8Q% zLJPJMt!V)pTt5^Z}Hr$cHM9TsFIlIsjc(V_%EPB{h&XZQ< z;$UI10|-(T<_f`;6{i0E61^j6rkxEY%VLj=5G9lAG?jiO172Hi!6+jLCtw zSxo?wkA0H}0tzQxazHH}|S%bA3a4iA>q^LKzQ--;k&56rzu!^HdMM!O*bieJADx zQVk^DhPOT%*dAg!ZI>=D?JFO36DiK|=rxHVf(@}Oqpp()b={teVRKE5+`Z#jY%l9C zck~7?PX3|}`mPdmL7B!c2#$GCizQU9Nj&#fv@doYV;n!{?6}J)9<# z$hyrf;z-t{{5OnJ_vJhEicCaMxb0@BBXk0(u;!4<6Wj((jP6o^7xFed8*42@rT`N? zx7cZWack(J^K47|f*@xGJiyk z_FQiM6;MAt(wE_*CVZHF6u$Z(cxL*|v;tK5c(27rD&N*`Z+ldx*}v4uPC>B>F=Z~p zICOh)cNXk`!@Mz;%yT{y>b_c8B(AIm{WuDieNvS2BWuvNG0N+nU%!vSAP`cl+-ZSR zbe$hEvZrl_aK2E7slR*#-=62;QnLc;JAc=h@p#egG7VCDA$Xet7%c%Ov`$?7JkHUC z>#*K0i8@6;dt;nAWz*mJgZ}a4wS7QvYdhK~-_;9mva_?A@4(dTqZMuOMOCH)4TZJ(0WOn@Wx1J6Tm= ztrLG2`Yl~ke-6LgpT3d3MCuAgCT*i;Ke^y6A|sD;0Xw>thx^--Ld{Pn=rgt0WH=o> z)+)rJ7YE$VZ-Pif2QLiFf(+t-*{TQEFLmN-vmdvwli63t1 zt|-BnDmmC%vv8<2o+BK*A8@5m}Cj>-h6spX`kDP`HG}%_Ex)-b*Db7i8=`HC5jzf0e|@5WG2>=jrL-aA#$4 zIGY-O_j6Lc8{QegDb$AoW*cFN^@7Td$|P_nC0emqx;kF3r#GcYY-%#?01H?zbBGCY z_5*}4Cw@EOmS31yu49WT&I&TS?RsmKs=l^3CKkdX7v|>CEA0EIQM~hZ-^adV(H8YG z)o(@hTAWWb5MGV-!>cj4E2K~^M%gbucxj-*pX|pQ1Q*6xxMD`A?R>`2Zl_Op*xoUR z+{UBhW)F0^VAFB5@tUhN>A}ctdJEL@7Tf{0-$vG+ld(7@Uwqpy-Ms+;x_HhOr_e|Q5DpN#5ocpwODrN>ZVeZugMq%;2qB$rsQ~-oY;JFSOlXQl%#Qjqc`Hv(tNC)560lE~W z@3D`BQrFWO8_yKD1oV0~S)yt$rA8t1zq%lBGi{@b9ru)-(bnX8!huAG`{%}gdb4mB z*9>Y{P5dVR(kSM!X~L>)HdQumH}fb9eoQpk1hW2Pis@4X0?r?ESFWM-NU?mtfcp;2 z3`KHp_p45y{V;eI#c>9YQ{0X>XGdpzVKw(Nz4RP$#@Obd`uE9jsJ9P% zyl;!UT@YIf;y(=e2kC%om5Zb-Hiw1FCLYOk!)zM^ic@hberot8R&6ckd@L5ApBp3Y~`J?Y{IQ0khQ-uDkghu`2jKhN>-4F18>o|hyXAMA(R(ekVc?H_% znYgByR1L~ak4OEv&^SeiG(Y&e!H-p$HZh3~m!-U?SRu94n4{`;(j?=4?k zxSLt52=k}f2R+|*R_`spspV(ztn{fvLq2jU>Y_@3{smb1L5tuv`qJiV+&Zgap#EP| zkJH8faaOMOzQ!UE`*5+;sr{dKsEQ6(#oOE6r%~|d(+(e>usMIZ#+}vqx%rvp@xerc z9`)CLa7p-;?S!KF5Ik~pnzYNnR-!}Tbi2mZT2|~oD`m?C55%(yHZ_jE*9|Plf8Hze zdkPa|eIOZpjEwK`%FZ4K*e+^;V*l0r`{Vz-LF|SUmfP3Zu;-!{ofYV5P$TTF_6xi|G|&I-D3W6DyC>N*e>(jSNAOYx#Z7Cy z{f{}-|1Xx8T1MDhec*1pJ8a^1TDto2)orKx>Roi${oPH7_s*{7fg@uUzzbgd@p|x;g8t)uJX2K zrEfq}J-m5)kGNsZVY7NDN;djp{3JU8`qRyd{&zK!^@MQOQ<_Pjvw6#k$?_sRU%H`YopE zt-hB?ruU$+4yAZwhw)HZIOVtEybZ3lt)4j${Hu58x1+*arYpH+Cx5rE{x3`O!|X9F zNQ0NG6UAne`#-<*-~5kzw&QMw-MD6mmgoJ|8X|3ZaH;>I&LMx%rmlRhKAVopYL>kMiLKe+ zT+oRI@emw3^i5v-f4k5B<}Lj7F{zzV;7zy@zl(XIlHYfi8Jsx_TO7;I9k-)?ry~Se zCcS7|r>T{UUr?A4qCV2*|K;U-MbyVcedlSR(k-3;zwRd}8+)$ZrXzN>M%GYkd&SpH zc{m>0R_m%N4!Rlm)%+jVf4rNeuP8sAdp$~N0ny@@*XTqe;AhCHNqw;@BYV?MH52wO z&P<4PoGWV1r=tO><+@cFFR~ib_olyPTC0+-CYk|F8aw^&MubXU6@`_U-}#V8c6W9w zj(oQ0humE*+xU3yC%X8>K0frdv{e7h@3pe8%m~FeK{>6Lna-6x!1I<99$4AinoqYG zXIBS}tC<{&UWDkAl^(CPN>no5&7R8)(pB4!Nr3!kDc#mLo@-S)=!;fCXIpJ%o?qYW zJ45!)$sB%jQi!G_vuZ;`D*O~J@7w%m58n-X9wvZ8_s8JOIdWlldYV{QtKoFP3Ij%? zNvc0)pqHgix-X-@p)fXN@-_5Phr4dAbUyzWD#)8Mw&*Wb`cIol(clONY$U6K4W?O- z9V$&boJMn7&Ay50%_pQLB~^j%7mzx1QEE*Cyyt-?t>Z>*S6{zUn;4WwV>)N0 zj)2relU!}}{>D3%(GeOEGtQ%5r$HF>dkXyY3;eDpJ}qDM#w zg$&kbM+Mifju^kxDh~inQE1jyd=E#WAU;Rl z$dK@@Y~HVIg@|7+;c@wc7-P}YtgR@0)M8k*KYPpQ(~y0`S2Vk3vWIMJvEe!`1qM3# zQu*PVHn^s!;%&Iu(SqK|SzY`NJ>Dsf^Mlt!B^IrM+wcl0&Q#sgbsPH9uoM-xkuR5T zIb&J2D-K^p6xxJWfyrteAkS`VP{1VPWUV(JpS#(EnI}u&{(=aD9H|seQxUc>{jK9E z5tl#;72vxVucy0?+wO!IQ(KWqi3ITf*cJcWs>wcxS32kOJEn845_;E9Z~fi2q0iqL zotHJki@|o2b^`)Fr!(&PXHJ>pr@~{v*(y`Mhoh~0ZXaMV_IH>1PUy4+#E^?HFR#4}mmPe+dLOLg z-CO|}A52bYuvrbzs#hk^k`+S{iR^+bU&i10oLA4?&N?-o;+3jpsTwxU_Z!r1;4y#* ze?1SMmQN$dk_`lGa-42$RRO#1uuS|$leq#rS}XebajwolVnllvb_0Rk=taYFw}5mz8N}7Vql7?~ic{U( zKm&XzTQNbZ6j6`JGCW?4tQYk=n5yN4)!I(8VkwmH*ssWd3~#>YFEQzyvg(wXv7&b( zz$^hwp}#&?Lgxd&BQXRLJepq-9#=jXo%jpP&t<~e4nxjrN$f|X)+3UbEM^RfxHn|+ zdummz7UPyr3ZT`oNG{$>1Lhfw1?z57)LA#!sTA~62clkv8zzh%>x!$sJh@97R%{hG zkyoJ@YNUc?&L)7pFp6^#@Fi3uIRR9oM*iWw*WWI=b{tki2j?H^B zpOzcQbXivlMYDwgJU$Yzi4(NS!;Aicm^0MY(pE%tA$QG?nbya;xBg^ zS3w>O0W`r$PyKzep8ES*VQ(t@(6SlL++qMdi8~pM$VO3vfixvAceh8GpraM~q{_{O))VSWra`{p;^K=wb<*IqU`RZKRgX7mnP7O*>J+w@Du2 zJ!kGKJXRl{41wP92Z@mil%-v`Sx-{QpNp=6ZRcIxt8P@FirGwQ>|M= zjLj|x4=lLzECZE_KcHxEU4dQXE~?`4#5nHC00s%i`R#RzH=Ta3X-TFna%_gMO7b=e-+AJv)~hyywc0af$6w;D2Lb|~ zPqx`@Chyuk`I{rg$pcY2$CExPn`$gGNh<7ig^GnHp*;BR!pBz3mv1o8FM+Ghk4`xyq2VI> zLe4X(MATKaR-Iup-{Fcy-vp3|aDj_;qqowd}N0|NgCzhn)0T9dm=HgO?4< zo3-Qp%c>3r$^@7YxXTZ`>hjju{W_l?^ST{4s-!}^^(}IP!5tb55 z^GsgLA}(`8#_3)P8t@~DMlLM>M3ZcBVkZs(73`K*XT4ZmJYn< z#71bI3a_TEFanLTbJ9l0G!2$;h7=n{*aMrFm+U5I4y->sfL{%MsKkIwmZ;>VaHm;P z4vCGkY{FJrSHjA>z1@uzQ3 z^=4!;D!ph1w^^~V>}CfBXoA_-g=YB$0$dSifUCATRf}f(a&MI)W-Sykh0gb;;?Qut zb9_Vsk9h|z{ExZ2u$)&LVo5~b*)kVop(F@0sb|^E+H(d`w0bU1-pe4Ca)hDEUE6_n z`u3BkN-D*WDVMpii5RIy(j8x2m|rdl%qYz#>eE=s2=`!P?`v>>)2V;2%*&=sk$}Ie zJBMYVj_wUDaIwRoFrF`pYlb#j-KuC>p_+tSp09j~$Gs``AQRMw`L&@(44HpA4jr7d zp1T(W1a4*4TX$X6d{(HKEAKSJcVmcHQJNjip&bMQLzmMDA#6~g+iHoIFJ!J2Cv6BT z`tdB-6KiU18Y``@)}BTRwVBp!dPk=Mbbr2$dB*R!G$3~UtDEE?gOM?DnJY%HI=(4j z(%KLyqPAF-*m;h~%LI{{ku{_KgL7p2$YrS?xyTAT-lp58;tQhQI!13cDVo#n^e0Hr zt_!p*A;jiBans3&+= z@n0pXUYH2jx;c>CMm_fpgG?q$GUx241{bU)BiS}^Q#)*P>d~(H8kF2EBgR(--hp%- zPwpiQf_*$xb}$`BsveGr##6+YL|phrR`1T$AKW>{S4snDgg$v{mvqal&7oHd@geD%&`f5MM5zvZ7|xsbq7iZJOH2<*_s~Eo0Qk#) zK9{AD^M^H;R`8m>IOe^IrtV+g zjWB!_m)EH_!f-YEqoM%g70hJ4O8cW)|_BYOmg`=795(2SJ9OH0xY^p+Q_#%y&edEBgg1NI$ zu1z<2+{Xh(`i)x-;o0p4j;n3VlfZ)okK-Aic~XKo)}eUpcgJgPd3y0=baDumzsqp8 z_o_QDmL6cMftQ)udX&t$B}Y(#WMvrzB?fG@)hsb>;CQ4L@$vmfFB%Noe5FS3MbGlJ zwhz0moYncBy1 zoIi6M5G~LH?~e~Adl#}^>b^XF7x-p7 zo>=caJ>rrj!mdYOwzubs?$+YsX%UzHEhU~>^QFkrnC(Dd{bZe zA{;S?UHP>Xn4|})bC{94IfOw4CPI~T;H9I|1gft#F&3*SxNUb4)GCneb~V8B49BH8 zA@49TF!+lKSU7%UYEXy60Oe&xWc|^|g7em1ed?3@^(njBK5q%9h=JZxNWyCzlnZl{ zq-I6$+LVsA+n|Hje6d)sc8W+qys@dCS4J0$wf-gUXkI_(P)`rTwb<7aE(b|(_NLja z5Bf%CqL0Zw?F^4eFr3CJqYg^H{VC43vNr#rjJWC>tBKmYx-o5E9%fXU?jF-^mh>hd z!7cD|r-S}W+8(kSsaP3f5;_r=frZsnxsS(wUofSxyWGqrRixtDzUSyvtv;V!v_Pnw z?+|pZeG%74x5C(FpU0O_+C@(1BHpfaYT0e0m$(QAP#b77p}2j`#H4WOfjBul>m0*w z#t*DJWI9|9{4rAwFNrgjeXZ1$PzIT8Q)W)K_uH3Z+~E}o)7`8?w2&K3%=pwdeXWH} z`P0ZLt%2*aOJckj+ESZ0hceqn7BEQ4;9w41@4Yd8jJo=^NyvIig9oKq@xl^}D-L>A;+uwP7rc?Acc%o+i6ME} z^ZR-P3$l}E%1gxQX8_ImLS-^W@E=&#p9HRn0-IS|MLe16>ME{_YUU<53XR$9c(%?( zRQ#e#ndPDe$%A;+Ss8Vl1GC#2tWI2V6{?V zm`1E}g!&nLeG~I5^n_uEe1722{}yP|7gi{R#~`#%QZ(eoSG5`$Wfb>O({?FoS{^^> zs7E=AOt36ZyHeKWJ`tNRiKU#V1=6T|1E_#oK3SQHRQyfy=i#2{F(hJo#CvM8d=;V+ z?w_>q879jvp)5CVZ{*spe{gZ=n@qFcTBhiCxFqdz3>7o4rVv$uH=fPNb9u_mb1ViR z;j+=w=c31BNqk1$;!GZ={xz!KG*?G3_IxmRxI;1Hx}0uBTeB}m#I1pOF_*_CRLoE| zc`0dd$i8tVS)09+`Unvgdr~P54qPmhhvy~7|ybG~D>qwVhw)abW*X_UreJ_-+F)uBX zH6C`jOx|xwF90?^hqKdkhW1P0D??*R`MCl9BU5@W+;n$#ai6gQ50YM$oX&B}>gW02 z%Zs6lL9ImNW0`PWH-KV%>K8j@G}7m9Fj0VbxYTm*7pVIJZq5%iF<=WaU)JO6<*{h? zRFVYIGvWfwHo7+_hA6f|CuQh+o^;-u|cXp+#KVIvcA#(5sDHUr3N9s69* zC8u+9!^9}&P(2W9)>G=bg$!{&ZbSQ8uh=+R(gqMypys9|!0M>uX`92DL3K>`yv&3K znb+Cel~kNKt>`ay{fy(Rrui`4_ zgv+>Q0(Z?o(3#T&*x$R)O3D8~PjA1%KpCcpL6V-Z#CeR3^*IF`b<=-CRZw|!3(vIe zHl60dlkNw}WoE(a2;l-Qzue#iDg~;6`h1Pj>uB6Quw>hDYWs`M$;+XIGHLPs&!T%r z%FFeF?tX|l&xDHs#`qN8CchuwcDge6M(u+XRDX`@J^ktS_6hsf7&IeX3w1HcKRq>#w>NsndddvL;ea-qqyD_t&f>^nJ93D- zPp9$dwO+?BUu@JI2{%E8jPr-z9rre)-Z!Yf8y{{i@72Zs=7;mwQ1Fu>QMjut`bGs} z+T{xMN-#_@hA2M0V+93FlH|9qo)!a*l9i~i-`OhC)CJI@eJoYWYw%9pe$WWp!RmCA z7POq<;`Fy8NT4}MRd}wC4~WkbvrFMR?#YS_C(Av6%!71Z;lo%nB&aA^C5WE4FUVL9 z_nxv-sxx~XWD3ZX<3(jLh1f=X!lde-??G;t;CIT6=3@zC_FwmqqB5{4xgrUE zX9~R6i>I0Y83^rp6;0v6q@>~EoiAsg8j<=HpKY66&ocTW5l0LHz5VMP(YoQ1rek4% zLntIFTg0EyYJLL;Bqsgpkl;9ze7aAg4gS6+V#EWnB#D)&>5g@iSc8Vj{X_b0bw$IKgq-o1d$*8X_ zl`TU7-1aL7St@zC;wrv0K|g63#N$VbfdNKB@ycF~YyGhzW1hxdeA|MWV+ngV7dfpX z&};Mxp&3Ou*ImKkO|SX#`39f5L6r}*6S}-Q<0GodL`lg<6BK^HCH1kVjMSd+0-1Q< zKd33l?0Y|s{<8m!Gj9>^F5spH)pHbIfws#Q>x)twckKm6qCUl7aBqtbvm8J5}Mt@mzsm6^SJ$o(`tP`?*7q}!-b$|C4ciVuN>m# zLh)q`-g36r-z+Na_uuX!!(@Rycx)a|CvLrK4#%!qjQf5kGJ;aW?gWoIMMWt6zrLQf z!lGV2x5QqoUacEzs4&!{L+!k+vqGtau8(u=$f9LXQR;mbthFJ;SK5mu6RbNR=>A~7 zslkoep?~YdN_u*8v8U!3pvdwypJ*f)FUX5 zzZVAqLU)&HD!c(pO5I}0^8ZIy@lTE?_0b=90QgJ3-gbumtbmnp7r_WOXWH)-$@L7~ zo^Gvk7|!HAOVS9_^%;Douu?`I(%*pa)T=1?ArKMA3$fP{!DSn-cj71@>FcsG1hRx2 zhUURCGhOHk1-&@Qqo_Es@tfE5z$+_7p0m7wbJ5rIrL4pqhYA6||EGxNA&YB6PaUw1pM|&h*;=JqNR`rMp~#U{|E4 z8YsWxS39f;_ycbChIRBK+Gs%3;9@`BzJh1BWccs>GDO_s>MKJckmi}?Hw(Vki-4Qc47@@D>8s8lkh*iyOFaW!6E7-FEd zGL8$fhe?#b9rgd>n>K1~`DAe`56DwK>l_gtta20hP!$*@pYT9)^^)r;^D@w&b%k|J zCc$H4WpVA^Kaur&$+u`xIToBwUlr)w)CH5MCoIl^hm*S=St$s#naq(1fQXop5}NM#=~f>`cP+U%%4qt|-Mz zb}^<1V-E^v;Z}I*UBPU+q5WLmh3V(@ zxlyCpqPnz0@tSfvLq8)@){t@72?_?q!vifX62QUy-@Z_H%WN~|QBJIMBcAL*ucLvr zWe=D%hT}*yD$k#Xn%@Xacy#&p;7z8)Kknd26Z%? zSS#>!f$Z34$JgKe?x8}Y0cmxxfua@ihDNQqK`L;vP@nI3NNniBk=Qg{Ah5vEl(~|2 zDp;{UzJ7(6yrh%vG|6rP>lP7h!EZm6qu7+vI`%!{rmzG0%Z9|ATz_Qt=I`Qsv&@EC z$7{keeS{dEa<6J@rt6bXWme!V#}o=IA?HP>oszosj@$CQ1;1auGtLVRvA&4?uhO`3 z>!~Hr_Esv=nX1=Bdo)zRiz#GPB)$R1MKD7Md~Z3u60eN8j1+pAY~1SZ0OOo@xt-qm7NnX}5c;BuY8=v0>?c`A^uo1c(>~Lzh(hk$v`YR%?`p;Jp)vz zdT!MiMSCX?RA_#Q;EK7uZ$z>%?eV@I@^nmKcRu{=B)H$AvMu;^NCWqhCVIrAdSGTf zEyOaDMIUc0Ee4cePA246LoCe=CGmqLb>2>)M@ZB$&ESE#O|2cHh!th z^&sOgk_4f^G!yi>jM@s?U?=xC`*SV1o|UerICHg*;T(u2e%&*50M5`4;X#;U@i)1e z8H+Uc5~1xHXL+Yim-UTlo2z3xFa3r3s+jTYD^)^y1|kiMnpdX+>1%UPNS?#9cf8}= zRF=r0t)!9hk#-t_-Sms~0=H;@+ApRSt2n_D87TfF9!2oBzUT6!5q=FJr?K^H z_nD;Ca4JhG`PE$#aQJg+2tQRxTd&CW9gCJ@ddKb77%qyq(>=5~N%PpQC3;&*Lj(!f{3*- zg*7x?g9s+De$ybz!fxL&yLuyaQ)Zu<9j{2bfMXv)r>LF|A7AZwTBMt+e$|LN|RgAyA zkVB2jZU<8gRmy0QJ@MZGBteExp1Ox%l8e2=j0`Fby7LK5y()wT(oHA+`VgX^;(B~G ze_Gp5fUZ%hb_RZzNPm|4cd#!$f)Ic3i5_%B1{&oH>*~g7*7qWB zvWIc4$h)ZhEY*w8b~mxRR@=5k8?>wlVYW!?8?`&nY(j?|s2mjo74` zIv=q=W37uj672uNme$00!D`Ea--{u!S7^C|r+?Q*#IDat_32W-*b;>l3x($?E#*+# zzWgr6@YR_3vEHxSAgPvj)9@YDljHme|11!^RZ15LUN@f1NVV8`x zmhhH3xLxR)?sQYv} z!X-9-CcK2PjZ`#^_@t|W6Ul83i`+L)WOoqp2)?{u`SMjqGwMQFGQ5^I7hhvLL(DO3 zo+(NkE1rZTwY@#;xv!E0-!=r)R5Yb1iK73B&XNW{87J9)Pw5UEKmqG5b)a+xUj|*T zoQ~mZ{dDbd*r;<{9_r)z88}74E!(vM?|ovvCfnbu^q?3cP>Oosjn3^*H#MwROeeXX0GGU0@?SUOd~ru;36&SAjHXte1fIxabOcrlwy2Eha%~=pg$5pGC*ItNz(V&CVnuO)*989Q zDHZsgxNj@}GEUdSR0ilET}5nS+n-9r^)3Fo=dkEYj+MN;Ncv`cSBGK@D)?;bgKj_v zM^FB<6-MgiLUuGH$yAAFDqZyRb9*=P$tygRAcYv+A8|BJaDGWx=~W?&)_kzPih)el zIA5lb?WEHXmb?pMf0gYm=Ej{H>5|}0zq!A|J4;57B$xpENrkqtwle^PIQogf65S_0 zzqN>^D}HGY-2IXP;|`EafYk}K%OSp*Mwa2Ly3)I$j3ZtysIsBd%Y4pD;g5Xkx5I)) zM_wavkgLJFZ+9Hl%SXH1>OQo%S-}268fk-S^STHtc5paAnr6EO(%XQK1576i#|iK= zBMxRp_Wzt^@sd}m*_B7efYbh&WL33WAZGGm3+H6BsvelZ-!YDuz&wuoc-^JILJWyF z-DdOrb2GKC{FJC7ZHP+nF=33iGt>mYC-f1Casv7F97GxG(`FPxt+z zwH6lkJfkY<?h@AG@Gz0+5P{!s==3>GQY zKatze0XJFQ^1`E&n9KJ9{J9^}25b9t~3Zx4j&uhyzr z=Iu2>$H5odt=l$LV24;x6Cjk_#+XvG--=B)i-AqY04h>vZ+}^(x1&WxayuXHm;WZl z&(LH~jIK9nU1WHCHS2V?CBOjPsw$F)wkgiK@~noqdNXVABREnyli0{o{l`|) zp^)Us$jls`8)rzsppcUp*Ipn)*!NGp!v}!L5^c9rXcFFrV~0n%H5{w+(*nEX;B7wH zhs=4EWdK&^YDH7uc;ByK{G!eJNi)=sc6fg$~e?p*K+3~}o z1>l9+y66qXq>WyD%Hv|LE67&7>2aLgL%P0f!!}&GHj8p#hJt=#4b({oIiSuzk7w`M zu@D+d`U=#m?@L{W(=@(da2zlqk`X$$odz+?spJyU)e6{6{fHsr!jFS0iVe^+uNDj( ziY2uCcJr5*2m%J{cJAZ_bn)Z%vgcxrkvRu)&*O&8Ixac1&lWa)!oYch9|*8DZu1Le z(6}E_5443{KNJ`RGih2LJvmNhM!Qz~?y;5WQQ~#!boi#m;#*6E48jl}^PEWZEl=@A zW+Ayt-wdBk23T*yyD33!IUUYk)IJMFygTbr4Eg9|VS$Ez9=)|(*b`S8JxZ|B`-AgA&tbp9<}P3%-xs5pVJ65ay@sBrDuU z?e^8?NnJNb2qUafh72>DpwirZt?0Sbc0{sY0!q6+G$!cNkn`ip>zis!6~2(Ml_|hP z_;%VPEu8!vWz+X($nU)fAV9ruvt*bGhJ2Zm^N9Mx0 z??X}Dp59X@6ashfk&X+|G%;lnvwDkM8}3HWUe?@yszIqr2lHOP;~9E_^OX09ZA=o6 zi*lucMD|!pI(>X&oU$OWFo2dW93zH)Ze`D++?UDUcipNz+;I%s8$h+zMchNIqqfj| zYW;!{FzdbFj#c5hMXITR&gxn8G3sx}?}y#oW? z0`>ta!X@8E=U{zcrSv-meST#Uf$6zDN<^-cgn(mV^_)uPo{usSzTu(ob(2YqY)b*U zeHOWN{);YtCvRULelJsGA_>p%(lWijAd;+!zg@lW2y8gkuW~<_qOm}E6Du%EHVz-! zVj*rL9amb}8}aVN=ABI+i4Arq44F5gL>eSTt?f_!YCD{`R=QKdNrh3pCxJZ*KKYko zlM&5!K$IMH^Z{6`CQe8F<$i}_mn0yy%(bLRssGb4hgqZmk&u5RZRI92MT|aw;5c0kqN?2q4=JDLdDV)95MU zyQel5vXOC8JiAfIkzYsnPW5>o&lLAqN$+N!Bz{`r0~lWq;8M zZ`iKyXtTI&-pS2QuY+Aq1{`NF9gGR}qL6KRJ^~SARGY}C;>c7p{RxPyKV?-;E;EK8I;g2&I5%^9$ zC{sQim;n?>_r;JWaub3W$wbQX{KB(XEiOFy-e2X{-9n5ejG1ozm2Mk2Hsy4DrSJC{ zPh+}-B6&fD!Dr1%my zUazx_&BG#ynJ|=LxyB%3F68)sEK*Q2hyA|N8 zu4mx8d;Dw8`q@PN!0Sm^|LkP6-URI@i2+ZA*|01Lf0%Qp(r# zwXQE*xsHV#790p8ihYysyOR?%p>pOs$BKQStUf=G{H#lcqC6;>8v}yeYHFE;3lWEbP4D zfUZ5TCQn{M*-^#yqY&c40fh#K zMRQaS>{TiPD|y8RH@gC{=lUq|KasI4KG1G^qh|NVkz0$$SEhl#L^F?1fSIwB7tN%u z3xDcR8G2XN(1kN4U}(~YU10@92}U>K_L-VqV|2fWDLsp2>M2*#IC>k&D|*-cBPh~> z0P(7ZF6>pu5z~IVpHyCMsvMb05xf<4tuc$;HhObMw9|QG_sr*^nAFf%BLdj&>)sf( zcxX@%2S1|%ycJ)Ny@6QZy5aQg1l~`>fHF6%nh+|E^%It_eWW`X1aydTkjF{DKGBvLp@iosn_`r1-iDX&W#?X$yoMFsWR5utUeW$w0`JYMR#Bc zW1|KgN&<(SG(WwQ*XMMN7*g3-?Hj7*5Y$F8kY@_JE z>Ew#)Q%sZCBvI(>+Rld3)iff3TYrw3s@sT;P6!={ei*QFj)(TrBoNMBqM@y3XG;rW zSp8t)%|(i}H>3Z2zSuw^`h1aMqM;QzX=oo-H#2T8q(|q=X>nn8uZ&R{vYrFYqB`!l z7qRv;8$3zro#|h)$u^^!VxkWwSRKa(FG~fmp9#rkW!$HI{2r!Ske2Qe>28(op{1Jvq`MI*>F#bN zhVB@87?7^#=l`7Z+-u$Uc{OWZ&#b-ocVE}_*^b@j4f|$8c89U4VTgu`-t#~(F7PJx z;W-SDl&y$2-4`sx=EEhHKQQ*RN&S2JXN!+xWJu+nd)m-NHZhh=yORj10ya4a?LEn+ z{BN_;zedBl!sKGH%k32Zu6ZRa{^m=KLn7kTEB}AG43NNj8GtW|md?9rnpD5p`FQ(| zY|?z7*lt1qT>DDyOY{qcRgt3Bv)nLxM z&WEiGsG0%Nf`6Mf6gsGqNl*}74S$g(Scgv$uEHTN!t(z<*Btff)bMz0NXO(@F3?1nb0EVLwQj%~7$(RmWLYu1a%l84N$?D6%E2;~>0=z5rzhaJiYV`ZI^SJ@>!J%9`)G5*wr5@-C8X(JC zUq*$6AtZ@x%Spb1-k#8q>SP7vAr6AG@-U#9lNOSS>~OL(~+zgfgN zR)%+)PgOOaN!2sovdY2gFq>gtkgH#Bcr#31H zlbv&rv4UZu@v{Nlq-guhKlT*f%y9$-Ta%C^yhI?(+>8!l&v8ypT-avj6&69L*+iRl z=ix+|LU@b{2{9FZZ-}DR-?G;c)fnDbdTizS$G9w#zpMr!330l1b1gioyo3I>&L5In z>o!PkFpwwm7fa$sgY<+!C~F^PS3FkiFcue*=``Cc9zOhR@)Gm?w`RM|{!+U;?`Jdg z?XjrmR8~+m48c(7*WQ9b2~RZ_$N5J3jXRV&ZMI@x9eE<`C{cLX3kOMo6l|h2aL5yU z>RMO`tHwsIg?FCX3QCWHl>W7fGz*|rSOTqrEu>{|71izA6@dZyr0%cziO!&4oN$A& z@xb$EtIz2pYhe>RpA0VfMRjW@iUXLg*TIl@itR^wsot~Ag|P@sxL+7W8r5WGYuBoo zZBK@w+B}KL{qB6NzP1pP^MFOF^)xZUIZT`@$Cf}z)*puU{L1LTgFU`OaGbtI>k4oHuU0!$oqRu2l8N)YOOPvkz-VTmEJu;DZ<;epgv@ zwGPe9S#lt8E1+b69l8*6G;BLVaSR2Cp&C3Pp{236P%IUA5y}ed?O#GfZ#)V=R?10WRg555%F`mJh|y__;e|( z3ax-wnlB9UhTqyU)@F*n)+*zqUsbI3JY6TPr_)>w+CQo_ocRzKps+m?krqUEbSGV* zox%6z^Q%Nd9deqOCk|rYrAoHDNS6Q34|_1a5VGLB$o1)rPRQcnl%6Xx4CW5~0roQd zKTZjIju)>+)*Sv>GpugQZ)(D-CFOo{`^qTH{xAhx*uhTy(GI3WsrAgk+J9f$%(di+ zMNU%**K-&*%lt)c2!HH?@nw1|L`Q7I*B8VYY5p(ar6vtL5Gd_c2!Huu&MWvU(1ZAl zWacYdv>Mfc7W(ASz{pS?N11Lhkj_>NL1J*`M`^BT2E9m>yIkvGlO|Wcg2Ejl^1uxm zz4t{Smv&|rTG>cFWQ7<^paa@Cy=(ML)#qWpXgGR``=yCE>Y~?WRGqKWd`X>xVdTJn zZ1vG3L6gnhbNvQ+r`8swri4&9{Cz+10YOg~jH^h51dY=MB$``CEYtDTVEg%G$Rw=V z=Z1WwNPS-r&OG6rf|*hUbJ<6M>{!h%XhFgO+m3 zXE+KS>tb9*gx6OvRP#Yznel#qXZFc`X@c21pQ3fkI!PD!)CRs`Q4+x-n7~dMl)nfJsFcKSXt)n#(DBhXZ{=yc?6;4gv$8 zx8vHiAI!C;1k9vWERUG8PDUrx^U4u~26$V8S-EGI6>OZy(ML z2DPRRmTJS&{vuq{YANt=^&IA%aDrG%h3SA9|H0WTm`>Bn?^+^LC>mk3rPy*j0{MQF z2@wzHJ&R+rZMm8vvr+5lm&D8V_9$Wzfo9r^ShaN-(sQjpkcd&q3g1gaM;{+c!LzPCt8sbFhNyv` z6Tx=LEtbwbn(VPjw?&90k%;e;*;N1Y?wD$&-E}d7s!9!ZqoZ%F&$qiusxLY3eHp?B zAsJ;~YS8Z$456=0Zg}8Z>{F-txMdmuCq8O!_7nX~Jv_)ZciwLH#!mK+k4;}(H^zf~ z<(IJjNm;)!--Q$tRj2LW-R_(T?HhDvZ(ocRPo&yWrZkNB`L;O2+pD?Z>3{XrH*W=@ zXVJatJz!*J(FPB%RWB|^wnN~>;vcvF9&;ajzZdWK(LEN36u-DRQR_aTVq8w7dgRYO zkXV=E@J-X91WZUS7wp%{V=J7c<}HVNuAMjP_PdC&#z2>=-A(3>yrUiEEP7zHgnO;-vv?hh(0|YH~K53^)pl-WgXGEjUd;c zJ8CcQy0~utxYgc+Dej7Jx70;=Q0oABWT;nSfWZJ2)t&nL0E^esiSn8xSt;m=-$gS8 z>9{p+b`mqxesR*L)$0sMlc;@jG(!65SyxmQFm`a;_2cb#Lor6TW#%lw9rm_^8*&|x z)%6Q?s(v18{N>KU8$yUE+lUr-K0~KiDi3?YOusMk3kE51WL0NyYS1t3TABJpdJS^s z@h6}`)qt}4%M0(+zbg7a97OR0ZaGR=G&5ZJaflMZ!Ii5zX%d&%sZP9`c3m)O1Z3}P zl@fgf@+~NR*4iLJVYvFXYQHabY~yFWfStUO>_lV@$0i|7@1_VSI{}XfWzw(-4^H9| zLzSG;>VD!g_U&0`Ee$)YU}e4iIYf=A4>(|6rhRM>c-h_Ru*}5=RZ~Uugg0u*S`|4+ zgy+OaT};nlCi3-e!~Jh~6*Dh#$OFw-FwTklP_MsvOf}g~cHflrWja_H| z9j|--aPODxmF+iHj}0l%QO#Tu%1NxD59}fQJ^ko2n9N4o^fv`d~4e z9OVm(T0XfXJI2EnpE0|SNla^3S+rh`Sp7((Q|&G#X1dwQ4N)`rD80GAw3Ck866#K4 z74FQGyuf2WOwKMF>5!4JU9JOsC8ZvhCvdkW8tvD}uG|IuX2OvjMmeH|%r1nKJ};5} z(qj%KUA|8T&JMaTIIE%}Nk)6}D=Hi|rJp>~TzqRsldZ(*{WFU%+vaErnBgrV$|;r_ zd2wSXP4CQxB!X@)s(qQHPJpOudIXOHM`D;3zq^#@lZ)jAg!BjzbNL}N1Hf%i*iip| z_BM05exm9srSn$MnPOa|u<#1mu9)U&r==blYK$@w zsCgLgfXhnUOK}mW9zJ^hbl#>DyQJhc8y0*eVGYCKuJtW9#ibPey31i;aV7RdYxk;G zv~)F(wU+$+-f4WrMRpGN_IPz=YluG?8_4qo@o>6*1R{l7;g8+E8Yv}*l7?@d_EH9S zMjKhKmO0oika-S$c`DK1fZ1{=6Hs5e{Ib{9s*U>9uoA|)var?GpAsqxbdl^9TSN*R z{3@FsjT%}9@XzjBlrz}z$t3M#0qK?0F8*q+!Oe!()Zn^pSXFWI9MT1oM*a=j1PKTR zS&DChYo;tRDo2sv!|KNFhMq$?r?F!PP?X096EZ-r;RA~uT`gW@9PNxG^)IO$Zw@&2 zZ0^1LFK<=A5@vv}ghjj8@lRo*fk|4KANl%mMt5LA^+pnc+qeIN{OLN5d3HNZHl;1gI+a^?{cMkd6Fpmo<81chjn4kF z-hS7GP>Owd<=j1r?`UxC^9Df~E|tVh<@uqfTQd9msIj4Yi7u7;w&O>YUruL+_@20v zDdQO{YT52rL4(xXg5b`&hmKd-ll2vqS%P|;@j5E9rEkvGLdW&TrwN&Kdcb%~7H;?` z)Fn+wH?KKq)dM}ui6W*oVpw}OJeQ#>7dAgx1Mb;LnG+Mo4j@66&aL+(b407{+{|Fjzn`XFqH0Rq)F-X`zg&a z_TzLIPv!5bDH0$@z0uS-75t6tac{0Hj3GBpbmUWaODKt{ToQW*KWB1H6)1(Hd%WJ8 ztRW$?OdI*^P^%L0-SU9Q2;=)0rp%|;H=p^j3$fojEP-VBZKj62LKr6*rHvZ0=H^oy z(^5CvG#Wgv@<&ql-J&(5VoGu7o|?H&)$%092tRx>y7)SW)4mgJ=y)DEBk-82YRT$x zG_O|}K6bai*lezF&7W3$=AG%e05o?m+Voj8|NDe@o;!Cp4`=_jxpIx7Y&|FGg;c$5B zFPyu|cQZb0^BlJ=4VP+WO({)O5sK}7Y(PCrYW(qBr*k>v{Ul&${mGVAet7n-QC_py zXVqgs!Uy-D(Erk4EKX_snd$87XnZ=}aE zI>rQ&hu{qTThANhH?kuLc&p29&$QKKz1ZE=34_d_UXwl<)nP=D)3#!@AWrq3-$Fs2 zt-6aDx+%ZLf0akm{>WFy2Hi}bP!JYt$WBtM6Wq>q^1HD!wWd?Xw}-L^%Nh_7Hp@Oa0|t!M7NsZ-z_wf5EyLfhYG#Y^MP;#d%11NoCj ziQwj*eh(Ck;z5b&U}noK$Yfm?t=JEaR}5! z8AU|wQLuZgI6s#3nNjbqBu6%uZVte zzUl}EhtxcX*;v#aa_eWfyCDu2=mwW;d-v5O*&Ofu=|wBSqKeKrfxFbW1E3_mmhN)` z9P{G;*S z5^A6XoavJhJe7B;Y4ktZ`v;vw)e5M?&-~Q5X*KUNJgK4%Eq5y);u;ub>WQHh~)8b@PQPWcr7s(R=Tvpp09JV}D(TAhj>x+YI z56KcD46TAN?1->tLlyfXm}Ik~$=WK^g*p;GgOup7njt)Ue~-4tT1=Ey&at_Gk?@M1 zedw3E-E+<%5Kp|ZG0hu6e2SDNq7m;G3i&#;O#VZ4!^oVt=&OpnNxhq;*tqndlV}P7 zVf!S}EZ46QW@Hk^k#i{$c-zy=vV5j8*yNc`b!q3s=}2dpAlEO~ZplD8Y z(2sL53(6?h%BH+vBhhuY*2&k7yL#C+vH0Los$ z+14!*51)|f@=yB6*)%!PMQk`p4X;t~sl@(J`CmdXnFRG17gQXlmHy#3FiYF{_ubQy zFdY1{wzO4z*MV3DUFR9B=n%5kUqBvHtC2O7jN%%fXutAIS zD`%VpHSA8|$;nY+Sbm|#()J-gb|VBAi5u+DtzpAXvoeG4Fxvq{gC89CGh6eB`qvYZ zT^BVOEuFXCK1j@D(yshH6M?=jGJ&~eId9+HO#AX-rwF{*hoT8f^WsQhz;nb7O6Uxx z=hkeUEK&dhoQ|rz=}v%Dr-GdHU$HnN(h~f+Ip+yI$DwZ>nO5u+i}kOtMj845{6oEN zt?73P!hI_L)2VZTMno`{!Un_uin=eVBKV(SaEe{EqoE;+>G>1pB*cTIb`6r+jF;-8z{k6V<31Qc7mO?55PB@3qnD~ntG_jZ5I1c01e zC^{$J5a33PZBY+z%yKAk-J+#kB!F3ig5w=Hztz!P3ge?IG^4vPW}{SHPS`XlLfn{b zGB9g)ad>UN@$-!YzCTNk@~(|-uB@nK=SOzMa;702a=MG{{>!B%+eA^bmi&touB>W* z@xXV9ar$@ku>juwK3tl{Pm;g~u(-BP2GcY~)PyA+3Nq*loTufKiB;%_l1rO(Kucsj zXhtWAk-X9D#I^_4Ei7GqvkoVFdTOddKb6Ndk<^>b>M2@Sc07!2L5kbqTUTk#(Y2tX8}B**Ql|1S6^=H_bI7y9Pq*VXHM>y+va8H==RK|rORg&|0=k}+o`tHrKUKJ* zgXyCVvqLWISZG;Yj+m(?Yg~?Hun&e)rv@zD6sfq=dkxn3vt1O$;)sp@ElL9jK6Jen z2kCh;TEHp_M*J?)NJu)?A(M? z)b;UVgUTzBtvj&gdV&9>y&-RGhCTA?Y^-5xx%2#IPk3CY)wsUNzg2_C#kB(MezAO0 zkVLq&+V8)^CMx?caXavmGdH8KeaP=UYI7#{e_Kmfh&LD69cmS~;GOv~QA(;u3e${4 z#aay50B>N#QE+bXyw0LrV%=xx3=1p%DpPGTt49!xCyl@q|59sHfI$;yo(!7!yrc(f)6r4?*o@6kmB_Xi4NQQoxO0xag$F1sQ zj)Vv)hc#3X2yaO5P`N*-Q0RY20V?=70Q;#`_*g@x&rG!)NRsU(A^h=S;Zh!idS?Os zWNuSDxfkvcUC`L_M)Y-tY$$$$0hO|^9771H&FM(w3=ZCRjU;LKnh!EhDXyxj5B-;b zlnMiOm-?jQjxl2+pWBE+Q^fl+-{XY}YL~NkNwf^H_M6yuCe?n^#I41bqaLkUuo}K0qNq=YEe5-MX~_&! z8!N(;s>aIX{LxIq28+i+8Bpa)ybt{O^VQT)wo9pU42KBwkg#3evD%Ka4Lwcr^&PY$`GCXSE z=DRn#y)~hMdzT?chTHloqr!vZ)jhkWX}hwW!A)-s?~9dRXJBXWEs4z9^H`yv^kSkL z*=PN!GD%H%%!1Ur@mUsAy%5{I#cL1p%~Y(FT;fX}8_9n6a$(|p#kYp@urTg0*yo76 z$ytU1@%jjBKZa0RM!Pe=2X=T1Y(2(>zKas@tm`g8Bs$nQ2hLtH+)TjFc68PfY;ikU zj{K>QY{YX_U^nSQ^J79?T$@e{TNu1j4g8|JD|?4S){rAyn)VhC>=o1l&~P%6X>7WpQ~KCXle1jMD!UBex=#n z~H^Ktp#269(q3Fn@ohsPLHM(v@;1!nKPdE+D^N97HUxb zcHA(BaZC6JI!z-M-@4tt;r;*~d2Q@v+cSS31TWH2UzK#E++N{uj~okL2$h`y?MNq2kiE~7wuku36-lm#Q5@gl{+ zT+&~VP~H+TzkcPJG739Ck>3>IqPoe=@4!SS>glp95h*%MAXann)s*iasrVJ! z42rVwa@E!sJl4N~xdqI>q7m82`2(CEC+Tim;j49B0%rZo(4Jf%HAiKnU4VR#xa8g6 z+Sd~Ie%!P?YxR#{-jhyZU7Mr1sW56>WfZcXK?l6>QEgV$;F}S!7bxi_FpwhBU0uq2 z^?9Mq!Cw|;^Q+3Nu?KnK+IV;0;ob|IQ88V*En|e=cf|Cp=|LW6^T5^g%WU)1z-rQ> z(Ry;e^b+Tl}H%`yv7pBS3Lo z)*C^ImgY7Mwfl%GXuVC7g}`Q50vJC1uXtNM_H6;Qmx14*_4S+5y2!=Dsqkye%C>>P>UP==@8kEX1ooUq*8qJ_e6SzrRi$FIp9C0N@;y zHS*547NQ=?zvoY3N>p^8dnyW+X@fF+w;gyb%p%G0lGpu%5gwcxLhVM2MKYfu@JKR? zyPbMOK_QZH>sa=@K|{EjCMmc)T`7;!VGr-w7i-c~gvcx}Zis6aDWBW!8w8^$;z)p?0?4c|oAb`7ws^Dq8S`70gy z7=HK8rtk85;DdI(#``|(|^Q_t-?kASLQh%$lPJ1o~3dRS& zbd#X&XW@g**C+CN;;zhN$1#-w1c_9*#*s_cwF{A;^bUW>r8VaKeP!Dt;} z!3;}eSH-ukZVm0R94p{#sWyv$BiVdu0@(~ryKlJu#*KXmk5SWo=?#4PvAc3SUicI&|qf1{{NdpwIK0 zR9Y8)Ym=D!w&Ni?TF=SEiZ@+^6;Q(L0Aw2Iv*0HF2qRo2-3%ihbn_wMBBV_gmx|U@ zw1(1@RSV=0igQH5NoWeV$;cHcCHA|E4olS2|5b7UXAtQ|d!BcS3Wj6eypZKB8?W;a z^|a`^aw;w#NQ%x4=_P3RfL4JOw5jf)SFC2`u-H`XJGWgcnFzbP+W&H1Myqf!CAM^ABhW8S&{SQsFyB|Rq{QI>=0^@C@rcDLyTNuFl zXpB6{w5HW8>m#C8gWepSa}uC)mq9D}F;h5H)8F)|v@Myn`z8HNin9;H+OuwtZ)%7l z_zU=zmpb^qm=&Gw=3BBB9#wOj1kXq&>ZtR%Mr!)Sdm0^R3ZBKxbQ3bEE?vp4gpJKI=t&9#$%Pl`G5`s+2%yN1AK`XEJyL^T!Ak%PLApR(i!L#N7S_X z#ly;^48u^}hf{xj9HvReZ;KTLz9pVhCn@BuWm-L!7vG1}i_%OnxQSP;r!(lyfdt%Z46 zRL=09c-HP0A+Edn=q{3S>EQALqn2vX?{%2?$<@9yzq|ACmW8v0&xz1o^ zrdp@BtTd?XrN`bNO5CEz?N|;3sI~IAAaky_$hWA-K|J80o+eDGjyX%pw_oSgO5yM9 z)?}WX6>jSpcyP_y%sU=|EN8;hsqC*eW3`B9qXw9CD37a>T(-E$@o}%2tn878n#JBbOF)XZxsQv|sov3SWU1oxVov+2S z#P_^2ttWqNrv_EWTHwM`kE;GA@$ENqL(_P-7d(j2t!? zL%C4ZOPG=FP>ST84fV|p4msg|CRZOQ$%IBY3jIO(jQ~sM;LGM0^hnt3k1;B z5|Gf5L@(+0#UYyqoVR#up@(Q#bW+j!R&GMLLslJvQbX_|?soQgm-8j26!>pAS6sNm zZZ4>d8xQ3#DYN|#oIF_1fGMbbm}=n`J>ar*a9*bU>an7x1v1oljbv8RutZ~JV6)}Gz-1@RULK@@4yuDw_YqjLkk-=sCH{^k4K)yr# z0&mN-&Lh~*9(o@F-HV`F6$1dS7t`yRqOi{Ul{I9}ut%$hh~FKlov1lXZ$GA!l)WTI zH(KFdSJ&v6rYD5{`L|3oSDs%NOmFTWjJ6RN-p-(UfE>cmuHdZj7Resb8;l#3>-cD- z8Mj?_@q*W5TU@9$PUpn(*2Om)qnu4`v zslaR2OxMiunJJFJ?&|#Oq&0-FWm{kw*&%n6i==@;_@#P{dIxMmzixSz@|{ndPiseY z4*4sXR*T<*S9Fs~8oa$RnwiyKG`4Z7_OtGP8WD&Yd2vJ!u$z&b&f(ivI>3HNBdIRj zv7~aWcn|Yizavol2$wk@#tv}N?vI1Ed!&`kb@E;?G;U~V&U?~Yh_hmNV&u|zB#;z7^ytY|1@B<5 zfqtY#`K|Q-=w$!R(WeFg%szic-&!S{`dEZ21*UxCwVu?2*y}jS-VAyAU_`}E^+Zhq zThnUU{?!c8+76Pqjf_V;qhTt1czQhV3|nq?itxBSt5wzA zu^b7FCO|}GH-JMT zDm+18fEf7Ja~N1OS~2UxrY~vXLUn(4c|e1lr{I*TZl$C}(gCJzeGNSOdFb+WU z>Gt1HY?@BS>y~o19lk1q&bK`tdsOM}P$YuYVr|*newBgb@_v1fQk$eAcroKGQ9K}Y z3Iyp#S=6-ra%=t_oX~v3=(|V}30{yZxwK=?Bans}KOrWQQxy$mnn^>1rUryCCtop4 z3P1k9vr1m#3*X$#*@ggbESgBe#1)LyUtoA+VEM|g;STeWkyafJIC+)yNw3K%S+=)x z{;_}HoI2%QiSYx6A6b4fxqcJ67Pve36ERWZ)f7U951_jU`E7`Ny9&3aPn_UBSZ_Pr zr0Jg}T=WI(c~2}jh(Pb$Mu zT(`pr5iPwk`8XM{Ix>S2VnVM8yN>`6@Lg`Dr=g}o^az*3z2m{#!=y;jC3TZsbi&{mP+3r_JE|15pbSYvBmb z@jlrBYbA>!1MsM+NRv)DVMc>(lG_I~MI3pp-Gn9>dUy1>_Z0=#te5Mkz!I>BzrsKS zJ>vJGlhSw9f?Q;fd!^GV4{Oqvwy{3MVsJHxcbPqGXJ^rO=j+2%pwlGhys!Wao;g+H zhwPYM;WqJhFS9r0l@-1F#3;_FTi@v_HyO^KZc6;4o^Xn%J3Tmg0qDcErLdCdPz8SefVq%BR z4BGpXc+s;pRaJ9hV(0$%+DXyGLOB4<2$~a@MWe3l1NM>dyT3Mmx@B!Wapli^PI#-^ zW-3HRXR~6mdXi07`%(2fh7JFvEBD`kFwMu*&OcTQhq)BZ?{33sf{yRsr&3OyhT;jO zwLXqIJW`XsOW^aJxb&>-9H~=5J|DV*1KHBntAS`aHI9Kyi@V|A?p~HVh9jD{HdEn> z@+dFJfiL^hkoOL@q(5^xcUJE6K2u?F79X*5&-v~zlpHcpmdWq77Eou2D|8!9vzL>;GmSbI^Zv z>~S1R_5&-q#B+MUT{H^q#(Yq9)W^+JaWBu)6;rCq5*mHpWY zwUx!JOpEfPd8~Y^)Je1cZQkG3{3Yb)zF7ZJFL-31IUmG5`i`4%;~ebXVz>zGD~=|y zGsy&bv&O%30lynimRh&h{hru-WlvKZDeja;X@LK+OkrQ@n`p_Hh$UOvg%J0O^Tpz9 zhPK)+M8dCw=A%PLtNWI!jRti{}MO-Pu~t0!v{Kll3DJraVn}dp?U|V z5iBI>_=8}EJbG3}1s5^!z5qhBsfT6c)F}h^VIuI|HqlT#pn;`2vMLO^SFoog;Rb;? zX9!THLEWz_HLkHkogys_I%!{qcC2F{f0=Dd^A~V{@k*HnLozD*Hw7b~#Je{e&Z+rT ziPC&Sq)3vv@DkLlPTOzu(_Pf4fB<;@kFq!6!rLI@KjZ=;-flc5fp&AO9= zR!ouAAUuI0_fzR=DBllkK(%OL z5;|tboQ@Jm9Fq%k8M_X~Bdk;MR+Xr{7XP4zTq7Kp672S^PP5V7uCzCc0DXZ27oBvb z*YzK3)JPGfC0zq@4QTD73^IYWT14a_ z$Ww@ib!3cRe=o-m&|9GeJVsf}8LU*->S(BPYdhKfp2D)FzpC)if+U_~jhpoo6G|qt zt?y__=0X524~_ag`a+~2rYY%1KA9KSlBp2aS05ESjaH&$#}Ent@LsZ&PwNZGEb!zOep8}|8RuOMqh}rl! zBhMf|&N;|JjO z^FQeD8POAM`u5VCpw2mSN4lv)v0}^R(-+EhQ_r~>C-v)QDo1rQ=w{H`xXswQQyeM{ zA|N3K9xoI9Qh&HCI<&WNM&Z|Mx+_^^^%+cKxRh|4QNQ5PXy^fx*H`TrUpdin-Woo2 z;fhGU+1mpVcopQ*0q$(11Q_8`Wd!)B09{!^Q^Mms755vlMzb`CX)d zx27vqDocSO&W_vfgSXk3qc05tNG974N6!!r2=3NKdm9W=Kf4Q!1-vFXKkjbk`NI`5 z4o$kUnNVX>3wWPklhCvPq&~1JzYS@ujkqbCk}m5nqzM72;M<@(4nDaoT&>y}kkkDdbRMmPXzxKj>3M@gXH z_7vvkosg)bSCLFJH}lD>`HQ4(p*1z^nt5?PvHK=^o^#+m{c$iK5hH118SBF}rm9G` zRt5w1E_>Wv>0IHs-wWr5rNiPJ`DN|E^(O$CDc{+;`1*tEn#RY9g253#`raczLFT#L z9lXN6O|nd&^5z<|Ud{{^UTUyshufHJ^NXTSn{dnS-m)-2K4d`I%kvhhw!KFI*z~q9 zgO|~24ZGD|zG+a)lo-)f!icEg-8@+Sf+5AV)ouT?)ug^HkOBWPTsT9=L~0lej&g-B z{xNKjg=PU>yXwhm%fnbR3usVSY-Bft^LQRwDv7x1RiPQ~xZ4<Cc z_V|6u8!gypeR1>Q+hn}u=&~h8lnCSQ2kk1W4tO%e`SWYQTvP?NaC26M2J&m1i>s%Xi8kV5clVey_nD26KS9RV zRIhlI$%}-`v;3JNY3uu(&mX>kq%6}1k4YcfhcznBdi)=6AQd0wzj(>8#T68WzNn>B zGIe@%p`o2A)nqO5m{_lE5QpLT_lJ{! z1SGE4CB;U#?jN3x+R;*NwLRIJgkWzRJl0M+1ej?9=!4}X=mUmQO|S+?k#Huz>}Ggw z@pF!i@lS%Yxi4#-XkpX|-P9k@kX zRpz~O+%6aP?+=abOBjqK8xCpd5D-8g0HU63a51;wwy^*%YxWi4Sv*MUgO z^1%Xg!j?&Mv4aa+X+^rUbJ!maUe&OnDF9oy6e|Bdr@CFG!N}^Y8Nud?V^>H=LUARR zX&+20oqzRiu}E(1HFnmJ?@=!!mzE`!Obz4RuTKMI@d8rz5hV^gx+>lx1xWO7b8Y+k z<6~iIC;HZ~-3y)Um(@m6IYVWm6Uk`{0vA_m=>_c%><&|XuU8dW1`*$-GQe*xJRc7Z z(qCt#s_E}JC{wIox-PspAPzd?>O!3^d8u4{n5r!lS9xeEs4BXHN2!`6oIYVOoeXV0 z$N9sQb-q~lientZ7mfzmdq(KT^4DmAJABOeaeS?mBd%eeC|lY5Y4C%#a!iaKf-S?- zT8NW{*;zFGMBodH3sS74^&iT2)I&0Kfj^trS3R@tgdbJr=6C6GZp#^e&Hr>Yx>;EI z(Qun~*BSi{cQTmQyO8I*z-4@nFsj)aE?ysuDA2Z@+RVr%c63%X_KyIJzM3=JQ;KD- zyQOPg1XB&phFtvvWwN}yyD0Fp*Tjx*o&x*ed-}fD)ET_c`jbSdR@t89I#(&C!id>?kxSdbG9d&3W1*t13xH+qOpy3G{ z{6KCCBYPXvO`Y)zs`g2$#^92D4puKkuf5w~{InZH9rI++S>=5tv>IsLp|$p_qb%qk zi0UDl@v8Io6{qxR&|!}~{p3co!Q}0OM+dfBCx@+n&HmYfC;U$I+eMC@8|=98XG+K3 zji6z}A=T60hxGChr>>lbKSFc|PZ}`JQdPf8#oKHCr-8D-?PY0D%u}zJKap=wAmh)D z-K49~q@>7-oZICT*KGQ~9f*r6i@sO&5s0*S1;b!I4LnQq917@-^$F*i1}C=D396w zoZ9qp8{Aj8(B~X`6zntV=gip2)=k%ZI%a(`RAyKD5(3r7St(z2xq@5OqUW`_-=fv^XOl*gr%H`nm^s8XGS9^Yi6nn^IfbEbAjsHJq=G ztmj_9F|~nE({5(Ak`a?fFJe)NdZ|4iFI()|r6BzUdSY##x) zAjLF3C%(BzaiDZf^4QI+TIIU&Tl1wj1@G}X-kW79R&S1@g6R~e`;kwcg$^;ex|gOE z+-61x)5OI*PO`48%opD!V_9U5EPo47zlK>O#t!c%Rxq!U6}(BR3M}jHl#rp0B2WKh>_3x=@b1 z%TPkQafW4#IaC@J*F37f>|bm&(#)XEP5Wk=Q8I|`>L9A77NXc5jv9+Qd=c>0(%ZCQ}> zaEr7C{aISo3^$t0WiqgC+rQ;S=i;wKqU8>k&QIRciJv5$Z6mDyw0fhiEBxo4fH~t$ z1{v~6^W6fe7-Y+E>zmh;m=Aj7^2%3=Zk`_&BW__OlFu)`aL}?UntjqILk>__n{1&o zV0gdaO2)N^UF92c`xn-k@Ar4loVv>B4@;Q{2>1Wv>z%_R?Uwc7Boj}ZiEU0MnAo;E zwrx+0iEZ1qIhj}!c5K_muiyQhckg}nIp_N;S6}H~z51!DyY5w3HBt%)_PxQ=#Sk<$ zNs|!iI!T-IQ@WIAs>YWQEOIYLsabpaVqSCvTt>rU3|;iQd> zvAp=jMcS^=)tAmTiFVv+mCqxkxxOmsIGFfbAzFo-|7N<16J5i3kQZ8mCenxu3d&#B zDDXf2Yzq;S885UtCI>PH3&`vBM3;MSAI3%Q#YG39V#}W5itE?R4OqRh4eu3!`zIx- zRAIVYKVCJKTNLM+#|IIk1pD1q8Q5zl@E6*3QQBE)ev?}EZ~R3Q1_u$)ODOI)sRkhP z*)56bJ9;ZQM$dVJ2pDv>Dw;K=j$tD$#mu@k%O&Q41;ym#`HI3QFFb};?jMt+ub}?F zozDCOlhi_f+Z5LXe{9yWlk9-#=*V1}oXYBXl_P=>drC5%Oh2~V2$IA%>ZQ=kFVx)- zF$p1#C50G(jl5c5?@{&M8q90={N9~D^{Ir2Z@ z5<-D5*o+j9IZH@+qt7Y!y!QrbQlv{wH!dVo-Yu)G;jxAeMO=7Pt3vwRmh>pReG-1| z3<&UY*<<`q~lO zq-ONBpV&^GP0eK*`B6vqDkdAe)LaF4))&`)?XNJDVjGyIji;v^#MTB9F9xn|ianA3 z8$kUxxM~aV>xz%1Gl_RB3YzvT!v?27m;8NU9Qu+} z4(i?ZQ=0l58wI3?0Q~UXInjkvI!MOF#s$Y|gta`LPz;Nn>N-Jk!H1kKAC-DlaW_nm zX{XFnri$Ly;3}V29~z{}D_<_B9oT@$Vs6EXJ!=1649C7p+og{llK|Aq(x2(urYDvra$^5GY2a{>in}_03CAge{CimYtndbj|`zAYVhr zChEsqg>E%+nw~lUJwe$LH@tGsq^P{OSd(=QNaI17PeT0y5*@O?$Z~-(NOZ_`Zyus+ zSupdnww;0m=jY?_U`vzV=&+jPSqucCwCuVid7m4G=2E($m2UUYxfFCMBz!V5GJfDY z-j_L9+MBCa9Ce;2J(OuiO8$oQ*B173Y3pbpJ2z}>^fkJJQ`y0~qeaT{xh4dZr3hc{9D)C6|(i&JvzsX}j z4_AiD<+HW8CXB-;x~7e{@8I?;X{i67c<6`)PSUV{-X+;?Jmq_zC7F{Eg^oj!W+^l< zAi6XxcfBYI>Da|^PWnCmj!<|Lq#a~vFGswpmNe%+LITeVZ%!=_P8;bVDfJkz%R&YH z1?n6SGe|)@4pnbl9#Y%=mUxRrJ)XF5)l#;XaI^{Ljv+>`hc5`yH>q>Xnd2ME>Og96y1FYhgm`_MHiw zJ2_d2N3`x>WT{7!^n#K9e2U&6u*k^32c8x0fk3I0eMvRkg{~oBlaBKLM5Bnm(5Ry* zFjF6-raI45U0!B$ii5=l=eD;4G&OY_}kwro{~a5=~rQ5agJ8f2dkQ`t*TNiH0#i38=nc~{(=0M|Gl#Y zFr`@G1lTz9B71a0i#2c;5yH;S*OauzK53~Qho_-IB0H3~tqC~I@p7p|A z6hW-U10GtNZr-yj8)?$8ziem2@ntJ^TCwt9u5*n)c&U(fflV* zL!n-!gGzGzr|0_DH~qVc{I5?|;2750#xCv&rSX-AQC+)SwJ+;h^zUbi0$J?=Svrm@ zW;(!OV5FR=;dC>&qr0bNo71YMMI4I>HK*C2w{OpErJAInXm_(}q;3_N;Z0924pMXp zX#W`B^LJyp72rYV-d)WS5cm!2pH}-1dDpg})~d4J-~W_2+7Z@wF;w}govETSWC^fv z((s}hQ4A1({>wLlD^QH+fQ-pVOZLrKJL!8;`SDIuhf*+rCF^mzbXjvvNk!ZeAnIoF zYvikTnwqrt=knZ21y z#F;-}UtKN?x7@CzlRUuw2*c-};^D(^5bFOb8#g^m7VXGZIaO8hQ`*VEGfeC3?#^Ik zH{>4{HH{)s$U7?5LFRN~?k0T?vQ+_Q*xl=cozn1!U%>_M_d&o2wBOe_bdeR@*_%0E&42J=ho+?^I;fCdWkjRRuL+? zd9j%JTS>!wT+=Z=P67n(8fv5^dvkROF_C4mw@n`IOi2i7E@f`IvR=L-81u=nX{;sZ z)<_s)&kL2S2uE5C5I5B(zc{Biu{`G&o|`RwJyK}76}qiV;vySjdLM=REsacd%sRV% zieXF^57=>|vZo~-AQ(jpyCiv8;Y$cv`#6|NM*qO)YRi{8mj20cqU>n&ijV(J zp%Ori-gtqy{2#WlE%=uM&vrGtR;=tWJZriZiL&&Z?sE@Eag7|W>fekd3h3SkZHZJs|6#5-Gzl9Z?LQv%;kAXH40<4ieAO13H}>#Yd^y( z6!OZXVSv&qkLjU%gb=yW!EqX##Vf0$%>HwT(uDIGpG^FutX(Ha8<>qcolrnN^1ngP z->>{vMTuE2h7ojub4KWvgrnpY1!I;@4rHmyl163aT}6tlYt!Y|5J(YyS&z&&W0I`T zXZBPI=Oseo1}=HDm9(lVRF*1dNA0yYYC~#(3vq^J$$`7BXPGR^o)(sBqge+OGK(p_ zR8g=xf>`S%lzLmWv#h`zQOp7v?SEhQDgm6~ zE5ZSHb-zE2jlceo`YHsM!3G!qyHCyeq6$jYxf+?#Zd|u|(aGXH4wVxgb1nW1GRZU+(_Xd?lwIPD$x(T^_45AiQ zhOST|cGO-)8hz{jfE#E;uggy-b-8_zqjHq`U5BvB`czwxtLe}RzX)db=*b8>(PW7L zaNYm#(!9K8_!#4C3TTgyWaV#B0hSm~>fmUoO2Zm9wbkY&#c;OJ*w84OJkn9id%ur( zcFeKI^oTY!b(Tq`O9A%s(jNnca;wyTYlj@%;LzH98)};`@9ZKyE>bdagj0`4X<=(C zmbm&aaS3O}b+`T1lih?Di})yBqArG}sy;oHul zkVN zvhUG+>E0u_gF&)~L))qf4?S5H8~tED5m@*pC-*PyUw}%#nM1*rt4;3Ny+2#xxOCR$ zOC&WWmbahaIdiApXR@SjEZ`>1Q(bkN4jch-r2!slDoRdxFYRgP+ew^J)0E4P;nh+q zF0Hs1j58~fD<0+KFH5!G>o1MAL)3wyT9Z9(H!Pm#yaNl@$qa7vrN=mbME+uyu896$ zJK5{sUO&TwxG>N**qHur_Oc5G@dO*HVO=Hblv$g$Snq+PJ@8H=lt&rE1swBis>8Uh zS;hELd2OtNlcdy?%qgf+*d>7VTEXA#Dpn2zZo$H zKKuo?t;m-0K|*XmLr=QWnjI&hpwDtZ{?L9vfw~cvjg#P(w%HhdW{9}pAXljQLXtgZ z(X_|t3C{t-$484QOONFc&*5dgy@%YyR(n*S&HR-nNB+RYW;<6i1ZxiK{N?yU?dR)$ z`qSKAG=%tT^#S>E>|poh)sa^~x&&ukEk?C5=1PXN)=Lo+oNpvV6|g6me^d zr^>VCS4R{3C0>L6y$H+UaWy)ko~WdSn2A^xRA&U8HmSajniDx0DqkNO%l6F(LM7sm z)?A6foq7jXal0AH25yLFUhM(%y)^ANO(6l00kRx46tl_Y*wM3;zFdNLzW4E(0nGj@ zgWW8Z>De%5_hv0^Zbus{hr*qs#V;@udU^^C#5RfFQ_6&iJ$~;TMr?E+HNa! z-GlS9{sQTW3c?b%9=o-Hl^kUaWbe1~bYDZwrtVtd`L)wC#tbPvmJod&TNhDwu)Vb; zNlq^{qf`F7;nw9oFDjDjm=3UOZW2hPGGtwxC|(}4N-_*$BMM>bdntDG6;nY~Mx&A? zJ-M<^-@rzgSFw6ov}h#|z4EBs8IWgo6e0&AS>8z@BAvUhN-2?+!%~?_FFMhG2Ug^C z<-^G%)Xu!uS8sC66&lSj_DqLEi5i&Vx1+BF9B?xjsuq0?H4eRIt!KSroN03 zLtyvF^KlhZ`hJHN`hLK+t{VRGCpZ@zgJ623Als}U8_Cy6K=#d@-teInSCy5WO~9aI zCE?EmdMn@jM;l?2pGR#S=dF%R6&!e&<`ex7>_9>U-!^{roN*H6N9L#D(qs2K1A@e1 z>lS{Pp2ZT3oHLvAiyzU6plLafBLmqUDD(%nJc-9M6j3HjFD$aoBC(+Z*_@Cy8A^B! zle{PHZ*h&W8p;}W()a1BtHfoI<5UX;Pl z_a3mhU=TV7c+p}M%2XL<^>-mb*Zxxv=G!HBAy+UVDsN{KnoLz?2~VnN-&qF zUj3*lQMrXj&BiZF&M9Qsx0a~Rc4^?mD7$fjUsu^8uKHm2Ta$>34h(C_pgnPgHv838 zr`pCj8=6_*SGzGP5l>C!-@T2BRcsdxE`$ZeH zY4&^4)U65=BkxZ9`ej(5&pnE>qLESm#`QcVtAi>H>IW{?v@3@514~qgwx6kb@o?wx@f{DQ;L})9J3~I+EdMyPL(+~< zxV4mBM@z}RZ|H*Z5>(bJnAz#m{-4q%?H?R|+x;LMu6E-z24+nVtVLJ%_gU{d#2N^O zW^r>D1&@BlCk2Q1xJ+8J507+a>$>c2xC~iToRFA`;MrYC*20p5DSemMRy~=m&@@ON zT})E2%q3exz+aQFCkSD7bFh&F6a~O%Iz8G@WhN?8mh)BixBWl@3djG!O>BSRa}^}; ze8902r>&3y@EYc*_-Z9IsdWE!7W|8`WYtySr8vJw&$kq_)O2NHhEd_4Sm(*#)B|2{ zCppbU4MEE8g`F5CQgfkBTBamI>^0wfhvW)%y zTJI}8zhIW7uMI`d90?bo*Yh(6O?ki${!_5oEQXT;8FYRsIRJM66{^J0^XA6IN|m4UqQpMmNMWZ_ zDgB%fPZvEWiS@PJyXJQtVx=vT6!a#(-RmCER?7JBmL3RWIcwq4J7wyb)?qTb`^smE zQ~9XVp4g1&D)ST>`_;V1_z%P9b6=S8&egRrrsM=LbE`zwcwKgKU3XC3tperbVVWSy zMl_Ud=8X-aWhGnC0zrhddXDh z?G|G`>`J#iu!nqg>RIQ@P?WYX@`dVKw45b3Pyt7Ru?GdIjqoXzn1NfNWkPgC)Afu0 zbp*V?k7~)%Q056ixv}Y{)OdlKoFez#|HY(I~ldZx5nnbY~RGvn(p20anba^V7rTbFOvQ zTlnVmM3W7G3Azv?z?*E;@il+eEM%zz^^SA*eyr<0T{Rt32n_${&DIJ;D+b+{Kp4}@ z&jdsf5XV?2_Re`M;>>KwcX+hhfM5QDQ?4htZ=($vD`(FQ8GFN~>wxUhV9}}wb%V^1 za-1yx(sZ$db_xzWI{}&z9+)NS9o-2d3wudfUasm0y(oP!7(H(ZI<1V8yirz3*w>67 z)x^-U3Pe)DAp_1v_=L!*o)^4>+$yElG#xT6-lCH_gSdc~h5-+5W=7b3?R)DL;5}NyU zO01VqO(2IE#o|#dq}3d;g;m(UT0Tyu1^kx-*4)-M6SRyvfsP^ttiT^ch(?_nOVr2` zgcb4Hdr>7TvTJ*~E5~zbQY%6Ux-7eZa8v!a;%i2UE!adxNM;7fswnbT$bB91Pw`;t z3Yw)DF&>pnXny5#b7geupUA?zj?K_H*k6}e{IT8T4y_k}DxXqENks4&+2fqT5B>GMoYyw5OH(Kp z9!=Y-EtoV}6*9SwZ%82Q*`pD5$*V@u&8VRL)=mCUKcgS^dirhf@j2BTYTVcdop{5s zQ^}=W*JTxmK`(~7j0wHk_UI;m{D4v6o-BN7PG7VF!+ck2h zfdrHus6@c8$G}kU=0w=-t}s@mzslqJVLbBD+_2+oY8$PUmUPkj`+qQ^{|bZ?yP*!3 z(%tUQP$nuIU?iP18;XP!(|%5(Gk3nWN7t}{JrYO{Pe7EcVrA!&Vht0 zD!09i392%-21V+skiK!F*6JW-j1;DvI~}ezz^V+GtRi!zP#Si!E##lwv4L_-4UwZ_~1lRTrq&*^ID zX!_qY0(O`rOPzzl853ZRImXbN&NpvSsYy_3wx5;e@1FY&L{#0D5nI?S9LqgDJA0Z8 zt$NG+r)-O(kXprkE)tKQqsTyS$@+$Qtd;vLW%QQ{uW~i10OeJLDx!EXN`KN zS8)NV5;Cuo{k(u4Tff7^S8=0YYuK@^@y(Dt{`hG+9%xZ8xnfhTwMMi2WK9>~PX4kg zGC};5Rh-pa^Yv-&*OW2urc*cAwNmqNlxMdcR8TvCSkERq+vbg+%+kj*0lE)WrXIlA zuxHp)C=;z2;jYSyk3(*=l`^h=;8<%7Xl6hA2D&wAhmS95Bh9N?rcI{f%M`Yi70t8z zxB2ctj+5eiU9)OwKBrIq&QZLB-TajYMkX+M<52ZZ)hj%JS96u2a#qfa4VPY=Mr9wi z=RdKOtco5Zvc^qQdBQ&VBC2S}!5@e_+UorcnoQqpYZGK1J)4K>;IRH&fUcztH?!SL z3AXaAziD5?UTF;{e3(kplP`;_xjO+f3)%3HYKq0jG13X@n?ve(BVUNdRd^`5>txSH z@K}#?%ly4j#zFVc|fO z0*cDi1HV+*<9{#+;$*mP4iE^&Q~6>g?pEygcBYK}ynjp?1tOrRtE*Io&G4+VEqI=5 z^*=|!jT+xmWl&AYH^18>c|5krb4`)$-`E;?mt3VUm!lP+=9N^p;DEhx3 z%^8Q3w^IF)aIB+GmZu|KGA?o+3B&tP3N)DCtQ&99)54N=B4{+U;PEb{gwdkgS z&sD-%HJF@Hqpi<6P{0gy#%7(X?i&{GAi@{&SeP7fXqH^ojlgb&zrs1p;k|4U(9Jo% z*+u6`(~%SoedJ`Ku1f<=(0YZ0xu>@Ppbs0Wep1DNg?ixmE%=CKBR+c+4<{yr1@9+ zQeSr*)uhQc-%JFjxJd_EdP

    Cb|niLY}4aGvaV1U3JTT>J)&+l342*Xt=)33l)H z_LafOxX3e%5RPT>(C+F{6A!-d+cqJ<*Ank$DKzzEDkC^`N67kgVrS);NO75g#9pw` z3SWZDaeD|mB$Xs|R&;_W-V|RX34BXCyu0#JK1cU;{gkN?F~@kV!@-VToT!L=puyg3 zq;Qb>eC7TJ&cVq*C+kg~|1)M8{DXXjzy0Uq<40uix!wMPcHNA?Q17!@+}zx0RFx45 zRh>-lUk=9Q@FHs^^-H7@-cF0>sOdxP}mC=zgK4=cq6*xbGdpiBH8-I)UqP z)K=7`%t5~g)qb4kx_UPH*BAVZVqeyxVjr&Rg+bL=@lf*t9xlQC%ENGgda-ahi9Q=Y ze{*278V>rBSdtnvjQ*3FqFj;&353pfT8gJMTVR_&RCDx0BP4A_nYu&djT)7z3*TJ} zCsOEn=b->QWXY3gV^6X74ig|VHy~QgQ$~q!rO9P1E>dnU$sMWGXUp~6=^*X+>HPL| zgZwjS&JKqME*&+Ef~>;C$hYXT+w4oHexIwZo|#?as%(=lX)^nd-_Mv-hU7o&*hL#( zoMk{qXgw>K;RIiwK-z$O+^>)HhT4voDj|(G$BS+uo8)0X@ZHN4Lyfre0szyBc~g@2 z{LnYp&?8<*XNhn04PUune;Tk58a$tVD@`Bh7J6V}=SEtkwBx$b3HtH*M4Wf7b@>9Z zj1PL{gLk>5{bB@RS%6mjN~5?gx{m+Q>x@K}FAri4B|8a;~_@bXD7|E_AE^T>U|fY;XHxrPSvY7vqoUm)w9Kf72N=hd<&{b`Bgc?AA8p=3?t za3h89vXF9zRp2PX2ySKdAy#hd$fsJyR|lY+|0sjrpcvU-`rRzD{OebHe6wC?BWquV z&X3CFV9WBdZ)AFB%C9W0r^waHMPH+F-Dtfh?#p+0bkyPUmC`4)`4cD%${-3QRsvk+ zEn$#8?%`fgp0Xh!%I07a2;k7rH$qt0vX9utL$GS}ZIk=O0$$Mo=$zdp;g)=%j?MyR zy|GXnYs!k_Huw)RA9EyI^SoV5GI8_X*7-d^rj3s0w6PyNDsV;B%~%1evH2#3_|gcJ zL~Rsg&f)z#!(>3UW0>Qb?+Ut8Y3A2G!yBZ@ia0*xfuOzr$oPPt<$`PjDj5p5u)KUjm% zifQ#V`GCCB7G;>411qd)J$MHzPmKHOG8B}gh1UY^It{RP4Vrz2QIdTlJLae#(Ae0< ztt5I6OC|;lA23>%Sh-T%rlEGZXE6tKijkJ+A`fs8s+6p$0yIiTkyJ}DVgY7(Md1WtkAXXCGga7^w981oajv69?LJ`06$w^V~VvxgmJ-^Z9-zFQ4wuasJ4 zv`;*Htoe3RxNBYvF>twX_&u#I(YRpx4kGDyh0>nej*08_x*-9skut=Mc#AMDs*mgB z_D!_)jO?>THlA?mZ46WlZ;T9^m|B8etEuTO;Jd0dn&BW`;QVFZY%virDv7`GF92Nm z1<8Q_EZ;cQf@CCD5LKtu?!W!XAa|17GujPB0rM?W+Un=eyeG`CA-s$ZNmG+BAs;es z?yDN~LMUBOlom=%^506Q{++fDV?yANG{i58Un9bKI;)c4SlKv=P6!zUe5(yCZ0%K< z!4XH#yX<&(<%s-k|11N;ffsSYg@YE@nb|oL2Q!;nAcz5r_S^V83`dWdE(u4{xG1ne z6cN{W!5l1=`aRlH)kgg908(62Ad# zh&$d>Dt+`jjV;u1GO<-ia)vk&K{T|2(FgDCc&_&l{SOb03kpx~;#FyT z*mQwgM?-b6p&Q52jOwVW|n^xZruJ>5$bg^qJC8Pa?_CV(P*yjTE$T zea+Q#MqgCHLv*=1zdd|Ad>ITq$s#s?(wJ+96+e3_F}*9BS0FD|$*J^Trk2npCS)RN z0ezYYIC0T!io~2l2s5b6j?LSH{GgLidevEk;& z#jp29)118NyxC?v4-t_k2XHAjxL+z1%+5?MtkNh>#WdNEnkW|H*fPyzefgOeDK7I4 z@}U{_ZHpE_tM2y9?=1=gQ5zOXW>1>+N)pP@3eDGILabOXr=;S!Rc7G=C&is*4qV-( zNN_7Z$Z+oDD>G%iE%FKJtUB>dvn7*wC)dacYBf)ua|&#(gD_6xyk(W{x{>mT)kqO| zFR~F-Nk8u0tM+X6Fe=`5V~=R6%$9=#Z~iOaEMg10#r7NZ@h1W<8y9*m zh6d#6-%nUU%l4)=>CL<=Ezp-cywl#qnGu|}G%yan-am9&YMdwWt>~v|@4W3-3yr=H zr9Bn7H;bt`^AfKQq1FbBuG?mKT8bf(5<`td4w$>8v8}99P&X?1U!o<(#&n&?fHFAx zAfK2-TW5{+_}YL-<)39@v@z<3>1%QCFgaUjvV!T`u;NQ!OXhH5D>-^_o zL^}O(s7?IfglVdla80C|?o^$J@rZCJ+Ax~y^N^<|6aM|%aA+~UHcTFA3lN{aXRePo zsm*_6=z=B9Lnq$|ucx!2EsGow|BlrCE+IlRo26oo^MV&~(d19AKiJ`LPY8i%6K2UtFk%G?imMmCu#8-)5m45{c%N#R>DM=1)zO$Sc@oBn(X<$G(G=7?@ug4C$3-U zE=z~Dvl&lW3CyjztCwmF90yvQ2#H^=X8ZEi9Qdq`y7;5rZe`q9SwcApm9OcF<<_G6 zdU@N^kL@roO)NQ*(a20B3?X`OcwWWWVw7W}9 zMK@?<_rfv*l2wKi*A z?q(l`{V8SIvbyGTM@U#9$q*&tSs>2N(?xSCdUF^ktZ87iwMw5^sQQJ#G}6K(v^Rq}W6|i| zYsp^z8Zun5H&It9hPVV~q2bS!(s*fRLqUr? z%aC?U);Y9z5B4$USlDUs?P6DRDAdI0nl2)q8aOrp{F?UC%tPaj%9)by0a^ws@k(<7^`HfzrQ?lf6tE$}hu zY?LjUWLBLy7Sp|n%;vR2NB!;>c|oIPll?FhlI00%hKkkznB0@jteB+3;N+Zz`V3YQ zIFk}W7(`E|JC~mPxS6iyi;>g!(miF4f8`202oby6b){}ee=YO*^WcE;P)hyuQ&%(1uf>#u>V=FHd2#j0QyLC3Pc z2H#KVy3&WGRI=r$L4}u2wRX3Cg0dpm;7r?x7gf`(Fk04;SJPWHu` z!9;zoH4jz$+27Y)1Rj*3D4x<`7m&9}CD%>^OF`q`sorzbS^w*D}Y3$}?Q3yk=&Rc*e(^8;5iW$vnEd`~+9m%*FRT<%al zd8D%8Pc4!^TQ}FSq0T0T<adpF!g?rV@0)!S=`2&KP3<#R9j5YznuN3#D{%_@Cn zN#m3zWsa7DdB}jV!wcf>*zwV)&L5my%xb-Pps_IK_~J?WuL+pKY;QRi(>7SMMw@g! z@870hnVF*JclQMZ7nsDtOfnCyr^fpv^f}-#>{p!s6pQFybx(q5IaVq{hLtHELpWpT#8><>;qJR_Ea6PS{;&LrkPYfAmQLcvl# zB3xoKZDKXt=C~H52ON>XqgTS=g--g(V{NU^ikaycx1|+*E_<>-`P3M{%PA2{1DyXm z|6zs-TV+{T7piOAfJv<}-izalF2c&ixoE6P+l<39skZOd&R^yu#7q(v-?)s|Pk=j{ zMKo*l#R2W}kw{<4{lpX(N;cxcTjMig^TK-U`}Me<63`TnKwly>%Czn{-yyRm0Jp*h07eYw zS7=vrKfuQ}e=&UwO+Z@)D{%5WQD7jVY-_~b!qns6Y^xpR?I|(;D-LvG+5rbAlZ-;* z29({CQZ@B_nM`L00Fy5wQzSdVg53bHy3$=$!VFhBAC5NGtr7M*iO~t#L$H&u&pwG4 z#<)7;K7^=BqH!;y@oV9zD`Bm>sa_Oo5~;LF*cVJIGNeD3@xgw?y^cCBMz>vTU4Bx# zyOFPd+1l_Y!(H}WS)D?9DMF>zo-G|TACB0^N6Ts7a23)KNh`bLyuY-^#-W_Qk>&Hr z0pO7F0ltz5@{QLa-8JHz!Dr1P-~hAaXs2ZNnAWAK#mH7q(B{%`F8i>J zO>Cj!yKRbV9+YBdhe*yo9XeSbh}qV3)CpGX#=Cdp+vhgSMjgC-mlzwfkEgimN|4ew z4-5=`mK2)hfqe#O>-w?e(U1koHFkBPPo|`Nw`h1@pnUUUQU@D($LRIjyJ_fJZ@Knq zSp#BVBKS?IpL|~ao|fdTU?58q02jpV5f7LI1CLPn!YmiVCRU(8UPKCuv4c#k;AcZV z-rUnEt>gZIZn;l|Za%gmm>dpZ8UbvK)>lSxk=JUjYjwQ+V%e|Gab0@iRo$?R28yW0 zs=p1l@3$uMvpi}qnaoa|pMAZjJd0d;E8aPN)#I%W{3_#TePOHIFDG#;j_)1iuyFSA z+e9|KseyQ;<~`apwv3zf8R?}J(Bs+ssf?2LK%5js+*`fxF)0Xk{UsWRKYHjLL$B|s zwUB2y&V;mLq@|JQkHgVyp!Yq-rvmq^%!aZwTs5(Nh3O?1vPW4q{j8m-Tyen&$#Bka z;$DQGg*AGQ#~C|38)018t;<(0%iJR&3r}Q>3UpYhH_9;I8@}a(Zq2<`e;Gre9%P1& zIC_E&fb>__4D4=!Hwty3?zYN@TM%uXXhs|F#|D^SfWs>_mHCkppM-fLnG9R}#Akoz z5x-B3-_3R(t|Y&=N0OpHy@PT3D#mJa1^;F1o4>*ibNtnc?leL3{T7#qPcbTZwd5c8 z(KnU#2`=H-)So*OTqnnu^@bliNE`zK0iR$|gm+$}`%+}MM)pSeHc zd4C33LVu-&o-X?Yen^W#O^+IJA4w4}M;R^aVyojToj@6IwazMd#*3Q4&t7mauiaA@ zSP=^*&P?~|#X1jJTI~;PP5s`lA6jQPknFv<1)dX8MiHuzBJHma^2g6Zcgx|`zmKzl ztS>f30#IGmA)g|ZsyFJ67PBpVv}1T~SE&A~pg-QEBo|{Rl~Ceb_#J|x5GOJatwtzS zT3Y-@OC2$@00qI_I#PK`7SnYD`0Z^Bj@?)fFG!84aBy&iG@30>wK5ei!AGP3&x$p| z>V=zwH+!Dy{aPvBi-sN>)dwuK9DSdotiR@Ho}UNK9|_p9^oU$aR@Rx!Y-fna{EW)t z=;j~rqa(DMpI(2u;#$<59z^Ad#cr*Fw?R2LWg|&DSu6How1a+_^~YPI8ngSWcLona zOG``8_rt4F?@>-{Cb@u|{9k8{wrByq!H)q2(;^4zp6J{M!?q9sLdxNvPg}M*S{-HM z@S)e&nb(IJPR&UgIxXfQ^}qhDXZk-XY#LyLYAlUwQ^`pa8xzYWvJsokmB94pylm`K zzuqf^-vlTjSa@OYNV2csl+6u*nKiNj`=j2ppK#B$PWCF`FT%rgb zmr;C>R7t~r$1h3Ib7yJIyvdU-$iAK9+`<88iO$?n1Hm?72+d$4R)Wh=m05yy_*_=| z0PId8+A6a5=wiev$Z5<;lrrorS!>b?$%s*k!Q-;Wy>_uT`tQl5CL94PsX z5y_ox6q060rEpy^JTts)H%e+CRs2rRs)$MONVB@KHc|xFS#6Hk=6x0H76j6ZmRWf5 zOJ$btDD9N0xvON}v%Rdf(WA;1SEW?ushguxz8bv6?#YP8AHxnm0q}AY3UU4YZ7HWa^p?gQeh&ccEkt^itJ2Aqn!fG?`+=-*VVXlpK?m+p`FCJPvnn-T+V zgRshp4S%NjFl7B$jm-U|D;8p_1KgD8VQbo68Z;}nGw}##WLU&q1~j;@TQafToe}ZC z6IIzYAyaAO?LDmn?TAv2A7uE{T$f~g&FyyF`0P|lteRMA70q0U5>78ty%E>Ll5+{G zqpHoDGLDTDYQZ(aAj>Qu+x0g_o&>rn%|jB(6TT(@J*JnzYj#Yu9AST#Bjc|qO{bly zvdrb4Ik-D=njX+Y$82Gj;6oy1kc-EQgBsL$bq|3Y5+sYJRJoO9>CKFauB%yl+14+UB2 zFj$Vz=y-}e8I&Nq3O91mNZhM0-B^0p8@M^J$$$Y{Jai@YQNmE+3aMIgT0SFE@iU^q z@(Xy&HH2KKOpcUS@OZj)whR@)U5J%|aty~5^z3eH&d$9EfY3v?(kyc5G`+~cR<<5U z0$D9RD09gH9nnyQ=8cu)<5CNM+jYLFbzfKJGZl8nrW+*RGkT}Zah~E|kVHn;iYx8mt}dY6^|DF*6rnbF{u~@y!8E$>RtWe*^ z;A@3UBzT{bDN|d~0gY4}6?_`=AXZeJN{b*3KpnAs0gS*<3+GB7BR&U(Th4q=d?f@|644=fiR!8@9#?V( zB4~=dAb7U?-s~^YC0!+7_^NJrkI+Rem`zrUaD{L=Ga_uhasp(T$eq4vd3o3riMHqS zKOJ&Enj5HMXz6|PIe&01>x|L=r2qs3+^LE7!Px~6z?qM-;uYeot3(8bF2?{>-@6Ba-w&z?d#h!jXE{E?eftB+;lZ=*-0#itJy1q3sMn#sT1J~7Y8%dS)cR72SP$rA*&l^7s zgmRboBIvAibUTvtlLt#z6Weh|LHf&k4U4Th@G@=huc3pi;~h zaRFL;y$av!G+xI;A5Q(KDMc1{Ja0ln$6C{Mnn#s95R(~Y9M_cKywLW!1tv9W3ME|2 z_n~jidE-NuKdf8pfmf*y!|d0o>ngkhH zH+|nt)WH!!4d3I|rdgxpMxImJ^p*#j8@|d;xG5p*z=i3eC1I7(Z8<$;s>=Z82P>+d z;UrOXrN7e-c{bdrQk}+lQiVygo>}fYjEKr}&I}8zMyh+2d8^m}+;<4UVg>4;gin(b zrI|0ih_E365DXIHbn|R@dij8SEtIC@@q&x*F3_q(T}add8fB*1 z$~}Dcx9N!-&PU<1W_r?aZ;Nj3LNhsz`Xw4PKA7lFhouk~<-4j>-j`ifl3aHx^SIK0 zJeAOk+2Ohyf4G-{MJ+Qo?}Ei2P7O`Vn;&=UfPt1%U4WA|SWAuQf_R!C0OBG^g5y?# zYjkPKnJ8{zSZNHnh6XZuXu5W8OL;?*?R*>5T{ge7;5}Fv`IIu%KVue}aeziFX-qR; zV&TUVe%Cl2ujOpp5#2NV38UK|(>$kw1gv$u-6> zRe840b7j%67SScu^w;Z=(i8;QVC%PCpTi-xrrb`4cI-}dWKL}ucZ$@k9>3(5PxBua z_~Z@Qy~^3|yGaAp3WOf5xqc_|!vu5$cYoU8tk(O9{qZAkk^2P~p^CE5(lU6u12^))6>HAzeL|o9i;Vn14{X@b6 zG~++IdY14B??V}_vl3Uqbcw)vBLWpBy_ov?G@-^!jg*OgxT{lP=Q?EbRmVN&;h8+?CFCxdVA-%{aMZ0?^w9l z@t`lx8^v@1_Dymqlb04THfmQgF zQ)i2ahhN2`e`9g-tCbiyD9KhRhOPgwN`S^1DYlu{8k<+`Sj+Cg())*(e8IlEWqb^L zUmW9P$yQ7f;$x@Z?0(kD@?=ET_l(5tX*Cwu%*KCZ&Il&|bNR9)F8}CoNe2sgwA3bg z-sL5P{o{G7T=v@Yt5=AXkN!?+kW0j_z%I$^*%#UX#qO_eJGmI=OXt+5GbIu9o0Li` zXx)tIWXacUU(~lAEE~DDZ<7WZnw!25f1%g6?nC~mR=f+hI`DBSM!j(!xWZouW2G9d z&6@?zAHj`Q^f8B%@7MGhZ6#}>@b(Le+OP<})5aqNmt(b-1*dHWR658u{xC}Vz9ddx zqLmy=ik1rBch34bMp93noJu&Lt#n&Ztg^ol`aU2<^mLXX%55rilcUOo(Kyv*Su zP|PnbilgGAm)YOZUsl@5ZnwzfuZd$T7>slZmw~iQqs3-ehp6WkxA1MjnR zgC}4aRFFir$vLFUJBqFZ#RqVj^72-+LE%~`2~FDm`)McRVyo=e=Wm)APR-Gm9TXo@I zK9fQ3I{KnBL8LBg7zRSabsxO(qQR~sj%y^tC)a$gDV3>#8|@YYccAC^1J}Qb3N#(o zmh6q(Vpa$#kcs?pmNOj`Uz6Ec+xp=-Pkyl8LDEvJ@H*`(jJ^CzYkPCYW3YE>_@YYn zaG@m84K09HD!(Y&$DT4TQeRRa*D*7wf6*pkdbY+chnaNp4`3tp^%$^#P8jkfYmilk z6(N$>0dj$4{nXZxn&^flo(cZ)a(-xjb0OpU*`pj~HjZS`#7oHYOTwP?5l67-q!Esi zg#;jU!1d1iD3JGhGQknP-K0%4@{{4z8GArkZzayIe5c7aA8p4w7O9crdM(7M#F;cI z>sZDQ8=*0d3D@*`sa~qoq|>bnp99<*Esti2=rXc~K(|!efc1?UAw7tmRCQ1{X4}{4 zgXs+DiC1ItTzB6C?dFXWGu6PX=VcfNt+|I^Tf_dP5>zb*&v5fk+Lw+A8fmqExoX6V z8d}qQ#I1(7ypfLhHxq?`w`ic?0w@A%BRCAU0P2GNg&;Ab-E)J+!gpfFNF4fM)X$cO z;YTb3tKJ(wp>v$N!uUw6 zQVZJ>E5O+?d_^gVVuk9`sfktKFf#+O(M&ZE`LBJ2^)m^`yMra_$xeM1q3Rz^aLpRQ zfyrPSpTsw;L2A%WkCH4><=s=ZEgvnwHJ+5+cYbddMK9+$*Bexw;W|`)bAg?nijLzi z6U{96-RX=>No7(lI~c`|AuXU<8*L^$zo`j-F#Imy=tKJT_p;SD&$={;ITzj}6{;V5 zU3dU=*N%O_$8(XXvH&h^y?m38*jbk4LdXH0){_lF4Xv!hX#>*er)MshZ*~``Q*{eM z0ttv0hoiojT(E?p+vwuvSYX|Opc-NzV1(60?59=sk0jEWKRe>+-2&hmt^J&FVDkyK zhGZt9Z=Me8C`E&j0UoageX3zqhVi=6i$sm1QbaguRL>>w)OU9}NBe7k4WdoX3L_$| z%e`(OuYdZuOsFdhB*UeSg1 z6y(^%M9xJ%cJMbVAdEMU@rd#J6?Exx7PPfeP%)Qw)w?+v#WZhnkjG zzN|y~xa+!(oU5$aI36ll6N5e;)EBEscn%|arD_)gKbb`a8c9jg$!C`5b777(%O>~h zxq>b~Hgb8asI=WEn)lPGyZ4Jt-b#9CVSI9h;Tbs5YZJHo*GYz#zt&y7JVWun=nAu>} z0gnZ`i%(h*S*#I|L?cPdcb*ELZ7`t5mnemZ$9OwJx@`*N<8H7oV$4*nvIf>tZ-PLJ zL=pz*NJr+`V|P^eH#>yO)~_9id8w?V(Vh<1Pbh2D;0G>;rqVKzipfez`gHfSghUI^OS zDECw8h%G#9R*#KSy_pMtemzw4^F~9pqvmP6i5{98c*qW^-D$;#PHjCGhYxyLH`wtR zTst(oaW8c~RUNfb1*Q%fC*E=0Z)XH~PVgy@2raa@kzncK{N|x+-fXbja{ktTu)j?; z4{uDJO$JwL?JL~2ljcDe6uO=I?WF6Ej4hLy1ZS{y5%1-^JW)y+(dbQTy~{o~EEGjh zb&?@de=Mqe3(qG8zactwt*KL<$>AoryOoS(lJnW(SP9!xs2-y6nx5NgeKv?*KI{;l z!?Hnv^0pR8`f0V+n6zN4v*E=IL6&c5ohiQTL~rGpcIB7YN0Qdr3&n>IJ|tbiPd-JZ zN5#rzDhv^&hc<_}{RQ`O*FLW0Z2fhc_QVs11@Kmzx_@TgP9 zWOIIaQu&`AIQ4|T1|5&#fP1}RKtlrO4~GSx6Se62%sV8*0GeF?@JYHmf&OC>1DYmz z+iIaJOHz}UW(NKZkvJo+@qykth^~^h=Xq=lL3w*xqvc`s**-3Tk&*$U>EL1y^q+E4 zbQ-8QeMQGC+;bL*H=)}FEY=YM>gd3D?+SoKNcZv3VBH0W$!OfT1uVHaN#KeRzq-3< z5dOX!n1Bt(5s#8p1Qc}mEQvyxMAgt2oN3*q^r>3v+g+)#qH*sM9F zql9R7;LpO<9;&$dMJ6-p^fQ~vAj)>I+>i>q@BnaQ)+5P8*2&te1glso%_kLv1ss(4 zHgEmFRL7^HMafvCpY!7Mi8ei{3{}-Y*DhEvu~enNXZJZbP_FiL980Bif0+zc;bpz` z9L6F${lvXA7VI@~{CYf2I8n?Z&k#^5n%B70w1uwam{;))U6B_|yPjpijxRB^Ei)T% zezzl@^|jwo@K)8=YUeK^TI3Zry;BL2zHOCC&Y`C!`op!QbkNx1Z!B^*A|=w{@!eyR zcb_Cfaaj7X|E*#JHY3tluSzy!0=idIFWbo)wjx$1xk!T9_WzZ?stN2|$kUmKR z#|x+(Vj?fzBx+1$^c_^J3cakk^s8Tk7h0JtNFnH$wp57t$H$s*>j7*8PtMT0La;f@8)KZsICHd(jiRevrVXAnaIrK=YO8y!O*Qqy zviDgo86kHJMAU5O>2<1dozzgPF#jV9Ky@=z*mU5TVzK%|XDy%gE<&mHy-`QvsT|6$ zr6+=KKVobHB8^_Xgb!SQ7XRmA3^pR8`E-mNO{WHWiS-`%D;-^yqI)uC}kb?@cR7!(?5?CAs^&_(+Ie8=~cG$ z`l_)ertWM^k@(xvWhooJ=3JX2-k%=Iith%J?c+#x{R(B+=Ngv0z;j+g(lSNr3kN&~ z+q&$B2#}eoqh(u@Ea&EIrF*mhOEHnlxMR4lRUW279u`2_Y!KyZr+>7s@D z4dxd>HxH#1k>{$T51$mN4sm^6iHUW2tEFlVJxH}gMl7%DsPgue-KzaYF?bFswM70C z3ldc;%h&g+`T?>cW%02oF5sh1wktyHci_Hz`Qklzqmv^ZCC6g$mQ|BJzU!!WD|DdM zArgGI%g#dIgt?;m=-=c8yA7eLupez^&2*Ib*v^sFe0otBHYwF+0wN(1*VJW_s1m~) zm#_>B%&&7+7}P{wZH`NKfl^M%WSzg4RHq)sVpe4t*-l559-WEK`Mtmv0|Y;HJK36x zv1K&AB-etRV%0?ZQjll^e|_>goOc2;wWCZgth0Aowv zbm~4+hm#`z$Y^2YR@#pNJ=UjYEBMpK`652 z(k1UH$61lyc1YPr?B}JQB++k{s1HQThgUnCjV4h~@MR^6bCFc1Mk(z92h`ULKrY45=DM;-??4gLS z%N7PZA6xYao|^#?CU&QF!5qlLtKL$!6l0V%>AF`y7Sq5|*`?>QFi#&7+T>cuY}4Gq zV@J*~mH_hQT5x?jzkJ!F60gY+P1z*@zFD`QLI}3a#kprsTQ%OleCXofZ-Re*Obp^% zxv`WL&bk{Z52Cr5oT?jN7@Y5ZO*t)ZOd@uGnoVs#_qo;s7h*Fp-CX+%P!c;b4(H=G ziHiTIE0^36{67>&Km`Ciu28&d0=AO>TD?wTW3bH7Y0Ld?Ww=eqDO7(@Q%68wqwY{h zM5MOxQ9_{AKF&zft6_-fLZ8*~FgIN9Y#W^K`TgB=Nkk)J1HR~i zb6;y6o@b5ir89!t@6iBT+=ldS$mtXE*`KvoOrNe+&J2jO)U6@K@v}5kcN5b$HK0I^ zuXa%|J&jh2s3hoCJ6Xo`V0=tuM)Ll~G;Tls2E82+l=*T#yTr<7y^LhZeV znCL>Jbd8iMZG*ky3l^4cmh3pdFt6LQ*Y?YAZumJ>AW;b_$e>}7=YdMkjKl|>vKL5u zeGPIj9>E+=OphceMM9F)AM^s=QLM~S>^eY2pPy-=?JLnc{~GKHecm-vy-ayC%h78W zY?oKO^Nu+SOg}QzS~&qE$_pTg(^t6>xZ0R;WKg2ZK^V+jty$7Y=8I!85Ggf`&TTtb z49diAe~s)ww&K}B(84BdJ|WV2uR8D?!b^EeO&7x`qdnU!{-!YRMjxegUb)RRwC(-K z@D@SO(0Sk*;*8HT1Ai9N2x(L(%(Z23dHGXZ^Rfo5)wJ3Mi$D{3O zUHnBu3aZ{`Z@*xZG!lze^j&^YVk{mKE_lNUZI&H=SB0>l!QewH>li56bhZ%Yb-8n; z3{TjSDszs*Xo+^-+xTGctDjV`C^Sdi;^S2F<``~N_*+7@<9cYsFh}KGm>F@vii)eY zt%Eh{-YkU2@}z6BE42D|e+B)Mjtftd$;?NvJH}w_gkd2 zN^qFsLWkM-xn%+IZlP&*O;}WIJsFX+^`|6jl<9?#wI3#`iwrS~VpR6(&O||w{?%EW zFRHX*Xot!nJy*_qyts#`CSMT>lR>V~4-IX8?C5YzF*32llzx*zZc3NlsMaLmljmA7 zFM^I>{w}kX`1wT>@sKkCp2HCP6yxEcM$`C>MOcBpnI0mHIRni0y!_mq><<$Ktdt(> ziR6=8%GP*qT^vmVR*$M^_yq-7Oangof^5zwn&%PMv_H4rC#V^=f+VV|S(?rb{i z1saVbb()(|OC7b{K{xA^domEFuA(`AZ($U2LZP?ztHajV5^ppvbQaka1z3#n=%{RItrued*ncPKIQhUVE1rvJdGZ5)W;r6DMX~m1l6_Um*3)m`~rOvsfEHI z_KM5hALIxXPd2TPwT;R^k>Ale-Lb9Hht{RQEL z%T7f-^;t-ZbMty3aW%3$j@%R>?lO!Lkg~=qSo6K8Lhq~lL`t|j4Zk8`Q91VK_Tz8_ zTJDJN1rD&+xXql8G~uJLE7e9H`P|*X;bT+eNXArcYU_zOFBw|BtFGjq!^+@!DoiO) zLANhu^Cv<4@3KSAq7B^oe3w{BHI_-A*{tV{0Ec-*T4br5EE_DKIO2iPM`~WG+78mI z#!IXpbMZsCVWr%S^ti}oAE%ZVrdy^p$JQ-Xw$7xDZ;qPYf_S7htwos!${_yv%W<6xlsQNYtcJl2YA^*`NX?@ynZ;d(lf}yUNHK=4 z0XzkPaupG5gRE2)yvHyInm{1FSJcRRE0}NTvqV0~K?q;-2YGVr*8zzvI&)aIYZ zO9DG_G7shqoE)_&l{#;;$(}VZMw>HN-J?d+curd)UUdl;}&g2oK*M7I3jomZoOcPDmNMnQ?HuEN)xjFBsAZ`x#%|eHotq)`!eZSIo47Y0r-6P`D9{xf#n;-A`Crm z=G|xc!X``x)Yx+=Kt5YFtFJ{{ogiN6PjiF3?7tG$q@rV$CUpNPeU&{A&&-3%?CzEL zIt8q_PlGIn(?bSeT7bjEVUTk{*ls!M_uOMpW??+QM z?2y^!=yJoKKRB8Kv1d1y4^u8j?&}V`Z${j1p8wptxW2|zvZs3saZhK@JlaKB*B)~O z4d6=^UBwD6Z?0kyVk%RLf}#`=*@$MpmA|4kkCRf3e{-3v@{ zFK4c}6#i4j(HEaEf^GPE5VZ{u-}sUa&V2F3;4UeyD|p(~B~bkyi1;~F3>eMSvAP9S zw50#d#QuNLH}Ro1w#!sxRVD)G8c{;+c;i&=!W6!RB#Ap%7Hixn@*+X$YpDkMyFL=p z+#*}?j!oyc=v%&Nal1vYxqs5EA=bkKrCDS?!UE#`@)ut5fzbqQkG3U<`TMKCYO zd6%8g!s_K!_fKc~;?U7=7m*^Y)8WNuVf=I@DtR+2cm-8GdqCa+5I>iGpOjF42uRNu zVrBM4EmysYRi*^hy|LIBkbxVihI0r7y1}+&M#-JbIDMx)u7Wy=j8vk@Z55O4Moa)T zxRR>Q(q+Y6Ft=YWYLCP!@T-OwLmXG2p(zKI@OBxVH|)-bs%{PegQ1S!uZ6OL*s!JC zlr#=*BnBYx=kxIS6P?x;ckj~CN@g7f^MbAh9)wE3*3%-0#Srvp$PsAna_LqGs7s$zl(w--xbi#! zpbJ~(Q`Nq2--}-zuk5qM1n7dEt!mUit*1B>3RozGmJ9-SofV5bdDm+6e%@Fv_sIB6 zX@vf0aK@P@`p|j6l0ix_Vp@`AfadCOKhKU(+kM~qq?P-gM0rCGsAU3XhMb@C6jN7E zP}3Q^nuff70pR5M4GAjB&N{1j)JseNIU{xvcmvgQglePr&0Ex=!Jfh}SE}^2ilciE z-L4OQJ8Xo%G{BaUS&RD&AxLjriE4{E31E#bHE^PjW<&26k(H8V~4{k=lca`s+oP5AH1*>*V0@8?)2e=IKfw)SW%Z>^o@ z4)ko}>u2aC^I{?5f`;+r+_urzbLRoe#oK6d}*;Js?P}KhxiTepYPB zUHP-e;s<-G!xz2<>MXsA_Vlo4FD9l?Ldf9rLUWNml#lAVW-YwYtnO4J!WpHRLz}+H zAqin;h57<~BHX3Fl@2NTAtSjB1ohE+!7LfGyMS&^Krm$sDZxar^7^3^`-w@ zRXd)y)sHzX5B5{IZNk8VpjSIGkS(4V((Si32zpMGm*H#9h94Hjh4h>mD?Tr}zKh#I zS%JhIRKxx;y~Gny*aYqT=lqfOz{Q4=D7;3ponn4d$Vup%LHoB!+kR^O&HGh2iMy$cnrK1h>#2Gj#*YB0Zw;DWNi(8jZhEk^@T^ ze}AaGE&ae1!75%1$wtL4?c&e??Rt_}IhH8@t(uHH>MVz+#&cu*snnIoh8bX%cXrC! zHAph}tJ^l^AYI@VWsJkJ+raaRGQhE%VD)9|^Rzr9O`?$xkp?gSd-y>As7)Ey=mQd`~WI8{g~K^%D}Dwo__v81}s(DVsANezihGvpO&< zWdmPFe%{4EDZkF6!Ym}uM?2|31rLo&r`Em=XxbHf0V@#`mA9zaF|-S|sTG^<;LlZ9 zD^86BV|a+)S(rz2S_Ri+O`H=`57Jp+YG#7`gI}UbPKhF?91a(VIA}=lI2<7HR5TRp zTZv1mWlpJ5gbtM)nhn3-oqSgAgZYYYDpD_5-K#<~+KAwstoJa(iyQ4&!xh%x73Eqq zqYYm1v_VYv-g?Utq)t1vU9K0*lo5Vq`78;HHrMu&W*Uu!5w!_|qFb=7?zCtm)?V$o z>uo*Cgbzf6CGa|1y@6_{o_>T(jS`IA&-O#8K4uig&5-OJe6OF{$O)qGBd)l@RSiZK;##p9s#- zbD?!wW!agI(PJs{xD96`o2Yvpny`r47DiZ_kKqz9JUJFbZ?FgTfrA^V!TVw=Ec{U; z!YG5Sm|1#b9=B#{zFJ$4y2^tCZRJWEPr}o9)>q&Bf*yn(`@BhQjj2myO)i9QOz7DI zLGrDj1>W2gn_4z}5+glMLx0^_p4S8nLl_ZB2r>2g%tv_HK(?~Jd#2m9M-&Nv@E^F4 z-~%o*LiYe;NZ=zaLE)817hPD{%22jR3{ZycNd@g!D=g|UTxhhvjEPz)Om5drE)?pl zwz{#Zt_l+~M!&v^X{5&4N&0?nd)(jS{r9l{a~Q-8D6YWmMuhQGS{(oRaY_?9<0TxNX&xvKBl8eN(6xpi((2Z zX4b@B>>|}58wq`T-Ut00h%TN;VFI(xPh@MPl|>@9(%WB>^6egIJwM<$7}Ocv0-gdS)0@VpIwBJn`& z-^C>V(MzY<1a!sdX99TkWfTLq0drqtwO)XUG}s8@>I_tvqoh0wJvDG>ZVuhD`bD~3%y|J57eyPlAPAlB?QiT4nD6UnBgzeV?^z?SuK|w4T-%hgDnKPBP+fgn9 zIGkKc+iPX?|0p*hdh6ok~bkfHMQho$}NgYnuPMn{(3vmW*3gq9T z?on{V9U6%94?dAriLBEmss5N5s}q~X^b=_ug`N^4ndg}+F@$qP1b}iYg4y-#Gp|B# zW`?zDZ^^Ct!n#WdGEQW1Sr577RCQ10Ad2bZq@m|!mX}ob1b_s+S zo47fVteScvhPmzoKlu-mkG4v0lumi{Bw=>TdrG=aZY>Ymi90+QNV3B%t2Pd|VmgTX zH`-1c*^Z0+C3*QILq27o?|^}sl61MYl%f2)=5Xx zrH~cF5BsH@2D}y~XgS(V&xDglBKY5led`9`nZR?RCEa)+8LLR_64QQeX}~03qCkLO zDQP{l61yupY?h}tK9tX?;a*YXar;S17+39Z2!~(?dnOEHxp_k2(we%s?I{AL976>; z%+ONoYEH3BIx=3)NoKoOmTbyEor)2NtQdY@Yk z>rPB0*%>`8)bK1*!CEqDQI4qLJB(8zQc6(!?{OYZ{w^&Jd!AeEOx?a7{^McwppTIp zKuQsGxJ7b?s9YCFH2U#L<)`49~@B7nZX z!6(%SteOoPgr%&Fc)u!1!Rv0755YV~;0WoTorjOa{`3e?di z7%+@U3#w_HYiN*Eg+9iOKP^6eHDqKzelGdA zl`Bzrt|#knJd^s~vbBUar(SKnAShm`xIFcf;fN;`(qicHwU~bRfuR2Te0A+4 z0Rq-8{?|F~0LsXkv~@ng@#8Wf9-X-v&Fp9}>gR=av%se18Z|A(7t7;v{<^)`XO4oP z=tf|H0v{a8N&PkATiF<3xZvR$g!^is`T@=0{kAFP6r?aUr;|p@MI}h`@2hu~wf27s_N3SNGzX~7(xRoYPRl^5L#9|DGRrz<~ClGvh=pelbZI%H@=o1 zC%L?coC^H1ZZHf20&}}ro=&Bd%F=K=C}~7ccP15`Z9H9F9R|EqBIhFuBU;u~dW&PU zg52tw{S#chXrM)hwV)MQ-ncZ;j;eslke3DEQPf7LAc`$xn5HsO@N;$XJq~SPP^TIw zod29NqjXlfrGNB_A1t_-SA$9DSwgw{Xf=Cf{O!Bj>otW|#tGk{jPV3Q2mfW^J!UrP zisrVN+Iuc>nhAOjGH8FgSz)f?J=vFaiJkORlkr)<%#Aeu&hsXj2y>LDu1H6i_h7Zr zZk1{beZ=ED#UN2060O5Ysg!56)I#se3WNxEMC7V44qnGHp$tjJV^N&0nI``l&dImQz?JYli+|~OAN{t zi-Y#ln+;B!hV^@^O8%#!xYlPByk=|@KHL?Zac90`0T5EXsthD@gfs@vN6}dL289@Z zA5B>l4xGeu{%9S^VnT((>FxF#QNKCaJ2+wXYG07W)QA8YEOgP(;7j|bMMuU7+sqAa z!AU$Qd3h#d!)~E!c`dCipn#8$kMTfRRU}9u$Wo}zGfk7juTo0Uc|(Yofrw<_pZ9tf zF`#n8=xXpg#m*IX)YkoeNXgDUGSa}f^sQhLQzg3Pu@&~F&$VP;M_UlWaLG>MNx}e; zg^#z*C}57}*`A3elKz%bNJ#v}O z>jK<8!WYIJAG@)em8HdS2WRistDyncMLirlHlyr6O+)X*ZumpmP)+)N6E$ul-W~sd z*k8ya2?=GTPmuIm!$B!JF@ZkM%Oar^Io%m%3uU_DC2TkiX+Q3fy!J$cRRknGGrSEm z{cv?S2)gHGqh6F$km|WW3@cMI!+fl(sVR*jhp{i}7TowgxWoHBb1oSL2fu0IeHb$2gBljUTRTXA16rDtR!)}|s8_hTB8in}%1B*>>t=rX`jCbPY<-O8$;yhEy#x&& zDPSQxWe}2xfAuAxwB8_sfL!~*2w%GFiFw_EW+KpIo7J;I&=K|QUamLF^*M^|^f)^5 z68E7^6FH>RV-!~}C%pl6Q9Syf6|$O!Jd;4=H-@q~tPxusTRkUn6QG|>sK^x=65IpM zFzA|f+1E`8qzyJ!%13ioz=`y2yf5*If8~36ZeiOB=Q%>G%-wnNr-h0NMwA{1eZl&kWk90YGX8Xbm3{HCydMgs+Im# z)NG|l_!sVza}ID*<5JT-Kq3wHeey#vJWEcrZ1x5H^1&y({$PiBgQ-JbH(2=*5wX*g z0pSDJKLywSR7U?AGRy!bPsBWWb4-9s#?91%uS5kk9MSoD=Oxf%owoB*5k`jF?!*|T zGARMjih)ZWvUc&dPHthBRRDE0?Q`etx{CyPp))jp?xdo5p|hUj7M0m|-p%UP_vC~J zDM3mKwdbMs7QUQkq?T#veN(;UwaFDHAfREYY8t_z%rr(SGOHwnp()K&o<3Y@dzwZy zZF;bd3`L&?JD%Xq_Sg5j8-Y*{K?yj>tNe7qvqXjTLm{de>V>x2ICHZY>i2gkGKl>} zLGmK%0-DG<2>Nc)AJod-RJe)L{}r$>Y$(yb{ErnC6jZ%W1~-3V)OY|S|AZop8SL#T zRbW^<%3^v%G6#oh1@8yqfoj&E3zs-zU(4;;W~Hv9#&_0%#-*SBtt7{Ky`cOI<-|WlFeQI6w@APoY&R16HxO^_5FH`F}+krsV z$Y_i(Dkx~8H{+Knp=)zIg-p9|Z?Tr%yRmY}lyRxrtd!yW914kv7 zO966T_lC0d&Z&N}I8ouAh-5iacsmmM*JuntDTc!=G0#_bkkv0xPvq*u5wWAGbr~&B zk**wn8Tko9KW|sD!7tVfVHx!ZDLH9O(T<=zTb0P+bOXb1iP;mlX<1>;_?h@`=hK*E z8^d6jnieq(8-}I8Pv_)GP15=q!dW!-H`qqez+ln=4Uu5izrZOQ{(-=P#wNbHOC^m@ z_!#gJsntZ@i@FU-lcj(e6h4^`uNd;6R)BA~!=sb*GRjIoXBa zUfjeOX6k2LG&6?peN%RvpZhXIV|E?iO{?@xbPhvmRO-|I4@JDp552>D^?SmIFCTdJZxxG>Cf*Orw>hf@ zxS|-y-Pa@L{oXTXHS$!DQwnqSWo;=o;lP5K8#d5oOxqNsMC0$&AO&DU?USX9|L(3n zvOdPPJxv#%dRK^xO{Aq>2k7O0$|w41d1b~971gat%EG@_HJ%m=6hkz0Q_tA1OBvZ2mT!>3@WR&&zgs2C8baEm>e1|xhU4^sgD`M-rqPr#sWp zf(Ck&7_@*~sjV3BY}J5iwvS%q!+LTe8uPJ+>hPD#GNhHFw#3M#h`D~0i(3u2^4QEs z=C#-FoE5Ez`FR*R#J2G3_relIY)U+?n-2^kWd?8bO#%H?)RKYF;a_v?DhNn>{-ya` z;Z>LO2c;@ZgaRGnU1BvT#?{vdJ6RDe0{C<-P0quh#nH7&y28ydCiF%btBC=u+PhVb zQ{Us)jLw4dv-qq1)XfAXDhdvhbp6JG;~)2Q%Iu+kVhH}?Zb293KT5#%^&=AEQqnLt zK8pH7iP5t$5MIHBnn;5;o1^&RE=q(!kw1P?CjDth^4n}!)+)>5GD6?j2Cm}PGUE`?!echuBG-Ri(9c((j^L#!JXYOM*}B~`?8N60@Y3ghCq@fZ?I4okM z$qnEDG{VZUwB-)Q>dN+M!bosM{qk33`p-bf1a+Vm+{{Fa6ZM}ObN_tN{^_OgZ~To2^(N!TxhAQPF`2bb~^23|HS+EnE?(M5Ogf! zaav!;S%63tHdo<5r#xYE@nh5K_!`CoP5Qc>){fC~6GW)PiGN<0zwh1sZD7n>~fmAxiIYZ6(z4B2~sR2mBL)HFDABBH7D+~yAH z)Y&*5FqmRcu#{J{bs&Eie%vQH!XkvPsw6d^ilU?Xl~R+ggOn>4ziXw=N5c9U{a5Sx ze+Os>9g@0ka9e(x{Qt+Zp9TWy-M_Xc?)6jViB|vn0Hk8%s7T7-aEaW0U}0q;J`Ksu zIi-jNg<-G)KI%Y~kbLIIx~rwJmK`#S z06@v#`4}G#x_uF_52P&_5Bv5EUSr+*2w54a5Vy0tLlhA3ZdBuDmOx7{i^cxwV9?8v zY3$)55;d>NLP{=^U05zP$<{+IC)%r#5aY$&CE|9RvSu$Z9vFHp2%cDB$e32}=25&hIrDD|zf=TM8n=&LNxVf3M>|cw{Y~kZBr&O#cr1xpDY( zYTyqu?dtJMBpYprSaBHed{M=`E^0gpurE_rz{almO#uQ~FGJ{$#`@DgpXGmCj6FCo zMxt~W=c=S2t${AA@(avnRTI~o)~;@gVVrB*{N-*@E@Yyil}uN;~{uHcx^sl zl@;w0h-az)MP<^cfBM&KU-p?PmvXa`tCf>D*U|%q4yDXN617jm@lePcVb=%)RHrZy zgH=$Kj-{u$|CGVmV*`kvB`#$vycjL6WxnLM{&lF*5-qx`Umg61ZEnUULpg5~B@_Sl zvz0d{Ym{fBdpXXt5f8Zf)ppP#tMZl4b#lC)aD*&~Cs&9`mk4A&JTPbt%lEKUSIYOP z5pDKnX2U}nu0(NeIA|XKKr&}C%GpZIuWpo85xe44KD)V6&{4=$y4v?PVY_zn{5sXP zqcd!f_(OnczB)G#8~i|9my|dCAbJMGQMox>w+P$OtiKCSLi zY2-Dy?M++mi==A$u_gOn@}B1#GCiu2;9r-3jI6Xj$ig^%{yF_?uJHffv3E`+D1uUN z)spagjk&m7g2uq&bJjtG$^+3)%k9$1W?>u9uYL zO^1+Yw~XCezQMxseZ(boPiSi49GCNJ2JGI?J6Z80%UyZa;jmV{!78PNv#f(hE|<)y zX(dd)Vtr0mmAG_Q)!e*f^MT)FH+?gIW4^MkgtgNhM;-EL!0|Jgf_ysRho(_UgM$yF zp34rx<)zo_YsGMW7Dw(O|X3Cve4V?VX41V5TiYMC_`h<;FjfC zs${9G{#E+adF7BcBUA_o_zU8cIQr3MU#niff7w`5VbsUoSj zoNt(`>Q5qmk2n-6C%9bQV0i5Q&`oYT-%n_7JB zTF=_BRa%;$z}iBk^v&S-EVuCa*og_Rs?#Azu?nHYu+UuDKm6o!&f!^$L9*1i**kfM$~OQ z%l#Fj0Nbh)>IP5xR54qWy!ZHwf@}`piXzfRGf*ewyDm4sT(+A<2FZ45sSW0!R1S9! z4?yKt_+O7`=*~`Y%k_MSf|`V-ELp1>b5o0s zQG{gq7~4}I0S5Ovzr<7=3tuZUqa40QVGrRhE{{(XGYQdcc=U)79{$;x3ZkHg4~YfB20>amWP&{Z zMta3L@~_S05U)unJ^mVGucL3U8Q86EZSO8M#@`k}8uV%Tj%Aobm9h4tGGo=HCo>n< z(+|d$QE%H37yi*bvPBXJO*>um!m(5v@O!O#NmV7zX@qTxOKQ75cggwDZQ{7hMHerH zdtF6Y6b87`*ev zvQ)6a)+3OAsn}T%Vb5v&nKmv&F?|AeYI{Z*qBn;)#*G+8Izz&i!EROz;=;>Mfv3jA zEviFslPdZMYUt4VUH}iFT1sNAFoPWUH_RhTVj##6xLm%?B!``BY2Kc>ezHueNSt4W zJ?_!^W=HQOzPI9`Ns@Xsavg_af9{Fgu{zIf)sAO7eAG6NJel>x13p*=5g=v0EtWz? zac0%gGH!TH=yEEKlKk-cvvXO)Qw3?UI7A zG?H#)bG-5^eH$hkhE&?-db>Z*w=9-vh9Jayl4bsHNFDZ!u|0m=D z(5Jvy0a>F&AxJ4E6uA2xP>WGwG2!TY%jguJDw9hT)#4Lx+2Hc{S*+3D&I3LYc;~*b zRUSyA9`}DqN;Y(NsA+rQR2xln58i+)B8fM$S z{fhl5IzCu2K^G@`gU~!=K-vR<0tULfl!jhtvkvOb@I*aF-xYHNMy+ga@|{#NhxVsq zhhQ(5^={GqP2>n#1+)ZGgHF?8ddNuD#V(pP_~Y#ratRR5wrwcO`oB|osq?#Z3{3=* z9w-}pp{1@}sGVH7c9b-(#@y*~I|nV**FxVUeME($VI}@j(03AR*Z!W_JYdJ`H?@Sm zE>-b4v(E2U-UDr$6aOe;k@XAG{l2Rx#Yg(x@pb_vhN9oE@zaZA^yp;)q-=`R4$3cj zCB(7z#M|~aaXCx#ub^yOMiRct^Gz3VfoBlW?CK@4oGD3EJpswhP3zkLn)6$)dSsHU zQ39=g2vNviTQlvqHGqKU9uG*IIZ*@v*K?98D@;|D8|F{e15Z3c<>+^AC$tUZ3GV)3 z3-93;`P+Y>%JexF*OaP?B^=p$=ME9LSjQEy_&Kx7tsnR!QoO6@d_Scy7{RN62zV+N zPli)vWYZwvwmavmx{_-yb25ymohK?N706?ewcIqj>z@1_y)^>4RqnB$G|u$7)AaK3{KQ3IQLCJ z7}3VoQ-T4td@~dBzQ|D%rx~sOf43C?!5Ii(?thZz(`6gD0zae#W72U!JZ~SFi_jDJ z64J6{=P0~QQQPj?V4+;)G|YB2%lR;F_TUmzc&ksHNl&kn)b?Im0q#Mj$8UdZcXrvf z@CFR<*CfS6`9lS@ou6;w6A#M8FFaou(xS;>UBntyr@|Qb=BT_wT24ze#kILxG?tfA z=unIgXBG7R*7nSw@!Y{Q4jHSS?UV~=umx8_J2Nm~f>NjCoO0OPT$pk5Q0redBMV;#}N4`yKwVT3{4QqJjun)scC; ztp7x&=FJ9lv*}+%K4-%7^2Fe^9Rnk_fsKuEO0q{_$%xI+EwLHC{1VVdJx&$7v3g#Q3syZoOzDh}1h zsDE?~7-1DHqsd{|!sHYe41`BiGg6#xs2f!W@XbtXjEqx_k+*JKiQ8%Z7f<>B5*je~ z59}#qKZ2l$62{@fSS4nyQ|c~HHyk{cq5Dja*iUJDwM4y#8|AVsjd_^`e34AlS(FIW zeyH)mAJ4-fSG>pgjWpdv@x%XDgTz;6A@?hxn+nI7pVPj|jNPDsqssa2Jf~6JH`8S4 z8F?zVQJt*|ATo|okv*%B>X`--DqWrZ8H&`X&d8z8C+NwZ>wUgd(86#`^{7;gmi5R4 z-2(CD(q)i@LSWXz2biV0y9TP=Aync}jv>a$JHhPa*-Oyqe=RB<}n&6gD@z z)0c2^PJlHP?mal{u|`9Q_xLGiEmY1SheoM=FF9GFcZ5N;BxuIS&o4S=!D}g{B=vx@ zTo<UNcK&=+1|#`Pj`S zf3zf7em;p1F4S;WdUBBpP~&#d7va&)bZCBli@fG1eIIzePM2$U$sKLcK+uwi+{mQ# z61d)NOLIpwW8KU{Y?cEFoUdlu=z()ZBY2_Wf9Rr8HFYsduk$@Tz8ku}jxcLx=Xt*7 zdq;S^0r3xz`IO^e{yB^-LxojrW;PUN@2RA`)uyc2DogY>-RvQ&xJPL21hhD4G9!NueY1gHKg>Hhi`m{Vy!NdyGKuPYAfm zuo4hiL&vy?BH9?DBbOW800;xRM6ET7W0ZGpjn^4vU7A1fn^2)DvVW^)Q2({nA^T# zYZ#eD*?uXKRwaB>wjXRbSt@jvc?qAtk)=`R#edBNbxRzZpgld+J&wCEyLL-tXlzRT zX<7h}{>fuP@;+R8f7#z`U`Bm-of4mW9w84-HWpRTA|;>p&RVxs^@4v953GX^=#i?G z+w~{&@D}3hjf;3Rvi-?89PnIcu$R1@%i?0)RX_}7C?NR&|3L7YbGoxuzh%E1r2McY z$5^NI+9A2|Jg`~>Y+7eWfwHdgOh716XH8?_S#nNXN_|{6H|YHcaE~Si$O<~#XA=U4 z+VFnXeRqKqOdb+XSH#MSu!{uzh~I2f@;#4MV}=&)0ww*MD{zy99DT?h>n6CDbDUM2 z^cK>)LhArpLs@S|BW;!T?|n*@S$NbKyzt(OLxoFp+Zyu{XL2mt(I*`_c(J6pYB$hb z7qY@uQU!&&waLtLS_CM|LVnIfe~j{A;3a`l+}j*H4&WL|&2~s@S^R5tv@8!Pn)hTM zOS7Ek>0Y2MI>WdGg{gl~5`JbS@9_f)sIv2CMzNkT^PfM|ICmS`y5 z@9Y_iXuMm~2;vYyl=A8V-Bvs=l(2~7%g*FQ6GgRzw3s-t)CQu42&pgPy zAapBPnElUk071}`L|S^td?BirHK)?lHxxNPC!K{?e!r zH?8!2GnHybgqY!FlD15a-nM8{GtB*s9eiuCpVw}IE_?h>WGI8Q<=L1dw()GH@k)-< zHI6~GfDKl(iz~_`iFgQS`UHC;M;Ju#W!Bwdn)!ELp2zk!t)!vs#_xnEfIu++A~sqW=I}qPvO~;LMBWGw5`|x#!mTLWnwsl)#%$FWXu!o5%QHz)TyDq4qg{S3i>@yH<=fvcgm!m_QnrV#WDb|J zlUEn`XmSk)BuIap7#N~I54aktW{4~|qy%HOL~*jPrr3tr1pHEJwFmtA4t?P}Dr2)xQ1PGj__QAcT~3xX7g-=*k;|C!;9TIh4m#0v>#W zq@ba+vt~84F`>-@fxoVerUi&t-1Lp0B^}2i*SaF%JfO`azm?1u=gBhRI?!DEtVV=f zJ)$jg{)zSc1Qvb&yHWDru$(wl5Cl%?b+VND$!1nu#M?pf(zb0^TZ^3AahQ93bfM<} z%d7SfJyn%I8Oc+>a|2|B+pKk1vZ5g0;c({^mD^flEmLv;a^$bW;kYK$h>lDL_T+2= zXw0sSV$w+0{NRo>b=C3_BzRl7r+KoMFC7UJ##?V}_r4 zdnC*gt55z+ie7VkB&=q+q=u#Mnq;7+Gv)oMl##-G6D;Oe&8)o%Fc9?3m)g?OPy1X1 za&_{+uOR#`Gh*XP#E9OM=(%d4)qCLea|yZlN1z#!?A#!?hvE%&L*~5DZVZ)-BDbT* z3+THlqG}K711Ou1nqOfZ$jdvnc|*=Q=$2P?OWiPTW}Wgr`56UPXTi0+wSyhQLW3A_ zPLd7U3ni`}TgG5u`arPQ5_U#ddR0;WAx~K~I7Gi3@A5h5R&1>euY987SV)H&5B@7Y zNIgfAs5URlIHPT&Zq53hj^2_6^$~OCnan0Sg+qd7zE#$C<{voT1K8YrYHkWfsUwt(kR7do9lqB>b@FT9b4gPoz`F zvba*sa(yf-4=}EOQCsU&fS6JS$x^iXig<*t>|I>x<03v4OB5pPRbwW5_iKulmNycQ20HXu-$4(|~=-MCzo2{)vP`$r*%xMGI( zIL&jpx$+zAr@}L@;K|CRfdyZsZSu=Ur=|4`L$zNVRX410oq^I$QKM1Y6{nGgF02u` zM>niEo+k50mOhcyF3@k`pw0*!Lz8eD@58e9yexKV}n=R?aBE7~Q*2EEj<+ zqJfSay>M@P3dpPc2epOJ3i0x6ncvtZSpt6k<)@*pz5>;)1+O)+Z^er1EX#iUp@%&t%>qAM*^){WFa~9q#Q*L2g-2OxnLbxvnW|s=`(O`a-vy zwI~UnXEQZB#@|ae14^yd|M()L{e|GJ=S{yz;C$mO=3&eD_}U)~tenF8Fe)^5$-%H_ z(`3~0Kc<{vHaW~VXq{H3{&fE7rE zMme=#UaS9It>pVJ)#D7|UAvptH+Tu#WF=e)j#3o_dX>*PiQb z*W!J865TTaMd#yfFH@^@-VVb+7m4E_Tcqb%HvdH8ghm z;0NR)Y4LGQ|ID?Cpk_ z5;JH3Wk1v9>CvYqdxskyZ~Go$RFH6E6)d;GbxA+J(=GW#hm3-pq{Dx&o{&b3_5xQ2 z8VsvK0^gj=@FQGnS&+*Qu&P-c4Cfh>v+G9_9<#mlMc9jdiJ2<1%BY zyk~&WqV{L#G13A|ITo`%)dNrAW{*^SJNAXY$v#*uVDr-Te%N4FRNg=Fm48dVn{msl z!%DV%S#Vc>cS6&cuCg@H0eZwQvM0X&7?&gCvnqO3e$Mc5p;2p9tm1;OZfX-S5PB>x znp$YVPfGFt4|X`;fUyzS=~Iz@Q=tgCBV6kkIwnMmI5OMC`TM$!mSkogzAg9zI2p>L zq1`g#6QLjGUPCx&5x=XGBQf*4q*aONaHlFxC2u#kZm#|XM(4Uhm@Gjt*G8CxZKk90 z=YkKnBX9HkJ>jjZ*yByWC!;f0VXBj_c-cfbMw>at)BMAxQI1@cCpDuVRm8 zED{E4jaUF)Ym;4~#U7#OI{WzfmT^r*uq}+O=F?tE4rMpk{^>nW7WCIm>O)HbPDmf3 zR|b7kTaj#Zl*nCQZjMYNt9W#!83KKwe`M;d7sXK_wzLd-(R|3(f6Ev%eNTeAp?WYP z-f~Fs6UyCn-0D;wruDphhc>+!|2E;TdI$nt*{7Lz(cb9hDh;Uy~OFrToHL88|S7uS8o3!oXY#y3+3`9Z!mTQ&AsGr7&i@(KNzU=9J zrbRTmh#pbGI4H=BxDixWPgcXCeMXBr0~->F7fJ0Kyzur2=Yx&NuQ)nWBVjL{u}41f zZ{Bt1QkG2F6l0VHr6xi`MElytzm8Ji66MUgv1jPP3@p3u-3NKfB)6nTzL}Unt%{Fe zN#f^t+Io9XZrY+p>s^QP@E6TzaHbo0afeR!bN@ui-^aVJ9a!Elo7oRr{3Y^6U$wxHl+j!cO)@7`t zDQA@ke;|{%aRO3GW>?U@z@(rc%gON!cT{f9qN}Wn5)N=BATd>ccj-8MfyN1|?`Z{Q z447`}G5gzwfLh2BZ4tFQ)y!?X1$6irxR%C>oqbxhmC^%$)}WZ_tW;xLZ73aQ*?+01 z@Ak~=W8iYn& z5scZqzNKbA0-6ac+xe~ia4*o0;J!io3wQf8ZustRsxeTEoVZy`nvp#(3VBWPTp8@} zl&S7(VZ4RZ3{gS;qz2bbN1?&pE_0Z(Cha<{uov%Lg~PoA3}}b)yBeW_zxI~2V9<-$ zCH$LYNan*dEp?5o<~9Enm6+gqx8!!POm+JFVs;Dm6kgjGy$luLf}kAQY*#&1m|M|I zaxS*+ylJ3jZ7}k&ie8o{(P2W!Jz+Ir=+(tSAGF&~>?gZylNnc-oOH)~y-Yp*+DB1k`@RJ_8$#ZzAPa zuRj@m+#@coUs|(DT-e9mHY$4ByA40RPlJO(i6X|H>05~2VR16mlSG8de7OISKLilW z0T~+9jiz=#_%s~~K6q0g&0zL$Az=n&O89;>m}GiBpzF>cFS5T9`oK7aCvkHxV0I`$ zhfE(ICo;3ANNPov1RWs^XVr5VrIa|cEH8ug9IyW#!TE30qksw{b2)loVL4Habd;wg znUlsD1M?_r)C3Llv0l1L+L@7$cYlXqW$!LJ_^V&#UH|i-C_Bou_ty&6Y_6JVSaxic ziJ-Am?u!=lrMCkBG}UJk;^RaP<6=}`kdtP4uHyv;p^d*&*wabQ7a15>nUVAS2S&bc zf!{g(e!~b`ato4^G{%xtIDPC+X6U8M14{+x5?wBAr&KTQmg*E+yM*iBx)rOvYejm@ zLpxNgHNtr3WTzJoP+F4^WRwLvf<_B9m)y(&P02;q10S7QNyPv(c|}DUskGSxK-Jx1 z^k(S`+072CyIdhH$>&cBGcF4E9#GMm^2A&2j$7B04S=@)RWuw4@YgtmXGD>Z|X9 zWqwD**P?8X6i#ZPU0xKr#<-;*l{je*=$VPXDjBE50wX8W3NH@69;!S5W8VmpFhJq( zU6ebT&8+fb5Htw8w!(ws6u%*1m{ zL-K)m-S+YO1O;Tk+n7TIeKEi5-pbd?N49LvfJp8FO)`vvk3b7LK_xZ}uaUL23bINN zakh%+gDV?zpD=GIxAfp7b=AJpM7LwGavL@C zVHKA*@C=SnM78G9EkgD$TB$XN^I%Rm8?|CFkVe5AI& zGSrsENDrOrL^jm;o$h{>E7KB0)tW&S6kfdh6cJDFNN=SJ_M|!>LpWI@@(?$e-*&NLuAH6KV6;LMV`(rp0@qQ(oXtw*u3V%B*;o3O^(SJ;8R^AbtRc6k}b+ z&FWu~tfvw3M>+UY(XWmi6~VjyMKc?(Lb2A;W5BqpsV3TXaqtOD8gT#dckBzwBYqIc zpG)$5WAgH2m&t1*S0?EGov+XjH8vY_66R!0*jL>fGE4NrA0MHuWzj&^PE0jGBY-PX+Z`^liHv@%!?`Lz4NM=V;EKk5JMo+ za6dE|8oIQw440x-sgrW=g;2-?cX>ElOmm_2mXAqd= z_U>LTik1*wap%U40KFhCNIu>YHtYuNs0o28%(D!IR=N;~J$HZox z-CvU2X(~r!%YDL{NONTOg-S5a?;-LlArr*&;X#Qmmj8VNiYkA+OF19GT0xV1h*fFu z=O3Irf9@JyYU6{lr_aC*S?xpOGkaEyL|ps3=DaM^)e&mI@Ow@U)LnQa*YHp_+MNe3 z0X$6`XI;AVZwerd^B)!`75e4HE%V6&w)lmbnu?jKakRba6pg9r)ImUX{n-1;a*7V= z<>t~QB2=m^yHaq}R}@=nJ8-qW$dWnVChnmN{fF43561WW#3kP&D~Vh`Q5IM-TZ~St z*N6`LPyCl<^|aN#l_G`niYYUQ{^1;9r9BwUduEAe$-)$(I)=TQ7#Dsxi+s2 zlV;~IgF|ceIGA$=K1x5nd8TE9DDA6}<+FdXESH@zNUw_Mx<1dNdqlCmdd8tl z7v9Ey6Rx`b1@awe%#CZY-mQ3UAa&r|V9m!tPU2glbQL_%Bh>F_p`3A#(ZIsO7N>eM zL5cIKSCk&7pk zO+%>J{us`z+wq;j%l}dN;#lclR*(c7On18Gfn!rP65}8GuOcc2AWI zyZQgo77*=gio(d~Hvr5QDG61rTuI!G2P|Nl9Q<05mBn!44IGH!pg&wM*)Z)P34J>KgKzLxhQ|3!t?m!m!JInlhw{M@#mWYiWB;PVT0e)Bx+w zh0~NWT$>^m+3lmfF}sH3PM%StUzKsY-iYUie(J2=zIe;To(JV!CXVOWO5eDymSA^_ z@4D79;6H_ko{iI?%OpK&eRv2Z-Et=QZ_!AguJ9HAM1a+|h=}-Dat-eSw%ByOgZ06} z+G#*x@uuPAq!H9DbthMhU1;~e3Kn#_- zRb`f7(U9~_cZV(HG`l%e-4)nCVUd!IyA4c8m)8S!5zySoL8d82UOO;wZU6&5!~?A+ z2DWlKqwn)Hj_=CC5J5y}XA}dzxzg7Z=BHqJ(OdWE%4#|H`3zF#1$uV(Q{|(qJ<{~= zt7_hJ+P18jlwR=F@AwM+4uZSt+09+!xzTmbwP1Bu4EnCa`K>{?~WEuI|JnE ziZfkZ3st_z7V&feM${U*OWli1so7VBo4?(*TaB=EoJp5`QxR}v`NIEpOYT$A<8r`t?;~CH}&HCSNEqH^KXGP-IM@Ny6)B) zzPTa?kj|K9$oC?)hW0qt)-ro>AL$~G8&0XA9n)_gje1|TdLCVh)NyWtWsNWf5AqIt zBmEJq^Avcij6bdXU8#E{0ySPa`IgpMTi z3wmzXf(2`%mHQa#a6wp#oo(9LqHbSxMGzJQW5!(%+X;4{TsRj}3UrI4w>=dBR&XlN zZwFu<>9CEs_?P2GBOv!JM5!iiFbvs&37c-#msTBlX_?;AV1t$3l6&c|ia`+sg?MG& z7uJz=|FBF)`jN8>0R8bzeE~H_wGhP+_?yQ=Fy~BE$DHYC(EnC=)KhRwkj2?jeFJlz zpN}+-VGQZ2+ccE6VZ%FF<@aP)$R~*6CfvJXNUBP9a4o!Z|G0f|2j`{Z>tA61am*%j zWim@JLe&?~fdL93JZgVZb7h+hx_m#?@0|}U15-&O(NgkSu$v>i5o0f}GjdNis&m2$ z*vM>M5nfTUl{2K$>#4+K6Ot<)Ft5o?HV!DjikCvIw)6**dZV zrD$RM=1_sYY?ESu+J*)@ByUJ)=aQ_^s0?K_H)HPgSZ8NiC_wsT)ugQT{%T(vkt>A2 z1O?IF%i_x5cgBwxD5&zRUjRn;O0se@4R!#{z*jn`V}YkT!M)5G4X+WhbS;N6ODNYB zY^xaaFFa=HzL{73YxJJ+ITnXRkV5L+S|b%Nxu+HL;2%08!z>|hp?)R2qYhYY@HWCr z4!P!(O-L>HrX7Ip;o$cJs)PQ_3UP`7(-p3oxbItDLRk`}+d7GAs~uKW`Bf`bqc=Lr zom^LSt3ny~&KrNnkk@i*R7UYm)!LT?MDq7VhjyXsFEvxfxyC#nNZ3eMr|ARz7{%;( zc#8gaR#P9$y3r&0SA@F9+Wu-KkEGDSk|rA=ZpwF=^7&_~Gj8Sf5Oaf>MvxCGxN;ok zJB@Zq{2_iE1UOTODy8UJ+>t}aFGs$NGgkMFqwCXXS66rxjs3Ol*jJdoW=Z?jiXeEZ ztwNTqmu8!ht5Bhv8bS_mFs4DS=P|=awHD=tiU;lq{qfpU6`X#N-0nyciXYCLD4a)n z-7?zUgnKEboro6oJ`OhptEcir%XS9ql2^3Sm>Fx-)dprH>N*#~c&ZXoeQE@8W?-*6 z=!A+*zz(KEf+bRhz~^~syC{pZ>>p!{>6?v**pj1Y!#cwAoC)P)OxB5i2l;auP!3IF zGN=cn%>EL&pYLGMd!JOYqD&6U3Kl?UJGxmG>9Msud0}%BzhFdpSba$SRbWa0Ya~lP4Y-BS%^h*+b7F z4vO-s01l6uqCdS@%AEP>ZS#;s;W0(}dbv9S6Oa#_0BGh;2%wu|o;5O;5f4kRMHwdy zb-7eSZJMo5w7>DmYw1)=F&RK$`_F=iLXNRNS)-65Fc_h7l;)N|S1VWy1&3b5&>Qp& zhM85)25 zM^4jzMo3+)(&1WTCX$PHDpyxbm$fw|km!9MK+bTjYyJdLu^+rBAwd2sV+2%X+xJXh zzBF|ZL88F%`4&x`X9otB7HUPo=$5X*lz6oJLcLW;dzA^SF&Iu#!i%tap~8EjtY~Rk zMx9+{+MJTKz7ET>CaUCwC}HgG^4{->xy>QsDlWh1GW|?4mX(lxkC+RHZ5B+@jyx#1 zwET1+$)E=^M~rrJ{4(~&8A(&UZz_by*!%GqTjO$vz<}w2mFO0TgKDfSj#BK;h+g;s z@RuMii8#067c%lLRuuYB<^VmlmVyMsSJ2DYHV<4sc#1Z&>#c&*y2|53_`Q+&nhGYF zQT-NLNU1pFGuD{bAMyKnx3dza@XO8nZ*XNPTLX$Rak5Y>%}LPU`{r44r3c?cSP7T2 z<#}gNrSq@zAPXQKQPIAS@0g?Wr$l&Jxz6@v`$qA#L3Zi|#eZqLxd(@0pmxH-dks+Kj}Aw zY(B%DVM?084~7jqM;ApNx3t{vb{u-^xo?qm-qVrTk{gWnhtY=!LNdcm@N*~?Qq_+p2)~=xbZ0eJ9JwH$X(A>q;~W)JsGk zU1f`rd-#hylYZ&@8J*(|zuHkFJY=-v=L-gZL+w|V^BH;}E$`LTZGD!>C>et24bMmQ zn?_d0)Zci37OfvnZ~DvIr1#w2o5ONC2{uz#(kf|b`pV)#Oupte%NB@Eg%?~Iv`JDgjn=0(t4!yjDIVTpN6D3`^`-YK36WjR0gM<5f85Qo9G)etZ=dZc zXrG)eppm{q3^WGkD|7^%zzyhzytJHA zo5j@R#kX&jv}(Q7_?>{BeataW>^P;Plx>?ho0Uk<1mo5m5(?fWMGS4MyHB^S>~QIF z4CvRV=@|0y5X6@QdZKV3q*&pryfMZJN{ctj2i(MSCGyJP_eC?<0{m`Uxf?vw)qpnj zIk_*n6;SV^SJc zIP)o3X+d8<=ps7fdC_H#sG(>KK0y*C9-g0V@YHt90@prh?g|ALnEALx^qu_LpDsK+ zq&WgEa#^S>UzW)4&8Tlk-Mm3x&0;t?N4@<74#FQs*Lw|_FUug35FCPgOc`JU%ky_s zKk@^klySTF)RM&*E7!k3WEfXR<{KY@BK)+a`a5>MAeDE-E@9k1H?8~Qxp{z1ewwRh z6J4br4}V^-LY8A6q^}fT8X|_P87bBsi1^KXgWlEY8umP4WU@V6a#0Ria+eZ+4dW@e za>V;T4#7FkPW9BQp~dR-h(l^sU!&bmqc6`YxWcES>{u{$k8cBe`RIHgkw3zYr(L=OUwLK$Pc&vwIyuK{wsbNnOAZ=1%N!pFedbYs{i+olO1BQzcQW8 zSzP+gQ2ON8Lgwmtih%#y(bJ^y<0KXHheX$}@Nw9m^xs3E3%FdwN5_a2zJ-dRE>KOS zO2OovjUrBr1C;xtJuIBkgH@5T@6>TFlcyPCY!z8+urX95VO!Qb)r-< zaIQUp`=#WG8WT@XoJi(hDGULmq7_Mf*2rim6p^b8hlEJbAX-Ejxw)&exx*n7Or1n+ z9gVOOPD_~*)dtf+QW}Po;7r zxC^1NS4{~vxk?!tCB!G3$_6kZh>DVI2DGB@pGInzWtuN3_{>_8u~|q4i!6eGa0^5x zi+7FPfMlM(7o?_$rJujOrit&z>C^V17taX}8MzQTR)C}@gka5*@L7|n%qj~Pjg>#XA z=@za4Twz5kU_%pr5^YdA6c=q_y@iFULX_cd;gA+eG0q)Mg1FjmhAM?>GhwOpF~+;$ zsY`U?SXr2C`(~rn^QYbPpM|)G{hjlk>jht`^OgvOtCr_PodH3knUPo0<-1wnmv%39 zhGC$%CivD|m+$qBPPxPfTC%d2%1!I74-vu{xJYLUts;~#;|y&d&~M@$CGU`7Jy0_U zYa11=)n^`97yQp8d9vaDsMvHR6aItM6vgD&C!k>uy@Si4W4vc<9YMNXK6vMkOblYDl~@4ET?J`)(h4%zzX>AOOg81qZIpwe&)wg+*=pU(JrUQ`$3~GIZc97)e3VpF zrfAL`mv39fjv%ch#bqIIKET(9^=;YlcL8x_q;kE@CpVE2G!~`XSNxhD+%bIB58c+f zESyB$NQe|+oAdm%fmcE0?MiawvPe$WN6g&^zrdS5}geybS)%+G%vvN&h5w?!f6RXVyz25|@&2AAVuv>PDy1C zI6^XiW6N;KEzM9KKgKy0693scQ=vbbCOaWwZZ55->WOHDpP7+8++qsuay84PCu`V+ zdrsvRbt|WP9V9;#!;GCf6mk<4c$1=-F{l5GWQ91Mj)Z{($(GigP~G2v;x?d#Y;4>3 z*(O}{#c=?Xa<9-0${Ba0^bvXFE0Wy21g_v60!L_bYMOLCN!TM<-ph1M!gBg%3nUVL zCn)0Wu<2X;i=82~+xkr5N5J1Yc#S3hJmsZ^wQvuCF$PKg+F%o$)$$nAZs;&T-0T;+ zJ1blmP1b1AEzfV~ThF_%dr@2!sd@@$)6t5|2|?1B<`8EDJa81k)rL@qvilp0|N%yrM zs&;i21C1Lj1!noy+i#sbqe?NA@9PjE_?8*zQW@pzU7%YKlMI~W`R@1z(yzORov{hpqm{+2iJ_C%rrU7ZI|8>TVU@@d$FJoiF?iY&Nz zTRGZ&!e|kGWM|c>@*0Bf&l~?&h8n}wxPD3v9Xg4M$CkV%%-;}z^m-SoAwE>`TIhTsh5|3KFBHmNrP*$CnF z7F>F*FPHjw1Z^Nq?i8cq!tizp{yp<6hS9N+^9>1xV@nsYLGK6xMBW?E<3I670!L1_ zXQlc9gEiUT5>g)Cfc)9PBHu!(@o$xp(ElYZT3^Hsfj#nBX8|>uBYO78kL?wP3Cv?m zK9Q5&Eo`Hooy@_@D{(M@a@r>Vc8*2#*SAh-mjhpss?sxeD;+g1EF5{7D{QPia?LrF%zQWo>K1q?P}Ws?&U@p6c@tYuD=PdGLz1Nhh9tY@nyz)-1-PS z*0c$2VB;nX0=+pu5qOFcO0?p6v#7++=9(wJG;F>pPv7!xMNL+gRRn0c!LD!I*YI9( zsS^0V@^G6j4}-^bIyxpZZXhc%AXLWSR69znjbr?v&Xkp7%md^heQur6$ZvY^8xp4v zZKo|dv#tOGHzF}gt5o$w5ez7cawUo_VZUwv3iErg1UdvoRsC5sao$~`)7;UGug{D?-1|&U&;Om46Dus*O7PsHP6Y;GL1bTA*OoL2Np3_HCk;+jF0Ok zrSnYjF&MpG`mEE;UJQlX4sm;Ej+vTFO~sw3D0=BI1dL;wT^n`%tx{Y@2WK~6M^{4d z?#y7xM~B&5N&C_tZ+sqQk}wczw0m_r8)WQs)>+2#9g|+197M+HEdbz+{^#!8I$g9> z#Cc3J9DV=zSMeb+(u^MCV13TEtB;Gp0w^~z9h=G<04l2POpVh9MPj`P!AKQ19c~xG zI<(+t8>7mQV;F0Rbt?8&cfH}QOA(n*1J5TDZUb+bZu)37p-oX!INEfX=1ZP>j$)m@ zfl%^fr2y!YeGDgv-|Qx&O_I3G(2O@gxq)4uHVTR<1})v#L&0GkqUxUQieLIswI+w^gn07xvd>;v zPc*2G1teZ#v{WuUvuPw3)RJ`gQn0P~l9+i(%tcX7JE_Q^yVxoX8=9EW{xowcuy)EX zjI-MxuQ7g;`p$y#*@qMV5Gmm_kIKSmp$L83>K5*c7zLUItFW%pZ=<($=H~c+b$bd9 zgcytr4INcgEX)b$-ea|q+y9ya6UYfO*O+{mu0o2souG&m5$vI17X#>`hEOzStWjea z$Og}{!X>)7V@T0SeqI0h+r;N1uRKlP$XN0lyT&(KE}RV9vt&k9)&GaFZ;Y;NUDob6 z>Daby+v%iZI~{hMbZpzUZQHhO+jjD;z3;g=_ntGxmp^OFtTiTTR@M7f)l*N+)2A~3 z8zKe_Aa=;-s^85&dUn$n=-GoEtm`EpfX@dtz)*^|Z9s!)4lr_ML=CNtD)!$E$>=83 zI0W4v-Ix~$W7mC-KYd`X))KW(iQ5Ze`;_=%e%?x$0!M}%KEw%_f~_{4riJFSXl*3JNrEVW}312{am zQCy8<_mJnOw|SVNQ0A&>apq8$wh7cvZON43>gBvXwyG?lQo(ZbG-K5LGBE~$WHujG zagi_1q?DkYerav}`9mfgu|gn&tDew0fSPu8~cRT4McrfQp_Gk(}q%>y1QEEz{x-UkNY_(SK(%J7Snx zk{F`lUf!}n8Yz;EU&8;nZ=dIPBa?#Y>-dv#3cH(Eo~#R?^^Md8cNE&BSq>MJQYkk2 zq_el??xE-uY#6DE-$9M*B^5P%$G<&DNzf-E1!K{s8{Ii)TFkab(|HPA^l?5U`q1E- zI}4bxdmCk4q~FuJ?U+!#jaQe<>`pf>=y%wLmBd(#vsLL4_UK`Qn_IsCK4n!bGZT)C z?N!kv<8YUvF}C62MjDs9p-M4Cobr^@;@5~@!&f^!SH0zPYphbP;Bix}^@ni|$t2d2 zWjhJ5AY-yaOljiEG<>G*iU?*h2#M+X7Be05aQmS_#wXsC2X!T5GIN49K17irM*7>= zRhs6(lfah)9@K>G3f-1~OY)PC$RFC5$PR(;yv$nT&P*@+9){bRrNj|A3VXE=W#>uwsgrl$h0OwYWfAY)eP?P^nmf9 zOwXyEGnFAIb3BUB4zwtCCm61Le#m|5OrE7Y-=!<*6Q>7gh>P)ct(SCI1Yq_^d;90J zfApK-VAm5Ma&%9u6R0)jjGqxPHbSXnN?cw2W<3;neTQ00GN7%)Lg|>Y3vL5PvsRe{?=?et4RV zQNcc>n1f zQNtvT!Iwybe;+3(_nUU<#T5JEbh71JNFmlA@N3AesZ<5j7@a}@0QaA3EoCj`0H1En z#j?3Bbfw1UT&KRMeUOz3=*OH5mIxJ4>mvh0YxPGl|9Q>K-(ecq4`I_)L(Q)2zbDL!ooV_Um_~Tu<$?m(*9<}fd+*! zQt`m!Cb^c09w|rDyD4lI{wC~)@|~d&MzI_StEg@OA9$pdi8|sEJjla*n30}VCw^C! zwz5B2?Mziz`V5$H$d~!wR?fX4pH*Z)@X%H#(eX>;^2-1%qBGD(?($-#p*G)L(jm=8 zZRKR;MT-X6IIh!Bp?5?SCA~(kXZe$R)ITk-7ee%o0&Drx(lG`0>zOuszQ3VorjqM; z){9%H&s~z%U6KpIrPI_zTdKg+NHWh`t}S6lrXCKz9lAJ?$*K7y|b z?G+j;K&4qqv<)yh08T)z&%b8}p%lmw+`)iD{1E*OW$HXKV&8ECh$`8xbJd#clACra z$eJz0WQ4-8*K0J{G4;Url|?aKIB_W|omcz`jUpoYIE|9Kc{uO~o=`9OCrA zgD#Ac5=O2p3CSh98u{J=6eitw@(g1{Nl+uNNAF2ebWX1d?4*8mFSF;k5T=ZwB zrMMX9uAnFe6;Pe*H6n*nD_4)&0!YKur6TW>BMH4jCg4xD`Kau^g3p1Kpupe2A$_A! zf`>is2P^dHFPn4_TkBU}9H0c1<@pvdeg=og8ThLe3k0weSP7e>^+^o-7MUOqgvU*@kZk0EchxTl^EN2x# zg5dqmTdJ@(#QXAV_+)VdV}w3Fjfl<$_4$jln2&QnItEZ`R#D>`2C&BT(aWDJ*xYIb zD5Q4k?2*!5wp4DSN@ATZOFXikq>P|!lS_kJ%lW`W%jQ?pvM|sb^UY~BA6#(PC z`fa(JtQGqfdjeDK@wXcIt$U4isf1Bs!ABGb#z?RDXn%^RG-!JeAwF5Sll^ih!1ceMWc3CjfMn|2!#MFpSa05Tp62` z5fXF^Mk7n=r(``Ya@)X&<;(cdfO*%c{RhIRdl3BWDkKbFQu30NcP`ZtH&=Q)7)x^34&E~W*stB zbD4_m_@U9IRaN>C25(h!Hkgr3rq4)B3v!?ZAcOzbDbf33-+J7SNMhzTxU9q}28JhrQ zS+VW-qAx9CsFC0BhndObXc?LG0IX)L$9$?P_5d{ed?b|^$=iy22*k3SAj!gB66)~F zhcN#ZEnV&lFUrhK z=xSk|C542LCN>Z$!%wAX64Xs8(y}3b_W%R`d%h}K;7d$IAi5I!KY`5;WeY&c`1f;y z3P7AS5hxt|&K;bGGR&^(LLYELet@?9!{hpwB9vfNyfX4XLQUSQzv;B^8sTAUPqc{I zMD_;WBjKYYNX04o#Kv?;ioyUCUqA;z7iV0-0`sJN&Ut`@`kHhn0X5a!0&EyH%m6-z zvX=W8iNmn`9cN}G{3xx7j=ZE~a)JJxTZMkW(92iJ)Kf%=}b3huvVUelpq99=tUHl)}N*QvLiaBnU}bRK;C`k)R(PPq@#$2*YGd`aAh*fQDkMN4Vz6$q* zM!~`Y4L$haQE<4ftq_0>V316~g9>`@a=?7NsbNEwRwGZ+AV4Q+#C6vzZg_dAYu7*&-+qv`cjks3hmV7AtGWS!_rR{pye=uPbdtIcg@2 z);K7#j`R2lv$$3*bG?*x9(xw0(D59XuVYd7S)4?Vv=#FN@Ib^%Dc5lOuWjx?+(a>* zoDmX-MZSSM8vzMEdY$$zP%#HjVHx$EVU7}2>?%iknjDthr6%w{h^T{s*u9KWzp2lM zln_Jkh&X>3i=6m)54N8^!l~~!V5GcBXhfL0O-c+#sg}X0!c|=895OA5mFzC z2w{%=NP5WocS<*$A5<(({YMI_&&~>lkFEE&Lq|E|f$ML1V4!)Y>=g}>eq$VDp~>O4 z>BfG-c^+Ez8}gr$tl$TjkM#~UBz3!^_o7e+$=C;H2mAlc+xO4poB%FidD|5L5+}fb zsEq2_Gzo32s==)lEZ9;!&j&Pr-m$TQ87IKW{kPT`FuF*R9)@5-PQb7%vuw3fB21!K z^+&U6+NL?Nc#Gi=I^HJ@;hr0#(0;fhj;6+;yw;@5lzj&)K@HLaaLT5Z9QyM6jYR6z zqlkxx5{C{s$Q%v`k&3lW71|vjpK`}2*%EYYzXj0)4t@u##$+yX&7UboAMAE9@<|If zr=9|&BI@jvZDM2DLf`xpTQYpEA&Ml`euy5E{E3#kVf!XoK)@WJ|66<;1;S6^a>jz5 z&wW}hBj?rW0XGh8o}1XHW5X{uZ=yK}nRS+8Hb@aqBB3CT$xr~;E;B&xt{WS-wnOvc zEX3CTrmK|PrctySgxk**o2H9mLe`XF{qEv6&LsvUtyaKDaY)TWY`3K((>huOK01|C z$=3Yo5n+T`I96FXQ}GDf%<&D<2?s(r3mZUDRDge4M1*q%J*~4`03~!;7KMD`b2=;9 zq!c{H$M-{LNW5&Ve-aivW%Y7hU@dB)v!R-0LlfR?c!YnkaMfS#6)@w;KWMFsGecs( zBsRR(F7NA^Lv0J|)SpD|6Mf9;BKG8jM$jB9N2tLae zf%J+Ym^u6vfeD;5ACyMPF~)wWuJ=~tTl+IL^Mn_CZ>xqSx)9K|Y=|h6%UHbO?0(ZS z*xpfPl6-j4$yefy^0XZcuX$&Nj3}DROg5we@+wZ@-_cbof9S;Wg+94XprqwHan_5u z=C6?`fm1b@4eKD*p<;T7$O&BdREDA3`5c_HohV`^(u%veMa$>jaqv~=gmw*pJ$~YA z>W#&c#u)}C$6@R-FdO&}(}5wKrBJ_nObvo6dn_1qR|68?zbcH`1&K`E`HY zWo)Q#bhM6TH+V$y-(Kn@r@Xb<1zo;mWL0dH@D)MpY$3^uiw>Mm$MFKCw5RxY2|LqsXU9l?o|OsVq4_pvR95QYk^K@zN))R(7i4P9kh@7!3V z5NH_bVChJ=w8p13i}O979B6#9+5RVrI(-ULq1(^3N0~co`8_TA{C3Ze0QZ-QeGu=$ zIFrUZbEqcT1Lc{>n$6mlx1rX=rcdXn8?UK+Sex&C5Ls#&EHMf1bu&}1Rm3Y>P4A{H z+DXm(w;_l1Qeu+PgJyQA2;=;3dm}0nfEf_dY|@c-`HKFi%uH82uFk4%V2rC3;)QpU z^ztdrbPQkb&d$!AI91XT01UM%D2h)%q<}=s5WjxuIF^r47X@&rH&HYzs5AadRqk^9 zSn78%Z)auW39R;1IXr~w({j?B2ZeVM$7kS~{T#?tQDD$)Y;<%7FcK60rOlOw#<=bK z_4CVxgRX`;V)m<0eKel-NTZC(SQktjX2Oje$iL5@{eM1M#0XZU=%re65O&y62vd9a z3BLF|r#Ndhf^`XVm)Fl!v2nni#MAIpntOJ>L$FAsDCgR6I{v9QP+_aw;N+A^LB-+Z zo#U~QbGZ&&U`xrH#52R)J;c$rv|;PEz3sske$zidrq1P`w6f+^wU!l*88Xi!SMeWq zt(ANIy}UBt=B4_zIkz-4hEm>8(pQQNzH<_<_WlOGa34X@$3ZUX^s~KH!d2!339QeV z&B5{a*DHJzR?&3NO#(<8nlO9i>84{jvhOxGy+ve_y~0}|E{bbnV*8rmVp!;pXx^{w zz~DSR8i>8@yLGmB(i%Pwzl{FflHObkw_Y&91zcbd9@n+g{9!J-EA4~3nzg0mB>WA- zFt^u&ym3AVu_ojcouN3Vsyg$vd}f7VLmgqtn51w;eNd=vyr19WDs^2A^C;}dJmncA zQAlv0URK`xehgELwSTx5Yp*|TAG7h!-E}M-eMR=Y$`M}bSA9Qh&%ed~;Fg4{p~0St zushai`%_k2*SKtAQRN{7H%_+&wX;Qy#wMDcGjx^q(YbJtbZn>b%r}#j6PXaoIdJFk zgs+FoEU(JH-5h=Q1hQ0BGg?J=*>0{aQc&_1)3Or#V@pF}bq%H* zs>9n}roS@&=2gY%Cso!*-{{yxgryJMuyyFHe!B^8LVK;Pg`As6YqXb^E|;(p&c~vx zQWk?a)x(uj*W8J@j#uj1*ZHi~q0-Ry$Wv06)*TXN> zKJ!iUoEj!j*RJpNPZ|V30BcTddIK59@#=vrn;+G>zBK1!h^F)nKY9w`UK+kz6%B;t zqWg_Ks>$3p^*HynLU5PZZ%~nK|EB>K_1*%>McXYA;7n&&3%RPs3@ze{XSuU7rnrQ^9;?2z^-m2wjQ5{CbM=-D6FasHkNS#Bb)L?Q1}E$2 zP>3A9Hy%&W9!_?7>J3#;e9OG8k^hmws~I1|=o|R*#&wzq{duS?V-LkF-ubyCb?zQ; zbicQSfiLgnIP}a9HF)oeqt79-wG@yYiTG2D_6ZU|Xx)uq=-mw?nB@P=j>t+%0Oc$`cZYcgn%?CnDI0_b`XlN!b)Hmjj)&Z0&K`XUSTL(5S) zX7>eVe4<0!%^}ukHF)GEVqxFaP#3<(o`PpPyY@lMP5Mf9|C$_JPi3kbtQMXB9>=<* zDw`>f33ZKZe~BV*8;COW?%9cOlfjtuXU}Abyv4&^NXz{-bK3#t$~L6a#9)I!9?nruxztQHc{q8s=^NEpSNfEt9Y3qeCGdCmB zsGOe0jlbTj|M{44u!me}&1E3=DOJJqQvaZ64HZZW=J`sqH#~-JE`7nl z9#f<=$7+Q8Ki9rCGiiXa@^1zMrZnn(I~99lk9!t>`m7VxbD`jn#Isn(FKV+mYBF4HheJ?~>VvFx zHX`C{rE%JH+&Q>5ibcLZi_7Y6Wv5mSpXnU2Laf}nNFTy)rJ8J7VOQP%Cl`Plnk^^2#or*XFeqoYp1R zt_K{NGBr#=Hib}l`Q7Z}#^X_fqgVMCfny}HgXVcb5PJ%?SCbM>p+@-A$^aBrZ=qsz zM1l&TRnYVGt}r2LB{DtMuSaAf#B^A{)e0qGT3@Khv!)N10GEr9XRO4^!!l%y1WQ2rn=@JB@>lCMw zT1Nte(Ks%y=Z=gSca$Kf1j~;V8&HHp64k)r+t5zh9ZaQkf&@K`b0zxAARb>vdo6+& z5#~h|&UY!(A1mfpznQ2@C+msvX)GAalKcm;ksnL$6Ue;yZDsO1x=*ea`70RkL+?tr z+>dOgGB?S?Fmh&dF!`qb;4aL@kRAl zvi)l+>TJ+>$FGQ&HCeQko2o34+sm+r!<#iQ)tVC_M<;(|ug#Tb+<7=d;tgKb2>a8% z%HV}|h8vU9^nQ<{`D$S5eAg?CxW~(Fbw-<|QCJ|U&1g-5nS+7@4(o|dnY@YjBSUfsSNK5eBZHGGo9wE*oC4~iud zR|R-t#!a@hrhAhCo$Ab1gg2m1zPD$qJq_3_wUfv7!h<52o*c6K=Ax*QfQ|(s=HENi%r7qG~GS@#|jM zLEof*p2=VyaW}{&(HWo|iic-W%-!6cmmJ^$RXh4@T((&y>}+xg=Ulcx2$|>B8PT0L z*ebrRy#PCG>y)&uJ5|lVt<#_Uitf;Ui#@>zbU;ud&jaTC;onA(Q#V0@3W zm|)&jtw9~S2`wr(uO0@f`ZC3u!-#uU>@k`SqIC>8D6a+b0qrpVz>xXTvs z!~fiG+2x}uDOFm?acsV}#wxlP#xE!^zNLxHX*gA1)nkum(e(2*Pjur|Dv8YUAz@2h z@AKAA)G=aaB4PbATh>=#^L^-`kxwZ5V|i&P?UxXdNn4lpBY)M)B)!fC5hUo9xtewD zm0@lh?V1>`%aSW}58rLOE3Hl|PLdV&_TjQMk1~3Lrx^Fp^K1j}W^MD;$zvVB3H{In z*zY8h?(0^^MqzT6^sicAyEhAPUP58^t_>1%%n?up1H21M>mPmGQDo2G3`kArJUokT zC!-#ZD!VFB8lDSgtt+n3yp1nH(d>p2s4vt)Y{!6v&L6P$IDCDs%Qr}S1m{tJLbp!*w-lLA; zBqSS46EeQQ;y(X=u%$Svww~q$^rx;LW2?cz9{w|tpGJhb(J5I=%3-8Qk_P6y`r(#q zqdDTWttSxQl4w56O1WdxH=0R4-8K*;zm6a3MljLmT=x29w5lzgJeaIY11G=)hljyx-y;bvi(~Zt>S*i1 z>hux!AZVw6^Le6uP^H;ns9((>cfx^V0bnqh`Zg(eKinNACE=87;U$Y!_0_o7U zVlA5K-|@Q)IuII0e-3!-*9k!qa!lw2g%~iH^-Urz47gMU;0zOFOw05$r+3+`4shZH zy%A4Aa*8PE9^&v0a}89XiO%wSxOLJxRTQW1=EID@LX@<9CZ^m$Pi}K?Y`q7eCPKPS z$fBqH+7(b|ZJtLGv&9-H9J6CTD*in*>%4TZ5!4Ikua?eO^?nJAU6XYtmH-;WJ*#u8 zW3Zm2mGyiGdY8-*c?+}K?0hkmwCtFzp`=NmW<@9=&`g=aQrCq^UHYxfx8wF%%q@zN z`e~Pwu%F-YZ8iC{4m}K{X=^0S4vNfXU}@{9tO0j-|LhfwZ9CtoyT)%qo{x$54Xty< z5Z8Z?l6?ExbCg^xl4R}gCxy_P@!TlmpubX-|Q-$q%K zH50|$w!|uxB#U{0be^`qjq9Vypa@FzMaa$<6R+2K{GT<|F2WD?i_08xr-v%VM{K=E zwAoQ#b}X2wv)LzXS}5~21c~2uZe!3uIN=+0=d|csbr0)5lQ(sJwdPlBL8tuf>K$HY zreN^brpi$Vpu_oo4F^a6N#`cZ%F=pT`C=VhClXN!23ms33A$YfmPB)aTYy=WM!D{M zO_P*DxTea)cQlL0ndC5dw;3H0PhRU^iDJ`V74#d&FlGNJ^eDrRnYKRdQ%{OueEPf- zUZc~YoFX-d>gjt;CX?qAdq6)RKfQI)#X(P(NYzAssNjfflw`R)5=p^4k{cD>YEyY- zp4a?ch3O^DT`Xn9BEOX?BQMX5!z>b z*$Z=wFBt^Z-|b8Qz(uqA}w zPLRe7&to(qU-(_)C=|FFc1vcj$eDay1%M6TMdZNgZQq~o#IfQt@!KpjGLmwfL+bb) zvW~iASTUY-3GSfNTNGGV?}=r~VhFySO=wOnj-^`%W0HAT$gZ^~8R-yPvNh8nG417$ zFBxj(|2zcu2pA4-3$w;&L*-CH)r=Th~sj%^e+B#vdWc(T)UTz)OzaKHNFd* z>Og+E;LmcXFyLq6scp$E?N5JtsY!%+yUj}zfvOq-57Uo`$RuJFh3kUn>RgTVBXGPKdQHm+8eaBqK%G6@%zCduY~*WU zQFv)TkR6){eIXRWJ=$pwv5a*H`U3PbVWEtMn~XY7U<#%{I-(hZRy99m5$cn@@qg7( z4#uSUM#8TqUbUvfS3}PfO1szxCa~uY~h4@~b_d$qe zjJogm(nlo-083r0hwE{HQeZITN5Ss*ei%*UT6YnxeYVCiWA+K?c^uSeq|uQ``My-6 zHHi(@lQXQ&=CqWT;?-H5| zoE1#uO=gnX)hL~b$nmYLC!Vak=1}l)GrhhcZCeiJ%s6!GF1}b>G@g-yPN!1z>>y$K zt1e0jl{Q<>;`y%*`pK5pSW=#ugIfIxA6pM;a9jMMd%Gm*dsQPdzEY1|S{szAralGT zi)jnEejeRxFb*DvQ2oJB{}M0~@>g=tq0gD>8f(Iq=L7HCyCS*jk1jzy?p29RSul03 z4+}Xi$FzFZ5sV&ixrzcC}~-B8Q5L7)XOZzV0=Z{dL$ZQG6e9Gk3pzp{DuoK1matJShezW8SXY z+q-PuCB2?2rpdQzqNeY-8TWk#QZ>kpsAKF_U{_@ zqh|kDveiYq0WHhD;}shHTfkoAtfJY|h{v=Ben=@&dhNs(XL#ss)AYZhDDh9rEiZB! z!X6{>I^{P*E+*ZR3e0)RM$6rPJ}f~ig93IT0*G?|^sy|8D6wJ3G6Fro=+CbsF~y}H z%_%=WV3QvzZcKV2k32d6Wm&3>tNCbESF%;V!8N=c?Y(2zoYWrAFF4!xqNiGa5(HBZ z_pO~=WC4ZMM-4n=y6SvCGT?rfNAGXAVp~NkRE}reH9P?dtPmE>TYEot^-=^XohFcp)gFtO>Nd|wWLAouW?soa-y!IJQ zk5vzL&VbzYf#`UQ2`noFF8W11ZWWz^wY+_AdT-nZ;!zi69W`x=Lmo$6HssE*nRpg& zuNc!nO<#1HfcNh&UnbPM9PQt`wT`giwIm*>QVT)DOtuIJtTu$^T4 zSy5t9%a$-36i&twdXDtHb{^GJHW?WX-9GgyLf-qQKnC{9QrvLD}o!Woz7zn}0rAY3)G4w!5Noc2gp4Hj)ZtUTqHsnqZ zwG}Js5Xd>04-*eLRh0k9M15qk*OJ0PD`HIAF%$O+7W2`oSRua~kB%4YraA zI4o*>u(39wBE+35$tH^HKa9%1Va|!(L~fH8Cov^v;=`5Jhz*NJtIu7S+uVNiH<4*P zxDJc+oaJ;O=+868e}m_@R1-1cQ&n?a^U*q}AumiDSHJ*oL52}xE08tb<;_PxB{~hxwc(PJHiW{<8P-B{{6dTQ43y|cr-qd=?jUTanA#>H9s{-ztA&B44 zSv{*p{T?Fbr&m_vb9yR6owWlJ$X_Ga9Q_BvdZ`~Cr+QYpOj!?1CN?4}k+LR}y~$U; z^jLSpwB{G>>)?^GJG*jFp?@Mln=q+HJ$2R^8;Fq@f-ZRMMzE%?BvP7qXiLwbb%OfW ztjoz`AJ4eaxY4qvcV61SyV&n!vO zE?G$_6b#w(Goz%UOWi&{bJX9hgnyrtyQ!!7H1}hPb2Mu2<8&gV?}v3In`Jf|WTepd z3`5H8hBKa&_d%o@7i1Kl`i)Gy2Ow~j-_W;EWv7)$>Ouw}DBH*IV1k9?F#aTBYe zW^5{&@I2gf2;@AAB{`ZY(q&pjLr;(MJR<-2cEw&ej;EKI&~Re3BalpoLFR!(vVv?> ze&`$e%=k;jYvdqqT$sJ$@oIMKuL2V(DA$^i$V4``g{WwUUtqFTlPA@Gq&xFNzI}dw zuHxv6%28j`g={P>ghb&24%SKx+%jOc`W@F2sIf!n4G9>)n$wtq>*&6eY0aucPoiKF zu4rt;4K3|buY+f_n88rJ@Gn-azt^yAJ;dXO&vwyA>4na>pOYGd*!?;7d#M$VqW9A? zA57o(TeZ_~7lRhKf+8sE^BT2Hy%7eLJF10LM&#J6!QhHx_jDwj}j%ASwo0Xv%R%PuNbkzl2%FJY{k* zu8B@bx7*t29uUNkU+I`Eq#Kk>Ll{93!hrSLDvnhsI!`lluAp6hoi16| zl}6tCwAC>tmPUWmT7hd%7+L?uMEzZ@{zN#?$thOKQD^C*)5zB1Zc3M?k%X;^O1tUZ z9a|2p^CgPff*>kSM4LOYg8)b!4_zO>f0LOWJMGRJxMnow(sA&(I98BK+M$AEfu{ue z{wpF^Lw8pEkWeU7)J3+!1TbDJTLF3p8THg~sC--)YEWbwXkHxx{$kBioBVim7Pn7M zW;M;S0+&gQI_!rK?C6|T99uxcej=H?D4K7B7cAAGORa&Xd0RAe(`a%hivjDR^M!ov zg-m;GtH;S?1$Ke-Hv?bX{WZ0@ZM=pN*98pu%E?3bMK9VGEq0q_=A)7rZwEmi*!Kh5 zE<#y8+aoT!7*JplQ}(6AsrT3l&5B$C(Y0PzR=aec1*W7+KBEYMIE&Wt07Xj>xu8t~ zJFZpH;r9X364}$cnW;*RReh`152iLa`yJIsCR*5C9hxabG3}~2+uJ`-SVzFLh~KWF zk@NgS#lA({1iaY%&d^zP$FyD|aV8h<1l5%^!pj)p=0AmgC{#Zj`7o*gx#a@Na_#-@ zp#x3kHoM+Pe%gJRepD#^P&h%J#OCgN{%9y=kmqu`f-Y@mPO-c-?(xK^NG{7#K&tH{ z>uAz`zxua<52FKLFzI@pcU$yjgbty*MQol zbLM2d+^*CNp$B;Kwwb@2YOw|H8;q|+e=LlBEvCc@x9LCwXA1c&>F9q^Rs2p zA|*?p!GwZen!jE?eYD3LnP{f1Jg-szJUg@2;IG+D#bSOvrM`=G>SU(8Z>g9xztbjW zn!1YQzGOJc^71@{GD_}^Xxd8n_1$gr1YOtWGSg|sb>N>1J8e9`Ryt|ORV+A4bqD#O zLf=Md%|?j5^O#CqIH({H31Vd_o5XW{e@Vapw9(t9$6ebvlaSTImyzJ$g5czCbMv3W zD*f5F5$HXYSsX^88o5QzProLwEre=voa`MoWrG}4tbmt#xn=O!sO@%JaXDE`31V z^(?(`p>jdcDkuUjcV6>R8Z9590_GGXkp0k=8O;>Sy{B!d288<@!#C(-u_;SnnQdX~ zn(>bqLtRg|Xmu*XRz)ESB$DejFtr0{gd78XG>a`Q%bDXCyqYfx=zmzdCR8w=`t6(P zPs+oX;2X@@YOUfz3%CPw(o-m&nc!rSvO`?Wsep8TWK7mG z-~s`rB7yE!X|RJ0o7f+g>%d2o^+KxI1X<2d=yiLhKT?<`gDt?&`18pA<=6zj*(yQV zkXIH($!}Gy5@_M#As0hSw`~KwMxwz}a38jzXQbhU$Do97Dy98y?*rW5fE5rqQP6;L z$r+QT3M_|)hSp*tPmq>n{jCA#H;^b}?kTqRx|qn8aSr71>Y5@c%F>9-s0ey+7xC(d zE$cIdl1Ty1TjfmV^2;BJxy}s^_cqvbWy2tGb};uAqX()~mdi>h3xqlUamN1xuMqe} z6phSe{!8aHQ?b8BCQ4g6aL#NZAhRGo zOOcwH3rGaJVF0f-1zK^0O8#d^87e>o3Q1-QSk7U-S4{C%_pD#9871BaraDawcOn%R zRISjN`4yX1>1^t6df3*T=ql@XzK?ZS-am?*oO%9780oWt52Qqs)ofm&UYEOo6`TpSah#>IdkVjFDj6?c4TjuM56 z`h$Dl_Vej{w(1}Dgh?c)B5ycK9R`TXG^N^N)tUPhWvB3gI|9(vi2Nu! z5#6~>*cL!V<4OSGZk23w*=wLXeYv$4Y3%^Mn2@=-NM>sLBimxwA@i38$br^d(@taO z4Jd&A3GY_k()`U9rhp9IL#VCUQatef&U_^z^L1MFgci`$tD+x?qSAHF2FYS+%Ix*} z_&)q#Zzxxk)neU2S`xe5O;~U!6)ooxsJ|a$0>?ZDhxZ$=!=XaGr5ex#+%@79&xr7& zP*ScMBnmrMWYOyQ7+h5ENrWKdks`&>(M}{^HRBbaWK(1{!31OD>i-O(9)pSLdxzHJ>aD?@ys|eKvcI9Sl{# z;c$29XXwmZ25KWqRX8BC$2D*WqtTL?&L(+~&GUTgoWa`y?Aq!{{@GRj_Ys1az({%9 zFXiH5DmL+SojjS5ww;cR1-c(+yOoYZ_3bqX1i$yri6K&c05Z;4(8L>+0DXQD6nI;F z-!5;V01gQ3uEED7i+RgaR+NkXxft5tvOhjHXwThVE|dh?>XVE8CsJU`Q}nQ2Hu5qN zeo}s9b$Z5VLTB?r- zq9xtpW!mbF2wL29Pjudtk~ziUYeqL;nrGdQlA_F474HbL2Z4s*80mSn%r9Tpf{bT0 zBaCm$CFi{KN=&y1N3XG))Rmu=r8HrObecflVj@J8eBf1-~}O( zp+b!Vs-{d&_k9ETT?6Cjp*TQUpHSdxyot&Ds`AxLDWyTY%*XgIvHiEy{|k;lA|%jp zk3?7xeaU~>9s{O*dOJ!qHzRK$`ATwCtTBzA9}pHIkh3(947P+D>J9I+A`Zj!UIJK^ zL%km>0$fdMbt`#WtONFybTCspQNr@7G+2OF;Y9Xo+CEm@zQ$2f< zDK9c?Jw)U=;d@(32L5 znapK&7W)DJ)`J)Ke!@!dYaK41fVIq>Tr~zMyV#jNK0xOHrl@|TKO?FXGPX3kkYh5> z*&!@pxq-7j zM&jF#N-+Ns0#N|BF$4VW<)o8p)dV4fo_p$8XhW5|#&G&CW z!KFdzN(L(Q)~wQLPRWUdj*5i-fj$c!;{=4&k2f4a=mNG4kbh}r5W)233}(f6=zS_F z{yE{dW23*X94!tiXH_u~EEdF%iB<)0{lAlZ}#Cwd3WVcyeJ zKfp!KMQN`ocdvXeDr03P0&PSq;8?J~TQP7{*5)KH795vf0>KY#zo3G!X!Hrx|G%*R z^OxO~K?34IfI(~J5((-5K-YJ|IygGkfdRsRvBN>fIk}?I@)dCn#c?*^ zUL)3zF`>$U((zPF`?@$CdQ%n2P#hooq8;5f>++*rs36?L&@-l1YD;jgwxyMrzF<8nGFCZ3iQG#&qMbV1y zZP37LYmTW!n^r8uf$^!BZV4B$Ne|cNu!Rn8MJrBPDJe{))}u=dUI!|5HVJZ=o zm(+{D{C2Xp!o9LO;R6U?0>VcK54q5e*EJF#D?@mn$w?W%KS|8W#tst4MSqL0ML4^? zc3BCIiuD1zDU8Xy%{1OY;O}w_1tPGcQCIY-Nf3oLoAm{QJKBScBWcrgm5mbx{%HBk z+1EveWnnXg{no8pKrZqcbt{cZ1}nW0;Z=2ESh$^=%V6uvfPkIiKNV;GOUY4u3=ys? z2uSQCKciOGY2=|axytVfG)cz1C>^74vfRsNcs|AQQW!h8E^P%lo4+sCeLl!ls5e@* zaqevc9d3KfW3oXoU{{2-2(UV)THlrx8Go9rL*)P-hd&lPLoW!W}uNxnBwhzEmQmHVRJC-h3C4~aqSWCfb(;>uB=4b%t z3_sqwVdKRyOHw+=a(KzfoTBKAyW2OPH=T8THY?L6PT`*yC?+vqPCeE6BjqALyDoR6 zZ;7^oL4=n$5QHqUV%O=Q8Q`?(1rtikBx!!sL2Oe$(_8@}kCvksTKR)=im~yTd{fP$ z7Tr_WK@U0-CiAmLIJ>z$hnTYh&s`CZ=laiGP}CKir?Jyj+(wxvv?59;nFnuu(z-!W zX{Wz@BEFB(YM$Y0#*pDef}Uy}*&{PL4(o81cJJ1mgH;9sF|PpPavXx%b&rbO+1!J1 zbu=EIn})Q8!3jVqE`K~JR4F4;h=8i8GUToXpgz#mo`F3EeHQ(PM~h|#9X0Z(|05Sd zTGu$j<&{^SVr6;BV1$}d0diR0omx#q0C~-qQ`=q^$>p#Jq+tzR=;^=tX5VqNLK${p ziF;9KEqkIlyj(k`_I4$~6>PNRV(NJpUG5}#jrLdM`QM80KPq?>TA-~e%-_qyWji(u zH??N<`If8uSQ;EfB-xeFocg;`=F_Mdn`Of)he3q(-QieM2LFgHp*Uqb@>rqqo#`GxNI4yJO zHLsnxNYi`DO}VUd6>m!>lcNzCpYRIXv+{|Ra6k@5#*I1sQFl}Gf^>&~tcIH!O>sBP zktRAb2Nj%WjqXrC^Kp`rP`Oc zeQ&qFa@EW>6X-LBFegHM@tdS5KMJ$l!5Sa_VbZdsxyY_K7MYK36d9hPaVo{*4@jcY z6w4&Q%>W_LiM4yEp{voz$wnsHfykwpUJ#XJW+Cm{xkxC2u+tFFmCYCb_r9S4mNA}+ zkOOj@D!kfNOT00%sk3IAXGKO{i4y;4izN|VtIaf?87B*WpG-z?C>Ts^_eTGbw37XH zF@s@CF6z4rq#u6p=-*CWW^;t%HL+8 z+8PiH9>^At z@w%OIC{%gQUR(S^9*4#iAXx#P)scuD5$S9H+=ZVB0z7oSYqOYNCs5Tru6u34-?4Fu z^ms{jNnl-4eOp>jCpvdlAHL&0A{L&BbCUPm+vcK2GP<2~#ql?P8mekry8`Jv zq&(E`CV7mcuYxbLFA8WT^N zjJs3LhDjb360$?r_8qWNW}7J$4$HHaX|mw)rw1NabF}p8-vK>MT|>=|7oXK(qir*L zE#^3Noa7&~uu^*sq0xRnvgvkD!j)!mR>o)Y)PnBLh~96_rYRXGDmKla13}14Jn@M>`3l?dZmUJw z7ijHUuj`gXUgib{AqFCfnqnKiY%Z^_9Wu6)0u$nU3(9$)mqBUw6fNr%znZT5!Y^7y zwlj-KN<;vXc_WR&U!=du7I*++cuFT9j`d=Z{*8AEAcQ7@1xOsqe4oT7sds^fvnhuBTiI-}P zYZLn%4LO$-{@3zL{uQ?I44i6v2{p1e3XX@jn|wAoqCyk1DRQ!Lnqy?^<&EUpyij(+ z8;QcZjlB)`pb~L#I>V6}OfZcV@t!xXPd{qQsHEe~p3V~<n za&Z8BWyKlcX7^Sd?I`Ckf#kR@w;|w^vC|{WBLXjqVJZvbenjZZ2j2`FTx|6SyppOL z*be_FI}*hs*jFs_oNCbf)rG>Fx(GI|=~Q5%^%JoL@|O$7<0toQ%9&&Y>#fyW ze>>&Pck%J(Qo2(R6S6KSOa}dPL=?1fg2n8TDDia_)6;DRwPcDLagUV}J#T!^rn!f# zh_IDQn2646E)^_i1-h72*7LaKZ9co0+s*Oaskll9aCz8kHa0geux=>FckpMsIY28M zTuAOd36otd5xq7HXqJ*ipCTkHdtl2qFq0lzpRjX8>2Q~MbfzB>6Kp@-M7131xZwFU zK17ESN7?8SBOh{QDE~PwY$oH>Z|UjThwIDUGgjtv^O$R_|M{}mm&Cf+QE6pFst>dA z-J8~=-rarkyl{0S&h*8L@?rbj0bZuNHh_~+`1`xXvQwITj4ynR5NTO!7sF zbY+z?AHy=M+y0Sc{e^d+`Zh;)xoX@RvGjrOaEv#WkowdQq(#99PuPcFF4Ei<;QMC* zF4_QMKuzXQ=y1mq2OC}0#ezn<4`?!=`|eozX~CjMf%%WgsSYv;0MT6?jZ zcM|-WKy!tljyjBMq{Pv2zzvNK{uafcXN)m}oaLm4;;~#j|65rnFWdv~WAD5%tB`kn zCPOfEqXiIjm3vTgn2+vo+Nue7w7}VZ&YT|eO_P*-{&{B3^2FO#%X`_4I--VaBc#NC+7;^#W8gHX$pa!?04EKvW+z&!e*?yMMRRpL7`I3BUBe z=Cx$v>12!Nx@W*%KxE2DxWP#uEEX*h!~hN^LX_I#d|s_Ly1;q>P{)F z>R7fnRell!@^Ai`Nurmm#Q%1TK<)%2XoD)kik#FUGe{5H0-K|*RhCru_mavDgYVq@EER_%9=T$C&Pxm!k=?IA* zI?wkhEPa;=w@Uhs0^c1{4d{e#>!A)KgRGy&0N#Z-w=&L$HMlfnIHl+r>(h)x+z8j( z;Kyv{zI#?xMk30oZu+S~&Ya#J|33E4>=xHZT=!l5*AD$@s@cOjT$!q^D78mrHj+$s zJidLHr|LaRcgkfbW+a;r%B4^% zh+ERqI2XZa5ln$9`|+eH=zFNv*NrZnpq`5oKJtM3tFRy#cV!&C_B6GLRJgS z2Yq65fJ7&lD;Y6YmKcF$V+mDAyGVaERK%GDz8O(D0Qr^(@Oa! zPoO)z+De`_5^oWVS1C~#SwU_3JpJs9aL7*gVg71%Abnb5}p4kpT+>4(~ z8SJsUjxI`SN(+q=`R;$oscK%B9Zu*jONdMQZI>kIHm1+WO%Dhn3|TI^TPmMgjoEy4 zQDj^yOLRQ%amJh~R8;3|CyM0^KEc5SXVr=e7OZ>s||RpYT~r^J$~G zef_F~9Aur%G-2M&vSBB8sEa4>`}{1hq}NaJc>CuBSKM~PCJ)2n!J?4%rAlux^D5A< zosfACwNA{u^lfGfLvZhV5xmZU#iSF+V z2ZxG?%-2(mi$X51Tx9qA?i81E0l|-hiT;55K@-@}CSnJKI2z)9NUW$uAZzwiv)AW6o%IjlOqke4>8Ba|=F?r}B7U478 zx;}+`}Jysj~E%;r62W)ZI!jw8#bAfI1-X{%qYtQaDrnn zL~AA{VSMm00essgAbNNqo&Q{8zv4xWbj~VpIPEI_uAe8g%pO8&x!(b=H3&FZ(A7ha zS4%C1vVVZm*L5loeBNhlu(`d>GxuAW%~qxiv<(gpXJD+FWh(e|>LoG1k8&JxIem=y z4>a-rKlhLjp{{JDtv_q*_0s%`q8vKfd>myusO(ywCrd3&O zOZCEq{Te23GC4^^x~r3~Y*FH8Hrza`ys-4W_h&@9$Hv`M*1hk4^CU7$S}8Bxf!y}GUV4@cyh^(R@Tt1)-%^5|^GGO^F0ap7Y_vFX zu>MgEk#j(X7VR+o2G#C;%}FUxAa5yW`=f8EKPu8OL7$nhvmgD81dns^^kt3YqP=() zdyRdpkP7ZoXTmZ0#W~gOGFW_ag8#W!w{77jWD|XmWX-dd~Ad{I^nVa{@7YUG8=GsV;#z=!$aO3Hx&P4<`&0>+b zSgk1shXOtVKqrT7*PgqP4}##EWf81&74OBYLcags53a)$FrAb5Cr%b(5Su*Fda zwMN~t&CzZSnv$rFKR34`C#oD`Vg-GGZf-NY5CN{o5YLW?{`5P{wWOHSL=5KB%_W!x z`11#g_Avf%OO_5_bV3+ar#dB}uM_{yZ}TpvPH`rn-HAbvKK_YNY`Mq9qXqTvo{Fh~ z!(nJZBV6m^&$x$uJ;gNU{fFC=NLoLc>t_8!`<2$cFu?hZ&YvC4CJlt4F@pV4F6v)c z(%Ji?U1K@fAddemRAVV;BGXQ-uz<)Jqr}$v7StvS(zn~dx`xLjQH63o-`G|nN+vPx z9{0CD2n8;M^Qlf>{9mr!t1f`eeIuaCyF6QnqNZ5gYAoXomJ=*&_MF9a;O({~n@dB! zM0UAujOgNOWK-XVR)lo++{fk9c_$*s>x;_P66(H1;r*&J>YAs8AGXMjKswR^*njg( zYsr8q8%9G&)B5941X4+i81)H;eE=N%QV|$F;6M|!-+DWIe%E&haoBut%k`h)>b}!I zeEbP3A2iyLUrk}?vzs8Rs(MetT|Do1wYqSK>uuz`Sm3~c`-_UsyGK+!Vw7KU5gUdKPa8gpPR28)RAkX#k#+l2B6I@+lS@4}X&0nc$) zDeUdqPuCk@4fjiJfsg2ySu+N$9xEDh7D>%v-hWt-1R_{|e(ixIu~zLwNn4jaj!v=n z|HBAJ388UQX~eBrxU=_j1LVyjAn6V|mJp~Z;UlcJQw7TOl5DL3ASspwe9bg2e*Kt1 zU6`VCP`TEwDr5rKzB~1)GmV}x8`Sk`&MnotV&(Ku7A^GtU*v%(J2YQzLy?bUv46^$ zich3fdE-yDVE?;};El-YO;-wWi(PXr*~o&3OFhSAN#~*&)Y)a3!Nlm^fzqL(sG**u z$;k`L5&-!$fzjfm3D<)Tcr8Si8VQMs*S^hGy<#8K(#=#!2e8D#6#6Q0?b)-x!T6hP zj=PcO(`y7#TZGZ8hce=k`A_5`FRI0I>1r%U0= zmQK;la~+l6Pexx`c-CYu+|Z*R58B2z7bG(t=fXS7_<2i^KFh|M?29E|z;!Ix#-ioV zeBHGR!_#ssqjctCA#AcWa?aunpgw<7&l^gO z>jZkwKWfwF#2J?lpsZ3Pe<7pJH?xvSF7J8{>5&j@-)Z3QJWGM80#Q6$N_UJs0(g$I zW>*e}RmVW6Ns1yH-UiAs4oTg#@6K$~PCUF(ro@bCB3)kX2@W0DIHT@BCtbiA&f5TOWJ7xqfRE5HOW4uZ9((u9svsWA?p} z$Tr^q&gGEbgA#X4NsMIIxN`N%d{{u}TdA!vR{lmdObG13nAMb9nh0kO`mQzQMLjI8 zY>N+SnaZ>wOJEsPP7$D+Bu?T&?nu))PDFa$UB)V+g zI3a1wWo(S&6syS=tV(*R|1S$bT? zFUtxCSm6;14=S5`2VRMfiUEv>G=Nq`gg~QC+tihgZVAIK^v`8KT%{AP`~T=^)p2HL zuTGA{5v?mSP#x!Jv=*EPTgVIjQB@5RRV>!dp$&YiNYvz5(j$!|*y{C&-LW`MpQ`1c zPXIhy{^bw61^&iDbu3G`It7_JZ*FVHd`8dT-16IP7jwQsp2Fi5>=>K(#(Q+$6};H4 zr2h(^&K5_w1+;7l-ha%Vj%kDyC)T9^n{LX2Yh#W``DVOGa``T^fJ8K$e$n8&3W>p} zj^6nVpWAcaNzJbruc?KkKg34}BSJ5^(Rg`9RISo#F$?YfPO_g;0=$aM2`ALLg=d`) z-ZqY8^te;A^Ovc|OU38!G0NgFN3P?a>$@L^CJPT&aQM(TPs#$CDAhhwUKPQ8cSc3; z=@uQNayc$GWI1Vam4wL@KJxQ>8GIVueVb z%;^gZ>E!sE!_ss0S(Nk?4|~#EB~Smv3-k5=B%Lh!Q%GMTNb#er{K<>tbaW5dN;7*5 z3hkE@s|^R!z*N;yWD(m`CGc}30tVmNPy>$XUN8{EoZ}%lOV3!zEUhkM6B0hwlvn8- zzy^;0_*SrkAJs7X3ycB}+S~U==gjGM^n#?sdxOE&ZaCK-KK|GAY>QjXwVr*}=}n5um^rmEgCOH67fTjpSB z{tTr{uGsefS++goX@Grq-ryDP8SCm0T}Ac=9>ugC---idPL>H@ho&9%t)2&z^L0s_ zc|4y9>-?0TG*990r>dTgSKkd$~=b8nPgA&q~7!Zn}^n)<}Irf z_QsW09#YT1#|?y>(NZ zzdwD^$|E!C9ZtGdTJMz`$Ekrm?2Q0=)#?W-Zn=iVxc`db8`M1s$r0-ZiVm;KOr9fs z7;uf$#V{w)GIC5fJD6UXHWw6Ntd2|nVzh@K_Vv{?{#rk;BB0~!sOu`wv^^6*+2QZv zzOaU-AFxxe(e!)IAHiIWOkWLy`qHL z2sT7g9tC9L=v6Ju32&xwacEo}4M?qXJA#X1m-94CA77?bCA?|K=NRbCFK?sHyh0xP zjJu$E67E3uGS%pCC{5^j&SIWxj)PC}Fvxj2^ha#SGD=YS%tCmFyWU@@I;-6%!|!xs z+B1c3$#K}3&4Q&GYs*iQ3I2R1MLo(5(2W+sU5 zBIf#FH>Y*1S#~4b{KCM=wXfP^@DIA{sfSqK`YhpqxtbMl*z3Ngx!6vh)R1*Q#qlAf zI(vqfBT!=JzS5W14dSOJT3_}xQcZgaEDR?yhltgk@%(k4bW;eFLkZY!^EJ*L2nCwZ z1Z*_O1g(*m5pnw?Vzr}4a#(l%8Y>1IX(p4P^qXv6z~J-$Uk!@aae!*yfT zLrEXp8Pm@k_8Vjuf0AL8{I(FMwCLU>gguKkk8{=k0mZux7RxX9A8Aayi81RE37*MN zc*kO5X%o@{10fllh>eZH1!XIk_#tuEP7)>A4d||$E6E*o@xSQ`HbEIF9*MtD42x<7 z?^f}u5adM#+R_F%VB^7!UGd-SL@&D()|k_k`18ph`+XG;(Yz0}J>#wL$6m-jdVnUI zD)Iy07Xh8I@3rk_{u(3!in1s9y6u#WB8$&7vpcy14$2EslSMkpKv|7YZ#r!A>bi~h z*~64AFkhoXWnUh`s@UsK{S!D()DE?-+^wc7$h=bhN%$qxKWK?j*9`!31qz_>89OS z#54)&4r0!bhF;OyE3_y8J?I$!{ z*%g{&Po&A+eT%HFt|D&s#J}cRysmHzRL{ZFG0x$s?R1r`v~hkx-2Ke$s9iyGBPM;n zMlwCG$2s-cdGm&U(L28L^~C3P?`%=Ktf>Gd8a2)k1A`o1m}Sbkx7*)e!JIt;e7G&VGS)e`1# z>V`we=tG3SQZ_u~J{(abUKx?58l*UIq>w839PK(<=cwb9R=w#sM!{_&BZ;6M*k$KW zecB+u=5OiZUmVCU)cj_2Ys@EaYvWa*pqZpV+}IKZA^d#^AlUB7SW zU!Nnc?3w}m;%^9F0VJt7kFAdvTTu6XC1X}y9g(v$XVFDbOL0M{*Ys&m6`gdOelJ@C zSNkwm6#_h&9ew`qJML8uzTuQ;7sN|OUk&X^X5M@PYrxMTEgV-Y-rQeEBo$9OWW>$w zi8Uln^KKNrWQja2*;X-&zn(9(wFNIn_bZW6knHR~oarsWS7KFb=k)C_EVcCL!1DPA8P^nf3X&d% z`^>v?x`cikJDw<*<@~En<5Yq!i=vH}+-6@4T<68_?K02diBsrrEyGEFx{thit$ccT zS>8S8s{gmaGt~e0uX%H(M<4C;rZ%e0~BttQ(kJ* z7tz`uy!CDYF?`E1GTLG_YZngh=x@*dKzdi_Ii?M=14QEc77B_cc0?U*It8mhVKNhGE)lpqh%+qOWmBWe;Ov5t8~ux~ zh7mw%a`L+=EU(^4cm&23Sc>Ogo{SbTwe|3`mXuMMNS?TYGe3SP@`q^1-D?pW+c#Q5U`1^Tl8Sk1W#ANK9zH|6HCRGL4BQt)|hEKh{>-pZWn)E~6 z`$&er*69oPGIsC;`(`o?w&*lz^*cm-E;{~C9+C7SjPFEf3SULme<(T8OU&{F&t-^Y z`tX3Ik#znT*(>c?%^K~8s4apvXBEhZ<8XRPhPp-L?pUAV7klks6^OB@D{d-}l33oR zFtUZSXoK+OQW)Rs{0`L?(X{Sgb?<-E9)?AUHXrGyje9|^YLwIXP5enfiU>I@KLQ_s zc$X8jZ+asi_&gsuNXGe~^Ng;Ygv~vHoUrZhwPpyHyT^+cMjKpQ^H#^k{&+2o(+PdSxqC_C~HN3=))*mCpx! zLFYei8A6A3Og^fme<@v(;V-8vlye444S;Y|)pwuBrNl6#fwv44lDQ1y4RNEW3_nC< z4HAN#cc13chubNogq^?nLfE5$b-u!1KTgA%YLed8wOR@4_+!?Ngs~T-BCP<^qUUay zkIt=O#&A%7zdBcYCY~5%t-L5`@RsM(KZ!|)!oJ?$+Qw+ze@I#g%v&>!vZF5}<)&Ir ze3t;G_5Mg=9#<=Wk3(bX1KRcJL_bGlHEE&7JiwF*5LYU*bFil?q!;69(og z<{tKO-u84nrxp{4d{@b{Oz2VL6dKre!X}yu6TamG!gfD+)XrZ?cG0F_sI~^99CD9- z!?-5bC=kABaSi`e*52m~2yU@oBeZ#3h6Cf83QE66?UqiE?QUL`OcvI4SzglKMo&n4 zr~{_&{IZ>6-mRBvvG2v9QAH`i!`=gwuLSxz5fDujsWW+uMW!I{9L1q?{bXpN^dNqV?DWAef3x6i9o%$qQj}4c zOugom?sCuG$^hewKg1D&NrdVA!{qb*%fC$Fa%bpCaX-o6$462>CQ`v6RrC%CAC>R> zso}h^uWF~raSZD;Jlz^}JGno?_Ae36jsPOuBBNtijzh@r8($*|Ero>rrd5}~lMZJr ziKiT@@$gi2JKw<55vL%JC<295GCt`0#~I8!MXev-vE`!fF7Sy-{A=auAUNL1y9}{3 zAu<=5B*c!kBWR{9hk-O))sUkKNgS$unf^l5CjEURhjK|dmz11(Q$M5g-z&relL+y0Ku4XzFkncAJ2S_fSdrm< z66gdv?GpPOBFJdFi&@Zr?9*?{T0$qm&o!mB9IfjUVsw>8)a!7OD9)3x$c_i{c zzUnRhwL+V9xFI1sm%~um9OMdib!8uZFIR$a10oy1U|y2yw}^JDOugwP5>)9-If<+7@fgFKl-&-%q%#oW%vC}R%z z1Qj)Chl0h%Z;g4^p8vPM{apIFX z8oPE5lgyZY4mUC{&vjnemTO3m`gzs)k)5^B3ebHZ94`vtkmFDS)`#NT6fV~OK{VEf zU$H{vzU3%f5vaOVvQ{X`(fsllk!4#poblbHYM*;$hY!W><8>#VQ1+hxp*QCS&H@Wo z&!I|!dMx?dY9C=Qt^<@0u`;>p9fSBdnYL7Y;p`%7;U}KlYpYcj*AVd9BvXW^lkZzs zu*a#~k!Pv;xH2Z#;{npA8r5K}R(4(b3l<)VkI>e%-d0Il!C359!RbUS!Z9C(1N0*W zmAtik$(Jz(jo-5rx^6l|i(btzboiE9++%0yykoh0v!JAI3=7tC+>bvXsw^y6HP$L6 zFpUrT#32(5gMrpfT?aqzbAt%0c}??0sP0JnNPGulkZwh80>V5l*YMGz^_$kn&Tf6l zkZY)3br;1{%JQhp4Od?zrAg@O@o$eCgS_uv8xOzpSMf%ExJvnz!41~f1gqjLN;ljz z8bVs0nD)B|H4#5^>o3aF#W1Hnk#(2aqNlxk#fPIN@BN#b=ZT>GUBc^+2J?8exYeYA zWme<=Jyo#5YS*e2I2=JeoOqkGSG8FOLc3|JDnS{eVoy?-^z5)dELOi_8djNJl({Jxaa#lrvJYM9?mT1@`4R}BWZT0RdOE9y8=tsOGsL|YIio(%U zJJ+c)fNB?Q@w7VnCyZ1YJ36J#jCd=MNJ07X48nZq>M%F2>h(Nm?V;I8;g1 zK$Y2tBprhkzgZEP)08a=T=0mtk)SDp*|zh6w)% zd%zp%D_KOwn0Th%&?25zoM#-N7)@)t`vF^?VdfyQoVAB>-vZ?;Y~7A58D$gCQ3^;d z@A`DOX*Qe%yNvhT?noM2X9W(3Pxk-`j^(81noM+o=fmIYirGC(pRPsJ-=^d5rQ(FZ zY&j~*tG(mvH|DwJmZ3OJT}ruel0ma}wGK5Hcq)Vg0{-@mZ# zDfREqS5+4c^HzEA8e|+f1HCRvu>fy)0ri!TeUYhPQ+yM5e(6fp$|)Jn7>7Qr zz1ChS$>jHUpHJqm zgnMwM4d+bY60?McNoV-Pvzm*0M#v_4^+-x|_IxW&EUEfwSAfwa&VJ z$$amqpFBeTR72`sY0M=P6!f_LTDzZtb;@ooRkbFQ{(UTc=mWsSuYj2s2x_#H`Iva>^DSC;v4I8yT$2A zJ&GI*>IuBCCDhW=-JRW{)?t9W>N*C%T{OU>6l}M=nCvSc6g<5vS2uUoqWytM z^A8+W6~d#pURF`m^PLrh#WQbv9_53rfn+s+;5z&yz8m-)%|R5;&5Bpd$Qx`ozxd5r z(5bwruex45GGLjVp;nHK2#sd)pSCP8sIP-63Pc^J$2tye;AQIUz;ly#o@%$&I_PFf zSM~{ABH&zD{c924^Z3W9_*bRO)a}eSFvqj%yhY9-YU_}5m&^d%SG<+&5;@zJZ^wZ@ zL>DoSQ(9vJwd*!Hdt$f|$9$%;e*WJZauB4(Ro@Y>^C(tjlg6FRpNZc#*13#@xLgg% ztD%GjIaLaZxzzAiVd%nZx*3@<4&Gc{TA1dY(=>4o+g)*3XcB7rM70FvPdEOhtrHLs zaL?%M6#3xe|6l!9oy^cDTwt{>bK+JCGiHpDC4QZR`%iJ9Q*cNp15T20z0W`lvDr?i zdH~K-D&odl?#&_*yoc4~ZVvpI{oA+$rd3^$)GKSc>XE@FGw}`zI6=~b4!R{Dgd`S$ zf5-e`3bQQBeofB=-gsUXYlTOLI^IGI4c-d5;FLS+SDYFRk}JSMbY}1b6qx(wR)HQM zMdpbjr&|nj4Jo(3sV~~#bf#S;Kdl0&<EcZ;qHHAoWo4sAMx{S94Qhk#4x9XI$05esWo#tN2I2mFFq)8G_=Hm_K?NNA%Ue^!C288t6af*8^aaoD#A0kot$+ zop42Se#a!N@*_8=ld-fA2_`^W?_6h$dVRj;uR47yPvD?ugQQxuRz23}Q^6Yc7Hm&- zhl1qDj%XS?%v6Q;SBi6QbR$q4)?K}QVyt3}q1IhBsu!I$F_(vrDxFHI2jFn^4iXb6 z)jX5B9(`W6AYp2oX3p#$l?j^{D{pxdN8@^n-Tkixo=nrt~=>~m<*cLqY-|3}&QX9#=*$w^VWufZif@noD-PI+mnYn%n3ZQA#vDvn~ zvRm+p9s0w18Be_gf%qsw>CC{SBT9Jv3(CTLoSB6sZVDZ)AzqoqT>tbsVAnANSK%3u z7rxFbzna}-oBIuY3eTVEcG=P4o0FM3`h9;Jh+g{)^|$?atE!!`c?LW`Yk(OQo`w6o9LB#SsI}pyH^txO^_h zzy%MmVlGneMO~y2HEm)7nWmgE43tIug;ifcaqirJUPD^FvT^)Qs-#d^_Jb4Zm zLGiuLmov?*XA9rnh4mtH46~_26-Ps>O}Z)0I$;kcQgQozh?elGj7V6D#d(ja&HNoS zY7^KGA$9_0pkFA6Bn~UdrIv*Vj7p@Nign$lSIoe(x1tPk*|?z4$gGyofBHz>1|uLMk$dBs}T^ z{}5d5f3TXADZ_?R1@xW3Lca34dI1CuZmY?sS&>5Zd1k~&AgumLYv!HqS+hUGhJxvR z{*9R!1lW932bH{oRdRu+Rc2j7zpwt$JV8RmR7R1Ubu5H`j#uk^|C+4&k1LX#0rqKv ze@;|~EgbZNdxC!fPlvZM4s(W~Lh_zmiaY`m!BYIQmu{4_ua&5#?njc2D}wLrmq)A) zOJ}F)?gS1)J(+Ye%UMK=15q7%_PoZ?y4+kaGcwp0bcV9OQf>5aiok!RO)wEG1uZJ{ z>3{px2SoUr^!g;1XCq=_-pb68?o$|IrmkICBVf9h6z}m8qnZ((VJcC~2P(8+uuNmP zgCzXNliv1{z}*tW1D_7Ud=tc)f1Th<*dQOd*TR%B$zo6C46qlc*~|PI(K+2X(3=D} zf4G%y_we==ab5=QA~y)RGd{bguH=r~7g%}>U7Xc0l2wKiV7AB)f|(6M5Vlc1s>)an zfw4pCcjp<`oj0d@uv)N2;By2*MztCueK$YPzFxXN#idFII^zuJySaGkx7JRV9KGKq z1(G9{+0IX!g~7gs(i^WsJWy5RZr|x+CW+CnX#;GdPItS|TXhx$M-$a0y=UyoKcJ^s z!C7(U94Yg$d#ax9EuiM4*E2ZJ6MD5WFirh#Nb&s!QqbHnsgnF_#2os#usb6u@DvOv z<`1bqaA1Q+Fe%l{@$+)VFt4&(V=fr(jpyM>E11G8o|5o|`a(tn(w@?8m_=Dd<7gB5 zN$}U7#W0=2kR#QnCd^^VQW+c`=9 z5k9`Bc;&;#+VwUU|GQkJZkl5sD*v-Zm0b6cKl)f7J|sz}#!Q-SpvxdFlaJdsK;yN4 z?#|DAs44y!{C4_mJLx;2CnaVVOJA-j#~%8Ic;;Bxv!JI$>@(+$*JWe=Bs6!#xbh-{ z7W{eHbOd!QFZsCL{OJ=hFpwHajQ$ZL^cUg%Nl*puZm6DJY78k($$`XB;u7YMBnK^h zpQn{u1Zs7%OWZJ-r@HGx?Nqz45_?MM3evS{eEneuvz)t9*q^|=Vr%i>=wF*NSQ}2O zgR_6p&+7-uv@+sJ1V&TF1Ys8xzE_Km%75P6Tlb4K8)Gs`8($|k%$&_!vfyux(#q36 zH$#(pzZwo_jXyYRi|){^%B0_u>FsaJriv$8&GnmF?)8L=V}o64f7rq3Bl&(I=Icop zlbRMvyWse8F@FAPPP$zkFNhx%<1KJ&eYnyPxm`Y}h-3BTJLQ;#bd2{L#A7H~@hz67 zW8=j<7u-+C?IEJ_bD!A!wOC(lJYQ_&&{VJBp=zv_tZ}NCw6BQ(``zEM zrhx7BW`$c7ilY4{P*B^>17YXMoKS+UvvrsKv|`#GR5QmSFrG;N+3Xn%?Q#v8#=@Pz zr#2=DUD8j|L3Zel^~n5?Og5eoaxd_P5EsRfH=Wu|}1t`w}QUZS(R%>B~!U?$^htv{x}DQ|v%Bo^4L zaCyL}^SCcui}BH{y3dnr&Y?+L9(cEZ{L3_k!_lvL>JMw^FYbW7jZm((IBYwifsgw) zwbvc4x8TfkLH-rbfRxuF&NvZ6Xp6rimrdqEsSnxPBJ))94hF_tOC!#25R+f)X^o#2 zM2H%EPBlYuQLS!yQ3DAnfY(;n_;*yJLBAes?1(A{-ANPLEk!$}Ch6=2m3I2zK)n#; zun~MTVLkeeb8p@E`4$7Zu}58Po|_Jxq3O`kfr{e1yUSct2S6Sv*Ao+_RA*6NdZj4c z`#->#AEFPK$2W~rrmN)FJvh2&k$rv&scp=l=MeLt_l zVhMZ*#rTbAg;vJ;n3ZsyW6QwGS;vMZr7tL%v)bek9kwjt_KN*@RqKM6qD48YJ;OIQ zcw%3wPlr};FTT8Dq(Qoz7WFmN^Q&TEc|0yB)gfhOk>uZt-deegkCJ=joOb=#T!3F~ zpYbx10$)KSDQ9QVsdW<~nacBabXnf1V!EIZTnq1*) z*u@YmuC&|1x6-IXO~*G%(Bb8QWrGX6ppu}0HpB1Re!tu#D4#!7bcOhrF^^PMCxgYO z3RtB2*mEEHvs)jZaoe0C6ZpH=Mk1-e%4~!u8#wp5Kj9!`dI(KjJZ>Sx#O&co9?eg& z4PQJ%Auhfj&+~kF#GCy*Wb;6(3L{N2S7#ifzTl@?noGI-huO;^waeO#=W})ScEHG! zKFd@G0eHktq>@ICO&Y(WH9$tIGh+wL^PAwkrUsesL!-dT*)?ry?2%>19UW|lHtSTR zlDB)s9n>0)ao;>gL62o6hQF#wX!9-tGBPuGL6RZQXCm#`x7xXXW7?@96hi3My*&Ze z;9NFdOfYb;brK4(V85jJAcCL>T@&Wxx#r5ZHoH6zdQ#71=OczzdpmgU^lwb>1pSy*2_qQ)=kln`mODA8F5#TbP!t8RUAQ&M)E*aym$ ztS6*#Kxd1-+_WuJ6ywU#fqTO3?yz(W44bb2RYkRmGcX!vB$?cq9z!KX`9krcoMz}a zZAU%p#z(#Rd^P+>)M54o;xi?wJ0CQhkZjx=0BODsRf6gSk{q%eTtZ?(>wsV&(S;7r zsxcY;MYs$Ptx+HP0G-Yy1ZoVOje^kZ8mA(Qy!%HOjSn>GLOw5qBg;m&0ZRX zW=V&6J5XKyo~651f5=QII8|q5Sa(5quP}tQK}e_bEm8N`PmJ-kPCev6V`EinVzHJl zTjZ5inVj-!c`A^&m!gcXWkO#06HZC%xdo}B7AUH) z6vexwtXk!H!F&n50^WnPg~dbY8&rjUuCw;m-(^~Pb0Lj+ZV`XE7KdoW;g+{QJg!wS zTsm#fXRj?h>~n9)MY#NPgahp_Ts2MeO*lOr&u1mtrW1#dNIOvT!VWC+#j@_Vt z^jJ`VD$OXA2~t=L^=y!S8ZIC)juyqH32gWuqE`_fdKySHnSly;k&?PP9` z<@`aWZT3(nqvr-InMUNBh_4eg+1S^-v;51&+RWdwMLM6Wg!cT0c%nIs?PZBkpoqVt z24jqd?bGs7?2t8X(To`+`HqPo0x6I~F%Se=SJFkHGd6g5c#MhGyfPQKKG_ui23b3} zKfZ3-e)V*(&^vAEpb@eP$hxep&)ltn5o8t!!$gpFE-_yQ^lF}|D)_Z!wNXWowiE;b1py09s)%$c zDu{wK=^X(<3`Gb%5Tqzgigcuk(o2BQ14NMCd!&ZmLK1p_kmQSNoxS(B)_Tu%_W749 zPv9AI%rWnKlsV?KxDYVMbk19mQsL~JfyjV@_|=g8$$j&0$_@>ZeY)}+^Oo2{hNEoUA+HXG_%2Vukjo$XAalF;fc-7w0w` z8qP!jD1jzCFHk+FdO$qZ)TrmL8^b>I=?vIGb~V}hG$4lZI`19MIuH(rOWG$~TO3R9 zA>Q0x`~xPt`^hiR$?FV0ICfY3X{t0z*9=KIIJ(f<(S#$)9!Vf&HzO(}cP(5MO7cO8 zQS3!FkIN8~1`pEARGu%MRgDv*GJj%gLZgO++~*d>kl~00S^Eb3Z6dA}bUW$5k+f5a z-;ccv%C(qt_3PM|sxNu*s|@cbCq_I)-Vl})4D?qnijpSoG&ejosG$`SB#Ifw_w%(t7;dAbjeM;ct@{Nw<-4FOhkYMB! zLV?&PpWu~HDt)WhIEyUMrkpqCNDF&|7x9aMh>uV;ICY*}eh4evXu@LRW;}Q(TMFl0im9Ca?&Su=yz1B zrQN^9-!aj2DstKBnNP&+YzJI>NdtF-e!4&Pc^W&~@7s3lpHSL7!Xx>Cxl~MlVeDh; zOpAnr85z(oCQycAng@T(Kw~0r+Y4iJ7AlIer^;@;>h0SpzA#l;ahBejD}g-#J(_AN z%J1&>nt9)t<3vptjfr1I@uJMDElhGBpH3g$W_s~|!Rvp0w5FwqN_wKvZ;ml?ze(Vv z@4RGhZ}!$iR<_seMR}H0&AEue5c}k<8FyqRw7HM4Ce%wW*K(Tci7}*iO1`EYBbn}$ z$-?t}@6$dg@}i(9g6NJ!I^-%-*xqqhY*C`gC|L zP!d9S6@%9DLH8v&6fMC_x9{Ob)uS?|=K+tpObe{)U)@VdYRT|hNiNQ{-DIG+oU|%D zv+WCRiA3(>TOKTNRUhidO+eCqeN9Z)$wTF@&`-}ZISL(%nPbctuUYzHY#uewzB;?@ z-H}JpPdGV45(*nao=v}ei{-pNbfk)ZdM~bJa+LtS3c5u1$#b4gzJSe)Po|6u&hgqW zsa{R=aqO=P77-?acC9B&MHUqjsD8oLPZYW=XMBsn6wEA6$k(mjc;Pn{S*s^rE@Q}- zs;IJaN)6NIukjIsq@-l|{<~{#%uJcDm(Ek|^%N~#VEuF=Bvv;6*`_Spj8`~e##zB4 zPiU^ZZvY}**{OY+K#!z`KU{05Zyw4pUDXdTeC7ikZ}4oWZ9@tHAnNzxL>_8?Q%u>1 z6@N!BvmSrVZD*^w*fNPXA9I{B=9SG3d}oF|#IY~HKQan<2kM+F9K4X%o8cT7VHzBa zTJWhIcaP4(L|K$KJ7vN;GvXszA~xX1bF>sQ?yY$^U z&G;Q@tp28POj?c#yl+8fhD6k;9V#)mSso_OEJ^WXpp-e6a*Dk#S&(BO>8BWb9aZMc z+2&8%DFB6R(^g}i@n>eFyrS$<#8whz>7I&vgU795m-~I9 zNKL%mQ;tES9lZrDo};kAmOiF?7qNZRUHqFDEd<7wTOL(<(rG6^S7sFNXTZV;Pk~}b zLoAEFD1wPIX~OyEd;lNs8-Hj@#kcvg6;z1>NPQ#UeZn^i8|2v7*(B!Ic@c*hif!ap z{H$=93db>^SGp4GnTP$b+43b)31T%=0GGhfnoJy-xy~=~u!w zqK^qid~Yu&3q_;FQh(IkaA{rLBFrmcH{|4wXO3sIV|^Y^s%Mk0B~#)=8N^Gq&?G+s zx$qfMqs?o^f8DMB1AFUSP%q8*q_RE``)3O+9KXPyo*tt}(4}ObUN&g)>X;OfvrZ*7 zg#LL)F8Wg0BWp*BMoZnc=e*}LvTD{jyoL@B9^Y@xOIeu_pxH>%Pv_;|jlx9p>P0P^ zpJ6%hZ{;JkuWnXeN^L=@V9;vs)U6 zKqcpRBPtASKhtcS%}xF2z?v{tM5@pT()E{le-`=~PPK2P`!U3^#u;E1fVh7q^yCx; zwiymihEn{DofEgzr!;htSGxbGG?L;0%lHZ#a`8#4Fu!JD^|U$^G2|WpN5`RPgo$+d zDj!ZB!oIhrr?bMGEc?JXd4>&ppG_`_?6|||(O7?=jqs88Ny|32i=i4fLla4*rM=#9 zi-}`ZfT^#d?CAXl%05u0#w>CR=4alv{_FZPbtf z%t~e_HNt2`=QQ+C6_W-YPM2GORjY4o!Fo2m^-JQ)a+5CgZ$ofD`E(%F&*DHRL{)AW zg3$gR-@#@QSEk_enLaDEURy@EkuZSHoQ?E=wiUlD@wLFh6b<(*R9rQ z%DL96*#=1&S6M6wRz)p_+$_m>ZR*DL;m0Xl9(nivFKy)O+#Avbe%XNWx;u9|IM0U! ztONj<3he@q%Zyw+1MwU%4Px_($rJzy^zv9)8~Ih2riw744uh_rq9DI?JcP>?TXuRc zSxP5`uZ3x^#cG&|9IFQ%3tBZ`Kd}uj1r(Bd-UD74@(6hkKxVy*2W#pqJ&C^XFMtqr zgna|5(X>-4=_rqDA8UTCi@WmX59?ivkbv4#t>r$3bY)ZChhB-{k1_Sra3Mk_7@_j2 zMZOp;FRdt%9oGX8KFAK`0N;mTMR@Lsn$Hu@zNslxPFeyB@e?V| zJr>!aC_&u*E!&)2ynhm3-QBuNYkkZ+s`5-9K6L#0qA}xR_h`X$YQ9P4^dK1Ey2EAK zGPx~p@$+=aj<;2?E|LA`xNWvYF#~uz4O?5hJpOXsD8L&tSemdCl|7Y`?!CwqP?XP3 zv`{E7E7@F8h<)LUU}|*rnp@2q!Va}OyYC6Jf3R4B{a7^QzIxb~c2TbUcs=#$yesIS z`;)g*;+H__L-IZO(Hv6UXip{QAIw3AVB!W)3o58Skd)Rm-%s7DffFxtm)9fEOwE)CL$LY5(vkmmd0e*)b9^1yDsq&Nvr zl#`lij?5zUigV`k=c4}Au5Sg&jJrTem9H*Kis6s;jBCNpkcIE z_UruW{D*}up~1UZvdxo6E8V+dR_uqxzEhGM3c4eQ+A|BXVR!3K2&Cjo7eyvw=Bflw zlT;oeE39={WcMihWciHuJN&rs*HujzerX1S?@e5X?Utf-N<$^`@cRM)7)6`@nKC~K zmUpB&feQe7y}Y~Tv5EL<6lkUB;XlY}Ml(NtkaK5!aioL6_lRN0?>b)6Rb0WD*`HRS;p?TTQ^jH6&He*RwdZ;-VF9?3`c#Qw7m*;vANfKE*`j*>%j{9UQwL~)Y z?krfypSZzki$P*~8t*jEQ!r*XR9I|H)Xl&@vr&XuC~()Elc$430gI- zdzd`(2WeKUaU&E7=+gJ%O``7MNhJxr5@G@mMJF5m>&q4#s7mp{ma<=2By4nJ)1iW`VmVCkrltErpi6dqyXn z3_~&vwgxZ#yY>sRktizrUJJqpBfAbX7QaL*mRg}7kYUD_F!IxYbmpE;Xpsa*RE@j4 z7$h$bc!3RMIqA5pwq{KSPEVGuW|BhAZ-XtE_Jj}3aO+W9HRC7fmM%WcdtX>YEr7dA zcH$^DiY|s|*2wfWLiw8cz5!4_Im)x>_G0F2oBom1J(yq7vT_kvJiwP?Vw<54F>MMw z?y=#?VSp)9g9LBoV7I*Lcui=M&B{4;GIl&fj?0IleJENNaO9hJ`Cag*##)K%+G6Vl zgJwgP$IgLmv!YR)klvyK-TiJ&y+@=&!5VW8VMLfZ1oig%j4&eQsXrX>$e3AiUJ2E< zZG1}ao$a1oL8&YUCv1jg?UjoS)!IUD{K0)MOe2#4e)z)?2Ul`hV%Do~*=p2e@!SV_ zDV~N6+0E%f8tgI?&VA3g?O>M#6LvKA*%m=upuJ&0_7&Zh?pWM`qwjPzp;!o7$hU2N z(Y(om>Ydez_snw(#gHk@O2DJq(r53APLIpID2%dFJo6`07l}IenJ2*OPo_|90rGP6 zvmjne!;{@bCyJZcCmZlr^q&s9Zo}9@AAiY97JpHLFO6RgSPq*C5E|FSemiPbV^w*9 z7kn2dVrV4vqNJ;)X7b77))C&$^hM9+8I?o96tsohNCKwrxp5%?4dNQ$FBz^$q)Zj; zD5vt^I+`nJPAfixaN#Z`15Udi>-eaFnhN6#x(GHD?r_UF{LWAxv3k|AhMSZ}Yr+jh z<}pbTk3`Jq!F3uU+@7GkBN61`0X6-4woQOyMI58*U;*(EqGVO%=SlG0oSMQrAbl&| zy(oIkEh~f09P-E($)MDa$kcC?i;N0ZV~hX8PvQfy!8O}OTTTI zoRvN6q*W$XL3v*z8rJ7%XW@Ocn-zGeY4(1Sr5ZO!_I?a~aowl`m|goMB0usF6=zYW zR1r7s1>r5FPbLFyE3AB2;aEM;yVcK!;XF+uU!0*%;6m^nuDWIH{$;2CIt!p@H05%h z&@#uLS-ZkWWk?6jC8teyH0PA~-1P_(jdpJRiF5Riat0|qVt+{2>|~f+flj8qJDhO5 zhF~4Cx&s;{pL(2aWQcs(ps|Ku61nnK0>Jm6kcIxQ4CCK(fo0EgZrMz}yO|N3N9>}` za*yS{D@tTyfBhLeK!5S!x_HBdGYA9MA;s2JV>V&GbHoo=d({`Mn?kDui9xG%T4JIj z5r7=J=&aBe`aIg_Xv<(tqZ1o0swv?^wm*CbD|MX;)anfpf;3N5gS2~HAP9Va zz!4@4`pB!JcE|xPv8cG8nK0;WJpAwutJ7YfqWgH?O8HH4bEn>h`RJ?j!o8;Dx7Cbk z!EL#Tcj>oJ+vYwT`q7uZCy{a048O0H;QpC5g<6DWwT$$xJ7B%FW5xgvg$D}I$QUP2 zv#o}IvsOn-7$q~(oO<5vE*dyG&w?BSZum-Z%E0j|M~k(R}K+=#kzoirrdMgFVU(B7HwgH&GI1fwa|wbFej|-nbrN- z9WiEk6xET3Cb=PDP9!2c`*=tYW(QlRFZW|o^V)~>CW8Ls7$y9$ zZUNA`LLe1IJ_R#ztWur#7zQWT&y{wt{K0NK^KQFCN`C$H#Wx=4rpHrLYLxYrh$~(=({Iv$beIy)|T%-Cdm1gU$?uutw~$Euo3KIRJm;jC;j?hh0fuUg2lYs z%3=sqYQDSCFLa&usAL3dRq?J#=iZK0;8|({{fqi;RjCcSd~84$kQ2p2x7R0>CA}c) z(JOguYbBJ~9|SDXF5Nk-S$kK)bN(CO{nL_}Fbg05BIQ?9W7N#NS0{A{1*?N)?q5iA z-ke75@eZbd*qMT|SL4{fZ9-|o`9xrS5eLIl?A_kc*A0z!x7WPHotZ-`*V+{k zPQKwWV~v>tgWeR3>?#;*5jJZ^#8}S_@{_^;SrY^xoK*3n_6w@C1>vye&{u zZ!h0HYH%bFS5z8j0B4&hG2XR72)E*I6_dV1AwWx)4)H&eGrGkh8%{vXU2h+js>pzE zvyniwMZ9+88&gE2yohMCTHm%J8|RCGb=b3QH?b>eohOSZ)@a@N-JVz2me|g7r)XVv z?VLGde8k*5Ai}BInXuQrc)ftqRIritax%S_7YIQ=O+8OI>7b?ZxG_vD&ka?bVZ|tn z%Y7`e&z`upomr~{U#I_uX-?+R#!-Z*BGRU%J<-f{9G#1yTCKiDqj*zD*{bD9GF*NI zjLQtm=uGZo;r%iuS%>UrZ~7v-oR(-?K1i5V0@i@e|ZRqXaS&W{whT@UP^ z4F-rD*1J}ibJt$B!`5vcV`D*QuRXHY;LW)3^cq{I%3Se&EF2$)7sJYK&((?E*km(n z^Tm}Q!LT{{d%CrqBCaP>A}+XOEmyxvTVO?U;f&c^q+h-TY{R>-G83cm5n<6)7%Rr} z0oH5YAM(1$PZC*^;=nB9KZW@@2Z&H2-1g4)4iE)3`F&K7-n=YB3E(1S_=oW_({+B| zlEIY^ae6m$n-=Q>`vnbCVkWRBbI`f$7DtCl2z`F$gRj+*@t2NqrS$fNi$7U80!uc$ zjkRt|TBuQ*0Hzi^t7pqX5F_kJd|tc*SdJ)A)HSm$8;eN^W42O!g`mV9uef1}A)O*( z^|?c?ouCOBOhbwtWSz7qp@F0qGmc|me6Ku3IqFqK%QePp-;q}1zJok;y4`5 za(9THf_gsiGF@k}a>Uals=-nzwftj3a!23ft4NZLih<*gC(&P%HQt~mXdL!Ki} zNr$x`FN%f)kUoLB+j-Va7(d+?7NXIBc6H^x?!kku3R;VC|J^kCKR6y#BhA0aVR@h*3i3 zj!PBmdMLEcs%Bd@sN)=wdlIszeIKn>1;6FYVa?v{cyp(@i^2IgUX%zE5K7LOY*!%! zVja@=7y6nLv>&~nrL#(5wk8Lx*JaccQH9q@H~T=@-Uqe?i;#X%3ei!_*`?`aKa@)W zW`mOnm~xQmD{IFFv?V!pugzGJLqZ1PUU>&KNC|Cd#_-V9>&+?o=T@InL!V6!%*>X# zi`5_(ezMjdDTZvASJtL=DkvmxsR({~D2 ze^|WeKQ5#?2VZq%Shdd-KVJ~&GNMAY+5E%&o~xgh7pZsPjVlJa(CmJr&(dK~LoTpK zVSxb|(0-%;#V5x{>-Yf^eOeCvt2ej!DTcAcf~=8ss@c2YSS_HOJOa#2w=i~JPPF&Z zYKjvn!AH{ktp(A4X2_$w<>w%>*}~0xSz>=3AvCneZOwHVj znOfqd)(tQ!9Y=WsdrTCxf+E?~Hl*DKgZJZ*`GXQ5;cmkweGjKmMP%s`1UI9kxcziH z^%So8tCZOD!SZ~vd(DWO9)4^F+%UsTdy#X<^}**4p{L-H<43{$-sHoKygv<}5FP@B z7}y5OUfrG=;sDLea#%gwAoi%-l@lcfaau`wU1NU(y?IidpciSLQX@wCl8N0iNtUlu zJYM-E@7#qOnscPRf>A;fTD*R>lq_gvj_G^E9B}R)_%;Ab){3gT-PB?%-|9tN_bBK5 z755I0KFbHjA3c}9Dsugrxnp`wKjh{Dyk-G$UY@!yKnu!D(3Ygj(LBw|@8urY z%^V#{|G|njYv#Kcd*OF-2xWl&QdYoa8*3Qmt>22JkLffLp)p9=f>J`V#joUaYOD7$ z{^0}fFwbA9=9MGR7Zv!0Ba3n6nJg1gQx68 zGO$r95xM{r6KPt{L4B}-J3@K&<<>0H73^uLc=P9HQz}$M@1{$w&OvpS>KAANT(VO(mH|@4iR@0-#;ymigW?+V_c`sXoS&Dnr%`?%mwMBFKqhl+jVErLNT&8ocGp@ zEiNC5!?E%&`){>p;7T8(j(1wG%hwJ1wT4& z`6eB=hjzHk#A+4Kt%fJ{89pC=zgRKqC#1;(sLSu0N$+W80M{jlema2^`?nk>Roco2`V=ZA;PT_Y-2m$WreDLZ zxV-I_wcQV)^;tO+o$BdMH?)G3ORl^DM8OWdoR`_=NaOpOyx+b)$X0I4FyVch#9U8L z#+0tMN8Nukie~M1TuPsn za7srK2i(N!b(W;y0@rp|*RGu~8?n%Zr%bc4e7altDUyPGhX`Hg_vmLoy5FiH2Uedo(;Nw5p^7`I6mx00MhpVCbuunNX9 zbr&74wjjXe%E!1gP=5Tq?%jhCDhn9p=ujOGrIK@Syk81AA|0x9`>Mk0KAU80ZhJe> z{5z6#$X6ctL6(O)A%VOf~kccbnBIubAX_+FcK#^9X+RmG8`$06Y=*kKTA1dJ_aVwjsM`zA(Bj{*F`d zTjsoo{ET55Y| z{iK1-$vx0c3q-QPB(pVD&Xfc8yR3LG(ehp1#ea2VYBV(S;6&4#7^q1igxU8A)0%?; z$3nJKc9#qbB9FqX1l=KaexgYlsozcx(uL|aa6+wp{M+oLEVcY*kSjajR5hXt+*wY}*lF3H+Kh$h1A6lQd0Kj2)A5?%PH5x$!N=_B6|{Dv>jr&X3ZP&d zM~+K%id{6>LXg#I(qFMJ)mRah?v&b$A9p z)^B*m6Q9&+0Horh@y_+K2!KQ@M+3{G!N|7BE!THfX3Ik!{7$_7Jp;=4Muei0n#;eP z{-p6ZcYkXV%FKEUZl_do@13n_fc_BA9euiD>Gfl_>gCZn%QU10Rlboa_H8x^;=+9t z%8vEf&kBA&U})_6`liwQX%W`;TXpQ{_e|l@(_mn2L$2`p;B`ytKyKZ8%a?&l!Yt>m zSi0o-yL6sT4GjRbVzUm`ip;g8s7UDxw&f zno+T0oa4-ko68aC&5lLn(RMeeiC%S;>@uPPCY$k0D?Yf7YH-VX-Bs^CXgeLd8afWZ zR|a4Ec|NGG#$sTPf4jbPA1y~~;>F4qEa;qJ!@cD?1oIK{m)-Q>eAO^NGvJ*CogQp1IS*gBW&a3ms1i(sz zLBKUhx76A9wXRzNf(N_!D~G5G-i8wbIUD$e;0mvpW1if2)evJ5c`0Q$KSMddAg7&; z+PTU4lzTaYvuhj}&^go(TW0mS?}&aQza?!I@l#7O;I0?>Y-tz5!jo9PxyGGfVDYBq z&y9FOv3V`U&U5(T%C|)-S{sc!s58_0^!or3LIJ9i1Q)&{hX|yIe-b=vR8xo;hBHLvI z>@W6!DIJY}!binkg5A>aIpV0rzA&z+%PxI{Z_J8%_~!ZYnTiI}ac^a9{O#6i%pH`B z+#6g*GKjCh!5NR1lXI*bShZ9`?Df0)NufE{wWk|}u73xERxM#hZC^8d%?OQ#sEXur z=e{ZXem%vYkr8ouZMI0(j~{EVmRbAQwFjUELBe;Hr9KZxK$_an@_;%3KuGkSm?z5v6Pa&{o5bKi?63ZQIxqbQzbC;^a2u(`s$8mCr;op76f&5LvY{z%W-C!0@ewFjSi*veQ&FDD68~JF~No1T)NPYQ6LNARochiX7@HRdu(?>T2#(GywzN) zaB-^C&DfU;=Vfp8x-A9M4LzqDj7RzKs^WL^Zmmx~6%&)5wuz~3^~#{)7D2&@_!Eec9|cu^sX1;SfK zZA+6AF04|)`m|rgnRBQnWg) z9HDz)RGbL3Fmjqos&Zbf_Rf|y>s38|4qQqQLci3CWdcOM;7PGF{G|3RKS|;##As%k ztqCiL*#>S__Tl@;xcv@g2;1Ud0n20K73qG-@7TK3~d!%AL7 zLy)sbT|&^|fC-gz!Q0NciG$kDxJ&Keu!c`tkWqJtEObk{VaPo0n}BYu5vC=Nhk6|? ztmLTHaQN2zD-T`L)d(b@Ilp8r)!;d|B=&bn187Ik7DF!aD$w#TjR$lB#LU)-KjNpO zjVTK5@Tr$oKkGK@aeMGeT0n3v7T~*vaJ~$4XDA?ZSlN@C&a`^LEe3L;I=@>7lyhXs zDcybic?p_pvU0YArO%~ZXw3u|->6bFa$RO;7C1k1SHnle@e+l2R1-kEy1@|@)9Krz z|0+JA+pp-=>mZ2|xvO(vCR=lksr>v~|D?(H%4L*W-(<>a*o3W&C>XZ0-S?9lb|Zm) zTc)nfLVA4uy<2=0FONDVnpuotevEPZ?Y3(6a|Y{&te!W^)i}n_2l+8B%$hPpmRV4* zPfpG&mJQ3tTqvtx*|mSYv!M zma5gkb#<^HoyWdr&9+impqR__DX;zEM`){fSxs}?jj83fz<-6t=V_|12L0d+P2OV5 zEDHJ&kj!hDH_=flS$aFFoRMNXO33cX$iQpbS=aK?#v$;Jl6!3iEhouwah+QuY( ztA}O%s6M;t?S;pak01C4|0w3w8Q3k%SRO2nZ>z_eYzln%%YhRej%4oUZ=87PEzG=o=&Q7?@-)Q97Hj6dF|`yU|v?>$2z z&)*kLB+m-hcoq+ttB_Op+>d%{UfFHRFKYbt-`yl$19s zylCIKSiokoB&}vBEq9y1^uc2O8!dC@ynQ0QUx?f;H_RV^T#XbHPIJZ5&IMh(vLWtW zg+Nv~>Y{gbKK12Zr(pJS#u!&?Yn-dnpK7d}BkuvP^@D!O%<@)ynt(}_tt!XHz*?i7 zUEbu(?WP+?-b6mKKD<)6;-f#{=5Tn$EFIu6f^qRCj7imN3`VYH>5QYx2Fln6OohI7 z(wF~QXd!P9#tja|G~$I#vY6srs_pLg*mYit<)5qLmRsghx{qR#!lt`UFY&I2=X+OE zoTn*$O_AoYNcd1M+W9z$1W02*T_z*TO?3(s8a%OvwkV&&0C?9&UE119zjfKqHl(s@(v+Q-$aU8;qQIVm(iH|96P}BLA)@o}jLF)1i?kBKx-j}>$ zx9G>T;S@X3q|~chpS;_Uj6mi&sNH!{%&)W+e2#KUEhm0ix2!Ak^zh4LON z$2K+>`j*!mSq2c^r^qZLF$!pm|C>=f{r$Lb1KI=2v(jjn^agj`KYORkdiMUrc?h#*MycW!0G*Q9vf>QF`ixSPzx= zWEs!e?&k++1EVBo*(;6`nI>g-#QVy>2<4YL=ch`BNCI5+b}s&{W$pJEhixeLywgle zZ9S;}_3StQ{Y@)EY07OjzC&{(HStD4Eo6HhS)t)EkH zNcCjajHNrcOwNjiIm{JRuE`0BPQ7X|PoR{0(Jxn_C}yvD+ur#r;pW=(OjWSYP?{&W zm8Hf}p&!ei_QIoWTGe)%#^P1_$1flM{Fi^9>%?<5lPryk8&dxP*#DEPIp!2aiIi0F zvZ;>A5yos(1Q2qE`_I2sd2YR&e_2{)JPF|b7*nd|;GXS!A;|D%&Mcb+ryBy1&LXJZ zFXz=bmD^)@bL>p9JNCh6;FBZ?Ktf#GD6ZE)2w)`K(w<=yK+e{9c2cK`K%DU%mNJnWJ ztqfZqoL=(;XxU8ajnJ@|ML+2~>`LrJM{Ilbq=Fx$z1{PQ5Rrij2CAzJmQLx5#8=x# z;ifv9|MGSIt3dykUQH9o3MJ3f0p0)iLil7+0Zopllb*}uWO??&R zn3RAJPM0yAu)5jyUm5aXpUylzno2v^Dc)U)-PD@p2^h zC$JCHP-U1%q*XUS^OG$Lb&=EOMUqibl1$e?QpL|FP|%tu0OGth5A`^ zWA0q42clKq{wp$@&)6fwKYH20IK%%EZW}14# zvRXbFvg~RmDsND3Y2`c_Eali6#5@Pub1Jb!8&_h4ppk-~(j)-hAa9pwo~4Y8^xlA& zBhCIBFZVIcqMQT7&z|-oHxQkS(5J9Xasc*Ez3H4k-(Je{!^ZG`DfSO_o;a-n;G7l@ z(_r=g6Ro_tLcyWg-}qY1s9c&~Eg>sMQaVVQ@y&S(PV_F-qY^Ux#$?7kE&WANGGAp8 z-mz3(AaT>)(@|Pdda8+dv7$^3rjeQM1_Ee+vM7W$X& z|E=NiG-bggFR*psWLnMOB^}PcAVHHn(Pc20g&9a{>_^XtiTBygi;2U^o0RO2egGXT zkZT4CVW~A&!iCBuT&AA;$%BA>c>|w4-g?R5WcQ@Sk!h04_c!0Pm!{bWU4ORpavPBU zA6EJBL$V>HcV#92ZzMr!oNAEUBQfSEEt}vCc3o|OOty5q=${@NNib936;1og2BM=yfHsBz z*?UJ$c1jO&S;DOEo?Siri2Ij%&(l0eSC%p`g@PX8|HH$782WEzlQEa++P2cWUz{@U zhA+itypykW1%igA(an>i&qNNjGUU))fqU=C`;Y;yuHFvVsnsL&+;StVF?su?i1R5+ zY;?8hez|RJ74jVz^3oB)|SI^|Ns0xVtRdoq@P9awD2s@wfstNBKIe6JrZ zayO&L3?}g3X^ly%PwE;P8tlq@N2Z$@n%fCtS^Sb2uw5Rq}Ise8y0{IFMn!K~_3K`U-P+dN{OB0^9$kc6WY! z+oQd?TbgY@(?nArHM)Ny#(q2qe0?d>B+tbF{~vifp3K|iAlBo5#oNkonToAh%C)Hb=pZZwm><%rh`|90KTv-y@k?itT4z=}Xu=YY5jscLpW zdlA;ar@Pt#t;{uwcoS_LIUiZe^QbtqbYdL8D&Mk6Ya;fLbOZOr%%#d2pNiO~TE_yh zX9W!Pi_70IGv4#dmS=im0vlQ)&c(S7N+>~Tg-p+UzuERC>dRbr3*qBon`xz~hbKb; zwXoy1ef5Fb+YS0&4Ug?)}87xg)V~UdpXsgAhXDHYPPku`O0!Wr9iQZj4c{l!o z@9|e?wDZ1dd4&vg&3EVWzvyb%lrq?8w`@Y?Yf9@>}Pw2e9lmOwRnHunv}UfQQh*FFj>_IbGYVnF?`c!~8+ zg&VM7rV1*AC`lmWjFd)dGU}r1&A2G~)w}z2<?cUC%#Ub1`Sm*?aA^?sc!& z`)FzPu1zV>cKMm8Bv&YZ@npoRTUkjdo2eY{pAJfR&ksqzcS8h$=d?i%tddOo^@WA- zst#GB{0X@{HkJ;k+pq2tm1({~z7gv@k4(mU0AbNe47bSJwrSdL26${4zqmu(n{71O zg1>SKNvSde1Ho{zo`fo^byIT`T?H8gN&Ky1%7h(QbtSKlbN?^w`wwD$mY2SS-G5`y zXLs#m*qWqIRMj-KN9FtPye>-|T`4YyGr(~lU6l^n`s&p5-bpF38$fvcfP2rt{}UzfkSW9L>pz9qBgDUyuEC{Y`Qr zGW#0^0zB$k407nCG7Q-bE-{k5Cq~uDL{36%mvQ5}kZ*_eX(S=zf`8|*S0Sl)gV4ck zk2bY7$VX$~e;I~M34kT??a7$`ne_h~Q3O(CJUl$U3wcwF%J)Qa%CTT_>0?)1emBL*VpLo`+I|%3^McbI@HwFE2?UI(x_>r7aQ!I zoMLS3yPm^f5hN`IB@^u(Y04cwKP&@kJQ<$pV;* z7gom3s}4Zp?-gsx3*b{PLX=ra;WgU{+5v|@Y70K8)kfMH`&sFPN}v2VsewGXbpu(mz*69$YwW-I5oqps z1_oo4tPLGlg$m@_ruYO!jUGFS#X0NiimvU((pz5}Xtp{Vl=Yp!{PC9~mANB3#IEI` zMWQa&Fqh_Q>Q(KknV;?fd|DtzM!^Sdzt;ZxPG^TzBi!e;Sn^Pw;8cM1kJGzH_2EGKS_eRbjSxRLrv%0 zjL7cWCjSLxWFR*-*6)^p*B5U6-*P^E5p}*Y_gUYAMYB+n;uF=lO$lsSOZ~++)VYvw z?r}SkpV{$3Dmz6fi^0kJV+Q`KYU6MtBQN1xBk_>YUM^J2(NWdJdPr1M<%sl-*PgiJ zZ*2+_`fFaP!4SGn(p-CAUe$QG(uK2GO_2cx~pppdyS-=8D* z?7y{_*7=*t_pGjh+lLqYz5%h%?&I5VwT!<61>H$F4*}N<1WvgBhX@V00g@~+sq?Mb z)0jJo?X1VXT3Wv?zqQAG(QR+DM5b6Cxj!vCsY#g)+J1eY-7%m3MzFjZr&M^{3FWhX z#X(TFm*2r|B{Sw1GS$+_DgEP;yh1a%XvH;R&eD+xCNr+o7IAm`(ZTtUQv2N$5WDR= z=$X#b-N0TR5HEw63`U_Tp34o=UZ zGu_PC<}k?+$t+PEvH20Blger$2y^`NqFq%$LSt*uMn;``3;pkHjWk||xe0qt@J!i$ z%pQR`T=Uoe%f9e*;4Mj{2>llT`#S68l^BO=MiEDqFTQvq)R&dg-A_@tbf7n7sWMPt@n z8EIYCz|P&ok9Z6zFS_>#l#`I zf#DGkGVXtPZV$v%_doaIU9c3OYzW9<{J6C%p&x2x=mqL+gp-kFj~!NLy&;NKbL&i{ z{+v{HK`Mc_9;EjdBs4d(R9~{KS52%^od_Z3GQOwi^Qi7q%(=^{WMuL`B!`vzrrxIB z(_da$aay&XmE;;P^f$_X9fzs%e>OyhNz=4+{h+aonH|Z=>d;*6Gp|f!)fsoC-c1qx zIC&`MDq)Xx1bvYtNE0URd$e>Zp6wKj1m4xv&8e5!dqnyCJjkxrX#(oa8}&8~aAFJf z{#num>#zz=>JbC=P<;8_oK{npyr^S?JI@Cr@LCUYnDOOsRD@XX=sM)MAy^~n5eYK0)3v9tw>KOTBrGLTks z%bW}$1$pv~M0zVxcoiUoT>JTbFC+Y@C7!LcEF|WZ7$dh)2eeL6QoAo>#H|^gEMPLl znBHj3i6knerLl>G*gi)g?tR~sC<0IaZ97zu%H~13_V~jyPQ!##=}qR) z-pMskmNTZ7qF7lp?XtEIXyftQWd*-x&=(c`1xkS}2#7PP>H$y>X&?1%YJhLARl|;N z8fVX_LQ7pzi2GMaHF|6MlZ0^^7936T8H<20UzK7^AP+U9b2>ayH>C1!y zfMyrNoZ7%Z|AKD`#=bRGK#%}gk|6p^eXDR6--6>hqh73K$^~!m3det=E=tuEbBiziIGnHbgXF|TYy&ZI=S2UQC@`ZrvWnYxF}H%vm2;

    M05dn}?WTYR1tY4%)f`;#e&?rAOWcNp5WeA>hF}& zmjbS1`d?f4F=tB$mYi zZ*kK0OF4D!a`%Vg&w@@C``#+OLKt&(2TN;n*dzW0CHi_yD-D391%IgZ8A-?pBKkwr z<^lX3iwq+Dui@x%`H@s9#bq*vz3PmZ3<2`u0*|O>K5%;>AHzuAusw7u^U@b|?&B*g zW37_f)BZz<;y1clTqFtID3ECVb=-%E7A-jpS@7%QSwp$4D>5J;QzVZMZvTF>N5bo& zo_lBEtc^za{(xmOJ3_S^KvrjeYzWng){4>at1_W_4Z~{gmmqqRi*~@?M8ii6;0|%| zHOR$*rX?AxP3EQ)pJI|JXGj399X3-D&wl|IPIfsWI9(%r-os$yQwnM62|K`%g=k$o z4HMgx8#gI4IwifMu7QH+tkeFjq@q&x?=ko!g90C&T)XL_IW-x>?P+8rh16*GU4T4P z8vL*`xshhbG$e0gp!kz{WkqkrVR*qeDY5fagQQA-@1|J4h@WFhB>QiVm_o-zrmM@# z%0@zhDxt(H?y|M(jEZ%;c=`$_QK>=V@^V0X#_%tTa`JvA!*}MN6U_8Yp_{52YG(epR)VKG~X2A`!xD1V31wr|L31SA9Yw zrtP2zU(??#Wk7pqQ?=>PH`AFbo@^uuN&pLu37L|xCC?K8+;xj9=cNnMAA;pNG>hHZ zr(K`!vQezwsxObnqZb1tpDfPKnjD3ZjTH4K5RXT1LjUWR|Az+opKF&t7@)%Yf5kl# zY_y;dYO0i5Y(DWsd&F5e)$v_`4+YtpOCD~dGCp9X+Ym~g_G21Ar?s5pDoCAr;_F0G z_@FS|CsJR7vZh+vq(;h)&n&qic(j@_F)~cM{+ax2ybywzj{CNZ=TSLTp!|L#{oy)1 z?9zd$)Jv_SR~nE`ly)u6!H#f_1sZ>BrUVSMqgRtkt*67tPSYTC+S+DmBcl<1Hcph^ zLJ*PP1GYn<7NpX)0uu*`QWK@Wr@Biz)pzkVOsf;NUhr`bD=|g%o5T~!%Z=>Pd{PgF z;gxZUKhJ^vIgx~{RN}^y&00FeK@}jZ!vLc(Y4#<)xPUSzgx~WRBT+98L0%DmJc74l zf))db(SAhBwLOzMELj#r;zUWlWWdTWYy>6$GqDN{l1t->7T1bT2JJQnNxzO(<*t9t4R!0aijR#kx8PLKB>$4^YC zGJ!@KLWXFo(ZL|yjmaAI+u{ZNPM2xav*U)~>sE&(d>(639x0K_M!S3;N9K+~8^Z#EOimi^2YrM~7{!((;aj`4VEV$@6Bg0uP?M)0n7?;Kxti5JSp zcu#l$8&V;kFf-4+XH^iAgm={}bNz~VB7IEqn4TnL=DtarMAW$OgW2w$jSKrx&!1eC zj!2#IKzL?ju7SZ>@xo7)6(T21Q$UmAL&B8%&3YmRt&dJ~*MHF#q1CWVFPoM8 zdS0K79etl6a{1hTg@UXQ+XBOv@AQT0Uw(1|o-lsZ1+t!}(-$5Z(BX!%l_EFH10a}M z=eh*v$xuAo$|;4z)_lRXS*3LvA>UW20X_9m3bd8qy5^{{+pxP*sYVZV2cBEImt65! zR=(%U&e=K{i3nX@t1TuGa<0?Ojm{Q~$}7llZ0Qjpo~UAg<}?yB-Tq1+bF9x4KT=~M zu7Wl@dSDe?AKJ=cO4aq|9@{(YMR0+i*XCg&ReN3%>G|*a_s>}ywpa0#!$JNPYeyv; z5hD!BGH~dtutgP)_P>J}G=odZZQh(e2J{3?zaTZyv%KBeI(9H!MabDF>b%wTa`-$P zmg%*5f4FN!Eo)tDx=C6#HzZRUxrd5Ryk~*?nGmyg zQ*I5n|AAZO3`RV?>7YWxG0W1!;ORwwWj$f_xj%Z&bT_Yn`(u41OwQXMrb@I?U`!r& z^W7eAoR1VCIEmaan@N2!VuiZ;Uv)cf1wkfgfmgpnvE~l(dtUR-kp24+gZgk}n-`G#56V|f$+&{z5 zyKw#LnETak_6sIm+uFR{l0tr_hqJ7H39+ff4ehcu7#ul0Qt2tt)vn}f{rjnX1Unmi z&x8_u5X)EDjNCGFKXYXLFX;<@0{YmJ&+-Jm(xFB2AHYY;g3rVHOz>KcNZ@IWeYV-= zp{39kv-CO@V%FKo3C>t{EP ze@g-I#kT1kL=LTDO8=X5_n*&*DbVi0XQ4$}Ewuk!^3z4v{ehIZkH2kO>%?N zGJ9#n?xSH8nD%sie|ejw_GgC>ag%%O)TR@*&C*ZVeg@IPFZa=TD5gQmN7qeug@Uon z8c=PLGSY-u;%Hf(l8-S$yU#*jvL93uz}|WJUnIJ+gPB=#K6NualMUJMI5(t-82P-r z7lO0pMI?ww>S8cr#!Rb-l#gm`Q05DpPO3HlWqXF+RNm)i;CIx~!*-6Zm~OryPjESU z2eZ|{riP?Tf4@u_$A@IZ#A%CFuPOAd(-6onSMW)w z#V@n_-+LJ!27PI7tm>{S^A;xsz=p5t61mvbXltqZ0P9<;S$ zZYliz0W4tpS_&-s+sCb!)>`{F>Ok{LfLqk-=X2GDj|UaFo9R!T_m2(O8$4eCv@e@k zG#V-d*qOZ@59@{iGf42H4T9*5mZ31)0ttMs~OQ8rG(K??!#&H`Q*@u$_nET)!`6XxNQb;!2Qj6 zw53n1>TJ&7>TI&=Egomt`+<~-v^nbFxZSdX7&oUyjfX-N+f-LWU*CdN^abJIV2~7R zzBu^<>TRQ_3Ugsi36z7zw}3($1MKsX0PW`IzKi@K@o7a_{KH29xCcnj@z^Mwlko9eHwM^)Ut)8?+R?ok+;bcYGC9XlOA4Wwb5D)kI(+{Lv$EA6L1Bu&nI$X@PHkNqop(8+2af zPu-k3@Llp5PvM`$Ox|VZZGN8fqa%8I&R&u^>z68pA`{;3E?KoexjVink1v~G9vXPC zjlceYjnf6OnmdU%1I@cL86dx?w8Tl?Jq*^G#GtLAt{lJmr&k2LUas;z8tgLhW*Q`z z{$>J{Z#Xd2U8Ys}Z|*ogyXd}-_*I*{y*$69mkK5$6D*~HKfG*H{dPDpu)qy7r)8S> zU7JcO-FMWmX#RoVS04}As$dmvg$&%Rc4BIifx2d8X?E!8N|)=v^JSuYGeRvF#d>WX zZqCWK0ss7!@Y4a9M?+tpv~=YK58+XpyZKcvd@Ez4#=r3?1Rf!}6@bSQ{}bKT)3qr2 zlOe3V_<{G_2_pEo;G^rP$iY;?S@EC9+Xmr&zF_>yiPy(>i1!a2gB}M+XYO@`j={Hd z`~#PF$?#%6#s_aZRgq#pIk~nwdcWY@qgGsl6dPV$By8}pK`WUTU5oO19oHU(!A1;G zKIK1`N}e}u8B|BzgycHdb-g`ZtLL<{FSqncJh?6on{~msK3eUAme;bGb3J5o3@Z#z zjP!6O;U9R8@_wu=7qXSU#M)$2O_o~!uQ?h2C9a9kjwu;V$J}o~|1d450q~DfA(XsG zuyV=G-??(WLE#q~K#k)PS$$A`MXvaJPgt#ZhHA_k5gLic%Ud_zGpr^#M%TbnJO zkiEnxaMna7n-|)IvWMF`4xK<5-Y>YrG+Kx}y$E0@^${I^{wrs4U(x3OM+GrbVLdJ$ z9}t-C05C>YHmOw&r9T+Q^9mAxbuyfUC1L6#R?YkYf|pkYCKwTYdi4foQmZ#cHeF$k z$W&di3-VEY8dO|CMxhPn43TuFUfv^y{;qQq#4z?f<=(1_Y^twB9q_8H0pln3P-su< zfedU`c;-Zynaec#h~;ECQ^-8>jfj~{j#KxqI4(v8{qk-|q*SHY%^co7bZ1)7GsI2@ zp@*y3jVEUi^J#I)jI_)N!uWEs zGXaybW_EgyxvTSEH@HKUJ80FF0)04jnd%V{YQ=$At&!*jh|*MB`P7I8M9h(lZh-8C zOVuS`O}K$fca_iy#9Fu?&^7mVGICd)61BdH7Gs)>&)lzZnH&MIG#uNbL54teqDi*LCHL^fj;lW?4rURPV zaE} zUexPhDA8fz>%Sx=&mjR;h$wr434JctPmpYmia+jG843)?-b{6-dZ~|t-X%4B-;2k+ zf{!!-L&a`t_~!vgx<2^W{grj~4IRd%osVn!jEO;hZYX)8`!!3jwIJX}`o)BtLVAX) z6nxTJ9INVuEckRFkrB@pP^2O4=cmWs&wYf@{QN;ZmDr?gPr~BI`89$y&Z&8(`R&q zJ+&K+10od-5U{I#SOqFaZn73-ZjpM}S1ZbwCAqYr#nB7T9CSLNfs*h=_%X_+XRV~u z&LimOnbfEIQBx=3cfc26SP(uwzkJUZO?Er(SbPSXRJ7HyHxtBCjB5l(y};F-j9xbU zt?D%9^^wu71|o$F2Z$?4dp>f14vTWls-?FzDSK8QJwc|InQ$ygz2uFwFF0JBg1~+j z9y0Cu`l$I(ukqUeB1``m^k7GGZCg2cW1OQf*T7~XM&7c5(p9@);H9wsaJbNd=k|t>}YL_c~R5Hk9g%mSAQ>Up8Jbzl(QQf+6 zEr0;+wW*rb$iQvQ7L7VUW5l>}W2pzAI^eGP@Z#1iK4a%xvR2Fpy&A5!$?-0;=oFM0 z18EuHUGdJJyC2pRA-6pb{sK5)R5~b z*be$dMd;|>sKJA=`{UBS;4tM*Nv(@3@uVwfXUDeJ#rpSv!c$|G`NFwASH=$44vnMz z$1~Fx-L)*m%6CcN|7A`iB>NvPJVuuGasD|mP$S?+@vmXm?nhX+D2u&K#C&=$=>)Y9 zIBXxtr=jWFFrtyWb<7{)f>3;}BjvMFRBPS6m9nL!w^EJi3fhf$6N^yXN(r$`ig-L1 z8DV~o+0$BA#6;IeaQPk<$P0e^_44FSod!Wkg9O@(ujf$Sw&F{qSJcvhT03}rO@%M{$1u?=ca>KgW_C_0A z)9WL{Uz*6OtDtMs04V841`twFez9N+^i7CJ++umD43!b>0KI8Dk*qe>_aZoTx^xI{ z@u&_9o>|#nnEH}ylR;i+F~sd%T1xz=c=%7~CREST48SZsSTY{;jo5=tD(ecKMqLG% z0f%1{HHQsrAe<`WYDTQU2nbg$>*{3=giIaWn@_^J%pAxc2VpzBPaRyTEd~+6b`ckr z(9@}Fp_AEu^mE0}##kW~*+-;3UVoCw0;9J<2-a=AY+0zGr}(NCKRzC_l74H)wqM9QBusH!`xdMej$p z--y~>Zfq8*MZK{ypuW+zfsNs?+g)MkJwNLvjevNi;zGjDlgB2ow9>~Z%J?~c-2 z#K`5w8@%Oi=SwK*knqa{%Wv*iJC${vV(5+#S7pCeg_*aY;6HE5<>a3#4HAcA05-4a zgZ(2(FHrF-%Msc_yL7(_`D`6ia5k@@qBV&J2j)+Khg{d+@mk!DBU&As-18LNR_60+ z?G4PaM`fM6h@)Fkmcutd6&r6X*)KO!Gyihjz;|6new{%Wtn#*->!s@SD`fOQ?#}O8 z6ePCd!WUXJ`!YO}?^9Yxg2-{&eha86TcIrvmqE0B$Y>_U+s7x6!iujkleud{0Z5r~e)n7NM6pi(0Xfo%E6KO>xgep+8V)08i7|*$q~jkorZ}QRSJo_b=m-SwOM}tL|_dr=r3BRtrOOUD)jz2+=hm^ z`5&mlDv^g7%OpfS^zb zgO*}#padIGRPi*2cQVT_JYO+3tn`}f$GF$|dh-#vmS@nh^CC(9kKTOWjVU*>aaXOD z+uo&-vF;Y$`j1zV*y`Yp>gtXr-K1ML_{N6@j0D67D+0xmL3a84&0b=-0G$o3WUOx9 zViqwU7lN0&04k}x^vk;mZ_SS9z5o;>C+be~N#Pb1CmP_Tjjsx|*vL6&h2@+gtd5_- z{{P|YtD@rSwl1qE+}$05ySoN=cMBHWg1ZNTJHdk&?(PH+65QS09h&dnzJ2?l{~l+I zI(6h>Kb*1mUUN;kl4fXPD&8})n?N35J@}jEE^Ne{YUZ-3>-}huu|f6PGk=iSc(qms zv~SW_Pucr>|C^5Z#c3MC*2YY5{YweQFjR*nCoYzF!GdRWvtH*hs;(u$Hv)33z~wNm zAwOY8W6n>F9Xl@<{#!a>8r{m+4dwo4s~JW3=g`nHBlhQ=NQlHLuN(S<&TBaMdfX%| z!(aQ#JtTv#@nMi6n(qt8;?@o|9Znbi;jiX?Fjzgig(v$#TD%5i|LXAm14gZZHsebW zeNFVk{B6W~#XH^Gaih52NB`C2jh=lzf1Ea1b|OQJ^WLhd_t}gsx>Q^|jJ-BE{Ox{q zwe3(I*)n(^6kcH%ULz2ptt|9bz)j-)W~kOG?!9Buh~VUGXS&|qyzbv z3vm&-;}#7;cN?~|0KgO?e-j8}P-*1SQfNl=?lzS^^DJ?QI7w4MBtepmMqH{G_$lYo z(n)W+=O!RTQecReNh3n%AL#PQ-0k$;lCiD@y%84o#z%T511u}82c%(Kgh|(IG z1^%mQe`rx+hDPd`W$m8Qbdlg0T=BqmvUK`MA!U<&p{yc!uNZAjyRb=mrWAcNU3KT( zy|M-RshvATJ3tZDdX-D`5{Y69{)o1frq4xXW4!^#5PS1eMdMxR ziCU~nb#@CD$~}{rf!{P;sVR3T*Pf|OY$=xOMK!mvXLR7a@O8LZh?t*(F5^OA$i;Hn zuSd%>eJQ9%1L)~t3GWF6jrv+Y)Qyd!Lz;Sl3r%wEH-ISdk7VrP)PXE9p+@w2f1u9{ zQeEvzG?vKL1uxm+QNwt*-TRTgp%*vx?5H8ZM_SDG^oM5aNPVEQ1_|j?5SK{YE%Uz3 zRn^74ViB#d_T5In%ATV3QB10e<;SNj=tv{Dp=DtXEd6c`I4gm0v?;YA*rSFkY<09G z1c^Y(yf$lwdu9LrfgZ(D@Eaq>hE^}06Bc$RP$pTdL|0LIMwl6t=IZNGw=aF5ZlM97 zH_;R3kXx8T$pOF2awlt+j)6x`Y3jUoDdWnn_V`vu$}`6OA+jiZv2vzHYv69q{+rA$vF$6YC1Pj9h1W$ObF^iNw=Kif+1uOYG1@KK=>#sW;ONJKEBxpe zM#{ZWsCJ&s;R zXSE1(`JjM#YdlQ~wN)U89{jyZFq@k*zk3Ni`+d4DN?2sV2|ytsK;sB2bnFy@c3tSv z-FF)MAYw#e)1DwdUaA(3!?Z^|4lq<}e1y^?&=~+DnY2=@0e*GP?KL#NjlLMCIY3cf z?~@bwM|MtjX{X@$zZ74gY_BH^KfZh~n5@d%9WBbtDwA#8xa~WII+LnMil41Jmpw$H zF?omg?G2lUem&&Lny?-1%{U*aKOO}@??e*!khE5g{A|9%PQjU-XuBg4aqT^krw5JU z)468*XDlX|`9Cyra94uWVLqNcQpSq*>Dx%FYl5basPThofrwj9P0^>rJK+;N9_WAe z71mYdw+!i8S1AA~o*Egd`kf2uZ%nVXjwEUiZzu&f+a%|$q1ZECW#{c{Erhr1wb|$V zI9^>I6Q%OBqv1Z_X?@9L!hZ`zV_|qEQJy*fPcQjH-C; zdjG>mSU$JGxpJ`M?NxUZMCT599xx%mPMY%$5h!xhkvBpfFWg$1h0bCoww_^9{jM(XdpnFqQu)UOHK$aVJJ-b-Hzj@vNU)rlsJje z3OWglXmsI^!3=rqE7Y}Ain_wbMi;Ecepa>GRno0P5G4%~InJt|IY>gx@hHzSM~uWr zE1A$-i8m&&aiz`4hhTMt2nY|63kDeW0gZA=OxBaQ6D$z|3ZX*P%zLmQ9{wO%L28iS1)B$-C zg$dh<*H!1f$lI8-ga#PtqPil3p4ia}{@ki5TWMee8TdVJ;3962WY_aaRO^7`+~&>D zu6X6*mx^+B!L)IV{uu#;;`S_b`6?LVzHFCymX0etr-{$hjj9P!YUi?odt+srg-1s| zSQL%WXir{5m>8N$v?u{ z-}0j}K{F$%#(4$q<)|ZOZEZw#vl?&(eDN)Ixo*s!&d&-s_21NWC>l}fl=)Zn&k)1z z_6HSI(vp|!$sEuuNpiR5|cNw3W(0|TwC9&XC*h^xcKoYdT`w4)<93j_5GSVOfIiM!_1&F!;jk z0oln&F{hOlC4Ubb{?(zJ20;?NMixSX6uLj^@>8ZnM9ISEEFKs^t>LLNs?EH5|5v08Co1E7ZRBfHORVd^XI>-NBOB>Jc$CG zw=#sTZs|I$P8VR}3y=T>8+=6AdiKMf(1C|j4ID_{ zU2x~=cHM4w>H#aq>+@+u%;)YWvQxFzUmC{7CqNwI%SUGIZ}b_2B#9V<7L@T9(2**m zGxw+>T3yzR*Lmo6%=F?qO@Yq>XHsH0kj}|YBt|V<^a!J)?nK`{wLgv=JAABeetIb> zdAd^?hAD~Qrh3GgEp5Xtw1vVi^!GMUqe(|46(dJ&u^PvYzIqx9^HeXR<3_FI!jHe{ zud>cxmEYDK{G>jZ4GyE+?@!)WbTgkk$5s^dfMWvuO_LZ4g(M|3a<@7`6-{6CV@b=A z9w4N(-rAD|&hG0yX`(G~u&1WJjk7WmO49d8ndcqdPE#mJ2YEo z*70q$qEfDVl^YmPPIg~|`YeVE8m>dNXtr)O3D_GZc5IU!Q~QZu9O%6|F~Z|mwu~JY zlbS{d9jdE4Ix;mBbOT~&avpc)7ZxaHK{&8sn)!4yx6Ib@BL3Jvy+$g5hsbmDX1m`a zlT2Z@N7?Bq1Ky8cS2a+@axkV(mh>g~U9Mi$S>`7TIOXudd~Fk|VxX#AY`DF3`-u%r z!Qxc8{-eQhjo$YnaO^Gr{W5GgV`1P9{tm|{pW%C_#L(%FrR$9iZ0Jhs%u%`B-k!d4 zu5#U*n0096tHSO?@WWR`9!YQqjQa)pYC7$^Ahf}dP>r3C|5DeKI|?*|jHE33f*XG8 zt3Y>XOM2LIljG}JcRB8>s5K6gtS`5O1)(xDbj;VvZ=&yC5hN?G7n+XupN(gc1X62| z{^+dcL%r^E?dAV{n43-f2RT5ggqZ=zqSzz3{`n6|MvRT}nc9@?Y8m72`=!avRL2`X zjLM%kNAbo>OH6f=cazJJtO`p8v?OXH+@y&&I~3XSU$Gx~U^qqtN7UxcCA%Rx5jTUxeEK{(oz% z1A&dt*F>#f4k7;2Bs!S~Sy;oy;gR_91N}6eK%(0wkW^mG4qM8aQ4nAFzY=5yc=a7L zQl||Qine@O&Bt2nzt$lX7gbq15tz%5aatEiK9OZeWDR6 zm$pT5Jjf;t`V8!jRcDr%De90N0?XxwOJ5`bc zS=U^aGhbkwgA$dmt4ogC}@!o z3*9h!D1QMf?<|@dJqioFz;i0~Z|$bw&{TOU!?*+@d*ik~<#fEjpBNl6C$7i&iK@Ow zI=oQTUp98o(iJ4k%U9wwKaoy}Nfc3~eyWAw)F~*2u;#&Q!ctx)IO7BsvZnV7>9|W0 zDC%P2S(lHPLt5i2SJ46w+L3pE#>ck{Wb>mO>NGJda5tz;J41S+hio>=L!O6q=dw9+ z!IIYfgolN4P-cW9MNT#>dMZ5xZ^4oZAuulmks6|5&;cD$phdKxV+T&T!d)lrg8~Hm z0QO>apl5%PixTvId$IjB)}I~d z`>R{~yE!&;Dpi+VFNPhy9^DrMbExhj!a$F&98~$VknK`Ius#wdk0hsCZc!}7KvyMv zUn}g^?uk`4sY~4e8V#?3(y;I;-5St7lV{)nS=1fBCkNTK6ncOky_ivrg{$okje_%N z@}T;7EsX0-q6tjJ<-8^_WooWg(nq4-Ti^?_`LXQ`e{tpRhQMWcOJnx^+0@MKq`{gR zpm|e8SO^x!k%fJ5B@U2_Ozo+~tA!FS=jkklZ6yM4frKcGi;WLL8@`)E!L{J!T;!0; zJaN!1JXan6gB0&rB^VcozF0v6^_0i2F*iGl#Mx4Oj<$j$U@Peu&}{|hh?d6{z+ZpG{Rf(yMT-H>Or zx0uI+9mWPa!|21!{q0a%;KDpoi`u;F92EMZr)zLceVT@BDx4|croIL6)DYxDZx)8# zL%-YYESiAzO3D2^_&Q|=`L%`Qt0&;AEevG3l|(r^e-{mbvOr<)$;X0y;`Mw@#vz-m zyaXWCZKxCkkvMeS2j1M@4`1f9+Xd(_ZgVe_&T2CownUzKimLGL@Iv_0q4ET~$)YA& zq=nf63h)RIQYl}g@)PKn$)n+$AO#mCwH=@umk@kOJWu3`6huSQ2A;8`$9t`=Nv>JW0_QAY`~GQ^H>zfgiH&Jo&XKKKb_YJE4q`?-@@VKf%*42j z?)zou7_~EnhT8lJLtT=;8=kvh#T(zuCWP?^1Yd*@gh9%!wQP|d%LP<~nl5lbuBIP& z)URASOSL2v|CaSf3@-CoCXO3g2&r|jQW)zQf4+srl#jMU2`Go^RfYd-hWIM2TM`)P z9=V(VbZqrItR?Jghwa`c%^fFbNUm7a(RsVAI=xv|?qJk!XDAllLc+!bYXTfn>~OiO z8y4BCUiQFd786bP(HGx-OTn)CP}#o*Yh(B4Ic(I3K;HLH!9TTjyadg98s%ZK*~MXH z!6&Ee`+TFYt4nT(cN5eLhqw>(p%I||L-4!D-hOoQ_m8Wct~XE5Q--IegM72dqIQ=; zhKZbr&6Qn?s1*br`_%O}k%4_S)w0|ie2(9(DXFI@+lw&Jz6T$bkv#5@$2oQUBaR;w)OnS^xLDUa7Vlk)e+B<&%=+wUd3opM&(BOo ziH}Ba@V!!f%@Qds+ggqW`Um4pj!s7Eb!By*9fnf&w>5ZU`TiO2H zHVSZBD?J1rk5T|nG9hwo;@?oL851x;v#ZOaS#*vd-@njO1FB&e_@I64#>jxK-~oQx z-tw;)6o;;?Rr!-FF^lT5VdkXgJ5#-`Re`>FFCV;f0ul!5N+(yu1JvxF2Q3|%`Ba5D z{=WNjPVZ(t_elD3PVCh#yn7suT~qzJ7#vUJA>kmmtk$y*Yo7C!^2+wE8^PV`>JU` z3$s&QCh|WAb=uzSAPK}thwe}?BQ1lL=>i&p+_xOw87R=&z5Nc|sqCv84eN1G`Y+_S z4>4qT1JlBI5f@gwz4uTo%4)0E;!H6*p1r+??@6aG{OV9ZN3nO&BbrX5#4|3ebtq|A z=cn#;b!Q-pTO>lSIKQdo{|@j!1O4xLVHK1$<=0j`E%%>2$r%PpK`|;aZ%B9JtU_D= z6kkxLH5m<|m#AT|vbC&e&q1!8@fYc_VPFfN08`$!8v5z_v$+*MU2J@8W}=0kB6(p7 z=C)M?#Fk)lTxW-;i3y&mjEQFA$#*dtscWJtKy%J_DbN28xy?+Cf6|SWm0C!u~zbMKyJRk&sasAQfHx*BgsN61Dz{7WhI`nS#%Y>P32R6U> z$soemcBZ|b82WlJH|b9oR-QZTT~-do$Qo-~J7{=qTsH5D0I^f43)WiK1<~f#2(Q(# zUjcYeu{oHY=e1cJ@DCJ78aeolO`(0hzRPjizsM;{u_DItzYn^OBtm=^yeOS$sl^*$%5AnCrs;ZXnX4NiR}5 zKeUVFG}`IFgE|4!WTB|I6|q)S$HcrcwSWM>ydc;O4Yd{ba-A3q@fq4U?Cq>6Ftotn(uyHe+;gmceYsFKpd z-s}(oq$F=<$;rAvHd=&@wC@ln4!JAWkRk@hB#}d1w>g>6X)_5q{x-8=IxVmg;e7P! zS~@WGrip<)_Q`lrC-r`mkVu5^KoL!ZRUdaBGLd;PV^(+i?e7TE!jQ8TZyO~GH+HDZ z?N$ulzEgoh56fb_bXszMg^x<9_zwX(ZLWDaXHa9-o7xABiI$T^Yki&rt7$T9RpV>l)pk`_I$35_wp4cr*J)ti9x#Zq12EgBGfplN9Xg_33%~>Lh{euT?>ST7+hSTrP zr2D4_CO7)SabeJiMzh^BagF&>w*lNPJo!QaAWgQX7?KKZ2D^`Y&0n7x0Em(-VNa#s zFdPD!z_C8{sA1gGg{XK#seUNz0oh}Kr*Dl?8cMyXsa50fse*1#Nl4s#xlGmoAsd1t zrCg5hY^X1;VX7iTz$$dAg7@Z3{|<)8Km%xi?Oow?0_1nu?@6S#P`mBvPDY=}O;}mu zVD9ZKyqBRBbg<1syx}Ckz8Royu~2m=WmV49-2x|}cWsB>oO00oC(J*uY2Nduu<~*W&8M&vJW*8~{JG*VgR$intkY zzVc@UX-K=F1v+Uv8;P;NaTl|)xJ2a!WA}LNa7j|SS=2yBY9a|^0%c`8b)mS)B9PjE z;|r9N{Ax)BBp0{2<7lAS@nhx;_Fmz`kc>S93sWUY%P+)v)@CJGU*7snLr;xHE`X)i z%~#OArwt3kB@&{e+HEC8kC|NW^uc!PtH>RUte!Gd2OG6`V*VZ1D@iguw8Dl0Yt()# z6<_1;Ux`Lo%K0wg@O<$y7Zs z&bS=KYw^{>*m`nPr4wpcN_pP;`(9|KfXiPz%z$M^yDDk(yUdyWl^2$;>X+o>YCq=9 zZ1*S=zon_r$hCI6FgJprmTaaluJUdCX^bNLl`Q@q2#XPMg}kDNk~8`0h_)H}KO^BS zhVb3KziPBN-et=&e0y_}x?Zn|9# zVRIZJdWi6~rU-P-TVMbDPCc-%)IJaSqR#+;zbgw|o6~f!)bC;c&bQRghK#M+*Rsr2 z%)6A>tk4q;0ZrZ;91yYaX_oMJe}mU2KG^JKL8Hy|gq9DJ&J&*WwLGQzGO8t%S2Kv- zaig5Xjj#Jov*471ySHUQ?xAhLSW&W|?QOwSZbZbD&E9hP?q`n1%~mDG)O%U;`gp^A zIRBzck8*0nGo}3EhJW@2C@2rl)b~;muC)JL{`;^$2?RWm5T0;E-5ZCdPF3*@y7$GP z+5ho#S9r>0`swF3NgSuiIKBR1?pw>uFyZS3>Fc9r_;!rs{TqGlyyLQu5pL7>aeUxa%K^q z=)N1;_VEw+BG~gR9uxH=X-y*hcWD0^;(yO)EiiYANUPVs4w0sRr1YMbh53y0mO&_}R;ya<|)j*HOsyR{i_u-}C z?F4@pc`7;B5ki3&k+J;MVdUL&DdDFrB4(}k$pBpC24}HNw32Y+ThyRgLUT<^mqss2 zOo^Zr7!T;Y1NV^&|GAG=YjV+QjvEq=WbYwD}T}!W5`{x+y#A z#3DB;5e)pI+X_*Jx{dIE0m@clx1Ykg&CF>_45H;pRD?f4V=c}Y>yE%0npfN6_B7=0 z{p28z%MpF6hYwno3zkXOT@qno{JX&ePikC?Pu~39u30uyveF6^=n5$kmN&$)R#01( z2s27O(J`}T|Gtw%=y}JBhGSuucQ8PA>49)ltAJ=h! zF;{^hBm;pPL{WNusLD$PCaA{5L&b^;4yU_|ry-;qT77~IK&E2TM9er799LeeNzSA| z^IJ4*$dKA3#f;8FrtuGZqpTV~Q81sVC>tr?ci&!ssa`tPq#dq^_!kSYIV#}Ka+Os$ zczQi^yeuAS-p7zE38LrF$ovqU;D|WN80x|cL*h@Il<4BQjRe}O5XLj|{glRoUA>+w zn7P@Gn5Toz97%XK`uR}q7MFMvICYJWk71i_<)@J9Ff*<2z&L^h}_ zR3osMtO(HZq^qHEq3fi(>Cy|`;4OnH>o>STo}@6e#;tZM&;fe~m?BT=UF;MTYTmhj+vaA48(_{FVM-K>j7K_be<-JzklYv^`IR)UN8YM_ZcDyB zpRb=zhSLGYNlC}e_2=TLLuphSlgtc+OF=IyQ%qClr6_(+w1UM)t(7;;K@;0)iIY5V z_V9pRA8!ChdbgR2gdySR^SEx|hAa=-!d@(`5E@H^d)y3%V8ziF*wLasZv>j>1u;$2 z?rMX-#N_(C`V#Y4;p83%FB@qaLO1U-z;61PI99JiTcpkl%<}gZtPFgeH-@6)XQGhz zzJ2m(W73y2Di=4u7#hP&I-$;W)H@C^fngPySins(hHWEZDRm!!fX2ZH2r<<&&B(h#kKVX=X)#~+zl*9wsZ0$C zV!0lLkiL~Yq}+KkHq`&H>alJzN`|_Q`P}!hA6WKhEp}#EMxQm%iKRJfK~Rp@(=l>* zmfh{^e)6jq4k}ymx2>wTuCNa7eQO1~+TUl4R6%aR8r4PJh(g?#au6KsIt92%ZL`W5 zdsgilXE1N#7I9AEL;$c?j$7|!QM#oN23t#WX>Zb@#ir2_vY*};z0XNyYM;l?Q+It` z-5a>K|D26^flhiEYN{dO#J6aOg9kC3%rd2EF^k_Z;bqVg#GJ@N3xs8|rw+fOa_|89 z(I~M2z!RowWE)HO!IOu(H4RtK)Y5Zqb|Vcn((Pl z+X|caj9!K>P{?KBut^uGL=BWyCQ5`(WX}HszmXOXk~phd31M5_zkw2j>;A+#1!Roc z!|s2ej*5UrUtTfQ+Q+8!-mEQHjnMSr5&n|=rOd8^kzx4iCdyHrL{fOsD9o9vv$*YO zGXp_-YL)t4&ziD;P=5Hch3t{bxJUdb()W4Xd|Yp4vE-@W&Nj_&xGf69nF1q}L@5gl zJ^ljjz@9INY4TTZn zvsFXA82g0at- z3r}?mm0TU}V6T`mJ4*XSq^;r~*%$6ED$`g&d*=H$IQ#M;I?Qk4X&clWO$-gCa%@){ z4HJRzyMzz|O)5A&X)8v$iDP-Vop-ukweB<Igsp^R&ThRbq!V-rE9 z-w>^yCk;5rmviLWR$p)bhF*?Q<1UGy?-R=xUUb%zp%D5j8m9S_o2q@x+~&ueT6&|$ zn#vZJ3|lc5_}Q1ajJcI2@+UX;Ms6m%7F{Lao-4Voy5lppkx4 z{ErU(`5b6S(eR%xz*m%&7-Comi__PY+z?6DB5$iYPEm$+7jyqcCM@}apKsa&ek88^ zbMLE_YUD(5u|94h_-V}tIKplNYMu{wIKsL;?Be3YZ}R*70_Sd?_DwH+xKhH798^l3 z*{Krk+!b4klVv!OSBKAuY*Fs*B%inW10u$n+l8*FB4Z=cSU3L)P>~XVHd<}^*r7p* zN)1R8q!+ekRYlpZ%Q;7bhU`;^|G%Bhng+@NxAr;M+jZ~X`kV7u82SFKpJ~ZS*de;2 zuW=zRU_cJ-j+SFjmjQ z&ID~8^32d@6<_*68a@GZfAZXoFaXSjAjhH52#Xmqk7{O_W*z+IfB?;HJJ|ThZ1oC{ zt_rl6Tz$I$D)g0vuRbbAf@E+vykG?dngDfRqb3&`1H!EoG}Ls-(V*XKt%^uN1?C7ujXUf+}iVufyu?gGN^ z!Z4rE_g~5e0BUP>MrR%eeioK`=y9U&HNmx;LD{Gl$n1QtfHjs;nAqC*FWB#QKsX*32(uex|a16gY9BNWaXMsj;BuG{yCWtqyGB$ORXTg@Xu1MQn-oQKC_8vX9s*ov`OogIujN2=A}=udW6%<<3bf z9^FhdGxIe>S~;?-8g}nmteICfQ8pFnFzDGQD=R)L7YTsN8SVr5iAW=QDx4l9@oa{E zVI*1G0@HfWp&@MY^D>E|%KxHS+=|PVMF4bN2qx7EO`+TGL`1MyzUPmD=bT8u<&0?> z75&8>epuF8$_sXQcN(oC%0+f7jS#uF2JAZ%zTeUes)__&4uOwP?~agxPu;Ao2`s)y zr#qcW^(HGZ0yik9f%l6b$VS-Z)4O-KI<0dzz2r0B%B$zb=e2Jrxy0W@^z{t%&H=&s z0kEyXk$uh7v^iQpX;r+0T*@x`=9swB^^Sd3n42N-DU*9RIm*jt%n!Gllz7=t5vl}|KI6OPqAe_Sxns&zUT3Ix zF9Kf=!9S~6^xQ+0E}WM3J!YUrlQFqmS8DdjckW0DE2>{HJMQyD9HQV0 zT*-Lj{i=Ru>dvv!sBd^?RLKQjgL>k1TK9L&v8vIn^l9>G;_xq?)!Hc!QJ}yjxK-ik z3RhYqpN58#gC9fsX4-TX>M0q0X(yvP5}`BrmEy;BYUtfjND;e#=D5-*6oU{(RbTG? z>2TgoNw%B4D={3UtMwDt%3bVAE2@SE>L%(NPalGgO9H*6n$kHHci1hNxc~#tT>7Xr zLC_o>Ygz5v?pA(^Os^TjDUg^h;>rR>9GkChOz#XOu`3Xl3)q=*3p6r1|-n0ZxOH?o!BNrl+R+)+6XFd{tFYju$tlNlw zoyT41pGOqpl;^9)@}5S*_d^f#o;RX#Rc`Rv6FkW^Se<*ux>O~kZXvAqHS}KA7@rUe zu}m^ypD`lo3%j#<^_TS%J3c!fsoBlv%GY!KL#Fu%LEyPl+a5zvD0G$h&zdE~L42_( zXnSQMpL?$)`N_HZRWSiWAobGF>h&DE(Bo;zX+C^F)S>Q|OuvJ`(`fM3M)48~O5=sk zUJJG6PD757zt0nw`)l5UKqM{{xQ>o~yJhLxa%sY4_6lW$I_Ew1KDse-6b{95K5L(# zBDC${%fn?h<%gwgw+nVzt9x3q6XyTLwfiq3@SlJ7b|?o^&lSEt-wywM%YFppN^^ns zH$RX+K?DyqpJ3jSj4Sco0qYvnf4wa+{b%8zCk;|DYUo9z>NC|=!kJ4@jssQIvIzmk zpazk9P&%}+CPeDU!`8S|GB#upz1-cMbXA#dVTumySQGRzX$l;t5{B{0vz;V4qok2u z3pKfN2%YFA471l4_xfN1J6KfLDv;Ps8!7p+55~Sk6K4>>odWry2iDXD<9$BUr8sQk zc(TV2uWOqnK&mO4ys-vuHV*x-t1H|YN#O^sv8DwL- zkjC)=RzAuxYS%xNCem$q+z+#oBl$FNhl1mg$2iQfMLS*t2OMD7LrF} zM0d}&V|_kL=3O1#=F2vvDm~EvL-jS72`o+_#NedpNLbF%yG@|{uH1p}#p)h~w*{)T z{yvDl&2HHZ27uCH+3d{3?o$A!k?un%`gd5pV5?bx zszDv~N@k>+;;EY^D0>lVQ#v>ox&qrVBBiBjOs(F zt(W~8Q|gl-p8!LS&kY67xcj1=3gg60BUZR)xz_yrsdMS=`3&{4*J3hy(L0~9W`NB| z*9r;nh;XR}$8oik5bGgfAN@IGWWjLuQe$*&oI2UplYuL*T)+3SZI%J$c8E~@Gb=}R zMNo|Bj)qQFOpC69ql)r<6Rk}$JrYPVwvxer8)uw`B z0G}%QQX5Oev?t%(-%!Nu*E|1XGgOGWah3UCdNcTer*~_vQNG*(w){Q`jYj0Lo#h=^)P)&oI8s{ zIq7+z-+gCMqYRXbvoeb(I`y}p+0CD%mA+EMPtC6x7^FxqOTwv={j`apWD4uM2v?7V z@8X@h&tX1Rk4n3f1nWsg0_?PUxcA&CdxiHEmS_tHJC>F~!kOQ}LUjxXRhL zre&wCJFj@@(f*Lz5?z>|m3+aZRr;c~f((fzEnPmw1p*ptT-nd-N`9HW<_$KAm zw>a)0!CNn3gT;VPBjx3pf~&Imoez_hLAV(18KK@XrmH|FBtmXss)<^piTg3$s6po+ zs~HPjKa^#=UL4*&d(CBvoG#6!ly|&rHCnmXRR?-|oozw&geo5id+OrpAoofg0N4O?Y=l5CkixO*lKgkvPy+M#XS0x%CsoQaMNb3#(vt6hM3pl zBnnnLJ#Gu^RJ%Fb&gP)mzS4ib=Uiu5a|8M|igqC!lqpg3iW=|hy8YSn3jC)?wc7`- z{g_R=X)MdEjQg)O*&rxy^Sha!xwIGx##txyE|e-8@ivPngm+g#ZW}E*@}Ck&-u&CV!PSp+GFrnKLM?b z0|E!Ej);AG_G9i#p8(&5xNE}rM4ij@ zO3))Wa$^%;-N{zmCn3;Sw$FFZj+|?{RiX0D2bu`iQ=Tj4b)q)g%NB*WRWK3#&Zd?5 zpnHcbu8}EPmOmlIF|Hr>cn2>^2c<=bFI}y5icCp13}IY@q>EinnqjVeQZAG zzDJh()VWak#e#iy`y=BbQ;(q{ay~H23}*05RtJLzq`N7 z1_4w9jt?o5%(4JZ(4Km*R7hd4t|Gg08fX?Q^>7zZCRty76c-Bjkjmh;F*emadTeU5WN9;qjwAMd#z%p3Bi3=9}=?FWDxqv)I=olPU=$iJ@2F2HCH zAgS+}*KK4%yRooYlD-Y?KkZI2*9Xp#-@PB{8qA?CjB7toL4X|wVC&}Q(XvihnQTZs z-X$d$)a-0TT0YI7kmA2(y3NXiV6#bj9^*%N0v8ckVvAS3A=!#2=!g24glEHQ%42fh zzRw$JTZm-S1Q5V8lkxXtgUdq_2|`S3NMixEqa9ogM-f>ZK;n7qrJiL5{y$+_%>1NG zU24h6RvnU|l}iFk$m;rS(Z^ueD?fgVmigf`0fwe5=N6=sFGi~aT>tx{3?M!T7Au4x zDm^P9m9MNzR)SOXWQc;}!ZZWVXx9N$Fw2JutVA+obrQRR0X%leT05MqvS|%SXz0j0 zURCB#UbfZhr{Tx2n`rVkK}aF$9ain09$gp*h#SNT)k&HVUuDc?=ThN+A#&w?$~#b; z9bd)2{*=3PMKYoVb7M-M1re1W%Gv_0p<@*ij@A8BR^%fQt$)!*LNY9DCspXOM8I)< zwWRBay_J4_ARK3fl2G;_g|w87;Nga*susx+Z*zhiQwc7P?x6-I_p-x1E6M2b#S(Ds zC+W9M-$iHie1&69Ldsl7*Om@77P&2<1}?QWY7QRh-9n$yK@iiJVyaRzX{s7F?X{y=@gIpyE8q-08(7V5B5v^xZID_ME<|i`H_D&w+&Cymi&D zIFex*`MMR4o#w!;$3)=T~(Gy+%Dd1nz@O>odHkptr-%4rEbe82F~$^S;;6 zVDl4m%k;570xOQA|7zR8yCht3I@Z~>V%4a_k@Zgk{7 zY@RO<0Ko15Ju9*ul5=Yp{qY9Ktz@(ml|(kj#eQSyw~bNIRoD9u*2i@}clo#Z^_b!L z^Yu8u(}nD6UCMV^zjnb!#3!IgtdjIur74}BJWr_8QsS}mxsByc;PCa;!U^+E49RQ! z%WaO2k)fxbp(=viLsbJ>=S;3g+wI8csRZuDN1=_>X3Buk_p+d^A49n2P+a=dTZ{iy zX~;~|(V)vwb8aHrYFDMV*`l%-sXfe5@Eo`hm64NGguHGET=%MIZKR`D5B8^sABrGNo2g9S1f(Jsw|lwq^-jMGB>8XEcU*VNOMSOp9^qf0 zsdghds&dySlP-JGV~{=)u@NbY`&XP_q$L(@>F6?i(H7wwMpw(V%x_=liI&%9P*#p@ zrrB$6^GV!o?aOKn0YhN;n20?(riCNQewX+Pol@rJK>|__(qQi+5K8^krgl?c1gO$d zmMqiCRzlNk|CHaEcXr;2K@Egq^7Wq-sb*in?qk4u-oGyXO&fb)@#xM$mVHWnDtw=p z3~?-D>l|dv92r?!-)foqy(;)adHKVF6=ccK%d|NNw*%gkM6M$080T<%4%W5wz2(-bb^@D~tDgz3 z-GT%sD)0P4Zm|E2-u}OT&r@LdD5%#fIu3mPQ&+F4q0HpV5QwPrn+FiSPyY-E{b?|Q zOScMh>2jhhFZ9#`EZOOB5j`aNEL=V23Ry6x8_66t6R9i%ZU9xmg{o8}M>250#sW~k zbuJ^U2zN@yLc)ai@3&i`k{EXdNQF*k+;zNUA1RDZcB&*T)4IgviZqgYmi)RZG&}iR z%eibnt(DXo7$ATXX4}imNII+b739kPv=msEt?{?+Jx*TWiIz3*VK(3>olW=(8Im=a z*^r9!%2ltlA;MU@wr;nUmQ1T+P|Ude5F*uD?{+a5@Wu+cZ!fW$YtaCa)1$5WCD=6k z6Td%F@!!8L8_M*exyS%CYqcJAQ7_XU3EDO=x6tgb3GMf9n!2~25Yw9cUjWh)mNJ0A zR5H4r0PYCRm9LiN@SnF5a91p;$rtePzNWGUkS>1{WFD4m>wx+NClf{N#^(C|GcRD~ zN#BUAwJl1-W*vS*rCzF#JZeK8qo+;5h##=r%UO09=rk5*ME7bh#4{k-$pH7hJJ(q< z-pX-0lf};R40MO!#}1ip`imQ&AB0y@>$%j3GUGoKG?GTZFg{wOYJmg@JCk7H!cq{1 z&%GUoS;c_lT{`qItU37#;xe1?v7DBUj&6VQhB{$~6*MU=F;;=nmltSK-@**U4H<9K51yHv4qoSv0 zDu%5(z?Rri8R->+&WH^*Ow#f1{#oj7wAg8iIThq`g|u9pOErmXgEzps!+i138(-+Y zH&0#e(}s-cD;y!6@#kF=P(i?-2N@4#yjdH#ov^8UO3G2b8o`@96#0krMkrxaP4hKK zS5N^LocuL4mKBCwd$ENHU0r}B7t+k}5TNAU(bF>&Rwbm$7aX60rZ zL)>*z@Q-@$Wb?`Cd$xE2Hf@$@F)hz|oSVg+w03__;Lg?gO7akCL4 zfM>j-J!@Ax)g*6?=ze3@RGpa`K@he=IH)7{b~|M9!rabh zz@nP&>4;PVPLpP%3`_CHwdt3nrN_D{Wy$VRSQ-PIO{-Vz->oxFg+7$YdGAA=c5ouc z{MPE!->PU2D7Qme+>+}-W)olo|jea`I3kCm)nnaP^0XT9%r-&Zb<} zWxk*G77zAVmEyOX)UKZ!AYeRY+y$9)Fhcd6`PKv z=a!-gjJB5!tXX3PLh{rcgI#H^GFyN$!a&~V>^R1}Ig#%AvCv`qTW7h?^LIc~`G|j5 z&?ZtNo6q;s}l9`g|dY@MN*sg1AL3E|Gt|CY)xv>C#KSQVm4T9L6Z z8>f=Hd_s2*Qhmj*4o_p@_ZH*x`!WRHuJp%x_unw%s_OqzkLmNnaB1RBq8fGJ=Vqn; zd&%nTx3mS%PYEo*S>%&gN8OK%LMwUs;M+LQH+QS?K;LKeOeEpePu4cqWIkxBqP;%( zmPYZJ7=h`{pPlSFtnJLUUN}oZC$;uP%%kIf>wI^Wtx92s(?71^vgL|^=ZgNG*G2Tf zc~3T)%v8O82V0vV!b*gzD=p}S($z#o>4)^xj>yx;QAVA98oxK=;j6W`3!}UoEw`Z` z{y)IVEM0Uiu?gn=2p^^LHkl8rH>G4MI|?FyoR}6*-|iA(m@Vwd(5`z$52t^9tB@cx zB_@Qt3kqEuiT(07Sh)|f$5lFm>N!RArl(v4XSCjVs~o#MTl`$x@q1FC+e~xDyQnh+%AM*72B)rl2m>e?1p|tP@)a7;F-Q zZ_NAPw6NToH@4XfZHnj#g>_O^r{)X6tEN$2^0uD-3($h@3r`Q#u_NK3%iG<*8#4qR zvn5*XEvv`U*RzCIH;f=)(smlU@UZKweAULCrN8cP`$L4@(;eF%UM|xKX6{`4sS5rn z<2!Y9ff=;+Qi-Y~s(fNFHzwhCe18l1YC2`%tW=QoNhk_?SrBvAPfe=WP)=#T_oz!AkiEzGOYALN~4 z$zJ_g@>;#~RctVq<5&0mm{;z+Z{<-(ztkznRH_>Tw*n2jiCUoH zXN}}idl=_X2#WW|1#&Nfcg^JNvAV(@_Y_uT*g7$DI@czO%q) z(cE?PX_|emL5D$7-CqI3_&f1?3eYuD;X?He4Muu|lSIgx{P8cenOjoQ$-Hn}EmUUs z18U^Dbtu;T^S)aSJKDVib8p*O16<@{5ljou^f(_>^}4a`KjNIb97M-*ao&f!pzd#M z-Y)*=zp-EO*vwYN?>No+t42;A^rX93XO&#Zb42UjbWZ3?pXC|ASmNCZWL2W``2|3{ z(pvCADVbQ#J#0)?ZC|(*J9QGO3G!lU02ZW0LDi7^PsY3gqR3VM!jP4IPyhmNU7w^z zZh|OF_rg(o7FDhDV#_l0K#4uat|l7KD`$Vimc_s1QMtruw%kH~3x6F|)g92s-{21*7X4W~={P>M|SG7I}Q8T4p`8Uf}Mg*I)EfQ9)h{LR4<+1yja_@C2 z99bO_wy=`n8B7B$oUHC=;Lc~!m7Avm|6_3j3;FL3Ij-7R23|fM(kXkNO%88hr8)E$ z0uKIaXmA_vjU}NdMlH#fx1!RHP8vCq4%1`CSIq6x)^l@U0-UVBF?tgyH~p*YU}L_- zgM#Jt!w8uZ+5s<*jK9ogp+upzwq9fevLzC5{Gh95%(Cy%j#d95=WSfo#eCaZPusp7 zNFNY|9PYc6MHG*ln-j199*BrY9+s#+%)9V1Qa~Z>+mtMEpGwBJEjsmf-5X>bb2lFT zr7DncQYb<;GQsbu_=8z`uqmymmOlK4uVf}Jm=%XKRMFol#0yw-HRPg>{LRMpk0ny= zTc(6|Sd}rAx;@beGW5~uxGWJCU2sgAvONyxF^@3v{^aPUF1i6!ZA2mW^_oO7dsX97 z?h7aZ3;x7~OQ(6>lzQdRei9FlS(_)A2R|Eqy9ntzyNwpFs(B(bPyL-JsnekgzvY)U zU{AI>C*h^}y@@h=InUoq;c-r*=x&5&?x(H~H(;c?%SFA9h={&lWLTg(iY?zr&n`gw>OmkVBl^p z5|_&R1dqAGH!Bvey1lHNhuo&<*0a*^K2>#S1ZriWFyb|tSuX(dBmjFdCvkeFfIj?n zRLJ^%0Gk-`0~7g=t>Ym02KHn4f2@aM|IKH2Ad?{61Q)_58irC+)ZIqPcSt(H9sd>+ zcbwW@`?St~xu&#}$M0~~4hmUelT~tUkus**(blMM{+SFwyMYV zxA8w+r4JZh1`mtb3B-h{L)b}tZ5830zo-L$H?Fyom_r%cgp?TXz&L$nq@_~(YBMMc zT^1$}2#_yn43Magt})4m^tjF=BiI8zRn{XX7yvxpM_fzSuT5NN9n~p8vyA_WBv55y z{Mn(n7xE_g7z4LKX0SEt>5Soak>e|@g(6#{GJ4N_G})tJ?-zz(lMY5Ghh#MsALU#g z<98I1R!-}M4McksFetzVN(>+;SrWss$`Q73WAf>OpXuKa0Ty!=(0+0BXJBAL>odtG z^a;y+BJ!T4Q6c!r6$|b2E52Fdh|##hx(h_8g219Oq9~mdN-K>|d0__UcF0*rEJrrf`CgDn(($uM^?v{(O6OV}dwQ0S9v(llhoq3|TP05jZwR zMXotY4}$XOvHHhBro@d{PZ&|8{&jSn<2FjhzyBX>mDAJMVB)gWL(6epx_YRC z3ei#-e1Aobt<=M71#x5W^M#`a0?SF~cP^2#2~1NH)^gNscGl0sT`@@HxQK){i%BbW zT|r>Pw&j8Hul-f$s!~)pG6kQ>aco1^ID-t1+*@_CU9N{^M@3jR|2=ya0x%lGw~JIY zv?J;lQ(Pi*Tp zftNIHY%jg8LGR5R<$JT?$8z4)tu8men0{7$un?|SJtr?eU;O;u7FMd>=H0qkHXhq) zw2t5A{9e64bk2D%^L^Tz<4>rwdA9x(>+BMXh%l(o83hE!^&jD3Er+VnGALCrBOnY^ zmKt>0l8p;7UKTA`o@oLcxnz>c1l|o0YQ~326veSnEQWn$(pwHuwgqgPZI5{rm14a{I{VH_@)UwP9)bLHV-irdJuf z`cujEVg#iWz>li#(9|CRMWTKHnQK%+R$c9tnsXx<^!8*vFc!U4RizD->biAr70_C7 zF-fLl%`_jqpN!0;d?a@`bgXL3ng`#7I!{JIrJ5?wT?_g{JgG~%43n=yU#6Vy;^5H; zebJDt9j}qEDMDH)SPG|RhXweFc#tx4na|D;i#Yx>3gk>;6Ls$}FTF&T*T9*#N&(P( z+nBM}zb)IZ`pI<5ScxD?RL!)#nwFJrAzHES62uPf_AxRd_F(%xMj+C}oUBF?n>^=| zgv&aQW>PSfcu56m;)j=;IoG32d)~HddM+vF9yj@FC0kxz=`<2h2E1>ssWA?07ob7l zRdJcO3fj{5zU4YRKTQ;WzPCY!nm700?__8Qj?k!<;#q6n)&p_Xg6?0JJm?rtQ;Qxe z9&|3ehNE5Ep?pz*NNSzNF^4h|kj<6V{frB@)!<)hcho&~V}Gch8WYKeC@|MEnrXbZ z(xTLdTh#{#04^92YT~C7HfxXgJtnP9dU3%^5{ScD4dohDBSRchW4>QY!>yOZQF{h~ zfDN1n6IhY^p_?S88{(G&-Cd=w-0;^>g4Pa$hoq5h*uXTBoL*}{>}WPl|(u>@#e{3dQH2WFON ziQy^w_wTE5mGM!=nW3!QX-Q)eI-eY(0$Ql?_nF7}mKz$|$m4H`EO)Gr6>p>rPI#Vf zJ$QM{M`s;1x#b=LZTqQGu)a%-(^gy2CZN6$kVolHTcgsVUbhDMJFWFyHxDj>qf`{6 zR~_(|VgGDG*yQo(D|($YOyD5qMB8IBXh&@vDt`zb|56b-IfVtCW7!-7{7Svrf442)Ok$9MKa1T(7`fA%% z%ucVXyP_OU>~+uNvaNb6yg$bcaZudRKTHkqw$6O{LXd+0zQ&`V$VY=a-kIU!wM#Ac zL^SI+quv{jc%v-B#k}zqM~V(9U(->y4`u zZS-_j`fg#KEy*u=@9{QR6bqLk;fZ&%WzYK`jFEnsfcym!4RXdiLI z#}~WkE$}ZYjVZpVf`!!btHADzMEzHm;oZ$_Pa+4wcbn#zql#Pr!prQj`(5mPqDjZO z_nxyTNw+W+xkw-Bm+j}U@UBbs_{UYp&qozr0bn+FJG%O1-mN7Wu4bCFMm4~z#Q%K5 z|CXKquj6LQtq6bJNea_X=CuC;0ebWV<4&Yyxv)~1M&$9BTS~6XAMy=?M09PM^WiBV z$OYb_4YnGS5TqitwE)jNI^UiO6YnY9uV0SnZ}F`g`E@+ z-f58w6R!9;7$^&iEdzys&rsj%hi%fhI6ucf2qH}Q{GvB3C*C{{MY3dF!%r(dLRCU5 zX`%qlMW0fF_?%zTP`r&jX=^Edq*_kW%8W2d4+*0We(4|*Be{`=X8F@GWP+l6zfB-s zL_(Sr9)*;T*8vabytw#1c^*EwE)JfaT7ZE4!@jf%18osTC%NNBEc*9!+iYtv)g*A) z6%!FUjH*F6w#R^UQ(HJdB@tq{{oKzzF{V0L*{0K1(c>nNmV%_3Thc~w2Zav^@C+TM zN?>ZjHbpH&g!;C1`K2b^&3`D#PgiK^*#~jCH+z!pwjhQL5kOCxRl zK{6VakZfra)nne^H_%1_N^4wS2N7#qXs~75NAB26FUiNkHcAlPtHHO_;Al9U5xA2R z^1`_6*7HhsqED9{(r-J;!jTM|%Sdyfavg zDgDU}BDj79HMRDR>x(ro-h0q)*9iqRDBQqUbE$g`KN$5Xt=Wnwhecghe`Lx}1$)v< za5fjR1L)HPx|7&WE6ph>1Dk);+Agu9PMR9KcFn7gib-RIx1gY->p%|$vlFiL*1_1tE2BnhX+-Sn2v;)4a2Ax*d^E zWN&_yb+L8pny*zf0c2$Na~gh3h)szCT_=04;R!#g$x^2#^Uj7ISp}c>_eNC~jMh}2b-88C*<)@;PzF4n4A(4xn&9yKym9FK<*;hr21S z4Y~WTvN@J`;A>Mvnauv{D!^EZN`XEF^M?tEj1Twc@R5t@yh!{>hZdrG?- z9NqHS^17y0i+*Mp`z3Dd5xHHpCagQG64#H~1Ue&<^e(-0ea(RusBM%6Lfc{kt6#_c zPnWfZ?cU{t$Drc1gsvQt*8QDq_5k*?A(@d(m)$p;BGuKv=jUfY3E%c-s21F+y(U<2 zy6~|g&ovpf<6)Oxt$_R=HHAdS4mZm95_Y?gCg#F#y4ww($!m zqvo5u6L~hdN~$wSq<(%JWKCJNXm=eu_M<%L0N^M<4d}_{{k}JQI^o{RT=9tNhLccw z3LZA}p7k$N6}p)#8kGjaKa@pHi?b&${tZGJ{Oz&?BW2ryqK1=*@%UWu$BmQf`{&oC zbj+t~CAy~q2sm40*wa@2XCu?YK3e0kmC*IU%G}o(Ntir!1hUR#@u$Rv!uC>FQ|No+ zo^{#Dc_7g}9<}3@(aBCDACB^jRDf<`HbxR>oylP&_jI$h!Os*A7@{AtQHtiMlw@k8 z08Weu7o&aYn3I7Am6hcrDF@x}Xd8Ac>{WVDjFD_P>^e`VQM$Yy$}owrg)^eT2aS~| zFHp1UY=!|gAOKf>*K!HUl5lxrP!eTlLYoHz(^E-f;FC$;QxDe(KID4yE(UEQzZialmK6Xd4bGr(y{_xzg< zTt?F85zfx#BSN{o*c822_cAPW8GZ-%Fh3BaJIEm%_p;%uVQI(PwO7*|?%?J7LQV%r zxh1|5^$CqgPIE_Wo_V@yNXmKmg|gu!qCSe$N6=FEZP;sPM_j>+T}+Las}-tefkt9A z)<}d4bv;1Phl9Qz4vwueHawi-Vr?Uqjc67uZcZ5fLzlvoc={lU$$Z=gbjpn&^+CKP z)yEwFIGvUK&~Cpc??i`KW3MEF>``p;)A4w4S~ma(yVzBOjYPQu&KVzy6K;O0&e%tj zyIow?+GN7G=t~Ft*@T^8rO>bWEoWta=aE5Be#czAZ-dp#_)5PUfuJUTFHx-hu5d-H zeLKLy;lOro-CyiW7-Nf3ry9OU00d`2UQP>yYz)#X8G-ZErbB@#3)xEacdMOB>#t8U zh&!z!_kD20GrIeOUy{{jA$7$dpB~1vyo*SH!%I;e*7}bKTF8k`EVvm6NS)#v%d#c| zhT3g;51o>0aq=&_mn*lw{P=0t==>yq^>pWix;1OifqdM$*p_^0VTGx=<*5$UmxNl|Lyg0)1-eNOqh(^v)*V zj`f95KM;u%{(6;A2(-)d^o>&dweE7BcQ&QKy9!!|7&ZTVn&ny+`OaqsQpY-8`X9jD z=G`>}J&~19qi5xRci=5Ux8I6jVDa3(Ud3TcQYQHEMuz8R?>qeV)^(h!kTxkZ?&e2Q zwC0dKw{^QBmrcb-8^`tYl6%{l+BaSf4qCWWI1jg*7*FjMB7cr?5yE|lYc-pi+kL2n z$HMV@sR|PC81v}L9Bf}V12*_Ry>$wzY+O%nY@rgF%+IJ-$qTS#uVkf$?)uNy5&HUz zw;8H;unwD^dfTOinmLXT{+dJlZ#w?}@3@YO&J%BNK3;XJ@INquJUT`gheI-r`5qSs z2Aug2f|N~jCVT)02XX0QCrOu!luT3~=LjnZe3?_QJR;i)mo``F&H({*P2Y_XeJ}ZF zPWyN3|IT?oB-_z$8%+faqrmf9_6Jjbm!=;_cElND3_X{L5dIW55+C%f8ZLkKPsnng z^h5Hq>2i!Ky=88A^sz*crB%W_v^P{-$hc z;u+34W*3=G)8LeN45>TB7w|Fa@!j84jtDg99ZHJ*!UeLDw6mZVshgkYMbQJY$(ZMy z#XgmI^!YViAjXh$X+a}=&`x|3U^>3-Yp(nQB=enJRExWJ(9R_MDHQ0E6UjyWEVM=( z2Ut|7Wd~>42=b$k(lZ*VUxICMEv_|G*LWO(Y*e=B|I$%Y|5e0(`o`ZB2>!H)WTH;+ zYO;`iJ6LMKQ2$eb{I6i1D}6}EznlOOlcyoo!m{?NZRlhEByjvWbxiC)HuvaLlPb8D z+;ILva~4jnK=KFogynuHJEu-ss!QYtftHR&zDc%-P75I8&0qeP&V+aqT6hui*SD3) zPz{a*R0_tQTKG*`Ju-hynBZQbk#Ju;h}-`8B~Lg5Qcg3|q;LEjP`gtOnIzo^wUaYg zm}%%WtGXx1$oLpYhy`oxyj9){#TlA`1UIj#4)+A_U>CNN0qM(Ed`AnN7*eu}5 z#Xh?G==CmQE~w`y?=92U^}vDj8g`0y$OutT6ZQAOD?EnZ&(!F0zJG)&f}NVbbE%fK z(yuH|EVC9s%K>I#u-!Q9>D0)L$Qm_Mel5WeAvM5l=Cg}+>J2K3^__>_7OG&1g0U06 zxu})xUp>st6z5;Iq_>D-xEd34=gos*cnZBqB$vcVaGuUQYltPbkozho3bd^5oj5)` zz(jlim&S3BkIlVAIy`XJw);yy^Vn5cyD1{1aP$qA@l)hVjT7c>xcJNray{CeS{ptNxH%^lDd7zBLLRD_Ro)6- zk2RY%4d%`}r*zLdM10Q+2i}KEmIe*+Q>^UKdAq;`6Tk=j2M6)m>XT^wg;OA>M*m%( zmmz#chz+^C@)GO1WQElVZ(l=`V7~G7l7M8K)8Th)TVj{yBUf!owZZta3G&!tu&fF@ zfJ;$}GK9k#FOfH2c!f4kIOWqZN+tQi0RzQ^rFx)=o8cqbhb@D4Ph`K!Wxj5}O&l46#%xEf zkcCtCPO>s7hZLk`f865%&>SQO1{4ix--q%iwV-BY@@BQ{&ie<_N@8kt271YgfDpE? zh*Yag3f92ktB0p8D}_B(D;a0zP#+el?S0U!^Zh}89Iq$MJWXx24+u@@w4$b@=WP$a zR_Iz{G?$M`#qVV}fsto{Uu}{4&JHR|;O?RLL;?=GmXcx82{)izes!789!-q(3` zs<3C(T>*M|eS@p;RV+oR2ifi@#8i75r)X;{Akx&h-;E^ZlQrCe0m(Nr%tY#Ewff21 z?%F!~d8|sT=@)M@=GI)p7yd&5 zw}1RiW2&FF)#QZkXM|{5f1TdTk~+GbmOZ_?Ye-VfpT8bfOz1kiv^HtRcfZ$zACJxz zM;^A^ECgE0|M*;BN|a6_rQMdQctkBT3RJW0wsMe)e8>eB;FO^;k%tR|%%r{F1mR8Q zJMnV@gO?MtxAGZtT6vBBt?d=G_q5+8DHQ(~JMsTjY+by0hz#;|_)$%J{dbX{yM_P= z8NLWjjZ7E{iXX+2_A^A?ra2cLGu^ioE%xV+>ZN34vbah8fI4zLR}2y>b!2zLX|xya zK!l3H$RHck-eKy1AuGU8z!rB_jm`4Cy?pVq1hs z^*$_#6jdVO!96*!MPKMW=ILF+#Sdk`mIio8)%Rt&E56_JN-~5Tnm=mx=JVs0A$>!7 zj0tD)6ZuBUK2H)}9LQ{^^0D;g$ui+C*h1|wkH9&&P`qyf7R3=1P|kTspkB4hMgV3; z?c^~WaDjSO0HesV*od)(J#q|;pm)wq3MCl`noz<+N(Kjls>vs0l4VyO{2}jg4S`mu zB8*+;X4+g1s|29NM73||lS4Xv0r)IC#T(9(PF*FD zTla6LAHxB6{CAj6s~b7|+h~q7T;DuXS*dLCkGMLin1MAcPE(pWjd*Kj%NZH}3c|T{ z1rS6M)$hj>USqMQqW^(2!B0f6O8^9FswzHKPKv+B(ws_gFM6Vg-b26^rb|hyiKim= z(0NW9`|ux?NDG)E3#%N*#AVG{F8ufPYNNv4%%h=RUwr+lVgBS8ej5Az)XFrl4*y=G zJMAuS0%n@OL`KJNYx{+84g{YEVL4{6V9U;jRa%#B$2n4}>D!Y#MUO!m#JO+*ZE zWHxI26PiG6W$nO;EHPdUpaFA({UUt;^0^O)pNOofd?z5V*za`iui&K>^u!5EOyi*t zAzck}&P0XRiAR5-%6Zl$1Z*J- z!X*xu*&7@J(<=#?r-oP=yqGFuTY;tV3JhAB~ zgW>TUmt^By=GVCp9NJ@D-_CP&wC6?h`{6P5Sk6Fb%o7PeEMjV5YX4ep_)Wj4M%Gu9 z-{@ou=(%-@?N1ZJV|}sj>QBOty!g*A!nF(6iiI`gwUs!}ZbTwK9Loijq#0&hx>RLA z^(}mqKE=gU0=290b-mKoN!Z*ABBe8DJ8p%2Y(ei4Ah`}W(_L!X#9DiajBt6ciAw5w z!_V>~CG$RbA2Y@*XSQ#}*JCkSDA?&ZIn`_Ck z8}Y|`*v}amS-{3eVw^S=;QS@lY7QmbFp9t}S+mj;A8RMi8G8WwaK-#$aiU=VeS6tTe&Rh{Wicle@*X}u*)>`4kE{aL`m z_b=MAGJN$^y$*6@tnkItbr8u*eeee^(I>Lc!wV7ewG*Zm&)|WWp`G4M*r4b<$(5wo zeG^8T{tCmiEV>jh<{GzcsEJiJlU%?5rE?EPg8xgdjf zDvQj&9??>$Bn4=xqDN3F2DYho*8n@huvqdO6QPjY4dD^bTA)A<9wk@4kY4wbd=j$J zzIpe<^I4xXI?k-yh&xsBwCjjQqWD(1a;qGb*1o1jb>0L#pmnHv#MwgZDfK`FJK^0P zHFo_`u?UElW}4sm`BJ9bY!nBdxibJE&Z^O{*Rqsu+WgR!ir^*ORY?ay)tLIileNOqzdE~mopU3F!!{K_$+YUHf>?~Oo(ALMF zpl7dw(Uxb|?JKF&d9lzEu>NVBYw6)meZdu28Rsbe^w(eL z1&Lbh7vC|>S=XJ%4u9QWH##igp;|1TZAZgPiX;EpgT20-dt;dFeHBab^1Z0-xpP9F zWvsw8+IYSHoAmRd%M&nz`ahikwvygj4txjbV^w0GnV^Rng_YBNe^vB63ED7Q zRxwyn_n=9<7E^Lid`%Y6%MZG?pyd6ZcW&W~VMGvxyR?SY$`5Yi zkiiq-zGrr-CTbif!s;*%Kw3olRSEk5RWlHAt@X!N#j?@A$68;%`)TRvhBB8{)lr+3 z;bmVEHS5^S~OwugK5R69D>wZ^0=0-YNZ2@wiN( zDnIoRIMBpYYAlICrP#LNN(Rf*djEM?qgaesR3miXbr8t}iqmWJ*adu-e}W2Y3wg4- z9LtyxxMkwKD0-g^iPK_bxGK$GTkha#q)YL0(>BuYyqi2g#;j#a#_766U>8CF`xV2&Xo<1RRoVJF}5TF`_CYxcMeUx z-e$xej?W@Au=UEj^&z)o2s$Ac1Ko);!xn%d81&Ag2AJ^DWf_2rv%8e3>ArRzjT*Qa zngx;FiN9>nQ&jyd)3OXn{S^9(EJhRDbYetBMPNb7xD6W+`6&|hW_yt-@6$mMMS0z} zFldFp34dFKz5srr^SvGqC#H^g%SXD zwlh&t2;5~S?p3s{W{e3}Ui@W!(Skf3S>Pwg&Iu6VM&9gS3W#$hISN^MOkxdEW-qei z5dn3X@RLO$FD%;x3}+9^K7y0YiWO60%gSJsI)|Pvb&S?`9gAQk>nmv?>3K*~W>A$* zkv0)=9xnU)fD=5)X&Qh4;@jC^zy-_u%8)|R=%_wgI7C)lS=8{)S-y4t{mrgVOr`6$5m9TeaWr$2`pxU! z;Q8Daa_UYC89X1$>f~!KTI0_}Tw^_r3vgPU;TLQAj8Xv0?lFYR+5)LTapcfo3@vj6 znaq(cXwyNm*(AKDDS#HCv{w!pk`+A%dtlm0ks_nqh>?g5M;IFjCXO9b7Q3wM5C?vO z+rvYG1O0qBh@~X7Zu;|6ANAca+Dgx41UeQo7aPfqp)j~4Aa$dS_O+87{L#W%v<|+7 zr6TIv}fv+aJ{)m*( zE-cabv$0VT9L{l#F|q}(KrQWPtfAaJu*q|AU9iNAG*vM=KjrCWr$6tx9^`=>L^!SVHg7s5wUH>rH`2kxXE?Q|4w#VihEK->{r&}I@x(q zK}3X?6wFaP?MZC3f<)RUq)eTsv+t`9O~MNet7AevWNjixk3E1FejiMo6nI?!eqg%U z9|SVOtFsEI?cK^{jcCdt5`|dLgSGMBipfp^tnDSiI$f+LYaW_?#I*fHPx;}fad$Z>aMknnjrqS zLD}l1#L}0y^S6|>?zJjbJF{Hv1Hbp-?gzql;mRwa+aa)5#og4~VfBZ4W944g?{4B6 zVrw=_z4GzHG}BL5euPfOe>a&d{BM=OLQy0aG|&le#@*ja{ya@&sG4Q(<7e$OIDeWt zVVzMvGD;)+?K5@r;VQB{p16+6HdH%GU2mwD5xs_+7-u??Q{_)XEJJ^z386j^!ocvyXuLn@^z{1+6mNiASN0ho$tfF4)gCxZ zW}kC7)T?6n2i{;p-L&Y4Dxd=Tw^sSTe3*dnzv=VmXDr^*o3bZZl#6HzrwZ4;FQ%^q zO8c-d$9ND-(@*p)sd`hRb50(m;!L$-59lC@XoDZJ!A_Y_&s(9>Jz<@{pd?}{CmL^U z#N;cguq!Zxtux&5ah;(hYpIgKvL=x92Stv2A!j_kHs))21N@V9!5Z@#yp)i zjdF2tagw7x9D5pm?{tIA7h3IY=!Ao{M;;Dcgx;6K{7<#32bRJ|cz2L6IK%j)`=U?W zQ1*so0C)at$)A|`1l^{|q#r7+Dr|)L%nUl zYk)*V$N~-{RwBRecp^=QnS1w7rpKtbKh37NukCg5lZ@aFjKh@|^zbsr38*}W)}m0_^v5np%C2x#nl+=WOOJ9>+CBce zaBOV!Vog!yBri|c`N}T?JUIld3>t zGpFvmzoxS`)5>%#)2RL25<(>W2|^1cwlay3h{;!^B)V|CVwucZ{&z}S5B!n6eE<2f zNuxMgUY(obG8-z?k)sr!Y=7qZT_n}FHeY=<{&Ropfeno{C8@`S5U&S3gs+)-&Es@%kiD<0yWE6A7S;%xEgOo5xDjjb7}6v>#9y#!ijA0m#(0w znQT4Z+IIeD3axWuC?9V-uf zm7LvyL=3A65gGcoemrR2?H|-t+ak$%w8#Bl6gd0Z%d4{za$kB$$7H*&`5hmRyIGDm z9vJ2-oBv+i?e>$F=Y(GE#uj<*#R5*UL-hgZRZ#0I&STxdm$%j^Cg#pDDmE_jjdu@2 zlQ(dF-lgLIgfCW1PAai4{W&=W_;(lTtq@ftW1)zBrVkUfT5ks zH&HeqoY(IS4%(S#Atp+Y*6Ub-ASMiecWn{=R*i-@S-ybS$h@z@2UFld4RPbv3KP%q zEHcZPp(88x&sE(w9au8gPe-+3j8D{B?VPL!S9W~BIUVz@IYB0JyUjH9ym!c`zd5E50?c1isq^Jj?V`~*ONn6GAIG6E|PMl$PM$G}0Zt0U~%0;_DGzH)8ps6z2Y1ZubKwPJVLTLMi- z%lWG3Y)BU)!^^fNM}tOt<%-(;-?u8VwD&3SZ6Q~?F|JMLe^o4;2g38PS;Ca%zv|=1&-Qn^qtE%et}Iqa z-_exUy>&{IR`X1Gt%p+h8n{ZE$Mqalzg#;UlDd#QVHoK2eVbMEHd0V$D>(g=UXYaXaQ0YkV+{)3uek22jKJD)j*g1FN;tAsezhV`JKMLw^+b)_m{Uf_+?}} ze>>Ibk``dMDmm)>Cmo`-I{k=jA3BEJA0NSg3jekHNEq2;5bw(9KcBz^UgyxxWyo^y zdCYUQv?WDYH`;+Jm=zS{IbWx`Wi&t~HcQRYJI($xHhVMAbeg3xjj*VadwGika?R=Z z!tBK~CkdT?H=a$*-D+Bk6D{jx{*bXNB!WN=Q*}PX*pO{HdrDAuO2(s=j(%j*TXh`n zZ{Yrp#_G(PlCTrBt&$x2)Q}8?^cNOqOutC8>cGAbYdEKg|^nb@wWq54Dn7JY@-vj*OF=N5f* zLW#9=sZ^=Y@>m6l!~!3|_Nv%(5oM zJzz1Ddq+CXmBxtVj%(PVIXjAXw`nP47vaMLWD1W#6y-)c;|NAMD$}b^B|g^K>u=28 z{|gO@wIofQ=cbhv}S z1XnlwXE14X1Ud%E^7*IKw47-4+DeFn-+Y^l3xn3J1t>8k)99mU z8Yi4Aju`UCnzXgYF2J(n%+YHX&CheDnzH7o5oJ~dk(7aukR%1$n;S(<>CBGi9fX)- z^V+Nv$(rY9v?vsY%eFIE$^d7KQ-D_%f}3c%HL zZ?9;FKIBd`KDn#`DB>`N@Y4}LSuI&<&KbF^rOC4lIV(iz&0p?UfgZhJb9BozD^5Jl zfNb!hM2}hKQkIOR-||>h8mAN;{NHAQUt{;GfkmG5OtFp_VJ04vqVnhP!aiEQLGO0D z9oFSt%i(QrJgiphdSe1=T{+SbY8|@p`|=n-n$Fk*g{c&9OR!1;1%xCq6$+lbVn+PGcpoq9wBs z#39Y4SP=2IZqt|nC!>QcO51BJjU3boI7sr)!YLlxz5!#;hAJ3v2R*mv)Rm2Fz)`J- z^XCFrOPZGO&?!hiyX>XuFC3f%X{4L=b%Ks|ts6Av18hkNeZ;Ur zi(3cTz7BI%o;VPr$c{)!u}h`hzMGxP2kx|0an`cj3)a`&bt0(+X9IxtSP~J~(x*RI z)MbxyY%M(T72Qwn7EjYivHl;z-a06*DEJzL!3GWP?ykW-XmEGez~JsqAh-l~x4|6- zO>lRY;1b+|FF)J;cI(^S>Z*6E-hb~^-+Q`G_dWfGMkwwQa|SD}j;`NhjO*U5)yHRb zgXK=vR+VT#(*V|iTs&yG^~;(5xo`Ka=lNLLBV(dwIY0~}=$ryM<8jk)s`H}T_grx6 z(O`J(pyBgbyq%C7Q56y5svf0T)z7rFx~I1>9Rcq`)r7pC&GjPUhl$Q9t(~F0QcV8M zlY9(4$S4L0d#{c^LdPrON~FA!2w<(!z`y7d?73A5?OsyZXqFt91-fGvZgqCs8V64g z@8{4QHq9~MA2JN|q?BvI@|XLOquX36SG0H%nM?3g1s>_>bDsiD_6{_>`$zRSdn&hd zt%1n76u&zl=@7cra`k0hEQ{Winx8-<#l^DgtIerz2GzszSqa**x&VassVt2G-P#x|?W?^fg4! z=@xfC3WRLEF{JjqI*wmE?iinUnDo1%J*rP|YZdXKQBC<;qOdMNpU$ht-b|P5Md z8y>7vtI?eOZE~zkoDdhB_BHFc@~tvPnY_!$dj)40WLgO4@lB>05xbXmqA8E zrdBN3f}tmB8*jKo-pjaqF(6LQ3oop61;J)WnGrf@(Uc9259I2`xd#zP`ivf|G&-Y8Pq!<9KB5EJN$y|cLb>S^r^oZdIo{G2Z)<_i7{Bp)L_%sr=6JBl55@VCa`f>8bTxUbK{#g7! zF}8j;yG0jI+k81Ovy32C-XmBU(PqL7hW+p?@r0aYS#I8(xu4S%9N5T|vooYe9)h4U zg`sBvNy9IY8+BP@?o^H>pBGL>3VrF6)ktrKq?0>*&Yd$GxDv-$*u^?44o*i4(u#F+ z>m7eV*w6AIQ@&4DBQLE(z18ndAU<7I3KX3*r1$Wq*4hZm((;o7hGay~n~n=CgyX+W9oMK>y)MuQGo(M|#SuezjnELpP@} zuCainz@9IV5RSDbCz-egE!XK&L2lM-91tbc=1+l))Ki^D>h0U2}z z5>+MN63|4|jh@mTuXK`r3URqnHdV&gn7Y;=`t*+Gw;MpM$&CUfTeSwdM2V^PNIty) z7R=zTH>Cf}U}8l8^1$e=4oW}s?ig7-CO=YtvLeAlcLo_T42?NcIg^9k?i(EUI)gOf zP;y4oC-so!`8B!@wssU)t)ESh&3Q-jpebP?Jhv1GF6$Sty!xjxurp>Y{ zU>fcg$Be%j*R8|+X79O%Cz*POaNv$6$m};ygf(W%TAKC)nrCC$G)@0zG2A5OniMEa z{X5D->YkU4M$dTL&MD;8HDi0OS-#A)eZ|r+KqI98 zZLHqrnCHuPE%*?I)OPsYM|RG8O2RGwV@!y=?m~bkW3)--NOmRWRZC{zp<*$u&7shx z7nSj8T1lft_v^81B$#_CPXxf;Rb{o3)GcXZRf6_8b?e6%myTd}B&iZN=y0m=>1K&9 zP&mdrMBkSNnnwBaafmf`ZedWSBAq~}x-U)N9orew=H{JlkS?m|OB`^gX-XZAZX%`B zUuAoFR0e3}KG%XeH)+u+Sy_R@H(82K9bFMg0sNeql1u8UXiN^|8Q=y{4{h2$aVUm!z)uQNG$sm1Hb$;f6A1Uwo5i=rJF?@1KMrAsrI^75uawzN{FuO?B8#pTnCYs^u*~?x@){ym2>Oql?D73zx8cO+-T0Vwml^g4l6$ zml2I1DiLk3$7Ch4$^Ahmr*w~*k%Kl`R{UUp4aOBY;`$BQf!zKq$KphWh)bh@Pu=W( za3xG?hAuLxR2E*{c1E>kQ8)YYY=vAlo4(Yn;bmE% z@4Zkjf+C?`w@@|%Z}7iVFuIVUDY=%rtj+u=mOqf5ToJPJdRAS1&Z8QB%@{4gOMIJA z-P=kQAa=s2IQ2fH#JqUJ3Goe-$ZWA^z;HWZ-mmk#BLPfK_{!glA^k={7?3P|ivfEZ zAl1$cGNXrOXVR`YWj-iR!{XgA1kr>9SYgwklQ2+paz+^1g?PV}_OJI^fg3{0DqU}aFg`uM?zv?IxxzSS zDYE*q1rXL=Q3r34K+v*{nP7`*v1*<h_Y*4UucG9b%I0#<-r@IiFdEb@uDGt@4Q=m zf)_cx)}nX+Iyk(F+>w};vmATMPO!x){YmH$L?4dnoMP@KCfLk89H2vC{28n7;_DWo zW%JVnRV@scvr#X8lOFMa{u|`Kq5AzJbYcf;2KA{63yV-28;ai>);}sT)c&ly-<4X` z^p+$&3CjY1OcDf%g^0!~gkL^x5e#mcVR4(CH`gdw=fD6*O05d|rYf@n2^<$aN_h{-*_RwP@Ai?QB!z zdp!5k*i6KmZ4JWKs+JIHKVA5LhE4wyKK);-lP}1m>u~jKLu*5z8u~vi*xoi@`cO5L z%_LikJs{%_5i;lUV4MHt(CV@1nTBMYI0l>-QdI|Vr)xl zQN?u|`X;{oL5?*aKqqlP@MV`QJE@k9?Ntc{JEj}&3sS_Pu4(+CxB-D3y$m@VNTY9K zd))+EatcLIqW>&-EI;W{MXKGhuqO78BZcVv;N|WMc7uRt|1^xIWjS;&H~dzNXrb4Q z)*BF4iSYG9a5^@0AJwHcKYt;RMDs%uyolY|BSM6$EXVkae!D{1f>~}lPta-WHVdZb z`8+4S_Ek%(HVs7^t|4P)ghv+F2tk>9zVETg%LNs5m@C2VS3E;~jO4sqDH7T%TNc|_ zo>)GC2WQm`RUu?-x}|+=0J=2V*A|#lqa$K(#FUCfGA@pe+;zx=(t^H^tkbtB+E~7t z7gBaYU1LCz-C0kw4o}jD{>%@bZ4tpKuP((%!>7}oDV>=f|MYbYfos9b@XSQT^hjU7 zNZTU;wJN%Nw+))LmoZ_odPr6jKbXm?a_YRAgg3-ZjH|OkIS#vc+D*^57I4uDV}_ z%gvvv=}=y;n8V+J;?UuGlpXicq&nK2Pkw)+*~yWM zs-T3!7c^^rfITJmavArcPY4nJHBt8VcC@v5q`ESF=U+lh*G&IL^RCAkJZ3?e43{)e z$|N?A(l_Rjcx4mP*)Fojjme{(fe>ctPL+|7rM+wI_ASzT;;!>&mvpMJyah8Nl*Bn* zJKAh34QHbspKp{B_eDsp+=@T8HHy(aOP z!q+9U^2ZMTor<9+u23c>d|0y-ObqOth89p9Lnt{X^bJ3v5ElAY|J4Kg28P(tju7Nj zvx@ODVPUq-iIfQo;gs9xS(V?FwC|?$%qqtnn?#4-AG+#-G`zjXRRRo)%1#wX>+Uy{iprg`NaP!AOYO4lk7 z3DWBCmVLTM$VKEvW$MM=>yk2^%J`1}5`HIO?o7JVRydFULQ|(wg6bFtO(nKAGWe2l z^HkseOxe%^p4ETf^E#_t>;UeZ{+7LEW^6L{Ge{$(o5J|OLCBMR)@?fn5o)5;67$xm z@oq*pbyfFm7Xb&@LpLxSy%1>xU^Zn!%v8KsWLQfdIJs zYtVi$5+I^*xV|iuT-{BH1LLel=dIA>p7AXWa0sX)!;z zn%3dzqszZ9Db})8=gyRytk1+t5t96^{g%lM7pugZDdbt<##|`72?v|~m;kwp~&sk=S7hWh4@d79{S z=mxBYz^w*lYj;_*miMYN!VLvirS10ehGSkFem?v2XsB%RQjM2z_j^HI2j38GJkymo z$2K|}FJAP4(>AU4(g+fBUTi=9^#ORv_I68fz*Mk&!>=WEpe2ILMk4&9k0Ru6C!$vM zCI6Iv8sUmMGK?_~k->;80gEzp(_!NJz441V=o=KFE*?h4_!YN-C z+iMr`Je88E34qa8V+cVzXaS&XmK`4gPr#7a%&Z5Ssf)sY{fHXIUeIwrzG9=1@Ng}? zS6}I>d;cjp;0T$FeWkk07F`vx@~0{vZx6ME9o#<0WM@*26uF+T&W7D$M2)K@b9`hb za1n-WytX=M9M>76eD6cBs+jekME2t=-}8!N`8PHWcSbFwO|g6_@upFL;eb`vl|o z?D&gznb7-UbX^Hjq~0P$*+1s3o+Z17iFxlon0fLBp(C6IIwty3i}OXlB-(_8VQlh9 z#u8V<~wGy=`McXpZ+f~#|tI*cAW3@V}%jyU?^;JO(DcWKt~pR?N$&G zyYwbfcyS`XQj4$F*y8j&ro&)e1)pmB+O|HYr=qQ?AUS)J|;} zO@Z%PLj7iD0;_xN^=u}KUf#0&g91Qys&Q0J0g0sYy*O+~Fv43fi3JsZ?J8NRxow^+ z0iGKGOMms*xijt<-hd-Mh|l%w<(4Q38h?j+o{8j%W{{^}3FNJH&!*=lL;)bl4jE9UhULfYUEtaVs&Y)byFII;)v_Zi{OojXj z=WhkvlOIRYbR;6d%AF@BT3^R95@OgbBJ0EKx+Vo&?P8SZ~Mg!uSJU2C2H3T`i6 zO5~#qIb_N8jE&>X#O<_Ia}|8F4-bLaAf^5*9qd;W4-7;{cCr?xZ^%PRIf63iMyc_v zH@cL%A+S*Q_Z}3a=JPl9k;$M08$kCptnR9%my;NOSgUW`+MNBpaEc<6#{_4_6f@#X z2BbCZv^wsrd2aOG)GxKt66{1mnJ_jufwFFkq2%)D zcW)(pP>J!!mzr6d;CO6wcRL*P)O+ilZ?%_5akKqovh-Rxul~6oC^Km3YLF(+iRLlQ z&k;2psN-pd2}Nf39dE4S?}A0F>vu?lJwpCX zmzR{t*==Z7U3dPHT#dx zGV!l{%T;pi;o)!Qf^ux-zBfN^|2~kWh`X~%&B;>6f(OlCI zXKAr)_@|yhS>|ohD)tQ4tG`gP$iN7oq2{3)!)5?bC=@^nHT;SwW1lY-ANBkIR2sbDpW z_C+X){Ok48WYnn5&(Z+68O;>hXyJ2akp;S(J7?ktiZll63ZsOuUwewyN3|a`N;`S~ zkt8h;7X3ml@4GqDOH!4FvFmq4vAjwVQ{wKd!uyAnlQLe|{$u0%qCSApQS@6EEaO4V zOi>n(uEUCNAi%dQ5WH8LBiG;l@mV8F-orhsmannq!n~NF7hh&!Xf3(l|siA zy!ird?KJ{F@$OgrSD#%%lKv_ zD7KiL59nXalXXs?f~#>RYL(zE@V#ys05Mrp=AkBw;a>PDE8OnBL;xY>j42ui19G=_ z49+dmLnAo4B2-B;qBz=0K5lro>J0O}%;E+-*LxHnk%wcD(lLx@3p}kyOv_pC?>G*~ z<@KpWh_R+}+?}V)8Z)D;Oi4HEmP32Ayj2sRqBX~h04TOtlvX=#@&kNV9;!>jJ{-Gb zBf@1l(S<>-`9xu|C{-~>i_26+++m)f`xx_wN+tS#LHS->Zt@F>(C4D2&KKFY80P8= zUYO*Rv}m1YiU;PJ19We4G|zEA<)(qcK%1m?8*|0K@!O5=VpXt{m6tg?CF%q(+iX^s z;kBQr7E9b)SAv{eMQtEk5x)c_P;@LC8jUj9cB;8_*ITOyWNjNcMNV8X!H!^%WUwx1 zoy>5IHetLBeOF{IHOyDN0nh4db$_hLikIsTvT@w7$L9Ykzx?l?>zjn%?hM<1Tf{J|vrB(=2LIa>o7G?3Qg-79R*6^Q@y=o>jcKVr&WT0JC#PXFc1%F3jBJHbZ z;qL)&e;3G%n@oSNtR-Q@Lw8Y}H;Mo(QR!|6u@hYXrQ7}I+16)Hh&5a8zttrUA@PU> zo*k+&sh>*9#mJi08*Z1bIAcg*x`j7Z^pDQr=QMWVhue1gS5R#g5gpUo9n3Rnk^KdwL8ukWIg<8xLZAYp zT$LusL;}M~i!+7(PZ%{y@%Ll}T?-2w=GY+esu#C~;4s@wTP5?;sJrR*MprjjoQC1Z zwtt-Sisa?&F=*#NvxJ#sDWEn@mWrzaAKVL3Z9JmDM3R|iN#m>!2}6kF@cc17<#~pE zc4If+k;Pei0;SVGu)8e(08+Xpel@k+jh2*H!M;Fe<(?ek0ESdY9_}SSNoY7+lU!<{ z` z$X+0uYnYC%Dqm}H(pR;_n7FEn>wDX|IUk3)F&LJHDj-;Q4vDeamhFol#YLfGbu~=$)te=1j83u+#KEYeDw4$rFkG? zwVl#E`+)-g+O%8?E12YGiJ^8z=IWg`=DpK*BOmFwtug}GvB|4$C6<--rLs@fNE{XE zdISY`qMS9B`33n)!%FU?rGk0^MlF)7^$XW~tE1d?k~(rU7ZXOHDxp&@rEM!mPu^3# zp3>v<6Umr|0cp%wCx7y%3IubEX^)y|*mKyGC8(iNs934ARM{TbanOZ~wonaJrB+}# zi+ivb0ec)S!Mqs7`;kU8|2UXWXRUXktD28E?t5L z6|GTKTsW^$|9Yb+ZB~)@@9mCtylYCytoe~`z0|h012=asd_n5Cwoak}2ZT-h&1kZ; zcWCGAo{)?IlJbS6+io|4nX$3UHT=T4Cig|kiK`Bu6MWpwfDy*-$FG7D%F}lLqEKTf z)gAm-?~+(J>oT>(HPou4;AwX<<4(`6DWPo|PY>(R@fv>}6L*{j?VudCeL>Jwr844F zOL|NePyUJiPwT5U;_sUok@8xM_GunI_%BBA@fb2X0E}p~xfOOPK_Mw$j(+}v(A}y) zyRdETMtbOfn)bcCn9}_l=DBW9o#+IeWI0q+R%UBGGzb}7InH}I(9{NsYlEqQ2#VJxB&C&#rbV>f?7+EE*{DS@r5*Puo2O7V_7ykor7y6<=E~@vUxc$} z_NTXKnNJOie9oSGO*kbx{Lbm<7{7lyBFg8ul4v7h<=fr`I#oEa%dD zs+`U(Fbxf_Xk<%tB>tCZ&5NU35JzIzTi>BUL<{aA!yotCj|SKm4}mz|4?w?p{{-^^ z@^tQt&2c5CUK(nM=5AjZeI+Zn11GzzMJt>v%%^uGt^|9b1UNBGT0%YRRC2KB#agK5IH z1;TRn2Ga#5C5D5+8bE{XR=9rk!CIwIgeZ!P3h@rEMkjGIfOr7K#e#U$LhDa8*M!E= zu$m1t+`1OYN-+m3r{Vn3c=mpDT>{kTGTnt564jfz4ryeeq})W%4D91$5lQ4Nk8!G< z+HGl>t()ZMBXjwFxwfPCLvoN}T)?5-cbBI-dMBbBzGH|Ejr^2HNo{Sif}G;v$k{=e zwcxb!r+M85oG=Oc_|_ZgjW!KO`lzEzr4ZCS!?bc~1leB)kq)92lUJ6ROx6M!m9DT{ z^E{@mT)@l<%xQ5vZ^oYiKX)$A7|WF(jY>$vAJxT@<;vv-813IW4z?JB>)`}`eLXN3y1R5LVi6ZzzpWnvyNrg*Sxp+YLGJx5-J7$9n_ucbc}8YGZ<%!gfaR5& zzde%+%14GiU6m4R1MEO_F{D=)U0`*VE6cK~53O$G0|yGW7|3~kOsl4$i>m^(un&Eq zYY{K!0AG-L-37^2W)CP~VBzvLyd2FF2;bNFT zx1S`!j@=wg6QO2lG$BC(ox{Vc0<>9*B*VQ0Z`&HXQRvd^+b$6~6$pa!+on7LL^#HnJ*VEAnDZF2?am zrdnd)Ij+W&^@SDwaW>Wp9(@!Gm%>mnt*YvzR`=|$gcUu8&-?3IzmJHnxc7Us7c7E;Zzdf$Qn*l5}Y9tF%M2opn=oPc>u_W^Ewd?{cUls=q8M6)c+cY=_DBKB^?{2JHv;pwAk@Fzg(I_|N4c( z+$c$YYDvqJHp`fI7awJ5!2<&at(p9l=ltIEj~qg(tuX#~3yVc2ocVuv;4++2MaoQ! z;(Gi`E?H#gD1vmcjBGkt-xiy|J%TXW0pCvj5ULSBgS_ODEook8`X)9h;x`Ak3orC@ ze!K%+$QF;Mwymj)_kBU{Q<^+gWC7xJG6XY-9&0mLWB==0r5IG6_-cb$)QC^Zc(Xsx z60=IW?gxkqA|fg7Ex`=sG`R~K&KF@Ta9mrs8=9(6G_<6q;V;vzK+au^%{-A}!ePRqEauBnDPzI-X=9|G61+5fiAAj$HOuTl{wLZD){3-IY%>~7)Jk|4 zf-3^?X~vWB*?|3;Ix;%?9J|{u`~+tt!&{o>!sp*?^MG4PYA~XlG;^!1`{O7AuXFH} zd5S(n#Bf6fXtAM4e%x8k{K64j-4qHecK|vBL1?I(Idl>s2YrNldX*_@=7i}MWfo`q`DE2gEEdY zqA`;Uj>oFeeGx3f#`m~ehhvO>#~RKQ_QC!mL)O*8>c1Da5M=+e`ZUyEIO9KPsyYK@ z>YPZhV2i#(Wv*xAUDN_?6=}uTQvj8_9+YO40Yj}O*i}pK6f!Wbw@-K;_^SQh>fUz6 z0NR=Ie?lZ~_YB-!u6|MbBdwvxL)=PlpUl|_W@xn)G<1yZ8Yn%o(1ZCO^9+Zlmeo$b zAfTZKU6)wyDU}etGP2>+KaB(WA8^sTMJFFTKc6>&=Yk?G;yp>CT;f4M(SNjkJ$}69 zACqd!t|T~{0^UFpk(Y+0srQ8F9*ZO5WN1pTSDWF)jO*`wy!yTMH{QgVkhhL5|B9Zr zqU$)*CQ!L*L7&*@T$Q%7iA6NX`dqOZrm>+|kD+f{vn!U9_w|df_Wv5K{6EVIy=T?I zSr3emuIIA<&fYKOtD*5SC~`0al38rLCfLTrF0Xx_1%LOQflnI7=#y#6#105TbZ?m4 z-ndKTpp3(n4m~S>4$>ubUQmP?`6{~i$?~IT>GyXR6s4o0&ZxGZ&&$?Us>u9AN%801 zSRf=JW%6@Ao*<0+(~?w6L6ARJlciJ;$4u;w@S8EiLvDxQ1sRu(WL?lh0m?B?GQSL= z@7IHU6`X^Bbp#Ma)*DDoLqDHHn8OfjGS|K=z?uiJs-C_|O^6vZMIeY;7p7Hda-Mz* zkDJqMke;K_CUF-Soi_~$XqG^NWh5VF#0XUi4|AiZQ!r2V?Bxb9jQU&X=JZsKR#_#} z=ZZvapZlNC9M4#%9XTn}t=rBCZYU;*%WUP#v5nJjrn_5DfBN}XA)p|#plw(Iqo6TB z&>5HUsI5b;w1yi%B!K;+0ub78CY>`=ZrJTUL~UO=3)OQiOr3Q7NlJwCW8ycBID%i} z$O*&@tvOUzcX71Ao5a9P=z=hR=*j)}$SGbN&k$p@SR78;ierys7fdsD$%ke<2+_61 z)}-@V_j{1ARM?LNR+rBMA|og?LGWziRpj~$*Om=y3H~EP&!?0tx^Uy8$l`R{rw@Hb zZ(2QLM;;j8Ow$-wQqI(hewmaEAgkneTb>M&%ZWYFr*mSueR)4cb3mt~4Wa*;9@-D&-@Td(?86(G>$pbH2{T6$LDG0r%>--O^b>2TaWVNoQ1a;`-@Lz z)1-LIwry^j-5ec*2<^RIF1uwJ*Fdd2sB4cSw>xXwT z(>+F`Kv1mho@NFTNc1W4xBZ@Z>Nbi{HYCH9ar1{NtVSg2zN+2UC<|$UlYUGmcGo*L z3UoWeR^PV?KcFBM44&}zO!OIcRU)ct9Lhl5=}DyT$67ZZNrt|r)a7LQCG@RDyE>T& z5iyl2s`|jHmju9Lxp7S4TDfEzw?^ELLOEAN%(r1ChZ)b=GhT}5Wi5xlAi5Ga z07Eo$hSeWzYQ8ZUEXz3&%kM#;l*Yc3?VH8yE(K#^)Z=WSxt$CC<|#`Eyt%+2chWI~ z*Nn8Txp%D=vQC+P<5EsPt2C#B*EN>eu)1~zgAp3F>ZRFFhn28i*1C%iR81{$v~=ZR z%h&cl6*5g*lc?os0$5&29?UPkxyESYvzLF?mUL21KK?yQ4c*HL{i#Dx0E6EogumML zUO9~o{v8+`c6DfqUd$L%5qt&g@Z5~vBK9Ex-RDYvTL}d_Ejf;EXR`>Vv?0zxYK$yh9t6`0+xaAyWB7RT z*Ll@Dyhj;M^1zY$q5m6upJGm#s#>4n{hYoqdu8}$W9>U-;j;bN-4Nru8pd;G0r#Br zzMeO)eyxtTrtb-y{-X#)V8S{yNV^Jxz)bXYQQ9n8qe@fgy8fAilM}q?b%WA$y)#y8 z&XV}M_3hU~yB3nEtru2b zpp)`Q`tPqZJo8g4j?iNYX_z2Si*{fI&d}tT(FBa(N54d7=CCWUfBzqzV7(arsnny$T!olfaRehlt9b4|8wctg1Y}lg+OsBk zSOqAwuWh<7zlF?IT9kg)^}SFqL`i@ydOw~r_jx#%MyRZ^@XW~Tdz;R>{HZV`WO@s3 z9x~;*w!e#J4@#|mj<|@lwFqPpSX8N`LBCVxjp8_pF?a(K2Yw+j#Q*4PgL94~J(1X& z&y~(9D2_5(e{cZrEbH4`kY<2@}9&Y+am=T$UIfK^+vu@;}O1HoLs$BxMHE| zAQ1n3*N`zbd5LM70qs{Q-!b{$stmM_ezeo7#z;o%={gvuF84$02I2N3n;0o04T5ln z0&28vVj}5HTZpFZvvTNU?|w=tS6z@7mV2sx=)R?wkt9bW=B+JWa>l|q6SKx+K2b@B zL|fM8;l`_#ATq3B_jLfgsJ~!rQ;v9A_>~&@H&LJPa*Ym45I{7=bU9O=baLu?+&Fla z#}qXM^z8>gE5J`<+cpk*pve}- z*x~1p0PAtM>LMm|Yzs`V-l5~YI#e|YERt_UEh#4{Qu($I|@k;4&RTep-m9>A$g?c8?&m|;xtA!v(S$qiPbMbjjmThu_$9 zeUeP}#YOuddxgx(y}-{)+FC1B7Gj3f2$k$wtdE&XVp2HmJnbo{kE~=b4`=;*RK}8C z+H*#irL&hT)lS`oXX4VT$terjY1Sjd-oea2rT0Xzmz3jy?w;n~-eGr8+IGs}f#6sQ zu5s_LGp#(8n#PQ14IG^?K#cO^J>y{lv2`gG`YHzP9j}XT*rHEk1OsAvMUz!CwCYuw zNy)#~H_UnZk1y0rwD`4Y+jwQN#*zSsgSMXNRLR8J93h4|;uqVibSomA!IWx3rGOhrgj2h}20aF^ymn3tXHQcSvI#VwkfklRG1Be(+RK zzEap+AYs{f0jHcm*zrti{c@Vw5>hf<@P;b2d?^VNO3A{u(z@qAQYHdas8zFQsbV92 zaqktL^iu{~qY$pn!k6d**ZlU{MdtjsiT!XYY2SvSbr4s2?fS3NuhidyKG%P8&3@ly zyX_?dnieb~C-x;PWL9mqjxFG>Ob%WMg?saJ`>Sc0z#i&iB z<8m>)Lle_Md6UHsmiCrJA*YKY1g?n2BxHf~)!Wl1-wa01!QI}5qeZ*z5Rz!8ajtc( zx2+g&Z{LyPN3>)ULd65B)hq)S{q@lH)FuwRjd} z=lxpr8K!20#(?Ou`a-BEw}j|;w@8S!MPn@au*fd1M*lzamjB3Iwm-7|P>-y4^sT@* zIXsA<-GzdUz7{?^?Q^mw_NGZ2{c~rjU`uOj%b)(~j+;$(kw=&1ttu!6hU3g+DB zEE@Pu!}qtF#l-;YG}^ylEEv+i1RM{rdtLhjkDN*d4&D0QIllDIn-U~2l|SJ7%DF&| zYu{M4>vL0I8|3DeEayTjpvOD@|LWw^6q61Yr$zsrX*;PC0zH4L9HAuHB%;XyR9t3) zl^xeNe=0dXjK7hSMC?Li|1L;15#-I8HaN!2Q0H`OHl zpH&(64+P#ga%fpgx+G{0)eoi_-92FSAL$mr99N@-0kvo*25mFZ3j zLO+?^*Y=cMw;ko5DSqs#M6^$6e#~))@iTqR=WQ&L_1szUH4$PlXq+Q@4n2DXn#?s( z1q&%yvxfV0j&+FP!L0bcN0ComWaS55W0ok)lNerNyn;5AuMr|IVXuuQ}jj+uW0wlOR)xHQ9irM#f|sm$xe|f zf0MCq?@S4qYnqTiM?5C}<8Ft)1Z}!|>A4$A@8Xy>=VIyrD%76ys#D;vaxW9Tg#oI%aR zj`k$(0y-K6LGlCuS+~_(hBXG>O>wQ+5*m0|cCwxRgEme0{L=-h$-7xCQz^_u7hV`R z<66k4QwS7mc*-0m^gq2`F@9Q{u8pk*B>WRr4b8h});7SxAfYDFXFP*sR8 z#2;Y6jJ#V8#Q<0SZ2AN0Yk0x(n7|(CYNgx2$fO*uhN*P(N0_q;HKdjcRSWluZdu?XNN1@FXWtlYV?n>QK+ukyg!>HYsg;VSk_go?l;6Hy zsH)VldFU4IWS=A#seVYjEpT6#eb$d(S!-Rm0DMd7F%X$*!5hZ4S%u5E6-pXpezDMS zU%?kGF-T6l11g~ppHO#Fe5F-nXUf4j+llhoKBm4@UVJdX>lhxae|N&K5h2TcJ>|E^ z7NH)oz@Oq>8scWN?k9&fn9R|U!(ms~?Mrx;vOMJ(#dX7taEzt>4$J9seT`lgb@Yo& z;^nFgk)FnPbW)Z1J;vnorVnY&f`1Cq`kyF7tLFL!T9RMjUG&*FO$$n4z61lc1&#K_ zQ&Ix4tv|wEyshw`?8hOUWF`LQ>}ZPSe-wFraIaAfKsd0C9Y8Bothzx_(ZdA*4AsP= zu+B9)Fw0}78c(ZEL2KGDBInZm4oZ6*#D*gPJaNY&+r4cx`%s{B2-4N%rvpLT)K<9S znA+`;idpk!fUK`&+rmb8!W^80r$B!)@!{fxBG>mzUCSdPr+{mk2dk;`IhDlEq7I0Z z+m=>QuG}#x*)~0v1ST@QMQ9<8Ils&$J2LHcjIJmWA56jDChbeuS3=2rVSPp4KP!E- zoNIJHQ+Xl%?WrXTFnVVtgAQk|a$-PMj>p3lJiRu64)??PA&~BVWeY8nXQ+1Le-TOM zPi0|nc9<)(Y?sxf40tNm%2LWU6F};n)u9Ttr~jVL8?PV|#MF)$vlc>4Fs^MSe3EQe zlzX1apWte4@AZEW_RhhP1kl%LGO@jJGO=xI<7{j@8{5{#w!Lw(v2EM7ZNI#qzVFr3 zUtLq(RnyZwRp;Jw?>WbhwxmOG;_oi<^!qZj-63%%(WFX3+fvb0bjwYRr-$94ckR(E zqT?-pZK`9KWRbUwOAhZ*a5~ZpYEQ+(2Od?SR2vnh#bk9_Jkk)R=RaRH4*v>0<5iHr zA1X-`DdkWcwQBSrhg<>PPcap6!F*TlSV1mWhN-I+f_?foVH($ z+UaXUNvN35Fxwzpn^4<$W6W*=A}XNN76ShSw< zP=&7)9_|p2$Ib1w+D-Q;TqblQ|xgYtF6@ge{#M!;s)SZHNxbDbU6__cNxRN9kdAMCWPGc2Kv+f8o_J(D}L8 zBGfBj`*2TCwM<`=M6G|dN%zRPV#-klb6JVRElZ|$ej-AW!4x*(xvztuHk;qCq3hLh zVEJ^7%Vcrba~%8F&$T(ls&4GT$l)sC#HRwrl#Xe3>1oRYi6Q$1`!j{Yw`gXV8=a{|~&C`HXQ$T4Wh95X}|4lm( z)ff9!{ncF8VsYVK<`X! zkt*A!SjDgyc0JsWpkiN_w(eAa@sdU{l-Khz5mU znsv@#r;_#KQL$4pAOR#H>SwB$w$pV(5U=ol>IjwTBDJ0_EK51DLW5WJv9^Y@tq;Ng zb_14)@;)24DvO?VOQaL`uaHV;{A@6+HOcL1bNkQTa{9)aUUto0Q$j`y^N9KLY6`Bt#GW({KKp!f2!{3o{1#`FU2_1 zh5E1Z@#ay>y-iGIt2$lgI@6bcuYjnihS8LNm;w?;Q*p~|2N4CYa9;)ld!D_^&cA6G zxT~ppdAp@g>jT%_Cq#=~n1YCweV9%uf}>QbssU(glVg-HGm<+0RUVk)JU<8i5*#)? zd3UNuw$oZDG5lVm_ebE6_ANMi*Iug!SYm)tpLL?3FlG(FQf*6sa;}1Uawz^yD$bE9 zbi9i&PAyOnfk7TOTQDhV!RLo4q?F0+wV_#(>n;PufFH2Z2|c3wSAd7|=hxeno)ckV z2cR|4#oZdXURf{lN+r0=o5pjXX$ub6kkVL^CNV*Jcp6iT)?Ym2zV&$d7K0F}JKM}S zJ;vYn>aM~{g)-YoGho|t7^NiD&_*@`mUXpt&6Vjqkh@Rjqi>uAwFs9MdZ(%NRYCu1 zp^}G+q*`lkvRalb}13NoH~KE=s!vWFmDdI_I!@)g8S-THBC|M{J8pL*fo#)BTau*079;I+XAeUS?O*GxIOo+Uv03EIWk<6mq!&?vU2aZ!%( z%pD8gShv#`5a<79ci6930NLcHVMM%nB6e5DwXL||+H|bh!+A4NNu=}*HMFbMY?4fndirZOZjBux_Wb5+7r1vqr zgd+`TQdAEBRMG4+vo`%ZkbX?FkakP!0xlD@xiSW*R}7G9V|u*Lj*_F}?0H}HK9<@E z$B&{%>w5$?-H=+51fmCy7=)|jjy{K$iXvknp0dNB`8gND8gO#JYMj8vsgH$c9Np8p z6`~o$ToiMyYAi)Z!PBc=GNRK(nd@|E&xeSGR*g17l+#)LcdoLVR&4RK0qhv6YA`lV ze5R^Y{l;b=q<94aKQD+`+Kx$A(L{$P<;uAG*?-MSdeRxO%M4ck^AH=3T{J%hFE<;cZfdC{Q3H<@(^kKcnnw;>4e0wtwobCKkdi|L(AsdI6k|*?s zm($ldgqz-)eoaHey;~HwVbm$UuG?&4zniqDe9z7L5uQMkeA@%-*gO8CV6M_u9CC#t zRsIk6QQU32d(3SL@(uC797_)3VSy5ay|pn*3f7eib=h_cVV#}}uOa@eVfuXkmpnIA8`SYJ^_nozl45$<6&i=s> zy!J8El-iAjeaU9-KJdEQtaHa@oLQ^Rgehf)C$6AU&im;=^^ZB*bkRCp1XJtZFICTq zCf9L`K=#?13WQi{%f!miq^h5bFBi!)APXu)tdxi!m_ z1lAT-=ylF&3~KAY@2Nx!!PHsLOkloj?KN>oq8yrY;nk9_*!O>#3VaKkQX#eBInh0v zBm5SS;~D4*n?8Yv^6gT3`R*7KCEw_x3JisFIG=d3ymwr)0Sw!Hyyw=^>5psTi(zCK zuM^OP2ln{-mo!=;mXj47>bw?AaNM#ha^BLW!qCS8u}dMtHfgr9zI<5dSuxs6a?pHE zQtO^`QtMcKeZ*BqoPoQ2c3q$bhSHhEozD+Db83bTBfV#eiod<;q|sw$lSNEU9*8ZX zKW4@Hdsg;1@TUg+IfmG&7%eOL|7zVrvy=;kEJvkkkhK5;(kIMECVZ~Bk469(1&;07 z5lJt^8lmH_WcI-iw&12oPTx8`p#a1q4hT~dKG_?)rZ5FFltd9PLMcJ|>R-}?`^C?1 zaIp;-O zbz6ti9k#8%QJ>g+nyPazJh4y<_=-72AfOuO7*X%U}|S zrV;oVDfaQts)$pXKtE+`e?f5Q)s2aLf@ z*6hbS%P{b4_eZXC5|Fd(4o^m(EfTEF2jFDKEo!pf8~qQZz2{tgfCgEPp-EA}-(~P0 z&+EFnQCQl=;2=b)DnE|gx!oamZ6I@Rc5Y8VjqwT?1CF-G(=5w^SpH~3`KQ7Et810- z51$crnbN%j-0- z)&(b~Uq90Y#Lj(ljqGt7k+E!s?q!9xBb#RY2{dEh3ON`h;Osp@=3@<`6dE2e20ck4 z2gs(Hm|Udq>I#()Q)`T)B7vp?1=-yet-1=e$d2#E+a%Y&NU=9P44YB!^BE{ynEYc5 z+vQ=wGq@q&W_ViE8n{{%yOZuVhSB+d;YL^<#-i=KrfXPbj<=9z&KndaJS-FmdY<0b zu(dy(raPD1XcY0)@alEHpi$vHyP>zpz(``tp>z0q5tRlBOAuZa(F6Lynps&!oK(hj zN9_pvD=2DIyKj`Rgp#w`^u0eXc02qRDm4!Ta!YEtH6wJ3onAYdR;rWn%yg~5d5AN z@?!_`jgYUi7F}K+2b?G0M!4dUcgG?`ZkK7!dwac~An7KxUa%Rb@E65rPyPpvlho(y z2wMY4SkN)&eOdg8%K!e=*3y|lOEmlh6pNO;35FnH1f=o3>-_yp`X$+4IG-Y+dfj1U z4sTZ?;6N53C?l3a)uUTmcZoglJVG-Ll^Nl(JJ4C3dwRa(Ok$V(ktJ}}-{bo_b$!Y4 z=^EMYI5}FzKexX2KDpf2EdPQo;ZgAuKoU0mk-Iiu%gJdt}+6zOR+-5MDm zYWXrU5_^VWEoWi@nwLN5$gB!zynoXp#PCVAY2#&bJ_F1Mlag&TyVf)4H z#^61^PMGyzPZag^2X)6OFCLtXY7uzJn|s$Et?Bst1oW5 zOP7V-YY(j;(M;Q}medviRy-zKHEF{z6W8`WT9J-Dum@XvhBOt*USf}0(>(cg7MSc>`{p&y8?k*pO?{7mGwF@Mw+@2!?nJT^!SQ{CMeT}uBYFug1-MPyZ`Sk z`v3jdT*;c-Q_-|=Fdg-umo-1&P`(C$*S#JOrWk>LA!mZuTeP&X0j&yqjU_QTJZ7h{ z<7F!s^o2n0m8VXyuudaqK%4m14zo(|w>~|0zSieTyq9ro_2=KLx2{{QW`}bu&^>UU+Plsc+ClOdA||;ZxkEOq?SPS5ivRWk zkfHfv%c&?Hb^0OEFe+B4g*sFH1Dlw7&&jX)MrpRyh+qIdokz_K%+%MS{YdQ;ah;y~ z$mJp9&Z{A|_mpQ##Hl8$JX<4;v91~RHimO6ZX^5wny0c!(>1oCDQpDYv9equ#N_ig z1uyM*#d}vJG$&Ud`y)I`<^ZDZ!I)QDsN?HE!mqzZXWD5XvwH%1BH7U^m~urU`H4N0 z*siwfoK@h{W$I2VraY-CqZKvy7q7ta3bsydj!=t+oRpM&{(O#ndxqIC_SUwYKc|tJ zU`;uYs?fW$w>uWyz~Sh*a8&;~QnXS?iF?&Wt5|30{5&mU;EXc$uiF}gLpPco$Q&{N z{dd|WarDNC)1S!iB+njRpMo#d@?Ch-A628|ZTWr>w&HPS0TAs;e1L~)w_$V!v%d^X^xT@lbH7?^a^tt7j-YU*chXx02@VN~ z@k6>}r8g||^Vt3~R_m@kbe(m}FZXx~J>=`T*Y(~|a_%vcyHnebl|zO!FwM1PUQ3^Z zd@y6wUuMh|Vm>JC&RGsxWB+`lwe&pC=q91y{RWuIM-xO1b&R9jcJHwpTyO~(1J|8I z9~1B$QF(eH|2F%yOJq25iyjKI4_ujP{ieWJjkMMuz z?hmC?gwh*?ldu@)Ge#R5Agvgb!!PdwQyl=+Z4~y`amWOmG9YTvz_#^=D)sV{C}0D`qZ89x~=S@bLSavjfWD?6nDM$ksK%V91YnP zBI=IXLod`tJu1V;nU_unKeTgr{sro{#fJ=<<8y9caCN_3Amc%Oq#p6U4aUqF7qA@- z)28XEAz-$)430C~I{zlj=ZWb~{bCP%`x#3;V9?$bq!~Y6qC-!sO-mxxZ=U+<pv%qeoarK-J^TEPi|aX-k__sT>SDfLyfj!r=1L!tf6PtYuH0w;u}X|CfQ5dW&Xh>qnsM-;@`tHs)RmZKovr;~Rp@u0__ z0XmXDb&h2+UWa)CQY(LTx5)!b(m=liy&kGXkH9 zcJs~zeesV&46vNu--$okua9ux>ce=}?u@~$X8Q5J@bP$SR z0)t8>mg~{Z_Vx9Z1M;}Vg5uZyy8K%&AyCB66;*-OtV=;aDA$T8%=Y~}<9XG)E3AjX z_h7Xpi*~e1mKkHJ?xNkcJP_<`c3%p@xD1X_Tcy3h1Mv;aq4I3W{|sp41WyAmwyou! zw(Y9)6G1@s`n9g_t2xFI{_W!ANT9O!L8_>_OM`GJXyM}x6tphf?sByKm3vq3GyX;E z%TwHPJ8;en#?<>boj;S?TNZzfjm>_dB?{3AS&*)Kx#&RU8?sA}E{v{Nu2&(dXogM3 z*=lqOWk83+m)D$FoMnGS&Nc4V1w8gGq;Wx#VEfG!j$2K;Ad0v6~?9gl( z`Ua?1c-hnlaFMN+v7Mb;R9Y{(cV+}hT!vm3r?SCAH}jS~G}`$8^B=Bw+VAREEBhZ^ zyB!!#i72;%k_;J_8?iz&^y{pHt7We6pX^!?$g^=NE}QF=I+KHAYQE~5ldHnO8qV*i zD^#LA#m5;reGgO64FQgxzh$;g?5;o4yGJrks$&qxXEyoKhf9Rz)R{MCHQHn-oSDP8 zPb}&3)$LChJ&MB_@I*5sMa!Q%S0E5kI99L&Y96`o9r#c1uzI{Y`q-H7-oMT}pkhgv z+{vF)o>X*HF1+Ppp1}0=Vf;X#1A_8UE+lCTMJtI_H8-m0A7wFd?Ons0<@+}VP9h&k zFGgt_+R$AkrO*+MeVznk?cdyXeqojH~zc{iQyOeiu#U@ph(kMzTSSEc`9Y!3^%_;=P4p{Ms?n>yHncdAU$Cv;yKytfqv_w+`U z6p{>M>>Vgd(RiV(=l*{FSJj*=+=B_fgf@dZ6T$|n(9ds^8r^hj1Z!9?yC%>d+qhj& zv9wQJT8xR@A2l?`1l-&+H#V^=Hfc28KNqfFo}*TP0gsJw&v^q{MK!%RHDSC3B~UyiFZ=OMAdfzhv-b=3_qOZGBPpl!S$DjK#8X2;G5=P z_r9>B{B9GQ;6gJsiJ5hzMs^4jmiO^^t&?^1d^u$xsnxWF;yr zBv}sBUy(V*z*-BI&=@-NE+9KvK%fc?NsG9Mk~Xn5%-J-?)yi%>t<0oEm{i_M*>kZF z`dP%t^51lLyoQ&qEEsHlHbmxfA$7;V#w6Llam|OfooXm& z?9^fOfif%MTeAm01@qkl1xLje0BjoT>%b{tc^w@zOC%v@qAdcQxAnpu5ZlnNM&Hwmw>3 zlOHhVSL36OTLdX{whU6BLf+CqR`wq$WVisPssfR8_6?#WNH2PN306?*%@ZzczRZNa z-=EJn`R6wfx%--l7LGEj2?nKE6O^J0jFVaZ@kb|r-ikUJMQjE#tNoPqsFwLvIS*4N zKNF2>;RFp1Jl5fQ<**7zev50DFtgx~NzNIH{@vc&XVSexBQ_BD)U~=yIjH3MDYk>g zMZ0Z@?rWVAl*iArg&OPMbRpOWTumn5TVW+Xzo9kE5Fl##M4lDHE_{AxMjzS9&{$l@ zP7Fk1ADx%b)qTSb6OoV7KICjJtXx7|?T?iVVqjE*ku)W58k>QL?W{wPx?pMBRy9iP z^Y}PFPCWVc_F>hCV+pamjRTulq*7~l)E0+wxQ*Gu7m0_gM^Hz9PT<4lr(FYd z))i(HJby?^FC)w{pMkxz-H#|mjJ5PcqF>|xiqE0l*35>2@eaAzBO~!~-|68fvcLM+ z_2H)zz0-wt+pWdB9%ijtOI9vuW(vvpI_OFk@%^|ed>Cjl&g+uC{OIr{`MSo~lMRAf z4i>!C_N)Gz-c7ZQFcwy14RlROj2Yo8AE{_3TMTqW%uuQf_k@I7$z5-(R(o{0h--q} zQnmhCvC`-qFZI8BPR}!44JJht8GLTF>mqge`iSRu%Slg33bvDe3U^`O|wIYfa*fSf>CLL z3p?GlhA{u_AAfg!)7r>AQejI6HLX`YPUa%*u!C^p&v*vg9dqE9dI-Q~Z!p93DZjW| zz#)u-+j{jxuB-#8pkC4Y&0HS56!rP}7`Nwewb4|qo^S{!zXn;F=(0xV$#6f~X z+T5c3_f{p;<7iBdkL;#(2hX+Eh%uhby~FLj-+8}lafkPQ>g(#pf=^uhc~W*l=fvWK z_o7fTe(%+{ABx$k5V^*uIo?f_!ty)D> z4ZexEewT~Lnl0}E>dRf7;7yStyGiGBEeF7 zg?B*%ihQcn3^|acZ@pcXWw?SGm-Q0jhG3;N@=R}iYxeJdeLD_ll%JU3|J=;?$L>Kf zKLvzUR=|Ip+9nqRuFjC;)uW`MOL9md2dP?fZ~!nUe<$L@whxS=0_cz6cX3)}lTypD zx7}HYW>(ke@a4Re$s6zI+YS&&K$cLT)wk?}36a5I$%yv!IRIoL3lQK2D?M!(ls?J% zS9m~=;Nq^skQSjJa-pQ20#=&sxo4L!wZuwdXzesY&;@Afu1t4Efk0rL=HXx;Mc_Cr zDrusIrM$9z3!X)m+ptO>Qwj^iqd1DT8dEU$26X;YB)zczZxGmgcW|bdMWkw>hdfkm z$c75q*?035?Y$QW_Km;p;fT}nk2=fxSucw#PJL-(`(F-OlD~8{JPp(=CK+P^3u3? zSaJQSdENc|%{+o7FRyAvM^m3aOi(<6aaxOipXGF{rLEBl5ehBQ_QXoK2sYU_*qZ z41Q2-rly$J*)G~&J#Ltk34@w$hhySwO0fP-Qj1g35T>fj-@pK`rpd0MHg5+dKH|L9U`(>iHTWoDU+vo_(5UW;OK&5dNkMn><$=I?6HRJVbUVDzvc<#JcOKgOrF zNKznDsL2>LeV&P=17Hr6 zmEjP^WM;A3tDb$g*GFq6Pxz8l&0z62SfpL;m|^LFYaU+E8`cf#&1k5i)<}Cm++A?O zn8WUJ{5~J>w?~27lZZD9wPbzR;GLK~HT7Vp{VpPxW0mwRXYz=?gIlrtbq0F>^E2B0 z2e0sM<=nFw(65svEE@MKo=(Uda%8r4V1;<`5IZ(f%Xp~g1Gu|7WOl=$5BjO^-XfIk zV7qLq%SdIKq&sz9M@GxXzQ)ALVj-o)71@ zN7-wS9Ttb*-pc*mACDu$4$@=hPfXyVY@?q{GQQVV_4y2xd^!Dc7W%wL`n!LS^W*fV z>Y^y|!%Cm@u)j(Yv=5la+iRTHIE`*No(NGBByVkS$@9$RpHy)vo9c1T$Tt_{Gf@Zb z=ky49#y?V~lo9)4rjtak{$tmCuy}TJyv>Mx(h`lqaEs!QrRJjgTSO9~gCt^l& z1l};xMk+Fm=r;pTw#qW?^w8@2Ne20#2&4wm{YwpedaG$J!N4i{yZXpichod^KRF6n zZ{e#7SWc~u1KP9wnOPFt4fv8H(tiqR_ja1%GN@pIVJ_E8de`ex%MU!VruU0#vxt&X zr9DSfOW4moUo?%LJmDgNetvw;$Yob@WG+o{bUh@Z!-`50_elBb$$1kBB*v0n#(X|c z9c+EBsO|H}eYg$q`*8iUP0@_$TJty|#9NNoNBL)4u(`s;v(xj^W8i)X5s2rpM)#(I zeKhusv(D=-WBe)$($($sL$aAwY8jyRZ&#fk-VfeaRn|0%VN?8H@$dU-M+aM93u-2I zZpu_TT2RSC_a7lS{1hQE*~%IGpEFSOns-_z)lWQ3dLOv%9czVC)&C8C{uvmcB1LRu zzj7{O^>ea@he~H#K@874PffpqvTrK|f`j z7qj??RyM8I!3vQ)9Vi(>@SjB4iFv&qrxBQGqCUSo>u@L3oi-B^(0Fso|>u~WOeoFJA%_$H{7tT**n$qf)Xl2rZy1k=Jq zH$|vhc}Ro~9F^TtrNIyvEZ|7d-h6ypNEPCgwNL)%%g9pE!Cn{9zzbCk0dB&$W<+X* zS{-Cra59S91W;kLp&4bKP?|Or7e@C*B2uUyf>K07tk?EvG0Rz06A}(;R9;lriZY4k zB`@6XH)OTmsDvrq1eUSt;P&CQp9_rtkgSmkHnNDhRGKr2v1%-(qZxcYMI9epqTeCD z|0sMFMOp(Osc0polWansJKz4&wAFAH45Ap1052K2kd``$uDLAscNFB1u}=$BejmL7 zbW6y&<`}ZEHPzZK7~qGaTW(~xC94;kP)JBgxBXDvPmaLRJzWe|S?05GH`zGN;vv|0 zCKZ_#DF#5Uv^;NR?&aN3%6=z>rZU{qgQb^>PV-u16zm~29ewdlF1@? zDE@y|Y2BWpKrHYyjIJzYl-AXt=>=<&8o4-)n>W+UrBBbM)^m@}gl!2jjjpY(!>!PV zdHYgkrtO`+Ezd=36gvi$M-N+D5R^1hA;RINaypy=kFf6thx6h?f(w&G%IToZUaJt6 zw4?-{+NEc#8cp{_v=2*4YNUVyHj#i0FiqIw!AC)c3@BAp6r7DG4RyEXGtfwf z9it7+Vp%~P0dFMIfDZLl=Xt}AV2RgA$wOb6HjX7!%ZUJuseX5o()Bdeq%wE-8vfn( zQ#zw)z*m;M<_CXVr?>|$LDF~YPs+l|AXHI@rXE9|46kUZ)=y_p0jJgGBIVYW20h5D z0Pg)Au|VpgD#%F3kes*Jn%rd7CI6ms!jvF)pyMBxFBlaGR))F9j*P~~iWQRUl zK1C`ho0WLtrFe+MG5S)O#%=Ft$H_3&*$ABDGMdnI2qrtTK$gRkii=BB7Dt%PCHxe^ zI%6q_Keu3`X%!=SbG;YfpOaqt9!k zEt`NA4#NSRE<0JMG~PRKAcX2{XRhU>dYUs~oJNir7hkYgim}To2HSxvX+Kw(Z@@<4 z_O6(ibQrx!&(!T~ilBU;#>BJ&XY1Gk`Z*wb;vDt&Gk>M-TO&q{kM4mj$GXJl3LR&G zCmY;bV_nH+kQh7+DaH4;U#?pvOqTYJ{^S|b)jEmcjPWj>lD^)9j5aks`)s}!taLpW zhUH%i+CM1-M0kPzbRCbQqVI2rR~!PB!X(YJbtO?jPfYsT5!v z&V*{xAmH|*o#*H~mqsjz=?wK10ftdLk!BFe6@Z#Li+rx19=v- zJ-McED0(BrwS=*~jH?RD_4quU`-0#rxt$UtvZ6s0g4m4qYz?#b%Kt`phJ4W1*E*3!&a++M11}YEL zbL?{i6ZBnoQW8R)dQQ>&q08%TZFXaY=$=kChp!mA1OI~Vc6o@ub~n@0ScpO?#i< zE3Xlw9We8`|LDyBpYU1Vh$Oe;DT z>V*E$d5XI6Y=NvYSl%!dQ2dh_P{z3pEGbjm?OMmb(27W5Ev8Cq|DnzrH)-!Q~_0Q<=<+lYZcgnH$am4M9q>TpkXkt3g?(DF!Ic@V>_O1`;jB z{C+SUlxeH`)mV!v?}|S(ZH) zU?Y@cBDkz=DdkCz|EsyqPH4~-8`xbgrBYP!*&Aa3>ce=+^%kl`11XIg?O$?WZhD=G zvgt#!BXqnd5sB)`s!6*M%Sf)tB}dd09Ing(ebIRMwq;o*B9F4h2B$dgt1t08v}vkkz9uE`VMDhh!2_2;_f0`@)srbqK&YA*n7;+NLBbH$Xa&uZzp74M0)+b|&))59%-_?Wi?961 zYM(V9s_v_A<+bYVXw;(wr2+dIkSnpI!;oDw3T3s8fthNcFExg!&lTLylRxQ5>#7<_ zUwv>HS0n-S2|%KiFt{KA%*8O)ssu}*I5*u2UhGAas}!u{@Q-$(*o9qY*doQU4n^<8 z(E^&|0VEAUT0lVEM^eWlo&Z2@$__5yoX;EHy} z{6p6fx2WOuP<=$v!B8L;i!9_1@5FkunU#4`0$^@>p7I8mY&>I|e>b=@R?ix7NseQx z$Qh*$Y3&l+r2o7luBuds4Z_^k{ydK&8!<$r@B4b0DTllD^)b?p&UAg=D#^UFToQ>{ zkP~T0p`t8yAecG+dX=ZBkvCSAxk8^b(;#@0jk9a}6F2?ZkTW@tp~;mqnD z)do99T#Re#IEV{jB6sRHw)oOSm?_ReNW$dnKbn{pnNF(#Wj)gqtV1!d!P%j4O-6D8 z!8>EbvMC|GZ$?m2*`D_GYmbWbm()64fuqrF9Sg(g867_*bwvB@W%6D@v2qLZaPg7Z zpLw&uVH`}?%ej)}=Smv2++AKb^+`EB#<5)VBn?kjNdy6k1p78QTET;%CdlZ8F~!n{ z>G9^9jVPwyZ2XJEi5!*Nw_`IEA8Y#$h7*aKP=0n@Wx*Zn2v{xus%k43S!tDg@5RaE zFfRUumyz#4SZlmLHkq7E<1P}ten5(Bs|kEjkw!@j-;d)#rw^gWH8d1ScE0ap79or< zsA;*XL80Pebs4QSJuVdo)#1V9Up49yOc@i(eXg3Be2>n*M6P&%ycaX*cpSGgdl;r8*qlV**DB`^S(yldH#WQ@pVyYMt9<`+7&fQk~9B~|w zAbCkNH!}Q?F3X~jL!u<6K`ZjwVUZg?@;!h#GQhoL@#Y(n+y7&kV-x#7w5*LPc%sx9 zpS=0UIu3Ou3D+ylirWKhdoQGdLL(XvJ*{P9S%<}=`3d`bWQVQ z?z1&B>n^(jgY6ynFhh@G@@9NvS!Niq>$ya%R9_o>`s+g7VmIO=75Bm?&;W zk^9)dDX3Zv%u**cWtRl%ihM{dkcqf?jkHnsw4d##h?+1c_@Cr{Ohl=67ze^D@R#>NRtw3}D!&@$`xh%6^7)9P3V^mLd5o-_ zR4-;-phV~-*Xcv8LZr}zs*N>&S`=XzC8p$oH!ZNU$BhB*mUa_WNJsHH45BneB>&MIOJlt9?VjGM(%tVWt(;0F;nRXIPAoyU)! z36kiLa(?iJt7bN$z~V@z5}zymQ1MwH|9%e|nw;B|=#+zbYIbtyP9VunFL%k%$wG z#cCNuphqWzVa^{*iZ_ymjI&{oC?I;)!k~18oJK5DP3TW%Nz%wVIyf4MYj3}v2S?3Y zrcNgyAL^0AU5ZwE$x{EfM@S2eNizCsXH(V}|F-D&;LMVm!3WfdUzb{#Mjg&&%S zmeEv#ES=4v)BD+O1x9t*IO)e4fF(^5rV{62AK7rZpWZ0d1q!Z{BAIEjvRP>z@1>v8 z&TJ=q*bXc{v}AU3(D=M*Lz6fMs7x72)Bdf1v0V1FwN8<+6`K5B70~%~ag)3D)&Jf3 zQ>K=nE1_MInH)?c;#%N0f5k+g57OcdP$^&=v)i?ps+OkFGj~ zT_CX7(EISpiCuEHT_#%J(hlQ5+&`+1*9ul5 zt{L>z2`eek@qu|`T>v>zg7M}@Y6J@MpdJo*9JY}}{2@TuLk~3U$)tX+7WVL=r6%D0 zMM>#)n$F(+91+>f{@sEGkOC8{L09lqo~RKwPYo#$vl!$~TrlX3O3p5v0B@v`NdeL2 z0=SEA{HL9C)xpZpWECxeZM>*}+^CkI@sgn=SPW%n3G}6>A&P97E6G1)H7n=BT+OYZ z9ZU=qIFIK&WWmU;++sdYf_%P%(Z}%fwFv%TMs|Z<8h^CuI=@(@t473AET-D+T%|G7 zOK3UZDg`NMdW{u?U5MIS9J!f6E0e)6UJMH+s?%s5q`qiAnxU#%_uj7iT&^gl{Bqkp zntjV7c*yF4i8!c;eDE}3*Gx(-GB=N8!+F?&GW&5<7rFT{@?Nn%g>anE6*vNJ1+++g zuJh-&!Zm@xs@i-6m~)b!;Wq zgh>+BJSaZeoHd-@?0la3-nGd2Q$%eI#W5v;{OxFP*Et)d;iH)Pm<|yjMcUo!-$zFu za}+L8dBsd&zVzJgY+DdO^*_S=p_1{<_Y@KN6Z+wHNA7b?>W$#$&a{U8Pxsa=iEP(X z4C5qR^Z-**-&q5cuAW-Rxw-M_>m93$SE83@{-`ht4{vvWB0;aT8-HkzA}O*i*?I=&NlLkj%%ht8TO#(pmT_dNCgGHU%#kX>GezHaGY|F1c0^9af?NhmvE zKzJ^GwQIuDbKpkDZ)ON*7;s+V=eR>r{n#mZ?SrSXB?gtQO4MF!jtqr~3hbc8&nd5y zQ@kvZfLTQZ+72J_fk@9C{jtP-7+x1b7$o za)$#pGmT^}^)cbW@5nmc?J=A|;xXV`t!%vB9dztpDp3+tR+M{Q1PG#Z%v8mshT3-E z-ITYM1(^{qH9z>^V`C|^*UdJ^_tvN0*QHpu1s3e2vq*p07!E*LLFc9^&{!4Rx!PhqjXZCH|c5K`Ff{tyglXPs` zw(XAd#dgO|I<}pTZQI7n@0@qfz3)}M_vfx!yY{SFd(XM%7-No|(Oc)z#W!g(9_9qS z!eGSDIRbr#5;SZ90JJ~9(QY`%&Us)WHR`=>7p@Fp^+N=)YQ@>TtvWGBd^<`Sgv$Lx zqbx!oUTlxQIu*3Y$QpLtVM=rks!Vv|lmj;Y-jihiadospLCmzH zb&rU8m!0kR9GgRt&mAe#$ZtZYtg-#6^q<5=M&I6#Pa0t8zS?EB*LZK!BBZTe++{`5 z4(+zDTnw3cQlUOeQ44vxVHvj|af0LzeFWxE2V1A{Em#Wj>@O@a`GKII%L?eiXjrcP zi8hb8Q1ILI?!?XB<>MF#RX%41u0I3TpeOM3VKiSs4gPQmL~LS%f0d(TOB*9$SPgr8 zbS;*$kU2F05}5%^zZL}8ve{M_~l2k;T^jcWrhfkqvZ&c8-FxYs)pB`WmW zc4GwrG)oQB+uYB&ULGP*Jn%?1j=L!?%6pJPzD}12I(P#-e6%C9T5_ip$_61I)dJ@| zOsBh%6S0m3bGxVo_yywQwx^O|y_WDd@ZgHq5m6m@PJCOW_I6=7kY`EMx4_6Wb8RIS zk?$;07;#Y=CxgQ6a8YD{sVv^qO|<(s=Hj|J!B^HWQ`YDN1oVA)%_!`zE`JJvV3f5EA-ik!-!pPX6v|k?QIjdS<9{p$ zrd z>XL^s*TO}+R9f+j1V=Abe$%O_-FrW&dQ8a&#G`a8p+4boJo)OdnYtx?WOzW;H|J)m z?4ADUF1}4bjbRL_nT&ikAp3>=VPRKXIAjYEfFv@_agf|Jaz;~yg@vj(Jvuw9$j^~m z<>tqsY$JdW_-MnfS{C=LsHL0yB}bcg?IZ2pBh=z})JYz;bcum6_7pVM@nTmmih6WY zs=k##rSodzDJrNm(GgWWtX1=reJE~hkz~Jj)k_{e;l&c*_?=Map%km=*r5B-nBJwos8=B{UOwz(MMcEh_ApK0W@)DHD! z)ASSm)!AP$j}rmJQgDUEuB62-h`WK(@z$}@N$$)Pj(Ab4f^?|hg^PymAXNvs-SzMF zD-Jl~j)s-vtqMaLO+WwpvHz#|=Mocy4D@K7uxRYRs7HhlK`!(#ToLuM7q!n&+$I{E z3Wo6970x*(Y1t$wkPxYW-m+;Tpyu>-&G<^y<|Yb=92Km})PW~gnOf__(l#|HSrVwD z&t>#;57dr%VSygo1_U`IN1_qEmHd_iIj*3o*k_(=9ib7U1)n-bZQ@o8ELV?IQTU#f zE#m}OV`y%WptX(ODIa#UhtBE0Gt+djY<(yW=AQ6^Zqpfcp9SLRUx38O z3IlfSSYoy`Pq8^bpWmTBif2%RJS+&oIf}J%Rn@2@yr-dK_FWn5kG=P1-%7^eq*pze z22B|=;E={e$P2oZb?-fL`Atv_a85KGz~66>84`ctgpCWnMZ!bb<4Zy!JUKc}Z4~F- zKmq9z189x>pRSHgM#k|#{=)lS1XIgFgA8`8A}}{AiUnHE957B5pND{Av1yV8W-$#opU6fqVX)>gZ&(4x}@F0*PN8IjBCnC@V4sI@{ zEGa2oOcfj7YR=PLHqb>}5Ne{=+Kc=FLD{aA(F#EOKVW6N;Kolq+x0FiW7dXQZ9MMN z9nUdCL44ktOVXLJ*pmw4C$8vI<+A$D1P{SLIFdZcZf#0OI?O)J$Y2>waN{B31rrbC z9~S(qgLYK|5;NM6TvK(ZF?=5{f%qRGGH=%3@~$^1HTO1#v5XilZ%8GReuf<=JEoBiM|SR`F3gd%-1k{{ zoc+w20d=HT(M<~Yen>%r?&DyDVD-J;Q4hgtztKBn!s~Gi%lA>unT`R=}ac0&`Vq5V6dR_Z5+_slU08 zKsQX__Q~MWbP4@W4BZ9Ed_c72msU{(QqD{;5p0`zo@%$pG!6Dr@%8yD#o~=B*@P}> z3_3h=Kn;M+uYbEL$9tm`7$k{e3K!0EK8;Cj2!tOK6>PwjC{-Rn@W)<5taE6ce_&`P z(g^@%KIvC~mHs+I)%jLj;t74z9UEr4=oU2>B60)K8pEdbg6tfQ?P3Sj>+R+UbhQyc zPN9hE8+RK7+)^Y8>Syx6P05Wd>18Gy+r^IT(0swXHFdUQNju&IoUn=vpq%9!&FE{8 zUtQ3+(14;eDY32>6Fq=bNwNXEhbK|aO3a8~e+d9V{+IE^=^Nr9pFGGrsjb8rg!uY_ zx0ul0U9{I zXltC=#`clg7lR(ZVvA)}*&=S{qi@BTIWhO&l^%G9cey!ZtlENS5qPRhMuv7M%f{G(bp+xDJ4kaXx!rd>lY_j09ulV@-+yQ*TX66-ZkXw4M(Nh*|H9rt z1iKY2WVZRm`rf031ItaM^BeD0GNpZd)axj%Ki5}2g*++qTwW^?q<*pKZ?xr3xqXO= z|D!NS69Uz;kuZeJyjV5kl_Mv-tIhFSl?D9y%jihfBc)a*!d4FH=SIOo6WQW2?m}S* z%Q;-pOqrjku{bNl-ShwJVZOAimp&p$v9_82f-Rf5p`X$Bq@Gnzx)34TnAkS zU8;d;cWGkW8#>4B>V3I#<_Q);U9O>u~PztTyjdMTAsa)p~LOSAXR0O_Txu6Wq3!a(a z*AoRatp@B4z)WT$QQ;w7EWnCFRR)B4#K{EB2O8NiWJxe4KZ2&_P{T zRoKyNYghB~kqbNzmPlu)*Z5^+kOD2)2g7tUM7LfSh0Kcr6HP8un3W@>qqCGv=`;eS zR1IzkM2ag^d$7}uVQo`;+&?FY^y2IS1h?5R`l5g@D*rk-SX+aaI7=ld0X6w-hc4eC zZ62!5wer^HBD!!~yB8sC%50OxzXc1%X$YYnx?qCAteT-}ap1Bl^Z9OxxNWnNB#?Fa^{Hab#{pXHDnN0=4 z{T>@%OrXLe0a}Xjjg0Wj<&wV(EU7Szd=Mm-3h=GsFoXujbAFH}>(NV0Du>esY^92N z5x(EcJ`W$Rl7rj;BvVw7=`Cz4wD>3_=R)rJVDge3DkTjMG?V*ZTpqI|#TJ||kiZrz zi63pnUoEJ)CIwX;i4&^WfrNC-G>2teXPd74rH`Os z=0k2#u~74TWCqi-dfKW&@)g^Qs83^b6h`u5lU25>6Kj4KGd;^ONDRr|?i50PiA$Kv zkBjrZr5sRo{tl_FN8mZew1#MT5G+7itXMu!`O%MJzJ}y@G1nws$Y}V=tjdI%1}9Hc ziN^>Ha_X93E){qNh{5Aa2`R^jiztUE%cdVmdRKH2issw@;zETcf8LSf1?T_vpgk?T zvBtFtxBfMnJ=Hj7?0f8O#R5wZs@8DM#U+^XYRfDS#xhq1>1cDhITUyHR z+RR|g?Wm08A9xHEDA*Ol6{e)5FEP4tTk|AV{t}e#!nOWM@yUx zrzu8fG+g8`3bDG#lmAvT1mIia=&UWU0yXo_74$Jkw=q=*-UtTn-^XS(^0-ew2pM_R z((q1;oLQq?5)|}G6@g?T#PzaB3qMnB<=VJTfxlAJ!0(>_WZ%;uN}I*>TMhqGVb-r?@9H$YZwA-*|X{R@5P zzy}RwrlI%#H{kGZOvxG=ocC&I*R8ndI5G=$!%_grXjkg0-pr3R?}O>rJk)%mFb#BN z?7OGh`v*i0&CT=1`@3yb=2x_<+v_r7Ja68$O_NCnsY(584CMN(y{W8X za3P7JPsu^& ztlYOW3)9PR!gU6~hD?4qvjvTK+AwGe1O)W*h9 zS8W7dt$;xMzMwblQgKR^!b6nLI$-Mj5trHm#fQ~whscwYU! z3FdgG!h0mTm^E5aA|4Mq%_nB6Q*I?%BC9V(RHAFFfpAL#b#o3ECgj}x05a69ZgSW5 zucE&c4jtD&IDSO(CbQf2Pt&n=LbCNJ(ujUKu6}n?UZ>lkip$o>&M{I<=p*`YzVe+; zMn?YZPKbb;WR?AK3@w(I_SN-L`!-Se%$d3#SRt`LagQsREd($xEw2~OVnOsPg!!fR zrmh4z5lmAN+dlklAvk$o(04pAnA_b_L-&28K@-g}2g$fUtBJyxK_V|%fv?^gcwJ)a z@gqUbmAoT#=fqq$buRb-MB^`D*%6C=nSaw|Stw)_mNv8p2sFH%6rp`W6hLufOldcP4G|OC&XyNU(cEQfy zk^;kljx3-F)I=6HNvll_;(f*^!htsqJA6-TjFTSZjZA38!rF-PKoH!rgY^a<=`MKh z`bT;pSVWI3S-cA#bU~xuWVQ0Vm={-Pd9)GHGDgZ=c5{Og=n|`;E2dP-euN%oGG6@f z2;H)<=(Dea?$6)MRuDqjk)1WPX(I`oi3(3r4#1j6asarn6~0qH6f+lHj8Eq5D&r4M zq!QVmA$yF&e;v5>5RG z>Y>!zWpcn5-WS*ww>a4M=Y5>?x=rPy#6@2zrGZJ_ho~LI14f@FwxZ;EesoN1qkWBq zNc^D?cw3GVGGvnz^jlzF5UZ=0-BSHY`6mIn8h&$ZLERt&E7}%>-?p7n0u}Ojj+uI; z-I+Fs|Jm4Um0eQ&Y!fvZ#uM!3?>f>*%tPe~GIc#65#M(+imCF;W=s1zaTtvEnOH6y zUF!6&*0v!%nPGolOWwpq6XQ{cB6*NJVM^atdGD;h#;NIL+*i)M9CgX`>4|~ z$SHDjV{moOVL|AwIJTn|@DomjtF8v_5y_seDc`1~FH0P>Y^E-8$d5jI9Lf?+ZCGI~l*w8Z6|pkaU4x)ImE2$@DVIZGi*$!-kZ-{A!{rw;3!ViUM+jJ>;V%YN^Q57Fu0MvqZJE(T(0~?W zbe}Bkl8kh7xtb@4`B!dfj)p`N8w`;UZ+aHOVrZEYlup>j35sNEXmr;?#L0=cZ?z{Y zBsDUcBeS-21#V5lgGSSl{c42i8A6bE%b%e#a{OZWO6l=d@Z(9tYwIa2fTQ1g^kslO z9&@o>IBh~`IrjuhnXY3-cIzi69@j)wBYsfy&qL{fh?Ns3 zYu>BmIX9~d0uz`Dbml=PCw>=b!@OtU>BWUS6*0*)B%91;=VtoQP}&fOT!83+zNX> z7pt@+{c_4TBj+)FWnrU|V@8n%_myQ?VTt}GxzaD8>ic|)0^<)}mS{V#y5=N^H+##--Rz>y4`hgix%7jht zCz;;n&CF$te@ULkEFXV-!}}Ir*E~Im>m37|=tC2E#X3v>gl!6TEzWGka$tkwG{uG} z&9r!g1b7!~xnyDT*u{xYurtQ^l`{QcxHgAf2RlykFyZu#x}Mlsbt8Dwb+8}D*kbuC zVF%NOr1ZJ`;xAGCNR9GJhg~EW)<}6XAY~)AL$~_%@pq4bNq9+HH*2{w$)7R%upa{|6LP*PDC{-3~uA zhE0s#@Tk>~L2?3y(d0q8o0FUJ6Zx_c)*l(hW|m>&^y4EaT)+P$gmY)4v6qKQ*%XQ4 z`VVCh3*}_fj-kGs(Gz@|44Y0?SuC8VYEKc2;&Z;CJGMOnq8&nw^O^J79rU;{?4I+y zzR$!*NYpn{RJ%kJzcv-s;3tKM@hoMHigmp`N6}Ik)~%r?-V)0{LQoE*lu0#kF~m20 zP;SI)tlCTYy$-bTHZP*CbnfpEct(C5%W6na^uG!CBK}*o^KZbvMhHc}w|J?i5S&BI zc$=u0+^d&@G&$bg{DFu95Efir%cB(W)+*HIFp)k0uDy!b@1Q5M7v{%8inB(hKa;&( zKX%hm4L`d3+3+*{n)g1)QkuH4l7`AFchd7o8*wB$&fKkmIP7bn_D^Lz0z@R>c-0-PKXne|Y1c(rN%YVS^G)v~N~JY8SjsiA zQ!O6_pH)MS`kn~LuXJNowIb5mY0fz@H2L(}kw3h8IKd|JmJ<}|qZQDGRe3pBI4^oa z@{SM%$G7y~9j#G8_R#|+8uiCxv46x(OVulDL|49DBySM>jU28J0>`*)trx;5fR(3# zu33Gc+yf}r(HiL}Td%vt3VE7z>ym}n=6tT}KEHM~`wRL8<*Go!-M3uI4v$Z!;@2P! z@(hLkPBf9sRPRf@ZwmH-k8yR)ai{Jxzz|JPGC86Q!`!RoiDIu-NZKewX*NKSD2}!| z3qAL@eG-FKFN??5qHeVO2l%Wa1pT>eQuMaCu>q@?Su5V=*=QG)K02Ro99j#Rt%u35 zM^|v13O=JPP_+pgQepi2ppX^B5@W$QqsFv4fEK$3xRy&Tu%Lw)pp=G9Tjn=e?@YQq z#z$lY%szLW)%&Ai<5YnIqheC z>d)4M9*tqCgQ1LM`UU@TDYApbFvnTnrESly^)})U-XxZL(?V5b4HA*Tc6=E={J`*g z@5u`8=7pu6nMX_6YRSAm*2N>pJ)Sqr2-e4_-ypI1V!28336xf(JH>SZ8c{|6OiCGYGvor8C*+5 z$#$$PC-uCzgy$xqiV!5B7`>9M1bbHcv9az=g_u1fyA3*o+F4g&MRgefr{V&Oc1HYJ~Fl85NHk(7^kK41W!#~ytKOrKjhYrVfhVIWpW@un4OpG(q<+aU>cEG3q zW@Q-7%wPi_)D^@9SE}+ZJ;sY$l)o7(zd?Qr)m4 zB{%!Ow zZup{3SZANVFK4Wg*M~^fqf?h|Wd{$@v#4Yy6Ih63%~K^pqF_e+QN-k2fw*&v+iJq!R&YRTqx5MkA zq*LHUd(FjxsnpsE5_lj_dAlm_t7FvzQeWsr!5i;?r)uE;FivUkpMDl%ds!pez)~_+ zHS8`_nCeJdwl`M&GZEBiA=j!kDT_$8>1OX4p0Q{EsNd1mf!yCdT8NlT)Jfr5YmIkG zBlS%A7lQGBm7ITHE@B1Ha*{@v3JRlk$0s59mTPqZ zYc`t@PSNj!VuiJHYO-%0k}y7jN-!8`WT{C~WwQ5)E-JPdmgdNd=TKtV9?L8F$~vPT zkttoajC4UL;8rXbXoqf^2#bo9Ewe3>KCU7YY^k=@c4kL*CCO}{{-x5bq0CkY?}qIA zJHbMxE+Pi=vzWl+1o0~-$@E&t9DpJ}RXr|9zG&dy2yI{!&KDsUG%|fqqKG(ZI$CGW zKSvSIk|lcFH*{)K2Hsm&vwQL4b_i9ri0uis#L5|=f6@%D?Yh*gXpgWGaF&Up zTQVu1l-Tdz7!`12f(+~nAjdev4;UrFP*0%4LMv?UsNaOfHq`;_KiI^MT1}}NOat)R zlBmmLfg-;ZvRpnpM!3_G(mXP_g>T}gmW_Jzy;UU}7a86^OrJOWnu&)58H<>_ZneU& zOkjJC9vhx=0VZF@R^&iuhQDksw5?(Oen#4JknLaS4X<9>-@6(XvXr?Nh1dJYuP#Nb zyLv9&Zf@b!wK_x)5CwpBKu)=O)Qi2;gg&;x4WRjN_O7*aEEIF9fN4FA~IUL}cFdSvnXLu3^knI=vo+ zAFh(uDMXV8*O+2P8g2$bHmcm5gB^|_#4kqO(67`;qyBkptSU7tR&gm5BMQV|c>7jt zvzn7A3s^{d?rB~*gUDk5%_myb;{=3Q&W} z+nPbV*~-H*HGv*YAZZ9HlwkPCZp-DzgAQVamzOjp|3eaE;JuICY9tW>u|o-yp^S45 znRdlF+}@COoo}p{nL;|84*D4yQA{TfM9A|lfjDk-?Ke5T=Mn=&Xb3$?EeU!C`jb*9 zlfj8x6(N^sr3I9opfaXwivh$Hs%CYF7oPKYSCJnj?7Sop^CtM86dcCm^;uBe^7IIY z(;#v4c%_(WXupx#7>y66_^SL^qv?d~3SQ|`v>sHClbFbrpjsVu7W0OVJ0G|{pGful zjkg2wmm}cY&hW`M!@V;YK@`pvBDKSM-FD)b-)5+2$bDKX@FR6`a zN?-6L{P%vH7J~3|NwyliEFK9Yq+o0#cRFx0#I7d+&k3{yZ)}r~i*Sx*-9~(PE)-tP z4)6*q^JE4OhE%gae0>{bR(QJ!D6IybAoJwg-~~2SRxXL_d%ymhCRP_KNq6IR1YQO5 zv~p-hKa7fQCUvhzjI2wO)y!cSr%|f^MW4evUB9Yrq6zT`(>>|D$L+}qKztEXLpL0?U|mg=3MjsjJ2}r7zs@#VZZco~ajB`v zLzy0}!1Hu>-u;hHLJcU$hE+JBxge?5myDMY47EI~zYJb%=DC=X+|N{F$4#UN<57TL z;$VQKcf!l?W+@?yxFM$GFkj`#Sz7;9hxhTdKc!6tbPWlWLRGJ$8LA$_cmph>@8;(Q z&6kbQ5Zn#yg~(cMpYI>5q^|5sZ_LdgPFSFZmBb!vN>8GJVt zcyl6aWjn}%gjWgRkJG{2x2zR^<>5XmO~%l0kzuC!np>EKB@}((8>6OL$|)rdZ=#Jh z{Z^-g^AfdLFzM^HQFFPjm3I@FYc|$STzk>kXG_!Mb)~h?#z&zekDC9CLi{#;xH8wa zQgcG?|DN+wN!apgG;z;|f38p>%ExDjJX?8f_$s;3k(e}oWS2~`?TIRMmu89ty)3WQ z>u18E(G5NFzs;_?5&XO9ULPdByi2m)9Kl#MI267J?Re13HhBNT*h!e*<1gTr478~SAk_}uM6!KQ{xs`XiF1hG`ibC zpA&7sZ|(_*u7E-fTnu_XVtVG(xjI3tFJ8(rc^Y$dihcs!1ctgatqHRbI-fGy-Mu!M zk|J3FX2E21RCOE8IF?~_TxqxN>865~AU$CXCaA$I>CnprCwLL3;_(>1d#!sIjv0TI z;H=i?J1Q+$7KwvV$0-EYY@JjmDZIN$ax(M@C-lzl5I0yIjInyTo~{1j*IV2>7b)0< zd~)s2dCHpn7?>+nkZ-|oBC*rnFm5RM8#0+3G&4wfr~iqSCZ8%inF>ujAGd78G3p(= zZ+@`#ozAtboOtbEiVjXPu{_#%5m;KH1;#WR=G0VUlj``TO&0D*H~J zqMb2y%t9V40sF#$&<Y;B~46F&dt6gnlmGhdp8n(C{B4j zzyr7UX}zw@*UaNZ|2atjtjuPT8QvAqHH7tzowuH^&U0JpBVHV<8`k}t0&%RPkM7~-^3XYlL zd28{v;El|VXa6|(cMQp<+{vnd=LFu+@USmC?hr--EQnSoc%N8y)TbqyK1`I_hd(5~Xafl$;) z@z2m6%k6li?55%7Xqm{-jm&Mvj?P#)*~2&2eQsgrpgmkhjG=utD53tglNLVZ1bX^> z!OS16y}j+_(c?mXrN1 z?xoPU9sf@~D{jE)m&L1reU?_9+8z6|bE#JbU9L{vRtWgM5Cmx(rWJe)25etbV?vH>^Q~5bMn$`VLf} z>(LcM2g9E=4p&H}Ve}NKSu8b ze|Qtu|8d3cDWv^Dj^$TcXgHygN&S+Sn9@AU3aNLP$7zhCVs;BHMNg`qIv7^V1l~b} zGdfm}Wt~r*BgsGly&kfq>*pls?e4Dw(&DlLpQ;9>R=1Pxr(q*(P-DrQRN)p!Y; zP|Qb`yUjn4r}2aXiGj6Q!S+~h@z6CkA$IjJLiufU1j+^FWz+&M%b-ODNZu-%D3gav z-Z-5KV##vG#yC!9ll}20+5A8a2b;UgKVT*v_#+h#)@}hmCH?q=9SU=r5UBr1@ARLN zLSEU)1n@?;7-q!|&2VNZQSe4_*a+6uG;<9_qo(*4Iod!S1elyEu$A%uaO+eE0*L$7$8|BXD+frnfbI0yFmh}!LBq*+bRqm9U zudAV3=ks$H|46h&1+=@}MGKv}B)XSJETcQurLXwMSR!}%(0)qi5!ep3T@O2<(d$8y zlm;Ra0b7|;_0u|aL#}mGOC?2!qpWOfj!PmFKk&#BTfu&tKglO+#_WAXS^eYnR^hoX zzA%c_S6DWBoh~evbUR_FnY{x+X2)Vlp}Kc-JNbaY4`z{m)cjwfPEPPd=JM7Aw>~x< z4)y=&@rlgAkLqWqtlp&CL?5f$7Zn2hcs|b>qSta%@02#iGNms62VwhvO{s19h;{ppes*T7=1Bi1Slj@s zUxI*a$nTRBxf+LM{@tn%D?Qc=uJ4_`;O*Nb;4?H}&(yv-iTQuHCts{y^^@-e*V&Pq zB#QuwFIEs@)Bu4Q71J$pcsmtvb-mJobS;y`Vtz^u7Vb82v856pU9%2AeR7AZJ=3EC z#+NX~CiZeC*I@8gai9p7kuXeQU;t@EHgtsAMAf^5xVzPMX-e>=H!jbyKv^S?myJiO z3Tr!P$-RvF_E0amBl~ipDZ#2Gs#}e}8!?sox3XzgR!z{V&VGLLAk@iFB8olFazn?5 z&plMrB18P2j~7$wI`J~%{+%AXT+K2&x7|n5wHdX~NhCt>VW2)7$xKqH%{ZctNDvJ) z_CrhV?_isKYoSo2PtdYb!e68{>&(Ycob?CoyDGoCi3>gA3O?nd`cMxIxka1n@&G?*alYEszV zqMe#j?ok=7q9Xrhs_@o#Xk@;l&@Znm zpl^f9O6M9)o{V@L4Fkc(h02_}{^gbtZrzGe=ng z{AOxrHq=n-7Ogl7xfo!*j-hHULK6G8lm*m_Cyla05Vk<^v%DQ;Vi+BqVWw>m@JeUG z_gO8+GWm16D)@~M_~V`8G*Bo)Zr%0WqCP4D_-c4Epn(m03Jai4jCo_J-RSnbaK-6Omy7Z`s_^9 zdirO2b@=9zy0Nn8Tj$-)sPSz4w5N>+!8q$@o6rL30;D*i{8|>MzP=}H zYq%hY(y$+~!B!;|VULt>$OH6)y&CklgQZRRBI!8@96d<{ItASVK!oa_ODytfDe;*#QE1xR2PoL1;$&bQskhdHy5)ufw zK5*as>?%yf@NubT?>KPykkj`Kp!f>vUttJBLxcP$CP8RxbI=*D&FQ8e)Rg~w!PXn} zkt?iW4=qqk3qmc%RIp|-y}`>@t=TiS&QXl|4zcl@e7Dwjmj>&Eq9WQ5OsJ%@!UE zv+l463v3uV#Z42#V-hBX^h=4yco8_LVjv;QhvM5OdSIqQc-O)P&E8x-j7*ptUKds~ zPJd#l%dc@Y)VF91u$^(^XVeV%qe-LVu33A9Hi?xlAt}%Wmm%10N=3-633Pg$SzA&I zt@378d+yQc8S7KxM}aY&I-)d(dhvrkk`m^O8`ONXY5`d;*voeqQY9d}K3N=g>`ph5 z7N6SB^>e-^vUq*JDR^P-QDhA6Cn)?_b>Ioy^7-|S2n;%9LGX0n3Ui)nWE$GLuA}|$ zO$n$J?{-6IsS-@X1H~G=waHnwibuGn4g+b#fxEbywn@~Hd?nfPM#>~;s_i33#aRC9 zT&bTj!M(=gQHOzu5y;`nY!X018B(Xt1dGX9hO8^zGg(0aUTx9rU|JQcL#4rKDbzm# z+G~7uZNV8Vn!mWm4?)V)Nl26uqGWR1Nl)#YTYr3YLYXh@-U>}QPUZb$c{(Iftc-hw zDCAOo%up5j3O@Wke~a6EAop`8nSi>he(MwK+$=@e^~xF&D2U){YQjbwxXybfvAjZy zPFu4}A8EnWY^T)k&{l2ZLM^lF=1tlB^caN^WKfekFEz)qLg-URm6GN0-raH2EEm4@ z+Lv#HDdkkN)C6MSND$Mg5`NjNjr9WNRuDUf^N{cs_*Bq0!0P9P*9m=>Dz1V07!*XxsKWn||{+WlZY` z?4BGv->Dwi{-X~(iSm>+o0H_xqor&$rMb=#1;a&RcLUFzUA<&f&V=K_gT#3J5t!q5 z;iwmMd1?E{Lnb)7f3@S6SL#qm=Tn;Q%gyD(Y&0_ZWlO1G&?VgrK(T0hm7)ylQ;rdN+GE~Yj>bs$c5H0sBCQJ3L3P)52W6?lF8KtJ z7%-=c@1HhqMSl`TPsxR-=wDk=zMR<0WDkJL>R$0#5FbD8@7e{M1R{Mba7r@K@O1tW zeY+&EtJ)?QkNmMwI7`s=`}3u7*FZ9S`Icw5fY|u{1dj%UyDF0w286_(9rVqvF!-OW z_pfwQ(*k3!M;-Ernw4`_{5C;;=eM(Y|2AoK@XG=Bb&#p7-RnM_vxSH+*q0PelLytv zhj-(p3AusLlg+oOuI2LM#cC=c0pT&#d2C~g58;tUZfAW&E_+j*mbJMaG3=s({v{)9 zk<4^}xWbbBb=}=gIr?RshGplY7U<5<==gn^`oBLjkpGeUv@JldGpcBG&{{hG6Y7^K z0rnpmQxM7SpN{l=ec~5dQ<6ux3<&!qd1q1C9MSR2ywfJ>GP#)RK@q zI1GD+P*Z#Sx(~5Tec%dvJ9VxJN{(P=s+jeIFs-^6lM+8LGm5VUDkL}wkp?PYtL(@e zTIvx-s?22LbcJ?o%%)wF)o^iIJ4R7>Tg@{P(Nkrwv9H$dMCO`l0$Db77x z#^f&&w9G9+7#X7br}L_R#XT$Y=m$rE_Ho# z*CAo~Xo#w9JDZKEz(v~rfHQX-Xm}4RWZH^GsGZhpRm>BUs%Wul1BK6^fvk`=o|4z5 zFmDHAM3@pIiRORt^_5|9X3N&W2_6X6xCECF+=D}KcXw~xAxIzuhv4q+ z?(XjH?oOl4$DDI#=FYkIem~!*|GeG3t7@%Pt7@-Dd~AcFo7nvt`#wp-q8pr(u(;kh z#_!p_%!jZaGZ8$?k3>0l$%A=r0b12tGY9cFWb{Zf0df`&> zPTmZ`Z~hMCqTmxG(0-kX@;I*x$Uw~+R||kT`i!p|WM>cKwO~bM!s=mnK3PAUCF7mZ zZSqw(Nz!xPDaE5dh|3(>PYnes27fU#x!T__!nfDKmOcyhWO8znhWFLxz~EO^nx&kR zO}dZP1ze0uV=MG%)%o+V;PwnWzQN=kxJif(UI%L-k1Hc&+wbCYqryadAK}Sy4o{i& zcGWND4Qze5WS73Rtb9~nymS<=i{9RfRPfve=P$$zbbX+Ym%T+A%D%VYw10a}x-mEs zysb)J~sTl&WyN$Ep8Cd0v!OFmdl9GdUB{Kad!*?rQ~BDPM!sooJ#JLItl8J5f-K zCjrVQ$$Z>@uSrP>);y}$%@N{xZKAH=M3~1j;ZARlk3z?Y+eYS9aUTPpGIa6I;u)0o z4{t^iYb}G;tAC}Yv>cxpsDNQY%}eL_%1LeX+)ArE&m$+uNO1BWtpUuyf1g*iP7%+k+C z06OdIzWg{YH@zh-5d>P&2)!i>`lCvo^=S$&9SN&O97sOr=)%lULQR9X&&*_6&FGFCHtZ}M4KzdZ;@t2%GCRdOso(scVq*QI=Cmr z69sEnbbm*}011AJL(QzetmGsk3CTIdICFRE^9I`?WyF>XJ6v=49txIbK$c{Ldec~0 z)SO(>)O=4@2jCI*x;-rm0bjYe-_IM8C|}^de_kr^{iVqnv6i~pDNZ1O%T@GvP}UjS z+Rhl+DYt0tuxJ#y0!c-~kdth$~EBUc~h1qOR58tGhbtCLpKPZpDK*o907z)o>WE5?5H+NDNu^G@w2ZftX;tPIbdDfy4bih{d7xs;Mj@FEbm95B$)b`W)~ zt^R;3oR~XSiLHRmH|GPsAfR1Y6!1Lf0$O9dsH4CgD<16A>D{LGNK0%!EXBC$ymW06 zVNz0Ogq?02`tKU;kA=bO`bQEfwN zCUxsBZ4$4FW>cnWkf)iE_7# zvngle3D*CSyNDrh2b=sZcBlRgD~X{QK0(0wYoG`_3|4%Bsk+8yr)G8o7aOa0CL}Cj z)u=G>PHlY6!m&fYn(f;3TpsjS4unyS6DJOiq8zpJWGc@86it@d;27VKn>NnMoVr?b zosAl0lBm>i^{vxDrQgPhey$=7cd1sv&IgZ3vAT=g3@NQpDd?-?$1d?4f6|y4dbr>s zLgDCH!w~7poOB?s-ZZI+cv?tGbflGLhw60?w{6UqUbCX#<*m~#Y#|iI`d@sIgB!$j zB4W%iIdT%N8Ciz8l0=!UQ%7%SvD8X5LSUQaeSXx|-Or$-a58`iWPYkx5-aF&$AWMx@@Z#C)*VldISK7Dn1u&+U zd4Kvu2}~y2$`5RBZyKIp&wu2qlK6KaV6~IJvRq037&b&SK-C5-0v4N$_bZY#FW0wS z_Pv7VIltF+Qg24a7Os8nRC{&j<(u9F{V#aUxm6r<>+xM}S_vQbZ8?XD9IKJxsjNA? zuiU=OXQ!t+w($&aPTF7<Y;?_PM!U~f`r$!BlJdowth zgxyucPpqK?O@H!JTN`g5J`MfG)=;4q^&^)SCgek6IsnYJ*917&4?P)^3``I;JZlKZ`7uk8!KKolz+^Eh{=?aTr6WN=0#-m#;)GjVZ#-;0E*+t*)|^$S-^ERxd9Jed5Cnn;&+t(L80Z(GWpGHP4N& zHh8;PrfBCq3ilw=ko`b^O4BbBpkeZqa5ruc|Ds(T(K-f$p(?;gz-Val>pti<8}pG9 zy0qPu^(^R#`(e{gG6BB`%1dOWSTw?Cz3%)cC(Kz1?osWfwkD~`u0*aH$MWe_CGAi3 z#l+d;ybtiMx3I$V4(wSg;Tu!;n$}8(5n!otn*6WRvx+O=wIXac}?oHu-}I< zqE2>^++>ZWAGr#Cf0m$E(ff$#?myQ(%?1vkdHBoPxLt;hoU#!9UHM~m*g zrZ2PKVh)?wtXH@Tf5qi?jMU^-GXi9e{%30Wr&`w0M=?Q>;q}2n*zQPSPf@}xi;;9T)7%f|1a!Gm^Q=9b%iZ19M=_1FCT3uB zN0^F;IpbhTVqLHc-G5`e*#zecs6HZ5Scn&-au84^*z%4J{CzHVSocaUoGVGszS zjB#?~kx#1(F|^)EeSSAw+R2_8+A~s{W~12p+O$_$Oz)~EU%QoPg~?I1cVgGj1uIsu z@P_CN$Ht`-8i?hw4ODCl1#gasd~%?~YdEU2(A(bOI_-;dRbMrAqNwT%a+W((x*x0zX^DQx{ykd^lq;ILJE2ZuBQX(C z>c2k||3_}ol7hSo4`mEAG@YUXR68a4XCE%q%SM?AfTy^g-ua5WF)0o(o3)6a== zH0(L&z>=uS_)Q>PyP|NZ;&>O!@8vSL8G25fyjHK(P=6&qT(8jz%v!B}yljiC_42}K zsgOiOR8>z%L@GP}R#u`F7{NrmC!NjTMJkLHTZYx9#`Y1SDM2qN#wbKQYjw;(pdYkO!7(G5qA|S#58t#pr7oSChGtyuT9(RkDhSzJ9isYt(91A{}jHUccWoE#|wj<5lq5zSE}tFTdSAHmnFmQe@-MTdwQm%(jT zY?~u^v8*I#t(#t*CB{6yiq1_~r&(XPeEHqEm@ctG3?>=(JG6!R-QXnw)H00f3oF&^ z-t`$@rQ!sh+%H2-hxJX9QF}P09EHUI$bd;Z6qh%#<9+?SHIV9Vm~rc=_2Ru0-R!t2 zvGY=*PD*%n@0}ft4sG>b0cEEdc2ZSEODX+K-I6-~QI=rb~jC~!o z>(e22*>sh3Q~&TUY@&8caL`v;H6!&Gvr?mkfO!$C9?z?T^LTb42Ry3CU+lmS2gv&cMEn2KKquKh{=Z}~hWi1_ah ztpp%ZqBoD+Q!)yPaVH0I2BdnKV{AHA$y-eS+y!8w8%N=mCu74U9_dJeQ)a3nbJI*0 zyI6wVNgN$yq=Pb+s7xGc)d3@Xdn(W^svoGLmKsQl4X#u{kB8jHj; z7e=my)nNj;NlFB7s|pV=4UXZb6?Zm{FJRW?{zmzFffl=tB;3@tj1oPF*-#Rm9&RdP z!epu1EZA5{%%t>#R>s7;Oq~eMU_CJyFdNz8br8#$1VBW$RR7($ABdlbVW&hmyc7&tTJl_%*{Al9U7(f4h6J6noX9v$M}yj4#UO;14Y@Pz;< zs@?uh=0}H2x9x=yuqq!Sia6wbt%IFOuY?-3T0?2rH=+{_3PNd}ZL49=(-{USuG(J- zvXVkO?3@jw~P?HV3^jE3{+|sT_k!h3B9l9MT=)# zy}|=CqM9qE^PNvj<&QWUR|1_b2~r(Jd81AT*2?g;4QJnDDjCA$+@(My^*Oz=^Pyf^Rx zkGr|dyFdccxxUYFP^NP%WwE&8HAGtXgc@k_pvnywk^@QyZF{sW)UT1P*?8miaBVU! zK$%y@v3*B5z`4j02~I!W$EqQ^ojMY;7mguO5aPOssV_y1ZcM&dnGl+Q2VnJxvu5=5 z?BaTD2@ZuJ;8X@LF(#eN!@g%dz+zHKJDVA}7v>-DhDza*#JA%!MDzzF;Df+?^(H>` z*;ZAupo0Fh5~r|3Aiu2{F14O{X%vte`xs$&_1;|v@CD>*Bi8Ohq1D&Gu zvF99ha_o2uyu0F4<9A zzsANC{aw@cWjk2u?4@NgG%sYk3|PpJ;i7Mr-v8Y9=`B7GpC)DO_xG#bQD2HEfC;9LI@(A^32DmVkb*6E5cnPj@Z|*QY~%N2n04R)}*KE zDIiu_B8I5v_ggCUA~ls8PB0-L4xT=(#ZZ7Gu*YZ+8XNNF%ckXQd7gku&Wn`MOBQW7 zE-UN3w~T70a_hvC$S$yBC1@uGv)!!-0H{r-r9JNIESXXW7wj6^FIG-5ja3XNTn>Ip zGZCiZdrZPubvnTt!hCq|r{#gPM+Ip?0Y+`QZN%)8EYqT-UPHM*fEn>2BK1ns$2((L zOjgeR5q%BcX3M^Q=Z#TsKOim&@kH<){ih$-IJEnEi)Tzftt5gWm;JEk|Dv ze_xmdZ+M91BZnpzZiLAyi~_|L7&e6T*qpp`l#^vSp}%cm7P!l-omG0jSb$$Kc=CfP z+$ZnK*4En+wKF$b-keD{yVnVPUan~!nf&cJVRRd0-Yiu7a$e89PAuz0Ya`0`%w;O<39#j zA8=ne$VsO2Pdk`b%?9tkn^hC%;p5?5VjY0;USDHD*3K>+bjLCn$HxzUf}3N&o-IVt z-V@Ik%a2lADIl(_LqBb7NVgfvgnOiSZ`ztHpJ?-Pm1a|hy~e6E0giHUc{JbUn{5Rt zD_jm3=M}T0euGz(m+HJ*k&2~jAXseB#f1M|$mDlEY6CAG^1GK@Ub5ZxxlQPJwh$Eg ztvowjcl#VsxJ!gQ9)z5vv;a}bOS9kkV|IEh;GJ5e2JxF%5PmKi2eZKZY?~?=TLwwb zz$b3ASAH2ylD0s@Gx1t+F=bqiqEA02J$?ydJA>-qCC)2b6A5n$D_OMRSny8A;$FWo zE@ouhS`Ysv(rta|k`|H7ZRdB=fy2fZ_D6H6g5-UL-1l{5iNE7aGyVITiQF>O>hC@y z^=esZgR8;}|K4+6nsWxZ7pYkLx4WJj3ta^;?5!Q6MWo1YD%S~e-nv&9H+>Gw-cI|I zsxO!IGWuF`xyk3pD+kgI8;_UwmJnWJ9(VhP9JyKL&C3k;&)Dvem{7gl zXw_|wC&noVjSyh~f@BU9-{THIe&18&Kw_*UCu^v2wyteY*RC;KNP-hC`r^NM&7bYC zKOgk6KWx>7puxL2~{txT?fMGp-+3~OG zNN^erGiu@r?FPP0>cc7On?plkg3(NrfFKiD@2T|Q#SrVCf!PIQ4=JyPIB`Eba6Wl# zV>BKJFY0v*qE^SEL25>PH}5Mu*DQCMB7G|$Zxv%f2|Dn?qXiYCnNILo&DnD@yT5v9 z!!*?|SaHVcf(VgZ6^X3{zcb)7@EZ$ayeWDr=5Z(9lU!oRM5*JXNxfP=?4$C@y*Wy@ z?F1gkAx>xAm~u2a*hU~#$~i&`4x^fme!o2X*CXMLLmsgOFFr+|V5HbHkza3#`v(lO z)O}kR%KS(u%Gka{jSVApgtOrOYNA@!0+~_+L)OxmA8kKOY}zD@wCbu?&!d?&Z1vB+ zXZlLSUg@ox2#oynPF*yQu}hPB!K)tQUknxMZiQ`iuL>_id9)S@BWz=qoJY#|a{N;; zNT!~5MdCCeXRs$n+!WDf3am|W+Rbc{wHi;;QX@&QOxLvQ8YX&4*9uowqGGYTpK|Qj z{p9F=8O*7X;zEU3d8E}W=BfMKkv)JRiy4Vz{nat;N5k*8UZa+B>=jg-uMH!9M!Xjs z3qE}QX4*a$6>qr1)_MU_cOpoKLu2iqKI zqutAU-I>W&(fLtjgzq-*H5PmvjHI3>^Xj7cSwP3}WX3hH&im12M&nJqOpfnB_5gJ~ z0NNo+VqaUATj*nw{>FyIALN+gVgqRlInAE@=3stu@)8aUav&lnc^mj)t)*mlrJnv? zYYl`4)*@O(pTgWo5Z!mZqc|aTKTXu*iKQ%l86-0;2)}G!=AWvS(69&3hOVfmySk<2 zF$Jq*8cxZ!u*jY_vx&VBsNNC$3Ji`5;tos-b9fUzw+^^^N^rz&MK{K14CW}%)c-T+ z|J@L$exJfytf$B(Nkjh^zQH+!R9ANn#pDig*k>pm#!j%uCTiBNQ|;WK;W;tq7=ox^vX5xu4@P;nhJh5eVc^63U{b)RHAC zop3!E8OoY5X!{m3&MPYYK-L$}S}SUO8g$TegVrdH#_!Owk|L2%Zkp@3WJnS! zERJ=pZDf5aoX_=Ezi+vI(^E<*T8_~WrswkTnn zPg+LBW{7>wAAsYm8*;ep7+Km8OR*SAaNn)dJrfWm;C<=Eh~vABaQ{+r*wB5A_?8>z zEXU^_*ICD%B~F3F#;MmZnLk=ZzBm*kIuSI8yEg)6N580}-K5?6>_?p^)mbOGdxjQ? zY4+xFJyJH7A??j@@#mqCH|Xf54#_% zK>|9Sk0}@y&Dr|^WC$=QMis%UUt%MFmKc1=@$^Z%S;>`Uo0W<$Ti%^q&Q^4Mx%>FC z_q0rG#v62;C2dWB$4Igy(!2uE08f;jF~0rcpDJ7akdki|!FTll@i<3B;LA*tB27Y= zdWz5{bTX8bSfqde+FJ|qIE>vlJ={<5{FSQ0=?JgoBAC!L1)^$Y7defPQGkHlqrTxr zbZ`O;SpN1KV4aT5ZF~NRZ}^11Dv9~>Yy-dzoT!C*)$b}%1d81~l^;Z|xmk0s+N~1Y zoWGXqa4|DOZ5~7R)F9TBL(i+>ohuWM(64>>!Iz{S;>>0VIhVpeI20@>`hEV*O2Mpc zLGm#8rDg$)Uza^zu0O}be_yo19q4c_cwPFVox$gi{G0$?m3aNMz7@6zD2)Bve!&2J z<0p{o(c~FFmqlH8S4kNXmB%3gA8TLo{!j-huAIyY7Q|^8>inXUWiQ z5-gG<&#NcX0u{SMsP@V>yP46EK$ud{$@t{@Go6{w&e&t`6Q|40&<_NrW4W#$v%L#9 z@*LG=xHt>SCp*LZH>$}Q}|4)B1j=vGcAF7})vNSQLUuz~UO)LiFpEBWC z*mnq!T=DlF8D2RJH}Rr3z)fSf35~=pS$N05vFe)B38PpU4Mm0R4KCh9BFyV?3j53y zMrs&=kc{iXyvXYFjLy@i_2<1fWS|tJa?&!#$QsTaLOIWpf z@;gbfm@!gd6_n*=n#>VcOlal_Znqix0Zf2Vxw-BV4-CGukg1B5f_xrs#g@~?#*7+|$P1bI(Rvgx z2(o_Z88^vmcRab5@zyWU2Nf9zg0e31Wi*9h|D1q~)P@lTnZFe7WjmR5i6jq(F(6#n zjW5l%Zi$p1+YlsC*kf4q6PZ4rP)S5RF3Vsux+ zV6@vrb+0|*CxO{-9{gTwt|)&I@x$RN!SGlqvAjZe?Krz9rMeV+POkD7ca7AY25k)k zK>+fWKrz2lTl^AQf2HNv)Kj+JR6aU-gzc%IE(XYdxrFRJ$UXq3@2M{L0Gy0)Bg?nu zWFt5?EZ=$E66hD4II?!(2C3_OqfP5tRk-?u==f^-=!reYqosx^)z;hTwTJ!~O13=h zc=1ATvw;n#SAY*+;?~}Ygjz;dvFC0yfqnL4*z&v*=9|7~?iw9Ma~J6MQ+~6H5fIG7 zwLW?2o8a^Wzs8w)tnvMV6=COb%KYb`&&9r2;oTTyY~P!I4)fq~iFmd6ESviX?jY^P z+Gu;l(+V~1lT;}tLAb2vKCO#BdeMQ#CIgJk*df$JXJ2Sn+!M@Nnc2)h?WZxakUvlC zpNHE@7!u);_3jSI$+If`my1(II>Bq73oXhhKf83cj~jkEGhXE*cCO4rv;fhg713!s zE#VVABgaj#SpUI!oA*|hmLXYskKp=s8J_(N6be3T=ppzq490z5iKp-;wxHcl#sUJP z%4Lu1dARvxgo`l3pt&ukY;V^pI6IOTi?GUoB#6L!UqRtmHF z6-Ed9WmV z52D0e!tDpGxyGcjm@h=t-0qFO3>=S}dmT%HtJkIdn7k-nVpzBK6&wo(p$|Go?w8LZ z-FnX#+MlL~A6IBP{j96UDZJZUvnVhP4y%6zn?7u;o;nVDrM6|Tax3~zd@Sp^9{~zp zbbGal*T839+P;pxOPtLRGT+kvnpyL<0;D{>4^%~osg?eEXW&=#5q-p|W8H&Z=d;^A z1DDr=zbrOLpwbL&omJ+JzP5iHQr7u)Y`FU7O3`kMll^?j0BG&FOI=HO%pP;TE0Y4n zR;*ubcc#G)MqI7w5R?st?J zH!tP**OZjub+)~-&Ze!PcNEeH*?w)u5$TOrUR{{Ci@D>5S81vmi~%{N(haX zW<)C{yqvO8%AO=zvxQ;~9j=QaNx}$S2~U9;NvGSmFTPvtm%{1J$dnKZmLb6Ce$hum~Q=%gQ#|FeR^A%T(al{t@ zCk@a2-8iA=Sog3M>^D1S{`>4#kl zW6ip3Mt#)WLZDdkIpL5Xuk#0E{gWTi+(Ra1VHj`Ml~&5BFv1w9lvPifLpYjMMq(xFu884Yc_G)Od?Kz zqUGBty9zZw4?67nySqVM2T50XT2Y4sopRc=6CI_TB0 z0&)#@uqvXBUxDia5|(|l#~k=m$x*#jQI!0hu^f+jJRmKJMN-oVI$W4t_mav9|31Y(aB&!u7=3~O&ZoA7Xy2GzQCtegcnbhd>-TVjf9SJ+R*-)sY&H~#=d*pV!&AR|!6_MU&&wGV&&ZmFo2hT#>mKA7j0S z(G}O~ZLm$#cDD9t-X+Irh^)+{%j>mW(e;k7CW{ou%~}h)5E*ahQw$e{sQaZ{q7J90 zaX*u?qf)&S3VUHS_wnW*9ot<*IurWdWH?W}I-QUJd`_va ziaPC3U|&tHaEAg0$A?k0^r7k-v3}X9cNRj=B774Bmh3}<%doD$%hH-J@yy0^Rc}$D zo*FVyg>adh&)V6}Rw@o#WO6X<%1SF3W`67Ue%UsDVjP4;T63_s|Hz88k#)Dzk%;Nm zDNY*5eY}Me_Cgud2kyzdch%3Y63@tYFa_6{|X1C#; zu@j4i`4E*KbC8SVIh#~keA8&-PH@OMerb$vnzdA3n+RiJ?27M}DvElsW{L}h{ff2fTHW_%Px5a?f zq*Jf>gXDyzv5-hKXhV_G7JR)k=JN~OyGiHe@x`AH2(+7Bb|<|O?lzYHs|EQlK5wNC zNkdUJbjElevuOjd8Y0wdo-a)cHR;QA7jWQs%^ zhcOAO#E!)Yok+G)#(M_SEw$IX_0&cYOqoRO^C6jv3&*FsJn4p{33ZD(q*YbcY%~DK zw$pNoB`v`GWpS)XLH9 zb|)_keu)BViUoKLX--+o7iyh9(#s(zr`Mf+$xK4tG z+BDZEkt><(@6*J$1NJDR*SbA0G9T+Og8LFYlWI#6oR&y@aQ#-LyhgAQm@!bhv@C0; zN+9C(J8gz6G0W^jPOs(&ZfqbUVg7qs^Q9O1t}Ppe(jb(TZdcYhSaIm(M$^`w-MhN0 zHSSP%b)ry&Eia}>hS|Qo=w&YnD{472kcjy{U)1y@q59LZM@$GzGq@+1GUVH({ zw1V!?YlT|8^u&^?{D;$d@zygVR%S)(18fS4Swnhn)nKiw(E33MAH>#`jP>FM>jSv+41^UmPI}I|V^1kMcF^}EA`*vvQ*l~t4?7vx*i@0(&AgZpQ@G1kvqoKHQ%A;^ldPvLXnGOo@&7(f~7@^}Mfwg*swpvv%dlLMlcg)}2$9TwH zEvQtY7yU`*vq@z|&yTlhjACEYKz?P8r!>g|G)&ahDN&-8xGYdhyH~Fe{ZA==dk0^e zHHSEfN@BuM!I7r+?wm9RM4QS>ZmyQ_j4>IGQCAA;SZ?v#CMs3}+8WKgb<}-BpL;hFggN-|ux1Gqi#o9sEbY@;LD=VrSGwxH2 z7|-t?+geUomj+5fF2g6y=(vm0(YnIwwHb}^OK!uruJdwN=Z+GWu=WVKznuJ zTbHpvEwQmjRr~2C&SnzWAE`Uwv^m1%Y;+0*?ElDo0=l}tQBPIZd>a4%xm4dl9JcC0 z9N0_znN!|qR@zu31ZWrl?Sd%{HhtZ)7u!qs%i#yhZl%002$WFaOA?5krq-@%q3J0R z$jX6hSu++LM#-TwQI<-sS=+p=eP7+rrtqE}GgqLFOJEl!NbuC~nrFgw>RBo^ycGjp zRhpav^{i$<9f!>LixqoYUf8`Ou%uL8AY%2Km+Bl8%$LZ4QoZl93GlEg`f-AyYgyLw zC)PT)yoiKg>w-LzP;J0Q`5^#v85q&g3l&fGu|UD8770#8#w6@V^2ywm(&oqzXiB3~ z0ZZgK204kti;4@g5NLSH@u3d&X;nT*RF=s4tlbFFjHo))kBHkQLh}a>yH#{v7l8f4 zWc6TjSO0G*j$sC57edPT{}jYGve4xw;HqG!jBd+$aK}Yj7chAeA$Lg-)x*L|sTW!m z%obQdM87rX^ZG|cIQXVyHu(H*Z%yU$ zvHY&kTty#irN`rNC1z@NlKduMwyadyGM8Tn*R!uJ%KS0vk?1wfpRu)-nD zSVmIOtLAhv^6T;tAX~Je_QFNQ_I& znhgPB6uY{wjDiNXNDXI!rMm|w#v{>zTzz5jdvax>jQsf+!Li_OXXVWBCy_>!+*w?E zrVI5nTLR%?1Ga5L+aQz@j%MTHnbrM?C|$mAfqFf;Q2_!IL_2%7oGK;9z9gqgrh%q? z;-37CUFALTRM9HxuHN!T0J_1`lHsKJbw(`~ao_DPZQ3}wNbb}tIQyunrkOO0WTK99UTz9-5X}96B%)#3|~H`GcL;lgPKdekPdn47-B)DktZCPXU@uQtP=K z)r{vv)k{QJWothrDzo(LEF^BlqY86@&VyPItlMP$ISd?1M7BIbWVKaIL^(>@xR+{?her^mr!^cKzWvZz%WPz{Dbl>8+lAYxB4K3IJKs;pF(FOCFR`}qo zQN$9%PXCJON@gQt1(q7id9LuUNrg6%D&jRz66w?$Z}X8Ml07B0za;zDLVHUnqGf}E zG8cHCO|CV{QnPxs+1cjRcarJh{PUVJ6^NpBJc6cYI_jnwpP}4j$o*!`xKLGN9h?)H z#BGVJbQm@COtrc?Op>2nl#?0sK@xH(Cw6PzAbfPF2<^7<{EqL;kCg> zTsod9cD&k6S1_bAgF06wMo~uXhS-aSTX9kdE<-!g>Hk`*>Es#3fa(X(p)=V%Dj2K_ zyq!LNdtLd~)KaDlB33cJe`+}$u@o6uwgz{dvsJWM_u7YC+R?|=3AK8^n2$gCdhE&; zCD6p4os)X>+sm_sNQG>Apkdk+X$Omz+ezb*6Z1F-Tzsu$rS&yM?>-Y24ux&!w$)hk|I^TL# z^ZBX%i{O9G&DQZir1em`RETS*<7J#6Um9_&08KLms%2%S4pbv1w~lG0Oq$;x;v6Jv zI~e*BKKhf;IQJN4OmDM3-VqUHvZtr729#>F4gt9QmffY@*PS@qBZKT7CJ$kG?@nUP zP2apqQUyJj_Dpr2M$ZRF$Y0Yr4o8kc)Qc4@<8;09PyR!_{PU!-LHY{Af(|Y6<|mPoY}V(rbkF?|qAc`*psTHo zb7H1B)RAyWldE;q2$mvPgSmk|79RhgTi6iw*sCiCgAhn|6-xtB4(n{uN6qoN--@5F zZEWZk)~mKj%P#_2Mwk&s){nWXn-YwZl6&acTik27@2!9qcEoji$SpnNC*L;b;_;gm zNERvn1LJ>3QxohkV=uhfWUw>Q?%K6bdkgcnyA7{^Fv5I77xm!g-Ck0PY(djo6^@2F zeS11A=lpVH^r-P6ID!Sb=vYq2lpP^Z4J+zHmlpxJM{NaF1wfXg!v9YH*z}Emgxp$g ztQ36!W%?Yaq0q6T#{7d+LA)i1I%0@cy^hS<5H9{zS*$*$Y`sfEuaa*u>-<6(>zpFD zlNL3W)81EVr*~eQk30Npdl8GG$cd>%(;)EMc+sxv^kg3WrMXat$2u7Q(M6y?d{v#u z%Oq|OOS8Y4QOr=;(f;XM>+edx;54<7raETY$6^zvGC-PGIjO-M&d)#0L&X2%Ct2`> zW+)fbyS|J9{AnLpxPi7v`WEZ>i^rq!xI51_g+33xS;XqZIUS_L|!wNlMgMb2|pPOg?e=xl@WI&wqlFp&@hW zJ3iK$n%pZRs%$uaVKyBKef!tXjXszD5`x)lxn?H3^bAQ$;=Z!Im z_$Yxm`BmL-owhF0nX}Mu2V;t2tC{HNGuZ=Qah}>eDhrzmIjxqy>CfC`xaC?)FW@S;v!hmPAl&j|AiZ0Cig=6%^;#c%`Kyp-Q5j9H% zuXOsOZQy0FL_&8$a@jO$yR%=~)1gs~`LvIHdqw_eh6hM&dxLN^Ufcb7(RFHfcu^dy z#^xCRj`8qq7lhbc15bo-7S+LTBH#xAg}DrgZpL%RV!_9JUNw%{dKJo_ifb+g}`I)oXNpz z!UV2sD)ElA@qUb9uWKG(@FxGBWhZ$>Qycw}nv?`MbNb7N~#V*)Z#BWJP$@n%qPy zcArKWl}*hcD}NidQ7oPRq?);u2%GA?pLlQ~c{Qk7i{XFn`2nj^)IiH7a_rZRa7Dtb zRIh;wmR321$*26=g&uRG{}9anP3JqlKmyiiPY5L*&hk@xKb;Fr1hXO>fU;l57`Q}( zBodFap?8e)HR=RJ=B~mwSuXdTS}%>1nCs!>%ORQ2yhNF6p2Cb|O7irJ_%B;*t zPB;DAx%a*kf>Xr~fNeGqBrsB>n7iOx1hb1=E90--e~<)4bW?oH=~N_>=10X+&>C{B z{0j@YHhF&qzpUNdLdm~@jg?3v`bxbbTR%`6hR%|HYz)m^;f*jfMKM!{s8@!fT70dr zaIZDdR8(UiUwEs)?!32Pp8>Ev`G*_C>Gc{Ib5fhgVo@vT__%ys zEh8E9HHApPlKMl5l6=HD*dXbwbEM1qM6Fr@*sOT@m14db*7x%`aoF&ECii>7g}(6u zYv~Cd80Ob6F28C^FO}s~Qq&Fct7>lHM@FsBdVZI>h~Kwl*QypJ8o{BqpV<@IwWz)vSz>O;rfH|362J z!{0+lRq?OgPG8|u7=5+RTZ0Mjug<5#EkSAL)H@s`KneH?y;-KNPI7XUH#~BokPtS@ z%+~*%vOfFmCL+OsLKWWiL?>@HG)FF+kyJqOcSE2~t7vN^C9?GbiKAH`xOBC26Z2!1h0o6(HN0gqaQG|3#Mn zh72VU--8FM9=V;4&dpD=vqNNPRutW~sB1@w9c)*`bjK4fm)?z)u$`9A7W#Nx*y!{2 zOZ?r5{j-R?!;oX5w%L2+c^hu*z%Hc$8}&lqVPiaw@iR-Cwz-Gy34J&)$&Cxb0d=n9 zW*plhQ`GG`|Hj6D?2QmToU~0@DH*jWtp8E@@Mcafs7AZtw!}I1I;w@?a->0{uK(B; zK>N0n!LY#%2CJG7z&)j<+xAYN7^L;RR$ouY#H*_VT^>x9m#^liG19Xa6kRPrUuVIwVOROlbkoi(@ z=c=J?uS{TG?&Jo_2tId&!1#L=nXbiHCcVD(wKj&LS-?!=zDd?O@7c5*NfArFAMSTI z|EYo)R=d9Sejh-R2*&AC((!&%2Po5EnS+in$H>=mb<4Ve@a#uoIEJy?{rCky$zQP$ z`@M|aaSNcQL@u?;zwof@2P+b)O!Zj@{A4CjZqhu`ZHrz(2F@Dpi+v|rJUM1iWpgR7 zPG&s`c$NDjf@|hyubV@4gUTbtat1k6*cg6K2a=|M*TbzPTcYwQc5lzXJ8U26@)jrF z|Hs%{hefq^@8g1iNGeDRh!RS7rzl+l0}S2W(v1j6>)_A`NDs}>-AL!q-Q7LJZ_elY zo}->~-s|`Nx4GDG?ftB{?{%-WpGOEPB>uoPTdVsmgG&vpygX=GCrQNo{!^FOjd#E{ zE2~9sQ5|{c2W}&zW}9c^Hn`)n5;H!He%FtYwLXb56Ww_h2PV+S|E+vcPZH%{9?G6C zf=jno)8MOB+nXkjbmq9JkeT!!^IXQ>!noAyB7WJX<3Aq6&4R>~r?PmT{Oep|dORI% z&UL>Rl&xPoIqTN;E8%=xb_;y!2kA@_{-$vln)gQKRhNv#$;tc!{2jHP$VP3T^=BfRWZT->GE{q;1Z0`^{)+jpq?%U`IT60weOaLy`Eyhb0e@9XC>n1z6BcLa z^NpO54W6loqkat~`ShyD?d^av5>bCT7l)I)zp#owL$*!-;K`eahU4tc4Tifla*y3f zjqvXN{Q3Zjh{h|-BfIb zQ14cD84xndUEQj!zJM2*%-?G&@gn{Y3e@k%ZfsH4~0jVd^d zZ>UCKJ$0r%FIbLIV&&l#Lz*r&tt0qgSqP~4F>V}}H9y5+qqju;2Jm3qod}8~a(I-~ zs2f;(`Xxnip~Kd#!E_%)pN+>mD?eus&-j#!y0^D=_mFcFP zP)sYC0QRvQy&<%ZvDuN#!yAYC2T^8Qh>a;|RSJf#typ--mPrYj&UVknf#$g+jI7u` z5C-CP!1{fdv5jqLq|jOH3i$q_v5tyGoq(eEZl1lhu0aGUZ#K6hc>!viFTdR9cE6PP z$o9gWB#1js$r$vNY5P!JuS9`1hY%GCOg^lFvMx}=ZTrDgdN|F;?=@m`*!yW$qISiU z>-G#>j&rvmjkLpg?Xex+c&gO9l|nOgyTWG- z{AH#7O4v=8kCZ-JYgc(n&);|{PXE}OdChN1h!#Pyr%zs04bB}7+3Jmw>cn+=&wKd<-L6GN4z|Su-o(CJ#J=s?%Ri?K zr;`nx0kW8t1Ky@Do;9!000%wHug%}rJNvii+Z@tXD(Kp9>qf=6r6%7+|7P!>KX7yU z^LqY`qaYr-cokg~WeYeK+Nb=icywfv;M}GxfM>tE z0`fVc``FPU_a}X|-p`^Y23o)T9LQfa5)5Ap1U9ZV6H4SHOYFC3v zTacN58_`=JvEpdHM>`Oj*xw!yU#AzqVfD-u8=Mk{G@grQTd<=BsaPpPh7j^bRUVTDN>-ygNOge?QD*x-)8UR69UdNI7kN9!L#-N{Sfc>2c$%A51M0 z;+AG~xv9-I+l(L2-VEx6a-MX&s=EF@p@m(;b#ve2x*F)Qtg&D`T_;+lPcs8FVBD5m z>4lWCyGl#ZHvgA7~-6CL4}|6u04W+3ZoMv z%s#*Xu`WCIo5y!v0ueiXG>NMM{y2Mn_?tTx%CyKO8DV-WGQskmXIMNvHN5MFCH2*N zs?SWlm)*ZR28J*Drw-p2E)o)t*Kt*$U1XgSG*8Ba*eYkanps(3QEXi{9R{}NPkPPZ zpjra34FQDS+{wV9C%E60UARo*Au&$0fpOqjDQSR~=;^Zk?=}T|{=^lN1RxDg-7jpxf zq|v;fG=~TYoteSpJ#AGdaUV_MfLX+b8$u_U1~QoeE{DXcxi7*_`tID~aOz`-S!O;n zY1Avnqz%?TIO;=&kt#l$;g7d_zfPl}#8!P&5lN-Y=D-;o*7X%PfbLr0~5$!F?Fh|CYei_GexwF(k*VWtq2 z-J7=WNf6U45h#oqc~&5(C{Q#oVs^WUj_6_pzwW3Trl>?|ssvxna4rn8X~l{W@u8WAC(w z_ReiWn4=9!!eA`Ms`V;}$Uk+IPbZs5*FAXdquaPL!-;;Us}%f@bC>T7*;^=u?f##8 z0ervHy0xgnv3trCovXU(zw{!bVd9fwREhmfAOEbTThg2`Fpcuol6#Jbxd74abU0B}`EKHCoDIpT-B-gy2UUr`^%4GyQy}t16HHXJox1dLdMXDV7Vvd)oWFmY zk=uN&>mmQKnWXt*`>K%(S$vcMq-;Oy`AtV`aL!;P+X!n`r`5@>2>006vduN2X5R8Jr1@xkgk}=oXP}> zwiGC+@JZCpwe%V>&wSEtt(xR~Ciecw~u{dBhRkeYs+M<=&Doeh4uer+S zkq2Z|bc48eY+9lb&f+rsKxA66b=%vqP1@__OyrM8+?< zJoFCD%C={Da?c78r(5wm9k-2~EPmU+?92@n7f>l0i3rwvUX-M?vh-B5UaT@wNX+?@ zRpW8kE8lzEHN1-5R+%669*q-fYKxGX2bhi-l#_zPMVvo{@*#FisC>HL)YW^7Qsfr( z?_&1fmi79X&;{k=3tw8VyUav~V@bQiv826i3YYC(#Gv<9AkkopNK~ z*Osf)i%nO&SvK&+Th0l3Z$5B-yTdDV*Y>8&LdAQ^9)=LYN)D=!UB@VIJ%ZK3QR1WT6nGP3`Cy`{^TFOiC`ZQ4X7r=22{0wauyD6h|1SIUI5 z=CT2&<_4*kSU^O+qsz^Htar_P+UC6MD-YG}&hGv*7W_}I2|kGz9MW%g1{`}l`c#u_ zLBHokIk^@X@k6XfX0l8MCr_YNY4J*vvEzPh-6WG9*mk+ zC&JF!{z^xO#${@Es-8#C^Pst~K(jlm-qn|N;$E)*7Ae+{lG%ZD%RIphsM z^ZknFDu;Ut!|ZVyJXqftD;>@KI~u(~S(KD7%(zK4!ey*`6|ohv?hA079Egx)iFrws z_&xNv+uP3GAQI6O`*?W=b){5t0k78|+|H$u&AlTEAmY)thaF#`EX>OIt#qDUA@e*d}j zhY<3fk<9~UbzCBU{&vCovC_l5hAv2vjF&XP6#LD(1bbd5tCAMJ|85`MH<^Rp=mfzH z%gj#J24Aw$*=mtiCLMbPO>1L4$Lv8o`r@O2!X-98u|Rip+m4 z<1Z`v`=x#%#s<~>lDw_K>-IqOA|`6NH3=gcIjqLXsKWiHZ4<$^`9%qxkTo>hB%&to zMksO5c-6ABl;9DOqRZlHr$AmcJo>DU<#0HWyX_Iq%F?h!$X&z9Eqdfh7M+j}Ef1tS zKa7l!h$R;UJa?viSs2Sw1q0H2$7F!%HPjZXvHQO_z}I{AtHuA{^W`ujolj50U@T`r zznGnoXpqL8?d8L-otzJmE%#RHI(G^yu<0q*z*{+4>Vz45V4w=*&1MOc0RM;EN^T<8 z3?X5s3dUE3vCJBv6+Xg#GfA1}nO_x{_7u#itcw~ktO(bmxx+jaY^YYlmSM0v+}lM^ zV-pX{6^EevEoE~GWmL`g8t*jrJk9aUrCga?Rj?vvpG=|xd)`P7NLyEh9tz2#FUgX@ zH>NQlA*zCDI!j{|>QxcP+uJ75yK@dEi_O;R6idvO5bc>|bA=1p88ipOzwGwTiqn+h zRKg7A_HTU_S+k?Q2pMS*zy~L@J|y0}Ah~whe)esfYear4U)Xc?p!t6lrn~CYd}1p!ad?E<9ZYU(MWV{3=_;)QgbUbmg`!$ zxAq#cR0DlRpnxyGIKy3CHpx)8zYCAfD?b&2?XUh;z2qx!*` zc~@_BU;t!?3-5ek?7_7CrZaZ?w@j0kWmmo3*5!FYzmhLol;Ua2mY`*;@wY>oDW0?> zA*hYz@3E$FX)%LUcL1>uzr}L0xD*!ALn(FyKC03ZMKRUpwk311>ZZEebPUfCw?G6|qT zK7)!WYrM(5T~>kjNSrD-m6|OYjknj*lIF6oF}YTd(OIVReeZ$@je1_a^3%8%S&d>e z#kaRYeHU;MjgO7+?$_6cH%zNUUzqbC*$}?^*n3&8Lh>r2uj6~B%Z-C{RglAl_Ys3N zm$~2aH~SuZr{N>A!A$br(p+W}Sz~f#fI`SCo*FR=yY%)8OZ)T*yHSqaPU=a$w(6QV zZWM+fTaQnfXx!$?4|@5G&K>ft>uv=@Vy26IiG6+M@7dOI;)sJ3lWAEhzUakZOOwQ^ zI0Gnqc2f>UK*nl~dLJqr6%~cQj7vADiGS^hJN)sHqe|xI;`}q`{RjJ=pq!-o=csbp z_?T0*SS<(k_2fF(GrS8|Vb(RJ|%xRvSE?jwPIh>OBwdN~szow>TKN>_t;{Pl#c z=V>@Cl(5kI`}lB%qSU+Ld{qSilv+R@GN`E>5Bo?By^;aBtpW)_KxYy=FL!2#ads>$ zmVTOwm^qd6E2Z~C;LOA%6BWF`TAuulDFKmSQ56og+t|CHTM&It|NcA9k6}LW^WTh%N2Zg7*$MkFa3DVdxn9SZ`~$8`N@B^XA{AJl*veqR>I&#HTT7Otei1JuyHVS z!?%!ZHtOkX$N81$!5a01*-X)iCO{?EV#mM$gG<|2-X(wX^owSZoz5?3hf8Ocx(70= zGd=ydMzEF{s1iQezHm~@T9hQilub6f)d*-wh_UCj=lSiDhlAMbo%Zhheu;QDzDd2A zKGc7C#MhX#I;_xE|Hh5~si8UiQ2~YWY@L=iAbGQDCFM{}aN`0!%onKLu+{=%X~Y0rv)f`l-Izx`x^2wb>x{9ttQn!(-^;gYkjBWNHTwrGwAzP~cawao_}ddp znQ?dJA|1uOGx-Tj>)3kduTZRJjFtP=1Mb!Yl5YB=7+KE@MU+{tQEg#&n0uD+$Am#= ze6{&?CNlKUZrDPgs$O#0B>I&{vXI2oZu%x*Nab>zHSPOu#`uE7zbxQy)D&TQ(c>-j z!tm6?2JTVqY4jK0O+WW>X_=16QxHq;NAVBdIT5kn#J3beJWpkN9b2=|W^TwZ7mwE) zA}yp@%a4_i=iyhvJ08AOTUUa>;A_0&y3Jt?2X$py4xhaOgPgW887>`wsf{*u$^Hz~ zl^%Mwld|GMdg4ch86;O>ogFD}3Met={+H*3Xr&tN&%qdhiX!LluWxwn$Q6Z-k>_p} z@arzeYA+U}d_x;AMeaUU&ckzROeO?={tN$5OkMyRjTp2n*QNr6d0lk`+ql(Mf1uZx zbxs8lk8;gBR>8CJdB&Dq2ZHkNv^`EuI?1zS)NJw(zXEdQW#5j>jb(A@fdL*CF)^RP z)M_EP&!g#8P5Wq|38h13U;85&2iQN$9>LmQP7Qu?5)F8awNR%x&$~HcBdsgXiIU_U z7@Akmc$sL@xwSOt)~q?|z3R$Ji?rx}iLG(BP!vmUB!#yNv9jyXkJe%vCmm)z;Qj1T zLph#Rv7p!wr)}_j==!&u@a?pj2xF)TFp|V$hn?Fn~ovC zS!MVuq5Zw;_!-r#s7N{>F(dWSv{iMYXI=(%J6N*FZt45j-{Are*p9~hP-qxpHGnS5 z7KLs!rqydlnqFs>L@(6VX;&k0Ob*%_pSHn^C==cI`m|K9?FmwJtDpHr5WPwAoWv8| zCH=8bZ3C9?aK_UlBjh_5e8-+gn8U(`;7$aA4-;e~SDydN6T)bezA*adTP$hKUkdBa zmtxoSx|AE-@OzzcR4#J3&VAq_N(z>v4KwpSdmkPN_8F?L(Vrmu#lqc~6G!9bW6^nq zz`8`p+N-Y6_YimLXC%%6W-i-g6`JF-N)3uNNs~oJ2Nu%SHzgqp{>_Fo3uXA%F(0)vdInw)!jC6r#siwj_P;xnefFn+(EXy@#qtMEVzH%T zDt$bF$-ST}8VGAh6~s&tK@|WCNX`zQxF`GBA`Xe9DEQ$dw#EE!g!FHvFK;a>XB67q z(j!vM`^0y~#4ofSeBSJ7-?I)`&5WWQf*U`0z5e!5eI3r6gdV_uLwUAL9?wDK&NJ*V z+K#+sE=n@#mZn3l?#P&^LsDM*t{PP32sqWH<&1e`9;Ow)C>18$fLdLhi)s%E=$QDbhhT{_h9FfxDB3*#k|csGq3MN5$J7zc$E%jR>Ue<^|E zK%#F!K;OeF!u2(td3zh+bE`B-2tNbaPcz>`Rpsmjy%Bu|Xe%~J+|B!ulH2atd-4Io zPHN>85Lsi=!V!ZhClf=rlg3O;ZV@+_1_7gOFKhU=V$lhH;r_HF7yIIxAzJ!kH$aWL_DbwKC{tasSm2S1%1@&vOB#{ZBh|0 z9Ly&TNYoZOq(Bnz_bbY_hQ5lv$XSd2IBz{D@5m(ZiQx&aEXwdEDB}QXEcX=c*Vlcy zfI-au7;KOe(mCSQ{x(cls{20q_uk(jM0DEhGsFX~pQFq(M197Px5^*OZ zrtjNU)I%jTrb!XwaQgu}<`(Fk0}5q5-@=IinDub&Sco{9KD$zQ0<9>#gfvl)zIV_b1UTQ$`^Xzuo@Z>D1WIANG_*1 za|_PTGahSr2scqSrG-)$wP63FYVxSCH48 zk8j!owy98^Ns^TZ+_N#ac}vwU8{3`lPbDond&y+(OvZX}$by(4{qTh(D)I|rQ&@?t zgLWl7^lzxKQlq~S~Lg&3)neF8RRQgsF)Q#oefsy z0-ITUG#c3|!e9+b)GR(^xpUE6Pu|gw8i-<`yjEo0cfUt@vuQ`=qI&h}&=y>|b)MI@ zOw0~x%jYa|t}|96Lt+R_RsXPhnFZ39_hmNUrfv($KG4RbI|d`aS4u5nIS&lun{alC z|4j~v`yl%A$s0k6MF(S`oA2Van0`B}q=Y|qmr3fb_{bo6-R#W!!8n%Pi z`Fi!op^OvwWI@2i>;UJ0n=8Ppid!|HL*s=wAZSNLn-nvHb`8~e@F-rcNg?H(y|{zAJ_6M zZXSf@t99?xwfuRB5lZVHsZ^S8xsqubxCghO>ky`p z*a_voSXcZcGVopX6# z3}xgiGy%wS&5E+>njnZ%91tX0Ae|l1;fKGb+EqnAy2*A7EQlVwECr%)0=mivIXB)_ zURl#pyxLj#pZKaca*dv(o6b{tTt-oABXXxGI80K&h!3FUNg?#XpU6`mrxyR9e{fgv zlKnQ-KcbRzkvq3-tLuCcv@#fZ|Do62Haw|qOr^mMRq@1=E4P2ELrjH)Y@h+azPXPz zrxhy=isfWx)*XtevguS}zO1Zz_q|A%4s-;rcJ;m*&(|qv)yw*43-Y0=EgTND9P*aJi%iaZ5Sq zb79zVCdO#f##WDr`d#UKUQq~keULOI!I3f>ah$k3N` zYgghCQ%tkbBm8MhOWLt~29lLPO&h#~9UX2>$kU~9BDFccuN+U%s$`(RMF9(cZK8I% zfkhV6)(mLZfeDh}4^_d#VLJ^H)Rb*tlTr0aw`ASp2ftz1PuBShg!SVfy|7MQo09lA z-o|#~!O~G#i$A0(jsL>#%;@?>?(kKv#~YRq*+m5&T>&j*Vd=JdjZ?b4_K|tM@Gs3X zuCIb0uwM*`P&D;5&IDl=C23+^%0-B^L&ljOE`bOP>Y$!jG1PGsd(z8+u0am1G`65^ zxf99Qs-sfcb9hFKiGiKEfLdjNBTLaz-7P5ODt@_j#*bShjd$uk~ymZgbSJP!mXKmbA6_cBD+9o+nd)H z#(HhB^htZ+Lc0(7jLKB;tZ(X;g#QPf-4B`I(dRDndU@f|`Jqrm)(EORdEw0auI>w8 zO#8y|Eb*N_dI`h!V?h?>86GW6id?4}M;)xi3B&7?dEB&Zdt2n$!y(3R?klVvLC~tD zn2yP;K3hKo;7*PYToltXscAcN-AI=(HM(p6WJT~7wvM^n7cx7 zI&B?#Xbde>CSy$D>^`VS+uX9$xp7W!|)80=_6v%JvOs-pZHf3d>tVYKEOOb#(fiaMnxyGNK={yT9C^-`E6l zWlXke!^xG>^u))jtE~k0GhaE)x^_vMEb`Y_xiWJ35m{qKG3Aw4vsPq2u>H) z7RCuoeaf0P}F`r&Six7KyV5e--Ixi1;8H@{1dG&)EVU%oP6OLwPIDk2nklr@ah znw^%JTk|a)p+n{LV_QG{_|0 zv8*%oAGHIrZ0jex{D2}R`y__s_`Vudf#gQU@PWHq!R0sbXdCmkM2s4qEalbLYZzGk$r$Sm4jgBqaY(8mMm7n%9i$0J%v>=GOgS zExQIBN}G99fS7Yq^niIaPrzborcV%jL$*0;|LZCD3m)!mJ;RgB(?QtxE zGkVUbfnX_}`(DC!1|vysHo50Z=lh6NAo(H**DOBYP~H*X{NyU)Lmqb45DF8o|G_Qk zFx0-TXeC&Ht$V#^X3CRe39o1U&|K;R7X-SkOMkF(eZA`ToS|8tfD}=}bQw~+Y1<48 zf&B6sE7GhPkw)ZsD`yw2lc5;>CZ+at3}-VYzCuo>;w zLkW-7PPKpbZ(YPcncAOlH?g;w)Dbh_iHa^eA)bKbXMWbm zA8#Gr@QgNn>epq|giYqyoGd~c>xl=19Z@K4$nCG<6*MQV3|S=5II2!Dj~`RsADi7mh|atQP7sL6ytnQgR}sG*RI$v3niH}wyS ztHhRyAi2h{VT4}M;DV4NW4LIcF%I^D|8Pj9Kx6`0)B2U~>P_`cxE?*rRdH zCI4fk5@K)KZ2ja7p%{`i<)I}< zu>EQAkg$K~Cd^v7RNj9q`P#(1-=* zA89jo-;y8JL-GC_mBWmQylm(+z zRkPlD>U^qVWyzQ}H3o{dxsZ(pnl!4t%%>mdKiZO#E!WTP8nm8Vm#^;1ALK%LUb1vL z8s5wbv0rGXaP5qnyNM)J0g=}9M_ATr|!z{Ix?hB2!)mYl4JfXF#h0&p63mBO=-(GUi#(} zlj?eU7y64;!H&O>r>5#>sq6U90J?bwP1`KEFTY`;0E8qSUXe7`hb!}=jUGc&e>-#`Ym+lM zivD7Y$)!0))NIQh*;`~!bUvE=hf`)Ky#M5bi21@%BB{f9EMd#~K)bq4Nd1`$^W9wkadsl%P505C4n~e!#1Vyv2sI1^5ho+GI%0bzX2$blt>A}n? zMo(b;+FY20B2SrI6cPGiK4U2fX3e~)u{Tqxnr&rFpO7ec&qB?BJ#Q_-);rcWf9Op6QYKH_xik(wERt|USF{u?p z$>y$J)M0dYr;p}*s*9Gqc_*<2@f3AxIX|5;yf0Ca1xe$cW*p9FRkzU63HZ{(leh9C zWRt`JukGeH?$rc)$|x`u^wO8ilu?gplDLX6Kx+uKf9Z<-2Ub*b{}AX{s?7^HW+Z(t z@UgI?o&`tyrUeTe69|vqmMO@d!7kyN|z#!Nd_(veRIGBDG}~H4khmA4IJhMZKz7bE4Z_ z<{`KYO4|x=+pJkVZNaDxsq60mx#iFx4xs**;QB}C!9YF-*ONFw`pQK)I8co(CGsez^!D*dh(!_^lLr;o=e^6hib`+ zYSCtA>v`=r&>Ls) zsU+igo&i!ieN*@naW+X+45T-qWbwEv`cSk(|I&Lf9p8tJV-Fq|Ug?4FVulh*l=fHQ z`HLZcyKDkJPnDJ%Y<%`9XiY7ApS1lT#>courFux3W^hLz?PkEq@T8ThEi=JN-=Yv> zGoG}8xhswdGvFdqL}N+Scs5xvUtYR&cYDCe$oOTzbVKny58J{G4p_mGw`$?W2U<$o z2F}R!3Tc4qgj9E1vtlFPr(j#lmlUoLL>J$A^P}-t{p{{Um6!M)2(J-_OjDxeZmL#_ zD^WZ-$)duRvQ^+tf%??H9*!e}l!8xJR>{qiR-S30X$Y>_E8}kt_(cHuBO$1Uh`w#m zp};{sBx}5qb=!(3I@o)VvJg$~8QQ8-R*anj@7>MGD|ltQ{A<&pnrY$iQfXw=!}f*h z;Rb-+iuLJR=>$4K<#?{>3dj1G)U@3P$P@&{yN39i`}f+D7raD!I&$N8F0 z9tsI<$=ff=%|p4qdsaQDXv>)PJd1TwHmte50y*dQCiGtBMCA8szHV(-!E;-M^(lp9 z(^)1KRvS)rb<@-G;dF*f#yjnM~JZ)A%LsfRR=G;CDJ6RomTZT}5RezMH}EdAOyKMKxAcdA0q zt&E7h%8869YY0eeq$$YZ=I9|p>7NRV#GJ$n+HeMD-c&Xh`zQKVV=u-%D7q<|Zsvk= zpLwx<*uUNVKNV1;*vq9MjO0N2bsqJl&-$fqw%)6d%3G(NRBC8N(iw|bA^)}^T|BIm z>FJh0c*!Vliw{X2ihU2Fi$A-gD$Sm!>SB`Kw22;02kPInIZk+|g`aYw+$WDu+xqP! z=O_YPV|%Gn^|utHZ;P(V7+=onr6VkAZeu~8fAqU-7#D_2@2k>o`W_#z^)rbJG!UWV zEyQ;jFP>I~q*m177H2>#0Nf;Dn~3_vakj@C(S8i^o;uiErqLNoeGu@8n>*H;L9f>> z&nbL$c7^o+pFg~0jrVKAi6H&ISv@zKr>$C7U#uw^D01m5aBTKkbEs=NielQvoulch z2o&R*Q5&hJb-m3nzYjN>-7TXVM;ct3rNgC==Wl9VW^M<)^5>K|0$h`(GV3U?)wLgW zR3sZs_^=|$UX`tzW3V*xe6V5(RoH#r$tN@J%xIzgfyr~uF;eMa!R{=)*rnMLs162h z&v2|ZxIhMUV!Og7U)u8{hZr-*pN^!|Am^zU2qw#dleKcUV`e`x{N{`IP}e@)%l@A5 z0ylK_r(}niVzb7akYK*ZKGJmVd97wT0gy+PhbqTA8$Qoj#pH$O?FUm&T^$3AV2`NT zcous4N_R%^Y+V5hn)Fcaa=CB4^rsdEXb|V{J=Ur>KiT#_;{VTQNzxFFM(Jxp^!)ad zgRg2e7p<5>?#svxJ?+c*J3W-yN>g(t70=|{rIKiz zYZkMQ?EM%Gn80Ni(Kn#40A{d>;MycZ43 zJ1Qww`SS(-2(uhvBrSj`K>kCy=2(nnezjs+=|QX-7r?(o^rfp;`Hkvim6)$| zk@d$q%MM*KpRd$VnNkQH-doWa%d|;$Y1^X(UJqZVC-sU>&k*w;OBxe;x$05Wk%ihstt@xV%4JN>M#Z@wLt=eK z{fT@B-{opip4S8w8nj>E6ZSSp!^?}H0_(#&790lRyK%t1HK~QzRC=~a5cX}|+}Q(%Us+){ExR5<;2hb0&r@6 zbK3q%dqeFRAid0Zc>XXdc6tHa-S}B{&n9aW( zb$9t}I2%Y%?l$y_`644>mt1}H8Qux6nK5nhDaX6@8K*>HK_W|MUVowqod<6Y<#{AUr{KSNY{g4-}^$k#B$q-MLTy}p^{SY+waA+ zXgT^yRvRQ{u9))IP(^417!!D(%CH`e$?Ru*epQx_f_)4I!luQN| zS`Wyi2cR8h2z@*(0YFki@Pk*juoYfC4OCsOIAJbFXq6&U&T@Upb$+9XXVvR}c-r4p zErjaBgq$FEwXH||M%;>_!!=R{dHde5j_q&`m^JgY6_3O`=Z~Mi&G{zZHjy+c+@!MuwKN}RLTY36*o}T&F%N-3w&Hx?}=55OJ|_T z?Lr~51jCiIKb=z94tS>=518!)`JH?#tk)Q-llxGHk0_#3lD7iZqCKVWsYkzmufk5` z@dM$8@~KKTe6og~=LUo{)kA z{}rG6I5O#6awOO+!4rWei$*4;9Oc>^qD$MQ5EJknvna@Es5{5&QnfMEjnhqJ2u zX=R-;@&JsFSECMiCvu)b%Jt=f2!RE6jEk_S_1^|0e~p7bU}a&4V6cwA#mW%9@RR<@ ztXOy**S+`DM0#a3Asn;GZ{}6ViEorZnml7v#V=Lau=GSb2`V{dUgtgdLoXx=nT1I0 zDfyC$Ty*%*O*-F{o0%Hwrl`06!6GX~Hk>*VL0kEXwiInwV1tj>S7q2v@|~q?nXFFY zAR~Ec({e3a1by z>&zmUB@;36(-vIcYn8U|28?<&Zr!#BdihnL>HgobCjJks+3_zH_|<`P)R0t6pE7gG zAMrDFnQ&#k&Xch$FC`SLNJNK7>$5$vd-+sXr%2&@0YXaWCt}4{pcKNxV&Vj!Y(;VM zip__k-SoLF3LzNFwK|lbM#wP)B z`qXGMN{ezDmj{rkF5))S&&`{EOa~C0AVK-jY^Cg_+5B}pb9@;oo$jhp#sn~ELQ;;mKBa%tvzvN@2Z6am?xI?1(k)X=TL`-n2c zkp}@mobD(JCTavfOw*eaHw47i}XK z(IVIEPL~Z$s+yUF0$|T$;$OiBf!wf{2voB4qTj*)6|Nw1O)vra#>ogIXb;fS2}-%1rrLp1j|V$-x2&0YW++>sRo;ZE#t8;u*}qZF5^k|$8qcW*b&F(qZ?MYMQ2Z|I*$?b^&}aqO;l*VfBFE9V`Sh> z#_1~^u671|nA;0Uz%Wx}YaZnLCi0S1ER{2)+$nZ^yS;$NrlnEnnm8y}a%f%z9h|`r zSe7S<9APsCT$tVW{As`qpgTYbPM| zZ27}9y~ibVZz)QsqV~H{jg9nM7GE{p>ypk2fRncC*sSW%c1i?M(ZwB0v)O<|MR#QG zuy5?mkvDkv4(N7j1NNb%W2;>ZB2&^5I6Wt)tJVhWe!EnZgjl-mVEt;-m#iGfq>R!^ zOx*Fk+|gZ2yRmTl8*lc+BI+uSzH6*XJoCZT33EnG!MX%OPU&uJ*Z>2f#%+Kjrom#G zva*e_*unH({?B@yFeVY>BIxWA^7OC{!!Jj>+W#Wx{i6u$Cr2Vsu|7+`?N|Fa-obWl zDx&I=E$_vV|I+rN|7s zUs?mFt* zn@>@2iC39A+enWx9LbW;cG{ct8SY_e)PA~DKCY{!z1Fhx@X5D_DV;k)0U@qy9PuyN z;oDxc&^s&{PbD`BOR_sV_1xYarsMeDZ#Rc~+HSPhcE*lr=4Jei0YdI}*_Gg~{E_3O zpa z^DzVly@!lwx*IW;1p%LQfKi`SRgyOe(MhUS{khw=>fBaC6xNSoRHcN9-#>}Eb{5#X zSn^`j5;{I-3U6gV8!rr7NY+W8c+ZJCQd)AOgjeLl8=5z+UdN=5 zopP9kSQ#gxF-_d_TuEhAlrjac7}CY)2u5Q*i6cQO>Rf(u3pQ-}VG1LnHO&HS)u?RF z-ki^ifLr@7jY!GJ2HXQIkr$40va-KyW2QVzkv@H>suy)J|NpMk5m!>qLQxz;iJ)N_ zvn7|hU$l9ILdL3&+D#9&^S&G!aamVz!+@c^}*gAxwvL^;!fsr8=7ztPlx(j6szZWPWGxqQ!ePy2A0-t?7Y zv$h1Lrnfz%spgmkzn)j*A3WYLpvJ8s&^ChXYZolZ~q?wE3AD=WDBCdlcXXGyl7vHY!0)el0Q zQIs*gXsLVpTL!T|L!ZDqfvO^#H*}E-b`XQqn1Zg@(Yst;Br4(nx>8`@HePv`QGbS* zIo>}qW{eAB-nRX~fYAzQ9@G#f*1Y#wT?Ki_`U&qi3O|3%KsnX>y5Axo8|8J{v!%ms zz$RZdE1$-1NVSkeICxW|;?GrDh$dciZm_b8^zi$dlyyiR?U_{msVMRy7h4;1QDAmI zkCCo!h>@-vfP4vC`2R*08U=Q>{}`K`agCBPv?3k%u5j+yQ}>_}r%x}`e9{IZk~rLY z@B(5>kM{pD_7y;Fc3rm>S}0PAJG5x=;uLKQ#ibMt?iSo#D!2xBcL^3;i@PSc2G`;a zH~s!kc;ElKbMH)sd4L(nb8`0DYwxx8dBiF}BYZ!Gbu={GHYMX65kx1KW@^#>0|>?P zB>XhXJ|B&F)XmiQYUR?&3D-%C=DGCRfO)EI-IM`~Qj(7mz{T-cMYHbkYra%L6hu{L zamC{7+S9Zc%)cH{y06X2J<*%JxPOjfyi(si7w}Ni_(SOiY<+td9S09TmR~9yx#UpR{|#uVnHg}koMz>hpVtdJ`0d@8Tk7U8l7QdSVSfEu|- zA!fHMDI7~dv8V;s0<(M6RSRc7FHt1!jq5L|hurcI%?Xo9-YyVr1-7`?Ws_d94pzre zb!)Ws2BMM1)ES)CxSO1X{q!6mD86WRG zw3Z4Y3Xdu>=S5qN^%}^hVb#mQL|Sc+EoG03L%9vRnuV$tIE;_(yR*XK@V$lQ4MAIx@X|(7Rv6Fvu69%xv=9P?n)85sOS6LF!{DzN3nECa%ZnD0Ic29-%d_ zz18F=K=Nn?WuMvWwUSc(CA|I%@6>u+pt22ok-l82LOt2_t}%zAKqPQQxg;7PgSfFB z4ZU9Zs=!s>EM^QJlyB4!Fuy`$AOM$6hhMi;AR!p%pY)k#A6cyAck8nA`pW%l?QO<0 zQ!T>eyH0r`*WFo!L&JnAK(QZg!O4nRyx?Z!U9pc&)PzR`ZJ!91kFOf zrR>^+069#?(@v#dbQzuiCyReIPlXyvlDq#7}Qjj1j2?-1IEX(Gpa zwRs;|{Cz=Jyxv!xgcqv+`VapFssBY+eBo%O*x~4<0qI-E#|q9sv_?m(8|4(EP!qO` zPndjnuEJeHq(XN;bL;3{vD@^2x2!$ek6lTKHG=mn*b3YTpX+w>i*)XBK$c?4z1*~r z`Zq39qzkY21d=-#DQaWP;ksIE@#do7 z%})O_ZF}THy8p*+{>Q@mMxmJskwpf>QM6ul_Q1SC8U^157S{#-u#}t{ILRHouTuK% zVf`vH)DcDU$?g_+nohTJ4Nq3?Xt)Wv&<~|CJC^rC<|~KZd_D-VUmq@Og}tz~AfIw@ zY>r^~>NkU!{3-o89NeicmM92J6+GrToO;1}*~vN={@p(;-BKGZ z2@lGEaQg6UA#$-LRKo^SOUf;*2FeVDmx7R%Kt{x6=;I8_|Q=Nv&L}5HU&$$ZAkQ7T^i1b}045&kg z<5T|MZ2JQq+9pVsxTXGaiH>^6W)2s2l>b>*1<3US4cw zcIYC#GSSd*{^?jO!`03p!YocN?w^sZXhsUgk9~}Gx`vpv9gaA@JxWlR~|5MB`sKJ zX0eO*|FJvYsUn30{4w~vmn>Q4I;^mo1)v$s`j4LWyg_dsd1}wAQSDvNSf3#o!V|*k zVM(M?YF{L4JceK0cZxWN){~G(pvgV@te5dcm9rKgSnX}xYFrHUys9u^=h=5f=^^Fb zVn&Tv0#-8sXkeyJ6=uY~{=w6$fx43R+OFSD;4dRH|3oV&W2^H%^x9&IYK+QWO}P=PCY1#wc#7mSd<4tG}kUi5Kh*$)Z6X9?I;K#%QR-=r?^OA;Aumtn^#} zarN3WCy6S9^hACV{pf!Q{|mxIN;1?wydM?@fAqM>1FlGF1q#u|<my!g-XwA`7`#7=q=UnMO(Fg|1a;2 zH|vRvv0&C*BhkZ8c0Jc8cjq_Kyi|5I5)woP)lZ30t4ttZo!i+-TjPnpirET~>m}eB zkBHSHwMyn6pj#tBs1!p@VS?wIN?d}ZWNXp*0riuIQ)+(+<%jRXX68v^2lXx5chg#D za^n32WN%>ak9IppMkD>-G1?Ay$6g`L9FQL)a7QUwO%M3(u}!<@agcvLug!TR68Yth zlSkFs3E~`%eDN-a!Ji7rHSc)vrT5FPC9sEE1T}eAe}L<@i~ExeECP4P$$7smcc7g3 zBn1FX{PdFOU+f$)meX~t%01i`|A6;aCbZuv*G$5X7XH!?w*UBH&}HXEZEx|{INYA> zlu{DE(5^0--gipuMJS%sX?4;9(<#k>A39_R)tTeEc26t>^9lzBfn{TFNX z|0*mI_?IZdK&p=SdHg%t7k3eyko<14Kk(??x3;kDPfb*raaQ!a{B z`}ckN`Q{}*W2NF6<9c#uw@VK(0m4j41tdMbspuGaq7Hz{K+ zSXEUmw1R1O-`pR+N7Y!_o!XPS(X)4K!2=l=w zHx-I?|34k?FH2R=I3>A5wN-*)i<;l2Vopblnno=8#p*c%o9XyRj+@#g_DNWG6JbSf z0xLIkNW1@Zv#N9Uk-8J2f2yW#c0?`m{gCNOgmKZ47faW~)m+-buUu?Z*)8F{gLL4# zLJ&Zea|1z-qf+`x=RpI+Oer2yekt2AeZ>en*b{`N_YhA1Bw&P3bDxexdFuV&KGOV` zk7YB1?`*P*IqDPAXjHK^20P*#JEE=c>20dA&}N1A3`}Zvs0NG5sln2Ak{0U z>rxUzV8avPGFGSRf)6_kF9QDNH|$0oI6W8c4hztbqb}Q<$9%`h83Kzn*0i|`!XZ<4 zyvFXO=C+<*KiR2}W)NS}o$ z?pc_V)A9z`q6<;`dB8%v1tMGF@X{^JyZaF2U&1Zg(9F0}j{g`}+0 zBl`t0W-&iNS1I*1FV<=Nld{hU4oe!wg$rC^E4^?1;BHyxcboWZxbqaE;S7Hch#E^6C4ue(y-CEJ$cfy-&yydX=20u0zEurMFUef$onSF ze{^(OFiRwLdbs1C*|?M2 z!tV|raSO*u+cCHFO~3@~n)+2gmstTyhpVYG$oJ)zdk4SMHTtj|&f_un2Y0KLl@+y+ zBt`~LOKGW0^=iwBYvQgva8>;?Nv88XL!{4?;h#+pjEIPm&S}>?Zd%W>IPpuyI%H5I zgJIOG!T{50cSdoZDt1-Q`n4KNJS$>=!9b)CaU>H(5i|dp3!`|LlYWxNVO>KWCTsJ} z%;C$V!@a~1I#MueLusUf-i4;(wKR?$#tW3iS|PudyQ3!?i)r3wLLS;YHX6L);0MUJ zk=lAq;k6>8Y5)7TLc#vwe#nyn*kXSkM*zI({(TDYNb`TC(MQjGq`aQnvM<(M5OsF2 z?e)n2s@fLddGnQwBTv5?=Xs8#zNxMxO6d8pc>8>b+&AFX8L_^xl0S`v98)l-Y4*8IXKEodvzszlWW6F?xgOn}DU@rQ*(HupE+JIp_BMnU7ngoShvf0IxG zwEFkUPuOY@i)K(khd`)I^%RZrDd#KeR;8=OnYn;!*TEI4wh=T{0AR~*`u z`B7%vGi$C3O2r};ciO8y_XmZh49o9v6CH;;W3P?da5~{7bPFwd?+tM(J)|dI#o7=% zu>;;DZ`QQ!4`Jtb8%p5r;~`65?~6QZ{`=7LQ_SSqSEPb}4GkEY$4Rf;#_e!;n%jP@ zLCGp2#+uB$x;1Y?G96^hSC+{d&h!!fvz2nwt zn{tR!>op-`pHz!~1h%C)XTk6H+c){URXzwV9K^EpRs;_PFJ%Y={9<`YoJ3S0?|AEmKsy@(3i|3 z5&WX5Aet zo@I!q2YYh{6~RPl1r=o@8Z%vqs`uQ*g&5057VICvDj=N(dsWo!=F2O51b>z%xv)!y0V&EpQ!x0?VDakPS^0LoNN2Mz35HLUv$-PRq@PCvZjT26x2ARV`&n&UiA;J@RN|i1s|y9xzt~ny*YM{&fzk2 z6|w&Sxg5{=s6VgsfheB%GOJG3Wo_TD{&N-X@H7eKMzXM&5< zU2d>JZZdwk0gkWj2MF5^IGGO{kD(2Zi4Epac_|C7-@?A450Uu)1W9JD-a6&-{JcwZ z%G_B-oMAv_@8(*u$c*USoSqN1mta=oiRY}R5^u}HM_U&COITMyn!)^eWeDf$L48HF zi=o(h(e>fZQ$p!l+szd@zzAYA5~!C%tZWSnJuKX-=5mi(z`N>HeHbSHM3^wvFBc3~ zhv7@)4}Ymt-rBnK%4y$9TwX_b*G$P=fn#%j-g)GbCzaQA%#<>rqpe*$#T^49Tn1ox zJv2X5UNod3v}TWrOp6`(Yb9Dov7m}0S45mI;{DHUr9S=h1y9}uLQ6>{X3oI@WWi1? z)yPhdam4fJSs#xnZh0>gce^V?&7BYC!>PpiGYW-Lq!e4|#QE#X4wm%Pw4DXshKOq^ zfv~Y4|fo01HcVAEBmwMf$O`Cfpyvya?v+tqj z!dV6ebuBw<2X>0Z@-Wn4gUM9SF9&OKkOHi*b;5B0Ehj%qp5?+uGmbheJO? z56)B1Zo6eSm4;QC&SzN<5P!O@fLqb)9_s3TPS&clfElN)dV8MkRrB26FE|NfctC}k zMe{;wL2+o}Kq>=M(Nl2#d}Ui`{Py5XNjNWlJKp6l8q$!XZ$f~i5j#=Ufp7qoqRT!1 z{JDvurx3u^x=6_TttSP~q2{T!eCMK=n7w2u&Usqk>A0tkW5uxBgb6;^A)$IwQm$aK zqAU0IG1}?f^d-7JG28B^jjJo<F554F*=8%sVZ@X7VY|i$7z>)_@co1;iv3~AuM4x++ojiiC!|f z#(XaxOj3a#^{uZ;Ytz5TLc3Z60_nKr@1~=p24+^e=AkAQ$l~+Eun7~I#O(dQG5CLg zJYu3hb$gHaF9oP??=z)W=Et#?E?WdiG39oGTbFK=E zFDBDlcbkx1G{j>!+*_0FhZ2;Adiy}%nM~d{kk!wd)n6y= zPxII=&emKHdRaG5Pfw?$rpR%S%~>|2Xv8wlf{dm?#mS!)aKFE03wXbQ;^!>qOMvtt z0!cCgDbRZwmu0_1qh?!`ptYp%SXdmtOg@=U9M9GDMJjlhs;w#JE`|( z^A%9U{rUA~+Wndd_oPly_iaPX=ovGKI8@OoU}R~@L??o2&RnuKSmL7G|6oI+OlZcc zQnh5a^O*jZmBDZJf8Z_OaL#}ppHnUuevv%+K-(WPsvqvO})xp>k9qZw4) zo7<wW%-Jd&bF>godUGNH&Ee3l8oSfX zNIE6W{fh1Ad_3d{gvw8q%~`l z0hR#jqR#}Y<39H~Z}!9U3hHxXyB+;BA3ydo*TkbBDM>N*z!T|FG+`~YjVO>%>XHSj zFlV_tThUOE>zHnJ4L)%@mq&)q-k+u2uO_ob*5M>*UJjps}(sxp8@c;Jx z=@!`Q#Tf$c6EIJFZ)947#hCl~KOxy0oxYb((r= zY?RczwYQTck4cuZQ}@i!e3QY^&3KUtmmFU|ao&QTjwkJUg|PQRNt%7g^CaVXIw8`( zx}OEies}F7T$2mG^}fBd4mCo%zTdF+vhZLfQEXBjKrKHKBYkswb#8s=EZb7QVK~&r zDQmN51c4;O|E2yNdEStd=y z&z!V6bSdKuo|;Z#<@K!8O2hN_Cbc21Nc~*1y(gh6vBi+S#M}T`2=(y`{%R{FU_ZMT zkF0u{-w&iwox98PX!|vJOz#5HYpG_z^SyUrr=2gRd&f5h_X&EbPGenTFo3Fk8U_`` z_uUwD^r8eG7`4XU4jLJT0Rtvo8&LW4o616ikc-~#MWRb-Vp%hZ&qmPWRU!_yXSF*3 z$kgw@i^e;8ChVt^+ZmAI zqtjq;b1qn3=Szd9M`H}{hekbB73kRDjRw^YE^c!49$@dF)-%nChQz?{1kvT4*)(fv)W6C=U*MfUvFqn=n-fMJ8f;vW& zr{t*BNUd+EMpYNr%40&br~?VvQa`thh1;b71@>RTwXu}UB~#~hysw{{TUjZ4;ts-_ zm(m|Gfb&&G0iOljfFf{BIrN*>NQpSc(@x%_i zF9O*;!a@U;x(4)DU27VH&b&{EnmXkC7vR;>NLtJBgcL=Y-KhP;;i>7ZkG%1bS&5Y! z3HO^1VQCWR5=Ju7iOC>H4Qsg?K<6gnVy={7yUs1JxP* z%3&Q1u|_iMw@XVt+jN&9z)X!Ts==knaYGNLwHdf}xr#86^IfE{Dy5-TKip0F&CW#k zieHiJnmzx#D=KiWL+Bv}zIx<~_2f(~3Y$)jT`2A_B!r%ZIKLPg*iKHQ2%F^jeX#?` zs+bC${8|{QCA@icfhLiT96%tI@0Z#bk@jOhAgut?+cR1?~Da7yOI5t7H0oUElYL z*u^7A`iO3R(_6;&qmkN!gIP%Esyt;?A7urF_?N3;?@Owpjh@USudsXuzy8@8i!Tcy z5Q%bZ$BmS8;ml0_9MX&}0r9PAJxIt?WX#Hs_C+go->4`9Z7X6sS^=dkM&oYkDcIsh z_cfyW#9zBs7WVx*`!h4Ys+V*WKgY4d)Dh{R)z}7WnG{cE^3#aDDH1Jm3uuyW2&jmU zBGe0Bt~&V{;Oc$UZXJ~r=+jmu1Bj37kzO=@dS{4#9i;FD4Eq z!l3cggQhDCN5sR5Gu?w8sU*Xhw36R5mU?bq24Up2 zoDHt_Q_4wBTCb8|(18sX)7`m?M}t%-rEEdeTg_S;$`iPIx|qRgG*9|sZ8Ou%UGt6@ zx|*z;oPEexX$k@=uzgiPq7E@Vr7PJRI$mc$8XvE-UU8+xo(pgchJp9-b&c{w-1r7d z7CoLO0^hdEYeom2XHMeo} z^?BuRi_)#GK-)_aA~X7J4wg?AW^vgR9_L1=m}Ep?U|jiRQ)WzFQ_!PpOLF=C4WWdC za&hL6+C4V>w7w0e`eR)6>))vVVG4gj6=Ie%DuDjz zOA4&_i}hME?E4TI%6$*`o-lp4vixxqRT!kI4yQM1XrqT00CTTo{~FreAz6W*M0z-;ioYCd;GTY_4y7`9f zcF*PXO*t)BF7CRi1RzgwlNs$KNc5R#2OZeQesH+Xe;s3B4A{Ul=r#wL1D0-gCKo8T z>EEc{E17Q}s!~G761DjpGGfjUqR3^Hx5I9S98P5)WCppmiN)eMziB`ZALmWsTK9iL zGanxVEJE9oHMP@EmLDD={pY2N>s#+9#^-njn(-DUiBCcW7@0gLS=)#IWR;wL@7gz_ z+C+3GF7Q%XJvqjT)-cAi;<6rrN+R~zV z2~y%=l*do)1Xj4%33Y$c#;Q@5#`fH<%Oh8gH0I1QWKKhr{W~oe zzCj=0UC$4{YsSVl&IyJFPw}q0yv$%qT znm0h(^YT57g_XH^26#H%vzS^RDKInUSMbElFK}KBU7fn<0d$Ds+bjjfGwqN6zHi;O zR}mWv%t{Mn4cF+nhQUg-t#GsZ?7@VoeJ&rq(un)A$&S6WMWHs^sCEz8=yiXu zO1dk-=eb|O{#wbp>(Zq7Fr3hWoCM!r%21kK)txqOC7U&Mb+fQpNT_R!)jk_5P?TXS zXJZg?bJoZtGo?DffC<~x%@^C-+-$Yc=(LhuSeTWHq>4}H@cQu;lniV8lmW_?pd^W| z6zpEnF+|xr7$!+gLgZ}r=}YBzC$~zsPHdV5MCgCr6&#T5iUoE2R!lyTI8nB|p@Ub; zwtv$MtSE%)q_o~^4i4p1G{h8!Y)rkLsk*as-Zi!@53!|53>FO@1)=a^mbWT{LV#)B z>n}Jq#O>~-7A30^>ZHyZRb^&hkD5yNGfdPT|1_bXpkc4+R`Xy;o81~xjIYL;eV9EF z*HlQmb+ix*$yhr=qwNUsE^W-iMv+gNA@%vZ87aNg_OHCpQV1k*D*7O+Gr|63zpHDU zD@2NsCkRzxI5dInQ}g^a`IbcPFc=X`5*kK-M-`S&OXLv8Sab7pG-?+J=G{b4&`O5DnWWT23vOr1 z%Q&XKbc#xY#m@<5#@!tm)2Y`^`INJ~M+G5!ma=@RWUeNc3LQKvz4ofFxn3p=OJi&y z7o2+FDBroWIVJ;XoKuSGO_(M(35;mQaafsi?s|$ms6q!uH?N;>^05K zrlH(dtqVTR{xNLC9>U&QM4e-{w3XmHEoBa8Sq z)vnLgnmO?@aVj;%^^{h`^C+N_gk6pOij40V-!lY>V#GX~_)OH8&YOTk7paOK4c=HU z;^+WYK&buB6-YfL!3U~XG9_z~n?;oPWQ=B!U)C-nW`Do>G&g7WVO%0mjb^r8U!|9$ zmE3%0o|LrhG6&3s=GKf8PDZ0t~I2c9D-C%oO>xD_t4I~qq>wK zfuCk1u)1=sslcz4)X-$i1f8Fz82lL^YG43+YDxhjZoHba$Xv|~!~^H*tWJXz z;=+U8@p^%t#=_SW66mHBzN!qptY9h73>U8G9_Wp3`bYPpr;}!^I3$L^C|CXqynq7# zaRKkKowIY6yI+bdAh!BP3bQmjnX|_Fbf4Um;hudlCu}aJ_!;m#dEi`zd&@;@LU24- zXj@n^57}<&IKGp੢YdBVEbw0{ogImPN%w8LQNoAV9sj_5uvTc zdLh7a4EBX%kKwX9{|oTnVC8?leJRqazuu;|{WItx1NCWr`-H2pj8PI@NN{dmoGgy4 z!X+k}FfXk4Q<<51gYmQ8%v`3G+ccfKux0@-Da)jFwt4)>JWXl*R{-G zU=L@@R<$S%sLIquZ|o$#shVUeT{C3Ms7+pgSh}J3_Ok48Lv}73`iSCth7#F8<+|KR zz^q0?L3Tw{XFYnpw^~I%>r=$}w7H+Uhq6!8NP-RE{({OH{VCDW=}tUcJoeFrsafVf~&$p6^`#|b_Hv+i!hnV~WhVW|t+3``8UQ6?-S$jCxD;X;>) ziF4*K+mHmysdu3gGlEb3if3F}D=gT;aD5A?1AFun`1tt&RVMF{&BStJk zuk)@321>9|l_Jn+XjQ1B*+5HV$K<^gQGp7a{*wgIKx{g19xc&Dxv!vzC#1l2Qb($e~{#7)-jXu_Fc8?36hk4viV@JOK+Me$iH+Hhs{eP9r zGgJCquxvk8Jht6@iG}K_>M$kU*lJp0jiA?G?Y|kmCr^;U>vV zokb^ydE%Qq;HJ)<>Fy}OZzv{KSCmnYwZd74u@fJ zJ$-$sd;w5ohfK#K>)Nhr3(k?6~1Umv;{f4*q?d%;!ZESWF*hj}E%FJ?G zXGdqdFc-Nw+S_NTHXq`W-R!6=YqnA=W>g5fTQg}%D4Qz~i`Ygql^G|kMnZ!LOjCw*2UrNk4;`aE)#sN)WxXeY zi34DTSzh>)GgM}`%wpT-%eEn(Aa^R-&LqV2w0&-?xe9s5lzBswx;Iv?=#c>Cz$L-| z6yvFmCmlQT*ERYB&0`Jbr%16}Vv>2^&b_BM@yG&7{SdbT+?CxvrxvK+=G8%)M{m1mY3 zSdB?`8I-7@PXBHsjoeuLW2_l>m$IF#Ab(qvmZns16PqQ*8#g+NC=AJH!Y&!8KZ?DD#&vSRGNFPwh5=|NaA5w z6bTW`+u!A5HMle`QK%^AtZ1&AM5Kp*TwYl#sk;DJoOR&E==UpAVC%Ss!Ay4wkZ3iM zB6k-^)qMsXs}>)aw3qE}+b`vF^N>clUCB~cE=~0$@^+|CP|3ZBNDFU!Q$FzBP%Lk} zbESX9DS`Nu7Ke^D35TuI_OJ2&KU>>(n(h(X&Mrqc-ha6JBcxo34;mgxrWwloEWu_K zE=Kj*G@rUDJP+{jjJA%T77Vq@GM_CPnZ0*ZSo6Bcrww9tGN}dcPYjflwWQ-p4HeHq z1wcKYkYAFk1pfp#pA?UZl*5YmQ^^?`G}^TN#!V)cRN-&LrgkPvCs>5nftnU4@!ef~ zueHDFxWff$?l5d774nJ*p5eGR8j{<(`G1>H4`=e=k+wP=)QSIW|?#j==uq(S@1vTsvkM zHy>BikSLtOaxN2Y8_X;j{G8EMHtn<|rHG72uv;B*KnhO#JVI_VI^33W+Km#41FwGp zYVea?UonM0t(@K(9mmD7v~qYq4IEJMQsnNBH3WeV0+y>Fd>vgOAtcw+o2p|jzjS~` z12CcSYP0iSYLl?Aq*Q~1a}n|l=ZfFX(;y~%SAxnmdm?XctFgj0>46IK^Ank{&OsK_ zLt}Q^V%>Rk+KUv!6nQwx2)ELX;I52>m{s4lKKU4nmhyh+8Ef5xavTER7!uKPw z#1=|MKDh;WK@Y6vNT|A|d;8X>f_;p(F}kJY4d zOCpb%%%NQi6=Da!zD+^-KhX=of?wV(vv5IA3ha@u{5My-xhB-+Mcj3Q=;RMZk3X~j ztPo3oH4zbc*_z zN4x-K`|)2O0S7A}1SQX}%h=r9iioGkNq}8oKqkI0g&NHaKvb{|&VhtBUu*-pVMD5D zMUjzMRe>0_%u$34!e_sqQgzggZdhsTqR08I zX;%vLY>`ZZC$fOYq!LvYKG;B7msg(~mYGGlNfIvj!L{W|l_j*ZI9c-w`E7Q#ALzCn z>fSEQM{INc&Q3AGbMh*}p9htO0{D_dIn8#4The zoB*CFrZ-fR5w@BJ22s#sX6L@zuI}03`YB#TcUa9pR?}pdG!iI5g3@pS;>oc$=Y%P~ z!<;@GMyePozu>YVf1)aU*NlCquI+5ovlQKJcdyw*kh(CS7@|dBVuVn3gDpy>zC^TL z%9EhElCe!+Yh%3unoCg%Rvzf_d2bPWo*@+QfNKFN=ARy(jiD&U2256sQ@kWm33-Hh zgrraYl7L_7KH+Gl<&vpOm~5=V9ir^`8QZZO!;1={ep%U-s2E>jtL)XrtQ_|_fhtoR z(mr%=dLEus#3;z_t5{Z`Sr3+%Z*jjR+uqyDf-=UA;4ooViz8_zZU3@2x{JJXsHgYm z)4w(q?@`~b?;Tb4j|xzCm*Ov|X$=16Dfm{MK+0;F+&ol!BQr}hY^MG|T;XKLyHSQk z3v0XB*Y{1mZS&9VBDc*9)fIZ~X>rD(i|rx(DNl`W^c!+K5NFK%&C)!pD#pm2ASDW# z@dwVC^E|7C0YJbOI&0U(-vwT!}s z511H>7X2KPy;d_|a3%EioezdGMFveU`^0|Z@@Ls`f!tYgQb(m9S^5*q1!ej1&!0b+ zsr>R$FH-$Of1d=~tne0(zT~jv@jB4e%jK>6%-ZJiODw{KJ4?v;%Bxg$aDTw5V+kiQ` zH=pnnDox7tMkk6b#DaQ?jYS`p*QBvN_=g{ZH`$;;3k>yHa=&fYftUoy1zO;80#|%rW55Gy`cM+f-VU{q11OUlOF4OlJkdJ z4@F88Ff%`=S_(j696p7~C9Vk74T~aOZA%XqVk5-8f?tLql==$H)BLDGD*F!Foug6N zy(cZ@?2 zpQj-n!`dPtuPlSm@L^Bj(c?VY^iXDFb6KhH?s?fVmQqS8!a6#ZHEz(Z4=eASn{&UX zkJ^r@GgKPlrqss!BSHeA95An398i&x@knCb8TXNp^)eJIHN(RT*YpgIFgh z+9H@FBH417z1a~xt#MO`N~u^E>bS?_j{dszeHr&s|A}i}l6V}MzOVV0Bv!#)Crrz2 ztxujzN>ES@812-5W8{NwV0Cb|cVU11*6F=%lHE{hk9>yMG; z!g-tUKY~R~V&rA)TZ#-n{w)UoC#S|EBfU0GK~DL>@aIFo%IPgU|sjMlrU9gh}J6K5QIcrtQoHh_Gt#o%@6=&JYc6T-Y$zreOIH$Fs zxZD{2fS{m2&-{2W_Wrb3OfWY3w2P5VzDa<~Z8rPR@1x|4gsIrW;g!z`3@euM z@~kO*?rpT1{eD^gK~vR*xv=f`2!&e}j>qSo0uPAf*{H&PYfe_dUG2Z zS3!G)`{n?;<7XSp24}KjqQIxoX?8OBnmN(&gVze&MtHJ&DXjvWn}&X-WJ(QOblOM#O{`3@ED6KX3S39mIq( zWiKlw>IMT9pK3KHX1y#?Z#GoIx70#B)^Io#6VM7 z8z4*8xDO1^FUC2Q{p1m1D|;OAQ-s>xK~bxztu30DW9&yk?wXT>6T?Pp?zKz}U15uv zHgMJ$c%oX_%h8tfD>Wn|U}L3+zjtyMm%uJ#Nkym7^O%Lp>#OY5jCQ;2{A&!P`HT|z z=Mj060lkII))6o7GLr4&&^I{f2V+wBthLwWu>+MsDCUei*PE>c#-|HEC*3%*#-L7< zK4_<={a4R*M$Cop{u-?RL|q$!CsZ6y!&G@DP#*OET3@^{?LxR%d)coFT|DO-ZBsJs zR{O#9*J7ze83=9gWVx-~JWX^^Qd*LeoZ$PgaO3IYrrSGTgx*OD zgW41`_N=aMZeQ%}8JjN8(!_|c(8RIagXorebH8v;z{Z|&+r_PJQGDocxu)8-qtJe~ zyRliUOw0Jq;y3@M*{j2uuihmdK=A&q*2QSkbNKSjui2C74ITP3^#$XwUnesp2ZcSg zZEe0)2D5@QZ<4f)$|rfIUS;&`BALD9`1~;LZ9JaLaBQTkuy5}Z1zx)^g=U~0 zN(#@T_bCD#lnybQwKG$@gEi-hjQ_R+l;TD-6n-R+s zQ8ZufdoeJANgYd15U1hEMzy4|TB2mjyt6k`?y5vn<@nxq1~{lNh6o^2{{VX7yOV&z z(dxk+$u32loCI=JJY4(&hAERlSeY@O%3}P2@cX)F=Cb2;lNK z7zy+yBg9RvN+BSOjIc7D%dw82>BxBtH_GBwzSM&6-J3q);YbspD#)sJt}d~mD2#lO z)DZXZ%zkxcm=&U6tIR5%@0by?$w~j|vxy~Ps@pP+F7KWfOYj8u1GE_-ZU5>)zg5J~dhV!%d=LnZb;=7> z$DMh51nfu8<)q!+)MyUB6=MA{!^;;Y`e<8Mei%`wjoHf!<_?vgBKoJ)nYxFWOk%Un zQhjd&uk)%af&{3j%v3)QZf3fSe%Y`LKxj|5;sm|s~0M@nfFthY(5i3(^WVd z9A;|*l>3>UPut8VrMe{x3RGS=js+5M_|N%E2oJ3BMbGt#l) zYL;6>5>rSxjQ)&{Wd>xY7PVY#Q)>-2e|&uBwYsvBnfO-W>y1m}8+bV$=`_dNh|rhJ zpb~i6wUf2#H>JQJ%1J)=J9f#5q|$xz!u(=|xg3wcS36ykc-%P;)p}6T8P}Yj(0robg7Ap zNWwphDqWs#p($xA$lmYV-yjEO%txA58e zoz2yVaM{y#uNc@SZwqv)t?3E>t+Iaq*u6xuuD(9M9L#GO6dd^VWq6@WlU2lL7b`0S zsiJb9QYdX>D>nN-l)YtC9b48l3?~Ny!QBb&?(VL^T>`<~T@#$(1eZX7;O-jSA-Dv0 zcb9yX+qe5B{q>LM{lOS}FgR45T6-^-9eHl&mDqcecm^K_gNEkSi__Av6&VHU6$#Sz91^+2;~Z?^dFNAYtMcqNpXb|5 z-aF2aTcxr@sX4vktDoq&P)}52m$BbpTl*H3tFBNaQ^<#j_?EJp1<~(km*Vb^pdoZj zOc4_|+QP!dpxG8uYA&jn9l4^?Pgb%|rJ3ecPC4wJ8d3dC6v;alEm$gol8V!Dq1q1@ z%@$dUt~B^9vQ4AH;rg3E_qLhsUX`RO(@I?{gsZ4X)(yO8>0MY-5y7H?Emle*_Yv;q zP2kLn0tAnT#^ERYjH~jjk2&@~nZn^CzXaE5TaGusNk4#yFs&`#!1&W`{tJBVK!KI} zwy=nj7Qpj6@tF=CL_#$pC2MLtj~cft<2H{xr#ig<4SK7*?YxC0o8R}mL_0fPoJSf} zd@B>dDuHO)We!4T0~G^B!S8NoCD+&>>+@lH+F*xfWMpl^=q$jGVaOG(!X9}(ZDyPl z$8^XPa8rVPD-K_rXRlA>efq3N1{a?%O%e)lT2iOyFK^$~^-$denPp|yukdEWM~G|E z`mI-LMIQDv=uX4|eLN`~@t1-RfkBOVs+}SRck!FYqY~3CL_E{t$9;sK`oD^{tZHXt zN}0r}mbDwP;^n@+OU}`u%Q0}s&})5~s(68R8h@f2C9|mJ8gEST8*+yYYiqK4FkNWy zt#!z%BaVs{OQ((U7wgd0Ru>jHbeB9oZuQ$qmz)kWd?_#ZTy01Fet&4zF#)9%dqO6z zGA**p&2h2DA%BJer5G?=`8Z@$W?Gttkj)Xnz9M`D!basoBct78fy|fXW7&NFY$EBr7@Ny{qVR4UtL-BqUGYJ|C~rnA^+)?1bo$yv6S`jBq)MP911 zx06LBzoRQAcfhVQKA>=2s1aKl+E+GM6jw1w_AnNdWD`>uShp2l-GV#ub(DiO&< zykOD-6A-3HD%K=6B`BwLfd3!Ic01jl;l`A>uf^!aZ4ZuKTOm!cF?753K;`Ipf<1OUA~oIjKhnd9U5yQlyK$1tC(fPXp|IRaFfJrC6c|`XPocQYdIb?%4?0I>u0CwI;Sv!rziZ- z7cN2rw@`~8RF|;L6FM z*7t#2yV=-*L0C%@K+()J6h;y01dXjt6$@8MiwPR!o10 zku}wKO>)oRP$n$PxSdNRuGGGdX;7ELZIY^9TOMAaH?pYEsFfEWNmZ19@?TRKN$pQz z4)|;tXVGGf31>l(#bWOY#orS|)JCGzIw+{V`w&jVtFG^Ltg%7c_WIo;70MeC4{vfK zMVnDY5g#$hw6Wq>T|ocm{Q7*|I95Xcx@6XQD30)xdGC1(<#$_v+CpzkixR1E!KJOg zad$y1vNZDYE2O)O`q0JO$BB#gCyl5s|=o;;#>1hB{!Qqz$XA zlz(1M8Yt`NybL39rHz0yIaC{qay0oUPhzG}gITB)!<;lM->=CAz}wxBN&GYSHRpYv z)0x!W9!gxi$*D*0#@H1cB%E&}HZDi7vOYUU*;|#frL^JTR3p%sS)hiwHc4+=@4M@@ z1nsDibatt-%?u=Pr`#*59*bFBqQ=&eHYf5h)W)<~h)oqG7CX6p_LpywXy=IuKj+y! zeN9`~>eBsCK?|ojYmJo0&)8e^ko5vj#aD~-sr*LrLph->!=-^iRL*)Y?PYVhpW@<+ zro9wHes(_&p?htlR}T;Els`+HkK*9gr@!~rVb3yzF2aBOz02`y9e_ic{n5&7Y{IZN zH8Ak^z78fXB=b83nb7J&Es~jp_JMo|zjkRHAz8MG&F)(V#&`ome1vjPPAgCsi zhQ#);IBfM|D(0?OBZ@)00|#dNPlEIwRZ2hiU%F0wOy8*a(j>ho&x56L4$oA+1Dpps z^GE~~j)DS;Xb0z4%_&FOA#jZo{__K6%Nr4CBP<&TzvWLI>1fDnhB6bkV+VDr%IUq` zJ!125F3URuj8r-4KS0{8j)lyfgXyU0kS!csCE0^fB+z zGTzC(lV^E9*N;)-v$F-uA-*2>oek{2Z;cBCiNYX_8iFoG#AE=Qyxg9@cu~v$^#&%O zuuR(GCLpqvL=&CHFJ7AEI((}!Y;GX^T}aD1yg{70A1^w0A<1E5FSQdPi)mG#Zmb0< z#Uv*yvniUA6Iq1jmK}`+%T?Cb$FP}Q7dnY612i#TT2SGJc7*WM0t;%2o0OnbDJ9<5 z4M(^hTloB>XNn-iRxRl~D1!Zo7j(g`&MmQ)G);S-lQyVHW+GTNT6dpo1Mp*O)wcF_ z_+DpN@c=R*7^OZXRq*8{wt1f|%JAp;lliw+w9jp{Rez3cPT0+r8@yAi6`B)lA+`-w@yo!KZ&C7Kg!_K zFCoi4)IqB#IABe)?ODV`o2?w@h=)OF@smfc)o-*&Bk8t9k5qvMv94 zoDiUk5FuGByy>Z#$%W~m0j-%nz8`wjRu*g-v4h7)nM)MN2EIf@b`l>CPUoD@@ zns*D$X7!caqg>C-9E5lrC{ofgM0cXU!(4w`Ibfcwqe-gPG%Y2QB`v9^)|xE)!v1r| z(dW&2ikO8__CoX0UX{iRXq~Hs@dd?bhVxt}7a!G?@d8Ao3iGSs&penlu9H5D-QQ9pogPr+0PdCN;J>t2lrjYRt(#V&0)?ORc!P%uJ`HSf? zFto~UVW^RP&x@NU669vi)Ya1~Krpelk9idlUf5h9nOG<;ndbV9>NoO>H4!o@zSzp$ z_J#9pKQm2}icw^TrKhF4CX9`jJ!#nKv>G%-*RRc|y?FL7dits^*wj(KqfsiS_U|K6 zaSD5TcR#VZitOR;PWRZ$V-Gi5qAPPAvN;FVefmLly9|R_3$64JM6us6V_CvfzO4M! z*Z&yFIqzaq*ew{U^!%S>myy| z|9XZ|QV>S!l7|>m|4JYK{mr&vA?+~BYz+vcL!#^oyp$PKq}rqSxQz}TrFAu4zpmcj z*Bss_K(?$yvP+Ne_|+>t#p)Zu^;@6vkdc+o!OB{cJhqL0t&9W@|Hi|F9V02>d*UJ9 z)Wk&opx<$!B&?A->y-P8f1dy*I50`4$X=oypL{8PSZ!TdCfWKvvg9}9&&7Oc;EMxq z&!#M>cMagb4!u7bN^+P7BxPqv{@>2ie-h>P`;ERs6G!ehA|`qe-`*zSx-hXK|35r~;Wn(?^118w!3Wyk;U-``j0Pe+=F!evmw%~nKe5C7uuHuJ%K|&M zd0VvGJcqQ)V)oi=Y6TdQrSK6@_)lEe2l%?1?xxqQjNwN`N$3Hy$tDRZ~yoMnjFsz*O0l`chgW-`-)$BAvr<_^!Kku8>PtwJrhS;FX}!K{0|ijR?y5w$WV&^8SX_p<0D-98EE=MsW4By(?{`7xo+8KP{sUXE8#~&yyPS| zVAQ~@@rdsJii*iKV=$(gCXj`m;6u9od7jr9Uk{{*HL8 zKEsy{sO6Y@LPwl&B`d`Rsg3E*@<&nhBn{zu23;KPVERAy1$Ky#on2NH93DPCJ|z}n zfd2&w6O7hL92eBndU1JSOHLe5b5jNqSewrVWIM_5783J~o?bnK&X&8prKP13^lFp_ zt65R3+}xhIy?v^eb1|w z_FWU1iJYmUS6Yv6If*T62MQN2O+GEi2H9Q{#3D#*VQwB{zrth9`PnF*XncNNTu)E$ zFUQ*90As!~U{vz)L<(5OVB+Pur&6v4ZVvvw3t#6vNa3b$e|V=)kiOctyu$ z>WyzuV2F|uN>MnEe=he|z%D_xcO%!%*w241F|{AhKuwvRWDNC`KQODv zVp)1`LrK26!GBE_fse*)^}vT(T~|}13~oX$D_b(~#By zZYgOJm4Ty~(n0a}edSb^Fd$Ybq3ty|+ojK_m;p;odLpepKAU@Mqk#iL z;UA;2dm>n?2W6lb0p99 zS6;FHR{=%2*NFR2|IT^-N@AmgA&jB>c3(~z9m6==i}BNsOt%ODtcTtcZOYX@LH}No z4%Z0=Si*g~0YX7-wro-#13?6JzX;u8JoD^|LHq3rV7O7iosAHV{Q7-|2*s}_s$#oL zA0BC*Lmp90MFYAfa++Vs@_EK+Y;OL(FilHEmfU4SI>HUApL1+G@UkPCi?%I0D@sSAC3cbsRAJa%l?%$Rr2_ z(Om=Jit%?J=!kaN!Z)xNLlyv|#7hrxiwhPqR4b$l$IZn>bat*rq#(c}0s;H6tAIK~rm_MksP>98z_@bHNM z7Mz~qkF9EHQML`$9dV}$GEU!N5v(}@EeBJGIu6tPG6T_B)0VM z>fS})oUOcp=xvUgbNh-0BNrs3=0fk`;ZcdrV*g`uWzjF@9L}y<(TM;MIEb4IBOM_K zK%*R+xccFZ+P78`_SwaX$ULKXdHmq22q`LyPm~14%EE@!z~HrZ=Xaql5mBx2sy7E9yU%~sxen&szbUzt8-eJ#;O&I{?BfVQ7&!Y+LDeBmAVzQO^ zxTKn$gGt%jH<7AFOGQpC{Pbya3#;&?y0MU;j3*0L&W7Atpx!;QsrkGc3Ujy5-?(oH z6-qo@G|*~2uq+lawmHHt*Lpo>fZrL^A>#S z-CS|6iCG;8dQrkrRJ>R*zQq~MN2E|&*4pZWA5b?*^YRcJ5F}>^#2T?-dg<+tKYq-i zLIVa%7BVo%vTWr$kyA5V>%JYogp%l(`u=iT4NG$;hx?IWEVxaIXtf(GM@ldZoTN7~ zSB;@zP2&5D3kw2D9QZH)=u{LCWgAm6??>C9yBwEI5o{izA4!l38-7M>7>>Od)P%w8 zQG(Yf)hzlDJG1ckEkJQkT?rNn3LI0A1Yo7oV6SzU!5%0p-9Q>_{^dheN zcfN09oAAMdCkX}N^f6YeiabMcsO33^q0!*XkD)wet zpdk~8%iFZxNIXqLTgF3c(Vom=i8)#rMSc8Nous2dksg_{%({|4G@r z8G{Bo@qr#@)$%ay7l4hEfQ%(utih0bIYp*M9Hx1Aa3EWMr}@d-b&eR&37M0DFq(O* zoS8{T<|oA}n_XRr>pF4wOYvhav?%KKBQxJ2wzD~J(3Zn)Bgfcq9i-g zNSi}h@f=2ZSAc_j`sO)C$WQ>~YpD%32gR%sSx)3|L7AwYJj~_g@OTjPcPRKI&RlRNh$P^&CheR(aLi#H=62kv%o9Ea8mpI`lJJj=QHL*U)~;b;D4 zYD&)NpRc}_)kQfEVT~$ ztoc0yb$@R!k9DWKqPHiWPjd$q7ypY%PjLoJznu(DgciA_vypeYQhyN!V>>yZ-z1}2 zG9Mb&p*WUBzQ_1wp7_@mS+n*?JRhX-y{5@o>@h)q{aUp6mXq|OA8!r%$VKDYTjHti zI<*XISsYxJ?j|xY@6_$8i#N!gkyYR|el08HBqOD_;yliAO?X#yZ_VjPr>Z@EK?ROK zc**ucAQoXe`r67_b$adG_SbyJTE-uIh$_ZmPPM^+JWU4^q&J~Gt-n8Vhd0gfi(hU- zvmgBU9IRg@DUlQ~V_y;;Wp3O!v_B4I?TJj^NQ1IFjilMI+QwX+?=&>( z!^b_tdg3SArK&|&hZ|)9TuW*o?5xJC4^zhEt-G=j{?S5Pin^ofpeE)KAi1FyD2viU zGN0tn_5HxK@80&2ChD@6Mbv(02Nw%VBRt}xI<0bFyMSzB;KQK!i#0mD7T#-IT-9~8uw$Q8LD9Z80nu|U8tURz-$OzdMAx}2h_>d#G3OlAt9#!# zpbK6W6u3pZBRW*3m<|}}XjxN;N+UmSzS@%8DF{jD_AN&ELyQBfhG|}HZN5~N$ zM7k8EoWK|P2%pKGSAQtbQdJ94mZNsQhNRJ_JP>wgPO?PzXEdw6D;<2wBdLI zlNVNLDh*^PL%s~+&M%0*vQtHq!626eyg6vnXCbM30|qNaZe(X<=7%VQz> z69aJ78^*SHEQN9&Nm-7cJ$|{V4{j(>jj^MNxfPaSP(-4_Wm64KaHT$1eGB?6KqWP- z0(yFSp{lVWB+#}@vtCp(t&<<2;i&mRV>BmR$aDQWKYGvi&L5g{$9rqS`uh5+Be7b% zT0&uL^u&evxxT(qGPfrx)on}u#4u`f)+6QI8WTfAt{i6``hMeae4vF9-!t`X#(5eb z4MYt&oHNW;Y+@cb( zl(=(pPn;HxMUBxss&=|b>q))=+zz#Hn=LeVPC;?lQH*zz_=k|f&=Np>;&JKS?efIe z{+pPR?r?YI^ZWbTaX`edP^uRrksZ*qiVzVE)XEm*VX0Mz)~A#tVkZ97IoB)O?f!|w zu!r2O(EkN-zg@_siI^s~p8)2CS$-N#Lo`|VS!^z4LskoQOwU@r4}YUJ(TzL!bcvlX zBZVKcb4>ji-$yO;Td$13<9zJO72ueii+2&pyL%!NF&>0FA|dFkM(kG+V7Ra#J>GDT zZ3Nf7D0d~vmbdNb7*P;Xg8wJ`9VQ6~2oUx>Sb8zo_1Sg4TrsO4uF6rrn?$>-xAfEA z(GAjg`_n%ARfw68EYIH~ffYv{|-&x3F#CvfRKGg8$Z;M=xu>(?J-{oB<ol}@FYxCpikfCD zyDOZq%9TZptEY{7;of#pfXqe>@_`0ulk*+!7S=N11-|?^E*i$?51u6EfScnX!H<%* zK73)6MuKwID;p=-9VdP6-vR6eG41As(_zQ$ALLEW>>;F;97<+pSnRbm6IhbbDk{ck zppY&U(-uJw--RB*i#yEeil6gIc$y}5dQ(Ir3G&o56H8^)oCd+22B>>Qy zL0PD<#jshb5*CLrh%*UI^}3v+wpyg8f^!ec08iQt3n71Z+7E#kSpbhh`48gMfsb-}bCU-Q zuU-oJ-P5u&YiX=VO|!dLE_tmOBYixFDVjZ)hck*TDwSi?a9Ez(l`=jWqAmU z1p^(3Fa@fdzTjYqT111v@26xSeW>zpvYRa*Zn1H(@reu=vy|u^jUhF;B$uoJk|$!+ zAx*nWe%BktrhWVAgD`NHmG{~!W!d3uD0%M6f4`7ed ztWtRz5b{a`34D$9fDvX*KK9s&F<$cif@hQa9?j8mVPN3C*DIE$PLOXH5y3EMHsQK zFpt@8#zklHsdJ5QQkJLmLh%~GbM0$LzJ!Hna+!acegX3orx|u_LRI$U%c$im{TfL~ z@TUa;_cd2-C@Hci%AEwwuEH&RNa|c%JUDGDHVCh^ZN)n4TB4=n{AVxQI(kn79NynK z#VRuR6mJW8$zNueIUsfnQVh%K2Sx+mFLyPJ9Z(Ttn0naK5vc|Z^$#f8QG(1A>F{=k zg|@;0<*$k|fEqb8I^y%P0Do9VS!OOKM}3VF5?m0rInR* zudH?1`mw4a$%?jm4mM$=-r1t8F&d<@?{_!MQ31oI6n6xba_3&ePZL8%CO&?GKPhUJ zm5tI(I9jqm0l+ZkHr0ecjm49t3rAaXK7_+3#j}g6tp~vX%Q?tE@J}jgdQw4y*IdHM zTfWJ5N$-`?gM zk?tJ()%<%_AU(Q6bsH)@8dVz}G)c3$4bp+MJh%B@jxY%{hX!anMY<+f5eBOi4n! zO8v{()xqRWVSc{8Pn)+^!aIR$X49@<*nrd~o+wPcUZq^E{=pUM6dVh(I+0J(n$XbT zaiBjQZdg@c5WS$uhL49P4P>+UCbPeWelj9Z&t+dDI(~Trqcw6WcQ*p)>?(Pvv6DM( z>Qo}25x1IDQM{WzZ*bbQd+!9cBVTyGszb_pHM@u}lmmcplDCKjP4^N#7%O@zJKRgG zFEDZ@MuUn zLJ!&1A#xAt!>qy|@Zg=_pQHOX(gS3F$^0M|t(45vcau;iom|wAnb+2VL2^V)(J+Ru zj0s&Zi+DD`w*`m*vXkj=S`jezesbj%1b~dLl7T5$8CS`q#^sQcAwk&Y!w{_5t+Wa$~N7%m_7ds&Pur0~zi+x2!SbYm18`>V8xQ zh<6>!n43jptI;KiM^YQjeC>C~ye`eY0tP;!>xmtY1EI|ddc!`3^Gd$Cy17yLPF`$f z*@sb+Uq#f?ni~(GdBp;Fs0dR_OKRfx3H@cFbP4DInPSB)J_R3cnU0ogT@o556s8TB zdiq=GR>aBkwp0g3&&887VlxMoLUQ|Ta$#bOb93Q+m`8KsFCBmxAyfrudjnB6ZiX2L zBugcmXot(OSD4E|&@T&DkgP;z9N^a9Yg9W;?@;rxiTOY?Q!=4chDFPMxm?W_I!?ot z3AJOAxbAkF?LIur5cOSwvs=pN!Q8W3tctZpsX1O~(-NmW5{7xuGY?(pbNd?G&|C=!IuZ3xF2bYyNtCMiT zD-fcBkN{e6BmXcdx4Xt?&uS?MC*Nuqyrloc63rN}#No?TZx3PD3N#MO%}m$M-rRwU z%#Qx2h?7l^l@x^~h#&0=OHbx4D#$VYK`e}Tn7G5=;f=XGb-1TFIdB$j<;RPxx$X{x z`*JGtKdIku@IDkK4xOEy(|NOv)|#)t95K zuEYF%_1Hr@H5Rxou4O?JeK{f>^*lfS-6V9iHxLVjnsi=4sJn|Ky=3ESSK)v(`nSbd zZk_{o6zw2CMLR~l7VnRPTJG~>PnvJsLj+a6Pc0S{G(R>L0XVtG@lsx&$&5 z7F%Va=V&|V*(FkOa_ z6XC7g0A-Jz-D|zJIC%h`r9n>?dn&OrCkYea(??f7LS0x`1v2G^$XDlvXE$X8Iz+cK z5b_u7JT!6drMgvARN#0s8cL6ELQmhQ(E#2Zb~bpj_CZGgjfC#g%ANMJZ~Tg-4!G3u z`TfliphfWsJ!v_QG0I0C5)z5}?N6~4Ck|C4Elg}kzBBp?L~$n}Ll$sNm^t>YLh!=7 z3(NQ~Oqc0&+?#&_s(+t*dwfuE5ZtgGmqK(lg_E?9PQMDnz)lrO1IKTLh!coZLJr$% z_@Yeg3J4#V5CO3#n^5D}Y~R9QBl%P!hgZ0m5!nd9z%7I=@AVnwmPD$aU!EE&C6(U~ zjnF2)8K*(HG14R=T~7w4>d3W!DTm9DRd+CrSUQmv6^VsKnhIONMyS(p555ISesj69 zSGr=ts2Qx5z>q}Xci5)`P_@d2o-9(WnQDlNbTPS4nW^x^2k4%PBwmA0^xn-)&06EB zU%@NfoVMD)=T36HeP_zF^GUY#cDz4+98v9Bp9V#BP`Yk0(A7A*-DFE%FZU^-bXkY@ zat0MKCac^(Jltj4=AVblV32bPFt{efsDa-eW$1-FnQ3ss(Rgv=I()c)4CPv33ET3} zv1-;1Zt}ir3cuao-xs+#SuPu`9FD$0LmPJzWM{9B^X6F4^gbDDCkEa*K?k-|&4Ro` zTx>kbWG0nzb-dgg9ks>WyqWZf($&?{%k93?;k5g5T7e7`NqG!Pa}3QV2kKC$gsXk0 z>4g~yfB?7I8-;kfWP;X7Z8mXmb=aL9U8Xmd$MBp9C?+fYnBQs~2~@wMkCU2Mtg|Jf zE)Np4Lj8_$fuM5y3Cf40y9BPui=f=$e4I{@C2sTcD+^oLej!JICiMXR6$OYRx4vOS znuRkB{?}vxL_i8L(FN85esCII?+B|XPTnXbh=IYj#h>4wyY+6$;_E>qioF@*(EIm? z+&T}RlPhUI8~zha>@h*9?+wi?7#IG+oT8*UPW4K-d)>Nma}B^=4gfjCUL_LA4+eXM(nZ8*u-ZN?tpi!&Nk(0V2OILhEN%?y-U#EXD6K^y z%sdL9cq3i`hBh$?peC!xQ+fw21JzSBZZ`Xi-DHj@60JC=>R|50E~_4vkI=DK&G=s8 z)uF|#2;9ltKHMo`MUabQ7m6p(f1L z?i|Q!rR<@j?~WY=mepzWN;?g*N^(g?CHTg`^i0z&L`8+aSz!+pD?izxn!r$1(h_$g z$E8|4iXaw&^~3zq&cZqgM7MMtev~)@e%CmtiVPWd>Zb~)MblBF?_Y4fO1Iz`u375) z5?{Oq8h0>-2jqGiw0#l8&|5gZrcLkLKU)1TEh z%ikPqZZdX=g*Ne}!5~A5Z}jguGM5N(6rV-v|=+4TyG9wc^^3N7+ zVPW;w%exG^P3spz4E6UTvL#D~B5xag5mT9JhkXwmlhp;Jk*Vv~ZPC!8Q#u?)mztzw zXpB8PxD}$xdV-*iPe1$UIAN;t#VvmpN3Zm#wx;@^ljNe{f zI?6&5r?|~LwY{L!vFV)tywyXc()f&SPQq4N z$5LDETWw`yQ|rjGTgJiGbVPBFh%xJQ@`{|}#v`b%$$qcMO?moZW_gL1u3s!uup+%6 z(UjCYapS)C>VR-a$B*b)*f0=aLHfS>Z3p~6FJn}}WYDtm@@L1fsQxAr$T_2A=8114 zKts{3rzK%m`QdNl_lGF%dc55)6T%c!^z>|M-A?2<*K*@-%k)%ihh`how)v>nBEX4` zMhJ0YNz-bHb$TgS3S}H;$v=>ss;Zxu-ybE9X8Mb{HB!LaGOxL{bMJ<=in*qV0AYMl zg|MIBc{a*MzaKYUdMGhz&$b8;5QEwq24Cu6End~KQ_R>p6A)xj7>-=t5f!M1DI*=O z91w(7v#{zFCFVL&lv&%_I=JyPwZL$2Qd&(gYX{Eyc)2$I_>$D~*hYGn~$KZ<8`BmSqPL!+a^uvd)6r7Z@fx*dzUeiTWW4;38EF2gqYe62KeZo1`oE)Q4w`jx_ za(!W|TZdctfHxo&{uI$!CZzxNpZ`xL(FB7~F$XXdjYwd-$eKbwk(n=PwU5Ac9y_~w zgmgzF;pNzjo@*h;=%oGNt(JFY<6=9sWn*uyUw#^oi1hM$c#7Sxw>d{FMUq2z+CKE) zWa2`-hF7R-Sfu-EnYgs~p!ifuXbIP)1Qyf~-Ps z(}j&n7GIE~a;cWa)}UMi&%1T$MszGpO!u9EVdMg%jJLH*i*?outLGz-XV1GnCzgCm zPxQ~l+RI)t{zVVp?_@Pvx{mnKUJ8DalVarDed}ayV_|bM7C&x7Ce&spVhNdNZ;<{_zXCJLvUPpA5UMzBrw zc8iCK6En-lJ-ufcv90-}*=9i?;$63_)YN_TJU+*MgeAqm8F@U*rTZ(#0r^Vr1ADu+ zr#3f!#HAP{R=55mW8=$`;$*lDL~q#1;}vs)>jKG=6r(m^FEs zwQ@J4-1dmYai-a&#B62ac`vT%Hk zfXx?Z`!>1;I|L4-86tuFVbMCn9y}}Q&Mtmk;ykE@yF<$IUwn=`P5kTzhlX5Ch>z>O z>7(xE%^4IY#sYZ}OgR)`O;W7jocjoevfKI-BS;SQcRv zAx9oLEh=|0hJ32ld_%H^7cgMeV1O)nSU!s%b<)hf%~X0*CM69?KT+bnJ3nD(B0cs~ zVW^(Q>zL{m1{PF_46|SRXk_%chvUBt`FDczfBP*HBUm9_;H0UkNmt!8AB%7RijJ3e z=nLa$UJA0sIJju3wN<}_3`LVoI09drT!Myy<}xtFqj3(y_03tCmd0L?A*U@zGbPf= zjcq8Zy3Dv|bLrBKd>C^%N9F7mEyA(q1_GMzM>tvy>@tssCK^(tky<7}LAN5Q#qPpl z@_Uw@0YaxDKDXy8EDYHXkHabGwVNL6IDtm${KZJ zfcP-8v3`kDqmB@*PD+fdo}a#%({Y78^}T))w;yKO3#-u&T#TfwU2w)^o(L273ohU2 zCb?jKu-DhhZQy=EAHw3^lRhO0>6A?>PtmlC4{MicHE&qc&b4ByRKEGZTGq{QRJWvs zrgprf_ZmNV43F%lH3;2+2IcCsT-+$+m)3@KjBfirJp zhWlh7n5{~_{I~D^Pqp%YvO21aP)=xPAc&KVZ4;^_gQBLv&8{1c7Ftf1nPaWnv%0Zo zph!(5Srt&gPN~+>w~g`{Wjhm(9b%g0Yix{Lr#xN#mK;{6-6~bv@YPV6jjRq&9v%S8 zwu1BsGL|SzsU3BOkKTyP@)BemYa7Sov44*1?!P%nBbrZlPK*v%eEX=;lj&$Y0%VO~ zplQ25aUy{1Ol@~rz`|@~G@H}u?RGov%Xt5aoKmsC=Hr06(PH)Wn5KLBsWwn%* z>TFO&#<}-K_&`oF{-NMU{#JjJ#e4j`*z4hL>$(F`y%r2?;heYpcFxh34X$0y^5xWw zP64;Km-BGjFbjt#IxnwJt^<9-bklTfA0ATh_0*uHS<%LysA*{@FK8;w#MQR6Uy~6d zR|UToHZjY9?TRSHx6w2(CK-S>Xe&EtG#XAc8!T9D!)g=EGNWMPx$pQ{hf$G}S37d7 z7353F?RK)zxCA5;y)HULu^W_Q4gAe@i%#UmgZ%QS>ndiW5B{%!jQ)AP-0Z2=Y8Itt zG%}066Xwe^FqF{KK3R2M6<49={o(&-Qm%-i`90=6&r$EJ8W zGZ@aktv~lRuzZ@s;BDu$^Mf){E0Jha8LAqfJl4dJHZxty8ZtYTcaax=Dx;#Tp0Y~2 zC_R-|GbrFi$&rS&=!_;Wqe{0H`+*4+b1U`W^9Lif?GnnDZ6U@;^2tg2nS3oP5q*yV zx15w9RFtbUdq+dV@#-WZMmFghPHTnu!7RWb1BEWGHSJpmZQPszf5{e=rSi?u`GuL| zVEGSzvto;T78YnZ>-{%GVMJRp+#DR+QK{yK>axXQ?UKK$4f?qJcGW3E35ji@K1 zIrFUUmeo@oy@0{i`tx{B*P4>4B@W1*)N@ou$HWe0l6^=t(yLVWsxdM2m)A$v zBza$+jbu01$*vdL-3lI7`(YXB82ZQD*biu0=M-|LC*n|N-8b^xOycJ>scBg!;abQH zTc;*j*pzn{hHzmOE(aU z5@mMny_XMnY)o;X$D^#6qj@4sU!?sNA3oxSf`cinr>Uo2QGX3g+^pXXE0 z^A5uC7=F*gd_z4S%UvcmQU~^w=|YV!MnChizF?u=vzzv&Es%{^^(# zYHaWBsqzHs7t`3-T5BFL*w9n$QeMu#^%lf9#9sOPTgjdD#IIle_s*tcJy$xoT1h_! zCGk0>aF+%BLVu^v`)&K%(c9)D=P2eLKI5v0oyZb)AMH^psAV`4}OoMBM|;%3{Kl8?B`M(E^quz*|s^fq}Y zU)5$PGg(uTQF$a+@_J2=8WQacy4OJFEs@sWF1;8BUVlE&vrEJ#-F&IGrv-NZLj9MtQ|SUH*HGLb9ak-99u5iG$6H z-(6g5;8$OD;bCY*JJipP+4~-I(j}h*H?7<(W0mMo1@pFd1jm%0v-e@@KjKy{dI1b< z3JaHq{C{~w{^NIpal@vG(<6?6WaF)Qm zC8x3popfyL!Z)yz8C{)(bThjwjyKd0FZhB$-`HE)vU4OK(?Vfky2C;+S7Fx*h(ofGIUZ=TIxbnE&2#kQnr8%ATll zQ~7bYyQ7jChfC;CZsSsh>xRN@C&J*Jv9XRV(kxXo>!J%SDditnExWIbC$bsj>hj{%lOX5ZMTqA$T#krxfk$9! zp7F_^lf`!i_Jy8Tx@6Gw&bm&E&tdwcn*%xsefpliY8#zc6Kod-cV>9{rlE#5cCkCu zWx%NE8_9M0D4@UHnA@w=X(ox*4oUS0-wV-iUeDK=m9=Hn}xto@qyjB14o#O`f)uf?|L!B5CE;W45 zfxp2g)FAu_>E@PpV&5VZhzsn7e4nh7<2z$G@}s?)aO@->T0bZIGhfp~z3CVCc&Y5l z#(8-|U%0RJh76)-f?)we5kD;Pf#-ytVm0NzHT3`I>i!=`7Z8?TJ`6vlSK1YDR##KY zX61ZOV6%d$FjnK~*=A?AcPKrTm1X>o5^hCK?rVB7>@c4ciFr!G{-(890UGBroY?tN z1@|Xo9NVi`{JL#!y6-eT;(9;F_L9r6Wt6hZ8I}3feAtm#kuK2q9{bbSxFr zsv{O_$U54}l{exz5#7_#yyvS9gZaVUp66A)pTB)1D+^nO`Vc8aw2d2xg(+QN>Iz4T zyC5t1dJCpzuTv5i*j@)xq8GZ$Y zmCY|T{NUeM0};l+3bJSkA!X$SpY#I;PzDYYE`sgvrBST9mHKyQEqogrP{=%qu6}iI zg)vIuwmC4lolT7))|@8csjJu;kMi488(;__9&j5r)E-z(ZDjZP$^0*#vo~>ppS%g1 z%&n}N>g)zuUu?(S(!IIOV*tNVQ*(9^N@#hXEwlk;W_4~75_)r+{%T<`+3R?@eE}Na zH0JJkA&m8{OxcS`!07iXKJbp^qcU9{p-{$SrL4@%?8od0UtP6XZ^{WPlhXCHqkcC> z%PF}ePqlsQ9EkgrG|(4Ow{hO87~ZT_uvYca1GxFg*4=ZMPoSIPI}BFCCOqB&FKET`rl~LrauXuJ%8?u zgg6aliu;a7it%d4ey^>aY*bP%xt<)}!lhljht;(|%Sot29MVo|Ga4S$yfZHwbT-f}+B??Po zItj2G!H<04O&B?}%Elx?5-MWF$K*}(|M;2wv)BFmC-pym8W+?lpXLAiEN$BksEEFZ zRHnFyAA|axqn!lZ%_USTOV|f#jEBLmykgEt7NlCQ>WjV~g zNU)wxVl}Xr9gDVVcfaY)Igh7fu)gs?)r+(=M_N>ul|VC&NVzQ?@goJ& z79z=2pu+<)04s0k60{Z8%o81w4LkcM#9-qMa^2#%(R5gPVU&Ym3**ZgcQFzidNMh z#lWPG3xrIlOtlxCk1V!}+nuo0dNe<7FuNgXu$?TmQ6p9}STMOeTWap=q78rtob`3A zJ+6p%ZZ5!t2i!*@uT^*hSJdv88w%$$-0=e?DgN^QvP#(n{-|#jq<>v4tS7UQCts8M z=JyI)I#IqJuZ_g!g?;`!ss@2zG)@-?G0!8VkcrOW1(EtDyvsf2)P*98ryn9Fg8H{^19w8qG{hz6(h{_J?Kf zvL?E6X_Mo{P=d=FSHto527F7l?!%jgzS9{W@hE5Pt`ZV&Nr#E0P0@&+@#UJboed`j zZg+)XAWC=S3tYp#X##nJ25FUji4<>(|6bK-m^M!6Ouh2!O9bFTCvg2&KysV zsY@n8aMGio9Agn!slAY&L3?*A(Od8r|q*V-D{=4a$U#m zg1|y%6tsU5*d%-Mqkrk{_twJHFga5)4-ZdJ?9Ii38sxlG#O3zXj9gb6>fCX%!O?u% z3h$dj!(N0S!)|s4E77P?r)Ld|p6UEkv+X8MHe6xvi}qVTNx&Th)RAY5o}2q<)8>yB z5&%FblAm%GrK_P+# *)u{sF?+FEBC#jD$(=xn;2O-vXjzyB=*Sp1y#7 z<6C8H`_xm8LZR3v=VV5*I=1U19^;E9%Vzb+tHBE@cI&PDjO>TMot^8AN<(^-A`VD^ zV8PVqoL11tkr|PiQh5>Pi3?o`;4wZ8-`TF#L@eM%H6c^R+=tV+8+tR|$;6|d-UbBR zxKB?`u3U~icaNAC5QkowKydUcOQ)184p6AK2u?X>_Rr#UV zy5Aj!t@kTju2k0U0cY9mwuomP{_@oZNK9hd~Y|B``h&7_hVJhLp*h?(}AD}!qs<;*78pSbubNp^CY7RmCMJk_ycbmr1O%<1)wQydS|~L0jwAx7v31O#uOau)8jh3mX{6 z?E&j4P4DV2Jz4|K@WqlP{-~bmU#6+) zhq3>ib=q%!OTzEO7cqK2N!_?#UNX`hZd<15ouVoURqdW2QU06D`R^v4gdOxAyDC2+ zseTTm>lk)dM|Ud3lov&X`cxj&xdw(@pO{=95l2_1Li!DNM~|YIQtl%3TH>&JMB3%- z(@Nu6Yz%d3mzA1?qFWA!zdtEoZ(VJXXRJ!GPlLz5*ur{4i?1X}mk(6^X{QS3-?9u^ zW2IK%a~geA-RKOjh)r?s1~tJs$k7 znT2ajX3|q#K@W9xb@#L|v}R-4ndaf`TdSB#SwoCWRQ6?CH^Z9-?86|xrzD#qZPW~O zvDD!?zqknHXzrMt@}PlhY~CQ-r(z%Sw9oiZ2;KJWsWfYCMLcdS1Nq^)GZ}+>0R1&{ z_#aId`YizgrE3f$b@iZ|?j@88(S4*qAh3y`OWrn4W^b^veT7`5X#ZDe82D+blC;jo z&+OYf-PSM-avfdh*(2NgJMyKoygvdib;oT;3t+rbIFJF2J6+n-DCxxK2#D|(F?Gyyt(7iaD=Xrh6V?P9-NKMSS6DR$qM5FtR9PxUh z;yGSSSo7g^vpb@BGFdH4T%;o9>`$ueozJdqSsSpMn6wWaKP)6rIKju~=GO5@p#89n zaqdl~=zy`S*zaXFA=3L_pos%`^+9!@g-exIU&R~*vL_%R@m$e=eku}Cd4D3&2(0ze^!)2{$Wo^iz96Iak;BStI^a*k2e%mz)#@-*{*4sUvu1l z9rbQNEQ_}6Uj$5`Uv7I=aKGT%i4Q$TD?_*VIll^`pY=kZPavD} zVK|eEi;JP6V)Y(x?$?&dRXltaYwZrh?8pi&30X)6)MwXD^Y_UPkjX}rbl;$??o`wy z*I=6E9Ipdh3CKTZf4;?J;%*q-#VCfBhmncWC*{X8Rlkk!PshNcnn}52T=Gx{y5~(4 ztw9_UE$3>+q@M5d*V?ltN(1?j13|p4K6`7EEQV?#KTiJ5n+p-PXLZ5mtQ<9rpa*4b z=H~0|nwa!^=igY>Rabv(K2j5_ZS5Z(PW{*rfmG1Eppz&(UY=Mb3=>VWcCHMEad3f# z`M2du3L337evw@h<06*UB!DYGKeL|;%;xv0-B^rD-25qs;hRwiR>HZAeIs+(-o>w6 zMA&GW==B{4?J<2W?;;^Z`s z>-4h3^r$9lVD{XM+TY`tW2*SQyh=_*EetXUyubtZPPp>lUj68+w8!e7S<~ul7U}03 z?;jJ2{|7rsAr2~186x|$Mvoc6z>vq@Iz>lWW_I^GanHe8TA3W;fc2doYNV?jTNM_* zvbD?umNf;0BE;z3=>GU&{4 zMv8oo(~dRntG`>cMvCYqht&(G&VQ`qE^)I>dw5h-uFP!DPf9`fuV$l%3kF%~;MAu9 zckWR7^KY3MV1@OUzA1t{F znGJ6wi>ju_DpuE06c|=u2E2NXcHfJ|gh`jHKc=dYU%#%OGFwfj?Xoa6j04vhxk&rSWk8(dLT?FRFyphLkQOJ1qciO9X> zs9ScE(zr>F135EW8+pyu*o06k(Bk1@gDcQdrju_`wx5- z$jQ8VHqS`C&weZF_eC-fOV)Pp$q0el9a;z^cmwGF^VS64Ci{!GI+-Mn#{1wg#C-SV zIOhHRBY7y1z|nvDZb1xcre1Vo6CZEliPYsyjt?(EZ-OOT&$gPrG3gPSHvVFeH}&%J zqewZy>_fcCBRMqzAvjgM>my+gU$K|hFMe?`l_ zU4Ir%*GV3brvP1ED+g%0Pte}M!3ABv)6+-+hif@)+C=?V)(v6GN*fm6)lEhOl(ou( z?NPY5UzH?7uoAi12xV%u-k&3t7Ie5Ct2{K*?~zw>8F)s$e>vtpV=Di94GgSA^Z%}C z7rf+fsq=72gwp8lGlw4aVUZ|UHsxO=;=fQ?N&ra8h^@Hnb2eAfUq%#03bS1TqCa1# ze*}ozKz(hU7VFH?HH#RqKwsr@Dl3<;_1*7&1P3!G=Pny4jSqwfyO)*qVwAyCt*6sn z9^B+enf}$GN0zXYp|5Y-FjgJHjhXRQz*ns?;A;$YBfELZfPvD|0p=pyEIe2HvNv_O zrioaYdw`^&IRYRviei_li7dPeq&zOk)r>0J|50`nRcV|ohX;oNyD4H&a~|NAz(^xr%BOL2ShNa&10S70UX}msZroH#c{?iBVg=y`2L~(>( zUJ-yoW(xqjjc=#Me8(t;B^u-6;GFcX&Ji+dl#O3$8^qQpA?owH(s#3@8$xzVUX?8I zJ;R1V+?A~~Fb@4D#**=^hT*orE|~ zG!h^;&B89ER$_X2c{%f5Fy2EGxqi%^zIXE8n3!Cj{~fDUae@i8Ns}%W9=0mIi7=cP zCZX0AI3lB2u7@Q`?{IdEf!Fy<=--#-NrV>Wv{^NWjdeBW;Ly7T4DVf3 zj6B6^ke3ID(fW7h^kT8~p;>Hk&c&%$Pvn$3$ zbb}+tmYSA9=0FdSF28+GJVPUjHn=X>c|x0koBdBoSg#Zu{4j@1%YVdhJK%0$4-qPH z_1X6##0;@nF%@RQs_uRohG9huCRCrJA!xGkOrvYei`<8ov3j7O8O$O1bjFMInHd-{n2rFGxOO)+hbv}> zab|FEII7mW$HL@(3J-pNioDPNt82&^j{cazDk%Ei6nrb1Y=bV?P~JZIdye12(3VY} zPU1uE(|dj(A>y$1+?AWlETTlMVTYUJll9~j&ubi|m(aYNK>5OtMvYD~^?mGhXd=f2 zTi9La@>^Lhr&^if5phAmeVM*KC9nu z(mPOJ=Jh*%ilNqNJ@ydsW-yh`V4&vK%@$EHo zVH(YFr4CrprygcMoh-Oz{>UbxU?^R zlbY<;S+sn^P#Q%=5zP$VjQ{U$m@QuZ@lF4uJJG(Zcd3w8Zl8s6*(ADO$ zFaij{fYjxKpl4V^x_^)9GoHX<8vM_J;0meT{eAAE^%#M!*;Z%&D#6`aKsv@~8%KBS z@a&-!Z=6niN-DRIMa)A~4A?Q53>-%7yX+=5oKhOiSn;LhuY4@SA6#+qEXjV%U23E_ zPg8M%1KKazug4FFnPZt^i-+t}Gq5I2@&jATDgw9_2#+RH!$vPmaB<(oEDkzsoxfVN z5w0aq6vYY~_Am9IJ5w*CD9%>y8rC8?K5-e|5ZS1{YqqwcSIdhHZ@pEOPd;17KHV^F zK${sZA4qY#(Sn%IxaQ|nb=IZM7=Oe*c0J&@uRa2IjZ=$0gUyD<_l>z6)8`P)Ygyz1 zQ5V#LGq8FI0++QNV9_-aL-aO7n_YSzF(c_k#qGO#8uSC)^c9cR#n=1jA6v$ybmlX; z5Sq{@C*6YGc;GPKA`$27)@V$p^IMA|2?m^~o1r4{^O@vlwc8Js2=Ce>zca^QXwKmJj`1{>%8Ip(TSxGCOytI9vJb89;Tymp9)$4so{qGJ;r{CYx zKQs5me~yRNQp|hvwjf2jOf|D<`STl*di$e!IkI0O5b&LhDFh1`eko{iK1p)_U6|&D zyy{aWskKvH;^(4D7;LZ+_4p)@{j{y_mvAd|sd48;5DQDA#P8EZ!!ZxoF-iAe>NlC9kS6pSn7hb2)`LNdi|sng3O^qO^%=s z1CB9FQYoG$7Yko&SoNK8;DD+pNvSF&Fm8VGWSQE;tgs}LJw(&}9eB5z2XfS-IVwfJ z641Yih3jvSoqL+x)ZaMX?MD}ZX)qy!;np(mY=;^bKQnQf*r#2xb8FYn0``)vFxA<+g@7=9PBbN=4Q9OR+{wDhAdEM!F$TaCLP0KDwP zTv7(a*WM++hc-M`2KMyKjYE4lBP;9FWqa=FDX*?ID%G~Zx?|$~L-iK8Gxnbr z`;*tnsfjCY(HeAWbOc$b8em$+;m`F?)mg6yU=?Mh``Zq8Rn>Y@{L0D-XCTwT&cVU7 z+EwR>pOGXQ9`g2oZwR?rmKZs z{!A>Xwmd(7-XfKej>A*5<5zqir<_ra_zG-F4gT9@-hIZOiO-^y;xyG*>BpL(84K4g zG;@Ampkr?LVEjxIk2>ngG z?WVUKCn3}=cGKzbfZL9HSU_N;e5VuQbF}D24Ggvr9dKR>BnDF8T)(hijXfx}-f3@x zvwkm9GwF(~zjraAYGhHqNT2;mv5<^miWN;jX9GGwnx4|oNbb4>#wHZl*}&B zh_h$cYHImqqDmO_5Y2I)v+FN~7h*T$f~G4gh<;epbkxfkpI{`3PpNb{AmfM#cf#xl zyA6A7c|N$r_ad*imyvoIxD}YrwtT6vpwUSKU*Aao{fB;I4$DcT#9*)QR$3T`Xe`r; ztmVD;`knvu#QfGsiots;tA=SBb4$z1c3LdM7T;@lvwI@+*QfY_kloV9$j2;Fc9jR7 z1J{l&t}cE{0VU6sVwqs0A%|kG<*D_uPJL+YbakVtMRIe5o7EQTgypl(Eb#DfT8@vo zi-eVI`nCEHTs*Szu5pH~bn&KQ(O9Z% z!u9E<3l+EtYjnzAc?O*Nns8r?uu*5OYuNWY28dDo1BCN zhjCrw%l;f|t=Bb=SJ+X9$VT}P+GHozvMoOwbbMzRe9*VWjK{ju@Wb=pNdhpyWSz;T z?!MBtDVYDTY4@2X#+}gD!s4fmZ`tn}HXrNIJ#=@uMG3U;Ar{D;f#@aM%(qA^>Vs*!+-qe4e{4(6;mS_yB0p?|4#K@7^K zr7>sYq~jv9q4Lr5%L_UZ)~xujFH>BE+VTNzI-4{xdZ44J?2sjI(KRf>8yrD0FcraD zBR&ll*u&LPQSZ*_RK8#Qygv=lrK7tD?6*ldV)ted5FUW6pjdDx)=Qix9UZV*V^in0~1AYgp)Y z*h=xuopM~%+5!kRyecx%QDD05+ipY{hyggj4*S`!u;D2?#HtyhYd2=}%kkX+S^NBu zD9T76|7F_}u7OagzOF?u(nZ437&q5P!rIos`Ql@y;X8(QI50lu8PBNpVKn5If%qw% zZ}_)qGFP!jZc7@*JK`R9zljA*yR04SYJD!o5)y{*ya_1YjK(EsGaEl?0jA761sK&p z!aQmWxFZ9l7Z2&BSc3>h?-eDZ#L*P@kP+?R)rJ>LR#n$d-Ing`vwy#ZX4Q zWIM(MvN~|;N>mLbIH7D2G3C>;;O=oT(-{89SeY8k|B`LUGlXA9W;2+YqWXLyz2Wu* zvetX%P$AYQbHXn88DXTv@=d}=R}q{a~A zs7@qt79Oo?BzUCA5v;Uv{uVp#Q-!Vh+XMrg!ic0VUwWrwsbjIwI&AphqE$5n{VIAK z&!ryMPj8uK9pjI^fBf&%27qis@PcFK)97G`{qic-^EhltzZ9#4BQ&^Q!yOc%? zi2bXtcU3p%`w<>qNiwEgn8n@NlpFs1DY9Mn#sPc&7pL(UMwY!^q4s(6yKdr5#BT(B zfl0i3vB{E`;{yMu_muO3_DWg&#d3#1JQz0OyW3sWsnRs#eLVFW$HWRv{|AK?yG(Gh*f5m(EvoVxHNEq*>GymLVL zX{uuo{`#LywvQG{176z$zk;Vi9V#M;by6L;xxVyP3}_>PNEKryi>~Op;KPC$y`?91 zgY8ypTImLJ%Ej+Cqv_wL$i|CQbZ7)81F3O%-JXD9GM42!un9^3dY5+^7;#8ms4w}% z3n+um##k+A;6IBUlLK=8xZ7_ZJYY*xl$Fw|6skFY*U`Ob)Y1wJIC3S0KzkBga@l}< zej>5mD-$w*@n5nmae~Z(>=Uy@RfuPvL!ZixrUdi>qwh~ZcDxvSkEXcdywl1m@ic&; zzG^W{=`O`&=rKeAVNYk<%0??XiO5GtUQ+e&Q@TiuY}8~0TugNiR`3(-x6$jFeHN*> zgqei2D`uPH>R_6CqWEP{7LQb%lSvw;SUmcd zFSj>Rjmvy9Njon<{wNgU3)-(y{)4CiDH?N}`Y~5gR@jd$O82aQ^hB(`={VIjjvg`I z9b*Q09kfb83+Vm@V&SCg>B|?n2`OqiplHO)exJ{Vx*9}{W*oC*hxz2|UT9}WgJO-2 z4Q;$nQ&yK|RnF$t(Y=W{=pxuSsjy95CfAjV+LT$)Vh@>gIf+OgwuK28{2~0#@|B+` zjki&U*_5hT^<-eH6OwU@S;TWEbh^6FsH)F-{vW)`?m$H+Iq-!NHL@;m;c|5pI={PK$RRa6b%SK6vWu-mAP##TcQTq=zX z59?zK?m%}cY^$x&V8aDi3DCuy`Y%~Eh**Z^*-J6uYtsQwRX0lifv&3uL14m^a*m;z zJJIALFjlJ{E?Q6u7Wl&M0RJ25Z@*Rx$(uMDeC(jl)_*naH>y>`N{miRvIJz)H@kL@ zEv_WR8hKwuGL5E20wzWF+RisyjF(GkD5c_Z>go&U(wD5u0ew3i(Vfo(KO!n>bJ75% zP@dI>qPK2xtMJf%?RiajfRh@ia79dy6jdP8HMr`~Vl=@s-W7jGwaRd6Ap3drnnMEzvh`x~f$)b_?ZoiE$=E;+OqbF}ih3Cq zAEw{2nxqe5sY9E|rvWVI|Lu66q?mnt$?CNsBgt&N{a3swA|Ei?>$l?j@b*dKR6S%@ zh&nLuePn4vsLi5y|JFR%2@M|mjD*(iW!LQJQtdL~Z=RPkU1gOW2S&~z`3i51hV~P_$|@f|!sB>3dmq1B>bO zhQuaV@4k1%T^jGp#Me)9Q)&jupW;eB$71O5<0b;A&QKE!47DF)kyeU0<}q?h;Ui+t z@p<^BBw_VN=n*m- zHM&i5=$I{%`GzxzspgF3A;FKh^PO7-ySSDqDLxg2+SlBhConcvjwkJmXGq&Cg6Gh>T7vDn z8?f$MeYg1cOcjqD5+cya!mfTatnq+JzDh1r22cVDZYE81i>TaUqxn-ClcZr~$-KO= z0!bP|Ogpx=W7!S!`^8Mg#rgeRTRTI5V~Dt(2ir9+t}j(QHHVdz4LXwgP;73!&ROEc zviB4oR^c1HP)C=aXxr_LwDbfP$}-8r=hEDv1J@tv7!tY89Bs*VWuMy^0Iy@(3HGS) zsC;LUHwmCVb-C!~%6>(G`3z*r57!{zPJYhMA93q zSfDpSePEdm;?e*wmV z>BasWReRc~I;L`D!=ZIP#(9j-N#+GAxF}n2IX^}8-ChQTTO!n`V+9=@Lq_+?TRSg7 zt4XOkSz5y!guy77#c#Ya&VfjCZxOCdz@5W3w$2m%^xKr`U$plhu)XsKeoAz6)ts

    T4DL}_+NM}z=%R9z8hQHlO+Am8DbiIH&-I%7UcSqmCf$Z5L%DJ}!X@+$T zGC@=?O`&r*A|kLwRMzdKLV&oQ?(O_o`!Y) zS8hpR?7QnTOl87tKAvE_sH}`zz@J$%EL*%kY&2neuVJ0`<~gPq?g>67Q`OjC`q1qf zILaCnI49%lOa1(}?9B-E9M5VLmNsz?l86J+gUy^J*f>ooTs1w%=jknF+fpIni^vn& z!nqL-T-M74HlrHW3|DC6JwB7*MDXVDwv@=qxb)b4Y_?GK5nCL7#lbQ9^(!)Zz~(qs z#^ExSAw_-uf++O7=@I|A=Mutau$$6YfoS&!RXcTd5!2yR3-xvN6y>HKMcTS`cXcKdkEKw}{o?z`Dt0Su3( zHN>no^5<1aQeVLgW`QrpwldL2yenbP%QpFoSFgy)$j(QHhMTECywH=WF@{lK+O6QD5D+N#m=I zNr6~Za@i`(d+mB`vOTU0t@LAN$K@{qEp@^66BPB8y{AC}PMcnwB_qwsp@iRB3u3FX z6~S8GMHTV#WD>{$^64iKZUFbRQ$5j9QPqAw`CuLcuNOYue&sJfrXl3Bc5vZel#lL? zhT|PC!p-wJkmWk!*9+$c^g%R54uYj?gv8CmTsac^k89U+rjzSh<2(s3@!}Oyg8tp9 z3@Tm2IbJ>5w*K72UzXC!~5{ZOKg0ZlP8;D#~d zQ^!59L|R=h#Qkr!{H&ojq=2$kHDiUzrx0N47<>R+W8}kz6TJjG^HxHV?L0aVh@HA4juOQ8Xo z_{~)t$(Sg@E`E#_o+D4gC4@r%PV{ezP+y8(b;7ZhV1BU%?~9k&IRvIb-8#P0KMJwN zV|(qVC$&=d*7+OqA!VL|N%K&+yEf^R2q`lJR$zG3NGt4C_!8Q_44cLLiw z!%3S)?)F(dqbF_37KlPGZZ67zwQ7|x02qxrhwEuWRWpR`K1@p$nPM_CmylSu_SkbS zgx72O8@vCtojxvof8K5SlsPrHWvRHm(>$1E52MZ{58*XBb8;~Gubxvwyvqz|8& zpo%J+9el$zziW~p0qNr$vrsv)HZ3sv>Cg$yi$?tFPdIX#_ zm^Ac;NhO=jJIuY^n?t(!ye~Z7!p^=SZdmF`W2Hrn-p>b|61)mVmL@-nME)^_jYX_k zNu0z*>k)q}D)}_Kba^gu6*coaj*hxbHj!5Dt%iA=#5IbeC4w*7Fhb_y`~U1-LFTkh zh+B&Q=7`#)Z#zGJmfh9-bMKGdGtm9vFtWL9O8pr<-{L^58uKt&NC-W_I4R^{qDx*P zvVom#3xsp2(T&Q!?b0oz&sXFS&Lf|d4~$PWL@gK{mudL&^Q&gumXf;T<{`K z-!~vgPaFWQ+b~b%*{yQ-lO=CG4DGmKSh^$1(8xRpxTr&3=2+Tp&9_#Mcy^jJc;F1eQUf8BBa zUCx-=CE)hP#F>0|>goi1Uu$Up4cH~4LQN&IaknEc;c2UOi@ZC?{sR2~^nZm1LFSdy zyQ~@wDw3~VF@Z5Kbu8=UaF9{Q(K$x^bgpN5-4w?h;6mPHXbX(z^hn&baol*q`f->n zS1a;EsY6Hc!#)T0mb6}9&&Q_^&@wv7YGmr2*)rf9-XiZ;O`-x)E~y<+`Yk|cq$}a; z@*^-KW)Coz0slORm|wcDrlyB>ucw(AtGIurmMr6XO(qKE&f(&6S;ba!Wq!2Mw+i6p zG8N7BVjY8glDrG&~3^$!wVPM+X_ ziv8D|7YE`7w+)Fgw6!}rsM`!z@GT#kUtB+vcD zZIM$gWxcgZt#OX36$}hN#qVogRaxuX#7Z=R%LO-8M!os%&TtFbL-EQbR5hE32HKpE zS((t9Jb3S8-RtfJPSS-JeQMW`2SPQ#AY!WMa3l717_~`97lrhsl~hD_=U;DyRGy`? z5@ENSAF$PiXViVh=xS2oRL5^@e9`fa=$YeO?P7Epo>=-Evz7n_7uO~2Uggl}=!rL3 zGtP-7)^_`Z+ewlgkbUj?ARyTP9&a_& z-}p<+2GlloL={_aH9G|M>oge@09H~Ilcl*)Yrg$-Ldt&S%(g3l0;wl(FF(BG)gH**5^Qf$;Ky32^-zza=E?V>QgXpwvcDnM%v&J zj6_TJf(00BVFAWk(yy%o?nv)dP4Yz7a=jA$7qKv(S28eT**vZhC~3jvofI*h`AZ;e?RhUzx^2&5=KWEYB3^j^Y+D$ zS7oi|r{_LdZEdLY)#V|q+FsNh^0pAzUq?oqi2l}|hZLkAFIkC8{WG_d%B$l# zRjvFPZP|m838mO*-nTr@9#m}jVBsa%G`kPJ*0tbxlB-C-$ugB>;LRCjBnIisFcw|+ zyO}q+lh&dwi<0ujWat)g5y*$z*d0+E;)x4>Id3yXkh>T&kPP6LpA&51i6PRzw zM4(Cet3+-|Jx;PIz3GWMFbL6xkU1J%0zUEJj@r19e~ZE|^-1RP5r01ZP$s!{9oe6m zB*69uZ~aSKaT!xpaFjmyl%^^Jw;+gaf}shc%#~h&(1ex6P2Or+R>UY2&$K9PTx~9g zxDHR{+Xjxu(FE<+Z)IVJ3;Hl-JI2I! zd725jv_T_29-i*|DB>W4kT>)nCR@>qBrZ;AFUUMFwjU619u;j3;bz% zwJBK%Oc$R{_$DuQ@H$BPPqlbDxOQ9^-2VW1$vRklYlvzuHpz(R&<915u-y(!?(x(! z*(iTQfBdD;Yj0Gz7x2?Q89xG0``=(JQhH zDrR6hK~|V*P=`P0r|rbkv+o{is^`tG05yzl(1%pw@6HFyA5g0-;cbsJeHD7zJpfe< zQt}+VwKGU3RUAku>>bjgW2mbr#oLF9$u}Per4HgRlee(4x;|>> zn@n`b+~|8p9?xRMkEPf8Yo=L-cuWh$d7<1nJq79OEv=vDX`A2^Xh_4aAD&t|>6IpX zq?|?`G`-|Ey^(fU3-85vL&L1AjoO?nD$GfHY+fQB_L!WhT#^=VKeKeU_XDe76u6H? zKxtQARlw6QazjS~9RMef@#oxNKM^+Gs!?Ulq~WkjlQu}VNCgX?`GY#=wd=6gZvg|{ zg!~Bz+TFM{&sV8Qg6fdGFM0TR)pW!TcPR)ub?dLN{c>DmR@3jB1mm)S#DiqbI zf+?UZ9`pY;vsWVA9u$nQQsk|f;7ATnM#v!3Lu$c4a4g?Wn_aw2B*1zX;v07?bwRf< zaeaM#>#saY*c8k{BW?Zrzq>y;)x#^p`C+?7#uUP+LOgbAU~ z3*T#yJ{cIQ<|eSq6a71_^v}^$Jxq`kFMWnSX)yoR-!mHjz`&Y2L3dU0mv4HsO}7KNoKC+A0;Gm3B`j(cqq^9=Ec4j0w>}AZI%)AxcqOn zE_xCHA+nlatkVN4AN}KDD>`x@obfl!PiuJqT1P$^jF0@wP|5bGM!kZ~%{tH6>_OK! z4e8t80o`<|2*})>v45zBlr<#;>5X=I0J{@94I$BQR zyvHpy+dQo6^S}MPMQ0qCp%ZX^Z1;b)cinMKWm#B}77?Y&peU^LCZY%;bpUA+MIqFH zfPg`yH|YjJ2c$_a6G27kQ9w$7#Ha{_P!usi0tApq4M;Dc?9*9i1N)oZogMz%{FPr` z?!D){^WAgrJLlZ*sSCw5K!F3Hb=n7IPBbyvzV8GM2-Ryp(UBJnpAlxviD$l^hJU&F zCdX2JU#9k|c)hSSY%`W&b=+2*Gl^le<);1}nY~2?WkegoexxREVVbyiRtk3Vvz+<} zqSrZ)|He_Rfse#~?{ry?_7M%50XKhYc1Cb&2Mbebs`Uc~O0WDS*O=%;9v;W0McJw) zsesDCtON6Tv+&$AjH3e1qVKGoA_fFEMxAKuN~^T?Qx`StwWhNS?9(ISif$Cd###p6 z_guc}Km7p>*7^#BKL3gYd~HDiXb$WM{-uCvz)}N zfMjNG{*@qkhulbGA+8C4f%>KIl2L5FYGJX>2!V7Z9%WE$B&q*;tbz}m+dGLT0*P|B z)u_VVSG5zGLVOtKR0!HYxS%TS6$WhF0Z!b+jEJ=CZ$`-jRT10q<%-SU*Q+KB-?7+ z2GMR45eWKbhi+og=&o{%*%QM{XY_r9bD0^+Ox1O~BQnQEizW@5i<%Dr#H)0Maxrnzs0Q=ooaW0xRUcg`S;^VYvCiB#HOaof zlb*}{q3h`kaMaq-^w&SOVHGJ7&9O$8=4qF*!wXS{MWG7%-pffRjHah3Ul<8pT{q&3 zg>Cgy9GIn60s}8KsT-OSx83UHaz6S9is1u)~8!Q z#(`it@IFw%lymNlLFo{hc8sz%9`zhJ?3u~MZb5R*R~kQrQ5o8^7g4S1hWR+e?dTDlbSnYZ99D31!q z(kO7Xd?U|5T;!naV*R1e6XO7_4kfBZ>RR3>ZsS`u>ESbX0Lo`2liGdNesj9h%tf=E zu2aQa2T8XZS`3Q9HzxK0z(lO`mg(h<)J-XmE$&M%sOe$&TNkppe62Zd)VE|8FIdTA zWu&>0-S15BT(VUsPRD*KlqB7h;xZ@x;jmtI(^= z>tgC;Ksw*6nFHUVX04q{=q#zrDVdiz6g*v`V;DJ=`)D%6ZI4M zSgk0#6dh>>90v{F9r{J`nId%QjwIYzOj%MUuK048M2;D5!Dk*0)ZVtmwn{}_)O`-(wD#%IE0NIJvkqE<{Aa6a|ryp?qguiV61($6DnuCFCQJ)PFN zx_Za)5fPbbJbc*$!KwJF;GAEdh;q5Qx%j*hGn`VmU)Ep;yjpMrlL7D$t9ij*TAP@~ z%$V$ZDTdM?SDuYm?l;Ri{EheQpwPo^fL0`Hj~6*M!i*n5p-^gox){0DFgva8ji0$+ zS5*ev7wfD^S>KhtnBuWlK`SN!`z>$cZs8eomd#y=*!zv?mYGI^h40yOTfO5fKuucI zNAmS(L3r9Hh`^|0RKLUHtt;JCqHp!=o{rRYGl$%ES|Gj{~4TmcM^i_5WtZtvT@@^|9Z7y~HZeU5#gp>Zg6J0ln&9L`Err z-}wRc9RbN9UvRhkQQ$f}|D=f=dH``?I%7}HsU@u^ve}9mk6qJph#Jwd7cy1I0|U>a z?YyBypcefZ?qxziI82M6JN=-}J~U3nXDVqzfyEFcOex}%#NCGsq0>lIvDU-$Lm0a{ zm_vr0+J|lsKzddm8*4QEfOuo-zD*zC)s)s)`NzhPt+SsC#rJ1el#^@}AC^VxagCN< zu%^!44FTmJq+J7zl0#@};sLMs(ERIUvz2ZxX+=Z7Rqr2B3+q{*jjIOBWKQ(4`#CCK zIG<;3YBAv6t@e!gtEOmfO0cWDJCX!*0&6+;YFx#$p!63PmVyI@z>7xR6+U&K$l$d( zfazWOpuod6Ac}n@XeRGhtgtA^N>nFo1C>9CQaFC=$m#nfv*s(J@Ny^ym#jbUVC~-o z{l#zkz1kS++JlVQ8;Q?1h+!1brI_f~a<=ZCd{lLavBhW<8J-Jqf+0ZhsYoS7ls3yE z5Vd)jPyS`ic43x3a3I;VYMUBkFg)GoLwm!b_^;2RF)J)&bBmDf)QqFDiugu|~#w?G@ENnRu z0x4ODz*|0e&?d~0m|YZ^51&H>>^_Z&`{H|(D4?Eyo+}c^SzUX*R3&&LX+nNx0*P3# zRZ#8=E6IL^|7ZspRo422S5k&1_pT=0Bs(RVXDs35BlZKAT z!BXwKVav!4L5fGso2hwotQBW>b?{v#wSp5*qihEX%F9S$WW%z(R+XiU#jDolHiN!7 z$yf`}tLi3>B@K|NXifU5c{aho7T&DGzA(cwGP)C>I%uDA5}t}t3yv318Dv_8XFw)I zVF0W47!P#!T_|U?+*?`*@l!20WOIY1saH)}Pz@bw)_)nDc;%4?@F>g%hg|Eh2`Z4Q z00ME|D;GizY@eO_LIC$i)8kTr`$H102kw3Njr-$(E?UvcNco><1i#N=_;>$X=>YEY zS6&}lKEBPwKpo=u_y~8-hydrdypH!VjxJY|v=8Y+b-&kV{Tn~9t>-`XBqDsf8%l7g zf>8FYOT}O6+ro^IqvCk+2*WrYs|v*t8v_IB~MApXCsf&aDgfOPm6Ad}~miU{G~ zZr4BIb+-?g2nF;PGY|7Tz5U8xkk!r-C>^dXDNmN$^OuaXRj;Ibfd0cR2awEL0k<_z z@+ZZXkp>v~(joq{PJax%KaC7c?{B5k(3Mvo|8&zoRtD~LZkPN-dCLXNz8%EB{&7J6 zWND3>Lpunn=OwH{&NgPBaOpdVe;lB{t2c$kd?!Jb#tS$Vyp+ny$sNSM5DL&Aj9eSI zzJs8`mVaA{EuF-#>?HpG?^Ji%sbV+D{YNUFvPbRyNgWq=Fn{8wC-a+|HOEd8_^CXH zj$Fw-B8?!hwJFcwN*yq&Bfa(SK5z&nrr&x~5?FfIg_T1{^VY8242=KNm)l&P8Ggq2 zG5i0i_3vW_g1SD>6l52;n7Or zrA7;WxmGbvmt;H0ywFU4WjRt)(yiE4qbviXMW-iRn&&3Da>3C%xv0nx@08r*9}p#S zG;_<1Q0MwZH`XlHIo7{*p|cW&j78CF`RhWXo`NsW=L*I)yGQIl#mZ6 z-M!7d0$;MR?r{EKO7sg_{g}5f=hH(KfKdaYdlB@*uCwVJ))m&3);*=G@+D}j2Ws93 zrSE?9Vs0LiUDS<4w)MJ3Z?<76di^YQgEsBEN}2soyQLG>d9im2 z#V??W;pFDCh`hUZY5iZ6DUG#@JhEg_I-p`tFd0tuSa7QPuRh=ZD=e3&4sH=>$g$6fRDks3pz+GyNJI72|s$N literal 66403 zcmeEt`8$;1|F2z=6h#P?P}$194-wgwvNN*pJ7XUblAY{pp^|;yjgqYiS;xN1He;+~ zne7bfd!6$==X0*}2b>?c=6dFt`?>G;{a#+L*ZqDXwKbJ#E;3&vA|j$ud8+uFh=?3T z_^8WzmZ{=$R zF{Jbl>B+tGeD6Gx zcW)8j?GP{OuP{oRu8SS)gQ-qc{Cpa2W#UKJOzbNv`Lz4RYj)Wo%H-BSX6mR?vz7|jUQ;u|Px4z?&J~4wUg)>HDoAjHR z(0s6MeEVFydORsAI6(Wt<&gY5k{6+pzb_cJlm8}e`u5V5bej0E{azZ8^g|lc5b|v{ zUWF@f!g{V;Pkept*2||Xg0Iu!X_mj3g|cc3fW(83@TXQ)Gn{{=dX#g^oP=ZUvcFNxa=HJ z{v9;`ER7QKn(=Dzm8))wR|lV@y(L|YB;!c9z&H_k?UvP3zMmPZamE8keqfCB18TsJx%RFXn#FwCvz>&N5#0TUWO3;tBgABH!l2=&6`W096 zp}!B-MEAY+@A)xf*+Rd)wHSG1eM3+E3Kv5)tIivV5WCKO^YBMHX-{|=c)~`z3@v0T zUyTbp%A&+kH&GY8$)sYi!5oX%T$RLaj34q}O^xE|5LxhZU*Y^hUqaRPM*ln8V~G!8uG!mfwo{(oeQ>oea`wlKMQ3w~9qAn| zx5T3a#oKlXH=o`v@EnpKGI%bqes!HMU0s)BDj~GjZaH*WX;)-dXqWQ#*Z7<(M)!^Gsy7co;jB}&fLH}**vw=XYp-AhPaWmX^TGQ^-CFZ7c1>h?bC)q3uEpSwuBsN*^7bFIQ>Lxw}ML+L}@h2?4k(BGE4R^o!>f{d1I z(C<}-RkBr6R$9=@Kly+5LK}NHp{%LRTw($O{4U(=JirHY9DvVqFDstCeRl68UoeG@F zg$65?4@{Gbm`nZCz2v-9G)(y9oFIk~lO;_Bl14H{tp?ag3WE+^jwkuLOCelJNkvSb zeIM`TFqCxuemC}Yq^X$3A=AFwHupE%nEYni=G~2vjp)rpPlG0=roN^mzeT?-zeYb+ zzw(2Z`=JN4o3v}-M`kyEu#7ON-*|or#`K5Tn+fM!VD;nCS7ARhby088S*vL~Oqb%+ z-S5ieBkti_*d}UZ%$g*{KUyU)}p8F*!XOkm_51tW2#F zB^*V~05RAw>@pBF=r=4fdZIsSpkcsc5Mcn*-*a3O{UG|;(WmZR-K)CVx&g#{#COC7 z!dgG0M1M?T%)y`3zbr5ln;06_aXG?N`Nv>qWxHEXe@CBMaKaRvOaG4!(I6-pA~N0KE7>{IU;7|)KqKW z=YjbHHF{J3#;NhfyUR2eoGzMB?OiISF`E84Ygp?QveYio<$K+k(Ue)1@fGVcx<~YO ztOj%vELN;uckTHnMVom2WV>#Ey7Nb)#KF-E{yfWOdZJJ}S<4d}(_eIb zs=t`0Kh+}I>(U)p;X+f-rX;4&?$ zVOC*0PqIEv%`pcV;Sd|&dO{UCKXr1`zPilG#w_LN-|#`P=5Dj)#@p@$)N?;O`Ss>} zquAL8LidYhmBdqmqV={+XMK%Vjy^2$TpzripNvW_eOdLA@@tCbo7@``cA%<4!?LD| zli{K?9a;xt`$>mUd#I*`rcGARvsIv&Yq;;HO4Fvc3f$hrkLnhR+;E&ybypYH#HU5; z28mYRE#sa>|2c^u*u=-F=fe+X4Sm?Mv;g>B^lxN0e}lOQ(^#*WO%}}2R4%7JW;K^t$TX;Wts<0 z>zNG;uwepAdve6B&Xow>n%_{92Fd6RD0P$pVju%hzpL(>%FW76nd{;s`=~SB??7!sfM?iOphw`R^)+i|n;7fn z)vVR6*6Abnlb5^2E5{h%6L6_t>CQ>S96X4{_=0g-P)A_(iT|KcQN7JvddsHKzF`1L zb^*Kl{2OHvEe(4*JEbgnK!!oT@wc+~;4#efK_>#5H2P-rWeD{JW4U<}62>bH!R_Q^ z{I^_*|NEtF{3c0JLYEToCD9#;-Iw#LqoeiqL}qHlN5(BpO~%ztCyfk5mL$Qm`v9Vd zxtl594B=}8lqYVbuVSsCLG*y|`2rDfgdGtX;S(|8kD2gCL`0HD^6v^cDv$KvXQHFi zFT-+@=ZJ{ph*T8iU-%HiXDEY>L0)Tkp;qU)WZhP%{npHQ7X=H8n7YuB)&(lgD^I_0 zEprNly=f6nY^h02{jO$vxwGLFOLTN)Ow@F-A9C^|LI-4h2<|uOv-cL}K|XCB>i>0^ z8twfDg>DQ*W_t|C0Q5vNS+}; zfB*jUhquU3$8`UB3~eP6G8A;ptg`4oJcBUCmGM=n|I^Ut*W{R}2$RSU?fmqgCn0B+ zi~NtPAR@84D`y6YzRZyM?myoY64*xipQrdffhYtkD5>xG;6K@lh_KG%%m3dHgiZVZ zhWP7d|F4Ef=Gd;qA29CH;`axVniSZN58`^~=kv0DRcTqzs?0DkoTnAJbSm{u31N_P zuKGKHfOY8Z3?7HGFd5qgyRjHIL^bWO*Ylvu12dcV|8X z#XP(YV-MR5+~x)(YXzRvU!>e)A}0G35)5K?B+U$#8>PB*>C)RN8Q;TE)Yc5QmzUS# z-tn?!N}%AL?u@);#MRS$bEStq4enu!Ym{*j;l|dLmut~Q9;paKke~a$wgw4!4+Iuq zi&K`xPb3f>UQY$`hsc?IClfNX?%@-gyGuhuL!V>QLbX?Zl?#26KQ5_PYJ9sxL7`Jg z+~D?JXFXh%Pu$st$%#;9!6k(8(>cE#UbD|0a7eh$=Lw65L_#@t7ih8I zT`c=7r%Ut61K;^?xjjbw%`d($R7jshb>GeG`>7$;qVuOednw+CXnxBdnPiQ)+>1NSAVa)J9d7E98=XIcMBnqo+7?B(M@nL`!H?8Ho$ zre_R%#%sJY{UMB|u5Fmb%5D+2egXC-ENPY1rl{}wN-F2IOTFO0&8b}n$jPh7&ZR`w zpOSm(%OH?sc1lBE{_2`Zv-jrl{@`G%L8Ez?2V1?zx(R5GucPA*GH{@DNYdlewHyoZ zUQ+Ij!$Sz)_JP7I!>nn$c|nsmzZj`8@84}!>gA|^^D8VZ96wW9+-u=Ew`M2hCc{B=!8x7%LZFn z4(*pDNvVRyFCF2$BuU)cxpSBl%nC5j1QP_oq_*L!0eCf=g*p5}+%mt+D?Zbjr#b#O za;71v<<9Vl9|N05)B+9jX)z~_6ez%9nAE;9s(ANkNvp>tU}Jc<9FIa)IZcTk!&lqB zFkbTOe5jcjjbOGJyrgti+w(q>b1mJ>06XJ1%kiZpMu#O?zJHjH$Lb**t*wEvbM+99 z&(k|A~lw8zt;HVoE)~}HE!8;W2x~PcKq^ZOK1w;P-zEfYx3X|s%}~! zUFZ6}|>e_G-?7f;Z>PX*JXg|e|@xD|n1=0LnPq!}@7yV|SI%cani z_3c}u@D|U=rLORLK4crr(SMN=J&N3lv-9#<_+9Hy+}YXQd= zzDWN$gt0UR4Z%ePEO_L41x+Jo>RpU;lV-}?7=K+jH~$0UdCP$hVXG*_R6W#4By?7A zX7yJN=c+B2(bf*$q0F~AUs25>LdjfBHzw&l$79j{T5hWCIO6 z#QS*az(L2W0QA)EAP_I&yIb^m1+&+m+~{GwNrMtvSz^Z?46g6az29w~N<8G;-?e=e$d*rCu4wiNj%jGwzkPwji3>0oOXKPqOwlrOHA`*#nl` z_^xoJN?w`bj>B^NuGW(Xr%^asC`AC)y=&6n5{CC_<-iNYDc;f4kSj*8M@CC)x)cV! ztL+`<_a_PYTfJ+k)Vm~Vwx^u|W{4tLo~cw#TKmYhJ!ON{{_M?`dSx`9+JGv;7>t$5 zVjoHQkgHm_FZt0FN=c@?n%=`wKX$dwn$rjFx8Q@e8aD7Y(-5|IQO<;Ix|L};GMUZ| zZeI2O7<~Sy_I0|o^>r`Rsg)p}Dde{K>ZdO`FgPX^Is}+S{Yep-MLJVYuGSOMLQmsh z;3}wXKtc%vR*LvUUX~8o8ZV}D(Tyq#@ zzv_POw3p(RN11?}6lh9^x-(}NP^8nE84kMTiMb9mQ_wm0^s#(w(b#u~F{zPKLnCmMj^2Rgn%Jy+F?_6h0{;5D zJz|BnT59b0;J^ban~>etC&e^%V10nrvP7r-b@lZJ57SqFfA>KD=`&icJA}-wd)Q12 zacUm7J*F&%;4#z!KLZb^IoC%9*%ujsJ|8dys72{(!0uLFTAGZC?f-7Ty?^Wjqo6k= zqCZDp4>id8hlacfH3Gko#ljc4RSN^cK2HXDcy3;{!+Uz`pltv~oBL`uTD|18Z9$o1 zdAZTnP<%54uYA1A3Vd&?F>;*k!nB6pi6Qt|Sd&b@e4biVucZC*uw7`JaXtYnY_8%L*aKe%?Fm8n=0+sZ zVNSCxjhgfL(0cqk>lSQSO^vWr60dQCoHKr&vc_~9*~8f8*{y;lVna^mxf&gv)7N%S zh6=hzkdp=AV8(WCDsA8qhi2kTvvZ@JXA*mmPpPrT42|afjB-qWYSx0-tZS?E4wQRk za?`~#C!8hwh-@}MZS%kFjI!wbmQ2{M8KHVFRdZiIIYDPN9p+*&smmZR5-gdH zrZ@Q`Rn|JTK7aTL)H-#aJ1tXSZl>Je1iViP5x^e<=Mp!a8w`T57P(h8$ctC@|1{d3=~;+T%P*e|ZiBq02fUOr?dCgU>?OguSxkMfj~-s3k4x!* z3Ur9g9>_a0aib4yG8$}G!b-M+jHQ@y#pU>zEVnME-MZp(_ne>}AnrB!o)xOZ;5anv z&|CV~)KBlF?}syi<>2CX2UTICkvoc)bUD-5zv zz8fAKcX%TvImae6Uhx=vFB~le@@E|?)@}MgZEl_u{p1G^nO1+BucAtD!L5gFXsg;3 zr2j^dO;#+5mnl!5XF7aZ_!zL|r0?RA>p0$UUu^Q&;UI*YZd}0I>7liJz!_tb%ezEY z@~}l& z4(?54Jl7rl?*uwi9}vpK@-X@+mI;a)G$&;KAa>pp8as*N&k8vghF(;*cY$Z`r|Y8%$igGS ztza{hTJ}t$9>^5p39o@(fz<-BH-B01U+Jos@DMdF3kpJXm{kbm15SjvYd<>LKc%2l zN=dNSO06mB)d#EaSeleRsVkOh@hi&v+mQw_uetCLp)Z* z7j8X7kWWrZamtA6drDdCG~A1DAgqiFN>5JoMWmby~0ps znUQ_Ule>}vcPZHs?+sjBQc|c{y3{k#@oP_eW2Fvzf6rI^YA$-Yp{K}Q6O@Tx2^@gt z%KAjW0TnO*{4(VNyDq_2rW+amF#MbAPu;ww9n%Rs5V7Qpk2YelLN#Y}vAS@ea5WlD zz|ExXl04qVz_JEu{)t^+vsHx{M~n@ux1>1J3O{){`;*>lA(E>7j=iiEueK4ss9At= z=q}R)DqE(%aR%Yh3wN+t)4<9z{k)`x7HGEY+^O#u0UADNM#|*DXcPo1cy33@ zK2*MpYw2iAuY_~#Wjin$-G*IQsuCgz8RoBW+tZzBbcmY{nkuEcM5a`4ga0Oa+_|Ks zq7axZEIb0WV$2m7ki}QkczLRV@xjKv$DD#L{k)Cnw%n&4Ivr_N*Dg=B1*SUA)_Hgy ziy4Dzt8@H=Fcpr|HBW+&PzWLbH!+awienqgq_QcKR3LEdM@=#<@@G+&H%ZKQluQP)BXGGa%OZLiTiSXpmRi zkwZ&&K%HF^5SjEzsG9<1fleGM6}zY#=v?V3JIq&!RawQBvEQsPcvlyLuTWr^5mnX= zvQBWsf}6AFzbTHnP(#TN5Uu2g2qBt($lyZF(Yu@}k5#nU=8n)t>9Q$4B`xbf3VQc9 z9F{0fS>U3q4limP%>~}+V;qrEN3%Q9Q>O?tc#Yl5oGGZ!!<03npqFUI%8y1Ke{Rw0 z*kG>o2#m;C8fdX7@)0IB>zA?-he{nijkx5{J~uY?nbXIymBwZ)mcM#Dla#hzZ8N#A zYF?(=5*ZkgrSi^%ak+fh`Qdm#AER(jNxcZP3o^H5+(E`1Fm!A_ux8(&bfa*dhGbZC z{woiEGIhU<8Vb&Oe=~imJtnj6XP=lckC-XmTOv28q>oa1&&tf#OxNu+eXev}{HmGp=YX>mMDbQKx zDC)18_&}~p7D!6=NxNMSnQR66C9xo~k9@0!c=vXk=5+@7D-}?TG@TR>Z02(EIs1A2 zg9lNwQpy(rMqP6ivVV}fO|b@`rA9B!hqt7;zZjex^rHuP)?XL#5yfjfZd>+d~eTaVHM-na#lp-^0LV6^};$$xhDzL$sbzCp#|-0D0s$e*0wEE zoxQi~0Iqz=&<3hI+W&)fr1=HGjA{&hoj_L$_sCXo-@Y2S*q;IE6)5VkQxQ2)5?XLN ziiR?0I}pX?#7*Uh!{7f=+w|P2E7*5}gSw(j?~K!g;Y^gFHsdYFeLy6rj2NDA$L-$A z^o@JoCT4rfV|GZVwx*S5k?g9fw$4w#m&&g)-}y9^;}8ZT3(NHw-eDvr(~Og|(8+h4 zY|%&&Fp+K|ro7CS*fa^v9$5FBybnuKEzBAa8GM^Ex9&=UBOqqJ`siS5EaPP;A50Z} z6d;S%dOrBaEpGYF))YwF(5C;MWDL#Cv0XrGE7qdx^Kj_m5P5SPZ-` z>+#Wf(r>?ev0?wmwM%S)%v+X>8fJJ)@^to)``0cAqqEqD>Mj+@Oxac|kdv<-3;3uKSjMzTrbpM%4>3b*! z-<1JTwVA3fp3*SR>T+kd3mw%YL-`?on>yWppd!?+1`a?NXPaTJPTlH3zZ3KEQ09Kr zcSA#EQh4fH9e!@H0XIBbqxB7BQ~RI3QjcHmyX7LgbDN6B9MK7jq=b+#v$gH%Z5fz* zNF8s42!xO?iacD{@wr<_j5DqtFzp)~80{F~%O8K3*^O&Uh`qGozl^vXA){_#a0yLLf?t4B&L0^Y39UI zKn-h-3P5@WFLj6t-Gy3;tc>PT7dtv=33$n+P9pA)Kucg(@u3xb^u%dWc9VV;=zU953HVOttGAly~V%k{sG8w~+- zlAW~>hLZYM#C<{aL~UNL9heTPt?PO$w4@R67tTxseG}1*W;f<3*5glVtTyU?@^piq z7|^mx+_;_%iym8IcQJ3S)E*C@G3|9^PBa0}eA2pqT!8!dMcCMoBM9qK8TI`~m~p@v z=({IG3}e|8qb)bBZTTgd+4m%Gmt4@TkAHBoVXP#jYzoQTH(virban|8mon9hZ&?R| zBv}7Uc#f4U88zS6FA*N-xz>%#gakpy>os>uNau{9z+4(0Ojc$rgaB|3NC?m~A(i`&(FWZV|l4@!n0vOkdD!k_Ga`!M_@r(bC?e!q$F$L8Ko4H%GvuP%IElQJxK?+l$Du3<%huUP(?i6{o>u{C+jEBjJ{2b4zr~M?DRLG=vYT=?Qnzf-oR!?1wKh)*7OI$wq_{UQ z{XhFU{(`-qPuk0f35O}rL363|1DqSmW)Ff!;;J7|aQsbK!gPcM67iHJ{0%39DL}@m(RmnEGm%fhEPL>jlP(1icVHGCC?ARgZfN@3IU~ zR&B~zPs*{O{9&}}mSdcO4Y8)`ujj>8iphUAZ`S8J=Tk|y&0s|D7t=4Pq=fvfYyVDN z<;X8AQyw^h+~ouPUB3B)Ott624U~ci%q%=cK(xfQtq9||qJp?Cp&)f&%KUyTlcf-O z(BLUE!EMx5r)Lj)Q;0L~TLcK)+2YU=)>uJRIG&?wP2Kr%q3vG&gyZw&pHs4P#t~w7 z&#d-Lr>L26{YGsmBIQyZQ z|1-ybSx1|SB;^}&$eDb4lFD`1XtTy*#!zKT z%^?^9yo}qfxY@E&?Ef~Z>n@3M z)sYNy1Fb_{(9_T(p;`toe-s!^D7WDT$2s9q_CHuhnq-Cl+AxVgx`xRH?Eg0t60(ta z?Gbj;pkglH>!|i-9K|kQ6y&gz*beQMS|jtD&ui6|0pi`@AV$=)n48m`Py2L|)+ z$x%x?ZkNt*<7F29I^C_!N24yCwLCYSf2v^kQ@>XT16MGg!|EKfz-@h2)+mRu-}0J^*OvI*UPvS8Nj|bU)IqPFy=il!LTdfOB1B$ zOZllvs8dhr0#%sa(Wo6oOIAjUGoSnB=-qkut0My4O9;yy2kxe^ge~}?Xs@fcej=rC z6h1kvMrMSQ5-9=qteJA9GQh0(1WZ^H#>|^0(=;%HOq-3HoQ_ds~=wB z7rw<5d06-Uwf{pAlE#x0O+K72ak^Vz=stukHHVSH5 zOQQCKV5qM~!PRPB>|hkzIb0s*c*vWv;_au~XZUX`LVD?3?Nm7Rv>kRRpKR1#(?U$V zQejh7n=5wv%!aQkN6>?l^qBce6Z}fMwTg0j6Ky2UX7fLtPPgA@bgkMKY0&z+4TKGP z(!Mfmh>?wZ!${64%zcXwHll_y-^iMkLM0MhtO~Sv}*M@r-nQw`+>-}<%2xlI{y@MMZ!F_6eI2aj%emi zZhVrT?1=3Zv9kNOS5D-{8Sh;V4iGOvriven6%w7hE+0bnpH8 zb`KS6UeNxfGfRYS7Qp4p`Txd!M;0#f6`-gzd2PK)|DStwk)(7YM-4Hi(8b)9)Gi(e z&N`6=&j$Z(<%pH-r9e&ApyF;oVkaO56s~c$BjQe{Y6Xi7&)<{Acc&C?F?68pZvqeI zh)?YV`a8e)$kQ?8?o`$>6SS^Bg}77?V}coMssWVbo$QOFQm-IqJ42 z2Wg(NlRShJ`B@X(m+@Ao8f=K9Wr;XVd=IS%Lw=4x9j)phhn`W_}NmvwMAcYRDFts zk`5jFlq;rC6lyx`zzi9{X@;2LBipJ|hiNY$bSdQut1`U#o44t&-xzy~dnk{MbbrK3 zSkYOfoxh+@*|e4SwqCdTC^OJ7k^4cVlVKEC!T9owm%e1sBuA}Hcs4I?o)^;4GP1Z@ z9C{iqUkT2yDpLh#JRfCPzkS8MS39K#CoHSz_&0UMw#Uy#BG<%uX=uC_507J?t#52y zh0JjeKDC@)kK(g%b=7Wa@-HkbRGIcOG2_d5`czWp-ho$bzYV&(jA3!7LGk;-;H#Er zAuYT?E*c|6&46(cf&*4Ox^1F$sOIy^>~0iLYKAi^*X3-(eF?k6gG{SZK3~`S2%_xO zlx{b3p0k|v)o^|b9dxHpgQm&@#@2|fGvd5)A=*zGje!o04c^CMN!?dgQhQn?std;V z!#bw<>jMtaAn0Nk_3{oktk((Is50g<&YM|}xM_l;ftm0HHcdAp{hvy+#RVLp4r9ww z#Vp8`?05WY86bKN2y`LtJF^O0MFnELQ;%&>lVsaU$QYemC>vX(?W1}NRc65J%gdg6 z_{x1kADDC;i&4zkCdnC4UuqV{Ym%Z$R;JId(_x@Y>^)*Rfp|;>z0XM-tKB+mq)r6kJJj=4L z!$e?pAk})%R{WbkC)v9)wX2FhSPr5g54+ABNZp@M%Z71RD*q{xIl*cj#~lgpb97AKXw>v-4+J=J{1f|*P zc};7~EF1kg%;rKuGg<+2$2W(oq-6Pybv9z~(X(%EOoel1_PtBS^osD1uX*dKbXPG~ zgJ2Cm*bh_#6=FtgZvh$sh>=AXaD$N~+q&D&!;u`ugx{E8{V~8ucoy70a?`q`TINzamyKCF_bwPL1 zJ?q*@j#FB59T@Iv6%IGXX*On=4>Pj-8EX&P?QxHY2=DwKv4tHcrf9?1oDu@JLuTb1UdnYw(Nl_V3oy}zkRuJb~` zo)qAQ&VV<>8jfAu`J5HqOwm`;3}4(U)Hj)lBoSvqPx_Y4$lYoHJQMkJUSj7*F8&%C zl8lqtsi_|F?(4}F1J&)x7Rv|4S+`LNKXoR?=gi9TbqFcE(*QcAk)uG!iB(HH=;l_U znt;SyBUc0dTWyu!ieS1_b8P`@ZRVwe17)e{+I0HJ0{!iRva(MFRxdF>1`}kW>%uwx zO?4NWb?BxN<8!zvvh@1CR_J_<;KjYeb5(>A(l_yGG|%6vqCSsi^*Xs%GE@KLEU#J- zcUUW}L@`fEu>pyy7yTlHkf;v!T)%ph0cIE@e@zAV8#8i@;C6~ zuG>_ z3oqJ&v3X_1tjzHh!`GH}QYLJi9%d)xY_Yi=E0u9$aKW}9D+;$@Zw&n9KIhlTXXtT%PW%2_8u98cSej@he+d=B^wbk zV##%)pZ*@wgj8LtaFn9{gDw@4qEYk_t}Y{loR5_CUijm0gM;bO)^8f=7Tp>-Z(fe0 zKsibJ51yCUF99LDs)+>2)}y;q5bnVT*$Q&jU)UE7_}JkNoktci_E_zpMcUa-ue|?d zRZBOO5+Bk>HF22jeTw7=C4D4^g#j39LZ5TuWEpejUCQLKEaTwAK3MOEy1Kz4LO#8< zn$z0lJH3UA7<)&4!Z-XEp%IZJ^F7pq!fV!-c2O|N1|D5KkTh}g)%jr-66MU7<1pZB z!*|`16&~p0mp&umpmX0P6yo*kuD&LVqr5@>#~C6~U@roCAEq0=4D!!Grgp?RFyNU8 zd~W)!A_HabBA_)Qs+J>I?mG2Nu?pZ?6=;swpv5*ssNtDZ{}gCG&l))(koD_0t8dZ* zWf8S%sCAI8&gJ3n1&pnX$+_RN=D4^7xRAjpv1m?E2I>w}BOIFmyuFYf_kk=>`5 z({g}mYAU=bxhV?Lml~*r2i9k{2i^BpKa+K}aU>`ypC(h%_Yi*nhv_72e61D&z>cR3 z;Eqla4g-u4eH9)e^`kr;U29_$574YI#5r$d7I#q^w6mtjJK@8p}{SF5ZOQ@|fI5`a#zjM0}A$MCF_eQs)Vb+pRvr z5*VaT1V#2OFjDPNY4;rW!V~$@Zua+vd@zd%-1$6k0Ax&N8Nsd#jT= zw{knuLh`0q9WjC2Xq^TS62I0xG&;(GIwa_jw>_-ue{#gh-bNT&h+Vs=5)>hktlYu& z7ZQC)b_65k z-74XN>@E9<=7Vk;d5sfPr5piu<_(d@qJg8dEr)+BPkPkNWT+6V9~m#7Cp8XhPRZ?LALGTT6L60M1vntW zdIw))L=c3M<{crz#xFQtAzWcA98r*q5H7eNp@LPlyOE-oN_plF|)Cr4|`H5njduzF^+CIkL!MZu(S6j=E&g(#OLvVJnV zJ!r=FStLL0JMPDPxQ&zW3PMZ}+?e~?c=k9jH*}s_PNBd9 zwOfEFfH2^e*Qyd{rPPY$PW=p6U$TQOvh0$|jKnjlnv)-rYv<%d-8dLHL|<>N`WOAD z@e)(RX=xoKa&GU8u0K9z6*H*fJ@4*0-4|Vw}TPO0voc*iBb>_LQCa(L$LbNw=4CH2^HdqM%IazhvjcG)jor(-D;!v z?F6_=6et|$lm831FKgy_2I|`jF8WA2s}5>>y+_T%NDOM{!!ZPk9N!hzsy^};8agyI zPNWB6Xtmdd}PJlY(kUliZ~w$CpqenRs3!oewWG4@k=V9m2%pB zsrIu~zQO&!tl@oQFvY>l?fT(LL$1v0FxPY>ZRX#GWhxTq)^8fwSi1JKnq4yCq7I|{ z+9SkEcp*|wiMv!C%_oB!)h#@>9zNh)pzJjJ{VvmSCshx_>l<@^_4f6p}^V_#?flam-S_ToT0>a(!Szp_HuLWOo!ZdrqncX@}An3xUq46R)7`Aye z?z3NYoOFh7@*te|wh`oO+aaOd_n)t+w|z+-P8}EBZ-=NLhgKYkKH;*TosqB^Z1To?6q<~0iv%2E_)TLVY+HXXo-2p9q9ABsMAFrhuvJ#w9DE47%XgGloRaT zpmqvMzdp$;D2_DoNA)(XnkK4qaoFTG8e$CT&zeNete^EZSn8f&0Ta4Ix5*`~9T#~+ zkoJr)`Cc3!MtJ5J<#flE$_A1+L)0*R%*!x6+gRXGv5+QT(C3aB8aEo$XW?!3_so5Z zv<3`dZ3p!4QyX4nAQT9QE6$hzWnc0h%vK%#6JUbdBQN^gVJpqdxSl&FfKrJ0FP9{g zc||g5!=Xo(tXuY?U_|=W-hJ!4bwc;C<(&nGg=u7~$F-Niiax--qB=FYv z=>&%loT|4_gTJ3(t8V=3F%S932yDNAE;OUm+bzkk+gFJ;mpyfX>|WE*{J`X>nz;!>p@ zA!QgfS9O$c7~3s)R9bDqGi*T|M2&-$TqF>BIe0kHX%@Jre5#K6jEtWunWrgPuKhY) z?wdGp6BeC`l5_eLO4PpYZp+c$ZF5|YdiBi3NT0i(c@ zKB_#Z>$hdfjupG`{=qzRQYU-7r`}UK|L6V1hsiCxxY3MG$Xj0A^T2e1;7T|{COC{# zdJr<=u=m}H+RB15TT)!W8g|j!LuzmTrsUXPnbVj0=p#q!?XnuR7PdewM@S3pJ)^S} zt$me*Ma(5RyBXDksGoW8P0q|S7sc+F1eC_T7aoDx{cbbmZ$1j}MIEOWmgei6pedJs zLIo0h35xN|AyzRXlevpG-*p^?Xtvy|%g%>F7MhB)Z{a}@FoC!?$gzRlv!_uxnU>xC zGF4m0sg`5En0BYB%L(t<7)zaYe9SA7^ETeSMr9k(Z1`z4I=+C`Mky5IMC%z@-@kOJ zzTcZvNH}V)wZ6>8sOfm~vjIh)BAxHf+38 zD^wvp^ha%_uBb5*D7@RSZleL|=0Aqn@lRZ-EIIiMYrRz}Wc1^+Rrqdd<(a-x`cF>j zI+Ihjyiw=eL8)KDp>`Vo+mQP=gn-{0mp|?aj{|^hUaKbgQ{P$0xD?tE^#mn*)k@@W zI6E|P+Q9&0w+`kgJ2;}eD}MNt57#Nbqg2!hVA|p(ge;dG^mvybpmHsnh3n)UMbUZJqN&tB@ma49zEd%2<_3kW^twd(xYh7 zZ<3-Ol+%fkQzEZ9$y}=;8Etqt&*@F=``2w9?cZx;*2hr6*~NXHhMuHKv1E_Ywc7Dz z{_9JqsRi{A_ifZi2RSrakqK`?T=loYLtc_^+x46nO34Z z#MH;uohfBg$D0OFW*a8L%khdDDwn&4ZtQL>Pc#q)Z)~h|eJ=Y?8xgsXq0Q-^%`aGM({3$`e^L})cBMYYMe+4fS6Q3A^*Y&NfnC#Xw?xhDt(!E*u(?a0BBBi+(afSTHl zH0sB2e?X(vH5Tj7Si%{7@y;77L6CQ`P;05nVNPJb4hd^S^YV~de`+oORbYHD2F;4b z!C)?Qo33IvFG>tE;yfHhf@N)_Iu=|P3Idb}-I%5bn1pt*JjS;pQG$b~2qmuq*Hm!_ zsy4ABQrm2n7F~8!5E*aZq*8RMo>yU=`=@krW2O{)Og}Uu2lg- z66X$T)GSGwB|vf)$8p|Npm)NrK_dJaLG#t zM8J+R{YD5KdC8tJFcjt+#$d|;woxKL@WkIlU~`-2=2MC0A}9Gg3Unr4=S&a*XV50= z;@@rtS@-8TJr@yf68<&LEzQeIe4)>!mwF%GJjZ%T^!kujQ6`up%AFB@>#(p$zn2dF-ov`$mhFk{S3x8F5KVtkvQwA$7P3&LQ^V zyGW}}ChSL7C-Z{p1f`veJv<7o^LJ1%vZQ3n;$0Rzy_3G3z^Tb%opo^gORZ;T{%6PS z(PC5`SH)k2ERsk+09XjZ~;mqUo?^y)AGAF!rPB>7l za0pHmL^=6-?>w{D#%t7t{ykB>MQHqIAlrT-oPXcxz$Fl~UGyDEcn9FL?monl?S`Jz zdFgs{!+3p={nQ&%uE5!&Y0ofd#X@KJ%BKOmk$AR=z<5-6yC&gx=T8wySuyl@*J$+|9!`J#`A?w_j1nJXRo#9 znscswuXfqPbafMoG6sx}4`n5_Sg&pugQfKX<-g%CMKJ<0HxChS6}+Dm7MFWqaaNyrJ>$;mwm|sv}rBgS=j@5=}9J! z?AuyzWTp{C{=PP>7SwW7D;5+if8g1#eV;i++#g>PQCnxvHc?+lMsyM;|SF}TnGDpnqrr{J$_h6!z>1bh3sZ%Q2Q z51RZ1nX!iu6&ovwz09C|ycH1Y5vZm)0B|aac1A;0#mxvY&Ja!bluFtRj$zjR^l5c?0Ok}4F# z9QgL-wUG?TKme5@>nTzH%At3auDk98z4w0b{LLXfuFL8ZS9$}%^ygmzORZS=uV(4jpYQb0HW9|hlcrF* zeaEQ80J|XquwXr2W+7IxJm;516gns;w<(P)_M^DUM!=H+uPbW)YJU^ zlJaIGLW-q}O`~HAQo(rKr9-dR)L!6rN-j=j#YY@{Zl?NJzM`>#b|oWayC*jT)W_IJ zdRe-)2uzIi7@vb~VvIRq3ISk1gh##t7~h^f`CnkLbn2Go@PU2b0n#+%^pTV6WR-H9 zm$q0V!WLt#n`VXo5gQ17dUAE|3|Y-m{Tc+Cf3p+-S9tjHP%$FTjGMTiBV$D4?l?{I zu>pA`dnP?nBV8Hd78h5rK|v3JsTgdmWI&RpLcA!lsjrHBYG!T8#VEptKkQyB?$Nrt zvtyS727`&x2D4v2vF3Ck5EpiR)E7yCnOSGz!i|k{cdBQ|p?$fNu>RJ^x95mYqVQBt zRX_gM&*tm2^sLyNOAF~PcO-X!p&@HcpD8bLQhDyYz@3!Lvw&pXOiJo3;tTXa<1ZEL z^_qbRrXBMrgv~AkI@GDnkGK~U|IrGGzS2HHt`x*QXS?sc#ZRGIe%q6Q;0HLeonQJM zWdk!Q5Q%Y(KzlKg44}5DG%VdTfbAlJlsI=1Bz~1-S=f;y%)enmg;J(bu>af$wh{Uk zBQY*q+V&^30mH?GKjo;no{zbs`hSNv2-TvfwJg9~UWC%4o8kCqVBt?y(|ZGycBRIh z16!DG*C%>F@Z1EbC!uwe&xBLD6}IU>vU^;v$7CTB=zA%lCy3n_3RYo~m0(LoRrwz~ zlY$>idWmLh*hj390Y^mkFhMs3i=#s+v#xrx?J-Ml`657-<~hIWSpl~&LDdHcr9XCz z4Q_6KjlM{!6~5z05mwR?`TPv0f)lS)R6qZ_^sneaQNwB0Wxd)rT6d){$(TxU-Z{Q( zCdPg2Qk_y4;D!_hnt0y$u3E4l_?lPd;{ZRFV*GyGlkNFgQG@H<%@Ni;NGJTbqdnPX5;;k{kWi{GYd`pHGER=+F)V##Ptzh}b zMi&1{kNv`Aaf03o-C0s!SHpa8JizPbHh-+G-}?_|3tS@=23lUi;@ATDF?hDe0rQGk z$i4&*l}D`nS?K25sP_eeK>KZYobeI1^%X`U4A2U{_w6PB(?x$_^8*;OkxO%#5h8!` zwwK(HpqU(yYCh1v`-vh1@VYj(HwmoS#@e&O7Q~2;& zJdli;2>wcU*OBj)CHC(Pq)fYHab`ySPNse3pOp41|0OCx_pP8aAW0~cUs*&N9(TD= zrD@!0%b!ag_PwU*TZFlrIE>fK%_VzmBS=dGy`s=Ge{T&TW#p?0O8Fcnx;DU!u~SoC z{^wl)`!!dkCHvUD?`JhAenoQoeR;wU8=`}!mQq|yKIIcDy6%9u&k6xp7ohA*L3DHO zC#(;*KATd~$Jn~w>|<*FwKN!!F=tV^uq5qmRn=L{X*{sfB?+S)+y);+wH^T^DiF3S zaDbv*qviNa*A6j(=*<^Y_&MSJl}S{{l2-qUZg%C>WiD0IsRYi$=>5p8l+2q`Sh=M1 zCBo+4VsTE;V<3>+0kr&-0#l>cz5VLBO`HNV7ym^-(VrUOSBBMXT&@q+a+nM@#KxbJ zHxz|AOHXlFX8M&s*O6=H8DVbYUDT#2;CeM$4%lyY9-&DsJ!DM)?L-sj!5Fw7RiW~$ zwlMbYW-i)94$5v(IgH;#HA z*p+r7(E@IhG$gEC{?i5IN_{D6e2>0r>dwc+S=8N7HnntMYr zP(oIeH#uzQ`jaITEsO>$Rt{atOM8E-3;_`_`;Ri|wH6W>P7UqoNbEeZO;;wH$$APb zEVV4)Zn87rCjr+Q*UD@QT-EKW5gZUiqV6X7l9b|Ao$wl|H|?MDLCV z>5`vr{EV$XybWRFWDi#tTQ1q3b?Mo#;uQz^v^lBr3PT=+OZ5iZf5K%CpxFvG+(Mg7 zp}@l0pYnpal~D@;Za)&q)q-EnR6HMJM5`A8gGVIQzkSj=wjl- z=wHoK40^!NAgZo!70V_<9O1U&S=7Yhq*7xk_7U=**w4}WHT^-2Hg`B(ulsAhpy~gw zm;Ylk{#r#;WTDR78e^No-=b`iuv&J68kM<^;|;Jt_63{iCdvmE?h)7i*FXYyli_wj zRlC#MXE^-PEpzqFSA??oO!aRbdjGrmKt1G)4YtkiXE z`7_hI!!yg*7LUekpoJYAi*u|ONH8gSD}6#JFvqY8e{o7uRK1OeuyP^Z>R7YOb@YqJ z1yUv<-BQ7+RF#cHs_s=+q%qQ;7)XWj>&y|Gu5gBtt5%;v9wAlL!Nea4 zrin7tP6VTwZ76+Ek&(c_4bo@oFWhwp*W@wk(8nYfQd0hRJ^W_dvsex~3+_f+6_1uI$UR8II^m6XuJ@6J07hR;`-XP*bn97dZ#8U|U>F7o1YE zC0YYRBl>NY4e+af8H)W(RxtcZ5!!B?2S-K`xY8*w)#V`RWd8vYL$A}IS8>tAx*X$U z>5*JmZPs**i(`WEPT zo&fy_+R5%{EdaC=`TR;vb~x3Gopc;-Zx^W~8|gB1Ee>Uk4IW|bBSax(Ci81S&3JS& zH!`6~+{3@Kb3;CqM~eJ2o(Y^E127biwv%$vU6a8~;rm}x1QWm%fn)WDiRni8BMP*Pze2+#!}SH)THi)pI(Up65Wb%0la zf}lTnr?n~ngcl{x-Y*vBmYeS54L999wn4I2Xd_=d1}VK-xh2= z7#Kjm(&;fSKM|JQAs7q4ZocpG0f9Y_N6Y@G@)qi8?%WRJ_9l`l=nn*ocj>NPIAx1pIw)Y|M4OyvP6I%6vLrNu^ zLT17y*>^!d*ltRsRAWE}Ml`+wyfw)sl$U1W^(tYyedlLQ%%t}=?NndbuPDL*FM_o% zJ!I6gLCwqU@2q!B+=MVO-9@? zGr4WVC4R7FrU0g(qB?`xz8WMrULEUjANk91uSRW#HBZyXEzcX|=>Ni3z#o|ocHj8P zQkgrgK0_9{z$P!^krQt9drPNd2P*|ZO*~?cR>4WY)ZZ3{)+d|sP;JZ8cOY+(cGv2C zce45NVnOP7$_XLcHFTR9)$ctWb1EKj&&pzoJ_jR5f8Y_wB-1bxUyGZ23c%mhaC}YxbBex= z2H>xjeju>Ujnk%E#ZD%`(53j$-y?uP4{m_F(iL(eVyeRVwVHnU7)S%>(p_!*Bj+;B zY|WK&S^{Mk1%R3r_dnP8?#l}h82BwN=EoKx16R%rqnW^B18A|7sfcLLJ<_qNpb}`9 zHhJG?0}M&QSxfos?{CYP_#JunwL+QX>B$7kxrU5!# zIXxse>uEMsl}O*2sd#-O6S~AhXAN9MxQSZdRX|0XcG*?#7HP?p8Q>H3(ICd*c=G!Z zc5sf^U5z#XKkGizrF0G_-9j+Jh2#wQfC=8?ufPFpf^mQ+k$!^$Ge9?*5nqFFWsuxs zqkeZ$_SBhyoA$er`^O>yJHVlq-^?^#atZ*oprD}aYOv2(c2lwQ8NbcHwxZ`*Fg;ob z=T1y}`sY*EqCROu9~LiloWF@3cwy_WmWqxg_NOrcql9!m!#>>gYQ0$56(UDRhhQ5x z+gQ!^}h!0VfND1hi!u z)c+OK@Q66nZRn#9C%eoaZBu|B8IJw7lV=P-Rew{<G_`#M zV?ACki#jsw&5wf$e{K>s!vXS=fjrD8#&?XzbRJ@;n82V;dFa`t{JiXT58=-<$%uXd z`v@I09g-cp@|`TtF7Ez3**j&D))db@+pVEvY8Kt-PhyXG8_y=IRDMwpSX>}z8ockB zI<>9KE~H`p&d>i^!F)xplAV1=0)AMZhSW z7~A6A=$uNhnB>UQ6HDe{>>GAC4BPULetS#~Tw*Qz4%|OHP?jb$aEBMM0VR@wETj=| ztN9uod!+FIXcwTN_5t12&mOP|l-MM>lH69#cI?~txHP#`@A2A7Yz$a4)GgvvxNJMt z8Xs1QHpgv)2#kM(-vMVyT>qvQgxHZjpK=dbFXSq|++_ak2Y!7(RxaRf&v|9_&5|SY z;i)urT+ELjKZdV?!9zIQtE`K=;JC^RS&LR*8%O)UPhJJHZg0?R={Ic)gxV4SmKI9s zA>WOUbAdO$ZL7*Vzwf19e#J>cIi#-+j1u@b3*Gf6cgQ#Ec)IjT0GyrotM~*VJ<2u# z{D{b&Zl<}-Img=NQnfiZf^m|EKLk1ktqpYj%NVI#ahF!>t*FU8>Xw$4ZYN(^R!7Ry zg&MX|*42N;r(0gbT&&floY;5O zFh4GK;wxxX9UPxW_tK#hut%_OPs*;?+XO3s*3e|N(fN=wXjlcK8+;mq`^1DAdOd$^jXR3M^8fE6vG^ueSiM4ZH8RcejaxyZq zTCd6IR-F6;a*#-7TbBp6)^|L%nGkF9-7jN|p*%!}5u6Nd|LOLR^zJIME4g)^I)7|O z#=#iew|&Ww-Ga;UM1K~-B^55lRj1wYa8BQv(IcR z%iu367T`{}U$<7Czaw1u-+&kDfdLBi(2k4FFTWNrC{j)JICz)`Yj!+g!Qc(}Iy7;b zY{rY*FF1S1gVl~5jFhy*P4%NwDbEI%+y9rT+iZzyVeT+-X&+$*h)1E|HT3k?fvB-vL#MA zsjb$dC80}uCMA5&hmS1Cpb<;`*~mg}QR-mI&lb}^*-EugZ~_Fa~gP+tq`!D+vp@6_eW$$hqP> zM{f#L-}?Ie#9V%y<^wXWGP#~`gd)#V%eu0Go&>z1!X!QEG&MxEzwz>$CDD5YHz!;z zngCB#cz(g!2w;r!OGLm};?rn+dMLq_bgL4UK@Z6rGD*V5fp)?sC$a|0kDF$jcgUB%FQgCB_~zdS6ib-*Wtx$9JSkQzKzh#KPb zX87K-*->yh7up+ad=%6nbvBj8%yN~xQ&O(=XDc`RE`o^^K8a9SNw`W%jM((^WZx$b z6vPds`o2dj&-%pt`cJV7ryH7Bc=>Bk$81DfIIN>`pu*X6ac+v4C#RrZ{PySZZ%fPe zsO$e&TAbsW0hr*Z?(gqEGCohoz#zLa5$jc(Fy>S@$A~ObVv|1SdG|Qilq01FOe+GY zT#&U8a5|vv(NHz9z6`oN=eFzrGO<)m00q5I^3z29P8=APR!-@#FOX?D-5$KVAQ>Er z+fb@@U-~MZDvjHhR(xVzow)G`^9Xy;&wdtY|G&ihHgWWFKAN( z2Y;tdCI1%S8x2Pz1t;xsyN2tbMl0vU{qS=>C9)8oiv?DRvR%GWzIcK1S9E=jmExRr z$crU#ACP^a_+Lt7(Z)P8>Dmi7excFPN&AL?U}mx2=K`c}K*QmzIx`$D8iK`5`BssX z7K;)Jh7?KsMJqesP*a1Xjm6?-y<0(fS$Teem(wNpyn2$`>4L$6!$9L^!=lT=>1u|+ zN=~lH{AV4I<>F4d_XYgr4S`MlV#7&Y@EsY5;Ah<-WU0%20F=-%?6nucxo9UZc=?Uj z)3CJ>I9KD~q|X_z=fP$1Cf$NV^I%%HdF?{W>yz?_#$j8ZW7S2+cSW{SO;&x$1(gc4 zGhF`1f#}50QNH3!GUQOrcqAdUkUmB0rThwxsfgZ`=Izi!B`qVLapkVvVs-1(BEynbXsF*T<6)zf9nqhFuenWrpo-0Zc$vp1zBmnOEc(713q7yVqG&a?|M&s*@j5M@+~7CO3-Hw*=71b zdSC0Ii((15e4z{65|aZ7vF)_Qt@gRCm1LNY@nVD9F-UQ10 zVjby)NWjXBrT$v6@ad9xiaF%^V}ab*&jKGIoqLHh(a{Be!S=ppkKfM?tgxq;@A}31 zNe@Gx^dc?G!^_Q_^mHeROEmNK#hwlafz{2^gTE`l1k*!_f4h4*&CbHJa?y+395*HC zcoi&hUypjGQM6(19;sTOe49?R<+6$$cM#oh6BK0cOdsxF=1hXJ`gSI0+=I0Ap@pp4 zV{%|g0}tmZA%{C)$QObasqYqcM@ z#3mICs%fkw<{1rUEH8HPMP6Xat?~leo@q^?Bg5^tT@WdE2>Fflue}4TA&i(Ff80$= zq$P3sWttzLx10_y9L7QTvpkgHd|q4)ls9VSO?_pK)al}awqHd==UWpvKla!ITHXq< zwxwk8o9+8JzRSa=Q~q^>YaF^-PW>KPm@Zq-=Og zrtC8|C^}|OX16_fYp3i{YiSJp+!Ifp`f^*C)#n8SP+&_P^+j9w=$AHGXI!0F(I_Nu zj%lJAmi-&S&(3+$SPRL|KdvwGKc<3T%Ns+AF}%}=MnYmwa<>?y9Ut#izc#;8dmol{ z;r05@$ChNsj{0qUbP1+7=1DxSQK*ehHy^f{<}>cMN$M>wwbq-?(@3OLOuDsQ`kyE8 zxnDh8gCBl$th{!4FcCCE@bEqiho$hoEbSupkxDseG9tbLH(UhPOY2bfOmk1Bdcuc-HEKFEAnUOoS{_$E)FvpUtC=9sL@-Ls4d_7f_m56%om zRzR`He5MZv6%ALbIsJ>9+z?jf=bKIqUf_nC&Xx2Fk!zU*Lv!=aZ;K@#);`C*Y}CZ8 z)n&HIzq0*lbZPk0L$H!zo8~Mxyg2G{BRpom0Zup_uYFmUbLAiHm!c;SAnjG!X*0hT zjM#jb0-{ocDi+LJ#bHK$_Jz2^eIdEB-KItY49t=X4 zpWkwMe?}!CMaO~?fZVXZRj%8oML^dB>l)7uGq% z(sV*?e(Lh(^;|L&1bTT>+h*v?!aYW|-C7 zEtJtU=-7V>l8zkj1@ioQ2NMGX?PP@{v{nG~K|pH`DfLCKABvmC^e(ql;D2UTj9Pgc&bAKIMc@CnpseY67& z3g5jZIrX1=-!ZOx`1>8Bz)?;t(6Uj%F~MjMJWXf%&=`BOiQ+Ag@#`^{-iZ13I~)&@ zj9XkKKwDMNF?n0nXu1}yh1!6KJYsXIX=I~Frqk6f?%W{S-RI7Xz0Qj5z|jbhTIV`& z9oXyIW#Daklq9Q%Q|EDSvzEQX;QIVo<|!le4 z+oT*Uyzh%SlwFGQ*PuyV=8ie1csJoM@qU+NV!zRkLggbYh~SPuhm>a5oicDP)q4S~ ze=jkW8t!O}1Y>jB=TpO#3BlvucrczGlg0vfY3n;^-A3iWFnGHii(@VP*k%*<(GhiM zZHFuX*d4VZE9B+z`58DkXo~{GgYnA9fGGd=TcfU(t(zU7FI9>g< zdXG2!UWBxkd-)rlTb=}QO9O|d{o=FQLw_J!|JqMiNEKP)e2|vHuA%k2E7pf@F-Oj2tW`>K2t_r=J zwFGpL(9Ix3!&rBjDJS&zwxJ{}lhoQfCpB(IOnlmo*y#Pq@N^zlGgOsIR07|7nkN_d z-+sPd|GUlekwh~waYL>UNAgy>a}-6YpVoZiQ_aVAs&;j}vde#WyZbA`Tj0Z>0u>HY zLNFONo6qhj>TpxSG>WJ~rYP9EMB{3y+r2DFcdm+3mNwNOz9dJwdr#SC5zaZK$43qzg%gyooz?PycLA> zv>n7+Y0hifG;iYI;LNp;Rm|3WaD;a}O3)-;lW>`MLy2cs8--f4$BuWpl;M-Jl;fjQ z-I+E|@liKou29wFwY*8ukpS6$JBJ2|zO_9yu;kdG&AOp<>&c0V*|_G6RthC}18 zYDO%lJCfg3KY(epHbS#zZ^Q`_PNEFu%g2@Qa6!HGXRF*-z70jXl`cW-_FphvYn0c>j^xh1T*>l z7qzG`A)<3eEZd_vZl3aU{e00b9l5U3oLb>RW{F|UTVV1Z^K(7jW1rNjQHi#HK^J3d<_UwtqAZ*!56}-dJ#hLXVF0wtCpbgUhtA2cxT*X;$N}iTNMX- zFKqOIy19PXDnBKyx>%iow(5xUNTK8XP_u6r7W6#!lV8FWagT*68jrX2t@c^8o!oXe zb?NFym)!!pZoW}!CZeQ9UQLW#XhV z@4zgH|J>zecglPTNJitjF+atlcq#C(JDKAUuS;Yd0hZ&9ZWs=;ly%{Rsj~mgPS|;^ zn;W&*nYH_Eb`Iu?pT`MrNS1a8z+p8h49m-ce>+L|2q7gpNHpyxySwgye27F5j<+SbbW8kLdoSS`+3x5APmBAjrGtd^a1xNCEajD8d#@uk1t{*3Gct zYRJ_{Rw@~e{oM)wt|Em#aAJqOcx29N9-|jBoFA?4O>RF9klEv%bDQafW$|f$&5?~w z85t3n{HlzGG(B^c7Y{y@n_jk6B%0h9%*D)gh#2h2vI*qEET&;W;Kb5eV_et>Uj)Ik z*LQl3a!lwrdwyby_!MYXQjU2!{RFL_cPNy^M{gq5@IaUwXS}JKSL1}CQO@Iw^o;ac z%?eYas$B%kDM@y=I=L8pmdtQAxI}n=8&wL#;yi&1$8RX&q`z-A%_nO&-3!ZYGH$Y4 zeC0g7=KN|$9uZ!{ny3P9K} z_B+FFlRj=5tJUEV^#jkT-&)CVT*~MW#B|IZrKz7f@(ec!NlhBv`F`%i=(r5yp5oFl z*fy@+#ve(lUSGB?C9fxd?`P@YW_{g3F>A;MpMbe+j!Vy)0ou43VmM(29q@d83Er>O z-F$n+F`uwVP-#$b|AB?~KFm#W5w+%ws%Kn|->UgTyB_RRbVHY(VJGu*ED`kkbORGg zT8+_h5ZUU|qWnw%i2;mRY)}sK=H=HN-@ZMqrl6U;wJwP})iB2VTaAnSaEs6}FkxX* z=J=77I%uKk$we>g<+E^RhI8B*7*Bq>C+0Jm&}}3RH90rRdS~immluuYXcUR!Jh7xr zCH;jOjM}+~O32zHK-kmFlSLIf6D3@pK9u|G$ z>>l`Zk$Y>(7MI`qu%kKh6-vpAK-Dww4}u$iu|}{t?9aFTulun@VKpeu z*d_JcOfnX>QBk4X4#PsOaFZ|i84aMC&7gXqM#EuK4FY^p_RxELRwyE~UDCjV3mCR! z)f~DUe`-9J0n!#G{%({hg=}M(Wg|vsUitFgk@W%zaz{r+VKp3Rl$!QFgLFwb(8c2& zei>^NL(N*E3v)cBIhS_d|A&b1sYhc9~4qDi9v0n+g;>%FR6SicP&Vu8veBU=En2r7T{0o;k6 z1DBWB2CY-i@ZFK-vXfO=!96`fvBYfcr$#E!^w--E4muqMTZYbto{aw5BHel0xr+)7 zEePj-fXV{yg#(4^0BhWM=ru^HO2g*P4{!f%vh8--ijJBYBo`#)z9O6kx4So7U}8(x~Dpf zjn7x6YjWj@2b@la0u`dn?;NRNR_pPKHbfoRTrwrro^GYY(tq&${tfTie6=yI8&34q zsw9$$t=^p*Idqgci(L*gS4=}{f`id?OEe5lVUAHl1n>*ns3OlX?_azpJs*}0n(8`I z#5w+X&4xD9G1hjkJ^!8YuTfPPAAQxc)6ieC*}w>r}_3W7_JT@)_963b~%-gFgA~`W5h#$zzK0 ziO6H;sHYW#I;RZ__k!r^GgTi++YLU!z1*xgP}ZwlMd!q`b{(5d&ECaF#aTx;tik*t z=q^7He9{Zaw<7I7FMwBhU6R|{E>gCX@6LQ@OHU>u`oX4T@=bx)gNdB7Yq{a&0I3|F z{Mgf{cl!Z&K8&~J`^G=H=&!3HG6 zPqKA>eTld37O}udOOqVRt@Nd1um*|7)bOV|CQ&?g{GzSP{a6*rr$R+s)RO52)WfK& zSDQs=*sv$-Vty?c7+4RePmJqpa}BkONqEV=NmyW~HBvyo9()2tO@*p& z^%kmYZMwqpt<=d>?oBgDPTO0@f-TRRmnJxm0)~5xff;w)ETHNUcVU;zf()fne~yn#;m9k8D53kCn`mQ7TGZZ*XjS zPlx`l;O+9TW!(xQnrc#E?eHL49i%{27j1~H|LDcKN%_$Ch=K%A^PP8pe`B=Ey|UZM zf$^Nv*=2r%?`|(je}TF(M0#v6q&slCfQXD?p z&O=PqKH?}S?R~RIH>wPJH*4CP-K_I2B`M~Zw$6SQ2BgtQT}dLjNO0GcP(Yoq?wOE;1IYLigLZ{aSk76( z1aVh0#)m+fkBonGIG@!9prTlyy-4|)SlXJOAobQu4k4y)gY-pWOE__8OxAOoT6fV{ zjzV8Wl01unSV3|=QA6;vqdO)~(uljZk=hJ2$Ad61Sy&F*wKL&);^s|V?_BIV;5V8b zl$nYWHY8uX{tO#@uPC4vKT5}diqzpxe#9>BV18R01)R7)BC{7-J3+y+kR^K7r)GRg zG(Sr1{R&oSp!Vz_z~XxpCauBywTJdq*5n{~IvlAuBhV{wB-3SR{z>@qQcE>WAw#&Q zy_10un8lVpX{@oCUr=+K(C0tm#)JUcwR5i$kJDCmCPHp~?juJIld1B2L#LMOaryQ^ zbj6LEN%riqib(qYov!6}5g$waR5Tatx{*DhWU#NU5x1L58AHM%{xlexBmHeO1mEP0rI zkey&OgN0_h-b-!pzm?yrWA2Y%=di&f~bAz;4s#85RdESv6t$jKZS;+z+T5r}aZ`O5B zn##a_4n*oP_3XG|Y+$ml%dcT2hq_f3+*S_!Zm$z8mKi%o8|FH;>F(I!+U8k!A#kg| zz_Cks6U=rJ1cb`-=gwi~k_?zdr2%1PTF(Om&yH@BVd8f%nn^&^t34H)$!gcMt8XLv zaHR1ElqZPgA@i|)f)v-!Fu{4tS$zu=B*r!!mottf!Z!LMd`Dus$3GoTMHm(wOAW^w zzNMV#(X_8c#L7@<(@$9Aw({Z+zYfU~H)qM|_L$+@pDiFxWo!#Z%-h@X;yAVp<%(4K z#;6lD%T)u?p_k+A@EqR}p)klL2$^voi)Tx+l9OnA*w3f4DXv4~L}zG=wBsS(mvk$| zgfU5pd`wWz4xH8{owM(RZu0%`tW$v-93c>+pKBmy!E>JcX`0 zLC8<90#DKFc})j{o!NP_*g>3XqL^*DYnCpOr&XuXgwd14UaxT-Dk)A@xTp%vF_afIMxca+INsIJLWc=b4TgIRw11M zl4d1eq^%4$GEO(N-fHEgt|9QGOX``%@{=>P-s5Pn@G#)C$PnM!D9yDToRwJ=+NSw9 zzM_*#^dOQ?|9;ByF#WH-Z{kCz_k>$@3UIH9RX&N0aI*WpG#z~Ul!d=rpPDmlZ?gR4 zqdLEPv7g|NGu0&aG(8!}+2}{81z%*H{g5A0F=bdSNIrPbM;gc>5ALQ=Lv{(&lby%3 z!@-KHswsx1cNVa>7rcf7fkEwRKdCrPTVH_*If&RY1CV??Y<#MBpGE!G&!`O|q0X}` zMLRMhCqDgQPbv4*+hN4&Zm-1b?d3Lib%Jl#ZoA-MGMT4n*2Q*g>yN3Y+d80Of!b`@ zCKsdcM}Z9iy0nR<&EE%IyF}ULxODWEh~l**GrGAbg>IPQI9))Q(2;P;n9lU25ytB@ zbSzet%;((=DmYbhQzn5{6h&n#pZMOol=I5#p}(_IWM3K9Apa4wruNboI=#BeUC4m2 zlcPH!fKqSM>FXNWH9Ps)vo30hY9&0TJj;2%J0_0LF84S&DAwU2rnm`+0tKEwE*H+H zh-@}pl4;dC*3R3@OfJy3n)$}#g7+ID2jeL;&cE`N_2A*3S|B?NK2usqo;|z9bW`tO z3mh|tjySKI?b2A`dB2%>7Tyc5{*?P?D4Yt7P{V3`)ZP;(MfY4wY?)v^){^fQOLt|# zspy;AHIijW_!1fwE7OFW*&JbhVBXJy-=}iESA8g2G2hTEX4ZBVhuC%Ujhh`-#LrYn z+&vwN;e{n70nV^~Y@f0Y?yVuDyy!+lU*u=ik0k&fpG$I*z3d1kvYC`iNqc@Mt7xeBU zsDe@z#TaS`?z7}~w};vKrt&PDB85iPr%LRv^WowQ&=j^yWQO`PNqR`or0^Xu*J$Oq{qM3BC& zo^I19IU>;_9VXFaFEjG@>G1TI57O6MZ$1%28i2eyde$TI9J(LcsSQmfiD~Ns*WK`&PWPmDiO-2%d z6ae-=d$t`)_9ex1Pp;sW7mkv+QE!-tG&VvhP+|h*qdmk;@|&sFSnJgn6jB5{oA@;V zk)YNDLAPQkI*#9%JA`P7^F;?ncNPBheGy>0_q@fA!zwH1#MkaaB?iwY3{FutN-G(o z9HJ$=Pw6~<&Tex4JMjYe*hyBDoML1>>3a0v(zk-fU=i>*;yaJvL*yTX^JB~KhQaDj z_ggRCbbn7PaGwsZ{&2HD<64VY8r8`^Nsz>VF~v1eu|7^*6!FTcTc<>m{9+uUcHO#B zehwc|r0kR8ZP7ksEwUJs{aRnw=o^3h0{BI#yk@c411hV|aNvjUFH;WQp{0QoZf9DI zEAv%maEy%k25&WJ`o<`)InI=Kr3xR^*L zcAr?~wWS$nedS!)qtpLRGVv?griV3RYXZ~&o@5Hz)Sl)hLc1oxbt=w4BZaFR(^c!z zB3!a-mXkDi{^7TP3Sjz{8!@Qy=zvbG%*zdcrtvQ>w8I{BBouc&*G-E3z)vJ`;mQOl zS4>Ku5~O3!s25E@h0?HYRMdlttX2`h&q38XPA&*T{Xp&yBJo$69g`(WygEuP>!Dt2 zMz}VP8RdmFMF5=Qg4oX+!V{m-$Du0n3H)ISf4f^sNA5ncML!siSl#S@p!j-3`vZF} zn9V@bpHK)x7=P7`7)vzGjrE%p;$CdzF3%u$(je`dmi7K%DE(6jzN&q=~1);gmMa zkH2RDc#u}dfRK3Vs8#ih=rK zJS2IRgSjevisT^|Pc@p8m$-SkGtu=r?y&5F&i9OH2%&(6+(mdkP8Wt!+l5)5L|#j5jB0&MygE;hJ(q40&0WAgY=k)Nv5 z@q8Jcf!a-~vbu;H=wSCWHb~v|UGioHK}?h-eV-KCz~Q==F1K2_y9qM;erlEIb+CVn zlU{U_QUSbs`1p|`cu=BCOVo&oLdgXZc{yK#8zL>JX1w?HB~sM)nI}v$PuKXagMM}m z2ME05{$842G0csp=n1%t@QyEpKNj$tJxo)<;IG7c+r-k7XL?xc^<9s#_X}@}ux#ow zz3LYNFCsZDi!>eQ&|okj&{Z7Q$}{UYllQcFh$(jlA0rxD#iHb_k)73bIrwOKMc)vp z%#TLCRydM_kOkB-F%t)g=Jp!|u(bK!Sb=$}WVX@Gh6H;^j@geOhYPQN9u4=n(fut8KS!bWZQC`7K&A=^@IUA+x^uPUxBIO*-<@N4Y>| zi()RU-<2j4ON`sQ>&v7PNhZI8O@#$BoQ4i8R{B{_*?3UeB%^QS2h^F1$_Og}p37xO zc(B6Qqp$Id>tgtMRk|Ca!}i<$JdHN#+Ui<60PhW9bcYquu4r`lu|D&f-QexT&)78d zG-8)fA5d)f>hWnFZa6a|&}`1(e?Df-47vKR{GcF1r^?^3Lsr}bUgdS=Wei#<8_Y*E zvoPrE7T_9L9)kRz4|)*>dGwbvQt}NbtMirm6i4fdZ?kmEg}0xRzwa`?W2t4{S8yv` z>QMr@U^4NMB^kc1{c-!r)v~sPhZ9>v$4(YTU&M;Rhe0^94@$2NQuWgA%OQyz!UQIz z0Nr2lGM$2%R$ek)c~vmoDX59xh!SfJv;7^XUs(7&q#56(9%5)A4>Q5X`yt64qd6Dm zeIhz*$?B(K6EhclKFtU{7XLRXTmiONcy;;;Y=)o{npwKrqZgmfR+1V?VY}wax1_?R zo}2Za*2Qh@JO^VsTVZ+bY&!T&Hk7cDE15YVG(?aMhT4Q0;s~3jPrefKwoy0m<&I<9 zK5MVGfK4A0NOoZrB|BDp>c*ckS$PuHPP*3AV;0O8pk$EBa_Ojp;0Ip~jiwM|FGC!O zdju84+#x-=v|r=v+E1Y)TDC57&H9n07!_HIZ=J$*6N@c5@XmiM4SN_lWYE~a)m9#2 zIKUUo>8am+L5;+y2ASlm+Z~ylFsDoo-zdiilBJL00uXsPc14%gj8+l{cJB+}v{^Ns%hcCJ&${>I zl4AU;I&CVRYodMi)X@B-6cmOVIQHCbWFv`HA$QJ)`3*+%ykG#`xkw|yaIt6Ut0@(v zf;4BW#WNaKiLJS-?<%@^&C8abZ;yl&6g!Nv1n-M8i^))#2^rB1g@*CIE;ct2& z`6^g0%~wesGob7L$W}lmI^#3^S$bYTVA-G*TR~>;%LP++7Pa6q}d6rwhu?)8;>zm*nLUbbY_l_-T3c<8~>aTtE9{~$EkF^^|8X) z{PHt!PXnY)Ohfhprvjow`-bMNY&ql=uab&)c~-fS1~J64I#3i*KT66=g@u~D^4B@F zT*eU1S|AJJ8Qhh@ZAE$$WkSDp`musZ{greRjSq(n9V}@$**7kb31bG?-W#yS%4+#` z6-CF7ay4juv`nZ@D90tT#fU=xS!)N?cI53|H_uL15DB)&4FkiZA=Q@=Rl<2PQi$W{ zHPT_HP&}n)gLmcdbpMB}zY2>p+O~CJ+?^nW7aAZ0hv4o3g1ZHGm%`npa0nJW1b26L zhv4pBxF6>F_gwow&$+G3@2S!I7`?UjcGLo(rAXK9um}`A-gZz^aMK>C6Ol|t<{jlX z{j|9bWQ(A@(wH4+5C-hkgkT~!rpHL1ecuM%PfSI-6cRDzni1H_eGA?u{GV%q$Bc8N zra`Q2mE5az%$sgV{3U9LglcFQc$qaP@Na|}-#f(4{PB>?Hm4`c&ixGsJ)k>0%*6iw zlx9^l$r8(uVt`}Acfz;C35*}8qk;7{b!b0aYQ&{=(`0uMy3rc$gL365u=$|?REv(HZQ^4+XisW zfbTs;1A7d_SnTHwkEQR3IyEpX|{ebT`mPlu?8)=%$vr%5yu3h%4 zh`o5UNl}v@j_w2ZbD`Yn9;j&n)P!FyTGNi3_Oxg#(Z}T$&*X6T`qnuubl84|{FaM| z#0cz@KRPUGbB$&vR6hKG1Ak-E4Zi-1%vqH`UbM~BsgcpttK~4B|BdQ`a)x32Lrg)| z13$1>2Gx;=c{v#}1j1!Qnmp>LJ(y%xbDE2f{>`hfoWyPsCGzBUp`>5CXg*dY^^Af} zpW+!?pwQ6V}*hA+Spf=<7#Bl2mGj4RiP#*L%l_O59l(OK91Bw zW>{Hxq?JZ|ik5`PB^#$&C>ry$D`<2x1lZxLX!;w6;X6iA(XpVYC@k7bU0V6*dqN@) zpb~U9%CtG*34fagsA6Tm;t3;{=uMK}sLJ?5=n~O~^kHYCVuz|GoFNLog}`3fg15{h ziB7=IU{|bjj|C!2GRd~()^L6*J3KmSBk2XF2-`oUcfx{W;n0p>otp8GM@D(LLZ{QH z83M_1)>xDTfm1-I?(t=ky_xrkA5+Kh$9{%!LpHv(yaUgf?NQ+2Y1Tc=7t?d2hx$O4sNy5eQdr`g3GIX)`|8YANFE+vmvn+z;t81U0)qf0;gV!)bFer} z3yMwL^f=F1O-c7@Gh`jc4wky|ZP8I$dobyfuK{c-2+Ln?CA_xzIo8hfL!f0HgP@sx zg&;uy?z^Mh-+D-L>Y-_Gog;F;WB#>^Y_jiTz>@8)yzfa4<0Q`^>e%VK4#|>Sw`dTp zCfY2aq?JZsBR(3NqK6iTKZR};ZU7F#a)$Xbyk;xt*_)_M#ecs$M{`+fx3fL?0|U4W z&qC|wXf?N-R)4Gfkupi!&q=IHxNP0suw{dgl6t@X{eIGU**k>a09|1;@pEsEepRfQ zz=6E5d}{xXRIlcvXkH|QR$NunK*h!EVQYQ#RpP)f30f-JSDIiX3Aai0prKt7u6FPX*F4yN<&{7 zV|F;28r7hGrNJdOefXAgw*OC4B;bIFFv9);7IomZ-KM3sp|B#f*w#Y@z>A=X5ZU7D zLBTqrP1|OVQF=xX@u7!rK^tUP5CXJjoGaDTSNjxy*g5Y|18d7kv=RVz@Q5i>(AMlH6BR;cI81qQG?4~vqK<&OD;J}q%!!Ze< zXm{N*e``aAk~`F7YUwW`DE>vXg_wTgSj>mZ5|gx8kirA_VL>`3&f6IJDWT%6($_Kb z{Q?XW_aM%WOstES+9etSRaKHyw8UH)IeTpCvh*Xt=UQ7zYt!a6AuTA}VnYZ>SOz%u zi_6OmU7)E3%p_MCKY|k^(I)uQ3go%`jPMYXv~rigSoTok95^JUwi&)blLd86QmV#U5ZrDGK%uTr^R{m-n==SG4lYoER(3He+c1Zfm)UWWm()fW{eIr;Ka0z*e7T?)++R)eAZN$V^8;{*zZ>^EA6HD1UkV|Yx-0N zD$LyOL4`STNKg_sqtpHSZIk6YVkbRDGAw^NI;X?G>541$i_x^o7P^TyqT?cGqp?R&)4mIGnA=83Mi2Q%$OZY5$xpFKV(@XKdvk_KLwHfhRZ@f3jKaRxGr5AByp73L_4>K=^rqU)a zY!PkrGF_KFp7W*ZGK?t3v`1-jnV)K{|4t)bKsLuO*8dg8&_ zIjLj;k@lBVfNQ%7(uE79c&(O-njg;bA>oXZLwx#^TQ99N1g+G^`y3CUKZ2VYV&sx( z!i|Sy=%Pnd@HA99tRPSa+Odt|mrOAE;T7sNdy?nmE_k*q+6m`^2X2dAdjFJ#h9I{k z6^3ocH2vJA*QnhCxg_~$_>CPL&*HvOWwxL720jHqi?83A0=@-f8;8t`sx$=`e7V4K zpN*SromRwbflw32p@>QAF-hB~W>h3-WIlGK8A(E;0UZjS1O$tc!-9Ga^p0GjT->V@ zs%AXI3xJvm@1n$r_BdCbQk^o`YG%}G($C8}4uy@Zr{Slne&IM$UnvrRLcpNwN{hn$ zaILn2QEj~L5%{)T%i8^=Rm6o6RJitK?tr^>tH`QWEM+qBIngK~pOXc;4_OC$+ z6>d?H>@&(CCHzV3%6e8z_~d?vm34`pTDNTE(SvCo0G?Ktl!{bG%eUmf_};ycXj*fr zotkf|MV@-5DqePfv#`Kj^zzc8NX@iHv#RUg8W)nF#rBMI;P2R~ShHFFDqd@4;9P_^ zpXvdkg1F#A7o?L+V1}i5QYdu9Rzft3yrYnB0e~yooHS>NVd`o{RXI3!Q`yig2_CS` zylboLS}9uxSfC%CXJcCsM z^-t2=``pXlzhdWJ3TKqnJo_W|h?2oYWk~apkn%SX;_7FV!y~TMZNcsuimkcm68m&u z6nPMZbMzA|V7G(Pc>dd&OsQ>u!Y{Llcjfs_+?5t?(w5=d1!&ji;`XdomZIqii z3o|)WBPOz+i^dfhgSB()3*&duw9CWJiuxEY&Ic}LziV*9o+5X=bhJ#%kT3eoc<)By zTNc*yZf`^QWrZkPwGPS zP@~AvsT>MSUd}jEOcSnF5nVW>)I4HnT?L)>niuaZ5?kak&F}9fmn@9j9&mZ?!eimM<;pLO)rm-XJUg9N-%~AkE0B;u!uBA7iCXYFe3A2$ z&&_$}QO4qcMAwLpn-UzuB2#3O{--4*M?C^Zg7`OX6N1Ge8Ua@CQgx)?ORpa5Xs}r2 zF_uau*DOCNw~NGGa;us1rYO|4=R~YDnL);>8)~@}*sf&y{RKfHlzHAO``IG~64@W0 z#YcX=TmP7Xg%T?#=t09rGLP*PwKsf3f+wbLu0A+rM4`hNBCgLw;T#-hOK$belTNr| z0fSy(3z=Cqy{njpvQhD|B0R}hObsI6&bkmTr@#V)0sQ2w&b9Kw@)xS1w}=s26%qf5 zmDkQ4#gK#K`XO~!DNm1Jx7`?9>A=oVGwEu(naW&JY+&cwiGNBF7v|pGgyKOnvSMR= zirxzFi7aChYl*xF{!&*RTivJr<%8`B%SiXn&!0uSUukR|W=O{9p+_aFn*3isyY%@Gm?j!{fldn$mo`?yk=DAcU*&`ZPe%PC#jP%F z$$Qt-Ia^$Z0zZS=&E}w7ngQtjJ~cdc5|X9ANv-XXnAjUBY-~0@cgHVha!fu)AcG1| zv~#z#6B_9cO4YxHo6%Mup|e3;lNkHBvQwaCjCHRKH6YmYi+TDtK@y<;`(#x~{+^Xv zS?}HAmske$)%gK5s&lVnvW7evTN3`KvY;c5`h4;CY7;wlk(F7A76OiWNr*JWW&s6t zEBKyI5`7u*HY1?`;|%5qaGpvC)6G)!(kLFBsL}RdjC@O;%s%o@x3iebLI`;`$E1o=4_7S{ag+MV!*D>GA$#umbBOHG_fe;1h(yzz!2Z&;pqxms zphfX`!fljaB}}^RL6twaUZRDg%d+;%{Xgl68&Rl)xW#m|qw)YdWAS@W=|2fWj?rDY zpHQfsYAJd?-d{ySEms%gEm?kbOKn9bxRAO}S3PExlDkBf4C(OL>sUt8##j_Sm!Bn- zgAR}<6L#RGO|zMXpa4AOlLZ=+2-!7d9M%yJL7ECtX)6vF;Eb1}vMd9S@}D3SGKChb zIMwd(IwS<^y7TyWmJ4k(DA}lPcDQyytv6|MNUAy2u6{|SRYtfdR&6dG9JfcEY+mIJ zpy#WX+CZE2ryxm-hM$DpgTLlUtI3s%Gx1HMDu6AcixN3Wxl`0R(>Jg#Z}=I> z!X)29EZK*15TWNZ0tm>WnF*Q4xB zm?^y`!oy~2tmJ%RHARC!zKK-3$$OqvN$(Ro3VmtVE^|??Myw*xD3$DZoYfn zRCm-JD2SUnU72mybaa4X3m`9~`7S*wZO?kV39VkI7q}PkR&4SaH7kxF?fch=-A`d?_;XAHQaFPii;R^< z*|s!L21$x&t|mOJpXvbTR3w_PutzBCg@3p6QQl zpLlIJb-q3-Hm@2k`24@o1|2_K6a{ZHSK2Jc)N@=1s!wJ}2T59w|BkMaVc!OKJ}~>4 zH(vgK!3OWR_iLsI7fyxS9oKO}kiHJ!{`>zZ;Q^?K&;uyn?04-F#6kzM9#aEfE%viT z{>I06?DbcsepH|@L(WTK`A8l=-#_EO&4OZ~aABl9^<_U3UH7Uy~Zo{T%p9;_R zQD@_4K+JxrQj;vQ^pDy_4Q9F6T~Ws$l1t_QPX%xW9CA_0AVWSTP8&nY(MyE5CRqL? zp-6q~U_BOq`bZdMoK9n%+313lbD zF*8^P{h;QUA~UZYH4RXMrX=Yd4iE%rq2yW&!0yG|q}yDsh6xwO_?wSOy8skb6(tXy zs$tgGv94^eT0c8HW^qJ2gmumOnZN_02VnJa zEeLHup0+@j`lhLuU=Z%9C#RQg&dG>57u}18j`>o(g zg~}r%FfIMyWRSF;XF2ghj%ld6_-r}VU$^gdez?9gqQui*BnO`>+ab==1wsM49 zw(|4KQOmHdA3a7+sBm1Fe2)A!d4W^w=&173p&YG7;mHv${26LjHpy-fE2D|fU_6QO zcR3%4IsFp%5u5B=wyv>b{UB_d$%o_aUm2jtLf)sz$K67WsyaNnftm)9aL5z}i)tD) ztPH6hS!%A8u!lct4SU%Fl-F__K-pjR8=+Z2zbP-lnHLwjly&3hX8I8^omx-mJT4`& zRf3>zo~EuxXdFKcw{sle@{+lhq@SeUHh%rhb9}U7iy4fR9>6%|q zdnQYycH9knc{)uJoc{G(n^DGubDx)`9NT?2w{ccvAQcBDs6I-fNy zScI49z7CULF8ghQSK6)39m>&A5iv+p?%yz;eK)&|0O(xH2l{>%ov5f*qoaZ9JF1L! z2>CXAeLt5+#SvKkYcczHNli71B;@7{HJz@pHOMJcUpg(3K2Fy~S+4#hKG!dq|Mjtm zK`IWTdGJ;vA=$YvjXwSKtgjlP#`wM+@tRzH)u}6tltx=h+gu_Sd&d7=`tnCmwBfZ- zfzQ9GMhd<0W^-7lU>T0}*Y5~6J+wOkM~|bk;Ndj+B+g+IC&NCWM0m;K|D7iBNI)r@ zUK2>ALZG%=vT1tpkIPI8@$PY7@sXw$%WxMqgz-fW$twFKpM}Ggt-7O!8L*lDNkE<$Y1vD*u580wj~FD);DcQ2g!~q=5sngKLtm-8b0;=CnF? zWtSJRGea2Ak}!Q7+FSv1+ge}Rfuo-vkzw~jJ~dp6!zk-#S0>^0q}4dS4NPJZi|}3t z-;7C+;81LM6d8Z^=-l! zzG_%#RSiJbDq6q8L%ca;qn0l<)IaB7_wJHYYE{)HV8QU%5T}?|4ED^#BEs;xJA>ua zQ{aN>4-Vmy(IFlsv93eD8UmV`;`x)7AdFeEV|zK3U=P>Ek#Js%^w~9_j19ZMjF{WdZ>_DzvPV>tHs7F+&5_A`D(qtLWI7db`wNO zHpQ2%5HYV1>Y|)0S^LEi+{luEKr=Z)lnFx7ne>2ZkNg zZ|d|0itD`Ci5Mt#%Jc4`fYABqu}n8-nI@4l+Kyr*H*^IcWnx?|rSoJ#P``)2Gg^YG z`qwRFJHv_4-i3w`zJbW1;ibdcCmYoj3d4e*o$$wGu&@f>Sakq-o;XEwe6KFwN3Tk( z&Q0r-9CBlVNN*)u8ny$uDqS>CcR}L39fXW#rfr6JB82vaU0q1{)oR*x$t`!cl(y#q zgh%)^s6VFl5x{TA^+wBEMH^NpBaD`6XwEJ6ze%OlB)~*ibtbhY-6T#)@IXP8`h8p&Z(Wck;#Sak~Y3I;2dIPxQ zciAO&u0q}!R_SJ2=)>d_dV&=4b6DVRclNyV`S@b_Uf*v>krwx~8T(ED%W(&;T+`cI zzJh^BhnmsMW(NQ^-`N{;{F-J9QlQ{5S(KSFv@pdn`?cLhx*ckA{)jRKCVhy z`J}XO)ou8bzDUq1`NnW8ao^;>O?B?o%Ik7g8;WlKmlTmn3>R&PXK5^ALJNn{rzD%9 zjYEO+Z>TmtE)79|T3$L&G+SPP*M7iY=rvql4w_ARLkk#s_YE!$r0=|OfdbA1Yf7IA zx6tgGVj6{IVrB@4J{gHiLVBpVmX#>7pi5+VNhs~XqCtz+RJz!<^0sD3(_H3U-1r(Jr<)=0`Netw;f+l&4) zu#@c_GTC9KHOXI&McC}sC$0OfnZ=Mqm;qBc87kgdoK!Q?oOLGviRET5Ua}kNCpCv_ zXf+%qW{33N4zG@_aQZEvCry*SF}@NwE=fP+wtGO zZ@OxxeeH;4cxgyn&xfRH7wlx8OC?pqu0k#(6B2OuK8i~x&Yaq)ya4}!SLy=)#O7J-|t@1dG1j-P5Gys5oxEIfjkEaJx^=0CK< z)7G4OU~=(v1Ihw1X!zCGK)k)sCTxUjc0Z z2{V})+d*9jEA##v`C~ic4)rSuc@wSe0GCe4tnnA3{jB=}>X5IG?is6X1^yUsy#?{d zgunhNLdh>T7uUncl)_uwW4+W>H$7$6lX7C$9Rv4e2r!*c4fEV2S2iA{bUt26FEyvk zlIih$H4V^ictPOp-?o9|O_-5LmBYX5DD-?z_)vRPuS9=0N~1XGxLV$;lJ-_b1c0j~ z1~k|+m$+Y*R`M>U99Tn@xEDH_$mt2IncwiKMyvv?YZgsy9tAD!Uu#yOR)8!hQ>-vO zS-qKx2Dxk7fFaqS5u;nqwNh5!_~I&ZO)_={;sHDWd{5Vf@iH?!RbRk6uhCUHwuly| z*lal))T@`v~Z{YtkuE?Z^UaJCU_CIo7itTU2MsyBh3lUm#oTbHbV$<< zAPnd=Yaat=;^N`>O!@^_&Vo-}k%(9UJZ zivKRc_J;vP5pE_e;9zR+-x1yxpZrH=1q{s~wSg#szwL-Nu@>OHRMgKq89wNrgBFec z4dF9}N%ChYF;^Q`364({jD!(?dL=~}$k++b!0dtiV@nzSE%|Mz)UJ@JsskaIN8N5SO}BdnDTez_t|NNbS#{pzG|RH=^t z`7x!+*c&cF)-jUun*alP3DIni}h`nX5y34JL)ow{#5~PL1}OZ)6v$RB!yh4%t~4fj+O#| zt+h^DVxWrXOhid^p*FpaDQxhtsA-&{#~kK0=#3)UL(2dmv;y$?jF00N9KkWO(r5H# z+`XN3`HWnqjMP?NbCUKH;WeA*pz8eMhLsu`O6e3B)0tN!J->6Qo4Ag~xsA%jV}#q* zy@2Wts2}M&32QuDf5vVY0i~MtSfSuu?w4<0_!o>H_8U7del6R)VRpR0Yb**pPrlC@ zv7+xr}ZCCme6BpWm|7p8mPC;eVA8@6s=b6yL|*>>(HzFpp22;TkMQRLw| zVoDh~PhMPia9;zJuFwX)n_~*?)0*&07uWw4V}$Cs8owAa?kwAh)A8)26L_CM_gSR! z6;%AMr}zomu+6{QR~SR;mflvDxkM^ckq~j^uN2QmAc13$oa8IVT@#_`V%71_&o#{X zQk!$whJoM-i=F~xL-(#(Oc>}%Dq35Q=UNfNpqOEf?ud_@GSc+KI+y`W=OMpUXrcew zq6OHBlggrn`Tn~%!pB|vt@XdfhXRIZL*#pUSnb!nx0dUZ42i!4W7nGQy4;V6TDIW~ zJV@bG!}K;!FMoolImg}OIWC2Z$fzI$Y@Z%c!$re~j|TU*)X(^WKkdfEFYyxeGB3YL z)dA?lEq8aM5Hmgcs4QpxL>;QVU^xafw#`_@$$;G8`m89VV2$7c%*FFO3a3;Em6NUM zi9tub`Awu@@iGYM{I4|q&{`4h0tX0vCSdpN;NMj2XnAL~dW|n}da%v3PN-#}_LINR zkWGG8LJcn^%waWYQ5cNuOnnS7|00ˋN!x)IZWl)(BvN;qGADBqZv_7ZEBYIKrk z1M{|)(U5HlwRgxli7!QHN7|NM0R!mNj#F-f9zR7M3%;iq zJUesvB3Y?hiOxXM->hD%e9+CceXkzO6V+n&e!*{}ZU^&A9;04(BKZ`Q9>!T?6!glS ztw-~v7*a&B(?xuK5_w`fnjd`#3W#d^+14X1CHyO)LqxbX2M;+*|KLunQb;z@(7vD& zGFv2XCbdVwOo*QXbD)_XBq%GyNtIY3wqRD@)jTI9mZ4}5Mtm8|Z()`m7(`R}D~3jy z75r%B==v`C2D=mJl>1iw=}RX=;xfko?r})W^PkFdsJ9!cK!8Flj|ix5D9>wLne>R{ zXP)g;M|Z3m`6+kxI?OIl_Jh$w&fCu7*-mJ@(Y?@1nBUni^kG4n*s!XDW5!jy)C(@H zZ0XiY?>F}Qba6%V56wf+?fxuQ!~LPx@4DV}@omNLcK%%De<6NN;n;@yJ>D2}sCWC_ z-6n63dE;AVz(Dqqq)wx!lob@u$h;?}&tyFfdSom8n{Ch~(e_&`V_jKh@Ts%Ekf%18$7+Y0sy3FelQ+kW9!2dLO>xJ2s{ z{Z%Ys({Zrt!Q-9-o88~r0^VBJHIO`NiJ|{G1FI-{HIJqk3D+rI$EkEc2YfkKHOR;! zJ0UlSKbv@gC44eIn?9LINmw-`hA`bRsVv>$JP9gm@dQyB-i*Ez=*MyfeBnM^O{S&@ z%KV0Mt(&#+2aq{aKUket{{Yf$6*9@EsZn*8tlKrmTbb1gRJ=Am(LjWV(sG+8>V)!oUSIKnA3Q3J?lk5L$)XMwh4EWzj)XSLLj#yesb&`fz@GipFCK!g60VWw=PZ)}U{mrhJ~m6k zLy0f=BZ($P`1j2>!-KY*-_4vg+9rVw6e`5d0n}J*WN<^q=R1zr!ew_oqn(1K=%Dyq z@SG&!B_&N78VMuJ4O)Od*Q@yu&Fdg6A*Q`=>4$CcI|IdXlilb=EQ0=l07AWFUp#4Y zjuVxDlDEMmP#_1!DUmQ19lV{Cco2)-GSpj!6;6;JwngU+1Y}9Chaa0nSg`NQrE3gX zv1BJ8)?;^2R-Rxjj!;QDL~p^h4FTz_r$b$C3a^=TA(bLzQHM$9Jkwg93UuQD9BW9@PuiEz>eg#y#n-D+Y0zyrwyO~7qF==-GhNj742msZFf(x zwxKB_O4w{Y^^u5Vrh6P*=DO`giJze8Wwg!q!fe4rRGng4n~0N9B>y=#q_#(8!{~9h^BQw?zsXHEezLl&FaGOx7&VPQ zgCrCUP97VwiyX*D5*Dv29p0Fz=SjtrrgB}ifP|e+#=k3f5zb+gcqm8b?%id*h~#O? zGiE+GIV$+eyp+qIbDQwAFG>W*O)w)G%a*-~7JC+3>~+U zQNX{!W3Fc(|6TGRR9m#X zPoXGvm&TFS$6KIThru9{An8uXbecUN+o#6JAJr$gWpFEiS9k zdvSeY+yYM(&2=y;^+x%8n3D2x$$-m)G;h>02hLI2Q{cfP?JjI#%I=QlH-|^OPTWl1Mc`~l! zH*R`G_dl%9|A4y8z$v+(!shekB9iL^BK0q_H-!FCj=x87@|9{j?FrTBLQB( zFJU*y#pq$w44RU#4o`Sdawdfx1FR-SbtqI!mrq**nPLi?cv03BNb-oZNTP|2C~~m=?k4*iA&mVY=l5CqoC4XQ`2ZeA zXQV>=&l6~Y&anvK*aOC6`NZfzng?2zw&Ihc#rrurMS9mT=oPDnl zt%35#8AC1?U~>4G3^rL`ra*QMqeP`5sBe(v=Rv9b08N%SA4#+AlsYZh+j`QOlo56F zNOUdqttInq95jGjQV@S`i}ew%Vh7r9OoEXdDZ+>SbBSWboKT>NnxFK(3;YvnWwNLR zR$rQlW`knExEKSbRpPpyUyUj9Me?@R(F0}n-kBF%=-{f<(^~KyBJwlvR>WglW3ojc zqY~{>g>?*RF*fOOi7M`O#0S#Rx-7MDN!#|bg^pJ3&fxX>+HTZT#|GA(9l=#sA!qd} zEQ2=|j@%OWwi&2n=`a?@NCG8pzpvBtOMO zD5!5cGU6ht&v&FQ+#4=yj;NEhnK2`XPq5~JJNi?uJ=e?~HeR`CH{_&rZdpzW9ljs7 zkQV0k_al{T%u23oqv(${6!*Kj5;K3Um$_}z2|7uzaEO8^3&RKf%pVrq$sG({+n@O` z%SR`rq?oM7){A0@SIfCK4#|98WZYj}HCkN;A?FjHLqk(A$HYYadPCzUW?#%dj`gN0 z*3j{G1y2O_8)fvATE!?|r2OoZTSc3a^P_>xiF%@?i$imFdP*lKt^dp{g%CM~lwHRB z=R|P4Ol7>{;Dn_j&B2>?To)*$>;2x~eO=FMJlU+k_b7RBP0xLGyW{cfqSfyznK@fy z?=oEv08Lfs;wSD$juWx>%jS<|0@vASwD@d0Rsv?FS|D~U}0*xxm?ti*D~tvTBi7Ok{hshN26TEjwDZe zhwDES9}2nZ??ElFmFGxYYeL&xr1swBHIV3h1q#@1xDvvYou5C9T;bRI^6&QwY}*@< z6{M#SpV3_otRCI3|1BbAJZ+VrmFTys4=bsWY~Zo zUBo-El7V$D&sRfXF6XmpaFthoxlju2rzGiBM0fq(!)A(7W0TNL^;m<~Ec$U5-eC7? z-d00bqwU(i+7}~7_~t=FQsh%@X~d=GXAl!~<`wr$j2^Nr`gq@S!^quJTSo)0mS%26 zyjPt+>onEuB%JB!u@$L_x*1IK$O@r-(>hzPr+V>Rb*EZz#eaMgp3rZehHhQ_F=&lm z7Z;P0hNlCthQcV%Oz@0D}7zRr#` zVu$nl0HlZ5%x&}fO6JAul6BDef4WRpNJv))Y*_SI=gEYQt~&K|3KQ|7s&MvR$s7dF z%(|t@XQsP%=e(DjY{TDy(lI;M+5BMt#^T~N`mFER!?N6zzxM?LB&+`)3QPOzl&||` zvtz0+o2{{H0WBcspOsdOJwR@J!jfhegtZ`uqGBYw7dblAn+?5*+6kJMA9?l-BTD<9 zi`bKh|FHU>gPf#dAvti5eeZ9{; z-9v9&(~7*`andOuONen@A(2pGG2Xd{f=<`yUukiIFPGOFqvsy(?#Ud;JMZ(q<}f!I zbY9lTS+F-x`X1|p?grN;$zKZrFh#pDjw?*u&6sD)mWxk9QrLeMi7S*#$OIfPgyu8Q zsqFFDDSbJ?AInKNn2tSF@5fNSUQim%&XlJVf7@*>kDP(R-o{|^p&E5*nogK#7#e>P z%`cb7j4ztK+SF(Z?%&NLZl*i=ML!&;8e3tv%RSxpmBFbNVA;G|=s0^}uUh{uE5Z9V zymv4lL`@%&LE{{6@hgdVaZx)#0(((0+{9IJgSt5h`sg?JF6XKs*^B4tM((8LtWH1Z zftY=}#jNEpw3AQ&i4sHeoQh5qAo|-0=ze7VCJK%RF#6H z4xh$Xo2fGLP8jA`mN4B~{sS-38b_%=tKH3)owW#_!Otz$)$O$A*^Ms7RoD>{lQ>Dp zP)l$}c+cx2FRhq-T4Cswr3GUsGrIdYvB+J>G`;8p=t9La&t(alvzkOL=l*^L$*&KGP9k=Q8`ZO>55JTv5Q`3Mk|}$H#ROp!w}Zm z-8YO%5zpJ7-PolCU!n{c;>hd0)vjH+AD_xrrU;#{81v4hp2?iJ_wSlUnO970w;hc{ zJqb6N->1Gxd4YBb6k9~A3!bItZBvZ3E8SpI2tRkcuK782UhCwzk+=CxF%{siz8wpu zX0}Y5B0k-cEDAClJNIF#H*P!`d97dYIZW9S5j(rD`#MiTwq7Iwkt4koa_&thce?Xy zZ5--TlKXO@W5&LBi2t=`^NaYuO6P#iC0Z+MQDP4eCu#RXFFHCW^%!KXS2)RE9;xE^ z&?cgt??3ix@j03FWOpm_Kb8?saaOi9EUrPei2t(`fv;@CgUgPD*!6<(jVXhxrSa1u zDsy{e){5=&H>=LXuOkdE6SW!Ucx0!YAlJ%gNb0$Jw3o*^Zdzr>?Sz+|iN}!76YhPJ z{v$1Ob2s$;$ccT+$?fTKzPWX-ri&2hkbu(fsf;AvqEB94q4lJqec7k#{r5h-nhthV z%JHw{7E}xrLzS=h*@$e4hXYm{Z(^Y|lSY@{sEy(Ot-elw+%xYRH$WR_3;$1ZUm4YA z6RnFAthih77B7Vs4^oO1FYZ>{-GV#CtvJQKxD(uqTZ>z9_XN55zH`p~bJjZN{=E5< zl{d5U&O4czy`SgVv-ee#=5yRgS_pSe`a=qvU3m*-b;p8!YR4S){cecR`8a;&k^b|1 zw!M+ZpcG{W4D09v-zVtHGOT)oznXmvD{n(YhF@^XCqY5w$7?xy^EX_4<~m>JRx5rV ze2-XL&5cZ7%6(e;_H7BWnjI0jw3?n1nUnfSRW0^1xqxaQ;iIYny;Rg%uwzIw>ajp` z{J_V$a%I$Hrc)-Yl$3|d-LWg4T!FhTA1n3&C+XK6kr!}-P=pqC{K43FcfjUvOq+ZZ z-X!%~cwLH9As~oanDOBO5W~j}jO5f4c9^^5LBU4U;to=N&01W1r)tNwN_?Gt8orFC zOmIHA9FEYn3J8;_@V^#5WDZEJY#9CFM!->hpU+g-c88cy)DQAfN_Y2Arg z_B!_{D1~3uQm{Y74L_XNS))!OG|y+U6`iaL8nF zR>8=cDWA&?o^+DV4ghKGpTHp&d%Doz}hP$ z%%948IDJPSNrV7K`i-^N_Bx}Dv`!_zE_yC0pQJg_9tpifM@jO8!SQ7lVxvJ3lU!p- zIV<1^sZxPy+^YR6O#C$F(LJtqJU3axGY-qo;7 z@SF5k5*`%3>v3Bdn*4xWfzx_kXYEgifn3kKeKkIXHtkMX!?yvrjKYUQX}JuI%}TT2 zuA>rz1EDo8Caj-d1p56f=IKO62Wt`Z-=}b4`I+b9V&WTVwf#{3Dy%^*sFL34#iSfR zCq}M}PhP@Z68HP$YE!%WYPS#erM3LaXc~PnQVfNJtieT|Xzs%`5)+I zOViV0r7l~^B=0B`qFWVJUh=%B;k8wZUQ|~v@d~CWuYxnY52br}9(t;SS9edit<>v0 zCQVewD4J4&(*=F%mwxX6Hw7$n-f?CKM~x>EC5j*YZ#FtFSFVRop8dgUkrtIYwY(I-EVmBp3HA!}>vYrkj2t6@XX zZT#`WW#&fXo$pQ6558nvgHghlWV@c7!in?EA3o>DVeO8fU?8hTcMY~t5;RggSN}rG)F2Ny0GrN^4GOZybg~Aye#GOxYi#A zNS)7+o5!YB=zc$+qF?qG;BKu1=+=37__E)vyP%&7mn;ciZurSu<@3Jhvw8A9tEsL= z!wd{GTFYOfW`O7Laci2y<3_;_9G3`i>A7?FrB`}}YARnhjWsPcjXBWx1N>HRyFBt4($QdE<-kl1N^+3p5LRK`Z%FufyYLf}Op~Y+KJ1AJ$*az~33P~6I){5pw3 zG88}m>FG9g-3Q0xdC{<0;}xv{XGxvna4|XM4mbNl+~cfOncf%XS)t;w1KNo(4@2sB zj<{7C5(f$b;m7XP%i{&R^5b>f-6%?~vzzTFpJQyNtd$2~6=IpV-A8^>;cYYHa2!NQ zpaT9zVN+qH%%CSU#C;|%zz^GV(8$4h6F1tQGwm!8hR!)(0`k{+AnjIqD+|`=K)54r zfZiFmx{{=3xU{ymSp&<0gy%=s8;OJ!=xR;Kk;GBlcQ{U?2n`wd02 z3Bv`s_p3Sa!=1xNY*1G^34=dQ0*UCoS>ifn=Y>rE@9)K-^m|?n?^SrHQy;-7zpC2u zKO@V|nCQuxD6J{pko1INg!XVxcFrO3kT3Qjc+hKEX=lY0+)o5HilDbYekCtxEruWa zN+J3z-^)8A5PCkQ;K9@X(^tKgD4_C2F5phupqoY2VC^bvJroc$R@clRMxt3>t0*g| z_Ey1zAG=U~Hx#VAX32}IrMfKI@^YztP>FtXO@JZ{*lu8U|F^H26wI-M6SS2-8$#uJ|_KaXjs zW^r1M8r3ek0HU&=MHMu);!>9Vg(N3}&ufH4sU-y~je1z4$$_B_j0icRwsJa0ADAUd7Txp0mCnP;0njPv^WdCi$aZw@$s?0CW( zq(#98c#l=YQ1i=mNHI&f!nd)wugl=NkwqyR$?7tI#sYM_GxE6F`$PoJ1+6pIQydoE z4hAeQO?cGyT3p{#F0Z4Lm|OS2%iqY`WD>kkXRpqL68o8(7p1Mx;vkk%twLvP|$(0_cnYWJupb5%z*y#Moznq19upGs_8WZPwXxQzp5imof}N=JKTF=6>qx zMXZQWD6AH0o0g=0p7}%#TlU)5g>M;ZR`ym&AJ*6Cj+0uVvYZDl2jmW2U_?X!xbyr9 z>t4y6kNsF14SHA-1k!62paFWF2Aiyagk$DC#&&L+*U||0XIpjre)P7g&}!T*XrQl5 z6O$NNJ5q2uu=4JUhs}f#Um^rm2qX|SVwH)NGh`!Gy7 zc^pq^Q=N3#%(k%PKaPlSW6PBuv)U>15CrVNXX?IC0fcOvTJF~sCoUBxkXX}Gpv}Ok>Zj=ZF^mive zLpBnnI2|4ddct4pA&vB4uv}kzdEQYs?}N{1iw_PDF*pOSX`0s$0ueacdoZk&)!JxH zVZ?Hukk<4W!!bX{n>QSm{Csl$baX8=;r?wBAp3~l;L|WyQkOKUiv{5Ej!c8w)$-^L?#H6^t)PXr7+eryvTC=(Cum;_qKSiQ~pNmI|&rc2A^r5c-L`g81Y4`F`}JM^$Qs0bGG*+}eZCswarr6&ok40`cm8rjoUHu#aOf;+EZs`<4`LKSQfDr(ZiZf%pvjJ{AogC&;C2Deub#*(usTHj89pZ zDny<^MdWQ| zhpb;Ur_I@Z%o5kz&RK}@+(8!2$kfvj?RH;F!$$a4j3-feRE%&*Y?z`Is8u$M=m*Q;Lg2d*i7f7e_q)*cx# z+Jt+ds36u;DU6>OO2n$zA_pGx^DrXbXcnaPsP-RecDay$wXvZ0pqcD~Jyl(%v&mYD z!1?Cr^gFY^V23(QCuYUvx-flL2|i0jp#-J{(a7OZs+FJpq@S5c&kS}!P_WVf%GJ-=Vj3jZ2VVaDG?RK~*|AZUgn{m;$1nCqjOzj1$_!QX z^PQ955jT0B7qY2&uuY8GVvEVd5b-q4K_`$t)y7Gc|$xsQ3s;E;zEM~PEF)r%#nxCe^2NalB$QxCF zNYVF!s|_g<5rnm8ltKWo)-US^wgX+V6WB(4zljGuvMwZ-LY}Y;-2-=|Gly>3(Ttz3 zk7}Z(oF)ohoS;QqtQjrlmNa^uypL|#1k=+kr^X3nmT7?!e11gDX}_i|P>?Mxl!Yfw zJ_&y=x~znd9i4_^QgSe8Xz@SE_kY$yZ$%2&Q&y1j648nAbl+x~X z5fl`hqM814Z{kBV%pT}JLYlNxLckE!7E{D%u1SOiDkEXtrV7=o8|?)%8dML-fL(5Q z(>B*}wwCZkbHOb!f}VeWULT)2xuZks$LfnLRGQ&MLj%MO&$U``6iX+wY`(Dd(}M41 zHEMKT`QPu#edcGIvH#;)MX;DcOnj)8+huMT4$CYbm18<^2%rVGQJl+x2Jg8#FH?2? zd^1`Zl;!q&=sLeVYC&V3$N>go53d-m)*yDRdaWXj^`sg2scDiEb?YCWG~26^N;-7b zOI}%7*(9~Pd~H?2u=lCWid{lFUX&f2m9AsB{w>sOOp>|;L2CAagp7qM@lhWo@A$o@ z`>h0m3|Ph;WDNJ?4Szv_;%hR~+M_A(AR{&{Q&rrHfsps}~Vsl;vL z*(1x=xP4S$1IvNfuL@G`+>hrxHI@y0bR)fQ11CGrS)ql@)E1sFM(ABGh|4kCu+o5T zvRfTnb2Y)Dpd#CG_ck?r;R0?;Q6ty6MUVvAqa5ThI|?E>eK&iL zgiSRv7+W|B^t%r}XXyc9Yl4s+4W~UYpPwdqesXD{UYe3cPuWRuvcqK3r~f2C4rqco zxk^7fMxZx>QDqA$jD2xU#QFe7Ztten;Hb_M=4jbV$X z_p(pOvxSFH9)Zg{i^8jLn(r8fTaqGRI5vABh^^Np|k6E3g}ItUH3C z=*99Ez5BmvwX2`69EC2qCo@|v`+Xpr;1BQLTj_^c$Mt5v`~7rc=r-~@O}CJ0(oN)a zmQfVY&}5(ty1n4x&m*RaL}<48(w4n-b&wgVMe20uVu>;2?oEs1@8gNg8@+2~jI)8nQI*`m2Y`sD>mTi$$)cMCq6EgpeYqZ8&Jyek z=tRR?v5r%r4Bv?yk(zC6(IMtqhVGPiRX@-9Ua`0ztGa!X#~Li;{A z6WwFuV)XXReW151f{ig-H0k7vpI>*;>PPu;;8{MDQD;$2i>_cY6X~tHJYtNRJf3uwiJdI)ma&DcninTj zv(T5tJH`}2ZCimzMCcz-Ya$^wayR`h`DY2PbsTnIl=H@8-Zj;k+Dz`dK04^d5GI16 zmd|v6Dn|DRwF0y3as71R4f_cA#tY}@rE-S-fSM+LVu5B=iB0=f&-x_J#nIY`a>1^l zT-}KX;nI516_Ea8!{$`UvOC1He>eoh5S1Ma6f^L8&mTXZvERM+X*lS=rq}ooMNV<|K<^ zKh1Rme76L?d%-`p9*_-D?PY`2P2P^v0>bs;XvW4o=_MpI7)}#BImih8K(kh@p2((E zneSeG26CDT^FcHMORqs4Kl)qbD7<^29EplYu3O)#C@lh*tE}WMo!?Mf=-6p?z|_j5 z1}}wbt`)U^>q@aFY^K|Q_ZRnZ=^9m6_Qh7dS1Y@Qyd3@)GDLrBCk+(OTGt2?7X6m9 z3+89xII~Zs&Y@$ox0Fc#;s7&Fq_yW@7xCK&d(~VIQVw!zEP_oaW3`q`ggXn!LC2vr z+jY&sK~(s(lhlJ^h$$`!UiF1DKw$%b0d9Ne9IeLe@T^F2TeBG}V70QoIf$?(+_BVh z1+WihW7SpxrYx$<+*EC3-QLLvX?&dxh+uP<7r#uSe^0zB@OLW-yER!`eo*~{;@`f6 zhy^hE^{}_x_ON74>7(gCV!QDmm~=bbh;0Lr9-!it4@J2>ez>pX{9yst*t}@IqVL|} zz0D%xDLn%`)_)}h##04tPM_e8P*-Db6s8xV`5EmI1t$-@e=5FM2{^MvA#zPdG0Wk1 zq25JMC+BX9GbFlE8*j=&A@JBx>P|uS7mMebRFgvXXDM0TOZl7_&&^pFFZ`EVcU)j~G zzW@5ik3+jd?C-D4f(na8!+H&lBc2nQtjmQ(#R-a;k)bX)LDBs${bPyexiyXcLpF7a zS3ZsXv*q)}a8}3*d&?WSLPj%e1pA@Q>f&`(I(^{%f%nb$)h~+hDqDrxoYz%3#T{w87ki;R@|p_1PFp%dN`=%>U?pdIh@Xe z#NcwTpwo^jQ^YimQmG`aYofBLRTGKfi6ve)TrwF?ihZ;8`t?YTg+OcY4!wuZPfMyh zEE9HV?MGn~08c~>54o@qDv4nS-7zGEvH`v4aH+Ge%(Yi?8{)feZOu?jh^7fkCN>%d$$1{ zl;+&ljeug)c{+i@rcTG*`V#p^i?DgEM(E(B?E0t_v+azw)C7=)g%Ua67?1hWvGk|W z*jT9m!aa5$`@tX+$e=jF=?b?BMZ&4*XNiN;6t?Dz5sdzCJ-8@m1M9v&cON~?3u)O9 z|GbTtqM{r`Q@*5mOroto3Tu~vnu2Dymog!Q!6o*E;(s>lmri@o_jpJG3{d(x@7{P# zVzssc2_R&iR+R{zuA8HQ#ybUa$cb~WeMlnwbCT2|sXMq(O3k`D{sKmMk(fTZlMlOA zb9OhM4&CsI)J>Txi#kFuM{`17g=PfWqC8SL1<7pbFfdY0vpG8UTpl_lgQCsPcpg%LD1x#|BN** z6gHswoe#>Odg>#wS3Os`+HYk9iZ zGc{?RxJSv*`>_!l+~DzTk-cBNzmRcFHb^+G_Pg-M>7Mpf#F(q7VObAvRao^B*pOjp zH{kG%OixyfG(!@SAfpp0Tby`(<_ihndEHm2mW>ghU-ZwS!E4KKcpbKWl|1F)@0s7+ z$uW&y>xZ@dVDa&_`FTt#+892H{+2WHxnm?b$IYsL-3;$@iHj_!j!&8*A5U#ve@*so zqQ4J>MFVR(M=HkZqmk)1OMJ@AiSKHS4GrH!j36^4h}ov*bY18w9E@X=L3*zhUOFXpJR0Ta=Zg*nOn3mE zuOmmyh#&M=04lPPt&|Zl$<%HPVih*8`bY8KF8}K)_lScksf3EkSv$BEoUSt7+DFA> z%x%#Y`YU3P?+2i0A{y~^D}T_$K>-*O^v93`fs+TpnHQ+uO0P=U@Y+yK#sdOLDs>Q@ zs&@D%ST-x|Bl66ER;R)8`IVw57FL5%AqB&f4?F6+qcXD**z{+%m5O9xs(fO-lG@^6 zivDN35y$vHqc%c2@1=v;JdM4p1XCfL3LyTro_u z{;2Tz9)J>+$FeW5^efX;kX1Up`m#;6OiKX1P;EPquZy(45J9?2E91&II{|+`Z*kAU z#`7^dXp1^k4d5QqEazZtgyz#<&izk9g6vH5p|!2ja?YPk(AcxySItB~Jk@%_a0{9OLq+E}v7M8c?&z${6AM9UZWf z6|Lo;B+$6Dy2>$d8*BbK6bJjdVgM{IPE>IsevT#Dhub~;R3UosjEVA6-J^OYm$pvBo#!6VqAyXd{u&mq<1JOiP&Mft_<7q zsDHahq~(OCOrks@rmDm7cQYpf@(Q~QuKg?q0ef9y6R;&C)K(-~&+CKu97*Dy07??uJp~dw}@S_ zaL#=kPX-Z*5NxHYtkG`P1~ekvcP)$j%#O~%rT{1T{==ubr!(r;di>dt^`BRI>G0Br zmh_mZM4Q&}BG8cz21yxznLr9dz^)2$cyZyDth*{&TCX1-9@Gv}PYkvL#O&>3zkJcs z3*#3P*sUuqXVv8JZF5^ttEN8=V+jek`A!(bVW;%N!P2EKscD{2t%pplf9F!H`_J#O z0-6)fACdu_`ccswG^(zaMbhP4TS#sq88k(;svm;`I!&kZl0~idb#pJIXJivU9qUeI zxR&3JzrkW@cwhD_3o2KVNJ2DK+ZjoRM~UR+b-U6ZmR_XDf>Js8x#Ul}ul5hjQND*U ztyi*Q-G9Yp5d>t@e!9Px6&vwbu}8Gp^gBYNQ3(9pn09c^xo z87wAviTYK7wz$r=`Cs+m-~adDG5<$P_-~)$V*zBfxv`BV=Kll{{Kwe-4{yCdqkb@% z(f=2}eV?da+LKR5ev2y)1{>6IS}l$kp?Jffj9ecyot>SpIm4ts6_%Arv9gW1`J=68 z5Q`uJT-`k6Wu2V3_j-E7t?X>l4iBvovr~#-?LOXx%{7L_n(An?zr*0u^V9i3&y|Ya zm%@M!=wQ3^&|gX>&?J-%wg0&g70Fj|@&jK%6&t2##c2>r=vgKGBV^|3 zsa4L(wzG3kKQp^ZD4ZHC>i*mz(6io@pohXgVRg2joN;*210H=NZlT%ybF#dm%-ub* z9qe@)Nmvp*)S{vZ=S$3sOmBb3?J#VB_f|CbV9QHOg=ZGx;$005O|mX76+)5Yx$Ajn z=jqwE3}t|i))t*^!$CO4;AQuG$?_6SD}kzP1#7q816u)NE$^eX+hNpus;6BQRVP7> zrmwhVp){AUqOw-0y?Uyl)==M(*1@ZNi_xOepJKQ-xY#oC@|1%Vv;E<3w|a8ph5<7I zsu~8YH;sG~N-#wrQo>=Zp4Gd-3%=QsH!tRXvPf(uS(OamkUMVrE4{sa%#Dn=`CsJJ zS9c$uL60Rbqog7U&O1vOUhYS1MoOQ|CXT;Hb|mlmkps^Ma%y&4Nexm$vHr?hSw*f* z%h%XzZk;44fzsfjmer6e`AYe!1Qp)WEmY|8@9o`g*irTL%F<8K`Wy2L$%K}@F$eR3 zx&ddXZGo--oh0=Vuc_lZpM@BHz29{$RCxRQ$B7`wc@!1Drxx5%)^j+B7BSpx)lO77 z!6wdqRf{Lz8LuUGO$JQ8)cSJ{5Pq3Cnwg$v0v!vzzWO^sBZ&aMAAyXFs9EfkAcuz# zhU?($oSP6_=?Ly7)>=;J2PT<;r9gxLc<8JdpaVMc*NXa_4|2fL6rd>(U`8ve+ z{DSk~=p^mz?uk*Jgagmv$$4hYvr9IX2E;TZcCKk;GI61*Vr)zwZRZsi9B`LiQTZA- z=VFG<;b7-%TR(R!2CBp7Zr$!E?(YA9Gh-&pxi)<`jp29BxGW=xl<-H}f%^6a7l4D< z=HnfDl{8eEof zWU4D-6tP)+I34>$ zdIc}yi^tQxe2y2GvIPs`+i~?yyVW<+nWSuAe#^a&JWHj;Nvza+tK%rTbA9-?%iiu@ z-;mjKuh$!LbVuxXqUfgwm%U*H%9??YhIhdgOHn!ODAldr-?I8b=JBluEkixGaTZ5; zwWxFWy>v*psFo81tWJCm>3;66EgNS~k#$@ol}rw2%(Ly1rW9OcrvI+;c--&Y&n_l z?Eqe0iO@Juos_o^Z089Vq|S4HKD~XJIYisW_*6Z4XHDrfg1DT(iz0(0dV9{r#N|-g zNR8FqrUyBd+gy4mvoM_F@jCLDA2rtr*ZaMMJ?3!h#asr*PhE~@RB<@n)`{CoMe-yP z2ThtZ9WRPy6;-KfC6gnG16XFULEo{wb5WzHCZ2wr z0M_3760r?Uu~y+`xs58eP z7;X}|M+NXIDUaj{G+FYUV|iZz)Rijo*mV)5uy<)+?qj?WW#;+2eUHP-RhH8<0FHUb zS|aKS6}@8HwmK=3m`MfzN1zVhp?&x%i(3E-oI5yEMfLAXf- zbyFl=yfw;1mJJa|@2~^uNV(ejYSithCKV*wsV;hNnO3RWu9G%N+S53f>2Rj}15yK7 z_lbH^q(ibbboc9$HW;$5XO0s}Wn0vvbAa6CciiSF?k+E@kyoLGg%UZbpGF}AQKZ84o zJAd}^u>D7CanW}wB)ve{F}+w++#bOt4p$K&o4?pz80!5D{{E@A}6O&@nFsJbmEp9>Fo}KC}0Q50OIGk z*w~V05~FJ=>5H08!^G(E$tb91?s*^`7WLJZHX${B&BsqrpVqwzUB1}qe1@SAr%z-b zwkF_7_;hm)E>ApQDr{i*F z0KKSgPrKCSb@H0Cwrk8LmNkGQ=a{G9-HKQA@uO|?l-xSL&UQ2X-=44=nX&Fy&r1C_cRNOMDU8R>cbc_~Jly+~UK3&3w2wIoTS5NaC zkvU%;)ej+5zPps}J9Yd4;IHOTG(wH|cOqI(iGr!9+|+LHOFSBA+4qP! z|LAS;W=Ndfsw;|{D^yuz3g8p%j`SN9{ZZIqy0vZG-tq3MYNrwip5@YiSPr+SWFy(yhomPNn-zP4ZTD$4Ewk zym0C)pM?d$Tdu{D@wK~gUpzYAw0-ZQHubDN&g82QH9dPbi$!Biwb5ake;9eTX5-Km zzT_Kiwh+^nyarbNoC}h`i$Dg^4x*ULu;f!p2kRAIPn>L7MR#9uLGrWJC>th@7OMtS zuTdQb)aWMZwXlPugfab7nZpq9^RVExV{Hc#8>;~9Udz8e41eY=Kl4AwNEZ908X*=$ z^;oG_`=zPtK0{PmlikpEK;<1^i7O3#1sol^gx zDmCCp0wV!aT<~Dsvso=Xz0-$%$vX^=t2^*bw*9;F-;~Q9`yl}#jy zw+g=Y=+;?c^uud0pS`&(u`r0=+A?WTgoYB_zZ}s%1798-;OD_Ls*?8%y^$9n4**L| z&Xjx4qYr+k{c>ACyu)l4NWnh}HL+c`@(y=(v@Dd4+d$~3s?U@~afGdOaX`8MttC_B z_lix-UvL!^1-=C8C2oY)y4VO$ubf%8ZzvZ=ur_3i`}p3j{&G&iTfLo5Ot0NFSh@`x zY$z1M*THpp^|%(RG`CcMKbW9tXnvwro)a$PHq;_p*>2J@H9j`>k1AVKp04)XmOqFn zv@sfWrd8?<_1xR5d+DB}Xb!5y*P6ru=w)M$KsSew8Q5a>1 z1xm*xR$~l}>Cd3+_G*`s%{`1w-;c#m7`oi9TdEd)Knq-g|S@-^jwk9uIX!-nf)P-C}>kRr?LU zy_K`lyC;QQRbAN?falR0!o#USgj0%!6Rm%z@S#VCpCu?N6vMU5Z$9U6z;sIvLWe zbfy=wAbsg=(L@p`6qV0JHpcPSP%FLTxGBoX zU~-H@UOCeeTugdc%FUQ8Uv2!~VU}1JJsdWLX=D-q6Tv_ ze=s2bLgRRCz?}G75Webqt2`w#zs%DAh0ySCmdJqhhhuU!|AnIau@Eh=&tDKv=jz(! T!-}F1;4fJz#gEkzMnV4vLTRt? diff --git a/ALCF/notes/assets/cooldownHD.png b/ALCF/notes/assets/cooldownHD.png deleted file mode 100644 index cdd2f3337818b5a24301e1135fb55143fe3c975a..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 327887 zcmeEvXIN9))-Ga0K}3+ERFR@m0wTRC3L;HKdheakdlBhfKnO^+P^1^>kVprq(jkyY z4K0Mw0wjbR-1|FcA3ghf?tbq5=^vhC&9deibIx~;@s4+eh^H#@mo8qvNJK<*N#XIs z=R`#06+}d5#z@W)t_+Wj^Aiyfb6LO8a@A5+5;Fzb^BS80P0V>c?Hvi_L_`wOo{q+* zw&t!ZCgzsb4w9^>1_&#QwV5QVrl2yPvZJiImGxt9XLB`gl^3Sowx*(HtkO~!B|OCl z9oU<@8nbxX+c~(1c}lYW=vR#J`Q$QymE}hjS6fL|E#;>yvOs5Z79n0fUOrZ-i!2h( zW)@=4AIkmNnD9-K)ymb?Q49d^@bKXE5a0zmTLSn+MMVL8cL8_r@(^n9xOh3Z8hi3M zxZFJH;*UNan!A`fTRXa1107gS`ZP8Hy17cSvJ%Q!{_Nb<+T!26JGlJ$_yj`$PD%j$ zynKMaRyX&w{+sG2CI7B|(pXIExw#9_&h6y(lC1pq?g99JboHZ5`+q6>cage-HDSDW ze>6L}f!I@PPjfr%ht~Gy4laLmdsjf@r(&(Y6blFo{iU_Ki@CEb(4Nqk(5srMvpJy` zE1)YY;Kb;FKdnxfKES``kMQMRE5O`=FzwES`3)N>yG}$zrf4lI`&24h`uYt~Hb9lEw?CTqiZb!f;^^Ja!;8pkLHu2_{Z72t9aiOT$1EYMNN&va%VN$4!we=%omJ3HFSZ;7lGL=p;EfWlTZ;Af2~6 zf&^7_TZ#7kZF?j9;9Vys0J(`|c;UmXFBAT0R2AL)r?+8BASclyoC~;U;Cv2V_>sGW zd`_ITWu)iPVT$Q}3q97h7vZP3+Amz8u$5uBF%Wxctw!N?*Mc0`MAudRuUgzJ$q{VL z_&a_@_w;G54g(nUUvFKep`!KAQH-mi$5yJXH$oC@=|FnVZW z7%m* zL|*D&CN^(3ZORgh+6&GL7HYZKa_NEiY(JaRJG*d@>3#8$%c}#bQ*j5)#o4)B8*x6e zZu!vyPN%=WZ200OQVU3*q^_9uha_nmU;2jEJtHmg&R;x|>TC|#hy7=Ep|K-n!;XwE zo-G}A;_`i!>3fAe$}E?0)65lsVEc)uigs7riNwP)+IWqWwIq#P zfJf4&@Z=6;+2sqk_W1@>d{z3IVUsSF0nR_%dN{oGCM{lETTGv$zpK~O={v1#+nmCg z*qd7YD*Yuyw;<953< z?>OVY2g#*#j*B5ga0b^}>`ov*-+3WECd=>V8ecM#oO#DW*%rvgcf~vF!gw8-(piCK ziKl0td~q&2_a;c$h%2ACq1pfG<+ooH9BK53HNRYPq%FxL-z0Vo^7t zp@>E~sdsb*v8&l~`P5>R<{udDk;TYzW?x&r`k6@oCGVr2!6&Y?I$@a)0q=wd-FMDT z1)Wm5WNMO|a=N3VR*Nk;I>+>`osdR!l1Y*l&lpM8wqOzCT8q@&!}P#P$MdzXWH&wa zUaB@*&MR&@H(o4lRh=K-4D%Gm$H?5Iqg=m`d4|cD;YJW&($j~x*zGtOSknW2J_T*Y zK4!ZcG4{}-mNqe5yX{&Vd)uWp$u_Dsp*d5N%MlNUI42Vu72?^XZpXcNs3Sl1NKB>v z$tsH3I0t-Ts^ zu}`Q@a?h@Q>Jzx9fmEA{yGGZV-fT`nw|xytXGzB>_K?w*F;gX3C8FOULnFgM-(H_~ z=FE(BJ$1dj8#I+}II`Pxi4Vk&5dhwmz9ewH;~K~HYELa>Sz#+|?b7)6Fx3k68uDAi zJB&hB4s(81G2{Wn;23<$!N`)ro^RCXL-Vzul3B6VrclyQ+RJchoUrW3U_W1nW9Y0# zD@`lYP16rXH$Qz7`=BPL-=SnGJDPqvLwsx-& zIx={#?S*lD`{&V}%G4_DM(uZ#OqJNB7cSLa^#?ZF7TXS_R8i$L0z5X{{*@oKDtH8V z#&t3)eSzloNmf;Z1+dk+s!YLw`Y-hzhnbi9I>s#`m2>*sbl>wpg`n<1NNVws{!!lH zl9r7MNgI>}pc-xj7kb6lVj67ra4}q$e3hIb%s1?MSXiu5r(@@XPVY|8O&&RpI6eNY z=`Ns3iHfKSG<_p>h>2RP!12X={&>H8KQZDWV(o+VhakCuYcqU@Nl{HvUFIKp>@`Ez zw8le9`brFErf19?svI&lXf{4=tjz$+-6o!`v#)P6ow*zkC(5ME#4I@@u`Pk|!7na; zPeN7=uN}~>@xSMXc1Y|AZ7=x%(GD2ngO7)Hd)64i-9@a%q12WDhU(BBTXy)YZ zWN;{c(1W{t06};=TRN$pF*?57G~^Jd;IoBRDvu zh@`GvVD5tFXye9H%%JwWyw;Iclh&`TOd-&ao)-Gf_olL)CD(lbEq5PKK4ZSX>J?we zijvosSB?;)RE~~~ED86&E^x=fqPI>p3!HcjaQx0XOCZ~eSF5oRbMq|JDdAxpB4Rnto;LRsV=TqOUt*s?bFB4 z3q6$02~pBTHqboL;)zyqpanNYC zVC36={o4&C>*NzNPuHF{m|u2L?KUXWEXTz8eZo3yT!zAZU;1V5WiK}GhS7yZuDz~b zahusq-9=tfxMUwDds$f6TaZa?&eKdj(ORwj+02KY*k^G(B4a*{-#rH|tVEO~UgKmI zRgz)8-9N7b8q+v1`g21m)q zP(*_Z>dFQ+e7w={@XO}?MN~BeW$j@reaQFSxaY%!yazKW6U37>U@OP8S#4@~$dY5A zAWK{txJtI#xYfBYEoSi8(0Aj`q1xKOF2LS^3D<)vP2O;tVJtN(^)>`}eK@x0)CJY_ zNj{Uj3|jPn%=vT%^&_H95K;5vOYFhqPnLs|Sw>@DoVxPq)zwq???3c@ zw(|H&&h1xMSMEGkJA-%q$mUA~Z(i)%**nWfyEKL{) z>-DvCa0>8(c=<`;_vck*ZBj*DG*a%m>7?+_XS^t8Jw<$mgn~upCDFe=1TImec)+|` z9*~lgxb#n!imK#y4?Jir?FPOczWLXRCoOiIPocpbOhkk~jh{JwE~Qx33$#IlrB)t} z)#MG3P;n@&P7Ao)`{kCWi0vt~T)zbx;}4eXTQi$1Y0+SqOR4_te+}x-57`Z*;h0fK z(460H%;*0Y^^XcxU`Q%MS_+z{MdysdIYIU->CG? z+IXpGX@~ZeUq-lN6{uZl5rTnXSsIV_b~f?nnUnQwqpkaduyz&}JkPG6xaP~A{Ccy8 z#B+WzaI$&qb>YhMKO5zx41(md5NsTAgpuCg$pIM_iWIif=(Ca0@fkj9y7z#G|9junyX)F5DKzl zfsI46h{sib9^SG6!OWv5kf)&=z3xxDf-teob5aj^BPZTA%|b3k>Af(}u0kAEErUry3cJ#!|L;H^Qp2 z-QaR`|G7P3M0n1pOg-MHCX7#$a^3y9XQk(ff7ya8gP(~P%Zs+OG>0=i&u40#)H~0l zGyY$?n_?$)cSpf2!?pyET%GR!FIB`8vn7%RUB-&~#axG(@}{$=^Q~2VeWv%OgcCC_{{)n<<8$u_s;}%b$LMKtQjgKoO8?jCp|AoGx+`RhuhN%< zrvBQ|Ke{6GM26M&qHO#>YV=EG7U@$FMF-le5xYp_Fwd3C~^LO_94Ksg-#lKb zKP6Ru2Ti|&rhkwoJptw4F!LK`e#4B+Z{5t_xFHn(zjQNvKK}YEf2f;wqynK1#}Gb? z^@&JQ=6hE0Wi!?APx+eP4v1+D4-tI6`7cCnRY=PwT~vi zvBy?fgEgg?fUD0WFpn-ke$vnczJKX?|Ji%5fEpo{u}00o?Sb`Ad0FHb=}n{dkiC^* zPMjyik84%^mvDIXf>OE&|FnB&#%zo*jO6rrUIw8l$JVlzt)MmJ7_uhda3ybc!)N2t zKr{oj#gV0kmjCAWvdP@EV1Jk1Z^+W(2E8v3X5U+S{_ROTdC*cr$R-q?{*0d-IQ7|j z_G@9Oy)}r}ZWqbZ@?4QOwdDE6ePX% zGy3J$qv{&*T{`cE1gVAL&Wb5~#wI<@VVCHqo~zIr&`EOWIW4NM6eMp+`Jku|70m3> zo6|KDyNDiNJ6Gh=ztW)*CP5ktL;D&q9po1XW#YX7fB3^*53^m zG241s+D^2$9i&IIr)aPWS>qM28QJ~ZQgn>tWZ0?93hT`BtXLQ-Y1~?M#CutWh0y4g zeS|&fB$;-@NX~ZWWjd_*BCE!V9-G9_=c#7JU!#hGhL$uXKO7_ZcA7RIPjWN2p6$n} zJ$UhyT2vLW0@WX9c(%{QXnKsC0;O$Zq_J!zm39MyTFmypobS75+qzfjt$$IOE-MfV zSoD%`X8Jxk=>)7Qb-;$hUk)Hc=NwDA)$LSBYqFJh*!m%N$6K#iW{#=>8dged+djl4 zr>j+6xb@_^_KM)i6!jImzx!i~9ukCT7sCD&mu7QbAOnzw_DN#&jFXZyR~10ZI{zj7 z8=>gNC`TmC|{@80C*Mgz*+KuyVlNA+&6-Gnk1vzM9Ys6c?(e)#4}UY5p; zgRB&BY2%?z$~|jm1E1Q zQ_9kZW3iC5R{1xv&@_TByfu=G8;(Bj9_(=QJL)rZ`VQwGdSz$o6v=UztK6ULi4|oa zovZIUn)6A&MTT4BXz`i0i$9yO9HE-#gNP7xL1*b=s+Ce8w4# z`$Z8&`K21=U!QQ%P7S1s3REhjQ)K7y+7{PYc+Q4=O9j<*tja!sgO>uowsn^~w}s0q zdh7ZW3)X?cwa^}`*Os{ho#tUAP-K}}`>emqlC|-X+jw!l=8yHm$o2g=!lS4Z_6Mp2 z@=|(k9MHCiqX_ci9b?(BZN7SrV$kcUTnW>xAgqACLJ6$6(sWL##w_`1y&(y-Rc;vcW6*2!n6r6VzD&P?vQ zT;tYgDA%cSN?e9FiI}3%bCqL#WYM-817R=hxCVX&a!MXW>NEdg)j_T^auPa+8a0p3 zk(2dsH@zsWb*EMMmCDM##iF9|W;5;Q$sfcKCvvKGBiU*u?1sy}e&q^-IzS|&nWd$; zKM%Id_R!l?XfWSpt1-Q2Du0BK+WA>|fnxhPFN-5f}FA59gyTW`IA)%V{X!xNiPzH`;Dvo^k}Zafw$GR6wPABnGA6d6ij;Efea}H zpR`Ehv8!ue3dwT!$PEfT6B`eR!M!ftOihcSm*%EfZm?Yo6u;?l19^@RYL>^#!&#N8e7^~0agATX1| zXsI^SUET-28%>AS>VA~=@_leCVLN1S^I3mgP?37ydLk>8;UP#Mt6c5=e959d=S!lh z6g9!N$Dh-EvZVZ4_yj7dJ%0gs!eSslQgus$d zzR$-=lP7X0lM`OyAgi?KTYY02)p-rmL2rDxb8U85RT}x)R3+n3BSoa#w8Z)cly1{P z_Cv}D%m2yhTjRx6v!#h;T=RM6&nvETO%cs&1SH8*#tLt%WMVDA;{7rIxgK3(2#5Ln z1|v16&swoHBMsCRsop_n5e@L)>gFt-$@!@7djsc)iLoqa%o;=B-DM@adg)Z#$~qD;gPzj-(S_E{Q^YI)K%QaBz1d z%z0o+!Y@~`S`D8hK9 zChnMgFjY}Az};i1N{oiEd8SzTe01F_F8>JtPQ?5#H#awDXTexzlFQONk3C(6J|@VF z;I{i{Lpds=96ry6&h{9>_;!w`5mSD5OCeIcdVWa57Ov99=lfd{qF$$;@$U;_*DJ$v zKstBrQW#KchDT4dj6)Xbr3@N#(;N=uJ@;qjFCIbS&rC=?P7sOek=X8~8Z)r%+gywh zA~QVt3Tj12A0N!+3LP;?y}uXlHZd3*PzX)ZG)y~F_vAFgS}``~1Y1CQtvwOw(QzC* z4%>X5^mKJ3JtD&cRyI{o7@XlTRc`dG2yn~@$%ngR8E5O)fFyt>oO6n(i^aR{?E672 z-xtAj8U|iDWy9pquz|%*y*JDaH^(J0$bb#;E&TFN*yb8SSWDRXvtI4p8(<80^ULXD z@>1}j&epb#Ct}7!mnq_Y6gTCVl8H|gpxR1Nzp7>0rH|2lde(+dK8Kp@^Zi+pU!k0Gwl&TaIp(7 z4(K&*oZw?b4S-r`azS_Ye8a}`p-z1dxLWu@E*Hqad!-C549+aML-wk1;}AkC(W#m; zn9$2;+JY8^Z8Shbf%dpQ3RCQ8@|8F#nvy3bQ%<38=AtMOX+YS*0W(9g;zsu#)K8eZ zemUEwq3{ncadVo=OO>U0^fPRFNqJ{C;!}7<8g2?m%FE%&?M<9isbja8z z`(kwcx?aE47|vBD{m>cB({>w{idJbCDc|V_;SaW2FkfBu%Nsi{yZ1~d@9GO|RCav{ zoQ3K-Je!u>_80@tEc)0~p*GMDgq1gXD5C(GUVY_TnQY`jd+oG^4nVV31_d1Ae1kt^ zw4p6{R2)^$Iqf$fIlE#h7b>Fgz$pU04Z_aRk`9dJrrD*p&Czf<*o>n3!y)Zdx8vl= zRZQ&pmj9yZ} z{H(${`r86bk_zVU=9iG;?LI<3s&juqN{D zd&MTmunpFsp-+gKDzh=6?Q`id_R=hvO-1;2>OD4m%}3QNsSwD!#RPP6A&04u$28tY z(A2uPP7}T87jOV|n{8j%o=;*6mGZ|~IyKOd#?Fgr>e^SaZA_OscsSW@UhKJC0xkJK z@7jTogE?X8g8VoRaOv@lBrF}0^8vc^5L+{z8aO^+c5(rJsUle!l7y=?=NZyR}!w{F`+MrUu$`8ttd}VtCQTTD1DU3bp9mG(}jjMqyTp zmG>9bJ)j%kI*g5I6T9%%L?&BMkzslJgCB+*4l~)NfOm~8QSclTr9(zgLQf=bR8H?97UVflh3eXj=xET6zmI?R{DFWED`1H zNR>0hsETZWZk!VsR>TeYk9_N-c{R3jxaG`Xx3i|EvRW1>mx*;of8Fi(-{2aMx%2~u zmnDfmWHRbImva@@>UDbqb-gwY%Gp8OFwGX@kdre`?5Rv+wv>ZqC+XC6PDR6Rft&In zF8z|Fe2%>O@Ug8gfuvz+cz^x&4ejzeI(pylum&BIV6QNnSi_{!abX|(soLbHPs$DlfX9iS*@t& zkRb*j-MIN`0AD=IT!ZJr9ex&gkUp6`Mww_na4@5N1K*beqHC498^JGqFfzQd8tKen z!h4?I03%?I zhpCX;y&GDnSUX%FFgPe+Zce7> znB2RXjko)RilNmL*&pVB?f58&$CP;u1*i8d-Uj~w(B}M+R1`J!+Asv_0vS5G+QW}W zhmjs%bx<)VqDiD#ya0bmqa{77{t z!DtOhThn<#zoc<>)p0RN6Qr~;^wzJiN_4fWwN=B-K@2M1Uz&djYf$ITmI zRbml>1_2CYREdCq&&R5F5^k*>G#*ab8e=ouc^mETaA;S%M*uUR)(HK?V5l^Qz-#}f zE;VYZ(l;6W7rwo#QYb2~qx zT}wpHw{rOYdFG)4wy|2-ds6G+9~6nLoyMY$0-TLz2-l96&Ez6xFo(vY_n((53vg#* z2R0t~WMK!qpa1mk*o>}BEI!9?~ zEBQT7y6|n=b!buT+(opeaRjzQ`xzJl-3VD2mrKwxJ@3RaVjw*&eBwXhp~gYS*sjeh z-Q^f7bxPay!@z)J@%5b}{npLdI3drEOSV!@b7>)ok>&Fu-*=1Pnfk&Jf*X9Ish4ha z)R{7T{`SKu?RQ$$1yUE7Qf^C0PeCkvtE*1mJ^V#x+Y;9}$~zLn(Z|c8 z=3n0=tHliLX8NF@1%%(KSrKSa4cOb>#Lzp~Od1CM)tA=4F~|)+n8Ot950&jZXYW(> z6ldVX>@hb`8!F@i@U5H3)~l+{L}@}z^l!>p;A z7a0LDXB6o07$VG?vL%FVjS!)diL;r~0~^t~3FRKz2;HKn2p|@>+&No1u5VaWfb}5+ z86oAF4X8Q~0;hHUNe9;$^dQ3neyNkqsgg=XEN3Pnj`2#|YUI9wWaXGcmIxX$FF0@4 z=!4w;put|=0nYM1Txszf-PeY9QCe~z!S=*MqVTQ?1m5IkC-Q4G@?d&-wbn1wLH}Sf z1L3wDMR;ue8!YivTqBH$KTPoP3@qUIsD-6Sm*BbA>@>-ljBnck%_LY!Pd)Rx_B>JV-Qy6>P)QNMzs`Xvgs51e^LoEiZW&1!59ALw*VShu>Mto27{R*lj`Go^bUmRxi_{4FAF)G!*MKittaS%j}I;3cPtI!!YlbrIFLwh^jwv;VI57%+hLx)<6+ z)7RleL;ika2jU)hWHSamUns~mQd-+$88KdUy+dlg(rI~ComWRYkcTj>)He7Efvzw8#w7rrj(mPN6>xUo~)cCho2N|glddtIA}@!8uh zF}sPWnDtCsD6L!^xet4D$G9Q$c&nf-NXyF8iGvXg{~?*?<&u*?|!T zbICe(fzJx}o%&CfLgqYaIPC&6z{)oLgg3u{I076Ly5+Y#wB%2cYw)IZw3ih+y%~1k=I3I6>}GRKz8`AnnHK9fdgNe0XVfnkGSqAh{Q&4 zJp#H)-eBB{Fr?E@vF7LqhgP%+Pr?aAaq(1Dp_Fhz#`MT1AT_1@eDvk2WG_UrTlKh=GT`Qxz1yCs|dlTeCX`FR7QsW%^rf-Ti!9E%$}}a19V%wr>&@ zl0uQpMcJrM4(^vE3l z2soi?s%1=5Z)Etn|GDJ6YriaN8mN1oeo@sEZD=&Ek5M%4vMHptp!Qozjh~I+%;F|n zh%gN80nzP9pcm9yyE&N*!Jo_}!F~8xaeS${r2d4_(QeDTFb0NmJQa@>cXffvTxTqK zvD?%<3h_NWn0Cjm=JQpbZv)x@JK8|+D1i@5Ui-!h#Ovd_#1EW0MpIt+p!_h@uv^kD zOT7;~!kGQ7ZoI2*MCq)d59dJ(I~Dz9#w(*LJeJprUXv9-9fK~l_-^;no4m#bo@Xk$ z{yI7^mvD-FT1dxYP==3LVlhT1d6Q3hnHS=u(GL~UcD|~2+m&!;IO^3G^z>w6pQc!g zlp?@A!2F-z?-wN@EMfYSCCuIbz`GaVpCo#7X;n6q%_-L~B!osMe=!!gRSXP8)?giG z`?d!pk9`&WxZ7w)G90G?t7QOBbuA7VO9@y#$6J1J?jP^pe zKS=m~N%&32Y~)cpZM)U?4f?6AF{7tU{hEo6(R8jCuoX<*%6=1&=i(CPQ(I(i&=4$hA)S?6l%XL-p z>ad?jJAo?NUtAPV)Rvy(&R@wPgoHO>n8jpW-W`RFgS74<3iLGxv(&q8HVWGzv<4!+ z+w_J2vc~A{;W(*7t9rNB4vVUh#cL~YwRz7bQ8i3{ z^7zPYwM4p3h(KCf1lu0mBdgcGe5-a|wIKZFRr<~|gGU&sa8XKIxe28$4v(L*E#>SU z91mM7Irb^n?m=-S@Q0r;APNj6XxjcNLHwBW=8GF0w}yJk*iuUck_l8KCxhEgxq713 zJiNbO+#xIakzPHY1%Iq619na^q?^5hgXN(0EN!UE0N#F7EiTc+cYe)~)C1<`p^!$M z%*>#nQcZ0V*WILR)qJ+Khd`&%`FO3EgDhk7qO)j0hEC=po1D_TD}_~DCk{{VzJYCN zWfii*K!n1y#kQnQPK&ROc|K8VDp6G z)EHG_aHnh8y$uL~`@I+xO~|DNZ4NBdg4pjTrQFh8Qw9!zF6nS(cKS{%ZG^!KWbv{e?<=;6jZg@G@RxnUq3`nBFr}Y;YXLRzG7ZM+^RI@RQKLECe1*& zz!hrxjek1#!Y+i<;y^+Jp7(X<%`QvsIG!YK+M8klno$a$R@lZ&KAgB&^ZssjcWoCo z*n`?lOf1ryzQ1Hf8QM*4qPEz^2&uL^>rIHv?(7d|ZT6_CEHIOZu-RZLRbH~qY1d)p z7J?6~@fgPi=g$m?p7gWMaBKP2iF?n-SJtS^V5I?w7wnkLyr%6}9g5;`n#p3`8BV6f zwVM`IEd{{aPV%lHY5{)Uo?>3{#;Q7Vc1>Y8oTYJM3`E;`27Mi+dd`c znjEEw^LPbkQ-Euq){e4|gqTw9mv=HyC+j$Oo2buv_-fA?%QGa2>OR6L(It^}#0FXM z{53JNOcrQ?>j!~1kTq~#hS$+6Z%t!GbqX}K2?VWj!t28fvvd7HUsmQ|%oj8$%t=(avonJFqYz{LM zii&9OWO8vG#qJUUEp;lRfX3sv7G80r7NufDw^I*Wgnbx?Q?dvRuZnBD!Q8g{sNi_8 zpMR}peZBf3!EO919tYR~s5%>1S|Qjf4Ez^Ijpx#U$yv{h85T-aB}LmN`3WSCcw|5& zB4fM3an;%~ScKHp+iR!1Qx_Vk-qT70r%dnCeSwoZUsm`fpb%8>$dRj0D2z?Xzs4 zK;|2lKO*6#`z&D5&7jNDy^SN13HQ~VUTRh?wDu(gYC@zIKlb{WE;0ie^H-}8nE^Lk zbjc!hDJ6tABDwSIeOX@02UMtq?b8Cvrc|Fbdh3wOO5b;^Q66U}6sG?E1c=yUp zCZ0)#pojdg$43=?#bMMSeVzEW>jSo5aNA*7<0&&j)e+5ohHODmu_~}J^~|>GHK^KF zOiMvw5oHv0Pe6k9y}6?Lz6#F`F4RLp0RNR~^smPZ!Qb9sXEAu{1)2JB%CCdZS6IHD z*b~-l2iz3D8$>RVO(9$aYSUSFsZ<72`(WWhr_g%?)(0EyUoG~#>_!IhWt1zc>fx~)pXz;=!fI8WQePna zuR;%$|C)f3IP-Q_+>oKtWwo0#G$_E^h#+e?d~4eBBE7qODy6ir1T571q+`n!(~cRT zC{ayFFeTO*tALg?T%}~MtZuK5N?@xmVppNK0>N6gKaWJ#Oq*=?p6MwpJy7&KKEg!x zP3<&%+t)p0Swwe8{gXboSA;!i3HNrYSG6?e31UII?v+%`x=m;SCx5FaY*C*Ge$Z8UrBcj>yA}=iinQa+aOQZCAUofene$XND(77Fl4)!1{7@p)1)RJe3(Sf88cZ zc>0Z(2LE7<&)qKtFiK0J{JpW#lgO-*&s)iYHp>^MU~#m}KZz8eAOD0~NQ@KEjb_q6 zI4f0YOn+xy7s2dwo%1yqOm=jV#CX6oC~b(^5>@EM^>`^MN$YIzn8QAti6 z5jK_O9c|FDHlKN-0ue=-vFG-iwi-~2d|g7mF7ac}_8Q0K&Op#;a(80h3+Ri-u^R>Z z-lK_6?86n_Ih~u=2~tt(Tceq`V<1eQ%F>sHxqc^_>^@=W$66hbqlq=US&=8oXDaWeun?O#R2jWif;cf2M)IA~VKC;}(!$5DBjq>CwJ z%a67|cb#(e4SpKvh|ws}J&!DEKh%Qw%N>x;K3{_1VD! zRa~j)C&hL5s6OGIbxAzWN$MsQYGPjHQ#J8EDs*3BREI{#>K)&)@3#8AYvOzDk@Xu- zza-SN(lC|>YK2cp~w@EB?|jI4Xw_ggz{m}h#d z)*Nm77Euc*^wL_mtl$Wu%-A-p4pG9G9nQ#G**lba)B)IV5(@Y3sp#QWtLjbz3dyq; zBEYLW75~8i67&H&r_ir(KtG>3GhmY-&vr14p9&$-?S3F_G4T$Dnt-E1Lmr{RLvaqj&ftln zoUx-Wo6^|7GL`n!R!b<~V#WdLQopF{0N{0LDeNN?d%?B!BX2BbGMqdFh+k}NfzS<;gUsp*cV6Hpw zNo<7*CG%?Z;=8a4NnyC}um-b~t;Awj_fhe#NrQ$*rVU+IoK59Rm6VCSfsoj6c{f*I zKzo7@I?eD{r_g{;lPg)s+Id%78}7YU-0NasSN=#~t%r+tz!N>!1Ash-6|Iazv31|N z&W=UKj736)ouj?&)6kU-Fi|K$trBQ)qSh+%=p)CzqO!WfNsgvyGD1yRtf=-2G3n<9l#SC;x8ONc6*l`M>+r7yA@d|V6|$Y9Qt99n zv^!onm0OwMP>?CZX}|mlu2x#Ry83|oc9at#Z(M-ATVP{)cwGU9n?^{1L<{F@g<)M{ z#r1B;7)T7mu3h^_rZqA~9ZW6Ko`RDy!l`MCy<;Ktl72&D{i=5r8Z-~1Gjz)ua|ow$ zIXM{!bYerRImj?o)W>eUP-DFuR$hOXiMsP$H9J@n^;XOj`|{N30;XIIyNam^=!$^F z3nZY)*t?^4G7fr8Ruu!+UflC{7!e%##i7*wwFa7*{S>#s{7jfNB&@yCwqJgTVNNYers}8B9x7LaHGEGWWIY zTPS@n0om^r)og$Y$SDflqqh#Xv@eGWQAFQC6>^OsvLN29Hn#niGBHP%ed(ibdBqxI z5K_koa4Ycb>7N`havBOim-G*Enl%KpB&p^!Zs=pLM?(MN1?WW^Xu_O_v|6B!w}W$A zFOwji*qoSG4UWGlPwk(plrdGX!3h>Ew`nK`fK%%12}%Vrf|P7vI(G{~AWp|Vsc~8I zrwA5$D^+qy3$M=9C<`BHG;uIiH#PAud7AL|OISZ&jHasEkPp*;9s3u`Z#jv$50_CI zPfuw%7A;WQpDtQyQgeZ-u#@i2b&vVT%za6>^$EhixS-Wf_tMm0uF#n<&tnv*$>B20 zzncr_$9qAM#~=rtN`^#jS?vdUx>VXw`T%;g)yl_fPFoR=Xzh>ay57?2n9dV2`&Z}e z$`%#VpU%b0%e~Rm)Ia{-diicm481UKQTgHPUi;BIrVnhCx&kZ~y7=xWZ2l0L=4JWo z6niNiwYmH9x|&PeF_6)n^WmX|g%H^LtkTtQmi+~|bzUK+d$)Y%Qf<)aIe#=pp4sne zhiZPhpGWgq=<%!i9Vf95)3AYCFj))3UsvY#MO(_2p-0hW+W7wyJ-=hgQN3JJA5c_F# z4naga!4g?mjSAcU z(TM)J-{0yGydwCIj2wE=u?GH3)M%q$97QIM+-xc+Zx$g!5VW$NJjV|Ako4y+8mFmC#oMZO0Knr~T_;OVn$j7|>28RV3zu=UIhZTGn zoYxWDJ+C5iTfnIzAmN4gI={SnOa?VdGd;saLvGG|+P@>!k~*#G0JAUD-!%o>hi=a1 zL=?4(84Hbt-?P`*qwC!t`fSML;hV-7C0^|=Hzrcvsa!KnxzV^?LdcyXA{&Hv8YBT% z*1meaXzbd!u4=R91t{O2yV2p-!$B%E_cq2m`ta35fgh9vZLg1jD6H0kQqbhRLJYc)TFAv(##F23r zkSRZ%JUKZ7FA3hYfu&Z1)%wGTh)Gyp%ABXXwL4t?NmWB*LUR%zvsDP$URLn;?nF9I zd8l6UmO5%bhyp;~=sGmbu;m&6#47Y|nNO6x+ZseMmHAs=)-aU!3|N{(Ec8|mB+?q< zqXf5Jl$lCk0vfE!k#tBF;o`);u2nB-eYZ(Cw#x_JhyfRZIj3n)?)_r`e~bb{$zos> z07EizisNfd*>Buv_J3w!-~LtouI-+*Uv+~J4wxqK-ijT=1f6Of&RTp2nK@ z1968nm>w^LAzQ!^6wW9fSG|+)IWZN|jKjtB!-rhlG4tXr12G$Tvf6+%6f#7o{^w72 z)tUMslrP?5V)z>U?F`lLsR)DDRcm9S2K$!-Je#bXb<5^ZN968$`Pd8#&P%%!OJjxk zMfb{OeYTc%R)*g~^|U&xYAjs<%VP@yD&jpoJ&R~g{PJhTSgF;=ok)E*Q{6V?P3v)(v@NR_;ByGU7_Mhd5xmkC zweQz|47NT)Mp?OdUQ*nR(xqs*oPUsibhN+n-Rf+4dAY{jwfQ(n);dv9CZ0n>J+MiL zo_D+Y`@p)VU*B{mE}#PZ6#B)sqqA@JTl-9{C|U+KEmUx^2W&4rRWTk-6#!-yZ#X7C zOnaMP0_}K$CEbLHO9w51G5y8m)c>;5ANC6DXP0UGa^NK>Y%n+^lG@MYG@1kwHyA-N z`@dlFU%&i5V_5HzID8uOU3P0hEx_I0IiRDht?P2YX28a7wE=;E*Q?tf;#wg1>Y$De zIY=Jr+lMeR20ocsi8;r*{r99fnxzK*Yhp-}k^5yHD40rES#bz4<19js6d2fg#Qkch?o2YTuEV8-CD%~_rJ<$Nynq2~&_MNDBEJLQJHn>2K5g%5 zcyVJwtsglKlm-3HLEpd7)T~*xCiYrYO3>8D28=B*W}=_xk$&y%qVQAg7sYa4O9vrd zgxF^&eb?LOcGJ*6RoG+Qxj33Op}8p1jK>!J(_UkH@8?WVvJ3e)<6d_sava*B3h1Zq zI4~7jMLGN@JMQ6X~`+-5QcXC~GoH!sn z*$S4bZx{?fZoNT&UjOJv7Okc%qQDADcusfC9vNZfyKoU31+5 zP><-y$jTU<_X*N+2USlB zzj^YcRZdQcS%+qFG=y?(bA-yap}+vyV)=?wP_Rr#AuTm^ZrCIQUO!l%$5uLVkmdD# zT8`YZzr>#r`s5uM<6V=M4;LT-gWi}i$Sj%9%^`i446p4rgxMf$q{Mp18qu=y0fsmn zsN1YLx4xw(=X#BjGJy3rkLcyTOpk(+u?MH!uuQlK*N<}TW`8xh$CkI-_{7YaupsJ3 zca<|8n(#iJg=5|=DrW2Ao^k@vskm<1JH8sL2`>*F9FaN}E5`9LGc)8>Esc~e>7Lls z3oz=0Bd7=-#(zVu6IZ~v;sHU}_EhG&vI`U|DbAe{MWcdSCvS1x{)?sm?c}&tO5Pzd ztF-t)$x%sxHLRS!Zd&DVLt(_4yeFZZt7429dSEJoaU6>_KDd0qfs)uNQ}7mVom~G2 z_m6u@r)qLbGfsGjK`KGC9CLtKMqSEERd|O*;Ng#AwVx{mHj^i|qp^ zp`$|$$@fL)I$In16N$1KvrAHStBa~yMp!3ujAxDAxJNRkfW*PUVKd&-$BZO=2|XAk zHOgei$c{ResH#wzAFxmYa$ok1w7aYVOn$nMo2I@n&c)v}t(NUtjI2)yuGw$bf7o{J zIX&>Q$655IptQ=s(|m8Y?_0NoQkg3wx4qPqO>gnho2D34CiE4zqn;gzN84^S6^%sc#xy zsd9V3JZnnZ(7~J5K

  • +i8G{vCHyCBZ{Vk~EQbZZ|`@#xIBO_U%#{^{ft8L@AYch}4RskFN_=bRZf{d8U`0o2rxr?wo!dgh=CCsr_cToZiqAJH5 zKf{_!r&(1X$`kz~mls4j9jCDBEG!wC-Az_Uk?awPLjJW}?Cc5!)iaLn(<2t9UtSSI zTJp7vdmTKp>9Vf-OG#{e(zim+7M>RK$HA^pG>1}v%rl(D0Z|Z8%;Z>Ti)Xd|)nBF2 zMRC)xXp)P3MyqD_d@x^`P<>r4B-GnmedQspP{qOx{^{7-8S zaEMlimK$Jjg@$cIRHzhjc*Azl=SW;+BQh&+~C$J@eN^)JXe^2D)r8^9J08DH%+cB z)V{9DU^X?ZS`K{yBpp%d<2DOQmEzDhn|pti&uZp$@&!xhsaC&tWHAvcqR)uCG1T(~ zVm`NtXYF2=4-7~G+o|x~g35KG()A-fjG|Nx!LR5XFbPXKAl4aejxGIRmY!355;3U; zk2}o|_|z^`(Z1Xd)+s$iPI-L)%72Q1ZbcbnOc+tYzJ%4At!j?)%!EWe9Cj;@-PaJS z5s5P6(7X?Z*l#HxCJN3n+zIocW;JcWSiAmZafzs2qX={*8Wiz=oE-7h2X8^9RTRjR#Dz07xkg_yY?IrQj zb$Zr-0i%;0^8ZQ-eQ=Q(5`q|oSQv-&;{=MG7yoXDl3~YX)a_tc1O0U2~Ox!FfDg?PN$sk zoXqM;z)e(8!vEpZ+|RtH-jl3ZXuNcrk}WTR6zYa*RH=#a!2&(`m%aKbQhJWkTOn;pv@tdKBqIeGeq*(+`cG6h zvI%GFD0H%v!T1YzLPR&ux6}+*T1AL&WU&Dj!ZZyb8MbU5J@H&NV*lH6DGf|ww<7t= znom7b&Bcqh)Gr&a`YWiahL`69Q8k})>+{%q-U-{w5d#>D6>wo?o#Y43t)sCdiA1%` zLjnT5`xWeB>3E|*wJ_?`6E0N&5<@XPMwX4i6>S2+D;i`v{KPd zFGYy89jzb75_5C!?r*59U&Cn6s{?Dy$jrtjhADv)2m?cNYg^nOhb9wPXm4K73=9l> zL}5{v((Dt61DklzkvJ z;ruRj2CHMYozQYRsm<(3 zQNWVxBk<002yY^Vj?~`wv{x-9{Cp=x8)i-+JX?#4>!Am&`ND)&2cv8{!La>)HDw5_ z-gy45`xvv)yWV8fz?VVrsD2RmpvsQmk$ivVgSoPF*r!iZIJDo+Xx}4EAAbFM;z9i8 zvX9i(C>%!$V&%T`GU@G*#?T~gseQoLBve$Vh0Ja*j8mt&bKqKXqW09NY>-!hq#(7EqT z4ObhfH48$$SDM%nc<shs!G_cX`(fuXkC-+ZqTWHGo<@rAw+yXz9xWwX0L68XyI z$Z4@F<`Pf6dJ>fQ_7#13bfGr;(ZQ5V3;b~??R=$yFfvgX1-A#%TyHs*5)EhV4IF2A zE^f(z)ZaA^M1;9vM;Gd_nmlyEL4TK4x7F$Q)Xu^8bu=`}QDRzJT5R@&EpRf1-oX!f zeYJ2Fot)&Q3t30cJB^?pU;sD7J!Ie`)BULc^_79Hpnm0Is33j@akP~)=Dd}ky1|c1 z<-c#KuYT@~ekmk>kvrP&e3_P)M5|db-7kjzkI6O+L^QhMW2sue{KkLpRBq&7kVq`E~Nqn0IxJ*X1w%V)iV z=Meqc!g1sVLW<+X_xHyU+ge>{$;Qr-Wp$vHjcjh!eqnE|*QOgI`f8#ghzh@EA+vZb zyI?ms&FA?{Es-^IIl+~tD$fp8SvMgI0Wz`pp5$p2!ahy1B$`M>peel}Vh@Xsy`rUw z?~;%t;}2bvPEoqBN0%dKuX-wm-vasZQm4F!%vt=|d)qOkH9i~sRVynt8R!>qHIgKK zdvZ+|xL>i!V?F*0oPtb_%jb!{?TM~q+@fl^YCW;~KHfNq-tO+HiTr?kpt2pTph(Xm zcf;Y+twF+*)Yyr90Ds^c5Ph8*+MOo@THd{P~SK986WQQ>^4|aaZ z<&{8f7SSxNOR>@bjwd!eEQwY0R3U_Xi?fHfCp;{gd)eG4E`aB;Tm>rmRyluc-`XAY z*X=Q^(VZ4XpM=Z&i7VDNHfGb`?9Dd9GX_f*>jW&^dmAtC4O}kE;xD{U8Em$)F9;E{ zo|CtM00&eEi)xtZoj-}WYD6|&lKuMcTj$CN-;hJGWY&t>jRnqC4t}PFl_7Oi;5dq* ze4N6Yjr8vO$65I_g~|pjt%FuYWctmlY{X2hQeD8%;T*CCw%72I-~IKn2P^P)yW#_v zx(zhP#DI?dyDy{IdNw1k{isPMMvuL#2>cm|Z8 zW{nT(MiWve84n0$c`08Ozq^7E-;61S=!ETH`rgYdl6aQ}UO7eIvBnyUe@$Y^y7% zck%Xw;v0_NpU_BoCp5M+#aGUfa}b9a0-7bMk?-CEr(*~spL{buV(3n9-9jRshDfYm zrxCxSqN=Kv9wYP31Mt(7GAMb}JqO59=~cPy*J9EpfPo&;lbWB4IAje+9L6BpVvL|* z!UqKf<*-CC*Y~U|NBcQaZLdxcz>Wsf78RIx+7AF@jaEp35?_xg3wK!k0MFi+eW@%q zAqeF>2zCX%REl_-NUpQ<<#VRmfQpX?z)Pb>Y>$SpCVL$9t`=^dSxjbYIU`@@1<=XN zRtLe7{;RD<9a^xWdN0Vq$h~pq;3TKPSa{t*$DWCg+c$VudFt0B+{97L&ODlGN$r$6 zqJGgXZ#F@#&a6Z0J+4*L1JuPeLO4)}rWq(ID(cRqlia0UwE_XJmzqH(y`#Ov(GzP-=L>sMA_C(u?P3f)lAO?i=>$Ou?t7u-Vw zC56@R6Tz6Y_-;8vbm-N2p7z)b;lav1BaDYECDrJ@XZr|VzlCSI+}(p4JJCkGT)za5 zZ|(v~)?(j-(pn4i9akS}1GoC6o7$fvJGiUlB?SnPhh68hiqB#98g}W$nAya{A{TOI z<)raT64nQUQ*jv@ z*F;*W#ZTI+(DgnVFgPNH=7S7+?vHC3`7qfG7Hq}0=I*Rl5`-f9>)I3*2VAhJD6P?+ z!sqiUDiU1&^d9d`X7-C&t#(EDbNszgB2mv5Efv;!mc49k7=76z*xH>?_$0lE=t-|F zu1R`t_)aNQV0&r;Iqc7`OUo0 zu&&niZMk<_6LlMjRHft;ua+Ui>Jd4w;Ph6K5%C+FcAu2+5G|LooV<-y)xrvg*=D+e zcl?I(I%ZO;LiX!@P=dtPW&D!s1e?*~PDd_|xL$(vv_E2V)~m;j&(qIJ{sIJxCIhhL zE&yJMDnq&kv|?#a&FF5ETc0Zhk{WlXK|RO)sG(Mt7=z{euk?eIXBMYkG`tIJR*Y9u zzE@f!N>4JRMg&%xgIsd6PiXYOFN-~w%bJ93>3vA-j+xgFsX*^~D8to0yPk{-MkASHeqz(6Hk1ZdRd zZg>#NwdOWExWg8h7p-g1BdfhP6T23=O3wvlDksi3Nle?_CwOKlq@(~@t6q9>G4Ggp zjVW{jpV@+r#zOmabRPZy2u({d6zxc!pz9HLUlZ*`R+T68Mc*#>o_!^bqeElPzp>}Y^ z*fj`j+QN05UDQuMyU7#M`7|`m5PxxLEkVo3y-NJ$q>bQy9}@?G~HKukgDQ z3Q2KZ<5?=2LCR7u4)4c`nmK+E8nhq2{t|&S-=6|ba9Ay<)18@}JvHcp&OymUP#Ad- zbwu12E@#~x6skko6jt4xOFqMEJ_w>0|GsaL6#ZGoA84yTbqMbYLEcK)IP&z_QL zkT)o1^_L>HNVTo`zztMA(hDT|A!vR(pFx_ej$q+K>D1KJht*?ELDcyCAsrtGzU~&} z*j9o$0ce}f&q?&y68ILd-Rok0D!Jh=CrhE2s8~d!W6)x0ylm9@UeERDkef!*_LEKN zB+46#*F6!YV}B@9@AS>VdsPI}{RQR3de0HZAmnP=ZP7`SxtQj2Dy& z-KerSWZhp`a@mg;9z~8CXt`mb$zC=HT)lDM~iExF;@*^@dnQ)szZy@l{eQZR--es`I{R)g;Ts}Nga?MbkG$DT*Z>T|}w4?`M2 z$$=uI@Zpq4-{J~;&)fqaS(6i@X)BXQ2jTk0tn!x1J)e4cKnQMnPLAP6C*va3w6um2 z`d&UcclG|v_l7x_KGZr~&yLjU_GQd!7L>@~z4X9+W8Y$~m5G#Wc!;~VLd1vYr56F5 z!BCzr>^eU^d8Dk&8)3P}@p-#JJs$9D+7ljIf)Q@s^`_=$L$|DYt5~~^cIK(`zpgtk z^go~9RN@JvP9l_kxNliGnd`w7cO{Q-weTL{w!sDGQ}2)V1y`SlDJm-JnN2mHWAxAl zY8XSYcOKRc?uLvSJvZJEo&LC6N2-~cmYM;F3q3DK?v{#8brn$@NBlCr3?0#hkRwWu z&ZViN@74`Av+6&}!p?VnOG+GdN^Em1zc5Srlny*#=jWGURf)UiTy`&uZoJASW@*W6 zv;zBaCY;=cwc9y@GcPR$l*q@Do)V>2qde-&`bK1Hp<^&#FC#m9_~1-U;l3Go>J^di ziFJ0(yHaGjT$*P=`q=n!6Ou!_6ISNl`{WJXr>Qe4XyOq?41=Nyy5kEq3b& zdaGh`Lk-7Da^ST`6a$J&a}$H%Z$%@Ds!apjqs@g3ph-ZbC_FCC*xckglm88Ie%Q!0 zT}S%FY3&KmBNg^uY~2xm@iE{DY?*bgBCjS|x-Z#bU(dZQrtfaj1a$c5S&*9%wda>H z(&MyS0mxDJe4d7d9DBCRp{%w{Y%M+s!|KfX`dQjzg%>YA2Vf}`0w@c|Rtx&Nhi_8O zQ>v=!-W7GMMT93La555B4SZ_!%se$TqF}mpx)XD ztHc7uu)mpCXciGbW24{i;+^lxpyo`*D_M*=+AfLv7tE+6(1)A>eRS9nGRwb_XYPR3M;(a*cx&dSF zIB++jwc-fMGit!f&dzRrrBZyW&gGM-X}Uj@7Ig6t(qp`eFB@iN6w-yBP?JT3Hiyt~ znY4sN4gcoXb4K*DD6n45fns(7#d09K{<)E{1mGmSMj{`8+R!8GmG2gLbm>+BXIjV} zQO(>HwVVcc@^|Mmc?vi}>RY_9LQzmv4Gr*aO#FQZg+6otjClL%p18}+FSKGt`jZ?y z31Y1wiDG1jJ}!Q?BPC6TJ_0>nSo8P0@kzEoDx9E`DYl7P=iZum$pk2Y)fw?wC_oNE zx3s=x`W;u_^1q==orcI{d_rIrDctanqi?0lkloGAT>bN4X{u?gc@K;5$D?XqVR`cip=* zv|0n$yydi@7kQl-$%TGTD4cKkE^#~aj9vU`?VV5gCe}a{h~}Oa6*PTWE`DN*RKEnX zKSp9n4g@f3$@U12;NV~o+dUW62F5K6`QFfyS>ht4?~=2Xi;Ta#t~(K6$A#x#ElF3$y^xawV<3 z+BXJi%`MD5 zs(B zay}WSUg0t-H0q*~EHnT|UOZ`_^Bge#z1>q{Vy;fJOG?q4E_1Zzol;GP0|HZs z%k8?2`pLt%M}zWdy$r{4dX*-v(volcXAx91`%CTq7FbKj;hS4Pa6mEGD9J1mQH}=Sd@F7MoXJ3~S61FXX6iuUr6>NeZ+@SFX#JO3_uUJ3 z0Pe@lo}M21qi+j*f1A9AE?K*%6BS2L{}+ z(ZHE`0kO}Z>cc;bUDWBlArmc+xUrvsn}Pmw!U2h@s;YHO$MSI%$hYVhfa=F}^JjFK zlp&i6Y;BW7Vs7X-kf_hS4&n~WQDg1kZMnha?srzf!y_)8IZIyF?Cr6ILMDpMNBA7c z)yMJ=*;^!z+SQHR5bMr@3JDlZeY=E=88&^n^?ptst<$?zQ z{1=U*x2^hYS7A512hIQ`a|{qDk-LCE$L1z1*G*c^Kp<^BD13Z~JU`Nt;z9~*e;W|4Qp*_{0R(w%XW zlNZn35pr?&r2|NF$GZz|H@e_d7?*Mo>Qie6-Cw(0@}hI8XA{U>%Auho*Sn3l_lWKf zLQI8!+&2+0AcM%A+DhG@YFqxL$|9NNK=xnEGmz}vw|{L7e+=TR2=s(GJ?$A+R8=MpG=NtwF~jilUx=Bzn6fbFv? zY-_s_)zsYB3iN)k48=WQH8C;K8)G2(v4q_{XXH&WODgYI|Kc3~Vh8@*6e+={kz$Ls zZEbDQLx!F*RGf#J&O(DPJN6t*045iLZBk^~ZF!{BqS9ZD@z0C?ypbif@6spBwyrh@ zaaVo)gjtBt?;wOTk8kIP%kG<@p&dayN3xQ8-_hW$jTRnd#TpZ&lURnE|t=s#Dc64#fQ>rnnBcDB98@K4I;K$lS-+ z1^!V3|1*EtB;T&yoUj{Ibo@g<|E{yn zx+gHto%S z&%bWY|9q6jU0nkBk?9#3F3EHY%0FD|d6`$!^xwXu-uz5~BjjbYEFSxbL zcv!S3sq=Wk_#7?Le&xUa!-6l7o*ZY|`Ya}{u>ONY=&MPk?KpWi2d$5h?D*=F#JK_W zs5&0-mv*WAnAfO4*gICDM_GEPQn6t5MliU9KuBB%Gud6gTh+4bQ0O{;W&%-g=v*cI z>!YfprCmd&%LXFGh{+Fi2rWB<%V!fujoumFOUYLr=R%gQO_!@y!F!>FYTm-is_)pU zhjjk0`uV$Bf1`Z9H1@J7JM<4%SN8f#eH}jC^s8Oo{L&js1-gi_ix!_1ejUy47mFgK zrxVC=Ak(fVPCTW55AXi5?8gu2<~J^B1~Alvtkk^k!xf6p%p0SAm(MUDI!xhSQ) zRL;@6?R2(OzLHn_UsvgiOp05N-H2Yr5IS5$V%&r*=wAbmzdw{-0|*>K4ooY#f36o0 ztTt^a!tj`@<1!_zeJ2X_B)sgX{$*bP#Px3?_*qL2=>_MrZ!Xb<1pc>=lNP%Oq?;jo z3ABIQ@2PKlmxgL--rzTR`uN+=HV!F*h5yJ!|8B$>eWABmEF4JK)pQ~rIZul%H^$H;Mm zow1EQW7q$=`TwCT^!k9vNQtj1#s3f)`eZqai)vLtZ*^Pat$)xjGYv1}QX{Fd- z_TLxg|lj|^P+Tv$M_g?TkI8OV0`3?_dP=>OBZLrI6z z2AYmtHI-}?Va%io_l)X5qZh($%_T^(EcMQqXZq)Y2Qs^Ktzgz4-}u(CJQ(E%XTXd4 z^rwAwn788&I_1e0dmrZNZDZ`L?hV0M*m)k1JI;=`U`A_HAW#F&Ge;aksI+oh)W-WM37~S_8|(i2{&mXRJka8TFMt8bw+jTsZ|-O zR!KIA?neRv_k{#C$G$Ta@>#d>y# zeP;s0ge}a>h9kpt!!D<@Fg&+wK=TG?b0bN_yV$+J0Hw2m=U7Pjie)MAo1EoME~?eJ zW9amj-Ne}GDm}%RsE$E0h~hxB^$3v1GRIZwtZzVBz_um%Nj4e9kU|%zEU-x(x~fY= zNAyayW}IeNZWUHA#}&eIthkgXxQvY0#Uwex+si#+z^R;$ADGFMBGacYgiCVE&Xn%9 z#CkNl!X9bHyQ<7Wi^gj{=RSi&OOzC2B9*`|^4IeDA?IeImRToagqlaj=NZ>ZGqVQ+ z)1tYBFw_GIP}#ah<|cXD#~&>5(9WI8wb$kIXhh-HM)O61x;`2TRWPc;S}WK#sZQqZ zAU5o*lVq@h#V6OBxIM`N*2bFP?G1Kvn}QAv0S3fQLt?a2mvY2uyKmtRI;?QZB-|IZ zQqWp$PQmJRLO-O!N7r&@j|Csm(zW{0(zxS*VOK;z#se3;1w-=v^aRc(C+G{^QArXd zS-HC86rGc)o})+OV+;bGxb>*bK*byY%k7^t4G4B23adC3=5wuAWx_SvJ2+Ul3u2gA zSmi?*J_Airk@#GD%a0|S7niNjAO+>{?Ck8Pq?dA#hFf@*w;P3p_8+&)t^p{J3Bm4a zJ4l`l&8 zbPiopuN2~Rb;Q{6`7xAUhI2_kd+B>tC>6cDbZ}s3_=Cmy_7v>G>9y|*qJvv%V%bq# zH>_eAt>U=laVu&Rls))Ia z;6o|KxnNum`dv->*1#oZgOCxMXo(<0?7FPkAS)rw3_)?w+I{JqJZ#xb{vQCNFOo<4 zM#DZ0RTsMpc4;>1{>sF3`BObrHXEZ(;Zp#oa8IB)mA3$c^`xIG7hO6ehG!RZhn9GP z-@^9B(FDCG$(xe)on^*;w0%0n@4BvQ-7fE0QRYt+T=J`>R@+z1X5&XqTl{OhjHPvR zGE+1asWI>sd%lf^V%Hn}FyG&Hsx^sLb=N*IYwGcyjbd%nBC_=I!?WgMD<@L;9hpVB z9qC8Fi#7{^U87}T`{>I7d_#F=Tb^R34IS~&L-E~2_90sZ9VWun#n@m=IBq=h+ukF!}^7`;9kpc9f_dknVm%R!7eGyhTpAl=%eUG zJ9fe17f}BlqIgh!!O;1jlee7~eKl;IX9ERVfOV~Q;F4aQNovyxs#A{M<8N3Mh%NrZY3P@@2UuY|1r!eb(|6(08&8nvOA|2PT%rDm*Io1)714 ziHz1-%w9=TvvB-ohgJb3d9}5HEV^_1`&n_vM5D>%0h_^&+A~3qlAR0;)RL({$n}UQ zc7jvs{QCy;xJJ#hE)G>20l)fkr_J#WiAqFN)j>->?4J?(mhuHO3a zAyIkj!)AW$_1B56sCN>4*aa58xn5ts?~!ZI2l5G`9jn0A-xb^-qS|8e+ysoZ0)Jv} z^SIEM$cxeI+y?`S9G{1CNKjzWG=GS5KQ?}PXkp)3d<)j`hXwwR?Q`}zF>@V>$_BD6 z=EQZz6d3M{TL8AyXW{808ed`EX+8qY%OBA?HDI$LNzu{sMxFkOnSrGLQ|x2viDLc7V^i^7OV7G2o>8R5(E8`*4@lz0g_k%p z={mo7*VJ>azYKsS4dryuB7&8$W-N>AeEj}rIT!YqehZAdCceASJsUNUNeq-5QWyy& zoVt2e)SorM?r@cgUC2))q{_@q9-HOrDcX@2t%0(Q#9%So6X1Nd&?k;#>*#E5-NE?A z(dQeE^y2mdCad|eW(F6^#1MDFqZX1}9&7q!^mKaHtPuxTFF>;A7={GIV2A_TKbpu3 z*uL-L89Q!7yj)hdwpk@*R^pDD4Uv2zP<5}wDc~SU!@`Kr5^vv;>9t5aX=X;bxx8r9 zA-oWWk|abHMX=M+(#;ddj{0(O;wWa{dVHbE6V)Q<74ka0#21!ew8_9LuEZnIz;D$T4}ICdB)P~pO=THZPmtFn|dA6KghCWC_)(31P8#Z5%vLKTIg z1jlkY^))IzfGu3WtAJ#h>yHR2HOKD2s8>pmhc)KoXZL6Fs!E(<2)I}55zWFJVJ`7i zxFL(W*@&^#YF3FX){2#(il7*sp~5!wwSUB>mPtT<)#WsM=lDOqb~m5GZ9}hltKSXw z^g7b+HmqPY3%^(zq7o755SCn3D!880WI=>}~#?RDfz3HcUESA2^nX+z7B#(Bnd}rDvUtx~u*P*^Uh$%*X<_`t}C)eag zv%USrx3N_Z3=Cu~T!)>MkBTgWT<95dP?Y9WQ;EsLNAD@k%p>J# zV=e3aMANKB0x#zIeYSy|ghsC>jWEwI(1a46O(DT(5!)sv^WES~y_m)q0pXs;M$3ux zY@VPj9;jA(u*$}vsw-O&!!wD52<@(58)NvYfyTK+5++=v_n^~C6a$CGs10n_S#$7i z-_>VIX|Z2#)ib6cbgv)&;5Ge`8A_*goym3WZ$nlDr5B`F4TcLL&SNR!F0k~OXUZq1 zb*{a^v)OGI^h^&1nC$s5zukJG-*kB?gT=|?`~5i9XzR|$CaeXi3a>gT+)tQu&7S3Yj_Bv`xr?_5mcMe;OOgmNQvaJfQei%(~$`^~yDTYTn zw6QW!+poVURJY9ySEPv8`Y=N2Wj7PG_>QU&CUEM2Yp;zDZ=1=wej}N}qUjn1=!uJp zcNC=-3~e{#bk5jsKJTvSdT@~0Cc%Wct^Y#Sl32~otO1-^GdPWnR0#9Q3CoL%q-74% zI7h92Fh$FLnCo_w)8&w3q6Vvk7{mrd#9z-KEnB=x8)}4Kd)($_9Bvy22Kp#TRU%Ed zxP|%!u6Ic6UPn|-2D844b#jgtq$tjl_@m^N;(2~g$#|#MlTd+gIoeNFS6>gq#x*t# z-i#P2U(U_7pWQ;F!{L8gG7!Ju(5BtuYS$d!|zBEiJ=uR=HBP$1aL`G_!+wW zrOD)THWO%sZ4<3cbDs@2;%TZ(f{2DiHgr8hEC0>#bG8wUr_3r6csn?_ zNBewXAu^m^?9yw&)d__2zO_z`x0*brhP3z5JP@8}N$l6jzD0dyY^G?xl=fmX!d(j4%r+0)Y2o#~+ z_oBr}$KL$HOP6cCqD(#lPaH$qOEj=z{0+K1hN`SKNfS3WGG+Qwu=B$!m70N6YjBpU zE0@Xo7;f-9a{x&&sF69F+uTz~6s;zNghJXrjgp0bYP^>eP=z3O;^sm>;D*}veYjj; z6R^FwZe~{EX%#zQn-E)W+_)pdVv-6S98EN$YRi>^xw_i*A)j=3S0L0ram%qi>L|+F zVqTOtu}#X#i~IzMiB(o5p(Mcz$Kpo?Y~nvuz4(-;81G{pUg}D2lc{i>!D;nn0AJhZ zBr~9$8|6PF-!lYiYdVumrO-326D~61?6#w}c~sSq*a_1(ppRhzdbrj{8dy!fd7FUF z8iLK$v-#+!o5$wXdbR6@;%vLQS#SnWQypc5V~0|#11xt`W z*iU~tbj+=y>FIpch{?zs0}Mk`&d7FI4f4c(wcf09@3E%k!)aTlkgM8xmda|KnW)aG z09Xg3o<-2eTv9X_6E~`~$+q{2xH1A5qQM~IqrJf^QfHyPH6m*pf9?GaIN8NV6fEV; zmshyX8SrcOcHVqR3~ZLGi2sz@-0OU$?khFvM{j?NxCgUqyc`ns{g;sybHaIJr#hFV zo%ALrm#>?YC!!idG~?5?KGr6v~!1of<#l3VCBaZpkXmekB_)2 zS*#A0GeIVu`ptTCJ68%EX+`#-yw3hBGRKdX3N@0fo*&RG!^2SK)2tSH-=s)LDBcam zp_wGgE&Etsc$iA7A8KQSjNxJpLu=$#479$kiLS9o!bsdSd`x_qogPx;L$-u#dq*u? zL(%SXubuvgV>6(s20i~rHL$u~`pJ{B&IBQy$2UyPtx?MsEVz9AI^_lA(coeHiz9*$ zzw75YQ(P1~MT_i>M;3Um`JEBK$@TY}#`3y^f4DBG9v)6^hN%uNaa(rOyS+kJIvE4& zyu3Wh#=j&8-CRDhZFziY70@W&T2o9Wq=k2LdjEV|C-@%vei zz@agn-Z&W>hz@EtA-bcTV-8!`1!u7)rTxee!5UKjfws#Qz!qUK?w6Ss>TxH> zo6d&!W==-b3t@LGw?4d^BPpUs{X>(_pQip!u(AAZ@)H%7j-kFI*p`a#d=f}syyUXH z)N);tdRm`r=y%sl!eif?UGT#p<$A=EPBmFP^axXAWC`wOJV&H1%6It7oNztmid=(eV$kkZ@tcX`qBuHwih%f&6V zN+{UjtfKKiAg$I7>u{cPMgr2L-qzt!MN5&5xi01TJ%)a%VBnqCl!bi+Vmn;rVxtp8 zgZ>_PE%BTPplEb)=S7!e(oEkd824AxSASYe=O1Qp9>2mdxSH6QJCr0MIG}iCw>I=O z3RKG{KeCHG*%PZ_z#=4}$r_)ojy-Y0057;8-Xy+d7+YKbO5IRTG!RTs{jL)x==ZXW zhSOowzC{th-45$Zy+;MPbpV!iJ5zO2y@ItYW;up0? z(?02NI1Jq)!F6H*zNBBeR-rb>gT3K2(ug3W zloHS>G8S`?KT#7)DWJVS%xtsHTl&7b{S|SFyr2+;0kf2qTW!6}ibOlFAOe6)& zsX2ae`YxRiIBb8C+a({qzgSJO54WH*nV1X;-Y}z1v@s~HUXr-HmTUrR?jRX5u&Yx_ zqRwrrBq2F>c7J#5>p#=W)7O@!eBNGm!>s)T)do)Q4hq3_N5d>D(~OcQW^>tuheg%Q zlDoA86!@O@km{S*F_?#6j3)-9wtW=Mty9Yxu5B62;e7GDk5MXNWmO0yw4GouP-EEi zF0)a4JjOI!*e|Nc+*qPp+qZ?2T*o}R8-CY{)e16Wz?2`WK}ak`xND-Xc9vR_ucmv- zDb>cd4HzxEzKIq5CN85jVk1g3xE^xI$YOvwcQE$7Su;l{kqSrx$gexd8crdN7NW`c z=+EyM8f2Jj<=c9VA2FWb5EiQA5$yn{xONPfr{FS-p=zGGMi!?@HBXjPuW9f8pjZE= zfYz;g0yV;*ht{;Yfhb&;HrKKdVE8N)U~>8$C$9pm0LN+Kp*yt~VwLM<(D%gz?TBn+ z9i#+@y0|Ahq{J2vXb5DgPxHDZ*?1PAK9yaMb;GrOoRL9mt`|dLBSr+4B99~<8{KNI zXC>aQb?Vf6$!Ds*c5>EC+s8}9e>`Ck_o#9C>+Hv>mK|Q+oXovX`C+6bsT;FB6=X3) zwp|i8S&^3QVwU+8+nk*!qFDTpMt{cv^%fXRTI2YU`OJpAKyk|p^Di=}lIi&_RV__y z75iR^JE$do7;Jnr#rwb~cs_On>Te)^ZwjYSu%Gmu8yeCX=}^Hicb}su&(4QGJINx4 zYx{`QW6*@=X9825^u;wqtXC24TwN;D1xru^Z6@a}T~0PSBOmBIaA{F3Liz~5ueC0d zn(ER1c2Y`>ueS?($xzmZ;(iOWzd4To)Y}_(=!EQ|gZK6H^For7y>~Ofzs<-`ZhdR& zNG~hPVSK~tmvl()NnW)2P2lrpCpK|Tw_NfFY#v0S9vG-(*3%ocem94`&9MiE3I|B* zIH94}$V`s_g}s3}9j*PDRd-dj)kAFE7E%(VA5hqr`J{*(Oz4`(#gNd8%ZG>^ywvI& z%tAWSD)x9HrjNLB`@N@E7NTFsoNe1^Swkc>zL{{`F^vxiU5UBup#RUdyVn;Zxp`tr%>%@GF>5&ey1}+yS)YNi;cKuD=-DV+$$I{zMLdZ%kLA=z z72LH65@yKZhvapUXdm=h|Jba%sG}p1nCm(y54P%97cNL95IFRDHeJ!Pn%e)MYx$D0 zO5b4GJH?*RClF+QC*YHr4JH6~;nZt3Gmjy5BMU}b&OMF5IiT)s9C4}z_0X!NI3&Qg z+9Fuu>1XuiFAIgXg8YrvC3$&|*ktjE0RtLY{7`d-aBipt+Txr`HF0V0&hq#q@z5ra zPy=@N$r4Z(V5x3iY*awVoFqdoiP}%n&@AYKE@YFizJLH2u6~WoL8s__8C5w!mk_e$ zo1u!L9BUt)4m@og44=?Xdxkt4A|s)dDi7Iu-!ep1!AZ5^!ku#AgMYk^ z%BiRb<@>tE#&P7Vh**UDZIQ~EcPi!KIfJ{6tgk-lwrabJM}E} zaQZp91z}AL3Mu&xsL4*Ss1MneJq#N+V(Yo|I+?F$Y&PWt$fspc2XrYg&Rc6!(supO z9vo2F!T|kb*_(;zW#ey%OB^USqWCC5@_#6M>bNG?wvX7Nf=3AhP!SQ466v%+rIa2u z1Yv}fG-Dug009A!ZUGt6ve7k^Zeiqz2}q8P4H(--dXW~S;3Ncr z<-(lQkfw$ogvNk*hC+;#u32;Rlg=132_66Y`J4#+QY-Uu+k62oCC}I07GC1M`_15Y z%Ws5YgcSUQ-?5Y+=;(54VtV^xyvMhMIl^aK5oxUc13C>Fstq?t`AXRGG(z5C(JXJA zsArP-By8#3$mV^or<7MR&o@_%Bz_(|NYH;SD)*YxY>cOFmC$e|bu8tQ@eBED^&QDQ zgO*RB4xft*>&n|%PoikkYfaYQkQ<2Y&rU0!{W$9eHg@zPNlS(Bfx5dhz{LcLb&;7W zB#MhnBoK#;vSYaVkll>HkuPr6eG^?iPQ`n~MUCH2U0=`MVhxIPx3uXnj_OFc%yU!# znNf6iCSrC;>O?avR$PF!33JB#*o_)1-67hkZ=(xdMNHA+=I)<8FV?Nka*Wr%asca@ zOt1nWeLA!7S*66reP4 z;nt%+$B^Gb57)v6C5jz@qs*uKv=Fr;H#uGJ7-|2pY^vEE9S())-|`c)2;hyt-+=ojhakbY&lpBnedLC z!k=7Qd?;h{1uKIFi3>vAv@h3Z_i2E8>1UHqWN9J#6YnM5xc10x?PwZjua^S9#yWF{ z>Rb;TNb>AJK;LI3se2-e`)eE&E#)NH3CVtPaa zF?h*A?b|HFkAbmiGa7*=(}!7baF7IM;=Y4hgYpQFTyI1SR2F@bLYiS}PtBZQk*r6z zaNFg(L6_cjU(FM*w2vF}_R6!RjC>Wie?edOij_G_kfr8R4EvxAPxv~}Oj7Vf$sIvG z4dAj!I3tt)Ac!*}577{+bG;=!F-tJ2U!7PDb!2Gw&knlGh5TeG3{Hh1&SVJpn3%*F0^5vhN3lddNfWFF0>KSoBDClt#xKzPxP|-im@{+&DM=N2=g6 zix+cvpqrl=vwmS}L%PdZP7W!N&d_G{prAF@nqnM8yo_5!6ST( zhg8QzrQ{a;G2+8}<*R!BS(n+$pLhVl82|g?H3?ybXTdY<;_L!He-`D1XZ-vQ{*}vL zRXPt*$yYbEEtTSSZ}L%w^q@r7fiY&UB)L%$CBd~P?T2`Rd!cWo7(w|N4_?%LDF}w# z$DjGyY7e&LSsc)`ItCmQ`a<*R`%?+?*B=h(E=s-x>w_xIB+)?qw_iz;;J7}KD3Hv7 z?2GK)Jo<_mD^d`RX!%gw+qX3BC7+6P={ePR--r@Cg&h*LexJyD@<-(pclZf}aBK#? za$o=w7t;MIo59pci}WnQYwt6wCZF7SaRQh0zTi1b$(M`;C-vf8Y`A-$HJceHtq9ScfniY*nx zf2_+%Xxj*BZ=MCYt(j4(_9yn8dc(%$ezE5Dp$}O>Cu@&e{H+Ae(MTI2d-joy~AAd~0+pA@R2Jryhgf8r*LJLle zEJt(e(x2J(Fj%++X!4g*=-~}!Z}3}Ft0v>%;(L=HXWlp~Q;rXu;QdU>|7nivE$;A9 zl-2X_R_-Xwi^jI@F4Z*F%g#JjI*7IASLWtwIc^ostzM$O3FR&elUf#+rp+IT1xxTNsDVM_dz&Jvl266dq>>ui+2ZH|D?76JhVzbY@ zUu*XdYeKPFwY~V@iVubl2+*B-$MU+U6~ikbef4$4DzT+H3`qznVtfAxnP-@|S%cY= z*gBK!SVh})sl0OEnXLK62TvbtuxM6dL5sqb%3LvFsvP6~_uo+PP6Omo-i(2s@Xwi! zSw)&Hm~@fmB4_7j+f}lOqT;gkry|y2i3f5968){@IF2}S{yb%%oKCd7eM@=6GYEc< zz3n~0LIE)ZJyux^9LkBj_Q$DP{eQra_Zgpt%~h4hJ$+$E)1q*Q>yM|0SZLs!hM9*A z&yHKSXhzBDpK&z5k=oe`P2_G%faXn(WONlQ4zNsiPf2zV^rlk92;1#4vB-ezg!D2> z`2!6h{I)H4jzpN$URrU@UujcdXEFZiuJf9UuZQDeT#d&+#Jd$F6|5zC6Pbs#);4la zmO`(ZD~;D3Gx2^uEFYhkW%b@a%|`oFoXl{zZs@{z&gPPD$sC<#0)yflaH|$Xx=2i? zD5+5dxz>C4k%rpl-6xv~2L;fGbRkwxEJn0!bk3q)4H@u3FxTu=eUDn|@~4c_s8Opb zDD>&}R8D2$vig0FK1@9sINWpnlCiBH{tc5(?~OG|qLWcZ)52ZG10aP#7HyNd9j!t-Fe_~-6%#=2LR@$B11(DI1 z)+d+U?cDB9o?At5tM*%`uK&n2^KExjtJ_ZPZg@waiowY}xerUZ>L?{{4M-l-fPQM`x+T^5TBSH?wml&sPx^h(Lqud~n{Fp~5mOinAk=TO3}{W!5Q2>*|Gh z^Too_ewp-fJGip7K$r;|c+Tv6Q53pg@f%vq3^W-=CR(oxkx0Q{78@~chXJhDIWg63 zlBZ{;Q7my6C=rcrC8l{74!jq?_CEW^$=SVZ2 z7N~-NnNaDVnHog7KCqw0vu86Oi zL}l~p6#dPmzwlb)J>Q5a50(1!g0SDbJdV68-e&SW7_umR;uw?Sj23ss#hatyBTuk> z=0;W;*IHS(9>~ZSv4$umnTwM?V`7-ni`c9~-hL)nVKQ?N}Xsf6;pN-KI^67$IqceDG zsY79; zzt^K3iIS#mGQBwflaXOQi|lZCTo_!_an3)TYys0F`vI4GsS|&`ley-~7il@yZP)-%*cY*Pb@YsM}(yc ziwpxNGK#rZk*BIpyNO)`Y$el0`n|SbQ8x3MmbjnQx0|oLIxjuc)!kLGsKXSYzX!}-JY-|+;${&6J!I4q_hi{VV{)>YY;kmRy-oA249n zerIXy{!k#DTLE*NgDuHQZdMg`+O0T5e=s5XU!!+9KPdxsLdE7}#?UJm2*@y2*;-3R zO4&Mn3H9C_kE=qY8oe$Q8!YB-6(!b_W!$4=LSq7G)DRuu(Qefh(A6PL>ne_{fJdS| zbW5vItm@@gvgj)|CRWEnDP(mrDo{P&MOSl_@CIA@Vd+L?2ZT0m=njleqbMH8&SErHR0-FEej?|W$2!PT%=V+RMzbF~LU!|XO@&=HYcu$ZmQurlHG zO~S}QygV)uO_!_%d0dOX7T%8zH+btjn#UV-moFB&a621x5HIDxA8aL-xY*nsp*CJ7 zQ+e&Oa%OYh*=|KAidk2D*_8@TU@#@?yd~fJFa#5wa{LVeWTNoR27?5BGSAo-SXgA;3QHn8IN<#}$Kfq{|o&9@#GV(mNAStx0x?+2Ph zy6avQuaRuvDc-xBk>S3>F8S@7<;-=?!tGes1pjYH*U$*&C2F2pdP*Mns12BY80)f^ zkRLm6G=*hsw1!fG$D`YQW6wQ0N|Zb0V`8wxOT!<>U06sQx1dyb{Utg5Ht*+|BB( z>AAUAO|+bKj?M;&`%g(c<11HqT$Aa4(x@4+W!C5c=gsKQ;FkhN!4^Bp7Ei$KdEJ+X zl451ySvz7JUN@#-(6qO za;5_Z$;!fde79(F0o4P^Es0=WbP_r8fQUjzr>?FKF{&sc*or%2L#M>k#wvY2jToi* znnwy^Zd*`nAHD#_7>HLb#ix;Ga!6ey1t~X)DhHx}Z|}ye2r2=ruc5(^=Q2RM+LHQs zl9M3CJm@^v?z|q$%@@QG$oMs?VO6oEL&v^`Et4NtdVDN+%F<-+$0pi$NS%@XNzR;L zxGoWYgYq!(k?voAtBz)h4wKR=S{~Q6rwd5?$m%w0d#=&LR1edio^P`?RFpz8H4u^N z^1kDpuoHP`69}N!>uZ9xc27w8H-*HK-EOs(J(inmy=plL&mOJ;#UY_bq)Z26J0$l- zyeLKHMhm~Umnf-_9pA5(YVBlIupq(0ube#11IC`_9w~XZeRIEVY*oG>ll#1mhD53` z+$2KafJE96?)8BVAP6ai2a6Qmfk3C;vYN)NX2;yU*~^X#K8!eyza^HqOAQXi%kLow zH17i~;*LYxt1>+K{!rM{?@0L|=4tb6ASX~oi@nWSi5E7pz3So!mG1GEN@rV}$B2&y zR1FKxV^Dl3HHJa&5SHe}GM;jgqBS8}qgy>+7V{_1vUm&5R5?9=++=8Dq zIOgyTd5+a0*%Godc*(`!QL?_3`j&o~=S;MMcgj46Tm(jW@Bv{l17qR%k@*1;txE({ z1dl2nS19&eKZ^~%?ND(Pk;}EuQe2+q^n5c`hgD^yRM_*yJa~|y#UMe- zK@$vIUhd&1D44Dzo-)H-M;D*7oin$pHTf*X7SUa*{Hp&ubdO@*%yNl`C!F?x zfVs8)$2Xu|J`o5pYiUbD?0PvK@R7xrre}!WGvuZ6FsZR=EtIKFUm?prw0o2b+G;(t9LoR4>QCF1Z=6x(@8QcIYbpL>^1R{Cy)Co zi2=Pg@DEMwNu0&?utoPvPTPhyomEmi1rWOHLX>So4|MIdOCa9HvMKVy)!&g@pt~u! zeHub_6bA*m^`NK>R1ymC0s%379gVB+{2D#T@+xv$t>JN9km-*y9v3ko4xY*4?6fQ$ zo|+wW8)*mkuo01vF(6{Oq54~S@ywEMY9|7B$D`Pn0pj_wa533?IjmUb6iy~zNL(53 z0dvL<#@q0<-N;?>?Yc-$yK4i<8Zz(Og3>hru}ENdeC5}LZhv${oTN6V=p$oe$I4jJrlxb+!Z=kXoin%& zN5GF|jG2yhQ}Em&rB0vcoe47QNKQ1ejVp_vmK}DK5)E%%k%{G~t znE@ihTbSq@<_;CnrGm5+4{{(lPzqX(3fYNHTo#bl%^~KO+TWd@en<`v&R;-`QKXL) zvu%xb_EQFGyv02qLno>3-UKx4>5)_E7wRXOg95%qBQ#!^7Y{8g7QP5Zctd-_IO{X9 zItC7tQlQssQ*C`Aum@r?kKGtfs1%c-y)I4qxW4?E z;(nFs;k7x-zGY6t)dL~J>xu;o`^=`cZ}l3)FvSSV*orrbKOXxdT9^LAoJAfV3TMjb z$z|)JKVkVfN?5){nnc=8v&$jR(_YkAn~zoH@HsfLu02(H9?*fCp8bHYY%q{{}Lb37)*I@ownHz1Sj#1j4@S7j(d z<0Vfcyd8$m&TT8j&LD~16fqY3$X4Av=pHdt4|Rlyajeeh4b~>A=o(SH50+`xef0@T zc^q@zlC}2eUv-D?aT4=oFR!5*JGNA!(L4!6pdFnYQg^4kgXXQ+t0AEQLg=!^)ErWr zQfX9g8j*|ru|>t(4YCqC%Bi*pT9%dWw7zVnrjSkP?seG+K}I1>2p5SG`nI??_Xrou z;p+Np`IN%TMedIB1#{-}IxHC&%Q7&&9~f&?TLg<(-_F)QkJeNQ^5ILu7BbK5!5?EA z{N0jlYW&Vp)x~phx+v0RYf-pv3xy3I%)`Ci&lhtRtre2}pjscc8u!h?+e+_0&L*bg zP`1?+77%$5@emB&Hi~AK5A1l9(No^9PoYwtx(5xL{q?(-$05b`kuk7nPTR;A6nB=6Thn6h6?4+hy;?Ij zv)4hd`AP2{e8|9Z>%x|MtLQ}Ns8|i z>D3U=4QraYN{gq9=cG^gAz%`iGL%b9d|}~MF`{^#z|VrUyQqUmx}j8FAds033^y63 z^%?d#6HMoe>c{$8L)1%+uc@RNFSbnaX{_icXgqY8%oCAoh&jlh!YX=NWh7?*1LR^< zmyi;AnWrI;^@akjYiczGTgg8=#o#H|RdgfGb~X70zq8w09#}iNqxM>@)cPl`(j+yW zc{}$(f6;h6&!PNw*Y<4fjcu{?vl6Q)t~m_op--X7l)sB)*7MM!4$Xd_#i2RcBJBxu*XGcDK>_vnvjloI4NV@M0!V56bkCGtGE;oiVK2H zI--sGobmW|bz^H(1gh*tunFO4dG4b#U)pWv+ezvwIl){5kgwY4<&>!FP8Sr*-w9|| zIfD99u-|-_mmSY*njn2OV`IENt`%uh3aW{JHuGZU;YYC}ngmhkm(R&5>uFYL zK}bZk(Wttp@2B;NNm>G*$fVc+b8Hc$j-oH4!Dh)vChga%F+E zym!;Ox2LCGcVulVT3tjuvuRSogVz*+N`jD`ZqIEk5>t_iz7?AX2Ly!&u!Sun`Cs(b zWaD;uWB%s-2O?9LC%k)Mf5m}sDKxa;>zqZNq_rrnndwtwiZ?rI&IU&wL%>Oz5U);9 zpHhern+_Y-oqnOTar@D#T!#IqDQO!gz3pJy)Uv9LWW)>o6zME!*Hynf|2FUlt1iU^ zD*Zkhe}C$J#8uiX4JVekk=Lb5J*&pc3R8|b%@k$2D{XC=nYkol+g+GKa$V|xC|X4Y z{ch9IZj*({gznS&_C!}=cFN7Z8#rGtDfg)0B>Ul4dani83|m=c73kEIK$-r0Le z+dM7{@zF(@7j|MQ(X!kxTP98!w=7lA-PictHraBUVfAj9#9mK0EYSxn5bW=@fQetqDM_ob5nzL47A(u~RIz{PrHgO}ef3*g-dGILO+n4LmFoQ3z zl&ujS+cb+UeJIqc%G|Pxr?Q68`wqD6L^#=QkDMquIM%UOS@63WO8D@(-{3tqU{W=}v)!$Yph1Lvp@Mt$=f|WF|~W0Oxz!Ze}ruc}ZdBK}RUK-8b9-hKoIS#JZZF!!s`47`vbVCP1R0_3p=i9=8ASG`hT&{;JlM@PlD zUV6XdCVW@h;Utn7D$h7EAt;%+4kNhmedTIKHZ}x{2eb?Qm?a z+Zh{MC=X6--r88INEKPA6Aj@&p+)0j9tvHUqyW=H_p7&~LP3f5=gdK=3+ z%_XALoViagQX`^bn@&^XXucCm66=%lrDK8Plc=i8lO`I!tG>t z6DH1hId%HshO8@!EP2TF|3Y9qAD*> z%0Qe-hqF1q$2i|Aq3y_O!-dgT1R^tZ0q0@nhF7Yidrz3U4kum?Lqc!uO%Rt5H^!2P zDOxD|yLv>3M=H&-VadFUJ-Zxz(vW4K-p&lwEpKo?e}W7A-< zT5y%ssrDu|k4st6;@!rHTAw|{AAuV_V=#>6gt<0?)k6Hhqlh<=9BPj#XgJ)1*~%Of zX~QEVWn+!4@{=AIVNKd@WioDPAKn~?Jh?Eq6U9qw#kH}CDhjIP4by|SS40BZ$p|JH z8O^-14`vBIa~+9uV!kmpHZL=xlz3b63AzW3a1Blo!`6Fug_EyTybh~KZYq)o`&xJ_ z&D^gt3;p%?&k(N~Tu41oNgS_DEP9_PE-*PFi8OkNEErYeVaZ?Y+}nvD-6>!dKYbFH znZgTwXtBj{tn)*KJ99xnLSm$ZievppA^EY!-tp1^S{AE>P*!pr!t7Z$U+gKaOa0i4 z!cv~tB+*k=85eK$Vlnlz**Q(6!s*xI^#r2h+Iu2dR>#2;73uZ_oo^;c zPD!y%8Y;*w+1jV|W^F)XO%XrYnLQr{a`e3qDsgO1*?JF_Ik}|4;qdocB zLRel!4s$-Uf**XmnuMCO*Uqr8tMj$G^rzsB?|o@X?_~^{d=sd}$P-fRBczhYcLqU( z-97hL(Wu&$fq70Ue%Xa>yb24w6b$K+<$R)|D1#gCo46IRx6}+LUBwZ*FcK=^pyRYS zw=$?%w>;DvDC5t$hde?HZv#nZSb-7VPQoY**= zC5~~B6jyfmY7p65Vqu|(x$BNE#pX;G4DRAQun8WT2u|_%`TDTP3+L?_8I_;z&84^jfHAYq95iPgDcWb0%VQ z;+bGa_8>iFdxmdPL}qb7q)q|6R=)yQWW>V9}3jAxn2+ z&UiZY`K7QEm8FyvnD|~fSIqO65Ee0?6MeLSZi#lE%}`h2)iZEqe=NC8q@nfn)TEfN zL)9tC`PdQiS&D$_dY_lFWuMI}6%V|E-K@HNmH(79uKwQ5-et-8q0;oRrKyg@lQ>EW z*KwI=*>fKD2Ey#>raUdiyY^ANkL)O^YSvD`$onrP+2^bH)^4Kvp3L3^RgPv?@AnUM zW-4ZKLIinjT1o1dJek3++xKXbfC^L(`CJ@+nE1s@@1;;19X#NUI8e8b*SMp-_9ofY z*h(udb%J0h_g6t-P|hjM6(x^uXw!#xG$~ zPvp%~))#RnX70MrKnz-EaToTE@Ks*Q5fHpD11BGT7bU>!Ho8&4IVi&g{Z=41*8h_1 z$F?ZcJrKH|3-F6gl6S4Fl=MEh-}i9+Ajp)X13LADv{@5U7nn$y3us#y=!>VzhqFZS z>2?{v6tYra5Jg3rwkzJ2nsfzm^vs~oj(_=BQz|IA+%32qnCD_QKnB{iqu1`}iRM(j zpNbNPGwsbWaJ?zHHa%l$`{he=e@Q79@7%;0$*ql8cc5hmS-;DDM5S-sotUAv+|)6w zf68cQI$*nA7wgh~D_hFx2|ah8Re!vX_9E|*{1KO7jdAH(3{%%(p_4`vRtiSL(Bk^g z29<;93Ly#+AoQhHovX$%)CKBswYL`T#|PhkPZX!*4HsdH=eip|edS)06 zbiuOS*$p2gxUx!$x&OEy&Xk^0NzcLj3Ix z4dV|z#dL-HTrAc+H*c(x?awY>#(l%U4upRe+k7cCy-97S2GnTj9 zMWegNQA-teUSeT7e_?gqVJ^gT3z^c>r~dwG+|_7CO@btoan)pV%L>WdQsOkl;9;dO zD%tR|&!aGA4AyB;G$Ls_FEc`nKP`m+N<8VB4yFcq!Ym2MXfYA>9L-2@V}9v8l-MwcCptg&G3m zEA<}ED0uP)4(%UyRmK*(c=A=~-o(xR$!W8BkACQTUkwcr0Zo+E=0l-$mhW|GFZ=UL zF&&;_9#iu+^Z8?woyHC0VAOi$Hc0LK%4SC?2Wkl2I<6wkl0@n1Lt4CLDQ#7ot@3Dde2Vl1fMeaH}E~BQ<$1~2%c+*Km4vLUE zT4T`AW9C&~Tss0NME{P7kkhjQGPuaIV+}x~h^sahzDQc)%r_)gl-!`5iB#t5hZ6e8 zIq!|8SCgO$@0znF;0#RHf=m<`>muRvrTW0d^k*pe`NalqF#Z!OT`56cCg{}Vh1lHo z7e!p3b*&$K?CWqJZCP$D3)+{ z1q}u(`ITtjEE|o&lSXy(3Pz*^B|8VD4nA+pYTI6cWcXD^l0`>enwBz`PLZvSA%Q@UgE|fSx@9Kn z@HC2dzsInD+f+yMtCu5Bf__^9j~&Y+)@c{mdgQLJsVu^+B*qlLo2Hs9g%XWUK&ne% zcHx=p>Z@{9k*e9+F-l+0U)Zp>G#d@{*@R=1*!dThVYpUk=n+~EFf z4;Cm>jpz|)JXL{`KA2p0?{Tv~mJ{#k|=rFigPgVY69cDyXl3X)h)|_tOlp~VtYcOXySTa`b_`w zLWee{3f7Rz+98yaFJRgcYo~cKqj_P5-jaa5VQ<>;YRV)Trv$2Aul4WMqo5)0TNbnB z+OzL@jXEgR6K~i()5xw!eHQIV(2BKJ{FFUTLK|&U<*TnCdjOB?z8;kDmkkh2r-8Cf zdl=8epM@l%?akwNnT9{rr`n96pSZer^Fb!k=FA4Hz&tCxzDV#87m!+Zw0XAANh1pu zYyyLlFtJz}4N;wKgezcrsFsXpyw@Wkv(o7qTwkA&@38zyZ=-;}Iu@@}U~5p3qPYJv zZ^Nm3ijSMS%oj&0)3O30ahxgvB9)_2W5#Thr_%d~wQ82^Bw&V21TG|8>zjSV{D?ZE za>g6nL{u)Xo`#dO&jZTVqBN`D@g$F#_J={GxIa2z__ll#T#>s5y1P#Wwy^L%D=P@t zsx4RF2{UQjS87PJD6sJu$2Dta&09#245}RLD%S3H0Mm{#t$9^bP?_diSrk^LYr1c; zmSV3KMII3HoG8p|_VZdj$k_UB_JfL4Wv*e{%)qDB_`>|wfK#zNOiwB~X*U@)-FeV= ztk*?oGk;@Rh{`4G3OFa2)jNOMQKoBm>-Pt(W+S6{~{amcI%93!fOo+ zgSE?}*eds`x!Y7Jt5^Qz^Nr%Yx$l?1?N#>M=b;l%SH07-L==H}3sZd|t0!v;eBZY@ z71v+!d%*|{74+8hH+xTO-H=mkJm3uf9f$TTypYMM)i8%kCy0-E1A|3EYM z{q04Jn@u{@xgN;Xz)0&wo!lfzF|9oXIXR-X;xM;Gim%GeHXv>9S+2`I@^o}r!g=0* zO-)=Mx;{~Pz>+p3MitD4!cU6Sa>@|XF0k15iFL`qY}&SxGUIY3)Ha2zEUm0?0$0Qx zAHXVRdi(G?qeWSyHDYN;8JCWZ&e#Z>%xmy)XJ-uU=A>49Oa{<>(W>0OYd65b6k&Z; zXQT)Q>0B|>?JqAbonF0;*^=oZ#?c~2sxI}C2O>qQ7Vd-$4?7a7qF+muRxBO+lr2RA zx=$?wfop&aeED*dPyCfG*IaoG!>z@8mU%EseCeiQ`FDFN=T{w3Ovt&YYLl7)jD0eV zz!YrC*Gphqv%V!0Hn}r)aS;$&#dvi``~%A$lB=(%mcTAsF53{c&$GJDHDzPWiJ);M z24LQaJnPcAn#6JvdN?2M&8Zko&j(#z9k;|*aAJFh#E@f{EZ!DAt=RiL8o*`2EEyN4 zwvNRiW>X-G37?8PJb^lcBj|M@3SE&spUi6inC)^;&%`IhPLi=4Mr3`(E+4{R-QJe| z-26xs{Iz*aHy6Ze(dzDsev!ldq18&Lao<8B_b{(6;)+3{2BT7sqtdAQGl=w zN24NLSY-AxO0NQ8=32p%?dk^Wd_BC#J@(=6=d;@yc|B%)e`a!3 zLh8Sx^j%EfviS2pAgz(2OP6^vxg|2^lSWRKS7G#=%8ya2$$5UKefvNS?S(gf|NoP| zvkv>Mw@%7RjHdeM9BD0E@DL&=p*EOrfv0@4f^+ zJfWz6l#Z+c;GHni+E0@^zf_06d0+jKF}mOb?@st{VH!b)w5HUuKJGmMV{io&p-T6ll7k#;g8*a znJ*_W!g=j?Q__2WoLN=<*7xDZo%g(M0c^Q-Y<1Tb`0uQVxA)qb2vhb0LDb*3#`=Kb zJZ-1r33Z2e8H6uGM-Cwpb-_8`XX6*4CUXam&)%}Ut!{&FqWd3`LnqtV22QjQ zjv(>X5Qgtk*?B`QNOlvoCD%sp5dJfO|BSy^=aq-3i&?E(3u%uuz~p>j1x8brpgzqc zYR&u1@sP|{l`FM<`9DLWM(I!bs>Obfc3b|c?I_P^nxIPVG=1D_|3}`gX{Y#uUMrsS z34UP`|C@f$GU<0GC}SG*a^@(t2BFaQq8N zzMk{-BWm3+JOwCB@TWldPkpU_SCzkHQ=chY6H)005~=Xiy+I)+=RYviqsE$9ZO7!! zK0EW%V8wshH~?g?PQK3!rV0FZw+22wC4Ua?r-ap*6c6s!2lm^*zI>v0u=2kE`&m!k z$^zbqufpol$5(|Fa!#w-qvWxkG4{MwRPQ2iQ^W#1q3~|Wg zm*4d!yBxrnTD>DI|16FDM^?S%^qT1ts%<8y=eCxO=33cJkETkoShMUx^Jwbx^Em!7 ziEpio-6EkT^QWa?C(SA>EV4^-fV`BXuVnW%-xy~3*djNpb_&MVC>^vdi3nt+U@x|iCs-O*F()ld&57sLrGUSeBGkA4cWigRCq^H= zK4i{Nlg+{@J{CjC=b%Lv#d{+yccZ^`5%>y+iG#T7-CYDy^YI4`{1^AT(IVv_wPg~; zz^R~(S?Ow2B)>Wcc*|wmh@jB_4(SI-+2_^FjMF@=7`qB~;cXMOb66Lnquk89A~ZQT zpwEpJyl!{@H*U+y?05t$NE`nWUz2V`(g#I5hVfg1E}f;O&i43@M(@eIOzq&P@cTDn zL@6{#1Ad;!DwxN>$)q7HD~{}Lc}{;&)roBvd|5-Y{rfYJfa1B#f9zTxX;9@$J$2E*U40r@2t#!%&2D!dW_6U-9v(44S&vjN?{^>>_+-zV- z=bv~3jPa{~#QncPKlJLnS1Di8=|HS_@zW4hlU_$>>!9&CmGI$)T`?y;a|IyQFUJc1 zAH1eERMm%0MH2#%Mol*1d97~GJfYbr!K_xyyX&|^74BuKsAdL&P8N2iE`Dlp z|36j})iqq;Hm&a}?vR1j#0YWiSX#K*@POt&IP`_4XRl0msIijaRJ#;|UxDGrp@6|} zebL=qGu%EDKB6+T431s(25)=>J<`<*0}E3ZkpL53A5fhQ60C22_{#fFt$QWmCB_$X zB7svPbtcJ)Ad2rD<}@BA#v1j^E2IE{aR^bHx&*R^X75RW;^?^4UlaT`UKVmkcd+T1 zZqCo61>aw&xyVn55KJ9chdia+>8JziI2Rm79&d@TQ>F5;^<{wIv{9excl)$&1baQE zf0%5_{5fH_$@#v*J~xJF$}Ff9cGVXIbP;*CQqvwul*9Y*bGs!|3HSoM^%FX(b$@_3 zL_h5(3z3I(_qvs&vpsS8b=2>*dgb+}?1cG*{Bc1jRMHYxRMfx3G-x!zqL54iT?d;L)E_k13r8`;^4Ga7RW+{#V)n{aO9fH@SOIN?|=!>)*bD?Q%VqZeOvjn zX_v(U+ln~(gzsM{%aVnFoeS7@dGx={1+%o5k4?0iKr053A=HcQIW;e@d<@K)u&Wzq zEDLz2qk4C4QzQ7pj#a<>9ng|TQMVz#Cgp7fkYQ|bGh9Vli=Mz?`%a|?JBQoUGVa3n z{gv~LI-w16yMFw!+;m>`zMH+|Hv>sOR^9dB`^a!K$KGg!ukGrfx3Q%zkSpo_0cyE8 zK&=B_M*#QDVEgo^Z@Is#8JX7M2m;U6c4GHN`|cw6vuOgclEp^R>(jC0`DEn?kl*@6 zF)=E(rfLFQ>NXH(`Q>J)KE&|)2}?4QT=Bnf>CFz8Vpk$SYSE6vz^UE%+w&2z`9Ztr zd;o}ALG~K;M+o2X1MJiJ{pwFxF1@Y!wc{}?*~f@7v2KM;yh5bQ6y`&Ppby7qj#CMbnjYrofIn-8X>2xU%lDlNS$7t`D{ndVn~w7#>KE!m z017rrd*yyff!FHA&jT}F39B^om#AU?Q0eow?EtGV+tGwrkV!`vQ1mDi5?3h|F<@+> zc5C&Oz$zt_L%;p~B9%^8)6=?WLSEh!`laRELsP@`VS~rsc!9e`%T#K`MKfj{rqnqw zdSKJmfOCr4^MBpR1&1tCA-}|xIe{h-eLA5}DespIm3^CC{RW)GW^bS=-RiaxYMf7Q zhaVy|ww%FG*ACQg5)#*^89 zC%anCoIXh{pC1Y&hXhz?R^_&<6hrl&3_vx`5g!PT8Xo`L5mM(%u;KnG&wnMq6N$R? z1nshlKTw!bm3_UgH*4`P#V%iv;eZ5VZT6HB_h6tmfzAII8@(wH?Gx$=mvLplb`0r1 z+e;;szc)s8DZ})sCsM`LI=?h6zj-}o(k8e2MAT#x0Pjk?+eJm)KDk!_`7cXPiT^UU z@DjNASwOD-G4l^tlQ5#P~$(?X9JtPY+L7);{PiM{dGvohr3NWSiR_)EY`!flqA7FPq9L=i^ zkkLMew&7(nm0Aj`$=}R#K5E06@U02@*OsaQJdgOR4dw5jp6%LbUIJX71un~*Jp3hz zKp8$8SRwnIsTFxNxqn0JDo{x*psFN#3Ge{G;e2@?zcrC|gZZ*e>yX*#M(#}bmlVfW z|7|fp&WnvD_!wKi@Eoao7kTkEfLgBdEV%E6`i=iUZzaJP_WajD6#&PLvC#))^K{zP zxCs9RCU3Or=N$f14Z1!BP#x`x%zCOQ_-`ZaHF^rz(8@4sL-M!I0!qoKy_{Ru>CzUC zs5Gl0JbX&FeB6tWrA|!wCjp1nFvKBrp2{`+*tmy|B4wxkc)ai&OVo- zWCfolFg{t2fC>S1Z*1nzW$*f(ofcghe&E+T?F5h{i{Ik?!}{;&i(1#JZE5v3`~p*! zF-IamDCa9h9#8vhoza`r*3woPNc}pv^6II7{9$AH$@|15I6w_p4H2VX@{|4w=ZBRd zj!qZHO@7>n4GUm~{21Bj7&=Dj!+Imo7;j03G-k94RSpHzOv8v&VbGYb=+ynoTlka7^_qbQX`C5=KH?Bqa z4Ny=HDQ~terrK-3r^Fh-5(Nfgq+c7+dkoXe87%03iIdOszoHok8w9F&m*57f7h1P$ z4yMeEf6A_LiHMK=!Ghm_Ii>TBZpR-A?M{@w5#qJX!~fWw4mB4kHT1{&8HeB1Ws#TWoh9AtKeqipTNk~XwFY^@t7ni@!SIuzj>qt#@ zIjZYgt9U$9Qz;cP%sY*l3lrVURJ()<{GEe)sYJF*#&6%qf@`t|Wst?Hf<@HU%uxLh zfN(&|yD%z1e=7y^U790^;FQ3$tY7~5Jii_dkPwd*)jdL4s1ESXl|jMTq%fYJO21y6 z0NcgSzQ>kJ{z>%)zMIue7tx__4;54aN-+ji22xy8)m< zwcv{^zpNNmU)f2AO_&s&sW?itQw&A-XP(V6?qN~j_60e6n1i2Ex)Vp|;P$)ZTMYy& zd*Sxye=UKvOP@PCJJA-nUCgJpdDm%#1l#`rMMdPbmf^VD0GG3jq*l<-9hZa13mKsn z(hmTDdNt<%yLz4HpZilB!-2+KDL@FX9^=@+gg*`iJN|&M^FZ2ZzlB7?e$9usbTlWo zK7{E1D+#JzWu=Arvr-C1U0p{W_woS^xH1scrZs#5Bl;?3E` zRIJ@CvFR|``Rv$|ZEtR)t~Ws1=IrifGq5!G-&3FYac`4>XudWQSj1v|R_s-Y4kl z7oCUzwRH9W@Kr#%dWar?zK7}N|M7LzaZR>g{}Dk%R4hP61SJ&^1f^ppBB@A;grqP+ zq;t>Yrzoi)or031J4Q&ih>Qm57$G@eY);;`Jn06Pp+@VC&C13BEG-{83f3#N8QHNzf4u{mDd=DI-5Ct8 zpppX%I1>}MyA=Nmp&<(h{=kR!PYl(NZ0cF|E7qTvpmY-MxkyUNgVDg~fA1C(*p&=5 zA0WmatiZ5k{LQdcUQ#M^x&ITqhnBV{`tN}|{PzRE(mZk3 zrvnjCQiw@^jB*`R3C)dozd}Oc2pyvHh@Leg`5TGW#sV#+PdLqRZuc7>O!-+gPK*-F zG?U8P>`&{0%@1Yuns2cmT$Hba#E^u}mzLksl0QWrzCiScw9|5|c2(*Bmh&C>sv~?s zwFD~jpd%Q@>eB9!Ut2n#)FkR1T%Sb3vpVtyLdSED-{j-hLJr=H(PyRzJ28+z$#|2Iu#h|OPT z5sO{d{@u}j+5%2f>#5Y2By{^lAyE00q!1N*XTTV(Hn4@bCtV$QiZI3LOX0`32D62e z*#luWU6wq6yS;Q;L}*Js;0qU(@agXsF6#B|{Ozw@@Cok`K~=1a{7s`CREk;fOaj=F zpqi?UMZee7*@usGJ|?->LG_qT9nQTLU*IHvB%uXg6on{-M(xs=kR9m>-TuZ&Bfttp zU)p>gU#i^W4Xdo6EQh)AyQ}e@sP*p3oBxH`Hz~N!d| zFLvLJQj=86^W6kz>jrt!PAt(DH-vfkuH8~)qWH^0fPwZ=Q}FQ`$j$-Kq)N)XYB-jL zW~YIMXA*sYy!7<8UH1U$qBeGSL1onDua!}W{Z}4|eCz~4ImVLSXttt-ROh`UCMzbn zb9apj_mIgzeE)tLyidYiLUKO^j9h4PntUI}65mRimniQ#miT?5uIqGS+cwv>C<89a z9f>MM@w39NDh0^m20O^nb6M)#x0*O)Z|=q7$Cy&!=R+zlCtm@D4f%TJFLxND5aoC` zXhgjw*S}rxag(*Dg`Vab$Ss&D6ax*Cv(BJ)8AMB(LxsaZI_>a-7Aw-{Zg1mrHV8Z= ziq&O*oeUDINYou5VYK0M$Z80t^`MkK;XLX`(O4-oSKOZz9B#-RGP+E{CwtXB;~_`Z-VH17pWY_`UOdBk?ur z3p!j6jLhe~S74dMamh5hRuzo%h@Kf%HrR@RW5vON5@G}Keq#2 z%ehtnELkfr^yoPCL@WAX|NP*DQxgw|LKSEVp_78GWp))?!tN31E``u{>Nu-B0x0Es z$`Tp1GGbY87UKNUUq?$VKQS99O69Z>H-C8Jk#onG)yPUG7#Z|6)*55@J~tIKF3cn$ z=KX5Y`mvs$^4k-HYWljf^*S1;N5*GPe`u{!k|R!_qJ-g}e6UWK%92&Bt)@>s=^3AU zO}C0lt}N+X5Mkg`*XDahiPhA&k1QypB^2R;8Gce=bN9E(9K7_v!I_pDI-2=^tLtn_ z+b@!WWEDL;+VEOkIst8;l`+{Jc)i`LJxU}0GRAuSBYib**?8dF=N~4XnYLaGa$C|I z^ulKAnkyl|)3-%St0(nP-+bj51w zqnb`3XQO)EuB3!u^O>+WKV+C%D$6#$UaK-LJ!$Dw$y*KD9nuP%lbx`{CYQ7ZXf-3X z;9OrWha|DJf(WlKyjVyP-q2!aq*+R+WU75BuNiC7>x<}t&MO!)WW&oj=>dd;yD3b5 z3{)>U#E1&3=GMH%#U{~)^xV1a?VeNs9^ueedy8EykM}M3@`;e$&Lk80{aek>rD|vN zkkg*B!$i4mpboAr9mi<|HmwvkB!Xs{(T9Dd`t~mC=)ne+^_j{G8^}icQbJtp5vaVp zZJ$eJZstZTZRV4q%`X2JH#^0%CC@s1AFAU|FT9|#Bu9HKy?Db@yA{Y>#KjD^jU^q> zR7gaorxk82sv}Kr(e;Q2rigE3M@t`Iupgkw>HalG^O2~8a@d{J_Ivb(4~PEtYHa%12-cBAQ3$5-W6%y2eg z+CH|Z$h2I5wIXwHt%c^+!b;Nb%bpqCcYh%MquPl4e z4IH!u=Q7+v!Stf6v-^Vpr(@omz0qsx5zea{Hc_lfD^bznZ%|9rGPUKAp8;DqMOkGu@Z%TV~Yr?NlDQRu_->NiG#2fLxmZaw`(N6L9LQ%={M zAe5^`C{1y8H>FOsUN7c-EjzHVn!{CZy0OL#kI{D>uZbUbz<=!gBGJ(jO&`6P>UZ1F z(69%kc85P0ufX1Qh3e#7vD-pp!pH+nO`VmTPSzusA!#PX-T?0;`%M(jy|S1~=4h^W zh6nPEp8IFJ%jX+rrCJPje->h$Sm{aO^~)B}ungd$D#bM=z{Ke2Lh|S1DQt5WGW}dQ z8~k6mxk_M$Q>+jlbJHI#=a!U+<0rGroO@aW%NEjNF_X+n2gtkz)aap-`6ZM0oE288 zf_Q~)qS2N{b(chy{+k*8)5I;3lV^>?6aNxT?x)z0nh^T*x{nl){nPv4i=>EhnC@P7 zkl`Da59LlU?7sKfWIp@_b*q$+YPvX!d4FEqxx<2jA6MSeDMX0V@)}UU*}LziHyNB_ zoGco`&A(1g#xu8ZL^(?Cb8z`RAfO#L+Ky|LC}ycIz`pLa@I|K~vxGFw zjy)0U%C%XDz2nB4mG16wIy>c7V?AJi{l&|=7_&}Gxa%)9%z7V>NCdPt|{aS=+6Bd zNS>{2J!W`Mh1uQg{enzyLtr61hw|`Yo;SH&E3tc;3TxLNd=a^6|kh zg)?U#xCM)nZsfp${RTXBv45$vk%ni#{H2pz@R@%b4SZJ6sKX{nha5+=nV2V^-8pV_ zs3DpQmWp^UQoU)lUJ>Q$Uc1rcBj4O+Q&4DGI3SmFgKB?jq!ce5s`DepyuVaHR8;hf zlifM&L7AADWZLR%!PWj?^1YVo`SIze&CO>stvly(B)@!$eL?*!UigZqhLbhB>NzDn zGU~K2`V^_`oXI3HKfCInV*Uz*70ViK0TnM= zo^sBI@|>#K>~7@_Ez5j$?~70$`sq?zgywj2c(}6KIf$+SnO7Rabv|@3!|>=_-dmng zQuir77Z^_yoCD*I-kU9@KEeyI<{Gs+{hnKY`tJgs#P)`fQ^b1QJ25Hk?BX79<|e$G zX)>Co&gijG>CMi~!E{p6>;{g^yI8BQMuiE$1GTiY68fc3#w|35QyFVcWnrStN-`Ig zy%7nm(bzKzVoz0Xm=6dv(o8p>7M&T+Gw-#Y(*xhY5en`1f$qq0K^)$-KnJHfF?k`pyqf3FDmq^`h5u7v$Va9GCujuRjofOQ zgHUgc4a`lZ1{Je1*Oy+>*M+&hjEag9b{jTuD*eEB)>smE@J6egQ;|uU^s|>iK|!Nz z$gI`~LAnb~4dh&Es;cqG+S@~+YyoVi9Ch5xtwk-fT6#LTz7&yxA|OY`U)G6KTle+bbJsq+qlU=o`R&Ts$X7h6m?L3!q_ zb@Fs_a>Ryi;B?7Mstn)AopfK{KJMo#DwS9VLtERqQ;V1@XN@gd#VVgZqU?nY_ku3f z%12jcCRAx7BQ0CnE+1&+?rzdo6p!pyBP_uts_Ab!I!=u_)YIs-I4bEYotj-4Z;6!H zSA1nmeLpucsnotlLPt#f=|&!M7FmH{D9zwE4LT&zZdXx%g~Z(PrINqIC^Y7X7FBK? zmmL@7A^TaU*miVl{HHlQ_m5IKsvZ`2az!lLIP0bzv;JE)O7oV%_NVSwQnIpq*uBOm z1IJsX=!{#m;}G27eX@F=tucq{24QZxmh+)K(sVQ7$lK1U-%SS3zQp$U*H2@Ck0)mt zE?{i;$aT!n1o`ej_bK1nk-ib+fI~$6=^OMRuFc-+U8)Q|gOc~Z$Fqjmf*G1Mo>CfT zRJ>dtAesHQpnz~0mTz9RHS_n@(%l!b^N#Jxv%5DS%uo|?-1Gd|Y>KlDdI+Szp;)SYKir$?j|#RQ&J1XDwR> zqt~w_O$-J|8M{e1yS&Fu_%R{eFcO$C^2<8*vnuxjf>Z4?M3xl~?}29qjEEdGrtb(= ztLhsOh#S*Upesq$hZ)aoz%bG{rY5h3TQ_Kud4)r_wzr2&pOYEDE3KB|>FUBLV|Pi} z5FJOBqYsm*>6W||1mExSwl6nvC&cV^oWY)P93vMOvs%Un+lACJJ7*SXX<4vq)9tOLN7BuE{4eFY zbo&>qeaio6`vRyUgLYea3AjjplW%eZi`CkkK84VrZH8DFAy35!^<4Dyc2@aaRC(P1 z6{iF*2T47T)bXt#ugW~k{(&UB{*#|dU&uN-R757`XJ(p`XjJQ&bR-UF2ro_4>F8%j z*p5b?gOx4_Td#)o_e-=zu74_A%*v|B8|anl!UtSi?^cl3dk}yT7euO&;nx0r#+Fc2Oi^Yc!zOz#Q+jkROW3xAJUICt}LCuLMMLHKh$;} zX6b!MNBz{*#$;;WwmaLahM4RJ-h4t>KWt?Bq#p zW~j8ZbW75b`%P<65A~$e3sKKn-;^aR2ro@RZ6fGg9(uD7PL0k>6{aU&#IPi+V@`w@ zr%PZ9Rgl#X)Ch+{O=syse_a$BH;)trPHCKFsJ{#2EaYFPolUXf*=G)Zfa?h*LnwGG=pB*{Uu!~o8 z;`kkYLTF?^i$Cdm&)Hd;RmxSZ)LsV*Ne5PIp{x{#6Mi1<7i-g}D@P2%q8J_L?kWRPFS`(0KGDjRUJ{^JFC zAGZbcmqLvHUwP~X#gt^gjGtEj4Vdy&U0d7~Qs(hv)FTnb%EYgoonzjT&vK;H`4i&q zusLs*I9l^}%PO{t3Zsm*L+oG1RSMr&#RssU==qxM&$etfv$z;ZNJuQq&i5Y&sW64; z@;-R1TU+cyzwXcNAX(N)UZN1OnQu+R*`em% zc|$bnBfefE`N4lg7ziu>^a2M*gr2N^V$pI)f$<3n1avvYv9aQ&P~fQNKTJ$Uc&{%_ z#OE_fy0G%ls0uqyKZp>v;%5wE%m@=SIsihz7u4wNV6XVc5$}_xp0*&@ucgRL#3GD? zzIq732m|{K`B|2q9@vq)o#oU1YgpK5E^ctR*x2OCe(W#x^YFm zX3lJ~)h6ueC5;k3la@l$iB>O=p2#szY~?OIH~2*6Xq?R=;*%?l_!&jq^oyL2Sy|^L zt$Lf^&eY!ew|w;Yo#z_CgF#qhm(@#(Mj*QA>o^jbs@2`xK;p5-2U2oi7^Jlw42Z9aSB&=tXb0n39;L$kov#v8qHg_ByHTvAp&6!3Wo;mXH zf8g~T88iCS`t{9)8nM~-j(X^%u8oP}IP?dYQk~tsZKV!^=7xl&p@^LT^)0ruLm<~t zRFM4thD-elsz8dwl#09F;C^cAY-{5BVoG>yD$@1y4ROq;MiKp%o&w1PN&f~7$BNum zNz9j+c3b0E(`Yce&gPM=tqMXbo1Yj*ZZyfnOlfGqky)(GwKplsv+)3aT zE6*c*3ilvgNBlc|7@i?vD}7$w{S=Lq4B>jM1A9#d!#cR);)|rRp*?Ta*1B2zs$0+E z412;1JaLoFOrF`4T{@!bfDBah>I<){Ge;|{-8D3I#ExptnH^~o9a3!Ax3$In-hM!` z7p(Gt15F|gza)G6dnA8(!}heTw^bH8JIlD>Rf`p{jAzt%!OFMP(J7R`vqC+s2PXfS zAeDyJUI#3sOz8L;3+&0@P>j5Ih1EPIB{Di-${O49QjU2tEvHOrs5)53D=28hq;q{D zk};5$)->qY2nlsbWbG|x~>vBDj2I9zXo7U}4ZbK#6LO12X z%K@jATg91DYq71-Qj$qhtBncPVZ24(%tcnN2YSHpjSk<2YM&KZY<$CxeA=@bB%uD{ zAGpl9jA7pjaM%X|>vQ_GH(Qfrtd4|}DMV3FjWknK(r03mEiN+O}iFUjEKe7bHp z@6GCNGPv@Y#czTme_Y9V!~TQXq5k(h@NGT4gxzB;6Bv zaDz)$hkPW#cq7>A4db`IH0;F>SR`-qDiFOEV_Jz(!xZgdF`8AxKL` z9{&Np6=3iEOc@m_r51veK zSBuHD{bqS9T+DuTS)wH1_sWIxeK3tWKEjpvVF7c3wrOAuf`)<37~`gsgxlvm@3=Im zXtrKoZ?ku7sZp;UiyVx2&iOm#ZyfvXrgKU8Ldl?-ibglXt0G%$%kua1GVq@7f$>l3 zN^0R$2VDi!q}zwa3$RhqRq@TeU#w(A#!wyePFIm0qf-aV7qLM6KuR-hA55J(8kUtH3TR*T7 z5G@(ePRO%v^`z!c!_u7bbj8wC%{2-1gI6NXLl<4VaIAX@J5+G+FRlUX6-~~ovwp0V zmiX2dIl2J}?BV4@K0Q5-$Wu?2YT4YDCYt4I<@2xOv1RKMOO9V$j;>UYk8AOXiE0)N zR|S0v&FrPIsaoLS(61O<+P7)#nqj+4;&Sm7l4lLhJ1;T5uFsV!Ch^X;9nW&jV|OGW zcAIMp12m7DLnXw(V6gR|@l%uW%I4-srl!06G)<8Ly_6RZ=&metRC9h#*{wiRR<5E} z0(_F5$bCt6HZ}%iH_HYY#0xBDr{(fK$47X-uaB&EwzhZKWHE_>63hxeg8ZU21Ap#b zik?Nve!E7Wg}^d=t*{f~V?I~C`i2^8sUt+lQUTFGo-@}~DCRHsp}X@`V{p}>PMz@n z2RY(F#tL*=*`~KF_qAisI)(98Cp&Z1 zC-SQGZD52vM&|Jg;cfl6@o+dVm^Q&G)fjYX?$GzXuzK=v@x~dh`q41*k95UR&E#Ml zBV@HJSf~G8sIziruV-F*6FG{`$#3nWnzj^=dR`Z6t0dtWtJJ#MF>h>bG?;a7WJ^ZA z?}++}vn&fgZ01Z+df~T3{Z=MF;Ez>p@TPfZ$-RV|=SJoWoCANPkjfJ8uzh;b?`rCqTM`%H=;h-;bH0TcSE zXXDu4{O@A_YFqel_+X)!pk@EtmR?XOaV)d6emKw76e_`GQ8qh1;GODfUacMA9YINR zbsS++HG?0yw6fgcgUbhP84~goA^>+sWS%I794u@REZ^_ zVjy~xVbE%*ci*Aq{($GKd>Cu+pT@|Vf_O^vT>?JN{DAS1)};!%7>N_WEt8|xs`u!Ra$K!m>#*jn z#twjy)FA0_xo`fv+eSD2a1d8Ur&=~ z1@#=a*Z$pRmA@i4`ucTwpD%uMAz^rdBkFY!0&@h*uJD{E0Mf|%?HS0Nx;F1E)pl;C z%`}(xKegPP}~8Vy*QOW8MY zrCP<9n)XocBuxRP6Jky|Y|c;<5h??m9|02`zcO^fi9gzNqerN;m*J-9SL^^T&Av)r zsKhd{a^$U!_)(11+T@$s=u8)x&5HSfqk)0#1Tk`8tNO4UxKN&p!75c0uqoX5Il}Ns zk{JAlG%O!*DC5P@Kh8o91{W7+rm*Rp%{T0Ad159SQU7nx7Y1%o7=E)W!@?FEf}DR) zGyQBjS(xF*^c6v^r})Cl@4OM+7>+TZJ2M>{SDyg*i$a9(DQkXsj?BN~FIS*(rY8od z&&&!;%u6MK0m@RVb7F5C(^sFc&Ns5hr>I`?1Fw#wJ1serS7p@e`%*Y4vw=G|Hh%45ht#pGCP>Gy`1k%8%dDcx3=BA8Lh z$qfz|vB9Ip=8x!&xphv1S$n>ng66Li_$zW;L;oR`x716`WX5|5txPuELi|i}vi!6L zI`k1R1$K0$Mkc=LDw64gvxIdGE5pn}lSijN6nM)E?#^>wL@hnjdpAA8%~fxTvH$g- zA?R)PHwN1MAyh}?-ENOeH}7{EGh_+7SRdttcQp2U;tRJpJ1ao{5PEA|qJ7jDk;V#H z8>XJ{>sTIVgXo?2JuPV?z15z}X#!vLmqb>Qn9N_!A~PQw#c3KL zRfimTupR3Wji<k$&%lixfTq1t9Kg3FCMMoVaJY25q?qOSP0a>1 zf1b`=v)(`ax0836sf<#~iX4u(= zB$H}|^@=3SVqy5%&e*uPrJU8FE1NTKYx%Fm<0||+%jTyC{%OB%{RlrNJN)f^jl+01 zy>qc*q^VbJw1en?^~%_bmzX>W?5D=IK9?)AQqaoKg$lIwUOlX+TB`cle3!MCYb}_$ zDX1r&ee>2f zOPmTz;`IOWSooUB$D5)U!Ju%_*#rc_>?VhV)#8Ja?_Z!(XKZbkM!=9_dGw0|LO`S4 zQ|TZe-4w-_T$Vj5qTFhaD@ue5viBk{1a#ZRxav6I3p1U@PEASl+FhxtsEH?Lgnk)> zKQ5Hibu*d}(oq&Zsjy5i@cAK$jD{Qo@km0IewNJd%8hut4RS%t?x+Bt6=LPm{Y6bn zV$$oSRF?>drzR$qO3(G^*#d=#7C4Dv%DhQ; zF}jK$K3qBTx02P7lQ(rg>!LUwl4l>*OWf`QC^w)xQYRW3?Th@~u_uO;Ps7U5mS*)P zy%`qfAOXS#>x(aO+?aEg;QhxIrx_syi*{WKu0lhmEM&>b&ynnS2VdYr2lg8=riGA9 zXN>m+VQbP7VrI8dUU2X70ask;gOrpMz3wcw8_zVdoeWzF^iFAHPbgZ{mGK(;BO1W; zN&SlFJvoNNdST($0}}oHAMFe#b}uUfD&M{aj)>ON>djaI#Ae zS8xBpR?siy3sGZ-EIwU&n`8bM_necLOBTJCcDy2Yo5BO zOxQ8iZQk;U665UJHtyZ_=IsuVZCr>@8E<9KW=3y1;syu$+uM)VDn>Mg^1d!JGoRIA zadv4bKOg9*vBYuX;X^|Mp6efX?`|gEpD(6dw*n#W2Sn+J631593YTG35_!F8H~IN!;|HEYql56u^m4u z-h&>KDz(os(?6nLg!i=w?^Prf%N*u zM$tlVS>Vn)L5}Xx$$3D2`&Z|O?m-GCca3?|S1heLbdNqxy^p)6t#LL5-|^)+j-*0z zB*X#Ur)Vy}>BeqAnb6ZS+?9z6xg5tn-V(|PfH%jpXuNhY7R-DZ5NB!QmVYyV`hzx29rSO!i$94|E3z}PWT=OB}L*CpdbwnjkR-&wbf~k`Aormkh&Gq_{$CO(h-`Z>c9v3b+hIozrGjy@P&Z-dgpD zVl{rR$Fr#FYI>yWO*leP3VXe8X@oL1duf@b(AM5w`j~L|X31=yk5!|k8=t9ZvNZ~n z{&zh62T67H$dgm72UE~9P0gQ0yV67#YbIKQnYcyz)|YB0T0yO#K!V4i{0;wyyj?;O>N2-xL+eq?U*#?Fp+o+gx z+d~R@0+*q!1CjbwH+)w>rg?2z`*SsE>gg@NJKlM{LnwB=u5HbI5XrJAaQUDKnfq>> zIjI5%FSLwU7ZrVZsxC}OQmQM;$;~hSct>#%?A0{Q z`0(_OB+!iQ%hY~~7qJ=;F|OVa0I2KsAqSif&%$VX>n;DIf`zLy5=W#8CQ8@3nwd^# zyM~7B;w!-4Q-QY=3K;9kLZJ$-8Fv>Aow_>_p#oVc_f%Esh-tngWpm_gWn>wfC4Lh( z&Mg?pCDFeAv}HHT^piIqAR4yp0%&@*pYR-B+QUW?HagbxbQIYNqXAih+nUl1zK9=Q zz66A+mI#@2%94bvFa&F#$?t!r%>Kq!|LfNWwZtggIlb__q$5^3E;jZ&h2z`{HND!j zeiN-f(3e2^;IEI?wJti2PG&wfXZ=y2IxkfT@x z4-IuirqzpUajp0{lZ3K~0ZKiPnr`p$X#YM#%`0tfP4K8|`H4@2QY*Zj83CoRBZKP$ z5+($9DXgQ*hH`#_?XFEOH99fLS^PlE7^MC3mM^)jeSdq-a(FaJ*=o3|-p0kQ+_Wd; z0{5c==F_V)X!EntB4(2d5Qz8!c3~YakHa2nu$U z+o-s_mTfA zKkp3Om??h9e@Zz!8?l;8@6_Ls!I9uN7T3WQwPJ|qA``JR}l5=yH zzQltx;|cTGSrerCP$9i`%akZ|WFx|JZaro|$-8f)`)Z|0olwd+{ zn_|w_5EK%-Iw3xFHH8}F{*0C4`H9Hlj2J=lf#P|{LlP%%+r|LX0x)1iDDs!;{CTri z6*>8J zbi6Z6;V!zfKY(^Cy9G0RhsovvSU*CgbYYgMYG|BA#N0Iyz#52eT=}Zzv$&X50A7{z z#p(U0v0$WtKO)t&^M;>z)bYfg&ZrXhZtXHcT=Y#fHTxcyW|d%Ywlp5{R$o`lzk|B( zGS~7qv0=YY3_CFEw3ng%-m$%1nEo2j+*yIR1-IFDMNaO8H(U;}{vz*~m?ZVacD(7- zJ!LJ4Q3>{L#;o)kp)p;K))NLo7q(68masD`L3GipkrWRF;L##`d?oIg^?hOU(pyD* z8Vo_=n>sSx`op@iY^CQ^xRM-x9m1Nl!b|Ns?$QpFALk zg`h?&O@R$j&IRw?BjP;9$9ybJ0^izlTGVXQZi^Ia*mO6j(*NZ6+c^d2z=F?AL}q7B zo{>?SQp0aBjZ9_+4P34d{*3Dwbh!%hWye!gx&0W$>?g=G*xg&yOIa-n@IA5CM)r?v ztBds~()u$^mhuM1eSLk`Hu6iEu3qiP1jP#BuU+rfrwGT82A{o8UA3yMmGZCOco-$2 zA`=u~jNVEcZ7u58h1~xjDq4nSYs1N1&x~(Gbh#yeqXRfv0}_Oft{i#1D>HO)C~GFA zDEVTs=#xTGUT3JY^PG#j0^iHMHO`>zLfdxbdsOww`(b}Ju<9N-z}o##%C@`Ud=Z2jFadand## zg=uuSVzGJ4jE0pn`{7dNQ3LDVC8A_xj8H=Or8MCVW&!j& zhbMwt-Xxzn1#x3;-wh2rzWYAX2o;l*;Kk(@rlfwRnWFyu4Ex@Es*9 ztus0QHC5v-5ryI1ibE?_E4pnw6(onX@X(K4OdNZTZ}Jn z?XJ^K^Hg3%1%e-kVY+;Np!nAm2}4UGS?;V|%i|2`el(Bl<%t0HFuyZC#{tqRLQcuV zpV78*Blj&qa`363DsT_gwA@S{S#^tnfFZK2%SPCj5&D`HVu#;c2T{&+h=ArlXp{K< zgxn{;r0&mC;#~}B_Mh(h{vt4Qz45t5E#u(xZI{Z8hi`VT#*#oXC~F)-4*`UsMa|2{ zJgR!`h953uQqKJ`k`;cL?he_*R)Dd;tjShiQgr$FiSW&Wee_*r1l@P7mzZ2)ej8+{ zdw^d-NuKicpd(~-W#ddF6LcpKy<3I2V+@A$rU@XwvTl9J$*A ze4 zg^hLGvUeB_MRgJ4;&-N;PDYjI&=|-I({=FPQF*B8C#`00|Bwmsbab*U;v9AMc*B$u ziO)lFc;F9W5@cDSp6~y5wD!wgguf@xo?Gv!5O$a!lF79U-@X%{3dP;?Z)}G?;UTD~ zCgX&#aBFvEN+$=<*z3nJREr||z-~kbtzhag9F790znhiXb8F%59$)5&9CZX1AM{s|>~=e} zI!`vN?p6MT#4X?XHFht^v8E#YQcA38#Xl7Y7>xNaE=Nk_i$Qm>>-H%Pxl6`&w64gi zGE}710PCwIvZ5ErSbX}tm`V8$JXXh_eM3K%(@oajO>4MQeUhxReDp*}VHP_IW;OON zF_JPW(?!*- zN`LVRlgMdc7(`~1b(H9Vt^PaXAH_&BL#@t8>@F7o#QA}Upc8U=tn#3RQ8*-bp<-Mn z-=2Zw$ahoRWq-4>*f?+sz<;Nf946n^T9c`?`^vo+TcS#tCrom*b#z1u>R5Lef-pUZ z2&N>Zz!n+*0T{f!-zK;3Ou_`>m)PUV5w-$){n2UOiyGBZMrBLs80mHvX^xTA*=^87 z`3Ks>2syFdB3I*z*)vjB=M!982XLXx?P(EtANXMd#CA;kht6O4tZTf99_uikl1Uz0 zhC@0JgXTx?Y{7_Aw6vCP)(S+t%skvfADxFsa%*LAUJ@ttElchi?}FDLsmDm)MZw|A z+fl-Adkna*+tMwdUkNe|0AqOz^BP}ur#rlVArN^q^o-1kT661;xHRTf`G*4rIN+6~ z)CpRL&m-sG_r28VyyV;|HeACnpq9aFQZnsnU)!QptlvIvj?@P|O2g>E4L4#Mq9`|I zL`h$HeBx;fj+hl;WFp*o7fer3jf`M;usijcW4}*`*6qreA7_yrS^nm^20@xVq{LLT!H!q>G1PDN$#z3@ci| zk9kc*Njw5wllq{*PcfmGl^z8!919-N>;38g*ivzs#2q2sbxZ-o-s28xSoAmd<`HO)(JUs z8PnHI3=Z^`f={0>I^YFd%6b)8$L6QKIol%1X+=!Ui4L@$2kwRyu4f>6LA%#!SB`|V zkV$~j%S0`D_whnt7QqMdaipt)*(N4s*oSxaG|&y=$D5p9=%V~L#-6JG(x-3`+2E6B z{&uS4xIXi|v9-C@{Pz@HVtHSF>)atDAJ5_5u7$zXnztO9o;07gr+>uUqyPc&q`SS@ zuZM1#7UG5XM_|zPlc1z+T{{S^d!GCyurDiRbH3voe)3hTh%N6MqWk5*1BIT*?!S!N zlhc4#diYk=yK*GV;LU$~2jhuj>Um~X?~Tm2D9pzCnZ9YGCfM$l4Xxx_9-)Y%Wz+ts zZEy>hID~6!(ePu(-@MJOI$JQZ%{mfoqAO1BBwDLa3nh_}EkXEs0TOWFv>-|llfyE6 zDNTr5j@$9_yAVJQw$*aR;BA@Sdh{N$k#l^Os}brs2n+F`tbHUOY;EGlg>D~(zb*8@RCun@341R-t|Oi>d6L9H-Mt36jk@p9&_WWQZa)`pMF)T`rzU85{x%B{ zs8$_#kJm&^C}GZf4}JeLwHs%g-`kd77Twkra`r^@HjYYc_YQb=NKEk!W;A(*=8oa+ zr29E(3@?(MICQ2dY@H*Z8Q}XHy=rS6jf1u;6?~4MH(4Rd_%R3R@3y!SB1i^KOStqD z^R6F*@0Bx?xr2`WTO|u0wX9~?LNzGgf8A=et~i5*%Oq+{PMJuilEzKDp2S5l@YY!r~eXs3miA`Lg57L*|@ts>9~E;bwP9?@C{6ppGsL-&SXf-cp6BHu?wW=qGE4vLm%R! zAY)&97wha@hDYm(CDK~}wVi1^2!8^F&PV1Txb;>VXsz(QEw82Ejc#0 zzt>k6{l?X`xqi%TD0-niN?CE-T1Cumb`G~*%^U3*Zn+6TNLjC#j9ej+RuU6eUQX7} zpLO`)huuvWSQP z{ruLug7RTBeN}@Z&xPdA3{4wjT2{{#HQO$4Gitva1>C>BuE?<0*I_v`;fwKoRXQso zbho!fOf2Y<#+h2b=vOcMoksxv)K^Wrw@)mMBOrD~a}am0DeNj&%cE?5q}p!fMX=9( zbO7q5%Y^)mTiXW}ep7eBgYKq1sGdcEI@10xZixbU106>gN((NU2C2}WiQ68yP;Nj^ zj5d8+T0kQKw%z!iipr_{!HpqdXk}53Iqx3;TE*8MxS@CPx zW%Muj^W%?m?#@rh#W&LFjusRpI9^?yU9%k`PBa0D=_~F~ePwH6 zje59Y&_NBa1mNYnaha3(bALCZd?-d$Yv<>>qCYZ{qKvf~Tfbv}TOLz^0KJ=yATGm2 zr@jii#hHUclSQLB61vWz_}u&BRrvGD*TbGgm0Y*6=~63hhpY zo*=FTM@jQ>6+ru6ow|Azl*k^VnnJCn*yyWUnPAEGcdVD%x-zrRq0+Q9X6R;jlHk`t z_zs0fwY=J(usvGJ_<8=%`6IvNO}CCrR_YN`e`J5=yrvJuyhvd-Pu}LlOLfA!93T38 z1%l%3c00VBD11>SKknXu($d>F;vAutq71id56>c=4)wM+p30i?NfogL?K!{g{>SO& zy!5%fO}RvwX6vl~aRD^e7?aisVtXnGrsY761D15IXLSXpP&%ngLR@w@F`Fmx)j`;! zFzr&n9^F@+vA0(4X%8so7|)lU8!*}371?iNf52cxlSFI%BEo+}bKoLtv;-NIm9KMP z@4*z=rL>xut@Ghs27$dBUrU#`#MKf8APV9HJ?9;q4a?0s%*R+4h1JE53?^jk7=4y)Poe-cTKm4G7H6<*lcA1 zu+|BZ>W1C^x2E1GC-e+gy$;yD3Jz}WNo`}lW?KILMA3W5+^Hk1u-}Mw*~Q-Giv3~9 zP?B^v`yJwgTOuyA5Id0<=LWKQ|Lc35>4_98O-(h==q2sMAt5v5HppnD4DQ`sFuPel771V-OMP-kd%m%bCnC+FV1%?=8!w%Kn)q1am-@l( z+hK3UZi~LH5a}m}Yi^g*i}?7YXiyyqIZf8j2!bvj4#VJGbpVmf502i+%CfVysf9CL zq3v6I&6A+l2)gP4#_V>r+)i}^iaRpoS?tL|15dR!I|euPLHZhVK3cScPHL@9(Wx;o zFz}wXHeGkODuAj~&$|rkBjKstaD;$qvh(}`Z1*LAyjB7Pb?tvPe|6-^U-D!OANOM6 zCmgi?HutFlk(wU)ipTzd5u+et`>laBaxn&E8Uf{B2{>!QMYlCxOf58gLjb{^1*$wE zqbQl;!YAb#<`L*)j=?t%D1#VSO$8pLzkw;H8Esc!&X&LaX9rv`gRG)g2G$@Un zQ@RgFxW-IGAs_|*o9QDpo6+m*U?E0!Z5=KW`&F})zwq4$`hJRe#R)g@NC!NB!de?QAL8MEUIn!XE=eiKs`DFko zn{#ijjw$^j$&Ooq9vq2CG7ZXG{J)F+_SgBqbVNQVE_cbRg;QWh)rC0iC3pW^06@#~ z@YJ>s5aXwYFJZgWZT~@bW)`&W^^ILa@E?qeBb>wDz2+&RRI?Jnn?WS<2UbX=pD+T@}*7Uct18<{OUn+bL_=9gS+LA}+ zr$vx@@~pXTduBXZ2XAZEQ!;Dw0ue;S_%Pw8daV{mlX6yypy{54)<%B;{()@WqnKi4 z@Dk6vw|5YHhalJ7*5xGH{l>XWjXuOS=d;K`{W6z`Y@O~bJ1*^)Tct+ml&U38b^y)) zStrjQ7GiqM5$!1Xqgnj$)w2p%)zn94gOO;)zj`7s*?}V{kpsWIYpDH```)jK*m&tE z)7Po6b0M~J-{rTm`EL$9i2rEv$YEiu^L%|ykG;f3m-5?9A_z5zTl*yCVZoY(;~dzc z_D(}t`dXq*AXd2YF6_Sd^3NchU|_%)_U_*Hrhn&6Cb8TWkTUf)j)Gofu_~OH)tmL) zfj{?t5W`BQkzP97Pn_IXim7Ch*3@)14I)ps=xuw0%BD6p7Z#9QeaoN^fJi$BU63q{ z?=6&bTpvui61s9^P!R& z1EH5O|2BHzIH-~B=txQw_ z!m-#ihh5J1p>99O=l*8Y71m#mXDB=L!}}c6wa8jFtBI)a6Bi#79xAgl{`!VsG5eKqNe`4>Te&H0|oQyf4lk#{uQUbk$$*$UV%g^*~JVt zq+p{!{R7|yHvpS(GiQC%vSd5zUAuBeL9(r*y~X&2zs~C0`2P79AM5c7sJp9c^@GO# z(;BZPck@@OiL3at3=KtyFN+-ZDt~-}?>f0G`S?w_2-~#V9T_77MgK%xnOyA&*t}=| z#3)_ab)9aVE>cTwCI?ow-$`}!7HE!&4z?p)|3bbw0(AAgi#xwJ4&D>N)*bB)&o$uA z0=cfviiSmG-(`dNFu6X{&70q2haZmMs9bb%EbM+)!85`lj+1pLHL1-BvxOOn;I%J4 zxDKID8oKyt0piB*IpkCy5c@$u#a^8na``6~DA9jw6CE2(i0S$?H zMO#cV2NG~LAlqL5F7f+S`To`5AZcgd%aR7y^R0(zMub*t8&gGZ!1#FP@1%Dkb2M@ zuz~QdTyI`r`lKB49RUKy)ax^$96u8_@_f>yK){zO5ys^YU@#BdeUU^_bk?D_N5K1W(o zL{qqkSQ?;q$=@wZ)~i}I`KVog6D2}30N2Vv?G#t0X8`86qlDI%B>O^|}dJ!JxZ9h3b55Zb=btyh4 z1=b>S;@ncN>dh&<$8CrG13;G*HO#%$X{oq3rbb}UbWuyYB_1ZNAK&W zh@rL?sNiSHKGh{77~N=nHnQs6_KLs&4`^7ol8>z8;h@CFR=Epks+^ zO~q>_*{W8Se%R2n1x&_S@>Lji-Ty3SK6WlK=p4$SbgMH@ih=LyYrQ^iEkFhX&*M-s zMRf55O9+WCmO6eaju+Xnug?Sg391O2A>ySKKx}@5xyg}Jq+q+1Vl*(X^nAQRgnDZ$ z!ul=h+Lc{*s+~&j&a8%!!h4P7;t&?SOVJqS$@19*kKIZ|I zB_>Ak3^Dbn*{u&xTA3jEIp+i6-9-~|ym(eGWG(NVjpFo8D(sZvt{kLh+ex7$L7d}F z`O@fk+4+PZ-ObfEV*kvPG=1-d+|v=zwH;2cxn?#D6O+ z@1c)>!tGB_%v23lh0!ykvJQ9aGz=YSb8+Is5Tp=)4`DOz4~StSGE{xVWof z`GN_1cS4lwU%W+QAB-OwPvv28BMqWT7jqsx&|0EXSa-ZzxGTUt2dzRe83`;@bd_Qv z9*t!r_`^P8ZmuY#zS*9o75f!bkuWu-4PU3oGGKvcvLr$${tzs;vrArjmxnks@ zx)eaWU)=HtvEKgV-{SDTWX@2r*p92Pw@|N&W>5%1@fKfq=aVHp5{2%N|ChM$d&Tvj zN=F_!GG=Py2WNAfZV6~t?mD8+V}F3LZ11C9s>t5n(EMSVuF5BLDc8f(iLTv`* z7y;8s)U8b0(&A(pX4gx!5cT6-+V)PL@?d~kY#%HS9j`k<77F~%~ zjG|?{Y$5A*N9S44<%dmZ%!HW4Gbh_HXCn2LfYTfjnCWKEwA9r|0N~!Te$|q_n7_R2 z37GOBUV_Ogl@)NRtjUw7iJ#$#QlJZY%FtW4!zc!h$Hw(@>P07sGCT+}Mzj@J*jZ>Ayu) zu*o_M({-BHKpp^)>r#`(;7O3#Z|I7<_Sxs$pZOBoOFOci{oawATk!Qens&pX_k|JK zrBSDV{gWp{CY>+1Z>wtiaC^*KxHQeh5R4sj!pJ=fg5NaktX&j+;IpUi_xzTO!~H`- z#tL7;%*=i?1W>ml)NnnGLaEbgR5Sx70O2T1&2{0<=GCRjorIX8`8iqI{!1K5@ke>( zDtlCuzhhXxoiMxq<4M;~j<6I1eFfr;*3otJAVa$Ic#a?0E}*rRX$$WWtMF5bmw!$f zmOCbd^XwxZcL7d77p-%*BOh&IZW0^7e5yP)H+F`0pqk+%bhpEL=}kUQe;{2Mc~K8} z4w`p9A>APoLJGuCs_g7aL0OfyU2k{U_Y+-Zm~<&Eu7XKBrl_su0WdX3C%(OXC2w#k zOtz{!{$`2c*@}FCr}P|`ooE(NG7XIo5Ca?|$CFigeAM`js8iFk7fV>Sg>zB;5$P7% zg`9_3OTphI1qjc{d z*2Wt3g(*`BQ8}{$_{`6qLse;cqouR?E5TK_Op228wRKJdf~sSEtw*D%B$n@&YCiWD zAYa`*#N0%s3^E)ng*yXc71@|Nov62p$rdoQsBX)L(iAJ4`V=V7EgMacEw`}zV%(eF z3WI}++34bMEau$XYp*PeFq9aZ;o8hN&(Q&ti`fI^?WnUF4-wTBfghJvZD*IYTBEzD zhIS((S=(I3(-m7ECb3|?z+i3~us-JPPkPpy@;biemadza-Lub}*5^XRl)<*uZ7d!d z^baW!oX>V1Uw+g!t25A6hW0tc?D7^Zuwo1BSE!V5JTpajUe0RZs(Inl*C_$Ogq|FAN_z*Qm*EPul4n4B=JitBJ3_2DO zRFRSfq^5OLU4T3Ps+anI&!SxT)JcG{j!9P0`%$$kI^ zMzzwBI2*hI4rXRtNn6C4nyLn48D^r3t^>QAQTi}Vic)rX>|NWE{I%m#H|GlY!XENI z&vmWt8$466dUiCQt7mk1;vA;4QNyuv*PRDrGhG|&BcV+FY^$mZE3Fvc+X^ok$JmCA zRL!PO^Q`L|@d(rio#o&+lJ^=8w zoV!2}kJi;M$4wUJK|*DDrzzaT)4b1dqW%d7B#@*oa3Su!<*<`US&A54nnVe_k@XTX ze&<#-M9Zj`L+n`Y8SlUwPjz_DBI0{Y&CE=f$5Z@)BCYI?@4-wcXyi)4V6ak`G!9U0 zIL|O5DRJ6B4y{s%jsc!R+)n1gHD}oyuCBXvpgOv$|E5t@Km2WvQCso zxOgXq0-$lG9J_b%S}6dauhi8PV+p1Nq5mEkyEOQ8)-=e^{@>l}-T1l(UWWLFNoaZ0 z`4Yua?+VtK^m4EjhP5jjrl_;tU0B$gxEe`Ot%$yB+EJ9( z+B_}W`B}u)a6H)&HC@jcnl69aLlk~RdRQG`xAm~dzz=Vlr>{2FsXtnE96EyLJBeo1 zl2Jk$g#P7VCJnV64sFiP`Bdwlu&hi$*U%g#3alpJp8>T0=9)d|bAQG$U&;!6vCkyH zo*=o@hO9M#cKv9Y@$0ZkBN8;R@XYdA@J`F(?%0os_kZvbMGCNuL%mP8B}g$m?uMN) z7HNk$-;O$TjMgU$I@o{BM+jk|ba%+zptHTFg9eNct3`+P32<fU>|70A16g)~!-yr}2AY2A5|0nHk*3T*TF(5=!a zF=}&S{pg(zPsZi0w;^glyxPH#Z_A~->8_@Voo)@zd<(I)fA<{v)rfstA>jH2sS31!KbJA-S3G=6}27$0j*)hc!2R^~fxpj#jiP7Hq|!*)(l z*(s*eWeH47OiN!r`b=yMJ=ziUBg^`V6pB`iw$L_DDTx({ztFUciiLKHLyY!b9$AZ{ zpT9I{4YYC%4h8NA5 z+L<>yUwgLH1$Cf z$=bc|MRvP}TrPK7_P6)Tv8p{oX64>`YROHh0nW>H+4SNkeU5YzPa z_BK1|fL3fNy>>&A8?^RZoNvu;@teiVz(Urt^YUm`y%STJdS?TO=L>n{vy4xW zJ}G=PmpU%rcWa&RwFsFyqT*i|rI#o@rd%PFCR>e5Oxd%`vm#N&152|G} zVTg!}mmA5qFpwc8GZh{mMja7lrl_))}|JL)B3cce?NZ(R#B@iHt=YLnlr(@WTS zbzHvu9q6bo5eyW9)S?;XIG*&tW%RUS#fFUMYR{06-@5} zWXzW2*=mL-KpX^!@1^^K++lD7L5s9D22T%LLpXPKU=KV~HE@ z_LkWduWo)2k?Z3Byb-q}Cq`CSn!U3c!oM`YvdyUzVaV?~+h|!6R3GuWNv$y6ri;2W zU>&l`IoAan1fy?!u5Rjs8FJ})NmDB|etLj=TiZ!N(`7i+rAO9)l%z_fv(1>K>VJ}Y zm2*R3v^>~H7p|Zk1GHKzihTarl%i=H{Cp=@3r|@L6~IA*J(lZj$@LY^w7UEu<9%k{ z$ydO&x4orJXuq=aL)Mu6nItb;!3|XyCg>Dsq)VR94Ca+qxEH$MSw-0EOj@Y804v0pmI=afu~up;mOX)v{88V)bO1i{o5( zK|(@dDg0Ldx<=mFaV-#9Hv@Q5!jYnCVk$9)r~c)g8biB*#t#!Arb@z&NRz<MGgxHHi{?AYRN-0$+I1)ji9N?$)b~~isw(q#_D?`**P`aUKjRe6$Yg( zX;vv7f-t40={LRc)&YXG-6{g`18j}L>H&=XSq0alS~FUg?|4^7IrBz3S!nIPGErh2 zrA#k5uhVqp{D9JQWLDa2D$5*mD{CiW^F!B4$o&NN*{yENG6~!ORXVNrSG1#2)_Nv& z044U~oikVL-a2kPT6jXIH?&I(#s&;Bx5ay>zC9msjD(5vhH2cS|DWfvW)uG_uRIkT zNs_hebmQU1*t&35Ldo3jG5F|B0r`lND?`0lI_B4a2@k(1<@qOS9_1MKryy!56v_hT zDxShm1KNa};*WByeA^P5D|4BOoZ*o!6PQPRjh4?qAJyCUp0RPDnWoTKv0KEC+u-Bgtvi8?W&QSX_S0KILp!@A7%)mGX`_5= zkjCVdAJQ8nkxNz(D}6)FhccRiFYeZfLlZ4ul#$QPC6m-1cNeO20_8z0fVly4j ziS9b-fK=H$DtiHvXcNUBjlc{=4>uP_l)yQiiL7?M# zxvTq*&OPSaZ!$76rZPK6KiN_X22!^|)Va4abKbqAJSX<*q4hcY8>c0h9!85zzG_G` z;WOHd)O9{;zukGVt#vJ4?{Z|$(^D4>n!aQXA2Zyx^WOCa)JLC!%68|c=H}`gVYo1q zJCLU_XiN0`K--;{WQB*O`@?b{deU`S)ax^6q4VTqQDUB7m47ZWVX5nIKDBNj-!K9Bj z^isWL-xIZRVOkucj?dNgdv}k|&Z+T_M&uQrbEf#?kMO@8e_!`?Y_ z#=6P=coC@qAKC_-Q)aa>)e_SSZg^EIM31}wM$4zuPd@Cb!6Jj^9#p4^iLt#s4G}FW zbap$(?{OhX{L(YA26L_^-5`d>2qjHb0pn|LtaeQ7m7Y7|6f=(71ZQqxC!ei4;Hobp zn`WTQdu&650sW`QC!^X+BG~B-RHT~|Oj08Ob+XJb3r<4FI6x;Bu9`%?_ ze>vS2Z>Saj6;ODKgh92vI<$x+EATxhzOcdmQnEw8$7EAMLn0qty6upl7{!2CJ;cIu zL|1{)%+wT@Gk7ks(fu@r3v)Sj^=28=zvwG4kT6<2; za1%4nX4h%PyZKcD6tlS8#pt={go?1mgSbYg)YWL|D#qgyN1U0uZ$|u+mC5mR3x!MZ zD$dOmN2(M#80MI95kDZ|AM3h|80=518($tD^#8DyU@z(c%{H%Y8hr3KlXJy(0mcTw zF)u4Q8C1HN+6~D1OZ(=|)~pkwU|rE!jS~?0+egrMreVHh0ttU6mdJ#f*mG9TD09Ck zL_zX7q$s~?r_bqlOM!o@lIa<=4wY$j=u={TBw;#$Yk9QZ_ zZ$)3uG??{{frE5B7Z+8YuUrgwSDLn8**0*;+>Z$i6x(a;77Bz-Ck50mkALu|=uG3i zARUKxsp!@9Qe4D4oKpje!VZoa}*_NR2(#Lgd=VB_9sbuHSipVQ;bZSY7@nMtbxjWvi*!*HZhW0QkMvE`S?L6|9$m}0VX4P32YYbMcXpUf#?wt6 zwiN&uU+FIdV@6=!j>Dv%{w#^xX|s_rGIE)o9rAt|w&nIeMuwSDM}OAN)q(n9W;Q1( zTApCKoE$++-29pARlTb*A5!v7oLWBA=`-H0wgIb!UOFOQWc&r%nTo)1fa!0Kzc>13 z51`kN$cIXHdTefPQ(m4sI7-_gm#krlEGyW2$&T zMda%DRyzELN0p)VxlF`{oVY4rOK)$``0}zGkxm_?ItlHpWL4tJAgnkDIW9 zoOT>tDvxsls%5Njw07(94inYRNJh*=a(Zz}qUQZzF*C!I`?&%-f+u06n=j;6M)Lq2 zhdY%-C?rT!*8Q$h*dSZukwP#4d2mE6N>$6r>F$6c%cDe{E--{{1JswAb(E9?JOTXt ztPZbZYY0gO9E`0TcyBFNDnhBgJN4AUQ{X;Y)7!83kg~FJ1{iF+z1*2PL=9kj6ZOOq z5BuBmhO$whXilsZ6M@r7WXI|4_-1!5WgU}zhv%u~n;-?+&8#iF)Niq}XEA~s#AAsb z46>9X{#yqWH0 z6^mGr)7c7|nwqNVqKSMsSjfhV$b^{9hL4GTK)MO&V0*)kJbiWxg^6d>IbrhI2xUFg z%*=b_-h8T}L|gfwC3co7HY=-n%s#!~;-*|3pu^6*uYDX^0I2IXhJE1b!Vg<8#1dGk zw-uEo)ylHb<_z+o_HU8tIwf0?k4!#JxCDWs*2X+lF-K_7D(UkH#;4YoUZKCZxuq1D zTz{QDuoH0;E%4dTu-SFtNzn2@ykZ!`%H-wo`meVZxlcujUdtR@S<7V9E_413{1f z9=c<-oG!6pefP^ur@83p^kgT^s_@ooJRo#3kv;wZm7zBonZMOHuE6~(&z4TKuA z!`A%QYp+NzGjCTVr}-#zhps3klocVcA&@E%$1K}F;h2t8DzkK*JcOAM@ZE4lIm!50 zm$u;2_2dAj#Sy&bJyzD-(Gz5eT^$tOXj|4v^m7fM#fLk{jKA;jtI%F*oxT5BU#Vof z)zg{zcARY-Rcq@ZE*2N8rIf;24`Nh-MPF=VF&1MGH#8Esmp%M`eE2&80(Aup-5*7w zZK|gaG&|un(J3j`#z9P8)b4cv*XIf3(K=(YX!76>v%+RjZLid@gx@IB5Vu~|e_Qc7 zt{O>v=vMNJx|aYsg6ta%(YM@D7bjoG=%o6r8nt}DT(qrP?S8HQmpMe%&aagCQO`L# z+Lr;!kA{jYpO;~tHdnXkrHW~BF)T{BYuruq50 zSE1TLhOs+$mP_>*_^y1;(Cuk`RL`h$3cWjKgh`VqMNKXt77%FkQ^@*A%T7LBl4}(?ScnvFi|&~PGp4foZVtE$_+r+_(39!dVO#6D?xjeK zBdUiEyP>G#4U&@V6(3ggc8d$d$|Mf?Z}}RoDY(C&@M(AGcwz8Fxd|uz(L_XTkT3?i zZ499JeLjD}b!h&Mz;bR`-zg5E2ycTjj%&&%yztO)=o#{(+yCgDd8qqd&IRN>=bb)( z?!(4WhK(Y28jbXJs(4D2G0m`2N6WbZ^fGbAbX$V-CXdB|j;#HQ27cg#_|}u_&znNv zr_;7OOnglf ziiy>n=-~Q9{^|(Dt*7U?P@veCc=gp2U`M9AS+F&SM%)CsJ#&4@3?!DP{FiX4pWS&G ziI>%pF1eh~1S#3!8YiQ7W-2?scDb{$@$I~Lm$BX)#-^&eOL4DFjcR+!D*Ni^py#go zD!TXf@6IiXd)Q%?om#{+lihA{{9xN*j9P@gcFEthb%m~z3&JjA<7LH>07w?;w4j>d zTHR!Pli5qKF@%W^wYAN88`E(QHl6&CNLan46o^ zgmG&*>gnAmTlwDNfln76J!|zvQtob$lWp}Ne|QY3x79GTwqD`Xj?)miEd%@yp^i|21yW_sUmFb^QRS2trAwFtl|m>xqMX#7>$T z!LF;E_Wc?#wi=!@R-Y@K2b=-1CAygIFJKyZE(|8NO zo3>VNI{L9cv&K1NZ%N`WBzl_A303-4tH+9tQYiUc?PXY#i~r}w%4>@0y1894BINViM^b9q(O|`!x-Qt@p)yVRHM$6)7GoT%Rz$+3pZJ)2O_UV7XUz^ z(HbKK4Qzdmg`%E*o#9p)rDZp9fiGTdep94g50LTf6&h?#FzuVA^JR0ONibt&W?Cr>_G;Q$;^R?84muoH5!isH$+5vwdQ-*b0;8{KkXT>kZ zS|__j+SwEyW;o3PeEX^~u;bNUr}=e<-JE27bzilZC%Okr>r9IU-S#6Rv+OVC0VUTL zdeyJW!_RU)nrH}-t?C@VXl`miMCQ0K$XJA)lNjd({r0WfYsett5i~R<6=l)E>Wx1P) zLu~ZUh}c}v*-28><~MY|zv6kOnq$25n#CE*>el_E#+xeuUNS{$g4}AF_goMKyI^qR zPbI*rcal4=&s$^S=#(F>ar;BtRwp&Rd$J6xa#S0y2K}{JlG_U7jkrMhs+jdK$!&!6 z%>7V(Lj^?D?8R0zYvbR=z)_BjS~Ari0E#!*80h>+1~b${?=?h*SWMRT!n$k1x%jkq zK0WE-cT3&(kr~K~Y`#mTEx~F!QgpB2dY$-ewNd}7M!m&Ws>ZA3v`2H#jYj@ip0|2} z%T|ZFnDuAmugzON2LPrI?2|Odef{Ezo_XMR>_A+>Crx4_IBu(9m*@D>k4F91WId0| zADs+__q6AKX%70dsi=P_%VHN*q}zcw0+9F9RqA`hBz)(;dgBb0n~FLc&f=$L+uyxh z${2zpa846%A=JKjH%I6bXx*fjaQ*T|*lz0PZV--7*$X#;uUf(xCLeZ1P)CCHHWw3l zZ>vrIkh!vRDTw{xlWm0H|K%GefD7Nt%Qm}F;Hk3Ta&s^!|9VO2HHg~MH|+V$LPG>c zqasMMyi;L@f_sJ#X+z`>*yqBAcH!dB?5n&?mNI<$nO8z5#uQAGl1D?iAumA)lg}YK z>xz~LZvI?|m6UUtu+@1fvAyF{b^l1|nfV;Abr($7D4|RpG6RB33b2j7bjv7X*0GI*!%^v_W7v9sicbVkmgYuaf0g-=5 z*Wm2k#EWkOzo#DEZzdHGIa=)9)e6w`f3$;mIs3|My|pA7sR1-f;?6ySG~XZr6GUnw zSddAr#RPgAJ|zB5Xfh=k8VWZRgA8?oNU|(8+5zfvLEnLeXMgizx%h^O)%tq-JS8=J zqByyz?r~DOh+VFnJ;b0xxZ{HzIQG)zrb~LghqJU8|BI|VaBRX7L|JK3%-qw6OMOq7an5CxBxS}{QHMKT*p}&9|As+1SdhYanI-YQe;QCd9-wp^P zao2iKaQebSLPH$(3?EdVbfRB?O=Wv&q(Sz|ZM&+?vTo+o=3_jkRVE&P@8Iy@!EHct zQj+fUh?G;b{jAU@5H4VrB{>*{%+Jqcrig55XuP<>ML-;Z)9?xhWOB$e(l?FJ7tj42 z6@nTvzN=wQC|D3gzq_a0)VPx^@Fl$R&!p@wtY)`tDteWjLU2{7u7%y(b`F%`e}2K} zlu(0Y!oRz{Ap2dlU7+yG%fWd(@w_lZczy<8kIIMlK??*TGdCg-xgvriLIkhGf|#(| zwwhZw`TKBmkH7iO6R9CamAk5wXd0b@=0zQ=hY%B^ZWwjQ2%={vy%~Rm}qE6za>BWf!OC^Gf&ubGSHNwRg!j4;OGOJQ3RokZOQq)f;xh zTm-3NV@pdP_d^=GE1gU&Hd9tD>O;#_U1uxa*T0zgIXCw8N<=HN1}#+<6oD+Q_&KBf zTk`V{Y=7~g8nf7TMmo$9c?oliT1JJsA9WU<iRWvKd1FDD_I*t^5t-hbo9CB2x(2D!3x^9+cxMQBAV30HPt?@a z^q%wWK@N=^5-qU{RXH)FOH@G zCaq@&F8+7xv3I^+FC^Q{4A&dl!1TNLvRhzu$fQ9_zIsY|Vn4P2-o^6QW*yCmc|M;B z0I(aQ5*rc1_M3KRgdcHYOr8e@-e~#!*_WiK={Z2o>o(1W0A?V;jKAi*e_lP#Go%LW zhHk+gKc|lO-;6$25~TH)wURZkxd4Uk-PzQ=`0gzKY!sMf!VPAyG4?*$TQKt45&1gF zPVF#(wiy8Wg~6R?aiG8N+r9q^<%=YlL~2ESMNfsIlYh4zN0p+d?fGMqOSp3p&XzqV zt{ii$s#08dCg>MZF* ziIc;Ht(_+!O4%$kpdutzux4k-^8HLFd0Bne`YB%$S{7K}J4uDK6nC!_1GRgEm;QAR zyxf4X>D|sC4dFj**!Z>u|7D56fXD5{T?WDpg8mx7ro*wIb~Mh|{+n+2(MRZE?x@LVuLbz}lxBgS-KU#c^mS?w+6boT*#izfr7wR{`er zqpdB?e)Os0Qc&eLt;9$OQ%mBLzTL;bzIEdc6Xt&0eJZ zLh*f-=O)kc$AE#lk0dKHNL26NZ(;}in?k=>MZhobp(SA*z;}|UAawb(+UzSymD6U@ z*ZYGBf9~r(d^`cfZ{E9mEpEpEBEbeYL_P5t<8mv~BoEk|*)8eer9>E=~#J9|r`{G)bpUdsla{j9+YN!)=hFUv#+02|y|IUe> z)POFTuYjXKg+bwaNAUr)fkrImb$_QuKDb$6-e1URHBo@NYGUG}AS{JAZ3}v;$xbyC zE-8t|nj^NXJRB+UE>kP+`&Y||{y6xx_kXzftULZ6Y;7=kwEx;R-0e_m}vv-*LPyhZyv;-3Arlx2v z3VQSkwD`;r&wBCiYXG)B@Y>y9(9GYUt2wG~j$+t^ru|NSte8p2jv5ZwNb}7)|Bo~gw;g)PemFP6Bl zfh2QNf$++aFUZFLaQRv6H3Hwge|8?_uc#V4Nl_2$G#hFg6ZA$C9Hb}Tr8S6u_ox2` z&*1sRGwR6O$LhS>27v6V(FYuOdHe(L&(_n1PCuPMGd4Ku?!4>XsAU4PFt#JICBI#R zFD^&O^~It#CkUdE%SlewCg5@LG05%IGj$ zeXn(#Zxi&5%MA77=OF%gnSz6f1W)(9@)IPM7X0xq2Nz!c2&d~xa?aZp;LKH7?}?U> z9x3Aey9O2!k)1bKMNrAyQ3H1^ZSJua2=NMk8{+8z#=EN$6J_QO6EPZ5qNf>mZD4ke zc1GVdUa%Kf49BQe9$}5-^%#II*^&{7EPp<{@f!;qPKTTrQlnq7XaZg8>#NVZ7oX3! zI2Fzor!>FaJ4<8+4k&(Mkj9Ll?^XT?cxLuj2L;Utp4_(^LLXC}>4Roz2n9q-8_$_q zoNn>W9yhVy3!B*!!7V)OtY=PGd&holy(xiZ2kD9v^qTf0jg1RHJDOZaQyN6P)8*XM zBtl5C>KZ~#-tN&r24BT(S#wG5k63T~@p>iVw#*DneXQXMwEt(G@)wgIaN7bzsU8cx zD^Ox#I8nJL2x3_6z#h4MQ5eCg-3@g2F@mDN=S(Cxo-h*_K&T_@yOu;+1D-D`I&@e_ z&RPL*WfK6K3h&T1P%YVb)iMbfZ`Z;%G{5uH{$QM*5P?x9MXf0*zfEYIy2IMoa0NBd z6Cx^#lLBx%PwduSXuG+$9KH-3)%4t!nIpjq4u(c*nq#I5`-RbGw*k%YJEIv^bOKA!S`;9tL4kg8&wo4Q&R5;A*rvvR3>b;!Xm*9?EM z^7nvqaB8iomyY1YZh8u;nzF#h|Npc+l0;>@V#L_l91tM*$!O@cVUIUlL0S}~?g^kf zj;xCOtk%9V`r!wt6mq&Oyatoj{J{U? z$_Gaiaqm8|Zh&`qH4HGx0%lBXv(^Dx4Nx2&4TyQ3S@LWY%vvguC@9lqR4>qV0xUmq zG>pxX1O+Sj4iQL-^-V8KacQLQ2j4#k__zO8{txZ_HI~52_%u+MQmJQp$|6py@v6o< z08#u_GwaIU+RFcTJa`eA0t#=q38JcVh8^wg@yH)eYVZ3&!XCI@eSk{N7~Kme z^oczIBjb3#DZ%?v0jpst9W#mdJ z9|c}Ke=wBg^v7#D0uUa(uF|Zt^!Z|e-(94imTeVD(hBWV6OwNZ-}&6^vrbZE=lAMmV>z(j z{zm9{R=I(~wb->R8d;2$CF{IY?I)uSb5m=B<8@8>KPPef%^$DmiQPiOVfWE)Wg=ksal|k2hbNJt%OR>o!1r!01prA&|J{qnyyTEc)&(vu7 zdnb-~$=39e(9m*uOZTRbn;Jh7&xbt6~{CN;VX=>fU`lz;rdUq z^;L|si^X^EHV&#I^mb0mC?UqVj@tVRpF>1n=gjJ4JRiiYnNJ0Al z@iEzp%4a0B82HmdLTjxSIb?LVjL-+njh8(zH>8eByi+=? z0EYl5S;3mOg4%vv)MDie`6gtV0Za|fUj{$Asd218o|Sx!GLY81cXnUKSeL#V=Ea6g zc73W~XRvf2^E*#4sE0fvh!Yo$F*NaF{T{!hMVJqTFYI+k3VCZ!#zARStUb2y1yDyQ~dX@br^_q^2 z8(&OJ1Fdd?f{n}_Yhi>KF_CwV*F2i(@8}df;ve=4Cbw_0y+K>kF{Zo5)R?h!N_1qS z5TLgF$+9L)ZCkeeNDcaw@v5mRRI-X&O3yrmtt=7{iTo`gYi_Qf30oV-EZ-M1F*9T7 zhy#_)TZgK=xjbrrv?k@Oq%MAKtZ(c<=f(%}v6(u{xS~{-lKp}(q0J$AL7Ft1tlOw1 z%QwrB1=_U{s5@gxwi&B8Ry-jpcvs!)8j!_W1rhkjKUV=Hb~$gd4p3PD45~{{Z4LRo zD|`<6Dkl4YmxIso_}e!*nX9j47NF$~9{f)89Oq=aGr(uXH(QFzUe!DwoO3jKj|dDZ z4`{{|EUIzU--rlT$XzYFiqMPp{_Hcl`fo!L9Xm^Z+y|U0WzUjSk?u^(Ej4=Y-%~wE z_^N(WS4ULeKb*~Ykst|+`W5>(~a$~2&dDM#ECtg*d0If#mCL;YP6>c#w)6kp{ z&<1F8==Ncd$TOfDfG6>jIg_caU;j@e3q5-H-sVoo#cV;j14aPqh4ZNfTh7d4bv45K z9~I7V85r`^$Lx73^f!QOU2`-urcC8O_0bFum6 z6s3yx_V#*;a^wAnpEnBphA&!jtv%_!KduZ|xh|YDF5k23rdFSck1eafJTSmQ{%Fby z%rv`%z|94CE#deJWq*S@v3@G5@_n7U|AW|{XPthx5aoXmIse8YP*Cf*N(0>nQI@mc>0Rj@IegNsfk)i$4N`nko zhoL?|*j}-dS3S7V-;eCu^Eui0f}x!Qjx5hstMTJrgolXyulTNQES-|tRZ6%_8}O?^ zds9ROh0Dm40rMLy^)$cT#@BRUU1p&H%6w){5;@;KT0V7Henzqr*Yfq2k&*F5-7osa z`0Tzmoq#}zsqME<**pQ7`n&pj&QR9J>#v_|2$hZ93-jsEA!V6roK>V%kSVdNW<~@9 z8}>|`mTAgudR}Y}`?-+{FA>LCt+YvICfe6YvN|Et5@0O07|*1ED?2_Jh+9*uR|=>A z?)-wfw!Gj-yF(Dun^}3xW@f~`u^=ewe8$K^C!o%mk>Mueq^5;^K$H{&*rWi3dAX0Mtk!pj19q-7fgMBO^N*oSDem zIoQ*W;G3lbpC3#7ioPd5o>(l2$$QQv(HwjxV{PMkjJ;r8jr!Od{H6T9PvAGmVe+x~ z*JU#s9!g8C*t}Gi*xwtlCN=2$ij=KV2>+~z9^>Bo(`T2D9n&}G0cdf1)6wW*JHki^ zN2Ac61`+1(y2Huv z8v6gqF@TC*7dCcsU2nW>$*KEP;eWu5|Jj&KqL;aN#x!8G!GqZDL?u3}=zY8-LJBUHkO7zs)zJswxTr^#=~>|^Om+CysKXml{QM2?1Uo)@amV)o+rMS9Gwy1pUq5+^kb6F z99`Mc#g+iThrFfmw?hY2{9Q)!U(vkt9YrpL?m_@POb&uI_50iZ{3vNj=uGcs{uitk zbB=8$_L^n25&9F7g@qZ}L%A0r${PZo&L6YTM;OZ^Eg5Zk3kzyS-u7PQDISFH&z>2A zvLdSj0P({F;QPr#IHyB9k*qm;HH#Z3K{o3wZv^3s1^#DJ;Vz3<9?Se0WAgdmEs5OA zS90`)&Y`TPrly8aFfgPaVfQ~3=pOqoH1qH(_Hv7U+j||t*9WxT_wGIg)qe!b?{`l9 zYcs?|Vq_$fO6HoSqf9-qo0`m~HvP8=n1HEnuFP8AW1x@ifDu0M;`cOJh3>GLC3~gN z$Oy3#{>)biLjH9`Ani6lJoIEHnl>Ec>DlaZQ+ryIw09nNq(L_!qrsJA{MGtV>~Ks; zV>3284ki+vEtW;ngjc<%S*2y5y;77J>uCT#W9sk-P}j5+&>4PnK~JfH=P;ad>RSKB z7)&;H-{g4--?fIc+!JAaf5JrqZTd$bM%{CBO8O#B24_<3O;9P0U91*g=}7EGHU5dEP$x zrn+i?)Pwm}m})Lg)^Ac84s1(+ z+_?^zkTD>tclqm{^X&rd;EjbDAHqt&&`jrg)L6*3=F?{Q+8)?=w~~DF6SzgLPjdYk zCtnRYh5P80l$YpmO9@EDoWxep%LnFB&9g_f{g2krBcFI3eRlRR^<7Cdf7_YJ$dHA| zrh%an!K9g_84afRWLRNBB&_KUni5UfoN6o3Ur@G**H1s(f2j1~dmY|8Qw6AdSTFUg zPd^{hf!`dz-bOP@Zn$Gxky5%hCaS!WqzWEptUEgWpZO>Fr_k{70jDQ80sH!O3eB!y zhU>3I4|$r)m(noxmU4jRa1$6e^dLk{Zl5dd2q?(XT5%qDn*;uUqzFmU%q?uJ!&P{h zSIRi_@^28vU~K$}XYOE5wEkM69%{RuczELmxUy2MdwC&!@Sx4VfAO3i(tqt^CEyGG zemQd!WWQ5$cWxt{RE0b!u*MyUHx@RNiH9rM=401e@YyH&9fiNee{nm$Tp&p@lWDH$ zb<9LvkhcHYc18Gwfc&ZS(gk)?Ip z?CWiNwwPZTz+mJ{elcSnsdfQO$+%GJ6!@D`on-X?B+oQB53{6lwK&H+S0|vK`YYOd z_2LP|@2G(woqHDEM>xM3FBM$~E{zHD^6!owX+nE{F?8Z0u%g)M($BjqO(xnQ)8AHz z4B-Ox)A+&Dvjpuy`32yGh^=fC6!Ky&+96=r3xg4Hc)h#7M_d|E&R#g^Ocu6DIcFES z5PE{%-Q9h-vQ}x>Sk7!}ex7?PG+wWHKul~HFnL=gm9QP9{KOxKo|;8b@l^MHSjs*1 z01Ayxai@}=6>~}w%g)YyvY{>i=^_)FwP>MDYaj&WxE;Zjr|fVn>Z@2lb1A$4jx<1G zfIHYecgT45n-67h@@A09nUBA?0myL4%WSi5jG@&=2(U9VIceiD@XbeEI!60E(W!-FSC>x{PTK<9fKOyimz`d4Nn{CkVix$*v_V; zr5y{qoI=&5WviS&7W1-T^LeD5z2j4k$Ze;M9wiB?YIScq%~G3BnO5^1svdxB_iFoF zj@>IJXYx1y(Z{Yf>yVps+@9DT&B4)wUr;ONWPr)#KU(o{bK+1Bo`LylHG4uVsU#2! zpB9&MncgHM2x&xMyi(KpZA|jT3t2#~IH~!!U^n^|X%k77$NQCLs<{E2S=!ljF4Xvrf80uI|I#h}*!ha% zhtgjzIBI-y>1mr3MsiwEXJ!R!5LPOgW!<6X= z^Wd2%h-X&d;MyAO14yyUX#5+@pCDe05<`Vs@fjdqXRP)OVJcoPV3s(^R|mHieB9h* zn!TN!RxKvy0qe>$Vh;|RIWRew>+-q#GH0TVZrDL+_y8t4&cd3tC_C|=)> zY%i>j*J8yw>8xcpTU`HfPw-|=7zJ!k$6sO)nhBjY>ej~Eo-><@FaM9Rs|<^3`}$V| zQCbmD+CY$2x&%Z)B~?<=OG|e%jtU48D%~X@UD6HGB?5!AbPU}&^X?(M2Kax!+~-_8 zbIv|%uf2Nx7UIQ#osZVb-=!*4=_;!{fkCl>WPJ+(7U%AP>;DF=AHYh3^_w0V&F$%oic)w4)eE`T^7ovDajNF4EP5? zAeJ1=!4c$3Tg+aTt`&Ga;Sr8lqbg%fqlV=}#0``;idvm7@l$FSSREYkDzs%2JuEoV zbj0*EgI-rzIu5o&q$up@=qJ2z7VZdI?>uKEiA)K)g~rP~|+QTK+Ap z0Dej}du@knZ*!Mdf;D_jV*0!?8Q2b&ucKoK0_7GQH&4 zCgf+d7N=nSsur?ldF?gFTzh{%Vd+X|QPn(hE!_nGkk4piM4Cmx1)UJd7!{eg789?! z`b~|#eOj!{(5u)nElg%07l4tfb?06Ig$(Tl@R7x`%~T`~a;!`6Z8F)cR?@sM|9a_3 zx5^3Rm_c-UI!m4(_sjXaq0Bg78Xguv?Z#1UD_o!HC&``vMlp<&w|G|fi__GY(NIsS z^6fB}ZwGO|u(OJm?Zr*?b7i-J?%lMtXEod&9mtp(F%SU|Oy{-_1IR7v&<-lcjk4YJ znKe}KSYTBNibg>Z9y?cEAyA{5wssaT(|F(kmaHC7k@njAhViox$Y3gLKRAs^l z#O88AQ-YfhIW`%y3dwjY-||WGf)gH6bGvKnl>@s~C$6IA6{6jn2RoIDV`Y`|RLaT8 zvh=cwb9Xgph*NJt&#V#oES#M=h|yMjzc#Sc+yr=k3chu%8RylLWqzB83Fcd^QZ@DnxYO6= z8P{GsV23u)vD&CQeN>hwlj7&xiM_H}g};jHk}D?d690zmCek}-a6D|^dE-W+FGa4) zMM8SBuN)&eyU?cMW%Y_Dn^>+V#CoRO7em&wv1HKXB>w3(2HByRI4zUKq(p}jl2~i{ z`7ie*r`P3FU*v{t)r}OG^GFR-f1wgsXz!Crl(pLU9-gQBLf^taJ>s+7&H4OY)u+`< zaRu?=MQg6y<<_4@+#Z(C$#CWvc4YSL!GGUY?yc`TH>~l+1|)?G%n|p)&aSMfJ32Va zZ8-ZeS$~R+%VFZ*oq1EhZ1x52Hn$Te@m4==sneG`bUJ;vVf$u9YIm+1r8B?fIu2&H z4t%6l-TOrl$sK6;q8UIbcg4S35f+z~SL4K~lEGXIz_M!-3PzxAl=b0(yus5B1g+b# zm{fRRN^>34)AkzdN}%=zzOdRD100pBc-)Bo!7YuA&bpNNYdi)YhjU+x%3{nPF;M)xN?j4-A9u2eBiK@hMc*+ z5*BhJ{bv6+(v%Lvi7$lul#LNC67fC5&fQC2tKU=_q?hq3H)-iMbJcO&pv8SD}={B@8WhaS>qmn1=PESFp)r{fI@qjLbWrcS&#^a!z;H< z%9uVR)jG#kf6Ribz|1 z&O=r`#NrYI-_je536`-OP{2dn=kvnb&IO)UK%aj8cBX~1CV5i`)ker*fTbD*hFhBV z3&Cu$u$eW*#zjN3&l9qPFKsJzBOGpE#Zw^Om4uqutDX~Oewu1YmTZ@b6S-rNe$|!=0N~`B=`IpfXc4qaq?5_DW!ZFEZhf=u-(!K; z8i8{~#OW{cW*g}18)2u`_GVQ}{*bUd~nk&RV>KqD5RW?8D0Tr?XW!nK=0RI%_ zXK8SvDmI&KT6>O+c(P)nKMV0?%fbsgSI3KO97*#>xcDX-gFGWy;O8wX1FS87hARdR ze*GK>UjbN}5>Zl8O~YA-`1L}ww5jjWLo=DcFNp4QXrU5RQ_7UHHmQ!3^um2QRJmC& z{nE{Qv}IYk#d^oqb0_C>_X8hFgJvrGM-yi9Ffd zpF=CtT?fU&52mUG^@LJO>tA|%Nu+XYJSXMCrYxt>Jp;~P^oN&LzcFmI@M>&smRziT zt;stlY^Jg5id`G1Z~OUQxdX-|Ja<8w6}~1tkDFj)DXph+5RHa*UjaV#S|0<$VJ7b2 z{(VL)PdR^&i$+&#R`*n>%xT#b^fhuAGkLxod?J@y_EX$ zxTDxhZ{ht*K~CO)o1@`uG3UrnJJN08Y&sdnvl^Y2Yt1jByEW=VrEnf|t;Xt#S52AQ z@S^gQ2H$ra#Qm8DCdI`~@1zncLDyn-^^4Z2e9X;@>Lc~ysc{-l5AvQuO6WBp7O6*t z(~DpPGtQWTau~m#+K=n9NleeY` z<=LduGccgU+E+ZjhzYGB@y?mpxS=RkyFgB;1mSl!BW_=U2J*QIQU1iCq1 zNMOAiejOZo;l7`27IMM*T#>Svxz6TBh_NM&^J+G-kA&0NQ8K8g^j4jL%)A1KFGI|>IuZhXvhpxdcn`@^z z_+%vAHR=o<7Ill`4XF*(EA6+o?6H{>6NA6O{WWI`($l|0$+GHJTwcP;x!bMmE8*~F zjrAQ*%@fGLqidF1RhZ6?z92cejF{*qNaD9BKU-K?TaU$imWfzd>1GX6@|v?d&dxbl zocLBCl^!9hpx_2@$#3;Jw!?1=^O@Nky*Y0 zmY~(X!D|frT%b)7^vyRQOWNe1Z{jy!i{K+VH018om##+}Sx<}SDbWf?+k!Dc~^lZ?x~H4X#4d<8=7I~=RT82&W|ba+jndQpamjo?L(7Z2XlN=viK zaJJBeLe68>=!`dEe@XRlR%G@kMOx$yuaaQ;kGBd8EC;G}pS7Or357ei#bb+vz02u% zXec`eES=u|cHhv_QiKEn92G0EaoSbBRL|aPXKKp2Gg%>4X~9xVK365D1VxiurAf_f3;UG^*s!$$w5R@u`}-~`9@#nXDcmG0I=KU!K$mO`Y)CL z-WiV~M{`z8Aomy{i(RO%*+^?Ux}yLff83*%3&e7IsznVKBCnhN{e+z+;7&&uJ*DdH z$jsh`OM-ckxb+6ic60Yltw;JSQv*=6kd(}kwP8f{MXfgzjCkY}2HxO*Myk6eT9t|; zGHY1o)MswMR;y>t$|nT|5yok}EqYsPV0@W^6z}zjY+K3uU87w@>#gf`Yw)Ryc(2pa zldT$bsRiVikMg$tP8t& zCOcvq3-i08AH8E;D74Sg*=QIl5o;5mOz$9jDBVMc<@$o54lXI$79z~-xvfFW${1u1 z`Par;UiTVrc?I;kTj7y82V`F*ll`r%M z#?bTf6e1u?=@cF8(Sn7snVT3sc-(wX-1@GIW0AcKLeH_(Ss!Bg)bx%s^7Dx|^l!FF z<_oy%vdZc{>gAuU9!A{nmT$Lnpi^so4gS?*n=!=?8UreT9b=#$!|kwo{S!& zl5&xxQYM_1mq5(>yZasn4&qgSE8<5O`k%o0MU^Ez0P_MSFTVN{J-{>qP=^RJ%NvJ0 zd@{h`8J4;YV}xeP;?|by$>nF~(3HTRy0{k?#Re^AKLAw(Den3pRQb+g#bx-P?~kOa zwr`EX{{hHw1gf$tN;NUDJS7gLoADl7N&eK z;=|PYMNQ1@Fx8uaRa#YjJa+|d)1`&hZ7ga4^#!2Dse>$HM)F94iPO2%s`_G_a%?PX zI!AbG*bs^*tLN1ajjDsM2S9LevCyVu1eW1EJDthg{E1{-7_G|taC)xV1nxmb=#-;^}Yhd2&=|vW;F)9WHMr;M+0`{$()Z}xJ62`Lh<3QV) zWK`AtztLSWxcsBh_y>EHrx3qBNt!KYVWCOa@+sITv~B`G9%C0qHigDUU5frLtb$N8o;Nk+NI#PkQ+`sVE4c2z5SKm>enyK44}^&YZ71*L&ld^e zK$d+XvE6*GBJb)wdSiOy$TTAT9h4|HTaobrr4CC(8Ir%nYNBqV7Zw&(P0@Y4;qbX4 zbc{QIk(oRJlz*Aw_x3yq;$hR5!6&CM(0N$>iW z0LUhvrl^+j*1edF7ouiSkz+(#13_Xj%N}TFQ6fdN?ct1J?`oQD%h{ zYS{U7eI<7eCiTLrHdYRPH*)$M)B?Di`P*#Q323Le%`Lev)_$dVBVr{dXPnyYyR%wr zwS7(FlkOm5hV>GqefQU|6;}$p+wQCTb%H5>oA%et?&UfQpV!)ku9xZdb?anzRYz_* z1bH}QJR9@anTokeD~8<=OC)83i>X(;OG-v63QzOeKM_H%${Y9sJVAGy6E&93E?Y&4 zFSYeP_-&9ibd<|2-sxgH*U--qWv(gU%M#sHFz(5?)iwLs{qYC@$1ZS?o-5=IodPpEoBfxiIZb-rs^fq-$+)Bo zaZ!zap~HOXWPkNgoY8=7h3Tbv!!Zx-=;)`$Y4ICsqof{Eg%0+W(N$F^!cj3-i)r{9 zdoRYdGSbh5Z1WLZ9GrLNH4x_KM|q09-zfy7fv{YLb#ns&1T(7^U`1soXgU+?t}zi2 z2#gEy9ypjI=Rmm?n=>a04SNR zd;sHXGO$Jz;nBS^S7~^pQ5NKcCbwGODvxOon$(V&~wVZ+kv{b?Pt?AmKyTykFPXvJ>0?-I+D3i>jy6vVH>VXq2%hMLj}Tqj3k zVuaew=k1jI?puM(N}nHqEi#{+nr-$M59eQPvcQleF=Do?ewMw#Z3VN93uQDH-o@U#}GrbYcWfgZs;enxQk9dP6TH{`EF2 z*jP?t$U85)y8w;!oaSZazCh``;@-#6*(o>Fz!W&%aG&-hm^l#`OAP~4^~OR*qQ=yb z-zhHP!NQwB%Kv*vs=|538_s!_DX{=Cjelw5ae?uaB|W3kBi@hKy7s|^fyu8A?JUb; z|JV)5DIyAGF=S;N}C?@)Gu&mjJ+p;^UtHI}<|R8IW7ero~3w!FLen zczLDWwgEzwiX1t?W=&TaCm@X7+Gb#-`ffiG7iI2MmED29L#}M5-w75LM=mJs3=E>8 z$y~DhV1j_X&E&i^-jmna?`|!Qn!;c%*b_b%KiDmk88>-RMS`dmSLs?I0j86V4@fm7 zU_g_)w>H@DW_D}x5NtR&IHARh@m;Rm%g+)ut1t|z3OSO^oB3irW09NvRO1cRwD@;~ z!^~LC*7E7GK6$;oFHT|h9j;Ki)(q+b!a)*2W+L7Bd*JPzG;JR5#8S*qF0~8HuWv5v z_;AZ?4An+75fciQt?y#q-l2T9md~?g>_hY+d=`FCG7(fn>FjgjU+iP2Kc?blwPlao6rI3!>N-FKR*2V@})5$MwFDeS9@XJR1 z2TixMZ0P1f2j>nlIfp{M=_-_nA)w{Mm?Ab*pM<-=lcIEC41#S_yq*hm&r;Kl8D1v< za&#?1?)E;k?r1OIP=z$hJq1`@1?FGt^GbJ1UX*wCJFK~Zq0%=%X3Z4#br^2!IJ-Io z(EM>JD0FXT&huG z!>np6!>M+zDF0D*S zQp5ZC&pZti%_j1SH*tqRy8OWfdyXn_`0Oq>h4axCj~Aw1iV(cTD{bQW5Eg>q3Rz%*I$4M7O9wP1d*WWAqC>Q1+biBmgVgjZnanUVU6YT@?a zJhoXb2nYkH(1k7a{7VNzK0t`VSBR4V9-W?ZwIvADFRm=+MC(*6Kw1g;k`KIsc-5$X z5%8VQzOAFCj}xa&CEYQpkBoSs{EH|6g)H;We&_bf1iJG(y{(Ry*txh| zQ3{BAgI(}BhuUY5gRkrz+=_qL)o!fx__X8MRg-gb)NTPkuh^g*9sXB>n?Cy7#GA=^a5Q>D3iLIREDrMt> z>sYUL&C(iC#T4g1ZOL3C)DIi7YArokAk<};0Fdj#3pBnPsv>m5C^%cxo;Z%SKJJrG zBFAhFz|!(cM`?M>`j{qi5QmDW8)~Abvfo%-XcvP7Y*MCrmWrFPE&^)^8(hQ>trvq# zFA>JvK`SG4ilh!bTeK|DY;vv*)GZ=gb1wC(+!(sq4!P2_UrpF@JL)o`kzDX9AF#;1 zb(d?f&CBQmk}Z#Ey3!`3YmsSpj$!-c%Ep2ilYT__?6w&wQaA#+lTC!WCC$Eg_Kxw5 z1Su=67h)DfAk`WCM9^|#_R{V%owngML%X>!o?x7)k`k8_3Cj~c5!Dk3-r0D7z&ZB>MI!P60{ZD_^L3^43EgR_$;`&hS1wJvu-y^ z*KzjhwYRq?Jb9OlY}^&U7%8Bsli%!31sSj$F}SUupkVu*)@LjU6VvUcaA|>ug7mOq z(!66Y(@>T;CkYpk-#QhytV~2hB`N_mtIS-)jpa&-HCVhZXfYVe4WEI&bYp{ftmlC| zj%n@eJXW%Lv#3g6+5wb>gby%X17B4@qfJK#l?iD**|Lq19f*u6LIaU!X;hK-?i@6C zw?Kua@o()6A{49#Hb~-IYBjqeoIurwdY#cMm@#w)*nV~XV2BME$w2NOb9~`~b;<%& zd;p^wcfWYh!{wYi!30Y_%d2+k2+HtCKB)UGX>t&+XIjTEn+#I5E4)q=&{s(P#hk7C_J~H=@RgPe_9M7mu6zRyz4_-PK_%5)B~f@4L$CF? zt4E}M={7~4YsUrQj71FbKt{YkQD~rZ zf#gx7V7QbUL!JuhvuDpL+hUmtgc^2knLXDJyUU4(qeeVy6H{A*-6hs-7U&jR`$SZ$ zx!F6%xH)N|ao%Aa)C-v*fv}^$@y6IX#oe)HeZwmqGW#)5VpZTZtvV2y-|WZ6hGZSY zuAPL-a0(a=j8D=}j=hB3fIpPaC=1VSf-1LDEAizq*{NftWJKw7CWFEi;_ zz+mR?l)4*z<(V?5`}vp_{B7_gA*Sp>2Sb>@5MsGBX(9q*N5|Ko9q5fPc+%~hJ-h6N zzrzm7a@@=t5#fV7{!J`dWAtHK=y(*Qo8p$6>{*vb4gP56(0Nh5%u@U)!&97jKv5>` zQ{@{LBbIMVn_I={8opQ)5*jhBghlvkuO^U@LxJbAvD4M%NSL=9v9s^)S33UZ2nu0f zql!(Ib*Gn0=OQ99jJDIh0E?E}sgQx4lT!;#l*_1?y5mkoFrmQk9im>F87xUj$>8#~ z9=Z&R7_s620I}fmm{!_#ptST$Utic68I>|)NjkA_w+q%)ip#iJZMRzlqT^O=xvi1} zuohCWm&cp2L2V+o7w*`7p-jPsoAwJ|-kILK>GAUQ$YNKTlAZjCtZI0q_Y&&(;G2o| z8MWEmK-{vjj=+0E%+#c%u9|CU_WGp$jWHxjoQinlY~@FzgO@8LvKPNsx*z+uwZ65 zs~IAE-jEa-+;mpAzIZ$-vC&UyPJm>d1MhtwIR;uSBz{_vLR-fR*#y%Bl}6PKMGp4nv9PYlGym>cKxyd^hLzRTu^tcUQ<#`l zK9r7E`PVBOhg$fYZ=BaPXPOi2AM^H}!qp0&+l=pRcl^~w)SyS-dD|}gk<|AeXgA&4 zByq6lNy;#<-b|g}!ae^T0MyELUYAd+s}r9B!-6#j^~F`%5)=4<<&UK!`S}1KApSmz zx(@dC7GbMaRzuK8e$vwIT`ITeNjHRnfkCC5GBd6LECma=+f|tXFjP4Mo$YEIE4S@6 zLLpR*;eUCHv?XkEo~HW|r#-{3Mm7H8y=usx2I)Scy!y*$2!c*QlR}bY{L`>R+61X5 zsLXf}gn2waM@|0=uOEI5C_*qu@*>t+}eaKKAwIy%Gazmh7bfkH8fUfeEq3T?v@T2yi2b*fwDjQl&3t(5>i zCG`4sWIZ+z3lh!*SANL;pDtJp(39Xvl|m1x4T?9Q5fyKx5_FnxG5!U^7nNL~CBV$a z!^Lq4*p|M^%-qtn)et#zhD@-O^TTv9;g7Pscu~|2BqzHeU!2u_>gA(eJexYSWst`I z_}<2#)k9Ypem@AVhC((R(^j_y?depCRy|&ncd3lYaRkM z9x^S8d=oNJg=xGvmU!NZ6&g0OT_Dwt?uqqRuDNzKW6nn5RrF-7l`_Qg=3=2YNnxe8 z^gMQ`U?wv<@=mbcK_*1PMd&Vi7Obnw#k-}Qrwrsloja&uhPD2Ufi8{A-RC19X3FtN zpe}@*2X+pYrWQPFE+yDVahg}2+Mr;9FTwWOpQ&z;Q|ZA-65mN$)bM-PdDz)}=)(N8 zC#blT`RNqoOSUF>iKxVisy^Nq&EH43OeM=;SIoY!~5?u>q?Ldb=}( zo;`l3aUV0toQ0-KpJ6skr;+T9ojfvw_|x2A!w9N&XC2@=hCdZ7-Je zT-sGtF`B(_tGZRONsqxm)sThCs?ujV$%w-4knAUk4<*F!GDAX5qklg2^qT?J^iiSU zrj*DY<*Y;e9uTSbIIX2xq@#T=&NZ)4b7Ib{F()^dXhk&_=*{DqU@uW&meJKpM*CM# zci^8fqQM1=%{w$;zJhQZKE%HerSY4m_bJ2rNQ1l3LaW#AFpR@H$4;dag ze;9+0wK*8__F|l^fjx2?=U}%(?8M zb&yWjt9QhDR6>h`8{XY0f-gwvyakjMs6wi0n3RSu=4Q;74}IzC3hXtsvXZlHEgB$x zWG9?aAbK1+5Hs}AMI}r?vPNmzze@iG+KeHjDuwku^oc3KW3=L~!oqgdmH>7>ptB-a#VybQQI}561u+ z5)yn&OO~1u=vba+!M}D?hV*9{Q{Hjflc=d@Ys5F$9V1)HtL))^@)n(*`7K9j4oEj% zO#yY(eIPdDN}|?7eCpIG(Y{4XUZS*a`@6Q=sU|oeKzya=upEux!c&h!EP&rw7<6Ag zTCInsoa}!|NOsTUuk=#wVXxeVwejzg9;V@@~c@J>>r}W&BsYB6J4yWG8(N5Qe6>Gapgl`_tV)lu&>@ z0<2L&ADgREz&bk2q_P@AJ#e*}u|ScjYa>eu-JkyLfp&a+RYvvl+Z+qsuJg@$2ff_C zOOnH#@RZi55`(=mf@E=p*%z`cqE)20Gwv_2JRki7IR%NY)NkZGI5MnC(CHqaK6Ys_ zD_ejI)jgzoOmE}C@1gh4818I@d~uW$slPD{&`uy08CpdLc)8R{)SIc%qovP&u{8f% zb-5C@Y*(Q3)&8}*MAJzFCS8GC>|<{IZ+)Q6mpaHQ%Oh>yFu8*9pp1h(EB&~hM^hM@ zg?y*WIDP`Nw;u<+NyIV(~ox{UJ3PzNyx z4LX|V60dRjI4E4GB6a^;saF7dF=X-(z3GtrzuAu#C==ayYa}On%AAtR(Mmsru?4K&UP2o?OQ;&Kc#Pfo&RwW6zkka z;|mw-NKMMhETUeVH~bG#|G%4lP!ZaF4aI2?k|y5Xx#U$;v37X}s(|!~dh8$FwfmHP zAyjzDUxwi=$Jx7oXJWPo)MQOPrfp&G;k=iFaz5C_aKztO1~9xo}F;psEjdm8J;mhZ#aI0U#V*`b~R}E_;?BxW^R$fbIne z0}fN6N04TZMi(AacPdhFI$Bxx4-n9#lqpXD`zk$pJK%Djay5}j;D}gVd8FX?u%8h$o&U5Y_mtS37&!sDH zocJMMC z-6Qj3`6FjRV?8B})82z00q^U4k}U4_F=h1nX93;8s(DzMmP3!h5=)d#Jpp|0St>w$B-SN#`7sbqJGS;wBGm9ryV_ZgLjwGF8h`vnNc%LZZF_<1 zNhQb}$)0+xy4R)vy+^$%d5QhU*Zq2-frataYwB|wrk0kZ^MHS>%q*Guo zQtek2e?&*|y=ma=yC19g`2tk4%h6wfPgPvZebjiQDEV=B`Dct8!X{E(ZK5V#Gvz%f zv+@~2fI_o97ff{Iq1Aw4`vHC*y!6%zHmE{LWB98+Hbh!){RrKKP40X%@oP$rsQ zjy(|!T^0uP-!vteJ=&MgmWZ-9mK*li=G*??);cGul$`|OBNuI~ZB#!-wX>9Ix}lu< z<6YfgDhueR7e45x|8*mtmouKH^M7Xr#w94(axopwHv8`ze%f(p8~Str%oP-0vONR* z#6k&(MjEq+Y!cLohNh&uI-6~K^g)n+Yxe_Gq zdF^=>1u+twNfXN@0c;UXQL4OV20r-zgthz?3-!>LGVgUrm z_QaK;Z#F&gnmh+3Qjbwv-{3BJf6 zQ>NVJCUCoK-Myi0&;}wPy9?F_4vLk3As6TE2+)8{lvdCts6S!3@J#y|S; ze*PHfP3)$Us9T0Rbv`umjq{j~-WNUo^j|UxZW$sz7s_N)tG?G56r?&+TiPXm7)J_9M4im(@J!|-PiEKz zTB7TySjWromlK;UMt>QXf)WH*Fvg8v0~D&o5Si=@OMoQVN>1Fw`(A1<$!0a^eD837+`9RUV@64Z{AdjekL z)im*PIdXr#?+^n?mQ$9E+!@I4Kk}9zFVQ%Uaz<-muU3j>>(xNyZ9UKbFv6kp9)P+A zTXLlw;6wc;;)$-l4GG=xNxPXP|0S^eXp`lT74Dr6au9aNj+@VEdzya8E1=U6m>NbWOJPzOzLC)?0|)D$UW0 zI^;*8k!lnklmWnlx{i|Ntl7Y5AqeR<$mKnf*9f@r(UX#rD=XFYr25>2V0VcB?jg!X z2{8xdL|b0?VqB3zw9%Y_!V76aog9o%0f&r(!BRi=^xz^W4h`Ex0#8ezB$dx{b~-q& zDC=PE!0&e+ZmgdIs3Py*c-I-7l$$hK4s^q!o74|>Xz>T^U_oP|cJw!$mJH;4wZ&oO zelAszCs^^XBLe`h`_mC1x2NNMQqNqHvtKVlObh8bfU!^dmzYo#G-U?%WCXUA#W5Qs zr7;A5pPjXqyEeAP{rAoJ2TBl<%l@hvQjFq+aL7fB>|2+EI@kA@TE`p3;ub6^M+-}$ z>#XaQTc4fqKjB}69hUy=vB;mNRfqxfubO$;=@ZM@@%<()wYVv6)uC5X<(?%S~vabQ;x zJ2;LJ_?`etD;ZDDO76-Q?kc_uwEg9k>BA`Dai&yxMtLzD6s`N*tw_0o?B?jwpHQbf zZR7(6s6zJ`2&#{gsi}R}jy3?GSD_*AmHrjS}z+)uUvhraW9THhsEWlVqnlz+@5ZWH~@Y$ulK(Qty8V&<%7D!owCUeAdwd1c7^Puh`-C_&o-GBoBwz zG07}l8X0=vY&$Ka)L8sJHT_p6K&O|}K)~WNhOQ)w^&>#a5}!75`a+J+0i+r;aBjxl zv!ie@j`$Ri6_B*xUF<1+yn7_1!Y@CG`m~3r<^o~2*MZGO_K@^I29)dx_GFDvlmxI0 z8qxJH9QhssZU?CBWBSEm#2WN}hVrOc3(H@0vHMUqk?vZ|FhB|wSmBeNPf%wk3J3L; z=X$4)ehWyoNLm10;CnMmN(%OOmq0eFJpZLqWMJ?Q8Iy<_de&K4bEIeg5A+7cSIMc* zw?I<#2R`{PDQPtCgbgb zZ&)K&D?VqQlE!zP1=(oBjmhaRmiI*t8(Y5{qSMpA<-q&bX=!SrrRF+79C|>{7<{aPQqF8(*-pU#J8IbW3_mIR2_w1xWsq$V2q>6}JLh z3heh2Wcz!tYN;j)D#yoNkXyZF_MAFtM{9oQbr6e7AUvsr*%J#9P-Vv;Bt1G-|6@0Z z7&As;7{UF4I6AYGSeT`bj7_S1wCrm92`}Od_u#m%F~|sXpU`PvRqgLD7~|yR9BnzS zbD@CbqdThn0FdE*Qs>{z@8{)LF@Tu0Dl3Z_rF)rxS(Ph2hfBd?5&H*dhwDBn_zgpcQhjUf6GG|N+@td_*1q^#%M4b zPi0#h`+_LL`@KU>cO8tWPDxnK%B&%k#I+EQIqX&ZvC{nuQ2MJ^`stF{*R?(85yF2`_9p`?tdDQpk53G3*uG)hZAyrS^QVw-;$%MgnlqCxS&ffSFA{Js@+EQ zh!8eCQL~N(Jomkr_+y_COo1&cVs0918;=}o4<@l;Z7;RS%Fb@7HfR}oq^(mfL8MoF zm50G#8#|oMsvdTXnmZig;9{@maRcjPQ zVkwtF4NoUSAgJRhxSpb3WPhjVn6joIQfRz5X=U4Xk0{On;_K8|EBZf3{WAf8#WRhv zRiaCyGv+PC&e6KF0T9@Z(u_qygQ~DjD|e^;2`W&7pX+xed?|pf1zGva-JJ z1`sXXT6dNMtB3;C$p@_A%}+rt*g5c`*+6eES1O!t-4Wo$v==)29(9sgmoZ3LBmp*& zMHJUc?utuKbp%7pPLbZ>wCiI`7Zx<}!Aw>hJRFz3cJ8==07QlWfFdEs$c8bcOU@h& zJ4YQz1%>-g__!6+yxRG&T1k}6;{q_ zg7^3Y1F{*CmD}T9mu_6WYI07b5_=Q_hxx*~t3$f(w&CDr(wwYxNqP_Q{rR2sZw2A~ z!T$0+sMUZh}KaWLspo8s+b7~Gg%9nPTb zBKhFzTA6+B0x{Ld?UV_I>(@ONu`c4BKqnCT=fD3zM+v{*m8!RqRy>Q=NEBpNr0wW< z`zBlX`@qmEoS{YqBIZtSg+!fQwYanzWJ4FnUL`etj?lk;i?1M}{S;I9<6ocl-*bOl zX~lr#|N_mu!6mtF!W(w8* z_kn>|CLkN+Jo@m=_W=Pq)Z{jjbuWGRnL*on2i#3ebS#rL!e3f#%r2E4C!SwK065gU7rdvlhP}-h69txMZ70uQdNtpp|+DofRmZXRa@9>fnwfRf4i>WN@tgoNeC?vN6!2em=oB zuhCWbuF|_a8nG=ZvNJ^u{S#-BW6=&D{WGil`ZwwuX!JpHC9;LvT{VM^m^$m2+0N-A zoQ~O|S#Y+KosO6H2t;jg?X6tDL1^=}P%cj9>0HRdif>Lz3g)Pcf!hEwm13Soj@lsQ zq_6;GdZ|LTuVSmhqpV`a3pxfLerHtF+cjoSdlp?AtqxQV6fvTBT9KUj$QBXTHR_(A zuGnD;5O)IKwcXepfv5X>dUC8zT0OkgW@#=TjCs<*^|D}dAp^HngnpzNM2p0 zGUxqv*HahCNdro@l}A@+`R-`HlB*q>yH$W3svYxaWPY;sS>^iLxn;4gh!eAU(d7S= z9sfaNpqH_;OKfUw5pV5(84q8;+4?YsN%7V3aX2@_86A$hG3Nk<6*bZH4Gha38s)_l zOU$grss}w%mk13H_j-oId*w)~2~D4AYHqH7>2vGSGFT3xqFp*5nYru^wpKr#RDwuS zGOXTiB7&w%by@$uv~(BWob9@kK78_PW7r**x9@zJ^?8PT(kGmjhbb-vao9iHr|$7w z2e|)V1{o{uU|F?9ZQUI1z7jQ_C~Y09bz4Sz886h4%m_gD7S&ISQ)Mwv=S)@Czn{6X zLRl}%%f-cV@(C-Pm#N(U{nQhXtye*n{QcQQ=>LwX$RL>V>u{c-qJoYm)o9|o(w!llZ zOp--CJ{(u9-{sOT0vNRx>I!g{I(72yug{*0XDccaKxn|m$)43WJ)^%vgZd_G^h$O8 zGV8U{kKsDHx)--3B_u93od4rFdtaiVjZThNgp9M8uC}NUq*tCc_vsq2Ri>(pFIk*> zk1a^gY4xv}As^yGVtOh}89&GB0rLA>wehR=C&b+CGLo>g#{NqF{g>XugEllih}$YO zZ>&atp#)ZX|ISI&O0S`VI?(T&>+ATW_M_ukWXG=3>I^u_9}@D^2W3aL?61=^erc4a zO$aOQ{PM{JDJ&9v+Y+ti=JIo``3#mq=;Yqk9?5_|#4v#Z)G9FGUZ-vYY^gZAv~(Leo}v*abBmX;nR7H4&sBZ8UpKJ7$3=D&iT{p4~R z>O)fgg>sy&n_k}3az4YgSfc8Ht01S4>N}IdC_;g8G{-2aqV)@uxZ-kjbdfg5yA=14 zUuS0S2rY8hTCtFdAu;vqkz`rIdH=5Syz7IbpYYS7HC!-~(iq=dXyWZp!^j@^dVS!m z^L$Zdp2rmUH0Uer6cFfepmLNjF50!ZDZrMk_q{FC44_?3;OO^) zk>%gNXFPp8rcL8&{^7RWR9h*fk3x+q;k&|iV&QUq@XZv%Bq3+N>7i8rf4@8GN?ul@ zLIWnB2dsvd8Ze>~aS3lgxvOd#otRmr1)2%PMHW6NwQjsy&oXM7cO&%0#eyyKZq;>u zy6HDM(AdJH+b!I?qeAbbkA9J29NI|{v6D*KSvRYvOd0JH?z&*OmY%#MoarixbrxfW z=3@NpoUo6+fdBtPz}8=idZE?gT-xaW}Ed8q$s8R z%=M|M*)I4=S~lZ6L&&zkrl8U3rUZY%~I{lP>a)kIG8|{&AO}PuTGioO! z^xcACFD5TqXk@Pt# z5m-7#o5BfAQcX&0(R;hAv3feF?5p$YjC5Mr+_TNq@bw1<4hay2?Ck8N9x{_{h`D(d zjUbCc$Hc=xgz?=QD=(jR?7z$0x-)T(sId&K8rc*rkxD+^GL*tU5R z!AsrM#2LpQD*+W+%8IofrEoKuIPyA|;Xm|0FZhsW0%)St*f@E3uKPZ2e;?|5g{NDo zk(7m)q!lQme1qHkJ#3|wJC0`DF%G$TUr5ho>iT9^s7jT%+_zJl&kjlNqOHD*@t)Z2^CkCu%y(<|$bOi6LMv#UU4HdDeZ?yRtXlSq z9TgXO7GC1%mnKSm&AxX_*JEF1KR}mvqE5DtT8?W~XG%93yd*8K9h2WrxeJq&lG4X- zk6$sLysxC6sbb$&t0f|$sz<1O441l~MOOq~T$`kf*3r7K?RKgr?P~XzFZXUI_*6UQ zIvUkqD{Bp#2@d%5^_#OhOEkBH>kpZ@Co$hUV zwg(P5(9zz4*4bgs)9#OQ(k{zTtVadB+h4D>QxiPbDDds6g(@{#p!?4Ve|4rqMR_hg8%rm= zkB8Qc26A{H75DE zB3&1bt=9OTyFxrBS<(W8EQ9xWQ0(B^E%@Ch<$dq1=O{F6KO@78MiVG9^O*4}Rt z8Yl*YaDQ;h!9JJ~e@x+D8tdhju39*KHLe^OOV`s4P6X<>JzzDvs%V}Nh8)Ts7152n zwCIc*+)!%YC4m1x*&(p|y_R7UO>5~wlHi#dqCnO2R;+O**7B52JHh#Lc16VKTML@f z`ob02PaR|~>a0_6k0N>Ig1pQ@Ke-3i1@Cdad45Kl-8o?hv3xeX&|zz|LM-CT+D0xw zhSz>1|Hu%=lwvHe-21q-n3|2@l3*HbavIbd|5jSn&hkvycN&WT)cIRXJ@b^g< zl(M@Fd0)*mh+5rqO*9drKYvLeb|RS?_Win?Xa{Yi z-9r^ZUvqcj{0r%t|C&EQz$fxHU&@h$XxiJ08W)XF7K}GYH<;BLKKWQwraC*aRy4Pa z`HtDL;?erfj(B=Kp>$csmj%R{YSH>OTLfZpwd=ton?&*nS=XLO?*^m>YBD9 z2-2W{gmg%Ew@3>jNE{mJmJ(?W2#6rvjnW*t8-3KV(jJm3eC)Bj`!Ed(v;D&bxiWu>7-|i(Gn65auLNl)ytz-N8dJ85OUzz#&*6 zrNY=V><(7@z;3bN?UnNz@bi0|)3LMB4KI7Fe^hs842&Ysi_P#JxGfZJ9 zCixkY(mhF9u$gCBi4JdQ5tk8Qgb;|I)4;FqhbZHO;GSYV?w!u3*tcirE8*j2NT`W&|vBel1Eo*3|wZFe#wIL0_jKJd; zN@>nZHp=)czFA4Vw~G_UH!781Y?yByDE0tbKAtIT&-b{A)eWuc$sr9n0|_^ZOu>UuF}<7h7InxNC;c;iw~C5@eNVV#EpY-2k>G`hHvn4| z-2;Wz3D$kL9K+i;kKSP4!R-)%0}C8{pgj&+S~9j?Gi2a|Ex3tiKGqf+^iX`4{bu68;M2DG#A@Pl zsf3}zCNZ9yK4Nq6@lLx8UT{a&lZFE32+vU1vKAu~FIKUIyAHh(7kZ-o?0y7 zoZfTH}$kP;lwdUYMfe0`sJ_J+oc;{&lB|$_s~+D z+-!Y)=v*sgGh#@(E;s@5h6=ywr>)AaE z2feS4(iNb=s6e3OlZnsknl;rEp)MzC$RcspeM%yuklym-DRha7y41n#c{}*x(it;I z!PE{A?E~I*Pe02)wLv(+QKYqvT9wIYv%vDAVdmXuXG^)*{w1afG?A914}iKN(fGdN zf#=N99-65ip5Z9kDK2Kq%*-8#9bg>>H|W$5IAU%%q4Q`Z@$s?H1Pbn{C*8-#Gnp^f zs~?($b2v-@gT12{3q$MCxg+l76i`sBTgGmTG)i-vSm1t*2Yloak> zkxqfRZn1IZY7Mj|pOM0eGMLmS$tDZO*s&`L=Pabt9F|rT2wOhw*&5@m+_m0pn+g&d z?X7`5KoLM&Gb~x0*~?yujWEKi?5K($hocrI|OKMuc(A%df!_{6+ z-w?6h)T-1AaD#i}r8eca%*>{=?^c!Me3UDtDYi4J<3c*F8EsB$aw+IB;ad(w zPyv=ht&<0Z3zGd3;>VQ`DI5(gEjGJ`h)Yhg7KEj+vUoLjndZX1F zrIC!-UQj#HfphClXx6S@G0N#Ne=vqS%KDL64Lo3`4G;) zzO%D4AU+cO8Zfw}q=<%n-}J=c?Cflsxrw5l-VPI)OT4x(M#jlDurP8%MzVu`3@PoF zflMcg4mO|~#LY0k%HzAX>?q2oqH8)&y!QgqR1Nj=fM(xg?~cmb^gky3wVrH6&QNx@xYw%skn01Y)R?Qd$AK0$9VK!OFVROjt`^81T$Ay1+$D)Nt_9k>sBH2O~D%m`IMvfo(K zUqAHY>ed~u9e)KB5Ltr5IhIJ0$)@~5u_epoLGMNhkIpYC)PLv|pl z=}UY4ITrnnSHRbGSb%URzW*f+5j!p}Mswdook915*9T&{Y47G5`P2dNO9s31_7zJ>rxu->8XBFVN$ICXecj32JM0VuM!p!MSl7_(&?l(P(VD0s z%)F3F)hEKc6hZu=T3?3?J=Usw1adwyw>D{K>H<7lzdzctpHV7>sCO;+fs#R5oYOw; zc4I#yY}Q=afdHMdhHHVD$rSO@lDOtM2OA$L;E3XY+Gy^9n(`CM@c==f)KKUQUfqjk zu`{AHBUV&?IW}A8kHc~8ij3_Vj1|UnE`RVF!oWTI9_%V$_B(~#szjOelK6KOa@rPo zLmc`|3kmavb;}zq0@qi@&D)Y zQuH2|k7Ja+(^-E9GKJS;}&^drN z;xtzXUyR$kzGi-^am)u$#B~)b0#-ct`pRoH0Y0Lb=*v!F=c9Ds6ryx;dipMFb-Z^t zty4kN+qQ>xp*|6bZDh_*O|F{b2wH8lcGs+CM3j8>_Cvy+Z5SEP;kjG4X!i1$$;;Wt zghpq!y7me14kI!2lgyrrf!JM*w9UQ}nsI#6YE2)*=6Ntog!i3}FqBGt84Uh>5Zn_@t5MNs3P#k9$j0CbU%qGjTE0!vRJSXr z7XIf7fIuYU*gCq?aZzuVB64ZA+5S;Gn8;S*EHN~w>&|N+6;%nl#yAAJqwt6a!~E{; z<`QO!u0`6$Cl-<27wbh6`71TlqX+RE#)thTZIf~Oqk%WC@WFs(Xb8FJFJ(Q}BDJ47 zb0^>$Nn;#d+izX3f#0=u61<`CV=sUia!x9>+=^uuY6cqQr(3vW;5d|p9=vqbk444J zMpE9tf~sFuqv9>}vUfiCQAWpI7P{xGwgFzvsOy$^u*>YZ%;gskbhGz+&y#WWL>7@z z2sDfWfPcPZu6QmYYxF$$;Ls16i_wYyI6a!c@hi<7&2#&P7QuIsU9t~=&xDS$ zH`pbgdGdRl)SQmR_I(=7QP#-F$p+4`XS4}<308q`rgVkxeEmO{v4z=M+eczirjS-v zC6j*FO##W^Y!Q{ZR-mUi9-7x{lYraXx4kNR34x;kK)SFMO5W5L4-2@4xl`7}Cui|v z*(Xb@rfE;kfn$P?OK7OMBYk+&B)+jYe>$g>S3+t1)LP~HQVv_;oO4)f9WuGTZTWWD zjN13A^xvxk^=5KLPAW+Jn|Y)|*1|N?FR%DQp7z|ItRcK$?U?WRfbi!pLcy&7Gb)~4 zptVgx8O=VzKE+4R0VQ5H`IF~Kdb*FBU0;HZy6*D}uV$sS>OZjf)4vKBa<(t+!b?Ua;8=;ikuv>r*r(Pa@9)F&q16eQ!SBQ76)IHI4LY zCYAYMM;yfe8g7tMb3EvKs3U;|e#3~>rhOdssG7sPaxU7QkL9@I7WIWxurcpRzR(`@KZ)XR>-0)7lD*sM-SHCFe zB>Ex3yl8YTj`eY5Q8$;ODr!kb(!OZ=mQ|#l0Z(dDl28o(3I(UnwV_`8*{zo%*%6A} zs(Cf~WfZ6MRk=45p3y2D*>L}9>aXjrfjPD~4ZH_QX7+H4VtEJx^esDaS-ctb0=y^@ zG(-D!&{ha5o9F%WZlc{|+}#K?YC&5g;&J^NXJfO&!YRmTJhs!-M9?6RtV_ReJR@Ma zK0>{*KJi|8enGE8T);L>vy|u7q{7#PoLr;<#$q3bt}jz3L|BO%5=nMS07l_&AbGK>AR#a1RZ{5u6l<>dx6DMl1S#xPKUk1tymv=8O)cJWJO|j=L+Xpx zP}Kh*Hh&w4mvT?}CEBnuAP}I^Omr{eGfp%8L9b0u5_aeEU}2Hc8+u(Uv_?Ak_g5DA zz*u+WqLn#0atXQMrRknJB}U0&|<|PqsK5KHz8kY1PE2FyZ0zR(Hcm zFjQP%G%byYO<&F!6<3vxztyrw2J~^Odg?u+W>G9`;n-R;yO^Ns96xUa|M{oA?3x%j zOz)o0;l0+y2kPLubWy-x zqyn8Ea$N0zDG7xj+)Z31Z+oXA_6EVoDs$zcSe1n~uFvJbeg2(~7G7vHVHxB?EL=~C zoBrfmVfR-s@$)Jh*xE_Uwg|J;$BR2NqN8P_qq(>JX~(uS zNYav56ir)EMjuppg+E8`e~mOwe5O5rVVsqtVI*xIHi zdwyo3IwMK3$WvR16$?7D;5l?46XsqQk;;3w-7aqhDUu$~8H;US=qh8edoH%`lCCj; z&x9!3&c5-G{tEsQz+DI)Sfwsl(8;tEc@q2FU^1E`u+cGB(dPCD!rx!We*i<1#Qc12 zanS_q)@+szE;SO6mVSdW+2XOl#1?f&thpJvBw07Jx!3SfgAXwir(L>8V+eQ~jm<-r9Mi3VQ*$@^T3J2GAXn4$#iq7Zz30 zlvL@v`~9+`lBoLnw0yGV>LyfjUzvz_8(V5vv6sK26|FR zCddRdFIln0L77%@$)7)oqVck^J^N}}O6er#edo^H(wX=ByKV(e3@BqCt0UN+={4Tf z&VFoR&2qQcXLpy4nvM=(MG-FqmyPGW&c~Mq4Yj~9!Iia;)gA6xQoMXigR}zBR73-& zpZgz9czP)dqx}pYz)#VE<_s?)uH%h^UUqnvBJUxEZ#``)+}BwObK}mkV!7>gBh1b% zJc~`Dp_xnZgw`efq1rhM!tqlKo zfZjxs?fv91(rc(8T9I5m;1M&vWczx2Z42$wC;d38zVc+#tbCRJ%54^0e48U5Yhzpm z-I_&2^V1oTH4-*Iljlw|>cF9AZp3n7+lBIyMWegcGY@kF55EpHYnh1+>~}Mq$+O#v zI%H}F*v{@9%?;a6kWKQtIF^ypx9$T#rDj_SI(x82u6d9I6hymuz6{J{V%yc)yZMm_ zDeFhu!Rp^-P-%3#?HBV}_Q%X!gL#5PN6EU9e8EMBU;E_V-iZ3Q$wha9MjV!vY2ux& zJwXeW!;&pMPWm%eU({ATbbyY&;AqXjeKZd|$sUxRcjyz5Ha}Kyx>J_J;-Kxg2Am;!^|+}N z3#}XMFus;qpU3_34rA#@#Q1C45a^!Ftli_qskWB&D|I$pJaH8C`N*(>Z?d;-{g{J; z6W9BtX^WG0At>4Hxdf^`Q5W`@oEELD&qqkMSs#!_1gdOQJm^SKRe8yly_P$9(zwPQ zqQmiWs(0t!#yZPW^D!e^1hy>gX&GVRT!6vAB+$eAigOID;InYrv&p~cZ*J~o(9xe> zg`6Jeo}|q8ej@%1fjp|vM0!RKMB^)UBX3?#dp6~Iy8eJUz=T{3NFN)y^4dT&3^c)M_$CzmEysb;RpVq%b^Vkd83Xeq$vFPetmTIT7$^lt zI{m80XeM{`$YUncyU150tU}0es^h#RNv#2|$!Y-vEmse9O#Bi!Nr$_%f)^4|F1FI?kiin7w%C@g!Vk<~}t;NG0Binf!PuA<@_63uhp+C9CZ2?rO`>|gPC{FI}Qb6Oy#q@=w3 zc>{45n%mdT84~g#Cjo%-5ruQP8DN;;)`Xscm!X_UlYda_OD%_!$Yd>OT5t=ZiOTQrxk*_R?lydYxoT1lB4r8B)0vff_S1~Oa)-VuJ-w8U zMWB2%prpKP*TD@$ZYtel<=8)wiP!-SI4-QAjXMHq*U;dMIT}O{{zl2n(g9>~92i%V zJiD1%t+q<@O?PKt>!+P@;l2{_JMVyg?;qjaALGF&kB?PB%VnLF%XWqd zRxxu6glE={hRT5xF4G_1u|(cP#Tv=-(6g-iVh^UdSZwJz2}%(~fKiN(;TVus(DkgV zI~FXMvQaWHP;uX|+;ABo>daz^zZsDH8KXi0=`R1v41PI^6`FRvahOqg#l^t*3lfq1 zm5#g`>K^0k+5N1n$12cZq;rR{L!qYzGMssBJn$9l_VU{?UqqJLoppG)gHye8*Fg^e)TSp^yE!%y zIK6)c9{HmA_uXdwwiv^c5v4(qxl)B&{8B7+Mge$*M*#(s%K_rVq!Vt5!7c5 zDQYN^;5eU>VMKCqvs2JSvFrkeDI(;sgcTLB%*W6QX6jqkD4=;vmN?aAB(~el%LgS= z!I#jbXyy$+k8huIoM4Vcwo~!~T9Z{=cSE(>N6I^!XHN(;qog{(X zdXHJFOYI3G9O%E8Zdx_z2pK>BT#Jq6WjAa z*%ESwdr*`e5cv4eQ~aZ+%5a-q;!&|)O1ikZ+zX|M4DIIL_xViL*vf-ia3-2(OVS^5 zH9|L}&W z;*8m5aBL`jV%0*w&dROz+m5c?(dixBh6x5439+#U@JFS_!o>~35SrL!3LMv`0h5KF zdIgJn?YOL>vu9n<9vq0_{xoW!n{cGh?41XQOyA(%OXr6?u2WTx99^=|6yy%N#KqWf z-IhJrE+R4sMB!&_)(i6gSh&}+>rh#{r`1s{Jh;aT-MgsZgrPaQZ>7uSMou-H8S?)RN2ruEcuH%U&jQa(rjOT zh1qphg8{j-LVOmAbl*i^1A~D-3usdNwF^mN11(FcnQAL)G2Rii$(Z@1vcdiG+}|I06!2ls%EnXc+~oKYHWxK$>YwTd*Q~Qf zN_&+~8wIAJps2o>upb%agJv`mIf1P^J8h^zvdeXOkx=rjvygZE~8XUi?mK z$Z87?iD1tM{&1LQe9SM8`0MV4txu!CDjxx7*3$=h=P7y(DuDlTu$d(FI|5%^q=pt< zfKFWkBja#jbxCwgOvJSdLfYQ5VGq{#{6MghI(kWo=O4$GyHHtnQ@g~1U8C7RYY z&qCebWmhQ!&3bx|Smm{rmA?!RD$!F5aoC>u-fuY}B!7mM$<`tza3MoMHT33xBDXOz zZNx_2xacLx)flp(r6B5a?K!n`?TI_J55*#N7otIg72NPU&wN6&9>q?T>i&L&0l!Xx zaqXf^{AMDH8(D(t%V`mjF?^**&9v9?|-& z5N`RRTs|w$FV4!&_T<&-nQZK%vdTz)q8mS*%5RS%WP&uup&xCboz_7wFbpxq|4uyq z_L;wQ!Spe4S0>OVuR>d)R!Gm8=K&hO$Zjc&|Ayk=c_j4KN@o$V*Wmg}X=%NWT1tL0 zC-5GF%>ADu_@@EAQEthpl>625@|$-w6(W95ED|?}+8-}l991VeT+I4kB+Aw#=BJg3 zvB(Oq{sQp?!6)K|?HNYMd02D1TjKmq{3xZk=WhsYC}!b zvlX4C8aIi&FR;I3+mFC=D8XG(X(%%yy2Pg;;8*^1RZ#WqHB!VB!?d>My39y_58Dn` z&&L5$KpILvZv2hUSE&?g?qi&nf_=sqq_`0#!#NDFL6a;%b=KPBW7&aURS(&oo4oO; z&o7_&n^*uf5ZHo0+!l}WB6#s*dK4EE{Z<06v(lfXn4vfBACu2Ep8Or&NyuEM_uhMz z$R^!EAJ1~$ow7uDlaOd@c-X}u{s;3c4K#1q8C$4hiC^D~@N2Hi4bIg%ke>D&idKeC zB*e1M6UVF7qzI6Al|*k*`|!Og|EJPCq-SEYor`pOqE+6CR81k!pAbr~fVoj2LzMVer}|>}M4|A2*Szo?Uoph~kt|>2$B{o}`(MuBB~;Biz!kH` z62wf2iP^)$$;@aRY8>(;rA8|9KP(uy29~=XE3s3yJXbIH;`JM8X5Obv+4M3g`;II{ zF-LMffaN@dc5)Oy+kKoNG%D^#ZWAwpI_aq5OjeX;h^E58c;IiXu41RTbvN;$<7 zebl#|Iz2%rMExEh*1i;=MKkG}FQ!uva^k!O57R~;WbYgRtuk1b>4~cwoji6-UD-c> z?PBx~$p*=SKr2j>own)Y#R>oaG8;ff7?>rHai6WOdE?9QS|9I}70$*i*v>=f??=!8 zd(Fttmop^8bX)tF9Z&DcTU!V>I-IN%g@{}@sx0vDV4Qrqf>eM`W+r10=nLCq7{{3>ERp(9LC!=mQRfo;aI_6KW9&({&M7_b7x?970Q+mr5Z^Hk> zkU*l`Ywqg;OTH4EZly^D_ivLTXJon>NoxR0&GKOB(g|{QJ|M=YR1EmDh zZHH=n{vt;&ucvsKbkZK}(2b7wo6SP3fGHb3Byu0DbWw|u;W!O6$*g=wH;XNWo7R?%n=3-nY@p?QfB`L0B)m zlI;f~tC0-#|AX34lV~%x2QA>QkXJSEu+laF?Y*=hascw&fmUBy_wF!Xtoz*QLL6r< zo?OGD^i=yYol+fwFQhg+MIv0jk!@GF=Q-?FkmY&DgN-Q83Y_krwbZLk@Z0z09{`P2 z7KbY)R!ZD-!`U7nr}cZqX6ewHn2H~?Al+0|d-EW!id&jvdV=^!&4zUSd^7JP^{md3 zbU*p;H1FGwhA04`^G92q+AZ#Xw3z)V`)j3o&^j<0CS<@qH`Cn2HE6+R^r~vdWbXCL zSb#I!n^Jp$O`25OICRQsCx2Y~xa#dD^W&*R6p3^YO*a7f(iP3`~^uum!O-+7XB9!QhAkA#JB z@~B#n-pPWG{I*jX>HhnF4TO086Kzfne|!JPq392CS2c6~;X?EEpxum*TD*WY+c$xw z(Hfqd@$`OST2JZTPD6rs%~md8`OXAJFAnapY|bHE-Dyu81OtN$x6pIqtXq3Wgd811@6-~e4v)1d9OD_aTCRo=re4C?{M+IDZobT1KZwM#)-Y?=R@iIid? zO-)&;yu_xz+|4Kx2T?SU{M}vR0Ordj;GBF1q3Jq2Gb1S$4 zn^{u~=~n*L-7T*Q+$gNpg^uvI(E~KC{FdGzSL%|ts&PfCLiNXm21tC}*V~^+E7}Fc zm`o=uU~iDQ3-uuaCp8q8d3#Yv3BjdV9Qv6oW#vmxpNT~D2h^}rJDsP=;MJ;9kK(iZ zyJvy4qQbXFjx~{=v4#b+SrAAQxIv@O`ll>9{$@)4h9Mz7Sb+UXU22B&0J){X6`5=p zFQ4gQf`EQ}{{|+TfU;2LF{YdX^{I5qM@^bMd6Ch?Gu3}#MIf@kf^Q>AwYhi zLZlRm6BrN?jPITM{CN@_=Nj=N(vZ5|1)DA@AY1KoS#-Z3LiJTAviXzC`)*l6dSHK; z9Yd$7Ct7nz#-f0>{E0n(!Upk87_hdIQ}EG7-ryaLy5-0zAMa;O_##lG$Rq&<$Yg>_ zDOw0zt!u5f(}faZZ}y@Q3Y%d=MD~7`q=KTo7P0GGNOCd>%j}{s_na z{`W5IcjYJNALg-L*)%`vJ3wfF+~h(8;K2kAa>1cqqGq6rKu0%ozoeu2LXk0b5ld{9 zZ@aI&*tt$24XujopV)*+7Ni+u!4quhkJIegLDU@s_!33lNENp z+9^!0wAc_=2P@yz5z%F+7RnG>Z>BS@l|(Y2XK|x0TAto2(!_lk0`v#*0{ubxmCzJM z;~;dhQCNFYpQ7+HV^-5bJ}6)vceB00Z6ze(1peD{CL)x-jRZOk=vvpEz>~HxEQK+J z{{^>!n%^o)^f{n6R^IcE?g>&v4Pruh6QJ4%IMFxbt(3F$a<*c%W)%)KdPCef>D#`x z)6B$xdfo2HAFd>#|A!cyU}Qgk2y4xHc~@Kg)t~uYiymUOTC|3x;hT~f8c;;BW0IHy zLFg2E~&zTh%RQH@6pN_ds(6dfDGc%ChxjJ)c6t36QKEJ6p zoJ+k`Veo~COyr9C&|jkE=L5mI`MFZO-dw8Mi+w`_7n8>Cyd1G2;MZgZd0Q%#A40h@ zygCpqR%F=2xaqyAFoM(1I=GXcLqG@WjkRo7-$-Z`7BHg4w8gm@LHo99dx00B%$Wbb z$^fkz2zs0b$6w_83Ufp%dCo2KkZYK*n=)*ns1Tx6e{5MZr^QSAP`e4%L_nFr&aIeX z_LMc=!nmqL&AS*mrzFjhl(MZ@V7)d;D$Amo@m?yY-l`)|`ft zJ8$hUsiOvo0v|A~VA@sy>|cl?;A!+~SRu4aaP0z0$w=UQ*tk@wY~gn^(=5SjEv#$&=z6?SZa0U z4JK{U%>VTJm%s3Hw1Vz~fn&xTOYL6bM1yYIaYn&khn=O@KI7D=HZ*XCRq}Oz|A8U) zqE=W+;3>M5`1^6Hl(TexmI^mj=iJfPF8_kw`CDd(J0@+d(r82voq8q?$8Ue$8@N({ z-wg`7ZWK6+?Jx(gVAMQ<6=;pu!Yle7AxpqPbI#8U18-ZEL+RGIkiAAhh#*X)!GTa25`s-@S$L8r>o?=qMT% ztVkv2Z1A1BQAF*i>8e^x-2js;4Z2d6(B%N7&yL#fEi8YDRNn`SNC%v8p<6UN#$Qn^ zT!Hxi`|fLCE(AK7s6DMJy{}#cT%FxPfW$xn6>zwDjPLWFfa4bEMz<&TLfoFoJ_H9x zHm5F@-C!AHp-&*O{I_k}LYcylx~@~)$q`f-(EI)r&MU3%^60n=Rl@dMfXX?Aaqqv_~DQo$8I}|3KmTS=j9W=APE&^#eMKYmQe(Uq6ld zn@t62gFQm6Q@3K-F$~Sm4nDGM<267qf&oewoB~ju?8X*wNT4nlESw05{q6KPQ) zGd%*?o#G_->D6QgmWQS#Dn{uhPOL%+|* z{y4^>kEKY}=Ur>E+b0i@7F85G?mCVs5ZJfE;0DJIYp!%39JHTDhcbU*a)*^Y4)XBSh|JXc(4n)k) zD`P;l_#yW|rbkN5;??w_mxl0FI|uwekicmA0_ef4%rW(p>g1x}YIj+o^YH``INJ`vOwabwy2A{!!%4h`H&Na&&VZi+Yy?A+hdd{c(l;4mJtWYx zt&(;mygR^B)ZIFej2ms^h5Nth|7D3Kse|kv z1PDeAMpzW7Cn6e9ux3+07m^3T!szLpeU$0N)G-l;1il$wk?|P+fT>jEPXCc|&c?du1jp%pEK z(9;Q`Ij!khU{yclf<-G6zp*-7Ica&Bg+6MK#9VN_bRwl{n6~BsRjN%K5ePM93$x;9 z)mv~@=o+qMZzS-*zP{v$dUhGgxZXqxWc{Jv|I=5}^vzvG37TZS3r5SBJ zgsNesy6HFJ(@KGY(8Sq-R#a?!Oz63fPZQ^QC}@ocvj{lrD?1`yQUAr&Obf7Gm9oQm;F^M|g{r4%kpUw2YAIG5>*e;RM-GK|34}LA9mD;-~D1`!oR|&IVXXqCKBfOh=8YgdRl^aO|5VN* zeyH7`Jp>9qyBHB>GJTCN3WWP0FgsD}F|Y}jG_WI=D~^EGc}3r3e|99dFfj(q{yXH+ z!?L};^L|2HXX^p$_APs=Cf@Xt6c56?Gy&MLUquolNIvVo+$)ltun5#_xdmk6a&=tU=jslP)am0B zeKylcj|ih8GCf&YCYK@{_WU^iF7$o_Y9_4KR$0!R0~zDGiZ7R;8>~ptn|&rQSz;=s zUKM!y2-t3-&85HZqIENCj4JlZF3hCOUB1yxoWUVlgp_ln32&>g@RRm}AXX|g3o(=UA zRntX(;XYs^00Zwb!NPQFZ6GJl?&Z#oPdhyA)=;rYJUbqqVC;8Fqbox81RV}nxs^Je z@vg@P@lM%4ZMKlAd@q;2G_e?=cl4?FPF@(fJq=^3SD}+*092qq?hX}1+Yj@#KX)Bs zo?i68bH4;kHvC^7(`LQ>JUTC}J)JnVV%Qk}H_3lIVjpapChHQl0XtG!+-be2q+sl1 zUz`yi@MyKLadlH5t=OBIC#SO^&g*_!<^wj`7n%-}*&CLt_>@J@IkfW&EMS`h$md^B z7))-p-+GK6LG-BOG%_+gQc0wcYA2*lefLq1@VCMITXr*cnBaofx}jaw9XRU&nZB=n z8(R(-aMtJ3IISG67Ry&0I%HQ#5Bw$%#@G7>UDV7g8#K%-p|3%_1#?`$$lrxglWkSa zQJVj9v1v>lY0A1L{vP8fdE3V00z%Ll--!36XQ($Io>_1h zq+|B3#XXMXxOxK1Z+|5>$cf0Zo;|}sFRCP)Lja@~`ldJeANClI)h@PQ$JITr&}ojH z+Qf^VTBFA_BR9AKv+{6eCQzm$IB|RtRnD>vg4{q>M z6mT*a3fhXc`*kOf(%_2ZUVQX;RAV7esaYKv zW5KDE(%W8)%jS3uEF)U~e5=|7y9%dJ^K9G!+>d(n{`)#@Sv2UtjlHy!AX}0i)%}iE*>d#M0R8ABQOe;vK5c2 z+zzQSpPInbzvQD7-0d!tzW)?o`zz;&PMH1P@+wo7@xZ5zV*`UUeEp)&f%T1kWo>I4 zyrhc)&Y~|Lw7U&13#8gohfh zAY!N_E0kX~V=n6=m=b0%q)2=4evHhL$7-jGt7N^|_8t>ZEH&7Hn4-(><=SxloJwLDK*WM!cxnx!DbBuTB z5yzy{t;dWV)*elUxpxNLQXy5ssWD-^J^Hn+Ki)&FT0wR89|Oo_2BVOtP)SWe!hMj; z&aW7XvRpFUhNd2*yTR`tRdNfN7u~&{7{Tlo$;!lsch2Y1TBe&-XCqzP+fM5yt3}tr zZzZ%_s>&qw>?QgVtq(d?YqcdP<0A)XO>d+q?&)aSYCU|X5d3yT&+IZ3T#{c%8J3iR z5JU^&D|`VL_UMv|%fJ#|WG*(fmS$6-oy}=RyN>6k6|W*kxJM2TCL~QX7CMqL?Zw307UE>ij%LLu+J6|;iSF*M1rw{;nq^pbdPpTeA1<}#nevo|dio-u z?08!zr>tXF7UA0b5X({vxjSTM7jht=5hN_!q6E_jEB;Q6#NN88SU@ z-4az78+YVN3KkX{Ppsrf=srHMxWr**VdF(sE6MEPhN4ud=>AIC2h`(p; zt<`e32kRqFPrF?Y-`ruqj^7CvT@Wau3*TVT-(I(BOQe63d~>rnik&CGUdhm?@xgCa z^JC`BC}BRCV^BxZn9HzO-+VB{LGw`PEVcC6kG%kXeF>HX9kIffKx5YPj_jBaY&Bou zvd1o$gC5=CLk>qwyB6)rMqWxv%zs$vin#XjUg$HWM-`h?!vt@8pA9JxRo^c=%M&NJ zBmW%bEdj-}8uXEY2SXJ{qDLPh>!xOI2e!-&MB5DYy|x`Gw>>gZSWO3Zr@-{fNp>8x zpQP%qN{{KwzX9GOXL?~Hzy2~2v*>&}Qg1#Ol1(i=ts34gC7PyLB_cK4{*1=*cH0dr z$yMK~h0m9?`L|1s8{p?sE|A)RDaY$;QoBpD{DAYohHjD{q#RE^BD?&Pt&DLh#=p8w~aS^o&Y#1Mm!_ zKtumyx$Cf{gxA^O_<_+Ic?64EYnVrHyi}7QL!(L8R#K-G^kFcB;#pTbjf4{Y8Da4b zn}_G60lR#8mlXV*W==(|q&{*nK>PKj;OcAqwZNDfR%De2>CBd7WgNJULGLu0+Hb}a zydpEc+UJXTx?>j+o{zfZHV-sGDs13}*$ALRf*w|%w%=OmW{_9Cd>#p}PulW0F6787 zRjl@8y+&=eKt&hWG1*Bs76rkd6ueZn8q!!>9(S^d^(l8U{((d4yGF=P;Ry<_-e~|c zwWiX$p=y#g&s)A5s=U+j+n85}+}FTw#fabW74mpV4IQ0D|AF)~50jSfWPa)8stm?e zss7YPW|~_id&_sHk&C3C+f^GSMf=x;{Wd4NzQ7YIbtqOG3g+@gsNfV$9Z!j{ykJ}U zpJ53jn3<5QeEJA4$|(G?-K0{o^|*PsPO9|nTR3{}U33*?m{9w`j(Z!tG|M)%WhJ!L6&d#_%q7ZU&JJ*5hvJzYp1Nvwoh`cdo3somUlLbuQ;*I}7 zXDK;2$SQsz~JS0JpX#6xgQ3 z;{8;K1_-a6C>Ep8VB(w0-$KfI2XhcP8%)v#UeE@<{KpLWBZ6I~c`4vspN$R1RYtMf z=g!1<-+-$8DFAweptEblD6ky-;GkBSI%lB}r`y)^mJ8~!vHa1I+ftwxbyn$(_Khje zGgEHLgc}asu=34xf%%6c3LCu2ff{_$vLd@!n=SvL|u zoSM@*%&;7o*Iz;KGt2U^VEECa9EL4?H~iWv&@vn3A`7be%tWEF{tH1SYF%-8`Izk2 zgssx|AaC~9rG9|aFOTz0ck0C`&_@?$gV~iwGh{wk{FEX8^DR1ObJ6X^H}BvQ4MZh{ z2T7&wzTA!EhzRPRR7-cl(GOT8K+6?~xAUj!Bf2gkGX&^d+<{eDwMBw*j?EB#zxyms zT?NfnW|aSQFrFOB9i&UT{Od9w>lU%WBcPSWy_% z&^4Lri>q~dP#0Yklk;ksx6Z|Xd8YTMEv|hSHU}OP^9?Od9_&xKe>Wa8CYVpmhJlbt z*VfvGn81pk7^qEtEAHY;Hv5)|pWXN%*L?13 zC@_xIew>1rGoE@Mt-mTLwmOQ-2Qo^zu1sQx)LzkUv?AkQh1njlcK+I#6xK$|$h5a9x@8;On z{2yI!9T(-^bq_0|fT)y$qI60l0tyTr(%mK9Ak8p@2ofUQ-Q6*CN_P%1l+w&FAOkZ< zzvJ`#&biNdAMgL>5BSVA-`IQYwbtGj?Oxp5m*a^-$IgU*j1sr7@}#K+e(cA9Yfaw) zfq{4o(hoV%|1Y{x-wNQs((HZ{-@@hEvxMSN~F%%IEQUuD8SfXPRHEWh6JA`vgu`$?0Dzl!AK zFuO^^x^oM=fh(X#q)z$K_4Ct|SFgztVJO(ZPJvfD`lCsCYOJ7P^fHha)Gsi~(fYYH zyzn2p@b^<=c3$-TdfyU#11n7D+q6VkWnWRAADbK%SzDSTH4| z&!BFfO$(BcayLlZ=wKH`^yaG-&+VcvraqXXB2n2+E%C;c?iHzSaMJ^c@b^}&GyHi%| zW=Bx^)uaV??RcPvH=#nfu4kun$XBZ8BW=_2}%>70faa`Gh`a>CZ`V9Bn0CO_*1hpS}m`Qs30^AS?3* z*R+|xGd+BdSXC&LkAbft<1$oLbTw^TK+sYy%{(BERFrfzVdLg)rl!5E4*17r0 zS7vKDr1IUtS7bfx@$AC#yb4f^~?Nt0`Kv1veK)MUjpRq@^6doH} zJ-+n(bE!>q^q>{A;+;rL+@dQ!amxE|4ITo$e?MP9H=&M~OCR|AsE?BL5pa6wJ?BJe zpTQ8AqEFM)DK(rdsi0}VVtH;Tl%2BiL`QG5{y2ZE!g@HR)-J2Qu;Q>)5w{{L=}Y%W zS<(`sCB6!{Z|G&KE6(Xmw&dE`E&gF*>>B zfc%(>(DcQB_n%GwSMU%m(YR>xj(->A^hnkV^llV_j%wGY8X0cTW|SE+u>Vbd{%@2{ zOBKDYZ;h_2wH;&@Ru*u7rx_8~a#tM)ce9gWuxhf1n-KdL9>TDiz(C`bmL z3fm~^ukqs_(CXINx1gN?{+;vgoYi;TA2w$gUsc8-|Nf2EwB9fXPnl68=#6(scYh4zl(FIJ)|Zs4cTUQ8J<-5>jE?o z-lnq}PAJ2#WPe2V?fLd%IWg~MDVsn8gPOb3fB{hIfivm5{^6R3GkycX1aFb4HTkwD zOqP&u8zqLe)7!Tspj#?Zm0^{6fFQ+FVMpbatmL;EP&!wVQ-nhV~(U=W@c`9cYd#j3#cahC)Hm&`_J>R zkc+*3H1#4GdK{eLazr=3{pLEPpB ziE!YGxH}k@I>A1n70EAuz?RV@$wCxnk57Gl@~%{5`8j%x2#ro0PqaWb{-ZhEWh1O@%h0zQv0hSL7JRLJE5H1 zXFPLQUTQ3{jWW>lq1nPee%n7kevA;cPpT_gK+*o!J61@eD3DKkaUv7uk;N z7m-<<<|=5TWsntpSrZE=l1Qn{x^H6drw7hQtG$Uq#Wvv+@(z0JrC zw@-ptDTqGyzC8W2nmL;Gs)nAS=%8;jtQ0q5f4ozo@1wETThoohH)7nQXJ7uMumn^{ zQ@aW=CEb`dR4nGadv|cQK!E`b)`D~7CgtR|Y|bY;+kAjFcX9?RZbt?XQsjAAg0!<)iUaQm%uou_s~0>n-d{uVOS>Rg42q53bU+!4DzS#!46PW#4V zm0Q+u)O)`@grQ*un7+zQhhb^E3K_aG-p_cjKMWN0$9r++$rHOk=K117<$S(jSCX(% z;vSBNtdDY?Rk&>lOR&+jH+2p0HD|e0Wnvd^xnzajE_q@!aiiWRZNQs|esE+m@1Jh_ zN6{#3Qx0)0Gf2JA^UtPjUtvVti+_QW4>T%z%;7QGnbNL>$-o}HfkM+)e)_`<1N{+v z4?UetuGLgonZ5S!qP?9<%Tal&%z;h(K4e%pPfG$^enkWMT zYmrWWdp1?(J$p;$)qZS{I1Oqn!?bc9Qu^Yu#wM%`t{zedKpy0}46azg}5;WyRab*N0WWY{Fd-;WTB4z^gdW(G@8 zzX|X?OpXmAXA{2o!K%j2M-UB>FJmKjK^^X}Pn2&oeud-#<_pQ~dE|q)rog;8==IXR z3ip6A^seoilj4N%9~bTxb{j6{pxHIdBh@Z&4!8Qg^qxB2stMA*^Y$L5UV1NyoBk`~ zc&;rG>Y7fvn4D0$8^&(@3&bB4`Z@V_zz<^Pv4pJ>V*!X!N$L#B!XNbs{VRfUgXVH1 zbMVa8&<4zA_Ue4S&6bQ(HzFM;`?;H3xP@J=iE(wqhg`Wro)Imf&*4&2nFe$lCe1&s z=IQ-ARwV4e*2C;KN6`t88M-pL9RHnQ1_8bV!wQ;XA5t1S*2m~SFJZ8>sPg@yJEn6v zG)1(y_Bt4>#cw}v@&){FSgR^{1MDQ(0?LTnVeaqnjOp-rVVKt<|8W67-norPfXEK&>9Nq6 zyMLE@jQOj8Y-|?RC$b`Y%EhGh_R|vY^N1&N9S5>=QTbJRyX`6Bkb>sn9|c>0X=VM0 zdVLx{7Xr?Rq(N2u7a8IVOTy^I)XFw{W*R#?wu=hv7NT0%t5-~}ug=r}i3k_OV(WPQ zVZB~CCR+U254eAHr7?U;~DtOoDW~Kr1Gh?6zS!s4vEp6y0n z&zC+fVe}X{WXvpgH4Lv0L5+l5$qi%SjTkSVOv@5_|=MMriZSK^gy6yaw9!MTvxGHd{-fR zX)WO#xH~J-qBX;4#5=OOu>yHCAlUfL|Ne~FiwsS1t^Y=#%%x$l`N37kOKl#cRk#cC z4%rK)43Dwi--leCsEZC>2vZ`z6Yzs5l)0*?@p1xZYub2p7^Te7q?Hcj3?w-@0l5Z!{XaO6C;=7#zO9=Vq+cY4qO@ z40?wvtPq+jLwd6`lVnp6SlY@pwWVFM1~4tBsv}LD6_BptU>=jd(th}6;o6Yz^-l8#wXJXoc{V_qr&4uD}`{WfB6<5z`wRyV~{>uUS zg~@Y22V0InF{Bt&lc$gk`;k#yE>z2CDZS`zAK3K~*UK95Wu7A5=FTfKoj!8kpO_a8 zFFvKzEeK30utUTZ*!xd;4OC8&oDvI=Op?&Mk-#vqC>GxeR6}vz1pCeaex7!@>3n>h zs%oT5RC`YIj{J`}Hb9KzC!Hk#)gA7#&riT}wwEm!h@{~%#U`wc)Cqb22%C>K0N@0D zDa+dmPa!Z&9FFysoM`5S*lK=Bim%pWN2)+m-QNI3@$|s+TI`M#h1C$p+*j=!F6@F~ z2I?4$&rTHfwZs8ztz7xz*RA#uD8lToto|R~>URZ`&56AnR1S?J`hA};M4C#AQ3toG znWXp;DkFJHe$bYg26NT)V?-evvz%j29ahn$X>YhhceJ(m?2+4~#15>_ao1E-lVnk5vyV2E=0pqT`qRFW%AlSwE+Qmeo>qPfb23pv8R< zB?^())kuUl@{MvcMs|fm`oBet|v7N zMSGWiuD}E5$2akI2+p}L1`ZY@cu!*I=#t0oWP8tDd=L*a02gb9>7xYlPr;L#ct#RG zP#I7B35A$R{)F_t3Ir^0?3~K?*%her`;I^xkzUeTSwq8qVi|X)K!AY> zF_m)L=9o7-AU>Z|BepUof6}~x<=EDWzjwpa$mp|zzGs)F8v4+)S*rVlxAciMa?THv zslUj~{XU!L4Y4+q>TYryGW)Q2C%}dlHz^#ADj(;bG>>KQL z4)M9+TACui!`yveI7f=tP{XyhLYN`Rq_`IUp}HZNoHVXIEtX?$qKdVz(d*W*Pp^?* zP#r!b-9<0yWWJXe3%}(2wUS2P;#(j5ZR|^{{J{FMH6CX2=9Fg2&2X*S4zK6olrfZ3 zy(+^#(rt0ZwBJa=rlXO&-w1x1wkF?ps1+P@IBG?AU$?I4IFSofEUrbB*?~%S(TmV; z3|582bjjtAc9sbEm9(C3M*A{dJMuTxruBqeeIO^!}Hs=)EDXj;R((?&MsA zm2e>y8{KA8v?~>I`8hW~YphfXPd}jvTsyranptws zDCy zvhk#n1a0L|0lCb_+F2v@VLIR(@3x7~Xm>^IUHoVX+N%fA09_T5T}Fz?NF@y>rwpY; z*wMIBH?vpz>x@@bGe`R=ZBK=T^3gLwBB+8#3@^&q3|z@#v}^4y$G)7Nl}|uS?s7T} zVA!Fc@*Ws2>_2qY-{iyQllInQ;RUT_v+g3 zlzkFrbU|&u5G~V21j=Zw2xXw1@XQDeJ(D&knfK9UiFEwD=R;I=`rUJE|D`!SJW`VCpf&NPOv3F|t`NB2$;&}e&P`}N!FMw%4=;bOm zyg!{W&%(Z6H~#GfJ60pMmPF>Asn^R@^Xopca`+Z_CEBPT#i`ZKK7;*d=5nCE?^B8-B7LqsS&CvS z-U$(rRITx59PU#Y#YEdyW-gzhZ(L?eiaYu=6Po5QF=!TJ*N7zfZ$1VTq9f5u?%%fYD`rE76jUCdlnn3B+@ndsuoGL=CzqHm`Lv{z)AsEDbFV7GVRQZY;P zo-=8$zl)-XCT*VkXEE@GC$XNJ4(tEs)Br_AQIK-l)oASK^-OLaZXetE1+$;J1bIw< zLDryN;x8hfuc1SjDm#aSc`VTo{+e|-ZMk~ww%zRt`N5JQ_N?>I>JfRKzMFhuj!X21 zd(SIO&DwF}w{1Juq%D7TZU7m~BM`V;%LiOOW;+^WKcE}Z4z19fhaMmh#adZX0)YFR zlJUw5yP~nE15J0r>vH?!JigI@`+-7^D(HutZhr{1dBv}`Ij}YaLtVVODF_bh;3JP) za%(D>WwIU5G%)*K_KAb|nZ@a`c({RU5&kO=`P%T%`CxbeONZcPFGRdOtjU_n|}Xgja_|1 z1KKInz%0~&f8|Ux@DJ5-^sj^D|EXaV`ZmAGRDy||)7i?|pAZEyMt7>juY*%XCy0A@ z57Ex6Z>;wtkVzfwEQQ397-7AZXrOHzz- ztc#e^04{5p!b?lW?`#b^V+;^j<*?Sr?QJjuss7B{OmR3Dr$?CCk64 z%s0j}tlt%@bjU0AuLltnG=H(?I|9_K;y>5GGkgu^@Hu&9lLv^}9}IH6cH1;vO-^eA)^OQ<65Hc_Mnp$pp`YL3+I#)xQhdm!Z{gZIV&$m3!`NcS^eoLA z)y%%+XKAkLWoe3QfURTfs&d*?h}!eBN)w)}<$bP@n{5$sd}s{|7*nKpJkv4=F2B7b zZm%tgF+vb$X;I;1TY+Ot{s1-u#>A;xdd^m&NDqnTvbu&zQW)v#nHginhoION9|b?# zF}yTjL(!Z6b}21W8P^2aLV1cwHDMMzS4$S#I1oW&@d@m z{ti!9|GB~1;bgG)t9CxYP{h8Cpo5k2mWlSsY*D}RxFEpMNl}ep?$L^sX z={5OYie#x@&bDg;cwjryfNrsMQG%qsan5>MN)w}FIz`h!{-ttJD z(4ekhMhudb@TBUOc#lu4vi3x2xg@BV)_9|u)qYp*u z861o5^?ySS!aP2R9zjP6*Rq-P=qF4LQOl$sP z=njV@`_8h(<5E&TU7C*{YoOh9cB)}!G7P)AKPe=l(Jb7?o#JT8)zn)###^Y?bn5l1 zYaP9ePxyZT)KLaeScudBI`+$|K&skX)KD#QpK9(fRQWYEujL{3yf& zBkj0uN3oz#c{wY0erC{uuMP zxUgjWO*)A`hqC!R;f8lnt8-yXhjx8m&5Tl^DlddjA&_wmW-^ZF_7le ztN)YuyXI&Kl>|DMO4eiCSCp~yvR6@n#Y-CDt5?g8FOB9Yb;EN3-#*W_E;x!W^o6G{ z26o9RTYIf{8bt3ME;5Ap`BXhvT(stYdM~voj%+~bQ&Pawor)Tp(rMQqv zYGPM(=mAUi$zpX&$hse%E~-i0Xaz_AbNQJoZ;d9Qvp!RRXOZ0^)29x4S9l0D0

    a zDl{iWSi$^xen*!Lj@R>%Ope!#sfrXz*nAE;&+6Z-98hS2+^M8PmENq$e=DgDQm4TS zk;olz*#Ozbq52DR)r@{}x>cFm*#w3ru@{TC< z>t~CBc9OZ=hdi$^J27Hrr-iY}7jR2>5NMHAv6z(UVd_(z0&Ay=k~_{@U{r85@21e; zoJ2xijAq0P=`6Jyu!hYUQJ@91jwlv;|K}Sth)pZI6Sc2#@$fw)(+z+Tn2QRn3L}xDH|Ai=G{Ees8tE5MZYQ$w*y|S9&WWE%tuO2 zjqT~*{>rpjya1?2T&Y3_vuv)}$rEL?m>AE_~#z-)afPEuN!k{o6nES@|FZeQRRPp^00+~?Yi{g z#!|mI8~A$1O;qS#mYuF(hL6eaJsuPo`GxRy63sVJi|D7nX}3X4ixaB}ob@-eClH^+ z8VlrLDwA7IRi)Gi^9L9OQifJM?6)jWdB1^U`l>+$6N8;)$9qBSew#@6ZGFH@)AdtH zd_DDH`LdsOuUxKz#QuH@#fuiW@xfJ3zQhICTV{MKQRbS3DADSu&%)oa`)zpqssPQK$U!0qGOj@O+Y z?ABt|BA;^Fo@~zos(ELblpfOWzZS@XL-Y2Tadpe=8W#%-ahkK9Qs&kditBrvG^=W2 zh0c9AbJ{5n!`Iu8F%bzZf*S2Bieyme(%7NhDobq64PLTTm52ETQm{fO<1)mN(RI-p z__}oyYlOLi7WQFfBCgoYa}o6;cW2c+8e9D&*P(aZ216XnoH* zdoQUT!WvwV7$!XcFCE~o+Hw5T47q=#uBMth=>3j<-DM~IAX>ZavY^&~ZPojvYaT;% z_o`0}(M1W<>~I*^dM75Qr$>Vhyla*{59KW`dcEcw9<+=dhsz@~LMY<~vb?pz0_Xj- z!GtFZ>@r(hoUQssA1*2km$En?Z~lHygm?JoifhC9>`nh=SF-hl@XN(oNRjxWPzkdt&#;gxwB zdKazY;dMo)`-y=#mYcj)-UY-B9{fYvlMS($W`h47GT{?4Oj~mvU-sUgqV7Fn%NNWKZG9CzjHB8rUb<&)^o?rzu zz0kT)FEVlvPzDL&yJCdk%4rt^61l6NGyv#t3yG-Rig zsw&347!5w~4!M{^DYFf7@b>e2%imHBpb>0k%37yRR#c zZ!r??YzI2|We>}4jRy0!&--npKKNi`OM2wXjW-}pyh6GM^V_zGi?S>8h;m1l+jHtG z8AfYYjc#kc$m%DCH%EnB*J-zim* zKR8&gKZSH&w<-@eA1A%Q*U^<7N$Ny9096(*B@TT+@3}4dxHf09uBJtoE~AM9{9sv~ z*^=D|kvGtu{bQyHosn>3ue8f9!Bs{*t{3#_onE@v=T0IvKoj?2&n=ky*`|oE68K0J zc-rY#w4Z|Hf0mMCfAKA;gVFeAPU7KFX}gcZ#s2aqu+Z@`9)nAVVaHB|Q+?YhbjBS@ zX-;LJ<#3Z0bz_#+)aV$8<}2^3@CNRO6|rA{b<+M#@jQscoW|~&0(0E2SOMSZKIXmW z<@-(h?2_CDzz=52m@1^mN~~IxM9P9HCVr<*&(xugIE>Mst4bI-9scNtqz*9;x9oxP zJ2^aq4G`b;XidRgRCh3O9gLyGQZo{~{V?K<5^=bciT@!7R(8ZdvA?=&!|( zurXbS#E0KNYD##wFb#706v0shrw&7l`vD)(xmpBzgHVU!#1%~=3x<2}ZHf}(8Y%Y* zyp_~W8sw8I5P!X%^I@PEFIQ;sNIh_YA|4Pg=#8xqb*zRjlNZP5-Ys+0o zTI6wG`DXZ~GjUMo?oHbw@v@zm-%%;?=GZLu9e0A@76W!AFt!=&)yz^%{4w@3fr>16 zu0CgB%cb8vews4u$=JI@?NEnC?A{SOUmFpRvntE21*aDc7?ROraB;a~X?y1d3{)VD zKRe-AA!f|);%R4pgI;N?DKx+oH>n%y-TmBSJ`Se$sz6$H;gwfcZ5Teanz*;~gs!oX zanHtB8GHI=GVF|%_TV(n`+C1wuQTEsy>La`y;+A@yP$Fpa(jvQS9ncfI*3G@Ls-_$ z^gKhx(C{hCkM(SEX&}edhR8*+b&YLN^aR;v9@T{)@U*o<+2~T-|I~6Nvhxz&*(jNx zF?`rCW_;Bjh-~Kct583hJ1WWom;i;YJVj7u#HhWunxNvVx|^LU$#6Nr&+xiqRDfwZuM+0at{U?$tfTUQrR;2N?zwghSM5AEr}EiYA{esj{HbYd zUk^CNE~V=7Zl6jF0p- zLf6e!4^Q7%`$AltpxRc|PH{Oku9nyVO{{>*$;e#P;*>9?MjZ7ZXG`0_df(LfSb3y{AIp2X(|a(h3J5Pf#L-QoGz}jvoyv%Y6#Z1~j9S?j!xt1V zBwP718i|e^`k7m430o zntMoO*{ve4Ss&f-y3u2zVoApsj+gL1F_I{>5A%k?q&2$6HIMKY3!M|hr-0MTmuKTW zQ7;eEc|lZ{<9W5e8+!H+1TR)^v5#IZtoh{1_O_S}cmVs_ITMzmR%l5`He8Uiy1?}k ztJI)k|BEo?k<1nKT2F6JnY50dTUGp92+uTS<_qB}p;Ty(p=XDv(aEQ!t+}))OlNTJ z3lZmG4=Eq+WMNwqe2oOD`XJ)#;iSrQg01OE?S<;c~fG3;|*XTPXmShOhCr#*Q z>L{lRvELwnZlKJy&a!?DObkw3rg`$OxAEPB;ACB_Yj2gU{Yokd;oLy!)x>6g?fOR9ohUsNPK`vIeT3zGLRBpjsvd?){Fy3l{YjLlDa{AC~Jq-A8lpwvfwh{S6fU zBK&@9)k>mcljL$i`B0Vy6)Hm*K`F3N73YbMp>Bl^N!Bxd>|0`TI_^ZZk)IlisSyi< zHu!u|<d&~h8mpMaD$(q1#khM`Tcw=4md^u5hJvNjN3IH- z+rl5#Eu2weXrFSB8HDkdZv*VxaswN5+=^fSx>&w*&GfT;N`)ZP?BgAoGfPyM5a z4OZtwo%a*dBsrmRJ4GTF1=4M+el^5b48**L=!iQ2N1^jx9+!j50_m4#CHa>p%9kJF zm{i&C`-bUVBdbI>CMZQL3fPlIkR=m7Vt(5Ik?BErJ$F@aYK8cKUxO~bv&h-r1g~T# zH2BcFZ)6kU1k5&xB*leGg-hI=`1>f-|I`Q>&{#Ue;Mlj`zuwbT%=GUh=`GN+y_9Ph2B<^uAsk`u`|O)!}29@6;djPv6wL2HhjMFMZ<7G&o%Y4={ zBj@E++Sa5pB7veRIeJ&}kn)`(N=)ZKLyWYemQ7ojkBRL1vUJqfWC(d*UvKrz9k*a) zXm(#SX1K;w9s9*IwiWRrBE`@Dk?{Y*(C>S3o5D1YhN1d0Hd9fLr2C~FZm9W40VhX< z;IyGntF1W6;u_4|rWnm3w`zAXF7ioU^z^XWHj><(lVZq#9cW2HN|sdxR$&3tLwxFC z9l;czyqNZrqBTL=b)*dT0;ub7m-9fftz|01nB$wPQiET0cCT)v02%Zq8n|8=N#0*` z(o|BNC}pOMvIOsP9wn|wh@&0xwzrA}n5Yqzr5kw78@Fwv@75J4=UTaN&6Zz>uVDVdK2np_2o# zRVWlx9=6ggse!ssT#@eJo@JuJHwZKt4Z3)Dj$S?1T9@m_2oj;rD$9z z*iS`7wGEEbp>2s)oX^`FUBq`mU-tEP%=9O9o+~oEG6ZGrd8SeMuI~kZ(3r34%XR`C zKTlVv>5Uip>FeYmfR@@~pJN^6WE_={gg|bHAOj=6$s-g|0FiSEMbix@p9`9o zk2A7B{yJg8o(JjmrT%9l2eDNCrv!XqI{Ju~twK5^rs1;Xr5@-l;xQ(AH?I0fuAT&^ z0x)#Krj=I9QmaPN|>JIQJC^ z)uUS8Z;-VtWLYEu+D%zVGSS5@Q)wINlMt^}2{kwp@Y#K6^l(e7sT-`twdu9D5o#U@N+HN|WCR;}ezI6^1E2|gg zE)JiGf4s|nJzw5GoW(gje(@sCTiP3VSjOxWwu0T-G|JkU6gtPcCM}>P%YPadhj9G# zx~Tr@3qI|8cu0sCyPl!n!qsZdQbD`Wf!}+p?~Fzm3%7MuuU@EmqMcp4?+Mu{4oCO{ zJ}k=Ae7VA7paW}3jc%#~$E*qoo!kAw5bHTd88erATNUje7S96x43l)7wO^k%bIfgc zyPIxq`%YI;Pc(jqMV7BW!uJ!GWvd)VqnNQkK#_dMHDj%)kEvV~C zAb|)+6$s%e-|0t`DFT#EOq_2MTyx*S2eilaCk%`=w(X&iF;Um#xhfxh_O*<^W&yM_azU`n$G$%~U0q3(!WZyIGx=)( zn*lakHuB#IowJDB{5JI0Y)+Qer(-o(p2(O~j0^7ydtoN-Bx$fPG_H8Z?Jxhh7V~L; zS!1oS`m;rQXw4s)1y_9~$Y0c8Jo$UZ%0H38(-#3snYwUOEup$cl?2Ssu&!t)2j;h~ zUWK{*K6>^BSZ0m%QkmOD&`vz5cNu;R;dA z>~#=5#^2;lO%vmgcSd7u(}8K<4hGYs`{)}2v$BlF?;!u~5BN`xm^8+U$@eRoX7iik z{q`Bt|xIqqT&|g_X>XBBri1lx6UWE^{g6AR7(7zI7$`FlynY~n7S|a4TJa|UZd`9H`tHddO<`+?>_0iU(#I1JODiU#6gN4%h&!_v> zicL(JBR7ZTF6@h$=4*mQv%PT#rX}bXy$g6JOBk1{Ljqaa;ll*TwlC-kg^(M+G6#bm zb23MD=5iD)FY*!uqh9U9OF_|t_}nP<)3idPxVbk~5oNJ{q zhuFNfVxlP%F6=AU`vv0eT>S@=-PWLI4vBF4U33sC#tt@T^+3R zRy%3v#edss?#YXW>nhfV z&l@1x7a}7B$}z%poZq+J-+E|waW29hO1h@dX_1tE)x)}e5_8?454rdzxD3c%aDbRD z$Wygf3TL(xmCf5O9#?q`j}ja-Fd;4)(6$S<{?~1dUL#|;oKf%nBrTP;=Gyb#Qq&D- zhQmP*mmHqWHlW?+UwWa$irQ8vPd9+IMh$D%x0rx32B?g^$Uj|X1VP>szP2W6yk269 zHMf7uEQm4m6JkfMxWrYNy0^4MYg$d_4r0_VeW}zCf`dcrIKSFA9wVw}<8;b8HsO7- zm|0n}j>6te?1x9BIrg1P_Kn8ces;GCJJzj%y23;HV)qPC`w^wXr6UZ zwV6xnekSy#KBJfELi8%^0KmnXIHp*p2WP}Za%VK z(n4Tue_1?wKmXq#^j!4&w((o)BFF_FyPo@ff$Py5cU?@^;dYpH`tf2A5E<~U;-Qf?d zG2fzi6$ z)11H+S9w?c%T65q%l^{%?Inn$JVE6ngTc~?&N-DfE%Zthx-D2HD?uB&WHkYASQ#=i z1Y~Kipk4>+Xlu0AzftPEdExeRqum+MMwWV8k|m-10{G#eN87xS7@$WY?cH?(>HpsH z%$>VT&Jg+R?kOlCq(mP!d3575PAm=P6R2fh-^qFMpGlk?e=gK)eL z6Ru^F;p*~r2d6#W9kr>{=J7%0XOE1&BvFmeDms5d)^8sc3736;M86|jC2v*Wo!EmPbH}q z5I}t8whp5B zt!g4NLLsSZIjh?nYd9|drd^n-Ac5z=c*F_2A%Q`HrOeKFplaNbzLe1r{&<_J)c5ON zVrBBnq2dOARLa+UN5$EFSE7zB7hp+Y#zf;BpEk?*&PP>lobmTtKCQcwOZ@&mvi*_n zq@fE-YDa_@Pe@zg6Iz^M==1$0>8Yo0SV6uh#F=vOu}=HV%GEdpirE+rakBi~M_$rc z1(b)!*CacT9fG%CntzelQ5kh_D03{rH1!ujHgsi|ON7qGyo~uf+H_bvv&9anjMno` z43x&Sbe-z|XOn;A+2_IP*uXE`>>u}_>1ypK-uDq9azsvesaI^oAti#O#fBojd(>k|dba8$< zVY-W4h}IpFwW#HhuM#--EjNIzWd025v>HFL*J4=$RaF^3NX#Peqboo8ar)M`;VO0B z;l5&sK~=YO8WXvG2cMSZjM~d8!2J2JxqQ6Vj=fL0NY?h#h8;WoCk|=zo%PHvUhI0H z*O2w=vw@_1@2`#J@ARzBYJ4=8T28*M_No+}1Fu4?1@lyofg=#GO zACsm2M$_jC)Mm$#Q#-jAM(IRG7r0|aUVc(Q7Cqj13naTXr;8j(J<$TG5bdc( zh_zHJmUQPA9<79_NfG@G$h&a3Z<%!y|Gl2lyRriZtzy=aFpV=ma* zUVC7+82^o@VU5Y+{a*S7pfuh_IO>=TGxSmgIqPiW6y~^-f)RbYxQPCJ-c@Xz6eTY3b=N>?8^zf z_&)$WSRO-mYhH1?oFoA~s2nF$o@2M+o$NR?ds4jk9}E=0fg40Z^Sj4W&&>nT{rak$ zxS^Ec*xty+Dv;noW2$PK7HNCAu!qD#iwKA8Rtc{=R{?gVD(Xk{X$Wq`7OK`bU9$nh za^r1f@RqdINi&^=!9Ftq|8JWdT=Bh+&xTj=P$U@xCxUX0_AJ!kCW4_E)e{mNsSYZ2 z!{0Fbg7|5))Rgk$nX#Vz^`H<9o-t*7?Ig*;u1h^uybLv+Q$vd}Sh5W5o+q%4sEwJB#UjOMRdS^x8tGRBnKS4LjYg zAI-Gifp-EosiLfT`B@mZN>*)Vgoe z2=idB|IQ3k?p-&vwy4o(4#qmw^zQqir=t4f7}loN^U$o3Y3t<1LslT{HG=W;*KMjL zXgwb&P`ADlktsS5yFD z7gHY)jeajH+XEZ%AF0>G=l7GKHtA}N?!R3k?`^`2jxnDlu_+?tF@^g@&N`GQmZ~+c zAM-b7%2r4ue4lS(Zv{82xTCRKuNuN4cVx`{HLh&mfU=w_WDc3=WZe;sATD<~_~WFn z41&6>h_fv+M$z_k}8g^_Vhp(>ZKWB`I#RU&VWJ%er1CH zA6;J=7UjaNt%%anozfvK-Q6H9-Cfc#gn)EO3=JclLpKuA-8q!BbmtJ?c=mqZbKZT< z_n&8;=ek&V-}hR}0IJO0*axg_K|3A~1MeL`;~(zV$PXH-U~4uXuqViv4#ek5(`GAE;M!ev14BD&)(+^JC}7cQ+#qQiSnr9}@7OH2!M+=DjIe^>(Hdkg z^3*e}G-w2)wf?uKsC2>>DYoAOHgm2!PU^T|V{n+YYQT)=?!a|onHbnkPO6t0;(rvR zunqu*IZ`fl)ogCVlyBg&a*d>7F1j*Z7)fL}3b*=HZ84z*TLUc#8wsYb#QZbq*<*hA zjYz+Z;oTYO80)v-x%D(dKjNH&%J;BM#qzK|06#^x^iBeE{DI{E*xh4nj%ZJf&Q+Gk zr{jE^+UBC_z4%_kLeE$6GT*ElH76A+5bsQrgxVWsr!vvNI#%hIS$t7CzR^3+XY$*l zx$LX>ep97O5Rtt~Woad2oR81P+R`6RHYYx3E2}4c66ekBN_LW}OJsgNg=HpU)qY4}dHh zsVPH0-X-&Lc0xz(^C>h(75Kcp4yxr~2zP76H!1JFyAX6!YT8%tEP7eoLY9QIaGD3HkXc%hQ_f?q}pUYi} zk*+S;8SS)ZWgV|q1@Dw51W0=G2@|A502*~R$s8hnQ0nd^u_X)kG+wz39e1d^#*cL| z3+epy0crOtDs&$>v{sX;=lYh)T4+?BBz9P;`8ci5pN(PP_r_Cxsg|oj-T8SuAV%>mamtg67Omk&>_UDL$0)|E z;(;DJyLoxb`fEKM%@ohy@}DB{Ay5`vrqvt@92*g*&F0%xL7i} z?ZWwrhrMXU66*1z7f^;ht)@U6G&4Eq-^wi3(D6|3#7z^>J8T50b6O*GBVRZ;Z}mIw zqTvHPnNr=sNUaJo3*v}>FsY&9zmlK5WR;(eU+efSdsAi?IDv-0edDpFowaIVtu=;+#MOETY@vu+!A%N(@d#SC!2=e zS+N!~lFrhDNu`YQA4jHDEG*WCnH=Pb40#x20-OW%e1*hSc(u{*6hiJTp)GfS+3U>) zg6VCNiH$h1uqoS%u%B1=^7F4g{F0`v4Q{jGGBGmGXrw4L>@Xzu-gT9W0+sGVtkFWx z-}BnPv!gMnW%sPWko}x1)xuvcoOr=(e$=C#Qq;T9{Z`q}Wbs$IoyH@?f4po>=8|s% zH@f_SZ#_u{WrhmY#a#(IGalTFunsKa1L;16Xb+Zc_Xd1ckeO^%=gZ))Xww7lXaVQd z&YISU`BI|Es^nddZs@4giR~s#2X%YNUh|Ta;sr0|^lM!@0jF!Fiq?1?f=c`o=LAw) z0tACVfXMlJGgkfwFN&(KPtj(D!^nG=S0;V#Z3$ZJi)v-<-bD|C}bPr@dXH?Bn(k^D=xDt1aA zH% z*@|hOaQNs>J>}%advy>Gpx-zjm2DjU2NM^BW8Z$9A=7=Gd~N36s8CQ3NFT%2u(oY1 znW1Ey-L6NL2u8Lz+Sc2y50}>)j(dYMWE)p=J;4}Vzh8&sHcN7i6~U{5VG64dR%`>x zitDmj2JUYord@UpL!^^o+KKltWKGZ7V&+8nbJ;wz;WM1yeXWr_|HF0Tk7}C$4UY=L zEYebj#|kG)OP}Ivb;AD!b2@k4eTGd{4VIwmyW!XRyJ}_&XQkjjuY)s+ChOqgh?7Jw zGGfpmh$DY(p1yMNx6<(NyYaA_6ugun-rIFa@RtVIkIJ3tLS;zAns8@jY<+b6*_h9S zhSbZj>S7s8?C6|u#P+R~?DCZc3sBUwg66giw{bmmk*o#cL>lqB|Ie)^r_>m7tG8d; zX*6vbXtNnB32YN8lY?KaUf~pZ+qJy0PwUt~Nl6bWZ~tn!KP;-Syrmj48B#m|ZgM4W zhs{wt;Dh-Kj5Oi28^7^oJ7EdoRB#1o*o|9hJw&P0^}jEgaAJ3|Bi-JXR|g#S*LN~j z(oQIl$}Z#yH3~%V)XB0pi+uSuE9RAhR?NsAtxgbn+E~BZ@{KDXN8#PUG$9`aKsU;L z`G(5pUJ1&jB@yxlP%mo9CfrCA9eQ`G0seLF+-9~#rIC|oi`DHN>3O2`JZ@}j6yx%bX-}r0=_@qG z+r1@a_XB5Pp)+wQAy1zWm%>N=8~L?icAlgz2s-o4$uWu0fkfvcv9>{L2;-gno#)3Zg8#RGou!QAF0M+**1V}1Yg z=4bUe|0mK34Yof^(M6H!qF96+KL|71;5s(UO_mS|k z*;6TFaX2?KAw-Y*T@DB7-+lSNdE8D!Ohl#=1;d<;jjSmqmCAla^q-cq6Gj07QBDAG zBl%raEe*|v&6BtRV09mdAIZw?$??Z#^K1s+wh`hf?A!^haRTW!(0a=dpnVeMlh~q% zB`a>4_&)e`>WVQj@hi*I;Ogpr&P|>*z#5NP^bwC4)ca)|sN>8_@Or`PC&36e+$&22 z!k4DFkXdH;vYWQ+(vhJdA)^N#CJrj9%{XEn&a#$L=+`8_&AxM)o@VDw=@nwS*r*SQ zLQ~2v%tkwFt#`!>CHF2-S0yZxFCq$79<>%Dd{CD#;s!$6CABE8o2v3!G{tAQNAH)C z1Q^gT?d&CU9V4tQn^nhv;#q3$*P~_OEO%xhzsSy26RvF5ruK51F4JthUm;a*(s4tz ziW=RbbIqbvKb7P0N9ZAUR0(5+e3C#OLAnD~UJ#}vf$47=hg@gL({ehLsf|Ow7P95+ zhK~!0?kT7r#|Q$k-j(5$BXh+H-=xIT@X~BsQ_`eY;0bn&PF%b0=V?PiVwUP%b{}o$ znst#?B)m7?hHRxSFU{=i9}hbW*PNU9k8U5lD{?P!mNN%Rj z#Bve%(%EYO51|-glAS}>M{2I&U*{#Ch5RTd`K%OZmYxkKrkRK<^3qoAYvpRn2?(0) z)icvp3mejBAjWfaBkgIcE9Nfzk4bmI*m-aV?P9 zucMlfbD-SrcS{cAJPBNulUW5H;)L;^r#kBcMJ;7?C!B`k98rn%n(9RGx! zMA9&4T1xpX>TQ5>yhoaGjGUNtqYdKzgvPwN#_>KJrXJ#$gcB z<$H{zmZFaADs^Y*wbrq+XlX{H+yvzt!>Dq%P%l*x4L73g42y4JL|NaAvdxS7o;USd z1|(cQB~ubM{I?MY8K$c!~M$cWZsEed(vTbVn#^ z8gObf)Ra7!)_F~i{peaTb?#BMv%eb@$0#u0?s>~F0^Onq^J2`)cG0^Bt(MOm5N5+% z^)u1qlWUvR*4v8R)v{)|Y)0f(#!Xm(fV`Z$&DZR1#+`@`Sz{uZ*u>d(@90C9Io&|l z(+oA`aY`PJJA6X6^wy2sT#MK8a>XW3)Sply%Rdy(ZIk4E`h0_q-TGS2eC2*9M*R8h zNX$5HbII2-*6BV1Kj-?sRI=IR@dfNz+NoDBDXPv9lbDXiBi&kV+^Gq>rF-eA$~Aqm zIFdWcM)|KzLHni<&gbMf+ciZe2+wf&pmQVb?xZtc+_f{ma3M43!6zMo++Sbt{x;rn zV&L@K?PSW2V(&5U5$-gHo%51Y|N5Lyj1h59Cn_je%g`FYF`ijovUegQNq;uR7dSPQ z(|4SIne$&J)X4!S@CHC7uCoH{!NXdl_3cYSdSIz3v0n<|jN9%982$kd*!L1w( zTPq=hCnO;kQY90*cxgvA@vGU$9>vFF1THWCw&53W|LJL>jC_{MQFlX_vTBcA(^4;x z)T?|ejWPMvz3$`zl3izqTt2cnUv~TXJ!xdE;Nes1SPe8 z={v~vaTy*s%eB1K9tz0}q*2O>QJ6+GqIKgQXU;B`)KWjvPNUZBD23e~uMY5&25$2w zt5wEqxJ^-G=KF=0^WV)DmD5wdBM0acW$?2U#8@3hZ|-xbsxD6B?*2~4_70)? znM{RI@@XTHae;SELPO$eYxZ#enyc-w{FBJ15Krvm>f)ELwuP-n<8P#V z^n$CkP~ZK+jsW&?$NsQ@Xzi}7vgQxix+_*?H_XXM%@qY4Bn^<-1W2pu@H|}`vfm$? zv@LK4eCqHVRZV(l)J5oNl$gbv_p3Y(So(v>-&~_uv(5CJaC`rvL@5T=ESf+GN)9L3 z_gxdu1+N5(dX%0mItYXp_*7)vtbyVhk@QDqz^ep`ImMrN=n z9p+%!f~yj0CfhKTu|!y@XT-;1*)gpaH9d?n^DFKA(k*k%{is>clDA0KsgQHLHg5ZM zQKxNT!6?iAxK4YMWxgP*zMB(~@t`T%#t!0KkFIeefi>-~*Knf0kSRX0G@^f}ji}9F zEKoZdCLQJIvQHiccPy&m!IFgIK1++AXU4S9lG7|`-=@2+7PYy(kK)CF!_1>4z~TDp zGV}Rz@nQoO#7#N*$>O3K+wag!jVXK|C*+NSEelAGGK1I%?_z*jiKzqzr5e{5(geA%;69i*W%`lcgDN{R}Vg|dr9kWmhN zohaqE22#t6jXg7qCk?SA>Kx!edGJz!f#9s~cL}2Zky~E;`SHIJ zO{XXY`!B;&e<>UtQpsdT`~c!n!|Lgc<}s;bpgu34sIWxU1z0A4SfEeFl8Hxi-y-^2 zA%PjZ8?ttVaMjA@G_SV#nooT)NM%50Qks;EY4pe@6(7jKtY!QLV3{_N z42SZmLil%aA%*+-{T~g|QTy9JF!b+y;~4ev0F1?`vlK{Z;>G(yw`IkacEV?NC~HE6 z%6#{&g^3-mzO!94mlkazMpU#H>>4Qrq((ZpP)`dln7wmtgg%B8G-+DhfVZ2+2tZK*c;loTaO@x$_=}$hE%sy7o<>rC-Tg3F+3FW~ul^^JhTh)WKMPd-m!M zD8mSY{QfsxL4u84#p$NVwS0BJiKE_mEL*LTDm!xP8`9UhaP60~&|Ycx-$4k%QXV30?YYbZ?_zEoeu8yzz zG-y(se?3Y;1lJw*rpMiBh0*bDlenk$>C~tC!m1%W=W)eP<2(A|V*ka)c~MdC;*Am3 z^v$^3>rcj)DDKl_gkis`6NIdC2wQg=&;&{<&s^hI>TN32@<3bk>!bENemTSd<{1;dk88 zN|1x4^N^6Esu%6UnuL|6e z9o{$O0D0?{7D9Fv#F~p^2rseDZoAHbOYk>;NgR>Ua?ISC67W^?(rq}@AXV0mGOG-vboSq>NG98C$r^D z8RBral2&?xy(>E6jQW;+ZoX$zf4PSh5nqk>EDhXNEf=pn!v#PA0x1PlzFZYzGC8`P zjU6}NZtq{M*@c>O*qpgK`Q;G|)a{^;%6k1TldeVV*$SsQKeuSeC<|{*5@S3NRrRJfr=b_B-!6o;2Hq&SzX)Kq&GCxu@s)`hg zD|agEHZFYoA=@xB5W{teYj?T#qRBN z^t^2_#^FcFw>{|9=YNv!w@TfUvBlK(*3Id@#Ya*~Xw5lds7MKaT^I2)wY6MrjAb2!kxYCY(qyo5oaj$l8$JtsDl;>?_7(}d(QaA6^*jM^KFEf z#1(WE;&M|EKVJ#_{y`H@#{V(D>Cum;A+0o7*9Jm4f^CTnC!Z|4bhE|XFF(BnUcs4BY%2yJRivg%}Mhq(sLYTO7F#-L3I z2net$p4;e2)h9B5(IO=EqvJfEMw={#Et$hGzD@=n+|-9S(@jrvI;e8jIO)M1<- z+L2uj1a<%1BVC2idGNhN#c~DtNh&9Sz?ZGAP%X{ADix|D0+Ev%K zrfn5RjQKW)41)G*0W3zp7e>Cea*Do=TA5D=QKsiiwWeo13Cp-yM#X#f(Y`-bP>%4( z^pc_!P;ygNBC@tA{!w(n+~);6%&D77F}}avjyFiDWc3{`;oQO=>PW`bLBWsI>eWbo z@%`n7+Lx&WIZCsD&PrhJVhNEsALU1LkTpZ}Z}+VSlIC$m@OB<-o%|?$>Kx~h4^ChQM`XLM2GghcM zuPQ&|yVkrXBCE*o+G(RK{Cwe+S~JH=uRh_`{S%jxH5^on-E&`o3ZF;Ckt^Qm6N!qs z3rRWhde(z&eV~8;N3}#r*?HI4Q;J!+d4k-2Agxe%d5TutK@mXTCQPxI=eI$CH4AN^QLJVP@J!r=f4xK^~0;5rJVyFrsgM(Hr};y5v33V~JcW((9`jnqr3Ob#XH#7w$hxu?Y zY`USqX?nJhP%O3lPUMT4zZtXaHag^v$CH=UnjO?$Y?0L!WBXq{#D9LupZ73*73Gnf z6Roxm*_W`4*y`p}4vbXjNMGl>6Yk#`;`9Y*_^f%vCm4y46devOvM|?S@43iON6fa( zb#x~w)9!`Oz}YGsbMJrsigKeI*^EFe&z>I92{;24O4`t7{wg@XD<)FYr8NJ>XqS4S zy2XjqmBUR~AfdcBouY`VRpI2zHSd3PZiU>g^D0D0$XOXn!JvZGLC^3)@$1sLVqP4` ze~kt+9_CDWt6`IbK2SAdM&mfb!HgiH0w6Ye!@3@Q~LQ=k129WmsJp%t%F zpKS+$V@R_BDtaTHb;f)ybG3&WJzj6G0`}-Pq4^`M;8(9kUB}dQXsfd)D(-#{RNaO& zc1T5AU;R2?>e9X^O0^*K+xdFiALpkwW6S*I#CtMpL&pDBnUi7cFRT9VTIN$2BE0#W z=MgT|xa}j@ocooz7B*()t3aTbXuD?EtG^v)LI5s})Ta9JDx+7aqs?gfdKaNL8??qE z?X*b;sKDBzGB;!_RrdGR%Jc(K*ZZ4Dn}ZD9Saho)iED0A*cS1LrIjcu@83jKHW?&> z?=TVn;f#NKram?zJjw)B!%V?P+{q-a=BO8HmEZ?)54!?EI6h6m*qJ%YsYJ7UY-`B5 z3~$c8B+W}LG&PJ8ZE1J@v>spzPPM)=Lp7|fjZA~$OZFCL3<37kr-Njf+GFvqpe+pO znp&XI?V->qhfjAT@xhw+GfBU@>~E;r>n4=H(r2e+9C~B5j68{*aHHVg+Nm8$ENz^< zfl6x!6hJpGp)A*L!WwS1d6G!#62DrjOKjV387apDtHAiWC^0~^>)!za@U$a#T@WB9 zjWpNwau{vT9lFZU+SjJjNjc}UP4%GT75dnFp|&O3DT4S>MnxS}`wBN*{Kar?>d(FE zE5#?R$61h?QtQFs4^r2W2?VXWvFd2$GRsZ16}32&8rfCM0^IC~OBj^p*^$PX07lki zWd#}!l=UUg%{R&A@<$HgNgj1SbPWL)6vS^zH_SYv5!H^}nlB423hm5fJ!sH2y10K# zprzhS!b+*q@@1-mo_7m<2f2!`hgRmp$!_oF^E%<#0xLQ8_`|x6^BK1K&)T_>U{_czpo=JY zdAbS4?$b`ZNR=TAs3FWyF6z00-h3DbTrGS3HqMA3hmDIZH6NLq+UjO@OsJu(bU4Kg zidw)~-3MvH&&u`}Hyy1f6efRNsfh<)Dk)e>GHco{l-=9@%keHCaJ zn?^rgvfP^ztGb+9i@xe}tdSRCXd0_YIXTLLfuzY>B6auZRaGf>ldTsP;{{)&ZIXPp z3se*lbcXO~1`5=BzZh>QoKs3Dq(lV@fdX8@x+8uF^I{YBQ!h3R?Iyk5B)OwT9-P#y zvdSzOkY9u?Po3#s4DUcuIUQ;obM7h3Qf+#6kXAl zldk>`PkH^@+kD<9ncZ}E1Srvl( zIt;woef4(-<_s|9>)%j^9{I&)b+c495H8rziP;ni{II2ph()CR0)Ca)o?)DB<~r`Q zvbr%jw}Z{7?_il1%ed^ZRRLlK6XMxiAamAwII2I>glycAZ}o)4DZa^|wi{fB^{<_x zzqa;jbC3(-gq$VKYr!39POaOP@rn!mIu`yew$i>gKV^E>i(-)|yIPi+dC0NS6zv`! zvqw_&?15IpP*`uihWD_%0AAGG;CUW{`NbbMDYt>2BBm+1(oC4SCx-Tr1^^Nl zo=-d?&hVq%40a#nPV?W!~)seex=={@9p) zWUq20Oe>P$Bsz+EQ=AX)Je-L5c)*^MZlHP4Bar0AXBuLEB_F)*?|` zs@znV)Y?aqroZ^HpBgx?XP0wj`vIxvDDb*=dHRUj49Mu&Lsqd-d0DNsPQ)q~mjJLm zIPu9+z{XW|FBU4@rDnjd19b;0nPm_& z3~gb$oxP z(LhcN`N((wV^yo-sJ*&C;=g!++tdChxsj2EG60jHOO*H79qVe5@@q^=qkAe(o^gll zevq&H!=BLt`1aDOgIbt6B0k{ahCF%b`v^JiRA;BC^s?XXMS5ZJ2rPwQ=c?`bW~)-F zQTd(bB-f zH_OcW8O9_{{*?=jyDf!j*mUFJtiF3j>AQuy2Wz(t6dctqp0c^j@p61JJ@e=R51^ju ze&r8|)QL!E0Y9!{Ovl{?GECaMwC;&AdEKRbf03U69}Bt8XlWKecO~1 zq%5Xh+iKwciuoCW#YsoMmP{s6u{YFNIw;y5f*k2o(uW84#Z)^+8w z4JVsWrkW#Fx$p2 z^SNYN|H4iZ4c}aPMfw4WiouqgpP@Nqe|xzQ)ef@RXIU|xdS|eE8N(jERea90@qV;J z8YD#G>#E^wB7=W6klfe;x8G>VC)wqBWd+RMK`T_h7@bP$-+{vlJ3IZ?E&y;&_Xlhr z!^E`|=(;^|+;5NI-6XOXBCt?Qinx`6obckdqc_8U|7xSfKp@B~Jv&>-Q_Ocm9hA6V z%sd6n8t;&>hw!eNNTA=2?6xmU>kGNxXTm@;stxbPMh%6l%dLRP70<{&)cTf7Oh?*S zb}PN`(_jEzrT6x5I=muf1DW?d;aP-AKsz(E%;*UV6J1qY#H(A_`LxsjY~)N`Y}LQf z`?E0x#jXEdnUT}nks0XL>TCz(-$ub1p z=p-^q!ho$!P{IP)d@xKqd~YhJAB&KhGHd}Ob9o%nr)=~#wy|w@QL>jR$jYX`TU->t zl{XRfT&$25i$t#oY6DOb{JeM$!3^mBTbIi0uppL6Y6TL>KNH&l!b>SjznwO#?VVn5 zhshNlMjCL{@8mJ0Z>`OTE;WM9Z^J4MvC*zWI)e}b2R}*mVy8?^d#eDvWSpHRD z8$Z7L2$s~tG0=;AdZI*TWWoJjESXBbeVyPDAzS|TQ`bO8JAAD_J-fS^1l`D}$c!B0 z+0Tld#c^IP5zRLa$Aw9i8Uzd_G@*K18U@J(;jfb_KB&K)VR1j7EC@))b1s3GF{h0E zEEm~3>$}?vmgj0PFO||)b64$Wt4WeCGArw9qAJoVOY~Yg_c>*NOx<{ys*}AvMni9P zS9mU#g|m9{5f0@KyzbbvQIg>t{Tv-8SQLu9Ik$2kpg35PEtKO-}#1d^YR=^jsn z0m)lw?p+2MkI3@lz4ryr@tR@x;1dGTjSh^g2IG;(QTCYtm;MV#yJ z03Ccl$;3O6{dY&RM)|!v>S~Z@UdrDa1%e7kL=V< z8b4s0AFi!7B2M~>S?CZD6_5O|aYj?^RYP%3N1djWn zKk!nl-WP(B_QB*f3C18^UyTW;-B<_H?uBAr7ERfRk0kZME!QXmWC63RWc{}gi7rIqQ zr^Sg3+cINiX}`$UlEedR%6qt)1o?0=sz=+!=5H#Wt|b(e?taa;?^{sRoS#y>`A_Xa zxemOp`5b4$NQ*gmcMIP#*z*dohQv{ZO92lT$b#{!Bht+TRp(zFs$1|2zu(K3-3c>{ z0o1-&#dR#$dRJ2J?F~`*TjIiZ8IN;81j{LyR6J-A>%lR-a4f|6JD6L*D6^>y>NKD% z_`E0}+^psx4h(j(Fk`zNN8VO9xih(3q4(EDR zSAWu0$?&VG&a0v<%uF)xN~>i3%KeTuqP(ky(D^$e&k!r*#=(g%sFXGgO~%{;$sz}Q z%pyiJzbL{NRHY;C-Jp&ZH#GrOBaY{aj<$p}X3bl;yC&f5xU|o!Hi~&Tu+YwWa|ldt z=yZ6td0noN;7ZP1IctQ$(DiWF-Uzkt?-lbC=qpy+Lxn!rt%&v2S!iv3jVx>I@zePD zIcess!F3)<#@yY3`N5`c>e$~k|n@->o6K}0w{ zkZYj>@wio*XZ)mPnN8FSO9#o59G{yd6ti7!kPVv3?nsdG^G1i%v)HFqLvhrjWd<-> z_b9w8>z!{|ky*xE_mgpV#0QdOF3)7&39emEEI9+N3;cVAm}0pZ_vU*mD4Z_9qsKK z{;2-vNtoXAi~bF_|Gf2S z-W!DcIhc;%#{k!eXQuW{tEyn5j~Ax5-O*K*}?-CA>U-hPgnT)dYh^MaH59tyq_AXbP_+iwS?T zx%4LwJ#b90j}WVtl@EA$Cmc0;7R8;7#$C?FRPXD^_>psv5&;V1uFUPO?8EholYdHe z&LnXC=(Fmke;iNosu;SYye;h!A~D#qSg}_g<8r68Q(e;d&Z~nZ(TzuQXYp$JYWm34 zUmLV_zR=apy?{9krB_@Pp&JV@7I#}FFI4429a-+6z#}ujY4!IgTeuW&_U0tPXo|$V z1seH>-FBsP(2{u{%9VIP?ixuLLRYVX1bY6Iaq)=ns<9O7oQMCAS7&MCe$5e(25?62 zg=0s7_wo6+mi-0?y`J56FOb*Hty~EUBrScQk+p3%toMQ@o~v|;#o?xP)Q9v9N7|S7 z&=@B_72g8^^ho=P3nyyIRIKJC*}^<3a16Yg0E@_atJ4YD**3JDg~gzOY2|pyHOnhT z1CPdirG5w28kW~z=L-L`2`guTyHq|6HEL5Pd9K`*eDeCK%Nwt=_qdkb=+xLpg#k-X zF!hDe)#C*f&EK_TAg+}0!|OjeXKROij4J{RY_?_+{vAxs9GSb;xf)H-rdnFJN>DK+ zFYtRhP6Jyu&Ei)}o$Om;qe5+*mjwSGchGtjxMf^7zs+5vD{`<$Njf70RDkM`lk1s8 zmIHgoURSgjXlGoxwacsekWG&#M2~!h@Hx~pxaC;WxY-V78X8#)vjr7{F|trwZWRxx zm@t!vTyf)xR3Z-(V_sLwGqWA07rP|9LNdu?2wVTHpjBq+x2Ubg5dBWFsPFLPF!!6E z25joDMHYs&MCC|o$#vJ@WdVl9c#mb4ik&cY2Dg1rlvduz)W9UEK8w5`7tf9bhJj^r zHF`L^fN{o*0&la-&7r7|F<0LQN+BnHm-YcVSe<+_e$f>w-hG{!!&Q;%T%ivuXx_s& zY53e^2czi38Qu47N3b6$moN%Ex zN0?w{+T7q4EQdKeF;{;zWxul(R-ExnUgp9=tDY`#Eer;)?ohOii9kP7b6);HP1emF zVsym&tk%zR`b}OsBMQ;Q`ZXP%<1E#F2j@){R;CZn^Bn`aeCc`OFV# z;wAqtR?QPJpQo@VXKk4&_IZ*oUchwC*kc1^1>c6f{Mg`dP~lh1hVe-OQ4(NRmuwr` z^5W%x!%W-F;sIzP+T=5#} z)$c{2X)g1XW8t-(p5XNwj%VT|6wZuIyFIZbXCgZNB zZ*(4lslx}C%eb8KUGF5A(oVay0_*0SM5R>i=rbwq*~U~+;p_O1;ootJy@ob_yN34e zklVcXIteZRn4D3~kRkwi3IuA@l-`xn8#>N_9=3*>wa9?;f_(@Zl)}~>BDSwR{bvoQ z3%w0>PrHFswyLCNHUfmDWp)l)6Bo-;`#CY^+{^464LzN5e$)Uj)o9^XzjYUp@N(Ql zFPjn*9^}5Db=re`kYRXv-PD6>gfwmTvJgh0SBa(=qx_-{NL8-6 zWz-)^Gh4Yv`wC#uADZmWXWxXQr=2gJE!|ulN#)f#Qo1X49ljhc+K`q_JX2|{zwj`H zC&w5mZ?e#2U0RlYRSU%ZW@Mo)bWTk7y1t{&p5GWE&t}F`5GAmoDB$LQ+>QN{xAl&O zS>Jw7-vrwz6&P8Ub+cTE?}XS+m*yRnqL5qnV9&J~j##GOCALuDq@0)vTGxoF{D!c=CA8tML6t6B-^cOi9U_NK11?NBL9b>1_TgI?7v zM)HGLM{h@g_7+k|7b}WV+QzeTSW&|cJlo4mxAW*f11Ns38 zOVF(XC=<)95fx!AX^Ha}Dx$stB1ZLTgP!BDUS36tx%&Op1`NahR$8_){<>@E*fV4F zT-N(aS`NN&O}=@h)dTS!tlc;@Z``&KkQf-Esh4brJbj3W*X83*ii`|iWhg;Djc>zkXdGa< zQ;!NYo4>1pLR!J;l>gt z?tUaPlH2E6AH?|zVA@f_4HT<+qE+ZNj02$kMz!Q&rZC>UL5DqA`<>T@9NUdpyUU5n z`OS?N{(KfUD({ogIP(1)Q+|l3>CHHohku+4C^RNY6=_+6wME-{bMoQrdN)pIH`cY^ zHiUjZ**s;3@J?MJSC5j(fQmQ*Xup)rF5`ccKLFZVDk}?a;g)h|sG(iht=tS0!=1}K z5S~hSZRitA6G}|nQQ91oCY@D#!M}0hF*y%30PSKe$+QNP7jH;$bk#CAVrb`G>NKE@ zrW24oD>#}*d+ZA&&*Yo3UOnd=An^B@-pB5{9FqJ5+VHX1AMcq^%BqdE9PmOvaxbt^ z%w!*)T%tzL4o=G}-K?ac>ti82pRAaMT>Dt#z9Ii7O!e=WC%T65-6@u$_8Lu@01i$I zobgmotMcON7U}S$j(;c6555lkqx1Mr^ZP4s-%G-{ry~P#=1Q=JVIXa5(zP!V zi0LD9(IA`q0NXap+Uv*1Po2k<$gM;=*B>k0wu=tZjb~aoU}bFMMRydq(@<$54#QjV;w_L}j~9$q?k-O)+n;T7K}3%WyhCV>|0K>=dP6o51* zSE-pB*Za$X^pdEycu{li+Xfl1nubHu{47$T#qx~ht){i9C4Uv4>WHm#%{TE|_d?Bl z&hJa%wdFsq%(FEoZ5y!6eqnU3l>;tH1?~!rx@Ig99^gcn-HRIgiVE#U#-x|xK7z+_ovg^+v}AJJ+!UN#@Ipo^NXKKSuXrxn2VSt-*0#RFY(Z~!DJQ$I@PM>~iA(>vJAW_q;64Y!L$;A*imjB@UWlb5X5 zjl6X~9QFXr)lb*TdJFe*Sgn1Rk5_63ou7_#YGoSNNNtB!K9K_4cpV#E9a0l- zWK8y9-I-SAWUyQn#f8^kMjSuwW71bBZ)Z|?<&8HF$`UAP_GG#UfcI9b?1&IrIDT$? zo{B_nZi=z@vNLRMc^K0(f3ADv3xz8Y+yF-~?#oCI7ha038WS(<(L&l)&pPVHDHH=O zven)~O=+Bb^+6_HRTL3G8lMRm<3bPinDZ@-cLZvZA|yb*#sgQIb4I%82DQg?e$F+c z44-oxx6U*}+E`R|%2D*>>encN4W2JzFu63Q6DKOff4{*@p1<|1T%c{yXtOM^#{Yho z;9@qBA`(FhecUMvAh8J%yU$vaUufoVVc0n_D95)t64^bxh! z+MxeVDn&3`w4R_*&g04}v?k+Em+#IAKkOx*+>&q`sBnep zFO9)5s-(#3M9A-J0VQhS-|ujLQOx`^gsbcyxfJ#9y(bCkTxPlg0Pw7fWTQG~we6Kv zG~Kg+>!8m^vs3q$FRf4z|LZwDfjtB5gNTpehJ2SXFIjzmXVS>s`Cl>W(?x9-d9pI% zXoD#8Ts5Cwxv^E;nBH9}B^=K(vudLaRi-gog7Ga++HA|H{=;_);Qagy#I#+GE5WyQ z;TbaO3SB|)9Wki5s{OfKfiR7YX~Y_xT7+$9L)ikWEIEq;SR?N6bskI4KiSc;zQe=8 zz*72nY|8Vda7r_sHtMQ45s5tf&{f$_{M9&iS&u`ZVo8DnvKT<|o<}*TAS=wadZ4tf ztf7mdhb^cAQV?6bnS!EVFBjnf(}x!ft02(PY1OSc+v#gcwxjP|vg{tG4_c1D;wCsJ zdj(Jxw)=iw>PQ!GN}RBd$?J4v3}RBClY2KTsZg)56k)o>_rfAZ_B>T>_fCiB!dKmL z2N;%hc{EU}`{%mHUiwl~B_zbZ$E+Q&Bek6W=B}$uAe&^Ij41na$ZN`%=k|qEOVJUT z|8*P1;()R7vNbnt!ttWCDF#}BsBooehV1Xf>q zYI9E&2}cInpB@ZQ(H0|ETaE}09`B@0$92DGX@yS|b+Jsbe}311-u1k>e_6gkfX+(_ zOY8qgp`Ic9VbpnQgtGHOo~zT&%|&ldwxfOp*_Rp66EEzc%A-rVf_V*ufbg1`8X%4m z)t|D#+FSg^B+T|T2hEG5UIR0SCWpCn@fouQR0 zOTs#6xFu$8g)b0noDvl#mqd8}K1`_PL+X3o5VC(o?K1>BPokI%ru)$7L0|DA7DOwA zQj*t)IO!IG=^#)}e^HX`6YpBM5gs`VE-W1jbs@Bj3D086#{lDo4-rbdKPXK_qhaQF zHum0@wx-;rU*=@Sw$OOLeS?VhGZov-t+tVAU>o%`r0uP|c3x8|5EQtlQ5UC|xeUr<{RZ%BYzp#R2E^Z9zXngf!IKm9;%d zE)fS^=OC^gWeQOIYf*qL-ChAo^-SA6Wl@=XozW}7WzEedx<#jG-k-%v$l&r{zsf{4 zfa>?vy}b11J{LXFLTg3bC4DFUt|qer*9PUxw!s@twwEf5A3y+TJVz&ITnv}$X2j$~ z`1q|2l6*xx{;|`_N*ovBe*YZ}qqFHziwMI>@q42MU~`#)2}&HyLN9oNp0!$;D1WRT zt}`hc(8dahxbdX2%@%#+pAg7tIRfp3@m}&0C9mP0wJoUK8=4?X8K4sjZnhoIbF6>< zqho#cPip+{C{D8*sgTR-ixzwmVdYG~SesuByH!^kQcHJqrJbH4;U6sGfDE_vOg4M` zVT^W<^)H^+an|?NNq`=t|tC#E! zrDHI_IwMNbhW)(KJ5t>730E($wU{u2z(povT-S? zHru3v5~UL|V`OeS#_YLcHz(N_&IQeA2j&@m%9C@$pz0@$l(X0f8WU$*uI-kutyYME zgcb$N8y|wZx)P1>%SNqLhS~9kB)uu+<@(30@U|7p-E)UVT%PNRG+>GH&9ORM%SeM3 zC0v0+Tc@-sjPi0BaCt|wZkNWzz2Gt3O1KJP~^$~L&u6PP} zdqqCFr1_pO>9#zv4lIYcZhPD~oJyh?hm5!u*vafJOJHLhy54xY4C!oYMPcHa%Zv<4 z6<`2uY&;k}PQjlG^8MZ4B?$lc(G6ohbk|l*^*{=ZX9%q9#dumQ_iTphnw7s`)hx<#C`*sxMqAp3ZkR;2ENEPUv z@E(h>#p`}(%5P`uTlp;NE!j^`!-D7jFz^3BT(uFROTNBW`}%-}(HTNpNCNpQjuTfA z7Rz0|4{;ctMOYA;9Gx+HbzZXT)Bwy0Y9Vg&Ha*n|Ow@RyZ^M)2k&%)m6W1n|ZCItU zp>%3mU8ir)4L8o^Ip=%2&fi21(H50K^f`{=Fu_UWFQt>@Q=&ver;xQW?TL#vJM$_a<=+4M$rOQ7U(uW;DPl1?D>#_=e@fF1+^+qd z$&_u=Eu=JB!;0ums-|5v0`72>nj>j<4Ial%8p;-g(2O`eS@9`bnUS!+A1Dm8Hs)Z8 zC&B^@+h7%(I?p$O5KDGX!AY8x@tbsKCvEGGg0%?-EBk8ivHp{t0$lotj2J%V^+V6Q zujl=fF9zJMWK%g0>*(3@-_1Xca#>3Oe|XOBEx!L#X|BNm$#UP=0$$(M!k)Ou>CoBH z{)?02X52v16Jr6V*)RC8(KB;dR0T=fEoDBJ?DezG^fFXKU(E)KdfQw>g}mr2M1B1=nT%!8w*n@v#d z0g44%-Sn?LAI=q00oJcEVQa)qbi2N|A(?%I*X<{oGt?F-l;12PD?2f@JjLhpe1ibl z#`nh%V{&k$2{)Q{g~9NAUs-pHVT?+Eqq0ikWK5(8Q?W-IGu9~GfZD;HZq*FngInQr zbyOK*SxJmj+`--bNu2b^Dd?_0%FpzdUm`RH-Ps3iHTN@Za5gfr(U*Z|00SxzwtS;~ zh9xHxE>q$^w7f*PZzC9({6kQm)Q;C8 z_nCWy+$a&oUL0N@*g$y6sGW3iN`P-pN#oF1lKte3?m}uo`UcN`IK+W-lcqnsR9@2e zLCRe4KASzCX1!{e7_4R4+n!~=XdESh4@V|ez!wFv9;AQ5s;4pNqnC;F@z3G$ul5_$ znW=XYL34yJ1Af&7=8|Hng6!U7k#oO^+(!D{LYbs~k29G2s@IJ#s*Fbn#di=9y|+DJZr!|JWg_TkTg8&wacGUHr#k@IrYsUD8r^ zTQP#yhDic|Y;6c5@JfjVEFXT37YiW@)q|_kRVm@+%fZG3NnqE6xD9!}xow02r-JRu z?U`F7T&7BsawRD!A1fD% zD4Qm!mcL;K{*;QY{Ds;KR}n|E;Jmd(jE^0$eokdsho|g5%dozhE;WSB#necdOif(J zT1@+HrcB>|?tasKKd22@|#k}bi; z*Rv&$Mo{CPeU@dd_8mzlaqL*-og1ZYqW7cp+?jNu6w#{&UT=468Dsvur|q_hoh9j@ zp+hpc<>PY4^Wz8Rh!7Ayr@f8@lk6M+UHpDZXzxMmi4!Zdbb{nCgEjuO~vL}_kj z7$3aIYk{RD7ou)0I*chw!ehe-q>O~rJ~gj_X$3*+ZzFpHw%;rM=dfo8A9L-I3Y%rvj%rTejv+r7WXwB%#9cfzt}MC^D%hOrw6Y)~R#PE* zn2$r}q#IuT^D!TRi173g^J@7q_C&eJRY=itJm<=QQDl3r^~^MAeJbENtIsF6Sx6FU zX=>bl0}>Wp6(%!Q{Qs`JKSP8sn0<1D$kOJ6YF5^V$I6r)>1^DzTYpDLR4tb3*!KYs zHMbCKsFvB2dgl)$jatp?In5nLL44bgydR zg>&Y-joVei_0!+Z5z|+C){HtYA+bEPn)HiwkDFU`ji`|v-E4Al5^XOm3ttimP!G$L zU#uVSzCj%!GLKTZk+!EyR_O)D=Z=T0sWi#z1*_q%aF+anwp7lLblS=_jBipNU4ZD9n-L{Y(0MtI(;MddB8er(WaPrJTb@~quL@W zEo=fRSQXpEq4X_WQ!^ps70|Wr%b-PbdGVb}$af-z#QVfO{xgfD^^R;IkBqG>2+odV z`94=k9DCExCYCNO+u=F*rV@Xoh^c{041!e#ldtnZT!AkiyT}`GdSZC_Fd<2N=?b>7 z8wvEu`yE0J)6Z|wIBa2D`?(yA9fLuV+!C{-@UhZUvAZ%l`VjbHZ#;gTpE9?hap}aK zFEB1(QdYjQ%So7t9NM=U$6B#jFjUS-gB78c5m`KtVEijqX&kHL-?mOT?EX7rd4{0z z0wFEyK#}d}4nrQ1=;_plJW;DA)H;^y3hytbd}KtM{}c38awAI0nJgc1z3e{T5SOw_ zqN}={^!5)fzS)hFNm9Q{SU!45NWG69=T*TE!#75lpFtsQxtLglU$timRj%M_iZzWUYNeF>1fjw83GjYrorup7(uS|I7o(|?WkJf8e*T1(}2>k_v*q>|p2 zzn+ouA=#IK?MR&xp?ud0dgIJEa)6#;ow(y1M0_i|9bIwWC*HJs0sB_gU9nfsn2owM zV4IX3p2@Xif-^~i*G>j$d$%kH-KgCog;tOr<_5IU3c{jKvFE~yMQ#S!an@awpzZXR zY9d7K5+^FEWWKimfJIF;E(G({!>_BHSl)_M{A84~YThHK)AM$IsmO+Dk#T2(6~0US z@}U@md>(-Iz0l-q3UCr}iN{FUs?+tx=|jk%)8=-L{(kAgsehxK5YfrJ%chq}efWgS zu5LNIxCc@*tDMdR()M?Ap@O5cCpdFg8zORH_rCaxbp96t{|g(n*_b~*)a(D;FZ_k8 zir2o@;3~+Cw!lEW=Ty^%mmJU2D1V0zaVi&L4>f1?Ask*nc~QAtzV3Vu_1%b#=At9M$jWZAubxA4-3<`cYUSL z@bZF($cdlZUupG@-%}2%pIyFT?Aejt z>{ll-J#T1CJ^|-B68YgLtrs;?xCb)r+8y%sEltjs@2PkCD^*Xn!YhdbW+@<4lv<@Z zc8YdnlCk8<2iKcOKAED~`jZ}PRL0R{&4d}UV^_bnF~T*`p7@5L@cGAX$3Ck~6!gB| zFzeLYR_KlV4Hb&50TpNF=;rO;kP>#52Aq>rBMaa@upJEX@h+N~hYHmB7WGGW<`p_D z?}HsWjLS|3!oMBtpU^pPyJ>mW+X_iv@zGAs`>ogZeWp;`MW>>NmT6)~~RdM4N# zjUGrQLDkJu2>_p+=paU(*B+d)hemBV8SFRc&KQ8BYy%jR#G7H{+>QLXfNZHNhVdc$ z-~+j;J6TQuSWRB)IW~Acko}*@NqvNO(jaHorgvm5;rZsQ;%= z`M&`;&x`;x7yMZ*xVcE0nY_+=vX0E`=R^023|>jRb|`Bzd?@#?vD zOY=39sDIqQ)?nqT{h@ze!!|K0Ihr}~eN~av(aNp!=C{s-&EwZhQbh~AU}4ILy5wuB z!kIRge45?&Bati9lp>^`|3i5b4okAeuW=Ok5*r564vb9mq5_;l5>M zVTwQ`P9uvRA{1xkh;C3$hrCExf&u*ciU>=_q0%>9#xl#cnw}pz(6SbJT-PgUGHE5~ z2j$+T|1 zpoUT4n1UHctAKC89q-Z)y`4$btx|!WwZ%^(i3sIqi&bj^BGKGrUS_%=LqsL!!@DJl zq?3D z1W1MH{Z1(iF9!TeD{f+p>X#1ral`K`g&N4EDcwL)|L44I+T;NFTiRty_hH?IBz4_2$L=^xBGsFATa&E_sXMDh1I|K6plyInti+so~K+I2jusLc@X z%f<}vWuupWguY}wzasz(cF@p84ZLBzcwwu=WS;q3M1D*-9{Y~{<%agNLlv%>XV3rk zTb&^Nj7PMVd2dtVHmh$@e3+6}rf;Smnt`lwo=F)@;OfvhSgP@BS&dN@QO1cSys~~> z8)>Z_U+v(C7E=&A8#vE`GqGvL59X}G)KW3Im;T+wRn&{$7^9ZW&{C6T-<>Pkds>1J zCX)fE$Qo8^7K2UC>m!u6I*5W}LFZ0dzz_GY@TIN=SH;I`_F@jXM**&jMJnP-bqyKh zrKwhxs$qQ&{tr&@{F>-%GSRozA7)bsje<*1&A?j-n>l{jq_`q9LWC*4n;Fb_cpV!r z6rD-huWm4N(=5}dMhiaw-LNKNYS-<%eB35qK30rQr;spo2zTDOOCS95v?fa}SSZj1 zkBS9L4@%^?u2`!HJ_rUq;NKZz@N8Me0lkbypT4!`SH}5og{K|@2GeXWKY*791Z332 zf;L}_)sHv(7Mj{SR>vTIDhRaWvXL5L?Oe7l)~MThnkFyGArHMpbxxUK1k>MkI+rLF z%v-+0*=Wn;dpD8N7bs#{wcWe-t_rJYa*Hc(j~pC5n?XeMxvv%7D=k0h&hXm|-pOi3 zmDfEoZLCVWEmQI*d!&Wi-ggJBp5b)$@$6ll<>Bw+&Not9o_=@8A3wl-uZ3z2p#{FC z|DUuu18J(gIjFl;!_=*bYn{_t+H^kZmTOLHoCogSV2%Df`^LBGL%_`+<-`8-=l4I` zuW^Wi4B&k#@Z$n`y2=T65L*(d!zDH9*SPY;B`}uQ!>geF_XmtsQcPtHGc`2AP9

    7;-CVk= zo--mUhsAJQNu~n6zSuLjp}HrOFnwl^;Ih=Y*)3o)NK=HVm0(hF>lv2#9j}H*515~IEG{RLAwTudLEGh8@v&lBxm__?~#*X<6-@@_7 zeMaA33c=_9=(3}GXJ(kuDlcrR$QbvcO28Ou5wS+SWt7j3N>qY*<@=SW!`*bMmBYy! zjWbjb87{rjEMHu7?da#av^iB8CcdW4it+%RS8T|}J2RWj{9HZa%s=YA`KnJxec3i- zTYL~Zj%zj6&&5u-XSC45wrm*e_{i8pzTIl_Za7>2Hn|Ew=A@&nDExh(VOzp8C+BqR zT{Czy+Ei(dF{bh>cQWSWIEeo8#-avdInnto{4lY`pinv%e?-q&&BAs2REf}kwEo6T zL(;4R_5J%YpBtxf)vS^(0cs~uGBEnE_E*9EsprED88Ya{8E8rMvvooUqcGq5xqyeT zc96{h{ZeGDr8CK~BXEbFMGE4^J|HdPc^k$5cI$WSCvO)W&K{wCW1J_e3pz zyDuV@mue;IArXK2b3StMkaCNFqDgMrBM(eXvt! zc&|4Tq~51sI+;k_+Q6iCqzi9x(8Ma_YU1M2QG>%X5H*oHLef|_|MQTT5iSI^WJdB< zV^y@=iMCtm{-WX03&X}1%*Jy~ZvcsV>2{V|zSk}cct;sNB#lykLse&%ZZ$3x`MI(# z9MfDYk~q_sZ(YjNtgR(xZpSMuEt%+7%qklLytDQdFsGIdd&wIfzluO|8327OuE_N} zks818U(! zaPbudWIk-F6is==npexbHvRT3ZS6?XM8jvuP>c7_bOO@~Q+3d_ez&t^5NtpTaq1hX z+xx4y!jQhy#I)0nr%ICcfkCZ z$Qpg>)!&)FT_yL4#`hh6{B7;!83H^G(Dg<#bt~@AGg#iDsL%c#7V!-Ki&jnNQvh!B z+h^d$ke5Dxf`$92&%f5%4*E^L#-Ul{Xui9|V2tFIvZ2I*+7X|OW%@FZ{tDJ@=jo`N z<7?q2$GQk&IstSfGDa39h_{-!=wcZpIANNJ`bo+mE58c(b2{`4@qa-2*ITlDWzOIs z*paARHI-2z&Zw8 VpEf{YzPczJkK3#-m|6rzy3g zKs~nax0)>Vab^oX&sa*0lT;&#uKa2(c8I2O!e@bwnf!O&?zp9g=`-r8!G|0GvPig<#v8=1O6M+(^_haL@KYZJ zden#fWJ;)&ppZEGUc^++US2|Zv69fVrp-<6QF}%b=YH(VI$AjX5GFklu$G4M!tw>1 z{m9d=vSB&G#XHoJ4y*1E7+r4omTf>b^_|L7R$0gQ_ljk{Z6?d*l$)wzjTHL&c9k31 z^~Fl^U{^}6elke2j3u|?Vst4N%mxF!%!OcEaD0N`q2Zpgf9xRqyYxr~Q;BN3o~5d; zWj?2SsyEFt1q8OMLl<|lFNQ*W>4GW3JrCmta^nmuOfve`hc=B3{4k!^ zDstTrBU+s^PnD2-0VVFuR~sNF7BP-_h8zE9U;hUppFRKAw)~w?HdaJ!R#VkWz+cTC|9bJCY7x9V{~18n zs;#SK!zg6n!9YEby`{f+3Mm>4!A!7Hmiko1YtTP>k!f=7fjTj@3kJ}aoNA|)g=70W z>3^{@diJVxf*8^PL=4F#)NY-!aBSxZv61hgXGdVD zvDM7r+6q*^16O`vWPT1|j+bqU9sij9jf|^&?RbXDtKkWAa2!u zVyY>)wxDn+2r+Zs6$l{e{-pADLe0!OUAj3Lcr3w_+<=6D@Q3^U2TA?oW|eG;qP}Y{ z`j@`#IuLWDa+NGf0N%?>2GM8Kz&d&Mk52};hV*>Tj?P_=%~@oiyToY~>sGb)<)OyW zan^HY<~{!71yF%Fg&==F?#oZlwya|ObLtDe5DyuaScD-a>dCQew@(L|G7(%|89K+- zsz#^bz@J}3ud1Z=f57@zVe3)Mu=f(w7x>FtkZM5-JFk>&xz$ix13su!m?`|lZ^t0nIK`=TFEXK~IKMy~2omd?x*t*Wob1wt+bctdc1Px@JZ9fKI}0Y`G(>EXd} zd^kRoa;Kd*3n}qSqy_LNippV>b}VBjV9KGelTWO0<8S~IYD1x=mh;W*;hHO=l6a#| zY*>}xM!-C0NCr}zv8}50!^C*DQ`r=@r9v{2Lnf}c6jO=6>E*et$YF{GySnBNY+yCI_hGFYFO)rJxBIH5G!(^qkdbtZaGd*Q zQ#C9?gPP9aQ(R{o4gr@O*w?cr#w!ZT@a{RC6z@cm1#U;;m6VO7-++0+dQQeO!LIJ~ z-9zQC2}q|2x8spzHDiz{wK1R7%#`g{Z{TwdK95s|J-Cn$Su0HPJ!%!EfKE6Ek#Q6d zH7hPZv!l>XF_5PND!kFSrM7f+uK64gl?7wsgeRf{%(Gl{EPkKHmqNFn^>Re*GId?z zuG*yvv!*(f$IRZAe=Z!JqaPC*AJ>b`Flo71+>A#9Cp`%vIj;yBk7~pdkzAs?cNW5!E`Wd!9K{L$*J#>?}cxb?3ZxE&@L+fjh~31)+qooD&=d&Opu zsk(NQq1{}034BC{ZVb}Z!6+1pgbqVQ{I4Zo==xUDPBLHh#J&Z_HWSix@o`k{Q$ut=!5r%W_fDR zKnC7^8rwvPAJR_gLm|qm(muD>+yE$qbB^jBIA1eSavs7pCgK7chv|f84Ozqq!OStw zzV4iG>VlCrmBG;6@|O>;Td-lUoU3t*)|--Ycu`8ZtVlqY#QOu|)T6?-qx^{?wDGVk zTOt@+S>bs8B}z;Dui-5sxZu8DgjS4RD*j?7^Wl4N>fL}(b3f-LloMe8v1jg`yEBL) z4@jq+^K;8NK6vX3tp0}*FO;9o172CYr(<)iXTzN=_U3X4hHSUSCQc(H0w_UDmyHsK zW46I+f&~LoE_V8cdmcy0_eXoF&AxQvF(sX;GFB$6w2*m#fR64?u_Vs25Sn&w-J{=9=L%7PJCe*Is2<31Y@k?x-jX+#jq_qj|YP`w43r?0T}} z*BNY?IPECru+*H;k5TXY%%Kg>S=8sRTb+WqFey^x_C21L3lXABGR zr%%?7Hi^j$S#&7(a_hpRG!5~}N8eqmLWmmZeK*_IzDI=UY!z-m- zfo&o2yx2w9G7$Wq6|f&g0|h%vQAhs@r=RPqGR`^m;Kl>0nP74L&;yVkb+Fa zxq+N!I%T8fb)b+66WPdADv-9X(rpFDDzPhD?J2cRYeCEr_JCvbAU`ro*|fUe@*!mc z)ga;HS{M)Z7qn$G$qA!$8+ZQ`bmwuEqMs$FG3G#^LJSw9VJO7QmRB^^x0;oFWRa-O zYks={^Q$x;;Oq6FXx3WBEIVw$u}_)ZUzYAoxz|A0BeFwV z)@C6^uU0zfR$iC|eMGxUjW=ZO)ke<2lubQD(mWB>CTjng!O|}4pUv~vg7Mh^qol+X#WK~U9?%#^?$4{-c-z?-sh(w}SIOfP8 zBO!K1OjHkr2Q6J@#i){dkZ`{%0z&#v@q{=Il9;0v_sj|>BY4)Hpo1K9-l$0qmN(>R^C&k@)-ZDZ{?XOZPjpz8jh`$6N{@<0 zPVmY=NiuG72zay*tO%yW?N1v^i$kKS1Ms0;E=ZZmFV**&Okj|WXP}n)oj9!G;~VL> zXh`L<%}Nxc0HO93sKk)cR$0Zkkz?U^5Ku^)IA7OXMcZzBF$2NW+`5gL(@YKTvXoxZ zX{$1AP$1{7NzXUoIL#z5ZQ)k;`ayHXiCl*K9zSo{1#{ZBi<<&Lw9Sq+UD)4B%EnKs z#JJwF2-1VSMPoaT@7%gKa+t9RK02cq(VwVT*9@+DD(Xv4 zu}AY0t$vZ3Qm77Sn0Yg<1m*seve)|B%?ITRj)Bo;-c7X+p9)$@VL&r=50 zcZm=!MPdgGZ1MyqQCfl&pMVA3bjCnX;Lc~3{E88fo7V>j9l|ROQUG;hE0T$fj?fm6 zd6EJ$lL{iX}8(i~OIQ!-el1@2-9x+^cETF78 zUw-<$09VxIBwu@i!RXr2{3hJSTlfULmK==q>59qQ!60fVeTD2S>auC>$!Knn&RIZY zay=fb+-}zDI6pM)sBJQ590l~R=A2T`mrFy@x(A#Iv-#!a? zl=}FKM)0Y8`gjC*VigAmo#htx;c#hKg>pLOI z@9Ner^J<|cd6^u{y|rckSnTCKS>I4#pFC=WtXXJ3Cr3aSy`Z-#=c`M5w?yI0f<_Qm zj8?wH1@MaivpM!xmU^}EY~pSa#pp`i-k5+OQr zea)(K!UPnp)_R4cq1`9maUDR=j1K8@-&921dg}OgqPJjsdB1i^H8R4D8!QLSA8^Oc z-f!*J^%<4#seUK;ahF}1M@*rjXQZj?*f74#5cOcYWjr?`6Oj9^;o#2Jgd$EVp@w-H zRzM_ztN(-S$9D#tZV;Q|VcN|ke^Tk{?6`OvRLCsATMd?(Be%~%y z7CQT++l9^&k@GOFv?xBpu2!8MGL|3D7}$iPEZmictd4RZKI>Jdj6TOZk58@x7)9;maJ&;+Zr6 zj{II38CBF_ULB(VsNrSpF)J1SICc<9BH{JVMqTbulYcA__!{}w6_63%vopG2SMHWX z&GAyK-IZPe)$wRm&kW-?& zs9cmarP#sIq&?wLjW>US7JYx@3W-k!&cX<(YlXm5TkkoJOi1F4=mVG@<6j%6xAndRnj ziglHDR~KI@zPW9kzayen&V_oEJ9%OBNGRLW&=`M9mCYHNiq_7FgbWW;+?I{TsC8m0 zc9d|ejpCitsJ?4q_oPH!77fMz{Ro0>6>RV`&COgxM!x^3w{<(K?!z@OMx9F?x@mm2 z&^LmlJ9(I&W(H+@#ifi6OI<5G$iZh%wC!?-2It*&r!Uq{W*bimd!@~=yE#%?R8*%f zQ_67#ryA%}$1dx08&vc|&9vKXx|@qSX_jLkcU_tWY~2dBm7iXe=he1()M>$3gbmP# zI}3I-@cPe`Z|WS|nQ+bI+-k%jsksBYhG%qEJG&{kv%@#w`$S@TJ~V|mAcNBXPPX%D_N&>!SC zbVeu|HIt~0cDK=TWboN2XP;C0*3-4Uf7&w=ERt7vO^*DGvEHaO)wU6NhWPPsFT+K| z$dU_}!!Y~2dQt)EE+S5U1-B>e8mx#KUW^WG<9VaYu+B%3Tz9UXr&4UI=YJL@%d>!K zeiWBuevIi+ziTeOJ2B9w%_n73>thk)>pJwdd+^J)AN(32g9NO@wq*}59O6V+2Xg^_ z9QscygjK>HsCFiFX0A=YJ^!;n|80|4(L`hewI`fH(-)YKF~VS`w^g9~ewU<@3RMHP z-}wDEHo~f9X2ZUu3OZ~jcD5Oha!t5;?R2To2XWM594q$-u)=77FGLMOf|k)I%_5+5&tw$4m5ww%ay&tPQA4<8lNZzg$pL?tO=r zc&1<(4GB!5a&3pfK*QEIPqY!${J{p|RZVYW%$i8QC@^*CX*HTJP#5pCGWE$D%No`| zVZLW3;uz$?m6&a~^yL7io?Y?sy=&HqbUFc5@E0kJgEYGwg-`nC*9~Qkmp?PdTx|C} zU-2drtJTR7d!ugzEaIl0R8u!MxhQoQE!zMinqy{_Y+7axMcp*Cms?FmJf1xYOVIk} zE;G+K2O%#jW1n<2ywH=r78PW$4~s|l?nZ0(H-*(!9pBDBjLhWNZux3@UUZCr2EQ=-=c{kmAB@hNvl&0K z{TrM9i&p>fPq^IYy}O*dX%qQ9r9J->I=$Z&MT=%+_^IEP^En5l?16L z7$F%#QmBe0K(+d?`1*goAI|bE2S{V%BrUBl4p0M6Jubs;AvZ(wxB8w1zUduMImd0# z>dk_>wadZXOV&DJ5iX{bnFUKVcd5m?#rA|x6jg3sDyM7HJ0tmZwPUKV{Z89_KKcHF zyfpuChQHRKh)&K*T&^dqZ43yp~|8zO3SA?NX(bfvhVm7w!GXz2#NI0 z3uM{%-A>1zkI3HqF05y1m4;F}d)^%|*p5ATUPGYF-024mF~xiRYDmo~k`4EF*I-Jl zF0lt$_L^N9)VKFZsVW87s+mVOQR5cP2g3^CmS|+(o!ebY^YxkMDP9#$B38 zq_eH}l4*tG*|WDyC&vwr3z={v4yIOI_B2M(Z*DI~yzKLKg*I)hOR>`K@3X#7gw77X1&F9q^+N~dGtCF*u zhJ~!1F~qGnf0@~;66$s_gf6C-Y8Ca8OSJ10sEP0ej_4DjE#VGo-}w6ZEUyeq?AM5O8owM{#~UtnBZ# zR_%IMBm54d@w8&c*?QCJD$cs8?Pp(HG!Fk`GplnJ*DHL|tBZ>whH>Kb<935<7oou= z-6K!MVYL`vwrQ5AaVPTKhqV$}d1qTMrD!gqf$fOG9|CrmePskrqXH9?{-NExb&58y%?xMfjzvhv%C*pL+ps_QMa@@98Dn_TT(QBk< zm4wc4o9bw3dj1>|DSub)A{PxSX5oVp*5K{+CH&A$b5hYN*JF<6SpB{;IcsJmFy`qrww7IxT zl>Jiwj3UkFNU)sObM<~D-ptdMajSpD==?T;E0F<}`6|Qu@l0-On3ZlE$2(HT!%ci; zc(r?Y;E<9q2b(!dx+U@acZSSxsNR2EsQ7dNAII*u1`pD51VN2SPiN_^wa(pMlr+va zCkbMVL<1iUYmi&~mDY}6c|L!h5|H;+5*dTI{w%d{+3%2N^*S$s<5rHnFMqDo=)pJ* zSEx>O&;OB5?=k8n%wvM(>&HrIeayi5#zAh&M=91iOV+jRmQq(SuRqWGx3#-yiO`?$ z(gagy^UREMuD|ehC#*f8RD;wSxnc2zqU#dFvvS*1|6Jvhk;kl?aO%ZnD)o(N zy0F!aVWY1SHg;VKdw)bcLxwmV7ZICWqOBWSXbVthREC)ud^FLrASiWUX+nZ}Nrg$j zJVL^Di9SK13WHS{7t&GSD5NTPU|Z$-lGtVgmaJ{1J!CmTLJy^6C>90QnRUPWkT<8D ziDxXdwTXmxT#KezRk*@@=6^h@w>uI2Kfc~Ns?BxV9&VAMZE@OCXpusJ;Dus=7B9tJ zgBEvpiWRrw7K*z=u;6Y@ad#(BG{ND^`Oe)vcbvWN@4ttP5%P|VmATfOYp(UcoO%)< zNV_qKo?q_sJZF#b@*m+KAd*ua;`WaVpULcn!x*=T(A|oa=a*-Ak|+l|yk;4JW6b>0 zQFH<{YUMu3l9FV@6afx+&tQH3+C=Tne7>G+F|ofc6kYQ2#iPSX3(c(I$!^gZ+lk?6 zAMAW|;$6j|x(I~_c#E=Gc7Ut%)iF7j{KXUzFcznGZW*T#L&EkaI*P<@005O-0Gij? zc?F--$mM5E5fK590X3?mxT7=c5pdA z0m?I`0%^Pkus0hZ9&09gL7K}`yyUi@LYDM6jm)dC{T+y)kbfB(JI#2k}mm6ytRg zL*4Gz#N~|(E6#<~Wy8tebuSzJKbXWT>`J2|E@7YHD*^WD*q+j(;igyRWP9u~Xj^(G z-!##z{^XW@W;9G{ec+93jZ-~A?b4kuWkE{4JOQ3=&bPAdz~JyUjcjtCTTT`qhad3*re;olfrGbISw278(V?Sp-SZRF*@TUNYoNe_DBME=O11 zFyZp9)~cjxlOY>>@E{|nOfAp$eovBoXD}1+YxCX~Tz*qEbKj3v=4H^Zv|%MuY%KJ# z$H>G4b$M0X&&DyHhxj5Q4&#@^+T%&vZW! z#)4~^3mk%n z0*iY_0SlrKhY6n-Q`6QwR>$VYUJj8s5U<_~Nbd~2QmOh#hLOT*TGJ1I5St#zKt$~} zLyS$5sszS8?ie)QNJbpvl5KO@XpZYVl#PZcJKDyNHWGzX91I?NzQUU3PTF>GWTCe7 zkb$ykf)}l8VtE1pGe5}VQOhUZtXG848$YI~M>woeodeK1XcQ!477%nr{_fC;@qzOW zgYy)y13M)$el6tIE)t%D>EerxGLl-kqYKPA4)_x7C(@io}-`OncYQ#JG5am!q}S_EU7 z`ZWq;UY+<1ISJ9EfW?c9g|3r_#d zc4zc3E`sTl3jxTI*hDIw9O6UEM@v^~uMTh}FAvva$8VSDEaAEx=|%Y5<+En)A087= z0DMg%S#Q)_9qMa4vSX6&h3kj9{OTb%tNRDczTNvAwWaP=*0n89_Yz{!UXi_59lDEf z67U+1JK`EE0Wgc8i&Z_Rsn5d!=K<4_KxqnHbd?KUP0aaA1Rza@eCaW- z%0;hoPcUsNSu9zmcp=$K+F03_uSzeZ!v9R1tNns4)wb=Hwl-^<$&7ey2nC)_b^ROXu@L&X-UPfQ~ky zp~cV|R&|7lve&ss9s*zMW`b$)1T1F+a zABMm67b5#Tk8x1_1aKD79OG9@U4P;Z0CEkwHOjst>a6m(Z?L6E`qEli2AGbI>Wi8S zTbHmfEe^xPyYP{YPV-HwWJ&#zR3!A-YTs34$CP5)8U{{{j!}i;J!vj?;J}%#kw{Fj z^o*2G8uN>p+~(h#?s5_ww$JR#St}CUeCkU6h$Eq>Z<2Y{TM0(Itw9ObImek#TtGNj z;f3$)(oI)xSAp3EtN6)^FF6l_FeZ4HmU)}Shn|*UPD_DHPyifVkHB;j7n#2} z2b%Y9U2vc0;T_e<+0$1!DE=w=*t&{Glbo!^A=&#nH=>ga!$t#z#g;ZRTR+zR{4nfs zQ0sgpZi>oY=AvYrv1|f}ji7l-ya{65tQ3`JVmERR%;XfsW{;> zi~wp9!oZZ_$D@3@k4tA(cG|1?pv-G zx6Rn9wfS>#*P#$BPIo{(PLE|yM0iaC_<^kZY4EU!ftstS z*rgY6`4e@?UZWgKQqY8X_*>+dWcm1-C`!ulzV1G!%5U`<<1s!hH6yl2hEAa}%xn=^ zxEjN6IBB(w;P`b*^2jw6g?$E`WNc})C1`iMD=JS?O*xIXVvo83rl#&Y>&gTWu{JUp z+kePC{LOlJti?_O(Vtr7W-O}-efDjVYahz=%mN3Rv#hQ2?SN|4v2K&!d$+~ZeOfp< zSn=>?vxHRZa)F=nB9+k+oPm7TPgRD6^NDOg`|46(WNNBLl#&>$s329nh0Ti`f90nB z{>*;Ob9?sQ&d+JJr53jIQJGrAE``U6dEUkoa_kw(IMe$@S<5Ir=X0(E3;Ddh+3BVA z3(u6D{@Yb7 zcKECCH`WfA>yyi_vh$>fBmLzQ!}+$iGBNCH!ec(&R=pcHcPb5i(2z(DQ2_@r03E0N zP&u6_UjeQr;|5}i=gEZ!Y0}z8;BB4b3QrMCZBoN#>gHz^vRfjiv}wWn(KJufwU@)S zk~lZ%e#s%1Hv{1xLF8|8rKgzwSrl5v<)|HpC1*zNZpF*x9Ib2(f@e?sVk&Bbm9b0z z2;&R97;c8^eidL`AwkU;@uKidt=-c7!};$crYL37oP*HdOvW#)p7gCiEDk-=TCemM z(~TEuSe<^VxuV<4|6LI zUQcN{fUoU2Kl5IZUOIZo57Cf*T*;9&z1@!riKd8?X|l>4k@qQVgNeZUo?|2f!{Pj3 zb;c_V=r}7om-#(o-&Slwk3PU|W#As#ui-tO)Aj5{p~n&&TLpjDBU|j=kS%s3tTBg! zEzQml-pF|zEV^4xSmM_B8e=WLU4;4WEcBq^^2Yp}fh@knI|f+2XRzM5#5+k_1fHU2 znjAm2pYYVJrw;n8^s6}2)m=DVH9?ywIT1a`7qC6FvDy>;mCUB$nOn! zH%(*(Bj!4o<+110W)PMW3gGTcWo4YH~fY4{X2&k?A|9fdrP5#(09c!5>SRoN|kE&+5gx zEu>FdYlTaXyEUZGN9P>pAR0ku-^y9vPR#6h3$Kb(?@>g8&+nTy?+zGAu>K(Ge;-h# zhF5nf%=bd+2)J{W_S$4I;Bz2Nuvq(v^HEG`7ye&p-u_tjD+4CgU7}{wHo5$`wXl1H z$rAr+I=-<5;64O;;|h1$buAx9Mc=Qhacf(0L)LX<5NDZ}PXunN35@YqN@Fli(+D~!-~@Y*MX(cWR8r2@ zZ910gpT-x2k2+B^d%8j}k^N=@Y<*9)ml}>o#9W7sR=+H)(DH;EY9k}xx=++j3q!={ z4boG4H^tRt$z6p}&w7-z#ReW6-?%kWi%uH@9cwrphNZuMKkp*iFVaXI95~j}LXJ=1 z9AiXTFW}pK1HcaKFdU{N43{CZ9l-*vjT;?<3!sC|_y#cXl&JcFhrrI_PWwrlU2yhv zjo4OIi|IuDizU4Hca~+#>{e8i*8a+x!42A~uZ!naJnwF7byagj4h>Rx3FyLpTdgyX zi_hk(*uoRF-=tU}XvqvtYXo(raidEUyA^S_YmY(BrFoU==&iPE7-TholQ*y+B<$db^{P(?-bSUL+SIG!ub z6a%zc;;{kMiB)F@yP}^hFqh$jvIm)O<*&LX6*ympz%L@*kydY|5QYb7XFXmDI7(O78OeW)!ZM&4j}B3n&QCwd=cha7m|M%d{$-+Rlru}G~L6=u*Mb-(=$ zJwN*ar4F7J@ObMRm+g|07W>=^4^T;KxlnUnbYmIMaZ z<1%d@%SRYcC3Fqev*0R1>t2Jsx3&WZ@@08j5h6o)4UExL449>3TqQoBlc{wqAcVts z^|CISHEdfoL(Xe?^+)nIf~_lz*=eF1Deku3%jDfpw9D82QbX(cDf_KYx1N3p1y)}H zY(|x1HrriaRvIslnrRDDI6Aeb~W1U2aFKpT38SOf(%0kj`CI-v>tLA5BzW z<$w7d;;0craUA6I`!cftscNehd`+4!eZd8mL@}}{=24ufePI}GjL1rW!pq#Nvj{_NOpokV?E}with*gi9*)pu&sRuj16d+h0N>vbS z(pd6U!?#$^ukLukMY6x?Qs@3$=c=W2ZrR*EQ%mi&&%P~)Ar#L4iCUVW%DT@X?a29y zcqWPCZ1wz_wFrf3M3*-=*QjUPP*u0_dx6O`HQ6{KBUlthkndOy+x#A?zvEh)5lwu* zzy`t>oOo0k^fn|Y$?-5X>#QMzIEjl5TIFD72j=i1dqDKUPxLjt6zVr`ecEKKJ~)ON+9d z$TP7vtq5`^Y~l|{z|UW{B1C)hO})9T-QTc%Op-;8#iFJlO}CTr5|QheWHJWuP6*@N zGlop`y;CvSo%yGefOQUBV_QKXY!}zFn zE+(5BAa)hv>zLl4cj!Ka|3?e*0ftNX2g%_pE|Fm-GVXVnA)2x(1g^9MN7p7-5@tAi*njmLE8!S!{`(yj z8Bc`Z@-{dh;s4UxCa=b|_wSn=Br~P@nf$;oCySFyas5BjGI#1Q{ zKbGzxcP!}6QY~On-6naq|MsDHUTd;?wOq@bt zXl#suG1+yqu5vElFK1#x8>8O?sAM!F)@fa;LR(KIwL!{XG_R(nmUFRyNO04oRhwPt z`bFz%<)u#BFV(H-p>-qG8&m$aGsSoz_7H7Z-xIdt%DDWx4|Ts&n{d)O0uFoeqRU@D&eb!;{<4Fc z|Ju*dY3G&f(A#Y_5cxEhD0j?D!p(}tKA*Z<;tEzm&*se@$fLA9h;W#PUWmwvF z!D4wYJ))|H_}nzY;#p1qx#s!+Z)_=+8fbmezT5aG{af{l6voVmxWn44+aP$ZAng~Q zin3?-*3<&DK3q=NnwCRlBC@{9Wx|k86tJ_l>DfUM!T%f2$G;65z_m_>4>?9zXTp#M*;@>govM@IuhFo6fg zw={i^13?ETUI{0`&4w&-=dXd$J~}Cf=hC^j+IXGHDYCU4kQXBT(r>;zf`UHubA&(a z@+bK<69_6+myG)mc`y`uskBd^mJj!G;gN-e-jsV8xaU3-t0(N~dWMJYZ%Gqj2b8~S zXL;w4mTAV%AXvTUC*EwK5ae$GZSuWt6CiYRKrM#$o_AIm@ApQ&W5g8Zl(Rt>prQ#L zH1vICO5sl3;Eb8}S^`;*DBy{hRXyHo`gHUIyTTcL3_{gL-Zuc%#a?6Huf+jva+ler zRmf_1T8mJr>f7wW{B)_x@q6+gXA5dZcR#tsE%ta5;gC7xF;Hf`5Vsu+v33pZ^WETU z5?$E7O0GM4Ka`O@!@Qlh+_J>CG?{@!NcSG4vxuu1AXr{HMwf(^=TBHM-5HHrMH`W( z6(P|Gklkfm^c=JM^YmANrS!2{9x{xTx(X(oqR^FG<}-lFFJ$6xX*O8O9uw@+N_I`m zC;wQ3dYQ{>e`YBZ!g_(HS#di{!{TcbY4O=a4YD0Af7>~B1l-|XQiIP=c+Cs=?>j9-7%(NC$%b2RMtSrQbZHYxz zWNPsHI;PxsOtg76PlN290z@JKF+?K>%*A5f3&l_%M+HGg)WrU@za zDx-0YobSy)3VFU3JDEJnRozMoGG%;Fha<2zugh&TZYYO)_=vRnH}GDfZlr+ z`lfKSo~Isy=HQ+Ocuz|etY>v_bkFi=rLoz}^9jV0JHlP6Vw}YQM7?X#`!Xfv^vxbG zebuvrBM7>RsY_Xh+v(httSe{2D)P<-XC37P^gD877oLYO4)=SM&I~fXw8zsNC!^M= z8QZh{{Pf(jbi;odys#hNW`1rVN}U-09qlYz${W z5!f%GsA`38Jz~uK{_S(-rdV=UQHqf4S@h$HwBEpA?O6ZJgqoK3-#%ju^5`1jv3(uI z-sL&D?Lf-0!GTf&lT`|Har665ww_<03u=j$D)a1ZebZzk$DC}C|JE-42yvCmIPoBU zB36@IOllKv;qKDwTdiy#7PJ&HUD)%9v%-gIIrE4LHHnOvnL?{xIPh~@vX0=~tNkRumA$$P8mF{VcD z!JzVh1VwJ2q-)uzjU0IH^-aZk&Wl zQ1+vT%GvQ-i$S(^2dpMT`MYwFrxY-o@hb=Q{z9<)*5caZv_h5%C1; zkxQ!hVx=3@Cn~2~+gCvu7M^_8uSrz6%&!f$zc`Q;O)$0Gs)Lp5Wk-Q|!J6FQ^G0F1 z@1sE$DE(lYjMMH#TE9pb+mJdS`^_lkdp!bhV$T#^wl(kEfupe9q6;nEODaw87PYx< zy6=7@wGuYb>uS=GxV8b7j|VAn+j=b;%UdbLtqSHQ&JXY83qpc~{EmPLbmb-ZtV^7c z*7NV4dO|ji*Twk~HmA6Z;6Y_n)NmidDu;FBdznbFSFf@|3Qp=bN=4KbhYVhw?Wn@J zvLp*Z`sh^}+7o33a6ZTaJ(q3TK4En#xJ^gGt}V2Oj!$M$SZ?TP1v2?PWub@~qXk#q z__|lncMl;{gCuy+XsNm|T9hQ#YP`BZ(s>dVV6LFwFHFBgtqGNXI`6`?K`C3IVIt_I z^@-^lcE*q*uyJowf-h=}EoX*OVvaq_T;qA3XPDEGWx>IKNN;}8^#S;52Y%~@JY@5P00CO_`#66g3k)>(aB;EgYpDAOPPJwE>Dm5i_T z$q1~V1+eN`34D%0ok=dyL%INN4&$JR+DJIH4^iGxPxd;Wd(fs6O0?L)KWvjpOfm)B;8V zVp1iRD_9$X$oFF>HycVoJo3n?p(@I0V0Zl(R}#tR9mrlMI%BZo{Q+)5V8}R0?ptag z{|xMTih19@n_}hXBQpAUdmLR&V{-Z@D=zXaHW6D(+(sePl4FBd#{M&tKwDdh30ir-iXHH+6l8a~S_ z=~gdu6~EPn)zA0w7&t3Db$!31qKd1=Pk2Dds2|B3SdU{o>!lS!t}8OtIr`EXx1*5? z^^-d*{8{zs!8G4)W0#Qv2#{-vC$|KVIdDt+Dom%nF&^m=S8N?UP64xx;~yLL1>9xh z$1)QoW&vg3>*^XjLbjeilTXOH&j*+@wHba#&jyGP4$>4%-M>Th+AAl@=@kB|iyQNm zY-7V^a8S`EU5|7VQwvJ^Se{0u^!#*0(8JSDnvG>_jXx^w8S7t40jlo3e|l5&!nQ($sfHg8NDEWCTI;_|D)9x z%jfoj^oJXVL;(&HY=zKot>gI9(K3xQ}X(s2IBr*2BEuwL%C6qh!v z=6(4vSD9N~#WKm__G!quEiuBYl8ww{b;Uz(>l6gx&7y9YBS7Bu8Rfy2>Ax8U6b4?D z#opUk@Zh7Dvk_iel3?tU*i?~A5_tQ!%83_y( zk)cV>$f;DY)Vb^eeog4S|EHOI*Pm{u^hzPL?BUe$6`q zPSc`Xj0mfS(Zp1~#(W%_u4&hm%}aKCvV6wl7f5oEE7^rfQhv2VC5@;pA2>AtcUp;1 zvv#TXzmRNrGvA)hrM#=Z$`&;=a>gvu8Q9(mn)vSC`oxBid)iODl!q-3KxQ^T&HJU6 zfZkZ&Qt3mPcH%NEH{@|~o{J%J*z%~0&lV1%5=$g3Wi?MH;2w z8}<8zU-KVns3Ry;j}e&%qg= z#(|ylHn{{j76!;oN4flFxT*Z0uE+j0n8PDlHg^Q3`jeL#TI~{8+Bu%hnA<0Q(=%1P ze}nqQrzfEndgoN110Ti_sMb+HY>Tz~Xqm%df?e6D^{m32FNxX|hrRMN8-`(%9xP2i zM!4?=9oB8E!{2q+F)Zs^9hx=if{u zq>AKtb4Kbj_U3j% zg|+YaR-s~c`Kk3K!PCD~fIs>DuS2t}{{EL7l!ZBZtVY>$JRHVr+gUU4vy4^loi& zzVq;mr9Y&|DV>4I8NcPW05=d905<;aif}t7(pQuaam%-|(KfLU{xJ%yo(f9lvcQJc z@o{oPQbl4+Gh+^jPWgsinogkR4w=!*<}xEW9kmf zrCGFp&j-eKxF6I07(&->njzZ$qzyUvu@(*`xK}0DCcdbtxiySK_2{e2N;qx78;8t- zcGU_m`e?qSY49g}#sm{2Y%DbkiKxvZ^x3+huTz<=gct;U=4>Rd59oe#*Qk;eatFO1 zUc^`WU;|K|=wp}>b9hWmwMpQ4V;2weS0sxxWw^e!E%q27Psj4F8pL!wt~#p{VGcF8F15GJwcTFT)x`p!7{1edtgyZz)fl;u94wq!(KXMgK5a4wtYO1ol-bPu z&X-c*Not_peyidX8lHAd<_;mA( zz?x!rFQo2Tef&}9fd9V1BgTlA-6qMd3gge%@4^)4E>6fM6Gd95@%eyKEt60{ckFW2 ziEG7t(x8aGY_~}3EeX+q!JN8A#S}2*_?0N0q+`dk4FPb`W{haWVjOFy-+_-2v06nZ z^+<;IHAcogqQo!hqq6D@9zWX^?{R8kjIxlWtofI8ul!Scl2tp4_HBx!K17^|+HPdR zIu_xFsVU3uSIX>*)A>#wIw%to<#?5~B7w42Chw&UeLX}l<ALFhBP0Yv#Ahqa?D4>V3IzQRXo9l_dk7<^HP5Hx{~G>ix-Ezv&$3O< z>I|KdzO)~%c^?V)trCNUkt0vr+{O3t4lABkjBX{>7eOq*r9=fU{}DwAq@J!&*H-_! zt~YyGp=XdwA@Xv*a85N7AU)%eC6flT!sFN_D+5-4V!KIVHZT#`hHJ0J_2fzQ)3@qZ zWI#hmg9n{!_S8r(@HzQ}WF8bHj++~MGU!&yqb>k#Be_%-_XO3Yzs-Klow)#GMEj6t z=DH7RmMmwl0s|Oij%dcY1MPlpguWxt!u?1TEH=v85btuv;+gOmuX=fRS3{%uZX%zG zayR+9l(AB3ilmh0owA{_j+mIi?dT-TGpt^Ji43o=3VUp7iHhPIdK|U}{W^#0K_912 z%)yUy(j-dy;W3oIun|Aw0?c<}X#HC*;N5I;5^zU4x-?z?S|?~r7-B5boD{3{_REXt zUXa17wuYZd7;tT0LD=o2Cd)jZXkAG#o;enuX@t!< zmf#DWOT#+x5A1ceT&=N?{Y()hMnzWwev#27LMJahAr+c}yx$hj$G)}J0e#P$?4uA> z^ihwwqe-3xb8KLW@CCL=+NU;%fYP@0AjeZ!_)%un~l@%$E<1zuUG)emoW|8$Q4D{=(cxV%zL$?c0@X z_4=Q5_kV|>OIrIzN#D*{^pc@Yp!!r$PbsR>85FhpsINy5axZ(*S}4Ty$7EwyL;S2I zN9u;}#JWbMRm{@sjC6# z6viA5S+$xG!0uW0t4Zl&0U5f3TM%h9LIZC!qpI*ADpP1QP8^u*4iD7$kS#!Ub*ApD z7#uHopSW2B*7>3{QO2JbwJ=8+DB33|`B4y@RFT}$Q2->*`ZXeX5bWVIef-**dp5!O zvG#*3)MSI&t66LxVUAhFb)~2v;R&ZJpPA-(@RAE4@mFQW+^Jhd&tpK{nP#^rNPe?H$DDqz)!nB9JnyN9sNS~oaJ8t{5l!Ke2@rS3Av-vG_04Oht zi&nabWoIn$eAU`2sod+nAO&kq*nN54y3#RYl!-HeNtlnrKRMn6xVbQi&Zb1zzk!n< zl71Na53c_8>2wdtm>v@q;~}DA>L8=S&(o}HG!lCXWmgwDEefl6c~T|V-x}TyjAY?x z5?jkV@ErFb-_(2+vo6^)2m|sYPR4<+1sjE}3|mwIbc0g4aovVGac=UEQ8iLM>XH%S z_>pp|3XkgHpQ)ycjuZ}#noJZ8+tE%zl}*-&8U9f{SvsVqlr^q}fVXv#%SCf)<&VUy z{IcOba$YHc$4IEw9=X=hnt8r{NjPH}BWntg(T5SM`FJ5+yXepKS>XdwvJaY^%Vmm_ zd*Y`_w+96-=ZI!8I#|=n9Ti)gQmuj(yNVycFRq^`eggFOeWYxbaS7l({beglbuL!Z z8_wV)9**MRw8*)ZWq%TnJX~D^Q>Fb^_uGzHu#daCWA(B@>wFf1l((~cdtK$#vjG-0 z=B%(;^`^HvTePSiNmOxgpD~lzb7tus!qoGp0loYrK|3lf;9fDR*gdW$x9m;o_?pLi zEV1N2-(miCwHHZt9p~lw0{%G9bkJXByhrrLVQvga8=s<8&1(1h;F#y9W5#LG`rB81 zPgWOsA-jHDy>O+ERiHzmIpFX#qFo>#T71x>=iXw65YNPvEKR_DuFk1!eoU1{c*`LR z_idXr{YDN5rLa2hkXstJhlhFn1Ht__kNH9T%yaf_##L<2**;A7m%MWnwaLMc_Mbwp z+1u5(u74m4(f$&4+QSnc$V1Z_D#ZGh1ti%n6S$nVv25}hU11og(HVt7@_!YIU-<9g z>bqB&Y$y--wAt+d3(B$Su@ah3X5@TV==|aDLoE!1$+cv)o>^0$=F-8+ga zS57372$JULrBvXiOrqxiGwCk-Ry0vuUK* zT~lO;jxJ~YWFISE({sEfiNhV-3Fv+h-HWqU!_!^D^9yQVHPua%d8+&8^ShJdh1rES z(WMJdCY3W_G-&Cj^#S{?&|g3B8J6BlVp7-EyCV~KrM%Q%z9F$4P(F5=t|@R`vYCX| z*mZU?UK>6vzaG1a2<3SfI5oKqQtYu2+HL$rd+)|mYn6$ycbe^%?YX$&`7Qdf@(YBn z3^4PBGrpRIh_$d4{?vC#{#tj6Xk)QY=>#H@Ps;gEftn3(B+tEeRN$BJh7q*migl=x z({lMs$w}h%ikENh#Bf5}54_@EHYce?+1g8=5eDxeO_zjLSpafmooqTvfIXT3_}zI? zNo|g87a6Yp$HP%u;TcVADVUxfS6j~Ru1p`~oIA;E+8E3J{|vcGbqsotYIfYA@EQQz zQ;HH)`Nmi1u=gK5Jg=x|Rz#2glXoAWiN014d_(^#6t-#$&O=RLnW3f%YA=dNi*3cO zk3H5?Sk_=720;?t#`&@1oyzdg^rN4tdHYscH&Rj41=KpM$^TKw|E+@hQGh?ptt2Q~ zs~l^9nWm>E3&FQ1UekCQ){6jqBgLE zI}R^ea-6{BNY##MBQP}I`%!JkRpRAbG0nJ`I&sz}&#Nty12kRkU?zytx@E~0f0217 zGL6hs4|3fxb>1}?yR9t3t3-S6k=FfTjLcS*I_yQoVY6V2Dp|8p5}57Gt7d&XmDagY3`_vEZMVhp8~vdz z%ht+^oJJOdZ3HqA#4mCaN~~c;ikPb^SZZ>cN~$>X{N@r`Z-XL_Iu{2eBoRr5!!mR| z!fW^FdXuEGedYn+K4e7Pu7*@;{_(2!iY2}I{+rRjkZM71<`(9fm5g|Mx`bEz%~!<^ zdWB!|KOgTmmFHOq|7<-xot}n3zm;#t-^r|+ims$AD1;o$T1w54ClIX{;&a(%{Frr> zA<~26wK=kASgC;V=6TG0qpWofKVE%{=2dACZiJ%Sp1-2?T6#EnuA6kh$Hh|q25xAi zgL{d;a+TJ(>RKOo43|&6S1YpC0FBA?*79oKtIn5w>7f1#&Hn04|0&7#1<>Luo)a6Y z(U6tSb)qC4KU5-Sd3SfR>g8+8jCK5ag@)iz{aKRu|7yYb0qGT*KFf#^Wg0Czha?Ks zFkzIH)ZQr6sv6HYj?Cqxbp$gElF8#L!i&%%Lvb~@-LR}_No(Do7m98_2j%@ zt3t%}H*um)XRlt7tlNcOZntiRw!X6u6uOnTF-#?68}Mm#Wl$A4zHq&mH#JAqef*fs zy=G*xyENr;y%{sB2JCBgo#C7>@AA>lHuX4VsO<4e0Ca*#`KJeB%pO+KS!VLN4nKs4 zCdBFEjWcwSRD-v^8Xv>pc!nLHxbd&{OUqJ&Yz+?GEGSV;|5Hg>Vpm}LEyTF z`sx$G`@!t{n{VP9qTU|bCrk8xm zKO$?k+CG-x>5|GU(p2IF(rz`*Wqp`sR^%i;w4FjQ@Bjl#7q8}-Xn|IbIdODI!>+a@v>;T9`pNNL1~;(Tmt zjZg?G4Hf&mQn2NCaE+@u$*JPmMR3AhY?5VHW2I)*@A;ujGTxif|7;}g$+2AFl00i# z@$23M*1g8UgHQ`S`B(Hm*Utb1#Xr+}th|JPFYXH_z(Lq0gXo<5L^z7k(`osUl5^Ua~ z2^mfm2-`*3TyunI+SXcn)=W$vJ9dqXiPfV2=WX4&wa56B7kemP5XVg}Tu0<)$)Z;z z(W$&__G>RkWP(X3>3&-iumm280QM2HFw-QT;0GmdB}-#LvI6ez%6!YDPX zB&N=x<@&sEBtSta1TW96OCgXe3|P%0v0jur)q|#h)jDVnwpx42okSujhA7ONQ^1tA zN7+{0tEMSYF&+T#i`zZS%F-I=PG%TuyL#{6P-k;T&G~W$c+;_&iP?~x;Fp<1mqiy?%bA6}V zV=bgFAOsc7jjkiLf*{<)y@suSA&E71N>T2v0`e4R6z;rX3$i1f}*k zS#)W(WABr_>mm!fJhvv*^{W57dvTrdp)gq$(2C0e_h}1zZ+m;=gJ^S~mrsyRP@9^1 z8z(n@p%I#0{=kPg=i~VQ#NtZ+8wK9P9NNAozhZ3#K3IyxG`)q^vSJ%pItt9m1otKq z)cGieX!UFd>9xoZv$5eQ!k@`Po1Dh0MZ5khoKb&jVJK0Hs3E)g?gkZ`&OE0h`v*@G z^gFNWgth77V*D8*YQDZEq=|p`f;=|Tn!3E?0`E*kz)~K&;uCci>pm*)M1Qj1`mhv1 zA}ggsC~3n3FJ;P9Vc0o;L*`uR5sEw7bb%{oO}ulxw-%D@4bLqsr2TlUvim#sL${LN zAxKoO!#a89Q=$zqe+kXF(Xx9;ubA!&kpKzA6J^2#!#hNJGnQpcy0D9({eZSLU?rJ9spO&YL zTYs$=OEMEvK+G_SQJOT~Lk&&WffrrBWYc@Np9qjU&ur>7m%-5c>S96lv&y@}+6-oQ z6-muS!Q(wm$hG}Y*C*;ymgjz6t@9_j(u7!SX<05%FJweI;4E(g6Wt2beoX`?OuQBn zFh^b#nodW1LqEXXy`smd4>RNhvZE>QrhK9c()+lV;NcunA0b*~_hb(H%}MEG?7^ej*E9t?Kddp@`HTTVf$K=~av-$yJH z(;dY5vSeQrvVyG9m0NP)PkIyh_~Yz5;o4Jqh{nw@6N|S+f|^i0Rb!z^E^)`k{_;o6|0PC%z!7Vw5*| zD$?KWsyxeCZd1hqn}^i(@LTH@E8JD;KA4+vz^hitivB%1q*l%dB4v9cxgbKJ>=QLa zsbI3KP1K0tkM3!Sbd zL)Rw4i%9t$Cx_V+75$n|nv&1@qg&OY(DZFa$huC4lRS5~t4Mx3+UwnnVW%bLpYr3Q z90@<0vG8}R%HlezX~m5a$E$L@hoW`AQqy`Wm#FTqtFF3%oUCF;ztaKNOf24xy+mYR zi%&&#kiv0(o_{ zDI)IvB&1}=@25LZwC7s$bvMcR<^?nI(3YdalcVi(s^0U!5N>hT`_2InXsxk(O()Z* z2Txq4#U7wxRyQqmLC6LrB%26ww!-UbUZR9{AFNJd7pH)qjX7$X*B%JXAhyGm820}@ zLhwIDeLFk03=|^9(0a%YydO#$-s(P&IU2T*2$;#BWna*XbtwmnRzV@HAAOHvh4WBj zSr)i;S{dcKde}nIz&nyf`*}u>zWG-FC%6X8|B*pPP{U_LGHnKe7b-~>wRB=-L%#J> zC;bFfyZc263;`#>49&;6!^hVW=4%N@6E6tgeE;ulgzc?wVxBvO)8?_qFyslJwAM(l ziSZCJzvtI9V`P{O*IUp(Ioq?e)OGvfjH_AC!gc)Rp>$nKAU0}RCmpRG>OD4C~oo!6m zuYNr~g_;|0EMt{-y-mC_D!nZiHYXrv+Bs<+24z40Ou8^CVpapE&K>DdzHDEw9*NVl zaN_lUuDZ|$$G3?=)5XKJBffBd-IHk%_dR7*hjudGQut?`%$H{8jVi0(i(F2-_tu_P z%i-4|mio!bYOiv-%2hp}^PsXqjbH#hb9(>t(>*-8XT;(m8$n`Y96=2Nj2=c=lz1LQiFAHSlXnaiZa2FNh&OFyTuZaRTUqaTPzvu&vyqEA5oPBryIL7W~_U^{-^tKPnQ`{3f%HCzjB^SU&ld&>*?} z%J&3tJDPceX>DWUKD{zfkeTPW%{F#^S9AWW`y$y@i0(fnd=<_S#x%Mf7pWIhxT==B z8<3!`+{c{Oq0558IrVL0fZ;duC~BKj2tx$TA6ok7u+V_v#@Y5+Ats(mRh(F} zR!O&1+`_79=I+YptL-rb>8Y4wW>~3kqd1B_*V31Ti>K>+L`-e0e;PDl&N4ITgs!Ao z(arhmkNy^L*#~)%#EHYfJ5_RqI~0Za(gD>aESF>q1W)>&uZ9+f|z}YB4r3pZu%FzoGHcA z#H0DlG|99?D?TkC2dkOgBu)5ugY_YzJ~tKbXnu?ETJZki-ZZR}X5Ci;k+TjL5mft; znJ)@Dsc73?5FW-w{LJ!lbmct$ST~nBgVk661y{mvpLbACaHh~L?NKwG)^Vm`Ixoca z)U2j*M5DuC?fM*~BSMbh@y>7`$D_(m(ysdAVLZw+f6H`JXbH`J0$W0@rqoTpEtg16 z&AjKix{weV^X~AVkfcO`zMLi9G1y2UqQfyWns&6!M?LynxS~?9AZJ#{eOOAQSp@XJ zngDy+q?ejc^1uX(K93urmoFA~F3`Sn@gq!Tz8!HWxCOOnK#l8VcpkcnvLxljqJ`O@ zK%X_CG8f(}2^D8#LF)lDw8b&|2>iJ{184dFXM`t$`6b<51mdD+yat;N?YiCd08S7Q zC4KBI@n!JyIw4!T9riuwtt;db)+(qvcE~i=DVPX6M}L?uRB@Yjbaa?;dR-@-5_udy z{qFW46;v_far95fL=N?<<7fKsS)7$s&0F7Za-Zv#q)gPv^s5`&O?D4x1*5)B3ZL7O z+hXU{rHRcF;L~vY^WXm*bP1DfU|^MEPS@AtFMBNqCP_JFVyU#yghi0odix`e4D2oB zt^TAjR{@wdx<&xkJJt9mG7iEqG*6W?@N&`XG1q+XMm-CPcndM}kIq{3*Igq~UV0A5 zQl)@O=cI5i>s{E;o3uP2_Jwt73g;z~-$cGygzfdRU=@2c`AB!ap6~7r2&Mwv4hm8i zoVp_!FV@EA`$MF>;`fv<9xX-E4aP){7)>sE^d5Wu`m!p;kwU~;)tYg)vy(yqA4GGw z_VLCW6pq{_`|!Tj69@Y7dhek6DBtd3u3S`fM1srJ(4>|gT?XTFJpEMod;HWI^>(3~ za##<2Dk2$PtMcvDR;z=1V<44aKt=>!~aTVyH%`O@x?3G?n~N97rsUw~NHTffF1a4g5A zXi%^R(d2q59?F^X(f#3}XB*}M_ZSxNK88mN2bAmgiY{NA{7q#2`OyO{De6#(U5MbB z_Pz11B=-x!*Tl&ecJgs|i?-1)+$KO68*65xUr2ufLQQ2*qUJdX(I?lBDmfn}yiJig zIMJ)y(ZcG;@(<2r{VLGv0EuaDV`KATa9Ic9~mfip~{+cO7GpGUr~cY6D=wQskS3rq@RjeLG2@*TWFFr6To z3s)#D5kb0pzFG7nEjZUc9`DW;KK?vs?64QG=(%R)Vj9+7zLy`!#oShtYW+KQ%X>~O zBg9xh$@^SDi|x}TIQfo?P${!5aP(`B%Ie8hA>p`UamcxPLV|DQt-H+t&F7T8x_AE- zq5r|-7A_@_cgLMOo@}Umjrq613&dqN7&qQf;O-(3w(?M_ZX?7N4(6i;Fxp0-D9d zHj9klj*(DkX0?Incn+SNjPsL4KL2zrSN#xUqN_o$8^=qDqqJ?(IZ?RAWh!nC3LuF* z7ht|x$p{#M8T0Q^OuV}kKeD@6CZf@}Vx-qAZ>h=4pGtn`v)O4>3^?c zTVEJ(dG`CZ9eVGK-Y2P|hx7EgGsCOuESpxC5Mti=V9Y$-J)rF~ecTj6PwobM7=Qrg z{pB{^P)4116({yA$#vuR>oVoeD~q020w|?J5X00AXNe=iLgIv~qPum$XflUM=>hy2 z1IywS3WER-nfI!K`O*zz2mZKMwbwlA87%p^>$#A{zg`Qofh_~~tlbYkC80 zrfh3QI^@;IVwaGnc1f1zj+jH<1hw3he0cHDoV8N$+KqG|*cv z^8P7M0AFKq0F;#;d~{muO2p$rlbNXLGdpt}x2pPm)u54! zg{f+^f=LR=8n8?;R5f> z4G!UaW2t-HnuKa-n5lcTtMJrg^M_1X%LoB!d?b9WqQY|3$CcXhpZXpg<|Gp|aHQ?I z2jr~KGxc~8YC5yNazoMoZ}ZT_+dtiAo}S+61-jUQn%-fz)xjUF9Yy+&m+4_M+VTKJ zO076-hIa)6n3wjB#jpS2Fg+69sGMa4TzZ|AIdNGZ{KD1?u+GsE?j$M0Xqe+?_Vc5@ zvp_Ajd0NMScvZ;_Pe^685te=Y{~+1nsZB8X4uZcN*@tt?*bwk#lqAM5+qq}W5R<=t+F~(nongpIGShev%bRHji-Fsdl1zwT z*80BQ&L^dE04TeXIqA|(^OBm&uCxSzoBoy&-9JbpnC`z~VT|tQ^bme1&n#Pxs0-dX zagu|0Oh`4utRB=kw%k5#+2{q-x)bYY(7&7AvBO^W^U&u8uim1&oeQD!8%3H%4Agz< z2{cfZ8kb9IVp-0Qm~^>D7WnM94=xF*ySsn&DWlxN>2Qss{z2ix%|&1%gW_|#-{@ej z&!9M3N$5hfhiYYW(ZLS95~&tYlM?95m6Hkk-sZ)xk~ML4)Vg* z)&2vi+GCTxfQiY^ln~!{TY6tZnFbcq$Lh!aYZT&7x6-^5>&-pl(;xZ?IX;EA!7EZ& zkrX;WYfGe{z5828{a<;X2O_Gfcky~jW5owgT)rC*I3t1>bD94-73i0Ygb9j-0d{_s z8TG2C?R49*Rl&3hjnIOuj0+8Y{vT(OhAvB@8f&xUxcg(d?`laCV&sQdB|_rxBXf9; z!0>hC$41Y(EX^tFJIpA~Z4yR3Lc4I;e@oZjsHz#}EBX3%k>SnB^)zZ8dLUf&1Cs8F z5yy=ACL(d}lM!eiH*V+xZprAO5d8Bd>^*~V#q^-Vz3AZbDo0w1OQf(yQkIi{@v=zv z!GWf;NoiB{nfGOocUC$(=Z=kNO_57B1JI%9_TwP`c{#@U01q;MzHz&UZAMw}^Imxp zrxbVDLWf*p*ZO$zI@^4j2%#9&NKBk_sgt|ZQk+A5VoKPFTto@)^LnnA(yM#(&0dWE zwY${fUAecKxU3GNC%7gz`H>0P%-Q*yvlO;}rVPC4nqehVz(CU!4)mA^6D&RSfvMgMA^#nQv&8ujQ6dZk=z&*DR@uj+6TNqbN38c@j?^~|S(ZUEp| zEm*vl$0(GipmS37%*GeIGV}YLBGa!qTxQ)Hre*3`Mcm_4F<-4savY+@QY73RjE6(b zA>vazQ#o)X=CRwBX3s640Vviv~~n`I$7;W$k`^ z$V9gkQ)8)9dY{|0(dzM&F^eZatGe;UQB%q;l>)SI=E&`eV9R*espR zIbG@KM_QUd;t$_|1C5Pc?z3?^N#Y}XPM)l}WNBfPthT%*339vHe&oaZ@l?=GlDnjr zV=RH~xbG%xs6IEXhKUgjZraw&ap97lEOR#~fv+t&}dtG3>$JNcE*8D3JII@(T`j>MO@RyiG(Q?g}St72eo zd-6ghgYE5`=Rw>;n{w_cr}jizf5Y4j=j$ zdZAVxQVUtXZ|eDxYrGNZ1tQxIJKtaV+HU}EGJ;IR2uUg&a^k??jY1x)#h9+e^@Xif z6(2+Af2K74M@sk$SMoqZy%4)U4=H87g4SA~Y}*OZ>dSd+MaZBM@+#n%Hep9!qTqxT zAVjuX$*#G1;Z^uTJJ2`T%_{}sDmGP>?YWNe<&RI*eE%#taYoy#{hj5UDA zeVjr#!WCIkxGCT%9GOGLNy-AqWNCY?{AQZp*Vu?rT`rJcjku%YHqOZPW?DiXpYM z|H{#)Z!3ovlG*3YP5uu~dS28c%9hY|)*f2+65@+OJxhFXO0%Y%HAX#07U6)~h=smS zc1_z(&o4c0#}jF#x|kQ&;dit55S_{j3UUoYE=ij|l0{t|z2r584S;;R8I>xO%-32h zGTq{J+X~v(vi&7Z!SIs2Q*WhBdRfQq{mSj|4&vNlm2dOMLG@{^NLvcnuittE>TP{$ zZXkhj^HdF+*376G4JFG>;w*CbD3$Kb7Y|&wQT$-AWu*SV>r9VIZkN6HvuY&^Cou8a z$Uv`HzTxqNPl#JC!);KBiBo2=F|>Slk15Ofx(~kZNM4QaB?HlwdEZn%kh1?=NK+dJ zdTIr&F+C_#n3KJ&qklW^c$w<`X|kywEv@53TtYh<Zs4YPNrBXP z;ZBqP%z`?N>-N*edYPcdKgrCI4%(a*pZbO9v09>AqT+1&p$oApz?ss85VD-7{N%ma zlNLq$G7f)QQ`oeb5I#tj^j~s`aK)~Ho>C^YXj1>s%2sLudQU-shtU4e!^J!m$+x}2 z$SuSHR}G+*9vHMShj)+BE22%QJ1ek@$%;!+a1nLh=`U8jC*lM+v;FG-|tE1m|vm znFImS2t41k1*=T&d-(VvVkMip&4^XP)HH2tmu{x7TFjvc@8*6AKYs8n|Gqd7&-nf} zD1Uio)`in4cOMAbsK1)78c7d|Prr8}R#){(!vKHR`ca z>LkP}sljnea`BlFcVU0=b7d!5-=p8a`Wu02&jj92^_*3FP!|x%aQKARvE4}Oc!Yzv z7h4>G<7*SG_3`<8vxF-_D~dBk*4WVw*J<0s+t)Ld{im>zoJ-K<`fGALQM_=x2eOWb zc-Zjc)4E#L&c!HkLuZ+X)3@i1HTxN(6*_Y;8z|@6Q@SC%jZRJ+{yu9Mn&!~^e|6w( zckuR(4yp|FFfn|urC0)3i;niCQ@|6Pz`){?E;M6>odyAMz6HmFXP|&Sr;ZhL=j9`5KBIQ4 zxYJ8j5VtjtEo2m+=#Dsjf3?Pl{XWgjMwqd;7o#P4eK=Ae#cQkLYB zyn7Cz+Z|kMcZWDF?S-h6vkE(~P}u0W;*_T!xYgRjr<^vRlc=&v3bzqi?bb9+o8ewb zlAvgO5VA}4pBb5l0xI_>_jB?2Mh}Vmq68S?-0;)Ax*=Ns+XrSpv#>i(k5hPv?a)^U zqZ|E|*ybEy2f-O)=dpKCv;I+W*i-N7pM+wVYh}Z0#2oqp5)Qlx9VRM?oUT7I^-3Z{ z)N7L_0TVm{AJ6l&?j@JkNcs3;9NfY)Gdd*FZkIZGdBl2gN&T~54$T7@Ln|wKJ81*1 z1=huz^7w2VGA0^+ewtT?Sy=TjjvQGTg(ugWOBLnZ&ZiinAVkkX)!RZv3}x1+PEb4KLwdxV@lr* z$|g@=f90V~2w0&{8G2^tL%N^9&k{({1p!U&uEv{QRlZw8^*OPg>jA~}34azO8FEn46;$MZA(^zw~v^mvatDt}ePg z*@V^yF7o0g5Z3ooo>0k{2`t_hyYsgbe2+&D6T+4?^)C&0Q2H<%lDK9vpZa*_-hyfh zGXnWjhAJEBE)Zy8+VsbGA=^<+951$HeQ;PaUAhc@5(M5HGVXtk+56$9vv+foL=A{{ z^Gy{&q2+*%eM8v=&jyW40NriGd7Ej!T>}=dK&>zrdaoCJ-mbKqHuEf+J2$MN7WXDhr5E5M5Q5Itlx7T=l3} zCmUVY^E3{RaK7Kfet%iTAI(h1s9Kp zu zBm$7IgKmZIddcd{o`{Kax?6T5U_0qW5*z|qXZhH_ZhKJH=Q?3ch$rg4Mv&EOc|X!{ ze|_F-R3n5;1pF>rdo8-EzJ2oMcRxf%OftD`$ z&DAdae4*YCX!ZSXFMwM?cZbvbljC--Nk|i91U=Zp%*FSYyS6ibR%zk{j;Un`(M=SK_Z% z#GmI1sdcvBLh0;Did5|eYibzLUYNWoszJvp*{`khp!`VSt;H+UT6L%tgVn&FQ_CeN zkUi+DJh2q&ZU)2XLYT1Py8-Vp*FaB8Q9*LO*Z^Qp|K$){*D*mxQIUKHm?`b-c_3C| z-n%iv`fMXp6?2OZ^;c^7>!+U;S|(vm>ix-n-L`q^AVGp7BjOh%!Oe~QLbn&cydY>9 z^Y$*Phpp^za~_ zKFO3i>iIs&&Ms^rCSTYe3GA;+6eLe<)v@TwOB~Nxn{AYnv|XE)d#;*S6&2^Be7%i{ ztaf1aRDH}c)bP^mK>xx0(c@PKf%KP6p+9d^4G)TrD)6$q!tHL5eWD$v?)o`lVWw`I zcE6ag`iEeIJDy@|z*D{PWS7a{Ab5FBdwTRxe!{_7`rn6i7XF4J#hV<^eWxey6Ljgr z{Bg_AXg)d8VPeVX(4m%QPA|#OB1_q#w2~fw7TC};h)7|lh!Hdd)Imhk)l>tl%Qb?`siWTC z@i!ekLs5;eTSA=($qOSRbG{9Bavwj~!i(3{&T9FqPLfU0pcd{#%@38CCWM-|o7H?L z8+!Qrfbp+5Xp&mahW(WHNWoZi@~DPy3lK;1N272+dVSZb0N|JLw&{HWUFQ>!w))-GLVq@gM0|VVf|j<7B;_b#*j* ziL>Xitp`%BDPy`mpBz)J%-rd4H4THxk%`NBO-1 z5)T7N9*PSR2;*^4pN`ox&uP`^n?7`$0)2A)1Ygi2&o6K8pe)X*FA)OEiG=lJ2!%o7 z0-mRZFE3~azadw%f0a>xl>&EKTp;dg4%g6OaY~yBe~3fQJ$39XnQ1YBR-O0PeQk(~ zkvZum(y%bHhp)7?#OZzbss8LcCP~6u6|yA|Fq2i)Z#*^ZSGP}B)W&yYza^d_AgBz9 zWITQNwwqlb9R_1I%UEf7k(nu8E+xflHlMx>GkV9 zspQyeem8(>p-wW_2gK)ZXq^-HB*hEdIPs_LV6Ek58%AP5KMl#t>;v1jblI<2m}R^I zj4|@VF{P*9n@THTgADNSbr2{3{8Gz}43sm&r@5jn^=7a`aaT`YB7Za7J(vZGqOH%gubN}P z%Vr`WFsBx4P*mt&$+jCkla%JE%(9;hC~sYL>}arPPUwz_G}Q2)OvVgZsrbs2WV&D7 zRlYSjEm> zihlU}`)A}1E09=W!{{h-d8V*KsNU`;$mv97On7fS+N(7LjokZOOq)_@v+0kDFXS+7 zh6>*OzV*H9PmT_Ws#l?}I*oBvG_TWHQtAtTdF{otVr~U3tgAeB2FBXHxDh1lGNW%v z5<}rS)aCm7k!ouyQ$IC$x=kjlw2R7smGRyRYlCDb+V2R29#S6F5x+`WHwp>0|2*v?_u0Aj>?Po_XpY#gR7Q!0 z%=>qrm118C{|%$M;E@S~t^_?B{D3%6tdyXP(X`w=Bq#8gOWV=`ZLb2{MV}!N@j3jkmO@-%430`u53Ia()!><~lnn(f0boLo3BXrn;oo zx2)F$&eEDf5y=bag9_B58nFwEhDjuJ{RHWfTTPLhy=CJV4%TWyZUXz8--@<&jkGZM zUUq`n@yd@+fSkHEW~RXsaCOJm>Py~8$*=g=_vN1X-zm}vP^7c54)*5Eb406i-;%E; zSZCikIKb=J2DpPrCESLyhnX1Kn4EeG<_q##Z%-hGaGxmbVtX}szG^g67Pes?T3w+! z+&<+#qRthrZhyRNY2XXpUz4`bV%Oe;hnd>Gh6fq(+fe|bZ_9P4=6#)++$+2{(fZ{2 z3F)WZ1qq;pf4J)}yK+1U@OSP;)W=;DA6WD{64iCohg90q@^JzN^GpMcnD=(G$u=hA z^-xba|4v+gc49@s(8N7YIKqJ^_T3<_BprACZOhdtJU9wbdBRv-&WeE_n4OXvz3z|_ z5Y+Yg^q6vLVDo2EMv~Y_8VkouxcY<+OV;#5M;)>~k$>J#8Z7&v_)-2zK2f;W{CL|J zdjc_$fo^zS@(oA!Z4IC7+>!UN2frRqzLP%xcD5ymmUi8jC3aYEOvYhO=TVj!>-iJ> z{?+T9hl={-<_(`RaleeWuSxE$+ZR-u&!f9uIX4~eQ$1M<8LD7bvleZQTp~zMG5x2l zER5)CCG|72eLB=++|Qh2RI)jQ5)3qg;Hpm=+S4&LZE^x0$Av3S?McO&idyv&#(6T)#A!-9ehiomDCd58> z*tByUhVRAN`J+X0AO7|x;QSm!#UeR8I<$Ll&rm^p7wj?0;2iiqbUes{WDc$>x3mW@ zN|*Dcl%hP^g1YN`W|MLt_E13~+FIPk*pHu$%Zz-J5{-J(M0~zZ4an81RK^ZXc8IDyhi!Ckt7 zi)(H^zhTz~f2-P}+=?idDv|$9pXig!OC5uwYP@+s9(`@+zKfkpY17m?c~o@1zQ*2{ z7zy$QmUXNDm40O%V1R1g?TO2i&Zk^3H*_e|yYutbR`-dJQZJ1G)zOqdwOlx(B@cd- zhWo1OfNi10yIJ+Pgg+8+)etpU8OTae|Dq!HDd)sqz(f~{)`$(h_o3`K7>Xu6S;@|k z>rr*R&HeEmX8;M^zxgSa6o$q?ThjHlF4F`p+`7W($D?FI-Nxe0_4Cim1k|ai0p+O{ zPZ;u>NFMRZ1rJ`*zyA7#6PgVm{}NT9E$lL;EoqmM7GhViWaF*4nu_8kTgLC_(YpZkBO$LR#kMH-Jt4q zSvn9EnV5UZQ*iDbg;jPtG-7;pz8!xBq`sUp;-!5O8 zKz3v$zRfpJXB10ss;hVMqYJ2mS<2f-+65S>ZYyZ&l3XL3(dwiYYepA?ZHrq0n^YV= zMNjR~slQdfFJ`1zOF8aB;>%KKtMA)^~*kSFV>nLQh$o(KXb?U{B|c)$SKQSIDRRNzkRyDoL$&U@fN%&Bq914 zV2l^Dpd+~7@|;>{)D$HCc$={?%H-nH(Kr2i<#!2WN(E%6{ol~m#8A`rnZ&$>r}gHD z2Y7mU509>^3wiV=S7;l)kNc(~7|bQW4ks!$*}f0Mj!eZipmd9shK}L@{HxwxQvy%W=wfM>AY;c_{?_{%w8mM&PlGQmyb{n~qPuLkg+-%s^Fc#0+ym z%T~gCSTq@f%y_Kga1L?NMl2KJ_$ZEa+>`AA{>SmbKeCG`GW=`~uUvVRe&072KpksC zmAlKm{;sH_an;!ZA1Y$45g)|V3e(iyedgHju42BMVRkxc)8iPcCcnF z;a{nfD}gr+dNQq~J$6`(d%2rd@3|xDzYp=Bf`Ju6ETfF1cTuB`>z)3=g=f~D7Q^W$e@?_e&;K>r>kcYHIx}6eOxPO z{iG?GY&_N=9ls&lr5qdiZL5gzez4q+Vz*PXJxX?^N^-0;M~HKm!90#Yb}CI~)FECn zNDR;c8rmPFoQjZaKwo&mjK<-nV0~v==OT8BUpyopK6Lz!6TG&Q{!BesN1E4obadY@95Tv(sZ!!% zp`{SY>uk}%7OR@=3aBq76G_n1lQ1s*%xQnR|9#cub8LKM>)5|_C-#l%diMA81^-)^ z^1#-p@#xN9Ms^a^0&TE9u%-UmY3H<;UA2x|Rn%OY50ey9#h3qy7!xmeuBLQi)FLr~ z>^EXm@Uv@kYf6TOLT*>qIc4*}ZQ3vjeqdEWH*_Y)(t3SnCCv57YJ1+zv507#5;Np+ zV2;R_`b|q_8VE-fE9Hp)BC;}P*VTUQdUL$wD@vj^cV$DdY)>`((N~7n;n572SkNsr zGgE)x&zN*)Z&yuMmwqfyUxBl{ST~;9fc)RCh%!>Ni<*D}xV1p>r)fjo%Xk1k z>tLl82I~1lIf=@NwfWQZ9>KlWz7NbfUT9gHx0vn9lE#S1nHNKED^XF1h2?#3iHl%O?ZdAnRGe6q=;({(Fli@yqbx%MEUa2q2Q0$O zd=E!;ZSCz*N=)K3bP`F0=RNg&m#?W*2KHdV^rE{hM;;Ed?m38S+Q%oY>B|yLF0?WT zrWX$3*Zv@zH%17jfApv!TTOhfH`#X;Jo9nhEn?xitC$VGAA1Gc=P!E=V#^3)$4caX(=f;=*^V7YbM09Zlm`{#<@_4Hdxy9C7)Ca zf!#L5Zf=eg&Ry-&%k3t@;ye*{^)81N7xIc8zQWtykl|(&36Th2q=HX__r~j*CySc2 zd&JF7q(YD#EScl_pWof~(!UdLl1)){|AZhz^Z9IBQZ(mfZD{M6X25IUGi8JTpI@YH z_CLVj4`L5*0A4f!x*c)e{=M1Q0?qsveDS>Gv^2ZQsy+C%F^ls6{0bsoH;sL0ab?*v zLtG#M{p8+;kk{Uc|CYWhC;1`Cd~=`FJ1(mmrc8B#nf;K_8mC7|@3yy|2sxT*wa>!7 z)$89Z6Yx~Digovbkv}t3)fitp%>{`9!I0av*Ks<+*nTkEBazS!^E~M8_A`GPNoBz~ zouV%8D2I#Eb1etHojomDfqr?&R5(sv2`HNcDPq}~^BkO>mOR5Hz}w$U>*gC^{kdL?QH2mb=vfq8{%6||brc9&`@AWYz zq3oRMXT8@_cicxrtSv1hT<_eb|MW3`{=`zl0BEc}_GyO#WYF{`5IDdsO@uwLS?95Z zdq>{uEXU``jQ4nf2^&TrW&m8vfVCr=pTrBJTV%v^T8(jyC=ZohItb6 z3=SeOoQtx?8bn3!^6NGcNczeG55UKKJFC6`)?=3ST~aTM3z!Y|fYC9IouZoKuRFE4 za{P@wWvea*80IX!jN~h)ct}w{$l*;x+!iI>qJ}GJNv|or6xQLqXeMK*khgG;vE_k` zV>k#cr9VTaNlE&^2`uZ6qx= zq0q}bHlDtmM{Hl2@VB3j`}Vgssfs?L_TzLujK%2L?w=*RZs@DR>XCgcJADm&P774L zlZF!?b;=E#V>8iU6!70Anj5XuN{)R-G2OV^8~`8G$LYrcd|)f_Z;0XXKLjCPb+tEn ze*mtw?I1l^JQsgX@m%-7WY%L7@vA%~;`5G0q&JH#%WD!?_prq^lt)r3nK9@%>3Esv z&m>Hj2z#Cfwum}I2$ucAE9lO<LTi-zki8lMk)&8JwG|O zHud1`j5UFNFbCh2BF)G31GRr?PlRkk=#F45O$S^Q+jC)F2VJb)k z$DC#-zdnC@$>8IP`J`uAuNXr4mJw>{l4qim2|lExP&tIk|M#M^`H6CdVA4s4dE8Js zPx+VB@8c6?-EVGJi~|=3r)ZJIjs2gAmz%v?qtGIfn>#8aD;?3^k&{K zo#~&D0{g#QPxm41>z%j$`aZ3k8+h#6lgX97ujar{*YbM) zJBVqt1(n=@uzEHVSD#BhJt;;?+Mn%Jtsaw_Jj^`ErVZ-x5qHGKKG)1*;;S~p5iL5N z28r8YUyi(vH5m^sb`;>lzf+PsUY=@Cw2RWNRKEX#0>5qbbfESzQ6}&rqO*`TD?Zu9 zzngzd+>b?B_v7!T(*W7ifHRwBE@Yg}^TD^9<`u+?z&2L5Po4g+$)#Z8{+5-FMF*@b zx0x@ zl=&3!<7z@lF~Wd5ffz3@#gBE5fn;{vo|%V9OuXj%`rBZoc{ap(55FP0%>t z*uk`O&5~&3b%X7laz@8uOIkEc{`JT$$Wq~`wmq^kKg?Uk@A^HC87WC=sJ%dpc7<$2aVH~8>x z+9_*#ehJTZ(@%eUC$yC^r_F@vlg&sRlZP(3U6SiP^x3WWQz#}_Lgp}*Y9FVT#gO_+ zYxcoW@_d#$g1K4w_)CK|?mFka|m0$K>KwcOBf$JhiG)flU9BkOpF|t)^_@)b?r>> zFB`)KT^9KUS&Fubnf--e-6_A`iCb;&#d=$>#{aY$oKGpjE4OEa8#=$DQ2JSJaDU3q zdwk%p_Q^#0yFsNCI-VxO@umaLZzVk8;>@ncO*^M*O1zpD1VNDSd4&}M>B-q9Ln=Jy z$Lu^bRqp;^ND6?I$_D$FAADf*Hsc6dYe@!VM0Jn^xzJc!nnp3KkxY6v#O`f#om z)-E>ib;1}=nk+IP(Hj3uhOo@rrjWZn&J9qS_NeEp1&8gSkEvTvV~HoJmFOjYxeBCZ z?R@@le84L@GLx!fg;v)7Vl;dh(~-cJrq*DZ!5C*J-2TC)b+;UJ%ib!*wmmq%I%6jxU2+P> zSiwA>8*1}_n?AIphQiillGmY`&B!z@{|F;E=`uED4-Vd$k>J^y}gkr z*OG&X*K=2(G4JIC!Fz-x^FZoY98+E})#QEK`HgS4iQnxYx-Msq@zvISRZf4T>kv@b z0Tg1uzI#|b^mJPZ;J4rSA_=23c^Re@JAaLr~t^@1R9iap!vg;kj z70u?xFS!RlmRkV_r7u=}6S;TpI;6Imo(Y|_hxp%gv3*pp1<-WcrQgeBdF`4p(}1EX zAAZ^VFCeE4W8(VUuYXFQoZ1$PKCV_9&;oAMFZFrQV3ktz7EdA;h`k-jX{7{vs?+Se z>1`Q!xkLb-R_hxU7BcinTjHLMqq&dSETC5sZ`$)SGO8y^u}-V7T!-Xx3!;0P)0Lk2fTAVBDOquj);2)qbe!(c|ONtJGB?I+)kCWwyR=64i~0!O zMukL(iNAi7G!Z)R4kgPEIYWF#fnOf_W?|GhbmUn8R!;3;T@Q%I^(XC5*wdN`#?iY! zeQ2M-W*QYVG-J@z=f zU2{d@~(V|ZMxNgh$H=g}dykSyF_7PD2@myQhJ-kap>yqE;6J6Q}4^8SGuzIv;V_eD|So7}Pk z@p)M263Ut1PWB*Mt|yQM-nAyEIVR@GG32&Q{FQgv0PtSXgPFt*Kl4(VBP)sXt@4)5 zB_+ZJt9%@J9}n(>2N?#HnMoKBe8|jq$55w_I&)bfvx5yNj#Ob`GB88OSM^T&X;KZR zV%M()tI_qZq2(Y#iA$Qg|3lSTMa8wP+d8;g@Zg2JyF=j?+$FdLcMa|q+=Dv=cL?t8 z?(XhzE9;)U&pJ=_LMvvgHpiI%=>6+`jv1?*al&R*FCcDL#y=*wGv--v?wAudfKBTAySqxZ)uNy~&shT`$JAC#z~x~TLp14FzndM1q3 zj3NXQ zf$v$X4%fmK;I5lVn<}IXim9bxLHg|C5oh;vb2DVqDIhcyz*PzX_F~mY+s+#f418a@ ztIn&9_aPlrb>I(`m#ZrORTh{>QGW0WU?G`L$o6Vb#5p4fz176AmCsf~6X=(>2&-q# zrjB>{Z52A0mIX@P;m*|8?u(Yw`#$$7D~kJBa7<)&)pZpQ{wPL^E>Usc+T)B2Ew=oE zm}EdlJ>aG?>Wm}aErbFtDi0LkcE840+uF3(uTVa3talE**mV}qpS+Kr>z*gXjW5Tc5hR ze1TKWgdZ~GGh9ihQ1zI$E6#HU9O3>YX!L<1xxy4(@T<`UGlNO+ws?ai%_jOdl@jSa z`6a`qy$a*|{8%NoK(-7A?N}_WldepWjE82yB!Suo{U8%tCLzmGeo$a7NXeLf4I|n@ zzp8HBYmyNnK+ntjOaTKZ&Lq2e9xX#sT0yo)vTU5)$Jp5_ufH?p2io+ zqi3R0rkRvXHvHn))nJb`bF9T$@?wCraE$7<|0;m4=`8!0^1p+6zaY3qh<_;b=PVj` zcj0Q6Ow00bD#q`P zyuM=8uJ9*^qSD^^9D+3VvC1QYvisO*y&_fSW6LdUD-cd6#MgJb+<^M-gMyBwN~SuZ2w9m@~v~f@Z@MZ2s>zE z<;;4uEy_-@ge5{0z0B_NU+_kx5#%)k^5=KrcDaKn*org%lx64QbfK?j7q>Quvn>?? zm&uxhzYOF!pv*!$Q*%16&P%ND70?L(PV|vX;8wQHem<9Z*gPf@QWq@KX(K;>4EVmtKVl&Es_fk?r&kq4?9T;Hw4QlgPMP(GK&GM)ZJc+!VWnxP z(ZrjJs}@xIKGl%_jR1uR`uNoNpwoArDoKd8*^wKr1Rh5>$rEIwC*HBeuI3 z8MYqj;!t{ubUDZ2Xy!Rmlw`m1K%LJ!(zf$X8@g8$&1pzwn#jn{%N7D5 zc9cLZIZH~9D+-*Cgv)Wi5O+A_qrvZD)0Ti6%(jwSj&zpQ3tgnS>XCT^>BC=FuB38p z1@%hV*Q@tDT||x0-)2yJ@EKrwk_BJaszPFn)}HV;PJ=t{zGd2vL@flv`KmeVR&TQ` z6L+9=UqbdPEa*8GgtDkx;kF}->t0eZMuAm}Ezzpc)tB;|*9pV06e|~4Y(Gqfm0x#U zs8q@U9qQu^oH;V{C`5p2oe-&JlY&(9>FQtiDlKLoSJXz1pI^OIh^&82NB{ap!74Ng?-^;>2r6-sPY7NHi1Y5Ed9^9?7IjWeX5xfYaAbYP`v ze7cxHO3#xlXN7ul?i-aUdG8xcQ^CT%)or$F6$|W&N^eXbpUCDN(cmqzn!?nfv6ZE7imz*uXUYIyB*HLgr`eGIl&-(N`lQ#X0)s75|uJvp4Y-HNFB; z9C+El0M(zp@wBflNp9xPr6 zCIn1878j6UOwCtx@_V-j+n=Z5p?z#$=(9c8{%819C;fo-9$`gwX%`+Ba_y@BF;>&WBTqPNO? zIU(WeuyUc{g4X%L=yR9BWE`!NGFjUdyCS9MaVLi!c9-j3do-VJOU#uK3kZAld?0-u zF57ywt$O)k8Q9fyr^+ZJ(ZIx&%v}qQa2YNH~pGTj!(Fo`+%opN73WqW3mAe{UvU&AKNAz2e-|)5ytZBO|JhA@% z>9?MmR5`Kl8pZnC!TfwmTDJ#;{3J+|W=B(W&bF^SZzAS|H>5r#Ym+P^)*}t4&Nvqm zblBUIv6vg)^AA1xH8@SP&bmvCG~{S$W??oKqLJKrd7ZxH9$rhl2@*%dW$+M3$IQe& zkO@liUM@PuTx>e1=DldbFpVc*#9|q}1#brpQXn8WpF zKl@j=z<Y^`#N^O1I1G47Ix%IDKiSPX;fXgun@dxJ-f7s7FbJAD8!z-*TOKTo&fsGkAg4%Jk+I(+H zRvjq{*?TZ?MGq4gqE3C8|N1IZg@D30ftqlAofpFwL(dnx8NjYbtom2Zs{XAsWV+7EX~1X3IIg=Y-RAvWH1mGF;-{4P2Jq2-D8Jog$t+ z-H_Hjb@A6d_!1PfeDKJfy#U~k5WtiPN&gXzssTSFp*i(@omH5~cOi0QYVOF|K!xXLbs|__9Di3uFVrY8)F5!88W7-Fu@{2x zh3a>;=d51AriyU!^LT(@MoQqHGuz+m7XJjsh=x4ln7fqhGDSijM!)&lgn2&4&rZ9J zZbP{`qcrqeTM;&xavR7uCMXbLOQgw0nOf0ZV>7-=R1s2q@l9RR^tZO5RW(ZBL z|3b-c-g%;wm3xGk=QKsl06C*@N~qM^EmoCv?53i|3aE+1O#VSRD2P7WiTs}TArbTA z9C*NB1;n#Iu-JhNss`mRAaENiSZr#-oISfEn*%uRpF;*~T3U_~Ed2nt1I+9V=R5Kd zT=QJ02>w=}wNEb02Q<#mfa>qraN7mj;hd8yFShFEq%C&#pe3N?lOI5kpN6|y;AoDb zIS=FbBL0pJ_#FLQLC~3;7>9N3kKgQc1^&FV@sUI>{Mr38^%bfn?~LfP1&&rv3NG|7 zfh}Nf83x7y?w)~88Sj?>G-AXK)g_w2CqpgXVeRhXuTenA&!mX1XzZ)<`{OaAXA>MY z9Jm%ICLFeJ-59!<(Df*5t|4c(kp6&5Fd$TP-=w-NXOfWwfc8A-zVWw48Z%<5#PKum zcHhVT;uIZCdg28@rdr`6SDv>i*|v8F;YlEk5r(F9bszMN}W0jh@` z7ss)pSMBxS0Trjq?Fxh*ph8U~7_W!bYKrjQaClTHIQhqC-K>CPUK9Y7jPPtwdWTcg zo&o_C+eXZdC4dahcXI35nTvcywJ}W3ti<`>;O#$B>Yy}{SDU~pdt3aLXE|mVX6kRX z$N2@qT->?DEf!*qw)gp#OM5!rk-&=k3$LA#Br>&G2#L|qKWSNnim^XtVU>(aBU<)n zU*HeM%GC5sr3KVnr-?=UCnM@DL7R19@MovXe%^}+*c!ANJp7X%izeigSHxe+Fb=ha z3}>avula7hiQUxXGXe34Yf&r9TBiN@!n4|tTR{w{CkYz!4U>JRymr0C$k8{e8HmDF zkZ(T|K*lqg@I8slXWHxCgS}~Wq7L=Zw-7m_guUvuw(lWmnyQ7k3t7c8pdW~D4p2kE zBnGKL4dS3D3+YWMt*$dx?v=KTVegK})DmbJXxJ(kEO?+iacDC! zS+*WJ6%Rc>Zaq6?hFoGzk7v2~K80Zdt*Aq<5myLlLI3=R?CYAXAONt(5GU4u#*y{X zyVw1V_V zvkj(V6G{Ah!el~7c_;Ozm!1Mc1*-A96=U@WN1)ko&%=h)#@jq;Z#2S><;MgZ)jA^- zNV;E_KfXc{NrC!cHJ>*&K00q1L8XAVwUk-8f05belg=0H9UMI^l~fLAmNlwK{lAFY zKfCMy{}6%QRWVgOVi#k3TZy;W+Nmu$Mk)O2Vq(2M4*ddM;;lfL>8*-W{nm%JeA5~I zwWr1O3;t!P>8^gj-?OD30)n534AvJt)Cz&(&45YTs!q{MGK4DQW0>4HAviRVO+I*V z2oATwps9*@;?IzXx&_v&7~v^oF_{gIU%|wf04z;IG~aeLRWwk?p+D2|IoJGme8>y> zB#9P^wM+JMvS=S}UL3J@&Ixyq&B`(~uqX>}Tp+ZZ2^j`$N=j&PBnxZBW3r?pWun<@ zPZEL4h``ZPM+nvPl(}y;DZfI>N~U;!qHrqmLUr=?LY9KFyOxpm_B}( zPc$of8L%Amr5!uCZTgY(tGVymc)PjUA}Y4W7u1^y$mmJqI#_p~dZrYYR*qO#P=e<> zLGn<`=LToDn1~?r)A{VFHi3%gdyq{18PF^3*FIvIv*Ga52Cey-J&5Y-*VqG`IQ4mU zVJhZi<&)pb3XH!y_sg#x|BsJkL;%L+9ycY=oJ1bMND@nz z3AtgjRdjugFDYaN^3?8lZ$^)gw%C8tz4&cAX%-etW1Y~0gz`B~ zV3o>7R2&SCUzPV$M_?s)#WJDIa#xSU5$l$lwzQ2+EXWP@aL1|d^T6~g%GdE7qoF|q zaq7Frb7++Fljvu2r*CV+c?7XJMwsPO#&xUQwWd7Bj++?=Ge{%@UNxD#Htk`k_OnvI zY0~&Ca!s(fTDr9q6Vrocvcu4&=kkl+YNkK8i0LSowmsc@=8FrdO)Q;Nbr|BVF=RyY zF8%mpv35HQuZ%fHpB5mnSKe|j%Hm{VlY6YTdVT$54#~s`(hDqj?2GI)pa-7i!707$ zB;9f(=^o$=AP)4DfpvzQUF;pV(=xJ4(Q^cx1e80(xS$>EIrWf7a|7G{HI~jh+p;tC zf|^C-Bwt&`{JQ;qqD@EGl>gdSUU8yCZy-DK;wkzv@+qDpd$#|zsT<-l4CR-bkbz$S zwvnn`*QVIxA5doUwfE?5?>uJ&WK46@`|kxu1Pq@Wam^G@n2)u{!dmj81=^%psixrn+{S6eJcF@2u83_ zU*L|9#L%h+sz31s86TN50L9vdP|{?7B#G{4Yd_3X?Lpfer*^Et)l^@E5H8|{f9dIJ zyE^}__kT7|gI+Y)T^>A9Z86l+_j8QWcUxsdC4;+%-j2ugD+x>$2j`rrj|_{exDS>Q zhy+&F5vJ?$%3-z)jJ>g_Sks7P349R^u%Nq5^Kkkc^I~OzKGqNOdq=+dGa!BmR&*uS zTT2qd6NuLDoLRdB_Vs0>KN}}%$5&9qEOwBw8z%)xZe5A1j%6kzJD=ky=tVjF|M#TI(<*JXiuu)5i zO7U3)Rf+_xvXG<~rXR{;RigX{bs+T72Dm1aRf4&fP~^?3Hq!k|#?`~CJ@ov9!i|vL z$5Q?g@d4CCOz|umXp3Us($L?( zq{}{Y1LX!fj+x*GB%;YT4&l$-(17qDnHWMh_*;>_^RS6;S!AgIE#>WLC=YYOkK#!h z65Jtr`l*Fty3X_V7*OE`z3G_TLQo7V+qK?f;NsBuST2-wU3xsgp)%2WwJvHI%0gdS zUY>&<8;wj!E(1`ryk=!c&||Kj?cvLb5HKAh_`;XL=Pb>=W`7d3Wz?E^hu4_mGlN&v z8qUI+aZpe$GdIzNijM|p8|>?T^c+7;t<;=~@&Czp;5B?L+1~jFOsEyYqe%u_oY)06>bXz!Zjg=C4RRiDe z*kP(fEc43|*lC?Zm7W(y{OLPaY~;&@M&kCLH}X}x9xSEURA!V-h>S&M(q?jv{`~&^ zr#z-pFLq#)cv;F&*O6xqbAo%3itlCHsxX(BfBEPdmHc82MF8#~wl7UjSB*sL3Q_LiP-XpiH8?6=Zddtm z?lgDdc+aY@MUFhoALAueyl!r!c)%_;7`8zV=$cz7UNAROW+j-xn-{C2+bLS?Y$T|;P_-^!N>JMoV@`JK{rrG;mlb2% zyXU&%OL@+{B=ho7Ii5J0X6YF563+!``1ty>Nn9e2!j0-}{lYt7(+_gkhclJERvHIJ z&=43mgd`s6vO0X_{-4$RcLHKC{wK|GxwwJr2fTXfru#g&DSyV zi(vJ`IF`cO|3Gfl(VqlHSA8F}p3fVJNPxzFV1Tqs^=5)^o7LIVGZXFjd!T8?x6R@O z1)l{oQx{Vxi2pxb0IWSur%peLrTv6@56i1dmZ&{f53s_`WlPyOZ)(N*0{)XL7boRXEA?)Da;GwN?M0)LBoUex+}S+>fZifrGm*bCEw7Aj+U^Pi-Atv3 zq!-P;>fgDj)9yZih;4Hmvkr?26A#LL0VW>Tp@gks6tWR*mXYS2{fFrvWDl;pEiwC+^x1U+tKQkpV8oMmaiNu*%q zuUDvMvMOHsbbG2QusT~>QRO4It}weKwsfnajvVF;7+&{g(@{Cf5M-xrivYC{Rot?= zHEFha?%K>Kq=z#>^U_>#FIKfY%+k&nYEmEGJ(Q^ulR!(H?UMewpr9uVroW`C7+7-04`i{C1xo{`em~WfATf@kI0Tp$t(1O? z(ZZD+4oTc-)BzePUML;~-9Ol6hwkbqeI^OxTs+jgC*mvTu)w+YAm>kmr+E6q2`p~3 z?k9vEIduuk^Y$)%<6(gPVC=XS=PCg){)BB8$Y^uJ-oIZk-tY?@#v)7Y_-%N}$>RoT zAs`_9jn5Qq3w2rL&vr4maMmN%&d}`Q2v}YQ;DXF zz1;+2Vy;^DRl>=v9}+p&nYpd}2~p6@&meEW{fhHP(=^ZOc9WL6SCmYUj2h38%+y>e zTdp0{S*m~Xw;`N$Ws$6X&Yd{2O0pBLHYz`=m+aM(`en)g{4Q{|tkb+*L&YDgCr&7u z>GxWC=mB=8Gb=!IN!6UzQB^Uv9;TkyaPJAqh#uZ9ITC7bTUP@n8RuV0^S~22wYXx8 znMqMEyt{QG-r{^>6mDHZi<9D=-o)$U4dKc!NOMT#J317yCn}cteXrNhQ%2Ja7sP3t z2XuEbrP~Qi^)I)+2VE}v3$8w*r9u#-wkW%*sBmJr_yy=SrK;WNg8Gj9;Y%49=L(03kavbZR(1~s!+`RS*XBn8 zNgW}Z1uk!!&mz{ucT+=%nnsi%v6P_*k&y{`zKjQ< z*v-%y-7LbQ{5Uukg&vEFlG|$e+#Jt{_#yNd#7vmqo;SnPe`g4rGU#Kc`yX-j2k@7h zn7|ea3l-sFCX8qCjS#joK87z$3)1w@QM?fjGfS+yjL#o9lV4q-NIpBS zu#mzM^(5BFnNm*Go#HA*nw#kM>4mC&V-HZh8QG0mO0=Oa|25{D4!6^|*^Um_(<^Syv;2N$;kgisN#+Cw@)tV_EvLL-aldwLmHA-t z+FH`YU)XE@-ou3Kmca;ds=7P5{WMX(Evz(;?5z0Q4ns>EG2=&uupz(@J7I>|Xxx5f zSEJ6!64-J5@e(D&q$aK{C6i8#HfO>{2-rX4y~mN#uob*J9f3}DzDuM9`ce*+lXf_V z4{Ett5UPG`VdNiBXJiFG2A~U;yXDApM=lr|EoX`G=)U!o@cv z5%_=Z*~w~4HE@#m+sD!@Q07LzUoT7q`MwAd3!!ZI78yZWaO>bmmrYa;?V(eH$eb*f zwNgju)GM7?(w!!??VlvP6MloG{1I4sEE|)K(g6O-FyHruY*bK41~^NP6DGRISt5^PP5S@9tpK{Ag#OwS&=P+)3iCj<{At{;g2p82>YBs`JJ}%v8Bx zrg-We8fAJ?D!083iNQyrkq#282h-O!O06ftHTa_c>KLE}YqtgX)H5^mAkXyvyaH4F zCPe9ceu4ZQ1s+rzhh~JYo)G9gnVS41vh#FUmZk-eEv<$Tn3vy_bb$EJY6m&124CSB zv(({4r^Z%XreTPDO`i=yazvj#E|CejZe=y_kRa^#q2xmQF>h9+0{BAM3PPcw2y|pm**tMqQ(u(9xsq7Rxcq z0_uuAam6w|HRlFXATRMC6-#aRv}46fd=h%9hP(q>}(I^*-y!r}tWOB>XKD$MZvY)Qr?06;@=u-^IgY$d4%1bFVG~bJvzRgWfHjt<4 zJMUgCX<=@pKz8Q!@%KO9{OIH6hK72+ipd98T52=3sqN`_L$p7Qdv4{Aw$Kfeo*axx z;LR)V2<7udZX3*2{OlM0#qYZ=8_GycxXx?dj7l-u=Cp zwc|r<;~mOuk1o4zh8t+`Bgg@3y45Vov@r(fCV4NGvCltr-5 zt>eaML67Ua3!pv-K_ZS*Yi#JTs`lpSb-f1?T|Qe5daS_t8vkiKB1coxkO%bk0IKo- z63E zJBk}=EfGI-At2BQ@a2!ELgk?7ki+3`1<^rcT(F zPKfS8H@$@IR0HZn0)m=+fUIUJQw1W8Y()(eIr7M3?M-&kX=0zaRR83Fr6!VB!~R*q z=@33$uZ@WgJ)E(5pgJ^$4%8c@l)T8J2W5TzfjOC_ItApk3MdLF`ovWfIo!Y=TgrYQro2e!-gWC&_9V{(EoPTenkw1WcvQP5- zWgMaeQ|G;FLee-Jl;2v%g3Sn0&l}EomPK_XCZ>k;Gf07PDO9A<;X@Bsu!Q8ewnjEI!e=BU9a z-4?OJKF+xIaiN1AKvXtpPf}ADWy0`G+1+VbuOi*k;WGe4{{l`jIJo@oL3K1IU`gs( zFO8V450DxkzeBwk$AwQBHjsV5#F1=8gC#4@%ZOOx0)Kt%*}a4;8Smm>kCEHTEe89W z7Ei__dt$9bI}Pc5OcPC%H1pDEA*5k6l$0oSAlugF57DjS8-bYxC@k3BE3~@P2W9tG zUv`))#$E&aRI{s!CJ%psSSXf-;0!f@1Sr#m?1RD{!#tUF?U{lW#1yvxNjNp+-1i7d z^DsnW_~cJ-0%=@RvcX8Mz2D40)S3S*X17lzlb0j9H(^3jT3Gp3dZ1bydY5WdZ84S0 zv_YIH-|Es*$(_M2iv5|_U6Mj*-wHocTcn@qXNrPNSh zn>kq0A^Y-FzK}=6n<%$?QME1xImPaSmUnDHMD&2UPBw@U;2^;4k1GA8U)5PZ9#2@Mo6gE~tcKJ&=h1%Z zN+~6{aB)HD+xf&D_^!b{_2=!m3}sD}vHiILtpa$D7fr)V00?#QCaeqMm6 zloI=K=S}k^=gs)&C0Dyf-k!7;S_{JIG+Tbg7TZS>L2ZoHnkd!UHfqmJ$|UNuABgtL;?M#?{AR^i~w`5Mv5Qjef^;P0M zmFnwZBte0{9P17`7?s0jYK_Z0qQ|!S#Pk(hcj7l77g5vvN;cOuM!it4Jfb>$$}j?TgvWAOcAgsV)I1GEH+A6P&l4; z1%(TABjHh^X>Jj`sCT+|)e-*O$>u4@C@m|TnhmFBlR_m$K)62InGnDZ%yBhI%C~;; z-iHlMiUQtvMp{>-#vPW*JBn1yfCybh4DXK%+Edbc9~we7e#+M+j?B+dNRNRMfg~2g zHie_Zh%2Gx+)J8HbBVN(<{nhwq5}tFv(c?Ko!rBZHbNR@?Ib=iwFh5}@0{wX(itMp z4Kdw1la+Wq3s>t`4Wm{T%qZ|rrgZ5L!hvW8Yib@Io^l8N+||6h2>ZJ!$CD>)nfP4+ zpoSzZ$DR6M!tHhx!5m#r4{D6jk{*Y=u)`6!jDJgdtU+pmR!S(Ib<~l86M!G+}N)j2w=?}Z&;qHI4FhM_l%(?1&(;q{@ zv`4;=pA{fixzFBF4&OdV%;)KyYWv`-4 zwrrT4W4x8q@L8rBwU*J=f4+tc({I%v zmayr-B>6lj#_P=0RQ9a`=qoF=x5Tyh%>cGB)z87J4*Mn5(;VWrJrA__m0_{4*kH4-YuMd1~E55~sW#$9Ev!as&67LBgZi2e8 znCgxwF0Uh7(*0e&0jiTiTk4YLVc{uR*pR_}Gc;UA^|^H*zNrWOP4o6nXM)hT)0^3B zlZ{`}#$tHxGPNZa#qCpv$Fxa!Q}(W>%T`WsE39S{aQ-KIY$n5G0mUBMyB5!1-Kwg1 z?WJAP3)R&oX%sosz(`ys-bKZkQ7yjN9g;nN!FEqi)261>CdtzJ0J53KNAt+;^)d6^ z4$yd{TGuO7t0e`#?!^xuzPH+v0&f=S0OU5Sbz>M}X#IV}kwk*-;T`vTFWUfeP;#MD z2mboqK>xP&d+z#`_{+1`!CFjrtU?7<(aOA6*!!6vAqua7=T;&o zRO-zXRkv?b?BWzs32l))+QX=cWm3xh2&G&kb?3sU8qg}Q{LLsBmEF9+ZWw03D+HMc z{UNEspQ)448Xw-!Rb2mv&(%BYt-%w)zEz*qfX+2{J7+V--ps#m!$@k#-xR^5_=!3* z)8Xv))*C8{;)%wU`cE%a;kScljSclPGF9!EHcWDE4=pG6+V~6fTcd^&`aR{%M?5q% zkrVR7s5t#1`ezAa9g+RvGmfSUi@V`uCsxn<`$>aCTo1Qak_%|=6>5w{3n%qi#(U|U zW>{Mn*4_}_47bwhjc30Y`o^I+7#YI`E968TUJ6dsD9iA68uZq%ir5qHgpCf>-Ok3G zgU6N6STfA@TI~E%CfE}Qjc$$W3D;wYz~&#sjFRcR^_^tKsr3fSM^2I^2RiQlasETj zTUn`}1rGGCm*Rwo@2e6?r)X{dG3@{A{crjMi5(50ZNA=1SQ2ST7=D^n%O^o689N3xeC8i@tOav9hiar!)0Ij_1?D<=)&d6u*=M*2!Q}y zThh$EOaHs5D48Nz0Z@fz^?44a(W9lM70z2(>>;ZtYv+HBGdj{(R{98@dsj_lu#;T3 zu;3t>>p{?n91at2v}m*RjWKL9YLhW-!yrLk#`SmPgOcDun8+t%3toFttd&HWzFR-! zbSaamB{pG!TbOv(g^XO=eoy|Q0}DDzXIUdY0Ig3wC`8@`G|W_EPVIQz)KwIBIQ}hR zxRLZHkL}zEsouD8I`CdeniWP}MOxyb!89eT-0l^t?6KzT;QrTySLQA~CQqaLcy{(h z%NYfa`!S-F(Tw9=)Ga6qw&H@&Tj_bOK?=xP-tatyC*pVEv%cieMg{XGJRTex%1ufN zm3R#f7Ee9Oe47lQ~_JNk$Dpzhx|FS9HQ}uiYDF3NN{eDlxC415KLgJS} zw+B**Zo6fw@SrPM=%=)cKv7dPfwku?M~`=?(<<~vkGmV<^)ePJ*ri%43_uxIvu&4R zckMv(7WxBZHT%04TZPL{tyu>_M#cqVnEQv2Toh7E8#8rvjIHIGOLMaA29v!T7U9^U z;X4p(@sEy#Mq)NMJYv4~l8s}|$zDI%@*t3lSE>6wqC>z;h-%>=^7|GV-xMQ0okm#l z3vbpZ)<6=fcczXQfh7Z=*UFi{yobEc355$5de>Z;Im zqR}IZIR)Jw-VqF@$%DDTgfns8{?gNv#_DP6wqZjlF|#m>&^~P^rr!we`AV{!r!MT& zJuZLaO4_bh$l+8)BzbB@#SG24(}e0EQX}E?pC3;7QRVMGvqPlH)U(?JGz81DwowL_-PI|Vih(7O=_v$-u>2K0 zj7qG?uD8e8tw1p20iJ$gnR*2id$Y+4LqMphG(OK|7_3m~p?t}#qst4(<8mHj{-GTE{9f!|>gx+H_mMfoc6MU=7Nfc**<+h0(PCl2 zyg_?994+3WOSYEZk^^xl>a2lqn|hPZZ8l1pYL0{N9oDnOQwwqP6J`U384J?YnVHEsvgzsX_n zK5v~vQEH*_6C$?pPS;Xkd6;?g(Q0u6^S*fKv^Y$eZ4kgZF{TBJ`D{c97RaZ5! zv7Z!M>CP8rpFeQjHISW})^<@@FKft7!r}qbV#6@+N#RkMJayuPw-%Ip;n>r4&No=T z=bF5nS(tk+qQx*+{?6cgee2y&aT69N^kc4q)Q_dH>tD0+CsI^SaNKS?r|IwH{Y1I& z?B2`LQACZ?PZYGAr|an!OVjd?yw8i&G&)%lVlp>Wz8#n1i~z*yeEYw&uj@^{o z7)q8knOlfd^|kQ~MX?tYLO9ATqiAf~{4H?O-#2~HYlUhwMXGM$y+$;>F#j_MFq7IH z&n~5m*!gXdPO9Uys|^=R7x_!nl*9$bJ(O%U##yre7-U*B`;O?)@~#tF9cP(I*zprU>uXs*tNw9D^xw!V_gsH^KeW=iFY^5@Sm^@8bL!i3zJA_x zoZ#KvAa=9AzJRSQ=&Z1y+igBd+mYVrmE#g*s-{pB?BQxA&;s*(e=k44!OH`pm+N5j z*op)IzQG%7Qrdz@z_|ePu-N(XSbN3RP~gFA(r66JnP!|h6qn7ZMQ)w)zSZoA25Qjk z@c#Mh4|ks|Ze;}ma9&Vz&@E2;{d{X~rLb^5?Jp2T6gR0~PO81K44hTX-f}8#Dh>_r zbMIu*;N_2e2xD|PXKwNLxb|<3ZW)-LR67QDiNb>8eD$^r&+Xm5AQIaLprdVQ%z4p_51t8h^=L*Ece#~l%H!p z$_2$|=c`M##pjRIIlQ4bEjDWn@ufxT;)#BKF&-$~4!4kZP7mUHGw5F+?G?h&{GJV2Ej+Cz?20Sv?~?g?Z~0dpIh2+I1OxMF`7J=6Krp${(4WUc5)FhVJ0XR#9(l^oQ!Kafmb)IMo&qT zZ-41)!}j_HG`llN$u7VW zzogyyMx1doh^om3ii7Z`7-_@^w=D~*z!rI=L^#;oKcGzNZOx8ZxtHY_t#Rd*Dgi_I zQ3cx7b<*QA_dU7H%wZO4Xy#^$`d?d1_9PqxW3%KzkTc9%Hqk4Ll!Okb+6*lkEj6T6E4u!p~&ve0N9FA<=aT zSTL{TPzlRk$9+(~8wI=1v9MJFSTlZ$HS|TXT$LY>MkG@Y>fsFLygBo-LRh+`DY>Hg zO3MQjt&Sq&)6epKa(=#F=(GmK4hTS3h=v~^-Q_udnr7+FwcltHr%9fAXJ7$*;M|89 zgMcNC&8ULgI(85FU%2L<=cXh+cvqTr>Eku9<#EHx{9a#y6Hvde|2eoI=4jV)#Ld|J~+v=u0L?e~awcz(_tIp<_+t7paMvzss>vEyZBKS!hQ}F_cLG$ButnL?_;|unhSBh|_76b$pGSB< zF8I4c?XsYw{1534(m^38|-~RDi z1z6X`Rq*HQX+)-Jh|G1yoXbEZKZWS#>MDNPFPtuoM5+2ZT-6y_T^>;3m=hg}F;Lol zb;+Wm(4?h9)tN{hIKLG;F>PHq`D@~Zz2sr#c#$%JZYBtVfttzk*?NZpwDRafw@a-t zk^cnak%)r`?iZKG<0ITO5-kOV-d}4af*|Aq)TzxP)ALRvEE4ps8UyFZW}{k$Q2cut z4XGnEWllC84o>JWXBym_#rzv?{$o{6SPX{p!EGo66mTj@yg3Zt>gZx`X4jWGXXryi zWN(A9Fzln1DsIm4zh?YjK`|c@yuROcW3c<-GL8q_qPwi^rHit#ych)-X*}pjl(qt{ zI9WpfrtPM_FnCny!O{@XN&cm zUKupJp7-Ynt_3bGa&8+I1C0BgW06ONbf7pFaXAWWX6TlId`0tHjKE8VP~A1a4zfUk zY3}4kX~KNpFUxMcJhhMkgs=5GMMA{tZIQ{rhxTUb&dVPe6I#FYJ>$wB^w(~VS4s@N zYaCh*v_#Ig>&BRCtKv>=ff9(yB_C6rRh+yHQrq0hpyokB_|tQNmEYW6rwz3}J6Qe! zs?<+U&CJ{=8lEF}-C%y2MA0p#3!}f9>*G9^!s5*iXYTGj^xu|l=503=xV^J`sn#pd zQtyxt2FqDFRh1Rm8_o3E-ZmXbV}6VsB*k68HvZr7`44QbztH)qbvIFNy~whwtI|m{ zve8O|G7dwp#mi)KXZoe<^V3&b7D}_PE@BiWa$VVr*ejvacv?i?q3>VVb7o z@kI%WTEvE@rio_d`Slq#*wx3QQ*>v~S;}ddb-W=l@u^2FyZ3T7Nr1`2i8 zWR$eD=R9-r-K}+zJTM0==)=-fYoyH8Qb1}Ia^4ls>`hHQOw{+)a|k9wpss z-n?5V2V_3dpTuheFBk9o7(nTyg;h5Kxh!iUlm5SB#Mghuh)f@4VErIcS`~Kz&CGdo zF!YSQSRaN=8%0uLTJD5MRkcH(lCtKTK~LfLzzs=E;FT<1qz@sc0#&~MHEM|lkyWa> zHoKkDNso_$50P^iZw*6-Q}&+5t(P!TnHQSsg3b79!@kU|@5deGH*aV}4|TeikEK?W zNAb#&$q$$ygNyG&XYHk@!!${Bn(Z16RdECZaDcjFUXSHe-ONGuBY@%}jfLk44+ zTAg^6i{3<4$F-7sAEe|;stGjo=tkSR9=Zys`R^GSp9hc`jo8%k13I98B{`o~mPOR; z7vuhh@L0?C>OPRl;rEZJ>Z#BL^zWw3g4t6Str4>ObBBm5F1j0U(RdXPehu{&?#D~m zd$|ZMeXr-%dh=ACa>3AJg}GueQ(ga$u=kE<^ZowEYm{m!rK+gWYN@@cO;y!y&Dv_j zrbTQ)wM7-J9lQ33Q6px8+AH>m6{^IFEl7-SUhns7y+6M{e*fH$+ju1ReO>1|<9VLv z91hi$Dh+{}O#5G1AJ2j&ckw0 zE#xvDibd#Z5GFn}pY$!%;QXGsI~O|6j-ZCspRF| z(MZo#(=7_=?8%-8fC0aj12T)J!N0yV-M{Hc{4)R2T zzYo#P$7)@A7wXpJuXSx!d_IxMB|4v}<4?q%bZ16T?!)&QOLaH008_5BNm|%;sbfOP zW9^oL@53|s`b){!)WykY;aNOl)-Orcj;yqL7Juk7U;|JfH2ak-d~clkwFp*mL=sIa z(i<_e^SSwxl~Fz*@A=J(F$(0#RB2vO(yr=kC+4`g_N*~Tuvp>Gt!EEWT)3g{!cRx& z=`t2<$=?L8Q<1p9s5?q1mD@3E+^N~gE$qrWUMHfPp5a!#I8WF*-+JSBp^UdZ&SDkf zdeObo;mR`VD$+-2Xu{nk*0n#6AxiG%*e|y3ijxh+qd!5u=COyT=d_h z-WH0stY&;H;rg1_pGd^8*yD?TNrf|4Z^QhI4L^4d3wuUJ;?aYZq2N=+`epyE{DQ)b zYuSjXPm!y&mSc#ay1?@$j}q7I$7xtu|H}wR&EVZ{C~2-C{ssA&#)2Ld;KAX7sEg`s#B|(!dI!HHO7Ethy%|qU!(mz?ak!Q zz*5v}yoF#uID!M4eNY#>8)77y(kqS50+C=Q|TB4rL=v6pXRW3aw@V*pg?i`_>pDt#smZRW_SgFyo_2NRv zY+u+tqxsU2_@}mCeNH-y$Y+)e+>v~G3q36)KcM+6 zZ{5$AF#Bwl=g@Q^V{jZjrUOF>v0z0vJwp2XvJjEdC|>#FVr9sVYpj5R*TkKupr4Jz zoHu>k@oF9YM&=;3pKb3S05bh@(5<~9o=!pOL|+@%IvFw+0MQbFkOq4 zL3kOXFQLk>zF97uBbZf&lvhm;We1QJq7E~PULiZrFb}cN!@!p$h1JG4Q$xFIgFhwl zbI!U;2b7S6w)9d)tn&GsZ7t-X?fnsjL42ThDEIUzlh_Fyi=;WjJH+uOIa@r6vsb8y znj z>hE07p_x0kS}Xd4Hv<&Zr&+TP&h{_DpP}3vUMFod!QgO1FPf`O4uJLVi*s$%n5o+H zLsq)Vif(msOko};O;DiY4qb087Oo#-%YzBxm7jh%=w1=q)VS21>#@id+uvbzhRS`% zu=+9mc7$icBd$&L{6F{bkBxkxrYM7OLjRzbl}egRwJSnUO;*p%%y?51 zTa7IO{bJYoaCQyzW?RG0U0^A>(ei{Dy0O5A<1$)TAQCudp;B|P;X}Ed=3bCA8r0r6 zaf_;KhGNU81@t)Q*2Kc|!=Ig@A>vLYo{a#9iJQ$PS+A*nv@N~yJNd!2l67%(=M?Lz zpv}ED(QrY>HLF<&-FeO6IWy$-sqXuonKuX6CeZ!kbKD#BgQ|ruiVc3tF5OEJ41X#w zMOF$3K7OTYHP+<7A(nPWz4;2cJ5!rQWL<0N2r)V+xOA)0HEIiYaYt$=+_cbLVTMUA ztu!m(vu?|W)zU9#gPFZ7I=Am_Y30U@0Qc#;ku8+wBnyp>${n0L0gG=&P#8oGi;ONs zh^IjO)?Ss};H_%evuQm;E@3&Hw@HKnf1B?wgFy#|>lh8!+wSQ713%ey4c1n=9sIz8 zg{7M1jBzXp|5zv_G_~i|FAT82gH5n?GXA-w{2x)fSL(zp&&2T5+cH0BPF*h#Di+LUni z@}j!-#lTGjtS`_efe1$lA`VkTRkaUKmc2})MWsB}53|ez zN0{^MjP-om78XFK8PL@T@(Cnfq`XPHrXO&sS}r`O80fMTVOqr3m>_pyhxT2pj1ZsXrK$e%~= zTntMyc#?fsKV6C+YGy~BAgCXZy&3g{YYa$fHQHHO_28n9%S?m$nC{6eIf$-_w3q{F zmfcY}j(3^CNr_2$35gz_`lb^%1Tlo!WMmd&p40l(8~IJ{ZW)$g(;yGD+%5Ebbc$8| z8@kp)SHa@BUsT#9SrrHV! zs00cS#I`IJL6IhtI&SVWG_TEBDTxT-6~ zckXxBp*V2hspGhL?DAX;HuUsDXZ!OtsI2qsYJFK80~j)j$icWXZ8Q}tP306A6+;sokW*=*YQTl0H zH|kPfosxwERCmUn52H1oUxhjDrEWc$S(BAwZ2IxsGWarJfpf(!|K0uC-0gBtpJ!fP zfxq_(S3V|NCfW@aEay3nE4T`jj8E$KvVqQg&=7n4q}J+#B^ATB8=Cl*wX1QLXREH9 zccB-6ho@tkV^<)8 z;83^JG^R4o>zHev8(!gO3*;yps>OH+yciBg+Af&Y0rGo!p3H};4l&lmPlr(}xLtkX z2eaTEcJ6rx%16^0S-x`|a3L>tlT1A2l z&9R*n-7&`6svI+(0WFjaIukZX+Rk~&T(@(C%vKxR70&WlI+a;TSr^+Jcv!-Mu2GIl zA`{&JHBbypAl@!7f9uA!wsCM?sU6&~TAPCci~^@-%Oy$l&PV&|vn$r6fe%4)L$wIl zpcrO&ZOgXkz!6F(S6EgijQOHr7L&%VYfsZi26oI&zaw`%=~7C_!WAqVRybU#^zNyN zf>NI$0#I*egsp?g*U@{#W5uILrSg{G;f>p6uFM9Udc^?8cL36PKZLg*de*H{nhhGA z4o2#4bxqeRee=FbH`m>-c~5XyrmaNRUrnY1wqx!tzSAHB!P1+*Krzf1=1O8{njj4$ zxHAXubVT{~wPaMv4$K_ARXL+!?fMcGJ7sYMXgD^a{$x4n{RkseEP1}Q9?%R^M#U}} z4+s}kl2K^fc`Vb^@soA^SR0aqQW^PA;9ZqHJw_N4guXoLAGMyTJ;Xno{n8kK`@N%Z zQ0Qsqadc3Ol&kSPmV_TD2h{yWb1r%hd1HUMSV6cL7dIk&(Dt%D}sj_x0{-4kIt3(%o6tU^!3#LzQU0);JfD13R^Y7O_3DMy7# zNUA@gWU8We%`ZQ5dGi!z(LQ-aK(0GDm*V}H=cB%w*7n#73G1)?8Mv(jLDM>Y!tBTW zipyW|o`5V+#he4xqH9*#+PTioS%bLHH&$l3@R3Ehx939Xj--7+Z^b~X9~=Bs&U3P9 zn8N>#6Ln91D0Fi$WN|$y#8{i0@iLU6)eEZZw3xF{bDmq9kGzifa&d3nesQEbD`A4o z_c(pfB4EXC^bKMR1J%h`;-=Fqmy}a=k+5J&DI&@lvBPX6dYH#F*HN=c0ct-YF;rN4 zi=oh?lnS9X=!`&)QYbk!TeL&K*m0+6c?*yK@z0?th{>A_1*h3`bFak;!NB!R!$mJ+ z<46b7_;ZWVos2s8iphxLWU|jrHG|qZoWp7Hx%p#%q3&nAqY}w$IjC1mUsL;hir@@6 zQDn=`O{JcSN@!UbP5c&gKEQSej>PI=nsPDmCgVo4hq>e(CNPhNBlx`gjGSjN---|Z zXv*wOt?>*i&S(D;|I5g3;`55G4bH&PG~=eOL(7zx{JA{0Aa9g|Co~x?550m3deAVp zz<@6r)YfW(RbW1BIZX4UNlOltckDShpepa3%)LYIPVsgcG2xlQY`j@lD_hShUa#(I zzY_l|f+AQ3T8i=J@4AQ3RqBE~bTJX*wA{n{&=sXqYmy}r0Hnc!Q(`*{{- zqhu!boQ&uyY2EJ@?3%STP(He{O*ukb1N^3xJho_n?!1O>Jj`jALHW0*Ykx-g`F{Bm z=-Q?qzNz)&b2q7}E=@{>;@p>fVZ*yce4mBImE&n)v2>&joU{JTit{hi+up^M3F3g5 z2=9-*vb(LyvEi1ZSGX}WMUqYTL|DE_lv_G0nD4?s%nn>X1Q+(TcttHX)0e5Q( z-CwtKrSqBznxMkj0z`6bz;ZeT+g#1Pvl&%;FnR|tC>ZBV3j@>ctb7`Re3oigwgicl zDH>vI3s{Q9mP4x@pv3kP^muVOKp}=})j~AD>CP7)*f|+DcMPfvI`r*aPeydG(80dp zxCLL96I!VtisXIIITlgvuv@T@6Xm@c)U7pTnsF=Fatv+1GwlQ{!1uB!)QKRc4;;GX z?8(Ri8W)^jA_FwL=RP%>pnpFKfN%!8&2VqykJ|T}dj}TzW{`;O+gs`7vEnuxGH#6s zucvMiAnsqMgg&NLRciO*>A37=?rl)cV81KvA@)}#&F^koL5PM5>aveK@a5jf2LJGN zBDr}Ae;i1ir!Gl~L1**bTpNk?C8e74{2su~36iXH-lqEd0wwD-^C9zxwAbu`6JW#_ zsZzUK2;dqCSq|yAWERiBR4-iTsgt3a-JH5&Yfn?9jdGIL!JUq2|A?aRNmcLGYC_B~ z9yBmm+@Z3njRMsU(2 zhkkNvTw>yF6`S1bEeQIR%a&l7ihAEYa7EqZP*!N6p^B_2X!m?4p>ZbFwzSq_O0TZY z%qO3|dE?$!9^9El*y`hIV!YPsfXf}K)X8iWmIwqRdts<$#_+&! zWDyffsJ!o?0}QiK{RZAV%y%{VfjITbm`YWO71Qv5D1Enf<(eI0msOb;_0LITd4=guyCdN~UC&1Tcm zm693}YglcnL)h7KAvPZKT^?SrC)@5>!oqFCJ>)S#O=-s_rO^xn()E&0SO##6bOTPb z_g>9$E3%i$#4^_sy1@OZ0eeD13VT*1;7DxFR5~CZ2T5)GkV#VO}QdCFE{Ciop@KJYpDJG$T0T9O{O)Pu^#=#+FBgetl#Uxq1(RFQDQmJ-*$TNd{C3qz-KIl2~gu*F#3)jqdLs9V#9_fuD2@639bgpq!Mh8Ixq4gGa9tI->G{azol%hQWkHfi>U_ zzbK%rYWUu<8KeStdi^hmH*^q#l<0!+s_Lm6j(+Fx$^(R4 zf2Qr#YH87jA1r%@!_+K2=4=^_(RztFc4O|VA8lfI#gnT)$4o(-VORrw9S-W`q$vn^LW=NDshzH;mhImX$yvKvJuoR2Jh_ z$XTtzrCiQMJ#v-^0UWGkr~L-%ZTXR|bcRUoRN}?KG)bL`PtEwz2KqM#n}60(W*`B) z?wU&=>%u;}`b@@tU}6YuI=~K`F)Yl;sCA$@ZLr^=!H6-IHorj|=d@)I;@mCB9=u6x z_Mq#?`M1Y`-{+}G2IunwumnBVS8*}=)6Cg7(xT3vj=EWc`>yu@$!eav8;Lrtqiw1g zU&sCz8UMo?e)S8b+MMjkN>P-RV;Ir==pQX%SX@??RA&}oXEK#8Xa5GEXcPz-HJazPFcA<2iY#R`@ZT4 zu|J|1?J*>K87F(POZSf4b#!Je029rS&c%@K6O(8z$}B$Dv~#rUROhmA8zh@WKGXG` zXiBiTxto+q-sSyca_jQ0zB4Dpj+;;xhijdtehu(xm@e>bYtwdvJ7Q4_VP{lf;zosw zcX=s{7Ku8l2O3vTTpWz@254c?g>+D#gjDQJS7-R{Sq68I1n_0OPAW8J=ki2b$~#b6 zNZ!9?5@PY@sBI*41(Sm@U#~H0S^st8cm8}pdl-jPS{Qajt=7z8=2n(4V+`iN!u6tQ zyQE*wd@hh-TL|f!RY!KOnqlP@d`rE;cD9$a6^h`gaWf^9(gIP*dk@`=1W0mT3m#6SRgJUA(^ zm+Li)_nvRo>dPugz-D!BGZ(i(@Rviv$X%3SL<``C_!!5-tp=LB|V!1&ws#?6<+`?3M&vu7C+9vJsF~;sDo$C}n-ZkmJ zxnQ(H7t9E&bQRscrP%uA!}L4{hTwU`kV!vz;t<_eWUR%kNnZX5*O-_4l2de?34rNt{y$d(Ll@T2p8qYV~MjMcm_7m<*J5J-DoX&I5f&jrd#_n6uv9 zGA6)s?A4u~ztqQMd>+m(Y^fo!cNii0Q5G(n5mBfLJ7J=yS! zzFCx;0lw382z&CF{9sEH&hZ~%}uV(8k&U&1yE-DOP44#f<*wGGT zP9%%zb+pII*}2-3QQSOl4L|{a38Oo!?+)ZgAbH*+*`-TMZ`Oso-=Z$UR`CHnpx(x* zf$I!lg!71ssM(y8be>bg$NOM>vOR=#rF~y0DF}_0S`16WSJy_K{ z3^WWQpB+HkSXniS=4OrwIJ?D(7gR6lW{r7!d8v5mPZH&di9YihS!`5Rjfn)U9-dha z9W!reF@IbSEOxz}3bc~9dwjRXWn6WtH0zBke)!1$9XHIwbK2sfFs?gY zV7z}```k~CY;X%b4sE>>o|eDAb%HM?+QGp`wzP1_5(uP*cwJn`y!vjQuE*&C7_nnA z2q-bmKRhT#nfqh(QG$g3Z0U{yy66PfwKHvr&cKThg7&B?VSkSkzot`W&{>D$ri|k3 zwIjI-*D&0CPKE+Q>BZ9ZJ*!uw4m(k{1u((UF#BulmA=F^mjao!%+Ne^ko4Q}L#cPe zq_=QoC}gQ2v_L#U{}8{CR(COcrp*GqCXY|fio4#;t|A8uTCH2;K4^l^<5)^bhSn6j zi!Tad(IqWMO#`ii{`_X{(jhNvPhA_DfnWzAemlpjhi2N<`!~VfS6qimC0owRJnwqC ztj@x}B3Og=lH5(u@&4s|lEFvC31lnk!1I~s{)Z>MyKr%FaqzKu^@LAOOYqD0ZR26j|#2^;K2z0eB(k^Ry1ri;2BB}vsx|x=KY@YfZy)t$vb1%B(Sq`>e121Wbd@l!Y<9$ zU#m4!G9uDWREG3ou2yN1t}mMAP4N4&$G_Z4F*acoZj%yIUQtlLjaEnhF>eyr;4zRA zf4>KpTi#&J&Xd^~XDx~CxVRo{M*f-G6cj_prDN(kFDECr* zF%lDm^Re;kuiDL-ftLmwdB9+N6Ii$n${}8~Ma;GRojS6Sb~81#MWps3{GDO@Zx!0N z(BoU6<`-)&WPiDVimgHaK}TH#YTXbsG9+fyQ<-_Q)d1B+xl3M`JQd-CyE?Rs@y6Wa`& z-eEVY#otlz#wD&zwOt8Ka&-gwZQ(v#xZkL z!jlJOEH?zb3)x*uk&ad6fLw>J`!Y{tm^ZVAq?s#9>0$Y%V)cRFv&sTb1D#ZNTp2Iq zHIIE?A9SC>Zt*sid!94BP!tN)`|wYiV)WXYpCvTd=(&C1M5xiBH#Vk>ao+K~tRgIT z4FkcLiZr?zLf;Zm2@U9febZPwvgk>mgT1oKH}zbvTE-N-gXI&t;RspTcKm?Uwf=cB9<-Dl?Kb9>a zLyA8Uc;-t$;)#0o+fe?==isDUq4_Xt@Hca-M}MVa2woXqxgzAsRPWMP%4XSz=M|K_ zVDf>;G27Kmu|r(tS{df1=teV&!z8w=)H&b9CQ{#pk3N`QVMnwA!E{TJo#E+sfgKIov$z3}$mCoA`L!l7a!q@GE({s@RuH zdSTsdNQ#i+WqR~7zAJ5};Vl`fGgs>$5Jq!>b9pmg*%r6-i_uym^ES5(^EyPBzJslD`bU$?Dmw z&3*YZVzVlL$E<%ma@_T;0M}sOOc%RdO7{@*YrnyYm#x{dCp~n=$Wl)qm-vw0dm6?j zL^JrL$(FBmS5gD_fI*|=V<0pls(sSj(eRzAc2nBQd&@=HILEt*i(P>nN0athZ-FQtM4FLbkf z`^BnDb@~kHcBRP^CzZM2dzjb1$~xdp zc*{tmRg|i| z%fZ|whD9B_h1(ns!V%b%TjL|YG=a} zB}Dz_it^>Oex?@h9rt}9z5PgZ^FfupogQfw;A3Lqf1mt6A0cGaO1JxBWzu64&?3!9 z*}EJ0icw1bL6~Z|-P-R;DP-z(RG=P+QBGVULuJC^@r~nUIXaIGy#4P(mJHpw+m-Jd z#q=V)XEH>Byt%Zx=Z76$5Cc_o-9#7+T?9nIX7N6)15XW%9}ynut8nNDR+xe{-|?YM zHssoWEPd!Bx(Jp^CPBT;ID>eqChgX=r`bg<51LxIpHR7Rv(bjoZ7VaV?R7iM?p! z7vjy4#KsMMbu=?7mWE;az^~QCKTf`xpEC8B7@9mtwg?QB-9H&i!gq9h2*gja^=~Bh z+f5;=^CxP84Gi^P`1ye;sjlWX6&i^N9i1{*CJ`7W&+nBItd>3^_FHsqVV%yF3!}^s z^41SZ9Rf9M?<%p1N>s}$Xv7~-$lu~iphL)B{rvgNa5iqC!fZigsJ8Mh@kic8VGc$z zVx%MWC~!9H;hpl`(H$vFkVY_NQCm%GNvvQs>%;2+xs%i7s=aE!*NNL?{9VAktMwhQ zlU0{ne>;&^LTj$78{LjtNG;E@@H*KoqTHtL8zzeH1YSmbNjE!Fge>$ayxtI@B-$RE8YiJyb(mj##OjLMsE9;MYIxjLHU)lRjstO}0vy2y_u zeb@aC^4$}n5GY1Z@4@X<^-ohNH$?ySvVXsJcJ|eUb8Ox1!;9*J|1NdC8oRO3X6ttL z)@8R%)tWy>HwhJ6gCO>uMPRiMV#-fWmVcBZei+-`35++>&O=W1Nv_L z1XxFi($#l|f05d2X=$qMzWojLG9eHY3fzJ?him6cY+k`EOArFd8=kxm#qPzGbKz_k zSF(=?W2S3I{XOGPtVx?pM0W}5>2^(s#A9)Z@sCs`B|Po1XuQ>ct<8jp)Lq)3dk%ew zp05XGZ2kBDDRnpC zZT^^pkK%gCftL%77peVrm{f!P0Z+jm_!Kp`b@Bx1h-n{tX~~;x;qiS-HlXPwP(DE9 z_*fetY5QT#@R!>ZYDW8lX}}4`$@XV5j{-25Ph!6j zH($Cb1+CU6|9cY%T7-U=RMyZ?fAEV*5a?rVT6_2Nd~`90QC{Q$YnD+#v)<|g2kV%< zN8iKC6iV&B;mvjtsoN&EllICf*8MAo8dcmYmls&q{rnX-1j>`>TV`!xkfRMyvkvDc zE;}dRzP^(Zyb(NKQ$3kMXGqj(!wo0vuhqTqEF&PdE_u=uqX(|fawPEe00csxW%d)A zuYce=g)LU^A3fJvY%v$(--6rJ^Z{KM=aos@ zPZN~lsY^CF?oYb{xbi%l8z@rB)VUp8e~%Y!;b%EUeQSH~;uoHege>Xz|8TDXyC$k( zm`iodKb#VD!L6&$weC1KQL?cmrXJ`G?ay596Ot8-Nt4tMFJo%!+T{M$dLekyEu#3o zNe9D-%oLf$(jE_X=bSGP3X3RXV@`nv#nCU2i!FFo(!JJKp@dT}wKG?JPtW)Ns?EQa z>Ccm^xIxS`AWfedhrkEe2-uQWyPa+E(9U$+HaR0WLIQ9p(`WxOK!v0{nw9wr@55DX zA)a%#@>FUs+&O0*9Xf2^@rx+sa)yZdzU{bvjU|x(u?@>!bvU0FTI`B}d*o-bBJG>c z$Qm|k;Arx)PJh;&*e2i08_FL0(3^Wk_&l%L-Yq6~Jx&17u#I>oPbXtCfs~ z_{eDT^(N~kqhzFsoUXpiam~h>k+ue!1m>}tELZh?ns=h9p!#=Lnf^^!c3XRDL-Kn? z__f0pvBv~`i?(%TeMeTs%A)1$6Q-)!wP{bbN?HZSF@)PrsHWw6s5X4)Dtzc5Vol9* zPr@-*rEe-xy@4WRci_D5h4bvdU#V!%s|64D=h$UcmpZ~e`hdy7^t-7CQv7zC>c7gL zyfnCdT&vllMBZ2a$E{?-4^^LrR7R;O(6C;pMu@=}-+G>`3gvNv|O z=gE1Gp7F8upy-GsH;Ev*-ol*k@rQd2D>YtoD2@xBEXwBFHP1bGjE!CFEXw9AI1^|4 zexLYA=gpGBW>#7iSdz7Z>-w(eRTr-8pSN8ee@JlWOHA?+|Lq}O%uLLE;~wsp+DKQ) z%ngoej#%g*z!oE}R2->}A8N`w?ZWNywR7$Q31UFS%0xjfzJ~Lf`6ie}8{11M=l>Ie z(q?jnGl-`pFNlS}J8gZ-Y9wkMiFImuf;iDl0mQQX^~IL6qSdL1g&4VW;s-@`R;shpmBf`M%!-kG1V8;iMDLK z)6NWX5rMOw%uy6K)vI%MD~kmB>X!_ae$}PBBC{a*#;$?CZpT>Y{wI(m^uuQM~L9>Fj?q4n!rt!%3s?%V|EqApu?Zdr1 zi_>_LDval5htVB{gRcTIQnzUbz04Yal2UYVR|Gn>xSWR93X9(FHp#XN+V? zaT1RH1aY*n=hXb7e2cj$@#8n6zkLG^7MkQ=R5{gc z5g+|aRsEy1{<&jElTHSXl2)hE3Z@v~ms!QY##5Io>gyn=G_pfoeB%;Nm8MF zXZ5_Fb5rShZ;r$V8ZrjLuHWiOB`5!8{*U=lv??Y&n*bSO37M8vi1PW$M2=*WLr5xF z*b(fB!*;lBv-8su1b^~A&~6kT*h0xB%=O;S>}l@(^}tHW=5P+K_?-%X+xTLm@5;2~ zY^UW?eB^gEit&>{H7~H*O|eIzGY=@lPRc^2bEI3o0Ao$tyjfaOZq{J@N_51mCa;@_ zm9ouAx`=2^s&&KWbK1Tb3XkcWuYKudLb8Xg+M1Ij1aY?8Hm;mV&O-brVBDY(+r~}4 zj#HhB-7i2&@?)}mYiHvw3b7tz5}BzKD@QbXa^Wk$V)BYxQ1G~=Fup8Prd>Hp z@);U@Hs@`hEItIbJW?<9JD*TcaJZVh*})p$TT3XG5^;KOqjQ$c-(1cedrH%Jd17(^ zR2UU(n|f|LUug8AHPG;{KvyU@w4SK+T~|`dWzy@aVNx*;=SZ@u)X}$E(Wfj=IQ71B z?`eV$#61N8*>MB}*w{`j0XVU&b1BPTnrOG8b~P|Y(J+?ru>(wstw@3-?A2dR+`kmV z|LAn9H^gdT##t;04D#4iO)}+g#XOp*_o2z3Y~L66J61_n zM%66+@IKwNlr6DfduO_=e-w-}t$1B7-Tu)ln$qo!@oD`hW0<17d#*bgYT3tNC1*I1 z=TFzXYEX2rmVQ2V4r7B@)n%A-JBhD1H2dhS%+tURudm10+ycHtx1C)dF2Q% zULdt(AjppPp{5C=hK3P+|B_&4+-Ls$QQtwYD-d{WO;IFEPYKa419XCEznA>5^VK4^ zGz!}Zc2}yvT8&x4ASslmdF%jL;%vn_4Aeb9VJmLQPr_Mm7zN;&JpK-N0gVFKOtPam zP_lM}6YH+Y8#6)rWZGRnXlR4K#No=;5BIuWvN6Amzhvcr!6C>pS6oP=17&(pS9dh8 zbzsS@@xY4oFtB=@axu?tqPu|H>PIHuF9m-K+6rH~<~#0;1%hQEb-p-XveYxq2w;9f zSyn&{v_{-^^Z&|e!!ysQ7#z@d~B(@GHR#ujCHs`!uNbr9XSG<$xj_=~Ep>v7|!TUSJ>7AaYo z#fd({BiLUI8Wt-7ZRC~-v|-eBv+fh)U$d`H(& zMAUrYlm)B+2j*&wnQlHK;nkSBt>NIHzO@Bq`eGUk?)^G053no2_!U@^S!7a8PT=?@ zqp!Dd6d+nMZ`$6QV@MdS?Hs+R7GM%XZH)72(G=0MliJQB$S^jpjD6C zKml1!&Xcv^%%qXAWsi-*jH)vqCcCXK6in`XanRW&P0mqKg3Ygpyar&AWPBx~S^49R zd3if&fCQKlW|+iXp&Aoh%AXF?cl8d&W!GNP&2v^xyLk4D7jCx6_iLl5^%So4t}VUe zU8~>d(q)tpW;rYdQl&1giX1dd<|q{~b(Gj$qGL41j%p2QO68|oz2)Hknn#A#?~Cx6 zl35thv8>rPSor;E% zIccB=OY)As2Xj^~0nC5ZNh?|tu>^K7g*@DIY5uG@L)<@bstP%Xt1+qxx!SC<7i{lO z+YAlSs&cqF!<849BCD-;)s~aSJgKZIF8E~jTnQkOLpfi)2kw}Xi0Y&3~-rh}N*P+6W=P-sV z=js`TsnmRe`-qjg)Sq-*U%j`+(K^d!4{HK_5kY zgt`fLSBCU21d%_$W@6#jZf4rP4#9R%h7o!O+@QQkW@pu!eq-Qv#p&$a+?<2L3TtTk z?cp|m@5{Zm6yWUM$0J=AkR@7K3Rt2n>tE|zASNjpG35-BMH`PC$jJ#-idL^8w|~g^ zJsutzc@!UCjh?f!%hxqG2Zsmr3EW>4AOcM_ZW3r9hgS)l1F_uy`#CZ=C= zfm$qH7qGvaCJqWBp`}D%-|e0@kB*POoPN9|M=VP0>dIo>^@TclW`>EIyEoREV=3bo z5%B@d+KyzT_$L5e&>B!227exT`^j0So}ox|PjQUBe~`zj%yay4?jNKO9TEPQ~m*E(-bWj-|~^~sMKW=sQ4DSq(&f{+Y%qVH+-%!xVSFtYS&2>8lfl`U_OTq4`Jyj z35b8JtoMwjRJ^)ZT_!$5!sKa~5;=B2LO`XZd`P^0l(C~^Z0kscJf2WycSk0|fGu1z zX|~-m6(LbNl-r-sYbHA*%u<=5PE#tt2AkiNW>JDjYbkJ;IW5TgG<>mY9pb|?H zw~{)pcd`E*@{mVX=0ef-Z3b{(5hB*{@6n(>DUQjqVXvgaS(T$ajWxsqA*~i9=+9 z*PY0gHZDMkLEEn*5H|=vN4W^c%Za>uV0*Pg3p}kj_*}~FQv|bu^nEy!s98KWQ_2&h zBzNe93N2H5TqL$PqE*o{+Ne)}|4tSlUsjI0eJHxmHW)y>1-bqQ-umpCLwo#1axv|# z4yOuzW_XT-@Qg5^rSA_+>Z>I2S<+ydGPq8LkVFD|=f*V8D|quMkQNO1SvQX&=lK^R z&#ry(0@zHPySndushibQy%MT&dsvekF;oa3dj}vl?kK6`CV-6WtvfTMPmKMG^Yghi zoB_3eV$Z4}-r~piO3ljG!J_jGW^x;!TH`1*wLjH^e%Rt%o=jbBYzp9*;q!#S$Eia* zX|H`enY}gCxX$L3hYko6IU8_jHhxgpy*4Wu==z5UEl{5s;2lrCR`{BfX>a5_*x| zR0IS>dPk66LJvJCNbem2q4yFnbO?MK9(|tYcdqw**L%)?-#PDp`$~3p)}A#p>z9MtpPojhfJFv_jUJigy8aTpBy$&$4~A=lA+kJaj+t3A}1 z#LGu5WQ-!gPdm!6Y@CtNvzx!Xan(;NqC=R5I)x-6g})uU4R@3*xNRy3S6*;(q4b`` zf`ydRf|ArMo@wo%>`9{2?wqd0b zo88B53lS@jQI34QK@)bbN2e%nt@#wb?Fh#BJSSjdiCR2b(pH96t@b=j}$Q zgZi*xBy58>diqn>R)>OFjJL}7>Ca1IMc~oHdjpBZcj^j0~Fu>c?tuk|0=_%2?5?H z`OD@zGU2>g-!ld7@UWac48g4xmg|^cela>lLJrYZN8eFnN@SX|9zn7y9SL*a`waG1 zP1{aXjLBLn+l?8*0b9AAndYi zwx3@AXw;vpCQ)or@rbwe%vK{NLxY~)T+9sLJw*8$Y@|9wN4xL+x$WP7@xS?u$Iw3> zeY)A`!-qb7q4v;lx_#-IN!Z>j)qgR170iOKT@Y5So(t+>tLf}J*7gOEAM7I{9kj_R z&i`@%|K&RKe;JP?iT}rQ@9eZ0=fRM?wRhyy{Z=QZF-7-Pq)1w$4S*Un6SLWGg3-S< zW0}>6x||=Yzy;MKXm3eefJmG%;68Wse)_Lk^FLGne}CmKMQXo?M6E6)yd!tL5Z^k~ ztWB02m8ZQ4(&qabzp`OrY1xSqGdU!dWJtpn7O6Aw^5U(RV`5@@rK?Ns(+hCzBpKpx zT3cJc6%>rANJUgs@C&D+h=HNDgZ+UPo!8$#os-qj(0CywrG4byC_a9n)Mjdd+?kqA z;*F|M>9Ur#M8)G}KVd(6YvwT#OOhuNF;n2Qdzn|UB}=Cp2$Lx)Sr>iu=sp(Kv$>Q} zSYKDHLoYe*941Kafjz+IJ?;_J^2rn z>^DgZ8`*<%^lK^!zFWsXhYnm1`YO)hhR!8*ZIj`F2$Z9@hm+} zP^s)q6T_y{SL2gom|fj6sS;lj_0$fSEO~h;V>(Rs_Ks4Fnb$-&=k8HBI?N9KA`#-vl3WZ&-loz4ddFl3)mKc0|vt3d+R%;a{r=1&8tj zm^#IcpZI?{hxl8wb}V){*p8}xeP*`wlveNmLqod_n{HJKUBr~!2f4EnPYbbfyM0M{ zisAq1`dch|?b8m8TVcKt+haMeS$g5}GvWolY2lR_-e>&r0t{upp7qS%{ebY}s9$}Z z{uloGECY&U*%^59+Pxxk6PPbDKa3P5{8@(b&`a9Xo*m?=r6^In-*C}IiN-o4|5n6| ziD8%n>ZIGbp{HKMQ#;1ytSjb~TJyDL_k)K{q>f_*9dm% z1myw##vd2xPfPjnY>V34`*9n)(mt`)bN}|f&U?1;3U@y{fd-C09l;}=fun@_#tpnm`|lN@a7WEtu^ilh)uuDHU&OAv0xA!51&fgUzx`p8Fxb zP`x(!4{o7ex-zJdg6k(WFngLf_{W~z*3^|6=ZaCosfMg=jRVZ2-lUO1mdvCDuwHX% z3ti?VYq@)dD}VeAhri^apM{(3_WARA&_qN9A8K2?AG~;oZdZTIbW%VoY|oE4aU}CM;J9%UTl7m zspF>Z7O$7ZeMiip#r54pUx3nZu*4G!84n|*3a zL{F9v6qp^buQf(sHJbomG=V|cXf|cL01beQqniM?`U|sT6F*?UoD9QXH7YN7%+>UB zpH@Or?WRAKfBW|WqbC^tQ@3W_z}vJ0R)6w&0FVyHz-r9N{8Q_XF#Z(NsbVp>oGUIP$Za2iF4n-Y~mjR^r%Qa>(2ZK>?WyQf1 za;5CNEKM_=-P&c$*k!Cu<^oBzg*1PxLXyD~bmJ&hOm}qC6|)yL<%50InXKdpbhh=P zt-gYZ!Na<`LOA7Tybu;>vAPH2Ucr?~wKHK#lXhffKGFYePU$7sx1WD22^aDVy;4uV zQydE!$BTz1H?M9eV<$d_enFjYT9vzEFLM(gYo}azX#J}!=l_$OfpO!qEGl%}K6NEK zQ%avw(+@@$y3sQ_b-!tbNQkcJX)(Z3c<;40;Rd`n12@iXh5Ot$*+u^NC$3xRPr8}# zLahFx-YBPoDrm@8{mr3QHo<$J=V#jBjhAKA*c5C{K@ML7IR3$npJ_lQCOu&blAMkz zeyB!0QMO@LA#A_{(6ssMEM{Q%ia74(0^C_(7J3C<18(YX?R=#WK!?r19Aw<6PTudxvfjh z47j!P{34n&Ud&wP)(cI)n{oiWHGBlBP@(88HB%xb=-$d{n+b1uQ=iBVjAgkqGM)aP zvEXn^GRl32;b2HM?Sq_PZ59kp=qILlzh~%=C&brcIH?!sEvfW-J?`>bf&OumS65mf z|C;Zfa>~~QgC!O5`+d@)g5Jya++z6bLUf(<>Ty#XJU{Jtgzkv1hXeb}``cgb*kNK2 z+c#j$7#B;t+x)#^;4n#wym~@s>q>E{Z$OM3NvVGXf6V=y;wiw>@e(Xo?8g0%(fLbJ ze6)$uUJNULjO{4PP|H0y_2j@Qzg7*`pNZ@veZ1rKf*;wpUN8W-A4F^LmorJjV0o}D zVm1uyuAC=$F7%@%ElTrJlvBhz{q(UkcmX=UCreTmbmsapI5^?zF99(KKMIGX;GCHm zjXCK6Yn|d9dbXh7BmYaN-Ov1bDN&Nfk@e&s&qVukc>r2mX!ndVl+TMxc~BK#I@CpwlrfJZ5Pu%n@pR> zJo4AT7|gM&m%s1@1%xmV-vzP;KqBvN>$+C0q z|Jorgm-{^jIt;&{AuslvG?BD1e>LM_SO8UMnk7e*-}`*me|4sGppUPDN+a+erI3up zxxFLS)l`W13A)j;HXkT6QqCnQ@?tc2!jB#2cKl#WP!R*W0)zDvdRJHK-;)mrn-a$& zz3*q{(-h!Z%)dR9)eDl9(>C>6+9lPXGYxG$yQ-^P!K}3ZD#}Y}8_PwS;R5GKKk1?^ zUhF@LFi%Mtlp+1$k4}?nF}4`;bRV(+9_e%FrAL@o`KoyJR4P&<=g3bgW%vG=I0KIV zYAEYKpQ;v4`QHK#y)SuiBo$!BclDcM@{f(JKV}ri=w;bbR{#S=one%F{vM{&ZWNBP7QX}Y1oHXW%|bO?YM@*X_As{H$fV&9X&TLI3x4GPcy zO@{Q_z*#nU`{@k4d_rgP$1gGTVOLkjnd3!IcQ9fKXDkkUH__DNY!aZ*yd%zLs0=zQ z=FDyG>HYnPzo$aE3FLIjOwNN_cSXlmZbJSLR$byC00ItEcY;ECK=$palFy!bA87O@ z0w&_mL4XayoP%n`Mrs4OE)#X5iL%zh;*aZE+R0HZ}-` z7&9{mXCxjlG5vg{t-?hg!IH_aKuS6|Cc$7(j#$r^-97fA(wAWWcd+aKgX}g9P29Qv zlUq*ss_NQz{WCI{0{ahA5eV_!Q}n32dKDXQnzy)1t-isGFAuxAm96cUq9h>`)9J$D z29T<%D(@q*=yb~q4nA-i`I0G>o%IVD^ZtsT>*hIC*~0Amx9MogC)KU23x1|aC&}2q z)k^hQe(&tcwB(NU^QYtP`;{zF3BG&8N=D+oZu3RfcZXP*SuXdc0m9GHgQ*VcO)Myh zvo9={Cx$r=;H+qGm3)a*-aL2)3frzEZMs&L1{we3&Tp?dO%}P?2!?5 zOW)XIeceZC*AK{k3v}wPvISVuLyfuTc&%yHSvTt3jEkM~rALVcmaJ@a9N%K6d+Pl9 zX%$=#)-V2=DSMh{$n@@*%%i@py@B?=eVibfZZ~8)86If#^ljb=y^c3Wk4r~>NJs@7 z>sa@58nQm2fuWiL*e2$!bRAbR(wipf@dD`NV*;^NbY%s^~`q#2$94KAS-JIa@q zFMsvQ2d_i>y+==`FJ_H{1s@{ETvHpsPgcTF2p)pvWIv)e`@o^0o!erki#w4EB>3-t zxC~c)Al%UC>UDB$*IwpJw8hYQE5#~)8|bFimikr)I6s7q&RkeZq{d=E64m-a*wWHA zx(l%B{ZOWrkQw#f8GVV>xwk6Yhs5BqZcLXF5nVjAmSa{(z5NHAbC0 z;aIGX(Uixl>>eYk<@loSE2CMYcXpZE45m}NxLl)S)~1uxnBxpWth?kcR*#=g1N-DB zA>V^~spU^#3q5kv_E}6SA-#f-1*4=I7UDQ-1wvMrnW=I&*n~OE+OS%A)a={&jFecNhjy_9=;}n+Z z=>s7|Jx(Y%rBj9b^}tAhDY!S9tO5)ltqSwuhBV|0GgYU4DOzDX>#d_l|Im`zzg942 z%adf2p#dxYA>mXhMr*aPC_QlymT#EbnX#Z%ZX2kAZXGaY3D9SqNIg83@cOtmT!I~Pe92n1e{7%{vB$i@|?k1 zI*|j&15OYt3nB$J1QU-#W8YKrs|kKqPxt{#-VG#df03+jC{flK&@dYB*&22L8ZL%- zFwo%JU;&xSd+D8_J-Sd&!WY99U7U83$!#jG9g;u^C00+;)KJZol0L*qEX^>o`zd_D}m+CL$FTeC$1ll0Dlng2iiY` z2l?Rpj7c1#^2`spu*)dMXRvbq%G68V!^Fw(Mon{+w=j3V)R2``OqxJZz|-Sm3|^Hz z!8&2qt#~!0KE^kO+tQs_$8*iVMdNaouc&&+Nc`ql;OmsJ5ZT z8)B6EbTOJvE87hSSWVR~iZP1WBbfE8H;h(Em_YnVL zzar0Z&A(aDVbgpF_Hg~t18o)fGqmm-Qgbm>3uAVJf_l*_@>}>b3iwLo zi5n`cx%=I&aRa?^xi_s1w=d4TGcrA#$M>T5QMsf4UK(y3q**!j+Jm&N?mH_9U=de$ zDlv3-jk}{qGC)gRO!uRv%Z8Tcb**PXd?Jwr~3p|ZC^U~0js=oNgsPlkgK^`S@pGsEuDPl$=`GZlIyv} zi!oz}>!!kr70D6`H{)i|_K9%{dQZN&SdAT+n=b7KMo5%@=hJ-(f7NQfe=iHk-xi+6 zpTD9q5V(0WM-htY!(`BW)-TH{UFP1qzcIomCDrq4v&bfPz4dHiG}^2u@%H|=gnWM6 zsYfF=%jAB30X<}o5^YuH3olNFte4kV=mqtK-Q?XWD`ItxNORa6;&iRXicNLk*g7Y= z8hsr;gS294iH|+~4-40qR?-|bUkVp~+VJ(+FR})0=u}wPCXR1z$b{41L>Ft<`9#I_ z?*rY}yWaWNRc<}ZasmTIH><$qE*d#OtKpI#9H@0|{{Ts5&_q4bq-382EhySv6SsJh#ir0NDox2vg;ZcxF?5r6{ zjMDe7*0UC77O>GEiMM#^T$J)i<&A;o1{P{>F{u7I;eboVg4N9olQ`M&1jPaG_(gInADn0yy@TxY^&C3;_0wGbu=wZs zVl-sJNBjq}K6(mANOgQ zWQk=#PiOfLK{5__rbqtv_}3Jw$2RtYME+o)xI=WER@||>YgtP@sIdFxdPfuIGrU6~ ztkc8FgyFgg>A5!Wv&bTYn}a@Bq^8d=oKJ4>9Y^9#jgSz4gG=6FMKL=R-J=4r)ra`^7f*ZIaDUVdH0tgVG}>pA>S2Bqjy1Oj~;#Zqg?Q1lSKdZH~?KmOXKbYgm&Q$Bsas9Pn_zK@v%29H&UUofX{+y%A@;Hr3U?kn3sdWg2*iN;lTI~0?u z%>s+;9@z%gtE>5@TTa|CnfBdLUey}bGC9*Xw$b0*VPH7TFY=qN^3F*(nb2Gw&MU*! z)rBH?L*t5gGo^_XToVz_IPEZVEXOHawoIGYeNQC4F}0utabbbYZko@5%Z|v(CqlD} zse{>k+)o1;^!G0yzUZLUMeEj?izBD<$peOsxM0C@5tp_`T*w4*`+;{3ZyUeU!5JVf z^vQmA%kDzG0fX+iwR9S*tg>+50nr)@NNw;VtYNqKIA^AD6DNjc5=OG7uC}Mm(=!@) zVrq0DZaETs(uE3cex+&gXeMmk$n)xN@~NVXsxsh>6+6(;t?>ZwOlw|O za#pe$&W#+SaV$fPI4Lp3IAHa~?3nI>GdB+p!?WFovX`H~41WA3Er>(Au@jNxhLiNl zj!m}JeJllvH#CLWI95x2Y z1D+qgx(u7StXDvQ6NwhOvWvY(zvQd^MBVaP$s*w{bO%n+-65H(c}YhjXrp{C+CRz~ zJ?Z+vgCdvyJ-SL-)}Y2=h{9=M=le3Qy1iQ2D*aAy1M?Jv#)$Q*`?E2w6nU&u(b;rD zosskXO~~F#%iH{NuTym1G80XA;_8EHO0{2M9hoBwSF2RSdC8*~e4J}0e}ltcPiAwA zqE8&QGy3yHcUSMz@Qkv=XG@2l4Ibl&a0DTBVqpp9g;4lyt6G;eT8<=(Svb;nkGlE@6(Zus z1A9A{r)*aY4X1wn;l?`B^qDG!8#&HSuWohACmN?D7#rlDWgocmW7+vRHsjeqdqNA= zf$o8RkTI-Kpkr&$u{7YFE?dYUi$I0%+A&KxU?TEMwrGW+?64lB@9CQd9~Iivaun$@ zXT)VUQ8SSGHKtRQ9r&;%(2W{>B6V0oO4PQNOiIr{daDG4KYr%edU!X~!Pr({!Dcp? zH}o`1KWWrV5{59VCpuxbptnOc>Cpk#l%FJGF~&$X5czZ0s2XP0Bw-nzL<|@v4M1Q2JnrRoD-Jy zX*;K|TBdfc4Vm|RvNKO?8+y@7Y4ZtiyTGr;CYh2fOegp30?IW>5au>YYqHX7l+&ax z7o!paoqJI+8;_eS)(qjey|5SK*7+U+=Q~kr-^H}ys&5kNTEmG1=G7c2(W{7j$zb-{1Kv$1}e&5 ze!7I;%W`ffD&Wean-$nFPA3oG+cxO9i9U6836v z6{apU>9%b3OS(4Wn=|I($--js7Y$plPvW@6Nk7Tpg-;@gHHul0bl-B_`A}0s5kUc3 zXcuZBD+_~?n5>gR^$2D*CrXZoC-Sz+utwNzz@n5fKv8$RoAm0{3~m*vYS`UaU;Fs? zoZ7%kyAD8=;(iv0`aFeLl=FimZywZ&KmBy=`0QX@_k`1mymitejwI8u?S`+?$5G4( zCk5hyI#@W}q`Ql_&-7s<0yPRS?rd&bM$zP|8HMEOiomgKRDP}pTKP#wzqNs&IEHTaySoHuSQrhHYubBh4ZM z^I?MQ&TLNjE8H@THdLfZTj$;xrShwSg|;pdrFEPA8*@M?O32jniS5B_vr_OvuHrkk zb%jT@wq?7QV3BI32j8rKF{7gjbx-J{K)fa19_=ro1JuRL1*-1}aq*WeLi+Ks;ezxa zV>gi-l;M)e+jPUXD0*k;%ypQRQ$J8e29WKPTh~GqUYXQzy|}LkFfQ^H(qAsL8nhZ7 z^1W=fvQbOO_|#mG>{m>|nV2NPwrV=t=-D=LM{Vhto`Or?va{KU-HE>th=gnIrQ>Q1 z^f88+XSyxF4~@gSaq`h~=X{q{V&elL{p>#dBBa85g&qPW_Z2;#yysBS z^BpYKKrOUsk^uOP@VL$V1ed3;ZU3B|rkIF^nl0^P{)%@0K$)it1CiA!PJ2hUWh6nm z(5)Tn7p1y1Qo~9{`_R-t+JN3)z zFRR08#pvZV1E242C+6p})cJUKdRB?N-LcRAnGhrycD-e(G1Gw3b$gQL(O!udyV~%P zW}gF@(Ml{+Rrr?xNm`tF$D$51AY8>%c}DYASon#gRY_AHmPc#e3Z;+sw-wCu7mMm+ z+XSlg(2LX^kDQQh#Ye((sij}n*1snFOX+@lX?MXq**x&!O5DSPI(sRYQe`0sZ%Dg$ zh6_@|eHkVXI;R;{fQRybn6u_j(>{LlVn;gnTy>wXBc#roQ3*4!E`bOmeXn432WjQhf@S@Val=q$CV~Lxr9HltJ;3jxO|6KFfCoVVP1!t>C&o zF0SR;bE2@EvGo|3?c__h)3KBe93Glgry%`xC5k*onvjNi|Lt}hDT}Pl&N033UZSR0 z;OcZaTwo|lq2J9+=wUW|$fL|9{*sCF7qv)v|8hxo-p@(v{uN`5r8R;XLFFyS(~ALW zl$`Qe$y5AUG-KmLx1TxLdRJE>s!(ko@E)h89_PTBHFe){RSd5yI#0#JZFVf0jp~K* zGWvE-_P1h2x9wp2*~%=^>>3}^zJ9B`MbYW$Up+07%h-J?@mP@uAjVi@zf-vkUH#8U z$8TS6m1V(-AF?l?`dstFENeQDxI@qevt#MPXfqsr1qsvIraP~C8sT%;0Z-9VK6*&{( zL`f}{{_QuXN^+mzInB#3N9tb9i-G8^($|5%3VY4_QVC5>6-RH5^o)4P&Q}7aGzBZt$PwCGaxe1rGhEzIz52Huo8vVix6Lv9A6;u% z(^35)xX-^GH(E3(cb>j9ZLP47M`CE@ll@cS`B!66Fe%VH_V2Qvm0~*ahM%C_1pNuU z@YFe?P!np~({O$BaTVw?p140ECRma}bR^YKkgYn5$^s5#Wz=@qcu_Ea{GtraD|Ues z%VkKOY_c_3+1)-D7`VGm|A1^Z2&tE5TY8$)|dqCrW zWL1g%`SznBM@*1*7w7FOW@o#qKN_#o^Hq^1qLI)neuZ*$I z`Y>CP4)i6OB!R60p*mF@_-056`vCD)}1RLEG1OfU& z){`xw7u^e%ILmnK5C7ESv`NP+Om$0&09_w{CpsOln@DZ3bS7Tb(b3a8?a3}mCYCGi zR?M$e>|Iy&-QJlYe1%p_y7kt0m3Lf?A6rw4c1*^=tO;796Tf(y6f z!~UVA-6wm{BAni%m{Oq(G_OwU{}2r5fLrQ?ZMDm{b@f4oH;de)-ePmZ5u^DofroY6 zwVxymM*3vV&^tO>AoJc-K3mTg@rGI&A&c5tfhiACrUppT6J9K6_^UqRX;l6;%E)(@ z?Bh^FGg>!}!sXrWS<+l?8N0GUdCK%w+3&7aZ|>SJt*bjtS$_Lw)0$VNUV~TUcO9qk zWU0M$ugv%uh@}Bnai8Sl!SlgU!ACEl6=40wLJ$5&E+^dr1@yst&j+g%HmqY`3VsbM z_vUO~G2ms*297yY$EOwbYVkVip#09YK3cSZQ_7Z}_GgJ1;ZC2Smq`(h+@7wDpH$2| zUYE!y6C(=7-bC*#h&t^KW<>{;`z*v7HF}tB*?2by%qPU+9-lh}y=RTwe-x*(_Ga3q2%8yQAigsXW0>~ThJAmVfeBY+ z8g>Hr(d(0Pw(?_J$urmyb zq#6_@Ue}DU)jJg)>kN`HCq?H%d3GC2A<4oj@c3}HOBXd>o-vj}TUd?QBO(DF_pPm) z@*R0Y-z!#|0MCt^#-|@5lDMrOni=}86-nhrPvm=Y0Qf%fSDmef>-S4+@Vk!8OZK2bn88KQVW&S*vSUk| zJH8zQq`_bdL!AjNWk)_!P(l52%@2(&$<0Mlg~wjvw~B;t1}+0r8J zPu(RYcxvQ(|E|7*??J6gv=!7__({;64n({~S;In`(tTj~I1h-SoclUT8r5h74H+SbLf@95_Mh=d`3wt6fi*+H?kLyDyWl zZ0z(!S-UTtDxU7rq7-zEFLv8t37)OfuLEDm#Onw}2Q_o(m$q_fS9R)c zKTfwB|I_i$h1 zfeiTb7FfFBtupb)C_himjcThCgAGBFq1{cmThoouy?4cH!677atB*qrdkw2+Z~h9A z3?pdS(iuX_9^PkiRh_VgR65J89iz++5j(#qH}^-!*NW~>gQenhORc~bgBa_}Pw7AX z&GH6Nk53%uJK=Oj!L?fW#PbSsG4DGD4X&AwomWeW?uc9KR6RJUK@`MjB9}%ZM_NvQ ziWev~H3(LWB^Imh_USd`HSSGCE!)}ETL&ULCk?zsoQfwJouJ-64t5Xj&b<$%>`6lU z*hG079%vxJY2dx}fu0$)!LfXxpyeFY-R|j+C0JUdOUAPmtAC)~yvzyzL^3v#rwaER zOdFg~Gh|w7K$%bEYi}P< zb(!4pT2|j9;sG=+|JCeNJ4ts z3=`BEI&Y~Ge+-qJDi)}zupJr1{kDDhad=_4wFmxbTQe`@%cJ_!Iexf3|5VZoL2A={ z^Rdfdx=wRG6u_p~3pf5&%4=#c$-ASNDPp>*3@_&ue7BvS!c<}C*HsBnd+k+d&S`Nj zWupuj^qk2ta}g#<>$`Hi4&+<;f$~>~`_jH!PoAnN9eJ);%fNd+iBB#{DzWAyT@8Hb81psTm zuiINrGXKqKPJF3PsH;`7oz>ssbDWQn>8o{VA1x`Y7TyVpt47cRecyzsaK~iWJm76Vy-pSAqgG&Ao)WhN}qm&JY5vlMFQ?wsHtn z$Bc!I8E71#nfLHUX;YeSZ)#HSMOzI4nBpkAN@!9QQHT($OUR-3ZI3?0*?@d!2MKT! zLr=rkuh~^Lu(02H7vh0E;BwlM86{t4&aoaj*AS77)9~pipArnYM+VR%h@sXBT(TuQ z2d8P+&kem%fhv6qKmdMOEy4M;q2mNQJV8yHW1UyX8;(2}M&8xf$Y(*dhEbc=QpNtY zx94)LZO^z=IVr4gHMcAz3h8^6Fi~OZDXY))@py>cE`9(S&P&=Bp-{uY=VZtb4nygU zXjI8`%+#M19wT9muhm!zGQN3rX%z3D=Lx-LT2UK7O7723=DvPrCz^7R4}?I%r=|&K zWcH*c0v5`BA}^WBy)PE}0!<==d-bkE&3d#gMknt_DM{>>1RpFWTBJB8=l>3Gt7 zkBe076l7B~=`%eJ*DXVAvUe6(WV`2@W$J@Qe6Ze%-AD$CdfA~{rlq;=<%Ma%9`}_Y zDjt3*xyhp16fN2AKy~KFgPo!bl0&EB)uRN%+aTLal9bYbr!(aZQN{YO2`r1z!eFY0 z&Yvx3ii?DN+6|KM`B9$Sqb;0p*F-xTZi<`J5(r&p_yu^ftfx>L`f$!{`U7bF{eE$2 zCMj2()%;v$Tel5ONn&p9*N{6c`K{p*62JUZxN!g-dO+NE+B>}a)gBJ7~x)ZX6rDwejk(#uBZp-&yln#LYA-lSDMCurH#QB}Gp zTb1rv)(KFCx_!~d0`OU%W+l*DJ?Uee;Z>A!Op0nq z(-m1cfE=YlGh5pum!9p`ajqxew9GA-0{*oK5OhYnY1#}5q%QS!ij-T-7r{K8bO*b% z7RA4bS88?i6>q!JcZice-Zd{|h0zk3jDu2aDk`@F^6-@WCq0J}9%*^`x_1QM`mt8I z923&;m!v`4>OH;#R1VxHnfXIUt5MwDK>6xRUKZZR7>@j#?)~C*M_mPWc)E$Z(-CHe zsSSb;$2kq|@zAYgb;UKT&;t(aA>1ZFH)ie`Zv4o>#5+)%`nu^j<`*c*#Zr+ZEfPrL zE^npLu*5$1)A;(*DE?~u9V%7yEuxC#8@13O-*CAd!l@K>o7a9E3QlY=E&^!u{r%Ab&G~8 zyztLf--%lv1*uO~qt+15Hs*Wjx}ze*Rt1V9wk@Ypph-7vJ}lGK9lR}%wwIy$oL1h! zR;E3CZE+te!S8uMzav;+B(P5bHAjkGz>_lIS|&S9KJ>4&rGFu}+;+;|d56Xo^9T{$d4n}H zX-cS4(>vn1FrUTV0M75}#3DuRf*iR9(Qv=XIwEcgYB$30N6c2qm-odE=ta0%8s{-t^?Cxj@v(XP6UH*(UzH z+LL;qYW4sIsl;d5modx|Cw^-qnQ(#p1)#3qn7A8QM5A-7`}9rgVYU*4{=N{w z^Ivuh#W2+UhKt#|r)TqwPRrfeHw`*q(G?aW>y-w6DKpYPd^HhXhsI7owLC!tSuLOj z+31UC)EGfUuKL{Oro}ugKU^(ckbaWM1~3x^Z^~E0huzlc&c2m;5Pe0I$;m2ucI9hs z;)KQ~!0l=VXds)uvH_3RTVH=K#|7IpXM+(EghE!s{mG{_P~)(IDZT?&KC?Z}gM^Ng zvg-%FhvPt)M^-4Qcd5n{n`tIbQr{K@h+1D^Qpt29?DY`>iHdz1i!6e48!_8p+rE@G z9w4w%j+CreiIA6w%IIP_k$23=8wf5#X6YVv%V~xz0UHcyK=`S?D`=&jydm>|o)LhW z!e1^{_9rvPAP%au;f}B!P`ZKh34O9;PKsM=gU*=cWClZKn; zm`I^CHKcLZhqz_3+JOVOEq70pnN?`78AL4=l zpiMl+h%PuVR@au=_I%$B#t-w_wS>eyc1v3y+3kZgF6~oqC@S&S+brquxlG1A!IC*I&;y&Q6lyTz?Vsq85@s@4t4fzGy9^S zV=m)YI0b5rSZ9)G*)H2o0G9YnIQOY1>*?StFq3>VlW=%SXdic(JsJ$&ZyO4%N+ zLJYiG&(W^?>kluse9wD7-4PNd({tNjS4T~E*GRBXv2954PKap4Tox~)){ePn4U~MG znlHJ35n|4Ut9BzX%ITK;!CQyr%Rs4O$`s;u^ct>Px_&{*lf8%dY5Z%(nLw@yQKxtbUVqmLslOZ8`_16GH zg?!b|GGMv!qjs!eI-y$VmE$ioAe#EJ%C|P@{cg-*EP_ZTf@R$cB_^BDJe`sH zs$50H?-HO!!P>QN!l11C zM5o=P5%I4B@!kneXX;YYbf0{3-<@E2is`uida)8J&Vz5IpmDGzVIdCCQPYFkMQmx7 z_RDOPBpv|d*gf%_^CdUfP5?ZR%2l%A8Mk)hWJ(JUNPvsDA$7qKLM|WIOLs}z__C#b zY|CdMX*g7+zz!EN(FP8PsA5S5qM7mZ+Liiaq1C|lSf{L&22C`j7}N+C>)NoN& z;#eZd_uRRaawV~K@1QLcjrKom#CCz<`zqI29JDXb=9i5KJAp#9l zntsVKczCzWxut%&)$rQ$4JW$B=NrI-P4YsnDbod?XSapd^qi$#oajPUWMRcp$nY-? zK*)ZS#ioAjb$`SsA#MgVuwmJ~mH8=^Nli_-%FZ1307=J@?Qsr}K5!Q%ZttWXPGH9Z zeAtP-t?6bmasiZJ!AUz%SEZrbxTrv(&2QPFzfpWj)fEg61}_!^GPp~0NMsl#{M0aD z%!@KN3dz!be3hmqtMY@tP$9|bi#|Mx3I7Ct79Xn=J;IW*RdNGRF$PTa)YYJp8TKdH zX)kk93qr$$ZhUG~D7-65V#~wevcMx~ zi=ZexSv0;$jZ@6ThF`7D{drw|9Rg`ER!rmC2;OfWSXc%knok~fVX===o%U{1$)x)e zS~Y-q1a7T2ODwW!U&eF+~`hC}f@JoZ@VjT#hNgXWn;{ zg3dZ{VnoiWY%oXDZD{&{$=gg&h*P@Cj*_0MN7Xx>()A!`W3DWJ`##uBl~8P9a5Vjw*JNv2hlilJb9KE+F10;dq+GV?Q0$iU z;Y(OznD!z`p3sAT^sY>nWPy`fV^pUf?WKxaInSgi~wD%2ufgIvL39{eI6 zoA00&RqNl_C%IA9(wAV=tWS5fa^%a$$&X^E!^TfCGIC#Kejw}tcqF{j;CYxE(3!Bn zX@Hkq6C%_cM0D_lT52qBeT8uV{m|eM5T-*??|y=v_EsqQ^!xb$D4+BDuaiynfO;E> z5%sAJ+9+Y{JkyP*Buogg&EbUA+d8T;2|dtzD^T(_PZ_cp;UBrz7LJ)EBMOfX!X3r0 zUgEC5M%9~6P8K=EnoM${x~>AZ2#cE#a$CR4EeZY~{ovI)VYS8M9}L%GqnMb!O6js# z%lH=S{^jQDd7iT5j(svBCftHRz>rB#K*lBchS_gu1i@favEE zV7&Jf4)P6(q*^6UNkZ=(%V%3dSq0v?H$yjL{D=7SMoCM}|5o z0?K^VN{$j(L-;#g^+3ob6=~W44%`*jm+l;l z@z>n3NcHp!@vz2JRYF@O$qz8)cYQ#{cJJANIeKFS_T??{=Zi4xBxG?dm0+O6Rg;%Z$A74|nQ|W}`So%Boaq z%&g}Y_nr$r)I~>(z|1s|*Q}!=y{E;1`&(JjXy~Zk>DRj^$3fs8>y&g98_8z!IEOSK zDy-J#X631^1~QA*cl|t9Ozg|WWi~}g?s%ZEkp;z$%pwmJ9)5aE;8km3A!HsRTw>ao zASCP%)cv{t$jv0FbJ2P~@zKN~ZNf=PD2Iek^}Py;k_T{OULYDW*V8X^*m%HpnxU#Q z)HIasH)j~MhSqC|(dBKBAXDZu%-wa1lD8#^vPnUVPL<1MH43?e6@?RBE5lU*5+?n3 znf41VJv3kE35s@p$t~tCH^1=pY}8%xk9#Ft3ho01Svn1TV^7*2iJD{p4kwk5xl>gM zs*H6-8uFZtO%pmUi1w^BMxfIZbF6XAt61GjtEP=!QD>fc?M9)M!sZI+$!W;lkGhpv zb84$iIf_g(A*{lL{f9Pr3oM@Gg< zGdLJ9*S;P5Nh|6Eb={OTZ5Dk2;4MO^J^a=P2g9SR)?MoGW(uR*GpbB4ZL4`%O5Mpe z#`;=^ywW;>dpsgtZu5mLWv@6eKDW>z@f#QJeEevcLnv}7jGE_N**kGYG{o+5LnzxF zLPJKQmYn-JA|XaMt9ZsVsL;4J%6>2}UZGQjoZn?5s92Wj$Cn;YxU|_sCVWK9#N&lF0gDH9>S(OyL0KW=NQZ~QsQwgx{ zr?o^m<*AziVZsc386Zlfzm4rT?Wm#RyIgbi3i8>85hCXsJYUSxi6u+X+YXf3O7fD4RS@E@9&)+|NNU-o?F4tt2)QyOy~| zM0!6!*_6GlSsV%3oS)gd2sV?qCdMM4#r>9x*axbsu>iCGQ22?J0&oIm1_NyYeo1DJ z^!Ik|jk=&u4Lwod9&)E)WaK<5*+TN#O6JROk_%DDyE%UYhaQY*aa4_LQ9k*d;Bgmv zGvsp7LAYrn!X3THdH&9jAdmEfmXXnSV8b9QwJQ_1F!x0R{Hz+>) z;0huHz3<2O-WpPM-|#oFk$~|aAXVgB#`P=f277d2 zP;S?-lDE4zH?48w!{e=0dBWl_^3H3v0S4zF+j5}Uzcf5X8=?dpqya2~qZQ>$UF&`yZQj}|*6ld4=RGWGv8ZB$<5hQACKt9pP z9Ik$Qhme-_Rfo4Dij$$+#rz#sTbE@8@3-BR)JG6uw;!v11hesKIo*JmFq>xT-ax4l zFq9~OyS7t)i3`5SXB@srr4tTq9`_w;%RyUs0w;q~3touowIYvsX$>&0$-`L4p<*%P z<1au~3oe+MAJNov+iS~r0brgaybAHo@!YkeN0-w7cq3vY;gL@X3nP3nx;NWQpVAB6 zOQNQX8nb#BWDgq}U#LC6r$F(UAlF;c5_db#JGXzH)%K@e@$qd`<#OUa;>J8@l#4?D z+EXt?+qo^Fz87`33S^d?+2maX_s8Sa8`+nn?C@ieO3RVE=k9#42#(P|bb&5uM5GZ~h!4md3_Wj+#UT4Qsi zfGOWK!;R^Hui*odL{KNqg$s)SB@ftYQw|OYbS8gBM(k9W0#^tt7Ezj(!ehelU+}$hX|4wpPMNPvt&sd*7mP z-$vpdl#XVmtp;KD?y)8h2JmnOF)wzl-U+RBI^vQ663{BwS9nWqZ8xri`aHwc6W?*@ z&qA7Z-cWzm2Mip?{UvRBO;50EK(YxkE@1zUt>r^A3{LZsIC?I`7x&MV@J3#2vE%j& zf|>Rze_mVhI6W+sNiV*=A?8|E<=o*ieNkD(TXcsIo^jZUx561tYH> zhZ*CH5MVbz`dnE!Rh}r@Sg00WE7n2^@*7f1YD zru!`3^d4+`@IdlWcdER&b@YSIo@$H!0Sn0*hY5=Z^jwS)k|ZLKm9Yi1Wrn$(LA6QX zc3C&3L2;PDJ%p9xFs$(G9NjaCy6 zx*;k)FJ%2Mq}Gp`>d80yu6#K4yxMXNqU`~1IvXUx`|jn4v)pCRhU308kvqJ@?mc_W zw-te}-b+5>gppZ?+X$ua8AAm8RkzTmc=Qv+4wcY?1xmo+Tro|7r)ZMDv}#ADdzrda zqih?j-y_np6dR8xqUN_Hvj^8>E>3`K$!9$=I)366rm-eFi!RZ#jBgQ8?{VJ&B&uOP z&wD_qVG`v!&MrCz!kyZ@S==NTc|%NEMvg~pw?(g!!}va{3D}uzSW{Fu1hxPcW@i%s ze%IA!L%)y+JKt^suMIN$z!^nP@Y5n!;Ph zbQ%)v2y?|j&RI3_jJdrWvr3tZVsNdT?=s!w3#CZ*tUnD1KJxrMjaCG zscF>sVg7qfh6_W(SihSb#`7`73$kx2;~l#_(1qb6B?Y+wboO#n-Kmc1X=j4zB^+iX z(0c_gij5XQWT3*yPpkRH+!~x&MQ_~B0EN8dCph+wU;N7zFYDx78A{ex4rlOpe+Zsi zl3g2);J!D@+$T+lG7bgD-)~#Mco0@iiB5HQ6%5Fg5cr-K5+1T3kP+-$Y9dtj;iTN0 zUQW7xEHmC?0dgGeBAFXV?C!F9k}`@IxE85JJl91zCQNLIF7CVkOOI_!mOH7w}-$C0@O`fpH;=rdH{0)OtNy-&x_&t?E33=OMXv?Xc_m)})7 zV_!227Hy}7hFTm5xlR+&ylgXAZW(GJGS!>x=(BRh^~>?DO}~%X#2_NRcGf#1`OZe2 zT=o-+>4t>;wH9-HX*bCix4Vq@jfFI01{-b5(Ohj{dOm(syC1Pt^tbh+>B)>kRt^7q2Au-hIANoFa)?5O8$82Jg-zvkZC zHR?Qc%*X|JE#{EX7Y8%-bzHaJh9296nH_x?4?S5{6L^?DDR@6xfkV1WLw9|5mE_8K z@nSaeLdTT!n*}*%cs6Bf*@vCl6hU?TzPQ4E9I_0eS|hyn%~X8*b@MbSWa{G1{eU;y zp_5qUj5|kZG-t3gRFVNBd7wX@O8&jJ$9sw=iTM{cT$iWn(ucRKQ=J`w z&IC*r#42t@D>QwUzqPwLC=OUDUtP0Wb{5WGFWSl?E(LG2X+-*utLu8DuE=NlRprf< zsx*Gp!Ah2ulPHj1W(#@O`EH9H3K?tJAu8;P?_Z5~f2wtYe7mv{_wo|XGK|ZhZr^A0 zjp8=X`VAONGN*KS!dj|0;zYwc5^n*-K^3qb(B)TDaFlBHz;9=|$FZOFu0a;@8>wgL z;-6Xd+K*RV{)3>CAnd^{C-V$R zj5+FV<VZxf4eVydur~IF)v-%c$*Bh3-Hx5^&x(lS)k0lJ(imwRcuPt+T0I;t_V60a!{d zq{R5a8(Qa?g4oo{^Xh@-SwsrAW-6UDD64UJrT_8+9=TNz`+>BOl};|uFEg|2@YlTH z?CNq(9;HEe>Vuff7?XxO-V%!fDSgT|lsFpazg6-n@5Fg((*&)+@jEe1k)@p-I6ketrqZJ6d!^+^p$% zR{}-LAUQqdNwVQ^NcAN_c$sGXovp`)H8x0lODyYHjWg$VR4HyVICw446L1o7%rcQD z9bD?|?^m^d$t2gGXi=0mcd!p2Ux>+M)t8stUd%7L)Juj@gl2F!NPU<*<67iXCPXvl z2QcX_!}*O@^0!=NLMhi9Z4B`?ndxczC5EhB+8RA$(HOspN%~#DOvgRuRwa(_sD8~C zDi0cX-FOAE*Yp@%Ndxz`UIe#H=G_YLCw!5=U6Kmbo%B72M8Qk7=DQi7YfiByWHpN~ zwVnV*O|yVJOy}@SOULz>z)$lv^Ndmp3^j&(GtuBpJ+`y;9!~6@n%aH2Fj{NQ6=0XW zJX{m1klZig4P|p5-h2f}(Wo}c%u*v{&rUbr^N9e0pL3s1a3O)E-=l|LWV@4p%Xntp zuJFEM??vMCLT*1ows#SBOJ)7%3HPA+diE`u@p^6Ls@E552dmD<0|^MU?IBj=0sR)W zoV+4B1co6IJ*i8~wq=UtPO)+*%<{dLZH0gbe0h>CZ&Hq|gs02HlD;mB;tX_pTY{!- zcE)q+?H$koJ0X>DwB|EpS@TwFmvOn&hxXaG8Use?GB&0W@=yi7_ihSPG_oGGFvNb$ToZVyAkpi@gp$auH2m!cY&}69TCm# zOO^7un>#RxVt)usv$+z(DeEsfbOWxZ7d8ic>+FXb9zag8E!u?c;gA6lNdS^=O(jVd zgzC-vGvB~`@=h!f#fzlvhVd(`9za%XpJ@6?lZU+X_VcA3-8g6@`kB~H@IJcbzQ6M0 z(v){~?3oI7p@fORz_ZfcXXaHst2;AXeuEb!2W zL-WXGB2V0m?K=sf4ep63qi=(eH@=;au zJrt-$IpW)=yeoEgn=2JL-DYS;>-Qs^lk1X@yGlvJ^IgU>oHvX)+)5p>&xxGblp^pH zw;Sdh!0D`*ux=*bz=pp#zhRZ>NecQmMBI!-?mbbHOR27XBwuz6vY?zH2KROEHEcy_ zj8cled-kQS>pd;6rAF$+inhG|Y&Na}fO8QO&u zx~LCF#}QA(sPK@`l8bfkrhwDfX}zq1j!Qq_2%m=L$v=5py^b9mT(9u<{DZSG`4~x3#(eAPwPS!LJyiAL@feZ#RlF>qMUDA{!4$v4 zs5Le>TJD{3bt>jL2=*JKsmH|!)AM6GJxv%PPm+$*5t`kA$6BWOA;8ri_e89sJF zST?wNIVT}_J!_|F=WffHG7?LEkR(EL6~{QZY6xWP?*#eE%6z*V#oVoAZLtg3zl}u} zUg|A2g3@uJN~ca5>G7tWN67S!cIgEX(j8eW$Vs~#HQ6j@!0Ok--2vR{lE#STBZ%Aw zmRK1V>Nms?e#FKzTc?-J>=}8+L(`V__Jyb~?dHmnG5{D^j;1W@d``H#9C=#z@?8td zZ^vW@MorEAm`g;b_H`v!SDJ?~mi7+r<`#BB&zWCOVLVTv=#i4H_u?03rSB-XXP-8> z8abUqHzTe6$REJfq{~|6Bg!~K1*4(RTiHB%!Lq|^yQ^WgPU1yVqfUL+WS+_SSFN`b ziv=(BnL<>~7`m4AP43?Bxxl3g^2U5lA=j?=kwCwcaPcY$E7@;C&u0I<_EmR z(BT^S{+?a;bXYWnFi7=$%>dG-mbApZH+t|bkjkTE!X9Ujnf1chnuqyEX?=k|t2iy(5UvsA z_00Z5=B4b>)@oY9UWrorA7=0_6KMt`J>e<-i=zVFa@=>A;L}>Bn%M|xOc)OMXh&xr z`_7;^dPj8x855`c&iAcAK26eUr+-N|-)wmw=cLkFOd2)&1wbM>i)rG~Ei9eqEsP5B z+Y*zwZJ{8}xHGtDez^Cr4ytS>&LtY&d~eSuQb&UtTkDPlX#YJxGA3;*+uRm4p(7o;GkA1iPc6j<4H#+mQE=(^!EA;T~NO1;q z7pDI~mXSVL9yrI~$+74f`!okn>iNQaCW<+coTQY|3DY`9cY(DyXDg6ut_W;SVZ3gb zb+#(6E%^0}@_X~x5w%ot#|r0{D?yYmj7H>RN(p4$K*lQ4|EV4|nP}S+fK$ z)(aphIr_KyzY-APN)(y#>EgZTOAi=Wd#mfRk6Kh)LR9pS<-*uX{~%Spq3K3UK;8xX zL)0-x#So7bANc3giDdzHJ&Eic4p>MnGlj@$0L0KJLxTsA0X{M`BNrS3C4mt`R9-5^ zLYxTx9c#4{E#|H}hQmGFv&6C~gUdwSZci+g zP*0h#QPs<@G6J_@+T(?{veiscQY#dw_5@Fn_SwlxoR&SrMYor1GF-?Hq^jcMt`;P7KAY~YgtVz^5WQs zbgOu(oXW^VuM49W=2d&6XOF(Un)!M}tRVf6W4914 zXv^d=CH?D6&lm9}cYP`53&@&FHJJCubPis@ml`EyV^`}+rt!7rMn(=pJa4AL@A(Q* z&A_1$oLqc8iVJ~SDV^AH8h9ui zmOoU4>f_=?89U9RQLmUt>JsLrt#{!Z!;(j8&{`=amD}$VCUMQG&s?|BUgky9!1bj7 zTwfmnuixfLpi(Ac1YG)vjsE7&D1`T$-pQzppY8R^CO(o0t4L35(^Cwr00{SnRH5y& zhBqnFj&r8U@(s-Ss?CN5I2K!Dwvf7U5DZ(~)%do-o!Je$12UFj8(+Ks9vk=v_PyXo zJuiDpazkpXk}uH@zN_;6jslP!x+ww;G~X;a6(jx2P3gB_$zMI9ccCuSV4Bu=VEbf^ zzQcZhEZ}#K%zyS*hzUqCBCS6u6m{(H6!Y76-<}2Gq@GK3({8`>5&yG>{8qwu&p7zn ze;Ws`et&8B{yTF29l3u$0{>mDe`#@nHTmC>`~PO-3SvbLce?*MSB-5Z?R0Y1j81oM{zC7edLXk_Hw`bxcRkR*`i8$*@P_~iWM{Vgx4Crxqe^&4sn=}6 zo<{xy{RBgO6)4|DKK@3#{`&>SFoP3M)vLGQoc`}O+<*T_Ht4C6;)&k_=>Gjz{`ZRC zlF~yNT3Y>j$@sgSg@K;t{1owBCh>1gJ1a%tSSyyO{3S`>-(6lI1)!&}qi_C+QzzM> zN|LON-icAFe`aOrfSv+w?B5!$@9)m=XhCdzN%=oCYz9;(<$Q{@I@}3Mqt{DtrK&td zMXjXRnZK}u{OcwbGfloW-y7Y{HuSjrsTGX4GJF*LM{V_(S%(e;t$mON(DCgITlL#CnqYpXh3gKG;|0hSBfd{rzDB zz4&^x;8@0$Pb&ZW*7JYawuD8#d<%5>REvhGYLPyguAOQc7$y-mX&GD$yVXU?Xc|3lKq?}dXzr2P%WnmpLi zsu%U{EGPMXR8%W1sa8~x{uh2&XaTJCAlZ*zC8KeOovB`RsR-h*S=eC~P^s>x($Q}z zcfVKn+xOy`9*$UtksE5Hixv*OG5p4d{czwh}A^Y&lMd`m}ap<4ly~K!3?E3bHR3t&gLd}NXig>W zmgF>x*VB3aXf}QpiYaa`)5GeCmCAlLZ6pjxVx@(@5yk)fbb&pYwS)GAZuM#S^y4$y zg42&g3veQFz8SwL4gBkb_P7E1{CG+nvm*?-N&N%s6>^&@7mjQE)EO_LkR~c?O@H#| zJKg;dhYv477>`UOhl~^+)1T8W4gGOjJVKtZiYHgu-spZ8%Qt7TxaCMdWnD4W9%>qu z9;Edn#|MV(u6VBeU__cI3UKdti2!T&lpOijM*q^|0`oI@k8_{`gj56J2KfBK z2ihM@`vc^LjCSa}c>7_vA`2oeI3=@Le1B5Ahw4wqTGD;UmkhJ>tfg&egRj;Sq(u3{ z+y7<8LoblfZk`z3@MKX>;|m(0Io}k={?tH0D>DC^C7ToZ&1E-v6aMt!w|^goAHWAP zw=fKC=sLY-`k(cHy@~PR$Oc$h-Va&Nm`Eu*+11Bov*#H0l diff --git a/ALCF/notes/cooldown.md b/ALCF/notes/cooldown.md index 4c9d0d774c0..f19ceebd38e 100644 --- a/ALCF/notes/cooldown.md +++ b/ALCF/notes/cooldown.md @@ -2,7 +2,7 @@ - Cooled down over last 10\%: - [volcanic-blaze-4312](https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu/overview?nw=nwuserforemans) - ![](./assets/cooldownHD.png) + ![Cooldown Comparison](./assets/cooldown.png) ## ♻️ Convert to Universal @@ -15,7 +15,7 @@ TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1 python3 ALCF/ds_to_universal.py \ ## ☃️ Cooling Down - 256 Nodes: - + ```bash LR_DECAY_STYLE=constant \ OPT=ipex.fusedlamb \ From 6c208091d7ed290abc43b2e7a1ab66115648ed86 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 11 Nov 2025 10:07:41 -0600 Subject: [PATCH 516/590] docs: Update `ALCF/notes/cooldown.md` --- ALCF/notes/assets/cooldown-grad-norm.png | Bin 0 -> 263081 bytes ALCF/notes/assets/cooldown-grid.png | Bin 0 -> 876567 bytes ALCF/notes/assets/cooldown-lr.png | Bin 0 -> 191538 bytes ALCF/notes/assets/cooldown-train-loss.png | Bin 0 -> 341947 bytes ALCF/notes/assets/cooldown-val-loss.png | Bin 0 -> 50028 bytes ALCF/notes/assets/cooldown-wide.png | Bin 0 -> 321310 bytes ALCF/notes/cooldown.md | 440 +++++++++++----------- 7 files changed, 223 insertions(+), 217 deletions(-) create mode 100644 ALCF/notes/assets/cooldown-grad-norm.png create mode 100644 ALCF/notes/assets/cooldown-grid.png create mode 100644 ALCF/notes/assets/cooldown-lr.png create mode 100644 ALCF/notes/assets/cooldown-train-loss.png create mode 100644 ALCF/notes/assets/cooldown-val-loss.png create mode 100644 ALCF/notes/assets/cooldown-wide.png diff --git a/ALCF/notes/assets/cooldown-grad-norm.png b/ALCF/notes/assets/cooldown-grad-norm.png new file mode 100644 index 0000000000000000000000000000000000000000..26a970ec0ff99f29df5408a177d7cd8b2e971025 GIT binary patch literal 263081 zcmce-cUY6l6F&+FVx@>QktQJ0rS~Si_ui!U&}(Rlii-3mHA;~lq!T(ST?jo=gR}qv zLN5W58_zi&&;32m_ul*0{U^`+zPmd+J2N|oea2roX!QyGuVX)zW2B@bYzJ65)N!`}l#(4JIjH2PX-A zMdg16YA700i@U&DS) zAff9T?C4>p=<4YR@cWzWV=?jnquKPo%|c>-Px{wrBR@x91uswBU>sEgdtXN!6&El6 z2YkPT&i9Yhaq8pySAV#de=WcffK$6KPQToSwLN%vx3pXp6m-=T6qs~i`2N!-?V@LE!=T9F3%_BAA8HOiYN0zr{Xka>Z=a2* zrfEcOsxZ}%s+72-ixF@68Ji7LhOeF7gP*dG=P+%p?N`Yhfz$G(M^Fsbyza#V2p%5m z)ume>EfY!i67Uu)2}&#Q@bd5OuL<5PRBxrSqo?&JTQX-G*8geDHmv(IHd^GmkL7Dd zd^uypfuk9o%b|Rp(WAcS68U(Nh2Gr6^-k2!lu z)93Zr)iu(}yVSdPn5Lqej9E$J8EeFOvl&FCQ1zO$8G_$rR|l_#>uyFGuRQcMM86Mg zcr99(AWr4T&Z?mKQmjNb=Ppt8)o*(HbEmJFv+MFb0280%G|7wKbraZSs@}O*GS9MvjRb$Fp_pRyWXZfwVY+hG=p0_@Z2==hWx6I)TZ#6V0OZl)* zQ8n;0rnlsx>*SegCy{&<(cRBZQ5#IeuO&mXLx0K|>0Dd9Lfc31VDa&{TM4QRTS=Yn zQjZ8a6j%xNlYE{{9kcBWHc%Er#{)6j%CQGBA>l&Krcu7dJhODQiE$@d<%>IB z1+N(0$-VgHSm~eqA5(UtywgUMn#8p%x#?C{6-S7U6khG8>;*)4bCqnw6EC&v(sM3r z&f$qI0=yYWc1WEAW4MF*yc~^93+k{`T_!;;eep|*@*MV$@b6}ivIuVp&8ZlsLNF4t zoHl~W`$O~_w9>LCG)^voKs1@&h65Eiui0?asBgWdsD%wczV61B+3&P@9x`B zD$#Wv-U-?^r}jBB$A_!hN7~Zbj7H%^`DR1|F-ekE5*81K2Y8?QEXFh2?XeQ{`fyXT zQwQOto+oS`9dibH+sMD)wrT3_z+VJ!cl(u~B2(1gb`Eyz(dCPmC;jS(<@U1r~}7}o5=2kpNcupYQ2e@o`2m{w2XpwB9;et zIa{f`zUlnT*gFqq#AM2`(A?R&zCg%edxtspaf+^b3DXwS{RcU(p!%Xeg_Ber?o__X z+cnc=92Y}8puIspG_&{w2(lb!}W@*=myQkozm>X^a^DW z=9WZ-zEX$lkoT7I3raHT?TQgkFmj`XruwZMqj#AdMI4D7C5vAAJ{!|Ilbg&pHwZ7i zK5>8Zfg4->v(T*27__>U){y3q29L2;MT2goz9(qLtp3~$HBmpgJptKt-Mn$Mx+yuC zP!2bYR4N#)aKW!LG{^T8uZ#~&4oo;Uj{f+wq-&hajeDeI%4MRdaPPVHM~gA>!_vbC zp+uKVmjrDlDfzBN{gW&1!?21esfrM*TASF9s}+=$X;t-_?io|!mBllxo?P2cJwFD% zITqPYnIWa6`I^(^c#S5Oc9yo7c3wh}-km;AJ6;YvY*-B)6ioN3sHjY? zOg9xaldhktT{ELLgIU0BL@n4~P~!-9OI(b!{2< zboNAeBkT}V2wAHwOG4gBUK2}KFf;h3Www<~^K{FbIgyrmZ|5c*f19O+HBA54afP+f z#nENgpP4(5(E|yK82Yc)Ju*n8-5ZDIi<6s2!?RyoI?tm{7-dLgATljsdD@@zn6f^H ze^Iazv}s2f!+B-XZJOaH5muN_m@w=;EHVBX{+B3vd>yfdocqIo;aIt1IkUx&ouZar zJ5oD)Ox%xzO1Mi{OeMwmn>bIcUz@2KtA?xElmORrMwV?~e184e$QBl&EQMbc52N0p zmEpTEd=aV`rfBpnTD3*h%wbT`-|qCAFa=7{Oj#{ID_>rSFI4Mro(B1zmS48}j=jBI zxjkdAc<)f0Bn9*}h^m7N6Y%8em$xrw--x_<`>laBRNq|RYAk!@!^&j8RZ5Remgq*< zTlu$QMdE3xS*gt|^^v~2D*-7j<8$Nc<6y1oJOTFsNav8&KwWm|Bn?9(w)^|^U>`+H#~v9s$?@^?3orzmh^i(QFXN%nge;@Vo> zkZ^8`%g$v;PKX79|0EZp{av?fRD)r3ocSYy?jrr?u4uY-y#NhfSI7rUpX< z1zIb-RJpX`5Pb*4X3^f=K^Zw)JrS_+st7dr!m%S&GR-74((<9{@nFy}+$ciyvy@** z?~Me?Y?C`C>c-XDYoj32=!nbpMW_)gO z}&8>6DKOa#3-MBT>Q;IF+d z721y-dow7X5!wEw#TVl?F*)r11_rRyX;}zR2wvV8YR%^JqFg|G1Ao$g$*pP03gX(K zT_`^rI&N&UVz5%#A)fb&B*QKqD|CHDB&;OJ1ZGHmk_HFc?HBHp4vtUdCFaFKAIc&l zc*8Lt8xMmj;iB!IL-v*#@gwU_v^J3u96<41M2!i$DiE`JY+@0qiEV?ORmrw;0L!rr z$Mom(nJ4`yP}uvkq{Hg{lX`s$u^8`as!Fyd9%k-39uE=TRVh(@ z_VVGFXd=B_S7P#09n*kU_qiGHggU*@Kdes!{OF^?XYoHjM;{8|*}34c2IEgm1{9Kj zRBIJDBU~zL{GE|UW`y&IE$$Umc1h|wyY}K+d%Pe@(-`pW6H?B%*gonUPGn11az-)rzfJEusL|Iu` zE2}0pezPKvB`vhYFk+1X+)Pn;1lRulAp;PM{tSGQL;CLrUINBptZMsNVAAYHj+7UJ zyss!Z>Lt-l>Vb;{84;2D;S)&cz1L2h(|SSlH-k7XpLySX zeXBQRDd!K{XmIBk+h7j6toxcK0=V!W7I4fM6A<=S8k)xZVJn&7<;#~felT8Vxbv`7 z-&vmMKMeoPdQ=*oye}8!6!X6;c~RIo<|r$6E*(6=L@N;d*)Q||eMFRd;>BV#tr9~r zb&@vAzoeK(zT)@+(&T*oL*73@&fUG7+iEi_;bL(|oUHP}pYBm|rq{_>Ny`WO`~O{5 zxXDvwgQvc=o-LlDvhqL|F@JI^D&n4)-@h`#CnTU|Qe|kO6ejxjl=WHQXi$yM^26!YklGdM zt#tCQF@vXj%MM>&?D?MB#;AvehQ@>%TUv5gwH`OGtN~26DOJ^c-rS;-(dh_RV8I*$tN=J`9P7Fms<<97FqYQ z9S+y^UiH%!gB})49&NfzfA%T3_#F0fPrAlu{9!2=OcfsOcyYSdbJCX-gH^Rl)c_V1 z))_8@oJc#@S5$C|AEl<`4sw}Prc+yo^^0k1XvB`9xOh7>A8JW;mMlXev6#!%I!4ga zT;z6h!2S%}BWekvJL-+6 zXE7+&PC6JCx5&@Ww=6$k_{S}ftsfAOa|YSFgG}&#y9%=Pmjt4i9jCLkStWq<`Qm&d zo1oxV&)mr`*UeIJ!UzNhnlih}FN6cKn;#gcfR zb4r*NUrF!4tn!7$ujEw`^d!g}^Ks~P$wWG)-|)vCH@iEmye7w#(azq>&RRL8(D6Oq z{z=T}ebWvoqHAc#oglkc7XU2; z*Q%M>s@Dd=Yq8 zoVBIP&hJMmFV5B8hB?wd#ZDJ>_=4|A_`hz&3}`6NVownY`)KTfH$V_Bg`FGEpAXoZ zi+mc`(0aLs>^TZO9qnCRHL(^BItXoB|8jFh)WqtiB^KwJ2R(!4*wFVY2i|RruR??1 zwuHl#4Yefy>NS>NM}Y*LkB4~o+daNWiXxG+@tZ&Lg|?lv5Ybg?>m``CuTIUb;8+9@ z`c-beNBYa{bwE27Z#~IIO4|gG851)=#j#H~cz6!xxNRa2D(6CApxN7rNcxNtHL=M3 zanN8siV9f246z%^svmCe?(QztTQ`s@7t07(M87SMU^zfUu87KQ&V8it8WVxxk~t_A!)A#F6k>wC^q@6D%OT%9L3Y`;+Q2{rPXPGJV#G^5+V> z`x~^LJtiMvc@@xUQ5^W0vt8en`{Gcr06ibH0I1Ia}2dGS6DL6hDh95@HbaEde9roift+?-t_0AgY>q`>)R>cMHrQJU{c37HhCiOUt z;``FSI5gU6KIc>ZFdfjcfS?i1mthk(BdUkdaF|*`d+PyAs3V9J2E`t?SZfrXJN%Wq zvNEvIPXvZEnP;S*IN=U;z%FA$pB%aXBUdrR42K~p@-_-E_t8S-<&y^AS^A(d=u4Kx zIcxYKr1fIJ-M81X5nW+!ZOsdVp8^|4_rYnE?JEB0Om^%+s=7|fL+7$$Rss%TVd4EZ z==RX_?I51=1S6J-%Oa^!kTiWCOh>BDZdds5D>O0!HXce_zCuAT%z@9!6aL`)-yG}jHv6n7namaU9UU21fIU$8R z&Aqv#M75`^73Op{t;3fd1V8WWi>dTWMfwX1I%%d{F=Vnryn{KR?@4_a3v61D8#${x zg53Y;t8YP>8+5Tu1Pb;X+YpHe?C(}y2%`T!-|AStml}m|?sf0I{QTusA~FhN8^gfq zxAtBCGLIW*ly{7<3D#S!vq4O8ZTrs)o1A~I$96`O0*U$pFuQQ;bGl5KZ87Aom1bUP5UCNg8H11RY@T;HBW zwGPq}PMyyK;j#`ru~^Dn_;kI8VV5Z^pMBCF=aM`nUWbjoM9@78l&+a{ORXT0X-*HW zhw4Q7e#BW$#a;xhdsDN%+g@OS^3!l(_KpesL2S!ekel82l&~q-SwG+T z5ueeGEPcrO^MiL`tKhVN#6$nIWy=KL_H$@tz&tC6`(i>s6JXuLVg_$|Jb3D3A14FyTX@;??bBN5HrR{M zObb^z&@yjcOnK>Kv}}Y3iah)oInep+N-pxqfCw!Z0hh&ac$K*WsIn`2Oivm+<~+gs z=Vb>g$ozwG>`o7or-S`MFnLiRXj+jLb)4f9eRi;#z)E1$d_Z>9iAni&%1ffYq03>J zgIf?wqJeZ> zQJzXCXKx~sFJ&r3kbBPWc?Dk}_f+NyR@Qp# z6J!REwFMxT5B~=F*Te{~B_F8$8@_*$ds?0qJX*j00da=NS#BOcUcRKxLDmCC^HHr+ zTA2dVp7T9g*(+Wr3$UMR48@^!1<@;^`f4`MtQ(`zW~UJw`iQA$LvxDoo%R9nMOdR2 zEiElz5PUO}VOC-WYO4@45Q50ro^i;4lMc|XO!ee_kp>K|eQB7pJg7W4su0vUp?eWz zyrtxYc6KVuX7DU^m1_B5C~7C4^I6vi4+duDx_Vn_2>XtBtpw zV!>7!mD>1swh69AU3qxgYOOHoJXpIFFS z!1PBNFWNh&Ja^>va-Z0Ld;K#)#{*sG+EZc&yH0mhm?pv3-8eZl^_Jl_*bm@^bBTf50~yjU zU-s1dRF?Y%ii(JYY8f_%pOe7cdVsmHQgd%fZUNIb-Les!ECD)gH*U00U7iJ9=Ct5K zIAooqj`Za_h)XNjw0g{Q8uonnI(Qu7uQIA5yDlV@hgd&8jU_XmqaxU(MW?;jVZ`l_2US3~S#)VbAAVlp-@7r@AoFD-tgUW&%7@q&M~UUUHs#r1>u6OkXHl z*qU9F2GuaP@VuRQWV9w^kSc9|joxf`kAQ_V6nKuDTMh^~vXNT~qScd8{gR*@rv3|0FM8xG4S6D>>w2ejHS6U$NkBfQAGXinO zR?OP*{057t&Iq!;dK%%Jv`N_yd8S(L%=FAQ>Z^t`!c(%s9`o~$x@TKg%WLhNWprSy zc-vR>f8_&?^lkb&WM_Q}qBbk-YZ0CYh+w5S_(V# zJ?-zOUAJrq_3POvR=2dw@iT}XNQpqqN0P(*rsSuJ^FpR*`nM}e7QTD1*SPSTW09PV=qg1})U%a5(#3J3ua&fzPWrC{ao>N=;XK_Ik$Zwq#&%uDY@DdCHf~ zY}LFCASWkR>x>=Izsr+z`4#hQBUMoU#yo19eFsxkUL~ww-Xv0L@N7Mm9)>+OKRa1Y=qtXtHYhJhppkip5g|?Ms-jHaa?Emn z;iUEy=U_dtK*j8kc+G@3&zAl>BqBLi{#|R-HF;kdqJPc)_09GEoA%OU+FSD$cSp=9 zcjl2sP|l!gGqnV%jAX#G=_JEh{E=*5fxO*{7*(Ou>x4b$)NAPn*-^I;PaRO?yu#Wr8iAjBEOMy#E? z>36pS8LDiw!FypLFGOv#d}bf51%*O{Bj zyeSdcI2}1Z+=U-sGsR(S$!=43V;E__uD263AJQU)S^sJ+-Z3QOe-v=GPq`FQuvEsy zmw^OB%>7ADa5ep$`&CvqZ?<{84mP?mPFbKFLfjv;Ju|s8y;D~r=1>FPQ;_+mi{v* z`Y_aRjc*zd6L_f@#UQ>gX-COwv)lhfd})_4pnfB9&PFG{4D|dFCuiVskHVUm<&qi=-(OO;b%F;&fe^Thu;pl`8xK?ghWMw zV3G6y-hCMMb~R>~g5li9*0@$I|}pMMWktUDw9Ju=Fg=G*~4m zC$ulMps@awhoRtPzV?d&9@Q-~+b(ucDELOX zBa}`K?RUDQQFSWStCcMREJBZp+C})Uf0Q|~{j*c$Tz)~)5K_77AFuR*^jhJkqR?E& z2>EuHM4R2t^T)^5Ys;s5_c9 z0edOV7uia&g{`e~o;J{@jU*D7%^I%NF&pcLbRq!%a(CYK_V9!42Y#c$M;)pPjdzKq^TaUvgD7X^!d7cN8y#Aw5Hu zWOYJv%P?WqL6>pX$g1|u8){ldd5>V_54e6O_V|QP)Nz?Ua{y$W`yXd^$o9-5jp0kJ z)ZuP{fJxl345C6%TC{(tqeYDuy#d9QJ4PYfE#2@z2(=}v;3plEkii(*6n)HxG0pXy zpI{|o47lZ36kPu9L}COp!jfOP!EfHgC&gup4hll0 z5cQ}!S7Hg+MiR1dF0dtu-74WBw!4Z~^dT}wDkPcTS4vROVE^D?vTq?{9$iKEUM4)N zmi{1)gf$zH`C-36CsS&AZso*?vA2Fb>Z0vLmPw){=Mz9(vHXBv!{5KIu(*hpW$5U9 z=TU)5+@aesnl6%ER$@zc)K!{uGGGkmUfmoYr=?{ba9Mde$;X%G9tn)brKtAj?6Q{- ze)9Dn)%L79Meohp?TQRdOj;&y_7c%Ebr^^DpqdT2F+oX3h6FVBl_*%w`=sS|j29 z$@i|x19uyrv#Qmu7zUqN9fU1reVGV=w50YoPok^BR+5tiT2;YiPj^`WT~BiN?J*Y# z@A#n2;6X%gQ2Z=w4qpeld21#ORgtg4hpb}&+62+gY#H;zZ$hE?3Lf-$;*-kFTs3#3vhQ=ONNadSDUHDnP)c>-%$l(3bhS zlG0NChg|hppAB^?%+1ZIRL2E?Q<|9}yYe;u)1-hScX#*ImhRX%zQ)^b@{ymuh(DFl z35ne6&QR8!JDO|s9Hhj};*yaa8JTjLwK55ds8IQXEVuE;Dqr*jDZJCZq(==fYOS}-+EUT9uVV*7q&50G@KFO9$F9(sgNsX)U_%yXj;q}nHn)S}$DEre#fKDy@# zFdj?su=f8+v=4U%1=8(#0z_>BI8z1|2 zNehOMLw>Wv`K95VoqG6QF9S=0^!|r+c6=#~)_ss5FkR|=BE`3~*(|9*>abZ(qvD=z ztAC+BBuzXjY}rRb+mJ;^8AA_km9(} zS);ako9{GNTy&|D~@b%U0`deSnNRYNywNZFlYI<4H*r>9kgs#*hb`^i7ag#H}RpkyNtp{bX z+Oqo?NiU+Ds|8I>rmGw6dcKX4%1b(W;-gUMll!B{BgId*Z^wu2pgZ~`w$&mcp7HmW zXp=6RYs+35Ld=NrqRp}lhke2iSc1+;s;51AzGb7$Ll(T(2l(sj^HTLeuorvFj$?&y z@~>=!L)M{(@vXVIMGRDd8%j}?`~uqHk!LN;6B%t{LBHj~bf4&2a?G;P>c2;f&mL2q z4#Ibmg5W^7*jEZ2G3}kEZ5@2jb_n%_;(p+c&?f2kVG(5t;9B5`hX?M6XdgLPb%$P4 zco3iBVV|Imv@P%c+uwn}ucVflyYpKvbJ*Ux#P5K2gqp~beLF|o A;Zc^12SB(mi zSbF_6{8NFCr~lZuH~Hg|FfrOT-t?kEJKKgzY%y$0pDtY6yn}bP2`@wtk5O>;18(+GP?71c8 zU_ww{GC(@BeAr0m&v?0~kvC4YkDFwH-POOnEG~-sO!il0!Cd;|Mijo;@aplxt$V4< z#*SUjbsPq#8m`ODi|km8g!Fr%FNg9EUS_i?ukKEzSn2B^@uVp4JTc2>TsqVGe!lEUx& z0;d~)rK8k?gS}xaLh437ZUHR zGV&t&dp0m1^7aoV;UshI3NEC#^(Ghp&(U=!(L%(Dbci9*K)Gi_RN&i$1W|%z1Slw; z|CzI#q~7&EVKFX@P{ygS&yrK(kKvGUKkMmHKHJRpa{T#4hppp-oVtd(QnAgOw{Hc; zz$+`aQn#L6aja7(_8pspBxlX=nZ|sg|83R3Hu$>ZufykQqK}e)7J07n0Rl1eSEsyG z`j58pp0%^cCAOk0-G}e~m#hEkty3Y&R)B$t`@hH4|26V!hx9lRgd@!TKK+^Q|9E1+ zF(fT#s`%Ue{FnY;982OloDbKpY5vFP{xy@odQ0I7Wx%M zW&Te_gRhaXp5WP%|F4yPK;b%|Bw<;CO8<+|T{0%s9sDm3e(&f19O3`Pm>is(Np&Fz zEbRP8pZ`anMP|9gxhpS}{;RS7F7^MLN|+$T*JA$)td<)ulUvq7px z0RfG5ItBC;E@gn^vLVVb)9D}8^riI41Z=>24zn|Et|ji*`!Of=+9(e9;N17*ab2WxXIJw;VNWc+$qwa@=!{CDDbe(go&}{aJm$O*tIYH&nm}#^6y|z82hpYK`N1Zvz*f?=LN9@#{+m<|jK=7wu&Xr(DQ zLsMyqwSB+`VYMyZRZ=Ho<$2|8>9xcT{YA2nPIqap3Zem%H1$11OVP)=1vBP{(%NXo zJeEsmCvjN7%FtRQc;o43=_e22B44-3y5Up<3omg|a0QjX+O0%+c9Qy;ZxTBAM<*9w zu#|lRUIRs+2FSx=Uz7DSWHGT@Nh!XHSKFUsSzJ|Or0jS0+H!5g!hdif4M*DBd?#tIILwrMJ1(oXx;`N^HPF!!tUQI9VTv2N{Yp*KzvA5> zEmsC(cxmCoLu5?~NhcrFf(=!9s=7ZIsD@4Dh9FucTjl^Z9-RxOJge||&bXKEMfD+Y z=y=JaH~o1`v@vrWNqd=Ab{p#FKxBIUi##%CRZ) zz#DORXHgZH{3|N~0Rc50-RdF8%d_E52y7T1m&!QhwqACrC`U=DU&<#5KDPLEgCxI3 zU&ip+87hP4Wcuxnf`*1hi;dG$xU+(X_l;CuuJjAh20Hxx*CncN7;_3SPvet6?O&Zf z0HdS5N3EUb<6Em6>&iYnlU}S+jqTvc_RzxL<_lEzCGzz=jA{(ZTi}34n0gTQEmIUY zEPf`mhkeNn*#B}l%j%WFllX}nIZ?(=Zk}Z^i`56+RzD{|he``~n6B}a3K6Va_t6vZ zm#e5=2Inh7Q1xv!939&|#^*c@laKU7RWTP!fHOO#G;*ed`1nNI>FWH1kAuzI0ltId zXFYxU4Q+F9qsa!<+UE4_aQ>08`5~BoK${k3f8nym_gtMtE-xvE`HDWVs)1iY#caLy zPQ-~dM^)u+J+K9NpuWGFN6MHw2x`-2u-9HJOkf2Wyl+yOrOSW-e2z| zH-v;l25xf7EsWr9f^BU0jzXN(+&b~~nB774Um*yl);EwqMb)3wq-|t7*l-U>e{}?T3ovEtwx_QmY`#AFs zr6;+W5wN>dgk$^Wj|P*$ouGqvy)O+Oei|GRR8cduEDSZD<0V2N)VlZYaO89tED0=B#3BlIc_y6B>X$;w~Nm6&rj;J>sB1K2BM zcbcH6%E(t7d)_V&_a`Z6-)BnFBeyH}(8u4moNg)Cwhm)50iT`YC+E!gj5mj~K5OGiuw8}yPM?_EV{=>s6vRQ^hP1z*qPzd z`ORRSz0kO6v$?u#o1QsmY|+Xf(y?=nT4KD_zvz;RvQt!Jpbn(!fh}Fjhq*iWBS6Y~ zZrqwCbrVwq3z9p@#z*(ka(TpqFlu%+69$2s*d9tozSMp8yi_!sM&SyCniWZ%@uHNV z_$4k9IDCgg&pe;CM}J{5{m)+QaO zy@U5Ow9;ciBNHKV3#XKitZ$OA(wp zKuk4HsPMA4v9Uwf<-V+|q3lNF=OA4yAA`Y}#?M(Gf74OF0{2V6%KJ>r?6IFv~;MMuY4j4B_Ed zavQKOVIOIBGB-OE(4tm!WeJv-#?vU%X0gidmE<*LIN0wu({S!?;x*1;xOP9074q`7 z?|kNTa?T>dQ&|VQnjaI0#UdSzq7O}a#^rje<0`;n4gTmAmWgYJ*U$4*T~I$f8mS&V zg0o}?dI^YF`+n;l#)DY%J9Vl;G>gnbc9&Hgoz?P^%ELSwPP815>9bSXj3P>y&;_eu zN}b4~R*&i0X~_5NPNJ;x2DN4D7^}?+_Z)oR`6nK0hH8?BP8`^MZCBnQQQx#)CkC3} z+qX~;d}MuOeR>@v}cVMT{z87PGaN6*BJP;&WP;J zcNI#{F8nqk_lWLv$=!y(_xKfi&yagx2+wG$M$#IcsMpclRuIIvNPW6>vKm^)k&=t=>&Kpp2TG_=3; zO)H*JeYI^gh<$5*lsU$hnRJQyNbsYXv(cE-g9N%vx{0V)aZ<0-q+6!JM;G1m1lw)W zb5b*8bBKaK%alx_bmDeyvs$HL!gljz6Fb1M0+xawl{2=vea>_pA~biFlUF|T*S4yX z+~e08OS#{XA`3TxmFr7#E!i(Qde6uqOu{rRC4rx(GvtP*MB}u`efE-t_8qF+t}DbR zm9X}ncn(piSXHS<82i_U0}YC5kNSd9pp)Ua*6XUB4Z8*f)^n*GcO-)oVKSw~-vx0O z7uSs3+UfsBdjt`U+Cqrr@vlVX$&(MCjCDMwM#+?)7Js@jczv%c#IT{T;`{p&1ICcK z_oM0bsyZ5!xdp8yO^Q<$aHbS??avj?^}b&}@hJ29w0U0q;5RmM$WgD(SE)oE@*+h% zaG5K^mT^@CTy?*&(E1)L>@1|5{uxhWJh3aYfQ-5pAyv*fk`g&+xMSGc2RCFaHS~X* zMj);6A)+u?EAy5vMab2HI-wtqP-4lrPc;aVv>!sqr&@zJ-H)oZ392r9Ks$e;x&A8F znJ)B2&|&*sn!H(w=2j;kNVSH!Q^l?OfXhokd7^%k^1Fx$6TPg5h*+76(Z1I?VSH=Zs% zvAMck2{jb|%2I)=i!~N>A9q`{i%v0!g_Ws$+T-}MU-3Wl-*&F?pA|U6Sq}p(c@CV2 zYJQrp+?z4?6R_G�XQ7SAmT8FEden{BDhrwuT9eVTo>I=mSR46|&O#!|DkH z&z;Ol{TL-v8CW%4s~ba;cP|L#*3RqGvnxU*x%ZKunX6$dqqBL>>i z@`n!}syZn*r>3T>g}k&+t<|6y6G?8h?~;UxG$Y%rir9Z#ehaKd(P_={^y$pB{tag5jf>l&0-uYA11W1Idh~Wo{lfm?QToe z(ye6nlXj-FT8ABBEKPAxLZ+Uzs=Ir(Ti<%)$iBeJ-)# zL$F>*?YrSo_h*0PV~7j({_o%AtTUa~u0QifrL|XP}SS_&cT$_u8zHCa_}uomqX$vpF@-Ys%-x z%Dq^`1G~u+fFUZI;=^J?OOa0;%VnNyEZ9RsW?QJgc_kk#WV1AR`#m!U$4i}QF(;5Emv01b-o9IL~N?KY&3v#<7s`B z9iVnn45NBz1odk(lkL};4F!E8?5(ysw{LLNo-0*JMsfaNnLXGqR4;N=(|dM;$=u-7 zIa7U;a6+>F{yYfbZe^`?g1bY9m{dJ#)@t}b*q`kg&y+**;+!iKs8BtSWWHU3ssb?0 znj0n7cx1AO2KIQQ4R=F3Wdzwn>Tbwv9&*>8iBNjD=ye^Ls?j7qy6h=EWTK+Dadj*K zJT=|ans0Gsr8zR7foi&JUrR?;s-S${sMPpAK2$O7(^nI<2~n2ABV8Y*?Cy12oAXR0 z(v4n6kF^B){gor7{^vGjw@Gn_lB5x{=q1OkiHIJ>k4~3)VVAZVo{d-gjvx~y&^7b4 zj`Lj(&r)>>V7$@!eoJx5=tlcncU~XqOy|dgJyxD6tO*P&^bX}lkddJ~l};3wVMSSl zO?7%%lV3TaC%#p%_}MwWYhF7q4{Hoi`$aZnb!Yxz_as|#+E~3m6=l7rXo=A z8h#8ZE4g096f$TnJky*V+vXed1NoL!K>l)JrHh z^9hk+BH1kKmE*bJde2CNvgNmw11LU}mQHguHdXnFM6YZY#i*%zzw@u36dSwU)ZiS? zYEuhnt1w-azQG|+^dkX1`q6i!t`yvI^c|?OmJEA8W)P zy=f)g^;WoOOn$M4pI9H_Gc)B|KL34KKn2yy#Bp^dbG~|38 zce6J3>R{g+>BzQYN2<*-hn92VFS^Hwg8_C3gNC0c<>MuSfQ}mIhrK2ye5x-(!k9i-6Ky!lX@q1hdv6j0 zQbC51dK~d^_e!^&`z=|G>|;=D!Ie$fv|p*93@5HUnqVEQjZ6Q(*f5zw)Mr!1AR^Iv zIr3ZKERgW}je9Fo=j&3d&r-9(S)c(-vo^m}DN}b(je%nOw4RoNqJZEhv3kQm0>DI+ z<(P^xHqXCKTZ=>jo1zn!v$n=sfQhG^@RV;R**@k_rJlPLs={`vvr3-U#HY$on6$zM zmt;`d33~cUv6OtN{jC)d4%#dx)>nQUnr?`4SS{5o#jR%1i`PO^w>|8Z#t2Dl$es6ULM4EK^BM*(gA*K;t+sLl|INo6A&aU>cV6v2W zJJ0(K33My_aP4fusDZ3*^e*Ic`2>g4n|*55py{c6<z-X z)=FC&uq++2Vmh92E1ll`*lgQoB0$msub5$}B9}foG$8}mUv=dTnoE4Est=Ehn0dDe z3p{9qR9M>iT|IuyPW&rU|DuH}qc&OT4gW&@AM9k*)aCk74rLcJMe}X~tn6ZMT!v(i zXX!F`n&;=(6De-hfu|!1yBS#!pnzunkBBDz0@h*s)c=dUw+xGNT?0l1DFaZDP9+5i zMY>hG1f@%)q`ON%MNn!eX{37y=@yBhy9OAAA%`3oVyN@7_Fj7}{oq{Z=lSvdarrK1 zp1Gg9`|T?X%2W$cGPeHEHB&N@Gzs6K0;Rj|S7lvPPChzP7WbGMbzfqnRz9 z6ALz;!epYE1xlFP4p;cHX%Dg&?<@8kkTD>UGRf3^0g!^z;1C@su~aH(G* zZj*4Il14ZGMbv8JRgP~?^06?w7q0UPuLRMqYrUhdchl& zD&{I_Wb0(3ezo$Nd`-GucjBXqQa)dQz*P+kt3I~p@iWjFppr;^dZ&6<@!t+XTG6$Fs-OgLX{1W zXy}!mH}Bzwxvr0*aLZC-r;B4K^ZrvpTTTL^{=S-0GC=*`BEobE7gSS{(4Mq;5hRs& z_VFkIt2{#j;u-y!&1}}r@F?-V%I=A z-H#&J(zAJhxV`QaE-Y=}vttv3lPy-p_*48JS%E~*o;OrOR1 zQS~r~>g511rsk&;8mrThF*xWudNFB+v7)#`m9lfhP}FzpT6{3S2p9&z4_I z!ceQt9UxaV087X0!`~nsaI?Ey_LZU&^20)_2}b0{6}r`G#-R`jI$5t;d{JbTRA}G0 zN8d@zln-$~bAf6L4TtCVK8Vv0hubhei5Xq$eVEV7ds5h_)dFRld@2P|_3Tjj=p#-# zHuWR1{uu{7QIJqxawN>3TC?hQmhHi}T&nAQy!q3Hd3jFn249T!rxufnx}U%{$qGg| z0RJHDtelwV1V1U-h;6)HeQ+PvcdT05)o;Kqe4WY^(ylWatKPez>(;VX4*%#7B>&WD z|5#*9B?P~hI`b5N*EVC*NlD?27iB^1tZ?s_8gmzYs0e-DHT@BH@lZk0Q#YT+i3a>vn#li}^D$RM^Hdh;f82*#0~t9yPvGL+X3urP zee*RB@Axb*@dr+OG6KQfE{p$j7HsNwZdij+QQj)A77^~o(xKewk?Cn`EV$1oq8Jsq z9Z3{de#TsY==6}v5@4R^3H;f(#~Zjj2$_xAKb-0qM^A}#&_z0&%rnV{ifmZQso&Rf zj?g0$m2Y4PgrmQlj9{gs+3(1gU4Nw+%bV}a+8BuTjIN5;6cHTtqfF^b+IdH@A|0W( zsT*ww%aC;R>AA^kk#~CieR|-m=_-z*Qo|<83+EJ6tq<{=Y@AcSU}$*yHIu0VyWR%h#^1ib5vqr|Y!PhZ@6zx2 zwU#>m@ngnXT8h+#wVZ}QH(Pfvubx@IIc}nNm;d$- zNZ(2m&&AXqkl~RnOO}^hkh(DU~nmARnmWJX7 ze^Bcn##`SU<+u96&}voS6RtQ`)7_gMxZ1)QKWXD1IMgu1`}*-j!>suF$~7LBJir`I zyF`2FsEway9&8<5KTpa#^cb7(MlobpE{yYPxXz`n?N(9|y*E8g)mnMBV^pq?#sDc; zg-YjVGgV@bwc^}+rp(CJ!X#_3d2~imgl_{)Cw_EQtZu^=wkGS`)Y1+mO_T?ZOzoA! zruzz>(hD`LujG=(^!00&ZbC`pyI36!rT^+X!o0mab@t|}Ag0S-;m|ES3HsIrLjQXG zL$Uo%Xhru#Y-BO=nOmBPnZW5l$)vTK#_S$>nH90;y#Y1eUy+<A9?| zY#g6)>mpvEynWb zFfDSpMj+J7TM#JRD|0mr-zJZ?2i@{5kJIi(bT-QF#uwZj@?S$^%{~~5y>;F&#fj?j zbu%+p=_NZ$@2Wm#u;0o-jj+SB06lxfpBSQkvkRR3CS?K9%w4eUCoa0y&V`vwg1+Lm z6XlR43+svU$%)uoiLAe3ZQGXQ_P^ubR&^kXU#<=wdA=LiPs_KaZ)W}JqdIE->RHO_ z#$GQ$)d(q-UB)hc+{a2Pokdpz8qCr_r=&X-R%3!NuSN^4s#z_P-gn6daVAhsKAM%WwI4^H&L+0F>a#G*W<>W!uGIFvYY~lu>CkbFbd5Zv;jhSl z*>^m{ruj{M@>|)ERx&TEful4Q%xP(?^2!N)IE1zcJsdYq!#8>R9lNH5d?lHXZ?=c$ z1K4VHm|*AX7G%&(57-@tn8kP*#662K)?c#9IUp5q zo~AaQU#;gIvflTdda5#2Bgx_8hkymPa8GD&Q{{aA zHy3P$(6)ECg9_-KI~wz9>wIRG+SNMFTGNMRDqIrX3@}UCmT5%IK9@Ew9x)j}Q6cRC zGec_C9TTzMBayK#TET$K)8G0c(cx>1Gsax;AC6v*Zr;s%@%f~3X+h2*&l77r$R zqc5r%Nh{htd}0UgU%gLe_7~F^5FIzfdBza*M_XT?K0uTV-A!Y`P>Nr6e^5Oaq*lsb zyCi0-ewf!CD->W3F+P%04B3|*tre|M+wG+t%=iIiFV1Mph^mJ=&cJ3(;%roGBJxAW zRH%WLL?y*RvbxZjbah|UK_S*Sv-{>;w;y z^X&SL?B&HUj^x}i@DzK2VhcR9mHaU-oI%!9@$f)137(m|wyj!S3-61BKl zuk0YV;h0zGax5a^S%1CO^Nwg-C$fGN390SRDlT2Wb~W7c0gfEK#qNr1pt+}wolS96LeW`Nt$u1Q)!;0@RPB)O`kq_XS^+|xx_%u(-%s$e z$m@v?f*LxH)3-a_Qx+uA<%2M~{nV_+_CHgL61dpu zwwhUiN6%^gQ#7=_J#b$$7Hp7#IT2dM;TH6nPIyYtKZn5D@38mYe04cN!rcrRg8kK2 zzZ5UkF}rC-cJ-NHOGfwf>)F7Bqu}yt1wu;YG%s+=c=U&oswE~14gJ|xRR74sDymTx zkv-O`YbtJwWSt$UzMLMb;ZfcsZ&f~qYuc~~$z*H3)PN*z`mqWlT`7;~_{?J!Qv_O2 z1NBw6J}QSrB1wRp0F)6Wg&MMZlJM-ARPJtO5$J4|pwz|{YspcWmN{ibgOZ)nq`MU{ zLWfpYtEtp0BY=#_{o)9l(^7A!P_wKQ3@AAv?2lZ2CUa%LGAJGFs_=d_&bF$iQd7Z`9*S?WfVuaxM$5og zV^F9<_SFflQ{%*!d1>Z_3Eq3(6oTYFcl^kY$qGd-)A7p}%lDSr%GFRie!sG`aJm@S zP__zn)UL~Xvij0sM_}|r$^nlf4AliSiPEH-=2RD&30du^(F~2~@z>*mX?AZHjbt_w z^20|Jd>x`FXK3yk2mPR06e2qmH-K3wexRVEy5IC=+z`8Jreo%T+W5-0TZoB^ZSo!U z$B>?FIoqHkBZbgoL<3Q20c6>k`xHrM9^7cBoz8rNL*39bU1=OFJsM}?VyqO`@v=_$ z1|>YPf-hJ@oXeR`jVR-DW8NaXV008Ya!h2Q!9N8W5gSjA;kcGh&{ieks0Xs1N@QTOa4sBEnjwIN+ApLBQ7HIB1$K^IPpppfs|O7C8(|eePve{}h*-8~b|8@Z)ENs1G2f zq1WuOA4Izuv(L1Gm1!-L)Qwk!?w_Igc)!<6y)KiMZGto*g`go!ZoxLa2({Gxif7D> zKnQJ8C6sIfo6S@k^Ks319Jiz}(-+S8h)+)GUVOR%jY{_Nx_ZIseO?5O|A`jLLq74) zZxxpSmTF7ZY<$;sfId$r&Gya6xI%9k2}r;z_Sp(#d9*h#1Upq81x%%RG4?03%kgA; zto8oYNhKkJb#Elu4+WF^s|<4mv*Br4Nv@WG_c}f>XJoDKh-0RpUZG_9ve90jS(%*6 zJn4}N_Ot1$>FNNI2bxZeM+VP&gio~u-%Pz+Dm#?Mg}{#9_@kleqd?7dZ$ICdk2eo5 zl#PYf>s!ya1BUqmsZR@5s0FlDxu4mG$GM%#6+)HT(xK#>=Yjk6fo+*oi@D7TYmeYN z^$o2D#$5M=)aG!R(F+G)q(M!B$Qn`T(fGa^t1qmx9^oweK;4HUgUL4=|6$r}p3G10XAL@ks}eEmsk za^F}ndIxI-JXE$kq-7V!pW5j}B}YqARIH?2|1y?WUca|760|DQ9+_j-t(vnW0R0te zhjD`YorjL3d%S>#LCmFDTjfKIk4oS}v@UKASe!+-onWu=*Vf4K>qE0pV{w{h-{n=K zzN_n;HXz;9``noDzL@p;cozwf-u66(bkP%+^7*g)`(3o7VcJ9^p@pY;s_87tBP~Aq zb1%<0#d#I7L+|~x6gu=OQf*mjYB+lfC)woAM`=nTzvPlnzSwYoZT8AD>=+hq7ya#> zuR*WK_Ch(bx7{ujb>H$RL%ac2)R(W1fS(ectyK1nKY3Rhg_<^1Ihg7}`(W<1Ti!oh zyMA7v9eb8#lAGA`a7gdhfhbvwkVILSm*N7$=B8=o{tQoM*3|Y9hwp(Rn}p#3tusmOh`EmH#91iIUAndF3d@(Wd20Ni$x96TFcn+qm!0O`!7s zBeRi#T-5vBx9B)It1M$~$>StJkkh=2w5O1GZ^>T`ooO@}z;xNwA7+2Gkfe~vYGfEo z7f9@Ee1DRcS45KhL<+ZL{$*|S)f~lGx_;4Fc8OxI+^VI;^_LhZ%i!SP$eIqCq5#I( zs2x+nPSEFXk0*8YUVmmCM*;h7GBYzfOJ!&|aaF_zkndd3(btSSa(yU);&0c4Ot|cL z#;l8++KHGP_F}m3Dqi=gHeZ*#T&G<0#I8Uddm(Q?zo&Dhas|;amqwWi^}Q2wFWgRE zU2Dt2XpR?olt41keN=i+_4I6DOcM3sp3A({Og*`K{dLPan6X10k>_Ktsgau*GiHg` zYt{2#B~hm5!>aOMKB!%Oby{+xUAF!4=9PE4#t*(-W8K*vR?8}t3?&Q9Sc;}CGr!<9Vf|{hSD+9t_Cr8-p(O#t}M7Z7iFHK&{yJ~r!g)cE~^O} zdi{a{l$c13F-#pJ2CpxW35lngso}BnzP@Z1X<-cd?0oKcYJG?muc{daj=q0EHD=)RiumU*)9Q-R>D&_I!4zI;_U1Dr)mQ zUpfi8y4Ie%&E~@M%+Sr}Y?q$bQ9qoXuEy@WR$R6_;wilbww8Ign3ii^S~V@orl5dS z*QdfWgkFdFCMDN+f5X%xiQ>dhVIi??x^ek#MUFo(#6eesYQq(03#4=Xa-43`mfZem8nVP9+k3PwO{Q=4GAc$$QVH)+(p~25mykq`{f2E|Toxd~*J(xW=@|{1Meahe| z%F3cd!KP6OB$G*Qp;NDWvyU>_um0;dgM@ra^@LdR(IzZjq8Yc^*q_wAJ-Vu;i1YxV z`&_m!Ei^%l53_t!L|=>e>4tv(67e8%Rr&)seYh$*m%d&}i6|k>agvallIt2|y=t>S zbPO?&BkNzQ(YD&&s>R!q9gqHz8cOgA5iH)ZMzsx8%`_`X)N>I6|^ELtwU=t+A%I+~XBIk6F6bTJ?E z_UCnR{Sz_kv;7?9ZXyb(O8N_G3!US=YT|XO^*ot(32c@5Y2IIJzv{C*w6MfPMc-x< z*gTTsY&v-*?S?mrH>6#`uL>3060`Tu!||6I&J4QR$Dyv`8%fBgR6 zFuh9$0R6w{_`m4*KUoyO$^U=2<5|$_9IvOTiHV7AwstNE`Ij%FpGi5uTJYchGc)qL zET@qH8vr)pZFz;?-i8@J zy)TwVdMqsRJ!(v{``>EGf6dNl><{j3P-*{f%g!GvWDg9;)Cq#lRVMuNh@@kjL`^re zL|9uH?{@(I);&)Q&Zk!lzt6g0x*EgXq00_9p#5URQ@cNo{qJM5{#*K*I$Q;Pk~g2i zsW=`+)EA_!h5yaKzYF@zeW3Cx>~i!9dV7wR$<1&zkUr_r1== z5151NUTgRy8-5`+mY<%_nGCdTM7{{MS1FkzdAu2^Y=7$#9YoI!oa-;B$}ZurQ777* z@bc;3_x|I0drjw9^TYq1kLwz~%T^>LBiIUQU8Qt-bTMm20Ux)oO=$S3r{fZll;+M= zP2`MmF@`aLc@o->L@t=5+yQXT3eh-yzHQQrKYBiRj|JI3V!{IxOsFe6j{C#0Agu_r zBj3ePB%VW2mwt5UO)cp#&Rx^hI2W1YE$$ZC`maTnzv*pd03XWNXR)3(v(ejDaFTbV z<6nZhv;qK>I=0SJTTIuwBUZ3NrzVDiY)vy zX8-$b)A#}Fvl*(%^zO*SPMz4zlYm*1y)aa0`L+$F>?oP8HoY^hs`fiD_)P`LQ!7p~e z%zA^@eFb@F824{}q{`@)Q*1OW#>Ks4fawgN%E?hlMIfdI1LXY`4_uPGOO6(00>m#T z^QhASD9&^|cNCUHiaDYLl`g#E6BPjFW&Lu4nR(K>ea7T4h=`X<)e7|`;;IU;kyb9zejlgz>2dvaIhx!Buu~g-o4o=%l(9&VIP**zq#fQO8cKM08DgN z9ki}mgbdd$x}keT`Q{&=`R^kzy-)8Ut6w||)AssA%ex`wl%swL@= z0^VOIK$^gtQ{#y{f}nbEtNrOeEb;HZ4X^`KtNUHuMBkGwE4*A%9N-3wg@1T(|9$i~ z0KBmJIK7YdKzj6(o2*_q=e_|Ir+JUZA6WkHqXg^u-kF7h@f$}QvQ=}mMEuMfYwvxs z6#o7D|JDb<>XxgLBqd#PwuMgNiMXgQk3U?@r{`0b^d9C56a8wc`n`Ao^kVMXUw=Gj$ z$ay}x02HvsQ%P|zuY#EZB^m*i$q=#||61zsZ^lX^!+W+LAofm)EWf};O<$jwZ)rH5 z*y7EbmlzKF5!JmK>|QrMF|o1b^C*@5Z};)P(F4FRVBz1{i3jQHL)M5(I;K9Rfkk}z zLJtgp(ZC*l-uoC92Eua@D>gz-RBOrg%SL^9BW7#loF_9*yK!Eq_FajV#gBqAH7cxZm#@q1p0Rrz5B-&Qs=2p3%@pJtP#e$gyjaT&SyLm z6mHYf0FzMe$$0af4YFGP7c*VMxz^{Zx%F!ZKUt@(U|F5CpkcKX^D@wl0Hq{)lXfrY zMT>oZ9ibBQC~(#qLtdlyv|_Q+pMB8jx6^TU%| zCg>oV5lPPMMt&=Tk{5pV$+;YM8r05?I~cau*Xf6H??3FU-%rLCy+pVg837f-rN3?~ zrpd5hEyX=Oj<;jGLC|}dhp-M{#V+pSb^q%J*FPL34DbxN=KZsi`*+ei#{pN|0IdhE zqpyPyDAKDvM`^cm{U&H>SbS+yNbuL>X-hD|t7s}8T|K}ut!pYYx zCfkF`7QDVqScSN~NQ&_Kb$@{ALZshXHa?f33GB_jVHEFcU?t&9o-*_jJ|O|vP=CeVSE4{k?b~-JWTS&)ij7u-2J*`-@TX_AS6C zj$A|<&KeRq>ZKv?5kG6ds*A~@q(}c-1@gx?TYxET+M7~S+*rrh-%o4;2`>@)2}6J| z`*GgnwBv^K)3cU@tvrj-mt1k=d|L#XWKHU%3JSdUL_p{boM}_aCyV`$-~RjYEpW}g zCHgm*VUch#%mp|TVFC~S#Q;le0qDc+FYB+$WZBqNKepMaAVgej8?X3p0O!umeD}#c z(b#?%qDPVI5C6ke_CGVlwAdHdM^p?Ym64vx*#-a>I$ZKH%g(dM|U%BV`4E# zcfK{lX>}*}hzON?(Y$q!#y9G>mNl?DSdLrA?sm8+-ntD}qZ$NZnV>@GJW7T#1 z$YIk1&TucP!8NH<%gg*(N|eW3hGci`zRdh$H-)e)Q_hN)y!PEf)$+1@_2AJFV z)!8B4kDA6u1ux5ru^IxZ<1m}WOX}J?=<0mUyxW3CCv$&=5yRhb{w7g)k68eb5(Ksp zBw;~re2{uuf8|5(-&FnG)C^<-8jKkwJhoR%?@^y&;L$-&PH^#|-(&$GKFt;yZ?2+a z6W|@Qe&|c|Vh-^0$GfvTUDEYB9of10@`HO_LapyF;|F0t(=}Zs3Yu1WRlz{iXaK&R zy;~2h5WLJW&3ys@LXywo*{jbB>~RxT<@|6-d?P~-0D?FBuGrZ~=3b2hv0HK7BenKR ztgZg}-h&u7Bwl-Ih-M|(VdNHLl!zbfwsSJ)@5=UXOg0Atrb2v>zcIvUlP#~z%8x%;}|3$2+)W+$g8A%SAZbpHWl<7^>(CFCN zw7S~+1bC637sVep{qJKk;L{rHKx?%gSbx%1c8Lq;6iZZK_NXkZ zdtuXtbEdv4PL-|;t>6-f{(Gm??_vTB$J+qnZIz)+nb2!y!0k~HqP`uIApFA72?9?# z^hYvA)x&XWE(rqZ?NtDH_pzq;cReEH8g+^q(heKPg|p{1L2@Z{3~t5`CR`Msrg2@& z5%#aof7!PcY&Nbdxh)W8wILloe)^;)#+$72;uRng<>%kpOYL3MXsNs3pojH8>$`pr;PUT~Yj|wGsN36F zB3nG%>HcoyhnzI$lWo-3Svc7jf9$0DA!{5#k^=oe^XR^XD&OJ4XobiA3}KIbCs>ba zRsnsn$Et~iK1uOF@5=qM|ENVewAM6#eB6_ioJ^A9?%@+tx7iZNvvaWBT|@{X#JOtB z*q0i`Ek+UE!wwdClrhLU{xQRDD zktsDKCe!)zF_-?>N|^Ii@ynDU!(RwbJW#exWzyhC-mjK)+buIBMOzpB4^ zNt$hn3()!P`&($wSSr@A`E`hh)tZKdQjNqx?%k7wa6>bNAFD zy2|hkA6F-%p$kKlch=Qkl0GOCI)5b8PeI_4xMTHP&@(ErAUpD5&-kM1Z~QV&=9t|y zM|Hk;uUKs-w>ozDdjH~|%#{Rw62Eik-3X201JXlYNQUfcrWr72Sqwt4si?ww3 zs%$E~y+?B$)Gt2b4PZt+ls4b8SE${>-TeL|nTT<`6@kK)CS%&)AA6_Fa|y+pIRHVj z!(g+jO8ho_gTY*L<4J^f#n4?xW387rN9oZ$ zj}q>^*w|dbtOGRM=Foiy7|L2hz%7ftDjlHQ91qb zSS1!>GRGSM7|ODehGY(p+sr_EI)5yL<|zQqKDjPZW8mV1;;o6Nzy)Zvf@p;jx$0l^ z_c6DRt~E|l%0yN*g>vcL(ratZ8VULQkfWw-RbmFk*T{8QP1F_8+_3c*T?P=w@dfF= zZ1ZgKPD@uAha;Aan}I(?6XlA<@doC(uwy=5#yLy?GPp3dD;4zm7iaHAK_+hE2KH9E zlR!Ng6XS~Xt_A+&{4{L_%IWU%z14hK!ISzt*-kkd$sKikyw-yQ2}+D*x>2{TK+Z_F zi)%K}@7plsxn@Jhaw$cwnW*;Ih9YvMd2|or;V*<8mR>CnbIrSsGa1&z3zs_|6e&7R zrmlj#lj!!Ehb@bU3V=9-XOD%nwhlU5-8TBDSsW@yfo#-vpNi1y>* z{KnCWJKWIvw$?1f#VTOnHekJZ>m$o+dKopfN6h)6yal2i*|noZDkYJ8xRIJxx{sto zA0}15_yQGar1VvKbBO|^JlEJ)(LKOmEDxkV`vt#)Iq5xq=>WKm0mzYk5)aK8Ri7h< z!`+TORQ~i5I}1oB@E$CKR^2on2$Q==%4ZuTr2n|%ohol+UmE-%dphKLRFvFC!$SS< zcJ^=WE zvtNEm+8l3$VO-_!)+@&B0+Eb5}PceB7^Q_4brN0T< zw-KbcUz7gf&6I)K;+yRi@{2IXnv@T#^hSs<}@-Cvu62*McGx&7+raEF6;)eYJe zRjYXwkp$)G`>63@)6qG_L+6{vm*fI$uK_G+8@kRa?IfCLGs+l;g(w0hd095}ipFLI zm;`8*r)vxV^%p*;|CRV+PJAnIp5>l{VQJb*z5f7k(D-ZE7^9}#STVCl;?+(oD#mdI znC~rmCe7V@{?1&WVz;u(X!n@@r_4jDWr^`bfR$^zIswvce9MEkN^(0ab;arq;#S^d zevkzMSTwO?ysRp4<6<|nnFg?Ku-mHxt$H;iF>+!Q3QCUmt6uX!)Y&}lRMazj-1@mW z`2blmeVy2?wkpN%ViVAO8z33@px2(*Enl9|g85KG30bK!O#x(*8X76yEocN7vt=(K zFF{Ik(Yb|Ys)b>c*Cbc(Fh_YmqO5zw?%391zpO71ra6nKY>-8#`cTqq!_^$6n7XMn zqv~z0W;9)V-hc%l?j(&3l$4MK@@uo4~2*UvQ zBZnOPm;#4PIW(q~Qo!idoJdhK5j}fFJJ3B=#WdUI5 zUn~?k)t5~!u70_`QuFxYQ|RqkfE&8K7FtU~Oil|ADouHmXW4ZZEeFgv-0|qBTz`Mc z-He*a?op^AX<&?pKZ8_UyG%6}zNKrkuH;I#5`{4K2YfdNbghY2_akyZZ`*+78m~Y4 zb)*(00(gmX%z`&A1Lqk)KLvN5#y#C8o5sDwCR5VxEVqiuCzc@dPbPq|dBZ(D!V+gw ziWOh(?&+yzZh57_Ks>&hhvU1g%o8Q#Sbehdg?mb$ za_F3}uiS-7iQQ&RH0#tD+1{=Nz?o24k=dN7juMG^=-NKTW zxbv^*0FoTNnQ>e*?@x`VtKXkx_VPAk*%)Dxc4j`L|5EL=5CE@wZ2zM!YdVDP3gG8B zv22LUNe*1*6bS<7F*Pl~PW#;WnKXejnsGjN%nxSsLds$#YI>_&q3HtG19}pl>Xy!R z_2jbFe738b4iS~084mFul)Ko|TK)zYb`WfK)A#IXakA~b_LQ7RhAbvb?&X2Cr;*2;Ta)2kq^$Xv9 zzG7UlDNgk??*_U4~r{}0jk6MaSM8<}N+g^O4 z1PgXqr!7m6W`CN2&jQKL&tvNshR2>_JTRZ7qr@rr(5Le96%FRY=Z_%jo`B-xm!*wW zTY)aS1Ma8fgN79Tj~bU*D@=e;*xX09u9NV}AF{ReN2j=avMKN1?eIX#=NCls6lllx zl)-Lm;wY^wFi4Km-|^(Wgl>oH+TQLZsa^uIt&DF$hD-2%?|Ia?{-X0HGw<=-sw0m@ z&QJI#BOO6d;+m6SyVC4^b`HHUM2u#jXrA20ndf_#Ze0Pz+DnQOG{jpOUudZGl3OVV za29*}efoCM`Mx2m&6Ccx>QVLLTZd2@!1rA=o$y%D(2JNBSG-nwiJiwA;rc0Kfv{_O z!O?hl$X|a=1ohLA2a?5Hqoct`Raexlqol@WEDLb|yf)ax`~Ow{oO8|zUBfQO3zu`&CBQ4C>u8ZWqi_1dOz{739R7ZM zqYP+S$e!KVR$?XJp(`Y2#iV`|6s8jV+1W_4W5PTodeQjl@qn%2LnNj*~0?4tKZ{Go6U$WI?sNka~kzwBcxm zE5Z-+0FDgL1H1iaBJ=-D(cr-W{_c0H86T@^+!Ca^)BmQw-)3MsoU9JPB8e=#f4^#I ztgXFs9rV806m$3oJykWGv>ifrh5up=>?}wB3GQm~5+2zd z9Un747pss&)+-_#@j;07NDi5yF;k`}$NVeR(K-~JKawpaMi@28FEOIcs^?MKvW@#u z5x;(f7AF^QlIbgeK>+(u!_IfDfhl@obL_ z_8253mMp8}7^CJhYz8C;UFgF!YYW{fGq9~})%p^B_dJnSqdYD3$z~O{cD*RldhCny zOBh%L0O$E$nG(x6UQ;>C^f>q_xpBYsD~DqG>m{|Y%pz6CYj44JLaAmE`*r~fD?jz+ zNaaB})R+0L47bQ>CFz7_-dBm8)qLN6zn}r65;#7Cz#6`A;obxlzw9e5MweL3q%?^; z;9g<62Wy>Pl_sSR?Hc*SB(Xnu-vqEnS2sDme zSNX$*wReB1Z+wPlV_Aorr>D3CPtB^2zG})WiXVl4B34#?n#P9FRrHd{D+uv{7g+vv zKwwS-yIcG@WL4uVo4h7kUv<>vGj&|7iE2>!McTxbbAgwgG+d_iNE2#C%~FY@s1bue`oG<3+xgp1q?>oQOoUuI~QRKp6{iQT$# zd~1KogX4~M-Wz^#NF`Xk8jjTe`r;yAI26xXs+Vd7J2WNjugmO~FdwoRc(g3y)KzK@ z8x?J*oJX*1hTuH8MzROE7nsgu(#|>QM~a$5Rm1b9_8sR7W`4P4+bMuDqwYL&n6|eg zu%5+kD!;SsAFW-wU+s+9znd?|Eh@}8)2DKQui!sSOwa1;xp`O`_J!V~b+hM{r^>7&Y8%YVkyqi~PE=Y4tvC%(dkOx(vW~}_0u7UWE zY4_IQ$t!aKogDR4?^4k#NL7nwJs|px3rkGB2sL;SdXPB4G9;(?&!CKL5F?!%58TCQp;)@t* zfD|LnJx>_XKcBZ>5+;##uB;XsPUeUh0WNj_y=He-ri2(q^Mw?BJ08 zZbrB`pW)#)7oAju$b*a_gu=CWcyV^e?G9iVAY;Xr@%YBycJf~dhMyx?jL+_P<-z%c z#=bxLk+aw6;!+R9HV4W0JT3F?F;E5?>)5Md`Ob3oSDhy3`BRTkM|+yqJNN1N>!$hR zF37ewU%Osd&;5llvW$M) zkqtDhtZLZ%tpD_4+}Zl-eWfzUH_pCW zh7T6^*jY~Li@~P4C4=Ibv6^xbB8Z?HpL62j4UIGge9gwob{wo1_WD0(rGJ=S!CCdX zRfb4kDB!uKQR?KNK87X0U9+|HHLHOpLEZRwZ?Lln$8kpK(Y(<#--#F9%PKNH$;kI3 z!D-C;AwP9yAg-@~zD=v^v-#$a?EOzZf0dMu#qdUPC&CQEoU1$U$J-a*^B`Akos7jX z@upQ?)vaz##gZE&wls&c%z#Pyq<^{szGW4ay63Yqv$?)+w>6(gi9JFb@&_KiHo6MD zoUW4EPqf^x8BU#OP*9v&kY-qm0JMg6>HTZbK~Tgh`}!ZrwFAul;@3NVK07ssFuxt* zstmiRl*F@LaKj>9Cp#drX(kxgJ|d*P>y@5#jOo9UOw+&+BQ1{csw=Gn@*8@b zE^>};UneVi8RP0?hPkggPW`d!;YAHXLAgzn?wYmBdPviaS9CaC(wAK3Uhj+ z2$5JvM!}i>Zbn$NE<~`gPpN<&S#)*&s&1?c><_4@E?cW#5whDC=d{j^1e&#hVw~r zI-cw~Awj{hsQ^U|R{OAc4jNEd92*n#6JXar4e#^vw`V7?1gacC0Q>qF5X%1G+ilMv zD*JBfEs=zq!yy13&K?Lp#60BOM8#aHPSqz+1wr-#^0Vzc2wI$BHA-j<%MQdB{P~Di4xMK3UyMx#5Y9yh8^Zu?Esp}`; zH6D96dhGulL#7d9qT~DUWo3GWA3nhWIu$xtOJaC!?W7f_Jw9)~+&}uJ8F8F?GF#Pi zb{OtE<9#s5vLAcs)r2POu{(h^mDhxsgsCta@0^JH_1hdZh#xTQxScJX^&Ss3o!wy0 zgEk0$d%@0`RyYR*;yK3FS;D90SmiYwHLneN_#T&#o~}n@twl)!o*=SOdRQXPg6Ma% z@KahMWu@Mn_sEm|P1w0aJolF%e*t2yKT^fEag@VCV(ncAH66_!WnpzPd5PdhG#$)6 z-Jv@{!XAMe&=V+?vXQfnvlYLJ7NlEm)#;W>&)N2Vlcn!3@JXVF&*qQHLHp`F7{}GF z#A+vt?)*C(<&)FC_@*;&m_4+j^!h_KFc;BJuQZlx?rN0%-YWm*nMA-#L>gR@f#$0! zz;X96_A&JBHzNs`UyEf4=fMJvDb|43d4i9B#Ptr3r&yy;1KnUeO?!#HHPO9&0VIc_ zXB~!T`c-a^DdPP+`I)ck|>@vUC(` zVjfH<-F(ix{s3JDwHD)=)+AHWGS?fIal>#LjVvm3st8yA-7)h=W5X#P-wX)?BQ{Am z?!9fkk5?F==y)x}#zq($uYIDr89?SE&m$Dj!w#ko2pIWo_6af%A3*3%I|>uoA?W zMA6q%hT(!d(__!Z-wRr^$vDW8rbSwG1&}^gnuZ|45bs!pk83=yc=!2YMXYEZAZ&-!R@_7WLcB8szUx9Sn3bu*qk$^?WPxkL4Ca(axB zFk_YXDg<8%chv%>nbT=Iyz3m03P%|p$$jqoE^EClz4 zM{BY^)aREWj|Z+swGD}4S=!b4lSUd{J`t$8pEZf=Cxt&;{b@-F+Trho+a^@~c5XlC zff=LfAPFspX45z)L`@KD4lUb&UIb7K91}xGpv#xE-=2&S>pC~ncV@@Yd$@CHJ-ElE z>n%b(S?xsRk-`BmX|OIXEO4cch+Xb1b2lU;vd;OPy*Gzv-7?#6)E^MU@Q{WQytKjX z6Xv(a0k&C%Uhix$UzZ8hn`R}LQw((lwW~#3YO@)3hL6bx`(Jzw3=N@MjmV!H1eFmk_7J4%D_YJwtPbPiD(>lUHT)v!r?`2| z=l9MqWx1d8=IlyRr(?m=|vCe4QrAA1?7*ZtKru)$&fP!~!#U#9g{@3}nM7d`1j zrCD3MkE#K>6(!`Ki2IobYY98fiI+fyWOf-JXPeVXCo?K3}j>56Bx8q3|pItrNcMEwfE%*sf>DJo{SDz z)+yhSx{D!r);HG3C}zd{z#Lp6FC`sGu5i_MqFhL2$CpTfMfQ6^eD-yl_n4$f*?0{4 zQk6Rvzf#WYg9VqN;zEp_-o(4Hj!d3|#BKz*Po4OZJtZx!O%#~e_*ilv)asO(KwcE< zfjSz!0$f~n#a&e#_zzfP+38c59bzxjTYk(dzMeGh~SP(dW?@x8P zL_G;p9O3%Qh?PFgISH%{mgiOx$1R7G%-siS*v+3l*c(~t<|xF(!!4+kPN;9fZ8}$D zWM7Fi-0gf8ho0r84YzKuH36A-D75={dJ`MtL1QV!SB9eG4z6|<32A9~u zMD|d4f1UIhp47N%dJZESacAAWY|lqUWLIa+7JAmk|Tds&anES-_F35S&g^02L12MQv5&1s(f$58&w&GBmnVs_TwCRvuM!QE_ z)MnJnS9}d+I!zgc=2G$FdH=9}?O1DWae+bdhkQ;4h2vWLZ^(f~1MCr+W5Hf3=CF+n zTCE~U$H)U9dOyH{6P_N@*^RZvh@bQcyFThR`2bBlK*sgF(Xu++EfJ1s+<+2d?A}B5 z4NI-2AjCBofd+~mGk8JIPP}43KB2ry@ye2rte|WFTn=4r5VuiHrN7mosWg*G@dr@U z7#>MCX%W*5%#%riTGmgl_^V-$S*iOr8)$FHOnb>QExvZUlPfBF>Lh;jBC+_yuW27Q z5oN!y4>+`mO5*$F;tkrHb0I)N_MsiHlz?XGHO1uUgM1N6<^>IB{3xKm*;D#NJWA!H zOssZoJJBntN4}8k5u^g=_cGx*ogTCB7_GEn_faXf5BV4Yo>(n=gl<%lTwvYN2fFRf zV>~l_8lT;vPm40MzTFkOHi;akR81$SB8{^{Iv=R_iCfdJN^#HK^urtc9d6i*(uX@6KXi~0s z>US{MkvLB0hs-;i0-xwhb@>`*E!$O;X=G}APuRiydWjx0llc(vKb0J{ zv!n7NgO!GRtRc*Co(o3Lg+#$!sTOoYd?sam&ex8JW6k8k=UINUf^Q{wFp@hOjfKlRf^AzO=Jswf<)wbJ! z*D5OJ$Hf3;_t=u@u4e&oQgfkXMkXD&F@Igp8a$aR-i(=FU#_b~Nj#Iw;zM0w-Ty<@ zcgIuxhVSP@Rv|(qBMuS{A+o8IJwr0HWpBqmR6^D{W@crSJ<6VyIQGci9D5(*Wd9z0 zzn|adlkY!XUawwyzn|y*Joj_o*L_{rlZ)jtuO(h;m7qDZR_6$yvapi{R^p^(?mzv! zSyG!ZFB|{hV9{3hxsWo_Uj}xpt4P!Mla`6rfl?R~A?Eo&8q#JWvVzrMrK&O=jA zk1Rhjo4qSFKSX316i0=f%(QISDnOsX$VEL)?03#1!!O@WV}dkQ&XSJk3OlXJGQ^Jm(@1A z%H!K`n0$iKZu71Mh3Fd2Mr*IfdozzGK8U=d)n;y?VD<$v$7WfU07={bkVjozCbncz z01d4h;z+7RO)CcZ(~U{ToqJkw>5EF@5d2Uo$7nWE*d{Y z3bFu;^$FRio2#pBe!jU^s*i1FFUrtB-yktU;`rA^n`<;^icVvU_}1ZGMuL9lnBr5k*-I(^y8{0F zcJJwAZ>N^I6)Jg~8Lg7)PHNiKX5VkLgnyqcOF7eI>*unH4PXiQ zu0nIB#Q_z1dAI$>NNUt2#S@ghuSsw0uRDJG+nQDO)k9hXEgN*cS_jLe>vqO!BM9`+ zk8Q$IOb9e$#LV>I_}5yhKv!ov;o<+@BKOa9OOfjR;6JdPjabR#y2WZHZRok(46?Qa ziRjZ_U77%*+ajvH;;^>HAQnDUf@mWo?^vywpY0i*x=%DJQ_pkNCyxJKrVAeW>OiQs zF%Rq0^p%kOsnuFa2aSw}d_OZJ;Om(-Xd8)owGtw2*Wwtp_3*E2tfHi^eBDGQKv6^{4>bch5LO@2b@v($Z&tGGWhw0i(<)gC1 z@?PJN(!!mI;?5o$tvWLLUogJvn627yH^tEOl90M^dC#cz5@ah&dTV> zC-;hf=-;&=H%%n_7hY9Xv%9_Ul4oK9r}F+LE+cyhI4)ts*QM=2n*A6Ur!W3G;3*>; zE5o97qa@X70FUTc%#dwm^p|##S4XBGN-sGsTKBV`6ljX}vyul+i_;2p?n5(lh@vbQ z&Jc^kNC$(y6jmZIhOBbRVYqNIfQLRNz(`O%3LdR_eB2ly)ed?g))zXh?Eq^0T2_Lu z1Q$5Mhwe-VL{X#yTLoG6{8!CgK{T_rqbJ?>fdXjedYv*GGj|oO-Wkfpq$(cq<5!(8 zV5-!QI>>GPw@#PQ4LrJ|7X1MDOXmn6yD}sF-91!X(6O@yZ74j&U(zVW@H{nqS-AO8+6-`*OzEYp(LT>SK*L71a9N{=831m>L{z0Ucs zaeUo7%^98QC_NsSWxil}LAOJR2J@Njdpk+FzX!!-6+U$0Ds3K%zf>6kb}33>^p?F> z8SB-tI{?#YH(BT=-hF;marD|q72+_e_#!`=9eg8h4SmTjr@4oRI65WJ&ub2KBb8EM zn}7~kB=l-l9b{~}Vb>xa8=UhLxPFZ|DXU>8oel2hO>bRP92& z?Pbwyem2v2jjtz_Y`$zcaXN!40>bcvYLffbWbrTjP{?C*B= z=wqgfa%gLYi8Yp%3L>ULjrwuZkQWgO-4Gi5C7NdD64x?CuatTROF-2W&?!dKwr0?7 zD6_3SAspgV%)o?*Yru3OU`dc?*c!&>5lzM?P=MHD}RN}E+stLT$0)-E;EYXYz{&b_K#lG z7ZKr?C8z7|zJ3z@+?VSplW`1)Onjm1%W0{AVGEN{i~a?z`#d4e5MH|E)9T0v&yt4< z@r%MOlJ6|IZhSU;8mPA)A8izM_p_XV#X3@O?+isos>oj~^HszPY{qklW~2Fv^f|-< zNv^YXgu<^|M0dx^ZgxDLREB&PrEipWr;J=#Z#0HynS2l2FNFphq)66bAuTL`?x~(n z|06y~&?CP!PtQOvq0##}bgMcXTquY=jxXj(PpwoD3E`1;;$YFxK3U>TyY$bI`6m-h zYlo;^u_wsDbn{m>@!KVeFNi{jp71cCoMAMMb`R9Ph}^4<^^z83HX?v-i(DUqVgg<< z!SUK|CKlx9xA*m}L~5)uGDf+NKW6kZK9NN;PgCAM`Hkvq<0bNa9ECqyG9%%&+vdI9 z&Hk0(Jr9frzygVC(rn~0zTvNj3i_Y5mlqZ{`5NO4!wPCPv1h35Y1a^)5~0;l?!{Np zE!EZ4rJZqm(z)iI@;Sf|f*Nlxvn4ClcES*PMX(GVM>%3Fou-KJt+Pmx@KN_bb0Kv+ zh(yG(}3nv4djw zBo*AZ&gOnM@rY$>wuUlnk9-OU)#bMi4@W^p!a43{aN7AaBYLY+*Ru?e1BP(9)GiCb+@T#D`Va-LO)S1dTwlH~fXVKCF&Q14e-9DTgzV z!LtJTK#Vj%LEmxjBUB#{*?P5yR1(TH;`r-ggGG0Ohu+8J=Glaa^tI>tmkZoZn5%pQ zPd>^q?b7?35xelFepsS=R{Y~-=z23<8!yE@b-ZBKJgmff)T%-HqAa0QnaHUgS9ZL| zfw|QTGjNMh<)=X=e0UY0Ma$9TLedLEHd@Ph&_k2kZP(*`Uyrz4+H(0noFPl8v|FfA zw~SlL&BGxLuO8n~(#tL6N z!ADR2>#q{1Pn#2Z>|TJ~uFZ4rnr2M0r;R7$-`K)A$PRIHZ<5c$k!L_8#MW&`qG|!RRMBty2sU#;$Ni>st5W%=4vq6Ss6jnRF6;kEk2 zdD!0v`;FVGS3V(E?}}?N47+wbIERQO#rWv5RP(vajqjRKJg%lvbrnrWC5{I7U&4jh5- zlvXd_=y}0nVbZQaK={PbMtVh%-LlQMTo*j-d~6NIfyIK6@u@x72F-6 zDFfXc!{D?m}(TP5x7CcynFs@ph7-^CB3eV9he*)*%J}P{d zU{iqs7WcQFw^>Z5vxaQ;@+qXhy$!bqdvl5O)NTD zlLSiTZqA+CkF-rb%;39z(N@rHQ}=GN;>z*L=<-l?X`T5u5j8llB^wzI4?7D_!4V&f z%1PB{WjYhr`NYAj!H5PE71bj3UN;HpRBo`9Szm078gL8yxKl%438<~IWiS)Il(()Q ze)m6RcdN_at9ff9zLobed#I2klbP)NbkB>r)G!5u^-AIv-?NE{CqDFQ3L|Nmm9AUr z;rCNnS2TBFX;iM}(*o_g_g7?}7d{IG3;kKVy*;!II$+V9p6na1vwqiH;D)F4_N z>91at#xYUQ$@6iTnT_I3?Xf&+={ z*-^|;+#-3&j^}dp!~Mol1}!*$E+NwHHJGb9tp{(_O&1Hlrp9jph7ty)3y1+ae2_V| zj>gKtn!+FJMC!NXc|=+J6d?x&8k9JBWXHAfysZq1FaPGx!3YJk{+;a*63Oz)AER|m zw_fD&rC`a7PaT*X{QXDW-C~{)e^Vm-0(zh;Ad-4Z7PN#jkcCWyYOl11sh zd>B7#QQB5OJ2&Po$qPG|Y`fbJZepzQ1s(;8rsrqj31j03ocHF+kaMDIWQ<-#dAW;E z-_T54oVJ110eh;qdFIUL?xxWax}-BSNeBP>lANzT;q5#F!ihX>E+2*~ceP3LOdM>S_cGWe z?0qm2(NwB`O`-B|G7wi|bQDNvq1%++ggu+6NFa`(f~dt;whp4I62;Dj;+}x;zRl}Z zn^RzMN)p$nDSubcW{Mppb37RDUEgUB<&ooB*=-@w7c{n+CBc6MD}^b8{vu}M66IAl zlnsMN@RZbe_+I}Ir3X0VGwWu?l9bs6DUy>bOS_#GTk}`8Ru;Ywd{f+cpBDyOU1Hj^ zytlF6Um_&Rag+ZEV*9&SeZrYQA4Kx@wVNB~!Q0rn?KY!(UphTrl3{%YK~)k&!e!gN znbN+}u5aUSZDeAK3oe>90HnVaid*|UQ{ftenlb)lT+4z#KfZ72>-Q>l-Fsh1(LNc)M=6nVHr>vRd=FKCTYWO$nP-lkua+w9URrFN6<0(%$8~>x#e8dX;hD zZ^W;Xi)nN7!&@czIE!lEtb1zF&90<+TgGpVGD4Yim5QzOKnCjj(B45EH5$RIptQ-W zAXd;i0_b~ea5WLBaRRbYnhVuiOPS;=SGOrI^zTTV<>@a`4N?=SKo$tB=5}18VnO!# z5z_cP$6zYRoj$|p>0=-eYR6j@~JGVg(2Y0;Tr_KSx}dgB9Ah1ilB zs`?oBlIHR~YCIt`{RX4Z;x@*IuT|eW(Rv%Aqw&4zzqkh)7Ajlv zas$2u7}y_Z1Ng=w&|y~ViIKR)JLHzrb?xpe2jE=N4r|?>01yIt(4=M z``ie+$SuP(D9SaBL_ffnIc})NNCAZ%DVu^Q3hlw^kTT%=)Xr=QaNrq($(BQxrF@1v z3-L+)Oft@5{O_nEvwrgKajZs(RjHxG`X)hFDVkf=#c#;f@OBls9y#I(NSkgu;!6_d z=a)}JMV*rR;Jf7f54`Lq7~+%@uW`S4SKgajogypf{gi=?tD*Q=t8gFTMoswUnJUpp zK?*VjPBHASN6229Yv7LIG<-xdVuK!+LHTZuk_ZDfXQ!E!xd~EktxUYugsmcp73H0& z0FIm2JBxZ!qbje304VUqWS9Len)(8$Ey0BR^F!(*#BEie04$TYc;o!O%9!>Nl~tm2P@vcRwXF zzq>p8RxmF&yQ}xuWmAU!tfqgTCcwdja-$*0#td=8-u0EBm3svtL|+W0*2I^YTCwI& zD9xTZuUnmXjTT=9mCgqt#g>eh3iobWlpEGD4EJz}Jf!EgMaO)IKIbzn-ai}hnd>1e z=6Puog&j&{Za>LOp!4%Tqx#?R;VwHMvJXki6`kt7nI9wmSnup!i7Z#`CUkM);6vl^ zXE(fn7^qI@Ih?jRx!hT-*7-c8>OiaUwZ*uY+#sszmmsJ8Vqem_qwYFsP}X?n^auC` z=@V+G_Uf>whHfz`h&@6{E*^CR?{Z}?>9L1c>6t1%`%p2O1{3oBNR2^Y6b2g}YzH9$PhD}SU9iw}AysSQd_DRZ zPb*m)EW<%rXuzOQp!|g~PhxmXm0DDnB7l z(W8Q@_}|a}=eHZLq`D71uo$hUWK*|1mD#~F+D`|y=bzucZXf zJlKCt+i^X)hMoZEpB1qo=X>~wvkdg4^ai|(M1kOq;XIZ5i#vaZW?-a*8EKdjdf7YP zJY%YMPF(!;DgObSX);68YvVIPiDb3`TYz@~a&l1(O~1*s^?|Kee`kYa?1z+cK+)rO zn$`OCnnVjFnOWSn;X7geop5z~MX)AH@#VQUe;XW0J%GS;tC_B{c6!N3=f18~C;@r% zWxNX956VlO`YSbA4u)&QHV-GqL7BF$^mLryTQL-gsOe+|L;y-0a|mZDqed~wHaGW; zz-`IVx#G?pxr2*HlQWVqc`Ao%i87rQZlVd847F2V86y1t<>_*5)pDkLs5qi%o>dyHtPAUS4Z!Lr@q%zSMS!%ncLIQ(rTB&DirJC2bA0_>uJ_rHS?W3?SMY_$%}Ao zOUCB>rii?+*@#}F#J8QDHacz7PLKo&gf@=-bnS#Won7vT`q|kYE|wD=K5>Ckwd|an z3ZHx~GSiIzNx23ePfk0Tt&0n0MwiIita`HHQd*O0Z0=9;;>kVo-I%Q2Fb%-G5;$0Y z%0+`CISgNgCi94STS5nGilgRf86^t+BK1a)oS~wwk2-p#_m_v-^oEVLC0KRWenpDN zvJooLqOaC{If`K_H$XDMOIgEBoZ)a+o}85zXIdCUH*Pz)QYU^gdM(EBm9GSm9ZGs~ zZN01X-f90qmH<}J&lZlg7MJ^+HDvRAUEG(_+!H2w`5tfq-MjgMFlFl$P98q3r_Os zr`^)@rWNMN83w&NKfEFA^H&@l(%?6Rjr~aO^IgickY2}O{gO?m zm5*k{61?9J8?4$lUHd0o{E33~93;(UQK4+}0I*nsq)yutm^?~jKztaCtioMc&CRm@ z;uHYcQ){Psx$rHLRW|=)x4O1+oZ)BVx67`(!G;sQJz0Uuf8hlTG|J;U?&iSWVdKOZSrccZHKwob=xj}k^_ z)ldLQodSE#D>k+!pJzJ8XPx<2#p+C#Q2_9zT=wa{Rq|KuF3$YnjP6oLTB<+yZeGLa zavg*?9r~R5CuncH>4@8(Jg+?r4VLBkXqI87gdyj^bt-X+iz?WRQ`))t`I{B(RBH_W z4H_}22KtM=qbB%hI+3r~dA;&6@Qvttjwoje?iq`_9cJlzHRe72|2SC8h&g^@Thx

    am;lGdS5kuJHX`#`g<;HR`!a7ru{a^!-{Iz7thwu+M1FPF5|D2i*)Ntj4gv z1NE@2#rIoka6C_f++;xV82S=@Po9@lG(Gx7)#In&aPSTJ#_|ZL8cnDBw^?#*lo`j% z=^OjVT^M7x%CC$%Vq;X{uA~>TRwUyDhU+OW%p=(Lm~^9Wn>uqk*tOjXPQ9fag5+Bk z&+c}T`4*j~yt^C@Q~xBD0D&DcL487`16M52MBs{=+q*FWGM3R;1F%MLmp~o?90it` zCjn`)P47pYCHU3W#$cH30BN(Eo|Dt~Ef7vXj9?_MpVQD4kmmIh?BxLepst@zFFOT@ z!Qu$k>-h+{5;#>OAjoUd{kr|2mOPJ>r$H;vdP64yK~z1zPD|4BXHou{?7@|np65Yb z64)c?OkkPj^sw=P)bg^|O=hq2qI+1Lz{-% zcWRVRnVQBFb+XD~dp#^T>h7SMVNre;w29!82iyWY)DFQwCZVt(Ifslk)DB ztM0kl{p<{$J9`>40kVz(Oy?JkGAs-}s!lO`%vS4FFf;6oBeLt+uCfi`vmwAs850F} zDQjZoZ8%GT&xg;F-!LmP!@5O~NuT3Z*Rj<;=3(Vlc<}hRZxz6S6NP*8S96(Croql*fHEw8cvd_Cb^dB**QFysyut~2|01B1uJN?6GXO; zjxER(+CVkF!<4YOU*(vE96;X8sv%Rh67NG>k&;Y@@4Tg~%2CMGypOgHY)byiV8j?t zcL+gs9aJZA?U{Y3Yg!gLXrGtijEL1a^9Rm|I==OgfqBrZ!O=h3UF~_q0I(*}MkIIp zF5rD9tF+{U!33Fz-*75FF<~+~>uXhsBZ6lt25A!LV_U1%F!M1#>1~38!7|A~nFeCD z5ooOeATPP8-`+R4g!Up2$%X7>GqwoyqQ%v{idPSA24LWK8~Sk)_vkz z7S6hM{M+tY;uQosUJmc8%6sQC@-R`a-2A__GJrfV_mkbdQitS9y)cXV4rh4Rafz+K zv0yaG)6lqXYeQgE1OctOXYPT7DL4|^clkafkrzIU!L_jOorWymcJ~$>;IB}dzM8ux z)#a<6)XH3#qKNzo!w8g(8C-Nk&5q^eotKxuAV(lVFxhu|t}y2oYw2^6`p7Ti_4xOw z$cXFkW_WMJveS@!>HX+O{d3kOzgok)&gRApJINIvByXUCB+spMOSS>W z3!8IxThW$Knxzbl|IWdXvDrYo6~`i9-oxnGvYjR`RczgHvFP$9>+Nu7H zD#s69Svk7o^%Ihqbv$tCmNd>77!GAp$#hDo0gqb5P>fP>AUV@}?I$`I_#CdcEThQb z58O{)I>B!>RwoAFOeF#Qu7}T+`zg6{q;FbMnIzVOn(huIzLAAIR~#AkVis*+_$F4- z-)H@)B413jaZZMnL6A7`T_YpCh(OZWc6_@SSHOMq>aq&Q=HQVNK{oe$rG~m)dgmSQ zw0LK$@4qjd&pg%gG8))l0eLAfwBN~r0mnjj+gQj1*<17tcx<(Zpq;)mldmD=Ryxj} z39<`T(C3JCLnaL0LKkF~Wwl;vnQw5+7AtXfHX@b=@vGH!vNK|F7#>DvZc|#^X9;8| zJXEQE@V<1Me=L&w-hY}e$lfzKJo2yv4ohJ7B|vtn(^~Zp<-i|Jko%>2*t^Y@U@|{8 zYEF}!E?XEu8w=`^L2l2l#~jqMucy~B;jCOq^js+b1U4`Sff98bILQY#`o0m6wrO?y zel6*6u=y9z`LyLni3h6z&Y+#G0+R06^|Z=|n#)?AtUCnC^_mEFS&0!RgJ1?kt@8VE z-bP?AjSiJs4uj)XK?=E_pcg?oHO>&~hSN&rKDeF1I~8!(bY4Kuu;WL7Y$pJ;9v@vs zgLl#~qu?QRudVK})BT3iOKlizJk}32N_beAjb4BUeE>OkdVOSZ+kl#^YXt6SPB$E2 zs+yNJENgD(^|hMCz=637+6hoyjpc1SL3|#ndF~t$=Gm~#v)cTchc2BiIe>J6i`4Nx zDC7O{zSZcIWMJQ&80+C3XywxJVQrwdPd@qa=jFMlFUq6m&Nx6^<*3%dBp-*9bu+mR zn~zp* z0j})U=20KUZ;dKMB?SyTr`j{J3{cY{#SJ)4Gbw zDWfjqqMDB3j|za|A~h9!tD^oxpH4aeY^$|lwnXo_XTXj<25{QzyHc_t2idYYNCRzW zov6=)>P~U2D_bdXqp)Pr>ZP?qDWMIc5aA;!YR^IFPvP9DW zeA)s2RpE#Lj0l9&yRHlX=uW&7yM32goz$5BZf36Pag98H$=TQ4&pH6gDwLDd8khmr zIC0GVoEeu;omwZ#~(izq)%d;)&be8449Kg0m7^CbxHO^wy)A5 zZ!CU9W{V^8&g^OV(<_h5O2@|~b@XXT^l56`E+hoo7Y%yv?}}k86*p3lE?>%8!QySW zZ-WcA;l2m&&O*X&+&_BbA_k{S(gftB@2-Njql$P0bV$_0uBe-RnFFgjFzE0#N@!e+ z>RTMM?y^sw36dn?yz5?oqZH%fs1?8(z4eW+2AlJECuYv{$;V!JLTW2FgY7Xmfl@>0 zG5PN<9b;+84NkuQUXbRXRvPFR6d^Ab82T@@i;u}-hu8km5$S^ZPKA90&S5IW;5dfE zV1waN)zC&)AnOwt>P(%okeH?Luo;^j_!^rGbEKrcJ_(1bsTfF+MDWfd1Jf8^W3~s# zF4hQdq%dotE4?D|nXoa7qfa~d$WaaL4 zoBe%VCvaT?C|>j5Q~u6hy1e-{aCYULyWxH`GU*e?gWnC@WCde{mWF-hyLcz{8SvOs zrV(g52=7p)zbOU=%am~&S{n9Mpd2vszx_tb-P~BSTAcVD1;Xut(S2m1TXQk&*e}LC z72fsrvTTBTSPyH*J!Eot=wS&QmcZ^yAR4^W=D=x3%H*@#w_8_cv8Ll7SUb&*g=)S^ z4kUMj8BKk#%Ok@>Zs5NapcTDu$6Dpr3|2;`!)kC-_QbNz_xtL}PPI{$sW)yxXO^b| z(mXfWv-cY$CrjC$lhf&g;C29#IQ3PHprZx-b^A}sl5!mcRMiMBD)6e?bV<-xHLtm? z4GtO3>_{T0j+(M&^*8GRK`&)g(=q8fAVg2z>+~ zIYD+Zg=xwf)%3fSf!P3(gLTaE<9$;$LP5(E?wi+P>m|=grxP6w%~R(vy%Dlg9e8)F zvAk`%0tNz_KLKT}<2;7-S{+3;4_zm~O2+)DZIEpzUBX@bq5KH1QD;xE9&m<)0`tr&>L}So z_WPc670xy~prb}d1052Z8TS)NSI0P?g_bL2fnNB;$K=9ekGfeBng{F2+wWbGyAfbd z_iwi3_1yxf$Jr!NW4jOy9vV%vAHVZp60QSW^pzd(;)2bag5xdG1%^`CuJT;TzH}2v zoS~u1E1*>RMy=-rfUfOYDKSQ=%55x%7XYP9*k ztDvt|sDa=crCu_K-_YmB1|gGu@WhCmd-h|FxfW=ZAU*YjmnRJ%pqZHE6zcf3SKkoC zIdrxTP^msLD926$q{X1Tf?_HfM`j7mtsJ!ft<}+n)lJY(mAgoEgmrBl@8JSKfUPW! z{#U{{49k%4$pHXsWfeGqia8%72a}x7G?sxwH!}pLGk^}C?ayLtM|naS1N{?d7uTRm zZgBTi0Du!KHZWa8N3hxLkbn-8+R_y;65n^Hb$aB!yahL8;q>Ug@T9CPmgVhN z7UcF-kL_8;eXT-Ec>gUXamX@cpZ;&8CKF%(TGM6Lz!ENvj|SUHI4+NXkwcJ**r{bYyJhHX z2&#iDD;Ebd@~c17Cr3;?9I`i)m|atdL({$Se31PdCb~(;yUAqQ=*o@_r2@M zi_-Y`bGiwIgnMuDza*pv4W-J4pq)s^cb*`5JRHNY&27Nh^l%36MB8Hl-uETzcK*$`!t>%Ct3Xn;p(clLz;Q{?Q z1vUu>{}%j;o{B(-1-?0eS74lgf(Az8aSp^b_ZrRXUVlfK>hAnLV9&hQ0g_uFl_2?6 zVBNi5Ppjhz09r88vDCfYyNI1ddt(m;_)68Yz&Nfm8>b zPXuzQ`>j^+r6u%tlvz&jkg^r`Vgw|4US+UzHUr;dvdGB_SFo4>HnsVqtZ)M5te359 z2ofjls6RV3)-}t-clV1ByTCIJUcC1asE_^vI(D7zwD`vSV56vtpZIJ6L=^ z(fc;16f)|cd+H)I(g*D316gOx0(&ofPGsC~XDdj@pu}n?5NPM(g~yu!aqc??&3Q^! zDJV?WYCjX*h{ml~aiJZ7)v z*wwxxhjt1FQ!SahEa9sv~ zK%BA<#*)2oeDxFA^0wCiASOz1a(03;=`&|zz}Ff8+Bi_$7t3L=2^L}p$|JpqI_w8M z-j}@hH4^{_mW9PK+Cx96ih5fyU~mlpS9y3uYCWMNg26khOD#b1(gGNN9_YnOX&9`t ziff_nI4y~5>&CO~060ObfxcQMYNh0E3Vr|pKmbWZK~zS08-U*g#ApPyml3~)j$Hym zC1tf^Uo5|bwBs(_W`C$ln!oLXEJ##n>yXvn0mn`(-~3~#%zsB3T>zhbYrxT4hqJ(1 zdnGS@a000Da>>+~5sA--mhK#pjqfOT8t-nEJi@TW%NzCqS8-iZ6 zvd96RLtab`Sgf?6ures~N)1eM7rZkBE{bscFCj?xPK`qY8=rzt$$#Z`nuRzP`P}O+ zeWOHR3jhr24b3@AAp16S@^QI!IVV5*PFSu4$=3i<9f*OopXj?NnTx+99nbbi_vgSO zf3#QH&pa8Nu8QZATcz|Et^8|_jL8QJh+DaNDRx9W76(~jJm*ubgnJoRMKa#?w%Ud3q0KnOa zm=0aYLX|UYMhMj{Y4woDDSNQ z=jHQY{N=UR_{TT47;pgq=pJ=jYWaq&d<*U34e$G&zz4sraR-=%*S5GsE)P^^cpHw@ zLa?}jZ%6rL_^$9Yg#u=y@jHYDu7dJqyT_R%jc&Z~zOgS+XVmTw$q16Kzxig<*&T5I z-X`p5)XgR)c?|Np-FUAQ*#%xvhIHQs1E^fb?4RnQDSv#k)2B{tNa}3>N|oj}Ni)V) zlwS?Au38#FayYCTz#`UOdbxEH>*z;{C=jj(@q5&~58s$*pAXkk_K72O{1*3-O>XX3 zFX?c3SOSM7aKB2xF+E#QH#k^UAY?b!5@li&SW+hW8a(J0H0=O8&oZTVbDo1rn*y83 zPWP;UHL9I>vKj65dw{z6VD9rp@flbr0s%^Ue&=++M@p@OP6s$W?x%Ht^`!tR&r5j+ z1x_hRpdMEaIte81rB2*L&*$L)o&)35d>sSzrOSUIYlW0Gpm`XmGG45Ki_swYqKn z)?+qXIwrXRD#;jT`$BDZ)-^S!!ET%px}D1&UZ(+6S(g2{7AsT2badvp@+A030+y>|qGw{p%gpZzdC2kUqsNX~NF$_V%`VA{XcYUiw*Rsn!n4=Eed z4`5qudS(u6^45$FO`}T>fDQ-uEMo_f8_=69jTjl9R` zJIwdTG37La9FqG2N zECO12An-k9dJqTWyVdL%K>6(fBu}7zCXLKVk~3k>JtiF^U|svh;GMN4#DXaqAerYc z%t4dX1FXp+ZwAT9B=?>h;;>3Lo>#kq`ImPHwnQX2y#0#<%C){b)`Er`I93!sUTj0<$cKo|p9b|4P~(&q((R zeE^cr2B#uOPmqj8hr@cwd%y&XQ^#Up^m<9@z$t(gIrfLY@22y6_k0&p1KxiAtwBbo z(1RHOyE3p@(NC%y;HkE@BIRo#`06B4;Gl_R`lR>Nr{tsOJLTf(5Y!`^{5w|wt@fV; zD?I!X)L0L|KvgT_JzuzaU4NpU0eAwaPIhSW=w|?!p8xH95Wyvz_-}Kl5|`N?*$%oX*_ibL1SMD4veLHB=yM80dW6_$0WBBI(VlB z=0IL(@%~mYMjrjf*W^e4;-AVp|KL~U>filu<<{T&+cNg|{zsYor~h1Lf9wB{rQiKM zng8=YmpgBKKR91>h2^t4HGvob-ZNoc%6DKUWgllUV3yR=Bhf?LZ^oOTv!)ht8oQl% zG;~B%O#A7WXUa{L^J?tfs&qq?|>LjmPj868E1#Y~--zKic~IOgE-ygDVJ z@i#3QVT*fPy}!|MswGJ=1;O~L%F5u}g;wQdGu*{`q`K|m$it+EC2&{*_gn%Fs6yiu zHOFG63vHK1O>YLMQzM{6keMI{2g1t4ZU?wZ&}a{*(5w$+Nifh3_y{s6&|AdRK07^& z4AO&SMiY2TliBVMKoz9pjgzHK&><*K07sr4d|37eQb3VR$uhR`HDs3Z1Hyk~TB?J= z%`k}eE+TpTB>bf0iq+L_#U}R{;-T|{03d_M(iAY~I%j2AKs}FvNfWrf4@`2Nhciwr zC`nLmdM*NzQ}5dVoji_4r@$bX#};Jgc1o?#!Ll1;n?!(H-v=kdfp>t887o^u_Pb?) z+jB&jDo&-G!fZ_p(N>z4rbR;*|4w?H~2 zcKF;9Exrr9rxCEp2Qh^`cn$!W#wRCTP}c44nT{Q{U21%$U7hVtHbtFB{ASPr$g-wW zkj@{$?ujT>y;WcIeoN*=x-&t4^lo8#zs zUIW_=5P>_4&qSXK9Xl*L>p-%3JGm9(eDK~cjB+BzcjM8sryZl5l0Isbd3pEJRk?Y4 zw8?p^`?;i~)UDCGGG}B~`XI~1_#kTR7uY(X+xL5!Ks%8Hb6BkqN8mxQDYIbJ#fhdy zEP-jziOvut&j4smwShBc0E8mgiN%FY25OP~w)%-Kl4QUfZ}(ZdiS@gZSn+_rt&aOJ z_Lb2ODjF?d$0FZ@CV3jqy&b@_TBfsCa{1Gc!wAvfMhV|JGRXllt2H1;L&m0(FUoZ= znycY43{Y7Clb7Z|a{N&yIl^5+v$##Ok`xBFS+p~1jSk2o40u3tJqy`^&X9SXf>v{O z3+LN10&Xte7=?6;_wA-O`oP&!^5jRK^FVSi-`i4o>B)x3KpK6dc9UlVfz0Z2*CejR zg6>?Q-Y%uCKtk;B9bNP$P@My!YO*X}4t28h~Oi zI49i3V8BvWJJcf&`yzqd3bL(GZUxsC4~r8>P*EB!z`w}l1wUEuZJ(Os0W*9E(rvcl zX_3DX>;?%Rl1kw_n}jpB(0vjN@_CuOz0p~GplotnRSe90g21UWzlh4%gK|CVuGN!T z3F}>@l0sYQKVTF>1@b1#7sCEg5^_M*H{X%#zxTTU@Nxi$0FtkUef>CQEToP;5|qPE zSIBs+A!xB$0{yAI()8#J9bM8HMroP5|{s+_f-m8h~~` zVj|?>JfP*tBLDhV0E&9oSVp%zKir0xErX63>k{Ec;@Lq-9{-}mI*tR_!#_AalEa}h zhH(e(y!wil-2cmfiA%4)Dy854vh-iRDc$3?Nj*Pu>?6TpSXEN5{v$J zEK(hva{5cZZ(Qu)hiiy#vY7gr|m#bBlocsy6=a@cZ^>gVc(i8Q-WP z2oVtH*DY?@fMdW)t?<5*{kT~y6Fn>$&zvrKlueGCudmO*34z4E#eff_PM*-88FV0X=iV941U>)-!ja7>iU*#xlT z0awP&yNG94*o=7}Z>-@6$Mo`cK>biYDmpq3B;t7ls9L-_CZ@rlx28*2c z=|Ecprt!MSZYIE`&IP7(QY)r&!EHZ@<^}Hz2PFi#EWpb$p^mU+h!eb0MmSl*mXS=! z5tZDD%B_&4yOj*cPRpO{YYUQ7mwa{>;1KGfT2zl*w{tlRHaVqI)X3iUq*hF4TX$}= zWPP&$Desw}J%LrVb};CzBSN=x$@1XrnJgS!G_uI8Jc--r{~;5f?U`C0&&A1XaOy5P zQEDE;*mGj&whwE``=bM*6G42{78aM90;cMAT{^ZoKxwr;NVaSyk;61}-W@-7)CKX> zW2jBO2sZk5#=31;+N|3uC1e@?1nrrYP9R&)=v-CZY&{p99DL?#lv`oF*sIYoYMIS# zXoFe*S?}z-$2zwe0KM09@Af!?&^8z@>!N)>JT`gjvf$v3-!wITYocYV`?x&w$SHaD z;uCH*hHhU2L-6&t-jN@^@uqwJcF;)>a{}xdGRf7tFnBzB`VmMFh2W!;TfwtFXiMU` zrT{Le6&f8kxjc(q2qR`(MiPZ%aSd5JN+UnMg;rw(>vZ5N=R<{e!w&o0XRt=Rw^ z?JVY`J0CKRqv`kS_i}zk{_@I8^2)Uz>^RFZg6fsl>|j{M`@b+5lX2`G7LDz>7sN zofrFMt(HXaPQN|Hj@mi`o(f`R1iRMojhXN2 z0N}FfSnrKm3H1yB8UQm6%4u0jB0#HIt-LS>xfXzeED%@~VB&JDLslwaxkC5Z9@pdR zsOzzISzKL|rHR`P)d&Qp>+@g*?Cy`Y^%$ZDwZ zkPPz5klhZZbP=FrJp8mv0&QXnef=F9%0KK;d$;FvLblCfqzaiP@ehoM1)oX%g>(C=tD;AQnFdKfu zk{xj7JTWBKPKJPG3i2bdS6-FdfBuJfAI89V)uU05AmMZ5+!Nu39j-Odp4CxD)e@-V z9Y*GfKx*pf0Axd+cS#72pFNgw!9Vqw0_drAiX-ww3y5;9FD2>_atz1vBu~mJ+IBTEbQ6WJ z?T-O?j<1>TBFBA0ke*3m_HO?Ot|zROMF8Y?-+Il<3g6rK+aasA29i+e!chby4ckv@ z`6rM473UzKlO)mmgv9aQCy(gW;I%~mCV%z)U_%`<)7GwBmYIM5+YXoyc5Jdj=5oj4 z8bHR@eXbhHJ7<{ubONV45ZwFCe+AI@HP;4Z9u3<)o`%1`o0mmXZrmQU?076yjP)w9 zu%7au2AJgXkTS__e}!@?Ly5bnNMI;01JoY(?wo*m;Ti%b6C|$!Bo7LNrGs0?vt0Pz zH@zSD>jKCQ!};Ww9Q#2|It}HgfjjGPxL1%*u^f}mV`qY6Icp?s-oIeSeOPHFwc<{T zYpTQfxk26sGB!?b3rsY`gTPFs~7X005{s*1T1Np;diE=B-SRxx<ac0aPPR@40*;PJjy6FZ6HLFICb`Y?eyl4yUeC3#wA50LgZudm8FlD_wBm&PWhpjB_?fVaq?h|adM!nvD{96BZm>T_Lf9i3B_x6T&O?R1i4 z(e7Gfz2giGJziyuc@&}|2?OhT`i|J2qieNw6op+13qzX?cgckCMz7>Kmsr81WeXvi{9K>?<+PjkdqyH$6y&VEJ+nuKZ z4cd|goCpR)29ew4u{bv`)1za7TtWfh%gvdJ+(A&q0{Wh-=u@aQfMt2QMrDVjkB3p3x9|8-IeY#wmv?-T?d*tUh9|THAr>fxOlr)e|g;M{DR!QJsOQOgYIl# z>=RH5W+&#c$X^eVmWz;9Sq_of0<`5nz_!aj2;a{H?mL6oAZit`VAl>Lc&DDhUlREz zGVEIw0e^G8)CxD%z{)S=I(4^@uY3Fq+RDH4D{{9dtlJ4lAtYb?krc1N5#eh|7x64| z2ogBlAGS+DM%%5BEC2}(H3shtf)V4A6FrazJpDAltK*tP`XNL0*=M@s2Fxmn18R3wj#lXByo%}BWl%13`z~nFp zC>#I)gU4_Cju4j`w+^C05&G$Y;aVG*5tR35lvMGOEywzU-2^`4Mf0sjTOMBve6F>_$p?8c>VP%s%d>7D^DXhYu%3_K++V;DtYX) zE`wMS_t?jG2}GfzVZBbG57|BKGxpI3qcqLU|Lgyh&U|=H%z(bK0;=mIq-bt*Azp%+ z#Y734iMKbh?A?ixf898z)A6#LXdeYjddBVM?2qhuWR7xEqf?3P@ERLR-^iNIlKt!( zc+OJrb>H14Jm-2Iq>9kdfkfitCW1K?=&PI71Or?yc5HO6JHJ zRCxbzBuYYp<$l`+sTE>49brEb9bE*$4ErY-IN*8DO5vUV99$|60Z5MHaEor1i0%O( zzCo89B%>1ghk>EY(kD*_KU=Qw+kkI!V7P{wkv7xqWRaIgNA)Kj5N1i;xFyBQSHf*r zv+?mT7!AfUEQ^&t`*RMCz?8q^j09m4O!vxlYvsub~Xv?izceeC^X75-6A_SHR>MFp@TsW9lfMsX7-k($+`0WL5 z;u%YhH43jMypfz*ufX75od~bnW{u^iU5~=-X2K4 zxB9f(GJBBIGsqy>nhFrF<<{JEJ*`)~_6>}11J)AAveF)8+A2WoX_71J3!PhPlEaCy zALi+jw;qJ`IH?tHz|`to1m2Dg4+m^=$07&lz853#Zvk*hYdDbH2P=8M1WUKl2xpr| zooNE(bP{d`?lT{IE(K2sj;g~i4c$xoJoa)=`+W@c-h088*ZU#>z6hY)GP3oYWRcTR zqs$0`)nwXl2bMDLVxr^YisZ)Cm-pTinZI^RugECNV7vprhqlAAre0w3B2sZXAll0j4Fo z;Pm%FgA^nwYNa;GgMqG=go85$Kq6{0uYFr){_Ve!l|TIpnYpq3U@5)6B$>4k9Oa05 z>dISk=gn85S?}BLQWcU2({nO(dr3+Qd08tMYU&neser*;HXykh6>p*E*($3X>ctY+ zR%%7OUpWBd9Dr8R?NZE!b*qjbj81X|U^Cg|jr@cIlzD6npjHLJBMU_y=^2A_Rw<4O zBfgWx^UL602LXx(w%0q9sO=r#|6K=Yu>@dMEsO7NXJ0qill~TfB^3_A&}WIY6k`;MQGlg~UGNUdmM3wCo`4o=E>1pTT5e6LJ{8EtT;0?@UY>aYXJc}sC0 zgyeB~%-mVLJI2)w<^qDQcuocGtl;_3`Jk3Sd9=YfM5aBh1`x4?`fd9m-0r}0t4WR` zWSKG`Y8s$#2@)rjUBNL zXmpnh;amjDY2akLnGUn{-X@uaGG<9G-n<6Qce1|cq=$@S8g7wfp@mP` z70$Lu41{%}gxNatSKo7o>K_7GHRF({kFJ3kK`OUW!OOA~KdQ-vK}i$pAjP}dD8qqn z-m@fTsP&(E6p27eqAvMRa47Wu%aVcgNdpoo<=5W~c2uG9z4ln2B*F{<`$9giiee99mbNKPMCa z{J)p=U;ii4^{@Yh4E!(ulT1Dl0*M6g*UZe7;KFzv7_VlR%uFPdPqHtpGYr}qJN29_ zoP|fiw>sK_TpQ}4!2`J32d9V{pLHP?zCZ5$MY~s>d`SvF_%+FY=f9KuxBo8I|7XWq zFJJ$Us304^kDA|yy);b^@D4D^<7kKBz$)Di%Zk_NvB@zTAy$*28w zq2Q}i;rA)VW6$9I+sX)J_nQ|JdDKw$8;k(;t=mnT8*tv{UV6D{TdnRnyL6+&PtZ@y zz5I$WNUj6h?|9$WsMv-CW>U3y=H-5zzA{;&iNOJHA0 zfP;R53&B(qKUWLn?e;_&07nV}@>sHqqd_|W5^^YvwZ-%+FM-*L2_G$bV3l6 zw=D7pNoqy!k8SS&bl`2E#myvFoE|yuNsYiK9XUHOvdNF&S!&&6J2?j6YqKDpEI-y$ zH9CCA(5CS@kh!nl4No$q;27@B&$l^fr{+cGyE4hSjXK?XVUiPgE*a31&J>jzqQith zw3?G!!MpciS(HUihtDA1GwW<1;7>q2YHqt>e6sB9anbbA-zSeB15kcZ=#bfoakj}> z$hFWxMx%2?%ay>+j0M_rEL2wU{j5zAm#@-`dfss&EXHA%hSY+8YOy2OG9uslvHaUR($g76EFl zC%ks2nnW3=27-Om)vhi>mmI-6qf#reEM~K0A$MRLL3QwOYvKXI1c4n$j*(`&|JkTu zTagblVhq|@fc&7fGJy5GO21%74KgO>xez4Lv@>_d<=Pu>2h8{!0Qd5dJ{dYPfNNfK zz`mw+k}Kx(S4S|=+RQbnFJF@y^vLTW;noGo%SpfF3f(R!p_MkNLDr;fB`5R*SjGTM zCXDP!1)yCa6}Ib1fHfTm*l2x24BlyjbTT^zwJhdC;FLA~ZG0#26dXt%Q- z_?INelPk#5;^xj$bD1RT1&m8>XI0t>NL~kXwQ}ccK~}L$KfWbrrENY0ZPtn%GO+~E z#<%|AjC}mB4#}yFI=4_jM?7UZ} zif~j-)`puHq;-;I#>T;-g*+d?aS)*t~fS1 zBv;5LPag5EhuayX^Ul}5D)0T)Z^+P0_`VdgDOvgHkIT>hvwtn0{i!d?$Id=4&zyQn zE}VHvjy?Mk9_lQHeoA^`Mt=rNf_p5pUX|^@wZ`#&Ccs3eQCmL!_^(L!r~k1$_BVee zJPt2k3FNqvX%sl)WC%L+YGjh<#k7%pf15sAh>qgoAxuIF={>2A{e=|Y`Ax?%*E}Ai zC5kGrbgUr#jTtH3B?#=*jbJC96?|73s7MUht1~coKpk~^Y5b0N@O}Clm0Dro(eA(c1vrfSz7BzI z=!#dy5es4YeJQ>3TX0g2!%^jz6Bco(19W`)?_&jhsy?r-xywgE@Yuizvgc(z+nm8LkW;M7}WR>Xu#GAQ`3J2>&a&Whu+9Vm2?U=#?eDNW zsrjs6)%`9*0=G7JCt2?KH%gX?a)WJaTnV`77Pt?h_ zEMKzD3EC16RE93k;eo|r=a$><#iGGGH7`2X`oLagJ>AOs(b?+j>v2IJd2M?&f~gh+ zclT}NYT5~GqrA2=GTTo>_JN@LX2y1a&O)7`Q2?k&@YR+xfz!ke&bQe)4r)6c0G3g_ zndA!PQJSSz;6OlIvJ+#S905C+&tHvTJlV~yjI3s>9pBdG>5(%n^Wrz5%{W^e10F0( z1=ay&IxxKzm`lf#4c@tq=fpN>w?^=o&L%a!e+uODSz8Ac-&I=|^cYv9dwH9Bq7C}G5assy~%AxS41%OQFV&KY1G7O3Z~%2o4jwLAeSiZ6clGxGV*d|Dnm zd!|VT-=^nt>K)x#=Lw3hA(5iWC1Ni@X(+oQWB{ zGijJ~B4jd!_YntHjG63#&42srvh+Luue8m2buSHPc4npZ`=AF=I-tG`*_Oz7uMJc0 z+ty6xdcZy{+aZb5J?;(eV#^EvKYMQ;Wl46I2mao^scoR+UWtx+^1I#ElyfH*Vax_xpYK`-yj48=k*h9GAwXUi-mIo=}#@3ew;x~wfv7);kNn4gBT zsstuq5&01_48Z{s1fWNW5K6Al6i9BcC9Sid9snfswz=o!_KYte8v)k=>_MPW@A@(Z z)0J%t0sqwGn9d4dQ&cvZS`@Hv1{o!T)7t>)6r+i4gv?6Y!s`e^_8wS_0G-m-buzO8 z?jwspdT9VD>p5t)y7(^SBhupELJ!{ua64ejLNK442Mz$Q>(KI3f+mXkZvd;+fVet; z6>!ekr96g33F0o=KY;I$+aL(1EeXc$q8Z4{-oZA=>=SY`k8*pl{0=CL#Xe&5f^rUI zt>%XRqXSMd6*awA7ZDWQ#`Dv7-b(-gn?Pj01B(UgU5(&gPhN=0Q{U4jjZKE56LKr@ zHN)Yv`IP&zr2yxeDYwFt48*;r){8O z+r9qqZh*)B*#0wQQ@Ic$I(fY9Ua@82ktZZ zt#4UraOrwBuukYjU5J0>Ym(P8Fvur2K7rk?6pK!aJJ73e3IQa?Ot1<={>n*s8kpPZwN-=UWX2NA(jYT~=G_ldeW%y6)AZF^ z<76@tY+&Q7jn{9o&Gq)ZUJ|4+S^jF8JUX8Wgtn*r5Zs_V&TgOA{a!P{Efb)p1c-Hg zZh29bLIXGVd%o}X;|XS&G9Ga-O7Z|oXk_4VQBN21Y=5+?}9PIDimc70CsLCxYHv*{|xF$$xc6KsSKT41?0fk|#MeD8MsmNt2d(?${4wlv~50}lkC8RJ2l$t34CCAnfaTgXUfkXZfQ zpmSv_AeZ1ZfpRV3!FPxbjh(WSDe=Ox(njZxo1cSg(;?A{l>TC z8!vxbre=tTa~TOzt)m% zQGai@Jo~;2^2zW2K6&vYFUXk_$AtmGbz4Dj{LsLFYPWPALG%H%^MTtkD0B0Abh^D- zdcbD%Bwe6|;Ae^b5DaQP0L*J?lJneN2JbXLf(VX_MuJjCKs#j^-=;BbK8_$%^MCpW z0LXt&LaXlV0)GnHw_vK4?DpL{sPgZ^o$E4v_3g@y_(f@$mD}SoKX!A|zWd!UJ$6IJ zUVlTvYgSHSu-GH<^j)*%V1HvRhrWjUZC5Pv9fIURJnsnlh1v?`sN~Ac>SsRXleG=# zlGBkvLm87vURYTKfDG9KXy0lhgE$Z&NbbIeB?&_770yx&sv`g*_-uibLvF?Tb``)F z%i>?P)wi`u>)>F;<|5knyx~aQN|!tc)^-$a|zp_doZW^6CHmm*n^U`7g;EU;DDoPJu)OkzB+3NTj@maTY;-^z)Lt{_gOjuZFRLRiTQTJg$pWM=v~~1FU?ZA$k0#{mk}UI0@8bLI^gyp&tXZd?T+ z%+0y)WEl>#LcQw3TXr2D)gjR86QWIF&d3#9XSF;lkjD462U>C-t& z548Kev7=zxB7kcTu7_-67#DS@2ikIY88^0`z4}{{pZHSMo`?7o*G`!!njhZozVM_K zuyfMBxvEPmlTCiJB+bQH1*&tm2hPa0dAHp!p|Dq?s29xJCIyypj$|BtMSO=GLp8{S z5i}S3QfPll^{(R%2jYEk>!)+fKCRS}TB&>T2{^GJxnhFk1;{47`3>c$W}9t-Y}Nifd@+b9{LndE7(OH8oa3LfcpQdtfxZ-RnF zXxGuw02??Utvs$$mPIAJ&p(yartrNmJF#%};Ya?S`-PE#KMa8q4V9zlJ5cKclhs#^Jfvn z6@lC4=b>@mC=AwX8p~2UG`OAe9j5ciOuHBB0o=a{s-z%_oA*OP3c!ljX6Ds851Hh? zM)%uFY`hGFka8_P+n2IFIxMUI_KrkuOd&96RW3+Bf8!$bDUZs)lh0S}w0{Epl`RIv z*_lyEk6afS+M3Z>JJapA2ByX(eCvjj$e+f*Zxzf{07=o{)-!!lNwBfvS!rg3KieoB zQO=gYdA4l$wO8c!jZ5LK76Pla9RbU=wfFT%Gv0x6`Z}Zr z+(EP_1jl99b6{`+41IIVN{+)y2-J#bV8|Cs8-dE=no_5Zva6*PssVbI5iaa2tbkX0${DH^3CTydrQX z2`zO=aOSt2^jS~6tou;+P)*x=!t%_E?b6z1 zeLDiE6OGN7Az*{ldDLA7?~G_Rv6IXGtgv64MVY1z85^F7to&rk15Hu}%Q#(~vQ+ zlH@S!WaIi3yq_o31}^}pkSY0uawwqRmqQsw>Adzs-U9CrXAFgG)@Lb){>L)FM>ufH zAPi_Lqc2@T>;Y{`a)rS=+pkJSZ#0%!fSbv_fV3B{Ia~Q&`>R;w9R92(3NpEnlgIkV zCO<9SrB|T=Pr%T9Hoy)mttg&QvAb ze;!VGZ3M~vO>jOD$Zr@xo7Sp#!5TqqNi;rK4}D!-w@@RWn`I? z5uHVy3E~-0X2c5wzHs>OVq71y6R{nGG{?vvqs@L*AF5`4zKeyq_XMhGQqZ41+*2M> zjNy|f@Es}we7@u6%VWFUI_d7HT7!dz?-{Eo6n$*NxEF1V-1jslvK~4f+URwisH+hr2h~~^WGy$-}40J${ z+>{V;5lw0bPP2o8?w28Q%MK=>#uKhXR?7)lLV3Ue$6r`*l6F@+J5}@JK&% zOKt@-2iQBXjlkuDa4P8YJ<3ZyFS2O$3;}{@1gr?G5%4xOucI}BZ?e2iV3v|AtP>T- zkx`9jvX^G*l3N2!oo^HM$nO(e(|O(Nl3=N3NLL@DWM&NA=<5AD{jKV@5txi%dEQ_x zvR<2Q=T7tA&-B!bCU~liAUW#^>lsbykps-uMm=&3pzCe-vTOnfYJ%(4;JeO)b=-6? zSWNPO-i>o(`ZDUaHnw#paJn6noX0fOR|fB#TUbOOQJX5OGrh}_&zEvX+E}+uW<3FO ztxHY@K7;0(AdfzdP69dwOj}%DmRDYXQ@-}n%W~t^9R=W<`**U==SZf!HfnzBXU#gx zf7+^9;{^!rTW~>( z>I5M;MkCO1u4UmwQX&SxR4dI+w=J*C%JPldQn!MEp*Covu8iH1xf}25Ej5>EOby>y zL%>f*gq2bkMF8hkUh$1}@s5s3?B%y)_Kmk{=CKY*rsc_5Aw!gw0zee!Y#pLnw&f7K zGZh9nZCeSiQXXv7Qes?h!n0`wJF{2(cqR-O=~$V-GMuG&P6Ww4c!pS2+T>xZ6cBJS zN6iPrQutP`)%cDD(0+!S-2ptA%Kk;f5~Fp3~}<@6%h zZWfD{;MA$n}L~ z8NV~44^hifdS3GL*Pw-N^}!hn65(Z~VCY@@!D_6B)^NyRH|7zrF2kw+h7}mZ&JNtk zGSq=SiLF!XoxpAeY=-9*+N{3wC><{opTJAbB|lxesUo`(kinJ%O15 z_Dq%x$T1s@klBh=5-FSly@!<&-n~4&AExC|?=~>A;$Ex~&V?5qZ;)rscE~h>%=!!l zDT8;$uiubn?R-JKp(ek-6~PJ{`fFzoklaU*+>|@P_teoPN2TFRiHI#{K{UREdYr_J z1`qbJt@nYU%|Mh{>S@_l!Z@z5wq(uRFlWu+UcDXNIvHrAmM*y;P6>y#>h6^=SigNw zpls|Z5FR+PyiF&?*Z!oDinBsX^7{3pynGqt0YAoQkX%vs-dT}M3aGEy!CvY8`Jb2i zMtAl@Wh(0z+oR5NPdK}LD2ImmN??y=J-YhXKvb~$wF=kb32tODnCrqnCX-1&;k zefk%~`)j`?%NxdVI?)4Kve_xxec=F@eqITd^>_;R4p!*CG0cRG~kDCaz zO1&?Cwv<{qj(Wm&2{Tedr%$@)#%DG=i9fit!n!*aw3y_0pA9Ctp1q(B1{Oeuw=x3R z<5O;)0LcUQN}z&s+iT2h;K!qcVA_u1oy?7mJ3rA`lhAR?X~B&Q;FmssubNCb>KEP(eHI!>Mt<*U00^B-Owz{OJntFj=4kmh?UB#reM5^B|~YGM))~ zDeZd$yy877CVADfRl3|pCzWhj>Jyv#$m%#lCS)(lOEbxJUX;+GtcNKbqxbFSl0a^) zbNazzV4y`^5BK646d-Sa?hNW^IwROtnmW~can8NUgY|@LaxJ~V_G=ML2|6ta;*#;5 zvt=Myo&+^{zdJS70iH`+JP}unbq&UIyPDPxXO2D}C29yjn`OuMp&qlEGziU(sM&^) zC2j)N`h0pxCOPGHwDDYIQ9Dy3^^R?pl!Vc2ceYoQpfT$j+hX&Yc@8t^uFjh8BtiaK zMrQw;@4hEr1z3Lh+I3Y{wbE_3jZE^Sp~=cP2U>%kjs}x)Z?VQ*z}^DQ-QUU)Py-;` z?)>+=k7a1eMp5F14hpl*-0wNe<2_*FdI3TzL)RvqZHF_Y8+8z%qYZ$Y>CJT?*kOK~ zO-@F#7vKJ@Idh9q-1yJ7VgrDLnEN}F%t1TWEpMGpFMYtr{RTLyw{uCkmhv)YuRsfX zuPl4>nq=O6T@7?A@?2RGnKK5j6+5jTDax&^SIVt0upcq#)j>PCKLXZi46l-vp>pum(A zTWUe+ksF%JS;(^_l^!`5=Y}+lK>ou6@Q6kxX2SSBQwz7S28OUv53&FXQ;}f+06+jq zL_t((HUO^Cct_N^jpspNas8f>0*h{CO4Id_2pYhJg(=Ay(iAzA$2!1*ESTcj0ya=g zz*rk~vNM<|pp7i@tT9`Qpqa+hIXs}wJ2-%n25a9Spz6bGF6#gQHK1;+!6;R_BNGB<6z-ir3|1LvV*+%^POr<=440!V2D zGeC6HtMst}a{DlQxo*HsXs$t?#NXTn`32Sn%S>JJJRGWVBfbPd;1Zr`3eFm;Km|~5 z{itV@zDOgesbe5Flp#?DWK_Ip$F%tnKOoY3ME}It6-;(GVDS0@%(pMx-pgQ%>z#<7 z5;W=~w8*Q2buxJhL18bCrezt=nz024qnC9h5!(1(sK3=1YH| zQ$Lg?+a8~%C>_38@y#HP7J^vQ_*@dW{&nBSWU=rON$P5Z`sLfm%aQp@a9|bWc-(S; z@|ZH(*><)_gnfrF~JXwRs{eRx3T3-l8thl8Zyu#&a%VfKPHOKt_i-QjiIVgdH z64;v(FoSmzh%iC!ot}YmE2gd>2kE*&VqOI0O?LE7^WX3Eynj)8pa1&TNI$-!`4N%*zD-7+D2xe=~S?olDyK+t5c;`K(7jNv^c9jfu zIyf|#ubox8+~_FcwQI9LahDxW7CG-#8=nQ+iwAA|&YI7Yvj+Afc<1`!tgOxe{I!8> z>WP;CcEH_%a@w9r&NA^)mmC_jGy<<>44%_lpn|D3+|y70n6&@MPXl4WXcy)Iz#AW* z!8MGl-{znB+t+0B;+wcOYikKUDyohYHCBc!7Kbm%%*9vD!*;!%$|82f?4rb$HrGw+ zl1JkPN?YSBWRfQVl6y7kg*4Q2j;~oaKwa+G!aNxFkaxOxMQ*(O%BEa$0ID*}OV0KT zK&hk+2J4MQX@aCjhyDqcHOd6%)d4d-Ui49;U6`QW0MI_}fN%uJvjBu?WIabX6953B zhw=c}3{?qCdk3o;A`GT^0fzhA-1F&uO|8w+Gw1>slv^pmUqR4P0p&;-gsf2=0s?AN zdgT3sNJEg^9l(1zRon*2bNHsDjCT%%X@KOY2JYk#u-zfQ;e&({fKke$IKeyXta`=s z715UG0TOE~WfmnjdR=mpZeJ?}N8fTiBoS<8qmKi5uz3S07dnZwSs-~)|Fks94M`9l zX9oJjWr?p8QcQAR)IFb*fxW`lj?N2htC-<4Y6Azn69P((AsrqY4oFLJ{xwLf{E?Dr z(EB|Ig2|@SOsB>8q)s9kNGYxOEoN{8?`~U<>o{d1g5(X&E~rdqcVf-$ zPdgdy(@Gv7uE`{)MtVy}7|vYO322h%N&SZfyalQEvYm&V5$hY6gF$)^nX^Cm#FJ+o{c9c)fQ&(tIZFobpT8A= zF=Ud+*sf(tZlwWCayS}2QLA$J(VqAaqhtp08;TfPBo+~Ah${Q1DkXOpGv<{{!0+ssPPO1BH^qSicQ{kt7Li$cWR&o5MA7Ee? z7v7Z2#b1-c^?xOeT~>R`ba})>Sh)1AIwX&KTNt^GnKkxH1-TS&Bi6xgRZhi5*BM}* zBh=9)J^$cWWuoy}YcmevWQ7*YxM)EAbPUB-#7vpIp*vnexn|;))5}SxSKw*zwVW6K z`Cpdh^nk2?1!=EY_1lj=NIOUZrTB4&d-a&0-dmlQge{H|SH^|rEj%UgMx{J%Mz9nB+Jq%>-jv0c0iwVCrbjjCWyZV=Hd7R zg2Zv!A=(jb1X@#ICeIe!!KdDLFH4pY^Ux9_3row=FfpUp={%n2GSA0)w1K|L>}=)v zve5N49fJDYtaG^x+Bvkj+gTaoV9o^6TI-9rj`0Z$fbbsO&FC~>pipfrH@5xeH7fSD z`}on{N~rxNFwf)XpkYk5zP8S`HU&hJ$}naOix)7B95^}ZT1i$n;<7$l`Ap1nmZ7&>^11D6PU+?EY~ zFv#>8fjgAx=XWHB>%=6Qnod#na1DIE8!y- zpn=X`1cd^Pk$XV2HGOl zw>gzWTYm#HGfLMaWY1j408GDLJ|$o8c)vVXctcvcdz|OLn+s;}$8ywV5VQMKL?1E* zws)%P`+T?jefgmH+8S{)_>KZlQ$aTYBxz&)kj9WK=VBt;oj>ZjA#3ww_A{ zeU~}Z?d;>v$SP*3z?Wzlsveo-$pV?=0Lk&52Jnn?IN7?7_hmVe7?`iHh*<}BuBrgJ zJh+UeqJWEBUrVZgR)&19HY8lv4qftu=r2P7FHWwdEFCSjYy=V)k^e^UoD{3>C53kC zjbW>iO+$WYcyv)bT{(Pn`o$l7M*^LYVkr3}^pP&{4V~Qb`FQ}22NLfgaOro%ml{)- zrzBed-e+%HS6fEkCUxVM$|H+!*(_#NXsFg=k>ft$h)nzYWVPx#tsIaS%IoH|-S}c7 zn|D8O0JNU6vJ5#ZlrbhiUeF94QbVT9>|jSyawk5LGr=k7-p@w*GBF4$dgY?;Y+A92r4~Z{D?g1 z|16|>Pyi^0FyePSFJE!5lQSeRLtuWz0@H~f%+D_)z&HBUkRnT>T`xr4GUI`KD{^3$ zZzSMk&UqvZ+5GoT*#rk4 zWxw!V=JD>$j6&Z#1s3-zw%NvDu*(b>tIw%-)LQ0&B_XbbW5_o#J&Aq`#u4bpmL?|= zw9{%^$?W%F7T7$Tbj8QughHTlWgPneB+tCqFJzHFJK}x|=%MxW^hjCH&H@;o0qCAd z_c*8Id6qE?Vm6F(g4jX16?HfG;{ov_bNm$2{;8E77|6!OOCTm7?G{EtW@jfp09)7r zNWMkE>}E5kJe$*vcaz7#F@G~lAPLXq41#Qtjmkh^c zV5+^N6&`;L0X&E++I$XzF;f^_rODEOna`Zp*bIgzS*-_Vl>p^wCJ_}o4rU9DvlaYc zFA$8?8pmXex3n}#E7)w+HiFK82ZIIzx0HO)b(86scMHmTKk8=!jQ8c$6$O&>J?_L* zQ4@r%zRvynJq@5`87yJVp67YkrnI#+V`+5Y_TU289xg#5zbB-vAKU;{nM z7AN=%2Rhq$8kdw6(d`KJ_ImIfdwX5B%ueqof{rW8XdO~3G|s$Xeb#lC?I%0f+Blo2 zRvStN^a5C>YgNo70;jE8ybC%{F4@k+F;HBkZZJBtY>#pn#MPlqzZ+XxwOMZ89a0Sd zE6~oH?G|T5%>eAt?=77peBaD8bTW@m!%=8EyZoppeLdZB^zb27E>mFCjDnrA#;M;h z#_J%fybPUZU0!O(7Jqndwinv?o_BY53MHJhZKRVjhe1gm0EjYl#G|-QN^`MnE8m+2 zbkZ#lUmyPp6S%P;b<%{kXAtl1_%NiPAWu+)!zK>4;u^sU+gYI4T#vmj}+;^dXXPU$NMkA9v)_~a{ZDD9u7bOH|hZVdNvj>Qd*+!;wTX%=F zo;j&ERV?$d23bN|R>w9BM#&Xu9Z9{P0gB=1yLCr&%b7sm*W1=4$2$>-bQXb!3%1h@ zjLKkRuSB|zs65gg9kNEPb2}htB7;Dsi$Q6_yRD&+BEXqKsZ|;qebRs+n{(rJxYriU zY*1OKKOwX+5%f=0y~-yfGrTP8lb9`FJIL1%csgTX~4J;zab>8oe0$2 z$U1HH*`U+{U^(d2QHY&D$-?!f5G1F>iZ+7n5&S7Ym%P{pX|JD3f#v6wPB3dvp7?us z0M>;rKz}yb(+^DBLVyBr@;9M<;x(_M9rOA4j#y>OctB0TWvO5LhH@t8b9w3jE)M;; z1Y4ibo76In`nn8uvHs$30J)!~mVF7*Tab$A9jbTk$2@RjF#(1?Bur38$t2Hbd^OXV z$L!U59uAUR+HLc~(5@di5|h*zFO2-zRF-br%^7{ zA}5>N2cW!wy1eEk@M^V51kim6$iFP(<7`_Q_1P1{yHI_vV51iin6Vrig+vm*R7uR{ zfOJ#z3>*Nms*d)6rS5NkLYn$Fx0NM;uTzkRV#8X3*C7W7SqlPMm%~p;e(EnId=$WQ ze^H`?N%1$}27=7jumXr?b1zG2{#!Wrl(7T%3x{8z!*~w?c6Pu?)Do+g6m+tu7j2}K zN1?oZ7^Zr0tklgrQ zB3%INgz7S6|7&GwR<8cxzm+H69aqP0`h$0`G(0bjyGyC$2Ir1M zlBhGY(z`z4`x;dtv0N#m&1P6g2B9>UB9ZLEXdZUpU| z>`|M%=odC%9MakfX-Mdu=Md$x#0Jwe{LJHfkxfDVsk2XVN5Cq-i1LVI2m8t!IVpVS z+o*G|VP-#O=eHo3CfYx5>xGJPDx(cez4D5r5$~q$pZqW88Ma+7qmRChSqIh~v0kAs z^wg8mXx5QDWI%7cB>5gVb#B_%k8iVgdfCa_gINlVCyz-~9|vH)lskc-t*@!NSFkZ? z)~QCcd!^MWNF&*K;ae2wKpPDV<6f)`kUY`Sidw4Q3B(}4x1^cpM!!3G`I37-0PiD5 zj!1KRtCZjWoGg6lOHu#t)5jn<)`it$K}06N+gW!N6*>y_41$7KQi{Q{Zu+=YIA12p@CxRzRm!?rcN z!UwijB|w&tsdKwm`5xr+Uw;V@jMD+=2uN@$^lr6%IS@DHO3c!uZfZCje5C6sodYJD zk?a;TKpU^0<@2E4HKjQY=4CD8K1&t=J*S6glGZ``zwG2rj2b29mRWn-+WG#6w#ovJ|AVJLI8F5`g}V7y;^pnlQ=Us=YZs7 zA)9rk3E&j<=I`e|cr3wM2FoNkOveqG*#zcThgnZGXw3Vn!*gP}R-ZEfMk(tyokavu zwW;(N99O5;`^{xXJJ=(O%MQ>^hBDu;yE$eF9Gm*$tlM?aNH@=^>KE#QHp_8|0Wa=f zTO4|zW`I&2QwOI2)9@byW48l*zj*aJuG3<ch1qn?~Nbm*s{QxNqdc7|A zS}HaGr0$m!=bL490q3ywnzIbvNhYWbl9Q2E0!U5@sW;Lpp{7CLRnu^QxPTcTMa3Wu zVB`tx5#&2OUw~j4$T1MCpeA{#TnAR9>u`wU*{2$!(!Et3;4BaPle7gVsNkJ&G)n`8FmL?gzvm8bn zVxb`(6p-JOM%i0s+}hP8#nWJvzxS3pIKp6qV860o=%w`*{ zd=-u#S|0bAMKx+@cFRNU*oQwJ2+*ezZoegnXVE=WC(E-rx&9`eBP3U($p$*X$}AhN44?cad_}O2=w>vD&KM$&Y{*p zNUmH0NWK~U@dViBgk)l=2Tba`%&+F<%2-xr5JdQgKiDK8d#XaYSSMxZk_Xy`HCG9N zm-7-_j;omvB{&@&!*!EMo<_M)i=5^Ka8-iF`--Q_Nv6CEkdrDb&e$#_9EWc}2{RuG zaGKqj$*LfggV3=rd+Gp=_edU}r$9RZLbN@;=DalZSpYdR^&rMTsQ1ID=aqGam{}6D zX@)1Qmzi=aYnTNS#F()Rr&qb-GyutK>blCwNh!SkKcnFBks%O@e>6%pmG!h5wZ-^@ zwvF;e)Jrccv!?_{Bi@6#%kN5C-=KIg3tXF{M-GrY*mgz%p?Y7%63)~4RkuzuE^`>| zK|>Ry+u*pW$BXWOOUO`Xtm0G3#Nzn3vnPF3ZSo{+67ZqEoiEQNAjBColC+ zUYGLdXC-vz7iuTmhpm_&W;qNcF1`tjqwIF5jQ(oz$;Vauq$N`dU~$hj9+O77s;-3P z6L5=5Icj`8ewn}6Yh5WGiVsd+E1pdWeJ-4>N|T(YE291}Mn=(tK1BnF3MUpM^``aS zl)+k052N332K52&N_puG@vL4%8@VEZLqDp{Lp}KH#DuK>-XDOqjmnByML}wV--8t~ zqmlaD+6a<2JoP^JIb!BYZv0DN&o4=^7tcvIL8zB`j3=~FQ#^Y3kb4~7rBcf=6_C~k zsh`5Ubw6ZagI*6-^_V^sbsjQkJ?!&VzC&(;%C$239>MlbcO&yeTfpNP*Mm%QZ64$u zJ$l&Mhi_Hmhd(4MpZQN}w-?8q9b1r0NIjy>N7!&`_UvPbC|((YFPt1_~u}~R+@*sjqE@I6*|x)wO>u9t_tjlM)yG( zbz8S8GlEB^Ue<%Urt0%((6)Lzf3u@N&-#H?mH>6htxlA+eGSY6N~vh8rAMy!5y0EZ zDG>n!JphmRepnL#IqH%(LxY+C@xwCelAC46>tK1AI`4aRZCmB5f_B z&auXNw><-+)dZ~hEa_OeUnBTy=0nCZ{ZDBARYjup-_7dQw{D^6sA;)(2 zUW^P6-mf;!I51^Y^!Dl{b-EdtRT~4IRs+_obLI>yg6bAfu8+b1neq$(<+at}+9SAM z(>gH=V`!icPNs;nnGOcZv216Y5wL;4tJ>Ix5G1euJ`f-`ne$fgPTU`)4H!^V#|U9y zQmtjK^L$XtYE5#o<4uPAUR_53)7|+`TdLduxec^m=%RQ4%6K@J{4VB$7A>{OX$$}v zhAwak485Bjhvco`$H9g_UXep9K;O<2N2CcF9Kl07Kn2yf_Ci1(GfIV-YxE8Zb3jTMhuo#<4C)2mp1qEd;P5QU!GD zebggQ07$`|qb;%T3p>kTO2%hLB|ZXK01Knihe#8G*JZB^%Z*F#sci{3tl}7a>Oh-m zlx5veOpc#xm8Nzu^$d`*6vA_gx(qYQoD??z&O=>YZUfXMW7IOG zN8UF8CLKU7Lyz1O0tiyyr;|xNX8v{byT3zlT-6)j{tO1}aez|+)J+DTZ0WSx!;A{v ziTk&8$@v&o0g8>miQ-vWhh&HQd^aKh4o(ir;E`jJF#;9(QRX4!>%r0wO@1=&X=Lwb zp{YRJfyTg{Ih#Go-$=emlRS)L^F;)l%o^>r(xXYBfQnL1%>(>@MPlA}q1nq^kO%5W z9!|_ruo&5|c4p4`Hl(loszg>_QR#}%$;pm*WbUO^x$(s-xE}ZWp%UOvAeE7NNDXSs z0mh1Z4ym6JJoh#@7it*;mzsb$gJ~w#-L}8hF?+Sm3f$?EA^fPD9lc@cIUL2e1UmF7 zx32jhQXk2UqMWSuD~+XO@aodCXf zEiQ96u0V(01<5(oETbN$Ah$x0ob{A43MC-itI~stYgK6a^*ji=SW*mc`>HUrxdaDu z9B~z1xtx&O0Lp39MQ6+dC2fM{ks&CeN8RrPwefxJ?}PU8VLKytk%BDMjY)q0t@9Ej zU)1--@<9Mk)5?5Lpuu0x#Swb}jAQ((1(1`uu8$>~+}B~hsK~5 zHU*MX13q>X-#WL=)BE+3*Eqjo)iJdVz+M>J_zw9{ppB3Y(*_=Kg=|PW|K~rIx&Pz;mNQiVc^R`73NOAOiMpVb9iB-M!7?*N0y}b%Mi=2gLLmSfGk1n~}3)C4NEDuWHpadRx2@oLXfSUvS z5SY^U18S1tN>+L`8<00;1?_Y{xDWhB@9~5LaFot48b}i$=iwaOYXJQov%9Zk1Usrh z5S<|zio0d}>pb>i$!j1hko8Ch#L)RFkR00&tg-|ezbe;EZmjt27;wC_Bf^fUdZ@0%;z_ok-d>{AAIY1DxmeYZ4#KVx9*=d=v z9czPp7Im`MA?s0X1j*^-x?dw$%euD;86X;iOY(a{keRM5_cqDoqI5dEyBdY|R1?%7C#NO-^IXr>;vMh~(Y2;9&44y>5La=$wfQltAWNlRN&KoLm z8Msps*qArm>IA?nlfYoq=Yj&sF*5*1L~L=bpQ{Bx>)@TmG_~ zW^q|jjMJ5NGk4gQ1q*QlI;|1ZyUu>JeUN&frE_(VC~20a=p4WxD{&gm_4(JoA=zbU z$uGEd*#k$7@9}3e3^pJrf|iBZ7f$b8sHf{BYvb5r{+? zv~+}|xuXFZ_?j1w{bdBxEd$81A8R5Xk=u}3X~(rcdwO%bzuT)WVZ4^ZcM24Di9nmb z>*TrJWkqm1)+0Uuvs$`CXGEzLX^u!CJSZ{T^E%2wTMmHYJGWA@gjpni=c6rh49*Qt zxI?wgfmVQPkUOE}Am_Gn`68SnwiYHgqVO0rF|&>t(5SO?j%84PD80g2GAc$def9hV zppEgaTaJ4kiVwP!TRUYD@+j}!P7zm=?|Gt;9|rLPB==%09Oy*9U=obpBks7B@3X%L zpj!<~;T+e;(OB7*$1`W(P6Do8faKohQzD~!@76okV;e@lu8WeseZP9{oo? zFKuP?wbMTLefOac6n*outiOFl1^@^jwc8S%p34#d+(0A0o=F*!aNR%mbK=7+h3grr0`mhyKZ1LO^8kT}i_4Vm!9N79r6XZG^MH|`pZv0fPXGPw zl3Dw3pFZ>rk1pME(w5M`We=l29EeJ$;PwSeWxphf0Fh52{>un1*f2UG0=<4&%J({3 zS=IQwQ4J-7wy%hGvX&8)SHU~cJ~ttFZ_m*e#%)Jg4gK< zfallnT^o|+-}{VYA@@ZiNS=QCZE5<{57s(W#Wn`)Oug#@>K=?k>z;hVNmB-kJU{W5 z$dnHgNDlf)!9L8Uop8%UCB_RA92`J}vD$Sf0~Crcc3AvP?>L(nRFnGMh51!yn*!K} zY+&Bqy7e0BqQ4c6FKsB`V#0dYn160;T<^hU^V$ZT-@f_=ssHc`vhlUAJKLNUvdEu( zO2TJOs|Ob}f^}0SCJF~MLeV%zNku-~!i?1?P&){&_n{7;EXX8x4DXV1NTAK~vW?xL zRh8@xk3L9#Pyz=fuxBMe@XHSXfQ&r_5Z_164#-@U2PF}_XpkOtv)JObuanH~`(fJe z^xg?}__ioB<#a#?_farv_)@32+6b&}CBIXh_HWW}YG<2FRRU`aw5W|h*H&ORL9|-H zEswPr==)bEjvYQEhtV;*SI|=5&t5JG>T<@4HuW0-=6)Di`?UaFp5M1mjQzd74+7P- zz|$$CwOm zy>F{!E94Afc)zr*0ASgUO-?xyUZZyKkwwmPR(}WCmhn2QNzNGHfO&2@ws=13v1b5C zYh&H41(Ngq@9XKR&=Q{~n_LG{(izxGO06t|1*f&lxtR`uTF;f`$8&0DF54FM4-6nP zP+l16OrW~zl>7RVWy>oCd2_!PurPdCW-%B`e>&>{}6yWma$ z@>*eAm!A1106gTHS|p0|tYr?#Qpz1X`he?XrdI*vS)e)7@5KDm6B4*H>>N^P#6S)J z<=};!0^sCO3v<5C~z7edD#nAcAX8g6C_97 z)>Z<=vfx4c0>Cr^b}7Mef^(Wl&Sy?t@;v%eWNXu6jIDuh+Ey0g;JOWTQrj#hIR?l} zGhjMj)0>5Q-66$+?5eE`&f`xNMiM)A9rv9Jbkm+j{xRwPVxO&4a z{}jq>Axvim`ap0lbo78h%ZsjIlsGZ#8f9hT1j+Z=SiV|Lh^OM0BnV(JKbHX@>4Q8> zOwtJ#jq!p09QL~3k~rAX>)sRc+?hJ*G-h4g?`4 z_tjGj*3!tB1dZ=2x(A9$PPvsJ0BJhHIKx6ouVB4sI6*zw?=_@ueBr1RTZZJu1b}b= z#nX#L;Yy%kV&)_)Jn@-#~ zSj;p(0PVWjC7FHiZRex}Z~b`Z*#3B%+;54Kz+U0C&8@g~UzHu|Dc>EfND=H>+Bk1!$60%RkRI99rb}{$ZX*u-2{;JdgJYDJk$IbheuBw2OwT~ll;B}`uvhC4!Fs2D4FEwvLSFOf_82JlH=*9 zcP<%6{0D%m?X6w1e16Ey2lX{S1AYA5S$7+B=L=KcQDE=J*I#nCqYmUIrvT!QxZAc^ zBS@Y!Sm%^liC(ZlayVj20O1RBuQ*40fD45mf71D_S17d-AV^+i9@g{NApqsZ4irWn zf!PZS?)RF>uO+NHzy;gR4X3^>Y|~0)NLz}pVv)LVJ|?X zdo2S32^#FEK8ErJj|!WWU{EdQGt2TpmREHdJ*vOyFf!#q2o9NA(H7Y4c23RvpnOBE zvZj2(R_~s9{QazxF-#`h!+fTc88SiBtt3(I=RWnZY!l7rq_@+-@u<9e`kJ;{vTdiH zyvc%QTC$c+iIlC5zt`I2*bMK$yUw0#m>oBT#-fBa>W%0)GqZl1W$tt^mL>YL#n? z7a&6r8C2jMr)_7@mKWSAENdgzW$D^mI{7^>SHM_bxpGUQt1jSK>VYK4(UU5V0>ry> zhBkE(+TyvX(Te=j&?4Vhvw#$aF*5^qMvJQwE6h0QmvP7@#KZ+?I)Nye{$S zH3_6$HlPRZwhyyDXasH>o2d_PjeyBwUIO@%2)S)VF^iy|mMK3?fV@sk=&WbRT2WTobs>Dhx#6hqTT~U3&4>;r;j$W$H^q8 z6>V0z-bXESrX@h<1=E~YKsLD-9u|W4+nF~6klX}JfcuuFMx`X~udfqt(!Yi`swVWq(Dzt4QMH0k) zH#E8J4rehW?E39wxDP9oTG`NyUF45PAR#z03fM;HSZ9ADoUC$47k)>gANhYJ`q4j-y6^o%sr%?>r0ygC z5!-(k;P|hJzwcwXKlj~IayD&G0Q`dk0rzb>5Q>ntOI&0uZ4JRD^OIlNHVJo@GYCrg z)}OiQaEp`jnJf*ud0~w~fk_NW z;~T$~+=^n7<9n2}9dv$wP!64VOx|oq8EHd4nQQJhjiM>Dl3t2gwgg;GhKdv;;U6onXXXuzm?3RfFU{c*hCI zJWO!yUJJ|u!rAg|HnI%200~_mv4BWZZ^U-M@r6t+b)qptJ-zKPizt>AJo-<0cF=*y8rq-M6 z;94>!Iz5+sFL+J59xFX^I0|=S%!9!(8w@ED1}xo*O-{}99Ds6doGoO|^3dD4q{GCV z;X%n2Jp&|J#Ukf>O8xb%jQ82s+oLr8b(*<_B{&wA^-l%t4`WL0)$2DE0Izp!#Uj`H zS0) z8k!X&HcSxJ1W>j)5m|(OfRqHwX&}bRlaDvb^Dnkbti=U!S0Tsp(x1)AcfJI`uMGB8 z2Q*j>JZt6lHJQKmj;zd#%lxfd;<2k-83FAw2nOm6^hvDM$}b53bv^6?B%$~k0&FgA zbCtDs?BzF{oh4-1a^bY}oPAnt%Y!MMYjOvwL4Yk0ytk&24HvgTL^8?Ez@5#h ztaKZ}Gs|IUv$j7c`T9dn7Ofj8DZeu!6PLyiXwrSYQ2^wP1^`XwB7t&knNU;`cxKwt zNgvqsZaeQS6xEb@y}yJqS0FjMA!L$!6i7}<9jos|$B1H*1L&coNfHdXRpTAe`IyT( z^owI1m=$1mzqH&62m7UUw~rWTZjlHAWtOu5$y3u#7L-dME413EbzVf-*Rn!6B00$I zusoXD5%3rHQfIR{v)-_5c-=}BLD0`8t=Wdmim05;=MT%tr=L?7n>m>Na>#{Y^*PMMDk#vhI`8yy1zZA1YxF4mJl8JoA-B!c9(r){Py zVi$NKmCiFyun~aa2NlcR$Kvn-U~SYuKkjU+wNkA|PVM$EK=QC%$LJiZN8Qv`+6E*q zVJ1ShyIl%TKQ5ua{iD+TkAGQ?{SQB<4|SHTXU!k@fb*MeYI?Nx{W1~b6C_@Nvthz3 zD}UvYTVM3a=-O%Vp86jpeBqx-=8#K!5D!>?`saQecUf*ihCo zZ078ZjTNgM3*vipa_FqQ+voP%%DIB1-n$7f*m4Y)&?axZjQVO_LPf3w(C?;8PCz+{ zz?9lN7{}Bv)mU*BSNRB$=+#|9qw4c*?zYbJRc9*cgK)teqkCCUt`WDYMqCMhMeEd3a}KVP3~+!8-OO#rV} z9aX?OK;6Fy#`>liy_O!iHE`dLd*`{>fi;cKz(<-hg+)tSgN~dU3 zOmc(4s<(5=Z<_}2`J3!#%^a_(M-Hdzc1CUTwpNQ}#dN83N+zLAPC%R2bmQi2xjuXg zy1(w=y9I&&&f#0#)780U*40e%G^AF@B+uBo>kRnJGTx5Wzh&low1a&Owqx3|xHgmN zuD9=WIrq&k$^4)Ei3H*<$QK8geJgxHX4WoPfgK5=-XloP^P6mP^&8WvAG^>f-}@6i z2oUH#V=}H^{qlmm{1@}GR6xuLBZ#bjZ56>D7bG(`A?b-*&TZuYJeC2DH$#`W4q!ZO zwb(6-1&bNW-$6`g56A6688_U6<+5q<%q&a8qFZ+|0K?M8U^mh=D0PFUoZP5O9)w1! zmPg_arAKbF-3TFQCV62M2BTz7Hy9v!0l~j=QJ2~3gDk+x*WZ(o*S@W?@L{kTfUc}IaPvX%!gm3K z)(8M##UuxFnoAn-;0T$}XuGojq*4F?R?3i>Fu-7)m$}SBTO_I5dfp2#iu@AvPz{j0 z0|Cg6K~A8YT>%Sn$~m0+#kD|kUlj3Vz&x*IWq30D{!VFX#djWc!vx7Qm|n1+N0~7C z7>x4pJ_CQt!#S8acErh6f#d-Hv=JbVpqwe2LMFL6k%AH{nn_+j{m$BeIZKb}D6!(j zF@9U>16hE(9|V&;sd#BO7w%jA)#}>2*Bb8q}cO>EW##a$yDocVeFo*KA;4JJ1le~S4 zpq)H`!3bH`wAMTCm&=bI002M$NkljSWj2d541otNX4_CtD$ zv!vej4?G)_pZUJl3NSW>2!^*;aL=eRf!3;eVju0H9y#h`3JwGx>MO^bQPyvq&yQzG zP4bi>%jcXFD`j-uRzVZJHtx2KX|PgnA$TW2ZHq}>1qxD&y!&}G8)u#1*&0~GHzqB> z-PwT^0?|vxtP^wJPS>^L$%|)3U2<(cTtI!sZ42>Yf@w~vl}hHUPOq1|*V$DIq~$hh zFjLz-VgutGb~YNIB$?#tFycjE=0Whn;}ZMPAC~ri^o!E_uYX;R{NZQilg$~ZW{?T7JAbW&plAUlpP zvF{TL`fS$hy1_nKUwfbU27BdNlaV13rEi0kkNTN-8S*O#=;;`+2+ewiULd5K<|K1t z*x6mNf@fPozcTsiYtCjRwesw0lx!7~JalMiQ%0P{Q2+b~ z0pMHUJj?!Dz4(4LRF3WVFwJ13EbVrq?^P%kToyUcg6A3NGujILCNtgFYLgF8UGj7D zN)Bh!McrkC)q}cAqJ!NBC2&vz`(FZkIrW+Y?`qv{PMd#}K;7+1lq}I|CcI9s`4Cp3TU-yBm-MV2*ya+@!8-*5M$_&E>?Xt3kHKLM z8g<$TuJ0x2sq=jlmr(>%BVd`t#aZ$zRUMr@yl8@^1mq3Eo&p!Yo0#MTuL+bdEG;Qe z*aEaSGuG~>*x2qjL1Q>F54ixavpGA0(jlgd5Zkg^jv@2B^#DBh4kV!4Tt_zcE*!Tl zV9rxwW)3>vlttn7EG@4HL1{WYOyfRNZiVODib>Axru+_B@>-XiAURp-+Uih83G#1e zbN~#XgT02q%c9L7pGQE?DFESQtPI~C#kIKuZw6sKf9ACGA`sSgroQI}Ku@i76Zlk2 za}8Q;m(f1%Q_eobK+%?kQ^0%LkvKX+nf8xia>$D5-GTqFZn2P*o ze=hZl?sLe-JaRMml-x;tz>-?wz>?spE$w0Ezty^0%)b#=Y ziLB?K70Y2BEL@hZXP%PQ=U%ja<8Q31Tjcm*S$^f4irMY~t8Zy^41>=N=qGtHfn%neHO35G~uvB>)y>1n>g@ zM%fliZUwM};S3?0JPhzYm*H$jzpSKVN{_m28&EY5fFOnUv?*r+YQ+uQvkg-7)H$$O zeUR@6c%Z3EfSeRU7(pH-bpY_U+QJQp2hoXlxJs)ujq8i!yH*zsA()%xKvHrmHhVh_ z*7mv&5;1jSPVi2hcNx#O^@)?x@j(|bVz9_87|b+2DT3r70L1*^LtP0PEJsUX1>ho% z$w797$I{sKq~n1`XHl@-dS}Vl1%?m3)!)ui_>!9#j|q!h^eij4wF3EqGV$; z0-ahfP&o_P8`K}ajUzBs8323AdqGx)|BDLQ%57QnE7#5)#SDO;)W7_ea=382X=!g% zmf=+3J3>ZwBj16Ml!VT>b<0f8brrl5V83S0Q*tG~T0j6z*J)RM%>Dk(+hDxThfW59 znYd;DH!^6;;h5P6 z(d?g|gnZn;|Ga$Vt6!If@3|mP$2y=(y`YkDMo$4jJOf95(mV_|0H7HCK=0RIakhIf zlPho(V~OA~Ya8wrH2HoGr8gZ$}V_mZG3a$%P{LLc~iB^+0aSj!b)IN^yL>DCI zteT3?3?yq9gw!#n@2$EQY@_7L#!D|-$c~6_4K^Rn73lDv!0pp04TSjwa%$WPQYCoubuDY-Hh7g z{HN>AkDa|cK+f}3OQ}$1g>oj^Sib!CFpbaEe70nfn}J2E zuWjnRi!%P!TM|fuHTT+OS@>7~LMFcQ1t;0U%}cWW<*!No*s`;&6fMh*;1hDK@P4(8 zcqxHWFA%PJuvozq2b2-SL}qO#niz`Z$qlm2a(mSiFTs z@lMS;=>l*b^?v9ha^%1L?Ha z}7MoseO zXP=XSU->6{lDQ{@nGoHFA&X$uQKd)T0%mF5=K8Kr7(h9kKOBWld7Fgj# z$qij{on9?bjmxOZX)wwC01`Q4wGPjpRhqySwG`r8P+3a={o>&UnO$}PV&>`UlCuo~ zupjNRP8DebF#RF5Uh5B!8?IuJTP+{nL9}~Ica>YIm22&7R`5^+@I>?AZ{OSw(39kE zf#U!mC(zs4nP4k}Q!-E~uU#3k^1)eyN1P~)LIDvWj;H)D*r!w?*S%Ba$X1i z<*Tc!O|`ev-Q#-P&dlzt-^F4VSYTlm1PKxlNRXfbEeKH{MIK4%NQye@#P1}E_em#` zqI4j25Fv;Lh9n3YEEl@~cCp^i%#Qc%O?7p7cke}JR7Pc2wHe#9sDHlc%!-VRjEszo z%zXcg_YeH>WAfTx%__Dv^D%I8{?Ze0B%ZF?ErZ!yuYKkZ99?@p8XyVMbP$YUIHJ_P z_@f4BiM|EA+QH&Biuf5jq;_Fv-ASUSxyX#xb)G z&R}iCt4;q&03iA9Eqml#Pe9H!+hrmTXD_8+c8(zkh}v{;yR?7!`w(>VHeDRX5=yST zSov+PD9v)0)JpT+cd51_Kwd6dpgi|s9}ABEFjlN%YLFZc3{@mq*(TfCJdT|K&j|$Y zY?U@fDwgb~23sxeq{j9#Sj$DZK~S#JG+Zpv_tW zU_~oIpEnNnuZDK*hXm)ExDqcB$!6NjOpOZpaRZc1Q>Cg^O5SChXA>$A5Kuo za6XVx==GjKz_`C%5wz3OJXEt_&Q=Uy+IPYLP?(W2FiC)OmL9qFp|1mXa(f;D9Ft{e z-5WZ=O!tJqB0n%LTkd}ke=0NOQA z6#(TVNE*8f15=VYS<6~GyJc|K?#iYB0QrfjVfpN(Pso=pJ}6I2-y^Tjj!JQ6O1cvI z=77{gZ@-Kjyi-~_du03fepJFg`~xx?_0d^KkR{ss*y2k)6(TSD4RR88s zO!6v_oQ|N?oh6mZU*?J#xRZ`{>cP`tS3S*!F9lIik)##I@LK^}ig3%2&xwqANhcNe}?zSbj zf^Q)*$r3(CyD&UFEK>uH(>9NpgBKqE8pfZzOrO9E zsxN-EQW)b7Oa~x(f$mA17KPgal$Q`W zG|qAe-Qts}5}sXPiYr4ArY1SlDA81cwH*Mf2@&M>NqAjr-M zMU3a+Xp)1_gFZUk<$Uw8i!2*e=6B!ro~>}YcF1%6PCiwr{)4vg<5p^YdlY>VF_U&| zhf~UO=1}*lh&Vu**Iw7T)v}Ct$?U09lEpieGL+%PcR=Kh17MKIMTvkl2;6aL<)iA3F>de*8Y0fNNJ6{jTUIMz@}&cewe#~QEF({10FU- zrHx7`xfKkyZX>X8BhH~FlbjlVoL{(!EYb{ZK&{g#h`owDIa3$LaLIwfyk+n>){90NIK-;hZD5_%uOugBR_hJ_` zYX|W{ZL+%bilO{v|Wwq&TXbsh5$LIU%K_k*}l}uU$6B-pPARC7K5DO>CD8cpUd8iNbJ~zg#Y+|O6G4K zs|3U3c`a?}mCW9|p}oAda+(DD$R?%)hbhBR$xv^FA7{|6x5*3!(z%bw@R|b0f)lNX=z$?`dc1>7V`l02oO_`JcM9!8YEvxZ^cyXUHGE3zM{Qp ze0Y8Hu(W*T83`|L>|-n49+J#J8?;qd_Q^ockc5VJ=+lrOq&C6;sA;AiIe$}bg+O@` zdeVg%=efQJF#oJ~2zts^T?=TJF0jeB-hV&91@7rG+#k~FvHSO063%QWhT2Mzu35SqT(V8ui219=tzPP`$d^psr;`WS;yMru0=m??*6{IY9& zBf~xAYe9eTORU>^#t{6Apnnk9Rr`TLn&3eVD?;w2xE1mg&`s7x4QwqZ!Zd#K2Ps3t zn*c^xpAe(a+D1@412BM$XRSxhP27*pB?kNSfYt6ryZUgA{Alwgq-mpMNea`_Ygwcf zj@glJI47)gpmU=&vxL|S8Ts1OBl67reXp>;q>q7wE-_|OBF)0BqK6BP#f%Z&S>+j5oGhJ;GI_A@WL4u8XjF& z1sNoAe8?K3D7iA7#moZS2L4*Qb(`XZU+)FfB;3HqfP-!VRyAyc#9b=uObKnhIt z)5VpuJO}_Y$mWGvFw6mxGoFAK0X+SEZ!>qVTBp)d~pop5_ok9;ku6*Al>5$8c09SZ^UMM{V0nJe7JBVbZf#c(RJ zq}k_{)QzuDr@n=H*tYrLByT%nQyyAT0Zj2(*5iI zLY8-o=*?>B#iQ7N4!Ye>ThHWjF$XR4m*IGOPHitG;u8P!&pP{{oq}M4`*#H7d}^SI z_(Dx_Z5ZazCSFL*U}jffyH3W{^~oZhc^DOyW1+GEyONJkLVGST-i3XMIP$e&4Dau9 zz8!d!4EhTlE!s%J*gd*ejzKSbjsu#@=n#$GQvl%mCK%=m7$5N+bi2VOM>{4Vxni0n zwesys68DMRd8-s>zafQ*C!BO2NFZC@Cq4vRcAEu~Tfnae1K8?Ja=c$P_er4Fc^*~D z!zuY&t6E$(4 z7HG7#JSn4*eBj*o@wZEkR$ZA%blvjL*4lgVx-aLYQn z`@;2Zzs-C+OPAbu14U&OBj4q0aGwKb5|}`w`VA?F?<6C0o{VS{w7eGQPlp7-UEPn# zhBo{8wK(RrDwAOBD1zk?K)zPy4&bSeW3^*-U)S5*OO7R!OSuvzIm_(qXa{KTFu}$2;R>(@Pl3D>vC*0(KrS-})(+P%~Y;tbv z@9lxa%3k4h+o-X<`Ru4=1g1@?lUjMMN1D3HTfi~%09;nHOeX>tzGtUg z`qZN`{gaQ#^vCXzneW*r^Y7g%@ptq|^1he^9zG<|gDXFLS#HPB7U|y&Fy6j3DGM`q z=DbW^cvDHOXgzY~r5<@H4gm5JKl-e9A?j!^dLS*68kK|w7XUVuAUr++HvH+x%KItU z(k4w^E1f=MP@9q~=`3OsWMa}cx=-HmiJz9PPye#KzGH6%Y-OEj1j%!F*8ou(AUHn2 z35?sIvdJnZZr&E^0iga~1V(*O?K=s8>xnN<)&~`iN>|(UUZ|v2 z$Rf9=tGoB(?>yd#Uw|56|~c}A8!@aj&R~sGsy`NSw?A>dQ5WW$Y6Y3 zh&W?$Fux>CiF5YeaEipCBYgHu<+m5lcAraH zm23gX8sH2BfTsC*1XZ4PSm~95u@Xc_Xte{7GF{D>xJtnl&Ex2Gwr$dIAn?+nDV32) zUbI=p;Ye6I274hlV%_tZOjssotn<@lZT=)|qYtyWLPHo|)?u6pFbtjN zZh+kc$!D`MfPj`RxqElsS|vC;QhB@!W8CrM`bRqelB4P9S1e=p5QI2X9jf+4+o!-# zKc5>>$$YO1@aC{B4Pf&m94sD04fVzLC@pdV=Jm{yTEW0>8>%Eb0dB^OM-*AxR}&;> zkXwJ(9suPd&}cRe#5zEV!_(V|Yp`;B)pg|3Vv<{cGJlgXt3YzJMYG{RqC+QYOc}w@ z4wk>`m@Hs}7Q*bDZSQyl&p)fL`v64xA+_|vzy38Dno3kkOT$a@T-!cbDDQy-uOw5a zGI*99lzsol|0!L+`16n?bFK}aNtwfl>G6tvUJT-i89s%jIl1(?-;#(S!9=iG_J!rI zC*Ce!x%6FlwpoWHV6B~5x)NlPqjq@4^hzQhR>=^ik+^V-M;|xFYPO7Gl4E5V_q{gq z?mF;%8!(*c+&2dR-jrdY~oV-9#}q zFnBIwW}{1Tg#c~JkXpe9j5O`(lHBn>sqE*0G>@+ffz7vmxRPeBD3E-=&5q`?)(E<7 z=F{ucCHJ*~We+)M5Pa@1p2aA-xa8b>@U%;4Y^-)Ep!Sy*-l*D! zs>qlxDY-ZQs45*FdvF4bRMMEQn==P-JnvjwwXxcWP7@e#0pxlw)TNqDUXrwn;jgpS zTRLx>^>Q@Q8!gaift#)ctm)DiO>85egoXd?i_6M_15Fn>6 zdA(k`?aupU=+^t>+~a>OslWJBX`NXKf~DJhGCOduB(rx*3EJH0l>NbslO@=vn1%od zH=;?-cJhJU>myJJfS6VS3z=HN#9_3>@B!R45RzlBSSgEh&5%rKmO&7IsT*FQ8`?X> z%Q3Kt`(+COasZ(WJQ&7@w6;`J6dKhn(qRPp)Y}Ts>Q3LjOFAC;UfI3l7M*-qOUifT zz-SNJ4E5~P1xcPhCLRQ7OK2$w#Ucmm6#*9uM!?S5V2fOWL_`Vxf)v0L0s~QiT{8fh z4yW2A?AamP|cFeMi=K)#yk+S96s{NcqT)a*H+IIpxlaHXXdf3U|H_*%TrG+R5Y|TP|oYK>ri#? zYXu{$IM}a(L|e5o{x;zS2QwVpMn<;^+Uee4{4R%@qzq?)%q><@izNVl;T~zXS>$>{ z5Z8w@4Hjl{lFqe5$97PLeP@9X;9L=yHn6i#1zOYjxl~MYdqx0%`_V2z-T>MF!uZkW z=s3y&`695B1w)jq>P!_#j{W#9hjvtQE7*rjavC*uDQ}U%zLY4)Q04?uJEn|_K3o}$ z{9;!x5H{<%5H6&oEjfd}!yI6-CezY7>pU|eU1*2B2lVb$%kFUiW}p1D{KIFTlH559 z%;0Q{PV}ApTu$X>%z*60G=l3nAet_urY5=F&vl-wxiqC#SS~R)N-xk#0A=&`I7Wth zb4OI>PBckpcv+ImVL3fHDEl$aXeWu`7MB2m_;02KhPhA_h9F6 zSYAFFl9_BL`Zd7YTDs)WG*O!51i;e9IwJfp`qZWErt0ctQSTs`!zFp>d`>uf$%D3Nx^+k* z`_EY=Dv zuTEail1y=bXTOsb&0EIvrefG0938FvtXcs`u9)OiiEJ)7hA(0kBH|AOhTm&##ruSk zT(R@g>DCJ-`DwIm$aqE(;0O*4x=PeVzzGe4OZ*F8vJ%i2GOxTQ^0AuPYbF{`>zdme ziH#O$w7^Z+0xg&pX$G1kJ1($>#`0!6%^+C0Za|!@Mh?=~+uU+1WPxr*kd8s^Oa^{L zvE8kYsq5WlH_P$d!8@A(u9@I>z4M?Xg{4bwwXX$wIp9jY{oYFHJGX6>U3ftzqx@PM z0eLO!LiVfgCL7PLb^Dm*w}Ao3K+SmVHj^z!0~`$3(|9jYVx|>=qG}nh>n(eB%l&uX zA^p$=*Rus`<<)8I2YMDq5ii~hPP&8-kdz&?lg|xHZe?R2Im_<@W3{WRQ-=EcRgg~B z{c_D>lCP`-+2tWK&h2_ax1#NWE<{y_Xni?=YBa7FiB$(?8X+jNvYIU_iqE^pmpx zQ~yeat`Hz+p*h^InN|lJ4nQ9^3=RJj&T+|R*ivQ%6j!iw0jMrGZJsRmfl&uE0+zWlZQVwNlp_;AV(H4{kzUS?- z^;iCd9J}oSSwQ<1z*C|&Ea;58B_9EMhNE=rGo~> z;GJ;>?EhvS{SrN}Uu`r05bRDHazJE5hKxSX>E;B<$-dX77|#G?tK?QtFOX9e zliUMFDIE@e0QQuy0d8I;xsroKg2{^1<#HKni%ILq)=E+Y+QS_w3%Cgt=A~tt8sn@Bc0QPrT)E_cmz9iIA@D_MBMnRu;HR`u+HfkE z`LA?c>9yNzah64VOoQaCZyEPh$#zT>%g~)KMDaWc%6YKJ6N!M#8w}Xh>O$uH;$t&X zKs`d8ZybrM#p&FjQ?$f9UUW+oc%GI zCEy1l1D*>30=#f2aj5bGIA@-s%THWFTFw|_+p*6*d!u-!h+lxA~&8V--OWdK0%$iwfcxmywA z(-LA5wB7%c0F=ElIn)ft1ANrR^PD?|fG_O+grLz*n{O zgp!1@t%52$b&~+5-)AuRPo9moj$F!BN}wlb=GLwys4Pnl8oo z37pp@jLDEsx&2V3h;l9?%TFLq%hc1*Wq({!&%RpO=JA#QQqQ4XET;lDP$DU~>qjK- z?~t#*oRGi%w&SR%?5Ohjq33-70J7yAFt@E(x0&Q_U2;mbgw|l8uP93=;D+{fyPgTU zDbZ0W0i)%)^e$>ifP4;`;E>?amNtfyGIhp}e@@BsaJvWX9UK|aS=BY}$GZ^6>ZUo{ z0Q^7$zgwg-V$5+%t?0ebp~8Fiz(Huw+Qq)vvuC9MsmJviZoP&zTACPAi2(~!8#S}a}zK_uSc8Rh+~kI%lnX~&dWO&T+0}AKpu*?28KC z38%w)Y-BIm>3{nT34G!wWNx4ps8d-^H0_k*fm;BQSYJ*Azx32;KFuk$V#%!lNW}-} zM$89E6h8nEe&o?gV5n1jyX^X)jzKPn1v2^1df;$Omdu$sS`}lCbyfMKr&K}sZ zcO2TjRw99-E;XxXS(ujW@S#N^%9P@_Ldh z0VTQOAgGou`O2|p0W9yYe`!Mg_H!5I)u$I^Zk>R&t_PRYd0&Pmxi+#^O+9x?u4rBI zl7zr6Z3jpmmW$I{)U}weBO9a3)5{Wo&)g&e8r%1*PQTtu4}*7hAb4jFo|9c0dydiB zG142tGrL)4a+YL>rAuy{1^N);C<3VTDv;cVcMby7evBX5$w;7_z;E6)&KCftE#q5` zMwyjCu*i3B3ClJ(H??Utx&B>m$$UpKW@w}1V<(ux7IWMNps`W?z6NqWUVzNu+wRtf zSI-Eq3fj5dNvp9&V74A;&u=z5d$qBzw!wL+jrVUFZOguPrRK*sV*9ooGTG}mXjvwo z-d=o9cK*N*%JMv%(YQ`Ij6G@Gvw7T0am)~_7_Mgsvyz5?`ez`AVh=7XBLkrP2>_EP zFf-#7;Vh^ailk0E4Q>QEpB_SBS&vL*+adP^@&&gW z`@-d=q^lVGxQKCD8{0L8_Xno5*zYIY#%JV&3-}LWw%Jo=z`=Zk>Ai#rR@BC(f+Q8QD!~MD+ zfuS+h9siPi{mm~sWtnSub~oL2t3>a=*G#Rw4nSKm$uSe8${6%>Y)4hnx_{t4_c4}< z)2z;}><43`&Tn%~>yg`OXma`H5}votkc{^@Nc|oDL%i2b;F{vuXgr(QS_4V+3G5eL zaNhM*K6lVg^SHd1Ob0fud+L=}%p9w(*E-ErcW7*Ev_PW;Hh&8c5D#K7)u3)P3TF9D zB)7sAwls*VhbXe|ZWP!b02$sQ-Tt^xB(;fXVDtl__+Hn{^BP}0E+-pTzsxF-zyAGXVQUJ{VdEMdx2 z5GdA0uyP5rRubF59RBE|vi;xwt_(l;sNQ_lmYpMGGH}Z+l{$JbWqszv1v!86xB@JA zph!3&O{qo6UvS=w3LVf-9@<)|2g#Gm`0$Ja^qaAP)?%Sc<~>_w+6Jcb07fzh=2y$e zq;G{Dt`1U}1yhp&6$$vdjXKhqMaYk2hU85Cd!Si+CjeVqTQEvZ2MA?C7;I8o8BF&Q z^ja-Y=#c)Ejk*3%_l z1d{T?<8$)NKg`G%|M0B*=U+V{|NiG*m(Ts~DQMYG$%$8&Dw0I*dg|Yn+=>I55_2;_ zat80zy5t3O&v`Hdx`fP_xmNtUlCm%+*znLuUJ&9f*+GJm@=3a5`})3Ziq; zp{9*8O)+~G3)6U{EXIVADRZGSU&hjlnV^xossZaFX6~du_%7+$iP##Lk(7t?i!wP` zT(3OF?n{T^tu>4wag+?_kAAO{ii#E&;1GG*Nusn#Vz^HZee9>f^mfv@R*_r5CDV>| zrgKIlXW_7IN(h>KA-zTNWgi&j{T0S}6~2M}3y_UTp*@QR?!ofgAydVcjTv2%^_X!v zP%K??XwC;hG7G26f4Hw%UiHk%%yPU+0;o*Ml^JNu;5cM72S#lv5_UU0hck2#oV5BN zQjibVcesohan|+qcrk|e-KTxv0r~8V;}~Oc<=leg zj{Y|s0Cfam#Si&K8lP1y7``JIUt8i1NKT9^unI^{<$)&j%ZimSyX$3#)n7@kv$FCm z>tt*sH(H?40ykF+aEhvy-cX?K4FE^3q{Tw4Ki|9zBjfExfxb!)-EKp+=VlEYNa+h* zH-bqw0u(bY*-f#kf_LJhmqsVUdZlzs)_Kzg80a@eqinKD(jmdw95nW8-N!dsH8)Ll z2uibGYl~ufs|lLx^y{@$+T@U28AgyDvd8y~@4W2#&`F{Jch=t>w3A~X@8iv279j3? z{IgD>5E%U~!0$%Yl^+>=w&@uL#bN>4?pdYtzTP^K-QAAwjO+^?EAIflaUX7T!Kd6LtW0*T2hxhZ!iWk)8{39 z?uaVBIdoojWd26Rre0HQDkhdO9F88^BZuC3U*%l$09$4N4qX82N(b@eH7@lJNY<9x zC;=cZx0r}m^%+o?ytTt&_@*iO;*wn99t3hzhMWkmuhJt2U`h~In_`j!1k*;4yljJ4 zs{!Rt{^gXs`I0rYx`aMjgr>G>Ih;p2h5&hF&34}gp!`m?nR?`&$e@|E<~rYv3ecHN z*nm_HblEay;ItWfirAI<|)YnZ~= zN|W4aD#TW>lTncESNC35<3iXk)?FBiA{L@YrOaOJ2+9h%sj^C?>f} zmwX9e^P4Zl!2)*xZi1sc$KwFU$uR#DfaU-3mk!Gx{hQ+|Z>?jIp`SG5R@kPtzd>a$ zp+9OSxu4ID=%}0ri@Z<_sD8cxpl~fyW#B$3qaANFlbm(+1Y_dse*o;)k+rJ6V!M)D zfi5|Ncg|#>?++|1IEY5G4LCP|GW3-qK;g6o$>F%cH4UI)>G(ki1@RUT#1X-fP0KVy~!fwV`OBOjLnzj!*fEj~#o`5bnuXR16)XI^w>1wGJ>P^E*aT;Tm7d!g_mTDuL z+zp26ZI``7LzJ7rXzsPgQxLg4gIm@tj!g17V`c+EKtK8@zj?!G3)ARaqt<`La2fI# z)AblyNw-&ubPzl~E1AyU03j9_@q z%Qz<4ud|pb6Z$fkYb!d%GR*C`ME!mUU_nhbxIR%E9wAPzxl-eOO9P;Qz%+R*>uEY$hP5vO)R#>_8=*H+*T~ zsN|0PS9M`2-=d9IvKwP~1tezzk7+sqbY{K8_Q(O_Ifn8xuK*-3IL{NEwbruwPOO!q zk=AH|Mho1OEx-X@Gs($}yV<}}vc`4&^#!OU!J@Za>!k^jZ&tY#wgJIf&9Wu)*6nDx z*6nmNmfwmglUk<6Vuw3S!qv`4gX9WiGVZt4_GxUpycQt9TVaxe32B0)msjFjdoO~X z?a(G?-=vYvY|08<@3Zgi?ve-Yy-V)B^LBjjx#xdwtTyZi=By1epYwUpc5J&+Cb=%> zYAiXP=xdoo?`^yA~x;(7sO2GeZ+j%N9xf4)sV z{K-)n+l2sZK&8lrC6MaH2d4-BFUJUy`w<*YTP-Fzw~+~KO0JMOopgb$@zkI!EN@p# zaRRAa;D|*0nHEYYi;@FCg_q|3`-s>oQ4cWVqRjAV{*HEMRw1f6AwTqZAAnQ z^*}GR4ej6F?$mz)!0x2ML}oj@?YnN5oew`Caa@2Pf>nk)YqR}XHktJ;MtPLgAe-E6 zWZ6?QU7HiM)5RnQIPBwvw~aC!W%Lnk^?>q;b1C`eS0?2&Sc@J2mmU|eAL@XEW6P*2 z#v2_G|G7xlLo=+);xn_7Kv%Rty zL~ysg12v^qSRMgyGV!%hUc>~+sY!0ht>_$-1p!FnVv-ZkRZMa+^lfC9U^$&rB!?UR z2q7E0o+deu;sYS&?;EZhFoNm*BZR`1<4XF04@ z9ZIf@LUM(Ub_VYxvgo5Oc{?FY7#gvRj+e?KfIYxnj_*xjv?rF71meUIg4CR*k$@A+ zv@GOorhUUoo`uKg1prL8bQhdHgLW(cN@gU`$5vyLTT&|y+nm7uw%(weJCl-W+;;?? zeQ;E1vvkQiilN{A#I}!{IyI8tx+Ya6* z_Nj0|+_@f;ocj7%$f4A_=jzp|2-&Xq&QWM$dnG&kc+wJd}&dgn)fNM;X?uh^W&jDoqnh-37&ht;W2Gu$+S*UFG=GRc{yHOKiqn}PP( zlQZ|qh13??L4rPq>S<%Qqf;zJ2fD$?Cuy*q?|v*DL8{ksnT?Gr>ymK%6KKLR!Ug zoG?uV@|<`=3g^EfruI66caCEAQUK$zGSq`XJJ*U+=Y``t2vUwR>{>!Bj`TVuSDd#h zYxlv*Pkp=tsXj{UjO)@!#r__|;S zLr)RG7=8HqN~JP4yxlm4U#X>1X}n0=FUlyaK-!RPN@@z5mJ>;VQn^mPVN3Q^Jsz_X zoC(Ee8=qsF#duBFKeQ$Ae4C!0k!)*ywz1Bumj!sPbObZD%L9lx23Xa$<67#*S}OO&46>g^I)#VPb0u(R~{dJ#xpsPmliOl;ld9(R6z)Po2iFigPFD*!{-~f8g)~mO+ z?2k&hYgpvG)i*+!f{aC9k-^gI(mP$v3e0st_xq6t#e3(YQaF{uxaia)r9UF=q*k;; zv@tRS#CxO zWiaf6kktw0DrFWY@{&KZEFLgaGu@E01Gth{fbMi5DD!f!%mNe(FIb=k>ucKLViL|K z3w%*$>Mcs92=HpoIyW3pF$PV!m24(23%D-`>eASyL;QpP5b>zgi|-u;Z&rcll>n5; zp9DMg!l@;hoh3;|6;bskkLmp#koJ*Ro>S?~ct;#tnnzg<%0y>G0Zj6G{ewGB0wABd z;Eb<+Ad*cuzgm`G%E|n(c>tjn7^L%%Y462-sx6EDK0TShaV@7s5@X40&m@GJUy0cs=ukX?S}&A3cY0mL#Gq>A}^(WxsH7)+@Z?q7mTOll@2CHoyN4aaM1 zCv@=bnG(rmu#=C^O8=;Fqv-?N%iJv3?MLHs9{0uDy8|+FYF51HH#e*00W=SEL2hLM zAUO<57ZT`#<7W|X%V@(An8)cnW?sVSMSXHFVkRucTr@k%!y}TofmJ?uR!<~qb>RPKR*Mh zmGjQH5(IU<1ED2fwxw<;_YzR@Ki0UN!@YQJA|c~_PM@=mbcKI$oR&TJ%kr0=fHM!X zQQ8Jqy?1!DxkO2(7o572LF>nO0321dqx;@VVm$Pr%VT1Lx)JjD&YPZG!YnA)w;kH0 z1bL$DL|Terj8og!cW#ATLE7)WPcpNZU-f+-QT1mV&rW6~iThN6B@HpfS$O zoS3Q8zdbVa13x0CW`9QpE?5_nY-|r4z&#)MkSco_qUt#uQyW?3i3IO`ycf{XmoT1p zs7MCy3-j{}0KPFEV|K&q@q%?+;!@c?^-ZL?)<kxq+-0XB!}Z%=`Ca+s&?jWzjzP7#oJGIIY|Sb|xn9Vg%;DR4g+Y=g zIP(&HEXw#~I|3HDHcB=nI^HM2_<8Xz!3hClP}I!Cw_al&X2{~Dp}nty3X$O#_vW8G+N*$Zvj81_3$M* zg2C)&W0AAz1CYO<#+5dLm}c-yoqoNRO1cDZhSaRy%Py&lKD zmd7BA+z$q5E#tkYr9*0MzusvhGhLTupC}owWRq*741^g-QKw(+W$WOG zbdB%Dhpbhy2Z7IG7qZejnsLUKGR~d-3-m>RB6y!R7>K!CSe8=`^WQuOL7zOp zCC!M`felsP*BNo0l{|&h)zm)btOerZytJwf$<)<`Ro)8$+;gqS!2dJ?TZsiop#2`ppC+Q>-!3U9kwp?4h2B%x1gPCNF;btyOef^tB2| z4vl4etA!B=SuxI$u4I>R4|sAONfro_Tc(Z#F9p8lphJ#;lmPKQ^P3ms_kQ`D{Ey$d zASa)6oX1H3so#8RS^mdwUy?8V!IYdo?tsc^NPe7z14~PGP{*6>;4CCsF50me_}ua1 z+G`_|e13W5xTCjU_7eT62;g`L<8l_kbk}1=bn^u|@@%6tO+Ycp(Kj*vx{Y_fPZ8g& zE!=<6c1W9fsl9*dTwcBf8JzNhb1#{BYOVW0rs%ulKI!?5efXFn~|?bf?a5$~U~cOI0M?tZHrc==2wInNJ7WXsSj zJim&uTiIBE0STzhB@TG7Dot|y=FE~f0aX?Z<053|C9cqfSy(>Y)7sdyGgSaA_Q7>+ zYigCT&QYCB3Z;Uan!6yS=ngnkS3W~2LloLl$zHYMp-i9|+_P#c&Kbb)nOR4!g_{Y? zKYQkR`S$5&UAfo!6PQIAL9G29mUEAIfFJDZ^%$K5Gs}w@7PE44;exXZ+QmJqixd zn5qb-722x{*Ivt1thMiV^7hV-7+$Oj5<~yu{jy$T8*>_8S9b>jx;DoD^|ssPQ}n{f)F?5$x{xxw0&4a4rdE zX%W-)JKH4>Nr$+HJXj0l2!tUK(g7);&dVDwRe6^Xhz&9rG=c$P1c7RGjoZUpy>en8 zEz@=YxLk8UG62V$vISZ4w92BlLz)3<^bGJF$$78HgHqnwo{@`oeaiqGLXZsT>|8&y zUiZh1E*E;Dl7g&(Kj8or3YWioOVq;a!+sHiaJ%wS;xvE zbjD{Unj}?#yav$dB(!)tyP9QWtAzv>DKi5$YZmo|TTWs)W?N@grp{X+EE#YCFZ;zx zpqvf_zsEs(2zEC|pxxioE`d(RY37?wKz3rWQri;bbnmchGUY7y7#fRk63$7LSxuM5!Dk1km}%mH-P)&UUm*8BQ#&jPx@B^n+T8QY1OIx;D*K4tYo5Fas1p}Maw zf!SPwbb&Tr3zL0cNyo8Ua6Lnvh-nclNdGW2ub;E-s{&3p3r6#5V`6${8JgGADi**o zI97JX{nBGxFMXUFz05)qhg$#N_^Vk31a8~iAzOAjK(<-G9LDW8k9Xo80jM3!LN3DU zn`Dv;90}lrn2ro&I<97 zjn8&OzmH)Ia`&@5*|-wNY!b^@MjA7l-1(NE*?eR&BhSNm+Jmv=fl>tbT>aS(tDR?t zQd0{_jN7;n_5h)D$F8!1{WY1Y3S4Ilp3a z7K6`Uz5Z1hAKoF`d$+FGzE(mx8gI zN%?MjpEBNK7ctW)D-O`?L4Rx8Jzn*EQuU(L%JfmmV5tljBHKJ9siOVN!*9k9>cle% zo(IWH?w8`+D|XhhwBvcZ9F5>Xk;46fP!T2<0{xXEPbcyu`8tvMwi+a`@YN{I89H zJ1!nic~f*VXSxJS>OWBp$Lp>f>3+ucXJIzl}gz zh8{WhCo?r%)*6774^Dxd!xH`Y4_5)?JU*o<KL>GV~22 z7zjeWKqX{D#@mW*(qe<$WNT87dD#ej$N~ujrwq0%!CXp%cmVV;u%u~TTr0AswUq!0 zl?+{Soz5kXd&fJwztYuKXkkKyG zB+met*T!q8f*>NF2TXg@EM0Pdm&nv%ltE4N2?R8(XT_#w$#?YD++1eg8y=J*WJZ#& zL#pqT13Ky*te~9^=-dS0r33otTGN|MO)?il0712JJjhI_cNxFvCV6< z2rcsCuh<|W(%TS7cGqof7C`Rou6_@U0z|Pbh2u<~tv*g3*TU2jrzWvmZiTyZc0{&Z*LhVtkkl9s`57>8E}~n&1DP72A0Zrjcw!!)01AyMgja+^X}a zjQGQ66L-_IQFJ<)8y5!^iJGiQh(tdAF z9z#VS>4dEM2I6W(`%620f{sxM53s5JS50vIeqzY`lS_pon1NMjbJG#&^h$t)mqX z>}EXoo2AbEy*+a4E&HUa1FVCFt*ZspY!bBBjSSkku8M9j6VIoXd<^Bq=qTJw#`^I2 zL8GR=Uf&FLj4vhRtB*Y{(~6bHPZoZpmSrKnC{I84g8bt*o?3UT9X}*GO!hDVrvey% zWq=d}%e5(x+y=OH+D4bbp0I=r06d4FeVI$ufPg!A=Uj4FCYQ#fgzixYHA%71Yz6H! zfKS#V1Q4JJz>GEvBzNks)33=g)&tGZ;i{q`!74K&d!+Q3 z4FcFX+5pOS;VZ8M;E?ROjTqU%trQKmjlgIEP7iHlVH32<*+4fXEXX1!Hs%Fj<_5_@ zs}>KV`Dp_VT!LgmJ!?k*zq{VmEeGy%#`-MYX_5dffR#({$gqsxYJBw8E2u^~!T#s} z5dit$&p^Mt34lx+f(r&E3>lKKJ&uEdz^)E{M=g9G+EB5W?J{!!MyUZ_%P8@Y1t6!5 zV0VZhxdCQ#8zn~=sL#6d0M-EH(t70F&izQ_q)ZC~BAVvKZ{BZ=sYj5U5*IW|F3bVI z;{RF}YHgR6En~GdNu;Y=I(P3=_iGNi;~4}&bQ=NZsdElUPJlAj?7#^OF4{EGB%whQ zB&Q?G*+-Yjz@?_vG#w8N2zvf_QuQyLkNsiiftBz71habp#z$;Nmp=D}Y3JN^FjQR@ zn~6+v#xU?2eNqM71VB$CDA$7VOgqrl3f`GSpSt*_<7iA{ygPX=EvMhGX1nNPU#%s9 zM4l`Aast3N*H<&v|MZcB{Q2*nk>?(ttz>0T-U}z|gZWYP&9T8Y1n;!&Lk}Kf{!Rds zY(%tO79cr$T?qZt4`&G-Y<@Z!?Y`*2%^Jjvi?a3jQ`+~7&?slQI{=c;g7vJ8jNvF9 zE(S(iMp_2fn6`xAso(tD8TtD2OY-^09mZ`4&ysGGH}SgD)~u77fdLTJpLg!VJlbIf zPNb`C9*j-?SPV>Er_a|rh6>tgYu&3nxR_cN#sbNmXOAv}<7&isW)jqmyFhYEhRj1Q z=#u@cU4(qmm)}I}2jkhR%UjQqeP*6PcU@x^+2lzKO7n2AWZ|4%%>rOn?|R_fvh9I~ zS7c)!HcextOw%$|YCDd{cA=Sh_#2m3nHhtC?=O_PTEbYOjqTTp*)}vf@d9|hM-7yo z(k~M=3Dj|-om;%W2yyMp9fI>r#1FmJ;?)U!ob*G6bFrd#3%7a`c~h; z-b$y^1kRN~bW?$8wN#r3W_Jy`%0OH%-|Mb4{3Mfn!AVz|_Dy3|`5e+A>ejxivaL*j zeeq2RY#FMgRjr_31nWq=ROQ1D0_sf%495k|E&cMV?rbZ5ue=jiWN9QcTA zW0V6!3z_7MRC5zrEfC~Jod%hU1Um@SJA?X-s?F-hrQ`}V=`FUU^?z@I=*P=W&p;(N{l@yJ@woli??dsIy+W>I8>cOy>>0NVmU-2$dg=K{@= zL_JePp0O*zV8wtmj?MkT0Bf}60CuJgcCSvq(j{4)1h+Z?7HR7o-6_6(ha~aRVU?6e zU5-ZfOJd-=<-y?*ovpGg0SG9eZUp_J1W{}~US?5A-4C8QBt8~zi6-`!n=$=k_>B%{H8d5G= z&p1QyJ&LLRJNjgB%wi?eu~WqPYIbx9AbAjVraX@}I!_3W)-uH;H<*mXcnBn#TJ35- zoP(uHj=-F}Pl)wsL{5I>9KAkT7zF;3yv&qfArT1-x9{up% zamhUS_c}=}i#U$IyGz=)ZI>Lt%RJcJbgty#fC|{GYO<+cdwKy3QwM|&qK>WjuTu!# z2LX_KQC}ZsYEZVOjJC_;9;DNOGeNXb|KH6dCqTYnGx(V=igVn)zom-hulMJYK=#0B zlbm?X0!bNclT7-UeVja;F;z@*TnjSE>8$g$_@Fs&fv5^3Hvm)u@EtgB9lVol@_C%M z=_t|l;Qet4ptx!CI9K+yk=+=R*4bCh{khIEp8D#v%)r6GxCrcXbiVMOFo8tQaM*o& zyX?86T}H>@=rL|Tp8Iof3Ox;gSv$4#aaac1dip{J&W4rFm;Gq}#GZv)^Wn9V+frS?6L{OrRqpw=P*;NiIXl8*Ms?k!;m7$Gp^gzhy0 zxt5?j`lDMUhq33kKSu^DFg6sKLwQ9Ugr{^C{ri=t7Qs;GR?8IN-V=~l{%l4kK^_ZZ zANs{9F#7$t{yBTH0YUOv+_PKzojzFE={l{{Vv_r9Ae)Tn<{INP$AJjFWwv9H=bvza z-_*o^^|htSSuoE@JY)X+$p!g=cXn3N*So^8h`Q(bSan-bmpqMod>;2T9be{*k@c3% z%g3Qk8?UW-kC=U1ZNv4XnRy7_Dot{1GSiu_fbk}UHdBEl)P9g3MZZ#pWIA-(&@Z(S zFP}Et>BoeOe$>a1+-9aXrJ-(-@bs=ju-U!D6a z2@E^{XSglFgZ)d(&~Hy^;2%XHFm@(X4bn51y)uIDt=ljzBY<|E?{#R6^hF8uAF9X< zSsC#c<3V6x00*|OX=Ov56?j*|`wfsBoiBa(u#^(ZaCSJ3arfTUPh9ub#_vW8G+N-M zZGjL5YEJDYYmH!e!>+sq9(vooN{@WAwO|FE6o5E7$h}qi_y2ZjK(@rk&%1|10nOr=l2iSqoHC*)gyGp(40dcQ@G3Kn)EXzB<)x^1I_ zCu_`1C=z@Iv%DFBGR=NbUv7m=atH19 zKK4@p(Y*d0HdB)2QnJJ?JHm8Y5ZA~bi^z;UyTwWcu!MHV_E{1)%1Hp=w}5gor;~8z zl;s%YPyXV+oRt>={A!~!sRR9_9M9<_fHBZ&hwi#Bif<5;O^))iE;hNvBsUxn0Q$}z zU6xY@1KVPew^hn_uS5WJ_m9{pC^oW>UNRW)E12W}Sm;zDt3N<3b+j#w&J^v`<8OlG zrjvo!lc04iOXEJAGJt*Nq2$B)lPP%_5+Rpklo*;tAWBM%oP*Oxw{(C}U~K}43# zen=qvv;Xd-{KsEDB47K{ONxolcnn|u%sDmIxB+rqC(aT%f5KtG>vX*g1VIk!u$ZH` z1tqiv(h5U#=paxgG|36zhIxPEx+|$2j?F~QVM=Ek0ASwk;8}swAv!91Fw=kmk`+EF-^y<+c*x;8K8{-bJ75LnVx~e^{6q!L#I7`I4NIy z!2!>88EY-6;U0%Grr;i&aJ@6|f-rdJEM^m3u9<;5S%wC>`QBkZTeyvZVcDa2Z7arn z7-G4FPS}bu9d*wz;;3!xj$KADYX0Uki~9F8z7LU^pRt_~_HJqs026y>-}XvUo{}pQ z$57!Q;y1iaLVJFy0+6drXrB_^`7R7}sGmh#ymO=D$2T@IZ9k-|erIkdiRLKDoC^y*Ttg1g&<9kl z%(BgOFWx#P^##(JX#8ZB_sw?GgYSbYd!wo#4coBqTbC!pAr z3MA*7Rl^!B&}e~13%tczfHS4sGmVtg3SNXan~^!r@wf@jiV);hHiKQ7oRTF3!mT#* zqcWK|XiKfsW+IW{s@ddZtC78Ff}Z-Am$M{*`Rvzb z6!=BZ?eQ;6$`fCjRNx-V%b}0VBFNNBk^AM10vKbU?nThHCTOjXOD$wFb&2yegW0Ji zH`)M%b_2AcmC^v!5*w2k;PF}wMk9f=DX>im*ggGsGga~q1mkb*L;zHRWH$Jy>!+7) z2^Km|sHMIhBfI-;ALy0A@lMHlnh{R5RnAPd%h4C(bv41wa*McjNdWvu05m?0`jZi? zjb;CnpM977&_DYH*|mS4PF9YG0`{pGQ7N#?2{3CWF+qeZz;lAKZp#BSO)oq9`*DFn z_VqCH66p4$uTgirmho2Nt)kkGOkJ00HaS7xE|jlx(5b`gl{COObJrYz&qRyO__H?z zp{v}y!vbV@zm^Eb^17p)AN$+|`HSB>BNts7;5_fqnj&e{)>^%dfv#rR~mr8HZtLs}nSt^}hfxzn(F$-q6s9wVB`) z0&hm|^rK#0$)!ogL7T!sAZkq0|uH)fn1e$EMgONY_U-k|jF zX^|N23k^VXKTB?9<^8J5ydFz}&vQmB2R^SD?|?=pLlf>Bw~=Aqj`P(<_gl&U$x(zP z;B+{JHgy|y*uU~jTwY#9d)~cI5##e5oH9kUUkTUKG&0G_?nf4xhEtleY%a&NV_5Br zLqL?rDh3M%`_+anXE`RZbQAt4F@xLy*mZ7eNdVp%k+477uiMb^VVM!!-+tWh+PKe@ zAt|QQ{Q$|`y5z?uPx4}kw|P)}J=HQQlt!6A0M4hHlTyTSE}<(QYcXr1OLmq7p)bWJ zoCGwor$j&-GOw9O@DO*#>I{MqQ=fb(Q{S=W6lAS=5tO>;ryc~OeAvCI=5I3Ed7OON za<1^4j!m4k@}lv)&COtp+s;j#V*oyYG8p!SOOHz_j<%suYQ@weCjj1o;I>G24C4kQ zZ7Rk#ZQE7($!i;ESx^#dMj|@`9Cd?LNo)OJe70)I70fnxxdD=|)ozXLjTU%Iwm=jb z%ZK1K>gM?1l5Q-4~qjjsYouJo9}CMT1Y zMo9-U$*HBRjZE=wuvwb{q`6H=tr+qd=JpVPUj`)9Mu7ZkNUo3t`tmpD71&IWm9`Al zGJ)rR0Q++eG8aDIewp@uQce`^Roe+_wE(cL^*kpVTd|yxli(18lLZKngYitzlt#_+ z0;bJx#K`K5Qt#R>gMj0!PtPmx&D7c+MEl=;uvPkOmZNfjU|+W>xe();Q^sQ>;L4UA zc%OZEi}<4*QYgj%(vJvPvL_9B0rxTW?;L>M6Oh7q1)8&DTWTY)`r!BU%htUuI;pZG zIGl5_h8eumhdvfH`V8Bp7Mr{PMq(D4z1mcOPP;928kZU*r+#uR<7_}~Q^8P)?z)UK zDttCWlCuUnjbkXu6_*4^W%njbG4JfJ{Nze=caQY#J4jqYU7I|BooW zqs?NG8>~gW!B`R?YXJb(4|cLP9l(>?#HQ2=Sk0tDBRHB2)lSJoT6Glenk zq7kH%mr^sSSte(Vn9@W{Tgx~!5MVzFnIX54am)50PDS7boI{ zE_n)4Xk`Q+j&}KF|5zKC!eo&<<237{gDf|p4{tXdn?;ngg#RTRLt7N*eE(RJ>_xuU zo>{0!;;`*wC~Fk`A&RkS5aVXR=(l-*ud{U=M+DH1oXyCeJuzRAV$#REnoGAPIUNZB zyU(j}qJ}e!Gf`p~kG1h`%kpGxQ!H{gYAS{de(RQ0NsAbZsN;UIhVFJ{A7h2%*9;_! zs)BYJ+bg={7$X=od!xpxl?|r1>MQp75TJF>l+c>wrgAKu4HxK~!#(Y@XM?aFymtzq znCFem?E4HZtc}moX7tf&kX&SPex;Lvav+`2PV}V#wL4`<5}AaYm^!74t6Js?d3oy0 z^Kz;$AW8Jumk=1_^vv0+JnkRVIzK;K%_R4iDRG3;D?zR){m#lRJU(^NzX(S8x(zID zZ`%%x$?KI`Da-VdBVSYkh<|3zJO_;LE%xjlTs(ibHI|f17~ju+SyiiCgp+0A4X28% zL&rWa$(`|?8>>Ka+~>YnHAwDBoREtM7+ht1E>dn~BZhsdf_I{$G??TYX~0I_Mhi4r zpwR+vffi_MZIOHLJcMbm?QemzX`Di%1sW~TXo2b$s4cf*$xKz}+#J8iB#*VVZ5loD zY!)AvtC{_Qjqi$z#p7#F+4h6&<~L+IiU1_YG9`dbqtuE!cwyxLm$4G;MF7tM2OJ{+ zmz^-c0+dOu+pZcKw3HN~mTxU%@Xip0o^7KBaK>h%>a>k6Df@8bxwt&{*sNlvoBLCn zoZ$IaK661%zOgK(1iE5%8vu#9UkUXp+4Um$N6D2g07t{y9j2(t1JHM)X0gc~hX+}{ zWRzD9GTQZev?(keAn=0X(%)JD%FywX7-DC9!S0E1Qzg z&7Fo5&x3J_Krv@KyaK@JA187E${nzuI}n_G=tF(-;h!0met_UQ1_8HKKyvl z`bQC9ojZVM5a(vH$*n!XP6k-GWoJ}|wxKTrJXhDl0RD77E}OwPuL8+!$rUTOr33yF zXVoBig-LG6FeoND&Yia2;GG2IFF}$<8y!_61{;{$-+XBan(1qgpSldP$PYies95k! z^K89^TbtpxEtmxN7MCI|R*HM^P>NeA?(SCHAwXMNoZt?{-Q9}@cej?{MS}+!elz#X zJ#+5-0eO<|ee&*a?X}ll%Y1d!&${jY;AG=X*VP985a-jbK78gWFy?;ssX;$*U*ag9 z7MUypciOx5T3$YDeboWi@fs9)7MDwOXf~eM*DS&Z$R!IOzG>$34*yux#nj2kYEMOYX1u^E*29e^g}yHVrc zh2!6@|I*-QtHP$sB4V>)C7#nBe;d}qAKptrqQB;vLT>g%^t3`9{xD|PPp2sVx>nA~ zmUvY7K;y^#`svcy%5sWE`|_%+rtgo7Lm$uN;Gnbc`ddY*$oB?YG`lYQN_LjnwPD1K zGWeSnjkp7a^yumtv(0M&`b^UT+$&$n$A5>2Lu`0S3cOid$RJryMnK2|Mm8y<%QY_O z@E$l5-21@l@Wr7FsxR4gYLpLG5L{Bz0U->ukIpeO8c-N@vpX$6=k8?~{i=Go& zn`SWF0xBzeT3(ku(#)xlik0eIaU7PRR;|V)l*tm@BZ-AepnGX1(JSJ3b>mW$RoQTm z*D4?oDDWS|@E;GM=Oys})spb(neaA1^qmT@2uYsxkF3Ue>CYzjy_2e|Vx3ZYT3Cec zV)n0G_g_}MZ+sXtFLd)<)|2sU>RJP8;YF>*D7eL8sfm z`Ti_Sb*AWQNTVpbewRx?VsJOTI?#BPw~otD62)c^Us67U18`Cr=%9TNx42-=jKM_$ z?)0eH6oN%i<4Ud#Mlcn$!LA8&wtLWc`M`aR((u%ZkslrcBSTq%T+|h=Gr6Rx&Qu!d z5$}s|z7V!Y`n|P(Pl`Dk1>t{OljuYPYWfj3t70h8L|QxI&hInFuL@lC)_2~_4BC4sJdD^f69myL>Xqdd4%G2C4AVD**M^{r5Z$qaSs_ew`QacI>ys z*M^QQUpEI6^rq&~H@!9tx%oL>raZgL^JbdG_`81!JQp0?4hwWif92~}R1b_Tbh-E2 zAi#(Di3{ks$Uqj@GXX#AUA~Xp#^X;`&0J2CVamI;^{z7L3mS)zioMHouRlsaAb7dK@~x)gyLmjo2dO0U3!%HZVtd>6 z#jofuH-aVWRQ#vQMBIok9Cbe{01q9X-+4kO^-aqf3ZYyms zO3~*9;^e#6pXp_NtDDvMXHs(DsxOs3ObOO6>8`zsevW_*>zv}mTvkmZ)W4(CnTHw( zg+ISwM(CS&BCL&*ZY@_2$z!@2MBcE7c16U#DH&wijjZiw$nWJ@{`uW2|8vCG+SqU@ zilo~=IbvT|2GPCu=>Xuz`_$xB&E|JqOQgG2)Ug@g2;WQa`e`Mq#_H`ux3ch^*6;&P zo3x|#Id&Pn?r=0ev(kiYOwt#{@mDDu>T`{jMM8X$-kMN%ji-8mv$d$JW5pBEB|h!q ziG%ElwNX&-?U5dDYplKEmi2$D3;(-R2TjSxORx9fUHir4;>jFp%9$f+21@z%N$JuRN(ztHdw? zCEIB3x6hmXy~mouqnPWAH>ZU(F66XL%j6fw3Sd{zGpV<&p}8qB9PqpV5V#aj>^q=M zI2eO+F)6eE{RbZ59v{cq7&Z$@Q9wGtX4O?ceHDMe28aAA5D>PU5(C(liR9sYPQ~p1 z{RqzfC%Ra;*s*h2@Lt%XcCC%yW*z8N`YPP}wklnv? zyQvTkB(LnNDC>~=!A%HK)i@2RttvD@m}N==QaeJOz%Nb$<5X{f%vg1N5Nx zW(M9~(bP$p6q#U}H*c}2p;c`Wc);j6VnrVY{&_IzmOr=!a~ab1*)hp(97YflXB_11Gbd`wKt zVVGUy*~}&?m|21%%QzDCD?a%8x+`k}@Fp;%d%B*`TOq$GA~`>m@Jqb#cV48ds{vU%ojO=2NWY7~tOO6MA5fTl44lxIeAvOL-5)>o*SwzU8OPczTR5SQU3Hc| zrI_d4Y(e`iw$ZOhm|%k+>__N*=Wxy#tR2OmJYxoD8Y)Tt9pB<^zy)9DJu zTr3Wtn;_t3YaRQ=;-8{~_nyNeshi`7Cb5ZcCI3e0LaT!pz^;4wS!2`w$bP2qbTI+;ZH9(WfL16?P~vu9k~beB zDYn%T9U-uupPx%%+b~0mV>U&=YV(J$OK5@zHs_ql{hVGWRz0}&{Uglz5tiWN5H|Pj z%+2%8DHU!+>ona!4RNNp?WFAp#}uu9%Uv2D_2&g*r2Rlo+hq4*`)%<;xv=5BVJL|9H9Naqr%F$qlrF$=JXvolz~O31e5KT4Q1{w8 z5`2ESO*AKLQq<-D5NV-SSc>$#15`g<^s9@Il+c8(1-b=3OYR7`U=uBVv7A6yIvnvY zj{eS1iZud1sJ-xBi{+K-yOr5rXB*;nF@uul?iuoxY#1e!v8yS)z3aTdy(_50f`NMz zr=8-X7A$hVDU?4dw?Ta({ipG5ven~=@sg+KEV7AO!IObTPBhk=?hO1lxDVwg1|b0h zjW^PVE`994c@5BK&t6aN-bJ*Mjxh~tKj}b@+nFYRtZnT@f(iHh)?X%G;?F&#d^YQq zp=&`?yfjBf@e!izFyxvNp?@dIW9e`a0~PoseaiK*Z z$FYBfU=m|9tY`C^>Ay&LCejL8#!hX9cF0^=fnI&Iwjp8T6faO)*X*E>BSd>UmY9c~ z-mczF(#S}RZgpkP4Y|w9O1de}x}<=8x%r8~Ow8aT17q54XNZFg%A)>q#Gd8w4d``7 zNm~*hN%iyU!MR^!o3~0QA5x(NWL}ZPlqea>eJMx!!dtI1?)=3lD8b8-5T>HK!=`W6 z&XTS`DO54a)G9@%crV{>QovbX0cPLj;gF|$YNM3w=x1SPrsaME)TP8+XVF76EE8#3 zCv~U{_s#;N>EH79!u)pLmnkYwQf|Uda+NneD6tP`p~SjtT~+e<_U@l1Qu=aI=CG18 zuy&`ys4mmNu8g2hz-mtW)mzv1#Up^b-gYlsSER06a9wy-jd}b@QoEbx!#9&iUP7HK zmK+H118w|w*O;ad$fNf`4TZMXE|tKC$sn-J#vVSksAj<&j?CQZ_f+&_)4ZNtAg zqdR!Etl?sgdVc4)O>~>;XnycyRW^@5jH~qC+z5wyIuVCnvzL9;o8-e9vm-9!{3OxX zkf`>ipC;bXksoDH_4{X5wJu4O;v#0p9_0Q;n-f#u)-PScAK5Rz1c18@BySc&K>;Ui zEmqhEzpgta0K1{fp+Q&>YE3GD>x}P{P6^9ECV2N{MX4;8X2;Wv0c%X|m?#{9ODx^ws!vOS=&D)Kt=-Q$VGc5 zTr?`zI*S|ebloK=_>2!q=M+?~WSTVD6aD{>${{CuYPbnx(4IE?3b1Ghm(PO0k$yIV z(UGZMYeCtD#|MyZZ#*B{j;<#|f}b8V!E$FjXD7D7KcXaE#n^!q&h?i4Ul2=hpd8J^ zm!?)YP5dh^(5JCfRWS7#EL8gIt;!nn&PF%tfr zXP%UhEPdR)%O$u>zYd7W|C}KcN{O+aN(3fo)3u4WuUz|G$sS?5gImo<0&-;myAph9 zQaUH!TFP9o8Ht$*xu4|-*fs(Z#bDBPA^ylU1|$APeu13&ns8a#LmLmY-c`elQ9j}4 z`2LH=2bR2~qpXI7(l%9Vs47eY!Vi!g48cCfM}o zQ|#alcGNjwx>aH`@zGUt!}VnfN)rw5$i52ilJ6{fXkiESQ8@Hv-Z&|K zN(NqdjYFG}a=-fJdAW>Cr@U~NKmV5(08$8mDcao_<`}t(O)6LU(XEFR%4!;GQUuRT zlIEax&e!1b-(cPG4g>2(jM^Hv`Fo9`z<|H9kpmgG%Ry>(I>=am!aCnV*5&8v!gw~(j2+HZwm;ffDDP*$>hYF(A@D;gpk3mWza}O! zTkNd^f5q(}tb(uXUwHSkL%?;~x)gK6@GrAMw0RLJYmu^DzDwl7=yJ>6UF5Moe=n_%iS=a(+EhC}(S2fwDB-=mlm82`!hC`pSvO zk{%wVOO~avz2@)}Vw+-Q)zjn13#V8l@3zE_^2>hplg?7w)xe2vh>oUK%{#zk=2U<| z%3^i7^8u>+1P;T-=bc@$-QS1Bk1CgOwBD+GUStg?zws)Xv8xOn2`=qavZmY0;K~X% z!7&Ou7OQ??Jf^5H5&n(2RO5T$!+L?+Vz0S(qx@D(db0hPN5Ch=iP)8ZR7}5EJn0invp^ zBypt-xutZT)|XjH+}%nTAiQN;AXJOrIS|zhfNW{AG>0E+qo2vCG2ibg3wjpTb z3sZVl>`1eQhtzC&LfUCD{p6O18K^; zShZc9>yFo8!O+hWcvF1~zZSE*L;7B8a_8t2is&bDovJoMjE9)6uDjMJ_ULx;!yrVxJu$QKmXt@5dtIwa!&usj36Jr@4L4Axn@F??k3Sg2*WHMI?ewnl=_kL~yT8+!e2Kn&N<#O*A474nFg44D z6`v&8R`%J6aIxa!lBmc)`=DG{L3*u{YgXaid$6%@A8yQ3Jq*hxhWSY8(7SB#qLF!@ zcCm|~Y^2>7>n(Q(u#A&acfZ6>Vp1PXdOt6Za;2h`Aj?t!9oi(XwH2!s#}_8``9-bn z9QpU)j&krslS);Z&sY^X_#&lyGyC>25Fscr@9f(SK1dsuVA3rLLVkmNU!H4(Qa`@V z;n})mPx-Ba{oPBWhidT5E(M0bO`2po)CD(lHcaPRrx5K;@CB(qY}M~DY9YxxCI&E` zt#~n9or~FbJ1nv*jkJC%9j@u3Kv}EZb(GGzv*jxb%Wd&f=i0tm3>KuTZT8O1*^JzO zK$aD@rb*sNAz%ZmJU#D5I~OK6(qCUvSSKlq>xK^(+XyMiDN&3wzt=OrP#cSbz*Eim zv=M>X4mpbWDSpLB$i5TadMH?NZb(Yhg8*4YgqGEn3sZ1pB|9r)Lv3?{kD8?jY z!#B;_S+J$#s-L`Lq3ZgLOsbI6d*_AlXL_~Mo+|sIiTJL&uCiWqw=$l0ec=aw&Gv4e zTg=6u!mhl%t!rvw3;)#t{WoCatWIFj|3(#ufs;6)o_C5qXiT*vuVYN~F8Df)_NhlT zJtp8JY_v$??j<5G`?1;QTIxF8K>!qUTV%wA(pV|2K^52M^ zj)0H%D<&WNp+HoBAQu|>;4lPk);hqbEjNuB+dw8|v`r!pc~AfH&k>ZR{VOSwfUm zK&r;s-k)<01v}r{T54HUsIW5@ymYkGdNY|40LiI_1xRgVif=8WN{O&OTP+XXh%!!s zxaM`~S_s9cz=f1dErFZXQhcNE~D)sLMB`-eh z@ORyB?=p`z@$F>f>Z<%!vA%H0kGdz}JSOK+l!J~w+~y~K*Y!8h+kHA~1%9|CGBx1bs9MiPj44PEL{(}s!qJOM*MTBTA7sMnkc*L z23T1vk`WXHpSD}vzk$V=xsI`$JW?KM#ZAy9ycX#Swj;qC1N)4;{>}ZXWuo}x6YgvY zet&d-$sR`v70n$$?IXQS-pYG$vItdJ#`$51tR$A-2FIg`7@Y{IaYT{lm|Pl{?DD4Wb9) z(AIkAqz@_cI2D8~)L7|A!YZ7U3=^Z?n>;L?-aBapC!4uI2Zm7du zzdoo;H2#xEKkEx{mAU+wsh!?FlM?v+x0`SaqH6(}710VhHf3?!md*Cvg_0|&Y}BoB zl1H!_-@e0BhHI4)Fx>tRYUMu-(jQ=%Jcoc!E@M@G=F1;vhEq39ru4U&lNA<9j?>Wr z9Z94*@X(0LPCA#O2mZgF9j5x-api*vQhsR`830ytI-iX%r{;JpiSy9_7XSnSR4>zI zujb36ox1(Bm5u;PUB77o5nV`grxr<(itN5YaY2TzP1jWty!nj2b{b9?W*Q2@qW$C- zhY3Xe@W#ST9B}3cd@xVl08wT3z`s%N;>ND3u47>mR8t6spxRplrgr)ZWA3Ge$o*{z zzMM{clRoC{smKV&Xuh?)b)c96F=hZ)Z+6=r{MPVa*>DdYLpkhh+|;EJM^#EVHW^-R zY&dD~pIlb#1uqzMzMnK}vziknH9E-6{lP7C(UJ$vbUH)G}Q3EQt01ZC4hKmr&yZSvYD_#@_{3M72@xbB*!u-&Lt$M8`k%g6xI!RyJ zqqPz#Vl(iKOz(T6v#ua@q=sbk;=&Ut%H^2g`{`)`QlHi$myO0U#lC9v`aZTrcC6Pa zxhe4mrjc2vgr3u~fbuo+MpxFm^`Q`W=F8%PlL zP{Bz?XTr(3S3Izhp$cMCsC~3XfD;RdDsSoJmR5GsZZf3>Y^Y2w%E~!;!IGPo^9NK77wEG$O0W^^iLOsvW z^4XxRU-NZG5TUDho^Ml=Po4%}cF;Q$X*ZCir#aNQM42d}L0)97-(@GRbAzoi zLfO@uk3owcvDZO-5Z5MbGLF>Jc9IUX;fQ=|LW=s^TT3kT0A*R_Fg1}8Q7foU7xbr` z#rVOaf^->5tbWSoDW-kh61c4Zf}R`{=l?g-_P-jYVU)Pl@Dw;`n8*+54d~c-_E`qY zMZpKC7A-EU>&(iYeQk}BlxYq&eq$aHy+^&7eRJB+4U?b6GC81n&IAQ{_+J%SLL+)w z;(MT?*MDn$wp)J0MR{JKOjFXvj~}U^w|UN(M_%fn*IT@GHwKt@u43n4ID7_(;i5Bw zOA1iJ$d<+#stBOHzl{B!p#evP>Cs5kSIxOK9j0dT)bZ0Nl?j2Y*%>>i-)@`lRV!og zPTWdA6KxcL?P)b*nYXm>=ho{H;ulM*$97BbRxj(D=Dqzmb#dm{Q35`l77{cJgN;9{ zLbid~fx@xAf7+B+_b-Vq+}$AhIs6`@j3<%2FqXM+j1o0_*Ka1|?{l4^AS zaPg0t@E!(Qq3Jh}_By7YMV1+U=5=hR!$C4x{M6^f6>KHE{0&~pZJe60qW$q zHLMJ9jAY`btkq$D?x8BfA37^?6&F}+^-5O2rXgh^^T@(}NC(nzZ6{B{W60fCe6vOL ztbhf9@w97+I9Go;lWsy8d8Y!j?965Eq%(A(y@;mGei~unMSR0T`g_Yn#7yY4FltA0 zPhR!}#uII2JxaIPO2|O0^;p^N6UmQ77ju-qeR3+mhSE3va;Xx|cv)Yg^z3djCXKLvx{1792zsi&JwGwvRZOE7e~?Q~ zOw3iW_+W+R$$9Le8HZ4et)D@`L>szevw~~skto?(p!zI}Nrc|Au77H5{$3F4QoN-Zrto_OEVk`CKfxbP@7d0SnB(7_;9U% zQ8!4k-WUD-W0XWH-_uTzsCj283GqldI#z22yV2*aGKCK%*v9c0W}~?z!F@LKGj}vc z4tSiJakQ}9PvuvF9h_o)hC;(6n2~o0_OuAYvh`F{(%SVpwgZu#cwd<}<3znL<6mW- zkyZK+`X8>&%>AVR!mndC%!T!Z88{i1S|6uQQ}VMJMm{SCmUKYo7j~RACK}W*xM8;H zF#{w&YiC+HwJ?cGgLOI%N3fd4a42_IM$8$%S^ofnG))hjlR1uU;EQVtrwh%C4H!7_hhi;VL?|fU>21 z=%JfJ8k~xfhRAmTFn{87M`in1uPbrMsi85U?vao2J_7kO^0qXKq)L^6&PQyJ0r{U9 z&wuN1eqyY5(txoUI8#aBTCPh+5+UoI>`!GIJV~SU9VWs0C!RpDo0Bmfov-3@(GChQ z^N}76G57kpgScK5l#k2JtsE~{Kz_E*VyQ6*{7!NQm+AI9Mqum1<5hS*wo#y6XK5&| zjqER>?k;LP6R?NKKzU--3-Er*AIM)wYcv*3vYtF_qq4jWpU!7^X7<~+{c$f#lPQ$< zLYpt?)@|$Bc%mBlec|scwl(~hzTN!1U%XC zplhkChd|ri4>NTQ&#RpQ#|fx!Zo4@J5k~@5O245%S)`?1d8(% zyZ7if1dG2m0^MTI$-8x;+QR{_213kpFbeFRq-BG=9YoS!Rdbp^%c!}zphC9k94!N& z0nw?#a|ZN&la^mAfphVUl6Gifu6HS!NE6gUUmamgc_G$HX?>kIuuJQ@F zw0&p_eEh&Rj#pr}OV}|fJ1XUGWqVgaOl5nu2~!}VHua_b0s-I+%FOM$_)3+2ed{-C zJ4>^6AAA_ePR$W|d|f}Wa&pO!5ABV)uU@oD!?m8DRzw}-)}T))D9H|tH~36px8e{r ztB@0+!lIRy-`dzUND)+_Fnc&!;cxr-F<1?_9gQ`?dr_ce-~+0wz{Y)b^__0>?6M%d zWe<2YwycYpEkwsYaq~>y^E^FfXY+Smsv7g75db)BM>jB0M305r94lyAm`PBjnsqm; zdGj+La>`du8XUSzGKHYcV?YKU(^vl)#9a?Fl9UyDZ6x9cIxPAj(2 zx!xBSa`X41KkX!rMyM@Lf4`|G-g7RqrQM9?c5ky0H7nfYEpy*bJ2A%K=c9>a&<501 z{mgyzVLXhaYW=HiT{Y;X-KFo@<~~n9{6 z&R5MiNhg51?MAez@ObEbqe~=G#h9LxZ);tvI&N$EWMTw@HR0*gC%~lGgBPZ z0DD;PeeO-U!1Fv2h&2F*@^%kg;;>rrTg26MMQYW9*)ZnYYA+i}%P&H3+IV37A5g*W zq!JrP8xu*pnCbp~=E81)*;+)axfVaEU;p^}{hi*F4=|J?#U#F4xzt%s{3~*8HXLLP zaAPp3@6OEnh5`hJ=p1hpwfefRQb@E_g{R(q+h{e*x~XHaU62jjS)yN>8X_msa$>$gB@#y+bLT0Hp0JRcKMwq8EH@Nn0ugQcq#SU?3eLl9B7|o^|8RNgkH~Z{(#9O ziO>cQaDtn32L-9|5~+3lNp0@S9i<%)I)2p%*$}EAh550{4BKrD*Wn0+L0G9EaJ8q> zW9^O#?B5sC$~Cok>7@(0M1Yo*RtD8*Dsw;eiK_QRV+4Q_LE@Xm7)QJHluNKHTrKT) zBG?x-spRO-z)Gnnov=xun7!5m`yY{D-wY+vf@I=iArD)_&BJ*CLswb`&cmC0$R1@0 zNtEKn5ck2+q>jWS*?7QRXGx|YDl<rr!`d8S2EXkz29r)h1bujN-(NNtOTrFRevw-Y>98uOK{_dSs& zf2va8aa0IGnXQ8;2e$iS=J@9kU*G{YdON~K+3A`8oz>g?Ge^mrtJQL~JpX(t0y(>U z9#4SJ#p5+49MnrZ&?b(9@?{aaJ^B2B({tC33IRNz*b%6Kk5S~+QlgN)AWT@|X_l;< zSn<;A4@RYr<&K`e*Co>TnRb#hq*Li!o%Vmd(^^+M-evPVS`hPSG>f;62^_E(lu&n@ z6sL_nEw$zF{3+28v-=2@LoXlDOPSj+iI!LK;V+o99!&tBQwsPtRGx+&a86|KP)$7H zT*`AU<|1aq4iu+AWjCpJDmW-3Q8dqL(%ndl^#8A%m*TW8xyJY<1yTvkCy{0#g(&)C;{E7eHT&G7i`_FzG%jv(qF4 z|~{gyr&qrrc=>iXv6JBQoK}+ zoptTtj4f!RP>IP8UVjv>aNJK(P92aJa4P^zF>4eP^2@vc>M(mUxQa+g8P2|4s-aYN zul&X#jHOT3!B7U#^x0<;Ye$pbStCQ{_i+f!Hq@&U>TY1aXW(V3BY7)vv4I2+Z z&7!e+L1RNg9$=3b)CCms;$dz2i~_|0 zkln0InxLE;^pRiT0Yp0A{FYELiO7*Bg)j3~*O9KV_Ovy#y#SO|DMnkh;>f8XLFEU*3IH`GZ9Rt=ormeiHAS)R8O%JfHz{N<`SRz4qj#FIZ58j=sCC;^Sg*P?X z;!@p3yXI1)V2;&gr@O#Ch;4_MnL!(3={yp6-s$Nq5bQ5~0b`Ucin1^jdKB|QRln~>;l@vO%xl{g+jYbEFNq=0$R zDWbAhTA2vY7|0gIkr(d?#)F=lY@~q4lETkYZ)@*&mzA;Qo&C z#)*=y>X2aTx?jlq_@^&p8Bfs`yrNs;ZI^+83)kTVhie>1jSlBk6X@Wk>-7d|^$F_+ zbOzaEQD3jg?<8;VN`qz&r9^ikkHs{tAGhx+{sa_UbWZp@WbPtYmyX;Ro1VWv>{(CP zCI*{&9y4@Rwc<>QPBIdBF2OT%3k5gt?uXj}wq!S-L*MzvIBqM;OaN2)*AY|-&B)^z zd>uh}X|PN1h`ym!vlw8jgxc}kIXh^;i8km`9`Ub0ea`cS767XnpTmF&3Dmr>j=RQg zrY+S15nnzC4CN9Zk2FdVPg)wmoe@V9lxTb*I;nkX5${A3hYvKwxz`XdHw$v z2f6?c=KI-smq$ATh+(+!uVZJQHyM~QXZph^-~a0UzsfF_)3M|?VEo_Q-&tnLX^|`h zKXznAKcima?rbAeM^g*Fw1`{cQ?t%8ye{|3T9I85=%C%N{ta{sU}~`xiTL@mhtMMp z=HHYy41)emf{unwFmWDcHRA&|!vW`b+&h2KBRb(VVq1f2UOIR6y%e?R$~4V?ui6~D z`C^ta<&LvYhL5~Pptol$eN1FFKWaLTDm~Lrr{>VW5%>*;wseNR)(qcPV{w5{8fos~ zOklW3_t-(eVH+<%73%+6>v2()Ized&jE6$`Z-D#tZ5-ZY=8~innsRvk@Qo|$eIy2h z?5|Mkw)7=|-$|%%p+NX98s<&p0ZXtkLqgPZ_atQ&-b*JFv#e`$bTH*@sRa04_n|kL z$8{Sdc5ai=_Hf1ywo$ljdE5&T;CFquF3(VF896Y<^gx``62F9+#SKfWzgL^=xG`xi z>t*u3SS%#{ah;N{mF#aKqL_-qc3hltoD(D(c;m)?5( zaaZf-1lLG;`1?`QD#TSXWW;j@@gT)di5w+ai@{GhCLqwzR$M$kAR1+)CRk8dnyi|9 zuKJ%Ww}U5`oS(U{cb220{#82g9&(U|(^j<)eN!dRvla`?+!PDYtd%4Y=E&V2RDItt=8RVb&( zORhjYz6Ic(Gw+e25A^lX0nW)&Uk-&B`Itu;<;H}#kTd>Xhmr~Iv)WQU@}{NQ&@#A} zHw{D3hby;${}NAATY4|Gq{1d$WaabIhh+cg?uvud`-8~ak&P}>ik?c`i8O(LlbEQK zu3LiW)UC9Izg}vy8Q(OP^4H}d1)A{@jVamuK|*;R^rf$ zZZa+|M}c955VzBLJde&QFcGD^`x?s`!?!N624+^!`uGFS; zKU+D60s>ix7GZxegcjQSM)o^%83cPP#N^=sfwNN$qVf`(b#%2?Gco8GuxNC0NfuCX zwUo?3(uzLF&OM}gDDMu~T(%|Zr?weA90vgyx^&~~)m1Bh`Q-5#u*;-oZfliF9|rzl zK6bFfzN0)Dox%RQxS8P(?oO3Unqb<6(TvD@R%S7^EI0Tdu<9P~yArZBB~fsoUaq3| zF}3&40Tpun^{MN$x3B#?9aZfUPFlye9j<)Mq6Gsq|L@KG|6tYkFDhoEx`F zCJ>|UDM&6Y8m-*)kH)Kmea1uWWl%R3?{aX9|8=|<2;6%xDEIrOQxz{&_2I%)T^mC` zV30`wH#=91-lrHT`W8Z|GM21S(bTypSD+ zg-^2hDWo-X++|Xa$3|nlFHd#ZotNP3X1%N}xfq@((1mydmc-hf<+0G|b8)jW0D~-z zReH%r(QHzAvt@!W?_ftIo`$2taggHIzH6V?M{z}k#E+X(?1 zEg(LyJNx16rP~t*5_kEAyr-l_e|a=4$h+@;R6^KPoVmNbNnLU__D`gfMEAK-yjN*6 zlr`M6;)x6AD>kO*=t=pqrw*2@7tgvBs#+Sq^8nt9E@^~7lBV&iwt~5kX@ZWbrs3~s zUo&FSkuyQuj?vK{0h~YIO?-23U3}zqlUBzsNaHi73bE`s&#sFE?{Paj^!zo_MDFr& z#J%?Y5OB3$HP-x>DGxA2P2iE5jm&o9Ch*UhNVpvT=5jo7Uzz;nv2L2_njhBE#2+K= z#5#{vZ!?J^WjolqT(f^Ebf2*h^xU`HFOiPNh=ysQ^_jxWMK?yKUAl;e>(Tu3f=uJH z$%Sc-CYeXxNL)eAC1)#}C6oQjoy+x-EV9Y^m^hsipU16K2EtE;Z(?<9gA5qgfh@FeJ)0{W4$#VLG(X)NkRhuCT|l(`%%{WMm}^3Yqj*hfmD*~%z&DrS4r`J%!SSYP$d)VkR})$%dD z{9v2xlH|dJP!3%#nXo!#B!Z=>^yKbLSmfFA#5f~X6kVbo)OIo58n`MA@G(kD#fVm1 z&+!^tUEJ!>LY++#^Za+G1yhnYb}wsYk^f|<2q0<{rZ?QLi?_0uS-;RZKHSFRZ7J@v zR^xTJWoIlZ%YQTP=%VY2QB;U~-L)Njt_-l@TF6HUuTW;?-Wm~9p2;1c3z8nmWi3-% zqbm57nvu<~Y~MqmqObNZdsuHOM63FTX)wpL4yS=?DrQILEFj=W#DM$6#-6()kC8XP zaxMYw>OCvJV_Iqynz<9(aM1ykp`G~;dg5IHB^Q==Z3Vj>(4|jpAF9E@Xa%M2nyJuZ zpR45z-M#2Lm!cn@CyyF!K0B-f8Gho+1mjcrqhAoLQOJk+Qi~vg1;@|wUN1$5oUfHH zw{GVO#G+E!f!8FbglkM%EJ;e^8cJ39FRh$VCv2_E=0uS<{B9SOa^Jtz;vA?ZPuRzP zM?YOznzqnlJ+xZ8l^kpo)*zThv#YE7?6x%O#F0`+)i8~H+}zc$#X0nP{CvpjJSu+Uo_{~?N+Nj2nMjF-MkXUS_Oa;YxiKEkP!QElYRM?9pyNy<-KbZ-} z4Ngq_X<4B+38ZltJ@$`Dy0Ts-{-Ivm*+}B{ur00b^Eq<)aOwu+?JOL}edxuXdg&7! zYx)$Wsrk@)3DFruLF2YhA9awKC|pOFW*xrM&)Hz?&JXbzW8dxMb$0TR+my(BU(Zs( z)#7crSYCm*B*q|7&PT#Oaoe4CY$(oVKbq@x>>vf$=u0e)*%T>~HF}xW(%fK{bRbIR zCB9N&(9cJkFgr45(_U*#xNjxIe#$?OAN`!r$H*8_xp%hsAjEIhZVa%tG`cqNntKQf zPNwd*$xks`*UeZk)~K?^cjG zXGSJ3hHw&N7=1;+rKC0LaqBrq(>!o(5}8+&6IgKlMrW%q3xS3gQ9m$=_|s4vN;iD! zJ->>ySsyK#;8{rQ>NEqM+YpzSc*{Sp-inUmlfXvUvd@ehE=sh5ixl;(bT;?+xeK@n zgPq=Tsgu4viB$E;trFZdBgg{Tb%!$uVPXeII=e{XRP4X*b+?+N>gxUGDmud6cnUB`qwsvB2(>@;bdM|;NZ&$FNLeai=C;6W-&rxmFH*=|IRpLeBnA*SvrG7nui3t{7Z- zr2NWrPa^iN9}=~XAs<%ghsOuD9%>MG)D!73aTb7i0sc7DEN@d0ddlwb5|-4I-4y

    F+Ax5A*Sk2)AcUCxbUnw# zgfjj{V_s_uvjujz&>4>__Os%eZUxHM49*`L7WC+1g3GG?k)YKn)sbjn=p2J(aR$ENd6%c6AL9Np|+)m*e3O|4!$9mnZ z#gOagWT@Eo;i``bHz5|h@Q&${tKKI%%H~@BQ1>O0wpJL zyI1C~4i;7bv9|heOdB{0ZoK~3$M)5!a^}b|4!I>N^zqpUN6w?-k;W*X6)aYGsCaES zMoW!HcWMoJc%6{bTG!9{NmC*Mf*`BOna&7DSScU3=FEk0Zn@f%DdU|GZ^JRgGHVM*Z|2N5pgIWB>AtL$t5q(#ld zd4*wISf2(<&ceA2?`EI#>Sg#VUga7+%xg$#*tFv5%&Tyyb;${$d#Y~0j9AE_?aYs4 zgyMd42Oqg@8)qD{AwI;B=CuY|S(r`KjB|EYvv;hp#a*|(ddbf>T7TsJz{uZti<5RK&fvsP}w$7|aJ*u`o z&Nh@g+F5BkUTNoiOgs!Jl`EHlDNmt0J#{z%e^@svf>=C^*x@R1s`riDzaH(~JvSB) zd24)Cj*`|s2rldMqADsJv`=#!|476sNRRZ^aqpJ*oh7Qnxykz_i);?&XAEwqk05u0z4!uLR%eLf75$=+SSEp zE}UofJGsXgM@#s zUHq4x`P(AY6gi04MhBVv3vCGpr)1-b0Tnztfa)HX_4s4RWqIfINLQAR>xxoT=;DvM zZXXrM2(m?C>Eqc@&p}SUb2A-&kPs0dZQkj27_aSw+M)HlIpT-V2Vs}KN4E=H#P`i3G@=sg1=z-c6fnJJa3fUSkh8aeX`3?^6{XUM*-Wa5!Me z2ecmG0!*y1L&W4YjF0OT4fS?9B7PR`tztEP%?2_b+A2_z(Q01P)6f~1BVEs}QyztV?fu9{Gi7_l1H(U*iNBXXhcR?=lOTSJCrt0^&w%8$pMNDOPk(bTKC3l&z1P3YxOyiHioP zD}wl#X^cT=8l&bp6Z~Fx4z&V8T>tg*I^dmx*5)@?zTC{3sAn1ylITcxoesS=E(qgUfz4nKp_t@mGiR*_$(b%!bKChcNgWc=bBWetanyhw(>)$ zjKN~-aj~a`S9)9ePa>@2^e;7UfRFd~?*)Bh!ph?h?zKvxA^Jh8rCq<(AYhDqoRBw} zP@>k2;kUX9`H+!?dr-rIhK`D9@ki5E$D9x;zyz|6s(eDnZ949pYjA>|Z_rFlpv?-? zRm&CEoPuRL=a4a#9@-?Qy{{m)-nc>}I`-s&H>&#>;`^l7!XW?r5T=2V)QV6Cyq@^r zX=efIwe~D8PybhS!uH59>t5C9e;78z5Wv9-&) zeM+XoP2bb=KI+-^K3GsCapVzW&|#`~&Wu)+2`x~k;i#*ZPY^vBKG7&Q;34m9py_gO zzI|`Esvi2ph0%kWoAOy!rRj_RD}>Dih&mtK`>Olq`6K-zjhoWskq!lSS6&OUEou}> zqCtwupb3T#U(@7yj}R@+DVe+yzQ64Z-^`KA%tO762BVq;7;_w3tkMCW-R*vnDWuJ% zqwGuM@36h1c?A{IE;dD&axLcg`^*DLH(&*6?b!$}pEK_)FI&PyR^0Vqu8deLw5_{a zDWjBi2Nt(mzmy%}ayW#^#2iA%FGMTU#$uO$*e~u9zd;SymUAtf-af#i@R{ZvYjc9G zS)ur3o?a&KN*b3|H=O-Aqxq6BSVZz-!QgtGF!QaSNSSyNG@+y23`lf#v zxG^w;6u~MN&-vH|IElwMIjr)oh6cYhM=A9WVdq)9s|Sjt&Y}D1ekB^6B?&IrIz!)C>O$}~hW|M799uffc=c+qJozrawd5hM6C|K%g(0zAYg?{44<#dwQL3>R8 z!qck8PvV?D0hV4vdM<%QDVg{eWEp3-X|Ul@1s4u_@>n}=H(07;T6n);_OGJv^Ei+qnN3Kv1-nLlsM!mZR`V|1(KTe3I4- zli@s`3U&x#p@Tq)#K#D~_Ef9)omNj!uO0Dtq=9F@S^h@LSRqHgAHB%=b$&Swn&RJi zQ_b#0oB~HtyY}8}A5Oym5n1A)UlOpbC#aHzS-vUlX7jv~AVkaA%{NtM83}5XjVlh{|RuNdnw8$S z=q}}em_?kY3RpTIp_v~b+WyD7S7uIGQOHKJ5kx~Xltj=%1`z8C;yXDoTV><>{H&GPX zUH;7f!qy^WYazTv*;qrZkEs*IOS@eW@PqoX@@6D&<#{oHznFF~8rOw-M#Ra1AM*#% zLOQSCA}?m4Q}vdxB1je6-(mqG>k#C~eT6b$J(E*0Jr~p=%Vzu2vN-lI!ew%Go5cm& zGot#^+Tw}AlKW4rg-Uy#bZpg6uvd=DH;ZrGPli^jQX->%i<3ts8zMa!XYOgLc||F~ zr6N~^Qx8In`*k)97otJ|hEX5e(E-61a_FOr@3BGtJuWqVr7UD!6P5%AT03QYmC@scQtUJAa&+rw7^5IT$JT$=y+0@EO*I|oKK>Swu>aD9n~BAr)p5dw zr=^HYd*6EEEk1x-pr3!)GU{sGE)NpZ1B#_af%b*|{hZ%FxrYU?D7rE7yv{0Q$)f=3 zab{t{m>h_ z6M2y|g?PsQQ;YYX;yhCtYCs240IKa?fBPsPry226U|}^&$X6MFocfuDja_Dg{<*XA zgkp5r{vOA-e!1q4PjAH7xZh0M{+Yw`z8x6vh4h`1_qusIR&a#4%gAr%BG^}~z5^ht zji7ChvUUrRgnBDKNo}y-8PH2t52}hDk6OLO286iNGD~VR;U@c^nR`J7cl1oh)%Sq4ogBEotqzYT#4XXKS(@@_u9LKx6h@1A%%TXi# zgCGJ_F>@tKX|cCe=))salh>S>kfTDvXtb7rLu+OQp<%4bbZo(j+o)h25zqSF_Aicl z1m4EN&F=v362Itb2WJz@{l#v0o99IFa1mnXom&{ExNV1o|BMR06%Dx11&{+8UHFSC zm^UQx2hlX-A5I(|o)5dGI(lR+kP3ktb1nAbg^;+|POwD$JVNL@o6EgCKqw4RA7OPT zT39IRsf`&Ld*XiEWFu3f9;R*v$S zuRH?nz=u824r@e)d%yO06NCCNEioO?p*&KJByCp6&qCj@#Ai}b=IIpdvpXN$5Iaw3KB}zj8}~RLk1Qx$r3#6vj<^i z@J(9jp>PW$uU`k*_nT~gT@DCCf%sWF`K2F!MIov5Qzvlz<6BrNXpJ76m> zA)BnRG2`r6@>T}`4j*Mrj_Ley4hIHr!8Q1|xRo9&K?>YV^pAY>)Ku5$!zqkYH z4=^PJ^`tWiu;XhRV-B*?i2;6*A$*<$j8dv)&$ycK45VGQhofB|JO^JDKF&{3@}ng{ zCed%E_T6O^y9JE=Z>ZD+_QB4&iR(y703(2FF2XhezOAKHV5==rA%_wloRwr4vX`4a z)NNzA(qrzXH`Jj6CtK;JQz*nAV;71s!&Wxf0du!lf-NEP)M2#Fho|e#7Rgx1(GF@z z4a8!<_M{AVdGXE>PQk@RAGv}@iDq^wZAJK1sgMd>n|$B(IR zC5^-rAlI+-;tqecd_L1Y^ zgkw@~Lr#|Mu&$R;idbaTN?7Fy@=3R3MP9UV)AxI^QgV6hKIr*(;&}>X?ul!P!g0K% ztSr-KXk>4(-IKZAiw)m#HfBmc(rG(9-Qu13S;l0Sxj-m_ng&A}x?K>?2-kmjdWSrq zFlrhGzDIn+VF7F+FoHtcO}kB?IhR>N@3;V@$}a|<3C`cr88O08pWFYL3j@GnlIfe- zIt5PEOR|^ILwQpS_iH#7ON`8Wy@~1L6G5>-I@8ruHF^knHIlCzv4=TP%AAR@Stq3g zcPhcJk)4nKjH^cVJA~!h^t(Ep(q=K^XA{x09i&Qc3@?n#ZaJzJ1J0A75^jM!EjBSJp>| zv2(a^Xp-JCwkz=CV2kzNZ^998bS2g9!jHGrewoK;DAv6BI|p|S*`+({?6`6`Kw@35R{{5XjBRqJzsiSEhA_SoE( zTZGl%lg$LI?fOjS0!_s23zs@3YDe_n?o>B}^*)k9QUy-5ou$Al#8w1w^Qg_N7FkUj znHp^8rtIXtH%PK%gG?;OtGA|yQ&;`Ix5zh@xvR+6$N9&-M#_7BqVZPlcwvwsg`&~D zoW?#h`=|eF9(e1Ys?xy`-g|T}HtqQjC`BmA^uGu)C9le`{4PIIx7QIY-LK#OS0USf zDp??*A@ZDHd{_t4v=19jn}>>U{tNdoAP7M{&+1RpSac)G^!|9*-rodE$tzMXyBNA! z9wz+Ls37O@m)3so_4R*);s~~JdewF3P6jY?_ImUi+(Blnj$>I7+Qeq-#_&L5j%^y+ zVqONWlhCY%T|~59H@1}|V$DQkG}CYwI|NH{QaeT+Wa3940ah@~u@sVXo*3D#o3}!` zeO9n5juuG!bc`Wq{N@Pxw6hX0G>R2t{W`@wu@{_pHdwG;gMr#1?6r>h5%G8Z6ECBw z)<8Q{)muE|{nTbwnv*G!6na<}%+U<=E^wFKG+7ol0papPI*b7l0w_J+B8kwGM=t8hnY+gZ>Hhn7k|AVG=_pJAhsj}gjh}8b5oXBK%EM zyv*7O(!0;H*{&CEaCCFgMtqv-H4_0)JtOC?_7%BmP60-F|Mntwq?t|lC$k|JrZY$n zo)FvINQdbd(VW$DS?yrMp(zpvx6Nb*+J`HzIJb?FfjsB z>TfE-F@fg`HBQ0J0MGL>R z_*s)q)ymSiJE$l2)Oc@}QCPWeEsH6%Wb@H#y;5pyC{Z*)4*_o6buA0z$&g z?3epzUZLxR`nTf}a7uZ3q`3WTICOTM0?dhmLII-y?9Zsfa9VvhpH{c$?y_p71}(k+ zo4v<>$k^}B5MbNKKHl|KGWt$oVg(OFZ)2OmF#Wj@7jcj>8b`w)yst1*B;5PW?Mx>nbLadF18OlTF?OM@zw`b*4>1UpNepKGp@uA{pAa^*sqwTBK|* zWRm^itnqOpC|_A-JB4t=i8ymN6I`d%TYCirlD(V@(qcfi5@x&u+de|p-;Q8B``&y8 z9aEzZe7-T3xAi?(k>O>dHRr&d1sJ!1=17qBAuYlPCZ3}67Q$N;a>!>DaR5X9uW#<$ zb5d3MJV5m9ww(A|WU!pWFyy3^lh6&PF(hJs?#wm24U)p`#1~Sufzv$H?KtDWg--*U zfp{KZ6v0u|joI(ENx#m7a5AU)ZU0i0D60hL-60iniV-T{9KA!pUnU3aiNxA18fNb0ncXA>sf5kJ z$O9W7o6b3^VZvU#VcNbZ?6?@E-O}5JFW6xM$j^Y(yvx}1{A_mt>2lqu+lBjBhDlC^l0p zQYtbtJc1^%NOf*FF-nEPc;RxfFMpI=EcGV#;i$!$FZ(aBks8B!(uag8A1OM>*moJ{ zly9uN>i9>}8V)~@c|{8$)BmY0KfNOlrq(EOrzF(gt5 zDL)wRPLKf=EBM9B)#2kuQ*`aYkIC=Qx@WP1$>9Nye({ZH0*5Q5#E^TWu$wV2#>F90 z_XbE{A|>}ru?Lt&)T=<~oqUGs#=WaSI&#)hnyD~P-WfCWt@I0=wtr$aaGDb%0NUF3 zd~6@S`75EFvQU7aoRiPlIoW@aow@gBAmM5>f+Sp$h~8}hu46}`6@oG;HOKne@i{R4 z7~H}Yr`RA~hS0pKoQMKBWYj8q(2TBNy@8+wW}5%JzWyITK{?R|G`u5$)nE(|Ge}AV zP$0j2j5J*rn6AuOm`hVu$>m`WzU#&F&MA2MV{yPMA!0p*{N)1a`Bq<#Hj2vzoyF7t`Mu7%EybXh)S{t1NyVKEbcyykR5MU0*19lSWEd&I4$}4pv zvo0`Pyv2_rb~F-m8`z;5U|c$1>ViwnQ7UsE2LwS33i;HX9Amm#IdXtAc2i&fJ}e)axH#{QSH9*=q(BlWR(8Ha9|)CbwE&s#cNq|XEPZy9vBZ*I&|L=w@(uJmeI z=sNRxlZ)CyLdHnel!mj>H|_esOIgf=Eh2hmMHZqd9a19hT&XpcX zCiXsQWvor5i}(8JgZOTIjk!%0kf+N~I<8)i5?{dEMd*pO>2?b-E#JWrOmbdn#|9 zlDdr1O)m;t^MQMxc4Ur%YmKK%*qPwA>S{W@WA*k?IVbexy`&!9o_xP8jKYA|(`$tq zGC;2#TPKfkMJ1))Om$st5BTUwt`;Bgb?Dl`-=<=t;xEG^3rwugt&bCKuiSkx-+Y0o zz9a(yr4!{tr*mBOnF=?8+IficjeB`(dqCof4Xf6LWi6hW(4hUo#C~^$%@yR70Jh9< zgaxiC!z35-@j+wdS6O)YU$Ix;lVM_myPlLSGmi9Q@zB0uhXQ>BdY=t^_R*PfM{kVd zKgU?#TkY+4_2Iv6ss0~{Oi)zqYuZVshtJE$Swsjr-^Q$kNk5`tto-{P(i)qHLE?}pXc;d>r&aw7|g=Pdu4o4C4| zH<|AcoVM^NxqAyJ+pYodZC;BriSz~;0!|B9PwdyV#5zm zMnkE5j^=WdG8La$@NZD+I)w%* zu*I9NBU$Vy-qULLntoFNbq5u^VE&vUW?1r~uV z8VpcM(pt@kFo12u=?>X#jMoI)C6V$tJ>=f|`uoTGk8AMuSTWvVDpKC26M!LG5a9aW zi^F%82~KrDax=-p$lQ~YswG+)fB9j+SHz~R&)AAgsYC4}tn=b!h>H$bK3s zpDuLvA=l{fff#_z-SM?)WAAt0I=D@~4xM|sqsw|1T1Mga`l1mghX;H-*>em-xJi)o>G(Lx3+ysGPQeK-z1+RPstn-fW&XU}op|#)S<@uI`m0 z!=Nfje}=bKpxy~|inD4K>NeB`x8x+?Wxdjf*j^#_2A1@73s?*6gxYL7=n=omYSa zc37~ZHCT*Hgp*q+vEhl%S`6N`tda|4D`rg^9RL_KhGgB!If&WKaP9lsLHbTo`{M3R ze_ncNXg+xV-jpzm>H8dn#D-L!bJGeI8tNWmobyWKksuJ2&ItHiXd^auY|Pu(_PGDe zAcyldB0SmG6*#}{dN8Jm2&NDIIUc0FeNtjeAun)=5r0t54HYQ;cZm)fBC#7s~~_+4Hv^~7#Ku;;9W&zRWz-s)^t z^U92H0Ddw(8wJSP&2{vm^GkwZtyDtp0&a-{aIL&Q{HUEQ+54$$@2UG}=f-MjT+GQq zce^^}cV3ODZi2(`skq{gY}(FGA%pGN!Y@cT5?BmVW5b;+7$=KEv48vX*iVT5e|8kx z-Q!3MkE6J!Xz%(zTM@q{0%g;z#<9~6jBp&6V?V8zWgSJ?mR#R8({!dOtFCqFeGf2V z5|S)p(Alxx2YLYz41Gf?+x)wf6p`AB*oGGs>b}#R54bQ>NV-W?MBt9Byo=PEAn@r= zMlE=3#50xtS`48wDnFzTw?1+&IT)Ct;gb%wn7@H-e?JaTb>Ul~TcGW23^Ma3F8J;W z%w|+rB{PgUg;;V&Va=Y#hjU-L1EgU9)F4;k)zk=?K29o6-2+m|HyWIPo^G-ryH88Q zya0cKRmc)M_vtNBg`$cW9CwyT=>#8Ho7;!k5wai@S|J=8y;!UI#r>Oh;qV!kv1*G$ z1Sd7D480@qiPRfvV{RvC*b}22NymI`bo8hV_?oV90nXt5`n+qfJcX>hO7nf-?dV%Z zK18*?AlHQ!yYWqO1wmc4UWAzS1*bL0Ekcf_q-iGaDqIv80=83n$>85}Ts=Q;u}E!e zsp4AaETlY+>p(DmlE>Q}=19r?0BAZT({z?4jV(^7cw-h66Pt$KOGCjEULfLVRZMP_ zzOC1Nr7S4DY6ClK(t}MFFHa_~UJmSfqKF&EO{g;18i^5x?71>^;>)vTNEJ^=`;YH$ zug3h@FAY zD?9~srwe|v_+RNB)@mtHhZ(i6+a(rUUUTqnv#HTmLFNOx2TT0&bpuQ^L^(YMQqq1Z z3L#&|W+3dob%V=hk(2~#`SFpdPOOKB?ogJ-($fRJpkc8s#}mE&tMy{=mM&o-4T@uo zZ3@=IW=xUAYpf+0P+-HreE0KFn@wzfZn4&~0M^KwCx^u_uyJh{4$0t-(ihAW#ky_%=N zT2xh&B{LtyayFLtT6N-7l*RuqyS?r1Riu}~NcnuM6c5B8D5?TNzA%``nHqn7Ng=2g zHxvU)`b`5KR_3U=ZPNFde5(=048s$-|-wp1>s@oXNF7;qx zBP!>Z$j50D3O+G706o;7^qjjdZ^ed}y73{=P%n)gUGhYQ*d}S0%_P5Mrt>Cd^8P-D z_`|Ii*KZv~27Zz<@(c2@zp7=L~QlVvq#M#A^CZ-t>E-{i$YB(RM@l%FtK@2*bQJx+uCCm9%q=Gq>@utMO94MDHC+FBvScD^vc#5#!94uVIX-pRIGbg8k5WD zJw<{s=^Xk#>~D@vA%g*IA3PO#3-7jO)kUBH$)E4yr)2LK?i0FrrShHJ*gic9|I#MB zx$K>kuUlPeF2j3MXwd_)OqJe5s&%tL<|WC`1A41Vb6d#EG9!!TElJx(3D}Jn6F>9)6+BcOP=vz*FPAB6=!=VcxF~gZ2`CK)l&p zTmfdqMRAumV ztuX1oYG#M5yfnQTr?fPTD~QP+K%upXZb)YAgyn8r&liA$f)Ig7r}VF zOtax(!@q3bcg3KSnE9S7*GY|kxz~Ma<#K+{^uKMW;Fy_BYJlU|k!(3i?ES!nvf|i0 zljtr<54I`705LZ-uHi@e*fW^Mjs##tkL3ZT{~GDm0QIsQ<*RngfB|#>e{W;?i=}Zv zkO}UAePV=8aUWpzrcn-MV$Yd;?e>(14f4;v%{Ii`M+{ViU#x#w0pc0X=-b*Nc{8ss zMG(^#k#iS>fXXd_F@|<*@nGnAfs^DP3p;0oeeGmxw@)6&)WSx1uwZQ=35ZGm4$O3S>ccBOH;iG~xQYtzhQ1|J z;QGb7a36%7kKsB0SwpPMcoXiEv;Jceg8tlv;<6vrUV?D$CCbiUE5OX#pK7AZ-R?70 z!*1$6>CZg<52(tuYOtf+SRLPPhM*AmW;&A~Ge@&uXwS|Odi$Eitlx9@)tI)mDNyv; zlOT3vh(T0NJh!vY`W$pU3O;#^rK-bQe7Ze<(3{HA{%Eu{=dw!u{p03_5afhHg6_*B z7i^@VlS4^x^clO8W_eHcE%wpueG8)DYxh8&D)R8{jI~gkf&3>BlZ3!=C;mgs6lj6x zJp9Qn=9N?M%_w#w1@YZ&r?UISBTBbhNTBE;R`NopFcIq3vfeOYt^laEME0~Dvgd#q zS;0ubvnX+bYd2O}I?$oJc1QgY!jE~t=c=S(XNA&muC3~@RHGm*)TFC zb0|$)lAPds*>9IA-2B2W_0N|PTpNfFXJM2%$r2Z>BM48|B@s@-yEyeIOpsxT8%KoS z?C)kTN}tdiMZ_Xm0euDAP11o+BM=C%c(5LUn^5ekJxpfOGjH#Q1JYncoCBP0!DssJ z+0h+PPtGrNe}M9(5=-o;7%lnNm-azU?;Vv;Pcon_)8k9?-M9qwuzf!!zXZB0e0d_r zAk5Q7=6A*fKvG_C$)_?eJ~x+3PLx^}=~8KF(Qrma8>e9mJY1Nilyea{>x;Q@eXvxU zJOB5SVd@W}4T4Ip?P^=>xvnrLFdZViE8rR3Eq-@&`(hucm1dF5bX zFIt<;K%42HRdDr>+=I#3^V<&HO+6Rw*e&X(gp`pm`Arh6^dw}$#V2F-60gnIO*!-E zYECwme73X+SG9xbM0Ft|p5GgN4~wBYuR-0=G;L6AU}^BU2y(OkK+xgLjXkb23l?KA ze#elkuwpXXnkQJ*#JC;8CT%m^0%m*g(G5)mkY`AUb`lL7uq7KHob~DbV!)HU;twk) zKcxyB9DC0=T>93>utNca|!_uA;dz2Ju_cL;n zm5U0z#dFq*ar&n=#vpsgl4P)QUS_HK-&WdxM834lRBn<9VxfCiScM+?$>@_m6x=#E zJ0tQ2^mV-2IZX5CrpnnvG^;0M6&xJou+*-kuI4!l{U9sP|E!Uai$vo@D84vKDLxC` zS(D$}t?b3WsikRQ{4nY@%LIQ3InYPbP9|1P{0c|1v)#xZkGn_JA6oF&HEoRUjlAbf)186qR$L&<&+|ERYpi<={8-hY|o2 zUX;`I|8{YMd8tKKn4oWIH5NQbzRCn&-}=g5oG2dA`tG6uDwlY`+6lbfZolU!y`Jxz zY7zJGz5ud=H;wm!8tXRbYRT?}MsA$7-ZC20L3u7wQx7%p0{apPk}<8vwcT>UWuLnj zWyAGKY%2WshtRwDa07vdZMU;Y{}AH1PE+G&avuJwt#zf1h&Bo2%joy;`*YW;P2~<; z^wFof0f=LGgAT zn*-pXn6euaYwuxOmok?WdmseU%Bk{Q9A{@ikEZ9Db_QjCx9n zXF^;~g-U$7?t>>}G&c(xu+Z}|A0`BAp-HckYrG0mW-6_5baAJ zZig!f8BiS~<($`J+o}u;Jdc88tIyU|AIq`{pSz@qein1OYIQeH!BYC36I6jFFCtB4 zO1eBU2XO04LX|vMgkL+h@Ets0SMYH!4eE$AKGW&5bHW4_A=W7!Wbz3hF&=r99nOV8)#=m zuxl~$gg((=jMR)fOu{yD^nUwqKIcd8uF$5p|4k+ByoR%rie*>3F;b_U!^|>kGZC$cCLhFF{mr zlH(kBCVYOX^b1rK9+>W^gtORIL28sz9irRSM=Vn1^U;&L9mlejBT&`?^PU3yn)Y7n z>u}g!1Ei0h=C7u!EY{ZZU;!Phve#$z%b&X1S;jh$6PP3FG^?tGH*WW#mjN8)-_^GQ ztxr;CO=<@*1w=&TjO(|mBHG9w9s53t=$&?i*Jrz%hrj#P80GM_zXD+QE2S&+&5@ZSCi?tI)4~TB{rP{r5VuBr3PFdj9(2>5H%a z3|fR1YUN-u^MzbHzLPsI^tv9%b|MzISF<>MA&aW-M1CEWzb{4k*?adcS zxVo1=`acCt1%;PxMu+bSQ~kXd{M!IJp1Ih3?8kxDb+WyjS9I?z*htv>gS+Wx7hj@wDN^hmZ(PSIbejRakB>>}Wx?FEiI z=O?;q@tL_>rA)E6?$iXJJ(6^n!DoB!@RKFK#(gjE_qRM@iCfT)(%?^jwOYsB^f4ky z?dj^{6-c~#&Em}qToP|?O=_~3d28jw-1 zcCmaYq;u`b-+UvP+L_0~NT{4oju&f7ub%EwJg?9a=Zg*8JBczGBVtK-(%7aVs%t$P z>^yLCD!R>H*E@c^Gpb(NIP_C4G$bL&%s>pZm&LbPo=;E`A{zovBrhIdKJ z7u{*6O$sLU#q$noj5F+AJ1N@rAL~|9sm@q^a`ZR)9RhUJrqKL;mK7Z1wH`}_u;!K2 z4b4WP!C?D?3|q<{ZLT}zmnPwEGy=UXK4E5@XoLu%_7VZmEv)^xY5rTQF%ZNiS0|#a z)n=vn&fQlF$^G(kE^j-TRV=7Rq3!_iG&joReVpi-)wOrxDV(wngG}G@*vxfCGvzOQ zqGAo+JAJR05wP0jS8t4T34v={Iz2m&P9Vos9J)YwkV>tdyP=Qo^dx0VfyfC z>mrr2a-VpMUv2o6I+O1Sxrg$FbhY9uB&t7W`Ac7gCgzY0T1K*zTS0%V+xenUT#o-~ z*WNE?LYjBI#AvI%&^#APOy=kOr6>m;?3*5f{FA4si#U;B5_r6CN%-C=Z@ur;WYwTI zuf$Zqj`$3|Bme>Ve?15gaT5iD4XmXHEN?>9R9cY5f=r@PpSa(wDCXzD?^trlKTrT(9{L*iG=cbO(0F zC4Jv+b2S)#hXU+$xSO;IJ@b_(_&RgACQ)&z1W(hCWrXul1DUmsU-liSA7c7ST0+G6 z)!lPzl?N@_k5s}CVZ}rW4;d1waU`uW{b=hrf0%zgosnvNMTcghiY4^;@XU?bhI|PR zh$8eBe(;C$;@S8v{JJJCFNFiYfcUztprd|iDx+c}9}6H$_6DbOd?OA-?8bTda_-`5 zy5m+xU&_??3k^@K>kbgGml#Q+F~13p2~mZlxobIFQA@em9COXJEi#X)36=NAhF)O> zfHb;<@LyY+{=hOFal<6;htveKUwf%>zB3R?Zww8@vh;t)qN892HC6c34+Jy08YR%5(Ic!~zvU z%tn3IS`1=GbgYBwkF1`}o`951Z({q36Hk%l+3JBDgSX$s|4`w~gvDHBpi#M5C0Cqi zFO7Hgl@KsvE(@-0Ot>R;#3~tv*&LMSy&5juv5S5urLpHxGp;f;dz>$0V?ixa%m zR=pP=-Y)))7|ndQ{wvj5;ivE2xK?8?50il9d7qyhUJ}>#yk6kPS<0p>soLhs?bmC9l#k$0_VvQ2@oE**B z1E7~Q5me8{Q$76G;EV0jyNq#$`>f<|U7n#~x7sHs3V~Y;Ij6~Zc}n!7bHw~)3d)_&r2Dhfp-Tqv}iwn z11p{QY($8EAzO_eh#3o1^1GV!xE7~k;PQ-lB2)FNo)p{_)^lb&H|a+7%}G<@tVM z7$YWRNy*~U-N$yV7yInSDGq#3x>@RYWp^l-A^Ye3(2WKX(uy0~oI)*&?ioU4v(xmz zVIQpg{D}9Q);0&x-<4t^rG3r>g*rD&cjGO!hV^3us#|fnphFP+ zs7@u+4MjqSL^nZns+@eDGsD&0&Ci$tA`1i_?tw6L&%j-olPt%c+&rH2!ZA)}k zW(x7`y#ql;<(jbX#kS|uzVk{kk!q$ynx^h#p=u73W-U)AP0`c9!1_JS#kH=F_Zs;W zWVRpba^xt1UDCMe=vB!@9r`F4dmRUyhtgO2-6HDDpi!tfjkCSHsn^M1a|=&@?$e}R zh}C)~pzcYYWzY}U*#zyWBPZ^`t4(DcoWN zQy>eIpLNr1qGB~jdetSp_%lYTHRwp-?36OQ9N30;@Y^`3Ac?(6#0M=e%bDU=LTTxW zHicLXl@Q}o>h5nJeDV7%w8FQuofk8Vxi_0F8$JlY`+=ZD!*n?SpFzh6;A1bic%6pna0-i0R zaUPr@I15-E$l7PVodjeehC*v3YFvK5}t;+EG_wy!4tk1)rVmxm0>RU zYjyNk^eDp1*79M<;;KK(KX-!4aQ^}_OQ z9LCi_Z+Jx_ZOn)H04Y-2Yn=9^M_poU~AW>nD zVl8Qa>3a8;#2xZ;9~a@{lY@@&ybY;ruDd!8%?~meGhHqMXF?i`tdAe?O`?s?O?(F0 z(GkxDk4(Ck+FYMh^(+RUxDrsr^ebc0YH{AB!`TFLbU}*9+%^MJsVOulBN^fYgU|t3E3qiMkNu4C!&{wm~k`T8THs< zRqWKO;`xjIzOJ19>M6p3$6r3#nU;^tn?*iN)PM9isaY#wz?)bRX(|kC*|HPkohJb# zxqLT$9MIUbRAEogzs(ZIq(xw9Eu#OBWj!ZTZIRwX@Cb!}X}~iwxVNCjPTC5Lu$X4p zj^pd@)(K@zH{2pfpbubgSc<~T+qrv#Mv6MFnG&gQ4(86z7v{a<*M)qsmQSS}N7@c@ znmT_?U&?;xp$DhltUFpVV3T%cPE#>#sC_a+QJMC>neQy0>9yd*hj~|1TepQXPmi0e z0$Vw?;9k~)Ow#Z1Pzv1wWLJ#5;p_qI_+@=iw`@9+%}dg3BzR~@b^?R8bt;XJ7m!ZMn)Fc;lM=n3C7Cz7A<3p^S5;EZMZ&YFGBH;4LWK*GkD%l$!4W*2ts*jDx~1+i<*@VHW6 z)f4bej6+mE#<^A9ihiJGW3?GQpUW4;uJ)zQx48C4o4*rXvQbq0!?j|@?;(qd#DFg9 z`ck#aIRP_qT<>q!`4e5*Ir-`n8+ruC7UsUeRP^X$#EzR)xa#ufTdS7XC;9ErAxmvZ z_#3OphiDX7biE4&7=o5ub^N~ORBP7?zks|ctTr-Ed9;p_c6oax1u=$FIkp+~ne|as zIK*|9=4K>$P$2cfbGX6|p0Dp=b~yt8kEFzJBU(Q^(4jt*$QkI8nI1jnF{yo1v){st zd>~~$(S^v8%*^P!RQMDQ)&*xYdz836&8VnC@oz?66m5p)ALb)gRCPg#o|TjKo(#6*bgaTJqdTw)!`Rj!dh^_(^b1hq%AVQJ%! zaGL~>7n0Sb!QL?L5xA3gb*|@J!857KkX^vAh`p#{q!LB`UbJUz-gz&Bgdus;d_={a z1P-t=)-Q{DR@qxY<15kWpD&w46z_D5*}ioB z4rHOypzk110Yvz{8n-but|1+1kOOjO6)x+&YNj_W4#+sHJx^G)Fp$wy5bxg0xUkAF z^A2tX%xSC)wM}xSBpKM2BZ2)7kPkeZXdAzd9KSt=)q9=%uqozM8$it};ir}#VOFSm zn%rE)kF`tNH$tztyF0>+!bCFnno#q++DW2u4vQe&MNoNQ-E^VoWPvD>lJr??<@{kB zP`)Ectm=Yno_kl1Ok7lBORuz>jjSa51qoc@>Zkip-lMHeE}9%n7pxn#CYEp(8brCK z$EH381gG|Abv^ZxNt%YtWc2bO&Mj!arJ6qKT>Tseue!;nrrAV1tWSp)O$Tg^UdXto zy;$3SU9yUDs#$|D0MJ|y#Ya{$y!){W-&kIwXf-BMN>M>0fhKeQN+BBy+ZX4@t*_&q zPhQ*+)?hT7@eq#{RHeY5%2*$eSJk5{!o$xIGu=;cBpCo~6;EdLd|!b$XYz=uuA?Y@ z-s^J;+X|##yAjr(7U^vLzHW7T^E*p>Uc%VN(;AA1y1G-2*qW`F+D*RQ!5w;H_~Oya z&wN6y34x*_o{wWU&)rH4K_oJ}S0h{hujGRNDTMIy8h^V^pDtsmQ4eS7Q%q5C0{ez% zt%DBQZ}zT%5m)^e6(<)+(0Gnvqz)=*pLuh(i9M1#@|sxE?d)*@y3aWpGm{Cx$c{rz z>!jjpAI6v28XxaFQ}@2EL|m5qp&xT{2U)kzc$<>cSvaOy9+!%(vGsklE?s|u5xd@% zthtK4U03p2S(JmiwQI41SbSyB$Iao6#%jqk!gu6@4bw_#p}qe zKd0?QKg|dixU`yEJ+?X)tlX~Jtyf&Jt)yzU*|;Pl^tpkjA#921k*(Id*+oqRYkcqf ziWez{ZkHO(3(ICj@?LuBBu(eEZydDJ#H}SLgsefnptaY1w;L40OuHqdivlVJ+Xj}+ z4H*NEmUdiCmTfuN#Hg5CbfS&bdsNZH)FW>%K(tnkTbud@4$lg!y z^%LtVHO??6ElMwuOfRjA7ucL@XG5*j0`Q{H1%#un>r-5@9)Sw+@$Hb7Q16~+4`|=? zhOrX)A0O0IpY-v*$R-s8v2-ExGQ;5HuC+7oGafr@Szl(OYL3d}^Osmar#tJCW{j2a z_)@VRU`0Zgwv4k>nUFYOZjh(CpoVUtbkq9wcN+b+kRADy{Oto%8G;vbYYeO|@x2uv z2sL3{2$U9=;*e-S5L7dDYlFo%yvm}yGiQ31QBXK0O;0Dpp4dgmp72Lh$P;T(vBA>z z>ttm#MjeheXoi=VjY={Hp_?YB-R=%x0Lg>=kZx4hIk9Lo=LIL(o1>joPq{Ja z36M6$s|Y2{dWD43&bp$I(Z2Q0MJTyFW17h-skUw@_qr{CYG8UVEAlzSh{rkZkESAD z2G)ah1n1nO`CZ7#TJQGR33((aTjCaHH_u>hVJ~FFKJ*)*(c94RyQ4stCj4n4mDOYV zoCJS}C%GTku-#5yI-f{#E6RYG>%@q^R@ND{;hYvTxdk~fP#@2gi8W=9k3f1I=qyS@ zuRPLwJ+u!j&}~M^vfFnuds>BUmpZLtMhXkuu8_gVH`wsE_Zhxsr=-j<1bH5>4Bc}~ zRsE~AfNwHLH(Xnp^Uy7~4~u?IHO+Tnta6csY^|y$=nd7kY5z`I61b-GN~@Z4bQof& zSx#Tw+rw2oY&c}dxW;RLjVPY+?YAsd$?d983Mbt>3lHr(5~b~oqhmmi2=76+x2kKx zRfR1D2hnH9=c867j{2^Fb^9NJ*FM;b4&X4W*Q`C{tu4q4{ZN*7v}& zbJm0|Vy05VbxlpeJ}Ez8F6Pg(dAs(e!bK6lu#Okoaiz5GZ&g>tY&yrfRNnck*!4h} z=qQj}^A*rkSDm1}Olbo53qf^(GyAJvQ4O5-)FoTy!#8IqpWMt)kBumhAQe|{HOXhj zd5L3SYMdg|KyU6$hU;oiZy#pR3i56aseW@_91ZQc0 z%sd_{6(QNfO(bp)yQ!A0za?#SC|r3dLF0DVrm=!Q6HF3lX*}KtlRRC@v$Q~-8+SZU zdD>FG|3TMX%2xdkQY?tY$t$2J-mEdm`W1{2SjsujRvDLf)b?QExx!jA$ZV8Z+@JL{ zQ+@9>mqJ%3$tR7^%>?!%iL<0&jop#c_93zjWrY))N}?lpn%Wwi!C$aZ=)*Zr!^5?I9fv zDc+Ufr)O)oW!y;IbdKBlCkhrb>M=coZmJw{~IZf`oy9;bRi&V>9VmL%?+Bsvy= zVW0q)YaxSj%D$fNoH%rZ@d`xDwQLVPHrsK2U#7u$=c06k8Ton$8%=MhQW$H*UL$c9 zRAG6%6TjwPYxi~}x&GK0I1;Eg%=nN9oae`C-2D~TYwqvBdivv2*g`J_|7+y~^2l!U zyX2J$f-48@ix&Cq-Tc&QWtfWv7{%*NUDbu?>&(M5ESL{_kpnr#+~JCSDSTRfW^t+^ z`qKA#dKz?9@W2|k@LV*}M7Oxj%*DZWmAC)HM_Egvg`RbiXX_dJgVUIsisynug16MR zr0C-|-v{*hr`QkILU4kUSeIZR;!Uu!8rVl4{m=@3**R2VgG2;q&S&!F9X;KqsyF>4r9&@4>lFzH?#z68?Tnorwfk z%t=wop#AXW9GnI4j-F@Zvm~HRXGgT2dI_Zd9Qomh$z~tmKVfBCn)BqQxS5ng!@who z1ymoX6Ut=FEKhzTQixgPw!J*tv$ur9eMo@_=;;)00oUsHgaLZzDiJm3NxsT1Y_qRe z^|E*oLGU?hX5sx>Ta(>>>T{`rG^+J6agvkTh&;OG_RpEFNRP#Zu1$%B+SZk8O&14g zov{1&dxht2$5r|Y*Hmep#{*j#yjwgKY#r~sxC|ej8ta@dnXZ!Bz43SX8sFt}zlNVP zV$bjlIC3wS%Pw!t-uvw9Ah^A8jaOlokz5uE;4>|_rfB@sF1$fddj!czJ+fovSzM;D zrB)B+5qMYL;OtG7U1()@&7t3Dvq^Q?4^D&Vc@wQ$ld*V*$krZ;px9n zWNmSpLYXJiZSts`lX#=fboiFMH(7-IsPk&(#*Mpd`jN-@YcU9yX3Oh7y_#&M;yX#| zz2_p)Y)I6`7et{Fo|gc}=L7Mt?E?fawl9Qv9fz7P(ULoV&y{Z8(7EFxc}K0bg_@qP zb+-zQt*Jvs)@3u^dm(vtd!GI8B`3jUo9jM5ddg5`p6^U>dcDoQiC?=^8>d8WaR3=H zI|UP3cX+h5QnKeOsUjDp1E+ANq#IeH*M&$wbH3Ehi|tPDmvLv?%qrE&&taTW%D#8T z5a8_@he#RA#e_QWI_d>Cb5-b`0gUV}=fk;&lW?LvMtPX#ueP{GJ(puE^AfB$i9LB? z9OufEtN6*xDy^w~;X~7SadLm;bAX1)qcby;ZT4X0d^0`UR?p4x}pjWXT z5ufg)H{f1L<+!iGY>AN9VNDcVXt7^^tujy*MOX7sqU?%qncdfG#Bkh4?(Rz;%w2T1 z#hf9Efg2k6UsO*cnAn%A7grY-S|lcJkKpS~*iweI#bN(OTZKar5(<`(zDW${16ta4 zOTM>?sp{d?Oy?tRIoppFEJa8h?@IfqB$d_LN;@0so4zj*zYQtTReg_d05qfJRQ=0r zTPfHkHNx$`AtNGlT1Cp96>R;p%Y9S&B@Y3hTQmmhTvNo9nebHz=~w6vH8OnLpH}U5o3^w+KaXi4 zGD~hti!+6^R&O1&(ZMxY;PvBcL7RuVQ$236rOYtXK_2~05_vT2lv+Kojajm4^S ztdSRz*>|i|dA6=S%L$x+iFyF&b7|5`U&h=B4@u+Ke?C1Ol=3otV!B(`fwHxDdce~| z~bBU!m5}7 zuWyzq7dbXT?mK>Gv79cWy6UWwrtij96}IGDL#pA@B#A)8rHx(AJr11l!&SK+)sATt z9OInH`$bnzJdLTY>N^E>`z$7Mxk>Ei4w%94p;_glah~<)19btX=K^rB-E3O9Ns6~t zfeaikeszk*E#z%{shLG2U%9!z!EnO`wgMN`?hlw8)S;Vg!p`V$!p!&gb+jJveM~SD zJkv9nNXx2b*#RjMV7J23e26$}63@<&!ZL>=xW3Rn=E6xGVIN0vfAx@EFV3{#d{LZB ztDBQ8Q|~!q-?M_(K6EI6zRu@|kJI7Qi4?;=qY>k=!1ickB8xA5;@>;x_z;hfAOP%- zRodO?kD_mqR3dp_#!m;XuEskwWz#k@duN5)b1fllh)$=uOiPB#+}t@^V+eSC@TP$y z0ZA-hPJlTuCRvf`fRD3jHwvbKQ4|~Hq`CtI>eW~ElHC3E zwax;BE*)>_vo%!L>su*-SsU}oK(4}ctnpQFN;;gG-?JxJ0JCtS;n2y{%G2lSjQ!0h zcEq5Ifs|N`xLHs42r4YfXiYI(&uFGWRA<(+HjcUY)Nt8aS{}SquIo$g~y8UDQ zsVZkuRu!wYr#Au&SrIMQdgP=&ZjI|pwS~*xIlOG@WQOplXu!M*kX6BE^kJav%YlUe*lp#D3JQ8{jeaz~f z0xwMl*y?|5>QZm+`oIUr`BxbRmFuKrdoPtJZjp<gvF zOGzPP-Dh2f+jY(^*Nkl>k3buRuad|!Y-*mc0Dh+ z#$S-3Q){Jw$X9oDS5otxxI)Wf;e6W@8gmxT20V!|Vof*j!ydEN`aBJ__Z?xMB}CtI zXastyPd#`|309&X5)ZRa>BLW4zt}U>O>TbF9b6iQ9`$SSm4A0SzTo&4xAO_^ zIv0Nc@Lrb&La*?EJR{XdKHmbI9Qa#U+ct<>zc=*hq9lRo1Fnd-z>WpKQ;O=wOb>13 zJujtZ9vH)ww$%@W^+CS$KhX8wWV|LN;m+*`pTJaSg4S|dXNQNG8PDbJWZo%eJ96YO zllcb*=cg%MyS#=UR1SqSu*xqR;_U>-*_Uu}%{~yHW}u1*T^dq9)fFTpHS)c5pR?w> zUXqBp^{|wv`_|PJqnNeCbuNo14vFj`CeF(!_0o;#Rz6**<;22*hM3^3EIHU_hNG)s z1L7On%V=M=&Qm@FC%{#0IKY?2wFDqZTBpha>iMNf0%p!D;Caq+_<<3ZVFC=yCsA8{ zd^M4av*t%?-EoXwnRDqi)%1+H(%juovZY2F*%jEY-Ek56If}#uEsx^$&E7FldF1$$ zXk*g!BO6DLPEfe%&prkmDD*QDjS4W9wWV*8^5m^8AU~GsMA`9mfj13Wl-}~oy3SkV;GjX;09q@WjKwHrRE1iaYNRJ8?0A6oq~26SBc0FB^QS^n-(edF@Q{^XtKEvA#dw#e9vRup&pJamt}UjnFju* zIOB^0;X2|nF$Y=f{=PD5f+j^y-R~dXTc_D4I|ocC9Cu8PjR0-lZrL=vAd17<^>fPX(&i|S z3gry=@uhiPYE<)rGdAOu7PbI(QCmrUkCzQlzU+ZKQJpvwD9heH2T06Z!rf&cEU_Ox zM;Xn>EG5p=Ma|NB(0$UeK-8H&b4=wpEGd3x>P#d>U(*+Q#LuiA#5M?pD@ID)?(%f( zj$rF_)%O0_@71%Y#RcC45q`ItNLgutLo}YaMqVNQA>D9gJ2t^?qRrPUI9; zbN}ZjSH$&^RqqRH42GQieU8(kvO-9OA|L@N8Y51J5uPGPF>FG^e5gq|8$YOfyrm6En9zIHqdcq8@{cs@5RlIR}mMk{ap^gi$uva(sE8^pM%*WJ}~*FXLA->$_v1$ zS)q0VHo>WOE~{gPJQ{ggY&EOLBVWa-D313`#_1s|W!EyjNcvWk~nRt_Ta9_f8 z;^RX)6oOlCkkX}`OQPNObEw%J4b#2)Lc#^y(Cfk-2YB}gYc}ITmvS+)_E#(Q5{AQA zau(w{agiK=S4!^~KM=2$zuIN^rPS@!Z^K`iNA)?1ZS%v3l3_Lnpi)mH$s}o0_ZBN} zpU3lJ6z{wkYsQ$+Ypg)SlH3$p-`@HhB=(j5ICgZO>oJ7GJEd1mE-5Yh6*WojltBG^ zM+nhu!Hw-x)v(0g45WB^x?ri}F49~);Yg(So_;zN$<|$7gZD9EyIcT&D20#>kck-X zVU++7Enp`<4O$&r&MVyPU+ODspJn7!FWc))t{N=hsF&JvWcDm_RrG``Gc!dF)&>`%wMB(+ zApZcqY=pFN`;o+h(yWOPTcKeH#tcV3m*JChZlSe$4}n&{>|S#m*U41a)wjkpG_1gu z?YDO+3w1tUM1A>^FFBizukzQB2}zU|R<>U6%L1Gyr^u5d^U4eMfD||Dbz`B|p{<`p zV&6LCf45V{68$fBcM*DV(T83P_G-5gG-u5EjYgvwQqXky`o&xtrNNt3#vnBy?yK=Y zyFj!ds7sHvGMo)2Y4;%+>vVb?|MnS<=AF)N*TEE7VPm)dBRt|gq_DZ)7 z@(#j#qslwlhju1MyC9EqZ(Uu%RTWgX7)E&+e*fC} zvwx8P&wnBa3DUCdoAS!%vqf28;+pxrc?;IK(o1eIe4q!W*e4F)M<9(_$aF0{GD8O^ zA(AF$p{k??2ZA4J zDOJaP7dqHBRJM?3YwA|b3MV5Fo|)wLgM1Co;?U)ORhs)&EYs;raMb&T%<7&deu9TG zyU8R$_y2m%e|A+a5u!f?S9dp=-jnO6n{VL@9TK51TIc5st~m4?p-qtq4~Jv+HFZ&J`)zB$tFM3T7Jxa5|eBC1e1vsXV}s5cFL zI~=-{L8WtIn7IYC?=C(+abRA(b_On8#6I~_*&%g^-ix6vMI7}{X{B#phFd;hy7l)_ z|DX9lh=)d-7hHe;Ln!`TLSBbbQJkiI1AgF43BrxE0;7xhh8V3N_NCD|^}<|$uUvhg!`2N?WB^#z4Maso<#@Jl)UQPQ)7|Lk zV9GYoaG@qc?3M;yinib*{F{U>0oFz24akRGufcBSWmIjDoEsJf@d&dySi{3W2eWwc zLNVe^xSNyfoE1^59P>mOil zx}Bpf@O8=VH_L`cOc@?A#+f?zU_szaQc;e7RlrF5 zo&P8H4Hf4(YYUUK>Gne`Y0o8fNYgv?WjJkqcjIr2Y%^q;Pvi<<6v!)hrzjR6ZUnQJ zH%uk9&Zn&&i6+9X>y63bdU)p&K|lImX;8BAMsQC3&YQh2+6*VpzycPAeX$$cBF6^l zR}&R#dg^KmZAP*z!QsuirE1O-gAAev6yc-mcWJ9ieJhRYdzuCNI(}o^CBo%<8rr6} zB>f9Rr^5w_NAYLp$QF)eGUP-n*Lo3N2^%WsyS=jvYBU?+=;@s$a?Tzs%9N7ZO$a*ih_NM`MQT;nJs@YN;j6*QYDl zFK=#cBATc}mwfF|7xK!=9=6Vs)i`(5obuRTc=-pANN>K+BE6qf`eDBDJc3Y9Sd8|M z7lGc{LVj<56K|>XrRm1K$_HlVePX-6fInT1|HU;weWt7y%yN_{w#@MOH(w-Jnp^sv zcea^Y6uKaGBOKV={vf-vBpkN@k-b41}SP5fB>Yp3idNZfDYolXu2 zk^Ijx;cp9&{2Xto;`qeF-$F*@Pp3y#OEdpmhF{#GbPsRozwr3Eeg1{V-x2#SJbohM zzjp4w@c0)V{|}IhhABB%SpWF()2H-BMTnGw1>xVs_mFDy+^VX3`~m_fiJSjjK6WiF zS>0h|1Ybl~g#3+7{)1*Ts6u*sm5AY3mkbu)z)auP;Xw`Nj7;=k_5qFI@2T}Y;Rq{l z5zo|bYdQ`GJaJkF_^mMfma(4T%RXA|G^gLzQ^_@W;;fRC@Ff3*%YIV)dpyO*l=jvB zp5l%0#A%qrBpUf!52)ZR{W@Yj^;?R!euO8^3Uvyw>|dPz_q}nGqX=*5tFTC+Wieb$N=nMB ztXYoT#HwtqU|>fx+xtzR&PbQ02PT25UjgpO4TmVwX|DmIg)I6Vs-Hmt{p!cfTMeP@eA|ebD)s+_HF&3GvLBuR8n3;>66n5B#dEl z$|$A^x+EVIG$*?N0WpmzYm+*f(n;=xWG1@g;0892k29M>yO=w=a1My$#f3|M!nn86 zmFz$AtN)e>+C(01Z(w;92;7;c;UVqYd)2&DXTsD%)qZhhbkDG)Kp37q?Hb=GVO9fVKkME z*27)HGnd!@aMindx>k4aI)G&M`V5$HVV`2gPj4?>Ejt^Z;tYFOKKdu=r&hEzmC2n= zu8S2c=IC^J`bPA~x)xz`&f#etamd-LcgLV_Ly>?VA7*yoQ=J~X&9~W-Wdh+}X|tkC zvp)sIzKxGW&W=-Wy7U!c5#xaqw+IVwj_(|24rN`Dm6a_VmpBm*!zUaAEP_q{Y{<-; z+`Jmb-K7B6Jy}I0dn(oNM?y)z%^X+mjz<#dt{{4md zy!#d=%c^R1zPp2+3UTr<$+&PqmxUD1O59Phk=q2_!J_BMV{vhRpj%`kuOBy$1OGs?OTG)wM*D~bio5R_Y(HT&_QIt< zdb5CCw=}Zl8pAcc3p|P*ux@_m#IuLxFC9pLb|wrBg?lc3MoLucBkh1m)W4+Axygk1 zDL&hjZ#DT<}Y|zfAN^Zrc zJjn|4IyV2)g1Ydoh={DI-57OA>@%O^33R~$ZJ=;(}>yfrK4BkVVDwzai(wQkzSss6B|zq&SQJ-j#v^iUjsWd4vZJ+nIT z(b}5*{fniBXk-hjtgn+eQyolqzApyqz;ESwr%LlCPBRH~L73^m?1d}-g(#{@qxPApAjS=kMV8B+^f z8TCTf`Yi3Fk+F?aN5~=cq7(QO+IzBm;!>uHofe4$YPVnq?JK8;WI6H6XA%pR%ngsq&{9qd;!?uz zHnFGpaMzLDLXtlaEpg)4)QQ3hzP`JpX=rE=w-v*SxvDOK^&hWX|AsEon&WCYR&kaL zlC57oE|+z3a>_R2-s)>@If1#Z;NPLkl}WBY0&$-6M430a_7xcx?SNw3&bp>VM+->o z15c~XNz5&EtIj{bf~Ka7JRtjDh6lpY=Kc?w3)a81gMwQ3Jc07aEfVL3KaOLUO*DCF z7O_j1a0MLrU}aGqe))<1W zAsm~$TlqE%g}2}P56Ntqy0j!+QxcUvkDoiw4i%j!WH5=(BhXn-lId^PaoT_m*fffnP+(fmI z|0b9Dx4Umd{q}5aTJKWn42(uZMDq0svux!K0$VSJUuIf)%=ZyAV*D}e3-j%=3pT`k zSOKK^IjSz`=B(ezjA2Y{Y^=Ol_3Yqv@_p*>W{PP>B0-Cd8oDW+xjOI+d*DPl--!u1cWoXe%2T1whifBIb?V$cx|kI*BNxvd1$H_?Wy~Ti8#%mbQ03D#~m%Ou~pT;Xz|w zbQGRn`zEoQzI(o2ND>m;5oyZIS{n5*0$9s;6d??6=~Xo z1$a7j?eBP>n!Xbc3Y-bE=bL(E*2-50UIze7gVx2Qda}aegZ9EiTGLKvaGg1)_nP;5 z*M|ic*9XH4&aq`|!sn-q%oGm%<4NU@_R;Ww0|4SAdxG5IT}Mzhl4W^cnP+5vnIylx zA-(Sr3V&Y)t#*>rKTCYqIMO{XA|LEgsTMnY9#@1NQMZfg1U@nWujmgXslIF7f~(Rd`sWa^|nJ}BYZJ5bg4so5ayWV($F z4BFgBgj7;2crC!h>!|jd?>a_#0zGN7Taz|is3+*so8xsZLB_*MqMfzVYB6IEUN<8y zrU++*ww7{_1!$H=F4Zog=XBb5y}eV{Q&sPFWQ>+S3kYaIw2QR|Y)i+c9bKIF33XiT z0WRDT(!R~6fHKLwm879Ig@UKw2ebu2Igo|}PYKbNF0>@T0H7@XSb(q(7oO<&rv$s# z>#~@6=3u($e38kB>J2b>DgdR&L&=nuscqY9`38n;9k~Jt9wmLjbd8*Rs_pEeFmwgY^&;EF5l+N-VVb_xWbYT8N81x~*}<-Fv6R4c>Bi!Py7u}7C$ym3l1pFSDUZ9gJS z#2|Bf5VkKnfxxUNl#CEXna?0R?20oE=hc3uRGZC!myJwY;b&a~PCzt8VCuDZ<{qQg zBJ4Nz50=c&`h_|^S(aAmoj1G8dAoVm1-B*vwXA9<;w|qV42!DKms`CbZ zF?1jA&aM=)X9r5wi-#!3bAOg7b!sqm-$~t8Y#14_B6=oN_U(1`BU7pI5EGK(S^P=# z)!s+=^p-35RDQ1!zw({>7aWwy70_~RZ>&IGdG-C?7r>jJ%ga5?YABoM^#3Ra1p)C9xCccBtq=*RHc>9+Ul$C&&cA{M&NcHL$GKh0 zRu4M8-yoRc>y($nKBu_Yt|l~>vYR4xtm6GqnWzf31wo?WVF^rkR07TTzy?d~6LU{v znk@wX7hRT>*hF%nC&5yaWrNgu!KO1AUid7TczweasiAhIXR**EXQC!MV)3pES|06A8oYs`Q(Yut_}m@v6qyrBRS{pN(i~wzhg4sz#x9h?XeZ0hhO& zqY&TvnEfx_reE?!$w|1WdXOm01P0B84O!*|eQg&83=49UgKRvCYWPwXQ6X{Z1n#Y9 zk1ecUIN>lHWwaor-zMfv?$1uNw6hMWIjspgicD|6MnNuI5>ST9Jz)+wA086PqI31| zjp|1Oh%P|;TbVTtK;2jD&iWj-Y2^(xN0s!K-#46@^zrOFK#Vc#f^;kgk|qjs?I-mj z+S3BCi|5Os{!R}v*T0&3wjZUZnbe8ud&mcx+*^!rfM(IKsPkY%e{7{?Vv=1^9#uyf zM{l3bFQN!y|I5;*=l&9=LH*P_`PK&p8|y$6K@;o9%z-3T%SOfz^TFUmhphoeL{H!b z9FU(D0G<-oFJ(;&Z&V4>aAvCC&K);#TF5Eh0DQy#u~gQ8!8}V#%kTvwMxo)&3h(^? z@T-XVDOvwj=VEUgK;Yw$jD&|74n#Fkj%y~EoH`3rlr_MRSI;8?lt#GRkP)U_uE#Ub z>~-o><}ZT^ANK$8ofCgOl)3WQE6G$q4td1a{zA810g`7kCP8ObuU>Wd_jN;CvmgpE zYY)LU^sVUb;_QGwz65xD6s8+j=fPKCm@BSP*M$H!P5ufnRGW+Y?OqVGkMLZ6IubfO zLqblDc5pyeAevLwn*{|0ZDytN6&3p`12*<4P>&K4nggd@0PW}#|4QxEl=@}&Z4iVO zgNo7#07%17?V&!>?GK`DoClC9CQ2ohl@B-xl3k1g10YdmXd{eDv;M}qknfEX0kyMPUH<{~W@|1sc@p8`(Ek8%u!z<*G zp=PmqZqWgP>`w!+n|u(m&GV6xWJOr&Hyvh%Z}ceRS!k0A!nkSfEe)KudA4|R$ju|z zA13hQU{mTQlnJ6jvi_8J$tTgoX72pJo%_kz1LRzGc|)zwbX?ARiNgi@2{+LELbuNG z;-?aB#Z#kFW^|)-60+jN@JXcH(*hx;y%$WG<}p$rL6K44eK4JySTc`3;m{2O8^L?ZM_gJ&Upc$HGZGC` zw#myW4(unHztoyZa=3NcuHM-;KW{R_bqorpCgbuG8YlQ?oqm;@IK-;0Zg9~nzzXn` zS$NxQyf|Yh+wlnD(aS)bi@0#(%NH`K5b4p)<|n}pOm%@6a#myoJDKRVJIvQKyO6tdr}t$}WTX<(~( zTc=s3ASUu=%Uz+^ei&L4Wf;?Kt#>L9wx3~V-vQTrG;>XLohD}%$ya=)|5eI{Eo<*O zzP3w84s@{tejMZG=Yd70u$0Ini>4{N9;Pd z$y)&`_9nYLVy7ZyuY(J^^~TJ5bKa!Cn1xOp#~O*XGn-2a?W(jiG(Po2DDcu^kn!+K z8(CZ^f9yyvL}+$!Fc;2&iN9iP-RFua?YR2qX&QY+6h>~!Ylu>4oX!*F|cS>FW@NX zbpIg9W#DK!OzpZu@wqs-&cf-CHs~;Y+p^z!UeS$lB)@vgj;N<61BgvoAPaVCM=j2mf_}?nlWF|XYY-s zB})=K8=#(&G^1NUfQ~8V_RbDme1r?6)LCTs@oX4bXZV=; z{Rix1(=P;?5&djDqnv$4bw_7@V!W#wwu0)PK^ZnrsCZD1x}5#8mSH&Cg<((Tt}M^a z_`y8)c5-Q-u7BcGoGL#I04+pO_uE&`#Dh;e+HJ zRqs_3$8ouy+ZwCaM*jA&P@m#OZ=XYEQW&mSqBRYRjHB09&ySc-_CdXVSk)}}&yXW< z^7!D5W zW(((&*%A4ak|a?RRI%Sve>2Xfp?0bFbX<7SM7<^sT~#4|x4!Z3!#^J$)TVAHn!T=Y zAGpSnau+R$zHRl`UxOdAgB_V&RUDmi{4`_g)dl6UDO&aCH$|XAG}gJm8eyxAN~nSt zV8aXq4(2_&iJ2PG5XCmmJQ<)~9N;HJX7Q<>{8P$%f&^-#X7uEp3T4Tn&Vw%<23{k~ zmV4aY*Rn1r_US3k+0Ff?h*Dl`!W}Sp9U~PNZ$=e7$C^`i0fA2!luh@>Gxk=x!RYIN zWDNFmiIIv9;3bO4t!a;D6XpZEFFd~s5G#}1CzMsg1LoDf?2{8h;*~QvU?{K6k)FzcmXNVz1 z)TIr56Tu-`i< zinKlsp-Z#;K8LIojdlkbj9B zH_f3628HVJv(*rps?^7BRr%&|YOKC5q+Y&g+CrBY*q=(vAD@y%DjG%~{GSa>z^YqW z1YI05f|8{1+M3?0;7iPSWH@M5p=>Zv?_xIa@@N`Xzp^mqP}tM>6~cuX%4Cml8?1X0 zefmhT$cj_#y0uy~V}TmE4PUZMdp-tLm$azrV&Y|Ox%>J{;CRguds@2rTXkXYM=Ffh zQ!KQ(+V%vX^b{4++oVhPgX)VLX(x0&k}`UZ|k&t6+Tf_P2^1u9tmmK>P(7LN>z%=Y`4@L`dBM- zg(JvYYjKj*T}e%?n1ho;Ki_Nv>vpTV!RrWR5oUOZ!>#HdoL670*X76=s3E?ZkFpHC z1@FHOGj3YuD$Goc1l0VN{h7XSPEQl|%*&wGB5*}yH9Jq~!v{?Ro!hesc7q=@D_2Mt zlxW%wyU;k^c(Bm{y02Q7S>_o>Nn&~(9Kd-jD9Y+dgWF!?m&E=LPbrhOyghiu@2(dZ z>0OTsFBgc z9iF5&_565~8WoJlS@O{FJ4DN{onR=CDTNNP>-0V4N$-CbhIPztFDVqt{GR_<(se2( zWZAi(gnDvBu=FF?VU*0N5;vNeR_F8B_q#?OLJp{p5t#$|r)lAx-~1wq$5^r6_NjuF zZ+-1g`(#@HAwrgWQ(wNWKM!hP5nuu)njNQ-5`Rb*qLH_!L+9AD?KzgHUw(K7@hbCm!;A4b&zXSi|Pj0D`*Q{C0yLW#}kXZ%(>;S z7keX)&-wMj1Q&KD)W7PZdhp5KvEpX+)xxbUJ}_U59|@+1yzP=p&`t!PFCF}?xkQ(9 zypUvHnQ!muE^RwD!eXYyOr(tHMGc9?GG{rfG) z(ulrQU7CK6*-$6Pj|+mUy0`L`jQ+_tK&C(Q$Vq6zwL$fmIUgi;;j}F*_oSa~K8y2- zs1Hhb)~ZK>*;!(gFu;FzeLqsdrGNo4Be-K)*y;D`4RNY&e1M6l{BrQ{=LvB^T}P*3 zl6c?!xr{7bEymb`acBJwcQroh#t1Fq!CN!b&%g)oXfGH&F=lwBfz)Qjb;I1 z%@Vg`?(#ad&pcd`%?0FJJtD7VF6jV=G)zbN65I3sxUrOw#Ct5Jo+83VgOdYYofR$W9ATcg;{XSz*^u6Q#^llBHQ+HJ1Sl_A8h}=p_s>Hzx;Gj z08JS`WUAP8ftFg*b3xUdiBuqI6Bv4^GEA2493LW+3Cex-}-qGeYco%;UU2;fx_0)x2Pg^(f z8l606@oVtrJ44$`WXud3{C!RJEs4Hbibu9ohPR2%$VevQ$3EA{)OWJ#mEb)3?j8pvjN_xMhU^ch^9 zt3p|2*H2rznZM{fY2y9sfP=OPE;}W`RcrslGy%g^M_pbcULHr2nI-CK|D#X7$8BI; zDdv{bp*jbTgMsEP+;172Lr&B0_GSMOdr(X-Ss6Z+<*|N1@&Yw+&cCF7rt=W>u0HXP zVl`1gcdwc;i%+UX*wXh%p!pNI>PkzIjF1V40;aYv+>-E=2t&bArI2w_;@Ib-#+s8g zde;x1{W^O(g|PjnBBxrp4boF9pso8&Qop%>R`Q1HW3moSd$?Py+s!61)}vu0jUQ^4 zI1tUf-Bz`wYMAyhMZDqm<>jj!ThY(^W)M0N#x&XVbSV#nHt7=TpR{puX~u>MDXo4ToSUfwcDU@NI&ub#umsmf<5}ZeK%cREBz<;b%g<^q~HpNwB zd_mCkup>zNB5hydK+S}%=8Mb(UXelx(+A6M1K7X5INI+OyiDr)xH+N51sC`TGwSic zKrFW+V|e%)@}3YeeTuZ*>m_pC$wzOmMDBAwg{N3rhq!+iXgZ5c7G_7;*{vIjjL^JZ zP@NiIGdCx~YXcnuPIM@@z$biqx7p=;&emhv+W52PncQ=!HCkNH;Dse|G7a@DgdJVJ zVK?YUXFIRVvBDDW@hhThHmw+53X(jXy*Er_I317+I zTKTKx6sjkxPa?kGBulXggaEd7VB^iX@$#*7ue~obwG*pa@XOvKS>5KM;`61kr>*VT z^_hY;kOzO={vVI*ry-PVzibE%vSYzKwVbfnI{5KJLF}3SwF2(R{{-&;guGV{KrAUK zGyLp=Z^u6pmvA+N-{bw?p@I& zJy=&qM@3D|-n8Bo&H&v1zE;-HFUutq>XKDeRyN_)2g1jWWk<<`g)8>jV`6lcwq_II z5~7Ods?mvX3hS(+^Jy4`5}VXNZ*iM;+f!1|b6a*c2JWFAAt@$iaEy3BXgDnHewSxm zN)EHO#`yBpt1qdDxw13F_ zzaF)`0NT9zzT8evr#fY_vaFyJ&i^w1ZZA~B^t!N3fUt<$MkyAUPlR`64Gm%Ftt~>9 zx_?oS=UuuaMh#bcAjLsp$4F7(>6-|N^ z(;kY*b#j##G?31|7NEazWS&zk=+4g0h&a7x&wjqL-m7@j@(04=NF!1URN}=*@#^BO ztWkJDRjEYMxjUQ5M)jiJ{92eo6fgwYEt>5s41!~bg4Z}S;W_P zPXuh>h2}(D#UmF7hb*SXUg5aw6#+;By;dP4?Q|s*BEc6Q1YM$fPuB5G8UH9^(BmKG z0lhb=fq`A}6dtn=za-A|Uh* z4Djh?IbJf^4up~cJo6|}RO;Zh7oyCr&mTakK8BJN9Ej;DvuR;MGTk>0OS;CB=C@O+ zt?kLRZ!8L{62eUR%}g-Ej*lMPb%Y-RzDU_~H1CIx! z2{(5`perP!WxHUFQAuN$JW_9TxPyx`o~7W`EFyPgDMy8lio-GI8y*GeTcmp~ft@R3JHJQ`fZKf&Y-ct9G-L_*)pEtM+Fs{#j2*DLX8~n|J=u@PSXruF=OS>FC6aOr^`MMkXgS<$rlea*kDIP)^RGI`PkUQXb_cbpus6|u9_wdDwfg(F!ku@!n2+PYRg z9vRa3*#c(Z{9})bKq34vf-+?2e{0^58Z|3bUp-(sH%Y7X5YY5oZ;llIGmOiDBJB_Ws3xnji)^x=cErs)=TR@d10VM(1d|2Y8eg2>2r zdYKfQop+=bE5tPLGd4>0y}f#`os2Bl!+V;fo9JeMY}_TnZl#$V>t;Z;oUhb>`FQ{N z!m92I!alP<7Tv#G_d)r|R|L7ag4>`>B78iYBEdSC?KyG9B;v-n=W`_U!=)bl3{Rv>_ht7%d{(WNpezoid zz8cJ#(A<^((g;wWgaa6@!Zs5f9u@!A5w-&$VcZpy!V8w%7pXeUi= z{coH~`rQN&*oMr--hUhD`FCsz_|g9V0uMHRh4#fiP}-oKUX!ieYg%IgnI20Xh>VUN zgY($i*hCCj_o@6uYC(V0Q?NdG{Nq#4=j4rNkyNa+Rz@WaJ8&i`|8E7OC8jb*C#EHF zffolSr~2Su7hC%*+R+UMMqaKxeU_(Vua>1Q19L6wL3igaUsS^hWoHbJUQSW{w>8{b zN>%KT?udl@+5~xMkG$`zS;mRlzf;$$Jp(* zaesgM1S}%*cH)gaeiYSId0_B$u;2EFjbrCDVXwM5-Q^e{1Xk3D)m2hca;WTq5n)T4 zNb$rSxJ|gXh3ut7w0xK)ZoeG-1d5{Cqbl=9SWeEQbof zphk8r>9=bnztS7>IbbHOJG(CgYA7q?-Z(y2x3xh5VYl6Mwv zx-B4xQ3A&GyvyC@aCSwD;G&WnA{`}dGo#+EFudR#c%6>cM7(~+G(AyLN^pzaUG9!H zaxmls%(CQi{yUPlf$-?9L<*I(;iHu(frZqxB4|||{Yv;WBQM@v-lQ2j*gu3GOcB}T zBBOqEed>zsSSeDaVcK#m<`H?R_rPzvlu6RH+Zz|VnxvSp8AMEB`-r(^L zu^uS~cN#`y+l+1`T={F2&H$PruiUQcd~R_ zL8ccnXfe2AQq*%LPegOt9E6+y5mu_Vi!XEIY8NOfB<|c3Booul>h~m;-+*e*SOiym zFzPPmahBPzva1Z+Xl=QHXK)DU_}T)H2vc_#r3YYSVs0E#5y_*}3ljbU=s|s@J>mO1ui?Mu7Z73@! zFoasGm(Q$a6%)OYKll<*;*;bOr;&J1w|ma*5`G@d35Dkb&5YHPmL&vTg0+@^wW;)$ zHORtVscfGQ!k8^thMlD5Y=+NK9lRoa3tMUwWQp&x9l$UipMck2^bmQzuB|G1!i zryJmR5FNUwSO{t7!}DX32Mh#lBEaqDd1R72+&pd4;Rc*v;RNpIZBG(jy?QlS#g=qi zIBc5I_O=IXePhAU$uU>Pz2z&us%P*`C*+l@rCnb$f;Q!sW6s7ho}W*ZAi$vzg`U83 zng)?Y$9Qoi4cW6N?yeCRUu zsKy+nU-R=86AMb5n;{;v0Wk)6ZMyd~FiJi*S%_82HoN?lKtJo`)p|UE4$$;}p7+_U zk7`?g_J#O5$3qyC6wDA0pw|FZQ*yR?XH&;#`EKc3wbJ@*nXfLe%@AuZ@@ zdQ_7AEJ(PLTvw;z>-E63weP8))YPn8kMA)gB-|8Nnhy8bO-4#?`oU@Ao(#NnRNoUr z%mmajKd)XtpYlCdf|R(7?BSCIbyNrSuf|my(;L`Qz7`$;E>sMGt%{ek0s3^cs5xu! zs-MdqaeDFM_31)bj9~SVqXsT(fc50#HVYoFjd;A?a3MSw4iyOfS@15L%`+V$X{FGU zo&7i;4*!e^>6e2}Hz~QG6UjE07J7jiyM2p!NOSm6yOx$_^3oytP#T>Wn|O`rJ)WBU7EtLv7y zt%gs(I9G!ZcG=n(xbxB-U>#im$r0Mx8lC-&pkigs*6mvRExLuBK0YRnQk)`eX}wdf zYJ&0SY=(QaA1*Ru##>YSg+^7!^5s1L1aOuAez7sayQj7&KZ>g}*x4LpatrG$OQ(y+ z4K=N(5BhceKH3UN8A<$`J;!%s zg%Q8x=ewVsOYAjiP0(~qOon2lo^BQ4z%=X5*|EgKZCF)zcp0I!`@=J8 z%NKBVQFVYJ2HnBXB0N_^kUIPiBP5yR)_@0SQ1^2=oSpuV&}GeCmUMVkRb=Akl9aoJ z*5m2nXwSHWb-u$~h4->Eh=iX?b6Q91C~@gYyB!zxQS7x#Gu7jDZ7sfJSKK`ps`J@- z4hn*VpV23aG883NHl}~=7?-x#X0sNXdL!0zSI_lZcCpoPSS+wlXd&xuSfloo5-$i8 zNKH)*gZjY-{kBjylIgeDD$@im5@T8kuiIDbLIUoi0L^u=Tdj!9k8%7GUA4e0{K->5 zxF^!%&u|YFDd#C7Y_R19N(<9ea8C5h1>e&FlG`G;#frKzWhg#=P{xb+F`;r-XNvpapfAa*epU%# zUi=b;Z~_)@SDqLZFiionYy>?t(k{o>?6eTbc74x5iQ|DL_}Qb(l|2T#{rHV%g1#$M zIBxsKEo*3DcdAHFzID#?1n-pHRD3nP^xR7;i)bapf1Tjmf!H9`-SfLzh2U0V$8g*sLdM{FOKQj$61mqs4mEfv2{ za_%hJs60=#6GT*l{n=>YWa4hKZ+|Fw%@-fGU5TvdtVd*UrR}0FodO5-tVC~oQG1!t zJ!F9VQhS8$T>-}k z;Giq5wtU*L7m$eo1a+-O81AvDy`{91`W*x;{F~5Q=0y zo2HaVc_wDSuGmqu^m$pD{N0g4kOG*`^i~WfmCPt@GNp@00Uo#0>d&Ldsy4l(YDYyjLnp; zJBcq#!AjZCuwW&)eObUb8rTh*sW&N`*!Psg^kWa>KP>d9V*KBnjZp?y;{g#3c=n16 z@k@AhIK%xEqQ#!(+sL=DXpfI=xs!LS$dI0PwNg&;+&HTpxD=fx=3`>wGY`RWXA(er zN(x_UifcJEyfmHHiWRjGLNA#}D@BK=jL!V`1n{|%%rBDZ2j#0mr@X1JX2|WE6 zJ|8$8M=?=}D;uYU@QS;kj$cBC>b}5mXVB3Up5mmamNme!h%%GHD;BRn)U%Kd)*w2d zgTV5$(MoS3OJ0{Dq+64@r>W(;#>L6mv6mlJbioVQo#}gUzL(=?kX~AvF&{VUJ(oE` zhjngO8m9pQQdP~NFQXP&Q~d@dsS*0b|1<~_E zVt@V8yeqOw8g-S%GCsl6f2_ZS*`GTSFv2YH%lUmxGmKz9g2H4leEH-8@S`NJ{uuu7 H`Ro4&qRhp! literal 0 HcmV?d00001 diff --git a/ALCF/notes/assets/cooldown-grid.png b/ALCF/notes/assets/cooldown-grid.png new file mode 100644 index 0000000000000000000000000000000000000000..b78119012d866c6e252de70009c69077dd788b4d GIT binary patch literal 876567 zcmeFZXFyZS_CAa#qM)E4U5Y3mU6kGxl_o_X^neCJk={Fq*Z=|P9aK6YfzTl|m4Ni# zOX$7#8s43A?>Tbry`JCu@&Cb3Y_j*BJ+o%bD$jam{8g1@FB8!c;o;$3mX~{|2K;fx z!@GF>!g=7Ag<6??JUo0(a}8}rZKWroMmAPF&y8&$P##w+Ti`w(p172&?Q>iUJK-K^>nn zxmsCTJBYeUF#pl6C~%GYnU|UAk4GFWB$&08RGDOK?4eA8Ja>5RFiR3KiQ5~Sh^jq& z{O^~6|0I~r935>%d3jx2TzFjW@z~g#^74s@i16OI%X{}OH}C|vgPXPEb60L_hnu)2 z{?*1qsDqKcxvitQjWrXl&F2ssCr1fpX5c>4zZ-WnH~HDRwZp%M4+w%6cZZjc=ML}b z$Dyv~zjz#X=jY?NmqoSJpbj>cPPorYF!S;A^YZ=C)F1bB{^8!wTOez5puczjcn$Xf zQB`wSsHM(Bb1SH|!ynDwy?5_FZ)=~reUG1C@aNkg2dKS_jTP`R(5kwTJrrog%*K(K z7bkSye@h*hKHk6P5BTRV4S-q$({2yUZ~nKNIe2)Nl+0yhROMx4m{jd;Ow28zczAOD zky?b>>J5}hdTKY$U3e(1FsJlU8IS!&Azo)76UU84X2{#$fgpZ8Z&LD5p#)eg7{bbNzkhm>OP6Td1yo6XbAxtNz{q{D+Snv(V`Yrr= zJ1#0VDi^$n!;ppT9S&z(!}~GIhNbB0Gh-f>&z_BC@apgJ;q7>Q)A}~fn{b)Vn%kC8 z%tQOrM~ZA=eYYDmLWJ~^UI#VQF=y$hLx|V#B$@Ee`Pp8%OeRefLsZ%Jc2WPl$*Zzw zS8o$cKV#7*%&wu#yWIUsVhqm=A3k<&k0ybT>MO{W4v}QNc74mYZ|v+VrdzMy-nrsc zdk*<<=_=LhpuHPkXC{0Or z`W^XK$nt5>_uLOs<#ZHpd@N#iF7BoB{epHHv#YwVcw{*{XdQ(H-zQ18rD%2VY>5np zoOyED2$CLiwyv&Jn?EW53b|`3sreT4OhEf=FaGK#G>3Q*Jv#m{!MDJcu=LgUwTJpI z+iGSfo7OlgC~az0#-rB{T^X^l(l@EemM)H-V|Y$>Bk)eRszN5y64Q0>u*KLgNCV za>itE|D-jYCpTuU%wCDUDs*+1EQDOlpVa^O7MXmqvC0O^qwu{{d^L9Sv*tqc)bnIn z@mo5f!wwF>r*Yd4^RPy(n}+8cY{PNjY6 zo-exUURf8ruP-paU@7u-k8^)_prEAG@}xzPTT>~gSQV~jl{~0ZbogSmyQpWmyKKRH zfoOYrL98bPhSBhOl+ux7cBW86_l%VYJlHwRIb=uE@AG)(*RCrsxY{#&&ALld*Iy}r z*6S46%GzRfCpcz0#;?>0|6-oj&@*Mx+LzNSp5q2CFbw=WokIbS%q@Ci5#1{S<RA z#ku^<>a(-|j^J|mASn&?_k^#|3)F!$Lo^vQBcc!KE$EY!gO$r#84^hnCBar;p^;l7 zI^{j(vW}=2>aKtWqZvM}JEV7V`7j}TwBKoN(iS*A`5vQaG&&d5L)=b2A349=gUa4D zNKRwSNrqJ`RLh?r&R9P)y<%~v>si59bH7ofsF7D4QC(D!#GRd}8!yaEsZE9JZR4M# zo>zw=pJs-3oB4Ifrx+Ne+r6|4Ue}#Wsgq#6L@q^+Y|sriVr|nd=OW<@0X2xC7w!r0 zU%m&5u@+dqbNtN2D#}dLvbb*vA<~mzm1JJkThUU{_N^B31+5?G9;g>}O|ckSJni0e zFUlq-2Ob8G(iYH>DC#Yk(V^1m)59P?59|#!=Vc+wAc}?1p_xJJ{7SQTgJna|Ai3B41)Bvya5&MheX1B}LHasGmnzfyU%u>!uftU2p@$~R$>6;^NAc*zj!G`4nDF0zW)QGKVsfwfF#OMsR z;d__NOvhNqr1^O4N?FIIsNOsJ?*?^}i;vcbwsgmO7Pec5zN2amU+ywWl1P?GqCAt7 z^OBh2^1Qyu7~V6iTGhnxNJSZzV|Km4*mSJt3GE5tnVB=+UecdY5iU-+-fG<%D4ij# zGxoVgNZ)2fd}W=9>o$KTSLRJ^F=4(^jy;3-I&zwFUUG(+9&-uplh0r0z0U(Z?{j-B zekM1#k7|WRlJ`jCwfjTQhoD+NIh34^annOb$X=}g`RYTR$MRp|zTD^MbywONp85u!fMvV0$^({eSKhXzpu89K`oo-$ZZK03EJMA`BwCeAvD*lXih%eW@e!|L#Il|BHum# zxPUX?1Fdftxd&C&ZjDM1D(ff{b0FLZ+uWH6-|X5XRV-vF3sdh3%4mxlWiN`3Qb3uw zwrLBcs8HS*x*O)5NYCS77&aHd|mjafI(FgDe)(qGOJZ7VMazeQ&ZtIhbXM75Gl@=Dg**!$oX7jZUa-Ff$J6d`^PfjZP z_H(5w*gaq;#XIzeBeA;;tI3`*`=MLNjqv4+sJ5%gtCyxoFUjrS#g1j*osW=>e}ab{ zzKF+7fOlS8NR18F`p%EwX`(sd)jbt$r?=O+81VROY!3Pj_M9B(UwRFl$$NFM#g7Lu z!@K2rrn|=}l_Yukd|P7UgV&~ht45O0Dbtre>Kdkq^if;H(NZjd??1ht#jJEi6`R=N6Pz*9^oZBBbT)CeENNt5U?kQhc$aUtNvLb3 zMF^8XqccBLVsoH+;R@58A$*y4B#Fx&3N~wp85-aD&fpi^Tyyh2*&Deh);UPVq{=?F ztCIHK0sq|0+nU+^k?@1rYM+@pE;mBoV%)8v1g#Y2JG>Y6j_>g{he?u|#p8RQ6t=ht zaNuvgdWn(R{KhqZ6L|EGHisOBj+Ch>e9KtIcH4%-b{O!_DG*4uttM?bP|}M!zNfioqR~)9ch~qw z^R>*FyHJIs=1AIfiseof-EkhcV>DUywf`-OKc&>|S5wdEbG<7=B?OQ{U8;QeM)I{w zN@CVd%Yy>$J4-a|D>~ZE`B{<~JM&ecH)ZQe?Ivp#<6o(%{A!%t*UZeDAtJAWFTJTj-!QXvazKve3=$zP%91 zYoepgU67A4UvG-ICJ3p0?>U=WIF{yWo)!k^;s2iZXU^y10(4T9N*;$WfVVT9d8_s& znjcZ>+;ZQVySmvXNmu zFwX_!C4@;+tZ+1l0ejv0#h^BWrQtD;6b=s7zLPec@35Zw5qd9L8{(g@!1jN9T#Mjw z5N+h@Xe3#w`|b)oxZJUUytLERq&0yCr+hA>Zft7bvn}R6&8W+g-&uIqK+%SrEH_W_5m@shG&;2`X=E&kJRiYKD%o959Yurm_fR<`CujIaq46t znMt|H)n>uF18?vDLfaO!Kk=LJ0kbJy1~Y9-3MH$!<=nGgJ5T$ICH$`eCxbu9hA?yh zYnNKcTK`FsxX6ioeHIT`{eNo_?%NJBpYECbUFYIR0WC;3kF`|BW!ZJ4=1DM1Cobc+TKgn)Syh zYN*)S0WFzmQVa=DBg9PTrr**k;5fEjJWup%vZK6w!QYG8l#%`idwaA$clIz!rg zd@pb@NQnGbOEHK;^}V|1MgyCOZ!O~SF)OvuKQ!!np8WPXc{<7t)f9CvWv+(O?_}{*) zMbyQZru@q2{$CA$3O)yziVowsUoXgCiSXY-3cA6Bgz?P`-2NxE{3n5C-aO+EcH*Zg z|MlwnwE%DpOB-E~TYB-qO;F(3EfTNQq*!l@YZw}bfYApV`2{8okEs_cgW^=LE))jTB-@no;fflbz084JWMY=P%M)dOgZQ&x+vgwXcJF(y4`B{Wa!OW2 z(Ql3Fs~BA{4l3*7^f`qS0mzQGbP+u*FgDWT$z(puy|w(6)ru|`BLISKT1db_bGSwD zU$~o)p{zfGzC z5FE&ZmC+d?5wq+wAwyxE+c1;kvJBjG7@?98O^8OGqX*?xIuLh;e1zm=zAE(5r1fo~f7-^sbM=$`L6iuODx*xZ_oed zbaN1#kK&>&-BMFT-&N!V57dgBZs(nH#P2SRX#~x8D3+ zS$)-faTWF&-{iVzE?VvE$(YlMT)@VJwAiUAJY8p$jwh!Up6jX^dU@t}Xn60^`_nI+ z?j3-rY@a(%N<%g&xyR%)=DFEIb=+5Cq_sHjPIjtCL976(-eEHAdKy(%WnNpaXRHrb zkpSf&zd%C2{lgffFZ;43?KyVTL~82uQGy*qY2g|IEbU13jC7u?Q+4N0X9z_3$6_wW zp6;*`QFDUGJ%-~|qxA_zUNmQJ1#ND{!3)Qwc0NrV=(M(ZS|nt-sJB^@UO%_t&;l&i z=DEgr^5NuYR*KKE9|f3#yrWp)Ew$8Skfwo%;DBq&T>JCh6Y}M>XhZ)Y>*|{xt$9l# z-;EmIsYB2dT0ah>l8V+J)73!&d30fP0*eOT(~&X!=Jwv)l|GeDeOUw6&YY`0V_#hl z6I}vIyu23SMzPSS);6i5weO|Jt*8m#93@&A1)jx$0#nO7&!jzV`Vv!S$rP`xDiSuV zj>|Fz=~Q>>DcfQEHqb_ur(r4F`MB%0V-{Dr^P+)rk8zD(Wj z6Xhj$*}!XgyL1zQnp`fy1b?=vq?oK%d5@-ArG+8#6}XboUV}vzTALyk@G=w0-f!qx zLsz~eYT-Ov+O$3?fH<;uSh+imEEK|ArzLjL5!RgR<5d=!f30oy`AxuXs^1pYG|5Zw zt#K7Qmf!dXUFeKmY+7{dL6?xG_VYJ;E5Ty5uZVfHoedS0JGD;^`;FFz9olC3>iKug z6S8tK2Q$ST?140CrNbX+?&)rnigeDzS>!g2N>%PHF)kIE@@a|y%!1O*6^{u<9nZ~? z-O-@=u5{UI@1|K+v6Vb8h78Wp{%?N00ck-rVuIGyldTha9$Ryv^~A&!ZQ_UeNAu-U zvnM?1)W!?D_X!!Uz;cUUjHYg2S4L3vI~c6y_>Sv1nrbq=diXkT<(B;ihI_2#JjL2u z^Ic-j%~izgqtVWpoZRznLZe%Ku1m;zqdnB>%(D8DGn^~EVzE{15m2X8j2HX4-v58wSB>B3BBlg|? zx9xOE!)EN9&yL9KVJo#AC%@%D@ZC2QWB%kC$GDPax})Z(A|6Nt`^u zhU34iXLPZ&M78uaA==>KI7$ow6_`v2V{g$@aq0S&acZ)LruJvU1jlb}8^*ah+i-!| zgCs6jx97dpd^%%Ug%ycq>9wecm(dD?S<63ek_1j4?6kBU`<*s2r>&w{Yz=qG2*a$Q z?Ngi+UwT{k{@H;%%d(l%5}$0nrr|(-juJdP;={(fCd7}hM_4j;c35G>!GcAoH*=GP z{6$vw29uNH9V`3#9FePRF@j$1;>XPeXzMg;83$;0W{2H+S~MDS;%x)sny%b^{O~}2 zJvY5E$z#bNsy`C)5Enck_jiW8#HuCjI#Y#b`jhK`=tQd`YM~SVU79?2wOWb?fx;eG z-v2rTQJ|d7;$n}A^j@i4^(|d47Mti+SN=|gn$@*v~Zk8u$ z$*2=&m4t|GpQaWePSj|(yhVMdS~JWM>#=LVsD=C1*A-uh1IQ2E6^PtQwL33$nlkm_O^?c;6;^J4oceSLuVoZqA zg*cO;^hg>#3&hFh+D$uM6Lp|tO{@edJ4aZ_eQT#81nlOi2;^@H+DOoe!H|kgWD0NQ zI5d(OH|ia)a-pS^h1aKq$5Yjr(e`ni=J(5_3=T80QxT0GYRn3{uGX`w%M3uEZ0II3 zqmeMS&xOqXXe)Sfv{^qm(E3m*lE2rFWZw=jj+pqM5Pl~g#RZJd$+6t}7e7*w9Ih;w z?JCE>3ZOccF2m;0r7LYh-A@)c3OPN~&n2EE5iNXVm+4Q=HiN8^(sy@Dx~1Xzc-mBf zcF45)_$CWpk2+1#@?DqxNUdpPOJgdc(>wiKUGni!t?P7;mLkl8q0L`gcREn*ivdTq z$YBD|X$=gpxV62@raPav+~pXe;brNkPh=s`WOV|KuFWpm{5Yvp*v3WU*ONsefFFx} z7UnZHOAVowr`TqZ_|8k&*X)#kIzQ>}7gJNC3e~mXVVYu3I4eLaVFK2}rJihA7KjS* zgKfR4UikVZm$X({=jcSBOAszfAe~DESEw=Ns&>8z!VK6QeluU+4CBI!z-LTjA!zK? zH`?|TsJ@1$0HdK)0O&*$haSr`^95yL_`Ge^Yx535)OS9W%_J`lF1rlJ5>V7@iuAgh z(Jr!QB!=c5^o6CSZhg2W7uyr?v1oK#XR-5$(W%)Ew5jK6@Q#FjW2*3*+rblgJ-`q= za$GLg_GHt$U!Z9YFH^Q=&w_1upe&Omk;C?0Hwi=x0WS*~6WE4s_h<%P7K(Ae_->KB z`J3MS%yG zTpc$JZ?Cloy^CTYv5H<6?1m3bKttlf?Oqk8sDDmPMMMa9X^0$uFq}bAMH;B=da?|a z99lq-s5 zf;}733O>jrI;^b5>r3;tJU4GPaW7ufsf+|H19u?fF|Yg!O)o{GGVI&wPdH-W$X=^#CwTg$}BaP zyS5Ac^aH$#9s5e6#MSBk%!z!9fe0G5+>&+m>d&O%4jsPvqpyWw5U7qzVv!MsafKX~ zTcw-yFAc){8fB=>z>_)Jln%ImZ;nM9Nv<9?I*81-vj`B)7ovHa@KN}RfEh%7ijrX zBW$u&Y&Af_8N-e9y%R<0BuU;g1CWglfV&sLw!`1Ev)LQXY6dUZ(M9%X8n%#i^}hcs zg9%XNsVwB;3>qM$XldgSfmD(@naxgOzq@N*4Jp2}N6xneFG*R= z_X==w*ccneBpXI{_oT>%2yd)0${wt6MG%}Ac<)D+uBTVm%AD+(>S*zl?}ha*DaJ>& z7=FGFgrJi;`4H%uU9QCLdq2|Q>X@}?{7_P<4YLHFuaEfH_4qbFbSYa2=7Hu?RQ?q8 zX}gU}VdwD8@1p?A5KJ#tcY}~Vrrs>dz&*?LUZAh|Uc1ypHY}H2aY3H%qg6qvEL>ap zg`rNX$e|Fj;V~TrEaOQcZOO2`o_?%^%Ny_GiBD&}H7YcPV*vfU?{}WUfc)(FEnWsK zhSXa{IA4&zu)0(#bKRl)D!Oz~sCb06DvrgC-g~O!Q3`>mORz<58Am6?Kg;olemtl7 z4OFz@zT<8tb}^Sq&!!Tz^<%;;<#A8|g_b_j^>7q};amZNLphX1T7Q4=zyr@kc!_dO z9E%T*pBn;C6%pQbnT=$5nTDqA(iS}2hz@AukZ=W@{>13v3NO4dMBUPMm2oMOw>O6D zqk)H)Z55r~v`fZ99q*?vAG;&N+JQ+A?>9@iyrDh%BlPCvpcGiESlf8Pz+QABPsX7{ zcO+3AS*W`zDxp}+5iwdB-7s6Un%`9S+A>MZB#^H)`&wAJn!RCwe7c)*!!!1}-M)Am zc9aAfvw|Ixa03UB+LWPZkm#9s^xY<|M6^e8kRmUtVhI&aJL=kHXcVNb#p@A#3U^4u zFTm#ZcJ>CJPXF5L&helovZ)0+UX>;)239E+f--#jb9G&S~jCfAgSg!gR0$<|u@4yg?AF z5MHhZ7L~%X)lVIpncH_wo|OVWwF8o`SWbC~!)27pQZ!35MPnl;JcgwLPA|4nXu;Qz zW6SMjUD5GY=UWH6n=M?m8bE|<89a`eRHit()7ttBeY_C2oFxwWjqEVTq!Cr3Dw3$R zBg~^*gUWzyb96%?Y;d6Z7(8shXoWri*pstP)!1nobh%@DoI_$*qk`fr*M);)2^U^g zr!)(8rXsiEFJpBBE{7v7m?o_QuF86VlmAz=^0y0B?mH$+{g!++ij6#}U|6f?5wd1Jyy>h(jZ;yc&FHn;VNF69+eYt7=?#BqFUgL&EvQ zAnAIP^Ofu(x{ibZ@+UEOZWuIBa&tvoK7Z-Mdk7_F z4GxlU9unwuGne<5z`<~GVu-}`K*LCbfI~*tJQIjkoS%b_i}%1$I$d=XOLXGi=HyM5 zqqMLYQS_$T5@*orl0$3D;^I6bds4(fLfA493wlE>*=5vempYWijY7d*r6G|01|q*X z3kU){h-SO6K7@tK1!ac~e6l@9E??SupP8Zv?LCpq-m zGy=;pp1;bb_d+9uW*5v;*&^A{4&4I^5x=8%xqv?l*3Lw5b+d=4;2G3>f3_M+&kMR` zSTv)r(&GGulRFaA)@hNL{}Q4YenxTC;a5pL?Ng@Zs&MCJc5cjGogfqlRQsy1Meq&! zj|z)wubwEZWO1oyM4fo~9BiGiaD$2!XU%0{)+3a;PV9laky05&6X#SUw;hw_!VKcb ziZpeR?j7GAyfbK|l;z^)gkF5WPLr;wW13jhlx^txS)4s*2|~{Hj=^5F08p|a=Y!T@ zyJu0uARwGAO_D4*KInstzoFuQXY0O4+OE%=CUNUsyW5j7Qls_+`LP>h7BATqtUbTK zt*MsGTRHU&HwTPe!kx;Rk0^0}4nSdb9r zLx>c%)bfL#HPUTMA0YmhcZP4hJ1!K}GnW^;Be8q5nFiWlF@RDGH01g?XyTlAWKT;{ zMim$GVMx9pknV81PSrOhp`<2)Vz{_D=3RidEtIn@R| z*+FkCyxezru2V|Mw!Zl!cb9Rajt=dT?#C}>*zk&9DY8&u)+L|mIn)inWTHU8+~-Cv z^JwX+)$Gg}g;?C3(;n>orEzI5tQq_2*Te%vBdg054m#lJh85JZnhr&XX6;2aKLJi` z)d>)U)aAB`ib(=9bn6BKZsVzUts@qy4-@I*l&*6Q>!~D zwlyvD6^kHlMPyMk#$52M)-?KF5~-zj*2w^+5v@!^ryfvTLhh&lB#b?aZ`oKAmU^p{ z#e-YNnSJ?zkljS>-@tVRf?Lj!FNA->9MYdY1860?O1r$DO+zcFXi$UIlDlB56RAsi z?l!kkU2u>esRS2^@LYT8VN$j1AYTp(bP-<6VG-qw6?RRv%4UTP`9BTDi1NhMq~E zWbaCXZ#Lu>&A5jL34SR2u5a$PvQGuc0J{W|D~qyV5Bn$46atDvr@MX10s$T(jw$ClCx@odU)r&LV;3WqydK|1#%fbED^{aKOttch~kK^s^#Uc(M ze|t6tpvJi{3SLfwW|DI5QTP_W+K03(H8;3Q@&W-iu3V_cXVnAMd{?up`QS*62Oso}iK zQnK!HKgU~hos3%TCK6UQQcg6I*~7|W1&T5}*`Px1P4rw;Y?Iv8Pw=keoutEz-tczl z65*`0%YLqiv)4LyjyF% z)uQv6b=(ex*>__^V9`zs-FBKvGe5S!%DmN(C^2nIcX8$9NA#Fj2CY7^XmKK9PIg{s z0CAG1(=RZ%u6-82Q#R@1;?IdFXO*Ous%lsIi8lPre*?7cd-&#kiH!uiFVI08jG@lO z3!%r$RfVb513qi__c3!qK%9;XZ8)r=7**oN>d1Dcpm?Mfsutx&zBbXyqD!oHpGb z!L<$UW^p-)MrrO>Iab+5Z=gvmiuT2wit0v(H$?$S4MY2W=7t&<<(9IUpPt%zA)RvJ z5EAp&tObA`{UUOW z@${YLp?VS1V^-&b6&Gd#EJqVtvM1Sz+g_;f(T|kLnB?n>Hy;P*SgJWm=J6vfYN(R? zK|NnZ;hB-I`@&YA5cO)dju1{tWzs?|TDR*45MZhV7wgoK$rq70^b#6v7`l4oY)XlT zpzwq_C5!$d=4#Kxx9RRr%#sb+cnwdtR2$CB@7DEPt~uTfGzt;npwzeh+Bmbn-;`TH zR(;fC1IcAPy3+}Hr4-p4OpZep*FjBO1&&;0#+wa_0O$*DmO42?d*C*n33Kc1FR;(C_RBRo33RioOXg(5a^x=R zM5bj$_Ti8^;gj9jI^H&6?p^~6c;nEct_g#F>eA033^x$p zbOK8FRklnDq3rF~y2~@U$O4ss5l@$T`iZxTh@cUdA%VETAP%$#iRe_`ZKcv|PaBK6 ziwabiho9BFY}5iJ@>X`M*=_hs@#^iZWMw4#k2qsEl_j?eg%=YXjB+b|Vz?i;KnE&^ zY^uSjsf8j>JA;khU3YgCKlv624Nr-F#|NMyY7u)45E%LhPLC{iUe+DR zSvum>vkoPEg>~JgK+2$r-n@CszP1B^!?zQ9kMym2KodYJ>_23)uXtou&G>lNdL(-= zEfoQPh-x6O&lCP(Bjr(s-euaJ>obQ2xH%v?j>?5lM=pB`!0vvj+kLT)2o&(pvuJj5 zzdzfWXjB4r@sGDId`)d@`>BNu$u>@~{z=CqAF#+R5I+v)4y!_i8KY zzKVCbQ6ErJgdM2Cn9=n+$K#{W=sK^p`@buCEqTYsD zVDoA!B9y`|aTUrJan0p$o44&f*!jS^O4UdmmDaPM2v2N!d~$*xSTtC&U0z0`Xt5Yo zC;O)dX&1)rYBEIxMf@CyFOjdej+#2u4Jf~Hz_MF0-l%e=WXVd{U{AK({gPw`k0-;< zn@Z~~HkNulQ>#AJCD{ngN4qN{Mi{T1V7>g=BVpnN(8}2Ml0%!Uu2Y_7><9&gg+$8J z#VErnQRHG`l&9&7!ub&t6cgKauP<$rALF@+W~H;aHk+^=zt<(8gkdxbnXK2Gp8@t~G6CURLQw`U-gPwk`c_0{S&=Rs-=mk((6W9e! zwrb0Y@wZANY6X(ZYfckwAt5*`lfP&lS%Wmf@WL!sZo!Gl>a9_A9kTRKdzR9o4G`J! zd`6GOV+jyM(3+Rq#P1~anA)M^Z*Iar>gb-A*w+Y`cYNbqg7KCe&)X}ri!v4vgF4Ta zmzid)Vjoc@P({MzdJkIpcdwsvU4Jt-(kT};>Y~$)ktv4^u6a6d^=zsO77iX4(C8_U z?ro&U*FBm|*4oaYGDHMl1PO3t#f%7lB|l3@v1;HH8JVIO&4b>YAyF3J?}q4eEIT{z z(wCs!st>KMJ57z-O^v@p+q6mE!@75E+qDL$ocZ%CfUe7Z=w6-XoUGd1nbD3;3tSCQ zr|X)FyqKfcs2V47vFea;r+Dv?K|gmX}5GZ z^zQQ#geo-Ujeno~?1P0BW-smaa@}WxXKU;B>UhAj&&4`QEMo=CJEuC+LG_C{mCm3M zV7WCebnfRklPP*-az(bu_ZfQ5r{OlCj}{tH=IOd5I&s)ov~Lb_KJ5LLd8yDF*u-cz zl5CUN!=LN1-P89+;SH69pJX^_9G+k9*cs<=4CD@i@u|YNq5@uO1m5lE>=k; zIH%Z@>OYN4bA$^_o;#~ZhB|+VTJ>OoH3ZlWOpgm1Cbq3g23RpxmVRfBH(zIiT?W_% z8z{djqMTZym+=KNnsC{VL8!B$A~H;u1-;t|`M}L?kyE7$DgdG>4Bw0Hx$J`nXs=zN zHDYeJ*6=BNj{}LD_AwXaN@PG_7qm|V@ z&h1SG!?UFT+`A4})dcy}gI((Psq{58lfzPi0rUduG;fT9rl+3P*C9}=S9gU96ehtg zC-9|F0xJDG2O78Hj~Dt(>ykW{>7&7RpeHcC;H{Tps~kSc*@~R3no(@$C3SLi9fEpr z)TO}L7J9DYeh4V~9VwjwNGMm6t6Y2Eg*C{if9<)zJBPLVgb$}(mZJ1*0u_0yzhocd z;u+cNpyOlVt&e^XU~hlc3ydxRUp06;)_h(6eoXSd`<*Uyn2=G!tBsSq^2JC-@aft6 zDU%C@0=@(ObTD^ zH-{(jXnRHvz{}~cuq8Mue4r64UU1KL-RqDw8ZNbi_Hmq4VdCsw{b9VkR`w?j1*!AA z*GtR%WEqZa0GdZZ|5Tdg`d8lkt)(8=(zdy^LOrN-Z*VcUepG6h7Z>2F_AHqyXy_Sj z_cRTr-l~=%6=NRtTCL1KDAXRa@$oS@n#n0>S{NE^NcfOeB)~cOQ%ygf6Mgn`l>s5W z9yze#p8@I{cI~0qWOQ{vuJ<%2av-`|LvVPz7c5kD47V}yf9gWDsQz-Hbkc_b;3_=M zDVvFNQ|ouosR7IbNM~ozAw(Ns@iZ=QF*zt-^4T4!z}AMV1kqU^DdtpT%aIG&u)2IN z_xi)tkihDZ&0Wu;+Rpf{R=qSB*=W@Am7R7eY^+cj;JYW1mXd;m25xIMB^7iAa2lg* zl!AQy&OcJ%;@}2T8sHcZxrH2348pe0kk-PfBW zVooB9H#a43#oB{gDFAmYvPjQn#0i3064TPrwLo+=CbK1YDqt&>V0J5qDQAtklmvqX z@*h@R>tBv)na+FuJs(_$$XysJ*TMYiWQ%q)`~q?|IYFu81|+h6cbDf-?9hkNL)>}5 zDnt&MQ3%si*m2)z$e8e3pKKED`eq|*Z1=V?Y%tm7>l?T2v!s&OLL{)Rp8HLPt&j(> zsla_Ffc<0Owp>)o5h$w>buIpO1Q52TPGvmExyemkIzwx5%>Yo80NIGB!psSs*#tWF zt1pn2Un@xlT#3?vgH$>M^o(uS(I}H`hnbVkew+PiqFhdI2B&6~3B`D8GCEVScR_74 z>_GKoKwPk>9kN~*;K#VU`yH3TG`+w8G_^lcoJXOjCtY>coSLhrlm?CwU(>b#tX`~XGUwvg%3X;oGNY~q97s+9(WQ33 zO=+e1Y&S-qBfF%Ca59HD4$j;{$kMT2Wj0TS3&!Q-EOeWlJx3FwZJvUwAR%QLK{;^% zODRbhl=Fi`jnSr6)1pGenj4M~l^ma~mmO5fDNF^|8hFl6LqJ3cMTMzMx(ciw01+jy zB(>-KiR*ImMh_dO)+;W-I9Zb3MzTE)#e4-R^XV3Tp)}}Ft`9eMzs+abUq1~vS+L^j zd`4DS(q@-sbSduO00HxsvMNbMX3yEg=jhoC;v=Vf4OpoL|$tx8XpdW9AUWz6ZWcz3Hdkno7_4d zI1kOUQ}^Wi5P{G@BI~O-8W6lZZ%k0fayHHa!L3-Rh%E3}MM<4xEvc7S2iEr^SG6?z z#4)yhXZSvy91K7{s6>HJfCED&t3xEBP9nfRl+~Iux!K%`oVnngZ9f@>zQwoh80yh} zK%Ndwm@ayn*3&mSS8`-lZ1)a@k`;3(5+&$+v_v9DN~%F_Su<`#<;!LDD|Q2E(GN2M zrm`Gm2Avj3sSIOsKWT5C5#XRQaQ!U2JwD>1_BDe`svJ70r2arbM#Rqi;z^;X)3`dpocT z2n@ZiE(-y36(?A4ThE_$TTVceL?S7;=00QnhIyiuMuBE^84o9-7 zSKvfwXCe-seEnP)Hqjz95<`gLpxavL;Z&B|o)UMBcomREsuamWV`t-&ka)YWLa{Jp8hyP9yfBMfhtespR`U14jL& zy7R|}R3UoyADxj3c42M0Jf&-2a5Mz3-XYk^&;iWQKagk7E0w;FLvgm*C*`uxdsuEM zmv;;cm^&7&qy|=0)V;LvdNL$%GFSqrpD$`U=>^=h0!n}byJG?UGZuRF`Ci#ZY9OJO zNeVcV8rlIusrymc{1Bcs7W zbGIunmtq->RcI6<_KJqP&HVaQL-^#tY*GktZWH>oKY5C28_Nd$CpA0{xM#oq{^*KF zUAtvpzDBUbOy-_-alwEoI}QT{NSYG3Fh@MI2q1nK?N3J?a)o50{(z%g=clzGa>c|t zM4mi-E(6+pip}wnRJgq7#$}zLN~0RSJx=db!N`eiY_m=xj)WfYL3R7M9Ck zj^GHNl57)Gf5M7p?<&>Qahulm+U}8bNLb(8N=aI=TGrN;2vRcq>7<(x(&o(%Abm>< z=BLln@^w>vHkJ7lc^oW8;em~&!|FQ-e1n|*4?_5-po={0s5lCS102H{OUMw$o353q z2h=wwvN?IU{|@O$$6vV}q~mBte)H~y(|UWwO!L{;C$_-8X#M=?RK%CJzdQ{J0OfS? zRW4oE$YB8J$?L9~gg8PqUlwQ4h-Cq9+g;#7&#R~hG9c*{fbvW$XOJk*2zkqCLqbn$ zVJrEWly1I(w__f>ixOVgWPKGN@w-zpTa)v z9RI+37Nno`=Z*6L$B02x1m`f-uHrBtFYvO|RC-~M3`8~+_bjysiY^P9xjoPs2-@6h z@FP9G$x}U%tx1QLXFKAm0?ZnGN1T-TCxs56x6fEq%DSCqv!A=rol?^rd(UX>lZuvx zWjCI-L~D@b=`R0?KK-kmg{NcdzGnkR769HAE@41#>da{kOU)~tr z5YASQtUu<*C5jc@08zW&HRW>P?B6y?v{Mo+_4iPk_MD0fuP^~b`Z_wBIv_)6vH(cY z00ZAJcPd-vZ&Q0JYQ%MJ8viVzce&ZKlYxFXrv&u-nX}C3)XoXNEKdM3S^|iQI95;) z4)}Bjv@4LO*z8H1F#F>m26|i=J?ykA!3PkWE4&Wsag_gzM0+(Z0Axj%KNi(?2oivl ztS%3}kOBBV)4Iz#45zj(NG}or-V&^$a+;+mJ#rDq^8`u5(t>eSEgbt`Wc%s6f0}67 zPdKDyuT#_TxY5$d5{SbA{tqKHmARdoyvHM;pSV++s>m$G3nC;DLh8f{_^aqKj!i>#~;S*rXgU==rn5b{(`+OG=qJI zR!Z;jG>4&BmE*d&|}(!HFyHyE7)(#@Ju?#CtE)bE~+!DKO^& zceQ@6bG)`=$D;+;hkxfX^iqB7?>R_MI#q7_%kKQ|3tEFu>KV$&9nL&nl6xJmAg75WPf|I_6E?^b~^-iamHjT--%|L~3H?8z3%(>)_8R8}^zc`Mu2V32?>#TH!AKtL%uDF)MhSYG?s)(QE6U?`ftQjBqV<7?hH_*xI4w&t++cBcXxMp_~<#`xzGK3?~~uNCp$Z9KPo?CSNR`7d?a~9IUe8o?R}y7VPz)2}N*A$`5Bu)teS*eCxb* zXuDedH*uDd)qOJwpX;>sd8LG<&ff9xeBsr@@E(rJX!_^GgTt)W_c}A7j5Q+tfuHX5 zzPRvzlQ{Ulf5DHbPix-`oG5>JW@;#D8Vp0Q?Fp0azC=n-T&B)mF?~9VIHWD15YAi$ z$~9VEjLD>T#wDY0oXSu!5sTcw1NR>Xg)oQg3d>4A|;!s)L?3< z)nHizVrm1+=l9cCp+R>U{TF_^E%2`0kj~48B!;<9g30RbNE2*TTdeM@F#j~!qX29% zRQOu22?oo{!@j(|nD&V6Kercje$|E^@&gWd+>ibLw)aG|L6_UV?(kX$t-(||sCQ}Or4o4}4_XFlTMKOMh5sqM2} z*@j{0lD}JH9f=N*lLv?T`!gb}!yKNSqo%p45O<2;@EMJig~%E_)2y3W0jGSs84(xn??yegG1vbe3C;x(qtZ^#X87L9i%~DQ zyCwhqURYk{zbg1Du#Pjdmt=SQIqNPCgIFM9><4zs(M61~X(&l?d(&WQ+xl>79S7c zJXm*&^7|@6V#hg6rTLLFkFpZNcU22a2?+@icP&C4$4>Tvkx@cHs^kqQ$Cyx0z2Tf;Nqr&B|9iTza2f;huTnEAOOzlFvMGYwq>JXFE%Y%p# z3Lr+CB0tOrpM5*jC#_LB6%9-!?Jk?T%4BYnPX$7Xq=`|sqHV*ovvwfq| z^ZoZ?s+jXiHv}z>J_5CO@Jd*}G-Y%?>aaV|FlMwksUo zKAkgE8uYA6BKx<8^l1VUnq0EZ_NUfQZ>-lkL+# zaA4g%0#?Tpvi^<*P1~*l7B8!-NCA6e?AW_?m9gl*-aRfd}% zr|=VJI6>6$nhradXEbDs*5D{&WqfAkgv7~66v&t&g!g?cX?Fza3rGE*&Kj0B;v|86 zekuVJo92rqkq3S;J>W_Z%6L^H!7*%Vz37)1LS;Ntk;7#UTRom%Dv`t}Smd}cDre~@ za}xa#YGzLjgDeWC)NUfY-}IAc)qAdhG>~0Lq`sw}*iv)nGQJRsvWisn`;zKbEOX#2hutQ` zSd9etib0hp);o7HZ(F^XAcj*0e7Nj;#OwSXYF;KixTCt>L*}t;t&S zi60t4MdJUnV_s;*gNmBCek}kl@~bXYzMplg7xClrp57!MpZ9(2u6$jG8=s$nD*9o5 zzRr0Uo^qxh4ftGixOzimS4-*-NZj)(RuujM{!Go%f|kumpMyDUc3FNSwS%eUQ;w0bNV6Hnx8)xUPjp+y=BihC}YE0CFxs1XI`-}T2^i%?)P4uO}&2IC3kH9BG zXkcP4blU}8lbn6eEKd^WEL7&$$L>SPpAt}f+D0H(u-ur0vbiS}7FOeV%ZY!LMYK5j z(-_!GwR~~$_)lA^L}A%wsujkNWS@omDM*hl_QhCYyKwSV-}R7e-;jAVi5i~Qw`-pU>B^YzYJN2LD!XDe|% zH4cO3W#;|8N06KEg$}dtk0?mvBPZ*rrxcXaM&L-&G=H$@R_)TapT7knvsX77qa%$n z$(mf&kamY$@CzAa@?cG>q2`^PIQyw4BUY!G2*Cxg0)+g;n|X1g1$!7G8mXT`EMRk~2B}#>Wjb5`u$wMO zE3^vkmWrb4th5?tf67Y#uvh7tiD`~cwfR+Q=Jr;_)Sy2Q74~p)A3N&cN6&1?DB0X= z9^s)v3Yd`&=~N6qLT>!z2m9yEh&D=;19yQ3Z@8Aoh->)iysr$9v{ekwTE6ux^~LpX zIcQ-75gNQ0({~+6Bb6}LqM$?Nk2&N*V!e;#f;e?*_ctoxjDG5@ZLPR>QGbZ5Z|rbv zSrH!h?iRkB2i^Giiq!?ZTmCZ_!6epX=YhSjgc!{8Yr2qFeeeDnU0BWIDA?7mdu?5N z>>2&_Vxm2(JSX8`-A%lb^(S4?qAj=RUcRkw>jnC&e4}ud0+fU&qF>SSx>-3SBC;Lk zW!70b^gJeK^XyMI1lreE>_A>%J@*`ZJ5y?1Rvb9%;R9EmLKrs01vSJ!9pB&E1sMb- zL<;&P2nmJV0p=upIL}Z8E#E=RJB+DH6t+`GGG608`xjG$DSc0xZ#y<}`ZY!`^Fj?@ zHEWXh7-|-PEN!>2QHiCAdLCG9p@nZ6UiYh5dLDbayN020xpU=>$cR@kXkfX*Os|dj zxz6z!m?J$OB%AgJ*ug?i=~TIMRQARe;_qX~r<@E_oHKL51^*CapDOT%B?DHw1|~ZW zk*=i(+bDEP!)F)#XXwy|D!IzP)kpl!&ARa#QkV&wGqv-Nme=3f^iIQujx}IU78%Xm~KiBBzk( z#Xg4n&yz=&hJ>fF89?@0Zepzex9K823M5g`WT# zfZp@P zzpL#xlK+DOw?=3;lFP;bDeH!66O8vtCys@UD2$t|*+s||%iFm8MFb5x-Xt^5>s|8d)9K;s2d+!D?s@T~r> zQV<8hoUuv05ycSGoiP89Um4d8Lz#)sA; z5pqM^+}d3CyGr=E;^ZYp)$-ISS*q)%NnhG#22-f|ao6+qC6c@51%rs&wzmx{%Te@# z{FN&et;cR9UL69o&bH45|4s``fE4{fOIa&|Q%3KXllN<32odpZn)5al;%L=I>w``o ze|#lSe9t1=_n0>{f`QXzZ!P_AUqm572{T`*#~Uu&$}Fh)x>>^`hR7X6Pj68b|In7b?Ew}fT|EShNlmI(Z#0} z1%9oEe~R>MeNw&{+_<`}dp*ouU0p!oi$D}^W|-Kmwzr`qzr}P{9QARli}!7Q4e5tF z`+K!$6iAgGL5jEh=K1M5%Zms4L|i>R6{LdXlzDqRG5^(ER681A1%K~;2o3#wnqYP8 za{h`^Bj?(-mvCgb@WoI+MCWZADr#yR^Zu#tO;~J?pfm1o$K1k$JUH7j=s=?cne)SpHXnD?vP!3|+vN zk2Rs$_hV`A8J2R`sgUQ-x^y<{kBn(935k(<75N&vl+eIa^r+&UqhQbcT3o$)_<`X_&S zS}IiqpS`9HBfVp*T#ApYMK)GH`1i+`h-jI8m1H3-oM4(Rs^(_=LfKy-k4HYLe}#2> zq7v*JM&$xVafyz>QT`33Hfv3&LioXV#%oeg^FK%Y8mYs!KM0YnG1!jyb_R^T(ZGu$?1+G@4v5+9A)aH$H zhVv(Ku@i1;-nY>_^YbOs)K1@tXTiQgUHcBU3G~9QyOCPTGBUJ(`HeEd`CXqB4Q=g` zMq_|)Q*})BM)R-@91i%9GJA_VVHq?6-%)5?;Ty@o6^TE86C2u&CV%}+dU|{YKGMcB zEboDg**99&z_neUYVU1rWzzEyhfBo72$q`Z5!$F61nmi#FY zsUCi|)c2eNObd8j#BY3leCn*dFrL+`ccjnNaZ652#N_^F`ud1Be(nF}3kWNpxVpfV zYCN>J{jBi+uFrU0P}`MWf|mS+M-3}99q6H*Cg`Nw***F=98s6TG3su+DI66&RyQRW zC)Idr=3Nx$p`#O0fQfo7m_|lFnPjHdigAXsa%y5F5SLueD;XVNF88jx!t?o+&tPJMttS0I#wKd<| zYe@=z(`?OmZ^1KukcO#}$g-Z%-#Ow*YcxK$ny+t+{hID3&z|P>Be0&px1s#W5aao2 z?*3W(Y2!SCx!vBjmjaWk2I^Y6r%*Xyr!&0;wKEVe2Q-kpI@{fF2Rt}7ud1~0rj(r` zvoE~!=Oj2!El5djHGP^8)TrxFKP|(jj!ww@2@`v*IAsw;rWXyU((-{EcMVn(9u9rW|a&Gyo>~i|gURLo=++7-NAPYdc^K)C616z)AgW=1Z@l$J(2jtx(~u|3ZR2wf5!0J@(D{+3ZD^JakMR!=-0q^1ht3lSH@?drQ!>j7`@<|qu9&U`iO|Dsp&q<_0g z(l>CDLI)AzSA~gqMze;YFA71&rO2v~-k|OIfZxd4ON*Ixtynp12jp0tpEQFXadEUM zh&&KSg0XGQt-Ie&zDy*<)1QwYP}Yl#25fvwjrLJ-K;lW49~k-zX-Tshf4{pE!S233 zwtM4e!mYDe?**d0O!w!lUEdziLQ%@!@B0ThVfO^)Mb(I6n+t}lmABgy(k0P%?7|&1 zl$RVd;hqw6Wn|$o2ZgHCz2wo*0;)6I93sXJpGst}Mn*=BOroD1n(}eK;FrgnogQ~0 zd0s-ks<8ti)KuB?%o{qGFY_gGMUuZHb42Im`3%W~P@q2TA07Go2SpA6o5gK49DVFV zNNl`z`F;{XWE2`1aS{01U zQfScO#<@SZM3!~K+JwoA=oQpzUfwYcu1*|%TRDm*8*nAHUwZNtRux}FE>myYZv4CW z-9#nIFCn-9AxA@Vi7`{e*66WIL$P3>JraKkq=qe6hbPYjYasUUrGVqEU@zA%T-)kE$~)ly>~M9P*Nh>D+@-F zR6(I5+yd&};WMEa7LQFN_->yNs}=hwqv2Xe{h>MTmt&qK4$Z1XM=k_e@3ISObx^Ps zBA}Uoj8t-P^@w^8Aq71TH!QcG&Ym4%kF$gI;ys!TAG0?k&nQT7W6v7r9BhAOYf#ze z>JO=b)~B~>mLTAGP>W<}^|)f-ojDyjs)S{6zv&;;rJIejAA*E)i07-rQp9a8;@pcn zgnZGGq0m|vtE1v$Ari?S2YO9ac1<06u1=1fM4Vz#4wp#q7ZAd)7d{eK5c+~PP8xFA zV$w#*$zoIh5%Js9?xe`X228VgdwP+U*q5qwk|6KwVD*nW&7doe`S7E-Q$akB(>!w8 z%UrXLx|~ejx$a4|ry91XJ66m^Q0_sJghPr$x*3-zYP#O#43> z;EmX_fKRnJOz!*U+{_Bob&KI>%Img+%k?Q%ZR_q=t$Oa7h?50KHMRmZ?uP1i+_Gra z4_!pOOsS@c2s85vaH_FtGO5ep?PF__+sHL6Q>S3-NK7+fQ3AdG2+~On`ft_(2ZHY+ zM+r*>b|n2I>0VXAW7fv%pVBP&P)PFS2NXsFs{dk2h09$4qXRcZc22#yfuSjx`8lf$ z1=-WU0foU%m*sVsn3aGt!{bCcCT1AS+RLWjhqt`25i-*oA8?3ci8&2>K9+PNy4Rsr zV|o%XXt^(TSk)Uef0LR<;Kd}QbvrCI7k*&6T)jMo3?-#vF*MJ??g>y8s{$0A;#?OgBz z>UiDY>lUHsFevyZ^uZYF9nuS;c;1^N+wPk4LI29b8 z=8ydLA0bJ(MLfI8aZ~H279A!qrV2Pj?aGm7u;I_2PFRV1d7(8T+-2YG(CBVCj?{#3 zgLrZL80@6e>XfvGB%_UzgoHyB z_~S=d+=mLL#}iSH0O$L=+<^=-G1(65KajE}*kjJRea!~E>0XJ<a?aG(7Ut;S8IK zEJc@IvBP!oyEod__)e%hKd)|v5N-}W66~2xkIt;1$k$$>o7-JgEyJR1n&0slsbsbK z_EAgD#&$i?ejr(6iWyL!{TL^($UY6@cE5_T>T82!(|;-v8M z@s)4JZV zuT|W$QIRSk*cs6mV#AuwVf2#~W?N4W|NfvqrLoL!Memmf8Be%8umJVW;2SfZG5g!| zw8&q~n6Ioq)PTgDbV_jJ-mbS2g&D{W3pN)#o?$y&og$Yg+*@|xFkxnxW}w1r~K{ac~{TT?gl*HQ9f~b=5p)WFWzJ>pA!o?*pdtLbzkIA z-1xmOOsR_xXr{|nKZ{t2V)YCDFSq!wKvaRzbs*W#k(6=}3bJGTvy{459DX66aNm;y^Lm%YPWc1hIr0dGk(f+ zNJxut3cER`j%TNv|F*b6UE-Q+=-2kc>KWE5M^OWnLY^O(t50@R!}|8Oq8imw@E#0h zQcz6=b8S7<<;|D&*OyzN)E8Y070|wkHVy75+k3kmaH)Sf46pem3|Qz#k~y)Mb^c&; zTcfUGPlZ6*%9dPw)k=I#WJhnIBc$Cx`H|VQjFdnqOnk+N5AG0MT{C=C+^@3bBs(Nx zTbxPfYgHH`&3>MQ0I5pnmTl8=A?0?)#! zLK>JuK+q@r$QI`|EQvB+2M6FKeFz!JHpY%><0qQL-3W3PZU-fqdlsO&GyFt(G$P{t zXlzWrH`zrYU?$-gRJ&$i4h;;^TD61#5^ZaQX1*4fYsS&SfZZt1ze5n`Z)}V);%Ec>lAN0b{5QL*mV|3BLto zxBkd3M~36e>kgysowGlR1!;G}t;!)vib-6`&o zG%t3{8g-6%6Aozk8YfUFUC2wYRrw|QBwJ)yswv>LE%oi!i9|;XENSNcf>N**a)LdK zNQJG6kUVqJZ|204$?rw`NF_W+6gfkM{+Fwe5vdcT?pcGbT=P6f=ZjNTnC5er3@%)z z#rMkHd32>IZR126`aT3*o985!w^9yped`6!w`;mG=Q+tLW$~Muhj1{Mx#t>t-DxgA z>)8f`@iTcIwf(Ao&hR+rxF$%?RF&vlc}rtP%740#|G?9_)@eaH4Kvmp_DTyXnM6IWH^Y$W-yF2l#VWrjom}G+WK2?xe7 z0unPKPYck!r7@hQohA%JsY+xGSs3qCx<gSEcIhXz+lHa(xnSVsx7bQ=P+t>0wtLC5^ z#pE)P8K!&TB2c>D?Iy zG14#EM~H#w!bSo#|Kde{PdyK3!ySv5?>CD8MX+mm!bVS=$5@tp&BB6;f3Axz5n68f zSERgP8iQmh%quclb5rDiPWjJ{yoD*NJZ(<^t`V@{!62c7>+!-+PWSzQedfv?$8?lEC z=mt)2+HaxlG%t3qwL5lK*_P0$qL#LN&!;1B8s3^}1kdg_3UD)7VQ;!`c>A4=Ivfr( zI@~_t^`o*Gt*Aj3Xk9clEP*9^dZ=XTHUWt)?Xb?E73m`C8thZ83nr;JKq7|lQGBbd z80{*wC#B2H_(_>ZLBfM`n_A@F5B@U zyap9*8NuY{I}c*2{@`V^2ZA7i}+2Pf^tOpkt+ z4&RB*XZZI1n^@+~5wmV>7M}30YuAl=&gsw^xM8D?shCbS*ZV|6abnD}0hkOBK- zqCN=0;q;cF01G(jTR{IJLN zcTN;M=L4A|XrMqUL^vzDr3rkKritJ)u3s)*+kC%7B+cwQGL>$Al1Z$b`G5p0=ei)Y zK;Q)yCrXpUSirQcCkeJMYMY{hjq3yFzSkCtsN@qG;70#KT{oQk_i%Fg-*!cNTm(`~ z=gydSo=(;ai%)=e7%K&4nO*kyZj9@d+YK zx9R@d-%5&oBi>%HjH4f?GK81AI#X-i>#c)M`$F4yRoDT$e#}{;tSiE~j zr@U{Fg41KSnJk;n?pPtC=aXmeisgUtyY2rC8lGkHFAqKSLn42PSURsatrYmHrh=r) zQYxlM!0Xev11q$iR8N)s)BK1lfhau&NLb4~5bfRh&rUg?#Eu51;8_{sOZu+qQR~yc=VQGJdFh1|fN%l9NkfhtyUb<>@lyAW+ zyLeDKEA+I&+p7j1AuRTbdSHCCL&4i$pWW50IE0?dxBn{`$vRDoy|Fc zVQ5==><1X*70e<&8D__he2{Y$7>F_wf3knv+~B*7GTa-Xlq_aCOE=sjV7T%F_6-;^ zMk@>PHbWb;!}=Wa2r`_?5-?CM5iTTAiAcI2t%?4k|2V)HH7Anj@gDV0_S4GxkQ=eO zrZgA$tApk6pIb40eGW^2`z*xa)*+R%Lo+7`AekLp2;DwJjjt~ez2E%;PW1rrj>J;g z*GhJjqdMAYOd0tjh#X7XThp_|2?V84#W0oORnUKCK&FnNX~V_u%~9!Pp}f)X=sY0h zzm*E-gdwStfU%Jwq_|r7no*VT2{9keX_c$l4A+YH7U8&T6|UY`mF2UfBh~RM(u7rq zVb2=1&x&a-ELg}`DTU)3Uh-X0o--0r3T(Oxj6+l2s4jd$ESX=KH!b1jg^OsPuSVrV zc!E#+ifO+p><<~-(zA;p7o(rE07N++SnTi#gS^ow@Kiuf6^4$JQ&@n_tAQ6OP9ZDZ zPj+rR9gRfCZ^ZMxJ-u)LD^-ds0`(haxB$k25011SUHjx8MAnOC)z-h0f!4(RqzLLx z!|StfTj4Pr1+ohiW{XH;k=sY6lSE7ZjnQkoej3 zTFmf`N_UVQb%5WoFUX<&f3pB;fdDfy4-xs?mK=EPmHp)CIIlDBb0x)&sJNM+)IY4} z=sVGAbc=Z5jR@g!>4D6uvm|hYcj55mG9nc*lsMZv?5w1NO{fBzLx5(BBNz$vI*Len z`;PbWws?nN%zNrwU)#B>;KDklC`04Tv9|{ZRSk@8D|~x_L!-F=i1J-7gSYTA;Hl?{ z9Cgp>ymIA9V*{L9|Kemg`{PtWF0{omjl2R~QPjoKykK55nyBpAxFlaA~Gt1>h@ zE_qzrQm0hG+*krFo6-v-K3{+jdAtUNn<-+18!ENwBU}#gatup|;j|2BDhnq~eMOdtc zFZg+z0UxD4^I=sh(SMYTUZ?2iX07-~OCm+*ip6Y@4ps6s!sIltm9Xu=KPcY)h2<*; zn~rr~wwoMlWuq!(KWXepr+lh?_@j+3=r&xg;h+dF#D>1x)-1z_ube{!%Gpp?MVE#8 z%WbPlPsCq*O<8Icr%1+1*;27B^#N;$C!$fOj+wD+ z?So5Wk=Wh0W-+Dql(k~G`jK8J!^bYyU& z2Bm9wf>hS(tMn)1SZ`)`tjLQ&ouQ>2r4|MbnHJ_~r6j3`NUqjfmV&IuBwLkHvxmVt zA=9sfjP30A9_r?lxamlsWTIh&fDDitJ5B=YAM9kkq_Nk8V);$3;HM% z_Ep&F6Y!Z;85ZA`GHOia%!M7z8`~>_^5;%|pFs~REf~sywPU<>jH8p} zzC4w0qTc3PI_eeL$dzQQhnuMX zGXS8bi!V{$G&qW^KldSY>|BfVTZ~xbcExC7y_$?!6o*lsSGY11(LU1TNLPai1RNmPpRm)7{K^ zF7^}y#t5)01nYipCOGxp(c32xGf1Q)1bD&!O;%wQljo-z@^D(yK0o!o#twR;OThxf zU80wc2P$P+ydPK)Qt6b5t&c~ldG<%X)$r*VfklU=$z4mRje*pgAD*YmIL9rldw|m^4H9aTUGrB>VL}W|RMCHI|9U=fT@cF-;xxILUI#*z?LMJr#D1e(dd3afj zB^$0&(rim-uB8_ZcP)e_@=9&fGCs!q5<8TNcoWlU(yNb?;(UJ9q-&3i@5H}!YrBu) zb`g3x-EEd^lZh}DayX?cH|(Ca3nY`@8CwtTx#mM4tdtDkp2rgtA=87b*IEj?a5|Ah z_{l=8N#~0co&l+WZ+8+Nwn4S~vKgep=fO^QXc(Cs&T9xt_d3P-YcNC84;Zm{zHVRo z_3-XDVB3?n`NK|Ck;_*0PSZW?%;tU&Y!QYmy9y&h{E!mkcJI z!g)OXz^_cnJR3(cuw6wd8YY#HNO==rWM@BpqWYvU<)#{jXf!D!k|o#a9ciqZids2y z5*QIdL~gRMhR*txnhWV@pH&rNIIMu-X=HeU{pfmZ%I>|FecA&V{rYk4lO&@UTitXk zQr`+4^kNU#UPh$S*@yxRD1o6Bykhoz<>Oig{0cpVwSGlp;RELN<{Ama<+PqeaqFj4 z)}P+~=H)F`LW4}S#=lwuapQ-v8fDVyeO^}Zf0cUNwc-=NXXrl{MM{rI3M&vtoN*xY z9K&0cz;3+;kv|;2-Q3cC`1`fOFYge`d2!1Ve^QN~2{(Yh>5E@^+$q4Mw52=_>TAe7 zhqIh+eqYox5J~rgh!1+~hqv;a$HUr%)$iqPfg&B{fM>Z6kCedXzmgFm{baAfbCMcY zuC$n);@DvOA-8)9&p$77Ra7TN9Xo!zPp|M1#*ze~bm6#0tTFMJH!Be-gmyD?{wBk? z`@O*^(k1F{%?FY}W^#ASaLful0cK&lnIBn9JA|JQ!h&a}pj%nWpotDgYkZz8HCeT& zU0va@DGiq!5Sqw*L|s>Za7SV~ow7;vuG?6`(Atddzux&M4&qntCv41t4T{i&zR`S< z#~Kz+P6FMk6m+aiBL7L}%fi_EeY3O%k;!(MlS>ul+g1Fr4vVxHunp0j?Urm2&_dg) zivl=nMySF&MJ3vNY|Vj-Eh|Rs)MW83X(BWsg=UMbv~|~?Rz-TTU4|Fv4TuUx1wjv< z|J6G4LJ$9wRGVrGUrOn5YGG~sQ2YzaWSCxs+YJ++KTp0k`iFtKqS6KMKUAl8veTO= zg4JT+(>oK{meG7M*FH(wGD0Z^gQ_Y7+k#Alw7dX*euk-t5mBR|nK#W8R9$(-T&iL4^c zu+JU?MrD?~gok$!&atlO%f=qrz*lN6M3ixm{N?uzeut_q_-|u3mYw)!R zKY5?taP6nt5JxQ{1wFz(;G`?UFE;u+6>d07x~~FKGgDRbMi3Wu(z#*}2H|w%?X>wO zE?mLmO_P37=VvbPM~!qL@)e5G3qBN+a(Az!#FAd_l~wqv?*p<#33-_anLgW%RZ2C{ z7v3R@=3>zpBra`{Q%G^X(-Qwj7iOhI2zxy;5?_I}orXa2t3XalW7r?sa0FnKc^;(3 zBgSssJNdCf5H~I>bdj-fzEA^+qq!;R3qyqwK!NNO!J)$75*96hiXwnE2Bwv#^?7#* zxw6i3Opv@S0G<^w*P748LqylXbj%M2Us@*YuY^}i`vn#c1l;v0T&~jiPEW&qv`ZLQ z?DyRUzR+@QvTqYSc0DXglk!3`I>+^DP--S(cu-bwhY5(7lCpRGYhuOi)dw2C{>R5h zjqS43isr>(`2g^s4#;uVEOk>W*G9L|56?a$F zfu$lDm7}g|b0s~bD`VKfm=F9$i>a$Z|6cyOvxZgQd~t^gAIa5m$_biL<#_y_#SM%( z*po;YYGZq+xU>+p9@6d|hd4sI0GBT^R}qS^u(UQ-ppj>dmUbC?<=#fX@$AGP zv+z$D6aK}={$b`n=)U%Bi2E*0ovNSk=OxzJV`%83NOZYazvf(Sd)hkKiiUOIaEq+2 z3*!*!ml#e=BHAXym;T<&W12A0~h+9$@rXG*LyM~hb`xRM_4_xuRW zogQo-F4Q?3`TL7JUJ9zY3DuD`6aPX!ZMXTY=OS3LzW)h_0E)Qqo|eVVG1oF14AHhy zA}&<%9FnYBaV4ftnbPoYG9&GGz7u~xb0DVznqm~n2K;_tege0V!G~T7>CC~`;1=A! zyUfq))hiU_X$dfd$XgXe@E{1|{Rx<_=LzB#qbxLd`Glg#%7*KR&k2@sed^*<;AU0C z^L1iwWg^QFr*SLMKo2{gT(ii;0MPNV6)O%-yV9i)pPydLqW+%0bS}Z0<}{ZF>ijUJ z2o0h6!_$(3BCW^hxHAHV^S1$ReVuuy!p+5OL;hKn6G&)p7vWFkCC_E204{^2gj=+irrIQ}~RS zEao-DNdRTS^#*Or&Awl=OaSdW-lm^GTBbC(l{j1HsXh)a~ zR)E8_@J0`h)adyOF0#K#PjQrpQK^V7$GmR*ZZM!3{@e9-puEo1?N$mGVcQR-Vlqfb z%U0RaGQu7?c{KmQFGu`NMasewY!PLl21;mOs5Z-2dOzxnhkG^9XF?UQqO!k&QvF{REAJ6j$CQ0O$5h)$lC9`*0GYyJYa z)+Qvk81Fu+6qVd}c~JhGx0l_;fiy1aBsyxCH0n zuHpEJ{@#L6Fz9UKDP^ArEP(_nBDsP9rp`$geb1&5K`Ai6X87lsEw)t!iOeKorJna$ z_9%+MkQzP=WLhvU6~Wuy3M#I77?j=<1}ksHGPn2UOsE6;ufwUAQl*FXl3x3RD!9&X zlmS5b-h*sS&IvQxKLjVyebSK6*_-4_6ilR*>;yoB4zO^WO`I5T)nx%5HdUzR{3VY+Da!0ug@zvwYS(SNyZ$Wn?^E0zwyyuv!wmw|T~G3pTd zlWP53XE)hvtnOREkJZ?kVRF>YFO+h_sa+7=5&Vl#N@MBa!qJI7L^Vzv5%hF;&tfY< zn3~IiUn2`u)dwKHK02dJe%V!5Sp(ke-dNo3Iv({ZvJlFn+Pk$J?Y;ybZBzY0n4*bn`f4qyEpi2c%rN zv*8W^i%ChYbXL_*;oWzaH=?;X+cJW5XcV1`&VySfe?I$3y&WLn@rq#MzyiL06v)*( z&#kwbM+l2BUA}^2DP*wOMZsq}MAdh=K3w#7U{h~98IF~{(O&w&=wnKb24u}kdol&J zKU`O$cC)cswGsiWzaO7fWR50b1q!rSB6r0DScx!CvY;geF(denS*4@D}gxW zZZX@V$LlaZ(vKeu6SuQp{DjzvFUB~q+0UW1yj*MZWg1VpfGTxfOxJ(UaG5zkI3{)4 zV z^MO{BYCi&;#EIC7=h#|M1qy=(MFeya%xdtqUihvTrG`E!n@^RG^Y%s)+N zdC!`tYz9*7NXZkSzS$R_v+8LJoa49+F-8v{N}EUmjHXNpN3UV?rq?6Vl)@uYDaXfc zI8eSZtvpQ*H<;5BV%om&QfGeZHmfs0tNP6YQ+D@^i_^!vuiGwyJ3q5#`R(~Y!6&j3 z_Wb)#so)dr{drb=`mG-)+rO@b+4ufY+5ao&U!$Q;6!0w-@dm5&+bAeInA^QTS-14A zLuP0qJk+yjb#Q+7zdZ1Bw=kZnvit#*_JT{})4;m@;LLiw8@7f789C`LFAj;!iGhlE zaZ+Qa+S$~8+-V7DNw3PAT>AItO`yvCSyB|=xkc5HxqUd6$vQ;)$_4pTQSOyqs&)D7 zs6z{DJ8=exq?=eZDBXvZ%|YNEXrB1}UaFgNYIp<8W$p)`US(pzg&89rW`EXHD)4zU zBU|){l@yRri8LUCuYHmsID!|J+!W^vH3vD1Fz}vGPi_i0cxGUvSd18Ser63y*r%>_ zbbls?ODOH+d0x;r`L`tHI>YWhEJWevq}$X^TjJh^^#s`Y(bjd}xC(3JoTyv*07<$+ zI4G8UGDvo`j|DwqG22DB$qxO$`$LcSvwh1^k+DbB^qYIW%GNEo3wuX|yAq?TyyFDg zS~`b?`SdeI{pOzTfx`LQ26EI?hw5*s5Z%*1>6_Di$Mb6{LGFK@gJ(C-M(GB^-8G}L z#SXY9*b<~{QVpX{)fkcZ!G1pNx|MJeTQB0bP?p{O>tIv8ZP&20vgmScg1vTiCOcwd zr&Dn2?mvidoN%%PpoBWMcY1O(3AtZ)X9?&x6!D3yc-0t;y4v}_gy;K1=?>$nx6Ln} zrCm0#qfsrnl74jjFA*z#hQA?_@68r}(WWK5uyMA#dl?d*oFqH&fn*rE4bS{Sig-KG zrrD|GjhHy9eo8l9*X;CxMFsYhA@?^6ZP1JY1mU50%tCYT9{&aCs?=^m(-jzz?s4m9 z2m;s2Y1czbcu_W8PF6hlHqo)*_PdVYlxT{x^CnV-xZZU3r)nyRdh#zrFdAzK-y(T0 zKRUfOcICqXvq?Uy9rRngw>0vdb-2O#mh0J%nwQ*7`v}b4;GLpn3+-O9vAtg9l6X&8f)>D zWJAGg2I6xnThYmdAQUzAlr_^!dJVCQL}2qfIP)x3Yfk$4=5Y#x+iE@b3eD(F;QRXf zz2X8+^x(5%(bdb(`4TqgL8biScyb0}LsUI90y*iD1Dv8672Oh=SfQ0M2B`B!y<9V> z$V2EHzP6DShF(eyYIz=s z!Q5dwsf>i^NDW?1jz2VDmdfGg<0s5kInCqWhh7@V`uWrPAz}@|23r!?X2VNw ziJZdUUQqy3q>>dJXIv2B01@!VH+2sX_APrDcb3wMXV->X2a zp#*5Ti2sm~@Vd~R2c3dWCt7<)4}D) z>^EZ8UJko9(iz+Wz{JoPKlp`-Q)*+-HsTOTs1*Hn*UVtJHOy4$6AExP6sHgQa95$9 z4nfnmxCW;qSqUZVVJ8{#!zX59Pysojm4f%16Z6r4?-(w>LO*HOlUT)V3~vChdat*-$}&3|&DS)=HR<4`JuTKVXT6W^nrQz)y^dSnIWEIp`6 ztGal6v^QpKpBc z!n(+55V@Nl{%kgg$be~(17K7>twZ)Wb&Dp^O<}sp?-^w}be9yedcxtwv;d}PKXCey{dhGj_e32u5kaeT_JJ0LKghLe# zb=AFc3i&U(l-0RJMQNa+Hb?ha(4dW#438O1QnQ#n#Bn~KbI~?798xkchyb!6uM#Qn zcb92Fe5qR(P;ScI;`-ov{QIIS#U#MR4P+5&6=EDY9+cbf3yRrX#I@NT6-~6JIx*y> zY2RKC@J~1g@XJOaAd{-@N~Xk6=i5-h%g9Axt=X=gF-#=p8@@V$dtZ?cZl*^}nRiEO ze-HAT<&9N068vDcDWgkAiFEOg+uay&DCm;|i@du6bGwprx90tzmi(+VvAUTJ+(Z`e z`J7O|!rf+Z`?BwP$p?QAWgxX*S=A;2kI1muF+O$tNVnPfh7(L1XKLscS?U|Q?Sy12zgKG4!CG?ke38e*7T$IX`TfpAEbZ`z z6gni0egp9=KyV2)pmLUO=ww{4@9=rEPk0`lci}Sky1(5BC8!Z3=3y_$&sq^4#FigY zkc4|ji)!wP?^AqKl;Np`c+XqiUU2sB25O0p(0>4YHX)2^JqfDITI@*(Den1EICo$l_JtK2x*gzVjpeN94z})OzgC7MCgj3v>I=$x>iUaxN=Cq| z#@^i92&#taK}(a>$xN;;DcOc;%yOSGVrFKSKR*j%=FS-MtGEZ0TfkYMqCtU zV z*@)-;h#^fRjw}_-A=e;+(fw)*(RB;pU8^sEK(;>*mIpNVb@Z+d1YGgJH6Ii|yX3PZ z>HqQP5zp7C)+j7O_(SjuUqZ^$_&t!GLU23I9k0-{8ywy=+clG^2WO{ymaD*JOyS8h zfW?+)4S}UF_A{uNiBVK1i{^GM0)}lMtVzm|e~c*Tc+nEOr^2eMAW(mSt!UI~!E?$< z?**6Um4N3v0aLpjkpZ^KvD%KDBDqe=$uxfP?tLnvf_U&};NTc;r~Jv(NlsQi@NeUk zRo3&9&)nU*P{I4QuJZ0XDK8Od-3}(+)YciAym*n&*%i$jrma(;!@+ltey&3JyUO1k zqe8Hy)(4;aKmBL7zW>v~{_0^nQlY*@ljyTB&1nBnoFaDB z>l!oZ8ysHPD+=MpM6xCwCq3M%pV(&%ClABII`}E`o-}V+jUf3o5NvdTCa8ct!xt?< zmsG_^5Sz>!-|x^XC28FKuMnIS&I132zgImEF&MEU7gdMct@7u@?45M0Fd{2$B+wIm9hyUzHE~%^QL$LfA)$Fnu$&y~@ zXT~hBNA%fPcf`WAqVtP%fcR8$Ayq<5U{ipiYdi!3PSP!<1$5gKFJYa}2Bpobsm!VK3%dIA&Cx3-C8uHcfXlLMA!?V4C(8|~{SSdzball>vDn(Jv@)hzg}x_0mq&ceTO zg?qTMqhTpT{Z6zR2?H5X&C!t#NnnkjBXCu5TGSM@UwLYA?BjS3&+>wxjf+602x zDh&a&kH>m_l3~AVx3T3QRbWs&FHHm%ipGNaS`<)`$R}(2EO^Jx$-E;#BbK8wZ_(6n z4dr}SfVIM#(xCg5y|@AWvwEKhS-+jJfHa$w#>@i1?g2+8a4(AH}$jMQ4A z?Wck`cNqhzs@O>de-78)*sI<5z8_>bu1OX%v z5LGpPlL#a4@H7AqhB~0}-h&{zXE& zM&KJ$eGFm%O&zm@oNg#7JGh22Bz+V(HC2M4b}ck)IaW-qK_)+Jm!uneqA{u`)myFy zOtoft_K0-C>c{(_C<|@;ZFnJZe7KYVcx!MxeDBYjZfbN1n`lV9?`JDnxCybA{iXtQ z@8nYSkpD0$oRD1mwc!~5ZPYNR{(82Bto!Fi;YcCDi%~Q;euG=pCH)?U!r3GJAf{Zn z;j^LT!B7SWgoM0D439J7gYeFv)ZuiYf{+?-<_lOrhej`KP^52tQICW~5egGg81i?6 z#*C^A=3hcQb69)Ko4ow5p}tX;^=dcq&AickSh z5haW*tqq5F-IuxBCuIM^&X|p=XNb9`0BJppj8#9m|Fa3EDEwVhx!8X!96btCoKtNZ z^{{0n=kFC~9e#ZhAkb1?CSs$996@ry7=ddlD$BbQKuU_my$pdoKr$q(lv-oP=lza( z)x9>p$`+_D?Bb3eZB#@XTW3(0?Jrv^Gsf?dG8ZPSTAyT(WWcRs02D@gA@Vb-+8x}m zzMUBW&cikMs*NWJr%>>NRyGuGvX-xE%4Bk(m92ku8#!zqPnU%jRe@7*t&>qVLoS^* zYG-tt^%@fK*0++FeDQhOLnek0e=`&Rqu!;?eIZgGSM5Hwm+j4*OK|gPD}_V+Rc4{x??TI2pJls;)pJDBK;`6a zaxN$^e@)Q^9-)w@Mtd0f5AB$e$u;OOMgPcE^4W-fC zc&>s?zh6vQu6MC=$6RUz^gyvP7hLP)WyKC}`8n#JC*^Tyyk6)15gPxK44XEMp*E4h z0&6q*q3+4PDL?fyj|LxRpb!Ut%pj-6^niF6kE!xkm&0*e=rI0lQXj{*x=YP5jKDf12n`W=f(xdmai z8WCgi-;j*Gjt=I-?ZGN;%xs76#vddS(QUAMShaHy#kR1Jrzt-%|4F^!_+Ji1UqBTk z?ZA_#Qn8?6cOOHHD}Ge55AG0G!CAD~Kf&KhKA7dTC)r$%@!%=Le{=ENhd9^Y8jt$U zqM1WwHEM-vx*OAz44gk?6%s*2{8MMK=Xp2nTZ0CzD4ZT>PRA@!4&}y_-ZC>9&D3L z95T0=^f+Zxdl9}6OAjm}-YBymsd;8gwlE60zHD*?!>Hrcte(p~Z6$mAQ7#+?0bo0O{EFi%vxk40Vc3q7jlKU{U%BsCco z5}TP4r};LDvPgB;Qe>0ogXaS+e@UoWrXr;XH{8qq(gtBV@s3X z_}57G>ni7u=cA%>dRj&MRp$Xd>ay)$fE17+L; z(X)H{x&Vo;I$&UPzrgnY$4(-r1KUNem3AhA0$fy2{R%0P+=xYwb;mnBrLenNt?d^3 zeyyBL4c>(qJ{E#5;Y&`>#0(GbTzDHT zIV&R42lB)7U9m?~Tu6AdpRc8NZ3~{f-`Rc& z>tc>isW=!XG~B~>p;P!g|FobAO-MfVSKNE{JOAFGC1Q3sWygQ3#|E9% zpNjtr$tavgJH2JKmM(mQw$lLnG-VtQ3qF_4zkgLF4Uk7Go>2@hKBfE9&JQzjl{s=1 zP`Rl)54A1j+lyxWTFxbDnMUSFj=r=KF&li$tEh{rr;WL$>67F@VffpB=wK0klq6E=RTCKY`>Lw>DEo4-)Z(^H>NC`fo_iGFs@3NPO;{%c?H$7O|m z@d)p)%((!X8C^&*v`^}oT#4M{TUn?+4Q4PMB?C^Zwu+ykZZvWi)t(b$fPiWm{CK<| zw0>GWzp^%B*m_JOeXmFThNkj-6t1woHx~`FzGe@s0S^+|@X0IMK(EO~qqAVjAHZA2 zrz?^Pv}D?`V&^=xxJzO~C{0Fugy;s*-d6Bq_QX3p;`~MF4Vg_2_m|vVsHE^uIS%)y zK$ynK;ND}{6OLx-wd}2}Qms`%M3v6_r!ua$Td&A-LubN$C&cf~otV?!OS zAef+ky)8zIg;X9TKOfLOGClGvLu6AJsIY`L*KlFCS7R1=qeo9|F)E>Io*=e;6q90t}u1GTL@bLoj|P%kxPtk`CGhGxl#AwGd(pB%%mJ!&Qhu zr5W8H)im$B#3}2~15W)uPoo;-dCJvd6F|1-7lOL#WeS(bEpi9gUk2Yi-v6MhLIks* z=;S^;qEC~`@-sh@Phx=Q1N39ehvDby^E#|u5_DvCyUo&P9?HW|tI1%=jb}|Y=*BSl z4}XQ|sM1krp-HnV!u!GxSKJg^$|c$y2a)S%bRf@fQ64`AZnswN&Z6C=@%E%OFSC6o z2SKW#6TPGc8c}5(Fnd825f={(>3&W5AMpk#q}JAPq^d&HgFk+fnq-H2*L>`LHx?u+ z8r1t9aSr8lNq0ykADtUQ@i6n}_Je;hYKRUu#q(knAu4$tEu*xtb7E3T*Rh^kKyARP z+Dsku=A@cb`5Wm(vvP@?2q1ZYyLoa^vig`!Bl_pu0rOoc1cov^XL+M}gKgS8s0Xf^ zCLk`i!za`t-V+f^Stb_--^9n)dEp?3pc;R2^@s^;;-+l(H{StV9}7fb>EI?fa!>fT#uaw~N=M!0>f1mzJ`x`;T(t_ZM}ogJqV# z#bODg6i#YdK&*<>U$7s!IEre0PgW+PBUCJP)3A6|1_PF+yJCMghEe%$B&7}BP$}~~ zdW}p+QSkO_cmGL3eKy;b(@7M{3Bd#~bg{)ij`s6^m3edW@l)RYK?$u9v@%KlQP;X$ zZKM4>$nneu3AtdIsK|>=B^ZBN>-HAb=l+h;qVvDuPGtWW!^^g~sS@%royD)mIut)$ zLq&d|6;1UXL9B$nDy4SE{IraMC!;s;4A+S##A;lhKjQfsUZnj|0=pXHAoA`o`<6Z zdvexn#vOy-ruK^{`THtRx*}*l^1dDVvSc1zDlOq?_KEDrmkU@o>L65y=ZaTC9aB>G z?h2(Z3XMF%yX_xF*!rBTR9kUB4Y>YG%$+@DZ5vTEf}ThG5BMy!~}k1zgd zx(1s|(?{tTP%oEy7d)zgP=CZ+$+?{JuUm1rb{8OqgQ{IjMjsV3SuTSnB9RC!W;sAb ze82jAvhXMbFu{+@=SlJ5(7|Ns zehN~oEBSVI#O}}<{)`-uJTtOGN8OW$XVLi<UkpO5JiqFcfh(d zPYUKv70x;;+;d4Z5KaKr`p+0`nK&5TsahQKxuUEUMH-qMtLS*F2rC7*d&&aW-Gs8? zYo3b~yoL4!>tXRrR&v!K##3^>Bdkgs^A#(89({IqA)dD0ms0B1v7}F)gaf6D^Ip<3 z>PmtVDjjLrqFQ_h8xf2khB2|{0sUK+R$ic6zvcKk!e3WArlzJAk>SWB$D63b{>r$Y z{tFEp5jnYmKHQKdPD6YmRAu~R=Gb&o* z!J^A6+WSKdI=4}E!+`c`-luSF63@#YS%i#k;6@a+rU&p#v^NyYp(o_RE! zIFe@YlJ-9{*k)F#wcZPE;Si;Yr*G`yP?o6;z}b>lRPzhDK@FniFR!eQo7;0PO`AS; z#SbHFew!?^1w>$~XkCo{gX)Fk+*+ZGy-|Zd3Y?;n6ItJojxObcEtsscLZjA{l9P}2 zH{(dFszt%3epxn#Y-0Ql@2dG-7-7AAEXayKz=4JoCQ0Q`5oRU4L4&-5p2vG z_n9QBP13YTTC5lb*|@J!M!5V-cB_}0dM42jL$$bZx%F^RD8Px37Ji#GhMHy~7Ri?I z;|;zW65YtUI`x41#P79Woj3nd<4YU$%sOfx9d78On2vX? zcYQDU%UQ`?6S3np0Ghb*W{^xr_dkwXh6nDBLk;jXFlr*H`OB$YK@@{Hm_{Cc--YubnI_csM*pIq-fw>qqw)IfA-ZiPe^|REr zV?VE_`0Cy!y@^X?ATN17CBG@$t;vhv#X}z^><%sjc?C&5O;^NBf0-@G@%cdJWP-o!rloYQI5fzVY#Lf?cx}KWKQ^M37kK;atOmUhZW~*#%-ojn?v#5` zLlQcv1B+t2!hzs^Hzn0_)fgo!B8L8~-Z_5p zU}uD$4wZ?MhhN#m;9`p>-~U4S{>YjT<6}rIB?Bz#$Mb_qTTTbR5rbLjwH|V7cjNrs z4sKW@Kh=tpt6NZ#um<{9YZH!dy;y(2LhkW;htoNRu<#PBNOUM)wOa`@MBzai;)`J% z+*kx*yug2s@b;vnn1|f*r6*yayl2Mx*ERU=04`#n92as8UpVWVkv)fxP@fPhN;i_a0eKFgQcwBIi1 zOnq4ZP`t&C7b)jU>vciK<;p;~o$|p4n%Ihi9p8{=pbI?A}hxry= zs3G0H#DAJ4K(EOJe2gxV+$CTsa{Y11gwQ4c!|IQ%=&dqr^eAZz%6~om_Z+a0GTk)G z;m5tRUJ8aHW-M44Z>XK4f!OM4;7pR=p1s>xqGGU0S`j- z8jhD~`n#s>NA`ZNLC`>^CCMN~#pvh2V7*^ZuvU)cAtnO*vj4BBzc%ab!zx}gp9w2` zLiAjMLTjFR7#s`|9tvYe5{92qh@8Yr!qYTIn@$=A_8Y=#syr3h^d1*JzQTS+M?Q_? zP(N+uj@f-uC)?7(gEe7=zk#EslYx(ZMmvV_{&Ogm&e3@cXvu<$c%hi7NK72()AaE} z29U?LrO(*rPpB-SAS25-ILJ8~jXtWZ;#gP?Zm`55PCUowT(#{8VrY0BAzgT#t9o5t zy1d#SCyu`ze_6*ObZUr)>Qir=sXmrWykF0opf5xcQ@^nFr$!cIb20uI2?;+4{Ai7k zH;18!v%zqm=3;LV@b-HTDU9@MvJvFJ_$beqXueJ(x{Yi_-vmL+*>!sKQl|cFw%=N< z);Wdktd8&YdspES!$4LcVt4Yfj|~Wx{kOB;D7b(@AU#Hxe?=o&VE0U{Au-J#_(h0bjT`0UUWqi6(2di7)@1ew#A2OA}sC1DE_h3!*%5?H- zr(PY>wawBp{`X_;u3CuvFTM62av&xP{rs~g7?CFlHpcFuCeE0W?)MECfL^J8m-4F% z#j(xkU34pYJP^lk-RS)${RWr75%6<RMr$Rl%`u|{JQhnn6C9bPn7&DLmE?n1EX zcRRmlKURyKdkxga?W8Z3GYkxmU-IXZor6_%R6~)0ma9ROx2`eG?xBYXj-_%7BD57yMXt%V&4Z^1qEYQHDeHH`V*hTA0b#bjlKNfSwJyE!-nRRn+%vjZ-E_ zFK_AcN0CjoTkD0)*cElW8D4P%18;%mI1{eNls057mps!tp65mSGhuftX=8^Lgn5lt zjl72P<+v&^_wy|nhkqj@n8_zD6z>m! zYaSH7m(Ie0+6+qyjQvMJ&|9_MFs3+EpgBZh|^Rmghp=B>QW-2Gx;V%g=8zxMD zPoq){Bo(S}CSg_)mEE-2=p02^<$CyY{gCn8L0OnHDjCgVv^dX0~7 z8J9XU1Bhd$(T#8^uJW98fVyjyj`v+$=h(XK!er`{gu)wSTj3=^${wXWVtptm z_dBCgZ{wHL`3)M#Wk+03wRvDTjdy^#$^vg=NBUP95qqm-F_y8S-&(5X93hN%=a zom}axh6DeILj(Akkw;c|0UmpKS|2lA~$a^69Q~Xt95gX z?JhM0OE*21rrsHcK-V@_G){KMINwsW*EBQWPhIY;RzM~YfnJLs93ABWE4Yh|&{6@i6sRnG6_NlE|J$Fhx!)(jFLeU9$1}N)NgExfh)kI1 za^JEE04NM;EN(BRcM@{qkCM7A(9M&1BnY`rk|Y82qX^W4g^c3v&IkuPD>hUX|J>1!5~ zrJRJWA)|6MYbM4%k>@rF0M}6cmhIr+o`eBrK67=exVzm{i_Q$CNnZnq9%0`|KjKAM ztfl;R^U;0XyN&m82%_5wLGHOYsC8X`!dbT0Q{R_L0rcH~Y{9$1UQ+{O$OB09?^lK{BYZ7}#Hqa(G zx7oCkGg`naqu3W-1@>)G|M4|n-7;r8ZRDO;uiybKy%ko;Aw9v|zNmdmnz8*hA+B$p z(sT01{URzN$(7BR2_RQ8JD>g^kP2ui^H(iWT&yJb<7U+a(f;?Pcedo9I=W1!Z%ZcXrv zG5R75+rFx8fHrP+SBaejGybm@0Cm@3BE}dX1}P%`MciOqcC_V%-7xmCG;hf|qnlqu zF&93W|34_yve4}v-26wvBh!2+mJuuBAhk1OxB&6k#_vYlM!_8r%A;b?Qj;)4c9Smk zTYqxfqKyb(8F|ilvI$hsIE#nvzwl+eTeW2icw_?V(3~66k`Fcb@r=@4;76FAH?%}i zqbW~Mm^F0KgFXnAMYdgmDRQHp3>ScFvu|gSrfnHhUKu~e*?UjMH;)1?FQeet-X1D_ z*F9c*&EDL)5C52};>C#j-L2j|miaD&s(EmxUc8OTTaph|FRn@MJYPGEi(dJEv<<#r znc$Vio%FdEz}udY&mUmd(qd1FE>{&cJyL_(x7rq&dVO=`KV;_Txvd(77IGtjD(itX zh`h)N`jstTpdR36b;Rg``)eF7eQvB#A#ZGQGxe;Qnf&8o9#JJjFu3!@1n4svbD-VL z5|T>ObNfiulW8)-9zd|uq6T<;yv|m@hI8!DIG!{=0?IZ#N~+_U7BV+_^|;(mc(1p( z6FcwDic#!00M{ZYpmhnywc3nDf-n!v_zIg!BTn2$e`y{xfB79vH?(#qzh;~JSrw{y zzfMHOJ9rwk@|yShr15zfTlwX_&8H;AOCRfp4{5qUduQxrs(Xf)P%G(P-Cp5cBq;rO za1pHk-4W$AQ@WF{(82xSfPan~@jg>^XEvP~$+bYZvaYU;j*;&QxHadcHFcE2$%U=>_s8{4=30X-e^5K7Hr@=%E(*AUoG1>Rdcsa*$@p>OqqgnsR zZdhX-4Zbn18M^gH(^5DsY}-EiHe2DCM@D_YMquwJqVDi(9Iy!TPJQM13hQjYX#g)^ z#`)acf~WT*`=n*P;V)dpz@RtN;@aAu*B(%)kyX^l|U>~rDD2-aN6wbI2M)hG@W zcfDTaLv~_1EZ@>*dgWU7Zf=QoYC}~eA@W{?qZOM{wl}>r8^SrR4>dx?-%&;-RT3o##xN|~wXa5#oe(fdaW1HROfo}TH$v>y`2PPT1z!ewz^SnY;ZYey(o{s5t{-=gl zTiUIrc?N2zEd3je>qE_KyTMFWTYff?VoweD&WRlVxMy%QNk!@g)Z5ERGG`}p;uW(0 zPWO#r!W9k;XMc^poAq}2spg3Ng~}(AUTh&+${J-|eypj5JMBU=Cz?|JPDwXC*?0JK zt(v-!*Pg)oO1S&vocYhn!|UU401p=58s^q_vcmLvr3_B-?H z%Tt}4S@Q8mEn|Pd+jd(jH>g?pZD=w{-&hTo! zG6krO%5GgD1+PCE7(UhxIcKbN_vo(B1z^{iKJa#K-PKgv55(Ui)Um{UYCA7fFZb}zvYZd(RKJCMxFf& zV_ZvIuQ|9Qoz#kY!x4Y_ng&45CS2xsZ9zM!WH)g2CcP`gYrGypjTvQIC#P#}kTYg} zT;lp$flGMMU413)y5h))3#m?f6ds&e7Ee)S0 z`>Ev!Lu1g$TA+e4sLL;CfRJY=Mn7a?W~!lx2fSk_MBc=&X*S2#a=nwId5A+Nok0C9 zBab&zm2U;zYKbWZj*sb8jrZI(S|I-D`Al8ynCBl=)n-aTto>v$`uypXO#~6&wo|uX zu4j!0`~WAQ7Q~z6r(eJ0N9=mWgJenFC+2g%UO--KW+PM7-@zodo~X(w7G!?8apiBw z*ENde&cyy}m9g>cWsTEnE9Ir#NvZKNU43~pG>F}cz8meAq_$Zut{5Xv5R{-u84an7 z0h_;#Sh0bYx;{~lt;S8sRE)?$N}Hl^?^_Dr;I++!6waUTIM&SI3!)V6)B%h>5F9TO4k4Ug8@CVJSjo5DnHhfJd4Os_9U5U~=9W6Yk zqjKEp?%kY-oJs|o;_ZLID$|*V{u(;#U&s7q*6vGADs|=hWJAZp%e;&8LVA#Ijjybs zx5Cb|&bmYj!X?!Vmw1EwrQOc4?JnB*P;pi0q5eobPiBUrOVFy zPMy=QYQ3g`tgi39T|1{D@lj9JfbnXZI^7NLEx!lBps#qg<*SEuo=j_pncLkd_!vwQ z)XLw&;*EXawy4H9a{;*77sqLml^Or;EZ_Gn2%>gySy-+2_o!xI=v698S)pV~vNcht z7hEOJ>o}a2P(PptPf}z#+bwbay7XEJ~C$hNp~_V-9j2XKUjE4 zxF)Vc(}bj%NfKP{FckEJunBUUT+9N5Saj@L7(EACuZ{8+S54pHVRX%^S23Q4#nr~Y zS}q959kbgBsjC#iW^P+p zX;?*4!lygaPf*dI2ijYXDja_o19XA|2ByDi5oPf%&MtY-DZ4jQsYM+k&{$0=cF7y%X$-`}+s497-xD zD(T$XcoQX>4Iy6LOVbri`hWBBZja9yJkt@0BY-5>BH5(K1lr%Eh1ZmJrDc?ppY6l> zwpbkizAqZ8z~;VOcgDs6Jb3)GW(&KGEf_}5ptFcP;NLH2_Ka%NedJXd%h#^k@d>Lwmm^zpGkZ9a<9a z*8>UK(~Zm9=T~b#FsjAnjOf=}=UcUO&C~3cIBF0mfDM^1)15DCx>xDyZj_Q?vIRqm zlp){cEo!TsWUz*%lz+}K8v2{+mM7+D$yi*AN|CMu(iq`wJ3Nei8f+7<52xX*({~$O~ZUwqu|(Vk}>EztYGLVQ5|M zbk6>?$*cUw02s|SxwE#}kMcSfs04SokM#cse>%|v5j#&S!xu6 z6|Z*p==|37J5F8;G)vy-+&$FdX8G{#ZeU1l$Iy%;v`I`zP`0wRR|eCn_Y~!3&;+>2 zByM>f!HBr@t&j?X7NgVkdLEBE>3NaB*o~gg`H4+Wu|C|PKDNGe zqTns3!uVc_cX8g(T!J0}iyn7D@FxU~Y^lL**vU_;zMp7_eKg!m;#^fEVm<1C%IVO2 zFv@H~r)|IwaOunHr4$%<^vi1X(>!a|1t!^|0ceuRdb$P%dYRFb0!z&l4vC`g2E7*n z9=$Thm$1lvu%Fo;1a5TBW)-5(Kxh6S9mZR@_Uephb^4Cz2)n$s)D9$u*P-*;MHZ=D zmIrlJ0?67}Q>t^Ne)WA03eF8FU9p-hyNkna5j;YL4}sOws`$}=p)M`&HV^Gs5fMlr zVgmG8krsvD`u=xpK9@=`nPzb3apFySLI^;s$-)-jp!2Pr4w~o2adMj6UcnE% zVen(R&1@}0~5?6{{Z6V3g zt=TQP2Xqiibv9tOEwJ(VMqiQYDoz*lcns}aBA8e*YG`y(Pi9gE$daR9ocM zt+^>R&Nee>VPRV`gr%7%&u|M>1vR=4lt?V@_}DZH=Yk872MD_eo5 zN`L0q^JBG^?B3gpVzafU5!A7o2)I<}#~rTIkRy$1*NHD5(+S5(7rJvKqOa(do~Y3y z(qwLBbHC_^@EE$h1B7y_oL(z~48C4m?523FH$+Pm&Lb14mOX2oMKYb@Pcmtu=^R9> zJAMKeu~PaW65fLX4^rhZzaO&v*sTy+Su7PXc`+@1FtZTo{n*m#)CS`YNwdeZ_tFS? z={mbWNY=2TzI-~is9wlwRckwu8DF^LL?HI3wffSeC(mP;sP{Zw{yAB^q1AMOZ;=K! zD#>f0sKe-az-r+mJY49dex8{6v(>5C&c+qwdM3p%8}z%2%HF8S>4D~600U#mjQ61{ z;s;&zdK>hDob^e-UWsLCMVHwPu;2^vRfSy6&kH4>b7kJUsdZ?FMSB3iZWO`9!eqhJ zKpXE-yrdHSO!Bq7Y#(NWoiRvo<>IfD{-3L3j zfS6DO+}_hum8%QUEbcEDD{yQ50=0bt^&t?moghskk&w0LU;hnAV6amP;3@Rt4iOP8 z9wkp-K3ZD2*ArPT-`5+p8@)@kP<7a|aFV9Fc(_u%Y~FP|5UI4=)i!OJ$FA#MeVi&? zXjPh@HMR}vd+6~4TP(uQapd?{cl7b+Ho(WM=a!=6Ygu^fr%YD`g3?p+=sp|Lz@|1B zZTos{302lh@AS#c9ag26xYM-US0`(G=X~ey(q{N)#eBBgzqxNgHGcnLKpxNS@@|Kg z=aTl4gDmdb9QnT9vHEZ~1(-=WeUZ&o&wKO;*gfK~`zV?j=a-_Cxg_h%Bpof*V zd!BlVFfTN0tdG6Z>@;4kq1*H)fI&tMSp4k#@aKzI-nxoO`xo5bX>=p4s z<0%|vHMA>U{gB>UKyYRBwsxq**zS4oDjs|%PE{SpmG5|^Rs5~P!=V2;-q$$pM?;aR zXI$MrZdH2lFM5G#is za#QA&L9%tDwDnWB5hl$w-TIAgM9!fyvfZ?q2?LKsBb)9@?8-!u;$-?g9PJ^TH@J)X ztv$W>XXgzrAOj-c0vOyna>9JY(KnGBn_0%Bo^w_TjQw z_lyW~Z=Thv=TllHVZ_q3ux`|5N5bATqQ2ly6Mrgy!<96s&Q;*JM=?4rRputq)2Uys zY43q%+2*-J21~Uaf|E$mM(r3gsd|&xBjBXiy6i}gNPNO$)5_~wTIFV;EecYH#9bRQ z%DMQZ+UsW)L#*jOz8YCV+HmC$@`GD?t;gEzL_ItLT(?kDikz#yB$}t&iRt@uOJGEk z;5;RSlrcnB5|8Ba>6Q3gy^%Dpp>>{8iI0y32fWVVX*0&7?omS=&PK6M7V?&jWTG5) zl<6wlOh&cr>!RnKLH8quP+C_9Q#^6>qeDy7aaT{zI^daJfq&n4)oa1^J1vVq!i$&f zdI|sJG9%Zq!QKuHcC6iO)z#mw)?J_NXX94F*a*9sUT*_6#G}jJzxW=;OOD?e6}OiE z*~R%>DvPY#p}76n#0D^@4p;5x2{`>us?K&7nc}&B%JGh&Y&{rL(u!xjF#vIMtNKkd zjHOC2=);W6Lx^Cxw*zO*&bMZ*f$0q7=dqnH!;>rOMQGO(YPc@sHP3(Z`Mo@%U0Z2B zcQVe_%JffoogN*4X)Q(d<@N^879^H#*a))V&J(9s2VM?&-8wBbWmi^c7@bs^kuU+z zKp4{>vDZ&Gd|4$ziJVsJ9TnRYk)u9U0GuO>Nr|lO*Gr|zh++&YZFjm^Mw_LE=jWNT zcb3PWngbqifGvmIi_`>VP^|l9`{9oHF{dF^u4eWr@g_M$PbDzO^l{({%*xi8=twh1 zy*4hdf(MIv3)4sF{p$3})WHW2A1^Ju<&lcd*LAb`xE)j(8F01ao|s|A;&Wownxl?z zXV%QZUZB}S$|9@5)pqenLXe;@wFLyju9J3eEMpB zln97Xw+spDwN)&wH|Yxr$pxlZf_m4^Z5yWS$^i5VTbEm69zv^3ubFzpkx-09Tf0@| zRf@nt(P`$_-Ce`7f5Mu!X-#=4rRj) zWOW{&A#_XXF9REam8yV0*%7!?hMt#i>T$3W+0}e;mx&>lFeQ4!7D!lEKmowTNSND9 zp~UXi_CMa)6&BZ4FEjI(1AWmIzX-VA&xGFPg5)#MW{7s=fa^&sW`g zuJN^TOgf(*%Dg%ir>e|t33*@25ifdpuDmUHwcHIV4p`hrD`3>0xQOv%0EKa%L4F=T zHg9ok6E6lL_n9ej_=+7~Q8;`yhw3ZAr&X|N-Fra~z!z=b%?|7{p~`=~H8YQeST}hr z``*Bo-OtfXxA#4`hc2YO0)=kPfU|0IEbM5CgFE+I-XP&;PT<^F|5f+)dt=_Y5BFnE zpW%-sA|ZbzubLGG$6g$d&x@mz*9}7pz0u)Lf2Zh1I@ypqbv?h5fY$#C7i#Ts)nA%B znawk~Xb)A0r=DJH=K&uQczo_0>&zCq_2zRE+69G5%^$nFOht4f! z0fpxwA&Ih1;QL&{AlvmsRV6fn>_-^V&G7xC{m0A0OiTU?b{NI}#cwoaE#JKiT?7=Q zpP+7#EKqg!#6rY#0VyrJbQXM?+I;1`laFSJz{)CRuckJ>=wLyX`oaR@ziYm4;EmCN z+@QaG{UHmfrahx~y4z0J2$h|@kw%@0$GH4OyMQ<@dBSM=vDq}vDoq|5TNheJOTexz z^Q6XWTycx@voqa1Fu}j`drY!HlA+ydG2!X=0Y3m=s4KNcW>wvgrDsNrOGHFe*Yn|%iJxO%^=O$V6E>i0HR&~4w%>fzC&n*&aHyT_ zR6ozEUbl9yb9YDS2%n1v`)?bdU40#G0AE*Uz`NAE&2^UD_Nwi9d!4yamM(z76ixd0 z6v}-`9wCO0ycypTOtV`nUuh@{4OYq8WAtiUPh389jE^s9v7RWQAOK~U_q{S%|DDF;y z(o&#EaS!gU!L77VinbIdSn*)NJxFnPcZc9k2)x;6?{nTG?C$eE_xL{oy7z#W6P*1(h;Ire23VKgC3=S9t}9h~L!0 zmT)qqFlHXJh|Yt0I5NWPj3<*=rw%L3qf;zB;do@BUt-YFoed$_!?x9k@Ej^8t%*wg zdlo>vQ_CglWn;Zhsd`I|^#N{4gJu^bhYGzk)(E>xSbaWh?e@*6rbnKI-&Z=pmW zY>u1|#3HA}3Yj~|pFdilaf}swZCAVODu@~H^+BWLG_0F=i=Efg8;%#0OOIR1Yzn

    |Z{51WS{v^25SYWl`^-Ma zTM7^1=&KSLtue1%*|g*y%fk;2z<{qdMgTjd-oxB@_3QSsK#xN))3;dlUI?^y(yPYi zJ7VMpC4Lx!_LHYF%k})sG*X*$vG5)N&3qLtV&6tBwON3=K(3=VzUzDBw1waRZ&Dz+Tp)Bi~7S(W@z4EQy~a6?;pry%Bg*^M}1} z-8Cf)7Nr9Dx|5HWXm%X8eJPw1AffV4be83DPo~PpRTH~Sc8GWDPcz=`*Q~G6a2$k- zc;e;dY@8z%BLxvQ-P)6BzQ+5qIaJEA%5Vd%gm)lt?*4M zQ=iasdY*Y3r|-1r5PbVu8&>6+@sfH2f&jA3Axdcs$9IMFW>|_|&N7U>@3+3G_m~$x zohC9HrY4L?a>3g3jCh|neg@EZMpwa6yO|7>ykjIz-!GeS#8z3}q3%2P{D+5J?=e%R z*HxG0`shb_y=&C6LdU z{2Bd`lDeBwH&+GlNk#2=rg4d%s$*GZbn-{}+T?}76{d>*`N3q)JqRA#P`_DZjbN&Te)DEvfyRe+G*eUPG^duw?cSuUhE zf9iVJLT?wYRyLpR>)a%HT>Z5GCH2+C(vDvAY%lz{dQn6GT^02-fZm7hLBnaJ@_Lf$ zdRJf|&){VRJHCyE&?k4u!IxeC;J$5F+d7>uham+yGgrqfj^p4n@I()yj#zDQk_)aE z$%eSJ8X6-djdH-IzjEJ;AD31QhSzQ-o_WhqjP?L1=O zoUB`8jvBWlbd8@yY~2%l%Ry)Lq^%UcPSct2AGp?+8U2s2 zE~fRoC44S=Y&=*R3j4vqdMdKQ6Jz5Pi1{6S{8LnT`gB34iFnJTZNoKaO4quW@tIP_OSseni=B>}in5j}K(7>0vh9X>!v6 zz9fVK(V$a$L39P;f<5L9t{+$VH=AA;>_I@dcZ$m0D7huq6!MtfJorUX$3mZH(h^#C z-!QY|fvd$qLPRiDcaKMss!`4W7kCq9kNOT!_m2{U;^~Rm_ViejRW!Eo_?tunu%~PeP_O8p0>yM4xt)SMce^CQ$ zkB|!PBR{>KH7D-g7`vG8jXfBABx$Wt*Tb?-p5far*JYRD{PK`?)8}BDbLKQ1$ayMc zRCR)8QF{}`nD5Zu$JjsY$r()5Zxx|V$lni+9@?Jme3AA;a&~*m*@1-aXKNB)x6uEqa z_2xsI0O@hRY?*1NM7s;hf!m8y#GsJ;s-(yY>lVE9!tO@?(uUgcO&P(~R!q3I zFRhK0a*&t7bHbJ(i?{a@!|L;>x{F_IQ?;Bc0|}V|9fufDBlF zs~$rZG9`-Z&oY2py{_m=Y9!W4k2lt<&3GRyput5lO;L))o=+E5&g=Sm9Tgll(nNI1 z3Et3XrfMYKGcb}QXNjw_W>85v&%#tS{{dpwv3F?s&_}@NT%T&vX%p zs`B;g-Mv>#W=;myaO6*1d(lVz0y102aMqY)v6IQ8q`iceomT?R+t)c&-5+u(E(*}) z6)8(KPXx4Bn|kJJ&qOPQe0IF&XE*Uy=)WHng`Y;eRp&y(vr&dj=o>-Ft}b*1@_g*N zdrZ6!d*U$;X36FK)auhR7B4deDaCy|3>mOq&dndFBHzRGNa|rqndQ_s5 zuEVXz#oI2OxtV=kT+C08UQM2=54ClFY z^ii3zCug`hl!XA_nlPbj=zARi4Zz!|M6Wb}_R`UM_DVDvaV~Oy3+%JVhB%L#Ir54i zvmib2p%C)s)*FT*ro3YkMYm{TtIt5{DdWMYTcdl$wu>HS?e9l!4yV5E42r|2@w(0^ zF49VEUtPqEvWD*-5pjmyB+ma7MYrotW^28A8Vs)ky35H^WjhD2^y;m{VD8u2oFi8% zWeZa6yqi92l%SKDsmyQct1%%&1D|Zk6ZhOt=QFn=m}-UO}RG6#@$0SI{#YVSrn%e>pNhdmt!!-ys>7j$P3V( zP1CPJ{scBAP#thKx2?kL1E1!}(K!_O(y49w_c9Mh)N|)}vh1?VJ|Q@+S!5TgKur0T zr|4xo*SV^e5;NWD;T!fC#m)S0C5)IsHuTn9R7`_>;*0Z3!QHu50meKenGVx<Ct7PDbzRmo zk@Mo?+&eo#qVP{GxfxX*i)#*AHO3H~i_;M0vh-^e!WvE9!dcI=uCxv@0OzAPlP1oi zu8~avTKtMk;Q?--TYE`fSOP;2zLi`S1F1D^#5{*>`%b2bujksQC#|TOba|Y-KsK+9 z`n|8|=@Z6$MYC;~7-v;M^}P;7dxIT$qVUzCjNIy0_k#SK0c)qb>lEc5fP3{fv*WF! z5JLQw_~M@39XO@W$Ye^y( z-cGs{NdR@lh|H^16!T2e?+dNv;LYKYZ6F&xC!aBi=HlbK6nBfmETQ%jYM{$G3Cplhy>(J0@NFR6z__(LU>7;II@pu{t+OP)Dsf6g-z0`#s%DT)x zuGz|G^|`^GEc(DFl-)+TJLJSTKpD=Ms%mO_L6U3TNm&&o2fE)N3fvJc-ByXBo{oH>{KX~(Q%q67(zY(9=xFdBCpyx$E9=G4`5 zlao&slL*;E94l2L+5$LJ@vkVaT}ts+^VL2{AQCdI6fksqt@jZc%c;!a#(>Vz?Sr{~ z=C9KYE(#ukk}KhRl;3K795pPruruo2UofJ8^E-!6W2NN3s<+OK+X1s{rwg|mpOFFoF_l2d9DsKd4%@a6E! zwZ3J<^w^1KUH4VBX~5k$>fMSrhb_5|cmH6tJR*DGaQ}^3cx91pA`|cTu6vcdosYt7 zO)F&4*gLJO@sgp3#&i06N4Cw+CL+aBbEv3Nsw>oN>%hH_`r_`a z;q{i8G^Zyw6(37-y2RtRqjG^sX3mm>^{)53NF*0VGWWFcQQ~Pb>-e=3RB*HOn8>*l zM}yepc6A__-YM?t1zy{Dpc!y~(%$?mt7rrCH8WRqu&4MV41krs=sKP8K`3r_G#d#KJ_^ z=C-Y3WxbL$2VG=vR&vMTCgab}_@+GoqF*bx9La>gje}+e^m?y6=At^||I1I=;3TUavEYIq6Ob>wW{!C8Z;S zkC7KH_c5iMYZa!L%%5|eP5OQ@m8+F6+SxmvvK`L1Kb^O3NRGuig3X=ud~_POlYU6x zCgdJQQ~XxqEhYb+x8rmv6j;KIfP3F;8h_7YDvBCQ9y4!a!FP|>Z!6)7XDQHpP6^%r zai~n_<#InKaxc6{#^F`)cfuHHC)cPW&ylVI#aH9)eXjH_(>3#^>0yCg7lU+A_nXSq zz|uLE*f-SbQ+W73-dD{L9j=rO(r5)Vj@`r1@7gJd=#iMwJadTLx=e%sqQUWq&~cCH zOk(<-PDg)dtY|Il{i-NMm7*YAXcZnY^ITMIJ=cMgBMNJ|$MT)kN`HA0R+Z9Zn*^!x z^rDU7B(!IYE$+ocLeWk_WKX!sWScdoaymb&Dtvg>cCtx$I(PDCmSOft(f9W{#|{@w zKSaN)ljz0tUu`&*WSHiX?ede=2VF1NAm6smW2BzGqp+Ajk6rdTU2c|T30kOjt+xlS zfJGRy%H-RPk6e46vJx)N+P9U@F&0T};4U1>S2_NNmG3ie?oGM?drea+VozW7KG%fPwgR+XqvGJVyvcF2KVyaPqL+2iEs3VcpbY>&h&o@Qryfd zRb1dZw0HNheEm`?N7fu5+zyo*Ns(*Msa&En0)86qi&n6M=9aE-bx4N^zlAHgy?)8RLXMq-VT4BqeXAZ1K^ilWYSaX_!en~z}<0%EbA1mJrKHHXl8anY^!bA`7 zr2N{T?n8$V^362a0)?8L()M$9s6r9j-vA5Yik{IjDieG|km1d!U{vz7RCLZRh9dO} z5Mz%6G(_{_o7`U1hYRhxyt=Gs3Ma{~-5pBvSl%JW<%D)_=ajfyVD`Lq%0xv1*Yo8PNJ zo`P4HbvC#K3hy4^`JUo%?%||Fe;|aINe|ZWC^`1>4Ee5uI7Qg0Z1gVj%ATyq0Npb~ z%ejSIoQT*STbXU-W+P@pv{^M8WEn+I7&fn)OY3(?lT_VqqUp$ zh(GIm|8w-D$awR-Hkx?uGTqV1n_fbYL2&R3w@|FRRvjm@ut)Y`JMEaj+{xn3bh;dF zeOCW{^68X(Rp^-qR&I*H{$fz(*_w97o~x^LYN4(;#;L$rWbowH_1TT!;yU}}wG!Un zOV&nE54nGA2hIBiaQUXzW=l4>FQ1wCc!SUhc4O}%L`*8=Teo6$?5{t*W!|t?cXf4! zzXDMzDihw+`zW@cRG;QCt+<}etoQ7lXZzl~&qfe|o?9I$_b}VTzL?0^a9D;s;qCZj zC-k1ecZ+z^H+AsBPRK^SWxpebYyKxwo!92$YN>9+cq{@o*01QGZ2M!ru%XIQnF`+> zT10{R^OoLM^ltg(>^Gp0Ol#GPk8Nc0jw_zU7w{m#z@f`>$x+`;P20f6v~s&dwkD%j z+3!L9WuFoKLsW)+NvJ2;o<~&Hg|Bv6kD^Zdg@}qcP_*Y1*Uxu!J@S<)>I$TjYqC}I zMuVJHygSz_xhUs?M4{qexH|ypnMabzo~^?4FS1Nv9=mr>NVjkNE`u%CpYUIs_q$pz zKYIc|c9pUB{xmA4c)tKzSCax?8dHp?x>btK(VjI7Fu6nvIEqQAo{n*)Os~(Y(`QQ9 zDoFY6fPw&&AvQh<<&%D1BV3pU5%0>$fjzzsFe);hc1 z8|7F}o2^m3!)ov~v&HN9wi{ROov*%^2)=J)Jff{ne6B@*z)Oz9E`Nw3!nYs#=WOtO6&Z}mx4YfMmL1(+f|k5AuU7r9jwfemu%N|7S~I$ zv@>&blA?(9Z>7e$p)Q+lWj^Y>O#Ej$xrhnM47<bW{sZ|7 zP*>fQmOxft{$@bdY-D zhLK_BgE@QesqY7BwxeH64C7AjTyR0I+QEdPnO`sUi%AIf$Sc=CL2@~6?(}L>uPC%B+?_VeDrbq~9>XnIjdR)iwZ&pwp)6)e*Vxsx? zEmoZz9i0bU%CV~jK|Qu~jAhqa#w_^5-V2wm(vUB(#{~+|;^YSRD5nA&`!UYXlP8td zHZeTCeJ(o;;MIG3?&BZfbP$#?BUF@60!rQE$q=(9X7+^*-YKpv$?E2RH zskP+%HB+Xc_to<;KfFMdWVIiP1up4j*ja6oErRj1E655gGX`BXY8lazckdpTDz@UW zAUp1C-II?S<}&fY&GKV~hjOC{oeZZ9<(Qh$iC|Du+B5Qxi{%F@tPSbba-oi2 zd!XWg)s?bHZMLB=G@rlkUyPghGvNoZ|c)G7P^W zjCB8yUE)Gw;btSO4Ypbzxc&}rnI2mQl!!5Nsc~{~0M?Ie`U*>UqLt#FGsWz{re|qh z5$aC56bzqU5e-PgV=S05?hVm(SXG%q`Y02x9&Ey9tm`H7p67@cM&gABp{p!pp9Pp$ zGwC9$_eAQn9F=jCG&Q?Ft{ra=1(x00Bl0ge(sV5WZZfPYmB?bC48;LryN0b!uAi#k zo%`&$6LfOTdyp0`Xu58xR>^WwvT~1!yIX53xWodFgWgp&uVb?eCok!)1AhO;jI5+J zPagh)O|jPwG2vkvRD<=1dI|4=lBQ~MQWpLxTvgUYGluzxIuvgrp@CIBg*JQ3_~L;N zuv#T&fi8M=v)*GX#hh36VP5DbIYb^G?Up_YL(rALKH^d})W^|OQO-EYgCz(RgTOQo(pvh<%f zJ-On^XGN(z^_(Q4*L8H6KDBnv@39GI5Ho%+0^l$2Ryq=nX#m+nywssN3aoX7b5NmW zM``!Me3ike!FvIQlf#!auQ}7#HB#mMdbRUb04Y+eHH-=d1J^e7edSzDsaQu>@e?qu zB%fF1K9ZO+m2WM%Y=IzjmZJpSZZg8yx}7a|Mk<|{kyqa#%T8r}!AYzW8q{>f7K2tM zd($F>E@2yjBpO1-RRhM?D%x+R}I)#kS1<2$zNFJ;ds2uHuA zEa{dzs%~}Enk&cCJU{%Y6BFNg8JXQ&RB7|#E3$;GC_Hgu9Xfi%tSx$Wyew70qotGsI!#@iH^nUk-uWH^yGXw_K7Bk=fY|!?z=f3d@aA@$;qTx2KQe=X1{@ zA&4C8W$fs3v$e*=FOx2Z^Y?%+`vyF#-nZ87Tvlg9Jm{Sw0v48+TOlfpeD=RF?&y8P zlV4qM%7Jaw5IVLE4X=p7pAq^4g$%Rj^ox$PQk!#__yEstPDf3ntmDAgPx@EAJ za{BO@nYQz5lZ;lepntxDd1DJ0CjB|F8a;}t)QvG?Bg;@!@6b;2VUn$0vhJt^f_zXAEZPP+-_ z(I+=i;avIA{s1U2?|c5uJ%pU_!I6-H3LCQIXVl?S%O(6G=Pmo-9#5Ne3JsBFsr|(# zCGSDeZM%{4`5j%+x^XexU|iUBK_%m=nUB10f9#v_`cH$qAjVL^jr_GO&%BlI58P|% z>d+s|fxRuqfQOmJa)*vFx%s|uwxpVi6Vm6tbD9o26K_7~BYRL8b=~9Td!`>(iVnZi zT_qtrQ{66~$&Cj(i>H_j4|{gdp?n9OV%0zSCX{#2<37eIqvvY8a=F+$J}=JBH0z{o zYOJ6s)$>e44nkvr9jnfyE{}TA8?{4A`4j~2`wS+8|foZKpQo z?7UZbtXCz-R(V@-Z{Fwhk(owKnSwDC8dXw(XD|m&C5tIkENxSC1W8gs(=_AFIlfX9 zzrd>$Qd%?6vuY-_uHwtbvfL(<`E4%w6U+*tdX}O$^BIRAbpaze=?IZ~HY-*SYZ~a2 zff2Lw^wr%l(oJ224na(6dQ!-;VO*Q(EXo9$ORqsciF=4|nLZ)ky4hkcp;vXG@stv+ zR+C%WFUEu0I(tx!B-TMpErMU&dVX59Nw>H6mb{bU4w=Nr+#OCC5qp~prPQ2Mxbefa zP!KKkyNt{eIrlM}yrEBR8oK1pNpRp{s#!pR97&nXl6;l}F@5`ScmG|#5WNLtu|@Lv z(?>;wdXA7NNUxs#n}8kxan5?yRmv;B6*jGnd63U!VJ7-L++1IS<2ZDb0bEN1(1KwK zF`b|-nHa%j?JN8tU(XQnhZ(Jjr z5Dpj`QmQ^q$5HAt(M+FzZ#~zhZ$Hub+|NFF6a9JY~{5L z?g)ggulwtoMKlC?3N2w`9=E>aSNs8>ig&^(y#MNem$JDB)tuKieM<#o2X{Ef~4NQ zj;fd!J$SA&uync15(jr{vi+RPh2+oY+1{!{tWyVlf|(Nx%IQ9>Av}q37ok0R7CvM+ zr|TEEJIjl>Ebu3MBb_Ce$Iysj3g_^ho?%Fx(+ItjhBpR?#J)(couQW=UqLS@EW}sW zo1Boui!I)S7L%2ejch!10lqry&xYuOLo0UdMKEOeyv47Y`s#kjRw8CfOqA*lNkCEAM}3!x zNgP1ZW!JX;atBfjW?k2M6+mj!!XbU%*Q4-$D`9^!kPM*My%cxzskuh0&f+es>gs(N zHN~$40OzkGzNcK_zSIpwR(~8`Hu3l2%q7l`L*Zb{%N<27f4}>5njfFxL*pVl(VOXv zVpx$mLseUecKk!LJkxetOIsK8!nJLP+;=YG(x##o$%swY$VZxG#m{WUXuDBO5zNYp zQ729{yly#RgMXg|kcX0GYD52&%=xDZou4O@hO!KpV1vrkMB>5$BoD^3Cb7cbk)s=q zyYH8vVVf4yZw$1fu(1T9%NAUz?3S~-oNGa43DnZoHNbt>vf_2J2m@#z&IXB^|9Gk~ zcmZth1G!H9B9|c7ZkC~pu~4QfZ&64v_GI%-LFedErNAujTTe;j{m=5e1O5p^(W|); z1p$SxcV2Vm-_-jX-W+6?kSPrgEkN$B6RQ(+HuE4$)=TmaH|=1q6R6G#)-|U0b_0T3 zl)@_1%X0WnPWH(gHP#4Lp7-jmF#)|oE?HVi3d4?b9!k@%J{ci==O4WKt(Qu)fu=Wi ziVj#JD@QkyH{w4!6(}w_6OHv@)aj#g!fdt{hMLtA0(mAUrC4lNEj){?2ZUI?dO9ng zft=*ireT{D)IE@1`M zX^jH*xl40?k*nO{11Fn6rm~yW@pIAWa73leFiu`pOQy86nFg119yg2^5t7VuypTEJ zYDogpcfmx;J@b`8AAmj~$E5`xBxhT}Tl1Y$g~6p{X)*GBdKvih+uiG%>SwqQJYQ;p z$8$a;&!y@1u9a(8I;Y>L*woleJRG$pv~kI)vNkXYU@nw@*;Mr&je<_UTk+(ak+5tv zFU~^H^h4$kfZR~VhY~cx6ZybgSY=~S3>tVo=B{s8=BrdDr0f{_4FtwTeYeI1taWy^W`>9E7*LOAI0SfkM= z(lN`x>ht`E*UMo!HwJDucRN)@-!_$G_Br)IeM`CxKy@7XN#cU>G`<%lSk$?l6+}V0KRo5hn#Ht*aJ=HKQLvBX7GY9tXO>oH<$uY`iI&qV6Va64e`v&IU zlrthJW4`Eb0|sNe=l(UUFv^SX8c|V z``dIhWSj6B(Oqbivn@!#4Cl6g|2<#vYqX|bIMYn3ikvRp4Bf=}>xKXE&3_^HFQ58r zzPW?!%&&RsjP}R-+B8NV*j6)F8zF`xlq%YV*~qt zgIRVyOe$Cc>3DqU$G_s^Z_V3(v0c+gcRvD3Hw*^|fA0bNpRK8H`%r9xhp&*!?C;e( z|D*MKs8KL$O$S?3nf@Gi|GjP+XQ=#&t!#D7`X2wUphV&~l%$8eQvdI_@aNL-uYQGu z@Jv--67ByANT7r6Z= ze%8>D5O!J?bc*?3K?yT85+y_c+!~bsehdG!n}7SW=ROj`Yi%xT%zwBJA=ClC~od8Wu&2Nb~UHmWd|0_=#cNZm1L8^K2 zpehE%-wze#Cr4O7^REJbk~69TjoiUEw1C$`)6!yo8&vy7Nuu}8+E!fD(gBoG<1B0V1MhUBZcCTAn z9#cd_*r;BZr7DEO?Y5T7m)Tr`mC?Gl*32Df7 zgGoK<`-h&bg!=zYhSwoQB{cmJAmR{f$6;iG{R_o4=Kq!xsOg{!-u17f^Y6382ez8Q zu<`DHg=`nNR|_!xrI7g|#U?rjcqm>{W7Vd=(D?J5%YF!%8@UYWFbZP(M@5L4ndOB7 zI537gUDb}C{&hZ5>L>Q0I}DksEhG&5Li5Vck?eeFe1UftQ#M>o-8CXKJiM#w`Y16G z!Xy$B>nqy*Uw8pAQxr`spDMQ^owdvFf6fqcE>H?GE#SK$8Mud8M~meQlX#8k*#QpivC3v1?QK;}+@46OK|kByy@sik1a zwl+HKsy3Ao9&Tag@n+jkctDgn`|?_=!nhFk1J++O7@rlcn0-t^8bbYofi(sOn^>mf=YG>x_=241 zFTp~^?t&6t+^x>yFk2ZS@e9+8%S1JN7os`!ss_nIk{7CAP3?1naI9KJf{EHLA^Y1IP zB_<6q+0Ii^RCjQI7@L?#ydU}q;Ezt{|Al+{eR?CZ-Qqje^P@`IY2MYq+`oVJuS);@ zQ@f|gtlRFXB_1mB=+a0CY%Ot61s8)es-#K~&4v2x34>O)3XA_m*2`Xs6!5tR-Pe2v{J2hOn^d zu(1z+Xx8?0a%U_R8uw)_!%#;5I?rdP$NbpdvSla3uR81ZfYl!h{O2yql;gMa`T5cS z4`z}zs$lK11Yk5e`dcEYz5I{jR0S?zD}_K)s>#l~V{VYOnXK<$Q{P{EC`Gq`9R=C$ zt$s@O=U%0e8|$YE(Xh3jRxSyd5nZWw3&9< zH0;_d;#%~8*PQZgg_V~qO&3Vj{JH#o^C?MGD2vY zKR=n&;g5ht$O}ArBlX7K3DwAyIEdBfP=KNm^M6LMfA?JamPkw&t38bX6HrvzJpDy( zl*Ej@MF26YfAJC5kHO#A+=)s{Ga#mP1xaSo(k_{9b~$?Jt{B4{V{KX<64#nu5N3M* znpx!LO5^?3V~@@TD_UjhuRJw>5vlwq-y(yLbuINV1`0icf8j3D4o3+IAOAKHS()+0 zKqx>HBH5?UopD@t1y?08>z*&ntuVDL{y89Jlf`6BI4%Ug z|8|kZ?k{E*i(T8+Jh&4b9sRfi@69+U6foHk3TTO$Wl|H-iu^^^^ACwo(fAt^*N)tc zg;1dVyH5TICHkkoT~m1$qP_HQl$J|vc?1lT=ugYb)A`MuI3_12l_Fy)d#89ZnhO

    OZ~=He!8h^t9uGY zI|>BWm*}gPtKIJfp-y8|(f`{T|FWvSB#L%HDFbZk5h&na{_%IZ15Z#wLP9Az+Ocfx!m{ z3e48mCrS*lEeg+YfGn;0v8|Z}Dla8}MRF1)%)@ehYj<61o*H%ndw;-qD|?v7O8QQR z>9Ic?Y?11d*SEE802gc@ued#BwYA-ThtJfmuWs-HUanlw)>ioT%y78z?o;;LSNa<| z{3$T4370InNuFW4o-A1M)86`Q@_Q;74dHoseaIbu8dmno$Z{9jFC$~BFKiKDlYE^m zOIY4$jFFMLj*cVEE%N*a-}>q(xkXm54Pxa8U1>M;{|TA@&>!ZpuKxD1#9q4zDk>)K z=SEmyTVZGLfx=7{1gwizV`PkD8p ziRojYI2|3O=@z+$3xhrYKJECHt1xEDS6P|3(B>Bw*vv>3A8AtjhaLUDyF~&atkJ`0 z7zfhQ9dnHMbO()Q|*d7SN_2L}iOxlo13523C+OWRhvu zVoz?%j9O~gS?fAao!+bal`Q?lX1pAnLYd0lYV0De24*#I8f4B~esf|% zgWt!RnZPlvpkPglp}nmz4W|Ac4C!7PQd9dSi^TR`YAln()`fnAyn<}XY!ehJb60`qbw7b8rdi|2Xp zt-3?7QcaBRsdt=X9_J!4DxNdG4H^U`KYh@;;9kju+zg;Z}m)4EC@I2>VPd4c7}4)Jt7`TM_hH63t>OM=xsh+J;&P zScUP^l064*>&@7*@2DjYFHlo=3yx#!ce_y|<>xQ**zM|E+U?2>67kk~|Fx#Bu0>(x zYX8+rr&3(QnZD#U@Z!u}_8;!(pZ@TFLimCpuz6%;@(Ci2=H$B6tREXlp37ugf_f^Y z?`}dX>~+PTL4Gr2<81sqNzjFw0DFFRB08LsNwxV$8oj|w^qMh4ertCV0Wv*GoY&QK zMb1g2w?pZ7A608&XAC@iot^ir?Aj%La<#mrZNrISQ(-Y$_ql@>9tWy_~%4N5izLEx;cN+G+%1HF*}o$$rJD*1fCR$d{pcia-}&EY>OVjw(dKr_ z>?Pncu0LIn6}jC!P)}v3WYl2e#9pj-QG(v-1AU(5kB%>WP_U9Vck~j*jlkWr3|~Do ze>ycX5slUgJTbtd`@3TPJ?%p(&mI*mAq+seHajp;{_1DMttPZ#hfHa%8${kprtYDx z85V=DQEUJf;>e5;_x^pCH8n7c6>Q7g?H5$U7HZK*PA;otjJjl?%75a`b`rV6_IpqSi*G2%N@j--I zyud%<;NKQ+v_&J?-cFqVmOQnz*^A*3c&jE+FJI}qT7>#Mg>c}0Y(!viEzTqKZbXE= zjy0;Rs0t_gFea^GmMb3s%N!G`%)3|jmcG)ov>fzHWtUs!s^NqM5@rFO3C7*N_HRbA zV2<;1wLI)TAZ$%A*`@LaqLtnA7ArOu zmIx1hWypJ-U0I=s=(WZUT}4#;`;Q|HJlx#JpuJf)u}mfGPibRCsmJcfA{)2-lwTRv z!w%G;%PBQ*@t7$WO~h?>OvR3gR7ik*G`e}{iN{Oi&@jrt*38VcU3=H;y>_~+wmZFz z3MsVgmF$#T)!PCB4vZBpFXxNCLhkz)Tb7o0yRK?p>;5X3N1|$36({}xY+m&CbKd@- zzYn2X!<)CzzP2h4)_w2;5!^E~Q$rbbaDvgzrO z0G+sYP(MTBlK@+DcwDJGYt}~n!>`PMKMy4rV%0qj?6-3-roUZkspCjt)Hk36Z6JH{ zub1uNU^_k@=h?7D)&fA#vN#`U?Up8Z?B}`L3!>z`aVc!z{re_+a$$h&FdL#&Jd700 zUpTpk6Zrl+9{~583SSL!>e0Y|J_evP!E`M zH><7cO!K&tccdnLyxq8fEc;ivp@5VwQoJ9HJuvHYxbKymhIY3&puH{plQRALz_e@| zStcv+;+%H@AixK)zP46NGD%0@_8Ow$wQHfu7(Fo&jlj-R1C!JyYY&tDLO9{yev=Ks zxzeP|_Xe!Bcel9n##9{R+%Kex<*t9DUrb-fw_ z>`$<5m$Nx%+#T~B!Tpf3Ao?Vb9_ERTkvi9es++&QdNn~{4f1G8QhU1Zae>T)Zd1c= zg7R-+Dh3oYy|hz3-Pd*Hy3LFh_OXmR6I=NBxTYli%bxCo^?aXFy!0V^f?ks8pHhFL z8z<<>t5aj&01n&SEo&?Et|NnA4;=jpbLg$L+jnf;EY%B>a)#`VGsQeaFj23tDWVGMGxzIux^kbhNkB8C8; zC}k0#{-}8qc}`_E9j`fRW~Ef=ihhb}{pn^%8_Y35dsQj(CJ6sls((W}nSj5C?Y+LM zmGMM5%oYX*-Zl2l*fA2DaS5c8F^RmY-D?qC*^ZfVRPU%LYSsVr3;xX!Ff~)GTT_^ zzpg`$DSm;!A2b6q%nl(Yzyr+LT8q2&>6qsI4>+e67_YPu*sIgqIBt+!Reh_j7b20W zCNIIM(HoXd%h$WUY$+VedY4P)a~*Y-uvoCu=ec-QJuwxH=DIs8QvmoNH0^$}j7(2P z%1cTf^PO7R?C?2ksl14N2eDwPE$kQCeJ8*RTy3Bn-EuVFr9Q8t-1y}N!~Mjxl&ACJ zz~y5VqfC-_S_y7-M(lpSB}yKHyR*IzMJ`}R4VbS+ExQ-hp%?7T>&-8Ncg#3OT#&B{ zPA1&1=I0>{5Sko^O`8^Xi;JiD>{O7A!LC|ZrfR-uQ`An=xBt!kd5a=X`&WH%AK2L7 z!7oH8DNBW3TTxTcF2`HxV*>v1XLnciL{E25?j-q^HhNr1o<9)1Gojp zw@eHZi`&r83OrpSM^I^aaQk)PrJBCo5V5yVeG{1mwo5M?^H0#oSj&%A`o&lX%l6H6 z1Qb)x*E5?*$K_!uy#0X>>25|$fzHGlsB#W>T({vdb~Rn$P8+8n=VMk`&^ET~EU01r ziJb9t*yPbHiqk>uM2;EG>)&o%(yI8v0IP?{TRD?3!0eR|^)J%sV(HYuneSZFai!jk zK>woC?cH19A(je7MXkNu_V{RL=(at~X6uZWJM_?OZLOK@`Z{*;$;}02Gcv=)c)Wkv zLhp;mXFIoxZ2<7Susm-{EDBn}69?E#;j9}KGr3;PdP=fQs1})pssBHGeP>vcThp!; zRGI|@>1+$qRisxNC?c?F5ke@Uv=BN$x`=>+fNn&Tjz}k=CX~>tN|O>2LKBdh&_7ex4bL-c*33Qk+%rMM+$!}iEamu^jc;FDeEdw5TLrnT5Jy7Fm*}q> zW_^&at&YY==a*x>_=o2cB08P_#fy$@J~FSoKp`n+7kg$NPz{3`Zs8U)wIgYeNS#)D zGS*Fvk(l4~4O|Z%kB(j-NBRx3eI?08I?fF-Om|>JV(R-HE6->p$`&lXQ`*>QYJ-l9 zsqD!0a2fG~FTq|O$xzv}xH?yeny%~7XGAt0>*$Uwcbzxg9#E70DKVUwNtwHosA$BP zl^`mWn5I(SiSXIhmsPsXdnL~Q3fQyVT&JTUL!CxA_zC>JV;dDUwzxTu;4MB2#Ra>7 zoqU+{P1=bRc#LwTN$|hacuseZ#nz(5ZkV)P$rDRd`9n1Q0SyFkC?MdR>`(i)z18YH zJ{(=_>sz1prv111kULekeCp||x9&VP&TUfaG2ZJky~q;sy^_G&hsXU{R*17sn!YAqW{k@eDbqg`V4Jp4F{q>fl3ijX-kG*I)Kv2^k4zM2eOz@rfcZuvD-p|JrS}zwi1m0L>bzv5Xi7k$PlkXRA$RcJfD>%l+ZS z?<3|C|5sayNr@s06y1POJwFy45i5IzkO{!-%lV>JwGH&knv{Lzl@iWdc6_piA zA{2Ld=_E{CrS3UWm`T(;=kkvqcF#Ei{cKCB-MB9Bo;)Y0J?F{cC`yN08YURqQ9>0? zmeb76{}>}5Q&Ut24WGCY3a#!aq8(^{njN9%q=4gw56f+JRF(Ffbl({_tKc)OTI%ws zvi`~PH~cWPe4WmdE*u^ayym~Ib@luyn4W^G*NyEqGq70p>gFBB2TGq3!rIh~4o!4M zHScHGU$%h+_()n9JCbdC1D~hDC{+AmuQh3VYH*=)T~5e+^^X$J=&@2oIHXWF0B8T{ za~#8DMtkLgb@jJWUqGx~NF%yQOSwy%#K)dumxyvrhK8 zg|ol(?qA(0b0#D07dAx|Ls1&DI*+}&HSOPT_E+ycD|OrTS+9jUzACzYdKK2DRlPB* z?#w;-$b8s;2eUAP$?@*}=_U|ds+*rxHDoti?&?^6K|`1|tiK%tkx=y8zBKST;ue}? zLSxos<3)E-(@no6{rk%mZWh9t2@1IHb3Q)eJGR5RBE9yE-EGb9Ag6TG%mdm;!26%aGnlO}cs3h^JB4;~}Dy9Za^qhQ(FFHAq;B3hu<* zs-4`w6cWF#o-CFow@M$LjrAVjzFD~ogEzqD7mw1k=82WN$$2kfd()14%5)~YP5$Q0 zBh?#udkA}e&gqjsZYA7oqt*}mb907^D!T5>L=$z9p6Yww5+fe@@3Pig_^_UAHs^AakTMw)-n{wwznt&?7QS90KV@ps zAkvM@?~WwJa{4but<~XMhRLYb!|^+Tk-Ioj4Q3uA5^%9 z&k+0X{$SADRM{P!f5LC8MCWsRom{;=&EV5zULh5fm2P=R@VE`BWX&>@>eAlV=59 zodaZK(%$c#Yl;;rV#e(4Nnw__!IOWb^`gHo>TR9nn9;POXX*OX+j}D<+)>VBLem~E z&n^y=<-5|>KKzF7 zl}1{RCnAF-76DuI?le6e`_`40XQ-O4QlsoFuaf84-m&=?E50VCrYD;t(}wBk$rXBh z&Ht89rT}wq-7Y5O!F10O!XCo+k7@ravHT<)@jU}l+@qxJjj$^{TzN2-u?($MDN-1- z!)C4K`g)W$?j^*rYST{g=E|2_AI~*M27yegKC;;Yi3L}6eqdIV>5P_GUkIUEiEPF8xh0)*Rdvp6+7;NM0Y_tobDLW@8&Ip!fPgpOMzS^ zRWpSgJQv;wWM@5pG@2B~ep`qxXH)FShdNV7|GhRVSE4$f|JklJ;o%D8STD-S(W~6E zwFM$B11*r&^zG-fw!HXw7kB*v6o0C*4%=Kk#USKk$RYenZChe+zy2FShkP+7gmrnT z(2iUC7q~`%^|51dFCH6fY;1jy%hjHq3Nt6DH1=j9UGW`$o4e-UDr>H)Y)`QiB#3T> ztahI}H=;hj&5u=`g*i=7roS(A?9Et@^ee2@H5b&vt&mueU(~0zw%v?=t7B_*pSv^G zFqeg-w8s0 z!g>_53-Mnzd!^#PdT)IyGU-(`KDQEHU>uIJ=L-g#v_d?F{Y0b-d4}IFgN_j+DAiN5 z3{d$hVh$(sOvF9Hb2zKdwKZlcZ7XzkhyDKjUmo%Q0Mu2TSx+baJL7Tx88@ zQ#q7oAPZ4To~%Vg4%A_A^O;y(> zsVFb+Mke{4*X7bxxch-i*CnOwG*#6hP`qWt)M0~k>X-?U98@pN3BxV?Wcq4$Om5wDoOrD&Mj#SyW zDJpId7r3|Rt)lo7R&NL))Qyf#Ok@vg00U_$TU#N;eLMlk5NwHkhlW*gb5f~J^Fx7y ziAYE@M6G2+aK4JGYIC*UiJX&HKSMGW{N=vUwN{rkCRo3+(u!7I_GHl#NK6@suJR<4 zohr@-!WwtruSgELpwmzfR2f=veN*3TWR;Pk1Jx?HLFnf01_2FYXddNu7$~O}SLR@n ziz>>`Dz8tiwUPa2e?&XE#I2$mXLOgh2AB=QqeVu(r#9x7S704o9i`L9dlRfCC+aJ4 zYxUp0eXC07rayrG3#jLRCvBzRG1qGGDO3*VNf?0g0UjswRVU`4lKGWtM<6W851D$@ zo%-WpLUYJ$Ey-BP2f5tBG4o^h%ueg?r=Y}lY)Jo77w~8eFqaH@rrvTs&dY}D+Y@b0 z0@8{*3tNu7HkmWt^)a!NXcfe(SU})|{=o;0w4<&zN#F+gk!+bCcEH}Co(XCvo~Tzk z@fp($wvy*nyNFe|P2XN$36rq&CM3LzT_lzCZERS5n#B_}SAVoX>c0ge3xm`n%d0Y? z&W(*Uycx4_b1U9C%ClkYN=di%aXL?)=JF2VM?ku(}n)3djt}bf= zF?9Cc2GovC-q%ZUuP42EZ@t#>4hlP9`<;_h_*GKXnuFxsn>~S4Y?4+L*O1Q+-f5fo zm_z>8;dohR(Revu`GL5qVf!C;4Rl57+Pon-QDgL+F1S7OD?sbwMk?>R=e>g=TPV`u zm`P8&*{xU-@3-qYhqU*b`E&WLEjEvUPM?;!$Y=k}Mt0}j<$YKxqM4x7U0{}NfYWSm zBxohd7YH4~(%+%@G4>6lGYBg zPf|^-!9B$FwV;s5I`; zS-D%v^7=KU$Hv*gW_zn^DT)2pXF-WqPtG%)H_FWB^N~nQ3av$pULx4f+dZ}Tvco%; z%5ilY^t~3#dlWLt{l+z}GOtxMpd}M4eRMc^m{k1M5*7$5dtpvqPyS$iVrjCwngM!t zd%G$-i1qE3I12&NSG7Dhkr<7U-*3iX0u|d^t22xD-T1=WJZ962?NH|f{rYq7&f*{V zGSl|yzrl3CVKy7Zim?O`O9mMM1u~2_i*419-^|Z)6DPh?{Zv_aSi+HGF#pCbMrAN{8U*nc6_lA5C4&P1y5TE+o#q+?!UKfBEdf%HWQ8`^J^ zW)zi0S(JB1)z&xT9C39*p!Q4vce9}WMk9M{b5NEmsNhKMlXGs2$XS0l;m3^K`sf4- zZBD@C#MCb=)E@g}dIG^RWO9|Yj-`6fa6QI<2O4ee2>p-4XeB}C&ZLEga96) z^kb)S+x?L=z%N9YgptaMGU}^$fVtF-xcaDh? z-RF4}6xi$Tc~6XR+GNd*!$DC$s0J^A?F* z1t>Vmbe<8_*Xyo2lGgrUYYLFim7Ke6)$7(HOKD-SSy)z|2+#+P;C8r?Y5EZ<%@l|( z(9yl@_KV%3>F@tJpMY9co1hXb?^cm%|0Yl3Wm^7rD8TgOQ=!nKytTkH9vd0W##66J zA$_+&30+B4j^3kRW}V^W;hAnF1b(auVwfzibMVLjSQJC1l!MjIQlUjd6gy;9v$li`8V^{NDNl2>*CEBrjwh zR>9H#+PF=B%WWoLo5-N2Hxee{2z1u?WOo_5(IKxfj4y-nVCjX9XgIlicPaXsPs1J9 zhD3V~0Uf+N*&aOXzm%aP6g}!LTS~D`Ghq%I7=E8da(V1%L-}FG7fCy^{(Zrbf(v`Q zgTu{I8EUROi!0OJ`;Ag1R(N+f2Zbcw_bdqEF?>dyCiK~l%nV2+ukEjr^Ons`@{#mA zwSe{ZwU7U?o8<#EdkaG>!*5Xfw_#9NKZD+6-GbBgTokI&J%oBqQilCJ=na3%((~qh z^-#=C(seoEC#*o{1OeS(PH^hj<+Q)owLGcCt#W>MCPPJe!C7B8`WkEAgDpc;?fEU@CQ%{p;paV(_hl zL+sCtBk&xQD#ZJFP)dE#1?$wR9eGlD+qwRj0VUdxY|5Y0JiV2&x^}3#Qif;0+o;8* z&pF%gSal}7Jz;QpB5U^}ac4>#BuL!ov#_gKMk2#r*e&gW?TDbDkP)p4HEPC4fm;J6 zrvR1nc)%y$+Ut*ZxBb#|h3EvJ)gSFzT2f3MgLb@b@C(wuqJKXYhfM$&I+5MA*K4n5 zeeTt35}OBh;Y`787HF1avJ}uD^>zv(hpVkGc=*2#*Xyf}NK$aUQn9=vgK?j&Sp(E{ z!ZmIA?e-P7$3~CBM7=$>Mtf&D&i+9D#`#R^a;Ui9o^qIfz4klm$91CAhkhaFEaNbN z>RAu2fvBj>(=?;{e^T-tp?AI9%i{(er|vLus`{C7H4hn^we+H8s^+X||F(9hX* zo1Gcg>FLFL3}|B`r4H(-X1$3Fj(`*$q00Hp>33XSP&3! z_CEa4#yh>#KRXb3IUT*eo5Y?U(ZkPg`m5#?i0r*>rv0!#^Ntw3{9>j~Y#!4Hi0Y`S z=)-V0fp+r31U&DA`|MIVf*CnD9Myy_wpkpX?N0$B*%-)InGi=XY?5tIzsY-i+e`AI z^}qPzRCLV}7(h_FsFuR~J^f`{+ZjU1$49OM6KSa7Uq-awXzDYNiZlF_0JI?&r(J(?GoIYZU;TrJZr^mg#Oqe^6I6SiJObOeJC zoH|bPF}eRZJpGZq^b;Hoht2!#a|!3JC;wjcIl`Xa?jAHcH{=bzdtSm3VM8u=xy%|8 z)*+)hPx1&T(75bJA1ptawcqozR;guvO;7u5{!YUR=QHsiWqA5C9r<39+?@DSI_)VW zD<3M81s1`)e3RyjoLC6}_Tz=Uijk;hMOXVsF)}?%`Jl8pb#eJRS0D^j9@RRTlW__j z`!zdpy~jBI0?&nT!jub#_$efn93w1+WOMNk$p&K-{!I>nJMt`JGy%$lvu<+MzwgUSBQ5=42-@0o3Jw03`|Nb+h;4#x&bHxB{!OG5R@(t3Gg-PrrmvZ2 zrF{!TWP7uh0!#vP2tNZ0$47B-juy1+{N6oIEE z3m%E&Q#l#QC*j~dw-LfJ#;J!jBq}fL?tIQSfkwnJVWs2jRa9)&TH1b*dPh)=X?CR^ z^S3y7SNE8&*%0v0Y2XgVLcj{Jrl+?vFR^;vK<@pu^^y`J9?vq_k-)X?zwUo4Sjl@U zC@X7WBRmew5_rjP|3VS&Xg-~t;Wis8=jnApz$DE4gn+1_)twam#F1GX@&pdG5O|gG z42~(y$NISpV#sadPKGou&EK_)*(m_zZ(l}>AUvSC;)sndKrc9Lt~Au-DM`r`YxWh3 z6~fXw-(QoH<5g2r!`@uDMAJg6cg$2 zH8$qOy^ve69KAYR=;#7v}@%MCv(Qeiu4__%Q`nqq8_gUPQ7BM;9nODZ2|^+N$W?3pN0 z$3823W^>!|vGkXLe(>B3&tj|Vtat3Dt5?Mk$yA>Fbp;Q9R-jVi{@~?ZWXRXhAnWr!#-uO%{#*7cnohTCyQDVlll9 zu1RI3$R8`AGK~~(YZAIi#=CFsTr<<76*6AbSg@U@$mc?kFDBhB?pP95z&Iv0{=8|` z7IJK2Hovg2%joN8S@#cZ4E9-y+TAWDuAu(!-n&gpm~-?aAZB1lpqt ze@B62m1A8K0Z_4Pz2p$f!K@2IUK+5saJZ2fMEFww_aAotCW5K7AMZ~vZgfr^R=2&S zX71*83GhS9CMcNM(5BrTC;@-X1Yi&ub`ScaLJ$ouAj1-{mbUH!P!~auYo?*8^ zaF@yb(s$&J0GU*|Lh3`Ldr+zEli0P1GQP#USJSks1Bv56+q6{3(F1NXDC84xYq!S~ zkR7-ks?12|ShF^}@a_9;+8~miyjP13*t~6^kp}d=V#;avO3*=B0kPq=G=uhkpwh~1 ze$;hDG{QIx9uXq2^ucjln6(gtc{*sr9U-?m>OB@uGn&fU00b9!-lK}=3A1n~B7<6g)%7f&r#3lyOcFm8zB$(VI{63jLie zfm|w?m%s=yGz79W*H<}J*z-rt;*#b|^rkp4^q$t z1sa0fH-Da1cUp)hk_@gs^=2^4!ElgTB(!;MvFg!Q$Nueu#P_Pk&`^v4lsC>yZ(!-g z;pc^&Q;AUcBPl2pesk37(_jc=O)lg6e{%SLA!l_x)qJvz=~}=iY9r9-_Ye)V!%lNK zu<$9J8usxPtok^M_Sl&jzh?4Qp}UOfE2zl0?|Am*WdItlx1@A+N4cz{YW>E*<_sBx z?B4l1QU}Mgnv^rott<(L&~k{1YhY}Jpt<}29bTIkK7Sy28>#v+ek$qug_8m4H4P2V zWj#k9XSnYdgr7eHsVcs*av&UP)`&&hFOGV%w}*URTx{QRVop@O4D>ePu!oBzmM4~k zR*K21DOJ8)`I1;r_Pclz#xIFqjLy9~CIg6r)S2BC{7a_o~VqKE~Zmd3*w} zJFBQZqSQi>eMb}Rijuhmz_^jBs-r46KwRA}c<1{F&|34SzKiDNzLKb_VC$VRf&zXH za_3JdLDg)_MUD;!l${^x!88?bbrgV17S6Up_LcBaRkJ?=Y5LLMhq!`L#TbT>C7~=5 zs1Zi530vR#--Y)f@b+mcBE4EF^dMf?`iRk!bn81#Pmc66n0$&2ZF|&5SeaV^4B|5Z zG?$EHGZww#3b!3yX?R+C%^giv>h~{)Y#Nx0N+X%r3NL3GyDxC!bfE;$dHKpCKwp5! zHLK=oI#~X;E*?N~8Mx^bXVXn08T)LRHz!Kd;ZEJ@D>p9eTi^8X^h}`Xl*haT-(9xy zlA*J{l8>tC>61(`2^C0MH!BC?HdbC`uK}aC-(JE{k)?iA7kly&#=(|xZtvs6zk{Kw zfu3yk^~VhMMXq493Uj`z&LXvwV<61p;0|ZN;dA%7<`FT2b_h;6^K+zY1H}#>&E#4f zKYV6J}suSuo1J4TT#WBrcQ_iDshWO#1HSUw9cC&V`;!3Jdg*|-e-1!A#?`%(b zTr(^)fa7=LJ9?t}hOf-nBMa!<*OppSZspCrAQ4hmmsPpJvX;&62>-sjRSx8^m#?R-Z(VHe$gUOIQ8?roi zOod#SMj7`}UlcqM?J5?WhK0T$m)Lf+w(oPu6ZwV&!+JX(#jBR{G7rc8<#7%9hlx`1 zid}48b`IQfzfv+*Ic@tdS3j`wYA@G^9&p(=&A=q~US8*Wz7fFEnaKJ}vnB=moc#s_ zN^Tf;SR?|L);0CML|*d{7&-?9UG*1ZFr!lt_~aB18V{{SyFIeQ)SgV^NuPgLc3_RG zuAOX{6QpLTblwot1x?eWQ7Pwd zf_7ekVouI>!dRJSW(3fwlm}CeRs=9khM$olF5leL1Qv)+MD=Db#ddX9Z)h_ z-^fDbc5=McF}D8sFTHO%on&l9m(+|#8mr7X?iodhS&Q@}Sy9fMCAJndP}rw4b1BSK zb1w~Q9P^`6(Y#2~gywDH77Gv+axI=fV2@sL0yz?rFQT$70=oTL39@|{{wV*!v0B#q zwPuKx#pN~vcP&q2s6AXP9Q&{zn5ITP+V z)vwr`Th$1#(14*)rxg{0X7O@C{FHdNNE5}HvmfcYhO2S0aXc=rmC)bQ>JIP&nrGEZ z;kNbg!vKx4F8oi!92DbSvH>8v=2kx83}Rxo>?ksh{ycMo)zv~TK5`X?9C(fa0j!xQ zuoENtVVVy2oauQpR36js0aBntZeDZ9v^E)=RUlHCSUI5FN;^lNSV>V`t2SR7ImGlv z75b|uD{7?Y`?-Ytj;tub`iC*%c4WJwJcU~Y8gi!nDebx07(1I4PfG%59oKj1a!-1# zX+>gKDKn{qe0;5~``YB|L_Ciw+1&d-c$StyIdlSmj9&7QQFfNTIM}*S_$=dRCd(@4 z_swAU3|5Np*Cup^t7f<<7qpUY#f4LvW&mg2U*M5a-4FtQIcW)pxc>Qv7MTPlziV@z zgRhv9<_T~^05#(sSo_*+yl%b-RJFA-O^`(|kO=V5WAMEI zgIB#@PJTVe&Z=e%Uzyfi%FsG*!eRCG_0FwEg*m)|ut07fJBJ&epkADzahI`#9um9t zyGcS~-~r+1I4B-couXMfxg`=+h8dC3Pkv2C!y(6U%$=A)J8)agpQyGqS_W9bJi?0^Nnh<%JIU6yZ#X!nJ3&%#w#gZXcCTg0>aeNLs7_H zSB2Ii(HdZ;f`_gjSe5grZun}XuZN0c%u84tG{39Ph{WSE_6Jnr$~)uouB2!@9<0Rq z^D);#P>U&tydceU1P=1E`p8BTn)Uhq>Vr|I0(cGh>N=R?4CfS5rNIQEOo}Fzp9(5o zNkZk=Qb^1l%QVkuN@9OQrE&>iPiVYPE4@Zk!B|NjlEkB2z~p84!rbE`WG}j zt{+2y0|@h0o+;>wc_CPt;GI(83Wm$K#>;HqHMhJG4D7oT#z6(OvLEtH1_5Ug`{F%eomNmAlqTX$8X!> zl~1POKf)7U^bOo*MBO!owk{+ftmfPC0+C6dt2{Xf7&+ z95>GE5)TVkHVM+e{n6gu*n~FO!h^YO4sm9W8h93hVw;@3{FH?k%o-|imjGoK0Mk9j z6_nb?@wC_5?As(i<_M;=(^QnRVPR2Yc zkmyZgHf-#p6$SaPeFA-!K+nOqu{G~cUyT`i0v)i(_k~L#rAR8PZN5P2*n$D0u%Au{ zM(~k2y%EnYJ8ToTumvO+j~aUi0W5O7X-FoM?)HDxumAhdh-s>%U*bs=jc}OItiC8x zA`DEfBRBeX+}6FOOeMHNQ_!q)p_Q-%i`6l!Hwm?#2ZU^%fI0u#ImOw-z=6FafXKrb z$DqukA%!ue-jcOT)hq&i{UOiI*57w!K}57YjPfPNH0=XqTx{ch_cS^hzj|%(UxJn) zKkD*$%wIp8Q9^7+h*LrAO2(E8&SkKcnkl8~vMJ7tKsR`hT~i36)MyT?`O) z|EV5$gBASzQUL4Q?1o-+oemj-cd@VhA!A7m(*;90s66*z*UgouMFlzUpsG1dJb!fm z0nKv^SF41l%mZQ8wJDtLDye6Kc4!1ms@>C>VjeQcH1~8j5!TSG(j3=^@QamZm}y39 z0!yzX=7?6Ij4b0n|I+}moxn5!Rc zvjCu1vwW&&RLy%sR4=ChewC3uY{LYS*&M!5&g?UEZgt8xeEZTEN0%`PNEU7-nngZN z_(&6wDV;e@nK_AzI?JE07oZ^or)bP)8CovmgB!CG{CfSkWkBBm+Aumg+}%G7w$~^s z4wxu{?NSNYg8eLD=(ysx){H6L`fbPEyh&HR5fqeP{!uCdGdX(jqy^Y@gf|vLH$H`# zq;=}Lt~YW}en61p1SFlU-)w<{3HgFBtMqZ)rfq7&YUu&)RbvIhRoj(B0oCNlW5Nad zOIjQ$p$sm*Jwz!5zCajNl-1qJ<;L8m>;S7KPWAlgZR0J&hhcA&xV=efE;;uq$~Ho- z$A4@~Q8Hco-&gCunA3n;pm!-yW&*L1L9&JQc|D5NR&ofI`BoARrmS@YSY6?|whlYR ze~$7)xfEz7fmLYomRbJVmc4Qkus)78v%#rHM}QqP{VZjL`fZ(K_(~>2I@%p%0?xh$6sxXNH(TrxIZ3K^}YHfH66 zl8PI3g6mbSzM7$`ZH=f7l{6)x(KWVxN{cgzC?c}-0*6)qVXvpe+nAyzE|gBds777Ed0SVMwkMSl4_Db0)S}?rm>`JAuU%ZU{{B?6~!o=s{8B z*8&J4^7oZn^u@;kY@o(n$~T-JRP*t6KF(?TrEI*gm^a!iN@FrJqixrvnbPR#OGn$> zlzr4b6t{Wz$T1+I-!7&Ki>phiU^uyOy<6D@A`JZCEZ>907~T{_gt7ZpQ|{{ZEl;zD zDWwR?1fRHDP|vC>cgH4$btZOygGRAl`T_dreKZkr1W$4yydN+1G>5wjy}Er}EUd&o zG$lisNI}E%aoYlX=I|;hE!1f!ZmGo3GUh2{sDXBRSxE&Y(%kCUiSRdI`Ju$M^YuWg zMyHq412_TgxV*08p6lz*kKYOuA$4Z3+vybdgsDtk20_OiDQG6j7;pq@8pwTP!l3(hGDP|5xcL4 zfQDrK?u`j;oNGdRr7Hh?W*E*CZW#ocAG^ODH>6Uwmw9QYTd?2a>6^e^&X`VMq2!A? zSs!pm^ze2lFr?7eSB4JGkpv<=Swe3#otb!=vvrt_ZN@iKY`7E`&ZV{x>|jI(7i$Munb$`a{^#%?p9 zA4nzdN1f1lBGSk-bN!Ep;e689*21lxL8JpxO5dA-O%K+9p@1EgynK|VLZjyWek-jo zW~kvLV}z=odgINR85dNJXEwX!=g)DG0-0WuJz!0c%yFEWT5!FSYf3)boU!Q~%o3tR zmu5Kn(h6rG8r0q%A7L3`X+NxM*=uZnmyAbPS)u!yu%*SU?D~@Zrs2(=2Ll4u$6?{n zlIo3gcx9$B3W;EREFS&jXBI{+>bCaAQ?TF#pQP8{mkDOUDL-#t4=OKgqw5yoV*bv& z`kM9uZ)yM9}Fw|gt5XI2ltg)MVbD6ZaNbM$esH?qIEM6xxDNb zVj#M32sUuE_N{c#h;F9nobtws;TxQO;?U0<(kj)b4IxExm`J79TRmM;dsBbf&lod+ zg{%Tb2e2pm&xJh)RJJ(h*^5xutVYbcZ7~Elv5GCMakYM8$JjNnj8X}cV6I^yllx+| zIUNLo%Rw7gbRfDnh|VMUsM8rn+RYX3OZ-HP`a9)d(8@h`U*Kz<&MMii412cMmBD?> zV*;h(DNdKHyVbkdJ)Xb6&8MKCVB$+(lg9_OqCx&{cuPxxln8r1>G`(bj!6!(Bt8u$ z+}vF%5*;xcMf?y=D)hQu_4#vOet97&-tW|`n7APZIZKGCoupYT062Gte_>%U)yc!F zyeP#K)mvF+T2+w4!G&6ihJ{<@!=27_8Ebghz%z?F&AkmPi8lGZ4hNusd{@BL6)dmG z?w1Yp!a8q8Qd4eY>&-DhmRs)ehvK)x7*ctmT9zRc0tl1Y#hZ6*gD4OL>ikhdB^!zc zPJiMk6p-Sap<}V!d@MjMG&>T^Y7B)btoNeni;66sxu;1+2tF*p5&Q&*QC(?S>u4X}BmVmnZHg7mRe`4HxYz=>FG@xJH@0=Q(sfQcB7 z_-?v*kWW6~f5^4co%wcu)rMR^IN0A=Jobcz3kUd=+r|s>ZzS?Zer( z#XlH^fo)ioHk44S@OUFRY%<$eDcHhPzi?lih{-?Q%PdR8fCZ>vBB0S60>W9Dl3xY! zS$yIFLuUKApd?%6p0fLiwZVU8P4J-YivO?sP<*vqzt)4Emo znbzqyh(H+E9osfgJJ!r3qqPua{kemeq_6vB!zth|@uJ1Z(_`M(qQLp;K;$QNG1&!V z{)L1<+SmI8>d)PRG*1l#;8a7g;wZg8#UG|knC1dd~PY5y*zqI@%~uH zaPX5meILVN8-WLH1FB*Lu)L+^iO#-(z*#?8x49%k+;Aj-6vH-B za>~3Ksi|@2>HEj7#u+3l!} zYHZABWmkdXWUDmd_I_14v4P_a-EbXb52jkwV4xieLe&w{7CCM z4}M9$NM+s-$1nHS_Sv?=pqqUPIsWhmvkNLi?!?|5+GT9VWqu;n#D3`Ri<(;u`+Ire zkW*^-C#u=(3PSCzZ6O#h>`xCMwE`8rW6Phy;nGL-XODi2fq0$O$Z-HnRf4cyzR2K} zD@`~k?)>@kI{0kCJi!-~LTThNcFP)jgqZ7`S0@Wi|Jbjd-m}R{%3X*EHz;H}90MQ8 z$*Z0v6{B_Efd1HNPJnvL1nE6_zOa#gbisl?P1)sb6=x*$b{1YFb(GWYR|fc<14mXO z@5#T`YjY~ZV&*ObGqy;f^SGEMvh&4iQ~N(k*+(~|l2$yk zP$EINrSF(q3H_$)>x?v!BI92GCz-NzeG88UIO*~`q@g|i`>qB!^i{wr7x?On%dHZM zl$%JlvGxdXb4F&laQn-r`0his4(F8K8ND^6D=QqXB$bc`wRO8GZfa#w{GCXY`YWz} zd%IQ&3Dz?(iR26-V2C)Nd=e1L{+G=WOGm)ymm`{sNeT7KtFV}(KdX=$<5KoPV$!b4 zT^1cw9I36%xWw7uWwQKTLWZgF9n?=rxNSa}klH0TNRzEmHD3eK0w=S~3`h|EpC&I_ zq|6LJ0o*irV$J(;(K3)cC&J0{JpJXN{UPOD&viLa`n=(|OIW|bUc-+pJ}FL?IHRI_ zq9tMD0Q04sWtW92$OZo$&sYOp=?i{&_*-2GZZ|Bca0_!jgM5?^f5@eMP~u7zy~Zt* z|F?Qs`0K~)wi1o9c)Q$k)2sHKM7(>=p`HAKa1qqvZ<+g=dba50r$!}*V-r$ne%dda zea;L;MqrQ8876~>dOXYiF#3X_7fID!Co1+cs9`a=w`Nv;KfRAr@FSqOxeAvo^T41t z@r%1oiLvCdn1nptHv(vre35WVut|qtGNB76%=8B2RC2EEcL{(_r*41y{q7H!gf%}i z+@x)XcT)Yuh%a|Lx$nODs2NF+xo`SZ)n}}s>bg%Lng8(1LmLW9*1rvo4DhSI$f6A9tyNTnqvsI^e2!@9p<48_ZGD?_M@|f!cB;Q$nGM;rLK#X9 zw)@bhYPGjtB0e*q$iEAPWliz_3J-1xjc+|w_6$>L8e{j^)gH%AIfZgDVnW`dUt?Zf zb2x@GHZE({3Cea`fT4gZA$927*)z|6`|B4;`lc2Zv~W2@cMS z^B9mBh;Er7YPyMD5;CZIi_vK2uj`Deoiu9l7(I8w&dgm`1=q67s}5BYB)vM|)pB2g zS2I}uU~}=v2@Vty^DOASoiWueLQ=@^@_nQ`n_`)qb$c4UH?sakf_U}s~&kl*@L&rN8X`S8|Sgr=IwPMS&9{d|$! zcN2M&XAXF$zh71-z`A<9C;U6b%~zZ&LS_mMJNw|IvZwao-f-{r*K>AixX~1`gr5=? z8RcI?2H9rgsAm>g5BNahiNh~pT#MJc9~?6@<4@|iXJwKLZ$!&6-%p57jKiKYjFIoL zgMH60??Kd==8J5X<3nMKhfDfblU6FrgMZ0ejEf|s-QcB=V&26DXJftY#Ku4%>4r%O z1)T9P6OL-($GTSUx9cn;0d23lAV+u_XO$yG=HC+X&MIoq2kIJ}+>zDdFpPDV(CTRagDsbtg;`VB-Bl;w zOkv=ZQLnjis@!lWjOulh7#$M|iR7ir8|~NH-0XZ~NVYp3BB#8R z4OWu}-DkU*jXZu=@zwzk0-QEL`G~g4#`CAQaWCIWJ+8k|%L~l=T#n$JzL4APInn0J z(S*eoKZ#LTJr|7o(wR3MACr3E1=ckCHIcB>twOK9|LW;~yz97^gT&iW?_bW&iho*j zORN3slNc%2vG@8E4X8loVO&`b<_X~=+t)4)edKXe|7t>$H~9731Kpu8r1!QfH)Zy- z8;KbSi$%hviwykGwdiy=A?61MQ|Tf^OpAx_#yEb^oJmJ`q=Siedpe|Ld&8<@Cq&2B6hg{hC;?E4S!3mMy3|7|O3ODXQhaN3F zOVO|oGAjvJ>n;tnEt$R9xHeP3|Jm>~SNape#(0{okpie|YTSJQ7Jz*rzCrx?8 zC>u^VI@|UfCHA~h^Z%S-WkTrIKf3^E+$HnbV@n$qo)cMuFP*t~=%*nCA_BGI-oDT# z5kvRshTBkR*L9w)f>hWt#~rAF3Mu@6I3pUR`swE5bi>Z`AgCorzK}W1?uWq7JUR(& z^UJP{>Mr8Km+NNbXX_`Ws~JwWzI(44ONa6jDEJmE`8*}Mmh+MO5}hFUD(eSLR$o`N zz4NQB3qQYQ-1MyXV+W_8LoBA>z1f5RAnoqmeb+86U$@{Fb17(DR}~(^aa+678~Q^3 z?Lf_uM>>>u^si+lLj)`(T(53cGQl%dtDn%sO)L(xsniVZk^`W51Qumi1cT zB+Y!o|DvXx;clKOkZMo!9XXL(tSS%_dF-oQfrft3%)s<#I^0vIyb^9+?oax{f!-tDXzLr(zy^!0~ZiK0TyV{pVnM(Fy+i&?6f@(l<^+a`2iZl3M~< zK+R{-Qxh-m!Cy@{Babw_0GmQ(#bc#t@ij(Pc=TC-rvks1NPi_f^Oy`DN6g4`jzhQBC z5GCGROh2e5+jBpzcSZ5W#nkEn5KV~vVXu1ICB)M?De_Tc(U)`a)4mRAPrg%THKMO& z*-S_{;L$_BvV(_(qfZ&&e7OJ3F_XbouMWj$#l|sOSnJ5|>s;P!xNV496R%;(muRKM zk=~EY8;;+J21FU(iO9yggr!q7?(pBi+Pv@B38YxR;BD{Is^-b6%6v@vZK+QP-}^ov zi;GWw6hcoAu+c&jkM@#bqxhhHAPgqrsWbUdH{9lnuQNnQ$NB`KR~NdB>s?Y<%Mu(5(8 zTtX@vZmA&K`c|`-w%dYsJuC)7NXI*r!8;@*2HmGue(yysPOSheGvrMDwHCoV$Zs^i zGIysy=jM0QMse7mD10XI*-@~3^nhmqhTl8|68@Mkfvn+8xn-shp+%9|D&SNi+>BXY z|H~|))qkZ!D#YUIp1$HAG*GN)nQH%ZY$TtpMX*?zmQ}0X+Kq=pPhvxAtWIkB#q3&n zc0!Mxew_RwIKt)ZoN$bY16#IQ(V47tso8Y*r|C`48_}spPa-eZ+v(OPO2)NtIHuoq zWI^EC(`Ct*c`3Ro{u52NCY6c9l1VT9TGzE2MVVaA^1(It(SE~dwe?d2^a zzCM8%3%}}nnDS2MLCfPc*Z4kYAL5tXLH7xqZEnm9nbp%;b(6<&hIx^L#Pz?prn(&Z z#=rf376KZFTTZsu|NsAe?P!1%4qml~Y}{tq+kEf6B6iZlcjPK!cWcjU;nD+18@iCY zhXSC->bhh~>>HrKvuX-Y{C1VgBC#=-1^+~0116-WMgE`d&DekO=oc>}%(aCx$nptzZtp2%szKA)N@vLg zq6GJ?DC4Jv2`WwFTnoTO|G$sO|9r%9V6)=0&I)J1@sqDS+9fg%$Dap|#c@wH(BLH5 zj2+|kh;Hc1(Z{LW^H{{79n3I!N+rgq*OynyLf2~n2j z>^$_DZtO2Z1qxzka*LCq5wfEg6$*QFvh-28qP&7+o7MC9)YLms?RRtBr{zMAgIr+H z34&5B`oUr!-+@jYz{H;4%=_UB9_gjGg*vY(c%HH(c*(7~BljsG_llZ{n3!=GjWhtw_$UG)L+d zUYscyp#1OG>&yw2_4QS3xSRinuj>qGYFoOBfC7RF0!mi|q$9mZQ55MYCDhP+uc3q@ zMO37yNUuth8ahINP^5P-By^Duq4$=2hx@+wu6lj%U-&_C&fc?U)~s1G`<%8r5SNC1 zCP$Mwk`Ez&JZ%3TN*jEL77t$5iKdK%^NuXUx42aS5Od+9G;2}bqIv~q%q3a(IYlk0H+<&QQ>mzb7}-xArojO(*Id1o zrHPx4-m|#%S(9h%v95@4eWn@r8cuiS>1gl)q-%mx@8utMl27UUaZNllm9mycScx$* zqTv$(5L%8XzMlXt{k;qim1-VJuvYN&CZ$?mmR=m?ovnvplYZ+=xvo}m5_1rT#sm%AeufLW7F393k_Vc8*;up@#hAQqGQW7Xya!lz%Co!>rraZ{7Rh%(05Ff6d z%hwKpxUe(KXk*01#A1xK1w^rRW<(H+rtdl%a_b5uPe1S(c=-dFc;1DA&H7E6|M8ad zyv9)yiz6PcwOZYhEp4%tK4#M5)oGx$t+*l~xmnoEpR#RZ4T7c>Igb7;Q&cwgLHd*u z2XIx1g7l5V7w7O!35Xp<*-w`9hHdBBoN8@scYSd5tp-q62o(`OJw4r{Ak`T@Zvc9s`8#%%g z<5!}*Lq#~AMdtCRotxRV0A=sadIMv-F}*|vPoV~9S@wiP|8OA zt<1~p(KJLuRN&>G%s}Vac^tzerGr$Dobf6_ll#|E&ls}X&l|)q&AN(dbWoFF+-uL**|VDM?Rzp)X112N2Wx>^isb58f^ExapKfym zy=c>xpf9F_jJV|kzOu8ok5?Hkhhx^l&_a5#&WiUdf{(3Egn^%6S> zvL~o@6aRm_>oADPL4^YaG4E!klhaL*@-!`DKMv&abH}Y{tqvx6fQq4<+hYAs|A{h; zwFq_+pJB*{@*poD%*DcZ%~R6|VQ@Ax0MnQ85V%p1?%?D7+l^^PxOtIzecXnmn+G`c zfRUChXTTQx^7xoF=JEyAnWH&VWGnEjT}7XT)?6;GiYT=3NM?gDgPmZkE5t<%ScE4( zM(kfS7D1mpVU1Qjez|%aubQJQl@`;7K2BpVbR4!84$d?3^E74oXK4G|>o~uWxqVZW zV}URYv_Teofax4g4tkdd-duGwndbiI+noy}r7(V&PI}z1v^Gd`6{B|)^@{NJa~Xql zO|E$9qf=D3hMMm!@-IGKuBP?7DeDPC+uN`4WOA%%FaTY#RDI;`i=7uKF#hmW%&=aoZ2@frZ?TlKhOY2AyAWDsVxY+x@S!Z=YbV zNiVIMTF0@Or+YUB3h&mw{)1aBp2TzIR^VlS9DA_Ro>U944Bo&?Csv6r$es(#y=2{z zBYiS4bL9G`^t0)QMBgwK%<0zeJ=5&Bj z5`mp$Ae)Um=!}u#9z(Pw%*MH!zun4>)kJN~t8IG7v=$Dh2w9-Zap>*ky)EoX*K}W8 zECPrvQuhUY#!aY5H&}p~kODDet_`DN#ZEPNg z9dk-xJekc=5;!z?k{3R*KY=!r-rHUT`SIxhKIb29xcDC4n)w5WQ(;FG26_Ld4q-01ILE6uvb2ua_uJlNON*q#uc_xrD#d?wqNw-6 zcM@ao^aM{A>n-Qk-4dkb__**UK!pIA>g1-7a@Mq<5Uqt6kg92|*6?w2?m$o_42nRL zh%G_VgoKn}g7SebfzysRISj3pOO9D?{>t$O`~3G>+)$SZ2J3GHU_Viuda&H3G?EH9 zX!kuDk=-6?emTGWB&r*}j&q&8mveA0XIFf3vha~Q+3Icw!$nuRB~(w5O8!l{i~s(v zpB$R;WBeWRcNFZ8ZfW4{RsUQm=Y%5N5magL^v1}}J@+QRjuXsoIm>xv`ENKPb+;%rz+?9u~53grO?npnRjw z%~hUbsx8w@@xPWtwt>O3e}X<|X#XxmMp0zZ{m2RBxu6RRpKzZ6EiGFh$;GLz&klci zU^UE0(Dnz;6!e(M$J_4B;d9Vja7GF(^HtW%!r5K!0O33Q?E4&bs)!Nv@NsmPdGMMT zeVgz4OKE(R`t5P%BO4iD68SWrW0+;jy7YH05dBV0#-~ zxoq7t5%3DSIyo7dY^+^i8b=5}NgqykJ6>T-NlpL>6byU#3i?*?NXnJy@QJ%+BQ)>a z6}EnkoC^ID{QJ-L8NZd`r=(|al9)CR*ppK6d=Gbat&a@JO+G(1dfYSPecyaOwBJXi z?wb4x&od?m2GQgT|M}m)Aq@ZkCd6=NPrXZ71hDptkFD5())41?ldtG{i(7oQVm)p1g{(}E}dPJ~8^Q?-(P;Id5 z$&oD)kH=%$PELGb9YAnK7;xZDALw!2YcC#Mlb^W$h*R@r@3=zwvY08mdFEebxZlSh`C zZ_@N&`s!4)oQJfL#WUYimvnI1kY+6mRwX>m{%hE1pd%>C zi#)^`B7|{mQ)zH3m3Dl3n4}ROZZ&N@$&6F<48nhU?x$9=XJn@?x;5p#{(kb2?ooJy zJANy?D?of<{x0y44c+)aRm~o|#CTFPGByztra!Gcam?s%UBBPVn53sdIAz+^v6K)J zI@&mz5$qSOs49Y^fuXa?B$m@h%j0T53}ul?T~=(wkgUOv%?UskU5}>C{)1p}p(Uq| z$NE7#=@Tl>{_4Y|5zY^PO#QWeuBPNCgd#f z4zYP@OsRHOvDv*U`oVaZ5!YxpKRNJGK@XWHbu>${y^lNa%bnLE5zo7z53p$JuQ>|P zS;uN{Q*EATA^qWWy^Mhm0bIHnP7mI&Le-6Ux*U*e7@GDL+TS1g|Bunx+W@GM|0dv3 ztD1>$m;ouy}52-9S*Pr#XiOoPP2k%F`^8ers^P$M}-%gSz?hU%swEoMim6 zeu10Q5-r{(E*Tjf)?E@856l~75^Jg)fKYnM!$t{j#hv#8>J)>y#kQ!Hv%q?E1(KSYmJ}8;FAj#0>*neR!X+fG;Y{e?0MI*C z5};<9J4@*Qv44Lr#aDtKjXWuLU{j}@ryP)__+I`ySwAQUN6mEC9=#lPe%h9Z&*#`^ z_iI=N=p=NRb9>aFKSGK68^0>R&De2WMPz4ANuD+t+9}JRuIktV#Zsly&3;jU01pLH zU4eiDIKnK|dy9yO=*4vAlN_sg{i(wuZ+BPAznF(00AIftQ_Z0F2~XA+9X&swk^FGC~XxxAzk*!MOWZ^j(I>SRt{Gcttm^> z)%3Rnz3xTdyNj#+o!<#(H^%XKTq(hq7H-23!*4xru)XAUt4K%Rg2fH5McM0h1Kt^h zaWzeY`FX<$z94$%BFt~-{ys*4+*R=+TYYW4@Dz4!zxyI1wzIr|Sn~U8TohEuua~j2_Kht@T$MN9foBVuZQ0rGv6P?114l@2- zv2Qy0Th++}GskbVb@TfWu_vUhUInO?ci*G+S-9?vL`fI=RMvs{J3`33XojqXdyraU z>T}AdxGqXJZx8x{zOXXA?4cQLX--OF)+Ms?<(so~Kt%*AVC|#7f%~6{RTAEwS8)B9 z%TbgFngV_JBp0>pCu@v7-uBrE0#Az7Y1bG0@HOV!nPw4hbbCeveBz?P!g%6vu*M+>I7YwZbW$ulNI3vL&<~y zu8SE$Ay{67%(}4uyez5M5>pacwvaib%|<&d51G-{y-><3_0~$rW-;d?6goehK`|p( z`#`mDbQgU(Bfq#j&|y;!(8?b>UhUp_1HzPE)Z;|vQPa_u5u@)om>B(fXn%%;R4nNt zMkia%Q7ZXJexKEojyA^d@ee>*_;5p;vCBou6}BL~y`wSzMopSMx_mXtw?J$BR>Ebr zsiD29V|=*#iCn}CvYMtaTbcYWWj$5xml?(3- z5mNauU~+Sa+x9Cb;1%khAA>321z9UK;DDO+rL<^ffKagp%4TxgrhvoIp$Mc#Y$Yq? zsGku>7R#gwCIfn8SN2pm#Ne|OiZR9L z%Mr}<#Z{Zw$s+9*Vq*V`rhP5k47uD$!WC^kmk02iQ!e>jen=)Gn%GL{Rn`!aKTh-Ro{ zj347hdN9&cCM}Sl%~z@nb!}{K^bY*b@BbBq`C^}+CiO}_PuZbad4)OC-;8rRXNbr7 z-t>@{Vun0e?e1;f%WCzg3fqY<5Py1Fnu32&{AI8*oAx zz=C3jHaUzAA8^4rv=FDE;G73SP(nGJZ}Xb8wL~F0%S1F_APl?;84F+!EXo!WlN2bd z1geqZZ;36UwJ<>5MqW-L9i^BBRPaRw&f@;hHulq>I?{n6lfGjo(gDW~)qDO@E(=%~ z0xL2Pq?RVmbcOvuD%B0#!m?Hb-CkHv4A@1Stk3h}wxuHsqOU-7qS+iTbrMT^9BE=~ z_i~5lJg1VloZAAb-?kl``Wv=)uW0&yk|qr-`EB#KqMl=w-<*J=Q75kYseLbFnTmVy zPI)LE*GCOkZ|{%>QeM5Z-ftAxI{4@;N{{08-QUfmy??ZJ#os)yB2hE0Gd}nfc;mNP zh&|gEoKV&_xw7f($wG`eVj|Q~J1$Ib?{c_fu@Xm!3UF(ycNv%?@l03oG%-IvN~(6}9GoD*aSB-?aKg}I{$liHY zAn(?YK|sYz)*0%WX20?X^pvS2lbpx@gcm)v#3;!CJHvS~6kaEdX7gz$lXjLlB^k0?sc~# z15wgnTpM5bbJgK#_ogZLqkM+}guqK1kud=IFO!J+t!Y@Qrz3F&ON~dD9>EM?YKTQ5 z;D7nvH`L8AZ>#y^gto`y>hGFl@_i8;d=NS=n%1Sz&v9~8D6;&HSYMXG^1J2aW)ovT zHTsZnP4p|RUJsYUA>hj5EAinw7Z)-ph#(!tIS9Hg`?Q4a*+w2o!`!E@G@lLSf%SyK zfx_5NXYl948R$Xjg~671&&^Q>;pz{%hnPKgPp?P`Qd4WM`5PTClgu}o?K{%RHiH8; zpi}Oss_&hWgNIYGLcvRTJ1wZFH5)>A0tW*tHD|*i@k>U+6tDi5}hjAAc zWoeaR9@ITprLnA|#e`{NY%6*)h6FN#6R@>WPzg9c-0B{=G41?8a3wpH5SlVEhLarg zzR}`LJ@nrk*A1?XQ%0nDNE&*DC#8R0m)hO)Wq&?>nG?A)L_E zE*z#}j13dwxi!P)ugNXNFdIqTQW>=$_j}RI?Rr1cV~l5_Y951;NE?+4Km($$2BNj2 zBs{KB!i8s>bCNi3U7ch>koRQ015g)WNwcmuiS8fD+P8jNRxim{dp+dhGy2v!hHUWg zJg3Mb%M`n0v(e-0S+23zg)6?p%R+8o_M|>3oI#$4P_Rl=6j?US+aY7^&XmVP5|6Xw zT8hpkvHb?;e){wBRWCIl6?or~*%O!Q*XlR9j-zM5BthF=uIZ=wVcQq+?N+{I3gT0e z4a8Rx!cE=Qg(kwlr_gm(SL(5PTx^;CO4Rv7tKxij2bNFGxbYi{oItP0&qs#q<#TLi zc;9)L2}#~OSKERt=T#l;^L44;c)wEOr1VhOAzpmY1128}Pf9&yS^&QVQc|~DmhoHP z@IMv4a}~MHs66AYk?$ck49s2Pk*LVi^@r~5?}dZ#w&S+9Nw$})uBa$+cpi3b4t&&S z?+%b$GP?((gg5C5=JThI18TO+0!-9%s9URNz=qkO#EBTqPp|)zLoBw$1A^Z-i*Xrh zQpXXB&$}zjr_d*j%PX#8UiGSKiG+Nii zTpprwO~&}SOgz_(;`8}`;n-}>tFTX6fRGGy!|r`qwZLP&wyj>F3D{BaDL7j7#jX54 zhSd13e9prsY2CK$O1I)fE|zSd?)VM*RhY)nV>q;P(L%ZG$IR(?7S&=iHNM;rI{4yW zU;cS_{$7BBmkl1(d+O` zgpbPPo2fvQQ}fND755C!X|zevqlZUS5Vgcn>2%@dc^B(nPa2@Q*s)6VfTqE-sv@VH zs){$M0LtkBli|oj;g1&WzU$^<_;^Ft;@2jyLpBso&Z6b8p3Fsbxvlmm?TY*n+k{F9 zXHoAek9igNk+$kX5Z0Sjy2djk*e@3yqk8cr{7Y@PW}mo}csxKvNxmceRyv~G9~XuX z#}J!5F!KT;95c}ePD+j3;c)*eDxAjR>KD*A{_Ha?S+5d7D1qDD+Q-gf)N??nl!J)H?xGK#T9B0T|pB)g4KD zsiaCwN}PX5`Nm1D-NV9n7EVDL~-uVlmd9;Ar6d=x7GUTG1&+bq%sJ0slnj?-%eAls{rz)Z2>7ga8FX4($^o-lorV3EE&?PK` zCF!5Er4)oJ&A6+>vUgz7vt*5reL=|-zwFK{l5%tYYdM~M>A2MTA`v~-7C#9oB0wC* zXAuZ0VcgnH?|_Q`U{1xBxO=uYR@Ot4dGGs?kKI!!0*nP!aY*R((LOUi08oRE&)%;`V}G%x%j&AQ2y< z#T3djwY+Oo3Igx3HMrr2QZIJ$3e=UzUffNJ$UfdzzzRv8oJW)Y^5FjoH?ev$Z0=ie zCWM0;xz3ATG>x3N@va@6eB_-nN`UW}uP!SiG25>L#W)n?8HkfjUu-PrjZ#%rfg>Zs zx4X-t92pBvVk6^{Q&Zv57l)U3<;OVZ^2DO|f5ddoj~Mp;wslvDlffR!6;vshM{IOr zF;m$Q1_w}~@n9>!CpFCC&G~5KWlmQ-l>9^m`SPVUc101{q1JHiYB4@Jf64IaaoaFW zDjD-t!xTnKyUM?yYkZhY?R(IT_QZmt5Bi5n*pfNkLjzU9t!!MdQ-BkSezyGLi{0e5 zCp62Q2RUiWW+k5~;pC0?07ulD*#R8C86pFv&-u3m`j!jI#Vm?A`b-3ya9yfzaHE%`3Ue7Q8ed+Fj8R`_{z(9|uk9y1+Rd z)`B5nf-bd;84f5^V!nopvfIZb@8K;g1}Fwis!nj+z1h*0=70NR=|;%KWB7byFw_y+ z3*(uSq*vN(RSU}s8Usmup<3%W3I28{G^SKwOJeunpdAy#3iouL z^qckthyZcCILaqQ8Itb_U!rj2>Vl?fYXtrq>BZvoAG0nMlr>ccX3u6`gSj+t=GI zOW!rbkTETPpUzYQ6Ixbm#UW-h`Qx!OcB5U5lc@?hj-Wj;sr$eUFZS4+$AGaStKA7Z zD>X4^+MrE*4$9nNRt3p<_Ck6`7Ryk__tthdt1`)GC#$*vshwAvbf%H;H2p1FtshR z*En*9>b^W?!cJsR$7jY%k^lti>xP~6J)`%L1DDTYU;=^rd|QZw*(a2u=hc2<^N2A> zVa?DIsm!vm8EbUp<;4i-_<{e}7>h=`k`Q11m71Qq3}&$fE>6gcviBc=LX{{!{6`uo z?Ai^`cRfIojFj}V-`=;;5wAS+oc~xGV-|c&zeWX3XeFUgyTN5^ z+OTk;#V^6kp5+Gv?DX4r<%_$Ng1pZORgK2c7bfzC(+-!au4^rBjT*nKbqCrV5B-Ic zNq_l`FG?~2y-OkVd!MRE#Rs(Z#(?_sS>M6;ul3NKbOVV*V(yC%j5z?>BfF{n0q}5v z1Ng4uaI@QV!aE!eDpYaD2i%~`sMuI5+ z?T`9|x_oL83n+r!xRwWK$1b(0jhDnsMsW9lIQiXbRMQtCmvppB%1cLXJNdj$lQ4-8 zSwcr6H9h%d)<_M$r3v=_s?!_4-k+9M(lp@scxlbNifr=!HM_I*S4yAqj4~AXWE%J4 zs|G$5*VV~)i0`4!K4-fhjqf@Vu5hObW=v|)V)f$&l9_?bEk8(i;l{3)rvvWA@jf%M zI`yFGLUfPi`l!RF{WmQ^Ass-r2nfVu^i;4W9k9{YPe;6&33lXOthkN{>Z?@Yc0&R3 zK;ygn0&>kVH{jf^mSS+mv*^CRxWd?v`uj>kA$(w^8 z`dRjdMeHmcRD}Rz3m$EeD>KmpB#-{n)+%tds(&1Uvx9-Y5eN&|ZApho<65AN+XvA` z{s$#C?f6e?YImC4-mTy^9kfq9K8XY-PhEiy=4UH^wK8G@nw(zLYDz`pwy9IBc}w5n zT9X;zB}D7B#&IYqfDUY{6I+;9xBI}V#81%UW~yFGo+SYl)bcqRtSlez1~3(WER#^xQLAQcE2h{zOb%(Y z)G4vkwp{665;NaxTUGDCtYIEK)Gj-?T4Ac)N#pDh3|9AkWr4*@lfHj1E>633+GuaP zEv@NhS4`Ld?RCuKW^m+CrS^Z*U4sn1)57Y>cfMOWVebQZ;MTmp)5ZMy&1F>3!pgx{ zdaL3!vn8Dext2<>ccBWO-8ohxU!YmwwiQRW-N1Dv>BSLikAmv3U)x@MMpr}z1T6tw ze?P(O(&yO36&-{`1+OEzE7ih0MnZlJ0RA=v$x>4S6H;2Wv6UADFaOlk11Lf%+viFj zY;-P(l9yU{M3p-;Iev>>x%_VLdxqc1vnV0V-bFyY$r7B{_VOjG3g!I@&#RI-CebRq&B$B2OlQXOt)zvu9>;NaCo2ET zq?H2=&Qx0b--b7jU>5>ux)t>(bLHOUq5yEdo8LzQIh6IzptCt}*x&b5Ihy%U?dfaH z&;MC$B}<1(&bKorPvX1y%C_stXPGQL0h-H$0gF|7+|gZ@8MGU=O9ZuH;DM*9ATs+Z z3t>H}zAW5{HPas!tl@~{FQTrjk5NwC$JK|bR!RcF7qGt?aLI5bIw#e4{Qhyb#ygLr zA5*8E*!Gj<8MRp-$FUQDQZ7^$L(M zo{nq|nPFQC(UyD;;uOCXE1e7GNfHoeomT75(Sq~yPlNBv3thR5AjMXXmt;@A@q*TX zg!luXzjujHv878(r|7((J553HI+uI_xmp4=yu|s#wI$DlyL%qxkcn*BO9k)G+B=WR ztvP|EkCEZwrG!^DEo&I^<=eeC+bhj()xr(O^h1Gpgn^i+x%ofo&qdrj&eg@N1H4_5$mqj}y}3 zQvtY)-d?1SZS78g2}Q7?=_p~9k9NE+Ns#rnJ>$Z|GZXt87~bGYZ{#`cGpsF+s|#HK zOt)c&z|nWVU5F1hdc4!uat%1s?tvqdo64^M(C&vH(NHrW0S+l?pN7T6S``8z4j+zNI zfbC;3YB;m)2pFQ)pMHHP!<=6n@+mt0@C`_y%*j>3Dw89wHrI?te@(zi2{@y$f)<_;ANKcLU$T#s~!i(a-0i(ei)O`Rrc$=y*R-u<+ z4t6qC*TMaL%1)}Uc!6pXNLuN?aDO_r4FP+7SlJAmiktUd0$WeO{?thf)Nj8z%70aL ze7Ac@jizp`UO`E*^+x%_1Vnk^hDI53OB)__vk=XqBo(*K0jY+C1J(~YS5W1{b7ewR zv?DE*uX;`;G^cNsh+wyZ7VnK?QNHfqtMu0=sypcjmKg-wt`)8H1l3NUvmU;%JQGAg zq2BY3HH}gA0|s}>Knn_>x6?lD(kcM{Bd)buT0q)#T8dn{5^d#Wtv-AUuwMG1wNfg_ zf|9Rz3iOYQCs{h#SKPiuS;+r^RYx#R?Qm;(RFg$H=OYgoKu!4Y`5@`6=ntG)M3Bc$ zs{OKdfGDd+St4u5<;Ks({kNIx@CVupo23SSL}(Av3byRuZSuID3C>8NbY&?z!JVH$U(gG^s;95u|-~WX$o_ z_QO2`O8l7Wq4uPy%*8tAs|8a3;au6!AeT(`8)X|@#6zBC>%+h7?r0+6XBMROQUOI7GelAkNvPuO*n@XB5o zRKP#2bVzU>SD4z)gHF`iJee~44eP`so`Y7x)X>vK8MeJA(wBQ*H4qXlB0tg^36u`L zj=MU}L{SsgP)Y=Ny%DKvqd>8kmiaB9-XOBO`ZRwoTz|NnRX}$1xK-9{B}9+ox(Y|a zX*%d&g+iH)@P6J9rD3~6NCdK4fgDIKS%CN8j4i3n?Bq3XYC__*nLT6tMW zJr_M^D6|`y7nlpmcX*%cZ18XjxZ8_A5XcXN1#Bf+Ck&4rs-Cf7w&=yz1iO5YD1G1} zS_Zn%!IOQj-X^lx2sH5v%i49K8T3v%=GFD<)E8wj}bdXaR0D zA$@wdujUk?mNT?1y&0DF&KMOI@LUjRC<19WkV15)U+aN6AOH+KKSvb8O9lpzSwqfsDf$3l%J$4wslBS29hT!Pa0liR8)(+>-}K$VniuXE7JZw8 zvmK=fA}QR8Bd*Kesq6+A8`_b51#w9nXHRr_tl!!H2$9Rnf9B+zehp|ZNZo}JfrMc1 zN{$I!|BO(+%N5sk^OUjG!{8g&3P@sm+#RHxqM~up#A9ALc7zC+QSb#>jY*K%tq!e= zjM8J3ZSsd4Xo$O3-)`iwBOFr}p242FBu+*NoHDrX{J2(K#QxgtzqpY2n1($_#|Ooh zcURWk>7%`zlvwMwAU*T^K;fwkc&Mewii3n)dsYXAXxgj%Tc0URPrvV`5iZy&Aq$T8 zqh{+4jg(bR1@$B`_iUhstuNd==F3>$Dc=}MisI&WK=7`wpoxD&t_?p3I#t1W57QK7 zo0uMRB=ExlMMY!q$q!n25+EjV6AZ>VR1a-aEF}fZDJ{dhBwO{YryGM$D=-2n4~>!d zvw1v!ePhEX{PE}SDI4+>D(K)_!^Nw49ro|`{Jlw~bde3AlD&EM+8o!%&w~(k7I~1v z>jHlFf#Fg#C$D)?{1-lluIR)Xow1E+Cb!Z=Z0zK8Q-H#Oyzy_|;y=E2Sc~Z` zfW?}q6S-JIE*jCuQW%}EJ0CsNX&_!b0&;p>_snx&H{_kfNDAd7Fl62k+inU;ah z7TsKaN`Fd`L_KfLqh^Qr$;ES~ZAa7~RV(vwQe`r?WUUc7E291cfm7ga1)7(9VSV?? z+P1rta`)T?&{>r5M0~jA2GeB3zh?o=OZ8kZYU6-7#3?!w+{e+8=3;XM`bvNDIlqAY z6^`!_ygTIjg9=W*A30Ig9`iq(2RO9--zuTzGdYGqP9iDxFdmc!+$?-4#OlU+f@I`} z4cc!KnwTv8K(4{uP1O#=-&~spdH+<(u!H~FvpSDG-a+=1@aD5$kCizQkM&;>-Q>Mv z2L;NN4ZZ|Tf{ng?T_b-*g8%hU`Hx0M#mwZB%vOX$_N+sIm31M^-4Gsi1Vjx4!S;? zqUKMh)7yUAm4E$d|MocuL!6=Js-fiM`v8_UO7Mk$$=zH6NL#<;NMxI>^RdZF6l9qx zV=H_lPeJzfpJ#e&x=F>|j;wzx$UEoAAe<; z`_^zxLCCHI5L9fpJ#9AAsJQ=gxPQTZfx1LrP?d!y*XfP=uf=9VBgSa$2HdRpZB!Hq zdfL0p{UC!%u$)NIObR-?_iW-FiPEgsuedr?KY(p4iqBu)yZbM(=}+MU5NWnY33`%| zt?V3)4JO4m>vd?H@oQmtN0Jl>#aaLTMH~QA0<*3HaPm6ig$TexR4AQ;!LOa47K3Q( zmfw-ljmEs+sy4_ZMvIS9ZFxtlW(Dcq4Qw1EyEvgj7qYW@$Ck1C7NmG89EP-CKeUCk`fpno~jxe_6@% z_?CCluZ0~r35ggqu2`uO+1W$aok^_0oXq{+Hi_;15&N; z*$qcihXG}M1!u7*R<426Wa0)s-Ktfheu+|$3ii{5H?IqKaX!41O~!G1;CzUf=IR#i zhVt$};V|;j#d3fy>DinP?8s~m18bvD{EPIVq1mJ7$b?H|>L&R-axX2q)9vdJB-7ZZhdMCEkB3F-KTeFzHe8V%ss(!~b`M3DDiamPi#@Neu%ku>Z=$-3gM> z`;iP_811gU^n`$$AmH&p#;5BijG&-Z;O7GgE-l(e(R@ZPlM)QG+&9ynqRCdM(39OG z!Xab2<{dAOlLGN>Ec-4c&~WR#p~VTgCYR8Y;|1j1iRvvKMth!~JpiQnlJ>yqoI(Em z&6@`gyY{`n?6jxhCs~>ti3Jj41@AiT0Bn=6HMJE4Ub|e;$njtGhIY(vUE3_0)Tho zxI7M+sMe|#AMnVak10F@fnx-O6LI&gPgkGJfB}8Y3xLki{p_lI<{R+O^S+w%qpwP- z#JARCCdD77-L)1YB^sOv(!#3xT72dCOvdp;ys2! z9q(FOLB}$gwcwrn8}yC~PJJ@~>Y7u$ldR1_rNTO+Wzf~hRR)nSAru68M^&RLwa_$+}i@6V@<2$lA|gNqNOd31&- zJj#3J4b*IFOx+v})2PaRn^@97;*h__LNUMXvM^`jqcgV?tiX#C;_z&Y6i`YH_;7e%oZtVtSe8(ICU%U+iq0OWT4BC-6C#022o8e44NHREcmj6g1jzma0oCQvW^F@jN@=^#AlgD3bJjVTec9J(@KazB z0ynwe!+J{AxR1jAOj49lII~5Tkw4X|P-?r0`L4;zO$j%pEO8aP z8C>#7L!9Sc-(yZ^1^S}zAH+eXypF|Pq@La>WyfYtZEn9ZJ3v3|PJS=aX~I;mO&1_3 z=?TU+7N0}kS3}Ap(?tFBw18H@Nv&0aRXHG5fOMGBdTx@MA1Z-^z$v%PHUO?S#^WTw zV9r8jfb8{Qn*r(x>E}#x6Q-8@Q6Xd7K##}qE(o9#*tD@|B7HC#6goDX2WE!_#gX|M zfsK-e3ALa=_n%`7|0#5(JmqEp%?v$OZ`ArFa9Ucv3w&c{&C=5bS*EJ&VR?iSH8C`5 ziGQwd%NT3nle>byky(rJy8xr4E+@iJDia&p#jViy(f8;jFj>*Ihh~B$Dpp(k{ybIO zAQ=@k+p0c#Ct^2a0UL^oXJxpbtnY?l1F6ne4rxydyek|$SDe~o0cA;z492ca;<@iF z(!5!rqnl?eE-R(HiqE-!jp&P4W=iUb=Z2fcb+0XIf3I0-_m~(a4z!_UARV*sRNn!H zbA&z8%Jg|pR-Zle<}^*@RJk0$S&PFyxFxII7j0`vbrUHuhvC3*rphVCyTim_n(Q)M z_@NeD&tYBm%U~X$wF;@DI=}j2sr+P@KMzprs9J1AiE)Kx^$Lg}BYKMLUw%2MN1Lf)cl~gZ|0p7GrcZ|f@Z*gDMUf6>p=sor zj#FTfZ`*inP?p_MfuR$z&&vjK68ly#O6lZ$KM;pym>5aSkmF%s5!^MTTHduW;T)j% zY1!y>Pd5k9P&MAs0EkovfZH`4XjChF3yKErI787Ze z_otsQCQTjBJ0&#-pSFi5lv%F}Mtb|ZTqDA#$E+$a-K8-YWmjk^D!=TEsueI)DppD% zW|fX$$`XS6k2XsCc16XqXzlL%5B5a)Bqdn!;65C=HO88*eF zpUmVZI4=WCPAHJfMc(Cjtz*4%R70cen%$<*HLD~3T9Z235?LQ_B{%c+0p?%P%Saa zfvw+G5%RoU5;%1j4*A`?P`*-wTxD(pjl$vXzDev0pnG#4Pv6lwhoXDfdu)M!u?|S+ zw@=e*d8UiBd$4P3={>QQrwN*v!RJv3%=&7){e5%=_?zElgo#fj@R4cPI3nPFkF4X@k|xet+5xcm+2jST zG#OC7eFyFp;1=<2Z!$Vzp_^@9m$>4361Z1%Hdj%r6(EMvi&~t>ZMqmah%bTjneI%V?aS_&Gh*UN7AZ} zyI!n=Sv;-gG2sJj^0;98Q@~{H$TPTlMIm*FUMblM6T)BZ;KJ!x%lm;&|II$R3(&nO zmQx1X7Ds!X(xL&npS|4Q=8JX?bV%@0hCPca^W7MLb!QZoF=fPbb$MtUN(T(gNO@sI z`ESM6$H_Y+04={psoo;Z_8sQXeG3T*9iR$C6(xR4-<0zjogS=DnpH%>JhLetvFKLK z*uR=&m@h3QCD`(CG9c-Xi)`j;hW>WYVLLR zB%iulT4ysb^$V4_X?o@g&{{l72i-?>N-aSH2$9Xl7jdh`+D}V3>u1v|Tc#R{mhci6Kx3F!ly7ykJKKhI0SfvSMmLVR8HFq=i(+v%hL=GRC*h z7p7+fa#J1>3)s_^;ZuD+HS+m0xH1{9ciT#h7{5)M9``aXDTwVqe+? zP=L!@j`{X!+95AxgqjGRyziT<#cfSyB(5yPRT=J+piU1WM0$Yf6oaJamI9@*J0gy> zh`lW(OKA7a$IPV3TW`v?-^epr2RLaTr7b~YQ&t^B&~q1G6%}=Dd6!#+_G8EPCEJOM zxRd~_bo%XazNt-2nT3A2sdgEXI5DfNd+*r5b!%iaz-p(=e!o5{DCK`7*3Ug91q=F- z_JYiZ1>xPGaW)jo%mwuk&hega=6Lm9XxrFF^wo}bZRJ01@kmyIb_Nv|YwEZcM(TC< zQp|qse9q@dM+_aE$}yuWmeaUapBPsY?C8DsODmZeRh?CgJ|zQ&u^La}%}ne{t%F>p zfI|j|cu(>)B9{o@ckT?nGqU%5TrQ|=s*UMXSo}Q2G%3DKnRVHNKSA()u?sxcgqAcD zeJjS~*r*WMLo1D|uI91puyY!jn7jvc&Q1xwym*YwD`7xjx89JS*d$$%I?=jXemM7p z5$+H$*Fqgm_$}5(4Dsa^2s1jiM0;sa*vTpGwpKboN0v4{>iPz=sm!|lksXl+Fd3w) zHlO=Xw?%s>RGbS|N<%QSq4)uSm`wdyQ9w&kCjOn%tZ71IZGR#n^0(vqAAn>G#{YqG z(RV$aH00;M85=W2JURHfOugfSO=2u<@*q(7?B_r`%c=bkIk!hk5e8f7+_n2X9bvh8 zL$gFFQTVQV34UQV(O>h0s{Hp0d={*w@m83fwJnYPd=vq$;#?pL)dmOTg=`3E%1zf- zG&zPAyiPJaw=^EvB2)a4pJKnPP>66}g6u}lYGY^?8qU0{8^oXeqj;ZSf_x&Z{-Abb zV&dk4^E*Fxrri}IzmEHd2Q#b=S`^UrIL5@PNpEeA5ajJK-mS-A%Nriex^;n=@^0lC zL=`oW)5z)=Fj7mL-m(Q&1Pk6stedEjlt$F@+s^SC62OYZV-fz(qQo0uw;lsyCwd2K z!0)dbrYYL$y;H>*EOQmrShw(45plld?tH)fTOZ878DAfwJJACsTiFH_6|x#WJ>^av zJi7U2;{MA@Z2!mCcL&0?Zf}Q>AP7l@XdwiNp6Fc)qIaT<5=56MGkT2{Vv-O>CyCLG z-e$C@Vf5Y!QNrlGe|vK7z2}^q@BIE7#-6d)UhlKkde-xYs z1<34#?_LuKW5a!S;@$U{g|{I*Cy1e}Y0)d8UAIPU%O+`VL!|SE%4|OIthy<1fJgc5 zVsmxA=tRqbr##LtZuhpPJ|$wK_xIVXNa4SF3(Z^dG9>1$oFuZQ=8~1M&phk)$my}% zWAn4EJ@k6K+thLWQb!7=Fv-LS_4QUA_hm1%=KQ_6y6Vj=zdaITn_BIMV!&FYDW9JYCz$>#opGpTjpu*_%$0gwQ9u+W%mSpClFlZ zgFW^5ZB{#dG$#6Q1^4}d==}fa7VETX}y)ky4)%R*77DP^0I*rN)nrzUG9E0qHy|9=NDX)Vb zKosPap_zM0LQQhB^01G;8yJQi@8J@zWc)~d+I`;A#oP5J@pjGTc10(e(c79bjOrxM)}0VSBklS z((i~AsHP01Tl2iHA~elr3U8k}Js0NZQku5D{P(Qrf2Ahh8-HM5r!us;VBZsjK1%&c z_l|hFUtxSfhmzCx7Fk?y&BjgBWRy_S`cgTT?@(B}riTK!HIW3UpjIVIo>7>JS67{XZ zl-Y#V?IcsX<6Xr-q+s0~!R#A?-j#9;)9x$O3n(Mvvr_EpT7sUj_g8C*7gQL4fAij6 zDFC)@PNq$Irlg)>xwpR-NX2p}MUXv7HZf`n1bbBgEu3<@)V`mkQV?oDrhsDfHGHx8 z4Eh?}+pX}}19Kw?Rf~MWQ2;`qz;Z9*8A`3|2gntM=nP`)59*X&!!S6F>fZ~+KYj#; z;YPiA6~*b0C~THA)iuj4HmWbdG+&o#E(R4XKpD}I@9$J3behXQGT7_%ee*0#sXL)Z z;X;0?o&>2-Vyu()-#<41O|Y|XR5$(Z2;G@7bBdhX_%)&>%HMZ?g!*G(@;o|o{mAqp zz#syXN`!dv_~eaLae9AJ&i5EVq?}WZ2nctqLt%#SYQn*02TGhu4i2V;N(zd>&(&?7 zHagBb>);DD5^7F)ALFo+GU!LtmT2|KiQYt7RI}u6rUBtfVu5cbU|)jPwRW}rVb+OV zU?5gIlXEGpc46<^RBKa($Bw@9_wxD5Sw3;2qqOSSmV}_zOUp+F90szl)lT;e2O%-G z^DmltvOf#kmR0KSoxf;C_T}T%b?7&RgMstmjF4#t)iu>PHs=55 zj4hzT)*CxFHx|PeXQzI&dqux>*#M55&|XFLMz@Ol*7r zpk88Ut@JDw47l5O^(1QY@s5JsN4FV?p*D!mKo`enZ{vf1dhgD=%|LXGoA-}%*%?p72QZWLKz83 zpR}#Wz1H0}+VLji9Dx+py%gBs?A8K5gb9&S`%6;+bomi1gpU{ZE+ZBMV5^MHR=Y>g5`Em_PVw zq4>z8cRyc6MU`!(|E=J@smhtOt_8qEo8Abpmjfo)UJNFqA*$vc zL43c1PS676S*MLO%<`i9IWG;lsF(vjIwPQVO!J);)AutMsUKVn7mmwbJ^aBoOd+b%~2XR`n?tI5B#7t!t4EK%t+iW@W0dBwz|>&DjMOy0XL z)y}JaN2uPO!8~f9U09j}SzQMv)o1c$L52?nM`CbO)0>gISD-PRg8^TWA%^p|n>y`C zR_~!gpx--?Oj?1tRG5{wyWYgdMDsdbFD0xYLg~7j=@PFw_AqhV42%w3( zV*A`xJkhFcg4%!;nF~cHeHM)i${ct4-;CyER8Ce>?j>lA+duw&t`%;liq&h5xux|Q z_n3u5c~jgbe?Hg+VCa&DIU~UWt~)d}n}i!Mdz0yXjIHPg+}LX<*-d-9it!Ki!Fu_2 zE_Hp7_r>bplq~P@P9RFmTR0Q$9a~*x1o#qR3xGL5ZtEo605D?COOnds9DCpOh8ll6#zYM&GaB zmb~?wAnH%csp7Xy5$K&B{V=A6JB2=sFm{<7>ywG8h!lS~lopv;ZM^Ubetrc`cbmrp z1GW;Jrg0rMb-($KK{yJgJ*WQ0OPI4};mk`;w1B*sDBv08L0b3GrH#ybjP@N z^$Qsn7UnQG>1)6z!cyRD6`i77T{@$pT&moKd7a}iY{8D^r<>W8I@trt<0K>2Ju8tF zQ$oY@eDq}%CFaH+KM8?Z8|0^=o#u@|X=i9WqMXBihynKT*&3gZ8up!q*V*R0GhkTT z!5RgEBWEfBS5w~bjAJ(_P!v}PI|4M<9p9s zpAPhqg}n-5@uu7Rz;edIC_?{zW9odZ3Vlfhp=qP45Z~sZ?8o??iR%5G=!-HB+N>bj z7YV3_HvEYx?eh1%1jEvRwji7JhG6WamRJ0Dui;bpUm#$*fJY>A0eD#-M4&D6>6}c_ zv(oLk?9ZDY-Ra1W|by-&0~5wSMX+Fvdmv zcmMgzjn|V8$^mjA}p z|M=g3uAlM7r)>cf8oiEU!O}WAs{6to0%nN|s&9Mm#OH%0-As*3ci(z+0iTj$14KZ1 zLyq#-Q{`IcUV@%gD)^1i4B#1*O^$cjt{da*_(SkH!^<( z4RwPy&EMm9Dl$w=WM5S@u#qz#fM|@PmM>(EynvyARXf(KARf1wcWlxx@(D4 zH8kdk(Y6~7Ga!gx7#hQ{QahJ5B|UR&wTwU>=kgbl=(#CQJi8fUxyk8 zHbLH5*1%VV802)QrL^|u{rF7{xQ{UwXPN(A2fTcCCqQGC+h_X9|CTuIw@ueibvdKr z@;XCND#@AlL|I9V1JItij&K1TG$jJnZx0W00ig zWyonQaN^{;6UMOI8+2$K%zz$CF5hNYaTT8(8gvrj$*CJFH^^TreLR{iQg9d8Bf?Sn z&GhE5kuKm^XWfq>PKKrG4$lll$^XrwdBOqCRnOZk{Z9OxX~nuf=iK$HmBm;O87}Ou zxUsaBqF0eGR@uo~hA5qenJ(x1W)MMP?=aUUCruo*J{(?k$X)HziMe1|egQkz$`XzF zf$za=O(OdXGXLuGK?E>6_N`mvx7>9e4=q)D`j~PYzS@BWjEWKg|Cy;he&o=Gm73m= zd##6ysb}P_v7dyH@&lMG(}xj0y@IuOF>qxmJimFQ-&@IFc>L@F9!M*_`I7?sscS<% z5(a)Op-Q)#W)3Jz`QJqDo|z%A+Vq*xaPW6v40@-@`DTsAK6CdW*KK=%0Im%~DPE~; z?ANO*(|2yNK0j-fPHdbr&rEY$myVNHOHuT@bra;*86|b=79o8K&#FD4Q3V^fm*Lnk zTXK}wjH*2&<_9Xgb?3}zM>~6f7oUW6-hT8c}z4##32eR-0DKY`x=Igx>npC4v{t z&(D-^@$wq}{&jJT0Z+>9imSr4-E}7rcGJ*1_4}%Aeszf^m@ELbo>}6Tz38?mvHUqZ zYV{5&J~cK}RM@6+K{uYpCo$5WYmiOl;RUnd8$Fc-I6yWc7y!%r-G!*3V3kT>wDu1W zY-33Ts!vH%GF$wGWD%k!7LXkc=fL*4Zfrij(582FEs<^*8M$hFqtg8=Bhq+O*iHJp z!+U@3aVw!x;&$T~K$9_)Ss_JS$)z$jaFtQ|QVDNk3)z4J7`G7ee@AjPKi&rO6W&m$j?JVGjE$&Ch9)-DryNZ^JR| z6On#{Z|+u?a=(eY=o`Ekp_yH(Rlto^RNCM{JJ3CtN{lW4dBFRRFhC{Glsxu+zsxM) zv#`afhT_!bv!R+w@h6=W&s6Xo^wrS!?R|Jna2ix!co?KCX6f;qY%Nbtvm&`j%~{y@1z?p6D(F_W6!F+5L|B=7wvA;8rXoB>96@zY73cz!ZbS*A*7rhmD`!}~ePBJgg6r+GNxa1p0>G_<_D zL)hfJ-9G%=JTrS_#^F4*TW2g$X7!HoM7gi`$$++<;Cj!K@dOd|m}ZUKeh4_;IX-Wc z%5Jh$*+?yMpa_#Z09ESQGK`HFD4pjy5`#3_mU-{zd=_cC#}0agxPzYq*LMH8xdZF8 zTPf{*oF`LINp2luq*V^?DJQehub;0@+W z4pv4&XF7b}3H_0;^$M?>>G3Wm93(<-J?Q$)aL>H(bAHl(UU7zn#u6hXc`pE8(@RNI z@OV#A*CoJgv0B3ZHgF0ih0KT`W7hya4vx9-mrqK}M4^K#53P8((|fH6e4l1#5mfL{ z7mzq;pIERb*Pk3=E;BE4jE3@J&(5YSoo*cu+}J+!WxOTj@!Yu9*IbOj#C1U0qYH+D z2wrpqk24d9`Sp2f7ujMAo!5dn$+gVX5eC#nT5bcDN`75V*y1Nz%Of!#vht3Et>vK( z;r+&Lc;tF2PRluiA4_=qIDc#flt*3^W#6AZ6W9D?;DvWelbQb`;7=3%W#!^t@^RyV zpoFX%jtcSGV?oUt-^KdwIqeqnIARV?Bucf6d!!2_>4^(LxNLfE0sD$(cFa^mrSH9UpgA0X(!oE;x6kln_c=c*?aL!CZ~&;VP$IdrRaq zF}t=1G%#<2@FR~v07NDzj(`9cbhdg?NeTCi(nR%fvy~+({-?4Sh(_Ck3V&$k4>rxn}me2Ufc;k88V_s`LfA}O9-#cvGF7m^FF9ePb(6IAaExLFzf zet<5{*IyC+c%uGcWP`C20&)r-*jT%7f}IcldSW+3V=0yD62x_@(NOqh7lFgWJw3s7In;q^I%|ukyw%R?8F$jw(C&w z1hRUL+A{YdXM6delRoP;h1MA5FU4Xo@V)7Ksticfl~?S!+MF3S_jzJSw)Xj|juTk4 zOOC$a_mjiG)bqKD;TIG+o+PaJfy`KRUbkmZ%~5D=n)$50E=fb`(1yAo0A zTFZ1-C3QBO9zn`B$^hA}*jmP2Pmr)|+hHPsEmZwoh_|4uA zjnLO#SlMAt_EriyGAgO^pO2xv+a{=c6HsyiekcJP&LSC zwNZy~Cc7Ssy>Ee+k|xF`Tqp9XwJfu99f-f-3};AO1#zGPSMWI0sYc~iVq8{jBMy|B zoy{~}O!y26jlsZ)h|ZruVCG9o29B>HLqy&~T8>qlC*|95weSDmbQGX~pbpj`EwSr2 z?1O>BsRftd+!yRtZZ6Y*1hao4^nbk&x)Ybb+D9nnVx15g!u|W_yk(;B%-?otj~J(z z3{DOGl@4lA^nCcFVMn8;+FEmrKJKug60AE3nD>4TK(RX}yBw2C{2bt-~kd|E-Mk>mVS-)9S>}$#6 zdK4!B)Yl*=*1-SbX!EEd05B9NaHT{v%Vk4Upnf6}tcytuE+kT;OM z61_6~B1^^wBW*2XXXjs0s~|a8` zUQ?em74?Uwxm_^UtnCtk-6g!Ca;g-V$$Dhx&OkJDHXl~q?`maLsZ^y z`LV~2xfr*3maY`TF?rVlY31F($2IB)dy$bPPDbD&__PSS|BM7J^~3YK_da2=ZVBg;wGuyh;w9}e@@eb2h+%M+6lZA=-Ef>;oA!FE)kl2yNOf$0&xSZ6H8 zLV4DeS(PDBEICc&3AkES)!R&u=UpMDmJmp62PW#=TC<)yGV~!P8C=fS7`_Aw@k)B^ zb)0O?>eQRHi(9?a85y)MI|?n$%kw_}thhX8DbLLf=UgnmmxKDklQ}B_J@kucH)LO{ z`Plh_K2v1f<#DsRi))qhm=&-(W5f9;G4^u7wvOjw%J3uAgk3CJZB zY|454`=pkhz1;XVlVbCV>v$}-Yo63FS-3JR8z#ogy`pGnw7@`q`6S_SNp;RB4uJk^ zbKJ{%B79mG^Eh{<=}7>->M0SgNN~UXK@cKoA}0xFD+uQNwcePp3Bk99f5BIPOqbKl zRwkK_ySdtcAZ@3AgdUH0GN&4;O6E60%7fXp#f1!!j|i>W@u7F`QfWb&WJ(;%n()l% z{+pl24GlBYzp2e{2;{+3Fhxq+^h#blIp%t0y4OEgqNfKni+bEx`d{lqu+CI?mHG+G zBZBhqTtx)!D-)M7;nz>qsCkoC-|^b7yf(f3_Z#^OOw|*zzLy|#v=pAEv0_Y*E#_~l zz0{)kRO&P}14b~_7tH6x{$~1C)+2DMeL#GQD+Bt(UXuOUjD*4BaNDil*L4zW+HAgA zm&nnOoGGlNLj-VPU{6LWZ{rh$SehWVO2Di0BY=;}kAO<*PU$xs;bBTnFksUT%NU{9 zjAX|yhuha89u;V{B9khYD=AqQd-!sm+2R8kd?*($Lkw_EnzWRFCFKdJB9J4@qkmk9 zfmT$=X5C0Ogb!6QIAg_-M#`-`p!)2zVy~l1S{;j#Bz&uFPN(RyVD;s^;rK7+@F}Zf z-`H^3=(dALbq}WlG+ImjQq18CcHtLCC64)X9Fl-c=ifk_FyP@fXZ(Ezbm>{C;H7aW zJ^Hit;L-z6NpScayH(QIJJ3>6k0{%54>9LOAJczOF?aEZzcP5|uT->9`jCkZcQ^f7 zZuf|TT0df+hg5ltdW--OWR~5(kkhwi02fG1nenTVL|!=@yjw)g_soY?=T(U75xh|A zR?!B<>xI=Lk{PyVBG7V#lE$7#+b-J}3_N3SkI-nU>gwzW!ts^y*yO0tC=Go|wKLX1 z!ziErGOB;w1z{jaMo9!NRrq0BMrHc!M$1oEbpa5SKHysw;?kjmjeKYx4 z@nJ)v@g$+)c|n8_+xmXcAB@JI%PUt_#vDR=U2h3(D14*Cyv$SWC9s%+DSNEC+h#U- z7%auUkCqe~K5nKcBtB@zJH89B&cKr@Ukr=D`?nR`&bX$#GK*7iF662gN6cNEyL`9DOUh!gW&&&NUPU`}~&beZR5R&LzlkIdCXotP2>;xbW z&+sEAG5jpN&*lmWI1@47o)Mwff1V!#@RhY?qA>6SyhX$-<8_hoUDl16HHH#aM-^jW zE5!nuk&$N7vkkJc#S)q`uBV$cJNuLa>-;%3h050m8TPcjKpZCag-2_$&>)jpgd2TF z)uQ~F$~)E{FNG3S5-l1)k8Lb;SFI+Z4-DK8%&T^X%7bOx{9ld@gc3n!a9(XK;X+k6 zsYimV*Z)Be{Lf%A69iGSBZCzaV)@FijUy8q@ahe8{R^EiaRMgYPS~vil8DM3l zS*eTnRr7b>|L@BGtXfk6Jb5EZ&w_2_O-L7}kqQUUr34M;il@%nuN0rANs&XD3ZLQ= zp{W#7QMh`@Fy~_ZT6{k z1BW^Jd6*@r41Zdg`~HC_V8W-fB)0)c%$8f@+&b#$=y>UgItR8`_UYWmq9utzgPtM` zrK&#C+W5GbhVrrK6}$-+vhR{t?`!J^Tg4VCt@{Ygq02p16D@Mr9H1uJBZHt!;s0Zq z>D-{~;E+C$92_On4-2U%$ff z*oVCc83Hax7Ht>*MydldBw6^{W0bm+4>qS3hws_~9KujwJ)a&A!-L{#wxK}6WE zYcgbod2S(4fHAUDs;9Kmhk+pwteB5o?r`FBujRV0^ChCK_pBYvdN%_I*1VzP-p+F< z5zE65vVOSLrq`}|2DEDM%!Uz@tMS1y7wMQUGqmz3S-AGnqg|P6*?^R27$j(Ep;@Q? zE&k#4v#M%(wZrHojwUR0)~$*o!W=I4)2!jYvj9j6sXO(8dqDfd{&fQf8)t%EmdZ+)GwWOwP~u&y96cxnX!JMGb7BR;~#g_NboHH=R(heNC8nnF5zen3zWT)Oj^lsGUi9w!V z@<74-+qs)_C?6s0Z;9$NfdI23V(^Flxl@PUiqNCmg~bG1(r@^Eo@jesBEHO2NP$~` zfys)$79e|CbjVLv%?13)Q5neT&*ZZlTLIcrAMS^;{KS|^JNN7)ePMi{y|REg-tWi= z^RC|6j-bZn;!mFG?32eUiH|emZzb0pu`^HfmXAF867ge9CwoYF{7vyP{ynNIa?P(^ zN1aJtGC~^6j(|%1^Ep9yW zlCpe~XK7O=2i0l~uc@N%1eo9EYNGB$-YwdNw}hcV*0dBoYp;zTf8%*~^LJ(?f+cOm zt6EU%yYAA;558VK5#Cgs6_ge0`Qbz+ixOVm1#dQNqyRe0Fl7LRxa@PCzF%wJ$Ow% zj(2uy*3T;jgS!aaKE^a@$C6~?!@$9T%Gj$^P)iq@Xhl?)#aNQumwS)0+>_gY>2)^c z-W5m{N+_A#bb1&NXg*Fe=4W0nQFrJQiemitRI-#HirDW&3#0x)yN`BGvcRZy64O3V z6d0r*$}$O|zHqetL=0{bQCV6{Mn+`KaA(#ly~4lYHw0H!k(piysajY%o&r|f0Lb|^ zZeia}>rO#(R9<59*dvt`-oVF&#ru>pBcSWg!v6??0f3Sbe3q9e1}SU4C!g%$$&Qqm zq691ReHBsis@$8sR_V77tAsl2iTC`39!=kttVD}nZAEIz*t;54i8LS!0i=13V12%l zmQJpRX9gEDYUT>!&tcgUr zL>i~Me*m0X@2iX(L^3m60nyj}B%Keid|z!u7y+*dI3W75?J-v{6j^Xf@sDDkm^e$u z!n|>KDkn_|9BXe`y)q_%!+T}+puDEaMl+UE zi|O$b#&X4ntw`r8`>|EkCx}){Dv1 zoG^2}TnAAou^q@q?nCjY(G2=i_b11WM`d@%RA~NIIsGN2uJ@$qxsTDuC(+A=R81LZ zuwOCah)H;hcl@9&0^Tfn-0t<#VxV7$UHVV_YtAym!xfeIy^hmD(?{6uo!RO?82wzi^CWB9_1 zkAX487%W0Z-w3_OYjW*-&z5b+V7`*97JFvNkE6#fafa{)rSR7&C`;Ds3jM2)X0w8y zL?foXyqfo$T7LYqo_j9FJ?K{-YT_*&nYIu7?qVcLU+~$~;6VRwXfAHW=Me-LNwJ?79)8ZxICniyU+6?%q|`fn1f&iQ@?kd&5xQRu6Ir7bGU?6Wx1@OV zk`#W-Q$ed7c2F7|VTGK621IbZVV?L>Ju^64_dEy8*EUFpoq08TBTrkq<;(cuu4Pib zm>(kQOik7&poE(2t=1)v-_YkTfvic^RI4RfiMb_#_4MlSIXauS+oGRs^K`!@30bA2 zkd~LLw=aK*C=pfMf5o9Ujs%S~o&-)7=26dsF^sz?JEHfLbz2I4rQm=45Ewo*Y;& z%RG~NiAOWnh7*9q01$gxAc9XLsJ`3qB|B z>ROh$Ud!>#6{wwy7F&r=uZDgwa5z&Qq|3m)Ea!vj=~5=r96lZTaDJt!QZ>pe`!*8H z@Fs*-o+H-~-rVD~(k|yF?ZSMk?6~jaQianlE4kE1Vj#=;jT`yT3TINq7h9;geZH+Z zPE)YE=u|zY;V2kh499xhS&wrRwz!`*&7oh+GJK#nDmwZP(DS&~qbCMgp#8X$1={ve zr}{(Ykx)Ow{?4i8ZHPo{1~TFbfO|iQ0Ez`hr-xeyVKD$hEX36tJH{2tfZp-|nkq&$ z&mAa&3kU+ngx}v=u#h!@~x=x+NB`y*i~@c3^%!T&QtD;T6dqDR>ylYHFF0q zcuj9Eif#RiEpY!_l>haP4o&qkR`oapI=nc?SD&5T*J3{CiFSFgr7>G%seR@x&WjH8 zzxVx~-Ncix1`#!X7j6c)0xjc%(GeUk!6%C6QSN$r8^J$(^?FiD+c#XgsS6}i&zqIL z;}AE4C9Q$ZuoN*HX^_vO0Xj9&TPC$wj7PE|4TSOc*Qf4_j{}WH`bpHsB9gWNhK@uF zDIWO6&*_%cS7U5dnyhNg^%S^5`(s5M)Dw5tAJ`D16FLu!kU3+uvFn~0tC*zz+J_Fe zARQ1UB~g8WjNEWE>b~gvNZU}KwZRo(K%I4JO8ySE+W7nZ{23c5;y4484O&%wC}YG^ z=+`+UAk$l#&+6wH+!kObFXFE^@OW8|{8scnyNl-ooMr#beV{5>D4`wjfGcV;^3fr% zt8EI;@xcd%%9nM6{Lz|jvSN>gf1wKrp8SZVtahF-?E~RT%EISei_S!7-T9F2BF01S zE^o;Bl{fW0f}>%wGK&y%JD2`yr8?ox@|6b=VANwQQS@gjmgHfWQ&(=k?&GtykI@<4 zLW?5FHUtA23X^-9*?{7Lo$dBVl5PVBy%7Y`V7VTe@?|8e{HhonbU#JR&Eu%(MXrs& zavrxaohvx0bwCh13Y?L?w->7cFLn)2Epgh)u>ZFuWXgxfz^=k}O;v`U?Cs!M>HPVe z66&D)aJ@7(vT0^>dBe=A*EESNq5$)e^zZA?6hX9FWCdFiUuKwV@Ms&14y|BN~W;dCxTqGa~vg~ zUY1<1^q3pN1}kec483S`s@@6)Tgu0kz6JU|0thf<0@Xr6G5=tBWd6L`xyg(P^W>wb%gkrzueA3I`Qdt#AsyVK3XAbif4{nrfZ#{6 zQ&{;Ue2=BE`kN*k@U&wAJm7Ze@qAiHl^UOzG;X;7w-0px>b68uw(fy`WwjZ1-n-CO zH6bv9q{|WL(n@;+BiASP0l>7gN~3ZLd_x4FPx0Q=sM5T@s@0kuEMWZvF=)`~l(AKO zW01^_G_)j1lBqH?ZbA5Q{a!JejyKv>XrFX4%N%=zs z%)=FzFE2Fb4!_W<)m#232Nd9cjTAtX{4cWby~L+sfa}cV^*7&N9z)FKIHebE4;6}j z4l*B|M|%VlnmLt5KKMn?{{@!;+1$~x`-3Odh;_%us>vpNqPUNL%aryRi9}%g?z4JT zf7?pDsX>y-!+}S3w6im*rn)W2G?0AS&10*lTRYiXaKYHa!j%F?qIeT6HKo$&Y-*2M z>a2v<22xVz_hTQK-XpdLSVE6TaUMPV**oh859~Q(Y-L<^ubB;tz@7nIc0kNx`7lLj zuBtK&yjs3a0Dwr?4F7U^S3p~#I(Nv~+wrxv>@186hu7I`OqHkQv{5&hrh;he-*cR> zw=xES@ebe4tM4n3y~wy?9`XkHB74O6)t7nFTo!L5mRhgXM1(K$cTEP694>i*-NNi( zwmh$}36Z>dK#|o*d@!3&razwbh*!A4(S46}5!&0Gm9@eveLi@?d%o)@zSp84je42%?yH(zI4~W(vq94Vg(N9Tzm2<@->71NEo{Ee^!7{*HTf@I$ce={~}x zS<8J^Dh3n43sF-`H;V#$0AiHY`>H9OX6g`_13otVvC#*|2e{7a$hfX7+(8(Q;(K6VJI`*cwA$ z;#SbOwNVs1=OI6eyb&-{$1Qo<8dmy|;1cJh*y3z~U)8B6<+u37Q%lLVydg1t3b% z)Buv^MPm`!OOb_kyy(IeGU`m8=SR{ijZ|bI1>gDqaq|GJ^>X*p@OKet(TFb}Jq!pw z`~dCzZr{DP_wz=dd5{$UckodJ4V}@E;lN9#9Dga2Ukoo7H6HL^o?KZjeX~p?(9vmE z<@wN#iQ-P7J_lhE0)~n4-`3{ae=emXD`1H>iwO_*zqwkiJO4mGy0l3>QclYVDYgg} z+VhA!Fenf{>eA<^ZExehEt;`mj}5Q<^6?hN)9 zV#=AhlHvi$@!zqBj~XB8Zud&(sLM)xnY61}0|!O@luHKYkIS!|ydkA+i<0Og0UU%_ zuQHLdS1<_7=|uOF0N~@p35rK4Vc=P!k3e`^aEN4j0&_F)srAd)GQP}+FkoexSB$7d~8stR55RH@0 z0EC>gfD>OdQjtz|d2DxYGC-rEOr$^Bs=H&Gn)hw^In}*a5hd?F0)^XuVe>BtiUU41 zrfAi0b9ic9|Ien3PtE{cz>%a8@jxg~G`*0i<0ZG(KdfMa&-`h_-sKT!W&BnHlH!IG z;2R$(dRkS_M_J~ceJ^X3^sBrVb#&LFiYUwdo$GxdFGbzlf1?ZhncY(vaFx)t*Npuh z@*{!=6SSW{W2nRGhZYE5VXuAHHe{QwNpC>9-d)adbLhA^=-EkGO*TYk7|F!^u1nOO2P4T!t2dOWy-5UVzRW(+IiUd*;m-osFhe zw{u7Qj&9J>kop95Z&ApA3MKXfV{s3Kh(i6o>~r%CbFoAQo+FYQ4VQ_yXz)N!v(?+L zlaxK7*P{FsMFNdw)!`WQf0d(s^wclYBv(`Be&vZ2WI%?+j8DzS9e!?K*ToUEh6C>?if~>J81m2l6$#U z<6i-UPcp!s&3pxwR5V{XB)5aAk$LQ}sXG(o{nQ}fBKYAmqkNL#4y)gt>!w^iB{aEq zEc>`*cX@%R*7H6g3C++g7!X@>)tK(yg&f;mLAy3ReV!Ao8z)mhM^D}K)z0JPo@sOa zRyjNt2jkD^@M)Z86xgnR{YIZz<3^7T39B8tM8Qx~NFekW3GPlgf6fup-nWyfnICSQ zIr&meBgi_QI2c@Ru2%xeH`hZLTq!L`gMqJ?7EBl2cPrlF+6r>_RQNJ#6bF!6siQEk z2W3@8a#*FKm9NKPJG~rs3;~hyJ&N{l+xVoG$HQY^Zf9^#ApX1i|Ib|vD4!xU13Z$T z^7xOMODVc3t@T&}+BqmRZQU#n@56Zt z`;0gtN>XK{okuh%%XP30Z)g28#Z9R(FAYzbU&^CAKD^d4&K(v-Lbk3apH5#?G>|Th zEmMwMGt)C@dSt8#M=;)!ystya#Vs~y3+cBlJAQq4nJ0r8qYxFIw;z)}e$LL1 z(QeiFRT%G{oe*8i8NCkpX4k&U3IUpWA)Oo0?dMLWacsXR z-M^qvN2-4d-sW4a1{j#*h=fM4yDk8yYR$V7P+$3;lr1@_@(o@jN2%QxOf4|B`SIJcs543O$ZCLk8M*Ys?CIy5We@jO{xWej{9+{)u_kmLVhJgl7IgF zm!y>u&@_Ie{#XZ~68U@l_@C>X5T*CcjsW7Sv1=bW*&2CVW>31;5_O7vp6v;4#Bm>s z5r@7YpI^iY2>)hjASIcmmJA5y#1>Dj{4g?FkKWvWOAF8*maJ)?ds7fAy_u>qBbJiz z*s`moc9-zG>bQrfWx|)+ZnD^j5;I{9x{e>^;yzP^z*(QtsqIH%@)mRP)1W5fJ&(QH zg;;{}it>)S)3b_&_Nf$Q_v4ET$*h`hda@UFWkIx zP2PrI^z%H65aoa$w`Pp?I@kIgj8uD_pLNXd$~|uK%&)CqQ@%w07f<>7OS^|m(S4id zeo5%b-c3V4Qhr4uNp|mI6KDgiiL{p5Qqm=+ZTyhrHq75_hx?4UD*=VCIKS7aihbrW z^u7XDWiO%6h}AEDScYqa`Sli#2u0#(ZJt3g5okLA?AnJ*l*>7*+4}KgWR_ zrwUUdhG1~B&k(Xn5>cZ|i=S+$fDebdb^0APu6GQcp4RUlZkD8Wnmm~R%4V+g>E#j2 zE15=S!#^6b^G=f4v;bZSn*Ip}rmQs5vD@|E$b0rJ!W?+FL}3O61N)k2_YAzL_py%> za`z~kO4((LkCMGkXtn{&PGMdf+bfP=t}pnsySU-`R*QBk!iE+ePU0Hi(oEwbB%n+U>6>jpbgP=ljf)O7?rKr+QfL zv{Kx3gO=2Ada`<0*awa^SfuWiTgFMs-up`!HOQ@HTfCCE=j}^)FAoAo0WD49xgezz z2Vjjh35RCjfczT}mcAu(!u!p?GU~nu2)jFlEIdl*d&e(3BklAo(i+Ycwssa57u5Vt z(V}80olZ7D*B)z2NRK9APD(HuiMei7Q(&_e~aN!+3FSXqF$So#%o8-a&YPER5Z-Q5L1%bDFNiJ8td z;T6@|9H7W4Jzl{dDPwDnQ%?W7$*ul~;+uuV$w)%6Q{cm9K`1SU@Q061kX*+^6c0tI z*0UQgQqBL-0Gee_D{w7$e>6Q7#f;n-!V_Ec+N^kuTdTXwxV5^w(R-U^y=v}5p{iHt z2$D1)y53W1HTEf5H^VJ=h|&+Vo3s1&4Cl1Z@bIhhM4-WJ(_4Tqy)o5IOLR4hB({I) zkl+XG!~t|1N1y4NH-YwF<}7lqs@JB>AOk2Scpn9XyE|WOikTDmMhShB_z|AKOtJg~ z==NPpc~fXd`#pXb=~SUl2}MNd`T-+*jL<;C6hk7yinKnG7&v$IbM6NY{@OwBU~=^q z1`{ioYZH{!-8Rl%djQ+yh_0NjPq7adBz8OcJl|r@+tsY;<#SN7SY)vJ*9}{FR|87< z$*~bxV%JmiFIQaL)-#Q<-CSldRqy(oDmaDIe-*$nw9CnK?QGOx6q>UWkXDzQZ<|&i&x`;T3mo z`%*kIjVq|#dw1y3r3}MJY{c=nPBi8KX;4?~F8aX^F6Mf=vGhjN{k(R%EV+|=du9qO z?;Te@F;2^jEuNe_qAvW`=$c1{$Gb7mk`2*Sx~M`p-2SA68H)lz=ZiTN?Y~V#6$nQ+ zzrHWnj{NBJI||%Xm6^g|_r+R^l%9a69?1AOlyqO^4dcfdoQ;J7wtsyip~vcKJF?B+ zxNu_9*g83LVq1T>*qD!iI-}}gS$(jL5E+g-D7-~5z-slFeL62)Ri-N_3Gx<9WaNqu zl&>e4ifMpXo)q}0jharTE-*v6MB+P)6KlI)zyZ3#jNKaVlfK`gs+zPL zHOn@S2qF3O;s$ikmP6u7`O905;IYe)tIGQ17WrZH-z22w92r`X{lY5XyWfe>dBQ57 z%WG@)`w?E8`1#bc6Noi~xpa2vQ+wk&A5Nr$MqVG3fF_FTNr zo*+fa%kgab{NM`^rzhyKsSX^y*2ec-HSYYCw(^Rec*nRGg`5X_LV6IP{B#wmR6Ye zOePtq4$D@7V3#`pJRx_6shCUf#uz^vLmo|Q5?z1R+YERQYH*s-h=d`6gR)_zPVg> zc=nxT`{1XR%f|P{N3~k!!vD%v-h9NAtV%dhB(^%*5*Vt@B4>)iHmE7^=Ar$v>M~WI zLp58GGlP5jd%rVvz=>S6SzqkXvtg)5!xzRd7j`Gf7LzL13FaRIisdDD^_i;KFvU}U z8LO@W>XfpjeCi)lYxbAT;F0Dwd#_ngO8s24f9Qo3W!UE*(3sGFp2auh!e=GQqs4<{ zAD_PTGhg#Li|_T}tq~RBJ3mYHd$2C;y+bZrtAi7|iStzfy0F{QmhCwhoxU1l7J2~# z=Cy>|)v_ZR%FRmsV&N)dtToge>06qOfC@|Bqjhcq=r9L^r?y^yBC?aJ>Qt# zTE8BeY5ID670{pE5)0&EU~*?)vkh%+wd-*~0Xg#TPKaBXwgdt_@KqQB2}a%IUA_TB zqRk4=%7CcWEvkr0C=vonDAEGb z-5?=d14s-h-7$l-s7MJY-5`y?&^eSyr}WTWGYm*ezr%f3-@Vs&|LG54=A7@_vG!VP zC$x|HgtC4tyNn5Kz;fc57?4_C7S((W*Ms(uI0ae8n?cphi| zEztJ@@{^!zO}LGlp&{xS{&O(KI;iV(Gtn@SZF7dBMzL~sz-?h|_adkog@(WBU}%M? z>rs}tHiI>oKDz3sst?9kMV{AZvv7dQ9Bsi=C>ZPn_-cW1}sYHs{ec?_rhN_^a@DuDVI5PppsZU zv0FAiC|j3gQk4}Vwlg5$NM4OFPqp~1n))-^8e@>Xvc5l>i~#aP zo9X@7p5-0RDz4|Q#GCEasre%wX^nz1dU>GUv@2%d`zjTYH}IHa^PjeZ&~?9!-o*DX z6VOCD7K1o<3mihpsf)j>>&@(jg`0V+@=hGe6?p;WVACVMdmj3ngx2_= zej_l{OBRLj4czo@RXV-u74h3Jck(W|dmgY53Op!g{S<+pVPEe%hHL4RkAay`~xpt|@ik zxSMlXsw$7Rz?su5wO`UF_Gbt=%W35T**kR#F_F*s$X_#~vKyIv^d7kdY zA05S#UZb3yNv?%H+R#n*R95+w4*$~eiYH;Z*5%Pj)cvfgU8%lQ#pAqLTx8U`l2S{psxv?Q|HP>@UP`OojgvI?fgw~NL%V!6mp9mOp?lz0?fIgLFK;91W$CT0q&}5sE;nMWloZR-&KgAWLAjrPL!IEmA_6kO z4uB9DnPHPFpQSV^Zn~|kk1kiJ`71wmLR(jO?pQOhd2xQoJd2O6*FJaGBW@kXy{q2Y z(&10Qx8%8fFxM0%T|Q=AMWHqY0J*)(3UKp_9l8uwJD`l|s5|-o$#DNXt9vB`o|=d* z>^Tl5Y ze{lws5*6SOsBX2*j9Ez=7o|Mi;VjHxR;J@3*Y+rq^WA;3K_@4Ru(jd$`M?}gA=WMQ zMlDW`*CPhMFQ*x;0Qoki4h!x@|4)j49D|=nnZ=HODQF$$E#DRLf;W+KJ*Hu)#vy85 zv1?<~KC!pi7L~R#&V~ereKCpRa2FGto0?Xc=Q*r?u&R>UI*B@xtk|I zJ>b;pCJXmj{(Q91M_E>}j=PSBHxn{6@zM9uKb!KOSM9ko#$(y%%i)@}YL0Khbktnu zrP8bQ{HBs?Xc|Y2Zj%!pf=wGeJ8MQynl1_7f4P`fFjthtX9>}6f}w36@5&N+u7{+M z4zXU6to=H$*2-iRx2`>Y)c?@^bZ;n4_$+cm|0M?435}CuHCFPjd)}oS^K_(UAV$IJ zziEl!Dlz#>R&b3ijBO`4_hqnOlmDad%V(bSj|h}6wx`3h)m>~o~T@{B!8-SK1YPwsaFOJZ3;Kq0@Uc^!->Dl(dyxyUchMPY{r&EVFksub!iA8Xb;vtIC$?olIa_;g(n(lGv*{F z_OFlM977P#_Hr~@XVTWkzU;F_%l|+~f7Fb2?YZOqSIjtNTKLN2H(39vU;_A3{OM1+ zd1S}v*7NlcCHWTljR18g|MGhMz@$*i6Qj>ul#_G+j!0MU?a64Dt(5l+3+&c)S>G=d znu~J_LoZ-q%DR${;aUIp9X!Gu-j^*@;aP7!bb5R1Y48F&rq=7GdZE-yADAxNX!*1A z+}**47smi!vQuTDc^LWdp+$~vVTu!K=dIA9z8F*ZqB9)^trC->wcMnw(_pSmy$jQg z{1GHTzmk#b$Q%s2B1f6O5)aDFHeq6`_@gjmbAGDLkJuQ#ZI`q<571F^6BEz8w~x~% zDG9tqs@=u_2nP`muo(b`lOZQY67*I1%U=@Ry4>;}k{D73%6NBu5`$}$ccNq3kux}q* z2nFfKrl5hj{rRPMs<+|0?j-CYF=coe=f?VqBh{v2{;Z?RN#7Ok7B;t94KdE#|(yeNT*(IM>dOw>vd<7e-JG zik*mwZwoXYRblM4{Rt*u8n3!BM6K*xvJ|#elR9m|uJSZ+?x$ z1WjpnthSSI6bRlqzIm1Idrja|pToqW-h=F|y$mt_i=9K?gsq1nyrJeUEUKCW;Rpgy z2vZW_BeT1YV|y|6VPw=8NNAso_&rVS3SFWcHbcO|W=Tm#DrHw0DQYeR<;h}maoZ>* zc}PoVtA-%-xJN{srtu2wU2^jJuNDc|$(< zo(*6vRd_hsi&{C3_9n=wp@yNfVS##MIp$@cq+FdNedV{1hVLp{R&cIyk>uQF+_{JRG<|q?)pj~!H?;l90_zjx*6hVxAl!QY_*=wCUOoVv6W_K{ z-&Pk*jr5DO^r2ps!i*}NN=pbUwsCHUK_^M^@;Wp~aagGHI=CM%uqrXBYWU2zVmehQ zNWXxBkq=_x#5Zam(U@E|$belGIdQwU`$ft)+{CXj&za#g_ zchADZd*%n!8ZDS6jxVk0fJ*<&be2DHnzZxP(DZNpJWSHr|vfNcC>1VGv4rSB< zQAx2grYm=zO}Iv0E+|FBf&yN!I@Bfcd>0RL6Hx+b74|Fl!c5nxsPj5Ym%gn2w1UNX z7qQ_5a@EnuN6CIowwU$?jppaelMYhlWkif?55-O`&$rLQudY%FJ(ZrSwH=7U@v-Hj z7uw=JK_ap+XxZ|L^MyWjN7p{6U}=Y$e8RUPudfU(gXqI0&*Eck#_j_RFLXRZ&(w7e zd7e`9Xg0cN(|4XawJ;S%`reOptz>J~+@4oBVPTz5)mpr3y@z((d^G|ACob7R`s4IfF{Km>aKyO{Gv$ z`c)gRWv@33UVjaBPXEr?XHH}M50IisAA8L+yn;BLi1 zLXdu)!8a4E7KOQq5#|*`-wQn#cfK{8y|b^fD_a1q7}P?n9Bs`rR|&9Y99>7g^F(?< zDPRSSlgDp?p*op0uO4UdqRcbKC0v>0(B{aU$a1Tgx`~IAegy;80)61JD1priU?3^h zyvz^nR9MMe=QH%Sdf?>*@V?e749@HLOO9qEiH1I$)NaDS%kr&9LY_w}p#hx$!LR*3 zx~(Gr%rpaVOycb&b~;p-$jpY-fg0RBsbJvt4Aw=8|C`RPV3Np-UsGK_4Q1GLZo@hI z)U!505}YYyo%E9C?p@}!ePVDV<8NBnHd?B5Y(4%@%m%=eN&)<-HaMm5oUF{U^m4*@ zPrAvc0BBnRLV;$?c#a!q-d`l=0=7mbEVN#1_9MN$-(<-RC~Kh&+NrfeIWd^AIUpPcrc)2gR1)pmw9yr zVRe&8<~_?FT7M6Ip*dj#heYEX`HS96I;Zs(MT%EqwX=e$J}N40B6xUzxNv?hW;XR! zL59lAi6*-obtT&eHEmt`B{C}vM@JF7@=W!$wx6vN_{iw)uvI#aT@CfdHy#`4;>dXV zkB05fP?|5@bO+N+QLpWliX?RbqsL?!x;9fh6ei5#FKrk$Iab{Bd#{1Ou1+sB;=v|> z(D9yjP?1GtY)3<58^cqz-VAO)od%B=)2&PM#v?(BszU@6l+Q)eq+ zLMQc1oh8tbX<0~JdSFeTmpM$@C;qk*8VPORWrd9*7ciG`-<5jFlgmz4>+{m^akhSyA=M%6d9=1^+EY zrP6Ffp6>-kK1;$Jo9QfTCrRl8-6v(E^xKqs>Ah-HGB(KLb(?&t$kpI$-;mkQ)_iFo zTCIw6V$s~*nyKdtXm41PZw&w-QZvBjG~0gle7W0P_S~m_Z+oqSBCtAtL|3Kl!Oh_x`7yJfTH&I~a z!kVGI1KQ7mY&NQ)?jtB0y?#QYm+k+}Ffp%0wUw0%B$$tq?&gShvwXv@+zb;^(!(8l z!sTbr$FiaRXm;GRdny28HMyYG0>cHw$bWWrP0d)jsQ$*YO+A<0V;(2>o$G4-KJJee zV7>_%8P1J%VS3uP?4RopL)V}6SXsYccg_;aML`IlRxh;W$hIp!umwKr*=o1?=rYN_ zXRJx{tl~1N|zl zB6jBwq%EBaAG#?DSfpI^}bkv(z6DLg#5-53w<>dT8$Q6d{&?h|T5*LtYa!bVfXEDE`(nM@`#0EK@hw~b9`@IfKKmU&ATPd4NgM*=)7+zSB4$s_8ZFwk#jrRWH=t zfwcG68N3Ybon!?stEZzb5sB=6k?ahbI+Q_v zt+1gl(Eo>k=+Co*PvlpQxm;QSqP!hQ>;$$THNN$bW~5^a>P%`QHBAk_@!R}*P8fq} zYA1pLhSkCovx~X%9qZEcI-M#lJh>?@Lf8wWVtwJ#;EUV*sc4uZ>SR3FUy=ol(q>-?TLeXy9uKL1nm9VUU-rXqyxdDj{w>Zj;e7F35-h?O)%6;U=R^`k9xZl_clySV>pDh#0;rH_+#l$FX`F!D+a zfBo2qM{k|Cn-OU5*}tfonm7r*mFU?#NeTK|!pzbq&rf`8!kHMrHENfP?8JFsQxldP z{TsO_?%LH6n`Ebrh(q2Ju=?!D#y*FOC5a}$5wH<&f^H1Xr)NHg)-1~VtUld-drGtc zcEn(~Hum`)xQ_KDgLd{2paERb8YK96H`^Y4AgRi8@ZGIA0-U(K-9DK-au~Fl^|wxP zZtIPV!Y5^VCr_^rYz3ZnPD*Z${64>9UEv|l2x4IkFx8f6aWcMHs?1g_3LL)Q_M->YV+>*LLONYRWYZw(O4XrV(3eFex5#EqJdO{qsf{gag zfH|wt*SBjQ9A1Ngw;a*IgdAo+FT&s9NGU0*U%ZRoz@V|(LFLD*)b$l>p7=QC{Gr^13 zLOv@!(ADcOE7!(VAD$HCU9|+LTi`N*pFp&k?jAWAJpTFy6s??_Qr^R%iN(UQ-(Kuz zjoZo@N(gCjik@_+Y)IDjU7SDj{oh7GRK>iQ z8zqux*PNlSsYj?5r|BKJQKqiiy%hp|)yyqb9g28F|s_F|S{pejQxbx%iR9+ocyqkcz7`q@}+x^@B$y>PXX zo2o6BU{Y5IM#aY_6b^WA@AQ_UgB^RqC_YAgcy!P-%3`MCX(t+s>t#U}nr&5a>d9+Y zGXnwq3bUDJudz{j{jl6?WNbAx7>~4_4bGMK*EA3UIleqY5|`Ci2*`1muJ*3FEKF2b z@U{~7D0I!QSqezm8zlZygZpP91)S4efsMp9UN@ADs14sNr^XqxvGffBIAX-&idknv z5@Q`fmBsVv-PAjZQnqd{S)2=_Kdw%SGhQ3GQN$|~k5Tjt`<*bdUC+`T| zayGGy0ed`LenMg#|5rjFeKbYRlr6oTdd#!J$FH;1Hcc0!+zlRhe=N>;ZFv>~1Pp%y zm}fj{Vgwkomhl`nq>QPYc-7B**b?s+eE|YNy!7!X5)W^-7j_eBg2!k1goGwbcxg5! zAOJa3pnL+~X8G>Fc+C58U}b6uDz6o_p~{7J1SVQv_EA&hcZ9xusK_$a(jwysEX;99 zhB*DexU_otTgfI~fVo${(W0f@!JOH`gIB#`nF2e(AhpqBycFy{#1JIA>#HQuWfi4%y{+J4 z)a+(?FRE$_->p?LQSzTC^MAj5_M^1L>`K3wuq~N-3i{$Js$}c6&bRIDeoSE-$KHIi z!1#B7#m}AKLu|{dYDut*)VLX(BR@N2tjc{FAL-Xri_iaYRo*d*=@If{CeMc~3wS+b zzQo|nbqhTX`COmU$P1|tWN~uewrLHYO7~Pwbq0Z z1)=_}!~~71T%23KZu}O+OaialD*Q9SG>Q*1RC+W4)xYFfZrvx!G1WNg{h z?nx)R)Makji_FXlQvmHc006B8s*1PJEJ_Pz-o2;ooT5Sik|9u#HTqE8eWbkCSNrk_ zcyX*)VpQcd&xQ#2kICpCS;vKHPEh|yGFPUI2ZRL~EzGB~2cK8CSZ>y6z5v9Z!hp=a z=eE-Z@)p78WIe>a`+z7YNNSKDm#D3pw~xi006MO0&D1t{oxChtK8rdGd^`W1?{yiS zzIqXU{P|e#nJ%&+*r>auzza0E))rt2gaW#`EUglz_jRR>Y?7sg=L2sJANI6oWSm?4 zn-gI%BL?zAmT9(uQF>Sy5t1ngOXGP`w4>Ji57PjsreCXob9(-#cpCd~@WX#LM^uiY zS5GPwUBO|w3RMQ`O0g|(=r3PLVAQKZ(h=D(9C>sc^4PYN@;XgnQZtD>@VFGnxi@}D zk}E9VFeWv`_<6Lb@VTxM16x2ACg@(%D)q()>O&M?QXBMZTJsE5wnzXx>r5^ykqYj| z%xF1)CCy~6o>NjCQHtCb`NWr03;EQNuE83MiLO7T0)qycfnC$iq`z)yX;6I#Ik;G)F1*CUmm z#1=CrD<4W*cWGzzx7}}*kTy*zR=94~O@$ovv5+pvc`f4bQR*o`??jIqcu-WqYPZ@d zbXXDqpOgf2p;FcTQgi zf{iNnE-e462Exys=$n(i(4dDx8xBOx0F=joYrE$Qm~D;$T#*rg*>Ya-6}I?yhj8{J z@vR?0R$^yjON|%Pt_G)uE13t*FU|YJsiedWyOXrbh8B-yau#ri-Q4B+GYY?376VNt zHDC=!jQWzULO~owUtkeFYATjWofj)fqAcD{5Ky@8+KxjyBSnT!{Q<{&^~#yw3n zN|&YlC5BGZb+*&RKge!z5wF73{imy?#+HSB>!^8D3P*V^`Y(<~1)}8b;`9vf_Vx<- ze79p`n+%aQiYvB3Y=?}%M{MjI>9&#k;SYX=Zoe$558Z*qKz_5Xd?y60m@M>39LZwj z%6gV-_^d?;R^7$tW9KI5En66vPq?kez+RlSP;KV4G1%}KFwP9efph0|JfqoGX5RP= zl19+=85&HmL{nI>Kt6kH??4O@`+z~u(Z#su5WQ^WZr_&CO`P0{vF_x%_orw-63X?r zFbW&%Ee2u>QB*d*YYb?7f7U!tN`x2CP8J>z!IRF2m*>CW43P_pt%B#nZLX>!IIoKJ z%TX+KvRv^m{7l%i^}*Xk$bDytZD5@LCh+wQ!F*d_#Qw@@VNG>iInnzG24c;6uk#*m zWH^`O7)r~gL?cdc3gRzt{qMEjul^)lDGd~nI%5jfruG}7KQ`xLkWJti2$nN93vG^9 z8?-k_@OSe9!Zy6!4VU5u&F;1=W|5ZrD_I=wCN#FcS1Hbj6ZaLb5YHYx6U&0%{)Yj z9MlHK>=<-W0lw7UGp@7c(;s8Qp0zsZ5JELugL+Aw@)&K{RrM+-mRI)~W24MECPU{w)eOZ4$2S9|xsXw>Jz;F$z2#v8{Z;ns2^CcDu}v z@ZFb8;@f97750kk&-ysvI0?5p-PQ*&b#;Beg-<#{bS{ZI|ASL;@z-YK^c~obz`7a? zdnY@S-evVooglwg9Owh_1rd{W7&}D*tVQV7lFL$XEz*^CGqZEG>ya^lJ{QQPLAc80 zy1-l+AL~48cQ&^P`zL*eZ0i#9&1@swW~4d7?evFgzMxQ}g)p0>WcrfE{?&Z!JY;wj z|M}=`hXK5|V7QQotAFs?3hFHZ}Nm?mHfOstU0) zZMsNSuJ)(I0ppXnD$*^xvQ^VD*a;!j>ST3mrn|*tFTNlUHFcBCFfQZkS2KH?3<%HA zcjJ0{JX)$18T$LKCZC=xz?5^C3rxO@$*{@)u`cje^7?g(;%U=%ExoZNJLEe%qw-R1 zru&HHIobFctL()_AU5hRQWS|4dkI6=lM1&6vI=fATXPYX$+12E*+E~vl%^T{j-Vge zjI;TRt_@^+4-i}9p4a@y!6C2OuGPH6pZpNWyvN18GX-)*pM-J~hHT zOHYmSnlK80G+Q%9>cDby6}bdoWPw^mx=3g9P`XQq3a>kWEua`Fj6!}Chd8WeXJehn z#dSCi3l2+AVy^QzUSd(}A2D=0T0sL-aR`r#Hvjq6lEU*vFauf-CB6X)lF`{Oqi3;U^QN6_hO_t2(4eVvsbT`CZFFw2S9(bD91Q>t4qd(5IAm(rbZG4?u@+iOC1=d<| z5BaFh6+&$`4)*OIeWS)dDt#Z~ZQjE;RHWi~LR&;J0F|RoAC}hGN*Ol)%$YbTE}m-4 zpL9NLHhOMXT?F1VJuX^ion<}*95DYZXdi?EBD-kXNo7$|?=$4VG4+DHkCXkZD`9%t z8(m&X_+68wB;x51LFjty*~z_cdHZEsH?lEk+P@Dx$ab4g7uJTD4*6KE%hBtfEgp{N zi@J>nLLnmUtjrto=HI*zr1arl#Gcp&MGV zT&6g0?!PXgyjBdUdQ)tJdT^ieXUAuxhYeyj(@3g@hDu!-=90J<1m6pQKpvFp=r7-> zFgQ!2$M|eQ@1X5(eVq*=3ou5$Zr9tEe*ILw9+FwQg2+1j3TTl zjro^C1tH+-6(%N|!h9`G8-*g3%l>7Fh~G_G{NCj@lg|x(1on(r^yq3xtZ6L4c&s5e z`?sLx)3buVgHpJJ#m%r6I|j@f&8;>PuRudy zqGF3FHv&C9$2lKX`yAE~a(Xx-b1XpiowZ++T_7j&hp5JpDS^GtL5~ zpuE8^+**j`n{%0)yv}>UZrT0e;l=;EXP3v9MN-@=%Ko}UIB0{1$H?O&NnV))CI~_7 z5d#*4F$b4Qwt``n$)B{_(H55@Y9k0nRq~bnv&BZwa=VMO^5Y7AE*ng90hjBirjZr0 zW+!ionIpsfm4Duq;t))x=-|h9aU=CK9);tim*?t5x}XNp@36F!Xg=Co6lPx6LaGwa zd?h_EFIG5OK6916f5p@voPt#}gr-kfrUrvz3&I#S1Pk~SHAUHmy2=<-dWOol;^#sz zq-*O_-1k?eEXh$Zu)VWL?V?@?7fG;)?}5}*E#8=&N5<;d9pr*82`+qM0h%-Q_-|qC z*Av5k$pYAsLfyAts_if+Io0_QTDyjhX|eKMSoqw=9j2l@ON|*2>6xB*cXa2kspi1{ z6)@|2Y(I>~XbgP0A9eFvHGncCyC=q zViGaQrgXfH^vr*A51CWO(C`N@Ecly_C9mbO(^Hw2M!~vqm`}HpYoDBOa28(@Pm5Wv zCMV+5$A-qPg0!;u%D(I2GT;MLvR*`?G%!$%9rRSH;rl5pGDiFYOvs8c&5f<%lj1qzK1h= z9%rs2{#BY%hX8Lh!>%Sh9X1qcS`{B}27U{M90@Jhe%aqrSVUTHe^Zd3OS`)NE|a$+dm3oh0LO`{px8#K_rMmZxj3@t>obA3~Zjhe?W z5r$y)E`$*IL;vn%Qf6`YJ?!bU%lwI8^eeB<0Tg<^GtXeWdxol|=>Dg6~%6|Ty5)OJx6n0%4eW27_lCM@|F;S9F zn_l*MwX@)mAXwaLs1K1*uJw?G|;xtC?gs$C@(%J1xg>H;p?2+#tpL$V;sgG61_NRutLK^Ar%g*MUsjDS3 zKOuI5n2d2%rT+hR=?dn9*WyvR%wfSLovFN0fs9Xjx~wqgC|irTqNqR^&YM@-jDJ=V zmtU=ij%qenb84BC`9N2AJ>QoJbs{Ss5O9vK762iGMzXam3Q<;ZKqnjfZ@As^JE1j_ z7S1CLS)z!oZd}qFI-1(AV_#l~7JgG$KvdlC_^!+nSfmgc&bB|s8R?y=@hhenF<^n1 zi90Ka;EwypA|)JnN=37L$p+eQ)zl`Eb+fF??ME>c&!1K=1LhU1kdz5mHF_VJ4~%+h z!g&?zsKUXN8_z(kp@krp@u4^D?8vt4^m1u=0DQ@8U=FxV%%VB9(z9HaW3jT9%OT+7 zb3R(vSgpS<;I#wD|Bu+jIJJ9MGV4hfMj_5#+~a){!L?Azm(8rivNM6XVgHt6+o@Ee zjwF0noY}E>Nj{MF*0-tgf3)mwoInd55V=S<7yme&tQh~h1x61Ydv{4l(^?IJ{~pmd zg7}9Q=`@0no}ie4AALCN>M#K(+=9#C|2*_ngB~!q%?qyGdECupR+|87b};abHTgIywj#|C`Qm$`o4RHmPdUGfAN!;`22BT&HAug z)_~>IIwA`aj(4XQNTdu9p4H97Du>nzW)h1UUW+HuXts-a?G-kkJyVzS^XW`oY057K z4eaf*B{Ya7BN?rNGMz%+_0V$>K>5Qn|K;z^#?Lys=_}#4S8;$7wuHNy_>2)d;`!%& zSI5w;HXanKU2-mS1hG$4$1O1w#lc_iGvMNgm5aN{7(F$zsmH=;wL@5%E(b$OUy}t1==)Z<|G^To+Mtl0$k@r+L)d%P)7F zJ=RD*6wKHJUGI++D8?x0wO9|F2z?6{OxApc-Brgbg8YmFm(SYCyhG^}hP-neiFRi$ z)MB1TWWA7uTP1f? z*l5}P<59X`ueqH>3B?k7WF7R{y?NuQ()5t}bDY9$mkno%&cdgSj!(aK)Cb(Y0@N^) zO}<6X)$+VAXc=M#7uZ5EZV^D0?%AW{*XV{1wE?)3Ki9DQ2fH@&J`sLnHczQWhz)_8WZvh>5!fGFIjfN?w= zMZNj?{dLH-qg$U(;(f)!XXFHb9?$=!o68JYtc~f~*X1@u_7b~(^s2G6j)6g8ox-YU zj+)_c)(Aed*RS=_&)awzKc^&&UV!W(wWijfU9G4vD1TiermG^!&4G5B%4aV)NS=38 z_mdx^*;qD#=`Z5m)sfAXt(0e~pF{JDhuOZm-gqQHU>vJrDKrM2S}vUF-fdi#^2jF? zdXNA-+yMH%{h#RvVFRQ$9+G)REXYPJun8P)y>K37XRIxUad!t;fEH{mDOTTpBc-(W z)}z-9q>gEL#@0_$TTN1J^7z+vwK26Co;x^GRSepO=xMV-@8SSK!oSw6z*%;aI{~Pp!)>cED!2Jyn~>3jwkGA&E%-p3_bPvxdDOByP8u)&jbBLZdAM!t zh=(_()^#!WXzt?Ci~|K`#E;lq0m@n8wi0%lkC0w=Z5`A^62N%d;JfO5Atf!9wJ4Y6 zdMkzP+_}R2p#b3|{Fa5c%Ex}GrfYpuZ=?~?HL{Wczn-2MMzp-nRGin<4*)2WxUk7x zTm|kREx`SVlT)hCy=(iWms3ubrXA)A{k*?Ph1F8nO)mbtopmb#RnahBW;0^lY^HSG z6Biqt${nN-bm@ZsKV$+~{EfI;q?mh69!WfnerBpx z)Uu|l1pf?7FQ1kD8>RU0^0Q*V+U)1|0}OdEWvG#ZVnLjChUpmO_;L8E#{NXwxrVXi zW9#_p{syo$L@##JlxWOlN*;YQH}m+oVL*7(etlw??*nPNg&XR``kpN>$v1HL@>c?^ zp*v77EAN$0v{N}RLc}X+-j%fm&-@8M0F&qn>|D#7=PZg7Y5|Wf+Ru(94yJ3%cbU%B zK54SX#s=kl!h%NtJ??E-3E{g5b76PQyAt|Z(y~kl=GZRif)c|xsDS6*RC0WNG$E7+ zNxkv8MKVkJd!8O5WVOF%(v4JxvZ*|&(c4CJf5+dQHJ$AVX{`1NgYTin?zV#=-3_xW zlTNQA1w(ejldjm%T0=X%JZ84p5(R+079;cjOUYMFrudBAUWqr>YB;elP!Y=*2UF)} zv8+G(lP%s{nQnV;cAVftSa?6DnEV$$b@_eckEb*mXCsVUdvaq78?iXN^0wcEd5AwmKpx)cj8g}*Ehr)h>!Iim)>nLv+YN z>EZzFc&Vv7-Ss`+vudtDouWN6`oMx~=RYE~SNeUwYp*=?-F1>2hPb}8Y&b#B^34Le z)GGDiEa5v+!+<>x(luu}dkC;BZ*3Y{m=pU|&Hc=re>6RJ6w`8D;3Y7dPfIExt{jy5 zkz;#Z=ll4T9zC)GW>p2iCUlFgq{FM%_svE8}V%UV~5pX=b*!|J`VSqF@~{Q*Su-e zDq~*>iuVyZ%aEa>Pd@J!f`I6q^DQ+q@B}-`dX`GCl0)& zpqr#uG+)PXq|=6o?pDTuzTSkrC>gjG9IrdNH2#@CpP?V5P&Gz6U|@8pgt9El+ElOa z{PdJyeTXo zZW7o$2<9OesjyG?v<<>FkOfS&&))xAN2uS{C9~&(8uJy zj-Qk*&i6~(o=JN)OeO?}9KZ3amLJK6R*cx|5X1=OO48byj-)rkW*a=rYF9r#-{5QX zrc;`(>^dJ@6|?X;L>VgZdlh>0vLtC2g%m3!!IO&fLyMVl!~I2n1){%n-G247s!=tD zjkjL=+1gWKwcXfnMK6KxhWVBZ*WSk4lQK_;3i5;Lnfy|BY)_67?>Gjx@gGt z*(!Ygc`PfBNlZi}M+TAmdL{kxhi~MULE0}A|6XVW^}y?fSDDYberywSzjK->9{^NP zao-MDSAwd%xL%{{zfL2Mpi;NDNL)5={46g;7*q>EiY3qQ^5yv53R!r#NUZQwim8F0 zu{(D-nQ1F@=XHtcSoU$jn}APzWm(D=<0Mc`>2gxv`pw{Cn{Q)Hq3c=W)(jMIve5mv z;C^WY#KYOTZx)PImC2~=nTAnT1q%9ky!D;tO~PN)pF?{Q0uI~C8&Wpf z;ulqte4tzIbN}J%E5o8{yS4>U326}}q(MLdk?!u6hM}c925CW1LOO;N0cjYzW@zc| zM!LJDzm4~EqrUI+e*eb9V|4bl*R|qY=QQknAqH#*0^r#{?Ie{(nP^{_y!Q>O&3 zo=dmsZa1jJea)p_;jUsCBR3m=C4*MaZ>$QaQ4Ej{xvMxFPXT}+{rI$1<5PXC043Dl z4(Jk~cROl4Z)#P(^HE=|>jNv0hw~z$q-0gwB!2093EKERbP+8ljDFXKxkJUMN(Bk9 zw~r&ij=9z*E8N^{2UDl#i}(M{!`}-S19IlMv%MXk-A1xTD_ZiU;d$H2wa71#mHo@A zBs;Te-od2emIl*0x_DF!&F2ZCAA_^UtTfHaBCSE6fGSxb_s7%Eukv4d&AbNm75I1# zo#T2v&)zI2dz@f)W$M&UaVkQ7WO*_i?u4LHlx2|?4%fb#*6FoAAo(-7`{S1KJn}xb z-yA69=BuFol8T#ouk0>28A=(4p`Ht@A}QJa|y%%kHm6C-k-=6=rW1m zH*Wu`Y|xtpxRBXu{tJj(`skG3{y0$JV=nfS0Dp@pImg@(kg!jx8U;+(MARVvolFJ; z2w7_ zPVl|&X7V0{JC4q0d9OQvVYB}n63}k=iueo_8=Da2A%z< zA3t(BFQb)z;G8>|hAPL;OIs*NpRohnkGnTQ`XpAWYEunYoYrYG&`*3F5PDP`q3&=1 zl%+hE2UWY$twGK6R@igY(8zSQyh7=pAZ^gNb=ggB>@MGJ7Xy>j{mTK}I(%^A`<^)S z_h0A%WEHmqTKe1qSou3KY!T^Dv3+&Z%J}_$ZIIQSp^kSBB@mZ0x%D^E@vR*Jc{V+M zT4w7m&Q}nc{egvWWJ{$;yqi~If&C;S3dE)R%|Y#+{5Wf|gT^s`q{SMy#S`iu+L4BN zmT2DLmG9rrPVvAhTr7RHs;EbIrr+SR z4Q#TsGU)o3ZHqRN$~^G_v_yOh0+ALvt)fK)VvsMOd@Nmus zH+T({L_`v|kEcg!$!cy)thr6j(Ok(pdEwoM3&6W89w2&53P3@ir)ms-#HD zlnN@w!FxRAKjHtujJ#i!d6s%Qin@JY)Obzci=#*Rt=IVF-(8R>nYgIti&LRPwJtAh zExk~_HEVKZI0OW%%tT9ckqPrZgdd-jG$sk_r6C=9%&)n+3qmqCia7(CAj~di;{aAb z6=yl88Ve>&n)>eEKl+V;9#g8`td`_3Rr}%XOWAylurUVY{!5~T-{zyN>^I}7H&HM9 zXX7zB!}72^A8P>3$j@}6)%CoI(D}UL6<7vT`-33AU3!AtXlaw8)Q)@*%sqQyh^0t( zUW5(E$sb~xi-D#~vkIXv1z#IIjtlsF_tpUZu!TE_CjtVI&MNdO6jh;C8vyb@<>+P+ z1pk*U#Pe0;e6|(qFvAx&C;CEIV1`EDRgMzrm868 zHu61!SjF=%$@Td8e!PGwlsDE=huabUWikp-I< zV`i=ZJ=rFXzM|&Xuz6W9mrGS@2x$FK^u`uAo(-8ticM@04mK27G=BKy>2CjzfA0qn zd~CrvA)vj*0#vvd!ETkptP{_n+TcM?axC4D%i!^)d$(EDfC)_+b>8aHhA`0(V+ z%%C8Ba>>_&Y&I0t;}s-hYZv68EH<~sbbJr<1Za&;-9<=}M!iI#suI&r{M=Uisrocm znUx#UFJ+LIC+dMUb0UR6Yte=`Ned!Gc&zLWWzN5raK~E~mxGs&ja9xYO+(wGB1Kkr zhi_tV);o2DCj&9V<+!HvnCwybm@UsM7yKV8bd4;T3 zU&SN<8f1}5`guehA!ivvik;Kb8ySe;L>Rx4h<~l(<{kfIAebVv0jO~Mz24mf+f9Ou zREvwRbs81v4$+?WMUI3eJ%+dS<*$(OEc@=BIsqJ(c>CMO>mPD@YAfd2uXcn zLBBcR+}A%xvQWW$RcO0`HOA!as~@?X66%0WH3e5zj%Of%i;wFdp2>VKeeU7T?b_A+ z)_R9{Tc>y7>C!ZueH;1KbsEJ8q^Dwl!hec$Yw%}LC4OO^+*584W$O?kL?=Y^6C&^2^6BQ0D$|03YQ9v0?*aQ&Muv~3V9G;&TBF9#BJbz0 ztUF=Ls9JsC`e|9N=?`c708#zlCvx+gk0}-CqPzC?=0v7gk?$|>pJ_4^%cF&B6!s5Q zp>cv;yt!QQh~STnWPud+U&BH`{)S?za9CPzYPZRJ_6gDZ9`esdhK;D#s%m3Lz7{ds<3cBYeh<7}cQx+hwKId9=6%CBIAwyHo`$l}2#w6Yg zkYcELeia1vR+TU0TI6f8zD4I{NI0}g50?D_6g=*OI-eJ=Mh5m`Vc@FNcCJKeNXTUMTpH!QegXG_ zV(x&OlFl0s`KM5SV%H;42dVIoPOlTw!K(!E(eL#$C5_FTOW`OI5uVp~!`5Y=@5Mz~ zVvmKaj#<7+W52!gdRB-kc6j!U(6a{bN~h*+kOlkId-O*^fPh>2zJf`PZO?nXe&||# zLvj|KKeF6i?D7&z`}Im%?|W%$#6 z0Gt?DRw7fUiw8zGAxiMP?{*SaFR_So71kA4)e;!kRz&|Xm4e?9^a^tXIGkZX;D$z~ z#yzo6TI*;u6Vx0!od8bs7^lJNaomh&JnOhh+A_whpx^?tk2|RSdnlW)yaOv)@KD2J z>2bi#LSL;K4U$Uj_tZK_v09&9W8 zOOxG24**|L--t9YpO<*B6yY2O=>*{*w`1&pzL_fnSFM;l;$=e`cUVI@YMJ|ka-9Mo8SyT%1|>o9-kDWkZlF6Tqfry>Gi!TA|(?U{)=w!L?)R3R^(n%(t0ZZQaRkD@T<^wtRv;V{lET zT;u;d6)c5OXd?a#2Rpkmd!;#} zT;^_iwH-2uwYVjL^i#I|x}%Y0I{t>EG9fn@_G6l_;m&D8=S=}sEKkGTFsmLfNbvIZ z&yFqCTkhHG5Uo@jYPol6O&@LfhfJF{IX=5J92pt~Y=kh>wDkv0cJBg-K4y$|ZR>2+ zF;=1q*-cDD8A3)wP-ibTj35$2uV>1nU+ZR7Zp_aQIP&VN-PIqh;Ss}=8ZnQSoCNOF z#s5i10beB;$g>z9LZLoP0|Yo6l$54$|e97%hLGaa7X{-~R$x4pqj znuZ+S8%5IeStOxMqwBN{pB0f%Ju;Qo=N&jRARhE2i(Mx zhu@6lVRKPLaT(I#Or-6{g{;d8H@N~}Np)Q{Gt807elsvBeabUt1Kac) zFxWFvgr+}Ar|fkwc_ZvJFFa^op~mTv;^0HdWHX~1>;2-to{Js?IWg%Awa)>Ak*Dn< zRf*=Zjxajqk<>(RCpU{m_fka49Z&F2Nvgjimp2@x3s6dNK3Y6Ef)jpgsl}EED$9$t zNDQFR3=PP-Pr-`x8MF9?5T_@Jc%`g7LDpwhO)VQTBMq~KRf}|lJ!nToc@D-5H zZ}$knr*JFlCWbYY04d}0F4;^&MC?4XY?a*kA+X+I?ehdGJz{tct0Z}3eEl%I*t#t(3(AVj!v}>ocYhI0IEl57o_f|%fV@{hbmu1S=o;22Y=7A65Uu` z&qE7`;p#7?&SR!+PPKZqd&32m(ZVx}_y1z0-zI<0yaB?JU02`=%Z#p$6f-^gidOA6 zQH?6CaQ2GHLk25M45|p4=x?XTlAC|u&J3)=;zZi>w-6-J(552#t7ZNg0=JWd_aSQ+ zdtW_ureb9c35V>)@#r(utyWj#^=@c`^}yb2$$2O80)i7*e?G=De{+fK|74v3c6wDgJ}^2aGN1opgH(fZHdz zmw@!6{1=8|x%ge8T1=FRMM=xm+EnZQ?8>nn^Iu03gGG{_I z@{y!w9en|md}X+tV5?p=H9V_msUp3C$t`MM&?j=99Mv2PrTIe*B=S*ctc|oboNNE# z^B`PP)=4o@bx)vW%*Ow-KeYS$^7<{JG~{dXU;hOOgv zfbpa-L07U=#t z*CB-biHO}n1cb1`Yss{4jYM@Dq9`_f1b|WnbS2)ly#54@&T)EF)pXicv|{CF;4ik4 zKe9->PpXsb^#U3k)sF9LrBJD{s17o`g{k}exm4$++^)XhN%ro`5`-o=W@%`~c1eSG8{HRF|qs+_VvB2oQ(z9=O z8;dqPNyKY(UbDK~hh`lpXMQN`+*3Cie`FWU)h$bpuLfD>zuGP6O7^ec_#WzMjsMFP|EHo^#tQptN6%@~C6*}>QLJV+E_iON z?2)O{Jj}XW#tRFjTwB%b+93};=jVE_=PqQqeLVa0Os&?E?=l-xX51l1TK5ogy-N?5 zF4KB_&DDwEe1WfYeNwwu=J?w3Z<5x3Znd-k?&*to+vTD#y#0IJ&g;6U5iLhiIet^l5|2jJhh+ zeN>#E{esuc5NjIRN+bn#d#cA*Sgskr^BUr2VE=9=C ze430-<*Pk#yBUA9qOT>UQ&*?H5Z9}Ey0A2n!S>sS)jz}PBSpGohWapUjFkzsjI(5r zK8y?j`87_F`W>EXjH$L}*=JjF)x5KpJ9iq(N&uKmehk}|Bpu}Oxj~iEqGFc~^;e3c z44`Bho(T<7=^yGqx4xq@I&y*HV2RPQVRfbx2v+pi9<0-3aQBm|M)zy}t9l-{{U-Oz zT-{g6uBQe#i9=h*UCqPMRn^rI?Wt`DT?r4QN*CCs_T(xSz(s%ZLT&?^mgMZuT?L$} z%5e-sC^Uhd6D+!YM^WtUz3`qwZuAf@5HAz~U?&dt=#l(>Sm4`CN$=mibZOVdb#mkf z2+Wk0GGwG5iu#{FBV1>)V^)qjs&)<}hQE4u;9vMp7QI?0f9MhE%xz!%tPZltdLf&g zL9X1vhWmty2)e4#*qU=(Kn$VdT#Hf_ZG$1MyL;mIcFwA_cNA9cB9pn zMb6W2Ijp}A(_5w&<{I1-@$e~-NT(P;BgXRuNzx_Em=86VNjEvsEe{Ofq(m4~iH$k2 zZy1+Y#T#PBjHtev4_cA$Zr>|9y)fao;yWAWj|YNSXvVlX6#=)`Nlzj$VtRfG4S1*c z+Sr)|Sa@&LmW(V_T7UnqGx!f+sGj(qtDN2F2l>WQ`&xjD$Cc_-2_|WDh@QpzxbtTU=sor(C5cr!Rl*QQplmj1AR=Kk(stI{kM1KHDxhv<8Ofjy zuuSIROdKXU|AjY;e4Qkb+TXLtCZtmBS=k3OY_&&HyfOcjp`^rK@Ej^dczBozGebR1 zP4X%xyfE27E_G8xv>x4a=L1#{uBu8dqdIwpM-U^{_HlF9q|*EmthIzmU&08G?&1cP z^l~)^eAgyiv#7k(&`}jtR>o<~a^&Q(Z2MXj|D*f(BMN}4-9M=us0BD}{-2*a{$R4-m2U@`^hq9$dgE_3Hah#g*YRaMp zEUL4G{>S9nf8N0u7}3Opu0C*u|3^@X%VK@m&m^-JaNWD^Q#kHMs9if&$edx@Pi;$P665Iw?!K%Sy^9{b;=0s*@ zcg<~x^>!ouI!`r|M33{9>MnXrjU{O#6Z^a8fm&*0yb;>v4dPh^P;G$cPyShvoucCs zM*#s)zpEN}#tp8riWx*HZ33nW0ilhZ8Nqzx5RuR7k4*ULR%>mpf3SDs^IvlwbcXUV z*Il)S%EBEM)Z$8@kpoxUmKkNVi6<#0$qyKDX-=D`i zQJ<6Pp-kbZwe8G-+o;8lD)xNXTL!-o37>rDytsi5^CU@6x!Ri>8O+$p4*M0JF-nJh z!M`s;Jt<;&n60fNuu6cLi3qHJG_40Qh@II$fpu>9`K^r3U z1gSw&_ky)v@A=&a5J52EG%V_!N|V-G{RN0v4H<(H>(d7trm^NCNI!x7(e?RFZfBn`u&$@GujM%F;ENoy=UJ`~%J{_kWMdmpCK<1=wMt*BvuC-#i5@)~P_ZDwFcPenPoek1DxkOHfL*<&|CadxoAhi^{ zohd`pSKJ6aUerq?Yh*ZeP*dC+Il0=^TJF?6^8pguVnng&%pNCC#moY)lS}>tWcKM- zNr~Vh2?EDM@#_n><)6EK#Kt$J&p!#(pT8~Uea|Q54!pq~%qAuaqv-*+=`WOibn?r> z+tpPP(X81S-cPdC6WU)@L;&ro55S=7-|&4s@lCBS7Eub8&^shX;(Sr}>HIs_rPWlN zBH9Hnv z^C`Yp2eOL&J0R4vVLyM2My=~#r;6dDK;)v@JBIyHpZtxY;ACAiIpw);X>YU7A;{;2 zwqlW5m9l-pi=8+D;i(LaCbI76at=kOn*lXuQBtgmnhj$Ia^Jz;?Qs9xsD;!e<;1Gm zpQpWXz|05W3e5rQ2ITcriyB!!XOD+H|U=do5%WTw-_ zuHX{O;fOQaohk-Z#7K)X;ia~@vuND958~uKq?UQ{;nExBOU;@ufHkL1N(h`mpkU3v{NQ_e0ik4* z$(XDp$wut&{E<$`B1kM?z7QWp{bn9O1>W1`t^nyxG*FLzU+czG79)pS)g;Wh=hM5< z*~XaN%`^*ITo^AuVtQM(Op`h4CW%&8zI6pIsuStjnt!wO<@P$%-^FZCsvFh)xoNVQEV{lUlyp4o=W_0KVUux9bDi1V-sQ1U zXe@V?aNXd3ta}*>w{@Gkd;7|M2UQ0EZUTnLD%0Qqf6o-7bQatV%9)JOwZ8Y`=Y!S_ zvpuz}wYkV{8Q|#tI=K3mh-V&CLHZIzz!Dtq55a3lj?b69d|41%)@U!Bcqaqv=fj>( zi{2nAg*ZU#^yjO0sY_CnPo-#x_A`Sw;mEDf=t`3%N^bmBC+fK|s&jyVW6NAj(EH5f zo3P{E-W1w4>W@+&)_8Sx%`*4 zw6(->jogOoB+fFIR$Xqd)l&XjQ?*#)riZw1x1ST#NsqD~v?8#xe(+L%6!9SaDly-R zp9%xM5&u85_F9mX!(Tr#fzrOEw0=TC1kVW5SMz!Q@cQFbtV}cB0$J+LkMxyQ3A=zf zgL!7l8xJahEAeR;FEK}%?nk`Neux=28DkntPyWQQsK=SS+$)lE!L>S8vKd|G6CPlm zXgj=*?Dpapf{=i$n)V$>q9drJH=op(YQn86hZWl%;M1)&=>rVb z`B_mHbwnp@YAS8GwY5yLv-Q$G;PmPFcUiAF7#6qX?g@j=Fi>%dNUFVsgkzSx} z><>g?jyq(dUr*u#!R;YfqYR%+a%0Y33^u-GXI!&3N@}`b8Q$Xyeuq=Q??gwt6KJ5>FT`tJtk~Mp++o-a7 zb!djcct{cjK;T(m5Aqu;lHO@>KAZ?~o4C{=-<-HK=_39k!KG4oq_38O<&{tL@IKlD zN|_CH%`uhdIfkPkebhls*9(vJK9z9GYAz-&a~2ukMLqC(e)J(eJJ2DXe%gBG@I0J8 zda`vuy#lIwZtNpOyi9O(W|R2QY9MM2PZDIZ`|mmM%Aof!`*fc)n#`Ng z4)*$X1XWTjT!o6|l+2NVlz74dGa(FqO0iZ?4si444ee!wB8+J>0us7|wHj2NcV{?3 zd?0t?L{UJm(^YhQH*4_tx$+B+X33gQM*0{KNi#p$GD0q1w(&f=Zq$589quXmo=aE*btCDa3QAmrY|lg>L` zZJ33zgC6LJhe10ROFJi{PCcqC&ijVg>Nb7sN!Dt`O*E7Hz52g(Ic|JPtdKynRtYm3 z@Jf;OM<1DQ?jx&@5_vtRo@15kN>5?CdLqB^arHTmYaU<~8(6!ByDY`MRB9)qG+P^T zb#3RS4cYY9FGgMq>BZt$n?20X^#L4^`!fx7@v&&XOaYQx`>l_`ErfzjSAoIGpXOqp zRPG`T0Q7x^%^a1S@IvXBs82Ls5nik(s^XY%zWku42h5hai|v>I?A!H;YF7vA4@S?f ztkvKKCBwxbAchTx8n9jn24=Gv1Ls%We5W3JU|L^AqWhf_^hW}rF9-wsfAcQu<#0KG zGyyLzbHH*_^Cod9qnyw%dNw?==W_?mQ)-EGh{YvTx-ET}rKp#DG+~hPcC6^eFbh8+ zEp}F-R%YN~djwxBh77J3+U%gL+V)&$bC6|)1y`5sC@G!U9kh74VPf6^?p_BH3&wL9 zb#S)U?9Y3Vo+n+EQcre;;=QR4;2(xgaD!8Vf;`!b-75msWLeH zJt|yV46~-r8)ki4q9i@#II$=F0FNPe6Q|oKV?71g=nKDxkKrfon&eH9PL29?sTkllKrIrqNdbr8diPvjBPzbk-rjRvE`5!Zm6cD@MMZbw$7=l zSH05tHctGf{C;1rgZ=LGBD*WCg}=Wp21zieEA7nSg>IG;e)V@@yG?tu+4N#m_=QZj zb5gI}U*8r$Hr@O{a!m@YteWEBAZTUaBS*sm(0v0t-}TBVby;zLzjv{wqM+|lSccxx z3OD1A^;U@c6MNUE1E{rOj^z6TclUJ&PRZYlza>W6p|va5iLkFj`uOEJ=ve_R-uxJ6 z4YPyqNvBuzNKL`r+{7NvKd(g?;}xKmTEdmZK`#^f5FZK8e-c8{Rk#%b0^ogPzU%Sc z*bkakf{4nKEZ*j2=5$N@-^QUF;G?q-?Z?!BJPm={`F3NeO}XsY@E1MPtU~E41vY6? z#>kzsR>nEmN#CuQx*(^&p7);&6C7I)kb;#Oq3megX=e`koyc$?if~YVrET{RqwksA z-Zh*97z?+r_on}7x3vEu4NwK#4==Q!3<2G54dVW3r;hTIK|3cYc9q7c8DDdiu_HS< z8sTQYw%%LBE4Vn}p5M27D(COsMoo+eU*N7FOk}8>wl*cp4?*$KnwyA$N~S(MAo$&5 zcHHR{+&@lcQsS$+49q-#VT?5jd#)`Zo$lLIeR(*hho?583bs>pr`-rPp6=0kg7c=5 zph3kX>kcDhh1eMDlZP@T<1EB$6)`?WpZRbgRc8nxpoFWTW#zHx9-|@r?ddIS1nYQ| ztv~Uuf6nAMxW03xck3buzM}?ENu*M9xr6Zi79LzoJN?j43rEHRC`d*NHE0BXtuzZG z*K5`i1NO6JX{q&~g{ay=XfEuB)rIP@`^b{;?^CpJ6vay4)A>?Vf?O|#^Ep}O$@;{N za+iSKQ^YfBlp4REgyc! z1AEz=`xRQ)knpnqRtIJS7=`eC9jQ=KWJ<$N>_t^g>3Ne7IrosxasakD*E@ZOL^m8r z3jsvLz5sS3+E==k>Qe+oZuz)>S_89=SLFsbOdOW64eOK)%N>@GHSh&#)a>l5q+=V+W=!En@`iJ0wNYx0fg&d!p&C!==8`Got!oKj7n6 z{S%ki5`g-WLL$WE8E44)ce!VEb)J3u^EblwJ%RCmnAvOON7Csc8RHG4m2X$L7lyRp ztit&u+6oHng@+RRBO06U>n#xpm=?IfN3PRufVwKU8AV3KP*ZaYP)1nq+Ex||$@(dC zzTTla!IL#iIj3kwj%3lH-3Jf5)N}A+qCCUmm~YyL+jayF_TWOLjuog+wCpjPUi&w zY5;u(c=tFM?<2y-l4$8`+d8+u5IIH;Gg;W+<`57FTleS; zxTx#?fG%Zysp;`TLc(UyTOgi(0@S11NjiA3VyQQAOujejl{qhQ@m8OnXH=be{fM8X zJ9ATT)npU*I&2!Mb19W8NdyK4{ynaLy;D6Rk`7Q=?hYnDg{&_fRRN2*h`i$tI%@za z7cWhF=^jXq1q2nQ&72>@Q8{X^vJAq1r>gwNqr87l1t9#s(@KhocG3$rt4hKmpfvqi zi<&zRypa&Phr|0++Z4SlGc{4Wo%1;P^Mrf8U;8GmG7lgeGAN7Ts>g2$0Ug)i zl0*>31Ro|wRs&`3#E)JfApPvZmsK=)mDa{)=!0j-T>+6eb^}D7kCU-^e|JRUXcUxM zp+brMg3TJ=xr@8x;IzE0T;O0D(V$8v+3!uYR4A*ySuk*Ry>t+Uwp+~?6t>N%Z*qDy zp1nG0RH;zoH_W7+Gk$q701;iAUCWqh84#f9>YtnQeWQ9_lYgmH0o67y=VR>A@Dnf2 zKf&=pQ%b6p?i#bAH$MAY`tjy;+jzWbGBcOIY9@ScE}#n| z^XdTHt8K+XKt3P%9Z_>D1}TYpGSP+k!Jh_s$6LZ|fv9S+VkHqtj7C}e>@t4Vl#kW{ zSzRA;qE%#0(=vhT7Bl@U*=D2X2}w^C(y#1giu~(q*_4y*{z5Fga4{-p`Zal1W2wl8 zR6Wj_#Z#z=!tznj>mK!1XyfADLi>TM{L0gpTSga-<%joI|Lle>;NFCA#dTB(CX5AIxN52#OXVfy%4$V1 zMiyYvBp5;PUEo%KN7{6kd{ReKcAEHYxd(nYMOgb|YWeW!MQX8RILUqEd!e%Hh5vy3JX7lR-_Wn^PdB*<2gdvFXGlk4mA1?xy&d&PwBzT~a!Nu9HVLxCxjed8gmGbLruaxs` z=52qcO&q(?Z)08w?`moMiPXjSJ%22Ih^!4S{G7-EK5b}6F5=9z!ENe0j<&*G8dT7n2l`&l# zr@hn11t!~!zcW57RWQgleSEa__!H;gW0Q)J;pQ#FlS*+>P?|udU21Cb14I}R`rlad zO{wL5PP4n;d?1pVr?#>Z;426W>uzgSz1@U;OM!Krn)ZFAkKZ&g&2OH5_V;&zaJSDL ze|@Rp3#xM0xN6$N#^EAHt21VA3g@LJO-}VigE+0yH7&3ZmozhQg&X2SL2eh`b0cJh zWQfPSIG;X>kPo!yir>qK4Bo?woOE4Bo%14G=j{e2A4D2yqiuC+zyI7zI&)VG*wHS~ zP2>%|(+G8OtU{AI+lEcmJ)#pq?Q>ogH725wv` zyK4ilP2LQ4E0oY~GhT^{d>vNh&b`#eyA>f0O@)R-RLoeyl0GaoSDp+B$Nw$;cjMeP zdAy4$J6wUObpyGhqlAO(hDYmfxzIyasDNN<&dOD@zm_lraIPihIg zk5d2&X>hLY+Zj1cb$uqt8~gE2Wi; zNai=0VrS&!RsgaEz#T}k9yLu^8^F<5dklid4-~{0nSe#_x)Vgy#M$#B`N*|C zazEc<5bnv*3Qa+-{8+Qt^qK1&y2{iqu=J0+ar1Gu;l>%hz?Ru-PpEhk`8FL?t}?dN zdOba?FWcc{tB~Lihh2VIlO!i^0y&5OE#vJgpgER4({d1@zybiQMc4teW^IC`k^)`< zvS#!~pC6AF2^DZGwq%%~OSWO(H2{z@^*Dj~X^I`<6^yI`(h5GaDL2L{k6cQy^GdP1 z2;%;W%f+(^Ts$Yd>P7i|Du6Lv`*Q>PyfghJ!cOiTkv}6@Aprt9b>X(=LLNo~kZeaZ zdW&jeED?1JY)Aq9WZ?#TSwyOKA8p`sr`!}koo%op8_9tv0tq)#^S2pU6@%t|m1Z24HdoEXMHstM`mACww1ODLR91h48GXLFBck<^)NiROL4Z-lxb$ff;} zsQ&ZO`zeqi)@_X~^#;QRKW6LIR-=<9;g~JR#TE~JCE!EV%qbUQ-~qM z3Xob*$MVQsv|q3qmxyE6P|TjGtqo#w+z+tqhw;Em3DpFbA04c7tw%c&?_lMI3IJ2A z6ap&|x4&EuA0>+NAh%>=_%o(pnB$D*;+Hgsc}Ycp@SP%-B#NR;w?08NKKRYQv^ndE z#TzJy?$L%EpoCzOmj{UFMTDQi#-pypgjk3xNuIHF3FenSms)=*sYglPs9(T+l$VHH zYd`%7_d0l}h@NVb0N55QnNoia*@TjUkd+AOT%es(!f_ zm9Vv`*b@jOLY5t)vY+qz;ydP+cJ79k+#rnc=}e~F)^W8R#Y@Sv*D~KM!TBY>6SkWN zj6kCF)L*12iw@I54?HG@Q}P-3E{Ng~YM~&i5@CO893uk_6w%W#(CyJ`lC(l&AKXBk zfk=xWW&#&Rd^75u8h3ZWY>!tih=Z$v^Q(`jSF%+%Q8IZE5p+w_lM#b4>Ca|#jdWd~ z?PJ5*sd#~PKRi!vV6A@-H&5=X`iqX)0wyz3WPoi_qV`e|2)wEO&HHPIq|&A0D#$;P za>zH8Mj6yx$(ts8Eg;59+m7(aa(m{`oi%7ey^k6aZO7*KFsK3udN)_DbHQFhZ~}t~N&3E;UOUOTs5tChZ$? z|4rPvP0x=bZmMbO_ToqNx$G+uD4fqKt_NG|SoyD)#Kl%LL7b09o%l=Oll(R=&l?$5 zj0%q8nfX~X?YF@?=R|*w_SHji3TF~L;5fm$&!d*Ot9IU{G&1U5l;yHi7O0?@&TH^{ zik0DdKL6~ryWjL!lRx^va9G*_n0YN+L26VsAfZ_v$fDjU!&M9vN-{7Y5F-!pp`8>L zNY;G5XX#1nzmXLPzrX?HL-$;Gj=2EkC{IM~-(x$Q$lelTc%i%Z0T}(|=@Ng`8HEj) z4(JO`PGKugwNPLy=tCSf2N<#pRSNjk;q2waz4^K7bn!m^IG`dZ2(Xhd%0r@()g%(7PDIptPr)kNt-t0EEbPT#P=O1hm z>}l=ObswenTvpZ-^NO$Jwza{Z^RE8^er&QLI4$f)kT1&Uh=BGzc<`^Ey#Rv@T(D^j z$1HTFE|d`o?k`x-Ule;2FJ;4x#m z0;$kP$K_9OSV|+qHBmRtoJmB@hI@#x-ZSriAj2Z*dm5MM98^xt!MQV`#8u|QIZ0!t z*|FkM(!R=D34{lS?c>$kykkXLjZk51u;hqw2QLoRo@!5L-j;4QsSn-dfr2Qo$JL#k z$?I&g)-2dhyLBn?6d&LFG=%6ip3UQ(6t1JT(We20R)9@jHj9Yb{O<&{UKm@U6`gGoI;E7JYw^mjpR3vMk)9Ww?e+o zvpowC02Cx?7_>YX7xr?d*B#+Fwz%93*>__+W0L4&P_rl4lT;FBR$V4Sfp+AvJRwvi zXF1ER(lKTw7;KJ7(Al{{mfYdHV9m{lYk&!qDRG*6Cd=^5>z3-x>Q@wTv2LKN)K|^d zIz+i&;MQBi%$vY4%BdzoE40RSf9l0=`}md5w$dm(fAl#JezmqRBgec5$oP+3Tr%>p z0NYn2yP|WlHpT~E$1z4XTtuXMD!dkrs~gVSrF&#iK3rshqf%*P@u$H0`+7d?u$}HJ z=aocC$u8+)KkdYygIulv7_l{&X)x|q24y!)kHJ_&&9+NIed&%*jA!8c5ZD(a18fe| zL-j6tb+PF0HoCeQp_NP&m4rJ7*^YAVXUcz=plnRFfrR7WwXa8^ueQlSa5>q;&!*uT zat;aPmOflVyM3lHGQ?&dAgRTkA;mh$r!t8L7VhAK&Gf$0d82asMr)0e0&uSZ-O1wE z@BtAVgHqv!3d@-_Y`Np~)!k#7W$U>6{*G~3MtuH=h;7EptT$>ko`sr*iDgXQ)m>3b zCe@e+`-Jj*%+6YsHBNLG&Cwg1msO`-GzGG!3*p%p#zs1c*B#Z>r`b<=_4r4sKFKC? z1C294{BGTtZT*#5nOSY=Z*2cRUi|e)WZ_Ej0nZo1M-LOe3_HOwpg|&{lkKc2F!M(;Iy?VWO7jC(?lA)ddP}QZz{&g98`J<5sh28}V;P0qS_l3Vx-{_;+uBCSh44FT;>^Qt|U+=Y{zGQ<(#&KD&4H3a7i z*84TSi_e!1gA?VF+lXk7k*Wi&zeB}n|b!Z5v zliEo)J1x%MQz?sLek{chfM25EMSTY>1d1h-ZU_^&8|ewanGE+e1_ z>Nmf?Gw^&VrXwOARlS+-}v84qJ_mOW7q@M6M6%i&#MfZxpD#V1x{lDd?*`k(Wmh1$iX{ij|K6!DLIOA5TRjYw3S(P$(!P=ZrSH!FY6~#rB z3mV!wUiQ{kd#(IzHeP!b#$y;=kgFvx{n}l|+Q?=o^;s0=Tl=nrErNh7X_}^eL)p3R zCGIz_*DXXWAvh~muEJicBBEcEAow5X)C!S#3brRmICh@ni$|muM<<*O4IFs-xeI=&>lAv4x2SlI z^0Q)J%sFom84lj`w&%XhbZ3i@9|eh??vt2A6F>CP@ka@N88RlyY z5dV4eC$R-V(CMX(`*%PgGFhst3TBx50+XsClo4ytB~_msW*e=_`B$yfj4T5j)-%YLYUWPGkt-Xv9tw8>8YPXe?ReYVgSK+)wy57 zeKFKucOu%SLIFXF)R!tpG9o(AxA}l@`eCwwte)(+5aRx(PhWy*lvba*%;js^a*ST* zMvA-i3?VP>f>O|KUj;xTmH~V*>fWfd(+n!_r+E0b>^%_I6UR7GROYZ-aG%= zlW=<+diN24U7MX~J73*R@D%|LN_CjTgH!9Rwe2RO+il%LnWr`FzQkLgxBTj`HMQSpHhL|ak$l%#u%XT&Y^BSoDFga9_?n+%Zh z_qwdWXCm@{_}TI>0Ayn3x?}qmAn|IP!3+mu;;9-d?l3%#z|b4)(I-KQq52x4`4-!` zYGC82sm2tL4APnSe{8)6G+b@>K8#GXh!ODqW9>6Q6gHDXft{*(Ip5GM2&9r zHlz10AsD?yZ_zv7$@~7EH+lZ+Tk9;#vM}eI`@Z+y*R}V)c7@%npG+tiqoTYR(7w@FC| zWC_3Oqt-{KGu|EG`;=%h`AK~zPjISb=_4SEbmFa_C#f9K>-Hs&=>lC2S^d`1gYeVh z8}{ME5_eWd@(7gwAGK{jhec%O;P65_bz|%J90@p!X3J2M)E${ZgT#i+tTrQD?bssV3=&JE4SD%ehy+|#m#HC`V?!96a}9*A#6@m$0I z(@k#>*6Or#oA0Ut3~rlItjACe{m>Y7XcGS_ofo^wAlVY8GS|a|pC`NiWb&n=m>V!V z>n<@~TnhQ|XWL&YFaWi{?``@J@4u^W@d1QS14X6Mpm-9?`+;DyJYXq6I!2ucvYqRoHqPN1@`ZT*pBx6?>dcd z`x?m7&Uk0d;eS&$fvZ>rcKVu~`m9@rgoGRmqdR<#4bVQH57cJDSE{l9;F;g<-pmxM zpnTxw8xxK!@?*lnr7EYo7n}Katb@ts_~qt&LO{v=YkP!_S<{yk{o*v|@6OZjpHkrd zmOV3*L(_Y@<(7LGsnqDT1HGPiPBy!J>hq5l05ehiRoK&Rh)u0yirdN~ROe=!yL$2n z32I=X3Q#^ox6|LQo%v@_qnQFGQDz2FAgD0^}Er)d3kDSflm#JQY@y+d| zjdAx81%$fIlRTm3Or9*8M*1P|gl-J3lS|4|jQ~m9AU!Rl08tN&wjDO-vp74hM@ZVZ zojTy}?r134%)8stBv?ufo{}{3)&8rg9(d_2Xk2p0lMLVWnYGOXCPMvEosTs_c-t|u zGa~wO<%`n>wtu)PDYIlOpppD3yU8tAi$K(N*WJi`s<={R?xe3NdxNgkZqtHjaGRM` zE3FhZ9YYFp&fsgW_XoOhm*gj{?kTE_d|FWv0hsp`P|$-LT&?#ej7l4D=&l$xpWBuBOV7}t& zEAI771&t3ViUA;r2$Uj%$(xZfU5T3HJQwWe3P{UXYw)qsj?iLeCvhIVIBMr6FdasI z`7!?4Ls(fTN+{T_Z8$0(1?&BPuU5xLHd0RJq2N#P%Fz!dDg9A8SwXITFA5a%#3mgN zU%@71GRN(eT{(9~iAX0)HJnt65D*B>)|-(zxc#u(}gx5F zSJS zY);db9-vR*#ASYQ%Mep%f7md-^{-Vx!P>z_g|Hr`2VxKy+w6%{1v~Gw(}2mP4dDJ$ z4#x4VyqMKS9JbGAEiFXS`|P3qb-8`7BUh9IqXGZ^rX6;VeuwqGSn$$|FgECph-l0E zr~$3=G20q|F?tePKU9f4wjAA39%c0uTyz-^-}B$sOBva;p-C`jDIyexdaB`Mzu6(_ z`;~wb3&!Jm@q!>k@{5vdTXdkaNIdf{iLDn7+ZOWUadEmm8L8!1rSmqAlNptgnGt4v z`j<}+rH(x<|MgZ;ux8_0C8mqStWn;HG>!qS$2bi_|7Cur@Foy;9>ONj#?#?vEGs zVj`}~Svi(`q>?61Ip(lVV{Gsiz(J9f!X5PhKR%T!YUFz<$YQDb;P20!0O&8yGYqM; z&T_ZrnammCK#Sz6vJ{5wLyK42PTXze%9;03lh}5&)=Q_DgrPA@KKv<3+1Y|1?NY1a z`0dBCuUWI-2)fWt;n3y!Q9q|PcGY$EbACwx1@r{_4wjaRI9q;fgXZY5yn4Exk>ksR zvf;}TlU#>Xhwi~)AhBhH|!nVPg0iD zAu2(5%XoBoRLeL!FQ*bk!4 zZt_Fc=W}iI>FIqoX{H+1w<6-tk}SdzwLG{cjq$I&wIWnlsBUzY>PFR(aGh_K$(irTdSJwZ%jb&W`-Xzs zbfw0snnRYpzJT4IGW;)NipotQg61pJkbXP;* z#Dtub1&SzV?$0T*?6^3W_@=kRY8KEK|Jn-BcT&iLXFwU46|~@C`;8RTVrTm{6Lax) zb<$tjZ_&V!G4KlCrbDh+!2R3kO5PY51)>Nh?*1$T?zmYY7;Ql!YjJ|yKgmU`V{&t?K@l#>LhU`SU8 z1s@n28JQ@27 zmCR2^uqDQf{*A%y7JCwDHYYZMN(QT*A4+K&!q{rrDiAz*|#QR%07V~vl*iZUPE z%K@7wd%jzS&2fM-QuydcT`sFmCXv@nFP*LC)^OJ*<_f)1TR)=pW~x!Jm{h%ECsJ%oOBNp9$qfdgSco9vwUAz(0}xb3T5r zFUQkw0i*-zQ?!xvfS{@1qTZP+27bsV+MsqA2Rq2<{m;k7fS!Zd6e_y~vCWWj-~_lb zldkjNHClezbv<{$i~PADh6dfOMC(mJknnApw*AH#2Co5N;oyr+qN2$I#oNo%`8`hC zA|mW2Z&#Gjt7=Mk>iiVwJ5m??ILhu0ierY9!+7IGh8r5PrPpuR*Ck@JIq=FCZgGcx#yD)?V3qqyh?`T?GIx3bx@s7$$(%D77Z z;GOVa5B=Y-t+<5W)NA48Dh$FR+|yiFVXmmixo-I#ZVoF4R8(SH-D={Ety8-<*uI;$ z>SGqO2Ye6I2h@HW#$TFf1@c7@JQL9&=`{mNPYdweZWv8kTn`U94X@gWo!{yaQ%5OmuL{yI+W^u=*0|OslKnn=|GffF2wrGLv2{nTkV^h43U-oAgp<-G zMp}@Bf~lW(?W)jx3j;=rn>?BBl?bh8i+qeE^<6ZC)WOsd9A3k6SwTG=!M>Z-uFm?A z+u*Y+yuE3NU!fy7*`}Z`qix4Ib8}jkfPg8M>YX2nacQf5QCd zVgIuu=`$1XJ%@>gd{>Y)nH6ROhh0DaJ}&QDJMdfGEqh>{KFNBTB)5%`+o9^O2l{%c z)~M{yUdEIUT=ZFn5Lvp{^%(=*2Wtt8th!V_a%fMTG=Qj9JF>ERx)7kgqxS?II9)=) z-QvGuS8n%GME07|&_rnK$wglJXTov>Ux;a~i^k3ku3lokK-Nf{HuC;4xI2J0VYyBy zShr%LhqpwM(bzm0D49nm_1Q6KRzN#r`1YnYcPijyOiZUhAh=)^#*?yfJNC*Qvi-(a z3i(y^1P*t|h{jNfd*tpvHQWZWzq303C*3}0!uUB|8l%jEtsr?%4mfnmNT!FsPHfzNzb0V$ZCg89cx)t8i+m4v~JDlj+X%H(yKFs zUd7Ool1YtQQXpu4DvehZ+b{SO7_}ccvZ6XHHb+{Xc627*hwQ(yM?oC~B@Y)p9C?Ye zL`jJdTwLSE`rFCR$eXpy6b}xIhLp37e-Qak5c-#kO$Bj3d!=T+gybHj19LBo2b8w%m+n8v=D>NTOvV*H;H>T z=3*b@6W3iWBkFrX8=r=i7qhdm?8!yKdD~1wrko-_J>L+bdd!vF1T<5=7Ln>1Gg~UW z7oLh1&Ip_geG&B~TU3~+_~g{EDYx!MG&($zH$mRL-h)(_y26a_pMNZZ-T5NWU|P8JXS(S2Yts+hJPI))JGm> zXd4zlk5N;S1AtOQ=o6@ChX4c{xhL+~*#C_+O z-{x`exfDW=@YoZfS>}B#p1;MzT1Hbap~1qq_}ZQ+5UIKPG9UI07?xPOCjxU7#j-ao zVjEdGWyoM*{GY`_eT009riMxE7KTC219}L1kUZCHlk&-TfI3(LgtvzhEZ5^ij^)D| z_&CI1C~q4nfT**WtlmSeVs>q9PVX05kf0BJVtn}l)#|yP^CUPU?()LX+ZXDr>AVeu zU5_9;qbbd`;_-U=@m6)c^7pa}<#n+0BZUNCcfBLex}8!K2g*L`A?D~UNUniQ4D+n@ zL>9zdWzx&UwhwksrQrrIHgcp#QlBGFFewnD&UCEvf&CBph}!NC>+?*C_PNod`dg7c zO*s6*u}$?$YzEjy##Q^S3UZKe<;zr~7JH2E47rcuEZ~avSmyd7%T;htyV$=Lhn~K? zr7RX~N$%*+->&hq5`6k$$qWj_wSS`H#T6h;vMK^ixz+Q@kk81Nb=v7?BECMyy4bUz z&upoE1BVsK;LcCJf^-1m5PCiAXFr(&&U%C)AVM|0KcADP#+oaPAcay4i`)tN8INOh zEgR%p8SU^ip!xWIJLg#yx~WqqvA=@_r#r`xWz;TgHTr_~uLUxA7D$7`9rfaZm@g72 zdj~zFR(V0uq1JsNNA?_`uR0~5>}MkrGH#!k{RFb}d9I_vE*sZ$Y;>u3&QYaU z7p^!ymJ?JIeB(>MSGiwUJh2GG$#mgqKT5ZO=~5=j>anpu&wrJBG!3yypBV(rJAVj8 z--^ZePklmV{J=NA;;9@)+%1D2VNmwyDu@}y(-K7JQ{ zm+ObYxw#}CfiHWpplI*uAZuQMSTGTEj)2~s{;#({6Braw8?l+I07c6_@UUkT%dRh$ z>LmWxNn*y@AVtqF8>xpB&fE_HQ!%=-w+llnhTw~BAAuUBr#m1x(n#>#2WUv@x10;+wk<&4ipKVUbL6mfGK zM4l=N^(39#tdF6S>7(p{I2z5@2{fJd%}gBg&HZib$YnDD+M_E@v_7-b30}W~Cw7yS zdC%RUKylF3Dz;{(;c1{HzD*rq?ulNh(Tsjp+h;7&OfEfmt%cdvPLqg>-VwrlM;-MA z5Wqc)1fzEZMsw32R<7(UR|3N#zHRr7G394_&W8$n9BAR1G_zJNuhibJs@!9qN521v zQhF@WA02~3DV7q62#UR^nz{EQ5Or(t9X2uBg#{+!QGS|T@oZzvu!$w1umqo$sb>Hj{U zi;ke`ff=$e>+UenNxy=|>V$^J=KW9DHn#_PZ@*Wl&^nmfh?5dYVZr@(`%jvk}&+W#gN2kfpGgJ4QubNat}7LNho}a zonULCX5;K9NB+{xQzs@g)^@FT92|BcGJS677_g~995YS#(m1)f(qh*rqh#^PW^BZ% z$mv%=JJAV*+7`cZy^Pil8(1!3>gjRusU~*>e=f|i){t%XU?d5wrOB!dRA&iJh{Bc; z&#yqBl&=xvZDP+(RA$n?40u^V4!+(cDSSXEyTHTx8Cwd_e1Z>sw#;LidWgTs zj1&t_7$-vO;nlsEL7cXTKwoAJg9cZ=tHSZ!V{7g>(pzhAR&4`Czrmc5Vb`+zdAyfb zfu7ABH(}NOaHb{A&D#1MF`I_)<_tac;ynjFOasqgEo=1;K`m|%E-&&q5c#v}@}1h&tKMglU(b@stOevqNKEY{!q@bVUD85}3&&bMvUt=f5jYC=6lyRA zQqx|->zP$WT_hx3lAaKAOa4YDNR!cAh(Iz$30t0Z&AcR8>e!D_|L)dU-ak3&Sz{Pi z)7d+hYe_Nh_aTLBIzEs4j^7n6>QN$=n;mNX-m4tt))7ciUeb~@;f&a!nd`4?I@W8= zhb{9omqAV!%Q#<9r={!^hZ}{0M{e`EsOfRnQvSHfB>BvDf~N-fv-C`;}v`-OWY?8B87!) zy&gND6o0^#YS1s@=(b^1?-lme4=NoU_^4CfztO-?y0Qi5c3kL;4I-(~CQ1kLgpZH* z*_-*kCp0Gy_5kjv-k1wj!fJbUr1-BbF#Nl!aeq*ixC`q^nuLUuM4(0F3^Q>E5p>tM z4cUrQL)-mEP(^wi6f0h31GQ7ueJ>DL{2>MiL}6=l=!s{G(=`}@$uYN14KVP!)0UQV za!kTXPP)r>OXhINEUQX5NbL6@DboxSO&QJn=hVHy7c;HtG|7vt( z>I)QBhe&4lbi~Pv5ACIPdW4Y8n&T;@OLjcGt{cvE-S2bnI+Iv;l;2U`eeq1TbIJf} zyf-RLs~v)?T62HXrT+(-86R!5RSdYx|17Z3_igQA%uBWXgJuEK82FL!g>)?6d>c-{ z-FplDt|_}!meu=0i2;vk6(;Z7b`3>*sL>cCHn!mtPHh)D4�Kw44D{jqBXAY+JhQ zEo7JE@>A^#4`jz>o4z#K&6wH(ftgEO(J(*eBgiyap`vmRKr!3RTF9E0XTgPkqMpD~ zrf=NsutPhmM;0heW8WH>gSRGDaVat_Ch3%CYwgiE=zCdz=lcbK1fT=pX2;+b`bW?n zPVq7k=n23$m@x-`->6lY-@L&Zz1zDx@A%TEU$+_q}l=|~2zNd_1&BeDGgq%qMjVD+@nr>DZG!I&h_rE?dq)w=yR;NKxah3sHcES!^t zfcrSwVFsaKbrn7Q){5d+V-?!rpHd#u0yOWFuQV#5npIOkCpdh8XwO7&dbhTE@fMPd zd}pCQNdSEaaJV-06y}o`U0;ojlN?nPFfa~}k8_ump~?J?9TJ+z!~#utZs>P|CshYl zj7^L?Z=mqe>1`zouJ_&M%!N4`M}hn}*7~DyoZs!16*%{q#-{l-f>oCAWoejo|;HV;PN6?D)Pr6#@x$5#cXKFK$ zE9vXuN|~s!di)is0yR(KsuH<`_%A0pAqD03Y6G1&&f4+t0$k zwdD^qT!l7->?xQ}u8+?5t?2SC;imBH8J}1=5^D=L2tAL)*dTP}Y|{l66U-D4*_ke? zx&2Ab;$Rl8FzMREiGj;n@k9K>(PMq!=+0k#E*}tqyjD&~WyQ za)UR@1t{MN8Ex~d9-L>GM!NgAFby?i+CP?ancI(w?2-@~=BcV=+Yxyt5x=rB=en(> zu2jsVRAoeC*ZE;%fiVmY@l>auGvQ0h^)%LE!p(^#Vf*N-HN7rL(bU($ zpV`87*!^_5v|;Zj!ZPHRrD>Aor*+FyvzF89Gm<}I*!%5eV#8Q`wqZeI^Bs#P=pWMw7|_!s_=>zkR6+q)X2N4jQL zd>s^jC)Y|cq)6Kx2Ap>yfwEi5=mwey5wq5ttoNeOgW@eTo!M$xG}g*D0xJ2Xebq$v zz8ky%0XZx_3*UL>Ith#QWw`-zggGOwyFgd1yY)J$PXZ#zX8wz~=PN4{PR?o>HtBuE z-v}`Wox@^K!lJ9@bKUHNA{y`1)$MpP`_C`@CWc>v`iCU|G&$5FL*B!g&($`lm2@`lH?ivRYnVW)>Qt?V(|=y_*$9W1 z1ov?FRFdaM3n(r&AMihZErQ&tgZX;k4PW%NKLSDzKg5oFX;56JqmIR}uGVl;9WF-z zF=%nzCElV80k~Umf>;CpM2m@Em9f9!C$zr}0VPf5uv{ShPNW4+cG;a)$tqu)M9Q$4 z0NzdDMnjC==|fH>vj~t*0P+n$`r0dk>X?XW)${r2W!=bGW>wS~fyh{5RnS1q{ z0$=K0p9(JIWj`YJ=b4b93Es{!dMq}Gzn;UX20QZ39(Ke^I|~w*DBEt3!Sm5AMc-17 zN9L`zV_#DV-xm}A=Pb9P6OzU;<#AXCC)9se8f&B2)g$Lz;Hw_8vRcks@V&wC^zt%H zskh7FYwP^mE5C2OM~HcO&o&cJjXoG@B^HtE4-IUt3k<2M-_TT9PC4cYw0M-65h0C+ z8+hc$n&Ui$_wkiW2W>Z?0x)j zm+3l&J&s#V)Hp?iagzg`m7_Tphp}<)?z7XC!s8Zk)sOBV{*n+=l z**Z+b<7d!@Wo~ayU|&svM0WG|Qya=vM$#YDS&eR1lbE~_OX+=WDCqMOW?CA0!$&qYg*m~@UDCY=Nef|&*9#UXBVL2eXbY_H9$g+cM5w+Mt-8<8pNBCK5~qJpBHq* ztq)S@NSTcFzZQG{eG%R$a%?Oi3-+TY@hUXZqbT6!=F+B;4VHK^3Ycq`pE)~%qldFt zpcv;mCN-R(DgrY`>D_?Q8+V=`rEJX;?Ln~w#F3dDyUc}xRu@ztCp33H=9V8E{rgE7 z)`;HM2}g;d(2ozhKkE^je^r6*Xtho3ywf@^$cCM2yin+@_A$pzY%DMS{f5@S~kJNkH;gSwC{^}8T)QbH15l(a*BaoHXCQ+fXF7jFrLwC z>NYj2$A5N=*+k>bXrJ0F8nwq&1&qecHzO4_0%G=RW)JncMs5%M;nOWG+Y}iKX1mii`wp^H|Tx zvyG*W27p21mxNi2!V|-okY5Q6_kis1=a>J3TJ`j(5Y7D)xch7>f`jf?kL0bE-`$W< zX!-Wd6xno_;izqLu-IeD9)~|9%Wwb0jA^FTxxq^`llt=_kpFwn401nv5j8vxA!i*t^#2BSzbif=y>l_18P!Ww-D(>QqYvmKy7ws@Wf){$p@|2Txj9 zA^4@|u@==1?o#R+bDR%hluAKbgCv=pI4i4>ddDK&6?PO#8{zg%r(gIp8WY$2 zY<>5N%O4EOB7m}{7~QmlBVF@J4?y-~x8gyS!DlByZFDlw(`W$#sScn^VvA0O{?stL zTqTtD`&FUre%dfUCStiGtkFAoml@3>0aqAZS~yYKxYK|Kl}2dx!1EoeGGje8p-xud zN?!gd9a$%vIw@BjD_V+3?RbVb2Z=z%WYK8H_Mfn>6-MB3`IG1#_Sql}<&Y8e9@u(4 zIevch3&M`#6ZK#50_r0JMX4Tc;T6*8oCs^C$sZDuh}axjG?||6_A#fG6~+Kg-1Dr& zEb%T@kedu5ZmuCt*8;VS(wzkM+bjWPRJqe_Tb(`ykyZDL_c#rA7juS6+;-p!>^K0n?f7{9(T#J1GH}VOsni)x4MUqDFc~dCI6Jo97g|PMis(W#=7Hj zmIVxvk)J)}2I(Y@+a$$`u}*7E!sN2<8Fmw5wyeb}Z>diPa;tVlpFK=hGsGA|Zs3LS zUW3dhr!GR%z{Y0f-nY?(D?qT(uwFZftg(!SGq#%&BuGRj9j1guI@p;g(GW$`?p1!s zcKjhp?uR69sd#Iuj8dWb3&rsztH}Xv7*=|B?OIzF#mxrO=rb`^*nUn+U+|~C`b;7t zcWVasoqqDN>u&dDp}VZJtzfk%3@t=? z5$YaI$q;bmx8m9-0^PMqO5ymkb_N20`Xd?oX$$@yH-^AzeMT8O$!2>7l#`wb%r+#U8_WSMj_?HLz6dNCGnc>HfHx2Gbt{fJz?YBgHF zVur1qbLZ{pgW``rI67`4wVf?pF>_gk-P_-77szv)PS?fvVEC#Q2@cup zJ43MUu%>@T|Majyl2qqQaA-t#D{?wMo+D=(k^~A$O1$1B#NiaX5m)qY)9Aa zz=5x9JS==(n!>gdWTbzjs}hF1J)Mbg-g&ar@Z;MdzA4wO&nKq1sCP7GvVIrhlXoW< z>DlyrP!Yl(4J>4O))e01xxdov=*&8OYc5f^-i9|#B#N0N{Xx*sz}A7}k*ApbsD+Y^ zo@~=51AkKS|A`l*2C7B54ZDn?Qv7LLVl8PX{2hKThbV)u6{UWsesf?ideT}q`6hhb zQpa)DdNN0n+BxOgeq%&gK$qxg4GiL*4Rtl#4RYOL(uH^i#zC{Ske0(M(E=BpKYx`I z6NcNpX?)>;bTX1zM?^8QEV(BWFQmko#{7n22Jql|KSbZrIfd@P1P z!9B3x;CvUi!%D@j7?@vlN0F)%Ncrlir&G}12rmjvv+6J69m0NzDiKY`^PjGa*YY@QB`kxzPc4Noo&}Y-B;?bQt{meNfD?AS2+Rl4&KRN3n&;zdzKVU?8DcU6 z84iUh*#uFSXon-JXD2Lh*1z^uQz(5w_||M8<}ObB>hysrkUB_+drx82kxD~7oE7Ho z<-*Pcv|0o^MG8UP7&a)CzKL?WOokI;L8|Z2LrF$xcG?%jw0=afSnXGRM?L^Z)k)6- znb3$SC&mIk$;Btuafp2vVMJ2#!F?~BLp?Z-x=>0JBWK3cRMIyyp(hV5xwjV19+9{B ziv${xgn)sY|NK02%#a5Ea8SHeB0)>F@*riV3$#pC*y?D$_)H^taT*wH{{1Yr;DIA} zm+R~`zxwK~kSDl=E}LLBlV69Zpod#erS&*vL@tT>adsD7_gO#fmhuQ3D!rJ%>G4HB zRm)_z4Ri5X$pb_+hL8Rn19ddLhUyTB71H6BwpRMF(wpjS@ z{~$$Eh$Xx3dZ{z@b8e|i5N}0+!onSgS=zUgF*|6awq)a6NJt0y)jSQ6CRDwjUhxdX zyfdBV-In-gom+?TG0ADINc&sfFqWSR#2OE?)0{+i1D=;X;2v?9CQ`Sja}rNB^WkLa zYzuZsZnmA?J?Jo>`rTRnuJGR|mW5%O94EJCejtXs!$ zZ`0$z`>f9*`jGpgnP@ER>Y;7(n}J^<@pTJogN0^W3^KZev^oay$12WcQ@wXoUr^4C zG~-T13y_@8!IBh=kKB&>RYYaA?N!QSDl9-wBQbpXH|`;h8Bv_>>^+0He>`Eye`GXy z%IT>Wvm(ytF%iP$77juoB>$?Y!9gRNPCt1|#kzJ_@<}DRd92UMik0oGrpHMMZMqZe zJcRm4p~DiT@Y{g9;Ae|BK{)aML2p-&dU-|hpU0I<@Le?Mv_>oUKP-*wXQUg4n9rIz zP`mrX-cq5?!aB){bFpyN7GAok=8^BoUhhF+W$yxxhFrw47IpvNv;Yt^PKq9f$NL?g zLa5kV*5{g&Fz%9D;qqtt?zulMvJ`tbRl){z@ofZ*MuCn7bH5sv5HO!GBgBqGD2S`x zbIPV}W;BR-M>}U`E&(XM5@CxRPRt(Zlk_?Hcwr~a4fSzj&Szo*XnEN$W>cHhQ7Y*& z*AUN_a7f|R%ut++`fS|Ys!m-z zygm*0OXkC3S4=`38_QLbgxjytTq8fvFOpT1oX{RT;PMU22E0i<6_ke^ zDH0J3l{V zkFXkir@a>M8=O#}#yuHx^3RGa0vZE8x2iwJO_)U6>o^)3n|aj_Mp89|Q76||R>}T$ zFSSVlr$oYpR5sMP>7AOP;EgLbpvV{PzJ3u%m8FO{$IHF;-Gce0wak_5mb&Pv`0q-5 z#YIj~H5HI~+YG5Yr7sbqA|EK~v7bmr?S4;`bwY62OEx1THFN4l=O=W1fe{CE5a?mIhFQzFBRE&*nF;!3x>k+JpS`dJ*Y(>?IPu zb~59>reS*Z=?3T}*82%T&4)cuKtfy&>u2GK5%TC@g^7Rym12@=`lj2h-|2Dp!)0qp zf&H>8$rYd@kW=$>5)qw*KJx41QK zQD>N&w%bRMhwqOquq0)4tVRY_DY5Ef!NRZ6v*Pf~Cs9UR&r4YJK3n$#B}sIpqF+8C z9PXixE#*#lo=DZCb55Hq#Qlqq+ml;djaug=jpg1{xxIJsPU!8>YZ?3ECKCC*zg*E% zV2XxLi7pvwj6_a*#e+RCu}T66##CEnn@q#am7zjIqAzT%=^UDq^35H{0}yL%WcrB5 z{s&TI$h6*Ij=fi_G_1D^T{^z-i?$Rbo}X(gQauds6HRarIVWLs!GA? zn)h~UzNdJBB33qX6`T@MZv4s=Tl`Scg%8=xh&i|rB};Fq;F@ec)@-?mB{vUydBxkp zZuhwr1hL$4yUbnrkP9>ju>F|peaFnBn%~9ZYOenUx5;nWpwxRxRgBnu*C2NV2?Vk< z-y-Dsrcx6}JVsl^Z9cZIyvQ332+a^dDSZE~!V5s-Q0c}hX6zoA+}nBQ)X1v5h2&K+ z*P#(SCa>Y77^JDryz=f7Zzq9kjf?=Y`*oa5< z9z4)!6i1)1D_{B-;x^#czoCEu=WpDX<4}9;dky3w^5&oo`b$%Pn*&9&D9j(=dTuhN zAI6sGG>3;J-&B~K2a)RF_NC7ZnZ6VvoZVrkj>S&x4=vEwksh0SFq)3Jx`R=GkHZy!9w%0%N$7Jt=9!4 z|I6h6(VG7OD@a8&NDquTmZwAJ1S41tsB(n&QwN+_?D_${XEFpEK!$gmy-QHL9wMg+ zS-h2~)C@RgcA6 zJtv5SCo18%w_^G~Oj#(yN`rL|xFRk|&k&94IW=?cMr}c1ucd2dJN=jmYGsPd)Hid| zJgc;vT#~na?5QFA2Tb^^&sJG%tDidTrgUh9+ghxM3AIRAJvz=|kyT6aFrP9wtFW`b z^xhQyaE@&Wq81>;lw{@m7hjl2VgeYAq$VeS~n%EVse<%zi9o(lev zD$Or=+S)6sPcwZyvt!W#*2Zta&ED9$X~a$G%04_KXPY?T?^t)<^!=g^J3K`+!tPd* zTarw#&)=>}-fk`Q3QO>lya?1BowWYydiKC&WQ5M_a*($Cho~b&j7LwN-luFQ@!_z1 zc!MQSAMQ@*dH1$O$kNbT(-7bq6YdVitq-Paf=_O_T=vIsok}14ZgBkHHNXHUVe*a1 z3wDGkT&nC`A8NraeQmQJ4!Wb0a-J25k^hNZfLPI<7uDgN+vU^f^5Qg{w(v$bG=*r- zUY$^x9++Blik(U-i=28wE!=h74b13pS*4P$qdG!*s$mxtU1J&WJIFyQ1JLtFF@bd~ zG=WK5;8_CkwtRZMctJGrcdVIov^4wrRdP&5rWoCAwUiC*|u3kVwFTb}YQ? z25>kC?Y(hyaM-{vBPH?{&dWuu%s7qdc#l?43Ot znIZU5v;Fg2(T9&^a5rD9rC!#1BD7IdG{OHO8k28<=@|B-YcYJ!m)oX21R3jqE3QgJYO9?yYt}NE-PG7hd$iGt?##4 z1j zV&IgjIDdBf4mvPT{QfXMH2`-dzHI@;eQ7W0(13(17dARp$$KesK0G0EweUJdD|B{Z z_aUnf{Ojr_guqZ)z{FXNd)lV-=?ayTjAVS46fYZoKMLg1U?!h!di|ir#i&Ok6Rz|1Y)It^fQ1JcYk1#16At3fz6h16S+J1xn0i)J8ww!Hq zW&ZXBeEB}0$7xbyiP<(e^Y&=BsmadM?=qObgKU+1(52dr>ws~XmN!}Zom6=m2Qf$2i9O#`utG_8SASf5z8E})tjF4=PC6<^SrI} zg&A-EZZip7CG@y&J~`=sUP}aR@?^XFNg|ohjb=VcihE{_?_BtP`{s(T3Yal|?ZSs; zm(#SBaWlpBeJzi$pZnhq?+WwNe(hd>b#&mbl9f=a9vN@g(R4l398LOw%6eh0JY7Auv=4oao*{l{N4kjgI;U*D0W|B0sQB!cPr7~*qY#5tluPL z+aeasBa2f=e~Gf2MGSal?JBh#*NgaGze9`mR`BL{wS=@}KUcS)799$DYw5D64heA>4 zqciEX_9?OB6>LdZAw-)V>=>GRv3XPB2=Tt=^9ufXje{+=gesIX4%9l9P;gj}$gTDy z9i-=T+_hnN$%Lzz{$tRNM74Tk4f`Q?a@|?JtdS`l^)ok!FUU9PhmBAb72|pk5qZLcdx1* zItnLVJlfM=pMK*coagMSLD`P; z=V!Oq#(qaLWluw9&Opy$OE&EE*R^`m$?8bmizU5#XYMT*3P3x*Y46w=;h^G38?Att z&;RN?08d%|5abqQXdQ2;kzTJeXqj{SgpM0>o^76dzw_&(Z?Fe$?nLHk+K;d0d1rK0 zvl2Rn;@EJPv40zr*@S=z`UsY54CfF>+B27LTbGk=h112y<6i2F*GlQ6`+{P zokHPenuHi~hV7#qDY&G5$Z6jWHhc1CL!0$rhHPBpq4B>NrVm2W1sa%X<|1LiH4Y$J zKq^*^;FbGgJcueFAhNn7Tt!1NINT0>lI|f#PC_4PnJaBTB%WyLWAy5hv-QM0&yUNz z_bfVYMLlk{;*-l{Z*M0VEm=*l1$)~%FcGRUFXuPLcig_#iB0TTf=ZQZ3FwGj0c)Fw~dt zgFJUaLhzZ(qRjut)msL|nRQ*mf#B|e;1VFXySuvtY1~~KOK=bF5Ztx#AdS02fZzlR z?(Xn)W}dlc-mi)(_;GdDMV)=NuC?~m_4kiHegRipvQhaUfjFhAKd!U>Vsb%wEcWuu{RLxFCmNI@1%L3QeW>2 z9D5(A{QvIH{ohd5xOUr=G&?%8_IJq8mWQ7^t3u9b zu|Z1lC%5lb)CY|go}*dGJK4wTJ-^ad(A2PJxcdqe`z?rJJ4vPi~$@76f+6Y zp`?4KvPJ%iLs=E?~|JL!>GJyo=qRO z#~TW~dR!d*)TV|AJ&pc6j34WGj&Hc`=!sx#YpS77SVZ5V5T~1fAk+v>E*F=}^m~U7 zJzsURb3QZIbEl8_;>zwyk_J)))X+ng*t45c6xUQ3-R07kj%ZdEp*fT*wgZE~bZ`1H zOR(I;5x^IdR=9=KPP8i!IM~TThyrikvIftP6M_}N6nqWIUFO;cC>KGW&x-#h-G|hF zlJ5KGrZp3l!Dsd9<-_;V)w2rs3_AkCFBi#AdtbHjaekw`Z+R{Ie-MLj5Gw=TEYa+) z*^l?Tqx4A=gKPY3O0e|BUA%^)sdC$)$_hYJ4o7wHEedrG)ltda0dzGg*vx;iCq6^u zKh|6dn10E++eNAi@Bez3h9orKC2mdaa0B|DU-F+1T#ZwG3xi{Z;wqIxy7b?{JvRU}$P+lNWkXd3=ncHQ^kk2Xv397L_O0Jt}p3 zrc$h|6iuSkL5{m|Qv@VH3x#`*uS(oO*t~|`z)mj?m36=#61s(Qt5wc799t|uNzk4y zhf+d=y$6dNt9sjMGujVMZ+irqaCz|0UliVJSr}Ski3rR|J(F&EX&9%5I>!dqd_6yd zX<)zsX+;Wm!tV4S{`w|*wOwG)m?xGq%X6{F_zgS@ydn(zU^jEoPm72TwL(Eni_CqO z9QHRo^Z%U2Iy+*U_x+|41w$6?{?*;5>|DSGWW}*Se;riV2r` zo6vXgZiLs~&1=Z>M!_||1G&b0N&E|s+8p9P$Dho*pgKI;U$C^Ri(cge!KHRoPZ&fo zQur??f9`{j0D=J-6%QSHNi$jU<&RY^yKI9f|C$ZNa>GObnYhx<{mG+>nke0=hda2r zmpPCYC&s+Cs7SxAVmR1Np_rWqxE=I52$780XhbD$n6~+TB+oRjKo%+^`}!V((mIOy zm+ava>f9XtksO&-bR!C6&1I}NBC>!x!I_zkLM8i4-0VEj#1B*%>W_Q5haH{pNq_c| zY$>293kf`{%9KvEnKp_0_{B+yXLNWH$AkFI|qrxqG|kFy5krHeH(?Oz9lIgbLZ;eJ$S91Q4X#9SqJ zlpW6+4S%ot%0>j;ydTm$xmwvXNqnE_Qr8cGaR1}p7x1B%zCRxHBeyiSB>*cvW+_SE zoDq6EIx-t>y2Dqy=Tj+LZj|;5-``LE2A>*2Wc49Ul?!n7D_Lh3l>qsYFTuY#)t@|w z6DEpc(7s8M^cz2`DPXTitY%5%plX;N`@smtjY52$Q`%4Z9@$rZpIX~3{wv#*ijwhs z&?;ee8VoQ?Bp-KJpaFK!{f&ep;edo0HxbNFF*af20>_DrA!}@SDQd!GF(nDMD)KPo&ub!YSU5oB62;jk=2^3G`GqlQJ=Rx+RDSX+ zgxE;Def8@XjRsKyirkOb5|a_&%PeZOsTcAfll!4J@GPj$GQIEg%nnFKlp%%jVld1i z#G0xu{oz*NR`%^}J1G>UTgG-4D%}O)25%dWwd@#uO z&P04)>V;D}pC8w{Q2BH4w@!U8V=ttIUm%6Ga`DL9eUK9=GxJB*~mRI)CR2@lzo78q$pj@&WCqquo0+9n|X5@cz9M zWC8Znx$Gz*#=TD%Qv_uB5I6#)o;HuEk+%%KKIo)sB7mFe;DlMK`dqx>naTZKZyrP# z1F%1W$HX^fN02ba?qO;tz5jx2iT$I~eGsMxG+e{X637;BEiVvEmx-PDf zgnkhVH_ry{csgR(%k}U>BH3RFMY4=i9UO=*pHwZjMyO?8iQovu7`X36=8dP>x#oH; z+01Fk8G`9Iot^yMJli~c9{VbxVmEAM}x^ zODRWx)OssgqH|Km9N(|B2Q-IB1ZGmji^EY40XV(tx)(aRWPFCD?i7Fd0lnt=E8+Qd z(`w`Bu0m;99H88W#C>TyX!->=t~s9Y=m={^#QtfzL~wLEQnJNK}RkhcX;dQOr zkJ9hl@)al1Qe*i87g>t}^uIm;Qdlui8@>NH+5x*Y2g|CBx%6W|wlvCh3wBV@C233u0;mu~$VF;Vz13KOZal?(uE;*I5Ck>*wdpV ziQ714e^kuaI0H9`en>s#G_>aX#T}=h+TpuLl_=&t@jk!(q3pT~NEPcc4`y%;BUlp?Z%xF7Dd>YqO-Pm>_Vo4<(+ zQEhgpxLe}^S}#UZSg-Te1uT=Fe;^Vu6kWg}UejAh-ViE|f%GW0VITAKJF<8t5Go_NQXb1Woa z(*-q^`pb|=R|4E5q33Et44z7okq2t%YpOJ2|I*JG?qn9=&g|tUvU9 z*xJx~Nr(bC^8&lI6!MS>K{8x{2iF}z+nat~K~Ri$mB+ytc_gFVof80bZki$m#D5H@ zOB~xOH~Kc`Wq)#;`SrokjBnJ>_7$>Au;WfFTS&0qs9&-dpOjBi;MJ7u?D>$JjYN5@ zkTxvr(dDySFc-k`?Ap}7OW=Q8H3+B)4VdP^lwQB-Q@OQZz5>>tN8Vb{w0{Jcn0S^m zelS|9&l0`lQ@zM%Wqn{DUIk%4f0%IA%lJYK!F(72{*WADbi9bwsii7XCHVC-?-m=;XMZh#WDsS9PNDo{mdyetpTD z$>DXTkO%b~c*(>VZ9323_rg&DNf4Y)qio-?z0Zpw?8+MrTh5k(Be%#0+1=Z-Epy!n zo6*jov;=IXE7~lt+>&m!Zl(!)aMxA09iTWK^`TZ!=U&@i^{#xxxn-98BJ|vWvk!vo zU?lmURe2*mK_&-O$~gjMD0ufi;b05bym@c+!HQFg0LY~3LK_47CFga6dHeNYv@_xL z+VAz&#mlSltB!g0|Bg{hLWH%LRSBT4 zdQ<_~%J+4|&4U5oMEz*uFG65*w8(*$(86Ms>D7Y9<3>Y5MwrqHC=A9}hhQ6;#gYch zo=T|j=C+o4ZaQMj{!T3hecB$e6a0z8)==woY$>r9=HDjeUkZ`9K-d@+kq`igJ$2l& z(t&mYeYlI~%eig&cJKB9#E%_W^)wC+xbY<)9ts5mF1yP-3syt#kRgS@IJwi=QxQ>{ zrzX`<7#ef<+*)Z;&oQI(YUq^ME4guMWb0-wgSHTiWH%NZulf4A=oLnWPKt{}^sj&h zG|-ryvKY8|IdQMl4PCYS+ZugV zteAURlu3H|=!`x}LkPlmwqZJ?aB>G(vNN0JdYKI&x%UyIF}42w@VB?1JTx|tUTTR0 zpdDjPSOPRbe)QnLa4_e-eI`p|EH|N_Ru!hCE$~x{4Y9o1_T;?TReF!okYpcb#E#n}1CJ5(tG~6XkrC zNjn~ueEkd5HWkF`1I^B`Vw3z|ir1?{rM)lH?f);4`Gf~WIqUL3>3{&95;{(OThzEk z?iyk=DhEGVzY23!r14m2XO*~5@y_)4*eP@&mm#k`+0crS>krK1L{_8%y~#%hy}B8C zv+-uw$3d{*8V${5@NwU{^$)*VL%*5D;A2hV<(9b2pvBS@|F>BQBM3x9TRE5LzUIO9 zGV;G=0lxxSmjXKIQY#qhq9w$Yb7m>xl+P>V?p$T*RqJ+}9t)xrO}7AH#6wivdO-BJ zKt;=bbIqrpHs2%X^>u^;3U-yZe~<23c?}UJ2q1vjtPzD`a(2%LS^zHIaGAi zj3`Bw>f-sF?@fr?>)tz9cuG%xU$wRy=cvd4$d<>A*jhelg5dCY$cX-qxH%sc#}xfU z$JfbTH~0EwMRPW9lMJ_W;Y<$$c+Bn6#1!7jFcBg&DEel)q7#E zd`#HX2hkfaU$vVqS&l35fd$6?5(|Iv%Um1nKQWm9v^!Dy{p&6fG1@_$Hq1!yeI`AZ zles!VbB2Cmv99V`k{cWG|IVO-{A}}{9n!$A?#sfxJADxafT#R>$pSuLhyN(v`Hpzd zRY-`9&TyY0K>Rq>apb)Pjkma7d&W7qWP+_DI>Qk#~c zMt?N@OAGAasz)Z_^rR50ViZA-jZTmqW+QLsk zf%t-5Lp+_UAG;}HZmj)=a>m--=D759o8cR}g=>Tw8rxEPR) z^FH{u$3gy<4kwVg?cvc-r)ZlDJSht$tgG6Iqx5-(w@0!ouo0!m*?70-ezCS7~a5w9-L!l^_ z#&Xa)E@#XPjQ)>LR{6Dy+3);v#7xfe-MfqB>dq7^#KfgadgZng3#{<`y#XWkthTRj zlejK@SvCL90&8$#$fnDAobswCAsw8}#?wjq<|A`F3@vKjvGDnS0Ar@-iW?sq?mgy^p^ z)tqR)W9F|9n?T3|PT&Bm_NdR!N9UoBNm8QtZFM$j z5<=gz0-8T)5-^_WFmMpCahlj4TO)gwI z;nXeDNuXu}IC2N9lx;qdg6LR2em$}GkzXr~KM>Pk^%i|R`z5#cnkOT=a$z>dNIY^C zZ2_kA342sI8zW?lzxDj1nJcHZpjo5WFBhuQt=7Ey!6{naa@)0P9*rt+&xwKET|*se z^0x!y&_UpDKD&FV1>CM$$iAShl`8trjOT~ncr}c*je-FXi7iV8_FdOt6Yf}Ve@43B zV%hd)PKT|wFUPEUO|y?E9vo(?Jiri9D#H^7mDqC+;3x*azr&Txkfo zmkM2sYU#EAgwk~_XsI~O3h@Jw|I*q)%Qf~yd7#>E!MHw+^Fx_qJB!Bx0`v*K;&GKL z%`QafV`CXva-Yz7k^>V5Xg^kEB=)~>EMwG!!112kcQ4tH+Z}GMFAE7~as14bMV7_N z_t)b4o~lvYpkqAD9ScH)6{q72Mf1>f15UAU7rv^LV2o}SuCF2}^kFPL>%N{JCa)GGqk+b$k0}BcSX|bnAn_qgiGL>u$}XQwub- zm4MO~K?kM6hKy#rL1})kC!IQnNChPT`urCJI*15v!;=t2HJn&e*gg)6@`%2);>^NwzS&bx(e*`yBMI&~$86*QD>m;IvaP;JG`FNiMb$rIG?X zv#2--+IQ2s60W2`ZTp|KKsaq8gjD+A$f*t?Tvyz@>Z}gat%dC)!=u6EK?RXznPaoT zk#2~jy9wx12RLiB)>GHC^>3Lt?d0(kW;<~SR=Xwmlz*Y^#Qe*>gtS{f5eBf>8NG2>EG;{c zwzjuWpgeHtGY}o$wzZGg4IyO=(5jDE43giacqaiVyqIF5bz|4LaCvVTGovM6=Lm!j z_;EcJWqy<;J|G{})%5-nmG&-Koo19nuRzp2@;a$kAhVK7P+L!dapH1?@1IoO5&vuE zTJ{A2`7`aO?_OPPNW>Z|d19GvC>s-nI0By1C7YdhbwkwvS#q2*gta#Ehl7bIqn?xA z?OrSvCVe`abFz_9YVKe_fRJ|mr>Gy8FvN?*A)0h&^)UbvzP=|QnJGf=FJEq{Ea7<5 z)Aj~tDajqIh7`$dZG3BQ)BT6jXxmvUCJt7UTHF>Ofw8utRoOBYZ0%?uB5rkett??S z<I+@%0Kzo!`ztt{{4d%xx^&mx_LA-&F;a5~rL)P8EF4d6cdH^r(i>V2rlnO^W|Bbvu(m*z-WXeGI|eN)mSvBqle+ z^v-LCh|S$8WhKdx>vvd9PQ^tIXJ1#o44)*bKPB)l}Rt=+S& zcR9aPOXRp^86xc|_2M9>vhR~uky^y|?U|PDm0q89Bp_OwHgaCHaer$~%1u=tZE9sm zpr1K>y|}F0?BW|9N!YfJU`^Xk$EuI{bG4Erx*f#f1K+;`sqrwP@TJpD4uMsriR#cl z#_`Kzf4I{bM~KGwJ<(ijroL`Uh0a0^MM_>NlA4eRw}#dj)k{)vv(=3gR>lqX+AXp5 zai}F*n4PO|v9H8_)QvfXc()#O`U5o%gPGLacSC1&u-dv`1pHDgv86zaL~;aWpfu*mFDUk>Kl= z$k0e1|Xzlm@Kn!)AZ9awUQ-R zNULL_^RnxI|1d_sX6U5+^_1Dk%(W9tCVc%pscL95hI;xrrKLc}?EAmL3DQOv&HVw9 z(78CYz%i|{u2TOAY1VD+2=uU25}7u+%0gFAO!XXfD>efek+O7}f7&s6e_ftaAloCB zukq6_qzHb<&qIph40UdIIcQT@Z54t{bj7VGFg%KiVk8E%-?6oMz7F#ats8AA(_jD< zBVGGSO-*OVudQy;%un|U7mo6DZ>aHEPMg3qZx_=)%GS`#YE2eY(gUQkE^LchVfo?*jOmKCqN0~ zUi#`mo-@|_fH1ax0U&4yEmS!7>G;K5+kta2^1swoPr570+-V8yrqaNVzr+LNCH}m` z{fSVf-QCL~FjYw;E3)xbLqQ z9UMp)(M8J@NE?@ThUCoJ>^8xtI9ShVBXM|e4IvkbY>%AgDRnhHP}@9MIWM9l~kJy z!xyO8ppH)Iz4W85j5u>~rA3aPxIylg{OBOKbpA;OO;>xsm>Dj$`wgV}T66}PQ+9uk zM*V#p6xbHq-b5y@3x|cU%+a7AWeKJ{V#8jhP?p^DM8p+ao|1O{DqLd61+*{A`f~CH zc07kYb`WHq`mb7VGYv%j>>{iCD@3)XNNIUkPOEWKDAjdE$j2gP*3#>KseRtjnk8%!*IWqk)q1Mm z-~;YID`q-N0!blkpWHV2!xo`xxqr<4&uq8=<~?e&Bd_q)CM={|?TgY`6I{G^hKXm48Mhup(Fu!G6t`=X%K`8i;&s7-Q39h(^cW!4f;*L>tl z1F)=lG<3~QeyZ(bdEF@uMq_Ia$W0*0&TXc4=)`ZCEeNTHbP%hR{jlMt&-T`Vs62lu z7q)gLGv|i;gDJ;m%i9A-qQ*r+H3Z^W#WG^iX6)qrEJ*Y4?zfxqCi_Z3gZhOk(3nW{ zsQ|M5quFTV$4WB_v`!$J+Sv_zcqvg6uaATK5d=^=iag)}JYs`pkp8{Z1RX%0)%Fae zXjd}K=J1W%Q(XLrC-jA>cQVrilN=x=Tt3N5< zHkpfz{&wI6k~>~bk{_!dY#A69jG^!L*l(A5S=d3qBJ^M2=<^Ej{)6U!1O3Q32NeSu~^3?M-q>7tHWW-JhNK8W^lObddVA4knUW6Qq|hHR{qd& z;@u)S;6lOXu=CD&Gj2iCQN1^w4wyyR_db~-7l9T!(x1ZV5rc(ZMf1fX)^zBEhwWX} zbw$YF`tJ~RFp}Tnn;EQ0F@KWjX~jtLkEx$FLU-@GT4aitJ%pkOwQg~tBtZn)w?Dj`;@=UM3>sq=a2^Sxka@E+lG7$&W4gH6Y zj%;|20vUua&Bz+JtbI&j0#kb+dDAbpH&yErX2v0golVS-kIVJFO!qqC2TQT$!Or!F z2&7eF#w}CCSWV&X{sEJ4iyH3o zQ&OVeFMACn{mXm`Y;RdQ3d6#YFLUqSigJq|9e6WucRax`f9wY%J;YsCw|(&2sD2u} zc+u{p0Alu`ROTC@F}RI248J*?BK? z{)^-~8-`Ut*H1eIdRO-)M$N-kl(We=DHJid*UD(1;fT{?h+x|lV+7!-e(s=B#-~Uv zsBZU|i@cV;bP4c=%??AgV8_iuLn-&u7Jp$PQ&RaHsmP*TsQDS1*#yt5)okL1k%!dK zws5ll@3 zHI#Vh(|hq93TM?o(AzG?l&3dEKRJ)f@uIYR_u9d+uys&m58XUEs=`n?88IK&T zi#_6jJ2H=cq>IR3p&NdwcUBrG zx1^^zAMemYA2q!ki|ezZFDRT`iQU>mr3Gnv8pJp@4VZvNP7l?fAi zf}0fnbwTOpKi2feJjkFcR&fOt+h1h5_z9 zsqG-WzfJIOON^1|YO8KNt^Ax?RlMTO*$+~{1;+$mav3)KMnZBKrv#p>4}Snmb=sVQ z#}}*@A%1E-384nvnGdr7Y0pihIF>aMFms;RCux$3i!s*Q(it+dac#wmnWlEsbBerc z_P+OpqxLJZz0EU|JfoIl`e*{>c-VhUDzYnDG#ARffUMMj*AxR6YuA*lyYG067cDo^ zEW(Hz4u)~45wbJSMy_7xBTem_W1wCHK;9scHM|h>rP93j?QrPh&DnHQ=)(5c_-SR8 ztj8Bq2O+G^53*BC)$n_y;#cVzO)WA8D>r;WPqsq`d?rA=&&6iNv6E z0iU_mD(^d( zv#~x(P)%0;^CXZY6ib#?Fue;=u`wHGWA}YeAz6DNY{UTa7lMsN&+E?&6&nO;U$)$rS1U3h@JW61#1jFic?!CI#G7 zxdkJRZw&?+4?(o-gM?TSABbX$UTMlR{E{b%KK-i~^;^CP9*Ju)_n0Sao`x8s~U>0r_JSgyuzeUn@EA*4B(BNt(cw{*atLL_#dR_UNqLqxts0>*9FglFv4YbzF8D;cc#N<_$z zzG3mo*>0H8K7r}B0zaR5ZVx}IWbAVlE~YzZt9#j2%FLDGv_qHb!MKJqs^56i;7ry%RSV%7;B(x_GCtTTu4)KZjgS; znte>J1)pO{g$K|0#=HX2ohO2VKW#dim zYACtqHtRw#Wg@uX2`JC#O$Epu_*woD=sw0fdc&iB<$VcD z9@n5nTu~{n6F-A;>Skj&3MW$}vDc+G=G zDpebAnIHyO#O(-jE>h*~P-jn(p z*-9V2WCad5m(rbureHhEa*dU$gutGqx3IxYE&255Bcmz`neAd7l=pBOOdpd%pKCba zunHi4Wa6{WKVok8u2%h2UZY%qo_4AsA5&?W6*p-aYbj>5iTvqjcr+DXKet5f*SB+( z^7Z{4w#CvrE$mN9t>^nMyyX6T2Fa7kZRD(t)=#DZ|EHer2aOkwAeSk-9$OaVU(*vp||4?8eF$z zP8h%6rX7VgPZcj+i;;oKoLZ2N;Xe%1Gbb)9fwVt58aQOl`+w8sVT}LRKmEzneo|m? ziQTc$OH)SxPTtr+ZwB%IU`FU|sk_wH!%}?aBA>G1EZV0Pu04W;K~DpFFHD41J*-NH zWCD}|iT8J4B*GhXGtU}B>a6ooa0KpL44GC{Np5tYECI$&4IHtU}z6BiV*GjIsdJQdU&aOnf3LEWUhGk6-tZ0xvCwe02y^okeI$ z+3bybl7>3pE?J6vXgG@PWeSQ-O9&ak?YscPq1ggJtB2H$FdI_*-ZGZ3=tTn zpUHDYQI}eEm$ZDdFo_HtkW!(W_t`&#&N?7RW)!{_SziDFNscGREGrV=X+cTbbS?V( zLhqixgj^utH?O1Enr5&067NzCMEBP<-9zXt)j1;zS5JuHVV~6Z;k)b1ut93i9>(GL zAW|ZIB|VQziyD#TLEi6BwkunC1XFP5^7auJUF&4YIOqh26?@$1fEnz6Ly6NN#-~HE z+EG=AtB3RSn7?h+o5x_v?ZWI>-2DhtOj>Y<3b#%A=lL#s5&sy*2V3NvcLcm@q_(}h zSl}^EeXdAA^6M>EZlSrrdnhhaE=t=LI&C2#^GOI%BHmt3U@#@M-0^2m!Jk}>=Ogr< z#0sz1kal|I5lg`P_ZtBGIHmA=QsDn)|&ePut@;m1umLt1yfKZqA6`x}#&p<6n&iRj%rZ z7rR=fNi)d_ESDK}dd$`4p!JtDC>pcMe2=dfGH1qFMmZKhr7tUgQ*j%-f1mKyMmGDE zJToHE(v(1$V5I9&|7JVvOk!uakg==tC-g_l?IAaHBOGozu{vhchJqk9-o2r3*PH2> zWU3z{Y9fJLaPihaH!493m9ju!3OPn$jGFp08kQMW*a9a^nJpnza-*5@ukf6tZW{W$ ztJ9_uxY4=q-+xyzw;P5IQ_!9oQ+%puWPW?aNR@&wS@T7p^4(D%mjnbRePA)c50m$B z4G~NjmC77m-47M?^4WHsVUnrSBfi}Zx<(g0y7&Oi^<~y|swx{G?yaZxlS%wYOpgJz z6SbpaFIY>|NG zPNp$P_34=vp!CiEm&Y%JG(@EOAz&qSgCggD)AjJK0~-$;<7yII@LizC=y7c4rQrS5 zC5=iA2#=;^J1(>NE)l;E#A|QOF+jm4M<&2p>>c!vlW{u4%zeO%(4DIh3D{*&Gg3}m z&@2ZMP^5LtBrohKDJ3`!#2@@-d`B^ngh*uZ&XHe$@At2~$G=Q@Oxpif%0o%bOn~QwR{Jm=Fumd^wn01t&Sp3kD^)Y9ZHd=u>^pyIVkf8UX4@%)SBr!9joaLj zVzn`2&Oa~mCn?ho7kMjzpIWL(<8t2H1(0M%fb1v;gwYZwpjJ_GpJDa1}HqVUtmQ5K1KkfKVb(4&uP)7YhDLl z7IxNC!t+MzWE ztwP2Ebb;beWaz)BrdUFM$Wit1y|?dzDFl^z-gtqa;AANW4mQXT05lACUZ9z|DEGj^ z_z#%nHEKsAXM2S)w{*RpR(x<}qtfkNKR!2@ypv^0#&&swHFh&0kgs6Qu1}A zOSAv$g$0>#?Sh8w2+K%ZF1eX0PIlqE0aA~9u%fARJ)sn6d$sKSbzTQt7nH1}6myu6mu;G;%jLx4Do(j|71H?rdB=+{+Tz=SWd@ zu@XT{mh&{*h&bT7rNR&2e6YZ3SpgOsQ&s<-Is+toQ)`W3MA5oSCT)!hHDhKrB%5uE zEtEg_#>8I9pBnK7*I`9lHGytNDPkinfW273u-WI-d`E9a4U+tQ3wSfA3vg}4XRq!# z^Rh9X#v#UiEMHPKP&1795&h{l6$whqy0Hrtuck_~SpLFf0Gr{d@;ea%kGg3JrFaJN ze=ic`+k)OC$E;|QY=&1@XvK{2{h%)m>ExBMk_8b$S_L*-u2?eQRW2lh`6`qq)=q+Q zhF%yi(=!c$$2)AVn;$f(Xp#a%BTD|yN&jd8n5mSR7pa)CmII`J_k|rq&72Vb0QKSo?V`>JB@B^Wu?Ch8+4_*k01PVB>MvO3))NK9;YK-Y5hOy-cg-#% z8Vceol|(M+7*0TP4PJvJ6RX8tAMLeyLdbxPB60GvJkTxj$IzJm`82X>>0z1;f*6$h zETQqU$vl&=?*_-tgviOm0i~&KJe>w=v_377fM?$*%L8$*nhW?%9mzHeY7gI=FZUPR zeyfGMFW3Pb99EjaUBzL*?PL7>_zysHm9fa%3d`~KSN$6u1L@Uaalg#~1jp#$UOOmv z8p4-WI~}GuQr+O1J3F*{Ub_wL8@RFtu?F1Pc=8vrG5qIJ&VrZF2LnEFp>%F}DYmWW ztT&N>i6I2dCcM&_wiEOwzl4GOc{Jy9B9UA*64`{H(~}%=J7lSVcXr?+*91%v%|iAFU@47s<9&i_(ePgt;@@Rv0{PT_LlG1yONX7pg!i+M7Oc>d z0oZJ@i-|HbEH$u10Vc?;^NJiJdD zzI#~sTYK)G+pZr+{99f1FrC)aP;0-wj_UJ&K-d~M1(b@`~(wl~86K&8h zcLHnM_a)L~aAjZp=A`(U&F>ys1ltbVxzQqNCf1ob^f zWq%6cM7#&-^U3OE{cLfDImSiQo*Ow7daxl4#-*bGz{}rIq<`k(pdy)ozswz*eXkK@ zpB3gfa^NHEYNsj?*ZOqgO~52!=8H25^Jzs!v#_pEyVDdwPvawWt^DuYF^r!l`Nilz z!J9Y;I*U{TNXEVTwO?V<2!zbAEK3*6mYrZ|u1S7(F2Va;+%zA&%Xd%@jcqmeK2bvX z05)!Fm8$zKw*o&56A1{}$~cg*KHY@eg|@eW>ejSjj>*qX2~>G*Zi3k}btqqta?$G| z0oZk(Jqj6tKj*viC~M_an&>wWKCQ1}*;7fS<;|c)Wh}$ajZ#Ra;SSK%B+sklKt%wK z%TKvYU(nL4QS_FLmu29>mPv+H&Tn`VPy;FyD_(=%rni;)uG#4Db&3~SBZ`IgZO<}d zZ;+FU8q`y@T*~X4@^M069t7cPk>U%uQ+OBr>)2LcV6^62w@r*P@n=uP8xQdk9X5(=tcQ^HLb6QH1Y*h2*g66basekx#pmj zEbB$Qd0x1l%z9s)-vcWSd&23;{9Ix4m)T`v1=uhWGt>t0(w{Wo-mvQ14Q?qJudti3 zTuBGimr^sA$Q{wCw4xD4QS#I22Rvp69w}pkS=Mr2#NCM_J;g8g+Ejl{VcFFFbX?F>D1U5PfEK7qc`<$ysQ6bifG&f?!eg8LX1TDSCp zqt3U>hZ@Z1qBcmxYFZ0`#DS$^R2Qq|;CxM=_SO?ZU(^N{3pz)l*5?srm8^ks_hUeAQ1nUVoZi%DU)TCAAB&8NZE_H;JPfpgJuznXkRe`ruE#=`XT7 z{0X`Y=M+nTe=CI;7u9 z>)iJ$_5z<7Rvn+F$0|$61f&RWC9_X54yhg*2NJFOnHpk$ud}pRSOk~wBEs{Is=YU? zkTDwX-*@Rjb*gadABmgy4l~{Uywgn51rsc=7F!hbkciMHgu+ya;r^$U!s+qB#tqfd z%>TR5_00l$!a|c`ji5dr-XR{eGeoa{P1)^v1A!q1KV-kx3vXN#HQ>zmv; zb2HnTA;pFNRK^Va&=sFVg6Jidhb7y6l}}c|?DCIBi(g%s?%9Kx+L%#u-&8|v6@9<2 z)J1a*Et-V`dsZfeuvfWcA)7?6UHzZ7O>;Yf69&cY?t2uOqn^xAaV^Tv@faz_Y>m5z zJ$Yd@tsyjfzkq9!Wd;I{awyVV08U~TMvLY_NAgHrvp>ZQDzq1=+H;JT!S%S^mi(Ap zy|G{_`%(gSz!DdQazaJD_usMQA+L1BIt*Gwc)xw>Cob8Aee9&`&i86VYq-zDKnROZ6 z+nBwQ(~l&{uk<6mw<4U_IQ&wxZ>g}D)02ja_m&fM-U^zBf2KhciwKI$-hga{qJi}( zz`H3G7as_WME1lea_2rBS!$QwcK%}NZYn=_8qlSd1}_xMJ<#W^kY)&^yf=k(^%8e^ zwaVJ!G=v!IiJJt{XJv%9`rUCgGfvETTyL2!_~g-i7(AW4=yoeR}2dm}$#x zk+6EUtA|)=lj`w!M%WET%(>D1C)su<{rczr$BABxw>dK@aw)mF! zqS@T|38YX^7SUk~Oj5&?%Ch_yB!g`9_}nDl-baNY>wp%Lo3y_VlPvu}cs|0g&#@mtnc5_6n%I@J+k`Hwb&vue+ppQb zUbE2f?UcsapOR&w`mH1m&m0g4s_K7*gX2n>8u}fPZ2J_$Vpg4x?o_@1@{_!)H9_YT zRlx86)P6_W7|n%+cp|DZqQ)IXlymMZ6^(y#@3Fl#De+@`dXr6Ev&)U={AoN*^!RBT5mt6YJBe`oZK&l2$#^e-Sz1XfL z*C&jC?=vv@ew308d(xy96*8-*Dfl+jHUHrc%ZX5rekit5x)#M0nCJ-`k!Tx>Ns=R> zg`(b7B@%=KmPoND6e~kIZFaf=EjNc-t0meD47If#A@M;yEfTeGvnDbbTCS1$VihN{ zlI-6Hg|x!&N33Jn;?0{|Rf$O&>LBK7qhVwHB+wBsr;I-<7%h;0jx5-ox00q1Wk(ox z0uKX3*IPbNx<<1d*+mI2X~T&trQHYX3N*Tb*mb7<{zaAP48KFnQ<5Qp-ZS>^Z3sP! znaZV^9Da?hEo8A9zEw!M`2z5R=tCVZ(=0jB{Ql)gUWx)Rsj}+As3^{;CX!Zk_L)V? zCRO@_HkI>94?={el=_JLfa#brSAR1I(;k_59UumeS2sg3(&U8}M^GoC#wq!EqInt~ zTyS4-^k+(Xu%Ei|0tu0DX`VS{iBDz&(e`S0M&Agq?S#-)%WD7w%It!u1xA|Nv|%a@ zWgppy8oj%s68ZeY81vzPzKB@Tz@qUR2jaas-tmww#(w5`GVFa2-~7)Ct%m#7&+O?n zdEy`EL78=wA?&Nvcae zqVlyVv(Q>llNBF8XSN1izSav@C}MwhKAJLhmX03wKkcZ;-q&bH|0Y zP#Z{CX2JCzNpxvBvivWo_+L2M1m+J`h?iQ9?O$l1`6uf4{9bwANGyz7(Yy{=ERI7s zx&g4ra$)(H9!{@F-zEYHq9&E<0=tg^p3ZHC3_b8&THc~ZoL{>PNXDtwN%}yXjBV~hZZFbb2 zDK^|HUpxgupA(TaRnu~RU-!3;Y*$_Zv#?s`%_jU zrKe(!m=+jRm1EFOS~FkH#0xKq4h@5Y-@G#iobjj#3xCQ)p%o;^iVJkSQ9Ywy7A7%D zRp_ym%^>uiO`ZCM`|x4_Qx++_P~vnIWW~!F0pL|~;clQ3^Sfg^X-B-1>?zLp2anQ1 z;P^A+q+SvAR}y=nR^~2-0GU>{NHveQffx2r38_0AolJOH4Ot-LMIW+VA$2MBYfHAP zr=_)#U=D4gIkfhL=HF%3mI%UmUSr2QZZCt2f%x6(u}H|AH0HtHg`Gt|cY(tTwfie0b0-+y1_m$=^=th-K2{MZhf^dH%KHdm^Cay)^GETj-gDu5C0 zL~8s}S1UjRYFir+%tfxLOi@Zh*r%a`09A_qQ8W}|CjK|*_>W|?Nlk_Z)uFZ5jnQZ{ zL}%ZCnxS&)2=n_1T_x4KL-yNutx6II8#Ewyp22LukJRNmQ3GVTFDJ3|5V5griCv`{ z03-A^{u)F}z;B@HMq67D11XkzgfaN@=q48CJZzxN(Vc$E@3`4yZ6i64`%6U_RI%er zNtkvazC~Xb-c5|`@H`ogXF6&>Y$!j&dy-v)+m&t9NN{x8xV4uN*z0+o_tTyhYMAgq z&Eq@D7=1ts{q$j?|LL1fT;jdmz5yw`HX9s;+5-|@r%80sV6lN2M|<G+?a9Ytf5P>7MyKQ;!n})zp=E_inuF604=m!#9^|u{X+g#8TXO0fNolr^OO8rt;Y~tqxQU*NTN6n!SPnO6B&ZV@sdiS{;Q!pkT~Sh2d4i@42Z8Wmk58ODrgWsT}@k+D~x{KKnnH2GSr&SifiT3U!mGJ~Y1)d#gUompz} z%rKcp$B9hrj+px^<`i&b7fiGg;NlaNnm4jSYPI6?b`SKujv;>`@qgHZYA^_UuI9e! z9|sajP=gsK2Ykt~Z}DY|{ng{k<%pD(VwD*FoGu4R20mgo9SLQd{jGb1iVb_BU0U=8 zeC&lS|0`nt7lvxVak7Suph8r_#BVahl0fQu&st~8{xnorJ2A`8D{6^>t6yf)qNyqS z-B)>{>Z=n=9QH6Tu>`;Q!%sUIw;P?Shvb>Cs^fVQ8fQe>cTmI%U*qVDcPO0+%WR{e zIiM%`&vkk1@KpC0{IqFYx`2Z-SiqRA2UEA4ob0!Fq7lS~Y$P!m%i6Dc13Mj@AukQe zqb_)Xi(DQ7av||A(~BFaurC6p)S;i4uGZHR%*=}f-BuN05@eQ1^;&L59zsech-eRW zvpFI$_X=>0(SUlxm%TaG@BJ^+KE7A%M4J0It->DMDO<59g->QW{Q7!AD)q%t%EA#U z^Tjjg?XJ?|w3Qw&CU>v1eMdWQ-4fpNm*t7G4l*?e05J77#w8d4`0JE$OD<88hg>UD zILUgTbgL>z591eLTQ?3#Lu&jn$zh&PLcQj!uSbe}8L^A8&Pc__n6y>Ep=ZFK;`$J7G<^YiOLytN=q4PqfRcRJfi zzJ(|2?HjKwRf9o$Pg&GvlmI1G0&jAFd`iW^u|fCp)e+ttd85`Yu8KT9YV;yHbFncL zZ_grg6uB1q03ngD`+%?^gTk*RLS#a3g7OdtCIo>HDEi)q8_uMBE*y|6DBideix%Bb-sp+V-Qn zvR~_^a10Q@COvUv-fRV51U^6lF#ZNYqJD*&%>zBXnImcgkc^6F+U$~0PF08RrWC9@ z3OBWNc}M@ybbt9UR!yLegP2gJoxqZh0q3IhniIF8V4$~1f%b3OaA(|jT-G$Tax&Jc zEqwY0=3jm!c2{aI*Hv}uoo$zH?9d9($wzU6&*g$QCTPsDL0sk;CmzsGn zaIP%4=`;NBGk?>;5W!{D$ENpoSPBJ370O zpFwYjd9p&Jxa6a;fD4BPoT{bAcQYz2kx7*IS(6V)su#2gf>VhJ5vMm$0~pQW%mZb0LuEP(|OFQAuuqKBPGcv8B9H6uhf#jVW&%%T$BNM|{?n zv~_>;OQUJt^i6n(nv`-%@W&V}m$NuxLU@pJQ%y<`vMj+M z<|sxKdEa;tZaB?$e_X$oYNWn`$t&dqiFa7`dX6BkJ7C|nc|B?O3#zHhiLHlid^Qq- zf-F$g0YL>m8lGAMK;{wBSLxnqiFBn$>G`pr-1aQ36#Ti$lud#fe2Icd<~z-BstRe( z=W(8L#Q|djm7M(}#T78moNnc$(BZhhGUCj1tesYCWu;ONUUSf^koooMb9QTKe6>Dh{JEIiVj-T+7G$Kkqd2#ah$f@n zTbG8(e% zv+T~(Q;Q$}{I{#s;L2S!i>cx6eguBS1PXs_|F*t3SA5DyEv+6jG5elWcsUIPGQ`BR z$g(Jrv4UGfMw))J;uJeQuCF2R;36^;)aYjh3GIKAz~GVwDo~F;TaRgCo7Z{c>O4;4 zh-EX*P-WMW!Gla20-wj(O~RM*Cg08zI^6AKThqKFb3i^~|3q>>SOy1N$o{u=1#{)@;we@{S%4fp)p45>fEFRrno<1k?c3H?*PxGJuYjI6t%{`)NWXLm zCP$B3NXrgOL9z>(y13`Rh|S6GB(C{OXH9h3jUag&FAOp8&HMFC!#`$e=d-6BX9C(T zvz6>(QnQC(4Wpz;7VqK{ei#5nY;2$jODB$pdSZ~O#){>|zwP|P6aHE9eA7brX-Pj0 zQon$n0?^r|o@EsmNEQ zl#hn-6U}*|9}XX7y=1|VXqsFB8{rs*0bs2mVA#j5{Ub)u{RX}5`00u99*r7@g%HnC zHha+&3nfS(l2zuiPn#r&zUT8+u2kN2-d_ z72k0Q8a-^t+F>O-6kW&{`a-^M zg^%;~@SnQ0)!>j86!v3p8R<8Vp=FS@Z)HsI5ao#;cMi=2sgyGr$p3;T^iUd_`HOpx z0iGalWcGQ{*6_?P@m!}W*oHH!;vvo;RGOmP)wVYyaCCZCKpnfT`X|Sk?9G+Q{)%wP znB{NvgEqz#RIV*?M;-q(ql|u{A>r?@Wk|x7(k_jH6Hep63Y90-{z{CTm5$yj!iTS+ z6ht{hm!-6;EDtQRgZ=qKDu?^oOG?9#Z**kYC@|Im(H; z?vVg2=h7yEU%^%x%Bix8HZDK!DN;WSGMx%C%_`H#Z8=H5WrD+DR`>F10->L`npgG$yF7amaJC$77JT4r4Cc5+4(`TPT+H z+oBzoUCm)`MGeS*kJ@PxLBqRHeym~boRFzk@J`BlL5jrVgaI})6!GDG$n5oh8Q~sd z8C6dY))_@N7vhRGQ^d2Z$IwsX%{a`}S;LrF-egROHxZlNRJ3%?d&o}En<_U>Jc85k7l z(VLLV{zmpM(>3)(}?N~kan|F1+QH2Od)!qLoAa0Mg z9E#y&`kRnmI&H7zz;B6&o2@qZ={;uysj9sBJCBZpdBtZ6kD8AcjC*=_ZT)vXttXC7 zjcnd-y#gS0#UIZ-BJdLosPHRwbYPqp#7{LGZ?dzOPKM~a*s+orpg0aZX}idJ@9hvdgmz-2rt6Oe*T#-@zZSR3#IX}vJq;`?e-PfFdJQ;|D(0MsAVkrTKWpH34pDMQf~jJ#FlgId7?s)k-Dyg82l-dJ6$UsR}Z^72LT5M&Z~&M#SUU zpP>k4LP&)ntviv~r|e$uV_qvh&N!atLj&d5qMBn7C-;-K__NJCWaY>`LU(_4#TK1K zqoM<$6d3V7*RDgRAic@5;58CUSCHBt1$B zO-D4!*mE-TPV2M4Lb{yBkf0JT!#9mrF$+#%zd!gFEcTnnL)F(2oD^^(J#wDc7rsPC zyjrROJCIFfd2~H`>D-iAXJ97rLDoytq4x2MHM*}eE8~dZoY;-P-jHX(iF*fLaG`>K z*1-arKl5z*+QqU@r`>v>k~UVUOaC{SC|9Cz0NY$^mZq&@p+l@v#Gnl_34VVi!jMLt z#cbC2DRUsMqB;rKL=?wDTEnz?Z`%d|AHBVtH1m?Mn}nf5oRGS)YCrf zj?VEc0gkoQT-*#6nBZBtufH86s*_c5z@h(NOCB5ic7hJ86Y!-3qT_2c6>tHRLHUxy z&*lueAvU3*7)TlAVVpHy!fzFSqYh69uEN;{ez~ti@L`A0`7=-)O&*O{PdP+!h5~(lwM{-kz;u1@Owu=b!}}6 z)f(`@96FajHJE8QBl}v{FCL;)ohg&bGhSYp1R~XIdw#%P=lp?zt9a@M?bbv))5H3Y z3pFkMrDlc`RQe*8Flhf0sOumySAeaK5qJqyl3(Gjf6M&hV}3oc597(kYJ!08T|DQx z`mrFAfQN9xneVgs{rjI243eBqMvfptjI06pFm)Qz*)BtQjL;!C#cjDK8t8CJ<{@49 zoJjbg?Rf${O2c`MLY#uMJq%#pt+A|znpXr3h1NutJG#k}*qv5Zf)^IOZ z)4{kAndYDIV+Jd`ZSIOca@@5WDuv?wJ32p&9n7W^uWqSXiaVdM+ic?YB~DkorB7IQbA%Rpc%M1Pgn7?FC$mPuuS%Tg zPlmeb*il9ka)5?Qd$wN}vBr`2o?iyG7yoMn3Z!S3tJI8ML4WG6{5tl{>l*8LF6kne z$iPU$dwZmoLXa@}QZLtcg{{i#7yE&te}8T29r=wpoDPba+B>yF^<2SEG z7BjmNLDN6L2jmJl403+0H(70lV%GWW)Ire6%9wKsWlYIUyte)gs+7lvlPsy0I^%0f zrm*`1LHq8pHq+J=k>sW>+7RmpS=paX88%;=YM>MI70Xvjc-Zd<#viJhr`tvKcneoq z)n>Y|3t1Bum|xh9k%n|9eNzy?-(BSy>^(xUT34Z;)1Qtmzo=*Fq^4UoE>GD?CifRKy|DuTEgJ0+gQu@)=Y~Tu3d)JO2Zw0 z`_*mCi}1wsfy1Jrv2e|XB}o|~xRK?+nqfxVt%V4&^EmUqj; zz|M{>c4s@en0W(e(5Dov4??k0pE^=cNv@%CVOVK*X`Mv&n}ciNRZ*4PiPFc&@4D6F z@}cATVn42=L4R(nXm&K9A2dEzsEt$!`2$s(SCn{aPxX8gC=P3eipXDOc$Ow=o-*TD5+AX&- z-@^!TWuJx#oRs>CMqrULCSioe6p)&cxT{eR&d2?c7t>w=oV#N4gx7c`uBuGVC%}t4 z8weDyVX&j|Qk-_8M#ulDtAdi)d_y~(ugAaRukQ*4Uo!T{wZvwrM$0n_i-JwP(?2ac z#Z1`47}?`+g%$zGj20TcM1em}MJlmfs5w`hg@<=L&}FeMn<^_6O7D(NrQb&ij`dpm)n_IVht&q?l2`6XU9mjSOw7U9?!$3YcZJ<& zgsv9yP0&>=um!()WlsstFBG$Z_prqm>E|Mo{qpHY#;1?bm`OmwKqOY`?|-uD49{fO zn3*7kXI~Iyz2`g;IR5lMRMgmF3PM_r`uk-cRTH*M>HaS+{7*Ss<%J?hi#y}bC8u}Y zlHd#Gm?0`by`0*t%<_D~4hD`ydSzw7&Y5)Fe@;&Ky+ONAUtCj^{Z8N(zVwsndi{ah}VcLmoGnzmWhG2*| zw)$+Dd(*Tw+POqU%ma%HGDAAPw6FzUw9Yxm;XP~_JrV9HE|SLGao=Yd?vHD4_=T*9 z5M0;SBjv`AT-@o8j@};vHq}0O1bM@UJ zH2K7NTJF9qA2E8vR)>6YH*2p*7(b`1tLeqy7Nwtj`yzs|_V7c_JOj$a3Ax_{=W`FJ zroqFjY$;tmAmi?P{5gBkm*osv*W-c8U_DthTQN`_c$OAlkFBk5zGEIaxcJeNPepZN zX4$*T)F7HNP2_)MODB+TvCu6GfZ_#gqb|p6L~_GWjWm17T#M9jA@T2LLj)Dje@!y| zcm%+=`a6Hj^3MVzHW-p<)}sqQ)|`&*;+R*2bzH*3zryLlT%mrDu9;Gh2rRXE14|N3 z<^?3Q44Y6fm}>qx+6kjY-HJH&c+T&Xq{e<+-~R5q%>1@87_qXT5N*LrNHO7MSZd9i zrEyhSdAKeJHGqi*%v!2dJKg>!@d>f7IdVK&>{O0LB|9gm?&57zy4H6xGu1`tUjvk* zQDCpJQr=}_6a&vCAXX^nrcCnFjc2Jt#1DI+4gg$Qjd5d95AU-K!~!G5QCVf&&MN;2 zj?ga<$fJ=2+t>u}=m(67TH;?A$Mhq!GbbE+QD-L1z2N4^kNK@^1F4pj|#|Ex;YgEGI z_f;}fGAihyH@fJ>8VAbY;t zqtUb5+Zs|)u%zwevqIF*&axJC)%?mislouzg~T4Ibyj_dRCyjfMEj%Ubg9T509Wqqp@K zG7kkna~LGQV#&Hwtk~BO z($)Sv5Gf1WI2~!$VXMfPwc%v%EFX1zBqEQD&Kl10Apr08H zsO>PvJmPF5_{$sZPPARu?o2MqC%4sZl8-SEuV;=A<2*`xAu_`2J0|;XkJ#M#sfU}@ zP;785!AafbX|mX3^Ch}bONFJD$JCmb-z)|`dWwovSQKdVrv>psjZ#S%)m-J29di7< zP+If^JCIlHO#A6ak%_q)@ExiYFn)7j_ConmjGT8JJ|houyL6C_W>eZCd-}uW>6K9P zj4l@cpKRcl0q)$_Do_bB_+Md7rn_P4;VbguD8_dm?tN~2jKLPJ2} z)&KZ!{BOkl5aalum}(s?G4sq_`?Lt-m8Lkq75bhuXCLTFRV*q~!I%&6E-f9Uv2mp*|J#pMrT&bmgrXYX4 z6ZGI-bbI>hq~_Ocs!N^0e61A$@LKa5vt8BOPAlVB&1hj-E!gbyzkrs1 zX1FL`$k)JMgw_X}bf#Wwb292RCuZ}VDRSt|2v8Vtj=bF)%k~GsNR7EGy%bF=6Tk-!Pyala_|r=tReDcU+<*8+*T~%N`1%oAD0$(_uo>)JcuJ@I=_$yg$?>}lyS*NI5H$qVr7 zRKxn|P4Umipp6H?qHYI#YTI|A!Df!cP&NZrb&K<8s>wnU-$!Ex?5X(SCY62 zgD7SG4(C`;vB_Ch`M!h-vvXLeS`ucQ?>8#^xdeotkz}rXpL1x>hFN`7%z6-OS{Hxl zw9RXHoHu7j3uIoOPqgL6mXNUq^;344RBKr2!_aaNt*asPE1A;n4{sU`JK_n)RG1Sj zL~SPu70-C<8cuko{*igce{=pm;%vRXNUUrdgB+zY@mh5xKkt0OZE-~J~cMT?|unhU*J`hkU_S23Wu zuacZN0eF9JMC;lxQj1U3QYZK_RtZT$BlW{@o#&aszue14f#9L^_S;1Jh1%{uI|a>} z*oxVDdfUpm4d=d~!*A#Aok7jO>DX;y%{`rYKd!7UWE$axBu|BA}%pJtzm5c zzs#II4|*Q(j#?$CCFveJS!(J@(PzE%Fjwe`Gmu(W3`R7`ix_CCaw!_vJMo@s4w3x9 zbpkhq=(i>fH!WL9xADFbc@A)<(o&6wl7b)j%UKTP1&q9m(Jz}l_#B^e5zANDr#MbO zv(VIjRA^#c5Gve{%KCWZ_WEL?+K@xVmy&*-tBVtBVDMAL2?xM0dj-H}LDqv?(XGt$ z%_HXK4 zHqrbY5531uqG9yn5yKX%6!2Y;vaw%nQXrbl=b>dE!Is#U1_4n+;>gE)jhLVuc4H-V zxBiHYx={8fWp`Jn23)m#0i2PHS61_|Pp{%;7D6FxY9!bSf&f56d(D0vD3w@tvb13Q zuQr4il!qtVdMx z{2IRI*QyO z{>!6jFAQgLIAhFMj4==Ve!D`4bJ;FCaJ5HAf;jltcJywNvB>2eEn{Rh^q+5YsDUwb zNz}pq>v82}fIwIC`3WP-*CNfFg%s>7y_9J8(P;hA-pNoV;?=Wsziq78<{RO0d1X(E z$@|7HX+-zTRqD!QnSCcPz2>M^$9m9PC4dtf#wbIroI;RHmQW3TS8EjD>v11O6eX=< zIA)w-|44Xx9`^8DBv=iDz6ECq_|$pp-*)wDJbW?u zKG8tkoE1WWAes@!HwwFss7EIK5&E1Gnr{SA<=mNc*oXcd0o!S^GRkt{fzs<{N)6@J zmb9$1nhYO~_NBAQ7}WjpygHEdRTYKZqI)|a84pd7~?F5=@k&~;uzM~D(fu9pDM zXH7WuqL+qRiy6z9AKJ&8u$nUWE#weeLlSl>c4eTQ;u1E8l5-l#0ybNP0Z$fY{Ft$~ zAe1S55$2`f23WVwm*zf7;#l(=$zVaLrKbr&cd(sSbA z-GSIPUy|#!%Qz?>KPW)T&eclpMnyHb=Df)Ly^3qI(=?PYI+BUIx>cukOJ#j;du@v} zhVAj0DJ9pRuts+54;)Xu*Yce9lHr2jVB50QLZ}4X2*QMTcTZ2 z$3z_XtEB=7mkC|Wtz(H{g>^_apCSP8-jvDPS1K0Lq)x~o5>6Mg@P^X;Wea~l9mIhJ z!CeCWsIRj_RYq{?4M?e`9?17rY-qv=Ze&k%e0V>N+r6b+(NwKzeyg)tx0L& zSnD+o_1$RuGo<%~;LGd7dy)0WSto;4ZXDFt=5n+w%*z(bL`^5hvB=Pa8Ti>iLJWLt zU;?O7op0o(^s$!?5f(D0{APHMkd2S zd~dQyL$`m{Pu`rTOsN_rLPAj6rTCnVUNYcVsm{+gzvj|a(_9tQa+bZ-pieY zSs-4i%$eio1$99qZJ?<9%Y0ZKrdDc@-ncR1|6(M!RHIlH}50`rn|!Cm5H6RJ5;im}j!b+b$>Ke#cNd zZH=v?EGYDjy8xX{lHrwDwYt`hnI{vwov+JwBQJIVUI?D|&oQruFCX}7K;bxQ4NZ!8 zTP-F@ftWl>$pBn+4j1XeB`j^fHqmuW{N*D=C7zlKO~T7~n%VjIrz(VG?n z{6uBGu$36W`>35Rd$6%tdT$ZM9SD!Cw0CDu?7uMV=On%wQ%kp&U9gREt^eG(FS;`W zQo6>=Iq4ONLF<{4r5m1NJVXfpU z>I(nvdiWt29zk&_k1%Z{g#n20I+1aMZ_N->&5m}Zk@7q(a+wk2+TfQHvMFcS(cAp7c_GxbtiYlqDvZFV^DA_w$_iQ>mu5k6AiTn9Myk+v@%IjB#1h8Q2L7*1ii+< zm}be?jie+CC`fOae|i?%KQqX$vz4x!UfoA8!zeh7YJ-(y4>k2V^5fzoCA;&@&E)LJ zkT31*m%hw>|AFJhA2*Z@-UIoNWcNDReo+}Q%RiS*C0t^y7B<1`eu-nooTP5iD590e zXMwYhocN)I7$i`klj)c74S7WKdH_xjS+;L+;?Y#3UyyK=o=G|O`tpJ)TKxRfP6$MX z=jvJueoYg2=@w&D{BS&&@KIQQ?6*rf7;nXNo-eE!=e}(Tr|-JZgb4V9w~LSQ{WaOr z@4VcC$i)}pT8e))Ij@E32s}(K*Rb}}tUW>Omag6%+4dbDoc>1g<(029ZN|V`ak;k> z=I4TMQX+p=c+{Oy-VLNnj^~~ni?Eyz_rGRcPP)EUTb}6lC7478BmK3*>&Hx}LQ$s& zj?YfgWXzfTaT#oXOfwF}h_MDqczoK8aYO!1A0mz?3!t2w#8Z!IJyBgMV?hYrD_>~aUXWn| zN{#-BRF^ErX9GQs2xlj3;H5S4jd|qE3&>R$0aziO^ibHKbzSnR1}HYhcCW!y+Nr>v zzd|}r2_Ep7+Y`eoL)M!m{TUZM-IhV$$*Vz*gW?&*3(zHlns76No_0zN0Xx|m64w7~ z@C-L2?r);gehuwgzneZuP-W1?=%Wt>crjwnf=b6WjZ!WGcC;COa{PGh26*yMKU~#^gqlC8yDft*9_#8+GY5XIzzyZ*foyZ?0MwjA`c8N);)p#5P5R#Zu-9xki7 zF5BG`qeS5Eg6(v-X@3PaTL$#s#t}l?3q8`6DLKhnb_(#oS7mCX^pO?sq)PTk^f8ak z^!ef?bE)UHUI_3YM_H*6Kg+D~j94YP{3-qWyZ$%3z(t&tv&>|AI32aw3HuKeWa7M9 z%_uo&+#l3Uqe9J}i=-S*eIgfmj1P}F78YZZU5)-&>UI?2EOI_8mU5ZUgwbdmJic~cgXe?|!>SrCxAe?!!q zr_wD8W4nKCW5TvGdgT)Rm1jBy=Nl?(cXQrZax)BeixTJRI>Ta8|3cBysgvyB{Go8Q z@p9Zky#^6pS*ITF3(*zt&&l@l2S#J6Usy}DFvRT_^v|v0fV<-|sQ5N- zqCYVhLOQ89P+ijrf zXl%(SiXPTj0-}P0;;v{D_tPp*R+lr~Oir>{R(gY*kOid~?0jT*1xPYHAq35lF_KxI zCNY2|)t`+PTGG3N3*D1~3*6iX3Q}pYKBS*>V5tOjb&F=O_sl)v=Tj!;q%lAtGnBbh zo}?T8-}~=*MrF+EcF{OM7GljXS2lA-s)9ajwP}r>{Q6JdrrOXqW7Gze?JV=b(;h_` zDYT-3e%0Z@TPETnAu-~*Hi5c`sMp!K6 z@}P_(O)8;GqVx!0x5QEY-o-1b*E&jbk_kFWhG~8y@R;ni(=Z;DMZF`uScGk9bJ`-> zR%--La?CNprQqC=9wN?dUt+S{GU3|PLO0M!e}Joouc7$M^Z!4{8|3qWfhc1Dcx-J+ zEVR+>+;xZ~-J;-dbc-9-oz9)uEUk#Jn&9h!_`sGWYl3YgcuvZvIHZ~Whr)~Tuq^DQ zC@0EyyuIb^WiY6+?mF{p==bspQ9#D$D$hr~w{(_HyXmF{#ag16I2?1iKh@wG2sBz2 z1LnO^lT9zPP>B7D(0XA|*g!$^do7^ynH(b5HrsKU%^tgq6D8@jj*TF=7@|+CB)m4r zp!6PKGuJ^7z4x--xyGdQ<$p91dj_n15TKAruHHjmW05VnKjHC+dUu(O4qsGZadPExJx@1V z)^>Y7I7+Sn4?upT2@6R_6VwG}v2AvL0SirN#)u$P)`Fjh*lpYY+4&$m!3{<5h-a_) zj@LW5sqs1FITPjl^rIKs%wJUvzGa9$EV1EsrMfSohl>Z=tm> zJnzNj&4tXob(@{EOh+U#NJZs|Y0{z6stVtgYOWRd8{{^wYN$XFF4^HX1?>R*bNhsu zDXHloScq6Vv(^JcflQt0cF&K)PuXRlvElRG-&n^i4wXrJP{3ONMq)xB9)fJ{=<(&C z>9Yea)KC5Q5U;(=PVF3l*zCCqPsf93w~>;r+<2(nJT_T;#8Tz>Mb=jbC8H9P?2%0- z1(kE_HX6Og(CdSmLRezZ66b6y7Qi-qRuWx)wDoZq=&WcGsJ%aEr2O`q8Z27m)5!b9 zsSE_hkO%%S$=N*a_2HT?6@!il=alLe?Rr&CsE0*-5|*??9Otva?_E^Ged7{_@*J)8 z%uR2@0C>-Ng2`RWP2z-1)i`&f5~!N&Z<&-7N2m{dO2Bu|enjJ>`>aoL?LwKu*-wL5nC03z#D zON9^++t6YyMQ(CSz>pM(B}=iy!Y?i~5|>hLgxecfJa;X6nxU5M6eL``!AR|K%_sIj z9H7>`zAUkE^d%sE0=?&3T_KOBS4M?Oeh-IDcp&k%TQ1#x=G$z12y35zV6_y;UoE)h zjV}whoK;lb;_F7n7I}l)ovLf!Sz5v$$3v8OlA=x={rY8=i4w>fBvhg=T1m26YYod3;U!t>{I z*YBohmqt~i8L5~dUhw);zozb1;4V%LA{)T}P8rbX=JEeNp3 zcsSg=!kIA83lNzGZix!NY$F=#?bC%4xpcO37t$|e)sHn$*jGCo;%q3RckmGr%APvP zdQ*P1*qpMf1Dm_DZ6;uq6Lr)YM$v%0EeP|05dhs&EAM|^qKk^6%Pop-5a(^2qj+T= zBp#7aH|rNtsgd`425^)L>6-2#vaTD7MMeC};gObDD(; zfj`7N)gARj;x6@x$G&`+%82OLbCG9BjQ3te5yX#Oecku@a1^lI`Yf+p=uPA`NSlB zN7wuMu+=UK^C?-P7j~*IS1ph{((l;tXYhr5qs%5!{>Y>(BY4Y5I}Oo&zWs~FgftThp`8>{a&QX~4|-1^iJbYb z3NH2EJqBhtWmiVb}iHSWi8i1DS~58lJxP(&e-BiN=|1relAU=o508pa-jpM3o_dfPI(O94uj#MSP9$)$KR2_qi!5go zE6_`1d4=|+ATboc{(*$gY>^P8c-q~>YrFJHfJ^9uM01SG?NN=>GkQTfTd_!Kp17Dn$v4)H_37$x$b58FwXnsZ z5ARNed)hlY?5=goE$u3=%p0E583R3) zWe_hGQR>`@mCBY)SHL1*I%+v$g4p@KByT5>d8Z35`~PB>Zgo0>%j=Nw%d%+OZ!Adn zzb*Wylz{3)tMaKM4Kwkr%xT7r9(dfbNhuGpNP+?y5MuDb5yuD@mlD(~#FXNebvdi0 z6@V^qRi??5SSK(@IB&C(?)Ugq)v`;M(kW(?S{EmU*vH)D9W_WA@TdE&?W%xfIAQ#m zl&LjxjVM`TBCgBvBcyEIWmf3NzZ!jJFmQ><5#L0G(=0F{VTF><4I8I(U|4zl zs&Iy0+W+I~Eg0h3wxv;AgS$8G9^Bm$+@XU9cXxMhT!OnpfZ*;PT!Om>C-~Rdcb{|b zd;g%<>N#uGs8OSaBRDCt^ZXmK?B_y3WT=-mBQ*vTKR0*spzDtBCbe_#zVi=>&e0d( z-d7J4f@ykbDk(8d4~Up(ifC6GuwjM*T8PNM_`z|bPw$pdPOUb`#;|jd4&W~;i~v%J zmj;*OBd=#g$)MH1L8g@r-k(27*~IDN)OtjX*hw7SGoY7!wFXL^WS0EzOZ)rDuNKV< zL~OZ`oIt2RQX63T;tscI4@;*j`?t}g41-zbm4H2q{}YHhyEui+_osg|HMTTY?vH>! zV7p9S7;<=bzuh3MGq&C7JxK(FGb%?fCJ5=dV~Vxy>u~^W?j#G>RFlN$28$WkGs}Jp zv*#Q%5l5`4lyXmTU^D>(3=1j<(Bf8aLK!jcvC?|WbU;}a9-LZnpY-ifgLJrt4bW;voZ}>mwUWmrv z)Mpx#`8L1Vq|rl($PJ4^79~5Xfg-@|<0Oz*R4l@+#digLFFsfN?l4v|(>?&agI)r? z$xn&9>)M(V3#l!z>a7nhyS)DPjn};dbLq|fA$E4^BzGm-b5l+cv<$VvY}w>Bd;fq! z5NwLS_>DMFRPoDB1n!6cz*vW}p0S)Bx;Qp<;2RgrC_IgXptUd@YyP1rwzn+ACnV#A z5Xs6$+N^B_BI=X5;n!04P-;M}WQzzLf;w|I>uIbx2cf?l93a)@Qifuu6~CROPS+`v z7ClfjDsL{O)KWP6P#u97uj8i~wVTSeRm@=z;i>t{phJ{}MXjUtm>7s53oBDi|$0&H6P%D$%Oo#zggCE#?H`Jufuv&IhF zZhu$Aq>1$`<ukTTf6+2{f1bT{1}zS+*^i2NzHSCorNv(oHsG)6 z=7*SJLb9C^dBox`c$Q!p{)qMQ66vOyY-c9$O*WAX!cUo7H8{z%z5JXZJJqi z9&Pb?+pyy{yZ-p@$c1FCJYy|r!JM6pVgXl?Xs~EF5ZuSh(+3WEs)o}CCxRjrMgeh) zE7z}e3hQkzf(9%+0{)l>$4mN+>M!~KFN+t&aKO+mG{!TyM>_5QVHC-BpS>>pxRdcg zG^}BuGPcRJ)%_6lBUVQ_uNPJLRcD)%{W4wLdR$(CS(kluf`Qiq&=YyDp7i?edZAg+ z_pLx(iBKYT^k9EE!W?EqqqV~Q3(rWvm{FYnXlKdu90L9|wt+Y@W&ODGs(SlEwywI} z2iNHFp4443^`i#ckfK2;>~7ISo42XE?mg~Mdsmpv+U#x zrUwebgigU2_|&lazOx{l_Bujx;>ksw%u+$ZKEAF}tdnssZ4k-sjhpZKWdXaH{^Z>r zKqNy&d;5EtkQs2Wyx+=fcX+@Nl9R$nWkDg`yH)O&y7z85X}uiSn+fe>oWs9O0aK=v z@`&M7OCWtH3Gp$ z3C~BMl<(R2io|L<65hP;W`iVm6dhmC2PdH8X z5ltC_NC>KczzP~jw&qmv4B8>*v&_w*5D|2?>c>ku@0K(T5SE;5!1!Dth3bclO4stS z)Q{qkZ5q$bS4?fOj_(%5ZQZb8qgKtx*N734vZ;HR5@-CLu4H3D1LD|jX|p_~S_ejIwG0v4S5PUlwdL-Q@@)_h!`Nh<$?5tbPxMXUpb zdnU9F$}j*S&a7~y|2Hgxc9ov))rWof$8I}9eR_NGJmNy*B~%OFdSaYcvs(QTwgmWLv*;MS6edNi+*Cg>Z5Fp9c0lLRC$T;$x_{5QD~* z?e97v&nI;XbhE0?$05oNt?-6s>4jRWRp^DdgT(%`&YNbi4Y9qsY2hMD(C=3HZGPHM z!|!#;Pij%C{g2<|W~3ZlLlFs)-%sX9DN4+ksN(0Wuo(4DYczKH2aXY$=mYnRe#a_u zXDW2vNlOYmoOU0XS9frHiLqF0G?P}%qDE8kTM$h3w{Jjl%bC^FbznaKKivG^>!S|| zVGAz!yYc71Mp$V_^Q|4aA@A;H7z66OsNSZYz85N49eIftn`I!+R`` z09HA&FnFOR5wmnw@v!#t*=O?9=IK)W@9X}Xxl}2Ur$g!HXeO_9&jV9GPU6$umTfim z_vWjiNA#6Q#zrYG42Ik<92yR*-%yygV8~vly=VG7cEA4K1Fx6D>Hhxq{{5lEAl1gx zb(z)f7vEu(r`OHQdPmp=E=mhVM6dW4aTF*>sBN)Sk>&9(Z#yDA;|4E-2Iu=tHg$E> zfib3|0q9o0W{|mVZ}XlzFL=5MzjRiNRZNWW-AwVo{r06$9Ecq8=KMX`+A#IIvnqnf zamZKk)8WhgifY+36*~6zo|w}TuXF+8dt)fTaA z#MYHf>lUg^f5IH_>kebY-Hz_O_+fu>sf(d9dIg7Y(q}x4k>j=hM)^^mYd}_T&CeIu z_MoOg%{f)}?6T@UqqI(ua|Qc}w2g^JR*3zll{SA=8I&0wrbkqjg^lpnCPCij49%*~ z@~#)+PGtdta$irJq61C+vrJo#GWT(I^97T>WPa5ym*?&DcWr>U5~B1&q?+Q#zIz(! zCO|V)Bao{MqlR6dMm>emx5y91nJQCAePU&pRa&OpX0);oHTlwIDe_2aOpN)w{J?1{ zJ0S{sIY4Qp`Iyg4!R&aAY1#`o&mvE)_)uDEx-*+nRD=PJiCUS(#l|Feh*$}ydCK9w zkIhSYTO5{I=jeh5fiA{THA@+%l8bl^aDU{0&@{;QUaiu%S44M^$uLN@s`XCC$KLnH zuSJG7?H?N+hh=}v2O4W72lP?6cdMKeANIOj3pNNpZSo^n43zt@_`<;*DLl8Lm4otF zY2a*h-aF_AYg%vp%GrW5SGqo|B$f^8KM&o^_kMCK&qg)75o8sZVuvxk(Gk#rS;jS; z?{niP%=?}7dSt3#<@^Id8-@uA7cx2kya|Y06chFRPnV3fv8#=LfDuq-Ud=VV80M26j`}+b-L`5;$XAc`F?5Iz0TVGHXSnh-^nV>R|J?TV% zWx=@4zbAz^%$}i>El{Z?vPQZvwjP4m=c{G)o}SH>3tl*!Oyp{Peu-_LvxA@1&;^EA zKZe=|3$+2_VSe%8DNjx5tbSttgB{CNHFXWz0aGkQ9D7UePlSyh6ZY29*Wr(`Tb=}N zJZwrKQcT_`ioXUumnY@O6KE|ns91m{6YP!d()$YV4QnPc01^_x`zqV8e_m(V5Hmk@KyvX?IS$z;MsA)G+=W%m2dRud`3FUZ|=tZyf^<4hwgjgc$J z9G(}dn2iweSAFSII~eth-}W3uXEEPmO{mc;V;b7=xfnAWsQYKHMg5LS$GAx~DK8|X z%!r*Uu~e&+XjRmGj|}anc&W_e&BG;0+?yyD?8dBCas*gU(`v~qA$@qmJsp;3HIbh2 zk3yZb<1F&0Ppjbz)E^4fJ)=w$-o}I@fb#XolAR7}L2%~2E~v+;6~OZ>e3$bO4v|j! zjj^oj6#rM*#k^FG2?ic3j#sjYNMz%v=8NEF8-Hsd@|vSv6I(H9^7_i79Rh=Zq~vx) z3RTIo)z}X*PzZ?fuyxMzV!9XTMGj<@R#}?;V@Ee>s>XPTLSksBO0WqzGfJ=N^Mky5 zl8Jk&{hqdiqn73RZ6*G3ovySc#V+jU?ksS%j3kY$*H(!}P-^=Ds%v7*6dh8r7uX3o z!E$h{)bnM7!p}X`_`n;64zaKN7H{DgxhLwuPhB)VC}R=}&`bFCd;<%P;^khJ-3TrC z5c6$cYu058zLT|_cTu>7I>6aNmxTRo za@2Q#zNxh(`AT))%X8~ScN(eU!evi+=q{Qh6QmXu)Uj}N-{Np_|d{S0-2j8uD8nL81m z9F5WHg(8FP)EBN}Ogb`qxws<`)>~8o*tv?HaqN7aKm9{WM{Q-uEcHStvaWJ1Y;E+U z{DWhyDke0CxKKo+eAx`WJsBKr?N5S8bHYjes%hk{=E=DEa!3O)pL(HcdpLd8>1%hw83jn_-mIOp^1d$dt6Z(g# zfL@~*&Qj!J5^-+sND0w^-xXF&0*cWlB|1)0+*p40MVkQpwP9Y~_GGTV$hL)$(4kJ; zp!^l0&&He+DEY9}`ppKZ!*=gC$&f=Be;PO=C~91amk-Zsbw67Qk^-t&L2JsWl)X7+c! z-kR9TPnne#ZGQsTMDSgB+nfG@vuCM1Hg?8~yBpXe19?Q52b;>xD{)q=fFd0@Di>-Ck@CwdfO-$u21#IwxzN$ zHng8YLX5O%kXAI_qe8ppf4QQ!7ewM>X21AkL6N0~51pt~v)BvG`;zP_R#@On7f%O3 zAl{I2rRG;L>*xNq^=Rm6yUJk~t&bj#3>CoHFCB|JHRLH+n{h%XOBhx5)z`D6tDs6| zgxlaP*5w-FMzMTQL<)p>C7Mj@)H?3Nw=G9oIUl9yIknkUA{~DsV;y};BU7-)8=OvJ zE}vZg_xR?(!IaR@a10P(>iyZUS?w#7QC3jgJgo(`LN(Q>bGbPMSS|SxtMY5Nt`Yyx zi~@!osuk;xRcWuQ6;XUIDm-`Ja&+<^Z)sZ z5|cx@DDou#V55kcm}MXPG)3i#Gy9;{q?^lrnD`eNqCDQ|50ll1%Yyl}MSE6{#RGD= zkK`k-fD1kbpA`oLvY|dMQY))T+_wYbL9ezL%Vbd6N~-c+^`cXr$9(2zVffBRw&sTy z_i5wDk!F5L4Q;w8%4st^2L#UjY*sle&a6;w+3vf>ZI4jMs6v847VIf-Mui^od1j)c-YQqNkmY|^%wLxEml7=VKuMM5q+?E!!Vyfy zM^G%_H6}fEIoD$_|7BvLQ9bja;~cKXR$hP&y>>xHweB~xz0mi|Xo0C#b>eb7s(HW5 zKuWL}WNA}%fgo@u3Q75eO2U#+(@Hf6X^~g+TTT)`c6no>;LZmks=2^AW^v~7y;8+z z;p9G_gruQ*#Dunz`V&`UDxCmH^q$8Ds+gjHyL0%4olO;ILna9bmGnM!G5KQhB=c~M z8LQ9Pk>ylP6gqH|T`s!)^_O`kbFXw7pVjtZ-8OU$Fgid3_sfW42MDV~JDQ8wy}djU zK)rXL^Xsn1W95 z@JPPPrI>iMUN=_m72kH-E>p4HF}{VY-pid87;Cix20fuEU0GLGNHazr$wnw4+ABi4 z!9?a zxfZOU+9I710e=Q{6uEPw5e#i1N337&Zyk@u1>gZOq}g@%JNPnjHcAhOoU_M z5OjxjsDrQLMhsRQlMf(pYQ-G?;rjof;HU(UT(Z_Nj||d^PpFxbD0VHn^~kj13~uxb zH8y4aPlJ4w@^%x|0epmbzxfvx@t}OvV#C`CgSMTGS(or{1jF?nt}LC~MC*#{5wfAc zVxen$tZlEln{gbGufjD%WXId0uG0n6ndt>d$AhUIOZKYLsnTo~H`ihCramt<7n5v; z_r;O3&jD;K^;J*w&5z624e$sBVOG+r7aaRHGY1s2o3c;pR~(y`Wh|EZ#Iza?S+zyq z&v8He3mX4cLXgot802|iPmr9YVs<9hp85?3ZFPJZKDW+aqC+okcyTS|Xr~WyMI91b zkrsu~@jj@Sc8M=A+6p8- z%3+lxmd&=oa{PqAOxnCGLJOsEw`w%PC%Z${B~mS`Qh=Z?$m^X#*SWgZI9_jIiT1y1 z)3=jmWc9KJxUVCc?tIVr)$&L5g6UVv!`lhphi3k3EPp(zVV%j4*>b}z+!Ovm%d)mX zE8{3OOo9E`k?{516>IE2Hup4 z;bYJriznffh&&2OM$j*q6v_(D1!V(WvS-rdDH_ITY#}w=oosu$TPd00cOKFA8!Nzl zfS?2;9ubEIFv0HuMB(0A7VHm!siI0TANb0&BXe7C~@0{*Kd=0 zPS1ty94g^E{BK^$qtV%};|fO3yYd>32C7UqT(WW1uMBoP!4 ztl#^f)0HcoiL8*JiQ~El^`}SpMSGtKv)19FcovL}(KznVX`t zFUrMqOeci~ZuqB-YH*wSjOi}%#Jq71pq73EY5v<7W}?}Ix|l79BSMtw;nNl?On>wM z5-m+jh7tVvvjrZgw&EB-09pEQG;KcdsA zo{4+t6A|n-!P!_2M1%}=NJRmK$bVMw=C;+t^`xE)rRE!834HChJ|o9Q7U80TgvU=7 z?U@~V0%7ws=NnVm)qSP1OUZ-y)2gtUl>7X~T5e)n%VaP7C)vk-+gx-@67Y8Zu)*cF zZ4;7@FyI?HX>_lQ$Sh8{-s_D=3LgAP0`*|KOoO-%cgK+AZEG@qD zu;HsD`_=pMC~@9egC2UgzmDaSXH<f6JE1FRUk!tqmS9j|O@MEpQnj~oSS z>n4(b_KsWjuQuNs{#U{N3$z|y=m5v@!4nM!a;=nukdjtVNpz%#GuLb>Jb|9tUyekF zaVchZG|4ynUd}qwa1j2^=aYP zCqxdp9+Tz#>%{iOR0d+#M%X)ot5P>izuU+DRYj=oM};M}0-WNb0I@CCyaeb1sV@YP zrjhrbtKe!heA(zuT8tk{|(lcSkY3+y{NaFVehvs<4VVp_VClCIr zEH6ig`@)gT^^eNuCw|!SovTSiX&0lRE8zg>r|ZQZ+@2~vb6{RHRvHoGm`A9d@pU?P6N9=( zAXqClf=f^bsg7R^ftj`9F1Z0AH!up|d{p&l;n!R-mf4RI5Dp#QEU*)jvOR5!wnr0_foOu|*cP(S*QM2QMx> z1-aY}J5wwQ5IdR5jl-U@B%c+FnMFZ+E+H37r2yh?swdr7PHy>8wt@CY6mZ<$Wh`ov1ONpXzou; z3H8%jil%=?n;77{^YKlYCu&ZcO^@c^5AZT8RFd);p{q?xoPA(r8f)OPS(m= zD3Nq+h=Pk|+dR3!wf)Z<+lG6}Q&PRE@En~ry)F2|J8=P+@SBtGj9KqTCQhbrDUsx3Rs5(CK?N;}YTh zL76P%aE9P;HjhuX|1bbZ(E+k_O=U)FUXskbxXo;z-5*Qjtyqo_@x0613IEwTu6!Y% z=k58SX+8%azG?<4vMRzqJVW6c@d>1MyBo3-?0RQ#2J}bhN`^jEi!0WT^U%^@(+>v=(orVD@7;t8L0fKz+UO%OGG&KjQ|{@4E74;Qi;ARy-75FhzvoH4)f9OTDV%RAzzc~lQ#hZ|dEj4!_Xvpr*aBuXYu zJ@Rvp1l#ChR*KKxw$UpKF&v*BA?9b0T_vR5+EmK>1g}heZ3I4GgjYF$AyB#4c4vnx z!}CDn^+xswRSPwYK%+I86~C=NbTDW|jG5udMOa?zk^xvH``ScLKQ(kF_^Is=fXbs1 z1Jd+|IO7i3M;I*(m)Z;BXpx&88k<{VFSoo{P#Rt6zM7MB-xald5FU{tdv>z>Dv{Ia zq-Qn<)w&y7TQhd*-0_{c!b@ zCz!m5(3Nybt_#hK1_kZOpj9(*w0xAL7;#JTkYLP}MOKL=ySR*|D0b8XKfBf&E)bT% ze(FrIAU%%rWHAtw&goB&_~kmkBSd5U7;V#tfNgB@Es^$}8bf-WJYSmLEwCu8Em$OZvs&2&j8OTcIM?u2SWVt&aKkBil++=NA9*syZ7(7)x{jtL0ZprPQ`deJ8Ez@1~mL*$J1IUh?Qu>uMz6_*eu z=iOTM+a@%&eY=Waqk2vs$RO2N$h-#J57xN?&z3( zxnEDAas~WfhMjJB{BP8Iddy3%#wx;Q*iqQ#J;sRJcMrn|{4FHCcP1#en1xfT{hwHi zZl<2iyJ}=qdeJwHIw$83#h6!vJi7-j<4%o{U&5G~q93;G0mmVJYZ3x8`*IQ?2L+8Q1@+}9S$-*AeQYspfp%Rm89IIZpYpl+VlD6CkRBj(9Gc8=A(x=gmz+tgxYwp?&DNB z453JdO1u6IOR}BHk*-u0#>%F4%*|heW`|uJRCQpNopv+*MJQSDUOwc+4D#A#7q$wT zP8oIsNmhWBS|vqyiCLl=+4f79Zc60uEI9gcg$NY&^Am}tN*7^j9jZut0jghgO*LbL z2s{?5&|bB=Nf1_2jpZx!IWK-z&tbGiMgd`87IFJB6e_EaOepQyhw#cY8pPzqJds@os4M_05&o2_ zh*0&_`eba!h}SY?zE$uxNpI1{p~Zj1$G}}Ot$_vpN>eBQMC?*B#=CqF!{*ppvrkO&AWv7Iie(382Kz}j zm^>2xuKp?S!SF5?1*EA*wyBmUuUrPJM|)=->UD(gFR|v5)`(#(4=Jn4XAJg%>PNEh z-*Y}67{ptLdBG{)5)MavCR;+S1rpX~?6}aA+klpvJZdFW7^N;MwmPh!lM9t6frr`s zwA*OM3C~auwDgCXMy!TAg(0US5-Tajcx!6z47dST=sI_-FP{+1mRa+rQgp@2Z`m!- zJgb8pyiiVzsTM3vx(Rhs!XF`_ZPZ0f37HdYe|M}Wq=^VsCr#L_EDo-4(AYc7(Wt=e zKZs4anoFM(EK7v{_dL;EazwrmD&NBV#Aq$y{=G8W&l%Q5 z&C2-a_VAh@bUa;03s}vUV2XVY?JoH!7P51{LbL>r*^cjN+_agsfMkk&T1&q=jNXuhqqOj?el zd4+lrM!R}|vGP%B^*G!e~966lRCKnU~EpLCmyJV=tgUckAx&G`ufVCY@QC$NKUhNlBm#1>x#ew#V`|5Tl zBWnqM<%()WG}38d7Pdc4d7G~qIY-#gUv!do>MAq#kliE{&g8c|E8K}=WN<}KZf9z% zW8+0fJX@?*E6Y{U7RP$O|O}`TMb!)lY!GfCF$l(P=#fVm>+{I5fwCVBmo1z zeD!Hu{vs?_?rJoaYBpow0PKO7jqnKhhWrHj96Y=YbV7*inG@p(yu6toWp%9pB2L{6 zUXu2--~IRNa)TF7%aetM8I4YAIQL6`tT*LFY#Etw|qRzcTw2+0M)F zMH=w3F>*Nu3Lddu8FoQ*G?O7)Ua>oVUx)FXr)~bK+^n?6Q&y_0`*yY5oRsKkA+%ZD zj}GUzfs`YRfx!z~yBrC>wt^v?(F^;114f}~j>S-`Waa9z`G^JKoB0he-8dxk^gV89 z`xBGPuElZ%KB-v-50q7lrZ8g0(13|h6xI$^x94ejQZCJRcG5~X>|%XUf?a_64nvPz zZ5TJ5q0}=J9Cwy7IKQ}yY{#oS4^2^{xhOg#68D1K%P;UXX3TLny#ijnwh<>ULMkY; znR`~RZAsOW>V@$me(iS06(OQ%<`ztqJjURZh#pcln>f+EV;pOg5bJ-k02pkL(8`*R zJ+V`yKPj4Y&p0dGy!k#l{{CUwfJ$Lz1J=$Bq=_InWj4t?;znb=zHu^f7-wXV#hNz9 zi!7%OeOCu~MVai(HYmLW5vX^ZAKz_bq z5c~>N1dv9|vJw*Kj9m~Qr^$w6VWsXl7OSIHKQ5T6^+>2*^HT4o7+Lux9s!86Cv(T) zV^%?D%jCeO_;~YpcApgmBpqt^67m_oF}p~x!9phWJ=d!pU-w3X1GS}378VaTG&G~S zo_w3WbBxX(QKKP1D=O_6Bb=jZhBe8KjRMMqR@XpljDDKU5LWElR=!T<=fO`~mo?zG z1v_Vm;h{QyCwi<7Z5rth537953jYVCm+_!HhGpm5)S+M<+TpsxE9L;Me4#GJsNV&M z4cE06y``?NP!EIAg-7`H@r@`+7)h}Ys>pVIY)w3M7RiJYyzSriwHMY0+E3j5PIE`# z-=FPcaO{_Qi03%M)(hBwAAR#z?g0N?LiXXB@Z@*5B#e2KDMaaMJL2CLhopP9c`Sy~ zpw&wvUGk;jm#lg(A@K!*)$n^?e$hIP6Dw}FngUf&kuO=f@37umXs|wXf5s@IV#;w9 z^Tt6I8j|6)|1s$)NhwCMY!{Cb_(2XCEj36^I9|ahF1N0U!!fS9h?^2C5E+j}+c*s| zV*USEyZ-4a@WVt&eX2On^ft6(A#t9p9WG=H`n&?>bae=4k)i|@!Z>u-7S7wr7jZ{0 zwGlvVYK1J_eakdHQWCqFX0FHn2CvuBVj&Z@e(*BPM?zu^Sc#^WN3uo zthK%TkMbsKwj*)zS#;xxyq5_RdFBsslB>kstf-^FVH+ynk)bf`XCk?&PH?wyAz3>y zo44Q_M3HH2vat#bb>Xr-)fhS$(Y{hz!h?z#mMThR!Xdx|O)gD8zd4to`VVO;H67coT)# z23uVvwm5okzOj}Kxx=l3^yU(6U9sxie5k@G2K*=-p!?Ps{!V!IY^*t%QDuX{5QBjo z%R(kCN(G*WVHj3eVb`3`1uu~qBO|{YCgNZQ<_fV^I$IfY^V*f6*IssFxpa;rI>)- zl8OosTRj6ViAv%%CPW{`;El>S`WMcZWmYn&2jyqWgL^`PSr9GKspMFsLKZ?`n_0&t zP56OssOizKh`*TR2Zv91-j6A3xl&0ikG&K4N9mX@>B3y(a7`_Iec5%b*V5MPOP4A$ zuty7uA6+nGu1$s2r=^?9Hf7ESe9D&xBTFcVme`DO^7K3gK>I~O@l8om5Vi$!!L2nU z)vlv(Bu&4+*?i84`sKJnb#A8VQR|M)c_w#$`fLaiVd#GdFyZ`V=M{bYJ@7FMdgSY{ zCTUjUzmW_YLM?-eirlF;FiDEa7!w3Z;z( zcu?$PfO@S;C*$hF_9x7e-O3$nb%SeGgkViof0PGz3u%V90IIRLPwR(Khh6p zG(_$w4SRG^4P+Op+^gq?OD0Vc?Nz#xILVlz?1BapF5AeHa;L4mWCi43!zi#!}A<* znh{|zt{lSw;qn>U^W@Ty058 z_>WvK6Vvl+pV8^!>5eK5%4hVEC+bcheZsCZI1o~LKls4J zV%0R!0Bm-a_POkzbJo%R&39X!w{=*{uJ56}9D8dDH~0B7#+V@K$+T^#D!QX!m&!?4 z?;?btK}VuU-s?rD-jzLC7mf+ zFh}ZIi6B*`xz!rAEVaNAn5oixP9KCsGWc5k%~!TuDlkDMfpmnP9x^t{kyKV7*v8mO zp5;qxc6QXrXHxZ(fywh=2Q6>Ox ztvxwQaFHs5&S9BYqh{7acLByB5Q)wosBWPGCh#OUpU z87%g6#X?U7M95dhD%quu>EEV-59S1r2@waIl+`z<^XOPlK7ysA#&zKZIcSEg^GO=s(cRh5U?-7k7kN4R4WR|?is#qjqg5VT9hev zw$m!xq|m8+b(oK@(ezQQ3|bYg+Y;3Zjg*B!b!nA9+dTF*r`-|gw|A4P7Gx)XP!qL})^Q;P=F-+9gMx3Zo{sWAdDtUAhklz|fgiatEo{&{ zFt>EE-zK5saJ6SYSeoVh>4~X|Z0Geos_*as2S0hBW~cy0Pr6N;isMYjSSw_N7OplABq1XuaT20XI04il2!<>vXZ=EzS7W$u_4x_J!)Gn^JtC~ zwoUj&0>1;GUk6zA6O~?f@b>R-L|D>B`4k zv3c$hOp$q~tP7Hf8nM%rRPeKCI=H=>xDjtt!A6KpS!s=hL8&-CK}D$YOze;z`$a4y z_Uj6b7RoZ!HMo%t#JUAi_OC*@nUQYRbdT$5j@1ZXPhwthow2K<}n5u73do==N zMFbAnij~z=h_P+K1jXp6B#?*!fF&4wEVL|#OIDY>oCQ@=5;E86EQfEkzhnKC;lL*M1>@YOd}Ze)`ymC1`Do_bUqW+b78r&J!q=?U?*iYRuy2vW&N zW;&Z%gxCzp;?T{tbsu-Dy#1?l%QRJ(D|q!(k=#O&R;QGm3n+y(`WvP0hVs}^6THVcv+pXw6nWPq2hl^YubkiG!(5ZSzEU*Cl0h2#OKW-=_BAgdJ<3?tbl#heA#UPHBFE5>^V24e* zWshme4Hf>biwGVu_r9P0`5m9xv@2esd^YpnYmki!BOAi5l}n>Cq{T{Em=bTAJ41z- z_(5`-w!cL*Lp8KdK&Tq|kl*>vm-1+4J)$Z_SmdNhx53G&XgTFzC1|HlY*gtn$VR-k zH~yRYGeG)EZx6GCC&N9+3A>K$c62o=zOD3f;{5_ zAbrP<_0G|?G%1`nS7Myj%73SV)&8XXkx)8)9IM5JAhJ;I*ZY}>Y1(6Mbh9a|mSPCB-2TW9TezvsKo`L6MI{+e@)8a1o#=c#ldpX`!FbRkWG zU6BsLXZfK9t%z<1a)BN4OR@>oRS=g=ef5L!9|`~pb<0=vWfF#pZ?t9=!ja<=_ZHvX zaEWZ8yT?-+{Bc(h`=M2x?TVHf2%bZ?krj>`$!Ij)qd(^=8~Lov#hW(JZtvtIMa-{` z3w8Z8w3HO7cERI7pF`?pHtwn!*^cbhG(P0D;$v407z^P4WKHvFCeI)}%KCbHb>4_n zhmQl#;5wM<$|*%kSOY+|Jsz+P&)Q8c;PK(SN9=LI_vKi$EN~i!9ltTouje~>2x<1rSIIu$v@f;wfV?>$Wivd%%BMThveXth%>KuJUSW^^< zRO{Xcm{44+5^gHX$d)Ezantj#5kG}$zRwEn_b`&ApyRuCkh9&S*mQwW$4nxL=RV{9 z<@?^{aIxN-6o;aV3AoAUi(bVD`Z8G!BrVWDoP{26c7nj}Su9O4ytq6af?y1R_?b{T zVAx5-6$!G^TDU)A)v4(Wf+P1Asmu%ndMXT!^f|Uqlaq8?S3K7NR>^`38TmY?c|Cck zPLLzpvrJI)2UjMkOPWeYZ5e@mBr>$(w%7MY)be2WL4->Y4*msz-b}HAkM!`LR~D@m zs8Uo5fw(FHZY>tn*VPi2LP=GWesP;1%L1;V8NuiDqp?y$oST{5wGv^l`I`1pPR~1+ z2ZVT(Lf?+NWFlx34#Vivytx2;HRn*59WTEkaU&3V-6cWkjHvGLr#<}FAL?6DwnPsGM#WR*g~Q8!Txp>|Fx2_Fwwov zk{j)PJ=!J%EtVWXnixodh_4*YhFw!WS7PXC4PNA6a1w@V-d-}qpR=!ZPci(-=XN`t zza`v#f3(Pk822osr%cTWjntx@ze;vM`CMI&w!Hia3hDo%j zfKC+;RE&R_gnF|UsifVR}cLUsFNsDx2KgKjPQJw8&JMYi(G z*Gj_L2#YwL0<6ZT4M~>hQpE-}Z#0bvl%(O3nSK2)VXA$u$RkE8n(6|>$^|n+_Yd3o zBJQW_kAZah`6~z5m~C4XqNfL#)uJ+eUmmwBx#h8?x_aC%8sxl%ytf+U^7xLM-dPu` z!b(?MKKYc7GVFCF^OzEdEc9rNt^??Qh_td^%?`C7xWF4Bwc0u7ZrCpca%SN)@p2fa z9cQ@L^Gf*7+T`xWE$t7T_ZlT&V6d?A2HAZ$$HxD}r~9f=50^;^akiR0dAeWUhKx%Ni=iHK>7JtZ21oQEBLH3hx`>Ss1a7~<&e*sqX!Z>NzMDzxR-KwR{R zpsvBWZKIes_d2fid|Pj-#my79Y9q4>{e?TY zsDo@|6%nd+%-zLT)kNsd;)Z$Rwd}S(-*W~8gZMZp`=jfl!=nv3!CJJmZ$wf0fg~r6h~^Ixbxu6(AWX^ z4zA4+z;|~u#W*K>$k5o+%%uCu#ot1vS{Ksqwn^e~54bz9CS9}!s-O}KxUYDR*IGqt z^4x%_fkJ;&q6?A~4u#y84m=`%VxdH3$FNo`Jv>d5Grb{go8LB+{y)OTe}^_;NuJk} zg103`FTvKF&ybsgHM0RJ=)C@8E*6)|IPF#^DNB3FX;DK}btwtHmnWK42yyxCoT4j89muCR>L_;Tq0xk#)~n{D+Nb}?p#ZQ!6bxG@*$JLFpJVj9qGDV zeRnv+fro&`KYUBFgzlx1$}PGH2%$PF&AloF0ezGy@+}eAhAbnHaY499Eb$T$sBLSb z!Q?dRwNS*FHN@s3Y0T;Z9ia3ud}~F3p~_n`eEiK;X27qXwNNSoFa-uu5Sc8E zDw&Nao~;~*zu!-r2bW;Y1bVY#B%HsEQy;o%hj-n0}eH2vhn zg1btK>!?~%r##FCt_q$B`6ayWF=O#Lk$klc5s&37qFb{5_irgpRNvB(Ww&I`T)=|+ zk&p9>FS6?4;)x1i&`DB0o>=^>m%Y4nZyt8UfY4Aa#Q2+0+9^JE&EMmBW-6oTyrx`y z+u(_m_vu{|4UhJ^0LnxZvk%aGUMYH@2oz;!#rYI=+~<%y3)2V@BkyBo`^yZXP!hvz zm3Al1;lY7`a`>Q&2d?ldE+VZAE_iq;*kpmHRvb$oK126PEaLGD%akV)s%|C4Ewb)RNY z_5-pa1b@suChU%G9vFd+J|Q2(82e^?Qni=x3pJ(; z&Rd=|4%0VCEGmMQP-j*ry1`@rO;`sq9i|D6`?<4LTy4P6?&Zc{+1E$MVr_yL zs7i|vB7w4@1iGk_oaD&eCF{MoW8IE-1RZjYU~7d7ISZf^9yKVrF|IIo`g4Bdwm{{; zXG#9qe-mEPnIitv1tjf$j3mZvF^&T&83EH?HY>bT0Nw)ytr3p1s68E1H-y9jb{T!<(kst zq2n>cnw4_0QjVq65(vA$g2BUB^2Rt?>X25^GywV}4j)!mZ(cu#cw3<3wT?Yc#xYBQ zNZHC1pgv|5M5qqg4Sf1Bs-=sqYEiYW)M)%>TARHr72l)9^V3cc=;vF)5J#LPW(F@G zLiDX#M?0DECj(wpGmq6osKZS; z`Xv86f=jh1&Oqw$S9#ANSN6bHnSI>^n;4fGB^rGJs+QPcgSBoV~_Za31 z0Ye-TPbOX0@YvtLIkML&IJhg#^|zC$OYhtLj z@~kzqDmN{_wX}F%M{BqmLC2_phkV78(jO0nZVmednH&cMQ*V1uap$r2QeaBE+{mB* zb`d~W{jmTjYLc@qw9or0ZNlvlm61NVpQE`y&pgl z^PGOzMtT!9!+vuddDABLmF-<701mB`foROCTY*FFFS$jzU2MmR`qEEY;I@{!7KWrA zWz@gPC9k%t^4V_%i%CjgwsDK~#y|3b4>X#!4*G#rc4qG`z{kSM2L_iF^Zcvk%VcmN zY)B_bS=sDuhyJ*M!ue?^W}Bp2SLGM7s20*q?AojDS<8uU48|gbm2OW(GZte|*jI!D z;g!M4Ac?V4tBnK7v4{Mw?HlzEk+o47lq09LgDF`a4T_D3~lpX-p#r9_{PVyn9L zYv&I3hnl1Q!NsnsEx6~b8qVIePWTPlO_Xb*tpS%C0d1>HW|6x5VE*n`E5v?0vCRbs zBrX+CD-~19l7hzz)J@~}i8=o5k};RLT?;4-FgQDVhU`_$Q_Z{6 z`C&Rm95@#mWZ5DU1F(>@4X{Q~zIMKsr5j=-kW6PJH@s3YuRmM1zlpNmMZ(Y=c&6Pw zg`@RfWP?Ejh*8MplKznTZ{7L7K#N~J$h5Y<%8vv=UeBIYFUE`jR*UyUz+Oh|k5?xj3*6%+(!HWUmPUg4G@xW+>O zmK_4Ys^Ym=^lRi_dX++*^&EudY&>Lm`O*aw^l}O*=5%|9zwN{|M>BrxR9YZlj{sDs z2Y)>a(vHkHqXx_pK3QVW1q4Y56OwW zsj507lJVewjYWx$g2Me{ei3^$C@A;*iWmLJ z63=829U&2MTvvwZ?|0eZ%Stw505py%9pL zx9D&Fj{a17SVREDbIDgwJCPPx_(*_$vcxFLBCDN?f;y2(UyRv?F!=6w%#rF&-y)+7 zYcl|0*p~BaQQU+p-*0J3=H|D?giuTF@MvtHQpkLRg|?LG5)IhLJ}+EGU~CFv|A_F$ z?LWfxw>6gmm@n5C4lCdEM3Xtlgh`aPU1ReULP@<%36;>SfBZMGs6m8?GgNy~Xy+xb zEk|;U^R70#Q7OE@>_FMC2c@kb5^p*?--|{%4m0nz%%}&!X}mp|0vN~kJ<{{1A`bjd z=LL<;CFxM?=bQAd6t&t|sL@;-j1wJq|2V#;#fO8e7~`~px~2MZU%MK$+_~?Oa|3s_ z+FfPCl3%2(8nuKRap28%(-j94krx%lSyslEO|t#==$y7W<5!|BzgrqwbNxias8k-D z`62L^JWZY;X}Ki32;Luq)FtT}iN|k1hNJp-<+Amq{L9QCa;r`Lm-Gr&TTR3m>>w|z zuIr(RsbGE$D34}cm(Q;X&isp)BMaRJwxA?56^_B{s))hsoj`G|7s7|Mn}d`@YZgwQ zVa#o@TZ({ZN2#{~Imy;xQoc`q@A-+B7Ria@mq=uT%AGoao<`K-0{Tsbki&sR!{A^r zPv@h9;yc1%MsZ{5cm(+P`WG9WK8P=e8PiKXXt=7YAx^wkQayxb)Cz{aK3dbhVuROeo1%gaZ2``EZLXR(Q(bik_f>aUPGVcE&bdhUWet~JY5+#jaCYR4q4o1V zi>_G+4A9^tARG1WE_aQCc(`M6p|_Y0GRdN|(LAZO0huyv1Q!>`kprd{vAjsX;?X`S zA-dJJtuUv{CH0Zm#cpN_o$~kMw6VvDHcw%K``vb6=?zx=(!4q*NyJ3@;IAqi43mjn zAioGHlti0U0|5V!o4Lv)-@oo25 zdboL%8#D9XzZ`x+`7B%GM}bfUW=g+jcX&!n>a{giS5TXonLNw{QR;-0vDhDNZ1;1K zy4YdRHx63zZivdgjL|+D2wdfbK$OfhmlKuZmm|R3Yrq#VB1btXQwznC`zwB7fuVn$ zOme*0@> zH;nSxokD5<16@dg%8%EZ#!zhv7UWrmi#UAfuPI;HS~(XYPBZf}jgk}U@o&?S_Jt)j ztZecr1`8*Dd!?{;z98?{r)FF>suqikJvuWbs&bSB2IbfJ8ypEok4joL~+SN zs*_eYf0el^$~1X*TmB-fHm2YRzqQW9tVDJETsmFpM*S6N=^w4lvhkjAT~rWAqujYd z3z`q!gPpjID6-A^D_+8tQl;906O(MP{Z%`ef-eT{xwv9VWvY~Hh9n)({Y2W-gT3d_ z91TfDb?+%BHoO(X)lrDKt)!T;!uM6Mnc}@r)#7-kcx;ch7(r++RX++@f=@HC!8UDT z6IF=|AAkJ7z-o*iK?}T|XdHqsyzuw^g5e9M(dT6xD}ta+3M40--IJZkw$}QCF%@?= zFbkm0FcK;ZkvO>6ZDS$Yl_HlR_&HNF-FT*?yp2DCDqWh11it~3OwzH4iry~mWZ%Uw ziEk76){YjYbC)3B57;YRjj9)uz3_(Tm8Jpq^kCuLV(v!WM6!%M|GC6qq_eW5K&yXw z3(S1o!D@K{-SAF6r}3D%%{X{31eX|PN+-C3*fywJcmiYJlemIs!|lqlu}(tY50y!o za%;sa_~(N&Qb{nh6C8Kcn+dL&;N8Tz&;D{>hDMoIKd7#ABHl*}mA8^<6ZaC%_G$6fB7c&IB@hjjz?oOmt2U&^`8v zJDIk>pV8lB=V6NZCtK0jeaEj1#O=E)^)N<1qoSq~IC3;W-=)A4QU|PgkN%*UL=e}| zdmm)q2T4!}wFJ)>vSRk`FWYG!aj;1bY&XMTwtpo@$NdpE;p9r473bT3pZ%v_cC|+? z|Mrxfc(JC6MYH1pI9HOm9TCw_i7yuvyM0ISMuu^qCrS7~mX)og=Ybdibarv3Mlc-z zo__C54QJT(oNz-KNx7}~MCOt&ULU7(uH@5BT;25gwOCfos~N9l=gAJQ)ZgZEUsDCk z^JAJJ?3zBGX-_M_@As{QJ=S7mPWH&fJ@dvRuFMAhbr|4E`qpg~S%-#Z@b#ejco*?`8KzsHAgo4IQ;&0T{?w(IsGMzzWqV z*?+)WmY7e^Og`0aAu}!6or92*P$o{PH8@ zav^X?&e<>9n?NH7wm=Y*H{`54u#Hq zHk9a@FZ3`*LB}%|okM;*Sg~zPZf21#z)=$meV0Rpvz0N%QK!!l{@8)3j|SmW#x4W{JIIfK=d*+y1pqiR3hpN z?p^u03VktrO=4SD%%i|^#R}I^nCYJ;@%zbELfRi1P8m2#T`M&+LQRs;rxUy1fQ32? z-uR@4w`=*GlW1@aS?O%$TQr_e(>s_~w{kuOZGjuv`D0}HTy^f2N@jE@zjN;`L8@_q zFwcp=8ixXEf4Gj{Hi0r!UiA;wjBgNG-wtPu$cAGFil}EzDpXPY(b`lq1bRcGHd(kd ze-!TF)E@E5G8d1kKN~8?AYTMYJ`dnt)kzB4gkVgjeDMNlL}Kr!U@llJU} z(La8Cm!O_6v%OTN;JrHN*K(F?oeQn+mkhGYJedxcd@|P&#R;ao@nJ3nZ-i_4-nUJH*q5Tb3n}o50g2>bEbmFPS z4wBtYu3GjZS&f*y>AKh&86hs|g zhOFI_WyWZdhFN6zmED&!@-lv(C6mn=x5@FdR3jQSPH%7QY@tQLNlA>?0322Ew{@@g z*Znk98RY}Ng3>f|D6e27x@Dn zlQT!irI~IAB<)7uDoHNbJ-&in-+*kfbt)&@^g7geo_yTE)@A>NH}*NGb-8t^Qz+2b zu#Ve?^f=m=&F=+AlOz!EvlT`qkqTtV{ZD%eG}E*2aWht=%*93wk_&)iHwzgfXPSiE zG)IW@9S1I#zpsi9DV0$^!ODRdh}wr-?Dnml-MXr9ax*&OISye|GjbKJI;v1%$QPwp zyqFHFX;mmk_ApevdGaOQdZVS@ij&)_yw3dyD-7vmO5!{B?;Ju@JX*|6^FE>$ceHhDlf7PCZ-h=32I#Y?Z{DG=*5ysi{H=%_Zlk zXp>jY@!ffu$a66{mrjgAqSSTxgpVtyz7;8NXaKT|EqBJfRaoN9hXyNRV!a|xm#K#{O6@#qteKZOFB3L!uoqQ4LAH)IWsO|LArk% z18{nt?aLDD2W<`Sy$F699Pj1r!Z3QS`4(f((P>clVv{9cpi4S1R2Z@R{msmAq3SKHG)|t=+YAL zISAI#jyBSXq*?07bL0sxc4WvrKXx_IpxCmTbb3qi;d!g@U2oNn0|4do*Dv>&@?GG+ zwQbRZ?vX*JbwU<2-5kY=XW^$O0mFOy@ZZ|D?C%E-_d(;K7mbf^`57>^XMx@}I+=`J zU+p&9>4I=R(J=i^UNDC(he;Ul>13izER9hyLXKW-dqN{>g|1q*Qs=O`n)BidSR3t1 zQfPl8s5(7}OCJwFi{MjY{N1)pP0C$R;dSNYlyO81lgC89X^)mBD8M6;mCw#QO8m*S zg$!@>F6oh%lvSl{jL}~O2U$iAZGQN!9xCYEg@fh>!Lg?DWw{JYve=8J-a@!u0#3{m z1&Rru=-w-KKR8V>UvnbfTA5JEz*2pWH=5pXaW}C>ys+FcdTD@C9G`aQqc95j`Q3zL z&+qN$5c{N^Vr}_OZ|w|CS%k-;>C7+I-8S|84o7f=fR(?nYpQytqXb8@@@GQa$3GPA z3H_{xyhm@$l}3$VYOSs}WJA9G_!W=hEw4e(yg&;K-BKVK+no2$vS8wwKeTKHziK_N zeNq2Zzd}X^bfeO&Gx+ykpl01k6#c@eqTa{{y6zQDFGFavjeES zaS}|D9KbkQzTS^eZX}~(l-;T{PV4{fj2R!{8ye>f;hQc0D> z->P@$Yvw;#1Jeu&DK*HtSw*y7&=;k3AHz1v!%>|tmF_Ad@jvSSiHWrUVHU{M5jB)T$m z-LIiz8@`w|ds6PY;`KM1*{uzi0M~!jFf}S-O4yiYB-O%&x2*h$D_p7E%9!bIK=fB0 z4Esj14c3)pn6eG03rIGmQ{A2!5jCzdUlX{Wb(==cu3i*8^LHeODqsRc7uyhcwb}*F z4?VW&1Z(2FVqlG%H(hQd*?GMpEiVm}T02=ubzZ+T<*fke6|_JQ>w9B&0#=OIDrJZx z*3(-pb-8HcA8fy30zdS*RyvsFLSH#8*T1fpN{AS+oB3u9{$RT%xa-8xFq@mwB8-C* zl6hZZxChe!aIi?XoKcH`LRgcJ*Upr|%7qQ<{r-BiBPNl1IZwIJJmr#{GY+Nei7P9+ zI;W}-Rrd|f#EBdac~5g-Vb8gPYR_)sdS44Y;)&}?VmaE0|PY zt26Pg+j$#_*j`moQaz{lyuapRui8i8yiDYkhkB?XQD8#7p=Dl|8`=WuCS(tgP0dU7 z#C7hT6r6TRa{i4Y1Xmw`h-~$AlrCF81g#bd?|j@t?aJtLx+UbZ;%3J@-{iK*Dl^e_ z?VacyEqW_3(obfWC=abCJ2J@CnQpW;D8%BmD)*i0*(Q`ZbFxxv+JG(CXV`e6{bhmF z5a<27%-9g~)bjhTuzpI@#4df{#K?byJO4wQZkEqs6`X z&pVRQnL}?ue5bGQC6vafaSsXJzSI8QxI-4~jZ6xETCEh7Kac#r)MB+WB36E4l{a`x zynw81%KOgK1w2`Nf6JOVp`Eak0tZC*iR>T{VoFvWGEffjP-L}M!bvd_`>vEhP5eO*57lD#pZ8siR(emTGwh;10Dbwt&wY$ziyxq^jWFX4(1!~$;C7rO zp$(ylB{%a^-!T%%Tnx9^qj*kzLWr35tx`3bH6T6uy6!EtqAU=RuLw^sy}cmFy2^#w zF{#rTPrUm^qdbXkulTqn&%Rx_M_#%>IUMIV6Z*P|xRtn1W@GbJx#+A(ah@ucP=Dg8})7%UvyO{#?THA>;ZewZ$h(n%g~EaZ2Rh4Te} zi7-!6qZ`U!3vRbV`bpn4)~tA6=WI&PBoJC8Ff+YJjjS6mqC|J4?lF-f@Q{%?7SA2g zZFn>KJ&P^W#uwr-R^zS(H336MyErR#zTiXxLaQ-@%AqBnNr<-uAf|+1B#+}{mM^;u za^U5b(8%oj>&Ye3%M$8%w@+~$n?QXRVb8>>2)*qP+oJp%f?`H^`B%4J0ZkAh%hfJh zTC}ZUm77}RNGf2(8)k8v5UzOTg;c=WcWK}u*CE~%Hcc3I(he;dYQWeV<}jk$>A{mz zA&}DHY`T=K)4! z*@9nfgga>2LOE2*APJD=ObO8wq-Ss>3@9+#wzSRZIOhdOS=DyU_#~G{hTEbz+)YYB zCK4c$?Q-c{mY__cl4yI}QR!O+G_@K!v-D|PY9lP4=@W|G=17T)PM4!dm0laU@;Lr{ zc87yZXl>&Veff=>i3#OAGm+T#B+kT=B zG%Xe&D@WqMa5XKkO}v&KKQ)6*TsLynf3_A+wYX{Xn}o~6TKHW|a`Ile-X}JkCnEO^ zIsDmF*LCZfRyU_-faR`>>4SvVn(Y)e1m({%==4a-+Rs>0%ZY#7uIf*Uw)=+a#$35#D;k)k zV3p$1mj-ricv6~jk6zd`(IRk_uK}i-NKQ0P6KA-M?%C(seZOv=Re_-Pm&tPBD`Hyi z>vP`V&bXR=WCwl@J$L+GMdqdV^_^OkfBY2W`lIga|hmzmz5CX-NXUdrwEBxYv*mVLN#7f`Z z_meT*6G)|!{FoYiZ9arXn$$qy;v{{W1;QCI@uHpyM?%x|eZR0fryM>f!JfRtcfvLo za>`_oT>LNY0$fQTPU#ZNA_*9~Sh9TbWM%ASh(9zFKy)*!%fX z@|N0L2Agq8oFBNecsR69=a?$^`=<3=(vxTNygjcTi+RrT20u1Kk1ngxHjjx!yQ zX)%gM|<#@Pm!vSA8 zEY;JrpLG9AA{l2LD{b{nK3SlOMwz@g6RaZg179`x$6Nr&q~8kC03Ua+U*C+Ch%$@Q zM;E~RsvUn&!kv=c0+{+dVH`V%c`tB|M9W3YSap2nz8qT!d7Ldkpag<)mrNZ=Ih551 zkhb8CrlH!pqy{lwwVv9R6Q8Dd1yRCgr9^t)g@E+BAfS~J>!k(UkVw~#9@ji4J4Zrg z-##i>yR^+WY01wz(fmE?rDD%qB9uoa){Jb$ue>Gs(sB_2*wVBuhOF(5`>X^s zx-Lkr+k`)h#VpF!L7IY+!J3QYR7hQStB8)Ds!D`(z=`5L3c6^8bGdTxT3Ev)j(pJE zU`M#x7Y|xjk`keg^R=i$=SIuy_N9w42Do^#7ey4W(ga<}@VP$(+tG`rC??Z`R;r5u zgZAy`;ceG8rs7kuwAse&wexw4fvgo3W`5oDaM65#Q~j9*JchM6{FTCZJ9wkylx8dj~a2 z<(0ptimrA%c_~R$U=M$3+MMJg0euWV^9mV^gS$%uEfIgnNgArEuX9KhCywxsAtkL6 zYyeUKUTo$DYE-+!on{0LJ~#$_EF zy?QnReF1MeCx!13beu(%cD83_D-A6TSqdxD4yAL(bb80T0gaR{l@%3VC;6PecD=%Q^5l3o=OCufvF{=X|L z@X^d1j4VOSP?fy{MpS81aSevl9)L*p(z%Y`cHD+0binW1&Y9E;Odan_Hz#EbzZsFp z_vP2mPv!SLF8ea%_rq8AP1TLF{;wCnf|%=mhWxE&NP^P_1c$Fe!&rRV&2VPr=Gp4A zRS`T^i=fw$_59kxAi4bmD$QmWqsiLhZDZ$?fGReScq-%`Rv>-`|wXs-Ku7P647m~0oJb55oRxxhuJGSGn7@o8DaE%pugC8ex znTrb(wE(?gbdxE=+Q8%cGG7{lpuh0qhDRb72{

    V7UWheaN~QKlikXuZ8T*D|qMPw2K^4~Hr`hF% z|F@B<)0(V4qRqFHj`Z@@HV?4B>tM{WPy!J}BB4}*zT~vX2tMj$+}6*1Xwg?O=ZxHh z2tFJPTs)rYj)B%eUq=b8l^|%~A;4q%|6+o(wSwPi&wuZ7@4QV+(>*oL9(THIdz?_w zsXjToP)lRKm2R0DNk}uhY)COMTL_Wc$4-#LG!a#t{C2qyF?{(HLde)b4|hEx@R!~%3MPTloW z(bQjrr}=;i?>;jguP^KSCjn5Mu=ajzA*|WN4sQTXd9|ZlIjHu zGZQfqd}}tfG;#1-e)-`jKgb>(H5zk^?Y!MZsy<5~s56RqFgj=%=WF21o>Z~Y z{iId8SCYEuUHU?Wve4j)zn*}F&S>`_(D`B_aST(NCnBrbI+_(3fAnf(-N~=I|D5nX zX!;DE^ZCFjFkom;?h#l1Fp$!`90*tdH%h_C%fZN>@}n301)3;;Srb7jF6`@r0V``2 zWC(Ev!8G9PsVd+LSdL+lhto(#?&ZMg5yiJ$gTv0?ZK4h8En+GwfqejcHAlvKn&f|s zpDs&Sb^&L)DPnn`K|!o?2dmI$#oV!j>xVZG=l;Vt-JV)V3l0QV*(I2KG=}(@YF9IC z(hAJPJ{w@p&z?DkI|TYydt?l5Nh|m8Kw%S$rRHogd<##q0yzCGU-WRiJukDa`h$fi zfANKoEa6)GFUBFE@UeeZjHw^B&`LMQ?D>h%PP zNLn26;va)bvjSA5R^J8HL^ak_z#2MF?ZMr`zU|0>F{fObSnGhV=;aJcIE|~>*dBYp z*)!is*kfwUxw(*D#KU2W`Zoo``{_X4#Fij>s)N+akeB}%;0=;LTS@SM=lh7D!t zPic0hu~?!lHFWsB&a$CX@ZQ%{=YC^5LN#XBU5QHz`;H~zZ1{tEQFqc;N9N%?W0`l{ zk%(lggKrSPWRC=nbzup4x_7?=l@Usf2Id`{_ie!+guhLUh>)df6r{n&UP;A;d#^I~`oIy}4(*o~G~ zf2w!hrfeoGdf&l-S-=du2|%lW?ZbeFKuVlVDOG_#TyA-3RJJi)pS|;f0K_!DwcmBaH_i9r1;Cgjl<9_&lr|H-s zUj!6gJ)aE(QqO`};{Xmql$p%h5}uOpJ!!`?rBIwImpLmcc_!YE&Ql|`GreF03u#{$ z1ZFK3r%wq3#+S$jNxqWTly^d`cD*crq9DMXs&if%PnCZ_$vrVv;QZ-XGLa0ltHi)K z@5&alk9hiYm|U_>TKh^VdS3GTlv9H5^ZX%4V1hcmT*uF!Q-x zEWC}#Nx=uh(e(c&X1R%D(Fth|B|z>Q57s?YxDpTnk$Ly`G5oxop9DE<^+dc=9$XxH z5n5h=(cnxGjNlCzHyp)MI1RVVN|^j2HSW!fU?LabZqEU$zo-tbAnJ5lLR0IuPwoKL z>d+qP@Qr@k?N`f9VB(anEaG);C!sn|==rp@q%l9M#_~-6>p3bMR8VaNG4>E|fXD>2 z{#+Ix#knWY$3qr&u`9n@#8@95-L$yJ^61Ov`qH^u)r;ZJ>>d`;chhArX*K<`8Q&6^ z7tihY);U)Xgrxx%OpJvzYi%d%*n#L?3%34Dr5}UPJw-aZ=_W_HI)?zxt{pzwiPEpbV}< zA)dE8cU)WyI85x@*WlT`l^@0+r~%HX@Vt|IssW^g%!U_35kl>Q~LHlku!{S;<0sun`Zt3IeDkyo@f zw-~FF?XPMS9WUc*p(DW?>_!&;%?a-OH;Q zdIQGncv9oJ%ZQd_Xwg%qgb&})Q|uMQ~OMFneCSgdXA z%uC&rE-IAsKFhxu$#K(Ym>+k&og`|3FJ_%s8*t+M+1Tar93Po7G*e)Fw&PZJ^JM8O z@RxXLU#8R#YmiXvfV1tNJL}5alGNw3bC0nFMcW)1x> zs?*)&O?R>QvK&_19p7w@9I6T zSG`%=q$D_7&OTloud2x4vxC84zUDah<Mou4S)kRUesMyijedcid6%~Z2P z54`3SdEt2IJ^XEajeTD&=UM4_4$uyaXW z`d3v4{fJ32-{OeN^qaG12Kt$nUrQNvfiWM%mtT-3!?hx;SQQGz|Ay!hH&d>B=cH(n z$d&ct3NsMOkNF7(4B}pX36<@e+O7$>Trbo*CTmNu@Qu{bsi-Z0@DhP4$F~9;w@_RJ z8Y9aDb`p6eY<#d z^PEXEtX>3#@U6HxFh3byNU+)RbMdCu23IYM&1qlsv*ayt<@90fvB$|b^DLy=&{^R? zx^K3Bv6{?&*3MAJS3&^&9Ev01Dlz7la+P~`j?^2F%sv21?V^X@BgnPXq&WT9peTRV zy2Xv_f;b0b;c3aT}1}G2oZae6`v6=F~5q<>#H5nm7!%zL_TYR0gamXcTNQ*u&+M3Gz zZs6Q6Fe+|c#GtqhUSE@Ujey!yS?sh{4kkn&gzqR@IX(&okGR05devvkUiV+^vTq~fo-k$PFu%(b z49f)GSfOxxl)U^s5d^?t7e^XaSr)bZnei?!nyW`cAKcFeuG!r4vF~buocir|U-WUn z^Lv1%&eE*N7xEJ31RpBhKX9Cl-Kl3=WMwhX($16}X@250iXA1ob3Bx9uEx*%c<7zj zMW&XxY{xt&d#!vYZ{j+ddIQr+Q3N0Cig;zi`m$_~X%9mricRsbT4JkDa!p)+V{R_6 zzABvFON>CQVus9Si)pWc7Z%mxJzSVIPWH!d>cQZ*Z)qxGr8?5Tiw7T{lr>T znW4G@ez^?Bq^%CJkrK)pZ#p$g(gkN0dCZ(5{@we<0_^QibNlbn#e2Tbv+urtSz;6U zy#Icy=7#w@V1?OU1mN=*Yw)RA1N_!_WCoUchrJpzRetcYbJIJSnSx1%FJgG8mp6CS zz%5i+>j)L|o~VNFE6TBh!3KIuIqPZ@ZhdNLV~-yTY?Ao!9#ExAlTr+Ex0DZ5^Qk7AOi$<*Wm8%a(UkO`|hoJ z>Q+~E|2Z{XUHhD~)^F{7$$T4`W|rjk=UXl#Ec(%6^K=TZO|~RhvmjCKvZRzcwbdR%{9S>bEwO^Q&-Bz?$ zR8>!*z)D#-)mz!~r~*|vL?sjcFq^fWWH8_Qz`Qa#YM;0DrZ?Me6{F8K^&nQ4fS|d- z2bxI<&#JG0@W4Amorj#?F>U_5jvt(Vu*KR|7POf(Tl)IKhKexM+i^1dL{NRREmArf z%qaoBtZ0*fBk2*CDf(5AH^ZP45J{!dvpRy`XHs)Fj3gM?&a0mz^lR9bhCxh0kngBMkaEG>ZR@?kE-r* zMj>holkI}HnCP7WjIaT3*xQu!7?Xpi{6Ivuww8l%#3bS`g73I> z#Mcwjj+dGx+sNRI_*6N9j~F#^9?*UCpZnB8ul9ohO@h0wJIW!k3p2lRY5Ga+5jky= z1@mL<(asw1%=IlAp!dcKb?v>nL}<=w1}~2<9WOGTRZT!?M)ia-$ID14nO26vp+Uq9 zdr*rlkC7XOecYhOODpBB2CJU<92)d#+Gs5DZ|EXl@x;fFYKVhO5IY5H(90olcZNs* z>>OqxPLun6Got62K-qdM4I+9w&!hI+(y&k2^i-*e)5>_!AP8Mxn^{CZfd!Svp5vlA zH+tO62VG2&6vDyvH-_J6Y%J+ztZD&?W7o-}*9Dl0qU^%ZtRgEdhJS2dETDWjZm9~z z;*sScUFwy9{6r^~rsY-fwJE?!q&=-t1@4M_LKMbFIw{Z>6e3hL?$Ov)LY>fw#f_Je zPr|D!fxGJlkPg~GXRW9*MJpZ8uOQM7O1bHRybWw!l$O|#4*aw-3SN2);H4qIuWgi9 zcf%MUj%V(!Sim;MO#ca{D(gU$kil*0&cNUP>hI{Rev8nt1Z9fFM1v~D6Vct>pe2`` zf7`uJhX;#_g>d{q@K-LMLtjRs*aoI5uX+-^vEL0$~ zDQ@zb{1I`bN!*h;jRsV_Ykium`lZ}88}<;|0=O?;1genKx7NdQN`e#L!JEcr-KchX z$W7vDC2*59=MK@BZ@GS`KuUS+0iOD5U0|t#AseESpKa_VX?HUI!dMj_3$eqw^qX}Ic>vN7|6z#0)M`uw-CT>%7zyw-^X{OgY*ec%*@2MaCbC-lCCi^CQzRb=43F; z&BVEwJsHB3j}g^7E8jrf-~E3db9dLbHnA85R=?4Ss;3R(LV zofHH|h@<&Q6vpR0gr*u&RMhR5gNaCG-@JT&X8?j}%f8;wTXHCODrC^O$sst$&yRBd zs)xs?13UdMV=qmt)<**0~ZJV45!GzKPzH#8R9*|hngi7$6iM3NJW;{x*P17l)i zzdP4lAFyOcH+bQx^n;}+v|>yQD5fcEysdj?^N`zL>-0|7VQmQ}TkrN;!<@hSe~W(U zSo_l7xOj7~FxsD%Hy7@5l>M+Ed2Z*k*Z(dfH~VY5kB3J@!k=Fl3E8b_lhXGimW4nN z>5KGe%nwDzAgHtW&vt&tJ|6&f2C~H%e&I5|WU!2IpgwP}ST(xp(ib=0-t&$4d(Ppr zW%r2f)Y50aHst?CoXWghK>s-i*~q!DCl0>8Zo=u@j3)Po;O<0hdW2V_9@j|z2?NK`5v?5RgLk=h318r z88Z`|W!)D6rlHUf4L35Kj83O)U|BWTC-#>A&qvcRRddZRUQc8yA~Ty<2c>NfyUZ`T zObx%k$t1({!E>9lbARBqI>{!@b6@zPVgCV-$k5MP$IgJOqn3KV&3)&A_~)JH#A+Tq zL>Z`1MBOaKR`QwE&fVm~?EC2_Gxhmde+ZSI`&RuukVm}PS4SY`v{jv4{mw3Ul`LON zdD5leHYdC6vE6?*79oR?}h+;m5^%N8oUUR-!^E*)EDq)Zn~0TMqP4(%5adKT`96(H}^<=X{3Y zdNPYZWA0*Aq`faZQB8{eQJWz>dM%y~)V%w@`<(yp88Zo>SIyLJjd9om&GM6_cvu0i zLjPy^K*)|uB@8HYrUl`M$u*qAeJP8HvwRZR%GT`{kS4;?By;lqeK!NZSfkZ8%~xXs zJLQ6s0gL?&sfDx6GiiT!%@hA1y=?PTP{BVq6HjG@pN<_mlG52h5Cq2*2la?%Z^_yy zoI2M;c)BtR@XQ4AHm+pf?sx7RIJ4z(G&a(JSe;?!CA&XA1!8HuH-4Q_`oj_GoDEN+ ze0H4=2jC#YL2x__E{C${Wcv;~a!kj|K`AW-zTJ!gy>YRge;Mhy-)viN26*nX+TP_1 zGAaSfs9@EvNoETHO0Q<+`94fL{ZmVFMFBwM1E|cS)aIIOsdjX*0VHpGkp_BvEFxqte!|XB3&v91w&AZLhP)L> zfI2_W!RL0aCO<(-JVIzI!?4$86r8D@<>=5t{Lm}BbbVH>ofwwM1VgEaHldU`X28sH zYMB4QIVr4xw1)`ttH0oe(9w+m1K6y0VJhDQpxzO8HVNfpEvo zbeVe-DB#i1h-3I45sl>%8tm|(@Q1M<100PKybeFLvDtA$eAykb@>ijK&qm+o*2C=+A)= z4Z!k_vhpEzFbPjSx|l3*`6n~5btN#H2MIhs&@PpY=gqTdT~q!?O;Gd2+HMP9hyhn) z&b9E9G2?)RSd(*s&h~VXWa2kdirAlR>8Hkq{sG_!(uG-C3I78@WaSULxHI>bc zF?nyY9u>=K6hmjJJ6Mof9<7q^n5n7KU;ylh38(Ep4=HvI8m6q+t#= z)DHZHMTnTkxxA@eZ1)&(w#bQTkWch>Njy4+tPtLE#6$u#6(n`{iUz5wo?3S+lcz6?Aunwq6%L1NR`3j zF1JIznV}`lal-i9ILv(IibUZqQNU;K&4btV^{pJN$1O9(VH`_&lwMA@J&kYox_taR z0)h6A_)9(U^la?@|BCz(H*sXT93#o6I?;xKs+gmSHI8A33gTJ2e{tEY75yp_0V9~D z4TZk<@vx%$j9_@{&755Z`A92`2Q%9`Wu?wW;LYQ^+vV0;mKoA&eyf}f}EQq8|f9_Xk|Q$ONkgXh&&*E6*+d-NKw>>U{NUBut+VKmSE@ z1st2>IFuJ5qqk}{AUL$%`5vz;c!>7-hs7d2BQd5&p0DI$y7vheV4e(_EbGt0o3&hY zhhBFZq9vLvAs&8`!zVZYYMFhpW&vZI(@KF^n6x*8PI<~I-#6yW5x)8Mk3ulmW4ysi z^!`X?zGsfMe8J~dsIWX=M|Azm}Ap_ zA1r>2zA*j9Ghesb<(Z#s|JS}<(VUP%pnKg+By|S=!r&S0xvc*uw_zy<)is}RNr2+c z^|byxmxwY#gOc@tk=?2gf0>}2Z&4lV-lDSI}9EXTeTeS1=J*2q($u!Qqt!O3-FdD9~x-g{kTR(rEF1CJ1 z3shzSWR<;ft4hs(m6~-IJ~;BI^uu_NMIHtWdaG$mnV(9WqRkC?hu{cy9yS(YRd0{n zdRr)FO7h%~VjwubjZoTcN9T&Vw9$S<&ItvGD~$ggKGLR>ay0*m4PnJTkn!L!7L+El zngRb6Go#~OYgYTtv~+V+5)|@|H-}duhq{Q{9SaBit~2s45U)3D2q0; zn!Me`QO?1c;GJAMO1y>U&$v6NNmLbUYa9ubkoB6Yo6r?Z%f6*+#<4Y+HOn#yE$fkvWLkpM^|kb+KP5PrMa9 zz$BdQ>MSd*8Be<8WPiU&EL3$|8P{&1zUjOdywXGqygQY+C_3@CiAlTY*eSLEz$d-Q zP_oTw@;lQNmOtAG_u=%g?>C$Q_yAqWr;|iW&c6L6tx^VxD7Qxj%NW@kkJNEr)K=X> zA~rngNqO}(4v!-U8y$>7rx1~!p@`wY3wR(?;C0(`-e%ylWPqOO^O%u8^ebWcAc+nC z3Hm26NLqrAD{6>t_Al~6x~jvfC=bspS?`wB7XQSae>GPIPDWH+G^d=e($63;hyNvB zGryl>voXhI-S@fAAigo+z&J&EX7%-6H0dzLTxV@UWTh>t`g^qm;?J|Lxw~B{J{WENkR<;%1Dwp)RVev(%BNjpB{u**vlv0U9(|^lq(oDgQ2o zJ@3mzKM5i1gxKS|`vxlvHV4Jmq-b`p$(IyT)JcSsfk(8PbA108 zQIY*cHG3*x`X|)^+|r2BFsB!Kd2+z7`5FAzq!pf?BiTVMNeC)?uXS@Jxkf0JGh?T6 z5cX>F+lD$Jxu_2E5F80UkZzVTP#AWB)#D6%tnfH~V)b-@bfhic%#q(#K2OWv9tPU# zaxg-&`H&!)F#DDeTCW#M_JD5dw)05v^Fp=%{m6y|vdcOKIys!eKKkroD{nx%nwkvE z4nLWkI9ua6d#Z*RN2V$?QqWO{(0b@POzG~T89?y!5&V%Qx$0G#L^7J$>c+J$4UzX5 z(obdT=xhosZla%>jsmG5r(o$P?R_`~n{y+3clzH`KP)mw+<~pcvYDub+PT%{H8PH525$Lw~r{q7$e zv1sOei@mbl9L14JL18Zf+fN0JAZa3nuhcOd9v9dA6sXqsVbvRqwrKWqmkaD24`*6S+sVFW0TjLUkle@GvYN5$3Lu{Ee89zT8!)$eDi!ea6Qv0wU z`ss+?V*S|!1E4%$_KZ+UomeVY^cZxJiQX<5jWHUXpSYNz5~w1BhrhXWma?B^-dR>!}|Lb)KxB84lsLA*FXzh^;{b2m^>}m1# zD!lA*)&Ig861a07zmR%u@5K*StNu+q+zvNS&o|1ELB-oxp* zw)RWT%4ev_O&?+nnAtj2sCkDuFc8JASWMqJUUuuBe|F&AorJg*V|O@kys?6DYSz(=&y$7yp43bViLy&00FzzXz@LWihs)= zL>^5SM8W%S&=oP~nrv)sAVGZ>K2CXUSG8ghlEKty)^2`3;;Tx7`N!9JAdfrc4V!mE zMp5mKFo}EL2Gq=S8uXjDN(n2yXzvbYG~)tZ&4%62^zY6fy0=oBFOPhyn{3K_^PUF2 ztX1aww=oMoo!Gm*)I~f_qu<2af70GIUNHw5NWE!MiJl4To=%3* zKEc2cKqt+3qt-~I;B}<7yZK?Tzma?D@tJ^`nUgPt5}e_EsNU( zX|N$nSy^NtvEI*m>bRUr;l4$ol&@RE(thw-xdxwqmhK2NIn^=$1?wPqtRtbsC=I+x zy!-F|!xxV)i^(I;7vCQAo|jN=Ma|OFa7}gtd+2><;May9XTnqXwpYD61RRW0yb8#4 ztcz{17CT7`-4d!wTbh_To&TvQ@2JHoA9mS-sSJp1H%ocMbLZ%qhKeS2<7Wcx>{RlS2b>Al&)1O+EyA9Dv~u9N=|ld4#eT#8%OCwO zhvYrTzcLD)yDk(unYu|gYikkvk0~|rGl&2{Uzox6BOwBX#S0gq4A)H~A-S##O+WgG zO5%S)3tpNxn z6Nd!|1Gc;YLUtKKzB)D)ttEj0v5-}6|zCWI1^wl;W&iCG(NS96=qE|XxvgX)nfS;cgflXfk z!51!O8Tr!HoUubT*GnZ3qV8g=wTI0MzV`{usQu$ltonIWd7||0db!FI*%xZ)wxy={ zx~+0<(6_{2Ji8YEb~%XzqSbW@8{&ulxI7klecCshqJ(a5Mtb1-PVPp+{U}>vu+F`3 zCPC=~`hDnumZ824E|ld-6TS5nR}pLD%nrObqm?f@e%LF49SA^X{OpR4!JQ}PgQe+A zd?00nYD9CNwZic#qvgQ_!BKV$7Z(ZxK3HKfQy`rYlFw?-6gQa21(l`{co0?eDTZF0 zLa=dM!I*#R)~>XENPQfQovs=l+2=AqF@X-BVU$Ee3u6m)=a2BVUu}PJn}Tz;=%7Jc zD`FH~0HOUxd^S=<>FC|+Ei~V;_FldfX8f{ZIND*91~t&v(65-)k|X9^2=5v@iGHof zg0EBB-}fw+eMJQA-6!4Ew=TJ+-$AQ2%e4FGXif8hcX`{YGDFzK&ZE|qWQ z?^k6KuMh|hzD19l0@EzPz3Mx(>%yld^Pmrvd#7`&i{R@0s(3eB)qQJ7xyQ#X@yu5q zVgx5YFXFkQ{vI@#|HyVPm;RLgS?z%=Cs;8!-ouf83jj&HeHiVdNnw}nmkm1f=5jN#P?1OA4W7E@hGAF!9!6himyZ%m!X4W&Lw& z&U|yuGQQ|#Q8_M!d0vE<35#d4*O!ZYMQ!BJ`|IxHz>XwXiDU*Gg^s1?y=uvnsyM zo84(HZ80oKsOv^tJpTI9#a~PJs2fX-vXOJimw%tI=zP11Iyb|=0Ig_ptJ~Ee@l$l< zea!w|=8#TV)j;v#Ay&1I8OvNWJHkjojxv%pc3x@nN(V8@r+ zXy#sH_UMux1V!RaxEL$dv?r(~wmWUdX;mKe{KEb|4>Nqu>he9@{}V;8O>u;7VR`R= zXAj;_0Z6}-T}nsarhTvAGbZ0KEmlX737J~2LawY=U@VLHk~2T_n)eevS;h(( zn;Z)oMH6U6@X-^_s)FpdbqcB;u{Vtb7_j-tiTx1R?+qCE=<@eC;FB!S6cHnbnnR7? zyLhi<$=YX>n+>AfOyc;w0r7K_8K(Q~!O|>={ao4`@tv)+ys%Hr0}avG3$8!Me{A`R zxo5tr!z;+_qURsqXPGZOiQDenJoP5@-=L@gks}BmoXyzvQPXB5x9s7+23IaYx3UIQAl;s ze)b6)C0KsV!@EXaI7;0irwT85z|F@a%gN&T*v5FO~Qi3vYXC|fZq29y% zD;Mm?D<7G9F|BWLF}IOW5^Qff^JWfukwOwQOM8nS3~9cs1P*U9$;&pZX?Za0?cTW; zfM#|@&=Wuu(Y#h)qgh1B(NhQGq$N=2chAI)1K)Ge&>}^e6HVxV@HVO-pk57piVUz+ zwv*O}^kmyDerz%ha&lG*zOJiJ2{Telt$8KNv5GPH!M(_Ama8@Ydl;(7jk z$dJp8M;b~*1+x$rNMG^0+6@)i=0QITPd7(_d{*D!Caxv1l;LovhqeZm=HvvPNlBDu zW#6nkN!8p)UUEm_6GkvJro58w!!r$Mr;FP|@AuzFiU%}%B{ddtn_~!$fa*VhfJC2I z8OhW{hxZ4fI~o6eKp`xOSGe)+2UoZ4R>g-)uKWl)=m)jXL4SID$otI1Ys@q1O@sXX zfPAC+A(z29of*?@$UBLmHP|6##x`F(hXE{8 zd!7cy2D?au#zKL*vvC=_yArTblJsJgQQX%T3&^>rg_Jz}+;rA{dVBV7(d{I=t02*A z+J2{B!=Ow#W;CMxsk9=5P*SAc>Msf@b9CPQpVV^?g*QYNt;-%Wu(UCi2N*jbavI5V z0>L}A6@JL?=(MsT*b7N7LJ$_UlW13nB^k?3Y82jS$Yn*M9-g53j$_uDOkZ`%k$vK4f{ zd{C>!z*?RqVXWeO)BnR8>~=PiP3X)5|N2VmNcp(=CyqyZ%y^S8qxddpHU)`fgj+J@ z)QRvll31spOGWG+idcX&*(sy9MOq18cSE?LXguazO6q~3mmb*b1YE=PFOBLa?ME)% z*PJzNp`OrEr)OS_(uOI!^E?n}#DC|5w8W!@&B$W+7hwZ<9DUtqd6MH+Oa^Xfr!TwR zRGeVt1(@=*zG34}UE{O%0{Ev3hf@^W13rR2>y_`WP0^cKDiL&{i1T87kXZ5yhqY6o z%861>NIK%cdT$R!R3N-7@Oz0@j0`c$zy#h^j>$c5{OlFNmm}1?=IgUX-w+Z-^>tBZ z8-W+TpPZ}+^3#NTmgqOd*PzZ7twobAzJwolzRWZzK^w`x)7E4dF;kK5t~CnfM{YgP z2mJU*@Z3K8*{m(`>0lwqEhHxy*pAIJy&zJZQ|1#FZw9~MR{zJRj_r-Py!&7up{2YW(mxN{U&KnJ1?{=Sps^)F?j~-@ zqpzicG(%!cTpogs<-hV7o@r_EK#qc+e0~@y4E56y)RKrSq9H@RAlDD%e+O&16PlgtA-)gMD?y3% z%42DLfCZFR)_)&KOkCnW#`*P!@urSQsQvA&VHbR)@hx@^QqGS{hQX}fd-_Tl8lH+5wUTIr#K@zPau79WEg)cF@8LLYOAChZe zJsH)^zRf6(-+cvT8k(Jgt~&2RP*o-_8-S<_#&)a%`HNH{T$E=~y{}-BeLh^NkZM9? zNUUn!nEb)gIDj|Dz?QsfvfRu12P~v&=flKHc3q78fUrx<1~iE-qMY878Jo+t99xOB z7u;!O!q6c)ZnDbx&^@GeutpBe%Tx(}B%`H?CBa9P8>~k;F_r?8lEG{7$|77TNUEz| zk^TBs_`4fG4lf}Jeq;)bumh70pwi88&>78nG@zcC*@Uib>hYgogahksCBIKFwu?}W z#8Hn7H6dRwsnw8h3!3?xQ;mVw+m$Sb8GeLt{EN(sz&VnwjIYCV2Od3s54N9D`Md{F z$imP#or6lZyY9^nc@hPLX(B0d6a8vc(qGe$iLNpWaQ(Oyu%4vi_nU@I2~0i*r#QA! z@d?q%n_r$Jm(Jav@8h)D7zcEJpFvTkp4+#O22v?KP~ycAgF=KEFi2iAP{|x%vNDzd z3um5uLk+1tB}?Y7_b``~yU*7{jeHN*6so#iXHmST#a*Pu={=(rE&#^${~?e5bDivb zdLiK@5L_a?vvC$xdx1`S)0z2<)T|`!-E||GI6-n-4g0hq-#8W)3Ce!7>xil4$Jw~? zOt`igqn^B4<}rjI8m&PrTQ;`VSnXhRj;!2lPP#x!H7LWV{YN0Y-(St> z#eNR0VOo$vq_{b$QOI$P0pj$rScdnsIbq@Cb0Qj23uCClNrtLV|c6jCS8G7!I4F%U!s&A*v>Sj?eIhXg{ zDA;(}srXi#3USuG!edK_OG{2Yb#McWdLpi@P>$pHLhZiK+|w-yX0MT-ZUqvo{IG9A zhJJPm(%gAD%b0tCkTnm;$`os5A<16d90zFZ>yO+H(|DmamR7{ncRmi5I{n@7_TIn6 zD{`-;>SBExYEnCJ-k{^{jARo#XR_K<$?+rW7#V6vzIq8W0W_4NqxHc{2BcSr!I$~< zk|;Z8y&abYWnV6_6AdK7(~*Wj@)Pu zryD2xXYd+XF$YO>&1V`PN2*fc_kSc~!BCpkVCRGl(GsBOQh8y089j7a^WCRo=pkD@ zjbckZ^3Xc4?6bF-LBPKFpV`4B&$G>nL5bwKF{%qV#zBWNfqpr}>{zpMH>7t)RwH8X z?yTas$?RKySyDK`EK&w4KsOS=r-CAA07mwR!!rXt zu55vtd_PqYQ40$Jlxrd?q!}vP{d(*|hI~D+tyq;5IXky%|4R*n{U`jGi;6VND0x-w z2MSa-YXQmnYgrC9zt%6FGecKw;aUVt-jtr3Sm$)8Ft>5-!fCGT!MId?lu3oxSD zTQK!I%JjS8l?*$@qlwG>nu#uIh<-=Qeaj~Je|rJ6HFMl%2>w0AmmZCdB^PtzfedAf zvyxVdPv~vEe>?#mu*BHT$lY6g?fRR5sK609%m+n(3z0V%0YE(Q%NsDzMA$VcF*iKF zWku7wjoCFfplHw{7*!d0H{<{W;Te)&hb7!f5vmlk{0S^j%lBw8cDNtUHZjOWeVNNU z!Z^fs;u_Ho&D0W%UC_a#W73n7*!~Cc+0sVDGL=Ap+BdvJK{)eD@cy2ooIx~p0-SvSSjV$=~&hY5E3ma&kex_Lh4*9oh!P- z1tS2zdFSA?3xqE{d5+zwY}}c8_HPD1fzyoGK2TpPv%vW80J^F{Rnc91z}c}M^N!Ue zRl{_Ex<$W54jxGPVk~yZq_5Ktx{*6U?zFCVca1*Xov9w+S|w?{1Eigt18q8JEKxlz za*Doi|CTZ_F)TfqEAJoYjx}?bs=Tq11(}Wls8*`VHE?VvRFbjA?6=4bE-*Y9prV%D z=Y*476}^&WzLH%FIO4*t1X2>AtuWwPU~cy0{9D6ia-QKRb|_F~Ydd2vN%W zgM86d8|Clt;#m9OcqY7$fxL+orMFBIL{|D6j_N z(ab~Z1cUDHdnvf2uYcoRC8N5C?N#44z z!md7bx0!FNoz1R@Ci%pdY>P8~{vR^wf9RwbRCIwc1as^5AtKS9>oYf{^43p@(T_n6 z+H=>vFwCeRIRlDdYn9Pym>6=|r%{4-Ih2>;-#IAuj%zO?F`3eZ1v4fejdVCC#Sc`H zp0jIv@Fszk3m}R7CB3Z}fJ{%@i7e90u#(~_B}ew3+E?wmu`XJb0Foo`HQx^Vp7mdwrd0Ypj%-{%&!ugSw4FOoisw(5 zp;(gl#M$6Z^u$+hw%CCDe(OMG@NTx4I;zy9^so-TpszgsTIc5{>; zJC-dA;I_}#)zz`jI4^yK!XEcDUlg~+m$vBJAa;HzHpWx6TQ!9~YR%wX_g3pjdU8n$ zbk2qlQSpWTwHBjh1^fC78`j<7Xb`yFLYn;JdSArW@>t~Kz5IyUz4#c*KZm@t zS2B(yYhW*4vx=%ea2%^1h1Sgx$iQ0nb$jFOQas#G&}`vb2E;;i)sX2hvqr)rQ+)Xh ztj``&w_no1uR}=F>xLRc4Dav0ueY<<>Mtdqlq#-FDrc>#i@CKr0I@E9VG=b9Q@y8g z8kk=8Rd#wb6Rke4^!<3R+-%nJ@7wf3rSknh&iyyz?E6_+XPaRI87(%&B)z_hR)-%e zFJRYd$X@Fr9dSbc0v2QpZ-HL(=&enz=L(tg%<@TrukP5am#uMB*k2T~cHflv?RmRY zz_p^kzus6sQ&#UDFA(RFg797)bvUr0}lFkc7Nym6EBMnJU} z##ei@HVI=4Z}0_{S@vq=m1MuIf}~XR2$9%E_}45h=Y6ftT>iW4e-dKYSjXiNdTr%> z|0S;de~ms~@`h4S`QN!!8dDWc%D%QYKWYd)R;_Z$D=_xnH16{$BmYEuuEm?wnd^PS zc?oH%ZoNx&s0FBfezyw)=eFHw?wa0bJA@1=DOYKwxVh7#p1;L;zRO+;kJV@%Dk8#0 zgF2ai6xkdDA6m*X}t4f5$nOWB}I67D!mS~wMR^%u4qaCAurx{i$F|*nqQ0${gNPnl1lqi z`u!`;{~44zZDzxK1`p8aQt#PzBF?aKLkIKPl9Q z6L`9`3xRHat?&U4qHtMIo7FLSZRsq;^0Xh#^@?771;3(~Aw0}?VrtgK zzF2jxHfKiIB_U8|*>DljA*Wqel^nHIu{~a)5F#62w=op>ilLaUo!Coau1O0Gn?kjfex`fOOf=Qaw zw)%l!AzuOeHW1_;z!W84gvm1PGwBb89~&H;--2TYx^eN*V*9u4U$GX9^FB)%qO9@G zHbDclS7c4HWv?o5V+STjD9%Oi4_sf)u*V{iiFq?4y@ftMmuuAq*ewDZ{r#>5G17fP9|mef8Gi=RolAYY`H58ycrAms$+bF$R~8jt1dfSq7Q7bm z>d>>qaa*Kc^4{X=%F5TPLV674cZ#n0pZ%RsDrylz#abb?LiglJXlE*vV*W5dLl?N5 z{1c$pB<%x3%_=ixqo=xGiAqi0op=8ybNi7C$4C4Ah+;J^m} zceCOMPsmXL7MwWz&~ArGgCkHg6C- zq^!<0)ue4&Lk?=NZG0R78|Zsujx;(7Xh0j42>PoC3_Qvd=aZYlXwrh1WA5_p!~VFk z3TMD5YzA`*%>38}NB?RLLh*V*h+@C`i!vd5qyTp)QEETNn4o5Tc+E!nq8T_hKUVl{ zQU)>v-W}>~ph@0UH3#n3mpM8xfpdie^slE_)2yj1x8GV>_=%xVR5$N&lf#8jGr^-m z=-{)r^Uq8vnlFNzmzlpwz{9j?hs2UPdehs^B*@~z94c=&UABgxKtm{7;2!nN@)|vh zwMlk#tqSO^A>TAF1%Q$|oLwc&vccCL?CK-$v<%0Et%umgenOi4z3n?3vQ=YwtUF&5 zUyW)A-uqm^QK%XtRW!yi@}ZQwavH~W>ny2L#EW;(0YxHx?hyiME6&Jkd|kmzx&Kx_ z4}ybn;P$``h?ap-<+g z(ys+Vro$z^IqgqnPNvRSzp`h~$!T(Mv;tT0jp0Ldj}}vAo_CLxFSC|Kw2yN10%@He z={9k@-y7oO6M0HhMn5Hs^D+80I28B=E>kejy073f^(?&e)DsA-;9r3-kG*UC`Kzb;d&a7^9G1~gFFy0)OpS15d3+M4&0Y=T{6ig-b zrjJlzdFA_BG3J9{q9AB=>D_It~fuWz{)bz>Zvh&dsqKP>ztky;oc0}lLf;{OBYj& zm^!CnRev&?+#WKmp3{F7VVfxE9OTow+{-&Zd-v~o0ydeg_m5R=0`QS}E~jUzL)Fiy zyUo4KflbP}>yx){a}OnaUxEIG>oUQT<> z8zaWb1Go8*C;)owNS=BrE|@6&uB(k9EpT5_B|df@!?xWNa{A$m2voWdPGa^`Z|U3G zVtf~?BQr_msD7zEkVw}bT9zs=gsG~cUU>Ls?`?3q04RS$ql}a*N(j)MUM3=4#n$p` zuGS&gluFKb2+2xTQ;LGCRsOZmTe&%Lc`2(E8oysG1?C*CR?4i(5HYT@TX#}a+`I>8 zDE*t)hy(>9<1#$}`nOMH=sDZ6WHUG{uT+r28SloqD13*%hEn-}AuQc$dcdtTH3_7N z8&GN~jPA7!G8g9ueb3&CiG|Q)3(&MX&}tX>wFZNpqFp!dc1N#JtUi+mQuP!P#X>n~ zZ@Bn_G2kTuqy44&8_ZK$mdEsW?*!>l|KLW`i)V(IRic}Sl*}T}ntQ7y>rS7l){R zJs2w&MJs>Q@0mqOP{)9iB1*YP^Ox6UxxHhq35~a15>S&&kyVB8XB(@Vv_GkdjMMek zk&q69eB|b&$2u)@)y_Xm`r| zM*eyl*r}gPoOD>#5+#o5PpaYE`xAL)(V+ap;nBXGMR1SJu#NvL%KGB;=Lazl7s>6S z*j~Rzn%+erZGG&s?xqhPSI8X3WncbQ^Q%L9Zm{ZZWE;<__mNJGEpPi5tCW4h7~Hsj z+Mk4k^SSwk++RV27*xAerUdU_%DAy`BTX&L8KkOMg3Dkugg0gRgHDFgoZ_S`8-^-m ziq<~60}|hkFLRtU9+6-O+55@#i4xdV)OeQ@v`qk!JAVk772HNsYgr2A4g@Rg>C|km zg_JC(+WjA0&qWAZ1vG0N?2qm;zJgZswpuPDp8*tm<^Iw^t6&!~=QL{GUlwv8%YZaO zOP`jZdz=PH=US+YAITG@x9aCM=(0KSKPJ!5D?fBZNwA>qnoU8rT=i5%iqYgqbBUdU zn^LNvP&5|qj~Nxxtc$AytFF+Szn{PMPzHa+K-eP{SX}kM+)$;NtfJ>DN5WH*L(b;3OpRUGrU4`3tz$&d|n(*813#p2@2AtG_CyTX6ZnC=4$DO)M$>a#QvDba3YJ z(~74LCYy;>_=0+cY13QV-f2&nN2P9#A=;iM47I%%W-Gfp-W!Z?C=cFOuouy)U5kBOl1KYyTEMSPnJISqb(O5<;9% z5cLI6$5(IO(^nnl{cQ9Pee=i@DgdLQl?oye5Jz4KB!k&gIzW|Ymi6A-gPo1h7Pj9j z!T!)Kl^gN6D$!~K^N%|2ifVQB;uf)RkEtbQ>FYvogG05>_3u#{?VHg78UjTPr`dll z`@IU>pDWbjJa&%Wo7LMBemqvgn9FD4NaaHU$ow9i&kC&VK@;I)Ai5`T*ea7c_*rkg;H!mpaAs*2vJBRrNvNIrxAg z|Ms$Z;mrKEtGg_+5OkSqj;(K-MT46-qmAM${ay<^MkCNDWJurJ^t}%=1t;TyEVraj z0#XEe=3X655YWc?_Xh%ihzWAFP`M0P4|mxsLi~Pw zx};o3I9jg@#r!{Xon=&;VYH>cKyZiR?(W6Cw1uL@ibE*w?$Y2bPVwSe9Ev-^T}pA6 z0>#~Ja_6qQ?#!$^$)BwESAHbld(J-lJX?HW51ug%B|a&Zir@A?wsCV>qk8rFM*aZg zI|Mtktoy8Maa_7!Li{XsOsI^HY#z0fXp~QVFCR<6Xb}! zFzg4q++cQAoISgaCSNFQ1<0~PvqPzT4!NtX!cb68;|clvCJerOMt21B`14q)avwH! zVRElf`A?GZIQ&5|OyXSo|2K1&RypB=9uiWn@;A6>6GzxWQLQ?E*Rzvx`!iuqalZ#t z{|NSW>rd#cV1T;OzRN>FBp=Dx2}!bS)WHx%#X>I`vJ39PP}_;2v|T9#rXPPXzaP3e z4!^&p=aBS#kbb2kv#!`4R&2jw?|j+9=6CXwdwk?^o4w^0$4*R8gr0ww>oYIDeP7(E z)Wsk5$3A1RkG}BKH>QcFY1Ja@eB{)1X+I*yFf|luJR#$7S-xmlYrZ`~#i`8KQ0;{h zf(7=jZ?--jqMYpLUCDj_HUOrhk)3>36V6Y(e*vC#GVBMcjGDYO))K_iTm~q;#^m>m z+Co4TJmo|o+}?1jhWta#luKyw+XwBkM%s-07J8VYX)TG^oMc{tWHo13tPHJw(|F*! zl2m;pR}BM3cW{R#T%o#{C5ak`ixe(%kOd^-&)(rqHVv2WH% zh`hMLji-7q1hndC6VObIB{d*S{~b45r^jR!7V`id?o$&pi7iw&csXs3Yz0``^b5s^ z2s5dO@wnytit@aFO~8}TqC#R*$MU{bKvoXU5E(Tcx!LrVk|tf%QRI5-S=gTAs^K_a z_2#87tj^mR7wI3JBFO{(x~NeK!b;cFC!TrAjdm73Y)g%51RH0y_f&Y*txf)^;bF%1 z;<;F%{)eV{6gRd1^3mrD>05Ghs=wWNZ)2sUO=OvXmoo~_cI zINb33 za0%D$>TtiB^P&FRl3O7n^wq5ifW8?N6ht>GI-mji%qD4yttM$d{9JML-Rnm3tFlm; zjAUm=#}u+FKh!1~02Yee9?*c-SEG;MZkFnl`?wK>zs?9_?}tc0ulw&(f0d!&YAHW& z`bq()JJ=bHwT}9=6#EwVYgs8>AFrpvjpB%<9uYPpdjm~J-Q-qwui~B*8?#L-^_cVL zH`bz2`KIpn0TwuBeX$5ILH=jec^SKSCw6ljTyh}LSKa@Rv|j!M{nUnBBb>IEnsW0b z0|cZyecTmRzXDDE5eQ1g+7CBP&tb=k(Q;*cpoGc>Nz_k&q}q+@K|K)&T6n%?H%?f4 z$>$zQdAZ6Soa?-^_ulrc8a{CBU_4oYXV0Z1n-O4-ndNMec&mqOK_`mYpc#HK;#M1v zWe>5iH>)F`R63FUfO&t!%cYQ;)9`@f(#NrvL-+_IL4z;mYS|!wC`Ynf%#{;+6A(^Z zv{vavj|o7s$>s4y=^mjT9$wYOhW}JcR!WF_(;o`lQe>?=f~CX74uCe2$Wav5@+Byz zJ-bn?%9Naraffde``SytD75OI(0o$eThu85^!b?;8K;({cS&_vr`nVk2OkDW`WMBs zWxa?E4fP~H#%CPtlP~ebG>@+g1u0Uo)W#3?_)6y{<*sH-r8fO>{YoRkwR$LW#^;(E zG855svKIFNkz)&Ca0U^u^S7nCrUa`2!70gPmDbw<<=p;tL2lZBxSsUK0rNIzn4o%zrT?_{87SbdNqC0fe4{A9|4j4v z^)Wd8C4TwDA6zbUtHPSyL*z5=fV8*%WM;ef46^Zhn{euFut2pa9Mn>3%Pu=nh+@(k zA&MF_Ujk7PCdIYXyJlTbljiJczyr$~F0jMA2d zQKn+2s&jnN*n5c$WTnf+0@IeZk3tfDLT+V>IvAE*4g<#M0&rzWTF2*^@0C+b{c3+Ge{Wvg(tpD@YbnrBAUlZ4D7Md2Dc^r zn3Pz_UIsx8&5*kH6I;qr<0q@{E3VOoOUfTn)eZt^{(LiJ4=F8@ddyIgdT~zsF`{_1 z@S)_nm_`F)llNDQN4P#&9iqLV&`?SR%+C~aDA2c|NXv0ZK;QPmH;ddlYt>!ua~nTgq&n+h7^9R{@~Ujr#gD~k{D|PR zAH~MjT`x!QTC}(SEe`ty)qP$xOUNN{jkVe<@|pKL9|>UIo!>fvCq9W^w5El~6vKK` z??c~BQ+AX5GjG7q>M#ECZ<%Vk2FVD8h6)FdvbEUPw3&11z`5PDu*E>Ml_S)EILeTh zqy!6CV2X+{OU%<~(}!iC_2O5F^$K&4m45<1hK4@v22#O&=>~~YDt_7CI(}STEEyLd zuO5L$d{q_FZ^`1@+%$PM9#vDHW#ZH- zClfz$YGjp$`{PLbp3HS%Up@fa7s3M^J#EMWjS{}SVn#juyo3hN$)F|r`!hs%pyOQH zRA8qc(@R_h&)9XfN5b|JE5E#^9CqwVl+t( z_IaY9kawq%s&Hwb4pPIcX2}br4@L;$-CB45Rn>)D61tuw10fd0p<(W!MclSpuNe1S zY*aYWhhWJ=(-m<3k>hZr`=U`M4V9*#Kd_|fkD8Jx$@+1hY{+cTlHp_?HjpcQ{z9#W zsIc+OAD<-`%dp~PL7-@QTg^wj{EvIe!%B%70?Vh|r{WoET?lSfqK zyMJz;mYq(>H+N

    6DVEwIv3H-~Sj~eCJ=i5{^fZQ>at0KKP@8CWdQoz_A@s`cI6w z>s6@%5{5o#YyUOsZnS4*C%0p#r(s~ZTl$^f5{6Y$5jo&-CcJ<)R5##Uoze@d^ul_# zN4EBm|Mr(VsY*Z_x3JLwk3+L+1D8VPjaqB1<4Q}GR&`?g*?4OW!VbyUbkPLXl8Xa< z>d^;0zMA^n3B8IK&n%P^+!5ylINj;xhn}s^|xSW&^zhT*DQ9 zJ7N*&-P77Fp>Eq3MM1|l`a)mN?J2hYx9rdT|2S5Xh8sKRG|U!0m?66&1R7SYd_C$~ zss^=tb}~x6eB%(#R1vdj4Q}AvSO{R$i=UM!ctg`c6{hnPARDAu#wX%otvMS%aL3W+ z`Gnp7;XAl0mAsk%JagDMzBE1IhIr|Gx)}{%Ks^jn&I60}s`zELduWuFDY{&jSjr{8 zuw*+yG|mA(3MNp8@qYRNpRe_$0VOSqnWeq%k2w>ZwifWx;XpXyl@4`&+DPaI8hgdQ zC_TxOZBB{f11ESG8!#rGa+6!&?W~yh?SUFr zV_s3qt>n*!#!Mln@BSkbUS=EqeeerznlAtUyj4<=(M|CQn{&-QDii_|^JO z*jGl-mDo(dp-eo~0_XQa3xsq(NC5o1daoG$`W1TSSpJc6V)nZ&xVYruz!7WJ`xF_< z)`N1Y3iZ%5YN8|Pyu@(1yXzr!>`v3&GlJ%BA`^+wj-g9qSA)h3$GN~>Y$OXcYXm?{ zw3L7kt)!y&rNq2CY#2F579k91LyDHZ@?`ZMA{vu*_jc!8w;QjcBH>HVcpuu|YcpLeO`Ta3CRu%l7Q*d| zA35At3eWgK{>^~UU*O|sW4!5lw{X0EQ^S@!=^`div&sfm0iKh>pg-X2t`o6q(e(n-54qfq7mHi8=A~N&e(1_MCJ7h4Y5M4^_hZoAU?4?k^vi7IqRPR8veZ z)qWhmIyp5uGCk;)ND=DqA5;w1Qv#?cv9FfD@0fgHSbta39gRP3%_Q4b>~?)iDHvzg>vNRJ zXW1{{q{C*fWe+E8{fA(rw0m~T0w?#E-F4@K+Pey5ss`j&7+{JUZ0?L^K3aKZo(|R; z{`{LoNdEllxb%6x#v*@gFtW?-ID^&KaaJ>l=5YCkKY)<+DG>-k0&x^pXKS0TUSW2! zlz%CmOe)$vsT*NlRbH;B`~?xmOg#t5RQa2GWrf2bnhzJhWM)o3QG& zOIfQ^4@Z-ddLQpuhD13g3T7P7o<6Jv2=? z8&omcD*C>vtFO=Re?aeN4^-FYGD_s<|AVsMC=ubG^Ha}@YNf+db$&+%mpD+Jot=CW zDj_|g-9uK}(y}VG6K7@Wsz^kXQkVo@B1?`PfSX+5WXj+IR6j!_Bl5KH8}Kd_C)IsR zjfU1bmn%o@1QAGh87HLooCx-}rmdl9N7e8cP6c{*bIfhB<9?Gtf6cxdEd0gOLB+!F%)HIRoSZsDkq<464R@yP2$tYcQ z@m4~V`14ADvqJK!A(183!I^=g$EguiKz#Hs;L)eLO(qRQ->fAyxSn3!|EVZhshH4e zD1hD+lOUk9Vy6Z*d;*o9KV%=Yx?)Rhg8T^U27XmWKoE7LtTboME>@KjJ$c|fr7&({ zC1ZaW7kCwYxlOZi$xBL>MgV$K&+?$0&vjLSWX;9(8P&{|B(ok1Y#w#ui87H-w6j%M zsFisozck~ymK5E%`rR_F73t^Yn0}+#{?xYrh#tsx>v}Ut93Gsq8XU721M!!`_2ec3 zzqx3yaT8LkiP+VmMTvXBJO<6HK7YAheckWEAx0p^u-4H_6(0}Zcb+wxG2GvGsMo7l z`M!?F1Dijce>OpUHi12x!k{KE7xTzF173kgQi2e#z`{WBS)J7QzwyZ53)RY|S8S9U zTUu<#N|)*RuAh!h;{BdG6phX$`+f1a=~9M^|CU5BPVKI&qvan+=Q$_Q{7}t~)XKPv!LAPeSkXLmfZ=0@A{it4$mmLJlz8#@aL=vi zx!W=Q^4GKAlK5ukTSgfBvD`?_Wdb!qv6F~R-`m-Qk{U+eXy8>%-D+az{ZOevCiTQ| zL-0+_QQ^%t`XhaN66APqOCT=*#iz{;=Ol{xCi?`$a=qSz^G~p0@Y7T55KY z>MjhB2nOC~C<>kiXhT(_aw@PG)}xS;vHIQ;8gQ>ja9*|w%%lAv(bn@??!r#(YMi@d zVi0REryf!i{!)uwrnKhIQaliD>O;6jr9|YNTo~(^x;mzs_Cc16+24Zi!{yW(jkDD) zcGt1ap!~tG5b*X%rhq7@v*DQY8^8Dc-}=YORoFuit68^-q_HlSHT8C*Tup}V#D}@G zkN+rib~QxOkjIb9U26`M6n~P*3&h#b%t)68Y9f{(t78@X_W^-5jdmvT0Zud%&a^XT z-2^6D*0u`yxYn~ZZE;yPDGo($toLj(70G}Bk_twK7&fK~Hp0$eOHHje&X-2B;jaQ` zOvZSg$7R3GUs)P`AKq>(?DSXtict;uDQd;pgZEVO>N-Fnkky44;U~>OohO<4zkUz` zrP=S=a}dB%dg*GA_S|fY_ni5D$JkHp(6fWi0rhaWAx^|8970{a0y9x~6&q{+WU=l< z8>37tk`UnS`vZJ07(=*YH^{19jTEKi$qoob$D2|GhzW9yL>d!0~A|KVZ#gr%iK?L z1fBCZy?Nk@Wr1oz$QBw{y}|F5xiT~k9>bM{;QIe>SEe4|mqas*LI>^DKb3`T{_ zoGsA4%`V7II`Ht|&0jg~Qs}e?HTQAR>pvC-T{hrKOCw5AVCa&h<{N!2&;N9FYZpg! zpUIE-w2YIX-bCoM37XfMPtgm5RqDKMiafNC2(B1G*4M@c7!@#X?EG1qu&G)1_#i_1 zPRELlX`GP{Vk7vI+P{Fr17nnYA`mSCZ{ zc!-8WA5@EEm%ElwgfZ>jG|?s9r$zf2v)zC5z8{j=acW`jn&G>H3D z^4eA7@>ZZR?O<67<`Awg?v(v2do*5pM@qk&w{lIDhgl4$6-}+JKW#ytue*#HXp35e z`k*0I;GfiM9c`a*Fz24;OzlaTB3}I&Nv5i zh>?5ZuFCW{M_#4YhFuN$j*Q&%)MsR7#*6r^QT9zvzw3^C__oDmWRMtor_1w~qCsQ~ z=zZhCX<@-MiGdYQgy25o<8}QGJUtq$m4V3v=Cj@qP*V9+sK*<$(4Dyid`nxCCLkb8 zLz-lbWnhk#ZMpSOn=6rWb`qOU&Ie?!SGw*@$>t=5#QZnC-!*F^YQh(j8EWwgVqL-sB7E+r=63E8Rd zE`=OzQgUdz9k&J>H)!wHgld-B{hr8DM=bJrv&(;iEaEu-x%t~-9O4l+yWHg^_RDki z_$0&LcKa&MHYT7i)HtlPvy%g`=;4TQGi0r(#(bzaI4q4uu0eM1H>g(d9!~SAjW+Y9 zmoKGWtZg>xt#vM_jK3MpU{8uh--_9C3oUo1{e$uF`X;(GJg^Wf?C*1cy4~cx2~q4D4C;VJg`7L<6=DGl+>B{rWElhk-hNO!~Fiv`(CI0t}s28 z&6b~8%kJD&6$&_8T)6Yur%Dv=+!GPU**6}uo}l^OT3Ke5{oVV2aw0>oF(6vqB&I9@GoDFU-il%*;3B8DVF(3p&NPJjjQj7O&wajSmHG_D};L7 z0d8F<7y$$0VWtfz`qrLUCp`5HqdT@PC3U5KQI&WDCtLoLJ{wbz^fQ&2AWqMNo)?wF zk2udeH999>}LuF6e}RrRO^&eh`aq; z9|)2XdXpjEk?=lKD6-k*VSrk3n>l-k>G-0o^SUwrBEzE)ji}-&9H*c=gm4V;cBCq+ zfzA5FZnKr^`8_W?UAi_E0j${8RM(09gC$PRM~3ORB&@Wjsoy8F7Ei`s-Ic-&FDZ}E z%@E#GbN$8kQ}&Ol5VRX@7hjcWj6WrQNcD^N;$~XS;P@OPxugGAcQ!sWBq@2!o58E$ zLf_p1d+q$I^643-m7Q}GN&<+B(jyGs?j5+SXJ>}`7pR-OV63i)P;aN!xe#B}Ung}p ztQ&MtT*>xRmw*13z^W|qmN#>3Y8qEEyxN1QugcP#a*Z!4SlDJ1tF=k;Gi3i|)Ax+) zAB(5?Cnh3?!GO4gcSoi#>|IQ1-XQ^U591xGVk#V=%NvfblgnszD_F4~pP1VPFK+*e zwkWdwoRqe!5bNvOqy2yCod2`%S&u9YWRoU1zQfpTzsYKO%ZE=|z=oKj+(3amW&4W) z+z88cN$v~(HLg8}OX6lH97{>^b;Td>{&Xx~#yj2+S z+R?nAjerbuMfq2&BIMDB01{D&J6|Zx4P?iSv0IW9JxxVjmJ#ROkDn``VJBu@VANpm zJ~Z=`Ena_8yFM0ixnkrQ2Ts;WJ!*;Ohuhw$WL?fA#8-qLZZC?240@B4cG-q-jF_TD zb@&U5_<5nA<^5KN{^@jidRIaZk$K(;x8dhi2p||Ws&!%rkC7vfe`eB|3Z#7fa=ZRPLG_&}ucwFvPgS8`xGiOH(Jf@AItTWdIzdyjt zh!iC*4Q(EL?=<7Q3HxP~nl1NBRZ|V=(58it_ZRaLYM2cvl+B18aobvT$Dp7u=z@tH z0q#Ytzor4`?}R0d?MDdV8uoP0r+=9J(uJMOrvt$i72I0?f_I&GtKxq}_Mc%t|6`z& zPj&E!NEpi>BX%0NHZ!*Gl3E_S?kaY~MPYfu<&gjVr1Z2dg}MJWhc+){nu&Z{916My z1*N8rm-=?o4>8QWUFoZ)((;;@vSS}%foKCn-= zpCi^cFJKDx%hMwu!JCQf=m`kbNC42`_SLO{RFYi27{m&HST(K{zS%Xi)mNXbv!ddM zC|_LErdL2WP26eRdWVji4(lX-RVHldOy?SV<_gmxhe`>c4>W9c9q7!T-1uzb_a5Pj z?Nq-G#ID)A|L1r2a@nkI?{2h2#e|91JoV~) z6(B9gf_8aj^WMDceh}ZOr4ij>EoXSh3PmXoZ60`4qbaG@Ww1qlB0rrjf!8`e?KB^o z%^JH*lrbUeuLO-#C=KM!Lq^RIDTx~@TB4+|1W#$6I%L3))yPuc1CLX(ehDFg`^o^A zWHb?Ri#xRI+Ru1S9U%@1Wo$Qgq;4=1LE1#ev6RsDo4`Tbal>1-K`1Y+COEw{F~4dq z1%wy^jfp3?*(S$%PDuil&ET%<$Qolr7Bp#kEX`pUvNR323xCMt`%L66na2uvE>)Dy ziUlVfDb129+=(K6&RbLe@($IEw~MFE5ebl>de`wht~zwodh>UuV(Cpz94k5Sbb8X= z?0UXGA_R|&TmA@5Q3(2c$sCl4I{aZj36`Y9^ejl5T8Pk2<%tEJF`pbT8?;juKNwJ=>k=oqU#FE05DRu29t_l}2P(~oynHjM5 z*4%!r^T+%EZc~-}vv6v(v@0cAR|@_TTZl4o;L#%&x%-nulzt4cME_XJel+q91w$w) zFW2DYyD$+>PP1$&g}G(RaT6rCN@2XfjCM>1jkuJL0)R>OtOUA(mO)Pdn{NS(;j7_m8XMQSFS}Bpc z4n(GZUP)?S@w!(pRvm)v%gk(Cn(M-NvA*zV${C;&fn~&eHlgP z)?}Fq!5rF=GCFKO)P4cC2F{Prxv5w5ov27KXjdqYZWeMB=V3Q=p`>>hgc-L;HM6Mf z4LGlO3H|y#Q1`q^*+u5isdOmBi1Xn7;`w5H2c9_Mep>m$tb6#jb4l0|gTs*&N_nvM z64U1V)c-9;2;(8cH!nncUKUIrpNNalnqlGpxBeHV)*4Da`xCPyJXb1!@_Kz2WqF%+ z| z7xX)h)%FdIFyjF}lDg^T-ZT3|IkcHZms3^SuK4faHOmr2nD@C@khxwA?)OiVeInw2 z$%c4jcp~mjjpu|MMn$+M>`fHcwkr>y?Hk4rW(Z$AAWaLz^aVasYh)qYNMr2D0qT(dtkOCf|RW(o>wt z4XU*s4el9|U2jLSOs+uGfY|;J_SCZAYR|=yxKQ-8(G!g%&O7+mg{Qvu_LD@l$DNb# zZQz>1(MjIRYsCRX-MnFy=1uz4if|YjryzU;l9vjW`}Z`&h}h#V%>2q!8}%>`Pdpsp zw$RMzbTMkltw@14ofO6kN&xh;;pntELMtBtpp!Z z__zk{&yUFDYRdzZqlmiIL%6$>*j?CW;pZS8QI$kVkkQ8=OdEd`nx?qSuRd3wTbt2i zh9-F`&-mP=X^P)@VTQP0p&l;I<`6}31Fyw}S+E=8?Pp3qPasBP$;nhH*AS+X7*WKd;K6XAZ5-J;ob4<4yOoOOrAAYWH0p1qxCKj!E~FX!76bV?|DF9f;?_5eS6rO@7At<^RlnaMU3Fx) zukcm^+=w zF`D;ggeP(n>|n}6w?ZxB6l(+0J4~(rVgbYwc?^-Uec;pId-{#PT})4Qni-Xqn%@qH z^i+3-(zF`^ZRjMeZkBLvo=j1;>h9d zKVz^)7|>5&g)Z8N1;fb{rwT$w24DF$t?5|#&A}Y55C`C~C5}_pNcOZeocNkEyS-O! z7?~ehy|Mo(T)8FN_F4t1s|B%eWtdO58dLO{Qly8)fnp? zMWUduADH77*s>U&C$yF@{P4f!PFTrn_dT{B>pXUY2d(k7(tpRqVn9TT z%Zr3TI(4`lxgbrmOgd@#xp$|q;)aBkk!jK@F|_%~UE#6s0A7OD*PCAPwQ5l`HyW$2 z-2fx`b+jBILf)#+lh&Q~m+h3l6>~-eL3ZEYSkNJ8pbC>QS+q7~uM1(wG<`7X93_JZ z^GT+4CTw4wpdWvkMaA6lFnSR)4p3b6J4%|sn(p(c=KdACcd$;p4eHk?4ouuqAkU$} zf#Gvf@PjY05$*T*3f@3*8L8Xl!OU&7B@znq4jTb=~SNw(dU#0e+{ zx9G%NAU-?@*M!5&5KK^u%EO`g&M^;{*M@f~Cr`Jy#gXf^LVzT{jIuINg}O|}xLhS8 zHN*#&tsflbd4tN9zgl-1fRt0N^WsCD(iY3aPl5;t2>(JrE)};+(;GBP1%JLU5OQ3L zuOAze-FgrQ5;`v{vg511Y##+4%w0MAvEC+<2fq!t&H{mlnnlTX5?}Vk)oC( zo75%Gfp8vXlPwoDN{pKm+>C+iYrg!7{pr#muzqp^OD0{~67veXG-7={~mfZOrFjzNO^P1K2 z)JS08@Z+54w9D4G;oV{DT4coEGjZReq)zw zzs2ilC>?9_?ce>T`&tiYxc8EncNWBU2hmln>>i(RpE)y`UGsr&M8Rp`b=HNGyq ztss$~w=a{imE*X=_&oa!Cm>GSy zuZlG(J9d`X|H)x@)Rtm=?)yXfNO~$Z6V}<;_dpie%ieeOyzqlPSR)z>tJ zIaciKtMzp{J^1mMG&X4G9prV+yvvhfQva*7*Nw9I=LY)Xi|q6ZXVSp)3%%%Zd1a~z zXHzfe0adq^u4_CB4Y?aTNeppG`O`iyUr+lCvc(7)HFBLZo|PhmjI$4%p-8$dczFpe zej(DnC$U|5xS1(9-V)gdafmcPK{Nihl}#3#9|1uO$ABd2sm*m?IYnxP8&7vV?e?p` zBWC9~DBl9Y&FUS*)#9*3pMaY@`dw6bZd`ORe91BeR~&F6rg279kcF5p+l~H26--T?{z=w6 z3c?g|-Lt97j)BV$+ZuQNd)+*%(=G4Nf&!K$%(_-19>mg&`5?)3O=Z!x`o7XovS--` zm%^e9cFssU>^U5e3lffu=p=XTBWtcCC}8X9=T@?=puM_Ow|5a^@R@cNJOZIVI<{F8 z?H-%be!))TEA6BX$*hbV!BS6WIj>5los5fXjQjAEmT{N#ks5uG-<_YN{QjvUMjnha9mBLH?eKYzvv zrP&*obwic)QsMJH^AQOEWE1^DuKOet8C`!6{UuH0pd--9!t~ZW2UrMO&L`LEbewj? zX6EAy8eP8~^P(QUF*KbBws<$OE(%I#@`^VVXmk>mcDMTDe3B7#Mp5)5+PNTRBibZ; z?#4v^9mgA#b>}CC!Cb_24nUq>rcgF*Ee~Phw}OkftdMM%ksjkTWxpH)34jRWIy~?9 zBYNvDzDH=3&_&nJdqC6E1gCX?ue+E<9dKr1!rNI&{;x@gf`Zh7rt*MO#0dT%WVW!3 z<^T=x{z@{aN| zNd=?xui#4cS1KYpo|}4~Q$f1iude0k*cS2ZdBcBulAxikwI`%>cc$BXv0Oof>)Iax{UAg&p>zapWcq& z;rD-zfnO~9c8%XwnQ?qdm9d7MZp-p+eMYe+a_TZ_hX{IVq8+mgBhzHW0~(jfnz+Ju zbGGWT?Wni!#gqM;f?@M!&~#mH-AYtxzLVVMHG?%Vn2-=uBYkkOqY=eI9DGm73>N3? zY8bspJs-Y0I3Hg!{*D3p%b#|Ny17S*_qGzSuja;1X!mb$JZ;raS$gLS|5`0HxKSLC zo)FhUTZYsjlVx*=c0nQ_Jlf@vA38U1UIb5dW|K%;D&l4K-^#2SobcZ|aB(mRfKc&_ zIQyfIS{15O-4uwv7N4>rMa$j`^JiO> zTt!doe~5-;UG1qaR0QH#IF<)LNX(b25qVcmXG;!VKr}pq?|%JxTT|grd?UOZDMt`U z@x%o+%p{plvJXx}9jz=nv_Bc%pNeSSGdlpXuNyx@2xa&P%p`td$KrNSTdFzGDHUz4 zxBn3#QDwHM7u@;u6YoL8XYjo6K~#UzGj1K?1G@gQ+|yf@7EK!u@>ct)9j7NIPx7z( z!@~6jHC6(;g|h_X;#O=kRjsm5W;h&Zwt*aALvhpG^|Zroh9kJzT1I)7KutAB{dCpQ z_7I@l5SJQbXP6Q(B7gXG0mMPpcJD+L#&;~s3Ft`*_a zFBPu~==+Qo9`>q^CxNyLq@~4wvLopd3-)w-vfC7XDki#`TQFfmJDe{Ad$TxZyR&1r zW;*Pz#nf+KR+Fe-a9*ODQsq?XFIx$(VS{p=efwd{566wF8qo+IM~-9=xoS9seD_=E z2_2KLxB|zeaVvIvPt~6#D7Edyv0C52>RiA0`Nqj}gWsv~6R|1vh4-t&E&cA5XcT#B z#Bvk}v44|4=4)arOPptVhZC|7uTc+OaMtA|w%$l(vidx4EdP~3(G&ix1WCziZCc?e zfuOsCkd27B!5+Vjt!eT04l(B3wS)mk^S9d-O6Kp!{Wfhc9F-j(;5EH5Gr8Xk0dp2_Op@v26QQI+!rQLib;@(&p z&-NEXjhqs)|AW8czdHO|0|6MOGLxc;tEWEM`&pqC_&Saf+!Sbf4wPh=4yy?$MSpWW-8@N`h&~LVx|SJqd#$HAM%*iEH*{W z{wfX4^2#?#Q3;ZNwoVR&|6!4E?1~SB1Qf4z3vTB##|@xn%w1vYm)i0L{zuFjVt-{; z`7#C{buzlUh!ln;1o6=15-oUqxQEevB}PfrPJ#UCi?0MKADwLqe}Wtyyc9S*Z`I4? z5u_I`M(uCi3IRD&A#B+wi73iYk05vnZyd3QE1}%mgaW7tN}8Yapo;a?Sj};L#i0!q z5~z#sB`k;Sn{A}xvNojt3^)({N(jy80z8oX^)eA}GRxmwrqf)0bipP)3$mq8EsJGU zK{_vS-})hQpyd*$M(pk>`x7Ix(||c8DOQTPqaMH~23fSRwc8w8)$`%1E8qQ_CM5`~K#t@fhRi4K(1A)@yYtP)AK3F^UB`Q&mg+!KTi3cntqB9{A6H zPKFVpIRJW9`|0-gCtI*If#d}X=qt-(;94^WX={`xjh4-=)Pe!#s0kVZ-D|HA@jOr% z5dq5^i((T9^#w`05npOM(Zm=IRML&Kok6GAuG)`ri9hgUE()(WS>w>`tRDZX33BGX zj~I|_<-rlq$pRtnNT5pjNQ}-vyq~#FPAub=&w!RSydfD?HlgN@y`=Jqlh5R7pZmFO zR7o`-1&O-=yvxufcYn*(fiVV@=GgbRYay|)lOvc8_7=+UwN#oYBHTnY4PXxcc$eM* zS73*e6p2v)|DG6My`}j9uQ5Gr7!?mXOTkF$2!nLw1L~>>_WX=OqW5;;!MW?&qjj^# z6*>@g=2_~^)vjXrLmJ+h8m2+W47r+@m(~Uz@%(BX?7EDi0-Q7Kjt)_EGFMgtJ^tOJF_85I^?Ab% zZ`vk)(8fIp#X?j#!BRh9m4?vjyL5yGHB>Tx9I7(F^C*~ z3y)^HFzkzRkjiy&0eA?VRaumRb`UOQh$@{6P6Rb|Q*G%V_+k&-<5a4s=0o(WsqLs? z+$L=&T`X7+k2I#*c#=6K;M+zt*f~||2UT|SX`x6_rG=i0vyJwLog&$TR(-J=w}Y&6 z4JS*YlRtg4P8Nl-DCFc66e$8b16&*Z@A$67y5mIsndLbsJDVO+NcxqJlk z?~pr2F|_ksnY5X>*W$`LpQ|FvR?tpY>W)`aRyuDj)qa0qIOh*Ebf3p?tZ)kjs?=O2 zPtEP^6)U}P3U7G4JOxMeCVRThOP3!bznC~kdbGDUGLlmVNU_OD1xA{=_!)Nws2PXI zV?FOIm1XIKvO32(!G=*>b&z+*f@R8PqWm3N_^3}b+PA~+E$c=-fzkV8ob8@3r#I%7 zdq#Es8HChNzxy6OF^(eq3!<%h&Zmxc{g=x}wMwlg#>?`pAW(%hD?QC?);c!8IOL2Z zzzg7(6{^Ijc0BqMi<7zgn!N3Z()#6@dZ>3iaC&x5MWgSJ2=dzh;_EDf+Kj?&o!|~F z?k+73#hn(Xc#&cOf)sam3KZ8O#fujU4uwJj#kFW~cXx-IE9cxhbI+XL`Ip)A?RT$t zt!Jg>Vvcv+?!!(LJK1EEF6)0WLye8a0!bBCk@Qc|b{b13SNr@(93DVb*+I9Y7>Do@ zTHwM*1gbgTZw(-M|1%6y%c^!HnP$cbpXPg=S6+47d&6o$@>VVMDd#NLOQvdUxf78y zfraIpwd5YRK)Hf{w>Nesk0Sq_+1(=y;0$`AfZwF1kc|~*-iEq#8$sBmeq}@{c;8%l ztxNxbuQ45(a?NKA?$m^==@yN38h^q5?*#juY9J{Hm2@tt(R+jvBojI}`Gd zMpUv`mFcJu2J0gQ%(;+X{4$=RO*v>cz4J{|PKbmVcf*XCU%jirkNTRfO!fNleAN&b zbC8EglbiOxcm}?TPXzEahTZQR3^V|und6X6-&&bJIWpV*Mu(LNcnH7R(!iT9oE-9c znpcSueoJX%jw@sU&k-)J{_pA4a$C7@tWd-IaCoa_x4unxYZ-ug@U!1#COLeIadj|V zn5`$Qb_Vw$^iy@T1SP+-fisGq2Tx;xKYlqgCE%9eun>j(ngfabHQL`oLqMU}XrHR> z=YD(Cyd`bo9*UAP7cvU+3g*@>-M?@-NOVGAU^b6%-l2m|0B!j@H#Ebfno9~()`N3i zA&Hx1|P+?mtDEDD%$IquWGz`k_}#I7S);vmxwH6M7TBluh`|SU z(j@JJlerx*W}o9wMF%>y2mL~OFF>Wvi2Cr+Q*3J(_2Cnfc%W*cllw5CiP@}j(*UfU z4{#_;e}aiTq58*s9)qk6vaU*o7t}CaMg1}?rEA$jLpq#AGzQS3^)veXpN8WI+x+NY zslHoBaX?{k0v)F^;=ldKqgw#DW^G50W5nViQIMmjqgesHUyj-bcxDa@Pb>R?n>IIn zV(2mK%^Yd;V9|P9#))}Pdo_ctcdfub%tMk#D>&5x23xj}mrQ8k z#ylaQ0d1u?Ils!xgKhh5*mO*A|HTeTeqqhP*vp~N=0x~~A_HN2y^GSHfV;Tat=XVG z(#?aD_PG32;a%rEL}Fo_h8e|W`oKkM5xz))4Q>cuoh^A)8Td4hecttb=+y$nf-VOU zV3}mFNRvJPOBKujm**VsBuDdZYoc+y0n;-Jg_`|1w-R|z9n7<+{tI1FhM@&@dS~egMhL1WVg>k zU{gD~{2#w1)iB)OUG<-X-*04izn*aPnsytm%(@966CUSCsg_!xa0VspcQQ!eMiljJ?Utm@-H|5d_6^ z!gLkR79i+hF69U{!j#kCaoh zgiyO~lU)~-L2B3OqBOZs{yDq#aVw9+JY`ALa8w;LJrY2!DF3o?*pY%3^1@9E4AC|W zR)-FKM85dQH6~cHar(bMjHT?SrTY2&$&+(~GGqWrRijn&-DgWL%AKqayq@0^g;QdU z3-BYpr*3a=EoPa<@2gao-8kO9zOFHj^`ck$aq}Ki-7z!^VDh>sI znjPyqk!(!kZq^eu56K*_&ehJB2+$xM8~fAP4d9{-u32Px6nxJV8TjdbRome1q5$m8_$DvTUIDDC?gZl6~oM4RU@Xp#!B?I9)B?@)ARc;BTBOI# zY|#qVab)-%)K~b?S7Dmh-Y+2FljQTri?NmM=^yIDp>zBWu}b*lT!fP@{>hJSfbTCL z+==|~d%$gh@nw?WBXKOUW}s+UDSDR7!UVnyl~S4w;ApJQdH?L^U1adTeA+U*F3FWs zPUnO$7pG{e2R_~rlH9mE-^cb{Z72Sy3^G70_3-;7>DlYRZ(Om0xwXXe^6&J?4~}`q zns@gRdmjmZQhOiRg{ecNtea4E#U~b$L5fbQAJ&RAg) zXAjC(^2pK(bpzyxzP&ejA`-^`@wWJXVi5oFO!w{;7!O-3kI7G$&5-RoZcHpbg}h40 zXN}R3BjOnBm(|IQN2K$qtq)fC_zJGKt-*QesD>%18zz1a+Gim9i9mdOc*UGh&A>2w z_goUj^}e2Ivf3G^&bdP|0Db>}0RqigPU>RG4#%MJyq?zzMkJzygn*GshwlFt#_nDB z2Z{FrF-KD$@TZDZyXmBACzGq}l}xaJY9Gm@pR`O!;)1`idMpW!uF=scVn5WzJyd3g zfs6Yi13r@~a)h}W*W#RpCrEP$4PzLfUBC~gD;Q`|)F?td7Of})%<+fc6C#*ba@d*8 z{U+|_W`uFx?j@d6piIJ9k3r;jb{R`}v2_Ne;j!ac&NL<(0(=DimHGDjl3=g%BpEKJ z-A4SJ-%*llW{&&xtLUf79=SNN4I=n+KX^gZ!n~NdQ(XXW(xA2*3N4xYUM!@jc^#_5 z-;ywYxzrw9jaVf>I4nh2AG?anFQrp2k5?UstB>y+qB?zIOL*tO`7@vQa8kDnh{KwYr@<&?p?~vkeu!I{1UmdBtX2isMnEP@ zU-lMzTn~*SmMdcuWAz-*7nE$Pa!iFJzpIax=?p39k3g(7_+Gxyf1CW2ITY59I!p$% zAjQ+*0PFt%5ieVi!d@jw_o?yOfm$YIm!f?XNag*|;31dS{4JXShvZN$N(%umE>BNd zFQs;`N9|S4UAPbu(`7A%#twPF$CHiZ8=WJHOU6vOcvEX;GVKyJXI^NE+8Zjj?a>{Z zCjv;!EKPi=rY)T_|Lf6uWKn;FjJ3PSib%Ik5lNO}d=@NrLaE4Hpq>5ggrcMo=DXa_ zv`vC$CGiUqC=h8Td7kCl6*TpEazp!nWsy!FZG(e@zY8!mOPd1*B|~k;xYvzIR}65P zgvis*qkE!hfAp!~#ll$nqX1=QbYY@UP8N764nYE@x#+$s3g;Z8J6p=j5x~R0pFxg# z7|O{W_+3{Hv#DGc$i@$F^qGFX^-&&|d7VMR^P^|%CsJ}u?83FC169)>j6eKg@63I^ zmt8Oj3b+rh zV&~FU$Y+x=wlFBvEcTGFb-Z=H?iWn%Dl-+jy8M4;; zkUo#G97gu&KT+gf^4qgn(2_Wi-eHTyB0ZtyC=g%O2w-F|_s=c#S1CFge1>aqL(gW$ z?7Sxa%u%7Y+8F+wmB$*XrwC6Q$5j7_vDUmZ?r)r7hy79;JLyPot+Ke-eJQx)i-TkT zbCr6b-Q=e>uC~mT?j5H0o+)jCW)*O+U%h=4?xrO<;oZbpw-9qHN^PL^|&2v zhL-ATI5@MlO`Ohlm+ey>)|v*ia_j&rUCjuO))G`Cqo-dtNsKuRsJ79k_<0N3ug&|( z6ki$(yJ_>e9LwsgK)w$alWwioJVv~8dgeaG{)rFun*9BTqmlNpT=i+*aGzjg-h&;* zEvf2&CHW!Xh#1*`NLOvaxdLx|W{~q1cOCmOQR9~o`sw|jeD!Uh>)Xn1^)9FS@pR$J zfNeqal{Y<4WPm+J-ZCLC&0ml&%PnM`i~?p-?e+6XzNDB1Z?Qo}ZK!kx<0Pa3?Hvf$ zxLwVMPi@|UUcjaw5%^KELJo!5>Dyof5?ZvUWP8(0Km6h=sizv(gO%kZ%Ffe@e`kfd zzPs#==r_!{4w@$18~zsZdXT#)FvqC$~7leCg3 zxZzq#E3KVcDW$^fWqdE8jE8`@9{=kiC;p{<@%@Ve1j1|DCll`YE$Jqr{YNJgD>RTNIV^#pO+{BaQF zpI%l&LfF7E!}S2F7;Kn`a1WB(@sif)2BgiwRbJfs9Jt zAvPpiHdyXzo_LN6T8mbYV%xQnYVhLe?m-RE-j_$4x9KC)rzg~^rb73!NmR!tUmRZD zN-B+5L8aj`QV?emC!@psR-`&Mg3v2!LNryLbkLOYHew)fD*b?)GG$$a>OYGS|3FdxpSwT37OPs?x6 zO7zt}qar8f;Xd;2*7rtsGBh964|v()v0^tVPPrnIVR`BoMV@XR}NbfzM=(Va9 zFt!8*4+cKqz`^06fMfgMwCZhV{F1IH|6yB0q}2 zF|xf7RC`SFG5s1OPx4@fQ0$BJ^;#tjCu@<6+dG_TH@pRSSV~`)vYOmmlqx2R^q4!| z*Qxbu1HH09Sj$hKJV*E0=JdK#unSV}+)Ylr50%@tlsdds zM%TvZpah&zxw*K%HS=dTlR{LY^tLHN;yQg+x~gf9`1CS}lYOT{ShqV|W_^XdjlLQIRqRGNjP zjdVFy38lcV_YAp1h0Zb8d607ul9G-Mpgv{>fb|U{98~H&-w;qDmF9V1!hf4n(w=-7 z0{q6clX{58jF`0SAD5r$+R)!}zjH}%@3Xf~rj8#AHw!Rd^qq9gt3kWgJs!HAI4XNZ z<1dl<7AdMR2i?y`Xx{+_!hX}m^&@LkX$!!K5-hxc-)45)~31Sl84?S6@@f!AB-F@_GO6SJ3VBWb^@q%we)W9}=bclsR^ z=UQGcukW_8ZQ{>`nEBEj|>>6V;{mwUHp zxly|Z%>ljIOfn$2XJ@O<3JXHFZ><%f1c);gi3vZOlw-vP-Ikkmm|q-v|B=#t3;6KI zwUy4PT_d%*CY@;ffU@TMmZ6jGCh1D`Aa^JIdGDbkqJ5iKZwJV+d>oWoh* z#YUE`K@2}BCK>eFveq%uUb>G*r5NtYzIvGiGY^l!f zITei22}E#&WT6H~b#LhEvlc)90l-gqeJ;fbT+oB$EO5|a#ux;9-dN=5j~m|YW8V`r zHsn;Q8b!;bQr(P_tC=7w8SjS4Iv)cLuAu_>FETSi!>BFj$GVkzp`QgTNU3o=11UP< zomDMhU*jb88PwS;2R>dd$N=<=++6+6{ph$dt0g|-qNp#G`t)8J{dd{MSeyJjZhCkJ zRihwssR(XriF>9?idxt>k9^@iAS}_~Hn!?!4RDrJtA_o}{qIaiP<}FBjFZy!dGqCQ z!BDvY>bc9I^Bh;hl_GaQsZ?DTTYbHm*Kuy z!L>%q0!py=`ju}_a!ZP~`qo4}M*mwxG66fspXR;=zl4q7+u%fLilDj1Lbi`ce0@;qO%J}Wa#)bzUXFIB?q<+y6*pLwDivZd%Iw%Mz)@osL7RrJB zOua}g%RrDGJ&h2`9&vD}#muw`8CZFTdmwrBb8<8_!-Yni>t^j!aUDlH2w)z9#mKI^ z5<~4PBj!w?ix(mTXW_Kxr$)c!wjh5ZNCTVpeOv!QEzB5`MT?bF{6BXRqF@0O&sudG zdH|yr@AX&0qJOyj0TVJMA;!PrZu{heMkc4Lw+zPB%SeeT;6kOuYUbw2!gAM_xf2wiEvc`kS=lqN29Qt}1=3K>&CCIrd-8|JeI zK1hb>f-5ya;is0x8p7l@QNdJ%ph>yj8#7R%FEtLf{JT%Yn7R?Xnz)l{uU3S2NJ)qQ zf5m4Zb&K-ySsX~OM?f5oto*94Proa;V)9pr?L>n;t$Lck4Sx`6^ zk(fKuqaz90h#FV&o?%}<2q7(MGDj14Un&nXc;nj zNqDxH^WJH+q1C(hzP`;{rhp3`M#i~qd7;U~8Nm15Dk$qNpYbE9`CKVe7A(oa2(VJ@ zn$}eI`;M7hfcC!}@Q?=FefR?JyzvEkdfGc~{`qZ})Xrv+Wi`(!D~jCO1Od_ zy*gjgEoLwDvH2X4PmsAfE^eM57b1T(@l-ol#P8p%SzeT`tsrM$3tBPnnsDW&tVYBn z^m$yZr5?1MHTH^)jeZucGV3Zd-sGKTEz&S*M3)VTd4F`Z2G2iPc6yKgApC##zF zR=Z!lu?6|Z^lBrH?yvt?Py81gr&A6a=$nthFw55;4L*$w(4Decpmm$uVUIqe)g8%g zlj!qce!FMDs42u}T;}=GogsC$U1`+9D`D@|P!;U@lu2qi)=H0k&e$k{t?)jGHhv8g zQFSW><+@q4_b5BdV#p_-NJvS z;(|zFG=XEtiU=gAOxOwp?cbT52LeU5j-9^&ae^#*yds>Lo3505p3J=Nfi03(W~G;0c0OB z&))#4HdEHH0ADMC{z|#2(u`CDY&ZPby^t&X zd*wYl_R~}?uZu_?klkIfDdUHH(JcinYXle<4shf%aip~b-DeXXP-RrAVG$&s_{^EH zSI0K*RDy9MLajtWBw52p0}jKMoJK1=wLr;X&~iZo(B9A#i`o*{EmUM*^~^i&U5>zdVIL?Z_c9$D81&xa2+oF-pqwe@|)tPt$d&to2&Iz&f7# z_#8Wq}t0RKx)pB>rsH{*K4q=b$*&-Bop)y zaT^G<%vvSCl&Px{p}0^p-Voo_1aPnS5kx&uACnxHqw*XMpbEVSoI;7*XkcMq7&s7r zP|(<5QT$~5N^t+|-9FaHnL3DZu7+VvqDLaIV{Yv2%Q*0*_3Xt2(EPL&ZWY%0d#Q8}Nd@p>4%HQ_f{ z!4RBq;2E>z^L+J}@2z)qAEYy>yb_}+j!Ch%@}jaz@87JdAPeGGGNzwW3=yWULO7~n ze@W8!(T8<*$GR_xs7P--kP7)&AGb45$l?Op!t(X8`tTnl5erHB-5AhMCX8gDGx+F7 zPNpVE;zAfHv7f6LsD{eK{38oE`8V-?z~6FT9XLynm_-yCj=e>t2^nCVKuzw%28cOWG7J`HXOraA%`S=QXt9-`;S=0cC0YSRu|1|Nc19e)htg zX(T2vfQ4?1LmTVosi7FOb(TW!EPQSy*ml zkj?l6nx_f(vPh)@=x`7qoyn{*n%IHl8B$x}_HOcHr-CjwoYMwdpo9=V=~@EmY`rh! z7W}Jj{iatEXL6fY(jO&o~pUwU2^SVX9lRnokise+>2}t+jnp(m1 z-ajpj09XsVZmed#u9OcG^G2_ZB+fwGCst+$d}>EsC6jvvC!`1)&S^V8suWHzCL($q z3d1(Ryd}y=9o)T4Et7wC;#`}WsC`rjGuUR(r)D zSxhDv`ULRPGPA@EpndM5KY72udYZ?zztnoi$kdMgb>LPM(({F+)^>mf;5ogP(9)Po zd~%-+s3DCCf3{f@bC!lq*w*yWb#W}mV9lE7$dzyzrRlvqy+}(|J*Eu@JfPbek~}pU z6-pgoz(z(_Get3tnTxRnf6w1^a;LBOvbNf7I#C8EXD4YuyTzwQ8DVC49Da}i?ME7a zB{7{x_41W%DUx}6TBdu>ygQXMkF2eP&o?fezJSe7ous$-BXuYP0xK_3{pvwOi;p@p z)c)^NulJL~gT&l|X`OJeitB& zCio>@fSvU6s%3(|u2;c87`XP@ED~d3WQP*YfIU-^orOtw@e>eYY#WI69d&wbj?{h! z;iTv!+`>IOh6ql@g`>9hPiAyvHS3$47ipGen{M?%e?N=YsE=dsuK z$neS_+skPZCla+udxLvD1{3q{qk7m{Y`8)^nOC_zSDE1=nSonB4v~Uo%vkuJzhq|Q zrSy5c2K1%=ON0o zz5NR1y(-pRJ)07bjt<cBUBOC?_UEkE6|f<0VmIwBDQ! z-RpJg8iP%cbs;iDA3c6T2HmxHUui~CBcEP0@?%L&nX82yFc1;^p!(82`rR+Ytqj3E zYAv@!K?UsAdcrSv!(c;S+%0{P3@p~bla*?*#6<-nV>-nJw>-`_I4tac0;c%zgO@IQ zh6d@o`2sY)C)L^^OwZj$Gg^SuZUugQNkmHKDiAW6vII3htNjlaz(oe_G7V~O{vBVA1%QT~e}K?5AWW~8g#`i(OP zk{-YoBh5~_&zKR@g&GoySGW89iNL#;`tVujaSvNRr?Y9IFu4GO1t{aJCkz)}8+Q)_&w7idlnF!-W zAxZvPKya2iv+(dzcYA{MgeD&pVWX1)@jumOVs7LRd+e>Y6z+?<(k|3X-E!Q-?Hnid zm5BXsheM(*s05GmCVvw=#(Sufpm@F8D_5*9cE(1_C}h+tA7Nq5!SA`y>lVEXfshZAjlcc~e! zE1Vf{{Up6I^HaZCHZgsYAFGOtW&RK*T4n&PxXhy9%@PMh0%1Nc{hKAMgaTgxG1UJ_rQ@sN$; z9e-%SaxjWI+dzj%c6#MF00m1eiYFc@CAwAO6laX|;f@yBrCOLDQv-1IBvm#@MufLI z0K(`a2J$;Rq%p8tarikbX)kS)>>`G7x;uBqANKv01i*;~843$%r)2RdT>4wcPPmZj z{Hof@r!_8Z3J)y+T?WSZytG1WXDRX)iekN_@zG~0Q$gx59)Lar9A89;%93Z#*@e}o zkCcY`FEW+gH2xns#Vb_lE%ZonjC9R_unV5Bf778wj-x0utk{rlbqku<2gwAFf}tjA zsTw!$v5H*|d#f#r%B*z7?^62*4dw^GmfjD^T)r`T4$`7&r4zxgPnsnd zk{VV6*?X&(F?Zww9{C}gi(xN}QE((!@6PP8a__ZeAPa(y_w~K$XJMH$;hbAJRx=k4 zQYDEEkvtsVobj6(VdpKOmjlD|pyx2dbNrWuh4W)6LsHs&7fwn4+{WUu%dz^#uic#w zt1twIQq_4j{@?-yF>u;RbYPR%i?<%kuK(sSO{pFFOLSh{`rvY?YUKUcSoDX%qVF1~ zp5*)DKkrT@79QTAk^^;h5Z#UA4_lt&s{q?WnTh9-Rh8a_qCy7eG4Ug8x2yd0{H{fkY z;s8FQY}uR>)Ro@G!CX@93ny5~+ni~Qgbgv6rgJ}`li${*ggr4t;tUY-6Ha{fP#^i| zoN?FhIET(o3NqEaA_-l=JhVM_7(}FLAOEzNkhlU42QMk3$(2KzK4%cLoY(#)iX9e}LH{EIpDvYB0k z8r;yIO;D7!O#P@g(1Jyf+Mtq(18mUgki!C--S=7xKE}pqF-YYfA}SP}9n$dRWm%-# zD&3+ds=j&_GJ1QD5(I@8u4%}2Li57%Lw{ltbi->2gcZ(`au#~17(oAif|Cp!Ub}G~ zn010BqEWEfyPt^%z0>2A-Uw)6H zinh%L6VQ4wSj;2DypKnfHeuRTXxBpy9W6gL_HRS;bmMod_+6X5qsH7$Bfv_bP{`@n z^~T{az)?2^w$V=Bw>LLC=HMQPzCp3{o@&t=A189(JM_kTvdj8lr&@LKg()WEJ1ZT( zBc0z2;B=oUt`{Z|?8~20Y55hTSBJlUvEWUTc*UI&MuWdPj+@8%OF+2UYgp{>H{H@K zZQKtObbfs+FMcmyF2xXlqEm6(U}zjJkDm<*8eqe!+L@{|j z&blqu1%G|RLP!>N^6}5!2Br;9X5ApbNa9NqS8moS?=`7ay_M{1Bx9m1a`!@*p#%zj zmIF)6n{u}RYTs5Kp^EuDl%>5VYZs+c+7oP8I1Yw18saa6BWI=>_hP3`f&C=%gz!<^=PGIo1MO#%S%$5VCW0oNpSAc4> zF3CcuF{CERAZC_2cor%R3NA;os=ECx(ksIHsg4=q@P%D~tc_q%*_o{K#LM#reV!Lt zgLLk($!i^Si^#kLA<)Xy!J-HmDVG(qZ1YY;%9)CvjE7v&o7~MT6 z8 zUER&mKj{0pCP8QWQE;AwiU-8ugEdPyOnnxmfh%Q;SfX3HlA=p<&F|e+Ei9n$&!URQ zsMJrmi{$D~nc?B2#A`-YAGGREutZx2GU512GFzlap8bGj)UdpE;60&EA$e8QEyYf` z5LtNP+QV!?3V3G|k9p{tlw84jb`c{dLCc{X;RNbT`R;o~y16*^7-1iDD={a2tL<## zSkHDS>Fu;tL1*8dN|8&*<*z#(l|FUi@>~wW8C`2FFUHLhd9iv=QvGyfyazbF=&Im8 zVE9+p(Z-skG8E0qKL71I0}i~05E40;LCA;p9*iDb83F7!U*fV)rUtYft|RGZ^O2z& z_sneZBVE8Kq!!Qf$uQ+qy1AG0cUmQn=r|&&Pc}wjKT*@D zK_3r&_{Yj1^}_ys^tDqRkZ0s1tXAj){cjOS+^A6$HmEB4LxB+2wxviz_$=UETbsks z^j*wEB7h3!X}xgPrd`p*U6B`vlK1iSsrPSiPE@i{=-59ZT2SjT)pn2ks@vWg`QmRs zLoPA^PknpLP4&}**H4A^NZy(7xIf=o)81DjJ5P;BJ?xyL0WA73$wS23cfcO?)bIea zjT-tGTdtIp*Ya(zhk4VIf(3t(rpR|9o2q&nUU_1>DS5SV9pnYj?U?v#J4w3EEib1< z2A%rxHKJWNYV7_j2#UbGvfNUP^8yb=+l8_n!+3oN)|fqFDbS7c{UG$aX-x zF>lh{_=XyUbD+oDMDE5aJPn;8zH!-dL^>@#=zDe$@Z8%x>Up}4G99-N2l{zVNXnR! z71JWM30L+%3yQUJ#Vw<5vGb&KmaT>G&D(=EB!UK4u-}?V!2gAt=C00((XKz;ZLIo0 zFYICX-Zc4&(hi?|dVCw3sHpv{bZ_xGkPVa1MvCU+&*lPJe3`ffR#pb!TaFRYJ0TEG zB1fF~SJ;PHYTrBB5ax%D^HbwHw*UF8nqw8Q z$F@+@+6N%d30Z;HCQj+Tg72cfkQI^C^p{YJ5uJ0IVqttPcH>phQDl2s&%WJKx!)mm zHE4C#v;yKNDPzk>p<{%4|B$8oS?a_~B|uZTAkm4popqZVYny;Yg?ArI)aoO@^3!%M*Ze;h$G<7)~^rMbcaGZ7FxEOE;c)^K9RqQmO0n zez2&^v+dMp?0e@`CTBD6=~&|_R_ZNe0b46oGu?x z!Yc5QTMsgF#mVfjdHRvy)AI{Sdyg!!urN^uM(vW7WW0G5g^q+;pHzY(E3}LhR67e= z97_NeJdK<1y9A(3u(+Aeo5r{l$@5KbH57>Oy^HTwp?g+kllS*rKaGpmP+tk?|bhldK6E~XouE$LlV;hzXDS#(+q{Y z=lf=|Io7~8cxeoficNt;X5vvh3svbVEc``u2)p&Q57~Z|t6a{IdDDTde-J0trZoO# zc}lIy^s5=AKBZSM>ObDXR2m;4H%HqOk2&NsADHY#X4nS%V!ui>cI{G15clyVFg>8mK{ zPz!_+8m2M6HSnfvL6)*X7*tiku{xLEM(icSO49y=%5I-0{ByT|@R13pqfm&k+9!M> zu^0kO6R8@5n%6h8{{RdcKaTe6}+|?9k=s=slA3Glgi=o zQP=wovV~S7Dd=q|UZW6&rK(^nVI#@x!7*Lp(~I@VM+@eGx(f&)dF#%QKljS#n(O^y z-}2))g<3BpH_xnc=LlXQXK`U`BuY@kP3?KgN|2h^f^L<6`ve6qy zc(eRXc*lQ-e_0`tW^5BLqNw45_NbQqj4JF41x()w)sVrj_<%~_SGEkhqr>vqtm6aTO;qmkNIV$JfJJ>`2gb70lf{xm&A$SjuLBnB3yN|h zqae3R2@4C6vsSvHi7`5ysfmzqR}K)3muMlypiKgF_EUmTje_yMY++m*)# zI{Y5{)CZkMrTdw7bVQL6H)DwnoLU;sg-Z5$#xn9Op`dHdTK%aEla?pc$SlM~hW?SE zN2UfAhVe-XvXG9ByuF6O0u-J;%GWVu++Vi0g>@8XymE0;Q{E=XzxzYm!7{rNpuSC- z9#IFO-ZH8h<>{?1X}AW(YgKAxBj1?;&GIs{%go7FCO=d6Q9d2A9g*uc#a(#(45Ut9 zZSjZ=BssaWnY9HZO}U)B@i%EG*yzUaP@&sUU;fRt@&vQPIt|F!khIVI6f8!XmgW^# zg>Q&(fAYgVD`!RAe)CBBJ1$iNFXca-1><;9TGyj{as!d445+;3FDYb7)U&Zhb(345veD7jruZM^B#m5SiSOyG?c+ptMRDWTkT(zmT5Kc$uUGT~B-U{JDCs^yLwotIqqmF8kr! zRPPE9$etr>z*i6RA{5$o9_uAZ`bKav<%c=9-(KZ&=JAIugX>d~``%o~dwI~TKhmlW zrw9Dju?7yA%>k+yI=@J(GLekVm7!Y1JvkkO-CWTtif4Yu0~#SGBjt;Gtr;~;Z}{y! z`d|7v_jJ3J+c8}o0tq_#$-{d72VZ9u)K=W}>jZ*(p}0eFcbC%Q?oNX7X#a)BDySsaFKlx_9_nez|&TS?ad$K2cum4)V^*nwpmLhXbn3Mun^}%<+Cpwl- zVA6lO;^=e# zGSYhIvK+5-Ox-7Tbq$=eH3!Hptu)DE`g31WC5DOF%NFP|Q#nzIY9GOZ?-O(+{J0!jlm<&B%Hje)J@GYAh(9x$tK#he) zh1?UG`KNw98|GCJ3C3)aZz8?l`Bgw*|9*ir+6$c zs4uN7Rn30YVmtBu*i;o)WI!T1m0k4t{3+(~Y=D=S$Nz+6K_Cd!P)I4IK>>;vQ`Ugx z^v6fMU`$7q@JMwI*2gg6zJcDSEvQubhkW7%m~G2cpbK&{jUvwH8e#tJDV(jZ0~5EQ zfc41T&9@eL0rC`bel^*9&Kp9<@zMu0N_evKMFv7GwMkfjaRPu--HJ3}zf_D><3DWl z!S6xK!T4`BcDU-?Ov565ACpWb zOG7c$ZR^RnJsVL#m-t*@=(HhIdkZ07U26Hza6jS!`<9_6)EC=EXTB>TfN%r%UW!1$Uow~4tL0}g>cHRdDR7p zOTDWLp`4VZHVlW9R*e0b6doXnw6`j;SpOD5Ek&uchIVAtyzf0o+!4$i**H8;uZBcz zs35z}HS=5Qd;j}14&*VWx{oFjx`ip?NPD;Hv3olnAEB3`baa!cnPCEWrO8GH3F)`> zu@qw9<4I9pfV6#Sx2i`pUNb@g%;1?CR@=cn+lGVUfO^d17zq17q$}(HHumF++9eCM zaZZq}75+B=LV_u(OXn|E!=b6lx&PJi6s7$i&ONw4WD7l1;WvizOt#(V;7T9Qz6~ZFy!Uy z6HA}ST6yn48Mv1w2f)teGF~kOp9v||`2yVYrF87u9R(kPMf(}N`g>hYoq^pVsPkfl zTCDg-!yIHm0AjpW2g_4YD^YjuQJcTFQ0kA4UHJ+}x=%(aO3DBEQ#kWs#m4nnRnpR_ z;Y)jE*wk(VI4q#qxj!sU&|@|p(MUYCsh#x|5HYnF=0jnj)3G2CIpIC&PiVw6ys6lu zG)8G?@{@hlC(C8PH$Y3hS6?m?OTGq3AKsHSw8Pzm%2Oehd@0wUc|KXXmnth5EQSx( zd=Dk&krBsaPD1x*`)KuxN7_2Pa%f*2UC68Z<%+9p%RD3adCY2T!<(!ZId7@cR}mbB zEQy!;8G#zHxA^W+ma5OPe;VH}jym<8*xsaxqsr)KP98g`pbkYp8TF9IAhDOL4lZk8 zB%d?X)t&pcWr_i$@V7USUz7n)2K**&PTgl9pmajmLUpwZLQ;&uyQzg}aCl4Zq>hon zIO-mOKt(X$IFH6V^+G^q5Y4I(ku=W=FqvRgG(km$+mqzjyh)X(77MX7Q_@Sfn6bO? zqbQw0vvMzQm7R-NfIbc66t1>Rrh2(!tfI8;H#0T_P8r|=0x}*3o(L-*ugCuS>BiRD z%{PEWm3tcy;>A_8`(|1v9pJ2_>gSE}v}-(d1>EwiF(I6E>OE{|qwpx@yNwJ@l!mBI z9|+`c@VBwIH%{VXY#f}3%hz;`FQdL+9;N%Jyd)oYQR>g2Ku1+>V#@lgT*~@JKE@+A zuXAn+@`1e12J?Fu@&MW1gDY>{N#u%>`Ycwg=r+;ox)=hR{*85_a77E_z&MC$Zd*Yw3{={ZkTrNz#_Q<89RSm7@y(=1t6+ zqNvG){Gy5j=r*lGDBzoqWZV=v1l+7VzIgrKLe^_h#lqC1q7Kq@j)E~M{xXb zvx-WQPwSXPJ(B&Q=J7o7e!EbKWyXVl#b28%UN&vZZ)J1Z>K$BZy|Iwi)DE^^@U0C= z{bBTaeA|uf@QCdwhlBP~_A(g>-v6+Pf8}}UdG%;sdgrIs<6Fi|QZb^t*35b2<NhcaVX-q?`u$3|6|`NvPSL z%Q1wk>Gm}P@Qz2Yqus%;>%GdL_3wlVE{)H<@}HACy^Ie2@eU$Se(4UNqukr1=;Xmq z{+(@Tbipg1`9$}ywc+pk>=md+x^{os!49kca{ID;>e`UyA7pa=_p1uv5lf1nwetG{ zr?Z=80Vla9lGo)Owb0?(ktIdXa&ss}_;Y8CMk7Fm;opQEV9KBNKhMVhQbABpz)VS< z%`v?qkK0!slHKjT(;+}mH(u3^^>^-Hx2X9nnmq0M_xOe1S>^iQR~dhOIgI0Q!Hn*V z`!lG?OOAMd^F5N!mJPETQPj|pg)V9QuB<+gJVgaUSJ(F#B*HAzavT;AI%R9tjDhGd zw!g;piB6s>`aQl0cPSo7gBrEt`T;0U0`hqqo3=oJ0#4QC28=rU_Sz{oLO)o~^7i zgDm7DWu+V5MAEc`8EE0FEcTgEzCYQ?q7w^ee9QB2Bh#d<2j#)Rb?MRCu1k~Bfr8x3 zY%;wmm=F_C@=RQ&r#}hF^m4d zWTt=(+Qq;j{%iI=RN<2rz*C2%gBN2ac3V-{;>PmWLQ)$J+`yTX1=EU;-stO~tBF%& zBgRNN6|w7X6_i5Xm55Npc%Xp8Mv?+TX{Cu9-Hs~|3OHr} z9jye;H}_MWOPQq&tWlKz1WQqVgi5{JLP%pCTM5IdKL55J#};cn1_(WNGc z-7RjpJ*CZ1Dg=M(WITxcEX|60`0&{Yz2h?ha^A6KyS=qx|*>knA& z#xJ}I3^%t$@bM-@NV}cfKH_3&3UP2gxs>65t8iH86pU$gXGAKQ0zlE<(5>8W_)q?) zv-bF;bg7;P(sB6OkSuD$;Un^#jpV|+Ae;kSZ|T@R)Ac`WZ6Enkqofe=`FaSsaZOQz z@c_wPw-nO)e`ta|@IQmC({_ai=4&H>SnD;RH1GVJ7LZf(L%>{GW|`8x9DXI)$ju8~ z4YN4D(dfrU$42m+3WjHB>j!RE63n@G^PbAH>^=IQ=!GWW-Y!ns9OGMGlidm>9?!I# zyv)o84&2^ZLYOHsW?EH$S$+uSR|X~pC&3g@0G?QW zeutE=OdvaGH@(!*%9SU`-Kv%rBP-bfkPUZ(>!e@RU0iS@ac%M2yan?T&%t5oZb8f~ z1z9eE-k1iX$RU>9EoaMt)?W=rX*!aMmO8egAJQQrRX=UvVy%Z4u1?>Zl`|A-DtxY5 zL_jwJG((X3T92fmY?seIS9dMo733|t92`X~#{Z74S!(KA!AND!a-&_YZ63NO+< zo+EqUj~U}ItQzt8AMAd&djM3hD;=3s@Ugpu#aV$vpFxNT$=b9JgwA+GJxVbhbXLli zIup2mO=z!M)A45gx*vTFPwhwL5EAj!3XflS??+$a!BcCBn_6O2i(^?b02sLtX2T<~ z7#hXAc*!e|M>KYOC?ttx#3RMGab$=r5bpOGg9Txo9~%~&OpqFB-*x3^E+HRc@7**U zFht&~?n6f@dJv|-EdELh39if(Ep)T_%x??P)$ikS$*CS){8wS}yyTtDrDp4l@6`nG z^bncy-yxrmY^tc$hc_bCHk!2J+Yxf;o-%=!b0q7-n_ixj%WfNU?@e@{`h)!Oa{jeK zzWCcj29EEZY(5$8O9%jKzCN0#1lW;mUGux7%i}c=Qthli7x7!(ecUt&HB?ej3BOF@ z9C%oJ3U3I!ZBMFhe>VEp_o$*1x!){rNPr zw3bqIeP>`FgvS@Pog9}pu2#Jjj{xK2C3AR7WdA$UYb`<$voubp*hg@6E#z+#FZ!$o zeGFM59%LX(j4UQ}=cLAOYenqTtkf4}YzO&42g7Z3(@*9fvJ`95bso<`l7^ki)5&dO zbY_TCc>BLUnu$lwJV@78=D8emfKKGkE~$AFV=T4vjPZ43nm@g%jjAixY7l3C=b!o8 z%y0NF8MnyP2t+m_ZW+Q)iA*J$SQ1k}`l?=w?QEpX zKT(xo{OgJmZqV zM&Z9TA^x`^8xw*|4stk`?0nJj{p*N9Qg0F3f)pI}NvK5})Zi_Xu@uwOgUvUIQaFT| z5XroBUqxV3kk8lc?3heuOqJe0*m>_(Wi=xIad-NZZGi~$(Gle26mnuGBGdfEN%~r+Xg(4keqqJ_4PRVVK5a>=@wijHy90LTl%np zxKNAR$#^3oSdRk>oZ`zYLe$~P0g$y7%JM|(U~by8dRR3ujS!+>X5mzGuiN64>tVW{*CQ_wA;By~72 zCnvDC7UdU$);r^aYadjddEZvB*3(9X+|3_vCq}+3@Add>F*1=0xk$?+gZ1QkJLWwC zdlF`P>XfrXH%JZeo^}1$fNDOxN(*4$?@y|^?Mc=aHo)c@i{oO)A=*=|Apk!JOhlGO z%R5;rTqQ;&R0UU}$JA`c^D~|rE@y4t52Kl#YH?o(rU9NOVnf*DaGSX?M)f3N@1O0b zg_5f0Yo{W&2LQEo?2OlP6W!VJYfTHlnIovh2rRO7fjM0W}6@@JlJX3*KQh@y@Tq)8~JWve{ko_tM2HF>lRFWd{c$Z$*F~Rj0;B4Oz;Zgue z1f<#07|pchFifjCqf#8Y>yNK(kgW@wjM=gei0A~mVT@KLCQ2HK$4xo2iSAFujbD7v z$;?In1B6o1nXhq4IVqa$v$ptvcJQXJF|r?;jM-cNU|+D-4Xch0We42O1+(bir1B6% z{*c(+6bT(I>Gd(*_r4?>()4Si?J23mUt#(;U-r-Rj8JHuiz+|7x7CtnwH>&Q9~6_3 z`~dJ{nnPJ%@`r}D_3%RR`3fVRr>vH=5>WAGWGJhB1&E(#7dmCoLL1pFhcPLoGETo7VDS`6FE=_tm&OkV&$TJ8Ivvc^5h^PPvyNMS`{)f!$AUWd)dxNd|_ z0fJ}HR5zD|)9fYzwyD%K2&Bm=bI~AUS6p1H!9jEcAh(mAf(h%fnzee z60$qfm+cV^P)nM!N39{&t(#7U=a0jW4TIt+DoR{iviF=Tw4dxOApHZpb#JRknL z>%4YZ(F(`ei|Yqi+F=Z9cI`n_j?=SI5*9y$BPREZyIU-(S}N&;E0!j9r8EC!y?N?+ zNeiKnmvp+ikTY?GQ@o3MdWGk4J_r4_w|wGQf5CIUXF@H{^w*Y5N1fb#ij{>nJ%8Kk zZBea0MwoAIczsGU5Jqh8Qvl-mygIElm&0ku*O9TnGCRnZVS$XRhnpuvF6SOF1l!>Y zz~w=HvQ82ChrlT>rc(ft+G+)cfF6JrssE#5giu8`O4^sj{mN_FPXMCoJjd(jDVG%$}$U-7_b<8OL;JCLR?9h>P8=+HJN3m=tvN zDN%$4h~3>WM35hK^eW>>t&Y^nj4Y_^7r1OOKrA|prKvxVfcJ)#I0r@tzf31%W^G`v zQiY167=mlrx;({|H38i9fJk!B41-l$2c_H5k$zKbV>05s^U<*+U{Mzpcm4Xn53L{T z&Jt4OD&>(rNNY)=mDTY+n0|?0aJW3tAUqCSOcc``m(%GUe!kO|!4hjDUgufx+b)?t z^6&94FSn`nQE2>H12z>_kcmrRMXYj_pfE+PgWE#fy%B}aW=IQxd9JuRtk~maD3F~+ z*XarX;eVuGr5X)Ta5eiL(-$DKI9HoQd4^)%zS%)Y{K`YLew*sAak^?6|9IwUeb#P1gX(M>af+P%>D+?e#+M>`@>KpC&ImO}ZzuJ8KXSX;SMazu&&X=V zquTG>(tI8awe{A}R}WTh*HU)pUip4vqyx}7%~<+N)XXFKb5ii0vLR6c;)T^{D=imM zaK^4$;IH!HsfP{rzXMFYH#t@pF9y#CDvggf9+kknDY_6?8RZ+Mnhu5?tH-_&i_6Ys zdD(J==i2eO(6jJ|-FWx!SB)KR5?|UcP-By#UToNcc9k$N;*ESooBl9gMI81=6S&GJ zkCS&i2F((#BAavS>l=p;d#B^pU++v!ye}OpW0&1iyu}B~mU~<9 z&dgu!A|n-j&oh|Pk?1I zv97>OryUB#MKbn(usqY}H-18aOrOT%ecdZgrEOjsY)1wO!4(3S3S5&>#t$`#lTc5^xqckeY)^nOXfic zF_%d`JWd=IyY7~0)gFg@$kStmom0J0DP1u@%phb# zvJMK8KM|1`&RHG$gHrxP(OC>UzL`TbKL)|slh;6*+*7ZV15p}u0>lDBUtC<6Kb&2S zX<{h9UqxB9k2~#A;B3}{glVs-BE2{s;k8lP@|4m}UyqAQ>x`g5P*DQx8|?7_{qe2} ziLBsX_v3-DhJWWwft4ZX|BR-!3M3Qv)##&Va*ka&9Qf>MblO3;B(Q4-MdZ-}`V#QE zmWiPv)RaadLI#r~=>WL{WPk(e2s6ts+8o-5T**;6e-R+2L9@uk*MQmvD=41YJ z#0dEop?2=jQiIpe{0N|etx&=fAGsYi`tOR6CLUWWz;!GQZjw9KZelLzd!T!q)Ht7L zOn}~GnfTRPgdtd@5ER0qNJs-rc!3TN4IM7Ch%z0LVI!jz_rPSOh<*m;j}(TeYl|xN zTSy2qi-P|Y#lFQLN<<-c_uDFX>UJC&+H=*+M2DD&m4-H2A>QgCHlxxG0Pd0~Wnw8| z1elTA^A%P7Z+&VPks;I=Q{?b9x98vnn@~VLrr{z0{Jch)7~H zrLN<86#h4(^@7_2cAo&K!H*+#*Jveu!UQ=O(GkxOC;z@o7|tq&&5(5Rc20SwRFXQu zbupuXG1)F+^YU4t*tF05T5Mzd!^b&j{DfdLX1HsDCnb0G7S5xLcMo=X!AV*l92=UF(*tbW?9damf>DX z!TE_CF4?z8!W4z)3GA@p!s3KV!>a)TV$<5IS~9W)NiDdF`lw+#wh{xy?L|tiuHQo7 z10TTX3=Yt+SNRHg3lGabbfI?qC=eSK9uRY?%KE(SG7d0MLr1pj+Zx}^i=xUzW+>JH zB>mj{Mv{a^$}NqG5B`vrD#hRfKfST;?BE#~(5bVgHdo)N7Q%;AZCWILqc$q%bs%5G zB6_QVjej-oBF?6o-EgLB;UW_Bg&*G)8ja(xM^yaWzMY^9snv=Uz_dYxsSsdQkF#06 zpNQ1`Oq7PfcMg7Db&-T*ZZ6sbfG^afQ0ea&ZLh=`+z6ErggMn^UYDeZ)EJ_?a?K?g zNF+tgQxPS@NB~6`F4K~AKI!(>`G@Q_X+C=C!mMRN{1M=%@|D;I`++F!P*!fpU)lJ< zP~|(lqh%l#zyutJXhaaAY%C^$0r{2um(-q-*wT z*uWI>^y4#-5n-%+*mGv*WvZKAzGkulS5VRfRjsmR^TYbCNt&GcpSBzGK3p;ROtbzmN*{P!1)z!eh{WH{3Gz7|#>z!~NR6NL^NVx@1jA9l^cp*-opibhXwdW#W&K%A}Dp zU4v9LQMb02tDXefN)>jF{m&AGSzD*K+>Z5F)L$?XWmKDf2(xo;@#@UOz(+ysWtE>5 z+=MG+jQFW;MOaS;@NCzl-xTA{e+u3}HeYRiICNQ(U1xc=tyS@B?{XCuLXI{%3!_3A zWUY_ukjJ1$0@$NndI@Jl#JEb}c$zukAd}$~usAg{)AdjC^W-@O+PuBtgOIqqZ!0_a zpfA>b2AoG)pUAo5it$>%We&X^@A7s8u5XlqUPzM9)-!7(_9p4$?SQJ~#r<2LT<;r- z-7H+^DwiX~47yl~`HN3gEPB}u^h;`z%qwRBqNVu$R3{{*36~j2W?vK^)jo#@smXjM z3$@zt9o!({^UhebeJWyylNB1-7B`PN>AOLG6H&Z_1(0xN0-L<_1x-S5k-6GlE9ipxmL+#6fU+a+q7qNFWZeG^Ku>EF7uR_GH*Psp$MowFPzi@_- z_Up5dDpU~K)d)xbDEew;SgSBB*8jdVIFHJ|b&<&*Z*@4H17J~&VcJ=G=!qW6d)C*H z{;pSVMw+uDEh7`%Q*a@K%!>KZ$z!K=yz`MquY6cN`gc6`YB3$%kz`oHzp06t zONxT`e&Lj%eA;e6IM~Sa+cH{6+ zmc52#DbJ7wY^DE=wq3fcvvVK%a1_0}EtX}Xe7=uxezZZN8GPnU&?bp75L~@AH-7Mv zR2=KRofqld#3y((WyPglxZ;Zq=>!wKus~V!8zc^Afbt zrCdabA1};&q}sT3WJ`|JVPy&WP=j@HP*y&0xJ>@kxfI{%D!k}1j;@;7&&>y3jk6@r zBLBY`%l|SrlvpJ_2QztA^HrdzPVtdrABl)OeqB^O>UkX@_G!bl|ac|9Sa_76~Qok728u%Bn~TlGvx zuv4#gzl{DZ?vG7~Flyt^=>^RXvM_4l{(bIw0t7r`;zwg4#YyvI`A?uwt$EC+O_!tt zUe_CVL*r=_N1Go#u~X`5yaC&eOLKN27nh7gr=#ly=oTMiACUi4qn+)2S|OMDS(pe2 zG7;MQC8oo6=pE%cbV3^^S$BXgZ_1QZX2P8O{PR1jlrU`u`M>V_hA zA~>r8mVV)Wv!n|)ChBov)y`?d=q)3x+Kx&he4MYC20jJ5bun;eo)du3Vx>PbRbV_0 z+kM_zob6n9Luf95Va~LY7}+*AaE?TzY%gwq6MH-ysycUSW3b)tK7%Tt=Oj#Y5@Pn6 z9Mrg0)lPPPm8@V5J|ZK*fM{#@;^+_1vU(z^4wIH;?`f9mfX$1jmyC%3$aDqLK%dFn zAfK9$BmFEpRy{{I+$W|^-v1uyY>k}Ova4NtX->2W(?H~$2C#_>&BhCJ@ zT<|z1RTi#fw5=AIKSFWYt zJD&-&oCL@hIS}xTP{)gphl|8*6WJZ9)6D{ODDA1|ZA2KoR?&=c(dUt*M)FHsASA%h zMz?#&I~4vqRAP(v_a(TIJ($y0o8cW%=;I;WAO4V$JUDC3_*^Je5wmKSbkrf*CFxkd>t)lk2DN-w0tTAtT=0SP)ki7xf`?`38b znFcMrXAl=i`YN9u`QEo?ErfOsAZ5}LbL{VR@oK6i0JHKUSuY!dMn%P-9tyoFG^oro z4?g77FPJ~5bryjII&Qq2&mOdfnr-Ue3`lFhV6sox0u5%*y_Fn;LPlVi5qid*%#e5) zSy!9H{L?Q8C!uC=50gp-eQ)x>e2>w7TxkXAi%=pe6+A=?3EQUDcLOC8> z4(Pjru@rurQvtiBT@E{vSn;YHl)~ks9ac)ujlFkECQ~m)Lbso-m37{{y?I`JJ-2fD z9i*0l+|L!?cb)SHR-xEM=$w?`*`YPWr=P`_EK6Fg!CIe}jG zFB8OjkJWtkB~v8D{WE7hvb0-4xwuRpxd=koe4YfxA9%)OkQuNzwr}DzHTo;$&Q%#f zna&E3FdwlXEe2Sy|8LZs2!md+F}Ms7fJT&3h1VZy2?1ly%I#7>f7F~mpdN2Yyf8xI z4JP=5=h0JvsN}RTEew)uSzssa`6cy+U&mJ;)f;vWm-`=d!!Uayz?eJ5G|@t$zA1(@ z2PhZ|(GyO+`|i2_Z`ueM!tU^R`dwu2h2falx_y_tVogl92uvB08Ev^Wz?x!(KIcc; zdBn>8HrH1Zk(KaqvG7;$^Qh=8SUTYA7KMk|h^rOs>%Xueu-*`B!8-&2=AjZhSD-=@ zU+k2ab-M=&$+?QP9FL=mX}+qD6RP6idfG3bg7~u0jRzJrJsJ5K&e^+c^LLOFzKRI{ z2pFnA)vaA$z@t7&Up%3yXua+4N^m3%^Bj#w$&v8icYVativ?FXa?v`6ww&g#vsE=!;g6gBN^@C=md~ZDe%m`MjUEQwFvG4_;Sa)pV)tXDxF}GwWS86+$LFSnFaBVY z=@S7&hUmIBg(SK!N<;siYv5=18V)xkZ?Whm1jU<`1@YK{p(nve!Ou=Nree^rXf|=YN!T zt!m3kYTx*9!Rn4YioQWCC~FXN4kue1Q^Qc%)Cijjh)9<%D@F(}UUTa5zB*jgZ63LL zpR~xyC_@~H+OX)Yl=OPLhtxV|tDrB~0I&ND{UtybNnHio@m0us(x6x2iih(o6wfM5 zm{Sj3G)hwco6B~>HeZ|SJ5wPGvwsxyg@x@nWP76f_!W%hhomlzh2{L>3`DR*h4bw} zW3APv8-{r1f?tpLSzXm?in~c>UFUZW?%TI{*I)e{y3Ib>FjyVW=ye?-4$f@M)Injy{r=w*)_*fS?X=ZN z(NCKZ$zr0|y_@N4vF>k6Bbu1K9reKx}C6np5`$J1b z3VUz@=8clsM2{Ccuit&wV;~m6k_Q2P1g|dS79oAaeNfY+YAAe<}@TQxKq0DD8maP%KhF_dw zWnQ_`O`uUujd<5d{JN*K)vik%Ul1Xa4W@@1zk zJRvQUcuc=*PnN@jL-j?$e?HJEaS}u$7kY075$`pdR~N3m;Bj@}RvqKUT#sASu-S>H z>A#O+&4)sbUw3_~5V=mbhwlr2hPH10cJf;F;}v3uW3^73dHxjo!Na-5ytT#RgV}-xk(Ba6aLB3tuee@T%{1RE4&#e>O zkEMC;lghcugljonivuZmQ`lWHs8Xk_5l6?=K`4&JLT9b4`8HVsUsQF~WDt83uK9}9 z4we6ZQ`~xFHsU+_80vfVNoNKw)^tJy^Zs~zG~zZtE1Cx(dB;42x{NRG%(YLE_9>jQ z?J-R}B0-R34Bo1rl&8~Uok8(b?O&(AP%Zrn;jPvTcp#Yxu$H%%_*S|SJowMt`#QEm zE#h#@*%Su@b?#GxHYTq)h!po+-cGPPI0-Ev#^gPtk4p=^#y@%N6t2tPEXqd&{+4|N z7$<5pmtX5JlW9S1egvC9@;$;RAB4hiu_~813WbEF2C(3lwePvh&bYD;J&EtviMB5N zaTb{9+L#adYZ)t8kz!RmpOm##(BJ9ktE+y)G2vi%M#Tlcn3yTaqkmX;dO$VWj@ZN>Fuso{k$xBEJS5SA?-+_k{( zhM*l`Z11BS(E$>${+w}uDQa+t*RwO>342M%F($o7CSq~Q>?)CLts%dMW41779CU};!28B z!Ke>D_BPBR8Th@%Lmdv`(uG}o;@QC+rw4a#w)($fRn)Z-&6ON!C=n?AGBMOnzgnLm zQ?na*LBQ>&mKh5bHi3O36rtLvr28E8yg~qNNCvEbMYOjLe0se6C;B?2av%D7?Q#FN z{Wa^U!)Yyr>SdvL*neC~9g)Z5$We+S{j0g8oHzl>qmYI3pXsv>NfxC>O5$KqoQDyX zP>P7h09?zIuci)Qf$j@&p}!~wDcB;{xov^_m!uD*=&RkHoGoVLJh!n5m$UEXR_`xn zIeLZgm9I69V1ke2B54Tb#j?Mg5ZwzyR`SYAS1Z@vpa{c5Onqm<*p=04e~KxNFM{&q zA^|{q*Z$AmpH$B-SpmTa&C)L6zY0ZN0FSKKSgKppCVLtHCT-hEv6Ipl47sFR&VNQbgewje zsU#cd_rqcD*g^PTU_-Mvs#JHu)aeMjSnN|^ad>f21(K38p|h-T!Mq7(i59kh>`(rP zn?iWQ4^?4eKSTgfvr8}N&n=JMU|B|NiOJo(Pnd~(5J0^3mUMqc7qeL=8}+f0TU07U ztWw9r)k3;Esrfe7cQpH++JV=qYI5nuu!?Hcm#ZCPnjhFawkPpu>vP{@_is8ukKhM} z8F}$fWa4H@0C5W<12k721K7hr9s9rxbIc6^yu(7?7PztHsj=PgpY-Pec=ya1mk8?k znA|idjb2k&+E+z^Kz0Ye*07?*ThGini{MmF>Ez0F(gz2%U%@Xjbsf5q?iM>i*u5Z05WU9Tq17dM7nuqMP#x z*$5aD2J^a5FFlNM|LYM70~4O(u?I-d4(h{Y-bDr(lK#KrT57Jg;PngSRqonpfqE;F z%Q9vMh>xc-05g2avVt7}AVqWjx2=iB&Fd^J^u(~9YgF@^_OxbI)b%W@rq9ogIKVXnpmov*J;1H!h?sc$-mkUhJ2h zs#DEmlLfkamhHlEvXiSl^;$C5{eQLviLJ7an)#2i&DM|7aEE*dF!L*7dz~>$Pkudh z5@+5klORb_*b@Ca9rTK&z&ZvHDhTO%C+;|)TTWl@&_c_4V&LPpQ}uL2ruV7&%3vVF zN8@;2C;I-a@-KOJK_EI1+=_>Lw1yJSffcJ{lr2$Xyz z&#Wc0*RIcQ_?8y;iaHXPA$#2`ZEqdw*XswAH=>xyn?Y{Yg7u#^R_cHA{HfjCYFy0{ zyb8S~x?P;A(trj~gT|~s>zPKb;q}8q^Z-xn=TH6)U4LD<_t1^Qa4}pvzmnL zM_Rp^jQKB}eIEnKUpx-xx*N$rgvn6uVOPp{O8d3;fti>OsC9>c{+;hziS6lgYpc!d z>fqbeXKzV{wBviX2`4I|g)YHP^^Uev|40RY9Pi_gmT&e&e*{Fkh^22v`HeQ1$A@{_ zHGKDPZA7Vhl6(y|X}&_9^9$|hl|E1A=-XMB;0Z{3Po@`h*3(cht5%-g9Dl7fDJAwr z^#G($`(K(||Jk=&cbS!Pv^PfIa{10m2CRH450O5zUT?+;rg{w|x5^V?u$&4C?h59# zIz+N6_Iu&&YpZ#WW`NZPTTTQDK?!waQjMLrl@(+eRmTpTGEU?4GaH>vIokd)+rre@ zQ04!P&r#Mf_5MM)+(xqK5K4i{1d)i`;vt4a8F9d<84YSXfXq37(OrXo#nK?mh;$j3#N zAC_!FbSHc~=Nm#$s(yVQWj`UI%vcZQN{^6^8y`qofop_Cd3Vr94=X04w9NR6GzFa` zne$D3%)_egj#n6HF&Lp_@rd*Wad@{Byv>cPXj))elH z!R*xwNobAif*bTDsy)%ZC_tL@l4sR&0tm4%;EeMTRrS4x12#jE1sF z1|uG~Qr}NPjHn^Nd=Eo$9}kOqqzTP@Ec^@P?2#cN^FL8n>3-OzR!DL+ThZHV&FlSk z_*2u#z4zOn62VCILyioOcBLu$mg;fbOTc7jW7~Ck%JwVuk0UQ-+__VM{m0)dS|S1& zBFC)GfNEuL5wX?a%i@0zI07z`9nijb@n|i87$o5aA88-Kh%Dr2C4WAydCjE*_}OhJ zi31W8lc)ZtSF~uM_kX%j5V#p|8>K}2z8Zz%{sE_M`}Mx>b|?FFP3dBzbk52;;#Ta5 z|BBHIU{78dVcuVWoN?|C>*}f@ycSJuPh4+$4suJ=!G$;|g19V2F?%_bA5ajmN4_G@ z-IsAO`N{1gCSwuG2QdC8GyjW^D;(0-0eR~Wrd%`r$^FqubXb#D=%5O7c{V63{G{(k zo+garEu=&#`_$F?t0Gd#8A>N{i(?P=w=)Apq99k&yxuioB!q~yzQz;!5H=}h5Ut$- z0M{HLzfcY9w*^TRfw(=GXd!3VEnYYjlZ^mMHGFG=@JTu6C(9MugI2FM9V)y`;3j3u zt3P-br05S|hoBDRqzKFp@r%YueD@EM!WxEQ`|Q2(A> zIWZxzKK(yy8{ciYvaR7!vw`?e(4s5&ak?2|0`~g-;=MBB$%?RO<-7uNPJ52`c*CUN zY-|moh|DQ_(;Ll%ou}Gk405)~D66&-sBp$n%fpO6|EUXk^EfeoCE{e6*&XK*xSefc8`BuGPcS&dt(!3ko z6}I0y8JwA@FkEA0zQp#eonjnxt$A@dEb}_%;GpytOn^?(#V|HDn84{T%l!@2FVOI3 zI$2e%v)LQYXfT1LWQf5rnzh1cKFe`xd#+2a39f00+zj%3~3SI|2B1Mx+VfPeb85 zIyWtF;(KQaENk52N>u;dO81P&8WVa#KKB~L@j!|9wa z4$Q40botpmNWX^+Dx__)vH<;iks??grlFY;FHyerz#M8DB6mB)XTinU4wn%{(}G@$K%*MBn78mr##dToXi%bjnab|AU_c`aFX-so`{WEh zeN3YwTZ1dfzk)-D$#j7(h}-~6q8<)>Z@lGxO#7KeL&Sg>Z~IGX1arABm9Jw6)Laxi zo=AYg#CiM+4}39?v!y`FOeg1$_qB_aP9G~TCg3JW$pP-;-<13Di*|sxZ<{pSHeTX4 zE|q>?YUp={^xpo6?_o1mB|B^IGJS26P-gdU>AFk(Gep(>yi8m8;x)tfr<>y_tzwn~$YwB7k$+tB=BydU` zn=(P1Y*{MAT_eoNWn)j_SS}dKSm(PWQ*+l z*Kf0Kt%$E~zg4Imytb8h__>skMT=N5wp?y@z%AqNrlDE=%RT<$cZ>TGMV6R8Kd>g= zQ|s7mM8DjRryMVh%mu#5COYeE^}X!|S{`v-Y?Pm^#$<;PB1ef64wn~;*)w(pnRr`! z!IYccJPHe6*l-$?dy?k6vkD@&2Z%){#sJ4o&EHEI?9H-6N>>rik5?;b{p;{o`JY#6 znz=gsF2pwj5`0XPLg`ewQ+{#Ql$*(k$%qA>mT@k)z9ALbwCnsoe4SN5Tv3!|t8jOB z3&Gu8f&>c^+}+)S6%ay#yIX>L;SL1^cXtn7xce~E(>*^u{ri62-@EUgv-et`v$#rW z{{yD--~I!Mh>;LwrYB$2xyJdUT#$HJ*-}emyvZdL!OzGi;-C0DYt@ET{F8HK9cN}mOZVs5HU%JUzc|x{Z@bxFL2pH0eXJ^? z4xhkGuw-SX6Io};v7Gskg{ab`S3AnkWg8P0AsLrL%2c;DSWV%fRcU{}unjA>7e;j| z@K&ZFH%R@!I{k|9&jy@V;m{(y2_d&x8=A}G_9%Na9v5N6mbIUl$-dW zSp<%<5POH;zu>^w21(f{dv;bo46_j>koHIMR@~TxKdz&&ao{0d87F?DzzTto9q1|| z%xYRX!^Ti@gl>`HiBX7f{J^hV8}lY(k+EsXKsv8a+C0uv3%6h0l9`sBlNkLcu8Yzj_H*fspJCAe6IdNaHUheZ8f7UGz?{LII}X1JQ_Mb#uR|y zlWcyi*SVZ=g!Hl=vfde}hR#Sc(_YWEz&!Brb)N2Kam3*o)NBTju6Z+#^0JqYNAG_Rd@#5|O0UAHxO4J>?>NL)YOuXH zXy|jpNJV%aKCVQ*&R?m1h7d|65d2b#(ptuiKh8XU$soWB8_I#^Aw7u|j6w385SkqU z`WzC=g)cV+U=MfO0Icj3GeTz(Y!6d0&gav1H6w^9? z7vUcA4p@94Zg;9<<4Wy#?%_>`Yd)^GMDwD&((}n3bnuFE$T!HxP8AjxW?snYSVr77 zBe%9LvNUTa0hS`Q(%2IMEyA(k(_j|06)7qZG#5e2^t`w1i8YaTP2{x-7tJ(}zT`ob$HtDa{7lV5BTyFJsjkIqgr#};VZXVvS0(<9lt zf&lkZ**PKbR3XGFH`1NL7@YAkzHFPo6~ z!_-5YTMlSlMkd3PMDoh<7ebSyz3i2#Qv#ge6RO{&x+bknuZ7q1os5CRpi)(TP@*Vc zRf+L7Lml2l88uT;wje1O*{hpl-q3JL<_DVw*xhp3mD27BzSOrjS8^Cp@<%g|jt8oS zJW|n8E8yO)&vVjB3smdIumggj;C>YHb+kkSynzl0MjOhHW6OPPE}+w3zmV zJ83u=mW!)DJ!5nefy44QRPi9r-_+j~49V}C`rUzzs2c?nZFES41~xctZ<4RiJOOMefc<-B8)hW%QEy~;*H(5Zt`L;7MiJ>vn7=Z@&1DTd~D{oj== zxvpQr0i?*9O>(k@0&sD-*)Ar13WRiqf}ww%CHJjy+9>*aZlA=XK)+l#eny|##YT-l zXN=>zu1goIzlLU?4>Ym(ItLhx<(4+K?(HY zx=>ZOU~a=AhiPmnvRIEEP;dXJ13NE^vMd+;MdeYR1P2Z0QB%0Ttvc2l!<)I$RQLoT zEq*<+?Alc7BpJ1NUiUL86$|3T+==)4uVL51hDIjJ=L<^N6}>e5#_=s(Ur(|ff3{^d zfag+THDm#yYdPJA+sNsY(*`#mwCiZc4+5lWy;^^ozCXRQIL3T6UV83&4Xdx}>NpGL z%fg{usp;Yk^}c2ne|is^R6m!^){mt%Q3PtRHZ#jJ>-?CQ12Z0H6J=v~K#3qayn&np z_iMc!Z!(&qFw73Z0poh_8q#^dcxyRdi?VU!8(QZQW{7omG-6fL113j zWkuw^H-lIX`5Nkl&X%p7Jkimm;CJLI#22@9TJv1vZqUNOmSK-R`asM2Jyt557q3xJ z$5|AL!|nV{r{fgn^_bXui5yvgnMivX4$e=;+m(}vq7o9b9c($Z#Bm*K?RC4|S0ew0 zg8St%y@0vj=wJfAZCAD1R~VrC1eTfYygAvDQF0MdKnrQQmh`8%IqT)NI?u**Jp3_) zfu!a1vBFR@v$&WqLNiDhgT21w$Yk&YF!AXpLuUy6t2VhcJtp3i-Q+2h!egRv zB{wG56igok9O$iooNl{T#fSOY{Fd`KH!WbVbzu|j41J#q%kKlGsYhnRfJf-xg@P;hT!J89u#csD#JmLy;{CO!8r^dZ@r3(wp z&d!R$3==xh`T}fgJ#da%X>$3(g#fnq#Tb_oV40?6ZL!%PwC0`vLXh6fJYgJpJYx^H z*3DV`L3%Se9Bs;@EUY&?G6%+@FNU(O-qJ$j-vtoNT!rdwh(b~8B0Yf)r&&*=1H=r( zUZLrJuq+U#Vh)w~J)B+#(EU+}WY^oE9a(M%>xUw#0*8Z+pYF-1^g{P`+J{b^A%=m&vZsR3XC&pD;`JKgc&x{s!yv1hC|9EF9SHtvr0F|N zh!;+aQb$Jp|CU64A>7V^%w~yUs?0!5R+39iA;H<9QQC0QHvLs3T*;Xdxi}%a%Ccp_ zWnu}r#9-M1n`h)gL@x?*ABB0Mx)&f43tM@tVCjdmU}S&Vl75jcm|<$Lg5oa;!9ZNm zPmIBa(xHJq%X4Zl@ zVs=+xxu;c`8lZMa`mOjn&P6`>glO=myPX8-9rozA$!1y?TX6}-V?pGt0FE2wcbiIm zE>=O~drOiQgjOH77Ki2#+7syW%7W7gj4HIW^~J9+?B(7+=K4kqD%u8l9O(etI2LTF zk5iE4&7PDIs%c_4m+FvOxy~qo;bQ>+0WjEp zPINsVDlX*l({c_G#oc($L|+u=%XbuLxv$h*yrwL1eAj-c)X0;GShsDt$NE|`H5<#I zLldF|w9sJ2%-Glg9)Ai^pIm9{FT8+@j@0p~DIUcimSnT?l-{`&@S}$KuqDaaSqam) zU%HL!;R&j8hfCZw*rNKn4;ZflG!QG85$98|VO<#EsA8zobwr-&=7LWl5>3Stas+!H znfYiH7Z?@t$?xG^2EOOvS#2Yl{>&lec3)TwY&4r|7cld(H}5b}CH9B{bBlxKpJ_EVSTIsW^jt+*K2pEH3N# z0uD0udg3OC828=88zY*_!BfOrFqO(g*=0jFZTqHDO`+EfAy_|bUr~>#~004-J5gzG5gw>p9R-y)Ib;G3g3i;PqiQ24oYEon z+8;Kaf3#mwzI}9idfy9h-Q~We)}DL6=2SdazhyeSo}P^?Q@+Hn`knYqJs@b86nA zUL5hZ(y!L$OAMwJQn4Lv-czRCKkqBMAXf`Mk`kPaDj!v*!Z^^t0zpZouiCI8R&JvJV}pXi>(i_~C&K8nyD6DAyG zN$H(OOsW#kQ+pa<@E!2|)JB2cSn#8EmEr4Ts%9w->c82U+EcL^ZEaHB%Z#N(we_Ye z^JnwhOPO?pd&`$0X7v!)XH&(PrJrM)@CZY925>&Kpt22&#RVRGc?@aOh=VLuJPBqr zn5mWPS^-Yh9qo5VSAb&dCJ|9ErS%XkLy-_~RT_BK6QXU|E1#t@7cYoAo}pMnI~Yt` z8Nkj5D!+8uj4KUyyTap4|6uOq zAnMec-etu2(;Re|_KVeg4v&b@1WIyxfJ)RBA|b?A6NX^T?MU3D{h^NJE{r~8jlkM$ zpJ&BDLiV8M%GyHFx<@)i1QineM;~!VTRAwz&FEvRh|s&5hAPr#AILm9sx9PB&m?m$ zN3vur=l)KBVu4*zHc|VA0#ZN7ih75(xM7#%j`QxS-*x&32ls~hq>zR9SNT7>s|9@P z3r`&2ndwHUS1WCtT`8ce6#g^E+J$g^pWa8$$UkU z3f~?1+oW~{?WeY2Hpjd6Zm}sQQ)33f!@+{4Jdl06QF0->htDC`x1kp(tU-@htoS!f z|8N0gQ@VplTYwbvgb3l;U;p@}gLidg)$C+yIkim;{I}+xWBB=Ni?aPyt6{I!A5@SH zeq)|v617FOgyqm-D#5_{JoFoMyAdmW6;b)nW0Be2nIEjaSkaGn)PTzMmK>v{+*X*48}T+K3Ig z!ND={52?F%P{j(;GdaBq>hT~yItuHoPlgtBVr82=+fdi6r$H-D@gq(8UEKeJa6O;0 zlL%%$KNV$&U`3fo&sRV=EB4=*mcoJyRzn-LKIiLZNuueDNAeKpXY`mWuB8L@p{ZM$ zv?F|MSh2|4o?*n|y2+x4#qXxk<@y;WIYXk;Q{weL(9uChRy+&+#l_Ty7FGK@V< zS|cWw7fh6)2xj_=5*2zdonkpGq-tsl!UnTM-+gWY)2p*jx7dn=A`;=ELc^dKi+IZ$ zgB*ga>_Vp7uhhtTMfR=^#N9FHf-#u{yXGHpf-G18r#=)Rrb?pjtA|hKImh`r`#0%S zA8VPhJsK$6@a{40h&o)kJ|oP|MY`l9lczEjGca;3ZZ|UOd%GBJvm6BbESaG1KCIK#02yNKO6q=Y zAIBNHGY{2nk9oQNwt7Xm`yQdOH^E0{K?5^y<-7~OIA}brLny{tec6>D1{|Gcq*-hu zH5aqEsrc^N@SCKp!UA_+s=H{LPFFSW{x=s(7*Ms|>1V9fE%2uhCDYH7=hpkY0|Lcu zUH#1NIErh_!DDM%=w?F2XrcnN>8}q6gC7)g44l{lU<~6y#!UV18oOa;j5cVSb#f1| z6qiZe?i*QuK;^x5Iu}>Y$hQZ>jSC|0^1Nmw&|E3NF&GRJ+nt2Kl<5Rg;>2%PwX41# zWjP)=G+Qj7FdPbz8R8_mA|-zyARsNS|=ZS zb*N{Oru;<+YJKFGNIV};G&vgE*%>Pd+7+7G|3C}+A3=tK^)1bq%)ieXf9E@b*h1!+ zoA#T``|-**s9sHnB7hQQyb3R7Z@;C zon}q3I4o$oL@Bm#SCKt9J((Oh9qQ=oD$I=8y7M0N$|l^I4 zhDPwMI~E%yk3LjSQbj%xxIxhkFxy1v?9K;{NeG7S1!s#37^C>7ermgSnKZfrCFqk zSu`fHZvXjbB_sN7XSu19R~6$9hA%>X(JBg?)ndrr|Lo~;zY6YCKpw~HrV$ZtAqGU$;iS)n zU@on~UkBZU?!*7ojKe{P76a7>m>=)Ks(#ke2cv%qOwzW4Sak)C+@)boZe9l*KZ77v zrA!U!x7wP4!S94ma(yFTBDLFnaOT9lJPn?&i+2mVUYRLBx49$vGtw&R z(}Ayi2i*=p_&>WZ{g7dI(}BZ$lg4aWzc7O7r3-Y3Q-3L@b|>JG3VU*LsXri*`K{w{jqOYx(07jGIxm+-*0cEL? zHyF4gA?$X|geeaELW7SBNm8N+Tpm}nwcPGN;+bvp>A}A#7eM(C$@$)gaQMb@J{tGJ zQ^LOpSSxCy&LbADF#?e_=`{kPv~}W7ErzV8J-+q$Rr1 zVq7t;a%le3A4drkxdKoBynYV)G`JY%5Otn09h4~rk|Kaq@ntnKV+z#=DGJRbx;be? zkk*+0&YT$D8>M)Zn@kL&qGGi*kfw;Tji_Np0~H*SZ3U>i>y4YVvIiOzgeC| ztfR??NuI-z4WQH|R{q!?cC~_U{`>>omD~1^inN&^B)9)t=5(k=*a9+TE=<|MOMoIC z5=7_#kz-Nc1#0tqs|(MHlKpG{zf0y^&2hHVJ3kkBC$E1Fk4}(9UR=aM4%&ksz-Nt{ z-)$VUF1HRMx5=L%;p95{VAUC)KiRMZXz&XmNZ&N2Jquur4F%$RVBZ7HALZ< zE2u)tQ_#2`joS5+uD%dp%aca!;Dr)$$kw|!9sQ|P$kQPFRHtL`Lf}9fJs1hspr#?DrlGH^AY$=rsZ18 z`X2+?Tn==OpFu?S^Hh*SupKpM0PZt~NhVOhlY`g?p(MNe*$;?_;a4@oFmSxrmP?RW zs91T`CxBO*nMTFY@j{m>5{^vq?vn-O2!+g!K?-6+DkB$4L9z&d?qZ?E^vrq(n0)&`MyLLk@HcDvKVloATGu!!jZ4KcoI&F+dv ze5j7iYr$Rw)mJz>py*cI;`ANCQzfSy>A*{r3zQf*FDNccj%SO)7S#y=d|Z8H_#ls- zabV|62OTIkoH90*9akx~O@A#MTgFQiXmeGfhZg3_XW~8RJvbETYCxT~U=hc(=?{H_ zdhaco=dPggc&@rr-Uz@HkNS;t+Qh_y8tj|dh~h@AGYG`&5j@N9Tzz+^Wqc2p-HF6& z#;OtMQz&2lDL=g0SC>CIBPd*DexvvW7+NeWM0=aWO=Tat{)(1HtvU^tc3AtdLup5Z z81ApV?ZXFG2d?|zu0e3X;R~GS5`sD5OtmhNG5MlZgO5qDYb?NGY5<+BNsG_oRh zfY^4Sa9V_jK~QHxEy$U-X?J7&KE;3kuZS6w^Eb!@l8Xg zSrh2xGmDbhg+aU|;|x?3B~yzA7iE-P;OSt>>1ZwCrDuXmawO&iXOi3TJ#e|Qj~OKm!Ud%G;&^_)M@VOBnUl6JTp_tQOysmDb<(8Vvu) z?>zfkZ`MUf0v=A_A@dB>a%85HfiO>gJ*A>U8_ckNUND<2_b+<=la<;FoEN?+Wk=$+ z#Sb9#u{^L1o-k)K_c|W=v$9xq1lFYIt&LXaJ?9Q& zy=wPgxIKe;1{1~UNjLeG%2an9Jz%8#U3IKZ%eabp|4J4BcOy=_K1CU7JRCoY>|sE6 zD69CRS|3nh(-wO+TsDRle)#bK#>-dz?`l<;l6+}K`V%>*a5dq^;Gg%#EcWP`$!CgO z8Nbnz3fNiClN;@xvh#M&4IepeX0dgBl2{6C@3GNb6)(7j%Mb8Jv|x7LMNM}HOuzL2S{P=TNU2hSj=zN{CH_B>oo%?#D5U@6aoF$LHn1#&n{OD{CR>$(t{Ib zjh~(XOdWsr@}Qws>*kWLFyiSBdf#?#eXQh~>JOb87$04*h}-^24=-AFRdaOje?1~W zz)8&RmaD}w5|lJ*h*WAu5_ctqROwL9Vz7}1v|JQ#papHR%TE;WqC+eS1UnJue7;Bq zDdh}V0^U4OcaP42m=vP8l!{)RQJ!yo!lykb!G_5TMk@jI^^5?Mrdte%VdJK7m+wQu z63ko0z=DuLu6>&&r`F2Ig0f)66yIcLAo_$GUeTstb_U>l&MXYwTPJ7>d{N>3MQGDM zGMJa`*NLNlEL^iTqs2^_|5=|;lacL;1g}B9%1=N+%~xYz&H2Ea1po@_KKLt-^Rc<7 z6!{J5zCCq~42y_(U=vTMb~~-Epc4A^>W^?;Pj&78tkwKC)2n+M8wuje3P>~nqd1R1 zFw5qN=k^OxYB-%Usgc;fcSxhyhfMA+q}${Pisj@eB@4PBD>{d5$P5gcFNnNIC(Ph6bRB3VriXAD2uXBg@W$Vfqh9z;?%AcIAc(L|dU8 zZd6?`ByUN^+|ee^(q#*F(r2WaaU##dg?j^4G6-iTLe+NaY0m#zouCA}4R(gLd;+$y zoQH;kqE{FH<=CBb^7C#bO(lfji>1ndaV3ZwPFRUYDm*RWN2o)02J>Uz*X!+BJl%f4 zT9}#c$j%eO$uB0MR;a;iT6zpC5=$`c_HxTET;`;OKtu%>Hw6{(TbJhU#}&yUQ_4Jc z*3qJLCVtB-2pdsix+o6;C`G7AfK9|eKhPg%R9DT>k?61eyX4^Ar}rn?B}aMZ(UjpM z*xnK-8NyS6IlEiZM#ajEoDmtdl8#6jl{wb{(;-Z|b3soS@tOM^+59*54K*cZbN;3X zzY6as517Am0o4MGaNlY4%Lp_H$q`HH^ltG%5lY6yigFmIq;v{h11d_6|Y5zi)L2cWh1DhHjm&nWjkFH%r<`!-Wu`5fDmLN3a;&4zx2%u2bjw$ zvYTs@s@&j51`PL#3cIf!BoA;lQ|WV5!Ir?@`RmXw;QRbgWH7oReNzQP;k~AG<*5CF z@d?SCL^e0rw3x(XH;H1yUR08lDYxz&9d#rH7HY}aEU zYzJNwel2IBF=9U(nfJtwmxn6$j~rH^sj)BsXVC)>;2^Y!6jnCp+i; z>bRT8Pk=Ki*kkhX%%cS>&1#+8suSZsSpZLZWiS4h9Op#czkD`|&Pj-&&Y_Um80lB( z&CoURh(c6HDxcqumRa;zJ4RpxZ21IfvkbgWH3qm#-ob-ls&B$JF+1CV&@<`NDpD=E zeo|5&wZ8$9;5b%9OXxPI#o81V2ba#0k~%CrU7zIJYL$z_u4&uki#Cv3ha7|zD5w6T z+eq0YW_`o2^UxO11~x_wmXj9fM*tlK@ar*|WTH3r@V+OHVI7c&>O5EVKk8QU{Insv zrOMnIW>X|!ASIHEi!y`uWaV{xuuQZ}vso{=vd{CPrikA@D`vu}UN9i34p!&q&belS$9@Uxne$Hm@`4w$a@`gu^c2QtCuOhB@faKdfNb zls`!|RuTuzJ}X8sWfB{fXw=H4J{*$fo{ISZ2ovJmBBki@tCS?>@Qi54rTorjLW zssS=1^MRPbB{>RpPt2_)4qp#GF-+L!mu$zwuc_h%5}5d^3cC50A4OAg_y@9_zMjA~lhAM)`&&K2L zS@vyTiVxZ4`Rv2Jg6S0eR|!>jKNMH(xvV4_m1Z*{p#JTZ8nowcS%D~^NOM=s9r^X^ zu;TE4e?5F+=U&=b_yIpiAle%e0%E>j1chaEJ0ClR*x=92Ojp`N1#Ckk&c~6ksQZAu znvZ{t5t?_tTWr#}|DlVwolS%V*=`NXXsWzj0IRBckE15OUdG|Bt*A8fuZ`!A-}Rj zZZ3!yReKvbrVu@M$zMY3%67I$27K7N`M&*)vd27&>s_F=;%N_^1N5|B2fdmTSLiu@ zLjUVMe{S0V2U$Usu5z${KhSxBWk&w6RFi=rkFcn!*{+#x?JoG8WB8gGEPgF=sx1$* znqJ1^y&qJP;yH-rFkZ1)<=Yr}Ij=tMh=;s8f$-RYaCLx{)|Q29&b#A;Tz3Bgdr3R3 z^l=a}j^Sjp#fgL*|32`qo@3;`^9Y#mM&H8(o&l&~|OcP=%{Zr*Nu_~0VI?K-EE=T3OJUo7 znr|0dnpnsqI3mymOon?rY!0>iZz(2ny?`tS+sEvQv*C!g28nfTbG3M}9pciJq@864 zntUVvoo7x;0FJS6KK%%(f%cZ{ht@Dpjz^1m##8r6rqm|k>&yW;3KJX9ev1w?bTiG7 zNm$?g9LX!`-pZ+@?98QxC@|xwpcNw{l#vURZ^9AFe?F~8hEkKt?UW(6Qr&}0@G^9X zcm5FtrCMOA@`yN*ed)+R27v5zm%f=YrHZ5#`@=(*lcrW_vdVtXWd(?G34e$cc_Y)g z92HgeD#3%{`Dv~O<+o8}2gi1KstRTUqr$!vXK<}9#X~b(`*{%MEcR#fJI20t)tRJ{ z1fBo(F-FuW{`wA)MTbYY;BkZdU;s+t>hqp?w*qB$K-CSqY({eY>CXK_!=} zRZpb$R*`iyAq0KWFDC%A3iX81)}k(k9mUJzqyv!`Yn@;YMm`K;j)#3+kJ8t~5Hc(G z$!d*d`PdHm75x)|@g(h{lvEMCV85{zw{5k^CS=)~7Noin=$==#@0IG{t{<8pV1)r( z#Gkh{qAAcTt-L!z@T8a9I-)J$tfl`L`e60f%O}y0`0{2ot2#x3k znIoJ;{Zmw7;C_ewN?suE)O8k?VZIKPA76QHd*ChNSvp&=QD!49Df}Rv?^)Q+*fo2` zF9@QjTyMM`{6R^L4Y0v!)oa~-k3@6rs|w}R#HS-Ke1JMD9J7(AasQWf>pgPNZrOhG z%igCNZ?n!*{B=1JgS9sHhwT4UGx8S$Yn=});BVsvId0aOdexC)wiDqdG~hO|(PXF+ z5M*41VJAJQ(1SMEabOw;+3>}TNh7immkQ(2LmcEuM%~IGO_?3zhEayt%~24}`!5A0JG!JnQOIn6K&x zHq5FM%P#|NY4F@e22%OX+W`a;P8oO+TJ?(nf_ecsd4UOg?}_p-ooL?r2Z#kPu;Lj+(_;Yp76m4NRL!l!r8w;r&~=d%-B)|RU#2c45q z=Kc}eA;5tMnwED*@E8sfe6RV^eh~*a+~6&0}NkP7Q^XQyg< ztIwz2D6?}#{LR1=v&0#UuQ{#)v~6qsV17S4#j@h&mH`Q6x{_9O#uIGg} z5`bt?NFLJLu*?;kp||VVX1b`L*<*LodeGQdFwpDKVnYhfxG**(xwx*lS{12!{vHVv3Q9HYRf)7y+S-h`p0C!2j zSXZJC>=pqAgsH`5v4&c`tO*$A&V{L`YDWwjk>1N<5y&v9utEr;6(_%k<}dFfJ!bokEH#ULAs z-T6O1AUXc4k6k#(?aumK#HL9}P{|S)x?aBC!!HlFsM$#)@P~hXw2o73iQIr%U;CzE zVTW75XJSnJ!s0pePX8s_yLY2GFVA1*$oYj9?4)}e;O}m!LO$SWwi-qB!YOHj`kAs7 zF2b2Tb3d`yd}8x;#FVdr#13oOL-A&juC@u4vm-!JI}Zi8W$ z&^Vz#ED|MaE(wZO5;zw~Oeb}wU%f|B0?_A>l zRWdn2vooS}k4YKaZ|a9I|TitRi-9i}0#O<5IiY=~@a@g>(aGLhW?OZlUt z5fOJkIu)%Bg>FV1RE5OFW(aOP5zywOQP<=p04>|@J*b5LjX1ci`<;40;OBLx&yXR~ zz1Im<0GG3_$?xV0BT0fSlht89e8&lzNo%albj&s^lev0;)~G-lW|5T-z$CiK}>8*l!jb{7M}hMG35|(+Cc|9#t{XXRt0I>%s(74Dg|L2r4xQd%uT= zi=w>e&F1(vo7X6$RPkf5Eut0@LQMJK5k+fA@cPCrVJ(``O|ZmazSNtlR^^X{L4PO_ zuBVs?^O)iK^Y|Wp7r9Fm|9y1$!X2SBhbjJ4&qEt&Q4G(4U%M7fI3JIQDTzp6Gb9aB z3>+0`@;h#wIs^Q8mUYo6WCa`X;jktI_IE%dXh0ELS?I%*6-^26m zrxuzK_#k1@0@Gna_7inuy+kAd(2M(avZm}CK9?vYE6F^0Y8b(De;JS&iKzK-cO4`& ze+m9PzQ?^p!+b21`x9z51;urrd4G&> z0aB`i*iqVpH6}2eeuus}f%K4SryMl=bI&nR7YpltZ|8V$zW4r{ZEYoo%HMUO)aDEI zeOdh{Xl#RWSNK77&So&&Wf_SiZvz6`oj>6ef}1&BJWx?37W3QJ_2b$4*?h*Koo0QV z$pN!fx#nFROEYga4#%y?Uf_|HW0Xb98V^Q^#e}I$!`qw+jko!L={bbcnQXH&h?i)U zA933vH*2;U?y())Bqn@YOk2meea&*Wet=UI8~f{4MEm^>{zEr&zvM9A@v*Qfx7xbZ zT+f6k@g(#cxr$hdOH7gULfwY>6cOc}=|&(p=0wlZ?-qbr)=1v-Jn7l_Pwqb^y@8*2 zArhVlf9&W`j3x4%lo<)U8`OP%|QOQehK8L<(&IB~mj4(8@z zWCoX|em+)d@FoPKZnyx~eTR3QOj2Qn*Jwe=us4oW=)~D&9S}A!i*4Re_ViqMp;1f# zDaWUI@{q9lug50qQdAiP>$w!2FAT6?+9ea#7)2Nwr0LCoBA}23U#%oVYn8rX3;-Zw zRT#*Ha7$6!c=B;35!b{LrkpG-xTu)EO(*PDxXRlYxN&>ynao+u!?KCuncRpIe?u{` zVF9HHUHp50-+4j>9uXr9mV!$lJ|`2q;;i!AxE5aMSnO}{hT*HD9?=5Vv=r*#`mNM8 zbkM^6vZK}s-|P)V(rvSvOnrP5{4x$((Jl0Ze6xT~597i{qG6FvBb^cGgq0=(Vvw|5 zdx^A=3Xu42K{WFo?eb<6>NFOl41) zoSG*bKxC43aIQBI%TAofs@}J*9lL%(xdL8zn^hmbQ}w*R|7AQ!2sx!GP9_W_R;f+# z<=jYob4&6_Og??AkGYobPSSt}-~)k2RUq(Ur@wV&p-*bY(ozR+75?ZI`-sr*9o~&# z3H{BuD6ByO#v*UNoaVRO;Sbb6xe}h%%RxNuwQL*iJQf5G;n}Ud{Bx*?XltKrdNy@1 z@O@*|eXcwErR`?%H2hTiZ5+<)Ogo6Q@zSbSq?!m^1u=XgpRKD#moRg8r(M@W#L7=V z+0Sl7BrlF;ZH8^AZ<6k>QVD>S+V*t9-ZK+!$u%>)lBD2wgaeYZk}M5|8%#^K{WC+@v4uo=CZ ziZ%);PYv||p3=gx_R}fMbVNmke~~7F^zC*!R8=mdA^%u2QI$7E`a>N*!me1aOhD#i zeT4~^Lq%AGb~79tZ-WpHZVw7;Yo<6zki(}XL54u05IuL}g>dSrDS@01)*rf@KnR78 zAQmT}4c3r=KSl2Ia8j24;(cI7v$vu?LelS%^}`P@X1f%^+RTIUHKsz%b8rIHb8$1d z_gVFw&w?{_N2r~&tgQD3<-Lgj`ZULH8__9;J#lE{eqiSs4B^K?IvqEPrH0%*Nb$tp zE^h31^D+#V-30NqzFAz3?%rpZ#R!F36i_MwVkn)=0s%xOqek(h_iXkbH^9fgG?KGC z^dJ05x#%6)x-aY~bxmGruo;wzVL_P`Fz%SXC!s0x6=!+SgPcmx(HwnLDO%N-yj`yi ze1l2VT_kO?Fj!Gg3}zSNL(dHG3jDnO_W1TOzYmNcLE1u4q@cKaz@V#$Zl{U1hhqyfr(OT`ROySnVlt7VlwPw~rER%8 zC$1JqshRS#<5v=kqd1J8q))v9fkZzWSyiy1zg6S8!J!G2CJ8L(40$XPM`IJgUQg)R zv@uK@Q*{SH?pU|q!?|`RTAcZH6>O$7YK4Afn7zr65PLL#5IjVW4!HK~`g<5^&;i6c zh=SRtm;`koqYNuzv!(`%NXsz!_D7+=(j&ha)J>IMl#^6j8c|u5XQJ5$aRL!xi48be z^IRN+Vw`r7eBf-EvZIN$h+pd&z$$`w!(Sa1kR5^Z@?H46qCVKLIcI)5VI7~66n3li zj;;&iGweK)Z~;l8-G8np{;O>C{~kc(lwkqo9&Y_ur`-pCgIKNBD*$Pds2>ZY^K0Sb z9!g5LII6=d3}G=rONdL>_N7_%0Hq;(oL%yw7)-s?SXaeHMpmFrl0T&`HEUe-dT}a~ z0A&RcBeT3Xd~seA=9QL}g%Uzc&aCF&5f%WcbRxaIUi*%8BMw#CM0lcnB;t-WAW0%_6YxH{S*P~63fqrpQ=X1O7uGt z>4=>@*%E+(!K$3Z8nQJcxoKp_fyVI;siG7FT=IGbZ7-_Ch9;Wif}rMDU)gz+7)`Q9 z=cCOFt+6LVPlpi2|X=(=QM|!had*fjgDrSm>in z`BL4e#Wljm!%&rWnwwMDhXnhUO2YwhuhJ?Y)${5yJcuC=X?_Qt?R}DCmkFzV&g96` zz^+_%P;E`fuOa`>Xf(?wQwcu2cZq&OR!o@JJy0z^mh(KEgO}jmMwAqHbnbkYwkEa7 zxDu6I7-A9s1N;$gca5h;xO$_?NLt7lsQa(FPe3?u&Q1Ha?;K(}qZ{%JKN-jqnOf9Q zr2#Jk;#a=G8TB9YEWVwZ`Lb7L`CTSi31*5y#WcGftG|w02_)8?h#+-$|G|rmz>?Ms zoo>JprLkO4o!+7Q3+&XMg-**t({pJhe8T}|GuJ&64p}M2Hvf)2#luW4-4_Yh>u!xSts6KEyfNYc$RUld0WxM(3m7rb|^aJlbNwBYtHy zfY{Gt15+%{7Zd`9^gjP5kRw)Y;MlD1_Q})gUN+{39XMaIU{5#1y{YL}u8Ka!KFg%_ zq4A7IAIT8Dq5Xp@uMb(mIKy`?c01-$G?e8xW&=jJ$tS4F`YL9gph9w!|w zOy+atmr5^n$9i|(blBd{42I@iEgwD8G2ezn=tlsuzkO&Y$S)AD1qZ3)wK=K+c!s$2 ze@+JZ{k^+mjib&=rhe+&-sly%((~OpAO-N^xcq^R2GO^WleW~lK&$i~zX7&(f~GBT z-Dd%^B~{O_?&@$05zssdhoEBOWb#`^C~C{=M5wP(J?-4vVb4aAk#e=XU%}gP7W=$U z=^F`+(TPBSV9VrFkSgj)FB&={vRt8L`*NtuID<_uc^?JWR!`R`bv2xwW_ET7-@A^H zuW8ix)iv57<}d!#TxjH1eqLHAXidY^dvXH5(n)Sd)aH(U;$I9apT$9EAb6`(8CDbZ zFwh^O_7*LI^7?fh*oEx+XB@Pj<-me;?z7o8Gaz3~4N;zbW#r^-1Iv29y;<+a#JX9V zu9GRPv_OeWm|yf_51lGG)A9Y32uX_{$XmBLCYr;-mpg}kD5UVV`Gu9sy|9qb|2MUJ z*#||=*L1Rklw6a1W_%}F9;f3;aQFql?3|SSd6|6mW!@^RE+tfn(TG>T7DS`%9wXWj zw%7#7fknuF$&Z0K;1ly+g|_vr4Z?oPiGS9CaF(Z?*lrc(9MpZye>(kH3?b!qNhcHJ{w8ccx_ZgYKKmNF$u~4_jFp ztbt&T=WS7}Uunif&z;ofoM9`c=inOwk0J&B2gb z1fwTl$(M4dOn<6~1Y%JZb1a|ifD=Ry7o=Yk;T{VJ>J8~GH>!|zeGC=R6d++feA`47 zZ{$N14FIDul*^qxOaNfGI%bl_8{WkH&t58N>R@|)`4kh7TwK+=QqkE4 zqoS1E(@p)%3++p=FLfPPA9weMR$H2<-7B~` z8lXl#GSm=D88HC|*9lXY@P*cCY#=GdF+fw)S7MqtBs4meUYRKuzBC zSp~|&`-vX3?oL-aWi5rP99-Rm`xxGW)7vMe$7Ud|XKFKV`_ux#4eYsCIqR)q7Z;@8 zi|xu8ucF5HJBH~@>pQ@T^sB88fcFo5kuvRuIFE((S?_UWYEPFsdL>X5 z!(P4#v5*4J77z_TywnMb)JC_m%WS9VxV7?uAT8CoxDo@`T>F8o6$N3|K^Kw1&=~I^ zLXmn;K+Vtt5GB!KROGd?nRX0EJbKT8N2_coJ5Co>z-}_(wF>*J{qyxDr*=gcJ*xRU zq8=-6E{bS@$jNoekIFG;^lp}}ro8ryYM}d}&>o4?jOCQ+0C1EKo6e~KzDD;-F@*>N-g@;iTR|UoALH70kqVT;YL#m)GAwK0tMz5!s65-*(nWC=> zSO<{<6$cM2|L(FjWE_^?>gu=V^;#CVzrud;Z2eY#`eeKE*t7tE)m!r4n65>36TrNmY4u)B@qDQXB^GYiqS)>|H+E5O zgsMhpf}c69l8i@tAP~`N^a*6&O(-1OA)^O*gah=rUXg5;Ye~VTlB4Hil$E&x0yLqZ8+Tg^ z8ktm{@1qPXI|~`rN6D|m;Cg&*j;DZ2GO?#WowPt1dB82bK&EhXOb0fF>A0#|Zo8tP z-^>?0M)?B#^dMKo!o_&GrdY+wa?sX0O_`}>Bea?-2j(#?rxGhKan_*&B(tN16>Cg} zSRJ;2Yomve{8-mcZPV;--y$UUr%u|t>Jwd`Cjy(!;4hyCCx|Ise0K(=0D9Beuv`Pw zgEiMFte`uuOVc4o4=Qdt{Fma21Tt=Yx^|hp;J~)aHk^wCFEt*borv6M!FR1ia~WHY zv}yITGBUiA5P|V#WU0$=vG}CxK_nqoWlU$&y8188hxfDbhTKCBOOrYDBz=~# zpOxXKR%+(sbO5v^D{#j5F>CQGpJ_1gg}v1>EGgzc0Gf*i8A*@-Z+p~od>1JzN*so! z;sMo0Vx$y=4#fVRohkA)ZbKKkSIuG5>_2RQW2*zZXM89G+t%#vkkjwZj17aO5~yZj zqK0h_R+iGs(^3@y;W*c|J2ec0Eo`Q#rnRIv2~iKK&yzcNb0FjkM>iq|y#(;_Ze;k- z11tDSqt9dda=4|{e5o2PX%S!DA?*o|;iEz?q<2pxqU!a(uC%c5(FIDQPgI7B<&uXq zd)(~y%kg`zA%C;zH6Iu&?8U55Q`*Rw#!%+UT=J1$*Xam9^S*p6{OpKx>l#`?6XM(Y z=p@x4q81!Ay}zN}6{+C;SQM86&km02-&M1yYjgnnBHb^ut3PMd*u5@LP7XbV!Fvhe za-(WsSe0@{h%&Xpm?l=oUwtWrA;|Vv%)-Lzy_nNH zR0hSkU-*t!Xij3+hb%z2HeNYGg!giF@G^N199u`)JX%c1`H8dmc$jk@uigJ0KWvzM z&AFZ+-ux^>mGXj}dbcSQUe3j+QsS#HlPoSe8)(vdT|1vQNI5+Lw2#){cw_>3fqq9K zL^co%NXxrw<-NqJK}?6y^+u4xWtO4^BSOiCiUu}p^|jfh7yEh)K__@YqK0BS7ErehK$<@N&l7OlQA;IyZn z67n;(A;V{e{a9iz?};JAi1gf>>Ns3~E(_$)4coIW{g!y>ru_xHtd{$X-KgW80r=@B z$96wT6c&19BcNOr;Lz9Zlz@O2=-H7m%y& zLu@GiG)^;ZT^i%3&PfZ>+(YzTTHdn1*6Q~!`m2mJ`^|x&ieYPUg6S6H^4p=E+V7{O z|Dno`R18);Jeb5U-{u5~8{b?~@9%xZN#>!WZ>0Wwv`24YZ*O!XY@2--3}9i$xG;Njsj{zYfqpFkR+tuwa zyTu2fd!fr7bDGdY(c&`;uQEh-P4aZnrX>mhy#5-I@O5Rnh)46t^gqd#K zMvEA=dc?Q>rSw&O!=*?q<=?jlUNr}-i(v(7Y!*=`YzMEVo)?zCS*{X-x6~I!2iAo+ zN8Zes>@sby3)|lfpq=uR@DVOzg1vmXa-yy*S7AbTe}hdPBBlXzoPH|@gS%FJ@&XDc zI+VYg8N4NI?i}v`gpXG#bhwj{TzXW~qv~9gp$@e@Y_nT}p`EjUNFLMeX z6M1eLc844&oQ-#1)K*GPmq&#})4I}!n7%^$-QQ2hb&RHZX$4#l^^_4IXXr0|by{w7 zH^YwoO%^IHTxvP7@H@nZ*0 zw)$xWp6(qh)e&r;T<{8vHq_e0R?VmeA09Rwvi9H7+T}hO`F&!PEs_7<6h1E;Y(?e4 zv2%w2QYiKV^xz%`{BrE)&+w4TwrB=Flj?(MG2GdXe=zF;eoWZCE1yu4;zqU&B2v8C z5!H$BWPzU*{Jn#JOoc>t3CR2+?_8XnPyrHq3JQI9`O?Zx5r;0%preEs9TzQx>HU2$ zvKz%fC+MGU04X>#9}x76nto$36UNAn1IatUKnbPgk~vg=1Igo4CIAXDpl6e$pRs)3@KMoh^>j#rH^@Oc`}(>jQ5@da_Z8^%f@KQW^|B=y z;@eF()ioudO)yLpsvyZe`X!UZhma9#A8%!lM8y9C)x&{PZ>%i^1xr|8K;W8Y>t_VC zVc}~6&)}SI@XyLZ7lwv1?S!1%x2Ax}#NGE6AGBAlUGu+KqdsU;oR)7Q+$PcY zNUF|9A6)1*a>)h2Uvs-C_#j-iV0MXBsvi`P#X1|tTXefC$LSwFDvfLOVF8YPJ{??* z6*30%m`Ex7*a0fLIyq|PNGZ4%LDu>Opyi|SA=NGovWOdY+b>yZi96c;#r%Ntiu*XM zOeFVBw}2P+=F0hJy&NAudnZ~A9VZe!<0=;2=Y%}snx*Hj*yy!_l)=o0<-)^Dl~t-t zVN|p$VurO+8`Dga0GP}cG$o;U;ykYz&*kYk!b~xTle=Ns(LRqqGRothY<^IkX@zY} zlcp9G`CM=pT)}R`YcKeIoCQRfEXFDl#_;oXEFqLms~AqqwDdAEj8KAciPa=WixGud zoo5GoW+4{NJ#9-UsGEV94GM*Qn*ru8D!3T6Ae+V>d8+3Vi}!xx7d#2~ygf#{RH9b9 z;`N%<?@DDllRFX-0exEs1?*kft6*%4ENT(iJAGZR) z(Ohxn{LH>$($?;Z^0DL58|CL8xPL}fX$Vaqh%>BpXNVQr` zd#!xZ2=wNah^`5IARtF`tQSCTmbXW=2Ro~Tury7PSG$r8yFAf+N zdp%S6vD>a-htK$NxF9ZFkupUpSCKLk94S6PJ5*zUPS#z8~h$K-^BwUrqC^_&6(4FYVgimP}lc51wN=QxAV zYl}O3-f6jxG|aB_iPRIrQg0bReKA8{qGoiSA5IVH0e9i48v_g6sDGv|miMZ`i}5&$ zSUEtjSAwqwFc7FbEfLRJF_=OYx_0PD%je^~IM;hrH*#lcWv8BfhtvT021@aR6%cuPP zflY|#&`2_t+u_rrX#Rc0qgX3v`v_~tX!4o!K&Jtl0lB|6P&ZK+)I_eJlkat7I0|fd z;%Z}OY`PllCG_P+4%Z{_!6#io>ZmZU0wkg(u1aRCzW;-nKJ@-Z;on8@icyI)?fUuC=-c>Bf zu);m7du2*dChMFaPV)o(smL&Kbkli)9H7R?*(uM)CzLP@2xjyrARIo2Hzt-*@@}D5 z1X=%#X>~w^`9yv(&U0f2#k?Q(&n>T|xtpXf)oel-Jq!94IMkHyU{|Z`#mC@_}L_J+6R&hbs@lz2BF9$Yz_PWq9Cq3|`5X zEEev|M)N~?m5j$p!de%$zf}CpagT@Xzl>q=h;cOLN8mJKmZg62T+ zhFBp+OKxd;j^P)3yz$pCgC#xoTaDz! zFwv|&-t~@EuTBGqLFz6!r{HhV%%nGm_8rhh-sib2b=~yJEi8iK-4kG)-Hs0ufF=jl zA;T-iQClJuiduXxB&zppUn zL_0(3aWM?XW^-S+UEkM6yv0N?hVx5f5zz3I8;*-=2~s>jqsgmdyifO-D@q=`B;s+G zWHvI`;6~*#$2Q(Dr}Dwy9bxOR)n%K@XOCqK%;eCK%X)XX|DVfy{1LH84IOob|L%tW zq>BIFEA@W{@@rW;Qx7-89_v+?@5w}+@H5;{<_sw?J&i&Z5n1sq$G4q_{%Qjhi$f;m zU}LD&ZzjpShHcfua{e1nF0vD|Atya?HE5t zxaprIkTpAJbscLVp}!W{yyI~$allO$uUWMIbVNlFtN?Uk(l^Q}siSGpJh zH7ORx!G!q!4u-{S?ba)jfIxP-Lq;l4&p*j<==4u*jMbY>DN^rf_dMX@@drR0MW`r4 zp_y#NUNT}$8DDK6Y#pt~1O)O+*XN_cu8a~t!XdR!Su4Epqxh}T z?JoduZz@Q^-jruDVD!cxn~s1vRtzPdcHI54c(KFbdT(JU@q9Wcs7H>_*=0B??LC1Z zT*hcXo;T^`BnrY8b=(nv_k>I#y-zhl5m;aQ=SN$~PT+pdTVYnspQu7ZSDhEgDWQ zLh0xvy<@de{6mwE>xhsVk^SVqPH`6kYJPf79lj`h$jevYSnyaK5eUZ}{iVD$&|*Cg z`{@=%bgbd+7*L*R(=INCX^FT|vA2j91|;JMQkSw&l^N^2{)$m zsd#>r1^Fo{_P?Tt3pWFEm}UPY33{QiF<_;HZ>UbsOlxdM0i2a?`q8we z5)gcooXye8nzg6HKg7WEp`is^TsOs?h;U-JtomfpZd3Jx;5TR8aFCZCY3eMV-H^=1 z@i1~zw%oLbSc?)CxLT>^PQ(Klk~Eihtu=J`DNNeGY!l(4vwJVQuKNBu7^H-xNX_Jq#GIO% zg?8Q5TfaSbq#t(_+i*witI{QBA01hC!1ODj$1`NYL$-T=wP?I-_ zwu$;^n$ZIMIJw=DlX}2bO=x>U(?Jc^!^f70*Uk{PN_chh;274}n0{N|@j(l~%g4|h zOYcs2yTNV=M!X=8?r4^gk#5#a!95G`|0@NLv-+{eDyhv#{^@ValT~nORm_bTmR`6% zksx+&ov`p9@ufIQC-dRd<>qkt$w|I-NWAF2s9^kc zY1(+pQjo3!CisW;&iwxIwWNEb3_PGqoPv_iXc#C)u!=&KK*|xc@=f{SwkJCFYE}4^iklf{a{r5m@So zN^^dDVP)(!u>R6C!vNh=CG}urFJj23`4^trU~WBKS@Ko<3S$78&Ldb(p~Y;=3lA3L z^6DUG8tKe!HY4#gH_=ls{YPJgvhW5?LWtx-j$NJdmpKn?MmCRil*O_COhJkt1ED`s z!ijaV#r~4OMy;NK>+9E16g$qBwSP>4DIG-H84myI40-)(3fJowy663q^C(w5RMDsU z_2XxnwnQ$>%fiT&N!N{?^iZM8)+%pnmr=afZ;#F|k>MY;ycK?0&l?xM^jv*dbS~(i zusX1U21IAK7kv9`OY{aQ!d9H{?NzpvMBgWlO57YVUmsNlnaiuBLvKgmQ5#VjhEPmB6k*O)vnNqF<=beZFe zN`-jCDN0{H5$y^KY3WU*%Hyj~FO1&l@p3vjlcKH9R-zvKsA!I{Hk-Vyr>U%~cFc-9 zquAO!fQd%kojEf-+cf^sCqiH|pU*MH z#8>Zo99aRY<NME1^oLPmwTx#(vjIh4m{-n}jz z7J2;1&zwCQ<9i=6Hz&B|OUmiB8dOix@?IfRfTSM4Y5R`$+RmwC>a%-c!uh%wNUYxe zjbZ*MGZJ@O)>G)Hf9o&52I1a_$i3=)OIVe@-zw_GJYICt-!PMo8*jFr&2!B3o^LZ- zt4>Epf;b1}%k@OVEpSf%U2nAjqP`q@&J1~E3!?y`t5HkUGH!#yj4~=R; z7DQq>uPtrT*5*AMg4Stui8*IkX*%4mIy|JJ1ePW2En1f(?4sdBV+zTd3s2xphUgM1 z%U)LDKv#87%+hpruqniX)v-Es^q3KhZ)sFbFFP6^+q8ccbHKdNRb?mEGeMb1seN`N1~qGtIZ304(*3%-%c16@hEo) zLu>Dy@Ul>J=_)KG?BzCXX~LvF%BO*yaNK`oZEz!(_tprq?svW742EntmU_2z8G1qS z2(24$GYaFgRkvmirKl{!Xei!Ca(BEhe zduC`4HZ(H2XRA+}56(6A#h2@T0j0yd{@Pkal$r3R4ssslfXl8g-N@vKB)39p81Z>L zimnFMcfD^_Q3Bdx#0q2FS*r4 zajbZo7_I+U0P`F3GclBzz0GNf>k!qBD|{?y$8>(k)u`nc%kYBoBGPm3Mw(jk$%lUewc{&_N}3p1?C@R_ zes1aDal`ARq8F_)*lDgKhYz=ZzN$?BXC_?9B{LQC=>fss@syNr_lw(O4su*CT3OTS zc3ZX3MGcuj5|$13oeceWVLFniyJ$qJ?=lXp-n6)uHE4(Cz6jNQkN!#gmKH{A{FR<| za<~nW8eb(77ISUfLld|>LEM3VOXkrM{52J9jvP`4+u@%}utz}*RYkv4VZjhY)#|gk zWlUm`Zz!B=(2CbdSb{%I3?h?(G#22H;7q}zLi~WK<4mgR?rk%Z@%0n-xqLATk`{fk z9eQ%H-ll3|pQ6E~{WrZB&rEZP0+&a9NohZjBZd%NjW(D0*bfnmKdTo@zg)0S4E4Vh z{4j@10yv$v?^U+m28!UpzU96%@NT7sm`VOcyo0FBxuefERuhrP=BTjsBa<88Mnc4+ zy9N#-&&uLpb2n4Z4HMqtEG)*oYSAh?57pnb;v{>cZ9#>np<(my%yPUlaOL9Cf4fNfefAtu_0uo7%b3ff78+hTUO^3>?{jaVK;(T)w?Yp-yhcQ3x$i+5C9%`SKkxTg6z zLd6w1ONBQ3%Mf7wbw#6TIt>JOCTa$$GE{#)@cjKsm#=lz1^VVJDatyyOCBPl`t_Dr zYXy(?I`B_GoHesjKVi&o|Is8+;UJgMY~4Icn1(e9OGwNkqN2PL?S8!RI(y0_nE(D~ zL)@$(ano;8=h-^raLU%!us^oAB75}3UBj2SDSN#qVdI9)!1xM`KI4@E$tu0 zwqriEh4oIMD^~@uFKiV>no4G`lq4#nOxF!N!jrYn8D8^}p}|-90OqoaMBI`5$$hn; z(x1-`X2f><7dXa23&2f9zpJpy_1j8=g2;fMKnFH6DnfU*15Gcy>I-!JxjI6x8SYbQDGdhz75;cUzvfF zuUXNdVT-aVZL)N-E2&AKKARPbCjkvvLw2tZt?yzZRE1i-amvDBg{b&0uB@QXF1%E< zz}s3(IPEitzNakGl?Y&=e)j8*;;~uz1xBH1yXWp`)azsvjh!mAbCrOWP!8-|@r59O zw-ip?W&KLSe7dUzJ1nMmD9xwFD3*oQThQ#mh ze7?S&Eg~K2njf65phhSLc_;hy3FAvpzW(6g9`VI^v(roHTU8_}pdXj!*9_%YRtU;i?)2>s z{+HWTAw~{QE5v2~1i967`_7=udcYM96e96&T*bI2jCdKd1xz{nt%az)Xzzgx*&TcI ze2`FTtVedtNlttsexH@!Wc{N21{um2{H~R*`M(-xc_W}n=F@Y$%UDB!fC zz?mRB2*e^uE4xGE;JN#mwm*x*ltAFMm|DT!_1k+47h%~I63pnyYy(NFUc&EOx>$NG z)kRA9=;9Qt0ArFcYQ($rU6RN@;vq&~xaCm_Tz*6wBz)(I?)y$Hn>9yHmDHIoZrhHc~ri!}5s)UxMX+UB!*yql)2(}CfD>IAcJbaqJRoQQD`RedyD0LrX zQ_4eY@?vR2ZF3!c+g^}NRfT>&@M|mw==8?ae8koTd*N(0NQ$9lVp$l>y`hNoz+4j$ z;()!N!rB>#CJr=&DBt%__Pmo_QC4*TPBlXHz^|e~2nDqeCBV}8AoiZsi&kG2v`kWp z>=Q*ls#ySi;HqvJTkDmyaxa zDIfgs30CpsD|FnIL=N_sJ!DgxdEztwboAjTZzFS?81bEH<(%8g=>tR>Z zbA@yx3jq4pJ@o~t-BG=-fahayb`Z=jed%D`*)tH#0|celM#tiMbFl0ldixdYHvP)= z2^0zSbI95|Y^c$UHkxjllhvLQEI+`50|9g+Z{B|6@b!58M;~#AqI+Up>5y?QZfZLGht3C!*@j zy|kCQby&{*6B*0Po(E+PKj}ze z{=Q1m)828z6ibtsQiDBn+hB`>5&&}3E*~;5Ws3LCDUw(fanD)Z46@=H;&%Q`zikp= ztHY`UXWk>`8t?HGCX@zB@C zya$;7h&N(V>EVVOQy@ibkJu3Rn_*?8tc&Z5sPMsz_uG?Cvn}b}#VU zTNh$Xt9UUd)85JmAJ}@Lpev>&`T?i-wv*Q@@!ej{ze+>2TKm~ywp0ixHVW69FfA7P zG zcQ%GNfA4=kPkCKMpKFa^TUlSQ^$QLc-T4NWFl;-lp5bESBSIos&!ao+2hS(4v)_Ha z<=CbW9ywPpIZ4Y~r=^5F6jWbtE21-9@%z*DhQEr$>-54Lf3;+AuezqX;=CMN#_@jR ziQL^M0X)(YZCeLp=NvvZKDA=NGa)MGCmzdTw{Owr7irnOfE4>54m_x~l3tNWr$Vy` zsF#TfxKVz7H);uUHFS)OHnK%-X^f@gJ*PL^Jz2m$sS1@v=D5Y*PQQ)d^fy z{|3}hbvPQ2evO9+&Q57;fe3Z&?lI(IL@i@`$$HOuwg%2->|`4O!Q}Vsv5{4RH)+jK zotke-u?n|EPcZ(+>)z(rTQVhj(-Q34_mXm`_$5m>n$2E z#mb*;3=f>koa?=BJsrb%ZS$w5v%kN|Y<*toyJ&B@L%;4}qM{T_$dm$FP;1#xLgUM< z=s4JS-07}dURv_4Ne1J$CpQc?Nk!cm9D#f5>6}k8&y{A)8q8oXEn|Py@}-VD$5nom zgU}jPHx-c=$JRrV?&C>_V5CD=U<$u4=Pi!Zirx`!j{3R~^aJAWbNA$|Exm6)*^xM_ zkFoWr(9ht*_Kzd#l91Vj5clC_c2j3iADK?YH-WMGknNnL&LR;~=Y3OUrogLkKHER| zN0V%92OLjb71IYmukV7Zf?saG+z}J6HDfpo1`d708oT-1u*7qhR3|wJPY=N@8_56h z-yi?~!7w-(pA;j9yn9rTS}O9YLudTMl*!V=Pl8xkGmw6Y%O=8o`jUOO_GeY++mw$6 znM>L`&Wg65=p&KeM^GJmjdLM{EXyTwC(&fo}hxIB|Htvgk8_7@y1WfDl(HM;iTSQn`r zTDp)RxmoS!;CofKFiB3ogIfLe@f=))7$u*&g|G*N%b=*; zu?Glr47voM>_Hg!5VFhC3?{P}F*_FC@+h@w%E^}N%~)u{)qPTk^7WDGYXfPMn^RZ* zf9LMJd?;S)6rArvlgGwWLOTTKo#Y7~%>~#{_`bd0PbtBMi6HHV)^A>bS~`Oizl&lu zj!y6Vj0|IUzC?% z|9C4F38WO!e??O^dKSZ?PYQ13c6agPqGxYKLP4+p;Q%%_l3b8hmj1%nC1!;q!-OB{ zbY~^Bo&ov|@JC2eR*RN&#Ev?}&UXswF-w-6AGjcCicoo9N|LB|crzb0PVCO#J{w~! z^jSv9(J+CEVE}D)hzqsIRkthuRhT2H`fH>uDq!$xJO#~#Ro;h9OlE`Dl!eW?rDhv|Aj#4%}-K|(W*?Yl4WnBF@NSnIfuUSmm@^xudsIUivWlfh( z8b6O10cGY4OE$|11A$|2)4qGGI$yNX9*{~t&3%H$rPDt$2`)G^*WMcc@Djo zYLJhQ!TSe*UFUjf8Yoc*G}Sr!<@&XXqN3QW>KljbJWw_P&agyn!Di96`}ruBm&$*p zTXAx+D5Ci`w5GUc$JL@(lXB3=(_a|@X-{p!y!!B9H`R+=#wy%*`m=h!hnmXG!Q+ow z^{&b#xi|j@%lc{T5@CHibtP6{>$7nL#+thgRb<)Q!;BRt$BM&|JKc1FYzeFL>&$Ua z`(=DUp!lpSf1?GBhILD;OEY=YISt$&m4LUl!%?J~}<%h~&G;f!!FJony2=A?`(WFDV z=k2ONTYlt-!Yj(l>X`Z!LKdRxI^;?Y(eGVt9o`m%jq7Sl?F0xxOCh<6VNm7-P<#RF zbXQ`(GP0Wv^#8_iJ2XukLa8vwT>W|I$}tL~HP*Smj{AX-+!*(`2tk&ExnmuofG+ax z=qsXdd|T?2hg%|o64a}210x!IIns6N%^N;DakxLT&44ID=fRJp-j94MM0|&7;-|)* zn?7jww=(#sN63Xa^}|>Kj?g@$YjB}X%F=#a=tuZUE)z*^$~=Rq{;^6W0|?2{-{~@iZL`rv%%7@n9g*U5ddPiK9rpwm z>(b|j@698Ts{K99(*qS!axV=?!D{^MpL5_KaU4zVP0?ebk4m6gz3~l!E!}rj29Qv4 zc_0m8rNs*m^TCnLtAfA>{nldK7H8L2VuAoA>A}^ujgS{5t~b>zIX;|F?^0&+ZR3PJ z@AfwufrSon4orjdoM=2z(GBzUlfTfag-p~wyBbDSl|oKN_vir9Y)`68!MD{Eak|+N zJPm)^uR~ZwT-&0`Y8nG5S-7OVmB1ss6Q(~p1H+4r)`*HG6rw=zy zmR)Gh6R%kQB$8&BES7%$)nOE{Fmda$`780LLLfU9B#AKn7n$Jv$;Q`4z@kHWW~!kr zl}fBC86`VH)9r+{Y{0LeJgp6UMOlCo=Ie-sQ~!*B+ALa)F8N?+YghyF;s(4YZ+xl! zsU4cOu*W>J?u9mZEW4kqlV4Lj`i1Xq7nbM`sK3|g>O$|^LyW3BrI^XQKdx1}ZsXC8 z`pL{9!;pT#eI|@?@h)pGN%GA}j?YQthFQZlo)F`>Nq0i#P)z1m3Iyk5oJj0`S>M+2 zQ+6imO5DW4Y)aV~vBk|)SXwIJyb_U_*}yZ+`=-TR&P~CjH|_a|F$yzW^jFBbZ=BH? zI9xIH(823;wk-qZhpdN9U%|F%FGey4XfQM?DNP7QRS^0sBrVT5$NO@&f6Xx?x<|_W zm-^&>Jck8XZa$d5KPv!~z@)Q4u6YZQ03IFYwEfXAAe)a$b;45~E&$my0{h1}nUg$P zYm#s}o9FtpET2K{!{d#aIV!QbpA_fa$D~_VB0OqsWL05VlAJKL!48Fts$IY3Ywi14 z6uvGQQBw_s|2WX_#(Zp01R_d4)4>(Y+NVyB+$do!r$a>+SPpPRdqn7WnVxd_jA$cD zxpfhxytoX$S>-P?b%g``J~FlyA&Z9-Er5aWNKQFp@|cC$@}I%0+Dkf?{WywpekHGO z`^AFI+*|1<*JnIAu15hZ!$NO@M(Z>=RNCBVGcD|6>J1yK8AhJ*8~lua23r|+Ns-6| z46eHZ`b2?;o$QBgz*sW0jdr(8G)eQKM1U=@2b{?*f}Szn(V}#5e^(L<$+jFctq^=! z$8NFT`G9-kFf0f)5I&6l0u(|fL`C(K59}^Jv(svd9GpB}z12<+=I40@OZ64Gz9ynU z$$NPGzQ36WyB>CaD9^lK(n`0k9j*owz!+b(DE9iBf4nHcpR zXvg#to8&<|hTUw7zB_a_9rD2B+zyt@z6vs+pws-Gv7l`v!nXYW@HGKt8d96b`Y+EM zp0+kme%K$tUpmBC+RQv^9v9J?(fCKa5(7$%yvZkY#UobZ2?L%oYCCzso521nyR(GqKWlmO5W_ zh66%U-`XlG_B0{#9ro333x`l4r{zhDIYSU2xDdb$9>kSsNImzbEcca z<)HM`&WcRLoUS3iNFmxjT?#!$zh&LpUfsso*5>@lZ#r)Xtekp6_I@x9dh?6m_z&3Y zZ}8ej^Zna%%QF_55+uK$rocB;f}oY#zopkfe0FnTXAW>jC;muxI#O|a*LMLu>ztIY znSYAjlMFm&R0=srZ}KjK(th6w_G(5rb@N+Wod9Gv3g07S->!RiojQNluT`sR&h^R|$x6Rl-Qy^lN1LCzej^tjq6uB$1RAJKvl!!IqWL_54+}mSWo^COp2irkCDsWP6=PW zUMK9%D!juz%M}6RzE1h!Zn+0lY3)Hr`pp=B>mZE%k&L%pI~0v6w>M{aSf7mSs}=Hm zM+!FaVZ!XdZ4c{F-}D&_ZvOq1GQdY*JB##$n5hROGrO}XFB%YKJ)^E!@HPX*!qyq_ zM|<7JX0I5K*$x$RCN97B>Gr~zmtAaNBxnW7RnbT{Sko0hmtjlq* zhML}4T!iX=K#9P?4@poJAk{gMg;7LwjRiB&BiW&$(JI&bO~}X}J(tbRla^*DB}Ky( zEanxkxP9;g5LKSmM6wHIA}$FR6pMyZ617%6a)p|9LKWJ3!w$@jIRhc)sKno4mIr)D zKqS|%9;qY_Zx|DX1P@`)F0pC3b2_UJZ!+N@_LZScx~97oWIYdD}UTtud!=Wzg354 zU>squ$Lv>AeOx-4RqW6LXT**c1(L)j7POAD`$E>QTEpDrnzpT-s_%4ts$-q-)(LKk zRUBUyhWIE_C@5Qh!dehw)ABYfNf!ue%=B!O*OavkZyl1GG zCPZBQZ=zKjA}Y9<+UdUznyVfhfpv_Tk)dXd6df2ZUpgddLggdhP1!}yK_GZj75V3M zAK=*mehr=P*Uv!K@Nfjh-lV%f%*9i7tq#^|(<+>tVbM%XgI3dhc|Gx(@W1QB(nEm{ z_&NWgz+Wm4QLOcbeBTxxAY?*p91|LCK9`kZtj6EcNtTbo8R22CE$o+Y628Zrj6TlEnF$s_%yj?Xua=%xnQem&JqYPnjhlBC3i_R%=HrtIwOtclq}j7D3&n8;JR zpvgQcT|Aphiy>Vm{W$G$i^4V_1x;OF`dW>2#JT**EsjKwM4AcZ3{JK==EbF=At}m0 z3@`+O+^e`V~q&_qiP_%^SuENHIRCPx%W?y+90 zav%O=F0>x=y=AV_wxMTTGK;JS!-2mYb_8T-eP?X$6*$w&1$Ca;qvyR>;BGB)H^Kfei(6^96Pb18j5nY2xD#BP0rr5JvpMz z)@?dd+&Lo6O5+?t!hlXTeg=kw-%X9o75qcHy!|iU4zTr)=otKKPPXsX=F5@GjiYWQ zeY)OQYS#!icuIP1Vm8ahp13&tz}5D9_F-%+q7cQniah_!1TMu4|HL#D;7Uy#&I9I# zGqW~sWp}Dit)gG(>8jFqGd~Y+T=BbwdG?n`4^)=vC zzSF<|Rc(7D${GjXMddq>4_8kEOAbVt--5{)j%^&3X!#K!G_cxo2JFlohtSKNfwWa* z7X$W0bqQCSqb}uWtn_)X7Q2hSI-9vF;rfKvG?s;xCB$Hw4>Jq!3(|()ZPu!{xEv33 zv4c6sK4OMXn{gd9iP%w~-2&Al0`;!PI}+W@ohsDNNI4XaR}n#=dudJXC3;l6)J;!~ zbWKW?)5qmK)wQ5wb`Y;7eL0y5h-BbI4EaKtz`n~0kH7J=|TG!5qq==ZYt zd(?G%K~1{52_Fo*(e(tkWV=$;Mju|D8|_X)x+=r#VY#VD)GqGL%pIV|=DUf-zI_{X zc8ecLb;a7yE3)-?+k5!eds*G@m*q1G3ZBdy;Zj<%*c{6`FtFEMN^ltB1*@np&UWV| z%QmF0I@bd*_sA`EDOsI0?pS8(!yBoI8!{<6)Pe;AVx{4G6_x?kb{z={8xOiuaIZ{Qv z9_D_R&T2|GQe2}rn$!yF17Fq(KLia?6rw&+oWO;nUTjSfObA$dL997M=!+>}C$Vgqzu^T!3HeKSv{-Q0~MFg3n5#I#wyADfQD@b5w` zMfCp!7x-qbGZ@ZELw;ynQl2&XIkrOd0JQoLA313k8Q9w)$>Mt1^%I)B@7hNgEA5u; zf5>)wvd;qv73s%P+?rqPXheWcr!(9-!3#y;l_P5GLocvFx{~xaY{4v*?*PiGH15By zzY90-Fv(IiR~f;+|^r*s@c%hbbyv0pa6W0phosZYrVl7^p5UE_}(&9v#udnB{f^J1f9y)?6tG(uG!A9`D`CloDK__tB!PMrhzZ_WJ>c;m5k@ARFGVH-WErqpMWGxbZ0ft4g2c zpLd!0fttz>f{piT1LrBL2q8V(^~|t;_kSYLDZ7>TF1oj<-;r?r5|ip}q(_-A%b_-= zZz}|z1@c~6*$oHsGjnUseQ!_x$ z9)^RbSQ_e^hM4N^W%u}QD8};Scb^+E7ePgl6%OxI+xEJ1hiD?1voCz9{rdq7N=d|$ z7SjX$L=jM@2{Fz_sRebf%%*RD8DoBV9bv}jL|M|uxr;&0{tz?Jfi$r5g4+62Nv(V> z#I!j zQhX8AP*+1!W6_DgL<#F?uP7+Q9$?*r8$jN!7*a~LZsy#gQir7G%rk5S82>D;4#!$& z<16^*DI={&TnZFLr6QZL_;tUDSa@pd+jxjiMJdzB$I23F>O+#a{qqYu2S(L!x>^J6 z&SX44I#X`k0Kr78gbU0-dS zLWt`ZSC=UVyLT)s*9DB^v|c?!7(1emawfphzNb8YRQL4AcN8<{%+5RbS?fCR`ps7Q z+ek+ft=qK7oe2$SF7t5hCw&c?;aBE_;^kT4kJdtZHgJ=$AZ{Or%KpQ4@$<5o{p$?O zLA}LLG!$tL?ka?^eOFK9hI$fA+MjhFtoKw0WaYL~;GVVKY-$gjFmEH|lhM1#^8@x} zcglYkg}I+SbQd!q!q_PKn)`G1ZT=4_n7b(*)3nr{Z3sei2Nih2o?`K)r1JwJ-@WXurC|o5lo)@U8^Dlr=HrgZ#ccr+b1%s~hmNp8kava!%rP$*mK7MkY-2^DI!> zfBp68ZLM_T8MHBeytcx$W|jl^x;SVZAu1@?*SZ&YWIrc8-J6ROpDp2MCW**eJ3C%1 zMau=|N3xJol9jLQ&cHFOYQFMC#n6|6&b8Cz#%OnNPbmxzmI{c2T4v#l%f5Nf(lG%}O%|_R^mf1R%vrm=>6=X|g+aiHeZ{I)`tqZ?h!DZg`;=GEbN&i9EdVh~!bAXpyPZM&k zvQ@5QMl&hls~`|}``K{63YOY$_aoc2kMv}da!o4XvLtdR&aqD=&}aiVhxgdZ;O~Fw{G9daVP6J{|KW*@FzXht+bWOihy8e<7@lvPMqf@uiw2C zk_T-)hS#hd_G@kW>iqf5cVt3dNLxd*_1Aes6m(W0&9HelYhcZYPhd^nk|Uy!y^UXTiFe07O>SONxFNDtD3L|F{K5jh3P;bMt5 zl=<=UxrUk2M+Lz9Iyk@pJs`xizNeb_P@b8uKc8~xxy)04rv17H{EPmok&&UthXr2v zO5V{+66vi#;U|WJ2_mwsJ(lZ7e&-#BaE$KI#+VNaJ4rRp0{yqv@5<@-W@U}}RBdwJ5C-e` zd)QOAB;jGX#9_gV;uKhCD|8e7(Dj?YSFBE?z}3HClHa~Xa2Z^4t~Nq-%09c*22Zc7 zVNcDa0i=>eFdmkd?sFWwlkSH_m>^<^>CO5-z_Km~Zw-_kd_j^Ps_N=Bo8udf{-TB$ zx{Mm=VTfpuqBZ2HHJ5N+x=PJqal%bmU5?pmUKV3B=0bajVJZAN42)hkn?n5{no~7)2TZK zO$!Ti_-uIqI$K{5;|E0c>33)Qgacf>J1}#8RTEWx^sZ#tx(k(5e0bxe+3KyoNZ?X^ zemAs~EWBA6J0Nh);S~t5fs^STS=Nm*BbNe@!lAd4dKi8=MM-6`UMH1Ky3!9^ru#c{ z4vOkP49T~v*DDL5)|x9(45?|lkVsUfs4u_gKH13fKY96at;&hCJAhkp{xf=JYV8qS zbw39}*XyxDHtxMVuMZW>g1X5ONvrz)p$xHCtNOS;F*fUWTk!7EZv?w5JQ$xSh?VHP zuP)5P-VQuP_0!MvPbRK&BQa7FlUZufh(iOkwRXVjM3xlLiTm!3^&+unu@zza?&0Gy zQ>gBs|72t@vWzgWb=;y53Cby~SJisCk>rL;y}0(EOd%6Il`BcJAM`9W*1!gj>q|s* z-g>n1;}4|>!`VNn%X3IT(9CM@S$Yc0PEeYwc5P*^%&uD$0~d@y3qN+bL`X_u_B8 zN>&Cv8?xXmBmCk@d9u7u2wjx!zRdsmSS`3;c8S}kkGR1!9XxwSmF5lVWX{6=<~PNZ zxOqw;AME=9&OS`xf7O=%tJutOC`?zd;*rcn=@I5;WB|H0p!0)5O|8Gh($7LW72y2k-kBWPCAr484`Kb84ShYXp14PFWH_isJ9|7Gfiu-f1>c*q&v zpzZwHh_3bMk1F!_q88%j$-j#_uYX~*VXiU@36p;>vcf*42;cb>h>`N~7j%jrqeM9V zwAT6#A;iFtm6KwpXSu?vF&2TP!Vb-9oF~Negp5Tvz>OZ>L;zLp(?Jr-`_InBO%~mn zSYRTGF!w%cAkgUhwdD!!ZEuRklpwk3L#w#jUM5TL%ya2_D+H9hSv4_z_*yHd?gZU>3!=NYxcQH&qxj_wzU z=^;u$`6U@@>MJWQtwJ=!fM|dhc%2dPDdAI;C}ds(ee5{lOm zjYfCrY<^wkswkm`do{eH<`vu?V*XB7h#Bf)??v}^`A>v#r?+yBoUE-_?d!%Zr=?VA z+O&+nqo#`ux6kEVO#|_*0H|4bD|Li}hfuN!^buB68#&a%Mq{6Ss6A((A+(uaZp&S8 z*tM5|R!I0_EvMH@KAE<6W4h*cENf6PAUD_7T4bpLqbW%F?G{!oZhTpiT9qa|bG-!= zw=JkOk^ZS~)qqC!n%L-5^wj!TM7a@F8sv8PdCPfSivxpO@D^ViOXcV1dull=Qq0Dt zE$iCx(N4MV z{jX<9fLPtq=DzJlE(1=5nrn)$A#rQOF!59$j6X2Q4#(5gWIZ(~fS?B62zl}dA%pIq zbg6~^g!Gb!FZ!=ihQMT9z7@ES}mr6Et%{esy5!_yoV-_Y}G0kABPD_+v}O z8J=hcW}FHsE1iXhkLeO=G|yKLATk(QJuM5j(&qjAr?7eU9;OtU_xpjYeq>_RiHSO(91vxnmik!-okFFA6LdK?6X`_+&PXR}|rdvkE#$t!pi2UIu?j zXpsT5MlhdyMJh_y+`g<0lw^bMPx008V zdTX@m;jBBN5atx6Oi6x>pRtXa4k4Dg5yvwlLcXWJ7qbD(ULp_^!Os9I*n!wqu3YhO z0Bi90B7k~f7CyBI(;a>yS23V}R&B5~jwS{#(j^}VNrK{Vcz1DCf5%cVzi8vQp3xJ` zRmX>*G4i&xvR+2=0ccaQhw#dS0X+%bkQ`7}gGhw0FjCY)O67$x3dCm8 z?@k}QV+vPDo7f|QTsQ))5RzkCKsV^Tcmm@TLr5}!f7$`t=FS$nCruWQuB8pj&u0$V zl70rPJrr4=$PdJ0KRV!fVRwBQx5aNP)wM90Wo$x3mk{X3jQ9ia`Q(tD56Hn!Jf0BF z!buQ`!1!=WMNG3X>kU}U&nUjp!hYfmZLNp%^M6?2YHJm_)vkU>vDe1;HY2|1?t#h zH@klFInMnXT^kvt91R>~)Zz_#3yM($6>dQ4S87!imnPq-Z$9Zp)p4aWVxPDeP)g>e z!0l`lxRram6!xZTZ3f^CiIur;Z>z;=9O_<-hIaq<@>~-Ecmq|0jfb8VQMdfzWdbkD zy#Lkq8YOoSoQn*o_4)K7-AaX|KJ}sW7*owYuz_F(oV`3W-2@A7C16Xd8{ zcszjc(_eScHa~D&zEXNdS5WqO@Ocls&&2V*rTbJ*=i1+Pa;5~{y)=P)eCSC;@P@Q0 zyxn1!^=YGg*Iz>exGnOr`aEnjCq7=bUjY>%rr{0^U!pJS(Du7pQMW_9yL$c{=M(*+ zQ~pIhl#CJi76<$ZIfo_PU^$5MiPmZC>hIryoxi^I0cw)}4!6Eaz52c?*}k4~jdW8K zauS`mCnSlp92;(NMF~384V8XuWuN%xrf_=ABw9^KnTFzcNP0hT^0}`v$L3(qL|7ktdtiS_*0!^3>oE|%u)6pa!2Qst{4k5r#tC~aH$(Ezlk)Am zwMt4<`dio86>u-<>|&PwP&QRq@~c~9KxUUH*3NV6<&Hyb&h0HSLg>v=XU)zRT7O+4Q|MLkuz1psD@D)zm5drO{<{-Hgf9>q1@ooWy9JW)$(3K zv47WozR4{yn~_+6oK(_nIuml1*r{cpqws&7@%?ulc5k~7k{-G<4W6{nVQ(zNZZCQx zjQ&sY#Ui|jhG<3fyrBw_6WGo@8vP6-c&a*&R%4Xwa&~Szx7!9BUqRKab(GejMfa|= zbjPI17BglZ!WFF;`_iBk=3@hIGr?u+ET~?M=`Ty%s`N<&wAbdJnr7dn*wqWp*H-~U z3`n;TLn`u9+M6O&tmYnLVnEGk=2$mM3(tY*;`L* zDCV`tdu0BusjEUq=yeVn_T^*nL9?J|a>D4S?^S$urE5>Y_#f+J9I-%}X~)^M-MeFk<;XQoKGm;tq#`A3#jm8BZ6g55Ke1-N16x6vPQ-M8acRIm1%F`(ae41 zi_3>BLqZA#0SH`?F1Viq@O%dn9m=S0lpU6k*w5{LE`o~06keP2uufR zkrPEQ1y-(a=_xU5S@Z6%QI+cCZ+f2qLagOT8sqzw%;26OU*4w$fa!~?xsXdL{8;0 z7vkP%6|b4B)Z&qh_z(HS5f6VJ5OSmT1(EoQvc@-(Vg!4!X_P39#+Gu&t38gEbZjt! z)a>%12^LYvwGq=0CtzfLI1RT-s|?O`#nZ_@F>N`v2%fX?H$H0YhLLll<098Z-gzkh z>Z+NNU;EF@yIlYSdw$@u|UhL#f@C}##%8-vP+GX&6WY4C(UR_|^&N@6(^NJkFg{wG=aR8wOj~wcfXRh0tbsQ<7Z7J`=|IKg%bH;Q zYZ)n_A5`IS!adLNZ_HfCx`e<}skrg0S#|+Q%8t2lhLD+~`I9k1OcBVInxnR(dN#MC z5UVV7XLWpFyZsAl3nDZ*Y^&2rO@%C)>&^9acHloC`^%=FQ zFzgcL%G_$lH{t5v%h@zfKJ4^FeV=@e^)^gjKoV?l$7zdmUHggMsud0iRIT7O2|G)8 zS=WGaE}L8)W-$82z8329Hfb*VkgVfbtd|}IrwU)d=;dxJjyGNmF7OQgao77)OZ~f7C0U#dngXad$XA6##>fUP$ z{RWs8ZzZUOZcMjXWjGlc)&l8mRhDc3eO!`{A3FYC@0SVe2R1d16GJ0!!;{zqtG>M1 z&_X37P{7*gK7S%Cd957PNxhft16pqp(MJh_3-pUb5wfn*Sfve6ZE)gssv-y+(}_xb zUbNq~U;CxkEd?n@$GsR<*B-G%{gT>dMj*6}WSj}1W9gZgwg@3 z+_*M{=8sqADN8{&nb!YS2?G6B401C&^vyK^`1$G9{S5!f0;nkrQ&z*dxqOf#1%I`m zqt<=9mr+2SP`|RWsGu#9ECutm2@aU2O-PI*KwQf{dJh5#SK zdk%6c24VbhMWk>sOS3~fyQep4wN-tDf9?F)TIcr&lzim0r)nHiziEk%1EW~f9bkI#(cGoV)*Y8p=+(%u53S`5!KOL2T3NThzeC{jna)R$le}hRph*eU)>ayUp z-y|d{@1~oBm=A@@e9qFedjE=WLnM;69qlBfO&yCGXGCi06p07;64JeJ|&pH{7}RC4O;gXjmxWZ zBEyq&wdPt2859Q_$_o5bM4cB$U4Ek+9Ek)bvlRa&@hxlp;RIft;rm^2|2>zsDQL6b z=Q2b%@Rw(&h5`lj@54!>+J~W1`rCit#tY|jiKh$?#G(FU4R<^leirs#pI?@K7|-+Y z_A4D-eA0 zG2ZkINW+4iCt`;{RFY@BAT6KNxnNlw@W19=|M~O$*ZOrk#~7}%M@ErM^Vyl0b;n>@ zWyhRau(*Gv2#-ye>hwLSACePvFs#%Mm740*&c4q9iCYT~j1T%O8X4DDwkW+7EwD<| zxFFC*!Xmt|#PzQ`e}+3{L}}6e791W_Z2sQ)A)M<;YmTXn40w!$#_<`5M2)9NRDQ) z0A3|=Fws%`mzHhjnQ@K)ROtnJx}dMzG+=QBb4D0vxaCDxZ&wJ`($=S)rslRNW zeu?}%!(zIKMl{{F127_}As~?9OsN^xc&Pe`zL(*+o>A6Q^Ogqm_~Kq6*fK%&P*M^N zZDY+9F{#KvJl3P;F6@k{+0ez-otXSoC~wvV7WUmX`AC8@RR?E(D1u^Q;B{zEU2zg6 zYZs2QYY23PfJ5J9{nNdcxR6*4D}F`QZ>BRRK+ zu**YK{~c|KH>a<~>F?*GgMj#`oMd+o92rJNh%#^-9ZVDoj>)VdEWkA>EsI5BsYbe) zTy|&APIS&oWTkb=B+S6({#%HyFNWeBX1e;3(gh40A%yYq2V(DLI0=arjiM9Dv4V-n z)9}NNRzpoQ;#D@=Up0i_dmJpbq^~}O1rnVONWRaTL3(SrfWwCWg66X?P_%%}9d%<)5;y3_+<(_EDOZy_Utj+RN3HJij04qzhJceZ`o#Jjw+ z9-z-!A}8d{(3|_Sr}Q zX>tb&*8Rygqxh{furIcPSE6@ht0-~J&Q=Rh;VHW_c%O!@y*zlIO3|!ywS(@azY5y@ zVuJXIYq$HD_5b`8HV@z}BQv+@x4xOTP+gH+sc3NGHy|uW#+hiOcQfVARgTTFSa}<2 z0ItUxZ+?UF<+b0E_IaV{;(=rFLF43aN`n7B56&|HAQ@(l$4Gp!jH)rN%T)wQJa5w< z04L;vY&(yWl1=~abE-|mR)$-;`>*becXp`E(OaL+;YR@EmcA(N@3^OpN8E4KrU!hE zG=}%rh(nPP9jH@(5Qg!61~wy~93cu?nk5|`Pu|4#Bi)Vx+4t5sr{$!kTwPJ#)AAJ( zdZSNItglYWA;W`$K+wFiy$ z@ox#St?=A211t<9+ihCB_>nGn+Y}0Av`1SZFwxNKpLUAU4$krugGg9<7McCY5S8Eg;1)8&e3JKENH1CQMjx9YP(!*g-DcSV0e?6Q102~fM z_{}#4C}4-*_i5Z2H%Yx;5ej2Z01;Aci80J!_3A%hsUSKlSz{Rs2dmZP0n@T*(B)cU z3)!BC;3}#U8pez=1FZDqJ};w2D+zdP3NsOL<))!ZE97{3%8UpgIU68*$#3I^3NyxO zRf8C79c3IP7nOlkZ!1F%Ku%+t840{kk(-l4=;(KxH@2jhJ}J9ovlO#*tQ#EDj!p)c zK0rN%AWjP&`Vs%U*(~)>n0r6gFCHIQP=`A~ycScq^D!i^o)E`+QYH^@M2m(MsUr5} znaOoo#=Ymud20gYc@1j_LynK$p*41m_Pi=DhV%ypHYZ)lUZKp!CtH+|i?n)0baZy3 zDutX5U!z2EKXes+j!I%d-LTkH^!G>GfH7VGVW>QQ`V?|;_#dTu_z1&c^_w#H zMw!97Zx&?m#)h7k1b`;a!z1bgDku-4zjCm*=6Aau5ZP`Vsu*{64?PvhbMf_nL|)7E z*74QwnUpQ+%igu;(C%pPlR&n}zed(JRP3?8$3-v7(<>&US2@_&LcEyYY4v`@(NA-Y z_vnRpJG*?k+sor@(!u?3m#L?)dhgJ-h%I84!XA>mmENsK_$&V7lAUSG}EWt zus{mwNkP~pJsMOQS9-Ox_xL<|0jSX`8SZ0LCWYX+@=slP_n$d{``WJTdg1j0@Ew`w z+e_P8d_VK&3`m%W%$)B%V~KA9a13Av6b1`EQ8r%G7JS;;aa?uqS?-{C!ki-SK&w$X z#TvO&4#0)h?QKHb`O!!^??nvz*2(Vi`|w}PR1!K=@3D#TZT#Gm*f)E*R3cxDdRJ3z zIpEFv--Oj`OV&G{=E(_M;x?rSki)+OfX&Is{m&;j%PkHN|A7nBoKMS0KCsXgkb4c! zqgS#RVyDItOg_x28?RrCT@9=s{d0LL?$M%Gh*O&6Cffs!dH*!ATAWRdQ*2+OA%NYee#v+v*KbkCWfPBIlC zlb3z>WGKqlt9uBk2oF_3>$S`C!15xrf<{=~9Vw%F=}CIf3Mcxwn!{y1)Beu_<^Nkq z{f~~#^?PWj#15}G2`P^iH~cDh1Rgq0O!$ZXOLv&NyX0yt{{`j}Ae0t)cozk|(L3Y^ zYeGvM+Du2g7$TS`JGeQyvJM4l*XO4awGy9ns@`NeK6w-Z@yA|6`tT2k@2q}ZF-I}8 z>ofwxESskULoF6%=3#%D_p!aKFqRczsow*g+9~CmyPT7l`)v)aP02_@=St$r>Z3k< zIgZ&}W3diU>btCq1S(j$Q##5~wtA+;i)S5wlWSJajiPI!AxVowl{?qVtmyc0;<^T+ zWu|&>J>`l!WEmasD&3|mCb<;C<39(k{Zg{EcSNQY{%&bf&w4gLG<=nr@Wnx@CuI}_BZZQ()k)Z7b$M`WnR zk`gU2FVq3j`?H$bdAf;Lu$1Wz>0~;?PSHpxx!k1=yA7hf4zLrQiH+__AYs57K5T%; zwSv@3m%+qq6gh(oQ6UUkN51SP8tH&%04R!#t0%Td?Hf8rbX{<8v#SnKSPgZ=FFiZt z+a)P)YdH?hiT@M@{vOgWmM#JpgCE^=XA9*#`i^%^IF2=q2jhC{qE|t~lLyB^wJ7(E z(n9h>H-sUJKKbY?;|D2Tj)v~grXaOV^YIvdPU&XSkkK9rZeG~0p9q}R8^xYK(i*Bi zPaG!&5aQJgAdrtJ3jgz<;`L>KXvoGX`3rbWA6~)tEpF#>6<~(eTQAUN*}Z7T7;KbH zgm;N;hSFZ=ZtpKl`krGe&4l+c8{ed1me=--U+NBYl|}A^jyoT9aJ$0gA)-u|I<*It zN0<%*9DZS^oc|NTr_@yi6R9=FBkx)$StadsEy)t@agprAtoUJ02B9t~d!csT8WACK z-{!>wPm0gRm>NW@qxlm=;uM$dPE~}Z8YoqRe`zMs7SH{s#elrSz;75Xh=Yj{HJw9< z({rU`Tz8qNaeKzL4rkG_20we!Jh=jBTC3qWgO<>)#jgH9pDxraFla-H$olcU|LGl7 zYs;U{TZNIM;LpfP!v*^~UIL7agPSfti{_nF9rFWNx(D+)?|j;HiEnQ>+c}=zW0nVO&`NQLLs?v4@^Et!<+B3$e15W@M zYZwlpBO{_gvbgoVl=w>#8_6;i$hMjQ0K6G(`?2j?Yz@YVte%L0*B=-P`}Hjx63?sn z361&voZ+Wml^?E=^bxTr(^CAYFbHXk;j71Mm25VQ>B*z>*jyoV^#W04-}3cB(hymV zU?!W$osgPbF$OfIf?h-2HWPCY-|}pAIYSfqDvQyMf0_4zYkF&nPiY3)kdO` zTe`L!xD`X{Er;2};%o@Z!<=`*hDEUM=$+yFZuMKNQE#j9x&jHh%%1z2nu!{Pt}epT z1PkZ2p1^!?iL4ay#{g=?m105wvBBGDs9JOuuw%q7Q9K$0>jjhSWT=>6Xk7oaxKb(+ zAG7<8O412^K_WANOHcq47Mq505w1{@smladQ2+2?d zT$uoJL|1-*sK06YQJlYgWX;`=SBB7j2jnwXy(~yYbRbWqzT@uVHHaNvX2_nFORsWP z+Db==HOK0uAVr4gcize9(!<^p^nZ6O>1DFern#WQT;mG;Vl+HkM$-8W(VwV3ts~k2 zg|z|u;)_7cPVO!v!L~}~jEfTk@PwD>pB1X3sR$FD=j(?a--D%H|G`(%cvG6~dYzBf z9yIX64dcSj@*PW4BZF>FiMh{5+JJ+v^pbGBAzTC_BaH%;iO3OTr01t;B@<#>M4~vv zI<>}NTGLxs7n0@L3C^UjwC07rvk%hy2RcCLHhVy_eJ&;oP};YWpQYnBw_Lp`{Z;B2Nyv1UnO#AJa&yx6h!zRI0%EXkrofAD%%4_7$$l<>;?j?s=<^2d&<3Pe+vmS_33wa68>ml( zQ?H-=aFVh%bJCyB5f0*gZ0CGKB*9jJaS{-7E%{}d0HATMM1%m)FW9y9`;B3XP0$0@ zsNay#w(=q4Tk__4iRu>NIhbPNg$`@jvwu2;@T;~6Nl;E|*~hEMa_HjCqv!{S_vNY1 z=aRBU`znuM)R!LzACv~0k4VK40?z5@cWTGeY?-b5r&rv54#O%lZ zY<9yeAmQLPpUc8!K=M$)1r1Ce-ls+Br44T0KBX4Yz92>QwaEslrve_Nd9cLzrMN!) z?NUmsMVrFkYO9wJnqgYT(wIvQVqL^fV@=`l`(9~4a9IkmQLejPqqT^~6#v+3eg_!#r!)UpKh|Bz!r zg)$Mzg+b)J`u4A=#Sa#FAIYrr*+D*mo&pbT({s&W8b{K&4%iyJZN?3l#xsW%JRS!O zv3N`j@WycVmIgF49DkB7mk$CAstC(#Uwd^0s@>#!cMYJS2p?Vr{u8|ABa2s6bEb$l zErQb_n)&}3s=ZnqjqGIlT0C0*j zgd&gx_UGf3UahJnk;FE{#*7wub2|aF6&ff{sICyZ7vC&qszh07PiD}m62RiiFanzP*&ZY=XGECZm zB{qPOFtj`|cTod-JB|P)w1lx+254czsXSAk1WGXs5r(~j6JTMWKquQSS=$ERri)=q zQf`>BKKMbn!E#or*gO{i*7^2v!I$=^YyG2yIl&dGj5JCb3;HCo5Y8kH>cQK)whzP;_X3}PZP z-Gi5m8588C9d~WB&sgEj0#^EW6GzY3wQ>89pzCa$0T-$f-tuYf=&07h5c|b75s4S1 zUWlk1e8Qngh>uj)PGXsRz~>9DAYPfMLoo#Al)Sd z<4oK|glGP`q?N5p@|#Z~jzxDzD#(kLQc4e^ymy9}=bC)8pCjLKOb_j`u|04v^9J4o zDDJ&pvld@Rd86U3&teX_R}cyGJTs412_Mv^H7EnS-6oUBzx+Cz z84-mGjX1vyO(R&fA9-fr7v6GE#2))M#$05*bVhf`J5=`BNQwk!`R=T) z&4w@SkzYJqY1&ozq(0Rc6=D1){Bn{Vxf8#gHrbO04=vY!yk>vHUrG~Jt6>uvwyIDs zp6Vq5czN#}3=xv;Jsc_%B4fdHx6fvMPBm#~Lk{B}){5(~ka!Eg_rG$`va;Ru{F=8k z-j{iFMz+nQbBL>Ka=}Qce!r#XM048;Y;nNj8wj~6^Frj!xm!kcT4>|r3Iw-*2&{nG zyX8v0>Fz&khYnsxnmL<3h~B7ZKr(nk)W)g-ge1;P@Q`c=uhllkF(2MXkl~!l_R8tJ z)uHlh0sHfUd5nx7UEE<8`=3XGD~&HuD$ki$3|G0wXgDy23llyz_6Qvo;h%6Sl8oQd z_FKkL6uXsD1?lyHKA2Lv7F(1fN_70aqqEs*AD{okHG16?n@9y3t>hZ!qZ#`SP`r1q zR%F5ayz2tGZu*8wbi=fyN8drbDwFLdrX}X7RK?VTL+?^F92$L|a{0JfJ-L7bu3IGH ztWyI}Cl9s@MApzby~*L@O?gPi0}r`$ZP05pjn?NY_@fGvKfCryGK7;cc$kYbWhR;@ zJI}$XU05O()`bDT(Y`~6nHsT<#MhKeG=CLxfn-Epbs?-Zo`y4Le@Dn;0~1NU8(t>R zhH~DsJ^L_bsq^>%b2SFSZdBoIj0Fc{dXhROes{JDgKh<0uatSBfk=G_N3{zVWITyr z=fg^eWfkBF>jeCVpQ(Y6IaT4t z)U$BD7}j*$8)>=r4%Mp^@3b#Nu+axhbaDrI#9JPqwk*+RQQQM=R=eIX)h*36?vVqr zS(yy!h}SrE&;SZuX)qqw3T7eeoGy@#kZwjL z={`!nwREz0N{kY!UvpF)VM8_~aPvfwfRI7Fxwyhj=Q5>28H#QQjs}_9zv7N(ORr8( zHp4+m5!+?FoeW4%gEdmbpAfijYF?q_Fa6@bYmkg@YF=*J#RYb9jl@=;{j>jmY4Ult zpT16^R}Q5}oyxD;e)=bBN%3r#5^!g(e9`twLjAEaksW+pUg!B%_G+RL(m;U>DiGb3 z8gGanEUU)8B)%G5wP5R3?D#f+r&slOtLT9iLLb(|IPLC%=g9<{`<4Nx(`|!P_Y1u$ ziuxy*wKZae+Vb^e#W%m@BS}a@6AGPiz`ZIUON5i>d6LXrG88jMj4`{Q zGNrTU-5Y%wC1T;}&8Cq}9MZo<8Qdf6qSuji!(?BPi&kOf0Z=$NjVyLDA%EhD7W(Ro z;3LVU7t(&vMK~+OxE_;;%(kaLs&Hp4ncucgc#OXpK*z>dBgRi#cS&?ba)lVycyo-N z7)b%ss#XtS6xdQj@EqXBiayo$#=TWI%0ruU|3-2aj2{$5h#+JFjT>&`h}i$oEyoOjo%J`{SVradnTUyhd>H0f zxf(!a89e*7wd?(swvUA`w` z!uOio&CC3$oZ}$;L_&IvbgRtsQz2q#J$s++$B`a^T&_a5j$Y8;`t|+GXthkjR-Q$I zUef5l!Fk-ALht1%{TnYolIM_+$fQ+EG%{1F@kektd>pX47BW;G2%5T(WKhwPeY`Y+ zAJt*wlL-DIS;!i^&ag;Ij#zr{CW)Mik7Cmemoe(h@qR_s1uhoh=P(Iw_!p8mxPfFA zR%He#L*Ynohv0HoD?SJ>r+JnCVztcewK+c8Ue)S6E;7gZ+u@}3<9cYn1bj~;;~!*{ zQ6vEYQ*xA3rCo}M+dfwK-mbQA-H1uKbU(Kmjzwk>`LL@$=ax4b6-#F?_NNbML&e&t1bYfK zSAHLS&N=!tOIaZ3-fO`L3Vn-hEdS^}(cxU-qbD2h7Z^uQnaNvkcWq0Bn#jswP5R8Q zZ*-=ukN+|{$OSUOfM8RYJWpY1pz}IC%Iif7B8aHu2S9T@o#Lo!%E7|;hjA^J&pGMj z$NQ)0NYo>#(9e(Y(OKa5v;_mGhJUq_yU27!l$t&JlY+`!JguAh0k_x|M{_v$fVJ-B ze47D#vMkWq`8!aeAs$D?5UQ7a|9Xq|((5w&TjBd4Cu2&U@J36j7Judomo7}F<_n4Hn3x# zob3IXFARLJFk$Sxw*BxH{^XPdx2SW9do6`+S`{vC5D#Csc7HjwR(B(%^fUfT3G;mNZfHP|x+~P`tpTaYL=$Y9{;lEj zavwBHs0}yJNTXP|w%Zl+ZNV7Ve2fHfyiv|iKMZ{+!ZV)|_vmEA1kCPFcYAnmTX14p zNb-e?F9Lne&sAsHy)lft5U`21gvI#SrI9{I4eO0PjpPPxb5ssVbtik!kb92)tL`D% z`NbWm_%Blp*{B?jb+b!uN<1Zv6=0pVy^}S}&k*_Xiz;jVDvhvTxlsmeMs{#4Hn<2L zAU|(Vkhj9aHosrk8N6X2@EsIJ{ZerI8(0vH)W#T?F$SA4hzYSXReacqQDxOqkJ#S% z(PhQ0(Wsn(zjEq>lk*wsrFO)^p#UeLz!uys&F#Qiz<^-1@&*q|K`gfE0-4`vnP#+| zRp&lF=IfQ`Q2lH$68jAPiomC+%q!M>cZrzgyhqS?0Yv6mmJusr$-YRCp zOMd%9kDj5kLR36OT2(mcK=O%8!2NvAWY?L{4rNl z_MCv)S}8G`l_)=bHBaYwdQ>0CY(Fpg7%$(O{csu)*kpa&e5DUDBn#p$Ob3_N|-*dw~I^le`ys71+=>-kacKPz{^-VWpb^k z_OXk+!R22(Ev0kU<5v0aRQUu&m~O73a*ue>WPVo+95H!H0C29Qo zj}V>rEjTGtQ<4w$6lKm3v?#boDTtYS{z;XHpa(+DX>8*~4-Bv2F??XZpe7`lI z2Klk|{epI3cx0pbS8jXHG>{?qO_;*TH!wrzEaGl$ z2I+gUimTXooU8jB7mG6@tIi|@yTE2m-D7$FV$3}ID!@t{QMWAXgGd8b?y=z1M8dH3 zkZY@0%*a+c;{aRR# zr(kqCJr-OcUILu5Z8+%1WNbLWlqzcqO^Lek*xOsEEs0PHY5C`0 zb(UFEee5XR(+0j93H3lq{8tW;H=Xrq$cDE6*-1DmK?Go zP!}QnbkR$&oADQv`_?Cg863jjpn6p!p^>lDAgds7S{bRScz<%hq_B8Z=*olnjTF~s zy@NNqiqlqTSAHbK)~lDT!QiyoPqK<-iHsfcYg7qzp+V z@v6o*Cj0o`9OULmy7vB5leKmklaQ31@Wa_snn;-GDIIl+VZ&%}^|=>!;#*a{z&_%` zq`gmIsX>>}mExK7hxl=qRo@TWHUZPI82n%m#h@Oho60A*6n}4F02?=nAVrxeFy)9r zqrYrQ*jXIm5ml0ppt`pzf0>D9yr?Emp#Qkf#J}iZ$&ejR|BV3~bM~{cd zM(CDm`<0`GWe~1Z5?K+AD$BQX}780Tg>RJs5ghYrJdMnKH zGg)}%wx)AjH&#h#7BezvfaU=>=uRK{g`HxB!rQkWbDQ6IP<;8OTmkk4(P~$OuJ0sf zAa=TN);xI>RB+H3Q5+oiv~e9ps#JGRU!8S*a<^0FXE13@a5{=Q-sK)rPRT* zb|)JiJ?CXv_iO@JatFL4TaKC%>0T-D0>?>8@3)qX%xumkfPJl>3(@rGz_IdQ@8(!Q z|I6wjM~JH0)6lEp=nJfWYb5{4LhxE!ocTk@L;qH9=nKBbh4~Y{;r&$+C(iS(jA0;U zpw5lhd2#{xz04;sqp>0VZMx2Ohudhz`ShU0}LN>po&@ zWXR!ePZ_tz7b2#y&%w^~IsFON&}QS< z=VSmk+s92j*i2M#a{VA{m?`bs9pGr-Ke45y$vH>3=rHB~e(BvLyv%@;Rwp1x&z?o= z8Hf?qrG|bnwfL##D%|a;AMeZ&99}Ni-r)EJXcfF>U>0%!jkWH=rqt^;_+E+WKdke^ z@wce>)tnSEE5FviF&Tarq<)8xejG4wXAH2wkpIW3FCysKCx#8iU?HqqX3DpdQae|a zC7!_;$Ma&*J33w7qPi9M{5}iSV)ZCZ2>j zF-b}UCRfHNC#*G2zQt4YCE!g&srs|9y zC6DX&4TcHF+S#1RvrRKc3N-rY{TtL&r0i-y$oh+76laxKfAP4MdhR`@GBDPZYU9*3 zCds`fYGY`!?f=oi52PCq(LKEU&W<`CZlkh ztB?ra&8JKyafrP4E@|_fCo&w`bto-x(CurFmw>*z%8?;MI4&j~*0|4}H?gpK|(`MHGzox+_RTseYOkTE*t ztSn(zN880tQZAb6A5__jObX)sA`(`!&(lw{Pj z;f+J^G{YZ1-1zqk-{r_H%Ox%?XGA6rm8=6lLH*E?t;?9f>QRCgIX;j^kcg)$Lmr7e zYjh?szR)tYVNGBwqD1PiT|_Eg!Z*v#O_nb1FjJu9pGT?Pcz0h{BDLcTq{WFy8BnHn ziqn+a5vPX(5oBeyn*6GCTrzHwKm%ih5wn(qqSzQXH*g@m!?$>K1WE)&^_W?9+EY+P zS`88PenPc7<_08~{N-||E|W(rh+?tpZX0EdtBO|@lzYf zA;_i`_MwoK_N}OeldQab+!M;XQEt@-Mwb#%vz%*+x=P3Hh< z;eD&o?bayIAnF0>n7;@9_)$pxKXLbSY*v^EnP%$1Tmwe>Dg~M5nj$1C#xFb;o1`6^ zYldD%1Ph$mb>sQutV))2I}!#l#VF*5=ME~nk{%~+lHK2M?|B;sE^woT>-bIX`ogsZ z9S;be`wh;zxHbv8qjeJDK-NvZN~@B2g?ayZ`5NgllcEdpiPJ~kY~p_XOIm}js(*ll zIkzP&L>uih67<|TUe$Tq1?o@)*yZe}qDNEiyI`3D%#h$(dTLla8>%I;Uj`S=fM~qHe3A++2k00&{ah4bwHKpG$QHVrH<+cq)B&LRk{$ zD{L22%1bS$t7@?H%|c|^#a>%__&e1uuHSfprR|KFFE>+?)lK|ghYnVQ)I5j?BR$K( zER`Q9Q&JOdi#?zA)!e=!>L?9Egb_&sj&Q~|2&AZpLn9|wqy(QNB-`-UDq3Tusfg!8 zGcxcf&GtKY!`Jjz>ZU1A%g$fK=7^>Km+UAw1>;;rVqfm zy-Uoc8r!#I!I)|K!Ec_2Y8=?_$a(__{a(X(}yO9744v z)mH&ty+t`PD#ZlA5DQgeryo7ER)|J1Kg!R7#2xPVY`A6Uxd4BO;t+(*^;ioU#2q=dJR*co~yu5eEkBr}V-n4_lS8DBji z^b2bW(m=#j&;Q*#lJ!+4Jpmf5l=5=?8YNWi989UNCS))Ozx zwhXSz5bJVe8x3@c^g1203bvR6h75BlaRK4(kZwUBsCJn0yIi~cc4)@cG( zJ-g{4=}s$RciTMMtJ-MqhM>C~Wn=<{F8PWOHcHlTE?n3$U>I)WuieBI37bVr^@);3 zTEZ>iH>s+XeQ;1Ba0fsS>yTzY1oExMjJpP$(XLXYtx3ou4i{hcQl;I!oc6c0%d5Q{ zn-5&^e2DzG?2`fsT=fkx>@T^XVXmoXgl4r2TLE^jrdHgPgm zOeRV1L6Vcd7yIw8>Ay&ICHQ_mmZ1f@0K^aPs6~RYJL2#aP=R@-0-s&^UDa#?c0LGX z@SUSQX{j<#6OCpE0zk77bRf{Qyk9H=%-ic!)We>RJ+nF_TR}Qw;9_W9GzrIBj4oLG z4x9e+tXmR}y^TVdqlK6}M;Ptu`r0m18)#!g6{@lySHM1ZR#W|zKJf|oeGF;`IJ$_0 z<8=Eo=9sf9i+-*0PIZb$FFdp8cGSfUvm8(~CohfLOCO@NA_89ksT$q<_GGsN=ye++2AoSF z4}te4#Ie6)pA9k>9gX)`{=ESl6-FQ$^`JBt;&{1$?d6ku9fl9msgDs`E(j`&+c7G? zDu;gYvLv`YN78Y|5}iiSeu}lUz_o4>$0%0NNJZGyz#kQSL%bwaAzF5aXYomZIuTnD z>YAHX3yiK928UWK0aJ=5ZL#&8SoWRnYt|UZO#h@IhQLb1PdLVy7vtLpY=$EejnM;@ z4`91rzA3H!r3gi#?jAw|9oz>)x?v=hr6@n~-3%NoWTvw{j|-5H{NB`qza5fCpW1rX zwiF4aG%)fIxmdZUHh#;C_o9l@GtIsSq|82}Kh)P5HYRw8(Y?1n5eDcPV~?F z_6^Fs4tBc^1Bu8RJqb#4aqEW?j+x&#l7l@&2f4`mYINhWs=yP&Z zuS)2I{8EJ*ANVxbi;bmp-n~xV`4On+(o*|gIaD-WaCQ&7r4e&I?=MjV=Pm@rh9Ys@ z+le2ddn^-dwMr1YR7!G}Seb2#-EY#O%S9Wb_3AHus!~&b`H*wnJkd7TrYqmONIohj z_c0t#jBC|c6cLqlM~dN!_6FM%qZ-2#qr~NyrTnDfJmVtSz7XAtXo@LmtJ4jOK&LXH z3Nz02+lupnK`iffTh$&o@U|^^yAxKjxeZ~kz2VIK2-}hnRZ=o zg)tGDS+`br|5+ZfVxP3z^3ooFZ4<{24m-Neg&?)73K@bo?wr+HWvg`&{DIw*W!gJ4 z7-33zP9{6vC-}qD#g2$R5tuv1k|^JPoHQOIB`&H`_)%9tu<`%GoI_wEfh%9Ckscr6 zcK*F*S2G%b&BUi4_?)OxzZ;s9nq?ocOqK)Wx)hk_>A9i`Z07d6OK2RHqZvZc1Hv^2`T7J^l`+MM2Nz^#E4cxd&EyfRhOcJZ zdrwOXd+M8ka*GB~QUj3%PS|I;4wP4FV*D%@VW}kIU6hf;*(lBU*V?42A)xO5rRz4I z8X?M_Q7BPTdHfptvboS}ftfIW_;8$6n+TOlKabZT<&N zQx9v2jb-XI{JgIwyyu)!wWCa0Ba@_net0@~ET;9>8)la;J_c?Q7W(nS>U|ZLpr83C zPZP7hgSxQa%=6=kI?y0Bhaxll9^6ZyH!H8)Z>Gm!5D_8p(Wgka)SR#^X^>5~L}7nR zQSfE4{k?MgReVb52^jWcz9q2txyG8FB>0Le+wWDr;g{puVwo*aGTzli$%EvE{4;#Z zT7!QAt`{}Xd!93KHq#g|`d9m441VEU$pZifYN=9xyk|CWA(Zd$mgzIY^x1TX9r+G6 z^^dHIkKb4xeCEEe0u>3u=Xm^;v&mNl4O~zm5@Qt?I&a?176|en30K2=ts7z7aJN>5 z1?!pct;NnmCaXV2B3^+4o05CpxLVGmwUxwSf&`K7A(986&-U3J6>r4aw_7hefJQcE zXT5B2HD;|l?TjCm^6^0>e+>Ma7>c%%tL?hcoRc17&m-Grfahe2O`hCJb>|uhowzaN zg@?mckQoQhA}Zi4#{I_v2siz6tPj_Was+Qs0WVxfxHDzd*B49z9Dr*?chGH~-QU}# zkyhnr!oV8pS(cAfb_cLjnC7Rbn>_bl>}O*SW$t`W(kHr6^n7W?C_n6k11#-jsfgd1 z4Pqz0)g#|oC_eXm{b$KuYnP$%MG)Zijg6zx1n+nSTLSpZH&fb*9BX;xfZ&7M@=1Ut z;@imz)^*MLViO9M`uJ#EXl?#m1zIcOTe#uRvfj5r?07j!&NyixXF+AlD0f zWBy?LgO74he*}OE%T+$HA>0vo$fMeHtfhh&S`qNfVT)k$okY5IFzMPkOI~WN!jSN| zuM{AZ*GzoF-;8v*j-yrP9}h)5=^~fB;jz1d13-tr`8`g@^Po_hRyDk)G7Zl1qA#`+ zKJ;&7?XvB09P$7!NJ9e-h_gf(sGaB|B3Wlx&S{TilL)yz$>cj;qeK)dvVn%tbZ z<6mkz1*-??zs`HKM*9*GH>DMzz}Ki$395x_l+iJ)O@)O}deFTRo2@vZb8361h)t-e zL;X=NHCVDXaiicLwWZf@X`ur#?kq1S!XG^V66?&{y}f<*5MLbEn-eAwmjcyl8{8;;7I^ z9k)c`7JZ`{pY?EjQG+TS26*TL!0)i0_493CZW4fJJ~29k!?PT>Z81lGx6AU3Ef3bf zA>y(qoujkUqh^+hAz>PfZG5rW4fxWZ5dQC~QQnet>|{}Y&yIqBUZ(BM-hmrSln7O|1;&l$+v=*(?h@SAEUceqhdZwOgq-aB_=#yB-H`cgs_l>A44BQAC89LPME(_v zaB(tf-$o~|)(47{g%wS&86=6sAivR)Q_xX2)Gz^9uXraGlB=k+7xqbmtLIq@Qo-7nlA)-N;tSy%{LSe42tINNKWpFT==G`QZtI~Bg%SJ{*+aY~_1hBTXb5@!H zhfd17X=&buCh6}4Tvs)w9Hp7-@wmMLvQf8zhJ;4Oair_PUi@4z0FSCST-)UhknUys z2i%MAhAPQ1U0Xzk**LNM+1bt)@Fh#H4wnR4pdm&k$tH}uqKS~w669tyrTN)sYK`~1 z-MlK~J>F-T`nNR~0(or!O{~!10u`l>i<`;Kspu9%5BICf4~Yy~e~6|Zo;MFpVw zF8Rk`N4xnE=-U6xOvcf{JK1=(i zLuww8W%WfdPsju1{Ioy`DUl%t3E9va<4e*JHMBFq7av=r=*IUE`$Qc;6Mc=H3H__> z92)WCpU|WK0^S~iNu@8RPw(S|Jcf@D!*&-d+m6Wi@Ay*d>2##jgP=~r(;ru_V=FmG ziXPUx;T*K@PlrqXl|3GmoXAn_xh!=6h=)OHqmrX;1GtKtDH|5Q{67Es^h<_av0O1t zZ^880;0yPB1-Q>N)hFS7=A@EH2DxV14s}d-6E*+=gH*q8LSNT+IsLYEbr+X%Jv-go zTOVO2(r7*b|pg7Ip*ByHZuzNgL8a&%tr z5?HSz)YOZ$$L({@MyXVZgyKX!9%HOMfU|w|D|@#|U5|$6D?9}r9h^k}+gO$RCHy@4 zNaMM~9l1D?HB1gQ(p1hAA5b8yddHfGvDj#AYXp9Dcl6b$;iwwIblS#sou|2?D3vsux6~C`nOU?fm*?Y8rV|A z4jvWtz#*cl)TF|z%G(`G=L_+^J_ri(lDSqE{{umV_%DZ!8D3aDOtOb9bXnUN8cXx1V_1uKnnTfzbd#5q(Npf)U#N;-cGbNRQx-(H*=K6L zDt>eyu1o~HtLEcc!+5=6y~9+_NvUTkMx2HS)?q-WhE47L0g_G^R^bW1zELJ=@ttAQ z#Xn09r?SBjbToQ5fy@NT6-LegaY$&G~tme}c za}#79Oc%f=D4|#f35g>LL4-HT>Xr~H`WBF|4J=iKp%6vy6@O{QD3-#N$*-~!A?&7^ zmlp9WTkcIGF2si-uwOVx|3`(n^LdbhdJ;!(BLanWO98tH`KrTjda-Y*H;-zA(Wic2 zTN3?$ z(HjFV>_lFoP8NfM-e6fY`>XCj#L0E8)q%4Q4>YTEja<*AO>uf>Irb#SP*t7R>j$u! z`}>$k1EP;Um*h-gMlSgd&*7keUUpb+MN2%@Y2et!D6<+*SfTy3^JP%>ngE&luRvcu zA$(_vL5G@cv5~qU{CoSyxDGw1aJ)e|D#6-AYE@-~n+|TLq-Q!_fcX3pBf$Ga6eo#a zV>zuZp(NxhsEbo@kyZ&Jc@tvQ5q$ zpB^rBW1tavvctWJBvUmzb-oz%R1A*26Ni^RdLr-mZ~}v@QGFR3JuU7ffi0^YB#?fAa*Y!S7*tso&_O&P zVLpuKl&qgILG?yi+P87_xY>x`FWT8%%tZ)&f638@!IQLf3dr z*xMyc{Rp!PVdD#1j@;!?qUiWF7y8&h{}mK##GITKB=}5eNYNr+Eq><6{eb=2MxZ{t zKjq*k**Z!O^m~jb-Bl}PFHbLmBYch%T4i?{=nH*XF2J}VKse0se9xRYq;1+sy3qHO zXs?DibqK&PNL<;@?EjqV-Ezsv!pxZ)3O~XS^Ia@om76NSv?tpY$Mf!2f6u^EOE+^L z`thz2vB28Nea*$Bmx|uQT_)1k2B^aZ3wa``JIdKxYd5C`P57?+fhkUOl$|bZ$>xL(MEN%Dqx1gfkFA?;JuDC{AUnw*|_pU z8O^lB-*63tA4vbbg{lAVAA%koslUrvmw_z)RLj{*6JWfUSd%OL4Hwoz;I#I^7=Is* z$7e508W0Bc{SA`@XgS~)PZkto#|R0TSe)&^KeQ@EO&nEw{xgKkar-KXu_Vi-RdBU8 z5sCo7fQB}$%=kstrJ}$vq%-38(+>L}cP&S%cr29WFWZr_*s52GEL23?s4gEUi2tbS zaU!q9e=7P#3^j<+B*^*vQ{fCQy4*2hh6PiPgUuS|;HRdQSD){Uerw1@4rLSI91^be z7xzAFvDRey9f3%WC4x1WphRS>Pq9{coy*mJ*cAv90$gH(iNi4evP6if3n3~37_j@gF2@>34 zyK_t3gUlfV@cB>n1J`#v>8M9t*+?uhP#0t?qT*5%u}P`jIA;M|e^bmq)BZloKh^MA z`-JjX0nokOnZrh3p*d==CoFa9_a*M_i!-XNdx(6)(&->2H$>#P32b?zQ<2X5ihSx2 zR##=kegMYJ5R@Ks`-;5B6WllWDaDEER1~`LoR3_uqTD-6L z0rB+$AW}BAf!#!BR3xazO%;XKmkKpItz0R z`N%Nq%FF3iRr-Rdj-dqB&r8Sb|IUUerm>j$^E(;5VCH9(7wYSj*FlpUJO@rU0BEiW zzg6aBdg2xMk>`4^6{}%drPbw|3dk0#HlYl}Wi{;Ha<|5Q;MPstFkDFjT|=zb?i8@A z`zAZpBGX9a@P9oGu5Cz-cdeOK@#xpi6}Bx1bTPfp&&2)&#zbwdE1&ie(=;1^Smb;- z*o)TH3b;-^d^d^F{6?X%_W_0?L<|~H&bq^PP~gA!RynBOHTx${PJeX=ts+nGai=Bw zm*|HgYc?1nOL_t6_T5$f(0b1P!O zhZ#0>!FnHJE~NqA5Zh&4St|$3L-$KB{GZt1wrtSg22(dTiM1s)@GeDe1!u`O=gSaa zGt_TvLv!on$u2vAWDJ0qP;e0&oJnCfoHYKW+IxwaV2~oH{aF96Yn$)`YNw7XpB2$! z@9sY*c?!vyfu$MszSUPzyvJ zRNtK56;6eOnf~pN1CdSbbglj{RXHD!<{TZc1{oL2?LrA7@ju(@dN~4*RNR?*2FJG; z`znYd-0Y#+$n`A@74rZE(1&6y;p)2|sUF8fV(oERB;yC_AAGWr7tyN5(>MtkW#frMlmjYNGc;|CUO+_JTwfUbK`+4tV4`ddBnGL zBOYKQa{VEGF#?(b8MNz%Z;xqXXut(a$GZ0gw4u|U(}YRcCRseXt{}CgmS%lWu~))F z-;kdc_}Qw3E@^5%re?HnZQc5A7~eXNr19}DSnNVA&(t}G(<>XUXI%A7JO}fnF{a4^ zvI%=(h)`@2Xu#09re*-}43bk74?)sD;7oGqhxL-{4p%xBgV+;!`RENBx6)%w?W|0{ zRhnOYWowala&>Q2+v4Aze#Ra;H-8DN7QkM!6{amMH}n>Yle4uj2B+*}-i{lTD{sn$ zCVj#{umHd%)H2#OI}`7Ee>a}jYM@8B>Z_QkO8c~FMJ=EN-}de4s`2B(36FU>gSGo$ zYrApN3z;#Y75wJHw+>64BZGDcrfFu($q?x6tjyOQe5#1QSU>v(AKF<8K$K$z zDh=ep8P7r8@t;=3w$ddgNjzlf(#UxUP`C6q?#x5fxp=<^$ImR|_ea@vqU z2Yg9Y=;htQwkBwpx6dnm=P_=HzQM{u9C4)=WksLLXhAr=A}g-iNM|`VhX7~>#?9DY zigeE@U-aW=^K6tI#@VqZv1iVwZJu#MvW~j#N+yv0MQ7~Q_PH5SNSS5RaN~b8~up*+S%x=y2aG@DU?UNu*rB z+aoaIDct7p3Di$7ujQQG*4!CsjFMKj?d})a@iY>Q=+F4c)s0))L3o{u?*r`VCmKF;qq$q7hK4WA%|Lbe; z|1pRB&*$NPuVQ=5x8g-SvY)i;lZ2v-Zt;$piyrJMD7-<1TvLJ3 zG6Lb$9I1@}a19*c{#7dG$Ea9;2wlwI?BJgkO}s1ful*2WD-bK-UdD*iZ_0kgWuWWu zogn}PWS}$3i-Q z>|0nuslhF?4UYnJ{sVbMO&gaQmaD-8>HqqhszXsOvlexue&pUyN{PE0LcAsZXsG)W zvpZl2NPArMJ+ImuuyCM3R7%fJHoauh*~&)t8zRBX4NY)g7?^dPv)JN?C^>i7hl460 z)mAiOUi<(6xk-g2u0h`-6ac*wshcp9(9sdRQ7qBKGZvKpWK&pHOj})@EK6v2Wj2@=XrmnA!u=z~U~$E?{e+v+sDwC!3W#7h^pCWFUU}yH&9O z!DZt#^(t)e_<>{ytuS+YEI`}6e+Pgj=8Z(yqx#0*$7s5-6*LRQeCAuH2(eI-cnit#qu%indT3Cw(Hb3@vHvz`i*u^|KrGKEp^gdeCew$H_Mqv8c|rE% z>`ohgn{6Ngz6jlrK;4+&wFUJ4`qIOiJJ@}51By$`odE>8hZ^wbYktWxEQr5;P$A-C^FnoG9u2Zcl^{S3r7% zrAB;Xaqn}y%?5z)ooGHye&&Ir4Imoq*wysCjUPbkGm~&pJoWh_+HB>u`WqkrcZdH_ zG%=!#TE$w^K&z`x$_HQ1C&^4=EE2T(~?tWtf zn?|k#zH3-32sSccw?3L&Kxz1&nP0$!l$y5i+jOffF3@6`rhuJqzOSE(=Z9418eMvBT|J z82XVBGEWARb<#nl-cN2M-DS}7vAg^EgKbv*)N^k;rpw&(-@|~l~ z$Ybm7HCJca4)jMjVhtweKmHsP9SZ;yMQAIik!lcOfKQZ{#_{kFgj+gr)f2T8c;nJr zU7sJ|WtLgxSQ?zNDUf~iUEC|1|LyDqC89<&m+*4>I!z^Qx(3D$i=0K7JPhY6-itpY zlNfM^5!*s!ooaRwTb`#gQ*wk}LXaGd%@4)-Z`jbb4eukUo(gPRr+ zQnSSmuZL#ZA_!94UY`OF77BFhY4{hJ@;NtY^J1pL2Qv|C2$XM^ zD^bV_$4|6;w2cR^+q=p(<=aHNu@qTkb8zQ6;F!o&E38a}|9Rk^-TE&v! z0CDOs!?)M+l1alUM^|>!PgcO+;RiN%Llj>YPeaTOGIypGt5-0sjy|@m>0waoLE`|9 z3`t$33~_;+Bv2DjroHT8+e$g*1flc7w*5C{sn;~VFd0l&%0f1=ys?3QEQ%y*1iy7u z1{~w3JUE=dq%hx7XMfs%i}sLF3Q_v>v1i^eVo0_v{Ogj-+!)~8$AgbXJ^$;FmNW4t zX;`7AzP_oAGc0W4YdQnGa92of2P`o{mkX(N0fbZPqvr{;n@Iy*8lbUa-A9x>>cdHUgfJr?j)sCTcnf7<`Mp}iWT)J z-ct}V4Mh-s#u$0#5x<$%j4T;jb`#-nEan{ohdNc}Q0b}uu3*tJ04Gz5q5Wu)Dif-< zL_hYr8Oqq?cmgeW!_^{)p|N1@3_>|&D_W_q$6m1U?k~^+dYt*;v=TNI@4FBBw{pm^ zl?Xj?PPNgWCynmN^q$LlGe8x}+gyUeeyDAHoIZoN=_60Uqb{yAdlbQ9@46*B4=(-2 z>^wojGSAE!WRu?ns^D=-Y)j}MhgZ`~VJob4&SMIw5|RggFi?X4&OKN3FXpD(V5spd>$l(ip2^;M8RU=;BKvydP7(&$Ctu zp9qiNX^+~|XJg;?*K=0E!a_XZ5R6lJ8c_FdOV3Opx2sU&00bKaJl% zy-LZk+xc-bqL})EB{G6mtS&D*{XMZqWqaRn1s3fH43#^&kk&sPoP3WKx2$a+EXRuo z;br?fVb`QyWc(|dAKiLvA07+mXXzriou1aIpfey=o&5Vh1K5W~-Nz#i9d~g}{|8}j z*%fEhZfh2W1lI&7xVt+P5IlH-7Zes;5v zmmhV^XHkTLJ?Zs_tj5sRi6VtMJb*`H0L*c8&(nU#mN+7@uYlh31Iz$*9mQC+PqA8J zw&GF`oX3;HzfSmlJ#_*%j8Z*of?mk0#g?i25b07wSd`&7ZEML)Yo83tQHOz@gpp@o ztAhn2A%v#`%AOsc5AmW%hvpUcO46+L2MpmtVzWo?=z}JTZho~^lhIsUYI6UsHG6y4|4d_Bi zBA0HQ{FE##sY*+*WTf6F)H#Za1Qtgz1TCyDY_uAxW6Nts)0(}sPL9nO&7Qk6}_%Hdo zx0$d7ioK1Rx-NpmcMyqgwC~}P+RdSvX1;Bb5dRp>Str-if#f6~JAaYQGO$RW^0fUT ze0S2*0x#tpf$iC2ny?>y>C; zmax@xv#T4Y1g~uf3qy`1QEFS>+-iCV4YPspIuYSt4y$6EtBNS4mbB+hMw9xib+z*9 zr&s=A;>5VKB5qFRkIN=$`P(!4<9c;^{bKRPG5Jk$c|>rxKSy%%UmF9=b_RsMHiVCV zZ9$19|GuN)geJ`Ixrm@g`iAhO!Ywaoj`unZ4tmG|RLoT{Up)meFb09xZ^&j(?VZ-X z8MScu$;;242=Ycpy~v7 zeQ#^)dp9^AB#5dg-NMD_kWtuOf!A>EJ2r`YWCO2*z<)q}H0Lq8%zEl}gp-y*xP~*G zAK^%9QQwM>PV4L&$5&-he+1WZw@R@!@9hUJ^8t>4sHwZTxCK!PD!BPE4vfKUCUcq& zL^`dTwNYfijh4TrW~DP$fD{`<>og_W(%~>H&o9B==W}NJRW*=W)IJ@wf{4q}rR`$0 zm?6Wb&ad~=KKilF)pj1age_-}@%31L_5X|=*YEhd`nS~bbbgyTLOGD^D^^4381%i) zkR zygBqSH28TYuM^;%&hfwqAqWQg5hIXWaxS!_;s=oJE5a9yRN=qE^x;?OZW`)!J(9I`3o7UJr+xxlJ9B1BaljK9q`#v7qX`bjgVa?sQ(S#9%h@TSk;fK8bXV| zU{l0T?_>CPyiX+lnw+`!z9?|@gdT4`fx3n49r>T2D!mSTfZbE%Jf(Jki#tXE1w(mc zl_saJ!?zc`mBPRE+f_fR*qSkTp%KUVQA$h8x|v0Zzld5@jDHZ>fP*BzXkrxn&RmoX zuD(`bm7hXBtJ8Q{a%RIt^f=vXo$pRH55W=O;H=(N%;$WUl_3u)dCRpQ`rfX*We;5{ zkndxQ?$?c?_|SUFQ7Fs#I+@*B94;pP&(c|3VNynSCDkNE_1}5T9pkvi_DsytOAz*-5!M$ejDKPyV$IWVgb<1p-=pO(b~Fx2%`M08?~4eN|nv z(oNOkt2!|~BNNBKPNa?Cw)EZkR6%s=h#%SA?r^*I{FRXcnwzXlOBFQsJDULe&xdE# zhb}Il|Hkh;h=yO(CbPMZs$X*}4KKYzwu{fh&gTi6tz)?_Y$=?lAmj_(6!+^S>uTl_ zM?#@1A}x$`QP&WzRBq018duV8lRYmVBKb9tPlSXWWlB<~+C`$*GZ^$kVm!7Bec((M zk0X>@f*{_N>+Kar$}+xxL&z^m_ohqLJr833xId#!HeL}i`rPZh>LWyu=4vpHN1z36 zg@p#VPVAGX1Sh0)cs5h`-1GwUcL~1Slgh97kb_iYN48%cKboYFd1@AU;@r7!$$ooG zSz!HuRV%u9sXSiN1BxOO1cDL7Ogn`9n>E`}$W?R9sa?pPGLY<(-QCbhA6bV|DF11Yt8wP1U2+SjH)4bbsHAbha1A7H*+_n z!cP9v6Qi0K6uiN-GyXzCX=EUDp*21|w!QEjB(tyFuvk&HwwZ-CrE^i*B>)>820kya zMVA@UxA>=+_;Fs~zV&nV!U9h1RPUl2<}c#;pvX{b*83a(KP{lxMmm_RFe)iNS*;dY zj%&OX)d?llnc7{2n(|H?n}S3YpHz8q*BzFZ1>Ai8n`ibEmy{`(+-B*DoLKA=mwA3Q zoJDDAhAW7}*p$j)d&%M86B@5hw_8k?V0Z6FAxQ@6yo%2A=2l;iZs$Gpn4HXLjgWx3 z7A+g02f>SqPKFapM?c&})OhyRl#LwmU-4ntsR<({?u<5#qiFuUxte6?fkkNY*TZI( zmD#6OVx0n<)XhxKI%n3BQ24-fl!HAcw}f{YvU)-5L{!MbM3OY7&*< z6JZ5pi30;3)!d=W;qFgDRP|Qhp+!dMaiCi&rzgia3<4?D9E`o*6|H^?@+oBCbnb&B zxtm_{`f_eC32N$TVpf?RR$PbDF3)J{ossLvdcnEKd~kOX!Ni2sSF>r`F(A$>RS}0! z)mMV>Bx{mBHom|SWzHJXq;-@$n!?lL^UeJP(Iw=JvxN{xu5%N|c#(;}qrIG3-gbCs zp)T6%LEk6+F6yLK$_&~RiiEK|3&*Mq_~zVZiZlq>6!yl;3DI;@iyns&C3>;p@XFqY zqLv1E8AXd`dJpUI)TVP{DND46J|Js&*nK!4P)X62binNOeDO&b6p_M> z;lHMDhdUnZ?k-~Q0L#HYI8~XY-6s^^JHK>_LibUi+^*99yaP@`d>Mg%?=U1OEsXhP z3i*pH!6}0Q3;te}L4Z_#DZ_^c*De>B$j^`2QHW0`fUe(za8%yLw)-G|-yjceSVC@0wX$H++TO*+KG3J3+c zyl;yY!%R;Yd~aC3^JkhUiSKW2sy9Z~BI5_ZpOmIL5F5B5B34 z$wYx8Cxn*wqB1T^efA#f4#fOUY_UqcdDXs;hdlZ%Du~T&n+;D$N4pVs;7Xg!(LXHn zvd@wq<~1T@vLET834zgX#oH7JPygpC12ru2u zxMhu1or+8PFF0i*2s3vm!Ev8y1=YM-1ds}M-mL6rAxdHl>a~SOy>ag%q65Kkys6xj z!f?Y zWyzHq(_d&iCR8-|c6;Du)dXShFQwcHz%xwoT5U!XtrEH43B`Conu8n|{Kpq;nnYbR z)cN$=uL{8xr)+vJZ8!anRkz1pP~p9aE3pmOm0W({`<+Ft#Vngw>q_$dsLANKBy?4t zgR2Q0?3bxdc8Yzckg|3<3*=xrUpJB9-0K0xjuN*r%N^Fpb{x8ZxL>|egLKgP4=4Z) zx&o)1FFI@MMnY2`8!)}8G0DU!Beqt&;oH?%pR!0_es?hTegzaws5p;^eh{g-I<)Z( znCmsdKYkHwb48_(vu&4|5EHBdzM9n6rm4T#|0$m4jaBxC4*JGBmZLi<+W4vON-nq z*8LK^uvZI9!If_P_>ZK6`WlC%p4a{yfh6(DpRKtydj{ z@;C+RI9W5qFgp2fXp{f!Y-gI20Q^mtv$?mGajZjC5@R`xSgzBLhlg0;qi>KyCoc49 zg5dxydIF_!e@-UTZB3{~ndhRT&Y9vRUFG6u*_Vj#q#trQ!R4fs zoI7$hs;pLnd4gv;*Yh;J`4dfm56M;MJq}#x&>+&YGDUO{;t_&F@x%O@L60ci*RVXm zCcObx1}{FqViW>E{epOVZ-@~cRabU6P0{_fw?_sfGB^ox5D{hx1D5I{*(`cU7Y6LX zSJl~kPV1j7-J`jZBV35%;aRmyY@t$}d#jxI&)4PT(Y9_oPa4y%xyoPSQv$1I`1+Ew zzo(PoA6odl$D{%o&AUjm@{LolbugfxAcT)IcFTF-P_2p>)xM<^4jj?&&98AyU3w=Y z_hSJxNkX?jQ$9|Od}seb5^cRxuivJs9tU^G8~G{+AHmQ6#o$GDS4Xd5MKKB_4jwTY z=sDER3D2#GkVIH8K+uxShs`Xb%-zzj@z4BM1pU zVPSposEW+j7b=o`f@mzQzf4%Jye28cd8@YfAKaQ)1d>FO>~s*OGhGkkhzrk1f1#N! zAoX4Z3y$pri8{l~Sw60B<`-2tbe3gCBs2qb1&Tr55S1EqXtYl7;}WnVhow3s|8(Wu z_=gADGn;J>I;xQ-oQ&%QmQ3&bn^_DwDi&irMCZH&uc!B-)9l zS>_*UQ@3z0Q`6_a9mx7z=4KX*I^XE68PjpSp+(u};YY(1AIbZ8X&U?=rooa^5ifP_9L3*ar-7+pK8O!&x^OF|we*n>Zi zh29TCNN@Z@a@xaa#d12s3IEyqEFx>$2 zm!nQ=-yE91WipU4BhMGimkRs#K9=0+^vJInE${QP`9oGb|30z{@(3`v`?l~v;#u6W zR$OC5;F2pip(e*#$ObbM;>oPZL+TX>0C=&824IHq<4=k2kG6TEV`6)@it)acw!Xa0 z%OrKQ<6{VqGcXpzQlgG`x8t#cO$~}C6K|c9%03-syC37-rV!z;@M(R6*0E6i#OUS= z`A=Q?*{ro7rMWA`W>BgseQ%x$py0j`-%!74-4828NDmjj2?3ECT0Q&UzV{(f;p=p) z2=g}nNmUM{=t|;uI8BE>_pE7Nx3X9J@JpCuWJ$#yC+ACq5*+3ES``C3T85ehtOg88 z406@2Y*Eaeu4>%kY*A?@-oh4K>|IZdZ~;B>?=Y=+Wled1e71Ar+eCV3wVHQ-?wbG2 z`VE2%HwNj>n})zM!oEqws#i`MZ54I7s1}b7^!D8j>F(MBh<@8|yyKAYz+nk5%zT>y ztZe7?EfZFl8%x9N)0nbIG*AIQ%fG>#xIEP}s$`!v;@mL#YKdOmS~7n>#g34Q=K7a% zs$lp+)mg&U!O`0s!Joq@&y;6tA~@QjgI;N&IlbSi<6C#X7-ADSYw`j=2m(+XJT0pr zQ{kJ9B=d-Utbj->q$O;>W;GFhg%|?>1sCmtMNEKGo&u#n$@Hbj-l5>$JGj!(_;*aS zXLKuzv3>{A@-xg_VBIO5&Mt1MfWniT7!OzEjJMA&s_Jy)I~ZM0QznuKO|8&wsX%h% zo~m^y+lN?x-g0BQf?F#yxqh7!tD!P{uvfg0?YL}yI_-EH11dZ2GM--S!2sm5=GKcN zx}hdMSFr18vTs8!&X+I&od3iID+}%6?jD_QpfG+_|KYsPkmWbK3Nz*4nzEaFBS;*b zcWHU2^%ph^!NQA*IciO-k$3hzgio$PFEkK)hkVL`H$#26xMDylxe}4 z`itS6VMrULE4jBL_*hpm1?a*u74J{3Rm5oyy!-!9Ab&9|HGX%BTmo9vbLwI&X!d6rT`x znEVlwNZ4Ijv-4>l07>v8v)-#j`gdaec2lDIDG_nWKd-ML$4`vU5`#Pqk>h_6G6XH4 zxbKDd)H5@XJ7U8M5S~jq*G2k9LQcJ2V$Siy29t(x<~MCQT$LV148S)SiS~TD>cL`v zui?skkR^tT7BV<#h)5lZ%~vVkB;OqdIEi4UCQck^=`#MSmN%}Fx+3%$e95lzf9z0ypi%cddlJ=6M$qzdA z-d|tVE3+r_pUN*JR=sp_cSr9nmrg~#Ff zK2&mvK{?X(SQQW1U^irz-f>N7es0cu_%%KDn%xL^)`9epxXw|Yfs=v`#e?p7ThuNN z)!<$W3ZJ)D$+%kSIoZ7YR9Y^sYF>;Q9UY117a#exRJ57?KLYi?blm@!QEjE=8TYRY z^J$X=Ax4G$4kT{GBcI+K(^7d+Xk<=-36tc@1kXRCqd9nsPRL4GS_ zB$%WJVfs*judf(d!~@E->N=aYH?IlmREYqRvP^E@!2<(fM>SrHutOc&gDtg#gLOe6 z18INSawnHduZAU2R$72mSBubz1UH0}w#(w7W{ITWvCdubmx-b<8R0-)4kAgJy+ZB> zfiJ>tnnFl-){0~{XaqoFbPN2ekKCjWo^!RLAwC@!dXo-g2$`0YfT5}(HWMuLYvR4% zqopQ)Hhg7##8KHJZjDN&oeyl$BivpQ$<@q*$thX|4HZK;hI|ClqM9TeYR$7+fILYglps{H%Jrjr1<)kjD zW$La?GmR}C`*?Vq5`tzqd5gicu9a;pOo!u}EU~a%mF4b106xFpsaa%3O`R{b{&)HB zB$@wZ`Nw!d)84zVYq3hFRo4s{8*1}{M6EHaDR2o3TPv5hi zic|)CD-GnxC1#Su!IPxrcT;n}o6d!cC=n*&cw@T0m-pY1IicBGXrTPrY@CgkH*fYe z5PFVU#N7SCgHKecFx?Kz&x@F*?`FBCnX>rgrV;ah!Y<_$k=)X@cYHy5aD0zt0zAsNfo#<&TM0BbAcXMwTQ+_M+kZ zt(2Wl*W$wyhig|#YGj{7F0nG>LRN1BeJX3)Sm4GT5RdQ?$Gp46>{OKHZBJ55N-@%1 z>8*2lnok)}$gsBIm%buGLY25FqEa&Xpw)0x#;@_B;Q$(7Ozqpx?za0(qTdg%e}r)I zaP;a5HhAhQC@F~IT=ZvMNfbKCizn&32Go*!VXs#-6i>y?$9__8S(?7Loer%0v+l=m zkVAXmyYJw@)9AWrU^Q1VsT0gPN9Z~ur>g)6riIO#VPtZp@0_-jar%)Bz^z9eIkbbc zFt3bb^^IEL-bR3OCZ|%w1T#BwtuZtu7NVAG2SQ9$dbVaTBsl{#hry>1G|2>WbZ-T}<9D@XZ11ThJI*iKKDCobPh zQ-#aP4h4qC_TU17O%~?}y%bc%7+qP)GHcuw+wa#R>m3O*@jN1w@_`4n7q_g)^CxsL zCG+IgkQFy+r-&q$GpZ@s1}vZ067d43!j#%=(9{RY5;8li4-{cMeH7iPx~q%tgYYA561tNyR#*$e$RrNmmZ}V7iSr z^IktrsMiVsnx#-3{>{dZ74oxe)oZLb#px0(g<@xvjGp~?onIXcGi5mJSUeL7+7G1; zM2?4bElS33Pz$&~x$FXk#;rvnsolDqO{y7*iupu4Sw=d*j?bOuk-(W>Vj_LQgK)K< zYr0Z5PK~)WIg@|VB0-6DS_rqRFvpGCW@2{<3_f_=d3$!dt-*2RHbkW4x4q=n@@FUP z?d2L`EGw&fd@y52JS|7|?*2pWC6_Vi8qWQIeE=jp*^WUnQ#GvemOA)bB658}KA4|0 z{L+-Z=$A?5vTdq5-NnrrdFyCJCAW0?@x7h7zv~}E($gsmaR@ zXWn8@tdLF_688V6ErBh3gj<6`j-~Ibz?w{s+6y&7Y=a-zunKKUG-|K_0 z9!s&FHh$fsqCdk{L%*V-zsL&|%Epjh9yX?B(hxS&DQ9a+Z-q{$l6)sq=GT~WF(Y;m zO03OG|BDq|*Mo#a+aP5+a(uuuC9It|E0Ji#%9jD3CPM^c>mEdf_jz~Qknm}JzyS%^ z$Axz6B=SyVxCvHKewt%@*y9oAE6$3bf-^78f$RqY!;}gb6EP z1rT~Hh9I9h6@Edqo;y6a&oRJY*{5GC=Uhij37zkr{|Yq~b($WA8!K-Z zr$2{IDiKFre%kMIb1CwEc%?d{kR`jy{H~lrxrY`jLy+D3NwZqkNErhm?{e%t8?1a({RKx&PZJWR^%N;CVn`f-Jo^A>cM%w$fz$iaB=}wP-u*TJzX8 z_OY|~Ik33z0WL{8psM+T+N4{*(ci;WB&02gs}JYGvX59JFvhx+3)_l!xe;6NDWb=|V}pk^uw`=p>bDs>6-=&%C~jl#@8sO^NlZYQ4@JUP z2<7a4wFA+bv9P`ze2W0zX5VGKRNUFLl2uMb6|q@eYFe@ z_Vu^HyS-46`y1JvV|x$6h(vgjslT-$m%8sILTh79Zg=7_9S11oxp<) zalIze3&z@r^ne9QIXZ1>Da1C%9v^_^q%5g@>fT>zqI2kOv*dMPwNMUOBHX;^I~`y5y^r zl*Hp41Z2qLD>2h9g7Eu3ultU97kYn}m^QR(GO_>_`vA*f?!H4R!Bn%|d zn@XU|HJ8gqsu&_`?d2~&`+9g+R(+=7U9l>LNAXMqkyGZ;o_(WGzgRj{mK9YpVbhs~n9dsmXdE{Eo#7mf4xk48z<@Xw<} znlWWR=ap;nD}zuzNv|8yf}IE)i+Vuu72!;5=x5ou95OpdBPIMN zqGS5ECwr@xVUDuNY+uCmB;rMNnA;e_=xYSWr9WNNc zVs3h5oXtPp;Cyv>!3XLW__voJen(yH`C81HV&&UpyK}@T_&j=Bs%LpEfW^_H--+?- zEqb9g3=(v9Yg3n*IL6inu1Jho$I0X4_=ipaOhDddF?mWA69M)I-#XUuD=4cr94E znh|&?l!wQ4cfJ;Tw~(E3=hYG3VJP8MVHgPq&p(vzP}&Gb#~fAiw3hvN3*l>V_wYb~ zNqM+oo=R1M$Njj0e6Aj)Sn~W!%k>+I-d^|dX@4^K@CjGykx5;`@?;xmy3gUsM0N|O zqAtpDk?{Ut$R09hT852!4~VZxyu3M4CtIIT!v2z-PgU_q{OiOe>h&x`xW!FR>*L;< z-7!B}r~=GJQ^p43zPIHkQFs5BmZiV3;mw2XX?s%mC6)hwTGOR@#{;Qr{5lY{;z=ZD zsb1dLfpz5>&^s9&nOWd|c)ZLzP~NOi*+3RtOfl(=Gs_?Qcgm%Fk`BCWa>FH%JMt^S z1eti8F8$bW_=DU}!VccCb%(rA7bP-hjGALjY2CW7PQHaAgxSJ=m9}HVw%rm#7O_(B zTk&nDap2+al%3-a$`y_Wn{`BVxb0Bh?OcKkAUK|0y_RzL zqUXo6?250T$fF*V2P7MA&kr(0d(9#< zkT!fYrBDJBK)FuD6fL{WZ}gX;bm7Oiq2Amw@Iy2D+LjRUUXgJigtgVS*Da9MlW8*e z@*?LGC(GL~B4k9yW+6dh!yV7^P>2CrGWg^Ndkc{2G{(eVkKc1~+cd;*S>d3L-ePl6sxxNz zX@fZBAo}ELBf7$G-W56N#v&6IHs`iIIOQ2_@^MZ3YaD@XV1i zH;7pdVk}4lRLE(lvA%>%`=_) zPM%$~ikZ#aC=^$mylih~VVlY~D9^;x`-&i=^k&(5SbKbnKvFn86m*1k(oMOAU*G06 z_ZDHSsQcN=5eJroqmv%J;T>1!tW(!p7vHa;?&%^$mK!>a@lBM@ijOBtPwJ1A%uYRr z_#H@J1U(}KnGT)y?L5FCd@TABgcd1#W0^C9sV>5qN(#c*2#1-XtNb^2H0au7OuC`4-cEMu*3}9aq5W) z2r+2AE++RDDx%&^_2m!Mg-=JN$~I@M)Q4lx6JNQ)-yB8F&E|_ts0o)itTP*Y|`fadoxlmKB8DBmr`cztl@?Hpl+zS$~01oy^e$MU}7^=q2dabd=K% zxLbQj(^F^0>BJ46%!g{F|KPcd{M?Y8a8BZ98?RNoYg;9DXIHg)gJap1mBzEE#}|b# zAPf*S4JP?PMiDu4&X^|Iu?XO{=MR~Mo(U$SW>s#=fGZ3PQbvl5T0m`zbBERKxbt@w zVpLGH;l_N`H-Ru#5Vfefq97`MdV`rw?^8l-I z=-eY*`}HG>jQL?OT_26R>1;=g95=hmi5dsu%pdJ-8tCJrTOIH*xcg@*)XZ;?B{>9~ zcQUu2Yas{Ew}hPPTap(|1Bc8|R%e73K3LKUPKMi-D{DeKUog;^ol^wJHP@3k7iub& zoHg)hE_>viM~<+V@=vOl@|j|cA^OR^|JAkmgyrK**!v{beau__3+PEJIu3uC0(sZ+ z!$>1Jg=pT7ab*&k8=o!AbTOhe9mdm+6JuAATLH_G0f6x$!7p{qrK++K0B3N==0j@g zIzT$1=lMFJ?;BXtP&X*Eb;%XBZit_y!O9zXOA>*oKX<*K$dTfCsfehf_oKtmU6^xz z2#?WMMblppltLdd*Gs^^-RBHoN6%yk{#`1ew;wO3UKm_^2+7uy9OpeHsqkOBfbXt1 zMlz{a$MLR7NiqAe>NXyEv0Tm`Abs={tl#H&UNKR`0@lVy+xn-Rgk+PdL5-rL>*ytk z0O@u1WR(iBoquJ*;8^aeAo)$wX*c}PhXmEf2s8OC3qg*qwPEZbayknxMJj5; zEUEnZSTuemm~KBfbnz>;tvJ%$1=X#qbl7ed=!tFU?eet@SDSdkP)91o^V!hA${DW3 z_zZqKpH9VtZx)$LdZN;zt^^uwem*HFp{#^pGTJuD`VwPy~4<$yhyK z*qG5oJ&P$rbfn76#QX=Zj5uv_TBFtFX<;%D08(CvAi2wQHTjNblOWo0oRf0H1V^$|Tl=&?g}y(Z=qIO3J>f8H)7RdQ#b z4FuM=Ye-0bV=_;N% zy6-YKzRU!DomcdC_q&dy@AE?B!*RIjwPqit>a*thj+=MD=v*;nCuees(f@7!S^tTnf*{o1dr9-cHjRgee9{i^XVn#e*batH{L`Rjy2#` zcXPeVQIN0NmgMJQ#WeH9nAo}Ag(J5(qejf<(y2`1K347TEeO}TQ-rs7$F0(f&S9)x z0gWJ9&zd$sJsYd7CpP}Y8e>(|u(7RJWxR0X?LikozSxskrjm|+v`zD=G(V1?RLHFZ zdu60x^)%A^<5F=jUU(?8A&pRHmw z4`ymhE~Zf|=3?Z18{x=FnAqAxldjX&1C!)bD#OFrR0gtUPjuER$BrANr-yIOD5&D| z&D?C9`paJAD>}z-!1LDS%2FBgbTATc%sMFogC#aarVYUmhviN;Kd)ENA0X_A*EcIX zRhDSx7)FaEQ~tX=f`&LWUW*Wus&@6rnf1z@g+PG9*F8Z1grdfxJ%uZACCFTNS!Z)) z#KoT*0UDggDDfSwY9z-Ay=8jCImYSzz?h`=Rx5BBbYK{cK}0Av;Erm~T0=nmNLmOm z&Kt7#$*DKEn^pt0&jm>Fd*1}Eu{^w}KE<#P2ruw7U++fa5dVf}Q=#9WP6;pS1SXhL zMhGGHf@E@12UATtCsG41R}wKvRsRzB?>Qr|_eRB*o(aOf?EO5#TqfkL113I40IiRclHPgEa#8hEE0Z|m?zwaWA~`0H@0?<|0PNDbm5+k^Ni zTLL#M>8ej7EcsX4HO}Yg%EFr<7etF|jX7Au=d3o?s6P*!>vcyk?TacM0|Ue**k$62B3?pMY3{b7JILn z{PK{1dUEg3CU^&i{Y9;* zKE^p}r;1}jWCH!%BN3;)N|yK-ftvW%gr~gvDk$C^^(nt5+b335B{+0;`_C~pyx48Z zVvIST1D!UXk|IIqsV$18%J6bD`G@M}cV!eX%!<6A4nNDKmK81zQF`ja(Jt&014i{I zxf2DYzDn06sbLfOea~B0g_kyQU)MHPgd5X_s)a%a@+~}0-!Hjb-Uo^39fxG^gBDK}s;pIMA|j{F)OWGv{cCgyz&7^v=$+g zMZXP9GCpX+P5-0>U}CN}`Si;=Uhw;PDR^)47w%Atb28nwicVdZ#Y@=jE8Mo>feQTm zTcZPzJbKJUoPwr+S&sEdxFL2SfYC`w=pYh+TUZ|PCD0npxhb0SlIsmx4~VlY{9!-a zq~yDm^-TIyRv@Z7YTsM%U|(!s@I1(w!x0aLaptYF(ltzT$p(Fqo>--%G+=9 zd0Lv=C`IvAP5LQofVdYwMifLrUd=I6F|%F5uXWkBm2+Chh`iu+_F@S_g;)ch`Y)^eE-V$$wVOz9Jq z6P;OC?pU8Tx9o}yfNaLmn;!AM#%+trbl^KR_?d@!x6(Z%O1?TIJ!6vD>m&a|uVFjs z|KTopbJ;<|AG~aziQg`svgXoZDog)WvlVU`5Mv;*n*FMSRYz#&?R@#meUSsDD1Px3 zQVfVUFq@C$ETTcFS0ov$#V>IwM&k+e+?(s%_9Z3Dnp;?3vKR_V zVm_>)2iMPwxSI z5t!O6nXN4@p*MhcELt8oz@_By!&H3hi-X0a9OSFX;jF>Yns49yMfwa7yuGY1rNx6{ zj)s%g1=TtUp~-oikKJFIQ?Mxd5589Q1vH?JMgD@joR1D$Plt7YPBAY8%|{W6Qyq23 zIEp}O-1v70kGQdRL0jlsJU*O3I zbyo&-D>mw;{5YFXpn?5oo!^)tS{O$H9NrK9y@sA_k|;b7T6GBA@2aUNWwXvSPGqU4 zZG8oN16Q)1w~9jK*$VC$ba?~{p%%OmM%ydtDgAZwFKt@}!dXezRSA&a-~V*DyIOxo zqzZta3zmv8UJ3;ff&C~2*OJn_3+bq=Fa_b{LSPliaH0E5(6InM{O%-~{`_!6>INwm zXo573&S!(B#jjp8Nw|*#26Z!EYY$<4+_9u-=p5S64B@Yh$^-uYP%%}sT0 z#e&3y{sp3bvi)Wg7tu|s*b^eREF47*9}Yi|ZeepV3j$Oa#i<^q%hJUNxB#8`0iJ!5 zNUQ!0!2!-k4EiAM;3^*@C9^#05k^hZn&o&nUxr2p2vQ_?QCeNm5K& z1g#x<+~=P`LB<-0hZ3uv)Z-EEUlkGk+p4%2CMHxF)Wl?PTetr&WQ+C3htX-=4E%k| zoe?gAxbY1$E|K?3gUm5B&81jJEpkKX^3+akUB#?-&@AV^~1FI zqhrh8Qlk_M)Ht*7xqag_Nlqk5(@#2&$gAR{d9A*bt-E`6Ft#`J6SA?w4f(zP^3l&; ze;X&=h2Tg>dpu4o1bM{6iGf7vXfeNH0ZHRCC@$^f)G^~~nV+{DCM$RPvt2C|4LP+< zKIo_9t|b+jeu}wOFua~wx4Zc-ZHlqDp1?SKZlt(hZ+QG z@Z)1Uz&NViQlL?70H)xw8Q}9$D8&hU2h>4gA8X&X(r&is`Y&a_JGJ$!pLHrs9qc@Z zkBHYlW&FF6H$b;#w>?)XR4Af(Hu;P!*!C^b)+7#1xGJb=bivHi70fU4hRNEN`OyH# z_2tOk?7Q7Lew{mTRTZp56og?Rosne(cjUqK$N3oAsNubjogw0#3au%gY{wVNU}SR^95Fr&e|#@nKJNAk|aLXs@Y|UsoP4i92imsBUcz@Axoi zv`S3g4^%NxZqwj}?OqKT-vYJ0*x_@G2@$C~jflyIA-@`%p~%oL;K}{O$jEK1uaBDm z0oGORyR@wvQH}Wx_Wem(Wx9JML+qSZU79`E+OEzHy0w}$Lnr}+;?(u?2ax2%@b9nTbDYatYcv2a!88;Wjq2@=I z;-W~Ua}vh83Jc$6Ye*53R@;pWR8H*VPpmn4y^1KsZ4;PeSWYl1XaMnP;DkEgviAN_ zoK80T3F^aFryKv*YY!qib%x?J7{~VuYq7V^Iit=FVU71K zM-T13Dq;`V?);)AAmJnmc<%DIIv7m+)JI=mJ6LIsC0wh*#!8dFJOW%s1SkMaylGe+ z`}MC-ZAncrN`ANb#w%dXk&;#nUJqA*6FyU`MQj-}H%8>}zmw>w&G{)`n9Ie(H*H2x zYyTc_r#rU*NeGjJ_Z4GB2*lwq@3k4+=5ovO=QMcdIHX%aCl7E6hKrw|gEF;04;}7R zKKi#di`kLfd4mvTLsx83>>bY}m4KSo;QuU>N z1@WMV^1+m^RwhJ*$CVXp`;qhm!Il&q8#O;D{j+h>N4$mG8*@M9@y>m>DGwxjGU*vMev7>WuvdTYce(mq@JcY+UQ1 zZ5usf|T&yp6}h5j8A7=G;I2Ck`=1;e6F=1n;S{GO}S6}v)^Ap{RXAvw|R zBYAbCN9Q*MKWo!716gDQle50gY*r;JtonPi3UVeS>N8ao0OkYT9nAOpOg@y}5H!H}GIOgNK5nH^eFTfBd#OoewFw zNWZI$WUxp94CRmQO|y9^Lxy$hP-hix@AE`w=njdlL%(NtlTHhd5}S&~V<3n+204F$ zQ7oj}JLWO6vuN&=0-mCc*xai#ySN*8Db5V+s!&_(&B_DaNys*ou+eR_Ln`zjqf7?Y z!(j3~@qXybEJfCq;hpf0>+ILJ_gJNYE%9r|7c(+|KUoN$$DExCL=&@u@h?OC;K|j|6b+7~O;DYp<`ea&M+<5e=&!oB7j}QJ@ zl7N~I8=I|7a>Pn=&Ph-1>Er|APDYL_-EDWM4sxH{XwmW!-8Zk=)D3;KiBn!B^ABJ5 ze8$>eh4Zg=UbX0(^z?nk4$*qF#o4`D;*5&eF>HktlT*ZX(tHm|Dx-ygW8V1cHbmea4W7Y?k+`x7b&iV zK#=0rLU4Ddlw!r9xCM$+Ae7?nTC_-Uiu>hv-t*pjX3l+Q%_KA5zxV9Dd75w;a&0Qy@zg2QU9g`e|GuKkN#bDBe0guL@cr;m7;&?#dls=kea{E6DNMsp1s zS@}uJ5)>V3ny)p%4Y6yecESMLiHMK$`sSQYm*(tP><679VvG=U zgANVk?K8v}>fh4QcEU~DkBW#{$Ac`noxBqWzBR39JAmXp2sW&~e-`sGDQ~g^W}$S$ zu_8`Pqw!(+8)a?2$yff=Xs(U`q@bCtTx95!aNHVc`r$^XF~ji9dqnqc$%mj3F&{}L zoWJZ}%F>jByU{&7!MkFZ^X|b=XDp*yNBKa8HxZUKPIb%l@w%p{e+kw~6lw$VbK0Xw zB6U32UHuHnMfCHz9ncI2wM0u5fl&iK*m&e`;g${6hR(aXaWCt#Qt`IG>F1t8Qqm7X zNVBd!RDb}e1tB0-{jhE=Y-cni_gMZqlLw`%qY1m}`@hEQ;eb4D7#a6kMYzwNIp!pM zby%OzgW}*_Tr~wS-3nzwfdV*1wp$hNyO0ctJfuaW!`4R^f? zv;0H0ZASXX#D-+qC$HE*l_XclcenQ*#yv-91mra%{r!J^@nKzrw0t4I{MP5Y&zh4< zPp4U8xo7$MUYrr;g$4F

    +d|8o|t?*!1vsD<)kNCkTJ+;6guY$+(XL6!NRyx|w9A z{Qen0wuX6FO#$JdO4$9@WzlnLE>+%+as7$DDmnuK^(g;fG>+9LkClkApXWk?6ZLbcf)Xi z6#wGd5J4rGl_Yit8=gZZzFb=!xfv>+w&xvT-LaLqG`;vc)+PfKRE-j#N&V_C@3qU7 z^{zQ7FctCBi;Cr!IZc;tJASiZeWp;0niFp|P^o)*SgQt7Q7PJzs;})2FXqk`fZUC-0iWdcx+u^<$%Z(~^1w>yPNw0vtTKV*BJSF_3uZz*_yB!>->1-!1_V zZipYgmUKH4BiIg+6*I%5|4f)jm~Gc{o#*%qdbg$Mq^-}PG~;qzPr-(NLhjAG@{yVlI~zEuygNtQuS zf}Q2ejcLdm`O$SGj&tct0w#Yo*I1a?2e*LfQRnr+pLu`f{~jAxC4Mjz3bC)nx2!*V zX@(j+pqAg>A56m;uI62su%U<|bKm$qa$`ZQxF|{F9!;@`CZT=7&aTVa)W=z;oD7r~ zH>!w!wG|FvO+huJ$4|Xtu2$6&*5dv33pbNzCv)=nZ-MJ<6sIkhdqJN`#|_|FEku@F0CD(5C-8>=@RQ zE`n+J5)o^$9-uRAnW*WH`HELEAJ@Nq%%33amOm(m>BY4hppDxU?EY=&2)^{wuL)G# zj*f(Pz_7{-11D(eesy4f{bMsBE*zq~jr{WEK7%gp+|&OJ%boEQK2jdNKQAxy8CblK z^%N(;&hWU|7TqJbpR3MdNrob3|5C{W9nRX-D|6EI#az?In3Uk8-hf)1y@XlgdSFCf z$JE3RGd}+1{DNG7nCB}%VBhFHS9I5du@n9)bKud**`i*oPo@0KrOL}L(a z_|VRC?^`C+nw!><;Y)0e5;a0=FqW6S)*`BB|WPQ(=qG3)l()s=zts^a5Hy`4s zAFj{_T+#J&yoE*at}cs$k9d;&NpJ-s|F^#S|6L&XKQ;=Qa?mGl2$d5Zh5A2LRiL1} z``_=Zx%)FqT)yIW7+(ig3Bp`xd(e*joMPl{tcC0#HL}11y57_!j5`}t`xll2p?@ut z_vZD-`a_;P7i4gQSJ|sHQ*MOqtcQ{;G6l zpkTK~0mtV&e%PSkxt6Pl*Bh3^_b^81(^~rF#wICL6`HYBj_zV}q<%82k?GW+2_V8b zra@thV)!NL)LbcWzLIC=-^KmxaeSl{HVKpX_XA(>OyFuY+CocX#r@LqKm_?&$I$E_ zHse}YjudGS<=1q}W?1lA9&46x$M^yOVp_v%@sgsg-mtmE_S+u62k=019-JjdIBe92 zF*<%zHn#mvM!55aDJ5+0*LnfQ*kAd0+g@(5k=AWU=jY6&mjoq7V59>o7&9v$wMlO+ zeYU``8K)#&{kVwDvSTqf z2gvp%Dae!Yb3pQ6Z{L#as>r=O_MbA2!|MT&@xxeI>EpCUq4O|6Vw@Yajdh>iq8>$w zn~W_3`(^9Di`Q$i_;rR%8a5;+Wg4juEEn!c1L-YA{O;&uRKN4<5E`W%7hY3$4&Xkz zo@u7Oz(GU99%@{dF4D8fiEODQVnT{=wO(%1io;+Lqismo73@cpSFX@;x|6y!6diK* zj+wq?raPqPbPid7r9++Ugk3n_yKJQriyi9al0!>E{53Z?NRu;$ughFB43qQX`{jYF zA9L4YgC2sl+gLYf$Op;IxfkN>B>MBM9m7}n^Ktwy3E_Xt1Y&!qWy=IXLg|3A?zzcF zE|vV>bUv%Te=wvjweJ1pJN8kIQJO1u#Pb3tmOc;m+)P_Wa1KRM8~&7H5y0>A_q14D zO~qR-n&doSMnxWwNqB8fW4`|2xPsw)@G;}$oXZyW(NB%Gc_%9l=}a`ZBA91*m&Y@T zwyOT~E?TkamA>{fTpX{L+e`D-!uR?yk8!i>QD!EP3JM)Y~uz&j2N5J1WwN{^8>Haebyr?l)|zYR1)>oLr``3%qt2xMxUoX>p|T=o3A_H5!X36~F<>3xcS0s9lTTh1T}PL${4Ex+^%5ONJVlUc&K z-^_LOj8L@rx9b@<=&yn5=8wB7$i!J6>22J*W4$$v(kpwI^G}>+a^-GD17-TcERAte z9VE}oipIhmo$-TOk~nA&tZ2XNdKgoWwtzq*dh@LaFVjwW%IdGYi)tHxhE4tWG$jk7 zLWu{@Jte4p+W#MnD)d9#^C~2#5~m^RjU)R!0d}9?%C7871V`5}!L6r0pmohmIvO8G z`Lcf^poCF~pS{~NS7*%O;hkPNSyM;=_;YIj8)AZ3H~H>q!580k;PbvjW$T^CQQBE% zmZyh}0Ajx96%m095h`SyRN1IlKDNM4<<>DE)h?r$F;lol!@4d6H>TiC0KeT=3lJm- z2I;k^otL0u*~v(AJ6L{vYHf1P?x?v7Us(8UIXh%OsOi6u44X3E6RQb(4Xi}9gbNDYE;Ipmq|~0>xbSD z%Q^G!Z2-EZB7=TL_x*5T4JnSkfBXkA602(SkE$i9Gjvb4?=Q{`Y<^bmBQ~L(uYd}gM8}&BVwHsT4q`)u@8E%!JjS{1RP*kzx^E3edQCW1>8p!o*tm7_-Mih;2S%>YzSexlCWRtwF7q02zp5pB8AGL#} z(!!48Z}(9vUWQ*DGp-z?1#?#97o0v*W_RPRH_d+>9j}&%@joug2~S0bD+{$t;M4sI zxDL4YY(R(qKHVIo@_CVW@nulzI;boJkG)dBF2(m6DqPh=5v&`#pWH>#IK%!6!w8Df zypM1EWm$p^{mJzo5?YEEP$%^x^ka7>8Mt~jzwd(tM@M^E!=Eowd#b3gi{qGSAf3G| zs5xZF>fmeBUutoO1(dh%o${#btL_p`-i78N!Ufn{JU!EA-9uWR7i~-&2BWwkD+lc$ z$4DDv=^G2iw7X^~?tD_qZzAs8si6NMkV}Rr?cH+rFp*bt*Ntoyn2u|~@Z|sx8+8;H zK3v;Rczqw}`N&Z=qtuud?9Yo`MR_)Uu4RTyh%VLoqrcR7^8Cjayj#I0FH}vj^9FZ~ zP4w^g(~3PL&8%HZ)UFko={x|Cw z))qgo*)qsP=I6De#t8CW5yPAXWR(DY10KInys8|$UI3ctMS26?(^5&LI*0ZA1+Vyz zD_(y?i5|U4cg}?W>Q)89(nf69#}5gqVg%`N@?ME0$GCpxUu0#%Ypt%wC`r38{>rY4 zL+Ft_w~grFy7gYS-LNLYYD<7I=|;jl0j-!UoujVb12yS@uOgyjN2^4$O5L+Q*E*zB z8jF??RO*?Lw9fy@q2#nK$SS zZ3SddF%HQfByCu+&ju=b_EX|e-4Jgx^nH(m(vCIsmSUn;KI}t``Z1&MD@P^pxQe-@ ztdEEXRYQ&kSA}`YftG0haf<@d`W#B__%?qn|NWOkUT7Aq+#ZqjX0qfZfUQ8#nfo7n zptECq-C9_KH{3v_E;)Z_I2a7Z^!eA^)hkGL_6Kf$K$KvvDBBQHTK>EMMVZ-5-O=RA zEo0tFJ!#*ym?yp20uOW3VarkG_s}AYlUILvMj=HRm#@A__e`yl7Aet6S4vdQab#{b zC?-L~`t(?na+O*VW;QkBRe^B65*%2mRlJ7HSyOtI(S9avSJQP6X zwJWChV40fSc{p@U)4^@{MSgb9UF9;1^}5*QNlWPCPyS%t@l*q>XL@>}bfwSL!IEpI znBxn0eLN&3Wd3?A4vN^sE08z@Tfr8i$BXGh`Kr2U-@wqqVV`$8q_8iu77!{cUy-jC zJL++H5lmlkKGbW})hE?f5aI?{@?-Q#0iSwGlf)oVxYhpY>Q0 zA!>@?Ylgg%l;5VvVed$|&ekZr&9vo6ie-0VGi*x?airrd}o;Wbo~m<|yjM zOT1dT+<3>{#hKURC6~+aVJqB3p%t4vGdP`CrBmQD)jFlRrUDC+tl0nVVv_g__OI5C zt-e;rPnus_G}vIyQ2 z`@E-iIGD;LnLWLE@_PF)%9Eq>bd;N|^>k~aUS@Wy#8VM?O3YJgc0WKiQ}cK}v^V#7 z%CWb4KU}eg-RTR3O5wU1{qgX-?30o%dt5U!a|R97zC8~d4oz|1^TAd^tG0m+(H$LY zI@AImz#l4q(`u9nSMn6SN>90%M`sNMaK^pFR|-ulJd_Of)kSghQf8X*G1sJ7&6N2h zU1Zel-g7M%(L;Ar_ON&Sc%3u;Vl*${{j@>%+r8^RG`BsCj)pvN0)K&wMX2({pa3>A zW=C2&WiVH=qhSu<6VbfTl=WytxY!aY~zR-QhYnHlaH?W@K4yL2XOyAu_YOMUd;y@H$Ob#HGSB~;c}?3)oZTi zTf%u8c~INAkovo*VYv#?E+uMC{_LA_LzmfK?&XH8Z5!70@q(YEl0+ zsKPq6_*0rsb-8M?f&1%43kixf;5~&#OjM6neKrfQR`>Rp%``2=tEksB!N@ zMm_q0%AXe-f@}X{AS7{-k2G~Q#xJ|Is&XKx%^sAM9wftSS{X?I zMN8Ok0do1o=3YAbcwC8#-Bf^&NnZ|jRfof6iYhkkpFUt8Y@nZD*Kk*)WWZDGuknnl zY|a%>Ce&8Jc1TuZAbB5nS!A$+BxPCNEPTBKjN%81n0KEM)1m;5?EfnSfBn~Vmb85v zQy*c`U9h)=K5n0huW5bH?5YXr_>4|@G^ z31sOb2e|65`~vwUK9Ys16ce|h3VT7b7^k*_2_wtqb>DR|D2?PdERFlc?%JrHD?mp` zLlOs#jN1e_A)5;XGVqXABi!6^Nb<3T8zOx`i6ZWrG6Skd5hcK5-8;rrj8ZA^m8s6n zG?uVlde&wCU;S2ez~(Yr_M4LAXU7_uL=vC&YaaR7`F|pAy%*o@+%Vj~2FYdN)uRPT z%uQT=;sg83uw3aontxru)^ZZMMTwuB5DOI>VJ2uy866y5IRs|^`EJ)9HG<>rlTwxj z*`2cdrtlkssK=Hcwa-g4B0ys8u#u~B6`}uOSf#A6j-)dGQD7+^bGUniD+foM9b0iU zL*%;%25fG0*2I~BO=3KLgt+ObjZQ?2d_sdmI6eLRBx=Woc2AJKmHeD(N2D)mOd|XP z>!4;J8Va+8LbE;&^{iMObAhu6FmHGS!RSS!e++jd+F&j*W;!_XL0a^=4ZP@+i zRgbM5h%opS>ICS?CnCm8ce0`}^*qZOr(*bt5!4Z78(}M1fcr9;j#%rN%`6HD)W>(M zK-J@S;F!=#L37I==pONYG230b{6#283KJys(|49(!eT_o)1A`we}$Ku(5K$*0zr{4 zp=^If&=uNSzBa=>pVoX=t8VXnF2t{9XMFY@p34*&+YbQ)uKzjQCyg@@kaU6Dyyk6o z@4Uje{R?CYhOuHX+R$z0?;DDcmJ{01jlc+x&LxPg%iXNb81m85sr>E8&$iG`ARD%7 z>|T$7pK(g=47Ca9gmU-p+C=Ao1B2tOVOCFuXP9pI5;zqmTEl` zmfOE>h*D^fd^q-E<s$na zrED~c1?cy$jR(eaDt*oi<@Qc~XO|PHmR)$^RYHv}HiUM!jV*|8m`NtMb;IM`D3F~t zQsULLPd~p{yYnfK0>W~oTQx3@yA@9%U=>bG^k&LCy2eq7Ajn6P%Je+_;ykuzY*(e^ zpxo1^T8`T-EewxW1cP5!zSXL;%7Q{j8X+A4wud!VQ)?gk)P}-t=ZLUjlicKjQ zS2ER5e`_iESv#CorMtu53}p$RUocYcWk(XHiHqBKo4ndwkb`1400w?)CufOQpo)tJ z#(ee%NWM6slCSf~@6Zlw7_vF2yX2_Lf&xYV7QkBnEwuW%=C;&GY>N@uTje(Vb&E^D zEUv|UPEP!Z!OEBQR?>hF83@PXg5W0}1R3$Xf@^*f(KcbQew7u?nJbaoI~5Yk`^9h} z_&>fOZl)R+xH{F@FgenkHIM|>q#SfW3TWIk`GO>y?f!5rXt*12mBv#Vc&4FV#`CbA zj{f1T%xdn#VTDf>-+|A>-RJn@#hXh&B;zHS`{mJO4=`{T#=Q-$ngA>dL?^x=axrOUZQitU26j^)@`xt}b018g%hR6> zX`PxCQJP|>9Y%DBdjU`3=eFh!Y^alVx|k9Bje}&c4OuVuP0OhtCLmu9=rd*6aPr%L zJ5X4ySf^h2O8feC1S}(ZrES7yWVMJ0rhpy3#@)5pQ|u$jwau}(c35}3kGN-M`4E8L z$pbp_Sr31!(yV$Y2s0?~cx}@9KNPk9DRv>r=zF}T;f{=5Bz2{XEA;e=I+n{ACBeo; zz{C|$TdkzJs}e=8glJ#=xU?b)r!*A~5l+Bg6BW3-QRHZq{itz!4iSWHC~>(jo+Zli5qJ0`{=;QwI1+ZE}&a7bw-ja@Llr}7#eXGWnv&r09? zJN~bHMTv-5D)?H~P)k0f39rcm9`^SRvIuFXiJ!$TCcUV7LKqBH1|Cc% z)}r(Vi&4jIsn@0kZE>}W5(*Chls%%sxj#y2KK;}pWjKRI9`fh_DcZ`T$Q{DQ{s7B* z;wj=52W%U^cokIEwb+S@gIKWednps@C;!xa_5O>Qoi?heh$;Nj=9x7OcLqeItBFh2 zao2rBla=KU5t>cveajIi75k}eo4*$iAZo0@rU-|F+J1`pBgkncO^oAQ8RRS8vTggc zF^Dh@^$)mTcabeKQp@srDFO+dR7B`XJ*sJ5fF?wHJziNASw!c>AXx8=QEYer#5m$u z2#pXO8OwG^GaJ8@1ZPxiNpQ3wu3XJjdnylcix z52H$rb4yWoKJBo##~$pdW#f1uI3sU>H)XmRo#ne-5mimpI@ooSW1epH@+ErmZFIo~n!)Rt*lapGdQqR+8xP5Gus+ujRd#-u^=Io8D0P^`W&78ZVahWc3jYW*FC12swfZkRT8GPjJ!FU(h#nwrM&cD+W0 z0nrkZSd$}325MJJ*i9PV8p!!#O|~J#ul($SpqYf~L3n@m?%5q*`gOX#UDu*s26wOw zYSI_mFz`1r@8^w&k!LmP!`>3{Q_5Gv4h;b)ap{!G_0r;ePCu3`5;K^lN!C`xXHRvC z5+Re;E7?OexrNJ4DVP@Zi69v+O&~F@wME4Hyq`W*Q%?FN}!4vKnI+X2QS2 zHTup99Y2M@bgY8XDGi%EU;QvEfhD}Wwj`>QLd}pkHJAe28bS#m9@2*6{dkIbL_h(p z=t3v{Y*?qy!XiWA&6Z7_ZTpp_oc&y_lwQo1o))}(i1_#3xmzsd_^Y!%mUmr`T=Qvrm>(*U zLJ=r$>^5emB@#~so2q7@CKMuL#I{TF43uzJP?2Q@Q(FyDLRORGhrox1 z%7{M`-0Vgd_Tx%_6$R$wJY4KLUqyX1pWTb!JV9oo2{pGpv}VW!l}tR$;=d8Uzutm7 zO%;zcMLpV3hEIiBJJYtHGXd|vz;vaWY8mP}n*+$#Dd{r?)x_YEA8>(~Mdkv{{dsm) zO%8qM@AqusR<89rW8AEZQ%xN8s$-Tke)l}Z(=Jc3ZhIn+HxzJnL~d;kCq%nVL6fEN zNep0n$cT+NKl*)lKjg*c)edIA{1oZYHCHzZAR1?bgku7Ld8?Nhp5A5lmmm(e3K> z0Xuvz?1p9)K5b6qf1zyoE_VSlB^TyQokCWNWY?;eEilzyn$w(y)XJG@3h!w^+9;@N z#0@>KsZIQu7)Z>#hg=g(9D#z+uI7g`7P6Uvv%|Y|xJsG1e-7Fw1p32Fkr^Ok7n?Db z(qC5xpq)H&V*M*918x+D8y0+C(IsmWtrg zV>YAv{_>Y^+wk-}a?gq-vcU@il>fAh`X0*&@%b?-T z1xN12v9H(X{Yy|uFKM3>(HH6pVy+bZ(+#EZb&o1;{g}Nc{|)J~A`4&M;;Me>Seui8 z#{vAH76*-U?V&N8rM^2|N-6qJW9}aJF}Y-eu{&q%sS!^(ASiRNu|v>6rP!sOo>`bP zt8328w&~^AOnN}FKRTSibl3$qY>5gyqd=fNnxO;J*BG#)X11Rb$6_KPw9WOYn} z&$q*WmENE0M!iPhwmw_)q<}^GxTrL zXM;}La_m=l3(ua7H3;jW-Wjx^T~Z0ST#DeF7>8c&Z{yyU1oXo}W2in8Mn~qwks*&Y zO6|hyJ)|sAf0iM-{tQ;&p(t1Jwg=WD`@CNc$~fK1JIDWTYbY%*tCoBM_m=hhC1 z4O^lPJwuJs8+Jw6u%57JHkOwW8N3>L&h~2(;S4zwq)ZM3bGvx8%2waSwuC&|P{$xN zZz~n<)dEYcjfpEm>+o{=^51M11^Th5e0P!D z-%?O08>`TaUJ$TG_x9CNgl9evx_QRvu`V}RG7HN;$`Nr04BMCxe`Bom4)OU!Wm#&@ zKj6fcZoJTaDIPK~cB8uP)5t2kUnv3?bdP^4t^Qw4+AMTuY1|iB-WT)*#H$HZzg?uA z{B0UoF?y?U0z6NBONmQ>&k=kqi2Es+6YCwb0EJPlv(igT0}(?!XR_YvvcDOm_8T?x z)_l@i4tdCQGn40Ozy|R`F(WaYglYt{nZHZfQ2x7^f)V81bb}rdkb({2kyINc>-&nL zZ7`%EHqHbEhC7Ju;iAKnq4h7jYuhTBoO+-)JKTR5ECby)Mib-D5PALCY3Gl~jvyq) zLuwFfJ|CQV`^wP6nk8obS774Tm?so?$E~w}HUwvJ>|OiVfKSoFw@=AI5gvjV(#IzV!J zkmb%0PfaNcP@Nt7chbyJe|z>GaiC3;SU_c7sC6xWqY_5LPeVo|2U??qxU@NsB7(je zLlJDaPK_(`1tFjpoPjtXroF_V|z6>1i1smqYRpvkS8|nkeUG##W{oFm?veBCq z+5qPyyK!*n>u@A@eWy)-hlcG^E(QlocUJQW*Zi52McK=~%_P2hmS~jv-S$R5w?&ZP zg~#neyaV?G+Oz>#NaRIO?I-^Ubw;mt4~K;wCMKoY!MQU}sEat!6-K#(54e@8y?#sYk<;Txu@34O|!)M zWCoOn0!2Y%GGur@?A~hw-VxNWi{Bwu;ta!GPH=tr+EIsB+z{_%?u{;LqChun9zPD- zJJe1AoC9rp3+FhN#8 zG3KOgV44V|gRULvviYyl5B-)Gv??h!Ms;$rT59Rv`OgSiu7$$u`beUKc8lq>D+zqk z_|8!HY;5J7vulU$(UUq>L1mOc@Y0Vb;r8LBwurPs#zZa z|9klVZcP1Gd+PuE^%8f;!`jqCbCl5L4-XmiaCBGem-x?n=P?_UHd`}}$?Esul2CpJp6hf9S_)LU>hdUKGIMf)nR1N7 zRj_Y;oLBa;)7_xv=DL-ZIEpgk#Z>`u)Wn5**ffmcHMC#BON#QME=CWu$xSW)0ny;y zA_5Nbz^*4ynqIMzWjE()GfbLuI2beV{ z0cA6F@5lgbVhV*w?vPM&qscd*+u^2XMA|iaV(CC}Mu)^R5Oma#B#E9FP(tzrJ~+Gf z=AEVd5QOqXDW&FB!Rk*tTxUKiR<8`_M67d}={O)9wa57uDO4YW%b35$XW>fEjYf!7 zw!@#xMZm8&6T(`;?9`nEk9mV!{`M&-*%Z_!C_r#zNqt4NH_sS6+T^ZYkIA0o7{EEm z8f*1dwXUlQ^*xRo&0I<};5`ZTcdh)XtV|Xh@`$+$MO^ucluXxhS57l8 zZ=<&$dEX_h)Y){J9G?(p;Y%D?MT)8H8mT+;KChOj!$Blu7opxwShG;_NilLm3>Z$H zh&ntF)mi+^2p8-9O8C3M0wq$pbD?m>-RdjmZ$V(2tBa!E);14Ygl$}G=5l{8P&X+F zqGeWB(sJmQytN(G$tuT7E=Uw2-s3=d+r|Iut%Rs>Fte7kw<)JpU2xJFsxWI&?O(1J zjo3l*dmn*#Zr@Mi=2Xo~Hc*xtXAiP)_gz`kQU1Shi9v4_6Q6FQ^U=KU=!6H*b+W~L z6#385^-{H`$AqPknHLVDjH_5zE@{nI=Tvh8+A0rsM;vX+P*J9>Lsq?B+)v}a|DZL$ z+C9BZj9>Z6wCTagIMhGPQHnWAhdPg8V0qfaXD3%%jv|G&q0=0bk}Wgat5J>y+ZDfs*%`G}w3$0DsX`7de3SE^7+UeebWr#d(^IAWHhMEIKEsG7&UnZWH4x?O^PC0ad$koP(=c(yoP42NX4nrz3v?nF0gqW@M&I@U(nr?Br2&uZ8>@&k&%5ZT^#jjJE-MIwK9sw8aA5{R zWk#$HKUew)0n`Qof0UKi7pz=gxpASiovsw3HDt=n$8kv*rm4}o4R`C$qaJ`7aKhd-dX&G{~^l%Y2@fpIG`ZCJ#;4|)`g6o!4 zSL?5Rn;bLkohG2HmH!KL6W@z6v=QOnG1}RYHrhRX>AQW;ioi>{BfY0TPaVb&=QYb^ zk4t7^)dAb9IhujTI|-$c()@Y1s08s_ajsRqJDwL6oflIWGwZ}OQVwZ`4E?qq4?q0T zbw;VFYU|6H{qv%g87vL#BzFWn^6@tTL0a{=)Qtmzy$nToI}<7rkpO@4eFP8fYNi&+ zaFJ21I&qeD(6i4d6^PzJ)XJ~`cg4ex+Kt!E=`B}#bKc>X>14voLzyoO#8m!xZC3!V zh+0d(taCEXbt3~lm%_TZ505issWWN0#`>K%;S>!7l@$|yHCqDCiN14(52qP0Et+$J zy}|nQ%^!5YPubKu!3!yitSGKl-@yJkL~$nfr0&!J$~NO~z*I!}ZDGwwSsg(A>%WOY z>w{;+>;z9$*dr4DgJCDXkSM*~gVXC-;$d|&{t)Jn<&Y2I$%@5{%LV#1tzGB2s(s<@ zR3*p%!CVyc&x&d)q}=eEDk_=PrNnv0i^`<)NR39I$g}zo?tcBo3%kxEJ5!V{AeH4k zYQF!!4FT{Ua-CJ$)9JrS-`<&+>2G2a^BYfj(>zvCVQ&iK<~7a7ED#4bboyxa#2n+- z7`Q&!a7#woG|?}dSX3m7cBIAs&|t!U(5$GSn%s@dHu;J(wLYP#I%CwJjWY{o=xu`vbeD;?ob(|3qf$%&$TP^29>H4m``6;gp7L3D`Ew)e~A4Vg3^n(vDIc)k<}^ zq_<9}e4tw*SF!OY`-4NhZ+g>W!jEH%jYyT%`Q4JHIhSsMhsy4NuwioHJjP4(-Yw%9 z{U2mRTtROtJAmE3l6Fvu%oQR$oUroWpA-bV-#c0j;ZnR~{H19?v^fCIHsAxG?wS-9 z&hq{{73)G>T)>Z%l?`!+#nmVwhl@M(m3T4@pm&uhimMn2i!D73yyDqWp=|=?^bFUv zP99^RbO13xtqCyLbvF&To&f)*prs}`*BcLZV2rTo*FBWn2b5G@z`I*0{*?Ak-(I{I zy4WL@vP_Gbd;{H;#MdRB?J|C8BIoeRo+rBoiwZY&9a2niuKVjj*8ItFeSW&q^c6do zI7`9L zMHfncTu`CqLy8|K_*!Ciu27Lt>eaetXB+a7>(lXYA)1A8g0iEx4{xgY7wcYpr%)~a z>&BY1IWWie^lqKyZ!)e(pDpy(dj!5RHC3sDs*Vo`O0?j*QMf#MVr*$?LTyF1KJ@>u zS&yumg|LT@ECCqE2VUjjQ?zZYAHm$VE=7+Pgs`kL8OjS$)Lu6Qe`~T3-ozh@`h@1r z@iJDQ*m6!*PY;EPc!)^tWkzOWxl6@g_<)iChnG3K zbDil;-?y{$pT%%f5!v4A4N{y-W|y|j7P2(K%@nKeiQDb&&I*#3hu%U=NWo_|sr|Dn zpa1wgX5IW7Q+fn8EVCWpLeA{HXkQLqYkZ$q{Kk^}N3o7$L|U5CB{8|sH@p=s+d3^G zziF&_1^M^`&&5<4}z>Xvv)dUl2Rr5VUjjr zHL2sf`A=dVf!KH4gEM>)}jbql-{UXQqOyf*v;6X}zyt8yhN{!*YFDy}|~^F3n%Oo=wgOwrx77#fyw zU9iLAn6C|XbU-`UlMSk5xVr5qXNTCUrjv27I3Wa$FpN4Ckx1(>9QDn#TD-hf1GJMw zMk$IP!u_);zus*LF=b7X*8-9pjhey((J zzmzU+23vt3g=}%VMUrGIk z>*9=sq>S}6a!rG28WT3lZ!gBw$~X$*hp*m+J3_x(Nbu*;2SZ{=VM|ne)iDy%1{oi+= z$4Mwq9mk;`|e~1QLA#= z^33?ZAGiPhvT{jT$|6u6@sy8b-{|W4i-}?5HIglg`$IwS6UwH^@pzsSm&kE9&rtg< zE_r1OL+3j-i|%H)1n;eJwkahGmyqTjohG*DGKpygPoM5-ib9I<>k7OBD*P$Ek+}xt zewdrQMlRYy#`)Z4IkaXPX$)x8s+dP$Qn>t%c#Ji@V=n-~Cz!Q@SCnMRsaJoDE1 z=ZtAAPZ}BdJp`hxd!9Y{pcM-+oS&iLNs0gz8Z6AvVsn0+(~CD&+xl1hc5FeD=vT%V zIE1g>Bep8Jg_VXZ1;>9!`YE~M#JT1~K68C|z!03ALAWETTHjiKluac{2CM-!rz|1T zVp^5KDtG8hRbdrg-}pO?ny(>pzu#1XBcIXCK+80^v+JKJyOLm?pD*nOe@_!07V1d8 zNWc^|qoDXAJ7UJpu>8T%BGV8lTz%7H|N3<`3|Z1Xo<2k$l8Z` zq_(1FrRw&5?k>u^*LE>q)1c9z+bTzs4*h)QPZsXS@F_}sp8fFy8&+!0!$soUwO*nh zA!>B{w@H!Jx5EMA1|j8#!1qKn9&+dy3d9WABi=eTQpTxV?5XH9JrbE$1i?n53+d&M z$b6W@-Z8ft@~nTvNo@mNl!2N-V4m}5X@xdr)WKj~agJtf{fd%IqI&XlPh|7{QqmcPJA5b(R!;H4QJy2oWb@N z^Zf_E_>0rt`QTE8J+?G)k(__dH^*K-R>FtTqimF=5uCq46!Lnh{lMiIAk}N?z>$X7 zlQ(JvWyI+M2HE0`Prwur=vP*9`mV`|T~>^U*e(-6Vxa)r5+G+kxbd1lXS3nC0kC)5 zoBOu>H9#UR4ixq8x%$!SqjR*LHR*XE`+EgujJFx?s(wLo@O`#&MX}RpKL02nPqiIo zSb_$|pA3w&9kZ=_Qs8hn%@9t|M{TW4!YZx}^$V{}5#f{G4Sx zBYdN@?E^{?l?StL2vFMfEh^~fp&bOf8PVCk8$B;;coc`%3I!Du*9b`m(!!kvYU(NO*g(SjElB#UM`*Squ zOU9T$ZH(hR&DAGM-)W79pEP@YPk#wVCgB_AZ7Xgxr5!)MwURl~xjMByBPj*O8(Ni! zkkvUkzz>*@C+1i*m=#>r2!Mpy)RG-oH z*Lc_v`u&0YF?Dwo$JX7Zwy0Xqw}?2qM(!Xj%mOvgiR!o&?Hba2ddobUUpp8d<}vZC z;o;f5$(3lZsM}o&wEqy&=@@|16g57u%Kw(><{3H#!$$3HH%;|h3c94%?3rOIgKdIN+w=> zUrbOT@HhFn%=$5BJzLRwG#k9?Twp}`#U>T9*u7sW@?zKs<=~a^91sZjCRL-RFWCIa zfij%wuz!;IV~S$Y2cS=pl%0qMCYA7#zykVwiW=ig?F7TiHeC`(Rjf|hwEY13v*=}L zCdOy)r+|bLQAh^&h>5d`0G*eU+B4#js`g!{3i8pc&3@lSkT-C)(j~bPI%evV9Rwv8 zncKff{KVe{?>6C##E z7}b&4M9HTe)pu%?z7d02lP=F#x8K(r*7F|IEIlyzKD_h7g_Ji?Y#@%JkM!}{A?&> z(t5{p4~)=>zPU-i$ohZyddsM`!uHEMxVyX4BBi*ydx7E*2=4C19ZGO_EnW!PLRwsl zyA^jU?lzg3HSd4DGtbOAA99j!$+=eU`@XKde;fb3Oi@fl#0-`NVsUn_-)RDLiI@ zpuSB+rHtBM3FO}cJ~UdpuwIaN-@zml&SvX~>34FUagoKyJslrU$^(e|Z(nk*Bx5O_ z0Ab%omJv36ynFlixG_fAFWHJ+PA<_lb?6z4brkLDvUdC_UB;(`2|=kHs|?3!0CHXx zylW*YC1i%fbMX9__*M~vt*b&8M;ey}uYN*KwRNCpPgDFwh}+jMYUoki&YW4f?8p8e zK{7Hm2OVnZ9oaOTE@0dA77CwP6;8QbLTiHzo|hhv;Xi{O*@edvRF8RELS$p4XT<{` zLo1vdrO}++NUPl_%(b7$3fCX-gqhK*(IYHAgtci2UH?EVy%^rTp4uLB^p$>K29XS4 z2y~LMScLa*cO~@?iQPQQ7$B6yAwQ}0XsU_va(UCe-G&11|L`U04hID&NQ(W1`Re2^ zXL)# zJ{_tV*5~YaenznIk$Gk>>8!c4SH(iE<^DTa3pOX}k_k7cP(WHT8uPE^^be0;h+SP4 zg@xJ#AbcKf)X5%(b)L%XQt@Y*(jz#|lk*<|Y;;idDEXKeUXW=Lzi^{477a_~Z-q;U zrnyHmNNS;r!>lGwd+w5fY^2O7G#eSN_RE<5eE9$d z0=&`6@3|fMYw#d=PTh#`1!Y*Li zdxr&X0LixA$ z`F#QKv^bb|rzxTdZ}qvPGhm(rL!XwEMR;7$G;z(Q+eaLN(M=`LaA7xJ7dUUN7_!{b zR0B{nm_(yS0s|*my2V!kZ?tp?Zt9ZckuRS_group;YIz;;&`Qr`NPoSIX9-T0Igt1 zPEeZEk1BD>MA#oWab0TSSk)oF#l&nk$e{vGk++AdlS(6V*d@WP77mKD%eOIBmdxZr zRjbYRX`)8-dJ;v9l%>Q4?Vmjp*T;%2s;6F+Z-&xp->;7+#33Vf=>hkM1^KiB$}I)D zBfD?_hiSk=yuFd&(*R-!!?qsv8791wZQ)0m&STR;&9=dExS=VXO8DhRv1sli1qq?D zC^&ZqrjrGPqW!PmKzdQY)NL-5)kUFE3xua|LC)tW(m zGbC`v(h1K?cI`|IwlYbx8c$)YPr`67L$~`9d%&ljkntNUI4`129^nr7dj^CB)1um-#qY6ng^e0dqE^%(i+!?`%7iRlUdwnQ9d%&Ja zF4dt#a*vQjghM4B^Jf7GbW4aZ@A$Oka@$aQ98>$Ld(qN4}Qx{(8p>rMhNe_)Y1wRA;AJB{17bZj(9PYxMH zRkb=(v3il-*@H;jr%e!Ir^HwvCQHqt24HQ+agLhj6F;w$AyUlEW7hP>ginUYs9*hL zknSAy3lRYPa_Op{7#=CHh8Hz(6y$3Sb^W^M=ev-@kBm5lU5Mn*cMHM{DHeqMHWZJ||0X5donpoj$Izj>H#Q06U?!rMoO|ml91mbpd`3OFbZ^D977PKpl6U`9 zKUV7V}E<( z5hG~7z@5R}B@bU~!o!2jT4s?qmCAxh&jdLPzj8q9Vno;J;NUkr_p>+R@@{Z)b##y_ z9N64(=z93UUP3*6xH7dWpB*%;_SX2_4i76$UH?(4ieY=3x7O}EYsofYgrjihEzu*Z zZMv5X`B||*-A2T|CY?$y@)QW`3K62h?L8$W#Iq>^M7wNvGJtaKm4bozX+(jy$WjZ6 zJEULHZz!2j^812HImWC>J`ZfmfBY-Lz)v?tR~SUIX*gErU%Hw;)yI&_iI6u zy9_)JS@%^1Ag1DbzRw`$8;zD5FLKp4WFsBE%X|Gt)~rN2^yJ*SR2f zkQqF@-dSGvw(Q4uTcu*Z$75mtAm^xxH`e0S#`7Pl5~fYbcc4vIj;saxAyWOZW?!_} zZ~6UO*%x9V=6klfDdWN5A~GylkMq~&s&&L>ngF2Eg5Yfzg5xOrL+c*O`&*Sl zh0OVl*Bb5vq75e4Fa&w268(O4bbSJLx@bn8LQNe5AdP(0b$J?!8JY>Xw|5$Cueh8hhAB_#5Sf1*VPTGuEOJ z;^OZ0XCkyi)n7ZTgc8gU+P@yjAV6K?QqBDP-}r$TbV(U zZ*mBA9f^;qFGheoX$sR~2}!3|#o36RFe)Ge4_>dn1u6ppKbzz@A=ZIu!vi6)|z4Gpr{rM1=PWOF^Bar1<&ZWKr&W0ygw46EUxX}># zoE;4^$3d#-;mh|ML-L!CYF!Ow`EWSRIh@sAkSKn$AF31W{ZQ{yv_-;*F1citCWOQ~ zvc4+ff|ofYZUh1idYY|fMK`Lc^@scN&|5!IGa+Ht?S1*EJ`PJF2PLI;aU8{a@9Os4 zFz5IGWZGiIMV2-;{Eo1wqgR3Cn;OK2oSch3U!;CLF^Gb1FFQKMcOO<@4`*pHcm0`N zQo=!Lg}Cn1(|Lg~r{j7|)vw!sFDJ`R9Ac3aCla~q3=pNrO_xa1(V88l;PKCKW z1y$)S=Cdz6Hut%oXV;z0kKu8Nt=@Zi&hKoc8M5*}!Hu>wL~`X%ydmJa{x)qIo$^U9 zAaHOjmGSzP5b0NA+si{zh#~Pqgf8o7N#)&~OVFv$?DeH8)X_{>$~f+L>;csn${W}} z5$^N(7sfFUapm=EpeJ?JPUxo%vEA_S-mk9sG6k~rZ!gJ#z3tzNsAt@mlf6i80Y!(w z6B|CwQosJ~tNmc06B4#}KReKGN0;cIcxKO&BQ{>@@*pboSO4%l_@)s5M$`Ef5Xv8O zlE4~FVU?g-@nb3#s_=goviwxSI_cO}=ezFksjdpAvfpyeYHd^!GFUXfj|OkMvuqp!J3RaV=a^*yu+> zcod3;vO1^K?_{hXth;IkWqmJSNk9L)jwnMzl5bZ+$S_Ob4RjHl^^Qw$u9X3i)~elz zbRTaO7QnZnPGnXf$e&my!0ByPU=d>%679O?YN|w4*w)Qxb`YXZK*jqZtEnfqS5M!MrN&{g6pX{4CQt6bR_(pl&pg1VPNT4Q}+p zBb3?ChBM>(vemu_mh4lf!iIX6gqnIbqJW5xc(C_7kwst@MJjh-Sbili@N|0V1?CR0 zWWNdopkY6X;x)?z$GoCFDnOKrd~gj ze&pn(kE;BufW(>Osa=A$jash2cS=6bI zz5K5lcLuUI-|kj8y&JMDX)F~Mri#%Tq$2(1abuqcDmbcl2W@oFG{>ugopV?VfSPfp zdAbChK00pU`TtYA`M+=FLKmt-3A6r+EzY8L423t0z(Kvsr+QKDYJ57cs225)-;iB& zgj2^Waeh6SLHrGJb!q~*#}LlHi9zDLm9H4|GQK$Nvn*aPI%MLUG`40?b7o+o4sbXZ zRlH0nB89|F-|#seeXnE%m{2>4JOiD|A^^-|ov~x-w+dl#9XFYuk9yd=c?6bxs2^QB z+uweVbvL0y(8QlN;ci}WZX9UitJ3w0c!TxA<&-JN#*io<{GrJ5I1GsQS?dRA@K9L@ z<=7&TSqAs+GL>CL5wN^;@N?jz>f3usppV4#Oq$rAUvlTVYybj`hobtTGdl}92R+*Z z$7=5x97H%O@hPD$0G8Q(>h z5UFu4YUwkLznEp~z*`GkcK-S}1*Z zzV5Ov(bY7o0qOh#+6ik$K$#qV>tJ3(?2+)8<9GF2feFyuS&L531T%0#6WB`)>bhRq(?tk3VT}VE^WA_WnYCh^Wx!$qaEKIvrY|J^|z%hnk+-d~Cc7%Liy3oD% zi4TSZi6<6uU98+a&*Ytn04WsLSfI_NFgSao?uBMupN_N*U(6@+P;-Fe$SxcgbKXNC zYjn-z43k*ruSko(p05-1-CP@Ni{A96JvVUi{CeIX7e4V#&s2xt?NSe*PvSAV=WY{< z+;{CR{Qv5YF4iH-i(hNr)R{hfxEcERQvTTWcI$97gd+ZHwFE}P>2j(GMaHc4G7G4O zBhswnFthmCNoSl6bD|r{SWpdgk?Tj~Io5^N$wbe-p1o9ZcTWxBYs0H(1YM@IpY;*R z&NU^=9v_6W%8+-2PD>UFz#@{Y1cgXoLC|-@sn_pAQBBFdSir}^1*9D`s-vd{L23N4 zsdN)z1g+1})In#*)7tP?_A$!j>4JMoC7@9CspCrzc-m@XjMahSEA5G-(e_;-h6bjF zO=iMtrtlV+*$YLFfB`oT4a}=A^MybWC0BiG(YTQy_S=Rmbt87VAnRD|l45B&hp-A6 z76FGt2->OWhH+I(HVT@c4|;x2Rg4Wi2APz&=fP)uILHVH&%l&E6+Xrh_a#*)&L1b6 z`2ZD?9cz`Yw*0<8Bhs38CcBcSHnd)jw@6;Nw{5(~Z`ii(ZBdLc4ACE13+RoQ;ySnhH72is?J|1TV-ndO`hQE|A<8dxArMb7?`Anaw zyja<6v3@Ms^=E+xhXF>BvPj6w2>1kDAvGZZlWM4Pid_qV)_R`PyO`!(y%S~hhw7l2 z6zO>mQTLhWv=1dP`4C~O)#c}~4m3}BP;Lfo>BA6j0xwR2ZzWD{W#0gL+AeeW2PC9z za}Q#a$fah8)zTRE`BHxrgSeFDs8m?-C0Lq$fawf~r)Hf^PPx`pSa>dAg>)BZ20VOt zKR>kr3tZ>gYEEX={|Sa*@j0%QK*Z5bwyX!zmRu2EhJ|$d2hzM6xD@mz|Ka>8_qMtj zAVx^B$hAzfAl}_0BEZTk%(i)`68hbyDv*Ho7`Cokplicszt^-4_=&vWiTr=-qVAI` z!aJtP7Ez7OGo7l6)E_w+ZB2#-9$C;lP}tjoBoTgr=~EyG)RKVePXT8>HKDIm$ev+N zv7stM8w5mHT6B*)#BnJ6Dc1#g*45^7h>2-%)+{JeUOx|^IJwhL;qiw;+}>ydF7Vo_ zYv?C>Iw7<}^QqVcp;EcFsi305Q`BII&g;Hk!bl;>c*}zS{9R~X1;X1$y>MDf7_zgz zSkjU;a^PvF_6EZzU@rFU*qmoC8-o5y)59rV07Of>St0SHQMJPSoer&~YRhK`{W(~6 zLW@GTRBV+Boj#JQPY3e3{sa(=R>ZaX53V@O1}1L&Ta4BgBW z{i7RZt=2~|I|M(?XxNF(zH@PPB19yG@6X><{;{u+!9KWl5vw_KZJ!YsrifqLmm=AV!p3-5W8$mbysXDH$mH~;yU`Udkx#%6xf_|kxBo` zvwS1b`V<}`6OX$z9WYrm`8v1(qyh+B@{1uHCVh`vI0oq8qr&1a*G32Q0A040!Rq_F zG|Me&U3^EW_KYDBv^)p_J6WnRActmXa{egl0vvJ(kMw+9+i*&2Lp<7VYmGD7J!5=7 zEDy3VzbEAGYmZ-`uQfUc_b`9T^&2PNJ7ZH0ctl~p)mx&_A|FxOXRH#yP2=TBSkC<^ zoBc&@lcB(t`wYvI)5EZk&M*6=epJB|_^$6fsOWM)JbiB&&ZJ>wb2r59ksfcfss(d^ zt<}v>MUQb5sdC@=(a&hLFdW_^k7GEJi*`Ka{S#_YCN?RQop!SK<`=K?k(T(#Gy^F& z4J!B+5%N9XP?u~c_&hw2E-Bi=lV;DGdZN2SRW1Y=`I@$j&`J82+YsQd<&f2D*=`6Y zcu_UuU=Mf97P(RN0^oe=l+9(r3;897=HbXHdTy5ZHn!T&ayAj~#E z97r(S<%mMo`jj&ek6t7stj=OTIGX*zCUKsVzA)TvcK?)xCc&k zQ>ih}61`*nfU9)Wr@-8mQnG%$uQv;65MWSi@>bJvq_#|d70Gy4^g~HdzCj*;F`a#- zCHL&Eia$^VFkMgptbTXai89;+W5RtOAYpFMD=sNO_L!e;$KfbH^JwbCu6PfR9iVd>^Rz z18g27Jiidq5Vsb}dd=J)UdN9b3S5Ora7bht*J)_3I~nD_2m8n@@uPiyTj=Zw!CDjA zq!_o9U+q4xQCF*O`}CIGwvq6G7x&sXj*VmZKMsBV%VW6V^bIbo-Ts;vaayvuQn*BB zYFHX1joC$wgRH?2FSr<+GOuYlz1m=a&?YVQErCWy@Ad@kV^Ry7{CwSg6YKX{f>?H2 zJ_yzC-x%a_eU9QF8&Hzrw7Xcn-fOpzVx#(Yy5FNHf!r=W9@T| z?#ecnKT?0eUqE2Xs$xlf6>)kqq-b7(jg7s&4>O)3nbNL=p)>4VX z1sOiW%3Z;K6zBx^g*}$EXg6@d_Z&;Hu3-7N$p@7wkf7tnlm#&b2saBC&mp&aVD=c3 zA5NpXGA23F#c0z430r$rUR~y(J)IZA{CU~<)=NHc`$cVc0gm5XP3^%=B8*L_pG;~6 z^~u|)i!p6K3j--#k12%f7vzed(TU8=U%BT&h0khgU2PlR&ND4Ktiz;7{%gvjVz=MI z%WF`>Jtuu<)KG#4ggy)tOw%%QK6jN^w`y$F=(N!NE!206&9cq}dh^ED*5jO(xRS zLxpUDgF&Nd`{!Xq!nsBzA;m%W{<~1_+aLQMkq=#JFf~mp2ToU= zF$YR32lV2#oGl=i_~hh$<=N}T(`@iXVW3vvn}T?)_`}qhro`oNV%44KvpOdh-wID_ z4Lv-QVS_i`^$hWS6U_g!vmNz%U3r6+# zcds0rk-srE>hKT~=<3)}axp_W=a3vQKwR+9p(7_*9`D@?3vpa-CZ#7uDq3VR@bJ!d za#3sK$BD|h%;)&Sz2yQ1znhEB9}25)9q}9D%Yb7@ob3hxozqrLq?KcS@EHGZpF4L* zLfgJ~0;Z+MfowGSngv3%8r=Kqdy?{557WZAE{e~eBqXU&zU8Bmp#G(E$RtE=+{)r| zq9@Qrr)3aIsJcP~HzDGZO~J;bKJ=U=u?-l=$=U#6E~}lX5U5rtaLSkgN*EQ7?5(r` zZgWofwTA}6hlExS=FD_$m>@xFTJl?XAx0n8Un=dp1iDvXN%PjqJ7WOcDF#Om#J+UE z^Hg^afS(o=Q_?Lu*+9~8lf%!fpWuoLhd&|jtmA!9N0@`xCF?XaDx!rEuFBMz`D@yb>eUdJ$Ix+cpT3CzGP!v6f&a$U`Oama zvy+LZI(FTo3KH^%*Nx;$xjHx<>!zuT6;D%Vgtu1s9`mRRd7`8n@`4HxSU)2C0(K2G z?ON+TXE4<=km%u1jT}Bac1`6yqR`N2pa8D9rPuon(w_cYWkgOHuN)~}ow5#ala;^k zvuU>weeYXMmdt!(nu~~Q!nRo%y^4iot(r0cI zh0~_A2KD3?HrDISQHJ+=I-ayJ9O}l(d&xxq;!hv+Kayg4qOJE5b2@wXEX$Fs3lDrF z0-T)(5LPpw7x6oqAimcFoRIU~+U(R9{nq0wP^F%y3y*ZCI!*s-fviZm9Yw1@yeLQU zI%Bf}4os;e8>LD)ey5|?_~Ae?8e)*#GerjRs*QG*$nFBkr14v3Fp7bADG^@ z5n}2Ci&$aqZg3}Clcc1wX6WK4n2Sib`-d)7Jr_KGj%|>xY4mS7KEpFIM zEo8#u{^8C^#!<&vyA|gD4_f8_5~Kc~uWQ)ki(SM>AiYv_6ns2v&c6eow9Jr1WXuXI z&MUzsLxj@7#G0qK+r&zuFK8?pc2o&fvkxqMSq%w)g~2Uzxt=>uL}$+~RB7TZX;ssPhQqr`y7gsFy6n+bxP<0J$TB8kN#4UwjJc_7=GnMyD@~ zS9)T0H)EqPC<2uTqIbXTtb(>xuX?H7_GG;q&Xf&T#aI+zmGp1IxOte>R&gpqtG6!{ zKYKo5*&8Pr9^wG%^y+wj5?wi(Irsd_s^8qexs_CwaBAFHkj`R$xIR=L)_8Ce^yYqm zbs-ii%4`@kn9YfF6vraR!f~|I;h75<3% z5(NCpMz}FdPoN7^*f=l@Q!hKu;Om8t@c8Y9fPJ44CcMjxCw#jN2%5ATL2V!`#{;CM zsGF9};Q63YC~<@j9|?#{e18>tH7(3YU;`Sz6*s9n*KJnvjwpsZ(Gv@I&_Ecu?&pkmrvj*g&JBCfX)10)O){ycH& zCQE~kA|fSXovP9t=t<~sgSwg}dO}Q-bB#&fqIc_F@fv!prJIU6$7m0s?YW zzg!P>B;`7L`tJ%R5-V)v#77~k4Jx@H3jH8uBjC}-iocvD+&3|p<{^#9>HD6NLYEmt* z%~9Tpdz?6+=?m}fed_m$8Wo~Taf%3~&g@c@5Sba5t0Mem5_8R4aBiDa9QWVJLVPfz zPKdTLz%jc>9?|ar;wboa340Py2T{ zet+(1layAHY!5n>$Pj1lH~k%xD3>)0UjJZrabcLnTjmOfH7HusJEewv50_rOFsSK4 zt~RiV_z=ban=r(pV3O`s8f~#-bkPim`23xHsiUgrUWc#<2_9qGR>5eJ!cm6WQAm?- zMm1N~e>g0|7EX2JH_P%Z`id$&X$|ZLF)Nr-mos}RcS*~EcRs|pPgN-BrUoFnjz4@d z+64prE!|8TOCufsaYo@MwcH+EBDZqXFNCeCzU&P(C^ECbJH00_LRxirqZWUMb^@Oi z>v4S$*!oFG9<`C%BdgUFK6@EWO5-#H*s(<6Epi=+*dr&Qa3&RE^KtPj7`Iu#%?mgN zax%c|Dx%ayI{1YHC?yP$RBH!>xFJs57QY2(yRC9b(=J#P=4!7$KbKK|%$%sLq#*fF zSw`?E`r0)R5-{68{bPT2%Yv9t-c%XSszO4nPPd?GyYP`MOb7fB?<$H1$X&~ysWqes z0iw3gE1Lss`Va={$n(hixyhgqdH1!r3$!0NAwbXx`X2eQc?b{uCTXQJmW3l-mJOv( z9}Wydclw9wLuGL%RPW}kE9MX7r~~xNrt^q?xdBwYzLNLIe-Pzq4p-6>3K%$=x}s?H zyVjWoBx(pOFxKO|z-_xmLtg4b56}HVR@Ju*-i1keHru3hS7BUP;ct9Eb`^ zX{`DWHFl-Doi&6=B}}veizgiT_16uey6P;u{`5__g-s&TCfR9Kj{oxYnf}OyAUPLd z&>QyHXnty-P@dU!)bA!aUl-2cP02I4iNv)q_Pg44hVm#43{+Ar(1bx34k}){MbaJ) z*^I--_s^-hF$2mAbQQ+)>))0l9>1FkVw~3ipMVyZJry{f z-6{b2g9b?SPnt@h%Ng1(mKeTew?B*J@nzhcY=}vD5}H$K(Z9d%*RsWZQv>HjBDg}f zF&}HA0J$p!_eI1LHOn_|AzfbQGuV%T4ORxCHW$k+ADVsxc^vc9T=dj_?!!W;{vVfb zu0s*_wh0#K1B08!w>~ydWARnetupD#b`L5wf<99LOV-K1pOMI4VO+J>MW9*TBR6Iq zi9>a%EyFV(ps~7QeFFX?qNm1Uz9EQnX}86A@zV&z}(0$*DKeV&4rd z?uy*11$tKAgMqHDXefEe-roeoQ&|>{6wh_YNA+7gT}?Up(@7TQw)!5hme9zTJvd9X zlwo@jL0a1H@qI-$_qqY072S}e?2{0MYU5S;K}I$pivt6bE`-j`+jr42y0^R+(Flu&Dew2<3a8?%a(pm_5M@R4pLBZ*B z(BH2c%8z2Z$gCxC^N?gamEvAK$M($+E?5>MkH(IITY{yZ3#4%k`2odKOYfzkM`x}B zx)E$cwl{wrJ)95t5bxg^cO49zFnw<>qQ&-dl$`9biBs*@=ER15<5h12k{a5Ob8D+0 zYYRd>=Rb{=l9bguq|8)JWMMgbR?g`3*g#dK{)nNm6Mu4I{(#h6d zAUul+PHeZel5SBNXkn@r2sqam`Za?ii2z4{DQHOwWxfBcl9ZUn8YD4zo~!rTu>b}d z0idLd2R$QfM<*Ekpk-WA8|p`m&gEc7k1k3BuqB|}#1SNpTBL|l5ZB7zI*fYCVtL9X zRcKo{7pfQW;taEDJAg=*{>)E2*7EUj8t4H2E+yCwd@y@g{3>v4g!PQi&bi#m{KsX# zfcq1i#mNQ`bh^B-ES@Ltf7{LBP7AuCJ6A$BnXU!;dNa&&!3D}^NnnAZ5eM;x_PUvE z2e7ZkKc&wd+GNgd)w*ocT4W*uV3J`~f+V`ffxqDjP681oFl6Xx!eJwm1<5h3?dWlH z>%8c|-DDDwfXvb1`BENgK@I3JTeP(sp;xq_var_;SwWR0m+vaP4yv($ubEf>5RK{c z_*xka$QhFjcYbuFAUmmmx28o$V2r08g`b`K24_-EJ$AyhXN?8IT~c5!6q+trL~*rA z1UYdp#IRhEK>=l{qly9EVe|+_B{Y+@#EqM7s|QH$$50nSDNuDiw^Qb9kn`toLgK8! z@Qd$RPS)~Dsk5T~1^OgD9;`2wwntzue*E-2*Zerz*;%bFW)X;t7?RRfi?M;t!nc@1 zLX+zRZ!cH>F~G}kEXY~<4#xmBvx5)AznpK!9osW_Aw)TR3A`P5YW>85$TS`@tiH_H z`#8LqPBBe$Ob5X@X5AM%;*|EUus*?W-66*;#hllP;;=^An3Gwn+X$4@u-{zcNJn(Y z#XCwp9X)B;PNZF!%goe949;@KDyWsQ2d8^djA+*L;M|kH+(};U5!ylS+ zLNdCLZw^g6YlE)#Cn`U^{I$F3fAf_2U=e@6Q@ZM_-Wf(t3GV^0h0RD?2#N!J)f>mU zK_8Y%w)Ed!0syT9#=2C}6W1yL1`=h;TQjtyEu@^)`r>>NI1>g7lBGedZV%3`HNILP z8$76=yV-1lF3Eu5Kw``n38al2nB2oCOdA3;7MySJ9{Voo1L$qJs>hwYxi?jDx>Y)}nI;eD0Gs_D&#+c&y}eoY2C;3*8=(eH zqKpfXKQvE>B|xbsG%W|Vl$N+q1OJl_gbX{@xF9e*loa=ZJ|K+J=%N776?G+Rqnges z#(Z8!*X8KY?Nb&xQ|)`1nN0K&D=c~%wi(mdb5ot3Hppzq1m(4rWdEj4^Iuqy|9t0S zg9TIV5OpOBdMCX5!=w}4ojI^Yig|%%cA^&YePm==(b>gj)oFhz8i!hAP~sxcp4Ekl zCY~O%uxUP|&S27eD4{vu-zB3WG@h);0j;UeOi*2S(>RI_8u7}ZyMQ-5F8qs7QkqLx z03_;-KLyzPRNE^L<<%R`Nk4V8{>rsrNNu*yuTB*R40g`;7?DljfbVVIkO5x_hbmU( z!GLEn7G$7`1rQzS4HJY7yB?G5r#GvYV3@KHhIb{&0-@DJtn+sCTKKus^*LDc%-~5#CvYoRmo50^%KpNmvv%+8;IsWjV$i34 z(--UeKFUwNL6h(dDfR(i9%-+7422q{bywNXZK3Rhgge&xMs&>he1*l8 zg?yDnPJEg^OhY4oo4^;7jIPYg4c+B;r3`Ar&t3aF1K7g9TZ){8-m(V9jN}90WTdOQ;81nltK9o$k=VkgaumsF1rUX@TCeF$I9uvm5_WJ+vY5uDab8Tsn zDAbr!b%g-XY1EAEo!co%%G}>#i$VMl(cX?pON*cVYRpD)h~7FEMy>SvN<*m4`x+)h zU0brmxbQT}3vp_p#*1a-lG<*tUU}>yDj8IH>+Yvg=AgsA9^9n<kBmVA`pG#R?^u1kjPiRgQQEqX!0(KONlIzU%Rk>Yth@@a}(&<2&fSn~9 zgJMsJ6)H6w#zy4u!qHHXKQ{n-Da#x%)J){8LDI@Y9(!cvux@MQoQe?+Frf+!GE$N# zcuPX{-R_o8S+Z~%Mtq+h-XJAniqwUYfWj~fB!9r#h&~KhLtIwV$j0OT`|EP{*W>ml zpl&9F?l(GumC|Cp?Y6RwQxFWE(z1&CB@P~Ht2L);`cvL$##$PrM-M8p=nxgh`O0q( zS|D;r<@Zt=sH0yNMZ(1hWh2ai_b(0?4dF8wC2LKUZf(L3`%JdV72j34!wd#8Ct5a| zi_HnqC@x@ePL!^Tl>G~?pP1}Nnb^x%6NgX`mIlvv4&`Gi<$d4YLB zIA4w$66K&J4O!#K%7N2z&wo!FES{y(ys0kpX@P4YzD)Csw$1~5*pfJ^B2_pCmr^yV zb8RRInI8;tOr&4Q!$VdR_Hfq(!i>Dh>b3c!Y>}r1_sd{Izcp_6;p4qL3(uQ3n=`Qx zCY8oxh>=eHbN%C5%bi%z*w5Ac{+@q5?4v1&hzbBkIYd-M#bjy8D05Uo77U~^TqU^} zDb5fOu2u+g5j_1UJT2g#IhOU{M8dJ%S#Yq+NapzP-R180yPA%}0_wieo&3W$liiWg zOuy;bx4XTzyL6(q8#X;lxT$1bJ3dKhlmwsIo);cWZc^^}RJOcd>Ax%93MA#Kc#lOE z+SJtSIHwX-qfgs_;qZa@9}qG{X8_2@Gapa{sx>=Zv^J%mx5bu68+)>5kS)E6jo;`6 z^)0nHMsyR#q$x=KYs4EAi~=ub5vIhB&^A$Ns^E?#Q%tUyYWV}nW530Ss+i*q-m7%_ zE4=ox$}p_+-eBh51isRK7AEAmw9*|0sj$mPUqx6via+mXC#kku-%r9ucC$q_TzQVvrftdII=`?o}0j z-b)t#X>31;O}$BgYQx(}vvWoqt)BVo;-yo%;(mJSg^l>aPV>D~K=?~hfJUQ22Zih; z8Ez>~|0mzxOGDmDys!BFmj;K(SqxvHLZ*RpXQIoW1 z$Axji8SgF0$C73N{k%Aa+b~QSqko#ht(0!8N|39|bW=3r)hrKo2s7b^QNx$dT3{{h zBNs-m$h9RAnaP@j{`8btf)Gr}db3>4BfeY;t?|{^LLW0%+uqcw%v01d+-x=;;i;>x zq-tB=(mC6DX*q!p&bs<(y6sRr!tE>rB|3Mlq&3b6T>NNEIusI%S~=!r`vbp8^jM&Q z+}FlKC{X0RUcDegQ1x6A$j;fQ0Uk3Y`coR?il#ilbteKl3R=7(mDM0`;2*(K_PaVt z#VXzQ7~1(-@M5AC;={k48Jf)Ymc*7#CFdIE=%{s0KU?n@Ebp$$_}^y*5nlQW2?s{x z7E6(vUQTV9h%z!}@fPMt^}A-N9bo}5$K*cCMsgUO_Xln;_lJ0)Spp?gyV{PnHoBRs z36+ZnCcC$(sM^;~HG}(>*1r!{z8Ek2=k^cB^lWe9Ovk|~DPttAzB}bt9Jb)>TEsVs z4y@Wzu;;v!oX6;>fav@Tl^?eqe~=g9ITtLX)7ZMP$?@A0_UlRA_dg z=)erZZ8DcSvsht{44j~1Gl^N>NBO(<>!@5Ke-@twze!K(@hR>YZPI!@KA@uzGZ~aU z8Z9~)d@l00Jep8H@e$Y%9=B zu6~nWp?W4%F`%^_A*L^JRM87+r&II%(eh?u5{oV=#1`HA!=q{!@2N=Y5ebN?U|#S{%YNl!Po0F<8|;KqnfD2u1+X9dI@ zYeFJ_OgtcVAq&HnR;3GVLsjU$5?w4S)+!0ZZ+$L2D52S4`3f6V$*)_R5%<&QD76%Z zijm6=0tS2-<%6$`t)^7Q4xh=Lr=J=gJ-;*wEPF@~pEy4@wPe6u;!WNj>(s|VneaoJ z3Ok?2Nq(w5YfXL@f*eDV0kXcjg_2VxO{@JQZ@_40N-?Qr99g-3F&0VyFMCPKrI|mV z$vXM(bkkL7k-;m`g{fd$z+lb&_#2AY_M>Lbw9;fxD_Y5al$eL5449~N-o4t@mPR0lj<`4n}!3woSf>FQY+&OC)X(1)loY%%rWD9Lj7aj50Q z7T&Jm>Zz%kTUK+)m8|GH(9vV(Z7y z3)fC-LP_DOTecKk``$qv`Z#9gdOkCP7w5EgNbxJbT-yyPTj)j@bzK3OG1OO>zrzwx$9zPdo0FrS&F3MhB3yIu_@{%)RK^3wMc=V zu{DY=wZPX`vE#0uDkU5ua!r6gqb_EcjwVapB;B()O*2Y{U#?dA%7YWOc%4X{5!>l1 z>5isFFc@>Sf_@ivv|Xwtilp#WW6>>v1m+*jLilVSl(OODB-vNuO>$Mn2fylXXdj=n z*srK7_NX?iCgzST%a$@w%Sld3)%oGC`&E>MR{X8}KN?~QqbNQ6*2|bbdK$?O%eLDC z7;HT;U|nhR%v@79;iC`Xf`Xel=9HZ+8ujUfx6<;m=*EXT?PI?=HPxxxz#n5|j-H{JPel)2I<(y3? zK*@8efqVJI?A>w^fW1>y-oqT{B}5^YVDVdry(FAKlj{Dj^p={2*w4a+dgy~Z_hIMb zJCa?W;L|Uk()d?yf7aX5D$TfFmk{ObNM&??(vwbb5fb73xes&_8iE{pk^zT{rR7Jo@-npOyNxDEbmEfxK+BHHSTR!qM#gLuvHs_fBS9 zR}q`7(6@vCw7~s;>)m;`bWxnG%xWIbKw0VZ@SQ)=Y6loz70} z-@;m$d!vLwv?N}L(*%7vXy(W+Iq1~k%;{X@&UP(7S_&6N}GQ(+F zZ>yPHx}d*41ma^AD!Zf2h)}A9v_> z%nrurQuNt5$Gdb~p50ck&JQQ)$zTs`2`n<&jjAa`s`I@J^22Jt`b5FQ$r>TM(6evW z5z3P4@MDbUpL_F1d}MQa&eqkIq7PK0Qhv0dJ+I|vlT6*Ve6S1%Y_&DcscO*4r>{b| zp}m|LjB*SdI8sSnKU#M#@8TBc`uttij5EUpoQLKy?5YY?RIZRQ`PgW4Lv3>izqE{2 zz7pUpx5l8}p@I=1KGY9cpv@%esAy@Guz-H3sXuABksb4w zPiXFGB`!HImu01)qTTk;T7b>cA>zm`xyBSP5vRZy?7XFYJk`C@ZZ)dD21=pa1@eDF z@7VqzaE*L_ra2I$YzkX6SIiM++s#eT-5__&bPot|f)&`;j5JRZ|7m{(5Byc{I`MntmpDx9TomQE?%w?t@kdUhy`wfPzhkC~IC7+^t>SkI$oPVeLMCd>9N(a* zvA_pvbCP_XrT5Z|{mn#vF?hM`*<$8hJ3DTy#3<2ycobrOUy7tjs&TE-V!q25XP$`R zx0)%Rq28HdRxHc__gO79F+WRR;2}@7?6&1wQb#72HI0-KKv|yCW#F`~9J9^KtNp4j z8pO|D7#JbxaU4cj6Q9hEV62&Wz-*`~aCO6R9hWQwNHNRbCTF#vF%nLnjwH7$hqK}m z>=H{b8!ASvW5D-fW)3DyWR-asr@8>D+2PD;VkjiFX(t9XGS31pBBhfn`QvywQR)qa z$yu!-_LGOZ$s&2o&Q$%Vlq>Q7ro6TV(~RBu`Co#FU|(41)Y1A#JU3Q%|M)X~ZrQD> zKz*OWPsAGT@{yGGrWgE?5JAdAdl>F>Tv(N5qUH^@ow5>F=_p^z2$7yqrs@75ZP@8$ zrOG7=^If(gDP?g|*^0xNY}?R2?^@SEGNOXrqzx9&LH?pZ$QK%KLdu7=!6I;wrI!0Z z@?ui()k2-L5I2?7Dc$i34OMQ%Y>8tDe>6)b-;W}$u1rL3^h92AMXL$GZm9!}o?zKf zCYGixalpycC}(b>g8Zm(BIZD=QZU7T(gn@qHo#0jboqpM3UD9XLv8<({c?`F&j_{VIFVpttXRwTfm$(RI75pu5)7GqAKHn zHTw;+*C>qtTH~lDa3v%6S8U1HR&hAIaPIf73J$l5%s{ydQlpEYU;Cu0RtSVi5QYfQa)|#ObW_v?)Jz{Xg%< z@A`h?#>-8`=X`Ws#|=6R`NVAB2Rz(;h`zEv64ASvuO%;xcfX6WsVWY+g>$dWW6m~1 ztm#nZv-dCm(&OSPEanBun+??LC3}qt8ov{=*SoT<;QSp=Ej9Y>cWl#xl+b3Vx=Ec* z+FO)GDma~jUU;*RM3($`1up0%yhFM$_Ypw3+kX8&46hx?SP_KTHZ}>1?82uyq!Dxu z&?Wy@mrAVfu|*7pKn5$>sP(}n)%ang8U8Ql ztsWH?rK0?}P<@@Hd|jzZO1g3HjRj^n$;$%GKSKd!S+-rY;+UNS%#zG!lE2>6mWUm8 zY2>+7?Coyvy~pr_-5Ikz@g!TEt#>#;Jfhy?Xla(ctOMvi5?YKt?~W;^L10CwE#Lmo z@NeVy?d`vL4|g^cBWZBN-o%Sx zzMa7!WoY9u*~yVM_Pp`ek&hmlVwHrkP0i)y<}wjb%+m`vYoE$pST`R$#_-Zzna;Fe z-^h~f*v%naXTA&d_&)^5sOSkfzCl(Q*kzjpiYrs;;h__n+gmBd)A#S_d`Fagt%qoj zXT4jWu8jLd_XZ!6bl%Ha{ZUf3nzuzQ`!rcb@i@BVFjUf#72Y zAILSSZ^svAoW&AhVlu?chi4`oAU(%rWo%WLJ3C@cX@npBD0b23gGa|Kt$R{5d9oAmwq$@^@&CsaGtj%c7FQo!38fJ4dpNhOo@E3 zi5@C+bFt+gMU~0V0ZLt37b^OqQu^BkET~JFCKknCET}2FB~D*joN29~FwmXEszjn0 z&R=pypB%19Z`Lq7++RLv4|VW;q7}FYo`~)K%^#+B#5sI$24F)+mq7)oTm@tiPN;T+ z`*DEakbzhAVmXOa5%HBcI}O-tMSL@gtn>>WU(e)TNBpJU6!OY{js8Zf`NU_Z4Dujh zp`k+@j3%U`!C#jdPX#^n)osduCpo`N^u1@zR&A5SzaX_tRkDHit}w#7_4XS#ZvbuEh7vGL!k@iAp$ ztgx=I3v)9PXB8C(l~2dg z(W@AA#TUeJSU+WU`5~8TNR&wxN|iO9CV3T9@ZlLRRO881Px<5*nB%cz%p1gvqZj|G zY!d{aEjv`DOB<58JB6V<&1#s}>Z1ljzxWBeiu`=^|Ecsj_4Ub8jXI+Eo>qE*`D1-P zJkmJ(uMtuei}7-dkjuQ#wTBs?Y>>0bpnwm&iezQ*hl3G1V}fn8_tB@P=;sNyeddTzu^-AEcxD9JLmm zv6k{K&k6&49XIs>=~;QI5HZA!sBtz{co-4tuIi@fjJF{7p1$76 z>tJanmQjm1Lk_KAv@%;-tonLM#0Eq^tqe<2br(O$qFJ}24jg9XqzZVPLTiC5Uu+9L zeGQ)0K94$nSiWZrRNBaK?@hmk4OFjJxX{B>F86lLx<1MswR#@ck2Q3!bFGzANl>DM zY-czGcv`|0_XGf}2vA%XnC)pw@2cz$SW;=qE8O13ReWCjgL9Cvk>W3~EGToOzS7ga z(xXWmCu6^MP;Y|Mz~Rd5{6tW4FaC$|)I)g_C=yUq;5+DrS6IL)u}>wa!D(GT8)_8n z&4WNIc1bMF?KFhCBM%0l>;|zl`A{D)Yt}u${NoIW!Yk4B;(M^8 zM?R)(2B2YxuElb-G53|(`G*|lowC%UvL)yi719@r9*n0wDqf6XhnwCV1V1*J>lZH@ z#fNLQ7tO?DSPsE&WzZoD#T$)SvjDVZ0%OwULSX@Or>R&jHax(+*KCFZv)3E)tW90H ze#*NWZvP5OvcSent*Y^h1Dzw@Nxz{ZSd(z-j0d&)a8x8<32JouJOm4g4Mb{H7*q+9 zSw~0?lzw0iIonit7^i)kI2uQ!h0#qZs{*F%9Nv?DAN){zOkJ*zp7uS8@X~l4je`gH z(5(%Xn{*nEN^F9{rPNLEL>}T$%K_>@W;zbX?XI)O#sinP*4XQ=aR+-iczs-h^S{1U_Z@|r zc#=LHSDBv<{*c2x@l{TXD<|1pn|NFDf<=3wCMEaK2_EttYo7Q`(|=GGUgHs7vSBMf z|9*MNFn6Jdks{1NIxu?5SoDY{@Jy4~H`1K_qSf^@VC54T}2+Fs_~eBX&ph`li+*(+Y}eX*yCax)IS%$v>DL>MmJ&PK|;dQUP-OF2H{;hy7m zOfTQwt>Zzpw^ZVsj3($0_Uk7derm`4&tSIaHihAX7*C}GaSh(ybbDR{&1fhvjFBSv z$EHcZGcr!jS_uKf6 zSTIu*PN~;ml!8P~v^ty8GU*a#T0o=y99j!UQgBAH(l@c7FzmA>zO{s8KS$i>F7~FEv1Hrv;u__O?1c+oQvLxP1*Q{FC?GNjx0G{ctSCw{d@KjpD-F4o?r}H zH1B;(ZdStd$6v~I5YYEm@5bnUKo_xi3DqFlFP5M9urZ@z*pU%I&5$K+%;HxNxBYMr zEC5^i0zZw+8Omz1v6(REhd~QWvJStq-g*JJzVJm&=k&1rV|rQ&>f3dkCd5OTOt_&< z?S|`f53{Yxv2m|$B9ZuAPmWiC^Esg+Y0Zxd?qp#NQuvu&F{DSk@i05iJ-1+1KIoR);z9D4f&zn2-I@Vz)}oo00A0m9MdD(FN) zWqI!S*n1=`x`Q*MOt)|w{4SM%P>1H&(ES4VzkRz+YP(~n!wnN?zUoMct|rh_DL&6FG;W- zFodQ}7S`DMza)IRXSkvqG=UG$R)0(6yHEEq$#+-LWWXe2`WT2Kil^x|6O*iUP_on! zdhCh!eHG><6MlJ`9x(tPweiIp*p`N0ALdcKjAK0@rX95l zsUli)Ke#ZYF{}*R7GK~lt#hCb9cVDdl4lEV^}GnIpN03oARIPvA+y77*|EH870yme zg5A)y&l4hSic((0Ni}2u|Alq2A{KPabJLI*dU>9%tTAK&G%RamQb zl?w4~hLV&pA@_xpG`_UUOKk0V%FOgzAx*6Fs#f3KH(w)V*8%27J2{DSHVLHwrp%H) zK8)tVFn7z+sDai1{|}JffD7#!DKCmy{2%<6=)T)dws1I$v03eB!24*Ko`t!>Ie>Mx zO+k?A(AUAV|2tp23Fh!#2NPNrUQ-3~{>EPkiQ0nM-;l<*TY5paRcKYjKNZ=mmxqM~ z|Le$eh0xOFVWQ%BC~`%EKfhc*KGf{Kv8nH9CF(*opm=fBa!%_4&(h{O0+?^=w$a6; zvQ>t*#(aK0MpC;6)~=Bx<$>!+as$*8v)wH&V~W8!oUvwx3x$w zG5)?iqDO>AVfKp=n++xf>^HRbnUbZ+Z_*!&{zQvJdLotMB(IHf{pqmtm7xx?pBdk@ zw@iMi{bqY%7mn$y_duEy|Du%axrA{_?7scq>9PM8J={TmCZiCckhEGM2BU7J43%cG z^RBn;+c)=6uly2pu_S#|A;9jqAPs~>6`~dQe}O$5`u34{V8=luk|7S&cs6nsKKt}F zq$fL5AmfA3${NNA+3#j9XT4buHS>tWKP8ei*88mC2Gyfy{n;bPh2FL^#JmVvf??Lf zOMRm3heaJX@qar$JX$sg`iI{wj@1%*UKGBS_Z@~(bYjlr;gM*wptl#1V2I^)l!7n* z@T}dkDm2<*toGX===4EYTKT?&U#d6AzYYD;5lzAa$eIx;eLi!SwT^Pr`s1QbZ4vu~ zRiaE&7;R>SXIFo6FA_+emvf?u<5vJBDkVvuHJ5=Z=fY=!z>61J+8=Ua5jg|}{G}6a zR8D|SHR4GOeXVz)j5H^)?(0H0ZFW_h z#nt@qJqowzP3}uwAe{H-pb-(U9DAaJjA$TKT(ZCr2Yts_^y#FoZe2{+7@pF-ZbF1w zC0lco%pW#PS59RwiKRXn8gBYFv&qbZY&ew*W)9d`{fvKBq&WtgizhIweR7b(y&Qwv zPmNS^EPC-eLo;lP{dD~sF3MMCxfb!QgE=V8wAIdd2uAoOI-Di?NARDqDNe)D)sAny zj@`D)$ai+TH_s~;%!craPLa=bHJ*{vLE~h*iC0K3ORl&`M(*zo8u9n_3 zU?0EKQP3=w%_zeCUDodhM~N^arEPh*9BRWl-$us00{2T76k}cjBf=BtHmP`?kS6np z6pr|NGJ=@*H+0F{hH#(6LoQYle9x!BNO>XW+*y?7xI|&?_}gnRQBQw)IK;PHlE5c4 z2pgO8`sU`%{=YtDh+XCX9jFZSk<44SmI*%&+zo~vb~5wwQ`fMk_nQ>#&i9AbBg(u& z+DAQ-&Q%&F-wkcSaqQuLC+?J^Q9QP;`xyn5iQ{XWp}M1ai_7?s2uL-}`x?PTOEAUE zy@h?=>ub|(bwx}V+9YjCGpN`Vsi$0K-{Zx1j>Xc$IpQ= zaHj>wCcY@@c57s5w)!N1f5g$Q_};Q$I%+x1)NEmL->8gN=%qDM;ENPE)qOEnZeP4# za~RWm--UQA6r+J`tyRGb??$V}-^3D^#=ZR5Xx=t6?P>`?oE{QaTu&C@klk+p)w{A2 zm4@nM-SwqakB<&laxx9LL*Ji{K{L)tOKT7?mzZKI{I$k`6*$*WR)oJ-<^@=ntkZBqA~~Ovz^S)8*%I}i*(z$x&(i5^qR!%Oif!S>jUexzx7V)| zrcB?qYji8_B(E4_&0rcW6JZPKs%nn;>}=LfB0RR2iL{FMZU-+_umWDX{54(&g3(<#Z@)a)`3sI@duTiJ6AcrA%{gJ)Z!)k+U7j^E=T6uYeE?s;45Gm^-UOaby zXVyGh%Q6s3z;LzerflQdcO`0wuenJ^VEM1O6Ui zJ4)@jnPm!hHWtk0HqjKCxyJ#UkUb(y#h9n^T*r1T2l(SPF_!r%fPoK?(AR5?o)@w>a7_WNp6+1+`mw} zJD%H@{*Fs`OHk|=AxXHooG6+IZDj`MxFX{1zOnxB7k5VVr|2Cb>|_!F)K&!xQ=Avi z(|*0pvR~8;Gt3FHGn9M1lQgv<)PH;2x@8^B;h|FfI#mo9xRlZB6tVxpx(ZmI_lx|` z@Y(+@tLcG73WO+NQ6G{tOM3lIt99BGnULjK+q$^Y7ieYSdM4H`WfGd!E@kLqt8z3Y zKoM@>HK>M>c$h8rb!D^VKg#@YN4yfpBuqK707%P9^R8etqiwx6RRku6dDL&vGP3Hw zQudhjHW^*&H%8O;>wbx?t0FEM@qXSa+WFT`V-e`AjbYP8@MJIxKp4j)tr$|c?XH2# z!B%Vn?}Dm^Mz6To)3}|aObj50O5d${YIL&B>z^Mh)c0_qgCHG%LlF(cy#hhedIGNWn>U7E=FOiGi)1X2$iKDR zVFDR%f5aXDk)3g>tnnQOk{A9xuFGWTM!i61Kh!@j5!*~c`p(ieT_lO9nrYyGGsLiv_= z7*Z*}gij;9h@mJ5O00Xz;7ziGF8YI^qaZs3n^~3}3|#EEzRA2>_JMrhXWM9_SKMA= z+=A^4uFE3oJwDKL4R*{ukI4@IYks5~0?$(W%pPRM=$Q?0uVac{GmFrhP?$@9)Fw(J zb?pDAM&g#{GHO&*-@v1~FRq=QALo?$70NC?)ZoY#_jQ5pDNLb6kBz?4_*A>`R+rnQ z_`i++95l<){MPPB_CGE*iA<*GS#UV*aUe z89JfkXCqXpn$d?S$nXuD9#h|1%%-)Cw#w$_=&6{$SYux0Q+db&eu5&_ z9!k`(ALK-jb%sL0psjw-VSdAD{J5zpRgVTYKV)m}I`4 z1V_c6EiB5(yW;52PRJh)Jftm0tgu_4BgzH&&{uaxt+uWB`!|3iQm7f$|8X)kl4iXRCAF!xzR)dk=4+;w#U8%sd0b8C$|WoKB50( zIP|a9sYbqoKN3`tO+KVR)~`T?pAS4^G(@AT$2kI^TGf+n;j5T$!?N>^|%}G;D zCTR{U#K~BiXo?BZ+bs*V9BLf;i06oJVtp9n zPwfst2dR_1+GxbDF?|^qBPXDaN2<;e(Gr8hKjcID?Tj8$dqgu#=-31_dF zu4&yUIz?}r3_mUSC^H&_nHEP&S-om;^`PnO&Qraz73@ZPzg}zPe#1HA9(~8*lK%E4 z-V?L)Irg&g-~Y=0RYH5s4mGF5W--n)kQ3V6gkndSd}K!iQ^TD%u(2^<^s{v0N`!D( z!8&N$!gJ&LY9gD5`DNNKT(|R~P0bLCtnPd|iqT>V*%)r`BHi9%4tdlp&8*6A0qq@v zwyo-5{2y5>Xes<^0|CxRXw?&+5LbKA6EN3f6e7&cn(0ysbA)HH8?%Il%~JUwX+iOJ zgHRJg9Pf#z7bPYxe-J(GyLBCh0)lZP;u`6{r!LkR7IOaBXtuu5#v5n(Dx|^pG$q5btT7D)Dr07r{5RZz0)!Vc=Gqfst#xW;pDoqVxvWR+|(P} zC4pm}MI#WUT9{f|Y;>`&^_V$}hif#XAu63CeQ}Xlmerj zHq12St&Yxuh&uJYrO(R((S%3iQhKS zx#a@%R4SBu-GvrFN*ySpX?PR4BPPuR8C-Cj8Esj5hvVOh5!a1lluU@5a?S4RrX0oI zUqFELD(8Uh5M`o^mkvkjVjzr<4|BA- ze|0NX?CDP70uOI8csU>CCKyOOKl1Nied2oFkORL{Ls~%DlX`dmoqihr@5=64T&T8v zP>EnOyfM}%R~QMSX3%_*$LZ4n^QVU0L=ec4m(Lfa4~kOHj3$aKtp`8Oxg#lyQTo(z zl}M;Y5sRn}(f{daA@>AMmdXw|`a|bpB+5WH`JZXKo^jLZwX{4KI=*el^zk30^}+P> zlIbKt4XOy~^K&{-7nsdP>6p%u3R_Iu=0fuWN8X6tScyrsbbQyK4#eVZ{Re9#w{|goVI95Ha+`IGaG{>3+SFv#AI7QxCi+ zX|58zCs~uRf=FYN#ySkyCnC9;R=;zMU-U<^&GNECE#8S|v{)h#o~JX$K2#9a?h(*S zK;}8J5=-FlsA6tm<`!9Ta%ogvPxQiOia0Bt%@?_w6)^QnX3qJU1BHIciB>I5OBZCy znqQ0Dfcy#1$gi5-J;T2CA}#e@3#ggz!orZP=_WZ~ z643uA{Jrar^rShqL_hD&EdrN9;cFa9k>-x_h~Kdswu37wh*&MKrF8FACK#qeg-D+q ze^oVG;S&LC`X}oE#{$lbPa@g7U;=7mzA?&th!qWs3+ir*lM*B(IKbR8ux`KpHR-mp za;(1_mQ94E3s-KJz9^_<<~Th3#ynmgEav?vBYa3=e2%%(lw?&A(+o4>`uB-}g+n_9EXkVHi#8D~_S}jftwU!PW7Q1sL zRDKFW#X8tM=X}8g>J&EC!OqkfoZ6~$Ic*0zd%&arxndOk@+l-oUN8zBs1~2^yw98- zJLwq>E>2&Ou&Chlbx@~Nz?1UBbNTZ5xmWw$c+zpMQG>Om8^_#{;L%amh1n9m3fc(4 zJ?t6%2{Z>uW)XJ?Agg2e6*CfKD_`j#x!TxgU2)Q%$4YNFiCg=vl55@k``EL-^)9~? zv?oRyKZ9MXXFHPxssl|SX^sFRW(s_UK~|b>d{SJ_PwO6^dS}}+Q1TtsY6x;YWWjP( z%6k3QkEAR)uJ1$yYm-4hPG&8}dio`!e+BIBZi5={;Zre|nNr%NTaz=1pFu3+vZ z)bWWMq%))lx#sH^TZOKt!fmGRk8===ECwk^PlJyvZ4>wjU}jn|KVG^5u7dbyqz0-f z3MHFGVk5unGWy${t~K8^e+FurSg3a^k(-d`rrA;l-1j+77jAmM1IIUrroeYR$kvwa zWz;39tZ}ylIlcy0^s{_55ki{Z3;lK2k`zPY0+ewR7PQgp7($&UZS@iipQY=|zB;6! zHc;w@A%wF2=iB-XGZ1WSnK2r8YxUE$VJq~hb||v31&@4)ID`hhEJq;V^E2eI>+v_c z3y?t3cdVn|kZ)zn)nI)V8oaa^sU3WTwUs%#nnzjjXIR8WKaSysis)&0h_4L9_@KPo(etef~npN>&{L2uuRxg)1(b|ZVotWv+1 zSIgXx4|iBj__%Ea9VSCUnuZQt^1EJ9p-kOPyt`8whQGpAh@UfFLj=oqrtPV{9Kz+| zA88t7XRTTEpL?)>WP(LfD|iZGxgKX9W&uJWQux1JUphAkgJ*u96QLT$^B7CB`qA$= za->QH$$xH7WO#Cf7qfwkK`A$=LjYp@d^@ zy^AOE*~_?bc}NN)-VAg74Y{vl%4*8m!S414K#UhEvo_XOQHQ;^gd%bgls^j}r$8o~ z?KMK`gJ#NJByKv$G9V^ygd*882I98_KaL@r<%!P77$_ZFuhs*!#mKjNi4-i6Ipfyp z_K-(O;5An|eiJ|yKk~zVs2YZ3rZ$jDu-5%Nrg)-Qk}IhS28c0*mgU&G zh$eLWiCZj$l;iX7_G$KanyR8a+ST6f>t%8YprT)}+IK{}LRMy$H7RY8xaea~M6Rlx zI7`eD=JG3L7_Z!6k@AN}H8Yi6M>U*1H~*W#s(5yWnyvWvKM_8pa8PB6c@3zE!6#54 zD3FI-J*Y1wc%GaAAkR^4C$kUpNk^;^(0nL5G8Qm`5rCJJ1r-q<2@Quj3k%w@Gb1_t zdAwXbUh}T#Fpm56aY>@03D&tF1MK3v&)+(sZobH~$qFA$siutSuB%d@gw%t)-QQkG z<|6vWX}>%WpH_e{KO|$O4RFPxru^x; z6EJk5Rn{1wL?h)qWZ(A^FUEnvuuzu}@SuPoV*)V$LH+{ro9~jelQ#v zptbfj4;`QsgyGuMWwoHr^Y=C9qlt@#;h7Ngoac~P1USM!+1f%S+zC;*7Dgk@Ko5D(>ff0wH3O$ks^F71Sdh)>~oEP{A) z4T{IwAtuX7*sgTVySJtNpq2lro_xH;N;{YOPc)vLbQr!#GJYq@F~yhLMUzkq^5R3X zvqrPmQjsvo6(F24iTz8xG>~J^O&`_mhlqYE;_SzN!F|5_s53wy$Z@!&3E6gO6Iofj zoz_v>^;JTJR4>#>4!Yo5I3(fYO;0vI52BMtfZnkija3#jTgYj01HUgs4XDD3IOTq9 zfFx_qDgW&UwQaR2A!Tis_nw7!4jnI~lis2MF%hBFq+#ccaync>jMKQvu^^=0a~~H+ zuCnX>r@Hl*W>>u>TxWkL&sJ`zA&b((0d7I2s2f9RijdngHwX1C^;70yRWB!Gk;G9C?mU6&-W~+8 z@FprVk%u7M9wU39lVQ`>Qh!-{-{;1S=M$|{-{qFEaQ`!luXpcguSNBL zOWDO(K&6Nd(j})6kW|+w&+woEYtzxXHQel1m73YRxMJIyVNIEVZpvWAw?Ql7jz1`< zp6K7ki;^7GD?{|@_oH87?wgWa0K^2p>Tmj_WD2`{`|*R?PXB3xK8|e&^gm^(jWvJp z#`wW8nuV<3=@8o^^^UNa^$q)c%J^P_Ly@55qc=zOV`qjGGwiF_SXK$5&^5|%MT87l z`OyvmFES^yEDhQvE@Zh|ik$b&bW$SP{-8iA;E=pEvQTKtP2x{Qb^Hm2qr*dUAq|Vb z`Gt*N#{eoQu0u(+Gk)hA%lqD-#}`@_azW&{nMFp4RK@|3apL@231w)I+zl)Mrp(RL zePOQ8lc@}(=S8%{>$ri=_~Luj_NFjzmc>G?|8pEaPFCwvr$(ZdV-(Q(p#T6L_`Cvp>cL@ss3ZkoAY-1 z$Z}_s&JGlNWIGVYz6w&0$D=7xcWq=`=q^DiUU~KCD456t=`l(_WY?C?5|Bly$V6+)%EyqBMDL_9x-B1JHDOk+Yz(*cNO z0k^11;U6%;Hynzq_ zlL;pIp(4fb!*Z5baXUzJFp{9Ut%Y&4I2hv&iDUi}c_k~C zK^*mxuT*CMbXnag!p2sw`7Py5PBW;Rn*1OH;2L?$;=lDOpdYq!pB}s30G$g{1YD2T zcyUJ=l4ib?pN1aI&@bf>&*!bOF=ZgKCa3>)yp!UZ-piLh9rDo$>i)ZVGJ8S><*t&H z!gF)R3#G===7%F;{+s+rK_svLDD{+Ws^isD1Cz_Nl=Pcs=85>(0_ ziH}g>j_owgndQ%RNGBALoCs-_?I;Hv3zX0GX|rygH9 zHW6JV`yC#;hmw2dKpA;2#SNfOS=5y_`d)8MI%YIQBzV5mhOcvBkzElNn8z9NsXAvV ztH0`Yro@ult`a)8k3d{K^$-yWi(jS9z89W4)yi$ zgnNklm78@p{WZzd8U~8jKr5@0Lfes7pc!eQhRibuvQeJ!!%v4|nxVDdyg|JVoN|hr zwP7y=gT?2Q@rtZsNZSR6yuKGu6ruluM3-coAHeH;Zf-a}^~N(hyO^LGW!GLFrK!X5 zzWl-G0LH&_J6hU(S91_MW@Hee`Ly9e zk1}3I-5;a}8Qo8)hr6U76_EE`gs3X%npT|6915m}3oHt-P7VUinGDGvaMP&@492*2 z=x0iUKiB1)ml9}xh0*j{PVHO(_a6%wyWgdp)QRzwr7zC7`UM`IrS>q_1(lu|uaP1E zxMmdJT?xHjX9?6sr@x;iE%t`nXJJSL@^A7)FiD~^BzMVU7*Pgpz`$gVt~KbsH3lX2I9{+7*N|K z`A6)dR#GIfL^#mztCq`5w=+8GI{}0$1y`9q9P;%nP}Ug8PD|)BpzA&+PXZ*PM5Lqn zhkS}p-kNMXL$9~r=Sk%Puf0EScO+D#pWAAb+h=$QdQ8Y+GL=n9$#O!NVxs&^YZ|q( zuI|Hv7_K>FrVQct`9hxt|9S2}z%w-P`*ce^15{KYN0jP&N6ay;>!k^kqg7^Dp>&7Z zkZ|XR`?g=~wxuyEp=r1rv+%@~Ci+(rmUvCDHi<8Uekdtr|Q(^DSn|c0E+=FEl?{qhc06q45+Mmx^~Bi!9yRAJU7-# z%)M(E^o;(#002%%ONv-5ut+wPGkjo=8of|XInCxZD|IJp@c9;gLCYNHB-^34=hYP6 z<1;O2rnj=5Km74k&{Jtfl>q?Ud81{XYf&0pJ3C5+l%)JDa$zM7>doXyq9brHcdJpY zB#&jb{-II21~Ew#6KPh_le4sJgySb=daf(|(ykiP5adRb!Cs5s#Q5t*#B(!X-NWEY zrL~)5bE;jL%_EW~q$u>|B#zJ0hTKmqod^6|Kzb9HWX3Z*Z-G=3AU!?`&H5N3$m*s> z>LJlIOShqgeU(zZ*n*$=+8G0=`??4NW3H={s7vzMNP$diLT!LTWs8jTD?nV8uMr7) zA<^w1klNc2uBdv{BNS;|umEgM1=`OS?H_@pPuMnX1N!lBK{^#az-*?tW6NYXRq{Hi ze32E+QA+7<)q#&6p}b4`t`CP%#FNvqEqJU4KZ{Oh6_4gZR&q&Av<*?{E}Ai#b9SMk zcjSLxvhmqnBYP=U)0OjjNKNstYN`+ZE-Anb?VMH#xIlf=+PX91Kgw163udHee42UG zI+}iAX=d7-&~6l>JD5hc(mo6nyeiI^5GynZC?EkJ-`2vUDG0%vwU~biB@VvD(o;CX0y|$Q&02r{0T?hJ zA1Lnn{xzF;S*}?+kmDx2^Ea!%GXq)_Y|_>Ot(T(C@WC(8;C|>T%GA+CuW3fZIbPsW z+SGHymS>ZYMgNg0^$re!8mOlbB)r2KIurqF6fYWM^L_kOZifKw^6?|8po8=jbA63x z7`}VM{IE3EgpDPu`L^X~2Il=+{NFS5u5DEd6t^cRa*ao4P~=s?OpS^DR|Ph9-ecKX z&&7yJ#A9<;69(C#cu{xtME7EnS#;w`EBP*LKfro)zd7lj# z-!%nO^n7dD^15{tNI!GRKo0dd)D)*J(P+c$BoopLq93BG5g;Hv3Jbu--0y?LlScbSJo{7fCl3`c70S@k#p*F4gf*N&=800KNQvx~c*dHpwOIqqe`#tbgQWMZcwzPiQQs$6Un=~c z8U%9|q!PY%Ri zj_4H$K~>12_83hKvlrAO(<=l0l)>%RO>x5pw+qkE>g0h2JjInH>7_Z(wL8Fz<3&}x z>qzi}C~%|hw1bUaN%FL`52R;La@t{(TX5U^Zj|fc<_k$|#P(a0z$MkBos>NAd&_pC zOA{4k`t-2cLNzv8IkT@Hax5l4HO3n;c1FzEj`E4utHgP|9l3Hf=*6V4bM>K9ceS*V zA{ZYce~x%QLyd6uh&L8X9134bO;Dbqmtcj-7XJ;^9KxH2ox9fl6@FB&xdsr?h@WFS%3BVwa=v7UqK?dCn)Z_JI~-~G zZv4KfT{nF_%o%*kOvC^`S<}#SPQFMgN#lEq7(^8xO3-`i@cTbQg{Eem19IXNAH~1lFtL)-VDsl-o+nH*^v?rdKB%Be z{Xs*Njk%dn)m#$8*ke&(K`rx78md8_9mrM0&9uUmeiW0+mn+7lpVX1>K)+jps~QFg z_7C4FdP0TzErGS$85xTfc_GE0)w{*H;et1shDq(v5b4-wpE3{!9+w$V;LAt5i3ZTCRc?(2dwhQpN(4MmNhjIEafSu$JW;c+OF=tYw*~0 zJo|f{siVaqkFCQjJFmSJDJn#t{T9K8NwVymRGgB*@iez~8Ir6EtKitRMTsnLkX|ih z311loVg0U0+2Z|GVfQ*N#Jb13ANEs*;Qtuy zwWSD$w}(=|ttI1@1u|@)qodt{l(;w(8(p5_3(lUruJH)(d4`M|WADC+J9t3-SlCo| zj4M_FZ6wk0L)*cOZRu7}fsQG~>5d2Ei@G&>6C|!~o4TIc1HdQ^%nO$S= zOyVNoNXa8$NRa+7!rm#ovhZ2sjBVStZKsn?I_RWh+qP}1W3yxSjv`72y4l}XyQW2rl>0=XA_N&~&8Ubc^9Nf-QOK%DP;0*6o$h;|&J03S(Su(Ci>S?E*)!944w+A9+7{XHoB_BHX4YGFVypnAnrRqV$8^sC`~{?6A7Ih={{ zj&C%C7K?<=stOW}OTesnfQV}2xZ>xXz4lLiKEXIA_M&>% z26wOnhhj>LU8UpqB%TrgElLj=?>C%Lfe=`-*BrvlY~c_K#wG_s!3DO{%Fo21!?_hn z3f;dMQjoEWX$^TZ<2YDW5x*3Pna1|TJu^$(uq8SVYnT4@A#C7~G`B;k;^xza{a3@! z_+3yQ64tEy4d8O~`uS3JJpoYL%(LjNJaU9*|KTEjWL6^R3g{u*z}X8T=8U5ZaZ;YU z7V%~(b&R#8V)8F5^im_H*v_0Azxu;C`D0Xw#u^u?Uq}}V@uqdbD3;GJffqR5@Gqu& z+EZB^(`R*r=TYjszp5ovUd=Fq^n9v_Tl>_vf)jxRul(jBa0dQ{iqbvF#`|k9X@ah9 z#jg34v4{6)?8152>1$%^hOISYo}6}4utkuNL;oHh+sQHG<=y1}WdtvuGO$OpI-iJl z#$tl>1?23Jq5`-(gT#<=y+ae?GDRXGh)@bTCHSsLw;gytpd2@n@Uag6+>kFyk1EVR zySFaA)Rp0b>BmS>&(;%FTS^hh7FR;`ou^Q#7V$j>osL}Y!>~g zNfJT}CrRS#PbtIsa9(1t@E`9|Nl?hL2HtfInzU+zRY)xren~22B@Udh!nBI+=P>U6 zBCA1i*w$aGNVFElY}KC54$AAIEkQY;We}F>P~vJT%0StzHAO1@ENL+}a`fAP+^?+Y zId=U1)Y(6)-uyg;O^rV+Qo|<~WC9z#Ui9MT{>brK=HD8MuJw@atS_ikE>Ge_^2c(x z-6?9xluJM;!d~ctiG%*(YQ&LPGddTAYJP*hpc8*=Qp4gi{`tO~i4-u#|9#Cyurh+j z6kwUBBis@vVG*g%5R7}oJ50?PS`jd_UZ_924^p9fXbnVfgt&{5O2^;Mq_Vp|L=jCY z!M1dU{~9OA(jw=NMNjHoNKui>V}xK+lR5cor>~DZ^M!fN{vBiFXDcc%YBMDWt4r=x zN|gN9{ny+JF=#~Ct-_L)vY|nlYFUw9bp>PbHN*c^DWRAyITnHJ(e$M`<#y#tfB9DGD6 zCmCZ^q$W{uOCA4FGJmt1c1sIP?i=;-^(#Dl^u0}XBMe3jq0t1q?M9Ydp#rR41Xfgq zfv`YaXa~fqRix3c0q6s{ST|zm8D{!H;KS$_sG_NZYwQSnG>a#tg-)CJY{Zz9FB;A; zges!ztv%Y6pDtes1a_zJ3-a^9%VMdNL>Mt5Bo@pauC-#uQoo+aLq*Isa)~PGkb;8c z5iPm3Lol8~lX~4=YY=|6kxnP_0spagqcq4u{*MPnoc1!OsO#8D{%^ct+%4cEdaXP> z2RMaUt~|}-yDl68JVa(Ks<2haNEAE%z)~ztgR{Oy9ul#b^FQ{gEE@NuUbho>**uZ^ z(40&mhJ2r0pNZF#^ebSQKPB(V(qd z>wmy0t&d$j>2v22u9tz%Cu_Od{pfXdcOncxO&k;;wzD-WMJMj+k%;)7(SBNq_f6R9 zNj)c|wfNyQ?5U||g%slT^^a+Q+w1?IkBuD)FRB_@x-!kRjoC`dTVA}0uvyYM$SWL9 zIoTyzGkeQmD;v%fEtghsmjYL%BR6b^B{4nq?H0)@MhH}_cVaxpOX=tZbpDrp+LF8kYpe1SFCXbv{ZF-&X{!*j z=EDhlA4>O~f-nk-*<*3k=qyE@u`oAJD0jO_>>Cc3r|oSPGMiz0uNqjW5?%kWp8_1P zv7~z5<7?%3s*8U)zY#uWvGvz7U|`70nwcr~&89Hh^-4%saB>YM>Ms6>5VEBECmSw| z*?<@!QgbC162F`V(ZCpRaQ49d@SIu`kIY8E;cPO4Pls$&dvaRFUeOxF54A^^4{bUs z^_3$MOGTR84gX^6nTP|9iCS)H=M?9kWf{&SmQD%pU_Y@UTHgxAj>Sc}#aT$gE ziru?HYsk|GpJlVo7$1K7jwf8#iBItNDJy)qaO7%4W8QG2>uObi^d5WB>*U_8BXUH; zhL7u(^n=kQQz9SucI_8qnxBUH$GKj-Mr-mpdubAivNTPqwRa?V zF>U0=;Gp^;%*k8pU$dWg|4A|bYCZ`;ybat>i6b>sD|69fl*u7Xm(Z^XVYVRdvyhY= zYCRfB4SK)r72oWM27K%OEKe9^Cn_knKh!Tp_CqOt`1J)b>k4pDqaEO}Hde_Yjd-3@ z;X=($@4P`1x=joeAC{tWV4M9eN}d&`Qd1K`D%ltW`CDUN=Jd05lK)*DCrKLbIVP$0 zf#N(S{>p@cBFJ;_k*XB_eE~-feX&9P#;a$FYU~~j;7HfK^|acEFGbV%1wl-TQ89dp zB3Go(0YoOxeUnPH;Fs}Pk|LA8YctqTq}=(wXFTgz@7IG4P!vpKiEY;wk`W8#i*IOb zMCzKhnPM4#L&i;FPUzr3YK$jmybz15cU!+$kh*}VoQFn^IO$JMJ(EqqVa}LS?qNOY&pesXyw}Js`}}T;`5p)gRTYNm zaSrQ@g`&BcWeSQ%Jz>zIKg_E^QVnynPxiJ}v=8P(wnV-k#<)cIH3~)=9}VEbXzf%M z;@UX1Va6f<^3hX=;4u&3urFeJ_rhv9{g30sII5=EX#~pX(~-H^FaBzZpn4PwW1Fe4 z{aOguPg;}?FT~Fjj<(&?Mz@J6&sB@Mb~n z5~O)RB39dQZ~IgKj&6$t`?{F-$sPQ#gW;gZhy?5~C7fI9t}hsc-wG~8=u^MhjDy@uopS7;shFd>L$mJv;_~`3tHaoPDh;csuZ5zl&fem z7psamTi2ZO0X@${W*xqQ>*El7p+X{Foyz!NE8=$(Ea%&v0co%DT>29e*Mu)4_438~ z{zNWpV%=mKzyWv?pp#b%+2_B<58UJ-{AirTGnYDtUiLD7V!4HjO1_|n#p;;f%L~vA zHX{DKogdK=R){WqF8wGi1KtvEkbrtEa$usd02EIskmh49|A@S66gj-sk})_*UHzmR3~0Yb--TkSk^}YIRHy}jkPw}3ybSK zOY9sz`sm#w3U{GQ{#UT?+S_N!j%yx*+yhjFN#au+}Bwz?K3{|U>LFe$>y%agFB zpU!#A-YUhX*7mU|`nZe~L$C!Bn80=}oE(NOOrMj0D$LCQn+E%fKRlt5?2uri^lUd`tD=ne*Y@?pKF0DxXvcHOgTlql3c2%(4@{6LEN!eGGknd z&>!);4j-Ui@p}hP%L-86vyasuatiujv{CGLI9>e2Xb3dsWQkb?n1BQj4V7DiK89>K z$)8KSofz~Oys$57xI4G_v9*{*1AK%yRViN34JB;9?zL+`awc-;yyA=3O?Q-> zcuP#YVzO`E+xp7a?uC_;xJV1pbz6Rn6TNaAqk_!HkCfpbIq~BJbfH5;zh^vC5)HDSad)>zpX*NGs()sYGYnZTr5Q?_{5 zv%WSiL(}XarZ6uZRQ|!~Q9x&sp9M4HD9A6{cJ05BqOQ%SpU!m%`|~l@`JrIdr9xSM zSl=+r+y5eV{0oIF&k0_P*uX)Zxk&DXra5^nzFbKG8ex9SW3nSwaj7_CDeA!txGsnd z*s589UCtAV)WDX4XzG8rM~0GBixq}^ag`Wys*&JM8B_CNQbInYVK z6)UnCFW*rIIk^753FSdJ{^sZqsrRH`^t>~IHj-k^T2hv+1Qmn8xNA6~Y{}B|jxvpw znuvwmS`jJnDDGUq(() zCcc=%T0U5vF5NqDwbbtGl#_|y7g6$y}I8ULHKBV;d`gxlj7$K}vK>jUD`Vx$p zt~tc1Usxu@rSCqLejCBJLkGx&U+b;i#~hor5-A-wfpx+Hy8n8w3tV%AQ`~e{W#?IT z*ECALb$3Gd;~~=vXyXd5LKXY>VE`O?2I0@;0Qg$EhgT@~<<+!$40bxzM;jIaZLCo3 zG+qA#;{M;H10;wIE+HQ1$@=u5*=eHV2nZca0ECPu!%tyr5GE8tDLH;NKM|p>A+vZ1 z4_#7WCo`Rdtbjr4dWERshd#jrmYjfAtOL{h&H&wH!4E1v*MG>i&}>q!Y@0zv+=+N@ zb?AVnP1_a*lLXP7-(Vm8FV8=-0C(J-UmmdE9Lv36)05Ve61q&?qfp>*sMeoBNn~+N zP%E|yKpL^XR)srgd*Cy<36bMxgmbdeW%41*X^4x8u9fB^tl&zPhhORj(zA@Ta1nbA zqKO@A10!?s)rdtHOcAmf+b&+x@t+%_Bm z2W}?}{wjfV_L!vJLJTuLjkz#z@~(crYqki{wm=p!iJdA%OC2Y@uFzz(g-J zM2=XtA+Lz%wrWE~KoYsCMY$i3*p?MqGx})a&r`Cy#C4#L{t0r1M2H`J38l=FaYy9< zi#M?o!0+Eb;bcb~N6+w2+Z7|vmiG}_;?X~0UXCAU@*%mS_Q0Jn9YSY){5!e3mwtzh z^l=7+UzvcdscBC0Da>YedPQ;`yzn3lT*~+>RfVbYR6gQIUFxjkM0h9W{^hTI6oWF~ z&#fW0Zs8>e8+ejte?&$keK~ElUiQ2?ZekopF9MvT(?tSwc=WZfNid3B9my;T3Y5AQ zzt#pIv(RB=R6>GSPD6guIl#wsgOwEOh}6e4@OTocuLI#)3NF!H8k+(@SRp|{B8Y?I%Y zBq}Bk2(E}==+-1|OBy3Pt;XOL5v|&Usf2KleKn?$TwR_+q|qYSaK%n3f(#QgEtR6r zrE)EGPBgvPdRpdFSAUkzjg!q1L{R_`?xLu$t=NA_?~~_t?+F)ht88I`|A93A1}mU@ z>Iy8wYZ*V;Yn%RnJ_wBl&y-;k2@hTUQ9)_a0k%1Qe) zv$opttrlW8et^oJ6!s=Y>96n*hLcd&STQ+m?}?>T)|O+Vu8is5@cYSpF|9)UHR;57 zZU0|w!$2j395qi1k`l=;dMdH-+8XL>tS?DZ__;_VH_+~&(gKJi&LpOj(tpS?b;aT% zhU2pR6x4|nWLp;YwTAUNiLWx$cPpTkF5poXWrQ*cMA{wWLLcL%+W#nzk}Fhv_v{M< z{kAoJcMhQ1;p_$a1b(|0NtQKeifTQvW#b-&JM1sy5FnI?^cl|X34k4BiJ+x4>_Ulf z9QW>NgMQJ?9{Id^X3l~feR|GZ#vZwjGFfx+xW1eWZHR!XPsxnbQ|t1#dC zrjqUvBR7EORim$+aoc~6#Y?@=v_H zPi(zt^g^2XAl&2_MRgK9=h$90pd>i46uULUUMYK`%ol9?ggj7mwNk*Wh!oUX-Ls#z$ttK?BT&MZ10_=Toy+ z7?HJjE0_&2{6=sl=&s7}u?uWfrA8}wvzsPKR>TnkkD6lF(V1$oRp!B7}_09a#8rc6y!3gZC0xe-z47Ebk&d z55w7h&&LPv$e)})xLxk)kyRHNQ}VaJ;>_0zL1q0A_A5`Xr} z$AooJJ}V;5-SEYlK?_R*a{D&Q&KTF~kxa}C`kMUzHF>UbbnIJ29fNo6=e>{A6AAAX zQ8_F5du3GEouAzhw>*A)o_AR#nrga}xI1szfV900rY*Sw`$S%q0wq!8wMxP{Ldq0X ziURl~+(!baK7WJ`xQ7y>yyi=yql_lw8L6l{%B>dBu?*}WMKl?~G=XLItuD`O9JjP( zeH{KDEP$%NtuSp(e^8!+-+^F z9-UN!9z3F`j0pb;z*tAJ7mW=H3H-l3w$sG}#pN3B3agD6- zJ6~NqPiSLpsFTZR`51Lzo=5iQKyY~ND?^Ef*qcm?;8wUWVa=zJ;-^IA=O?Xf2~fa( ztYLDzuF0C7D!iKA&84WA=e8#aa6~zDVh^f@th#qoJVDbV6+PQrdk@6o|Bk-OxI=*T zKKs#IEbTE^T6swf?MR5$2{b5WT|1q> zURngjKmu-(E{LkwQO1YD6;vT)f7Upsz-B!=( zP$&am;aRuX)^tOJF+e5bo)IN+$b|msST~&1z_2&{gA?JMf%N1M*CvC)X$iRYlYTEt zm>5wx2MuXzk74WD$XEFj&smWAvapH0B$g!yLH2y?-zsk9A1t;bP?Mar+fr4&W!+5P{B5 zLxnZM2=TM)>}L20gp#0(uVFI(EltY+kTvIVLtw- zb!jO64OM2H6ncD*G&?Z`*5xL_-z)#EtDMHHAMxKbTL)yf!8mH%-x@vH#1&XYRh~Em>XOgrmpI{_wurNueI5XT&qKLTD9x zWFepnn$5-FDc$SdQt#f5{?y*a>l|-2nko3)P5XKE`I!}j#44YOCQgbOk;(3>M+~}J zXZ91{cL+k}a|-AO;B%xxQN@|pUDE)5uS0&HFOy&%QrgeOS{i?@Ddrpxac)*-1Jw*u zez?!+OOY#0GofKq$Hs};h0Kf8H9H_nYs1gfFvApUmM7q-ye(LSG*q8L@Y39(tzB-@ zi|>|Sa}t3ty3)?(ufb(KzWi)NgYUuH@Jcg;I{{Y`U<9OikGr=`>!qBlOrBR+4-TaDXl+ zW|+dvaD&i>FJGBcUoC$PD+h-_yy^**4=$-Ji$YfJ;_`tK_*l1ZYthxHoq(30*0Pbi zWTII3lb2%39BD(>Y;6&Nc}*X=5KUQJh$sgL^K|D>wU={YmkM<0uYJ2}%}Hd!9GRgS zwDZz}eFrzvWQsR}5oG0o$0hETW&e&G8fjpf2xc@T&0y<{4?*RCh5DV0z*t0|NTMhJ zIfgPhx}6IBpImb`E{`uat{zW>cQcLLsTrve_Ey%vDx}$IouXFq6A%4+11DunXFDM@ z^IXhp6wac!?_o8PP?S4ckaPndY>ergISS$Ql<;)5$zjo4*yvL1Mzdz;6jf}&^>H2i z_^h6mHwv~?D)N>okxYyd`R(jJSkiI=5G+j@-+@I;yBzE(VQ6&Pz2bfUgm92-C3>Z* z_q;8pLZO+g615}57oB}}gRXouVXv37yP}tbKM#LXQtj2c^PH~@c`O~A1-KTv*iOgC zmrba&nIN^M%5GhkkGuYmb}ctyZbgp)aitH2L-0H3!X;|P$iw+yBMRr#pB?}`1}>?q>#v%`T^ObRb=YxD1*_ldP9s&iffd81te#=oBi-O(PB2gR7V z(gq*xcF6xde*drMEl39(oCU%HSgT7k7};Ha5sHQN_QQ8KuAewM_F?R}*aVD#w3&Vk z4qDlm$$NUq$vF4AW}Ix%Pl%d6mt8ks{guT8_svc4U;Kcxo)(~XAuKDVlRe~{e?}3M zK}K-&N`RvjjlU5v>R+aOY*tHWhmQ*r)RRfs{V6`w!wLCN!_%tb|YHg^XO|`oQO_bj0mr7{@&sGz0CS= zI2h^Tx>K-0&!z~-?rvQYKfIWx@Q#k!g?mV7LA^&sI))JOgub-b@@XI;C^jxy_*z+25`j-|qL z(7RLDK~uI$(BxgmgsBoxq(|A=?=phr(f?9ySw09ESxusB=JQqN?|UF0)5PLn2hTKH zF1_MJ1GaOM-aMQL7DX%0!0?&Tu{yW zpwC+;LiEs)9dS|K;WsYc=lwsf*Iy`j7*&(QANt<}Jw;FalRfBqX%*~qR&p(oZrtv= zztJ-0uz(()vy)YB^LLl-d%79QcQQzSzJ;?+c}Bm|4&c3CZ`kTB)}kf|CK*Cg=yxAk zc^SU$n^4|}e!RcmbVT$$?rHp}+4nEbYC&)3v{@2n=dqgT@rN+hB%z=qIN4L7PTyzm z{yyHD1{)X)AEC7*EzF` z(e$FfYt^H@oZ)3Dx?SEJk8rs2*h0y5tRp5Rr8ji9*Y$`Oy_5LxKrn+We&{~Whw@4e zDfXBCpYdiBYc!+312`M<(e$gv`@hac<4{GWQbwQorF= z_l?rgtCG9tpD##tOG14Ns*}%Z-kl%Pi!42!q$H)T{ zo*V!Zw1VDJ?38FCb()`Pst*$FATs0RAtoql@h6XAqz@DlIShsn*TN&Mp{8}HH8F^O zW@|8`X$bk5;3koO^CFwQbj>Z=gQPw2cIG7%mLLdFBJyh_eu}Eg{4r zIvOx}pt*SOb49S6cR0_0U-1>a={ZD0YX^C0HFyTjm7b_`LeF}-9fFIwEm3DknR{)n8rqU!9WC$L^8+{osu{>0x zF-!^v5|#<^y$$;je~=Y-w&K9@Ql+HJi;W2KOf;@P>S-dI3h*dYGap9*E=qf1R{uGp zSr;M;*DjI1VGZ*t{I_$sDt7qiS61kbfTR=cuepO285MX0uYt_0xXkW*RZ=6|)uPP! zDRRnx55Fn>#c4cPfA{(TV<3DOuLNHSQ%cVnFc#cmAsT}*@~)v?Yr^z$OtK@71t)%) z8~p&XsIL!p%>w-g|F%yW<(|~Iux*lq``#3y^ZN&6a^{S05vI2w;%cdb5RW!q;-R_uEJ!zziq@08fP3L38 zZ2DjxPT_FAR1d~Y|0LrrjW*%T!9^M6;G3bNAF=z#A1H#OlpeMsX&#+Yj)3=b8# z)(`6|X@Qw^f{>oF&71BQ_TxP(IPR?6TE|=n`He;WOP6>x-_e@zW(da~T%x|%v6`?G z2I*Lv%&SEI5hM<7agTX3OC_1zj!P)n-W?4t^B2NrzD68+`%qBEp>oE)h}#rifuqVh zjpN5p11)#ST~=vqOjPoRB68yt2;P(dpI0;}MF4jma%ctq z$YmP6(n09;FY(L~=vG^9LFw@)c$=T` zlT~->P9-~(acAO%5n>F1s({a1BeggmqnsnII>cHMi{EzZ6y@!u6plFaDy}~Ri;U%m z8lu*kr9Ou^I^}Q=N^Ng02~{BjUn$499&lrLYD4q+J>Vc;6vj>P$MUjO+*eOeFL9{F z=EHfQ?N~w+$inqENUZDo`jNlJcu8VIHH?ua-;|1_6W;TT%ShzIKhf@aY;`9EA3V86 zXI)}*U@h;0$Jz!x#E$)Wb9fr(e?IU3Yqi;KfPW^bPUk%?wyug!Eg>e?%9FE_d^{lW^sTk)08Nep@UxKC1efy% zwJMoegBt}Q-{IP}X{HvAhm~N96pGDj<}GU1NUgSS zVbq%GQj;`rWygpj1F7Ob7Ad%}`s&h9E2>t!on-DG3uX~945-gkaJp=Pq7|#`$-4L~ zzP;5|!O(P@bU7!c=MwW$4lE`a+GHs(hnx)2b(rqtV-ApbJ+G%#YmF>JUodn*{F%86 zQsxmp9GV%tBzLQBnxox{*v^Nrw(6Uk1c(j|G1E=fw-<2!%8_s8$uFm%9@-4?4`dG7 z^V81oYh8{t4N?~g#W2Jex)I{I@6M^fj*+9BWsiEOc&u707Ycf2s{T@b1OC>7Ft#3F z;@x_JDM1;v$w};z*pUKEQBlUG&#MZeH(FY=2VPW#19)kMV$|=4e zx}VX+j&D0GM!2ZML3zZ+^#y|gdgV^`15nV=AK;`FDVw@G3Ajqv{rd%3t6I-STiT| zryYYkB`4aZz}sL3I^Os-OXfh!BL%4srvLnQ+HmHMwo+fY(&t4GCIA;WTbF%??jzG* zy8|+brr@00ww>0rE748Z5!TY^=gPC-X$^4-XV&v?d-cU-Khx^jLT?>a&O>ZCTNIpvHGcbrLVOPek(O8PV*RZf@f zjoW=m1TmgR6@*0?aw?2#v3oP@-KIZMngB4dsM?lOb@ts-af_CywtCEMGsggJE9QhL zVn@3N!454B!nK%Un2i5{HR6Lb>%O@RKQo{`Ci$BOyXAx@s-u#}vON=BJ1j$=C;)II zSyx2tXUT0~#BfiezZ+^2;lO_9D|L*+vJ958X`K^>*iuYO9JD{Qw5EPt&)$!M0b%>QR+wk2}~ho6ELJ8(JtYwEp|` zA8Wn;{bi6Kc#x?8qZQN3W0p3?A1Iu^A#65ba&dN^smc6&t~-(302eevV68b%VwqOu zxSXJrOvzQX+2dO)V227ic|z`hRI|~L3^w`X@A&dUkt0<6Gmab{f_roQc+x6E_Cg#= zHw^pnK^7SiHXI+vQFXEnjB{vz!9I^x>oO_^2J?ix1r-4qN;M0MoHjZ$eCL$v7ey!v z;n7!ue11$eiQOw){JzmCATnYuQw)Aq-8dx==f&R*AE0h39^HUJcGn_CGgg!4)?bTJ zMCn~zV-{j7Z)*qFN5Ft#hhb}|0S}!`1DhN#&d$OA2Ya#ib1sfztFB5f+K_JiE}Y-kxy9>tjSoN z`zx4nH=MtAsR5<2wB;%D)_z&M%nDi{c-(l@*iJi6{6qcQYVpoVr@1KBX25z<-FIUm>*@ zHcISLY?FDuoX?6@hWw}Y`zJpnkzHp8kGf%Vin~@g1XhL2;SJ_e^5E(I z+Ib#{`pZ?o=b@CHFrP}KT4&uZlmo!Q>4=+8DWhfFA}@HaBz}UJ@K>+!2!)a6RhJY# z0A`QBB3FD^4@hUqX^cir;J5VsY3`WJR^LVr*;bin#f z^k_&~2``-D&VoB55HVMwAx@mnLoq7)c#LHrK(j0#Z%sWRe|%|eoSDp|~*V$^WLCgK_tdEfHx`#zB(!G~tovMH+VD@;@DH&`pNEmWxYWw9Dp*k-9}% zAR?#%iD(b&Cy5Zr>fJ%x+E+{+_SqtQ?~NdDTo``q*ksPn9_`_V{d~fc(%gP-tyCm96Y<9>`wLdLkDs65F_J#eQv5P<%2ZH{*k zBiM|N6AcP)wu$zK)qPX-!qv!j-ZFpsZp2edii+;eUcCg!;5o3vhD*xzxYyVDY8*;p z!j@#-x$j!|2y;Osc&%V6M^wD$ALP|w7IPUB}4jC=T&!Ztf8b|`vzE!(r z0fn-n1HdpQS{I_Fvawv5YnjK4?|@ig-xP0c?y#W>sDaeF;SDBMIN zI4&=cbW^GFM{*~G4gC4tyf_&KYHg#0ny+?{J@}LOKhlE#D_7_xjuk_oh*zvcqrmM) zTo!As{dZ_=F^>h-vU6Tcxp+P)UXb=pd>K136p4d3E$L!d2N5Durq)NLOVnabhX`D# zN&tY3&4cBh1a+;IL`Om)o~VHjNc4wC$nT>3;Go3D54bBt`K1{9QxvtRs1S~PV!)JL zs_Go;a-%PE^^|ztdfv?Bx{<0sRa&F3CU$b-Zt?tMa$fma>}uzNCM&a<41N~ODo7M< zxI)Cet^S_SNZP$jplD<^rya%A5F>_4qXDc>2Z5pjVNf5-n4L3~S6>s>k!oi1M^cy` zkcC>q86QB_dkyq(8H&kEbYj0=Jo`xHa{BjtP4??F=Xk~Gr6&{Yu##$CXGk%%K>_zW zO`rKQ`A#06ktX^+(PeQfmeU|O6fP9s%S50b4viWR`O(B=aDuBb-$_R!iE8C?Rswl9 zGyHXYmucX*{kJPM4`)Pw5rz5rUtbgz>jtr%3r$G!9B^_W#pQ*HmV-E4&Cr#Y_La6y4z0uHa_&`53TUnY#!|4vrgD3Zg zaLGCUV{Ozvgf2&Ex2Wum$m)?RKtSphS_0PY0^>}hw z+9(5|etr#y(_6*%09sIqO%=%>uzAD#lB*TQ!6^t!zZ}Q7;H%~BJ zETBQ9rs)Hw(cNlT%{N}$hOW`NcFT_Ubiw)+A7tiJ&caA=_ zOtBq94U@L-v~(p5!dx_a2AN6MXg!8a!-wtcTMJ}kX zjO65_<9B9iC8Tc)iKVwRdx!gNo=AnAhAfF_(93(kvTJ{b-rXUd){tfut z^Lh0}IUh&P*?z73{cK>OB95jOWR(INghUG3n*z9s8 zuvK^$aFm)L1lu>6q$cP?>XLWkuk`j7JMsRY!he;ZJTkwv!kU7&=I3dX_Xr92O7-kuNCrF^RPzZ^4Kai$=%)G zoTQXngYOg-CSfK5K%iWhnF3EOCMNESms+C}M*VX5+vqcJ2+=%B502+&Dg-DNXtY)8 zN10v(!;#G1-tkOJq?$%bfJY7pk0NnrrF&wocT;u2f}Z;D07tno8@H7Ri*(AEJD+V& z?I7F&xziyRR`TgGirL?L+jUarBezC6p(6>dM9-t=-Z~g@WM5lc32t(ix@niHbe=o> z_bdf4;w?OvrY&tdEg}NvLU(oX#BLEbK|9D8*bn_I&bWx0y!#ikj|XF225DbBQJR2z zT8CGfHv1WUk42U%3PzAUvxRIp1={86w6Xi_f6|=)ca3b@UW_aZ`c_K@VZQr=hp6Ju zHk)l<@w_U^0{}b&U>u8P+GND(wm#5!pf?tzdMa&D3r z_!)Ku8H|e7gk`~n{+?PQ-?(=-ygh16cDA6ef%`CzRKaN*}dI%C+kGTn)(XA6J4O#pos5h=jw@AGluc z9kxI}-I9Bp;i)%u)p4L2m)dIEF$%~U>iN5p$?1?m-IZ~288$%)oS)RBT>?-X z%Q2(B9Ke&vvktzxuBWzo^8D9grIHUEw@dG&Bq_8aq)8Fyh@859AlAAO8__5Z|C0}= zEmk7u;V4azZe$Opf?a2NYX#mROTL{ixKv-DnZqLet`1n#<-a6%ygeTW*rL?^GZD1x z&0prwi2N`FtEP8a4oq9KRnh@pj=kr|dJ7M?Ljk2g31upihYmp#+&tJw23Y*M9#87K zjQgw^W?a_9ZSJK8G5VDzLz^~yfTv%p-;=4lk_mYfueD76#~Ah}_iCdu10^YAybC$f zDeeLU=|fi)&WtOs@z(#aIs9)j*%~f*+sK>Lj9J1>v1RX8{mp=MWmO>O$9jC*Q-Bq% zR5WC%C-|6qR&tIUH4efLb#~jC*Z5bC0y1c3EV^f;pX;ofqO%9e_&?l~QPe_Z(eZ~B zi|ZQP!SUx_OfRAv-Jp9An(|dLTbFR6@`4kp!cs4@EtbqM`?9{&G_{Y!sgd0T`=0NV z)oKUuk^W@ObmVh?aj@(S001efbJce#!}>A|_tKr=9DdzI%14>qS|Z;xlZNeZtp=gP zjX#&~ZBSqKgsJyff5DeOEgeW&0&w_5`^4;|2!CzAF*qjMnBYPP`Z0HHDX$;AKzu$c zF@dAQK$jlK%9!nAZa_5T&VQ1P)lYfBG2SvXsYBl$f=4PJ8&4VKQH8?+I8J2^8# zG-lud$MNRw6=&=(uK4qeqr|^(o=(9W1?@2xo^=M!W9CE!nG-%X_(y@A-UyGDRm&IS zTTTEU^+XUW0$EmzQ^oZr*(+I`@MPt=HvT#$G2J9K!A%mdQ|SK5vK;C;`qpLyE)veI zTs^#kfNz}i5#6DoXnewr<_bF(ST8xV)HJf1Ci75f@)|!_K7Gb#g8M2M^vp5u=s@%t zE`jL<*E~JEPSy>6bc2?VV59@=Dy6GFm&4arLj-8wtPb(co2{^f>Sa>8%w1&j;YEWa=1&oB=(&iYU1v1iN2dH01Nb>qSQOPaon#zgrRaChd>Cef#X52ile|P z??}gqm;uj=w$03pg_Zo^I?K`3JI0&6%1>GUBh-)Rg;%^Vy(wVJZG#BSZwax>M5j7u6#^mT4}x)KZk6 zp8d7eLv#0{lx-?%bZcNX2X|RqPd$7)n{CIq`iB$a{B%w0dmp4Db6Q>PlK`O7I+^5> z3NqwDZqR)D?Fb?9=58tub?sNkERTT`Ku2P0b6vnv@~E{~nqy2H5WMoYsKJq33A_jw zAPsyaJ;REQ67oNBpJeuv^OR2MbU{A#@;EyNHNBf1CN+_y$TgKhF9(9~Z^wTXk9 zD{MeYR5X}uqzDkfe$%p27`FEUkr>yJ#B{&36B8Z`EH^;N$%iFVcu;XUR z+x6i5&+6y4l6YXu3w`S-F`NjYd4l$HABB|6hDc>@d=^A14FrW(F_;;xF#!uud8nTS zfi-jJ%$y($R|_hy5S2Qh=C0f9?8aowa7JW^dh_&3&9)PMq6Gv`OPu$uT}sFyygr74 zHy1i~X+x^Db`|=dJh*_#hwVwuZ)s=Ff3%iiuN)yEzWN|INge*ff7J}}?RBaCkiS7RW7%u@X z#4~v;LK?Om^>wE&&_p$NAm6rc)qNf3-%;ubr{_nmx5H_|s1@KwTAnDLBA0V2(E8`0 zxkaBh8Nf3_&=3vkW@PgRIKO>9BW-d{bbA!1OY02p1zKsr<)Dzl_qg${KHSe+Z1?WI zW7Fr=BK=XC(W9AMhF*kSDG`&K3FPE-KRGv_iR#No`pX;N8o&2<^VlN*b+`5vn^pdL zu?fxHsXDIzPkv_u%hK+ao3j`6&q1?W?-mAnD0X2UH>%ob3vdd?b=A?LsTA4pDa3f* zh=2q~6m`d`XwL%Yvkj2j7Oj3ysI8tFR$!royAbW3J<$xeUkA=py-N(}xo)}0DWd_g z4~kr-w6xDnWqHW|kSD33S}%F%D#fG!$^m@P4gW|t(i?N->!Fai?yiA^#tAe&{mZVIn)>#xS|?C_1gl@U<$6M?-=1gQ z604k|j^~z;FvrHPX7Ow6NI>}HqMILl?PK_>a{j|I*GrKAQ;>;lhiGBoMpnLoF(ysD zTOG~qrf%EL11iN|toOm6iRG z6VO;h^vJSAjkijRuhW!T%^W}D=*y_dusKYP2{Ndog(26PeNRoNOKOxF;m3#%UK_~i zK&Zg>Y0jdaj%!rZw$|H$Ey^DRu}jXY-vw9@AkkX%^6KUtu^%C+hG#95Qst4f<^_S+>=jtKdfKB{D!_y;$l zOoA)kr;Zlu;%mgEtcBSrWtW44=iqsxA4&*WOSF-s4Osw@N5D3Tfe zVoH%GW9nk&va3`tZvds80H>?>WE|fNqBQY_cz}6(Le8xXN&EE}*A$JY$t}6gt|tL) zljNJP_#CSHI?C4|d8g*txQ|+D@})}CWF|{+)P?RHLSXI<@}7~U64vjtf>s7B*uk9! zXsnN%(Xk^uuSss$9Q9}u341$8+6Us$Di&%2PG>t10M@D<3SNp4Tyn)8$yq70GUfhO zlGk8lyGd$22uP0e{N%xSAs0~~BBXEd_HO^m&}x+>?JpaHh~nc#U9fG4XdLqFHl#54 zLD6FS@|WoXoa7=&A`x0$H({_xjSHrqAmq;3@&7Y+{#U`-tU=aV)O%QsvSy5wj4}F> z)zh=e@MrBekL^l-pm zY3~&_lv%=>kR_$82cBXtMxz^n@esu^C_bV_q-R`Doqgy5FtTTU=x$$^!>-;9ErBq_ zH~}_n%r+`t(LasECF$9+J$Bv(S#{w~^KHz~M`_|C0|)z}^8J1?@x2tEZl1WK9t8}U zIHmVv#-j!v3l2%K#G;?XR0x|3A_Q+Mu0^>1tO)uXRMW7~M+9cw>Wv}vWuOULa~Nic z&lmD->nZ^H*w`hX`;rG^mIBdT;WO}Z}uv< z6yW72&`}k*MbIU6z?r+OJAUGBVYTLRrE4Qgyw4?pG>lAmd_ZC4>vi!#rR^cZw!)cV zZwq7Jb|1+=9s@u`Bpceen0}QU(Iq-l9r44|f|6F_ZHgtPF^h*V&a4X>uw*?4>hQOui3T3+7%d19n zKqF@;B=U{gj zKUv`m*dJ}FIC7Nw*0P=bFNBfBY{4@66pf;2u&9352f78)+N09$*{xpP%tI_4D^o6l z2qkJNeFXVXBj@n*gpBN0_^~x+E21dCl-mNHGIm?L3)5R-QkRef3zOtc#$vGVDAAxL zd&^mr{tcnrM1^-daP_tMOqq^|kVpk>cT00~Hn*Cz@PfCJ-~&0317kgvKzpfv1}8q>Nh zkX!!Ans+*@*(m+a#Pz9U9>&qrpFIT{?+o_I$*ng-MpI^<9$$Lm zN!-taU$^ydX)osU`6@9K+JRx_-bkiGFRuI2D;CxrJTlyHFG*P_ih5(F-Onk|<;5S% z=YLN!wU!6HM10ogaLkSwZuv)}#g`f`8jpcRlILBj_9X*%;NtZH+I(-s*;Wo7t;Jlk zAZoTw<~jS#YyP&*>JV7+eY|DbPp*zlvoqaM8EwUf`uZB6P2ke2#`+j|LK(AP#I;$K zKbyn(44OCu-JbckE=3gL2U&<%W!46knq;G>#v!gF*DUIU2l=U^f+XxcU*uXCTyGuu zKPd2p2BGh#Kke$0Cj|SVKvtPmN&tbRo+I?` z1vob$;7%$T&*PY<@Fh7mm|Vi!%K0J0dhbS(@D22oJ_e#j!)^Pf$nnuF32M<#1jA{|L<0rqoXG)P~ZyE)fL(Of>5k6!~ zB^H3k3mUcVuqdKqV#o9Lf_c_EIKUn0Lj_gHjn4g^&n+%Z!4)_18*Riag~;a1lhF({ z`iY02`NH5A?08W3WB*Ed?&aGZE$#CuMyK)LFV6`*uPfJFJy)U)dWn&17VKklfk8AI zW=L?L;-Z>T?@G&m-bMd=Gj-db=>?i5>U~7laTBr4SS^iSx>e5{J8B^-_HN^XYnRa; zu82J86}C62PbbVXZKuUUO1iCfL~?aT7{$^$IfP0HybU-gvb~BaUKOih^eQ@6cqpAn z^EpI;&dsI@HPE6(Z*c-ktinzCbxU?xUFBDV3KWGrlY0c&_uGk^cl7F{{WVuCOGMND z?`3K5+~MC(E0H_QjTK|^dN>>ySJRo&kW73pq3_ArbFg9gL&qn0gcHsib1)fIHbgdg zqKl}UxfnrUI*T%BE6vQ}jwi3oW*ijRqQKlKT`_d`hd}vF-h*q!Kx&#tXTDTIOnYyXD(x!C;1c(8WieB|d^8S{zzb2v3DJ>^VV;J;>FK3_9 zmA;0}$|4oZ-6WpQrvcr)0Ku$_&<6ZRV){w-pxp@D#AeCGi}7*tqju!<`q9jlF_JX# z&W#rKCB!pi0^KM%)J2E#%?W$fLqqKT4AK9QYyW#Rv6;@6d0GgErrHRGdy;U~fy51W zTu)X1=aKvrG@DLt^reVBkuF-6_cTL=N_~_>(e&Us!o9W|@q|hN?8Ly7=GooOc;?AT6FNKx z1NdRd@&R zd2yT_4YgI>rQaE;q>bWFOKdq!P%wMhfZNwrPMswq!2(0+HVDUYmkt9Y4X?vLC*Qs$ zG(eUn8g8d5b%JdWLn=m%4iFeFopHZ)31m;jpAFydAToMt$U`D_8{KK&@BWo0JC?_l zD95zD+~a2sxyC$di=@Lo6Z6s3)<~9LWu$TP>uF5X-pq85hz$Ee6ckNoksQ2OBA)aE zrAMHMh2wnV)CuqvX^zjZjVSjp;Mu2RvY=t+I4*pT`TTnLp?t{u)o+FH;wM5-f9%{l z3!>*PzYmFiur$xBB(d%;iU*ryp`VOVcZDkon8ZjCyJpQ=bfV264$NJz;HMQ4S?gjI z9abdlk7o)%>Gg3uVUieXk6*jnosP`EL}Dj>O8s1u40Lq*FXqlH^NXGni<1k*w0*?) z;hDj#AK5|t*Yx+&qu@lb;pnBT&5JSQeEYCTzoRYh+~mmdb|-+HOp5fi`T@R|G`@b! z>LS_1%Sk3HT<2aZsGcfHc5tZfB0Wg|z2+XAm|RZ|ci~W)Jb8cgX}-7ih0H zAJd#m++h0bR zRG5eKDW^8Dg!?sT10J9xT?1ZR6liP=B9EnK8kWJAQkYnPV^52mTs}%%EADJdfGfvP zfC1yA(C0p=+MHSfT0m7L1iByfq(Ntl=NJE_rtKrWvn28(_X7{SrVX_A)&lh8c$J#f zx&rsmhg^SK$G$%~r>eOi?xwtQ#r-Qh!u_6Ul{CaXU%p+Sq0WH^cdYgOMlwgVu-5M} zO8Kw&i9+0XcCph(-r# zLIvLTV_9jKmbl}KyigBivmjfXaw+^0EI@^n+a_ZVo}#>$M+T7k55_kd@V?;W1~J|e zkS*)a&^wrWJk5G;MAwlEE)a694MBJ|!0>a5d(KsN`Frg|mFjKe>>Aj~)V5l1sQL>i z60RmB_@K@6r}Ld%N?pP+A;hV{op>YPO^$8-+=zPLUpFfOKE-22-8r#Qv@J538P~zD zkT~eJoa#Xy37g(*3*6N2JI_p*GzXov%pqj8QpqOW`v-L`=AT6hw-}m~{{H)GIxC$3 zZiwc|PH~;%>l=uAV#g?D$u7I>@y}1ELO`}iMB5_HrafLluR)P{T6s@9tHB#2!u08s zO{WMeB*B`ds}u-`O1Dg74^kEJT%~w-leMSC{XxqJ)Pk>t#B4j?Pv!kd)2E=VQZt3> z-R$@d+*-#ob!inqdUN@+?vM6OJOq^Ka{3?(5(vEm3gpm-p8J+)EwUP9ZW09;T}7-{ zko1HkBS?hLnA1Uzmi{*i76!KFzAzl2(=8y`FeA`-a?-mws$XcwD?SS5zN9l&NhZ$} ziNW2Rr(gD|S{#qFM3NPwFG|;n)|g`8qiH#WREcB&vYfP?tqh8b+Szm#w(M{2!W0D@jBSnxT=nxyicT`t+ zMo!pzF6J(uG`9EIBbpQY-9y#f;qVBP-)Z)r9Q;{e!J$Fa! zsO739SmV}TFh6M(G=#6S3a01WY7-`< z@j>r;i@RD=LhTo=rfDAw{AjfD0li$eH>xzm)d_v3*M$$R_{doBp@-&sdaQdHv`4c4D+SLzh_?0{Dosq7CL7r1|Y4_GR#+HU2>+cG03{&V$3pC!f`AD)`@LjD$$sq zY{BCZi>1spD?+tN+N;sJbgX$Kbd0rTr$$xtCM4cLhnAr~F+7@AAV6;f^VOAo)+`R@ zaNTd5>x_l@p3}3VQWiv)hKnVtdbu!r<3Ii7cJ36aK0U`y?Va?%oSZ^qRiDm(Gs)oV z?B9)JWdkZFdz(qnwuIv6l%EMl1}wrY<#Aq443>C(gg3zh|FpI3PAgC+Mqdy-_q^x3 z^D9lZeyQefzY{dD-kaeXUy>Bz%6>fqdF+{YaS**;v>wf0bhDv_254t>^KmT^WrO4a zm%p{t7LI(cFz(kxpyw603t!uE34s3&ruJcaj5%=aOA8-rfJ;SPLPS=`TCd*}n1d0l z@>O%&_|jh+rabYc1r|1}Rl0;f^wrj>07m_9`ZSYmU8z=0QG1o8`yFwiFEsg-ScM|3 z2zXe@uS0v~Z|sr|+pp0f#|C!0?t`lP1ka)X%veO6P^&FTK0cin{`{iFc^eLKv@#=^ z@UT!5At^>Vg|LC;LD!Aa?t6!JRs_WDt-oS35C)F^%-q1nr$ZHUmwP)}Q3K-_iqCo$ zXSRILxs5)PH5{m5(niIL1eDViPXhXq7Bw3$@t{Tpd+uXJx8g?JbEexk2a2qL@CzMs z)$kulq#@KVN!vd|{{B1=-tG%Lu1+wsMi4ypN3-LjtPF?Fct|0-jN};ZKX);J9{<14=o2?zeWb?`r^VF{6?aPiOuP zjuAw5roVWdsQP4*8EjWC{>1`7htsm9b}7|!RzR8^BvsiZ#|(^!(+JDCq&WO<&cl>ozOjDCdG4O>{1}eAn^j;jN&|K*%Q0 z1bg+>+`{DN>VHlilfa_V8s|z4&q{yMbF9O|4rcHSJe}=Mzv?U)RgF92BMuF}aTQ#Czk%G=ROR*ks6})fY$oC0RKz9PHG#dJ1)8S#o@q~lJ#=# zvABXyu}YBTHI2D<4cunLBo{$DE==J&;gd?eGbll#+E}IQDCBBP@VURvss4ob^~0r{ zD^-V>X**_|=s04Wpp5fa5lsZrwP)(wL|W!H^iO6VWF78uTM*6Y0-;}whA2d|m~)cX zr>rp^M}t`AF`%~Q^JJOJrxS^hzLNy==RhRT;D=Lu-g)r49viplR9%bd$(%;5g9XjO}D4|12JHl7%}Wam@G zt%H+VWy{c31)cjtS%O%>QYu4Bx1;`!&Zl+K25%O3YoPbH;M;1$-`!LnQig5+&E0M* z3u~@v+YY_jSX1X7#x}uQ?IHP+Bc5;v=)@!dcJKN(F?y zdRYF=8bhLLIEW&oYgAC)oL?TUBj07<&Nmq*x=j8Zft3$G1TX#J>?H+%iyK}O=k=WT z7hyE>`)m>o#H@{AR|yCy$>hSCuFcqju-ZUHk`ORge(CdM0RTQU)}aIM z%A$y}f<{U#dWX~%<9&!lH4PSb4t0&RRc~`6wpRf(2PP>JiSJy`W3qxg*kB5hQnfTO zMmhWl|96@hmUW_T>W5Xl=3yTVq0QNNx^dvKsTK+04!{`sq}=(esgf zxi$9Zw}a_|!qS+iP?Bp}ZThV^C5^tWglO;o1QI<{kT2jYlSoq@6dO-TSwbr7LZGIrZZ>z6;k#BWb}IrYuaE zX5~!DU&78Yb)oN5F1iTJ5h_TET@$O(vGy9oa7vsBR ztgRhHX1)ae54QC+>0RdB6>{6@#6ofR_-QHL>;+S!t6;U`r#cGXElLXm&Svy=*nMIy zPS~qiQdYCL8mTl;h)t=dFz$C#?71i<2Y()~lz!TR`UIo2_)Mw7FvrwS{Kx_1(rCQm zUmN_IXIuH%3BQlLSdtOo^0u{Bm*2~A#^jB$1KE%?ydRslG*EBK8(cZ@2Oyv{&pq^G zZH4|EYP1Xy1Z(=QI?$YR2P8q#Wm-F@3C~eln@d>X49fa0@MYDJ?(11C{e-sU`^58z^9!7a;+|UrYKctrP zQ(Ao`D4%&N0ed!^u;l5EPl5lg!UsSW4Lc_(oCQZ{=TBF3%J=cXYV#TrLF*-~4|SlA zQK?TwKYKJAn#KTVa5j1N(cxYLfNcf(W#aP9TyGz>YrIiF2(%sN{=bJlb}|O)AUZhN z{2$$@-`Nkm=uC8f%2v?s6c;B6!~m@Tqb^@UgJtl|ggtCJ5S(}AYU;NR9`>;2%;VPoe)cCTt87Il{UdjK88tNU2wv5_&X)l2h!K?n;aR{309gOeLdq*Lmw5O zHLG%W4x!rT)A;DYF@|hdZ zsOA?n>mOI-#n0_8i&qiPD{qkR&w_6o!PwSkVr_sofwL1rgp(bvvhwPnOCbbz+@YB9 z+V1lGUHC&r4RJyt0uKs1gR7ds=`Q*aORMFYi=HOd8CDA-iWj55Nj7un+l}o@y&RB7 zmxLC+?~PbJ=Ar^Xju^1Xst^ty#f>mk6{d40HT%IvDhbGt*_V1hbM(VxA~qhl>CS@8 zFX3c;m)`_0we6Eif$SAN+_}5N0<*rTn>!%nYysPFWHY+5_0@V2T_TqTDw@6?3BEQz zgR)bd{Cs;79X>={Q9jg|6P)6V%|XT? zA_Xz4-PFk%3PGN5$Ba8Ka^Kj0WKdPvT6b%BmDIPTBMg(E_L3g=gu|k@nb!<%y?nJD z-ol7V-2SI?@&6`Fsb#2P_w;Fr$n#B8YB>tHRJxs}-4n+O8ZHn_-M%xGZ+uw!k12LWFYN0R!e zc+q30>Fb|XX&t%dc8A7VQJyGxwUh4phIvtI58+%PVmZEnFX*}bCH_TnP%g)(DHQ2m zB;iuzQZ`SWqIZD9`H5IdqU0FbD)tg^>YhVHn6dso5o)pm_S!UAv;>2EQ-dbIT-Qz3 zpiEqlCL>GBk4A6rLCdqxP5QM`AAdPUO0eSFobM%nhTi~}yIqVhF_lTbMiYNj!}lpq z7`(z#;Kqxr z+T+&xE~@a)r_^R-ux;x^JY2E4A6f!h*?nmVeU7AZBvC!i0)S^?VWTq>Xx=K@3# zXS?fJR_BOLl0thyD?#FfdC@D{=iwJhB`r|aItubKf-{Plzczo_mk{oAQnnQ&2!AgW z5qQobQ#Kx!dRr{(hD5`Cy#ALs_$YlBFRAe7=tgZ#8{?>^S(y3Tt&)xz^^bRzG?1Np zbE9;O8M0zE%%^A}&tkSHEDs=JY_&i$m$<|jCE=do3|K~MLMmbH1#u+-WOmFa|4UXQ zplaUuh%o7riE&6++9lu5c}tPx1%2yfuutgG$&*UMLk30@wO09szwv5dIHZt({$O)Obi0+O6;n3E{S5pY z4UJ~l@WpC8@{dd^VPpm~jV}}6d{0GYvu-Uhtl*`L;?`J=w$I=G?qFnK$Sv~8r{$?@ z18$+Ygch#>!?x~IRqe!5c3~tfLw{-9SAM1&>TMrJFDt6F-ZBJSf1ZV`L&kOK(BhLc zQy+CwvbhU*`JUhU*x|S%`lMF<0`N4sEeOR2!8F1|!6%I-_eSjFtL4*I-Ec^*ZNS62 zFzT2Pmq+`!tFiaUJIdGzkwEVo>J3Jk*Do$!4bbyK+pkhHEhz@~ZQ_gJqZ z8~6^_D6I}Xc9qm8s^3xb)?Quj@QK5Tga&~)F6J!Q$BT2EV9|o?@ol?3ZJTv7Z36gw z2-xaUWTyjX@K1)dzuQ{nySI75O{?}>ufAr|yf+(0}>Tj4f3=DlmEnAn6 z(0juOOs2?}o?9QoEsx{9#ZP>ADo!lcL*wM5DqScsa~wXRc@MYBcckaJefyT@ zcEZS#7HgX2g6@6BY^7CRe>f}P@+a(kcX)#VVlK(gsEiP?F?~FD$TxRmd?=<$?Ktb> z?TUs>1CvXFkq`5v);9^3Gk$#~8a|rMn1YgtB+}AHf!_|>?mrxS$YMngjd5D371-us zqGW{zOb*%15ik!1djbd z1os5B2LMSd4lUEkV!`d7Xja_IeZncnthAS;P(m-mjLl@;l+U{DaDDEJA2khF245AM za&)~s8bB}dC7qLQ$54JR#=Uv&FKv`%>U-23cc;DU4z;tH(b`}iJxY^I03ga6a~ zj1r$W>{<5;iB%7UsdD~%i~@D4Y2U}u22Ugc-+s<={LoUe0pjs&s;tbo*i!1H+c$|W z=C_mVZ%1@_82(QtE4{Kbu}4QwM?Z;t_MQZ52@aLDoQKH$nUYupaH;E?e-f#!9`&89 zVYZQpmxC++*@5_Pgh4MLQi?S%+sPlHTfh&)q;E_C23II+FyhinARg?s66TSO-jv8L zUM)6+ZIRkTCL$+2mO2MQh!%CrnRF>1%+ZFjB^t|H;6mvvh@xoAl!KtC!AE0FH@uI^ zKt1}K!BSfD!1d-OYb+<1&WbO;JrSzz(eqZs6nAoB?S}R%0 zIf>Xik2S9VJULHsHV`#9s0iwG8v{pdz#X_HvkWvc=sp*P|~Hwxv8kZfNKWE9vNbH-yofko?V>ZXGk#H1t+e*dtx>;F|=afC-<$bo)j&=vj-H9!yK5;$9?}WJi~j0#3fN*p`yJ zSSpXbJtAh-u+rzcE25QjeeRx)ExOX9PNIzvC1i6Lx#`d4k_W$dP?_n+9{r!n z-XT|6JUnu*b0*x%fwmyMAr1*kIYo!g29@Vs;1sE1+Q<=C(+;N$>_ zI;w;<@-j`~VKp?RDjLGWyL>mW--&0dw4KKGA9=|g;7#iA1mmUHf08Y$1ybDqSgOW# ziX5q~5U@Fj%&{8~38Xk4wIEp7BPsdrxMMF9a#ZR-GFG)`WDH8uuO6?#ff<=qLzE_X zIjhz`{k&tdpja9fh|yUzTUI3{_Si2Yt8u0}cHifCM+&c-m0cRbt+f*)VU#BO?b_pY ziuDm56Skz*8#%mLSb{E)B3OQRXqg^PkCVX~4vzO-Bpau*3JM8m^%s7UHmKQ1T3`TT zbU>_P^<1`RjR*A*e6>*MjY`B>9n`>?<>$U`@K$!ioURO^gZt$0q52GhKIU6sdE^l{&Oi5b_hx zinU|i!l`(pC&?TbaNAE|LFq73J{2t!O$S@;bG!2o5r<4uV*#M;PpHmmqaTkFwa?P1 zF_~|R>ADg7!_*Hc;;t|2_DBC-CcoEPHv5;%#ISN%3Om z>SjuwC-FuS`ccOAD2~=|N^Oj2MH}mujsTh_rPO77UUz3Z7CFiM@SM|fsuVFbXoYHH z>FSsS<+^@G?Xgy{Q@xZ+Q8V+Xoa!BKasay;XgUHV6sISw1WiFLu`1qDEy_wf8*je- z;;M1Vr^6OkWBH(e@O(F}r1QyGE*9@1Mq;}l6-WYN1)h_UJ+%QRbt8e+fpC7_=tm~v zB+Dh9ene2_N8aPF(gEY3B?mN7a#i19PNnzt#nRA^`w8FjgxiH!ihQA~nn8D37s%B{ zpP~3k6GtsccW^FB?e?~^1ER*bOJ39mclIyD|x zFbwx#vh-{Bjl;lZkBd{;O5!e_lXhUr5gEu|n4M|(k`XZzU~MbM>UlslsntuSQ zQRbyNX%@G2weLUesL&c@r|u&fh0omg>M6YL&1jDCTr${R(7|1*uJDq1T;=+1%H!?w z#XqU?3j11hUg0As7QG#11pS@3X!lXp%^wndBaYbPIkm{{vni6lY*EV>=FV0fvkSSx z4tjDhdeYHH7>tr^orl!edF0qQ|VUh+#Gdf zxp#b$mv|clJ`eGY*xX!y?tJKyA9E65fbX||t1o$84)rxPdpqVD_pnr7#{Ga`DW^3s zaz|(Hx@-Sum-l%5VtoG7U@3K3$saqQh@K-I5mlNuzd#`k*Wc!%_lJ$i>Jvn-V@;h; zJ3Z_~M1rYnr%mh>{vtooaRdB!Ca&*Pdsv71zu@`|Ia}#gPoXw_vJJ*Y_We1dENcDF zg4+Ki{_bf~8Pl4)bJ0Y={SBm7H39r+*O4E-r)n2R zvg!Iv&CKvy$M7~kOoSTNB74E37ezCIHr2kwl)o4;DI7=6wSr-;_HZ@Fv@lND8hLiT zy}$-u?W@ZVo!=W>gi4v8IAiG^3FndAx3W4o`M%if$NJ!R3l|~_VN9(e3Y0c4r+l74 zMrfc%?59htlHxR3X{H{gPSyK8hCZYw0mZ#q7 zHh!~{`^Ra}1yRNUMusyj>f#-HzAUV9t%jNAv(QYknZXW z3G=%OR~~`*i6(#h4G|lB_e1E=#Tzn%Vy!p?+Yz_zrEj>*P(r*FQS!)e*HjKVLu(F~ zw~P|pMHx9rvQyL$v2|stu}$Qhg_fJuJxy|Zt)!C?!~b`8KPUOh{umVEZO?A5b8z}T ziK3~9jTt;UEmA1&9F%n2e1hxEXuI!=aEE3H?0=BATJuh+JnLcMKRZ+mlB;g@)VzI8X8V z_%l2Gn`Q+uoG&+=GJDzm>(b`eUDfsvG5G~!$`4Ro`+H~2`c3xl+ z21H_edR~Tt8UH*|yRdIAut1IwAN3T zwJ!?Nb6~iBvbzikfTZK`@}?4;OWx(8v|prZv&cX)W$w6~4-=9K&?Q2GT~;L!wbx#D zLEQSNpr-JabRO~pkX}NE;b#V>9B16XxGX|?+4&_(rUF=4B1urY-<=fS+6jXWx`C#hjMHE-Y{9$v$P&xiQ4&dC7bghXIK0MqLQ z^Y#vU@#b%EqmrBG>VecpSWlF>X8w>@I}CSHYbGfdN0{aHCYaSC?#YlR`!k!sy-m zagaguKgfT7RK`v@t1!Ve5@`6QDZiq;-!;E5BTZ3USINsOwHoaskRZzRX?@fldY@88 z=4hB=24O{$Yuv!)2MK%)2`(}Gf!Q!dHnwo9Czm4V?jmpVo;Fnume3h5^{Oz`3ef#D z{{>$7xumNcA0YXhzvJAQAui5#y87t% zb+f-3ivFf70@H-rBV!9XvG@$}WpFEDbgnPOMT_`4umbC6B&-{76r7TMYo+_J>Mu*! zLTSW)itXq`P$xwj4K}qu3Q6q|Oj%sqxdDm5h9;kKO7SDKsr7-42JW1^{d8m6&g{DS zIgMliA*T&4RN4EMSk94!459hr%P7~TJt|wykUPJm`x&xRm%E8oZk zE?e-=AZt@rC{q2<^uGTq1jjqu^JNiLbSI9+#MD?rvit9S#)u-0drJ`k!B-O`Zf|Rh ze};+t&k;~Vfm#{||5*<`KT~w^Y@?fCq(IKCaG%9GQ-8l^SwWX{C1Zv-VOJME$n;_) zhgpK`WcOD}c|+Y5I!p!M25h{F*}F1a%T%gW z^Fd{FNl;H7l2s@%(2eugI~)|8!Aa{)bv-AU)#xcdsyX)LEM-yFzS|uAE(+=1!ObYF-0U)U z-~BYix%~ie+VAjg}MNNW22=oXK-B1YYbVy$aiockiYW=y7z|QnfMH zYH;t4^cBN$>v^Ny-iX&YVMlrv=srrWJA`X#^>YfXZdi-CG#t^;i#?9bbX0OT8se_) z`B?t&GDgiz?Uvub{wAPU@mP53o3bMRC4ZVFyS@_TFFNpYQA$K;`zzCY5tRJZL_PcS z$BDlNkU){cS*EP1fmn$dsz{@J5pn4kM(*F1`9&3IN0v7M<=)uQpq${=AZ_}{`fhM+`_D!QbbdN1uTTJ#0M#=7=d(n`Y)n#ky~-h5 zQjW*I`HV!JQq>#n-omoCNG8aaIfL#UKz^=BZ^jjCNYRY`*hFcs)e$ZM-^UX8%Wv0D zIVpc##pSH*wJ@`9wE!1_p9RS(bdiBCCia(!!$&G^u=GpGUQ%{3)h_^)O8;X z_w+;TSD|Fot9{(*1*>lZfpna#^zh!p!l)zqHGi3iMZLLObN#1Dj(Y=cgC1@+8?elv zvT$Aj>@M0H3xg-!8`q5I{T-WgOIB=p8o@M0TP}F~`)*fRZMd zzq5$?J~U`FD^=p+ln%U4tsLgUW2%7$^5d(l9Jtz^|4i`@^ECd&HW7eEE@N{z7-GQ3 zT_kZXOJB=ysDC`4FgI8_|2=A$fnIx1wQfU_BPASBk9@v3#K3)RLKQk3?(I#+=AaLV zQ0v;Usy~%SmV@iu9nhp@{zkO3$wHjjXApYb*Fc`jqrHL@pt*N&+;{;kXxxqMNR5mN zs!3blu9+92RreN|=V9fO3Qy6>0no=scSEs){3e&K2u>QPil)cYLjx7Yjj=z&`9m)F zi}p^z_GsoIlovO&5h(hWA(4xye13B?I~M@W=|7M-L*V2|Zgozj6P=rjW4}B1({KxC z?=G>Am)DAW{ua)y!Zo{=J_`T_y7IV7e~TSg#DH=HeS6mMw{+tB+s$+ zWrSM`l4ecE_W5C==B}BSI{*EW7sLS_t?)aI6#+*NE(&JI6a)Af1U&!DWTG>My@f$O zKzKDWMJ$GBockl#!<{PX1GY5-n?%6Yoq_5uVDX%f1J9mMy!*zd_`6Km;@mW%GH&@k zP#cr=kH}>!zg(~Ml{x?o1fi6FFKJBQ3yqiK7qU?K*TRCiX!2HgOOSQUD-g-?=a;cj zxF<&a|KwWdb=p;`Vyp=C=;90T7G~O7>FRTKa-JFAzA{p!OIdH^j!~OhJroc(iY$4C zZ^Evp$v>^{E z-pJ4!AXBmPiQph}hj#8{_4qy|6C93qnINjYjPF!gMO6w~Wk(WnH$4;&|0-P35NL!@ zvwq{YjbCi?{@pjLK?LSW{Wa0Jb@a(fv7d9DO&HCyntV{1Ld~bjeYcqZtsiRKbr9W8@L`PN_z%m6W*j*F@zT(M-g(Nrjf7>(* z8pO9&iLcF07PUpm(Q1PayU&G34y+1^4UXC?chH`@Tjn`T5B}4?{hS^_$56{m+AV)P zX2iA9!p8QxW@mDD$q`?l2W{=lQX^Ji{>-k~3DCN{RoW&u(~JNg_M=L&Ro}H*u$2#P zay9cRd~A1O`Q}_?X8)+0|wpwvgEJhU8^Azbu4Oqx` zBk8g|C{DSGvyb%Q!=BGtTnAzRtxB1cDB;h#?)5)x2l2@HS{G?#q4^iHork zAYe#54KtN@Qie44!Hp%v-FV=Tc2RNK9hnE8CV05N)=f%R343Unn|n|D1pfXgJNQrT z=f74m7sW8lE&=kGL47!!=$yaXt{rlN-!75@vc8`LXrA(zU*IAPf1=DKit_G3)%GyUA)j2LdLm%A)|( zLY;MN;fuz2WTBizKZAs!9usZx^*TSc@yjSp{E)t0lLQx`-4 zxB_?@V~^w23zhF%iV-*=MbqJTpY02gO0iQ9DQu&s%4>wy;cs7qdU0 z@;3JhKA0J{fzq1v7-meBlm|SdvEm?lRKWH}0zzpRO&l+01AD^|j+q#KyRpOcg^sPR z$EH~-%zFP%T@~Pairig$($~mpN}2ht>`o!{WA9R?YA;YD{8ACn-)H(~p53O!2 zoja?9bbbL=P$;@?u4GF$q59&wLaWkb(J|qaig$Vuhz}K4Q?lhRKyY1gb*P(Q>E^VC zBz+YG`k5*4|MZl1-|Q}?BG|{yZ@NP{eQT>L11<`KxpPRIA}Q#RajB^j7(XUNC~@e= zHc)&;ei*_+eUjEhl@xM{F((gc8t@s-qDW&%C}}n?A%tIhn}4WIVpty`|P^o6-VkEw6{#{ z#hhDc;iop?m7*=XqulWup>O~Gdf@f+a|)M%JX^Y;DP;f10_Z!?*+Y!uf#qeR8{1pT z*t;qH0KVx)V>AiGKi)rdE_%zn$=T(YvEiXwDtjU{i~TlA)^s&p^iY&-UeuWIbwkhX zS_GlF2f}s?UZkGE;>Ku@!e*P#PiE&e>w?p04B;b1qrg1?j4SjZn=HC=GMlQ-&^ z0rhU-lK}beFL0SW^D*D6NqaJbJ!$~qo+OpPlorUZ9#am-n;@m|Z@PXvMz_IueI*0Nd9?9ZwJP47)#5R^DH;2lUt z^n1bXlh0Xn=%+28CBX{)&;yr+*9l)DaA?m$gRj}^vCjp#^$g?C*0%oUOyX=Jrp>Cz z$BVQFIX3Jco_td{jb8buoFBvk%?=sNEZosYhfc%zKFynG=91re27@9={tsbq9n@C; zy^FS#LVy-`cPUN6)_IGFU$DTczotfEE<8E_7{o_U91^4rL5-_Jh4Vto@L zZmKk~b+uS&vpk`v#Gvjmt!`xgrNAYr+GJn67PM_-3=!LZU7$ zxr8_P5-`(=U05IhU&1nLf|D!1A{~>ypIaJbP7gorN|At{^`V>8^LajM6#WM={<`83 zyx#e6$g6BzH6nB0@hHe`1)EIz^NH+`v+q1>bjloKYG$=vk|_S9Q{SeMq8<#`JNW@$mzwi zg-Fl-YA1z?eh-O;s|@1Kpe)9a-b2iG+Y}l5cueN$Q&f~JXsKcFM!yVC(1@X= z+NPwQ`rtGr==kiD!=~i<b6KZD{Ygz#cZSOO}1|! zQ9e^Wg+C);i!;03Z<<`za#K2@+bz>Ue0_hoPHOxkWB zlULh0FG>kP1?_%64_gk?VAWTpE3KU*WY=u*Vw}&5bHXw5%$aFu%RavfnXBRkJ{nwX zcqUKUD@w`&85*OR`@QwCmvGA4_qo{V>t(1`tjl^$W1Xa)wrT|qm{M%mi(F27#@Dxv z_7qTUN>iZzP~HD1nERV^!Qf9^N8aK=-dN6=I*7D4!fA9gTa_a!j1;slCdd>OdGYke z(5qsxI0*m903{ZQybmj(v?wUIq+p$#R&{P5-J$)EKAi#rxea}emGMrL4{f7K?>utc zBVJVaxPXC#trViT*u0HhCW4lMZ7o^ydQF2p`!^Z3Mt-sfe4`9On149)$MBfUO_D3J zm$eX@JJY?@+R%4V`}O?4K*)t7N@Wun&cW7zdJWP#^d2YuQPO!ph;f9Bi~Doyk0uOp z>jIZPD-jiWfocql5`NQ`{w$YE@*Ud0fC!t3eaO!M+acusL+$cs|Nd4PF?1s3p01T$ z=@h~+t*)RN+reZvx_5EjYfL)C>(U@zKWxhGr)7N?PWp!%3@0zj?>=C1%BTZf(bG^G zUR+h5G36Qg5OLz@3cQw#IDFAMbWTE(h0V}~>*c_3g{W)09+xWH>@m;Lt*#gRfq8Ky zM3HyCz@2FYn&xsQ91lLTO#tBsy=~utb#En%X`j9t5(lxVc~t5u#Rd6@`R|kMl=D)Q zeCJej1GUED5ac>{*TN;4G&twrv@4$er}0U<7kT6WHfN&;S8?j0YJyMWHTesY*Y9CK8N&M9*#lC*X9se8**P%wzit(X<3I^K~ zTSgjw_{Ai@15=IZT}qPIq;pD+ zLW|#&c&KQ}12nUT+Ue}s6c*>!AmzAyg=f!*@D2z$YvQ|C=+}e)43nt~TV$=AThFro zku09(@i6ltt}HgeouRZy?79B+erETWZV27aMm8kiu@4ysC)7*OcgQi<<`_&}6aaFz z2iYT)A%dVfxXdk@HD`GA|pf=k~I*bZp0FAQsUF`d1M=J~0E?bS)wj2oFU z&brzc+f0I*JU~A;y%lAFpw-v-mci+#$ zw%x^Pz@yv?a+A#$WLGz0$gc`Y{$>JVbC^3WQ&$7wdv=L&xoa!86XOT|OOh>#<3T*n z*h4>m=*B^>(%CNjDzo;MSJ}5!n~Oc;bY}0JLJ?So38Q(&X*g(gQREQPPn7R{fM+S~ z{xaBsnaFeZAbAR#Wkbd$VBLXRemr>Ftq8$?yB1FpA?*wC%$}RZOhOl5=z%BtFSU;C4?ms)IqXudZ0+@iLQ8pL-bU$a}m#PLBebkjd;kmFeyJj}Efqt@k;q4}ys zZ-9|1SVz=-y=oDlf7xepQ%BvC7!jUA0@^nYDSP**ri2Oof5wGEMXSbrwC9#==3iqo)FNRSXAsBkI)>;TmD(7_o<9D z@hKL$J1nkB*Fd=;dTO*|m^mhyGQ4T}um-B}PIQ8+HPn^FhiC7EzzUeN> z=e!Zre&2^l;9#;K3e4v-Al*nj(yQ5r_b=T;&%GF~g6f@tuUe)W=T1AqJ3^L)N!CB2 zTWU`S@q28?!K5HTs=isyAVNL9!H=GCEHsW;UwxEt4#aUoVIpwyUb9w`S&N-uL!E&0 z%%Kp1IN1emHTdo7oWqI~lF1N^6kfx&RBF35PCe(^@KP&8H z9mbT1$|<6yzz{i~f^1yDtm+GE6A*9I&ZEci=KE&4ca!A&;&96_RHcnZ3()aLj-#6q}ON&1|U-`t_DmbsFRn{i4 zO*G+~?;N@Q9f7E3KZw#TFxqzYD|3HzQyTVYaC{$WH`?0l+&mYfNI=`V11}YOg~=%u z#{?T#Y&luzPTeh%@Uj7j_4DCo1&3bT{wIV@Fjy$@26d7c$C$)Kcx8W zpOz7Lf23xwLG@J7p=M-v=1>-YXupkoib5w#k>*-KL7t`dyizt??2cO57s1mzPDt9_ zz*CNSM{`@cVvow4*k~5YuA4CrVO*G4Tb#{0aJ@HGUS>4hW08Gv8EW3v;JV1jAb-yA-ue}7 zzqtH8+#ws0=k7YIGo+3buttn*r&hKmE$1vKL-qYhX9}HMoaYUKC+L2wsuvBCWW#m+Zm=Gn7@gqU1;_&`d5|BX~t?jbm=}~p~hS13z&QTmwlGEh0 z*>3R*t&&StLjTKpiT-NSOie?xn9i?iK=iV&_BY=PcQ}=ah;3iM!ATXewH_4fJNxkI zB!)Ku#Tr=S>Q$26;Dn#F5}{FRWR1`t#Gk9?Mdm8gSQ_4Uk8~Va!Kzu+Q|0hzCxJ_q z1YzVoazuAdTxqXfyE6@>XutIVBn|=$3v7)rttk#0-dSszbk(dVMXeGG`@EB?n@BK# z;Zk&|T(s}uDGEf}QQ9+?b~@w#Y}3WmCzsgeUQzY1rX+oPj76&xha<%KgPpL1FDhsl zclr@})&lO39&5j&oYh6<_Psl*w?4$N1GtrUekHVAE9-G}2$zV|!hR}czZop_SVAJaU1&`Fyip@`To8!g2`Ky2> zAq+qM5#O2-u=Be^-^4?cd+#+8MMed_Xh=;C+mYQ;0b@`96BWcS*#4b2zgfHIlr z1!_+c!l->R(%&Od_u7oY;9y@z#f{kDZPA+f#k?s@P%EbU3N0yN;gxY(i*lZtJ#*pDYyUyY?W?1NbaX-1LFS;MFwaw$rm*oFszOa`^T4j1HrxL- zB#ZP)wbunBJr_3n_J`4bRa*Y9stXqjVOJG*wb6)R@TK5_s zP9>iwAk7kmxnw?5R?W=aE;fkf+{BX9J%&-^E>m-%r&b7cN)0~lJS2ULBCoOpL*zQ;j{6BbKMmGj;qIQ76p14=fn*jwf4U&Ie87?{W;*<&`@DvQ2SrHXW{<w~GSj9-jeYHR?Nd^& zd8+_}%-`d`njWe%(n8~D77rI3e$LWzeLd|ua9L1=ld(|I^l?ft?CJoZV)hJTMnUD( zvK!e5oqF>iOTlq<`Dx1io1OVR5P=Xgim^=J;cM_pXb|skvVXyVAT&Rj_r+F8+ z8@wmatm0sCdr4ewqw;dZTY?3>AKIFtok9d}(^l`(+5*j?%lury;xdKM$4S1w9TE<; zA`i&hvUcx2&JzhYqSZw&3?_`lmfieem%csly2#h`X-ixk=2qFH<76>j_Dsolwv&Fz zfQPLNHHGeTXce@DfKLa}6qDf%VwA^}=HlY{(qhQbHaB?ec2{P1J+Pg6`&7yl6Dvs* zwY^}xwUFqL3W=M@$p_Y~9QPK{d!=0neBbbCMcv=uSqon=eUO_psE63wa1(Y47rGx# zZM38Q@Q~Z<=ibBnF#RGtiKP@wowW=ECvfftg@Sh3{ zUAsa<6t@<x)#kz<0I`ClFV(y`ZLA7$~ZXSPpX z;HmnjB|X++>^bez!28Z%bGG<+@``iL_K2h@ zgsRn-T1|I}pW4-X*Fq0B2~iZZMAC69T9awuuOo{ZG`o6o{fM+b3MO!{d=QR3w|cr2 zfid>SDr5$K6+}MG(SH%-_wd5qNwe!h#W*^Bi*b!VnEnD47%^fP38#+RX<*fuhub@opT)v4kCFxZka3KN za^9C)n#oq~Q>C`rq2jgIoBf3`#k$7`Y;r{UvT?m1_^xxUR5dK`@OY1cW*VCOllmCH zJ^0=IvHqEoHN^x@VBulJfzkHG;G-hKp&1IN=^k;=2TV^$4D=DIVELOTEvJnGPC)7`TLNV z;Rj1bB7#DeuJ;sK*fMdr@u2`IL5zG-SK8K~G9Xt2-S%4d2AX*NjblymOe^l|Cu0W! zzmkQw+&5{k8PXm{*xR-Yah)gBzohIYeh0)$ zHdpQ;d^*DAg;O z;WI=B{J0K*c3l!=&jo3rDw6|R-_ipA{4DiFQm?iruCUK6e4`z7!Q}dFknfS7VCtuh zYs>jjvO3wE^l;4oSIqiP0*0_f@y29 zj7DM2xo6d@JvaF{yEQj<_j9?~$=CgSu-sb&wN_)n*)&8p>F!g0C-}4ZsS!_d!f0N~ zyBWa)huI{>y3>7+pLPe%g~oFe;DontfGq32F$B2K;M0)c{=40fTb6)rY;%rX^kPJy z0%~byWpmUfZ<{rd;ThvT%Jkja9my%ygiafG9>o{YtqLe)%sQ;~*XH*_B;5vc zpg^zaP}LB_cG^{v>_u;aNXMmNGmU?n5V`x^Gh$94%BJKoL%vnY@gItgm8|QWmn$-} zwX>Oj$2@iz*Ae!INMXl`}2=1q03@~p9EwTzslruzkcyRM@*PZ>K1ISc~h@w zO>vLpdaSkjcI)_|lISh8`WcC;SHh(z-nONT?|tN4^(kE#omSD50LqopB14&QUBO(9 zJ@S?3sz^Co6ek$#qEUcz$A)gnU1n`{Y$SMZe)`rt78=Zu?9#zta>&s2@@Yy4#l@+7 zM9?S8>g?XDz{msd&4<-R0wFg_zJ zC57)`2^Zxp+S-Q;GdZ1J{El%Ia_g|}HI~9pykF-0GyR-`}5(HBt-G=t`)xks^ zxpLQ~9&%03x2i@8a7j!rlOB6M8S3px7D7d?2WHK8v1V9{+lMU0+^jxscG0_>4&Gml zM3aZdo7qgQCt;^4;k1B$R58VefL5JKo3z%Y^ZKl*4q?5tHNN>fs_gEPZ%IT?Eo&Lt zuF9X5|JYWdb6t@k!u;!T&0#0({?74k?XCIc2$6f*n*Fa=2S#X=L(7j5_>P|S5$6x} zr#AS22j8$L6(^NvI2nkR;Ht>aylGS_KBhG~@y)wQ(FG8x?$mw)mm0P{$w?rf6 zdwLA3lu-)Q@yVP=EvSo{`J{U!9gTcZFHk{<)topsF;J|(ddlpww@3LAMkcxa8I3}u z&-OI6edUm5AK`~dUzHJ`auNWPb7Tir(BJ=HEt}ijApBX2kMyG^K9xw3G^9q4V6>cX zz%i92GAxS}6CdL^-|M+z*=NrBH)aJ1Ui;(Vp%`dY4wCCpP(&?v$8s%6(4r*cZb>Adx!I#zzPYfqeq7|5oQGM9i^dg7$rC)P?B57nN+1260FZdnj*Q_#!b9=#g} zI}0#*KU(iJ_%-7&+uc%d)e>XU7!m2tmyD^CKd;w|-LZBm+te>C9o=%maEtWi9YfG> zWjy2lD-@Y-o~$di>>)TKQ%!rYJTW~Y?0T1u(xLTXm8{W@>y*K?X|0z<_SR1a-!I6N ze(Gdu_0-uqMgd|?xsD{%1h7wsU$@5}9v*TaG*WF5nEoY4W=|0zwroCEb<^BlJNlU< zI74*a1-Jh8DaQiKXVmN&`R+Q}aM6nW<@=?tNYswzz0Xiu6MI4;#R4dcslWb~Dxs;} zrAaV%b6yn0KJ>1F6?pGrZ1S4>uS5%aN@(?)imHD zAj@?%flR(j8@E`KfLyb}BDmVRp?<2zN43m%Hx7{h1cT^z#s0-FdAM+H`hv6faACgcQh7We+1$_Yxb=Wa z>~&s+iGvE&FrK43PucVaF135d9nj4j@#qwvg8kYUnRt7Dln{pI)f{K$c2z#9%17ft zzEXVf+$Dc2V6wFHq{qxs{@`TP^!L4YF3Wr3^v(U<9l=?>w~NW@F;g%9;v&QPbr`>+ z>D8_p_8juq5nc~o5s4m7v<4-a>S1d!!wmX$Mv3FlLU(X`3hz_EUz{^iIJFsVzC~&W z-;clXB%Qh-8h+=Z>q=idjRLinO)A+R8L(QJU zej$UC3^&ufC{qf_^iDz7yVoAHua@BQgt7?dFeb;&8OUctnG_(Y@p}y+G=b&WT zZ=Taa`xI$>-RHjnAFN;a%(=iJ+YA=)Dc=!S|8A|Vd<)unZVC#QyDTGmy~xL_R@7f; zI)(7ih2JBxZlb|u8*ke3gFf5{%OMs2c~a1nu4rx@Ej4=SecOp`{l%v6&U8cbEAYiW zZ5dm-qxZrbSC0~_oJDJCz5KC@jeKf0MeEhEct4+umL$kvw243NT5L2@H#r7)-;h>~ z)@tOIOVm%pm#^OSR|i}P2~|k-h7g%SpL5Vhd`NIJrX@f3`V(coU_hC%`D?}RQHcP< zydTz5`?#0-80pMyX#ylC&CRc#x!3Ee)DmjNoznp7uo8V!&?7PvFIdE4K1^%!IPJrq z120|MrlqHktqhV>A>_gQ%^(kp;DFCY0r_9&It*A1FF3y~{>{Psb3cI4>dyUmTj$Wj z`@${0nGd!)guOgns1R(|u>!rbcvoqxHcLq4CTcJh?kv2V$m;e$7v^!EmuZMJd#eo#Jyb!?pH? z*XRCHv}ZmdC01LN(syQUBuQLsi--{Dl8ncU`SA>c890e^#2Z-kkq*CRY!xmnul963q zU~ltx_6Qs{%7tA3g%{?jjz%{1AU2EpKYZAt$RKfH%z- zUI)#$qx8j2^LZ~Oz|QDav+O5hG2HGxRYuK^jsva{rC&*A52Xz3?N-T#>==xk~`t0+$+kwJ%JK0MH&to!zqd=562y8(tAE-k1#0jV7Z1EBNP**>)hu1Y8V<{a4Ic;ex zr`w|AlT!jU4mcR9+Sf?OEWD#?-!ps8n_321unvi(Dc<=-H>s+=BIB!WqW$-H z^aZ#@z5eoZ$oN(ulYq~}8Np~FGY@q+fYuV@$D8uU_1*FhJ|wHEoYFcyRJmLYp}jem z)gTs=p>wsKf&CFWAD2CKK;-2S(mjnPG@RN$bC5Z;n&DLEbu#)%n4=^Q&rA?@B%dmS z>JkZJvL+Oog0FP=OfHry=eI;>T!8i1q6`2V9V3NoqSwNE7te?^!`JbmY>l3r7yNp< z{KYgGCE4h;&b!BB#&hx6m&4Rz1#?`vLagS43EFY{t? z;n7*KMXb0^mN)C_k-D>VcjsX#{CKtlH(0c6&FmXCdrZ_#KDei;Q2a`_r>d@6cH~6M zkL@)*7Lm2&qA|fa?Y%*=*Y&ay$1q03t23oA3b^8hA=Z87R-16WObrQa57w;gD-q%j zjaZ1oDDe9wLcpZEcGaqzx^LHPOmm?S!#(pJ*|irN(8v>ZY9Cn5h_DfDv!&ohyKAWw z-}SXy;69pg?K$7w8W>^x3D03d?8^C_ms(bQDk~R)&iK(wfUb(voV;LQybI>Lj#Q+_ z@9)O4krT78ryjWbEwpNv6;aOWnKRUS%E_W*|W6UmIn@C-8zom$9gP*5d5$aKR}q;L86i>Ma%hR zER(mB_cLtnj{9%C+k6&vqj#D2yTwp4>(-wOuV#};KxDhb+i)@fG_aVGIrB>AQyqkW zf9{+>v?`>+VPQKW+Ol|zjRMpK2IK_HD>}R3rw)93PBTjwp4s3$-EQ30$guj+8?V3m zowX6&+WZjz=1;O|x}|38JA^miykmo_bu_&9;!sW#0IdTJJRUI}F~N~$$ReiWAg6wu zjD}<60JOg$b3lvHOpdc(;xw>d!`5~AjzM-F!9CSB2Jy*GNVgp>6FSL8XDh>QVPqFr z@*#y`Un>O!lk}jlGwgkw=9#>)n(BAPRvgPC$MrX>)Yp%cO$M4>t-a^*S#(QmXJD_` z?7oOO2MLp+9@BL_yAM$|YxdeyectT5%EVj9EFC-BeA$UMw*s=s#Wz=u_-n@VUr<={OVv#ozi0h#YxU1<<^DhmDYn^II5!*OVC{nvcZ_79%}#=}2)WlzURURV#LGaiPi z-^$dbZ9mCPf7d$J_+ns43Zb_1;2h?~i<&8}#S8LV55gWLD2%b%UndL?xV!n_X)$%W2smiqHFidW1VWe;jy*PdS*FXtu1@vUCY+COb+_cFy%tc|!-cIj* zS@(H5WlF>e(87V*v?}RXVQ)TWn4ImoJSFM|S4iZn;w2i`Z+d*%O5<0AbR@ZyFMqjU=2pyTt#9B0MT>-O;gb|E_Ig^X0f@e zGg-2ECSg=b9iLf5^Fp&InxEf@BRa3aNgWo7vZdge;dxUZ$wGP&~id=8h(I`AO9D0DUnbPjG2Hc@T| zb$YojmHO*@j=*V|Ga=3M)k5)0q0ZwMk34cDmsO@XdmYB-JEsQKmVKoiu*S?@&9Exd>Dd}#@QpA>D9e{{UPtcgSY!W+FM*{kRFMF zE5={e?mV`{+M=tpPFA!2x8vHKL&0rNtc}QW9uRE%x9|8igxoiD91S__!!%@MlDlF^ zygZ=rt1gO<=TV~c1?xGoqv~%AbF0%|JWEpk8;Mzrv*LideCT&{J@D%rsf+9x)#QJ; zopffoFzw$iS!L@1XOuRK+wquBJ7eoh{WB?2y1L4j-oLk~?th0dZXlpZ8!4n$tXI5K z?nv*c@pXL?3zcgdraqWTPvx;LxOz#7 z!mOnI!94TjA6}r6ZwYZsDv{1q?yud{*PHF<6GpQ(8NAuTSE_<{6-)}WP4pN-2^f&t5U|;T{NE2HW5yJvsa-r7KGK+=GE2sl~Czw~5p1 zhN3l%B-y;zOa1M1GoiQMSl^LLb3@d3nvhnJAMUM?x9*P>6tZZvGq+{#^yzhuw~;Oc zA}+Sv&0l!}3)OzZ{bj+CVV`WSkdB0|84l@p#1R6kn;OSXW|TO_(j_F zz%H9BUQq)ulYL5>`}vX5l>WU{RKtQOWzK? z1VqivH5+On#z8VDwbw|d-tK;t!JysdYnacv#Je3)}1a}%iaikjfjbO6vMV0)MI+yu&kx71U zM7i*_K$OOX58}iCwh8<&Td>#mY9o#LJ36X&zlsH3FVfLi+{zm-R81&XwJ*ke!xvin zHN(ih2f1Z&G%<8X?w84;0a0^0pGvmHiNdo1(`dy6QKn=rJ9?sp`rU0y4i&Cyq#9rRh9<_Yha<2}yFK*ht&gmiVmk^7ut zna*CcvxIIMrZi2J6Hl(U@#!ZB-8NSd8O_6Z52yn?2+f_*tq`b*+(FXX{dhMYDi=!*e4b}9VExHe4ah>?UR;y zR|PXrxtgK&i{4Wa4w7stC6j9OB({L?ppx1lY-Nq45_E`nHRv~!b^&p2a%)7i2tl2pJ{lJ8L@e^^{YGWQ1V#-?CLx^V1!69G}ytg34%VCdN6`k>& zDv3Qi?eJWZIldw}aY0ZQgZ|qc{mXUdI9sPb)hMq4pWzT*`c+J>w5oO_tlBFFzHcV@ zkaY+>{v~CuYN)W#>xD>a`D|AjpEW~3!#V+-le))D$c?50;?$i7-Z1|N8gru&kCiRJ zwR}2|@}4}#>5PjfFciAM^`(R<;V*tN^g&3M6q4KiF5VPPF#>M@onUUa_e{EXb`uQ_ zm{U(m$wurrba{0Fh9U8h<~YwO@Dg0-!w-7mVv8oxefKgm5j$0`Z3>~DV|cQWgxk|o zgH;t+@-u%DAV*x(=T(sRK}Pzd)1{Iu9Egt*v9?HiCIrl<#ka!Y5G&ansp2x~XFcQR zXmZh*1iiWtqRUOQ@7A__51J9jhq7R0_p%K?P?N^*HCS28&%`pZn&YRkQOTNl|3yVCoPfz6-5@XzaECNF!8%5dtto~Lng3^XDwxnC6zSihSjC$!+R$h zBW9&$o2q`Lwq0Jg`dd>3oG+lZ1t$cqp#PT;^G}Mm0w#<-wz*$; z{#Js}nJK%Da)aSq()uYRlmPPH;5z`4f`9Tbg*%Hn)x7^p@|gXs0|R3R@%diw>C^SX zwJrdfGqYNFt@-c2qAe9BxwyT*+I>NQ5DzZ=~Qd)j5fFdp&&Qgg@y2wo;FJaK`36fDUk&o{V? zPydK>(+wNO*8(Vjz`dW2aaA?wLENtl{a>chiG644n!vX|ppjz*ZT188!>yvbg890L z-fb!@Eq!Zpt#H4k`4)7N_1p~VBGZ@|70BKhco-4nsnR8aDU>HH(Yk5Y2i)CtU-E!8 z)fsTmRvz$|y$IjSO95zP*yYc=iR=9i7V8SbVCT13Yha|gF?uYb(pKzh5i&p0esel? zN98>G&E0gS!|k|_pc1)TSg5zQgDA1$Pb6XlE?A6h76B2j&jo<$o(4t@1=G0J6!>5n z7Sf*(G!M+sYfl{!hAb}MT}=_c5Qjt{P7YZXxZp)RgZ<7&1pi%`*8-D$u!${FU3G_r!T_mOG7Y2gc(=o<`oNv9SG%a?@-(ijc?aev_G4dihCR|;kJfZ9*IS)h5kaQ1 zcJRZ@E+fj#(weFAH2BMh@1*?4mpsb908hqKP`$vhrovXFxg?o+!{-}*%BYM6F6(Ts zIavM2@jl%9@t=uAxDA(>s1>@*>F7SwoPdNiGEGvo^oO()fk}Zs-#6Ti{pk~7{N&Ub zwZ|yG%5dKiMi8HTP7Jd4BUcP0>+_?blP>`#2IP0iNM6V_W$sOXtZL;S=BU<@bhNh2 zc4g#@N7aY?=`GoM#52~lHJO+jVGiR21=c1W~^2}dPuVEGj^zUU)rLspw z5c5sZ=k2E{s#51;1%dI@fUl~7@hVlte|!2u_Ja@}dU4aj63&oODU|513s=p@=*rcS zDTdygh}w$x$A`2ZR_27fN0IR7<++}&QRA%%)iOG!ViR>gsKuhpk>1_n8)tY9#Szb` z&g*f|%3;umnG=<=if^?!__!}Is|cS8opJ0=F}&dbBB+1g?!93{l!R`c4DHv2xjYD) z_+`3z4rr9g+fU||#jvJGS3|1UJrYa48esf3*B)T20napU^BSb+>L5#1ewX=R&|PT3 ziqckvqFguk_0-^uuKg2H@tRvt7+Xu!OJz)-S9R?o4ZaulSgWSz1vMr&1$$5e`ci31 zn>PevpWVA*ZO=}fIDnsH2lC#GcfDgivMOs($c=r}xijdMi=A8+1lYc5Tz;x599xmU zZpcXf@mYg&x{&e+HnYIArRDzNv+ToAF+%39=Y9nNEHfQEz@AgZx8h47NzAK%BF%+p zp(15HMsHCBwZCyA&83zf+~V0PxKPr|&HhphR*jq@XjxiX&;81I6jAs_!cpiM$6sLP z|8W92IKY#h)1iil?LA!?oD))nMPwX>WX-5-}y6{Kj!ZEY(UPyp+dC?IU!HSuqJ{^H~lc%O3f~+4h^GfyEBESpejc>FfA}T(fhmk~k=dOCCP=7|TPJBv!9SL3!~Ki>e@ zUJL~0DSeIx8}SM~s(MDm>wG>MU%ww)S++^~FW2pQ_pF0`+^Q6K8Xadx%o9GUo-PZr zk2)c(5_QQTrhp-^ktEqghVR-MT4`&lIMj0W&9`K7mwT%1`4ApO&f_~D~)|StO~cYpFOlZfD14BjbjW(Cz~VYa`4T}$dk=^#E;5_-BvJ#0eDoj zKlzM!5D0CE9|T5l)CGPqA9nP_^@)^k<>`%uqv+zzaQjP&_i)#<6z$;5$a<#_7#--5 z2j!uTFaJH#|5Ha^vc=GvFgO05NkT>cMbd*25nt4}8%PNeEw8567s$Sx)AVZeT#uJV zt(EqBlM4svVi1GqGBO7LJuUClS1H1GNScv_Y}|dSD|T*X2Ce~ma4;Ew9?u??<5JSk zQbHd1vB>Ato6~#Rd_Fs7!zm}Bs}jb|!6Fp!pVEBt9O*bO4npEHfr?FMs<<<|4dZ9e zM$BAVy6kUV_bjg-D!ua8)ykS;^edjjNljvGG3j_$P@i6Nz(s4V%L-b(YHX00ItIY1 z{o0666W3aNC_edaGW5offTqTODq!Wva|*XMxh0j1N~z}e0IBfopI{CROdU%PBw}zU zM<|FN;%}#D^Y~Me>)SpH*^$eXYu>2bWY<{LwQt}(n&aqOPxU!m1kde`VRaWTw%sOI znaO;F@!@}7`WC5dXMh{Wx#MZpOcB`pAdga`{4P%yxdC;J9h#IXOt@(tf_@@u634ZzF+g=g`S1q8|LyAyZ+@U|5I@d<9JyVZMN5e`VM^6gF1V_f3ps9 z65LHnEwsYl@@Vn7TKZuMzU2s6HOV75k-p2x(#h)77dzZde_UDr?}qWeLr$k2BcXht ztvb`|q7uOF2QfO>FKk%RH)3WzctcF^D(91RFtQ5=>XZq@ct=t(pG?AxBtZ8%*v_Ew z(A5AP)YIAs(?jppZ4=TxG=|qBHSTDabkO3pb&TwOTSYX#nT-)f|3!dmE>V>Gp+<&C zCFmftc22QM(%Ps>ui+v1bt5Fz)ZK%>uhZla4N@b66unw4_)?CDqi3y=gep^m#gf_B zhEvNJ9P5S9C=x4=HX{2nOVabg#Jyoov_XU#A2AJ;ty`swJ)e$ge82zeJF%)m`tYmR zcvvXbVA7%_=g$_56d}8#q68-SaD$cNhqHyH2<>Ij3=`8k08mnD0~GIBq6lg10)KZH;K> zYc1kZZDLiZ*B5Y{pBO$8!>42AM6^irB3d%=pf=xh-_T@A&QaPxt00E6E||iqUJlp? zMwxi8{0X^z0MJkV;=nB`0@qfL?mlxfGt4L(52!u&Hh;^$>TxR5=ZTu_Ti<;>zxu}< z_42{m2MJ?~Dyy)KLB>_m(sWz3qO{dEP@PP1;yIF->`OBOz^|oB2l`#iok}ZOK#8ODo4?RN;tBPukYgwto`zXFL z+GEtWoPY9RvJU~%1ElLg!^VW?t=sZ}tKnK3x;b*^6kDd@hmp|S{~(+A-)I1s2 zX2c%9>%0-%ZD_gX1Ry>w9qRok7x^bd{Od(B`9r{i9KaEgDvM_NOB$1j-+*#_2F~YS zVXIttxH-{bp+Dp7>K1|=pBShUlsuKN->1>dJ!5(fF#tf-z)Kp47xq)1pL2vae`|cW zs=4~+MfFsb4$*WbLyJ92KF8ugcN`7|39DPA%8cNSi%vHj9DEaWS~S>$M}gHr06~vg2%4lkLMvYa<ydt_hJC4yFpQ1#!VqfT06xeu4FOUmsSM@_?f|VSWI>NT(YB z@o`xWdLE)z@lyN2)tULP?i6EKb+?fI)4x9et%lmqZE1tS;eM4E+Fz6SLXp9A`JzsRBNg&&VX{g$|g;;dqOt=ahK!vuH|= z$BMc2U|5>26ip@TpD#N}Ha+jjj}-JL!^JOzDz}r(>#4?y-Q2wU-42+F%HIn^$mPEY zcVx^VeNS%3VrJ3uS=3BN(0uR(!#+r#23SO^M1(F*+8)BAg3^=CQ-wbgf~@a(n)pPz zY!{1UE~Fi6ovTI=4l&CKH}qCOOz&@|j#nH6l0e5cG=*Jeyf)sGG!@3{S4^cXu@6@U zZ)3|K0XmnCEmej}D{;i|@Q97+=?&jilKoI^_WvmtNQr-NBDXG)=Ey&ZIM|}LZwz?- zRCX3*vynVfAbD!1^B!YFL{iZQ1M4f(|CB|B0$yY8Y`uuK;>=dh5$~_|8^ZeBiMyyKn;7gAqlv^zDBc0!Ep<`@CD*m8&Ky zBdX{nW3r*d`EUh8o#}ZLbYw+fCUO0qz0frS<`Gpl7;%*@Q#VrFbHGh>UvB8!=s*=_5vd8E1Z2l7c4Ot|2PWAKo)XR@r{8w#C^Y z6j>e4CJkdkyt5doUg9ITyyQ|)8|PJY> zuIV2q`bBlwRciR12DA#Pnmub=dV_RQsl&$T#3)L78`G!MVJ&87Ah8#>b6y{Mr7p8K zVhGT*(D4ji`vE#4&~$8NvQ@?j#*^awfnsmD-!a83AI*KT>i$#$GMEk!^~9;wT)6*8 zKG3luXv&=Gr!E3GoquZJaf$jg{rO7|KI|iS12srTwU(%}jipc!cIkwK<3W~EtL^#A zS6XO`iS+Q*?*dtpVC5+4Mzl%1$DixD6+uS!Fmo%pm5%G4;2ScS>77o#rj2b~h8-U> zLHh7Y1s>2x7k9$YFrLQP2R-<`qU^P%x|mS{oqHzNY@!&O)!?FZg&Q;c4-$qKA}gFI zXR8v%D<;g2F!^rH>vD_3+mFGvGZ(D-D>THNH>@2W=0Ld+O>9p89{VW!m}tVj@(_y} zy{MZjtB;El1);}Ja9rE5$b+}kd~rD40^Q>uM+Eexm^z0aQ51Q0SBGB?DTP{tU%tQl zSP}Wx=ne#P_s!7;ZW`~erM>Vy=W3vmas)}GZ}}|>sQLU~X|Vr#pAh>lH(oNRQWd#G z0Wk?Kj1#g?WrJ9&jq)141eS9e@~ElwVJ*-jynlntbOgIj8(4+Aop-1rSkf)>xYqAR zOF6B`UTIQLeZB-Ral=)RZ?^d&dwf5?U+#4WNcbDZicfEv*u*fp_f9fXMt<3wk{x4< zK5Jopd=<}(*oy2ocX*OtLQWK4qAwSyCx6EMAVDibnIV(t-X$G%_F{EFL)7VnK2P`4 zbHG{>o1uxPMbd+j-@QFgWVPAvErL^*@%h8tCfX%1+hd1*lBgNXj&{IP^E++-1FTBe zf=)cfw1Zk9S9D{tvAeS`;Wk;^=-Gat<-5SttPtXjA940rk+s1TY+981FHdR(>op7` zFW(o}UVlIOyzDkh!Hqh#PyF2cg{NMr*mMB-5hr`LsuMY}mr%|QI8V+>f|uH6UbYb^ zZkeC44PhTb4Q`z(%d5C(M4()DpJ zKM)?4);Aij^M_E`!ie><#X5c1a6u}IqDhgsEknvH?3^u=4~@E8Bggq{HVr?3$@lP z`?HX2XY}ACx z8)u`#DBfo~hIcBrI~G;$=ilpFk`IZZ34RrsP*!h3G zjK>QaXWB(S%jx*5sipWC2`|j^83l$LGSoF`6n-K+H5#3iE*>6TZ0Fr z;)Wq2kwPPOyxQE@-C7rF_W^Btvxj%18cCLB(i<4V#jq=dxVuoEZV+UuW%s5@z1M&~ z-RQrO6Im2!5TOS-S!pDQ+=>Y{J5APb%t>a)zS{@PE7~8*Lwckw24uj7@mB@k6s z`_Q=(hrih}(p@;iElBra?0T+xa*O|5jm#;lEWtDzb$3-_?x?f%`9Y~C z-Upu6enw$?@~NF7rc~TR!Hxyy(jZ>XbcQjKj%y!AsX#f7~7bB`z(|iN2 z&`*)JqT%H;^+IM^!S(gJ2VY&B8ZNr?icnSlPv`aCdprZhAl|eFu~8W zRj+OPj?>crJjQvsm1AWMGq-V3Da5Mx5sQ^hWLM47b+TDVWw*@E$k#eI`78fP%kY<3 zOa8#Xj}BVkmt}`J!pW~eh-251ga*hXNQtW4+>|F}njw)^$dUzei?M~6olfPus~SMe z?aJQQW55RP?_aRTflG))b=>g=`P~dq#{`4*jDwAXUovVJUxOryGFvAJ&d`2NA#O8n z35XjuPAFZOzS(+wXo2n53H`%C9>0m*M&(T2<~mxp)pskPw5x-DzHhL{W_a{P?%@CP2eqS|CAgC6ivuWM;bl1i?RNHjsF(%Bnl7 z$KH=Cw~J@gUJYJgoO8|loS=D0XimLy^O#aUHm8;W9uMzjoiSJ%tO@>Gm(i18EJ;qt zoUdUyiCE_?5bsAlouAM@rAQoP&=J~^AJ9N^M^^AE`gm^66%mlx524nXNW4@RSC9s7-u!Fav~TYm(WN8~WK;ptuNVz3zvvahHX+B8m3CwT@bj_AovP z(+oPdJ=gtuvk1A7cJBafq}@NG>@dz=3J*Wi|7s^L!zD1|$#eRgO`IvwtuNi`4?hZM z0TXSlr88mM33KBVdrGNXR*axK}eFJWDq}? zs8LW{Tj?9jw3cCSd%@c@R1bx8MK5~{UkVy==2-tz1Dbgm8N zihZMi<)qVvcdAIx$1py>*D1u~LXN|qIK%bL>+NMzD+QETAsv(_2#@ii9;rxKWbBBc z(r(m*TtCn!hqfPgqRoRPSS*%UXzlKo*E&-(9mt3)!VPub+_=^8n4RyBD`FfH3 zYQ2$w0;SJ>o$xKEA)}&F5T5QAx#eC^De5QyenYt&w*S!0DInCc1zFvO+SFFqEd%FeF4o@e z`gYe8X@}#nW2yNb#pYbZlC>Gt*X49qsSxLX9kl=QF&_DSexcxJVpXYsx%|(UB+xFa{9jX<<%TT1>p}TB*5={Un;`M^CGE@=*va^HdUIvJm9Hdc~4U zpE@94Y+r1OLeba?N&g_$LxfEFp!SCH?&*rQAwd4Q&;PJ zI~C>U09{&ASLu@gX+7olDzH6xtvYrbLiFVr3d<>kl4uxcfHj;ujvS`1kL@%S3ed*d zE%B*4=J|{_-JA@7ktKKC$?i_aD)PHt0i5f6S(CAUGq~&;*}a<`OSe63WpfF$OTx?mht%#k+ z&qc6wKeb-4dGuI3bXr{P@!j9A{HgM@y@cl0vfz75x+mAKygGI8_yf4cZ)?_!>d-c= ztOM?Q4CMccnb|;l3>N)Q@bm#%M$Yr~L984l`v2JY|0KKr^9CM%R>E3z6&6+m`H_Tx zVmL@`+O7Rd6lU*_P+$-pC4`TZXLk?}d}+h`_>#HRG0*r^d_Oj=UqMu9vZ~IpT#!Xi z5#r?FdCAkN5%TWGfa}Zx2@|RfXN$AF(`A?%haNAmwJ{w8sW>?YI~;uP>UW| zc+Lq`>d^r#(KhhnZEWvnlD}cJ5Yl7L_50$|RLU4NRDwR#*Vq%f206F@*={T>s4z{Q z2jm<;WEHGEk1*)@uK$Ivz+4x7b9!JwBwc#6CN2U^_NnC!BmG-e?oTSG-dkXwb~s;{ zFe)~riZ#CWy)kY?CrO)DeRr6mNv`M}7^z}}lj)K$U^|5AsoCQKFr1NTzmI>^t{(s$ z!MdZE_WcDDYW0{aV9#SWY3lqbjx>Fv%XH=ODUS~6vN0I28n^VL{iwy`+=S!7aeXQ7 z1^e&b>VK_{ruUg~Ui^v|6TtneBv&JZVT?uR?b4~~a-_ffAP}ocgaHVm>s!r_IFPI8 zT*K5V7qiPVj#5ddRiQiXtkm1SE19^c1Upq~#$0?G7Cnc-UhuTV{()6%5#(7q*QCZW zTY@A?q#ubx6gfh|lS(vrk>~E%eJT!2H>|4B2_vDv;?=ZQeNYi!Z~6tPQ#?xojGIGB zBG%q}$!&QE()ANEp*%;h%5Fg%0je9ueLx*s71{y1auvcJBvg7o1~WSQqH4kVWJ)6S1u>1hm+i@`yO545xk za|F@p5WK9$;dZQn8iwv+PA(6W9m^I!XcPzJ8pnccl|y- z-Fo&fZxOnMq+HO(xV3(dzFBt-d#oDyw{pAb%Jg>dN5oDCSP0_`kP8R} zLKt4A@KKV4V6x&hhqV{MJ8v%%M28P?KYo?0nG zgh_CBzzi^xu9~9{Ma`YM`s3=Sezv`wb*E3bwlcDX&=rb&{AUh6OZw!-ee7&)f4tAB zx;|*(-uC{!I&8`2aSnyMt}pU*FjsGv1^b}ilZ~gTJL#o3Gi~ZD_8Ks&G5j70yiJ#9zzD>;s4n7@x*-Yy?1a&tvfg2Ye5k=6y%<+hC>rl ziFDuM0Ua{CLpVZxoYzbWaQP~{W4X3l095A}lN-(UDR6{2mt=m=8;Ba?P~gwe(!@un zJ+t+H8o|HZ`~xa|Tlo*5QjBnFF-|Lkvm(r}FgpR^ZwMc#0-YKy5O&&NDJiA*h^P}n z*a;*2c*c`nF1YCn6*mVHORk^nU|2bX%Ms}c;R3XM9^6jb^19tnK2eT_?#I}kc;t=Z z_|AW5io8){(1?*52L~PF;jm8nl!AK>=ECu zI}$2lYU|BOLF~@lC$dKtQcF#5ginjsF4DC>q}TWy=7R4ajpH$`j@CY_QG;9-N_U>s z3#n~(x{}Nm{U|tDJ8R;?|XoqYCx8Na{H_DA;`{}cGTonyYX41;OE>IAqpz#%43}R5 z_3SAC|6Jm3hq?K7G)|l^ z`PT#j#XBf0JL>B8WY}gX);mJCY6Rj=6ZpO#7qag(H9ZkEPvkkuG=dq&WUD?DpoIh% zZzK0l;>l*m`S|G>P8~9^ZbsrLF4_6hNl2o>~bE|N5GW@VonX8#n z=@m<+-r=$0c%f?%(sN()*0d=U{N388=2|-psAJjAZ`g)WaAobox~WmEM5pRvoQ_EY zJ1wtaTqS~+f?Y_?Vk8^ffL^0CpVvCwj%3aeZT*3eQm@xz@FM2>>QdLe=?adAvUKZq zIgBHg)cqC8#bvd}1ZVsN;#M^s;$2VyPd#S3X@`6N)Hhf1N@aERar4ZLu!?k0X?Hu)yrLdbPvd6bhzSYCi824;Tn_0V{9*}nhPn&VcGQ;*2BduRn!U^m8 z6U;=9qQgrA9I)X&uKEX2mDF|he+F(a%03&?dc+cB{x#@u4+AM`jG+69YyK+nv(7!uYYscq2It0`Pmd@+u2pOVO_>__lfbu zwZREwfrS2?2eN4M61m6VHJ3meDMr|w>0{l{;|eu(x*N!pG`b393#vn^AdDP4vU{3u z)G(agMjNQPTxX{382_Hg+1sN+GUFuV2#aExDA#} zhg9FmP}ykgOXqTEcIHNm%PclH9%#ja1fz^oTBg%-gvF4{}NsRK1 zbS;1<5DC&9xjwF1l!$Y(-uC=bvI+MtiJFneTL9ab|!LF z0xrhZ&xpWzXF$Uk*~y6`+!sU=Xl}f%jt;f?@+A{KKURomU)^FrJo6o%lmAgvXS=$X zgODW;0IBCpD$xULSMe)Xi96ld)Kr5z;Z$rqvMouoO>u#nx}^^|Adcg-iJnl#HSrHn zqSbB3FnU;yQ;3M>udJo{{h$*XDF_RP&GEK$W_|3;`wI!7eSGywxBcL5UrqNdB_XVZ zrAoEM>Fvz=wEi4^UI$m*$uz2kd%-4`IO$HP(bbk-JEor3cR@y?!L9p(H>E${AN*u5 z1?9>0s;6fd_$y)&=RaLvY0u4*!soQM&1hzR888 zlWT#z$Fb4dwV!DxmnI|L=C&xIFw`g@!1ZbV{bTupxRJ_-m-Ub0I)_0a)pe+?%N2t4 zIH0v7nGoyGuMw(1~F$ok+f|T6~$Doy} zO&gepl^Q|7j;+?E8dh&f7<^b^%x~oDF|>jB4K0RUO+@4A%$P8wc}RqCDm(4%&w00H zBJkAj=uv3*Xrq;q!1_bRdm7Y;9Q+Tg6UEZe(ZzJtn&S;N&6;cTBMm}T7uCr?y%w%i zTj2KnI>W}H`+VBU8L%c}Q77J-tGUc3nYziGR#+Wp?M)1ZIkXMyvZ z_L?CKINtmnI7v!^Nk;8oO@kzr05;P<$C|C98(WctVT zb96kY`bI6(dB#9?;nz)xoBK>g>-ODk#a>1)1E{0m#-;ifD{%A(%VC&Ii)Ze)S8&GG zoy(~;h%hhZ9YW&2yUu}7Se*}-jr$WafiTkl*JZQUb<(;HDh4O|ngtX1sF)n<3i0R;_V zWv4QZY?t%jFAHbqrCRpVkQkgqE=^gx-x5D4&5A$qcjlFBS6+NnA48O_!Zj*?Q{(l9 z`o74lO{*=&3#RSBCIg6rkfwksUv&k0ZAZ9I^EPp{XY zrg?6lEgm^bV}WBxuM32NDJr|pPrd8STaR+Q^26mUgf#Vf-ehh&hBEWiVInhTVWAD9 zz8rU=HScO-pXK4;DWwy$$;}aWdGrQ*%qYv|eynk0IH!|G1);(w3TUHjUYjqD$1)(X zD{c%}dEPwQuN2mLRGU+NL_dK3A_8lomA6_lBju~L$umfeju>G~p{9@$g;9S7cDOk+ zS&0H>lp525j1>kqlP7TFC@YdyON17nwy=65=7(y0nR_tDq;T?ZdS4pcMrtSga#LKY z-nzxv&pyiU_{K64FGcyB53{eyZ=BvEm40|e2Y z)1nR?-$yQN4h6Df6q?_m zgFRLGo+Q3%_~NKV_)5GN(V5VY`5ohXJv`b1YmGsp=wj#5l9K%6d@yhJ?Y|CIZsekV zc1gzcFd1!nRFpX_R&`>4({mN!)+%*n9<}mF-7%wpBpB)Hi7|+L<(p!)CYwl0(xP)J z_Fm&*Zrb(~?kJ;ku`>CtWHuVDtv?wOO59JWw$F-vC<3cJ$O1KN_egq}tA~`}_O-a+ zr0ufucsaxo7xr21e_ltHpced?Ovr?(GD)PTg>HJA*x$e~NiaPZ!|YVd&j%zZoB}4= zp;$k$$dVfexm)R`n~bkn5pr$Pzgm6PS~@!vlTKQN6&VZpiLGa%8hfmn&t;Mhx=B?3 z+Kak2S+jeMheD4Q`>(MBU0m`;IOU{ z{1&%9AYr0qixaG2&<9U0@-tHORLtyYI^Q7G(eKH`UkeJJ0T)#4z7+j3=S26WMjLa6 zcB+{@m1}L46}fc!3uC$QIFRIhaov&{3zlC@%C2B+d+g8wTLIC^lZo z7|BR=1hU_cx5uu+B9z~KgrVYrvz^hZd6+d5(SYeJJ@C`Pji&)M^VKDgy)pr?yEPUo z?q5EyZ>JuGC+(X2yu8pG;xkBU--yS)a7W+OrK_y8l+ShN93aT^#(aK{2-hGoA>PA=}8L|NYs6-^_bEAZ;& zV80gk30HVw+Ln9IHw^>qV-6}iHc~DOu8Wf|+gD3)vzRo-G{?6l|0wHJqli?}WfTdG zDa;pg^5E;d%qfXhA5n`g(8&=mxFZZT?w%@3I);8WKqx_oQg~dwP$F&Q4{4vTc90*l zUB&l@6$Le}WV<7L>LR^Jy}RRKI=z}L9@2)}-uc{dz6~h~NpQ8rJjwtqysG&;C&NGnPhs#){m zt4GR$OHQ_t=AED4LOxLXGo|x1xZ^Jto~Sm0)(OxxNk_f5K2_fOjtJ$f#T9)vdv9)P zQP4K6r0#$1s}-JDdV>*19#)<6jU}cCdT>a{k*cUb4V-epXYOpFo1-~*!JKQAtO6Yc zG_7)$np-SIz^At^y3(buNAeM0SKYhpGm7eSmlDxH4x&j>uR`nVmfoy@;$ARohVYYdB9uM>(SdEj`;@3D72QfZc!;)aX-`wjzMjNxTOb? zD6%N((R1CI3--nX8h<1B@;z4p9XFuH3P$JMlRCdO3v+tb$e=dCcAA}nv=RfSz?)64 zRV5RM(WCuWu9{=<%RCUdR9bf2z+PV4uzJ^+*t(=9fMz>U?vE4|Vej$S^A&f)(+fq+ z+wF6f^dkXgwYp+wn@U=lsQch2$0B>dhz?_7!^6*{QBElm`D?+JZQm7Ar?$RkDjSO#{yzUVv=9gb5)n2>nbeUU0yu8{fM?#N z`K@P)Ixs6=YK_-l=46E5ye`>-M}dD*FVsu*=SO~SDwn~ z=9zuF3)yij85B6WN2!Mr`!$n^p=ih^6dM*n?K4Ahbz5C)^oMa7n4l3vhu}l=OFq2X zHJccu_qEM`jGV=QrADSqTVPx=yUgY`w2q_8Hn}E4dkaa)xO0;A(YJRqM_`lKS8=r# zl7Mk418cufgz1|2!vs<`Yke;tf$-HyM$7P# zmJ3sGg+h>CNnvDO;6P%=TN-*J=~Eg9-hOx0u;llNm(z|1FWV<-lC;3wW@*>s3d?LD z6|^W?j%LOxM;zkWzlRCmN63r?z3Kvw9Qh1lP4mSwx(pq&7>3}&Y|}mk^X+ect@1sd4BN!$ZzP!Z+WV!TGsL?y zjI{)8T@fi)VHw2DoYoC<16X}ghRSReY7J|N$wV*M4Z^HUnq zQ}uQkjmgS+|DYEa6AW`wvN~ZGAnA~v;YyFxliRvVwj+ywRS&%eD1Z>t4hm}tc=0vPETuh7y10$q9m$sa2}O!`5bh5*ZV({!iMsmu&}rcV)Nv z*~!b`&B25VktcyAZ=AnC!2ezfA`E#qTtm3e_PfukBffel%$g+# z394qw!pxQj8fAY&4k zM!(i=2k**TnB9uzAofPGLt>L3_aKcfSsUhIZOghQ<-6;;ma}!_uj7Gn4BXTdGn++zpl`TcfJr$Vkw!RMotp{thee9T4xeC7rXmZZf=8h_yfY3 z3(?B~^^zgLm|zfrk`0nE2{Qw=waNqRw_=*Nf+cL}uoMX|O`;S)X_0tOHZn1pHGpN7 zz?`##YD3`G7m7cm)R4J1rX%*_m|G`%_L!B15&h^X5mGH^6&(Ye+N7~3@muaf1k6Y0>xdIGQ zkH<-J5|6EK;ybx2OT}p=n8LxGq9*AR>Ru`cQfdv0YN{&K6XGSW{x0Hvv7wa*X*9m= zBJ>60{#;;7p~r7-@?Tt72ny&Q_)l2~ljM92_d;5+7TUF??JHRv^*{TCOPe$lT3o!QipX8R{`0pusDL;S4^w5_c~hcfC{6K47RdN4~*t{M8LQv3%CVJ{@3QX zKe@o+{#TQgOL2y!ksk`TfRlX!ovS#5gcMv#Vq&>u8)#%_N~PZUM@w=#+Ewc_-NbPl ztIpggom0B>Kjr&*Kn_vEZBtruW%W4VXLTKCO&fMDt-Kt4wTMb%TOf-K_skOZI+_6dN zc=;;3Rfb>XB)edK>j*kUdO#($%ag__QTHIba>@sasdbyImg0264YW=JPFGHw8sFcP zw=P%cxB4$dc;&$I??o7iEddbU!c&n9S>m4s{=g=_N3%fy5XFiia787C}Oyn#E^pgbnh(_;a z=X5ukLloc@DWK(ROjZ`8DYKba1MS^BrrFjMYuyc}mc>odpEyz@{f7lpP!dZBq|fIq zlcx&Ke#A~%TkC2OmiG3hDzxSg&WivH?}&3~RH#kmW!?xxI2YSxPs4|Q)~kY04A*L#_gT)q-VF75*M@ce7~#7-K76nx2H~GMDIi)d=D=f`^#2R96`Wv zc|$W;oz=v9PvggDpA(Yi{Q1K8=)?DM_2I4we`FXw?xua!inAb_Tue_|HBTOB*mis& z2MHkP-D2+ZVifx8s>GuwC)^a8#4?Id*q7H~;)bhkL%FxTU-<;?HS+iBfqgf1J_G=d zgtIX(;8SR{IULuZ|>|mw40>_&$`5jzHpWFz=)AK;^Ks=6SPwKTD z>!=H!@r4g&5Iv+u`Z2+x2I=Jpta7yZgsbbFh+Yb=U}s8@e44-^~h5_-IYG1 z*gvYDV|X1c-MG%yYs9aQsdj&z(x@2pK1*9Ifm#|{H+89owvcdK`oRDirg=j=zL-<& zuOiSkS(%MZO5nIC+pp1}WLdfz_BIeA8!dCrfFwbEsw~mSBgd|0 zuwlX3tD7?TD5S04=2>f*mPS2BOw-28I{$@WFlC-lv`dsBC3jiF74BxPT`2w6K`*!y zCH8Reu{oyx+3FA6&Ss*c=pt_v6#!g-^BQxr{Tm6xps5Gc@1L_W>f&+CgD+<0b0n^$ znBS_}7>zy^B@1Lgu<1{ustcv9j#_*-Fz3qZ`Ge4ULG`*++Os;I zUu3%fSKKvOQpIu$)YD~okDWqNv>SewQ%iQCPX!k<;G}r*{B!22Ms?@-Z-r;<`SuUc zoiW_nAwtuFrrcMGI+o?o0Jp7l&Oy=Ttgao#k^^hFD(tK9yv04p(CP=+l>!-o#WY9F zbKV4)_9n86sg{>dCl7#yv4;1a|5C93>*cB#%*0OdMM0Qr`QN~!Fw}}9S@;n8i=?h< z)F5hHPEs)w`9EwzeTaU!tizu`m?6MvdRTJK%xh1U?`|zgQo`$40~1M^*yTiMj#Vgo z*#d?Gf!A>5a*6AK_O|<|O1;N=Uaaj`C2oR=<;?VymT7LZ$8RNtJ7f2UubUz~4S#M- zH0FNUb31oM|H^$UNn*dC|CvO2$^@5+zBt^voY(aFYos24ijO&KU}-Dtwb-<#$^d|mWGYVuI$fn4`bupZUHVGW z=Xu$eeDBoFA-z)U=b<($q1g)k&Gh*qo1?eXJ@%3#zCGJ(E&SgHqNKk+lO_63`ubIc zRsJuv`CpHovXFjmFuENsdpYBQk-+9~FT_tGn@Q9T8GPe9NKraM zudWdXmI3=#&a{$yb3ZpHG0Y*zOV8_~gQ)xLvT*7k3huOQ>%m-`jb3ZclPbiuH03MG z0mh#~bk*ni#1&msvGWG1rg4#-5nEIW4a@SJ)e=?ZP^yMW@EAsYQyi{TRBums*D9o<^pU&WY_1tqj$q=6cA$mO)S*#29Paf-1M*HmEeiQQPC(oLXo zOM|EVbMmMmXSvJWF={< zU@crL_(8rG;uIgs+iu}idpJ8KzmI2O!`<&zaInQ+Zru1`a58jz32ukUn=$fQ(gL=@ zOtg>&y88iNwUEfx>{dotAYsv-_q3W`>Srn8a@ zT`S(UM%5}>d<{wOM9D$q8VW1h*DFamrIp!qiH8i+bY8U`nFkzmg>L0hmW-9@PyD9j zS=$!4TgPNN`b_t;T?rsUOCuPrt;r)mSgQ%L+h=QFlr8O~Yj`?DBf7EouS5R>oiI{D z|Fna0S#IuH|8ES2Y+q2*fqEQ#@u^@Q^Qtf)d6N1Q@XX8yGArF1V_A3j2k}qm2+1-A z2l*T;1{H%sV?*!St_;HoJbf-+cbzN_1kGWEGpYf+w+=5sj97Eu>tL7=`}Vs5O6dgI zUws9I+hGR_o0sEJ zjbdJ}!w#yo*1*YJIsp7a_MCZ*0&Uo z)#zs`m#&rduPvF4QQx|aBfoTL8^ES3L~7{jHyP$Vnj4Q`#Lt2#JM-B)zM;B-d5ed; zP8=jLMORmK8(V6Zfoqbo;Am~h^Z7hYnGeD2M`D75PYco5~#iw z8UYkGCt~2dY=6%ty*9YHF=A(S?oX@c8npRKWJ$$mdw-*}rR4Df0Xcojb<`t1-H#7D z%Kh<43O&zDF!Iw&!;U^o;LWnRbAIMg0Gi)i^UJH%^ z*7M(Ng36=(3gj7=eF~dyhX6o_%fY8gx5zfF1kwtqefNRf0E(Kj;R%HQ)`%AOBc-GB80Wl+fVPkbG|_uTVS{?M-e6~Vj_ z99JF%yhbxgUs)nA;ggtX;7rT;!OoZ-E3VOz;KMf}yk!+j`m2@zpF~1vkx%+#w*7Z1 zdgmK|!k_4)H)i#s*cX|Ide+~KI+#PC}%i70}QD5k9 z;!o@Y!*ipVzIqo~6*5HIMgg<-LfXFU{hspNBSV<5p25~rQk_yiU$sY&LJKxXx8hJJ zop(?vp(#4tsNofbd2ctH!?I3oZLEjpDQW6#Af&gLKxC+S{4k#M1}?nJ6%4|pi_4;< zl;-t97)pv?VP9qK#~u4_O_Z5QuTX{S*-X1KHkT7068NA24%Nw1JB08lVy32NE4 znz0|c-;=YDbUgZbE$p4_ZEbs;*X-#+4-MQ)Z>mgaE_B2<8>aGgVa^8tHfV>cjE0jg zq*{mh{v6*ViPetHy_FM{UnFq#(8?WaQmxBwNO9YV$?5{f1T9y!sF4q*@3oS>Ottyb zIia6M8W#=$#O9yt8PnU{0;&3C*YpAPa)i5V!23xI)f*DMeD_i3VcRAN9!^~5dd3k% z^)#0xj0pE5iQ$?@Vq4i!j81jw^|;lNTKU1t2ob5of}2819Xd_blHPEeM0nRVMe5$+ zU#b-2hknVX^XLNw(LnC2hbhaN#pw>^- z=o6txU9rmQK|>?jFbzB>m*(imjDc>iFo9(DDcW0fT(P;G*;B3dn_wy>CzLZCcBdcp zOK2TlG!QINu`krmEAF73KLs2@dKQs#bMSDu8j|B-b8YvhP^}>#h-xda#^eD7>yQ{? z!K)?qTXcSF$p_?&{0yX@By*$Ly$vM@42w!{g5Oh&Nr0k^-rb+|0AZix%qlffj;fMw zZ$8LOtT0Paee3KSaA~1j(~=pk$>K{)aO+rwW4S>Kv(vFMw`HsUVmQRidRatb;Z3DIb7?F&o;v@f|D9?EL9o8yOS zAgvRlD0T3}&a+hS1WrTG#(fruQ%7^`@~T=1DZq@5$I_*ly~qf>rwU1>$&l#Rk)h1F zi$X9#-I!*}!#Ib0=dydQA=_2Q&3b}(4;4Af*%$kk(klCP&nn8lI*|W`5LfkJZffTq zmMhx-9-<1erTJp5ZcwCg=emefHw^HwoKT(nMZ@Flr<9ee)2L_84n;La-AO-J$D&s& zz;aCnPNaBc_T}G5x7KIQERQN&1Xbb7ISH`U{k zVM_lk{#jSbG;F)=3!mw=5AC7B@ne#KhFoaf{$RSy(@n0nI_;@v(aJ(Ac`**0Q!-tk;A)XPZ?JZDwA zVQfvKo$9W&%uD5W6Y8jy2lmrYRW%KL=35_|o`f@J#SA9@aE4OzioZlZU7hJ6yL0M? zwj3S&(j7Gf=;r#iH^gAvlUFF^4Q1`1*-97xkY_nddgcogou zC)=$5PtJ`^PnaslDu=EqXQ{s<$?!mE){%%+^D|VCNqWc~NvS3qvBLnHf_qkGJ{Mnv z9J<&;WJnJ;Hsx^Uhr>oI?$qn@6p2GTd9EW}m!0{glv&?M<;KYf35E-gi=m&y^(0MUqB!nNfp=@mNM0>9MJ1!?M1)TTH5JCCYll$Oy1; z%{SuBn6e8|RyQ{nLs5>+|Lj`4ZK|(9sT|4ik71A@iu}$pKKj@C1JP#q!5aqvvhQ z{%K(J=un`wG&##L3b@XW8ao-rMO?S1ZIaG8PhI8_+Rpy;tyYl=WR2M-Ibj$>Q}DK6 z(^0fzLw4hgH+)rDK+agrqk+y!UxIhjf<=K^?R_(L>Tr2ibq}q5B#?J|38PwFoUhZS z?Ia6@>DSFfrW-r4sVd6CBdgN+O~X*T@8xS*?Q?L|{&7&T?2FyKds8K@y^k?vA{{32 zb$R;X)181LmppG)d=ro=4(O-f^1pn}fBCq|2lLgJuiXP@2Aup?>fL!>APwj#4u3(! z9$&7^iEN@xIs|eEp?X<4yf;Z-C(!MqBN*&~e>uI!4k+BJqn(DOEQp09F!Fgaq*=Aj zqq9k<(iJyS%JLNAclY|03)2ngJL0cnmvN5{TA-yL=f?^>+RF~q7_|4B3ilHC!RDM? zJ}c@L^*fzi0Xg{c44=xs?Uxsrf#I(l%Z1hNaMEQp0A>x-*LB}jJ&>5sS?Srjd^ zJINdCER%2dn4f+7&sq8mW|#>wmDD;r5-54jO|%}W17t{S1X%{Pas*>|EIhSi-7OKhfG5#S6{%VcjzwSUtZ^`s`+wotkTqN&r6}IA9xh^CB>f-iM zd>4%^Z?dQ+$f;mW#4X16y3gjTpA5TD2~kErL_z;`5C!D}Gn51XrgJX3l}dDk1C*CJ5C8<>`$mPfFc zsK+eFhJw-AWJQr4tq7W{&_ACTWr_jL#?cG`RvwG%1J+1B$Y^lzxb67Jxj%$K?^b}$ z>I*0O&2B?#e5-C1P6gtcYW$ziDTqek(hqk7e(sFvPM%Ya6s`V#&>m=B`25QOcMrHs zl2Fyl&FB$uG@4xB?G23g-Aftf$Kc(|`*@a_#5hRR8*_F8kbA+H!dH_7L5* z$k-?|Fr{1)rRH@#94R)AIB!nXs_|Okz2UWRfOp1d9tC8uThf!fAzHb`6~rF37>Fw- zMVo{k{eKvHtFSiPrfayRr4(Yex307PR6n7{R zG(d5OKhO7k@4xlkeea!XBRP(1?w(2hOxflt4yZ9j0lciiy03 z+asD%sETd(E<6t{^SY=XV8}Mlhf@Cpr-Gv+Q+5ydJKDBV!aCs&i-ld&r4RevM=p8x z_Y}Xb3(F5mHfo>#wGjEgWEyvsPQOTxWp)TA?O&0M7xQy{+V|5Sa_;A5PlW^qrBl`F zjqxHByVJI@9MYz!J%XFJQDoFOlGe6@OC`d92#^rnj}~i?eZ%I=EhQI$Hvi=F3gSGH zbKC1ptFlb_!=28X9dB0e`UiPrn{%1yPYHqcp@=$Ec{1UeSXADng?!Lh15=8k62$IM zT$Qt{Gss-1$k4-P|CZ*p{1HPqnE&*khb@XBwg!=d?sfo|l!mzwNdf5$N%_@MRyD&xzXg&b|Hhh@l9r=*2Tv>07%u#6dr5U-`=b z-vZwYN@iL>XpmInD6BktJZFFqEox32^3J9Bmn^Ga-kP`2?&{TRWzGfbxzkiOC)bnb zv4c_RS8uth0bOxkmHv`ln9)3a{$UpHNWr$M+#Au{{(eeG(pnbxr57Oea#U_3vq2{X zXcTIx?ri2N^l9Y{pno;?=kfw)h1S$(9wn9D@)0tBaAGCi>~l!np?~C@0-i$D}+_AF%JqCg_o?m=?8UglwcDpTz@Mk4>%YE-nPQEZ zemqjBUmqMdr&*t~Ul}}YZeR%9h`fb&-0sV6Qa;|4pLhoDez)Rx9Mk?EEr8{^oN!5; zO?nvrUQ?$My7BW~8`ODJXmdY=)@LmfO6O-?`_?HZ{8%$HRn*UxQz2}i7BiGEa^ISP zVzcVL|0wDGqsQ~p7f`ZYQOs=1WRVYTALc%=<34s7#woN>_lw@<*0>yvn>2YLhL1 z7S~Ce{-$DO)OhOYPH>Q#(+ZKl#! z5swM*lkvkP^3eRYy2_nghJbRmP|*i@jnKNv@xkh^&`gCr9`eJThmyO!7`DOcFBXo1 ztr2!0lTHjS^wH-s;!`!OtUTXXU#r2i$O$SZa%S5_nR+UE;Ad`0W&9OV_JFIN^!jl* zH~&T!^Zf`E^f6JoT>NLwzotpwJ;joqZJb{|=BN!C4t*h2(z+wWu_Mmc?flYMzF!{m zYwZhKk&id!B6u*4`LYf+wGT%E&Qgs-{Dzo#uosiFIWM)fCOrgETKNR3R9_?Xqj3#m zpOJ1y1o=wZVx7TkOo$e|UM&Zl(72v-Qzoln#9_$X+IN8`rY17*^_jOn~!4^aAiy4$iO-twLx zVa5|#6({GeoFr7oAOEU^o@-Pyv)?4!zS`mQwa%b-8G+g2#Y}hu6-f0egJy0v>dI}t zEc)xc_$k|aQ~e79Fb~@6G(pWelf!XSiR9+H-8J`}Tmhv;f|Zkc3TGgukJ|&4U9B+zl?;z}xa@oI=$yOJ%I$rZydc`fdBsn*@07~kTX{CO5!ic4 z{F@#c1A4OK_{HW)v(?|6fGWuL-4ku9KGAD)#%I!={ipd89O@Fz8}veIx>nB_(G+BB zPTzi3o7y*wlOV&P(-&p^Q9`{hVP~@(FiUW&v=Y@;^4yi_i?Dz<;6_*$&MDf4|8^+E7$Y=ygxt!1s?T)=v_&>r)*f~9LyQMwqB<2*7SxT3S6bX7TA zz$KBrf-V+EE#2apyg*T+eVl8 z)Mq&>s3t<)XV((8_q^90l3S8ludVp@c1w8LimkEMYQZ#_#H=`&Smm&V*PJ5eW}<&3 zA!qXU2FHfu0DmSsZJEg@MuwX!5+xddjppq4(r|Ef)C+UsQqZZHsSvmWU$*9zKGQe; zqV;8xeBWu(mG=jg@tTf06wF7-v>xTUidC7_bx>Nm*xL{;61dBz63&-JhQFZEp#4u`ClG@Z;eGe2WZ}B!ND#7P z?EKf!{huUsvA6H#$!9)QM1YZ;L(X>Rw~L1SVu*-a;mE1cN&dG3JIEodJqkwzJtj~v zHfZJY*Jz)1kuxQNH5W4qkvq9Y%|FTAq zAX)QiO)nckuV&(+8WkO?+++mXh1su?A^Tx8Po9)Ar@IKR(3tR#7dE?E@2tu76GH>` zwgKnY()E*Zm-l_EiL(!rkBXoaMcVq39!ftB$C`~IYXdZq3-mZe#KctD+W*NHTqZ;- zNU9CU_>DhW{rk!%TIyX&f#3Ou82IAVU2qxAel5k{E)2(%^{_F-m*@oqN)A`MkZJFo zV(J(YaO{#b%TJM2BYJRMzz)a=s9X_BXUIKK`BLK(aUm3f2Ob<}9(F<06 z3Nq5ZwiPChzB7yBuaNUJnzPw{Ooo1-n))ZV5BNm{@gzY)YP!oO(%f!GK*H4SJtZWblZ^+YxU$Jdc$d z_5%Pj(K;u8WU&~2Ty&eifEzNw^RIE^zlvmgdxk>^F=eR>f$pKT6~mY^TXu# zGGGTyA*U3yQ#thW9Nib@N6P+Oal@8Mze*jIPwcD>mbND=`I=Mj9rWQn?!qZ4)?J^2 zdd6qB6A2;FEvMMJh%5Pcb%Ve*3i7hXh{RHT_KA#6c7@jOv|+8|2b!tV~X@w`1s-z=KhG*GA@lxi?A>rMoL->1D*}_;4cJ%d)$=o?_5?C- z>Fyh)mOzP&@DrQ)%YW+Ob9n<+)o&2RhN zx6b@Y*uTr#0^yX~4rjAu$~rVZ1YX1sx)5A~(Ca6boF03qV^o^-h;& zrlboi*ju))`{@>%(1Sv>?UJ?gJjxMzF5vi_pjx$!`SgG;Q!KguAG(Ehsv6Wzws_dO z`+SnW-2KFKzWMX3h!#Q!{_XF;@D0f(OpaF(+$`Lx&HEojI#?dB3^Nk&nRkz(Up>2D zq1ku2{gL|7Nr+ZD%GY6Jz(R1MX2?SKhb_~2L|R^;^)x)MzEwDe)`=T0`Nz?351|BA zN0}!}+=bR@BURHw!;>>dHCkei((3997Uqm1JNG}Mw+c!U z9}B-p0!}((9^HxqPI4CP2HQl>U0p7a8cAAy&2D3A&IfLdgE&pmzZ`BWOnb?(7fIS{ z*t!`WgRrt59O)Lk#BK0Rw7=flsM(>vNZ-`2n_+(H7GFgA7DUcrMgXF|BiZt#lCA=VmwcG z-fGPE_-c$e*rjIYO|NOl<>?Bfz-n6__)3V9gEuf9|I&ItRk8u#!D7@$cj5sDP{W0&2OqX)f;;0^d`|PvZn$$e!#CQ?!AUaQx0TwgPqrew)jFu8(`Z z-9a}#8EANT_!#42dn&c?B1>9n27qDoJY^ihkYe3d$mpW0W{^!1q&r@5IENSF^lf>o zF`+s)Vw%Ek(G_5>tI(Uv8DF!`8Q($ySu4)*b;+rj)NjfR2V1bhwYD$@@>h5!;@i=l z4d}I87}KlgKcm?#KOEG`Tn~OxpDn*Iv1)kl@$6mNh1B>nzmlX}+u%9gIcqGYcGpsw z7In+RC-K$m{PE-AJmxAX^||F{qRW25JJ3kY5ifuxvjW!Z%2%*DLQ;&zA$zwmKrO*F zb5?Dh0E@bL#p=UJuvx{a?To9$EnFaBF2B2hBqZf9rWmXc{4dP!z=ATM@0&C8## z8ff>W7-0AGSHy!+b$*StkN7<4+{R?Pvh4f5diZq*h!7Uh!{#Mc4e5g*@S|w*Uc-wu z6(&a5ohpje>$M8S;3+R1O~zD*E+V zd``!)gon*4hc^t5nhwp6KT0puaAYa(V$15e<DM)IhQN5kvtLzuZK9hQpOE@#} z2m8k6@){i$d&(M36-F~@`{HI@^s?WAke`CDX}NhA^U7Lo=~|%f=c!9BezQ zfKc~+FT*!Gp|yZZFOEp1 z4T4%UdKkbPLnZSa709LbI!UMh$SfpjO)2&5nD496=Q<~rA-B?lA6`GNuPR`z`CQ%g z?fkC9ZEEbyk|Hw>cg1vN^%d^v<8>{-#hP0MPg0#yWYr!4jvTUD6^`WLeEc?{0|tC% zUjB{x*Yy8?TjQB@M(z2FwR*q1{@x7rP!gkyMOFkemW<*siv1V%JD*i@xlKLL2D=91 zeHHTsB14`Lr}L|6M#Zt!&sYzcY%8U@$YJfzeAD6kYUNSj`otRSE49ZY_q*qT5QQsp z|3!|ny1qi9gb+U$Csx5nh(f3ni{Sw3q}HMj{Q?#9d9d{L!0PqLor|ZeoaM^eBHbVT zX1B9XOSi>i2NkPnc0lX;%?iV7AaZCK=>7Fq+<0k4*+Yl54|GNIxCk^DO!-n98Qd`q z7W#UHblz~Cw8KuhU$SJMMc9{mf>m(H z+9XA8E+4YfAJ+@byA}FB$Iu>ErLVw|*P-RT)!dmdjdu&lRS?UxXBXEz0m&e`?49i} zvD%p9Uc-3m*-JF*QyAYOSA|wFQ|EYF-IQetJ6Jbn-TPd|dn)$c`-?GIl{@O;mM%@{ z8z=ir`myHU`{99DJG?r7ZGgipM^V%U)tdLSvGW=vixmLI;IjwQ*56RbYraQcRNei+e&% z2)$55f%E-(;bdadU>Z{$t0W&YMrP$H(Q2G;WcYhC_$tNXgIrQWE~ z(=AtxE5<&i5>T#Q9qA+s#=eWhYDDC-Uh8VI!H2ID{A=C8gKzrH`&hxvN?5#HyYJ zy#ztPw^x9U2U`bhoB!$3)%f7NWhRkS04xr3ug}@HYbdu2bH{0UH4r-reB*EU8-N|# z3Rj>780#jH%#f!zjtw_Ch+r~(;_Y@2eBOsx*vQrjZh}mch0*X17z4<)L)b<26iZ|G z);QuieYBWnUVctzD^4Oc(#2ImVA#$asr}^CZ!`MVc&hpA^{Q_fwdnQhgg))MuC{wu zLehJ&bl7`H+_?}Ma1kF*e>{uf?`Q45|N16x;{o+3@mjS zW(|eU=-~{VwC8UQ*LeD?rN{ow(hg=QHbjbYc45orroFRX0{Fn`UW~i(8;> zU-MQ_2QIlum~{x2S;o+l&l6a9uqV?@J`j43VmbWC?H5+2J~pZ8mwfG6(fdP7YkF{+ zz}q98QZG+s_rrWWM_Hd)(O8DZk&f7*qd;8?*OZr3WIFoO;a}fl|DMR>W@R!KAfbg2 zkXLekpvDZUJo-lavJ_oKG&Q!+2LA}J4-1qU){qok;#%%F@TYFnI#ZI*{3IndfQ8kO zG~9%IerI7_XS~EMCwEG3Yn4=!x-2xHJN!+FSfH%+kNall28yYg#|1uvgN*^eCKy|*7EZ{7`L#Xlc_pI3j?OWQTlzS=c@n)SGoo~}bYSj+J9#=U>|ROXEadQEjZ>O2m6 zJFo@6tW57C-Cx>x|MS;e?(=p>li6IAkF0@L{ekrN0T21w3V(!RL^INt0v5|pTt!d4 zQ>+^=-hi1h@1p6k5ev_u2amBok zoHdR>vx45o551>;6}@fif{@dF0wb(o%B3a2>4`Ji9K2a#QIGH&P@d%)a{r|jw;7M4 zOg_hyDfHP4Uou|9xR0r6szo(}tgh5cPG{(jOetaVO(AKyX6(v&kl`J0{;sFc9Dek` zpng!GsL53#MPDp$+aKHJLZn>XM6w1!*G39R5hc0k7cV~_D_M~1WrJ_g1TnLxro-Q% zR%o7`uI_!`Qv{WdS(#F8%#Hw{c{k#GZv9?Vg2L@9DaCdJHoxNBX&u4!-}Q3M+mj|P z{XXSYX0w_Yt8{M$?H`(?kUJ7|+nmdc#BG#&3#fg-OKm9qCR^1vVeVe8$*i<5sF&xy zEA-yQiTeJY$fNwIu!_(T`)J9I#q<+JG)2=>&qyuHRv9pHZkmlAzo0Q(7 zcO)|fQ1z~t5)}J*@<^iKn2h1sQUOOUyKKUjworBVf@`+0Bdgzd&zb<2exv^@kcAZU zi1Q_CJVpz=BDkJ7dN7vX+jCo>3*YatEre%1%!op{nYUMZA>aiVb;#nRIDM}_-c;$pJ%2c zh5+?aAzZA!i$PI0T83}cyL%41Tbx(OQZ<>uL!fDB@I`CS0WtT>m6V7G@Ovg~2fbn~ zCjAUOZfld4ko(ivVJDgeChi%xRyQoOAna`I?4wum{vaE*A+Basco`_JX4GEx`bfck{PM}-+&?$9P}?}pV{!u=O2)>9@uIiyP557h|j zpJxUO@mtF>zruN{F|{J;eXlbi1$i_}zbY-5FyJZx{GyVEIaEIqvV6NlGYRhVZ}DuY zoj~auZK9~9R*H(mBTH+Qo!{m)nls{eTVLl;)2anK64{?7h*bi*i{b^b^L&j5p}pmD zHo9R+nESR*FQ65ccyDfhKY)g!CmS!27h;_Rz?&gV?;YER?v<1B`)W~N)jF2_nbw3d5&Sllw+}_ z+L)oM9%!D+EzzV5Ql;yB60eL{yhoA*w0$M)h@Okqkvn|?m#C$G&1TZbD0`v3I+47J zyc*<=w#$gG8AC5qNI|{%T^RTt3JabipeX??78Yy9c?GPp5PJ;|8g9*CI~Uwa;sFiL z>`SC_zlLTqL!Y?B6{J{)AL|0q7-5PcumJ2Mj1Z^k&qHE2kdNX_gh)*#9=fuHhCEs~ zwcQdC$9qZNw={eE2y1leo$@6<_Io0BIflq%^{4i1Ix1Vs2)K;yis{hnZr)>qZRyE+ zpG9gsnBuBUcS1l?JkZm3o%WJz4c+$a4=#$y<$8U2y@?q6ey-jTC?{FrEAq;^6*?^H z+wIt|m+KpIM(ucHEl`pUn&2FlEzKSWCCDITW_OAl9+3I|W7rj0Qf0 zZo%tapA`dZgmYqC<^|oEcH3S_>x{g$6_a>tgJ=Vrd#(JNW%@aq=RZG4A9a0xT;E@f zKjyzT+25eN{*6;SZ15Swa73oNVqM^`X`oY&_P1Ai6ZADq^PP!2x>T%TNc%Z%%#V_n zur#iKgVmuw&4KjCMg&3NXT;!-Pn!WsxG#zMrz}7&m)*F9Us?yVQs<^^fac{ZaRXMOM-)`Y$wKu6o*;7n8G`iNDyzWw>T%#53kU$)dE))U50WIWFlnRo!-h>K z91KcvN9Ww0pHE#n?F==WF#y|47_Z#rW6_d!UE> zp^%L~Wk`6x|?HX>| zBRDxu7gf$K&KkMaVm%WM|E6Lvr7Bajoz!4ch^%VdZb56ENk}r zcyuoDOk96lz7Khg@&n*q6bVxm_SPMa6|W#X5QRlx>j`OU^l1a@{td~?EMYWsRd2QI zLX^1ouVGZejKbIn3iD!T$In_ry=~7ks6GPsYixg6eT7C^*a>akl^IsAT(owiN@fRIRd%cawOFhq4IHZ%eGMD)r3 z+?uA*nre-K!YbF9C3n)1)*eAQ%70KO>MQxAe~_c}6jOlZV2ux?Zp&)&!%KwbqiCIb zRfsKX%XY5Z#-yn9kDEJy?O;Ba!LI2;!qW@h|ax#78rTrK_w)Se( zC%5%d_;N@Q=5G3GlkIP+TSmCAe<6DYU#H?&4DEA3l4cN2r> zAmvmcbJM=)Cb18=Tk#PhOzg#->ooJF?_+fwtf~R-@AttP$^mcX%m#&?{U27Ge{lX8(H0Su89_3QmR#wjs%-E-Cjj*a)VZ(PUMXJ_e7Cm#0*f9x7;cpR zwfm;cExJz*wN>ZcORF@nZF~;I9fh>mg3e>t;|AZq#9ie@>Rj+yIg%D<;&)g_d)xCM z7!_M=~rBz zh;T@X;vI?~)vRO2xLD&^n6P7sFb(9X&MLCpVqP%ZkPn~vBqA$$`eIT-_{t5&+ASzF6635Q%=PzJr! zJc6ehFkjRO_wjn<+GoyQFD5~IJ>E!6eg0DkP6f&7d2VWqMNsDPu;EJ}1wzHI1ckM4 zTiCj+exdB^xl*upXs#O_a!S{it*V=0pe^A@kmb;`Lbl54EMqCOCbe|M`>hsfzrv?yU7|Sq z=FeSg(!6FaU*zxBCzz*~3X z{yu{Gok8>4hawLrZ|1Gxy9})LFWo&AD1HcS^ExlN8D@BgOiXC>S*DZuxuNVtAO*@` z>QmjI^jg0$-$>4tA#{p_pOO+t>OriU9r0Fq)Xm`q5WP1k!Oa7AT&y3=}E7&^Afns=(VPcD8wHJkY4ObRNHl!^l z$Z;h>Dmp$-tNh(Q#@}231>*&V)@K;RfX6r$BC?X*8-OvAZb6wN- zj$@7l(VGO(1V_J9M(?#==s6g4{LaR?nw|2l;&OnYX~Y(K3Uh@|S;jIpWno z0!^xxQ^}eRFGa`v()hW<@+&J~?V4|(#KX!JY!jdTA?sv~xl;3u0<_G!bAPPFNv=ef zAh4Jfx8u}r8kv!GgR@OZ_H3+rE4r74oZ`V>f*n!H=O6;pjWy0p-Gv|Wom;&wP9ngm zZ!y|@eluY21?6_zafy1CCK1(TV1RD?P;W-fK#71a(j;CQ$xAZW-A1a|agQwfp%1Ir z9=guuw}*}cmJgAhgY`u1tO;3Z6ejoQ^EOiDHc>8veO3#$xw<8C33@?ZF~YQ^Z3HPj z_`Bqyx)-C*RBFn9&;)*W-M8+oxDhTD6a@H0YcV8_SMm3sO{En8TMpzuR_06)$pB`MQ zG`2oeDvGyXSQ^H=j%JVOFnd54#t+i=Bqg6bef~=9pC0^zD7O>{o87>-Mv&%*O2d2i z>n)8jM6ABvV(jX}GW+l>#bd#pui=%mt^t!l^udl7DW9D^wwdds_fsEC(}l?9Yrd2o zm*aK#^~eV>?it-#&Wk0he{i?X0L&ievNe(s41NJDOlTOecp|VP-?~$p`TVOm{g*BI z>x0I}sE^e*3@3w_9n83hW5=Otdg~b<*5+B2sXrrV?D6~%c;T7)uSt=)cK)mr~lats=9ITK%HtMLgZKPKwAZOt;T8 zBlNzP!=8|3afb)Kb{W8Pl$96mMx<4@4m$Ug<(uNrGzHJZSLhpNAp{)+GW}J~woB&6 z(Y1S{i;95RbfxwSMuIgC{GA0bUBBTdlUHEh*$wgZ&#k1#|J_t`vXQlf-4U zSZ>AJP*0gv2>!DqrS|JT9OM7S4W}ol7s7MqFFO8Q2FkM;w6Nf2bQn_jZ?4@fK1hCs z8aC5Dt~Um#IFywSmqIItDj&BT;cFF*Ln8*CAds$=C4n>4(l($UIqMA3f_9}rAi3-> zf`9PJCSjD_kQY=XcQPVY1;)W~^s-l;!(Cn+6M<*bvC>6rsXCyO^%DlT_RUdTzrnrb z$+0%Ytc^sfn9D*BE*x@L(t5!606X$r)=vMuHy!6%&|Uj>RcEq9L;8DBa=lDljgGL! z;HZ_)mP|D-aQI+5@|6mO!NdsR6W6q9Pr;Hg3lm`8Sp?1YXMZIZbLD!z(mBXZ_KOlvJB-IH%kC-m^BUv2nA{gf@~W^0@r_>P8{x zq;1*vX`?c-(+plV=l( zW$vqa2cIIru~Z2ncg|f-BK66bnHX<~qS@EJBm0khHvx<#(HwtqqBDgTGm_l!&!#>qnn7S0TKWT0hu)ucu}|84x3Cn)TZo=u`qMsUA0R`%l*4dq>o(l!I9x z8AHC4Ylf-~L3{~@S3RCYS;x3s>oawoH#Q!TweQznB>7OL>;9Ub1Z?T*1BYC?W zVLLj{4EzI%z{Lvw$sRsCA&iGC(S}Nnn?>4d@W%Ry-3$*IVFO5ZhA*699sH4j(4e-U zxwKV6Q&7oAH&*UcGiRO+z$D+z>%Yn|~(HT#%|%Mdl@70uvsT9+h2%a zlPc?8ZXuRm_l0B2rLJe2KK)y&R}05$UL0wr&cF9>73(qPge&*l71Kse&QeFMvzKy2 zwXFdUjjQ^z5ylr{J3so`!#d+70{7T0G%*yCZ(c5DUF4j9O>`uJDpcJZ4ok4hW zmqt+VJ#N5Zt&jvbU!in5@#uQfbGY(h!Eq}s)90of(__9FF}825r}s?Pu+pGpzMAlr zp02K+aQ82we+oCIC@BS~HNW0Ho0R&*<{dyKUQ2=1Y>NL;jN!1(?|Mhb_c|-?1!Jb4 zaglhX^!;fq$=QweNjPQ)gX8K_l0_o}(xd&+thUfhnTkcU9Wd+mG)mSI`fO-NTq#B8sZ9Gzo`VgQIz?zeOyf-5mi~174QsK74vLjAqoIoXai0W06ZLNLurX%{d!PHO$mj?#7x#|jOYwAUx{?h@n_me1!spbm*zB_6J}DY-b&3| z6*o5Q`%DPVWb}on<(s|mGoq*TB>Zp2PZvB6k{oP!dp^w7$)vTG{HT54sRCkeB!VgpqIEsAKp+J9e5_end!8ewKsw^n)c>`L0Km99y_7}(G z^-8o6G#FZTq86)VoYab;rV2p(%-T||)XOg^;>M;_s_cmbQa=(;r$;-K)`*ZGQniYQ zV0&XO4tCj@S_Ft1{{TC2X6*O&vw^zc)KJ0G+P1exZ4QbGQW*2M{=)MfAMDT;$~;p)B( zL7A$F|E||L(~k-P3a!`wZ=4$Rfi8A8@G47pPu!LPGJ=U>?RE=iTCJr1^~>q!buH_? zYX|M>?FTcCG8V&)Uogu?RK;`!J;La>0q%YK;C^c6&Uc0oMsZkIvgKS z^RTNwSHT~Sy#{dgrZYT0Yd>G^RC-fRA)p^a7rHpqKPu98vILw8&ecbQ*y!Lnx2%Z}N%K?q% z)=V2mex)Rm&_0k;DcD(onUQH=-mh|-V1R4a3`kJ%kqc;$1@dwYw6yr*gW9q%XkgmQ z(|p842S4>?WC%!0iaXPl6?%Ss^?7}I8P%uxR!ApX|^ zMysfZk^A<>Na&gLffA_1F&AN;TQ6g}T5I_sg&6h^^+XU2)pz4hQj)k1r=L|xMtr8J zJVl3nF`tuY?H&Pxt&X}6BZO15_+0cj?i&MdFav6J2scz%!oJ#8aEv7UL80SgtSxxrr8eO5miLN9q-mriLH05 zFGfCt_l(x+=?qIhDKT5?nwQH@CO{re5Um%AUWYA*);&OT=G&CJPD!vpGeXm6GSuO2 zVz`!5(YG8Js5BE5xu;L~y(Rkx+pS*B#t87E>(yszOJ%|DqHaec2+hViqoZDjkCz$b z{xI_r_El4EuI%fAMQoo1gk_N(QRNJC3S{wG<{u!;e?Kt)Ta@z;TDLrf3|T%~1(q$& zrb_X7FFi5@n?21d({Q?!|EaJG0;87gDGL%E{^6Y?a4+rbt3a!H9W+bxeN1;tGqi?3 z!cm;&9^Y1{H^$Fev&YUFfO*mu?{`atVO?$5kh@9nk7{@ z>s#LY)qH5qtjUJ?ghGa?9l+ElN@R6SeYuNNYprIts_`Lzo2)}0b!_O_J+g34uQ6%q zwlIQ>EES0IPTRO6MW!(;ff+gye>cZVlu^CR#sERw%nnh}!3~vjn<+08#X2b6A3Ir^Jw1UwE0rdRTkV0$0I zkfpPE3hzI_3nx|NK^gU%&TQjGK|5zko_$I*@P*CTln;RU^qQoq`jOlN*`)g+pzx}iQJ$Tq2%RQVq*0D?C`IZJffwajC*KTH0^*;jwp zdOcL~S*2mQ)`2ANlwq+si_+T3l7@%D@EmbPiuTH zFwE?1y+d1VqrT8O0KK>_nV8}^qBCeV@^>Q>-P^p6Igteo3+3a<6F4x{N?e$NFg=RD z6?V{$F&&<7uN0b*aK)R^X0`_wr!s`&HzySMx>6 zNo%5=8;C^JMdGv1GPs;AsoAl)FU74Zh|CZJ-W{VkmYC7irPXbjm7FnB;Gera*kYhE zeYYTE+XRt^mwxXI$? z9$xR)9)P~HI>K3#4}PJ-*lwqaETNm z5&z>yJM^Gsb(F?^zbIEvj5*P}&{5ZdAbip32$|IH;loDSzvVZ)G?-mb#&zmEX}#%+a!e8nUno^lI54sCOxjo=?|2F>twzp3*;$8J)bpaG zt8Cu~d`xk#?@=KHJ}tJ9tu`~xcW`yIUF?|CjTP5m4#l>g@JW9$K=@_9CMkW&NLx+m zCtoEl-gTiNTZ#Mpj)yH8O}W%eHHK+#i+!WT2A^KLo$Vjyj*B%Zw=KA`JmF36$(hFE zi-b%Od|MDJbMlC!8Rt?Wu87uA8f6XvtZz(Il(d`4qiyNj+*QlJbPjII|1)*O9l-3N zYirTVPV-ayjK(%`tPXUpJiuF@e2-d!*0)cJckY=YDbUHbdZR&2G|M)$dsA#7EHt=a zH%}PwCl_#%W~iGW6_~kH0L>ps2##hCgu622O}}qw^4Ixau!9cDbmhCHLT^-YD$yIs ztc0=qYn9=g(|2rv+NG{L?m{O}kw?iBQ_;Ko^~E${l7I4y&I5v~Kwo#J>CG>vW7Unb z$>W;ei|B#sb2MAFoqb(mDH?IZj%X8)o$4#BW5qiGW9VgGH)S#zZQU~PtI&JcOAi8V zV=blm6=!)ZAk)uuBj`Pgtt#_lgv4NaiD8OGS&i3wlHK#B5& zO4qaxoDuW$`J!-mMpP)IRTLTc92j$I>a7o{z= zh1z>Yh!H_-HDeQ7X|39|Yya|o<9xq!zK{369_5_mdG2xD*L~gh)48aKaSl}hm#z}5 z(gv1%4@eGUDNjd(v!pZ$An^3aZ~}6ruMlynd)|Ae`aq!$@_mD?{315KqWEPdB2-mU!ULT`^NDHu?H7igA?|e^`mC$ z{nG>AX}%gmEAI5Zgqi%0D#Y%`(Jur8t_WH3CU)d7mEE}3G@(RFbq=S`PD3v&^vxyK zAQxP4W=a&kB8+o`Vl%7dUVEJS*J~sYS7sN+f(jjdNXpB9CkG?TI^}?HaTv#teWlgM zhVNvdk1+tXJEA_zi|+UhT~3uw%)_x92ArK|U%gdB95LY<2^WoBN3C$K zTD%(Gw-Bo|i3evUB3Ont7}pG-sE8a}8kB_CfK{G2kwQlh=`;D|O1n7qp~r7(4l}>$ zu6wu6c@pSe8aO(_P322R4yo)3y2sHZ00SHI6GKnq_Uirw7V zi#+y}L?(}uB8nSWm&0tjb(lsO0o(mQDLZ$yq$E|jtUYq{+rN8AS@aF)3O6lRw@q(tL5BF>{>pNJq7JS6 zp@gdd#nQjjRv4oO*Nx76h1)ZfU>=>$>r&fTtEnk& zrr;alEL=k-@p~%rtFJ3mNlhk2i&bW?;qY?}DjCwT*0 z4_`}|4;C+)FyJ^vi@~Jz;W-<7t6sY!AO=ZVOyLuu}gY8=X!Rmq%;k?aIad+bC1NlsjLhJzSdy}m~ctflu$y9aN+w%GPg|E zA`rTwIWFi>2}=GIsOtc7LuOL%sH)xyn(;PsYgb+=J{%cSSGKP|w~ftET?DhfS%$ z2@+;}9ywUgdsf?h=a`^ccxCNiEO|6<5E0?Q$15!-M@@7gAHF3&G@Fa#!l$${lTMfk zb^ipi>#&86lTZ761dOSp2UVUT%FEBz|0c-x;TF2Z;481C`vTC`nD;) z;CY@|4j;8A!li>9yfZqq#`Ux7h(R19?4Ig;GezDudrneT*})S3UCTmA)dM$Co!v`u=NZH0hV zdamlMQ@&k#ja~U9tA(1bOdWvUmXOJ<>U57`m z{S9XK8%Mm3!${j6_1e*ZWq|P@UoR!ee0lXvUn5{c zMbTs*!d=T7Bef~Jit968G1@sNk=1!~{j%UL5VSTgd&{!M6ON7~n7e4lx{OH8#P9#) ze^8(wF=~W<5q{wV2aoedhRKW9_%05#T@?@vzf2S9P4wGaq#arj0)mDRW~ z4KXR>%XdK3M-^3s_6nCu5V)7e3pQr^yYQQC;X0!iF8@Kk3(zWibfrrVnOu#Ezui=% z2lUaWo({39*}yCnw(m_-SxbDjM;Yx)Bu?jrmejMaZ$C~6XBtQ7@&Pmo8K4vWT~?S7 zXwCW`s-|K99q>W|ocJG&4{nLAlsu>Hv^Pn-`1kJdUC{ta3MZMz{ zO+~Nd`DfTwLBDFO7llBuOXN|xl1AQ`LZw@M(s&|Ns_AyR_0Erl>NJ>Bu1+t47xqZn;ETa-Nc_~v3EVPeCZk=K)oDRz)=!!7+r9z@iD1&gW^R3=Mo-@2yUWD?vFl1X|e#nGL?__O+h|Uf_YQg19F89<@}9 z5ASbOMtAP3Fxf~plO@3j07lPYwEp-=Wnq#{MioULAuXls{>;VJ^-nvP!uY+(Xj8+> z?wc3l1AloglLwiWI4bx!**pVoce1Bzo+&p0A*L>j%|;Y1214DNlD0jsW zH_^2$NP(5>X%r%J-jsR}*{rO!R6nn|H_~G{&$7WwYT&g&aXx|KS9EHDiu*Zv`qo7Y zOV=^Y6BH~fU0y8KtK$ka>pTfoP%Oqq7pYCFJ??MuzT^jZt!y>TE4awogAt<)?WzVL;Ek^=^|1eYRHt`}Z$<#q~ih0FEpe_h8? zS@$4f%H0fp2P@g13hN53jnsMBfBwl#F^lhCbJjIf!cY*2gJhmkmD{jmD9u!F7nD0pW_Y-s)R8hLYhNtk=z+#Q`Iu z0{$kJ%qSl*J{61IDXi69yUien^y~G>1|uf%m7mQlS1}=PJl~h}S&XlFW;(pJKNLx7 z`ANAIkg1`*_My0I&S77A@b~me4L#@3ldSUg?S6d^L;1}&dD4l+mVhrseak0zmMmi_ zw*DngKf0QXg?%)wSEakq^sZdJ;r||#VI#DP_$Y+U>WMM0&yRA5;7ZgR41xL+TR$7) zP`WiLVcvyGInoyi$BZwv<-u>2lc}`845%kWp!5r9KU^2Mr4R^(N3z)ZHJ^~ruR89^ zuR@ic+H`hM+8s(9IXgoE83Hp0864tuS|8eZy)jv)@#JZcy221$@d%w`2({VLkAC}l z?FYy;+nIfDw|L!YwP<}bvZyL2qu6!*&SwF~z9YTXa_VVKuorSLd_p87o9~!E5wvF0 zHs>sSzMr!EONnDepl?6v1+JXOsIxflV69$l%4wqLIGl`Hyxw^(`ragb&{LwlEus9U z8ZdJ?;=><_95_;@#PqR-dcmHvpA}gJGnX0ZMcDB_Fngo_Z6{+&mJUim?u(7_rkFPg z;|gU~weoX-a@k`QLE}<a!YOsxs>#anQ$;<>5uDaL7z4c*BC*##=rQ=47|AJ@&|7t33af<)ScJUd1 z)NJHYRyyQtcIHXU-a63ZVHhl1_jYE%G)Z z4zB$^(^NO-&5vqJFMw)UtOm|teVAjWFkPIlV5uNky2NpS#4<)*Kex-_EoXbJ+u>OR zuT3jv#dkcmX1y>mc6!MlObCJaGc*)pzKyKcOk2rIl1;Zl8pcDyrq^t~s?nC^8Ycvk z4*fWCg#zJFzvnGb!sjB>(XgQ4WCQV-gC`9|1mKZ|?A^CSwb&J0nf^K4y}+=i;Z#vq zUT=dlrDajbaM9tr#rX5^lKPW?4-0H1JJhZcKUU=T&|g_5yK+YtzL@F4&+OdpCqsGv zMYu}mU#%PP(R@Ue{xYZj_}-YzjGe%vm8hqjf>!mo^%<;K76Q#M&NZ})aG=}>(7glr zJEUgseUx=0tll)<`jNWOZ;%Kz-~|d^Z-e*66&@{c<7Ax!pot`1&qGP3gL9Wcj(e+J zJ3y$4{iv6#qb62J(AwZ$)8ig_=(I1xpRzlMpi(KWFLB1kB{I2t7|Xpn2?&a}nCjy) znGNwJYT)6SonYh-ZH-qFkBY~iLv=BpV`wu5T!=}07#;6$AcPnbMYK}I@EZA_fl-#f z;+DEJIR0rqqem7A_}1{zqv86yeX_=@4A$X2n<)hL_@rih_MJ_-wLx3Y$PlQoEch}5 z^l#zR^s|MJS@wBfP__JXoU^l3*1y?@4C8zm)Lazn$#FlcP2v{~%rpH{rN%fwXgrS& zhr4_$e_uQ}kITWeeMs)Li2M4p^sPKWq(#0Zl2E^$YxXB4$kgK~LHE?TG8+qI&uc1n z9pL!_?Rbej9%0cvKuP7mV~lWAi~7G{M7Y-jXZ2fe@KO+Q)JxQ8(eun|*q)LDfKm2> zUJ%zbFuMzh`9jqK_67OeD`bj(HNUNdulHP`g8su!f@8-Yj!C00R(2YEA0@H{tn!FF zm3M=zE_V-QTu>$fm&}?%hJ)Mc4?Wi#x|`+lw_!CaaQke!W&{f}K1VPi?Il2t`aq|{ zL@+)O`eLf5*yKUI(H=ftN*{V{VZJQME@`!y=VEXD2W0y*jmJ#Fll^zsgj#E_kQ0yzL2(b|h*Tx4fEm!;O65vF zKO%zit~`=`#h&EQB@P8JXAPj9D9n8@iJiNTiM2aUr%wbl)LhqTf425tt?yb;@jHa7wc#P*EI} zKL!dy6agUPMb5fb8M?)bGng;QEOEhmks#F0s$MB8rqO!I*=@>5R8c;)cC7s)>dehgLZo47zdpMC`8DetrMzQ3T$vX?dY09aHbTQ>i!+x1+CP5 z?l)Vsnz<9#uDkR-Q+v8=Pz%!?$eFq$}zc+dJB&t{5RSRY3BNDtf1{_1clf|v`yU$mimK{-OH-)=1JblMKfCE0RzFl$s90q-s+Zpf^1#JeFV7qwby%gccPu9^ zsaBk>UA26Z<6R6X#SfW`3Vjn!swdZ_Luq2XT8nXWpkPow9rBuRkFy74x`@HMFat^!>qyk8-XimC{+r6q}G`^zvHZnJ7FpP#q$49c$7+$`h)&QG{zXV`(coGyXa4BWmIwwsT2`T*ryLhvYKvBB@2+8% z{_W}eyiSeIsP;kp`BDQaLvgxhEH019r`nk}hW5O*f&aeu?5T&W3LILyRhT@Jc=Rhi z??ATZV8Gk23jN58yBRi;^<>z~F|RX;iv}a>)m?BKyz%qyC5e}qLNqD=@oG)Aq|Qq+ zd1J=SuZv*2Vn%ghR)!4cj}&>nmMYZ=`!oV7++Qx$pWC1;$tK>q2z52a8`~8FS}WAG z?>cGjzOf}_GQ4?U1FGfC7P6IarTd;5nVe!Q^dk;O`@ zGhQ>jgsLMxu1>dXsUE3(%q1}mp4TZZdzNh3R8Wz{fbFu}8JmcnpP(1?s4lDETB)9+ z5#CV&S$zD{xqkwYW@aC&s&hMnmgXiF)gRo=2Y(@ z9rJ&QfDB2l;%(%oDL@Pa<< zQQC8Pb9`UQ-D@K!B|i9)POMrn4}J zuM|_28nCaqX)R|x!pq>A)cmUa3GGTij+;=2j!2lb5*t!?&E^lW?1+;E;+AGIz$aZ@ zkYx6`LMI8y&hnRc6;{*J6MNTghBSQ`&m*(UY)J?x>s#@l-X8x&05Zc6hF z(Gz#H9PDd@=0|y1{pdA`2VDtOPm9ahq+S@)I4nCkRQ0^C!cil@P*P+q!aUUsO*?9H zI}DmR^Kv#rbJhV%;_`a%NkQ#-qnV=&yhS?4t_4v^3=Wo_XWics-8K|u0hga?I$77K z7lmiE>Wxdz^hTZ>1kXY&*-D2)oR=;1(Ol|XAyD;SFX?$Gb;18wqCgSZUSy~6xQ;5tcpD)TJ-a6mrP(jAwkT}~dRp|XyZ=lRd_imy; z^}N}VtRWof!?ctToI6~N-1&XG6ahMe%hgPJyPjBjVgA-o1GJN#vJHBaI1EyF9WU#M z_`Z}}e&Knb*Vp%L*;T>{7#_RnIyW$G)tDTgIQ} zIhwt4goFR39^5xb-M+9 z*SjsR2vrnLN`cb)dgngg1BR{nnu?guR{mE;4QlZst3Q_iI7{-m)mDun)r%kGy4u%4 z()h<=@kwllv|Q(*$-j-d*sTY!$IA>Fjp@Dfb8s<({({oq^PP=*Gr zy0bEq(V)4g-6ZM5u|UZ{sOP6|ucfI*)z>75b^RP_&U~QvF#@zEQR=O+UUS;ZO{w9d zn)&ezikjUsRyluq(6etqC_b)FQC{lgGWVG5m9eK{Nb&M2&BUMmEOpj6V^*)o)c1NQ zK_l+BzfZPn!}aE0b*d|l*A7m1lP@L}uUri;eBTP5`Wgqdnr#*QoFW(MfZz4oa!n#L zw(L1*wf)ZgAa1Iw;F<+P4X>U#qEw1TZQS!K1E_@Re%PN6p!eJ}xk=Gduj0F};e))B z!Fh&0J2Mg`u5XpvJ#zlW2SD$tlNB!C0RYsGQ{&IW{}|WtQs38$5w7c3LlrJmMg_-3 z>y(Jz=@pq|XdkexZ5tHd1&7<}plu4C1}R9V>i749pLWYbJ*rSE5(YdZ=B7;+t{}hm7^*xuNVdL*@s|+Hf)f)$EoV+Bi*qMr#qw$>`kNQB* zTy6)r8M@(>ZAjmMVK6c?NqGdPM50T?8>W$9_R>?H)xh_K>Q=)#`#cI+Fto7(ehN zfmS^&Ir=l-HWd?UwteBL;8Usf?EHIRS~zSITz+rjaAX!_3j!RR=U+u`j^A1-3pW}@ zEHvyMjz3`>4k=ze9ZPl?6#t{90m!?bYX3X4J{Vy2@l#$CwzA<(1SJEGhcWIF7yZj$ zUSPKNhSct8GueNR&HbOlQH{^&(6>Xt(>zJVN|7ApD~1bj>VxdJrHi)jEk?X)8iX9p z5C!9n6i!V}nLH-uNE#Z692*n9k}-rt6W4T{ct&+?jPEy`9dO@+&eg#fjUPImFqD3s zI-+IqkQDXGLUafs7`+wKz&m=jDWD`W{5uBz3x>wS9chpurXyZn*=a}TkKiR5zyOQ` z%S!p{^$_SG!EKp4g+fi2@YG$pUlrzuKE+Vm>?cJI;)6z(D!GM1rGk zh@>cf27p@mRpf-r^aQEf7t}`Ucy3xG391~5^%TBq-T1p#hT2rPCbTzB6l*=Vr;%Dn z6MbE^P)dmkv(J1CileL8O*17D_u(BGmE^+ekr1 zW07;RJ1%so`7`C-rV_5`8kJwg7UkRqoYc#S$+APyu#u9frVHgJ|4F(&V07}`n|V-A zl^cFZW~R_$b>o5SzvGJ3Q-qW*C$x_)G`S0OFD8EjtPPwRnP#Lt>?aCDHGzWD3_T=< z>cqo+i256$=Jz5EMS-Exf|Mw?Jm1TGkmnSkD=Xn!?(Rh_UbuQYIyfz{s^{jDfJuN+ z)Su@29c@hKpY8lGN2dWEXyA7q9=N^k6=6xClChkbH-`3owy_vT6vSYnVn^(=9JD+t z)$oQzbvgTekJ-SQN?q-C2WcGw+9Y0`S9R*}*O%nu@=#8a(dV9IbT7FAjll5UbZL!s_zg*#^WW9-2u=9uzd)In09rg zxYEo(8JX$~XA<-GMAPaa9zBai5o8(9X`FLv>a1&uGsLhKVbv=d6QT|}W>r3X?4!_M zxTNpHWqC`GwcUm1+N2)xDOuzhgKv0=lR1yFtbAXJoTFuGp)Y{ndyi=p32K!#_LFlD zdAYr}lzGEkG!?u&p!yQ{smJ(fA4L0t@Te*yqNh0e^ zwc{&qeOae9PSDrSb~VgbY10H=jyP&$Vy9P&m_qd@1H4|K@s-J8l?$X11bOQh8Zx)4 zB*_Ma5L@rO-HoJlFY2S*le=O(v+lMTJv07J@{kfft*+K~iCEr0Q4Wk8%V{cBJ(u^U z*_S%qb35$*Q}KbuIJ#3v&dgXnik%uRy!2cY(-hGkLY7loO$G8N|Dtu;6YQa#1R>*I z-+cL6t?|#GBC--?xaFGMYiLpEJahMOWL%lpO{rQYB0!1fQgJ^-eGNp;*<)PGsdgE> zp29fT@P3^m{gh72M@~fsd*}7R#}$?a1JDwIH30^^=;i4)B1M9Z~{N`~F#y0g87BQc%o}=t!w}zs>f0bGDMD=*F@Mt_e-j{$2P#k)) z?B2EOS|KUfx0K_Z+9O$qt9VMEQTyZ0p8a(ZdnGncvg1g&ak%i<=Mo}J>@f*IuQQ;q zFg8g0r081tmr8n%#jP-@Uoq^A2iz0ht;O&>?G1^E8OC87jMb?1R<35?i}78xa1@0L zs6dd0l8_i4rQiix+s?+mJLZ8SmQvT1DlPS|J(k&$MK18j#=u*7JGa^bbtw@`#K#)!WCgmPIL00doZ z$5=}}fo<=z6*6Do^2@&m;cXmycZ{=IJq(bhb@4}-ph2507wSp>Gn8d8jU8I9_Z!nO z{hvAVzZ{g}MI*5}og6?+Uxje#9XA>{tc+_!S3t+EGTZhndfOaBE5KUQpbn$nWLu0x|Dbw{T9 z0GX$lJpZh|2~D*-Q3Z=o1g&I&>{Xcm8;H<$}>`5~JUEag+geP8nYN76mr zwXVMMD#+@R2C1d5Vb4ply|bw>u}yqkp>jY)wjT8kjYSr>S#H zv8>Cm7+esi{tI{AKsA1M9KEs5@#O7gCUTR6QJJcFT*p=I1V$(zmiYCje`>B{>6CJ> z0Q!bzBa)zlrp6iM(Hp^w9%tbpn)P5oAQ1jJuGSg&YQaA}n7*(pN=hk75}2$n#XU4R zJL8~S*YMWQinTM>6jVNvc|=J{(?~MU@gpu%bOar&C zXrzP4ri=EA5GnU7^!a2sK}z9toyts`(Hd2u z-&t<)>=jx{V21)b_t$8AfYh7Og8poB2Q>b}i5+B|_ncUVZbZ-&6Y8D$?{$8*8 z2i)6Mix)a3?JYbas(F60PB>YsJob1|tg1+l8WbUx0z;37*BRHM-#+06PH{rM6b;!- z5AwDYQB{qX!ftnE-9MY>W429^00iR!1Q%|4jqro^O!1b&<*FWEX=~8Bx0%Yb1_8Y6 zVc?*cn@#&)iZ?f|!PQhgum9>HzvQCj{_=0(kwdSdzl$$k6qSi*@QkS+t;J{~frt6i zs}qZ}A>+voAl$=<=Z)Ccz3vui2ntBW_tSYdY3UeptB^`b|7KUfivqx++RU%3Fb0`2 zW$m3NevLrzGxxW;Me``9Vb%n!w`uz!@9>_$%=hTMkG;eD9u`h2c)0y?9oCYew(@0Z zg+)4HqR=a4#oHjv;2OLj$-QsVa;cPeL^ND7jn=Jw>gcAT&O_@*#%lgW^@(G8zDTr^K-}`QF|~1 z`W?O9-#b3NySgkCOIX`2T*Y$q$mpZ-nXePlK5y!*4SE{=d`YCaf|4hutDoeuMLyoR)MeA(zY8ceCuIeP-u=$GG(~qcGRpA!TSic`jp_Z5g+f#bX{4V{oah zn7X+*iX!f3t*P!YCa-983Mw9VSKmtZ`a+Y2&mtxwX*Ei5AI;!WR`Tu?f(rQhVaUg& zz0M6uY)8Gb|D(x_RV3Bl;K!RtMb7-0aO`}_loqvAQu~M*B>_MGd^|tB^qnK^-LI@f zdEDs%PUOy|jZor*N+_`tKiS;DdP(hpmX|B%Fls3n%i(6BhD-m}b0m_{s#?{**ZtdK z1u0wBCvGlRKtT3$x1iia-k+}oG#@JTG8WeQNcK^eqJArX1Fy3MIbe!XsC!OCqvcJ? zStgF#xbl@&CZfo}y?lp{D8$WLZZwl_b*!^hI@?Y=hk3LwQ*ew=MJhP@d067wLzmsg zJA}rBc?%=Pr-HL|BoeeyYBY2{>JPbYSB*~D-#A6ko9@-~>cM?5eD~k&ws^VC_~0Or z$RcqXxP?AXAYNdhNkX!5@mC!gLmxb?+ZEZ{xp^P>JGP63dY~CjJX@8SBR1ly_&<0+ zII=r9CZmXEWPXdO{z1d}eEdS5%o$m=``(nAM$?4@r~i?|{3Qi6>s4?&Pv4MVa-dEX zSA|^XL=!Q#Hp1Q>vry`Cd!^(~f(j$eph>*3;)mSvTx@#R@Y*+ux6g(o6AbR~l%kVz zJ?wC2Tds9>j8}IIT!>?o&3o}YlVv5$X&l2gOX0;D-=aXaoNyW*}&9hVee7p#g z^zKks8VpwfSMq&!YzG4g{7Y(UVywboO(~86Uzsm+TDdFK)9+kB@oTLLj!%Z2bQ?s> z`j8YjJ1b`<&>GK^vhAKQ3s<4z;cKC21IuT{nH``3(CA%0)b9nM9S02HMMvkjLnrzV z-!=F+cOqcQz?N)#-OOJ9^z;2V4cTq(rS4SKOdd_*IJ_sZQ+YCAf-@p&*c>F<^;wHP z$G=H==@xa5WsL~=03<0I{YF80Q@6R8~ zmhu^t*S(-U9X>X0t1T)UC5|fTqNx967T_=HsUw3@Qtw-;m;K_6^!May&6c(zPkJ0a z$6L|~lZv0~5fvu>vHe5Q3Km0Ols)4?Yh{IogXoA%{hNi$s3jmV|6qVgK~>H05~2WQ zy(5l1QwzHi?@*o<-B|DnaZoRPpRwL+c`DlQ+&=;zQF~5d8#s|{#`PFMfr}<0O^Y&N z-T-Kp5+^KC8HnP6FEF$LO3dm#Q;8c@ra#aB!Oa*XwV;Z|q=p_Px?#F?%8bZ*x&?-m zUEH&m-9UUXBTh)Y9(hATs!eBX=>DJ3nRFXnvX)DCoH((PO<|u=I0vXw{Wll)e=v17 z3f>cmzkK~9Z=%5T5U>V`eO&QKYk)WWCpk_~?STDk`T)>EQ?AcY9_M(4O+$-%+nTII zpSW?PVu^?Y6#&;=+B=RXGk&aIQ|?jUo}hk)jyh8$9W6pr`&Pk(zM`^Z+2rBy zT5SegZRK4tdi_tKu4(Gxj*MraC#n*@h@m|5xs8H)QQX85^rat)4jZPS4Gx{8(VJ720+#=)-MXL-+=}OZ4vX#v8I= zrQz72!F<1qeQ$mM{_y-%+IIWUED5k%$PznuO7QvmDDVHU>;I2geV0Y@NAhfG0i*=O9XsD{%_{4ZBo>tnpLNhIISTK0Sogx4VpG;IypH)q%iP0f9t}w9G`tMx;5epL~ zpS~DV&M)}7m6VrqwOEl>A0a()CAr&mDZ2YX5iysv35e0J_APPFz;P;^?czE|ZkCgT z_>JupK(s2mVuvY;^LU@lU$$vK-CjgWfRcm=(ZU-Fh-{mh<2v&9ONNTlI9nu;$5zshr-C=Or&tjzOm1G7x$j*F z?^G4wp0?N>`NLb zsnB>_FZ)Nt&dI>YV}h-Ua6eP1YGLR2+x4K+q>_{pY(=3u;a=I8KCy3rgsx9vIZhAt zhdsnm8CRAM8RWQRbTmtn$VO!d6wvjSCs|*COV#v_ z#;~nazAT-W&YED8IsZ+ogN$+G!ut)hwt$b(4&}&$Qi+7TH<;F4PZm|8*oYBgOUBmO zWfE^xu|%95EpK3GlX$HX_+e|w3j+b6NdkkZ{< z`RY(H}1PwOy4u^rxiQ! zMEevCj+IqmKJ!uHTP>5s3BOq7p~(>w-CN+Ym+V8VUzxEbY8q+5y?2myu@1ZJzpNKZ zPaRg6cvF&72p>*L<0){Q5X`FSO#DZzY4u41SJ~c62ed8x6WG(`H?oKK;>-Gn$Gvyc ze;c%NCE>=<|4P^Y18Du(IMQLU-6dLruxO=wePp;q|71U1;}b|Lzwrr4y- z)9J;|d1s3?BYD@2FSstjiy;@$JBgg**eMO+N)r3f#n(w}1oO_;hIP=P#u(Z$p_IZM zE=S*c4bdt{-JYOi8w%!D24p{Rl0n`BFi-hYr!7VY2(ce67NfqW;~o+I1AJNrO)Wu` zA4y|J>-nr8>pXB0ZKB=l;{y#1SJSWfz+9P@L^fs$M+;AQ%Mmd2DlrCa*sGg6cP6wOwjmD^G$bYe zO+fpJ#H%(#;*++8`egz45m|+)&bSuM?iY7zIE;c%PRKSj7}jhFs5Qp2@)C+{Z9?K_wKSXx*%Rg!*83X!b5!`FeH>5p=DB=2LD!_MjWxw$9oeG0t z%gB1EVGswd89U7el-)hWeQD>sv)H_mrw zK;!khxj*VCly0)iz5gDXdi%v^w$JxGv?QY~70b&vL`;`{nzHpO7*458eeZ=?TFoe@X-%<2grBhX#s4uUKzwiJl$dy>Mt zGij!Q7EgT0g;Ak#tTW5L=}FklRA^`y&f{ywaZ9=&{?sXo3EoB)JL?TVIl2y|@;8 z1iL~_D4S(D9i4tD5`)lDQHl2|17w(xJnR`m6n)jZgM0pfW^K@Gp9FU=y3GY+Gg9m% zadXTG4quRZ4X#5;alC_Eu_HAp3@tw* zV-bksk7}RDFv{Thfb{EZNnOfbUWyN!8u?WQhSk78JU>Ik%Z2CTMe1dd!z-1qf$;UN zlfUZpcou%10Fs`1ashrJU^GG+PrW&Q`tRL>48wzgA?BZ+N~nZ0a;HA9|6{-u{R-6E zS60O_Ibr^pd@V;~$+z_Rz|{6{>}TNRvHUvXuPiBfj!ut@Z%{5hUyk#pZfuVwy8$m6 zA3khnRBU~!sX!V_R~o)Yxm`kLQQfWCG;lY}-^S4~##$t@T*;?wU}JFu5C8m(%0Q^! znxQW5RnL4a^5x_%b3Y&75sZ3EhZU`|$%l9~Pq<%zaEiMAR-Iqa)0dGczNqIwrP9)T zx!QzW^c0#P>XQAn>c-`1_a@cVkK`YJS0rD?jNSZovqO=j z8-K|Mp=VmEfjCanPI8jZg35?NxvE!+Y_%tJ(V9i?z(D#Tb1SkOd~m>N3l=cgtu-(K zJ0s(5ht*6jE12y`gMsd1hju~nK}u?AtF{7dsSp&6@MUScBFoTY_{we;~A z3yZcpPuF-S_UE?|-vMctq!uDp;$VMVDF2#cUP1ek;?9JX1aK#&GgPQcS>$U*XK#Ui zt>AhOq;!XbQ7wz@?`h7Ql_C)b*-R>vQb)#vSa7UPpu$s@a0eHe){pO+ZKU(~AK;hb zs%>u3L>fHyaTPQf-=Pd@iB8tB3iYLA__p^#+do~{%ebrSw1%xRz3KJV%1O5#+}`Ud z-F4N9y<7ZI0dNF%73cJ@|C+s*>DFK)IpZatmZ*JDev$qs-?*S66@Hl`ePwj=e?s&P znx9>j*aeI@@|lXg+R~{#bCb}6o-u}w-Gxo&8ESoMyFnA))AJM#VARMCU<|P@@+;4B zEyVqP5h&_9-p|)sG5zcpKbT!e6IiT)QFBz6Ekgk+@&=#HyN6DxsQl7s{F*ZQv_f^Iy2a2XB2P{t!Sy-Q##G0s0?_Mgqy zSguHnyX4!+bzYmdSSxEx)$#qK3J$EBNCIqczI^#Ov-(%=iO-x5U*g?*$^y0Q(`1D* zswD=R{it_S26uR4Gl)wKgxCn);@E@Vd53&4T|I^g2e@5q3Qs1s?!hlIi}YB>{}Z zP*Qwk6O!Xj>B2q5j}g9mKftO5f_QBSA9&%L9{vjYJ=O%)Tf*C+Bb2Sm>}VBI3=f?} z^bH}=A{fr5%<5$FbM|)NNHYj?Rv!gH|5Z_#$PE?98yS~NhmKOU)V`QxDCo3UbJiXD zV&ywci=fK1t^! z?@j(v?*!~ejZ5yH2pj3>JHj#%dM&UE*9Q?Rz1V>F!CML>cOrNlIAgkO-x2eE<)U00 zf7lx(jj>zT_o%MYdH6&Bu@CYc1v-nfm9NJA`;}bIoA0N!Ev(v#74lnY*UDJs2GT3d zGre~mxLK~HZ$oZ{da7g{-1)t8H$yE<9T{b9k81c#cKOF0r2C3)$;{dNYRz}?b#U;c zKa-~GA=|U&=cmk}BEj*R>8h0b1@i9$wb&cmN}hx>?N*kP`73O>BK1sxC3;|__4p%l zSr3}!b}OD+JGKTUx*j#p=wDmp@J_Lh>$I1{&)c4BiP(&HJdgg&a6M)GIO)@|{r!}n z$PZm1SL0!mbSg{{>M>RSGBg@_?e-USnoeG0exl)fYN`>T%l1Dc9-b-{t1E%{+}gy zAUu5^FNyJ*6C`o|JapIA26@%AB1`dPa13T~V<@~6zm1p_{vKZj-}L_it>Jk~vqqKY z;t$Qx)XU_!J(hFJFq%i@D57_RJE2UjzDUQs_juC&!JHsGOiM)2UXPQ-TURoPQGiYd zF}m``XaN3#2VZ0KO%Tp!JmMy4a0$TqCpRg+Z!C`X&=I=*zaaa6|DnIu?f(2&>tV+# zvb_)Ghj(ygi3py_MC!G&o4J%p#FaE|>^EO!AL7{K)M0J>X_`#W>6BvC+xQ2(x^#4o zL~jl*F*`cC-NEq)hMmEw|5C9v(G`rNCc&@YOUO}u7|(~A3wc1vp;#jQ9r}{T#n&MT z5`>JVOA^K-n=jvgFGopra|{$f8^+0YpkE%`aeK_{_oK_~?N!t`*QjcpRl8I%r{qyO-;}|-y`}gQ>IUcq^UIU= z3C=)z;NFeFo*#T%^h}&v-w2IgnYA5rGrz_$#~Lzqrabja7z=2v!QYoIhyLHU{bw`K z(6O?PjY^ra`S!1K7rbK4etlf#`kNu35V&9XDJ&AKWdFl{O{UyRI3cex%Gs|NI4o#Q zBPNhrjPa7xoK5(sowhIrT2^|4k@Zez^%J=XTxkv~4H31MOB&@uMFkreCy>3MMr>$% znWsTQjQJ5iyps4o1uTHM3fmZoDa%g&ZIYhwE7iiI##UMq>u^IGhs!0r4O>yAtW1bo zU(|*YrfvlIc2J`N-z4p|K6-}oTxk2jYxVHWkKCKTIEP})KRnUf&NM41ofCYVlfGS> zZ}a`!crMYr;G0aj5?M;^S`n!B&2I@FM~(hXWu1E-KEzAnAHpoy z{@o^ouV4JnkfZTZx%bRpfb@@_MNOSO!Rohskd%q`F|5%>or*PHLIOXlz*)NW+!Vt9*-NSHb9PfE&@;pS*S z;ps&fk`qbk=_~!e>^^!5V-k_uF@p8o68N*l0QMAqOw8(@QfrL)+f&{1Ou+(CtU+&{ zFTc`#ER&#m-Po$%-?DnmurfmRTg9MH`;YpKvNQbCgYv4~ocP4e^M%TxL2S(z_oq37FLq=Jm*Mw7e=l_D5J@Ts+AOPh!2i%EI-vhT#lTi>F+SN|_S?57gd4;-}Q zlA9cV@29u2KJI9xbvdPH8o@%!0Wl$bJg#P-eka!~I@c+soaaAj$F6Fy?fA4Hi6+7zqRHH|8H0Q9{*+x@vOF{3vfb=1zj4uJ*2}jN+8R~kveXcU zwKJO+gOAt74H}>V4KlALp(9$#fb8x=ZN7IYdQl3WTu(}{7+UQpW||nfTP~y#o3Qp0 z%`OtvpHYMq3g2t6 z_T^ktkMWM6t>iDVM5gwQC8Qvi7`)q9LBkiGAbivEL}? z%hvde@r3ha-KNk^85@@98nGLI+bI#Wf*9pJ+6E)U}X2L)bc zUccn^T_5oV<01eNOyV9NU2*^E<)56-C-SL#B@q&qR$5Y2>&;jSero1w&))(7&!I=DOIbn_$06Ue>dAx8%KZy`L@MEsb8@3 zDI;$e!|pQ{s&VZD{t1>QmvRovR&I44*oaP=Xt_<2@+6OzT|XuDJVFd{t3_JR9Xu{o zcg0CpIo0{JlXtNo=EnuG;(~6sb8m9vCk-varmvdSLp84%7ko4O=IiufWMt=g>h!-) zoO|(fe^iIfsDS*KB#<9C;u z#7?1qmzsvBzq&_$lOoRyPq)2nMX>>_FqYagscx22GGi_#b$zq;?X6<4E4uA@?Ml=K z$oABYf92=bV_7sj=RE{eINblPVhewHWr6X6}ty^~+$ymj8LT9QGB-56{sduSnOc$Z}d zVm;)b!u)@kG=H1(YO9!F#{}n$-`~B~xeg}2zDo&pYHLR{ZcSesmN6Ac_Isfrc-lPt zT-MWD;+5~!~YQD=Ta4@G+t+vhpP$k4zKothwb03<$ul* z{!#+Cs<%Vyrv9NQwYk@_<&`V(D2a>KS?fIWTr>KX;PllAEw#yX{%JqBRQ73e6Byy58zL=tWvDc<7yC8b zd=WTjKmEh&%V@XuYe?rOlEI7)73hO^aoy%{+(jd^<|0juZl?03QG?{f_bygR?Fx1X zZE)T_%jxx1ugM!#t@rD$CE*hXZEapi>lr!3KwWYS9pms!O{|j}WU^nON{~g~s5bAC z*zLbeDqt0k@3qJ8e$bO~o76wH`cU_h0iV<@4Yez$iVMs53N>Swd;?ca&kweHRvx=> zN&r{C*&vQ5n|cV0Q3M-t2zMm73=ue8*KFzT?g|uYvT|9V(!W1KtErpEC`d%xT$ zML~IE&#TNvG^{B-$$h@H@K#$8aG`vM4XR&XKD+8}Imq1pOUSNB)LWN~5?S zQS@_Z9Atyst}Ru7DiS(adM|K$_nA;!&R^~oyPDNDngD#a@8#FZC)nu&x|PmYzD=p_ zW{teOGtLD~LN;%bMW5sjhup2zzeXB?n_Z)*D*Vq~2;|v#`!uNh3CBOet11BOr}BKF+vCP$M%}( zuE7u&Y_qjEXkigs{+hKNEG+oP_YP!T{ty&fSy<10XS-Kc_UahBs&qwY-wP@9aGe!a zfN{-xvFDTHLkA9i_1yes{7awospCvrJwldVVXK*;3jw!IeU4{}OyBpspZIe4A5ne4 zGTiu4ZV!E&kT|_bh<$*0kT{u@66LL)b3yQE%i!RytN%fd=2#iH4lrBddH{Ya*L-{} zn$Cm4F*(~?ta(xpP%a?@kIO zn`eaBqUCQJQjfSL=9W9DEWSgKJAXd{Z84R-<)pN-ttUtliJtd7ijB-~eSz|AdQ;GF z#orviBaX6vZ+VBHbZ>Iz=hW8+oGbrPc={aTKT;JKu88_a?2~7D#-X3qydo0-FVnEk zboHd*VSWVHWY6i=Mygg^R-M#D73EbmJU^$Q;kkY9mrO_&wmWB%uYP$SShd^ro>fI; ztIB9k!Pinl9ivmTIkX37j|?*APlrNobsZZ(b}Ng1Ur0C-u7JI_T37uHi-A%nql1_Cs~L z=+f}-+P_DJt1rSm!Kt4ty>r9Zz`c{BwF;86?VNNu(@wSa{*5+r24bp-3x!UG42LJj zv)*}=gb$aje}wW)KAxNxSJR|)8a1tjef}6`w?~;8`PK6!;Nc(TELyh@?{|t$U7~pS z3x}ZStrEgdt(mm7B}LHIY0)hM$3HD};s*FU<^FL|&m9hbXqQ~lxE*>Y{k@NwSxd}bqmWW@P+GDLB7{49=Q+C)8@#U_b^m2@x&!M(5 zq|Kr`#5Rlx@sQ)99oMwQ4Uvq{Y4lR4x{FKt&E;k_9OX00iR7f48>n&^F z#a&+fq&w%EjECj6a8?+$YrjI~EYWP%?zyuK@u{Y!%d8aok572Xz^Wo7rxkQ-<)H@J zauRAyDxp74i6LLeRDxhh9C7o$rY5HNk>$Y3=EiXDRfeKo7X)YNaR=Ij zxB33=_d&;s?h0wlc1;kUD{mPEa-9H^Z~nY#)putcmmd>L7!MksM9Y1)_gp&#idplR zJ|&+y{0~K$s?)KY-#6}5+I|=btr(V7EVl72wBa5de&!m#`kM6+ZHI`&%V`yrCQMZ2 zp+=1?E`Sz#mrF*sJAez<6)lYvRgcRWm??>E-RbV-T`#)Z2@z-9w6*SLHg2NT2sf4HPE`7hX=g~lr~KXGLh@R-O&PyeJ=~vS^!U!9 zPk*+iwJr#)AxW|IYSemZlgbfO0=cR@1m(vB8t~s6m7lLrhR2&_L|B-^?{#REPRm~Y@V}1mJ`GcBl>5>_t#1p!%PbngUUl3 z>9y~n0evoGA(W|wn5YaCMsA=v5Oki}cT$L=bIBs;fP)33LdARJ&ff2s_sj~}n0pQ$ z^c||PCYIJKF@6`rT>3*9F$M-)ci(!}m^FodMD}uH$8e=QK0QA&_Kub z1YJItA(O;YBvaDMkS4!3{VjEFF5L|ap5+o`yDk1PylZcb_=y7lg=lFLheB?U$Q*6d z)r?}eZu&1mSAM4N;i%I!p{gCW&MrmK=W92oC1IpU!L5YmE8L#;VU|E%{&G$WKJ67r z6%-H5b{B}nU7;d2Qs})wE~PxNE>vA)dia*u^1;ZQr9~ALbf@K5QoaGGOSSw ze(PT8P)YE(O*#Q}hI%r1ex>1X^OMmt!@~KoFKP~MI>=rwRpL+h$)>Op6x0;{(qo!> z3vNjxQ_R=9SlcLTN;9Ekg349MDpc{?BQweOE;#ywbsrLGacX8gPKc3=v%tJy#_Q$J zOTY=^B0$o8g{KE!^HbUF@!tJ{bO)Z}n$^~@>Sgu>e&1|l#kfS}o11NN&I4u@M=;Kj zIZDZS3-)`x#7tkJ!HXt(qc=H5F{?1?O4lSx>LrE#TioMpreR-(OFYyfWH{u>0xFAI z&EH9Q{(<99cbwD0irCD&mC4G^l|lF8DC725i_bmYW)YZkM@T7`6EWn{OTg##dfOPc}KOY zXSFBk$7b9XQyHcVf%O_b#7e;hsShGG z$?(-NB+Y1|)&_!gT2rcSQKnARyTk>PTd8)?C4R&2;EuW6O^j*?vc>44Uic z4TNm5H2M_=GiCdx^{zx*%%O1eOLIL2e@|sqRavYnKld2dKD8K8&DPX`^t|SKT^Zi@ zxvypa19jc!o-*4Gh#3P}_p6F&-kQX?`~55r-u(KF7mf{es1h2nW6qbK_WYQxjb;@; zDRgJS{!7h06V#DpEzt<5ucg55GADAAIa|u2UaQR5yHMZ2utCB|?mA-FYd$$O`%B3mJ-yd*4i^V=TGu~m`dTOkj~QHGh{r9%0@j&FQ)THlVG8DexZ zKH#2%UsG}#$t_CfmOZi3ZnDb3wiODjki_}_veWYBICfxT1Hw@^`?n+VxO>vSr*1Ah zF)&CbTMCV-HiwkyVppLj1(YSCxvx5hzv8~?6z*jUHxV{aR@21Cy7y~}o2Kdcw2Xsyj7=r^u5`C*97Psd>Xk*7pAel0a+yQ8yZ%S5@9bN!fD_Pf+>eI-;s0v@cx% z0LsT#z5fMvY_P~|%G^O5Xe5k<5k0nfd=Axn4P6H3-!m{QYg#!zQANy7pB}cqJ6$~S zUE+HZ9H*kgd4I~ePfEaduIm*;;yDT?dp^zrqe5N;V5nP8kxQH1li$M4zx_mredvBL z+59Um(4iEkzvG=;t0Lp$AwDyGGT$bwJRx|JynNlg z=>Tv0E(c#fY}!hZl!1#&5x&UR<(DF{kswO3V=sRmRHV&Dl;s}-u~acu-q(n1U6U^M zyqOwmfdSfBpq5FC7J)1XBPbj%h>p36S)rtFJ|789++fd%EPEZ&SNUs1;r&2g{dvdvrLH*9e{5(5!~5>& zV;y-bzwFTz=WqqRe4S+|*m6ImRCf%tGSsmYEMa`f*nLpBOCE>KX|+p`nrNtX+(_ja zu?rVSQR-x3Yw~Kt7n{X*oDAK1y}DTYN*1-Vv#g6<4&G~mUp>-$v>vU>&>iCpQ9>z2(F&P6+L zIvoRVo{NCn_otQ7+DDSBVkic-r61#2^I8cZC7j5GWIJABJO`C<1N<-epzB} z{o84+#L+Q%AKSYqjKYpMS;5f`7ye+K7$d1W`s$KLOba?Xc{VsK+MM_XN}m@Mt`5O#%tDsmMoljLZLFD9Y4_9;d@EwFmP72GNNLO8F?4Q7h-2Nxw70>xz z1ZCI9CEvN(J(V5c?(ukNc1jiM1`nTBVe~MEa;2ZHdggGPBgIcp7#?mV4wNSDDLpfx zrnS;iCE$?Cbgn8--0UTFn)a&CMv-x)swik+8P$>^{~d_h;Jlk($oa2t*%;lgGU5`K zyJ;IFhs!r^`Y&;!4!t5!=V=f;i*Eplpb4T(hoq-V-b(7~*1Wrm(W?aCb$dzz@f|gg zC_pKYYfU}kk@QLEc0|LJvYAc7I%VoKEpfzEYx}+Wkm}CvUanLx7r)F zA>Z4g>=>02W=7?fNjxL?(bAHFcz&kC4@_0Y*9m+(`PvOTHR>ujk%|r-*Cc5J$J#66 z+*;mSJG^kkvy7?0;TkDbv2!DKi#4t?>Qfm|q5#w5=@aVPEb6`*7nXF3?%(l!FRB^Ti9x{e$`{S^nvZjI-)dJ4q*e zEt|qBFPIdh-<@cJw`NNz_|~(E0U8X8$`~`W<-rp4d>;Th(5h}_w~OA~?pA;2J;#Za zqgMAgVCy)CZE634@Cy{4L%?&fhgfH`}ALR~mpl8orv#88Q{K%HV@QjAK24$5Y z?K6ygdqPf7sHXluF@9@_JQnNc6jBZ4ADtf(?yPtu#eEhPOd=}~>r2U*=`XJrpo*LQ zXM@4T;Yo%P<9$qbg04j}Lb^Xb8n=BtcKHlsNRq(+_U8Iboo;=o7GM;_alrdbJ?D910*Y>+)H_75SMP7b- zOPP^n&~V!CsF1H;^TWHb*gHf;Z2m`l=z3MwS$;=JL)Qu{#vyWFPEYjD*#7L9&$8YP zX5jdQ{PCLK;UxxbWHu^~Wcuh+F4Sc?gGfXa7wG+Gh5YX-hBpfIzjNAtg zytb>B$l~*zl#K zlNHtAcdx1(oO;K`WbnyQ<4*}-w&Bmlt*tEV1~-D%s8iJlM*2YPS`WDX#t zrVKt!z2KimnNNi3D+2A@3cI-c*h63i9Iqg?TPS0*?8l<v7Nj&H}LVIV8hRI683cGC1s!`=Pz@P0j#> zTfFPDMODYU83{!C#>>ik@aB|9XoH^Q{LG-@K$S2x=98}9B+I`jjAgfi9&qxVIzgFgn90$4Fbd^f!KWpM$H-2b-%Rc% z@ppVX9hZl4xPHl4Pp_{+v#sTA^NysDLJI1#{XKKJWpJ2Ag-1w;!>cE{9c-%tpq+9* zh?emcsnx=HN`fE9IDdEZe&*zjfeF8{)yE^ zxc3+3B$a^a%@+SC7B0Y?xf;pF7U$O~V+ct4XF^9DO+x~c+62s=SU8-jSct|<TgK zK;Hu-Zoe#@*N1?IHuPt*wzzt(rtp_1@K}US_ZIB}lv?5#ZtAL0NxeSWthNZldz;3} z!VS=%V=>QCVEt zWWKVhV1Kr@l_I9Z{v(hwj8Rw|U0rpMNd=j(haS!;@6h|BR3g0jID3u?alY`THSc$w zFSiS60r~Uug~MmoI$$5ZnwVlPr)urg@G_S{wM1z)dWD~mieEsj)RpzLXzrfTmejzp z7|s=0M7r<^kKZ>?p|@lQ;M00k>JU6wE9;|Co^J6rH}xPu(eq$LtX`S2u44smm3Ulg zPBr4^Prr3F2P_deqN?c;lJj7~NM%6o#LjJRShK^b3$b(_=~8cXk61Tz?(xNhrJ5`p z>_z}yE2K&1wfvZAGUpXtDr$HoZy%Lgcl`{poTXZs{wqo9XEJZ8S+nu+s!A5Z8sz6T z+||A5W7l{k8PZB`_wQioAIrjHL$h3E8g#=}=C*!jDbHe2y{2$ftsnvq)095LuaK?~ zWFID_KK;6Bd1yd(P$HvBBl?PBKW)ThObra_Ucezc+TPaJPV1SKhBk)R~e(cR*F zoiR-lYXKZao%gQFCuE(>`NuZ)pk(`uK6H4`FIIhcKPjvho{vTJU5j zt+M#KlaYq}r#3Ikk~O+>b+Q ze82U?M4eNWz-|BAL~A*4inf(6KqkmvFvRw~tQ&%25demOdaG?ZHI3X{U=7#^4?;XP zwl5B$D(n|XZitt(a)pTfmje+!D;SWKPftSblv;Mgo+KyfoY<+SCPzsaD-j#WyzG3} zo;N|Yn@rgaEJLI@c!v>@hqd@gAyrw;DWZhMliXAqsryrC^G)nan_2H1EY5+ko4AGz zYC+=N6#Tkc<6=7-?^GSQwZ54lj(q*8j#rPvw>L(Rq9ie&duB93C`SURGC< zE~EA1n6dEWTZkMYi+6Z+enq@Iz6NT3k^6|+|DUwv1q&?I??X3Bp|s_NEBU{2*L@t0 zByk9|MYHXWyFx80Q(w5r{d}yb^4O(&{>fIdbg-fFRFe@FU^&*NYC3xFk9&R{WO16P zWm67?qGs?J;=2L8$u9UIO~X#sQus6GB#FpmoxZCGKccq^P(Wy%Klr8O{&r~QCDu#h zrY&Kz3RhZY+&qGrV(9X#M872o@fYptn6qXJi%6lwyVns2#v81jE(%c+8 z3FU64{_L3UDK%8#j_xjsPC>!!qI17VNN|KPAZHq*re6Sq^!;x!z+}7$t#BBf;$106 z%H6?dtfB%6{y6Wo#rUkFlT$L0w(GDpIBQcDfL22I4>Ey$H;f6o=JE zMQyoXDcKfh;V$w$qgZat2-39b5gzP^v@7%*?+SdNKu}Z+ej_0k`L4>LH~x}(T>Xu2 zxPx`xJrr{HIo38Z^T~kx`M4)xFcF>68>InxZt_d0g!E~R5l8s&fmXB@(7dm zXr{#mFSN#KnhU{pKRyh)@>|v`%3FO- zs-Fxu2a1$~S`)vrOjcolx42BU0mHF_srHPW1+_|PF*8Tk#E#6}kBVDu?;0}9D`DU6 zyEku|)Gxnvx%Jp3tC8H!8s^tdfrk_@Z}A4~4xc=m(c~G|_j??TmIA6g>&Q@MNxAW~ zkMlx=`;9x%hnkX|eD!P@t#M{GfcfQ*znXrW{JlfB6B4>ncS{~#r-=oM9WR2x>cPE5 zG-wg8^buEdLhiAL7#D%o9Py6bcB1+9Z(w#>J){d`xK^b?G^?DO_+#Hmi!qnuU`qmV z>4>&^?0JZGqhs6i`o%ueHqq#BKGLTKRpnQO=x7y23rU5s`Njxe%0YAv+wiG8)*dUM zG`k;9V29%ZZQ!(lH%3dEIGA}5z~bHs$-@e|XC?FP43C`|WOtRjGufr*Fm-|C(69L@ zfY`xCeNxH@$mzD)Ur{gm_|rGTDvaD93+;A;PhHQ}1osDNYbX2nZveW0%FX2G5|=jV zfroWd3O7z?Jlp`#pG3INMOBnOWyQs#OMA?6`i<1C!VBICfR@QxeeNLPS|L2`>tcuy z;ztE+6SQH=O{G#tNlTTSTaj6c7XT#;XkY%TtHdqHkD9MUy<0lB`P|R_DRcWjpS{&= zVNFM&Bp&D$2OEbJR~z|Rh@DM&+PJH_mAYRQ>h1cpR64hzXo$s$))7XP1;hNq6)q=h zpoB7m>c@UQ;Ip-FePKyIaG38`*?#r#yI_JLP@tuRu-#?KKduV{q)KonJWNEU;Hikx zZM}1AIT+Y!9;|`?L?g**c;$c;cS48+Rkxn?<;Vug>@`hcxZ^9ciwUS-T!U*b+$s;= z%Obv$cQ^VUITF9f3p-U!S~$9{CEE%U&C0LjGW{`Z+NfG#V$DOJLhZf=E)8c17Zk_! z=N1-%htiZ#`|-us#LZrfku!qK7H+1WbM0H>-2!TE=`qWy<~WD5%Pu&t>+gSoHdeMD zRsY^mWxNc@?r%bPk3Y2k7>G0k2n(&D7FdE5ZKZh3!FszRK!mH)bHPpgO}|t4jra$V zYl~U>rgONhdhbOV{<+z?50tUceB{PPq`T`JfIDn-a9IT7eed`-Tq=-pbReO+6W#T$ zSKKsqD3R>{qO5(sj;OP@FebFgSU26svRnTa6ASPt-{<-a{wV7BpN50a8%)FbXbnZZ z@`unzv5k<(+I-~6q{s2IOp`_@*4Fq@O0u7}@l-Y&lMs+0Bp@ZA>dn03`C){~GN zN{W2f!0HEnH3X<@%rlZr!YV}g33p?X2vfF}d`o3V9JUTTDcJQQ4iveGdZJ<5K-rJ#HLa`b{qVGtuF(c$W_OdbFesG(&V?5mxXCaXF_MkIX=|sHPjkt9L~& zY2`k@dt>Hn^A!1Wfu6iCEO>R!>hmK9B?ri7`q zD{RYC9C&-<2|Jg3J%ngtJ2WVtZTkvcZ;xwDNC?BoSiFpLdghNs>!R0C+(47Oi?*H9i*wDVzCJ^2h%70qO7qW!O6P2-SC}ue(F%%ht)_(n2y6 z#FLh0dO1mqyZfoJw{;0|vdWT=Ki#ahOdM^tqP*CXcA;1b(C#euCvWxjDA5L&ba7iE zfr8Sk9uzDy3=+nOcr1mUYf?=+SsQlsicaQwZ6Pm5U?8BspwXZ#`$n{n-zx@8Ekm?@ zb8$s;9lHCdH=y!ndM!|%KCKJTDp9|9nxX@?DIwBQRT^*4_(__>I#tl^U{`_f2y5iB zQWNL2p!{b-%4WAX@au_wg2?4Eb_&&n^Pm@vTePYVCW**|VO`-)S@hDM&#GKSfXzia4qEL&g9DFm}e5tV~r|AIb zq@mAd&SnaG#PnSzBVWn0g4()Ya2~>Ff>WW4F^4QDjSXRcZu{E!HsHCirh*&!EY$}3 zKDMSeSNE$H8!|PKF9-CkZ2HABdXzxK?3E(3S6Oz(nxy#L9)&!3e+|^~+V-L!dy5Hw z)-GxIF3NuF-tXo{?T*HtNYz%S|Era?qrF z`sS+6&YMf>3Zh?OTDl#9NBLIb^uhq5z$2K(1*PUQ+RC#Mf2v2b(K0;hvU%TT_JTg8 z%J~AR9wCsi+Yz>DI~clIt-Ni?AXB{dwbXH&C8@*sYlL*6b(ymvBt2P7vJrN&qHyVR zIE$~+b_cZ4*=B1SU=NaQ9dLW?5O}vKKW?=$qeJ#;k#YU*H@pm>mcb^`&CSffh2bmZeyEAf37y)8Q zossAq<`1fPd6g10_H#ah93(peK8LtvCU0y?M~2bw#o0V;t&|2m&y9&yO(;C0o`1oN z>mQN5Jx=wJox89*u{qKU-ymAJ#=gI~KE(Rd9Z3qDIQYfu6nEJUHJ=RIA31-=YwHWY zL@oF#A(}Y_+!f(jqQ80w*Wq4|m1JWCXgH`=uj!Mkz>(Exvu2Zpp3JYj4WiK;6+((W z#t~JQ%bJ3d!|FI(mG(x2%vj4AgwXObNRN0O{i5MiSo4b=>}bzek+z>DgpP)>YXzrEE_s|O%M>qy$PTOxmB~k_S{b7a_U1cG)eli zN(YAtyt=yKa35}k=ahRR-MJDRMF4T)@pjL7&Sew5A};c;Wgr+>g?S-*2W1-itTwdb z0rNoL3eN>}0p3-9+sP7Fe=*;I^N;DtgKHJ{sRyn@m&BV$$HshU9xbB#_cdpK5L-7v} zrGzh+aDJY+Y|J(JZrQE_wm+RrCrje;LgtAsIQ#Epnh9=CC5+PHMU!1MN0O@>#JtR^ zD#5}ttCQu^lar#yI1KAo$>}r{&gZ8r3`~t(UG1-LVC)ivx;~DX- zKp51@)E8GEnF~d?wd=3%j%(iy-7ImzdSuZfYeS%9Qu$yg-O0TNzmr-uX&HK-q3r6_ zzbei69c?b!UKP6eA5;;#Rb`3THqL{G(bsj&vawSWRlk@TdfH$C;M%fyWd~}LP~n;3 zu9XytG-HJ$LMWcU-4mkjbLB0C>9c_jZmMb!e6UOO;urSO!>o7*?xc5o>^ZYAh&*n^)qiZV@ z@0$MT>;B&`BBBW1iuZnHmH*(;Of|1jyBn~-D3lU3uHx)Pfq0M^#P2{I60;*OL z+^FhsxoL>oyi%@oc;c+v%y2}Ph5w%Kw7bf0);YHsjJVuU#91enJ3Cb_*`bPo`9?LglBys(F0?8paLp5GmzLF!fO#;{Y_krjdJ zci(%cv_%PeFvC$$DXCgSKW4pt+i>Fh;*F})^aW^n^KB0;y9Rn$m_e9y*j{@6f1+@Cplt^k*zEzC{Q-Di4q$-A!z;U0 zYoYy3yG1~~{@yMGSh3Cho=<=Za}ydCuC_Xq;r2b|i|S%y>?O#(ve&?9n7>UUo3m_T zt#EC*VeFb+QSs~I1#pc|6ed{t)+73zeWA<1gQ*?Q=`M#_u!vJ#5#N=@Gu zd4$r$-QPY)w;N{dDXm`BN=oh9o>(jRkNFMWoMkgVWT?*Z68u#{c)1)&=7N&C?Jtuw zJgr>XU~OBy9>I^eB|k3`$SO?|269UV=XUT$D;*IvD*Bfziea_gc2g^XMeza(2}J(U z)&Ty3kR8pF2;K{lOC=S?GwF8HR19+88GFd~MN@Ox7+N18eWWKtsS0&LZo819uh0DD zORO@b8s*K4jZp9aV3OVyzX#auEwZWbv}!cR8is03IWPTt3JP)09$UEm$3BDoYgQr2 zxZM?={i!SCW7{T5>AzM$eB)jdxjepidiFcejsrh%2l(%>z)_9f-v-0EeGjLtTF=Nk zsW@0hY1i4&KBt6timqtn4x69gcf@T~7wXLe5&7!2tZj?H5W&Nd%MXSxbzzLZ(@F60Sx?0|_<|;dCKo^Oy zv6SP@q%|9h-s)aD_WG8#1$}90sV4Vwnd{&>pzH5+Ddq^&$+#=#!7qMP_bHpo5^C2TeOYjP$&F%lQ`?JRdfEKe8W?3V=BH!Il zO+*w&w5f^3esNDiQrBO=1D+e`pRZREBWPbSZ_fkY%#Y?6Ai<4Q$DVOqR(QGf>4Y;B zH6oT|w>59_VF}8$|GY0#K^nlcpbIJ7kP<=5aq>(+^LAv1S;}=#4y<*sHHF;14Fja> z2G_8+=d-Tyb36$gWe0CT7;#_9RKz=CGvc{3z77XVRDI&OM%I5`_=zxxa2A% z@?Fxc@gkQxBVC7Oj52@l-LkPe0V~xCR?V!921W)NK@H}aA#-B&yF{UF-A_8d=4Zx; z$)@i9ojAj~fuYHjNw>wS2}idSQ~m>iDC-AyS9G&p_G4hLId(GmqiT030&@1FBO~A5 z*^y|OIkdNf^i1pfqqv3|8@VBaVngi$>iLuMLHj#kf@Zg!ZgmiB;+r@KrW*S$2!H;+ z9>PlDQv&HFrKRW(rvmB3p2`9H|KVR_)Wh4`?V+cCtoJs%nj5C1zqeP!N>c#uqXaS* zLaXi{2~u`(FNew&RTfT=WYJCo1D-ql3uVT4HG6wYz1GblT|9k#bCur(?qTCX7hu!D zTXe+KLo0`sekXLvr!YMyju~fH=R^Pr`7a5jVu(metQ6-r4g=iQ-dm8@&iKpvYzD<9 zFO;W4)Uryrx;e#rk>5tPWR#odgL2URK4q&jK_PiaUwR_~aqgXsA>S90I z7;hBTpkQqiUb=QA%}5JGT3R~ow1-+z#t@_z6MUsIgF4a8+^CX`p!sK z9#;_Wor|>Lj#>Y-a5HE0sFvelY8=j0ZjSwVVGOIEw&Mx+(1@c)a8#eP&uhoF#yfW0 zO!nQWKb|H$v2FK3I8BY8ew^RL2E_RgnWJLB-^ zHGf9ZQ_BcXyM9Wd)@7fRfjjFC1sHd<9HNx=$>d4?&Csv_Q5#Ev7faHpc*=GFMFdAfWS%h9K&RIRN4kKs^@>@7FoLK;TlQxLO< zCceK@iHzMvse;wK)$%i_9NSz41c1r9Am?lY)^o?qntrOsMwOToQV3aJ*Gy}(MX0=v zm9EOIG0MzJ{ZEC)vYPj;$I6C2gs(X?o(0wNN6}r=JKb|+DoGv6MXm3v?6n{&v^(y= zJGCh_Psw%vU=QJ}6zVoUW98!@p~NxyM_KRx1lg7&XWu)C7O@P^35J{VZP={l>iexv zil=I!?c(_wz>*Kn4E5-@TEBcc@wmV0IqPa%o1LIHnztQgHx?1EUt=#)ygFfPd}cJM z!`j>F&jrkXlg9*<{O+5Wm`v}l81PUrK_;ISAhAe{_7cAQ1G_HHYlcbU;`LulL=Vg| zEQDjBT2AbX-Ldb$*dZODKLeOh_`hcIhrHnWaVDt>7q$Bq*rDu+0uAX37iz4zjvIbU zD~@tUeHA z0Z~s1?7$_!)h(8_Z(aZ-4}wYjD0^v(W_j}k?wQ-4y-54AfNGXC-lij0)hUkt5vBe^ z`~P9@tD~yiw!W2;Qd+tsrInBn*ouNQigZdyceh(grIhZF5<$9KQlz^ZM7sB;`PRmB z?)AL)=zGWa*Ehy@_-haLe%7~{;oabfM11Z6T zNKjhp@JXw{eYvM(4An7$C|+Lc{kxYI=_RH*jvpR{FY%5gx~Q><3aM$MD&vBd2A@7z z6rRsW=;>?;D0(RaADaP-xFbxBJ42HHt~&rB17Y79oMc}JirL;7gwA}mIf=C~`;UF( zdVr+JICx~_3Ot=fP^s{#CT@~=RmgL(q)gLyWkm`D&GD+nv@N{z84i;7^>Ty7+v|<- z0v~oe=ro3tPxzhP5q0V!Nvux3e?yud{SrVYOCN`Ys*iWsxHFp=BD$n1$6U%b1zV> zAqkG>+49shEJY(x&@dE{VoIz0IJ5aer|29jMn)V|tFc&FxvP{`*J<63$q{NBraG*D{5P#RrQ~(TiUo(M-8bC5&0RM=1Tw>xG)p``vGu)nt$uAs78i@C=|{vO z7QEN1=s{_vn%a}vlZx&GhqW&MY^#dBrzSQxF(!aJoIapA_695%bXmI{^2YCP#enX8LBwdaa1Pra+y zob9=qYwzxPC!O8DX8b(Av=sI^{*C^B<_4e`?6@Hh7kfE{=@I37WCMdVZZJxFQ9t}3 zaZNN>!&wx7$+sceC;)iF!YG0zrocoHjg6R5q4?CNtf+cpuYSn_myl{Ono^&s@zc@z z%i-JWDuYISdW(a7WIp|uKIC}XB!>qWuCW9#?OJ{=eOJI5;uB%Rp{rKp32UFO7>#X= zP1sBktRW(hC>l>Qls5@^Ah*eq@7O?x4;|})^sha8Sp}C>Wo^Lk1;W4inG5$BV z!ibCOA_^*jCH`@wNtpbOMPzELTzQ-QldNJs8uo2xnVhb`We3~+2$e7KYTV)(&onRK zZ^;IzQf_;B2BRkYzxGMcQBd{Xy=9A!2_F_M$DO!q6U7UfY(9=79x=SEFzc+R8SS%w zc6+qT-5|1nHIjObm%}nN4w6g`E#ot{1<26jdH4h)v57*I%3MM8gcIDtHsdU3ZK`P6 zaAPbzsXdZy@8jMvciLzDq{9IZm)6tmW)-a&)hY46ZR>VKj~gIv*tsG-5klEBHpF<9 zm@J(wCj8A#p-6(a)DH>Z3@XQ6@ZEpSpD*e^U>Hw8*Dv!QA^K~4{n{A5^cxXc=4ju{&yK9RL-p3Dkz(B=*dd$-1e=k)+GdaCP$;fVM zIakQV_EU~VQKd9_!%>yzG5ekJdX8}uZc!YTeCckRWd_#W++6@nUkr@z9B=u%i!kZ| zDsH=+p0U4GfJ->|qA3PVet^?&J8O~XBhhvmN95#uwTgINiW%e`sP*{jLZ#;+?;*T} z42j^uObz=uNKIB$)i^g7T5h_+a!8+OY25_o{#y`fZ1-(p1kkR{QW2rw1mTl;Q0>p8 z|8C0vJ`MG%1o<-@vc^Qrv>2-$d4%B5XhDVD;Dbt|JkMC8wnH*7Yv;19VdJ z^$gkU-8*+Qzfdk&8T_uTtdus5`i>J9nk7wyRNHd@zJ4+T&t>RqTHQK3?TTdN7+G1_ zDV8MkK#j{&$2P~t(g!Vtx$A1DGj-ll2e*Rt`umpyy&R@!lk|`{>`&MW$d`npUxXxS zi2sXj-<4?WHgeD$N4@SiJ#ZR5zh-4=vRIeo77@o!^dxKIyf%^FNo&k$1=ZNMkyer( zUwBI9p-W8Nd4n(3*=;OJq|Q!uP^O1=(I%*#3mxU092l*&h9tQSH!q}|D*<|N79TI~ zs@Hib%n|+g;-7x^Xb#}z6%|TjV`KiaEb;h2jFC`-{GuwO>_<}SJKQgw?&%KW>KZxF zN1W3Mqr8nrf?h^P@zz4K=J%&j2t}%fGaQbz*rXnACfl4?OMD-?e^R!GE*Y}kzEx&m zE=sNL|1aWmsUwY+h8yr^utpbW3sBB`Ndyfc&sL&iLtlVys>;fd@G7mf0p(WKt5@fl z^0OZzKjn^@FyNdLhV7A_AVsKfMo6u`KLw^G*fKx$dPYKYhcgLdiMPa`7sM<&4iOgU z8D?6)Ndff*57vLj+=4sXpH>JWp$B?gjvEN<^8$$FBdZp?|IqzEwKJ;E$1-r(0swou zTb!(lw{Olagg>j}Aiik7vQ_&^O*?XJ|Hk5Cg`hS>HSbCU#1|yv(rP@4pO`Mq4k-(3Gg#<|`7BdUO4j2bfPRrhfu1MT2wg6m=O<{2 z|Bie$nH(;uiGra`bMGY$hW7ghkMTKen%+HUe*E5#%!FsStE($lxU> zg$EhV{-_dhXz1XeWOy2bf{sfAd2#77(jOn7P*nT-ck&-+0dq9s38}6&OgB3(4`;qL zgnWklIsuJ>925qF_4M|}SlUjM+oYtV)W*k^u_ILwT2Cw$s+$Tb`d_1z`#LRbU@0wa zC#%7HvW0SuT7)8)I;6Gdbx&X4?dqTdREMq}E^u zAN|(K?p7N@SVl&Hl9GroEqF!%b!EjqmzP2ws@15_&KJnb!{dC}hQzP-{)`MtWu+iD zR7+OU+OR*yyvqsWsn3h~+8ZCUGH$zkSI|(OBH7{~NzM3JYwo-5>5|1dnW+E4Jcp<( zWurb9H(r`0cf}z3r~Qscw|8_TLawq6Wf&y^&R?J86YRA&=fO<(q#W9IOKP8t&nkY_ z*>(#xLJvkzDKH$OGQ2&HiCF*(dXno8RK|&h3o-tQuf|0=w79#rkq2L8{H*u#e zm{A;?X*4z~0@G?e@Ms{(d?+iXsnJGx;iEKpMPfBb#Af-f)luFa%!LdDKNBDgmv5-GG^CLmeb0tzGL5r~gJ@Gr{6HnqfcnfU$vlja^ z23>!u~zw~&C8sagM*!ry90=L?fHnlh~kdZWmX~al+$PojVMqwNYsmiqK zEFux&iy7}~{5S>mtp!peuJr^Ca;J(o1j5yVPNTBSK_R51ULeL6iTBnY{kg-RbGew* z2WO~Zl#Vi6;|i}?Kz|KWFy$)xDxbNWk$mh3ewRjZGMq((0CBjWczC#vxNdYf-UpvU zkzQVrDvXu{A0OjxZA>vIk28SbOr=2iBG<=pvFvBbhnUbpOcI8SoOnKETBP7(%&ZS& z9U1yXXsk7C4~NhnUvGIc{fM?+{)I}jN;=YC&Tl;ps@@CbXM3OEL%BG&USvO2-#x4B z861q@s9gD)o|0=#Fln}!JKriMA+c<+r#>igu_7+E%iUJ~&EKTKo-qf!fD^Vi7}Fs> zXTPwsQxmWVL{LEv6Mc=WCFQM$aYO=DONP7eCFnDZ@29$|=Cw~}5wpyzXX)^S-&$7x zfSIlQ4n6SA3!2`?&&m{k+O&^S^>(peKXf@?ao^u*@9$_y&h>Kr{;sEL14eCpHAMK( zya1xYS@)hlA;Eb!VQ_Z(Ovj7h48AaJrQ<1F^@ZnO%y1b=iFQWo-Nz5&zJRMSQ3^MC z`0uFP^JPH-Ke1)%d0R!H1o12E9U5K6{8Gt={{bJm_RuzhQ1TW!HuFq5_1iC)sW*Bg zBtEBx5(_avlJQaJrtjbgvle_?i1|0mfwBBJ^XFC!XG`u(W4OULa5ihpL@+j7*)+@a zT%x(%o~zb9vvY?P=ThgiY)|YgSV6@dJ>6 zy7l$e(B!<~LLJ#m9E!WD@o?ABsC{xiBGJ~@^6qPu3M!c)JygshfTqcQC~^i=t>mKw$0@RbsCu43+)BiC3UQ8nv?vB2>EtxEZ30g5oMNN)4q{9_N_gQql-w zkvBv)NW7sU?zb2UAw{8p8$ya-7hbzWO zo@3-8)=9wVBYJq`F>7@`>t9jcG9_iuiA zg>FFZD`DX87%B~Iuu(XOX_xs*_!DQK*s{%dvb08ULxvb-%_Y67II3r=L2VE>iP>D! zH-n}ev&N;(TJENe2&wcI1A|<%NS();#e01LVad7Q=8Om~UI^rT;kxKlfB=EyGy~*C zeGF#&3d+dgfegDx5@3);e0~Q@j^XPT{ukV`XU;Hli}~^$NpHPEd}tAKj|%rUwTX0D z0uqPIt*0MUpxA(5%24xEYqdM(`b3O9!bdu#yMdUjFV^r865@~9>pn%D&qy-2KM}eISLJ<9r^=G0hkw3jfzwm!PqD-#4sF zs>(-~V=pTx05)5YgY;Nze0*?mPKc1i*T}?q`FMDIT3ReCftGI^d6@?ojS+O<3Q7S? zJlQl-Ag6Dnw_FuL_f~jgb#;}MY2H@71ZkV>`@wduf=*u2i;tfciZt_W=3D&qTLtgK zP>YMIrQ^C&AScH|gAR8G1`o$08&}>j{3PpGAT(1t)S^^LQAy|pr&2fmSVg#4a#j|t zN!Vs>sj=~vwv2#`g*RABBOZxRX~99Fo|xFYHQ%fdVnYCNq;jZ^6~&9;uy^E zTxP$U`40PTYHlt^NAxX=8KO`3Uev4Mk%)S_zdEN!Dxse#&r2W64J?$4kAmD~i?GFkVs@|=Y#lJ! zn@wK$2%-n(Le`2)v)bI|Q#G()a)6Iku5&frcNos=`+2~Bjr?DHcq;~&Paua-?a$r$ zH?e=3;%^2@u0%y~RB}3Hq5O$0e=yUZ3;fN~h_3_)aYrH=rT#*?3yJezK9BNI1N003 zXA}O}gnu{zab*5E4*#!g0@=qZXXL1jMvL*r&tewHdC@typ39&U>y$g!JB*Kq_{wUC zpvUo90lRh`XLqUd>Jqy;$+xZ;qLr1VzN_fQzOh97HggX~3e229DV)S?BTktwOuw~n zItgGb{VJ9^u3g>V_bxOMc4j_29VY_~Pw;~s?0p_CabPbcRMB)qi#RAumjP>tA>Xp` zThHjIKu#y8$oWLiNDn;8qV+Ui)gvnkQq3_I9WUy{xHLy65$qGgo#7wDZRE#5%z14- zHrDmiB)l~@H+Sn^;U{N~;PH4+zU|6hZ2h^KtF)kClT-akFA`F49vlk|v9W@PEA08_5VMh5 zeU-{H(6z6z+{Rl}qKPwNHL>Wjh`1Ldoku)eV5yWSV1)xc9f?3y$=DLUO2Y}sB!SWnqHUz17+2hE23K<8nmlqpJH%d2^)2g6Nrw-e&+c%-M_S^n^5jHy zN*5ZeH-+oSS5Q@9TieeTP^jDwwsCbrN#u^K&IgnH58M=31U96zqVEtwoi#-cJM7v5 z9&W;p#wz#hu)kIy4BB?isW~D=*RFQrO)gukDk7XKN!ZTP4RElF0UQ z=NSSKY`DmWo(Datv_oE-gygoH<@;f?z|Gjxe!2q&J8RuMWi}rvE@|B%25#f2`;tW& zwyTCP-%KMZ7KNziUGvke)Ggqg_>`&Cl7qlh9-iOUl9&7Wi(%u@geV00LzQ@c_KXry z1;ga5sZN)E*z{ z{eTH|7Cv0F>-N{F*|Oc~__QZhHbS~@kyI<7>DuOt965UE47tIIFqZ1o2tS+;O9}>z zcb{f>9rtBR`oMH0*2oA%9Cm{|B9{}C%o>n!M4q^RE~s$1A1&%6H}=hE^X$)F0IPb{ z#{BLJ#?J7XlG#EA;;SYD0U}Pum!T|psuf2yTOre4Tok54nrx31`$u?QEk4?DBk1pi z!M2X#SQ{7U^BICYRaly*s{3vak$c;LYNct;L??1`@`ehmrF6R8Cchez1FF0z zJsu1RO4wte5Go7#rDh9Wo&kS;sc@d#NZZ=qG$ivU>Dr`Uan>C@ z7kr`ui}}v0!cK0Fle1W9WSbKxGdYg+lh82Yz;(ku8*=iDF2-?iN+w9!?@Z*Sqg<|} z^T4a5n`{eCsA|+;Rq;2=t-MjOxn8ByVmpO1I9Jr1Amo%WRbA7L0w_bU5c5rxlqu}d~p>zrDKJGU360{oQf}{AY zc-cfz(4W4k&lPs-hfB{1vHB7wWwA#)$-UZ4Y=O)6!%M^9%5!d*cF()9L3d8h9M>`V zrTP9x9S92?iii89RBdnaHV+bMTO&s>#j@u;o?x2oMFaYzvnO7uHZ6&()XrPId-rK- zbB-bm3{n^*eVi!jxnc1KP5s^JocA_lsfxR6Q$O1j6ff#gupkuvR-^im+TqL z6m?OIj?sM3I{o$_)R_YB#@V%ItPMNj4pzzQ63a)O(s}9ajIyQ})8*$=TxQ~M--FK! zvo_ci$6Tbc2A#*NjrwiUZAs@7lniD1jL@Q+*4@OOOieS9pVP{w)RZWMmlc!BZ;{?O z+Dmk$gnPd$?t&nNh-}N)gr}3|RS8bK3-k-89JSsZ!fAw49oUe$A0LE?Vi9v_KegTK zb*>@st{js?TlTz-AzsZO;5aW9HZO_th1}{BCeGeidvqc?u#l{IpgyQ9P z{Z`bFyJz-e1^3)tS6uJP=D%AhlrT#mh}jmU>1nZ>gT9QVKN+6iGV` z#Ws8Ft`9D^HaHcv%YbkKO$gFrZL6I~mp*$t!DXa~8(SxNI@sePWAC2B=dilv1!e1K zDPPe!Qcx(aeiobUO%kD0&RgCvUzTXwEpusuP%4H(QC*pq=BI94#7OIn(FcNBi)32iC5*kv@uz^RJu4L_(f!wYn|%(FT*red99NSn7BlwD z5)!m#BZjBeYA}*m6=q_LywoVNSg&a%U0r$8{2Hg}Y>!-r9p<{%mu;1nIugmUGUajH z=w@~&Enuz$vc^nHK0rndO2nr)>FGQjbQr3gZzop`lyeRe}ApgRl8P1vjV<<`FF!IN!I_~P)C zl!O8&TlX2;@abEl05zrlixHA-5o!;?Kx{H&VX<*cY0*{**dr_Qe6@FvH{8tnl{jPt zWg`)h4U8Y`lk1}%1wzm}EY@(XGT(<-Oa9}cc6R0GwQhazILv!mx4O3zw%o!#i(sbO zH_M?9<#JA@eA>0qHdCubjG}?WOF2U1#%(I^Hy)V9Kq{dEr0`#~(3)snRp0c7jO*)_QE(_jDPVPBY$y$$*&@D!#^4b|)*z7OBm8h#& z+I3>!bcjD&8nY1{D|t$J9FofH;6N#lW!Hswg5vZOa*Ymk{#RC8~f-iy4aqe#AxoXE5S*bz~?^mRs9@nKe+Xe37f&R0*M+8Ku3 z)`Q9dcJ<|@=3^^EBIY^_=1v}Bg*~3{k~V7YP15U5>Ki3~W#q>ix(k8n*u8#7f7~u9 zuo*+adMc@Lk*N7WX-FapSt!-ZO=-cKhpBlU0swaOMx=<(uBCV{F+2ah`*TH>h*i%o zMa@qtzg>iOpeI=2^J*4si=)Qa9KHstxJaQLlM!0dUST@B^r z)xnFP6F=-3{yR#nAd&Di86qYob}(->K`gReuzKXa#FJPy?{>|%*kT;kU=Whj)hf*I zexYw3HJ>(bq@xk8TCa5cj8uSv0>}E?kYyrGWAuEkeUu6{@>}a5|5qq&aK& zWdvNXPhpG0@>nYY_dG!pZn>j@iPb!}pFl2+YD{PyFnW7_WPMTHqZj78Mz} zQmVUzeCA_)oNF3Zc~Vip-Fy5!b^5bsofR#xO~j>>HQ~-(t*J`hFzFW%UW_Jq44PEf z?MHi@rL1`$0p)|7+R9=FUS`;oD1b4KimvSlL9(}4@C{Po{Ov2%*!u`fM`_0$R z<{KdVoj@tUw4Q+3svqxV!uUM{?phO-x2peY_sKLjrx()ptTJtLUBPMkEy?V)hQr-H zr8!<_^MTF`$|^^HZ^?We8Q|XHV$xL^4&RjFtT=e7N4KfossLqxi?%y-FQ~S8D7Dl4 zWV+?%X4FhsHHO*5F8|J!3Z(VjY$c;rYF&!n)KeIQ#Bd)peU{;=1%G+8(Ft z`Bq+V<44q0vtuG&oD^`n$K)jiAp%84e!%HY_YJ~f8M0!VbejX#h!1eN?Q5EslAOr8 zK6_`1pT#Cr3<+2?8?@wfFxz)Eru(oMX)Av7ErJ=)Pe7mCox;qr&SiD)(w6Tgogpzb zNyKH8`A%g~hKsE{5}b5M^kv8f=9=>~__)gGt7U7!6RGaIKgY?Hbax{^HRp6EImy{n zay`yPFgMnF&x^zebc~Q+!$OdflJe&F!gT4rBvzF=V!c*qdY@4Cl(aaVFdi8>`TaGD zimk~ftFaP5@<Yh$kgg$oMR{IVV(VZNCM{*Jyna;vMgJut zS|zTQT#8&3isy}avzFrMtTJeJn}?%%Il6xj0gB8h6T!m%i27s3UOSc72bK{QdhUn(NQJ6^{Vs11~QJX(q`?nKI^vN_X?d;#sp%CRrS)nP^$KM|FuK=lS@l; zYb6RWRS}ES=L_02LayV99+PyH`%%l0k&$zA9>*F1{Wdq;*s#BCy3F#w`8>Jp%X9M_ z$yZRM-z}rTQSUV+>tUXgz#sV^gh8mgg+zYG5V`7GBCOeV66dOeP-0e+GSTyk7(MjO z^&f

    |Tk|@>E-~TT_)#-JPloRkli?=rohWA0U#zWk;b6*Bq4r)PDM*9)gfNMCt{F zvItOy`E&)VJNycc@6gaj*Lhyex3Bg-qf@BXA{*>;+=zo17zt@KUgyU-?!2~i27HPA z!+hpzb%&$lqLdu9TPLBT96T?Bs6|BrZ&tKq#n>d=7j*71z6Y)+Fq7BS`G&d$L`d2t zGmwJOcANiX-xLjl=`U9wG;A~zvtM(XaFu>VXTK*0$iwe-pV7Jy?&{k;O8i2DJ9V8X z@>o*07-fhlm0=x(A|_AYWaoZ&+7oC!H6DE8umOL$OYECBb?i(3VRaU_yJq@0)4vfl zfY?{w^Tf}3ssgJo87>Bg_Kl*q5@+P3dFd|h~w9TP6 zidfj%wXY7MPs>FE#;Lvw$bq5=0_yI8B3`}Ni^ra3)Q_6>zfLHl61BO}0sPeXF#ENm zURc6xP?<7gf*Nd%i>s8_{7_fpO_{rz6Rs*PbI$FDi-?Cav9P~^;KhOc9Xs%LkXPIM zG>Hom`m$5c1fsRAa;NhJCN)9)AgzB@qfv<|PuOj<@vyw?RX1X?0Nj>Y7bB!DKsN0N|7vuw5vCLazDMkA1D*Qj}}MW(92d=7fOhFz)6 zZa6#4P&6?Kw?d@`513dOzUF|$Z`9A2wO-QDZg z#AfEOi2XSyJdC#ER)?%Gz2G?0IOj`s=OMG{DpTlJuU?&H)j4rj1rfnH!O_<#6k!f+ zx2U?0Z&8{Oj6^u(G%c^nC+5k~2bGAr>}BDeTf*hy$~;NV`n?VnSHl)U?*46W=j>GcM5OfzN2vFHn&Rg%czRb%rZKj;@ z5x%UP6Pi8W;jA4sQc&rLTb+ORk@gtir!GOED5kDWxg<@4V`H)6M$>1@agEg#cCOAP zjC7`_3k8|?DIPLsv?e+;#@89&i@s~3XH&!P?oP|gn0g&MoQz!VlN}odV;K6tM~K^T zR{hu3BzHdZk#dHP7-4RPgkHsf#VgoJP`f6Yq;RWQ{#-tL)o*tgh8ks=5Pid0E80l;*C#rcti}TIRMxrQo3{a`}C^*Suxxxa}l$hYm0K5O& z)#cIT9&bD0i`{^`u6FYCH)M2>6Hl*(q zFU(;z%szXa5mhB*3ApB@xg&+oI|ci&KEsSx2v z*Slc5KtG6V$jG3A4z-9~dut>0q}Qg&J*|S<0AYclQv=5`$(KBF=OLYCopLu+wCf8{ zplwe4enrYZB%Umb1_q(WqrvS<&P7CBPMVX>o*g{cd3gXZ^8M3wSj3!|(h6B2^xOA^kz>)FQ zo}nXgRv(xid=8~(%dUO-~P~Rw`^KUoO9jMe6{BA zblcja@L zB(k$n>S=1whWjy_7T+VGt`2QtoNpv6;N~x02Ryz^f*+JC4mAw;iE_^{+{-Q84hFF} zGp9evZxnzQ?U5p7iWf%CF@5QevwH=o=CTF13YGx(JK zSwuUBwAWDx7Q%z)NM`KNezRJT8h(BI0^9_AfDGg+G;b=2w_o_uERK!cLh_oU)DVv7QZt?} zn$&0%z%i<$sZZJ>DY0@HeqGQT%~f@kEJsA?j~3vcoXEoV`W!TFzt=t??RPV)lIUU0 zA*RM8egcDMbLGLi*EUUHRW-xAG4+oy%|{DM%Bv1{yKc(CdEW^;0t^3Uj3^0||H*`W z(veH0DfYtAMI^_ZTwv!vi2~;XYE_Sf5rjO7iUBqde0747uiw4Pu`TMmB~dZ z!L|EH>DRvdZfY1uoBa*B)MJ^DYnyeRn6y$L?DPW`QLrbC1)OIB!lmc?MuJ zQuueBSJYhqXjK?9wuYnY?9)cUlW$@$eme?7U0Xr5E;q(jDLGcP(jOZ1LUfW8)}JAN znjN3^%0IZ@25?Vg8S-&sOO8O1hS!9^KFt*rujp+hvnfPfQRrj6|(DtvHSvwIKGZK zr`%NSAPo4*XU$1obTWxNx7atpHJ5ma=dw;$q(^tEM7m6)5x`AoFCoA18$BMouzn9C zox*A0sowth^);0MJI~8zyqaLR`jCF}f*8zqr~&SpH_r`K$9;7V*|&IZWg{8durwvA zuMY!}_P$X!IOX?kaA`<`;BJHSrUju0wTOE;7ir-%dl%0=T|yK-i$ktSvp}7Ar>^+R zc5j6pafCLm>=oWOfdX8xKU3nywUswe{xT-0W{u(+#?c1-2rHUxl)N`o+-msZ|$_RF(t2{x@s%N zue|;A^mBKbtWjGtu}^_Md4FeZ(MO@=>>ihy2o;h~+2LgfV&&GyOwS8Yt5~kL3tUEW zxsCt=9*lXw0mRRbCk}`&tA@{E-Se1-jvktCw=Pgm;s7KvOk=96si;UITI=={8XVkp z`79rPe#2Bl!--C7!lKlv`JO=|3(7|+8uF16{iQ78nqzP&r$Jgs^c*wi6?L5JDrSjX zwfwOv8jQx-<@fVmi&>q9ZKTFXr}C&|m`z?{Wo5N$m?_nfy=L#d{q|>Xi)VvXW^OsL!b#T%jeEb{ooLLB z6mdW34c4<+f??`));=AI#5iO1q~*q|Qe{5rv@)^WfRA*#cP=nvVXxoqg@fpYVK{5w z#)S&{A^Z6Cp$CD-LO6E`^(0>2N+>hFSH239zgj+dvF-dcB}5me459=^P+Q*`9|C!t zu+sDZ#dbta&T(TZO#8ekm;=>OX`eR38pTvjw;bU;@H%~9=DLiON{{PZ)uqMHIq7C( zWCXI)rBnhoA{(D)krAX1d=DxG?#)mO4dH=&#Re%Ij5+*NjH*tAomBS2TRs3Bj%GWw zo6k^e&rHC6db8qv49Z7MU8(0yx$8oX*Hp(2Xwu%2k*#2bnJ5eoWKE`et&i2*X`+d# zeWG$(urq_5IxPUCuo~0C8rNVYwTCF;_jPyM^lxvl3f(vGIu53r=ifLcC^Y9n8h47v z{1|l47daj4dXkE!;Zj%P(t4I~t}h?)Fe`7t=MsgRksp(YkyY=QuZi%1Mrzj?nJY@XFL6i$hX%Dlv%-*=H2-Q3siUUJ6>UroHJu@oHPK3Fhu;fDQSb*}vS*9QE1 z)WRpuBjqM$mA2KbhhK?BsaICeEOS{eVzkJ^VZ1-KoEIz5Ag&8ofr?)(BwMYz8B*z* z>_Gw?k^8kD^Fo9x?4jHjhQZ!glot{Ti{woFXwjQdaBH5Eji4MqHGF0ZiXgN2W8Im^ z_m4X?wH<3Lj)C+g$!hw;OzsAZ9LYgh(&< zINQFOoLW7U(WOT%yc00`ZVm0L_q-MuNS{2^ifD>Hg_7-YC8ovEWuZy6}VR<3aK(p%+W_`SRLcP=;>r=QA zJ!CR0ONgMjH(uGb3+jB3gT6EGf8BjL&nqKiGmsmdtLf-+)c@y3*B~sgJvzu4Nc+-5 z*K{wlvkME6T8reS#fGaO9Op7fCz@k3akP%>e&@Zx4ZAw`l=3NHOuy;(eNB?A^!?^` zP04Bi9Xeqov1hJ}mlkuNp#5Nkw3rqwXr)041nBUfe)Jz!Kz=+eJ*Y0ks=DN3cc>Le zf%lA!3FE6!kd--Ax*ifWtFfv$h~jc|eUZsgpPbaCGSaGXS9|_i5-F9*dP_mWW{H5& z{)*6bqdo4~^35x)@9$x-Y4!+rIBU9``UN3-E;U+E6Aj0RT3G@fX-duo*?js0CuV_+ zI~E%o4wbD^ZT<(z2JKfsuB%%XCTfRE=es}U*B!xR%};s=ix{4k;?~yDA(ldQZ1z>r z;oX|v9f3Ug{wSrOfR;_8GHnN9%F}vbVGVr0mp1Z;n+rCWMU@8Turn`_gq@em*ZLghq8jYysfd5bNZqEYAm^WJvOHK>+HR-SZmXU5U*Err}m(q5JiMCHzNcjusg7J7wxnH!gN%I9=WrS4{M;otxd-vr1WWKwe3oaAOlIl);7B^Hk| z{i!6y>0eEYMEr?-w-GPx&EfAyfszrFTQ=`-S!yb(1GLE8@H6<7M3MMa!9_EmfR-w-&j7}cR$)fQD6)Xwv6c^w^Z5=wNOG_~fk zQaRPsbGkc3hd`mCMhE-M%f=VG22|OBxvtcjxL0CLsY@ejTdOXGq#- zSNS(J(c|%7h1~bt8Bs(aEhLRvqTq-goPNM7OHwy^t9X&Iw0S$DOm~s=RAS)@wGdR% zE-3!syv#tHmVkQQC>rDuAgGuZAewR{oJ(qq5E^16L?9^f(bUb;?&%}XZI6s6qyd{Y zt?;J#dgNQOu><%(`Z|t%>}t(?7Y#r-K*i7~oacq{U>;KrYPIyTd_0V|A&uG?H{_;usA zmiqj%(8Ge+>Y^gnLrR=kr_D(pJ30!F<^qy6zIrbz&pB@^b z{NQkGWO*%0vzT8i@szU`cdr*l*~Z&5WewucBl#xM#+He1QMP4d#Fxh+(Ux;%yWqDv zg^e5!)-1}>ypa1y&$3maBr>OKoZ9HwVeQb!1#Ypzz^En&ua?{a|0_HAqPF7asN^9+jzt)o_>`YeLvbBR1R zH@Y)gRh#FO9<>0h8^gU_&P-vg=BZrF^C7D-P3(~Xhf3kB-G;&g?qE!k%}<3{nsAxm zh@6ime@4)NK-l5fGTzzw(w^rKVnDe|=3O*bGvsDA?9aRT32U>qjw2yBspx!;J96#S z>UI@C0DT-Z3J`MI2n>G)TN_=LKmUQrD7txWN8iHxCzCGzPz~uGlKRDSLg|#aesAd-Ee%tZn zwt)JYrt5lTP}>Y8dH-t9!Nz@edD_~{LeKecTaDDBk=5RXH`(1ZVF=Kakm6?0towvU zlM!NVBsZbvam~^6Hx3udG3ZK;-Pf4Jbz4Qp$isR)(4_YF9cCj#rg1GR4)srS$&g#l zVAQq5utaxbCDTQ-(04T=Nhq)O-J1&rV!z>ipoD_UI=Sz{nz$tuhO+t}YfVN|B`tL( z5Fpa${;wtL4Rmgz;kkpbofHzUkaRBoH2c~89sGkqmBbaIoP{^f@k#DM&*!I4WsNP? zsY*ZxKHvq!3l#YkJGGv^Ezi-sKLZqlT=(Z&2PJRSL&Zx;_=}vT&*&jA=Vy-PECOC< zQ=6Wp-5~u&Uk*Dy-?Oeu^=h6;6g?WNdFGCvPp4Fih!@Zg4}|zP)`%OgY89$(BA_HZ z&Q)Cx zUIp<87MZI2QPg#z+iBk3fjW@$`JeBE;9#?A1YGaJ@!APGmhvpqgo~9Nm~SUIRZ7TJ zo$RzrkHvzLF9cJuC~O#&ESmNo4mo*|Ov!6!1TA}=QlGo3CL9Ux^cyt*tiwA?I+1f7 z_o<5I;&Y&8z)Om3LYCa)%hV&OfyR4Cgvf;}M~n$bV43B;sbKYBNaWalQMn zT%PR0C7ofzG#F^xW~2&R+w(6TZ5q_&KqRV z9;?7a9&qovoVboacreA&gRZ96owFbh-Kga>HMZ2=8XGfB< zbFS*S+N0_*&nt=^KOA#=)X~O%XPRVgIy8m8^;DraM?*?i8{eA1P1M}XOu9SCCPBaL zQN}HWOx$x(uc`R0^yzcW^6g5qqnh9>=^MuLyowsO9MfLOyp(t)`Q6Uu;8wfC*;-By zZ3UIf>6o5!^5PxUw@F_+Zcf$EqOBPqU)JA7#BxN=3RXj3)C-CAk8He~L1Tecv;~+E zFb2xI9!x|9`)R0(!V-&h&je=M;S=m1vR~Xr64mF-2XUsAw{ZJ*CQJTLMnmQAY^S&d zl>Ic>YyW`YAZ&7w*J*lOHphIuQ4+w_qy%rWQo29QJx)<~FO6P5wdLQRyD08}fE2^m z)0|RG3qj)QxZGr&td2hyi0=Sn;DbE)Hj!TL#IwBC=t#hzF6pkN?SE;Vi z*qgeZNkn%I#d?t_%1`uQXOkh}3fo-^7e~Kc8|B7}A~{ecTH5*kxy0H^5Ty=e3Pc73 zO{FA^Bwl&#fiB0SE!TEX2d2gYddI2V4;~9HoHU@?H3lxEtGouGThuw^iI!Ak^Q+_9Yu8^?~lOe zK9+HR)|sI4eQ-X+>JQ6ruSfzQ<(H)TU<0lfr^gy$a<6eRkkOmq33^6NL5=R+O^5tm z4r~$#cSjsZMBn;c)0}69zpX<__J@?dD{e)izHp@e?F{@71s5zp$Nbi!h~$WipDy}w z)o%y;5&8SciTjC#AiSW~@?(p-65~(W=+`g($$a{AJX>~}eAeMXI=jF5?!Ox7X)17{ zO>>lfzazg{1PuUWRM~0?-v)uNpEl^fpU*#=fG}kL;e>xS;o`vl!wLUi9S2AS%E}5) zNMw7T#22u663zB^tx z*uPQ)_1`flgk6ivch~G~GBAEQfqHaw!m*nzLF7>zvbw|Q$>XTMRWcxk^}Q#>eyQL} zN5kIt#g@dMV%2|{J5-(M)jDB8&HIW90S3?}DO=*D?L`xZUkv}@G6>}dQ+?Jf_{`H4 z_G>cG-2{e#vAH>RmE$@lxc;*cZSAKt;eQSe#UtXko`HZiDA)u{ax&-5%lz`cwkHD9 zASs>iL8z>k$u%{?YokR>@%%P7WRgULPl+?UdKt@LP zruj|3jkf0X;a+X^v|GR2j&QZ)p650v2irXZ1L2PMqb47UQG8t;DbRlg##zQ3_5A_m z{KXLhfySFi>+9>ZrCu$)<$5zi6BTx1($dmjVD6*{ikXj>T0DbqOpZyjroX&^5&qrb zdin~o$rrFa-ky$l?@rKIXZP`4=4&)J#WF}h{-wi#3a@}=VoWK0-S{(|3uV-e^20Vwg`F` zy{SG2fas* z#JV5rGFx0PXsjg38x-Pgdhw}zrXGrH!lhzf4(_3oJa5*`I}qS|`RH#}@)HdK@`WOS z70rQC*p-Ju(4P5Jg~Ww5c*=p1&e4RkQ|xN{Mds{p=I{X<`ngn%hnF`3dyTBfcJXT( zK^1|?rQyBh{>Fv2u>R$*%(a@oR40f@DXb#h`oijHP+`=a$TPzlT+3_J6<^HVKBDnF zi%_UO@ZPiWGAo>0By4E*CYiCY# zFZgYTRpDdj4Kxo*kRF%FXk8mAm~YBaD{}4FyU2(AwBLWxDWE=vCEsqT%YHAHq`pM= z^Ghz9xh9?$tdUG;?v!_(xy&ak0b%p_rJ`VcQlCr?|7^Yo-Bo;C$^j>8P=!GsV`KJ5jb%gjgLxc8TETH8L3Fw#n<~Xq5Ch<3OmP7=EQ^frHw6K8LK=!|x4VuZI zUGMXbXsUbKbcuLEKLLh+S{j&|sX9%mS`+LIAkDx%T@_jCYnbwX{l#BcHu(q@SaT|D z29A{=3HHmSfksD|>x2>^RI+f!d^41u_1`mpF9EQ**f#D8X~moT`j=ctNCgQuf=32> zd*A(+=A?p-PSVC*zVPcA5}bEnzY-V-0AotR+`;zzO=x|tx-fn?q7T`x7y=T6K}aTP zmNW83`d1sZ@Q7wyn>L?Sf{5;q#z?;z?-ySU_<_*Q*axHZUy|-Gg!#{@&`2OcJ5Naq zK|ix!ujV&jK`hWgd@IORS?ZU2@tda+U)j=w_?AQj?LV9F&nEoC3IF%Up@2h$@XuZV ze`>>j zq+j|$URDYDuvipa$4r-^lv;OL(O!{CfGumlpR{VMHo0z-f>N? z%i8d=p;%BsL_h=;f>NdTVxtofgwO&???{bwu_3*L8cGo9B}j*aq9VQbPLLJ|B|?Bu z!}qYibGEqOz1CUhJ-_$A^Iv-IXP$dzu9>;+nSlRt?SCn{o2;_NkTX)}-~ZVVr@#F6 zuVDbQ4_|#pt*s_jD*xNTihuh;&)ftA{b2t2Jz%`Yzm?#>lj^^D?GE#CR=4OdTpRQ+ zrnCGF(aD~n3?HI>aPiH@zbS}cBKhr+?Di#=!dx#&iFMwp;;Fo`^}p&j|`_)Q!KuIxH?HYqmEf z;`xYbgRkD7zhSWS9X0Gc*HVu4cG%FwXmh*&Bzr==Z!8z0$|rmvMMXu4p7tYYlmXCC z3dSRzA_{se1I3H!42?K4$VmOxab}o2j*O4jy)}4FJ5y1yncu4C z3cHo!?s!bx9U68ussXkZFPUOoo1>QtZCMOAB6i!8V^QRbr~UwvWm^8vy?|d2^%nc{ zGUsf=Y7eD%XKDXM@<0a>jxT{wvyEmx`y(#?#MXm6er`eZ0oT3iQl%~oz_F7Gn?5sbmVs3k$u+O zlkU{-i*nTyv>F2@a%~L(>vYsW(#)1)4_+|YkCuh>6+k+*=&U8$5kk=!s)(kY3ycO2 zCAQOOpu4l~jRHgc3C|!3RcEWpZ?0Z@E&PGfy#LFfp>N#+p-sTQS?Uwv#pdu|fDONg zG(#os#kRxfez8gs_%-pXwnz0?cK9j)h-_tK`dO*sP=H( zUfE?fCC+j8)cP9Hd+bcU22%!@{P^)>Z7ry?HP* zk8V|#SF*Z?Yvz)XY6v~D=3~d3@A}6Kr9(HxR!=TwzFK*>rmmtAY`|xowm}nlLh|(+ zgpf^{^Tu2|LUA1Jn&d4Un)e4`Ib!4DT)!sOw5LeibPDg04R%!ffz2B#?59h4+5}oX z9Y3^c=~**&@rXa2)lD&WuW?^t`n4teKIu-XLzcu5sx8E7wl?W&OMzhot>n8Eu#-wE zuL$GRpVCeK?BD#HxIV)OhVfDssN0vYxtYc76i5E-42$}}DFCD`s)er}(bh|{EMj6} zUoPjgCdm5u)U?HW-UGG{5DjdeeNmgv8lL!#WPkUGGyR|cDeGsi`m>74%Ec=c#`S?c z#TH>sE~GJzqknyU`S=g{-3v~>M{LJQO%_g0WuILWG?#vj0dQuBCs(^5CJxQlgM#W? zjb?;gcUGwqbQs7}h&5Oz7{^D4PnIl?Wy6OkMdIV*`H7virW1Ro__2F{4ShIL?!e0= zQAU^Khq8&~>2ObG^d;|?TVXVTV-ws#v0ST?tl3uCYaJCN16Q=bRAf*(zytw%w`?x2 zZPOWv@1na=cKTvJwx)r%r^n0e$I(u6^beEV4tNZ?;2^bVHsu&4C#AqXl$J6=(2C0f z=^c|ehI`cgCCH+i3<4u=`5x?t*-jMITVv6#oC`ha#Rm8GBq`6GBigPM?@z1Aw7w5v zgl^Coz4UvUtKF`ZDM1MK#O@KKkvrcwMoO%>y}_a%znGu9$`e^hoNJz{S2kYkifysBI5yLv7F+!#y)%L*`_XC@CMgvX05QuugIr!66ff5Csj5 z$i9$l&V2pSNc%ZmB&=ij-9#~|ctH9@bvc{V)#@^+6KRQ$wDFEhZ@mDZRoxdxVDmY& zaunWBIQ9S>i-5qul7tbT%mpZ8V_9J@SP3dcWr(n0=VTXpcGF*3Hiabc?Ha%^BjPYFPQ zQw1SJaR3{~v~|D^G@+w4Wjt*<`HJi`_xYprYGY#J^=+W#_slM{&pp0qRSmwpR9fDi zt;u>i=(BpMq;pJuApno08R;7@njCP`Mnz?MJo+1_uA@DZ34|tWS*z`>*fJ4k>(Q`JTSw4uxU-Pee#jZBMiS6xPxV|a?EG!lgE`1fsN^4hp)Ol-~SA*BV= zp;B_z{dv0Tbk~oJl}%|oL^tNlOecQ%={t7EE*J#6FHsWxck@xoHs9JtTo-3Xf%EXx z^gIHT7a+k5y7SCgbO_S`09p(#S>ihTF5e@Oz(R9BKu?s0yzr z1Mk^cX-aOB2FaVi;V!40^)UriRcC0Yv|Ej0oD-W{{7J4(M=MOR3iWI3cbWki! z1z4Yhhvsctnm2rxB96bgfj#*7sV#0h)49ezQOGtA|fCPFhk*Wo)hnkd@^h&Ku$DDSPB8``p zQ@@tTfu7N2?Fj&WCCkL0iu&^9A4{iA9x@4aFO?2XTvDZ2+2L`g+aK>{R+ZiRDoH#|sD%h--&23W*$LA-5sVm0pC)4us_(!v zRa}$AJv$7D5{^HIUml>SgquwfZ{y7-`q_~laz0z#On>x0OCs+XAJ_HW-w2c3X?k;t z;#xJblnT`4so!H#c(X>L(IpAV3PgSzWtM7&HTXqr!A?$3asAQAFYA{=%QUE>zYNq&kfJ1yHCzx zEG3ejY0`Z1+1`M-FGt5F>NJT9+@Dy61(!`B+&uU9EVN&55I+U5N$t-}YZt+Bt34wF zqUjT14PUB)q)jB=4-cyJi0*WZo(paq!SNavJx%sraYu@NxS__kY9HcmW7ShpIOf%R z!N;WC!+@RaHSv8>5H-hyNs@4=iC1y15iT~igBUtio|hq9ZyXo@a&QD&y|{9wBC}EE zto;Z5BlwJE=(srJIFYx8b#VLzTumdERO(WSgth1BKn(YdIEPzzU*6wsZExX047uzk z!M0gl-I1At6$fAay(2BFMx!@J^(V)z;olX&ewdY#`>d1XL+F0zg}(_TwIv zdo8)%k?5+k2@O4huH&-Ojzn6;^hc?q z5afu2U&&gYwd9WhIl5m>nsog7?Qyb}PkE*qmN}vLBL1vusm)*+EWe^r3CH$IwCe3Q~ z9)3}~wBcR>nNTXN2u;S#KXM#I=nz-NF;cn8X1kmMH)rc_%WbRT*KT|%sNLtGGxCU! zF2Q-XIUAWMlp_M}RI`^8ZA}~;hAO%PeQV7@SJrvl-QBhKAk(iW6rmG#LN!F+)5e?c zN;jlqy^ko_11OMTRQG)leDR;4a`MBus@;iQe#a@(T3;2ai=^5!D{Hiz?RmB^+PJJ4 zZxbNcy$bbfD1Mry{?&P~G~aiUY&(KOp>J1z3J{GobgL{a(s^#uEmI*QKIceJ+=6Rs zs6UB(DR@%>&Kzngmfaa39gD4W&ZIQl>iCvPkk8fTPPa=K&=|IHMVYn6w%QcbF7!fD z&O*N~ck_h>pf2GGN^Axjj7o)CCsTbdA~r(9NTVz292@i#7TTOekeQ<0E>hX^HxxaU zn(_g`!6mu*YNb6ED8=vzww99q{?ffuCV~p|w6iO%(ngT*9^KnPdM==5W?%B-4AtFl zguD6e0-Tnm+3Dt`IjsddVbNo=<;sFw=ox!}At>wz{|TDjUnoJ-I< z7iHF8jXl#Iwi(FFUVNFgbY%mURx(jMRT0$yQ*dl}sDDJnzd?MzuB)Q1&wP=`z(~Hj z%#Pv*kYPQHeUe4=I!WnLZ$&5=k1Kfe=+Q}o%S+R04H7hq+Y+P~c){n~gDT&%(-8r> zmV4Q0Ew@2eGjm_krUxVMJ!wGD=Gxb6gx+GaH8VGF$vkGP`M7f8V@db5WYEMyX?0M; z#9n=>khx<)z(KxYb;Irzi-VfeH;O>JE*x!gN#domThpP@CSeQQ=+}uB-TTfRZUyHv z8-)ToBsTTam1w))J`8DYsIT{MGYNssd{EvG9qkWiI&yt2ajy=1-|02jzy5I~{jRQY z+-3kJM0p#z<>I@kYtiyq;Jre^Kz{zD+_xpMHDBJUc_s94k!x;z(h9T^NRoJU@a9cL z!PO#r!4g^?ZUhhzXUtt3QJy&kxBXH5I+Q}OgU=+}g~e+A<7q;yl>6=p#!m$FU^EL2 z04qNcs@icq41Z9Xch8*V?*Ei(+Lh9qCxf<$Y&*>UXlD4}(NP_Fn;RlnJy!oaJy#ain|_?OO?Ka*?6mhTYYao0Hr(sfl9^fm+JpN} z1Hk=H#dYo-pEF7PUR#w;?fnqNQUK^|qSr=qK@rMtB;B?*F$A;k>9GpM8LCS;uaKG{ zNlYa@WPtF>+PE1So711u@_vIc-DH=_r?X$H$fZnzn<02{=wmh3&0k}CJJx8sjr=m0 zzIXf+!R3@*%|7|G`%b(?(j)l`*L|$aPfOYi(W6?ee=5-UqcA_rb8GG!Bqn;Ui1lQf zcrLqL0SF4Kh%26{oufR6EaBPS&Y7|VhoaoqoAgVC zlBC)L=LtHdv>Jiw33p9|p#4(t_8ze#T0#!HtTfMPk4Nnj`oWAB(9v#p@qow277y)g zm@NsY>6N3++x3v9suv8CJy(W#0-MfYPuBNrQ8`|CeRNCxt}^`GT{Tb%&g2X}I*baE zPtzX#!^3kB=e*Mvj}lKx`CJgt>K;QADg*Y2+nO13C+4%FG*457)2Q0!K}UrL(tH>W zx*p{MJvB#Y%louSCl2OsR{|lA6y{t@8Sdm8AWV0jO1-}@mP;eikzH6p-rs3yZjhhW zBGOrotue6=m?4Gxzxyurq&u_ITUMc6@T@>H)bPQHm%V0b%fLyJTeo?uE1pur1snF- zhHy(i#E|0{*Jxl~rGdEH(zE)vSL6E|4c#|{bnT|DyLAsu(Wpt?cBo#>QR7E8QB&2Js5TvXQp^rZnk_32cuUp66Vj z6|srhke~ku`jH7u5O$hwFK_99ViHY&WML#AzY=X*CysMNSq7CwV{WagP6P3E2Kj^y5xq)6!&@FJ>Cjn2|*`>ATf1mX{P% zW%g|m&{;41kc&_XTn#_nXC=!ajt-UFYr2*R7SYbrSs0P&OM1;E$dZ1Hp)-~8`i^Z? z5Cy#E;0HToLkRTMEKPA}jSJxgBZdqHp?Eg(-Ocs-rtRQ+q@hn2^>&JY{i z$Cwy(?)Gj<_Oix?)JzK_7hjf>$_o^2!$RD@D;bxB!SF**tOv60s)czEpo*j+`3B1) z7dOH88dBe*M^Y^4U5C#=G-7Y`60*BtV>~$Mi-_KJ0IfxF^3Ij-Mdz4~KPtRU=q}Z6!cZwVz%@Rb0 zE}>E6!YwEG{A59ib-#vO@C$aLWzf&ra^`jw82jis>}CZeFMEC6TVDS_p89y_s9xw*^~;wohfGPMO_ zALuMG+wnMol3nPz;`FJnT;km+Ad2Diaz#1L=2XY#azo$#=sq(eHHb*AZm-DYl@-iN zlnvQA8s2AC9wr2@W$GP;X>>#e(+>}Jw6sL7(tt)V1DOqVatRf2;7J@*Z~~A`xd$Xc zt7T#{{-hSfg2;o8)K8h>%~l>Y$8BxGW9cf8Krw+?Lxs1iBI{RrLY5`!ki(Il`Bmqa zemzk7BB1c>;?f?Fv)FqwA)SG=NapKf-?rFHtQ@d@ki2}{PCjlX?sXu_D=pw$vRUiE z6kGAjWPR#UbdSaczJK3sd`RsN1;`eVwk*YNqFb~KY!esbK!)z1E0t6f&vEs!hO6<_ zVE6}@&HIER-~^v;<=V4*ieVS8bzyPz?#Y9x?#_e3uISeg50Cd2@K>Xm3A8A2H%sQ9 z6(ql?s4KeF=J}fA64DdH(qk70(Y+p<^G_z1UZeRTivugTDl52W%$KR)?E5E3o*!`@ zop^}4uXu8T1N!3^3$3p0Mnpfl%9n0=6MY=&GH&TPxT69FJ1(DgX4f%UO)h=XsFMj^ ze)TY9Q@l**WKB@ut_1bZIjZZTW2C&#WaV9nw9l(AuKzQd_?L=I%F>zn7J~^z0pN7t zy5|XjgPl>K?2lk|m=jXkXY^{i#zD>a=jRV0n)##X_UhyC+^;v6EEBbM`>C0r@_J>q zj%TW`>_^Rucc;SLcQ_H(`dLtZoi_<5MLH=$5(L-AO3P?yl|Ko4NlAsnHs>eWU3E<2&e3r@^)KKJCuXxV8(RRxP@Q{jY_M?tdq%R^T9ld@-lErl zkb`q9J5pp3)jRbqY=CjP=GQ*BBx>N-pF6na@??nXMMOk|_EbApm{z2VIl5jsW@qo_ zSSIn-Q_IN5`FcW+Evkxx$g4(>VX-JJ|0QXqMNJyR90aqkK;cgI}ccgVfmFFkPWAC^ccx4oEL+oqt894lqc%F{B6v z$G8*ebU5ep`HT7O1t|}YB$BN362GG~+wuZ0Fm6#`y~L%ypY?7?o^Eb#E;P?x7dba| z(=fFulh&21+ZDgHUI}dLS748}VnHkFmvi7wH?~UX81`~!Vbpiex#8N{ z?LQme^;GSKrFIf98*LRVgI*gPvN&L$Zd8jB=#+d&ktD8eY&_O%1U5NZxD^cqtb};r z4N&A`ZQ+-Y}`Ay865d8kzu{=F;fu5k9CqN{K@j#qC@R2 zG#_ly)JOWlL(B2>9w&j8Of|Z!j4!UJz0J0tOzVbB-)GVodMSFOPWF42T)pfhn5oOi z&~smvo#LY}DH$jviAbDh{B5yv(#Ns*DadTX{cqz|SzyPLZMHGoQT zU{79?9knE=cWGRrrU@7Qv~H?_d%xaN$=Wt~eqH>y?`Q8Ein_Ly>{cb(Nj z#g&%netz+Bg9yx&Y)$C}M-%;#ob3rhMCtM2+7-k>Q%}0GcVx4RQxn9{d)%X3O_VGu zMi_IRiM#ZY-0|gTrIV%UIPIG;=JA`qm>Mshp0VsQ+26T<^2`R(kkRXtIqSum*u3Bw zs9So52ZU*@rZzgqZDqu@Y2R_m$FC6dCFK}LC;wEBa++piXtEY+S;%G(z4_1%sJm)K z>23vzlZ9-QHft=%+p&qB6Kar(9=u|JIorI(uvITY)-WVi-fdumQD$dBj zl~Ke#GP6;7F=w%>rg}c`omuwVZ=P52*qdex?)=(+f(uh5&wg)cP>StHGANh!ZcrH{ zCsqMo6i(~^>*iOX173vpgF&2p&SMq`_gZG1yal7>qN#q+gbsG5>NJennh|)%fiNKDU2f ziXSyiS2QBE4X*$+9dtoOL`N^%(v?KSFG$cp9W5LOP=&Y`fO|Y zY^|~n#vc}sa#7c2bBGhGv8K53j61uo_`bVJ#m1$&HphkCQV|pnJWe~-qZov8=xn(B z7zVSKJ+TN|XHqpOsN=?hZffeSbf9@aPo>FetKN)KA&gP2-y5$H_vAXa)l>G>3v#-Z z#_3#!MSbGBC(JKOZhcmX=kKp5FMM5o5EtE7RCLjYS?9?&o)cB2Fvvs#sIj))Chw3H)_cjX(_#PHL#;22u^x7mJV7%VgsZ%jhaH(iFd5 zRF6EKuUGcs6t|PJ^J7Ru4m4NWFIIIuPOQUTC)3T&llWPN1|DhPJ+EY3YCYi6U#RKg z7qvWUwp83%3GSk&VD^}lzH6KDHHKTSqN6v-BH7~s8n%enG*W-8nCFsqU*aDm0em0j zgv3yqqe*Lnz&0L;lcS#8mLH1c>amChI6BK$_c1yvt4jDB>ya_9=?njW!|l1%i>7X` zZ>1EwNE7?u{A+8Rrodb@Fp}ZY=J~Uur%U!vkj8JTCI}YhJ|GK3O%$t>sVIRy{C^w) z3cq_fEqAeEiGo9BDsp3)kWwv)^5(2oTl#PueEu9SS*E_!x z#6XuHOpWdDB;VR8Hl9`5?#t`hST4r~fzEhbRlVXhoe!HCIWPnokB_W$)g(!7&esim zj|TGT(O#RwxKu}j&eUDf_(4k5>Ii3YzwMD~zxv!MuT0rH#PaVKjwHk{UH6*<(Kldh zhdy^>S}c6)Mv0o|>=>_J1x*mvYe9E&MM-&f z%kj6C&ylx?odf41H+^2dtcb5B=!)&;J!f?tzLWm5(sWn@2YUeVBtq7cqyWzn&%a8u z03`Xyh_|-TJJl8+_n4@QE{6LZvEsjU?eI~>Nx@Z64MYtxjy>2jSuVhSzF%fLj=$d9 zAK_BZ$f$6`_bUVb6}QT6G?z31r$F!*z*cr}TD$oOc7B zd%`ApudXoBtV=ge3`UQw)DWrSgh{N!dZoTl4xXR2B$kFp3oT)rwA0w$P&CL0UJVtTDU(a$J@(yw#1&%@i zK|vQrtc5jY66WywFod8TUNe*S^5x5@owhu#L1~G9xL5LCtCaF{9GJ=MI~yHcfILJw zC=@kqE#bZ6Vj=fv&`)Cw;HbgHoN;21;4s)}V&Vz^!XHWDYlJB_0I_&olM3Y0`P}7N zBOQc-qAT!2cXj;$z}r+!lDe})WmUO9pg&W57e3>+ONeGf)eh)2uOyWl#u94xl~yn~ zo`8UWyTiw#WeIE>$?m>5d{=StSU|-@IZX~wYM(Ye;skzO?%^Yrl2a5^*C!b!vA&&$ z%46gEI@VjArZneiwlM)MO~<9jy1TjT+z);Ebvu=2HQG%VZ5yAKX`vI_|2_G}I68Xs zQp@@V5JLPAAQ#k<%f4L8K&@}?UzFKNgpKhvZuQYV{TT56;utLKWx@H_xF}Zf0IMfK z;7jh!kLjA0%g3jjTO!$%z42K`6g-9N98l4YLEbR=6*As=;+HIiR%RofHtAfO4wDHW zcZ;cSPu7KbQ(d^>f>48|1dd609%qUi$G}>nvXML2Wh#M+h^yXZ{o3SQpT4P-7|@KCjz{7s9?Ot zpE!CD@e>bxB*fLT{k{RU3k;$!dm4-&N>AKW)KS%!YDbly}K7GX+3!qxQJ64~e zn&6GE;(V|w@!d0c>4>@bU8UZZeYwk?#ELj$5(~f`B_*Z6Ax_xv#zeEQTeT0$U?H^E z5(<y74`uwb(xPIc@Hi- zd7reijq3v>7hOEx=G&+SYv@20=>%X1UC6bf$K2 zVv^Xn4)9BZI`VJ@eIzCREJ`Fnm$NX1f{NHU z>(2ZxPS+*+-o8}VE)gPlc`?_aCK}+9y>WDtxbVoGy`e2Z#z;eR{1~g<0Wcr33aF2s z*ctI&)AjO;7dUVM_na(<&eay%m9LWnZZvu)2ITJ%Q*Yj&pjyI}^>zxTaDRIC+CL@p zFR(>+p7PP-$1SUk%s7T20?1wrHsO^C9=*OO8+ZS$>n z^7No07ol&}N~>%=}%BXL?pKrc#`=gS1sG?EGRu`1NPB0U8TMphGf*+TG1 zfc|OPj~7bitX&|gC@IBw5^gLf(&dG`eS0#K)B9qe9=<;p*Sp@^nV|?UUlaLeT#&l- zLM$%LO-~{d5*8ClrHW%ahorjsuA4on?gHJ?<+j7O76%GqjePc10Y)6C9U&Z{CcX2; z!mQ*fH=@KA_ z`Rmov2YWwyvYtOAQpv~dyb-Xr{zC_%#c7iZi=Oek3eXjSN>J~{CHsm6ZF--H%AUR| zlDI?;g-!u_$|+Ozj*q|S80Q604J*|cZR#S@!PwD7D|v6*R51UTe1Lz%)i>0gmK$0;m%6+slJA+JFE?#XK+EB zmmysrEX>Q}(yf>`%0$5^8_474f*MGN^laTSuS(}o`}r1q#0&FSEAw~vLYS=G8&y7R=ZoPgro*DC zGLXXg`tg+F#cW?$x2*DdiY*OtfDul~TK|Ii0NN)pES&L|!0sp>OA1(;zz`x6>?_-q z-WC<43q-WXiS_OjE>_8OO5to5`iuQxt`VE%Y@n7hW}0)hp2?I>z_?@3vxMS~>mWky z!N>VLcen0GAqDMAmOCbNs)jb^)L|{k-EVV5CcIh+%hpp`&mUXBi9B!AB(yRTbT2MX zzVY_z&*@K<@eAyD3R`_gce39Is|san;;E-}E4Ay@16=0jDQ#kgx+`n(og&x#jCcB4R)z8J~OVSueZO zg;*?frf8m40+Pu|b;PS&;Z;?VHZuJD8se@?lh-+IKUEZ>zU`bVofhZIBF#ht+%$On zb;o=)}*Ov)f)&i>1EzXmkL(uNfoHy(4eH-Zt;Y% zmG$*hVO|Xkz=sW@FfFkRPxHpOJr?h@muX_n-sWVuvheS9z1nqze zs6m~FZ95r2M}{5ZHadRgwaz&3aD^jWEkS*oYL)&P?U9Q=;!1#9Z}vczodD48rF|Op zfa1zsTmCNqJr=;u!90T6*>qj~UqN1L5rQ9rh@Mh}788#~Wq3nKb#A^l;aVTSc}sEw zO3UzjGJI^AeuM8D8Uh&?*YLN}V4Y&ST!T4~&1ZuN0Xs`uDbP$Le)Y#L;gEIVp?!L6 zfVTAT_=o#q``|umB_r3q8i@7GwenLNnLnl8nvL+k5nfTDgG1|)zE+f*0W?xD^f{w= zS#AOP)}(~t(DwCT_!cOM0KD+Cvzc{6zjC_|Y)v3 z@ZIxElzthG6@I6cZ2p?XgA7ndIT<8Zxpo@yZg!)3ih_0R*3idy#i^KJDSV4UWp zHf#uRZ`)gSq&<+IRa>?lz*U&>43pkV0q*ptCYl2)m9^nQHAR4`;l%>UH(bt9j`O}Z zK+BBnb zxNj`i<)r8NW<%i_B4<-oNaiGz06Azy#Sz9CW3`&)>3k*#rhOG~2>;PPfR}0ThsR2x2p7cN- zl#^kro8?I_sOD?U>zzrR)l&@sA@UFa%uA#k8##=3cP11S6(azE5O96#+yi@i)P_A5 zKc}nihGFc^PLLH)$ZE_LUUySolj>!`qc;}Yec1;ToScgKvNh$WRKhiCTU`hGQloCK zGHFCPc2gXYXA#F|%kEn?^GOnP4(U@x9<{-YN8YWw3}j@hEfdXi<5>Y-Rg6DvZb@SV zkjn*F+VFl5MS8O#kOGe$0Esmv-{Ujeb)y$wvc*)YY;8TXG9S58e%(GVw_=cUd6OaG zFm#?5$gu*+^@$pA5`$fwD!(bYxs?Tc&5xPo)YjpvC`pl!cU!}Qv}jgY0DGb8;gU*+U_>CAdeF4#aeca zBPZHr>1lrk>jLq#9P_QZU+3TM>>ejmV{R}KKb)T*EMLN2hrM`ok{TX4(^lN+jFNKy z#{M|@Zns!Kh!f|}x-gmMLt5q7L5PR}Z>)ifUopYqHjv0%@tnN5y-boZ@JBN4mL$H| za|G(8zswi#$nQ4yBt5HCkI9dnqsDuZ^XK?QnU=b+;6kc-)@UPe0)F%Z!#{JZT%Frd8Z=}109{RM(n4ra!*bQ7Pj)B^#;3o zPmf}b$<%%OHq+;K!iySXY>aOK81SacBd*6fxmWzCmPZxybXNm^)s>FmhyO4v@|j=q zbSXPR09MC>ZL!}!&CTH_ueTqPf!ft#J&zFa|L}8S3lG0HT8TB3qI46xW_yJrJT{iS zSaJgSK<%gHXP058FsX= zJU`Z{cz$}?^W9GcdWPSk#!HfZfnOH`lYbLFaPTSWOO5v~>+{X8-%%6Q(_ww;754O3 z&Ro10>%qs@(L_S@pD^+ys&;q2Zt1Mu;{c)`7oi2`DN2ox4^BatB#z#)e-)7I{o{xH zmzf#UOhtkr>*4htxlOOWJmCgP4e}e)vvWt+lu~sD@VDBB+?8<@sDB+l{Hbb8oJ7xh zOqF76eb_v++^V-GJY6}qdEpI(Vq@rXesl%LJ4obAl<6VN^xEhj(Ezc+k)*P8JqE8@0kM`C4>PaOr0 z;*e&xqZ!o$Q0um_#hnVk!PWe^#<)j(`+wd2;2q@*4kuB_lN22K`=*SdJ8YZL3ZIic z;u&Nov#zoP%w_phr8_wTCsLah3#z$*WOI01l1M?*plvU@)ZoR}7hBNbkw~-Y_Diit_1@g_1kQYK88IK~%rg)Y1{t7OXO;o&yPsY>DQ? z;D-)5aZ38Kkma471cK?;*QX*Qi=Zx#D}VzthsC__^$m))w%$M<6%J6G3;}RxI8rz5 z@tfek+@$2vD`^rh0D9c|yGXPDF8|*?`p5CFhsNQvG34H(g(Kmp-@{8;+NC-37Z1Vn z86m&J6Pd1qdyPVQKyEONpVO&Qat=H=-> zd+#rd*>5_jaahfY`dWGYe<&JRV>-fbhWe;U`kQa4w;Ub@<`<1UYTAG8|G)o8VQGr# z=vDsxZ~hXek}i-w)u6yR{U2NKzjLnvDFi^PPW%slL-y`rTE1n{fa?F|N`74v(8VPP z_E&0mzbHD$Y{9b+M znfSB!{;t4M4&z7;vll#n8ZZ5}^govnU}Plz=12cje*aI5j9=ZAp&FCSJl%3bCnqPW zVs>mV&{m%_q1+yRA)!nGsL(i^uvUURBS<$7Mct%d>@_nY*Ys2iLpWXi#+Is!%vbFd zN21uiP4l=OOc~yjfFZdIW7!fkHcxIi#v6GKhezEeLM&Q}F6c%gzQP~Ke-P|e{!!=s=eN?&>CPoXDnh*eKY{w|TR({e0n;BqkuW0S`33|4X4!N!x8I!J zgv(u+qFQ7a1}QGJI{1F3ZUDNAn{3eDxWS{!C=&TR6A6mAeEHZea%+KZuFCaT*y!o^ zNZ0oz*4<+hv&Db-?d5q5Zw>1;>m-?zySK<`p@GrgSW~T{1jWnWJq_d_ALv#)lih~* zw$aB?c+Wd?QKH$tPA3Na*upyD)$X2$_2v=mIP_c2y@EF~j0E)e3-N2So(ZP8ic`a$ zEPe%gPvcy~Gh)u8vm&0J&z{J830Do;(v;upT3gnmsf| zE%kG8Y2M0xr$}&%a8f+5q?9<{K(NSch81g=K|GRe?rbbRjeL?Rr;Gvh8*C&?*p$sS z!I>R4^42ip6^3T+FW&BwMjRe;aQnq`9z=A`dEHga4hU6n*+BD`jIu0Xf0$;jyb#vt zDQ_(g@`<5uHE&7YGpNKzJ#q5NHKEFEb@H^ILl>BmQXM0aW#fEjf^yiNF!97R4d5hP z{mTl~G;mV9@ism&-ov{G4t#t@oj*z$74o+vfEn;B$1vm7>r%tjlW$%%E}w|7^H*wL zDIdmoU1k03fMl&DMVRi9F-ywIqu8E3Y{L&#gu?iNA$R{s>D_tT(7$QHX0=nc4vv?O zX*f>*r_z4$@$;TzM)_>)s}()#uGM}V&7qa1-HwQiO5_SqR9tJGQJekd+!+Z9VYor9 ztF2w;CF!1HtE^JgQK0+j{o8l%nnRN$%r-_`*S>Bw^G|&vvU2eEI2_Q&(?8Z5S0=pC zsF3wBPtJ4qt<<$LJ;a|a_6h}+ti+YG`KXA*D z5eQzKVefi)Bu4NfCwuNs)~cy6^j>ly51)JkSS6}$X`eDS#F(QMOw8j~;h!rY_ONmb z7@?$YH?YEJdFgpY9FSG}szt0^65(9NA4&->#67`gPRGU@?jGNSD~DJ^FD&WS5l(cK zxj)t!<~@+*M_|EOE|hy{+xJzZ#?z?>`)6~n@r0x@Hpe9Co9?b$arfPkk4%c?e26~p zlBAiK)Lfj|oY(Kn>KN;(m0i6@DiRF_j|TT^vTh(_Nd+%;3R9ov^s2A55nI%-Y}v(z zLW<498rAy>ncjuMo6FHj2a;U6xd)%Yx$2dHZy!I_-^VTABYf}tln$=VOjA-(s~uly zAHExD^p-H~o!5n<>%#Qxg|gOc<(MMUc(`fXPM&;jUi7Bw0wb%#zK5chOYrwGSivJV zTY@)e_dj>TU*j9uN;&{zjny9i<3Fsmtm7&3J)u@bJl@xdSu6u`G24>-)k#h+!;61b z6FwHrHJ#}N8Z`2D|1y*N8YMOxRFId(1O2E>^Bo~%Wh;cOa!v4(w92u?7Cr>2B!M6y zRo@*%hKI$pz5U&m)8xxLd`2zJ_R+2(MICBWcC^T{?R^u%IC{x?c&!*&6{#1J@HUH zazx=4!{cF$H1c_LVC>LT?|reXCcb7_=3prIFo-@2T+PAnoif}EvWp}F3Wn%7PqUliR^S#ORwZq4 z&G$YOD-3#xvs>s}K5k?5_4wM?i+A}g+~2F@+~A?x$kjBx++OVA@l9K|m`y$>4tH;F zW~3)qHLWlBeDO|~Y0+Tqy?`nSQd5+Sn?kwvWope{(t%BWvzc3^-OIeV$K<`3oV+yL zU_HiZA!c@Mf&j{H-K$+`$F!Du7lh|T=P02R=NLq$Y2qurGF0|#BEJvhu}dYmZAe!u z>{909c353$7alQepP6B0Q@~?^Z%5^Xcq-&iHz?FAMHKPM$h)yg+OwH1C-0)(4q9^c zGi+{%)qV~f55m5PMCZVmpHqM55g2~-qew*Jv4&4&H|tBgXr|gRwhjM~*uZ{cFZ5^v z+`9i=e{i8TY0c_I$}O&2!d*AkYgP-Ch6Ga{sHiwUr($O#O?P^V9r=<3NhTG;i}TVrge*n#4m&07Mugl-@2+e&-DgK^ zMc+isr@KDpj!ylSuF*$kEtJKX+6(WaR7DPSs$UCAV&b2!D{VH*d{&A`X6{dY|45K) zXX6C4SaOhe&(lA;kJdhNlBcHhU36dijrT?#$-r5gGV@j}NY|3Qp44+N_C9qvP_syi z8iObgi==3GH>_VJ+DkaE+H+=eN>;rdb2Iej1kO>+6Y-<@Xj`FN>wQt6fXAy8O9ba| zN2_2`uRzBZOALifzZFQ#W;nNzEcZE3y)Fl60h5^@q;dt6?z`noF)JUMM*bpF0makd zD)yuVi1Wr*@aicagS7-;g%FIN0hYj#ecf;v|461l((0{lrA=<0s@`oGZt_Mr4Q54j zUv3y)D|6FfI#hKHALekyHzn~orDWqeJvaJ$ftxC>81GxnuXTTZ09v%UI#hR62+#L%g~olL^$)dJ4Vuk(#3=);Igxg-Y7=KR4kjv`v{6mIC<$%%kdPhWI3g z@hn`o46I=xn+thzo^UP=6JIC4@qx_lkJ@o&Z-jkoz%JV##T)zJS>`p0&$i9Y?qc4` z^IN>JbId{#eX!4yMp8{&143QdiSvFFCpHV>GB`cB6^C?}T1_*Uy$mCSIB0jhz%S12 z6~6E_mdiBnXC25JnuSuVXFb+ijC{rMhHo?>TGLKLlETx<^-w?RH+w)n4!4t3p2` zz5259wz3E|$YWD5*Y#DJC%Wxyp_a#r?+yCeJ>z|>kaWmVdf1=CXW;ll;U5P(#n)xL z?En9EPJZf_hk#y4x*Miy0T(i@qdBOrUQP7XskLn8w@DPT=?fq7*o8}0ot_wAb>r5n zdgQh=cGpk4e+^(59{hi#eRn{USr+erU_k*zBhoD(5_<2TA|Snn&;m$RI!KWkPys;@ zsR8NILkDRI9qCPK=!n#W0Mcucm)Tuso!Oar``-SSLN50^_tf7h*AtD$8&%ui0&pG= zG&`XXajd;5>?-jk?{DQ&@)Zm-8VB8F7ey_StP{@lUb&KyI3AZ4#e1@kHSKC|*i84M zFaM6Y+q`hI>xp{`!WsHeLpRCLTA7THq9@t~k;E&-ev;2wFU=vD0(jo+g0d;^kE!Mx z=fXdyV4zEtr-iQ?MdLp1vnK!!B-mzT+(W$BRaRZr!VI*8{jPdbEv0H#Eq9LX8I0wx zdyni+$*(7LYnMC$DNQDGUnER&K;H41JiPB=ojUAP+P>~~BP(JZxg!KCeWaXSN}(+% zAU4;AphpFN#$5nO8y*pg3&5Pl^dD@h#KpmQ!)Trk&v4+FUr^H4bzRKZC0ONh{Y0ww z_ge3&fzDZKxwaRuSrnz6|M*{&s`441;Ukh(1RL-BfQV|vt zZyRxP#N=8T?8jEwqN3Li8bg%eis4HKk8~0%_l?#x^)1zQ<~sNNlR$RU;Bpk6;|S^b zcw;|ETYNGlywP(2&x#8Jy*-fLI520^OiWAUI6}9nJm&7_1CGKu0KjuA1jok&9QE1| zATu)?e$;4+i|e%TQP~fti%v_pXV63JKKCfX31|LZCU~F9U7Z#X*eBeXy&FM#F!y=4BotD6H{A zz+j@E?Y5kATk%NZ6x&|pHRxL#u?3mwh;GV0j5JJGeA>~{X~U#TWiCw&Fd^jp@D20u z=i>U>qw6dHB_XJWXpEZe}RVH!Q*>C;!IFJm-MlTE&X zSfs0!^$17;%K&VoO`VbE4sPJ$>52Bp+QwsvL*#Il+SiqtEVZn-@k-7v0#J<{!Yk+p z{U{e_3t;0SOG8z}RCj%=Jc*g2Kt#5Ri!6?RaGg~ONqZbsV@KG`d6(_J&@{JRq;@;? zt_RyX1s-ZqmAELt?J|guxwsBM+P?l82B8zDb46u^uojVBMYaHgdnj{F?noyguG(^a zGRJF}8%Db;^u00sSYUg^DGb+x9Sd7nO%)eK$5ufr<)Q1{<})%euVeQPxXXES7aKG< z6F>t%&swFHHRk8zkG#+fMKVAIUD)?L1^<)b*YHw${^+z1_IM*?ms8UHI4|@G;Y6RK z9jXmW1+(RZJold6h>!_|ob+x6#A~WzyZH*L@5*jvVSA(xrB|j^--`fsBcp={WI5w7 zjQ9hW*{Vl&x^Yv-#X=_p?991@J1c%GN3yP~$v!#k)FCk;i_2CB+V*ny@wZO{H&oacl|PD#ejF8jFWQ`{mGSNhqoktc*V$@#ATT}9 zaaTwpear7iNNc{R8ZI*`Eq!LB1dnc|EH>gqOL&)6*j(_)e)8gPk|IBYPO-l<^E6QY z3xhM$BUewBkS@yL@pFxO{HPyJLBbiIZ>y|hFY!3r2*ju9oMqNz2-`3%t#AmcFY&u% zL~9}NUKsDJSIWs)*23Wk%Er!6dprhvM#9~qW0%{{8hn|(`}R6dCzEUYvZ4pIx?P@1 z+T&f_$KQav^Th)!RrmHSEnew4cQA3CqO3jRD$~X_>sXtMQ#~ak(i51G{uVmTa_X=9 z%wWiF?!nYIyX$&!N+Sis3)-TUC)wU!5r-crLQE*Aj9&sW<_|xGAazfA6LP=!#SpYm;^ z0uT91ycU`v5}>JMlbJ`Uv;f+9I*YP+N1F`4@FAm<1jJDqtLBq(gYuwA|2^SD#OuDB zrjR%tsyv{RWMpub=S?2>QAdNk7Bg7k`Nv;O4}FSaJCY>Fiib~fgh#Ruayjkm-4|z~ z$&kkiXTx>!E?lF6V5o!x>FeWQgj6};2aLTN~N4v1nx+w z&mE$AmI~rt=goO$bb)TG1otXG=Yh&KbI$MPD<@@g`$?g_a38`y+|tXfR)5Hzss*a| zfQJzs#W!Oac)W8I*>#f=IP^Mx#)+k1WLU1Gk^k3&uFG$}MdyETCckknhl9o6HGSYl zD+mmB=JIri*w!U+1JMW+11E?68?-+l*)YpKDP;T7ApkWCIW7SiFY;MD%gCnF!c(8UQmi;p^zVn*&diX$@sirQ7}?y zSo_ei@krrW9l8|i>qWnPs=g#GA?Z*mfsuB(R!8^x^0pM7 z5MwWum}xh{L%=@Je;TXBW6$%%`aO^Rf|7lY__xqB;nIdOUkD1TSE%;p}jCuPrDmhhray4mm6@oSi+f6I*J}C9|>vlDOKav%@n&Jx=nxpN~j4Wf%1qXSx zKT`)_hq&qZPa=m$tAK1>-=9`;&xt`JU1~nh@8t*RKHA$XcPP0#>Xl<|{Q^noYXozA zE?be-pS$Z0BZMAKR#`hqUVZNM&c@KgC=8dEuIFkd^fVV90I*lj6?^$WDZi1H-McSs z<)xp&@6+Y~cDVoWN;l#DyySJxOD~82rJMo{Xc97z6MKCG9fvy^?^zXfE|37G=}XXI zMiS#z)OHal(~^piPuTgQ>GhXYnu+z@u#FY%2h6z$FFlr(XnCe`Uo=gANkQCf1IdKv z_w-~5KLsjA39b~#%yx+Baq^4$+3ERQcaxA=>#R64(fu$;F}wxjjvcKtr`2rZKv{gf zp9V){155JP#@*yqHZX-=A>6vUo=P!fw4hLxAo<%VPtbnVD&Nj%q5W~ms|$w5`EdqQ z>Ic25kTvafc1Rg#ddZ&Hoa_S-6ZgznkX23l*y-qpAwO(M>fkTY2x=3aTEv6v#plDA zOLkQCXQ8#=v)-BII5ZF4YG}_ZQzPp+x5?a;h2-ZPo+fZv(&Lc6^tV)s5uU4~RRP9n z@Y^VnHf8_Gk4B)G<2Fv~w#1?}xaG6JO_GwSODjiG>b2-0TZ5r*S5o+8ih+LhyYuOn zLTI*M*&y6W7;4=s=$tvS_g+}UVOFf`{8rtwCp9GUK+;-mJdJc*Xcqogcds&;P89Nv zQujY%FN`T-d)ie>H?4H8nt~F2gVLH!;gwI&@dS+hjxBbu@h#fcXhArag7mH~Q5|(t zd1TXQxg4ItDGc?EN7ZQb>(r$7I9adgfp`ZRdFL0L3cAx)vQBDcVat)H)LSCOOirTN zrn&xvfybQ9ET^FlK%v5pcrR>W;>}Jlg!=*j8+MQg_3!T#FaI|#@Uxxz5B-yViL4Mw z)AQ9#{L#OFgajz*McPk2{s19mwB%H7el5iDnVmz*_G(P+18F}Q2}_WD4?UDIa(CR@ z*rva4R-$)rX#m=OK91X{T+9xIGIp=6JM3QD-d>x^E@?DP^_#bE~c%ZAP;GPxBn69$xrF_8A3l z_)y*&%#GrHzd+3rn{fswP+FkqEn5f+R#O{(km}iM7GWfz>x>?mkE&Z!sr>#;rK2&b zZXA~d#bZ%|b5;9cI{iu}n`;}6G4Vk%dw}=tRaiTp^%RHgPSMEDW~Nu=C+jj3we?^T zdxuWX^TXwZhU8iCyx+T`1cy&{yt=wM^gc02p9tYzqk8r$l1^5Zf*$%GpdCZ=D9lVi zxV0Ejk?V}h8Z1v^){!ArJMgawaPw-23Fq8lmjPPqR+*v+HY$RXlz%kx$Ecn<~lP`J^cCi6LFiMtKqW8lna`pNNnG{o~+Zssb-5!wTd&E2^jd) zsPXtYjSQ}Qx9&Ui21$7Xia;DRcN%zn5j@|WvR|gc52;)dpz~TzBrb@D1 zOKnJsD=2vRjq`yO5R%=qb>UE)1-NKcix%ZhPMu*qoBm%JMc+xhKTu;Cn$Am z%IG54Ytmy~PyEH=n`Chxb&vHK(S_#X;3T6gZ=DoYi{-xhp#H6wu!14!hUnyuP9NgGiC?)v7Gar&9d%B=KnsMFFK9TW| zINkCMYZg>-GyE)a*9Ces0PZ@9xs`c)@X?CPt+g4T{_xP1n)hlQ_Bw)wT^v%;nxex6 zqvQ=C58)Uc1PG;u_bzEDl&SX`D*%Jy5t}_Ki!6Bi$VaV86Gzr_zpmMQ+Ga!Btpun` zUY$KHJd^f0ToH5)`!u!ApaC)xp*P5IfxR7RERe-DT;!kJ?Sq!NT8}@DkeT48*Yh#}7jA?xRlvtibE9}}UZrYlW9cPmWPf>2;h(>JQB3|^kMe|DjeQjSsrw7m zd$&}sQyW!|gJ&!%OU0JY>^#``PmXJaQmCGxFB{&Aj!mZ!Shkb9wc}0h%d=IJ>p?Wr zl38%OJ$huVw_oLSRkp|+&s!6ZQ25goy^d>e) zRjHnTLpKd{b0#`a#gku0iq$fHWeb|BIRVi~Ho&z3V%UiD4>0Tf8y)GRo4(tczQrRKUl>Ve>;*pj*L z=>}y5%cOAkB?ewkgN3K1a4@J}`EHF#tql4ILBHlri2>S?8+t<=Mn~{yq$&A&!(IW8 z1#iz}{k9LwGoI3pZ?px;MW4d`=Tg>%&!n_DK@6nV6kZvk!;FLa{rN1<`nd)1PGT=) zz0&RKJOqrpVZYtM1XZ zxase2@(uVDl}>5}J>9C}027qcKy&YbS8@$^gviNWI9-T(4!o{{8(f0ymtGcg-Ju2fiQZhj!{w{JX4dB7xP&*kK38%ypGgRGU3#s|uU zuq6sP9S+QN`B$n}Q?8AL$-#UREhSwrpUQ%WVc{`iQppiwFtf(%AfW!@DYP+b8>FF4 z9(VR&pTQ3!A{?fQceM;uK=fS>%iV90ySGoW#~-3W$?*2{B0CG|r%2(sT=(okc}}&= z`ppEC&_ObCKp{)T5AD%4YTYSVJ)paMbt^Vm#$0dFsFaH@E@K&2<*>(WQ1;~#s@|tb zs7Lx4gnZ37lFq))8W#j=lAK|RaXMWL4}X1Co5`(}(lyE`pY~I?HaXY+@Jhwx zczG520)_hWi)Gh2PO!b+6A8hmC;L$c+sqhbhfHkj@xZ)uw=gXvv6}j#jl#{+!#m%c zYr|85)TLmaxfwCMA9_aTEa34oiPahzLCs4CHwxUGxh7L&rvVa2(Sse?5B+D`-GhRG zHBysDj@C0l_GuF}q<3u{>juSoEHz4AQ!#e?w6{7XK9EGUP+c(oblZvL`cFede=jw^ zKgc-X{n7&M5~-Dlf9WM!;>`lYkg7nqRnw_KaDREb0Z%fR!8OA(^~t4IMms;2QOftx ztpDbj&_|gTW2WU#>dUxjLWz8{A8yd-Nq9gwLb=lp1eA?X9N|ilR3GI*6 zIri8KV+B#zzr4;;RJ=r?&d7)cbv)ngILF3Mf2m7!x0Y7J=3R+5`K<PoOQVaMoDXy1uw-Z&GJvaq*U*s?Fy2<^t*MVW=3eUqv)KlpN@PBKLQo4roK>Wu`Chv?=U1+FUFbS3;xtXey?dc#51ycf&w)V+bUOpr zQU)Luug$>z9M7M{iQY?sIf+=COrM0kwwOS1rH1$HAp33(w6if}-4feh*?Ed`@!Ebh zr;O4(c!4yBTp&xHrQhP_Z*{*D0_Dzsp?)Yp7c(yx$>6iiJxPqx z@A#$6DapgBlI!5OBT0v+bJXhv;poFJ9HGhHs06CwI!QY}U&p$Bq3qAc!X6Ezz&er3bBP5`wa(L&`O|sQ zXL%)p$H`iv?K0cryb(Chyd`2W18A&DmatUv2)r;9$X7kZZBn^u#%IQuL={`#XaSSj zO3IB`KNA=rwYwD=roqHsR2pDJUZ|jcN#zs!ZMl%Mtfe)Ez7f{HSS;?ACLeCy%8Z7Y z%z7-e@SjOJUD87P&VRau#a@BITL9P_3_N8{Zu@8*g$Q8>*fa_>OHbdDFPtGQYa^R{dSQPLFh zmxM5Ho)B>9$+=vEkoF5F`(-f*JE*S-x=0$ur3SLvy5a_s!L z*wBbu82C%PBVrN-J=rp#yP^JOYU%x4V}Swv$R3vYjSB1-L)VL^*PJgo9#l3tQ4d}U z8QAYMaj&ONefOB}Ypmhkl1tJ62(3F`5B2?4hn2Ptj@g(sWCxcB%GsiEI}&>7Jer&( zSg2ZXi=1oSLAbpMLJ7i9{h7w)Xun7HeG0qB4FiR&F_O2ThC)12rn#r6 z>$yGc5u=AQPdA4phit;7X_&6y+#SEx)w!&4XZFai7wGC2_Ujj;(3o%&@cyRfiN5OF zW51DmM6u$`|NiFlv8R&RpooiDoWX5iy%UZ$G*zGkJhMc2TVpq}$!_$PIxVW79jfus zM)v)et=w$yKDqAcedF7ENUpdW+H0PV>o&9C@+?n>r8y~W))+NC@I-c_ zP8Xrk?}D*&gV3XK)1LX~Ap(BwNrxdrv}`|S0jw?^bg@@WOPbJ*aRkn{Mk*2Z8--iS zFiFNz5h`r6J`X|P1)yhbc?o{`Cb(LDy*rBj9DXG@HEXZ6dcwyRP;H~^=)e#ILF?f; zQ@&$yU;oiuh&iNV1h1`-SAbq_yzW$CfaaD?C}tkp!KA4yE}o#}fkDgHJRwNdU>{aM z8N3feXy(0xK;<-fs#H>LB!du2vuA`Fck)@fuv`se7jX)|GLPv!^@N$tS2 z%&&xh*s-Qd7qH_`8q%s=7TW11?;trUo9;wPnO{2k`ieT<`)T`McMF)clM&@cv&^o2 zeBwvqxR*WfEwACNL49L{j}KcG(4CSvLpB#RImC4^=i@o63EN0_dNfyQQ!;o^uEt=s zv~AV?2?s??jv<6;3vPlVwQzQpk3iIP7xFP?JO)*6p;B%H#g5mA-F>14URDtaHJyQIYj=A#Q_N=YUB(r$JO#_-&Pu#OIQSW_a5xu6uD$FMv~Ii?<|@}byHw8e4*2<%xD?5!?D^xvU% z$cxfZRItCt1~s_265506&z)D~;W94}>CJcM{NHp2JaUZ4Ul?SB5830Iiet8M#M z0Sd4+N%xm2anhEJQ9`eUE2~D(U0}C-6cVC)*s<>3S?O8TeFpkTN(aq3w^hbvbpIS& zgzKPZUhv8wsNz`nCNGVO@_ll~k{Nw&)7lCv)jXBRbWWdj33~0gY8p5qLr$jTC4{5| zbY*pZ#i!g;s5#lx!KWk6+FgFeeLXtBUJLN~tejG23k-o3$jWKOtmumRO?#)MA@a4D zOw}>HtIdS%3tV<4T-=r1u7zu_wIBsxau=Ge*Fw6&o$S>zH>*n;Ip(Z~^2THvSJ*-e zFeYz5T<^gdZO{#G5|gKoH%kqc9D2Kx$?dKe$GcAfXjpWg@B=yp z+QHH@lCMf$W4T;zVif=?Q~o8r8KQZ!O;N(1v8u6D_f}b*WQ`&TM<+E3NqLn z%%+j6n=>8HSkSQT8d}g;(ypEsHS*;>tIu9qCY1}XA!`v~U*o*{Qjo=u=Ca=402~@t zQ3l-?cP9aQsLi~;@*gPu5A$Me+JsLe2OzBp-p*t{QiIE+E0V}8sM&pq>pbt%;j_Y8 z!x;r605YSjV~r6vzO`65Sn#kP4^|vz`Yn4&oqW0wMTo&Y|$b`9yxy+*0vZuHD9+%g#35PWGeyhIzCz3Aqt z8`Swt!K%28+vFLX`xM6y>ENGhU)RehZe5A&ie8o<(z5_{aJr?=>+4e`OLEzPZuYFZ z0Wj^`&uHFIr-1-mo}sxL<}N7x(RGS-9y*5jrw8Z0In_VJ@g&_#PS&*qYD%?#j;VZQtNUDWBWj zeyb12As)?c4|u`JR3hy4>z zl*rC|#J!Gs@RvWaHu#l2241Hvy=~t3@0urmU_So_AO86&eVh2tK(oRt;s4t@X7S|b zJ^HrE%?^Ipgdfn|e;N7w;BqfgG9+;G!}Iq4Z5_TI;yTtxq@_N;)y@7D34RFOpOQOI zisU6E&a+SaFYh$`@0LxlBCb>6OEUhKSN{BW^wLcwCSJOc%gOxz9bEpq=<+9$_LdXZ zF=+S&%lzN%|Brn0gJ?jF05k9sg8(xWm+3Da{~p=-S6lyEhKge+LNwdO z;x>Od=JdZ?wu0+^sa-tpFUa5G@IPgbzaqiv+w?vEI&^E{?diZSa_XXKi&VfP7~L~%h(S~7n}Zf^)tOQAA9a>$l_|* zRoBH;m(X_ea(gt5L4#YFkf?C%>8`fz=baRYEtEaN@9ZlQ3)7x>o@fjluC&fe6WTqv z;GMiMG(5%rR^od1wv~$I6Q2UHls20UB4Cm%6>^8!?zP6QJCvx-J(LMZpr=mK{D(L4UvbNSs3gE`O<#jN z@$=gSx&AN!W+*W4U64cKotes&fYa4^)5pf3PakaGKFZuqVL27Lig6Wtbe-4mCMBhN zVg__)cW3(>=ak@1&|MGgB&V$ZVwq}Q0UY@Bb%|SmvMAHo@tCJSEAGszpkTiEv`wtY zakdG&ybAOHD=Da{xUR z6n8%^#Lhq;r=IQ!ZkPFN-lJzT?FeDPWiv`T>U}~yT@vtGy?UU`cbN}?!TFw?P^*YB zM&&{eqDMwYyR#WkOAHEb?=kaf(zo(;m6PGom#1gIR47bTY^G}ob1`eY=~;@Ma;kk$j5yWadLgCI7<)`$aH?k~a%~;| zyQ)*?73PUgPaMX?djB{2H7$~Lmz%?=7#*N%kOfuu^XQ9d{x`2;hA!H{3l?|M@UKV4 zbQ%0+Pn=<|Tb@`PKg6SmVthiyC?ewE@JygZAFT;9$ty7O+s`-wY4%()b6}xNUYbHN zU~1g+RH{!l8q~%jf)TrQ;FO+v&zZXtVmV30)9zotXjw=cKoJ!f}6f&`M{^39DcoVj@|Tlntkb45qynsv+)x#DydQ3 zsNEUst!nh!Z|fDh`hzY0LnQvIFPxu{TG3%8+z#`0_b6}J6w0Ydq{mLV)vSFef-$^x zFGb_1Jy$$9yzwaqGQGR?^T!2kn>$v^!&f=uxuZawFop^Q5LLSaL4y)e&aAg3J=O)6 z85q`xD_$VW%UPD;wgvX5Doln^-b09!nbeE-f z+?HVp@GTAxBaRY1Tp9?I!9R59$z}in%IPiKO$kl04_$-yn_YjW1=AVH-1Zg)SNeMb zTNN!kHtJmu$6e6ajee=7Ove5|i=S?|`Qa4&wTVk$_>~aeYfPNjEKU=E?Y{f|P2CLq zQT!U;h~(xvpPqOMagvs;RfA)Dd8Hjse_>CxEECV$js=CZy{d-92nBXG^3q8|ViqaY z@0``s7pw(6NBxfC{@*^86d*$OyErO7Ei>Sh`p=aU%F5lk?AG>{Sdbo;C4W@AI@=N%eExJya6+Y0x2RoV%Nw z|j36ZoZ&;J?-{7)L>Xr*(#f}8`j@yyhjL>3yC^(OQmH0HFQeMaPSy%cGb z_}!avPD#^ND-uXc&7hBNMh*5ZIiz}yveJznd-5LzG!-PBSKo*~IU(HIdvxGtB&mK= z?ZBXV@=_t%Y$7VNV>pSf1id4A9C7t>4+P3Lz;8t<)gNA^UOnAyie9SweR#41`g+(k zygeqlsdDd{0P}bT%j6RR2D1?bGlA;hki(}33%>4O>;C!(|HS1{X4mv+=J8uF7GvZm zb93%lca10*9J2*w(lgT2R}+WAzIrvL=CxkLeT1-#2Z71iZd{_@LG$TbXMZS-w36lV zN!xWRa4@a4AObx(NTsAAQ|k0TNKTM*@K7T19yz`y34{Huy zn*MEm$qUtw?FHg25pCgC#A57OFKBL-cHHe&uiroZ|3zF?<(s7AB5W*-d^$O9eAjq! z9-SVyA(qAIRH;^@q!-e)Q3Aa2et!PS9p4@yw>JYTJ(C(6;mqta!LFpw z+P%VyFR7ZDv9Q+@c5B)i_K6^EQvh9ru}kDMPaY_*V}o90qf383~gngTqxi%Sa0c*la@r>Of{Z5nW}wwB5VwrnVYpANt4@ z5zOLg9NEEbQerbLQ~0xD2<)a2d%j$!Ra7uuA>;YZ3|yx72P*tM!vPJ_u`eO}wxIpX zpp3T(#7a`ho1mu*C)-#YK<8F3#)8@-sO1SjPx*%2RIByA-$d@KW~>W#tvs@eB4FYq zD>nt>vhCkEkatr=Nb1N8byV!=wH3GC?(CiHfO1fymUbv|f6Gc6wG50^h}CRGZ07)_se+St0>2()?g zvzeg1`}K0j0Q)-(rmzpf0~}1-QH3Aj-S>-3NlVVB^BucDA zLL3Zvy9#eWN@%km5g1en$lJTf70bgx#Ugjr-*@KgT+yVC{tl@J**0v)6>9Cut^HnJ ze zyUincJFDyKqVH5#Vo-KLO#xye^$NjT!JL1#YD(|-n8zTxrY zTK*XZG21x5n|Dr(m1c|oTrBZY#@%Oj&xo}#KqWGWfPOYU&B-;dG%ZI)1NwGe6qi?2uR1lPv45#0f1o47}Sh0 zahj{rBup=jj*ceE@0ST&e0~=j$aR9BhSS8(ZN@l@-EIrFkQF`QU!4I>UE0NCZ0%g% zwR{*IZE*YgSoCB|UwT_(UD8?KA2+oU9*1Zd&t*Pb4G)Sn=D}qgEXo{u++c~dfesa% zG29PxI2!omrs6LDX_x)*_ySJU7zrZvhoU{29{;Rk`8bK4sw}adGSg2sjOLs^;4~r& zm#Fs0aB)2_x*p6XLIbl=Gi5wL_$S>vuZa2c014YkJmRAO;V~||gD#dC6i}_QI61ZV zc0iLc;KsY~&PWreu8CyLjM!&e(>ULyVBXGb6_LJCxm2GH089KA1}IPdW0;ry2JnK( zfeq#k>8z3H{gKwihmxLMwS-gfSY?GWg;7t!-y89JMuFuQJ*GGs{!o zxa+&I*7*ihRye*R>$jYIx;WS&7)PY+IAmycELu`%M){jLZAxp1 z>k$!(`5hs9h&h*%=rfA%x|_bH(_zX(kVC_nggPV7Ent2To!|X8-s7)d0UIiO_x^z~ z@p+OP&&D2-nkY&1b?>==Ua>{%HcF11C=*5f9yN*05T{9+X>=x(RyFV7)B*!B5Aps^ zTO<}3`-0f-pbWIW{b9g)IQX?OWpk`fGG^Y#K+TW&!S4;%A60gOENKkqTcLCPNZ^gj zI5M}HV26yNhanIEV)SQ$rfBpXw+GWWoH+agOllR@wag z{B9}yAq*s}%9}hb79&X@R?{5KMl#1e6|%+>avZ=t!R`fu9W9UiNgzd>Oqd~QDU6&= zN5W>n=~Vf(pVjDcRRCUeRTyl-!yJi{`ZCDi89_JYX5m+FZ`|nNn086=ABBi~He|+_ zW=)cky?y)JiP6rF8e1RB5eF4s`2q#(A`}Vl@^bckK1bhBW5MFDl{~ul^F&9(>#jZ&iF}m9wKp?9vdq`lj3owlOt6{ysI9 z_^}P~V@7e?q5-CYTXxA?k?8MY3yT}zbEt+BqKi&!+lnhPs!(>UMITQSmC= z0k_g;i?21Jyw50|Ln?fnE*z*nN4mtCE?e?_;w2{+tPXOK~4vx)D8tB4d0sOGZsOg!bDsz12zy_G=hkXZZrM6eWO zq{H04Vu{%Z1lH{HMsI>yu=f!!)vy3Q67yaj!GSPE`KknKw-XG^4T#&0tz( zZ_Q9yJPFN#^^=ecgCs?6dqsZd zsI1^VOCHlmLOLtW1&gg^711}thnB+_A`HWTD&fC=DIa5~7@1P4H<{xmT(Q8eK}VeO zx^Wrz`pB328dVSB6yNH)Z=74^t)+LDA|M&WRIq<&0StMr7 zcIhKPg%aNi2ckGj8ikj&u|)ef>wR#>!pqOEsE~0zZE6e0gj`~P?x-Ev?d=gqI4&29 zEQ=Nzam%C~&0}|XonRL1y`{1BJAwoCvH2cWn@*dVY56rv=F8*1u|Ut%-U!){wq8G^B4+77@4E$r*o4CdH)$K!~7%lUI#*+>wvRrDB=F(4WD z2}-Y}MU%|+v-|r;FNBYm+&(Qn4s|r+yWpUUz>h(?pfW={veNwxSxc_r zjJvp-t&!8NaVMF}X{);PqPXsKNHwTrn#7&FxUzs(+k!E6w!!~sgPJ9Zp`_7}i24F^ zbsw)09mn|04h{MwZsRX^lN308CptnOecVpjBxg0ee&7y3L|Kfbil0=038 zzKEao--E}MFMo?ePlQ1uI^2VV*IIQUmE4=6kTw;P|5R@0Z>EzctA`vu2YLkMjkk%O zBeh~nmxX>uNGDp1r71l!AMBO7DoeRDxa>I-FNpWA}5{26i(Q+x{%<{XNxZxI0lE_Ji`q6N2$Hcvax zd-~Npk6bO8_Nawp-REYiS9tnA5O&#R@5BY_n2_N6X(02z`8h!CJ{Y_0Qa z%l}J!#YkQ#d9M5WDwxaYV+({@xueAcl&Io&Xk#Z~ycZfdbH|8h#?0u0bUvpG#D5~E zw}o;|uRs$FvX=2QpQ;3RNqf!$)gmYj3X(B$m27Wm=}F*0>kbBM+kL;Hn)AgVT5fP) z0rs+inpzzE&^_B1+aSr#b-1W~GCI!#nIOX636-2_1DtE9Vu5!YUk_}GR0ddMXz{uo z#+$ndamVx;pGWhJx_GLmtm5Am=w9ysY3lZGM##*V1O^`md^d0!z*J*e1}{Pz9j8;u zXh_I%e7#82=yefAhBnGPIyh}{N#AA}vbux62C-00Fht$^CXd*(H?LK{G7Yx; zRk+ueR9x$MAH|%X!@vK8^18KuP4GN(hh*3i;YIX>QavWjG{!+9q4HVlWS65zP9v2y zH>jBpCp?-{5R2xW?7hhPJoP==BcG+{!b^8SXe{>1Z^!rV<;{xEuaa`fM9N9mbY(E( zW3^3h-_e4X_QtKXy4=)^4e+KOU)i!k3ER9%pm0y_ioM zPg~3o8}qZDP3wQEp&f<$Gcxyy9&SFxXY&i7vL5e`u+KSX$ zSH|3bEW`#}kSG39;L4XoRwunSo~WlUZFOQAd<~j;Dus!AXqgV_M6>cy{>jt*^qjNj ziCIHYSoGCIwZIorsg*v}J@ReETvz^-5najMuEYm`73F$!Hb<%DYo5riH*=Cchcso7 zikiR_cMBYk=S>I20LZ(ZGmPI#&mrC9fh{iAGw`#-(-&wQEZqA8M)61L#MTpY;N^|4 zlbeh$zhVCttxd>Inp(fU7N5MUZStp#P)cU39zy1eA$gF=o!{*qlu3phVy^pDF-eRp zJwFMT8{NF3XikQ_>oViYB1leUG!Y(nswDqCmqrdN- zTvN2tOVJpF7#w4umuU{Z9s;~zlLX#+03glz(?x${nqC%Sy~56lq3hj>d+&|PdNMgZ zto+6$-k$3y{xixfX9e)I(0n&^8ocY)R(wA2y|`DI*gZ0e^J1TGnV_DQsz~EF^9wxM zGReIwUK9q<^yZJpDQlPYY%Dq(HnQ(-Uu%9KN&UAW!u^ShmQZ@#{oK|IH~AzCGo#4isX zoRCDS(8g1;hb?;a5QM!wE-cD>T`}lB2??lF#h0hR9IPDVm^TjkrJ%`NktpVX2o>Tf zsV{xuF!NEli+!#N4!BQ543nkY+Yd45^_ZabqVlI&AzT~!H~ky;Ry6WCcGGN5sG0kH z@uQPNti-H(_AKSXU$dI#*01G0c%xqew9kJgw#v^;J66NT;!+oWYf}4N6n*+DZ&#Be zqv7UYF=e*+Q$sF!Ge?TmFo;Yf$fdh*T)r(|3BsMbo(aS$nhkk3`e`rpqWUqO^8eue z4&PMt8q0fje_S{*ZF_G`$a7QLtaoi7bEAGur1rEt!eC8Bcn7mzy(Fq3n?|=myoe}z zt+?T~-~id*c*3y?H&^(f*>tg5m-AQyAKlJKy$P2s7Ii{w)j%+rzJGp%3m0p22M+=U zyxBQ2W-Tb>^F9kS(A!-}upq^7n3}+dvnaA1K5lg~N_N)zieUzt3d0~6eWK8~H@k8Y zhY0}5Rg(@#Qy|jYk#ol;dU-wB_-|i^rtkF+0v%oUzBHSjZi{Y@U)6}4&MZKJQ@u8p zq9c0({-^+da_RZ>n)YveF5@&H4%okXan%+=OuudvrM-XLRUA6A`a8hjpR4F10*Z;g zN4^{C>2Na-&?8&;g**|qZr8)o5T_HwbuS_BOaa?Q~Md_ z{ugHjUcBrqE`MsW7VHOk&30`m7BAXO)t^v%lIkD~zO!z(W#3yEb^fwgyuQC~SJE#jLlt3q4;$C8+0g{59EeR`tJ zmT|~xRA2xghiG!j%Gs#}_^_WFTMBu@DoCNNmS*M9&~ zt;}^jX=6s<=bJeHuqObw^TxFJt#Q!c9BHyI2bno=+^TV=eLaUb$=sOkr%M%nxI1}d zB(UPi>=pf5@0kRA5cvF2x-U8LHs46;aVuSHSnTG7zQd>jSpt!Kki@r5vE**0oL`?? zwJWje(7baU_CNdj0@-p+fLu1OFSC(66W~ku=lpTu{GR1cz*M>^tu3w)jgYxQGq5(y zGpkR91m`teyOYvhTUlMHY$1mWPg(gjv$lT(i@;S9u?ud||3lf8$3wkF{hG+WWUmk{ z$k6OK&FJ{=E-p;oMeU)$XZ!jYJ!8UVHVm_a|F!n-` zl+^J;T*fAB)vU)qb}9q)CZFCGX`M;GQDDc^Zc_8)oV8o{^z2O3vZ7EL zBu&MSLf|lHDSASd7Gdl13onGX&UHp_vDo~#7@#M#a>nQkos|rY_!*(+$G*Ij6-Txm ziT<6R@M#N~<1zQ?TMSR%ZuC+NlpNsop6W7#eUO|Dhvz8{9W`F03tVL&)?x$nj6^TL#Uo5TZyN?$baU!VO z)$ZI>&#d#g8fvo)TBZ0(?TaSbAKWqej2ZeZdVJ#)8aEKU-COC;T@ZQ-GIPOalmno{ z+420*-!(zLpx+!-Y@mWb6C?E=V;q3YbkfvY%gK!Xg|RMRuF}p@`vG;eGr0}zc zV+4?Q@Rp!aj$vV-pbT2jN7_Vza(XX4UA%}qS$}B;4^7F;Gg82Lehy%BtKDf|E?s*`BYu^1_NC>54y@9) zGTG<>*?QzdjqkqO|EP76p9ZpZ`i09ac-uCGUHie&H-zth->c9AP1u^Sdm~UYq4-qb zQc0VB_roBgtdE3epY6`t6=uw~P27fELHI^R2E%6%yMap0{VqCh3$#E$vD8TYiAJ?v zrIbH>`$HGvDeEqmtLp6~ymHUPog~eR4Uf%yQco!skCyLF303^Z?})5yZC3nZT=sHQ zxuHWyweuM6Nx~YvjNGeXlEKN-x4S#8X_(|kKaWEQ&`%8kmnF3Y!QJ3VD zmbybH8~vbci=4r)2=eI_fLF|oQp>3HvTY$w&8vZ-FZV$XK@+=~W<4@C>1gP1+&*bN zD^%xg&+4^W6Vfsv9;SYLH%40HYiX`Uq>dGNqOHv`orbt3i`xvPj!c3HmY8 zC`}~a%<=KzqaSyMY9+{vo~WvtZQ;~P7^XbvE*($*V2^)^cbb8jh;#(CG7fiqFxGtZ^r1toj?Wa$44hR!E%(A1e0aqz3RF;;U< zWL$K^eGHHE)PVk096;Ts0fcJEYX*dw`#XlWMJqzYAM6#3QZL;{(PP_mHeDK?UC*fY zHf~QGPTjI3Zgldp078Qn=N?7ceWu@cAhC>~_p+W`@v(&c%8RUqCp91*JswiyWJ%)2 zO)rkdn;8`sCY>i7sP|f0I#j=vF8HLM=EtC5^?&?^j>^d$zTkFo#cyS?Tspn`l7jN& zn>`UjTIRo+*>??Npae(;@_x7FDvG4;fh_Z*-_XEB;e3VsT_8>TY6TUA)P!Zuc!Cxr z-yc1*u6Euv0&1tTT1A?g?6CraSJ~L@ln~u)TPNOAE8C+%A$eu{X&{KGrPlACH4Zxi zBYjApiEfE}e30h_<5qm%#Aykhh(m)S;65>^-H`5Vq3GQGu}5cRHN?2$&kvhgNAVwX z>ikmRawb^fn*E{YiV8ehg^~h4`PX;G@-J~nlta~))|+=mPpc(9qX9xfzSni|Z|;2+ zwZH3|GJSEQ*uWfOHaxq06R`g?re7Scfl_LY(5Ls6QsLVvO*rkny${a!&zWisTYK@a zqBD>~4xhbEPhz13|fh}vXxI>dA8XQj2<|vRtTF#*-y*qyk5d*)4*^dt`=)t4Yl7N z-sY`kFzX%C&Dkb2Wj*@(I^mMmKuVN(+47iAAei;9K&29fo5VK~U*<+g9O{U)O2>`P2w`|POr zF=|0S^FlThq`_^)f)rj*8aoYCPy1M%r?ICc9>n*%s%@U3dkT0(w}UpvotzwbCvqv zzD{X7eJ0-auEH|^Y(C%JV$H%X8ws(~(^IEP=qY6-w9|Cc6TnEUjwTr!cqDMB|8B5D zrV3JQ6V+|2e}gb4i|<`jPMucNI^-=w0k;c_NPc1m3C&2uwNyXPJ&yiosF=quRp0m} z{tj<8tl#BKe-KiK1~)L(Qr`c1xVqm-LkBjoT&e9VCH4ERld(3ry&hBi*@WaY&u$eO4G0xs#2@MYs_yOCFbUHPi<09Eja;@rbp^_cnm0ey!0OEc-w>*n2eP6 zHTIn!m%fv3N^NleBAv?UKa4df26vCvo=7Ov_U?LuNG88mk%=C?TrxYycB$b|!0(>q z+k4LHH$vdfgT@Yft;E>F(^7+s6ToBv@gxS#EWJ1QH?uQupRJ|)M)ob_rI`~K(zD<8 z?z14!pVLrPAp=#-D*sHH?A@dZs+Pq-b2zbC5ENyKMK5w*)3ZRgj4^oaAzBfl6Pgt| zXIfO|{WDC<{(+4h4Cn$XgVd*d-x2&|jo-ADQU0Y1W}8aAk9*l(CbdEfEzxH@=Eq=L)(3DaB)9x;zAF-e1ldk~7SJk8)U1kMZR=>h z*#reQ|DuZaC+hH+MZpe;QaSiD=PAQM@&zke+{fc0Ehk#8Y)5s5z*$XNo}UvH8HiK$ zfoFOzFeNE}mq*5x{%MAa~1cqEO{$2wH68^m2xxBy$&8N z8X5GmpHY6)LfLihQYPYmsoZsfdn{?|M&8Jq(G1*BRl0v(-C`e3$qq&%Jvh}O?-0!o z`ORZg@bfsNxn132Hk{d>EMCnKrgPVd)%+kU3rd1|w)lB?d%Ls+9kaVPu{ zQMAv>s-~qwT$;HXIF`dsJCFbMX2gfM8zT{kPAc4CyHJmLZ{c4;kwtpgP*cKNbGq4| zxWA1tH;qAU1`)4mHTG2a`&|G_{ob1LvvnNr@ejT2%m4rUHzT zV4o2f$n{m{tB1MY`&)KNF9ezvosKcY>=MNr3CZX}CZ2#O%?50vE`H^gWM&g&>ZB@2 zQ9x0lMMO{dk;B8-*}?jmex-r(8z*48lWhTTaTm_p>H}YIM^Ttv9cme0Z?6)eM_NrN zw4CU_CV3Js6}$DP7P<|9zGl~|9!!RAOb_Qg9H!~thEzm{z@=vNF3boY`>mUzEU7|{ zAHe0)&xj2}RN5nT4`fk$_kC6Ia$Eciz6&dGCHdil^x~!2vd#`SzgcU+*PjCEg%Hd~ z5i*x^w#(cZsr)3v+DE^X_jOZnx7_mUn|toueO!5M>$Rg>1WzDjj#kMWV$M;{QEDVD zb#>W|kzfyWG!pjYJ={++OZQ9zubq?ZQ_^+noR6u)VR5463p=^Qkt4Ro#&PYVR$?{& zaOZoupRu{4_aYB$&e$fTdSlaO2n{{^FE26z0n`|uOFGW~NcDn*86sj(94XLhDkt^d zA5Tn!jHHCX)7b07Lg1QlS}R?usE0*E{iO?U>d)Y%02O{?c9dL+(Zim{+Fp>1L_R6t z^5&$(I1_*Q%pa>bY?@fc@p1g=mt&{^UOt|aw``}U9u1BZW?B;XtGE7l1ve^mX?5_J z5i?he0ymw#;m#o0OV1aJY`=N_Z)?%yVW-$kQihhEGI8xEa(|7@JA=oh%(S!-y!ZaPcjICxy%AA+>!%Ug z+7fpyg&)DWae^oNooV!SghFMP$OCx@14@4)xXMB$A~J7sVRYz;mwc}*-I<+3bSE<- z4{ZADi3Eca@qC+ff&YOjBeKN_E=-F%{F}2|;*QoEA!}(9p4+HYJ-buM&O16r_^?XD zdIF)74IOJ?g$I=(bQ=1u#vaK1nC|ZwQCcI8Y-xF^k6V?&oUNw8p8AtbunS>%ApUUM z19!JN8bVGUFqU%>-HL14UG|&bK&U{{ED6Xg1-N2Jiy@LusRQw zs>1++n4pqB!G&_sD1|L(&I;DVreAeNz0AuUW$(VY@JkPk}hrqeAlgKm)PZE8r+JXdDK-#I<>nEm})kHP>-5T|DBtznJI7*c8Y?#v%2XXlwLLzBcpxlAp|~x z^tyY>TI|=~qTBaA%-W^^*Aloi6yw8jfM;v>KvCKMjkW26C$%CoQxy-K2W83YrIgUZ28hQ!_puQYAZ?7 z)52pE!mt0v>i>OA)g)RP+zw?Q1;1!WU!jA7V3dV%7WAf)eQ@ElcQoAT+AqI}3SvCx z4=0gbM#SRXJh40LkymNx5C0>?R2mFQBuchVR(HN=&nqvgSD0EkTZt(;*)A3}CN4_M z`_#Ajj~1F(3u%djXbJ|Famvx4^fGOC49A?pO)7P1>S2c2FTWCI_yKj_UNrDt!5_*q znL!vxOy?*@lXPAz^c{jghNbS|KWSz)$Fk+z@$*>%=pCMFUmY==DNf|p$;;6xbRBq< z7xie0`1Z>0@aP*t9X3N;-S=Xm@4zMqoyd8vs}@BHroeGO7mKr*{AwOvyr;Qw`JOFp zDSFwqsk<6gs2idt6lxuen+kz9$G}&fb>Aw>(h$u-N)jWr%A%@ZL+wvzFHS~Sq*)rs zb9gF#Jmcu_%SM0StO_PBc8Vt|+RP@0CW!h69kroI7;pErM?oJrehCRm;i}@VUy`mj znp^T`wRMBGoKcH_igc#rPR%SqO2|H35QuIb?K(1cpuFRKS)PNUaU)`>9zQ;&c6MLS z@8I#<1EccT*(t`MR}bH+Vi&WdfGXsW?UrY?)@R#mI-^$QgKHx%Md>r<@YX-L5cdZlFf`;>na%tcu^Y zzWtDikDgQvTVg_LzArnU4fRpOV^T|=dO!XJyuJg}dwvvI12r&hq|K0VvjThCjVVqd zO=~e?a6x1Gtmv;k8zcJOCN%`!GGxq+i;)fXENKG}Z;+)6YVD`WK};JP3yG%KxbY&T zX>lTg5h3tQ@?cf+&8QGK(-`+(phJ`AMJCM=557gMPoHQ*E@V6CzSY7B5OiS_h55%0 z|NQq%RHta0j8gR2%c8U!rBiHY-L|%92F9qU?GeIuL5wnpE!W)!0?u*cwo<)C`MLxs z_H=&*9R&HFSZLEja8t)!$^ZVzA6ukaMa7R}sA4~N2;PB<%?xfD;ie#rA=&=wCp7VA z664a5GnU2Be3#a%hSi8SG9o3wef)oxT& z(o?r)(rg4_#dl{hujXii1bdQWrZ}H$n*r8Vz?ms_2XP=c1X*l5%XYiF9q(_ zVVij=Hp|sJDo)WSd-=|Qb8^?k`aZQJl!^EW6S?ivpIZ{1WSN$wnYuO>EgU=g>tFlv zPR#!E@O|&!t_~kTo#5ot0_sF?V@DCtii3Z3Ig}4$GJ>FApkSpK#_fmHqqL_!BQfxY zsqzHKmVyhR4JfY@`2@ABcX)h^zGFg<#2>#CdMfO*m5YLH9l^WX=}SCUYI5>A@(_e8Two~$&qUef_)!w^ z0;2<^3zhYJZvp(j%TB$UAxP0k?Wze-R<`FvYE)Kwa#GT5i6Vbo$6*tpM|@56?Cv}$ zORlkeOJ@4ZJAJN8xB&8Ub58k7YYw6DNooYs za*a03wmm8gy|!a((x`lwIsIk9<|H{BfO|A-2NUT#j6xiwEf; zbh~B5Lp<=9k`~sHe#Ogzr=|ICuDb;a#wq9cT#>h)*nV(g zD`KQ}u(`W)6(rJ|ht9AyEzIa=_8PSpOe;>`k*H5kV2Nl16|$s>`|CUMoppVC=e;{B zB{K1Kc+Dny{#&XCb_BXtwRqHD#6l%Ie~NHlsx_dZjjmNt(Nx~&VIc>%;>~O^Nf#dNsTfpj zXXb1%rlIvn3kVqqJWczE#5HARH%%y#0hvkU8e&D&FkU*~)!cwi*G;;x?NpzOlu<`S z1t3NT9|Fqu*re6i_#JfRT6EiQV#s%wzEWJ#q;bAEIO!`5>nRY)())|m;tRma_&n9Y zseRvAdK4$n2hhld_PE%pGcB05i@VkywW5!ved5jt$ZWJ!xhL9JRtwWGIL4l2<;HW$#K^J{1|-Y{L7w%NSI+-0uE6-s71&W0 z7j7;VUyU#5FJ}Fx>-cx91?j|z#E$EMy4JWJ;_V@P+R2Pur{SsNg*%~cAlCpE_UX)8 z{9cwnXqyy0)Gh$7jZ^a8i>-qgDV^p%cSbqo!~h4!pMbelziqbkDcOnMsAQmSt1b^#Cfg9& zy7)iO4puo#6^F1mv#%lQ_-T-7+oDb4tapUND4u&CYI7JK2XCP}(tN+3?MX(3z`dTN z&HH4vz=;L?OC%3g&_|n!J^tNZei+X87lyLnFkNFW^vScMEDb+8!-X52L2CZRJmPqh?JNNwhjxTbw~$zCa3skO*unDWN}S&Bj75%}$c%$`3xG zkeQGp*KIh=fQ>s1mrJp_A|G>+2-9PVg%K3`Z;w)XegGb)o1@Km?;`L?E8VI>3s}>{6&9ZH|vPj({ zxEo*Z8y-6YoEC%!vocfAmFVXc_xu_l9F67S!CpS(_OYlu-l(a%>8j&@vo|IbJ~AG^AJ-h`wGeclP0oo4aGJ8^z9+%Qi}= zF|)ig@iw>0-tlCk$@~Xyv4W5mZpiQf!6YFEA;?K)WD7x47Ymh25qy&^urY33rcbt& z4o?V!;oi9n3Y0lJ^Vym2db_LR->0aDq>0r7Debyq*kg2~`y{~SOI;GIQQpqW|4z4)7Uw66p>-5O%{0??IxqGSM1^c;LG)5%4#Y2qS7!LzcOCX(L0Y6V9~I_@e+ zu2B4ZgNB}BBoul()~qXbEWsbXr6;+lMpxESMk>}|B&2)pHC{L2fKrqs)c-{Y0nm~*UwV^`u2+HR&Lw?|55skTvY}2&f$;ZNIaz> z@k^FVf&tojf`6$55aBktAmjRIyB5oRIg3ta>6c&QJAO?rz9S1(3NjO!=(;_^k==a= z0+h*e;ma6H!dY3jP3wND$#r_p?=%djrIgVgIg0jMnlj8AE@twMl}lJWcViPOfc;9% zp?nY@wyF)auf(C4=Gk6#9iUgrQssI1z5$IeD3&)YKksUzf-%CSBNdB^jO9BUn+eG7 zgF(J%o`PW}t`^>Ff2l4UJ`b*7L`6KW50=K7W}_p}MA!M8XLRc}r?2~IKFv;UMur&a z={`6h16UlSyXA>w>2BbzJxo-0H9jDC@H)_#wKC>k01~%I^#|Fpw~EHE$2I-F%$3Q% z(Ut9WE7R|fLQ1Rq!)t}>$@Lyq5|;)B)to+Tu%C6LSaXYEq4VYp#=#SOTTolJ@$n=D z?zGIe4V|HD+Dx||?o|^HP$N{WGprwV220(!orh2g8er#6u)parA_DoH6b8Pu7YB+a zuO#I5qO3W;w1A_X*UTWM=nR|f6GgQGnK#2tQEGN^dZsHYFZT)16uYLbyR7fQ&ugqW z#UPPMbNwTcBOPzr>#We(+#po>jEoN6Kx7BCqedu3aWW6bOvXwj``5 zO6kBKPTrytiOV1ZZAIc_6*sNpQi`+m+&a*kr}cB*g+~(!9By{ZRgy5=qckOI8ifjmWa`6!d}RYlIqwjjQj8p% zA!z75d=8$ZB}wG=o7nk+m>13vS}q4y=OCCKJH~}Bh%J`|yASk-VFBJ%}ehcp!PzZu))f$hGU;;9rNh#U8c`vbL zm-N`cfz1%8h)gCCw(`yZBn^H<^K3JWyIXslFZ2dd=OFRRqxroO^Sfj}-;>aUcXQv~<|DuzA z9|p8qTuWMqySO-ol?KP^E5d)jVwDDn_>*L}GWN#ZY>eEgvq{lbp!8HH#emeDKgf-n zojUqb`ai4L@L3liEiIexlJnnj0I5+g6aA#R_&H!miXQid6-pKyF0F5}COh$1J8SAY zx2o3rv&py>(uG2g%w2+=(3tl=K2&o-Z2^h(E?$5nA4lXa7bCZ(=kSjo2C22rnh958 z)=Gpva<4T4GmU}8<#17A<|Pd%;>~PZe`2veA1A24q@%aATFPxb{;8jdE9t_mlehuN z#ttXAGt0(|S4l%Cl<-1O1rYPN$7Ww*S2Z06rz{G6;IzT&9t-?n3y#7Q0Xkoo1RWAp zTFkT{3(6={?Rs|tcZ6jswVBE?XB0~{luhZ8Cy#@c1d2D%nSJhCMh&iTxJqfX3SF3Haf#621t#z+pF6-Fxp^6+2CxOCl$9$06iUWcKaV`KQdsZm4VDs=mj*Q>-tw|O+TOHn%HW;^z)kk1JOE2Hc z*(Rs>5P0yimaF^v3H`ipH!4Co0hRM~l%_<|XP{LtR3i)`hZ#K>sa214g5VwkDvWUo z2Jjak;0tx?WDoDRDWk@P2CmbDF?%=kPKLX$g(tpq z5USB{(-RnmI03b<#(&*#Q0t$$=Dx#1*g}q6);$NeLZpon`B+Yvh9(qHX}T@OB4FJ> zQKAe$dP~b`Mi{h@QEB>q;>_NI=745boT8RfJD5P}%RSMOaH@=#x!40@bPhQ)X@tiN zNxlN6xTJI$=(h@ND*7W5L%>z*$e~DX1hKV_fs*PZw zB0txzm0+K$hwS7$Fg%oEdFBF4+NW_nIYnTRh$7cYW^2{1evsN{2T=T!tezV ze7Rc@#`?kZ-gU=`6?OP*D5!-;7`WLkHcrGXoh&m82x?0iquuG-%72(A{hrxNq| zUY7xrCjS50dqB?|>QZUtDy$hlcojBtez7-k!ww$?u<(^pMG!!o+@Z|&NK8$Ms}H-K zscW`<&Sf2Y;1fWEn?&;D7ia5uU+Z(ER}63z(|L7vk18UC_NDtyY<~zmKuzvyR>UJo z7=>{k?lwtcC|0m~7p>J|W#ZqQ_D73Ivg!d-P0c(Qpz;Cg=z6=AK=^LC-j zn|=RB*O6w@!o`09T8P1AQsRlyQ5IwhdVZmBM&lnW##oast7kL_U3V2R9T2xCVpEt1Onf8%`r}!ymyx(bH z@MmXj+$ThwG4+lwxlJuK^%^gi^v4-ypT$rsoEv?W$)%vZ$CMcvr=(bV#Z^hXYS3V@ zXR2OYUdiYk74d*MmulGx6I;SJf3huTVhN^VvN3{#lAfOPD6Kc%p(#F9!sG>!Te~{k zr~9nCk=IpudkJFRplqbTnnbn!VH;c+Dn->?drMoCAy6S1er~4HHytf;5xixeekywl z3g5G1b{uF)NrxXGx~}L}V$>ZLsCa!;Ke40sRS^&t2zHY3QSFBOTST$1AVzO2@ulfx zy>#LczC~sGM%~u*IQ#G-3=l6g3LM;uxZKesAG^lFzo0O?;Yo15uqWI;##d3##CnzW z(I=agw7GE)DaJ!*n&sDB7%JjG_O{o_vu(}XX|U8&C#on68v}ABfFxCpFL(r@Eh0|- zRJmByj3THN0%1pAPbf}&)Rhe2?+b4L_sB@PE3_9-gOC*{jG~eAFpno{3RG_NM;g#7 z>vr^?BSu;CfGp8-AZp?tNbDYvR~JT`-Ywt^fd9laXZK)zas}R%mvM(B#F$~tlS5}z zhJza~_J0%a_M%GS0wX2{@gFj0X)f{k>35cx+$2@;~MK_#An2Z4kQx>=Mi8c7n5 z6jipL#@hN9xH!vKVwB5MD8F{Xd?$E@9$@<-6Lt%j9-#_W@szr${17<$=$XY6KcX_# z4P75wOHIFOBjgV#Yr=3kg#a+9|NjsQM6~{=17PZqUkNJi*XZ+QP}l z8oCg36>1xZ^g=o|kY`}w8+O_dj?x|U>M_k|KVmC;5j)(TLTWqlujH!w#nq&zvo?KF z+fR;_V8wg{-`e&6v{`z_6gTN3V>e}{(k`rH|{yO zbMW!ajnyQI!KW|K5r<=2abW@pW>}RkHQ-oD46>LPU$Ew9mpDN@gVPp!@VtwAGgvju zR_`!^bhWctr3H|f5vIy4>T1i4(}Aae#3O!C)E3q~rIYgNJY$a5I9MLwISa~G>VMoj z&=moq>>~3}w+6$VQ%N#{;muBy$^9OJW92)>cBgLyazp~e2+H*D@ur8c%3_#bBr1>xP?qjNd5-oqc>(oxN`0o>LZ!0|k^m9+mbKa3TH}wO=JWe~G(cbR7IF~I{0(FeWc!DyL zJ)$q?n5*p3dc`zIk?nPl-EtC;C}JQ+C7g!RS%DF~3!jhoxZ#u=^_gE=wMo3+(TGPA zk30pfXE%-OdmFd>nY^eXl-F9%E}lDe{8I|`paTh7fF;(IjLZMjV}Uh5>4P&i8_`@% z9<26f?U{lGIKW#lr5x?yz>RIC@1-N-ls`NU&r#tI9(L2iJ`NxXVJ)dDN&)a!J<5*s zH^o_3Z5sUH6LX+ivL9|7{6g|1NWQb&9>)D8a8OepaP6csT0}-7q&DzTj&HvQuI!P* zQtn1iVJGHPq)nANj9S{4eQiLmIE*yMeWM20HV6sWUtES7Ykt+-zC z$z=zC1^ch9r#?dCcR*6+`f5Qo{AT9J5WZY-PQ8bmLUNdUZX?lnnoP};45279MkD~V zV&qw@ZKg;YqE$->HbQoNUP)~e0?(f%|E`^6^OXys(IX{!DbwbRHeSk}euE|bFsZc{ z6^rzGXw~6EB720#o2ZJoF$U_=8)wkBa^lFwUsB3ZMaKmm;`Qu#i*CV}%VMUhW*Qe8 zmt*4B;C^_|J%6~h_PMI^t9mPdN?>h$1_-BHbwTgeU0j|+(-kE~yrJAx`?e2kIc~Qz9St5%7&*2S>bWK%XWUgShzQIPeO`&Vs+!D9&>Jz1 zT{v=YhBI2960>R+bt($-H~@gP{UL$I0%;oDv+ih8^{7Sv>)g1=zEa2bKUWo15&fVC zO5Zoc#UYX=>oH3x z?pkUXJs5RPk77JlbX4I_c8Ex#p}`%sDQ21Vpbe^WId0`D`c@0HSzZoJYh3D$*Iu`Y zic?IvQxZ=+@_4o2(fV$D;q~JS{l0~(kIAe%S6<6d)ti4%W~tl&0_|izf@m{bKr2up zPoI$vDjf#aK+;WQr?hu?TM2&*fKQwOs=z~xSvE9?mUmRGK-SdjZwQzOn#dWoLZmV= zsu%9?v}$yWAaB*b{pn=QZUJN_Ijr`!F6Y^qjs)8RVQNjrjFr4skmEp(^yV1X;HcWY z6*tQDln*8=)P7AkNfc`g88SnZBD2VP%jkAqO1e^fejaE@boYmIW@mti;%%wq8>No&4sobp;;LK&vqE zY$YdaMlHQFvRzK!vS<~c&TJYJir`IVp~hWp?k~uf00XVd%-7f{6SCp>HFiT`l*CpC z&xPr^qK}yew>oSEWbb;Aj(ZxcO*0T((Po#JhfY)gZ#*tx^7wqZ6pYA4}tDrF`E9l7$6I5IM6LRZM=&mE?!i40&~Kr?!u zkLh+)j{{fKCv*rrPAh|$mm6JtFs^0W24C8--(fZ{b9^x>JgDT*is({lN7eYy65A_q z!l%O2Q>ldG6P|v$w=vLJG#z1}LW{GL?7H~{&tHoM3P%ZK=;4GD9^?BIGTdbfP=@S9 z`YcZsA2Ui%^Fb?rk~jZjfej$#uvz8bQMs9s>tu=HK#%cN9Emmat@+a}pWVdPm$*~- zUQLL^U^kI*P*8tlM|eH3yf}yci!j{LUKYeXU^eb|RApDa<-DF9`btK~g@HR~y5(=N zK%jf?1ztNq3kUzOP(Y@)ffK(QEfQx+3tB4zxkfd_0>s0mB;bE{@#1FloQOPPNErxYrJRa|CF&N*Ldm)?G4rR`o_ z-5Bd@kZiWxy>(Rg^$UiKc&ZoLcV)Y16{%1fcXhZwP+hhoVA46YX8F^!qWNF8_ImyU zgw15x2U&jx#2tIU)IxMx(}yaaXuaW>%f%w_+K!3QH=)BXf=MT9MAxwjp-E#5sc}->Qxv2d&B{s4Qg|SrnJWW@#LMT6Fg!emS&ky*L{ss z(ZdfY3ZJh!ZV>uyp$?3ML+)+*jF-@iSPP#hGy_?TG9l zi{g4Ts2S4Qp}T-=zVmka15JRB2KG!!*9LHvFAR)?Vu5J>4i_yBn!6h0<6mxzfK%1+ zr8jd$4DC8tv_y?|_O#WyxBhoeQ6lt!Ls&eTcH9R@qF$q`GJqitcr(L>UuiurWX8r% z6=lr4mfm=~5RP@QB={5_^1HUC$Q?bK7?p}JfMwgVLxCWu+%7?rwD0{Uz;3o>fD9Bq z6lM|dtoufM)+yc`P~F{L2L^X8U2algCj+UsIOs29D;KG4=ZdZhzV9I?(}}Q9WIyP` zILnkqN7`Srq5B1qNej4jAfeo4Z!#I4HqyB>u<@8XRqM4FH`Xqx1UvZY0H~kxgxM9m zs!eK6U`jH~=pPNZP>4&islWKfJEr0(JrMXyy^URuxPtj3TRFLhWrrsdfuO%#U5LoN z&C(PLH))D3obT#>GgqLs0a?~PVGX7p(b%9>&zeu%+k~P81z+)I%b}{)+78G4unp_I zQBlu%rKXOGnK62!8tP|wY0XP>cS`wzrmA|~xjJx0hSyPMqD@6wTx7H%@E1pBnh zIqf;m-D1Vw+vT0;idltgrbf~QA^@b(O=NfhRu-?fC@5Pr)CpkI3Fw#wa}WQPw@}m3 z_NgQ{=w2C95cokYB=EBO23a;EtD-2>qzBxXG`MAXIfvvhgkyUpWA7i6YbwI3Wv=jGwPELJF+1>HUT(5M zIVZ`?apcdaE`)W0oCV>Updzpvdk~UjnG4b~gKS%?%j)Xui5OVlhdz(O`FHmndZ(vf zeR>uRVmHsXPC2XxgaqUdK?=Dhfr&tiJ34k3Q4N*8hbL>~APmS9F?|E^*J{yTF@>|! zb6ss^{f2nVZM+V3+@(#i`p=A8H3t}EjIy7n-wi+`avgF_L9K#qqrTuSPqnjm<=Nb3 zKIrszsjz*NYtmp7;Axur7)sP(6% zYH7PLWs<#2y}A5VO~F*tR|%;hs}#mEZ6B33)Su#^5O6v?EOD!gAun~J@PGu6shoW0 zLG4&Vyiw)i2Kjt5L|Xk*C!WeH!fqmutr)x^0meAAveYvw*}`NNR%k%&{~|NG^fmhD zWXD`o-0G)0%yw~K&HPj%vfaj2G0U3f=xLBZhvV|m`}PauxR z@tCxlsUsW2mu(-ww#Vg;NwHbx!Q0D6b#NZg*O1LjLEvfhiUVT_W_+o&Xi$d9rU zWg66aCl1EwRr^4kK@bK?w%p0xPUVuvmb?_e^fEPl`IuMS;yd(k`O10au1}d6z+-sr zXa81jI&6-(%Kc(t%RvZ4`YeJEq(0Bn!C3HucgGri{l4R7`4zVh#W_rnMjWQ+r@MCt zfTo1y4_bw*ZKL(aatYLsj^gJ00&;gXK4x&z9sY3WQO)9<9xyzF`B)DasuieCQ0xoq z1cP^tyu4-{>A-P&C%D4w^UKz^m&nmwx<6%0|JlANs^h2xpK?%zB!N+3GGae5l4tNI z6^H*ORmGXIzZT3ip=~pC@Es7;bJ*NO!dIvX0uTlgV+&rj^|jEh?kBp!v9``#n@+XU zqxdu#OV>JJyoOaArBc4%1;7i@@QN?d&uo7T(!k8TLc#lYYrF<>|82Vb57YfOCIWK@ zLh)$lm$ncF=FkfnL>OsvPLd@dq~YG^_e})A?1ysBnt6q`*8~I8xw2U^RjFgj^td_o zVHNj1YpoqcR0amqEK3kak~sJtZ)p?4LLfu&2oln*E5gOb<2<O1R5dWbS8}Tu z6UGC9t-TM-04b|>yD$C6oCe}#Anm6_z> zJ`y+Lu=px>vFe3bRe!kKtA1^uVt<|s%;ppV(CCrT>rReTy5^di>J~gPSUKnx!nY&M zP9xVt{M0!_9?f%Z9LOZn3qZEoKDE~>^Ca#(kC*Rg`)G8N|BMI`M!;(Rk&<7@0AFf+ zF;%OE%~*&QcNtv6Qh1}?u7ItWDQpAcR|GL7+sZeVXA78l;Q&e=2`ZNInX3${=Q~C{ z>9ayBO2n|Im_s$GFkBrUp9fZv)c_FQJfkt1B)ApIh1CPJn=E@Nh%F0{I-cO|okp~- z6XxM|)lw!HVq{&#-_n}9)U;-qYe&U#24+xJnB*gCFj0foIx%837J?0X4x_=%_8DD~ z?SLoD@7;>Sc)#>|4|i0e{-E4Hp#NuhCI>cdVSPSGJB#TUfA2%N!>D~IiD?B`OPWhF zb`@Km8y+0aeFfItEwXe+c4^7P_^u%<5_~ni;L_{_?S{|$kE=ugy(H$a$)e{K2mrd* zNsc9{s4hxo#v1h0 zyC6MerrliZId`lLV2K8Ril3?@x_^o7alSy&wiM*OTAS+R0)aZ(IU ziEW4k-3?I|BS-#RAfN+`%=LY57~wP+R0CCy5#NFiaEr+OLut+HrrCNz2;RIvj_v4~ zkrs$!`!ZFJ>!~s?*@YYpA4teqYoi~A-jkYCX|fF@ynE}v$BK9xhC5bT*47g|6X&?c zHb#}lo~5`bZxYqaw6N zX#5~eP|}6sr{t2nj%G#SMwN}_dC-F=;Xt`0*tJ}D-+|mY?Z0ZtherC$BPIuL;CdaX z59DwfVnWh)U%6JY`9oFyA--U33{%)!Fwfxd0&?JXcfECGJ$s`x7_43O9lfq)>Ea6X z(~FLM4VY8WuS$$?rEg>bXWdr)XZ>6C5H;q=+k5&{Ku%^CUg@lzx|d?|`KN&%(1Hr8 zL0P3fm!8@XTC|Z>HQgTUuany2_Oqt>v7V*EoU}Ot9Sfo&lFq7dFL7qFM*vkJNYm8T z2T~dD&Rw%}%pk;I%y#wPxy!|}a0(98xJpKjy7QQ;u>r1OF+ z@&%a1$M5^tX7ICcTgmXG>^+mc&&UflR3i zJ*VwGW8B!$vSy`5H6nHt(K(&i>11Wlugyb=K@b~ z%!7fS#$9t@`gF?b3lp+MDIWW{tSGNIPoUSnsIOtQ?2&JybVPjyUK@-lA7c5pn*G}) z|Cp6J&cJmt8wAKWe|X-Jj{)T1zKvLcFaoI5(KC`Iz~pgkMRc{)q-uP^Z8R8kyVT=6 zK?@T=B{8f3+O^ad!>CFD0t=_gmR5)7ENNG%b-*m$6_P|fA7#gE?9n?_!(-Hhs|v6I z`QA`artbXK5b)a#y;2OqW3%=7uWbKNnc56U4V~bw>>V3^m_J5U*6jV)c13k6;T7=# zGPuJ+QXEzEW6jiNU0WTUD7NKpB_=0sVY4oFxVcdl_p*V%YAet}-k6NsD@gTI(WOxb z>Y3JQ@Fgeq7=8Plpd0Lf*gWX%D3to^!0b~XQ?bO!k&YRH#sKI|@@UjTvSri*IGrL> z*anLR<6Al}hG(RAJ@#989SLq%;~fj6bK|wwInZ6X#yT_ZRv?*?rH=Wrn1zH~(s6k( zffsVz8~|n&_h}fec%cfkc@vL_>FJS%?_8P*fq#&?BikwEbGGr(qiKywAp9l#AB8_9 zl%?LyT%?lie&W8?BhBL*8lX^BJaV2&Z1%qWhH}mp#(8M5=Ku_s90swtfqCkdk3XO_ zo?sq>s<-zn*IsqM#IJ2h>ewEPF1YDsey}u36wHH3Oizwht<=d#sSC&1w$2K@k15ie zlawv9G;p~>FxXdDcv+a37wq-`zQJw&6OEwM8sVgFasXQBJhM6z2~1L=sMnQ#s7Iy#~ zMsF#*%3SDj9orQE3PLRGZ`62-r$XSu)jQ4}Aoh<9m_3&Q-5bphOo`JWa0W>e<0f!{ zr?maPrH$VCGq=0d$z(6F9F92VyZ>$qCLE9kXRcLC>oWw1FNo>f|^Jf^|u z>N82Ait^MGK`7JaQ>XmEB{LDjul5Yfcbsb3i!#JQT@+zQN?oR~pr@ocRi-!)OfZLh zz&K6tuQZ!qTA6+GsC8eRqwAsKnVp&2{;ZQ8RR{eW&I9eI6HXuu@9;oDKl8m5n8#0j z3edd{t9LWUXJx|zwMr^f@|N5}*3f3B!6NkISva*SisD@>Aq}{TK9N%>-F`tSlH0C2 zVROMK9`tLXPH??ZvoEqFydSgDPPpejT3X6SVccWUBAsyt+(0v3@Id*`w#Ndl^RTJd zJJj0)ZU=C9%ITqK?oRoM-RU0}qAdpJZqWS%sP{FvdB54pfHQ@*+>5DfYd+#c)3g#C zW`|GMoN9gi);#Oll?kj#(NAd&vG<&6#vKpfr>L9rsnean6|a)wzQY`|!@w zUY_?|JJRwBlMTXZhL5WC{kZ^qExDps{QW5okgBILaRH27wr^DcEeJmuE|`jwDgi0O z*U9@HS$6PUH`#o9%uerGOCk-;QJH+BrSAEPyFQI3GDl&U)Tqh|xMoJ587VL<5WB1B zP9;WB(QOb=-s6eXarFUOoZ1yUmwhP%vgOFNKTCQ$pg^G@rZrClaEXQNCULnmwPzx) zNRqZs2`%*+Za{8!_XidCFM7By4cDkKx~Cik5^nR{uAACrOP|hYPpvKUN@0b*G?Tpq z4em9-*#upZeVHog+3AFhlA;V2)pulsMqgbB4sa4h)^muLdcSBRJT;NWSPiVitPKo4hbb8LM2OO9octd zmu%Vhecy+%jcu6uJ>!1A_tN+Gz2AR&jaM`CdFFY}@;>i#o^zf~Unc2G#&c3NQL#$` zW6{>ydrRbF-v&|Nb84i^%;0juDU3J{FpdjVv+`AepA4d$gBt9*bKea=%bsLy0tOx# z^iR#p{<1UYubp{&i^@^qbkO*TRY`hK>U3-Wy}DzP-^w!?ot>sxYbr@}Px`}C<4d!y zC1AN4IMRastmqfhfT5fo0^8p0-o2~vU=_Pj>w7=mQWU%X6aX|I#WnTVmH+AiS z*ZA5D=!wR3bv1OF&Z6Ij2f?DiF%fmIxQMGoh-BJ9;$y*rr`RKp>NW{LwAt&u`ZMuy z3tX(Y8l?p$#rOh>MJcgyjAc_DS7NsEGILn3inRD#>VG=yWl|z3=GGCXUb5s$g?*>q+UTr_B&b*61_$U_}dt@7>nzCHT3F3(XkA1L;( zU#bJ-*EAS=9dy+b zTiIBtQ#oPzJ)Bn^1C@K;YXh=29)A{I{I-HJzl+wbsK+tk3)s^Ckheqs56b8y*Rjg<2*?TL3V$Ec}~jy``IS;n4{+(2}JaR zx~RSdr}6V7!+pZYAk}K6_PYt3#lQ9PKyRVzWq0F_o6k5ch8}s`XXbM|*CPx{0Log15yZmyIh&&$Ys3qKxsI@WkNr@h~@ zggq}mZ@n1V?kFFP3qgR`AIo&Gl5&!w&-wenN<7!-TYe7O{51MYyCo!8v)DZJI!5B8 zOaDR`hd15c6>S1}o@=VrvKhBlnUGXsZhO#|AyaRw#UEdysD_3QVH9*5$)8uGDOqTJ z3NX_OtV*TKulo2bvD++;-$_h>A;MtcVR&+znbt=ZC1iwqg{*N`!<|r{@HX^6g|Xul zbOypm25&*+RP*lSk~Yn`&i)`mkCGxgVSRWzvND~RTUJDfpYAIqY&9;yc9&x49Tx|l zI7Jch#I4qr1&2!HlAp)$1=YNzYFSCAce{iwoz2N5Cc=si^*80*rS)tWQ7-gbw3ID$ zrP8*Rmm9k}a z`~!tIXsT`&mlI2jJJTG?w(6mdh+&d(7-1pLD@t8*p>fzb&0iu5 zZrRG^Z@6w%&CwqnRU+~e=OsJBs@;A?S{9#Hj1m`sZ4MLOm2nU%VB4)*B%ghTJ7?zH?5Rf34*5gUQjsfn!-C z=F{<00zGQ!J)Pio;6%}L@x2xS|FRDmb7anmAg4Y*KN;ju=?Zh?_boPY%>3Zqx|a?z z$|=9lW1l9$QDw2CjhwahdPS;WpuF1RulE^v@1~$nFA(`01D`$+qiT_6QozO|AWt6czye@=wlIwe3Nb(ru;D*i$ACsE=f?Z&$;s=y$w^Sqan3i=cdF=^Jr^ zJDR5_465093sJV>TE^nhP`2)#eY?84G_s zXetkZhd8#u=jL8UPgoAx)h`~Mf*td&_{QUwfa^!%9e3B4=6r1N?#84m2nXZNf15W~ zhNM)o1wHKaz$7=JSE*TNRA-mA%;8*zFupQk$eG6t*ML=IPnc7Or)kHy8xM0JcesgL zJ2!%;%;NbyhSQ0{tLWZc{qG~f>_OAbAI*!CJh1EdQ9}<4m;B`RL?vIHUH;wr-g8qN zDkGTuE;1vWz%NIaf_JAM5`1))8@FN{hn-!WBc!B;dIzJhj%&3!E`2+UB4dcRq-w6x zSmQ5O(>@+(nrHMLWbOfvogWA~E%b$YV23z-g34@RT7(SL*g(?Z*{dYBPf}esCBmn> z*~s4UtQIk|EiRkdm&dj{iG*&V!)6L{94pDFsmZu& zid-eqAXhDkgTOBTnpGJcbR44LR#Fc;%J1%SW9Kz-)bEyNe&@e;{$P+p3X=IVa-i#D z^be8MF`I@kA?FXL*-Do?`Mb+il7$`zc%ivpnc*hSbWUJ%c%>;dI zyuA@8E~HSRPu6l-sP_CQ?3(7MietS)NQSL4U7GivO_^?vI$k#INXQHBqF)6ThR$yk zjYVERZZKHl$nQ2eN-I>h`5=fWubIB{1%{8Pbmef9?44Z@)uAVLSrQvZ$4r@~l!~0^ z9X5H8I7^=m*p8FUvtOsqUF6Y-ezT13esCx)ik`{Q#^u!F36vaCCJ{{Qg&G<22EH}gj}7TZeG>GF?dryJKnNOofd5N4uI2Ph!fp0R z&23AGp^Wo+lpA+P`(YjuJ1a`)f&ydPKuwI##s~wbs1}d=?guMsM5__%uYmSbf%bQ2UAp(ZoLG=AC?-oQjIH z3z=5jzB2;sYSMXK4@P7OL$J>SI|8(G4cJl$p@(C4;}(^6{6n}&lp?p=xTy{uu@|}G zSH-bI7-saS2ETG;Yj=k~)!uz}FrQm#Ysk`FYUdrtjw918#4L59V)QZMXc-Y-m8{YZ zDvndSZx_9e-yh>2Ao7bJlUziHLx?BeE2rJ)j6M?xpH z(v6|5Rl{j~f<%93{~)^kn&$Sv>N=XB$z)RVox?KEy|ioEQ20YhSG%;d6R4Hz$(*fR zX&<@CKGvkR=y2)_u!HAq+16yi<_< zBvsGsT4SgmXpCf9+&f$a2)Am>HJr~IZnjQaHBAO?t~=Vw#U}4A5=-puGrFe;&g~UPE$7iof8RDmsvQ^ zWBrkO{14LL8>cLpyFSh}dGV7&su!Ie-Y=OBf?cJ|m)g=80Har;7_QN(Y}@YIIBsjV zluSDdKKT^-WvD`tEzNZMIw)3*k%KUpT}M=W&$MVQh)OM>F5ak+g=WJg^3zM#Ug!&^ z_!adDH-Gb^m&{E;NaXniId2-BA9()%%$}6__Y|H!u8h1KRoB2Py)~T$(#dG8yc=f1 zVh_tcO>#@6SP#2h=m~p6o_{9D$jsndjlb$lThJ?CRF}k}ONJZ=);KeEN?M$Z@W*>_ed zY?-$EWXT`jKOacsHfL(t46;BpX~2HG^666i&9f^-4i-1_f()v&V%GV)8kvP__O|kI zii~2(Ft$HBJf9soke~DY1%e*BQq9!W#~37g$O}is4DAt)eSyGDZRP61-D6@>jwcP@ z2h0w1CS=JU@l{(^L7^njXpq}#aa^ttNj_P;mjKgVTw zty7sjxHYK()tR(;i_#Wn$nDaFgX)*-7xK=;B&*A=L9llA8P!iSXU{e6UjNy#7F*jC z`ZM*2YLR%jMY9mG_FP|4Vce9KWsyzyh;A8Xs1R{<8*gCZqPaNW&e^BazcU+yRAHBP zy8zq3#h_sET(Ie>srX0AdMM$XBJ;H?)ng%7d4;E#O0f6)rKDydkRdSA&L~Sx=St>R zi}92|rAHGE`$v-X%#q!eP_bE3)m|Ld3YiSyk-H;nq4hXeN>=dZh-pet%RD(@tM1N4dAgF^myA4K)m;XFo5Xj>$<+J1|-Xkmwi;zFqSW}sGj0*GK=DMJ|6_?f%g`?s zUc3V9k@<@+$C>*}c>bq|UIE3^8d_NkKgr(Hv8CpQ@q{g1nKgX;>p^ecRXGi|-152P z>P#0klo4#6#nH}dU1R(HohcpXz%9Q}Wc|;XrQn^hVFJq*Arf0V4a0x>ENx$WB_odM zT`6t>TjyB15nkkvu9d(dFK`EPgpJG6OT|x5m-Smq+Ov0^#%L}WiF@X1j+o}wbXoF% z5#&(bPDk$;@s5P)w8$@nCm(d)KX6`1uhXR8IXDg{LzsMpD=+|)H`xRc-4tyeWW9J} zqw0R9t#hV|R{c}c&lm5Kh9rMiDo=OGN*#WWZr?5$is-FIX%$Fw;qy;mG-Iwd-_s^QA4(c3!T>#d4Q; zKh3IR^H{T{00&tU{K($g8512JW0b)w?rYV;TD;5%9l)iby5f(hhR#917LoO~SK2g7 z{2&mxV^X+rH>9MvCv6I67l(!dPEoyG?Y5g1TAZ0P-&T?mOyEJ!QdvcP!VA&8@!PW; zod2GbGk^MC;+=I4Y-nV}1DnaJt=KYre)z=Ctq{zO4sIzHRN;5lk95GIdMTWq;eQOu zhr+PYpdc{3zH}hkYSYX#w3*sZDcW=Lqu?WED(=OU=Fig+Oq*zC|L<8}P^Z-}?91%K z@wRf|v#)pGzu8P}Q*j)Kdx3~9x`xP^9q!UddFB5D`9w!QbI3)Ir7(_6jW{(-K6vnU4iL`*}T`xe;-@t zMNVz@AR|Q;Div0&ic)J4RJd(ZI1L(V&^kyh&fllx4Lj3W@zv|v?xsbvR zVLbVc*eWC6m&XACH^hCp8^5&YZ>+@`*y-O}ggKpJd!Td9%50-ak;9=K@2Ca$DYE;X za(bZ+^6Mp!j>WdK-l6^C(8BIzjux_+@R2TrX@^u7YIdge!{69oUGCF;)*AK^>=U}> za+Bn|9tX)N)xSqnU8ytV)zGqupW9rqN`MMcr_rEVNQ1>h?v`JdXQ`!qNxAB#^jyw% zRGJldN&2zIDbCQ%&QA6ZANt*J;wC20oxAgVUNaOl%{m1a#zigVEG^UBrd0SFGQauh zZd}xWVbX%3MSt>6L{N9UDQ>z?;;jKLf68Yjb725u;%x4wVIHQLcd5TmE)h+V%NSa&s4#ywcuISQLs2>dln*9@S+a+07$} zc|m*N(Yg)C;@T1d&d!EO`LTv&aU`J-@TR5)o7O?WaedCcO>gFoP3KXY$<=R8g0?4p zD)np~3*pXZ4~t2^9qn@nq>glx$^xX4+3Ro}d`}bkck*JP4u$MxmhLB$PS85c($5X* zQ=)0I4el$cd3>`^FQ-*|XG$Cr*h(k7DdUH2jq8LT!T-Dv(O6%aIksOZyg z+AHRac@@{5%*naF9G|j+iy^$`p0eiCDn98j=u%85y=4U{b_l>%Gbz?dV54XQahc*% zAKSSr$M1^TB`wV1G&fVzgCmaMYrZ;K*5{?9?QAZ!s#=+;@UkxsNEclh@0sxyH(7>; z%%AMfirEzJo5{46JRrHTXUX!w^j@gLENpr>P0i55$@P933WY$g8(xI>v;@)h30&OX zJD69;$idy}vDEw!ItQumQHu?M%52)60Es>DJ0cGYe)la9z%mm^RZ_AESQ~k@-yz{O-_m3~JeG1_r*R9?-2CUwV5(D~Lc zX0Nv=z<51^+4u%@cWcK?S*?nt_rrM2@?~EZ_n=4CjOk};jC0I?iPyGwbeO*CLXdeM zaUXrjF8%3PeoT%)QWx92xrxi&HSuYee&kM<<3v>uANk#1t2jT*r=p5SUjL`aB$gX&~P|$oVbnrgMU*p8Brq5H@v;RQH3V{gnAV(HSzT)p}IBKPW9&9sZQgQ zsFA1KFLHli-k*EsXCvqkXYI4`+O@%E@*9Q52ZePeBWHKg`B*cxbYHGFCiDoev z3D(`5B^sf^RPvqT3>N0*l9A%^?jCKZOFac!3d zTNtvq&A9c;l)oWx;hW9EwnD33%G1nKG~?|tE?1J4a{E5;X_sE2XO}i9GPN-gALv%A zAZ(*y$2M?3wI{8uh(jYHXD_Ev&a)>~e?hJz2|Uh|)$KNO6V|P8@|>%3IM}&qSm|YK z63P&_egz^gVLxW%o z)MQg<$mdPfak<*p6i>&1+ms2Qu#rlja)_Ul!j+~yNygMQQbM*cb<9fsU$_{A4!2*| z2ZnFOT`lvCZa4SmJGW_49KMt<-L$wkAV22&1$s?HI6c#$)lNbp^R$?qZu*R{OiAFT zk`WV zVnZCp5|kMhi(Ii@GuJds+rRL+7jLUAX=1za3(Wb8vO4Bij-74(29W#5W_wX&hMQ9+i|4Hau<; zRMjY5@q_e#z!_XuLHq_&Q1Zvj$aHsm%+;LFw9J+ zA|={n7?BCPIntoBtp(hPBT^PozKq=&P)gsvSCZy9xqRoEd4I^e(JyKoa6b5jD8<*> zOm2@{H@s~tTvAJQU>Gc6pnIa{-EAcovHbsM#P*-)3#>Jg;+M$AhWy~{oF*)l>{iun zZ-t<4xM#ma9?vn2g}%&^+0vP&2L2L;tG8zQ*Pd3Vqx&`tkJ3fm zOQVZ7JF8F0ADCWxlCuT&IovyuO0f7F_lHy!+{Ak4*MZ{V51ze3cfbk8x>8(KyJB?B zvLUeM9OCTaC;mbkOK5MWWzj}}Pj1n!`)u>fhXSLgTE7qpZk_#46%W z^8|8hm{17v)o9$JE^#+|dXk>F*<6V)CFI5g3nePO#GCyUfByR*{)(18Wu_@UsoaTz zX*<1a_SlBBBBz*68))%fny}G;%=PWcS9~FqzSzkpHX^UU(}se*Dc2wHMbhFC=C}RO zcAoc%&esJlIxY^vJIn9>0iSDwCnMyZG#7lUX&p_|A%>Olq)pdFp*H||Cvq(JZO+TV zug$M5h{aSMv#qYd&vJp_ z@wLOOZuc1-LFvo4Wf7L`AA9r{igYifB zYR*$d*4u5(3I8##=tZJwEBcb>T~?kDr|2hYJl3dfRyLJc&a#Bm%>wF-ekh!y6t8z} ztF*M(4acLSd00S5g4;5wW>!Q1KYb{`&_}nikF3C!l*ovDllD_Rmxhf>SJS&1 zW}orKAAvxQeH;ltajgSVm6iNmzgv+tgzY;&7Od^?XV7J`TN&IBd zjBtCGNFv`Dv$?J9YAXVVm6Y|^peMu`(_)$iDd(QPr~kyh5f_zCY~+CR-};JBS{XgS z@%|++f5y@0TZ*Z2!VOm*lb_@q5!*I-`LbmM?wkZBx(5*| z((At*ajniW8Etk+lH*s261P?*%p*5hvyj+mHGEe-b|90t7>iRXJ7Y0VTe31Kn{L)) zhZVwPsISx9lO+GkMgHzPL2rFhm3^~_JiAZ%)G;v&xoW;QWrC5mm&Bm^fvs7bjODqD zluEd|RGy1!g(=Kh5W;%cS1p2-_FV^qABxr$Z{A9Buu+2t8TolaSI)VYgk#PqHsrcC zR74aqg=-ZGDTjsPlTLTgJ=>NUF(yku6e!R|BU$+e3j^}=Y&my*cnivR zo?TbNJYt73-;?w8OzRQOb4_%cn`C-Hg=juG*Cv>t*(~lb(c=FQF4*zm;{J0vXvvpP zfHo8{w$=AUZN_yw0d>gxed>^h-Iw;9#a=Tbl044&?F4A5~AR5?N7GH=M ziU|#&a&Fz*m@)2P+7}3^(ej9TaNCEC}Q#FWimtD=;FiJIvVE^@_#Gr7e z;`1*BH4%F=Ha5}^(_J9jewKQx>3wGF+Bi464g~N;eE``~HLBG_ixkmscdVQW^Y2>Hshk(hEkRJZ7uS+S=cbBKcB#n5_k%8d@g7F%df zJj^d4^;AoQhS1jvzFm!y;R5FVdPBP{-Z>njd8y0pKhbmGd1L%C89apPU>3W3v6}WB zFI>Lap2e7Gil0H|58zE<_s?Z?lAE=Ml$wn$XSgMcscBqg%I~|DU+dJty`s+h zQj;~fh$b>}guz5EBgOdevp1gwd`X$S{kwMXPoN5*Yp&=x%kdLdXWb{8?`H9YwL#FY z94=0*2`TKiVn!!wA!awM1o_eQ9VmxcD*pWRQlAZ)dE8cgby z>iLJn_F?j;P_VrT+<|(I+Ocy*3Y7OX0F6O*&}*E$7@KaJiI-6fEE(YM9qNmC5aAyf zpriF3-JRSQmc(=<^v)r#9%}ve1Ptf^zBUdGXi(ExR8;?ive;UhBUGP=jPNj+9 zrUh^&w1;=`iLGZaE1|3yDn3c3!u8Ke87a$8^+?|L|Mzg_ezu_@`gY0a4tDrzosXlTfu2Z(P8O|j59ys_=X6m@ zKKRve44WdwkuHe$lz-ZQi|Xqda^{O<0;jKa>XdtF0XtiHgB%@gcT3H`&PQRdr`6m;`$zTd3AXD0%rZkjHtwwoJGz5`neqX;XC)6*&^1yG@ZJ;vi zk5X1cZ^v0Coj#iMUW%nGgIRU4FU$M6?w*YC00~H7>uEhEKVyREUJQdv2C{;%tOSVr z6wSoZO2uR_4g7m)t7c|{oI^%KOWiM$97KU9t6`*p}Gk4g-5-a#;d z$ucJvOi~1*_InhW9r8!eV;MxJbDJt0y8);ro2}g&9}Cc0Wv^rQp72DAGa_>$m}2OF$c(WrMt){3Yr&)6AZU^jH>y%-wf1T`X?Ro zl+mI9V_$K<1Qr9SbJQ^)`5!z#CniC;@2n4bF(0D|i1S7vBz|U97@d^`mr4sRN5DQa zb)REk@K&j1AhARD`+WhBtGkCxCHvMyDc>kz<0#nMY2{?h*<$Pkrp|zUi|B1LOQdIQ zC(xcyap&U@!MP^HU^bG#I+u?-`0&#nkg)$ipv=^O_gJobHWLMTocc6E^KQca9Xv$# znkhq-S(lhu%;p-e(}sExVNp#1PxH7Al>Ia#_6W&#Fi;*wti=asPmDCJjgM}hR9z@~ z5?Hmw!bO$|A<@BFNHId}Q}}Ce$@Jd*T>x>1Q*)%x-@vv%H9W4UB9i8 z^CuZO`l;Irvx^}GTaCH6=UG+E`M%DVm2~?L;etA6ME>MR@(aw30aoVCQvQvfayFw; z8h|2oxajkv^zsKWvST3MF8Zu~bnv6qfOwBRZpfyf8B`JE=#)Q9t{<{fba^CobDA72 z-P?0*ezd(NX!Rt)HGxIr+ru(s`+i0JZGj}c+-K@R4~xueCjH@mgv(3VsTLii%$}Z) zC3xjLvs70s?&Ej&PS$OKK#nn&6T&q6RB`>EJ~HrQAAGMIDfF`9i>7k?)5xDE z1Bm+s38yHhN5v)wL@H(QtlZN3?`aj1-Ao5|b2D1v?vS1rippV@Kf5C`{slr6j4oa} z_`#&(G=XuGJq^2A?pIj{M#KPt8X^6e39jbxWly5b%)PXJhH$x>fRtX_S5&DJ&cfN0 z{!{BY23w#=DX`Bz)2@O~K2vY>CXrP3Q9izNJF0HS_-;6EExiwV>fS%PB8FG_GWryk zhZyzNj^#Yz1SE+Dqe8Dbn*Jv8fGq>q>uP`SZjFmtL5k9&*}fW@>MfKO-sq-<4)0G7 z4|$Sc%{X{{2J5qe2gRx6ZY=$)ut~=b?LPO5-b;b;v2c$X=oQm$0JUUi=6%W2H2>y?bA57~Omd8Cz?Z}nh8>%TqGtQ;*{NgsQ zU$cZMe1wQy{M`ljWQ#Fwsh)UkiMI*f7UWi}>}Cf& z%x}`VeCk!&p=#Ts0cX7Ej*F#n1=%8?9R?{Th_m7hgXjO=cJK;uUN>2{ymZ_dvN#u^ zTD1P|w)N1v6ho$o^{}TAze0>j!ZgSo3tNZk+jpjEUV}Gh02rzU<#QOWdw%!Ht@4K) zA{*-y7t9mAmp4^0E|t#8SVeVkq1`hQQ=Z)Fa#$SK_owmrT`d1v#*Zf;L@jRS_=XMk zoOT{7(w6GJn3wtjF$eoC1}o>Lo><9a6GsGRR+Us7aUYJri^|y$7BoIa$*(h6!!!b2 zfxLh|^R)(7A(*-2l*%0yX|4v!mg68CobmKb<-61cnzl>Lml+x)UahuCJ|gSy7~#v3 z+!O!hq)ujvLDa2OgZ~wd@B+L%G#%9Hc(2@AOD$CnX&S%Y2^W6V}7 zgdr9Zj0Ks<^FYe}Lt6~1lO(&2B+4Px!vc6E+_7P$P1lg#GZ<8=-uJ+o)Zfnjw?C~J z0HNZS{exLltK6CmGHY`6RRS2|FgopdWWONDxKXsRsL(}Utv!rO9GXgi+AQxn7dm=B zIAzc1L3-np`!bZac$;vx9*y+A$%e5IaGN>5ILx>1F~iIml7`jd{8pExbe_v)1Awn7 zQzDQ{j{adAqVR3_J6B`kQ}`d=2Lhk{D2}veoY4TEj5!Z6+xtIAL4H<+SG;3 zPc!N7j6I#N>XjvZcMloa$DAv8$GCV=6>2(FROvt@s=hzzLW z)ZfBC&1Y%k>4~csB{L4Eh%9iM*IauNj#G^n{Y_2wZx`4&nPNayc#cW2+zpZelo`Og zSq%wK;b^1T1JvNj3&C|7pknwY2%Y@0qdgvmQ)8Rzen>&fUtZozw>-*=dQAM3wfG28}zKtR%%rf`)^4T4$WVk*kvMP$DD&3k01dySrzvKj02 zU>S@$9<7<`BIu!~%vUPP-RYm-wdf>CdWMW2g^N%E+t45*dwc<=J`DZR`Zj|IgHb4n zo&=Gz>uLvOTQ(qng0_A2{_sZ3-FA)fN7zS>B-UWt71%NV)n1YQL*NuptE$kS!4mwj z-UuEZ5IM~iOUmw-YA*}M43z{nF^%AH{tY=O0?+(a3|re7bQp5`k=Y>~bvlc_dY zxBCR-XKrerp zUavF=iHz8}6+>cFr)E<8PtjY1O^kCi)noMIO{W`w0QvS3qk5ch?zVd#hW$Qe6+P-0 zZ;{RXIc&GJWK4kbJQWbvt!VB08oB#HRAq`pFv(RqnL6QXwxruC*r3L(qaq&%9wKu%yAY@ zvrG0yY`=N+n?3QGX1tdYTwW0h%=)yCWUU$w7x3jBPNifnY zk9xu`!&E8AZFN_(dk=H*%5^SZG*~^S(O)rpEt4$4xL((?Or9z6$dP78gS6}Xh1biN zgi9{Vt9AZLMFcb|*<{F-?Sh23-~lTkBS}H18zyo_A2IBRp26lJ&9Ppn06Gayc`rER!`o3_r^B%B}ErVn%kW*rIC{Dc6MAhpi zNoYfjWpn-R@}qwf$dNoPkph!5(RV*QwXO@L&JV9v?mP|jtA+h=dyanGH8zK-ln1KS zlhuGZRPKHv^y8kWhb4{t=ZP~c&sVdeJsSTf2M^N0@-Hx4DnY(oc z#T%hj2rAFOb(zPO8s z#HG`h`rDnm*S;;G*CCh5X27J|%hXQXW$3zX)6#=d{a11U(q6B%VvO7(nA`d|cf zn7oTGcp-OUY4!z`lQ8s}dm+GGkoM4F;R6 z;`)!|2NdbAcwhSF0STOm^12tF z3BO|XT^C}<<3o7h&Ahd5sgDyXE%2{qNol;t5uQ!-YbjK3WY_;dIq(+R+*exwSEyuM z$*-z;8RuL(=`ghwEq_(s7jJgiOBImI4!h)(=m5Yl;@B6unFO7_{4T-=*ZV^qtk+EG z>=EpHPjChO^UttR^ZwKJ3N4Cf13eB#$$wDmtLcukP*e0GiW@xReqzb$+JI_F)+TurPW?{}{n(Ox7u7)dVPlLMQ%F<>enh{EmG7Xdo*&+qsi z@*KvKMDwv z7jju@A$a_k6RAD%AU!_s^BdYdC?Fm?^@SY2?it_IdFj2kD;X?gSNciu za<9O9Iyx+hvtmL^jU*30t;8;oAua+i?jwKGW>)JOzY^%n?VtU~9)%EYo{kCFwv+Dc zFWCIFy)=}5RTAlFjZAPUxh@QVnso*#68~E)e|z;qIo5#Ua-fRtuVXpza_dh#4!t0+ z*2g1ndEhhUb7aXmS16Sr6CH*}FNWs6pyY*|EwpjaPSd$(W#HRcSO6iZ-nGMsU#02Y z@4vaaR6aWmE)F}n-WV_hcXp$7WRORtw3t!$u`lExiRXpld#AA({n5scpFZxjQZH3G zseHH~PT2jhcok|lP%B0Pf}T6Mv;RKEZ#xh8bRKYMrPG1$9Ph=VULaci5qV$|37zt9 z2qWVbiS)1i_yQvlO^=QAP<^***;re`Iwx2BFQ0vaBqF8x>dMOWu^?@_5@n!)y1=iM zd4=24+acG};FF*x^YQ7KWT-sjYUrp(nW~uumbK^Bf%h@YB$d}T1r#rB$C`1Z zs6DqrMUyg;*~2mKwmzE31}RMMolXjMoY`#$+C+othvH72^IL+#*xE+Fu+Errz?&@n z0!2Lcz&)Vo+sf5K zcNi2or`@Yddr#-^Zr+5FcdxAcZ%Jvs5d0w0W=UVW$<5)R3-$I>*7FS);c0Hy`MWz} zYGJ+6$uRbpByr5}IE=WEz;9TWnR&M}4ctLuvgA!IvvrZ-90iD``(Dn|2vp_u!fRnx zVKZx;tPi0lF^>xQU|tnzolu7@V88iK{f-^>d+tMCkEsBWF=_m)CF)d?cv|U@VUJn= z`EH0w#=#KBEPzB}zY}t|8PPu{IWnMsr>RVe2c)U1`u9jzh8{7mq3>5})2>G$Mi9~2fjFKS_dJ(5@VHu=M*VO@1}N4T zwQ@;A^#}`q>s{1|_3oHqZZI!!#}-_-W-TsqfK~G69&Mii^EM4m7y5MpY?x~y?H#;>-#^R ziL#VPV<}`w${LEumOXnI``EH%*I?{hAxVgkvdcCYTZpkHg^_*V71@`p*?;#q@Av1# z`F=W|-#_PJ&e7>L_v^Z^>v=t|=XG78ns0?j$Mg z3odfGP~YTz#H7bc$6)6rMP04=H=JMP>lqsiJfQD^4!ZmFuc&~D3h`OK?m5cs7q|SSSisX`RMcEj21YRnSB13XftP?#hnORLl|BU_&()R$!UgY<%_o z=)}z;CDqE&cPlTSqJrTqCZ=MHZ?|nr(*WZxY4oA%r*$&{f1FEI zJOAm?%rRcX$k}iI<@-2KTk+4KUD?{dR>sWlfgDBV>XJH(x$`YMT%}QnIlS;Jh%gbu z%*FD-=qc=41v8Mr#S1c2G65k?=M7lC^_95XY!p%nhv=S-s9tWn1PXJ`mL*mYkn~F% z^+*)2_dZ!VkIL0MR%81OJ>wi&P&UB*9})yV5A*ox)}v)Mh5G(#pZojw`h@YYyr!+m zadBJoNoPVg@K0vF5&^iQBF@KJvq@p`$JtP3oLY_O!nAt#1+bgUDw3Pa4e^um?jU~E zh1ZBn0CqeB5R`UfdJ8T5CXo+Gg%x< zta`?xCY=0X5m1t)L*1x1of4jL$Ol^_H6AiJVg8TB-eOUuvXry6` zKUSWu@_O?mLoe}5B|O31bM`}u?lRUtF*b@VK_8hLO;%dHYsZ3qKB?bFU(PRv3o%|_R*bnRHlru!&zVo`e@$EF|Ev5j#E!EH(N_^7 z%vodN3B;6&S@{cRp^Lal39Ce})4MOsBVa7)jRB2Qb2)N)a}fghrv03>N9jwRI7fkYZGFy}#GFl%|Xu6^)$DxD`d6|FZFu*3mzo0X123wUbbq*#xgS^5+FxQXH)EsDO$98Kh*Ghu;39kv zM|kTSgq5MD8gikiXfl|)7c||$%W%oIHPsCb7-}RTZi274Nu&Jyh`yyk6O3(~;7uS) zAgPOpU6s080Ok#Mhh}wNwogx;t@CN`t%cdnuM@|O`R}_W&OigXq8afhA$Ia2h%P@T zDX?mW=QT)XkNF$FT;_20KE&FjsGvh38L5>WifIZ^-KAXIU`)VYX}2KWAg}3;s{Dq% z%IVWTy8W@Y=0DAGA(Kl1ZsSUV^N(FAJQetG2h3+yAsIC461+0%rKtRH z^w0Kj_>C{nxQ(yh?RQSnEUR_)D)XqkcG=rr_v#V_TVKxV$4BJi3_6gz-RQl3hmPk8 z32~PVi8uQ};3+23MqTWDXne&pVn4^-#G7qMWhHPwZY1A>8NUj|mIiNL{wk2h94BPeY|jkKNok@_hBcd9j3;7*Vk+ zC+mu>cqYwMaE%aQBXLoaDVT&sH!d7zui+z{aOV&#XJro1=X|J|wT-YyVRK<+Z(X*b z2iPDNk`8p(D;f_iOpFWHjfsW@q&l`f53QfxQiiY>qX3H9GT7sFZr3}0wSKdZLoXdE zzDlebC)1)$4(f=c$Zkn^`OYh1WY(N7a!k$%l)-<0{(J+!5aRQ+ZUrhh!5k8`K`sl5 zs@SF0#&wd$?6G_Ul>}T(D+gAiW?#v?S3*0`PazEJZhhG=Ei0OC)vJ5mO&DaGmfE}> zy11JELU%Gdpp=yIE>yn(%ePG{B`nG{VE9dQZdCAzywEvpT`X$IQ`=yVXejsuMCAgQx+FtVH@=rZyr2k{)D_4AIh9x} zP}Gc9r~gnd$ZmXcl-I-=z;E`z5D4L97w3t?laxt%0STSU9J$!eH3(*uTymQsKJu&6 zLE=FBBbh;hkO{yb2-cTwcSp;k_4kdGxhe(Ng{J=s^wS}}O}$MPj;|F;2cTWJd^GST!})*ovq?#^-XRC*uLnqX(KJidaf)kr=Quz>sd6pwOn}x z2?g7Hhe_sd4}0v{jj{nw8kn^R`e~afUW{TxO84Adc$FLd{CFVMgNvUo53T9~TTZ&G z2y?&O$wN7N?QX;Y@%qqgorNbbckvv86&(fTh1S_V=^J4H`o}DBU@M6GJAkUm77}gQ z_;9r0JL{I(qio*F+m$8Cp1L}07>G2EQ5;7;{`|WX zGzWD6yVi9J*r7P_TCsuF;XV5MPuuv{j6Dp)@iX&+uSiIzvWQP@+DA{+Eh%A{p*NBa z86{xHW;BJ?S6Dns(w?0BFuUgh7{+WD>>;>otA|9R8gA->%|=qf(4koE*Yn?K#OAtm zfk~6RXW{MZRx7}65+OzA*n__OQG%fd1Dm!rMfgIj)$vido)lmqz3Fj8q`(A!`@T7F zTn4cDFt1`Qz_^`}dzcMn)m-dIvQm4gv+#x<+#^Hy(3g*8LCV9GeOe+qQC3=&oLprH zZ_lhMERc}7%HVaXLNUa3?C?e?^@<`4(A&KGNgQG3>r=KMxVmZr z7;Sb19Ot?|8`4rZis9D0r~myuWi5EnJnMtbwtE8)SGR`wn`-%gKd-;v^S3mj*jj`{IMpE9KEC);Sm&x7sF4oUNqc~k4H6>m71IG zc6=+}$vAoFJfa{aW(_5HdK*ibFvYd*r#4`Gihf0L#Y*iEh$ku+@)gx#^B0Mh&aojm z6$5q}hgC0k4eIbHdhzggcX=rCvLkPC&u!5|F;?7!$B5+L&iXK>WYlad3=$#hSG#y- zm5FYL4t6;=Dc@QsIu1ID zFXcy_t^I*##Vh$kdW@)`82*sNlYhLQ@#n#|i|abuFVcKdAH!ej$MltyL#`d8ia~iJ zHjV`pHe)&yX~-&b2SKwG=rD?TzV!8(`cD%WvnH7AXucuz4;S(#fBlEFZHgv9`>)Ton?R;RUJi*&}&?XB{&n?eSX6{>iQmh}; z0yfgkWccG9H}NCRa1IJ~>=@il7=_`gEn7FW=G9m*|LX()kFd+W2xu-Q1rYFv zLR-e3ZgbP}qlpE-s<>hH15*o`7Ee3X4ccY`*$rVs`XSwjbEB_`yzRyuZeIXSplux@ za$R!6Rs|bct8qh1t8z;8O%_fqi%z^lA(aG&q#NHNzzDZSJ#zzN|9Y-OSoyCPml`T* zs~X10gxoZ&`~pJLv4D8w<2TO#DRlO5ssbACVUaC+Irf+X58}u%gblP;g3rQMbgJL(NID*@Wl#?j^aUhc3IK?HS!xa@b^ohP4tR5ra&3!i z+GKB(4MdQ3j<92!9}gy+Y$Jidxr|WDEe}WKZYw;7GX_x4cs}^DNudq2a+MMR`~Mob zfk}i2jJz`^<&I*P2TK6E%|tckKWCq*FxwG>h;t*aqcOb-3t|i+ zILAx`B}5??u>M7c*W3^rGFjeAY0F5!DHx^s?&}Lhw#XTlaDPzuK~`!Tyi#Ksug<2P zU3JHR2&_S+1wpQI4loRKS^ri4IN}U`!x6t#gGycAWw7579+DXzW7i$b3Oc-+PLIY%&pZXyD)~U%U-Zckr}+If60)|bTWiV&dZ#bL zCVJ}RcUixj!Szxqz~(M4TypcgVu+euoofupI(vIXQ`8^b2L)47lDCOSkTjaWrcpos zl}EsFOBJkj)8yJ=zH_&YGZi)?)oRhv)`sW|kA$N|Lxy~5d|p8XfK9o%y}ec4ce#wf-M%oY-qawV zU2~I@hO_GloGD?cNeZ+lqI6+5`=I*iU+)1TF4|1Qxj^XutgNeX$47+@uB~aXfe}rh zi$>yT2Nl`$c|5vlp?#}rTCV~Xj_eKl>iRO2=Wa_57=Urtvx<3;#o`Un+SZs8U(g8O zXinC@?XT7!`MfmR4BfXYsQ>E(Fc*dk;Id1D=j84Dv@7N6qb3ld!N}_jT|jY2ZBXP} z@&n)@3aa-GJ+VxGa3D@)xLh_yC3=Q&`5q(`ysfQo37{^d^6i`G(L`qh@+4IR@uNr* zP}HJv*tOI`keP6namlQs$n zoXrSzqV=I6C{+*xw1k|oNp#j_3$b4hGDbZwoJkjL+sr1~iO-!hajGKT{mU60vDDw8 z>bL(2Z%gmvidj}Hx98hSIEul8lldRTwMkff4(RPZX9iYoA7eRWh}o0=vQTapbo9jD zk|H+s4*t3F`B!`($Llk~ix-=21ye~*UxHykF>Y&myW>=te;p&47y#9vj8o60r8O__ znXeEpE9qrtFEFD1iSxN@)UWulDcXW|wE_h~q)5gUm1A!AXpkuh1yoJ0BNuh!)x$of zG#ksZUz1nE!t0KqQ_%Ae&oenT^B9MfW-g;)udZ`r5R<8HIG`7sm{qb9zV;Mc%PaCp zrC(k}?(lZRbW+_vpQYhdww2FOO3%U?^fHs8#dPp$5vOd`nndRkH{(h{Xe)Gq~M~tHyr^lnW=fC4JdZ20s z37G8-7TFEc$2?o#ur}>cMZNNR<{@YB{c^7vWwxdlWzo;Ht`pR)dKm^|rLRX*TrkmH z|8yAI~1Nlz@jnlv&B(ed(@6%M@|0n^i^M>iArF zTtW||*onk+aocOYw72_zr7RC=NaH9F;1``bHDXD?3FVJN0&=cfkeb(hkB%1E`t_r* zIolUESTTXa=wf=v1sQs=8mcdjnFDE}?*-;RFIltoL}ePq{5&Xf$+(Mnw;4Zr0Ml_< z_e%Er+S!V)t+b*iUUZ+*nuGRY28w&FpN0wSeCDPleabxVHJ)$S zM~-1{6ymtAuO^_CB1bxPLGZNQnZ?vcx7yCG-5-X-EzF}5S+no3XSbcW_yW=vyXAH5 z3~v!{qGy5%zr48F-pUey2hW{hgvj9o6|%hl^;P5k!w3+JPrjJ@bS?3*Zx%>D>gOdl!Tq1FLJX1%L#llsuK~{_ ze9uq&CP$t1x56qh^Tf|;OL;E5W@QdVzpkh#j1syUFwpY98wC;LOc<)g$;IyD>Z*GV zk$+cbU&@PGuIs0-`R9G~UesjioBUx!bHw6~J`dUQJMmAEspf5j!g#n+ihCjZ#rrh`@C2l}KNCdGno%tuPs-;(73QJp>Ut zC^vgJd`f_I^2cz*urt5m%Sg7H9uBgosEibDd`~bAk}T{_pIH`J>#t@y!>>QvSyRD4 zNKoapw1x5QOb@%R2mSD%T+HZpb5;(y^rz=1{2=~46{x6c*Pd{y^5=dAO+`Q)vPV%n z|2tk`2!cF&#dv4KWgrUEL03FAegP33C&r)3=AtBzOsAEi{1F0fvv7&H5Ph^r-91Z{ z?h!qONt8Yvx6s|ZFQRyK3}cdh4!1d;#Ghl1E)v_T(54M9XyA ziU3{2E=T^+^6D|WalM~3LmEz%QS+Mo;R|2OH>gzidNd&vqkB?;`-nU7{eqj+&|AyZ zp9L{^`7k$Pbr)Qt(B_7+R6Zn>J51@>I1VPY>et+m;BBLevXv}udW!lS+HA@x_(<@e z`OD`MWZJn9nZEF1_>-3994qP1yx&{gRjDq5V;+8YA3dH5vo3 z;v)%XQf0>;;VX7p-4;I2bFoL~uWl|Llb0#~odxhei49?>FPftWsjjwB-OoP9pE3VT zD+e{CiRb^|vU1~el)l4_*Vl28oXO7>T+$~zG_x=IMRZXe&G8}Esbonwg-#ALjEVVW zl1JU5J4r&3E#J=3G+ah>pybrgrxy0$d5NIM+^mnusVh!-61FKmWxB9{y?@CIMks-c zS(y7Wtq1g5X3iuZ3<}v%fH)bxVmm%FeUw8LH;N*W`p)K?pRHt|HjW;t{U9=c$YNr) zn3|Y=8p+F}sXxhQCDIlx%9LBTWT2#vL z>~EE0h$6%$jK6!&a@GSs4FN-=lIExJmoi$XBOC@`Tt=A;hpvS|1*A_YAcG2qtHHvo z*}vwlWcQVi2Yh@18USFT|1tr@6(}8JO7kwK+9%^KY3a67+rumJ^=QiN4(O4vKL+!A zPwccH)Wvc~qlW!ruGcGGG3%z-IzGh4bjUQD!Qs*PO%h3kgy+nJHaX5&MUla-Nb&*_ zSgqLF&37Jp8;{wvVSr1k9v#e5%7+Jnjt?Z{ga5~ZfJB0IuJaSYaMa^}k>k}2f+Mc* z+N1U&0c@Uv$WW)h3IkkIx%_dvxvCut@rvO zZ`Y@DsBwAwvJL)#x^aK$t;smgD%F>ry(2vCNucUHJyoOs9W~XY0Q5y3I$9$YaJ?8uV{^+Wf^piB4HKQi!&JXR4*kv^c)Ol51 zY{n_sC`O7IrAt+YPRr~&>Xq;{87`i~{(vv6bZEJ!x!kb3^GX3)=FY_~JCgVcgMS{f zQ&jtjTzsq@--#<3Bf`XOgVtJjgVjOosyRGi3%lAda+e}VCZ#al(a3lJc0HD*jL$>h z;TiN+phkYSfi$dIm9w?XVDqO5C~-ioj?MHs{wDr5@0Y;5Q|O&gpr{DFFVNJTJ5IT- zEO6}iy~*X?o$W8ERyXiW)fy18jxr}^3vn3#6hgc6P~QBif_E(2Dh^j9WQq%sM_0a$ z?1&N&fo^J8NYSsZ+nJ^a$Z%94rbb!y4xsN^%x~W{d%*Xg#eJFb4Ok7G6WBNJOlWcUYTI=26Oyq!9Ru3Ki{pFnS8zXw} zEMvNvtlaOFuVC_1mw{2hqA9kO4>21VGBLGY>3J=G^%pYDSv*XH6UHL>>ZAZ1@aN z`Esch&)v1Fpj7m8=mROut(J*g6u6uqevVOU^j8CvF+6C~h;*CM<>_F(1~QE@1RV71B^&fotfgBYNg!+HAaE~S@U-hM#C(lyns?Hv z`emximGC+$xIV^)jv9AG{B{W6XM9}w=6F3yJ(B>sQXd>wnGlnuv#vu2kToryB+T;S ze2x}qTMmZfks@bQmQvTNpWK5hsZR51IM%Qj|ygaz*kF( z3`kiBed=wcS1dfGyZJMPceN*AwG2`Bc27UTiXep~pTG7pF&3|R#nXWlZCU0V*VJDJTC}MSKB9;amQyT6(b)ccRgD94x$gZhwrXZFMLc*nTBQAIe3tU1q?C^nd}TU~l#q*8 zw3g@?UFx~g)Hn$mBt(pAEW3k?8vc0fzP+6oM)DjWZEdEoNDpC+!a(bTtGHn z>sUGgx-7WJR4%Rzku0o4poY4KFP9o#gR=7?kjkOnH?C8ZW(leHn&6<_2Q3DgbJ&&_ zL^BIPecaKU(D3+c*h#$#pONhPXa*8$87bI;ZjrcZR3R+-Xa5~M7NT3HFkMe~V0;Wv z%o@}Mwtzy(KmB8%KY?ct3Z*L>>krmA&p!kd2`FzjI2RMf*d4i;c!?zMG>T(~Pv0>c zy)+zQmq$RcvWi3F%wI&`MD))Twi3L?^*iPx@Fcw?MB1QHB6R@TmS>bzfQvb)HF8~+ zX-pE2`2vyvBb9@Tncqd@W)SJ`=$j}l9g!LZTrUygVC1*j)9qf9Y>;%nd_4MlF7Q7f z@4qf34`@19W}_YqaYC=X)dZXK2+PpX#GSa8jwAH``p1~PA3s#SZwFC2@0z8a(mwEo zbG0ZvZcCUsLo7a8qGd(kL=i5lrk70Ogo2^l3inLL_li_eO_S}9hu&PF-l5K+o#n$} zs)5D37Mf7#Gz~UHaIRJt6z>P%>YDB7Bz%lUAm;?oq&qyQWBK?eHtMK5Y11!YKYw3E z+eK`Bg(%@I`q-DKT-gmMde|0|GFHfQk6a71OSdNo+k}n1`c6`4-|qpY!Hp zqZn>(p2%Zj>HpdX?hE-D5s}Rk_pFhtGH*Pdl_g(E{e@Hu8{3d<$c)TrQ zY(bPpJMn!O&sS84K|&0ur#C5~S~;*U;{hw_3i+w6ouCfI^sIlLZ`6xIg($^L?!J&% zzVVs|qCn`c5ruLR=7eV{@4)Ej6FE)I_{A`vM)pVFSt0LTn+0Ul_2=5s2P*nsZ(K0= zP@HjD6O4!>>y5j2Ol$g^0sHz^1GloZv(1Y!BYx!!DpA$-=34Uzku}7lJJrWnHHPII zeo{^03p+v@lW6*x#XMNVf89!j0HcfDc{SJ9jCsty;>-9>cyAIygXRZyN$Rcq%Jii} z4Vs>j;${hoe^T{J0J$g01y})x!e{eE{74ys;ZH!Bez&wnV?j5kFOATt=3jF+5Q9dl zyhEy_LE~oDouu^^Ci2v(kLUk;FaLW(!>)x}g*7ZbffpWNMLK$Dio>`jILaJMG;w~g z(?ubyj~RpelKY@0f_mx^7}*hv_Ttx4gPzL^i$z_ zO2~psFzLYfb$ZKvs3d2nYV5WX0!fkNiz`d@=H|cqhyOmufBqdAA$V@T56y}z#*lO! zO-xrL99Z8ptM@>g2h%Yx7Z(Eh1A@LIIIG^n$m%e*Hi@zSYo=%ji9)@w;( zs=TWe@9>%*yN!Q;tij+C2#*D7Hxh!T^!{a4Y$h?1e9nsT2K;E2h6kOVf8I!v=Yn-? zEmgR!zZ3H#j@T)F`zCOMSUH|NHo3pX718=_ykg|GRRN|j6V0s&O#AUrJWsrrwm3s3 zPKmFpm7lFfYnnxjV@RG8+91vR(uFUuc&-xFLO?AlI^m?Kbx7bxu67xV*9N+EV!Mnou# z7Qb>a;%sZ$~HWKgb59XwDX_qW6!NJ1hHmz!urTjKT49 za{27A2Eu<5 zHd*4X?#IuD>=r!kh=2FqAm_a{R<-lJcAU=?0Kz0xg1Mz`dguL5L78_xwy*)x>FD)+ zFU9Z@$Wouvt#rJt?C7Q{+Tnn*BPN^U=oT+x66_n@rg5Wnm&<*>r6b`VYC8vDMW|S$$@Ff08%Nje#cov#p7tzNBq=T%%aX(38UQH$4x}q7h&WLFL zk%$+*+UB(RZSP*}2l$6gh7#OY#k=rOA<%5J2XH9XiQQ36urkno3#Uu9 z(mLTo`?^qpfmRnZorynzzvP$wyzeh|4T|}wq>|3uH@QojfApx1ot*1^z?)FzU3_4d z1rkf;Oa&L(7g8Fv==4jeoFdHe$1XO$PEsX1TWj?Ivxnhg^d|`;#+X}^^DjYK=^)P= znZ~es1BWJRC_#p%(|aj$Veb1zIjOCp1!xbXl_12|1q1DM2f@yEvn8!&{W|&~vxOI31>z4B?Q1hjRR9Kbx%%fSp;v+hj{<_e5duXDGPeuAfmy zi7zQH%LAer%ej)??Ag|OPaf9TA;o3?kN3$_H};bv;oU10ybfM{OIqZx;@o7=@$UEJ z*dZP@+Zt<#l8a9Tzvt)YHQz~Ir+Y_O?wQ+zn;8t3!q6K=(;st3^jDJDozL1TH`bPV zn0+sKUu12bas{NB5zkr8Kk?2T7&s%1aPw{n!^JHI!k-MPw}ecQKm~6_#LU>Vp(q6L zIT(_N`Du9j%09o$%nVj4#nBKBluq4fcX0zYvn*YLQ5CUwsFE_h;*WJFhUhttL!%>a zz<}am#m@KQc0L1qT?V6QB^R_p3Sc+=*x<)528S$;{#kC}7hEeA)e+Dq;nL9P4#1?k z39i(>3yF)@Qx=%VY1n>gGLVyC0T)3t`@G-reV50?LRMj1S@h;YTr4%@9K{Wd`l~Ma zSd=!xd7rauD$<~<*R;ehD_Y0$r<9_DmSRY_X*xGTib3?&sD2z7)QBp_CCAGik%;f#b|u}%%y3t<8e+W81Gkrz z+OW^oT4?&gA&6jP-C{C6bztCTzQJ&q_W<693JW^u7jsTq+y%&!0JFxrH<`1PaoGwt zwFtY}8%Oki!>@#|IAI0YZl*xS?YoKpxZDZR;=HyE8+-AGOq&IRU1Etos-fFaXKQ{K zgRs7``9AQY8xGUMSi|N0%~~BgfCQ#CDi<|Wp&p@7N}O$%bB53_gOQ9hz|3~);*U0jXxAjj9>Y^8^3a;Fs5Acv-bc@G1W3;&hpoRf=v^@aUwL>l2aG?u zKjm~xuIO9C`(NKZ?^7fzRGzwVzrKPcruubcomhPM&rA5!(f1m`Fy1jfJjnp+Q{6k? zM!v42;6)(1jIRo2m5cVS_P>uMg=|6zvxAi?7b?$(^?V!y4?fp2?@ufez2p(XZK$g9 zkVIM^V2(=?fK9<>@hi#`>|A_PzWwMB=5t$cTafe7RN1AX~{ zh(M7BwKBFCd+6lA?WO10n<$<_R+gl1iV;Cq-<9U}ib}%>-mIimD4Jy?c3s!9td?h* z1%nw|x3K*%p&re%3d0!H+NwZv@_DOeBZ8w&=5M5n`)T>Lc<1otuonsAOn}uXvf&r* z7r6K=bMV_g&LQE2xU7oeM&$@6<{T1$bCMeqh3Ro3yv(v9ScCl_L)uK@| zt41+z5OOdcCjq*Dzqv^p(-upii7%O#`ygf*6Rpa(L5EXC7Hf|sL^YkzHtBaBr@j>) ze5JoW!3mvi3Nq7VxuYJfo~WpBg)&;wep6mgYi))!qzD(3TeXvz%^H&FECt zt<3BM!>!7b(=GT>*2m->6%5NN2VaAz0voBdl4}lj@&xie-sidUTXfZB1z;z`nXZ!p zVnBWa+;1E}mOcZTR-aSv9_;{SSXRXUh~oWPv;B_NHCu~+3FTw##-$z|DfTuZl5H$X z3x*q^>?uo1p~c=No)F(G-_evg8f;)YQ7I{`0^>9f7oc35fJKc8!THIRjpu zb(+g({Z0^38>ce%rLiN26KKwbi`$}-<=*mH?42mJ>OTNI7|Eh8*_!$Sw}-g@Ty|j? zBl7zU_!ny4CsIkp5OW=|@e1$nd>5a|UAthlaBuo}lIRB|l=^Pm|DKtX+o#enu%zqp z85k_t9FcSFi)dVg0;@%R;p{;=y!QI-Lp=OSoQX1eVUB{gi>e5W_!O_>gu042axP`L z-O$FT*J7Y0R!}L|izs+E43mc0zIpxjRfLZF5W3Q_Vw+a%Lar8nvDlQrR3+a&H1uZI zh57T)+5camETxKK>b=~}ZyZIxo%|tzS5rh8#Co>BJYyI=bn8R%#-uIXkBgX>G0cFI z=`QNG5Rq!ZB`?Ca{PNxD)$N}$?j-&vig_9vtgw7MtFj>BkOE^Q3;Udwm)lGKoYZ?t z!7a8yIXAdKB>rxufd4~zkxbih%^2Ry=#$HvDL;&2h}Zu`n^ge@4PY_7p)X72YToAr zCV*L4HMPV?d5s!2GP!qW`b+X8Nh^j~_pB*4f|?Q1c7RT59gPp4qk=56{D3*oQ3FCb zhtyAliNoV5xBxH@adn*lxU5~wL}$@{URRM2`b4TWC*Z<<X>! z{8N}@7M5h(tGWx_FE4wvS@JZLIQV6Ql`Nmagd&gC^^c@f2RPPrso2aMd*Rw3Q0}u} zKF$Va8dGD1BaRkgfWvilhh&um%?q_%=k`j|D&Kjq@k{|F^z|?#vr_Q7vuv)Ns*A;i zWD6i=Dui3Rrxdz4KIle2pmI!Mzt}-W80k*5@HLFwhD^`PSH9`YGlg)%P43zxOER zqNymZ{sZ$oC#P&;C<;mvtuHFVnpo+fWoL-Z`)tg26?iZph?p_e-Vlq4C^21NR5F*I zuuJFAvZK=T<`R-JDki2jmJD?75vnU%bjaD9SKU)@?bbuJ%!`>Xa18>F56~CUMA3xSByB&qN2=TVAfbXw+TlFLUev>@gPkBx4H3Vssi)-YfV3m1)%bHS4~G$TgB=?HFq?dJ+~#dGB0V z$(BWXm6vfBB=>m{N|&u)UNT7DYTq!rhjji9%_f)*Jv^$N`@md(g#3g1Ui$9VxqX?%p?DkWwMOoi{yBI&-O9Ox z+PsFqWcRe1K)U}(P>!OL|J*V^D3QOG(}mNFV-$XcmV(J#B56R7odOpvt57`Jsk&S& zuH~b+;Zr7yq??Y=uGk=N-Hii^yVi#Lutw``yT;yC3j{76-aH?n#P+G&Hd8ZARf{Gz zu2_R0kM=epz6^tjl&M!Mj#hldsVoN~vQ@b92rvW_16gAYzx~P|I`N5g&SE+&`9o zLBQNwPf7RXe2|2c>&JA0vkNj+wNt-6#?Y$!4`!)ZZb#fM9nMeD0}g8 z1`-R?dSXnDTXKU_qC9BYUv1MIbYN9{$QS^M|R*qEx z5^zzWj^r*Fxxl4`&eklzRw+UK`}t&#{N4$?JQDvG;W_=`uTBD_R$$bare2gn3~K;6 zJF7Rp*;5Q@rh}H=>C1R-d%AqyV3d|F1LfJqXy-@$h5TH=lHa9_;4h&zYf`BkSDshK zD_s7hiSKzOpL$+54y?E($Trc^Ccw->1J(@vcfnZasCnIrq2f`hkGEf3(vdWBgFvBY zc8jWalPPAmSHP-Dfl;^Yab+E$s8?Es-i@SAGY4XWy>r`qr4^m|R7Ydh|GZ<|%NG_; znptc;xiP_l?|BWE0ZC(+0+Z>NX6l~t9%oBOiKR3$jUOt=lX?}a@Y~uu{t&#ARtN7~ z2|$fRC6>l5aNusdH&851I;Wi#e9yd~$ew{h#$RmA@&(EL13bKHVgO4Q-PdWeb?*aM zx8+=}(87LU4pEY=4aC$+cjmnH2?H+KFTlEQWiaNe`E^=wH}w62?>NAp%~8FPfG>LY z=E9lhde4>%6#PlQ!bt|Q{?5mwfP&*D!qrl=8qso}760z(c;z&Qi+PNdz=zPsDyYLR*bXESDFl!fOV;~k`odcH!G z?#hB^Y+oeRZO=7-$B14-k%@moqTk<=iVA4E>@51Bn5{aN5DdqW0n&|Vm}?gG@vK4; z-{e`oO@Ho3xZzx+bKP~1C6xP{#!uRJLo39tPlYzi-Gdbz4zBj5XuCYMj+C01j9)jJ z^Gs&?!aZpwq>$Nd-}`oJh@I@Ylo@<;Ab<*`#z8fSG@l=#+bge^le-p?`BiyuI79c1 z`WBbZet)kEx5U@Y8vtalv+g`=WjDOmF!|DY!>t@6WY*hatFjvx_y2UB|B3&LAp3UC zBhrIh0n$)FzXF{I!SpsC?8sHCv4{M8$EcUM6fSAF5P9oT#Ksz%CTua}$7DXS3tX6o z-Mk-GK45>M8I1aXu!n2FFi2bZhVqe2TO-tFl^z9)px3R-+7wsmX>lya_M&r(KPPmJ zCcbJiXUB{2T{f~_ABJHAvk%$lu$J7gf^3ji#Pn`kDq5}ed~gi2DJ6Cecq{otgF{mC^I+mLBlOkG z{agBVNX=u<_4~)9b%;>Cgd{^gq}WJnFWo|l0OQJFYy@~`uqETSkUnV)SbPpGi{lHR}WckTz^5yOz#vYrfM3jeAOH+kNdW z7d{D;gq~!oiPp)}>{zPT3Fy-d9_7}sr55-H4%yK7a4d6gnl?v&d|Cuie}Px_?#Da! z3VLr2e9AqIH~x3547@3*j3-cdY-xy}#tfY+3XyqZ7qsm(qIhh8e$4^I61q6xoT&h$ zDC{EE-!}ZQX~9DuZgMB?O`!Ef64%8bd{QU;#EGa;!C#GBG&XE=eP%qBVRPGSHk!n< zdtrgTh6Eo`rf@@#Fk*oo?+y0u;Gx*~aL=r<-s=Ki6~I8D0vEBD{xhRz3RQ1-#hlkyMJ@nuD{Cm`MZdFG}U~JmpykKY0BIsSJ=kZS$+1k zR>XD%4kW()eC^2bkAl%}cDw9~wGDU#Ci22i159AZ(E*v_W3_*AWZiad zGhYp6=5xL9iZn7C8e(%ix4&Z$t1G)t?u_9)RgENs9I8z>`@C^v+&##e%OVQ}Uo}C6 zv!4a=j+#+25FqkSkXsDJU3Bj8s!U?j9;TRQQ{HiDLJ1{fK>I?OpteD>FR63oMibta z)CAWju!C@?U`l zku;4yN@_4H_u>y-66CFs3$}fJ{~-E>ueFUpqrCDq8}_!5dkGTJ%@dMz;CJK>L^iSP!vN4wRw+Z(F`8!_qY7^JKchn=3IS@Hz ziuXQe8#0T932xNBGPS^}HRvH`CUX2?lTVlL>=pLF#y>k>dk&7$Hv>NfkhOvPujdC| zziMqU>+Z{rXq-jCb=>T({qW=&R*y8vClr>#)jO+_O|+zCGVDbrt|eCF0I}d>IBeab zO1J5kDMCRpRnS#Em!p8|9B^b~sMoZpN zN65pjG*hvwalmkD0DU_4#vjE$lL~LxUncQ~^;sxQ`F=`$bUpgHT9z&^zoTrh`Mqy1 zUx!(sGeBXC2j(UxR;(1Dw3cVpUR{e;b-N zAK=QrfOI>5)<2@ILDa)n88>iHsp&Jqi)kLh1!byTvflN;QvFv z^Y?`XnHD~Vel(n3m}3lR@-)M)3Y=j>y8S}%IvqU*KQqGVCCFxGew`l>TqQ`tBUE9w zPS{pGwxL_uAK?_d-bMK#34J(%0;FYhY!%zHi~RRKSzUfY1@^T#0^?r)*)dChJ5|ys z0M~Z1E+W~|9!{l^r+F5MZK67cR0eXn5^1tlm3g^$^Fa!#Dp2krGWEuh1oW zwdre?L+^M9t+&0z;t12wx~d!!SM=g+OGcfqD#=hXd0CHwLj@f+aXw@;Na z`Q@Ygx+_}*dq6AkC3QR^8wkEPJpWAh z#CFD#4+n+%<|!slY{cdLX%UNud9*`Ie@;8Xz~-{KpJ!)kQ-QQg2oM|r4?iWD#K`8c z)E#RW*N*iib;;F9JVfYp+(`MGnmd^f4FWK@912I`PG~S(scc2l zwKgWe2esUVLjNG>@MFQ(c*X=aPF?`1ze}n~jNTJm?ew<*6vC-1@0HV%plP zaA6DCK38sL(X>*ZK_j?3;;;BZy^P&ejHa0y&xi06*hewdQ@_vzcb6%*>IL0aiSZd$ zF=FT&MGho@XYJWMBlwoB{98*`X3^px0o015UI^Z)UW-lbaC6gT06QSjoVyWYq?__t)AFb zMeVh`AexMz65|0>fV3alNy=b&_$n2wn@a(9xgCb7qHe^p-x&nBzU#TVNi!kGbtC_} z!n*{R&=A^U!%`gfqm7E;U(iGN9N5?ZVP$p{Ordi>Ho@cS>_vh1R%ymjP<&0ceCJh} zE&GFPpR4sI_v}JOqGS67bp&o3J+1usRm_E5gy>JxnQBCE%oU@tq$++>I?L<`y3;BD zxpqul?J+UC_z420_WYdDHxV(_Bdps*ds@n^{ngxo7>U`Rd#6WcS%#oBiD%P8e0vE@ z3c_vlF2`_MKR`TsKdrGK1gb6GWqH`S&Ih}Mw?QXBcrwVQ%VEW)PTSGA_X#gxU)(R6 zp9+S9E4uCioNVrepf3+rhN6{n=Oe0)yqhD&C_aXFh|2R>xf#d!{8(_SXIA=Q=0e9J zdl(zpivXOeq+&(6+D5b6R$_eMf$F?NZ)?URt{uQDV=Z&IPz?B)WkrIuJ320JjkOkt zvrh!s3Bro!x}qwE5;z9uc2DDp5mJxq=35t(hr+TN097ugxWMR$YmUeGYvg~m18s2a z8DI;mQt1Y0hM%x_5_60-I`UQgx;a5@)WZ2(bUS04JYa+Q0sa;6yx!j}Z?b(%f4rQD z(;q7hcCMt^XFBw9K^Jui?>2vZHQB`~G<56k2W)C@v@tW8sow{b{0A>dwZPNg`JbQ_ zZ;QgCnNe1NAE+jrK1dvnwTeM&NiTPyy->){9owD>2CWL}sjl8ap{Z#evAc5j*n7j`-QMWI{gshPP2^io$! z8ftc`Wthmh>g|T0de%EH^tjSo8fU!syP^xvS<2kun~w1GbiP9Ghw(P_BiH6moi_bx z`35`FR>)E2X5CERnz>zBo%dzY;(H+DGwJRZ&CvINjShJ+S{z0@^1svY)-<)jfzI67q z*Gm6=kG&<&@3wB=yCu)&4cmhx|M`mS=z(KF&UWO5>f4UBBk>E_hqZCPPbk(sA!c>Q+s(SwREzm?#sjkm4&oe zfo7@CR;v%3|4eRW>-~{oeG+8%k4y{l`{2qFB9@ecULIFewVB}E87fYtTvUtYDsx9) zA4=P&v3-i!CUPl$ov;XsHAF;LOpF1FW_;W8X=2v(KF2!Ebf1P!s6kvz^(tbDLWq5s zY1xS&B%c`RoEI6{tJ2R}(q(ZVkRfBXv^}ii!l1HeQ*0E|{Qpt*9OR|k+vJ5ed_nMyH`#j0> zyYKh?PoGa}=KH;_^E}SuJkH~UT3jgz^I%UGc=N*qT+Ua{Zm$U9j!bJ_@iJ0^|2u(0 z&Crfdl)CJ;%sZzHXTK{_ehfX*k7iaqlw&deetV_w=~C#(mf;16T2_Nw{f+BEQ`=+~ zz@1D5OJ-XO4-JuGw}Q8l*SjBmE`X-k+U$)e7oqng?+YqT~zdoe+Gv1Qdtl~+T4`Rb9XDkX(-nm*`io~DVj}& z^R)p6!((g9`jyc|C(P{f<|;ARgfS#(b9#s+wV*)7>Tb}Cpfw@PbYyG1+6bA1dWdT& zu_n`3pTxr53-+m%pM1+48VLVcs;y)Hy(E|z@x6K*vt1^166P_eK_tgklX>1HwDsBf z&Nf@WDjB)kf|h|81oFw=F{gZQ+tzHdHw@B)948YgI1ddy1pd{2e-RGxaFO+x1gWT& z&*ILYhI|ZPUw4gD~RFz&4q555Vym z)r)PHfIeNTDkL}S`!?0t?LAZLA8Gpc`U!xJ>wEQ#=TJffJq1PQ2kOZN>#LY86f2>V zYha*rX|{TG4RA%PA<1@vU=&|Jcg?yn7zAT4$~QK0JCb50`>g0M@OMLd4mAa@=2WNO zH!s@LBglj-NNpu$a+XkN-ndt9J|%ux2+@zWipVV6>>LSH_wm%0f|kvGo5))AS`PNX zVe0kqRa1yYS?j?6x zH=35;b}`PnG3Hdios2++?y@U5n3>qu3( zg=}##*W<%y2a4T`O<)l8qAcrEap`J~Fg0n|Mnxt1fnlgs|5QbBX`+hRi2jM3dic-j zC-wD@MurBe*GVOuRu-0i%i-eJPa1H|8bE5Zyb@vqWSUn#4nq%BcQWC=u#=a@iRCfx zgz*!!@pjF3+$ZUdwLFN}ukgZ45QXMqoRpmsmQUdmM z%BfegZbWuGdBt7!!eU+p*r1AN@jI*9X9VOqn?Rb3&D?cHta|c>hpvaj57vsYIW9ZY zqtgVtho*)j5SUY^e&4pY6V!pYd|zO+(FQ-IPPi!uT_gmxY{kejz^b>R`_Xm!;F0B9 zGiEgz%a*u7ouvW>&~XEC_$ZkK=I9wAgYx-1^RwKAxrxH}U_RYd+_2S6fyP#;pm(Dn z{FVicM^&!1mtXD#8Iy3^Tk#Pm@wuDDTua|v#e1x0nP0RLMhXx_*G>|yN9#zb?i2m= z{Z9l}YEp9mv2~vjaT&C&ub0kRod??c8q_g{tcbnzn|- zBc!ecUkAHqMA@G<^@!H3F5TFknV*RQHXMCoHES4VNPH>g-!JPnLlK_V2tAc^@P;5jG+_oz;tEkT=BH)FA=QV@x1^$dETIo z1G7$p*mkDb>!DWWUbY;RzbFX6oBJIb^#FDX%ku z1edtGpxgnT_YxU>#XvL-d<+P7=! z`a92EPrb3K@bu>SwY>{fgX3Prf>OCr853`7Z*T9X*hWr#Q0q1P#~QoV5nLpZU8l?W zt)VJp4w%{RjHmZ%mT1}Qpw^(ii*wUun=8{cbB@HG3Gu`)iS&W4Pki{j)@v)kLklwx zA>JeemrsBg&lbPN3B<*jodo zMmbHRsVUq;v*|~F|5OpE=HD{H0^w*%4JKXoWa}d(7ut_Pn_aOBQ9g2R2vdVv^pmpd zjgekC(zQbK0>_LKmK$UU$r4N_jjO5_LQ_4va^zud{|K7F=vcvq;5TN$RCZ+}7S1Bs z&C9c>DpI1|Af`+O(iq}wxqy$2Hf%J{)$%|@;40&d$ldYV{dau=p~QmT$w^M(@Ro8K zXBG*^xdtxlYweMRO%qF}wov_2Vla&?O4f}`xm*^*IJ*;gfk)*>%ri}G%?9^CH_OI% zjm1WtqT(=LlUVSK;6<`5I(fo~8BaX&`1CvRw3oKx%;ClUdqX(X;0YADB{|R5#7>D2)KFR?74IR<_K`paUPlO% z^dV@KzW#>~{VkUfyx*1Uvpy%FT1&lhm}Y!(99qe;<6jy@@!X0+||^Pt7vzCD>@ zWZoEkhI8Dj%<}3ACe<7=^hs>S2CvM-b@N?J%u8F;O1rc3(}o|if0FDB-ETNdHK@%Zkh4CW8}feoBTgCXk@+=rMW4Ig3Bs2{vW4@!!gzq)^t^mx=El<>8utTGg53%R^+8D2)>+$^t_WCc9 zrZr8=!yu(IPS2l?Gj`l>xHd$;6v_Xc8FTZ`#|KnJ40*2zFy7u~LbeCT1W%sqeRy+rAa%t%f zb1!GBZ1d=YPY6Y0qy=6d(0f?k-+u#DZdPgAry}-pYiaYYRp8ZZ-uN!A@x@YTj%(&P zQ+-l~?QPwzo7mRpaM-xSx#k1I()9U%hU`5whiEG0*=D9)qxFJ(Os=)IW-?kq8o^lK zx@s+@WTUo3CM9^fgEBPYNkbJlAO-mo9rK%-}+L|+Dv8BlwUZ?(_#ia|^gH9Rm8!D#@R%VDo+vKC2)cGnGc658P1;@-wY zoe;EcV_DGMZyk?)5Qy^PU&YvO-Q_<%^r!&!u=NiRBDj4*ix)!Utr;%Q;D;K9J+>$3 z-Q3(-&mgQ@Kc~?l_?rW(9-|l{H9n>4pKbEmxYO1mzdx{CG8(CnzP0s9=slMMJ#hwpc|+3sbczO>=Dt`IgOA=l1R3@2pdL1~R;+U^bIc8b%XX*a9cR$jIB{_yfTMzCekXWssqL$%P{Az6VNtXm>##MYfKHVlY754~9> z?d9R@Q=^BFs+Ug8mgD;#Pz+m@Y8?kcIg@RKb(|6FhCq^0O>;B30>!eRp{W@8GCXTZ~D}C$!xy?L}%SR|J14B16{7f`Zeu|Z7;=K6t zOwpWp?V-ZglaIpYuZ>^!P<)Fv2RetlY&W?_t zkwD^_vL0e<|7he(m$iYGVFiB|ZOzpD$~gu1#l-mdg?hEJi(W%!R?auSJ8zLDtID}H z{C>Pe%OxhHwsO63OS3Qch8}$~{a@N3z&DXP_D!-dTxyaH%pb33egcL=%BkH+*`$w9 z#WLX%h|T?9#T7g9t^bI>`&oCu8`yQSJwJ0>d;RCrY$GAGpjuAUUI}Je-jhJ8y+NoI zGo;6!A*}K7dX=tO8Lz!iUfc-efx2VNm(1$Xp7TV**7~u^#t*+HCU-uL-+<+hB9uwD9t6VZxl;RiHW&`8lQ(5>yygfOK|jf-BL%2Mf*<;$rfK1VK0i1O`hCN z#IU}H#W`eN81i1a=Xzepu%%Fp15&F9VHD#paESS{q?VIMK1TUV6biVR0PR$w*0<{< z4^JE@^E<%6Ujt&I2l5fBM-1}>QhndIO6pB7@N31_y*_l>THyf5^ zXtJ)?=$}Z`0dxXiOuB^{9Tu%;Rig4V%jta%*V7n>(*h?>Rkf!w!&@>2_D1%^eKT9@ z277Y42IGrFr~bA|jYr-Sd;q0itsd16yt9!Fo}TB%S80bX30s& z!SMOcj7@dvxkC@Ss3jliD?#App)>b-Y)LuG9OAOL${2(kV(gV^l)XpLOnv21cSi0R zef)9&ydUFGul%U|8QVvX*rGR$&)~tMnAJ^EZVH_`Na#W+T=OopNAhHd$KXT z%rrApkhv^q-%R@OO;cv-FWUHgDoe6pRN|BB49OM}qP-N^@?~Xn#gMnLOBQ(H>jC}YKMp2mtiqMT1mFMn@BriHE4fo88CPrGWJ1)OPF!X}RsWWVI`XI8>9*V0X2Mz_#pG9108Pd~_b^)s*3L;ctbeDh`f!Z!J0x)d_ z+ApOS&WQ)-Sg&#uUfn7;Hj_^%`2O+9@%8WOi0_|M*L&jM3{!Ch8)ISC9)e3i|LM34 zIs>K&)O)Pe3$5gOw$(2;8RTS)yrd~0807t;y0ITT%=b5}bA zdoe!Vpl-A-eC;BN;J=#z{;lB*qMoomd4D*4!#1vDICr1%ME*I&&BXxJ1_UU1`|as- zf%MVUqwB*N=o}%|n%wm2B~P5+Mx&nMtGT=L*TWV4h^M4eP~I~HodtHtthM13Ijed- z6MB7?rHY)s)M_&F1}aAr%@mJaox4$E{ zM+26603((YE7dqXe`kF?J#>OVie0@q+k<||!7g<7@|vMYlu_cm#`5iV?_#}+c_8DJ zcPCH_Yf84~s~JAYz5D+9$j{2{r4hw< z&u6GRE2v@JI8Mp#z7y^ck+QF|HqB!@**bEBTt|G$jHo12M*6D<)u$6%N5~V^!s2s7 z6T@>NvnZYAGV{lp9^EI>?|wKt@3xj-K}mf8&zSWHT0-(vY*1$dHv6n)pnXg--K4gv zEz#rCM7O)OMXSKL*D>b)Mt_;3<|1?}~5srY(R8yk2&8TX(U|bzEA$UE!BDi@#!nkkNBY1gZ zDb%88JAM1d+D7&EfMRzYRLW!MCC3$u;R^j|e7b4nN!(JKCTQCP7kUm@JM+aUkSq17 zS2pv2ShvD^L=Ha_vbJkFj6at`b2IDW1=pKT>=gNKoIcvak)9UWW^5FLnfVxPRoa|G zUjdiv6cuOqW#iBLM(GS&$Fg1uY{B2aj43fn_<8Hl*&bnpd=X~$D(va_$y{6H;%FpK3d!tf(jzfT043s zR!YRF$)VlOa}k~z${l)J)CZIv9IEI^O+x&KzT3A(p>fE-XG0s5bQ`t7|cpt zetbxIUMZQ9ZpEvgCT^^#UuN2sit+Z}M%b0FKy-_QID{~5$89hS{GQ@87qa?`ranB3 zABm6w;_U?=pa{fsdBWIpJ=t}Z@E zvZi!Zo3xx6bRVq9Y0P}*f5p|&%+Yln-i)sVy$LgWapFkt&lLB;ayPcRN8cd?#t{|$ z%o4%m?X6(4bu_Vf93Ks+XY((Z%bdN7vnVsxC$$tMrFlV^K9-jI z7NlU%$4G%!L9)O_HT~I8yhGIvcVKFKm0{MxZdT`K+iM$4 z%KO&hiF5^U^YdTHi0YQC87ujDKw8+{VlyKk7P#oh5O0rqYOwrC#;L zTlIpHY-k$36y%VsK+Yl^T*NXVnCb|;Or@6B3Q15Ks zC<+5R55$l2Id8>14ueuMZ~vXF?=VA(QUmaf&D2Zhdwb6$PfPM0wl^v7amBj+?yK!u zp+r>DR@e27#42yB#GTWLh&IZ(Qts!}Duf8na=waDk601v)469m-o~sBCIj4=q7;@j zqN_^fFpKZPgp3V-EX}`CD8Xm2hF;ur0+^DC_ZHcCVuxk7uU1x*fNu73W$Csr|3-Db zu753>5~VY^F=dM&zHG2z& ziKOD7wK54wb^fz7Y)tSntuisSQ@6Mucr8o|?tEEF4vZ@+4J+MR`}`+yR}VcuLwJyF z%rjJVG2Jkj8xO)MEOt@U=-FWIQjT@1Xy<6p~!IOOE0b!re&&~(L4Ds1CN z_?D2d-=6`*3(ROkw()1)YrTyThg#0-49mONE_g3rk}!cLKI zf6umla>VL>zq}jWEi(gd15@Ogw}?|EQ7C9pSO|7$t)O|ePqJI4XzE5@R$%KP`M)j5 zjZd`8kPE_iKp44tEW_9(sHD;vI%d!Mfg)(O`^_EV&{$K2jg|$;lb_yn>&r8BQz0ou zbIfgg&$+G>9w}L2%?%=@WQmdxZ_kIybMkWB7?o}j?K|E{*%8(A?P2P%WFN%$=~H8m z;RoH<8jQWUp>d^?7(~x~I_%Lv9hH0(H1%B?X^*&hdtqvy-#_?=(*lzn6VEmtdF#}B zcCX)GPCtXVNp`#>gf%<&r^y1?45*PP%Zg)dTOs_Mt?MtwHWSTygvQ^4Q`BW#>GBQU z5&WJy3}aJC_i@#I!!cM}oL75b;7PdDu_?{xtyb?b^c6kipL2XjGTY-`j>XmwJQVke zwMJRVUf_*|BSt_ZPmM;&3iRT@gD~?KZ_)Wi|)P+m9#fzGI-@}$I#WBJ*$C3l$%}!Y=M;8LIq0K!v!nB z!HGKR7{RO6?Y%eeadhfhmFCxvt6E7?oWT(j6ZNi1$Tew+<{QO-s|P3F+tbuuDKmI<`i0 zcR9$@2Ulp71%V$imz-2>8M zX}SuzulgQFg3R{EEPUmKqId%DI9cI8H^s7s7|>q>{uVBYSnEJ&B}pM5GCpmoD@M#m zLm(?kB2T1u0x~D0QzCaNL-=W}T=h0H8}=>}`;L$OXBywbsJgG$q5N4*-~l7xJ){%h z2n?G1@2UOK;{(zpVWjHCq)K`b)tixC-=5@(1x6ay-#_NPj;dO}WsE~*IaqUb7Ry83 za2_&%go``(A{U?3fH+fxu(&Z|6&r0AdGB+Yu*|xe?)D;XuiD4(J8%O9T%yEBPclK`$Ueq=n*(pral(&XVx{^x_SdP6?q%>Lo?cA_LCQ%$ z^!r4SZ66MA5?1b$Wn?`S%utx6^hx=uY-llgD9fxUJ?k}oi_~iNLkLbH*X5Z;(_v*& zF81XE#x!ev{MO&5a;lFjIF-ZF$qj4g^ub;8+1Kn}56lVHGMgsWm|W z+4;~W3ri>%fFohcm*ckOvQTuw@i_<0_K4(Yyir`mZ$I;-rIo}0JC)8;EP|xz)f5Hn zl_|Arm)_+Na@xH7eNTTddXG%rRPenXD)$hcHWw&U_#0kSi(TK0=7+aGhpl4r$ zn#U}>9c%nNmEH$U_x@b3@wW#~=d4maTF19jiu{6$4I6dL)+wD~} zRyHop&)>uc1e%6gbL*E&853R0rl&jc(v@o%JDD{zvME@By_#H=C?)9ln8gpNHIc|Ap8Wau>~Wg?sEgw#2m` zm|bssp4mGQm-(f0C4TnOP$Rmo^<>@OCJxC5HJxAEZvQ1}r&>b`IMCQnP**GJYkA6f z_fM5y;uhV0=@mO)t8x*SswA8yr*PC z_l^D1P>E=!AR>M9!uRa^qXE{Fmn}v~4eB>P(IZQF)WV15A9oj|Jj?> z4cAsQt$p^D{@MO{VpGns)9Dg4uETyS7F~BkxBp`L^O)=m%t$cW!uHlsO&zb9XrAMA zu1CO+4I7iDgeRPz1M|p#yw4tk8f<)7@D+5jB3E}8=3t*!Qc$6DJuau&TZR;RC-9E5 zd59&9C!iL1MB5ZF`J{i8q4=8blZP*#b%{h}J(YU8@$D$!mdN;PvJr|$nO&wv;nhzs zH+0~h%u@WPL`-pL)sarm_bAAGlV}b-yfNrNzW210Z16t@LIXXVHrF9XSU0auMwIXKU)sO`2}c7 zuukyx#e`COvSik2-AAOar-u4%uNq9FPZJ0)#hiQnHkQAmm9`71YgJ&9WV@vb`Wf4pNnZ1^w8;fU_}x9MIrZS@Z0Wv z>Q;A$Cs=NtcnAI6G}lG9L@@KO>Ev8c+l_geCRd&!XkHR(k*31h3v8O+Ze0yUGminD z0`h(Ac&tS{G#08jTKRB}yC)XGD$7*xFZ19(PW6BOcpq3M371YikH_3dG|%df(qfIv zb5w+Y_cXe|?k|Anc_@S#O81X$YUq0u)J)>9ksE|+$EF_(j6nXm(d({}wU~G0YUe>=#6X{$e6q~+k0j7Yen|4bxX-)YfpA+)Z z14x_Jhg=#c?jh`AcU#+xSL!pJsCNK>628$RonVsHQGG8l^>(3R8mREXu5lt>Qc(%q zN!NrJ)t}eA;`Ey7*B$5|6Yl0#Hxb0F&&K8b*vE#|l;F8om>LGPI{yz2-2bj9?ncsB z#ZO$6Z%fB$b)Z@$KDrSO{N_X%fjC}SPSN<^Z zSsclGJ`U__DL$;x@6v*Vo|_+t$yjU1%dPMAfq2BVWZn?)?O+#tqWhEPrDrMFUv}j( zz~KwQRTAtX)V-*owxsg!a9~?}YpTs-b?KY=KYo5~WZJs~8|Hgfr-0*!W~D8@GvpZf zB6nDZA{U-a{i#L2r-Wtv^!ny}&`Br@LVndJ;#H?OqFabj$F;xHR0ifGL9NP z<%rK~7Y9S0kM1Fe85C^*3LUy59V_JzM5Bu2+xOSH;!@x&?)o6G;I zPao5f(+&zm-G{OVzp_|(69#R*eVG9c`xGtm;E((I`yY`l=ah0?yyMq&2_OI@wfZY?CIf@N zfW4w=|a1k!Dv(w3a#s7j6>_6 z7OU^9dS@Z3QS_>sw>#to-Eve+i^ml$Z(gm4g1^Tqh9HHEbnD|r+n!0oZKx1Ycj!BOrsA`i zg0-|av6+ST+cRK@kyeZUHnLORf>oxXq?zbDEJRne4YBHyR!&i1WMZ+b3Glc}P0W0k z8)iyvqHjU4-vTLrO~`mLN=5fPD8J>+B^qvH!bC6pchCZuBSwLooH@I~hKywLUCYj< z$FfV37Brlf%Xeqv+qpkFBb}!@#WcZC`Sy55R#=n!zSmiw`jkSLei=L z28zT^&1HjjL!1lHjH9OBuo>g)4y_hN{G2pyC-qt2>ulKU7mNh= zs;Tn@o3zhgN9_}f`LS5bsH%L^jjuONHQS)E^NvgT1(;U7=3*_}RVTuo4g2bWR)CKi zJh5l5;*@A^PiJ;N=l}Z`;KY?zltaVW!>?uA;PVf(*Xu3)2_*Icqq0P-AZlM5Y38Rm zNM$C`ik}Cy)f`6RdennC@k6uAl<{AJ_pBot0w?|bgQIzQpW~(0+^@g_vy)h=#6$++ z&X)u1xoCTI`EQLS#v5HrmOn*`D`)p_V)HJS_jZ_PUaDCu1U2f=YR;AQngCce4+CgC zysGg2r-YUU5}HlClz*irQnSdJ^}QBXgRODiQ%n3&2C6J?$JKW|3I-dEB6#L;OrVBq zRqtf9fbn41^4S7OW6yNA?pT(HwYeQV^ymqo2~Hf&LyVL#fA;yNf6aOpc)4S0lDAHcnj0cyTQ{Jv_Nu_Aq2plY}wNQ;fqan;Wg{$Y(exce^t z=R`?jz{hV1Sdg0dLM?b6@rHi}T0nO#>Jcvfu27SuA0$XA5Hh?FU6Yj#~*ZA8+T$tgcBB%12tg-0dzgT zicgWOH403dqv{LF(n~V$s0zrmA?+hN8hOLz?)2w=8M*G3sLij zO(l?z=^nx@a)m1nA6o*>@$sc6#Yb^RTB;7zH3FL@NyX7)53KXFKJ?7UmwYCk<5?s% z@&D(s@A}Eckg>3J(^ZpZf8{$`Ky|_gU+E3jE`$&*ejlQ)7xWUZU8V9-;_1F| zyHVPY?1wq$c)OL>lq!knS;zA8$&Hru4-*Kd@{iv$z4k#P>5jYW{Tfl7fvpi5N+pMt zWWmj;@PA-mwW;i@oX3j)zp#?6i3njx0MSP~Ewt9MdtJz3OQgo>l$W7~{tMA{p!U4C zv45OYYm}4U8vgWG^J`;n=-fn_a4pYidGK23&%DqLA$>#^f-N8<1tSz%Y*2(S7g9#w zBgp)$CJCIV4qnh3>t)XZECBtM-PvOXXNRjVJu^3O7pN}etx0WH4!K-C%Umlqa-Yg4 z{`qYG{iykEK!~M*f1F;}I$#u(VGGvTkyY$>t|4f$HTHl&zmx&R@0J}M>Lctv(KQ(! zA3r`g^5Z@f8wy{i$H-iURs_^jZ-d5%?>1fMvw`i@C|hH8r@6QHJc4a^i2psxD4nM> zcL=&Kz285)MxF^MF+}C%Ib8-%AZCt4?fSl;0KT9i-uUKAf{SXp}QST+G#{lI-XUbHvT5!Gvd;%v#W z?_to@jl3V6-z55_|6zcS)(|k==i9g{tx}ch??ZLU#Ojj_j?9`1oQd5<8;#LG0ZXjd zpT!Y$IVmo}{P3xMUyctivItobX{1_WgU|fktGaG8XD;&Se*6Ju-xR36Ud;2?o!V5M zY{YXjy&a50O}9vxMK)DL4GN0Kj)W>*jOuRvFwYd=2|-2q_-*zF&gF$m9tvU8m88Rd z|8j1>G{|ZgG>lr}2mA=-X7V7`L zAk?3Fm5{egRLLxB(-R}l2}j6pHPi^|;;p!qp}y>4a`Xl0=EZccmzu}9e9f;MqXACJ zW(q#JQfVd$O&P)+z5!LHcSsYFeu0SOzk2U!&f*B$w*mgp&|e6Cp^5%FG}3k1zR zxew&0)F&{Rp8fN1Ghiy_RGvJvABL>^Jq+zk(o__b9S&}K(Vbi3dyi!Fpy1mBsRvpe zPp7>Y%6a`?GRo!~r7tMEgYMU-dtK^eHVS?$A>Umv7imvYyj(8lW9R8NTTr>FJa(@Vv5UH`rS;MG4ucz zE(9j8d)8L5!0!0=U+(p&2^%wbmN#&r{`RK}_0I2em>+2`p}3bI4OO#Yqrss- zhqMWHjx#*lR@`;>>g!muLTSfv8m=n7%T`C+I_Vz9-JUrv3>K zAld0kCKt2qB494FHpw(LEy_utuj0SfT=B&*J6(=Uxq#HTk|bY;-PFSa#1{b&--F7- zH?CahzC>%HdneyX-sUO0q=%F7&Hs-SUF4i&-)t&EUF&;f_njKG`qdrv4^nE_b~a@p z&KvuV(ab8NQR0g4d&t#XT{z+r$~b1>ytlaQq%;5=PP8Xi%NKv>Yg79};g^&-m+eg) z#J|pT#&xFMu}Pdib-2g|zc3aT6bd@Csd>T_|3r@LlV)e7d%<^bKUWHDtxBnZ>QKLr zyV;*S^=~(vI`!Btv&@=Z)n!3zETW$W5Es;9q&4jh+Y_+vu>P^@PT);FnlPl1Oj1L_ zOqW*m@s27%3C+`7M2TZl&N2wNvEcTM`<7i^yMOZWpCC&j)sVO3veQ>n3q36IDKBQr zITLaSAn3A%8I79xd>jAFK<(0*oN~T@%KTsmOzCV|I8m7-xqw(H+c46?+-Eh1knkMK^B6kZ zZUxwmhg^G%A9wWY@->XfiP53?Vq=SP6}U)$X}dEU)|PcQ)~ewE()a%2bS?2sXH~sw zxs0k0zFXc!s4Ogg5=LMdmsxI2No&xgNdAi2PrU!-tp9d%#hGk> zjzh@sV?R*vd5|~w@+Z4}{Jx>8d3))YltYd_&>CX#tVjE&xb*4FT4tet$o(&|${VOQixI>@bIB-kF46s<>1^icm-*6CtGa2ub7ONzMhT1Jm0w92qkedPP%Rqzs$6pzhu zV)H))nkGC5Ii@Q{hpq6VNxxL45NbYE+2i6J5~V{tGH31pgYVoh_D9IgPN>O$mq3y| z9B^&)DwC1$@Q~ZAVLS;9w&Zw0SvImg@BaB7WVHCZ___?mS*FXZnO$blX#`o&v|b%9 zgb+EqTAID{1KOzc637{V9hh>o6gMq(S-frJZCo_FQYH=M6DGIW?FEbnXdTX*8RyM* z{){9k%xw^>0pnLQNz3lO%Kjq^`axrPCBD9!0iLssKdcV^xj#R~f8I*Z^+dl(hdAb1 z_alxhprY>g>M+W-%FB}9xF-Iezx6*qC3@4AOBnF@7VEWbJFCj_=?HVWSgI;SV+gZd z1^(OC4L<9!?}hBsz4y0(zTapS-Hj)?z)3DO&f*&>P?}w?Ox~;`L7@fpWwt($oc|zz zmBB(F<7NC!Qi}g3rY>);^!Ahk(T(1XEb>(QbNR6_u}5gpgOFBD-5>W~lUz$7jBZY1 z)rN)tT_W`F%R#FF;Nd-8c4(y{5=DnK7lx}vEwuUtV{h-OCqOqxjap|X>yf5EjXdjb zu=1O=5^ll!P?ocGDP7ROfZN8)?5n>fd*bg!cm0)z3dYC}of4;ZcHn>(_aV;$;PB>@HSv^g^zuo`Tei`dH}gck#FfF0Adi|x@{C{HEjaK zjU$6u-LS2@SBaG7x2;uDsdGN|0@#T%Fbq{ob0F&n!p20C9-g*oFe!3)qjkzpXUc_Q zC&Hf_Pt)P4iVa1qhh|7$hVoVuF!YEMDV|;#nPl|)Z>n8jqfDjK^Y;8OC|ac+ZS!8< zAMhxg>D@YA$%=U#=;^gz(MoVec)>cCnqo8hem(K{G3HX832uQ z%MhTeNQ?Ie2R(yIZ8?HHXctsFZ~pIh2k2w2-m~GbW(17;jl~5=?6G-3O}w1#zm%SU z{H?*5J4Q2Mq1I9*oB8fFtI+#6NWC%_8v0fT;M*2K8~n>G570ivFcW&)s3C3U)U9+O zndP203PEE?0x=Yfl_9jQU-HL` z8;jQ_t~9`WTsDdsQ_Enolb`2$??a!Y*+jQ+&|c_;l;nGNcLKci@gMNkX)4~jN{|Uu zv$6k$x17r#buz&XFfR6w-saS=G=c9B7==xfM z-wzIm`LWY5_}o)U6HluXU!5(55UI5!pMio!#}uFT%+^TwOiZP_k+6=3&CfPlpni~i ztt51$Cr=$)J)&Q`svI>!6ORFsn9^RbGu!H41-DH?%JiFCuf zHzxPZw8iigf=FoNgaB#%1_+XV)dCNr=&?V^wlfvGR^;qBt+d)8YSifdH3Y4s7L_fc zoZ-E5f*t7FtWm$$VK#2GC_{V5XDv#ra|mj(s`(swI3=G?yWh*I$Z~e%Eaf#2iXoq~ zPd5Q&Sf5qp&)FlOv6F0CvA8;T_d^<}cnER9jZ%Iu2_9=AEA1tpm07V>k{3f|q%*b(;cgJzEK+o^CXjB=tuo{N*01TZCroXy{nW(TvBT+e>^-YOPi2seBZ zVqAoh3u!Be33Rwo7Lg9AxG3oG{Kw}c7QQ3BA~XjApuIrCBuKe!q$k&v<<1n~mt*PW zGci!{3ldQ%Z}P1|s8Gz+9GOC=2*7Noo$M>7{pi~wA`6iDnAT--UE0>i-bAFj@{#%D zeG_~+lc1eDmtW2Mg$~z@;FU&634flVKIET5BpCPAe zXT;^zY;eDNhiam|AlSw5+3su}nzAB$J(cAvFGM0>P88d#sls#+umrT}In-yS-|PBi zCcj8@n{Zi)MdJ{3>AuaF>MUK1OXqPo`0w4$jY0_hiy9^60ZE!`^(#QM1w@Nk#xU4a ze}%%&t_OMc$rR+m~;Jio%@PxyvQ}p!OEe2kdHHCHW5P2p_xtCC3Z~#S6hB)8>=u z1Vq2EzhTe67xOVESgI6=pA+9JZv1Wh72$ALQ9lie&#!A^(D*4yUMD}W%YXsV?qn+qVX#i4p-1V zxAA6pSa;%(hYq34`rs@GVua?WrxN0XX1`h(0y*2iiLIK+qmR{8`qw;_8+d0|=@lGs zTe^X-@!^a4R82~D(LD}c0&xNFc{`)ZMf}Po*=StvT(3Ewv&tGrevSFrj%uH;|HCW3 zNNoqGAFw1ZB-EvTznwddJOC|7B3KWiQ^a+mn_JSWyU+cq8^c3K>Yk@Em$foYh(hF^W$Uti?MA6S23hm*7SNd;PhA; zX7^#tNNu3$zctg99makPEQYWGQdgFF-o6Q96#&&vb79rIUqfCfH`b=1EpX^>x2VZh zg`L^Mi56@A4@0H*PlKDT;qbI-I9e$gw4TKXwdi``)F|HD2X^2iI?kiVYP*mQ0@ucujhNA4Evwydc}lpeRssVDjm6`Gj;JuY9wPaBO2e zUeEe+SpEf*q~KxAOLW*dcTe|o07fg)Rrwx` zAv%}*{EP4JUNqQ~a3a@-&WyiYOYn$krBabOC{eWput^<{qlUV?NU{!EJMuWFwC5sT zI9hAZoC?&Oo4&DVd$qoqF0w&$NeJPWEfG zFAA0LuUBPjr?8Wx#^tOW$Z$3WHI=Z4GjA07(}!7AntZQfa?XW}hcn%RHbCt2Qw9YkV*CP>*Y(mM z<&wnEYRTdfZqRF#yup8aEIf1#HWM=PjPRv-AAu_j&tjcbRvutyl}b#62Z0h7`k_m7BNBRVu@69O|t�G_Cl07ODK{;8(!4Ue9bLRMsV3Xj$spqqf zC#60tTJ90H6kP&E|EnA}?E7z@zBxQt?r7D~I`#hvT_QDc3w8sRdBVokUDHjg9}9Mz zA!HFNgWyL^SG{w~meu~&yhgwpqy7;qqZ(H*8VgaM7*Lx~i0_!Y&g2^x2U#~ZaPKq2 z>gEap1&l1x^GeDG(F0WBH&tx)sN~F+!tESeJdi8>YEVxRDz>ixp|}M!~*<- zDPH0#BOEti*^x<)MMp%+^4&Y4a2RX?PYbkI-Dp@v!Jwr99o8(v$}YLhu@at7Ynf0X zQINT#K)(}suG4PE@%=|$WBHTHNlVqsDC_)nW?lRV*$e+kR(bIbwntWED5*133^f244z zN&@f0E1bvxD!obR0vA0PnX=6fp83Bk4q(!!~5S2;aL`xGiyVh!f=+8s{z`O0Fhm~KkiG7N+N!p}$v5}LVig3p1Zokf?*>Zbx$f9>-9cu0U~sZUl4OI6d;kMb-F z6+VQQ$uBb?^Zm0dZD8Wy=3G)WZv|aef8<}TgyNxN=)bZP&A$kb;0M9RFPDD`ac{jO}JLR(GBwJ05~rNTY43e9;+}^HPQ#>40zQt`48s zpkwkg=7KbD2~tNXcmxiIKP&`;kXCXpjIVT;B5{_D9StLFKI`zzGe3P>+D5R@ViLJJ^Bm)=WgiUkBLAiYT^ zB%y^Cst5>3Cxi|vozMv#@-5u=-Xq?7_W6Ek9?{3V?^<(?Ip&yS_H*>EKfKzQztB`!N|KNK})b;y{>~(PL(uj#~Nfn zd^k!~&beD#@uW~2n-hFf9}R+lQNcz}R% z`gA?0TDOP!B@c3%98qArf*r2!JnNmxKc6XqI&_6wLnj9fR+LYQKz&S0zvfjG7u zk`e!NW|hALT>5;ZT?4=ZQBmau9r9320kF&d{T2R^&A~^=<#yo}rc9qM9ZIcBqo65U z>3kjV5o;m1K8%F|FfcrD^@aQ`>>f_nR?Hww??;<&M+|nMzTO_^z#y@=p-v+ce1js7 z&tnu@5b+u=^#ZH5dl$1>>hHP%f_51AroSzEstN>;AyfN0yB)R85i#?TS@YrOZlDpI zeL*E%QN6#xOF}hC@^}wDb!Wds3FEx8KKu;tY)gT7L{@K~2iFco$W>#oocW^HP71)S zD{g{Yvo*Rj+~N9v_mb7C0b-@bNqKKq0?Ok6>+*ZD27mCc&VI5S7=9H{e%iwe2sPd0 z3#z}no%k2A(7a?W2rd;;T}QS-45op+2!MF8@uBWd9ag1aRv_Px(5Psts%*KZZ!l~E zv<*{#`6ui}CHR=CH5hXC6iKh`a$1%zT{j<@G(?D{d>~7uuOym^7-qWE`<((>rVYsP zZJ&|rCkSm}t5=U(mya9r;kH)2Z;cEk87DH-eMZoBKZB=uu^bbrtpB^UmfZu{*iH>O zAAT~|tBU{OiQYl~Gr53@H?0vMq!|+s2>*iXzbC0bzWKSuLtwG9w4C>`3I_-a&;bk@ zcUY|OH2;1b{Nxm9o*SR9L+PqcL{XKi><4+(C6r*rNo2P#+f~_gQ+y<#t|^ zFjl?{o+)a=X=;T73orBy2;B zhGW$BR^J4{!15ESQSjPnisU}?kme+NQ($|x!N;nuox{A3Bv?UVZs?ZTeHqXlJEYY5*x&qtg-8XJ2bJSYJO76zTKrQ+3WSAf+U z4H37MRklmEYpOggG;U=QzM(1q+}(fuytegMgw zV-yBBN=n@ehhN!AqAEXv-+dBhe1|shLP*uC!R(af_)m_Us@6R`?&GrYkEZL(<1V=E zF!YzESJ4R5-$i+j^uXbs-w7ANxbyj=I+v^JKE`HEeQ>)%s;LT@-3koIW)Yw?Ec|{E za*J)BngLTrlN{^Wj3PcgZ4qY?8#cEH^$^Y1SEsl>J=|?)RXz7XZ`=RerS@5(?E|l~*wkI{{jn z@AHk{uZVNLTedV&m+eBDu(@Pg<)Y@>r%fvuf2>R9>%FW8&CFujUXhDrTVkmSK#rSk`CbLESSjd$^;$pRQpSlbds8iz7(KhsLX0y7|`1OO?O`ro8f{R2nj17Fom z)qNKL$oY2Dk#*{@_1W>NQ}~N7J;U>pG6+ptb(71Y;uR}lZZ5!_q2-}k3l}3-xXOh^ zCWt|g$phbqRZ{&ah0yt5<8mv_D_c9AgQ`zkGLCOFL%yqNUzEOc`u;j0ajGoz%F6zg zCyyT;SSrw3;t98)bM+9Sxkc;hfheA_Yyp)_%W1xUl;qc(05?8o<(?hv?+HeBV=yAY zvtD;lt|-_3cj=;6JsWY}G+_>Ag@GY4rw)QUw~?s*#&EQml1oD9vN@qEoFk=@7G zjVC}pCpSQU>HXt5>8Os^RcG)TT3C!*?xV&Gvym<+ew%o%;WqVgxzP-& z6M1is%_~i}Cxny7#F0IudA3bYk^uP@E2DOx9}H_vo-*)lj)4YA_)Y>C9~Zdq>k;#p z$~qg*!WMhc+y8IzI2{$Rg!_0i=$;8(ADsW7Z)|lsOK$Kjlfw>aiSrr@Z7(7`(p4Md z;X%$KROQRi7M**E(H~eBJ$x-6IRy{rGZvQKPgMo$3U-|o2_lV(&tX@Y3 zP`m5_aT`hBAmYZzn9MPH8cNNi0Z6={PLfl1_CCdRtF^iKf#-HXT$ARHzov9-!fXl8 z;CeC7;3DvnGdo6+_*puYI)Q};k2iN-)E+(A|5Nh0!Jj$nKrZKH11qr{g^FmB`?MmBwCLzVtlk`vo!c*RkTnh&ilUh`j z4TN2N`$LpH8vDI-!p9N=@wi7)Ws3*{SC=jQTa##wcsuDiNyG@4o*0?p@ZhY_a{S`c%oywA9(I+@hR(Ixo?v@k#0NGXx>9}5>G*R|^ntuQN)3R^?m2n}*O>hg z#hd(koFA15JZ3;!lyUU^(!n!heU4E@x7r#O>6yT(w$OZMMyA_@rGjr1tA=mD2Fvq) zFddhT?)aw1)P#^i{YVS2a!nQknKa-cqv~_)_EMqB&Rk|_`4j)vWcP}dP(+yOAop~w ztjbi1=oRP5esB9D_f(IAo_;kB12vTR1eA{!N#^IPbzEHLoXJtGdAXN8TT8pxW;(>?iQgNuGwMHP=5#2OZ_g$-njU<|*42h=EoM*LogR zc}@UquaK%7{q6YTmV>&?Gr0lU=NHe1Jbkcl&>*qauZHl*`>^>G%wWWJdI2R^gxxX^ zTd9$E$BQr;O2u<#vc;SpVp)^?Ur#N+H|Uu_NsR1ko$GGL2~%oa1dGMzmPAPb@W};( zRl?x{qYM= z5gWwj_d-XZ2&~xIZSa3O$~u*A71t&`jDn}oM{E*zJ8^5fI}NJ&P3jI#&)Ve8&k7hP ze?p{qjZu}CYZa(es+V#GHymQvvuSwUJ-+CIJR3n(2y0^_Y%3?s{4&@tl zL1xAU%9hg^lnU+BF_A*5)z(5j71t(4X;b;WY<|g;*(HwHWcd>3^UnO#MuxM)tJK`T&?!U>W}|E{C3^rN1A zbC_<)v_c7!OznZL_`U;_WyC*2*|TluQkYFj&B72lHOqD zukQ;XW(D6bDISc2g7m$d(2mNePE}I&g87{$Dbb{G_*ITZ?P}{}@w6ByF-Bvp8*~z* zHNO*|R(y&yVS5L2w#}*% z49_FlE=liq)!>b}24Yb?JvHvTA4D4Pq`dsS4Mi=3ufZ>Xo)HwI7gM};mBFupI0|~# zzK6FmfiHu#=qsn-TUou>r=DR(k!0-G|2+Ww`>PiIrT>_f*}VtwQMnOJ1B{*D|8ny} zKdH-6>fp-o4O$9%zq`~wnxt$fFtC^Sl^p0|f27tWc9Q5Sj|+<+oA7@7R7E9ov6QNO zo_So?bU(2bV&uw1vug#9P1TLOv07(3byI<)-bpV*3=EJuxl^4R zED_ceTcyBK2{`ibP+jb0xk3IYpNkB9_P-J^ruRFUf`Sd~fHlRxS86y~`CV zItDC09KM`y6nr$Rkv&*nXAF5W#stc~V6aShmdi-jz!J2+1$RL@t$9tCR}B}@v*uXa z3V`n;uKka041kma66j1cAm!$jlhjdtOpyYNZ|Z^8?k;X#In4YZwZ+T!tFx+GJ*z|S zzWen4bV|iOM^z#|3eB{dkBu`^RS||57@9)k^Zg>>`YNEww#GDt&Gv3EVDDT41_CWS z^Y=c56thgud2J%Y%vO732YUz^fSD*>ls@X`tN<%F0$NTxji41ltA*ti@9s_mD_E5ous?q+2f>e3-6{TQ`APKv z%k(=r1O5P2vvq6bOI}k2SS7!OYrZ%)Dk{X&}7Gg&#~%G%R}-&m6X0f zmYu3?A#6&=1Rj@$4*qx^K3ut~eMTUVPjgb#ZtqzTT!#8!FY%AZ0?Zmq?($fezu7}; zC+XT$XgCZ1PD*wF`K0q1V{9l9Sd z9D7|7Cn+jqZM)#ySr_lz$y@PuRX$nmDzo}?=Jy6)^`ASl6|?Eqwam?bBeoL19ea6f z^%&)S#^<8XZ3;#4&%`EQ(SAd(u!1K&RrMSgdq9MCo7*qIKQgP z)k@^On&I4X)}0tzl_*})=JWcI(>-GiCQWfF+2K)MwV~4Q#t9+zlIMi?rjp^u7_gy0nzq zx;8%8XX!aEL6Haz?Ju%=P+_QoM{De6%!a{bA9K-{znhP?)m1Bvv-C5H44g45R`QrQ z`;|Bu{ripj1Al{$_Ept`LbCGQ+b5Fa#G=m%51uVN~gAs*oyWa4p|A;j97{Pv3MDd@x-s&blFd_l6jw~X>A zfACYx<-MKwIWpT*Kj#p50n6iE`c%=p^5GQ9@YvMKy1dBEaKjws{QdU#69D(noE4(! zAr;9pG4XAqkLfk5e6A)xl5u0x@$x{o(z*@};jl6#>ZPe4U~S!;IN#4auJ_D%T*XCS zBU6VsiOcSj3ByS3q|ko`pzTo*r?ws8iIY=uPz;k$V$-qGss#<`niV@%a=;2V;eI@<8Q*x11znFRH>vbB@RtO3#$fTi=geG^6EjY+QD`o zqxP%nav@oIo#!5m7JUB34)gJYhb#Nz&Cggfud*JvG}?O*86@+Gq2bhzT$0^8+PF{a zH@Vw63h8aE*`i^5qY<3ntd>7q2c>9gDs;ICnES$llD;ncx?TG$&uC%GFjX@tLF#oY zD{RePq6)xbI0~ZQtN7eZ7WM7G)N;PD{LHfiS}fIYNXJ`P*zg&`3uuj1LlbN3sk6K@ zcvUXpgfXWcxK}fxVhPCja4?%Jw;XMOX-#tUmPxUyyOfN_XSmQ1pDDC4tWC;wAUtqd zYS<%VofwsN5KJ66-XIYPr|&<25OEqX!xTcf?=fsrAQL=d$7MP9h3gLnqBq7}Vu2@9JER^CRfhvM!zm#aSy{5u=qP*A0nTpR?{NkkwD~ z;(Q7%7gvK`W#yeIQ1V(Jo5>~4bM~}H5b3%ff29HY<{DqE>WZzTQODv>tUx7WirWTK zCJVO4?kYWzAH*#TwhJXP5oh9QT_eDRX+F8|4A?It)4K_=d!m)wJiIvmZS?a;cy?*I zQ=5ZklO|C&8(!mjpqf1JHbJ>Rz>?(LF-MnQK1-F&f-`f3(^0w);CwPVZXLKb)s)#3 znC)1WVMgeLzEUZ@E@9A-S(>J-A1=h|68R&J_|S);Y|e|Qc5_DX*+>Gb)z(2Ml7bg$mP7v6fFXAKH*k1W%e2dMoH#;{x!KUdb<5jKuEb8p zl8=6MpO_iyTwsJ3S&_U=toahs_G`KYxCQa;_)ew-Xx;->VpAFD0t|L}wJJ0Gi5(7n zl|zsa?rZDcR82zt4KCW>@i*=e5X+@!jGMqBgS=%%j!sVAJ8%>ZQMl2GJw|UFMp+@! zA1H74GnYzg9d-9ehZvY5N5?f+V~)9BW&k+dSl$$3Fh7{QYITq13tv>}N}j9q?zVJ) zS%FT6JL)vjB=FXSQDch4`0LtTr@EUNu$kPt+Z-Zx zi5P`eQ|Cp^{yOOl>j(U!D_ey&_#RJDrSbPFaA^%d>(FA#-$e((StJwHT)_(WQgPau zKF~WU5Q+Ln1aMl(XgE6Y#f)V9J)>KYd&pjqp*Y|vuZNRdU3J>|O(CY)Gtzo?%&t|_!@1RQvVzaZX$ltf^u>T z_l)+B_jT~2e;-JBX4cm#h?H@X)*qEFPYOSEdFoT8x}xGHfV(bq?!b>5-J|JMmW0gR z5}C7k6wI{f$m6a|RW1y#X{L3Dw_aW9$hv})xz0h4h6agzWe9^ zn?1yUU8UMOJ0Oy2vH!$}Ht_AqyI??+G`GN3^s3tB7qauozN6}0rEsQLl>6)wbn=^? zNRm^-d(Z92U5SmJS1&MTn1Uk-@`wEG?o^v6VrI*pc}~}`)@6it=_lv3wWYf(YHkfo zK?W@3{MF8_Vu3C5R-5pb`%;x^B7-7BeM;BUl+KIqw!6qmSC(e1^RQO_2VxZ@Gk|GB zepzU^P|*R`zo8u&oc1ey!qXG!2a`aXs{si{({A-%wi0bSiIy#QFK58@p<0j*1!R1hlLx>3PJv)#o@f z7l&*sLXP>M@N~3c9`f5{TI?DF;vfA%ZRNr6TZ_eYyINYX!3cNNB7FdK*CeGSS11n# z>O+v`2S(Weu$^1TQGi&}U)r8sJUt?tGnVWiPJuZ=ff*+m7zv|U7s2&yNzY8!sIZk@ z)z9$kk$*zyhUrP@biZS0C?wl$xvW-~Txp<3bI`}%F%qZ0TWcWg;j53#fJB)h>6}f) zFffKo;iFDY;tN}Pr}ad7u~xy}2EHzv?*l_xkNKo|=N;c;FBYN(Ws1YV>p({{`-!}P zbfVj^xXs$@x8j0}e*n`zK330xOFLNka%#kemKh1I{d=Hq1teM-zdGx|_;&hs^-s!j ziDyBwtKePoo>KM}%l53E1vy$h?ZwH9@pp~HX@s#y3fH5LNiXf%+6)d~%7}^VqT9@G zO^oUus3H=Qixq^uMQ~?dG@oR&5Ti&$B-l>TP6J5764o8s2F`4)d6$Gc&yfAk``^IfLfbQ zwRI=0dxZ$JrWqt6hyl7k(%EctKT~06Jn=$*zOEV6fgOzKCKmY*Tcm5f0Z^-*6C!aq~+>*ayX7Pj~ z!`|8%z-)b;EnL>D=3ROOQV%%9rdgv{BV)6> zqpp-YtY>v9u4i-kVDuPdN3s%Jj(ensw)_geA(rGe%uI7wgRr<(NxX6%80j1gkP{~2 zHP#!jo;>b40x)Nf%%lk}!pE{?ZNH5+ueXqBI$%)q$s;Sl2 z!h$<{f8iUq3`n!<@%P$(r+fc+n}__*(59+$Z`uhT#++}Tf~e$LyEY1B?AsB^k04h$ z51`SKqyAZbuq5fEuM<%>Ce`taT940VbfylS{gTx>N(mrl6T>*O*cr_y8v4mck5frJ zTdv#S8Na!z!|ZRz{HmC;aB}(>wvd{%>6KDOJe$ILntQBS!uLkjBS`@+dAj~Q3>0qt zd89 zUGXSc?tbZQm2oA-$(I6GU9_`?)yg%Y>GTtpB9FQvR)MfS317UIRM^0Tb6eUbCCVzy-eM&R6o5%6={SGF-}nrMxu*sw6IY9MHej@h4~$5M}0e z!3o=sFaUm_wIsDN^o$FW4!cTKlGmR}_)6&*Utw$u7L z3mnniX~|&m*N*;okSgm!-D5rVV9?8P?cNt8vw~|Q^#x371bj6UYicj=Hruw`GS! z2;C718xkJ(z9fOKOVMXToeBnOpBuWVbj+jqUY68jS0#Y*Le?ACAk4q z#4>GP996mQv!JJ;+)g*s1_R$LAJ&Puf;wFYFEK#vr?UX)W!m1^R`2ulv6RFgu7+_K zZLLMz%jA(;{_eqsUI{a5dgk{Nt7Nkun5!!dT-V{#jvCYNzYV9l49Z7xip`8JQ_+vh zgYA>Y=8tKG_))n^>5j0sKk$kr6}Z)he+v&6{OHmf+MtYl=4tCDq(r0n3_ifuOPYB{ zK%Q2~oz*0tT5r}?YhnV&-Z&ehWs~zB*uZ zwLKnMclu6}Q4P8erJHBH#`N${h14}YBUivqxta1;M^wDk(iU%5{%3VxceBM4&`78MzC7|SMS>Y_VGp$?@0&M!F5FkwBfP-2$}m{+ z>Q!4)WF3v4Wj=Z@qBRYg-`+9a!5j1K3Z6L>ce25x294I)Eii(a+4q>ZDSMV1K;s0& z$>yTxllVFPVI$rO=B){a?SAEJG8+rzShDf0m0^qYT7}TrVoXI=W1}^7lk0phdck*6 zyv`)=4C)2pz-c@9G)Gw}iH?^~Lz-bpx_L-fs{)DpRU=$sPTRM^t@&3ay#G7V6953) zP}Hp#Ws^z;M5H^>Uxyqv^$ssTgeqi;QDFSzK%tA@IP)w<>RNgZl+S_38_+jK4uA*? zE91Ym445A5LYe@o?_|Wt=gTV1x^+oRFR!dlw4rBmp`_CV*`A>80gWj1H|lpzMpq)8jK z{S~t^*rFbu>pn~44!cNMvYTTo?%yv!`1PfLX0aq8rv*&#Gd(P@)tp{m`RLXs@JaH( zL-WH?5pU~ojp7S4pzm7pN?FQJ+{3--;Smq#Y0@JleZFrHEtECmoeVY;@1^^F*}XSd z5jG9KL;EFW*K>5q73)UL|3xHyrQw6p6*`_ihhE?7DUD11x+HW%Kl0YK#Kgn7$*$Mp zZB9}oj`I-u95$Bkv~2Yan8XzNnO3Uxd@nda&4JKO*WFprLT*T?D*kqs2ZqpeF`^DTB3iYd7d+V$5Y>B0h6W#x} z0qVbwqpgZxcyo)#b+>1<->RI9ssWoluY*1!l!QC*7!N^$&w-?$qYE2^7(C`0brDD- z1{MN>Eh5e@GEgawOm|0|CEkxPo%%V}m`;)clkj#&d2M~l85($R>KIba=rTQJt$|G$ zM$u;IJfcH7JB7Mlq|*?0NW!$?=i~x#@RDA<8{4CoRuSd2vwvngsBTk)x!S#+(Kl4F z^S!PE_q_t^!704&FPC^q4rE2d=Hxm6h(#>Jxjm?;{T2MUBT)I_ZW3=beesu>K;nW6 z#XIo0rcs-7;GQ-KNlH&Y$2FNYd%nwb?z1gRn@`OE`c~08tj%n?o^Hc!HVNa1lJvH; z-BFz2H(D3*S$o^4jqMSyM{EKHJ)`ZfWcF5tDxTH|d|IJrF_Srv4G*BKN*zBTsgb8& z&K~7cyZUirFsoCXqOLLh$Z)dr!7!?A#2=l4dF#eFo*JUxEPNth3d8A+N=-*dL>tF3&)WMaL?6oThzb_jcc0Q z*+P;0)8E)==g1yaf#veT5^D4a{$PJ08cO{ujM$`c4awNTB3|t|oxXr;hfSz<=c}K~ z^IItDm4binB+AAxW#YcZZbgUaFs7qBD)qGU9lGV`=|_z>+pGJ+i8gO2DvG7yiXV5-xam;%pnvZ7#DlvW-pn=1rf1 z*vCHq{+ThB_yfK4nbAe{pY+W1PO2pm|A+N(F~1R;%$6=tl8KDsQ4s?jI?S(o{k1oZyu01y0iF8r3I9rpnFR&ZqRh|XaM4bZ6ReC?!lL9Bmg#eT6GsbCFaVV6$q@G`tI z{;*36Ehom<`7vQS8_psQ#i zXJUpUmfW3_gNK&AnO^JG-Ail}Ol)%-A{7>6dkoL8$7ZSu`Tqk6S}Z7+JG*vN=UFgS zP}(PALJWyxYgfi&Vx%sAkB8}xYf3X%?uY5aY@k`OGC-KX0em*gCCsVHiA@hPbYaVC zr~ufHJ7D5BO4A|SkI-UWKnXEqqoJ(mH;PzE*sSH4L!@iOS9C>arGiSOb|K7)EQp@@ zQbT*0EA!l{%dLc4qid7_;AYwcS&ghk#aVf(G>`FYjasOiqt&i2tT~_YvP>jo*dWuN z!LyIxRMrz6){}@k{-T4@T?*UM=(tlc9>h=FZ97J==tBj;u5J8omp?T2MDE@MnS;70 zsP5C&7WM2vf^3KEwG~~3?T25j_q?SJmel_WUf_5Bu|SvWvSP;9AqNRGR_rPjzzkWg z$h)0+n)J8>`p1!DquYqt2^t!CN`YwwjHm%}6o2YH*R>VW1MbHZm^Loqy@}i_GP!<) zB}Eo_zR`R4nAqkYm37~Tv?9E9bBu>;1)XP>7F?`rhxVI>*-V@II+LNwR^*Y@*-^cV zrz@4(Q{S!eBW(Wizu$}GS_rfLnQervZEY!%XS!Em<==E35)qQev>+!KZ$ucX&>P!v z)@=+>FhA1DZTM}(AI0CpBMpi=@#4_tGNE2m#ir6{2?8s^>pC1FQPpuqFe%ptFJCfZ zV0OTLx5X`IW*U1qz`UwVEgq9np^QTDlT4!yVzYx!YJe@ut!Hu#_VqNQ{Rg?m`DiK{H&A*z(7Vs&`b9#IS^0bd{k0e{_r7F3Goe4}_{ncaalwrA%7cg()`UbJn|8|(XeG4uY zp)Oz|N(p?Ysz1?)1(uZ!K=wH}tX_McWuW>!EOP2lt^oCGGkePvfP55T3@T~dkHteae`uL_N=AaLo z@+e+nCBkQ=>C9Ge;-0?j?yTVqX?J1fYX=iPY0UVI<7B8;q2bxTXTo&>9mM0)humaxUoUD;}@wDB@<}GVBIVD4S*ioqfdnduf#quWmzilS#(| z7lL>SB`#FK&&o>D6#eCK@&v@~VYMA}xySZ(7FNgtS`Lv|?kIV&m7g8X+msr#wfysX z$f`MZCGV!qP@g_()ulnAHt4waEy%4Z1(VlK`EA-O1kpv)gP9H@ATusmcUzM6#+U-?t6Uom$RWa=~bCv=f^% z@rz}=r0*kD*`-$;g9Q(!AIke~t{66Ycnx{AY0dRvO_GuSlv)|J{p9#qyNCmu{6@#k zZY4QT2CdAioLZ!Qav68p{Lm`%L11m9L*tGLP;>VBpqp4kXQ1kO)e2u8AbsA$9JZLB ztGKn1T0g?k^95I{#IAVNSHzm_goejsopRcfU*6Mv(L3z9*ry`B=_TDi_MFcI>!r^^ zZwizJJ#OaebSF!JF z#Z9LN_hBE{5Q-YEv&x;FdA(&qPzD6Ug&~ryN1R?mDb_gBZJ1jo08ZNN`N|ai6g`aK z>c*B8D|m7ACM$3br?kZ0yn3?G3Es1h>&Sb!#OVUTQ2b;S{(iN{C4fZ9wXb zvplJ5rr&KG)+ELhjsCG_vbtDsGz6&J^*(%_{nP|+)#P;_2W)_A^0u=d@+n(YhU#V1 zXPEqXr0t2}lNuH_sDMb(AKPn^Ev&~a?EQ1u-GBlivp_TzWS5ySbJ~!jc!4Mt6=r99 z@~irn%V07Y*Yh-pr~eBf?Pz85^xg@2v30yv@z1GylJiFzf=?Iy%(19xEPy$B?i$^e z_5P&Xu~>es>BM?;%ZLs|Vgpin$c_8N8&H*GYI4thZoj8&_Q8s-VS!k0XE_IkRI!ae zk7}ni1jL(w@x|!JeTQ)lakydJ3Lw*(dK7eP>BY{Y?_Nj_j@({dUQxAfnSnahheth} ziLBa-2P?Pr5PO6wgmUb>ZpO!nbn%s;Z_hL(P=>n`(4&7H<2`$f`y|cRK9xM}?o{YW z*O)Gbs>CV<+xiHPf6qQ>XZqns0ic{7h3DC!vi==jce{FU08?Dbi0-B)o z_6Y+m-Kp0L>lP#Q@wy0@=APV~b;r~6)cyLez74JQ%+TbS5cGzvE$_FGLy#^Ah2%sL zieb8sI)*t$rYA>zxPeQ3>xaIX=S?3pBSl!+uI@#Y?noZ`rZK%z8YtV%l30}-E_ONF?=rqI=;2s%Ux&kXm^D?#W_>8(?$;5TjlaiBF$}U{^-?!Ip+riJvF{1})dI zQhiw;&C8g5lhSnKn`wQNiyhILD5A(x0#*4M1NUZMiDP6=|GgkTLg8`QbML zVMJHOwSL5sSDtUZpa-nB_9=)A2CAxGfs*MpE=~Q3-~w_g(Z`R4Ra3pv^;qSRPhJ34 z+tkW{>N+5MO&bO`ecbuG*M}k#P5~0wH@0eHeGSLt8$9?%XyqW!`Fp%rREA)YY?YP* z1`0_OR6%u}MEytr9!k+i2E&{W&TDJWLx$hN9wX0+pPLIkx9mna9=t+ob@UCpR$0eH^P-~P+xLTnR^Te&jwRNn zOXP_qj=G-t+6dcTjs?#g_?d+eWR^&-`x$d(nw)F?>91gI3zj|jcL7K0>A=?LvTJV% zywq1#e`fS|pN*xxfw>mLBC1c;%$<#V>4ORiOl<;|^q}+(8vnFoAiBr#^PT7*Dtao2 zBrbW?u9?+sP7-6T;>*KwyU|JRJ4KVij(MjyaC$6SuNcg}CT~@3Ct^seHqz_W6=&?M zJ^tts{eHo#Z6LN%CKER=ZgnN?v#$l8;3~NfL_O4zw_d6g{}Dw_t5uuem2X>Lq+5H= za;lZv1au4UAW$99gkBc0`i}1KcO5DQWIbycXmnP$&~G3( zZ1R3i23E9e4XoEkX<3At=yMr6rj@2$Gy~VUEOQ#)JNt`u`&JsR#6-orATk^F zxcJz6pYx*-|4b+jNA9>;vG@Vs|Cjj*W6S_CEtUO3;QQAEa3PZU8m>t;Nq$x0accgI( zP?YXtubtbR8-1{UH@=LUe7NE`iJy!kdDp%b<6nDMsvUI^AD};S7;oRxV!J&fH-{MN z>(~nkUUPQod1hkr^C#<-f~tMXls`n4|N02zmXtWwz;_WPkCHUAhg!VL?0gd2u9UiL z7E=F7t`7onoDx^f{M(o{Y!;XTYhftn&cA%c7YQc8e!T7CNecpek+RQ?^d-WPh}^{J zE=86GdOTrJ`nKoGedQg|QF(+EKHa>>3y{G_i7{Iy5hN zr_(Kp<6OHRAS_{RQthcCTIbs=P3cwF!dOF37HxgI0HX4H1Jxod_tvM6y4n6X9q+seY|90iX z2l^;IP@Up8TrfE3a* zHc6Ieyvq8Zb+gBFg?guZHrx8zA8bjxJ>=u}D|**A7)=OW^-mW+kH1^|16`HX3zsP7&5%p3a2(XF+GP>3y~-! ztBie7Tnyh#ft&p51Ry}7sIGFM{_9~f(zzFUJDk)cE9P7Uo4O&ha6K??XCa{I*u3-c;HcdfmztSh&*Fptts{ zzR+gGd^gp_}2zf=h*Y_}XA{`)(Q71SA(Ug;YRbH+zRjhPHWnh&J3i9Qv%xAn)G7 zROjE!#vu&i^L{p2owJnE>9kH%$;`lGZ0QBNJ{J3FV6C@jyqs+pku!1+#@~mo$9IQf z*URiiS!0&v%QTz92wP?>^-a;|J$}v(cEBX@-$|{zh}Nu8C2M{E#>UjRCLNXSAI(f< z?pdk~b3>Q8yz!lbPi|q#(Q6DdwW`Nvb^@D8{(GZdf{j^8B30rnCBxf`a-`|JJlcxg zVTJyZ$Mf6kHPSfe9Z=>}EzRj89^x(JrKQ*yk-$f9%3k`=p^3VSIIb zoGl@`N}S`*X4$`jt{-3*7>Q*^rC42xtARPHPHwPvByRoGpY{_a;uMC1nBLtMI=dj< z5+DXkizseaY|M6zSkVArfT}$73g4lR>J|vi$Mo)0FokS^ zGoXq9;u`0*X2r|(%lK1FU|;M{!b_3C#4gb6C^XY}c%5Dk*zf{|2gt%X=-GTHlUaAW zKQ~-_{V6CdNX4*q8heL!cX8?Fy;^U9A?>_8Llp1>ofb_b7fHOGHiDI^eeE1 zVbq=*>d~gRf%NuRN`J*lf6dp|U&i9nH1ClQ;`1O}#92^|hj@U@D6)(<68A zBl(Fw$hSX2Ha><>9p0ZofUtTGmPn&phu`*l-Imx|ih2<@no9M^5eTr)#$;-(^jm&M z&w1(L^LPriW6h*#`1F8)J;h7(Icc@^%I?HZ+VPyGCp~-H8`?^-L~XCt;f;KD$-^JK zwH{mDX^;?6h>7@mX^gULs~&B2Njw{!t`d11Y?$bjMyj$z6x?yd+dW+2aXn6q!QxIk zchb7;7Wgh_fNDz-;kf7&zdiG5t_MTCcGFH@Fhd;%Gv9Vm@FZ+uXW)1^YI*$I>epH? zjJO?`N%Y-nBfpj*|FFQ`Abk0fx{KE6Xv3irimJSc^?qZVB99w)N)r4oXP(ua8mvk+?^VNZ&`4gv_jJg8_Btl#bNih(7tLpS zU;_15CtfElIiIRLE{Q;0#opO#WPnSs&?sh#Dg5~sXcq}=G}IOQ(qD7Tg!9Hz)QN{c z3A)$Dg-w=B#mD>43V!1~D#&>3c_GRTc`Qy+o4wf?k4D=FfZ5bi$_1hk!MRRAbk5%c z+R8Joi`5c!c~?0YQw^%I#!bNp6>EEoyRYx62O8fRg2(IPhtiDfACpz~%*ji#9n(US z-NVD_=+N1h)p0`RIJaTguHg0PLN&rG`G$HtbFm+epD^5WVPWUochu34w10TQpdk+C zD*QidG~Y?m&ZqAI!t&(-cB+FW9e@(bp`^%wz@Av^C%Zfv$CyP(M`j$|ysU1!UtrjS z@gRY}#z+II`z72!_w9D~zDgk=S1ek$MN;E}Sdf0bZtAKwUW2A433_=*W3Dqi(YM@4 z9g+)?SZhsJN7&2hmKM4%pB~v1p1C<_wbCzJGEih&nN=4X`@?Z>pxmKWnZM!NALr`# zkM^^WOK{tv(93WEH9pqK)%JyA$&+CAEtHnwkB1D>jwli8ZlZafK4pyBu_+EXHl-B? zfd|z1`Kk&pTa#n9=Qw}r2@8fEh`}uXD4yj#rcp4=ZunkpbL^Xr^Y(C1fXR*I@uVtL zb)jz3-{J7Vuue3kKHy>HDf1RUqpauk$CLqO%(py3e zy(mZ#2t71K6a;L5Nbiu4gdRc&O#$gOlmMaE5PE2#e3yCOcdZ%yeKY@=wa8lL-sd@I zpL6!!$3%6x`8Yb~S>A-G$JO2a5)h1NZcs8)`W@epL_Lq3~&yNnqmEJ3%xrdm-_y9^4Qe`l?V;=qdY6xQB zMcFMkZsIRQZ9U5a>TQPTTGFxyft;`a?M#3M(dPHmob0v)u0?7++Dwi+mH7VoT+u!L z945_4;68jT19KScOJ-Fsnn%8#_e)b1*{Pqk4Z`%1@LHcA?K9LJcSn&xsB9beI3NMA z7Qz*N21XnoI!!tldzmhEdo!Q1Yp!L(aNcjL7&GyJRN#{C&mqd{a8q{B<*@Wx_{;H6 zjV!h{?149yH!>-SZkXgxTrQ`U=4Yk=MR?79=l}4>s0DnzNFI#$JR%xNmcZ8P$raCu zc=t~(TjIXIwo!7m4&4RFL8$MxUQJBKB5tV7WQj)-b>og~O*)hLpy*JJ$4`AFFnK*U zhUt{I9d)O{debO#&H+F^uRu~VLZQ!7KFd$HUj`%!VXB9P6lMy2T9Bg`yw;|FI#xHljki^w*1-#m|pf(OsSPL@;4u?Z+`C6+cYx#Z}9wL+kQcK-b%f_ zHTG1lw~eh@Zlk07#oq;$3XsaNf`;xvK5-hL(^Cd~e$rs=_NuI|ug70kWEftD=*&L5 zKjwJMZglYE{`6xgQM)ft5P19_%$t2_0GqpYCrfu-t_vvudMe#~H1oY-5h22atApSt znS>2j1%JMhR_)CaunAOoiU6B?i*osfZCFqF_Zxc5^<$j3y;cfd-}a$2T{Yy0KviMH z3j$#ipO^`1WZagpa+QHo1J_K#Eu3F}V@y>{o$?(>m}_%I46M(}S9$vTKl$1J954J4 z@c7h4TE`Rb&~jnDq!(9j#)F<^s2@7)ID&eaTg&wYEl&kRUOn-rXVDBr0KWEy z4Q-BU$Xz#8JPl(L?@%Cg1AoN*i$Ea$;&IwPG4xV-n zg!TD6TJVeMf}!qvZ6x&4IVU6EoLMxuID=`ShXju)S$|$zu+ooyjCpn`Uc;`i=8;8y z0c!y;jFCykM286@eKk1t-TGs( z-8al-$HH_45s?B@lxJ36zP^`$cyb?F<-nkD5H}Q2^vcsT7o3(|m8j`mU?L*hn<9+5GE?{oLX?sJ(Bp^>(GC1#N%Fh|`A<3Qj&~mc zO`&O}kC&T9$3qq$m(C&zyM(zKmFLwf8_u?$(DP+V z|KyUn81wln;22wvm;N8#E1*-0Z-_v(a=lO@Wm`c_A9ec+@hkLYPLfswlTkN#Gv7H{ z6u2G@Lnn29#P}g}4#ckYF>mqk`h|)diH(Vh^zWgp588bP*rbED3r7vU+Nr$K&4;F= z(gL|fIo0@qvW1kztt-RxT)SsBcl$tiq!nGj^d|1giChbt{?fO`1D!b)JElf6FP?7R z*AG@pFTx0>O}W)g#PV9=({JQ`;5y`Ej4h@+L=I7P^Nb2Fn&<;_da&-#>u0`V28w^k z#+&3dwn*vR2F?b5Yniv?sX2pD$v^i0ZLXa_4}V3g zVvR74c4h0AL)CVIDIrOhkI+`W3DD5W+tN1@ID&4JO@y`%JWSU@X5IX%U5z{uMwCwa zp_tkbr-C$Pe+PEclmi5A*G}YFSVnV80vsZ4h5^Yupl&8g=?r~l*NnDT1Bt&QmOSoZ zw#5$|q!i}rdl5W%k&MF>$sTNQE{~NFSrPi~%fa`4- zu625e=}-tX&kl-FmaA|)pnMDTYpVeWAC;#8Hij*r?jk1_M+4ngjo50g)MXrtHQ1$( zx%Ck!>xw*APp{o8r83Es)%JSZXEim$$9W-U92?@;< zmT{l930Ij71K7s zr(z%AG$hW?3_^J|>I0wFG1R(w31-zt;hxnrIz}htItUGnYNBmZ45nW03tujq*RlSh zQ~hU^sgSe;ESeUl$feeSPV5UC*=ejBHuIOt{cB==U#N=p{* zz2+5(rFBS(xcN%=B+XoH1_kkzKmKpLK>sHw;Uqp?Wl8k z2Ok+weeEOwDR1uUr!-}YFAG;V&fWo76O@l>;XaZ07OlI3&zOu+cZ%|;!jE3H+XApg zrSztc=DVh}&^MzWC2tG%T1kEpBwXyB=$c$^0=YFyT_3%4uk_#d`#4zy#Gqa{+~vhy zYSUm!?--ELh6=i!q}gU6p4j|4`%$dqT)$qNs*i4U%bv(qdt@)&YnL4P7AyDG8IuNd zjLm zAL!}9r($J0%h~ZS?O~_CeM?U%1QsR&!*|5xAnJ|U z6}BtGd46XMFR7Ay5j{FsK>z9@-R)j2o0_{AbG1XXlQ+{`x72TP>69vDEL?kJ*ozbF z(i+;8Xv6edfc$$psU!jL(un9iJ%DEY(%`bN1w;2j2zgXA=F`OA|L(7gHO#f2N>kkH zNLwnp;^XW?q?9)DTc5RUSD*%KfYo1H+Rb`w!reBWY0;3NoPEbwqbBeMV51eN(0e~( zecqm=>{s@fkoZhjvDMY-Q3Eh&2t8iAN=P(rA36S)8^qBo?@tC`kKVcWGu*nu*f$F@ zhz^*graM34WLe%~L;zCwI0pA(547?$8y`UTa2@(aqnAq$(Rc~YWxF0TSAqTWt`H&@ z2USZ7QMBukEr6`Jr9b^bE!O=zK?=L{OLX|QqHWpIDc274ws2&*eu7Bu|@c86WgU>x@VVF-OsV11x*I>!?6lU19WfX z(`X$qw98!%{ZdBm@nL83rXHab8qEHui;>vR`Se(zKZ6^y>@UsTca5k(_!(sb8P;4s z5s)R~WRs$KUO+2bxrrR_yq`kH4iSp66oCqnB=hVfEqsplj!onZOZH>0qtnz~By5j9 zP`4E5wW|10l4>5^DY7uodT450@V~*QEC4=t_Q9t%K&ZFtmQu+ykB&1e`+bAhu;FQ^ zYJIg5Ach<(B6guOg2~7O!W9tXrQVt?HIyDba@51P{=wLJZ+nf?3t`s~FY?cCfu|aV z3{3h?Fe&q&c(+-&((Xpw&a^`wz!bN{%)(|YH4}+4$evxKx{2@IjCUuI?l(xj<^<)d zCysfo{v^Jn;bm-1mb&F+ZsSAtE7VM3n`n&13s22MWl*>+oC>BFy6X1m0z9{G7f3-P z%AxtY{U-02v#MzvDXePf2C$n|aGLYdnl6INF3Y-P5>TnNo%_#J1DFs0RjMi^8G)N) z1Wl`i&0#K2JpR^a{wpL`;sl0YmoN?TF$?8tm1REdbGZ(5Fd?~04{PdEz(EH7s4rkJ zLQE4OTPSP}sW2y*em#9?<#AT+4w`mTVigVAILL#hoQ2rgXwaZV^P#}i@exM>Xs&;v zvDE2y_DFfo<1Lqc=~q0_j5+5v4**s!ohVvotPo=1FE3(2Ykg+opT(>bc`IZ-XxN0B z$l3=)-p>=No6(__I?`oJZ+oRn%T3kVD`GG76jRWcv)DSjfYm3N;Oue(Kc#aJXNxtX zo=^1F`a51_D=l7nMWX+CT%6V_me$V68v85c>0lBD5)**kN}eZjd^_sl+@d_XYRavrH?@an%n|GE7^$ zKp~48`|-R+E8QTE3IVyRlBZbEFPV|R=IizD!#BC-oPA4E^dYw#ElDc6Ik%xwo$@Ol z*%>(|1%`mnNP*OUXzd^ShrrkhP}9@RN4~@Xe=od>1dxdPN<*wZ@9$iY%2C#hQQw+* zN2sMs(|yC|R^>{9j_mJ3#+V}uia_{AymL{$?z~br62XjGdDE>pj>`=lii(CXrSjzt z9AdhMm3R@Vi*knR(J>c(-jk@b4;UM4Dp4`1oDCjxxCr0BHAYYj6OLb5B_T{`j5mC z5PRaQw@gK>-Ut1YP6Y~prwrV|MUn`s3vjR5)QT-Uf_s#2m__RUYaHNvyN$+*A_^f+WUj4~4#FR{f7_7H8u zB)O*Xhxd3Zm^D4(;6Zu`sQj5%cDlS^yUmdb$&bsm{brQb2VJF5q<2~+YhX;~-Oz+@WgoQ9 z%d@)yo;HZ{j(@D44X`^3d{-4#H+raayF7;AV`Ar<@IM~K5f13DcR8|V;+W>U6zQn1N%hbZ7k zL*RV|ElcNkmaUg$H305GY!LW}bli^g&ZBCi`^gvj551rh_YWinAtW)ZGX$Wx18ix$ z^{zo&S5faNcfM$ROTvQI$&&8%Q%f`R5W_R9uTg*~Bn=pf3Z^?}I~s4OwT|e_g{=OZ z=h5x|L`TP!dS&9anyHDpdCF@s3cvrfzh98QS8ht_F*4d)2q@ZawuJ)H`|>MG5pWG+ zX5vXy3}{LE26$mR9Kc5HD>gWAQ}>ryY|1;pHcEgcne8Z=FN`RaHa<1h0Bgt+dp^Q} zadfLnNCSlen;OB7`h~ja)x79#N~TO#k+cw&`>I-_m9q59XGYjvhM@Gu+S@l6J%I68 z+I9jUZdCazdgC6`dl75+oA3*kASnLh&G`%E_s}TO9zo^{89v-LWB>p@Z zuq(4My6FJCw-_%3} zQBG*B&;pEec@ALtAVA?F^)9f8d77%k6kG?`F~~DVO9tLcu9WIl;ypDDbu}V< z?xvHIkHUC_g{nF9YH)wqw^-2DwygK3=Eicyer0K=0XMg4WQ?#*2-t?jnn$Fj3-Ezt zXCf-+arO%QOEGuRsUJY?4zVBDOH_g0mwvRha4)M~uJS{|KymE!@n63wdE@^M^U#3f z1-(zW=EfW_Fa0P21r^^B%sJE8aD4L1hDXl%MI~TV*Hr(M{mlweZv$|Gbv|+Akq@~X z_ef-~{yZ%Y=;>gYxii7I@tlysrB|oD=$X2%Z)ahRX6~YIs7Lf#9@A9i5Wu{eUdW7e ziELztofn$!;SINQ?buBC+WiD5JE;^>KW(VMz=>cf=_J$VXRv8bVg85~Clp}jwly$opm*#5=g}`pZXdft{z!@e? zzUTUDi677pbYdf;7Hgjsc|~j2pq{J$6-y-#k1wf`XvNEqYTgBYN{q0ZE8qt5Ug)|tEoH){6KLVKRD`Z!}^iU<< z#I6ml2Y`17Ky`lY3@9klk9eqZEfu^3{D$l2%YzN*@$xX#pn9raOS{dNJkl#r>woOUiwb|_nz)Q;=_4$(@ECoTB?lf>#1iy1}a(>)xy5$Kz&Usv&BS1+XC<^D!W2Uv@)W z5kyhozMssLut2It+Y zqLyEug0se*r9>*S8Z-Xw%=`ZfSHPI7qsy(2u{Z}OB9q3fTaon>P`DuJp zLq6U^Cb@o7egO3J-F79AM@8=qu?KV{y;?QUd+xISxaR(?bm!}gbyN=Md$d@BTGkD6 z{_9j=UHphzqt+(my5XJf?B2+8g^5spJKxIL{@IB#zY0N}xvk>yzGJ`s>c7X1+$^RQ zhbdeZ)!OQa&^rZS?PgO3AP0C&jPLt>xK?y@S?)mPf55_Pc_r7X^n+V#z~#F zX>s^5kM9lZjU^7Dhy!oKdV6IZu7p`X7X2oAd-c3l%KB@`DrQ8Ti=Lnc;OyXsJdG}vsec?Ue|{OS2~=#>9(42(h22` zHsB_uw&J`<1c`x6 zeJVm|GY1jy)#qdF>|JopNl1tViC$y8xM^cFdbzi#H`ssaAHe^YzEptc$*(U-HpW|N zrw>KxBpDUUSauhhw(V3wn@(1<41{?iE#ufW>>iZi``ru>Sx=<9l!mETGv&?y~oTtt9 zA7|YyrGT3|J;q1eb>}QxAReF_LM{u+J$PVR19~_ox;Mi8bdVr!jD8$_LuZwMNR{>y zAdIdrK>Gsx<&$Ips_?u+ACt18@C7tRq%RQmY_W!%`DPth^YqQYWG+3XN?7hEC*#OX zo_{-)H$|AnaJKQ$nCU}+Pu0x=bKReTu;0g=B}L3W(0cwtaEgZV!wDL_Aznq$U0@6{ z1&-jD7_W0YGX|Y?Hfc<31P`1lKA~3)dV{}r6HAV!13w0c;3R@{s@ELD-k2S`vv53% z;iJ}<_bJZ#ku0Y(6pu0_a=MUro_sdw`H@rFRns^LFygt6u>~oyjs9eeBHkh~N78GcPHHg-}PNmMD27LNLf-IEp=qGM3 zxkVm6rq+~v=_=ysZdW&dOlK1ScY$>h`~= zHx`lMCwyP*y76W(1V-vVDbkm7_UtdG_rRg3ApePcCi*@8d8O-a-dgyAa}#QSS#{uz zbbt(~&wm*Q2?<~3IC9N(iv7-AeGA+pBgS8byMS~pkINngq&II$5UBK%GVW?{Pv(v zcP{fs#)kxzv3gC6ne9|rgxOxKEXj#Lx61jGwa^BGZ5UC64RL`HeEYj}pSOX#n;K-< zU^RYma{{p8C;|XRDGC)z)QH)<2Xld0=1o<5{5@p;vyk0K+>*k%SE?}g^Rg1R2V3FG ztN7=ld()pq%IH%Go5ehc>GJ2;W~+3ppT^!qogen&$Gqo*+w+4FXO0E0M$2zs7SWNQ z)imZbO_fDlU>NKzu9zT=cq+IGWd@Qjn|P)ci%HI6_XZPD<-(pZ`YL|5a%;58#DyN%>Mz!u~_^~MRDkzsaiS9~hh+^GAtM*6c`BEROzD78vQv*-)xC)hE`c|D*jPQJOfUV^R3{+-tQba!)OT8vEHK%_ajfm+w7@-%AYoSFgW|xyy$*cDk!K@89!V!R_g> z!SRx3j0hPI!osqV6qMsO)Uje+6OIrD$r5+A)csc60ao#~GhMZ8cI=^Hx^aLcWUF0$ z@99;w_!#)t&A9cd?{%x|d#juZ_$3^Q{(b5j(43vD$$B4%G!_ELp@o&ldl8A%=RO8Vx%a4 zdPMlM#s3qrEf~R(xDzU#IhGX}n3)fc4n^fD2%T)34~|`(GjX+zhAi0{v13}?^1XUc ze%@7JZ?Dvz~T(Wf(D^(3Jq}B?l$SW|HzTXMKVKaC>m0j4kR)Y zpac@WHX?rP3w(7n8XQR*8c4Tw<55Kck9L#!x{UX#}_%v@XV)cA|6XiGHi= zN|~a@yhC1d_YjPwU?QtpIBb|q*vs8zh?r>R>2*Y)1AF(D<@budo#zskq4+d227;8g z`sDY>7%XIbok5@S+dgPdun2X3-Vu-GvSo}HAkLpX+u^sfBtPv3u-B8l+VO7YD|pAg zxvocZoaO7>;u0!BbGCJHMh-C2U3`fUZzI2CuQ=Tfj5Y7sC>pYoG7$jG-+#IDjevtA zyYYSW$psN#o2PxBpHH+}>f|~pd!}8f1DBUHBbws#ds?j6ZI4~I<;$HeEeTXLWBvq{ z8rT9zByr44ZsBe@5!uP<4^YynWE&D;i?9qOqFnoCVW=E0T>wHiZhp9p6FlO@{XWIb@huj9XGJG?7fKGqcNm*{%1Fn+`qqz&HUR?TVqw1z*F+!Syn0{}5>&jOs4^Xsn9#FpOUGo#UQZ6LR6610 zBI4jdWgHjJ;urT#B;;!^-8U9N+(`>)=g-+KJoQJkNBD@To;&6x$}J3+A4E82az_GR zdL=^6PoZ3|zpFPKBGE)wM~h5${3t4%{ZPdD7^(grdwX*Sm^s9I>6?^;kcg{1-cQ<} zq?O9ybDw{gw(l!zH-w`M;p@GF#~gDJYc;#GG8K$f9I~+p6d}Krl+wQB>0`!T7S}k? z?o>V*-#7(dT#TkHs;aV;E+h;VX)i?$R`!WQI@g80?QaXV0yg4UIP8-dMr^G5jepKHw*m)JuCR?&}?SKG6X(2xAM$0k{otKGc_6 zONduHA4`p{CP}`mfW0WSXq*(B|4}UCfGvk|nIh#p+4ggXwD)`p&T48_ zB7K%B%)t0J*Pt~fiu3Xc!x39;Ijv(N9Gdsu@h2`U}1DCjW6-`xrCtZIiI)i#X!>?H(SGH=CfJFlzb9Q+CuT z;Hl$TRiF!735G=Kh=4eTRX6inf&b_Ml*&31vOs1~{I`VlA<`fO&^^BnXkpZkFFcto zmpR3%^4i&jS+N3egF;-*A2*Ad3HWnkyA%yCFdtb0L|`io^buc{NbC53Zhu7&>kQ~uesY}7HG)x8>evq}@BhPNMkQ$zwfdt7e) z-Tg@BsB*cNv(B+bbo|&=bzNMaobua-!K^jzXO(l!(D1V&`0M_eQGxxSD9t9+B>>&(3JI1hM3p?~B}iO>SQgA26xRG*#Av zuS+6lyxm0kK$Slho`46=fAmW^^y4s=ZTyII(Ac_7?N@Slm#qD-*hy+p=W=dlEOF=SEN(?jPuNT_4~G4EFu z3)wxBq>r;-25n)EJ`y=DvO%*hLPESi3qI@o=`eDM?aq%haKjmAY3GsYQ);$kf zV3pF5+{q2-Ne5qv|MoNf78*^1jx?3b7ThUtaD<6d6&xzB@$eo$r}_qIjreY5e1V%A zE!XcBxny8yi0ls!SFPwAo1XSK+dOEBS;5dyXzQlgVQ~;z+$G`pO%sNMl#z9cx$fkO?L1%B2 z+}WfXP>>zmo0d5Iywm*aY@D{ z0SGirE-*w_(J1f+)2_c;PM@4XrTMh#V*_|(=_@ygGs{tl3BP{5?IE}%F zdXu$}$6-f=$`sm=!VWM_Qo(xWYUnzP_@}&uI-2O#aCXi&Ij;6lH{q)(MP@NUiKj0& z37^Pa>{x5DXJ!5lj$(BNTOmmB0l>SF{kS@;KZY3X=|jRuZCId}kya3)B>R!dbY$;D zEXQ0v;!3Z0#Nlt4U$5@p&OBiDdb*^%uwTC%wIp0`e^TkEUNPLU;?cRxS_o6M$&M7fsNH6e42!4G=Er>;OyGR5T zsJ=djDkvY6gld>lI6hUI_XkcCM>*zT}NCE++}Sg zgG;ehNbewa*3<9xh3g_beY7u{Cz6i8P|ots&BCpP>*@V$nMg@pKNgmqDwS5}CP1?` z^Eu|H?eIKdiLt$3D!05j7C$dzA%zTibNhjGJ3>pQhk`%!8R$JGOD|xx&aTyB&#*wQ zTTOt}U_Az3-l;}u#a(E6@t+XW^l2chM{5(zlZc#!U#`NCkAx#1n*Ubh@9kR6tkzRU=bcd#$<-+a&uKYn<;VsS-4kn zStCT)qeUOSWJYDy!ev+gW&R*W{9Kyq<4?aSydPS7B0BGoZ(XJcPy?Nne5^j^@ZQm5NbG-}j@OC_|Pm;n} zt%UnpKnT)?)#Q$P_a7_va?!hgg|fkUX2O7PgPQ)~R$l!S(`eO7G@0E|*Yi@ZxPDQi zZpBMsg6H5uynbF|vUv}jYnkO=bA{9qp{KcmXINqIp;uMe_O%aWbPXWC7@C_hpfpPSJ^ucmo=l|`^ZB_x?oXZ8L+-F#N1jp| zA^4VJWE>!f)KIAb^G4=4H}y_EI4>i1$uBqvAd-#S9sL^wqgVrB)>D1x9AwM3GE3@wya2wvhc|{Y9itX+kZoSaa zS9DzP%-7PW!Y7@1ekIBw+4nwQCQP+dri(ySu5#G3&NFk_IF*^}!8IhGEC2hjT<;~5 z@~Jszlr9WyRWRmW?}}5tGOF9Z&4umZKXpE?Bx@p^^O(A^Iiog%a{3$yIPn~z&rbtm}{Nccb-A?u21g_ZE_vDG8plM<pO|wlSL9Z1KsKt?*GAmy1BqaTQgj=XOBG0+K`y91e=<(7UQ_& z)B&Yui7i>Y0M*^e1TDjaJR=F0xGY%F@4yJ7> zKO5G$PN(w03PTiUK7Lxat!26$Ece8&1ed~nJgv~dP7BVayeuwO2MR#3iisKILeK4t zzBqT9?6KG{e)W~bvZSOi#YpG}WLh~JXZ?`33Ui)~>NC20~kQOr4JRwMHn91-KHYEY99 zh&KG$2zb|{&Rk$`v4NC*(aybJ1E|KZYBQn#%-~u*&XAPA0Q$)oKSb1 za;{6mz-y<|Y$9Z>&u(h*s)==1`-S$;?mgGfT+#5>FRStGkdt%6;I%7cx56d!fxc8B zRoR|4kl!ns=a+U6OhR{WK==)d_BNJ@q2Aop=w9)$H*dlrxBp|ec+TAbe*C9|dP@vA zvi^N=#kL#gsK+)7*p0+O5ZEg1NQ!CFf#;Ijz7P=hZFZQIgtgRiGWgX~|^{q%LY1?}@ z8NSygH$N{Me*Qo-_5$|eWUz5a_>AMBvYWx&N-uoa<`CV14?@B?aw0RZw zV}Lnboa2O^Fkn++qE6AyTW$$*Bijn~HWwTt@FvUD#RV*pQLAx=s!kjt?gr<5~k9LcBct!pq;>QnYy!PI*ayo+}A_s=P07{M90Zv_5w0q|2+;JKB0xlM4 zRYRVoAmen2)Sp?Djo{F&=`ds1-FQu5!l8Tb`8&;qBQJi)Z5>&lBz6FM1#?C{m@|^| z@{Kn=p1=;Bimu`NbK@Yn%rymspqthYL!1}2}jvjdDDLA zkGDz?ZkgMoH1beUudMzx_txy^Ug~(ZXaN(fq&NM(+4ekIHH@vm!8@IkEA~7}1ZH3M zIkOdtiw?QyZ>a^y-np5n!&7 zVKQKpmVA!o_j_W4qPL@Kl6yca?D zAQo=d1?ID|CyMFmq))3j)Lv!^1p{>g(vg*#Y-3380w`28cM=MdoZ%JCd_o?9QpfJB z1|CC*rBzX?Q(7XbPJC6pXv&*MHT>?iG{PcI!b475MJ9ozU*5t-(|q6$j0@8_-ztuh zDV~b(OI9E)?z`@3rR85iU0bP*hG zhb1s3RJOJr`7BC#fZ#E=)m~jvkK00NL8*@{s2}H`dFSr#_khbLJc;y-8pQ|%jOI1z zk-a6|+*9Vh5J!%_EN%fkoT%sYz`jmBJa}l9`U$Y%7_fIHY?5*HyO@c*7w%qj^E2xb z^nTby88Q^s*dOH+t8h~RAa}{Ls{AcI{NLw~QLZfaK5C$lF%Nhm-|flh8oLqVh}AB+ z0rnHppdkS!Ut4_V&%+1Ldt@trUsxA(d7)C1P!@NA?=|=gC=`4me51(9omftVE_ihy z-SzcuiJept(qCL0SSqA!Z$2XKIpc6~!1y$)F1QLKbb@uO(Az~teNY=)5?|*aNS>Or zF|k@*+tr+4J!FH;!QH`@KnQLapeW9n)61o9kIZh33U7=@;MCb>fYzB2Fu;AypBVcj z#~xbq{y(QL)<$&wUhZBO!0-K!;A6aK;T~a~xzQfW>FZ}$ZP&x?i$v57TPBt}yX|xb zGJ|p>NJMo5Mg+Dd7~zQb;WKlaAj;NO5nxp?po6e?HY=Z`u@njj`Xe^!?`$FY z^)Cf@BM@+-iRFz5$9cS6@b)~-!Hx{z7Yj4AJ7rHVT`N0dy-}pxumDCyaxf1n=?NwQ zm*A}%+W&IRncvd)urrgx2f=c*8k;(fJOt)nQ1XNjnc#ezLVrk5z+gl1wABDJw(|DW zDx6m&MLuUcD6S!Ojd+~>uCQXG3+URgZ~P3D>NOn$MNFwn7uu^gAU#!q>dvZxyiV~q z-fj2vp40WyrIa5}?`UG;IaONzI3O$NL}D*2lYsky8fu(kWhdg@bq=drVe>d846hxi zfMk^tEkeylzDmCK4jX&8{uF5VbK3sY-=f!ZqK>-t(U=8eF#5$N{?XRZxl6r8fBba) zSJtgiS!D_f!R^Hm^2wDs8p&~J=ZQ^A*kPot`--F^F*Sj*)Syw+-{YWtje|b4yT(iR z+fLkz-mTc%0?hQBsLuASgWC7<{dQ{vm25IhJm4seow=!sPgeD1P_AJl4eQ3^5`$xC z$I7}dtd_c=oOj0j-0Ko{Uj-$PoFuN2V@V@+>&aVbJ*@-lwB{jVX3rkoen)<9^D0oX z<D~RjI*H!P?-Y@Zjfg5xlIm{lbG&-g=5NB2KbUk+EE@J+ER~ywEZBNs zVD#f+;8Lj($l7vHM)|1AvekrIhWM&m1&ax-V>{)RuF((`op_g5MuSrYUQ^M7TZ1CPGvmpcVQeNlEIOf5ezm3I%;syJj3(lB zv?Fgt6d!0V4b3`gmw1e74u~us>I-<&3c8+Iq}+=&lvHiXO7Xc~AqFA55H5EFmN5d} z9IK(c?12M~<0@9?f6HqBr+)cs=2S^PsGEHVz`&KAalnlAhE~UDssb$iR;!v?p>qBQ`usFU= zHA%>I>QSa#aNaCC#=D~=4|u@}#ce$I5(lCb4{;b=O!#xQJkwx5#}|zSu!tAFMh*2%+_>qo@!5>HjGLlw_9JGClwA1l zzXc(lO#5*hsDO#<4_@d19hb;Tmz%bq(KDSia=%xxZTjROuo>KD#Bed1sodWULoAS1 z>$^@J#(xJGJ%)?*CR}sQf^JrvlD!ikm5Lv_d9mmIM-%SvzYv;GxdEwv6t{l)IE5PY zdO%6j!GDIc!nb)hL%os>zQd^4)-J->ChS)KAisqaA??45cH5aSR^&HjcXs4=?<-c^ z#eg^oWjFO{)cJg_S?4-X(0x%8B&M~lY_R3jQps*`#a8le=g@XM3(FhV;Wo_$taRIt zK&~XOrA-@aIazyzFk_O^w~XCwPMo$v9a{qibK}qxjOk7yWvK9;Ep$MkXQfpZw$ik0 z@+ULxuxc^(e{^$)txq5MaZqV4OZSDcR*IOCWyQX`uE+aH{^SQP4nU#N##Hq|-41u9D_RCYhucKgu0E} zd-zxten8eS8Eg5%D{Hcw!?28a96)B}Ua0D-T-NkJ0GeS5>esY?UgHTDp*$Q2f#t2u zddyFK{j5J=+S6pmD;In4e+&Ws)=4y-I4V>hSqIn*DGcHKHC(jE#BNm#vbqeCNya~I z;okj)J+`{@<#w6XBTJ!M_dxS=cpF_>V`tRC?o2L!XOs#5sZNpEN0-b{az!;~$SvF#=saR~SIgvfdyL`n&Bza9192S| z&u=kb;&|-ip{RA-K^-jiGJF#KZ9wL5?P`5qA?K<=)8@_Y5`XW`(qf(1ZZWu*dPx9s zr0H$>taRSbhtNF1j%Op&y+2BTfAxq+1=Yj`2MyTvTP6|Kq@sRH+y8)7Q5BV|I_;NQ z){_|v9~qLjr87EhTCVd%=x3o;z6Cx2z}(KlUcX;gPo4Vn&RTd#oq$Lj%=hrXA78jk zo%g)F81UUueiTdhC}eio#6~=Gzd7ROP(sD-jz&P-F2q+jJ>7v48`+AAnvM`iqJh7?$dBc+)Kk6WXl=GGR0)F?QIIz7{{N=*SbzZa$D|8 zK##tEQ+c2D(43LIAy+5BZ@ixH=k!3&Ilwhs)2nctjVkB6j70!E9Uh0STKlKz-@b#S z__rPp=QG^`)%n_}6NUBvSX-RQ!Kro!=I#%h!Y4lzqX}DVW5WHn)LWib5{2_Ent^Sj zQWNwsod#)-`P_K}oQv9y@p|ka61L<9d0aQeR|mGX_C}wd^q8J2We^f1ODq^RIs=8X zR{tYr{Ovsf4kNka*_@KYg?4qt)Tq$c-^qChD!c*re#0^lqC-xpD+zrO-UpnC z%;bRzd@+9B@aKIEI~kcP>k9(U=jHJ3i~oxQj{D@#)uJ?@kKKq$1CIyeqL&+ zMY{1zkE^f{EFpy3(P=eD-0+&ndZuRKV;7|)f~sqZX5p75vwXucV7%3>akUHX6~A}KVSWRz^Z~r zyB*ne#}n+YBE58HkIA7XrnWnm2T*xy`*kJv1LQl{cO4A@yL&%c{rWR?eVy+emOtNR zs&334yjZ=>d(^A!lfR*vgNu!h$lb0a7SzqD;{5nmCMG|Dws~^ilI>evFZtHi3jZM{ zRW5_>>~u^Gwah#c>80om&sZ4PJ}zMxA#vO<2Vn~WL8W3a+4LaeZ<=|i(ZvP&{m-59 zzxG)E4ybYgxvX)X9iLl>+{v$JF7}FFue$u(-w$*!t&J_1YU{z*rl89WiOpc|+h~B5 z&;U>eMfoG@*qXkufX?rf4MUbgFcE^Kx00O{A69LB7*H}5JH&DS;)NwU>fm*_gW9tN znH}lv`VX9=U_D?RAaXh|9=qdKZY&+yj#2#9WqU8ygWJSW#fnZKP}AS1SByFF%C&)qzc)a;g_E*Kczv zfXGRzO z>c$bMPP$4>TI}Wp{RV_`_hI_5C~CDS5#3!Z%Z7A}LsB9OT%7j~zw+9wn<`o#JV1vv zpmfHKj2gz^f??c6;9zwM3`@XWIrBB^%NIduFPDX^wfLV6$)5&N5{~%-9JTHj*zlP< zR@~)II{Y?wvy2dC_c(nz1p#}?#(HCewv%=;>&zLu&2EY7eH;IISt@U+>QZi5)b?B( zUS~|~{!dLs`E_7C$3FA^ddEra>S_Zi)Y~uepEZ*uvCGf&)yn6H+2&P2+nYJ8oGz?s zkI(d<(ftAT41zv7^Mqw0Q>j_>nV-=;Nm(L3ZV7RD_g5*U$~(- zK9Ct0DGH_e8~%TUeRot-iT3q?Vj*?~6hS~mq+2KgQbZ9&>4X4*fFLzA=`GX&6a-WR zq<08ONNAyV1q7s%gkGfuLJz%smwEHPnfFs?{&20hmRZYtbILyZ>~jth5L%}(v01mH z-rPzydXeyKS8kX5`33cqpa~x7KhPD?;mmh|(27?Fmo)*JQUI(6i@WOI;?TbbX3bY) zQBD7d#R`7+wMntlbvz-#^ut=h|5c;#Dg(!cna+cJU#lo8ryyAA^XcESb^cozQ${#9I&DI_&?5uf|7n+Xo zn>y{Dc#^=y^ zqw?^6dcTBqtV?N%YzcDL^c2y$FsB*VBpsS{yF5rQDuc*m+i1$roFS}KYR-cBU%&Z# zLI$c4UO8a=pLgQP?Y3FQV@LRlzACEX`a?v2@*UvVy8LdCHa=~-{&(fQ%yves?uY=;;JFOap^ASnsB+9|_UcR|5-Z#StGN7cwz z_{im8p^rSw&EjqosT_FWfC4A*#rTzW6@pNpF&AFE@O@*Kdnh*f44J zwpRj>L%eAv-(Zk98n;?U-traLfD3wSeRv0KZ@23b`vHFZ?ZItBe*EbUR zXWJY8$j;Qwg)M?RM`j8guoB^0ruQ&52V-?|_Gp+8%Ju5~X|3W(T^rRZ53eV?*EDnA z6($`h$R!~*V4({u^{-gu^_|&BX4Vc58nT_N*&grx4$Xy+Sv;!KK9KZ3&~iWfWi9?+ zN#kQq+2uy{Z*i*ZXL8uil`x6?G(4{e8t9iDCi^PPD`f!+dSNL7>I?ScRdnjmN@a`! z&@$*v3wvueEh@?Ektz>}VLnRR5Gn(LQ6Di)$4>fOzHm&Nu2|9e26d1P+wr*#Z$S+x zHe6crPKK77AaP0SuzBL;eOuLr%8SIWGD}C~$E=R&!q_h?nore#N~jP_FYVbVf$%^a z7GG-O5O-2keKApiv6-d_r;|w~SnVRbRorxLitgq#9}+vjHoHa~IEczwZ5XB0D64W?u) zI3xG=ZpevpT_~Ve@B|G#7s;Hy=`4`SfgU34H$P+EG?3}VHE0wphJ=qFoe5XE534$u zDM+vpp7pQNLF*;0O(nTI&OFx?s}&)IeUrf6Jac?= z^=oL_mELHLdqR8Bkegy|3+>1uX`u{{_rJ|DEKQF3d;7;a;rg{HS#uxp3*Tx0uMpHC z-10J{cHfG~180=qJ4ipm92O>2gZ&}HHtriT?^9eh_%r9(`}TtY~qCG{DEJ*wE6}LkP|n8 z;c?#3S6OQfhoS_mU!Ct&$Ew02o5pp?=X~5Ss~JCMjy3 zg;sNGA-^^Gg`6G%Zam1Cm;UcbhyNpdV|)%k72l4|B{g6*j?;rKQBTNFdIIQ2{n>A{ zMKK3B-@`rL5PFWsVJbr6z2-cIuKaNj541YNkYG75`MtU63S0hfS~|&-IdtXcf-u3n z-51r^%^7;(Jf4FV0dor>#~&GbQe2W$;Wp3AhRb%vmwR&pFg$U95=$4Tg6K|dmo5*D z&j(&6Sk%-$V%3ir{v@&xbHTRi4K{3>y5q-9JrU`UTV>FuA!L~7Ea)YaCbIn;I>-_; z5ZB;(VX@cIo45t;>&b&P2p}d(OzAJN-uKcvJ?ON5Kh(h=Yw6)tUT6rBi;x46nr-W{ad9jkaiRn`bA^TE0{gAWe?%maynvIaXf^?9usI3iv!@%2mqTXf7A zXP1GKmRP~*RJyVou{h?X7>@w?A0F&)bmnh95$x4gPaw6bhM}U@&zvaE6z#xXt{qqW zn?wDXfvP<~JtQHZ1Z*r*hjHR+H!*o!T}}N)t_9!yi+EaH2G2Igibeyh3L;*!SW;ZI zjyJ;q(e8QQ23+pzF$FACnKGdWzGN7-?$^$2ReVM9@{&0!b+Ow32Y#2IljL4`a_z$E z{IfxEudAu?r-jW1DNhMdW8g%!D9;)CjRLs(bH7yX-L20YA-z-DS-Ks2=e32e2U1hO zZ=UIZ#rk&W5AV=WwDJ}Amt7zEf1vZ_4TV(j#<$%(`=)Kd;I4cA^(>pmbpoax!278y z;W6fMYPm0j`AO;f*Msvu0Q^Ydp2D}K003j7d+KaFtiOhe{v_u7Nvg)FynLvezE|r~ zJ^TD|olgo_EEl**wbL-8)S)`BTMzQ0!)z%!sLQIlDzk-HU`X1_q#5xPenStW5ZLy!z%F5OS45MgQqo0OoQg|&Ek~wUz9Vqg{ zD9=$g#mJ)dFzxoQOM63-_c9nk#ZtLtO7kKGo97pM>%7+hY88*^Si+u1X{8tB4txsL zy@%Qsp+r~t2=s@8=Yk5grthH?xJ{v2Em%w39NXPSX)yO(d?O-DTrV(u3OO2smv4Vd50b^WFS$kY|8Haj#?4LFN1KM zrgtasHuS|Du^+(*_r02rZUG8p<{Yx3W~2Fa%5?haz;25vjVTlzc5=1Fe6YK112`TD z5WldJ_~dNvECS^#UB#)>0+i~4K{e03)muFEMlUV(|qy) zut5B86z>lvpqcPnm7SJr!`n;xsWR0(POk!+;9m^y{urbTqK|FJddd;Md~_?XLyv*? zZSxI_xGv<4L`sW;lLxaq!EY~#^ahHfujPp8+CTu)=#Row+MDMBoCtYZg{I+vwY(M`y{=pxH_y3Mzg1M@#9y7JY zya%Y_i=VcV+peB&vHj7`extQ#$dOn(tj@u?pyV1+PS2FrR2os*YFsfA8J^1nW1P2( zHR6byBN;WtH{I;MpypsN+u`_%&Pwdvu_mU;JxzN_7SY-5hnk~S%~-eWTzixK?L*>= zHAzR@vkl|iJ+*Hcz05(LhEWK;xX!*^+i(%fXFLR!8_!c6=EJwnj^vxY(_y`<}cjpc9-&`()2 zU4V!27!d4sZiAFn#t>4c8Wllca4{k`3As5oAof1!IVag6b`_{li+1Ha)Mh) zfBU4sNKsN(D_&x?vy#Lw11T5_F+OMmWR*?h18=vBZ@3w z6CEE|(7&s{0^&=r-Q-s!cv*pWfLZk$&mB3S3&qZ$QH>hkkkweyi8igNZ4p1SnuS0- zS1xokqeT7=$mYgo-qr^M4<(67WUf_8Bz5U$IFeEoOUDyFf*0tyAITN_OWYJyZBqpy zrQE#~1|M=#q!S8z!ank9toCJowy-)elfeJ#j`)gg`L0wIQ6yVc9;G;Sq}4rQ{Ya;u z_zJ;(-GQQBC;P(ub0L1y)OZYJxho zEo9&nQfmr79d;qpa!4ThcEpzaN)rc=$ zXc?tMczkwccXT{MsU~*1=gYhtL7-Xw7I=iO` z&Sa|`n?`4K#jOxLyj=u$nA$yQVRjrEV7MdEkV1E~s&Hj$f-R90h@(z-Mq}(|mhq>n zBa70!V{5c+Mze>Df@~D2eqQ`Oh=1LvIV80Kq1{?|UiZ)7Te3 zW;b9n$w(M*2M(55Z>mGsu#Ek%o_0a|vidEC4}wBeFw!VFZeXS$z_TA+sSoS?8hEIQ zRk=BP2UtwPi{KBFG-Q0nf9Av{<|;(CZKu~la>=l)C=Pk~ba1hKvvJebYKzn^x#|GY z9Qgy%JbGfA71nQeMfxJZLu7nPD3eYX-~MI?ZS+kKR5D#{lK;%i!G3S*^VIq#KG`^4 zXBJvs$-^LDk%|x}xVyC?+WLGrWq0v40w9W5wEsz`%^S5D`K`fB|MVt+Z65gQZiZ=N zNyQt1)p$@h=eeLqbMqzP&PJ7MMR}Y*f*bE@MfQ<^;;V4OAcy*o1w48s;pQMaV-4Gf zqJ{I{>v?&>aY28y1Yjy#?XR!S8U|S@+#I+BX1}DtZk-JB6!JPBch-K`|23U0KmjAE z&e!ql@t&;P9$F6>4?FNWNAQ2_auh%p?QMQ&^^7Q_#1y!t#=A4#OyjT>^T`1-dpO9e z9lB;*fp>ninUb19M=O2reul=?B-8n92;Uus#}&G4_eFsfW^C0SuB)@{yTBzrifdvE=y!zB84$cYOT zMZgip{`YMG4wKEd2X|fao$j|M1w`fys9Zb=?5$DVeh$o~Q%^A9*Jt0dpy$6IDF-MD zEk)`B$zsXmci{Du!B5_o=MFOW?qoaE1ghDKaysa``Kvt`HQ1;ZujnfiP|rnwD6z#m zx6C*uQn>!5X1>XS0*E+Yy^r{-XQf4LT%+} zN~;4?xuPAw)@+$QX`#|s>9;u#M%kA|PVGo-51aSj5azs05bethqAM#6baTkqk1H&& zs*Dt{Hn@!LVSsWbN@Ie?uPCE?B4%>G=lDzSf;4K{l&3e2oVsB1zbBvnPXLLG_zmI_ zNf2D;k>*E2Ms@) zRkM)@fe?fUGS0t7!A}lMKmH)cKaLk6XyapcKgqpn5W}`Z6(6LUrN;DMr4OjDvJ;G8 ziI+$q+ITvt+U(~w%figlE_odJLXSAjsiEEO3+@#Y5JAY zY1Z4AVW)k<%O9r8iJEyPkmwd*i39dcm2cZ@fdF zy)UvI@kEoU&HHwmz{NmiVAT`P_hWQJhMA3+vdabbU$eha{Q@s{zdPpAA4$18@jboV zbFVdN>-gbTvS@!Wjh9RrKR`Q?^{ja%LQlm@r^uRf9I}_%JdcKFOlVR{weqE^f-9@L ztH3M$cPoCv9Q^H<=8KfE%{^ks@*^1e!&eN%=r{PZ&RT@HcmV`VQ~bA^&O8+VlDOQw zI7@A72G*hyaQPiZa+7C-W)BbLktpW{Z6@^`4oikA+GWDw+Q{+BJDL7xk3pX-hDx5) z@sf|PbSYB?v>Kf%`VThW*qattB{}QQcbKCG@Y|!0t$HFg5b8a&T%fybZhG3Z$w|uV zgYVmIOVc3ImsIbhZ&~e6akKSi%`J$-v~RR0FFX zOL?08p(%Wc<9O}!_C0w@80m1?-0&$bgsM-ZBmxdI2ePKrdkWbHn$aCn#W^ ze5b>EtbZ2s0V3<4P)=8V2M}?ffSxLfmlK$wwRjc|E+0^U?X?{OA_bt|z}m`p)50EX zzo-HaaTMH53^b8jWyrT3_G>H9u|+Fw4^c>O9vU%Ux{0*ph*#_A$VyEubrDZxT^BVY zT~o9$54#C#Ag%I_jR-7Wzs8Z6P?|YZs~0*XbqnPiFX*o(x6m*4J)kV1qzHB!bfy!l z%zok643=89+Ml}MV#5aklFVdvEnV51-!KG-p;twF97nfqAS9vTrz)=S7s~x4F8=>= zIx_t?$lIui7cbQQW;TCIN;x0@AS_!ytxvsdYDy@ehcSNTc9_tpe9nMdeSg-tl3Bf& zkUbU=ha6*I#*G;xI;>fJyute9f_K9_pv~<5JD6Rz6UgRTA7HF5a71(IV6aQ+3TxM3 zn6P<&!bdgD3f@O%Q$(?+*qqF>=#BLQ@;2lGe&0ni6ztvvt)8VS2NHDIx)1fAr`Ps> z(1!;Bq}(9hI?B^|P7k+i2<&rO+<8%9yS{$mQ~Dm#w_{;Z9!DS-+)oH-N z1`L4~!U2v;q1rH%BYu-XHeRe8XKXj_OWr zvDmCQ1@PLMhR}G~Y{r4SMTphaOiOTdXW(O@MV+D}uo>bsAWukaMhWR+j?^%k_npZ9 zo!0sZ8hbk}FcD5z@hZ{mUjTZ;&$K-ET?pwZ{}+@!!`Y>M>SirIfY@WO%_sFZN9K77q2ISe%4yAog`_kjj?6f7NnPe_*Z!0 zn@=Ob)oYcGFX@s^EV<~DDkn)AFQxo!DD&$V)5iObPeuR9I)0;l#x9csKi+=n+<&3P z|97E3vW{k9^-Bg^{X${ZOo<5Ry3oy1k&Gm|+aAeTIK2wMI`Wu3GnIj%nGW}F9%31W zHbbc5kAsE+1yC9Gi_7MNl;eo?HsbUfwgPuT$vH=Lg1eoQfXALwmFH4on_CzHm;?u) zea#^4{kgFdJ_gm9l-bmS0jKnHjLi5+O;-t5Zd_v3T_jDd*WE8B`X@IImAda*HxGq+ ziA1r=A8ljROfFa95CawjhQ>VACSbqS#ZWi39riaDOPk5finCwiuc+-{ z!IAuGrL?<~yJGG@khrKm@9@|kTMq1j5G%=Aj0_jxM%X?A44l45?m|DnXzS_vY$btr z=yljF%Y(Yc7U-@CM9N&BB+bjm6fEtVE-3kZFJ2I;l8$g>L!@?uq`qQ1{QC0Y(`T;B zEuE25eW;N3I3`}QR>H`g+?El@a`{%+oe=%I<;9J~^+I<=J*WeiSKidx5)6-@F*uI_{Ot3|=R;PbvP%A~<-U!cb&;G5 z&<~5BVKpzGC=JvyzbrS|c2M5j;af=#h= z`klD4DSqAuNbo}#F^Vc@y|9PkW?gS?2CztC&BO3n#gR@1id39t?i@-3OOC&0Ue!3; zxJ_mVd$$$)H7&rGqY4lub`tL zkH_BnJ5rx(SpW8B{z8LiWDMMBJ%dTL#msE+sba&hqyYKJo4Gvci*eKVb|Stz5?}0+ z4U7JKh>Y0Ht4dp6N83uF?Yzu|j)_#{zgXx6+C>sNc^OJ=JK=P0bDK_%nCCoV0_EVZ z-~ulQ&U2Hk&0-iTNv`~~9$R5DOSOd+J!Usq2%oxQ*(5_q~Da6-*A2xh9X_srR ztX8M|qA1fbabgtS#gCT6jOD?ifz_^BT3DBrc&;-4mhkY4z_mGGW;(+@PrhVkZKy2S zjUxG~E2Q_IQaSLHMEdp~VR&Z1`Ja#U`@}Hj?31NFo!}`+h8+hjUMR!jUE=YKxMVm# zj95J$(!nSQU#7ny`sggAtMsi@$n_(H4|?>(Xz+SY~VADM6>pL?*|xc1l)x;?zf z`}skdM=lBiVC!dyC+mDcgf%(Ok7}jrwg(j&-}N{` zLwC-K{vHQ-Fb`1!Y@zEnE@FTi6+EuL#0*OlJo9})wDk;E;6K(x*0H};Z48LG!G2sT z`iG&+1EV^1=+8RkS5|b&C($qDgU3klr91Bwh&tt=F8nY^_x5Zvb6hZnvOM&^_!ca2cOgB|L+W4vdsY;*OafMululR}G1Vc5N$lp{`#yzck2I|;t{ zDy@}Z2;fVphh;~-8F5PFXlebV2ACP?1A&{9vU}$a{btaqiY~PN4ndpJ+nbG!cV)0! zgBc7g(Iy+nz?mATJ8L(DwazDr>K38Yqv+jSofF-%>E12X(LUfpwgijbG@$4W+3G>q zVE0gluElhe;Pr&Ld>XMvx^bsGB9OaYt!Gw|MV(x;RF3?-y!#Wt$|!=~?q1dHimLf# z>cVr5_!k=npAX_!Kz`un^}3Tf5RFlY0_`D07u@jo_vA_#hZ=?Khp=euzBQKT1j6n+^q0YR6DeWSlUXUF{VhQ=w^{e8Zm5ZTj7)t)Uy+T zl@{9yHZ8ppjMF-VN;SndkEteZl>F~xwSCTbiWDw_(rmgW$8XEGm9qZ;J4Jc`zhsq2tMsG&{KqFOXtr@He%$b-zMiBneQpeb zQ@igVd%8*yH6M70tzdrbugw9&V}xk+8b-l(255qchq(fv(H%V9^ahIbvutBsa#{{@ zA;;AAy=VbWEzc$iJI3d+livB4@-R+vLsR=pl<$!9)*Hz!C;QA4m8=-oo@w`ly_%wb z#Q$aUI}x~k!0WdCxd5+N!V`hsbE^<`b#1G=X4K;4jQ!G$<-2ggA{_5TxxW?Zh0fjOcZ6irvRI%6lYD1+z`qTwxv+Sp&63gF1CFTy=@pn&@u5Feq zzK8W}#a&JQN_C1JGlJzbGnYMXur(r04@PS(>C~9b#T3+ z*7(l%6@1~euZGReWL4;7CcEf-lZ zTPi{q)#D6am*C>H6MCZE?&{c#wlH0>*1*y2wCeM3NlXTVE0$r2F=Stz{Ci5S!;be@ z;l(Gm&9;+KQ*f%A7 z%P;v^Wq;A#?z>5p$zd3h;f0Ms^8UEk?otuNaY7chIcJ+s`YJL&tdzNVv0Vo~>9z;#mb@MaP11!2h}&l+ zCC-!8PbXK-D%rK?8jwDOXuiC+6F*aml`3C+A2cyKnLj%(lzy~=h}|4t!!)(PSe(7K z&@_Uh4gXuj{oj;|H1`>m9@pxMs{UpEaE|f@0^nJQ!;1l5T80TFkN!S zYhMZ9$mq!;71gzIjtMD6K;@=ssjZgFdE6g*)$SaJ*lk}L{MP)NN37X{M5Hm0KK6n1 z;f90w5xa%XjWy}KUZ@boPN(6Et@=7r7%r2+rVBnQJ!-yhC1lwxIjWA%Kef08CC;*) z;0ebe=QJI^#Pq`y?`;u7bK6O*uWIhvo`nIRsviJVvEAYGxi|$#)K-~8#&(=09{0%y zQtb3vrn=H-3Y~K3^XVREI|9f_7?)T>8rgQAyd`$-AC1HvxcWMH#Tmx%DR75JSay1qeMrFCuM$J)tsN|OD;aI>8J@k14+qm(M6|DW~tbL^IX05Q{R%u9!T1I`F%KiuOYlKmx$#j z_EF2ON$fW868juS=0>AWIyL4lOy$-K1UqUkI+FOzftS;fj{8l0{i7{0s|z_hlUAQ;!$lm*(N!S!^e)!L^5 zsY53iqH=Yd8ur$7G_-xU)s+R++Crn(o3m^H?HJT*7pQRBQV)D6`TfG+7v z*w9p+_9SJ8%kzy;pX6%-5lK&7YxG!)bdiUh^%Cg1<-ER2v1-LbV?Rfgo(FZHi4Gpj>%PO)_X{woh3>*n!i# z{Znm~Ht61GPVnKXwXM5%KDAwX@tytWsNqj?k#gZvvVN%b6z5{!FaI6s{4uRyU_lG2 zG%$v+TLyoif-QYVo1E))6&s1EYp&JbEh{DL<_9yoltD|BJ9(%VCUh=- zZ8EtieEm%RB>rfE1)*p^R<5r&Rl7xGDtU3yql`G|Q0FTx;Wd;l>e-WZ{^(KR9?0^_ z4Pl2japbAPeDY}}iX4dT_mA-+zvCpK;2Zh`!Qq*%XzbY3T5;LJ5q?{|>r#N~??JN` z`@IiS*G&yM&nWSnv0|3gz~4)bAZs^WjGT%|0d@j=O?|ZKho8E;uDuSdd~jB zOk@lSLHCH0=k=r2y43B3r|RfX!$X*j<+R$Cvtqd;Ju~k7Fmw2lxKL7pzH=5SG8wT8 z93IKKZ!nrTX4_BdIfZG_q-pR)Sh2oBQ0rXXCfgXz2=#6R;ehN1W*cvxi9{u`TEg$j0<8A-cRwUsmHAV!A)lZ!(Puuju6!Rwh;qW}**$TlhLB9%@vB z4Y7+GU63Zl6dVX14P9>>vZbN>0D27Tv_{+AjUS&^wvR9%3DNYj{_5clkaqFz!D+F5 z7TY7gu!ZUIVA*(ji4c|>5gJL2x$Oe1auE-|k;7sfnv>&u+xTJsaXpkV$}WBBwjUfi<} z!6Qz_{h~DoNb)ZnrUgQP_E)aLd&xn#nYnZGkxM!qr4YOHJ60T1jF(#bDb#^>wmFw|D!#%zv56+FH00Jm~}%a zRb!yzfY961nT8KoI}cjiIg`$w3!Rh2{SWaAyb_^m-d0o@BM0RKTLuM-5?{>e>vgCK^v68jac# z5cA@iEQ`@If*TY3GlM*3j>XFSBSY^Pl6e2H6fJqBChcJR;3ai-gEQ zD9&$TZF+90AYiN6E}M|0h9wWYDJX3NiCEfn-j5yEI3&$A5Ob--Q9dN#kedm# z;s{1>q2ii3Lo-%Eq-w$AlV-esS;7!8>5JZ#uMU_cidETG={hH&@2d+xt5B!kCvv@hUG z%e6Yem^$7rGa5l48pMN_^6)OkfLRoIN6~qgk22~_NIm7XQPh6`+5_ktRrqhg-%0@Y z;Sj8!1BP^_O=tQ(GV(o}22v=iTM+~zvC^bvA1hkUh`YwNI+KQXwgL{pTrMg?R*GRR z@+;@t%i7p#s=_DSYx+vIx{O(+qVvP738!5ua!pknMWSHnD5l@apn>OY3sMl#lf`Rbt zz4D8=QY=gh3x@EdeGfPy?3Jp=--bueSKg?K1|VmoksB4 z=-X43U8yi!b_*@yZQq&2t5BHtEL0+<0Lg#HtThg!&RA?hj3+zwuD>{$1jRI)6!1Vk2sy#aa(h zqJF5X>|ih@EAyr(oL~5*ul1&f;u;-B<4IOP&Xqm4SlC9hoX&w3P3IcDwtK9oE@)2| zh=vRL^PRKO)7{5G!1(;THut#O+jK0k!*s-y^$)_hmR#2>SHS7^SkEP;#8zFN*1h0j zf75BLSwej%m7XrpY5;{kEyx1KeoRb;;E7m`hGHYvY(1v>!Ne^K?|N+3crMp3P~;!4 z;np(%bhmtXGa~d05AxADW1AAnMz(OvP_(ZU)^|KD!drMfOGW*Wk^^SuDFMYZlh+w| zw0mW?ne2*hwprzgYeNm_t3?nqwM7RXfEHWyW7^e+<8hf)MKZi3QdVnUZ}~`+XcJ@K z_UdAiCpA&PjKr%)21L0|$oo6j_a)GDzH9Jya_A=*aN#S2_+=^tg?KDTja*%xm(pV` zcAx6VC89aAgYik`L&#@d8T#@PVcYAOyV7ekc50x8Q%3W58RpK_U~fZ+zj&rU4Bm&C z;o=?@`&c=l>nHNl|9rDpk{0$1(xc;Do?cq=z<&jhKLKeYDk2u$GltT+f}L8n4ra37 zyHk)k>}zoRpx2>yCOrID)0?BxU0aUs*w+x&v&ZCF2J_P&JR_w!0BsIdw@e5E65M1Q z@L9UfL++q6^nH}W$`d2EJ~WIh&B#f)y%#y~depM~WUr}HbI|Xd+7Mh%`6+F`1XAgm zxhYKA6%E<{j^?*OU1n;H3t@T{+{$f6lf@q_I%Bymw$D&b#3D>jnazGsyt^0#P}~5( zK=JiNNJ$@Rz&~_xwEUM;8pFum1Mcc-_+@cgq|a{fAgRX=WTD-!UdJ4_oXI!hD2@h5 zTppOqbsHdXW-SFkkvF%O-q|IlG31FLdzL@4tdP9(}0ASbqj(BiQROin_mq1s62{$odc`-;sfU=phV3KJCCwiQ9V)5r5gR?KpX%TxOD0v^@9n3{nyCj^v zBg3 zgj#@)=bwv^wrWdkhx9!u(0BU`RBD8RvyK@v@gzbs_mS31?8GQB6uY~vf_Z?+gFhb zR5k|Zwi)Kmkwx8)M7$MuvsAB0an@dm89$77Y|I!nh$;aB=C%@BoR2}2=yAq*1t;X4 zjhV%_#hfM4!FJ@x6T^vP`<*?)a`&8m-U9nGf114i&8x|VQSnHo@82gS{$^MMZ)lb7 zBfnQrEj$Z>eA=cyct|lw=Pm?!`zn`3ayF1rk02eMn!^{#uG!Bj3QtgGAwVP582;Lh z|NfWa3fi-vU108pX`f&yxv{E?Zw=!%87+ISno~VABma?-FXe!@9RMaTDxtM zC=-tm(Dx{K8r)y6lSc$2mS$O4`W_1t%i^jnT3?perQ3ehI!+?$Y`g$ee|bqP&pW*#M;p)L;?1 z5Zt0Jd@|1@Q^ssA5rZaP(U|^bZs1&m-AhwHWaGr^Zuuk~O$WY1Tyu=_S2F3p?yjiY zpJ)!$q5XBADZ1IJ{@HX*m%*Uy6RQ`1UZFt6GHJ@;yPqf3nF(Cje#U6+*kQi&-9s!y z=PPhx+*Ym$-Xj^6ST%@N76*w$^4G*j@%0!f^{ATck$Kl`>GI?NzKDONa0Jw23&m zHr?&)scZ8{4X>mBr#PUJN^9ihoBLSM``3vS`GSTkNySJjD_c>)+Dv2Aa7Y$MGZoq`p$jwsr3?OY7G711t%Wf=`_J!{${k1ZVmy^vTLY9%7 zn*93OOnVukv@aTBf}BfK$eJc^_|U{lMG^xjik_b#D+8~`NWPEs*EW~k5*dzZpRpMa z`uz?aM0$8Ivj&5h$Dn2hB2qkeG!ysgy7(Xl45o@MnzaI+X4@EEWX2)y~`dfHlcS|))Og1vL-|Fq(B^SM$7`#R6S)deex!UNb|%D|_WcpMl#tZGtVwtNf|x56(lZPaByUJb{jXBve;yBMPUE?h zH8V5*3w`8oJ`yU`@S*9)N5$DC-y$*!Z{2*TxlbPRZ|4gC~XDgbgzm>u}(thEMb)G%Z;hmC}9RF zbhd9EQo7o--bPM6%_sD7y_wV3MSqA-Z$IK{vUfs{8ctrco7>5%vMBZ~`fMg!%b`huw=-{|CgF8kZU;rEsk9QWq*f$+ zMI(2jyv2yLS*~e`BxMA~3VS`(Z4AgMx4k&NcI;8M@PcxosKw1$vY!#rdkT?Z*^6*d z$tq6GoJT{hx|s&PVF5kMo-KwfJ<%;ORI(-z-e8p-9g$GVNh|fli|$4ejibOj6;7rbe@e6;MEdX@2UiV8#RnGyI0;or+ zbM-jI@w^HHl9o|aUFr7>;Mi3B10#~dY7f&ty^K!`w9adZ;5ddaaX;Ms{x628eoL>F zalPk^nuUF;&fd#}m_#_l0v?_M8mA{f$HRerDvt@W+!4tiI*)D>RxG02JPGL>pmPiY z-@%t|W?9r+1x)0%);nuIyj{gAj%gNXg$SF%p5KP6UcLmdcZQl6s(uUmY7j}3$4QxjSri# z#?fEi{y(PwzkkhF@Ya_#B{;le zVxJ_T)uGHc;J9b8RJuLLeY*bIaVuE)t5rK5#(6+n4#<$QFo)xiv~SjO^_sOyHql1w zTqTUf>fw&z4hHtWpy~by8cC5r-9cEMdo}+{Q`&!04bMF;Dz8zL{nQ@z2{&ol|FLf% zX}bKAEcFFlJzg5j&!<7=wJ_69EcJ4IT7h%quJDraZ;fj@#0lpAt^R=e0+E8K)UC-X z-L!@(N#$To&WM7?xc908C0jkr6`q^B_HMILdcbNX^xQiVOWj5u&>64oFTay~P<9}$ zPFG>zis~@rY=aFI)o-)BmDf7ijEGP2&nOh=wRCMCtxldTJ=jvJ5YZ{5be)}xaNF7o zpalfpmPwZo$i+G2?`tpN|MVJu^Xa{MYYjAGDR!;(&@Tgv&!uyJpbDjp;8^rcK#CM6 zY|kJ#v1=kwG=j1?sXBy=)SGBC-K~S~FZWfJbC$9TaLs>xbA;ho2;Q@$oqLEH_~sKe zgSExeo+JhL6@N)FHwO_W@)?FR%FWCfNDWu=hR=^EtI!6nUzK$>GfMw1OF#MuIRuM8; z5<%XpY9}XR_X)Duy?serlR3U3rbD(? zQl%Jfva@tan*Kh84q<73y9c5+p)=d}E&jOtR=LcVzJ&yLr0VHxMQ%6BT{V&Q^{@=; zI@^hxR-<1ftnP)5A1(83wKZ!2I8MJQDUD=}Pkre+MiBPiWbeVP&+xe>-A{eFmN@Tt zwqHBLftMr|(wWVnD4XNIm_7hPOa{k+eyUFr7P z#(dAGLhKp|jo19d*PrUZ7vSAMdtKQ1#q&mS&$UmLdm&+G@V2BPHuFxE!jwZq;sjlp zLSeb9?UUo%C+?TvVQ;46&x_b$&ca)Ez!urDsaT(8l+Kx1@(7r<%0c!r)Bj`bEu*4* zxA0*>5D7s+LK=~f?k+(4qVP99mFPK)OMC=x*MJ-}^sj zo&P#Vd_TNjShHaG@Z8V7_rCVEuf2CsBlqCFl+t#%Tfsxc7g--0HTfqPG})L*K2F;g zQ;0t+LCNumMf~rm`~N!9!HA~hJ(TW7ykqtsJ!`pg)2?>&n;&+YRz7g>KZNj+;L}=W zxSEW1?{W(ZoZs0rf*|iGysP#!2AABeN*?H_jV%)I?9nXLwGzS;nBXWM?q#w@t34nP zSQWjxAlz`PRJI{W|G*S75rc$2CDy-2t&$hMs;K&r;+3^F65Rhp`lBko=NEDB7Kw;m?2aaV?D zLODK(*Ou$fG|ab7iVBEuCYD!J$eN|pl3><@b`s}UP^R+PY|J_dF<%V;tna+= z+J#xL8>*Da#9l^*9wA7$1ic|D7rP2HAKKvxF+<<*lS%9NK2zH|)s10$rhFe_WxATK z>3>_VRoeyHaP*%0BqGY0uf}FI7jUwsEo88_jClHx5_fodh3rO)iXj4MQ(?}^KlAG! zFwl+UOBWR+3pG39VkIrnih<#q9K;$upatEMT-`^u%=x+{S$zq6urvbeMh0x(#+CFS zN?bBwZd~`^&5tEr8*aE5U;Gl69@{LnivGtS0DLAFchD;FJw8ujlD zC*PtsPSxn8d==wOL+z3;UPfLW_$KyM3Qnwa*Wh|AB4$=Phra%__Bj6W_R7aZroQ1Q zd3mVRaN7LqGX4!F`cick|5(AsvENG0#^59}A41T#1aA$h+)#k5{o!-O8F$F)YeB}fdT|S{d!~M># z8$DQl<>cRRXpYN<)GXGJ89ffYU?r&#wHq$J0_J8&h>K;?K{Xw@Q!@8S?Wil%`o*cZ zy|)T_tX|%`qI$-eIhX`r0l-g$fV=n3E34mQsYHX7rh4+K!$n#ZNBcgK3;d_EK_GMo zb*kT$hvaqFz-MR3FPZdrFVT-h2ZK{tD&f=8%Tvj#DRXIXzr}O*9_5VX0ZFdh#?vGk ziiWJX;cZE+ync>jEohsJOb`2 zW3g4ofh^J`rHo%Oi(sGC(M@WuhPw@>b2}d4j!7Zo1LWMhf}`2HhuE@(#A^+Au2li8 z06Eb+L4t|qX7kXJBGx?Nm@8tkVl@=aBsps^kJ;?B*mc5L~EWqM?S$s#4Vhn28X z$8qYX%aJ7)bj#cGrz?bgRbJF{A2ws5`GG%s=+quPLV!cJ*kIFJGuWY7IshKwME%;= zR!aGOdk6R{=6x#YIu1sFnJJy>UcQ;g#&84gx-%oMT+7Fpu}~ZHMkx-^wki)HHp-Jz z{3@79j6?N2c*46sW?7VT^}FQRkj1exun)pxo>g>keEcydM^eX<)A0}P`pXGHfNCOf z{m9?z-T&78f9yQ@cXGzasCL$vt-LNWhI0c3y=Jz2UUcDFxhw0R{4?rMbG>`a8)VxY zz%fj_cyFuH?ddXSps&+I@I#Q7d#Bw6K>B^LxK7z(h~8;)*`h#8D__~e$EynDX>@d}mG$*sPKk>&_)6(fLreWY<@{d350&~{_=Hzcc2s5dBQns}olD^Q!%#F+ zw$T+`?bVpxfJ*<-_Td%HiD8Ieeo=#F!E({y++M4V?4qKMr|QPYYgB`WcRZ&X z-+h&*_n}(+u0u2Vy6dnijPxFZk`d>xWy5^!$xKo8!=su8uz|JMF56X154tvkhSPS7 zTvYVN5elm6-A3RPj^1)0@Ga9a6)q6I=Je=K~CiMqix#odS{po7r_{DM-cqINw#&)|H3^ zl+%D&N5R-1S8!W-mJ=5v+Hs-cyWusuKxy*K)fx3W@EX8sP}-{QB_gLiFTx_WP=M-z z6a%U@FrNANE5C^@`kb1S!IZk)96bNB8Qn>k7g)bVC6b`o65z@yKAd?}5^kTajemdW#7LpPr&~v&d+aDQYC3b*ftU2*7pIaBN1GMX)#di)#XE2{ zGTU`UhM1~zJV_B3X)r8wCKN(6!b`GN+aE7kydkH$BXpB%lmS$NeA}ac(v3>p^8bm8 z!4^eq`~Fy|Q#TV(OOnhh7T62q?5d3sxVtxV_>TI1GwApnD`BlHPjiojk14+z6KSk8 zkHZfr|A;FXIa?@F1RMIjzmI2J(`$KK9DP>hlE*GUp-z?6WHYub*qqIJa7wdp%b+3G zgk*dC{6#sSiZMqkE|IRtntJCf9VyQ?c@Q2-{<$=+)>YzVfwqfgWn3+QabruiCk1J& zG|EC`oFaZ($b7F4lb06#;LsC;EcN$#J~p05H00jlp)}(HwY@iS_W)^%QP&^_g<${b z^3HMq3c^u_O6E;YhzV#@L(W|1(JqP z#FGM@3KEHZqRbYXK-s&(-d&ohum;svS6>UHj|on)n!sAT?$h&}N8qxODF<8^n+m^? z7E^2nl=N`;p$;GQ*R^gRk{a4~zR#(`8FFXo^R& z+bH%gnP`rsAkdMozs{Iw*2_%X!@<8pfzs9GX3+5M7+ajAKBKV}0h|SglYUU+P&l~8 z3&NzDa$r!LSNaUPsUv(kep3$zpqi5^L3Ncc#r`ps+%!fjc4w_=A5YU7$rDpk-^>g~ z0;61Obt!FkcBLUB`MO9sPR?6#lChdx4>Y=^fw1}T0f^>uCImWSEkP> z5Dq5wfEK2pm1O%mgmApxIbu)R`;LX>1aw0Vi}z)Q69^VK`)FX}x^ZVS-WS}1u}$*F z7p;6dLF#*|!q||aKoz3Uz>$EsBN(HplKYTcdGT)dYb?2 zj;fZ)0B|wt9ce4uKeUhxG^!>vm*LdR z`GGo+1b~*?lfUCZm#{r*Jf@X3T{wIuJzI_ST0Wdwy0gg+7I_5D!Wrt zwgHMws(&L`NgaQz>WUU16vI^r{M1g7WAIy|h&Kuw21ZJc1mE(IL1>MRMz-Htb@!Tl46zf>3WymeFpQ{%!crIgbl3AQ5r1>IF?UE5J!?tl7pG+c^-7rp&s-yXG0+chO z2d8#38hIMEn+ZQHEeFaCJf#aj9Ce#}MD8U&Ta4*5YV!IG#C_&752=JQf#srkYGnAP zjPu+@5QP3uu>&Ex;s;!|e>TYYXFxS7;n_?4;6Gz&4NO2|@8zrr&>)tZMLITO|`T3N#PE>?=hEE~|^azFMI(s9%0H9z~Fk?{P7D+@0lL zUM;*g<7bq=>^0O3ObGMYH%n0gv;8Ow;}`rQrKSh7uB^<9z5EMH;kPCtl8%Vhu0ZPP-kTZ&*$biOxUzk~%!F8Dpz%&9iPDuqKxqrq_dhLO9m28n^llBt+$V%yg9KSA+mM zheUF8G1;|Ex9;%Exa5f;DBon#RIdd3DIAO@O3Mo}{m~gr ze4l$Qmf@}k&J#DbGV;=M9dhpRI*LWyly+80(cZ_`rU0q;s8(5Srf_Qf2K_S@KrJS? zlJHr$-D1A(gvO2oByj6!Wvz>J=u8(zbIbziJe%VzUi`45<6jayUFnl0>w9PyciqW1 zWE^hJ?MU^=6t|D+Bvn=N(pLp~R9rh~I#>Gz;mK{waZR+FP@wkry$=-K=A>DM&TA8>r|UcXNQgrhk&T{TH3K& z&-1Ob=-HtaB0(KXa?HVapQQ!Ps@ufCuzKbY~Q|YOgAPdtHvmsj2PLF}0`+?};s$WfIh5W$~ z>9S>ycla(kB*jZ;0Sk#-;^5?kQ8S}9K%6Lm&;&H@I!D%}smTc*Jq909xX*#a=loD^ zk?qORW^_duiOBwSz5J_|7$VH2E!c3t1E<3$?0Hg9$#4E?HCWP%wD&ySg0??*iG@Hj=K8 zKT5m5RsR6@CYbe4lOImm4eeXyFKO?)TsQA|X(|%q(k&LXvAT*B%x59+(sm zcS2nCYORYnEAcK$Rr`av+kgg*N+EbE=I3)pPSNQpF=W(mXZV3`8RMd$6O~tvJcwf< zRv8o`<6!1?_*^%cBpZ;<3|KQ{!c5a5I>VM-teToWGqkx*4d`WfS%#9 zO_epa9brtJ?*8JPnO`W@>P0VjowZdd#u6I z{RAmKmv6Uxm>&*AgjSxYx)*!T_Q{V-bo;J*Ei>H1&%QV~tQ@AW{Z4dKJaPvl-;8$c z)PMfY;1^(omRdFc8fV^Fq6RV^wE0H&q+CZcq%)sHzp(?l&@w44J7yx&Nv+DNO{4DR zuD)Zc5WFnMet(I1mfh8oLh~E^!aqkBao8ZYF)@@uFthb~7!P+iWGVa;e>&CZl*tC< zV*FJ515hC@eB*u0OIPw>x$a=N>j^?I^k7-FVx{=1K@Y|zwgQANEQ9H(oOp92W8Mo3 zAwGVJWJ+lFcwec?dkg(B>0xv}c&|%J`C=iUVohm!s+Y+wJR>0wC`>r<%)-G5xRCI# zrfu!$fr4UcQH?WhGD+l~EeN$4RiDHB-Y<2jyyibw^=8g%A4hLX*5Yt(Rzr_GPxl9P zu4*j!(l-dEk0LAgsNF8A?A8c8@pIuEd88@6T1upmh}Q7+?WW3}@q=EP&p$~k_KI99 z-t0*CSlW`ik2e-+uz0ja@bget+wJ{exe*-P1{f%_ThTKAx7g)BUyYg)IvQ@*bJDhp zEZwp=_%(fOs#y{RLt!GVcgE|^jr;>Ur2y4*dMD8J?WUr!L#!YRd-Bk>tdaqDR%rN* zb5TWZ99gzA9Kw}19iR|-p^$`$B(Io3htSYa?%#$0Q!OY?tYx?A-)hRU&m!8;AwVle zW+s2GtTBdz7!4Pb{D6Cv<-kR$D3;iQK z=!o9;nM2in=*H>n!Zh}sI^?;QzW+Bd&0ou*oTb(&&FiMZB*(R@simmw=ZLcAJ~iEX z?~BV-Y6%DZ&6u>!wsCQH82;%{aUu_jFLzT`a~j>0bAEmST{Wfp zRPEXL)=aNr78Pym$Wsq2-Gde%RRA&PnA`AsUH|A#pFnV{QT^fR*5{tZ17K&YAmv$i z&E~diqpNT8xr_MK|@L#Wfse{jBlqN zv7TW*KUjSk3qBRqVd54&+>qR@ntd;blq*JJHX+*p@s2^9nsoxdyUvfNhDTJ;-QC3I z&A0~@P;f0{7Cw|$^sT)V*39Zq+br~m(X=Of!M_2QxB3%K_{4s5{ML@{&wMua#+-7YZ z*p7Zu`%CLwnb1TUX~;4Wtdd-Q2=fO?a!wEG33@f>Y|1idO={%-Q;bA2ffK<68EnSd%8 zke_PI)x7dmuz{+V@)l2aQ~>+^^0h~^b?q19Ci_{Bqzw{)NTB;V)?}^(XepmpBEqTq z9&#HpT)RNEXG@l5F9ogQw_0om(&S=QETG`JwHN>}=(}=lx<6H@s0M1TQtHn>MxeTX zVq6Ut=?6saLH}M=6{hFLyUM_0ja?!y%U5?_N=KW=h zD2f~Wq?;GZz@nZW&l|6B>j`=^FU$9f#wxX7^t(~Cvg*u-dxuYlijvmq8$Ff2n{!2e zEr)bN;=)ZScDZEG-@4=k&0RkK$moHuQ4Jl^$3oxLuRn2?6!XFUrtq#LL$w8*ax_m$ z@GzZiMyx+==0qgJY9n7K536qQjd3a2p#Wu5=h`BnTwRN!B+w({@gTpnz;yrjV%%U& zvHfWmRjTjQseoffad$U9F?dd15-kLRt^@!RFtCIqzW32njiYZ)LWR$Qvdbwm;@cq z#T!?QTJgXB^dJ9FJ%p)&ibHh58|MjYcLZN3fiejE%?uvTGSG1jO=pGR%s&rp*`}oS ztFpbng#KZA-n77_<%^yFGjvFP2TGOhqj4M*{8nCKskpN_ZKK{J5a!Ap;D-@6`0m|U z-tF7D#S|^+2zQS_kbi~xSXMJiQHS;dqiR<2C8701H8XCr8s63hb&z1xT~~hRpzQ(k z_x7*}(=RR{xNe=+Fk35@e3im2`SFbfSG5rb%2t1)HxuQ>Tn0md|Qg!XMfAwcqk5iL2p7@Ea%eMla#?j199*}l|24db6g7- z@D=W!gwA$9{cnr<*V6+gSDk?ViGkZRDGDJ*dO`q8pWS)%1F{@aG(d-9STK#%%BmQD zI<%5S-f*YVH3OAJvhO{K{j-!V$W!{A2R)(j`snf)#GEJb%&Eo=8ejF~Hx6D-!(RO=$#&6PBWcT&Ma>4czWbJKPq0{(Uc;jPT@jPixK-IQg|j?Y zx$#bUS9e^}|F|{(@r=O7cI^0e?~BuHN%PxyI%TRyM&91~)Isq%kJ`_(ejnEDXFMaXv)#~M$QtS9|AT+|*GbK4 z5gL}Dv}%)(BsZN?lPj}fR8x2n;Au2)L8-5#m3GzXIEgpIFc}sY(5Sp$xG}6nHhFs8 zZHQ^*T8cL+c2@k+h4!IB?-84v1qrsBg8iZvh1}cxge~c|Lbfjj6yKt$E~?!iUozjb zkZYn9SM#}G?kwt22{pJ%p+ipM>mFm_=R7dOZn((9`yeFFfX2}%CWzJR@gQzH3CDFq zWYu3Y5lKBAg}cm$gREKSC@h48BrhMXRJ{P^-9d%8vJ7F_0pPqe^18&=DI9gQ%0`rb zIS4=W1lA@)(9*PM@0UM0<$uh@1{2Th_PS%aVMx$_>t zN>nl>gXV72x#1Hu>02uWAv32sKSPK0kbKy?lUA)yqbwK@Psy4PtG#Yfvq>hBn>y(zof$LBRpMTNEz2mP+6!$LZ0kgDuyO=rXv*D})KF2uuZpyTBR^ZL@BS9)9RmaZN=Bu%CLU>20eU zolsSrUy#~MTU`ADhgaL$k8r3{&_)u)w7YD$YG3y0)nxiJ3(rTBq2Gh-`;G_qk|uJ~ zZZ-U(7N-Qy^uG0G14-RgEGsC>L~ zMB4A>KsKHBj{012!wHuoSH5L~N3oifCRUatH;#p0qZ(BYzCHR%@S~xdhZru)B*JVd ze-Nw&^a)cCWJ{>l(vD=-aU#vsQzngg0z?0`_`t8z9*UJAR zF&EjrlF!=F<=iyElDSfsGFd-#<8CXZ$&xp`B%A3Z#(h2~E*#WwB-(dyMuO)vb-S@b zTlP-tyA!~z(52{^=5mdko`r39`{uTd*WpI#8TtA(lPc)w01Mt%}2Uk9K`@M1&bj`zStM)U>)8Am>Tf34Vn})6D5sCk$_#Il6RFeW`LK#C-#p)d4xf(e#ZYDaC(o(csTb}*#92x%B zN}JM^D}94c6%m--2Lsu!xBnC^bNk*hrrj z;mJKpA^n0M*d^1-Y*f|8yWw`KBB&qnrkXXv$QO?jn03d?PPiV^H{;5$L1yQ^NT{u> z<&Em&S z(4qW`qSf-|s}psf>xEjoy7f>XI9N3{V>X)SO;LAn%{LsAIP8#$Ll)=fVuUiO%^xmR zYx{QU7y%cFIg@Af{IvDmX?od39qxtKgs0VKKTtSdVxIlT?Vr6;8`7j-SXKlE-jKkg z)C@rP4Uqj2psNXY%Bi7Hx*FS|HD@3p+tlSBdCw%UNY^DnzDZ6NG2f=@+22Rz!Xb+Y;1o@*tNBn0SPS>P6t1B-~f$)(pEe%>IsV1-fl6JNLlsCt`fTfu25UA)R9?LK;78GMk?= zhHFwO`*@x2$|4U?j9)~!Zbj-lK50$r8t}cvGU+_g!X{SBO{${zc#;^lSqAsN3pvgt z?U^FnJdNP-O2mK<55vjpX$aRv+%CCV`t45{K7!*sb4)eE;W0Je#d|VRzH{caaNMbD zFWBb&U-|3*xggg!=m#^}Yq_4dn7@Tpdh|Qkf@5aF2OzflfsNCFQW>PpB$CDhYQ(Vn zF;-gxp?o^B^$*-4@3jCpv)!7A5}~49kX-UvH%$Qrv$|AoE`3SPSdOM->u7SBuTMGM zA4)Ik(LX~aEhP=L`j-BdPa&#Elf2uEy|aCHVu~4Lrb4fA2FfOg(QnOW;0VeJ-k6%Y zz1`uMs@pa0(dto+x@HOnZQ}ffdg|4CQjHYa9xsfRbI(!?E(5!>?=6dvml(Dq6&N!D zz_EEYESV%RhOkmMZ>y=};Od|Wp}+f0kY6*ryWaJS@XZHEkDjO;_&PwTH+bwFzU-`> z8XYH%G?jmJ6b!@X$wDRp@h&-@$#GPLHr1n`r$6zMf{^RH$y}tm2YxU{|EF)%EZW)9 zD)?QA4R^-%DJQ2{>53t4p&55iiGZ287%9C!5CSi9ZNBn(=lOvYvcLpxmXDm0iNv=7 zD;Wv17N0zrYj}2?m8Be9mnb5AJXJ}3zS}JQaAERX@_japgHf06`za8wKK?EPAtyT2M5pZ!rj9#E(6}0TIFeSAun6*6G6bR5!n9;34^OpIq@c^)Su)|HZu+>f*fc z>(aM)cBm_e^Yx2=3sdK;ZW0zW7%UB8w9v{3vUq|mz@w1?%4PR!rV?NAs3ET3LZHGS zX`8g^5N!8gPNEWpuW!ha1xlUlC5C$$499ab!-iw+UT>6G6RZvtBi`wEH5Yqq&r1&e zP9mu7n|ScN$#t()wc|uT(v^7A!vf4v$>tc2?V4)~Z8C1&J{A5cwrVR;KPJ17xHJRH{@v-hm;N{3X$cO>XCQglvS z;^PZnGHz_Cw$@*b^NKNsdar7zjxJO2BCkw|NeVk;wIUud?yaPRO8b+!bRp`u4c1Bl zMgvdLQ%7ysbd4(s&&|n%T7aW7zQvvbvj?yjTXrt=kH_-BOk#Isjo-D+3pw>-m1~Q6gCxopB&I*6S&{=h;CBkFX*n*+ zhbS_Ys$^L`Us0CbE~fYZv^&J>Yr=eG$^qb3Wv(U8WJ3d;SQ5D}ZdmZYz43?fz1JXP zxvsV5W{_bogQ-EQgozkr|dM(YO0Z!YCfI zM@Gf5QJ`?5xJQ-!QLv!U?B86SNGnqF437p}b9{pN?U8zmx>pcvGHDXQjlSi zft!6n#IZEp>XFOH))1y2 zf`Rmn)L@ensalU-gu9M?caQWpqGxKl$n$VW&^Y>wO1HMGIYk=6fG~XdSz(^LKl|z8 zr_4s5ESRPwU^=GB+U=;@K;lDcC^9gT0CKDCi>a_sWgS7N>wa2}I4v5dB})uSovt{7 zL{h@%P0)GX>i2C9alC_*0n>9D8v`SO1{8c_HFWKitQ zR!kkE`titzdvwF=!?C_;Njm{(!~XIBx*>YA9qfSFe?JGY(;rEyvJ=(MM z2)X5Y!~4$BY?$t+M8qd;cVRRg%q{oM78%~^nmU9BP&iZQ&Y@l6 z&Sqo9>}3Blx6sG$6p6GL@nHHNa9n%fCC#}Eo3S{q-X4Z`uPGjyk5k)S0CfR1hpTw2 zvXQiYJ@xkZ!HVEPL|Jc;>u@VTVYP&T(Y|X+rEmWVJW_8_)bVU*zwNv^)=qk%1qKDM z4YX%L;0W6wpbo&uX;~457j)34^)fN-Nl}I>2Zw)d6K3%xPrQd=%4(2ueP%}&tv{DC zy5=3row`925F7|n( z^FZh$S7P>0=Z{AkHFf9KG1Qx<8=>MtGt|#tZbkzqQSEx4F#}BxAi5ZPtOTtr1_}`# zOE_NIN7yaVQC_}yp-B~@cyzjzw=A|^fB@?Wz6noxa* z5K4eLPvs2jdr=;Xb&54Q7RKv=2Y_k&fq zWrZS*{mHzTLgz#m{hxMyw%RFERC~$qrW_gt!;xMj)q6V2(Wrzb@Y}5e=2)%cWT7EvW-N zM_3ef?H~T0*NOR8G!mLlF9@6b1)^{zWAIr{J9IlXbB_e)3WiD5V_nK|2H2o|)BFI# z0BV;{212bqlH<2JI^CkSY~)#d@uNU4vDm82@M*bpvXb52DZY$)qM6kDim<7()<@c9 z%|%$qUg}5EQsl@M2XU+C46~Y51vE^D_YCtYyx1pJE=2(Ftv z4iBn34WU2X2l&t-&NIVLOYj5SZ&!CJaxCxM>nezAv2<=B+I~^Z8nQG^GUzH`3>o{v z$B2C--1V*Zof{hqDyy_>g2k@Pnk)P8bs4|`WQkSZSY!kIFV8FB{=7s&zx*Xd{-r0S zF4)dS|E0@eRzEQifTovs2+)`7?>gTBc+hJeomdH&QS;&lYMWM!@C#6R@ho~R4N=in zv;k=Jp``<=V5CM9G$g23v0$BWdUvib0x~#qk8b4y9otA+2#y=oCo@jsU0tBk8#nbe z6e(RWJ)1RNmffW`2Nx!GIY05nB-~tUF9{kP1bzdP*a#lWH9FA8>K3G(H09MRg}d-e zi0))vCYNgC)uE2Uab5D7;xt>pi|+^oSI_9yYap=oKq4rn$7b(IaG>q1evYT$74`XRAdU|M z0#SDQH}9L{aB9=R{ix^BH9ztj%SM*y2C0LJ`vYe&8!J453E3e=mV4xuyJ1S(Op~Mo z2GrOd3Ak)PNh*eo;)z;Tp4yNFe9fbKH*C2tQ@GH~ha`VVsNc?n)5+8)5o#XhY z4n;g@Ame#v^c2yem4vAmemh&goO)aR8E1XxDq~Hq-$dzP;ToT9$|wNlqTXWps?_^ruR)`_iwRSgW%6JT z3#T--(*@jm?Q+2ziP45IzPBA0Qxk5tD3%LhjKyy}M@J1|9(U%g*J~A2`5Qz2a%U#R z(C*aPe&ug`|CcYm1j@92pAjwa#xlVdekkKxmvY3ryN^X4M*D$Yjw9PckO5_p9``mr zn{;ge5849x9n^37iIFpKdWnAyHsu#9mvA+?C>)OvqR1h@pWlvdQBl)g2OY*Q!1`_MCpjfB56UiEVWshC5kRbl)n2JnN>Ze0smuYCBS#wY~+K=`xf?*AHX)WL?c8B>LWq13`#BcK} z6cO_$HuV-3xQ`i&AS2KPk-d_1E1fkbz(!OrE(gsKhFs<{<%>DPZ88_{Lhq=j9AEe3b&EV_d4=(>w-;=JA4#(P?fvnoB76|}H^bkbuuEc2z=9>1B+ITq?#tFMX!Uq(g z$j@@90c%OwOWpdd;&fh#eK8?2Co25K;^Na`VG{AKI$CiS#iZMzF>9xHohC_Xsx?5%1didG9yx(Hy1c zy$)~~dFq>u{jFA~fMGkX+BYA8YC4e_m?ft6{&Fx3@9h6+OtQxp>7%8j&{qmQ`J2(J z(aK=7Um`eV7<(*lB-@#}YMM&Ee`rpd>u^1esCs#HGIjUsvVq4v-=m|eceg`SGWxg#O)BCII5ximK^pBu{3{jJoIxkyy1`-ZhD8_f6o}eYvOVqA#rW5%yGZWDfZ7^>8b}!ep%->_%%J2 z6(z%=~%St5aW2PDO4;A4I zp+Noxiz}79ti|3E8~NzyEVJF=l|fimJ22`x8%4dQe#T!_nU+^`pcwkHc8Zy<>TGyn zc6w;6B#R%A2L^9*(j26Ea;4AwQc&dzzQ=+D(pJUza5L|%YK%I3ek$3F!O3rOa zvFeQ0Yt6SE5*@^$D3#fj(wNil=DyoxAAfHZ-AdZ+wlC@o1WFXr^>i5I$h&+V4>s}X zsH6MhtoA-_f0kGn3&RVU6Xv`1bM9HZgL27;qEcK?3me1m^AE5JQ}7lLiMxL#*tBfS zur?3evjxgEALwWH=~f~iKm54*`ZXFS@mfACj@#3p6nDhjq-z+nL@^L1pir*utjVDA zRYmMfs{oVb#gR>jLc?;JbB)^A4^;wux?Eu9ZegkTyBWo56o`O?)&TxnH}n~8=dNO@ zVY?WmjYM{eTyULB@ZeBW+HyCyrWk_XOX zF&(2B9_;cc{a~%XU==QfdV$b5hei>LrRr=`@ps**CZ>w9kS%G?n^>s@n?{m@Z*R5D zZXw?h56Ee$n}!Z$g9*2F8_h5$-D_4plT^rlSZcK@)&JcsNAMb$OD*`KS}A+4ZLZ&b zq-NL;;B!8NW;(wl{{?rp2q^J6)Ez`nj>wpBV*|=Hv3F;2*0GG7C%j*+NA!o@9Wn$M z|BAunqILw?)&}YE+qkU*FNC&rHV|!}zQiMS2Y* z&W%|seH!t)_U{x^f0>kg=O-EPE)j;ukNE9*_I9o57D-W0s{^(haqL89vl2RlevVQ3 z*S_K8V$0jp4)`t&0I^3mYSh?`4XCN)EHZ0QPL8*JdNcQc35c5*l#NGt_yR*>hq|I| z9F6)UGehfi&p79uCLXcDwV87?n&F8WB}(aT@9w3j;<{Ll>>|Ye2EWw|6Yp*pi~m}X zvnRjtK;L6{H5A~wC_CZ9&o(hsl7(}8PoCR=L81kyUAEjPP{KXoD$>-I1e}LrT6aeB z#jw3~jGW>POI}zgF%*50A=FHR-wHY0PeTo9IbF>?4skttJV+=tNbyXcAF43B>VbEF zU@3IoA~2s_^#uhcC>8o~KN-t@?g=}Mu;7cxs5{-B_bvL)IZD|2h(uZ}9BhuF?A|~0S?0~T+x5|Z~>bnUy06%h442oU@ zOXX!Sfk|dQFB?G_(3pm)C5)fD2E3c35QX;1X zsYg?hzz+)sjp`3cxN|m@kn?YrNP1b2g$g`*x^T8Ai!8>>-l5Id0_nm{J^|cDz1tr| zo6q9zJlHQ*8~a*Pb)yB8Eq3`l-6sF1Z`Ds6S>KwjeUr4qxlr`VSL8SENEPNBgcAua z4euF|)^r0Y(bp2MbO_9|Omqe5JGP}nIc>s+c?#WfG3J||f9B=aD{s|TI(Yn;ZqDJY z1hdY`xr-Z8oOB`AVwmNz>>0H~k0>#|A1C z{v&qq6X?m({mcfv{Bsq-Ts1ZsjNiZ7kfv{B20LUlGm)E_8;iZ;OOeMmoeDTYcMqO$ zK23inBu&ixV}npSt&2AJ0xM);s|r*5oz?Iy3hFRycBueXS2etyu}b`jN*h0%@0RBl z{3=hOq_(;B9z)@$eoA4LihtO4~ zh|87Zc1AEXOhc`FA}BDW$ivku2T((0^{5HTHgbp45Rg0a<)gO8)FC`JmH2IPapJj^ z*)~CDsdRK*5=Jd|EW)e!$dOC1$ePUFqgBwQw}%B-o>^!LiP1BhIC{(-D|?cwdnpE5 zJs~J9Enn~b3N#&bZG72_RZaw*oY<8@9oG{ntFNrSKP1d3Cx`KSpZOh}mGliyhiP7+ z0t!ShO1?G6M+UTAZe1?5u_A7L5Ix!7guJt=`fL?lFYwJ~%#t?zyw0UXi9v@)9u%2B zar5Ylv={x@k_1rUG;jCeHXdej3+JQdaFXXwneA^(V+npv7j@9()$XNy zbd&}16iKL%v$uDAP8PhzTO<&@xfMNJd6mG(5^FSD%6d5-GKqPw2(KvvuVLw>MBc2w zlUSJf8*MlM0RF1`5bDy}7KWbi1S0Rzbp+U4)oKm&2<$BkU#_SEU4T&d(YNAhnom@; zv?BC`KloO|8>XIIBC|DlEJ-#3G9f-_Y0j(uD6o+#)}2NxgVWGfEzj6gjbBPJAnIiXC1vu;X7Be)Z2{{Q>|y z*w(FVVSmuKHlpZm9-tWyYf`y5KG*hB)y^=NY|Dz-a!k56MjRx}b@bL%tSYIu2{}83EGEE-l7iKz3mvorQa>I zq+rs1h2IjfFfozLg4Crl!fE_?dTWB;DfV`*jkre8I1ga5A^NsL+9ZunVL=C?4&_E#hqp;i+(t_ zFZQPI9-U@dD|IVwkPK}Ge%$uh{n32pE;UP26(r{+^RYGqIn<$0{e4>d z>PH@DUAz~IWZ1j1$gV8%i5DmK$=Ry2P`480#ZzjsuMP7j$*+9Yn^zAqJ!KekSCRQp zP<0xs)F}rC*LuMLam-$WJ5CihI2$}G0}^sIrJ*Y`=<1Ee4*QtXge z8cA)bG~spg7Qc($*myve7HoR-TMMsBiS^1R$=k&1yZ$l@$n7Mg9ZI(f3nYx ze(-4z0??C-ys`&dN>}(w`alyG=S~|07ZDm5cT$~mf15vH406v}6YhhL04TUYW3F4jn!Hi=PCq!qDHz#b-+Z zsmW3PhSqIN(VyMztxczPNfTKruBye7GMo>|BmMHI0#H=3_(mv-G$@|03Z^ zjb!8!IeYnDrt)11Vw*UzV&+ugu7goP6^^+^zM9||r-oLPi`7sRHxwzB88=Wp@Xf%P z=s|Km5Unb;b;S$X$*|B0l7&yL$4;N`_0#BXP^!{TX%^+ca=ZrD`?MyJlmb4m9i1Yc zlnOb){l40EQHj~~dpW!_{wwcj_D3TcT&-?59(a5=_qBOs>C(`7kMx2wtDIT(OAArb zg991DHBP)|p_JZbt=Np6AIp0j$+Gx??DBm8f6W?91VXO`Ks^6#_sfV!|{C!^lY-A|}9 zyWz3nkF7A=C}jcVE%6T!TS63x6;Rg@2-L=r?8pBDoclKAy0+YDl;=-zRHYT#q|iAX z;YJ5^b)E*~g0>;Z%ay0)i-}Etae3EYSUHvBeKuGAFnz=eX!I_G^I3M0{hpfuYOdDh z2AzA`2Z>yL7o& zock(<+ZPiqz1Me3>nnxQ1ay~n#@RFWQZ*;Yybb35K*@hmANxl3MVMhHIqB}179w@$ z=(7kCR`*^>72^c-Rc|7c{0JiFVIcPD7Ouq*TQ@R zKtiqX>ea?(2bbrNvy#LK=ePQ+br;nlM$V5%1gj}Vtne-|nSpHFKq#T8z_YKZpI^j_ z`7{+) zs0=#6(xb3*>A6BGA28H%<}+BwQ|ncVzs(Fx5UUj(K2$cz=vZ5}84xx(EO81Iy#w#r zmIP||i_Gl-N%Z|8VS2Fe&*oP9=A{YhpZYm=88KqCCJS|&q|KtbsZslxATojtsf4=n zjQ29RedWF7&70Dtv{c8@YZpLwdBZW~`3zMuQyD_aG<2d=Uv{Igpng0|vu+&sc&beD zuK6`?MW(V#!_BRgC?~)qD-)x57jpfjxNH*@UFA12%Wa6uEtUO)BbVbdH>dO3kfXtD zfq?z8KwNc*#b$$$X^K*6j)5OV6B|j;+3KO^(1?ou@_x%%PP2r(Fh&s*w~G11g73ku z$pbUEQkHZ)&G?8aL<1k*?O3He8)}L0Jk-b;-YWWppahoTkz5^&kc^TqbIxve`9g-8vo_nU1u;uV2EC?eEhJ+dn4yvu z@qEY++Gw)#gc8lx%=!{IFpkw5PBTt0q{~lY#;-!z)Tw*eMEoXC zxpyw!wEkcq;GAP$wI)(*yS3)qzf2s3@fQt>Jdg@)$Wa(+als=XfH&bKS3jbH=01l$)sTgL{XUTigoSu+w zg_{MZx$ldX*BQk2<)0aEelY$bB|W(|jdPN6veGP9b8+R+%%gBi^oKX0bVTDa3R&+g zD!?!gy8zPbUS@krqQ1qBiMtCx{Q*!rS^UjFoxwk&!iERU^GW|kWR2QUUJq0DYX_0# zpE*2kA_b+6+rBHb;e}Ic+KV%ew)>4hf8w_g#2-B3>w<(Qi|EyGIB@h)>ok|#pE`12 zk;1Mf&>4_K);XKakcg?^TngY*^rE{p+l&uk^%1N@5^c4XT`iF<^?D5lApXA zMND@{EaOQ$lit@49A%X=tu|RKdCI@isSgbuo%Sv4?UDLelQ*ywPE@Y6%H5He^ZrxZ z+IiNk?fNP&A#7dekh{<5o-#8Ucu69oHWWfky2kzA>^)uTSr=Q5V;Cdb$20A9;>c^{ zQMsi&N%HQ&F3B49Q;vZg&i-SWBkqDvK>ulB;RI8VR_Z(~zD@AK4p1(l*EAfCbd90? znXL&Q=_Rq;50JzioAJB#xdXnEL&4rMwu|wm*u0DHpA-H+$!->m(tFDkq{w$^6Wlb$=9%)@AE-?y7A z^2zd9O~=2ya|w%Vv_E~4;=v?J7QFCO;2<@cr@Vslx*WkISNE<>wz*34y^^`P)ek&n zy#cB~@$o{i8_@o#ZejRA349Pcx*&o=U812@nknLH<|I_=q&cD60z1RaM)H=9&Bi5b zvjvx|{nla_aJ6o2|50SEufmHokc7!ZZ*o-yA^VFl;CjepO3F`hT0RiK6SIB)kHvy% z*HMAuitvS*L2Pc^ugQ7|gv`;Hbm9>kb zC?1_L0rvz%_iq#K$v#6IM$qW#46X36!e0Y-3v+=*V#>u zAVgiU?&%QgpUAfg3h}59A)xvkgaUpT1{}5`MsZqzOCjKP2Md2xl2r4VMP@Mf&#U*{ zOkG9bC%uur&z68Q6b``xF}ESv7b-tH$ITn57um!D6;GHy9#J#9S`ytc<_(Yx8(kLF zi7EZSSlb=cVt!y_ArC^jO#}xMRrRtG;>eI~k_OLh9`o7m=IQ|Gtlla0G2ki#@B!q3 zF#vS03azzfuLuJr5ve`5RF?>YVuzO@j!R<@buBm!f6vyyae%beDD2GLDW{rln}=}A z?A4S)M0s$)KWH?dHJchxeHqiNJ&}KKm|+3Jv=61?+bj|$uI#GD1xPwROJ+%P3O;5` z#S_RT&}s28!!MidCzmN#6&64dgK+5gLsw9qRm8O{5^asuYmkOuw(v}U2fe@KL~!e#8@q4 z+~YY2K<4Hv>RkyvJcX}f;9S-Xeh#_%WjZ8YfzN{*D~W^39R*I7+&>?FnzP=0W;2ud zsix^BHt6-u^%W^uUWzY`uJ7M?nmRt;0?J#M&C71xZlU5xZ+}`ebh(a8)q?Ta7$P<_ zju}u-wmhEO=fLV^!xDkxgm>2uXp2a_S+Mc?!IuE(=E$dCa!d@uiu-H`V;!B+u0@S?CY=5 zCz@?>q5oXlo{`@$bVU#}h0!-aaL;N?5_w=!s{#&| z1xrhy%K`Z8IL+h0kLOh|k-#1(d?6vTLxH+mR}I_W)O3}S;`ZWZ1^?~@sJxj+ow14u z_sf33fIPHiSs;=FI<4z9!A2iwx{4$*XQ7>kd7<}Hlg4D|H*Yy)ce9=fZe`D6iK?lZ z#cO`{J!^Ga!TJ*wf8{f~fHN}aOLmA&r2ioIql!-i;~chrMAIjnp0p2UyMGyMV$Ti- zSJ|v+-$L_F^7!6jAYNYT6-0@N?es!|!{y+ER6C+>95K5!6OutRtcI=v4wv2>%y}uB z?3Cuu_`rY8O}X=XaFq^!c2rL``_b9@N^h)ttz0ahyvr^vNjP1uq__*|^!Qn;B(j<{ zm$a9AlSU2C95`FNww(DfwliM$?S;jkIyD5RnDObNF7%Q8`aZ^A@e4|mzMp;j%#F6~ zh7OHZo5vm`>LR86bk)U@l5pF=Tb*tx{4y3~d&ZzP^))2%GE>8)A-8k_hE?-zRqc2X z#K5R`tBj>AUo_7eQ{cn>kHFF0bfPVN`*t!X!=k=X5GztVsO0m=UX>4)z|ljp@0ozk zNr&z0ME^zZ0#N!@8t*+CM`ZOKyVVEi;91tOq5KE(vp_J!uSf(SCO@73*z02z+Au(c zt86k;WXH@z9HH2)DG?GsbjI$IRXs0(xfDo2zmP_?y`2QMu! zUd)V7e~_;qOV&-rsS(`!`9Wbt`Eh3ilhihH=2M~$G9kmUMhNwVF5vYP7m#`wrwYiqN*&&>}0tBwP=V1s@Nw1M7@PcJOwKHV}7GaX=et*w)+ zpyJNj4C=hXoEL$VQoIIQp4YBh;G^!E``(IazTJ%elo{UWqm-xx+`W6%jdAEXfIVx9 z?JRj>K92GYwPL!U|5oqZ{SK&j2c6HF7;YYSDrmmE$s8NUzx|b6%<>(tS+m{MsK(Ys zDd|k|?WVAg#orFV2$xFw$&b2k>_)d9bDicb7??e>NfP=3avR6Gda>FLtuCrD;~=9+ zhwunR?NIaGairoo{o^?X*O|WL>rCHT&|>MIGJTV{dEZ%7Ra@>{hyiIir;89uVNSsy zO1iv*j!nc9BI%E8BLy1X#ore8;6bN(uO3;~T->awJA5K;5Lsz~vY8})XJtAF9?S zF&MDX(D`cZS6Pr6Dzh-GX zZ*aBCqZhLv_MG{&=bQyadnE8!;e#x zVRN6XunF|AA=cVG>a080YZNH9JvsGE2nxEp+Pt*Fr|DGGBA)Y4`bLa!k!4>e z&~USxD=ez<@ZpR{yza@>Ib?JvgtyU;ZVbR5s8lZ^j}$D$$>XPOc(!P}pqZ54`a5+! z$_Tc}A!nSlVG6V#tsN2dTb$jA{J5t(0`8O-2_Eu0>sLV3DJl^xTBbH#<$S9{D&XkC zZ%N0|;aKX{B=!(FCS(<#`a#8{EkxV=B)_ z-n5!FDa);o&xVG?P43^{xRbQ4N6L2>`lNTxM}u9mr}Oj4n@7s-B5X@#=$ujJOcK(0 z1l^hN-KP4QV53gr)ST)&Pi+4QPQJ#0gy}D4Z zQ2>3^4^(>=;c&m0W!@nxe51-99++jD(3o9E^0D-l#!C?n!X3Vy{@&BWc&oAe)GJwW zqta3{)-UgI`dRz3J)sv##9z^>%AKCs5Z%oco>^nvfd5~#xJz8OxL1MWUjCsxZt8~j zLJAsct4)3$&vcXI5!tKF7JO%So_TK7?~x1b5|lo}2GKUIXHq5OfnOd|y{qZ-k*`Nl zyDOV%U~unLnGBehvdo(^<9e>hT*js@Z;wJx|^v#8;Ed$To1~DBQM#+9QO@A}3 zpl&>@8Z!~(6(V%XcH~@)F}xOj#}Tm9xVXEWiI?}Rr3_=tQk3$wb1*E56``o53Vg?G z#~dJ~Cs>%~Bg1Wwr&Wu_@3=iE>Y-C`m$`1=c*C5d+v^HNd1a!WpW3^!cz3n{>ppa6 zyZ|f9aa-{AQOvunMZhpi-p+MQ=~mV`o^yvLi}VoO@k=5Zoca1>;Uw9^E(C z!Q`AKy5A?kfi-=v*OwskiEX4gFGfi6FRu$^dvo{>r@+=IAsIfd%8k#9aKfzYoQ7G{ za_??qdGFB()9_HZkP7_>rm6kJ3t5erk3|`5DheyBAZJa@WI6pwlS$C&tJ-soUNEDx+$J|KhdK|)Os7p-}?Gnh2O{jw0za0WE*V_qU-G3wZ7_>-&RW?fSFVeKE!5#!GoAtif_ZHb$*?{aME+_OBd07Y^)c6r=!bveVjI%9b%&I0u+ z;T#G&c`?isA@G&&YCj#<7C=eQ4)TzyjquK(Tz8~uqME$hy;Vqj<$;9Op&mG?5!W;I z?mX;#MElrCmyC!$h}%x^9>g?txNta*74otX?> zWa^mnLsj7sVV|I2!;}fMx8Q^X0y!E6&mQL*kJ^JT>OH$ppA<|m{M_8C8yj_c!Bn;W zt#bRQLJT5mSz1>{ZRIkU;)H7ZRLlVRR1yd=`SQ3%>D;_JHg#q9?4*mbHY!|;Y_wiq zkput1oGP?jkjP%>fgO{x=hj#x##bZ6>BFg`KK9JwhsO6Q-OY7vz&Kkj(-X0alky(S z)h^ZhT{^w-YoAqpMI9~=XfHg&dKk;fZee}LGqNDRiH3gr-Vi3$`sK*S5lOBkh;l4^iZA0O#;>&^qUkVbqWy?@98qE(GV&D_#M9#6oGnjcdc#k) zRuP>gCF#OB{C+NQQ^V(Xd%uI! z^)hin!-4huOGH#{IaQ3BiMQ5K*?21Tx#i1+B1-bWHcEkqZTRVTbToLzF(M~{a~`{H z)mZ)XsTuIP_cCDAcd$rVa4#A`@tKjIe}4{EDufX{;~^_uKL<;gNjM3k^q{5~KU4R} zvx>2!_JEq25TQGA$yu)lg@LmDV@Ar=k%?D^$Wyk>rfDb0vqICs4AgLI5UF%TO#Tb+ zb3{GJdPnehN?Ed-!d9#PWVlWy@cGOe5wz%rHqV0(gqwv2o}P6Eld$0#R#t0b+oG;yHPMhBh96gYfCw#ITJi*cXbf6ZrQ+t^gv_^@doEx@(p6t z?S_CLMd+EA^UZ{lj(CYW-VwUz`-fb5WUjN9r1k&$)ri|-E_=M5II`ZR+hAv^>Z##q zpd0QQ$N!MZUvhefS|JPrIR9KD4unb1{%u)MP~%Xd{d+NC5cl(eyo`)DkCO+QIrzMR zREM~WOzY50|9yVPosQf(zRqCIab=efCXG@S&p>0{mmfx{=p&Xx`YpI4ieIW*XmH(r zZZ@-PY5AcZlskRL` zp~~DYqKZBiCmOCa0vj!Wb5cwq#UjRYxplUsB9Dt3U=}&qlKgRjeKj>MSAV}rf009` z*sB31iK!;|_tffA@jkdu!CrRGo;BNBU&+{eFQGUyh}+2;vq)AKT*8y($MBP0P$?8_ zN?p)%L4JQ)0~408)jxTBwW}wh1DFCHt4eo8fzy)aJu&1@b|;4f++8WH%ux_Z9A}sB z5%+8~T1~_u?rzMaHgTdG*c_$NG36j)1D?1vw1$wCl;{-(<-&r}41R|qg>#mgJ?0-LNV7+>AXpD@EY>q*Lg&SNTkA=5ePN(5xb%rMPjt(swolHTT zk@;V!0*KWhx(=WJrvm$nHXj=zR&}~^$8DAg2^8b`GL)X+29hv@jlT1#(J3jXvOyV#tJ8ueF|oyb8`;TTTaK@y%(M!OwoHCEMjk{sQ9k&W~i(2L6_ zjS&B57*I~mkssun)QD+X% zdd{`&4Su3n2cIg}yKs5Bcl$Z)CBtWl9m>isMqEXHGmLSuu)Mr{&F(;X+*}U*X!qMM z9SB#V=c+O7xtk5q?`B=-_{!shzhXM)*QZc%dAxmXG2y|N;Id{zCOAP3fwr@~kPDU1I{S#(j428K|xLX|q zj{ciDsN;Bw-wGtVrttTkKp~WS{MF8}XCHGGz0sOxm4Rtd!wMsFYh>Mi_gW8=LhNtM z&JAxnH2CpO2(WisbHY0fhpFfbRRS6S^XXi{_@z8V=Yb`q7U99pPd1bM81gcWdup90--Ry%H}ZS%X+w+c1rIyh4jc|vy8Y$dP@?_iZ*4EnezYs^wVc;c6zt59 z_SP11)*aMhEnFZ@k8W-jh5klbx;VTa_8l=_&K(ysoYG~`60xy<_o?Ag0_eD*7li6 z`3FoUa>qYX#^ZDe;)9!%9ol&O1b3RwY&6yqpn^EQj!T6RC#()vtPJ=7|>HhIa z5UzUJy-lp&u2qj3`}S6hq=c(ioz-;t^I!_Wgn}xoML+eryzLaFmzA4HE$d(eXm@dK zXU9L&Hg zu(>+aQXNt4wC5yWV^hN)Y}}|dUE5DQ7jATRb~?KgBH?LI3-h>0;VDTHa-`Vl8A#~7`m9nz;yeNxFOYO~ec=$vu4LrfF zoXOGAm|lu&;ym#ZF$SEh2|`*;+rl1PJ@{K!_SY1@*8*0u7$g=p*$`PA2r)IYu=rdS zSR-h?d?-D6N=#snIJ1G*-SyYg@SQ* zs||GAYk(&IkkaVNBNg)FU|G=DE=po$(PLgn*s`WxF-C-*f}7_;;4~ zdg^iL#q>fM(R4NoOW6jau_nEAJ$JbVr>3;E9oQ`4r6uG+vuR$C~OsXTGZe>KFREt$Qxf0|S= zGM+wSM~M6!%puVhx3#Zx95Xgr#84e8cj%ptc)99%&Nunkc7^^*_eXs%1vQ!~ZK3nQ z3f-2fm^sQVWWjJ39jGSF(QLaD?uYCxx46q0xuMx_aP)Eq%BS3CaN_rmHYkNGhl7jYJ~f zStFt{Y9|@5b&85lL$as`5A5S&S`Z)_v?ksOTdmToVef@rO^5?j+$Vf+xC~hh`s-uj`o@3vg*w=A6+ICo;ZZNfgtlJ*V z)Sn1{Khv%$uZf&0s;imgp3bKy(?XmFgorh+EO0K@F9k);s9grdwGu^8Xdx~F?uupi zwwQOiwNL5NA7IK^AK9DOXuF3F61sv?`pL@ zQHBN=x4n(w7=)HP@g+e#g>2Gge+^=9W zvVGj?_!{QeQ(G`T5XoSx9d(-<^VkD!ti0`)#Z7B z){OIN{Pniafh6tD9M=^fpXYm~qxl}YYp&Gl(izrGVBjF!5$9&0fip0jg41lK(ad=| z+e+@=|2(0A_og?5{UqKdO%5 zh6ei1xg#lX-i#_Gzq~OZF!0p%cM@dp638VMzI|S~xNmx+{(C-s@0i0SZ+Asv!dQ8B z&~NRBu-+?-vu6&}rJCWYJ`HgZ5eDp&{Cr;yK6IA?(LA^eXg|>A5n(Se<`j2#-;Ofp zcdsspPc0BDhxNyf`}}U$sE6LIFA6XL<2U$dt<5@SjKcQoKQJQJ+Nfoh~BxCH3IjNZrZVM z-w-l0Fd#4@n_5x9Etj}sz8Xbe{Pgdw{?^uxypVf!mTI-7t^VV#p^=f+{bfs| zz6Y|`X43mT;aCMikK%TmIEPMA6K|?^NSGs_2a={R91$~=>(m|9)iLVTrDapGx$f5@ zuHc8Zy=}6v)^F(CY22GYh)dK(->hc#;Ce0J!-VC*(5(}2VqkUic;wf^MKxr48ZwnK zz;BZ)!Q40^KR1I3ei#YbSlC_0HBjYs5Z3y`8kipZ^xI+YR`HHl86W4Y-&)Y+il@V# zAn96M%C2jGwReqij$Ja+{Ibym&qJpV4gLza2&ip#_=i8LWMCTjYXdm0@B7pa50S0G@$mg4hjHH~z)s+6j^O;7ghCIRDa#^9_{_n3D}T?w z@wL8JG_+RKyu`0N7gS`g7cwgb4DLjov^!^+sRqQJ+5B^L$VJ$@xM`J{_`T z5e+QYh<46R$q(yw*fo=F(TcIbQ%?u=uDR;p(2ela?YbKEe24|pB&%g~A2sh`NF;PE zsApih;=X#S27}kgXy!ittqQz}CJH#lBvK}rdGK|=(TA?5T`P>&E+iM|T-&x^mRlPf z$NiU|PbZrcT9i95kze~u+7i&vYB~-o-x2~tX^nuJB4DtqFu3EgZGqdxL?s) zoyL3WY(qvtza@>ZDlmndoMzs$9)>6oGprI-?AB}v2tne-LVp0A^JRp`lNC>E6VWB>10f6Mvz=_gOKBvQGSM#=ZYD#sBvQzy;dq%x0Zn z6GXGFItT^;3~*|0E`QxsLr6+V2|Jfg$o0?C#&g1L5fX@CX18^7XPeQx9;kyiApRA-9GRg|8V*Uqv)!$)g?1? z2pOf@6?`#cvmjpb#hKLg6RPx~>B$_|&d3dKYyumdXcj3#;4i&a?86I8=sd!fJ^0n9 z|C-2u-vIxneZw8X&?0j`iWBc@_tOp4J$aC`bsX>D<7*7C7+4Gm6rej+Qi84Iwn!y!^$YMSoaYrTm! z$T;WMzAl7@l>e`{G_;_x3F}xGU|+vBP;D*`5FhCbf&ZAqf9Igy1cq5=Ee6T20fB!- zz4DHUv5QunT4VEFb9@a@DTm1BxrZvHPCO8>G97%g2U+bZI|z^S^fUqv!^}5EB0u ze;zIgg>KektbEET9q?qJb7eszBrC7^jp#%RdSG z-!JO=Z90HIw@6s}`^fUY#PaV45r6rl(EPjJ7lPjh)qkdpUJbB2TIt`p{?j)4w|M`L z7X>N_>y6SjKK#ql{}q8tfY!s_|rmv_wZkut^kmx zJ)DVz|8Fn;52~j=1XwA>^8a_maI~1;@b69Oy?&2paQyKkw1#Z~Qo*G0Fo2kmtk?}m zaBNM~ay!NuDnWo#s*Bz{vPQok(K9q(yZ0LCYvNzKThhG|1jvq;@n zWK6cks_6N_TY=3I!L3yE;HA|sgq;&9<&H~+nLMk3F*BF|18GUUXG%;C5YHeGNdLe9 zS*{urrQ&2TnIO?Ql9EhuTxG3T8hi~z`!RqNSJR(D4XCc!AQRP4v`DkCF5BChEc$5? z0QFftXQtj+LLTHYX|OIFNG7wIltf=(;TzA&4jhMpQ>vUD^QtUV^Xj&~MGX!p(Q>7% z9+JNtE^=?{S?lVt&Si#k6qvU!BzsDeDlX&v@_qh;-TxtJcU~&kamD&E=>M7rpn8+; zU^wnfkSCsP=>}ICL2H-sY*$IGNJ~+Tq7n-wx|Phqlptrt=+;&MaZiP=a+hkC4xn9= zjn>us;hh?X${A|)C%eiWKkn+JC7ByJoZaIEh0R3 z=RS92ysE$Arwr&U7NN$BrQWFFP``+{)ea^OSoOs+_GIRl>wT`bk1?=%p6G!L z*ps-Lzw)m?>);L$ff$Y-WOVnoxF37wJ{k&VD^2U)-Q3vN-%u)4T%z%P&KFZeFd|yYT_-xSRE@vu?)-V1!glOIO#ncifLlQDX6tnA6^b*Ry2NN28IC z|I&Ne;=VUVu(WW z&7zkUk26APQs5HzFh^WX)?X>6c`eLmAUv+HE_|v7ya?&R42)Lg&Yb~!>059W92xdM zNTN)z@Wql!uG+OMUGs7YC(a51Z(lCsh7f}xRaQUA%b7Gd=FT{p_DqC4b>sLP_;fu1 z(G23Q-$M$7Nn3M$+z*qt4z@-e9j9D}L!m(*QN35|h}INeT+F=ASHFam*?W;ju9Y5+ z41_Bd)Kt^d?H^g2Jl{+Vny^Sa3z<5o8?za%{<)JFXip)@Ei1G?rfhq%+kbMv?8bZ9 zLW$BDN`uogN5$DO-PrpEP@3>rA$0cg&CQy{cqyz*r(S-<7hl{ns|L!fo$`^ETa3Z} ze9U<2RldInos5Qcm$JS7IT^pB)s4@w)?<&X-4965jLu{=D(Wut5v~1(Z3iNsE}1%$ zj_RQG?tPAK%jTX0%m_D5UC)+@;l1R$bu=+1iJkjL8#}dotGXyDd;7Td60x2{AOM|p zI`VQzXMo(j*xz%Tu5q?oAfUWbC}re@?nr9T)jQ3OJ|8}w!qr;j*C;PKu3sI+%4_>Y zlmFo${c^y(_pU?i#Kio+JPdEs2eh>>-*<0FUk4;1*;J-!=dTn-9Z0!QnlBs^p?D@ z9*UvH*xB>+W#J~dwX5Bf?Ia`d_Ub-^Im-GfbvE$oSYC*$2Bglt^t!#*Adt?7|;E@ zZe;?`J;WSF;$R3|u<0)wo7h{SRr*e3>iwbv@$gqP?^(46kF;H^UNUM|Pw{4|v|h)J z9a?&T0;sc-bkxx(IV)!6tgJocmpQ2FqD#$x_pDbsb-H5zmqfdGpH)|%yd!#d#DDDJ z)Q|*w74&k@?QBOW3b{3{8?}9mIti<%tMeEnwb~>eQdPHg=DIOZUABrD^G1W>cpIdy zuIe^ZZ7D$}=px`5kIAQ5mzQ4&V+pI1pZ3`7KJDZf>jw%L$_qQ#f2VZ;MS%|ZPcgzW zs{hh0$~1jCY!er8yxoQqdzZI98LZO74=e$WkpPv$;r;lf!U_Ih z9EQWN@hodAo9vLYDZ_a=-t`OvWuIU`^C5QYg`~{7<)!MzSC_Y1PMc&%-mUG?a4$0t zIxpoonB(R$xy&f+Pp8{^`M+wED9O|F&E|M+tFJL%Wj`%^^svMEIwkd+6+ePEjO{#m zbbe_e`LEwZ-MdcH7zo<6Mip(QTi7o&**mo3G5Z>&_oi>^PE^<0hdW>EzAOXp{K+$+ z18(gytDgxv?#oq2>wcZv{EkTa&A7pkW?bM$U2nW!>5axk6yeEy?Yp+D&+mH40j4#n z3xkA|G=OpI@e?gn-7Ty|F|Ij{%?FxVlzggZB79t=nF;p!1MNhywZKGxOnr!9d z=lgRkT&QOd(8xXH)|wd6jH7&)k#8w!yXPN96L4UBZ8MUpG34TSpo?!emIuDsH44{S z(ep+=w5nwIZrAfyp0A;ckk;)}G<&8|*KGm^9TK}+7H_Vn5=Ond-vj>Qpj-Uj*|v^F z*#%!+ZGSqouRQ3e6VQd^Arh&h0t1|E9q7T) z-UxTSegmmaa3s1TUgVKEa;q*+Fa7xo-BM-TU);?>$Ke*tI1uh;5#02IttC$P^O`&o z$OY)3u#D#G{Zfhh8lOgOL^Bc}I+>#&bX|S+io&)i&OvP-|gdt%6TKc%mj1b-|+rbpu;30i5=b-v08nCD98mb2Vk5j=gqe z_vkuII`hZZ+ehV-a5Xl_NL_O9jrb31tDVkEb(fO26EnOtFQcv12K9aCjC z^CB@S&OB9d&iOrUagY_jO5I<_%`N{{Jf+w**G2~qbeka>cWi=CtB5@^X={roQ3 z7c7rmy_!W!>z&ba_W1ag(I7m$meWS#0U^Ex86c^@^teeokV2;8{a+L_F##7n$9v8v ze;6;#R}zbqrdf?s!n?+%qny$51BhEb37sT3t;t(OU1&$y(#Tw2|m z>-CJrQ@Fo378YyJth%|Fb7UT%PT;IVr)G9tF@T-{nUlCH^c~}rqJ6SxzJuOiJEjxU zLU|qf<*OSgy8v+y`gwPLX~HY_1VDOOzBfd($B*B@q)8Pzw{&krIg1jO z1q%<~cA5ouCl7sPyTqXx@}uxL_DL5rTr@@GdfFT}6Hmj$a}9v? z$;1d#VK`8ItjlE$6HmGR<`*k};HLF&gHHWJ@HO{>Mj-@2ct^Ig7oPEvr>D&!E8PE2 zXICB%^}hGZ);T0~5GRR_viij{PcQ~<#MXKN8{qIKKx`tkqr;V z4-fk^X_xA?cM&!t_68TQ9mU*-DtSTEQS=S^Pw zOohNtPJpSnnVa`)IDy%%#A)(T<8vTv!c<|Kk{~v7(pREtMe3K0x7q|m;@j$Y$mtV^ zOR^({52<3;Q;OCHr}NUd!IAP6yn*F$Yb1#>R0mjnO-AKvuT zMsV+V5ZPPa@fN=OLblBG7F9PpYSI1u(WoTzvN2_QF{-}9h{tn#-3s4G7YHrg+$Bwn zzx6lrpZ@gbS`e%m!Zr8i?We+ne|yw`E1y8HeTs!sZ8Ic}dNAypbJXG)9b6YoIG$b! z-K|e`4Am_TlaoIEy?j$Y#ACk5;8m>UzIrrB%R_`3|a2ZVa9D(=GC8L(sx8t_yJ ze)Y(fHSt=uu=2uxIF|7iR}|*ta2$u_voF^dkEj&b@*lqd)NMWw=e_XKF)aW>XD;+x z`a^nfFpH)AR7H7d1P;cjC8k&q27;pD4PxqW?M$df$XYyQ)kcvfIbl7`4|hB@Po9KX-cX??1G(NX zPlN`)d$v0Cq<}^8lY}#NvUeU{6jEr*n#wLt@nASyaZokQ$jiHA7BAYtkUgB{yddW; zVPwy5$pHs>6+3b50M_s4e^qZCsMzV?zWOPz_Hb4GW@|O+3Fas^fr1yh#bSD;NHlOM z7X%Qqv-K{`WQLRrq+^#&LUVMtz7%-ks3r>B*^~}c+%4R;O}7~c`@Jz_r!8gg{cU5% zQPZVYL^g12J=6Kd{Hzz4iCX5hWXw1l#`W}7Thbr*4WEM~A86y!bf#CdNQvznVozpa zMjI#SqkjLqo*<3lTIVNar`8yztN`0Hd)TIxHEa1H@dS}E@p6ZaY^qUfyEDJL*!X6Z zVdO0AES<1)u83E=c#g_xlY#^7Jp}9EFsXLGUdjrw`-p}}$uz%_9rR(tr1(tQA72+SWVP3_6zc?yb?69%PFRCgl*eYOA&6m`B}874Pjx} zeFo{|Z23!fxnpEH!;PN|Q-9wIn&G}ImU9u;fag|#sObo<@&9go1J=bWTnI07QXn#KaWs4gQ zO@D6aozbIB`voIOwYCX?Fwfg7&iEZS7yOD=efc_9zxtwI$iUFCkoSo7t=!4rHh-_J z@y`nj!i$*|jzYQ4lO*0L+9ND&E27KyO7e25WYA8dkoNo9LYD1~Lr-EK;-QO@si#(qi46p7w3L~)PuBRjHWcFe+2yB&5>aMBb)Kx(6z?v7{ zZCSN%AxU|7UD0>4wNC)wWEQr-Z z>u$;D)W?if$H4tO=IjV_#v3A2|4d4bbE@)=DWjnK?~h)5_fEPr5aNt_4Z?S#aUKWt z*ChmK2pRW~KNF36<>QycgP_|f0ZLL+Hen=Fy}_i;o}k{VzMnO+(hD2mV!d^G7SbP= zOfb+|WOcCiX7aF&0i&|EOj#BoiFrQbZGCV_2EY;}n&|z@QMLvr8>H^9wMU<~H8QX* zMBQ_(xM+3G0CO}1YUr_5w^Ts-OvJg<(9(O=5hBdPa9#d1>>?AFQTJFjx-vh6cowp91jSWcL)zJ+2Ek`3 zVIiCCzL;W^^lXLHoV|X-5(c$V=R0wPCE~skBWr>a?EX(gEkT+Z!JU-1ieE3l22Fp) zZ;t>vlbTC*HThLs#M%*M_ekhq9AK4%V=Q9D<@{C+L9@fCQ#+Z_UL4fm<$UUr#OqxK zJNE9ahuTx&S71umG!C1EMy;fUJaB$6u*(#%JIW+}1AylB+pY7(03>7EL$?E%pot0- zIK(L*4I~=cYM30DblintI?7p%0y29*p-4$%EZVWyT`wJPRHxkd=4@_3!Mw3<(B{r2W{hp0tVB*R@oIlQh2+^^nA|qn-1T`!hhdX* zfb-Gi6N1i{6!48dJoRT@&PHc7id3NN$c1q1&*&;~K!Qtz!cVSmxUhZt;E8jFcBwI> z-0?LHyjT|{5Y8JF94MTEeoC??jo|I9<$J`$>R|A%(e&}$TQIg%uYP8(wVv!k3v05P z?D_TH72oyYmF=Fb5IwpJ3RP?zvm5JU@`N`A!nV{h(4g4zW1f9c$|6tjFCnXm;;nC+ z)ET}k03B~;30c*9KsFy=-j2Q4P+5h8{O=QE7Lhn--aWk^d{^Pxo}OGOCDXP<9tz_p zQ!Em|evDUAP~@-=+P!pe)rk+gRdBnU=;~->NSaQx2`c|9UV4~@Rd|`2+HZ=M8yR02 zA@?VzhS!-gPVDSF(cr)FrkjETZHe-v^~>1sEqsJQ&aJuKTh>yF4|{j{!t)FkB@P<7dmEcLqpJrP1^+aKe;c2Xqn9lH@nnt3_bMH!{7-zutn?B-tZ zlWcq5#JSKB;Tke>@Tf3QuZoKgagBRId4jPp7_Ibd{in(ct4sh|QS0SiMf_4=DS0LZ zyYXw+)h~RO3!mznvV_Az50E zk#tBb4l*)9OH09qN6*w>V-gOY=B_&SR0@>D4vSPB^sNY5OyKb3T@CCUlL2x#rH5OH zf7zoK#b#c%fC71f`c4rME$R_SFpMWx#+F@!H%6eZ38v)eJG=p1K6ZG`h&B}#rKB;5 z-Oj*htf_Jfw00|nv}BR(#Wu@fFJnsvFoB)QsP<;|etF0~rItxM{$kNO!bu5M#-@Ra&%#d{RE~V;PR(Z4K@}`uYm{nA z_er7&^EM}+`t!1==`>oVVDYU4pZ)dSXU~)cfcP4t`_Mxxj9nd+*3rr!^)uY{tuJh@~t?pbXdXFZC{mx;?*m&u_*9@58zL!dmf+Tp6Cw-*C7*DM|k}EA07+ zOI1p*nLm|?c7CjpO~vj7QcTl=r^nxrpVTit?<~i& zn+dx2=JQF^QR62Jemn;>$6CD!uF*6d!Y}V&tYS(FD((=r(wvLm3(IkMvLE5ay$gWa zfd`Q!4aUcvl-9M;r^^?}b;j2Rz1KB2Hy^M|kv!g!RcmQ@K(PXuXIw*5q+(*Y@BXw# zqQR5oI5F377CgD$$+FDpV}_6jCdis>`% zG&xMnnTbY47H9UymVyEsbRryOCPQ|_>)FCsc(~nz%o5x_gzN|4PY5M<*hG8u*ho@8 zoGwpazakT2-c=%rK3RXM=HtqybxpZRZB5R3R1^U;mgtAoF=|2X#D8(=j-Vp7GQp%5+fiV~sw7!|mn!t)1%#zoSuB9|wr7quKK&kr35 z(Oz3vsCvqxj#O{^iy54x({C-ZmS0u1TBuU=)&=2Dw|nxZc@(+s2;W$5B7D7QVjAEx zpg`_N8BwAWR7poSrD@@;{kwl_MYQ`?k9L+RkxY*(EJ))J-GK zK1S;M&AoY)EnaZsSM{CdQ=oQt0Eh@gE!t>>TsY3oh2>2Wo}K%D)>;5OX& zDH(@fIrAZ}OAC@B;Po-($_BQ!CrbvC=_9=DvG=*klJ{9!dW^jtX=hKjUrrads=aX0 zv{c^6E|ZMG4BiMbg7Y)6X<*U1sU43#l=WnzB6aA5H5+#8Uq^r$f9mOHJ9z`_H|5k% zb)zotzEJ|WiKkr}e!dr9?I+tgwL$0&5<2FPMuM1mVDqcb?j}DBifhj*vTEnO)6GjZX(Z8PTNsy9ZT7CV#z_2;+ z9piRa@eqH3HSNz?cw0u4V-CI&`W`Z|>v(bWioNS-iv7G6$~cFRU4$#|s>i79?6;}f zWYB9p&JaW{wiiN6yyLCP0ozhC@^|D{Zwk}o&FahXkS|y<#IIWROWQ2l4EEHVqlU8UCA&*cvK`6DN9VH(agf+TixF zzVkpDj*7i1W6e7BX}V3keoE=e>oYg{rZ0R9+0znUk+lst85j-ncqb|G%=RlvAfuQaIKZRp}EMqN=%VEzNjrzb37;%)#*yr0a2=*`B60_>+UvsmdM z0knAnCOcp&XYCp-w98H?8areLfO36|x0ocF907>!G-uDBX>(O}ff?%8c8V z5+PQ>2&Ytq8qdMpxfBmI9S2&fj@Bw$#_0&yKf9TzhxSSEq%}BI>R;Heni}6a!hoYf zO<;~h)`{}?%6GGK(MZwxZj+py9i+V6ty7bnQ)>%vv1Lt&x4*9$66Q9tXJOU&Lfo}?Uy!__%hO}M59lfFmlLc_0!9{hFc#dp z3U$}e%;+zst*wZKYxe7sCvvF>?SUzmq_bO0I(o+|F7x-^y@&F`BT{`d1fQfp|7U*e_vE+#XczgJ zjXvsyxRpEx#Wg#uH8PXGrZ0q4mfbfne0Ac=$zLuw9$wmgs3<#fC5ndhdop&yQM|T7 z4CS|mowNEQ4H`10vw}=6=J7OUW@he7V!p7v4tHFW1f-G(9JJ2wg3n2`ks3gGJVfV7c>UrSC4QXI0sUIR&=Wy&!Wq~E;aFh|P@)A|L(IhyL)=4~K< zYzm>Y6!{GydJqU-$b5v0%@b$7y8Wazr+RC)DT)R>j4TMUx;8V&oIsQb{cal_=}=-G6iX3 zg+aEjANj^xH>;szMthodD>ye{aAtt8c|a%8Z;e9>(R&50)I*ycDO2?tusZM=Dnfvc zNZ5=p;vplvG-nvHnWkJdJLjC$6R81GU^fVv27GA|p>yrkY;0f1EJujxN82qNb zQX9?~W;saSe>DCl8v7NQ`8P?R?imIF)Bpx%nL!MuWz^3DFfK+u6A0n`*zR<&1W}MQ zx2x;o)cL7(pI$x1uzy*Xumn>TDefYeHtGy8dmp$%w@eXzBQPh_c%POA3W?b#+` zP$`b~rxEQQW@h&W1?DI{N)JCdtw%`kx>WlYo!5(?3AI72FnF@>T-htm34Kj1nf%?A zM)T_VdsrY>&eQu34h-DXO996ug>V{ z>8SxdFxQCO+9mDmGo0o$b9;CifIs1 zkgIx9VX3xYnBbTiq}bWK9eZ_5agSGiZMLVoYQ9^Nv%SmwoRon84J&uo9UILJ#sy)g zgVxjUXzZVRgx{dXj}M3gdGZMbz+vW?#45|J(!u!H2m#_TA4} zb5B$4lSg5Nr#lQY78sPZP#a2^F}4aI%y>4g?Py`i6oI)bGG8%4L zp0t-@-LPh{Nt1opT%}vrP8}VrbslD!eQNe^ZwMyD4j|CyAQn^)xU%X2$dX+ta_5sy zaphDnd#l;x)7LKLn2l08G$$NUzEZbR6jNj%WOnC{JjlpzzOw1`?1J29xm02dwXu0Ovi{_`fAe!!jCF>1EkPPPcR!E(Y*z-pS=xU|KlU z0yB#F8$tQrLSc%q-_rH*uk(_V^MfLNlpec^mpBv0MK=Go+veNf2pZqn4`6UH*ce|XXVI;np*tb(^~s%^40N4-$_=QFukZ_haqLG$v`EC26}kooqi z{&lm;zi{fjLK{BUv8xfxN0bimmpBXKr8j@T2l|nqfL6NJUvj0`$C93U>!@;v)slM$&nR|X=+Y5rGNQiU$elD7Tdy4n8UVr28=r;=ESLpNF|8HH+x6MXO zX@lN%bpz}|X{BuiF1#`@uhFTJ8)6FF*oVJ&W8ePHu0yaW1|A*3789Jr$QL#9`j2y9 z7FbeQ9>vhg#`yZk54eor_U?%30eo$c+<)ym)>%Ry*2i!*yka~mj;oiVUjD_YxYc<=817XU>kO{a>S84wC*ANKAJrugO0l^pc>$kT?)t2jO z24fp53jtENTSnMV(rDxnk$8kbmRERq__dHPhCFn;KW?dOFl6-f^knlfmbJEUI^g7d z!z6(lYJ3z9&zPUaZfR8M%m|h*xO;JIDjB}_OyMhn-oM|GM(?fDQq!yWjSv6Jb|hSZ z0UGJ-?kx^2rnUYdPPy;VHv4%c3`_d#WE|yp&k!mzg=G5d4TAXdF2`zOgwP(^+S)x< zO4x1UkGq%}Jb(7|^4egU)#A9>Uj`-~5$?Yxe`hugh_5&`Kla&4Imz#GcUSTY$-S1O ze;gIZt@$1;2~tKzn5CpQ`@lin=erg|&x@ulg8K#Fexx*6p0Fao(#R>wmu5L|5YF2cJK_ z-TMHZF8yLHIOXaCKI=G|x}>C}e0Z_!Z{IO;7uh&3Zh!u;Ivu1iP7Zh2jf&ycOdObt zis6DLqMc8^I4LHq-#L6IOX52dpuQS6W$B+!Uv=k2B9ZqdKZJrFp4Up~K(n}^#P5iL z@;k5q%#RPLsv)_rLt^rv*Gp3G~;T0n%RT zNk9YAN;eNgQA~SY|5qxF<6BF}y#kx0i4rVBO^;zTXmYf0a-$TWs;V(U1D6Z~812y; zPeEF*s-mJ&@$TIWMuaga->Wo3dGdf}^sO(X>ObuMaSe&_CZX{0E@RIr3TZ>R_wU z^nL6iM$K=od1S!h@L`HnT3ohQjnC3CjP?tu0}jVGWToLL9piOveqniyw|xVbEU@hs zN<+wg=@*{N8Q{K1t25xfMwgOqWaRl#NOQgL21HDc^U74qr4Y(qVDNF77eDqjdPX!L zpLwt_Tq;j%oCI}o=_-GieF!l!GIAIU@&J%{-dkD`l(- zYJK|J!*h{4`NhS(>bcq+*BrW&VvqP`OId%PvwC0wVjfe6{{>&F$jszt>CrlDMg?c+ zM?#r6)R4+ah=4c;;;Gy*!W3krX9ErM{LU;Zm==4rgs)S^&ZQF@5fu@W9pR2-JzRwM zSSUvHR*LMMT8C1b#z#)h&M(;FrH7?oB)Nn+_B(k(npDD$wLD`zFG%#w3;V7``p;wf z>pL)^UpR#&Ao4m1h5kqG_q*?&x=*$%#lWvoUS@$d+*xy>}OYptEtGMz9xJP1qFpFFDIn|1qFW(1@)>A5gt-9JVc8O z1qH)sqowPvtE?no?&QE~YT;yN$?EOk45^2L5)$=xHZ`}ibSF2nw6bv&raW%#q$Ibo z5T?}OR(`AOENN+NBj@XCsp+etW$tTd&Tl~}D)L&$TL995gQdGExwnJ8qnm)YFy%k` z6@a||DQ2T2|HmipcEXgp%4+12POg^ZJgje7-%^UaCKqzG_#mJmCG*e5kYB=-*6!}k z0&HwvUS6zToUBf+R&4D2{QPWhIoLQjSRh}pxcNA`n|iZ2x>5b<;{Ws^W$9+_YUAu~ zID}ddO<;w@AlB>BoeXy~&go2U_Ptrlq)f~Xf zHqfAeMU)a#*j4_j0>wa42Q?l^&PaiTp%8^6-W$VQu{tjeI}}mziM|2%Ywwqq{lVPX zdoyuqJ^Y#8iZ&!@_c@&o-rMaFVBpAu+eow9@gj84a&L;lxtcQ684X-b)2P^Vexep$ zF~avBWKbI)X-vpdT+PfV*F@hgKDD=Xz++hY&50LZf}WoB>poL9LqSo)ucPD`enmfx zg<36#DK3M8%EdX`=0wg{Xu~rjA#lf7*QXiLSbIk^qP7+p%!A}&7)Ayy_HO6gQV+`d zLOffW{?{h~_ctug@rb=xVXw%qhCo2p&)yV0JfrX<@;dvqoE_r>cNri0!f1g&?oZm) zD0r@4byTUww}j8wufmr4&Z@>wMiu(m(Sn84h$cMNe9M+xK4-8S6?eVdL=I=;b*0 zCs^du!A+xl5; z-e23>_MO7Nm|T!TDT(PYE=Y7&Gz#tjTb?icrvT+`{;xIN&$-{;YajCIQOif^oHk|G zk-t^p#%oO2?nRl~_?X+KM&ktU@~Q1@fVaIVv|%PApiN63BjMW_cIDt&X#e+zAJ-q{ zy5Piv;Bcxw1nrR{gbDg)_^yd+t0JzxBKQSEx%##TC035~Fs{o^h#sa>k{aeL&gElp zJEa=pt!kbw4ISy4mJY&K{>dV(0k~L-5d=(f^d}t@MtK<_Iha^-s;5v{e_gn@Uxaj% z9-p(xK6Jh%{w0OR9DjSLFK{5pIS88}$B+YYLXm5=C}FlHlCc#S;%UQyyM}RZa_q`Z|^$s->#to%KD85xkXHqmUx|L@r zv)3;t%XhkYbg!y9&}E5B}??lrbdR3b+fX9Ph%-EUuUOZW_ZD7ttNNJId4 zU2jrf5xt2;K7kS;hk^}u#z4gqd!72a9rWeE5dOob=J%NNa9i(bbP-Cr@TyQJKMAiw zS;LgC!rl;MBH;hfb|$LGc03`z_&Brr>JvHj=Pz$D06nnPQiquMpQCRme#lW{h31p* z$>C9o=QjDifxP?j1){>%xkPz zY{78!@JDJad7y>r8I5$pO+JhUgUu@&-n}<_Sj8C^$z?J;6s=z+e-&FGHK!ViFH4Ik zbVvnAJc*6v>uR(yjNwpN@>s%I3KoR8ejHc66`RV{*8~(JO%m@@+S1g2^iB5-y;D$D z9#$GwWO=7t)}U6d;Q*Z1tG~BBo~)lbnrz;;*?)bxwJ$goTe_kZD4jP}W({4Zr4Q}E zUmoKb?-_gbE;wiH`;Q3>TjtU4)7Fzs`KO;$aty}#FN!ZHec>L-A7R>c5^`+{2BtRc zMrO*Uh01)4YE2?@w#snJlPc?#>{6!r%Zuiz9hi>ZJLGtVU-29z%%c;$`I-5{67fwW z!2&@M!IFR!i5*F{N{mYL2x%5tmWYvq5$_W9l3vSHi>!NV>YIs(0rPEkoww+3E7@0K z*$ICVP!ZO;EB#DWG+*A0om$nrbb%HU<`ul>tjT6*}dgE1(_UTVdF}op@jS z!zOsSTEHC8_qs1RTKMf%GKHuqKR7dWkw~y~XI3!!Yvu~I=tl*!@ zbjY^pm~IPLqz~cPp%-$tEo; z*8xUPg-`y^gwF`j+tA%XB+#mS4VlCvjw6v`MPhoZIbFPlPGBMMDLFGe*LUXcRJwwE z>`ja}#$kGL@8ke-Cg1&bGeB-3Frss-tvH2=rEWJ5MC14lr#WFWHD>WzU=kFhm7Q_aH0r5JS zo;)PpcSn3)25Bdq;cO*HJw|T5t&<)`% z{`_h?;e6s8U9pa)IbL%jx(JlM%ut__tkC+w8>Gvdr;10hKq0%Sxj76ca3a}hWSVy) zXUon+!6co;Kto#Ht0AwE73~1tDj_$Y{@1a%89G=x3h#^!wmM5Z z>!l^HkOb!am1$EZmRa**sw3`8REG;b0;UiI*c>T);D_ZnPblbi1 z*7^#FR_w7HTk81VyZyw=W7#rT4rF_J>YE-dkLG^Ijk^%G@VZfWIlJ!hH~JyjzUP}m z`ju0gA`2@W2g9Rc0Z>be(U;bDlX;gBciBret!!YV`H98V(JBu%H>X>XD%I-axu=WA z88j(kYr$V4vnP35u|(+XXi40Z(X~YP=`|TG>6+<+=%HkecBeM{7YQ4xf(FV4rLnuD zdM=5+b(_0g)w$-6{RRDgsuo$^vpnTTR|E2Nm6;L3Q!WDQ7jr<#01=-rZrihZE4Af2 zZ6K1TkVnM{pV;&0Bk)mZfXDNtTBxk-1{?@^7{8sh!Chrn$amm*JbXwzC9tEr)%D4C za%#jbe8$mGwRPD;(tBfXxGjUt33vI<*>8=cJFBKO-HQoKuv~gOeAU=)L~0}pMp$wR z#CTr4lKk;=Cw4Pd#4|;xP`JX|>?|K#JTx(#{WUuhL@Rm_zzTTEX}s_%U*YYj@;Tig zgAS~_R^C4dVDRHV*{RXFtMq)@y3#QSRC;coxvdm!WAH0|Znz@3UrM_kI0pKs-o{;2 zpCud>C4(@5$0(cVC~~(PPpd^x@QJb+N>ERWNKh$cG zUx=AWp}4x7?q-c|Jlsfv01MDnpYATWpvlr#$k}fLrjOtU-y-R>8;SB9@z0 zL4lfDR^ycqq)RXbB}8wQo{-;W29t`c@bN-Hi9yLriEDX7A7&x=8*EMW#j^Me`Sr(a z9JV7sOOSuYM9io{Y7I8}76n7rS_h~9+JV4fj*7w*B%w<|O>MzuB^Z;{-FYb_uhy*JJ_2T&k!miR~uG`+J*K$4~9~O2& zqoJX_y5%lKW@cr`aq7I!{(?pzF;S?Jv%31e{qt&es;VZfYK_}m zc5U#q)9icD(7>TyExn-MS4jenC%T~f>1QQ&O2i(Kik7q2cb6VBH-3zrvDb}O*#!lPD^!<$Uu=qtFOKoE^AZy+ zIzhN$uLiSO_1k(zcFu-aJ({AFGa5(TK%Kj4$=S*9)u4}I&zwz8k02bXPcP>Ig?NDi z?Vr_P)numi_e_{R@-F)?uwD#wy(>*J1{`_VT3eqS2tARO<>x0RGyBZ2S`6Gs_x4_l zj2Ap*GrsRPXbfo2iq*A4BjMnCU%8c1fzr(!NJ->=?EEITQ*?a|?lb4rQ(4U=>nYm0g(D4fPRTL5oIn@6=CquMQ)yFCA zolOFN4KG7()sNN!o|#$DN#b@UXVH5|7Cr`BT)EWl-fxNWX~h!rf6~y@ARfHamsuY~ z9URPkb$g4ZC>U_X9O@r9QntkxWt+|pNcW8}XT$0OEJU98z2EP(H+pzLCaG`+iQF^a z{WR)4{+RZb`h^uhq>lz225A3DQu|UX7n(2y1(rrFXA`O!vSVXo7oBXKxssu16N}Am zyHOLW2jjZ3H3?(Z`GF$$T!6DV{ENryAsfWj??MjC#qN_)d$O(s08iHNUR^^CGZV5N z5y1#!;U9p7(wj#S9~+zEmHwWfe~Sbno6!_RvBLw!){pfsb@i~0l$h+He2D$Fe^}t% zEfSa9z`%g}u(l4p=*NbuY~y2)Xl1P`JD*@ie0=<&Pv!+H^kQTC&yQA|S20HKIGd zPKJHaWK{Fc*d_q%f4l`dX7~<4%g!jKw@U1Ql3n7ejJs| zsJg5&olD%V@_dVf;A5L-dHddt%Z41TDIp=D^;QYcdZTc-EQ`uwa0TI+l-2?gA5s>TCuPE5H>yS-a>KfzhNeRF7jki~p|M-)HnKJUKQ zi?1#v^|?WgKX5N;=4N(VV_EM<80OEtCiB?KS^M+k@;cKEnfVR?!q(w0(+*SrK$Id( z|5&ALuDC*_?9J*;#Kt;gAY>hi#HBb_mNhKkwTf+@yz2oUHx4eIk*j%XV%X%Jw5X zTg{QIi|!vB9^L{zKD2FV0E|dbNi&B^}iNvZEYO34(?JL9G7Ge&^9%pMGZ_%O}XBS2`Fds#VHli z*B1|+mN4OY-nF&3pH+9I8R!7xz>8LAf@^e7Z6#ax8X{gpLrlaKX4e%JQzUmV zzZHL-+sMrZXOEB4rufU_r;7}h{OUHCvhU?#u*(f|Ch>#pxId0n8YR>p86)Fs$F8n z5P1r7&ot>Z0d$`XPq>K*p8aq{*j8N_eql-dFw58FWN{caF2-3 z>GR|SXw(V_09U~e;NarM8Rq4mgG5sTyv~(zu&{DBFQ0b&S<(p7B4!)Cjz#X4NW9{C zcxFZLx;Z#GFTbb{PFt^EMw4)Up`@Z(YB*8|amG14z1=aP#tf^IztRt&4j;YUr9xNN zFw1QSA8Iz-V5)4d2upKi~bN(7q9ei=*wKAb9TAm|I z8W5DLOdHBGA9Zf5nALjUP&AWUaS-_XG6erQZ?Vxfjd&})w4_w8XStZglfN?)cs$bM zlcFMm$LDr_NpRW^paoi8FFyw@Bt?$0iq2C@R0&ptfIy0QeiyVCjt~mwiE{NO$tW;7 z!sTx;6N*$<<9&Kj`}~c_v87jf*DYJW?33|_gJ!$HOCo6~_loPpm0mn@O3x*ioh4QY zpqpsg#uEamE9Se*1(H2~c^ihQ$;lJ$#SO{}{>rT*=}4R2Sf-?P>0{;`(Z%aPG)N1- z>crL^^lL%<>$EmqcH@BFsTK*~IpvDq@vFNjCr35rk<5d|Eha`L`q=W)0KW}oNTc}x zY6FuN#RKB?OJ$FlGQ};kXrBWIuTcofshzl2hdmF`F_z3|v>(e-EXxD8>=^5NPjO!i zH?G8GkM1n0f0kMUt898YviVd8BweHwJ?&Fo$Tx?rNY1*B8ZmxY<6%37G6Eb`8h#ep z5uqFvefUL|75SKa9A3|CdU{Ur#EkV(;A?y%=vQ0|$lr2i0!YzOfpGNKJ`{_E>JbrU zq<{4SlkJ|v8V6i*`|e~Fw7h!{F1z+TT9)*Bcyf$gPdX2YWa6*Z8lzPOBT1v&uKtXz z8}@rN>`5D{Y}2n6JG5sv)4u$b7nm>md_yNqemJHfWO2#x;Qp)UxUD_~JKBEr@ElD^CWB>l!qh1QYjoi|5?8&*ms zcu^*zV2;SCV{5V3!Fcf->;1OD-O0~bj|G$+;d6_mhT!t~iVpBh#A=aXkmlN;cB$dw zkO23d>-35p=5;6NX~P8V00|x4N^voAF9+c{*3ryv@$gU_?Giw^Bbn@3-BjHS&tlNgB*HR{mAVsY!ygJ%KJ#*;WjUIY zA&cH^YBr27?2jJM8&3#(Ey?c@%c?1x_}}6_ZiN>et~d5v-G4U9%66Mhv&G(`r_2eH z=UsF!jVJq>J;JIA%5xzu)yU=!Z`U21&T~2ebbda5<=}h#h`u+j$O@j6{yx8>j+Z_~ z?I^sJHbTl3X6sERDd6&hse*YbF+E(`oIJkTCL5fRpHz;Ej&^v!w(5N;9^|dIV zJ-+mGz9B)n>GDFX+Vtx9r*?SG)U?3#P0MEjW|(L~_D_x-mplkwpCa+tax_&$^r{av z7W@Pv?z1P!)a))@yvO2X)*4jkx0Qhrsb+zF@QA=7WhT zS4`EXt4(ZbS#D+1HI3Ac=O3ZAagX$N*rhoeR%BLeM8dW>G_Co&Urvb;@b~E!#mgquyWphNnE|ke~&%R)|&}3Oqro8I& zB~mEIKxSWeeF$8CKk0LAJ2*+lSTmc*>R^^3fsUzrVRtOgS_C}QuD>-PuysgiSJTgPONy|A#|1wHG# z-~KS?dgZ+mto3DW56%{03OhH;u<JZbzPXJ#aP2mb9}!@C-)zsmb=Xp!pD$Goj?2?k z5Wbt;Ks#OU!P@QcJ9crAXF$fNR>|nAlJmGNAfXIvP#M8D-qvl5DVJzhHXFOXFaf&b z+zITp#q`Q?D+5K8g^#XW*bSy_$_XsVj;&Sh8A)kC_);4c?xs>Rd@M>t;=~`4oMvU6_s^&u(>k ztWVA&=AE2t-_H-n&SFy?h`e;r(1>rZC@AQ4l4t{rvS!|#MG+U@m(>pO#RDm_%Hb~m z-aQ16Ff$8NLinY4fs|EN$Hpm_hTdy%)ZMYI?pJA6O^6cf)2hrL+T`PyQ@s_;C}`Mf z>*W_it_rB}^dO;Viz4B77r5^{VX={JcDMA?&U-d{wwhb?W|{9ST)$sP%B&~JKS()X zncCdd70**{87^{y=Qg~7neO9;DZ=BP#2nslZH$N&+L|HJfCF~pPEjpfF`FYF&C`;I zC(C!*z`g^whN?^f%LudZ;*l+X9ORC*!wAico-8Y!i6H2h>yRi-UER$~=bCH-r{;kT zV;@2#!|wv3k+y!-TFOfM7(4}Sb#+SRSNCHkb}k*qmN`gBr;aljyIMN42qKaG0GUN& zLqBP+>&3TMf~3)226BY9{Em*@Nc9 z*Cnc3<&E3=dwVq?j1A!X`Z`uGVG+}`{FQjjXm}X;+QUSCxX<+5s~SX;E=p7W{y}$@jxYcg5r=sf)DFc%a*FVDg)D5~Ei(pws;q!d zYjV$L*pvbi5yRc&gh?CCchh1H4LTzLfJ|NFVfek0V8`-p+!M-}!K=ZSo>;64AL5_= zb3*<)_uEz8n^cs}(tVP*Vn{Q;;inTe?QV{4^ubK1GXplgPMQVm?b$4&jg=nt=!5X9 zkLALt%vf&vMIU=VM(a8Xp9Eer!W2kK`^l(c1e~nNUZndnC*VIVhXH(9sf1C>aI=cM zD7E|u`b4?R^d!zU6ypy(Q#Lz{+qbP~`Zh{{?*yB)J!XoktA7|JDK<0A zPdyeY<~ibg{@fJdYt}nOG_X(~hOF76Tgg}C3khEBy@c=p}G1X{Ao$a$Fc;{5!+VU#~FVoy)c=w1;h4_)+OR%8$o;3_4| zG2U)A)6EaPevLM6lI=g3osz(wC)?We*3)_vffKMMa&Q2o05~sHQNNtDF@zhdOyPx^ z2!6q=c&myrwR5$BQ3#Zk;o*WU;jOdEOb^}0{tA_LMXqc(VL6M8o9gA+#S*h!t*Z)wwc zk8$EtW95Vu$(fm>vTx3TyAh?8YfWLeR|7mwa|&l0nfZ^;OVYSRpz9@tcT__(uN--F#vQy| ze0?;?`6-7+&D+`FbH4t)(uxX~)a@3#%0ft>B9ORe`2~gaP$RR}oNideFW{gVq)Uq!{ymB0$I!NEN<(sifK4ucI^gJU%I+*TQ4yPo|3{%}RiEB4DtNZj8%Qd~TCIz%vX z3AlvNL|MOHZpEM=g>TnS1*H{GvM)7>aq_Wik|Dupk&ktw$#KMyt6s9>sn1}(%r#)X zP19&_5!bprJ#%4ldV0L|xc%5rslTIz{9N-Hf2Z;w`++|=h`aE5979CWwxduvUsxVV#_IA@``RgE@ z>%~c4ECc&8F=N~Bu*=!y_IqM|zX;1zFLI!T3iUJA%}oO6&;^5LdY9h{SnO|V`ebvy zd#1=naVtb5_(7@bH9*)0SWr;dV&C^8IMi@36Zvl4D(svv4&lX)g@R6k#OFz{F)zUl zv^Y8BC~G;37$TWwG+VsnT~r!jb1r|!^cc4!{#3Ot8+rv{Cq;VM{mzOz#}1r2EE`MI z5ttYmb5{%lv8N{aS~4;@d^kvW4|MC(ezH&P_12cT?6n^A*X>mjn4UtS_c~-V#!w=s z8x~aJ)Z!2A0?Ymzm#pmACO((k`VS)-jom4Tjn9BvTt+6Qb6YeMqQ|xds_A^)ejlO( z6}Bt$k3`k$?s$QlPZPA&w23=n6{>9EeAjzBCSQ57zs;*GN()5N5Y46kp6e~CdybI4 zF0Fh_GU(V(<-1dPzcr|RrBeRj(+ELoeoraR;|bSkx_blo@O3tRBWPHsOPc$UK#J_? z$iLDVPs&h^Lig8C8!t{Oa&jcY<`dVuy9_4=w&NhcqZy{>7WWLpv__t6EG6c;RH8E; zRKd?b0j3$rhJ&rGZSGEXVBB)xvt)V~SOTGHwaE%ZN*Y+JSO!TV01k$ETS$EGE^Js? z8G(UoEABxXXc0^rX?6(xDSoR@e)4_$sy1WmG<5?5midOyt;Zdz$9uY}wsTTaQsYc) zj|;}sFmvUN%Sz~?XQ}N2Kr508bJo_PdeHKmi)+9zm4L(AMMpp^BymeSQ7I zEnszOqwSQcdmq8M&97B^&x16HP;|1wwpR$d3&$Nr-QARfq>0i_owqHKa9bEL2|i&9 zCB?b9@fB@%`su>=d;!a};~k5J`i+k5;N@a~^C2m;i=x3h;|dhYch0URCJ^kLm=^cr zx~i%8@TTl-TJBA@93mf6XLcK?P&Gq=DZp^A3W;~gmB^@S)t|8Ex*hR>GS}E^c3!de5?QwpMjGPDIU+P;aRfL|qBZ<9S$+Mgv?ERWE!*Dm1?dNVTR)93k zjgwk>+R{!{AtkNUtbC$R7n}k+Wt=4GhW=wA4@%{(ZxpjbXcQrr42TKc`uL4=-Gt9N zuv7cYrYe^F4oNI4v%SwMuiBR7=>t__2)X08Ysvl37O|7Ob9|5P6!cV=t#pihLLz7B zSUpSvHN|J2agCko)mYOE?^`(PnluN6SEG$HA>~-eIAX?Z`HQt#62a=eUeO0&${)U-93BR6_LA!`ITY-!Rclg2?7zT)cI=)4<@JcC7gpf~&wO#Y~xI-i8`!paz)h*@`Uidy}3ABN%eYp$()zx2|rct&lk~KCVI2 zVd|@&ex?UjL3D?oBmz>5*ol$;CDA$v%C@mteaZZOTfrX1z9;p&FcsMA;eX6Vg>sJ; zbCE~N0)S!sb_ya_t_Io*+o@FUXg*dZ1C-H)F2Ww?ojUepS`OnQ4}IFPBYP#TA6J4e z9?!^#OP=pGP5!_tPkKrCzeW}#KSA;5somJ*PJ~ns+~>A%#^x6KeK**5^4kh%HMGM z;uKU{5@Y>kBtc@wm_f5MeuwOw|4}YVUgdctaIdYBAutsXgN?Bp6*Om0>(JIeGH-X_ zb+}kCtl{r3gi6K}W$e49eRr%_$v5dici#1Q?a5DS>$NRy@)zWT`iKV8AsuqO8_M~2 zImEYpR6&%>Y_C5M=;o3iy28DIb$P&btyI8;RGs zF}m5L*RP?b_UFFJOJoIk??i#AI=r_$q^f}VTeRQMv3?CNsp8@$4Gv0Kr}20wkQve> zPo^M+iTR%dOJ30g+^qH2nfI#*0G5Ra1wHG`s`1TGa>i>k1_i&W0t4;{92V>5mW@?t zRo>rw3+y{EI&7b()!PJ}fY)?Vf!VYl5mFcM2ruvz=tTB@recX&bUA-4PiAZ-+xHV?_s^tl z*=lckQMvzIH!;LN2JriZGrf!$Ax5-o4ohBkQH_e!wUwVwJ%J2*P5`$)i_cfNheM8xp1 zF+A#!y8!>xhI2krn8%70@Hv*pX<5nAa*H}SZ+LhZhlofLLcS#?CoAmj?QtEl;H;ZUeE}Cp&T)C?4VivU#+44)#n#>5MA6|UTv(u4EA2e>@VM8 z(x7@e#FmB2NdA}QzWngd`!^HFL}n@*$?*T1MPmGdxH7miu<6nNE`+~N*vE9pL;~6$ zo4o(89r^-*{L?Re6oYUtY+t?5?f#B;{`aQ4Gk=(igQNc5|3=)f91u+TvCU6c@Nb{= zziLOU^A8DR;v~xYw-^6U$^D<(3*=uR`w@9_>F4}^qnbNz$V47}?_Ey+8~J@yg4iGV zG3oT*nS%el>1gF2$1t-H8~NYJ577c*Q7}H&Ri6L5eE&Wy0{(1-&gB3@q5npH|G%1c z7t6=92!+96K{1F&a9nV&@3yTU&w<1=qTZ>@c#1lq;-Z0=RSoKoKYU9>)npx#dH!4$ zxTF7=Ov|+8SXmGs94$2q7HLr=8G_t+$vg6icCRX%s$A%e72ZnWnkWRI)^T>ml^mSv zEsD2akCYIW*qA@PW2~&)ZI^!kp%O`#NU?bD1l^*7?W`r14Rx6NzPQ+#M$f<&u=k}w zSnAu>Dn-dcRT}dpsxa59FaHW^c9miJmNSXn{%yTMok;$*53=($!lC;I2T3`>pFYvk z@TaK=-bxJ;e_bx$OV{Ol*o>^i!K|*Y*CrT(FiGD&rnyh=6=`T$gu4@NlC&wOx=Oy5&=!nKWb z2O$)G9)E>Q0<>HDLa&8bwpjZ`ME4mfNeh1=oLYu%j38lqY>~ZxVW`DT zDX8p_zZ^=@G*FH%r+Id;|2>$Szw!`4DHDA%N#yabeF`nrb_&@tc?3rS; zkBw90@yu?%%_GBg405*$2cJlhrl3H}P(#9r4FF^nc_tnl{Jo2n*w;tvYq#g8rDd3V z5gQ#&5&T08Hz}#kDUh_i!(@*t=`728(nmzo~P?{s<<)v+qtCJwxCY{x%{y(bH9|u{mBWzL;eIiuClq>&2;)& z`+9}cr+4^bQ?Pe()B<~rGeNRA{<hS9W^-k}{@uRL9 zF7KBM_PL+y^kfhaH031V(|kIaJ+!qR&)puRlV~GI0U_<}s&o$TN(|32xWV`0dw4j6 zGVq$Y939P{j>Db6xu3D0l}raTJwoewbg=Z=ZiF0zX)0O8b|eC1B7lGxLfzd2oE{Se z&J+CA-?z6GBF;e(WIl=U$E_*3eKDTk@P9c@F(@p^!5@_Ih2~#dzX|K+vzpP0<2?uXfQMqk7}#EIrNK(s7zyLYxgdu3o}-XtBwO3+wBv^)Xubq zwFDoHifVPrPynI|5E%$*d*XJ8r&nA>n^w;Tnainths)lJdJBT9+6U zUQ{w|L&qm3Q=O2JQnybOCy<2g!x!0R^U`Vz!Kbqjc*izOWBP(PMg$E_T^Uo3!a$mj^xp-##wE*ze!v zT6WuN_j+}f)z+^a<{E-y_)GGZmq&lf>-oTutm0F~Zv$Q9jdm0KWqb*a3v-9m%pK8b%ph?k@i0ncM16I6m;9%J)(4!{li!7}oFd=+a(U2w zNZ@cvXnI0npZhH71t3H_ZMr|Wf%|}AnVF&dLj~q*!BV>XUB;IlHH$CT=dBsFqA5;# zMmdN889W)bc84&^3ncVDRZUg7+?sA${b88#Lfj;->j#v0&yIZB=viLxx^Q@uzJS|q zQ1g*OOcA{VaO{m#Bb%J|XAv8!du}TYF5H^Sr5cZ^Ix{;GmQ)f|ip7x(YMG!SmB^gl z7431Bozop_+f+&3ffZ*@4&jS=Q(IHJ;mVnr-H7l@;&XKk`W3njgjTQXy|;WEj!4Wy zx5k)v<8#s*(R9YLIelsc<;0a)?O$0QUfTr9vjW;5TUjFkouZE)JQr-RZIBt!#@&XD z-CIeuylK@s{ldq)-lQm;_6v;&7&gVDb9`2RhO5$j?Vt9%^b^tP zi+m%Hk=t)*XQJfnjYUKJ%#{=nKvQ4Ob$)yjOYeK=apilNUoPALq+m?y*0B2C%jX#a z)t^#fa8*Cfd3LkG;PQk2N78;Vr-d+hC`2WbPPSg>#TIQ3D}5{riOFkgYh0ce zaw*1CUr=+bvn9v-h7FAN3%(T_9dcIRgiQ*=;2tO>L8HV$JZ9=Hhwg7{JPb2?w#=yp z*60@fu5D=rB~iRo>=*?K&4$vq)~3i@+yr?iO9DyKIEW;IGJ`YdJD2o0|IixLWF7wsa*o03@` z{N*`i(ssj5tXdEO`)~B@Z_}0nVu_jU)FVk^E|1Lo-SdeOL_(*3c~hz_Xekn1J87i< zCgmPty(k7m{eTxRy>L8~5D(nr% z3hTw(8f>WAW>syXz2cfD$BY;0!uWEEDk1u_*fjjYIT`mlmP3-OW0XQ(RaEhy&rcCd zIHRRxgA9j9RLo_A#}=i}Fdg$ursr3Eqq1T8tRn3-fsT_vqiczRsF2*(N*`Bic zNe0YTrjPg)qJ+Jsy+#0&4-F&@r=jR1xlQoeL-qaYKSRgnQHdC=W*pN4_vP(-$4$$} z7sqpZIO4D3h}jC*Xp#FR?CdTnRaA8FKqhRQ3WsfI)J6oo%SFDC$`9<-L&~n$=S=Bv zoe|usDJh4bM%<6MdyW;SFh`Ssl*0yS1@`L?hHLg;chLD+ zxAxYI;2bcXxkLU{`Z}ZdL@aOIV=_^9iDI4qgMoq!UlOrxE{mx4_-Au}Kl#Ocm)8AC zj|0{^qL8inZBjpUqV8%|ZJpF5-Q02Qva7R52g(H;&%ELkjrFL+AAX+S>p}tsi}TF; zTD;s!XI6e4TqefPdh7&rI1QExF4o#E%<#?`Ti+j~*PGvOEZTExY6lM|<%ssVky24| zeae-%yzXmuu)5`LCgN*Lk3PTnk>H-8I-LFM^x;$D%E!{HCnYOoS6mWzL3o9ndTq=x zU)yH!Lo?1_yJU77eGiqor?r3+Srj34uYvWJT7g}%2(+2h>$Hl?EJs;Z`^D!F4a=~I zt(CYt_Ek#wk(|QE(q9?|KiEAR)vR+j@axLT8`vD-LhrBm!H&vLtc_e;&i0Zk9r>$6 zgc9p!4D^$tfxMF}&*EtUeC%=Ex5ql|INWaV+>bX@dJ%J%XBF)@?6ytDoQePX?+7d+ zBqoKfQP-vYCpaaRhmj^;0WaV_aP;Lq*G!VxA(>Ynqj-iB5SY4AOl7e%Ep#?#rp2v! zrhZI&b?BkJXbhSvZ2!0|B?Y?kU5LyoQ@Y-;I=!L1s;{cj>34z2Qzl{CAtZ8O6|&TI zVVa-OxEypXdCE~bR=f%@u(gIaxHU|hIyx<5dB-s`*Hiah94rk}j9m?yH_LhFSgC*K zVFisBbE$oHWNI0fh&|3ff?1j>FtA99{Tst^Go>y&1StzFY<#h8ay{PR(12ee%sp3j zJHX}j@qVXw!G}pfAUmrJ6MQWgY2Fw*hotfuR*a!}-Beq{M?qUVa5HFvydef9%Ndy4#=2Pf$SN( zmE!0ov(_XEsPZXSIq~#D6DNN35oGSBUR_f%8dpKeFu5fINtTqfJJ_jX_K4u3{7+UH zGy(QC8VOiEM|FC6aq%}hIKoDb$H2&gGrWnGfsv#8NG=YO#KYFF zF`659zG0r7OF3}@G4%9+q^F~B8|_E;ujweF*{Px)=(-c$!vw5#24R*Y4upI`bf2Cx zRQ^Mr($IF^dSY8@PGsmlzDN9%Wn+6ghtO1W3*KySJbQ_G_Ow8)Nl!-vYC(iIy?KAM zsyi=tG5dW_h?Sx=r}kDA+$*wRNtB0N&P=7?JO<{mc8qn;_o7-{uKq3Acy|_Q3-5I9 zr4LI>!QRD$xJS!ae2jgbY?IUOb=P=0nt9639M#Am1-x&bWtI7_8&r@{(R1@JoCiq> zl#x_U*V;&0TaTnt_iZ4nztL1_hfKsaT2V{YOy50UnZD9A(kzg6O?#B}22I|}h6*v3c-$KJm^-jy#;5Ek)a&y5DBXED&GP&s z(F5g`Huofb3I@rVaaaoc80J{S?9Bhe-dl!6xwdV?ihv*u($d}Csfb7mNXJk^hX@Qg zAR-_w-CfdM0s|;rLw6`014s?R5a0Mb_i}l!<@;{m&+o_ghwWmvnYpg>Jda~P_9G{b zBz)jW)c0G?1Vita+i`=owFa}hcJ+WRY2`fCX-wRcRzPqSSq=Uz#Qt=ctEPU;D#i{Y z+M|W2)65}eC3EuZqnDV0p##7H12?ju=Pl6Tx-pWArBpwC-LuT)3#<8^mLQzhw=DpP z#CGVOpTR`lldQ7F@`k8g5r@L2xJ@^oi)o59EOoguh<2-rX+&rP+Djm5=BCmLyk!Ba z>&`Du>BllE;m>q&&Cg5wmxKF82D8q^sD>e?%r8};zBT;2DHE%c5YuL6*GQ>-`x$E|4=phY z-;wvni;Kizn_W%^?7IIz*wl+wbg!(HuIXP1svEa(^w9P^r2cl3#ZADRCOYx?i5iEY zfABkRXnnD9b_u(4j`iV)3jri>&T&1WIaF|K@h;~bZ<&gLTnPnbyk-M;6Rj3&nJx3I z6YYIyktbi&_gBse3bXLi(DRb!`pD0-XQeXZ(F@{>d#YsfM`(Eav+nzR2qdf(ZKN4< z)#Q3QZjFTH5wZ0C7(A&0c}H)Nkbt6|9?wWl;Ry55v6p>`HFmz%_||Z_3vzi;sVbXz z7SCg8P3Cyh%u^OAOgh_{!};Cwc_$}au|{}d-p2;%qchlRwP>3phqPO~#fBB~N66P& zIROfHphdnJj!5Hi`ttUQlVhS{#+Wimsd~aGuWB+S3LO1Z&00!|zg%YXTf0P=(Ecii zA=W3=&d!Hj2OA4N4$YMd_Psz0B*eeY9Gv+E;m>z|QD719>ouuo9WsZ;62_U#WyVF0 zOF}j_4(s@W`+PQfLnrZF{OVoOlZbl zP)V|eubNwYtSJcw?Jp^Z;F6AKf}39&=%}?rc3(d@(`qHz68bb~zlQb6el6*>%%ZNr zptM}P&mLv@6u_RetS!B=(Mp}zaH~3q`Osq34ea4=mqi;n`cQ%|hF&1AuLI?Yt1XeX z6DxzODprCkOh$huJ3ORhDXQ=2`F?%-rUn;+-m+3^Q6|PgzPQk!fmZeTSmUndf=x8t zer~)?4N-Jd(Oo%u#U2hWhg?+-_Ts{tARVMvDnxAc>1&UJ;|#To2T|{0qAc+l zm9n*-S!6i7@qQ(4nC`pKiPt}GN#y#H54!Z>@xFORf$@<0jGCd7<7V~3EeTPnq&i!2 zw;5JAIPsZ6b>;B*lgDP?9JuD5qHEF1W2dVk=N`=(WR|Zf3u|cIh{N$;sBEk(D=QY= z3_2Rgs=IoWM)}{;#H@$l`6+YxIPfk=M+fFf5O}CE*Hq0~SJZ(0sAKHA82=WFC)sO& zZYh3A^k4B$jHf8~#wTnirQLqaTV~8#)Qp9w9HsomX4OTyAb&L`hoa$d^4`2w>XhBXPCjD zcp09LxwM5jS$iGU{>`XC19ik?ogwjh2&pyko+A$vLoGd<;7ILYr1s+ZLl^s9ZOye!+V1@Y|lK#JD0UUepm%g^M zQto`wK-l^n_57&93GhLwbk1=n{#y+9bO?)jcn%8*sRkNWh;owds3&M=2aj%i!JZ;R z`fd56FLU$NU(RY9!EKJDUHq{8fVDQqgX)Y+U(hD{r}8C~5rEIWi-tngIKD_?dF!0g z@M0~rh|VR=X+H*ygIcq|;y}*8YQ05)hA7*5ck{?$ANg8TgYZJYMfk#tGi8T z7G_~RzI=^9wjCrl94@+dbwhp-#q|r)B_dfhsq{fb_27oXDpBGPN(ze>4bOeNZ%uY~ z{332mr!NO5^W?RxBM!FnrV-u_qNc9rq%q$*sS>S5lgBRKerVbsl4MDPOSVLE4wIBD z8+bkKmeyYtI>#+SCKyJR<%LQQmy2V%6W6i9q+yV#M@vpeQ%Jvcas9eF&F8r}K}r9^ zJ%nF$`lBL{hiKTsNNq)buK$hM{m}4HM;TozjcRog*A?KsZQAgojJ8U5y(YXA{07O9 z=f@c-%MYvfryFo!>=R!=8pb}0lH?2{$YAU8BX3|CtnjkM2PpeB>-$K=q zs{6+`VDt%v)P`K_m8tudEsD%k&{OwT&zT3_BVX);I`Fj=XtqGh>uE}qbKMDCxLHl> zFsBEhm<6w>i~ImMLP@iz+XqoQ%Ko8`lZ5+xIv0|vua|_`d*D!FK0#YMK`qbx8yNt_ z!rDQQJZjwg$Qu=UQLCB_!vnZDyVUlLKu?ydU9UOUNV&3-V+-U}#RCs|6=7jRwsF-( zh&U*Ljg5XtmagYue+DYCH~Rp(>36VQ%`EO^cW&gYW;v+KP}%R`lqXjpbsj|OfgmQQ zcq|utzq#~4Yoi)TB}$wR-Y0x)?Dzz#=3%b!@yQbbIVYi!GqACG>pAm$DZRiU5DwS1 z54S!lDlagGX>~oW4A!9_*L&VDIa8HvvsbxLJ6zVa7bi{t_f;j(cO|Y?zf37Te}oj{ zO>9M_mwxN5l|^4?MQak1;^c{Z=7Gak^34r!p3mf~wD9KdkMI8Hh)~JymRAPdF%(by zZDN8jN(oa2vfZ;+-f$}pnAM}CE)jh4C^C}Di2zIvpkjG^61f|u)TOG?EJ{Zi(i4$_OC6E@r z1gm@T>2i`$M8F%snTZ>>aM+Tol-73$KD?RYZ!m7v;yPb620P}dUJ)qE<}Wj3huKeC zK@*b`nHT(uSF;;Abhz$v5w=5BYlPjez!nkTuf2p)#oxVhRn1J(f5D0R*1{2*7SxWt z#hoJL{%m+9=U_hXS4~m2Gtu^Pd&MQZ-t6zh)Ac*qC#}Q4)quBwPL-knEz*_cWk(Cu zAmreTZ5zWYKJ~S&jfG8IWF6zj=PTufU8(A-F8e$-lak0N+FN8(aX8l+W+%)ALVJtB==cqcbVb*XFwf(^%OtjzP^UhxZ(o znAqrsWe|DTg%<$TmTe*;8YOe`(0p}TA&4AoA?0xS!E5V)a&9ijo=)TaL4%!?bLClO zW6JK@KF>j~;-v<{mz50vcaKoCdd-IDr*h(uTlzH z-kVvdd@X&09{s1+%a{-c2#CD+z%2RBQgpW)1?6{f)7PY>AANfdKOA}-Veczk8 ziI@M0T&}RH!wzxW{pp(0JqhxygCuHYwx^vma3|SJz-=M>dmsLLGEa6imk0~RNVJt* zwRz81&E@H;MtL>r6OL;Yysy8LD3s~!iNprIJ)1K%-mxi(1qVMq4iF~OJM?S{v6zaaf9J~F!|>JyakD+kg< z$Lslmq`PMpwI*u@iIim}6?6o3N{*kBVX(apHO?z zzt?YAm&Cbm9?+mB#be?(3z43{$7o?E_2})d)vwx!Py51mg4^69YIIURJkIdhGCM}g z$O1Q1d%5J6c%l<59H1=$-Ao{^ThBriH%kSlBfkc4WJ@rmy|DNWm)b%z+{I1K6pot4 zMd-yj0!xcc zs#5~1Q};$!iKe{7az{}{@q%*DA&YHw#Puafmo!+jVf^Fh^HhG%uvmS_iGrP-`=UVY zl*-D+Y*)sS{u-9oE!PQf;M^djnd%yaU z5D=){#|r;g69$kygyyM$A|KL_me#SYCow6@TmxB#>rHW9K}i*vb%%+}u+}JoJn1=@gy(!-ye44Vd0} zqD7(=$3|O2TIvJJkKTq%p;#JnO#%K66*vWu?NStW#F!PBLags;iZ!T0<;YewA@{xVHz4Yi(iM@eo(H^rzQPJ?uBni71U2E{(aYOrp-!2E?=5Q}%tHyb8KkZS3SR zpcB(mq1L`)Gzdtq8sA(*Pvxf%qyO@jIAH%i_w82&k~y&qNul5#JW#y@`1{nCI*_&0 zbyY!B6oq514GG#5qfUUo{>!IAXC(YJ5>tC>>Y7HLi8IN?5&rC=pged{Fj}W*-2TNv z_tEw8<)^n@ZQILVh98m>sPJ78nb(1D>*LXuwq)3HGVVLPow`b?b~`0YdsRFVoyQS? zq{o*o=jAzop%eZfg=zL_* z=1eu5_A2(cMMEOI>25+br2$2JmN|&@lhukrL4Q@;^2xQ!jq$Wq@!Dm%WrcVjcZnezr&sKF&%BZ1#>$G)^rhW{2bcvvQqDC# zRDaR@#J$x%ncVU=^*AxEp#J^B(+_;!_HYv zb<}m|wM`Pc<%Y?QAvZbccJCFXT_#8;e@$i*T=r!=jH4Hv!FD0RJX*vgv7Zrf-b4;&*9wJA6a{o zW{4BC>(&FKKbEzAmM-i%Bht;-fEn~VH0WdZUyewexKRkB@>IgICDZk02|3Q|jX@sC zIhfzBR(ap2Dmx?s%=!rgf23+Ws)WLG0TPoV?T$IqN@)wd$u5|J+w|wtclp&v)`XSc zz01>NTZ`b_f5E*aBvHEk6@gGJ!<}Muxp+H=Qrw)r<=}A-nUT5;Hj)~5=DRc1>{r~I zdxic#l@fZr2gDaaH*X#jFp~I`bT7@4yCBPIYN(bau)=r4NG=gnVj#{UK+NNbEjK-% zgLyqw8c4jlDkpWZ;_7iw@O>>W7ovK7$ATlp9JlVJFRW3~&ohYncx($7#7Jt zzTyn4O1v=zGJ$JK!u(m>B0gl2JyGRyeA(|>YKz@s-v)lnPL8jxCBJ&qt(B}R&T_A0igUlpI;;m+Tf(+KbLU$Y9u=IYOqoBdAE?&o zzbgq0O)%>o@F~9d%3S|GK{0A<(1nicjCgW-(M-rL_0`?(i1z9{s*yz^)6MQ%-vH!9 z6g68SOa2%-p~xY`MT$8To&L1oW(E9^&pRK%_u}5KCcT%i>a5zEF3aFFJ7Rm^qMfkXVcd>N1dW?uzz0y#O~NVyfxZ6H_#0CH%}0;n#3SE zXMf1i8I0b*Z)0Q0@#G$`mU1k7L~22PJMPAuSQNO58ms4PZp@k(Y#yWsSjq0XCEEKk z#-PAT0^YGLo@aA82#|gCoVPntHTxOK`H)aKOtUNjfRQRjt-&jg6I%EYo>BZpNjPy6)vaSrgQPcPpw1;sVtCN%k$YjZJj%ay~% zC&oK?G%gAZU%d|?v@T#n$>V!K!s?BEHlR{XC+J`#NVPiab>eP`E78k$bgE}}5{agA&b3!p z1`%9ZqE_Eo%$(_)_9AI`i<=w7373Bd(!*^#F%DXLnz8=5`LLB-n8U$VT%6-=VNHRV zW1e<3>bhf*xU}=?*2~y3!_8JtHI&L|Z6ISMh9wJ5{(7>d_zRtEy>Hqzq88N#tkUe6 z-0974)W{_iJkn7sZk>AZEUh>?hPndW4qh4#i8!tu&Pmw-+lDD#8di8n+jTsr_#*Q> z6Z2h9G<_gjjK9LZC06^hhQv}$$a{KjQ43~(Tp>ZyF0Sc&m4eGQnkr&i!)a#h>}{9! zaa+=^>9@UcQ${@pCHk_KF>2&lh7@ZPtY*`J$U#-u^wU8xb0?mQ0hW&J>nkZ&Wwm5L z$FhwJ2l{+IxYnKzSa?}S|T=6ZP08SX!;&Yq^j zZCax}kdB0z2-GcI7qx(v*sZHShIaQQ?pn#8VyQ=FfA4w)Qbs4pkuXZr6XlxKQm4OD zh!X~zEHBjC+So5EaBG-;aE|YxQP#%DoUL~+Y&$8I;Bly}1K<3`{$$4BMfbU*X`&CV zQ1u6^?8j_o83#%4#uignp3m8uh5eY(dXs9nAX;b|z6#gk{DIC~+pU2u9o{n7k{$$E zjJ+415U*G~Mx|4C(32qQ?-S-~D;Vz914@cKn9p6jMIAbmci4estgS?vHfB)iR2^8= zG}K`29}hpwI-;677P?MzMYUqtKHoWpUL|^cd4ux7*3|nwq5EcNQFFu#< z5fhPHCPleYt_WRPY9vXjk$@7wdL(<2mTi7TIPmZMjC^)5n}z-gp#R@4vK8*EXieq4 z<^O+C$bA#kY49^dyFta8eeHEIIXM5>RD>O^T@~&vc*@mrlISdr;-}d0}8BV zrIQ}nbNOfM)4wF?ryPDwQlR}98Q0FU=RXK3A};ltrUJpXzqyUQt?mCB_^MFig-;Gf zsB`Nz_r@miwy?y0yqMtL_Rc4FRlGBI=arR|26x&kxoQB;&)>(O z1A(kMURJgG1cbl}WI@uR!6xw|_WUG(p$ZdlW8fol7?2>XR1k6=C3TTLx!iPoiwwX_ z%;k!#aqltOUxA}Pb^`tHiyb^kZvb2fWN@_XlCirhd3!gAR6|KKHI z2p&Kc-x~57@Y20K7dJfJpN}hbqY)NN;}*YF35$6D;+P*g=-g^Sdd)3*h@HCr>>-iM zes|LH%coK0bm80%7y$0j@6Arr`D(-`C@k!bKs9j0Uk{- zI8f8&vx|W32+TsyK`=I=y{C&w3Fq?iiUluKDly*&*!GG0kNCEDhB^=CT7$h!;|J+1ram-lX8=m z$F@wLCyA3dtFXYDWR79EK}qS(9)Zm+gCIKxJI#5I4Szy)b>d6SSRH6-C|#cmjA zf#tf+d5}Sj-1W@6aBdjYwWw67p%pHf=9{nSSf7$O(U$ zW*J!`XZVvM5)ftD-`8N@RHpu7cx5F-ykR7e!W9(9~iKuhdfvHYd3_KcgdsJqDhfq$A*T6~Q6NeQ1<3FqTlvKlsF z`HS~1UK*G<-D+CX8r_@zi`V@+R%52GBGu0Om$-LTIM)`L95Zwtxz3nW8ll#lQ8uz` zuYz1Bzvp-l$JZb`LudC+H;Fnm%y$$iN^R{AC;jLX->#;@KUo+mH|*E>}yhLx+Qck<2~ir=_KxBZS}s% zy^w#=lpp@^14+@?EAj6L9*)|Q7k4gYS0<#Sq|RSii89KE8ELLEg?-YEe;BL7!N%!W z)*AKw5)T1D02D-pADJO&7H<1E?AgKznGRB8rww)?O{;VWRlV9XX*pi6bVF2p z**kat`eSN32ew6eDEopw;d0s$kI{ebeEF*?K&`N7!dCA0RC=8Ug7* z9V=HEbi_v;v$z!bnlaU~IX!UfcDG98zdx4vMg$vE!c?f8SZ~3x86^81l2FjNU=Q1Z z$r?Ix*hBrKWuyS%f#Ay~r@iqbizWS7$I5<5!zt+&=mv{hre#^BwE*Fv;p4$J?3zJx z$=TwC?|8RA%xl;-i&?1m)M%#Pp(_PuyXu9nxHoUqkS2sIxie=z{rmLq1f&o)xvYWW zr)Cb-R$_$-x6}?z$O~71d^EprBl6eCP~V9U z;RQ1Vw2i&5fKLJ!?mqW1hM!*fg2dpibB;=ZZ{d>tZi>53Bg5Kx<3)%aDi@!PsEbml zKc7K5_tqC0U;mnIxt68bk7+uk?lyxfj-9FQJ6>{Lr3VcZnR35BDkgzV{ru)ZTfL3F zTrqTqF>>C=v3>e5eyh@`#fjQ)n~q4yzx|b~KdCX?Kr)){m#QI;rwmK9CuSgKWYrw~ zb3&p=Z#2!YBuLLAnDDW!!Q<@L&8=Gv^8slMry68#*bs|nBZCeJBe106x$47J{PrEBe%Q^G`9*MU&?T9`CE=1>Tv|H3D&5NIHSp8!r| zyS_?-IOxnAq`bK4{Y~NS1h5f&031$G`$h?sBWBh-{8RLA(I?tDcKzX7R%-%TDTElH zd%yG(xd9JIKZIkI(s;Iu9doqD&OKDMq?jU8s&aNQ%V}v#c|0=S?=ExoN0cg07!*iUcn?$0&@Lo$ z_&>?Co-Fi%f1*$TAIT_|QW=)QX;F~C=eT+c&M$>@yc#GujI)gGq)H~Q1pfpQ}qfR92Ko&9pUSWnuQ-YIHbZ2b=S(Lci-_B_&Yxo3MPa})FdCg zq3T<(E!s5<;07=I?W>(^(E#c+>GpveTgnFukHvk!nU)bu8-fV^s5WP~@U}Mz*FEEv z>O6(mDd(1{GFhA_J5`O=3lYutP5d_8PeS}px5n~PLw`X1H(8#MG`%*yrhfz*91Vp? z)@J$P%+IADYXu%%8V{$dK&H-=mZNdT)pC_#5J;iA?{C+CJXIoY(FWv7?^1% zs8ohNw!A&>PjCgZyHCpS{R9VT{R4@ z@SgEp<1sr_bS-RhC12p(oDG)*vE&B@N+(mtwJz1QrkAFBdg5EfMd60hQ9eER(8)eQ_Dwj6+lx1r#V`a0XjS_k2Rp+e!W(Y{w zTF2U6$zC~u90~*3`XAyTbiwwQT}Nx1L0efFS5#Jb|6|CX(SKE{ku8tS1BQacoW?xJn zAPu_0^!Rx7dk?RT3M2gsQ-xjG+_a=hqsUB;f{4vA;{oG?x(`Iky)Ei(H$7=fJ0?Uk z6)Q0Ge%K7tNv3FZyh`fKO4E48fzVOU#_7dXO(Or`95~Hn1t@lLX2*N7a=nRE{rfK2 zN6n2=K*%jZIzB5BA%*G-xb)C#50~GN%X2QK7C6EG-R!8OckINm45?rB;UxL-=$v9DaAPWU3)4be<|!k-Nr86E>$VdlN}t_}kcdozTxT3;9s2g3EbUmXX7J=k zgZCCH8S#EX$0uDtbfBlf%|^At+8K+0Gz*g@d%}<5Tdp~Oj-L3T@iuV3fqJE*lGS4T za&SAme?_=4>F_C@6mt62=9b@OZ@;SY8+eoBeaNGb)8{tP&&O|c^0b|;@i%86p0^tO(y+Q62#npnICR|B8|0AgX{L?wxHj@sE z1=K4Lcdh#JJxyoOg0JLnhW`#q98b0*CAXAsr{sA<(XgB~)QzVn{YuK>Kf1$)pLC=G zrRfGrhD!xV3|w0qW2gkZNxxUYSg;HX5mb{Uhdz!j^To5`N37?DyclT|&r6>B zT{*H)7>@o`|+w9&$c~2S0jsQLO}zbRy#^N&<9}p z_!_8uRE{wBk68s7GPI@nSdFu0nG3A`SbO=%Q6bjz{liugslotm9M7B4bN_tnS$VOW zQ|N7vUp+w zwI%|wz1ckhge7W}4yWMxA0j;5{0@DVZ6U|?&dDtvGegQe?aX)EZ*6Y+ecA`pC4A;v znF}OtVLUQCgAt8e;FaqNCuUQ>fYjo1n1K2Kzll(({dum556$)H{UM@gWBwV&SLZ=I z*WgCLw~LUGR1XIr%KYFMVNv@HWxGX%ar8Z~0YES5qJYQ0TabxtF4jSzjWB1%6Un-V z7Bc;EXS4(^k`L{ft?UZsGNe}A1d6|?{j^X2MgEchuA|jmxk9V~ZHiWa#}65OaSWY@ykr&I-U~P!?a-)ueSk zY7x)ln%hU$wb5yQqkg+YGIq;jdHIDSMRR(TX)dO!hxXq~mu*rR!WTr&d{%CkG_nt) zZ%!+7O&6_izq)SL`e#=vg`i*FZ17yCi@9D}UYa`DnfWbWmHHncS%m+!mPW_n9uLIw z6yWFEXnDtPwbzgL_SNPM^tLuju>DLTD-y*IX$}M_-VNd5obCuG4DOJs@k0T(TRf?a z6-&<#o&~hE#}9S;^?de?G|fVrnjV4(PFt@IN!z0{|IQo#5<&bbRKZ&3irCB!oVOo_ zqz6M3QJEjh64{iP2-jBbh_IMV^x9A|eTc(^_lJUN7zBEI`+;z=zpz}UIk*neO9Ot2 z;wNLzoi4*d?t2_u6^y>F>=k|5x?pS5v+QAbeOq_68Pgq*t~laZ%DlJ{l>Il_@E{Bm?*ofX!gnAUerao=_|`TU^%O5b}3uP*o%#r}z;ZAH3|`bg}3 zMA*OOfwrnecCr@AzE;eYOc8&yNEK;KKqhdPYfb-XmBe(cP!YdyBQ8OB+w8T`;0wSb z^JKmuG{om99STk-%RB?Q`KVr+Hk+HX_R9iRj>F@G(%6j|;b@ajHYBc{l^Htc7BW}Y zo5qKp((H!CgY)*jU%`jMmxni(_gpWA8;4pgo22OVgA=GzTwA!5@A=)XlKKnpmV0Q| z>>WJUSH~gC96%%VFbtM1y4#-h0bBOk!{;KZ+50-!k|dKqbxe1jVfIvx=Jw;5e?%ab z4XBt(<(%<-8jcK&FBRwA-~E~YV2-IbYOZNk#vF@OGR5$z`SH%2792HThxok@ydZ#m ze@?+_CA-&fBwu1~;9dOD5QUR6q$ZGHPG8~mA=0PsPqb%V5Kx!OT%_985x94silS_x zSaO?oaNIN_k#WlZK2_>)lOg6YDmkwhm!hHgG=Hy zUpoWl-JB}{t6FV`hMqDSA~;Pz09svy^&I;e+v8-bH2%zcVs_IikO&qjM9n*pRhAJE zp&h$tfWZaD%-8Zz zWhz4*{1rCyIY9xkGd1cJaJ2zE$*l)*VuzR{~T>$l+PZ>Fz5p&qAz zfy1yt-#bP2{&SRn+BBrd);vR!Xki?|+GzgF`yoI~jEEPqVK0=}pyj4nmharw0Ta%V zudtYrOXbg8B^|Hjaea|tL=LU`YWn(8i=#GW@v-SK+7(FFYGs|~8T=QN{p}W$70Vj+ z2uL}WbV9^|k_z6~dgLhgiQiws6V~`M$7&c`D&?yx{dH6g2IG(75lET8R3!1VLNe7? zukHurVA@z^Q6T{&X$X?DQ$zdjMVN}C3F=!QL9+`dA4e-=2(lwN3!9pJ@H7Q%Vi3%~ zyRaov1oB6A_8n8{De*HjCQ^SOPjLZ_FISAMnN@FOPAI?RH|(cEWd?MdgX4T5N^1S_ zO3ipH>xc9{e=uIOyp{p1wNUOg^5<4Goe(;XT2WM7UD*~f?Ipolo_L~eOE`|zfh9kJ zbuDFPFzHB;s#s9(EA6ZC$4L*pih1z8&$H1G&C=HI-WK{7HzlXum4Cd^8Y}ihFnlvnNTb=0hukp`hdu zf0%@XrJ*D0P>sTo%Kzvd*IG*A<4BWPTfFa=J6S?mbEI~|TPXKCAt*=sXWuYwbO40g z6ZOcm5`;;#64zM%`<(6s6yv`i@YIy0t{yfmT=$|QaNyek^<65`DI0M1SoUl*H20nw`c3R z^8Se);?JsP$M4ONCLJ-(Hl~q7>3E$`T&w`~E&8b`(++Vd(`a+{Pim)G*P2^mDfH*S zmhk(%$!r&p#lRs!kGTq(KvlP2IzD!DhD77`bfHVP$kJb{_ z(e=OvyKuKa3L&^tLWK{bTz@0ZWE(j}g)$rCdJs|eh8Q*h`5=!w;7L*p{e-7VdoY*8 zy?oLKQx;18(|u-DWfAaU#SZn5#9RA%31*BishE}p`0*nY3`;bmfH~YpGGd`Ke-TgUTt|Fl1nUTu`ORaPEnvRkHL*7gbdCJil$DPXR*;Cs}p zVtLD48*E$i5|;3dyHqS~tXUxGAR5)NBbJygN&!0#cy1|IQsLIPDH7$t#VDS5f9p$}Z& zwl;H%qI_cLZ|ME)%9Q8sFf6UKut+VCR?wXv0XJXsLhX4b3q{kF(@Mk5=)aOY*6fw_~S(--dGyEoWeJxqT_>F4;C?W zB0H+kkos7Wc%&lj`Nf2M@b?J#UuU1p9?)bB@ZJ^Ner^RQoh544LK_;PebtBJ&D~_s zFyN^~++@|_-e;dt(XC)s7Hhp9|Fllr60T31k_jGLLrXkX@izxHe2~CUX?SyZf598B z3a73<`x7eW5snTxxF(!V7NaAOg!B=y%--X#Ur`8)=XH7nXlF$*99Y4487fd|FQHBd zc8*2DQA8#yZwt?LXu{mn16tlW6u)dK#Q3-7*JvmyU^a~Z&YJvZ10^1;BRA}d3(3x2 zAEBaHuw(-z2gjX9TvqX)e*{V{An&^_D_=CRWJIq17EJz`m&O$%eDDs-Y?7GZVhYJm zv0C@~vutCY3Yxo^+Z_Um|L$RvrFk2m&)3vWBA}!8f2Oq~k8y zyrWFq+*X2=;+$G#kNDSBFR_1f)HfXuvDODY?W_Q8w#NF>-H4G2I$kN1sN$xn@})8H zS2?y5jm}kH?5*Mbx~9*oc=nLXI4>}A&_F^-?z;YA&VwJCJp#q6JN@^gWLI)t*S`9< z3YwOno~9nvut>GH{x~gWwjZr^F2UuK5V(hOF(iEn$MGKZNjS=XSW>Plb0DW*W0bwwr9pC%1#B{Ih4&5kqE>RMY3=P+OTz*a`;P6)o>2mR{^U z>G@vXyDbd7yQO~scxDsYl}7b@;4>>1me(xQ@l?oBfhZbC{sfD2?LkGLi($N0TVk^| z!3XuhY9AvNjxGf(j5mStNTnkQ`4cL2()#gM@P7G6e}<_`lD_l(3D>6hV^~@=IRPuO ztsPT$KoKe}1$>m6xWeh%!fbW#;po9!5zErfT#}S;fy37TQ$S*f55}YC*^UX{C(#m& zcjgr84rL2l$^O#nzac7FD4;3AdLB&?0T)V*KT5AZzl)IQMA0;_Hc!Zj@PWsq2#*pB zqaxN5kR0**h5#a6%|(V`JWD4Or;T;J@ALfLv6i*PCt?u~v?# zuL+TKGB*l+Pu}zua$<3DVy)Dc!6}ULjwb59X;xP?lDnA8~GAI z$y&c->OpJ?-Wadd?}zy^<+%RFWp@K&Fn&Cy-BZj11Ch$g5Ma&pZ(-_qh}radvrb1* zJa(Z8$U+_d@Pl{BqjEl_% zOOFdhj>M+17KyY-OL=rgR~gg&d&2N9Xje&-U99DiDt%}TC-$KCnD?EdLB%zcjjJdz zsc)e`?g2TJTY-iM#U|+}-qiNFRn`XcSevg^mI<-uD3a;oOO{GW-j(?d{Dx!@E!T1$ zo~73P|6a`<6fN59PKYB=vLBZAmSK>4DC~_DT)VXnyC9R#sll+e262Byy=^*&QTqf& z5IdLjx=%tStz|xtKeGY-e}onagnJ<)MYzlZrT;wd9*T0azh(g-H))P&)U^pju3kQZ z3;~lSLo+daB04NFZJ>}6q;kk&110Z^1yc+u1s05Pu&t$_aVqA<2tDCPaN;l%h-P;* z3`y%E{c|Q#+$K*GWbF7~i{ZV8qCpO|ORXnJI>NRxh0e*N7RuA6%@s|3WnB~dT3rya zHDXHfy04om9Z3fz0+H_cy$3)1;7Q!b!@l=`U*g|i`Q}a9AZ+4Yq<9Lrl@C9e>~7?j zbsxA8c)%0a?G92(nT026c?Of%){3QYF#iLosPhnK*mzTBQE{Cmoa#+{(jR;b0T%po zm@Zx=0+Vh@Tbuk##|}YaDw>W&UH!~3rQjP?{~y3SEM|y;`!1^ zg^J7g)xJvoCMklomghW0IdhHV-&%~^Y@{bnyxtte=5rLvwhG5XUbgC(>KPP_pleI(;(DHg$EF+fi(L#S{>g={j4-r&O!}JRVN9Gy@?rUe(?1W_zJFGdwj+H3q7p__txbgXDm(@`fO%^sl9& zmC2#~(uEG!4wyFCm8;$0t=u7LBo~)Gyo2P_MT`9rL*E?-VYTeIR#uNXsrV7k-VCun zzdL-wF!V1+RO)#_RMM>xix9@nE60OS=`U&7eO@0Sod` z_@xj}YUfxhR2GJ)?tsVw?qP|JH}Z_Y;^^bYZVfEYaSC_y3nb-`bghLq%l2s%&)OF% zR5(|*KG@|W5AL~byRMmH9nB56-}~0!Y!qFR)Z@RC2_xWwl`PQTN6Q}k|4n8*Ogx5G=u89Oue;jRK@afH7j4hLEDST^BwL~N6(Iz$|yy4KyDP_~-EKDHF5^C1$ ze^#O-Ans?m=|T}FOQiu9Hc;;~FEd?K( zmpfAK`V1;3)DnG4@5*C!0(*INl=dgs$pgSNu!1ja4JE~J_2ERHOHgdz0f!AnqVYe1 zGAV1He#WenR@(q2?k#FW|Bd3gOB$eH%a0o{PIoW|CEy80p4YgV*zWzyO^7dSofz(H zd6wBij=;WYzLTYuS6x0;2RGh6y%U&DmKJZQCuh2M{eP2+0to?WI{lyT>+s7p+(sgr z_7fNa-;*$QVDhW2k&;-{$dysRl608il-^S`AqJ6@I5G-z8EyKUvl5X3Xv*0w5Qu72 zFCb66UN`w4FQbygpl3C|4VJiBpgkUMe50-!6NZwb$H+p0$!^Qy3%U`Sy-ebSB=&w$ z!pPk{2$rYVp)3-m^in7-k3yT2@wTzPMS48`iqx0SUcCQ5LpW%39Lj)7-60KGv+C%+`^1Y2V|rfsVG|W;_tMmj^wVV1=EQp17F*U zNGNPoPLDx!LmJPLi~)q72hYeG``?fk$CjZpHZKIjhvtcgNWZH~hkX+z?fytWM_9?+ z*Lpqfwz`tIDztlvM*i|!6s`2>=cU0g?~#vE)$*8whBwCOL#*fAC6MW*2jEB$z0X8Y zCqMiD_(3%sM6jWcOV)!5LU?)rVS9|04##7aC7*qI>@V}XBewvh zCHX%otxKAKFLg&E!VRpAzymAXu~5vo1k$V}suL64W2MA>?QR2vnwL;ExG@zLwQCaP zUX3Lwyg{cy9;`%6d4*6ks0q}UsaT+IDEzMtColtfN{1A&yqRAP5Yvp~X?K4U;;fmJ zyoa)Q0Gt5Z>$=ZKV1;aBeq#2l1yIa)PU5d(Uk5gBlI{c=#6HRHyqgFcfcs#`e&5ZE zn#KodjztX!Ye<7>ue_LJc)D`&rrfvpzlz|-nA0%^?Smh6ONwtW&Zg63RJv(Gq&bRz~$k=HxITwQ3}tv^$+Q9Es(dM3Mh3gD0`fI~Ietrd9PF<0A;dRWXhS zv-(QelN!6?E%4iUXyhS};bwQ)cWq^qWR8D&dLIaRye>z4dhW<|d=N|#WSG5kecgzp zeMY4h7C6BB$J(e=5+t-mjf|^AiJiJ=gl3oL$^FBT^psxq@BP#uk8DPzQ{hN~JC#wD ztwf%vftOt-nm$cPhXd~DANlF@ze^o>{epz|d%JKINu=3Cv;ulrh|`Cf%Hgj1ZpO&HtFmI?j~@|;|P@StNKmE6-cu6H7+LYOcE#r2f#hIIxd{2*h;zaRr( z>RT7BYk#o@?fZ&^*IU@+XxiV>sel_qQR6{blUxTzA4J>baA9dev}Jsu@St;Gb)5E@ z**4Qjk*?d{&Jr^!Ve`GM*DNST#bJCq%Mp*Gtd*V9CYb@@Eb447!&HZSC^d2~=L?2` zOqiSQxx)Cii!9#21eO_K$@KfKdZlouvzRyaV?wLxPcm|-kkIgIdG{Wd9;e2a%Vxr0 zl>2PZ1lc@Lj{vf!~9ZmXj(5oS&23%fu1-D81| zxe38XLPL5+tgtp#wB~`CW}jn%%4kfa8FP1&$WcA z=IEV^EBNc#3l35ocGunHBr%SFQ`{|}lAW0G4fF1QwFE-J%>lADXil`eYvH}HO{vHN&NUA83%odPNtDn^p|*e3$1@A z3=?<0;z9v!;AK)PI;5RNI*7vq(9bVQPG9^jKogiK&B^6)@rLT5s$!I&Z4_$zn^7gz z&Up-1-SMpnTm?j)xSm^`3LCswr3ujW{22?c^3A#78qX`0h*(!P)s(6L73{JVBQDV zdg-Xis?SR4*JUO!)L_j8&L1lu_q_(;#X4Y*zr&m-zhQir80@W9;F=+0Q%lE`-=J)m z$Qd9WH6T-*3x?M=(*UY4Ru(X38p-UxIkSLZp%~9) z(ZM%YGnNyxV8)_V$qaUw!~>vo7aR0z*L(%psn+a;Fn|a91flRI9r6U@wcUB{Y;zg6 z8$bZlU}9{iEL9S}CxnlA9@yAK48%3El)7Qsou;_%Kgc=bF=F7G$Q|I0Sp!^>vX{^= zNHD6@=#m{{$u3u#FH1nc*xRi$q@k$_LjcA_h607+>pOJtdBfRAGMsgdawl5R4~P!y zq)iAvv~Ksy#$JLJyC;(f%Om+iY>B*(knGX(6jV37EeG+B!x{_ts=C`?>IP=n)h`J+ zBa##(r)Zw=KH~n$Hv!Q1&AgBq?z3B%jsB2N<yGSOQ%Et_zz?JIg1PWjl5oPrLbP74{ITM%3HRJ2F z1Oh0a>H_>qNwh)?s%i@#%N;0JZxZ~+UM2-cC0iz5U9V094Yj}D3C{=n_4wc(SFuaG z9DBIA=hYF7o~K(9VM}Yi?}+h9K+>diEjOsMOiy{n3Fn_C-w-6N4=TxcJDW;d?hUZG z6Y(awRHVsLhhiQ^xukep$V+0y`1)_zh*iK4qR0-{Q1Rjb+RN=A0AG%|z#=o;P^C9+ zX=4++lHFoT8k!Dn-7@y4i1Wpox#&c(FK?Y0xkqz3zmwta_)-E*>Qh((ZD2h~KXuyE zY9zM`8FKSxem!j?;~&>@fs7S9E~&O!9mz5Ed~k#g*vVx$qPy$qMxTJ|W>697`56p; zv|hcPYpR+mlm`1ENDgT|ZMT8~@-4u+64&ZGJ^SGrL>q}fVrKYe-&#PrN&O?6tt|EO ziWgrg!bnr*1z~^u9heM>3Ts_lmCEoK1k7V>e(JQIx5wLrWAS3~^d<(yN&5!a2Egf% z$s=BfnOWi=y2kY>B~_fm@eL z_=)Obax0(NkeKCAK$c?q`?tTYkmBx~2^Vvzc$wo%E8?uL)u&tXn3(b!kQxggrvDTf z{i8((DA9m%MfWdl&bA)Z@xt#~pqXuC!oV_g=LDn;6zYI)Jujielh;n*yoE^0V=PP4ND0DfpMuKnIX z*_I8(pM7p6$&LH(Ihzn4^MWE}7WaQanWaGxjvg6YZ0BJ< zfujpNIK5aiZBtXbwch7CYX^$-lYO8t1|;7Xdu~7%L2rVyjHmfAzW36Tel|zr_Za0A<@FvNo z2jY#MdesdTuf3j4Iqf=9QX@e4z@t#KUeDFNiC)Q_rN)F9>`+1Q7k^~8l}+Ny2w(=Q z|0hP(hupueQRj#@TpCxrjzob=68vQfERg$11Xw`9WB|-AxATym=LfGT-w6Em+^|U0 z$kEfM7~}N$3D?D^h2DU+Xs6WLL`ug-FS)@g~gT)a4XN~n(g1f#yi9pe<@_h~Te3mP=r7=~P4GyPv-{Fj<% zY(O%NfL+i|3^}@8Ma!OQrgds0mr3SB z(89P$z^50)d1tPF=Nhnad8!BXOwut$-6?*n>~p3=ynnCcKR7M03D`G^!I*QZv2P)B z7yf-x@K#Si#%yLj0HnWy!wOrdb$U+K_mE}KN$g{;5Kc~Hz=D)T(Ul9?W>&U2wVRj? z(PeN70!uGCT{YI>)dvrMZD5`3;g(OI5cs#I%}vaey4hu_`b?fp%*r-zt1s#1#UJB- zMV9G`7m!mAWd1O;$`XgC^CU-G_R}cPFX|(v76mz}j(1#lnF7kC`)8}fXs)GuGPZBR zaN9ER8BEdm3b6=p*0he$_}FGm3HOPz0j5^qgB#4ai1sX9OyRh*ei`2|Jnp4?3!K0gcm`qjd zPt1?{T+i@i^3wZl@U`vp!dLpVGnmWa zf)S!qgxsP-BG6jhuM{+h(D7-oF(MaR&om3%+p&00r(<0^i5m<&x+C;KC2E5~mur?8x@_n2~H8KAP9QlkG zKiGTSTo@0u#yuYuL0X73*$!gxZq?x-fyMtbub;W1sqfU2NbWMfsO3vNGD1@=o)wd8 zwj=oK>}e7q-MXLddX=)yrMs2SS+*|MI?@+T?}^-D`0NkXUJtY~@KX#b5)z=hC_KF! z6h%tnM1W4_jCZW32 zt?o75a%^)?Lg+#L_r5Y{>xvU07j}9&z`}w5vf5#te1wR$`wtc0IEVPXRvIuSh3Z_N ztMM&-J2Cek#X-9;rrWwM#dZsd#e_rkzN334k7=nxzbH63Pz+OuWzjYNK^r7L}! zG4xqH|IGERBm8ZqLpx&p3&&1s^0PmJ%f+M~!AKbY(tZ7J)^Cg<E`t)+JMs zzS%_hLa@?vaQLWJG(!FGAO`d2WEz-5!@1lJQdO`&RlXv9z8b6wt9Asr_n=1p{qQDOQ_vQlsYRrJB zD=EY8WXlY2H>^$BB+fAaeaWU{pYZ?6PlIYkyS;x3$_tndP+JtfOg>1WCX&@p0VoQP zH8yK&KP-T~NEEPEbDNH*eEN>d+dv|c_kOVZSwM@5Wgk^6a8$T$L7<=h4&H%9jZwsS zb^UA^(R{~Udd}cl1y#aL3G-?40;7va7keirGjt)efqV4eDJ|g4Vh2KMZUX4&{-q!d zc7Iq+VabH#Om=pHEGZ5}ugWHnp)gK0i)Zgx5$5Q{FgRWA{!=`+BjZuFt`r7BCp1)J zR*WH#7{6?RRVQcrbw{DP_WA{}-K8Yr{iIUwY8p_2E9-<^H*+0G*!d3`XaWD+KkA~@ znIWa_{5Cix1eBV$`Z4tLa`YQT#kOm6&N0)l~(9}4~no4>24mj03AfJOJU zRcyaeHpgbgQz(#S=OwW%Z1u_|88&L(Hq0VSdK{PElSdtWumk%aTb z_9TXESQK4MKl{;b z^9Yy;Us)5kX#mu#QxKrjM;3~{ zY&?X1+i8yL2>Bck@~!-ZJSZm@yrj6>9lN8Pdk-m1&HY4LT{jA7?zm zsleKcQ7Idk4HZJcU@R4j$-SUIp%`-(qkXbniGTmgCc^-yq68@}>};|QQ#h~^CL$w; zjVeu+Uc$BF{&;e2ekaU>)|C2(!m&2{xf}DfEV1ho)yR&CJoz0AD*z)UF9RyAKI$mX zx16h4GBq2s4ZC?#Y_Hd9JLTf*FQlKmcT%dMOk*#X|HN(8We;FS2AMY`{T z&@!dg{Y=5}*jEsvH+-+wuk*jc$RWwZ2Y}R5BAwHuq?oSHd1 zl>i9x(A>kDv7u$0iLjPEYRX?+;D3S%*cSAQU2*GnNprTucuIv%6$0MOJNF8YV{ekf zj*q-^L=Q3=x|`9Hlv_fFR60GE>C8JY8%ymI+oiq4pvqR6V3daQ-Ju_0Sor;{v{q z2o;Aw2Vu9tcx0)DY(RdamuKPFO(?@^^Rd!hF_#|YShzY z3@<(sKksrSlGdsUKne3*cJAGPIm>yu-|#i~(*Xf1lJm1QA0J6EKuL-~z_^_)Q*+w) zue%lm_i~O_#$&$3f%XAosCE#HdwWc~rWkV`GQbL$u2|s50@~5`7aZ42$F^aXC9rep zMx`bW@~@qp2h|^=kp?G7HMX%o@r|U5Us0Wghog2U zJmk3C7ifm0ZMqvuP18~AEm`5IF?DMU4UIw$!oCutZUp*mKk_5Pwn?}}T(#;z?%G6} zQ^%Q@pQ=}ds@&<2i?7uMVR%xUIVHIV$Nk1_xZYV$0wTD|e6N!3-z5(eydd#3$EH`t zKp1>emvR5lW_$uMCeY(8kfYyiag2LIZz`NPRiWIGB5`Hmy_wZf@m+2a5B6^a#Ql>7 zuu95&1aImAORg%N28_;J#`KeK>a@)7(u){CDI9gJ->*z z7PxG*+t++jT&t(LL1zuO5^(Vm=>DtU@ayD*phdyWP}{l88h(N|H1W3>3KoqSl4jPlVlDca-eMwks`SPdqXF3-yH&;@$hpzdyl0H)MyfHRE; zy^6zC5U{MWuYfG=EGLIC&UN1vN`nmWFKI5YaS1(8zbIxCsQG>t?7OmJtf#KP05QMC4>k|XZ;oX$meW= zH#{M_zE!Ejg18N;kl&;O6f{Z6Dpe$cC|p0sGW|9PE+IcY`zizh^WCSsgbRVZ0PRz! zt!JlzPyS^+%WdiZ`8x{_LJObM=517pzugoMbC}AdirL$%J>1E6gZ+*<^FY-3KDg%4 z8;foBIqb~=E4S`ce9U@Ej|Z`j>@hP5q)e9-aj(!yH3W7-Bt&woB!I`l%}E*MdHY_8 zW9U?sU1tmF?ME={I=LAoIt>#Tjmyrc=0C6u3^q{CxQb$U+_{0d!(F%bxOu|U#pOLN zIP7?1UMf;RI6}LqdOBr;V9h!f)}&`{0LJ8#h;g>kRX`@JC0(Dn&QIQf3Hj-4-%5(x zI!ZH9C$0qy8h^*~z$sacPJ(2)_q$~I7o_CBw9;(3d0ntuYTEvdN|WD};UeI! zjGc);j#aGzmw=@zoJkoi=f|hodk#1j?yz>YY+<_Pc0cWHrmn2)a2s8xh*AEaaj0RD zJ|>xMR|VR?d}syB^zXM^EKw(>WbKPHPWLOmn#t1yoKdg;JmCU%fi^v&RI*Zz>S-Yw z=Sgl@IY8|b^p6oB2)N0R3uL%dxwtf9$4cpWoUD2fa=tFGp4l?+gJpU=4l~d#7X&+C z7JQ8}2Xv}br)QgIjWP7n#Q}J%F|=XLe4ast@XHGGrxxj3e?|WmQ31d{OOmx0E$Q;4 zSH{&EQ8zdTmL5J;k38m48|}y*$Wq|4L8l1--!-(|7a=|&jl`tD-qL9e*#3Kup&OtRJr72Y5#FWJ^A9oM`|TwI-jW~Sm5=kEXEyG_ z_gx2Bi09kp)Lgd3I79XzW*mf-?<7`cZ|50Qp5M8-?{YaZD}1Y`v0x;mV>Lm@$o*Hl z9CQfamE$HA9EBaoB^6*-1hHN=8|lmoF?h4PT2=Xe?N(~jA_`s%y5fLd7L69=Qqb&gUI;7bwUXkvIHvE#H!ck@kkhp)}m zR57}+3OpvNnx!v3Z#lheJ$Df-fy`gB$0RFe)-Xmz?lkZY{)(-6gl=D(u?1l1ofaVc zv485GmmE%epW@aF+~WNemA#C7><7x{X8C@FI2?gG$Z<@MTQ>kRAM*&!qv{awTFQp1 z_*YSi<}dsrB4)Dcm9k&uZerNDg*q`^fSu>XQ@50t*6I%#xXaVRxlj z_V+g~!AOKf5-JecD2r`gsAB?yqgLmJQ_CYR0$iB_EF@aXfL^nzd}uXcP832q0~;Kz zeRQ7CJQYJW*xH-mXXa>)tSDGx&N8sM*pSH9rJ^dPXuG-%Jjb?Z`myurE3==}u-^Fm7z2YvDznBLTI8;1}1A<+|2=9-mAs;AU5-T_=V-FM0CmaCP6+YinK8$*BH zz>OHtCCJx{VXNT3@Z>k62oW)*Ht-JU2ix0t97+E<`k;v`m*r!UOGFE>oewQL4;c8v21LYyZW#%O9P<;PkOSnE ztoYDiH$pUsbjCGw^dfP2{o_`(uKUY*b8%?m9DbrEcn~aytzqRwD#A!|NO+DAd>lVU zh)H`qggiweT$j=ol(__fM)+q#meVN2g**YC_%(tHwn?tAtMBUQ$hLk{DIRC7 z780q|N*(!S?=32R>l-uVfYhpl9}GTZO<&2$rrz_C-S{W$9iPTi7y263DqkcAf>G0u^(_0$=H1bb=8uEHIHAWZahou^Qo;kin?Q6ivOsk&F*}PKNX8`?fwsRgh z)2xAc*b8YO4s3$|`8@Y}$j zv>$X*>`H(`U?gr=Qs7WAoHykH(E)GabT*NfEPTRdgATj46Osl@-1+05DcyUA*A5Mv zO9oOI=Wg@ON_6VL`G_ewG7k1?tzl&!K7=kbUF`Tn9RDj2Moys64}W^O;J(*b(WSD+ zVC0r2w@=pNabcb0gm_Fw<20AG)_-m!MS>d4P@hX#8#{rXsNr8I1u0Rpnvh$;)P_7* z4(PjtubnUj0|@Q_0U<8nyt3%$E5IptW$@90CYT)rgeCqWt+(EbNh<%8!1i$Nu--1Q+r``1zd5*-B?Y|_?1S29@VNrg-1WP5FQ9jS%L%}w&$y27iX=-r+ zY+)G$<4E;@C1?LOdGbh$oCf_{j+t$p%KV4j15Wo+{~t5bnSo?n@Y7@&efumG|@ zqi+i&q+QnKXNMQuRZ6a7r3%3Ro}NYYEAi(;$A!W9s{(po5isZbd&3_TL7~cV6`0F3 zKM?Q?S(_u!L`O#!E_cNK9OTeXMfE-&CYNCx>Dl`V(r=S2F^(Fvp+6Rm1vbf?C&1G1 z&jtjG3wp%Z+X3d2i=Onr+cbaLTF&`emi7zeFVQB1?15vB043~*E-6m!=&dCc=M4oU zvmUU95fpR~{VHRBreN19L9k+oCdHn~Mm(XC14Pv+aWgU!vhs2J4vfmtL!69WRLOsQ zE&mAu3Z)RvUfmux4gQ4PLO-=0Z{wnR#1jXlzF#c6zyV~)xg*7pOL0!d%~3yU4(2l? z&Jbp@C!y9L;qyx22$IC3RT&I(CfFiQ?AVjzD)9b4&-dp2B+ZQ>hXd#8bHyY>sRu-8 zB<@om0O(ds%TkXde#ZfZLvLOODpX-2q!>@cx5UC zXix3CdSvw(o?oo<@~G5!qtW!voHj=R7_Kzr-vPM|=V7+hRTvYepd&qyh+Pvskd-gb zr|3KZvcln%M)z|d7yv#NbW{ZsNF(MyNFyML!)a~j`v*eWHm12@ECm&SIehWRFb%7v zNWx65OP~xxA7tppB4GLl`fjZh*ztB`z+2`Mx-S$64bH$X9}t@c1%7<~JSOHYbKG-b zX3T(73$9y^f+*Z?cAZ?YfvDgd(2jvq)G`9>Q+vukMmT~Sc#|>^)jGeldE*BV7eeV#1}2$wZ%M;biG;a)05N3*G3pRaWAK*~n`fmMw7#z1M7pgwxBTRPX?a;Qxv#!%PO*e_^LrC$i;B) z`RVA8`B{`i9ddgHgJ5tR4w)1fKBn%S7D0>BwvXAAsAYIN(ev<>7H`e9gmGJgA5Rx| z3jPX|B>|FDe9Zb8HO@1kX|@)dmm`-@xd%%YGNzACP!}=9M)DPcI>3_+_%Y%@KS6=d zUGFDWNI86PtXGosZ?F`AoqHtUFsArqS@6lES!_>w=Q|YnS~e3?@BHd)P_0rZoW6G* zc{RgW)o>4IsGshP>G58h$=xIGZtw9zfhEB|N5@os@Iu{4{0>e#<*^V7cuf~&QD*LWK0#9 z;n>%d9XQo9u%}Snh_X$BhSS2NDE3AM9G7AdLjxsmFkD-gIoW z+OX?YESKudk0Y~cabpVD+lLPx^Obx(-BS^ZA;ApPZDZ(w2&EZij5cvAt$PX#RI4K+ zb=!%fcQRrOKCx~r=6MAHEqY)tXP9-=@`=D};O;>1Zv|ADz#KYa|B@91YOX?b@1(7m)m3LhB-ql#$MaLu5ilZm1>gbUKdv`Q% z#?a(NTk0R<%j3BVaOS}iE+OQy9QekvEgIeuUbse(ux&C3D^edE=rIGk&6Mm*8(`Fk z58~C^E`A;p8sIGnCf}jPs3?0b=gaB|-`_w>!xfiv>QR;fnU)7?rXYrY(#v`P70#Q2;IE4W@X>$fjnvoqwVG*~|2;a$}M zw`o@51}sn!DqwgRW~IJG6?`=wt59^Qr0fZp1olpD>Q5WpDumqw5Q-flb&k7_n?D`) zxM}-qcI!0GIJDsiNGqC8GhZSKrMI+tSGE%~L+nNSjEwj$&$lm$vu|QD4N~+&hYQSU z#p$V3lMh{d7|A400Q5DG*c`c5W-kls&zg4ptK4%X#1JTuY_?|zz7-zoKW5(li@3|fvmpEIk36iK>?@Xh5Zv22S$;Jy*-9;brl zT1godKuF969haH*ncxIcm9)@-kKn>JH2P>*MaP3+OklQI?e=raVrB8~xb-(R`|~=x z+i{8M#fj~Fh9VJ0-lIHxS$a%k#gboKSy?jjM;!02)N)kiYNQH8$nL%Jnw@@L&10HQ z&NA;Vp3LjA`)2j%f{TY+kBskk-*UADWa``g{9rv_qI_*fJeb)Dtqa&;WR+5o<2v$T5&we-@M z>f9xoKv1)6tyWltNIhvUPy)vB5Y9F;@^)9@tM`^Aip(~x@0@>^J&gb16mYoh^s<$2AV)YW$CGSV;Qs zibF+Wl~Mcs{QIist-57UGra;;czv{A>psrX4@n$oU<4O6+8_3KpQp_k-(9@Fj!e!} zoF1H--A#2EVX?bpIJIzu>$=~kx!|+c(caB}G>u%{jOw0&`u}9#X!pmBB(}TPZ9osY zpd#^C?@4iIVu3ostzd-PjCdl1x0u_3r~ei`RBIdimp*!7Pf;v@Nn0g6o}e`S}S-(Pn3YmH^C z%OCuNCN%8r(+V@xqou)U+XU#rmWJC2*jyPW%7s>~@SuK199#U=KkI*b0 zCt!7}_aZn~w(dqKdR;sB13+0z93iVdJT@@@wltx20eb3ie3X5?S2kZgJ`1YaRa++L z74^%U`}!g<|K$!KU=+JW4el?E^k`t3sGZjr`?&nt=Oa9}Zv?}(#YSpDUQ<#CQl!VI z+*J(nGZ7!iyvi|rjf*fk7q%zd;5hY(IyYZ#wdcD z_U5Cbtas3dD^NksxPGeC`pPHv+xjC82^vRu`AF4L0j{Cl4S77xJgQ3eLTk9Ik7(cS zT*FaPjgLpq{urnK_`b(ckN@?&kEjy~b|XFdRbeMjQb{dud3&6bKg^UjjtkEX99g9q zZb=;AEc6+W2f7((4*-^!Jt}FRSN||m0gVsHQe?!dVkfJfFyDJr6&L3URB{GKha|-` zdOz;4yw9eV1Zsec=EFU*1>yEii>JM!RP=$?_$PyK$NDKjMH^H>B-yyA7;WI|^RI;Q z;qFz2rZo9lZPBH50e9~(lHs!Bb2=Vedo8K%E!e!Ol7O%jF1_;I3Dek0>QR*iTQc%v z?`0VYczd!I%*%FYf4;HM&3`zdp>&pq(gF5U1uEG$Z#}P1z38(*j!a_N8GV_lkj__r zm8g9C(+YNe2EE;6#e`r32e@O|A{=Q=(L+kX}{)17(0_}mJfY4 z=tz;Kvut_h{LvTwRZzTRHF+YqGftyCfDJYrz957zw`4K(2-MVeFpw_cL|Zh4T3#PF zmwBH1p&QY!cldo}GIRSa6|AvaCw~&((J~VuXOiYUUAkCg(KK0GA0;FD1B>?GVyP?M zeHYp$J{bYse+VPGo8>Tz(S z{Iwe$W~2B>nYf8G!Z_U-EYkf*7Ppu^n-}P?5(dQ1&%P<o)c*!?8W6B(u>#M_EDzjW8Rm za6~I}*K`58IaHI-#snCYqrL18#%5AS1>0D7FpR$J8W((x`{+m5Vun@NSe5M7ZSMeQ zZ6pG_w8~aTQdcXymC$Dd)0?VgmUa2Q|HuVnwU#qeWEWylw&eN4xY-vkU3_Pt!(U`a zya0PRHqyZU$YkuOCt;xwraBxuVe?|IGTS+qVnDH!2^e5uI%~63e z!pOc^DpMhR*}XqG{8HL$ddl;kO*iKEGoQMrr?FH&^&0rJn|kU_D(591VKIVfGvPtt z2S>E7Mf&QkTb<%i_EGV?m(kcXC!#_#PhUCv)NH|*6=8T9gx<@BUGLLIU7i+qC7Kgq zkm=bOaoMO5A)}YBx#fIKB1{po-lfTEBesU%!!TyzaqFG7j+;s5$aW*xfwM;%_y|nF zk}7Vacz)Q!)~of5Z5`A{c9Qm=55^pPYk0cb=>TWmsR85ssw3lEB!Aq@wIX^N0H*7Q zp8`WaiT`jL2v|cbS&9PMw(bvSA1Ay-3R7 zKQyo!|=-E;vc?ry2{u1V<;Cp45Y>_=!fJ)ME11O~Io4tvaZ z=dFGMpPzBfbu)plb_|vyJnEKF7m66^T zymd8$-;MaeP?q1RFXJ;LojM(>0O2mMgy~B>TW{~}40}=ImYud+XL&|ERqa@*xTK(9 z>Z>?!=;D68_s#1d@~y^+Ptwhg1W)V;azmOg_MHxHr;WlO`gWMm6OLJep-+k(ZQ|mQ znB<>AX(+%ty{|e!%OY#S*xbF4nyF1s?CGV2}GZKE>>s{i*<8f(r24)1RuX1`RaZ%i}R?`(0H@Urh%}Ut&lg(QFwo zd^o^awjMLXrl3);$7jCfbN?*nHj#d9A+SoG!e>z>yPX#q5De3QsKK~rCjMA&ymM^a zK5V06hwE0VVz8M?d-R8GhhpXQI!6+GIqjrFrbTgA=j_19XkcZqg9oT1x3CwuIvupd z#+|BQbelEanO|YAai{k8LipX*+G(CrH6sr?pb2Y7i>G?BzTNTsTJu_1d^r?)iR948=jkNDrB+lm0xgIxud?GN$USwCjdhs+ zl3lB>^@=OB9?*Bmyc_RmEk>TnDrQ_GQVfD6B4F<}vC+mXrs71qsUp!RPR=Du*-_55 z%Z4#zc0ki8iAiI?^6pEi%+9?%=%S^6eusz|c)0gz5TkBV(X z`)0K*_{>e+RnrS9Z@6vPd2eL}jcMMaIxV|=C*sXeH1k##snX@^H{YM1h(w*eV-w$H zyO31lp^8-1@X^DnNyZRqlCTn`^1~zVWqhpb`aE}3_f?-Pes_)^ULxwXo|5_K@$1{^ z$#<$_@-2yzw#ABa<#M-%P1AgaKB;N;n0>8T^SYJu-FxM!d^Pz|Zbo8s#S@|Z7cWoE zH6rSTL?o^W!{O^Y9WLUdHbOi{tezfqXQVP_zS6Fl#Sd z25Xm>;^~^Qi3Y^AQyT0Otj6(J%7J5}$3w-&{t!L9BG2okn@Vu`IFKkZa^k4+R}~HM^w*BYr%WGnxIR+t&{R}C zZSN||BvhS-?UvqAse8NC0>MKKc;PjP2k1qg_6pSn`F)Q=#Qxp@a&Be~c8!)l#h%n8 zcFSzH1L@D+uruU1xn7(OMO<0vO^bu)YS_8bB>O>Y44$gIa@Y%2V-dJfvlSGe=4k7g zfLvft=NFN&grwGTc_dhT3?Wx}y!?Xy`^l@!zt_zQBV<|aokH%7aqU%1+YMf66UeTr zhUAOVu@|M=uwjNC3W)+>H`K*yaI5w_9cV~% zP3lf*{#L@vOu<){!u)1;$E^j2Xcw4%mGx;GZlCyWd0-4Jah=Vtynk)*1mD3y@)}*_*FzT<7bE%tl#z^)H2W~x#OQvv zz$3#7g>Z*Pg*f(KX(+Y z>u9^!uPnbwZoG0cB4UK4c)lhnBWu^3iZYs>Nd|oO$f#msgjPQqMeAFf?ZQCE6xYJw z&3971S?CXfzwb5V#L}Bt3;RTv6t7nghTF!gDv|TTbv|6jl=eZYg!%mOXp$18K&y$j zGk98&(FGl0?l`t6!`0}oUpF~MTE;f4t57d8jDAs5a(l%ouSPJ$ccAu^sww3a@*@ws_o0U^G_A@lal$_ zJ3y~9L4xU2f?-O#v-ZN*;mXMm-J|oKaBklYJS%@_@j}F(LPC`Ct(R`e!$&s?X#>4-eKF3}BL!M@VjIf^w_W=B}hZW!el ziqLAwP-N9ohP5f3i#LWbm{3gQjn2_XR`!`0M14tUGVors?wuEQ~EAp$9tK zbK)m=9v*Zpwhp>-i=!*;V=IvFYr5VK!;xlEVuO7da9-b>N0%PG$R}F+GY(3k6WEH6-przBC|*^X?!do@Zr!#` z3r#d~5M@mmc23dnkUMko%hdU-uDRu-b2R%YiFyJ%Rg+$)Mh@mBy*$P}T}F-ElODCJ zhra!O!(S8*f8<{iPut0^&0LOnnS*}6G!Dqe-}zJT8>|;2Ixq90yndL~zwdSFS*<(h z3F29(K086)o&WvKgr9n=*^W8c;`ouOJ&u}etBz)uT&h&F zyav8F$&MY2QHj<|$Y>S2;d8K)$-ayJwO&3=?T(L;$@8-P0=au)2Vti!t~OYx2UbrL083V)5T}obzj;1|jW;|Nw zQd>Q3zAej9S8N)7j)q`XtYnK^4(a?_Z&ZRT$OWHIibj&*?Av3IS4u7femH6KLaX8# z7~Pl)+gELj_(>mVbx_y!DgPsE|E6p?wT07zd@R{T#Rw+J+1}<&-{hn;ir%~@yL(~! zhcg~ih;RIk7T@<1wfzeFZN=BBpI?G*?XHMB;Q2)zk(+ztd0q`0x-e)FN#&BGuUpe* zbDay4{1v4Yvc^ZzN^Q;&wODv++urz?qgm; zg~CG~i!NzeBni~vp;-Nzk*KQn=PJGwH>lC$&W_n&#TA6xlc!3|a(>eY_t`!oI+nui z?0lGU?!0}FdxdIKa@E=%Dx?BC=r>jErpQQR*8g*=j+WhK9qaMrfk?o12=qYr6Q*Ls z@mq8w9v2$%dc1CRi)HDxo=OsG0W_Jrx38e8q}ww&8%O_8z!#MR-`2yNZ~AOggn4^t z#J|Z$!8}%YIclBT6OKQJ6fmlLyX~bN9n+6JxYXkpGX?JdWdh;S3z$-WLtNxlTBg?f;2n(BGsuBc(H`cdGXY zj(!T89>%T~#xdWVKMCdc$^Isi#ykpwS~>1;3rN3{^7PK+-L}2G3>fmJ)I96UIwVp< zv*4X1!u4BcDDbsw50fZ=(XMbk8}`fZe`uKcUeXpXK`P{UN8q=^-gY`zWaJ>4>58pg zA+L5F{ml1L=_dJ!@1pVdA}j)T5_fmw>mKJusdG(f?f+f@3mFHCMX%Kr)zu2t?R45N z>#kFVF8hUFlt1quij)<<#)eL-i? z0=K9Sp!{lnr4m$Xnb|TTZe_#2e%-Ckw^gNlW}7;gLJ&$_HVl!k_93M0PAeSnRjdXy zDD3>WSH1B{G^c5K#7W9_ZNmBhyLWy1ZBk2N zJItmZ&J!Omg!;@Lnlybr^uzc!V_sy#kj3tr?sh!bf}?PC8$r$&2w)RQii(5#Z;OB0 zD-PIMm&!ObOb@#!?@7zV$8w|795VHadZyD#L@-h%h8@xkf5$EoQeTw}Eg{l)wc#%} z)Vu{}FXdg+&x_ft<7g(>_y@j=_l$)TW7th@OIN$}-tQe!EP@1r{E%DwH*1vx5lT?M zAtCfjMR-9(Ip_m-E>%+n-k;)CCJ6+qt^! zD;Yy2`U9zI+_`653LwZ|XH~$wj>UWy}&(6Gk{I^}d z{!b5=%Xe!Pc4UfttGUy?)KbW!gsd+V0ux1UY6+RYxU|qYGJexCLb)3@Qp{Izxvy|h z-bIa)verWr^{2evV(Zs`fx+V4h4_f`5PJbFu?fCEgI!R90NbN5MG%IAy{e)u=m~ScSE&8$6a?fC&#D%B$ zuinwg%Oc-7<9szi{8~ih!B2$)^b|go0NWk9ufl1O?&=B&;CgSnpBNDICSAU_ut7wG z34XWpA*k89YJx{ZNa_>_%?D|wENj1_?T5vL7Rj$SCqXhA(4_GPO2LR3tmfwMMGQaMg$!+U>K5O8 z82`}&_xt*qc1{o>kq95Y4h#<$!{0W!CIQbNDleh=JH@pLP22eeOU>rFpO}jzNC}S2 z!Z(VlgwhJ+Mpw}&Z!^I(g|dmQ!SM&G!Ki%M2TO9y+tvL(Qbi-A>~9`K^zJ%s33KvN zUmD`J+7~e+p#w#6vj!K>Q50v+7IC9<=bDcQ$D5Uz>FLQQF`zMSG?lInghyIh2(fzO zsXzcSXG`o}jJPmF=+_rAm8r9O0%mZVy9oOeGPml>2UMG)qF!@nYx4s~GrJ)MhHRZj zSMv7==bgydBYl~)mVR*md@3_E2o~AsH~=w&MU(r|$(|*z{yK~S*r`!Vmqd< zG(i5Yvp3S9ec#97xjPa+;s+?T%q3F3RY&2MO=D)MPt{g6;%LDira6mHlI0C)nvCi@ ztswTwSFE`#G0L<{>`WVXMAu6UtAUhH<8x8!ro za4crFa>L$g80iMZv#M45>GmL>vyOGO zekfwhGRw!(V2&f(1c|}mEv@jD+vyF{QVT*^(%t^Tci*0Hyve0$6Bl1KfZTYmxcw$k*t0g7{H$O>H2LM=`DKWS@)Y1C6I93$}G@_+zO8 z_-@dvq(42CxSm^!Cg4je)xul){Ec%qiAn@di+X`cjKY-L`=4=O6U}U}xRCfJj&AAE z`Gt~UEu`~9=zVg&VKx3mtFAuFWQCRC{=nQpW&B_jUUxz5v6;raM}*X9Q)$})EKlC(ov z&^$9VuIydhz*o-hDwUh#M+D*b1cghhv@~nw*f{)ZO5L*GnV&<&Mu+VD^GRZ6ODuV` zsPG!{xiR)zmsIRV!yvA97RC1ItvJ32%+|0q)7*$5l}~X|iG6y$zPKEc3J~9!yD?J( zWa_mxy=6s`&Zfa|4_v{cRl2_$$eZkz&Fo2-__)m3@B1wFxvyZ0$z+KEl{kgw!z&H- zI+>+DP&wXmm)WQGWg{t}P18?Q$#MM$H{?Ez5i>5+9=rGk$%Q2ap--e$I(9HHE8fDT z4Mzvojcq?DbCcoxp~q_l38(<>^)60|-Ki!ri$^tPpKn_}yu!teZEG1xs*RP=b(|wUYdwAOhZnD@hD+@$Se=xG80_7{Ws|9Vg!sFtNln-+O_Vq%RaCy7 zT^``|r5XlR|KM9DC$f&|-*<$E>`h$t7b&hgDfCRD#VgJiENu6N1I4&ZA7&ID^;4-f|+WsC;WH^!PFqwv@YPBP* z%{|jwNqszIxTS=I?%WkR`-jB)`F4HGn(Sw(Wvy#PaKf@erdvAo7YGD-mFA_AEx;+< z8f-Rg*Ns6r`IuyD!C%oX9mN1_YG~W@Clfez6}9jbmG%a@z5Q@`zliun3!85-n}3EC9wAUYr(&uc z0gTi_Y_nLR@;cDzO|)Bi;f|FlWrUKw+o8M`bBdH8E)!~rEjaVu{LPzzvWQMHf;Q-c z)EgFpZSb`VUnhJ>7a-Z7X|8nXDJ$w~Bzn;OuZDW6XNg8UwcDax7>Zdx)NNbPQ^mMkw5M5w96w|>~4oljsNr!2y?m@iSkgjm>U1n)2{8&Iup+Ls4`>#)u)B)_>ix5{PbYDZ6+<&EQkAG6#y`i=q;De1K}$L9|N4620fzIip$87#0XZEE&=gJIbxuoe%|!c z)f4;mZI;trK_}02{4#QL^T{<^!LuRzq$I?@`b{?1+ctG^zb-Owh*oq;OeQ`n#V=>_ zyBl7Fa$zbU=EZI;>4C~(nx)%?2MNzIg23*E!RfuXf%ea6ua+TIJIH!>7 zBcrADzC=}hWk>6Yzvnj@A*B4ZP=7tYtyCX3?Vm~W_P1PnB4rF2rg~D%sF!$<#05!j6qsaEk96^lJ@MD~2MOU5sg8pwv4U=h-dhe&PP?qdy$;)3 zFXXptJw-8ysUHVEm{q4XxZlA_^I~RGb7O3l2PRuY1``OAJoBNO)pR+;fN6Tg z1bO-XCj2dsERy#?RdtG~F^f!I8~UOI&zR`grOKdqjG>T%^$3H$tQT|Y=HFad=+?=( z$Io9jsz`6`XPiFm$NA&c@JegUy8lDK&{&p^AjZru{1Gp1Pp_tZN}jO4sA}?Rds|(_ z9M*d)L5mV!O9qLvZH0BW?fAp-{G`#9E=pooUO$Wy{{;9I=o{^u1|!r}HV1rv%MCmEhE* z=%5yk`z881Unllk4aZj6vrp>gxh(uM{ko8irID`_&L{B>?Wl+ogmQ=U5c^@qv$&D3 zFTa{9@@ap!Oy549E6G0s+PvotT1S4hrM{&%zo5M!RU1eeTZu4C`!aGHoXVujUk)$M zoHIi2OO)=3TFdmry!jEY5EMKW-kor7kkY#Wbt6z=R41+42ZTfk#7u`g9Ccw0_Nxa< ztFfNz$*k%_Zl^7yy{nYrPBz+sst=5BDu4Y)Zw?Ooe`0gS-TG1v*oJx?`7ZgDoarAO zEdnu9%OF2^guwK|Ok{~;%*~Y`oTU#o!Gw_h2nEbW5=~w(yDP76@VmC8xlMYbf zrz!9syu3LYI^^kwU|%lbLP>++rv}1tK*GD2nkPG00+}xDjlkF^cgX8F(``9R(+|j;?E-O{S=92RrPK#2xEa&w{X!lqY#96#@4SIxr;i5RA zP2v*4e+tP#(|_S}aNgy>X?lU?1kRyZKXHEW`*%S3m~3^<1}~8!C=0a=t4{(edCk{Z z(vBhU7m9Ts__>5^rN`EOZWQ6xEunM)Rk2`b12S(Me;5%-i(3ZPDWe z5+EoQ_d?9=ut9YuQY~rdoj1nsDdt!c*A;NI}mX0@(;u~@~iSpfUb8S{-^O@zZ{e$bms@oKFsmLjt5h0y@chx!@#AhxBF`n zY!A7sQ2SFu!-0?euW}J?)!=Zd*}P?Hwm#y@Jm@Z2tB(8;GEh7UOU$j@`UaK~fmzZG zM@p`b^uJK?j#^+ZS%RVclk^XDJRH}8haXLFLoV@Yd$ySz8%j>g=9G0994rZW_3}u; z5^~bWx8!#VrW~9|PKGVP<5LW6u-RmAB4q|Im)Tr63}u~=8DC#6hf+dKvP8$?wymm< zq|C4Lh}5$$=DyABqhh|g6e-!pf{M2LBM?YaY-yZ~oXNT(=VbznjVer1mRyYM#g!hX zs0s{lSAJ-a-}@sH{^z6o*o5Sj3vdhCO?9^hY=!kZ>Zbuy)_!zpRO*?nZ(uR#FQUM= z+Zq;L@+k9d_jSKY9rMUh)d0rIzOjPWO2-r7cKKb|(E^`>#C`An&xfjAf!D4mW{!_7 ze2aZGT34BMz~E=aiXp4Dpe#2U*+ASd!b~rZCPe1nbS84R-0*Z`Sio zOP61d&o{8na>IT;!F3+jE2UR(`9Kr2UAb&Azs$DECsaty5SDfF>q@UXn*eIbRB?@& zCs6ds1WT-ThMR?cuq4L6(jM<-@2OOmKm-_i@TT`!Gyr@nO8DY0S-j$Q_G;|n-cE;~ zM)8{tlvLiY>wlLmY}$Zw#cXJOJWFFlX}w=A2_%1UdF9L&vw~0QeScsUqvSA z&Zb>hG+_}=0HIMUPR21@ogCD4FC4qUUmLiw>~S!^^u{#J^yfR&(BpyW)rXts&*xyR z3xH|2i9i>=lK+P-zF)z$c@ASZXnAx!rgs^%?^Vl0?;dO-^YXlf07X@b*VK5wGAJWA zX6RSX5N{PCj`GUSyVBQb;{%qk_P0v(Vsn8_tnm-7usGJ!l0)m@GP+#@k?;w8eS13% zyP~m@F&+D?2>$WHS{gj^f$sY-(X{ia2K@wztzi zh$%c$%;7++3~e~I!wBlB7TNk^$d)(bpaw(d&8|oS`qd>-#T*25 z<)qRkqibX%ovBLl>aNx3id(ZsZ_+LhOMnSQhCDWj2eG!5<~N;tXoc<8>>^}nl`ZZe1`KW}rF{S+?y#!CmEYho~yGt0yd$Af#2 z%&-0&@qj+0$O1?zPwkDy0a%9}5hDHZ?G->7B8yY)*LMqoUGVMv6BDII3Ue?qtwK+?@4qF1pQ&abXYhy$WEr|FJ5JB0TKW~vcOAF7$u6dM&#a;CI{bXg4VYTcf z4Mjo+%EO}WE5DF?Cs?+cVUJDHwd`*W9H(%%ODEsisBt*oaP}FM5N;d)G@MF{Td(nQ;mP8~NfHBC zHSy!@0fC^;UKeW5Cfl96ql0EnheXCvjVtnzy_@TE6Ty#nPzfvFYcbglP7nFcmKYd* zWV#M|o=~H(x0R01I#BnnuN%1to%BwGZ}!R4;hfsSm!6H;u>SLI4sQj;*9!j8U+WIi z0S(9uTge(>-I(NIpj?0pI#%GI#h1ZRu{;g|?$xcqz^hVKd0F;eoKAH>t@H7M34xPr ztK^6W;=GM}!TU&F{o6C^l>62U24wFx0fR|!(t+0-411Kzcby7uTYoC)H?{XW>3p+( z5o34|^uhMO4&?v4JtK7I`wCHF;J1TFf@E*bYik@)TU*L?FBP{KA8*?}oY2*WQ+x%U zJYOHG<4^lA4yCH<+c91|%zBp;?UYZxhw6!`lOJzDMd`+wqE{??*)nK90v~PO9Vu&O zh#8fx6A5~V_5-Z`aJ6G(0U89eS|4}PW zXa-kAcW5(rc?_tB=ZAGgw_wU2;c>C{(J|{AcMnNV!~B*zz_`Vzt{$cLzGX9{U1y-s znK)>84Hb86HL~loFxD3vfVFRk;cs|ce%J8fPR+hoCc#9y<`%8M-i3=ZDGr+tQf?Gf zb>V4-od_s2H)zR}159{O@&`y}xX)@xFNlxu&cU(IJGt|BHE zzr^QwOA^41l|q(4NP0yq>a}108Ff1>?t4mHyGxCH?ViBnUl#@d#srp#E^m&SHAP!m zFs6^H!w1P~MAJ4EOsqURK2X&xGTR+L@*vO`q_?=wnoFY<3#Mh&=o4#DSFr)txwn-C zVbA9DEDdR)8H!};Jm_GPHg~Ozm|5{`S(%LoA|G7oUl@xp&9j=P2Iw2K@kMo_y%v{o zUuq=Eit=9PKK=AMmz;N@iR+~p*VUOBB+ishR%Zjph_jwlPvGW}drPtKh>RK9&Y`3bmu(GmI@pq}YG(|H(T z78_zBQ4_1Uu!Fzb<}TpO-(99)*m3V==3pnNi-ca8TJWR;WugUJe(ZG9iKg9VdQpNurCS3jL3DFg z+?X41oJU_M7r)NE+q0hC4?uZ?(6U+qiiARWp^jPjXghA?d;V{le;rbCQgvzcB;$8i zX|$56+o!&}Ee_IXzEDvgjubCfx1g{p;*l1h6&{Q|Ax=r~0LSAyOc{8im)j7WNfG9Z z!nU$opm7@xc5%G_dC<#`jxd;jna!C$*GSm9e2(J3Lgn!1N2JodtatFzLX)3QZLu(UTVY@l zc(8R}rRmj4JJIx0ROGCe#j2`L+|*^TtMV-XG#N;4yB21@P$y`(|(1rC0qd5=gEuZzG#>2=e^!fceZfh~H@~7^k>aLSbUhH_Vj`@fLfcMYu9hSd zB{-f1AcRVWjV~VjAJF=cepNhjn=Kj1z`@aVv!4QOUtKIFUb%W_MB2yKLI?*m@!-}E zvs(hkRXLokc{0V6Aumi#&4=OYG$wAc;g+9T{l*e8&dcb(bxU`<&%mA&V=t@SR+c}w zs!=aw(@15SfE0S7D0ahCE(u0yTI*98Qr|?X3dTU!z*2}h%_mwVuZbVwpJ-bg3b7`f zT^sNUv0b^ma>iA+%YCgW%71-P+BhLIt6%VYnIm}oAF;%0WxwYWjuLpq8DDKbuIhvO z{+f|yh*%p80_Z7sp{4Kp#w`VwCtD81=#SWwGX^t!&x#g;!CqO!#Pb^fMFmOk2j#VJ zasng|E<%Un7?>se`)gGOnRc!Fn~d*|7jM{iP$jSfdNH8ML?CyBR@0r4a`b=BP%-Zh zup!+THlrxB7YBjiq_b|l4L*!aj3936tIJj0!>r6A7i)DslO6q9&TzlIZpR_lt2mcjxo8?1`zNZQs_MZ3ltbZDRwL zv98)*>op|Az-&27wiA6v7l9Y0B5l{MSt7AC%FqhOi3hV- ztLTSe`$PvW8jPJA9y}MSETGh=eun%kU=Us(N7iPmdp2jE5=_XPKL#VGE7oruIWET# zVW&>7%twc^QnezidA2yhbz_^o*4H}S>}*~0KlSvl9-2Xc+TevTNJCd<;?~6rJb3j* zvSqo`i*VfxMpTfAbHzaY1Bw-_&uuY64#P?OHrr>Km;G+X#0FOy6EL}*l$@npL%=%! zhtC&L2r4KUWB6th+@|(_eV=lJ`fz39xvkmUyU%n!yqej9=Nb{OAa+6t=KnDm9H2-b zOo;6{sriiie;WO)Qp`!}tIhVV&TJl4zOl85^fT=F{u<+N@XFTJb?uCou3dWGytx^! zL2QBwY4ekz^QFp#^RAn0Wkvlzo!!&`n1;tuf<0+KqB~X1sV0BO`~Ba?ayDdpyZ2VJ zR)URtGrgL|uF0x$ik!eQngz-h_@A=H7~#RrYd@xwHtq#`YZN)6Wq74(bceDlIaOxm zc1ARrbNd;aEq#e@!>DSr%IMtDFV*u~RA5lvojV0O_I^q6G61)K(M{KIxd-s*Cv~i# zPjoN#Uv!#1YQ%R8l1zX8Vrcs|b;wkA9@yNb@3YS)WGJ5+J#TLl?K}{@MGQ70v@bBU ziQU8ySo7LRx6M6HCvD{NbQxeZ^xV-4hJz{}FbFkzs?ahr8xXiPT3%z2=NV`q7hZl} zW!8Oc5|~NWfO@W2AFN5gDoR)=bN9pdvz8uxD{S3XBx-(2mt-TOs_LZp8z7eI@1p_l@nPbj*++_TnLl>xl%aqtK6#=iQMZgY4!{wZBDvcUuh zTpoN-3fxT_m*5r2k(+p>3S}PA%6z%D44Hj9D2xb*Jw^oRwykS_sk4h4)JgLK&QfQy zWf$-8UoKW5iN;8=fRqFW2iIP*I&K{9cO~wZSC?5GHQ32hm}8l4XISMc-kH&#Ozecz zLGwYm8!#l|WX(GD?^})JjbJLB#u0*HQ|iOmuD)K|?HMWeHQbGcz;Ggva11duj8x%aJEJ>;FKZsBc7p(d~76z9~uox z3R#+{Y;I*=NnPbVpYR&9z5ApvrX}IA1V@~bdcE6Yy>=g3su#~)`ygo&A(Fel)QW{X z?C*Ik7n?S@KVFE-KA3oKQTlJ7rHl5Ct+O=T0UDlKwB?$Z&ePV9wVj?2*x7pA6=}r< z9;h8sIURdAJ6*O@c1ugY*Ox5BBDA&U3fzZr`9ZcA(Kx1iW4H8GHSgNFmmxoQE*twO zJ%JO{GtgAOAEVFUI@AXNF6|q&)WIPE`C{I>~ z%ENl)f03q;_WW3KH!Q8O^dP}2qzYhit+grsdj-ac0Zq?lBrJ(*7*FD!1SOAoiO(1L z` z2h;z8Zooz6|r?#n{>8S$5ghmC)lULJZRlmbEea~MoE`QBC|B6;lDAmd&*y!!7&UPWJ z&D*Xc&|UesG`Q`L9+06v8UE~pH<`5vg%rOFPUC2}+Yc-Hz!E0*U~>S&0%4|ERrMYS zl(quS6$CQLRh4n#1n;lZ+3kpUYHlXK*%g#hEq>Lr1Br|1HNRUWh;FZ4oIv&>`}dux z!C#_yqrl6#JC2U9V|Kc;EE1<<{4sK@xfM|#NBV9!-j$yQ=PpKI$gL{>m_~>IUab-|`q#IE5OWRU`Nl#FkP9i@2t&O)%Qh%qsj4W$HCaFwBZ zI>1iAfpZ)sP`h;V7HaF_qUt3h;LEzS0r(V;NvdAk@4W@MC7HMp(f;J0TLwl+7s6Ul z94bDu1dh^K=$BN>YQ4zTMYZ8NH_aJ)wNxnwJ7U=0h=5wi^4{j&vGKn;v#f9I_a2qN z*BZ>or@(7`=D%J5a^>;3Q;-j|FK&UF;%=ZdJ=K(YG^T2yCkB&iE8BSfkT-r4Eq|~D zAW?rrsqVn^&b`}}%692*8V@-k|NBneZy~=>?^8Y{L{cl|yO)VpGf)!%DgD7Xin_f;y0+*vXXPYHXX;j_<&8fKKK8H0}j2?Z}C#CX$?ey&Z42y-~LF5iTi2?K*Sa>t_ei=YNg_;qg6+@^sMPLhF^ zUy|EWS%8W4yy>2{{wqxtyLdaPswi{+D)O7(a$e^?z}q(hD@$UJ@Q|$S0*(@(5~D=E zMheT{PY?u^I(hodR_nKw(}1YJ z8XXJD#Mbq$+g7hX)k#BgbeZ@WTWVnCsbgpt>-QomT!9WH| z@UqKg=*kw{2rpOq-)^ELq$d1;%>(=GQCC9;dm=U*jHSia)NHGLj0s!!%xQ8j&W+J9oWm0-Ne>XjaiRO|vjB97M*td0 zqCbjVsBFfUnfP^hY5GOOtP@R`wCLI&fgJAbYj`6o@3s9^KYSvR?+Hgj?K%GdMqs%r7<%GOA}(bYw(fs42Lct*9e}w<8*8DnI#W;`juI2`k>RAjlS2!giV{v z+4lgwH7L{O1)PE}ddYUZ#>rt1Uh0bX&UXwIw^JzzXAM;X+mb8#Vo6`9l&45 zxGYt@(`3838y$q!bF#9d(>r3OCe%>w$R#6DES@4^u_w=4N0g#S;hu7^bZEO;ln|^e zYW@@b%RxS=2jyD%(`JM-1_f1xv$wLOiN3rAZy+O!h%fD(%* zpZ@9f{40O*-G_uNftE0#*=(~ykJ<3U;|8#^4vD!Q5>w4f>V;j}!RLH#f7CpiHP@}! zRDNBLEx%5Ek~m4x=aI2ZG)`LT##F*gG!7?Sp&wG9gsO_9U&hHNZDpN6!_CI_BlcI} zZGyTTs|EoA<^OBVE(q`UxFT3KcD61ltc+c9%TXxWBYgZ`SJWPU1t?V!uC$=yL$j*v z^maP%pcjY1;@dA&4DzQaPuASXsCG-w1cfJKE_O9fxTdfHWdv5Q>do^#yIQ72{R_<4 z<)lpHz%3_2%?_YaW#oUvEM_Y!Rm&j%zq;X1cI@VX%*@P!lFx@gXoRi)c053->Zd92 zw+z;$xCeT+nI#@pTa*xZ$%~5^=J=Nr2}qgnfLT=E_tvd6dcXTyLnla0*lGI5Jth-P z?9wZ<3$6`Uf&CiKKPy>nGJoo5cpn#M0#mvF%(MSpe`Nv4iS`1D+Uo=whVV=} zTzL1NwVD5eG(VfxFJh(G+~8hDQF{e@1Q-Y{KIR|cX|e3n)1L{)#QeZ-zKZY=Tr^Xbg| z0BK0Dv=A7uR~8Tn(>`inc??o}`3HW*FHrl*9zopz3kpozHQ$!yQ!8PLZRQ|~gq->+ z-csdug)79cERWCkMu6Q7w_@+|(YIoDCMP}Of~p1O;#lbEHpCixZ~|@K7bt*Nwoy}A zKFGtHQfBUGRK6lw-rY_ZAXE7}w{K(eu=i|$nU`V6!F8YxIdU+@goLl(-`NUlc}usq z+@)8JZVkOnR+Tav?H7}9q&4M45>}y;;2G zS31#%iw`Zt2S3Ie#9gsCy^ZC~Ih8(^fZ@Ksh#53{sf`XZpw`JEj(a6JEg|rn#O&>K zrktnTB`6KrHM=8mKVy=3Z5wk5)8DCY+o&H8w2OH~518pHs2v-ycdyW6S6=uA_%||` zNp)VnPmxQBXz2gm`VFSfQp?}g$oS1QLEc>{V+mrQ$b-(0+1T0K_^A#!;N!$;AD9<$ z>jJKzkd?%qdR>N(;*5lyySLiy;%{2M2w6ZEpSR{fwLw-Xmz%$y{CWm*w9( z?wgq#K3&5hRUU|ys!NOBjjHtiv)0X7t^U<5P(N|yQV*08@WS_Zuo(-O-C^yM#d zA>LOeUI4He(k$<~C_Nsuj1^?TaMuM*X&X6$ff|xr^K@W|KBdM)lYvXz5?iqdLZ{mN zORqmWLnt_76R5EHUqnUiKh<3Zl2konr4YX&v)2`5G;G@TI(s|Hj3CPgL@TPyyzGw| z*~*A2cuBU1o@Bmp`N%bFqD5XySV+5Qx>L=W**XMhH&=;FCCg&;(umY_9m~Ko2|2>6 zi3JSzFk0kVJ|MS|$r;^(wsDmSX^(^9#g)G1mq)oo8z1Xez0+&1Cu^jTg0;A6N?Na_+?lt#m23AmHpT^Ivp%yDn-PmHZ}dRej)vHwyuYF$t;yxJFCbepf4P z=S3&vSr0iof4!GxK=?aZdtjt0xM!)vb%t}S*1bHpZ&JpmdNqB<0xObH<$An#4_EBq%=4JERD3mKa(HIZ5DYqxU{t6_3GNN zTlLvA^NdI^vc5bt2(KIM@b)@nK#Ft%G_eXfS9s~~`kUrK8B~@wc9KzSZFwcE<;VMY zpVCrX%Mu^uTECFJ?fs8^GfbsE#N&)4+abJz?ccdcUY?c9^i4;frN$^SM5xjI5eoT+ zt;{Vx2rO(?dGjK>meBG3L3LC27Ebqsdnc+SCN~rChX${nJFwR%E=sHK%inY3zcQTv zq4W}{%CvbUz*;rpbVKq*^T^-l2NVblcd{!2v3|j@iN?$GYTk@}YoWG87|ra!%3O8W z`qiJRQ-Y+I@iF?LI>?cq_`2WuMFAxbcC^#Y<(dB}P!KAWFPMoAa0lm+c)ckwDHW|s zi5^vb0Wn=F&y@AlrJ7Px3$>rtl}y+U~US!H}7nyo|-p+V|~D|)M9&IyT)?CyXitOT<25| z*e`$VOt>R6@L=a+KaL1LMACXt2|SZ3y{K?Gb9G|l@y~r9azGe?4y`MPGnZ4Aa=Wxc z)*@1;&Vp|fK^(i^u^-Ebf@vs7lwR94qycZi;-vtI}72VTqfE6HJvxO5pvH=X*=2u%i;QF zub%>86HTc`9E(uP6J)BkE>p&DHeZ5KM^$i5;Soc8nz>U5rM4(gbd8;^+b3B;O%o%f z5ZdWE_BDTJ{AVkLgeycq#{sejzgzuKoF_46Ys`dS-*)ZWE7sk`@@-L8((5;fkigd; zEP=eSuQ<@9=)L9%Ga%AcZ~PdN|2nt-oTHPXokj2I3*ltcG!Kx_YG+@3e?09UG;x1q6VMu2w8HS&rz$}Pg5Gma49F^1TEq=dg{M-#FU2f4<31DdW zR&vZsAarzrmN;VX#M5|Z22tQ8fPYCbSk&ZZGx^6G;J-Nqs&(qLgGF9dodSC^usgqq z)?}P+bDObQyY)iw>gH6BZ7zrf_aKQjdgJnG8&c_}A6qI;kQsLv?#==-0OB)Luig7h z4tci1-9>J*NX%%;xmTCZL^(0F-|1L?m<{@Juwpy(L8X0T_pyfsw|h^U9L`@B$=ZnX z7w3^v@-~0%qaq$KHYEtxaa(L;mf6n4dz`||W_12FeV-i0Y@q`ygR}?fJOFkCQ_UID z9jil@)$zL(AM$^OYA`9Sfms)mmK`}6&<1k&;(Me zVT*;i>hhXglDZy>BAi*CQ6yoIak|~2qh(_~R`^kI5%Iymyg0jZqFlg@tdzh|A+#{) zRgJJ3`Ba!;PaKQCY^dGe*9Sjr)5Nd!2}L!DrVXbQoO_o~*-2dMxr?P@0DU=VY>EKs zN|bQzKdV9ickQC~;)tm6_@1}Y_u+Q*Yuj33`1)Fn|6c2Bep@xRLn5W~O}5wLOSagJgi*-Y0{Kz`(lN+VbI`Zu<__Ztc;&Y6>XFcqQG|i=T zpXz-u_Hh1>>uD3Ivp{r>*EKHgEkCB-obLvnt&wgm%OmI6R!*e`(U;g@wZTl|XG@tg zaMtT!vfNa;vJiV*a&SX26$3^u-Ei*WMmHM3gR7;ezm)SU&$g#*h03PpD1`j<7Lh{j z7U(u4uxLCokMp?uao0|ht0?5Xhn~yqHc>|ItjHGqs!fKwCg8oDL^WtX`_ta5k-4S~ zk@K-G-9@}l#z;-9q)w6((g<%Ft23l5zU?d`&3Rs3yUbHGuGdrs42t(;0U3}(8>qny zLntZ@IxSL^0s{ zyo2G)ZjBRq-oMuLo}N{wK({iht$jn=*}MwvyRV=;6bOL5jG$@*xs;OWC2LT8C@+QV z^rxNtsPccjAk*o}<%$kIZ__ik`Gp?HPF*li{O zj&FxY-8&N+$WH8ZI8!%F=Fh3=S06Ylv!JH@Hp0AXE_*=Y?6#Je^W)dVTlvtljb?3S z(nGm$P$^z5U~xEwr$~gqCCFQbSa4{=)Vv4Qc9fh&3FiOO@BtMB+4tbel!J%bU4g<^ ze#x(}KMpA2ACzxzi`6RAY3UzPdTRQrmdL0BIo!!->Sbg?ihRODY*ze&t!ry!dlj$* zthiA0R3D5$fQRxR77pgUWjM-%M{^9~a@WqPodC<>m*N{~alQN2h`j-2U%Z6x>eW#| z07Vx~ z$~8pt*i$3oiAHkk`YsJ4z3&<2wLD&Wb{&~9xBz4?KPZioIweGm&^UKcPSczqbo3Lz z%$0E7KyBhA@oNFz(|Vwm5frKlps;;=BlTdo_|oLRd=F{7W1->~B5RYvVDllFu-|%h zgvxyL2T1`aev3U=)6(QtK6;Hvf(i=arm4z=d9VxK#(y2-zSyblli065f1(}G=b_O0 zS4J0b1;AC<=HshO&<+A|z*kX{%K+E1M*|R8Fjhw=*gvM{;lRC^K$Y5tGOR2H`>M^WMtjF+)EW6+%XU zj|?~8a3fKBm=@F;(0S#%Xu<`L^!IaAx7N4Gj7`qmmK;f1ht|nlX!mxUKt*ie+yypr z!~*sB_N@vS-hCLR8M$Fu2H_gY1?ymoU3zWgRw*azfFkWkVjvFcP+6SZ*sd18qzNh6 z@mZ91pQy)Hkx*N_LV6?^bLRF{7!hR@1E;QliW>j$=B*Rz*#ePKaW?`)??oZCw)f2n zzKPvuC?*R29pQxh9i&baBc+g{>UXdUB<2@?2$-8+d?c`F!VEefA{96Yzbp>av5D}* zBER_$8v9QVT}+FJvCB?5u+$#kZUCf#n$Z4dDa5j;eOk-z+s4`-mP(o1Ykf+WkmFLuig4dt=$+Dw z8mb<}{Fgzh*a@q5E#lMWl>u{M$r9A#Y;WS~?!UgEr5vK%{5|S-PjLzmJXwjXxGs)9 z!Kd!<^BD%paJQSph5sq*QPAu3kHI!BgF1!yQ2lCxjd5Dn!?LeCRvV(R}#$Mtn47>@Z7y!R9T$ z>4*Z8D7RZ~3BP9qYqbU)f2kXhy#r?6pi6YL!}{L=tD1-5SIiIm+OOdC6{9h^btFmC z#>{})?s#rjsh)2VGgS5l5bG(Ayc)ovrT}6GUuZx-fFZNuv;mZEPRi_5 zSj6Zfl(ghHacRaUVhJXgx`vYJqyq`e->;61(u9Hd<=KHEwWlBYhHe2sZyfI)y)+u9 zg5)S0R-DU8gMm^Gh4|FTW^%8h;+~ zn3O%Ka{xOiq?mVOOW$3w=gV3Y|2c*MAFmaK>}v2VVi9A8J*9PBsw*p>mfu=5^_~l3 z8)AHA-KFWOouMa~JXCk!s@oLJn5^}|{U&2gY+y$qQC@7h{|1Xfa3U1NliwFyNn@m_0I5Ta{-H%@-3ztZjfKvfZM z_+-oacKk$^2;*vSgFIEhb1o#>|9;Y21@-Q#{!DYF@_ti0KO){~HUK+@z*3AP)R@3p|J!ui9jKi`l*XaXGuahNX@D9_=p`^_*g!4RB1M8v5qahaVjJz1ym zeuFcwl9={ZmG5#LO2XBP(NQno*2+J#-oe|Zg94HGvfaW#>pXTAOsSI+$> zaDg+wX>pV(p8TG1(CR`S6ygDpOmkEem4(xm{Y!|fL7WMMVws{SSyDsM2Zt2^~8^*O)tyfg7(jl}d69V9LfQg+x zKO%L&p2Y>IZ_92&Tisx$*M2t1$RbqD=Pb_Y+LCv-3Q?O0#(?vfs-%-6D!n#>|B8?I zxP-5PjHbX67wu+y<+BOQ;cl?MVL)J7fmKO8TA#Ot`50V^J=>Oau0zo_;D z2uQ*Lj|G#u`}D$zZ~~+OrONp=5q@FSu?rU+VS}G%jSZ`V0C^Y4sx`!is9#0`CSLQu zkZu49(v#5r7;$Igi4V925=q~Re0sWI8xbr6vdc|Z+=s$MG!cTV<)NwKfO*Rb-rn++ z)8g}`^)opRH(X)#bq~StkxE{+6a_q{^2a0e+OEJpW=63)jfoAR`0xYWATIN|Hd{=*F7A%l=Nfhv*NeA>H*u5$oPD;U1jsB& z=vxMmbq%cHakdVYH9S<3eb;e^$^5Y0IsBh`PqLpc-(ReYh}@F+Ar-(j@$sWseoLl| z^!6^d*xGM%S7h>!b?~u)$mh!BU1@jW%YNlsc}%vsyjr${&NxcCXV4Ny8 zpx}U*cM>dCd(C$iEiU{V?0OdtczZ(=D7U+gvpm(y$iKT%s#~<}VwwcGbozBj-?KwK z@?U{_$Qbby|MvRca!^e!r(8&Rd^!_R3x%wlBi~I^$Wj{XdQ~O~-je}|JL~>Xh5)D^ z_wE_B$gF+)U{*k9w7XPB1dfe%C`CoCO=#r;ncJUk-uOWq5RC5j7vsU=DaN$kG&y)h z(4?ODn#(J1-#mDwOYKf)&A*2FLG_ZPJ7Se)zaQKg-)nfWLG8Wg8Mg_+FBLIPwWoKs zY?MvA6sEo*4qa+FA~bJJr_vGoF{j6%A6o{t08{Rz1mD@KW!$Dv{>JRytv|8@?4vqQh2MWRXd_b2rBAhJgCVGLcq-Ny=)usa)m7@8+!r0Y? zRya%}7?7tk$lLD&sZQr5>P?|zBj3{~Dv%^!UVRB5lHdORo}YtS=VTu8bF%I>yuT>l zzv3n&b1$YSK)L|FJM)ca@W8J>@HjCu0v0|wL16obM;mr>6bLj%g+6i0?Qt7J*=!&= zZrHCs%ma(~k}#Y6%->P)k{u0Q9Uhhb`Mv$-PA~@p63swHqmw0oxSV|#a4 zI9o@YJWiGyunjQAX4?0zO<-A~_iopQV(4pS{1#Mhpnh+^2IjUy!xSYF;Bx|T*4bac z>P4*uAyQ=XHWM3(2F%Ke6VJ{h^Yq>TbUIK-0{;qCZ~0_ui(<6EghTG)YSQ%8Z#270 zeJRA1mpjt+I^JY$M7xz?`w_4O^D=rhjRq(E_p1Tyy~8qF@XRNs0Kf9`TqfJjIX{?? z&|!fxCdo-$C>sR5MyBqR59X(WssG2;R|Z6(Hf;+6(gxB9sB{WQNh%hNbT87~(n|^) zMM6@fmG16Xx&=hKSwdiG*aepOZamNXJZg-#K1X*O$8v zvJtOu7k`OGP=DTWopInp>w}zG+5F*V-gXfn;iIAGd;OP6@%5}LvJ0dv!lCj5Hoafo zo4F}oL(=>M&63zCe!|2ZdNmL?#xXlc{WQzO^29}8q`q3 zZoFjt;<#Y`o#-SUFD5Zp)LScBMA5TJ7xIGU3e%7SCSW!KU+M=Z?gD$KKX5W|8 zF4O!=H?%4dA4?uyojsE&)oKbf86cymKb}0e-5xTzS}s8Lux@)l!JS79%+Yx(PoEpF z_ci|2T8O(KZhSJ7n>>^MhCV*|v!{IM&pmGHvUgeFInA^Gp)0>tz2tk!qpdI7B|%;) z;dX>8&g>tdqx&>For|bUa{FSMOWePpw|?| zJgzxn^j>r0Or%M})Gd}8ppjDpbq4;kxL*kDf1baj2v(m@SsG-hm#!8FKGh0!Pu6?aW5|J$fo-{pj z;88$_oujE7Rkj5JB$y~j1o}mckL|}pJ-QM=&yo_m|C-Ixukp~6rpaPXAbq03`JF>C z3G60etk|`vz4Ptxo)RZYAL0$v4sG0^Lmh0b@)9es^v6RUTDR1cmLcbwNgyFD?MIKg~?Aee9;*ebjXAe&>M zpmRt8S|RzLoH1Zx-e6Yh{}8`4jOPVB!d|s=b33PU(v9cK`8uOz*;<^x+=?9L)w9@o zpGV8vj@~u*G`7g6Ba_dgIOuO+g76WvE(7wXvOzFYl0op&&4PoHZJ@i1AvFqqM8#`6 ztbWoSRRt} zxfwJsPJ$UJ6>C9LFGS&5J0)wZ)Q_pL=-m$^?DW-R$GtVQLbXM6fq$945%`;93=n|{ zkfI_W7M+L)nc}Jc@gSGjAcEDH~q9&nI}5^gVXG{^!l{ z<(u_Tr#F4JH18(&?M`9LcAgB#_Rr0vMw<=LI=fhh`cfmP zXSXFFLk4s33FdQnF~`$IDL=gtgSj3spTUl^b|EBl{jFM7#hDBq@^Dmmgv1k$bMmlX z&F5_9jHg)~%*b#s|JCr|zfUqX3mfOf;xFFWu1G6r>r|?J*?PZTlpAttt!(xMe?2fv~FTpoXuEWR#vlJBf5SaLU!{B{cTL9R;vfX^ir(c<7z|smB**` zedJ_MN-nk(EHfHkLR5;2K5b5%x!$J)67>3`=Xoxl0Vz3s;#&~*xerLiReyZgK!T=2 z5z)@SYJb#KyIq%?mlwXiwU(Sr8ETrYC$0XHI&y4w2DkGh0L*#V#0|c0?PLG8=_p`K z-JH=XeMm-In(wp4;f&tQ$eTG-w_#GfG!298kvqxLdl->w#-tecNG@Ez1mn?vjNj&d ztg0q{EUL(qFnkrdF5liK9)u<%>>_2A#K2eb#k`$*1)JsaM}rA0?F_rQ#;U1|a0vT! z@}f!HQ;rZTGbV?vyjLHt?;D~@lW4X+VcyOT@$I#c+Qz>wP{3x2B4rrlq3T~akPks^ z0ep|w-Zm$gg*J8mA z+lSNKsAqoe*yY~Ppwz06c<8?NF7`iG1{P_~fl)UT^o^Dq*+kO3BF1Uo^~{ccF^W=G zoF{A_5`OPdtVMpn%FyNa(KN(}nQ?W~xuU?jk<;PK7H)Mm?DnWwY zGWWwmvDdZax_(3>_KpK70d5~D!m6-OT&<(5z>6t?z6>b6d-m02-HoEZ?qz6;=1zKY zx-{D;8#rJBuG-z{Rn<_=JpaXzse5CY|2eb&9^VfQ7%O?LMga;whlYjUV3 zDjvojxh$B}wKm}72aeh|2J1&gz51G-hq$DAM!y`$Fy~@zD}s zvlKgH>K_wN*LQ?e=(Qsn6OC&`RRl$+ja=rwvpUI!vKhp26je&HKbehi`OBSkO^Aw4 zN;Ka5hm`&#Jn|TfzmAXBorFFd94)th{%k;8SO^HgQu*4Xw8;Fl;-TNm6pP$TIbO6S zDgSrF{&V&96=0hcvaE`DM-6j+$>L3wAN{8n9_?|wD}#hBc1%d8Gfo8tT<+WmmMR|n z<@md?%2+#>mkj&W_GVOYhuNP775n|iePs3`dOt06g!i9d9@f4{{4zeoaH zH!ouAaney4k6e9k%OxjlLb2oyO1z_)Mv7+wMl23Se!@=pmzy5#t5}MzJoFeMaj>A* zI^Oc+jqDgXx4?V_+rw;EP8K^6b-H!xKK(R9)P}8YK&gl1_#u9L_15c*z#3jOJ|R+T z?Y0va_kE>41bfqP1}~aoi#@>p-G7b4-%FBXo{ySUOF-rNLa_Am-n-VUmMVYWYQC-D zBko_P@kRT}pfd>^Lj`=;EzCF9a%^I~&98qM(O4bLE$+$Zcl5CAbBa`F90z?WHhMA? zigpC!_)4%%D+gJ}*#e+!b=>?VgMtZ`Vlflf>W~-9Z~-J6yM#^|BNCBE$tieRis=Lh zs^5H}6E^$M->;ga-}0@U;KXr{RMKm68C!&FLmC*E+styBdY- zC|L~rA1`6sB}S`{=xVh?CZLqvThtUX@D9n+=blTZtb-Jo#0gjPh?*O{^!=em^0yID zAtwTg2@A$c8PQq%0nqg#JAt~^2O3(&rjU$lbo?p4_(;ltSuV2?Zt4Ax_8q@3&ETe= zFV<=09m{q<*Zbz%Nb|Cu^nEusWue0!expk z>}qBllmRp*fuA{f%8E?Or+0esjTy$x>SKmjO+?N3tGp&c|84I~$Z&&_#@#hBy|$&7toFchnyHOLWWK3K&Ym5$Vysqo78q$O6e zE>j}Prde0slCehgudvO&B1Ee3SLJI4T>s?<|NrZ|wE8O~xY8H7Vs@2Ab3-m~)7z4$ za*T;lg>C8StGKkX+~8EehtzN49wCH^C&D9|#!H!FSm(E*7`7zjbxhtTUc)xU14r6B zUR9Li56$L+vmTy)k1G3+zxTFZJ!wK?j+!zTX-lg9t+nsc$fw<}k!6QltWfTb|L&fB zF`WqBGImvIV0wGbS-yCnOL)YXfFsYAY1(|V{Km0<5NiR-TBvGbz$+(ggY-m4^uF~>}p{hVNShxV&=sCYaab$T$a zp=h&Z^%zroK)f|&TfuDCHU+yOp8eH7w&n7M%Q2<)xNUig_|MptH%kl;SZeo7&D2t~ z=L`hx9$;McW)jH|)djGE)-Tog@Xq9V@PJA^N0JP={r>A)X9J&R1g}%32-NuM@uF*W zYUdr?%j1^_ghdl*p0G_3rnm#Ah#irdBEXOhc`5(t^usrwBLO5LfiAe{VV2|@(-bzi z1v~WaLuNcP`QPsk8M%3kzItxr>bKT{9qjjl4GPSq)$ZQ6vKZ$Dk0LN6LVU1<#PA;m z`c7F$L8`9ljJj1E=gOrFX_C!JHz|Yr^GdQsZ>J9WlkkoCu!>gxJ5|j4B{$g>T#<2I_rUJCiJd%uiMov1uZo`bpekJ54p~P!gk^X7`I_(O!<^3D6h-E<+tOb|Y zQuJOm)0s3KF%J;-EKaXG;*W0<=i6=!UBAoR+!$;me+!`!58YCSZYgP;b52&-xePQK zerSDg{F#e)Tth|Q)~f__E}Xn!kR zGzMNJVS$nqk?>Hoa_|e*8+IIUhW0y;qwvh*p+N}sxaeJDeMl4(S4i(Se0+NlOBH5k zl0p|eyA8|KyPMr<*~ZLQ?HGw6K@@QD0t5e(I_&-gl-g5NJ1FJz+nz1z&s~c@S9=dD};VB!~ zk>y2mdIkM+Z@|$RXkldO5<)8UF}>-W4O_z=sJ>I=2R>vE59gG_b5WQF|KU4HE7-gH zQvvqlA#(s2ZTFm6h*EYxe0L>lf&-J7b+MyZ_R&3Fi$YlTAU7XB=pg#S2wDH4N_`Y z^zo09&K#o|#S(U$npss~_?Wg;@*7iC-WfUx3aoY8eVnG?ft)%Ln_vX@`>EN8osi_( z&B1q>PmFrGvirmp^Im(Xe?#QqS1l9y5eG;XkmVuu^JeVpZfkymC4vTQC{Ds~tKUtx zuL-uUt;nb?{`?L~q(q_Az`DbC&K0hN(X@LPSyPI*=1cO`_ZjvN`E#4c!Lc)DajCvk z;L^Ar&pu2JM?x^SWOL^5VqwOJf3htyDz5f!JhitH0N!xg}Dv6M1FW^jC-5X>{9 z3xLKREiW_ZE6_*vtD7LZr@ zlI~Eyh7Bm2B}S!5I^nu8c)IUX(VeKfmE)~4Qk&X~29)(}S%qDfjOtT9i=SD%o;CvM zb7X({uJhfRw49v0rasKn6{9d!eBg--epkjpXROwznwxnK>E48G2q#udYCR--aS;$) zJtyJo%oNZ#TcJ5thjZs+CF5Mi`*32S@UC-O3-kdRI@x&Qp>dxAm&4L0N0sXAs-?(P z3hETtx|*2gSQVKVd+k9s`ia1&Ug`cB=7Dm!JBNgb$YjN4zFrx=ZJbv%#86J|`L4CS zrl;B5G>H5rT_rXRE?qIGr1|lPRwg~3Whi@}%zW@zk8782?l{iB)>SeOy zM0TxL%VT}nmTR<6kfH6Jzz@jqFNukCBteGerHNx0p)|LN?%+5u6jewuTA1&Vs2jl zqV+wPDECIe^YrU=FpJ>@v0A;rT!59JExhP_g)X8u=6<0@;mmh+jmJGc`;c>tmNZ8! zkH*apAHv`lXJt)KA5s-*F83NS&%4y^bow0CX+J1kY>7_+6Yh2UHl1TWtF_zpyCd<%lx0rEKCQfc)jM|^dJ&rIo1!Z; zhUuL|c*L3XsOl+*1C(Lx8Sks|cfV9kg`PLxc61~{%#(hh-p~1GZSXAWm+7!9xgEF)st75{Kc5|b0d*C9&GPk+u_5@0i z4yu0hr^@3}HL;l!v!CmSQ%4g67ks6W$H?6KF$!?Si>9X0t57zka??YZ@`{*?D@TW& zkoi12g?f99xT0AY?+jH7ede_Ow+Di%I1rrJ3lb{^r0iUt{`pUl}&?k>cFz2 zq23(Tk(d4^bj=y(Vwd}zWNVfiTD&@&Sj)bxXwm=%(x4@FeJ;8olGw7cevLJAuS3ns zCUi|G6?fx?d;SF9slJb4oQz^<2Kcj>162uG@=xW2O>nqSmW&5+S#XNqhmhN;l2-|` zH)=nUG3&=Ar)Q*jp!DiR0vFy{ud$w|`{rOH*v6UTp&1{+Aov$>MVadbXcZG=4G}3F zP$H6XK(u4#P1kvo*%G#oy2~bmd1;6jn@+yYz0i2^)RQKJd1uw4d{~qv3*NrK|(owR1b|RMyrY$mw~l4CK%b{6L@jnbXIi&`4UQ^ zv~ey0Ht^%im;G9l+l^18J8aMbw5AdK0j-E?5R`o zyvRYsQVSbxJ68iaD_^F|ONMr&A;!xGSXzEh8i8GVB}s53y)DlF;|*5KLRJv|kzTZK zN^Nn;#U0kQZbFIvz)Rowu<}JlC2us?%qu&LdSZ zsa5HLuf$2f+)%GUa+5RZaCf@52YFrU$=XakRMAsM!|W-+U<@X7y<1I;eW%F#0`fAY z!BBtvT|E7gE?Stiv*27gzeCaI6l$GHUB|PFPai# z6Om$a>kcpOUhb+Hn-4bGyvbw)$Zi@NNAJ+D^qiCH!fdE>e9uPDB03V1S2hG}gB6a? z#_|hteCMc(=zb(?G#4t3;Q*SPpLZf~b#gM6{Zb{)y$8^Lyk;{gVN`wwv%m=u#>$8gqos4b( zXVK}Qzlz9CK*?*Wp^Ta$aBf{{P~_zuhZGh3}MA zXRXa$n&ZdJ9)cR!rwpyz`D|;fTD&xuFp^3S#?JV@3UyL!%5-1raLZFPvSUK0UB)uh z`&3jMzy2j`+5UCR!6thc$m0?we~^FWy776Uc|F;yiGY~X1dm*RX5LKz-Fpe%g3hhEXVh250i>f z@N2y@c|HO*6{Q#^HTX%t$X^4|>xu4^xf*k@Rzs-X$k$><-4Ea5^tsPeb82-23dU$V zGRkw;8c#@ReXJsn<@uSumP%37cdIk0^7h<-p@A_GSA% zIz6P z#demyq@QXf-)@O#PojJ6j>C+RISVdz#jl|$m7-%8&nBl*du(%on|D(mv@S%cN(c&x zVvbx@3Vj2$I*_YbKl}l=(GgfTp|_h$nz_2Y)v)K3t2npc0bszGTJf&;g=Bkf#&I5V zl?1c7=GjS-a9ECEt2~lE!N?R6mE@=p_dC9geJOeUNID1tc>~+7GYpy=SF(JE?C8JD zYi5V~#?FkbN);V!sYrpLksoKFX11h7*TzhXrrEI9@6Q*Dq=N>_#dbdyo5j@@zZ-QS zq6{FjmL{UhaH9kZR*e$|(xtk*WFu*RC>u3ahW&`6G*Rbik3?0VB&QG+V}4XQCd+g{LJ^NmceUJz8xj-4zdMqR4vwb+?->yt6| zZ*#@0y_~swg&hmged52;C;y0PK+>E^3IC{}5Yt;Hz@vg@q_ybNf*?N5DVb?^1q(>> z-BcRu6=AH~4o%v-!*roy^eCpYy>8fz?J>y3e9|&dFIFsVq75Zy$NB88KOo_9EboHN zK(WGBXgUJs9*Ch0It)^Hc9v1@DL(qkBg#AS1=%uMT!16#gVCfo6^Q>bR>mb3sWlm^ zcYb@fZwOUXo_wM;8Kxbkfz2AaztaGNRVRHv9}y7t{eTswMpZ<$e%*-_VMM9u&HG!d z2o8dP9N4KTwd+k>oUMws*<;VGc>{XcTDJ-Cj)-_$?H0c1MuEes@*U7a%#2li|GKFS z4l7D>E{8g#DOPveX!V?BM--zD>+L;10W47rn~HiY^mD_npJa6ExHI~>#*Swec;y9H zE$W7sJJmHypHM0q!FnV}G(Bf)cyO@325~g3Hgq0+<{4Z`=$_G|Z3vW9Br^WNvN#MN%>N_7-jIl3xiGHt>e|Lja-(DHjOOz&7i4fD`_ zs=G(r)DWt47_xs%<5+VtTFIGYbe{e@n#cd72HPz8YNvBUJ7}&l!ux(f>_S$|fiY|# ze>JvK-p}-gk|3*dxzKK`F>1nWOY;*LATNu`+;L-=A;e1j8i*~*bzQx(kmZ|F>2zc~ zL$W-fwp)j#W#pvMxmphWdHW5qfTA4%#Mqef^YHo=(l(Y()+4`*~m(Nq&q-5Oq!Y;MLNdV!gFYDKP*(yR`-!~ zZ=fwWun>J6W-zy8Ot^N?d-_yk*aZy<43p=ba&cXTWt!d? zHL8E{fAl}ta#}n$c+R#c?J}o{)5)bU-&sToh9+IGxMSVctoOy=dULJL-fX83xsu3G zKke>Fm(*C;Zr8`L-Iev)_3gn|wG!X9Hon`ALDmvyY9w6Q;~jE&b>D}~hAA0tso=iM ztMT~2h=B8advuwn>#mzaMQ4#@n5$>Jh!O8d`@S$@_ z-`7vJM};Y^=lu$UC+BrfwI5Z^y8Q=y%*5;0ca

    JLR@tEblw|b(nsK>;8$ood}#! zkBZR+sV-@!yGiM~!4U-3LmY)E?<_YJTxjv6VMSj3Mq33tT9-n`dF=YchELBWpjck>8Sc%|3nl1bQUBZ~7z_ilXW(NVg~UzR$ysQrsbQGPvV<+C%;=9&QK zED~Gmn>9yL$m*FKn0`S=wCtO!t#SBFRV&mf)R>LYzVX;8e|8x)@8dcBd)!M%vgD?Q5KFvV* zp#|Ymzo6e*ykwlm%+c)eXpy%#1>w3PWmSfRR)qoKcS_W1bYF?GaDSk6r;s3=Cwm5) z2*=Wg?Ir)5l9pe_URnn{-tO$V+brCH`{Y4nWaa06^~U~XN0=t@$dM&ikvi84wCJq| zhtg}RI42)ubpNbma_GYt#ZEm0uazfQt2v~3jv02D5d8ImuiC@$3H!GDP6M%Xfho*x zx7O<#Hj5HF#+!^{&L!AX-IMBzhoQ~_MJ6b{O+`b42cZ=+Ax)-Qc}B}Fti?(g!83k^ z_RRjX^o!lqs&&BKSwGcLAJB4q85`a|tj$yL#G`((WQ|T$&5q+`dy5D0@y1ME_*LSurL-Kb3YQi8=dk;8Y6;8F-^%k%bgolv z?L5Vq(^q?Zi}w{Q5mP^oT+BvnvN`Yk4Jb*5ccy%H;w5;ro@ zRbP-}6O+7q+ZeLm*Vt~eI?4OtE=v-2on}xpb<#2dUpA`;0sj+w zD;cn2Cvs7%+wH(Xj;+|lj>5M4NX3XPxjnS()(+a4LUH-tBXE5s{hr^n??sJvsL=w1 za3`wgEmeHqVgkFv(!SibRrG!vT(jDj46(s-4PDAME~p$P!Qmy#q*Ye?0sLQC9S$(2 z$#0Bo=c7Em1=28kj&@i z^XHvum>>2-!9edd?#+n?DF_!QVnWS(Ycg~yBg^beE{m_)sfePDAlP4XOym5SD!biW zDW2^bPgVzi{KdWk{7tL($|EM5`|?LVS_RDzt#6FY1_kkwB--{pzkxaM4Yl!w7)4A+ zPJCk*sx~FrMa7Q=s|h*hUAf+?q!7ZTC9Zl`h+dl(0M$hyzv4MRL06oGgzlFL83fhm z^zXm~im}h+Rs4nWe#@lxFAcIa{Gl1|-?`%f7eH+8hgcRX)!x+m?vdy{PLmveC);2y zJ@<`P-knjOx{>C_M*}w9{?Kd4&C%UL-G)ML4Q|~A7b%~0Sfkj9j?78{&3OOCD&G0a;1&qBf_ zO4oG62p$fKY4EN`KE75-XYHrKm7d3w)U~*BES-4jMiIUpqw|sD7{fl`a#P9fD)nql zm5$~$$i`sMgqS)Y90}tS7rrdstL#=|E!G7=?j4(Rf%20OHNyQ{6jULK2BJKTy1}aZ zXOq^%fV8YUtH>4jWb~2GKykP5NbrdauazmpN|u6p3}xw5!#v?R=jE!MqG7SHvgs)V z-THzouHgSb#J6!7ebw6^Q=OwO;1?|yY!^`1b=|eK_bDP(c7C1k(PZ@EIG(7RCSfY2 z#>vut5zj->Trm^fx#}=B5_f|+u1(gM2Gp1jD!lB=DoNC<)4i_)2UV0;6HcNgU3xIO zn$FM==^|K<$h_*6{Z*f@l@PBu9s`SMg)%O3K6Fo+YprT#`Xi}dA?t7U^2ZZUjZx1( zT&MtfY`7A;m!Bba=2{MotzKf=yYS$h9fdDsbt?jw(1b&qoy;=`cL6JGF$3OSNzuDJ%;3{eADH`I%FclrGOj&eiIt1zzeO0X}$9Hwu zEMV`PharkJd)CV`WkKoX$}Sci^fZvhu1T9zjJ=WwP z^SY@ENp8Pf(PSdxwW6FiqcWKWteaIH3B9{+9*;82q+*dpy!5hC zZ*R5r%q`*)3f&dA@&cR)E}0JmLQCi#X6W;jP_rH5a!pdMDD3YOm@Ws$^Nu^R>lqHm z9km2YjO^w@kKR%%h7@$ERzfsn$tw$|bx}q+y~K>3Jc2HG$4_}2dQDwnLnU-B5LZ1H z$VSrA7oIlplj>?}Jbz&kksBV{GuqqgbLWfHD{FR14XHiK=eu<0T;~WgpZknLWk-l! zyNZhsR^V|rMZl)Egk>yq4^>?>4097}b1)&pw}c3=&FFIRF>}9vzBpaec*&4io;p2u z(NZlaiG6auA#o0MmX@?@yt#wk-9A60Ta-9m8+affkcQ2nHdyLG6HN$9k%`!IhhIMr znd^X@AI?b~a(bOEQJycZUbv60vkhj2b!V6q_IprvtKz44d0Qh1g3J_>Y7E>|b-gya z(zG5AX|eO~kgQL~mC}y9Zj_Vg#7U;T9{Pu?9W2IlsqvkifH$Y1MCR=$9gf?C7KaFx zhlQtWZb|9OEG;`FY1Ro}5#i1hM7S^cy-KCTdh*@qpbJTFD<*f3O{;6onzo9T(Lsgx z6lt3%GVzsCSLnq)UiRC}ijVv!erBLpPvM>|Q91386>Pm_IH7D~dy9}Ior!}_UPrCv zvqi64?6pK-T50oW;o6YDr5L+D4JC1~zc$QnxZ};Ow|RsdK~FHiXZb-=*yo^ydi)1> zHvUl#A#7dlNz72C(G=6>C$GrBu2-xP9}QAuW;lsrI`T98Vw30pEmE^&O65Jh$KC+` z{PxtQwm{0H)}#2=1}h$#i^CgRdZBljDIfl_JF&7b#^vC#q634mM=9R84tOU}PyVp# z=~UQbjLRBq@JwzcZFN$+i-|ic`pE6{9aizMX?)AQa20F<0M4c1n2Bd1k1Z%6R zYzC)F(B1kYuK-d-&GoYLtR#~$hz0`O|04aqkVyY}s`KdumP zD(j`sbmFM5&4#?hp|f8|;CTwhdp?SJanWC`s>$|JEfZ4!LD;-3p#MX=@cxQ7tE8Xl zZQK)tn~lzmF~$Ttj@XeNx7nCRs0KPpsNUwh?W1Y=Vj{JMB%^E~=*kc%vTKeg87<^U z{$RxN=)=fn9u*1(L3VMZOlXkyN0eq9((vOUW56j*R9b68JgT!(av@WO!T6*78nn(l zQ_9Vc)FM^SP`OejeBwoKf9BcF$!u$gQH6d5HMMo}wrb3{Kf}`7*`Vm#7J9br{Ic!_ zH5UzaYnypV_#ABWZR|u=DOzBYxOLM6R>Z{YOPz5AJDGhx+Oef!#+Zp)7JL(WpqgeTs9h%-J*>b4+CZQ?zCTC z&uoK~61x;K1U2o0-UaSGA+_S7(mOhL#TvV89`39=2B@}YJ5eeE>g&O=fiSk}fcRQ> z&h-Mbl=HuUUH&7t^FxPzdZlHyYSKZe$K!de8K(DyXfje3wjK5Ql8gH`dhizULw=3jm{>HQDzm6S_Di&a3ZU{vHA8j;zaEV<<*`w45=2>E6tNJxl0*Rs2kq zl2l<;J!Y=1i>oX-X0l;XcjAFNb#yZHovNYx^?Y?tSGU(rC#IELCZc>itVpwu^ImU? zi@u4W0T*^8ICe^}9Iyamc+rSkOHPlpbUcUt^1AvveMBi02eX-s*@f1k6&U5gik`g9 zyWFn%;o`Cj2dnpmP$Mi%8AWuz;sOtRAqJ)I!wW47HjklIfSS-(z)Bdl&xCa+kbsDt zK9FGNwCkR2zIZi^#LD$kBkUrgO!X)9zl}M_)+&>3E-PnFkVVbYE6f z=Q%#+gmulgrk?dOjWfskmds=If7wd?SM(GvhMfxn0G2YV=X1`)T<7e{@`su=(X@{% zt&^f!sxFBrnu)#xHl15U;sxnh9%c%*Pu)QMz=CzP`!NO|Zr|2aTxY<|kl4NG`OJPJ zT;$_Yv@#j%a(6P+^ym$oXH=V>yPr)>Y*9IXYR&nH_qOnF(C!i+o74MAcD>uYrn7Sr zoTORtt%y!bmDl|8#Zq?fQM-Tsk+O<;8A53jWcRqwe1{zLCp{*8KIkGC(*tjrvEclv zv2o^vzU`)l5$yGsr^x7f!q5p)i+?=J1X;1>6hoEFqL~4*;FJTHy>3+B_{bBUr)svS z1fINK5GO^jpDj#3#7$)R;|RfnZ%if8K}y56uYrzY4W%;o1>&1CwmTe7%5D;h8^em0 zS}u!Ho0nXvZ{mQiP+gIf!`pgU#B)fRfL4sH((0>8njjBd^8MHxa)Q9igdXK4*bJgH zOpwwO%6Z1fiClkZ#CDnoF(AGIo;81wuX@Yy&V(yZnVmj+OsrM}Ypf6TiZ`iq*>jkJ z!cby}>6qe^%jeuAHP2L{f^#w(Vz&F%9O7ZSZ=dz(s3*pQ#DRK;Q@tTknK%o!)ge62 zGU$+o5ILfffc~rj7#{14q9ZbZ(%`PK`frAVqs+H+0~U1WZaphTQ)peQy|ig4mC7zi z2qs;E_1&Y{AG~you8N>Y6XN@#F#wR6OJ2ulBsl8x1_hM1>2vI?vAXyw~lHRI>nnNt4#kA)IXJI_l~9?)LF=HQrjm?(CyAP3Y& zhuy7c{QCx~Jp?}PJV7IRv`5pY#HBlHf`>lXO3AZ^nr84Yfw zb1@u9*u<@@F8?aeykG6g8D!_f|5k_m>&vWl&xJg{ZZ~sq8AAc$R5>U}dYqkNna=Az zP4TkF2XSBfzN4)R9u|$jHT4Pt3`cGxCj3jLvBY=AR;p8m zn;3Ba&tBX&x8TM(P`$ho5c<8!?h*#3 z^K?HvN6OGd;dx*DWYINe!ALNE8`yBy`R$`}INQr&?nif8?XA<(!kk!%J63jJD}+B2 z*o<#U&Qd*=e?xot`Y+d?10(i+&5cxxj>81xfTI~%tV7EN_-t0My!~YV_XhmuGCa)p zt?d{YUF!fe^6?9WU=Gi6qkdC-H%A|QW^vk%WfJ(ywo)RIYqS)A{O0ZFI}kQjg*`;e zqq({r4^ckzVlK zagDA8KTLdQ-Y;`{$Y} zZs1nb7NbL5XioI6)uyK>Q@O7}_)a(ihf~7mr!|Z|QItHNDAYc3NC5F?9v)V`3(YhQ z?0tM>i&o9Te6s<*nVodEBr*QOR7sxkDs_p{OF6(-GBAUPfi9s=EhS3^lP#U&dzGTh z=MB&iCdr>bo&k^;sDek!!WYEv1DyAVrOqV5trRNRIQ)-qWkn5qrjeH|w!;F>b;RE% zGlAkd5LlVJVFc?sLl{j0)r1GWf09{d92X4C`r60A6~AK!#~shWG6opsG}-XErTz($ z=y`C&OL_ipkGZ-(sshmtN9q#~N9oG?Y3h;~K&iMF`l;lAK>@d;NkT`Qt7`0>U2zrG;91Xsqigq6 z0i;5wKK~aJa+o6-qE5>Ad(A(6+i_RA5M`jTqw;{b^twv(qz5q$dOfEf+RvWq5BNMW zyT+C&AK&)=`65*a+|G!tXNUa-eZEpt0~)e3DYvYAd+DX+U_N)p4c}r|TKtrg)g1t+ z#8h%H_Bno$6R+`c_)r)JH-uLTuI1%^H%KlPB1Gg;`}y3SMj5BH>MNjl{xx!fR6`=7 z>cx8x5W~H0xr(LuMYKOs zE=V7T)gVENjN@+Kb?Jah$}b4|UoG!}pr3M(PRHOmwI-YorE__`x*&kHbg!N&XN_h{_uayj4^_)% zNYtd!Mt?FN;B>@F!QiE@d#Si94^K>8Ah4&R!9`r{A76rYOkgZ;3i@HVuU?PjbN@~A z-752I1-!3bkQjaX^!eccgT7n(Erb(JMmkE4P;y1HScXzDn5RiWcGEMhL%G0UdSBTl zYAOx(1-m>M%2xEoR49qKW+SKu%6R=>>e-sou!Zv1?b>wuakM~u5k}7;R`H`NeGfna z*e>My!X#jucDK9zA~VR|y%8z5D46lqUN-g9tHf_B4jZqPc~8f=omstVLNiG0A0-BX z!Jo!XTD>Ny2^w1%>5n1&39Ja&webF019s5$j){IZ)Ju5f)rK(7xQ5khTc1lUMq~Uy zgWmeBJZF=CqfF$p8DCDfX)Dh!c`r@I2m+0G&xv;weA4|9} z&7K3!qdbgJkTO565$Y*$bO`NW{CadWBGh42Un2Qpp{10dk83%;O1D@MAJMWR5WUB- z^ife%-vzB9e&wBw>faJc*diudM;54uujFj$8nlXsq&O>t{{ypJ8j}b22X8r8H8%o> zFj&%(CzcQwVkGrlcH>7@>#tDf`S*iM-p)=BCdL}Ihnv-vBu>7Gwi-+Tq86rYy4e)2 zLDwlJJI>KsZHy%6y80Nw(f+!_jrg=0Ja92c*TT>^ z2?JmEvREl1F+nq-& zOcLs@G6N4N+3*&I>~ZGwWGM=jw2cZA1BJ1zZ#}?AKlw@QYC*#ClAa^>!@^n{pOb?5 zx37ix$Q;$wrJspbpVy>yz>u4#VoTE_RzLy7ZLwtraC6J>QM@bQCp+nL{EIOzF7}o9 z*;aX#(+wW1_j88ktzN19kC#u&svR9MnZ+qPqR=X2Wb;3M1!zL)5Oc-0z^`>uO7Y&7 za#?%T_OAG61Dd?mqvzA-Em*3o1BXe_lb+n@)rJM{vywG(hE$7vRX`=rzkvX?yful0 zL|88Y1^O_jKq!f8Bzb>Xu*>z&U1KrO{=Z0Qzf*6#nh&eWcp;kyh*>(~rD-nAaj)BU ze2mlA`Q^p*pTjmPjm93)2_X9p@M7fNrtK1nOOQX~?ca`#<5~8MRfoNHHH2!UG!Ers zp|`G!4bmm}r16$1TJ4#D&gx)wYR1ANN9jOl*8!V84Z-Hs8c8ob4bC23xMBq5c7X#$ z-|zX8Jp5lanJn5DQe}Kn(Lx?!h0PlYaLgSLnX z=3T#fQh(@oyCEU`nv*lJ499_vuk*av>kLYI9m8wRyrUY-o?Cy4CI+)H72@vXjiMcv zf|JJ2o{RW%a&0Q9`GRD2HQ7#`i^r4RBdZ_6JJsM^J4i1zE~OnINs$X0R6I3l^W)1x zOp9cA+0jfLDVZe2vV$zQ{z?@aEWG2H7SZWui20RgKdzP|nebh8-IHQr(e~g-J)M`K zvI-I1KB#m#`}T-^^`d42NJDx}iaYju6VcvyC|6c(Lgsp$;jHWaADRlVz4>hO5#HK* zc~GI#GX;diWZ39mItiBQy!N!xJ_Yb5+Lm^^jvG;Xp1)3)DCXFn8CCXeBUsnb(nOW=LuG?;b<| zLoxQB^|7ZOBBHXP7D9J`0IO=#;kS_S?-vpP1fTDK;PW!m^HXbe4D#+$Pd~-O*?S4; z&fK{!f6MKouat*%!MgLprrzUh6QE?mPK;AdzS>9OeUlDSl;$gi`GQ@$?&e-ldBai$ z!;+FS#n75L5lVXiKqGY>qd~1|PDU^kzP9HG7#Mi=QD(m`-9j9fNa$5jZ^&#IKi>;=Nw$QRvjsmK z(eH6W46N=JMO}#A`*wT0$(33rq7!A?jN+7DAzHs)f{!_%XU0#MjShFi{|se+=)yK! z{~udl85ZT*wJqZyrL+Q)N(%@`H`1NbF@%D2H#2}rmqiH-B_Prw4FgK4#1PUkbaxCe zLwq;;{oe0;bU(*2KlroV_jTXbwbr@Lb*}TW8iMo9T=cp+CucEa$xqsQK8Q@G$~{46 zMYtcU-E>%M2KZpYE5>16@9D>m@MCnu>@j8R9IF7X`C3+Iukuf-=FD>aY`d$8aVQ2| zYVfP-OPu}VXXrGsF?JMl4ItY0>DjdP5Va&DnrVaxe~9sEhR4T*h>doe%2ExdD!nRK3&?_Gu=0~>C#<`vr8 zFlDSc#<(vhua6qZIzLFXO|f0w^37w_X)3ZBCw?vAs)nO}yrnONjRUs5Umsveene3E zaKYpgN>h1reDpnpEM90$LkXzEBJ_Xxd;YiW7Dh!Nx-=iGokgMy%<%$-7XyRaBQtM* zF7BA zKv>!DTdHQ!sSnI8N|hex{C*lD9C?HKOodHt%cV!$qa7#&76!4*e z6U^0xnSK8V{@<809V>p&oG%rcK3=82{Q7h;6k}!!j4Fu=N4(35n_at30g?S}-wwp? z>||hioF14<>ZjneBV1YR3=7K(s)EE1>x=g5 z0|bt@0C{a&{i4kf!Qw~zERQ;KEntHD@s*zY@2}hX0{ntUYg2kKVKk&adYBeefi(Bl zZ$2x_pd0-aC|yURENol1;s4ld8L+Cj8lPMYvh|51Qud3S^&b1`pj}7=jMWAq^?JXz z1_-|n1=@ETK$K(F`-TXT2f~~_FMva5qO1w^uJZ{w`a72?PL@1^Z;soS3H-=|vW98Y z0Jn(3s^{GKs&DpjdHvGv7H0B$c`0BWoa!9@TNh#U}90RU0tNK_9PEsPk~rB z4tneSswSIaXPe;OBJT{}>%K(~3pE@F0_oXW)p4T1gZl)9KwC+O%WX}#5{655QczGQ z5RZD+`b_v5RA_^6?TFDuUIzRu{wC$$KB-+LelgtN7X0<*r&{7ERVOJsVn)mF{*9x6 za}oA`fy!PdNn^$37QU&39mee$5j9{bkZS#1+z)E4TUVil7yaW{ z#i5TjsAiy>qM^V0>Nzo8aQXF2DFN2gr)?C> z4jHF5l!Y>k-}>_wi??HT0A7HbgY-4`9N>Gp>geN8)xAtu#|y=3StOdd*JS816H#K( zM{ALvFzD%81(Y=HW&}{MYP}Y5KuYsOA`NeQ+E!>^_IQ4q2aU%y5EW$n0$K-gMR~5c-tV7}(=nH6#6Lu1cR7 zSCN_5p%Bi;)@om{eN!L4x8)pmJHOq@X;mlq)PpzzX3UQxunRocJaCI5na zkgP(ox%~PiFPCx?og`NG)Fn(q*ldr=!JzY90CYrpC4>H7o6brM47$he+cXlxtT$5S zr%tNHIZ)X7MDic=p)(`-x)7-djRqKNZE{-}7Lc3YVV^x1s|ATv2|fTyWDoG2SNB_@ll=q8$I}KD0Uz4c}e4qDhfZfOWNZ$5-K^0 zeDw`sO5}w_-tK<^jX(xf+M~Qs->LqMSpLz|)GgXH|Jn@N=$Zy#JguBSbKMgqRNEv$ z@PW2`-R>53?m}A13?51ahtOEPl{u~TikTYN7-GNkE**?OP){SFT0cP{8;`NS7;saIgULYO(RWgZGOtNy8=+s_sr#%qyX2hOl?}(cBN9@e(5VqDCvR=BK&V zuUx4Q$qcYIQS1`F5}tl}0~aC-um*{e?RV+|bfHnNhL6;l-to+$7g+kH)r_A;WGLkk z0kn3*$XaAt_li~8Riu{Y9X>98O~)f6{Bv*$uu{Yx-vBON-hW)VZf*Byy)VVHPWu_U z(jm(LDZB!xPXFKIhAmO4$f(-vr|W9a(LSF4&hp6fUJDb|hA0s7mML^B6ig+iMRzU1 zk%-Z-K8ErHQu^-MsT5n$^EJT4L2Kh5v!&;jI4M&s6^{|@#4fj1@&5;rV>fAc7<1*K<1RAIqW*)zJW$ieJu$J~EjxV3hv-TL-)tlU0 zfph)b9O#C5A=6#_Exw!uj+`D{j|YU+htt#vmx1`#_nY7JQa`j>b@Bih!RKr^!&N)| z$=3{h>Z%=9_HiZY)E2sl_&)|5=m>Q`_Yblr`hX59)xe+kvSZO0wt84Wrj{h?BU6wT0bh%0q1{aZm> z(GD7~6nxaS^ftI&;f}tDZ4^u*X4Nn63!U;q!&A5StkU3rR-=Ox@*|qQyn42m(Z_lk zO=`Y9&sw#nKrGUyIllq9IiLLWdhu1aPxgRNbg?*xeg69~I(*inH(2{45#ZlM2*{h# zsUif*)31O>*^th!s%*6Y=}(gvPF2i63;1~RolX{~0QT!AB>)lqxP^0c7wC7Yf54{c zXQz%EU9wtQ=ZUL2KWW z=^(-Y#*`lc0OG$GWDTJ9go}@~fb|9w6OOFV;2^5%Qi<*=)v5!j)*!9*Y6U}t{!4bD zQnY7d4K+_NwV?#{VJgp#8Fe@aPd6> zdP#GvN;}qb?18*%xOnm&_eE*P`>bmWUvGl}pEFT_0KIukqxu(`;(ilaJoM5%mS{~{ zreKHd^}05bz*{BhOy$Sh4ZU$6U%zXGsA0SUM8LC2p0Ed^&r{x7Ts@^1{--!Zb^3Mu zSDE-~0Z1-(c0REWN?N}Tq-aw+{TOduDWJERx1vgR1@Q>&fdS7Ra(pbkzrCLd*PIxe z0nVd+dVnt_w+D!}Z(+M7N2g%iWVNeaxzY$Qjd<@}Hxv7}k6{CmD&1eq-~ubJ4;k%A zg(LC8GoH-#114qN+5~WQ_aW}Kf(LgDfuN}Z0b=?^Y`*2-6fy(UVFfX`8ZBM^I{zo4 z0GRTzMLC3hJjq;}3x3ORSnKv3o@{$B%FBDXppc@jAz+Sau~L`iq1JtY_i|xSe@pr- zv+Ph+@TT62;;OFLHb(Dxz;*5G(qN9Zn%JxtBw_A%A+zaj`Jv>MII@>@)^no5qZ$u@ zzK7A;s&{syH#{Wo0$~C;XX-rJAT^gTVeg2TXS@b{4xSHUa(?FI&lp(L+P`Be=p8bg zv40pX$7Tt0ew%M1gd_D`1RY^GfJ@*=>i=prWwU_*Ww@K|HS3>! z1TEvlEr+X~5#4_BD}`PvGwLe5E&LiV=Ep7Fx}V@7pHRK^LTMr*K-YxZoay)Scc3Ej zIKgJFzKOHiN`M9gnLi5U{pr_sMRkWg-FqrE#D7cEr1(T6O88 zS}V{7l9pgp)N^G3$lFpd-CL=evWE97?jfbVMul4dKS0)o&;Yzw43P*7lnO9;J~?r( zya0J6?>$*7`@6VUXPSKtHxJvg{@OGh_WCLIRQSdmezJt+;zvB>?|AB56fj@hQk2fW z_|cpS(9HDHfncWzV*z!y-7e~gHfr(a(3IWLtoh|Dy$A13FpODqo%O03>JETIlumX% zn9ftTg?)bTmoMg{XXxFZ-Z&}SS_>H2FZFz)lMRHQ^xzv3*h47bh?sSyUvbEnR6Y{> zV3#9aW8LHZd5`7S<%^=dgt`F29-t1kW2q|7m{{>vYuGmD+86u(dE}pzQceCk<&lOVa6FMNOt<@fFN|1+xzey&z096-)uVkAnzLPwQ}&O zuWBMn9Xcc$#C4Aaw|&trw4ArIqNY1O>hN+gyH$?}VARWeOgrlvYRXX`jfb;)C3_Cfq|9a#m@3$B z8WrLR+wU>%ggY~=6{@ytXyECaD>3k0Z0 zWUnT#7h-N<+S>Q*zyS3B64hY-JthZ1@OkcR~8Z zk42QupylQrxI`h)10h9(O z=+T#}Eo90OBG;O3@}lH%^a)QEm2%|GK&wf#px-~#4F`t?Zc0nhc6@z*wsf`cK6z%Q zH;iti3qdOY&_xr3F0ul^!8ZiJDh;-2$S7A^T0PAl`95{wh5M!6^SV&&YSX{aO=7P4|-DADQ!tfc(33-FE-i3!te(-C1>Q*ykXa5^}yEFeNGxUF3h- z(DjJ^f!~!|;A2_VgOb1O&jj(B3W|8{+-i;^izk zHMi_otsbQ##)~=r-fNnq{U0Cj>Q6ud6nQvLpr)!_nuW>{_8-Na-KBj2^VF5`;3;@^ zs;C`jy>56@?C0SeP^6S_&c4}10px-xX2_M`toj}`d~KAK270}g{A4-nhR56(HJg%g zC-OtqwZ0fo5{Mz?mIO9qZ|7de>LPUvrK!iNPda)k0)A7eu_MT0TI0AH8hvP}P7*#f%ksu&4f*E{aD>w$I z5>jRR?Yqn*Uar3PY;gqUNoQx?Q3CpeltK8Nbne2cO^{ZkK5)Y^Af8BTa)T&xY6BF% z2}&BhX@-UnF|>N(UwJ0dt#?OPO6TGKspAM!x?S;e8oSNFEBCf{$W!peEaFa`QJ}RS z718Raw^C_uC~fT7`SynK=k3{E-%XXX+}>Te`Y5c~uJdj?=qt*7BE+J&2?GXLvW%2X z1jVJ~t4KeVT_hadaEpq=O7W`n1`K2Uy_8AOX}*g3gH=0+y1o-9G2c)GBPp z0MLO1|Bve1iUFswi`cE4aa)T-viRKe4n*%o;AeZtJuW}y?hqi^l0(B*~$p`nczDvg8+#T1_JTQGf-0z7rs*9hA4 z0Z<(n|I2=>RRNfVa9^3Els)P7{n|wH*h(2+%iz1ya@j@G)Nt0_RHeg3YPLkx0#v^0 z(+E9|r@t=Q9u0U0d;-c-NzjA-+H)(enG01B;T@MZ*_M0KSno%PK}NO!sKE2sqdjGy ze#Y1mfPX58tHGOnvAW~zF@FxVma0feP7f(l-QD?~$vK2fqU6&Gd#G6kql`{vYf5ho$(zeFCb zf`IhAe+R`NcT$S~U74lL{+~$rDdE9_7*r%|)uVpwvX_J0C|7S=ApXVS|HKC1()Ni; zO9bA`g#*xlR4qcGZ5UfJfZK3_MB}eyAw5)i)UdZ-J70+r0UTR$sM#;5+s=Uvimm0s zRPMzE$})Ok?YUA&0@0POwh(BQ>6-jH=$p+nn%;cWjJnZSX`gj=BbpbrBFMCVUn_EC6xk?! ztxF>!@q^jJHyyv@Q^99SEW08xMNa%r^J|SdWZw29v@R3^lq$dNv^E{d6$0DuVGD;N6yAXq-|_()~VRmgJ?>1V|ekK}ECZD-eV zq;>MBS7l>zwAvEY_+Tj@Xt(TQ%$z9cnIzGJDqbPSGS0DLIoJIlNV#tHO5j#9U_bcL zwHb`P=}=;7$7l;Zd;Vw0HY9rur~fPhc^*P*%DPG=*-wZxmDxmimcz}txyGU7hZ4du ztI_;Gxh^ktWr1oef2FIx@VoDS-lkPz%G+|8Fdv@Cc`DTs=m`R9!B>-1d8t~9CH&TX zF941fmu?@606^W2s+e|Su|JJ8Fff0%1$LFTu`4I;2sa`J7$QKK1~@p$<6^gO9Wkd1 z%%Rwy5esWKTHb&pwF;UcrOf~WHp#Gb6p&tB6-eATrE!IGCV=vPj{k2?$@piK@6!uOj}-H>A&c+uMzpqcZO*@57&BRnS)=cA#_7@_v8X z0|4abb|S*tgx|QNq=!{lz3RB~gIh3>_L~d#7ehTXI#yaZ<&$Me`%?{oVX2v{_k&1Y zz`>OJm(0qF!os^)a-GKr*Wp@Tj>nH^@)1z^$~i@sLS4feF=&O@b<7;)BmQWs51)`O z1{EX$H?W8Ae&TKrKCN0WtM&wJ`hZK>tsJI{yzKs9LuDxl$MQ2$er*ylCNo&N`CVZ= zzgqCQ-8Pz6B`vE*g|M=UBK5=*yeB}RA`JxbhlE$1q~$xiDoQ(@nbu7QW?U63@xGtp zWQ06T))qwghPrmV))tOa$EBC0ha>|Dd+y+$ZuVrb`G@s8 zAfyCFsvw4G&`wDBKG)x#H&*ljwKG(EFLd`!6!`o;{w87D@LZTnK@;R&demewAP3{i`zP){tQwq@I?gXD0pWwJ3`V+3ylAWt-a9kcfSqS{%?Ev_cFIGk z5@IhoGGTjec5mv_<@l%TxL%q?*4W(#q3_KWbo)KGC5?pHt)Coa+w0P|l$&(z9H2C3 z`-Z{$pI4q`-70(&)tbmMYkdn@>U1VrH&mDOKo^mAA5oWD;dpZXV|Zih{xxQ_i|u>+ zY7-xAdEg0zhtIfOIww3|LX}|iUNOa?bzw`KA2t!-GTP~l^`ELbTu<1+L*V`9V zz7dtytxlSsRBDl{-o!hw+lq2j2VXmtEWerbjBastDVX_S*<1zfr=t2yj&I}QX7ND(3;d|i|M;~!-kY_NDrX9pUf7Rq8Vhet;G25m z*`8&s^VFjJFdl4^rF}0jE*CXU?MXrNp89$L8|Z$223Ui>Wt!J_{s%9n5~>ZxiMcFz zwuY|Xo2HXK9j4Y|Oz+%$N-@mo*r%DJ8!Ic1pL%e4J#GiOAq%z^Khx}9j5$Fo5uz-$ zg`eJB|L8-I+5m;VVHnxk_{!$6j@Ilb9C+`8U5)cPxg(BM>z&^eur2KVg7`UZq%GeK zd<;+Y@VSel7}nIc=0oa#tmU6|NoIr=Pru~bIH_&{F50CMFa4x#RA-LrYN~3TfEIB9 z*gi$EWwD-mU*F_Cp-+ghR1=XD7wEnfGf29H!pygJZj}*j%RAk?HT|qhJFM&Qr1|~r zwSfI|)i{&^y$r33@GV6!JT){?7tttkqr%=RyC#9f0@~Hr$&&)vddw5mNGQLoHscfXhnO@Y}yvQ33)< zKAv6Mzu(kyeXAUuyH!toA_ATx9Epa-Vbljt1-hRjSM7&JJYmAhM6(ZO^EhK< z0CKEjW6Pbpn%S>;E992W`dn=!hk*_I*Uujw3JR@7DOZ*cd1AtktJ`XYuN8(|>=b-A z?~6H%gFT8KBZz^etms4)b*7*h$(}2LTqO5eE|VUrADYz)folu(x_IAog_qmy?a9Q? zN$5|iclisrTE?4z9l=u%@r1v)qPUPX-^@x!(|G2Jy9vw@%v5f--rr$Jm#ue2*`=Bd z{qn)O-Zz@XweJ1Rkt{|vyvZf39>*>wac6i_Kh9oYEBNjP1qELp0%Of z3LLWM4?&MR(c6JJ*UQi5uKDWBiujysbXO*zO4U7rF&nzU2R|w8It75+lH(1*3RXBc zGA)x?lMLU{J_hV=1Oo~~#Syoi%!l@=pEMuK*>z@r$_tT^hArQl zJrb*Qu%}kO%x1(|7W_czmn|4A@=J?-_(rv(HQ#;A#vdp#EctFwMNt-fhm_AUGNFQ~ ze9J}N=;Qr$f$s-k6F3M=!5p)>HWZ3Its^!q%?^3iRPUQoDtWO)#*E?f|| zW$X+p!OL?8`=kolXQpmz_5hSzRsMET64+U}^=E1eLI~D$;>!yU z$?GyLBYwotJ$Jb0KzG?I>I$P$7B02PG!SmM9cFib*4q3ST@7ne@IdLRimMNFV3#_; zu6koe6k+W5bmI3*5Tj90qYss?jEMxw(D(Kf!&_!_3(+oI; z`})^J!cK~4p1QK?YW!LoZBMt5hdcv+r)Sud7>t*f*=UQ{vi7Gww>SOJwU@828+?LN za^H@n>EZunqzP8YKE(&GO~}*ojOGPvafQv3C_s&$*d&vU4p2@f=uzIaJ>?dH-)t z-pYSR0fpH)uzAVe?cj!6^l@lsncv~F`Y!H%I)`3{3N3VsEpvj)9Qf3QsHq~8r_`55 z8mfmuq9#C4Ml+FSS6m(*(EA@H;+0J52<`KlwPzL+&|y>fGceBy^*ur3wl5th@9$EB zdyIS_g?MiCQB36Aqw~G{LD;i+N-YcBCp)&;zty?G%Nyt%&43EsG5ya>fAx1*7wc^9HPXNWdcc>5 zY2St&qXX)*f;86k=Gq5)osRTFhQaSI&|li&Zq# zr$t6Ht}qx{T9+@x{%1f7iy)h2G;r;_2tTgLTNq~e3WT*#b`lF0%JT=MshmJKQP^x3 z%+1oc=OOMWb{QjC`3cyeRuGVuIBo-8u;_^$@8yw|6*T+v5;K2F*ErICcdfkBeQA)(7M-6T)|caoR2Q)P#~}d@O#gEb zC9cLfqUN0P+;T&45m)@YCc^r@OmVL$W>ADY3ML&tuFd}DBsmxL{4T5bwW|=gsm?UD z-B}gc=zg+tKGu#8fn?L2r_Av~kIZi}o)2;K(dDJeuqwZX*5)d)frtp|PMVv`YUx&J zc9X(i++2{Pk&_p|+TsZ}yyL*8bZwRb5mL@V|WY<){x5UA6L zbaiHnltaAI9a+?9?<7lHs)+QF#EzU8S-o1<11rpW5S-AAHUhqSRIY^{`!-X7LH8}% z*}XShSC6x^P~E^L;{XFWGq8VXB))Q0RHL_Utv#-~smy6BujKoUa{fMq??|?9-jhYQ z5!1NBN|Ny@zIGIf=)=z`SDhi=;~QayB18nneTathso$Gs-qhT|FSq|*bSSE-pK(oiRQKI{X4k^Xdy6;C-XK-zcU)v*i#CTe1D2t$(;cpbR8a94y)M!l2{R(ZJpL$!24D z`?78eM43#7?de%cRooHXHvssMB_>Zdq0_^ZfZeadI=kEt1svVCS5G zAc{qi57?(K_;5rz5T(b_&{5b)d`%$kH1q=AK*s;8ie??nz>uT7*rm|KPcGyM>wTgu86AOLR~#kQ~1po3mEFdO%EA zT5Weo^2P1RkoKX{8$%f){8VZrX4yCGrR-z)k;Yf+WL(mMtWq$DX5Z&UU}_s zQoWFI11)xbz|OhvCC1XHM0Wm^ypm!_#qxu^^oL>+N0YvLj?+NWPE={_K55};oZS7z zh+Y_PlPm~o8j(0NTisqGQZcW<= zO)4r0EQg-(XVpz7N{6GW&`^V_@`s&6ZPeQN{tsn8O8fyQt|BEVNQ+Ui<n~Dweo#dg=prZU zX#P|ruw-dnh1~l_=6UdqtTV@(DFNK<0T2LAw8Ftnanm#P(xEB7;d7ty@rpDEAi-2! zLL<8DGGj8viiZ1_c$uuFv0Ev9ID* zKWLdsd3I!4gfE#Y+b4=$M|+9b{Kl;<-1E&m>?XZVW0YX z$+1tjis$=*SNvxLDWS$;4-acKjDhj`<()r^fi*Jk9_k(1mdXsNb&z}NoAf|k6lBED zwfc;GE(uG>u4-{B63|69kp#)tg(DgIuj{9Z#z^GX`wQ2)b_#Y+-PlW>BXviX1iQXo zMbBn@YI&g7b%m0QE8%wD^}z(F%4AeDQM6JrP8Ah({PG(D>Pk2L9i3wm-!|=}W70BYIx^~IxYlfl&Tz>c&ENSdmW$ZAQjqP>!QYn|-XMH}KtOsZV zhgG1c`Lh-@gono30E~I+i!8G92C@mof%3@Qtj7BoUBMrkPq{1I(AQ&Nkly>ymq)EE zn9{S(A$)~U@D%0hs4KzTkDcR%*5c_^Ye~224qLH|3-7yK7td~v^7h=zH22r_J@RR| z4OD4jAhy1PwOcrJQ#A7dL;Gn%PYR*{V%EdeL8p}-oz%?g}3|b zz3FP%+(f#-J<>bzt#hAYegoqvX$AhB69c}ku{M$cikfXojs5y#6|Tbrj_g*B=wXpW zb$VMzX2#&gV#1u6cnIZ(2zUOL#D6mOr=dVRu3y&4!q5M^NY5i8|`z zYRYSWs}z?{9wn}zHxEhRLCtuPu!&e7@2$WG15Im`%+rYI`IjNUC>hd@G~f0Lu(w)* zUl?ZCQ`6pP8flH8bndj@_g~=}8HnEq8nHC%hUIyrb08ZUpaDt&mJBX%4(-S`+{h5= zUliejYU3u@u(E_G(8fHqwmAGV@*%dg9q~-LgqT4A0Ag(RwG&gW40Ql>{4YClEFrYG z*F`bVw`uU@k2zOxRzmBv2|H#f_D4binLsW|`CU{Br|S8&>R= z*aaNPb921!Z#+xv;gz9n-ir*Eqx``g#xv1DYr_O8Rd;aG$@!LAZg7sZoSTEUay-v_ zxm(3g-HT=kgdPxH`&&cuErNt^&a$_BxMvqAS5-^PbwjkRD z!lO@m9GDAGu3f&yV~=VY!An$bv~hBE=vDi34`vVJ8|R_~$N(DUO#}Q2Wu*miQra|q zdRN_|CKdy0Rx33EA6;tE z6GL8N#B^+q=HohJs-CZud45R!Yl4;V-2J_0Gt676isb`>{2x7Dz+ z3c_EmbS#Rvp9$+c1;`i;&6b`m${6zV562uee+uKy;G|k_s>9Wl#4| zyX*B5Ka~SXe|^%iIgJC^&D^V3i~LEA46uj%jM@7L&=bAp2Kp#~UoHHn{Oa&GWV3rE zy^_m4X_!&9{8!z=1maN-?PT&R#B;%NVw#l(cCIQH-9uZ8iVv&GWY(64>7gE&d^dJh zO!cN#PV{(YXd-{68TC)>I^(U`@wUv>?AKNw>V<=RAI0*b(%Yj^@eAA4uUJ77Q0co_ z_+Q0w7#Ae7rs-M{rQar=yo)=Kgt-#?v3;rT`pN`40t2V%`Z2&73+lZ3%!st*G3!GF z0NzzP9IRX0n_P`cP6y6RD~gY7UGPmw!=5o)QJ%=& z8Ssa^6U2NoLs78KH6ya~&wFr1X3VtaJztx<1Bq;`7@>2eQSgVf(C#JK%Z6*j=OrY( zO#`+(MsNM!#I1Os4`_l9QA&H$M!M~JF-OkJE~S;VH^P_{G0e!XCJ$u<^6oQ(b~KlF z`+fo>MNch2eb1ZJ8P~dLt}snp-x~Q4n}1&^&o(tio5w@DyHdQrw;R3E-Nj6(k1sk%W?Jgozb4ZGZsA#GscH@pnU!i z2y`Q)=H3{C%L^{mLo=Voa5iTM$XaPRf0)I3Vz6TQi(TheN3oxa-v(b=qmL$lkF1_o9^o%?07TDtW-H#s|A7GkiVt)`l8$h-P8D5eGzg|tzZL)WA4(ZJ;!ap&=+blZY@;}$cD-DEg zN&0FE5NHl9&~S|${L@0Y-5IC{egYs>V;UfzdZ9+Hp&!N^Ncswwi6WE6S!#BMWA^FH zsh0u*7Gt)$TyFKp>~EIRI(1AN9Y(aDa}<`1s)ksdm)?8gVPRw%M2r*S}vl(4mP8ZHt0QS-#4;#2MJJBpg5%>QbD$ z>K>|7n&Q0z`c=a0{UdYV!i+Ra>>*+!@5*F`JmXN!0UmuZN0LXCHc~DK8lk2;Y*zD< zB8OP&6&e%Ll>j9VI8W1XRTs)#3=wbDdpb^D!kqoSc#;> zuc{ua)$!!|IX+q;rAS|T4TlSuwBD$4M5}I?h26?|`1ccA0Y+%Ri;j$jl{|i*vnpb* z(31E-OC|=U7!Bk4I{bUfk=d#v@XPixZHRD`J&BLC^aw?2W>2!MRER{_WZY3t!R?Lu z2{cKs-x5KpDDr&r1<|}DMckN_udfyWvswb2ML-MNll*n1!ew%mp#@R2zETv4mT z4Mt-=9ywKs_u%M+JwfWsgKdP8IQG}QQs3!>yGH8d#&in~S8v$qr)H_@--)+1kL*(v z4E~S=WG1I#p>6!a(v|z=cRF5bAtMjk0XNa`O@$k9Crhw! z_>weX?u~d@$eL163RPjE=0cg65+@4))l=qEmpy1Qe_0=YBR(wKlQCfrOabaMTM9l8 z_y)GwAA5=qJKeiEUj*8yxNSWzS$p$zcVecd%d&8${{uL#3LvT2P9&wDdeTs;<4$GN zg+CPz&la&lsnXhKwM1fYa0NRP_my*+Z6#F)A6J$?`6N1HP=!QblaRnX;ne3+A8w}9;23ba?g zcG&3Ph)KijULlBX#caBb@r2RK_m%L-#{O@%)e|;DVi}8j{%5F8kAucKRZAg?mV-Ia^)KE1 zzgK%Cf6>i;T{c}i3Q9^4w`ZhWa47H18{p;%3!b4g$RgQ4H13oB2~}4&aRNf!A3Sdu z8=F-IY?|_%*gN+i-tw#NA(8W~HcdY=q@f8#RR)Y%=`j7g z>nXuk#~*r?4X9=|zs*K4nV0SJr`bsL9$;8)oF3+& z3BD&}bFm%d+i9XmUYeC6(^?Xy8ga%)+lZ~}L>p*}yUl}OKI)~lopeTai|CH`nj7e} za)kxClCMgg{(Z-QZFTMaiXJe~PdG4TbVC0eukHC!9r)*h0OX_vl60E0bGFhv49cPS z<^p7f$B+93yVn5J6gRX}oc3iX@u>Z(;hY>PTQ?nluH5KxNOeBe!2ZI@OiWEAE3egE zDq0(upnA8XZL_HrQxMVcPPSWsuG9c*D22n-SA~38OQJwZT*u8|n8BXdRsN9OP>KoGd(Ux+wE<9|e!Fs&WH48XIaUTcY-`uz$1`UTBm63*oF!;xpug*%g0~z4-PTD;Z?FGg1r!3rshg5 zsLJ81-TgjP@GLtlHZ&ezlxA*RnQX3K8|%KiWmz&#`6~UWO+&qqu+INM3<*r^g+Ny! zzd(+$Hm!^Jt$alSKzvLmXY^n>u+ z3;^(6VXv=g!>I?#T|!I_W9foJSB*ICsizD~r3j=JMZq}5K?jvnFIrSXDtMcLFFuP$ z@an$nrqoop(WU=2?JYND(eIcK&bSPbdA0hY=yKbW`Ducr+kul2#HBPoE-R1Kv!KMY|Rai zOb<9?h=hIVy(G>@69Nf5^012osc)u7$%gl4?WPf1FU_2WI4OzhBm>hP_TLX7AOB5j zCx%_Ok>t$2{SJK_s}W~kO7ku80ro|gqxQosIg4`jMw?PiFd|X*A;dWQdE>O}`!++w z4t=2ce_OL$_5;k;hV05lb5yE7l0NQu^YHt()W<_=C24Q-W~nX%leJn1I)nR|RQGsg z)C(ckH;N>m-U(C#e}IR+n-35P_oi4o9n{8DxH{omLQW5D*(_Rz-E2>G8I?HCG_5~) zH2#wy@`t2U`VLNv>ISMEQ zgd@JAJmJ0<1qVi*KVs`24V|5LqTGF*W8*b%c^<27RxR}P&)fc5xx?^EY4fT@|3}Qf z>g_RVhOWFY-}1(o_eA*L0;-F&)~nuL;*p@2$r9m(|2&XBxmjj& zcU@m0vYPu7&^o76P6=cUD>V5*o91+OCN4+7HQd2t6A+){B&#W?Z{ogpzt1ZE?|N(u z4IxbUHpfOL)K28xSZyFljbT;*qLm}5As6aeOH?5G>ED%EC_eXX94e=%@GKYQE+=mc z{rbaLn3PX0qjnRu5#;R2_!<>9BNx>nwphrD;Tab04>T0PNgR#7ea!No2kXpnIP(l!}Ag1~yY(hNgJHk*>t^vwfs-8zWzz6ud(Ab$J}Y zGpf74HJOZY8~PTIy*|NchC{Gf*;xI|JEVIlgYYw4;R5MR;wV8KYevQ3T$?HriQ zg!?``4$XZ)BZ)VhK*Zt&aqUAbi^}-IgmidleNAI#-vp4G=SsUiPl^PxhiiNe<#vj) zXyJ=o+bcP88zrKr1RJOW;kALyiVK9(8Nd+07Uy&})7W_z)){JTBZz6L1^G;iK?(@2cGH zkxV<$gTl|whD{BdrRy=K_6v?|e-rgq`k;rMgb}{V{f{Gc7nfXL`Bs0M+<5tNJI|4; zXX*8caLhOf^-e%A;m-s9MK*uWC;TrB6aZlY`+ZXU`U_KC({?C*L{ zku(gpkS2++g-hWCvlNPC(i4TuRZZak3e+|c2!s~NRYhvibkgyaj=7k3T5BP1i2tc) z%s~^$!~_3UYWEwHc>NUHHFjnXZu9etM3aqdSGnXx%;KYwB#98F{NE>lLWAecPVj;# zk)VuF_VDpU==SdpM|Pb>1JJS>j@RvI2g6q@j*+>iv461!0C0>8WyMGVmxdcr?*M%< zGJkBa`$#M-j#?ul)*79ss*tF8&sw2{KN@yxQf2)UD?y!}0;?_44Mi7S>nN~noz zbi>cTDKM-cQomtCWa6sW?MO3!)v#RGR0A8n07$!|Y1-?eP3c0;7D=K=XM!w<%Ix>j z-)c8ip%bB@s*^#E&x2f8I_Hk3cb6uhSRPTD1u;k!a+Rp#t5gSK-OnsO2T%b}nm)e= zI#qm7)71;=^+Z?q-x+!GcNLp0hA+A%NVqL{B`;An9!e~eal5=fkO)!$+I~GSje(sZ zRhGWa*Yv6fi+Z+Y{jS7s@8;u6GK9HD;vt(yUNX0!5_>6UuIGH2dQ0()su588d;~LzxN=} z)lz7s`M6s9G8d&gLMbu$vLEhSY?(DA1A!oZP~`sX-n&vcD7Lt9Vk$3m-}rQ00Y`?c z(5$ZMzUy0tB`Z~t5(7U~rX6Fe;LD?oT605?#JP!jF_x9(?r%>Zd1hBa3T7*y=9HYn;BWxVWXX~~`?nq$OG{l`108JM&M{(laFu8AMvgs`ky0;z~xv6d! z;|%^Kd<6DmYYZu0F2pK`dP}+S6d460JBc`Rzp)OqSTK5{mZMOEwlTTEb0|}S70Hn%dFm#2Xx<^ z5r-tPo_+)B@8U)qJZu}EFv0tfZFkG3BVY!vq=V622WK1K+X{Nzu&==hXRYEir*E{I zOq0m4RNpPW#g`?^DPq*^1P1|%L#PyvVGgY!-V+iq;D=w4vvj@Dw9Lfk!u=dyUfcMr&g>3p;Xf>2<UFBEm* zY;qOvKNuq};NHF7nj<0rIB|Gh{|q0%T0T`Cjnjjeb^qu8##^ln;EWQt#loRa=k6ln zmOF}*2;!3wyt&W6sRZ)4_5AtnLk_DPdJx(JJCuqoW~!9)G4x-%41aTc2oD|IX-~EX zjG?lKO}|5qTZ2mqqxU4C&UYEl7Vr_+2Ql>~1mwM6EQGXz$e$C5?sZ<^V9xxMr!6au;-^ME4~c zmwH*hwJv(Bn@7aj$;c0DHUKk(hc0avC+>Y5i~JS&Gp`9!PWE3Ah!VMbKdyBK^#9QH zm0?kJVb_AFgfs|}N_TflOCw6dAP7jeXPMDM1~pLEa8~q=V%@X??w)wN=S2hp&M}mnfGYeRfENSc}`U4cd`=$MXTIUk3hs8wM3nbs96Z5{^^rnf=)4Ylup!)Rc$C8`x zBSqgw-;m}sbq=43?#9VwRmV#xuseyq{s_aNhz7eO%R(6K$h9hU_|EV)CS*!%azdmG zma}u&|6R=FJb*Vxyu5$%W#58Tg8fR#d`mHifXVxT5=_leFS529&`uA4#Y1Pjdx%$^ zcGPHmy9v`b|2a0b+7wjwulApo#KBK0&T|84cWsWR1-X@i52WIY&u@15pw}@y(~8Jt z&hcvCq$Xq{dJ}*s@t^xq*yA7#`kE^vkyO1RRb=%p)-7P^<>^xV6P@z8`qsJ+nE6t? z)B1ZJ?MAFf{!UV*xrx2igO;x6AI=a(RNDohgld=VH2YY>rwP1V=)6N<^AY-n{^>OS z4U93UN%-<|!dX@4vJYx9A&H@8*p7r93)^A7Lp3A|0gc;S8_@KEahLX}OAWI^02wn~ z&A8^t?BA~qz{ji@ zJAMeQ_*ebI?pOR9f(`lj**sAYM4uf+c1)jOpWsa_@pU(`wG(U&K*^NP4g0p=?6^DA zl{H`DTwwA@)xv09&@)6x=zQ4F;xUe$U^UjRB?J)n%2;s2@W>`Mx71z$IM&A#Q%3pa z!5mfR&#b4EN5eIqh3{84R~kH3YWeB2-kdfzi^$((lN2!ZrtMiM^a zY_P+%7}_7~bHv(e*x{E3G-?ffW24^JXd0P|aN>ZoWu|U69Wt_1RK%Mz0pdHDQYmqJ zI9B5pMyf>yxYb9Pi`N`cZa7;dmK5wNzA~>eN+OTZL*=^JWMi4Zi?fE0lPxOPwAE;b z06DqFm;&Q0x?)iIv`|K0IvAL8lA)N+%B~jV@BD?c3V`w}OzFoG5;qmQrBVbR_q-a7 zU3)`YWzX5kew78dc9*g8lusoo6y6Di=H+3RM0j!nSQoplu`(V3I2`7>{L+ZKH}p+n z@;WIFG*^;9@hh`4lqBE9^_2iz;=_hvwY6*e2BDqfq9V#`%+Ia;6WzD)a8a6EL7*&J zmk#&Gy=<;%u%x|UaTK`gBW-UO7)8wu_tsg8p-H+Muy(GEB^_rq^b;>SZEN1=SEry66HG6zY$-4uvO+GDP zWJlh++Ub5J&o8a(4Zj~9oe9RDH>TC;mqABNJQ`VvsSnZ~08q7xi5V zSmA#|$PquR%x4rK+O=Q)$V=(>>r4X|pG84#d`E{P`~j6bE=2G1_1pHStATmVF9``- zQ-1#MZEDD{X15gmZqA%(t~)Rs{Y64_x)NW0WwZ+kxdb}tB^aRYBT-4uYx)C*J8&+5 zP>~kNn4(FP1Fy!I)6Md>_(SdQbT;^E(L5K@GJBNs`Wa>u3kLpeJb3yswpr6EQ>N^~ z&Z(oxJ|)YufqP$FScnTadw-ur=cLynGK>SgriNw%3aPTOg+2tu_YB2>iCZo=6{fYH zw)$I`e^q5ZzCfPXwV>*Oe1gk|HmudLk3abl2eU9+peYj*@LcF3a$)0EgL!SnR)3Jk zAi&3Y9X(HSo>8H$BVQL6qhC+6X}{xFLe zSTr#jCzXqasAV2c=xBJ|HSM7bc6)^H1fG?84AX$t8zPkZeZhk`M&+UxJC-z!#awD+ z0)Ogt9~P!3lQhs^FWww{__3)&A>r2|EQBLrr47azGKl|wSpZUDgg0)T-?v3YOH|vb zXaH9b1Ex&WGOrYQ=OTCTtYEPtF~H%4J9K;xH3}WG#4G%*5Bf@RcmH~FK#lC14M}x_ zv{#RfV8Qn#r|=H589>TyYjdA%cFO~tI9)`kov|-ZH8e)Vrn6TflG=~6nMVuz&*7w^ zMJ3^*Ti_es9*)uyFfi8yo`n?x1iIV`Qbdk?;mR zHG7116nMdy36$)sDz|I$xsW(76|9s9IgaFKtqr!v9xe;5CjJAfXO=ya+^)^ZEX|6) zYN`INbSx}O7DDNqYjrypmy`EPty#n5box2hw^yw8na0&(UJ5dRkj_+>Z_BxmN1Y#u z@o5oCte?|AhkxL~)K+WT^~j6wNr*guLqhw}6KOqett*wqVX>!$>xTOXGKoHUU?%+R z@HuK6r3grf#4fW)Hjao1#I`s383v=(1`I|VTO8M>rM9drGoUWUPR$vYWQHXN(GIK* zdV=x?{gURE5;<9QqUSx^9&9Y(Wo{BQS`B`H7}p)-G|ySP^rWIQl{K;1%uR7@?>BO` z#P!^h7nSg}ic8YcONUu!O*G`vmG3rf!>zJAs7!|GJL4Z^=3iV&=@X1>ZNYLVpOy68 zxeish27?D@K*#VN9vck7lsZ2a{SrCRK{<~F~cyq7p=2~(?8Bx z*%Q_D?FkY7_1Xc2f*)r259%)q#GWmG%Bb{Ym3JFqQK0z%=UDNlCMuQFr00OJHTWV8 z7!kc_Sph>ob&thsIFJRiJZU@80;G2Y(23Ce8036Q3H~!Wy+U#})0joAIqzH z2~mqBz^=3+@>73;h+6fD+$j2&;A=;d-I7zaDH5})gPf?>Z~Byl$EbdNO>P$G&P?Ea zUv{10Fehp7i@)#WV{4()*ux#OfNqOVb>Y*tYXlGaQ@P*9C^#va{B~e$dP!fBEcZpH zbn5EunaGgLA2s-2PcS{h(wpBF7Alm0-z0bSRTRXZAthtDtEZ&wSNfaodz9AFu$glJ z_@vj;1m9Dp#leRkUa{!B5nx_zq z(6TnbA=lLV0~FY$C&}Gb_cv!p)q5K4?loDt#91FvH(gsceOn8EDl)@~0GvQSa;8;Y z{d^2CO12-dqKk1%66#vCVyr1a8*Yf0Y2D!K5B;)KjQxFG+R$Dmk zL25~{zxrOr)Ya0v#Am#M%<&o7in+@po!zFnY1a(PXJJXMcM|8a$!;q|Iv0h)%S7y&fQWXW2sxOc+vn z4Cu91Is$2acIoMM+LG%TTIbIJ!^s1$r7p9ldqB$@tU9WBbCgEu9P`FL-tNmKju`t( z{C5v?MaQf(3p=sjor@OxyNJ>$}bTeY(ZcE$D469VW==udo4=nFYf z?tZJ4of+|#+y2)9*P~=CTQ{5{zmh9w>q0l?OtC^r;pbOS><=86zKL+|RnO$xjeg2P zwofIs1Q?|2m%o?dX5ZtImF(%)4`YJ~0u7vE3IakX1iBo~4eXa8w)E?=beyqo%VCxd ziYhx+XT+)n0$lH;Dhhi@aWu5Fj~=|ZB#c}{I6uTE7}?++^gY|BO|SgxMnzm+%yC;iNuRA9 zUx(;A77l@)@&}G0NMe*rh%n|z`9k0+3r;?(!yjCiCY(=o`MmWRrpKhTcmjl@y>Ae7 zP;rsiV%;wCSxBgTP`_GRzeLKPdFNa2&&=t4w(rJU-mXp0hz%{Oay!6{eOLb|GBsPN ziBIpBX=yG%O_J(~)dz*0iDajf?heU;7E09E3cGg^6$1=jF5Nyn9~X9%60K<_^UPB# zSF1xXQo*H40Qj66;nckI5i%0Bd{d^dl_n!4y$pG_G<$1XWkOs z0kQ;Wn%t>g6y!kLP&emsQyQ*E%o{=5QsC{96Y*K|%r#4k0fy2z^fOtJ16;;wn<0}? zvi{jrmcEkV%E| zeFkuwQMqw*2?uxaXA6$9+2r%jHuvYjsf-Q$o11-VKM8FX*i@YTTe<|`3c2K<$j#Fq z-5*_OcR!x}7DrQ>81^NTw~|y($F!xkW+>Y*g;}N($ewIUY8$4VXMENXCI&(f3R6JC zLG%jLq53UOAA1zPGwiMVmQs!mEBdEDit-`C9>m)$8&W8+g4>NJ z=v{V_s=_qMc6XZZyHno=3n2PHCp7xLRZ5xQwuRp8u1x0ZfFk0{7p2R#J6HH{4a=0C zo>Rm7{sdlMynhk$I(#k$&N>DQ1}X2-lhv}j#OU!OBV0(ScN*_X1}DzLcfN;658{af zhhy6IbcKKzs+!kc?ot(gq9T9g)24 zjBmuBCc>3+fw^0-7pfL;ew(o{%lx}7|EB`uGg@FtGF~VVc;}z&%hX82?IQgU1cWrn zR2tpesX2oT5c@}eIv+ovA;9xUAu@bpnvwoTRJ=Mnck+S)a-8bGrCX6yK5ehweR8oe z4m7hMASgTfv0_oxL=iv+lza#RuBwRE4zCPsc56>rfCFxzKg~|RXJm=6=pW1EH>*3> z`|C|U%!Ox~Jw-;%2XUG#%DRJarPqTdj|m-+1+1Ud+9;379aGBd0jXeV1wLgGpnVT2 z<>|wU3JX($Y7WYT1-(7^`+On_PJU|fURpb6I0`u+JMrPvYm%WkSurlIKx!3Fy|L<> zf$CbK6n%B}c^TdaEoD`wJR`y6)U8RH(KiEPLm5ox_`3=jAzNEVdC`bT7uk&NHSB!n z$0ncqAGz(mph@bm(}=bCGiLQOBpp|HPzIdNk1(Gh4qW!4mym{!>Ms^vzgV!jxISu5 z;yNqYaSo=mz4a&6WvL+&bW^@hP_G0^w@4`L6+}dXA>0enJEWq@m(c_gxe_# zd8Yeln~gdgjocc?2N54$Ow2G^j!@_WW1n%V_+;a})LN+a(J-O>lu&#c=niW+#E9A@ zI|SZ82eNz)k~E!`6NR6OlO-`uUeM4am24WBNBjm+yL3=oFel0cVr=8B8;2Jn-42)V z4$DEXJ%z+-ml@)cP>$dbH)Qgy%n!P%Oz@RLZj?&>e|+RWR| z^Ecb3eZw1Uh~zGO(p5W}pvRN^CnHC1%y&)bgPQyCMB(vjWLr_%%6Ex^nVFW3Br0F4 zOuHX=jjwwnIv=uZZwXq0QhGXTPdq5CJeQ%2kN1FRhlRTkO55Cj_D1|%u&u~FKIL#Y zEc>%kdXM6Vp+jKK_oHq&+5q`}`Rme`$$&pPOugazHTP!uQq%JjOghb)&@8KZTk)S( zr48Qc#~eSLoY)s!99v@>IUDFBa5%bgCOf<@B#xrG<6HZ#a(YHoN+Qa$&~Z~l}k z*Tl_OmE!ELW#_bB+yxR0o-D0bj%d5hb^47hg)p6?9J|&573b(2`+(LO*B4%*LaUT= zqkqnOXY)1wk@cTjF9|x(w&_P;jqc-Vp(O=Y+6wzYz)gAOQD1~uTGSBd7#F*`KZ~j? z_Rhy=QFM#o`aZ1qBsd!cjxgKmaqizq@4t&f`LiMv)f{;eQxr8hr4WVb z6Vq=(xoNk9Di&r8CRsHa7X$Ck%`PqQh!QyamcI*Rrcf&#Po=-Ck;1w`xcbd5(TTOV zE36QH9q>+NLuT>D|0F>XwAI_KR>eQJgjX?|x6*PwLOH{mdrCK2?scM<;Mczie59Jb zTb&`ErZ@go%H*nv_;{qf-`D<{yuW19(F7smEp5VdHCT5pcH2XCcn>KWC zO{ee0>FVmnbvHNQ>Mv7lQ<++9yLuu)Y3rZ}CENY<*gMD0Zt|46FRWZsa5RV zB^GYiapScoPb{M(fl++LmSclt+R-}VrcU4?!nNhV=sYQzI>ilX))aF(uMI5xlL(bjY7WR zJ(~nPoH<&!R1^ANM-rPzR8I481_JnkFSd=;?+BI4?v}cog#oG=7S^Twpn71+?Q+va zM9?x8yZBVX!Z-7C%9qu%m13iollt?+vmU+A7HdvlEVPtMfIHtw4h7mT+v=VW?5~{5 zuZyXkFeGRXcOUK7zI4!P>aw-gkX%^iDLpWE7_3p`=gfY-biDP7dLZ*fPJI>*=9gJR zwY0}JzijN!4c@#?L@e`(9Q^+cu6wX2NRL^MSS+{#@oem$_*Gi%+B7TxL&(W%g_D29 zP63tgtutm}Tn<2Z$YyCG2($+qZ8IS$bMfPxy3=fm9{QSq_!?IkOK7uY;9i@4=Vl(M z<)EfwKBx?9p!ZPjSJ=;348^9FlxE%zu(W~KC38OyjB=rzPI>w2Zea?%MGR`(Xt!AQ z9MH-m`EM?V6d#d}R*`ZI4`17v?^T9bHn`mu11}p=VkIIT9o>;y7MwBM@`Lxi^se`u z1%uM$9*U}9z|+aG{3bL7yMZAT*yNlB2{L+-((4DAOVH~>rP}F$)!{2WFeZrT?wRL(>+2+LO<*Xl;Oj@t7-yz+#6V~gG zAr5~mkx}JsG{rj<9XvP#%Es~#m>d+lKcSdAgVW&4b;|eMb+9nuRsQhrV|Bf))kYBK zhKIXQ=<<~@gbX{2q$LAf?SPKP)_D*TPXLQcn2S2m_-X`;?~#*w)lL&1K1hbfoH&S7 zI+!r?C*fPk?q6;{;$&^$jy<)7wx9U;qSRf*4KC0!6;C(@%;otZb0M3qV#_=8_iWz| z_O&tP3?!GRX36Lwa&nG>ul<+jnyhDf7%PF=RJU?w+EXxJ~{@Y}@@h$rHPi8rBJy4O`wlMLE> z?B`z@`wb~~B9`2i^6x;vVxe5La3%9oe)4{kb#u*YnbkWl9SATtJh<2JRyrq=RBe)I zgzl9E+#fE7hYsoV;A)4gSgjD?3c`(6`K-L^oZgpM_4LI=TY``~!e61{W{8{~dDk3b zR2>h>hY86oxiT^LtaUnd(Av$#GlqU=o%4zT*dJveLJEiT{|Ng5IM5mbjMXuZ+9Rfe zrd(!p;}(KXo-StcCncY$V*e&c!&FkyEl>lIlyJQ9sBkUatpRTq>1@>*Y=NNPxp4laf<)_fF2+ zCj6I{T};YtW2+CSx}bVqmheT}<+(biM!(WB1Is@nL%=Hq{Pv+Yd;&7ONH0YOLzr`H zC$!*0`G2)&L1x%G^{(l&Z(|7Oa^?)K1UG!H` zgS*T@pWL=?2UXhknxEmq0d$L9@1L*^b(%QN?;pmdFIJ@M5Y>zDxD^#fuN^x;ol1#i zGlPiHe(VBwGABDydYMJ+kc=y^(|7o}YmkWlq@s6r%Z-=QOC|Fz41`)xK~i^f;x@{BR*)o%qlYm zq-cABvX3iu9^dDlI@>+@*_&>yFGzvC$zgzVQmBg2 zi54<3ysdT47Fn+Jxxe)*rt1v ztJli??mVl>-Pkl+z>PWSU4i;}&bx^hQ_UxP#h%};DgCVO1sy`zWo&tq6YjEhdl{VF z>Py;N!9k#kxDB?=4!$j3uXxHNXmwfh$7dhBP`Z+VZJQCb0y+BE|wfRbo1f@UY z{AK&!fhz!AaYwsNe95Nti;LMEU?|Ip7=P_4n|)_PTw*aLOtz#Pu6RF9GPwzo_Hh*5 zCK%_MuBy{@-_FHe08^h^wwt2-bxqMsV3LP#NlH$dayG-SPCoTzKw34Z7<6A_PnTPqxc)xZvdKc)R#Da^o&wSx zoGuaMJQ?DyOT@pcOB!tRbZPBowz)h0kGj+xH_$dt+bZU$jve=@(&P^sSZa6Ct$^|} zualBLvJC&pw{IObGc8n`=S0{e#aG=>Y(&iF)}i{zE9OqhRdjQbuKJ#!Wo~*aFkyx3 zDaS^=8mVg&+naG zIv0QcrXO~zex;!L`{u*>^z-WC2o09#jQT;ei6=hJG0M~E1$p>?oiP?(q`CT=mS)hA#dlbB z;fc^J#VZyi*(h4yNgsjA?83)IpbxB(v6Nd9vJYezbZ_vV`B38Qh~(uM;K-NwtJs&% zWEew+TYmQ%L!M>I$b|L7J4=lrHjnJ0p4iIf{QRXkdFAf~B8c%nicBnbQ zdh1%H*b0}Nj)Pr3iDy?90*+Xg*%LT4VnsX7GfyE8DaA+41NhBvkTMUw9xhF;Iqe4A zItoPXl8RP33=Snz^JfJRM1gCSY4bjDsKidvIV{QB9X;^M;;G8)_!bGCF(tTNzeO>Z z!&|CzVWgn!D#Y#H{?y^I`Y^8#qGvuWGyUi7$y8vdAs5s_m>%qhXU1rL-HW- zxE0I=SqZ!a-9B*4Rkfc=sj#x|e%mPnftLIS)%Zt&i;1a)`{ZZQ>=K+tKBYa|4fR+B z2nfM-1DthCblOClwp9@)pdHJ35_HTV5W*!$_uHu}gvYQL5fP*bXjqPSL7P3JpogE# zAU!r)Tk21f*$!!}BjaFXKn^Kew{LSQ;N`G#xl5wx*d#1m3XIcU$3o438zVddv0yh+ zQExt+8J#Evc(DhYL*Bsv8W>_$bEH6NaOx|%0B)W{)O5-SCdgMtoOtt=mQJe_0P`;j z)CDI-*V6 zxVb^lL&7ZY1+bBI&VIr8Z0ymUn4(c|v{1k~UF#uGlIiWyO1!O-b>)4_%M*ZWkS?>tAi*?V1-L_bRWSNEJj@?+sX!OvFspr|>4wqesZ^yM6f9!7em<*G)%(^457kC51OzQjF$0oU{j}C55DX{Nj zPVQ*5Yy@8M2HHkKy6q;lp4lA;YRmAgb_&Xa7<74rP&aMCQLPBbqN}QZqDZ{*fgn7I0jl<{v%u@{ z+Y1BO&7i+NATqlG-Sof=bx?zy`c&ZyY1Nzfj00U8hx0eKIXJO5t8(pSy!3by&HoAuZKwUU~+1P*WFi zT)p2f(X+KbK=)=;Os4Eo#6i&v=o|N!8|)I-QG_}&srK{R8XpqBq?RI@ns|XwHfvgODsFOwkSYUXkv6k3-5k^thfSI-nu$^0*~8nCWX zEdNBfBKaghyGolG;J0&gue-oab5bpNPD2p@vL1LvsOdS3}O~>ibm#m{qo?geLh^AdqHenGT4Cb{s_<~Ca~sYP|N;Jk3whA z(b=*TLzARyy>pJ+IWs}AGkxcnlmLvfsh zJAXZG&NyOALC46e0z{aN9i-B%U0XQ)LMs;G*>GU4-P^%U0(+cXAUo7MCOag5$F$Nc zFaqP1!4qnn-y9eKsS5Xj8kt{dH?J?-W_+zYbC#}VQrzW*n|w^Y^)r?6_t)K6PUvGq zB*kLI;Q|r=PlwG3dXqLoSB#r=8PkUPcXL$MPUdh{k&+y9?wK3g?qchnbL>a{j!7>F z{00x_BNRW5c68jaN@N@{5|PQ1S`BAJCOb<~qO!bwLHpKLDffB($Mi)u&mMBrI(04Ah~(Dkuo(vIs>*IWL6#-&hUOE7)lSS9 zyo`@OO>sN^_agnXY_*DQ;~YQ3cRJ_RDZsF4Tq0G06LzC(*U<T23ong2*<^%^3Gk<;fLiOhPRJ@8+zDrOvnkX) z`k(fl^jB7>R}kLdynJxGK&gFkwX~NZabdR=an2;XA{8Jv1Rc9w$3jD$L`@xfx`;q& z8ig9C1=h*>@ONfra*X29gRnuLb*~)+>15y4%s~qvEO*mC@;g|R*2%QJw7Sg9+r|2& z=9Cdv)NaU<#IFz@@+>VoO~LG^g8=Nun}&KJ(dX*wg7fjVUjww(J<4h~)8|YaIgYn` zgJQ8%gZ|_Bt|Uhbz8QCw%feQ*;bD%P>a%zRZqq4LUYMj_v_RXjH zdPVXvUwpZrD3-iA?AzqcYnVTyPU-s8@15kj#h`%8p}YZ27UnoBiFno7u(VGtCv!$= zRL_4Y42yg=4qO(-uy$c|=O+0kgMo+G^8m|S_KNPVZ@TdIaxmBnim3z&XC@TJ8>|x_BEo{7zWzwGbf z@HpO31%NU<{UH3U4A1C@qPB$Gn{^!TZ1l&SR@caF-=Vj;kauSy*{zd(u{#~{2BM(5uRZy1A(fg9CP*-b>B9(QuUkTV=u zOf>yzEiGVXj;ujv-{mhl#pR$f=MJ~@R8rk!Gkn^yg%Nhfll+lM+$tJS%ABCWvCct( zf*gizj-(?5cbi-g^0bvH%u$|Za?)SB&_eD5w>L=;<6wu-QZ*RgDb3~jA+xtroHIwU8M=aQy ztLd2D^WJj#Q-z0lb(lOnMJyy6*VE(M z^x{3`KI^B#PqW^4GPstn2+Z$TDlgtHlxoC(@2}9+%EZStgCs9J*ek2TLW{v}mMSGk zmwoRcUOE}W4$fWAz4Cc>)11v2%ejuO&uoAoTVAS4ZT~U0;fn4Gx{uT0INzFiN1iqc zS5|BxJbvTvxhQxyBK>SZH@R7Woj)tQd7_1T7a`E~xwk9@bZf)`uFoyBeX!a{Oqxy> z`*3i%{8K=0tN!9JWFt#!PB%?rIO}wkXYS_Ic)-Sux_`4K$7%e=n0Wk>_Xe%KzHly^ zvYGAsf(IlK7{9BdR6n)oqp}#VdH1C2`Ur@y17O5DdPc8n`z4w`O+&rO@peBT2zB7+ z(cYadrzyWsA!&EikUf2HV9UDe{+Q$m#(zB1337KATOX=`B+I=um-;#$mP;$1yckIb z6RF|4s#-7eMMA2$MmbAX_TG<7tv^zUR;PK&7H)1{GM&H)mm^D(fmr<9M@LI~FR6(|or^FYt}vrZ<7t8^o*9xOp_j0IuV_HU zRqKO$nKKOL4_3Z8eO@6H=DX*NcNFyK7b8z~Kb(2zczjd*tr#fQJW;ZBIBL$zKwTnt zdnsCy6JnRJ^2=|JOKTxPgUiQ4Pv<8K?2ROzdg#sq#e)&GA?Weeufc9jCP!g2S;G5w zGq~%vQLlgJm>AFL&jg5XxK)a2+3)RJ?k%H}{s#A3Qwf2i`3!HSsJghdzm; z75R@f1{VICSTKPpeTz=QyruDrZ-mKo9r%3>KQdLgVVIU{8le^+>7Pei2{_O|%oSOi z`~z0LSMN1Jw9Es_FCeeOn^{I?ImYI4QonfUHXh$pX|z8kH>pi(S|QRbZ=p^~ZVWMn zd^f)u3R^nnG^1KPGc*mx`u;rk3$dt)sww0K@x8S3iRtP_^m1BDfY90}S#Xn9dvLSy zGvUpj(0W#)azeFngIFRPid>I%YZ)jTj*6*7I;tMjIHsQKY$3DA+PF@rF-$PuNLrP* z;^+}&69b9EUCmz%O5mdu@L((Lw!4@Z z@mNg?-&RwTJBfC#zAPHu7aq|tlZ^n+;(j|wrPEfuKboVTk8**wtvBahM|nVgKjyvCB_(%o~fVkFb@8uB0u6vqt6Q7*H_>oqUzCXFmFy* zA70lXcK8A7R19+Uk-7gH;;cDdpDEI3T_|pu(4|usHp~h47vge`7BzCj2Ig0t+pAx) z)sK4|T+$OyB92=`zzRcNIpSMsOX{=CGp?y*Cs?(mpBaU{e@MKSKuxC26L0{RE3(A* ztj@K^_c%Qph-^;A43-u{^>smOp(513P?|>dHXxN~X*!RE$m1)uIOm22-=Aj!C^(1d^*$&WSda|H8R)m_48F zl+%`&@Seu{*8U7c1m|t8{#~rCO>@VjgKOWSOy93s@@L+plPg3uvKQBx=o?wBzE=^! z#&hwpSoX0&gl8z|xuGFX$8$5s-j~lHx4luO;de92M6eqYeqFfI3sg>}PwqE|G=Z;t zmIitGjFnr_R6WQ(#gZ{&V?&R~7n^eu1`64kDHG!GBYKMT|!Cb?km z8G9*FawVq;RK{I_I2*srjPrb#GyY zikrZs@$v&`^7^q}9iy*{co`vqPV&0w#KtYL~>)NkQR zZ!@+TukT)R?J|;Gx(??HMwF?Izg|-&(X1Q_M|U*7^9XI)BfX1=QX-`@3(+d zggy!C!v~dm8W&6V*-phBdCl17c+cNwR2c}Hd}q&X*imYSOWbG!uv1Y41FVkpFo_7Mht|LgyRwW=0VSTwGw zKK|$zw*l2ct-pDei_$v-S(uzn1KzDvTlN>Zf24`)J*M`ky6B5hT>7lGR6BXg`6MJR z7K

    s0qE3E_FPhJ-dbq9A&nJ1^e0JTl2_9jM2A>@K&JD-WQ+L3#H1$*tu(vOeX~4 z6-|>m(Oqs=SKD3>mU5fF4!kQ!7dI37b|{#%)rO|*QbZKiu>Ed~ zTWJ(52>4whdcVG{)k~Xu~$DOo5#FFxoEhkAL^@cW~yJgB*AtDJC6~ zHf*6Sn|}x55$8~Jtk|7LjvsJv8_1azpK83&hns!s0%`a6Nk&|>&{=odMiS-F+{Yl? zIkwB(jGtIPmu^5mv1p#^+~Mlns{HvGz3jreo&_KHyy+Ij7G9Ylws9Y;Mg-1Sgg(CR z*uw z?%4C2n%UygJE@3`0CZ)z`l{*WmGQRkWfEo(Sr}ja+(#825ZmoV_NFn)xq9Q->-KP- z8C(}3s-(oxN_+~lSjF|Bi}#m%iZ1oQ-%+M+VEX07iT-H$7Fmni?z`SrLz zHFsndW;}ScO6;uebCrIv)3{}<3-kmdYJ*|tbp}#@#@(rm+tPm7salIWz%dj231$D< zz?I-Xn*;wRxhv*N@`hB9kqUh0B^|`IUG3-`0kTLM=;8x`_=e8b}g5B8?JG&acNKSE! z*n}z3w)x*uPLv_(Z;`2(;L#d)-HI02vG1PkBFBvM&6Yx;gV(|JAbfZF91>~wa-#xt zsbF_;?8Me;cQtcccfEX8zwUhtgtKVh?%s0aMXRqbz)u2P-zu8>uP%)F!%;|&g3a2S z&WMla_64;v)N*;jUEP{+;KU-^^p3T>r zDuA;zG>3no+w;ER={TI&1$r&>%?BhM_mu>8nko<{N~WWN#-pEnaUnK~5rFe&uI$enxoDy?ldS zH96=NXi%|oyIeb|V8TqzpL5D-1FlFXwt@$sz?b_6H-;q4IYEY(Wt%xXdBtfxit-0H zl?w^^uJ+%-1|}}vsY3aPd{R`x|3Y2FNnm)LG=}gTy+d`0-l-(4EqJ--u>^5%{Q}CD z*6g?kGv#7sk6F6!M7W-D2lO(QRp%Qz{i)PZ{kJd z%$BxU>62x_+u_taC5YDBa~BBDluo@%b)kGt+-CS;!>hSJY>|WK*6nd1`hk(DDG$^O zo*!i&h#aS=f0#ReP-p!#?)pAYd%Vo$d}b*kqq9kr$4rbf@JRbsQ3}zVg7S){neFdV zQq=ppXE-Gazm*$%+}w6htu4cI&3dJH&*XpUKYWOA*4&;-NGE?<(=mNv<&$SJ8L?#~l0A*yRklO^0E@hW&*g;B+8&6T4+pJxbE`2W zbM+gTD6LT)@Toj|A2ngTXn3^b7_tt|a`*&NwEiB8R z2|B}xM54FFeKBNG_6(n2d)hIaLN%O5H&FwrX1eBF7PKXkv#uqWeAh`wJEqbRx}(-@ zR~7gBBTZfNszjIH8qmhwsXG|LgsIO83cMS0h~sH5I;JC)b4q z#4Wnp+kWrUH}`68*Dlh~r6k_*BRgE^p&mpp?KMu{q07!l_&!akp`3ls!pGVt^Zcr6 zGE+qQDc*|9+bi$p{DNX942uM`6>dLhULznS$XOb$G@mO(K4pUXIdHF@~7bzqp_ zF24WdYd+TYi&K$fm20aS!xZ21uKINonR9*;_tg{Ym%xJkm(l`uz1l~k=ac>q`wY1~Tell0y?F)q(sXIZofK=SrzyHy{;np?w9)*g`I z*U3b#Zk<6>C)f{uUv*4gO;*%CWRc32Igi#;D{as_V|))DsW43SbsdUchlut}xjr&A z(LL+yB>)B(Yz+=46N{Qlwd6--bkC|p9->R6x$#9QafQjzBiUm1Ec#q1`lufPv@dTV z$!wLe0ZK0VXAqi<0G|qeCO)mN;2tu=PGCp5sr1oLMc?vS((Hk8T$`H@$Wh3VI(qCg zSeUHnE_SQKW}nmC11b}<(mq9Rj+NKi<6r~f#MDQMqiQi6%+gV~mETAZ_bSe{%^WL7 zfmOjy6AFI)0(rm3-uxu++qp1M!VHMUP9#E}qnQ+I=;^RAXp+mI<4$B8drIW{Wt{G| zB*UGy$HTdW!Rg_#_jCE{eTSnX995h?;&)F$P_i4Jz%0Edx*k2!rC$}~`6W%xs<=?v z^hZbZ?*yw35REm6GtY~4e_xUZ4!8CY1IRi;H8^p}98=-l?oG5#t@&G@f(kVQa!hIX z-fbp=^a~Rb6mhMHg>EJ>u=$*(>sPz8Dkma@Z=wUH*KGr>oDqgEGC-eKA(tPPpG}EF zi$d7q3+^MN&B&A|Vf;mP*BJ;tF&!S^2kweEv}9xRzBEh4Eby(@`U19n>wB7%#DeiR z#1#!*THPP_O3pYObmQmxTka1Nr!!A#>mF1jMlEfY3ax#_P-+q-q@!nVUbox&l_6~W z?0$UU!B|MAB}|#zG(G#?+-f8b_s;$GV{wMDrw#q-Fyx4z&GocrA(|p-$8eROP!>>xVj)M;+C9;x^7MuhjSPndKqaV6Ee~yMoC3Ljn zJ_#R_f@F_WY?wPUh{!L$N$a_ErW;+LS^BlKSGc`qY_1J0h&>5vk-2zpo$*ACW;D}} zW_^Nq%sNfwrTej)5AgI2Wd#VF@$>9V$fM3?#Y9 z1#n5MlFD82*5<94sQs`2qL~};_XjU#sZw4D1fky?3$%$c3-Y-9TAFRlF865T+lM+` zEHE86G2I{}$nQazrGu4c<4NQpPZ;kYT%_wae!(v(_pD!vqfsF-Jjjp^r4i}41z!D{ z(P}*=H_rQw_iq({p0Z7Cx&{OvtQNp2M$;r!Zx;|Aldegg}TKRm7&JYT)!AGs6LQk^;L9y@Y`e^Wf zD&b7A3M?=h+8rEh^DbECzd}$P~0b;8*`KnYw(exlK$I8x9ps?;5a$SH7ei~!u zNOs1)$=uQas@WoG>ZV#g>$taKDMAYxGpl}T7eOK-dLLZC7MjlR4C6_9l(lX5y8+$e zeR5|Z5cI-Q@OYr#5~T-fsZD{`z_P_VUT9BJ8eJIP8xRx3FQd1rljb8faha-b|1|vb zC;U%KEmDW4yX^o9+ga7zD-}nqN^~fhnsgi_3UZ>4*hl%#|9xtddjy7Ebx}5I2=>r?hkO9bYxe)DVF`D(G4j2hI9UPSX0| zNa2*IQ+DT*Xsi&$)iYlHxK$@>z6Yr$_vPOiIcHjv>dcNGrpGcETnBb zt#<9yFg&}s=i3EjG1Kb2D?~okT4*-)ym%9*y^qro?h6(WYmPO2LXo&RBE<@=ukan; zkqSeKwtK4_$G;c+80k0i(FJioyXg04Td4{vv?1>>+0Ac{Qfa=Pwh~r=(Acq|@b1Wo zP5B3z!jGh$G_2nWme7E}G^OVF6e0N7!En*N>8Q}g?}F>OFX96OiwX^&3b_kIQCYNZIrY!}0g|+G zDJ9Ym{>MQ``I(>A{~LPIhe?6tz|y+dgZG`hoVPW{ikPtRZF1AXM{jaGn?XkrS61G+ z+6HoiMJstDgQ$Z9?r`4qa~vO*9Jfn-$bD2cQ3QG4 z8FfZl{uNK!!B%!eGd_n+ipQ-Jgopa_97m7`B z`$t<)7g3a&8#0RB{zd#IOB!SDhR=5(MVq>JDuI%?sCH@C z?Y8kR^o`h&>?3r%bc{014xjpF3O_Fq)OlO+5dHJrq4ftsAm~$JyZP<)VWuV@G-^a{ zt7?H}YO+h(d}X>*w4iJyUxe4oj^Hs(XS8+gBR(jMgX!u)A6tE7I9Ty_v22l_sNTfR zXU3Q41z(Cd?e5*Kr?d0x%1M?CXV=G&tGOT*VmJ0pjs)jTo!5}Z^L#P{Q0|A-@v;R3 zTTaySG5vQNUWBQO2w;!D?SVa}gAIam!ERpcvbl|#4I+EWp;C-267aUmHuIB*og&KYIE8AmMY2?-_Ko5<~nC zF%PK9*3uuMxnCywEh)TXt*XdvGrsbEpXzp8p6T=M2=BJUjz#z;2#fEDxD)BZ#7giM z)k6}*9$k_4`WR2yJk$5jrN2E(aNb!s@FH9|Li@o40yJge&*>Cq#d)nW4kou+DslR_454@E8IF!OKE|wQFB?(JOFC36eraQ`DRn7S#_S&0S*5mZ z{Js}mX+-6HCrY_`6@)&1?rEc!MX>m=R>sfHEY3b|HOlN?E1aztDKU(?U59g5l4X@HQ>woJ<909L*PCEb zi86P6E*y`%7}x2bINY;QN+>EVtiw1%+CG;e8uV-zp{_=c)UNA%a2yDR>HVNuqu{vj ziucQZu^H-k#`hi7i>4>U?`jYP;t|Sz9$)9Z_g^#qu#hkt@G-9}#CC|U5X#LO3-*HxE_tT7?dF#K&VSUOJ9IMdkrLhu-34;Q zp|FmJ9tHXa+cdjmkQ*gOE=EvQ@srHLEv@SOt8$0)Q0&LgVt z8key=5pW_TfX;qt5e{@ZB`efsYMA77m}}_rUd>f^S15Ha>{7vB5`->6$%%I4pJX}R z&|6byT+5NGZyNUzw)@tXX*D@X`dyAByar4=_erbg-(X!I)`=Wd4j?RN$VRuLn7Bk(#jdXmC zvwjT3uPX8dxCqL;8pWg@kDI@iZ(GJ5A{F_D+pMVLP!@ydVQHtCu@blw_dr9dJC>hs zdZY}M4Smd-l(NNW)MT$XX?#tZak`@K;eveP+&m$1nEn?_dIgOtUr7cb4r zpyo93Kw4{Dajge;d9&Uz=pigh9Kp{$nD2g&7RUGb=H$Zcz+orhcj10yoQHpyXni+~ zt=_Z8H{T66lR1gnw#D^NUa~KK-p0>S|A<*T-(m!oYhLZ!rMOZPo=>JG>>e!Wrm(*- zK-r*li^NRwK(2@C)*{FT$-#Z|t!8Htq0kMx=U@JbmL*FKTbE;M3s#4zUGY-;O4Uc6 zKo^%5UV}VgU-f!fM)=>Aw>yiqA=G;h zudAr>ZYS6lt2(~Jm{8Pdxj1JA6tpao28^$Ge~en5HITDA8sMvlI9s&+W9UePgGH~} zv%VWqS`hjql^uLvQpek9it}wZ?r4i8NiN(V%e!hmyL{(Uo++Fk%3dhSre>sb2iS^_ zV|k#>`S+zy{pPdh<*i-O&6LK!)e*uJNDt}=eE}0MVrsciyW11o0LI@bEM6FWMuC~m zD(z2A3tk{0%}Gm3T3h+4X|Mi)I&A$J0zR+r%M=I%&~s*TKlz5}y%r2?1f6@$AcHze z3F(l8O7N-FSdUYyn~ah1FGB6uDs^%Ay1Twl8WbwQ`(5>aiDnnadqMH~7#0ieh~j%M zOMfVRg5CPGdl%I&Q7e}loC;sDovCmQ86!mtN^!8peKEBDDgujsnTZYwsf0owV&MiiXX1T5E z17Bs95gT8TqIAG41KP&WF1iw;!I{#iA5n^+K^B(;_rM9masv}SPm zifjSeR7nmby1`6Hs`~@n*GZN*eVXEFBJou;2aQ-7UTZWNVj!G5@9P^@*y`RyTd)R7 zfhw5+X-BDwbYv+$hYI2({pc zk+*zqD2=^R8V7e@EL?!qT={tC=T-%p!m$rGLAO8cN)sGn(c;B2*9QoWcd=gb7|F?#s;F#W+Req}AD0vyJdd9_PuINXvuj}xziTd;r(*VNb_UC-qm+>L&1x85oVA22AV9v>^2=O zr2lI`-gcr6bTJbt<~ub*yw%Hd?n zj`QOhJff)a6i!V}w8#9K8>?E{LEp51(vuc{%@i^&db5nE$2HLK?>EC$B>=w~H;R3_ zTN?QBxJlk@Jr_Sz9sM`_YG>J>AbJNQD^w8Vf!xTa+b(+m?lfr!J3S}M!>RR6T7pP? zCr{$C+(P|mQCyAJ!J*|p?m{Z0MHc1GjzW8 zxuZLeU~V^b-s^_iGH&SNL)Zh|F_)c@p#KtOm+JOj^%4Yn3pR+~0H`*1-?9dA;6qSk zTS4BZ!=C{lcuoji0V%S$Dr;xYKq<3-d{<-j=Hy2~!CYq&LF=EGbL>6Rfl9c7sU( zZpjN-JPBQ_krIdYzr}26jS|*`d9yo_?dI@TnNDB@8UXACSSy8AeR&HNkrLYk9?0$^ z3LA(Dqs`@%h?ja_#k&kY?knrTu23Ptlq3leGjCPkC$iyC(Dmd|pAyajX(ePro}cb# zBIv?`QlkNw2lmc469u~o`{du*6cksg|Js1r&tPMzv5dCQzR*gqFL@yi(g z-4EZ#7)Kn|S)vbL0;{dd&szn;4b(+=a zs!Fs$8#xX)q0cg(gKg!SuuY5-;yLMCa|lh~r~Ji0x=%K?Pc^^B$zt*y`_)}Zl{umuv+7$hqA5I?YR_TF}jxrHWR zA&P3cxmPq;xVu};&G2QjC-lt&D+ZHcec9d!~^1|k!gG{&YhDNrk?VoKtQ#rP;_x$bHa&R+58Fp5fkTQ%3 zOt@dnCBroecHA(f`%J zEC*I4Dt-I1Q%-zrg@XV2YRnk&LQ3_92xvTvk8zTzcAedQDU_@_7+UV~i^gHjNv7EZ zv=5&RSxYJZqJ3n~G6KnNpcy%6E6x0dUB!RQ-QA`g=_k?I^C1>1(7DRd$@*i;UK~U2j%UB{6_Vxa>kPup*v%*Hh}+%uYne! zQP!m4BGne(o)XJ_#nN$h-`jmmR>{Oxe}1*reOXBwN)CBnybL3jal3lXJZZ(-PBCew zRW-Wn@ulx;Ls^E)8DRs{g&P3Uu$>(3gn7k)-(F^<7ecP}M0EjCE<9a?e$Q5UnO|#z zHOkq?z5n!Um%$j-r!a3mWGNh}!)LC;E`knhrEZg<`>j+ppS9>v*Xn2d5Ur!04l;2X zf#~g+csn`zAx(K_M#tQ%B&pZ~>B%2cp?7w>e*K=dG4wO8u0N^zcKRUqp zT9GXa_5)Za;E}GI0U7%I1v$zsXVYR>N0Sl_^_+skOi9a=%=MNDk*xE*wsUuR$(wnFX4iT8h50Rfsx0mcb_YP3N{E+Kq;G|M zOW|w6*x1xF{4_TNS$|>rb>HxQ6ct~aCz~xovgVP&sa79+4I*EtM9wiH2vRg>&46xD zizhvD8PjK9?yZ{qZ8dP{`by`S8Z~d&4=Yn?-qtb(>mIQvdq>>UgR&75&I4j7(~&D( zrfi1tQrloFRe{)KZM|B(5fPl6FV6g+<(tQ+S(VKi#D4I?hz4GuMy8*Cz-(Bpf;js7 zl9=K0V-Rre$diIuUA9i;e+8HYyA^Bil+~hYkh4uo&Pl{e(H3JpEIsja2@1LtXxPf4 zi&gO^b7=3L?=djcE?B5|ae5(LVpeeyVsbSSf|bh@QKaN!L-!R^bFXiDE9P@I=xCcr z_E0O83p)EE=Y{M6Gr@}Qq@m@&vbjE@2)~#&V%5rBjK|W)cjE+c?7K*hE%DRE_~VAs zP^9+_30mH{7~QomKHu}pW`A&`OZD@|1>)gAm*A-*x;D$M4Y%k1$=z-2k$x0-gCoQ- zSUqhD3ZsFyKj7tlGDr?bO7k3RTVT$GHwJOP!z)+U=1jM7fn?NW{9AtSGKx0(rMnO- z-XzL}$#nF3VqV{SDhoMFI%dX1kXuL{2F=X>wo ziYUrc?N-q@n|-Kx;EtfytO7r8Unx#vb8-m$2++Eyq6c0jgFM)g-RUa}$9@#&r<<4s z#`j{$H;TXJY9&239x3MQbXvn|I!6ZiO@bAC4NczrUbl)E^`5JwdKrzIQ1e{%&9K*3 zTD5)GtEtFyott|P%9-_`o)qJ_y(&vmKu*pw$o`fctbErv>%J3C`lmP zqoZCAU;MJ*8065-lDc|RgfJpn%;V&rlItgrkjo#Cuhpmm8{8`&13H9hp@*=iIr`{P zBBwB#BPSVWU^m3PFH+Urd~@ryl7OfKBrb5MVwvL(o${Sbod#nsOBd#}x2 zACN|(9wYIht31Pg^Apw1UZeB+8VV+Er9KVigD-Cl9SNL1g2PK!7Dsg8!HAo>K>V9*Ld<8|LT>UBpK+mgr_G+>;=ZQd|f zuAwG~PM)FZ3ye(S8=Iumwm)jsqRnORi5intwU{p-Ypv+5N~Mw#N!(90Av>G9E}|_S zY~G$PAwYlq9f3#MP%+Fpu4vr5DOj|GUqxM{CKOeSaa3@vBta=+o-Cl#*So=dA;wwL^R>cX6bQ znXY0+@(b`{S4SvC5S(7pbJ1|6gY?JXMX#gIPfA|%utQfq-q9?tCUK3dCZA)k0k!MS zXjnTPU!0y1oVT@nk~}QZ&#{GTuF=cIh0sir=XnZ-js`ee*vG6piMLsXKXu#zcVMSx z!5%pLojrBW{mzJ>LS8>t^xK%ME=XM)pNt@v+xwl_z^1VP&Zer?d!q`~mw449Zw#w{ zOCJp9{#|!=vA!1iU!FuJS>I8P5ata46~Rk>tHDlb>8~S0&0c?%l9E>PVF*?*n^2vc z2*%%xAXfAy&V|D7#gf{KNk%JR?I{@b>k#zYNpQYX_L?48sNYxe>J|D~bFP`3Cs{?6 zSp*kD2_*H@(aIb?U*7i{zY&BLymQcoLN!y5RVjZ@V~uZDr?I)X4l6ibvD?kkKvG*1pYw zpeK_Pr?;ShC0iVG+AwZ4%03xD*G7~4Mv_19(;$I=G|KldE@t?61%~-4hI&@bN^Wn?}|Ox02URqqGFN?cuLd3jyMAs7@ad0ur{L zy(s)JV^xu9rQTTQo*OgUxIX0555At@~j|k_J=>zx;HDeGv6jo#HwZA7g6>%JrkZp3Me_Zua@DYBmyO2}}O=ade*n#n& zsD>Yfl1RwR0w69)tYO_p0^t8AxtX2Cx6Gvfou_!pqF#c;pGo=S+W$B2w~d!0!|-D+ z4Q=rhMJ;|VLh;=8igv;_WQSO7(Pe?&0gO-T73O^~0uW|4?yYZesMp_~`;4eD&6CSl z@>@mB3|fuiUwoQ%ov0Pg?)9>17EFx4kNdtg?41KUk9D~J-xS&PcIdklHU)a;4ls?= zDE;#U7*CP}u!&iMNt**l7|7CT`YfvSW)GFP)I$b%g;32PbRicC8AMA}_Wo~9^xTBl z{39p6CZMi2X$D_&D*O*fKbv`TIzcpi-iXnuNeD^gDRz)4L*i?wyU7@()L|I*?kJtt z@GmE1O<-FqvzaldfhjHpK>i02s20lSM(y7&6jB;i{w`|?_A|7rb)%%qZMA2J=+%Hb z@tK*vJ84jU=gb^D4<+8VJrAcHtVBk47@Es{+*BIs@-BY>6z?kZ^@z^zbX|M*HIt|YeMa~PegppUk+utogO?UIk*4tJqKj<}X4$HPqCKvW#M@;P249@rC8*A< zeKTruEsQbd=*f2+Q#wIG4_h>a`rrc&kO5owbKfRRWIjWSatzAO>Wn7Uq;kX_xBVLf ziS2lnSPY6z^=JQCI4QPjEoX+|-&YkITK2s^=*NL#GZl`BzC|j;42yE46Comyn4&Xz zwrk@~o4TdrcPNNfx+n2d3r|$yr(m+`fQw~Txo;9I>&MC40kt-ZuYI;*6RbKj9(yws zSN0ixAar#i(EkfbTA~yFn;2X1MWt^q(w{o~a3O&+gvu%SR67{hG@yBzPd<@eMQM#! zw0zHQ!l*Csi%~1v1A@sqvI**nUJ66WdH)Csm>CbV8jT7^g<$D1Ad45$LP~A#_jT`i z${SuRGP*|eh0W^L5mZYZ;<}3XfVNBGv+6b49Vz(K?M0>fyM6Vq=39PgI!YbVp_&7) z&NTt&k_Tu#t)Fe2-q2|~Uvkpt)j>mk8LuT!HK{==4%mWQU2o;j%Iv*n&ikCScVEJ7cN zg^A8GNn0*t$$9~S(^L4JW4O~;iHU4B?84Rn4B6!m1vF|3@1!`C?iz$!6Yp2UhN*Nd zP4+M(Y=Y8qRgNyNuh$-v0*EUKMY;LvcCyg-8H;aTf_cnSTz!G0E0(rg#;+_d+X=lb zbnRTv7?SzW2F-p~V_~!bQ%OfT^Wp8nL8nCQ>srgGzJ}{(^MaZEw4D`_^DY)on1OW+ zinjH0<5|QS9!v|tsvQxcqZ{HX@ ze?jwaY`U!5AKl-fspy3Y?7#Hq0_?bo3ZCE*G|*W9XG%3(R1B}k=YMU&onxTJp){pk z@WD7AYtAY*+SFPSBVxsGY75jCoDlAYg&=eZPOgl(1m+FM)CWS?Q0Xr;qUA_TovEsF zw=eiN?SJ(NLZkTNEr^1{=56f16`J+uiJzhuN2MY5et(GJej^*;DO8=0-c8?>7_URXpVPOvzq3%HW#;`nkcpxtV!hW-n3q(n*=vmWvlvOW-sBH7qorF=j z&jC?9T9nAf>d*0{Kw}lOzIC!n-E|CP6i94f%=hk~hJ~5^_!sws4t(#{p_#Z>=yFG; z9MFOyJ%3`x%35_lH3GK5(U{qaQs9?>FL+-#>?i&bRr4=x&}T46SH@#_)_d|Jjv2L- zJFAPGv6pxEZwU*`?BKe+gkcfxr3W~nncK$ko%;OmI><9}u>g8U<K%&3Buy$NM|vG9PJ)t8NB z<5`{ziuNyv@<;$Ed29^#-shtu=LBJY!-4NPPe1vV_qUOzU-d^KeJPJm_B)#J;a5rE z(~+tri@wb(sOD1ilD=a>QSS2@tOED~>AN!0>G>FUP<{(gqg?n=oOH_R*5jVGdq?IX zVXp&B;mikE#NgW>7Mr>gj#~GT7eo&3|H~aMHiZl$y7I2NlCFo_C`}?$8??V6E%{-( z0hyG1SEZ`a$QOXhB4bm#Sv>gOiMdJWqhEPd)?A$x*zH>Ne$Q~+;5Hz@?fk84@xPI@ zYQ}%QNQHZA1Fki5O`TPg5f*?LhYRun0>B$iuwjD1T%2CdEh{H&@+x|XnOHSs@Fqq%*~oY=O@{-FEG&%HjjG7~_l_Dwt)4@8-E zKEU89@*EcD38P(e=k(pY_^5H)xe{)a-*H5KnDS#(^81wu*1w@6{q2=;Yd8Ndz8WD| zMzj?pXZFeea5Z4W{WgS#SUA3)7}ql*=)a1+jA4&n;3ryveF|J~hW#fw=lUXof~y-g zjQi%=MM#E4%^%%LKEKNo?iT@;Nj@F>ucxh5V9mSkd4Q>jJ(6VP<3AE)P7{zIN$MJX zF>-_%gb}5B3nVjTQ_Z@SgPv(Gi%$4!KBcTdUIWVifE>Zt?!wOfUnay&n3Cq2lG)~a zo@GnYVi-uKo+pFR+bC4X)>qy;ec4JrvGCY5(NQeJTnCaoskft5kv{A1ny2S|GgTc< z!eqn&Ou5+l0-YS1`~<;uA~3d1!H5!DFv7v7T#q2Pmloi@Og6B9WZ3s;`)g*n!4A|c zWJZBD?Y3n?o0K`KU63y2D%E%q94Dbab0%0z&pQzM;$14h1dwFZRJU5wfF`x)e+06X zF3a*jEb;5J>lgU-7r<_kCq(7#nnRqp0lqAp){_^nhyQJSMq|ja5FPS<8v?F&E=u70Ig>$fk@rJ;gH&zcBA!jbB>X<=F3nJD8;F zA_iKbN97KCp5_F>C*`NZuQ=7w6ABuC7;HLNi8K_wMu;m~4?cb3XG`Uw``;<_M$6~^ zqT$MEeMIT{xX(uQCKYf~QH}Zdj#_&Qf{NvOT7Rj|ex?NlcKmpCw>*STqV@z~nF zz2<$@^0n4XRgs=TI>Bh^?eQDmezMiu@gsz7i z%1g7)4TwzN?XjL(rz$-$NS=G}*#mLH&*o-3mQk_2S9)Iu zn#(Gp?cj~NZ!5M_=Kwp$0EupcQ((gy>c9^S2*iVAN*f55h;G6JR`&7}T3i@X~ub<@5kr zbR}$xl@)YBsokMGxH9TR?LT8Q%wX+Uo3rU6!$Sp(w-zOOIQsn+HxFb*gJ(U8(Lv9` z(%s@QB9-oEX75>nl#}N)vg`Mo0pGMe z=*%IufUUesO8@Q$aZb{edH6~)M|iWHE$*Smf9-rf1-BkB{X8-40f{Y>4XL$8h?j6b z&RZ8WJ-fyd(YV-~L0`tmtgTU*$n7bp%I^69=I}d-yVJM;i4dxy6(zh~a$pLi;dr+Q z1);Bj(ISaQWV!@hvG<@QXDcwQKHVy7-uBaIBx-`j-)j;_aL^PRAV^6ze>BdMd)=~R zYb{m6NK6^?q*g(gX+@?Olwmit&5zymH%W+59!oB*8)i0^1KVw1NPbyY!x6&5{&LU7 z{^|9D(LDFnIpb_xW`gQKEH!`1JV`iD)bZ;Nn#nQ8A&LC;p3m<^JPKZ135fcBNn5lu z)AYwEXQYk4;>Av{YNQ_;Ho#yL2mnw0S)zS`xX!7f^8l?;Q%_S-xR-pv!eb)=~hJrB9$dZs-q+?ex_C@fm=h3@|fHx={qAjxN z{ESlt;+ds1RGJ-BF6J=>0Mb@KTcK(iQBf(#lXT*IE-k=vNug+;w-U`&-lT=-9&JmH z^fU3}E#3t-nImaAB#H?Cl66xN2-)pCNESyKm`ho$VY+vL9|XEtMg`OM2YUOl1C{Yl zq-v&rnwV%PKmffsnJ;k{p&JU<9~*MI7!B;n)v+sT;QXBYABWqt#`*D?AH(GNzck(K zFGfCE84X?SV61)J4U1H^Q;rHCLVs)SF|$D}rKze=qV$C3c9m>IN`D+ z@2cT2N?d__7lzvc^(s2o>~Z&GZe(h_1S4;ThAHh4HkZ9O=o#wKiuRPXy4{Cp6gW|8 z66dc9R!Jp_|F0{_sv>(PF6t+l3Az5(h{=#Hk`g2EZV-)OYif3oF{B&qb55WE15$;> zQ(nAqd^HO6VOqpxzUZ5pu%LY~#nIKs$Nbhl{<$Z4QdKg7Tnm^2wD$*Q-XAl0)lHvC zNmv+LMJZ3?kyfOYOdIDOp5|gH%EN8nT)yfMeVe^6l% z83gXfL_L;F*`&=H(?)JBcx8aN-(2wwG8lhV%+R<>yuI~LaPec5QNe7z|L472j=z{Y zVyWZhzc9BX7H*`2jETSOwWYl26ku4RJa#>~w}YCnMOftogpc{7?m4ewJ88GiG@fBE29~$%WX}z?P1+1o$bRwckq~vlTvCR$ z3>Wgu9^DsoL;pXty_0a_W+A~>ZP8C=T!Ixf(;n%kThY1XF`VI;xl%dVeu}ucTI`%J z&(1}@6XTP|Y37iva-8XJ!1>3U1e*P-STlbzi`F;gRApi=PTkR}WinwUcqI>(o9ZI{ z3Rjr3hg7@;xCpytEiAM&UvyFw5zF`%&eX=dv*I6y3n|L(9@m{;sY=PXS@!BwCn%fS zQp(KNeusCf%PVQrXgpuU$}?7#YBzfVni|_*aOb{)?F7n}#%KMQ;L@rG>jqiRwCvlU zUif?sWivK9{wz$l!xN$1C}#AnloL@Wc^pGdqKg}TXRd#4+hZl*dyk?rGm#}65I^zH zXWB{8WRm6SWx_6SiZ;WxbZh>^?H4jerS0B4xHS4T{cd3RD5;yd z+jejn>;iNF?}VS&1qGka3LiTC@H_v#944APXO~y}*?0Ha#j_p9@DM-Qq>Xs@i@8vYgRUIK+$1+I7ZBq-KM+a9YGM|Vm9|x)s!%{(44SbrDON#vf`@*Tc@(8 zqRy6}1F!N?&Q3G^R1Lh{N|`6waZqL+z>CEyI2V0dpY^HyW9+>>J%WuR!U0}p2XfWX z)c@VlM^-qxp_UrB1}e(c|1S?2ON70OQW)lUj5EQpR6qn`ZOIzG!#;}RXwy1l4p%5# zc}g|#YKM3cDq@yk^mBHwcsA_$GsMI<7sKVL3wN8$dcaki0c}ugfrJ#70)|tX?PA_A zKS@7ME(Ki{3#u^nlhIH=M>ztEQX?Ghjqf52JyCuI@mbb;T5e8!$}ZIb?W2NA;d@fw zyX@syB8m*No8G(U+za4KRbe>ruVs;_{bZcHxQFjTGJ^nO>C3ck|58ETNc)hyeuBc& zm&ASSRKWrHLD^q^oU(}GeQmBJzD%3m*xflF98h^rK2w~Reb2A4PCusDT&?%6nSj{C zu>TMsnG9%CEcxi1O=191kO|4ox%F_{_*|`d6~A_3w!ugjQukA!S_SDHTNl(+dU6x8 z#i61_@B%x}*2Vi(!sBikiJud#?SaGD!3x!@?fhq}5P7NVOztU!lx}{G z*!RxBZ^dpDfNE2ksaRBeV7_t_aeoE;!%g};?|1id>Ru)sL1KX%38S+yLAB;{HnpDH zcelyC^u56Y7^c?RmpVm&Ve%p$jD-G!VFuvZRgK8Ee65gqXDV$JPVxpDj3>i-T`Y0` zkF^%|f6WC30i7UfgM|!n7YFj?U(i@m$L5q z)!B7r`vzyC^KO*(JO4L9TpCRg+iKQ&;(-Ef0JSXeD^1 zJPpXQ{TxQ({vXrH0R{9^*n}#w~j;n&md^@>dm@#Ygo|LF0L2{-TpNXDNBZ}@4tVTboUBYmXM${niUoT^ z`NAtT!fUhGXud8v=_TrNbH(*^k5&QdrhXhxIuRGU z{!NBw*Toi`z zj|;_7!2DbOfn9$zRoq(i7_2K>*zMd}s-1B1L`(Nw$K*+Hhg{Y4VR5L*6(^PAUU{Uq z$UfVO&4i3B0dx$bWU2XUW4m!JE0 zs*jH>?UDW&2)F=;cHe7UUJf1v*Rfnb{IvGgj{B~~hGs$wed5{+L+7VOq~xHXbc1D>UMY&wC}mB z_?}hQPpj zqs33^-X0`iXTkxscKZr~eKj5?ojw8Tg#`^prj{O=#uC67&QZ-MtpjEZN+&kZQj-4)+(KnVOs5 zO38t9DI{{eEY^#1U_ELX;O3G_aAM&B|=-+gdDzour~8yC0kZX zL`hy8hjvHonHj{cmbgo`2W0r+o?U}Ro8C&bGXqx_qhl0Un3cV6_FaMf@31N!tc={> zie@EE&b+)ykV?0v8fX0AO)Ayd9>X;zfAfCD0*`{r4}n$zGgL&D&liPZ=MwvOjA(*Q zRzRrLs`yctQHdTX=-Ihs9(ns)>O7r+KKVkK@vlH6_kSS}y?E!33vpV)uIirv)XZ(b zMRW&oYMaC)J*ETsEbfmYY8mQyF5FCcB1Rd8=8t2qBYzUeb4J|jOR1ra(Fi34HfwuB zF`GXuT4r}z&#n{nM-MDsbNul~{pJxbKY2ZK&AlnlMn(w2736xAr}$y8+Pv9*0A*b- zapiirEjHojH2tk+q%lKCBuhwF&_(S1@Y^M+rHO(4wQpPVOW0<8GML@Nj0b zeDxDjB*My@;E(-+j@YOYRdujJf~eE4>-yM%wwPb5C~f6{>WUo)LeBr6d>}82?h*@9 zJ952@M-fGqKH6p_2*XKGZX!-TOi8^|Bpf*X@29VEM4GfeHxBX(-4AyTBn%B%bjR^_ z#Cc&+a%d~}sGn{>mFwW`{w99J!?e+#9ig9xz?R*?n&k{hj&4lS=W0+;z#J@%~u#g(3;7O_C#mB~?gu-j>tUFvXrC+ml zeQ_MsY9WDPWE1Ub?q>My5{1z*$<&xTXZ`@+327(<(mkLqA`#~Q()eGz)53i--AJ>9 zScnnG_6fQOb56Q^LMWsCzbJdluqdOpYgiFPPys5`Nd1Oe&p0qIa; z=o~r)q@+Va+5v`;92vT6$eDrn;(osG-=pvS9tZs4-)8S?oolVLJn^`W^d95oLpb~L zUHzp<*Y}d6o}}U0RMifx75Z2ts&$?kkg}Z~&H6EOGA8UO=P)m;U@<>6#?aO>V>Z%lZr_HwDPR!Juh&o_jy2!92kt5rMiF9hZ$UJuoip7X^) zmM0p_Wi=pbIYUo_LpKSJz{J3i=b&b+@%Gvp9~QGMANrb-v%f6%HBv2%_3CE@OTnI@ z3}I|_A(S3st{M)C7E`)Gr1)|jR)0&Q(%~gnl>FeP`r?#OQqrVsa;>1mK&e%`T-fCl zV=`<|?pDz-`KD&tu*J$otD1dVLJ_j_>dBACXt%h>;x_H#mc5k!-&t|~bC;~Yosv12 zC8l*lghDj`*09C!NlZe*B9qMe`JQURxlF(OQ<-3ey&q~;k9)@%IekF#%qnLEVWg0j zhXVbe#6UsK)ta{yax_NWedZ0vb7_q`oocPX~ zB^gMup>+Q}s9g-*z1~WAu`?EmfUlT_AMz{HhlT`f*G1I2+8xQqv)WY6w+f@$Y27Zr zwRmk9F`c9Q6-ZQGuZ-@74WFhrf9P;->Ur#q7xwHh!B@vHlV|g4^POhcQ1oYc43?M; zE%-V{Oq=%RKwD|^->Eh*^3e%Lkol+KFx|lpL3sVscivbk>NgNSp``PnUmuw9O7bfW zzor;-42m<|*n1P_)ax*G|Ebo`AEKxuu}2AZv#n-~Ua5-9%<1fpZWfb1k0kwyU0)Nz zuUeKG7>YWKyRgUIz2CdkJ1gBmv^YWHtjLtVFfm%2?+A4Teq}%Kjs-1%k0pC|M7+pp zChR@>x{>+!V9^gPZ_$~Wyd`qoo<(i*zpzsAV+&F4rs=x9v1a8XYg;i=7;XI}S^?8~ z%Kl%UQvin}m_|Iz^EFoqq?UQi1qm|$43yuEO^PM%C9=e?n5SnWB*3(b|J z{;K&xd-up=ihG^zRn^AY)Mnf(;!u@oRVGXYzCMH>) zyow+#Oi##oK7HOZ2#F!b%e@~VF4x6ah5udmS5;id=Yh7R1V3uipLS&_H!MuZ5j3iQ znT`j1xw}tplDqx#PB@<)aaX*m%sDcy8F2+oMx5 z^z<@rAsp0499_j-QpP0_6&G&BP zqGo5rst!kMg%IwjJYhW+ak*!M|q)=@Fm!;hcTreUit%{6@{|p!Cb8z_M_<-Ug!afg}A`MQovmH>9eI7 zh=m46vJ3Uk`@3Xi#w{`$U&niiTAXG8#yd-{I zaH^_(wECh@>pKM+_V#(Ujg3aR5QM9%Y{l@s#~l~>=X2s-?|O5FZ^3?cr^!G8jLjY< zr?A|L?}1Tc(y5@aXJWm>PTM)LRsd;xgu*nL{(3|X)^`KG zf(<=WnjA+|#`>#dSG+N}a1{&j%ABFK9yY}TAc{a#1rxHgwtnzJ4LtI+=LBq}LG4VY z1VT+LL8!NnTghuU^j`qc6r-hIkrkzAJ4#9)45jv4nLD+*YA9zyPe(5Y@6XoEKO=L| zOO+~Bvw*gR7F8pqtS5d^)O%goT9#zBYKsfq-dyQwi)UM;5+YNw(5!B*lbhMp7kcpq zLyVhY+#lUSVsc(-&ol9mSfZLp_)Vuw_9{BMGa{*xP(mU5y)^N|2S z7_<w@5hJBuQJ)x^%S43gpBKm zO7jQq5B=S>lnUE;9J-R;(*w6RTJ{Qu@`jy-s?3jrX%YBd|UBzVI)? zW4nC(=E`(G=afj>$NlRXC@Rdx*E1^%ES_9Fr) zZhOv{GzxL4w4j#k@1f!48bHi^2+l+>}StMi*xOUF3n=O z0OoF$7Z$gD&m2MR@aDuWg~p>;5huY<+si*54*GgWS%^1dkpZ5( z++BU*=??-{?h0CnTk=6RgU+OpYAo4`l9~%-)Kx>4spCHpbx;#p-0j^iMr*4B`P+V& zF?r1oe6P1k6Lp!^*{A%)S!?30Q3o#J_-2ZO0i_L#zaQhcf3po#wv&pirX*?C#A4{g3dygTdxxgj}NriamHeU5T@=1`VZv4mQ6? zxk#^X?tTouk0$q;CLwos1J|7)V(!zZSeT%69RLstpb{!KyS`dq7h!?FUu_sc3NQ4S z```q(+n<4%n{zfJR$&IArfJTBjl9ZoSa5lf0IVvO$FUHvfm>^?!YhFALhOomKHQ^| z{l1c1=yyI7S37s=n8`jxl?jXdu2ZAcX0|*!<>33AT~|CJMcjY!cK!6Pk~tA^zMJG= zS`~YE-rkIiZ1tat<+V3vZns;>Pj~hsDIs-9VV8n__LYX2+sXZXS}Pi9IA5xpEp%6o zyeVzni{6qsDE&};=+ST4-H7$V)OXb{NT=*LAI18$V+1S^p~X&1r7S>nsjCkq`+Iaj zhX$Oiy5xM#cTnr-1IdlB<1t|(@6^IfEF^IBV>Q@ON$^W>qkQ8~Q6va6+j^1$_Y;|X zE%BzY)qMj8Z4@RI>xUN+WKD!ruz(V&V7_F3{dm74E0p&KL{v!F+pc1&hut z)1n7pJ25Qe)~q?r$}0i8&{`;AF)pd)6nBsL!(8pxPEuO7XV&scZ( zuuQugUtz#{$sl{^1iVxQMWnNGwIL3+NPOO z9|nKq4O^yS*t%LGPaRZ5oTqGTtx0i=Y)PwuJOwU#EB}En#-tBMXD{EX8X7Qtq22uPIC%oA;0Q| z97R91AmMu2=HqCk?|nS7%oj%{fV>VSptQNB?)MQdRWtt@K6imO=<~VRmq-MeObIO8 zDpie-;c)0}-H;{Ff`H1Aww7rg{2-I~z@Ou;$Ru;Ng{Z-IAB!s{H{_=wrR>_GkXmTL z7%>wunSfZFAQnS9nsMf+HYab|#2?XL#g4;)0rn^IhC64u#2JoS2Dx0l)t3Sn*@gvA z*nM<xLt({N~&`%v_Mai&S zT)^1Tldq$wwO?J!%l6k941dYwTu02*8<#lUeD9brmlj0pq6XVG9a$5)I{+^tKeD!FZu#?>_=}rfq z8l|+YjNBC+8@u(o;hwbu|BoHYM#n9^D{=VEbielt-olI2OFeR2Gg^f8p%2r2u-L1^ zgA^HKzs&x&`|)c;+6sW5$Ki1I>Fa+)%bPeHn});o2RlTO{`3RBp&7D&$ZK1QL?0Zp zynbqdo1Ju`Y@{TQk<#KsLCNksOJ^Nrz}H)KaobP(MHrk}eYdlc7NU)3LTcTgf#=Fn zl@_Q)+d`Vm*pgsDGHWxYHW=vqH70;IgZhUZHN>~IOTJL*-Bw?~DPVUs8O)Cb9CN@v z+GNx4c%QcVA2SC#9i*9UG`y_8qDpwmmAGm&g(u0&V;JXa+h6h5wNvxX`o0vpSKk7y zam}v%kX_B1mY(5SWqizSRRw->fw2C!vB_Lw1+8s%x59c{l2T&lFpWrY;fhxyjM1T8 zTVl7r!+Mdiew7+0%N6} z)W$2O8-P-A3d9*;`M|0H+5V#7{hYng^H(s=G{V8{!ae`^dyYTyChF)2FB2j$poTuN zP%3yT&>|3%6`?Esr4MPV(?X(4*Y}okncnwqFV~&oM|N`4s>B*aY3qa8YPlTk@-KPb z%jeV05tMwP^RE+p(qC*y*RGNAJyH}kBldl3{U>*qVz|J$@x~)Z_lj6y1E~J#3&MzC z8DntB|HDn*ya||eRa!t4QO|JWhUZn@@OrYTn3Qzht=rhZvJsch%-zF$_hyP*TIn2I zjYZqb>u#aD8jaMTt9$m>wNG-JJvVb6r`F`(h$~e9(dr{z2(hF*S^5de98V?pJ!F<7D?<<3geE+(M7JrP%?VUbdr2?6hw}LBWTJ<)fel_ zXje2C)xzF%!ZjB0^EuEIafh?bUGK;I%jmKrb*$K7*+L9!a@_s4w!elrW9T)NJkz&r zclFnlbMDd7Rmu8CAZ(f%)VSGy{h~ZbH-tVPD$*c5-x&OPe@2?I#Al<&Ez9R9T?F9m z=ktwE3BIZVyxmREQ2Oskm#q{65~`r?0JThjLq5 zDq^t}-OQB3Y2_U3pgUA2r9dxXHERk4o){sMvq*#lom{1Y_?YEnX;yuh$I?U%a#9Ga*Jo~i&eDVCV>J4HUORoh zwz~!zZQ2ZRl>EKkwzpS=tRNu$i0Cry0|^B~S@inGoI#*Oa8wP|h8T6C2rB>RY#Vtj z(z6lT_M*c9=2tP1_p#G~m=w$;e>%*aE0il9sCopr1136I6wgEktz4i`Z z>6K!MoWOsp2nGJ`5pttDYd+`_Y-fCsLsQ_8Om6|ol&wg~r(X^}*(tmCi1O~MO^sVl zOqXBJ@1fD!f5O?}(~(*;Ki(fBai-WV=s#M+s!?!KDBlu0>L@Cu+vEm2XkHb4!`bj@ z_9gr->pV%Z-;%Z7XL5Qea`Q)r>nKb4dTMcQt$tY7ZfxrMSHpBFUQMl=>rUQ)r;a}$Y(d*d5A`PMPji;z}bG|#f4?Z z6^@VpEUxG90ky0pp-u;t3mMrN?u=ThgxTA7L(g6PF>e;BzPc#Da$;-uP*Hev{cnHe zV+KsiwVpf;N=EYD(JMtN)=!BCa9r-5=D?KS|0XC|-+KdL-KbEC>>g1mUEp4MxSAx_jXlro&nO?;KAK?=D*p_*OUj~(5C9Ud-ZB) zV8BGgc;W-U0>U5*FAE~V(RUZ<7;uB#HWCU0t;7?lr@&4LYk6^{f|5&+ zan)<33z(dR8Jo4D_r)k2wo$9==4^*09$Br|ikk|=HDU5ZJsC}weWD@9p&9!rq83%_ zQ^2^+&+Bz$+4#!mBczgXgn`~jjBC_HC zM71u4)hyL#_&8iVPd#^#$7QHy|3~FAIw2<11MVN`+uN#Y%@wi4OC^E#g5Yhm#sufI z&vx|55!RJ$N=v#NE~CxZ-xYEUYz}eq;}bd>Un7=!{0^p9gRLBVnw-CCQ|>?K!fdst zPAH$0D@Hu~7y`W3^0~iuieLK$*^_sQ^IayJyUVaxx;s%dqhG2s-TNyi5Dqw4ECY5d zh#&zwM73x-;L{l}$AiVz$`mCXzr3?iw9(pUND=tZxQMyxFt&j2J$oa@30G=_ROP&J z&fN*=_6($D0~_&PE8FO!m6vPkuz&Z2X_~a`nx)q0i8R2Uis!49*;#E22Tbwo8J|lW zo@X4)x*Ujj?RKnnFCEUAHF0=ln%!@Y-iefYJ5mpuI48%vwPz$el0U)D5)Xrc53*xu zVnPi-)Qi0bmhyk+Ho(lmq?6O5ovBMOTlpQy8gKvhjC?r=q~T%wy>1c;4yIfP!2`XAllAo|V`TC}cz_x(hOSCDhfH)qW~g*K_8H z0^+(j3bZXwlCX=F4Ch4cNqeaUcYNN~loiSUEsG{|J{Zqk=8C&uWoD~_lCckepih^>pQBAB9>RLw4Y*PXS%d-(1T~eHbTs#g!{-q?B9oDKIndxUGQV&(v<_}Jk4+@-MpfpyRg?aXsDJ3506>x> zlvy;6fR$shBVoVtu>(_jidCsG9G9yj2x7BmUW^=*fl2kjXPPAi)g-xV#yfjOQw9C4 z^I_6s3p??g<~qDdL32Ud?ERyhV}8=DPK}4wxPfI1rd;Wf*xDilmS5|vWIQVS-uAk( zjvg_1PtQ2tc^lv%Vy!VKM=pQZOOGROvS#`&%c*e(iXHm%C5D{ zl4~ylOLH2Bf^WV0y%NuUXHkL6pOCTC zS9Y|WC&=7NElh4sySr<}wZFR)C>r>Z;!}(>vUk$IxPw0oYyx~({ia_flg&O*Zedlb zvNx%BY6dStyswS^N{d>Oce-AXy`mfsz|uO$WZ}OJ%({*f=m?3bcBp~xy6zh2AqAy* zB3QSt^LKHFW{LI**b`RdTv%yBip6Slt?%KV7tE$7KnI995jI`XHZ!5liTL#8d=c$h zIQJN2>+*{fx%_&Vl2A-HMZ5(VoBVG0kvm@s<6k6zGxch5Z-R~N)(iM_u-;_wAJ2l0>+H?U< z+eWjSn}h-iQQlCK?;3bkb)0hl1w4E=`$(E$jCA?a*HUqw>N#{`?jip^zTky1)!Bm7 z`A;o>L(;W3t0aR>3%injYV$}TrN*WD6=^vd zR~cI>e5=`x;8X_z1(ZBHYaTpK4F}w%#3!G6x%1C1evVZ?;t^-&z*IyRvaHt(5+%HL z)A_UyZ{%TR8WVG>^Q1~?>9S9jYmQm))D!aM21W&ZeyG8>>O=kUw`dxWg`IWna#s?6 zJblUwoVCA=#STi6E&l=HCv3pKn`bqRe%(haLzwNY&E``AbYcb zt497j^%AUWmb=l8GijqH5l`5CuuHzU(k-y_n=JNL%(&!3NxB@lOhr|$y^>qea)rGb z*VxK_v_UKqdr}j%TV%FvOyg}N89tmTe*lc-SK{EPq1(_N~sN}LvhvTLsfqMF+-Bs_g ztY9MW;9ESO+rm{7PF$bhdGHWfXi%z`tU;or*%fkH%7_x7l>5LG1@>~!Xc~{I+LENh zdq7SWjLIGASslVSr@V9drqEQs!w=O(Ys$kE?N^tOA8C4b?F2X4CcaPIjcLIU0!IyK z$o;>U76N?doi>hOLcVJba~?4kiy<*0V*aj`ZB-5Ho*Z`}H%j(`u{c9y>6^QIiu2{- zA!@55tIESBPDlEHKAxhx5tJ1qIZeRI@f{h(win^Sbx@9<`xQx_Ry z#tz*+9ijgd`}3RT47HInK_HJ2hf~_Em}l~5)=drutw-FaytN85ii6XON92bhD5Q_s z>3cKXXRTu$ZVHS~Cx27-)Lmd^7#+mgzKu=x>}j2}9;Q8nD6pb+in;~rqyWOBv-@!E ze_eP6=*R#3Y%$-MHa-354}zW%;r7hZKDPj=_=Ebm62%K0OwJwmjT^&RHhu(mk0N-2 zAW(hM@m+U`-1QT-&u>{?%@1fU^fedKOUswwMeXp(}dl=40JS&X>sYgFr-sioDZPePZ= zm3mo{VWxz5sZlWL6I$s_p4x$gZ}D)Z$&tZTF9GTmICAqMO}{LpwQMGSGGeTZP9)Y=Y)p-bWuQ7MsgL9G4wSk{VSkwSt% ztNRPqTghg!Cn4$!aH7$TV$*3arf1|{J1X1+IpZ4>Z|<~ffV%jZtq^bZfy~fYfaR+F znfOlqwe#U=FY|M?cX<3b_3nj~bG9GiOR~oK=gv$smCRTvR*gA)1(4j9rkRTk)fD;d ztfO_CN+82ov~6&cxV`iR5Ksl&yM&GYnt=xN$#3SHfHon*Vq;NvX-de1x|4@Pc%kxj z8=`y+k{q{4ImI&L%pD9#DRa%Z%vGL@6#DI`U%Ua91GEpQPPAu_IT(cL5{c{cm@=P1 zS}s1jvVWUJ@;+; zscRmc0!QyKy8Vw`nOHyD=|GF<`Q5p6pw6t$2$-~zKk)u_@&<=Hj}E5RDG*uNxOBy2 zPf2a{OEr7wmG27wsguE2>%q3Q^htHiRug0Nb6K8iQ=~N||1O&BhDJ8)Fw+r(jCttV zZw2z$1S2#yfORKv3w#e!KixQDc(4(413s3;anHtYUBEThw&?B@928e^r~MIWOi(=4 zbbVKpOy*MbU69J503P*9=3p;Nq6+oUf$JPI`g=&aZjZ6Ns-pd_t<$S$)6@E>yn|1P z^pcHV zfoVI-ClGmH>OlfCPp@(#%y5V|7q-fIBFASp=Gy+4N= zAxEmjE28u!v27)v#!BpvuENXfzru8&irm0N_vXZ7!DO`zJD8R@kHLsJ0_M+nE}zif zX7TPUhxOWyUA9}w|5wC$w?)u#W(5LyH0xsU1Ccufkxaqnlrkni0#D)m)ZMzTF^gzl zQkLg(nvjM^iR0=x&^M$H6Db5(igEh4^X>{fH)KxZxrD; z5>9z9F6XjisqQp)hw!Fw+_9NHFWKMP*nc!y6d_Iy-gIBf6QRoN?_0Eode*sa(b8O? zJTvIL+V^mA3cdKCA>Dy5{nw9T!N1Zi1F7HDh_Q%a9#3v;9}ko2uIWZ~Hk}P0x06p( z=4>~shA7o3UZe3I&D!6_NBjqk>*H=LM@W;|ip6e=J|YEDL%$F7UMb!LN(HyDMz*;5QXx{eXA_y@2A-1#Oo+Q?87H9T2&Nm+pGeH1D!E5`@uUTd*Bo9kE zo8Msc?&>KV_p*T9-*6`Ma7si@x4zaZ_R(rS)11!#k6NpKA7jFqfcng%|9V4r)1ufy zTlJ0j1RHq^X%6O?%EL1v% z6#Y^sER|B2%$V?y1jCypFSsZ_Dh8U-Yi=GzoAh6(7Q4Wg7Z$=g$0j!t?_H3yj^sTj z9>q5?mt{>%{xNthOuKz6Y&aD`a`zQM zqLM`jq&DrrM+?@pxhB9n(N1muAi&J(w76QqjgRM$v$U~);=T!|GhC{d@)4hFKlF^1 zQtCFj83;s~By)#3)Y;%`l2d8nk)lpxB zScHV+rgcAchNN8=IMvigll>#@$MC6?G224KXqd#`CkQDW1EL)B*GFigPZ$p^2mI2~ zVhgBjNP(`5r0$iJe04&;u-c4adpUmr+bGj~6Q`nPxAm&q1fiLva?noB zF|bIPFeuTiDxAe{EtmH*e;yfo1H;>5+zH+Cd*tA-k)k)~V{0lkx9{F|GccaS;X!*K zbApIlevR^RqlJ2aNVCLwK9*ZnNv77WXPya23kHY3J;)8nzlV&}O1} zGJh`=G0qHV)U+164F11sRA7JF)0=lF#US3hf|X~dVcqpl8HnUxko)1?dBJoLyK2C3 zc_<%cA*%_Ic)~4^5i3PXEy29;ebb@z4W^{QChVTrp;l1CX#4ZfhZ%`nDai|ujx`>H zS1<6rNqwXodAC3nn==CebI;w{ffofj=%C3rUv=A9@xCAMIjQxLsc+1bj`tnTVLku0 zmk)T;E~P-3jFF1b*ysqjtK^_X*bnggLi5=)@DpXN=H~aZghkIb)`zqH8sj7 zTq~U`fLfc@O+Woo^Ve$cYpTCROGlz0 z!n~p(eQl1VESoUK^S!#6gHe`HCkvk7*SgqQC+J)3-HJoj3?h|v4jT$-6q7d9MFc^a zanI=G%s=vT*~h40B#W%9)l>*{(2im|)gEEGp^o>D)~Ac6@+vMioM2Lhay(;xEsb9DUcqA195S$ zobO{z6rNzdJ%B)J~&oNtvt zsyE?+)4@M^LHo0p`dAHGLjL3x&E&HpBTTp6@eE-R5RzCM=uAk0FC)vN(oId^4Y^I{}=7bBc>%V-sKA(U@wQi7s zZ}Tlu&u^-RD)!7vwh%Q*wp8BaZvr0^74rfPQ}hxNluv%Lc)ehq#Cd$vtaUzbMY76> zJRFqxUcc03DkmG2?OQU?*5^AJ-sY=!v3OWGQnZih#{=ZpsYIH!7wjWRvc@&=0IA#Jr{Nakx*-q=!+nQ5;=VwQ@d|Kem_%{Hs5?rFxiks zS%DPeR=f|)p!#B3K`Z<4oCgNFC;k$UCF^snVt-KP_G`!Hv`P)(ee^}DVQWtZk zaaE}CPu7AkmSobYUzLk`P26|_)0FRH4 z{JFSgg=fEvloW5DN{1Q$Z2Xj8`=nrca=nSG6ZQ!g;p=OBQB;}gvPmB{&o1X7{`UOq zUBRk6N`#z7A7|R09fVA7D?YYkalk!oTZzWsq&0P4vbTd8jSO5M9X*LzA2hI!aUx1u-g}UpqPlwbu z$ECqtCK=Da6OUECr_6^YCcn*3Ks{4vf4LxsPbz=GB`=tnEE>+oWwXf%YbD&p-I%Zb z|7QU-@FGIH%?$G3%otv_@=~T4trf339Ve{GwG#IBZk@z>atBQ(K=P!<52Zx57w6wbLShV97cHez zmFFk(v99f6!l;bdL;h z3`$eg3pBl%gfk|G!O0hshY4`sX^1-6sWiwLuwzgV`1-Vm`?BxGrX8f;Vx@mRCg(Mf zhb*xfDT1YW*4N%2%Ta9?p3N%p73^<21(DdeCRA5Sh1?ZT0+h@L2im8qe@_|$!tDkQ z$8ZMQGwW`|gK$s+W_%2am(b6z06I2%`}H2Vzx-GMJ{d%q@*T#hTmP6Pf zCZH86Q17!cQsr%Jbt2CZm;A6Ba2b+iFM!@SWp00((P@QOkWUi~SpJ5oPhcGgX6)3+eYS2%fi zF=>0h>_e-F-NLmn{ZzqEta9S#DOJtHl;CIJXyO^(tQ=6i$Bnp=^|z75|D$K#4c^h} zHW4EAC!UOGo@a88F(Qoh>cP;dYH!9^suCVpa(#G%Q)cr2g;{4M;qT+9XkAxQoi^l`?-FME}#jzL^RNk zRIa$vUaWtgu7qP8hKera=;TRd|KpUp7N>BhXQW6{u~ky(e{EFqaLXx z$EzJ%Mgj&nltvgh>h6lUd@+ISR*~}Ft5=H5REcXuFP{;K?Iu#;^}|eJea+^4 zHAc%`p|>x;k8MJ>sdKlV;VdH3uFIz3^)d1P5HY{aa7N!6YLwJ^zRKbfKb)kgm|3eA z&~3!9`5&ELaf1AglS%L3=I6D#oh$mLWo42T!5LNM^iXrP5f4nc^Z)qz1d_@iVG+hI z`lU5D4s>6rD=GR>FqHAHB)5>tmB3zCQE)P0YG1BrlKOLCXX4a-_cvq@o>)KnLBp9( zRM%2OnSf9*o~|LG zPn_#4Kw56UCw7g@1N&R}A(!PKR5WJ5#~u7%eca!Aa2Gt$ug;tIuo4UCKl$dodeb${ zK2fm!X(HX7Bebn<{o%raj1>6r5Y|)eeE5a39e1P7u=Da5fsJoDyr6c*1jXcA7{q&( z`{r*UW*G7eulo%sA70%81}S5ol9s1u|3D>Tm~1?*DJxz0g?&Tm=Ld{Izb>&$vuwWYs9rtk?ScU@Zed0Y}U)0jXi zW%B|iYF5&4T?j0qoD-Y;{#DGM-vvRId1ioOHjJsX{ZAn<8+ej8i{6iU+9t?D`DF3I z(sk*}KYtpGaL0FwpL0my6ub4Fi9C!Q*d9!oWY?Q)nw83|p*IND>Se7-Y$|o+u zqKn2K_DHhm$A7BEW8gnXOd>(VJd8seQgEJ5J9Z&wq;Y!As(Au`;F>!{pFggl7mCjm zc9{ACb&`HQ+Ms;76U*5m5>yR+anzcGKOysH=P@1BC<|Qm`V!?o7j539eXU;%3S$&n5_Mr`8-sc7kqpm zdOq;p^gv2=v!hb+b`kI4RhuXioK2~;y$Z&=E!Yeww&DgFZkI>oent48#Pc)XgKr$C zfktnC1`=gFr-xgvTtwJC{?roG<%JLH(fmgZ{Pi%Hj#rWrJ@bgmrS>%c1^+)IH$~i0 zeWgup& z^XqI~%~7G3qKyV_LWZ{6aTEH03$ceb)p z(e#0V^Hk`(3*V@=iadk1h&|mf|Ajh+yOf|bM?H?WbI;P2oX<+Wy}VkHMoQ+|*Bxff zi#0C;A7>4)GVJxrrX`iC;pr6gE9LP@yD)j2D+Q)QdC=>UJQxA>FV{nh&&sXPt#U0RPi`>d5$Nq)*0j;7$MP*X*U|8z15T&)>)@8Wyc*6Qa^ zgXn4S3tNlZv&7+C@&Y0JjK%)yd(6$^Le^VP^Pp|qF6+{hV{>di-c3_2fBbxW@`0`M zC%X$QCg8;*q_dr|>{qW`m5Q$?=XQtHI`dpnwF@!e6VYXDW8;f~yQGBVEGl03Vh}xI zni+>+1X*wNz>9>e@jac#MXJzG+s#&4A!}KFiZo@mmM0BaaC(|ox5`LLy= zT6>^2TT=FT!0+SO6suAmz9gwsSaH6h0!Sr#+VGo86K$b3J5S>nuFzSgBB&COG)&9_lNe728~m(%sD8os44hByMz6BKrNoJx6BC^=7c(wr3%L+TP`Zzw zT(j^8HMFWP<%%EzHj(2|$(Hiyk`}m(X54#ULA-sa4e6T#v7HrEtkJ0s=1``G%8{`Yv{4YINr*m~Bqv6wn>;lMXlb7Ky%s>5VTqfX9Lkg&Sp zFIswUZY;3Pa;(k%2~=z^)QRyuKNVdzjYp-&-F-^Ql33e4=4D!Zib9qd$=}fjVQXdRSobwA^w_WoG}2S{{U*1;c&6V4256vU6{1 zm{bbs$S1AjB|hV))wvMM$;#C5D-hUhHBeAB=$wD_oG8N8#ELT5g#8?UD=mgd;_Q|B z%X$4y4?eP{9`R>3>jvB16aP{V0N->3k;yX)vZ8k_btXJ_D4(&sJO!_A*!9z7(cW-B0=uWXvUbJK z3N%S6^ti}&C)PL8?r~tiYmK+hF>$dCjACxZDb0>t?EZL*luE|XC0W>Bt=Mu&tD2Vk zGX}g1KwGVzT@<_;H_{8MkvcJIzSPT^{~I_~A4g-P>jNja*8`K?i>>xnlhc_Br$6RC=EuqT2xv8NLzU#KW(&<9W;V_b3-oe zlMcoqh&>R3L5-eBCm6bLRfAPH-{a_NuYn|2>zanA2XC(c#=$R!fOoI$_w9|9N5YdU^;O8}ZJR;7d{CORI%;6sBfy)onB0 z_ggGSG5i+!JCjtX7Gozd8>S z+O9k41!0#ECUySva`e!+nl+yOI}Ig{qvQH_y_?+J?1f(4zXe8=;+u$0U7bDke|M6Z zmYS$*d{u`srhSgKR_ne(St4Hawj4-wF}+_EIrN+xgU@>{Z4Rs@o1havt5Nt=`_Z

    F&6%9&Ffbkno?rfbja-IGB6lyP3(z3KAIQBdA( zQPY21fryE5f;`@8C&P?v%!F&IGeFz?TSvO$7sd1tKna9yHlpfFE;C)p?f5TaOys3T zv4yvwWEI(J>}k0w>Qn@N#3!~83KjS4NejE4BZ)a z8IG`J>E(*`&7DvAe4HF}D-vxk0#v4ggt=~8MUgk}pGDJ`Hr0*5n(j7Bo~dQx@j)Wz z19cB2`JGDy34f12^`kHGm>B4a(2%wpAbSGf92 zO3u9Az5Br%G}q`J*KC9=0W|H40JS_G zb~Ye+R{8Sjt3B0uqqU>SMp}x;`TO?y{Va}|JGDvaIe2|8`VFw8K9Ld1K>1ACi)B*2 z%j`n_cRA4SrG59y{B}i~t+&TU?R)za)(ZN06N}DPU_PZ!W!yx?jnv27i1p^RVO+8D zp4%VC4Yv|Uwz}xrO9&#?7nShJE&`;{*bh#+ek^*zb&5N3C2A#)dkO`qwb53R(jWo@ z%IJ8=$t#AwkfCa(i>B9;aN81Q*5*FUUhD#tp(k{@hxg?JaPu97Xc*Q)>0B_F?rY7~MI$5wPNRa+scps7N_#Z zel91)IC!~tdFJ-YDBo4|6TiN;HuJKw?CmF-oEZ;IX+Tx;625H=9T$x|K8*$%juSCP z+gtH`Yh1}{3EutU;-cA>+O-`fJTYYrcY88qV3!+?&0Lr8p=NdA9HvV`#H{_~Q2$_0 zza7t9E}TOtn}hm?>U(%M0T=3s=;x5%IT?dN2dmdS-O7F09f+Tr z(K(uOHIv%K2dk*T!(EPa8O6M?EoQ5kW-L0bc_k7*3>|d#Fks2CpQ_)e$PKW%i-o@0t zlF4VOB&D4y=3j*s9a-vvSSJK?hEWu0&5wA&friUVN>($`PI0=2-)dIU_9Ef zoAgZfTRkCZ(aG#6OvNh?f!Pn6=d<(CFUpL`I<4>`UmBj?z;vnyY}sx=u;<+iBP26V z|JVcJm4W=XV_dFiSJKy7Uoijq9m;ddjPPgup@MiQe-=-j~X{yEsow`S-Gx2wvA1Lv$7K z#NS(``>ZG!bHf}bXgM_Vm-ggnqf46)SIk$qRlc9CNBSWfbW8KW9`m9%o!)$feR?>$ z)whIRON;VDw+=1$*N3hadcY*ht*3KMs(iLqmbgIL0JS(HHp%LT%i&9F&EADiPttrB z-+X&=-p2dnxaG&WGj7I;piAcTKQGS|eRQf94`i~4fx@~l~t@L&sennRo!%jKXm%)i0qdme$-T3M$3?KL`D zzEX}qs*?VD=W8>(=^LX1`6=J;hnZV72u80HJ7C zuWxam0xL=c2gW*J&kK#r9&p-GoZo!&)YVZXNbGpv<339ABr2=8D#}QYq8jIeZntdO zxfr(bPA;+V(PUBBAtOU-1skCG#ULl`nU9`?tk;mKu~l^DSKn5V+$dlDqbu^Er8|=l z_W-CNBSLl+L_8>>P^sjz9g&#LI*{YfLp`XA6}e+t!iS%?fV4hUib(i#iQ1v8Q(w?; zrSmVB&zD!g^F$#PL_LsA8FN_mqFiyP_nq!k%}|fsHcz(y8bt}1Vc`ChW;3Ir5`;85 zt&}`RvG+#IH8#X$M^>d9iM>(OFLqmLJ|Bo*MnR6klV7H9AFxDQr2DPz^=><6Hbdtq zn)fh0djdR6UnrTYf3VoV0TZHtBaUl|h`HZ5P7T>z3j=upyVkI3XfMQWn^%Kzk z-{Vd!#wY@Dm~1!GGWG*ok*ctQkos78E1TdkxkwtGSyn1CZ3rFLyLR2a9D*Eee? zRc{Zl=$;8I{Q;??X*>jHUv$Cvg zvB+i-Mn813L6nEzbH=rydwo4IYXiS)76o^lrn8=$MtL4AQP>tK9vn`Doh_AGGDfla z<_<03EBMl9F7+v#ZbT$)V+wG2a8V}Zieo!23#Ndz3YMk%N@qJAsh;lJyN{vu@&$({ zUHX_vQi;P>xflO`T~$m`CU<~GgfY2Yr1<{VPags_BY!yO6_UWI!x+?ndRR`yKX^By zf@pMXNxTpvLF}8=e#Q(9O+-fKEA|G%pP(-3K<7@rsrSTfmBui&I`AE-b z-*(-$cAjdYQ>_{D@lHv+jlt5ceuIsPO+&QrcvW6pw1K^q@TBwp?#%GIO|mBBaw9&y z8DKPH;J!!ee|)N!Ty?2i%kPtolm6sFwZ=tYPn-6ruQKA!iiaVl8Ve_~V@S3`@25PC zvogsC%ExTygfxZmYt7UT3*jG@!7e5qt8`wau&Z5+I}l!0$&==s&hjz^yEtPSuibZe zI1;Ch^IAV&9H)NTO_qYSFA8gpgZ(5pVSc71PtQbDqbIi96BYU_I7XZ^6oM}1ku_VG zOjgM#6bF#C^cn?XarYC&-Z-i!TU=WBCUrC{S_s=lOvjD>=Kb41(4UUXMN)66!0WwB zd)*<6xo2M4sP0H{HMF-$#gt#+ekT)I)sUX8f{VEd2iP+=h`^i)O3}J-l;)pwL*EuV z{!2{f>vH3Hy;(=Ixo*qBMEDE?4M-Rbemy)+x)pAn4a?AkM));+!T_4TpO z^L=tzTdK|8o69_7hz1;bKMUd83h>O8OAJW!fX$oibXTvcu9lV6S)KG1*AzQ9PaNg_ z!lxos7!#zsxBz}>=CUV&&A)wf=JoYh|LMY^7-6{rh=Id85o1@AiALp;i3 zsebY_S+H&Bh4r8VZL}BkGi=X=2Uil4xLDPYv+CKgW~?uTLvoP@Drc)vWM^eN^gX}l zAVuGpOEeX{Oio3GLlWCXy=2ou|9`PVeilji~u>Q_2$Mj@{^XA-E+-$zkc*Qpn4WWx|M-Vk$K{BH% z8Oqhcx79B9Y4xw!A)oSIS|x;ToaVG+?7$?)iR-0;_K6agENQ(+q^NjKPD$vx0jGxo zVZK_MBBopM!sBtHFYMt96g|-7nxLHunJP4GQB05O%fKhUon;>H4b*3q`;Vs&+>Ejm zywskr{W_qq=se!U&oT9Yu*c|>@No})xaBk^^t5WAubgZ;eRt6@{3q@CyU3LD_eK)F zw#t~jIfNsB!(xor&ym?)QQtr*ld+|mqo>4eD<8{bGg_@L_ zh@WV3^z)n&Vy<&JWW*{J0_olcu?n$O;R&SKRnk1@H``a#xz-JJUt;vC!x zJh+`Ay>tV+WoPB_V6yf$aOdOA`6l$-X=9|_&!}ay7@MlnmnYOCZ($pWh0Elk^Y_zX zmf*+;P3rjG6Bjj;51;^r)`wXZt}ofA1SiE$ago!K;%UTEH*R9z|BJVh1}S?(u**H& z#6+$8J&DBTPenx^){X4-VTp_9kgIsUr|ZpmV~NSR7>1{T-VoVQad<9m83QaWg(JgcrEv;c4CRkT&uyeU&u>@Y2j^vq|#13;)F_fKip2Z>^$1;GEcr> zD5?B8kYAPR9z{*<@ydaRjTy@Fh-p2Zncblm5C4P!2oK^_m=q6qcnLEd$3BU(X zX0O5uE}j1ktiO6|C*DQ-3~u1z$(-I5ee&e+6(|z#&^nUa(IhW7Lj5;4Q&rFp!|3S1 z`Ov|6iwa4<`b~CWr;0X_f#}5-_2@3QH!Ag&?1y|4F`9(c>#gO59dxu@gRoSgKO!X^`C4h~3$b|2( zUV{1z@pFEh_1C^@m7|I+JKH|!D7<43==sClQ#<3rWyWvGmH9kBx`ubar#fzWKvZ$6 zTdJLa>Cw?4cF}x*sLEE<9r9wrVjJ&!HWVr&G zS0pC7l}n=$2jEF?{Tv`socn_-HV@;H1de{%*t7*($qyQX>ZVP;M520Sm_Tb%OSd-? zoZVU3`r~1KCmvhzrbwUt@yC1U21a^%CHauhF1zM!>U^13E5f#uwL6XK7DLZGxM(Ke zY2L& zzqwQRyV%={(5tl#fcW6>^MwoEu>=t!c=?zND_iT{WlI zoO|9B73@>;jK zLti;T6L%;`xXIj9mPNdkhO(i)^#E5Y365I(CcxeO7p+aN*K%0jsGoS}80HlS+)ODa z__WSLE$;qHqzWg8T%YA7DcnlV(MVbvqK+1h8FYQthuMypA_Wuom}jGOG@|7h_L$~* zTy&iBvPUO|15FhF`ZtOBC^y(}@Muy#=rMk2_pGgV`R7oiE4HG|kVXBmR_wSKPezL7kewDAba9|je;#GR`dl14MsY#4d%H?O!L-)r z!lL?Y@lqqWHvd=0$%AaV+7=_RtTjZ~Tlpc~P5{%KW8U`joqN{2H=C77Qo^+O^T57? zdY0u8=e2WL`-fR79;@(|k)=-%M^?>rQ-U*XR&BaBG&RNv7swMU|Ex4+(ZwEIRv8I|K2;**!n})8X|{8c7+6sWFbN zHyfiGX|@){(b$gMZe?gV#HW}tZ6S3XVOfMjtftmJ{z2r$<9P3wWQ?%m<4eKiCmiYU->rtDncWWXnY9g$YX6obG_sC{yU6fU# zSxH}1NAyGXUFJQJ52@6O_*!PUCq;gZ@&&#cPgW+-&W0BzF&lgUktETTBf)tW-!*~t zS{s=XAC%+jPXg3hh+uVl$MWHhblzaj?Q}a|uzcD%?T_#20y*iMb@WQB2{gV#nAhXI zu9ufW2^oHOYN)JY<0XTscE{#Hn-<2*Cw9e-5vI19yN-bZGMAJ~{AnZYwS&Yw0e>f5 z{9C5*tV)wH2#~p03A%I$-*I0Uenaxz>f5GP>9>7r#tiXTHnt&Maesc#ADC4zgw*fJ z+Ip8=BDX~um@y(VSe z+B6-V9AjT>a<15H?wfN-c9yt}-uP~{Qxi@(sjxTx!WL|CP+3%5C@3yK)N}DsZe?5a zdG8TSM_`yUu*!~K-E-q+zu;Q_O=FH(k8Oc9b+h^0>{0VDPK}8#B2(tberMeV&j8_O z`qBCMCqoyfp%vQ=niX#^3@~cH_Cr)P`Q|HFq<(m*tHNHq%d1C^K=rn5tmlU>?}7z!F<2E}-UBrImIGFRbwCe2{2l^h{#u zd-K6;gb(YMSMdX=9vj})7{d?2_QPd9ad`ht+wgF5%rWJ7CtDYo-U9*FqcIL@gWj3B zS`#Sqh{x@YvNJrwqueq+P8J5QZhbf1vlpJSnLw|1Q2_vV!$Wdq`*OIVJ{0-nI~0{8 zK^gfqecYaieQNRrp!v64UQRsB)0F1BPwX{txH0zd)7qeR zayO8eor^t{jmYoR2q3Q*R`*Ms@5vMCUipw^jZR&()d3mmsij_XfAjfdve|iqz+lq_ zNpKDgb?Nq!O08a~d-6i81j2djWOF&f(V@&(d426U$+r``KKC`M8BviHZCVF+Be~?^ zG(Yz;73hf5X`0stkJsE`(KwCm3{VA#FUx*PL<3oTn~pRo%m-WHl=IgKP?&Ee#qM|3 zP-`t{kG+LpoRUEYSFz@1bg3U30Pq8~`Vj>)$H}dAfLY^A)PU;s4~g^Fk(jCwo`8DY zC^90i`70xXSxMZ`P+}x4)-iFx^)jQ`g=pTd_G~fS<2K@r5GKwSoW)4Ss4$e_>17Gk z>Nzs&tzdOAcbc?=muyDQ`{NCx2ii%G6{4{dskPZEJFyiu8&(i^1S*<6no~COWBlQ? zkGL;8#@HG0&-NEU_{)bEG}ES7!1xUxrgic%v90gv+N-Y(6oL;e(qfQg%?t_th93Z2 z6!{EYXof~$5I?2|2Bsek`&91D=3=duQf->q1WK~qF(VLk$cWt$;oX;C(woj;J`bV! zN%4xR{Y9DMz4B?&f4C?w$*dfA`13`v9*{%XALlSK;aq!$0a3VBRaHrGmo$823MpAx zSz%dsYoxEd#=0&Xesd#*Pymsoyp2tUmFgE~HyjH^t{6#ld#-s5z#|WQS|yj>Xth&+ zl(Pm^%8@fn!5b#Xw%T28Y)0k5#ob>@93SOs>B*qn(O$yN7&Go?|2{7Y@ntM&am;q~ z#fFgkOyy5Q{uX{+c@K};7`7w%A!|MB^0Qn?PkD)xW0&d1o$po)WqzzbhhizArLijt zqdZu!m<*62Q#p=}g+xI6Y1CTLL{ z=YfPU&jmOX-dPxlQ)a~{?6Eo`96v`{QLoNCASW08nYY=I;n`2ga>k|l%qE)xh7A07cCFOUbkK+~~zcEY&~8)T5}D&-m3^y-&`uN{6$g zGXD6u+y(&=5uDZt%04zmMnKLzQ?ZKQ+o4B6T;Gk&p%a2S*2LA8Q8~Wj-#b)8eDN*> z{ZNsU1$%o&=*H#?7t6{(g>WWwlUP7cB1<-7XS4;@mv)ZSRiLWUoUY|4R}KTSLIAuq z=O%ma0#;I>rn+H`hpUi-4E%?%=Rm7~@BS|ZEiPq(k~wZEm4P3Vg2Ub?G{u5Dxi&G| zRbz+W$56RQ5o&~wt5~lM3DqP@wc)Y83QTmG>4sp{?va$7E1LJ*aCqn4A-KT2A*USt zArPv4HZOQ*nGbx%zY7(XS{M?Uf?O^}gQIyxmDDV<0c!PltDW76K>+YHeTD78tLbwu zF#m+DisX3-U`R+KXPHq?x|x(|H#Z!+y4Lt|be!y~rQk$QRIf7oLT+C*PwP?WXb1H! zenRlD0!5FC{o^BkSb9)O8H5K579#nT9&YF8; z*{C$fQXGz@`IdY))=JWb_&Gw(FrFx3I^ONGA{yTyMIuRcgnv46TVK+$7$JiB*9}C} z!5oD%U#o<=7}G=^8kR&8dWnk+|IB;(7ona z-pwv3m$v~K!TIg+q6NAH-JK*9qTZ%=azid{pD!tEqzJR;va@wsI@8Ciq%$QIig0;& z7T@YZ={mdqt#ZZfP+s^svZQ5>zd8_Mm_5$_9E(YZthgKY9?MTbsbDD9k4M4XZ-fq5 zo0#Ss{u?fyOxd6`T&m~pYvO&9_#IY2;1&Ii3IWHg%#Ix1G3$rhXVDMDo!fs^?{7pH z>tPP9bN4cHsKC3EWRzyAi$QQU5*K-NA(!4uKI{I;9zAi>;N*JXnOic?1wtcThTv>K!1Rp6 zt4fD3#UZ}5Mj&BN_k47YnqsUpyIfxar2f&Dy!bGU;!xo^Tx?98ljs;OcbGr18auypFCV6L5lhfLfCnNn`X^vThrOiAto+zylHFNK~B=Zd4{L-12wFqKht zO~<`losJh)Z~l=Z~Y=r1O2Wr~syv z*;lcqNuPdzY0felNN=wY_0*RO+nriY4tt--G4A#`pdC#C2%3EeA}5s z9;ZD4u zeT~mO%h!7QgUo|Xl_Ym>C$Bk%-bC{;?Xo6c4VdI^*qTx4R#N_ zhJ7xgG`yo-UE$&aerAwPZ>YYQUD4_ocL#V3vc~f5WDFfB#CPs;MI=A(`nVpxw_Wnt zVjf|MV@RhQsR9caJoeNhkZ(T;$L#WKeGTOnl}l~R+xhYy3!xRN_T^Kdq#OK zVRx2DyUUCWV#>3(_Vn~ri36dm;P+@qiwrATRVk|l#fv(abPJb-wqHmxl z?!%dZ?Hi0+SXEOkrx)(^_e70}>Q-~yXWjM4#5+?`h^>2A+L8D(jaga!eMtXcSbU?> zuwyFT32BxA%^?ZNp@R4*L(kTQYD+1zOcpRr!#$DZJNp!L);`cubTekq?*uq~v9FN1 z&xZ-bu@%vmj_;rpEOX&Y)*-hK7Qqx05k`0M_ftM|6c-Qjll(WJZeS5(BU)fHRIHv= z8h6uEm~<)k8Lefxch>-6F+!zs78s#y$jG#={gvf2NLo^~Y*hd*bud(|pjJq|%NOth z_CR=zkAA)W14wzN**)^1-$lL#4O*n_rs4FnE<_`5bAsdC=*;SPVp23hQ7-wll9C%) z^+wI-!9FKpV%)uw#K1}b)Ve-b}| z^b4pvDHA|%?|Mlm{)8*aV5*l2l#7nb28@pK<5eAJZ(>?|u!(edl&E_vsXDsH$yp>h zoh#SgUe<{*-e~+_dg3)uow56(ahxny-m(F&9&Pzr$ujCVM(@R2@AOgBoOY>7)@JM_ z^M36rF~;HG3QQ=^V|z9gpTuhwyY2thMS(?CXgyHd`g24CA$)zkNgilfdyQcd>F z(U(Gs5yV%1$XhCW?EV2EXpGa@t;ON1$v8dx4(pr=muqP7$soFW<;1iIu&7!L)<1i* zIJ14#JB}#qdc{=#%$?j3o_jEYiqej1Xs8b<32W5v;P&3UPPb>#XlV?qRQ9MM^4PgmLL2gsppF8+TOJeEADGsZTU&s03UK5YjvCC) z9GD!Xq@=Ly?IFuz+Dog!;;t#}up}c;yPzrle*x(c>CKqIF+O_frW0<_gbCCz2q$Rm z)CCSMej_&j67G%Gj1sEUYSI>SVtQst^O%o(vw(*k6Niv}oE*YCYS=2FO~<9TwY|6V zYgSg@FY91zR86iS|M2keHSlyHppSURpWwf)g%jOdu$8lj^)N&@TZgc3!n>D)=xNUn)4 zizE-zJ63>jBu=HX(}0g@jwpfT1%{KlUN5wL)?*iqfNn;$wVmdJ zk2sz1gL0dHhV7t^Y2{7iqrof|LqEWsMe=M7Y5^+>WFneW!tQN}+mcQVi$$((=%VRR zvi>OS@JH%bDi`~Hfo_+Q=3UR)jvd1AgT zu-z^4UGVKW>GS*AfUX~nAq#=JDk;Tntc;nj8lS7MCi1yxM0|Y<%PTX|%`Vh>=Ef%E zx&DQ3&Tnz}cjjP`CVQOlnBqS?;t^~_8i-`8PT1&r*gaOMVMm_}10lAnum8Y0Q)A1w zHUr)BwM)80{1;AX6{mjMlVaz&ncf@Nr?RmgX0a+gz2Bz!a`h?GdBLcfGmABIQfg6O zdH3yU%K>P>2g{>kzv+ZE2o^wI{#&MTdetgC`4x=bHf@ex0C z{!Lmz-nTVlpMkr_uS9^LN}p3mqxr;sXOt@ZDd&wFRI7lO(i&uZMA!HUz@t}HcX{LO z?On8#DpLKnjY{xN$dysZ4Z$ubuqnP35*ixQ*4DJ=_MVF1dhAd4*Vos-b#+;6>Zq$l zoArJx!gKmX#c+M?_(7d(zOIOCK> z-bz02QA#JNP2;m*gay);?Vr4!^^&ZAjoi|zTx-D9ykcJXZ}w&#^Ef}iTx`)K#s%Gh zUY^z!Itw8l{sI08?$<6C+}GMJv570@`*K5KteDVYex8`6H`+}BeaMUH>k3W^>V*7< zsKRKq`zIOhRJ061*M2_tAkNRA=YxJyR5VwU@TZq58~X7GMk1mGYZB(`MoItEdkjuO&GmNdL?b!1?yhITPC2 z+539T%x4RG?+ZnHJ2b8`nTmQhBZF?!#~_3Hg0Hg0rYsAPG*7YlH?Yd?s_$!epKtUd z{>cXa!w~=RDPRhJ_!kc{XZ*`QC-gs4`}a>R3ApzWlq44JAO8D&QV1MeWas=hivK|Q z^<4hr?Gee%4Ueyj*DB**vgCgOONYe5C{0KC@PDI-zkm)mEEqElo_7|^ z|HpkDM*eS&5lZ>j_SoNp7{Eh}F{7FK5!nA5WB(2;W_9l)Lcy#8Pya=)e+3;KHxp(W zU7r$D|MKep{P#W(vufpn*|Ps8U;h_QFkpAR^nVLW;I1V3KkhT(k}Bi+|B3j%M1o@f z#cI$3srVu0T=qdiH0DezX9huLV9U|10RI$BsSv5?M3^c15>6tr9|BWft@V#xpGqu>;fw+58E6|-=%eXXneQR}pl z#AoOI{E95P!lhIkOWn9e=jJ%9t@G8ic->xiyXTh!Y;3p)tql#WPpTdO0LnD)&vgvr zZ~wx>zh+_+rs45lPfrP{;!aL(TU*=5VB(uSLqjnYql-_BRS+NH0wmRLUrX!Ch`T1B z+r)Wd9@4(YG3i7VT2_1Zl#bsB1WWJ-)&w#xV@#0eE9ZLW|h>a;#m(p!8Y%fy}iAC55IHcKsQ8CBu#w0F;D?3Ag!-vEO}w) zVm>oEG0Jzk;)1Vk^^p9x0Ok=Gj2p`|_fECj!I8qUe;-_0ep7L>;|UykJNf3B~2#dc*$$J& zj`sG;{1YpI?5c+9wt6q9AMaaKS||z@NB2H0G1k}S8j?UoV^T)CeIySSi?ym|B0jjoz}oeZOnA#TDz7Yvh=j2M$yk zcWgA6LA&{`muwvm6MpD%k@TDR#g$U#zIfCHcqg5}TNTJ7NadX9LP=vOPQ4kh!w zjZKQ?QyYm>rC7JMwXG#{taBmK z0NyJ@3WMgPkX1x|0lP$5rou>OW+##VwI`TDZ?V= zq~)-x@p4`?gQlCk6wG1C_Ay+}Vu&qXj^->bD+e2RYom-`67-vteLWZ4Q>mC4g)#SK zixhwd_NUIZ)zA3&4AZ<9iCKT*8v&~XPw~0H4g@YPhnq3`CGgnemfG#ANyaJisF=im zC^qf~H&Njg5ru-$PVuhIGfyW853L(M%|4%BrczF^rl__-o5aji>iUAcLg6dqUsD^K zqeUV0t-3*1baf+OkNtbG`B4{?Gf8CaNDx+OOo&GJr2>nj(jDHz#A|zAkgQjRM!*Y< zr#6m3Gu~Rw_aq$B1d8=9+XQl(aZ9gku;3?U81s7U7ClT@XY2Padh#cHFeUb>q7~Qs zMk|%wu*2Wsxn_=(g`3fav_scWXEJ{muL|uKc=-`DqGmhzj4Nj6Y}>K#^1M~0Nb#wt z`%X-d>Cg3 zKoN1+E-DAhnMA??5~;b+JZ8*(--{!;Azjn7ojJeP+jaII1>z9Ha%|8fmVep=EjO@I zOU8w*+o@cq47W)80Af{v|)p%^m@6~ zMOkEVWnmxxJz)#`_!z!dlSKyZy3{j5%kJ8y#Hi{{a0O5TN>zJr>`Go9ld-V%6B#P> zyP3Vswn+pH2*kT`$Gd-Hzn(q{zgD*)f%iL4VVAYgsn1N4bKg zTe0|KWRdB`^vWtF^fS&00?ntCc`RJ{A)KgMurBA$ma%eL!Pl;?1@1dD#~_E-wh&^`vML5R?1D@4zhu642WQ*)u(2 zI&*zxK2>1g(kA42RW@W0gE^p_OWfWpoRpM93z}`af2Rd-_1`4}Iv$=8;WnvHMQWJJ zxL4hyW4!HI7gd;c5R~;yWc|tZUBpC#LNvdD&rUBGczC2O2A+A?KbA;Lj^RK)s+bUP z)5?_n0QxQ4^bcJdz<$51yYWlt!IzpaYJjyh9$`e$Y(w|L7Y*%z$EY4oOavCm@Mm^0 z2H&$8J^6}r^G$CMYTPwg5}p?PjXW`M`4ij>u+?sX1OaG&>wwtE3pYorPt_OtuUK%8 z{e8}Uk<0sRM)g!{U>k^F*y+X{wwW>p0fDKytV?jP1t%PVsCW*2n}!sl%B$P?cqm$a z9H3?KFVBjb>L#!$l>eahmAb93`R%(UJgkKL_IiWgp7r$g#ZFiDw%;b-Hg0yXjQ{q{ z49L4!2~-TfppAa&(4u%vynrWfF9)Dqi~vCEo_=O(!l~ZKOzAsZXo{81wbm{n2#}(q zo12@P?6I+Uutn{(Fic#phl0YE~{PQ8?dRxv~-bTPDbb41W_M zNsZm5kOETVS%mxIaV441nCA!LiMNsbE#$Pc_r?fnX+*8^+%`Z2trD@{|6xu`DPkiY z3T;tS??+yq-5U93eIDAq%zoUpvEsbRAt-xxe;jcaNiZqO&B!nN#JfVIW)oADb&~4w-e4CcIrFU6 zp-y_TM3>mfpc!MvoouDVUkey*-Gy(DZ-37hP6#OQw0YMNNV|ReGb;#;ye)#zKzZn{ zik8`p2^ocgI7Z;D5TeRSCmgb5rB#v-pn@0w@PYyoa9t0xDRHsAi-?%AJ6}aRdLO{= z#1(8%iHJ#2siPO3QteMU20S}4y5x)m#p9_s` z2k!zyhQUEVYLx5cDT(M5xsghqQzK9xwULu0P6~`-fA~m!k)OqoY}GqX=BVUTF_bjFSk<;q{(J|-nkIaguE4(?%rCSDb^j-n05jsxpk6PLcnU zWvbi2N|uCKQ1}{DM3`b~`FL%+4!?9SLvkPWo&$V*)-O7g8H-7Yn$uK`C5o93*%sgN zl3Yo_tNA`^@UQs*TFvjya42)0ehN_h5+82_>eI8_M5azExmn0IR4N5NW?vf3X48m?5?vyNb0kw%~W>>lL|Ore?Y12RPDiW}=F5BGIQ>Ma6i&^Yfo zfw;SlDAF4oFQDN;8+A%!%7S91E;%JZ3B2r%sP5a{1dt>_X1&F(z@n*u z!BG%jY8s7=LB{O+fyb{S%&$47riGemAvZvhgM5=KxKYsfS*Zf@!tb$YSl?L|D z6iW4{GY?s|8`qxq4D>{JOKk(ge*8+WYu-$p?dti zs=kB~>5kmkhGGOT>7VS<2|fcOV^aA~my~G%x2T(B}RN>R_e%&s$qRK=2yu(+mL5wVuRK zh^!KwC@@6bL&o^R1=#gof}y(lv&2+gZRO_T)v`*xb-r77x!UOHOrL76ub<2}JGY@_ zNmxo#B236|mBhnji@wIBuB5oh$Yk;rle9Ocy=>9<0VUtCy= z6wPc>+^#oW`rf@jI9gp#ku5i)n#@{rgoCN}$nCCH)3XWj6l)bmij;K5E7c%RmBZe` z%wPT^EPEqhgv^SUl^78{#zp#~X7eb2%n~v2R3g0~&vkXJ4(rRHBGWa^GiCdq&Nywj zuIt^0o2d}JFJ4F@0fha@f<|HU7bhD!JHH4gcQe9Be)sb~)D>1NRoX+gKa-ri?Aobo z+kHYkmPsT@=j8M;X!sXQjsPhh7lC7y`@!8YS9C(k*xk52x*?Qc^w%f_>676xdW>UU zv?7x8*23bp_-TgCwd}rPKUt-P7n;@T9+p-N*OYCAb&_qP zWM7sWlw+8zmq1%%gPF>^*nO<)1yjW4A4YZQ?_3q~7N5gG?PDquPiMv)SQz`TXDk$n z<}KIjN+|L|wIZm_XLKLJfkjv0|GvnCosJAr0(o~#?7#Cm_Rm*>Sa{cCRsAnox&Iyy zF$GphO42uQd8%U5ml;!H^r%xEzg=rw3Ksq9pxCRp*j~+vX{vzDj`84yzK~{`sN+DI zk=9dm{SO}glhQuhKb*<0tBngK5L-q%AK#(l^P@q zKuom<#;sy;J5Y@>m)UB)^;VYtqU`<;izX?#X2Ub8OZt5PD*bj z+#UQ9_~bn3(ln?QWbb65cl61jzkapO{90Kt)wXn}J46xro+q9yOp|9pU_y)Ae3DKD zFC@@*$$Yj;1DhDz>uDU6u~GSttdP}~+MzjYESdun>4=U7`op}=-eIabsgV)nTq(=| zK=7n{a`P0Y<=}0~+(!13^_Zd5y&H3zjnKsCSk_8QKlzbdV7UMYQ~fnvZ<#J2 zZZ4nGzo8F1t<5p0dCvDov0!A}J6xaY%daWV#djN&3a3DK2o>CvimkVj-mXyY2m}jp zX-xc_2i+NnFW-5s54+L!8OMw15`9j5kiHp|j4>$cbzpRa=0+rh^I&{Ix5u%8)&@*p!x;6 zpO{}UOaeF7Rr&oLB3k@LqZA2s>us5hB9~#ONPckQsrRTU#pfw;fvSL?t6aYH;o$?L zHrp!her{5FimP8geOe%|#%#~p44oOOwTnkdOdI$$S=>x<4JqRjj0@n2CZl>^N)!IX$S6RS~xuFBgbMXx5pPK9X?2;c%_4 zXpjZuHp}^iKYCDU`QrbrI&b=-ri84_{+oprpWU;A{~^Ym)HeWIa+i(1U0$*B%AKvE zGY$3Xo-h2ZYOi1Q%J=Qw@0axNO=&EelD)wH^NQs&_yeG^pugd|318*g>9gBP7o@5N zKjq)|-Q`HG$#<+0Sa1nQNZETR*2q@!TN}VPMHKwS8Z0zls{maq<2QU8YwY zbk=!Ke3b^Cxp8>IQWEoi{@xF#I<$1=hiRR+`5JK~*VJ~)H>v(GlMBo5CyR==@@HPI zug(4P?@#yoPBXjC5MXhbD{K&x+NQOE+x9nX)dsL#*2UrlOr5iod~XYynwh%t{L0Ml z1~pykUjN&zdiTo}=@l<_cy;@%UbXF8S#bR3pADDau35KI@Y(sfn*D#jF|+)V5d~E_ zjUV+F*y_D0)M~5RV)fwftEmSIHGbtUzRsN67#&@H@zunIYvbQenmAD{`}&#*)$exY z9&nhy?!4*ou*aYpc7_DgErkuMc9li2Th~T)diI9B%MnhmN<1^)+WXnrnX3Ep=3kYz z5#We*e&s3_#%N>gTe~DWM4f{F$!0 literal 0 HcmV?d00001 diff --git a/ALCF/notes/assets/cooldown-train-loss.png b/ALCF/notes/assets/cooldown-train-loss.png new file mode 100644 index 0000000000000000000000000000000000000000..94512194c37e4934791befde9695d0a68fdb9995 GIT binary patch literal 341947 zcmeFYWmH_vwk`|=Nq}HMg1fuB1$TlwK^k{=O&~yk;7;T2?hxGF-Q8WA+u8e!b9VBc z`~CmMxMQpzy=qp?s`b?Lst!?33!%0I))2tpW^D(iLqPBey4e{RSpuAh3<0L*HvA-K&Fv&a=EnRa z>g;k%a&{sBGjmA~2Y|ANyo!;Br4hF=iJ$;HpBoRDfi=L%fXL0-%Epn$ji2OSe0jjX ze>XFd5dEu*lO;cihMXdih^+&Fh=YNNfr&%_o`}!E*n~$(RQ$gNgWvc`%$%I;co-R7 zU0oSmSs82{Oc|NExw#pcSQuGY=)pbc9o=o54BY5#97%t(_%{zxfTNLvxt)`_tqsv{ zo(6`t&QAO!Bw#wxe=&D5H~Am#Hje-G`CuRze^VHl8JHOVvpc}e{QuJZH|2l2{~pYv zp#*TWwQ~MFJwFLE8yh3@zgYc?ruqLt`yZ03jX8L|EdLtzdjcLsb2osMrl`3!z{c_4 zY?-*Y{~~MrCz+G`Kgg<%00$9UYw%z&S7jpy0GNxJtrH33Z_pY43w7}JG5+`dfnWYx z0suDPZFd0gFLzuFGz0{qoVkdIqLhdTk)ngGiMbU30zxt*P907|c>ptATZtI@t*DUn zzT8iF2x{UQh{-S_8sc{-(m&n__e9W@tuKl}4~3QmP}O1l>W9^=ZIc6!DJH$&DS$xs0M$xFOoyKu3?o}( zAi=4Nb{U^^LpLN#jhSkIf&E~j!79Oye59hG^Y=ArmaHt|Sy+gko=#y;auXBUPBhiq zf^ebHZ`^NWzf<;KZMutirT!#DhpZ&-PKs_R=qFCul(qjp3A>A2ijH-TMl>7Pphkuq zL0HAjkV3%8cUCKloxs{HxHAOnt$66GwoT!n@|fsS_nk8*k{c61NhTuujjKR06&l$tCno9T*hS=9U=Jy3>8l4v*ulYKNS0AhQQ+vdxiu1cf6>YkPYVy{9X zUSioe+T$oc!F|CWFh?JxxI(&kNFH+fH#ELxWqeUR=F%jS`k{W&kX}Q?B+rgnAG6zo zxUl7))uKpl3u_{;Q7~`zhnoDupbP1g^Aw* z5!`oPj}(s$K~;sf>u<38p-9%5x)CEK367&WEcvLQ+C|8qE~D)I16xTH-##d0YmieC zY^bQi{p6m`R~dkbBp!uBB|-+NBhpBT^GQNQ5|M&}CA>6Xn1c8;;$B|U2~FCW@cKoO zm_s+PHLne144Ig}e-uX~UfhIjAUq9}=(L0=Vc^2O!-?1m zQg=1$kJuCyrZ%F2L{FI@Wj|(HmQaap1n~&c=(CDnj^H0)^9$Y;A0rsxGB2V%Ve)=Y z`z8NnuR^WlT}6eF2J37D{^!joj5{;Cp7&#zocjt4)7Y&hZOfVfiXHhId4739Rd1Lq zO_;&pC|+G2ZHm!BhA;N(5yXZUWKezfbXb&Ft`Mz-C$ zzO3gn>fAF!+hcDp$*zqy(4erNCgmVZ1 zA_(X}JCyfmLhy<3tz$t)I0mz%t{d@ z;zd78CGo2WxqqYzjS{Ez%WVrx20n+hNHDA5&j^Pr(mL@CigTMmC;Q0@SC3bwf1OzuFG2A)xDARL5_uV{dEdoTdfSwp(E%;o`}S5R@cD`VsB;-6}MJ0UB`_ zQ;ec?0nsrL9!YA@m=fm(dz7RxT4_l7xuz`fF3KKC0xBo!9a`iE-Vo%F7cw*{ps~Uw zxme774wMqL`5SZ2gO3Mjg(=tZCE^^!%|Au@3yt43Ch7>UiU~-!iTZp32~Ff^D78>e zpc4Z)05AaF+;0y4lX4G2Gg(^7-i7a`@eWBW$ZP#Ql0AYSrRC&CWJhG^)#OU*6ibz? zfs2~8PZnp>wKFHvjfduk@Hab$yfcx-Ybw5C*%KvZkToh=kk;I#5iZd#k+*7rnHvSa zrcf;C#tUZ6rW=+;ywa`5g=8GwcCDYWG>R;x8tCc)P-o5a>@!|YT zxADXlW~Luax%%u<44MtvqbffaL`|CoPDo|z8Rgmg*+*PxZDsfJQy_j2{7^Zd6=Os( zrqM))L>s9(z|(Wc%FgzlRW;Fu{e%zba^_Y3{l(nh#p530|9_F5)?y(Lu!wiGEj(Hhz8G=rVu71-( zbI3AB^NO8mgMyR(=ISnJpleEGcVc~F%X}m0v~l8!M>`n5ORrbpNbDT`T5Emg@Md(W ztGVOJ?~YIaNuW`n*(+VXBAqC?!n;#MpH;u@Ol^%pFkZiD?aoIRln3&9#d(E;+=cA) z!-rJhs!PQiwHXZ)$`{gH&+OpTu|4HGy&$5aVk@94Al2aIVs4Xaq*DD{760w;Dd(e!G42z8j`?&i?!g)W8eFtK;(}J8oed!GSx)j0@653ry z=Aopeq&u0iov}SLpc~Vxkj#1D6)GIcmCGF$n;hFjTI=g@zU>^-JheO}Jyj-Gna*N4 z*jNawe3eI#Bo6Eax-g~vOiyj2eIXT0*6%tSTL)MT?|-m^If5Dc2H9ELNf1?TR=Y2i zXS;hioUhrYX<6k_^-@h+<=LZSA9oLs*BFgYM|#D&TTjoVuUM7gX|Uh9C^jwD z80JIoLZheqJh!>sLr*SFjvzze^Wjm>yHLEaqmsZo$-f1s#shvH{nn9kE)tPFDudTB> z(Lc&7H+pk%*R0TNnrgVeICf!X z)9+}n`Z9t*y|d$`;p}Sc7)#%m;OqHU(`#Oak&g)vT|ORDtzDSs8=Izowwj5Rgs z1~scqXWxWArmvVcGoHR%oLXKVuW)8`w0#h$P^dgx0A0V#A&KId@%9VMpJ(qx;vzF6 z#j%rwSK~eM1iBKZa4WXHD6i|cBEP!!BDW(2ToU;5 z_{-c3FLO={ho)xJf2N0xQ3xLSFnEJ9>#to)*EriM+%L8WA$@D^V4uT#A0+x?ciL!V!I1uJ@nmHWmy@G~w5FQ-?1Y3vgDYpRclt>16pc!3tDv5$?wo z5CT(bAc4Rd7bgUS5QLPdu!gVHm#qJ>Zfa_HeInvg9XhUsH=&o=^p2v zPJ#smc9vKBbTcNIC!}OMmE%WToX*!-{`x3vjr)9BvZyoRz)2xaFYjg8Fj~@O*wR z(90*~p~*xLP=7q@T>kE7Z4&R_zo&qEcX5FNC&zYeIygkTU;R!0=cD}IPJ*t5T4hen z*HtgeGR=jDJ>b);INtL`KG6Ioxz^;#br#!H2#<|400spD@{ea98*(mPRl@p@#@~JX zv&Ft##nV$5FwwEaDn-Xc%;r_%|FhR?J~R;$NRYo`p8ijwzY(ybcSKv3*M8aFNcJib z^3QqywG4@OLU>NGc#HoF+VeH`G3E$cy#%JoeTe_S3F0l6G{j~`_Vzduv=GD}k4IF; zMk*c+4FiWYBaL+OH{%SfcUFtb)JPFx(i3`Q@8|yJett&+VSj$++)`Qp7pV`ze|A=F zvC6r)Fkh0|;*&r2pIH9h4pOpppRXqqcz=8>cnKb9LY|(yrMmU$w>h~JZQf7jYoM3y zjZ@(_a7Y=Tm#3-k8(sSi%-NpP$QO;rcNjGmGcAqnowckOKe+6dB}%eWjn>6D=$mUwuCdYv(z_$r)-Jv znr3Tt4Br=P>;a9SqBNhwLUdf*Y`WRGxv5R^wJBR2iCbal!Tx?p98MQHU9SnN&>!Cw zQ>c|U>OD@hv}(+l0mjCP*1Gt)WRw>^`IC*Nz+5Ia%X3P)w|(^5^+}^i$;mi}V=uTJ zeQKj=G4g!=e0+Q!)2ycWM|pV^kB3LLW^<=2@Mq7@Q1A%vRT~DlS*`X_1j>W&3Tz7y z|5PCes9ThMka`)vo+8{oFbokgFd*iCxUQ~omoX(vYbwmzG}Llow1xv zP7Yb8++6=961<=8d6wUqMKZ;cR9HHaaea;W2-fHQE|{|<=ufC1HJ!?230Fz2c7{4nz&hg((Xgls{ zmXQSh5_)ZW71pYIcBO?Jta6aErfTLVRSuc&{(IFJNxz=&)QDPVMxBqNBP;)Hw{nfW@?Jy8;!P`eyXP1|({m9w zUR-^F_G~ax?iBmnisaSWYsOI1DBo#Rh-OFaZE>77lXNvMXNk3>r;r?(|W5bH95IvIW_H{M3NrkKnczAfZVqG1b z{Xtp=$|M&6*yN{LR85uV+#J(uKd+$CZ0$?PM0hO4cza)8P?|qQq`AD)-rSlR!=#z8 z^#FNf%SODWxvrTyd(Bu&g{q-G)=&&xt)?ju6aQk{@;ch0b53(E5ko;Y7>{=S_47Qr>W0k$>0srY4JD5d;x})~= z-3E%s$I2_~GEM=85-hQ(iHQ6Jp6=hkBOtKGg?(f@ zYlW%#+IQJ}JSk(9N6F5+eW4Dp+-{og5x~dKFSipzA)jH;6*P(Ox@v^kQ%^}B-nMzM%aF=GrJ` zM~9%U>)x2eLUobxbXb7kYt+V5;~32)(z zaXGA`^FqT3*e?s-ip=Tq2A6FyqpISL3&x2@;s|36H}#XL6df7!*mt^NmJ*PoGODI9 z`QYig^r(h@EQlYtUjG#+A%>^Z?8Eo-gcTdl$xCDI*G+02|4=MM_?GNT8z@>)=7<5$B zGV1{MKe2&C;WC8UEuSE$%PotOkmuC>;}D?v|Z{ zo6M5iSkH864T>u(Cu;Q4%+1X|o0jHzbY`enSXJE`G0oq`M5h|p_tz(752@ht-f}tD z&E+BU-xGqlto#&R4!O55H%(GtyOJwP4l=D>g5?uC!Pk_L9YWdoNs&i?Lccr`obREX0D~8W0ibXifxdd9%ZnrM{=5%>QJSbXgb^=q@6Qd zuGu_VGMv=s!6ngXA=mw@`LuD223}Pg?jK^lZUuz`mnMe}`pd9JeDBC~a*UElUHBlJ zuWou&cVgll5+0KID-bq0E-pT9Vx)dqoD;N#`M8k}lp?`8sZ`FFV=bmY{m$5OuitdO zIkcaxAdnj!oh~tKy;wD~t=vuTsT`Gz!|hH%MwTN%aC4}xu71}aOd3c2k_S-PFNj*n zH^$HZkkmM`nHmM+I?_6zaeP z>uX(q#$OTGx&rafGEyYzYcwQ0`&OWmN1LiHp6#fs{R$iabHF)8T@m#NcX>_IB)D9p;C)}o;nkChIE^(@HYFugVC6o@= z>qq5nvS>X;rqMyptIbz&Z>P!EUWHChPCRB5VgFnv1QgN-@cNl}*r5MZp8fYwkdTmS zHn=#4rGs<`*OUH(4Gtcd_!wg4FJ}Oc$ALcc?(+=~)mVFE!G~*B^2f1Rt^ff60jImP z*Q|%D-N~9lWl2#oe(hG1_qwhxayeK71+!TGGQD`{AAVUZ>k~OBo-FBm<9z&RHq-O? zcskW3O`{}pF)}qZG?;+Sb)^wWz#b|9I`?jRsnfO!{Tdx{Q)rsLMb!z@^de0Br$_z2 zCNUZ6Z*y{TG#=<_XzFOH>Bhr`r*PjWEZGcFQ%)_{>*XV34d-8t2`03(lzhvSOQ|1U zcE73}DcJ?!bx{)f#$qx)Mwhjp;($Zq+9J}<^*6`@R;p!~X3C)yl#~Z~0;)%b7V;VT z(pBz|j2Mp4dah~(j(Rum-knw@u}0Kdbu2BVFf{IedNkaiyk5fs=5JBT)r11$9H^#=oZfZ^7X?X1`7hx6cwCC4NoHNZEIj^(h~|;=~4;^j25Nq3BOLO zJl^rUPSxxC3Y(_62dTyZef2C<^VErq26cJ&EVR0PY+84KYEfT#T~4!Fp1oT0a^$Fa zY>qWnc&Mf+f2LZ|Bkj&Rj_yx~gjJuVaz`7Lt{{N*J(^7Cnusg3#wa#47hT7D>`Wk{=SVd4Iuw2o%K; zxbAq74pp=k!GfL=}G6!kAp$Qrpvu@?*qiEIN!;jEE2Z7)yF=uPCcCmH^4 zMjjQFwTUOUp8!XEMPD=YW8bG3)~;8@)1{HCs?$xUqW#{o}5-$DZ!d1NSJiSqxYCMaNd zqHmeoGdi|EDftV4?Gl^wAbPUb=DR$HBy4^Q^fbp?O}es|FEq9m7FkcHzJ5W-xYAug z@0Gvyr11p-MOQQzF%tB}r9wjDd#;xe7?YIb1g`syYR<11&4&{|vzd((Nv~hL8v+-M z`hY$ZA*Ykcq7b2Gg^Trgg$>Vl7=xS7cdPlphLLWm4`xhH2f)S1YzSQQsofE>&DdG8 z7jU%~-8zb;-F%8dd11b?FKM?q14GviiEQ!7Mz2xLlBoz2jcQ$XXW_rNb{@Fu3FOG{U&Zc?CWU+Ir*Bqt? zT!s;wN44*Zf4>iuR@Zm_He<1H%h(PwQDf-LN^dAAB`c&jYdn%yj>;j!xRA4)>?>~v z1-0eb*7T@dAz!)$q0JPqhnOs9V+Y&mAa@IcFQbV4Ty6zxXYLRE`wa;~+ASu6BUIBa zP3Sr{1M=5=pbO1F<-{U(K2V&N>!swozm|-`7}106Q1@6nx7G~HO(|H%OzA}Tww=WJ zLo=E!Zg6(mOMRpse*OyF&swN%DhhD^+q8|7LZy|r-4nt#=E3g!?vq z;II-2VHR2No)8`xxjtW8I&D7jqsL)LUkNmz89#&MZ}+ngj%5noz=-{Y!zkrACrh-R z+bQ3OCYRBfCRGlbMJpOYR*%?Ifn=Kf;Z&DH6-ZS7uVM{MU*~bZZVRh33_f3ot>Q}hXpWTb5k|K zI+iG}Pxt)wcST7|T(lsiB01<>BieQ4fl=%JK!R*3`0s+MAM3k@nriMTVz0j`Lh8_9 zi)DV+<{cXsCnwA&pZeqde7U}KLRR{rLN7I+8&%pZAyR&jYK9j7;+*1S>~haXj-ve69ednhg}+;Mq9H_*s}!mlLdo{eAn0C|VN8>8s8W{u~Mc zwSqFAgKc}>z--i17(VltH2;3J+x{K1aErM89nSj`eE7Ms#%aB3Wx;8)tTDy+zf`@2 z{_hj(qV)Io=UiUG=LUs^A>ebGIAPUR!vACC5YVY29V^^6>oETyWl%y(4-M{b$57D9 zfkl;yjH`iifQ*8G(k{EgYa!#;nbP;gnx=+G{+`hR8icjG!2I17xZ7Wp$b|KE9o%N2Rxz2^IW zOaB|6|A+dYNymTV^Z##wX&_W?eZh6E%P2@ZJUSW}83{+*+QM7CN^OC%u(2UgP*4aB z3vro8#sQ}R&o}R;bJUeREpgqUJ)#x#_4|01XXR&ti(>22iS@SePqz{YYoH)e?uR~@ z*;r*CzhRdaBsKD;s))=N@h&T8R@5^|o8;h`NO*?g$GZ&tZA@AfGAbL>kha-ra6{i; zWe7ktne-?a=1{^sHr@E_Rf4k|Zw0?uzrUC+ zXY&xPXY73b*Y?;bO`Fp9u<$t8 z;uowjc!ze{mb?I8Cq;8y9_Hw*u^+De$T(o%4n|camTe-R(Ub>d3FzeCJGtc|w^B0gtrCw5rUYu-S_B<$_{C4(KW3Tta;4nP|yV4ey@9Z`wSjm|0wxpvTGXE^W-;H?Mf3~v(@?#EM1UGnE6YpsoGd6CP2?Eq>G?*| z*>&(kIbYGrDOT(xot~p=ykR?EAAO=SH&@L3`1Dl%oqeU#X;OX+u>9ouCy?8Vs8ASqF)cyttG zFt?Q;1v?U`Y?ujsIdOOd67%%YWXfz9kZcbYQe2}mb3Jn}R_B7E@;L?WkOIX;Q1*!- zQBPO(wZwzD#Vk{(+^u{BY}jlv2z5EwW2YEy$20RDCr4(kuX<^UO4!Y$Lr9+7#%J=s zO?J2~ebz?9ejKLkWs5cmr^nnR;XBx%9X?~FEBb;8Ez~gpJ&U(2v}UFGsYkVU-`buK z2Swa8GF4s~4!9j#8aEq1q0k&Zap-MXkXxHn#`|EAJZ;;e?A7(|cLMZo+SFTty?Hf% z_K-!{i1YQ!C7~DfL$00~jiXgwT2WoA><#u%LF>f*aLm1Fu%{#&Ok`^?UzhlTlXkT$ z7_NYP0FILz3Tj2wLSbTGOO4y=0wf)s;M*Ywayd01bV8%JboKF=HK| zJc8@zgPcNY6QT82L4OwzlXv?y=VGO0eO0xdo*p_Lx@ZqMxr&O8EYPR%J43p=Ku=#U z7j$-JW|Y9gr6tWaGeu7T)SEbaiXmDWDt^{5f#+kHrp-;w?X!GfQDIrzXc?8r0x~!W z?wBAR*Rsv{fqZ*~>wPl_%OB9{_5&8ng26IJi(rh|5hNhpWCGz1eQl*v=i*~FZxdXQ zzapdVdX*nK*;YI8Oq^LK@Vo2{!h?d*yt$qTJD4W=%>Qh}$1);l#MC z>3hMMjT4{72kphgwr`7vDZiuU^@Nyxx5rxJw|Yn954&lZxeBIakC2Vyp$%urtuZhh_e6&de5J{i3YBd3>o zc2$(x#S7kS{3HY?$Sf3Sk3a}tQ>@lUikWIAs@+au2^+k1V=k{m#DxWw>-!Jrpbp=D z+kPSdf44Bq8<3x$03`O)XAIcR}p0beUsqk3k^!7rYdx z^fptY1NAlU^0$F#)Knx!hF+;vpC-e6E_>!w)b!*L+ClMU%}Ty4{k`Hw9#g;Gz)kK) zby?Zaunbq)=UDM{!-Gp8DbQ7df;n#k8BT3t-Na}fC~^B7qp!0T5Rep=&EEjLy?!da zGSBBBw_-rP1Zt|_71R%wl(>lz%mAi;niAT=b|j4{XT8z1ja9tLeSDPr>QFmTo3+b?xbCQ3|Qy^LbfRLDeXYa z5@Um(ui?@-Vy|2~IG4BQc_x_9Yx*zLr3>vr0{V9_H}69;nBS4&9d_njmxAvch(wIP z1)C#k=;Tn&&odsFunR3);>zdUly1x7f&^7LAvB!)x!*n5wbw}RI=C)FxO4lKN)}k> z$Pi(FGzEzD7Hde^<=6Ep(dhcnhJA%mX?qq3<-22bxTz$ttt%~*UdYo{Y(N8;6~D1* z^R_Pi1kP9eV&28lKTm===y_0o)3ZaeCqwg1r!8Esq|_iA&Wxr6Qj~M*@4VeWsREI& zZzKI$0m=wX>UZLi64`E9FH|a<^BK-7CUc!~#gGaL>NlD7y0hKAM8;mgxt^}d#PqtB zajG60=GBy>Rz19-c9}JwlI_f(>l3}ny&B^QY`Zt9Bq0dZpaR6tj?WjMi;0OVlH?PA z`yF^RLgUEbpc}}0pgP*LNPUb2v}*{;QzuI<07KpSw@|9UgKm>^NN&zK;lt||r}rw? zhPB!kXaaZq8^KMF(C4@!z_8X_ zOE;a*sIV(5y8EM=QK3v!);o^AgZ&UR>-%l`oaz%ry0vVU4sYG<_&sF<5ETR&eqw_3 zgzHDv*WjY`Oh$|!cCR`QCXg=+ld?m7my}9r1!gbi^9|FzTCwAed0z^W9O?XoIHB!U z_F)w{i+$y^Gv5Y&PCYyWDwmNexGLm+lPyp4f{{SMR5Y=NeW}}X4et5XJk}C7^&Lka zsh@f*m(w?GNv^yC2fB&$%i5i~s~)KA=m^Ox^SH^=j_jMPN3t`oEnuN%B-xM=6%uBa$Ha9TVl_##sK z$U0;!x{4*&n^54LcT2dwaXgaPGa3RsB2!pL%+_o#rHiNrl_O`~x}}Xx-f=45;K-Qz z;UI|Mpuso(C2B_mJ4r=*2zg=vj*4Bi@ zcpL(t_muD3ZVbO}4URX*>+qYMdVcjH*(;7JBs#vUQQ=g6lmE?%iYd?da^o#+`$u1g ztJ;f0r@q}3Y23o}Z@}J8SaRzeg;>J1oQKOL-CGGGt}Y2zHe=bArc|7wu|iT#Th7yL z!iS!RnCKv-bXlqXQ2g1>gF>83izE!kO&5WQHaf0{m}cP3jnFcCJev}&AYx2qwO7{l zn6|;1jy0a|ROCyPJWuIt9>?uW#`QSmWM6lEX>K@9C{Ap!rPhXw*Q^7ZuE=M8_uz-0 z`dNbR-wyR7kLRsZR0lZ|sE(GsR;Z0d3shS>grEHw$76B?mFg)sGvZi>ppdW>d&#zq z=1f14Jy2>Pgp}eXMtYT~{BAUh1}{{GmnvvNHl%U;R;oZ(q$j0h7)CP&Tlt9kufDPGbg-T3LBw z7#rf==SO6iMXf5mit4QF>T@Fos3>lmU9oo+r>NLQb<&Fh8EKsv;-=i%0#_n$NiCjs zBsv&6(OUn8-A3yx36iw2!D1dfj=%tjisa(+m^r-a6~eJ5k5Mgt-XrvOO4U!XR4dT& zZY-OlQ#>!Zm^)2?th!-@XgJ`?lp}0u=KRi{FLEiqV^7bx;-yc5g=;6iDl))?X4|ou zF_JW4dM;*4!sEc(`tZyTKR0&A8b(}0SBkl~u~x#hY*7c}wDb@^3aAvy{PFmjBoDP;chnf{use?B7=uMz-)ynh zEbpsc+i+eV0Q5DKVZL-%kN0npaZ=YsKzy(n)hCf@y=#ghAH0X_%56*+_l3mpVqccJ zQNP9n%|~3TFBMGLzj-ltDk5xL`k1d0p-%#|P4Ze9-Z^SYpK)%HMIRsWF4xnJb(ZTG zH_14j)EQ(L$_yy_6p9p)<+CA$zC@|zuJ@(b*+_;I5RaFK=<`-*N<-vbrhP14x=@t= zt|p09ZGwa8{r+HkNz9lvLw**oqSJS23mNJ*1-hHFrN~Y$xilIR=RJ|x`nrqzl^yrk z&(YSKvFPza7n#gn;ni)z7A)t^iS@LY1SrzN0*};=Q(@KU{`B*liU%>19h(XS`O{Wr zHbjDq`l0ZlY^C(;%%H>M!$w^~_*-@X@bwr1n$aA7_b0BAeA=E1P{eoK&a2T@n||iC z_(t7>p>2OkjY_AzlQh14Q87H(AvscPR>j1-X%Z+N(m`CAYBJrJ{n=`9~z|M3meOz!sqZ2Ro;iY@K(g3R$y5NWJyL$-y_aiGd zU*zEmbG;Pc@}V!^Vh?)D1LS9cg)Sm*YE|%69?wqu=3s4K9NQYLu;$G7Zl0H;pK?sW z

    MQ!ks;M|G|sTM<0u-8(NZ=y|Z~p_i})_yoH5kr0R}a0=JZB`&*5Xqc3SheZ|bC zVUrlZ?P{=tU*tTcd${FID3aFC-i0zzQ}1}f0%)zQ$clgnr`2p*2rrI&LsXqz+uOD= zSc7m_=ajr3Bk7RzhMb4bUKf*poUC{q3sRPHe1+15Dow2{W9AP`W??C;ElO^Qj_LP= z@U0`^DWCFq|2|H2+U*>L9bZ~KJ=Z<=niL44-tO(vSg;&DpIE6|d|2@5%0#r7Z;o8l ztiL$zd)YSKvVnZoiJn?ow6x~aQ4h4+d||iW+&f6;f#len>gaN-tr#WUO^X#`NS|#d z?lx7HV2U(#ik+m)_3r@&z4RVMN6=4O@m?ykOx^v+s-xtN{|Lc|dTADo3^P@?h)>R*>CdEY+lP5Mov59QUE zuK7F~9F$8xY;RFcZcBYAIsE?i3DgIq-bom4@_S~tt&y>iL+xN_w##EFTN@ivqcRr| zW13z}{7iaqO-)HnEm*t5qTio&J~(sK-`AItCo8Zj>s3=yth*Xlr=S^Y4E#Lb}SP9k3$QreU?%6h3*c!*_ITZX)g9+oICOaCo1A%*$}POQ&GQAS620f|9T0ZB zl7yW#dsEb$fjalXhSv+=x`J$?mN7@^_LfFlTcX#8=fX3Uh{4)<`gVo!7U$o3wPS z87dllSe%tfcpuVmxfi@N%7Hq9+3uQHb&I1&iMV8mRaWqnSI*pey;xTgCcc8DTt#n6?p#3mk z!O?u^tamK8PVev>0>@FQpss(nCENQ2^v1_8c||D7UchN;>y6Wk@BK4z(FXJ5GQaMA z+g$Jwd{w)5ltlBaI!)HDGx81R!$McU(im`C883gXdDkE0bpHk4yQ?oZ8b+japt9vW z9y-OG`&C0pGT3Eid|vsJ`Ml_p4^dbF^%Dh$T@A<{-e7{1q&Fp0Gh?B#Y;W^RHbt79 zZR)F*c*JNzrJK=3+))8-u?qp+&jMxkd3>+%33=3Dx{>Q3{-eIF2&fW~b4Hg8<>l$xTxJgFMH3fG8$HWVhCR z&o>3PmZ7+%k#@nJJv$%f^_-~N5LFVE*0CpAY8HC>pUS%M`BBufca;e7@sk@_CZR?f z*B8Y`bSX4LV^x;SUS+zato$LXI~TAFz9IuH*3nKvXK0wSfpML`4=$B<-8u4 zsK@fm&kJK)MlB4bbrPEy_toTxMN9BoDp@2*@fYqTrq1ELYad%L7ABVT6oy2<6qj)3 zu^gFedmcWyU4|K_Klb&e-_{*UYW2IS(JA%@SC_<~EkxRI?O3ofOP5wtJJPn^_4TRr zK}F?;%iv4QrP^^nH~^bj{_Veu3(9lUeZ7Mx5s;HroeeIwYBr{tPO=vzMSpQ7w1Sat zG%>%RLjf5bI-`FWZtr6=`4V*+Vhh|ki{?n?u%glBJ}?~7+OjfmX6C;gQ6BvCS^&CU zZv=;OQ30k2EuyL|jHW+Q9>}WnG%A#_?Np|xs zLFBlYyLiRzxHyk%&W(t0HEiD^6oS>@y-Dk|EtdSu{^sfPF?=*;mBWhi;JsbE?%g`p zn@<#ch7{t|&Rki=CS$xR)mI3`rf8Q>oL}zAd#TD9zKUH;{k+0ewQ&m}Y;hIXQ*IA$ z{Xl7l(+_t~+;U@UJ>=t3yR4-zieY)0ZxW%!&I;l`<=nTLEkAD7H!I|m#IPh(o76~t z*nn{*(5@cK=}S_sS}1Y>$;L9Y>Kbfwo+HBJI06?uG*OuAuJ^R#9LD-p;8QrpiXzZs z8Jcxb2cLsdC4@=Om}UxBL4uh3br<(M z8;rz9mk#mEFRDXG;PHG&-Gh+oOR}1m?e4Lrsxg5L)J@K0ab|HtmcxFLuCvH=>$>^c z3gs2VSCqN)BaGjH#E6r{p9JjlsW!h@%5ftM^ERH54j+bO(9C}LwKy!y$XMghA#Yjg zG0W>~M1UuoP04^i5q~f7xksU0ai)(@*|SekX;Z;_a^3TC0Nv$eX4?dttaEHOV5p#;GU5?;!PkV^3J8Ehhpo z8`GRwcNtt5{CuwKADz zw=FvsZZugcGeOT-H9zT1Jp7rj5H+&zy3Zxy2g4%e>-(7LQAkx(WI@R`zB2ub<%iT= z$8R(C7yLIWp$#A$&}rpt7Y-{+;d9J4I@wEvZgpJ~3V~#f99HG#OXS>XcDC3SP1|wS z_IF$8_7!n&HgRqVs~8;=yW@6>XO%3AL6LB+?{8{kp_{4|ws6HU?NSX&Q{fFZvU*~Z z;z}DnM9~G!tgdGeMQqZlD6vb|sFTs28cDdW>R=?qPvq+}u!#Nm=Gd(ztAeiT2(4+O zClJZ5p8wV`+EA%CV~Q13Nzml849P?H^UI6oe5k4o_~PX*jX?ZW2t|~6!f1LXu6_RA zT$R;R89$vN$OMg$l*>bbS^O-NGCvQLJrq*czUd%6=WrOT<@Z=t!zy$ft zPI{}S)T1{yf>UORO>umOIi1)z#;j&Z!BI$H(n9deVeGHIPYQUkb%vWVTnCyy!l*B8 z7Vmc)zh;V3QFQl3p)n5(RlU$J)m8gp$!7;4*IF=aR(KFVpZlaaY?2l8c))airsJRx zx2;ve#KN`KQXjavQSwcuF*Di5i*l5r6PT@gb-QKU%=o@k`#dY@=esvz@uqzsBRPlr zrC4(&nWgA4aIKd~Kzl##)V=q7X;!_5cclKUOcEj9I_K37Bkm?Rx66^AEw!7{Hw*02 z?i-Bm+X>wVAIY;yyKusAsbcz~S*O`($@NeLqWJ zIapJ8x`LBLa1prIjFooO1Q@JHem>YaMN3q0vSVCoOX1lkFxvvd1l+ zX>{!sOG#f`r#!z8f@ob6ls6`Ias_RC2zrf_7P>#D;e9hUQ_ec}+Ikt_U4HhI$U?qq zAIX~)4u3wN%ZKYD?bFrCXjENFJJVbr*{~ zA}EURCXuu;oVG1o1fZNojv{PO1f#i+?xPvkC(xkdSZvlY5f?vtF`a5V9`1%mdn?U>Ut90-FefVg`=LI?UW93eVCqJH%VPbyBys(m&QwxQ~A;qh(vGl{VgdZy~}|u)iCn z^Z0ND9I+-MmbAZz$Dgw$?uk(J-cO}u;B!=>U?oy9Hq3(~7h5@{tBt=GvI|RT6i4fM^9u^ySy9LJO6sM?ncstG+otmMN5sKmzBW1AB;r*piuYB| z)X2fWzhLBGT~i(1_r)Vr-HYO&n=6aLEbcfuZkry{FE@Wn-*5R*iCUAWG7-7r_5kKGHf#RDn$Kw2?&V6EHw%FMB2y@sALJe?%%H-`0@`?RBOS~? zNcWv{7e#BL+_X-d;X~_BZ(bt&Ui|TYRNEn7hWLe0YIG{qH`a5+V>T__@rzy2bZZW7 zp!q8o@HNh$)K3UcV$OOBd6%>N($`Uw_Nuc7hDV@zEUF*k{Ywv3{5+MBoE zRIZ)box9hxn9j^4+R)40`~jpy2Z}H#HZ(^6VR=!kShW*DsCM_) z#&Q<*Jo?0Olvy$vs`EJVES*|PGO2p|tJG)me@$&A8m@HtKX_f1Qs(L#z)^UEdA^OB ziuZCVxLT^&{s}eZucxpnZ}NjQ^K)2rRJr*Xs^7m}H|FP(nq&Dl(w7F~#m94mk1&n&7ots}EZ)Y(7&gEw(W&DC}+7jhu-tL^% zXMInV>zRS1MG#J*{K8LgB1a=;Dk(D!{P{2W@p=lLB}Y;a002M$Nkl$A}iKE!VyaR z`Q@*AkFa?R`T>Se0+@eJf z9vo5U&dGcw4KolW&CR#0_f0)?D=mT5C>98{C{@Z)4SbVw<4=BrVe#6Q~B#Ms(5Ff@$=3v5xDa+caNlGYAq_HmaHtQ zcdpq1h=If*Er!cpg*|>hfvdj1JqiOPmOEWQZDa%)T0lPr{=l4@30~A z79Br0Xr+4Ah4jS1hK?b|WU~#N$&Ylz!TOzhjJ;xCzC-02ZEcMQ53<1b85z>knYjmf z4B~Bbn%?G&bmR7a{gtYc6DgR3&@8QN!N;caz?HR_@GCE~tfISY;B1g$o&yHxu9rgJ z!Gox*V{Hp-WbQU=YF6a1DAt?8plnJ~e@Y6^WIn;2VQV&QG6_N{Nr`m9IcJhYC#oi* z)m{$&)Cz<0U|R$=NhK*l)wyS!#_1mU!_;B*90K||j{zs5-@1*PN&F_VNM&h%tX*#s z2FH#bMW=B@n;Ib~+xY3%<+PPgw8Oc1v}XM#I_ZRomM*$ouHil|UG}>o42rNilP?Nd zc9Qb*ho#HtAfJF;nFy33K8cA)HT!~d%~F4=0;Ht2lnB5pFS*d<&XZ5LbgguFN^I`!~YgvblUtp3N)K<#Y-&d_xgwE5aPGJF%Qy{QFD3;5| z;-iIMDgxn#&0BQE!FqmEL^VnCYOjeHC+62h3Rqv>H`Ubo>T*|8ZC$Mo%QnJJH94kQ zaj>ze^MRr}pRZf_f+b>ZE5GRx#utv>O1G-^r@Lu=T1~GCgR=c*PWfh#NRb|-;p(d2 zdEaH3gu$@MQ4}?eZ@~D*PcarbOV{&7mY21{popKjEVzjvCJ}Qo&YL!!q=BP)vRD`t zk+5t{Aq8?onSZjyLu+UF=xWD0)>ZATFbRVaC75_QrzGgpd0mt-@bVV%jde%5PFpSu zm*<&;K}la3b7p)eCw{P&`+YW5Y${QNwW5^;9C;UGa^=tIJS76Nt)-RId|HcxB2WrT z&pzIq{-lv=)A=FlG*ffaj2a0TkYZc3%l#@tiP}uOY5*zqI*w(@T22@J6JN9VdR)dL zT?|JMYE48w5eQXN1wVD!*0hEDt(rThC$wemEL5r{X<3~i4wi2%QN%{^9Xu>lb#FCQ z#~hTDqRr(zhh)#h-mG!nTD4X5AzRGFhpa@kW@QEFrAAUmI^G(^y>BvB-%=TuG7Xny(%|Dfgv z(@PfQP{IIz^5y6it?}tg-@Tus2!kT{oOew;N2_&;?e+XX`?LRLP%euYYdN~9-_Srx z;H5Q$@Oph4ux% zlywM;CG?+5J6}Lk>s0b}@Xmgp%S@T`#FPUtzILL^p{}92C#jbvs0A zwY@t1c6EhG7#!W-mu@^e!YqyG8h(l6)8*y-(7)@?y+l@ufak_9KnSOagF6NA8vcyD zvWP_y7Dj*GSfxZ+Y6ZMu{5kxAuh(LGIXcr)=-k~QD2MYkw1lszk_yyP925buikXQG zwJ~m-vWGHm1c50ooUif5naWSZA7Qncf9Zt8H-$mzheT*eq@AB%^&=1JG{#z45o6V+a0IhO#WVeYBO4~e+_-`$469ASHXhXhxCw6B%CvRm^Nr* zMxJrZ|KmeN7S$~2ABgNk{$dC*=GP(1`T;%xia6Mtk+qiFJQ~Zb z6+3%X7?gcz=jhYcqV-M_2PNgCDh#>>^kD(;d1D&U4qzg*hIdxoQQD>puE5(WcCUdB;t8bQ3ZnW}hS>e3c>>9Vb%nEUdo zA{2-S7JBjvETDEihn2LdWh?JvQM8V?_h0gg9LxLCc^~c_Jx&n^4HX4SWT|$YY`Cdl zgR*@&1`_=yk(}Pqlg;{MOb^OM{9MgDe2m#TbwR{bF`s@Xa|Ge-rarValu*gCn~j2N z_6BMlRm9nmBGqQfR2%Wy)@Z!9F{@x>p|zn*S>x)}_2=HrE09`=p9eH_Mq$}-X;$3U zVC;*VcPCG?3cTRz)p-q5`=hPX#wnP^Ts?#M8PTsw8r-&~Dt->b+gzD?RiuE@Cl* zS~3oiM4-A^WbA8TVNuMpUw{6bbUs%#`Qe%B&pzktiYo=rxtfAFAkNjxsE0J{J3xgG zKgt53fvT7Ms4EV(Racr^wQ*WmdrfWU#=&v6%$6z%rF`($tT*1UiFkGK%=68JLD@!9LOh*$+Ela6iMS#Hme}c2Pcah)rA)>NyyBu+ zB&kKkeqyy-qSmw~k%t>LZ&iv0^9g$T)RRrZpcINoDBJDN%D>p!jhm>94@6IG*~-<* z_=3`I)_Lcc34>DR&wadT#<@ylq%7;1MBe@U+waOMNu4RFJ1xfq964-=lAcs5iI};Q z^W$nQB{5N1EaCmO6y6GX{6a}n8pm$|$bU%%n%yA;YJFtqfL=0dhBEe_ zDhx^^5flgU+W}%Cj7dbErJcm-Ol9moi2@y`i4wJ}XSu9K_DMQYTPDX-Qn-qVcxSKG zXu{gN@~;&K)&8j4XlQ6sXX)J!KT)DcH*VRgL{W-xyx^0C9Ob&0BT4PQ`7k+j!pzaS zZa8Z@ts29NfzJJn_ugk;X{0PW$pZMyYS~+rHY%6JEdGV&`cPV6*B%^$_<0Mazhf~ zShkKI(z6Az2o=NoY!XnVkfRu4&rPI&fyNum($$63$m1}oCeezLs#8qmt*(=4jfjD4 zea40JOOWn7o*}1!=!kYxR`4Ggw@{7J^zUc;3r`|AWq*{YzmEGN@e$6)(Id$C{Q<}m17n@xi4?ujl_S@5ZgD)+I9wh0IHh1Z>IpLmZ?+ixRr)l(&x2)UB#RnoK`d~FZq|H#p795r&ZO(M09hP66z zJ4;y!*i~Vr`VLv8+zGCY1h=yrreKRad)zrFevL~tia3fOqN+{ z;D^G$aym>)+p?mB@>pyTn|@ienJ_42A|_sSZz`uzHU1d;@uOTtjFffunh0u}IPIR8 zM6O-<&m=QpP|8Of--ph>CQfE;+L87(uAiSGzRCt<3_wfE=!jfA{_JpNm7d6)wcmJc z5s_B>*|Nc7gX!vjbIKvj19K5yr_PKrHOXc1*3u+Wu>VMYTblh#q-Id4`HN7N_0?rn zu#{&$-h6GgWj-8TCMJS&OGmVtt-6-wjUDHxq)C1CzBKyv^8(tkqKwvYWSc~_zVX05 zJ|}Zc(XW>6dvjS{5I-8mx1dfl5yjM$p#MBC!b-G~$g!(W52t<{$*Wp3e+PfW0+d)E z7JBcwD9TJ2l=2ffMel_j(Q^kmdT{kF>vLVA+Zy@!%Wr*&E|jt3w3ft?%XiK;-lvFg zw~$k6_S8gBTei)(Z4ohVIX}`&7!-fU5A~;O&kXI-_3zuOm9&*o>9VPP%!ENHlduiO0U};m+P=faGEYt^-H#(XZP}Os zesupOeD9}8#VTnzXHV|@{?FjC9W71Lp^oH8aq&NWY!HQZi1?QEESJ@YM>&pNrH~Ei z5O(F5H}Ojc&a&0JJAbnbPw~-pTiYT0uGv#ZPk&NC->s>nRl92G`*oG{4vVJu&dXCG zE}hpa-EQa>${24)!<@Ap5yswQAxlgm{~kX7J~LrZ$|HH4w_B&(U?%)E8=C3E)t}00 z>ceB){rjvt%+^c!$pen3*%M~TEFCO3Kr8pJRp!+{&hVl7>S`(?Wlrf~rOH{Ha#Je| zN}-3FJBw&DY3k5Zx^hDwul<6raWZx|rwHf$)e3{sFeM_9?mpu-#Ya6yt9-&KJ+(GG zc8Idr&TkX#v4|cN0dqf#fj%5zIA!ojO1<5A=Fc{yFs)#}vaTPGlPChAo=N{MWbW5b zzS1NNNUt0D>oHwQD!|ky3$r8rabcq7P6(v5zsFD`3xL!z2$0~%WVzD`?&5>zq+zGDAv|wtZ&ny*@|{*Yc$>u zXwN~jDzf0>rtp*5w#HJKbQ*tN)by$%@K*fs6xHon%;U(dP@-EaRy;<w11<8^=E~F=KFl22SSbB~Q>Ainm{|C|cbf@= zQYPc}ML%>uxySQ!MjO`F)J$db<|}I@b!PNq519#rvLg7@DHL?+-j8`hR=I8NetMi$u>LK{zi%22{C9*QeLJU^xSWWpF)K+cW9~ryRrtvu; zeG_Ztq24s#GB%s=N!7(VC$7?*+ehSeJBN0vqTQ6qCnz_U_BfPiJ~2tDIl@Lrs0dRN z#~MFSWt<}SOwy#<8uM{9pXlnCc3U)|wbCw`2k%tvaE?)UP?s~ajG|v$ zOcX~YMk~b`90IL15pq@0O9We6*1unh&Haq-8)<4L>0d2P+TM2KwRGd0D@jtX+GF)B zYO2T4QZTh_uSWi91wYF=mdolsOSx76w5+F1v#X6ZlzMLb_JMkQCF)bs+)4!Ll*uP5 zDRiaIK^~M<#%m+?=MMoM_wXQ#>78R2SMTwl#TJ0g@HOV(2fYX5Tk6Zq7tI06`vX*qDVj|v& zplQp(cx*_|5S@pCyo=r)X<-dEqHrRo#3*AKN?KG~DLO5;tBo9V<8hR1nKr7h#m4q+ zwbd)HK3HiywtT%09cvUv^y*2OOUHv&nGYHfN>Lm&89O_f!cUIobTXbSTyvg!9=lMY zVUM;Qt!sv}wj(0V4o(d!mN>{?cT7P&hLWap&qm{0V3IC0oYN!gS>FU-N*xiPtlF`r zLP?jYt+%D)p}#~PjyggA+9TudO`IN~G%MwZqD&oOP3zNgek-SjRz=$>Gx#%YnccEuMe4J6BF$&+n5Q6?4`ti9`ji z>ky*{a#WtAzO=`7XVsg=b~}?(Xv#Qh}J~#0=d19maJeKG! zJxgH0aCpZUervg}B24zgR`Pb$7dB^dDpviT@^^4I2X943|0TllM2@1=Gf)03o<7#N zu82iRi4asx{0n2T{T(8ox=uu88S7CZHz%=BDdL<&!bvou)>MD}W4*f^gYUwV`dA5| z;wLhMQ72ay9`@AZf|STl86R@$XnqNTkDCbO5^N`0WMOT1!f>VDx&!O@ z8rLb@%lLiL2QgLTQx^D-q&*cuF(opY2Jo1GVtaCo&s3C!jz1zGiVck&+)-jZ3Fc@( zTl<+DmS}1qk&u=a#i?JPn)wjDcEt-cF=d?PYMaZd*yrC}WL7_F>pD%fIHM;4ljcPnLHQDd#vsReGnJk-eGYt%)vsb`;|rsdXp<8vib z0nfMxCQ`&1pHS3{@5m?N4DG5e=i3^M*AP|wvsKxpLtt*zx#rkv)7HekJKmm?sK2hv zwd1Ac<3gn4gM+-onBcNf7_+IFoq{QHrg!HT(d?7McocHUJY zBiRgC-d%n;3Ff953qXDPOolpiSZy>j$LvlZj%b91^Rwl2YTvvJb1>5n;A(0HPuzlxQ|-D^Yhnu@cQmGnh|SM@^y} zRnzX1TBc4#5KT!+R2D7!D&zdAk(8O)IcgJSn%N-)#`D;OaV%bmWn^Y6Q}vG;Nhpzl zT9Hp(Z_m_UGu~cZU4w3up8mM&NqsqY#8jb9w*wZ1X8t?cTU1dNRo}AUYVpM&=k_&M zTuPVo1zyIk6oFTwC?}0Sj^15f{E zw@2IHL)gsGXksE8wP(iurO{O3z^Z zhkR$r1&rG7Q@evd&}2(^80P9@SU zCfWD{wKuoZpC7(E!({wBNpmSSY(lVc#*V+mJkFMC14niJLu}w^<9az~O7c5D!}9WU zNrCa2B?4ksV%MNCoqa3GH?~o?-`%u?Q^|=*N=r#?>WDe3uPMWf+b`w)Rs}WbPaa3{ zNfc&0kM$)SVnq~TUKxA%##m#Lj#SLnc!Qd%HfQ}iqJkvdt1av4GbQzYKzm)#^R111 zlX2#HVkevxNf+OiNdJ6lppq_i+|+OyIw6Ekxgd)E`Sd^*wvC^Ow9S>AcDJG@u8+8` zq`XYhL}?zeLmLi8bckg(R@+^iTiW^gouEmP*`0%2PUotcr0KL}B9;wJGCttS=kfZq zmdj~xRU4CP+lSOq*UNd8&k>_NvAmA$1hQZ~fJHS&Op?LK7>Z)Ule-q?#%RXcRJW~e zQ%wabzN$2-H8Tqvl?cjIjxH3THKij;Q&Nj+Z9k9UDrT8(R9mOJ`9MyqDOSzn$l9_V zek4t|F5Q0anWJi!zXR0ro=!jDw6u~ER|HixQdQ5#owFk8x--M+g&Sh&(kXqEG`W*G zy7lsDq4dNxF-ioiwEJ>pIj2xGKUekqIkRjKr`PoIaO~W)w;z%kb7b5wd#iU@FQ+52 zMk*bjYQGqiQ=V_~dtA~e)4b)3r@JQ&i5sjG7w{N_T62aEtv|haiP9_Jbbeye4_}Us zbi{JC`FO+b$4)WLTh6%fScP6u#s}nv%FgSJWj+1nc21*d{OpO3?jqat zeQ@8aIyrgAoouU`QtlIXl5stMH?~xD?xRFjdi6D4ubT?DoB3!eSg$O5MULj@7kmzJ z+E5XzB$cUZhH4&%kW-e5dG($6f9$;noFmtD=6kx)XaJqjVY+*APtHTmFmss1OfiTO ztz?OoWlLVke)ei@`MtHh&$6YJW$&)nvb?e=D<_H)MTrzC<}l=tGvowva!<~k8|aMC z0D8~4RTou2BQ*v?HF~}~zv(Je-FxfSzX}DQzWUD5i>EYo*u>wA?mc$lO{v81FYr!h zD(3sXWJ=c+VwJ0n&$3)b z7B=t2J7zVLJA>)Qvve`{U9Y+vHzWAz{tMBE?kn9iEQTWU%Wow|_{+CwPSU~Ye3c?8pdxMd0(fOVr_dim&L*N-+dOQhf z`)%T2{S@{wx!<1Nezi!%KYK2_jrR9B6|>!z2uJkz9v+_a9&vWl#==pm;=GJ<=d4o1KUr>u8dWd&&)7-hroRdosuxaiTB4sK%l3ge zg?^o!&r6pGyTj6zsBK+YwbdqF{cOucP?h4Dlbd)`V-u?FeLFZD;@tNwq$B^jkHd;a zIY)#)er)XZF68_t|7t1qu;^C6!;&!EGQJs_&Z(DBF^3m2PF)(BbH7R1!=Z9@@y^P8 z-SBcANL_kZsB2wVpcDQP-;0$o_t4?-;6zHGk=~XchkqYz@#*AsKR_p zp~kvojeIvVzSN&QIHJTY8Iu+P+f1Qm0SgT$4vcUf!r^$GJ{Co_ptpFn_5Ee4`=lL` z_6K*S+DCGIaehjX9EIvDOd8!!a&qcYRXg?T#_~&3e-L00D*T(!WsMlR#Ke3e?&jN9x^^|}&T3>AL?Ox< z7NDd!*gex!-I#NiXU%pJ+ET(*W-7X03CYaxA^VIxQ|Hwn{y2J|j!6xT`T2?(b zUA3&s#|Mta>y%Yx?#I4f!mpOsuZp*IwO+*z#O>_X?Bm>ER?fdr*LGFM)qEx2UvQGM zaWRFv4^yCXD^)JKPp#8+N?ac9Gj~BIzHyl;wg0BGKD!g@Yzjj8!k&*aZVam7|Fd~dw2ea6lC1}j(G z#i4SaQle~_&&~4s74|9@biV5HjpTah>zq?<;DT+ZoFmy~4Bt3~Ek{hjuB^p)U*_6$ z{H%H1$iZ+`#&yCR!EaoII$J43AGYpN&W3E;bk5H0T9c}yu+|>fG3aDtR;J=)W0u?j znFw;!GB%Pqo6PSQ{Nuy5ZXE@;Y~lQZ`{P|qjE~d!p>~pdh$8w;NXS^;PPD5lp?z@g zKGnv>4oG#MH{Qb(W}Mddg$etUz*FcVxvDxB%N+dVMU-N!H)Sd@GxDN?btP*gnOg8V z37x3JCuXVoTU{+8)QGu*5*{?yGGv3MLujUJD-JOU1v$nEDTjuvr@>^~I$UI>8M`%I zAF^CK?(EynR9~+XaZMM=Etqc^8aBqBDXgXL+rB?CUb{HB*c1mv_#0;S)y(}2$1j$g zh9Vq_=qEy(AlS=7zcu2Able35XP#ZU%01fDL{9C|L-GUPLUNK5@no1mPkeN_#Au`|n%&$@2d z=PRnMI=ZfXDoiD0<=8L_zP3K>byCjrdv?%oY_mEKF-rGYsIhvR%Bh~i-}!n%D(|2c zu#O)vz7s7n!;ZFbh}TzsC8D8q^Xx>sQAZq*v(kid)#H7)r!pofBHjiA5e{9N`m4Fl zZRzsxO(qt(3t2><6vw47>z$&2T!?Y7c9%A^VIlf2WFLKl*88{4%mbNA|9&&ft>lo) zcJyk^@U_x!KeyB3VE+*RI5kU6n)5`jueFIj(nqdEKC1OtYi*%yb8WBd`E2gm8r%Np z_4cV8e+jQS!UDO^wvSS07K@-we0}X@vGD+h7S-aQ&NsNq6#Z5$@X+I&6Ys!D>j4N6 z3#Etyat>cVrc=>EpXdew4i{U&59g#RBWBKJ32iE&JkO`R#qQ~!Z}zJ$u3lY5w{A4P zYvtSLdUpDY*d5`pwdZ<`@0j`al}|fyr!8XlS22|SA)gv%wWoSYZZY4malEQHOr$s& zGnG19sM7use_R|aL?4;9E9?4(js27Px+&b4miqS6KWg<;V{_lx>*)Dfn^@XEVzra? z>*pEw8DxI{xpvyL^ePp4^jP;X6>?OcH>>8XFvYt@t8tb}4#*_8RoWe3SCw|KmAbQwJce*2 zzdPl=o_vBE*9jfg(xLKCX1;bfY+_Grtdl>GbGu<5mAYneaL&DaOJuYw-n7)!5T8#q z%~U4B#=TvAPGqWUU+3TlVWOi$EIu0T?d#(66A=}fRbJ;B z9yi{WXFT(7R^p&66Q}6ADGtguosI0m<^R*YOo*5>V2GHd_1WM5dVi$dgsDrEnx-DT zF4eTO)_B+`Hoy&UCP)UnQw`jQ4m$wOdfxx;;_5(xLsTv zYs368<07yg$*A4Br=6^qeq_PQoblFOliCXHyuD2zL@rREk}=%hrY6fj7(|N~R*B zmk5JWyf&3uwq0kw(Q*;|bRDnlWX-f)z0U4B8yU@TuyunpQwe=4MFh=kPs57nr%Qd( zc-t0QPwMF~p&|fEvF8z7w7~cQC%F!#NO)Au(OyrjoCbfiQ%^esQw_WhqS(K$b z&y@lUW7>C7XLBMAnogTs1}y@YE_UHeFSAdF{MJo+{Rc4nCb{ zGE@DFb)AS+y7+8DpVNNCJ#msNXiJ8@{~(qeP+X#r2KOP=~{f}p*RJg&oNqd( zC9J5syi^IBo#ulZ2hEV8E7@H*&EED@%VD&2NzRj~m&;VI)1gi6>!vb)m_`5pc&bk| zS=drezkj=RgLNwX*z3x@80}E)N|A@vUiV_3Z>lwTOAfe5Ed6ypnNxF};rNlzlrb}8 zTyh=#x_{LAo!4x0?-imCy^VBJZ9Y@Ii<9ynXbNF5NxSU6ej8` zz0~TRt(s~l;K9x#{1APFmd#$K9$c1{I&A2{a|h#_IoxqX&FgjX71Lm*))f1+Gcetd zhRsl>+1#e-+GBZP2so-*RY$P^-$%)e{)oe@L_$@X(wM) zwmp&gniH|*$1AJ0?(1dRuB@{Qf_fa;+N4@vR%2~58b}oUlIz=r#Iovn39B6y%~V!r zTth@0loIOMq4Kb~n%QA!hT0Q{p<36qM@`jy&jB^>YPmALY`V;7F)kM3W0|2ugNC@L z#l#F%6IN-bVN~)DrYW(Y_4!g}6KK;^pR~BF%D4|KD!UWm@NfZ zHa@Byi&|1*wkU(?+C^A2lHzD=QNp0i7ord8CtGotAx)*;`M5cq9Nkh%Y>3a>ERMT~ zjis~2Xr~(UAC)W66c@+LFri~+zCn*IRIGcwApi#6*{zy}*ts}i8q6xEr8sB^z)_o- zP4BawMr`ZK*f}`k#6j87CiAI){wY|8YdO|zImZf?;xpBH^&-We0b4BCCe3C`E@DT{z$R(AFZNUU$~4E9~temg-;{{L$j+DVy~+C{ta=ql&DwL}! zVrS}Khh3HMtakTia)0Swi(60^EnMr@m|6rxyb_0+lyG0rT9IW=^1DfXJSm|pj+~GP zEoNvLce5pG6{3$U!&IwPSW}tfG1#4wxwN&{ahYQe=a9ol9Dh=D8+$+1i}?eKAE=Nr zdRoGX#wfty7R%Uu=V`W_V^vFxr(Q>o8BfcNap+MgdbxTrP5zWR+0nz!J1L`P=+jg~ z)9W0bGCE)9;25)$A}-BbLcOLMPx7+yVM7QK7hyaxQgS7{PH>?0cLj1L(rnA&B`&y3 z#b@R;l{&L*?dlrJl>jEo{aN}L5J!jkJix3u)`JPsHe1mG=+G3z&*VM}Io8Fijq8Dr z_i1!!oK7F%YX>`;+g3=pcVo-qlw7df5Xz4qFn%bXIbu9W)h6QLG%F1))-gGG$hgj> z8saH4XJY+%gm_KWuUs$PewMPRn=p#UvHS!>hJ>9D3TRt6nWp0+x!@8cHHr5*7YkV2RYF#(oRNRgt)=B9Z$h^P3 zHj#C1iii_y7a8A~dO5G%5Q~UMIK;AWTrrGR-aVX|lipvcJI^~;CJAMY=^i}VbxbWkXgUJd%vl+)w|Le<@|0Jo z9D|bYuxqG?w6M0;%v%_*JPl@AYmJ+lgDeEgSW^`VquOhR+dZEUuc`XRvZ)bp*0vc9 z__udtqOq-Lo*_xqA6 z9I5#kEMYwJEza7d@urdC)A9LKGtps-W0|d;uIp;OOwES-46#p(gszs$brp84tx--w zacD|N)5!2C4mUWa*5{jVFy8qEn9<^CY;0JKHPt30f1-p_mEx$jgt%07lb1QFjq7f> zdtdUZ`0G%QA-2i_NA<$^@?2-^0<>caA$D`^er&Fa6Qp%g-Y`YJ39)o+&fO6h3HN3sZHk z#ZX!2V+^Q6NzQbJGrm81n4xTd@yJy?Z(^Qo&@?}lzRM9u4cWw5?1-NWj8o8Yo zUv|IF?J|_zuNm)E2_tD2k>ow1MaK5G2RUC|ByBJ6yl&l4Y?(Lty_mKkuke~K#{qsA zp!1~h!#Pv^lM%p=ojo>6r%W-eZB=E$rWdc_>j)RgO(^cnVHOpnEap&p z)2)bOPeOW1Y2TUp!Lt6Y5cTnYDeJHHs%atrMA^C3al$yht>=-Ot&n4yu6@m6VP3!Q zH}4tNx_j2)&a*j&x9m;-*(qTp_n0~1=GnwR*+6SO3%4AiT*?;CshO#4**BtAOW4Pj zddsDjsf}Kz^BMNDXm^tHo90twlyjO4*Q4a48{^QK>584l{G@%BHY@qM4wM`r> zWhc>c4j(FIe@8nV?KXaQH*&s1ox@SBoF;#+ShRtKed8g-9VgySIjUW!_VDM@1b2&=;il)!CA{fruoGxaQrtJw>x&iT+b)FF$xZOyQn))N5DERaU!*|(((3TS z&UxnD&QWQN>tf){ODQ)NIBR~cFvrS;cc!e9ZlQ!QRtr-@rI$gaWAA>%uRd+l1?&9s_U-*}@MPt511 zz!BpispJ8UZa2;^$Lu*I`!^%S!OJ(S=Nt&e58; z{A{nMc_WI$Zl`G_oj-RD&7V74H&;{Hwss(H-tq=TC3$61k!71-H{J*$+HAOFYA2MJ z)$Il5lK~tom8X(e_o)M~z2S#sd%WfFe$tobyAOSK{hHOPs+@0o_8-*E;!_!~ z&Y!7sH0qUooIv(FKYEyCny!dLv)vSr?GDa#^)L42O`BR2g4lSJx}N7J=q6GV88>e> z>9wH_qEP+U8v=aS8>8+m(fr6i)T{`U=w_!x#wkrg@f}VHC#R`WuH2<=opJp+D%t! z<{Q)v#6>8fHzlv0h>;S8REic2b-hAQB_wLkmQETyoBFz#ZhQ3`jE7UAoIA|-cHg9G z0|$Kk)SvO0!Gb$95L|=1yVF2`;4Y23Tc7@>>fAebYR<>`bN@}%TU}IlvERMyS%ai;hmBtS+i_g)85yMP7GKfLk1;i0bpssur^m_EsSPT{0$&z7Gk+=m2-VQ{ zsfqk1d#k=wT9Qd|I(d)hC%eO-GxwR^(4YKJxExv4H_nYbZ;y$z26pg6# zaOg?{$i+OC6Iih9qV^lQTFB%tx!y4P(~d;2ca7*?Ahz^~n_rrGJhsI{S6>EIMNw9b zYqsWaPV2PDz=25ex`7($`2J7D7|f1<1_8JfQBuKKs3f9^#=v`U&thK}m$8;7-_6g% z3yCq!q>2x;A%>Rb#p>1}qNflU7@9hcQ7exe%JsW3h zOsw&)sFrPSdK0is92Dj`aT>t0;c2m?O4UlpRcEI-%h~9Vv+lX+BZL__Ir{mr{&FsM z4$JoVIqQf|PT|C#v+U#SD*WX=hT}v>v>x5SJ*W4b7HW~@GFSQxL8NuH(^e3%mPY$i zuj(dyMU5V#yE3{GbbcofIUmNFxV7*;O}-f={UiJgWyrt zUy&?xeSpkn2KGE#Hpdrua_l(=HB|q(uy!8jI#G$K_Tf0buuC=Q|?L^sO9OapsU9{nBw8Mfl)ky!J%*#W@;oThH37f3t5Q1xe83F$?l8 z`sL~(-`!h;VQ=iQ>>G;Kj-j|;t+uU3^`;Kp0nTfKl(q=$fXWcwyb#tjJ(is;+;MZEwHh- z_{3mDc#fy{wLF>4)qB@#;~3hJnXX#gruG;c@AR}szya9LjxO1+1~^|B!giLOW(#~7 zSwzQCpB{##j(b)FuolgUDb%ni6vGL zZw*8+U|B5jl@|s)>#V7$w2#W*)K5j2$;^)acCae~>n{RH4}iQr^4C@#A7y)A3>&+xPHuAs2fN%>o^^B^c+5Ru$f4uO`M5sHpvOOQP>g`c$<=m+Rdk4iqIj2u&M&1ct zw^uj)F`Yk^`X;lVFOX*Sm(v?F96ama1Sn$`>*!2Kl|(Urdsi4{?13IZGk-qziT7#3 zhi2Z!PjZf-Ebl<=x^&`&QE%ovV>e*+Zb*H#xW)4LGHDc z@KX|LL=`%(gJ{3TOqjyrzrKgMsMzV+O{Ln}MGr`B@RDGHa(Q>y3pr{(U2jeT19aV5 ze#)(;*qysZG^N{Gi~T!M1k7lU2D=vSY|~lHtsnMal_IOVd1$q1%|RNsr;B{=<1`I% zY+9ffIEz=DC?-3fO-Ml6kI!ezZhBm7tWDVrJIyN$Z}~oh?Wwm*L88+?@agm_A?$C< zkT;Kqj{&RTJIXneE$oDT89fpeK-V~lrP}?;_H1RRPrANeZloB)bU}{i2k2Xoz?wSWZ`$&tN|<8}djy8^t|Z(Yn;s{Od7k|Ic-O zopYX3Mf*#JzW44%ZStGGxhxfipMHmtp|hTSB~WaYdy&NnQb|$Lu@q{z-H4#G*0Bl` zM_8D>ndeea>hcx-`iCTwhgW*-{sIGsn=KMH1h0&EI=+6u?E#jrh*ut7vd55f-kC*O zz{x_rmW{fr8OMV~hSm;uo`dG=k!XNNs{Wl8wcn#msMOuI_FrA4N)QFl4H{$=Nk(^j ztLbo7aPcLV?99F`NALt#cl+!o*qId|*5E^#Wuf$UwE6BTW+I-j)xm#*<#^?Q;!_b* z&5M?x15`CSjOl|geLGwmewFw6884t~H@daVa>xmrpR^c8wJa`%_sebG%GFM`Yy9?f zLXQFM{>MhH;*u~kgD0aR=k9&=w;k{SkY;~xr|^jEC9%9w{3xAOVbI4_lcH7(EqBHtDYS8^=@S1e*hj_|A{s^#fg!O? zZLbh;@|$BNVDmly0`_O5m^4Cb7k5CyLFtHP;^4ySXBp|ztT z$V(nbH)}uK;aR$)BjSvD@Hm^~I3hB#$DAz&LUpOzWJDAy@+C9je1QKFO=v$iyM42W znYDKty>XNGX)v_aGR#F+dTFf_&`X?@L`Pjr*trdH4=oO>tUURYV_^5`7v%bq$vZT& ze3O4hyr*mAqSt(Ay0xOb;GWZ>yaGBAC`@RbJ%80|9fTMd02yVT)?skme*$1~x~?_5 zKa@#8vN~=bde47c=v>`aKH?d0No$GW=P~o&nDLGoeGxy4|1$7()1%=jRe$+0-rH~I zPZ@X2>U)k#2o`jPdQ`=Jr4kR1Xl7-w=v2>#Uag64_AsUGCbUQW^ewm!L07Vzf^TPr_<?aArkWo$2#EYve+&g0NM0JpyDaGO6)D2Opx%Hgr-kPva@-KRo>R5Xd$rMX_23hy` z&yJ%Cdpeabj?(R`H3#s5B+`)UGuN%X{&7}NzHjvqKtNS@vbaoaw10K;Hu93Qj}{7!_FT<42grFed8 zJuWl}@AzCJvwJ^j;80_9&U@{zU6{?P6Z&)52xiuh_nll{(>{1&Zbmz(LHwoL)fu0O1o6tXswn>4J{V}h zkVW>rG!uNkf@Fjl2V5=6KP%{(Mpi5rOGY3#N{5k7@ zgw0`Cl$_3X{t7>@!Yrp>`swXWL4nXY`V<`pN$SiC$~4GPJe4QTzvdHl7Q7LJ(BfG9 z^s!hiy9dgW9UQ!{?1UK;+EqUI<8(HQfim-40`4`i*imR}WPx_?d{6I=vY$RFL{0c> z>F-`EA4>>Q)?)4Y91}qxxcB%*Pw?uacP|&so%b-}V_|`ABoI6vS&E<5eF5*MLa)?D ztOoOCfBrmcPIv7%cRn2y%>6t1SbOLI$@=Er-pgnHj(XT7I>~m<{0}by5RM&dn;FI)jB#Ql-fh z9*TewqSZ}oZcD}B4E#WYeV(O7XiJB>{M25-h3st7*S%*U_Wxg458qg09be0ICLn@%J>a(&Y{R}d21L{D>avj&0yRvH+Dcw zu>aJwl*Ws(%#g=$w;olL^Er6MyubBI4{2|lm!s}U4gQIxk%{Q0EcT)0@NdTKg(@PG zNm4_~aa*e%hS%B+veS?04CS65WuS85ds5eDga)wOIEz^9pNOiM# z_VRvW-%d&)zE=U+zxm2{JSoKmbwtG5VnQDi_n?7@FR7Ti2Ur#7vO$5?mQ6=jP9oc4 z;G?OPX`xg(J$-xmkyz+ufmsz#J@VE+0*5nMbI=W+@-ZPn0-Q4pWOc2Wi%gAG>8 zlV$Y3A`vZ4{M`tG2Infff7`Th1M-F<{X)OuRL2KrgwK6pYX5X-84bU@sOxjjhI9QM< zL5Y%GbnQCZ2x@EI8o#8BV84+4ePlk9++5vR3)xH0xceQ5mG`33Cy8F+Myj>n`tMAc zABuZEq`-UBgx%pJWLZugv_#X%W%^j^n0>8VX=L({;g`e^9lj-phvZ_U zr$6ybiS2}hda{9z#BWcFyU0Pq^J!}^qmeqg-*r$9g=T-^35}_6iy?hPF!O1ZY?T^S zXgFu(>zHgrRYupN3kO;!shpfjSPTt3Pq2_`(J@1BCGvTte9keama{MXr#_p3wYolyDxjUQ9=u0|?b zB)>*SSLry8jV(|`H=?4B@~pWAaNm3KpDn;$?ae7nk%qSWJ}njPmyl%T4489;#%w#{ z6vpoDJn5kMn!lCOoKoGXh?yu*!UJWjc$>SF77<;$E4YOPX7hZG+Mv4gJx zvYojRdjhF@hsi7;J~oH>sVxDNo`ahp8gbwqsrg~~xg}e^A*NJ3QbXfV4cF9L*%=1^ zRyPs3GCRhEZnuP!X>hN5g1M073GF3lyrQ@AW#?`B_~uV;nEE`2YizfxtlNpS8C?*` zaCh-&_zju0Ek{M`pVQ}TZQ;(%OPLED!JF~v(sabqnu<<7JRx)Eigw5}^4+EUZLk$D z47NRj-%ZwM@4XUzs*bR*3vRv_F(5u+{q zecr_CQsS3JUEX9atr{O&0{zo+EB<6#?&KvMz-;`Kvhh=Xyrp04E|`M$Qt zir;zEtA-XiUaLj;XMjdd=L~!Iix5fZnWw6CD&P~aOdeUDiFg$HOYu(0L#$c6$Jvy| zvuj%~_nZb(s;@wTcNtF0J_>4y_t4l@W4eLCknsdAjgf5Y{BP>V6d5Q9Gw?4?d)dc$ zSsnVm1nzYDfm4?_-Ngd_&LzGys-gM^LDr(WeZpTbfD$2(X;ZMS6-FFy#xLyd?t1Wd zsHq#lV_mY)=*vz_lnun$-C=cf@Mg}}_O_N!?M23F=NZ#U1Ei-X=Gz{i@<-099djBM zhF-JyVRE)pG^XQTW(Vi75&WXZvy^U=Q4;FXX9dni3?Xj`laU?i8SuIaKNOM|GM<{8 zXx{C(y3;8BqbJUhg@1s*r`jn|u=*Ycj-+j%iIt}ExB%U6Y$FyrUv_-&Fp0nvK(bQT z<4LNW9O-!nZQ~o{{KDftCyYg)=Gx-wk}E(&jxe`2=8|m0f4IrPAb>!i^^s>~Hov57 zOgzAMlw=CxwLx|fb+@-v>6_x;1yk&ZDIy!nU{`LRFSOxJ4pT^P!&A5S!uBYo8>Xy2 zk;U*(@7ziLZKR>^nx;pmrYpSy;+qmFqm<(A&LLmG;Gp$2yKUp+Rp$6w-U{_v&V)b2 zrS6v#S?0dbN1;$b?_-HQOn21JDF6n4Ya_2%@lH^XeGQ3531k-BAu=e3VCL+@Q+Q#J z@~AoV+%+IpEnL|90d)u~ph?brt9|t#@mYot)@37C6On*Vf6LgfbWC+x$b3Gw{Rr10q;Z&pbuvGZJF8Sbl;xl}M_Z;fWNiA| z_*0xO;E7EJ`A3Uj;hwk$U_GuWz{p5wLP?PY-QFkAEwhW>e4hqVvoIE=rnU;KHZjz65)*s=a zmjCps(NjUCjhLANJ-v||sw z^T^DKP0ys6&sP>Db#Vb7FVI|}^x%EKfZD2)#Mf3&1b&q;!I52%nx1@B$-!9m3(xq& zr4ZGv@u9YIwKbgDiXT@L|2?IgBVE%fvtQc7G85hJYvsCDnR>Z6TO7F)>Xm-lfr^^-mzU2IV7DrkONa4-V<4eLy!$~IQJiY$!dNZ3?_ zMcY?=Zo5EC3@Y6cTbZv(_S+vdf{R!pBjZJt#ClzGHIcz1I?cYc*N&clayBq#Q;uq} zoHEwXKZ!EJe{~$#bISLUGw{xadDporJ@k^;noz|@;w3u|Mxx^ zywDxVUrL0uNMRP3zC?Et5XUGAMOHX7 z{78$#8dmZWA{q2@DJPM!4!JKHQ^LVDr#jl(YzVoTwbt)htI)3g*)N>!oEm+ug7!rp zVGtWvA(PP33^<$>F;ZL-{jr7^68W{2-B=}xh95bq$4gT zZVR1yE4{rK)#^V*m#RLGeBoij98tY^AB`3qWkRxfsRL{Ka-n))>OmVir+1dqe{>@o zP)nr;!98?lMm9^)*QkWnS~R@7ZO1$sxc^+bJ*d-M9W8i~jXTAlLOKbg>?;UPAT$!G zh!K4gLCW33wJDyLG?|xjhBAi+*(WbGnfk0p3wF7VmqXchO;V(8-80P~$B{N5zX-cL z&CAy5Q`s^Xbr!l<2Z(L)b_9q;`Gh&x8n}dBlg7_fmOK6NcBj9tSrWTZfUSCi47=$H z9OwEWbe0xU_mb`%-xf*M0bkx+6Ww1jd^&MSR7UKB9b+*#>cE6e6O7c}^|?B@ zQ`L1kZiSz?pWM+9t=RDUI_~@ik~Dw0e)K~KRlx3WmJM!gpQ+)d zU=nwNNZAh_p4wGcf_M$o4)duPe5Gl$oSab6#AwDI_jdV>4!LHh{jy~lmzx*7Tzw(v zV0A+9Wni9mRg;ZJ@JRK1*M?EvQ!~pXf%@_`cU1Vh#2B+r$XIjE^NR?A)8DXhxu1;H z5ye$Sf@=={s>g+~ulSa4`|*wS{0fFmdh#UT{&@-}wk+h4IyD#`|NMQz%AfiEe&cNk z^KpKd;@umAkHl@igLJJxY9-W0UmP?w@|+LCO=Mw3oWh3jMzr|b+~u&J_n#SOG&3Gt z3KNBH@3z8jvJ>|x`(O+cPnyWjYdg{>TIH5wnn+u<@{_*|oYFecMg(3j_*hKUGc-Av zho*&vXjGJ|@?fZmFOgu7ah-cHZt*S%Q~(oeIyZ~AjU2h98#n8|^!v>JMVYPgy#5kO zg2Kv7+i39FzJ7Ra0Nvo78>-OGx3=H31&8tv+@QIY_QEqa`qnCXEfxhe%A=C7+#Z#i ze*J?oqNBfrN+&|Y7+$9wzL)Or#l@OA?Cd#kkVN^q?7`^JD!weg<}U895j;2xeC3P2 zuN^+#r=M=QNY48M`_<5D|L(-4u(h(g3J4@^nF=9nd)jWhc&3uPiE}LRy*Iq22^HHi zVH6)=TzWc_yXNnhL95*BV68`Fu+f(#z}1(g=4Q_dV3x)yT!Ay>|Z(++j)kAx$rwUf7G9q{#nyTf>?w{ z`s1DLH#LVA!YheU=Z~U-f#R+ux?Uu+4i5+~MCPsF-Z`~fAOSZ!&06Te+ISB+#2)SP zo!--G0{Lc!$MhNv2J3W;-SFvdT0r}vyX(&w`ZbM6xcgh?6wH*+Qf9B%HjO6^p+4Sw zpSl?*Rd$u3#>f!C7b4b44dO%{pM}?Oschx$@{!O2|j-F31&%V*w-*-+qdcrFysC*~JYWKXY zn>MoVxvV!&Z|;Q0K1q0Cho1P7cQoD9Oq`$Im)_Y3unA6rHWz8H{|uj&l#1?pMldNiYZ=@UWE>~x zZwQ2wz`0h1=&F3(`s?tZL+7x}Qlf{{`i357wfh;gWwJJWi*?r;x@-`?H8A&y){IcO z_2q$CS$xT~w>KZPS{NgiFk#l0d4<)3I3$`-3B--6CBe#ye?Di3guDL?P*C0iNDTrj z3Di;?D8ZXllt#?ro*y+&6*xSs?s*mH0 zoEz%ne57(o^@g>>wp)--EY4+=9jkjCja`E&m#t=hH!6>UA;{w55=L+966-@2`Rrwg zOS~C#P5Ngwl=k|hA|l?<0q{Q<6$Im_NxU*xeH5}-l0W_N8%W} z*H%^8n35!A5iKB0*W8=OC$sL-nPEW@$H>nJNFf%Hz_y#RhCn6lnRyM)rkIdEe3U{8 zuwwH!D>!yCFGq)NE$L^*P!#*{v6>-;+(@_}6OnzhX1e)xUSSejd1SC= z$0bliLCwJOixxKk+;ar`^#}vL>?Fo-5lewGUvB0JsRBP3{t?eFp znTYZ6w+dof7?hWcA~;7>8o7u+7?J4`!{SSroousgNeb0*Y*rc&zQ!*NarqBh(b$V( zXjq)zFHH7;c1r@ps670XI&n6{2u+BNDl&H%^LS8ptl((*XfRsDC<)DRo0{Gej63Gs z@Gk6{E!F5P_|rRR=-M>v4@}%dP4l>Cs$W{@7M^7o;ar#k$x#`%mIF3{V~vuM()(q} zaGncbXaB(5`N*dI==0u>{_`wgFMds)*KGuL0q`EmrS@rV0tSfNG32>Wmp7gRPSdqdB z`SD%!hz_U_doZ>?J*1M=ObCJqCT)>h=c3t|4f``sjzoCF>z>EbRtlCy^nDvDN16{ zEF78j)CY)cmBW4}7oj>4d)E&1v7>FR#c|9yI`u(1#k)BMmr^}t_;IZ5Wi0||zOa}q zA>94ollW#_&6F{Ov-asF{?le`z7nb(-Pd2!HY>8xbF3D|WT1?rwTa&CAd(5oSc$nU{3tU#<9&fdS2orb2V;_1R@*?IcW@9X(1T{wC_GmWTW z!hi96q7))TH;Vf$7XlfB3$8;Z6abw{_|F>-le@1LLZY%OI(;8wb;I9A8{(NP9YpMzU+ZVOu$XizeJM z(GZWXWIgR!Tl-!lKkezktX8gzR*lO7)@+faY3@iNR^@r&jOh06Pw%>E6Y65zR z`s+_D!rL%_m$2wZ=-uJ3KQ!N9&=Dffi;c!!ZyEab|7rZcHNpR=<$u)jKSA_g(dPee zY%%NS-rHs!G-izu%SDQmk(M6X2V%6QR2g`_{lCD;H!$~@sTc=Zr0DrMIdQP?@w z1no$pXE3nBftr#rlL-?Y!o7KUPJ~=`EkDc~*;TA)$NHRTQs4eNtOk9Od?VfUfxmd3 z3VlfI*P4*9lKax43=3B7d|$91#;(X~pl?nJ1#q^rU&blIq#xgAhXJPgSlTu_G9@uT zuw4hay^7QQQg1d5ort-9;=^8iUME0zVwvK>LxkTT9Lu1=|NaIs@LwL`bM!!NP5#06 z6mUV`(a-Mx@E~BQ*O3NJA^q#Wy+M>Uf##8-^X~3R;a`)mPwxQif`alNpCo>UsV-5D~vpnvJ~fBb}n4b_^sTN8^W zI=2u^$6FzBlPRf8*mrM1Jc!_RN*2xk7NP#rmux?wum9vOT+K%s8y+|zBssNbior4h zy_y0mHyo)6f$hAP_rH4hf61^k?*AmyXGG&C(pZY1jVUF3B7#UhSfT<{x{ZJqSOA$m zf|KH97uG9T-<c^7H~=u$$2Pt^9ble!d+aksSZ`$hg3z=Ioh7f#O|otDM7-*%YY%i)VI{?uY{Md(#I;*pxjlI z4^YxdO7ylf7%K|uOoxXDLrT_R0BmBeq~yg5RCu1#^ak5#LOlQ1vTQ2n)VRDrpr10R^03 zLrL8q>d&5{WB&&2|ACTYh){&G=rih*4-~eBd5(8ru^#2pG#CYm9NHOZYfGOvYb*=k znL4pxRv#3dV9UIs{Wq`?Nr45JSck-h%TRo$MS|rP+Z!4}3`ODq1`?-aSeKSc!HW1) zgGm#2;`3N=06vo-#Md$biwYuTDJ+{=Wfcju&6!gY)45+%7qdQ8YwO#MZ4wu#i(ez* zzrRIgjA#RJhrDiz)N-t`zV<%Bpe}LAF0S<=%dz(uP^VSz5Q*w@goeTQYT?*S06GE( ziR>%l8h*y8!U{^O)0oGF+U?>V>TNYHEcM))+RB*JI3M5_cr5E9E4x0tk}Cs+r#?cn z>{Tc_-y4_&B(Yp2e7xoLnU{A3mcDg=S!$Ovx{$AT$%S}5URjB7a34D=AknBdR6X6MEdlfSe35{`wbl` z?5AqJc_gT?Wa|V~x_}F9T|hzfjKl5wM*40ZybnL-dBUo4dr;tBi!c138S8G@C7B~M zL|s$ZhXf$tS|61*eSl~0+!Hes5buHx>W)16*?scw{6v=liW>S168Mi^N*HJNP!+Lv zOrVgX77lS0(|ht)GzsGoqfL8Ke*Nf8RrosL`0-B`aiOWEAT>lkbCH}nyW@`Z`O&7K zp(0b^j7p@f9FG$>IDXQtZ1b#1+I1olj3bBqf&LXc{HMGt%nQ>I^fp4A(Wr7&XOv8w zt#ej@m@6_=Jri0kC`zu2FEyGq8SwASMz|WYY9`Ql<=syLDD(zFiHY!3DzY-eQ zgCZA})8q-^2UyC}UsNmy-l_ktJ}5k3aY2!;-|vaV8$DI53!LB!NUqYDxQCY^ik7-Q zLgr_3Le(}B-yp1;?&g2RjmHT8!4gn)Usm9osGreOG+^YPa(J7DX5w&aYU+w5a(6Nz z!&~OyUjpUD&Zy+&mUQZx@UP^Bepx7YWhD{#Ndy~@cva+x!Yj?z3SWR6vm{PcknF+& zZAkm1H-CJc>Xps~3dcb)STi31wHH1}6-r2fP^jA>uw9SMMU}W|Y#!mUbisxHPWBHw zG^88&zvvigtF_{oPiQruOZ;0G*usg$JRETCrjLVA%ZH%Hgv&ERsM-U`hgG#y4g=sW z>;qQ*JwG=gLK9aTBI+0S^G#Cp_n}sfo~g#H6oi(|(^B>Iv#Q$ZOG|A%J2{xs0(eG2 ztXSr^>#(prky}UI`u}cZXpA_kpfKU!&nuwQ7{iQkSR&JESALB!_dq$&bNA6j`f5yy zZkmI!`D+(I`-&FFVE-|3k5Zpt;G2&ueImUaG{Oh%gvxd79!`^6)j+@>{l4ySq%t9` zA26?2mQ->1pNxNd-n%D4D&GaZrQRVZ-|+=BclOgS+q0Z_Zo@INS9aTb{issx=V=QXW(C(c`YlD)SNU^bU;*q{ zU}iUy>F3GmB?$%oXF?QI6j9Q#pQU?$>A;3#pHZPyx@4Omk;pwc={&$nCcmdUbnaD%!ZH5g=F~X}__WMW69fO626p_wN!*+b`Q+xAxa%Y{O z6JqyZGF{W|H;`n$3zR!rYrd@ORc&{a-P^wvl9<@M_ zcEm*O1h5Uq%i;ole&>vq~UMMWklOOM##k7bYhmqy%1!s zWa0^v9McAz3|}Lk5n0eQlq(dR=3zLO<>z{E)hs37Esuk^K|D5Aq4|$BW}I)msMZXv z{2lBt5M;)xa%&))gh1idZet#77Wr317?LIxf+MqOEM_4>tRvx(4!sLtH*X%fg5Df} zuB}>2vuhQY=@+l?h0`buV4o5RrEK%hDQh}Q$H_oIvWucK>On_j+Q)%c8NDI;kb=mY zb}NPTv1U|npHu&c`5w7Ngq5jqEIMC09|;I zJUWe?<2{DKIs_nL4A{z!BV}!F2l=l}!yqOQANSi^L*mpTxdea68*w|32o-8_Rw(xn zG}*18}uR>T=Gl1rH;rSKSH%|>cl$BIPHJ!YRRz|)l+q2 z|6M=ZtuQ)E^}@?s2Wuug@tS#weT74)FA zqw9)1wljjI{pMYs5zNx+qE+zw%&>kh-5PU_dPZO+Apqt2T2)O81C5cFpI2JYJxi`j z01vDt%a))A(02xg+9;m34)&giusxMC5Fo?y2o<==@BiPB5R!jLh{u?6wC8QdJJ-#t zPwrk?$>vmapN(bY#~(J<_yI0Mp16A&>iaR6^c(^cN?48|bA%ri@al zl0IpSIpTK4Mi!0G(E*~ekKUMq1B_tyHbGsV2y-N$Kx&XLTj0!d`G~NN2_@(kH-c02 zb}y#gD}2zf1v+R5+~@rAlbRaFf{)j7kk5w(Gpg^9%d_1TdYC~;OH0c)Eeiu|2+jOd zPYMyi*hKiBg4kr=(dAk}0>c2wDylmKbx57}ClO1GJP1*jq7R5a`gknusMq|bP+wu* z!t0Iy81OV*W*gM!Sco!NY z8%gNbZ5Y;YJTFd`AxzoojExMaXTVCD2O2=@*|s*P^^+Va5mFNp#|LS6phjvaTCuZdd75j5;XU3Bdj-~-4o1kY5gJaT_1%8n-FkuTF=%Imj7 zW{Ysdb*|I#IHH7f8GhxS{3o^+pqnnq$W)?o>eUW3Hw`@;D&G_ByQ&wIMw?B?22Ow# zhFJH1mu)@8|78`(CiCGgx6E~639QNmzr(V7Z3KmnLGv@6CCYv8D~foq2+|IX5n<>z z2}}Ccn=G$$Y+jcN3Xxx}ic5{R(25+5C8Epz4g8LW>c8{F?U12Wes4tSco3*=f9a7P z9i61%L`y#U8w(1&CCQGg4)z7qs~*t&beTMscmF3-;{glyW8nY3GX96o?haK@Vo!AF zR>ftRy*-L5N%?Gd`zk{CBv zH>Xb4^nE9DiErPebmKSM5+P<|zbh(|@!Q-~D_?v{T1qb(ahNz$KK6r2*8dp~X&D$a z#pZfH61$_7PUh~K!vvAe&d$i%FnJ%nKETSSR=nvA4T$r%OXtwXQF+;%c%d0zOsn?< z^Z8iWCc0Fd3(*!7phB|;QzpOP=ac0>KG6ChR`HOsWJWXUwfD}=BiE&uSrM_8p!)z`a@?+_H1+aAR0tC_(?8tLzNCZ6U0aX>%@B73emBguKM@?9~xPfpKO-pGlG zVEjGW(4?3`gcg4(hImWbr(|psR;EA-Ej9JjIP8L1MG1fIZ(ML|5pG}B#ESDFH?S2tp6S@;cr(AsWT4MGCX+hc~y|Gli-2QPI&;(JQSeV3hY);cWAHW|j1Ra=r_}v+Y2GL= z#lP6#m-0Tu>h$tLhB>EsS)pWnAqmZaIW!L|Zkmhb*8o2bqGI-h1mk!usP)u$fGvqL zE-d)nvDryHUsdF{fdR~>qQlYt=+WKX&*|elxE75KG=nQ$>#@d1myg(y8yi!wIb3=M z`>>)&61-(4aq0|uU*#?Icb8o=^8JZT{+SDd26q9RJIcra71+=!+^r%8ZoyA0zsE^C zRr9q4LtL_;@{xY-NEQYDYv5OviLZTjTn0j0tWdy`w&3IY2vjDRt1K9KudwKcn^Ba3&iChY&>kG3w*QuWG zJpI~Qh*(Yxyhcq3RE;@^It9`DxPs1}o?;>Iu|*Vhzq`jV90zS+yK&T0KJpaRr=?qP z$QjS|jBa-_ZGsLhsB^Q$RWP>tlDNObw;IlOlSxtK?=M8lYj>I_9keQfIy@-WemTYh z0aTntHnAZlJ{8}_->bah4R3-r$z#Uj@x+)qv9{^PF(NO?Z>8SC zfZsNcJtg*+9rCUP_NA*{NkA;zKsy3zJffU?+!4%za0WJgs+FR5nj83Ao- z!-N_Wwn}{YTi<7iw;`=iwD#QFJHK=1)gshw^rkxhRw#MWqk4K^?m>Eq$*L$erVC4l zv8kOtPc-z%nE}@gMgPQ~(@{-r%frUT2@jQr!Q5?SKmu}5c&}n#ukH7oCKdxDX z?3GS!nE-3eJ`UgCsOf1*rRdez$1gGH7|^FM8Ywvm9t*h!LzyLzlv8@iq)5I50{y zRzk&oSZv^;#d~jdTo@6fa#*fKXdLu4nJCd`=QepE)8ZUHC%feSgxkb?p$3bSB7%RJ|Cb4% zCxe@l+Tl2QnaiiBK=`kbiKgaX60~cFj5XZcT|||#eb0da`TL~`vM>tUvlWMU-jSOsI6_KW_vHhb^T*$S7oiXtNPRwAGtKk6+m^4j5g_wPORiI<_6Kl(_V*TY*b zR;WVGIZ-|Y#7fJegxFlC-`1!K5c7zDGj$rb+0HAb8>S)VDGJNUxhT}-(|VlS=?Ujk zY8MB~-CMq^X&lAbgVF`o$<7a#!Xse|=)!N_!v6OkhIDWs6oeR4^Hr|3RX3)C@wwB~ z`q`wCwk~_zXIqOo-|KOYoU@bPbNUH9ZW{2YqRA=FyIk46rYCLA(@01GaaykKK4a3q ztpy<$^q0NukMab6eqevFhFe|>c%V=7NIkw$_V&+iSo1yGq7WjSX$ENDa}PTzqMQgk z4bS5&CSM-xjDn834(BS;6N6opB1_m=d=(aJtl_o{Iv)o1z$3_d*8|Hwca!`{UcnE< zc4}p_UMu#CBQrPiu?L|3dE>^`+b6+HzNV%s%k&G0M&)LP?4{&z4AdLjVbuR@@&DKM z|NJE;zP{{u79f3{t|4-pp&%4J;w{k-Eo8$tWj|~wQRutEyNy_KF zn%Etl2JiOXt|pI6f?TMVD&;JQILniBGqhKl=b_u3{o8Q;{LD{&`F_@Z@DMZSts(pW zO7@?eJh034z3T^r`jGW(!N)Qeah;6AUC^%qnt;W|7$MJzWzYQ!|7W8}GRDaP6GPaIRjx!$70ut$le_5a1L=hA& zSf&*zcz6rgDV$DFPVdf<_*?ov?BHIW6-j)_t2hY_=Bns4RZmeLtF>qqAgW#CvO%vi zT-irJBS|ax84M7pYd^YOK7JHPX^un)mGD39M;&fpp!;ZX|?QJ%Wz_u0@@-+j9qgIuUy5;N9SvRqfw)&BhM z@@n|9!@@DzHir=J{}i=vg2Iq2ohAxobTD+1T7l6U>mF-sm$*KMyQ43cT0Y8^Tw}Lr zCy&P}*AESxoMWvH?L0H?xy;T!k5O~uHjM_X9whwiLj0bjF4s(HXr`Z6xB<;6n{$~i zK9}b=XlQc0ndy=Q6P29_3o+`Ea~aA4%{4X4)4mG*6H?M_~ z`8AOYc2YYPCQ@y4vw5fM?mdan<8tC?w{B}li1_7xrJ(EgVhoQ9{#qQrjp-<~aUF*? z^JM3R>WK#(eGpGAO8vN*N-52C_W$DRO#`8R+qdx&hGa-tVv-grdt~1$ls(H}tR;Jv z>`O>f6rz%S-wg(3W@I;1vadtNo-Ell_Vu~kzyI_6@B4Q*-xpqZGxM42I+x=-&f`2k zlY6f#uph1gU*Y;1^P&7M_}SNw+4=UFxd`JF-g zs|CRkF9jc)9THYJRRkcMH7jf zoY}3$p8MumoV(Ml1MP=@EDoEiw~2+e6tyWea-^Wc%^Lp8H7>yK1bPDotsHs}Y zMe*hDuv#ON@b>#9R;Eqb_td+z<)4o29h3*ziVqC5XX+OdhsFXwJA!b!S(kKSNkKUD z;Uv`{vEN-85k^!ht{E#$b#&~_)*ttK;E6JgEtTAo5iKIQ!pJ1Vql;lnQLAXNX6K>>4xmc(q=v0sx6V;sIeem1JCHP5w#bq7_yEzn2B&s6za(Y;tY z+=|w3qQZ}_^BE?qb$q=S>lQg-ZdT`~TtIVJ=DWElksT&{T6z-S{cgK4Xm9WXpY@?< zxr6X@kX*~v1P|h&ECwN9-)OQilpgt~8pc630vxAQx$MdV8y*}abz*99a|&rOSI0(b zY&rM$)UShzyhrBn={3Ix<$S3A z;kh#TU3nfzHqY#@+mxL<+~${+`oLS%dGIG^EcJfLZyW~Q$JpYSB1tzO^tMRMxehX&dmS;xZZ5Z>BM6 z`|E`|!ubPvqGKv@RZcezMH92kiA0e>k8k|nFN{Qzu1=4y*!>Fbl0&fb-NH~7CHxIqSe%0Ig`UHkTZ&_BL4wD z|HEDME-~))Td)+N&leIB>(-15V0&{eV+nb>RC(^b)tX5BBN5rj_T5og1@jue>u2NT zeZ`6WnbI?Vu2`%szX`#WJ6H3gMVs`_vl9H#P63Cj-gRPi2ZT_b7>`1ll-NvG4D9>{ z7GnsFRnO6>mtP<~k^Lv*M=M@~waw5S4C!o%0m2dj~c8Q#a|RDi71I6Lc`?wqY!tcl@U`AS_4m#2=Lrn^`eve|=}rqV{ctuTuI) z+^2_u>%q7)^tTv>9*m}0)I4qeGI=A2dVJ}i_WIeAJRd8i*MtoUrMx#8_Qv5e=jUMG zePq|VlQ&Hs9syYIv_KKBPz@DCTZ`N(P8Y?$$-7nctES^J`AeeF2vximT;Vy%O$s`= zaJcJHRNZP2c<|NGM!O*|c5fA@Ex*I1Pd`=Dk;RwpzKHQ#naE1ZQ7f9L-rD-@UiPV4 zb`759Jyb~TG9vQZBh2LV>EOparcy?{aoPLxvjulAyAAT6AZxb43CuKV&uvzif_!({ zo|5*#dagwCpZ+d2_JS5fPCu&QqE)jhH)CQ8bxaIm!q(6NI_by25Y=W?q57uxm~Ipb zEoy&V`qan->^AbJJ}>u!%H6C?Q`8GYJ#>a2_D|mYja#^EBE94LF3Fw25UVU zG$T+zUA5%_zOtF&#oA(2nw%&Bfen|d3Thn*m3Ay$T6Ue>@hrntYPima-oXSQVwzm;_p>4JKNq)*~(}mg#C#7sivCx?U3}T^Y!$y+p z>7U1T89MhVav=%_kLKvholwO)8)%N(E+*2jH{0@RDkk9wgX}3cWkfi&li!RA-{upHc7#iPMfBNqQ*T0^_(v$-4ve)GK z?R1*z?eOJg1sVf!kRu4y>pO*%NI7B}J;ZZqe0(CEbMhdHsC$zJf?{7-osL*tT#P|} zb-PDghaO0=O>(>P{(~b+GRIfAmg?{)BGA+2XEeV|PDeA0WnDr?MeVipyf?NmSvLuX ze_j7gN(nyEu-l`BG2-9dw6pddoUO(iJ)V}249Hh?iQ8zw>!NN=S*|&s;k!ll5YzdB z6V$Ed*8OH16b{Da%J?s`ngmf83R#?_D%4Jen#<)2DBm_fO1LB%$@Y&?3_AXz7}QXa zKS6ikt44%l&S4hT>+om(;m5qHu5WvZ#BK3On2UoDqbzfuBd2L*abg^x9O{!H9xT7` z6Yum>E1a%Jzs?@c;^5)D3|m<3k62tpNPYfn`4Y)(&*$fN_zg#+_2M5u%k9(ggRWpZ zNz%W(d0qVMM=6=_{nH`M)}l5_2ea^NEN!EBCA(ZJp(~pg4$c+>3cvBnnb+n}ynhq_ zTQzQOONK!G4}64Rik~-x(t7`)G<(L)xc6<$?+Ht4?Kbgzv>K1?gf34Y{Xns~czD08 zqLTp8t}-U5hTqxyuwFKRWptBZn4NRY09hj(xnC0=9?lc3wNrMLj}V3;SQp!d_h(ww zol9mUA3%;%0V}T#To0>o`^~g?2eRktQw752;Fl|0e{fejF<850;O6W_66~W7B5(m_ zEmNPhYwvobWrMu@T+F-Fw3rd9vg!3vg#0CSFw-!lxsQpMjeWQNH2p@#nR?*{PbzKWv)EC19tP3{d*wX!!!wEML!}ljIL<`!r2Y>UM=qF@^!u9FLaL z&qG@l>YW?$el^ZZ_td*X2)cbJ_K2n(BibsBTU<%^LWyAxl0P2nN=d%<$4i|i8{*jR zJactrtdSxbnuHT4#>-O9>lJ?=t`I!6B)8QK-Y4W+vb(fSV0K?5#&p3N1)+^5JlEdd zvu*o%Yh|(ll*qfo&9vqec`9U`@!0XxG=W155DHcryQ+S^xbj=aC!z%lA$ z?CYj{_UdOcZewb;x=t}ioS39C?lp!Dx2{GUhPhCMbhZaQ`_DtItAAwWvOK=@`IYXV|VWy9#JFce#McAu?OMt)cctzc~=(xeC&Grc!;~_3%)oZPdV?E z(Z&n6%c#ZJ+H6Dv>u(ymin@M0hvGCfJ%tF5RkKEfchm<=F_Fy<)9S!HYL_{U2@|I;LvGDT+tOT1srao4~Lvx-Mr08jOwA zxNsh!}*##DL6T0FE(+d0vdf@Zb6TH%wAQorOFsKTs|>G94i;^f z?^gN{p$C`x@wW5VC_ec5*>Dk|8#EKn}L6<}x*rHlkMEvC}p^!@YndV+?zpXQ!K$~&9# zlEZ@auPA9Cs+2OVp*7Vt);-q~6s`t&$b8s-uG(}}_w`gru#)5~?`#p?C?YDc@J7v{ zRvz$!*-`hJu=muDofaQu1$z4LLYScaniZLjd2YgUo8jcI>6FE@FE(4b?oYn(ImJAS zN^@AZ?3LYGy;^Ra^#s{qDsJmM@-!hfQ*;IOJzG>nRAlSjecrgdVzHT}T79Va71#%? zwn?efi3=A6kCVF_E(!ofb}pB)#R9NB0K)ylp<#T;3#9jcY~~WSdZ$J6jBWq;+VBrL zc%lZ6#--Dnd@AZWzfb<))fMpZ@%Y)+z;9Tf$NJdZ)p3i3+@T$#Fr#CHdTD;;a5|2L zMM!0&4cWNax^#k-II>x$VmYHpJNpvoxq?_8I6^w%m$dE6)D zZH%Z@Bwv?#=}U}vDj(7Np&&$7zuO9nR1bZWc)vT5eW=^?jckcfbR zh(JlBYb=^Ma8ZHnNhgHVkL2fA|M{uYh-bB!@X4ZpJh zRdE|2pH|C88T!=vihf!DchdSl zE>!~#KJL?jUeJbP4+2wjYo!AXuwA(3KK99RTPl&?8u_{y(_$^<^%N7SA2en<_iUjm z`E#rIsHnEyna<7^@fpe_D)L5o;v_f<8p-9VlH%fcLX5S7Fv86EhLA;_3Sm4VQ25iL zlX7bh{N_*hT|mWYeDuNX+Z#9HZ9EW}OAAYhww>PGuv4M&qC5A0ej*!^c&;c)N?zPD zmPU^t);8KXbno3IjPs9J>>}1id7|&tVe@g!Z*=QzN(lb7Lm4^wfhGJ-%2{V0J8(}f zjn<;D?J8@Xd>S?2cZ_J3-1~^`!~=f_Bo`5LQfWIYDP(|-j_$yO zx2L6yIk*B@oFQ}QdN|(I=V;rg5!5ZKL>3iU9s~Jq?*#0H*=e8ob;I^Qx`Jn`(?yaf z&Y_shOixkZGCR&k*iEGmfpUwOT%gMLgJ1d4l4XVN#_g8<+q4i4e4;H$?~4K1jH;mq z4omk>GHdX4=DT_CGZ;VX^WH0EP8kdkFYH+9^-3AVCAYpTFahn2llWXXR9uUXjUn5u zRAnmNyVd6oxd&(a1divek`#=hOfa3q&5E9@a?9a8ZnD;ZSIM&9*kHJ2;4|;&m{AdL zjrNPUYjR)665UrYJ-hYeX1t8#d{E41c`~+Rc?m?xGY@96X*5a{5w=i-lk!N5lDE0} zhhjB=yx`z1-8+>-JN*cMfPaq-^`(^0$KEr{zk)%IQ`#UD8?h1k44}5)(+#jSoK>Fsnc|ClV>voLJsqMa z#Re6(p0=c%<`z$)M_HcwrS*r37)A|~B}=}uWrCw-ydia#f^_L^2L{U`5O>Sz#??g~ zlz!X9#JdZnU#*^G#ULvli|b3}FFRs+#+a~|44!K`o{w?(3QJ4gnoq-Aogqs##GgOj zfFG=gGo_`lI`cjId5e;qJ?aXxkJs;^AMl>-J5{?7-F7z2*^yx*(N&*e{<|`L`f5X& zl?pq+l`y2o?1gqEq~oev4mI7vV#cbw3ymkHp^-W%w=e%ER{wtoM&bjJp>8vkPc_{o zI6@y9GUOhsStzxAQg-?I2J{M`Ov9X=9>?ui1~eoS4R zkOgP-OXBFzbY7ftt8kaR=5&8?PSvI6pAzgVa2;HJ3*q(2C$oEEny5$-&f%^sWrf$mtEQ_5UJ;^ig3C{1k($3ombaWZ_G$DF07 z5R-s;6jHAPa}ybcV-N{;1gqc4xXj9aWaDkULdZa7iRQ!IjZlBkA-2UxrBKSxkc%22 z{NQpP=|Ds4Woaaxlyi2R0eg`v3Nd+C0P`>jnRcR ztQavlN$FDtN|9dNGZ)N|_!{KxUOUR-nYyIq|7TVI+am55KXVms@eajaWPA`@gqbll za&1&XI~T$o&!Fq|3bLZl=#?Upr#~4367;OY<;2(){MVP`sus$;kfioz1G(6g1f-HW zJmuXiT6^YdswC+Qaye@E@<$=?_S{=|GswjF=WVGAWOBIhQ5Ey+sEWCBEZCUk`=f@k zUA3G)UmEc>H7nJ4G!7_A6?(3lm2)`E4RF)W@EEVp9@)vSZ##(mo>jo>XA+PyRsOk% z!h4EFH$5Aa_P>{s3-Nja;8Y3aQC}|$O9e+5NaYt^GicEbmZ&%>Vb-{GpkyaFI{hu= zHE0p6%e}rc{Ke`?9k+w^a}pC)(-w;mYwwLMVmEbmYDu(yGnyOYzfRk6et2ZWyWkex z*Go>yHAxNR_-Em-AX4i+B^gOmM4xrF-K1jnQO--5`9*Xu_?mhK+8+?B6@%Pq_fSE2 zx3t|EVQ&xV?$1Q`Zw;8cJ#XgHY0uW{h`1xuCHvMm2#avl?8t0d6GF4dO3KHE;VZA- zb-a&#NZvP`LDc$P8&t#@{F>J8?Ns}>US<6jgWgX>t4RL`yG)t(%$Dj0v$)Dn4?l^x zyRLc&X4a&BXj>;ubI;Akgy}D9rq6xkM;k4Ya`mp4<<56JdOB3CzhF;D-Iw1*sBAwq zM1=J}ip_S`q}8T-$wWYAQO^AM-!}ID8DBjc@S&z>ZPyv&l|jrgqRSWz;I5OKzEV7Z zW}!!?k3DVG5);B(MmwpTL;-mq(p=-E&Q&= z7r4X=sjXXHHYz!!PbF*N8gD=#>Ko=|2~)&S68RwN)+l0<<8@n|$%)^Di!-o-}?|IllwUjz7D ztFz`|4$U1kuISkQICwq0k0k-!tRG|F+HOYsX<@Ds-R^1Cxm31FL~7=_b)EdZuo?Ns zw}MHW#LGS~0Ke_%%?oLg^>$SSJy3kQ^sMPUUWn@U>v6INI}7S_RhPDNjPdt^ABkK% z4sSSWeW@|Q1`S*%tL^9k92nAcYE4k%(Kht*;{NR+Jo1b2fs`VQYk%AD~H?F z?1S&?RvtdbH5=(^Ix0Q>0+*H1GDERpL!htf^+k@p8xf;Ko74>Wk_%gm@tjA-R$*9l zHuyX3BM~KkE&2hE5OYV0vR$uDhU^h*Mud}-bG#3KV6h=@jK?j>WqM|NKIzw%dL|!< z!D%>Vf~!usr;`9XL-B({pwxQi=XimTWddj@jy}@aE7?T@CuZ)a`~$HVafEn1aIka1 z39R4m<_tC5TCvL3j}=d_Dsm=0+q?KWq=WpSRn? zdI2qVfO{8+KAu!O;%Hmoq+|7wt84ge3;u;CJp?jY2)f1K8YT*q7%p2}YZ+E5fXFd$ zk}2ha?au%dhm-l{cly=aXE?lRDeua#c)}$8;06@@{1L-J;)en}9U^Rw1M6waD}Ywk zF63+_wEQ?xvcaqFGMesv(UFgjC(B7iJ0ObWtM_3AD;0u6cY1qse=hP>rp^^Q z6@JvfO3vAD(gHu|9eIeRxud}q-Ep^z1vBG8A(S8|+2Q1^Z4yQu^Ein^#HFj?vy6*? zJ*uy4pqRweG$^QNl0ckspOFX&0t)SN76Rb%X1bQD=NH!3vnJ2DySQ!31P-{DG<#Rz%~!5xihWecimj=jW!Jn=y9$RcEiG>X&llCqYa}s& z(&x)6dW0T=Bvp{8k?95lM?g8<-J$`1zw`WXOl&3=q4WATMndz8oHzh5*)49>$4i6U z$OoNMM=4zRB1F+LkP74GiZ6zF*lfNLlH^rm^L+}_ulPMIkH(eRZl-v96B|FfamOEe zkIFYbteQ#UKXLNi@-Gt|Q@o22j*>9uI z2{IDoF&>icMbH_!tyM`13YxZeiagO!Y%;&&)Hs{XPv0+E>h~Q1(*71cL+`g+IbpeS zS%(kOWW4@k^cyB%I=$H4?70^G^budH+SHV*&zbU(ec=K%;q?t#z7n&Fhw#|Az1}o$$rX)-PxtW;9m&-&AoNYU9+E# zWXc#zKgx~+9zVbS4z<&YfFe7WFjCHzD06L51k8`N-4}e~eK_7$jlUxsALZR$fdvt|&GbKgpnkXZB-N(@z5snZ&++fJY=8K)_@*flCOtY6X!ZzUBHLvc07E1&* zeiZbS>;>i?th^#)f*J`R`Ft1=jty7rqfGA4>>tLvS$jAMEjlX5H*?UVsw)BpHQ+=G z!L)=S+hr>S?R%1Tp^3~1Qj-0+Os5Cga1k#(DWLR%IIdR)T)*u*`c~1958?%M6{H>nb`Qr9M!(GLkusW5FLmQi^<;5X~iAMZxYM@FNCk?bJEQzD>2fw3v3( z2SLu^SOvH)&zoazYtqjnCQ&|pCzY$-*z>+Sg|VDfA4{B@@zuwKrg#IX_&8VB1ZDaTmzl*Nwr_V#2=+y*k0`(b z*T+esBs66){MBkf0skT;$T2p_FZ&J``Ah@EtTa+X$d$AwX zdf*VKDS&HhGkvcseSI{ zhf&9kEZi1`55fRt2xW&ZwXXi@-}{y`{NC6k#zDC-2GF6OXK_x1V_2PQ03S|6iJPxy z$Oh^oz|v^;@D0r!Bd%!LP`Al^*zm}JFH%VuTk-Mq$Nb!1F5Bjg@nX5jy2q3J3Z|R|tJHzpNLv5$AGs8o11xhG4$L0$o;?|sLxGB@wC!^_0g)zvD%GRlF zv`OcTla{A-uUTDfzk%crac>=y@cwzF^=$72^9$ZyUZBT6WYVLH5n$`H5}f+P1f@P~ zzV0JGa zuU1ORWREMIZ>eQf6Iqk^!_6yA8mNoDAOrXzFGCb?j98Z;#%>NPECX*8-7#Zzdp zGyE~!k2vo8YuS#bi#fVD-fqDAFe26;{%+eTyWQS0I~0aP2@P9! zM{xhfvJhJC(bG7d&}VS?{JE<9f`hi7J!zp@NsPcKHlxo_+3z>n6VEH)Uy)bMN{&Am ze+Zy)c)*q*1&znWx6OA=w5!mwuXrKYhDF68d+}kf#2X2$`J~mK{rt6pvGaA@Ig5AE zVKr*T4fxZBP-(O}H`kgtnjJI4=~v?I2ZZ<=SO-a?lJzmfS_SSbwt%{vo2~6pHlcty zo;Yvxeh$C&knszcC#W>|YF`&e&i9W2>jsB39q_<_Mk7YCQeYeinw+L-cp;lmQ0_3R zRNyTc;p<_!1fT1uVNyfI``#r{J7ePfYNc`zFGS~E_MhJmpio=4kt`0 z&kVkG?Q+hj4Dz|cxo-%N-qJluqjhaxdy@t{ltf)4@#|AX*RT>A_SonFdQ+ZUFgi|| z2^cuG5&y=Pqt$gqXI!Ziw8C=UZ@0!-$cX+9@jYF&yPF~W7({{5N+RE8i>_S0`*BsQ zl#?=g<>USyYl@*sn6r^P1S_pqczriPX3O+BKC|uATCu06^;do&0jV;oPLsy>!2YK% zvxL7u&KR#slRFy&;P%`1%`VR0Iu`t$&d#LUlxMztrjJVD>Du?b3GEHuv%w=U?v9wY|2NTL>Ua%!%|Tc&2D76zza`3d?g&2nBfn3hkv=p91n;_?hR+;=Q^ zmX@;abkr!(=rl<4@1Eh3_d>@49h|E@1a3lBwwH>xje;h|6Tc32f*{eZEL|PirqU=d|B;URGLO z{3@K;Muk^FqQ5aC%X@D_HP26&g=EpedNQ|TSZNG-l}x|{l?V+aD=^gP0)^g{^cdT^ z-71SD&^o>Qj9EY9JnVd8P|V2&ps7_i={k5`KC+Jlj?kYAXs?ULbs7j-+faA*QEcLg zSSj<;`1c`Eu}VBKv0g=VtP7_P;I|?S4Gs5>;kxsa0-OX8C81+JctfHt22r@Z{9|1L8!YEQ$V@GDQ( zif^BgqL(BT#c7&^hZI5>{2c^#OlRm36XQ9H?W`RfYyAV}QbyO9pnl?HY2{unfFJ3- zUyEH5eo1km<}lKCrd=Z2$+pL$hT7%upy^O}lk=cYFDJ=FjEyqPgaaYQhQY0AcQr+O z@LqRP1~zA%_sQiZXuXQr8EQ!zAp0w-%xZBgZ(8>hTx8x>AlDJzF~}U<<{$1St-^2g zwpS}Ho2pkN>kCkC=tKQP$UoRc#yyp%(*3Qxrxh*3C`SY4=}Gu1mn00J(r_we4-U2l z)7rhgmqQB%s+Ya}hCQe0sV4^db2~UrHGI_C_1H!5a)Y?hQnQipY=ko}X)}h;+TNHQ zV*u(?0nI9@klMT&H=Ra2BF;(K+iO!3?SMTyGFR)M)0q-I`2L0r>_h3tZkpnIV~rtX zDiJP5GZ0^AM>*gj^%^vG{@qauPdn{ml56(2!-d)h#`Y}>;5ybpV}Z5ht*_P$MMci}2L*>rW?lfXA z%F2c6wv)E!2gc&TSX>>)VK_NmB;3L{Vg%o zvrwXw@-7ORBstK~U%!ldF}`I_2k3=5Wbi+9Q)f8VyGNoIL}>eGwJ_Vi0P1PC9@zy9>yP2Q{w zLDCjt=kWLmIzEWESEB)vn2JeD7mI6uVG_;~raO3UIE&QZHcuw5fMtNJ06p)?SyIuo zy{V5;ADX^!c(B|?oGTcM)fGAzZ44C&U2)grjp;O6zyb3i8f^~Qr~1052?2Ha+~fLO zF^;PO2-Z*9z|%Ed*NMo^2*J( zh`PJO_=lj#V1k;z`X)h+k$MHn43%FhrKc#Tl`ZueY7W;n)7VoyS!Gg4c>@Q%LUr*$ zyJ*SoIauKJ)8e`7r+`IdrV@7;5R=*d{M9~ur&{| zP1{V+C}3OMZJaxzQ)8yA^S;hv3k~;N)rHoY%Vwr64w^2sO>}=0wQ>dOPPFsr!+1{C8*`?-gByOMgT!;q7 zrz>R_>eDD41)h%>Ggqy+Z~HdAOb+IO2=A9EXEEq>K+jpUg%(+|yxctCf+Kx^cJz#zeq4->LTQ zA>t1jgCcszdsWdIo&5$9+=NbHV z9%@X_*y~JVw~CAw6@-W-amz6+O6vDVgf5qzVMRN9HIKBeGn^47UQ9EQV*?CDSruzB z`|#+8*i5CI+<vk-hV@M(`w_30~Vr3dG=mjk8SZ#f&|-qUC(0M z@gZ~98EB;fQxZD1o45Kw$9ewpmq@=ijWm#ffd*jQ&}^)o01PTW7VTQ9`>euKN!+8JPP_@dYrS%O0XeCn;#y5)`(=w%1jc)ve7WxrT{GaYw&`Sc>)l zYTkZqs4p_L#PoU+T-45h7+4z^wX6@twVy8a4nm6dgm5TnKf#XT49JXA^3ZgFaIC&nXY=1HX;&W?pZ zL!Wq2jGJeE@EjMTvg%*T9qJemQwx=F zFtEP$6hpi(ks^~1R7cI(gM&`s8Mr4b-AZFKN10-^3LzZrA*A{&rHB?Gbvj!~ ze_u^;6~5`M!wYfrEW2DrjcvC(YJ1{7IlyDxGZ;PIlBV5^QL`T?)_RL`d{bnMT;g9~ zW0fL1Ysq?RBWpT(c12U_5-2n>6s1#6l`{*dPmRANhQY6-_gZ$*{SFz_)kk?Aq=}wz zy)pegu7%rO_j^)aXKF$LW*3lSXnK?|OpB1tsn>}2X_LzfG`j|f%$;Pi^KV>nej$^*Cr$dxxE>(U^3^WcpPAQuCi>KTG zy3wCp1KGQk0E4?vYuiGhex@6$mZMaEwyUL@(Jt{*lfTayy+0)KdlTOY51QQFx1t(=?rB52H+N*azfYZ-a ziv7~a$v!`iGFEC(O7Z*-(kYVm7wBY*`GVM=Rw5tn8=e&+k}SU0KYR?3X5GrUHkXUfZQ8keN10Vj}HCbHb+ zEKqQKqV$%*GEDZ8unWc29=QI@OZcw~2lDMhBQt*g4VHqtK|vik8ZXNAto{SdXLks2 zYj{CSM%($m1f;5U zDRxqz;jO<&Q~{WEzm$K$gji`;0o0G@2=+kSs(LxzlnLL%7U`5&(4Y@i!6-<0 zXLj3xi4j}i)cBu2@0)G1dAlpA%}Cq{67d?5pWK9LAS5uypfqJ!m&Uwtd;AWtE3teI zQOLIJ-zw^eTU=WLTh}5xZX89<>r8jmt2(I)-MY|Qz?i6ZS_9}?G0N%A_IVDjTVNFY zyN0;a!;j;tySjq5A_;wW8}7>4YVkpGVtzm0>u$H_Q^n7cO6`tkIuf@cR&lvlT_G0dKPkX8)y zn>qAD0Qdg4imhHB7_!#-qSY#J9gScQH4`^#g8(qDGKW2}xG76a?2`v7zaRY@Wn)&} zM3<2S8MZdoY7F)|iMq4n;sVB{uH^SH-Uz7Ksof#FQ;+JTSV{}2V={En@rn9S=+@!2 zwvX51BL9-Cy6!!5EsZ`2;9$M!$tz5&I^h~lzXu%X;l~Z#K&T*up0hWrlmt5}fMhsp z@xItEQ*(p}K;9YS&Sxdye2~e-2`;pWe|Z$-F*oNI$m?ikxwUt1lno`)DCCHVZ>Fp$zSv+No;T1er4D1-kJUCuoy@(di1z z;?Ax^^?^WWFW`^lUrQLa=ZY&s!xv!@w~y|mSMaS?Zd*ep0fEtlVyPC_cYr1hzMjD$ zPu;YC_{K0@hrhYH^0}0xtO%BU zV%%pv)Lp#XP59|D=6Qu6TKVtmY1m>Fh2tHJRDVeV!>B(?mo0kTzWLl<&>82?10o;g zOHcvM8KMKOFF=bV0gU)vLCcqm_ouuq?acA6Ny0H>ZJ{jc+zt5R^^nbPKt7kpokLxQ zSZxtEqwq}VHGXSZQ!`h_AM-*Vsx?sa2>aj0_ir2o}R%I!cz)VWA2zvZ!uq|Kfw6wEEL&swttB2t?vd~6Yfy^@nVMEtu_(E9ltXf~w ziH_gzOpUJvaB2P776mdf-ZfjD+e_Cu{)%QBnP^(9LOkz89=V;y$7g3o0SBA&$I?l8 zu?7G6g)QJ``S~Mo==?t={6YynMPq(Vqb})iULpZWh315D=`RLQEqn}eou#~o@XcMk zZm%q5mPQK*1)tG$^W#p+dGqq05dZD(mh2#7=rX8-cJcm5JIfE+>gXHb-*U@wy(!Ym4kzV8<~)Oi z`m5u|{`-5taV7&q{OA5rGUE@%=tBh&P6{b4c$*+_z-nJ_q63DIs`D7VLyzl8bl4jL zcbU%;v?i<2GXDN5$j2^2GLo*u`+gQAHmBqCZgfof102rSnM`BXgV_SCoC}u6;5+8` zn+?rJp8x&b;3Xj-sCBnLM1i2@cl;_0XrOOOi$b(b#5Vjk+`!894XaJR>T7j%fUQG9 zolq^}e!;Pr8Yh>_2;LY?w3bPC_dLVQDTG-5@$~YqlTXU*_F+A|k8JMxN&gk^p5kTP zqkh_T`J)0bGVo%X9lC9Ok>BNqMc;?(ZL&IXnFuL6s&MeJUtgZnk@lx`$R)9}vJM{> z&F)nt>fg3^{VI+K-Ul9})|KRhvhAJ0u$)Kl3P+{JolDuM#y{e<|G63;baG$n%PH-& zzf;!#GiC9834Fglh1PfLmp{a+p`rQ?V3tHTpW zWL9L^oR*eW@-r{l!f}2i4W^T$%|?XhA^|y?yZFYwci#T5k_qAenaWuJiXPFc>1D@R zwfM%@e)1Zik1f^XkBd%siOwJ2#{+KIXf zrXpL_)wQ^b%m0MwUj##9y?>9;Iv_$TQ4L-IT6C@Y6nZwyVV|$XzCJV-i_SNA60enG zTX{seem0Z3Ho-;E%G#g5AJdNp{q`2C-RZWAMX3Qb`vFhojE~bFVCY?HCOTgAr71WT zw(@{(I9)F;bIYAU_ zp>c5ai&=23Q^W58$d8wE*P-~LGD6SL$r)_;*XnXs6QoKBq0u+TDBQWmfyNvZFa97u z)h!H&KNA208yu1y&S1I5k?J-$B!m;gY>?3H#14g9bEcH%^d@u-gRV#EZG()>K+dZ| zZ@m#=)SEhooB~2qw+ucbUYZt4@YET>|2H%^%GttaKrJKsgrkB6#Hn0Gf0J%q)aG2_ zMr>xn+T-%G28SRFGp+jQRPe+gF}GmTS4j&Wp<%J z={rBAt+D6 zF>!QA-n(?X?R@t6pVyW$^u+A2XU!}u#K6Kco&?arjndZ21~@fU*PID#4K{e=l z9aoSCpZc+0$0h0H`chSmD7QfAA4^Htq? zdeZQPcb@iBhUk8cLS_tYK(}%n--OglEubtYSX18Bp5Kfs!e#o=f7QK@xJZ5RH^$uX z>_80wZhU_#)^O`sv8zT1V{`G>I_m)F$z|osOdb9r)5h%FZQ7eP$yqYo#V2YrhJo z&GIYdSlNQ;5PcxY4ALV38_6 z-|B5DGe!CehCm*hm~?a5<)LRkFP)tE%Y>GV1#qE_nfKu5&xXu42KAS`r+1Bd^cR?-kKIgamt=liX^9?nU{K5ZdFxW`U9dAFHND&( z(7!)5hQdO4A>6R!@)jku5y7$eZvF)VJQ;KbKA_XiR6Op_?zKOhxq}F_`SMdqGQUha z%THi*=I!5x?1Fy-5lC|6?w6ysTpLA#KG80J#>tu%k_6{Y-}K8KDGr03B+;8`S;O;M2!`wY#(gXA;j|pIv>8VTtMd zRsGo4egH*yzq2V`fR-Eq3}t;)^#{^&#^_b2(@APq_Qh+6M5CbMzfUCy=PL@g+5F$^ zewzaI0W_Z}ypKsT{m?zpR#NZ$Z?hS7Je7M5ahO&YPTK( zcZ|_G?DqFuY_SIWT13P)?-o*f?xSO6PU!JbiM_=tSedU|qL)TrSfM$K`mYElFUYOP zU(O2bQ$+AV#9&1db3E4Xv*fKin>t&EzZVRjaWQF`%HHXv_ox|Io6#}7b&KA|)4f=g z;+8*4#-{n-IpbNlrF@|J4>a}ik?(3zi8uwDH=|%4jq$7(?df`Q3D0k$V~}v>pWz#6 z8_PmdpEA5&vG+w3oX@0wWInHZb&W5ETd{kkLIOb)6jI%A+}Y<{$o=97oc*Y9a>g}V~NTH2Jh zpGNS;bN1;Tn(Kq{gfkN@O8bKkAi>P2r=G;@TxvO)-M>kk+REzp_Hcfw%OV2+N667- z6Q?O1WE4u?Ool`IIj`he2WE_99tFeQO(4)Qx#%|}%m)E)0q`e{H>uiq2aWyG=3|1; zQrCzrEiEQ=${#ks&ZqeZDgXIs?!xB8H)oJ6z|SA=2}ayuyaK=Xq_2*9*_zEgt@@;& z;<R#xt`qUK-z+*OG_;5kfpkS9eiRgE{U!)3P8evIhzNnZF z*o%~UN6Qid_qIP?h%obDAl8*zvSe2Za+gsVwi8<8+K;w1g7N4V@k*W}(KDUxGZs(n zF)e}xNpLCn*Pfw%!vUD%vin?3aa17K7{RLjtI`8HR^%c*scX_c2S=%txe2Gm9qX-( zB8C|rj$&NcSBOeg7E(VBRBHcDkV1e&CsS_Wi^&nk>I7p`8T1{!PX1bMHD7 z&TZaryjK%+Zhl_t_~ML3pi|)I&3f8*YxPl_3QzjC@Kww4)~MQWA}+|4?PN^<0DNkv z(S5F34Ahh|$IDr7olpBU71)Lpk=Ub!u&L z3vBiu$V|4wB8cmyoPR?CznMTJE#Cd=03vAwh55Qk(c7KBn=`w4awnkN4E8*wzgysg{TH#c$ z)+g5gsjQr*EQL?BC0O-)@J$F-%2rYR&TM1solDa)aAZAiMXZ|%Mdq>oG=Ofnn@QV( zq2=#(U0EhzjAPV)m%gmBvZr(6)~=o1?u7qxd|Ziz#A~0G!`9(|m9C?GU%DuxlQlu_ zE$pfgSI47;^~&$aPIPc!C=~qz3YnlhOn5Q-D*Y-%d*Cv^oPwmZxWkl_Gxq}zB5IER z4t-^f|Hsdz*9;1netG53_PH7=#1UBxC4hYGPAOB*|6%Mu!{KbV_i;QCL5R_Ugds{2 zLbRx3^co?0FGGkPEu+`SNVFhT38VKedKo2J5{w?bhtW$0gHe9BXYc2`KYKs^$MJtR z2XEGOU#p(yTI(ubx)6O5KSu!;q{@)J3V^mE~cz==(!*|Zkv<*D zOW?e9b8tU*&Zr26Jv&`FTKj`5v#&fv?h4oPC>ESD^cFDe{>m$Vh>Dm{DxV_?iK0K# zAe}zFuqtNh;{}%P)t^B9fx?BPyY{92G0y{{&xpkr!zD0iOvibB6vn{MYj}CV2j4Tk zW(>I*b5+(vYa(LbaA$Y-5RTozc|B?QBIv!41uWGqGSQAoVBktXa~~g^c=hc6vf-pW z{o^w81z1QaL|Z`Zl*_R>X8}|_caQ&wAVo*>#=G&2QpG8#Gq!+rVWI#V6UeA6*?Yc>dGc)M{S48I8<@?dcv}p4v6&$x zw{ytxv7wW4B~RW6`fDOweZo(?l5*Ihx<5=k-t&bviQdaq)2?UdOe^YCgVeJR7Hjmn zk{RK-Wq9$mq!M6pjE{Zrhmw6PPE43IxXBCWudwQpm<5U?&tiBk<~ z0;zF-`aSVO3afx4b2eHkN_%u{1>WiSPjm-phXsb;-|nA01)v{(*~`^0fI5Jom~_3O z;)RJv8WheirUYb&d%{IwUoqzhrSxRo@)u~&Eq~QlEl<(68kUbI5)+peD5gKgCUKum zBREn$R6JZnjWCoyz)pWL(*qd`5@!`H&nM}FY|0Zf#d2wN`0~O2NB)@gq?Waj zo`Py?bgE&r;~>kQ4BX87)duUcPvfh*z0{Oaz^tSD#H#d~Y#e(vPnlBr3xaDMS@{wg^`X>NQZJLaa<3adZ{h)C;dmN14 zMYDGc4StJPb^C_CrdN6tIeSqX{(tcY4#Do?_nl;?eo7bBS2Be-XxvIDZ-t|+jM}v5DJbq;jM!>s+opaV8 zz-|ezrvPcL#B~b*RRwQkXkQy!Veq+#u4OqoXKI_23xS<(!2U0XM5y8WLu>li2nN?N zJ?j-+<0U$@i-}e}lVzhTAURU7#aWyY zpkm?=gO0G~ophAUOt%@Ge&oX$JxpRt2kghjtrhe2rN`%Hf4r@t<9Vd`+lNO7x!OX~ z&E!6wY|L7s?=ddOit3F&$%U(0?Zlsb00t_f9zP1wx^Ue_JQsk9<<*%zpaVE&o_B6s zXRll}@mNZGcCpJ^QcG9d zlSxvgaa8Bd-ubrb^DV1CY#yGjc<$r6ZG%p}*}00YZm6mDSL)^6Q*{q8CMW(FK6?Vh z-gM(peP+PPE{C(%O(6{i*BJfeMPvG@kf;X?@9*+)5vr7F7N!#IXluwCi< zrrt7*;pceL|6okOd{GDjenxc(s7QFJU6(+QuO?OlArzDhWGa}Nv8D>I0wP2J%Bs67 z;|&AQubVKDn^^VCsibgdWD0`7smbOqG;DMbE?xJ? zO`)h$a=B+|x+DXPSDZ?Gk1RyiA|6T)OzdL%9-yOki(@Vg5CbthAlaT0_yB1RkXIipdPJ#+=SEj0vuk=vA^S43CCkxa^?Si|MTnEy~jV( zRJqv`;l=*48BCgqtV(Z*`T63mgihbR`*FK|JZE>2v?3)uijsuLVR~mCVzmbpOYKY$jK_kg@;WJ|112Wxc7UEX3_>}kdt@fimb87oZ^YNto3ltCkuOWGUWv-n zx8DD1zh}WVSCut%0(1xkOv0jR+i|hKw_u7khd%IImEQ&(Tl3bUnwf-|RRS>!Fh(hZ_Zi;{u!D25h4I1xGCi zI7j$Pzzq=kTiYX6^f)S0O7Aq4j)=UQRx;6A=+-TlkMd7j)|HT)7!#feJVS)o%T^+x zw%fEEx_j3%4&pz(+sMF83me>Rt0-V=s`u-a@N1;We+za$!K*J?P=bm}iuNV}CNUM@ z_cl|jnJ`#EI?!XiWLi-1&pn65Vo_VI)`~aX$BDnQmyCBXPO9VjLj7ibhR6ViGszcR z#2jEx-;tJ!_9-;)o)a)Ll2+|NgjNdP|COq~_6^QWJTV?oRe6VL;VU~lzuZZYOo=ep%X4KF>)?*;rz8FSWw(9XE!`IMK}4+P z)tG+$%oTSfE+Z>6ukl=Nk=O=*{e+d`PQalh>kUr_pdCOPP2sW@Rkt9Pw^sB6^?*)Q z8?{XINnDzi^!h}_+W0{Sm6XbOwwg%om?H@g=_Trvz}0)!BiNvc-kV% z2{tu16VBtE(DZdIyWQK(l=S{rIs<8JCiMW>>F$=J$VX2utX%MyL1Xua!C%k{g@$vB zA80_HBig!KX>XFo(ZBGJ-xah>+gR%$5$}uI2KdbZ^T2vE)dH;QvOEhimb7{Br|6U! zCe9^Z##R6+0uj(g22i0Y7#TmU)bEiyzEX(L@1 zQ!5Des_xqRjBfx#PPdDipBwb~iwKjE3Jx~X&S>Y#Td!LR2F>TE`Xehr})V$sbn9RUdN#-2^ zCEpbtPC8VaukTR8OAhTZOi!*VhmMVhhoW2w%lhid%Re?iJH1HBbarRTYk+tBY*0j` z_hnpL7b&DxZhK_!z2)OewBYOh%T5opD$%C9 zkk7xO_Dc>Kv1OINyl)bFD(xdOeSlkVs!H9|_q3rRDpC@TFD)SNh3uL(6h4iT#6RIP zwiRo{E-~te+HN;6R-r2GVW@z?--Y|>(vwXt%+ebrs9dNEf9@L-^N+FVS9F06u#ePA z_K^l9U({Q({8{ZfeWoiFPXKJIk90aI7CBw0dd(SC^f0X|oi^|0(5+7}{ybS15#$c{ z6Vx7RS%3#ObL!VN*+L(F-ZS)%9;#q0P;B>pobv$wH>=y62R40P0DdkkFQs0fwI*Cm zYhS9d+ymc#O=Pp-q&VXvkGT#5>3KeZb}^Pr3y357$YyqZ8@I|)%Pn(hgCYZh{jQp= zDA{UjV4f&fO5%q`kYoQVpeN#+Y?1CTsS5J&7HBa4)byJyli%3cos!Q{PCSZdLHpCY zwNVeuu+UN9F|zyNAB$4yU8_18yCW^LBUmLPw)t72sk?amY+mG$<|~e zN^b;~rLA)JGoAM2B}(WluIIOGGNI^$Yb>&-(XT$su*H3BA{=`z;h${#$}gcyMQxQu zf9NJHlN%@+&M`_yK>6|#AXXL3SGVIlO5psi z*H)cR>ty(MueMvH(0~G1_324{G)BK(2PC=6TVI_?%T=Wb13afjYELCF|E&1l-vQoCqmdyU-mKf=?-i01HkHix?0)^%FyGeA zu7!O`$0>(e@FvqIcx4th+^Ewv({5~yakpKA))6(`(rVXVMG`hnCO%0ozUDXrsYADX zb=R@s2n!oF9Mk1;4G>|N4_SY^pZsM{_=W#}UBv%8&?4JQu^Llcl6V254un#3xm0s)FZbNu$x#7pk>Z@%Vk31{9`;k+qcY_b| z8)cD^jU2<0c2|WwIP&d)3SIbLz3}szQo+BlhEZoa&fmJE;-wWvb)(z1odu( zhJFYncC$Imw(L41vK-zp$n} z@42t2e=r2)oR4DU@qJ#nOt_g!lGB|=rzW$Y=n|>(@Z&yo+g+T2FW+)fMgFy{N8;XM zO0@8bovG1%wckR}6|i+DBzMBDbMC7+u_G}t+fXPww1``~uC)R2d6fz)jr zKkAoR9)8rXazx~@DXzON;VX%S%N>=^HeIX`UFw(IT*tOpu%VX-m*i#-?6jnnDke63 z1m2+<&m{OZnMaC*;;y~;U`VYQtcmpXT-V$0iJS3vx;%F0&gV0n31-jxJ=R=E(7!XM%@zAT^B}O z<<+;wvQEqt()jKs$qo1yww!R*;I6f`C6w2+2p8$t&hn3#r4i9I8a1&F-FoVUuaPPq2FYEIln+3NZ#_+eArVs$PY{=hC|lJ7 z$xnk}C}m6S_7(}S8+ZiY-tfOZ`Co$0%d;G0LZ#rA_s&1LrLv6lCyhH-T1v*K-M0Jg z8`8J5MBUsAQD9TF@}i&@POIKRt+Yqhxy6|W;ZS2UB_Y|FY%v({CuRuK>dX4V`P2qn zIlU*2}aJ zgDan45u(6*Z=~3_DwP=PYyTZT{;j5^lZ2GSarjdKKN1-Wb;4tG%dB$tjKFf|EnhU? z73GoW?W(^%sX#Dpy3s|yUD!rhF`g?My2oWlAsnht$8v%&v%CN{G6df&b z{k)j<#bz&K0S4@Tq0iC;5cC`M_QEB^?F1MmHHx-KH6~#J{NQ(H$~(h;GCT~j@@;FT z-f_CdLvRR!X^+%D@z&zg9-<=VGaqPdDW%!Ww2G;phJ(aAVIr7RrBdcFu~Eo6#nAU( zpFcOKK-f~PKk$9`<%rN}KX%S>G*2G;s!fg~r&Hxl^OOD&J`7ogMbdMRBTp*T9vbWi z|1*aGbRdi`c=qGw0s4Kyd5|^~yX*nF2Nq4OL-}68SjbzFo`h*3?Hb?FxbeH*;`RTd z#v%fY-LmYYQ?J;E*3C2?teRjQu$3dSqf+j{)&sZ78|xTe;Kl;d4Oj6#S(-Q7oJ78m zXTVo@!~1>%K`k2MK@L)mFjn{Jh1^ugJ$hwaS~A7EYOTj+H8rXe^|?p<)%qQb>SKzm zI+Y@le1W_oNPDLGL-YS%!gE}1Yw@0@x4C+mr0wlXQi=q%pLc$zfIM#Rfu>%$gF4o2 z+P%dKs4<79_0)W;yA)DB``!L~8(wVaVPe6ihQ@nx1JBtMH&ontiw!H)Hd`E?uqz^@ zn$q{lBT{`QHCcszS2>2mT|3z9o-E}A$yn5r>ow?(n9H@H>_cH1_ZGvDFPJ^FA6~4n zEc{%6cv>oDOoS}1bJy&=5~9d9_Uo>6M)Tb`Izf=;z2nf|{nlipetG-X6!iSn!Q?C5 z&im_!WN=N2wics(@^N@g&$8+14X}HD$>cF4!LrG*MWB`O3HqW=QC zLAaL2+uRpPC8d%xI1js_wNYB2G38F{M$s|#I=i?<-{j~R(Ap9gej2frH&j_|!nwsL z-G_^)Om!%&tkdn;?fpTik%fG3dn7_-<(4kmxoJTN1H9K7?K*j*4E{?XZ?HQq;Nqw` zul!P@biqp_o{Er#QitGBu12+V%+5#Bjn$2xk04zv;U?wH*Uj}5n~8cWr>ftjp}a*s zQrbhws=Y{$5aAN9zVOzHiFSJZu zo}iDs!@?r)~&5($G10?`f15$ghQi zFHYgTDP_QKI|D4}C|@>HpQpz(>V2E_zeh$NWCtL|)LhBs6a6%>Z0IZ?WJMsR z)NwUWa$q@Fofb4GD6l^q;mgFeJT9%<7e6G{7STBF>onveUQnao_3`e7w%fc*xP)C< z*v0-PIZrX@)GcZ>G_PJNftSuOc0XZy9EP|8mbm%;+;m0Ps5yLnqKjpbUn1VGYFCtH z<3B%sG|J$11fNoa5?2cLMQdyM(l7-N7JJ1V1wg>u;@lXECm4)E`RZ}{Dr3dSZz$_Y zvSJjUUT`Utog-5cW7LJ=gMI03Ivxie4Na-MUvjY>KZt*M_G91`N6X5&W&#g)X#hKi zrd;=wT5tMH!z~rb4qf@?9-7j_O(lX~?HUChY(ldq*1k;Mk~0`8AGYi7>2@4^j&~NF zfZnW3rEPB*-Whqf1zpU{U=e6+or&v2+KNTMW6RRHVB>~b9=JT3-3PWn>M7-foPoxF zz#f-oZbeAJow#WC*3w!WM1s-(+>vI@9RBZ;1BOqjGFwI4kh5}HRI;W?87XV5m zGFcg>;3}_D4rf_LrqQ*lXjxh=>gbo=tj(_P-&58EtSQ~;yoF<7;K@hV%YX~L=k>wu zvIgFRYrCch_At%ELUCe*xsOD!R$E1`@MHRF6$p zj7pE9$D5s)!41#f6nA}>I&GyBfJZzN3m6SLq)GfP&NI4SUn)z7w{)ujf@0-Zwjfuo z;8ICLxNfX|x6T|u9or*DV+aCTE>$89S zT1+C5eZBp!1o!`1d8(a0>EF8m{^ivpS%9LLWHkx+)P}|cM``|Y=>##1jjrCuj-5`u z;v{XsGKYN-JIs80L=^>=YvX791K@?fU2-D$;CHmzJxURcFWICBt^xb*oai@V8{*O5 zy$1ppiVZLpH56LxiscK;OWhm8E$g0~z5;oIH9*?-q9%BsDWBht>7%ripR?!wzS0zO zaiKzS}EsdGdx0u6}7yNfMG0L(Jf>;Qb6=@E?j<_^e#hPXX{B52KQh*pu*49 zmL8Y7*g}1gXl>z_<}H#oiQqsYX{pf+2PIQ-=c6+}n!YY;;=kvecI{K8S061g_wc(b zydYPH%2IBZp+{O!oWF>zc64es(i`dA*(kB6y`ZqR*U{XPLy)XQUcZx~lk$7}k`csH zDFhvIJeVeDn?qAX_Gep{ou_VUnkvU0SmZ7(ao7*Ini%)D_9EJt-_mD0d`o3Dw6`v< zXl-d^z!Hg6Cg( zqH93Q1{NmcMOz>Z&3fUqQ7E`(TlN0a%-m@LTH-0rp>pTQj)*#5DG!j+ed9bau}RVR zA2#SOzuHP1-@0+$(`##qZnd4iqEVXR31S}6%99MyNJfjmb~|l*dvMYc0x*N@8rq2s zmR}yPRG-pc6!%d)-+0$q2Z<=8f94XUCE} z(4(r?-3{-`)pumwaauRjq8*$x)Y@2y+&%)mX>YM)y;GyqBfv4T(2*&k2Y$Ql+mTRG zx92O5M0lQIN?>17U=2dzO}U{;v4Pm_w6W5D1ezT=hT1r?ioIa{67t|g>hcfK%NBVHB z*P~^lXrl%ohMym)itf8xTU!G$q9zIcr~G?FZJkT&8k6Z;{%o)r#zOW~lkFpDvu7|R zIdN(V<;J%&w6T{&ar?wVq`HcDVz?cEw?&Ehwl7RsWmpt0x$%b}IL~rz-L`x-l|)2g|r8?hcKj9hQDnDsKg?r{1a!6Ici98ekZK z)agZ@P&_)Ox5BS4<_uhzwA~UJSJEbYFsl0|si+a|DzN$E8sF=wi(GcI90?1Emk`SF z`VwJmf((oeTZsP-Q}?cd zYwO067F0TL@@kv;B&e~xjS`RZ*C(vhbtv^=XR9{rS!2Aeke417x2m1A_&vPuR%5Ik z^4&uB6g_c(LC$7e?DC@MOpOz7Q4t7~L4c*8)~h$jyZ-CRADI~Ja5Y8G(%Oi!XNfAt zLT59N);{KWGwFp&eL3b*$}_CeIi$~jFW6MdukCKG^S7rhuHCI)AwVuS`zK3v$tCDS z6@Ich+=o9*QxH2Z;r(^8yd7&7WiJfFx1isoxv{EZh{RB&_9Fu&mpjk@7czbT|G z2o3gzpc$LJKtysm5td9Vb0%$l*g7O))y|ZP2)8*+8|>=De>aLQ0EH6#?2UUg9a#)? z6MLBnm~7eC(=-?mabPZAoN~SdwB^!Oit}4}tCX6GZefjMDND~Nvnn{CF{G~^nK@pi zfv2?2cT3Z2BO^UqCvd4p$=VrfcC z6O+oGSyQrelb}nfqR^Kt}A)~EKa;<3ZWt%%wvJ+`4Q94G4CcJ2IlOPjVBY1 zu$oPq4h_pkbpbyl0!}9-;?Oc9*TlVdJ`tOgRQFWRNK?CsdvBrKu-Ua<7;>F=XalH_ zam5BnrKhCoh1wGg+=E5)E9LW(DzVMh6ShiA=@mXwW6^)>E+_J;nSkU|*2q?M)3L&& zd9@tDP>l4+v(s%%)0ZifOlHF3Zgk8Ia>6n)#U9#}O8x+N9f2b}WA#k8J_cW6xs&%EJ%w4PHoiCDdHd{n>H z6FVqa1A_@~NqSGvl$yoa$hXM?SAuzVG3H1~NfK1YREPmeQs9D;WMrnqudUR+2o?XZ z40aFSK9Hd7o2~(d63DPi$O&Mfw)kl>%KPc*qDD&e-EFm>t@Ol91~9ij-_4V~|8)Y3vs}LtlDuw}kv~Z1$Ck!vIK*MEp`?0V0(iOKKh_gRtAcVx^8_NIXbqmdxA!Vx$3+*&h@RyQX2x#D->8H z+%um%JbY@KL2BXLy?Cb+>~R~#=j@-?0W^4lx%1P3dO6@Qr*(@_;wY3Fa0%CIiE&)C z*x#MD+1f?P-n9&7_^7sU=Pe?FijGgK_)=3YA3U&Huul>oL9*RJ#$|un5ZG^JgcTDz zT_sI>3bLq-#Q8sy9=^;mQ{kU4a2OAmfWv0hTdUm!K%Og?`IL4`t@?uecU!l`x<5i> zShxs)H|UT5&`S$$H3Ucx-LAfmPCsKKm@DYv8Qda9!L^dAIXZ!})-QBMWZzZSZWk4O z44=d<;|eCAqt7EIoJn!1`{Z>@UhO^|5sx#or|F@)#?NtSW}<6Z5$}sBF~C1Dzwn-0 zXhs_+50BEL6^&kvx7owF11?QL&xBqLfh0SKw;I^p(hb(RTxx+)DT8)MLSd`!|Nev6EAjGFU^y=LIpi`9?(-ri3P7}L3*31_eviKcF*hw6 zMTb<hJZwm$xI-LB%GK{21CTS zqbW56qK#=!HQ?UNkel597xXb98qqy_{u9V}nLp4L-jZt+l(c;Rr}K!u^Pl!M`uSO02{HY<9=uGlp$ z%K5%=@U5xHI(Rojte%GHE3c$C7w`U&q|4d-V=YSqI6}QO93 zT7PJRaVRh6&ey#=ZklG%xXsCH?UARPn=Nr{QhDF)LP>ScGfb{ZmKVE4^TLGm6weYVaLzX` zNiXeR@$q20`=KqOV`+*1eJ$m|81YsD#f;xb;SEv*);?x(>ulE+s|z;gVt4{|s6?>8 zM>_lezSC*IdVjHILl>OnUMw7b!vjx=Sz)Lc+4SIKhQ;ik`UvDvl~#93JLah?is}pT zFumje(5pfd(UL?H(c;KKzBMM}*Tc6{PJ~=3@=bONTp9%#3>-U$yA1B2_j$ta@J#5X zS+kZn_s9tS@>^9xy%AaZK}7(J*$B+2gcOKFVi^5C2dpqp?oXR^qR`?OguvJs;Gln| z1Z}PZ2YvaH^ws_qS4%}JSNZsFawPr-)6gY$Dw4?8pIZ*O**1^6QqI_TGgIW=zeA9B zc)Q-w*A>S+od1+)16~0?x6_z;CTFA9-W5yZ$MrtlNICP1O>)L&0ad5H+uB(F^B;SU zC}tZvs4%G{pMMCR8qPx#KqS*;P+` zz|_R;vMUi)%AOYOfQ04d0RXXALd4sXI8vPGwK*eNuqg9(YTTD+3ZcAUPE0(1QOc#N z$Z{ZLpO@Y9>@WB%_GO@j)6^?iAtw?MiDFF(cuEL>%lDnj035CZ*qJ>o;V{sO3zZI9 zh5X!{u9bAtNR4nQ{;u=Eu-_hktgAM-tO~8rDT1cOXOE0+ip{5@U4D-0KufgZR>AJb zcgbR$mbT7ASJZ0FswX}g4yVHTvyh1J!L$A6@*z^RB$b_;GU-Mh5}&=bJR+RKtxE6X z+M=tc8yuZYo5T;SuGdeSIW!b;gQ&Sz2Ox(#T|G~*K(hSBY2|w5!cQQdZN<$9WX(r} zRg#zUIa@HTf`^>7HvB7sZd&1Q8d>K2l#3|T_H~3ydNo%K64DeVBu7}<{U~Nxwe2tr4btwZk zvu9~OEbQ%MW!1->zfKUDt4lJ!7aPIM?#`G@3ZAKbo0+6j{oL{1Sgu zTxCew!kQ6yTsNQ8QD0%~V}w!OmGzjEKpSSDSG1Ky*PV3uF*!NiFt}e?Vt->L7(fN;K;_h4`6rR9Nr{-7(JD^D^k8nT;%==#=HggY zC9|weeEVV*@3%QWis{&ceh;s`@R-mkM3ho%1@ij{jaS1Kl5)H3m>3}=LP7RB_P-kTw5mo!$n|sme2uK zU$Zfu&2n`hQBay~tZ4|`5u_qP7C(%dIvzBa7M+pu|LyTJ%J<<0?{lE-B?^)x0?-l@ zYbzg}C7=vkxlt<*8KZBQ^-HxuaOA7UCys7Oh;fM6Ql%m261My+3G8 zT&iSKN28SYTJqSaSgmcJvt!)&$?BgH)vZ+gY1xF?Zq{^(OJmTpWR=d0y2~3iXG=`* ztxY7Q{Dx1%i^C5pcTcuH1FvGk8pnl(^KajNj=P9U2UPav#<|M!{!3+-2z6O>bNV=W z6B0Du(8#ddBb=rVo{N({`(F6@O6U6yiPOb$zHWmV$=!KoKB{FD00cIK z^L_e{-*){aT+gTL{cvH>?Z(~f4(qucP0`}JhMW7z@IIfjx4uc(#wi>SO}2u(!7u(e zlq}TG4-B12GBS?~)HKrh^SPj=1?p>vLj`ug|0t$P-OpPEHe&^z0*oOq;cn|PEfpd4 zwZKmui^=>k-jp@?4kv6uOwhgeA+kA1b)Te8j>)i#BQsy>6r_@cxwsm`*ZQJA?B1$F z^VbJ{$O3TFfRn(w=m7D77d0+@DdwJ;0U48=7!Oc35mN-i%=j&x_<12^1n$a4=P}Fq9Ui-LAzip34rq?%`X)cN~uP&w7@wGcyegK$K z@HM-)Oz0M|>03)ThzVAYwS%`mxOdhrl$SK_6MjSZPMI@s(XOykYPum1X}IMZ!F$!E1YSvdffb&&jf5|h&lq#QPQ}i z%VqlUdI;R>JV0EFdKS2_&ja@f=*ew90#cLiH*H4I&}blG#IT@SE_YxC_zs3eXA-7R zs1|YBQz`E)CO&cZ(ih9Y;yNYfD2CNHll~{-?n!B8-ata|KNZV4w-vUeQ+2mSGaE>Y zWLO*sY71$Jl_UEikB-3<1a2Iz_$6{|qDCO=w_abn{FI=RsKes|7dg;9FU0D*?d)_d zk#ADc9-uerVv1leR9|<}z&Qumk*ow!j_Wjvk;J9ad|VAj!e3A@yMgk>IQ!EV-mV#l zcJug@_yaVL($xYLC4Iu*&qV|&@1fAL9Qz(0tdipbU@lwnd~Ia~>B&rv17t`2_^mID z^WzeZcDUJb&5C98X7UQu?W-*lV&np@GmJiXM+y}_GMPYrUWUD4{u?J9D`NH7Ba7>giRu3m_kC1%!w!)F@{ zOaq=L%$@ya(rBwFI`VW|d|<`$RWAzte7!CFdQKc4F(m) zuqv+nwLVaPTuD`%WKGbxa|6*o?@Z?HIbVOU#4ik7*fgQ@bdAk)1KjVabDHf4rRh`E}3WR<$R6Q!KHN1AH=*tBHAr0-Xnde{+J3pdEKaP>n>1q59?uu zxEp@M?31Hri_ZvrbS$Trei5wp2%FZC7`+A_f+!0dPs=vdYsTbE#cl>f%!9k!-^PZu z0bNRc5?&!KpK8P_2X{6%Cag}Lc6BJ#__?m_iS(y;?b=ur|D&o3W(A0Id71yGi!!@} z#m=T8LHITETePHl3SSo{fT0s}Dg1ygmtOtP+=E}66~w&&pq<~UfeaKe*6kHm*D!y= zrZ^>4Z|T+!H05PWRueY@y{`PCm=*Ai(DV)PZb*y35m5%wr#R-(=l7su&jApkn*PQ1 zx2rzeK%%y^^U?MX>ITUSZpMF5qUJ1COMl^d6rCQS<7t)mNUs zaEbpmnn-ql#}eC5y7e!&=$b}iTHD}!x-^pFEorNwtZ;O?V}-@^N{Qk>M~qnMyCwOk zP_b9^@_ht>pK8erhP8cKsB#SRE1lT0lIs-@+(_k(u8?sxdNz}a-m5{kxIu69)-5f- z475r|))-H6tOg$SFGNfJ5T|VzNSFBsAn7NZ11!pcJyMGwG)N*BMHG69QPK3h^9=l9f*p znF`-Z=Xz;iE#k4yr67 zzV4M^%q#b{K{qRH4rDe^-I3%siq+JZ^PCv=ch9~6>|<7jsloL!^K<7$k$QZvYYxFa zP~>ECcm>QCG;yI2Yky80JJHk0{oeihYnBhIz<)I~|92J<`@VTzs>MjjxclE9|59+? zwOstNOD19)cq@Efge(aL41dxDZLj$8#^Lo5(m zj;L4#H-wd?e|8vd@UFBMI{g(*>@p=-?L1kk!C{;)h&>u6cVmtv!n-umQE4m45nZ4= zT|h^RYwad|Ie=RKg?Nz>+zl5w-r$oKIR`YZg+Tr4v-=%D#R991vVIn{;=0{m<=b1d z2rKhRc#C!NIg4A?3Cs}ULA1YL3{=`8)jn?4`HK-au| zr6rY@(`F|fs^5Nf#d!Y`Q1S@?RK8rAiIcM=sV-gg;Kzzgo09J1mptcZXXtV)~Yur3dmCf`+4x}9{TJSqASAU!1Bs0rMNXbUHDOZ7at z(*>+arD70OlS;&T8U!C?CH2y>b3l0U|828>odp;;2}58Pd&d3-2V6<)Zu~w6t7?O` zip93SqPb67zlvSA?Faf%)9-T(3;|#@YlKb)gULp!)e|cXTY)Ba7PR4<7bg)_QB-UM zvYeI}g>GpagmCs}{}{3is(cW}K7_jMLDe;mJ`umN@QBaAK{&&l4|t*_m_MwU$sTjd zW!FK~Qp)QJM~Ww*!;RcV>2wBzVM$K5G^SN%UD3N2P{F~AtfiwGuCbvsg8$Z<^P$8S zV}*x{d_W1%u$TRVN&LgEuHwp{K}@*9u5$%D5;4!ucwb*vA+5^&XiT6$UBT~ID-KBf z!d0BY#d6V&MxZ7*e;Dsx5`ye4?XV-jAq9WbK_ZId8Lv`!fL7h~!X{$gi{)<3VED`v zrcP3?Xug)ECCLB-?G5T}zvocJD#STT33W`NTFEiz9OnP(3gtQ0vYhZj;qQc_G~ma>?x!6!GYkh|>A| zK)?%Rs>{ov-V~1Jm_0{>DAG|a5UO*e7)D-8Eb2y~6dr$p)HdoCWht}W$A19A>)$HU zB`1TlO*vkd$4(=vpRmS>0&%kE83DcF(8ydZ@m$v%^vo5jslFZa z=m{;c7qZrLqQ>%<0pdUdis6iH?N6VmueZ@j`lMAalo8L}`QOg)FA--sfe)RbW_ab_ zX+rZW61$LPpiXRVIGesw?zZ?-&CG`@H;=_$IfM1p_$j_E^9TsFYA3MK|;D z{vmXWPI0@lXwO$}?@NZ%n4F*`Es>Uv4#Sk3;Ur9DR*r%H(qEv1^1R(f%~U^^DGT~OfPB^E^~e_8!S$}$1Awe54E?;`Lg!b_{>=ldj=Yl)biq9ac}!J3VtpW~hmn9z>3Kz_!*jcrLmhRLRQ0!>4# z;Ra&PK7og#mGbPZCL&Zd^F9eFmaona>kgt;ji;wxGse{Ojh z6*1rRU}Y8+F^`m~HkaJYN+>SSbzI7-*>l)KwD?qzX_^IKZ!!X1RI1eG=L07B#1C(~ z2{=e|T;y&Ks11)f#CELDCu9DlnPQ!U{yAnW*}`9nXxb|>id$&i$cgYBmvuxE1@q)F zspuS5wCu)Y;_yvR(7n58ZRi=DWQFR0hr1Fjw{ekvR2?OP%G6O2=681|ol5}?UxonC zGI7pijemPCT$9}fBcM&)wYJAtTm(|jnL+@2Q!mC(fQ}$a;(vE={!27COX@cz5$;V{ zs!DgEtn#A4mH0qu*VZw6-B30~1AWr8<;V3?Y;lnA0e35H;Z1F(m~w`nTw%0NyfM*$ z;DBw1Sp^m@yQ5p2OKkTYD`iVkV3;UJMqu4FK z%5K(*>D&kyv_9~Cm)bo+N^uWnCR$>@se~sgua&t$z5xif`1Ys26}I*IMSVttM&EJv z*+YpXM=T^=uf}G%`~m8YbGK9TEQ)UGO9p4Q8KT}V*b%^@dM4jJ(xO$yKy*2qs>+@| z{v-KRIm^=m1cqJAsLso>FsID7A+DDH?Pr16g6}GX*5gt0&%w8O@|hfpn{}z(IJwnP zV&ofRs@bn{RW}&dd%i<*(=9YYIvateCEo6555;ilwKo)4%KN{FciQ_AS59cWwu0`S zI?g1E@9Agy4Mg`8Il2AmY4p=%Hw*GTUTyi3 zq8CxW=zA~4yhYfn{-2mh)ZbM#QW)C-m%;%dp4j;H@>N~dB&@!LjLd(Q%)O<=b3=38 zwt@y|*N7UgkiYs%Od(?YFjs7O`ZjI8@x~84eF4LLE|3EBXVwQypHAnbF~pUk@q=AI z3Zf9%I6nChT3?SKF{NE{Hm!Gf+%A(^4o7xs_a^cYMk{)Zzro+a2#}#K=#y^FC48mv z4rq@0Pt#p9g!LU`1g%p9RL#t2v(JNz*GNDOX?bXh1{R0A{ao=E))(>_P08&Y208%? zdo#eklYxB8Hg9zbb78hHaN5sS5Rf>Ay{=qSHp_>XPPdPB)d-qD4=x;D3vJY{kG-g?){PaI zjEjS|Nc?U8@zBe(?+2(mff$-smi{!<$ca;1*9=j-{S0Tet3Bm+KL@8#(&|Z|Bq{1) zP}j-q$)gm|=oXVRUj+2M1O#AapUtP%%?j!Nf^9v1+8#+=sjk9@9(1E{MvqG7cFy>t zoQ7`YvvX{YB#IU%>e+&($>7H^q>I2q)gN%|`ocdev-{I`LFgS`SFXQZ{NJLzY+eRz zOxd5swJAm1ds;NCKnYvJ3z5~v&rw<+rI^lrwL99yS3_DE)N)jI59+3d?)}5P_r3R%Kj8zWb^6QU2UX3Xgjg-G z;HaQ9JB8)^{#2f=Vwx+bDe!T+Vk?YUQJkr(&!{51os=-IKp`1gA2gT z1zchL#orAFErU6}()GgS{7E3ObW2__TB>N5~rr{Q- zNn)ByZ3KM1s(r{HvWw1cbL~UI!1m;TeLOYJ7W8RSLA-Ic)bF{)%znUY>?hb*W39Or zkL4F(dfS~R*)%wAB>Wc#r^pEvnI#$c`QDRj*9FKRGbn_4@W0$J4{$AVU7=8Um^lw0 z`r;?h8Jw4JYS_z{(!=V!5l{1|E>!9y$9Gp1O(6<3zv}*tyz7qu4q6?%50X1D62LV=z-Zww2lG8$lnI3 zs8tgXw&VqV#6vH3UyIY(4uN$egGLIYeB6c&eSFE8ipn%$j_KN-RyLv99rp)E{Wo~@l7RZ))XL0g$Qb<>xG@`)y$G)HUw6)5L5oVvD7Q9Z+} zn?#x^o$>^wHcLE zYmi!$S5+C6J325_V~;;M{_YssoixDhq^*TmH_t=Nw{Ln?gI)se4BL+=c*63v@Xg2l zuVP@01S;7&zj$xO0F%5|+giDTHmIGcOG=j0d{=5nW0X7F#(%9la8Ci0&a6`|u<2F# zV6nCa=+QD2R4)hbY7IbS5a?=m7y_`?BC*WVz4) zZ(%u5Pgn1cktAhW@i;oiUiy5wPh3)U?ouQ#m*~%1;Ol|~VsBgUsy4sp)1=39;y#!S z0(T?t{bv_bABecbACmB|iQPEFs8FDWQHQAW=w1Jm=cn1VNycK9ZZhnc(rPr_yr&SO z^@ro#5vy(d$ywnd!=b=i0hfguUp^!tm|SwY z;E^`=*#I+^i2Ma4KY;N!yXVE=+0H;ioO?TG_Hv!`d39B7y>KT1QTYpLfWv=gwwY`k z`QJeU$M5KKLpxAGy5|4RLhyg3ii^)GWhH_)$py1VfHCLsjA%k%{RH1-h#*{aO;7-s zjWgklo7akIE6DZBZC)pX_dt7WjO)GtB=jC+b&GfP`)%oE5x(sBIJYsWu_!m>`f&;B zuJq~H!H5_H$Zk4nr>768%~T-+(~ZH*F+l-A3|NG?m%C%jORag|`{+qcPp1*5OUUzz zmZ%u)I-&{&Qp>#l)GFCwt_jxPNoyqb`ew3){FMJi?`KI8{gt)&&scKr;p`F6GF*Nu zb+$0AVFuHBFZ#>}<_@P=|?W$7~o*DM^4 zw({`rOyGnUOdZ~t9q3Wd@i$jEp$C|yv`ot0Z8mL{e=^H2&wNG5anTF4stqrOtUQc_ zPiE^z?;)Z?mnq<4-Em@p!+|r>nKxXg6ffPf|H$GhZx5l;U|bt z>WEzaptQr<^GsVa$MDyB;x;5m4FabTkMaTLIC~y22zxB+QLW^z2t@FyN7cieu}2k& z05Xx^{Cz^uh*}py+u`W2duiRW7sjjp2pcZ<+2bOEn68>;qH-SKt9+liSc+K+p5QeY zTV3fHa?fX*urAH^Lx=@d`#D{4EB;x=Q)l#Y@@y(o;7Yje|9S;J74@MEUMPK*C|BU8 zqUGOUuCSa$5H8nR?cL&bNOx$b{;gyLC?n;Cr`9F6b0p$byq0Z+5xSS)S$&eC70gPz z>#^yu+}%|O_D6Wk%sSa5t!dPHZ03%+)X{v6Q{7I8*VJ4Z4qlfqkhI^}*!m>jaQs8? zbhVYLNoB;`ap8&qak?!SM)HGDyG-TTDRi&E?$@kn%W$KAl?rk5AMBWaE@C>Z8qb6AHSglC-`iccO6#3Nmw<6eieSr ztwYRFs1Vct7gB8n>StOko;}wQQ}jN@Lh&p1PELJV52XuJNS46hos5uLBa)<#z5c%k zNxr}3<+tezWFiD5k*`$-ndfKsDEjD*9OR%zIY<`}+&#p5+LbHz&I~)5TqivnV(=qn zZdRUMC$=rqvpWm#q1l2^<`&Fx(sQriH$ZUnP57|24ty!2P4*od91Ry8h0~qcrJ1v+p`|OwYcoJX;rZREJ@t5lU`f);6@@NVn zWCJ!U%@I&ZKNGFKv)yoz*?%XVJ{ zF&Fk;;ewP?fJsgzof8Z@>hYFWx9s{?xkr0o+J$q;WZ`RX1v(5ig9l=j1m?CNrC(DW z>Pa8u_8{j!=)BC(;6BJ?f1UHOREx<+tYH37z*3t!oa5@9OLksfIypFR>1x|z7aq-t zR}~Oq`Zs$96kzVHxT<+f-~f)UzmWS)FxT5bdoRMQd!`k8Q#Z#^CqW^+6}rASpx$9* zgT*9h-%<56llV4MNLls=&sqRndjeHbQ0of-P9p#@cHZ0{C?soSz`#g=n1LMgglXxD zMiqpAF>#`z`wzD@989(>M>2SC=*Wk-gK4fBOc({sm<3xL=*)`Te8^u8tUD2)22O_C zJ-t5cIVsUrm^~6J6d1Ua(jxT`dfUrQIz$MXD6t0fyF1qdQaHMgq9oa32zfI+t;oJ) zVyJvk;L5G@1*kUJRtj1OTrqbdTAjUkx-? zz;J%5I4k%+@(CYYDmD1uk(&**o&d;qIUQA!zC^I$agJS|y{N<#bad50wvdbuG*7{t zd!*b7wa0dduSXO=Le;Ao;Hk3Uf!DY%>*2V+Iy^x~OxO0c83^}pAv?(`gkTEV2+~WJ zX+&1hnM#)=JPQzpn|1P-b*`UpxOZUuGS{q-V z78?l)K&7As(~Ivbc^Q04K(zZKrgW&5Ubt!s{2(8(3t4Utkele_-OQJp(Y;fjKYwB2 zG^fX0I^fl!t%eN_(gB1Y$a42K`H+02H1cw^vz(FYI%FexbNLd1W|vg|>8*ggS_|UZ za~%jpgD{*D5qKS+!aNg}lo7wLabMl4hzg z2Z8%a1ei@`laoHu@JqJpqu5Smlc6rilQQv@2+YO#&x`XPE2l!)T$kZ*(tuIp+D|UESN|5(BT>#VBjO5}koq9WwHC+rmhy20sgF|Ehoj*-5z0x4X z9yME?N&iEeFq8hkW4YP58`p8|-Q=#yvT)wmd)IlyJ-RTd0CNrwaGZsjD@M%v2p!@2 zB-F8#rie^G-y{p)77PvyzgOqQkf0?lJOW*54*+KTa`&-v)B3>n;N1BzPI9jZYpq8F z=%9Y}etv+KgiZ0!NPK00L@>Q7;Y|&+hdtNLphlmE)~1%M1a!{M?%IGEgO6}KK76h zmZ?CDWhF%u5uMW^xdDhDzrX#Lr?R<%?(IoWZlL=U)T#vjI|!2gx#tfT_yWh^@(CdU z4#1qZ)k&Kj_F<+Yw%G_!+imlh2_<}y!KF0#RU?Tnt+m0rMQBAl@Z_qUc3(VTjx=cF zLk@=~lbL54dvP96Lo2}W4u7~L`tN)9707r(f`iDf>tR6jQ4#&2E(e&+t8c{U)Dw@j z?dX9Jeqp~-D==kEu#!G9cZ7vbZ7_}xikTmg-XsU!^+L#uVDt~`ZzqrNGQLhBH^MDh zAuHv|)Fo1dL|(`V_z{);DF^q@Eg*WP2NnMQ>ONV|H4&`w4ow)b<3~aeH&{`;+z3L- zymkmKpCI=q!V;^KE|fOla}}E5&Tg0T56kT-G9n?xc2(L=mI$bp;|V4xldCYQEU)bS zsxoMz2n?7fJ92Ct1^vEuCpS>#H+>Q0^HRdLYva)M?{M%@p*CP(SYH=%Kf##u)11`6 zFNnMU>_}}0eC;5e*?RnEAOdM~rrtj*I6&bUeVT_qNC;BBuL4Nlqyt1Uv;6FVkGx&Z z+TI4PbqGLk#0A9O^ktd!UzVN;g!X}i*d_yw>yt=eNBwOvoL!Werq1I1nlWbed7XN{ z$}e*r%~m8v0z$_z1>QCm9&Ja9-!%v&;e*A0k5-h-SSHQqMHdQKeBF^Rm0O=a_Vi9Opq%WFY1BG6HO(%&OqS;bYo4(Pt$sh?x3EDuj;|uGVA>8EGd8s#h91p(> zumWtTTy`>KsKIO8<|^5n;T!(9mF_epCTb5@x4<^ZXqDhU9KUs({suBLBs#zmccu96 z=~u5$z`S(=^tIg0(Lcsj&^K|nZ~OF^<5~+~$gl*4w)ptF?sA-5*3-GGxHhHZxYo7e z(7cllYg_1J{?{S{QSo7z&gxcs@0I zlp57OPju~Eh3rE-rvL+Z1{RxvRziRweabzbB0lvTmG53+nhs+7gNoml>-b(x2&{ft z5DCWtrsl^N881(Eeunej7P<2b zQD*&xM}t-U*@@=fR9d@nVX+9GPL)c;-sBU?`XjlOP%t{#zlVyWLjWZ6W9l*ebzo{_ z0ieJ2cRS!=KIgbk6ar5G0ykT70X~R-wL>*_#s8h2q}jI=laY4&S7F}(`rGlr%Tsg! zD}^Y9ODT3yuVoxMn;DwJ&@>1Qg}jO~-9zn5lwx&816=BDgmlDXbJmr7?>c!N?-wrp zoNr%BJO%VkW}KSm1v)14Y$rW0KYK8cCk)ZltXGSiA3Ooxbr6Ydxa9SGqt=x{Gf{;4 z|4V(*fw#VQ(_gzjYFN8$P)VP8Fd-C+mfP!3{%CB4b*Fe!L_NtmbhNaJc?H)w+ zG*C@#+UY(b3!TR)hXZNVO5eDI@WqnaGRg^TzuNvBGuj7eQGh0zGUnl&bC85haz>(i zI`BN9dv%EDmX)pm4EUOp(A$lT&A7{5Uy-;=M(Ur>{Lw9WG}|p545yZL-3@qj8-$BJ zik=$iViO`q1O%2!*c(V|TW(*YpX<<|E_dZQx}cuMSbt^b9Ab@j`_jwFhc!T#Bui%m z7x{sQ^0(>Bkb(6(UsFw?lD7VL=egc;$9xaS`*BHFIkN4)^ix-!Kk9HY^@QJm2kklV z@b0M0CV76K>qkyfFNqS0|GAr~`H!e19VN0+ZJnmsQPoq@?g3rewN z3dhWCS~ivToDk99!WD()wT%A)QPJ_D3f0AvBgl=vxnT%uxub$OtQWe%caiYD&1JN| z2e!UWS_H6 zl7-)=BRavGtqCYy$6c}RPTYS_rBRCU<$HsFxwy8UZ8e-tYt`Ady8Qyiey&Ogo@^^7 zTye&4ta>-Lwy<75KNq3HAgA7MsV@);|Chvo`dfe**5!T~`g|;9$qE9WHQx<7 zBxznQTA!~46i!R+SYu=)bPc+r-|>U)`Mi75pPIH;5U}#n{sdhnBSKM8)0fTjq93%> za9()Pb9RO}0<9y`j$S;QEM6lL#95xf)#n)PbLL7Xq7YW)jQCfJx9i2N&s?OX^yz>?uB*(6SGm-{gPMw-Apx+ zBKFT;RStU3AK(g>yCXC>U5wn%gXVud{-^&3M#F$yP;YV@XYa4ukB`&0(hazs(UPgv z`_vL3=60uKZG_M&`vzq$TGtBcVAuD+Td{;ZTa1JcE-b|E1pYYUQJt8@Oes?{o zHb$jV=_t9rL5tHh^P_qb;$;hGHg3K{cC0oY@shN`)foBt4tw{+(bgv`O5w+V=JDrwl~PFejOiXmou1gxd$G^;%y^ zLfVa*C~EieCB19&&vm^Tzrb1>d1c^kIdpdTk}@82wEaojpFkB?Ya|CTOuoi<*5G** zdV6Q4KGXA_9>D%1yxh_~AI^{JIyCxMf?Jf1wgS=mLDpV+bM+fd(SEE(N@u>;Js3|c z;6u<8gUSkvM(({vS9Uowv_2bH!pj$)ef4L~^*w4w@t^xhE~9_NS5tnu7XM zqg!^mLMlCun4`R6bGN5daaSy6W_MhDMs)CelO=+uD;*B}II=aT>z$j=urH@(x;SN0 z$iHx?YdwndJUY4zu5vw9Td`02|B$TF7OxRnnQ#>R>KYH+@>0e$A^tRhIcn;dX$wA? z)1zWz1IkZNnPUQzl>7KchDU~F6;s(B>^)RqDe!gyo@QXyu<8?V!gCi<;DA)&^6@%@ck`4d*Eqr#FRw448?0;IQ!nLSE~2cnBB zT8E>DxOxF(gCxF6@2{yYQ{#?r48*wO6hA*%k`$eZ?-YtTsG$^?3K_{7u_#~S^{3+> z@W?P56f%)~l@LSs`?c6Jz3o7e*gDVSL}{yo7i(plt#9m=C+|8*+7X4I{uT57D9Pzu zC9U=%u~#!4fBwqPTa+F@gwQgOZ#4cX1mZwJUshoS$a!v(0I~~DHE2kOgucbHqW}s9 z_Si1{sRzI1USsxT+T;NQ?ot6ut8$>vuuirhME3Lhz-#d=1Xs#XF09VXn_*7j1`YfP z6y}}*1UBKL_n&TT$?F>i#Gg=gS+7E1XUA6n6a!QOfyQ2tiF6C{DTR&o9B-Sq01>wa z7cn1?`97FJrkDws3L-3fHdB~^R{-mx2{4+WZPG{<3yh5Z$JGwQC#ka@*~G8E+7id4}^M05dvIXrq*Wjz16rzUuLcHgR!Gc;K!^lgH9 zTL0rO(c2+pdkV`gcJ^4+?ua$`kd&GP9PMP+7ysOWflp7~qu?PJUC;^bADf*ZgNkI3 zSl->OWW2FSYfem)g*VSrMes>KF{BNw%F%%xxT00axqS^j0bm)^+2)>3ed!kKCim9_ zN)*C3C1c&Q-oeHLT$*q&_#H&JuK;F`J)Y8}hilU}=V$mSvU1U|*G9D~v}$cDBxl1y z+|~{J|BpQ6>d2j5#sXpk*GK;D-7$3`)rW&@q*@ z{U@?~09WCshU*^Jg_^I!8h1Rab7)Unaam1;)fJI){AUHrGi}7lkjQ=DH3_xadf552 zftaLdD{@eQ(%ju@0s=@gfj-e(C#!${>3<)HC>NRv1fC``v35arg5x&$1WN#0byAu> z`sif`@Vs<^r4|y(w??LiKCx)~YSrWZB;th)cKJk!4R-zXG^dELyD>0p{Xnb30=!y< zp7{GDJar#S;>Cm`_=EUr53IE`w?^@E!2CbI0>=-_Yw=u+fd}Ig#$TJlfMRcydOGEELgp^vQ_`Qg$=RX9xA35Rw7rEI+F0)ainjC{Jy1!lHD^8o*l7%pjN2&K0*(qrG_k7)isz23JLH(WJ zhj>+CSEO(J6Jo&?mr2)@7Lp4;9s;jXm$nKdDc_Mz%p*7YWHq^Lv0QD@m5zasY=Qaf z$#qaFcf0P_xDU8#@P22oH&~E#+;TuJY(Ko)dn#-#d~*zdoob;0IF|S)n_cj=2h);5 zZQiPv%VMqbn(|kNi$O}=eR9&~0(=_~ zmH}Rq8J>RjP<&9UexjNDVTo-VCWV_Sr|V^J^yUWK8y$K5?T^fYCFKIp>W#&Q2Ia>N0Ua8L|Nar4!Fm?$$) z61Sp_)&RLa=xCI?OONj|6Cd6q(vR`A3jG5=R3m$(h2)>Y#|%qJ4@?F|8m4gcZ8^jqZTJo~A!YtQ ztAGOwV8WZOeEn+^{W!g3`N{Eis;7FydXPPR9;+(Hm7EfdgRkYNK9d0qS_}nTj5N|U z7fY`6t|D?s$#k~sdMdxj)9a#z6!SNX0mX{ zPvBvm>J}!vM;dwL_j|-=W^aivV8?Fj;(=~a>`)J(eIw*)xwQ-diEv6zH?f5T3yk3dA%l>XB%tz5<%v;JaxP%FB9TmXgb} z*$bmvET}%Ue4vssr11;f-BdVum`+J06e?Ic!X5(jc@$>!UI8RTLZj{vqFFk;mcfKV z#-(|unN=bp%X;!+jXEV{asK?6u1q-pUx!*umg=~aaqhMq-}RRO+t`W+)Y|L2w)^Y{ zP4nzH&VKGOK7oV~zQ|HaVQW?lLGs}6g@{T}i`!9y>9G^Db%3HPw$r5Wd_M5SB}0P6 zfz^9vAQ*f(9E+&*>9VSQORjB>;q%t0JKvYiK9>*Vlb!Rwh11!VYUV+Xe#j2z$P~?x z+G4uMJ#9gLy0x5q?V_&kX5}mp>e-LU-gN7|)l@nFWkpwj`ar&OG#q?xHJE zVhjq1*leTiBYs%Tz**8B?Ctl#zwHcdp`IV)+k#9AKs;LyiC=ftd3g%&-8ega3FwIZ zvjs1DHSC^d&=G#)%0xFXfG5#yerKTXtx1(;Pf+qX{fq=P&&^uL{Crfii-*=j8`irh zc$7TV7f2M`C^Wi$5Y!e>4i1gtxDXMN*F51pVperN&(7=XF_@OIzdvl|MUv4(Re&6u zzC#H5j=5cCSYsJsw~?{mf;9dmNirQ3NCi-UBS1o$IGjQ*h5g)#Q{}7V@E4x3FJzay zOK(05*t;SY;A>+R9DuGox-Hnb5$tUz!5Mo7e^}POzcEFEO?c9O|6rJM+H{g~UGe+P@$A zj~h)S+lVg)7692Zk^knXWe&-;`4E8E5+-#t5twU(f+wEN@&N-Q!-kK?kTOhbrY%dQ z&7llf30^*K)7kMwx$nb**7Ub$fj3|Vl9(dSL;Vm&xF8unM0lgnDi!rD`i1q%a6Co1 zh9&Mn={UWY-r;m^l44BRnCp8U~k;Cmd=Q?Nq0_%$Mnve zo>HpdGBfz_#l!Qm-`_JXn+-m%p^=T%5T zxLwiM-2?0eR^`H&`HR%b*}Si&RRmgNe;eZ`)ulRsRET9zq1k` zmg(Ho+nl~q+A(k6{HeQuVXLugB0h8fEJOO_q^wHk0+|nSd(@xWkkkR4_vlOVLulK8 z>3?AIgXTR-VkvJT@VmwzxER8ZG^%r+=wI(K;G{}03`^fwGO7&{p7f0C>TXMI=oAve zb$$YxY@`!G&qZ68w1mx7z76*SbFS5(cK6TV&sLRo9PDpvgS%p6zbcSeXy6j-Z_eud za#O@sA;}+l@0`siH8Z=h{eb~oEob&cN$AY;8)oofLK+{C!ntn_k)2Fn71}8nh_jEb z+`D|MdQp+}FtK7c77)ulM4z?&TPQNVqI4|*^6xi`P=|PxVz75SYDE@QV!58rHk-~s zzZHfD*5oa3cUM{T0L>3?TQwFJD#VT_5C4r zV!zVVat~FKBM*obAb{#L(9h{xmxLtjkcBn>t>~b|huRXLQ>oc==lX}%;TRAbaKOEy zi9Ofv5{rF`Kt?c4Mg)-*0#CqZX;D44^@ama2Q5kGAQoj4Ny&~sgtNX-I$#lrk#sK? zCy~jMJNtWoW23&G!vRoONz3AG=N#n)90m)qfQJZp=ge{pBObRKJvi;<{r<7xtkuH( zv}$#Tf9ay(cj5OsPx4Df*-b)OK46X(12KLOBRH6W?VuaqeaKv*q>;G#rLUoGrPL); zfCZeZuSxDh4bdsprx8D>1_Uszf{z+BiGztpt{wupsp;EXby#JBHg6895f4`?D+6eA z0j*8Efh7$0G_8ODf2Q;mQ4)v`$wZJ$<2E_O0en)>U=x}#!AtsL!1e^aZa$DNJ2km_ zH^3*#OS#t(d$`-IO0Xj*b6T>sV+~I-V5`i~&>cVT(IKOmG4;~%o-_CwgK-ZMK@{@jD>=S+KP%Mt3Y??Ux_{#Gc zy*D#9;JcJJ^zA>QBMv#r>7{F`0JY>bGYMFw9~PjW4u>eNp>FWYeJY^`gz>6#E%E`% z(Hq?P1L-T0)kmAx;zG;1=DX^O!336Eh;zVwr^(nywnn{VDY&1RGp7aMF(=_|P&1%R z-4OAl!MXR%fJP%D)h>Ty5^$GrJqPYxZ;7H@-SO;~JupeGlehQw!eej9FH+JuID|nt_i(=wP$F)XsL8pE=3#0qC z*9rS)=g3SWM$~|J)X5nC`zN0eRgpw`c7j&JI$H)`Kcu8aA`YoE2u-S*(4w2u?&_Z> z@S4C$E-plQ--m7ordO>Ot=SZoVFF1GbRu)lv?CA#+|W@v=nCV}mR(Xig+PG80@+4q zQ7&^KI}^#o=Ai*F!m@nNPdG5U7;1Y#jupHwM*hX1vCPb zzy68EM(xoR-g~4Vqxg@Xt5c#iad1=(3c^CT&tfvaL7lM0u|q3zVdVkNW04T8&lgkC zaU9EfPjWk8OAv_AQz&1eT^}PzpSg2oqcpgT(Z7-a$_Ju)XTloSIKUZTW##zM09f1XMqO!Gp8P_^_fz}x_Gj;rUu|(FjW;AC$QJcXP1g-% zkEyJQ$!RZlkMCyv%md@ayrOhSOREv-Q}lGP8<})NloPgO-vU$A4{!l1(+kGAUq>kWFu4=?NOb?pUq50*o%rz1{7zh>+Mp_` zNA8ygQ3&uSlg_8pJXqp-Nf;c&$%@!&3lM36ToCkPiwrVTN*K(%_)uCj`bD5NS)0b7 zB6eJ@=Xoo0%f|1`UA*-4cq?nG(CF0e$%yhB(eW0P56iPBa2D)E3NRPMM~uijd4Hob%P4AohIGQph{{Y> zlyHIbi3t6gaJjvmacaeueq_a-aazdUZSHQC&HdL&dd1@?*#4q7gxCCik(%#Cy`@hz zHecp~wx8oXmQ2ZmsA%@HJ|Hk3y}v|m!Pi;$czgC-RdXO0$>WDq%OEZNKJD`~^zKtK z6>ve5#{Suy5?g^8o+Wn6EIkwGV-bpNk($ScRfSZv=mc|!v|wcaEkiB3PG+}-=RxBU zr)7&4?@o1NCqF0GTa^Uz5mS0QM1~f)!(-t_IPzn`Is#@jJOl`>;yie&BCUemYNIfD!jV@4tIt4=vdyxHe{^P7ZD)AW@ECgPu z*`5q@{*OP>z*86on+R{NqaGWuIzD2OGMv9Ir~=wBF$xptcEiP(WLKL2Gsu*A4aeWB zaAEf3aNl?&AJgZv3|I+;M1*RFj8)t>EiU%)Z*;OURa!-}_kOgvBiXx?PSMhW>ek(+oGxqCg1)R!MOkMCZ{axCR?9#4R|Z^7q5hl}iKZeQXqkJd6e6zo z79)%po|M99J)?4Owkjp>&ZBGJc~3k!iB4n1wAA|V3ncJTeW(|SV1vfK16UuJ?sXqB zY>Sc(FrQI&BFBZ-P8i^U)W0~;;uMa!N&AnRyP;A)w(4PvvnCo)jIJN6qX?J2vR{VN zNTXMQN*Tu&^OCFBor1nePPVpgw31X9yqt%+?)0aHB9t$uhxcTcE5`iW3-9TanIjZ)NsEF9YBB^32(9~O_mzWwi_ zQ{95swo|j-`IOu6pg!1DY@6}o0%2x+bwM+s(P&Y2_c|q%_V}BpRkuO|9Om{#|M<;5 z00#Zb1g**BIn+;3h4l zzao=>@)4^e|%hnPx26grTVkEA2h_4xzqF8Yl@qS;SH!b=MiE)r~*NXH7b zV>o^mnFa;5E39=wa7~_0jok8I_Da{kFSR>0pcY9~3Aovh5s>@0!-R^MGh7__9a12Z zW2T|J18Suc%u#Oh!S~ctNRQ6jm7xvUdsB%J7`Id9r|49oRMga7>Rgr9*hbIOuENFd zGag2cg;T|iBpu{FreJY+A0{0?hM6r+*;b-wzcxoC}#&Jb%y8{+q3^z&rYLV9y} z&}`{+uI#$$A26(r1omdcf0kR=EN0l}lzY$T{J0YqA-t;kCr-C}YfcmYgMIp@lg)}g)*Ep*NSbZ%a#Z=B;I}y;hL?epV zqI&v5(`S0P$s7uj2D_Z#@A7oRXPR+Mc;bHet<@rEPhaEEIASf41(9=;s3N+JoRl{l z_=||XNJE)GuII=XzP(1Lt7DNFjJx;WR-1)2rWdixY56S8Dg_{0wKz1T9x5ofnp%p) z&)2J!Qg^^`GGXqaGOpUvqBv|!QsA~UJAWe{ln0_%>!D$YZOliYy*#Bc;ajVHcb=F; zBB&66iyN?hsXb^tv0>YPGDb*W3QszbNUgST1q%{@d`2B&vbLj71Ros zLw4)>)ymiA{|C_0Q^v)}{D!3UvhC}janh{G^cO25tixiRI#NfITHaHCr$i@T(Ldq|eGUh7+nNK7R6N=Y9Z+B{ge zi}q8yFm!cS_~2rG>;a3oIrmh@P7351)MSsvV{LD}R>a(a)*RlZ8YEgPepKX5?rAN} zEX?xpcX7BY6-X!>-MxN|bHyv+FoDx3^z%sZyYnQGHJPy9@{$9Xw-ZemLS7VVvWxZp z7OL{y3GHk}cf9x88KS~{lb0qO-(vj~=B24cA9Gw3s5mI+M5`W_$m0dsn{u(Wo60CE zbB7_Ebqj~=SLwg_UV|%ZeAEM${uh!f4%fHf54i0WEJ1~#e`hTWfoTg~>PbJg43<&O z*2tYS5e(^s#hBD?mQ>5*RDuoMvSU!0rSgtQ1z* zZS^iPcPJBOE9R$AzUB7BbOZz~x{3*CI7nHDIlN5jky&Au7oSB0aLb*^USAHGk9nStRX!3%(}QWv2($?d8`3pC!>&(DW=kezk-|Z@ z+$$csZ;gH5x>rkuBPLno7BqHfJL%xDW7t5mQ;>V3@msxquu06;qK!7N*kTwh35yre(=lCy>!atKS( zzFzY&8fwL5#+g6%`rCin#epPM@_492@-en@UW5{m3>v>RM`4w+^mw`?nP!Ibni)tt z9%6)jOZyROz}ifbA-=*G!r6;)Fvx&DU20ZzjyaTUDQe0(UT&q&JpNIqp|pU^DxtYE z=PQ8_WenU6PAlmgF+zLH+1` zrjsJ5u+HTPeFMUxZQvp5;jl$ z`z@Cui)wnbY|?0oRkZIdw)S|H?~_SKN|+As7l$N!OW|n6I`rqCqe3t|x?+~BSPylI z_R-fc|G>zjwm?Xk+yxIW{T)b&qDfUN2*tie9mnxdLGyTTor2fJ2x8V$)Cj0POb^JX zF2X?5_)fG7=1E3n!Q?k!y{p+iu>pG(`V)sJUG6O(6X0GRc!WSfN6B;Z$4dvWgP)5- zQGMuK0ZUKD+&yeCW%`w~NoDLwz9{FA3XaaaAuOlA+o&o=h2CN3HRA#R3#0w6!i9tF zi^K7<{)Jt;Z?Mo7J-@Fbf|16jPFlK@hfW@kjLEFxmjk{M`c%`mI9m{)NWm_2cYMx& zW}qRSIDfb-hqggM{ZiurhlRG7Csd#LyRjeK`*L?k4SqmF^&6+F_0pN?ynzW5g?-oP zsM!m0PRvfE1kD}z5n4)+=w^cGM5^ooZ`kY!5C^!fyfB-7Kb%ShN?V>t32AsjFTWx> zT~g!e@Zl6GIbbG?A!hrl)pM^E|0yg!_bN6w8q*|~r^ZitYPkO0Jh!}Wb-*aiSjaFf zcBL++h#o|eRXuSQPOSM=-g|{mQ!q-Cf%dbRQjdDz^uui`>ADmH*!cSllMX=#nX zt#`DgHbe1+F2DBAw4`vx`GlVqAX$~L#vL-K(0>@MJENHuX>A^mc6nKe>50D2-`lwvg_@&yKV*W)*Qp>8(mo+su=?JAe#)Ept zX1l3Y8ABI$@VcDw=4v_vbCOyde|%F^e@llK-=B^7ke7{S#KUcx^;-&QP?jGo2{9t) zkTI`6?KG1mXu#}(h>Ka!o31%`TypQpXfJq(dhpUjyXw`%T3vPKpU9=i_$uc#CQlvu zE3pRZ47eB^+sSz+T|s3ln;&^pu$H3C!t9k6XQ^Ezx)LOkO{MDMAnCT&Xp6I1?>k}R zVdET~Fh-nhJQeU(^#WDCc#U1BfK<)E5v6da(UF-})ciB`01AuHjBHEwfP{2Rop}-+ zY0hlv{+&Z^^&M$ZXhoynf?g?2umjci+oeb_%owmy)#z$W~hu~2A)6J0-GX20x>0f8VB|m%>By@?f7fo z1PDt}Ky>=gB7oPzK6HqtI{mn7Z5hWY4(1srH{IHJ?hOojz!uJN~t$Y%8M4-C(;Azo=|@Y2(AqtzucZ{@@V1=Q{KMQ$n05jT~{ zPFR>XX`c8yL(%ZT03kdvsO=lJ*KSJg(QZL>9#0UE1?QCNjIxxr02~1QrrF{wuv&<7 zmlLCe3MSjh@RaH)ii{T5v5rLl;irm(Fa|wsEac-$>HaEYTvZa4F-R7ZpNzTjF6!OQ zD>^CkFnF;R$!UMrJrAS%4v_QHw5!3nnp%D_L?TMRWqdKw8VmgZW3-HH=5&oozOy<# zUBJbp6j6op$uZ1(-e7)z-{L)lD>_wjB8!w(?v>rkVu?5RUqw<|(dev`s9?E-%;~>z zl;?N##p7|`ad_4NSp<6hY|UK*RlJCDzqDhm*HE&U7fO#|Y;<7V&Uy>A!6GXwbh9qj^hO(n#;j{l$Hk zZ|$qR>*ffal21$IT2Hfjg~h$hpZG3r=QG-Yd!~8Uoum|*0+4C`Q$ZQ{*yT8n6{qSWXhzsESIhWS=&h&bB8#>aJz*}^)7)dU%Tz2+s@{@ z-Qn;)b#l~%v(NOVpPm$xjNdc!ukMLrbm=P9t>U#> z$5457&7QxALUe6MWF+BM%=uq1yz=8r`5*W$f?}*}^epX14Xa#(4Iu0A0$m}fxXjY~ zb?(&c&73|T?Km0@tCsV$`w#*!5hn&Tsn@(@3fLcve(>UrsbEh_7(bcO6HK(iLF;mV zT`>?I^3kywAyHOhr7u`IR-qBLC za*)uYd&%v6O z3*`BmuMmRxl=eKs!FfJ!_22Y@Xq}Z(1TwbkKAyMnDlem6P!L4Zrl<(<`FyvXd7oDw z{iX*8W~T-wmFdT>RGS&TyLBsfdPUWIQV$QG7Kd96@$PG^eZGefW5K!HmyPr$u9Ca6 zLn?~B)lz83Z(2%I$tMbyA8HxWm0J91T3+o}_|%|pp!XjYiia3tii|23V^R?{i;x60 zF#48v6{~DaKw|=~Nd+lB^!M}K?WI~oDFJzi*=eVeyZT3F#~cbvae+ks?ODPO*fp{O~1q1W}H`Rm~jdlcfMLpbg3E|eK0 zebPEJ(gO8I4r4pg@|pP&(&l|x3Glrag`iqrWjS#diBZ;Scz9w8S;^Mg1Bd(^Ehiig zFdir`K0bMAe^h=@S}q9~TloW^xKmN(SZ4Wmr>26z0gB%q5;_0D5)LU(sqj%G=7&~2 z9S>lFI*}PxTgd@IQ;&uz3c+HFHbk&@Ynp%-`>G`(@nUOoAL=NuXS>zBRWik*`sl%+ zM$Aj9aM4{MT9ZdE!WpD)9=0sCA?47!a81Q_U2WmX>^qGk~w|BtG#jB7gT z`=&>OfJh7&-BN-~a&$;{gM>6Pq?;ii-Q6J|DcvE`DM&YrZln=-=5=4s=k|WPm;ZCl z&hMKiq%Nx#I-wixXRt$wdzDM^2V&5H@$8msMM7SY2<0n3f)H^Wm@o`#)*z0RI)M=R z0iG4FmJW7j7CN6?t)`ciwP)Y{#5@J(mWFD-rIt|W;{PvZ7C;H~3bCgDT9))LuPy$} z0~4k0`DGSR-@WEJx~CxG7M14~Eu_99mtK2`84I+317GtQCqFMw5A-r%=!fLO1g?;T-D|>m7*C~a* zPM}P4diYCqaL77(9P6@AlL9KT}#$Trca zD19Gm6)j_{3MjJH-io3J3+`?cy)|;9ZR4O+)g?kj3>jJ~>upJF*6Aez!Rx?5$U-6$ zjZA!7l~z3iUdWZ&32$kgtIE=q~RkwO>^x8fgwTKwicZb~+s%D{oW1 zn#m^s0mQPiv3}!iYmJR%xfzsDd~~5$X>`nh`hNLd%+AC8@ zTSt4-n+W`=MveF>a<7d-sNIdC>CGVSL&!$UG~w{-iU58(OrvnWHsMG_(^d%3!jD<) zL!f=ez%|*`@YNei?z#j{ygYc+q!jPlH&cD8=KPr4DRUfKO#N1(nKXfpny^*ySBS2b zK<8W>av0ow6&FE9PL@?tY7tagbL0|muqr}71bpKmfjhY-!WiP@S8VNBQG9ew@4T@r z&gvZu7JM9%|HD-*1|aU(#fgXojDPP~(LnKFVeLQ@mAPU)HU=%n2oiX%_uBw0|M4{O z&i`-}9=Ir$V_gd~&cQxTgLn=32!-ehRSQ;rVjK!LtspzN@`6quDH{UZZC-qa=WDqSrVMJkCf;ZuH3iQ?Ucff5~J$wo504d#|oN{p^kV&t8O zKK{nn=fP$g^9=nQvks(z`vdvic*3~fYxE1Ec-8!kuzFzP; z??K^?`u`|TO)n4|Zw2S2i0;2v@2FI+CjN}rgF1Wy`nglDA|D!?3=z|+Ong@A@=q_nxS^L5@SJ!I}c#S^%qE*RtgwEnAv+>^}3 zS5$}=th*?+a())In)F&Ip5MLyDC|R3-iBRPKImuWFSk$NLn))JwZ}&nu&|#$2f1oIT*-@cJMQmH}B0#P#V@$;%@cQd)x(jD|B?P=lG9bRjrUbzMDM$iqY-W*@fZQ%Y8!_bH6MUcPm$C ztG3leTHvC~^jZ}`l{p{L!}oDsm=1zPyHT~ucUDU=?}>T+hst z_PnIr$o7PJT_S67)9m%8KzRhs(MRr4!ZW^S6Bly@4egWsF877Y9bU~2XJ6(WxrxcL zD`X@@!q+evR9r9y;c24c#2(BaujFnxe%jduL3LZ55)I7CEl+Vnh25?6wM z4gs8VlJh%RPkI^h@3f5hh=%_}_CcTJ+BGol^Kb;GScb1Ai7c^;lu@I}CyqB|rx<96 zEPf~Ka`ia!8*woB3N;~%*0280>1LM_AMAb}R8_3~k49>mh8e6D)&|=e7uzO!H6$*& z;Vo%c`y4N|QPp2`1u4Zbn-7JvND;w(0Cn!mu(di+MLPE z&KX{PxT7?q?zX@HgGCA1m!6J$0ATll?#+T8<^l=^(t9qG;q`;@zUrGb2N=0P@FuEu z^~QQPK&)yo826gL!~T_M^*s&z{Gvhx{!@{yTILMh{ze>MI}S!TQYL>H z1kZJOR9i~wE)kvavzkiL%&<(cG+}r|{{=gVGxp|wHBNqGYSh||TL7reH@1%qCC}2?JVnB{oPI)3(+k6lc z$zKsSkbrln8}0gzx;)-t1)@4A7&4O|L-Rt>F)RC612cC#O5(%a38rIJ*uBd0y!-#^ zwc0TOzjylf*$eppucJa2h_Hqz5Z-vyv?CfOfQoRxkh zh~7;DAlCHP;%Il5IeieETP>lv5QBmu+)$kzUco08S+34%8t-56GoL+u&~;ljp}Z{U zQ`N%s4$3ZV*F4Q*a<$Y0-#9h#gWf8zzbjtvOvpN&H5bOcphF3JmZdfPQW8 zsUVoCF9IP+l&oC}(~qPwfjum?s`&8lelW#j4+2Vlh)FCPNLSOyZSoIvcQ#3NbjQE4 zz>5v~+lq>YphIlrH6x$fUG=mWk|8FZdN%cy6;?t5@n7-43wLL+b{4Ef6rogi?%k2c zisou;+wKTynlh7v|KA1Bwda2=RhOy-osgEA{m=R!MA!Y!fSK0p|Es!@1j^EV5IHdo zYJ+8hazxC`?WOa$Ex2cMzhdB#+>OT*T7TEBTW3K^bLZ|G#@ik(M4)axslTWj&W;NS z7Go-*K0K`3sxcsb&EbT$wUOa|G8%D`rV?wB&xmcq?6n;W>+e@lD8Y!ct40u; zw0(oEP9eNddqv_{!V#gljZ1yKh>#c%6OGzY@vWK;7!&PXthDGr_shziQ)$V=#(MG> zAb>QruV5yHYbdh&&|t_ zQ{jn?Wv0wkLf*~nznlp}mV3bnc#l++K^lUj4hgi zD9aAM_OW?n&=Vact_qPZRuf(qH+JI67D?!VJ?zCeJlvD~0GT{H@8Q8HJ}I){0MEjJ z7ta2_v%F1^w@lY(&7)kuj9Oz)5>Dm#;I#hefP*l1o@?sef*N|-#wxVqVU{A1)2!v0 zN4M>Cq%mmg$c{vSTh$DiW?clYL9vZbKoY}*U#|0;|27=Pzm88I8qf4`^4 zE0n1rh&j7$B5XF|BttGx?ei{@@bgW{g5zXFkWuj45;ep^8NXj}}9SGQY`g{b*s>G5pkg}GC6SvWo z-q~UHE}Qy&ZPoKweV;+lTT1ZG9SC8N=9XtGt|nKirkYyb=BC*kiLje|_h%kZsp`B= zZ(;?HEatFPSmF3>wCw?+=f|Gg|1%J`%Un=Vqio^Va)DxwX2LHSivweyGZfkwK%4bn zz-N5J^*RK-dH`Uxk>+`VGVzIGSh|Z15DLN?H{%*I!6KX=dKRWHiJ~QSENLqp#O3)q z{-T^N%uEsp7FVXV%Ed8Pj|I|+`4uev!`+O zXOsqNgx~+a3n?Qo;4ytA(r;4#U(x_!=a49=_9>8JjbL-ZenSi&Bj<(d6<8MFjlQ;y zo7k!$4C6MYti~*Qmt-CX2G|_z^&!EBK4``>FRgq(8_^PsS@XO`lrA3hi~Zr$V8BvX z2#IHTRY(iP#Sf`R&-PwPD4f!9)k}D2*qi zDr*Lv69PF#d>q+7mm)D3hh?Vx#$E`&Qyu+;58kbJ`>{CBUpyd=Z>5?Oq{V*%LjkvP zKb8)hIt~yERkd)IW#TeOXm3$+e0O4q@pJqX50dBciZ+ll(7T3U;{E}WnQo*N2KSvm z%Z0{*xBL)>2a~nl-Qes$-D)iU)u`SpvUA~IQWCy)f8a`-+2NEFH_q+@0RWSc80Uwl zQNW`d`2ttOnJgU{gJ@dm#MHhr5>2P{>r;>pUo2I_UzjOWO}s4Ik!Xs#ezOf+DqQQK ztZS+_gX-tE*JaL>5avQBYQDc@AtT^xZ0+A|`5)P}YvC`6v+(hgw$#6acrTrJP)swb z!{14Ap152Fp|07X_{{Wc{6UD%yw9v5=yPEdAf(V#&`VqhSw9x^S#tPy+JXN_(SsU- zCoPjT7t_Ote$krSU|~*a=3k$5bZVwn;@&ee6N8C&j8a@*a@n~Bb19ZfH_b}fi2Q8X zZ}DzA`l?Qa`Uw*Kg zM_bXT`-UF~{Ftj`+yA$ru;(`RrHr}XBDovR}fl>4rnQ6x5Gw}}G2nkOg27;?j4h^V(KI+Wn3eG2a@sWn|aCw+T;k7(iH!ho4ZbJw*q+MmS zSfqUDAMO9v=lxh)|UZ54pR`Z9)z#~O1Ba38fc6`3|5MhsWE@+?SB{S z{&AT}5}EV0=gl#J8v)~-h|IKtIy1^QQXa@$dT{8V-FT_;FKb2JdCvtlFE!b=KI#=e zCxM^guf!``JwJUj?+I=``S7Kcc^tmhTXs%DCZ+#mE4pfB%dsZW}_%KLUA24U6GQ_%p+C1uC`)~e?Qhew5D-dnTY z27u-%J1L_%rA1aHAG<7PdcxLD zt&09I;QI{em|>(kl>yNG^U9QK@!gnetZdzL{vVv4OL!iABeB^{vn^%~fj3u4Kw!`J z9XqS}a9%9Nd&B<)O9+{T?@r$C{}^cs5UD-Ypj>Tyrqt?j;DPEjTAedII4258!0qKX z>&63wLzDqUb6<}xcIPt$tuLvbzJHf6Ih$dFjG=n5`Z=Cbwm_{7<5TsRkI2@X|E+$_ z?7;ZNZ1rcUSR@=%&c>H!1CTLm9lq%(H`#O{29$#^eZZtOP1I|&RCfGrmbdv1Ml+v? zde?%;T{>g&p_<3k@tVEO;V|nBH%k2zncUxSOLnSok79`qHca42Uww4GQBngb}DU7fN@ee2>ph@D1A73I%*p|l=?C>6hzRGV=?94>t zOf~iW9I4Szf3ybZr)_?QNz#NSkDmlL_4V(3qXIwi`)=29=38k1&-{^rPyu^x!n_W(_%N#h`dDhI^GL7Su9VwD&mgi85{YvTa+b#+x_2Gb-J`gE*MdfvbopcvY(cY{HI z;S*~+&K4}VX(>>uAahM%)MX#(yH<3d-l{CZ_V{`A7xU$RjoMulw5sFy6>3FwZ2vOP zh`~n|6a<}cPlWYplYld=q>~|e8!UjdD%B)dZS15&gcQQMSqc3BkB9ZFU;4mQ48JuF zbTm25Gqi-4(0OyK+-#?q3}1Ktnf8s)8#!Z0=Pq?XL9?nOmJ0^DJJ!94t0SP5#V*?v zf6Ei2z7Buq8q;}l$I=`N@p<#88?C9`+>2u%1OU`rmk)`BU5_NNNM^341=IL@pz1*9 z@Of83^cSo0^f#AH&uyj>nl!eqBpM^-H3&yjJQ^~2R!G0yvMeZ@OE zyO5~P(&+-jI@(oYca>J689r@{NLjKjl$_&rRGs6wnX=xMpDAxR;b4i2H^e##^j^wA zW6d<}K>&Wz7^MF1x<;}$Ws3~snPw)q8x#G$(;mbcuzgF)9QzaMO680x8aMk+)YPms({@z@$sm80L zp=5HBDc1g2E&QN)%*76=(Lu2&O4asPF?54@yS2`jXE4Jye;>-oaHhQ=i{=_>MQV@u zPp(w@s`zgBxxjMe6K7sO%%U-lqQ0iYQN6sDNYdE14vR>nfcf60bNN=r0z{HfW%88^ zNz$i}5K*6r&v!if!v^1i2|YoPKh8dj=tgYE&cp*>xiw#WZkPQ4jO^fY!9F+7Q^RX` zizPBQ*70BkLN2TgZ7qfZGZpSa;HZ@bJ~XckQp};TLoY`W@4>CnxUeLC)NUV+Q;orh z#EY%xp?-&$PeN!B5r~#&kS+SYc*+p6{s)~N2vU|PQ`qvTCcfj~b>b9Y#? zQo#5Uc8e~n8j*O5VZ*M!O$iqz`m9M|TMj^BJx^rvh%)zJM)el)>9ntoe1p(XIjtN#K=23cTZ1C^&xl=uug-t- z?#WwTj5NOE+L;>82xRo|fJ=By3+C1gbM_|}58u+84%D!%0zYqvQacI?ISy_4^$>o- z>igu1wLR}SUaPQvJ8RX!sQwJ_`CKYbMoCbbOc{7evKoMvY2%_a8Eb32QS$#q(<{t5 zG5(p3e0|8jJDVi1pLR`+mRJUaiAw07#?Z(D;ERQQDgYMlPq>8j_D@TRRbVbr1!~t? z%5{EJRi-w$T;t!$-3GP^l_(1GVo`zs@-xW~qcGEm)V8KJ9&qW#7UUceIVS^5@ zZ#aT!znJy{(}UgVbuf7Pe&_r+Cn@3S@ue4>pC#XlrcnG~1)fWL>BXN%29%n- z$>&5R3}L{KRi(=oZcR+stzhui5g3n$jheFn#v$stY}5!oaV~sYH4(29g{f2G9+E1> zl~Hru9}0TQe8P%zkoug&aC?pYe&mBq*D6a+)DzF~-P(*-)A)S9tH=289Y1ng5ZmfQ zVo9r(bhKEo%-4!tGQ8S%m{Z}P>?N4{bLZXHN3W0S=|95$JJYBCo$1p*25bL!3i=-( zr6ovJKjM1_egch0pUGrh9YNL@^S}}DUm)DlPHZ&XWjubyjfVJ>kiomb#vd9d2&E?v z(j~_S$jA>=Xp%OVu`9Ti_j!S0`Gvl}D525F!$=%bQ>g!GOTX2$cU@?t)kF9}<410V zSvY`|NSNej`9QbNL`vIieJet*Q6D;qVL)LLILExuDX3GgTf`nIBUN6GFAU`|DWg7e z&S)Bm^XE8Xibv~=6TAUiZ4PV0l}_tqJwUnb<^pigTRy}yP^>YYG=b00&uPnX;_j6P zyNT);AM6rgz*c#>XO;AnyZCUNVB+rgp#Jy*?*v#K>L}F~NUIgOTzY)<$NBN%c_0cK zAV-#_@afAPRs4oo@kfTNnNMx~bL0U&`=&4CcW}+wKb$Ymd%d(J?<60(eY*0qx1pex zQKh@DxxXeLbsd^c!7=7;9D)$S42$@oY9dm7tEs5=nIJ%xsIf1qV+NL3j zsd3rxJMp{=jv2ONwI;!aU*$JZ%+p5M>v68FL;paLH^(Uup&u|@*_tQj{sRs9iqD9( z-eIAJ%`>Q+Qbi{g5tRXZbhG#DJodpg5t- z6aI~;y=niQ9x7T(Jm9BNIs32>t6we_EQLIR-F;6AfXLN}5?a{XJ*Yms$3Z3|G= zRIakePS4Ct2J9|Q+g_Bxg_dMNpiNu0Ez9?QhQpfRIXQivoNu&E1BcLRo-m}9ZVaBJEjs!T8F`YSRG++s#9GAS zeoB%UA<;a&lkdeT0S15Hr*58CkS3>Am+%eu)Q;SR)&5DJgx3Ignu8 zmAo8g5#N5V8S03_VXfMY(Hyw0hg;mxe4n%NX1%1c)ozs8c#6zeNbWTUCGv_>!U=mn z%7sbv{q1h+9VEtbR(8A?aeE&Gia(_EdHs_8U!V4W2&HO-T@oK-%z|$aqC;z~lBJ%s z{??J_IwKxNN@ett7!f@P3^8vMq=u@0fY z=?z@rs##W*`&cvnuj6107*n~_I-+u=U4t&S6Aj#48KIwl$DXINaFqD0CvXtb_t5tD*q;#` zc3lSv@#5aXxo#325ce<6h60I$?eKB$d8@)y475Z!QlgHce!_H|5{d$a0&)8(yIDvaP0+qQT*R)W}22z9BVV) zhv_7j58`0gmkF!Pk98{y+y>U2?m|CfzJlEGTeIu;62jNe2fNj5_y$H92s7|5t7^z9 zdd^R`T)>YTjAY%=pb1p4ZouqD@CI9f9FJpsx^9OOI2FLjVpfjKnjUojE)2j%P5H@; zPYyb*`!QO@;^emMbAA+*yF6*dm^wm*pdT;Br)%j0ViAJOevKAlQOZ$PyvR}9Y4jWK zcrKYvJq;%H^$7_kLg@Btwj2b3D4oxV*Z^mzs%y%B%*nh}&u)=9oTB!zeK(Igi^3UF z<#OjvGwMj96sAF85R4?)c>Z;KTUKW8)rW^yHzI2>g(g`sLy$&cgJX%5wn!@8iIV+* zp8}ya^jKbXBBjTBZ7J$f=@O9CT$9$O2KkGHwbTJGSChpfyG8`5UvRf#zu)g*J=Km9 z%6xLlJ@w@3#f5nCYV6w(j1{eZ7EE1}x8M#WBi{CQY=h{`pd(*5<7*U`=5@Th{N?r6 z=6e3uujr|JkoNDqr3*%R2CZpkFu}f8O+l-B{%SF+6I2SkIPLEYp{uq>i-cCIl{Ly? z8yl;EcVl6k@vuvt;mU$+Cjk_6!DYIy$CRl7UiW{~xjKmFbWUp;;+I^YG~3j_JUvFS zxt5-BEl{0ci$|Yc&J93p;q8XPLMq2EYTQ4v0GRo)`dlVv61gZL*&BbzuGyJ%9ek-< zMgXeilHj?}qFu{B+HGq5G84ZFxWp1t#J&!8k9B_ocKmtg!$w0q31&K$_2Ajb?K>v~ zJ&TnP>0JddY%gK9>(ZSd-ny5=o_KThvRm{m6{Tt!@DTDKtN7EcmYMk{CDsazM1{#8 z26`wAE$~6 zhjRCJAVvXY-=~e6!PfymjwG0}e?I-F08z;GuJ&(oVN3|+-bLMGd(HFBclt)36{lRz zH_zF_k7jgVqm@9j_frNJ_gl1d5B@qv6~#@jOdyJa-?IM=2OtLHTg+h9yr!XUk}G>K zR8UDa1nYwxI|WFVwgQ15GiBsgxloa#g=X~=iKmv0Y2e}&kG2S<-tuE_`*>Us8GN{J z?V|lcZ02_B@XGRKJ4Yb@dnIakQ|A^w^3DvSEsdOcal;)0&z6D2xX5Yt!SW+|xwA_5 zW6~T6Ey?S8K{rGFz))~aHB&q4$iS33v12$T71*9xqofz*BrEu_s^39ABl|~|KIR_(d9Pz9$fC@>KSlx$3Vu5Y=mx@@Idwe9R{^cx=71$_bLF=rE!M_J~f zVVKsfoE(4MkdKV8XaE33w%k7Sdl?5x3}7-c=&6EN5_CLqCpKzUtUiD<)sBX}L{<2R z?_>d?ds=~|YjW}?8AipoOvxC|dsm#`t0H%4g#$uz2lKRdGwn|auU2rkKntJlhwG7O zlb4+hf7iAd^9o~j^Q766gXPZ0M_C720~PVp0lE#DpiTi%0W;LQQKc7(qm3c|8N3wm z_sRZMGuN>GMm0JR{=u|3m)HjXKJ7Yn{7Lm+j3haPh6}&=A#3Sggtr)*g$y1NghCE; z;UOUPS2H+)Bn0#f(VSm!tE4chsSV9;JjlL>ij?570dCWuA1#~j0e{(w62Us^)iN#v zg&ZlFUQAL30?YG}Xp)5(_B{}4Wd7b{hyM z4Acl3^2zO4RaU=?Qh)*O6fd9~FyJ5VO^Q$Pv0!ocg%=>k-nm~CD8yfTeFTf5)5?#R zR?>bevFbJ-rE_jAjWtoDC+)vki(HLZ>sXURym#@D6)>tOJfTX2?-ODsp7WDUF4szx zxY9NDg0A$_Q!&h;4=HEG=eM^pmnMEZ!$Xt95JVjA_o5~Pf*pg(y=CB7A>?=vQUNDR zaa}Xs>ZowvO-C$)FfCeU>ou`$y@1xggf?@yIGIf9kJ2RaKGBauOipiBO}UE%eDWu3 zYIVYpI3+37DsrJ^liJg{IQ{w+%&P5-%AKFi#BcO1Z<3 zkblT!VY9mj-D4To7qeKPMGu)1qaRZFgJ116D|txDWi}KL5OfP@$Vq`h;wq4q0^_G_yO@KQ`fiTSLVCX zZngNCibh$|SJCXI_pWX0SI;#_H79*i=~o1xgubGAOWXB0>Z=(2y-Z?1-YtG}Ra{As zgSK{$P5w2^GDnH~Ui(dQTW|g!k&OcLZ5fgcd%DtAj-19@QK7}u0##1woW2{~S?z|m z%H8v5(etyJO2A4x?`(zknm&A{a7d-r&j3byIQ+hLF{Sw;?*n$`y#ex;4fP1Mfs(Fa zLo{99Q8Cs?p6j^i&=!5U8W?gMbyp&(lUc(Yh^;21vqV;}r6nPL5i=UA>9vRL{ z^cs~L@GxxnodN@V*X=Z1mh1rnm7h;qQcI*VVb1Y_3PE$>5t0oSWHXl2p5&(&}J&ufFTL z8NBdVLz(5MJgbBu^XPdSs6}pOs=OS*Ot)#$PUpi=*+q;_GY;c$#`q;U)zg$CV>w>tK*E$ec3lEbY7asr4Oaf$qEcwvtM=Uu5 zwG>Q?DMLlIj$)&NdveDqRkUP;qpu+kuLR-ciiB7oN*zH_Zgb3r;P$&$h)_0rk|RA5 z9_@B)R?Z5@Y?Vqxzz6e?Feq^)0iL z2ZDU8$)F#&iM}E{7it>3%#FFrD`4S7*n8J8$(Tl4Lj26s)n6c+M3cY?)-ela|JRN( zlh~HO(tM+KKo~A_fW@#bDSj7tJ3ZuE1zM!TVLHcW{=t`sDuFH+cy@yySq9i%`V$-! z(~kNX%DHOVBPl8$ShTA;>A~WCTlDeezKhX2+_u{d7aL)3$8qc+6asEMVypt{MA()q z!U{`fu81XU%^iWp^J6|aS}EQ_8zP$>)A$kb;e~$qj;fe)dYL75uny$C1JrCAH|t;Oc~UFt6Gf*CHv3t>2W3ReN^xhRjG0jM4=C zG+RY^sQ9-+z+(Y%9&YOHO$@63hm!r98&SaxWC_BaNZVC@&jL`Qme`5uCv$;;wyJ1`cnZhvrHb_6T2meGUIuPypm1386Eev@OBh5gKY25Px6Q08m8Z#ZTu87QHY%;)cu z3n^*B#gQZAh=8K@!jiI=w!;O`YB?g70StF-{N$?~{?9wFQ3JWPSA6-(Iix#N;0_L+ zFGT%WnV-&b26!+fC`WwBB*1g_Mni}~jcGOjp0jU%TSNmVWo;o{WzEV+H-7xYWsl*4 zA1;(|s5>0^Ha4Le0Obz2TfjK+*Q+{qY4tTq4!LQAoVl5NPw8J=%3zpmPz1{bHCZH0 zyls{+F$uyw)FOByOR}u@H@ee4Kt`L;nYACB)g3JJ{Tl`whOFJ0%npOAvZepInGqu^ zI_8I(`RhEia_HNUy60JAU~AC=F-n~YBQFI&px!>>ux~6>KQFqh=<}je4OnE$w+*5p z)^PH>M6lPIc`ZRsxS@r2`)7k+V7=@>n!DnJiW*Cw}sDeZ6B=<;|LsQcN6bKSU{sOLt8wZOJ1Cvnu*(_K-0ymT*Pj=WeAD zREWr^4Zj^ak zgkvT-=qkT&rVgh8iZxV}fk|HwOPP;vKo%lSxlSQf9dX+xJHwF+I}XH;I8idB)tM_Mtmi@Xz2PPjPM4+u3~?f;jTmpf*Y|H&c~O9v>3 zch+DPDR9H+y70`|_y5%E>c?sGIUfi9<<*}dC%AK`*AX~0#ySxu7f>8Th=8uKN!zZ4t(=I9{vo~l!R5xZnm?t8JWF&G;X{mG@H z>dHwyBm1pUR_84X1zd3Y@1Rxuw+ln1^{M`!3*{8;Sy*mOdIRaGY5jW`7!zm57s)l5 zk@$$GMH4MK3Q4|euMK9)+Hz>z^mL1erp2yV7LXar^EA~trF6W4=kj$ft}G%X((b6s z0UtOj?!J#Z@^K6;=J!w;e0A=V7aQ|*+|C{DP>#@XlMxPHz(+m291D4s_#4~pZwDQ5 zZ0Wdc#40Q6mj}+SQ%ulN@CZm>b4D%HIyc^t(2XKu z`2rDeCDT61Q5R{wHYtY3kZ4mNFzsdu&&kVoAt*Z!pYcp>!kO#qN#hia`B)Fa|ECM5 zf4e>1eD@{BWak>o{!(TYzc9(w@^w3zj%%=;_eAsjLJ~nSni7&_E0!-AedjU~80W94 zqG=)tXfBEzmty8QbwjW&LQyDbej@*_OP_7*E7?7#Yof0oYH~h^WW|p!c6me7__;}p zn1Jo=#>1(uvHw}xd@=~Ig{0jkAR!2<`=F9nWL>2Wr=aU&U8NQ}DWVOuiiKJ9Ix@=c z7{xIr_X`l4VED7;GZ#I>WjryXstU;&Fivvk?m_mIh1eWw&!z%j6(ahsr3~he4=c65 z3A-tPTeW8Lqj@Ka#KCMOY=zKR(gg~;5G47GUjMx zgo?%s|E)-w<*Y=JIuH|^VgDQ={V_W8YrM|2*b<<}4U|{UlBrf&(oTT#ZOnbgd~&uP zE$b}K;5A;z&g79l8I2^AJ2Y!UDQ@#I13aYYAtVu9`v!XX#mmpu8$$ww4wjpic%@+Z zUYFhaPXRHOvA?t3Ga1O9;bWRRqx_YzgS3;hL$nZ2US|A657E;22xeC~nhk=rV9JHy zs*-zXd;nTh&(^P#L#t!2SMiH@keWYg?_5S+6BONV#T@1hn9as*cuZO~AWw#nhCbMC zWQ6~?*bfPxVHEKiCA7r9wZtjZ%Bf$eTWqkx`aa%(*C+}156-gYP6q+5Gj^&Iu0}of zz}O1r(t8*83uZwWN^E7SlHT4jt(>fPrCyYiskfwAAiZSA&(Miee`}Rq)Hf9jdXXT- z-FQ`E+1-pJP=M&VB=yv~DDInfmA_pVT~p!+FI4f*@dbr{F-sq2@H2Y}2;l+P?#;rv@pJh&^(w>`m}$k<(@6svbT&-d(1Tljm05Q%O*i67BqRiKAh~L z(&B@bL3(DkzjQs{cj=IyCy7nM(ZV-KFiz$`;?AP1DGI(xln6^4OB2rDF25Ah+<|K} z#9ywA8naKkTGQly(dRqmvq3a%b?->b;YR$Fx+(?!;Li|;ESN`6Tis?%#ael&25j0; z^+f(2+E2dx{lTAlmdaH?VM|DJ8E%q}S5DOM9+k+j66S}cco6O%Xgm*^nS_Cwg$z=%kVBL`J}RqLNnWLsRrqy-!;iGFxU6b+lVczAKJ?EE%W5iq?oTNo zfG=Ejv?$aMKQl8wM9@g%GM5&>KCC$8>whN#>ptv4KVc@$ zpElkm#AL(Ry)6Q6sti3%wl-C7 z^(0X+oBot4caW7eIIZV^=QZ5V?B=Y#M2W|k%WwUh{J5X>ylaI7{Hq2GEuY8a*#n9% zx63DZo&mmM&^||VQ}bgQ{gdSo{X>-mFYfwGBQB8aEf8lNlS3-;Dym|a$?UAut#Z?^ zfTTytKaO$tyFOTVZE`S+oUsrXT!Z~1H^wi}clijxelgYuH4)HlU2s%#NK;*%-K3Fp zUQc1^(lyb%x_0Mw`#Q?P#DOts(@uQ;%-l|`fVzB~M{NMOk|;%lNmN5Cu(XIZ4ddgG zJ<6kOmGZW!8nn{j*A8eN>12J{c;gK)}t@-!WUU|r#}ck_zj7COSW_QKXh9@rmFK68(@kdFv?ma zP(XmDj{{U?ssZ<`+=I3##36|MZ++YC{(+F6mtSZR zEbzyJetnAM^LM{sjhWv=?|SXfRV$|fms~0kLeT)=kHQ+UFi7llLX+OF#PXk&I63;i z#Svu%i`Iw;({4flElv$>oGGu*g;JViWfHFC>q-&nsj*6i(Ec2IdHl?C?I;aQXf;+m zKeJ7J6xs;$GrbLDCV%QMBm}WcJ#fkPp80e4Nb^hUBHguXG!gSC=qagz(ssE^5==x_ z>1xc6mCNn%t#H|h0}1rcy;%I}Bf;ZYeIHYG>d|XE_2zXmL(lgi$aH&qh((X5E@TT57>t1}=eO~g+l+5W|SObd@dPEX!a2;%^tDo2s;hc1NXjGT~d zSf(jvfn~5eLl@io7j|+9^MMr>CxO;gY{JY!cshmJgL|kb=l$cGO^FSjz}|CVShx|< z(L>I(+w^Y^e=9u@W9b;W1#v^Z@XQ4`&Ai-x^7_|^o8y$s$3gotor{D`>RKhGnTEf5 z+A?8DXWofh8+Fk?#}85Pw~ooZ1DXyHP%lyCeWW#!PFC6m;hODiw`Ln(i6d-ciZZ$L z!4IZI)cHl$Q))z<&1_yR-k{k3NH>V0I$Y6?dZ(0z5!q^~AHyMPZAhuzPOK z)n#_xH>3)2fq8Zaay^FkLwSOKn0R0J@uJK*ga}phSA6epUjXbuEN7Uc+~VZ^WCYR% z2|n#YXlD**kDNOH7`pl2c<0!M1Kv%X9LB`psHNr<5}`&^Xugj!Xt{eXd8fvS4pl8| z*eco=u(?rP)pkWI)%n2+r65q)0?8-d756sLGMJ>Yk4CSf6)vD#gjJN`+(I>1U&Jd* zVpP4^9B*c4C8>ugx5otED=+W~qzIgVwDcW~C9i+~Mkru}ay8BqL6dBe-{PdE9A>_5 zjyo)1A49akWC>W>JKMg;~~+QiI5Zr+>!jFuM9K4ky)sv)!fKbaj5bxf&{!?M5q zz9U(Kw(*}lrAc`^DLLiZh2G|hu@csQV}}GZELwc zBrx{2-&&`z<1^|46F4jReq>iyxRO8dYOZ!H`>ONezAe`VF+b3 zC4uIl(8qtclR-|hfcwu7@cEh$Td&l87+NeFzzK4E){2{kN;G4hEnAw@NM4!sjN=7r zWG@pJ-U!#d>M8m44|N^4@XL-1s6g$j+` z9_)6BI~gDKi|4L+~;*5-#t*uCJVxwdrJ8!@Ke*c^%h()@F(FS$N)^}4< zxyCPtR&T^ET8!8{wnda^u;HhF&{7QgTt0AP@-HyhG>k7X_Xrj33!8p87nOs09HE%c z=!bX6c8ol2>PX7_|9pM>(j*Y=slSS{Pa@g#QIs4=$p>{j?+{Vh%55HFdlzb+b(J6W zV#et@q26Z#nL&+HIRKY5e~e!>$@**XsHWTEdjR!E-iL`lglqMls8|D?w{{UUQUxPh9U#T z?8TIHklu+o$yazQv8(XaZxj68!_+w9&1%#Z-TU;E>10Rnu>(1Ez>TIwwYDS-x~&^=xrfFEmnvi~PoIIjJDRo_XmjvGO=#S^()ESoj$PLx1TWfa=?T?5_b z4$f~wTU6zk{tg**2;c17ptQgWk9kqM3eY&_8Z8xGn6-<^wIFNDfsb4efQoe>NlHq3 z*))E3Z;FZTzEqL}CTJ(iJV}a_1Lek8J!54cMQ8`-OU7q1#=-tz$KXbdft8c?Bym?cS{&z8Qn5&aoyeyPgRxhi=58p z-!=+crk-5usfX|0;z1noxUb{|De-Zintc5<*1vENZcdqL`l%|lZ2I4&dba}}wJ&?^ z98M!Df2_{V?zRuL9CUYupozhXnYA5HPcS9Pw6Breb1JLwk*|{Ehu#>WJ9dg{7ROBf z^{ZF+{%x-6%_>vU_%{hmit=m#!iM5tA;YTs=xRa)+oE|`a^jFq0@kVSL3=Ul((y{r z>CIN&E`-PL&glB1@#?<+619L*jeM}dl2uT@Kt+4{!FU|ZHQ2q>{9;A(?Y`u1vc$qy zF<7c01kX2oVZ<}&=gyv~9|Vx&es)I?gZ(bg`|{I;glcLD4AeURM9O|g4ksbvI-uSI zlwfpLFn=|$(Q%89aITkCeZqG;@Co0f;?s7^%@+1J!xq#O;z>)(w~mLor%u@3oV8J3BZ?@psK~W!9Z6pVkzoJMj;-4AD29G`13i38C2fLR!_w3iPlK!6 zANr*gL?ssb?h0czXt@W~^PzwyI;ObDQc!|=XuIg*tDSjzFpx&L_&F>B!3-j*+)>vZ z=mCzkzs%j9m|iMQ>w2Eh>dA}|ctLE6E}L@eOIB4ZIz91)Y)npxHSU^D=D7>cf|6iA zMEw3AUvCxF2ETlPwvZBN@j|fx!QI{6i@O%5xD|?9a3@%CcXudOJUFGe7BB8a3pf4F z`LBDv`*0r#dCB_Cnwh<4&mKreSZqk0&~uCeR9Az%z)&OiYtNGFGAG#)35U^;Gkm_y z2$>+t&Tz%bE&6K#wv(HIm~hau_JL6+*%5iTUSzS?+9lrG1LcBsw|EBb^KA*AGSgNK z9^B7p*yjIbp(VdPH5%>mT^sxxy)a0EU({`uzkK`t(_XR*Ye;9Qp%pG%$|Cq@xV86W zT3rARVbp|LTiV}1o$BUvC#>DB6OaK}6e#X-gGIq8mA}vQ+;909(a+)TnLSys9zi>K zlIEi!%5T2xpJWHmW0fnf?hV@nfqW^+0P&X&88ymCadmBDE;@X_L)m|cWjE|5_<-ad z70YPz-+pk%9W|-K#fon5$Xec;#2j#3A%!LlLbUU1oYYp47}Ykh+^O_FYH^LU22_^^?B zaS`{t5i@Id+PHJJ!v-rkz@(i+s2vbEjuRI)4d*9{QixBCQY%ZARMoW7=#%gh>edwT zkP0&-4w`RN)EqBIo@&EaR2D!QqFWRaYvc&Yh z1ybIr#iM^BK&C@%5SxDdnPv_HT2p*G1+vJAPMGu(pek@H@FAyA0d@O-T5K2XxoA8N zz9aK1i66v(FM&OO_gJpP#a5OGi?+S@T|`?a2uU3mklp0{e1Shfzt^)Y8^Aw8i*bjv z`WQ*&*=o2gYu%~a9N{B=BBn@)?wmDGtHy`l%YBvW7IBDsr^klw1uy&U*NV|*^oRUr zWly&TbFw3YjxrW&L+$0j?_j8Hp&Wd?S@#`t4gBOWX5W(;ohA{f&c(~d93aEb;xfhR zme{WpXn!Wm;VQBs@QCHj;DF7hp?VAr^XyP;D|1!w&d9_|T+LMl$8+Owa3&A#Eg zLzqhw<73jjEaOuRqz66=7pHq3%z_)brVtX*C~ zcYGe}P~{64D{01t1d6p@gEpa%0<0u5$SeIJ#o3K6N9znc?Uej&<@`|vPiIeVW=p53 zq>WUW>5uELgMr6^j5g`LCXNx*Ax$FCY;Z$`);v1>^$_aJKFZsjz4@&wt8vY~K&9d#E zi~#M%7+-dwT3uZCg&B_zs%Y1Ze+w}=;J|SYS5s8^UqL`#>jNm3>jtE#G`=m0oNB_o z&SSsg-1~VUAUo+nf$+iax8BoPiAVy)cINZ~2oQ;vK! z-{){-XyUx2w#zaIa(}q0@BB>J9tLJ8KVw~24m#`F*poP~PZyMT_E*d_(BMf<-D+Lo zV^)5s_3RKCc2&Q-TvH+%`tFEicG%%?uI;k+oC=fB%0emc`6Kd3;rTo6X}?n`OMEz+ zciUJeni7DnHQ-XyyZedN^?+>>kK)OJw&eLS#^6Z-JLQ@;<3|~ZHY$vtzNsEq`b*Zm z5ZuO&Nz+vTh=ixs?#K7;u{NW|FL9_#z~Ns(+Y625iM~5aKD~k1^aTd|auIpf=0M5E zsP%huZ;pku!B%d2Ne@^VF^1SCPHhHLjk1}^!8T9h(1e+sfle116w*}@+vjU~cwv%1 z{@crJveK5K3eI)#kap`3Eu`&jilklGhz182Ns4=>o%;s>zB zEzBq+CYl-V8xgo6O|A#Rg$aYTri`w4i}iEXX@21rSN6IhOQTIDko(Ppt=34$pHOoV zKe~1qnl1a!x^%fXqk2v0jFL2x=f*_<)Zw11dJ@!qDN`IT(x;MBG+mEWccv|T3E2Qv z7^!Fy0Tm9col{1D*r;&PVKj=RJHB9V@{l?JlR@_C73@YDcPPmRo%pv}Lo#F%=*? zq1$m}*lf0k8!SUKgJ~U$_yi?0uuiP6C#Bg#!Bdii@!_d(W1wr}b+UR?cQEW77Uk|{ zOMEfK*g_T|geNP}Jy#kl3=xFB0-9(}w=~;f9_i^7qQTiDOM-lN_N}2Qd+&B_7G$n0 zJU+3w=jdc|P9k`Fhd@HU;h6`br2fHMC$wMy4b^Fji1vEle%*i}^`h|P%ELtUs6=;} z)Qj-X-!*7E5W;oz)S#vzL*>5+kuw!QuhTU13@z}yu%!-f=fk5WD^K21q#oy|ztc4g zses6u{<&qLjoaZ`c^B4j8~t7)Vd`-uaMf0+mayqXJ4VLXlZ8E& z*AJ^>_a#ryzPw;Cvc0(pqLz0KHI3^^nBjwJ_Y1gEIl8%{Nhj!qaW#tqx)F{)6Dgb{ z2h44tC(a9WKc5cwh}`JW#v<86-M7zY%6{3Nc9X2EgT`iK;FOna0Z8Tmy(#gEqLdl$`E1R$Jwvn0 zhwSjTdn*|?mTLy|3>61HYFhmKEG~*Okn`y3qJeSH<%~ zX)K-0#XqN4j~$ubn-K?+=qK#I2C$_CaQRA+N4zD)6Nu~Pvw+w+Z99)%`ni3U;$iW_Qh4xX_bMqv9 zDS38V$!dJzN(Y$D4_Yz8#o)xdw_gs_muJ`>C0sP z)_mb=QC=D!BD199>_OF{6}GSdiJxRr-x44g`~6={_y8aT^q26@zLEXJTzK;`6nB;n zs2jVTtA<{@pl5Oe;%mi=JjsOro0SfMExdL?Z8${5=lqm%972%@$ZJM6LFpTO;ye9y z6$bPHpMLN*Q*@ux`lR{bz#+?sz}rV~Ukf+Oa+uH}9Xc-bFe*U})YFTn0u#jODUyny zEemTxw6R2RCZz9Utn`Uo+?OdS4<}?cMf~O7Va8tM?wV9T61-LVBF~0F-^5IOf7?zv z7$34Y2M2qe19Q)puf=25gtI|rGWf=Mw+MBCDzT`19$<-cMkJI`$vi5Qq`t)MS;UBZ zWKRJx>|Z+()?nx3uvh2GR*l4_$>w^NO*tjVZ-|Iq)x8!HL=4evefH&ptTfq)>ZkL5FJr4y(Fai^ znI9(BQNwCX86ojgSbJF`(pj=HH3fGqvDN%0ppl9ovvMHR!(2cDa4jlRg)tx$RGZJs zJ(hR&MGd`PaBjNVAYHI}oWDEb$;TdTlGZ0|CS^Pd+<_6B`8i&JaO+2LQU#cdi*NjT z*r1@w@rzIP*CW+t4aQc}5_C*}?F(^VE1%yISPq#>R{=ihIz$2-yY)59KZEwkb!7w(oqEBy|N z)!vfW{2X$|3P-3GgO>8kY*NnBBSZ&hf1*6k6{mO1K@}mLt}P9`k?uw<8m4#&%V%9s zF@K%)y-aEX$3!@)slwqEEpSihtxHN9*OsO3Ie^%l^d1fJ84Y_jZAno|>7xnZgVT{D zj#I8g)-7xBeiXP+_`2orLciNxQ}{-2nHDdi9X*Y|$3w{*D>k)1g*p^LBq4NuSqh!~ z(QYc%q8i-=vzx-~{qI;QB>HKbHo3UvQt6}4s>F;S)V#nPOz!(I&nSM=S63-fTCWvmSxtQp*CT6o}aOQAHD?i zQU*Qy0>x$3_5T9lZw7E;!BjDks>+0*WM0u};G<|TEvih2O-{JEvG8vle1d4t_&XP%or!p(Tk# z{Y?~mr`?V%(PyUKoL#QBPn_R4(#zX^Skhe_U)Q(=C~REioK_Amrl^Lv?P2Pj8pe&L z4P^^z%$Mw@TAXC6g2WpE^-+9>vR+w7dYdmwW#wDUf~QszK0Yn0i9Cmn`8uhl_-zJB zT1Cr;Tcjb{A1iCc$2m3ih7Bo?;FL)~QF)1Jfuxy(ln9tFa1tb95aF12urM}1gDWD`QPo(J>oPP!_5-STkAevAOtkpr>JzJm+_Md+sECb6Sq zY=3Z1<-uySN&4D8#qBAwac|#9I&09dowSeJcT5GwH;f6%n&}%P?&Htat^RKHE!u39 zB`!9JeMg?=cPiLr8}*QI-2d5Z@$UXS}t5-Zk~)1mfVYLSN7z_HOB z1_JOhyhnTmjxWAJZk^# zA^xwjMEsn&teUc7ONM}7aRgUvEf*~rOo+tDNKzC7T0^tsM{F31*50dcF~1&c}i5dmW!$F`ZBsI^@N$ zLDp>YLdp}N=GkW+G+UyJ^ulZ1eZx138h0nJW+@Selvfjg4^HzZ|V7o=~0S_oX|&)pWv#(fP8H% z!TZ7&PfMe3<8N;HhJ^{UB|&DBI9l=h!_;=YMD%ukVc%}}kmV`a;CXIc(I&&8$Vv0^&!SxJ+`b_TtYG#Et*!I!2HXDoq0Sfo9lrDROR+!aQ_HBXaPQGlsvAA9 zvDPUDpX`pYw?vRvnm2YH=jB8T)pA?;pbX8Q6%ifij~^vq|Fu-i(iieG_yl;Pi2$LBP%;475S*kL_v*GDl*ot zE3C7?sHDo*q$V-$@i>Cv~S`E-=WzE0tT#i)9R zf{tGKifFHkBI|A2g*@SS9hQfaAIBy%YL@rlN*Eu8;)qqfRXKo2=u73MzE+;F8h+3! zYK6iooL2ONbh8>i%Ytm!K^D3e(Icgx!X@s;OZ3fK)g;J42%bO(=Wq}E71u08odsX2 zN)6*lXa+L($LjL~_yOLnq={xq(*v*1p-w`wtJOzBk}odBH4kaG(%pCAO{(BmWL}rE z?^)hFF8`BWNK6146Jk+UQ5)Nq)jb+JAkOE6UrCU-GjeB%&&VJS1iEly{)i^GIqkwu zOPA^Wt%7FTmbIjc?z?V<$L=Sh2qRB>jRc{%>m{am2(=7YVs+;&Hl%zLd0T0`v_PEK zLPC&*rm3|%2J;J%<~fdiiG|MK!Ot&QZ39%g{1~$q)<#IESW1cVsXt85ZIpM2%>8%2 zGSshhX}ML298zG+{czcD`>Z*H-mYFrwl#>{fiA8?{>fy2LBy^ztS$C+UWbIs)R^ts zSLXdC1;p7|79LHK-bD;1`s;Y%MG3#b@ort*P_xd|LDp~r3!|d44v0Q~SCufoS4D4v z_#$2|l$E33h-8jZrd2;3!@~As!n{T^`=VYR2#}xwrHTYK!g6_+6O`)JxQ@sS>{#b9 zYr1rZmmRJAti`H3w&-ffrR0!7SS~Fr*Y_!nY?j#gH}sXIUC*LF@Jw(8R0 zKtw_Hw9f0#RL(QL6%?6=zGxO2yeeD+_(V_IM3>Ou^f$rpvML*fI6-V&x{mjgdS^_;IC>|AOq!vGt*mIyn!DBTq} zRWfZii-v~`FbN%17RKxt;R_cW^zyRe-tVTr>Gb~vH~;jsrpUqY-*-p%Kn~9Ocf#o?x$#zVm;?ZA3(Ng|t1)a19tW{y) zP3i6B)D`P|6;rWmL`>5*>swI;8nj*1udly@&af&kvD3m&*?4t7c@)}*KDwh^j9I_A z?2_~?1uEdC3j5WxME+W$)@G>0slIxe){* zPO~;r?*K?d(PIyiQ{BL&KZ&i+X?sur@__!W4#ACEp@!7EESDW)pAr`Roc0QF7cB^W zojT^gz^TMjhL_G;GFU5AgQ$x#%!@U^G;Z5Ck}D^O)dr$5J?N8yNG5t6A>mLKd0g|6 zpu7ye8+8Nsev)L~OQf`nT<=`<0%uv$W+A!QkWAB_f&wqip!|PPiwCfCbp|6=< zU+2bUMs4^ISmNVCL5ii@LeY7C4_(x7froc~qp>CJ4frJRH%~LAkezoWo_)DW-Y)_7 zCUBcw-NG3!n+BPlbAjvOkuXI7-A@h!#U&?grwInRqYd?w26zX2_7QCCmn=-UdD1@+bDpk){1H$>1hfa6|r%8?K59qnoIcN#ma;YiR*y50G%6YUV#5Ziz>Y0HeqHy9fY zYdVozNPxnMCEXh`QZ85=j8Vu`{hecQ-ZU< zf}I;@^Qg0~Pid32!?sZ)P#cAql{33e!1}w`lw^v8w9POK(qMuKjl%fX)}EvqRf(}7 zX5pv5XY_X6l_}`7px%+U=h$Vh42=cu>34}y_2AxFbORo!V@oNUv9=8My1|#fY^eG9 z+1HIdi^y9Oo}HBY7H~Xc#C?x$(gRn#yrdmO_?TL9(o@>;iX&CZ`vWY7#;nC8-4c@+PwF(of^Ul ztw+bZ_CC(#$b9k(L%8)UP7gOEfpwPwT!zqvo36Ljul}McPI6;Ev8=(hHsODgjZT(C&-ey5B`wowVewGhYB!N=J-_S{cV9L!bG()pc)~&!I9`%5^G(^ zliAMh#|k+KP0xod7Qe^caOffgakRB-Z6YuZSeLsTYQaQ41f4b zA!6q_@Qb=;583lD7C-Ci{^GRwu2R$O9}qSwyt{AW~i#_g~XEWKs`2?d;dgiHUXw0-p$Jz}y}g z+;wXS$7;ASnDUKAq1Z8^E<9Gn=5G$wv36CWcVK+K!dfTgidjVnWU1C#KnyXa4U{Lt zrZ&jl6ZYIK{y7@2=0qqLy;WEys^0-@74O+&;)@t`;0M^UrF5aJx(si-#Vj`2@Q1}< zh;7^Nb{&RAJoiXP)l+LKA|0!khd&!V)#Q=*DA|yV=msyaHTqO)vTie6et=T0#1(ET zOH@r!r3MJRpjR{yAcxWj$hoi{_18N&NU(QA&y5N#k`eG43XisBk6>oX5YM#wK*~vf z9r+!XW&Js?u&B2X9=eOY|Qw$ ze#zIgoAJh`VK=_UWkodBL&T^T(tu#>%(RQ~0Z!f^JdFhbIH+@Ukd61`Bfs7nm^d?)jO~`ILfS0u!k4e^YYqe-k-xXL zvU;v9rKI-Z9XLlqY0QIu2*~5yI(f2JUin7$MvZHs-KbQgq@Bki z?~k;my5QFE^i7v~w%0ST!c1lr1#^unox_&)+b9!>Y&r?@mqbbvHOv z&B`$C+|cEh4KGIHUD*st$An6UNV2MikAQfE&I6bTl-e!b9n4wtsJg|Yq;7{M$G2L@ z`pFzspUzJ8c`^r4!i7UYcv0#f-zyJ4Les)<`i<2%4mptH->@5~ACMKr2B{_VgXQoE zE96-}|Iols*@ax8r9>mBlCC?Bu7vFeq1cZ2w*y$~C*JIuYj4B&ScNG7lO^ zxnOPx;^i|5+NL?OrV=w07O7nkC%ULylWyT#G^abxc)FqbG8plbImi4VgQ9g5d!+^M zC9%6#x?tY;rRgLH=2HX9Kn8HN9GReXquorMLF%6-UnkOe>#>(8I`-QfZWBER_$Lp> zJq{Q*msqeKaXy`uJ_AHbH5mehct-?-fr(P_lZ6PfZ=OHzNo-?5pf@oB8<-wAm|q^d zLrGq7uHX~@zf%Wzn=g(W?6JId69X(}vNH!5niBx+O|H!bMnsXo9Qd7AdB8HX=l z7920G#|6`qrN8>xHWNRoqHks*_b$gn+CS5Sh^%8D=R(iMOldmY9B<)`BD&hXTPA#( z6W+5}jg0NJMo@i@zSI$8mZ8BVp*ot#8k=*t!$}yy2!+7!L{KBI2*LPql^Q7VaTY~g zFiqTs!f*$~1NB_#k7OxqBDEF|UF2DhqM^PhK%q(v*<;Xj`nwVeUDxR`Xst4v>vkQf)qW8HD-aekZx! zjHIg}U#dcg9yos;WAKjuo23dLa(@&d?dfrz6ory+K_+8S%X4`;9LnbABBp=7$qTDRGo?WlikoybJ{~m++q73H%oNl%WJ4$R>steqix(K_MnhUdJAfk8wc~OGy_c4-eu2O9nK|=`U@+^3dtekx zt8x3e-X01m^-sn@9BHiXI~UOtf?Zw@Ww1ZwNud;v@gEr;!{+wp6Z)p0fVwlR=wPV= zKM7`oI~Mk)TYJs7$yzAAQ72ZE&Cpf|@>dnk!2!iD6paof=}lw&8eQMgRcWQBkfd8f z0lG-k@itV|Z|!}z*X`W66`xdQv_E7fmxU;Nh4nAD6VUsAlJi zGb}@H+3}Q{OJ>aA`{MeFg_o)Jy3fRb_CJPaCjKFAO`Te0gE)L1LH*a}6yNS8bMPHf zHdL+YIrm8Dn}cYX=eb8Api213?uXqC&=$dBdGaLP8jVL+7`S-vAj89*JX#33xw-?aSGjK8N9ov_Ry{ak{LN18_41eb9ZIAPAq(N&Aj*Mz(vDY=2W@L* zrh|M{gfqwQ4tlz7VQ~aZ?`~cns!gh3HUPPO9I?AcZjqv-rY>81O{sCZjavO?|Dfh) z-}L20+o#N5*y{6479$?ECy*)QI?HyTFVE%i|6(SL2;g$;+iMHD6?FaoVKJNm?hL>@ z^jySFPJyFR;obj`=s#(Zw~>XuNQ~#A8kH^AbW^n$`KvE26J7jvd@i>O!f%|l;`1rZ z;s|WqLu1>8RG#Gp@P~UMryc2=B&u&S6fW(9Q-X3EYKHllXzks9SBdZFLtOO|@)%<_ zSh7WEteA~fZuAVnuhP=M7?s>2QEb@~Q2nn^3J1CPg#IOY$C2GZ+i3$H^u6B~uNM7| z_9>ccA6!TQ)(bDt)5ChTwW6nzgeoI^-j{Z7=zTG>-y2|xz{0*}O$qWKuiK@_ETqW* zXT4|Az;JK!asuW_o%)0v710?VvjL1R7vbCKnKiLc#$R3Ep;4MI}oF{IM;@wUNpsM@0QUOyzNYAB%1ku&{5RWY7aws?ow zoKdH!Yw8+!6~ogZq0N+EH5}eg&_phgWooJlHk5I)gd~<%tFlm(whdkewp44y9u%8i znIO_DlCB2^R_#g@J%oOmP1ez$wHkTbbZ$HAqj*q2J!#@8i4`2CiE+RcaIT6Rr)Z6z zSY&5aBMT}Qwn0+Gb6ZZ732@D`sO+n{%z_1=FAWa)8qhBu93Rv;aUW^tzvF1}w5y3C z@@yfN^mRn-4Dh+S1DtU=>hgQMW+VxxUpn9{SIjQ6XZDr%(QSCx;Nzzymr^8*Y7idd;KuvGpzNIW$VH z)afMGgjiowEZ{s_%!n`Eb4eDxSmJdNW4S_*PyEzGnJL)`7pbRk9)tP_85G7}X_aGc z*i7+egV>O2_bnMsUMYWN5wa%1)_pGJ#(|&0ZodBXu}~FUDVa#NhgrvUVx zZOyNNda*YXrgvvh7}+4@;#7dl5V^Z#uA!j5@%{$ou52`Rx!ecBu>0`M5nQuA5UPfp z2rH>O)%f+Qpk6rz5%zD^apeYZ{)>9noKfa_U~ zwPNMNeU?cM3`@}`U8u^-S@X6kYBoqcIeaBeNl^CFPEe?-JnfdWnlH+uz-w}Xsr>N# z&KG_(7FGYfy1W9dXMrnqVI6b(dWt#e!ze-YT^?tLt$Hx$?2iC-dA@LVCB9PCuig4< zww%@DXdMf=t>%v$llu9*MHB7*aQ+28iVVEi8!6y%;xYYKOvmvfyv$tbl6tz%>#7~@ zdGR_1!dpY?Cb)VFZ#e`1V1-j)+exos+OMSyT0KeTPx|U&=2h2wAf;E{_`hfVGu@rC zC-VE{uyw7rhNH;xqEYhr z^|EFisISEpPmAXF!mwD^Hdg_R>^p(^P)YDo!G2xIFT3@HuXI&0E#@}k#>;yAG#RPc zAaD3ytT|Or<>7e;H|&cWrqP2puNm)NdyNDNkUaQxZ{N0n`W+GhcEbeshPFK#v}2#J zH>wLAgAuglcUs(7<1ck$Kw9$&Ess3??gT#3^Qjyw4o`=-v`&rIGY{&va7BUi{^l#ULD3@sh- zDq@eyli%x)W{MGyC>g(>yMF&w(aHDoh`|S79$%tkX{O_LwyX-Ri_SvGnL3I znde){xZWM zPV#!+tTX&_z_TFo-j%`k?oUhm1>NHxlBcIf{w^R>?~@LZd-7i{2Ltd|Tmvp1cANfR zam}}L3Kb%9U?RZrxBdtrp4QbZ+h;C*0Eru5vzFS~T(-@cSIWu(Vp8XmOdO1H|CJzuzRzwQhv%~J6%Z$f+o@)sM}dY-65 zn%JLP3g_8O2+FeGhryDa=MT1w(CFjtlb%X7;>F5L%h0pWVn?>z_B11w)+_ajmo7JD zgu#=~--as8hB^}i4U}x9U*d@7yc<;M*DdTTricx8(rPbA`7l|{Yt}9IGRn(<>r(ZhNdomFu>a- z6O8NIegd(|mw3QyzNu+^J)Q{E#czEkuF`~2R&H3q z9aIUy2JtIF-(<0RlM3l$TD*du#wd#?I@aQSPR{6uQRivquCa^&UxMm+q#vrFqIsKg z6?mF5=noh}0E3Tw;t#sz^I>_1V-F7idd2sQ3Wv(%~=O;)(gM z&R^|tVEb{iH*mYs%3rqWrV0Dk5!#5MQ`ckKi1WwDa7k0$EGwMoi}ZJAXRI-JUzd$Ewg(nqcNG`$}?4U zqU6(Q?(qt3gW(@K5>rw;mosIYW~18fvrdwQV9jY+d`sa&3R9#5;*DRbnYb_1=6{Se zC)Kz3sB2HM&{Jx;GAh`&j2`5>4dUAQ`8t=mg(B+k(QIMpn5FVmVSE3AqUC3i zEP5zh(49cRjK(cyiM##ND41Rj05uTC&!*-0GjVRJsz-_{w~(2ZQgj&#qNGc@ev3&4 zo~TuF^N{%+*FhKorJk^TW3^l8_f-Gn@zb7FC=4z&SWV!q`OMb+u|hEgq4?ez6>c0o z-T37bMjjVg0^qHwEP;~(KG^9^l($iG3o_em+QVbq2oEp$UNJSAjGJ-{qB%J!U;nHG z<1CW6pzSE8eQ%hJ)6>P)eLT4$L<#OF4+3E4pNzzE0V5k5gthUMS_=kE%8=3=<2TrS zI|A55(V@Ak5ReAtJ7k0EB%DI4?H1hq9LTtNsWH2+YO5JC+YC=7Keez|ma7{o1CWoh z`hiO@tkfp@%&;W7&<<$Rg^2Bb)bLvKHFj4KHNe%h>?U%=Ia*#S7w;6qdZ z1@M&RsoL#d<@g3I5E)%K6n%CuK0j^eqouHTPK0FUKm2!~nfSLHBT?q2frKXe%rxl! zhsTnOBatV*GBU@c=7A_Hi@WnT;q%&{bFk1w?gSj=cr-7Lh+&xU|ACdhpebYcIM%yi zhZ8M~9Bz!@GNtTIYlXv2XVej1^%$7gXqc0hyxXCwb$51lXRd39aZHIztWE$$rl%x> z)rn%Dsk`ilV8A#?|0I~%SbXB*(@4_&L3`ATx((Ja9ec?$kz{k0tW`v4&Q}D#-(WP( zQ_jrgK#Ts0jI9%7Lz@_>j9odIJwY*C);4;UPl>D6R_md-#M4FZf9K6(VFsWpSVHiN zVNhp(cv;D2>$#_6rFYF%uKYk;u%*oaq~Od7l24h;dP$%&7hDE=U`u+>1iuF|#5p;r(~$LUrdy9Uyl#RqC;3vjyP@Ou zXNGlfz0mW~o7oVbK=3b7&CrMJysM9tboCUWwHnvCU6J?~qV=~hTwTt*B@>75p#Bn4 zTuDw8R;W(Nutv6UyhdL!teI^@tXOZS75BVoFhc>2CE0rP+?!p;0NVBjk7`bYNiMzF zCGGWB_K>Me-^b#kjm4ghL{2Gm60$w;(Dx7D-=L@mp2iTm?u&`-!BLW?yw$3%cT_bk zdFG`gHr45H(-f~@xn)>YTwTr9tC!_>y%v|) zqM?y@(fsjw6!t4QERK{eS^YFQ1QbaSK_$kZ#eAXJ&nrOhgd(#CF`rf>R?dftI67n- zH=+G+K-|*gKw{IwqbG`Z6OvYUV7BI0{ek+~oR!}=;BOChTa()6f0P|Z?{lhT{o`PW zCAsa@Z!mZf?n{;tg;D2y{}s6Cfe>;`ao-JF8ER&wp8<`@5Gfs1FQWg3?v|MG@1s{~ zLjKm;7Pffr%zvl)%Vj^+2-h7Ni07>J& z@c`BhuH?TLBPfXh1oH%LCFaQ#%<%eYdZ>k>2n8rY@G}L>mb7BMc=HM9&;9U;M-P|h zw;k@iiZVyRO6OL-5$O2r*Wdm)at>rgKp#^~45#4Q0rkgXjX4O{T5zKg)#y8Am z2QKPugruULXi<(k!~Hzi-Ph1A4#cq42|9l?CfT%hBCO=NBKKtk;2_3?oaGNV zDij&7N_^*8>E4($=&LIlgxjkAvNH^szpARu_lXb>Ya*_OyicWGZ=-}Cftb3A^p;GO zDmF;F$i*3aNsmyrM5MJ^?#VpVE2GvX`3cuXf18+6?6hL4cpm=N z1EHb%wuuomqHnf!k|yefoLDuRUO9em&+j{g$c7(~0CFS|SD9u^EuV)`q60@nzy_j; zjxFl-%F}LhzT0r64x;9N(Vs;ibH?0hS=0-W?Iu;tU%inw zoGX77-Mr|4cPyEZ*O+AD2~mm!u;3;u=HbVh$4J2k2@&q*2vooEG32-Amvc;ut$wa{ zcN=eI(JW!enUxZE8ae3%%*}gyZ5#^q9_I!{2WoC7LkuQgUS3dys#w~N+Nwi&(C6HV z<=e{m4AeJwaT!N_)h(vJlQ^ygQmpb;V)KT0gIi9TLev|u$62(;0kNpSKw68A#pXBzJ%6-Mx~|9XSKo zkZP~7nl7cfxNWm_vHY#*d3VaaUGwxa#jY=qd!+ONWsgxtc(`b~PN4kJXh>G45ZM`p z{)IWZY*#7EHN3wm7_k;{VMQvY3lDzbzvP2?xKXKEcd$iQER1wlz(gsA)Re3han9@d zf?CC3V#aab6v_cCDZ;5%0y)w2-7BQ;@+hL2fUSt;lb53Cnadjn%ZJdX-yLl7Gx#wN zs9gf(th8{&Rq^lirDFM;(WZ|Z@#nE!SB}-X@*q+$ITe4My>`@Syc(+WB5}u1q3BWi zS5nrjLTEwwBfoSnH($9rA;IZfa3q@yU0;+QOEKN_Z%hE$ux;q(OdE$>beRw2O~UIM z0#rKS5Z#VO>6qFp7@y|0v9Rnd`DhvEr9Ic)ffk6ZL`rBtP(5&gLPY|OhX`Gc;Eq~A zqZhP$=B2mUq?#WBUp_RTLw{EvEc{Z4oXSsc8xnlm&Ni9zoeM{1vDoOL6oD*KXi?l_ zrv28N7<=IQjbdArNJ@cCsvSMbd#f9QmlH%ZsDo_;tl52D?(QR)wHn{Te=;6W<*3gM zsIG2&?yvHHD26*Jp|w4D^O)myyM1*`b2=%gb{;wrLDdjwy|` z#>pMK!U^Gk-JePnfj*f?DLY9UsKaqTVj!q3!Uc+~f@_E1ny=h-wk8D30n3m+ zBNVN^7_M!g`*Ykg<>;*?`dqkEU`PIb?+4?A&M;wk4Zmy#_Y*^Uo+5#8ZOe}Y zqmozk-+B9vzy7-YEif$V(EDOrE8wgg$Sqjy${Q^!|9U1d>o}#cx!K&W2h##reHo7z zFL*ZJQizq&((A1vfh4g(jiLJ|J9T#No%#LPoejV5PcdlF#%sA=fK$ivT$oMm!@UDS zn1yUC8(y9uA3pVuuC+Cqyc9QXD!})I${-J0B6hotBVFsqvp5*~-d>GJ1@oxB4uvE6 z<7SaY8L-B{L({=;3*~MQk0-bGYCFvDK8UjfxXMn+SISr({b4+8d|3*cLaDGj9X3Hd zx8b=RVQcp$-zyTM5?XbB#7#ViGMX!^$d?jTC$0Wt+CWg>P5HF&R_HsBZ3ciPMl_-yS#=(UR=2?CBG3*MOYln2F^YFl3^8&t7m0iz|Xi6!fd1 zTVmh*0)x0NJ`W(88mJ>SAAb9$d$WN9HvY3I_^Q*{@c;DF|7}Ki*#NRnuT;=`(aM%I z$OW=ijJ^W$|`wRO@v1G+?uJD8+M4qtP9$B6rd?Uny#1P1OK3EvW3uTG9 z$+$VppbMJH+Yh;VY|mPyPk&n%i%vKbymhTOD{gIVa@TIwihga=DjaJ3me(FJ3S#0h zy^pm$a@N$BL(t6D;!!v;k;bcI|9Vwry*|WZRrWEsy3RS*`WHUWv+lLl4FmsDNP4YPX(+G5CiAh;jF}6n*t@6w z^%+2K$!&VSJr*Zhlg)gn3TWopI24=kg7Dx=^`eAhn-^?vS*-PdPz1UzXWeY<`ux*u zeW8CKD)<2ZW%z;G1?lLP{pQtYo7PhJS+rfoxwr}Q)9?a+taJI#b^*;z1rl(B?Ibtvy* zAQi~;PBC%!@QG^C7{NVa=8xd~6$2wXhK>}~Rp<_lVYo6el~{0?<~uWA=npy7(MAE%4Bz^S@DR_IOR z|AV>1af0T^0DmoL^pJ)f5S%zSVy3bJA<8?7_=%YGU_+ocb!?%iG{$qL_*FF8;8kbJET&iAHIH9sKg za=zwVlEVk~YYKPL8NNpIgQrdWkZ3^T+vjL(j#qpAicyaM-VAd;3FOiZ$K@Bng!4Z5 z1F@)PUan2D*#|U#wooy-LGZn%&9b?LyNA{$;>6D(l{pxanc%iMHTe$BYAVt9!HOZ3 zZ=!Z5jM={=kh|~mzu%>yHncUt7Wig$_e}UL4{NE}x1#|Yn1rv>iepregZ~U^S(6xq z#}?QR@@SYP^MS;y(&ud-3vbx5oG2oBgehX;Hl!Wtn{Bn7Sdb*{%rW58fj$DR!Yrdi zSf(6Tx>_61ycx!JHVC3KT#5LcnVCsz(eZx=Gp@vVc@&*dhdZ^l&s9Rl#G-hH`2@-n ztD8o5NjVsAk>)!*y>>@7l-yjltSBnxK9qsR1yMtOfrZu)6rl6_$Vn(Y!%@MM;KNf> z!$X3QFN<7m!qBe2_pDoE&$R{KO5xZ@>C+yc{EO=hy<&I%P0jXK>G|D&Dbi78 zdoY3TIlkn$ouMiXNEA73A$50sFa+yxhb<jr*j% zYPR$IuSnmsV)Vi<(iMdN^X`+yjd(8ee>($G;^{9I;L{CUBG79-iGU>Iu4y{vnX46ef!5WKb*pj~<{)RZrS3x3M$=4WF zPSosNO!)CIBkMZCb67zXQ^OLy?=3PfSpjouLt@8Lf_?X`i1v+5?)(gLY}BJyyw-q2 ztJtQ;)lfWQ{7VWa&Q`*3doZfmRzLIGlm64Z;bu*yxv<}X#X_kV8PRe-@;ezhFRUB> zOI9Mk2#+*`vbd@>WkL}-5>PCLa~agK^eoZp90EmT?o~so7H?hUEUL5h{f1VM#_AON zuX;GT>%aNr=LT^H{4w!%$3w!Udm(}?WstQfkA|BI1>Hzu!BXhU8KxQf2?U~=o8+4P zy?J1=0ZQef^OEgT?AglLNpA@?Se8SkriHRZP)e(ct*ostF7@i zNx4`r2y>)T9^`m%)Qg43`*j^WsuHrgbUx!be@?PUQP_V%dPQm5C6VX#BQR2oBQfQl zP8DQlq5DD;QG@X4rRRassFb0ec-EB~y^JaTwzzccRZH~7^|o3n4bV?Pqrh05@am#i z94>=^ac`QxQIm|ve0nr_J{0(KADGOZa<;Y@`^)DyJvX9#if+~Et|T>6+*jc;%1wRH zZ?5jWXno$pjVP)xSWka?Xo4N@Xg%(V(fU4LdLv`2L3w?AUW;Gq!`cPkXt3p0#h-}@&N$<<{ z!IJQ+V&@_*f0l>2%XA1KAu&rX%r6opEhf82;M!^Dw@u*HAEYX?fggG>e?r9%y4K0Ed>Gl^6<6Y4NJ z|3dXZyJS(QmJauuJG{K2UwT)L>^^MlUS61)<>RkuQ(@# zV}hxCB6S8_IyJWxUi@~Z3E1j}6oXV^Ie&SJQh2spyVEYPFK_6s5WT5e^42j!bl9q8 zixl~1z*%c+qOL2)nqdO?C>FjFwC^U{udBxXZdesyn|?`u#bkl2vEdmPaMIFe@@tpj zoX!5j3{A-;7_PsGgo5RRbL8cRMOAkc`Nj52B;e9vqx?$X9$vup$9C*ssrvEY=;g5G zUdzAHRI@w!fl0O6xs*KIQ*Ovp|mdjOFsI@@h>N z$afia7b9-7@c8eEmq!R_t%S2wfH;rc_Y+Uybj`G^ z>9sU!ZawYc(g@4|Rc1ozn+^>j{q|<}g~|v)3hZG8V~DPBj7KH+VzeHZe@P(0(e%&I z$ihMtEU9ui;Vok0DhRKU*lX}s+eh`4?=#?&!89BzN82FnkeX=aH(HP8Lbs#na;|?Z zUs^5s$~wJLqZ{|(kJ0UvvOLZqra{E_^D@nY66We6Lvdq?Z?$t60kB0y3{$%EK-y%} zZx&ii_ic=_2(9XN(7bIMKcqAyXt5lFqz;X&rts1wm)BRQ>q|^*|5d}C{w>f!r+J$n zANK_duBGX8f7jLjB4*4T>xCHOJY9Ly*Qr)-p>$+jC80ZS+k?C?^;BB>pnUdlFXnEAKBrfzH{H!>MYL$TcgNAo|=F4C|pi z*uAro_+h`6U?cHeS)AaU41^}ryc=W%10hRpm8})R7rDDU(mB4nbfF}1g03bCold!;QdQMV$K=a(LTKZaft~p`%=Rx3oAM z4Dn5_98jf;X5PGzbz4RFP%RaXWzI8iCBRVVSW?L#scUn_e=N8qCH|Sl;nZq2)N<%%;t*@~Hu0^^!YE|DXT)>9Nt05KNIl zaXIqlKH;m{$u`?IlCAvlWcMhe@Z1u=^Hor;(2Cdp0xnYGfnLy1JTzhQhxo1sJ1h+U zrIvBbL)Gf~6yy{8R9pWJ7fuRD*4ng_?N#OZt1SfOGI-@4{*zs^l+yow^uR_|nCVpZ zQ&EsAUq?2zl`hQ4?f=1jIOAZLzRZm(iBc)?H>(8?dy%Rcc|Lp!CGm0&)dT`J!trZQ zBY4p(qk&=}*bP-7&HATdUfR~(fN>@NBBwTEQj_r&N`XVPw-S8nZdH^Xn1os8^MS*t zfhpju`fYhobvPt?Y9HfB#V00VxClI!wbKD1s>?wOo%)Nz5kDJ`(++=9=i$$IHT8<2d2fo4)SH zdw&axCQ}$fiRK*h$`}$If2Y)ySXE3C=6gMcAX#ms5&yP+U<5exa!AlaNM}^XgOdp% zu0eDeopW^6sd7we+s-7(X}*XRG_oele<8}~i*SJWzHjw!+H$79kif~q{S_8< zWyXe4`(pdhVtEQLNs*j?~|owgz;3e|44SO0rEXC z+G=qUknH<_wB-62?)j$!K7E)DNB^;}5s)+u!Pv3BdL;f!dvrQXd)!~$)iNpGX9Z2UTzcWiF zGP48fgXS`S8G8m4bd!4|R7(dx4GqI5xAtc#=J6RreNd#;SSk=`Xj1Vp;_7v9x?>kUT4Dd z@k6-&HV3V;zpS+C2C?Y-F`-%cw@RdkhHB7FYbSh}fb$PL(#j^$G`mG4i0>D|99k9fp#61<1c5e^jSB2=+bYDLPKKb%e#c zoL%eaRUDV=AmEUd6yKj_@XM-8S~4v~uSYq72?ik4pg67RG>a6C+9}?y`jjOY(b8VvyNdD%YjQ%7+(O2$Whh%jP(##O>4fS4P`QoD*(q`aj4)IC^@|L^pzP^L z)^YEx#g@`gDuQo6P=4^sm}rsbM(?dgWI3tV?+`;owgZ0I2Pb%)^alrb^A`l0d2y!N z!7ocL*X}q`c;$s0$8f1*Yt(gPrn9{-^-DyLe08jppQ(&)gf2?n=UU!slA_9seuFWI z6XjF`k=9Ej%qkQ8=ImBeol+`J*{o_BalT;xrm&x2IhI1&zu@vSlHwzp&bvjd_Pd|) z6ZgC>LP6g3pF3*1%B`eJztQxAyw@v-ai82L^Z}HA2Yk=(+yUKGt>Y|x8OieD%xWOy zt9UZT1Q%4y$PRctJZHqhnfz$Cwt|6o*2O)?Em1xyykK?!9OfZRIzcZI8>;X=0~nTN zsr^WaCfK?jyN1p=&J6OFc&5~QcgJ5+e{J)%?F89Zz=4C(kYzXAh9tJ^MX{|+J zib-eReP8M=WKvD$P_RiOsCCy;`=O!<3XJEz@FKK`r>96<4y<8TO>q`D=l6b^U;G zBkeT8@x4`vGXnolhOb$4#V#mMVrcct*}{EV70ah!y%mdN{9?tnP@f+_Jo$S~;ZJYO z>M4Zsxh-%L&wh_0=Y-?Yo~pbul$gp;i#H;2HVE1k0t>XiD!1Zi(VQ zUAX*mhTwzSa0LcK>7Z#_65j|zDg^*vBnd+W2b<{7Q@MK^2T1|2-$EP$oa8n9_bWG_ z9~{R29bfDH8e+Q@lMIU8XsmGQFApd{l^DjHX z>2-|$BIx9D7%02=mfJnF$A$swnS%RcRPE{9NXi^N3!Sr%BtnpgArNBxvj0A}qr0=y?jLfT;?@G#RH37C6oM?j{c?mr{Wf1w9O0W5u1=DL^xWzbIF@(5k@P}4 zbS^d73e>dlz@$_m-1cW=(Sb2qZ8RT?OqqpquvYVcj->UhIzX7l%kr_ik3s9WNxq ztcG9f6>Q6)Q*~CiVa=P6HO~9 zOuT5Z{7*gc`l+lkDOa%O*m;rVsZ`Tuc#4k$$eEwf&2+=CLgVe%5F>!Bm(jq;nq z1cJm8Awd(OJsHMXp!qIrvQb)O^Z-oOkPiL1^UFwpp_tM48~%?=xAynmE~?W2e<^S7 zsc5XFVLfgi+L*|IipL=~wcFCVLxPIKYWF+k*6t73D}K|fo;{>t907&S&!Ono?g_7# zSPrpg9C|C-Zn0uHJ&GLrvSK=@bhftWe+|C`kfR|Y9FnBcoVzshR~eaZOaG;w4}v03 z2>i2~*V5Gp{`j*!Pdw`I&7R~nuSJd)6i+?7wkC`8^f=q2Iuo#GIL=M{A~mm5e-f%X z5IztiBWCCCq7nRekhX*j72J&H!ZkP>k;p zbewbINNinXFvyc#h4J8MX>I)NXvVNo$7sC@(O2~JD8U-_AY27WhnLFG$4b}#LvuS# zQ#3`=GA6T32_=m}N8RP8>H&U;GwOj+oO?qWX$5{X0-ML*n>nnb{k=8wAMY0rW4gyP z8lXn`B(T!OCdZ5rG5plz@x>Um&~x|=2JsW&82ok;Gx^;Pj94VC$_eTTzJb)I_zS98 z88)NU2YoThCwxe{L-6ZV-^(ng*v*;-jF-|kZxV}9?J(4;pI8RyqCy1%w%_pI{uIst z70$+-&A4%9#^zH!VyZ6sMzdUGosS|dt4`aPe0QYSpt`cyuoSKXc`sZ;hZ;LPM0~v| z`zLNO=~>KvavcF<*)|>SdT#k_Han)SG-zF9-SSjkjF4Bi_gU1r7p>RTdPJk^Rp13e zz!T-UeFCv@IiGB^TqxrnOM`1v(gydOP3Fm3cH_%vUC8?Z+BIMUq&#n zDUH<0ZjiUy)=C6!!k&Hr0p}_dm`O$?D?xW^oFes|)?7wMCSnyx&n{?lA^NEhEbzap zxn!5Yp`QM9;K^0#kJI{s$5$WZPYKZTM z)=@IS0Se(KV8FKv$I6_g3;UjG+I`?_rVVV!;#;Gv{E zRGEspFh@K%woQ>b3P5f4{ZoP)v+7tGYg56!C`BMk{dmFO<8)nu1o>UA`!&e2w(%NP zOIUvd$Jz3KSVa8i*H5GX#1(SIgu~rqyy|l<70NF1DEg285fKG237`h`-t^=o72M+l zJl!AEB`$z|o8n<$6Ar}KCc6$}=Gi&6I_P}& zgq!sJ@zo$N3$~Z1@w*t6x8so)MO;|6bkRvZS*>B%v>X@uUtRb@<-<|JqMMQJ)ig;x zo&fFIc*nAwk=@0F{q?m=PeNk}%Ola0uuu%xzugoN*n)I{GU44SNJ7X+Qy8;dgcO( zVv}(HMwSlcx#y&?AcHou_ID|M38`&;v6lm5Wn7WMoR1ao6e0x7SIOi!j|v$n44= zgM^ZdV$i~r;o$&VA(6=4+-^J#++Td437vYSif#cQxvzY;l3|NaYg&OCx(a@ zxM84%)l?)!47efoej&Ns(K8kIRV z(}TwAK4IYmxvJQh$Px|dE$mMAL!IXY z9Hb|l_2ta^5^Vy%r}<9hK;;V41 z70Pcv<@uu2Y+XEiwgQmLiVN4Fn|S|gm4qA=juAuy$@@u^CK-(Fs4ql0&hAv^1rn(N zlggRkB53SL7XyvEdQDXm!5Rj|n^IdGT>z)ef15tX2XDe3$qnzjh6z=cB6~=YijQ6p zUTrgaJ-+3ome|;4y8;oz4F>Ti*gVi4yke4zxwa!oDU}6gMe%cmBSAuikE`N%i>&?D znqU0DK3vU#T~1U($E(rtPhXKoy|Gh``kpX+bYV5yRGQIq@YRrduBQo~5F`3GG!eqz zu%**e8TVOK4oFxEP2E>-h}0^7V`a-Ge85>2gcQ2%Y;$q|1vb5dsrkqD8K-y)#73PjF4##V8&smlb`x^@ zVM3JF@0o|M>QeDN$>4|C7#^9z^3b1KpX+&T{>Z4oFua8c(P8eO^GdIyLttuJt?Ozd zVcr4B1x6Tq$+eC=RRxsz2Md92p?FlC(ZS!Hq z9--0G$Rxj5(i|7c?4)CVrfHMrVQnKq-0BLD^`+uCw#`d?R@Uq;)ESi*qzqEL?)cVI zB+2nKMnqdrVWvvBO>e*&%qG9R;{9V*V%T zEhwA_f(`^mgU>`&$l1Ylb{4z?bU^&a*NQq!gh;;suS51Usk1>SiOoNDV$$D`R2#Ih zmHnG4Bq#f$KfCV^QhcwHiE*BO29VYncx$kc4?*BT7K~xu`0h@B3}%3chRMq+yyVe& zlr|`o!p_ZIXQ-4Y@D54TLRxn>h4a}H;=Jgly0EX#aHaO9+DFZ=C-@&u1_XJfQZ*!y z3quJ=U!7@uz7JORbN51H^BZ^5c!XS`2j01M6|FO3dG+jtH7opdvwGfY`rgjrKo}rmCm% z2o*oWI$WP&&nse|mVHC;ikw-XtN3t8DD13g3AWzC+#OGgq_&Spa_2iLxNkaPvD*1C zH5L;}Kvs#k%6R>{)Req5BSCp&dQK?JdRC@mhA69h9}}K68qFtc|J<{k9qVk)P9N5p z2U5SNlSPKk!P8FIq{5@f3^>-3#7o1L@_SYUDl8@KkUpiRPV`Tyh3omcS1MZXh0!8jOl-WHsiOpKZ;sa7hvkhAT$`Y4rIxtooePw{Kg%~(P$t8$9yNO~Io1CNVEoBwaHPNIsel{V85MHCyWSk^`MVw|F;RKWZC3mT_6uhR%l;}%LGsVpiyLmAKMkf@9Atx-TDax zFgVwvs5s{oNK9@-pG_A(w7CbSUse2rME(5pTl`M`Wc-|p(@CSs`(rRer%@D=1qQu* zjiNwQsDZju?Jf)SP4W+Hj`eE(6y=E`qeEj=pz|DLvNj`hyd}b38Q>}*?eW*T_kWe! zk$5#wr6{Nk8bZ5>O&oE_W|5QDsg>^}IF9@@dd$fo0zL-M1R8sD7z;SQa`=f5FDRcT z*vxs0{xxvtC7$3eUpMIu6D!-E6cUJV%QBotF}c*w`#xExwwoNbO z^PUI#FW}@D3_F$0OYL%uc-`k{@t$DntmYx$*< zrvV8UL<)G;jRJHuvwc?+SUM{XN_?`EsnGE4NwI{ua$Vh!{>Awg<@kPO`>Ijb_Q)2B z*BxK8unx=vBeP>RiSN@uW6kCY0ap@q(sX*{AlmUCfsP@GsQF@w$G}AiwLfFzeCicE z>R!(i0?|bNx1nRrcqE!9^@$uY?ulR$Z;l`sYUV!*k-kLi55B;G)tT@KZ}G7FZqIh> zU>mw2b~6YYCeXK~LBDmz0G~>yBp~C8j$2}L5zBx}G@YhSq64ZI#w5!RLcKb17~OQN z1Ear|nqPAzTIdc*21^R{pHl=;4UiI!iV%NTm%wNUif+I3iPC~15u~vQD^u?O(3Yap z1JUFWSRQWMP;jp2gSn0fHvzn3Px%EhSafQ6+b+=!h@#fFyrfrpJgQYGe<_F^x|Y)T zxvwO_=AGivd;S>|mn5Sx*KDUo!G5!|`M9BbTF0o28-1!`WGtfu?O%i@R1W{xvH?Ar zKWB6c6zkbPuX6}^ZTGg$2{x7tPr)v}Fpe-I@?dfhbyij~-SE@Ft$p7WYqOy(E{_+V zQ2prsasm9dZ(V03a?DFkihcb1MfZxnx6(rbF$EuSFWVeOSr+x=g-@x()X^Ri$*FkUbSFE(57*a$q2e z9PFr4IoMC;g~IfzVHm7az)*NHgxo0+suwH}U4c0vW*~3E7h2@aT*Zyg*2>|qSkoqT ziz~z58~Utb_5BE#`j!2b&22Zb`}36EPH;+=#pQ(}Uhj&Oki;smgUI{?%z;O>OJygW zDBBk`bBAdhOU|E8>o05G*f`+WIzhs^zxq}WdoJDJnB5VMyrjHKsO}@>3Y+{GTcJch z?LgM_v*b(Hy@r&~Th_|qNmUKmbf}|-*)yxG_mSxuQ{>%~{^_AWCYnXv3$YbRwM(iV<-7e!5WbvvK+K=7a7#S_gG7yA6 zXhX7@rMFX|*{LR0D_qZjwWG(@&fA9mns= z{ua3imj4B_cA7{4+hv}dZA819YA}X_N7GnqZRRPqMb8s+VLO_12qHAluploL_ahbY zt=k1*ru2{D4~Bl9fM@Ax3s0^x?EZ{Vr2x@g&Kc=P7AeiDUMl?%V9oDo^QxDYYAR;> zfj%mvk)?uoOUFvkDWTQBSaeEB9}<0q{)NFncSaN}ko zS{FTx(2MRPtjp&(3(Cxya98ZOHJ_-uo00Lzi7qxHQEn;iLDa(4e$Y!M6qUf0vgp?+ z<*%Er>~0<*P+qO+;=i-Q=$pma02LkL4`nOekW0)IWz~S~)C-R&iBnPDKZl6>o+8Iz zp0k3cln#vWU^E`U#uDAg19n_8pIhA#5iMoDWYaKBjv>e;gJc>*ynkE^@Zz4(P`zNK zVw=67fX9zDpGGR#q&vQ|==%%qhwARi&JLfn+!3P)JXCCZ(Cj5>6KF_ zQ`R3H?Bw<#;o&dG7fy{m7{|U4S|JnlA>+WFxAzs-WG1^)sA70*jfv+B2M!qNaHdaN zXrK!XqnB|&;Yc>4Mj%+kLX198Rxszi{XQ|#IMYo+QY>c8*Qit}5IWC3yO^xgfSVSKcm-K6brK34 zF~%Q!t3X-8@qni(tKYUVQpZ}*T3<*$%2l18375;Of}X!31Ows&g4HV9aGf(lNn^ zWvxeCt?~$xNrgmTnfkJ%Pqe`g>bbl;bkxV_y8E@w>>nDzW_XB`^+!$Y&E4RSu$F{a zsr(jg{)Ccc>~{+sD@LiCM{+3Se>^SQ+%_5aY{WKN^^8wHj20P;GLNZZ9?Sxq66A-t z(M0yfv%?rJi-3lUN{5~R{@Yanf%B;wfS&}*cIGkB$^j;ir-}yV$~3@Si#@F7u6G8@ z>bG)iGNe^Y|3#e? z<6j;HU1SFbQgr#3j}?-w35sVo$|0LO3kt=x3IvS~hU%i%^$e0ZYzAA47p9H7l#1kA zkYs&7?R~Svzu6a_#*0H46;5$Le6M*_#icVx(|p5eSbxp6)y|h^`kWe*Q~=+{v4dM+ z8+4rh5%ut+Bx;_!+1}4EdR-E~<{}2X@H@xkcel9CChrSl|9yjXrKhS(jy@((%a;iL zsaY+q@6t0&di^3ENgod2yXkX=R7xoo`2x)J!@MC~sAPoR$;_X%#9OP0gb7jNf#g{& z7l59GZD&Ogf%_BW39#?#V-cRMeCRqZ@|Hb0eVE*X*u+tW$BYL|pO7IeScIpA&aRVY z0KP*aiup%4KZJ$c~|6C zSTEU6bB#8Gc*{jOk4k*P^s;lnUxUEaV21VY!&5fF$Yu-ktMQ1l*&UqBj{^f^CyjJk zuQEjdKu$&U(_hH3td_)Hmk6(U>It=sld9p>rO|t?S?=#5KG2z%h+{$rGG+b!hlfZ{ zGjAUlMnn{H(QQpg2R{v6NQ7fEoB@JAxVHBBc0BcA*BFmK71r}aF))>j(5z^U0WpT^(0=>UQz zqKM()sb*X;`*iZEj*%SI=dZlK-##pDj$WB7Uk7Z*atU|MiR=if9+zV*DUSXYSJNC7 z+!L18&9>qK9t~shM@_<Z|QmL%^mnd0uU}P+hhBz^TMtTek(G?Lrz>!9AyIkWZ3SdZ{7CW zW_lj^M=PtX&s8lLEFuNqe}9}sK70vCnKTxP>ZRXK37}P0NZ|j~0t?W7yDICe?9t3^ z>=$jQohPD{ctG=o?n~)ZN;Nm=6urGa{x$zxDe&IiOTBjauFnh1C}i9Kz5G=K1#^eS0>Kea*mBNVV4pIm_=VB0K}Ll3h|0>R07pBjblU2wFB-9wB494VUGa0P?@4> zV6dtJZCyrhQ6I@;n2}2GWKetxHUvR+P3`#HjWX=-eIEWa2&S>19=ssg#lAj&6@a68 z=<0Y07lPTOkkgFcmg*rsL!jZ&UxbZ>B*U?a972Ksh^OBWN4%mGnTF7fw(6?FzhLgt z*@ig&@k9C4CY^)>gy>mTQ#7HY4NYZg$e7DoSaGXGJt)p`cSy>2^y%lOTV|ka|2|;G z=r9I+Gmi!^>xF~@D@~Bah}3nP=Tjf-Kct1e2gF=PPsQ+`>^4?&AcDZ>Km}Y?DjwV! zxG9Xg8>cu5JvNXP5QjFjaxF(d4v?AeNS^s>aC3z&b^aL2Ao@=WfB%ov7b`Way_J}q z4kkl;ryTeJ6qxy~=l1G<27v0p?F%iS zD{>(9O=j23h?~Y%*@ae0ejyH*s}Jv@LVM0!)7E`1P@jtg*Ix64lV=`|1j^kG9iGxm zt&bsdX=!DMjow{y(!#FuhtA~&0~6(kowB$QWg5_Lm)GvE0(CanGY8OQ7>{ma$nv&H zIqbJ4nX`RwGiuV;jW#WFUpu_p`zM(FDjxh0=o)jW97G-TaI!x?fd5%cXF;&-CJHPd zOX*mHHWQDa+rZE2qx3NX8reb;kZXl@`IM~zT~HGzbvj97p)mvm&(}}gnGjHIQRE$Y z15+#jj+SYxutLECM+V214GSNMo}pO+Mu5@ZNNeO-sC~O$6o)E{ZoZFlaU*$Ytumx} zxDa2zw;-Xa<@!SLPnTe6_uS@pVFYEKD|%DP9{JzT9)2cRezRKv%1d$9N32sJCK38H zaI7cYcM2zfSsuut-pv{l#)tdv+dN1haNU8RpxczAAitA3+^30~A^J0hYKHOs-j`1F z_pR-8K?JQyWC1)OXq)-~4(>mk5AHxQOrD5SOF1-|Xk>n<>M{apbHgWhL8xS4h%d!B zLSmiL22V`F{x6fDzZj|;42WiGtS%goA}u7H^Z#yN>OG^_20Lik)rW2T?)uCS=m1r& zs;sX>E3g!(oN)0Xr$A}fTz>RFMhS%e(Akh0*uJK;K^i|Cu9q`o7{>PihhzXN8Rgg& zw4h8jB*X>-a}lpx@t*6?+(5PV8y=!BzL_0&Q}OM^<#^(GNrf6cqQeBQUYe{^aeYMm zkg(+$;K_9_*ptHP@n6yqVBj$1R5bjH!7A?TL6gP;F?9>kk!*c~9?`x4A{?>?vKee( z5sDGAI-RNT@gaekkrymX6IT|ewAYMa*Fj$kJ=8{J4_i`r&Fl~g8(bAY7@?KkaRt7X z{)nWO+I2upox!3}^!{xz?R7GDG&KBm^F$B~TQuMaS77hIbfioJF<~b^QzY{U~1OTK9Q7nN(vH_)<>;(d0)12I{ zb!$}%FAt}azyrMYg#Ljbm9DUS6pr$fuyUZ03x(Ji)Op|OY8$KOU*Pp5UKUqP`wy^5&Yd$b%TgZdu6j99FDeoZu= zUO8|@dw6ma1W;Rg_FeJ16D8PQum|32&DO2hEt?w(bvO2n~&L~g-Tk)sARtWFo)lPe8hP$lxco&EsRL- z0pI)0gdKx2btRa^ZH zUR%}9+AmJ>qz67@{!cMepdAyfb*?c5@5A+jX1WMO8rK@Q@rFlRed@Qmo9+|)TP6f3 z8~|?3y+8cUda6aF=rV6hJM;sL?_P^s)cP}!9{IvLy1p5xU-Z~4 z8{t&%7#kRWBrAcd)9bvm$>on5xo2XFc#|$2JNJz{N9V!Q z-cWX*Ad?;B1jEdu@7Ku-wmc+Idmj*!?i2Fh+JHhP$UevMSo}X|Y(L0KTRAxRb@s=9 zcNn)@kuOUCmkX`jZj zmDWo6z?1-mlo~#bQl1rr;FtiInaFYTy6g8zU+?yAWZDJuH`q1LLeXI%La3qBb_Ale zjRCUbq2~kGnWh0T7QZ31+`kZfHqA3qn471x-uN;9^1_!VdBHIYnBxRxbKbgkt(Fvk zHq=dca{=6x7a%#<9O_GID6T23lE!2LZ@bj5NbGmb+FQ;F zhfm(X1rwC!_ke?##i6t#UuZ^y#WQwV7JS9)GS?JpU_!-hB;4*x+26hu191>L%M$!N zC_mU%glVicrlX>Aw?=It_@L~LV=ntwc}#U4_6toQ`peWL$Q6~@EaRUi05o`HJI6D? z!@T8FexlqvC(j1L`8fkyj@);1hGtdxGY`9yQ8?qHG@hzY6cE@!><8TD&i&#gi}Lhd z`74Ygmr&**{^tcLE8o}OUtCo81CR0KgPf*{!u=w^{~uj%85L#t^$Xtv0|V01Ff=GB zUD6_rq{NUzqkxoj4WOVjNJ=W*4HCkPNOz|&NH@|YzVm<2I_sSCtmj$RmzlL@zRa3^ zUwiLg{Z^9XrNj99DW7Mfei)xI-Y|rH5$}ioG~@1xv{A-~Vcj&=;u!z-C?bw^>nqRX z&x1lmI)>==uEH;~Y#tnkuJpb9xksc2;=9wjBgag>8p#dYvqQphgt~Q*My^-mN~dUc z-vSO(FWP|2TQ1kfroF3MK{nAavs4ujQye;kJDyF}30}*qzwgHD?3NJeSG}ri+4QfU zK=RJVmu17*N7E>PO>UAJ?AYkcK5JEjNE5HT*7GJDrLS{nEWH*W2}G&gw|Lhej7!fu|; ze`O9mV+r1bF#*O$$RwYx`nb@$%-0H)TvlDGJ%V_9qaUsW<0&{4eSGdUFCs$T^~Ibg zsBB^XoQPC`cucM6DEJJhjfjH7gu&>b_2=bhFIwD#U_F(Kp8NsikJFEMm5BLVqF0Lo z=dP=y=ayz7Ro4>toh%c#<^INrP^bM=v5@$;Dz$;xq{TIn*;c!JXg&_|%0ji1-KBYf zvSOP&;R&_pticP47T<%^jVwa#+~wfv7g$M!;D z5Bguyt?X-)fRX`HP5BV(nz@lqD#P7}&9To5%+tOf8PPv90W#vfUjoufaEVvtc{wxR zdvDeXT@9AI$`biv5m!GEOKJz7- z$|JRtko#L}mWPL7n*%T}03oYH0|PFOt=TRS7`NS}*eg?=aab z-=9uVu5h#rU9Tn>qjjb0GC$ro1bSk&h!5 z)QSc6sHq<|==Zo!!FjX<(uef%f$yB`^dccSoJjF!eWv`L$$!!#g(Y61X_gYhqm@P! zM5mx4#(;b$u3?)C@}-^IRLVi@Xks1G`adElrqul3N8TJeQ4LzNCnt2dPe>eJdp20)qKjZuVeV%NBHn()~c?Skh z;Lond5B^svnURxB0Z*q=X9i}#cy(9Bh)3aTac7t~bJni4D%-H+#-*(6*YzBY?+fct zvcH+Zc*U{22`tD;|05lf?C(31yQiE%++E}G>`GOFnJL2N5=sKOEhf?ksn8Ur^2U3e zRK{Ql^SjvK5z=+73TRMtZ%|7kQO%%-+R>^Ee&;Dhei&eB;L`RFb$hZJa<=#R3`b{3 z|8?p3mGJ_P9B_4KmZCp2ILwFOhD%~}k5B|^zLI{vV{l)@^H$9{><8hGPu*!O2ss{y zUzm+Hp*+G^xZt1fFwLA2pln&_QLxTKYo-;KkV|^X1xinVo6jSKVenYH`isq9%@byb zinYHdBK^h+)LySxDXkZVrzv4U$HB18Iv=5$`J}RL0o)1$wWm=E%0ld*bPBy_YQY<& z+m%aZ2hB~m6|F@X9Voq2l&;Qt(DC6?xTWydKLT<1Ft>%|YR~}3)fY0PhF~#C?F>z? zL!2B=UOSdVx<(Wof;;W^5J*m4PfzPD$-c1*)D`QUh!9~HIhr}FJQ&OS2_WK0TR_70 z(RA}Yj(P5K9J79i84s!0*+)!-MYs(j##i}EwQ=wOg%nZ8xa`bqJ(E~lhDVf%UUJ^- zNObW0A>kHx^Fu$Hz(@U$P8wO%;GThb>lC+2@)pu|^h2PEFQo)w_F>%ZcptatU?>XIAxbXFfC+p3h%c z3ObZ;3T)!_B#wrVq&yEW9(1H9qA6_v|#S)sC?hxk6Gg19L*W31#DdbW1(I(**xl4cs z@$o#oK5%P;-qz+!7PtK&arW2)Coo)qsT-_-LO5Wq+JVbtGjs`o+Cp>YD-0UV4Ho!pKYbiQ-_zQGmIFIVpkU?^WOivK>qVu z>kkl)*A!E7<7~}RY__8%`TyTH?JyGFl{TJolA|%?%d4R zvG0{&!08M6!d=p8<#jtD=4bFd_+AE-N$%6J+eubOli0TU_W<9b22NUNq{^tRTtL{i z`eB=(?Nj<@r@96qQ!Sm9wX$X|1!jQ*bsOih1poEgk?$O(ciAgRKAqR}c4WAQ@pK0B zju}l3$aob~`1W)&(*L%DyBvwiW>_zgemSJTW0w;HO16u3sNWvQw1D?wOKcku)^rOp z#sf=*`SsyZGeLMwcqhMbo`VQ9$Uy2Mt6-szWF8N3`xD=iT^Z%!)0iY*YXE)dm$J=K zPJ0J92)fI+2A<6?RKG5_2Fzw{-83Kz+&`d8P_i#t2A8nEJ#C_CVwp=bc+08q)BBtyQJBr-T`kzw3`}`p{Sttm+0^m3O>`l z2k#}a0org$l@{|LN~A2XVIZBh38Q*1P ztqv9i%$^B;Joq!O3C8|?M*ATP-#byRRRI}(&N|@pewtW&0pPo?_G{lq{FSesvbM(5 zMbi25$jb$FwwNwy466*0VI^WU*)%PNPPp9f|Moh!XFi&r`0BVdde0hy8AULrf&;I5 z>(1~kIdw+~8PkQ9e+WQysazf@x)l|O4urwE1Ibhbr?-n~J9L_nK`t=eI@+{OPYv$A z3`hF#n}C0>^}YY?XsFeSEi328a~78!h`5pdb0lX1_-XS*Qb;J7FE!WZP@_;PzDlRG zHTMzo0UfC}ED|^4?e!mh%sJ5$RCb$H)MmHbD>ygHMy+;>o5KprR}G$Nb2ZM4iev{; z1}tk27v0Fqg;qSoYtO5O|KIU`1iShm)YQr8gi?U3<06HBQ03}++V!3iW@#RD;BSue zzdoII5K4-wDIudN6-O9+#~72x5lM}8IEd}D&3;TYutf@eN1@>u(mezy1Z-FINvEjHFq1%Qrjn;Eqff}sS&3*;8jt} zfvduY6nk&BT#~v~uP^>R%u?1CrWn1^*B+Td_FQm0H5i2Ee5H%f*b>F*0aUG-Wd@b? z$jJLf^N<2q;c?3L9%Z#mk7lrkN1_v@Vz+9vTs9%|vS#HA>5BE#;V&TaF~dJ=cz|64 ztRLJfXo?al>TBFKz|`Vy36Xhoj}1h%CFHyWmbh`pH2r=%`Yy@)3vaJs7MY|_XWpu9 zZml;q5;o(i3i>8fM==HO&C;;@PE(UUf725@F6wzM$lM$gPvHhT%ZI3qrfHu9BhBG# z-UTTNdT+yXk7+-sc)dFU0;dY7;_{RDy9L#^H!I{~G}PTlhZJ1Efr8rgQ5ogFlnT2E z#T!s&Po2qDk|Q1z zDa(e}QMvN9FC#QjR9ow9;yy~~h?yImckD2bY+K@=0RL;vYfoED+WTXkBS8Mz%R^>z z@N_1;_4btws_~tlhq()5K+5EkZl7N%^oPQ;&5qB1+bV4lhhjsvF!0WQ0VIF>q8S9- zgLn3r`YlnTQY1{Ojjyq&EusA1a-L%Dnvge(FDomzFzl>FAv@x+Iz zQ%ZNsE(i}qY{<-1{-XAoo~r&sXck*W;h>f?+1MpOj7ue<`mvWW-E5%;8(?ObqTe`; z3{}}=q{l*T`}wWFya@%K*3V>h1TPzcZ0Vp`+{swa>BCz5QmAH4CyYCdmc#t!v9983 zJsI-Z$3AO;bT@G7y-B5KuXluERBXq$gswz!RF!QQhjeKh9_Kwx$b%OM0Tp>qv z@)4wSf(xo_7tiZ(s#_f8jYFT{y?dQXSxASy{)K=*EYF&F1c2Nc@3m!0lJ;iVk4Igb zCLE|KOvh!Ujp9NCEi*MvQ3{}(qz$sOTDC@cEhV@h7=xA zPggKl2EDdNLTYfliMb#D9nEN6dk<-JamKkF-f!NQZ9E=Cs}R z{zn6ynteZ&O>$xCIYsDaFeTsnfwnITf$hEw!FkdFKk)RLL9NZa{f3#6>62|bx)rYN64fGfAUwi$j?9o{BBi(5|2~C!WRhWH-l+b6;I{xR4u}Ce4!}Z1YafgD_$52m9|JPzm23 z9Ebk_Q~tYy^g5S+;Qbj51Urd=aOAhmz@z8lM z{+Z;^!7x*w788)BwT&aq*>OuCE*rW?eWiGJVaUZTxx95ADluZa@AQv0bQM+IV5(-_wI?WDhsrEDz6$HoJPml`Ww0=f)TsVs#u3HgBb#}muS#j6Q zA1e9Adudv!0|2+fftul%?Z?@FQ}PxgizG~Glz#AXP`QZSP2vh3f6ykkmu^SUyd7nl_6BMTloGf z01FqCC*QQUy%ooe{tq ztVuONGxrE$>j%V~m+%4Vh9K2kxD4U^qZv!^4Q2YEC;b(aoOCzV2U}At%^q` z<$oFU(9EM-%-RrLd)CGnhrNX}P`*exXev`|;f-<&Hc5>G^!TeYFL>N)eDR-$x^3AJ zpX9)ES5p)8H(Y7NytO9o;P;E+R31>p5lXw4eiox~CPg!0-E`ovD!|$NCtLLK;0xFl z(cbllU_H{-{G!NJ(tGk6*T!kd{#<*5;io?*(MV4JQrIF)?z@~Qcvo+Pt;(g0R07k2 z4H=K<6y~_n5MceOk5S>G4w51AbUs{=K3L+o?8#Zb3da9DWBS4k?y>#ohBtm?KZJ64 zH&JR+8vWm%;~9TukVYh7$;6HP8-ICkd)h*>YyEU<;)U5=a&gDZFMD_6V%yMA8AlI0 zGAJRkpu)#;n{}cxQZL+W@g$O|&+0t;#Fr3V?ExUMg~ADcXbWThEt)9#6Wu5K83=<0 zH`Cvg*d57~9jn>g@A3#MDy2_6}- zhLTr{C(y_fy$BT?k`a4%hMue7O3^0~dbicc``lX$vyH0(G)LXeM`DE20`d4s1;mSl ze(Gbt6)Qn2VX<-x;KH}mcFM5yJ1Lm4i)cad6mLs71r>VSfe4tWgJYp@2Vvr=4t*i# z8exmJ9}DuAJL-^SFf!>Y-Em=oHI{$_^+XsL6AMgz|9*--6H`@}a@D1+C}6tBEMnE< zFy_e#(Q7qbU)9fzihf%o8F)|i zOA~=S;t!K@&^08Js?3v;=!zJLLW`&0iaJ+5eb092TPb~9{F$nyfpa(CZ2MIa5{h$? z$8?t2^I&Xibu$qvcm2js64Y1dB^<$4<8Uc2p{f*%8rkMV-hFTLAqq>AfUT@DkbDr`Hbt6TZROM_3@ThuDOUA+Du#@GL8 zx2~RH3geaw|AmgQycSnT@|Nyr?+>|!3Fg8IcSDf^OGFODLY=TML~9eFyp}qGA~o#e z_%Dg%x3|V+z~2T2!y{Fn0}YR@Keu%!C?nJdu93Vx)V9rh|2;Hx)hzC%Wfvdn)5xEW zGv$Q2DM8AguGB|W{R@8~-s7Woo#Pfsg{D3R$1+{Xcp(Qy1viWT=35p#ZvWW${Y&dK zNM-Orsp5e2G0u?)qTnS39fIKDU*!Xj;2&ZasCmbJoeL|`_APHCGwnw zi%#rR5YtG35$R%e!_NS9j)n;c|CXDOCChom|0L^&-)! zmiyD`4|}yTddL7Rz0bU`h`!K~p9Wo2Jwo#W8|UK`=!2z-@SsH5Cr0~+2vr*PX_Rl<+9dlL8& ze9|F1*S`=rMON&ah!t;Xk)Iy?9LH+OEk;qAGxJNgtwVEHdkpt%*QwuHK;6sr7wdvy zVk+Rf8iT!5$QEt+k0pKj5iX72qLQ+y`Aor`?MdjoVW=ihfr06mGzZ=K1LGgoV{?kT7pk*y0P2tXg+^A6OE7;d~x9EpUXI0v6Ufv z*i?NK3fuFW5xmu!+``Y|={i*i37Old9=!i~XY+~A&qda?_||LZpoe`+gShYuk;0oCezgNk}C; zRMtJbi3`IR{nSvJ71bpT*G3$}P`9hb z@b_xY%Nu_CuyEVCw(nKRVij14afpcTn&25O=d~`cXuAnno0x%<2ZNTXL*-gYtkf9ftzLdzIXXBe4M7@qrt^H-84~`ufz*MC{h7 zcpC{@Ci|8p;}da`m?5wSFmovc6Yx!)I?Z-vfSUG;rogs(nr7xhY5+dDIQ_=_4 zToVqPHteb^Va>oPOD+bIeiC@SY9I=zmGJ7R0t8Qjk!(#<(TZ5UvzQ@Ftwt5d*IR z$zUfM}v42MdNXn_&v_g9*yEHSm2EnJr`r@O`>c3|EWj;Pb6m6dttQWr$53<{$ zKSQ!c&z$%Ss%i!ww=EmqL)PtX@^Jc5J1wY%#4rAe$8{5p zWIzUH3iAa9Dr!V`$N62i_ukwXUNh*8eD*55LmbTj&i^KP(mbQ(#S!nELRVxwhv|e3aHG>7wk@2Pl@^;-6LhTdrCqJxb=A*e zK@2)qfW29J;*$)DrpieEz$6prjX}3W6U*0m{P|6z)x^SGvYlezr2Z=w`G1hftWoR~ z+;uD>;m6Nw#2VP|_tgWRf4|C)Z9#E@d_VDF3#5mEN{1nf-4NlA%5DNTZ&gfsA6YJe zY`MRa%&d>(f1aWPRML+6-q_kQ@boztsvzHa6m${QOja;I!A@QKYd+;_s}ZTo*`>^j zqV))C4Kn&}M?BEQFOzWW`u9c4KU5@jt9ercZ@$#!fRleD2szy^fG^M@y#?1T!#Q?8 zAMZN(Ts{YCd9bfT?gDup20U4Z^2zTu+;T_I0jCr~3CZqLR3OsW*%>>$v3wED?1o85 z_+XzBZeKCADQ_9!-lImK`(!~GsB^?T=k5 z(!+6~EMm`D4fqb#$~)P<%i-qhTL`%nn0(s7IU)8ST1+aH4efXYJVnTj63j3!>QpoK zXHWCtuug;W_!>F+K67m{kLputX~@*A&?1EgUZ>)Y@pqkW@X{+g&k7&JyfVp9#U8)kdI6<;iY+zY`&p^501to?6_(ump55njMs-bGWG5LsY(Eu z(U}dWDZc*m#pLPGs;WMM@bD8B-kPI82s11uk+R|PScdFLm6|PR9!Zj-$#0-Rxu`Rp z`pwsFD?XX|oym0`*|JCk>XWtO7rSp_!{NJ!Wi%{oKyp0AJL`oMnW2m+|_f zTDJX@zI8k8x~}-a-z;}cFI5Llm`{lhP5dB=dmJPje60>L0&n?f0!MaILBtWh ze)5COby*wUOdK{O{K7TTi-T3L4mnq^0mYv=QOkz3jfe{T+tW*lSA>4 z()K6MEV>;rT`cyw2yVB#CN^=*d-0jHjghX~>ua8y%5*g<3A~(#U-y5V3PZ*ZDbVoL z#9v*<)^BtRSu5KvNdv{u0{CHI2*=WFH{gbF!9moB2oNU%)px!#THqPDSND3@)P}z5 zC*nPQ03qwwzewnj`Vq%W32-S>BdPTeXctj^=WYm}f?XOaSPWrHJGkLs{gJmKta?0c zVvuTh>Y!I>k@nsTn>Z;^Sh=H+Y^D_JIiM;?s%>wnCr~UcG(HxAUDI0TZj}0)f!2mG zLaFK*&PDbo$&z#q;GrmX45n0Zu?oJaq@mHHr;TReZ+8whim^&8hc6$b!@&^z_Jj7H zfrxA0j64rHhr4k*+HDnG>tO;_=>R_Lo6>0eyAy$hkh8Nv7vfH3TDDY zAKVBZ4)sn@NI2r0c$puIhFaOVjr+|{+I^o5j0~m?D=m+W`L!*LX8vp4|A_o!c7oUQ z_fUSU7f@Ou$X13m<-_hErM0Ys`*mZBd+pqT(|`9ie)n?0tSTYa7M%XwuUDc5B}QE) z_hx17k52EvgNC8VO{uJKHQd`q;VSoz)n{G8l7Nk5&B^;ZOY(}2_VvPlPZp8(=xeG4 zerO)iP<2p7DwO?-{1Z>@4kT`n09B?#n(0jD2h_}t+IE!;jvCr)?r0OU!-WlRw2^pu zZHR%;LhV4GvBlK`DVuc=l-dvQlg!07Kx}j6;B%;gsV4l83b?+#{zL$c?lse?UD%RS zV8Ocw%ZK?K~{}#@B=A(?(9G_PgfEp-UY5k=WZRqJ5Y<|3FON-0_PC;n9T4gF)`9{ z2^21&l>3((nNO+uleI)0h#qQ5U{0-~A7$;dGG4m5I!kSNQtpIntB;0@bESs^4HdxC zP+VNoc}uC4Jjln^5&(h^?+8p%jG~n@ zc{ez7N;tiMI`_w zT+w?BRA1N+m)^Z=MRP3*=H&l6V!&G8FYjcz3fv2k#xcwF_-h(8ARRUM#$Mm)i?r^( zrZl$12>D4CO|=sqvDT;ziN{nEl)Fv2=`MvXM?d3S{NdkCQ%T6QhXiNGH~xHCBpA;P9OudOM zfM-q4iMg|;sE}w0F}+~W4THR4O43)6gUdTX@KUKU4+F z$_bj!J5WQM-~rOpVpBPKLdk>REy*;6*Y04nQl_?>+dKFzt%8zy4ys#I_&?;#x{rbg z%#zl1aG^^os`VY|R@ZauaOaeqbzu zo@qf_%Se(hV^W8pB*)MXdE@=eEhBO&sV2AyyS)1MQ<%qax;F?hMM(U9yt?%L+D#um z9V>}tnIQE~#a=cryC2b@jk7Ws;xj}~1YXqx-bG6FP8`P-EYNaZ6vYb;*(rgnbMy_9 zHQDKpUryu!<_!fUjEVMtRT-BSz^b@z`{Xq#ZllxPMAOQAeCqQU!%%)U^j8W*LQ+zL zb-C=kS0BnRwcA#I`MYZr>hTwAVk{kCY-SP@RUF0vB^P3H8G8;&fxjf{-G)A@yB2|b zOD`7$JddqT%g)*za+hi z$bO^Uq9l2PXA#y=A1sjOKv@EHWy;Tiwhv~6`wM-o-acJ6l>I=M(|b8|$r2>%8(|hI zcG6!qUQ9@y*zr}G!}Ecm^d1WD-yAwGQpk4qYv}+cyn%(sI}lP~z=)AZ_7 z9Ab|Ii|H|{u+N2G$^@CrEZ>GR3GD?myzQl3S({ur4fOj-)3aBmAm&+rdQh9I?_!;K zaNQqO<^OBJqax7$Hi6xqQD5jZ#XVN1?KVkM_#cDkaT0X5%>92@$GI^|R6~039>t;j zbs~^I59_z;{+{i89OS}|JX0;i0j9$q08#YLvJ2X1MS749Q)otZWoJTctQ>gkQ!cHF z&8zRXiM+D+<*kk_L?)IgZs%zlh-sCq@00UM;evDFcxEIhTUD~W6{$ehfVFr7`BZ8L zW|5aHy0az?V+nOLnJx7v+lkV5J^ZP$mRN&>Aqz>lGHsD?sxY-)!!hy-@zEp|UjEkg zY>EohJH4FuTz+Qkty0ar)M1CTKxxe9FG>~yxJQ$$JraoaE#2WMCWSdc^~;o{M{DQ! z$WY{uA?up1&x3odl3NRL?$t5ZKX{{=@y$AKJ$s|CNN+B$onb8J8{w#abArJ;px$@zgr1=AKubX8`o5haQ9Fk6{>_?hWO_dfJqgL z{*8gu0-bgJYdsaU5$Mmqy#U#W%>%P~+Yt&+QZzG%lBiS7!*g4&RSx4G8CqGTUyeDE zr+h#JS(XU$ySMn5EMcytN&+>&0yhBqfRY#Wf4(a5rifWgNpx8QyMvcZDTV2t0^qq8 zRhTZq!~DQBnb|FQ|FO`5*l9K)h0qcn>BV?+nszH7eay5#J=^9{exK1dMJW-yOp8_$--hUoLne zu{bwSI~!k+gc?aj!cwa;DG3o4`K)kSLxi|&vMT$dl!>Z``WjH9sgKl@S#AfNw1SF_1t)=$&Q0i@C8Q%Mx($3j0p28Juy+4j9agqQ3sjVk=gb86if z#K)!ix%9)hdk?olrg5I8E$?*@$>i|NQ@vRaxWv}E#863s5@(g$OnavL&QF#B>a+-T z2O*t7EX%y?opZvKAuwU1#+j!!u8QbHT{&ErA%X|~atYH}S&YMpY-Ao0aS}>TGLmrlV^r06C|>BrHhJ_vXZ!rqZaTe zZ~#7S0spt5rIlF5%!;3*tS%F_l~XBQbV~bcIa>OqJrt@RGk}HEkT_OEyxh;fJANy8 z;B6eBOpIzr?yuM^@id|orn<}mFJ-SBFL2TV8$!)|_ehY2gKZUI>x@W?<6ZRx)!O}$ z#PAP(cMV5Dc}HJdll+t1d(5Jm3gIMi)}BKu7M*>go(-k73Oa)GqN-DPDEMZU{xvSo zV$wbfGvQ@`L=na#G`s>`9#p*(H2s(!`g}}Qegs8qj2z`t8<*l46=+6T2gcEH21Pwl z>X%PlF!ftT#X^}`0|bn#I5(M5DCZJ}C_n&-qDyaf$S4k^=qKCNi;$!OwZAEb0F%H1 z%JMtw2d#uKWGT_-8X~(61r5FAU}E|pzjZl<4h*psz3>sJ+eE`GA-4sFHB+MdmD ze#92?JnbMOg|3nHdSl}Y^M6HCfwteD!;NaU-!3Z%Re#ggGLOhb67pFnmx4$`*z;3C zL4g&+7sv0HLPp!y6R}4(Y>=h2(;5zZ0V0SX#sQEovj3H!148> zOl*kj4PNWTf4`>-L`f$wqqE;VI!^G8M*;)kN90xTnC=Ju2AMBPEt!25o}16ED+XC9 zm<%d^PJEbFRoDd<}~!JQ3uwJB+qN3vrs<>PyCGS;|N;|D~ts5b8jyI zK1BvbF}FASaQ^87Sb|hPUs#i?&de^K9k^NP`tLOlzG;|HR1*ul?jK#w1t3Omst3DB z7}GYYoFNs3?$do~dSzoOPZ)3>kPb_9;n~|M0+QBYcSE02DLhe54L>afHRa<8TWtiC zb%Vi@NMd=`!mhqs`Hoo|x8*kDE614ILO2*>I7R&72fb*w#S{%ss`BNWzpU1n|@-9-v|*)Wf>7IJI9l=SG3F=0;(CxbTD`Pu#(#V z{F84CgI`IK2>>W;R5hYFJ=``bJ1rY2nFK?r*Cl?~5>&Cgl<0KGpGj;RGzHKChi!g~ zYC=Q;kkrM+2+0o@WYQ;Tibw}I@k9CgzSk@sx=H3fj2f$-g1fN#5YLSM41EMi5Cg=o z#EPx;DfjKnt1%+QE!mrUs^BC-btWTGON6ymqQ!+nk??K_p@NV92T^<&skI zYtZ&*|5qqW+u}()iz1x8xLm~;JM)WaGiDuSeMOwvB+>M|c6(QHW5VMJg?22elPHjG z*IV|3<&G=;QTuLmT}SoFLUC|@Y!(TP-H?NmK!r66`-e=^QI^;&ae*|dX#y4plEYWW z)F&b26JOz9E?qDvmZHo9=HB42k%w5Zl2#A&W$QZP4c-h;jPhg|NvMFHq;BklWJ==6 z>D;>(p^BTNzgTN^a3INNcS-qwJF?wAdHPu5!_Ro~;4k&XC~6$`(w1k^QTc+4Q|)W) z!9-@#8h*PoutO($%DLs3m*I!lrhAEHSUcYGO z(y9Bp>H3znXD9h=+fID&f`^e_(vuWJ!S_knV{ajRrF#~scQ5jor)&uw{J$8@|J_*A zE=dxD1zutKUNX#d{*%f3yw^hmJ;dV?o~STo?5Jk5vk!&Y;mo=fkH2BtV#PX%2}VkT zb8nHt>;8KwVG_n(9*usugH`dLH#0u3_GI z{Y^zj+>b*tJ}9RCfr!?32Dx%Rzm_k)hJYOQAUCS@izpINkX0szC9Bp<5;4_*h1Uki zH6(&msixn@zNxqVT=3%+a^r{B_un{^``w(|Y8(sPr(J^SzEw!ZAj_JeItu2Tjak9K z&zi&d13%)*?C+k{wEw1fm{Stz_X`i{WJ4;7!T|c(uVF@?zyFrRt>I#0)&%fERlAp- zLO6dQJxmaYb_hzLC#kW(k%HK+@O!9qVBh7XW;0eZ5T@$phL1**Ue04XD&VVQ3EL9e z_es(~(%F)!1~15M&6r86*t84a&>qQE3f$|+!DRq zVtU-$`leBLC`Lr14I4Mb{V7u}1(nL) zi&If5FP*;tu5z;{bvkXe(;LWjHx z9C@}dKQ|rjjM3e%WR6Ix8uC~M)*h4<+Plo_`RmAuKhrQwK2SSwi2sEo*sl*hC+zw$ z*m@Hh68fMDHr}7uY9H%4{69}M=@;@3a82nlnCUWTtgNvPTcu&Rc6n6jckxh8P*|_{ zd?2n~|N2C&F~aD)tEvIme22{PIy`LLqka zJ}pw=XO5zV%#$}OiMs>0tyiY?=4KX2Gi#A``oju1>Q9^M`C$FQh8OOD+uApKAot`a z<$gk9w{%D$rTjQvjI0!3n3KC~l}D`?)uHThEQo^>n)lS@l^2XLc*i{>JOQaQ(<`2HEb(q8#A)U8`V{h)xCrOWuwB`B2=W#o(} zlGW{Ig-8p1(bGA}3nCbkb`tu>E+K(yoGwP_*wgL<0vxRA=0$uV_7)H3zb)VA@UpDH z>(ZKoW76pGRQ&;w6c)HN17iAbLo*mVoj+L_ritZbyCL(>&qDjVKYOzg3zt73>7j7_ zg>Z#e2-6U1rmEEt6Rx!Uh-oOAP>s{(99JQjXr(2+ue_w{^S{BaEMlXfK){0Aikeya zl=yc>(WE1pg!*_K1(GH9(xF2^J^zA>s|*$n#Qw0?W_*`GAv(~TG+ie2&D9v$PO<^Z zS2wy*8VediJV9CDTv;>88wEM%Ce}}40_s5~>6^pC!mfrsfAA8N_DS$ey z6-ho#%p}V+Q%L;FEzh00>i-!|jQMWGWF6xna`b86XuKp(tD7{*l1@GeXRgNeo6kvx z<}pQxJmII1&a{cIS(Cm^nyPxgY2-YYoZDZu8g8}~SWY8essz~|P4&E!elj^%u>0&~h7QWa!wIC3dgY{Y-Px#)~~(|T!s`(G7l|C?}t z@zroIj6+oR&r+~Dm^k8?z*cKk(i9-PS-%S{a9=-8*QD;)J8zNi=ZN^AMX|{ix99Ym zJ)!dWf?xFvUA;3Y+#KB}WgOrGyWqVZ*v$)Xe;>f@>+!9bD-!?fq0&ZjP!H zpizU@22gyUq?^H5$!SpMA)1g<%CD=5e*Q*DP+%wPhza8;Zv4P)N}d(b6x`I&tQe)P zI--1%*S^)4bQzP}zEw*|S~LB}C@i>%aI0dGQvkaHx|ym$<{Xb?oD##;5GdDDKe%y! z5%%7|{CTZxkePNuoYy?Y*?Ai*EmDJw0}&spMg@is2y8L}RC4>=uavu~q!f#^K~4$3 zUarNf_tCNUHMR*RXQxTdO_q4Cd;cvY%=yAMML!)87{G{#dW7u44aWUWIJ+?cf*QtH$C!`=h){Ii%Fk75!#iW(crju5xH>Z^_f$CgspRqsmQNMQ)z!>uf1 zrZ`FJ*JET0>TFfdCa5@6GJ``8e@k?F>NRyPC!G6Kq;aROk?ps;w|=QI@6yd0xmB z@1?L*v8Ad`#}EEZiWp}4G8aOGbl0gay`2(vR#8BDSF)F;bn`b{XKx-P%CGLLEwMk9 zr~dg0F6-F~*SS<1=XzW&E)$q|UE9_!agTJ#u0k(m7JZZJaYoDwXAi$w$rP$dU+Z7P zTu*NWX7?%bD&@_1E_e0`Isuj997$;O$3}qzthkH=4e5nP>8=vFRxd2YGvSqAhoju% zs4dlP+Rhg#Hc*4VGVCdMHd%8ftrpwA$^R=(W8sPNZ>Zp$o%`9^w)EzE;zC8DKij3J z`2T##`0|l}&&AT&BQnTYv1k`SlvB%6(#azoDUk_v5x4|?mA4_w)b&isnpt*#m6;FX z4cAe(k9wyoJPeZQ`8#=TL-9>i%n=ljkL>uyT3s&$vU5_>ir_qY!BsfkU5kzf0he}v z71`=dy0Ts?&ghyF9@lv9#ohZBF480=6A?h!!xTRDQ{=tl8x^l6BOLPY0;BU@h2WA7 z+Ak8L-t37HAsO=zeSN`VQ$P{|$COCC57|-AfxYz$V+@9O@w(j+WP6?pE^jW5Hs9Ie z1w--R%JdsDPC!sbhz7RORw5OLaW|4pA&(;O15pg%M@AA<&sQd1So>x(YLwv1t5RD| z{A3E}heEdxpVj`YeVO4Z;2YX*Wp93)cNj5l5zJ5}$-_C6SI_rJeTD&+G%B}KviUxk zRin}cMtEs=w)DomL%VSb?TZjHoq*_k(^u8ddq#oxn8wrIg7iwA*Ko6c;Zgi^AfD=Q zNgNXDgaem&B5(Lo5F+}_O1(se3Gn`~MI1YXiY@LoCH0m3ibZ;AaakA>&@!qF?wd(n zzcA{2RNC{S#v4mukbX6XPHvbHV}e9Yi@+~@Rwr+Y~XdxOyYdTVwv7)Y_o-4`klKuLnHdj6?1^)0sBZ0*WPa6D1vM zFNH)rjrn#}t{>c(tt=%o1wQC^2{aO6rZ;ijl-G;8iBI5<-q)eSzD~JKknK+G@e^6v z=kd;c{V;50S3+#pi3)O^ZAwSm|Dqn3lSkc~{hvNi_UrM!aDs)9;;ITqro<5r_-(&E zahZb^%9i&i1lwnOZM+C)?CZaAF3o>1^Or+#dW`BLtM9L}C65@TaTK(Vwx{d5y~}al z=4t-^ogxkUslg}euCto?FQ?T(0i$DU7tzFJ#S*?VKHt9Ir&K&K_AOk!`6xJDIHn5# z@_^cNg_lpX1LLIjCtor?CmB6>1sgZQgv~H9-lM9ZcWqhIvgO5Rq8ar~*WH6W9oeH4 zN)Ox9|Frf9S}&WpZ1tO03|jQM+^=`W@p-y<{Eutr^wrRtlg}cCL z2E_;Ci1NRrr~^ZTvZkNYS^loSS2xZLU^gc=^NOUo%CB_q9i6Hx2M&A~cjAp&D@9+| z)mS4jlech9-$aFFov)S)5cbMe14pb4A`inA%WWb=%s7q7Jky1HM#FPStEnJ{;FJl`IDeq-Ie1cVqu7W z@*luX^Y2lOH8z_pO`_G{QuypF9TBX}1)(DvoR?ZS#N?O}j|>j;Y9;vCi-*t;fTv@^ zZ)> zv=>!DgCph6haj_pnP!Q7%5@1=apCt~>Cb0_pW{fT^b*9sHaOesv)^Gye8KBEXMtwI zd8MZ@j~M$3%E9ff=zTAQ#sm`hTwknTGyA7G)Ax?yAocL^BMe5x_%TgS5sA8RCsR_E;nR0dnPPI zzDj*NAYE_lX6V9JtN{@HJ6G*pu!=!=#^s#$BpT|gr_tkzuXUj76meiU%e+Sto`s@9 zSM--Q^A~S&t1_+-8vRa?{4C?w)~Z{j4pXywwH~wh@E51pb_XBrPaR=^ux{HIpFtl! z^hJlPlG@Zk!UCM)$seK~}QuCPkYYzk>>wGW$ks%5XDPFXA@JviB+@;$bscd$|aVq9lmCd-nUfl(G8vBr);R%aWFMY~$~T7X%j3s^-zr)2nD`m_-Uv;X!NHRazP( zgx1kH#m9tku}F!)h)0gVrrfUy;WQMZd+Y^@Yh$)V*5j}@4K}glZs9)oE{-zn^SBq9`#XEB+R3n4MV}HGsq^;Wmrc12N%X2agUt5#eO&a% z)d_%9=SE&K^g^v<aeufdIH-)0&xvC}{L=MlA;Y*n;)Z+}}g_spgmDlCsb`Sbcp9TAUt&Iq|=^NzABgmPegK zzhTwk?Goj)h@XwZ3tDSqMBUQw;=6r=MlMtN!Zt+}+NwC;TEp(mZIyOO_^Y0%yEyk* zbp@>W!HcbV(0vTflYB)+$(OYnTda>$`^d%)aIeqI z7JR_$&Acb{y#C8QCKP}`0yOhVjkuO3c@e*G4Z1Bv^(lC#vi6=_5P>O%a#iRH%8vc9 zkEE!~jsC4=ba#EnxJB*s0(>t&_@*@_sSa3;*BUR=EW2lEkv(u~cn;{{8=Fc*(+C-v z0e)K;w&}fRcZh4vF87Oed8amBYqL!agj*z;M-73 z7rG`e01d|kWm51YhjZKHkR|IQqn8 zJ*XQGDpXUsFBB0hYst%KWfroaatt||JNY#R3e*evaimp zpTR>$jW5o*qMdS!Vmg5nz=>qzfVl-g@?HzjYv52!TCJfg2Z$t-Q+_rLduGP#-1i9V zGBf_ToY$E6*khUz`-4kFgmFO+Z9;GL(Ryj%`ly;F4$tq0lDF_%36 z?7XL$#fvmYn@X1*R=S?j!bZRUxvu}%-gX7zEz$ljX!3u7lmGiCKvgID?<6w$`ZI6iFg_italn3$`vV{cyFt2%$g5t_`%M+myj8=zU&_+;l2G1kKk~HH$qaB~BW@;w)*@L3Lw$N* zq_fIQjwXd|l}kY1?94&+S27&_;2oHeIRFcOGO>t5b?HEn2@ZR!6G1`pCLtkKLz}{3 zzB$VK^C;8zh@iD{!fTT&GKRn{iHMO~M%Ru}?pta7#WR1TvJ@c$$_s|(a6MoCBan~l zw@#4c2Y?NWLf&7@yDD{Y{+L_aK%X5D-ht{vu5&$njbH?Grj%$Z+(Z?r$tN6Ni1>)` zGge!gRwor*;c#0UoI&W)u)m&_gHD1ySt~*avHV@>~;D z?MDP&XpJa#`(MYR8jPQ)=f|1m`6?vqekXHg;H@pd^!^=^=kWKoWPL9s=}v-+RD|D> zKzX=9I>vL}#9+c1SE)NhL@$0|-D!l=;7eI%#Xii3%VLa-um2MFgR_FWz~i0NNB&?R z<;flCSGJ!_Eeo>DHb0ko;=vdx3@rWdbDd$~&YN64LzY1Es05rKQ^wj=0^KvVX(U}5 z#B_Lt*df+B=qme?^i|5gGazr?Z+2jPsz=QF!MVxTk?yH{(Jp{t%%HnP47Sg`T8j|qK(TYGm@ zoD~TAdAdF41~b!`Z$`I`9Le>`n{7~X%B%H?^;eWOLJeCqEYc8c=_YFh7j-OwTQESi z+57>dWWJN%xfOd#|7tc=yR%>pm)|I37`8BFFME5+33EvQI(mMa+_21H)SlT55gUWi zWM7d{t;;2_B*?MoASc0tV$Io{8b6Adchw(FFCKQajI$2?4~EMR%jJhO!s5zP%xc>C zaGb|D4My;X!2?t=b8IlU&NtC1Ms zUXwCwgWVqva@A&{MtA*<1I5_UW-KW^oTBmBEe>M>j-ILp!CLigdMT%o)HJFHiUzx8 z906)Z7=y}-eNL4mRebxNy3960*{=YK8=>|zw_-sBp&2Qj&HW78krbsHI6G)$N?X_; zj|h?g+nU`+n^ZD|b6`$ILcv&>ZgT|Kqj%3n#f`;aC(M1+0#upFBv6o~UyWH-VYC{h zv_56ZIVfj)+w&HlT53X{KPmj>-z4cuy{IXNECWm1{N~sA{IwQgQW~gwl^A$MRe26O zqTH!@e!nTeGmlQ~BK}2OZieh{ClD-Kzl|QGG1dbc8c5CO(asy66$PE`^_`qX>Xi

    8PM6K8L^?0(*>`;YHEN?5*CRyD@j#($;-LvPwD_1 z^I9cBP}euY#o?fx0phyz)V^vi+>zW12iwTg+9T4{6A0dCPN(dL?ndh_<6Qa7OB7c+ z_q%J9hl#M|~q}l4?#8$cD3c7zvR!!FB|DjU4307c{ z3icj;O+bRudD*}SLN&C43cFDrPjy6M-afBNtI^P?-;5fT-BfJl?ibs*y5CyDSgQK5 zH@oOEvIqFYa;uzd0zbn(tmu}?U86vu^PH_G41Dr)?G279G}mtP^V?lk_XBZ$1i7Ml zFkC1xBCC+T1qfT(BJmdx10_)nbtBv3(e~9ESOTaIYizG%2Ztex$F6azQ;B6zd{(U0 z8L$M)b5PnL?cUnZO?ul`3JmfHTe4#jd4q)8Ml^7I-;0`I^i=)TIHUX#7))wvw*XKu zcjOd@MbvVH??`OEc>B)tgWC+(Xk_Q4qvBA5hXD(@`LQuZWf+5$ z_SK^D@*^w$u~m_}4L?ukq1EobVfY`hUro&5@@cJP#^Oe4E}88t`ANicC!~Cliy*@= zYlSDLfaMa)UTmUq z<|EcT&^s_&`A_htO9shgBM73q|G*7HPVn3#ibRJAZvU)8dz)-7X_A+~LLZm1u*&C? zuNhf}Nch(*932aWs=MZ#)~=dZjooWS2mFn6hRCrs1`rwQ8YDb-Jt+CW-TIrih63X^ z>0@mY+(0=1{#4Q%Mvbg(&Tf>l0h%} zihZbG4V8=^-^i0sR#|Iu0@aBJM8q`kkU==rnX^q&j`#RFlLY@sX<8Y2f=}5frF+}( z{YfMxCf$#gocm3!)Aa=7pxO7}UB$GZY#K4Y>KtY*&7WsbsPP$GEG*A)_I8~z3*daP ztzQ|JgVG=PF)HZ?_Ep^`>bbTIDe+ufGc>|U@)joNX#bO|1PjT`;W^P161dd`w_RxK zj<1=K+k?A#|7wE)wp&=a^IOl~hk{(#Px!f}a_iICP6VAplv8gvxE)_YydAc;(TM1m z6#qAQ_Wy~sy#OpTZW+WDE0lO)EgssI{bIxgn19pB6z=I0!*j(`TLqN_OPDT}skBrN zGu}5JP=i_fbkn1C6in}M9rCcQ2iJkXuPHOO8nM60H3|*wDd{6gIu4rL&ZE~~pF*AB zz7a&VjwgeZ9xq;p>>G|44v!HdfwsabT~J?JTR-sB{H75gbp$q<5ySMqpAx(D^ms%; z)u?Lp4}7LvKU>;bid2C#JA$V3KN<24aM`jrYu;ECGBS~*>U(s#IwY| zKUYr!_Zd;vG<-j#6-$k(lq>RvF<_!|U^==-T%VNSBUJX)V@GAnMl`R4;5yXM8txtj z1*mo^i9S4fTn4mCMt#y4K!G3oEfPoW%`b4iUG?D-5tW3yU}L6cWyl$76!QR?ts%Phf_}GH)G>-4qlc;qqs;7tyX8YTK7gOF3+{@(Y=69e-?9UDR$K+u zdcJN*bf4f}-1IkAdx%1NJpUZ`@z_>B*7W5ztRY71a!}8-H}Xnuf%E3<*)|C_OLKyK z_j-5>^Sdei_kM2hcHLzj|8kN!G5*+}quHdbg>vRW<`A1;4X1H|$h_1%%6D$T&Ugfr z`8#rzd9S6qoTw;xKpuHrD!M}AW`a;^$35b180e&uK3xiYqA&TJ7 z9ECPUAz4*CZG!9>mK!fr@)1&7gFz2j2$6)+Yq*X5T=@1y_e_ZAZTe`ko+hKHan7QOsH^qdI6$VlI z(Qx8N1wE*mIl#2BuGeeR`s>xwhuY?8H2j$4~9=Ah0l&P*Orp;SvM8&tG4*SINw!S>S47A50 z`Il-R4ya$g`Q2?YYb8)mpkHadzn*zG8?oi7T^nR{4dZ&K?I z=gGXB=cjko$D3gyy&HXdEurtcT{uS{<~3rfI2=XnERoouKU_+AK*9BO6$SPyG%}sg zvQmj+j}ZEzkm>wi9tOUWNjKbUamEyP?{Xi*?)Kn0yuuj`9`1e_`&_prEy zjM5HVim{X&LU05;*f;FiJ@6t~hnr7acE+DV>J;zV*zf@3OanlX&}kjZV!Q-lw}Hgp_I=`v~scwT9WiEsW1s9cO;~n?PU>gedb-p4_L1BK1?QtgL*PxkYs}?dSM-=)A(Dgg-ZR-6k@5%oC zmx${7ir-V-QwK*iJ<_ty8E^ulSh}2KDYk% zKpGwQN+9;VMHFyfDcQNBZOf@RiEF==HchrSD<}hze2ikegt7}Yr1N8(VzHVO;x2?5 z`BrOB;6gFv&O&-){5kvD33bgzTKz(4(r}lK(h9RnP0-Kc7z5e)lpWbep8#!9IH$2k zL1nWu#0`mmuWcLC;ltkh-v;3RuVug-XxlY+RE%D!F#BYe0vPhegp`)wFeRBOu~N)( z*dbE{Ws!(gw9)iu=*F3AQuJBY^CCW$4p*m_zR|BYrjXN%#jWW1 zpi*q%Im5?(!^;zw2Z$FwrrYs{ez-RHqAZ;-J?2(3!;4Uxb^dZdkR?b#k>-vet)8}fFe1%pr>%HcCVCO6$+;P2u%*Ohsu>6W~ zfff$WvO5fb)$@QB8W&h=94#b%e8)XXa#?NeAfbs9Q`W0PmgQ6e9PBII?ZPF|sa19? zY2HO%@}R6)o^}_PatZFu;P5jCJR+Pw<&^BFWfha|uj;;CEtReE2Oah1ul7$ryvT32 zBM3WsLpizS{>0W(j$v?dRB55ms~Djg@v@%uh*bqA^%t7*_y0i?3`&b6FR=k5R3lQ( zWt13$7M17af&)Xc#>8|Em+TmC69N%Y#0c-{+%+5x<0u@bw-Rz3Z%83}C`=z?>!x%O zT@NqxD;7i^Qd9bU0{(fke;I4`tMI*fC~9N9LZ8g-U-8tS0|*IXUY=&Jz9$4iaRuR! z9^?LdjMZ`r+|`~vg{{VwteR{s84-htc7hTbY&ZX|5Fc)40{&5G5?YWdUNUzi!msj4 z6t(01R?78VsJv+b?UEpz-MeT6r|3B-ly^I|-D+r`0f9mC?7Ni%Xz)-8h6iC8ZKb~^-+KX$YJgKFlERG%;c|Ed z2faqFPk1z+@IE0C^7M>cz};M5jHnsWsDa&6*IM4P^~4NYt>Hv6R38uAr@BT``#tmP zr{|NG?lEgNj%9po!q8T~a36+d$m92+YL%vR=SmC7J&|NfG-0Sw6JN=P<)KDcBGyuJiFA#d6t8{3h1mEjh{9eyUUC$w(K3;HHaro9a#~YVybv$T$zjD~8 z$26arE0;D2TMk{@pyt%5we;jk$#_S^Ii&Fb=ZXlO!_>nWoG(z<5c9q-=wi$dRxZUD zE+}DFP%<;(d6S0d2TKSME#DvDV5;Eo?PN$NVY3(4^NfAvTaM~S%@G?e0BNmNpon`Q zQP;C0R^`?}Q6SN^UQByGnO=2#kp~^tdk}nN74EgQ zwysG^z5`GoXB#=}5Avw6%Ce(_c-rwaNY-L_3xI|LFH}xHM>NWWY6o%h>YUPN* z=<$(5{-|sSxg*j!v*%dwB^&5!@NPcpRl@S`mydvhzy0#gm+s-5FebbO+0zYO2G9*_ zuV+ zyLHlnOn5#Hb&V^mqB_{GO{3rHKFB5t($-HBgk4?bNH=CtCz{$@J>u@!uc zKApup1wJC2T<>Et4s1Ocim1O2biKpgHos@&l=)p@WS7Rg+Zb~H6kFuo>-F)CZQbIc zu$DdAb$amp^us0iiU7S<;0T>5+{=nnX*Y12f0B#gAzw?iRphlnS%X|tbL-RoplUmx z14C45c$b_xxvrc9J`@5W$L*;F=PibCTFLv#n(pz^`&drd`=z~)ReAl!w@Cr7yW2Bk z;X<_Dwy$85pWixFEV-|AvQwL5?r`>%7DApUNtWUl(mjAo@$s*iqhi7Yba9Jy8;8R0 z8sCZErAy)xCSX!toP82mJuLL^|IbtK|E@Vvxa6uWHKyi@UrESujGDxgs-=B7qqzoO zDHTm!Uy#rGg@WpJgU~_#PbEw0e|4FLBQ*Dq_|H$7d$GE1CPkH`U3q&@l~??YWl~t; zoHniX3O%1e@Vv>%YLV?$uwqy3RbCfZ@tmT6WK2DU`(}n?U7SfHm{E^R&<~8Vb@Q2i z?HL5UaN0U$fc6i~%&MuT;f^wmn@{zjEKcb=@r~tG*;9B8jQEEiuc< z03{fqcN!`ZAe`K9)A3|QoW%B+SPZ1$h^jzNJ!?xaVjAua_gV#p)5V}IJpctAWiNys zsx>1Npp%tKskpv#m#h;VXYWwX9263cnS2Q2en&XU6;g+eZW{gEeqrvS81Ir>%dGyh z0a`q~>IeCvc(^C%H)P63F6ZbI`89K!bMw>qdf?8cTBScwQItRW2Zs7oy%l?isyY3* zGtuz8_ehB3WE||rZ#MaRnVG4_=TevKTF-+i1Q`=4r^o`$dbxkSX>CM{XsoEN8(AVP zcERqhvtj>45^pAs0OgUGlURkg$!5B_~Tzr2lMBq6T<9W>50E zi#lP#AxcD{ixLu#$yHDt3?KhwiG|(Rb-QtLiF_HQ0s8!Ih1Ph0{orFRv?c^Hi@WnDxOL5G@2sf#XEZW)x zS@+`Hv+BBK31|b9z!C&3CtD+oWv!08t}<@xjAZ*Tr3Ny*WYz5VT6 zeUh2d&FI7hj5QW55*YEs$nH8S{t&ci3LztZ<3P|m^1(=@thO$Hru`{@lW}Znsvh=* zY}+R%ZvOL3Q~Q$IU^WlIj4qe5Wj&i=#h&n6G|~ljCVS)W7-9(5DuYmilb^PCV zDf^QxcY=*J>nV!=ped`N0_?gV)9DUI#D3J?Ds!Yk-v#J51?}S@!eL^5(NynvF6T)+ zNL4Z(7Xn8u_cj;N@&sK^vrD>g(90eUN*#aZ@d9yzOeRu7ty;8xxOCmXfAWB3clvd^ zY2lK3s0znFUg?~+^U1}kyrv)JIns|o>KQp_pB|kj45y=U&jF+S zj#r9)RI;QXb&M5N7CoOd9@o^%$C=cQ+fnOKiYgAwK$@d(FQY#utnn!37zMXP*y*b4 zb3x!&SSBp_XX)AS#pE_EQ@KVY*vUbn{3pja-FKAoYfs@UPK@Fh57KJ9aH$QD9UWh$ zrISPWqx4pFDEm)1@WcL|gc5p4pG20uN(gQg0l`A=k?rPqSpv1)C3Z0TVLH(4w-s`= zfF#*!OPb|ZjI!a`hTZBsH-=%cHGFAO4X6&kJZFl{ z7Wn*5W+@Zg`_6U0A`FM3SbFQREe+i6`iUbFa@mJJv87LNY?9HltcUz8u!D1J$(Z-s zT<5lAB?lLAs~P*nV*>8wn?Kq*z=edYxSq;Bq;InYpe00a2J_*C?~&TLa_H%u@d4D} zRDVV8K^_b=ujp4HzKyPH0jXhi)ezmX6Invd5DN-{Z)g>x9skg4cqBJVua2^IJw1i6 z2nmxaD&(b(=DxBp#Q;=vHrtmp&wX9v9IUB2=NNIhgdJn2?a>4mxw@kz9@KTTC%h|~ z^bBHYum6QKl``9M#&4CXgg2!Ge4UJiq|PS{?Gjm)gQ73n_G0e7kQ?VKrD{07DV<}x zMgPcX#IB{c%6@&~yBtXL6u9nHXx2YTBHI%ht;ugmeN;a9rC)mQLKn`xJ^w1U_BX_b zNT!-%=m3cO3H1+Kx5 zbWz6QkgI((I%K!X8@bT|#~_0;(r1qchwC1pMs%2ifi?Tm9-#R~d4C?P)I3zbjJZn# zWxrXtw+E)c1ovrV<|L-@B>w`}2!OJD;;H<^>Y(z+nli=botK)lN<7w`9@gRC2i8nF zamBx7I2gqUWrMCWw~3C|&ow)~vF2UMB*V?oZ6_9*gmtdqf|b(N8aPVgIy%*R)Q84M zD-Cdt;fMQ0iv*dHAOZXjE($D#$*4pSGmqNo@&k6%cA{R+Gb{0O74-0v2o(I$p1@lZ z0*sj*KH_q`pxH>tKf@3ML!nMzg?Y)kO{G+8HVhhM;{;>LJx=<~0C zKdU#XwoH!vbXlw}ZSH(SQTO^r2O!Gc?Tmy+MEdB_I%A%yjf!O|UcnXQq4kWpaz~L& zA)oQeC`+4;zsm{I2O9H@2l*gaE)Lae6Xo#cy>JhK>CmP4)}qr(8vn)7igeQNVp4zNVzWkcHi3y1L)xQv*a+i;=@)GW_|w3|&FEs#%a^bpfqlJ6geXk2CAoYcc z+BqvBu8>lluE8@W{F+FSIAuOir>frgNHjx>IQnQ+9VnZ2O!)+K63D7~0CHwvK|vRPzVsZ4ytSkG&}W>?`53$!3#LD4QWo+^2gmQAVCf z%x^$eH7H?rU)?R6a zF{_;(v?N>@+zqoueNFB6O(`zFgVmP%+v@V_mn1_ks@F6BWkQyPksccaN|!r+@*=*f zBz%g#FnRmiigvgZlY;pdumC^v9u;T#5V+tIl;1m!@{}@yar1l4s`42pNwBG^>+Q+a z^?HhwX0AxZ$2Mt{b5lh)xD3rl*8nO_PMp3`c@@amhKA0!bjEVaoJtw%_~{M|zJmbaqr`)bs};Vt+fH10||xdqJ;L!X|v(Fs#) z_K&<{>-M68*x%7-x`nT$;t3_H7FnINiLV`Ap&sAUlPQmO z&3YIE<#JgjcwAF8?vP*VX$8`6Z!B=>KPLIX@-+*^whY4DIQrKVHf;9BdtUT3i!S3? zM6^S2T$x^KRL~8YCFvxz>=3WCmbzV#M70K*|J8pM52PlSQ@L9qlF) zNQGKO%ONXhOedgdh3$#HP_Qmsi1-=HZ^YeJ?nBvf6&4F{$8UDR| zKHG1oqb?1Qd=g2K2K{gkk*M!et@<}<6|60<=IXQd4Nb?=>>DwoAD5AYd#n8p?W_Kl z$eBOAG^*4Oai_>>01;NqeXXtpM-L1N%dq7!(0foB+ zLRMv{F{jiu0uLX2)^si9j8Er@Xx|0l7jfagI=(IMVVm4j-i~uVSz;o->W@M8nw;UIBBAKRAxsW>Zfe0 z_uiZb^LS0*h7wU;d)(1j!P5rGS9%~1wO*RvUEM(W9b)~*1<~cVeq;B#O<@9Cs3N06 z+&gkOjRZ_%6Wph2Ua3?G`oc&{4!XK%-;YcdXS;dMO)&3lW>`ri(@`MU4Zc||i$)qR z^iAfCuz$drJH)CCqL)RAB_3x5Q%c4>tWjcEllSmg?z3x(RMdu? z$9Y>;t^D?~$olI1BZ6w2eN6-~B_hX*d_R{)8`EDcn5Z|aH%_@!Jwr8h((#v0ZoO|v z9ZupYRpepRK90LtM6^&wB{m@A&I!gK1O@szEE28PJQi{_d1?q(FsL2Tim0!&jyR z@I@mYV%8arQfag5`Ilvc3C9!jndop&fy7~$%_5RJ34+SNc#Rg~{;it#N`zzFs@VSh z3>xTu7=hnfjYMcC7W@6>F(Sk^0mzjn$!Vg~^aWS=^*o(F+K?(9veQC>N}}ZS#z9@e zRb`$qlnhKd+854&{6y*qcPmPX1Hx@O=a^p#nVh4Ts*o)qn|O&cdHehd+0){db^UU4 zdv-;JRo220*+l^WiC9XCxaAzC3!KYJXl@KgCe)r*5eV-#gEqj02 zdJcH1e(xhMJAa#FAr0Y;&%69`Ciu}#kQl)ys|eX~I4#T{{HyRjQ$?H7SzZPCC>Vc< z?*H>1RD-C3mJ}u^(_@2b{`nH26;Oj`Ni_pv{)QCMdD=NP?R%F%Nm7w!DQ!mLHaSSj--oX1!?=uNWcVSh$87 zSsgk7qY^n7-6qQ!lNa3%$%C8)h8?jnI7!_rw>Bj}G9!+U0WE-49Rs=LLmQA-l|Exi zLhF|}1e#P3*c+&x`@F&?5jl|!m3~FZ{rby6vfN=C*Dk*A#b!QwKjP2lsK8R4;b!%t zG<6T9&gEuax#F?zN@`d?!x$1FfvqdtfBPFODbY#74j<)^xBVKnvwIC>wkEm{0EK4RwGW z5=)^hK`K&DKXCFx(=L^9AsI~Yd)UG%5-u80xCG-s%buJbod)AH_LdyOEssxd;Bgu} zBxeZ%%n9&#GrL>LTB0uuzss`7IboX3AI?z*6;_;4ots7hC9hFGk9c9wfDzlhu?Ij% zknC61*#mzpuK~{|aJ&^A-$mgDnp;rS1EOr&tZ4by096iS@)dJ zdcjZ!nXOPxHQ!a&IpsF;XjHHn0cd}NqeE5DtH>z;Sm@gHfS;E^*CB(x@qCm4LA3`^ zf57UQ16z{jU~gLXVREt=W2S-Dx`z>x#ZO1nD9#lymy{tI^2x0#8Btg$3HH;Od+dk4 zLL*#1wRZchsZ~H6%r}J`-PHu?(#33C&VANbZ(uT7TpM+~7GKHpO0a6a=w zjgfkPF2UAawGogUFZH)0&Y`*vRK8wUf3WFNuNJr0CntsZpG!!EGB$)X=uf>>FsAY* z8v`)wiN$@74JFnZV6Xq|(EnPYbtK%Y?CUHSOo00}u?5ay89h06>d%udA`NmVmBg1S zj`l(-z*c|zh#b037yBQ7c3fitP-q3`nQWIkGCH#%zh@~FZqQsDKcaDZC2)dig$Q*X z6K-6cH{L~6Y&KZ7cX13ochLU{PPt>ULpaegcrR9Y>96W`WUSe$KNV@9g9s$^msPza ztb!Z?-1yZfJ2jER6nrUeH!S%x1Flo{s$X56#+3zL1I|9lI!zDsN7SRMGhSV*&7fWms0l@FQ76bN;vC zGe@r-z6?<~haePq*{=#SQNttf1M#Rr_(``;vB^386*Z6B0wp>2w!2PBWlCz8LDtn& z@&NSDL1c!Jkx`_>A>S@WBv`Yf1%~wCxtwZ2262Mtm$+k%Rg#zYm;OG!GWVj|{+g4_ zfOUEGf>7>Hm>g-d`0J)V9QLKN4GN9C7f7`nnE`APiCb8ic$^v}D;rwJaM zko>}(v>Wyq*w})>Z%~g6h>oUe=v652bTL$c)=AE>Iba_3c{%Y3si=2Tvdc`YQb)y_ zm-#4oBgq6GftZCvNDBj4!)7_?n{CrX>r_oHw}gd7&f-E)MtcH?snFTGhqi~?Jpbz z1mI!RFwi$cdP}y{L&Y>FMD1Lj(c4u;f9yyRD#w-aniB2nS+-S=fSfB@7tjsbv&a!; z?4jU|+#_Mw4D>W0y7%cy6f#G$QHvE%$Fc!TAAZ!GcrOkqCqeq{_{grpjYb1)vnjd@Cp8@u?8b^ylTw}A8I#W-Fz zC~(Vqu~Q9b_K?5gOD@P^d;Si?hNOi#sd&~?C2;e7wfUdpxGuTmy7*JeM~)x6rlRj? zTS_6Dw^IeD8efv7PdVOwv7i2IBQ@qF0c#Q8Zm)A7O7;RCKH1DsSV3n*LxURpCl1(O zzxa(hj_r@~1*gExj|tb`1r} zlinYxay@m|Q(zBbggUS)u6Q!XCqq++v}0Y?US=zQRmoLF({;&9R_6C`60;!k5;b(( z{C389t5x#8Pi%DDVCRV8y(4w_%M8rf95ZieDD6@!(a8t<$0P{zD5e%Q@@eP($Hb7T z?SBnRJ;1eIROkMkuD2lI@IQA0&;xQ5;Cg%V!=)ET6vcyf&e7f6=FbkB27=zNXvF^f zlY~po#YeK=`11Km*Im^|TZx}XWbtQ`egKLTs4p`K%pC{DFc68=NW2(lS0d{cli^OT zMU<4YkM0dOe*)DkZu$~1|11q)dDyNAZJn+yNayn^bQ4HbP`gVFhYn_wqSt?ZFG8vPNp(MQfTsk|ye>6wn-nUra*lzhti5aQPt z#zjfK^i9@P8~u6d#vXI={uq{SPp$IntZKQYi5{e+BYfqLl;z-ekh@| z=GW!^CZ};B!vq6wAS_;W=bwTxU@lX|-mQ@UsTlI=JyHuI-@kW+f!`}Y<&>qgxs}%3 z@LDRvkORKG*>_EEZ8T;=x6WLi6-CVAdzZb46yhD^oG9lOh=7VU`Wz*rQhye?RK%+^WT$;gJlC(3`-jlp=Yt3mqt>K5UwmbAaOZj(Z11ha9<#H{t$_aZ zZ-PYDplqFsixs!(2X}V?Pr6YQh7L8mbmL<;(;%Skm_~oG}CKH)5;DUUCQe;y=vMo&k5=s2RV*Oe(N!eh`IJ zgf|4kdf$jVXny09k4ocAlf-~KGEIIhUrveCEnJ?ekgF&R%+hY;eekc%7*Rb|$_=mf40$3kN~-Yab%4$pAUH?1#=Bc}W9?RG;9G5yyV(WO9^st-=`T`^{4=9 zR>vthxyfOC*X(YN;sU{{`!7e!AFbSKDn_vzp#YBL@Kyz5;&<$aZ=nu$WLOzVrnHY5 zsy?I5P?332F3e>8$9#j&@QEp(>3F;=!S~@!9PJF>zFcyMy0XAg4tg&PKI1d?mC24p z8}P0UVsE+17(|==HnEM{r@p_xB&|k!{o^+&O}s25kBAusLUWD5|gWMMFR#FiEfbRdADbhIr@kN7m zguWw@9O6@gzXNj!k?=fBlma+df;mq3P{jV-Nx!Ch+BTsOoNr1r?oudzvmCE)){lo9 zH*IM{5A5$oA0YWz0+J+?*cX2&QUo(UQ7VbfF>&L}x3rQuKw0%flAJ%SCiw7=v$ zUZp5DyU>3jmoNVEgW!3E6VtuSibxfz-xZ=wVyBJMgmQ_8r-<0nDb zoUnLmJn&zfDEHu1Nxzcg=Ve}Iyy&=4lLuW>YCy_NF~DK7>t6l0#wn!W9kR8dB9jDh z4kMV`R z@ofhrp?L+ov>`((e)NlaYbvmAi)ZP6$ID9-L-IccXzko5*!M5 zcXxM72<{LpNRS}G0t9z|b^EQJbl*Fw$r;u9_c{Oml*zpSN+BBN*NI(|iXT5kfFwiW zVi?f7xIAhI)EDy(HbfGL_Z}fjHFF-a za~$K3yrRbwZtFNPO1B`Pqs7z7{D%TNeI2n;mfTl~u3+;WgBm?OY{y(I$JPpEFh_}0 zj0CaLG~%YYB$Q`O-h^C?Ww<&mV&qh(!1S9XBS8TT7zh1$%#XwV4(QBViJEDY5=J^kbBbR1?>1 z7;|zlqH~vX^X|3cnq{-s;akXws?q+Bf{>`zE-EC0Z{S2?%d=qtA(EJ-YfN!&SY24w z8V0{{g%gSW4FaT%Ir@mEo3JwV1}#fuB)ydEXVpkTO7Qr3&Ij42#@7~V;g&=W!T~_mWQkxAMuvQ}?`saBjgLx1PBi!!*kgE&~nZ#F4Z}&QjqzMU|S zNn_a_Sc0vIDSG$F;|mr4>IM)4a#*7;_|}D9kb4c-4qCyFf}XD^Pa}S5_Z$-AdlgG= z)NT!0ie9Z{57XdQs}(-igYg~^Pm9tedT4_*oszQ#)=3!auS^U+UW{Iyo;~tL_RoC< z?!X`!&`nEQ2Yd1%>&e@^+P69(&sVqqdR#t4io$(}nBGln{*I@zG*enQ7-C0)f&l*e z_UORIZ}+W(`=nTmi63ffgre}i!)TNKj=NA?*w9TrtdJJzDhvC{7c-?bDm4TG+p=e z@5;fm2@<__5g_qLlyiCMyQ$W$psq$tHup73Pgs;o$}a-V2%bNGx5-)FO&=I=j$>%1 zTdPY8qsRGJuEv}iqRS!eX7_Q-mkMbQ#!T>=gE-jkW2a*!NU&3EgJ23XhYNF`Vl_9q9_ac1`p0;n%JUU!nCjxw9`kB5nya_y({5 zT+`A2U4I(@fFp=Fcj8HvwFgn9Y7`s1i%B!X5%XA?v4I*>=;d_2n9m7E`!A z*m_h+J@oc7!JlQn^S|A3MI`QRS9NM7!Mgm@xb+<7!*CTG3V=+`NVQlKM9o!bd_ik`asKtRIx3zK(BG=7K~%hav;hH27Chg&f-gVQiV$ZtyY2sdTEG{~)xW`DbF@vI(7UzrG3sXY|0e30)@fAJ9m7VFNlpuKN< z=&%3+K9Fl=5Kbg8_{vp~5hE0U52#+qd%A&>WMq^P4C!dKWGBT4N8{yypD1eOUla%%08tyjUcI6IM@F1lHk z)Qi#u>G7;~x_{NaPYL3CI(3~8`1keje{EM`qM^L$j}}beUm^T|=!rR(eJAH!?S|~k zCcilez>>X8);uW$-SVVSc-pA=*?He+fwtASf3?~ZV?gz#S0TO$uBR?i%7b!BBk*IZ z#Dfxd-QT*3GR{Qfaj+A(`5nGnTvbeb)iEoi`mLBGketYtuzJlUh8H|~ z%4lQwED|zHJhO(q2SQ@901whx|1KqepZ0B4DyGOqlClPr%?x@dSgr4KH|ceI8Up9$j+pJ#gHxl@b3Q%mw z#iIh2VWx4ZCEy}}9bUdZ_QybO#kFHve(I`-KK-&#y z;SV9UOphU9J@BiSYis{TU0`+7{N9TeGJWgy?k5v{txxnIx?r^VxeFKy-wEjmgy4(c zf7G99G8*{iV)S37EB_BzRVA7&Ms7ZxOFH=BYyHb2zsgbnlCq)3xDyGi4$OzXpBnRX zHC>=qPmIrqe&ghA-}-e46IutKFUEe`~906)1$^SmLZxi3VX( z(n%mbYd>;?YxpOUvnYtC+~UW0&>)QfvpWLrj^X<6GJZ|lu^FAeMY`FUOF_F{1fC6zggbdTzo{SISmF|ki)n%g z^p3z=sNen_+|a4n7}9jjhOx^rOOcqGrWoE5uaGhwpqmse@D+JR$%6FQN459#^ZH>m z2oQVjN;bQ5eB1D}##*1la10zC$0*|3U63i2c-M-+rJkpV77S1`6ksgmetnGVi)kv^ z?=qVftO`?d?uCCy1@iK+meI}PfdQvJpJMf@pLToEyXE}lIVP%vEljEmeuO8*&W*)l zX##nM!g46aoP9~aI&#aGz0xH?e+_7cs{gpR2uX>=ULL*u9on;em`>xs+UkP=RK#`qF# zej@uD@qA9F;fAz$k;0Hh^~3Bihoul(%8vH9w{SQc`(>W`#5`w{>*O~hxz!sZQsA^i z!0mKZd}(su%CPW#;`zkaFF|ruZ#w`7-p8n8_|dk$aX^=eRwq9&^nd`cZEB-u+cfTg z$HlM#aaait!A$FAiDAYh*xuqJm#A~AR}rDm3djX#dl&KluCy)df0$kT*LtyTK&OUt zKI61I9=Ri9A(@16a|0TqJ;@!-VUe95jMmIbkf#=N9d_1r^s00r<1L6Vzp{wODbFu9 zHo1#qHp}kNDlIGme~(Zpm0ar47xBdb>P^&L!Cd?HW4uq&2$1a|A^DD zo82Aq%-uP_p$a^i(BPS|F^Beh{FH=RyUKELW{2fPx^6uZ^c#uELh zBQ8De`o#`0=R5=f9)YYt&4^hr9*0@ay*xMyn&WjnAH#w7-2jC#X0)=aGiQeq=X_!9 z%`5p39BCu2WhfH)xS^_e5;@|*u_>Sf^PphtX(B#Qt_*Ae^13&78GcgrHbssr{KPHjWRA;T>b&pi-9c>;-h}lMio=OPT*yg5{TqURqD3Ypav+=ysX|1WsPr!riXy!V472^6ocmj`$l5n$?S`i zIW5j3O6{kye{o`TLb;io7`|-jAb~`?gI^ltQENN#1|3i+_TYDH2UKsxa5cE2`}>xR z{IGb0k&wA+#2b1V5>`>zrH8bNv*68+&3c5AsJ@MV&jiKUiL}>9HZe2g;&I?u3?|i% z!txah9BDEF2s-$&Q*dVD&yJNe+B_8&SoDCWX2PSp0dx@;sHKIcd!SK5#YAW(K9ANf z!Tq8?F9w}_0GCjiZx-zUo2QRR1Qy6qo$=sFv3lSz@WC{MUp!z`5GtVpS!4zF4+Y+3 zAF}Yx?PrR_<{vs+k*fxP3)iYY4z|EdWv3)44StHDj@wUZ}M0 zg;N9Mu3W#DmvHsko$IBBBsvy8`hJvZX3sd^JgS2&8uBw?~+0ctP7SS&|By+v7>+$2km1$7U-Bd#i*KH<2M6jxMOZPyl1~4C}fnZyp zvg&*Ny*#m5*??Rvq=||sLlL(!#wZ#tu}KN*&o7)Q7qyH^*Ai*;Kln>E-)Gk|X}S(b z{X12`Iyn32YgLSZ%K!v7Hnt$vW~NChnvWrMeh7CbjORFin@&ppo4)El=KvKpFi2D5 z0^(9I`ZK%adJA7Gj6l2_?2cqe4gVFSiVJZ|La)OCOh6_*n7%-fMFw z+o4@j7b{X*#ZfT7V)(GKpwqan^HxF~1DnLgE}r~jyIT;u23H%c?aLe%uSdh|cKvPm z7>~H$tJS*k6!mQ}47`KQeuU5us|i#L;-gyLm8Ivcs>4|eXuz$QWcWng$60BBn>N2} zyLc6T zUurP;KTHFWyFHn$IrSs|Fq21Vt@EyYr2A$cC36gb5b@9EzDh4I z4epUv@!LZ7n)9m!>gj1$XCXle(kI91u_1hN52y%_YC9RXr3JVXEVFW^S!m!b`wN5V z!}P{ya@n96?k}zJp;g*=hc~}knUE|jzzr(xcDcz_|F&8JiYFN*WXCRjWSASo>b0=O zIDM9QlUG<7we(t;v1Bx=(%2Y~6znGmjCtdS?-@i;(pzNQ=2$0VWn@tL(D|~;5%5pQ z(?XRjjK6noc4mFbArN}yx;dgsDW>yD*usY>V)yArUBkUKHQWDJiaR|D={Mwp)6eM%78z6qP$e6>{S-|rtgsI{F9!hva8Sh4AS;Rx+$CQmgv-N_sIzS;^!jamdabU% zNizab-5b_9lFvuQ{14>I!Jj|Z)c4)j)&B}77Pvy?CCK9MYiyi&!~A(*Axq&J$G*=g zLwRbT(NRJ=F_2TN9GVF-F3--?^g+8bIPAf}vlO9yF~_1gi5l|4JEh7s7i+hPd5pXG zsU)BbjygzkZwi4MH5YZ8i$mt0HO%-)dk3yF3DaF}Bw`U$ipRr`yX1eI+N0t|`yNss zvV&&0X0-6XJ^`$eqTY5m=oUm7tih>Np&D_IBzRRkDI=iN z=sN_MI9;C_I>+`{W|?`UtN9b&eT#pvDo2Tus8==P^O<*M#?s$SqX51$j78J-e31R} zw~vLCKmoYoqN|W}j6sXhrV>z0SvX4aaiDck{)XU8xh0~k6qU@<2iI?H)+;1>CTHFN ziFT%O0$=Bk;XmHQQAac_{HxZm$rS>cIE6<^{;)Vf-KZG=baRLZ*GH^Pgulv5GH3i2 zP{|H_B!W7LdyO&fkmO0hL# z*l{MrEmtb5(&1i2Bj21mVoU*}kl~=+fFD|-G+I2NlT7RUEY>)FhhEM}b8wm7WJ)fjDrO`PkI z2%IK3|4hiH0YoB@%bRy7qd&Pz#N??PYs4DV zg$bwq_2Id>j}-K*3E*NGRo59%MeO0dN!EsFw)nmMmEk+)(9wrtGyI-Dug({M57~Z-z5Q{hU+699aX>_lnNK zUe2g6b-k+o;_^rM>P5UTpvSwf_5szy+xX6_yc6 z$@_HXcMWp>Pm~z||01J=vN(nrLOvM2fe@D8CJ%F|UyK`$oln_IKR??8_yq*07Uhs0 zaU?8oK6}0tl>)83nl*d{Md50PA5U`mNOivMsMh~(yDv?*mxJxQ3B=iK`pV7sHdrp3 zy%E-(8N3~=|MQ{MRapG38!A&0tUKoMs(Ldt08Fg?3CIm4UNaJ|87#f3qr3t>=}rpv zT$sA?y5?yxgl$JDNp%SmYa_eW-<1<4Fboe&yr%tzQfy+I1FTaLOM#0=Rr6Q!w~tm{ zSV>#Y;zD?4o9H%v_8;At@#B{AVIC2!_C!pp&o%^qQdf&AVh%3gPt=e6qhD+*`YL+o zQS??DcPv``t1iJ}3}8kSbI}wvid2HR#W|pfHbi@&n_2E!{epd7oV0yoZjdxN07KsF z*}&dbpZ~iBZsX8ZaE+jM9aermF2rLz7v9~MWtf=As4tqkNU)zc1%w(pQON*5Q+tvz zOr@rLYx-06Oi^suuD#6XQZzN}52?q4T2j@hERD2}^Dm7df+XyzowO}DF}t|EZ6HHl zMYChv#+Ol9jv2lQv!2%uJSxT2%vF011%}^~qc_~S#pomEBkw~YAT)z5-8kpcD2e|rRA#05(n!|!d-vg9@ zbO`mF!Y>5WlkNrVbl0=he@r4y+sg;LiWzut;z*MD2}w05*u8Ki?}5-Q;|-%*dNvQ; zS=`|#Ve7U~Pp2%XIX8jyIR}f%u~^COm8|1p=0Oie$r@V5c+c%kB~bc-=;rik*tg(o zQwiGQ#{US3fR4U7d{!mh9?^}oPjH?w>LpEjSOF!#fKoleq= z(IyLuR=7C(y8Hof2oh_yUQn#W{CEZ40gs&#J`WF>ZG^W9w+#<$@aybJqK>TF+G%Q- zA*v>XZ?dF?4ceJ(Vvqz)+my%8xRb@e0vWlKeo0@z7c96jsn~6(lRhGT_d6%Cqm~?B z2mx7Vqf8IwNEb(w>hq0p!Tbs)wICjnr|I-#{G6%ZTLIZ znGHs9C0sJHlOC2s(j0qMdBWCU@(0VF)T7FDM?dNV*D~3Xz#FGB18pq=q@JEt#hJ!b%X!pVk^7Z?k*CWWZ z!fshYPje}n7394(oC9CQDd;m|q=0!~H$uXnmc<}|E3UkRqT$U7-4`ygP(xq8b*}_O z(3gl4rGk*cf-1 z`$ejQ?c9x8>7iNM$LjQeZe`F7g_axFLG#|rM@drafY3e{ieM`J)Ud;lPk0THv+#N{ zw{-7hziuTy>G>pw@J(CxiYXhVZlwnd2-XX`_wEW?&t8}#x)hElm4MdxsipHYtP4Ab z9oqIus6yJ_BYE~XrWSMquTtS||E-XT&C{g~j~uzb9dJ5+Seoq|ADgf#sU}{uBu^+6 zaCc&O!!;$MG6dKCd&eB=;*~HXrN2%AxFyY{7mMs7g2&x8N3T}7D~5jOU@#jre4fZ4 z0US`ZYy01?&W;rLs)>pkGY;aBq34o|%sbwHxI$rFZ;y=Gk$~hmwMy3M`8#ug>6W!v zJZRoO8@x{Wqb3$Z#jgK-`8PCzcY_>G)g@|t3PWYJXZm;7@2)FT2R`dS51C33MFfDa za|ouk9aV}NIA++~^9=aG-4mJR#V0#bfyX!h0u!BTDB{?MV$_XvHAV|X(GOH@ zmGSDnaKR#mz2Y6itaKqCb~PjC9byDIYH1zt;bd@D1$Y=sR<{ic(fN03BaSkB`Y4?e z-GlqpC~1ueNv3ZFW2pd%2;V$_NUVCDYg0S!ut~ACV%;|t3|JL$YdM@0h_Y$Lt--z? zCPN&*&Xico$p3Hd-LX0dzFb}$E)uz(K-H(jX%|IJR%7TF*Rkqm?mgyjHX#!o#_B$ zAjaH^XXe8L*PY;1n+7y%WwQd#Jx81roWsP3Gk8Q$pl`l?fUO6tk(Qs6y<{Ex_xohO zIs68{IS!eu-AwYH_+K(y0WyO zhJBX*`}A$&6%Rtje+&Me8aFOQqY0}Rl5iZ^>ukuy@G}Uebi_-`5~;Y8m-Yz51UUy*(c7pMBhUJLyXAM_f00{fyF;qWC{>jLUY&nFI; zl?qfepZmR~Af><_-62{e=m4>#BInuv3X9|h*<`SZ)AIiP-^zFN7Obn!GTvl8nE-R|@)tGxoXHp9~JoeD}4 zu4nt1U85W-BR9p`H32^!l^v{U!N&zgK`B`^<5VhhyxVw8q8_}GcuBkocIR%E3*sP7 z$5S=9`MUHlI`UrLgp6Ew$Z2eQsw#UOU3wV*xv8GgeYq4^wuO&vQY>w$Y)eVupj9PR zCX#<^-LaJFYt!3P6fR~guq;Ot+g?G5rlg=5)#1+@nI2Z*%F*}up06f2tfHpM$ys0S zc*$ypL;R83&nivLFOIDu8e_=;X|nLS z<7Z{RTHt*BagBGGy;IL95j3hUXREyJQ*J-P9_ampJ56aM77pJYr(wGst$J>V^QR$M z{6`JN5z5S8zxiVjiiNbCrrdLz_$=na^Pp>;AD%@3mcD0KE}mF1jhCpC z0l>+3=HAoIIx&Fc3$2+j@zsybz+IeV?W>pWqsyzPtoNUoD(`ota4Ta6=$M>V!xv0R zU|!8`AP&8GlA>zPDQNAjHaUTbukFJAaeA(GP*KyWn#-4Re6~Tx`T?$3iUxgb8vqlC ze#ajr%7G~wG8A~f);kShBg(S}HkKVMdpsLs^&#bp79Z+!Mo4z7V`G;}V z32~I!LLBNTGdJa=wXAfKu%qfF=)2I4+OW%jDaq03jdNKLI>hgblNU65bkW=Q-olYN{>JZ}k7Pvn^a&?2 zH{1=F@R`5Ky{bEo5-W0%icG=?$un*u9)G*&_S*`p%V1@Axzz14w8>m!hubo09kREA zpw}s)>EKPOS-(GyM?_Kg_jL1<=h!>;4xm_2Y+I{oG|t`7&=k^W%vDw;pJ@hFeF1EByFpD@D4WL6n>+XVQ*X8Q zI)`An8u}H)x%MS62xrf1SER>${U(=Y3?llfd>k1}2w%k7CEB)6la zIODAgIo|RWi(V7(7EH;y#|GeUseWSqYmwiR~fwo5BbU6CXl4*a|_Ramh_e zQV){u{+m29tEiSmZPa^2{)Y}Gsgd^2s8Tsdp60b^|6FcBnLVaoW%+W$$;4x1Evn&M@`jQv{Pl~c$&69Jm3y~6V&8gBw z&y7A&M|eFprm}vH+tlblsIbF+|(Dk1sl93Ux?*?BD93|g#-&zCkm6EW>aSC;@NiamsCV%iT zjKm4&&$YzWn&pazbL(X9tL0ig-)zHgdOf17?b>~O$Pj8gxk`Z(>{~fq2m1S(!!D7h z7d@+UBf)w-;5@fvMb4sZ>8mjY5p4Ml;tcfgSAw_bM)Hce zs?nEaCa>IUjt--It>T*r7{)Nd-I9iF{C}ADPl&4DrQv1|DU7G5vKt1`C#<=G=k5!g;Q|$_D$B z(0b-wW}q1*K_GW%soTkjW&q#VqXI`*!Zlwzyej&VPK;`NB{vrDmegvxtwwQxI@-{SF3lF6U#6%LU+~vC{PaN1WQ%!Xk<$z_T?{GwWo*fv z57_>AcAD;I_skjmwA(yxamY*C*zkY4q4I}Pa-^Mb?wy;l;4k>s#$B~vrz!0;GaeGE zG()bPfj$Q$wW^0%aV<_bS}Dx_OOhlwtn1AM>)26kx>#RcdQ18Qf@?5Ioj=hu)-=v? ziju|bkoqf6FUWF@Qb<;VSvm=1!{qu50drBjV9oiN#G(szdl(sS&bGfpWO&6=p0a95 zOl!Kc!id|ja7#F5QrR->;t2{CIf}!a(qlcy=~)!SC%rEJ{B&pe{Vcd&A7Hvi15InR z4eFCp_dqk#zd%uSI>p3gkOcS)_7vuUP!f0J72s9HbhJSmJ+e&4L{kazVLWZueabkB z=$CqmL}Qta^PY}q6p2AXN~(Q(a-_+~MnU zWgLrbOg#+2DP0g?#lFTgQu<%`c@?w>H-cfzy6ze`#k)*m85@6OvKhY|RN7yk3(xQ9 z|F})~ZEwg2+jTDuRV|hp$)UFZucKT335Lefjvbr=f3s;u`B#?{3kG~fswX6uDP;dM zlLu)W=3`kZ8}bY``?lj(#UJj+bs}eA_UAVUKyatQ0N~a>*hknX9G>ri%sJ+CbwuD; ziQU8>OJ)zHT$fWIMVm1nf+mUfK@6PURa~V9nPV)^z0b4@R%GNWBL5vg;=>KWmyh?c zYj2OEk%cr?wkHYv1$LDo`GAN7#rz%iKN0IJ&{uv#kNjupRfOJ>KS?B4o;vgH5#e|J zM+(RAT?(fz=o+=XL3%z5e74N2HvHoMAupFcEh|xkA7Wlp;1KUE1qg)QmrK(dgo#U4 zr%htVN=Dk@lvz2~j;zSnJ!K(PX<_#`LfwxGNsydj10t6v=IyH3d}BuyiGoQAD?sZT z<5_5M8cGv7T{bb&*wN`T#+mg85~j@6%;W{p0xmNZm|)zmVDIk?+a{ovlh4s`m0M_H z&UeMm7FpU(AUN;=3P>$do{Ae1#>Yu6amY1$NG}l-@2dFJ8i0dTQXoIFdF-TNd8Ylz z0WSAMg*}oy?&u)t1<;EEZ8j0r_5#= zEW2dbK`YCPWe5To{D%;>uxwcEpk_fuc8J5*nG`dWhi*M$e;Ux%dEAnAYuusJHmXny zh$XCi%ZF(=C%4>v5AON$zSP;d6!h}>wfSi*=sX*J^+>12bUVhiP^fdShz6*`oUCO8 z%o^i`Jk|4yU15?-9hyppHfDGOc}3nPK4^v^(d!cqMWOfhmq(zNR=`noCoGODF^U%9U zI)b7tN2U%Ar~wh-$w&vIu9i>@{qX%RmH(Y1&GoeFG5E0hzQZUHtgL0L_K!>rJtr@W z3U^+_qV6y)3=yr<^MAMy`W?hg+d3L>8@C+4>!U{^x5tP$kWvYLx4J7XXFgBQ;<0hnlQYk=Li00`aAIv<#;?&DTWJ7laffn_M3nR^I59%*hR*H~ zm>-fpJ4mAU_Z>?#;zC|Ya81zJ*jpJQ!@C1vgHvr{TDt+OFxpzM`5QE z11ZZGp5-<-F_f5(Bus*hd7W71a)~r1D{BU;0Bg=NcLn)W3gtRuY zhUo?@vv0GGW>9*_qdnX%D3_^fKuOYQCMWYjH8%7}U8e#L86QW#-eA3XUuuHxs>GwoUDmJ}|Q(@_m@b(n|%C0=%Z zQxjWW%(ESYR9HQ`)nl9d6x1SUUe8W-PYQqdBd(JamNy<5Wc{8|vrjx6AkW=QJ1*AY zxts~mJygnlB>eL8)n&!|e~o)L5TAtw-5I+{mNQYzH~0 zK%V6XP4}v$me;{lm9azYS4CP7Kc|8?U0ZT&Q{tG(#otvVpOL9p`%)cPX`B<$y(|I! z{b?E4q%EQ1188DuZ}6aFr3)A5u}xP?iF1%~QkqTA$DvvpihFa&OETlC1$PzZ(ex}` z-m%k)DBYal9u<5*?b8Ij*;P#M)whVrvWvDyIyGjCzb1sd_!7`uG#ZR16Hfm{c}hpj z=udtB^cWtVC~JSU+WaF$b~+zDJr;#I^llSS%gs8+eDX)zL6T8GxXk{umaFO`6cpz- z;Yw~xfDjN%5gnpG;vPshF7}L?JdTnn4#{Eq zoH2~3rSNau4lS`-^Ydjvs2My~J@wZ1@1OJl`UP{AQJ7x~zjCw3_hk`4iE<@!5fCd9 zLI-!J`JZSwKJS%0*jDI@9pqbL^arTUgtHQ$=ud4zwc$8Y-??ii>;6-iAs(AS+s^Mo z8WNMBYDP08VGmR^+QTLGUu7_ISaI;b5LHa*ZLV{ExqYs=U;WCE>3+7I=)I^2%#S~Y z!FqsuiSfWeXmd;Yl@8wk9`TCqRm08Mjc}(V@cE@$pL0pkEv24HuJOuINPXhYw70?O z)O&+VQqSluHhdLZHR99eTKmwjfZ_U`>zEVI-U@#f_=pG|5WEb9nN?Uax8A$qYmh+K zLAzjFWKn-{rgi;o--URHCXq!*SuOZTzf5JyG9DwdOie;Vr5|Jq^bN%D3xxPvakh+Hyx0nQ!BE=1YKKM45Pp2`_3ipXc( zY=lv=Eqc2vj!?1`rrIN`5ORr4mB-ptp>N;yl_fU6OQnQofG|BJNbj_I1V^C4DV;!> zhj^8P%H5w&Ob&M8FE-CHGO4*dOX1^sR$>bm8I^UHZM2p93)4FCpqq0>F=e&K2@8pe zcz(H4Mtt!8ppt_AKFCpn9i9_%xnW!_a5(!Vo)|s`wULmV=YV4``c<)NQFQxYyxONO z~&8rfZd_SINa}#bfOaWixQuP6#)+C&h$TC)aCuBQHZ> zJ0LgHW>U9;2!0abmF#z7wMwo`b(mT{AyRvN*P_c4N%a-No%}&X(*olbwyY5`Kk*=m zb)_dEnvQ4!ADi$eJ&k4Otp%Pqoul{m*G#&rGu}oV{suEkl_aqKCAJYvJksP0Ygg!3KAc;oP1p{?O`|G#a7TH+Nb|Qx(l-}5$Y7n3=}aW*%sLe!IkfcCdmx3)Jx0Rrm&R`^ zzA^<$`hP?0iZ*>TNIT>_lW3$wyPRFPI0Q2E$47>dyk}W7Gwrvc5dD&4N<+u#rb#X?1nlC_(V@W)g0%Kc(Z}bSJP<(&4__Hgh0L_p$)s4NaPEKBs z$hw{ZiGErkC^V@vd`0+!Sql`<^shpE7TObHidOzCsJ){iT=z7dFAlb4m-1%%5KQC1 zD!XW3unSi}p5j9lXse0#R4RH;4wG#YXU|l+#yxi6gvM?=IyF?!R#r8``!EMJ7D%r0 z9lg?1_bU0!C_`L~KXeWwyfwW9--X zgLdU7MJ)HIuR@T2b*spGphA6SZ*s8g z5L+b5ofLopYgQcE^hf-jD^d7SUGYBllS_h9Ipp@;=3L2wX`UgeX=;?L$aQ}IHb{>xRquBK8*0j>H@+usvj*BAlj%9To zb|_%&byu~C!$7A&&1Z9TFa!Ouho%5|YCejB2F6fkHqJqZs!2P)Ds$Y7={|rD^C=;? z`|g_o7L{qAlT?zXj>eOw`d%?axOhKDgLlDAljgkL8ZG%PWy%xNo;fkW6h%mMu>YSQ z(S&{^<#)`yel3M<(6$J-)OH(1zf1UTRs$x(-hWwvpleFzL`lM3k4TiJ&hoKR=1DWS;|Gg8czDiWF(b@=GY# zeScL?g1Dv!p?teL)40Wx=tqT~u^%7xpEj{b7svCjbIFDWfqInwj8NzC!&RGgBnuIz z4#gF;qVa8X#sy=q$mLl4mS+^&M)F~>Z4|{AJn>T49F>`w9G2QA#`!#ajwNSdL$xi7 zo}voMrqXh~3pA(z?Dxz=Umf3<8X{NMfnX*gq&^`xf0SRpE}b!s2kspyp?muti1?jM z6#Ppf@O!ZqI!y&ZbkYbu?*}^$Gdh*u-(v1l9I%SjA06lG+}LOVluUn>UmKV^gPM9H zQJKHqeo>R&Hyx@+OT&tUbt6{P0O7AHBz02(r9e{=i^#>cVHeC-?d$iQ$ zAkQSlvzSq!YC=ofmDbEVlb|JbCLZ@UQeDU4R*8oRlhuoQ_OpDx!zT(?B>f&$bp7Dp z6L1~>wlEF-S>-5^H>4VtZ8Ih@_=ca8!}zxNOlV-8War`7T0B|O)?JPN4_|K;7G?PE zjiQ1x5<~X@LpRdh-Q8V+N=SDP4Z_ehpdcXK(%s$NjWiMhQhV0g-#%IUzrLgQXb#@% zdgpoW`LBZlW(Z zB6t;k`JyBBB~K>NNbDB{YhMF>?3X9#@B|1#Gx2I3(<>MOI>d1JnWW@i+BR!-OSQ1D1tn>AV z@mWg?S(W)Ux(I`gpj0SmP3f#FNp?Zs(XE{_-#YnURU%_IMx1zTkvH;mqhmO0B~lb+ z?&m_&E~%SMA+Zb8Vs44_f#-1ds%6tnFETNte8#~Hi65T?2P-`fUCsXDW>zl*DSr8c zfCtw~W2Ob246Zv2-Law)R$*jI*1|-w)a_R?*p2Bj^qbkxzLBlh)Of}1lNn_^NBwa zyq)GxzdFuiAQYmUS9LJm%|w2d=+e%mg8%*WaB&AYq8pU(gnE}IK;Mh6tFP;cE>ACw zKRBc)UR+1$SKj}a5RJwPv8}wMHo%vO;agaLpHr@g6GC~Xh)fAzKkulvlak)% zdZWb^!Yv2=M6omJZ3;?3?Z(Zg>YFmFCTOCPff)J*0u z{s2AZ=QH)d*2PoLpK(cR8kSo%EI?R(G2XF+=Fzklta>+!G0M;Mq|@Rc*zWEG3yR|W z)|+5@GRsh~D3icoyI}}7B`@0v`$0JV8!A*gyqX=^pJa$^G|SdbAAV@+Yb-%^?oOpZE@T{!pP$OCYNtpL%MRYQ+TRiX{U>|i&37#f z#T$cPW5EM(wJ`MeNxn%U>}9{X`A@CJ4h)_H-bn9%$xlg!66sY0ZgX}vC(E*A4IUoG zPPicV*;x=VV{aM9Tt+Ny`4zZe4N&E|H`&6Ebt-s@Lur0tuun|Mrht3T2`hg3oGV(P z-PHGHF|En$TDZx|FuS6i+Yyzz&sjy$*+B5x?T@uHga=m3e7$T#3sy6|a^n89)t&pr zALa-4kZVOOfBjF#auC%;BewkZTX;%$Lb4hpECTXMA+Hb!*F{dJe~%67!k!I&+VU_k zK#sQxg1c#Ki%n5PR4XL3sE$#y%aXU(?aY~BN9HwOeAYC$K(x9|Z`oq>Zg~M+*!fEM zXQGo9vsXN0R{al;@}ednvHJ!TY;GI8UJm%uGy;^n!!gIV5=hBL&V`Jw^t@m2>4_>h zZvv+B9wj`Q?|yH;jjSbWm6P(xiC@9`?A}_(K6P9bj8#?&EPjCXug2W$-Q*FPe&Y<=u2JB3(W43+(AbZ*d2}m`aEL zB$4t^d2bVM=LqcrWNn17_!fK8jBLov*`PNyLXX)J=<=*$P6Ku!M>VQXet(6XJA(b0 zkT7dXd|=iV2d+uyY0!VgIkp{Pujs<}vp>+p$^Tt)DzCvAR#==a5df?Wh~g-jL{Er;eA)g;6F)lz>3yA=$LdwY z0rCXAWy2$e#~gz&ncsyxd*n91C-vrTJ>3k4|GWJIlFKB{*~Gmk%|_3haHMKV6I!=B z5T^?5QY`ao$LVAy)$?NWXS7~H;s1X^_^+`6ZjRjlm7u#@g-NVvk$C4@CY6&Rws~W> zG(X*d)h;QbZ!w1%!^_0Y9sqnH{23x($uA>J_?bMoYDu<;Qw=Yy(*YZFa0R5_SKcQ_ z7w?tbc{sz<9@JJ&$*%1i^7okTZ*B7Yi{fg+H1W51LMt^g)N;ZA??$EEgOKvWq2rCL z?d9P0sPRb@c_K6lIPlw!!vB5Pw)5?1^P!4{4MkhS5CNrj;x8SE<~^O)yOl8&#+mj` zy@|jHal6yjFDQ?e40@J-9+ACtsU%1qh6CP*xKy!rM@yC%&;f(gg48P_c)cdHuu-#6 zD~XWdw1D5&1f1!y_M83yf%J&3Upr3TUI8CpawQx|81%9eh`yY-jXeoV)8Q+o<=jk2 zQ0$#|WH16aw6q8*V~Ee$KLqOZ_z_7iWMF)Hw+FfMyuq?b3{9xTIxp$(EcGxT4?<*O z@AO2h9oLh{yxeHES34^Rq-$rmEZjrw{~RfMJU%3a;@RFlY&2T6Kb5GjSLaaEgNS~X zfb((qgN8p6@~=XVB4s5lLDu8b8H`?MGh#z&HY-+RJXEfr%XB69?RRojdccx@Xtc3` zZiCIs30IMDm@G{vzpBD%xM1r>%gUX8*s#(IDzo8LrX3&ws*Rd6G9+gNwDJy3`|!I9W_i_s7@Lr*S95rOjC>KF_NC~kwi2B^#+ z3+l5t3&?%-JriO0v`nBcYy$0k*5eW(%wnoe7s>S>7h44*)CZrN6qy%vj7nB`I-+2C z#t*(UfWCn@VEV3>GwzIz2LxNI*`thHpzk>$A&M&4{N+G)Pji9L z4J@+}Px@qNAtrrFBKuT}4r2vRG$KZ}U?~e^ghQ>ymSBKpskh z{fv>C42bT(vBAf>14$(AOc$b}kce&7cYboz@q&u!7=BQ0Y;)nGX zE{M*4TFnEzbEZ8EZlt{uO4I?`E2-sJR~SmZ2{K649;Zx(e#H_HnrcU@aX!cakfptL zf${Z_y88aDJk2;eJztSbl&5vUdV@zrv#qVce4i87ZkqxPX|?8`LJIAjU&}olzNqeG z6>%Gvm@UVODvH(dyY_ZY{AnnE0Kn?^UAVsXuqX`o87Tus1FcL~*+$bkur++NeX2my1p^TJ{ zQ<>XbwT%~FO!L73c=gJ26ZHS0AAHC&K3Tlc?h5;9Qd@1T^>-gzT^P}&GxRmzx-_w& zn*MUwOZ35>;trMr*~ z%pbnb8C45a#-@Cnbqvj1JASRs&mim#AAQN@xcC6XXU^T+aX66fecxGmzq%U{UZxhp z-y8TIhVQofOfG++sVY6I^gU5jg--YHr_B0so&Pb${-4?Q0ytV;meItsg*2|F!NWh_ zYDdnnCgcuex80Y=TBW>1;_5j`?;0dRkT&Vp*P!@Ee~Kgb|qls^z8+RN!@G|oqe#-V3$^H zCR^W6Oz0)P0Tg~UP985xCnuxqEwJNrXmR;lc0XkHggHj|Jo93|d>Fx2`b%Qbw9)v8 zKDdEX-+<#F=S92tQ351st^r_#_q#Fni2ue$q~IFl`Sft7A%B9;!S9rKY_X40)}nkZ zdeSPut8!C}!Y@`5+&e%EnTU*kDB}l`Z+qVITZ6cenYelO!Fal`lsx-o$MQK&W;Kx= zHH%YBiARTx7L#P2+@J(Rt7+#)TRZD_g_)yM@;0dXj&t^n8o6L}4(ELhnm@Bv`ctgu zaLkMGXyjb;2g$an{raXC+dy+ng*f(Io8(gWcTFP#^2&GU~+GGCLhp`LwRho2?t4&Y!G#x~!^TcVf+W5H4rQo{T zPTb)N(T4P|79ds2PYQ=WV2DxVWnDXqqwM!zf~sE5YU^mCS)6bxLO8$9I>I>MX+@Hpdd|T_4L3Os^Me;5i>EI1IG(P0 z_um&Ze25)7iUP6a*YP&ZcM-9&ngWB>nuK&KaZc!gUKraEkqb)`@d5_~F65~$w3hN2 zm|Vq0e;Sj+5hu)1S~66>nAV+qoF3>q>TyPmO01d8+%jI3muQZRb-5F2z~JOx|D7Dp zjkTs9!;PMm1hrzYCf1#OQN8jj8J6h&{=du!!lKB!8W(Yy$QeX6kn9JW-Kn z-#C$3rh44rM(q7*M?C=wJXT~ekY7Y}GX`Xo+(uUsdYnxj6bBBWlkqoA1lnr%I&_fD z`m-*xz%k*gL_Gt}y5B~D^$xNk41k2JF$2__MWABcfCZy`D)Rjuybs7|0FBm&E-2zW z4B3qmU#*kNLD~dh9~$?4IuR7yP90wtoK?Aje;2U4dQ-X}!O032!VGJu6zp~15DWK% zR>&p#C_uG=vGG8sH<)_*u?KdJhgcJ+o$_IswZD;su)8bk7ubgL{BH+f1A~})MdLBm z^JOkTSC&0)>Od(be&=j(za`f9m;<&TnlU@8Q%t^xLeCeUDYzmxv0-vA zm#ZC$hoj^kVJC6itV$zSsQYi11H`{?+1Nw|o}? zD>VbW4W{H&X$^6qgw4=q9P45T!)iKaolGDNF`Np9ECI|*K2l`@coskRh>n#AVrIBb z;>4-`3iFkrCEe{n0iDDnPjuG|#D0jVjs;)mBUdN*CG|xIR&DN}<{DqeUr|0+lL&Nc z56?#{a(OI*j-d{H6LICE?I&$-Z8T)=ZSaHbo^S1C&+8Y%fng;z&hIQFuLe9j{+)qW zewqG8RPGeLU1B`TM&}ud$w}p#_&UKFP{_K)G}34W)5o7Ik{+$v-57lxE2KvWc(|mq zI_3UWjAhU{v=4TxL=mQsmlj}%dv11Q zT`pR^H zO_Tv905B-DkfEQPdHw%`3I6+h-Q<8EiTUIn4MtcsHWuaVee*`=nxw=Fwvr3lHKzie z0oMEc>$vnBa~IX3g&*9E59 z+tA+O;Bn?QvV~VH9iZ$D9)M6qK-}Om5Wr5pZ9a9X#T#FzQCTM>K3P#~YMmc{wxjKQ zwi~V&K-pYbwGKhT`Xa~%o`cLkP`MnuY=7xsQialf%@yb}m&o3a()YC-rcmNU-HR)e zyexl%@{o&h)KU!zEC$EIA`tUaiolqB1ueoPykzUO)XnZ0{!OF|J0f^cix47Q-Si|o z)%womVp~V;UqTRRP!O6E^m9~T6S97m#fsQ_8GzFfJry4`2HYZHPx_ zQ_*gv`QC%I+5CF?`2o_!XIlG#*ROE!+ouK8eHq!!#_s6Lrwn;_d<9k*Cym_smei|o zY1U+=Mnal!kR~g^{o5geP9i3b@rTy;m!&c^0L&lxwQYA|1cXIc*$Gs|`(Wjuq&7g} zJ5P*D)}E2+!E^T6V~ek8d$X={aHwb*LuJ#WF+!G!S)6RWwvQj$BVxIzlUS8aVZeEF zHE_ETp#5I3Hd1Ga!86e)Xh^WP?sGkHXVL_pL80=i2zQ2Mr!hA$WnY|AMOqS+Cc@74 zA}5U|lN7xYIZ8+>Xc!z^u?%(eKA$mG8SyXH?_=&{SFapT0_z!^3uaTs_j8|J0~i|#^?OMcYdyowfdElO#`I<36>T^bUOR- zcVfPSJCC}Z12Gfr?vNV-K8STG{GR+D!0Z1(V>^4OVB2&awgfGFvz;YO`s_@yk!hSF_J4HreR7EMb>lRv(D4B^VJ^p1a^%iu~$5*ZIWs zkzOk1(uB`4eZh+R{-o&zuX^oGDer6^=Pv6=fs0c}5e?gR#^!@K&Y+a=1+5E==wP^y zd*FwaIC!XU@T#PEV+|u%gC_p|`(RRZ(~6sZ zOLfMQBIKtn*Lv=n6O`Lc|5_u;=u1zA^CB|&i0B|Luv!^;uc(&5MWw>jZ<#FTkrm%f zY3Wy-lgQ2xERK!*^=hY@(vCx1nt{NJtaRMk&>8+jt-7_c{9k0^|1M61(!W2E7Nx-%7u3JAEWDBCW{O`2s>I+cG!L zr0@e`eIM@f`N;jmgYza4N@GaZjY}Xx?m({v^0}meY@xbLGCslzr>~I*K1gMf!#rW` zviLlf!0Av_^atjnxT#23m#I=RfDGJPW29BG;2@d6q#h?a8{dx4Qyq_E<{h-9>*r z`qlI18@H~$e!Xj|{ya0TX>dP3rod5J0?%`_3Z$v|EXgqtL5qFQZE%o_{)CTkEOU2? zxc7@BLRm8~liZ2y2>R{7dM=U@=>cPD=bR`d_Zw zKa}htS2jQ$Se?yJXsg{olCRR=(ue5I9Z85|ol5x**&l76P*^0t`)jh5w&#i^rLZgX z_%43W3PhA_PNs9wGw%J}G#jw~7Uo&?@WyrTlJ$6T_ZNHqCYEaZq=1u#EWXPQN7%77 z|86GH$bLQ)z#V2aALa~wS-ZwUbAp4EVG%>z*z*AGAI=XP3F*kklwU2>?d%NX$2ES~ zcq|Ar{)3s=j^%cPq3qe8X%^Gb&e-2he_XxXO|!G?`AVJwmNrAIy4f1pa(R810L!)N zAVpeIUG1XIqFGYtop2cn7Sc%u%f zP6eixI$4uFvgI?VQMvv+OlSECiN}SZ`La%c;ts^;*AH9 zZVjbmFAff|Zt_@dRh~M&YxfMg13V=4Jf)arqt2rq{Z2?S=1Dsic^EYNIZl1+lO-1Q z2ISz(9$8_@Xk&$!?+;~v6pu3B#mJ24_lI&3xqp8QDhwr_x8NsGI~bi*9RkTpI2hj% zz}Z1}&SK=0m^4fGBdpF5dGCs!cig+zUWZS97C&}#Sb?#aJSZW|%vvNfVj-h3FgL)% z|La>e%5cQJt4;irnab|wZp~_=0Q=MQ1J=m=KDd`;bi?G5JIwEIcCbT~mz>#3*0xMN zCi;}*7&luqrXv4)*fJbbD*Dot|Co>upO#_1Di1`w#`HgF&HwjyjmHkg?tUL9-6#%P zGfj`cy-Ft=i~p19#vt>3T{d!XZ|*gaY4A0FZ+|kZ>$f(T%2PkE)nt~^Km()N2>uv) zZtQhWdlhE(+<;}sO&{9H8+(Rqb1SryIzq-ujBNYf7IOg}h`pxZ*E zNGD2TMvV;vI4U|m8HHz=_#Jr%M(>C#)K>q}OP&)?&S{%v3FE4B#TC85vZ~+z2`kqX z_)T_n@iU7EhqGkxuXUt~v6SoXfU<Y!DXO_d?5Wt#3)btUXd* zoqu3`G+Xu28fdc|mn3NcYi_5L4h#ZuTlsb~o$)VO1JxGiU9>V9OEsdZ1nG2I3PaEP|$n>Sn-UXgng_g$H zbt_zc44*f}50(oTCmmi$xv6ar6b#zR%IjDN=lxOZvsiy8HN9$GxE z%;vZ?^(ko74<|TnYyPk-RGDR)Bgrpii#~Jo6jZqMIb%;pZMOMyyyZ%0v-4?pO5QQj ztq2%-KFW*%pkPnuYf};Mv&c5eVjq!qBs=YXZ|PA!)<^5Qzo$lVe3Ic?U!9r|k4u)M zEHmCCMJ!(?j^$H<;Yw1dg%@`ASjXBEEvupc>i((b0vq=fTp?5H?B~zNg@-Eb5)l(5 zfu$oSdJM^W_m_U62mn(pkmklK289OOGa)-{CS&rGNW-)I;hM6-0CeW%oJ3bddlnI| zh_|99M>h|LF|8{|@CRzW!0#5R)>6ybhE_|*a#Ud-5#07Gj|%X$^qcHIG_P6i>ExsH zY5_}Kk2n&hh9PbE|EltS*G3#WrH0|-+ZPK_))R02K1OauzHXw`*4%otPwLjAKKffb zqW43jfXm_JiC}kX$pk;<;YT&?m)$fJdDdbbk+yS=0V$MX57!$9eU068^b0oBwFlsY zAgB;cfF7|6cve27hZ4${J(jzt**Em|U`ObFH4tPbw=-@%=_z5KO26$Q z3Gku9QC)2I9d)sk%`3mJ3)HpuzmIUh&a7qUW@5C8!NsgkD*o$d^yhO`uCS5opa8== zf8$=5eHK^Q02K>!1wy-l<-V=RFek+K5E-}mP0WMK#!IKp{b_K4c4C3VYm+@6oq6~l zXa4)+G49a@R>_!f_t4<*l1#qMb+VT;%Wsbtl$%Yf^8be6u@#T`j%o{mtZ;2Schmp4 zZ>WKTkxGDGEKkCr@4E?)a;Th|v5LAF=T17O2z$_Ee#B?5^9$PC&k2+5xz)wDx$=Zr zF>AQMKC1k3oe;G${kIUX?l=3FDRoMOpWWdopo}S(MDRC(mXiWW$!2f42j=0 zR}GKlpt-7)!A}tt0(LKkvVNmT=}pE+i9SAx0{c@A1{-eFGQV=dcyGc3EqzaT1uVjA zj`NFbIoEapZ6}O+zb>700mzK}I2VT*XzFSR3|8f>kP@FtbQdzGD3!ghusw=ww8?Nq zm~U4OL;}~abLLQT6El}5M8i(B64^t;C_pgrL`5c?y4u*=$e6fZu0O(cxg|zFdVhXC ziI+w{_$?WOvyiE-$l1#zC0ldW)&ePx=eYjhE{f^T((QKTf;MON%W6AOEbkKAhNs^3 zHta!@Y}DW2+j(G#*L`F|0V-yP)J`qNjL*DC?@5zAf_`@Vi~d%qzA8`O_*2d;%-^+K z$4iJb(+DTC)#d(f@h0_vVlN;H-0KsS_9WyGbdkFPne|QKIK7IfkBh`-)HtPGz}nB? zPfmsw1$wz)y$rdBZ&m!+0RE;o92M<`Ogt}B+`|SRVLM~z>wF2Va_JI2lMh|vm)4IO zE*I9Tfb6jU#Lsj}QDIS0=!ZVqe~#`P0KMYZi2;W=Id$pSc+B`NEgZW225p9_JT06s z($EfJyCnFZTnP4}$Plt+Eyo9x7YVFGeJ*YsxIAxFuC}lEuqrUO1dnf{uh8dYE%&NP z$htxrqPNNMup3l1aAcf@yR*QtZ*5eIWCtK6vU1whBJ*~c9E(8dp}#HzC)one?2|l? zye!{D#Qd%MJ?zQj77_iVmGxDeOPIsZPZqrNb_BFEfSr2Gu_$`*E(Ct(B%LH~9@o{q z88n_O17v2>F@NO($@}($2*2+sPG4J{EP|9Gx%p!dQA>IfCMBIt=~mC1e$IRgt{e_) zB$i`*8H>nUTrqPZ;DPAUJa_H~v3x+9h;)=DXw{yZF zCj$^)<04I6TlvZR0%CLBe_0-bO}zP*?_BBh6l1|tyR%tZD?g}0bL-ppP`H4FI@t_6 zM8-+IQ4cyl3}6fL3hK$8rU-#w=;JI}>9HEQxvPb`PxkY*;puxAs)EZWgo72S9CUt9-PA2&Am zC?4a#WReCRler#Htg)|bpft%t)FK%Lyd)h%Ri*wD_p-T#)p$MqbM&8@-_oj&$GYG3 zYM%GGR@&%V3>A2%$Gx%~KbSsToF`(_X%(9IfmsVQy@F6yDEsRoqRD?yC%#oX+0L2l zEIn8l>-+Xt;*E)5V`z;o-nP3Bz&`~Hu&~eg|I>ZL<#V|s#%`)4XUN;4Ss1|)mn6>U|bt$4p9 z4w5kC=J@u6+CNyyZ+MANM8zt*+SSe?!>Oh3IJ$}X#8ba2Q~w>-PK4H8W^k^4)Pg~2 z3MGZUz;wNVlUUKK9{zwhS21g{;W4xcVJi~gx7LJ%o9(4P>v0MJcWa3*Sbh?zu9Lb* zbm}vru`1y&p)T)wQHPl1b~TpfPhECbom$PtC6vfRh;~zDg?m7f6WiYI{xxt58Nx8d z=nrxn>(4j>k^OI;$vZElrYgYKH?Ibhg$peNe3c$5y4@ zp>wiON9oxb&yZ4lp!kJ#ZbT6&v&EF1%z8$7C2JENGvH^Np`*QhUFTG4ZxsgTH!H&Y z17CD!XsL_mO@RWxEMf2RA6-jFt{b>+4zT=@tmu?_A%4X+e=!;kfOp;{?SB7FcxBVx zMR#c@xbxv=Pg0_b$olc4-N&=%nXHP;wZzF%)`MNhew#Cw5v~c-A1}<0t>oYlTeASA zE~XnB-Hdxog_`Nf(Fz4$r%18h_P$o=$1{YUnoei(H~z#c(PbiuLdaH_;@`ka$ z=W9Gqp0_N|=%>rR{sZ<;^A1;2UyQk+M$b^TdT-7B^Vy`#;JDKNJa>X|UmpbS6Jop+ zGrp#vNA7o0EL2Ao_qBY_9BeB>ycXe#J#g)N69I{OB>cm8nL(@0ReT!Xxjk39fB5Z1 zCh0l5!^+iFqkJ~P@RQ>y+10(;`>fM1DOdb)>2HoWmWgiRyWi1hi2-l4FdGo2pLX5o zN~%HTOsPssRB>~en2~tr)B %ejM5y_e;{cXFj2sXK=P$6?66A?mIcValj~;L>Dj zsi%B&QlILB%YVk!E>p5szOOuAaN9PHw5LKuPT-D_FM^vP_*VN=W&78zl4*r zVvCIaiGUT$RexXSo)XkfuC(=Ll5#U#VYbn5>w|?e1I-#+qOPg~K}5RT8Gpdi5JMrf z7+(6!)5uEt(G2M+2EBWBV?=a}8#QV4lg!}#U~9?kmvwFz0;yjhp9}Y9k zUJdDlh#j4q+AM2sr~;EjCvEzouaUFy z_c%RTU_Yof)T%<(x&yB~)g`Z*o$u0Zl(F$!f}1@RA^~kY70l9r^P^L(PlvT$8#GN+ zGauG7stCF}o~QXTh`00D5KOsegfL>dmA;#UWYIu$&II=$0V9s>Voji*Zcn|@MdWp; zTE;Gi_sLvT;Kmp3p@ z*K2)@@rvJ9%2cwG!um*E^lC>K%yCT`wnXWNT!ZLN5yu z`;Q&B$?q~-SJcKINq2|FB;<7eGl2Quicdr%PLxE3l#_7yj!&aqwws$H01~l1>l-vM z_>?jqCrdI)^v}v<^~_EAX^(K!H|;&)g2Iit_>KA9K;c#~sg=|(N%tH6ulE44I9sA{ac5!Vlk=l24RoH(0F>;Kf`xBi?e8HRW zT*;j{z4+$KPb)z86|^FwKYN-Qi>4x(iXzgB8~gBy3EQUuuhbvQ2z1gUPD}{XD_MaC z)>#~mFH+*d-u*AVM4 zJ{72euRxvR9HBI+-Z>EmLKMB?UcBx_-xy?VuI?`wo~#I~ z4tn{>60SYxcoES-E5}hW^*M<_Z^1wkKflzZ783KiRd8qS62RPhGd`v)()U>0Q1Q)t z55kdTQJ_$=zw3<86^>sp#-D8=+v-ZkhhnjJCNV|WUn>Ck6PG4UaLzhZ{s!Z;X&<%6 zI*zrj$ZMbf+jHkc1p6;1MCC5jQNl&CapCsGyyv>|{$}jJ@sNIDbamue3IiwCGa)Ha z*6*JmZ+7l?;7UwQE^^l5p0}rh?r>|)iCl@Ni6>V>@5^TZ#j3TPXOP`FmGwu9-!N+j zfiecU;5||$^h=_v8a?qzb9ABRKSB*B&q z&pJtVPZ4wq&?`WlwWR`>X#to)b2;r7mm~XV9hW0E07XW-Anq{3+~<8(*ZRXAMoztz zB2y1Xio%zO;v`AKrhwS zUbPz`@jOXkb=^J)b9EW3gXiF8tYiQJIEa@Ay>0Z=K%hq6HbaS9?cslb!xAtQO;qTX z{3^`po$COn97w!KHexjGH(MFV zRB3+<0+3<=WVwEDft~*}JeiZwVCT;EsVM2B9R%kMo-T5OM1?+tcS&K>A^tt(C}ih+UGuq|w`6x?k+tU6LaYrq^u!3=tqlLK;j zulvv01d?XtDt-I^Ui81@epWhI#!Zu-GqY19?~0192+Glxz=5%#i|1?Tcv<&2DE(Ni z9>;+1TS)Ig2I+I_azp5}Kp}qE$Z@m)(xipD4Pp6Bp6`=NB=?Q#aTxW@Hpf5s=ra?% znjh_PDY0k;2GtvCD|wK{t>~*zl8DT6H8Bp+@mfCp6lG_gtA0wwbmvDV4!DKuwaY>i z4*lE!$X12yJ<;#AC3OY#*~1(|e>fRTuzMY)Gj( zbDNSmMk;?LaQ2>zh)w;gw%fG2KfsW%WFHJ)mETS8vO*P4L4;~*5Fsc=9FTF38ACv9FeA+=A7@_$1sEV^6 z*hT@r#)Ou9V?hq>F^r=zb7`juKq%#L|CyolP%*L1y%ywx<8%#xN&&O#_e% z@dIx8qrdgfx{d6WMgn-#2!FxWhC16h8|oSIq6GgJmXz0Vfwq6zV8&5B9E-BXQx zx!e2hU?rY?Y#l?ba6!!}qFsnJwM-V)`t>Z za{H^W(shM9%4NNFI-R1`wqVG>tgDD$0?-FpDdFs{oA?@+yER(<6?d5;0lgMp(3l;8 zpT>c$X<=BYPOovPhq|yr-O*_`HNILzKLF*H#(Ulrwux{xK8GXrqtC^qZ}HsHzBy#H zHMCx3TAS8Xh7dS9S4j1@?X7Yg7?SO7n-(&l&maqMwtbE~uMgJW`8D1ua8>Z0^lIK+ zh%ho=CFs$ZaDId5+$4|#E#oi}3CA!#2#g-AqGOUAQa~KE>v~OaE++`hr}UndI@7de z(?0wgk)xkZpgB7q!Fi|rVwT?VJj|yoL;~qPM6#4cY629@^5u3ezZ$Z9u`qDU1*=wm zU(%{)+@b@oYAW;vy|uY#V}5I_1*71S+e{dgsQcB+WI%6m zu4$2Ldapb511$USzhBnk*j`x3^2)Tl5$3YU2{e-MicyaBP*?16!SFP<{>fn1cm~fb z(w&`pj0rQ&srZ1g^e>Md9-jt1Wi7i6#D8fUg(|zzlW)x@wxTf;ZIKB@k6{8LIOlA9 zfwY3s>qAY=^ez2Pt#oYgWoq~WLVp${$!=(KH`x$V$Dzfb85}Bigc5!V)j4 z))sw!N8gSbAff1i&ujz$TTb=)png{iRE5pZyMx$Wa}_MRipTG)J#&8#Rtnpi97^B* z2#(^TcQxZ-6DoAEUcwY`_}b0y=l#0^*dyA2a}zr{^^blM%K^#ecq z;Lf&xDV3R?FB5^*B-cGq>(~|%*H3={!~&6dlo@YG*cDy!sF=NiI!=1_oZif%lL@@- zYX1sLo4!cp5t(IOM{2~3g3w!=jE`W8ve>q_AyG;2Z)#;gIcn{NWW83d?L9CJ#s!vF?TiJD=^^-b*w?z=0dki(+U__a z3%JM97MNBOWKrNQ9l$SqDBN^Jof54p_o}rb7g7DMT>pONi$7os&!0>*)o& z9^O=Hepq7pae2XesgcOw`>$Z&d*Z+>RFQwppx5nB>;8zg>6;i9!bmcvm+~#c=*w14 zGcKOc>R&XZ^Ko2kPdx}>CgE62y!0fSMC3vQ@Ab~0gsLD~LgvkIi%}7Z^W?{-1|-$* zY+AV6DVjuc9;F>saSVVw(|}ZIxG_cR*|aATziP|Ujk%|$!0&&(!n>~h0VR~sn0I7# z(0U`SK?&YtgV*bxoeu-mci*=H|2}gjyVVc4q}mX$dQY#Wh>V}o{EK&WBUAdnouX#Q z{puoLk40OpfvJ$!3f9>$H${m(g59Gx90v3O#2sG-d}_qaZyuYQu5rzH&MSTL z^7x~&u|AJ3Kkn#QaegzN!7*G?CckQMU5@1Y8;Q|Z5Kyao&#Z3QGq74dRB9W#<*HTr^8LRL9)sU)(%$dNYbAqNPPJVtz5iN#j z$b;soa`Sq4W?lQJTF@>WV-T30Mk z4|6{;#MtGY%)5;;k|^`aG%6~!^HVR>m{AY7f!&DPxqhJX zK|rSxeeXBAh|w=5j=pli)?BXr(WEV~f;|Q~@ngwa5afowcakZ~NI`dr zKHU5ZE4s`5!M^dYi<0qKG%Qb;en=uFOACJ+zGGlQ!SU~>v&wkPYK(n@7Ui=^7PP)~ z`qQvP{jtmI?IueT5@)A}`tSb$n9lw(LtRWPK4t$IJNj>XFpL^YoQLQa>R2=eD^XXM z>fY<@$sO{Lbpdyv->n9-Y-3m9wx={QL8U2=UcfVju zoHq43^i4Mq<#4W{6<}QcW5p;umx-BcUq;zbZGP(azRPs|FWC{d`Qcqi3Up^R%5U+W zYUSq=#$#>YPPTb{qGLw@SBlAguT2}4wJO+yOY;@^2PM%))^X@aH^uxYi{#?c{2LfF z)vf(8_V{hd4#%KGE{J7(vgxb)+si8UMCd~(=N`}3P|4T@?f4q@Jhx;yb5*L=N;t>C z&ZEzLd+1$b?w`Yn7AE}*h~OD$v}w++K9-$7%ve!i`VGlO_74ZOMUF@Mt2&H7rfV`4 z!k@GqJW#L)vM75W;{r9P~J=T$Z#k-qM){g6q9}G#8a?wm)3Rr`G zU#Zy;5I&P{zR0FiAo}r6HZZN3a3Uu)J9Z~w7BT=22aQ8dnm5;sOR1~!AP&58uD=75 zu$fq7d!7SGOCD2&B8jSWc<2o;dKyH_xwwdb{O5-BpOrxW zH72U^yrwB6p}Ho=w|+I)g{S)W`Ma&hK&(KT8I`K6`j=TI#jjczomQECr868iX=d!_ zChfu;mT#ojx(yAsxHS&b(O7b2_8{Gn(?uj`;_B=jnuIUge+KKc)THy(6KNl?_y&aA zwB}6EaXwp|+Am?F6%1$$Nz85v=SCxGItb)wsMLDtkVl>XS|TGwua?%I4XvH+3W zsXpc9#;>Cl2;=YNiL!P3^qY~i)!emu!n9Flu_vya-tGv@si$^e=Gh)R^Dj;_rt@>Rm*<%f;j3Ai#o_W+FONq%DN&gkc%iS52XJ}0C0~H zOdoBKW|xHg*sZ6le1HZhy?F2q@W*-=zU?mCx+|O?nY5*!jV;6kXe#*gJb2M=aJcnc zbK17D9qc-$-j3GIJAB33ynQrZ(rJ*cR_IY^L(mbuaZ7JcUQREJO=kLkyIpZ!30|$l ztVc3v`a3q!Y>0t|%4O4ucUpmZ+qd3@| zcDEth_Q2OZ8yC$^e)K7JC0m&vo#B}U_ac6?=9uM@e4ppXhX48K!d^Zwh*w40*v^?5 zMYcwqp0ae%WuIt+aFPtXHx$Xn`5{qBk?JutZ0UHX&v<*}?=nOb24U&R1NTHU%m^b0 z@wraBsNfC%6B?F#@kTxb&XazaZZ6w8L)x z`Re5@WnmKoCtBx0eddxE3+;xivDb*bz6OH^TF)NGF0&c+I_Dl_1+OecOu!fO2H|DE zHYrPITos-(?}=7@w}t3E@e~`Lmv%==gf@s`xmM_Q{s9ttpGF8lMyf#<)xAdXQ1GrW z^5K|NI4t|KfUPkG?&srWU#J+8-$`ms`bJzMCswgp*B5pW-J0-_e1P6?Nt9D3y3bbKbw`Mvi#4=LWYcS)l8 z&g*WLJ$A`72hR0)*@IvIU3+< zahrt;sJvfGrS>o3g+^4MaUriaIE>yLHGYTF9sU1uVA8)Q`X5(+*sv;;%SMSFX`mfh#f+9%z3mZ$OZTmmm{fx?|x~7XIiVULt zDikaJdECT~ri9U`es$+Pf9HP_H^sjs! zd`sjzBYuSzq8*L4O z*U9}A^j&dM0w1NI!&P}+o+L#TKf}oKm%kzhjb!un3-K54f@uu6t=qfc#Z~ujb4k#@ zlY@iQmqvOd!ij7_`6?{4Yt+|)2je*cEqkuN3p#!7x@Y!R_iVAr>l}|DB$cKis8iU> zjeHjAK~7_raodyOdb=0I9AR*1Fg??$cbR9gF3Ox$&;P^MTZOe5Ht)W;6Wjv?DemqL z#i0~;iWT<)#oe9aUWyfWcXucd+#xOQE-QQOz4poaf8Wu2T{%b&-Y4_S+%xmLqmxZD zO-;~l=soQjl<}3|Vm+JYVwLG2&t7D-wob~`_B9z==9oj?X*X)WR2+2((iEOaQp$>H>~hR{&4v>>#XI2o*?rU zXzf%oiR9!b5aS{*p>Wq;Y>QJ#+;b`C@mCC;{iEl*zq$C48@)MU!V^mz@#iGa&GQBT{r=NjN`T-6SF1%kkciEvKw#!&xUl6w_}x{ft*lu$-XJet z4K+rk@GxHq5Fq!l>jQGsr!cR20>EPg?T9>6?ZqZW!CR&rAUR`sO*UMi93YiRi_+j5 zVvJ*bjTsE2wV`>=$#Zf>8Cv)FyoPNdF16HbQl2#G?nhD z{i7y4N2J7cHp%ghmMLAc(i<{(=8zb2GM{Yn^`q2yCb?qbPW8*I!40#tx9fB=a9`Bj zOm@ZkGXwIY+2@-R!^@yQ{^qWEv*%5;0dWUp2dkuz9J;7?FIAC@BCPN=(RVwwZ1ndrZmZLDwv1`!h5@b^s|1IZ`U=Xm zR-wfIX%guFY{F1so1PVt_HYr2)x5PjeM2%pVO9JU?OBqntXQEsDIHs=HI7mYsm|z2 z_wOgvSMEw^GQ7m0jFa4rJ$f4+0qGv?k&~$=P|?v4m1Ey`iGlfYf1oFL$(6SFs0&;1 zU#NbuhzdR=6ok)DjDttOWyP^U*+F%kbp@CbRZ2y?x!`hkt`xqb1Tv9VParoT?47ZF zM>NIUu>m(19Hk)D+!G&UXqe%+0-A-chg>}9JLnQPOLe(3iNs~%C#_t2z9v%xEPD=6 z$9ECFNhK!UdYAojqfyXH3}ImIkJiqF1#e>;>}G>B3%H52-P0-}WtP>;UdtR6WjFg+ zwhDB*R(wMs(4(b+IJ9nq_lWe-sNufPGtnuL2OlH*$#6vE^;eDM$fF6ak1N0|G0Un` zYoSn)vvr-1QCp_I3i@_ z4;%v~lrXRiglZZhXa+&+~XK-IshS-Q0BQ z>H47t9^}x=s!)Ot(aE(|+0WF2?9bw(%T~vZ-sI1Gngn~`;$-tAh{6e*ufxzFb!EAe zC)XW9F}=Bz5c+=(Tlhg}aLxsaxB6ca?VP=4VtwH8f*IW9B7ESGyH^LOYQOt$h&+Eb z;Uu%x2bmSL;-J+;96x@o8tKgUm{A@E>67jmJ>>g1Ylq;EzKmbM(Ja_^gky)RTbHTG znx6eDoq95P<8|A+p2;Q?zbaA)*s-331Yowo=ck$(#X*5L2c`%S{C*qm;nw5S_B(i` zaO@J*W67d}=Em739KZ9Q>(2p(O7FSPR(r4K_?C=FXP)=HR`{}MjdoM#DL#0|yj4C$ z5rXJ;%3{dqb0+42Z?0q!IHP_4j(KH>r=W8La0wjuQ+V0Rg7R0m$ng#?8OSk)`+Jb9*i~z6ezGcxscX z?|QR%Cy!4i>2ezqVeUoM!Bmt#L5QY^(LORE?-@4zTLnnAmq6p53of-wA*K#G1D5q5 zaOcBiz^}ykuZGCpP_y%T_Cv(=6{Zv^n)R9vDZyo!a@QRawfXCps^d3)mFXxd*)5-q zh`B?-aKc}%wWQse&BF1o64(PLGhtX&cT>EKQQb$6gJRo)j3^A`bNbZyF6M^1pKu+Z z02f(u>2*aKoA0l=M!$*ouO+WyqMN?<;Wf1yS<#LKs7xwXl*k)^SdXL z)nUAfNdf}7!&zf1c$2Mj?v?l2XxyV1@S z1A6minMHY7aACVe>&UwVB zoN1ciOH#RwLcNun9Z3aJth{#Gh?ZwsWaM4*Aec&H*4t3#vDQrrx4B?>$r072Opn3Z zBHrh^kaJ=vxKECblQ-t%%A-MaIz5z^wFmQCKCLq5elhKr$Og?X8XMp4X~JV0Rg^<- z8~w|}0!}Mo0`L<3__yxCnmN6In&U2!tpXN$n}?;JnpnHL}76?_u4yuoIN)-RkM*dBM}wa?{Udp(2sV_ zhPUm6uGOX6hOM;ajgGx7qk6WPQwMswif{r%L8ULVOv#oZc+EXCYcHWwnx#Mgqbk=+ zj}OZ5c-w~k9vpfSgqby=UWg^_1 zRK$6wO|bA;QNf{fci%L#N$u$YbmVfNofcXo)8+Bd)y)s(5ND;}YsRWqPuKe<-r|38|jO#Ovv8G6^-)y@{6E40iSX+R4FjdW&3SV51r05 z>tYa?X$(~L9`j-EOhr!l)dx7 z-)l98LM8TYVvsp$cnnlUBbsKh1(8p4!p4_wIUOe2Ck1tR3B0KAz)dvTaMbr_|}@ECXd%Pwadu$V{6lj&Dp*KBkBN|F{V7 zzd!Vq!l~WI(zKX{qlH74^Pz(V55oOzfl$)Y!`zYTZbwRtpv81C^X@y7ceTSuleZ@1 z<83MfM0IRgH*{Hc9ANhC97=RS&_4v~05Uf`KEH>#qM)-JU4*jUn5;KFOXx@y2JFkr zMgJaU*XB$9hcCFF(ZdFS(~yWy0Nm8o-6FlmE1g!mxVfnN&2A@0lV@|E9O^C8fVJPB z$6^$2T{kNmQ6;?OPE|jJ)2RcbhQiUygF?~VevvZ|g4T(aW(Yi#gqx}QeV0%WVng5) zzLQsQ0|8^8THk z?6!*>&U+KgM2g3`wR?o)Ubw7r$zu_@yQ(RV#zJyK|rzg$T{Z@MqDGm9x1Uu@!Y+@NDOz8E1$4ZlPqpX^#g>iuQTAC--+NA*f{V^`y&LC z!D5Ztf*G6}@T8E+1ujq*qvYwu6FJ{YR;O71_?_O{w=nhx0h+t| zkDT{SF%aYE8x%mH&N&}HeL2h$Mj7}AF?8aaY?~mmHi4@sQ7s?!m-wcWT|VRNtFP7!bnw}Ts({&T+9J!%RjigaSbj>N@hVm(MSqnkN5RL-3d*rTocVG z+#1Ye(?!kj0)h61K8v|>x?lH24KIsbT!Mu)6I*B>qwo7>?^TkloQUjDVo|}Ll}kSG zGa-Lw&?=W=$2evQZJ$lnTNaNrbUUqYLM@7_RB-HU;0@(VgVt##(0A4}Sz9z6ImCpY zrswr3U@G(bEZF>&D)k#!Mt%ZJliycDKNPjJUa=iimHYu#W>>xe% zxb}|Cv_g}i38-JuM5LhWhcwCDXbmTNld{H_%{=PVxeFq@&?w1}QJqKUgrIJcEFN2t z@2V!V;nNbXjLtQiu)ErF|41xKrrmwMqUKcZz|~IKFZnF)DkduV6>L-+X3PX7)X$ zU0tUaCQeMcP|5Adjc-MzPqK0+rs{m7JKKK_VGMA9ApGbYoT#34DF!KOCBPTvTz8B` zumHI#qbF;7Zk)ureRkWA2H7y|?l%+!L;^|HP$E_gv;i0EZh?i|lmrQwA5IH5GfY21 zVDl`)y=VeYk*F`&TgzvBL;CbS}!@}H8e)p}00yrpvjp0yHa*5r@xaj)rn ze87Zp90}m}o9@Tg0DlAR%r-SChD0h>*jn0gMNF69ALi(1E}OT%5y=5Fe6Dky1$owG6Or_NZs6-=t9lVY=YZ=H0^+tB-PMED)L#g?GPZ{u4>}6|{ z7yN!`Q1Wg};>I7oan%hm3u^0r+C9X}FUg@Ht*s1F@0u>NnxetzD2if*h_V(lFyjhe zBC)LQHm6r=qcV@D7$)RD%~`r#R=g934BncZi;qb<1?lPAW$?MCDw@mhVjr0^5fJs9 z4e(9~(YtJI6{ZLX=cfFZU;zLu;|nL99$V|*?-Mv+?7lv*tVw#RaylqJD`^10TyK%$!6%( zCx>5`xjn19tcHVsMim%UxStI@KZK9-R@%F$jR+2W_??0g?Q*I?VgsF#mZFt!n>p<=tQwk)alqEm zB(2|AnaFFU+VQxC@41uTUzZioc>7s*x@$Y&yK|h!&{QaElAGt5_-f?i2M(w3(2=gc zFnQg}J`^pKbBEBipa^G2h_X%>&Q#QeE<49f*B6nv0AbQaroVoQ5BF6p@)g-#l@b$n zv|+2U_!epn`_QdLB^%YGWnc6u=Qp!nJ|(IB+cde`X-~k$`pX)4@=(0T=G*&8H_7I- zvr0#vyZbAV!3|?RuL8j8Xc%oOw9WuDfhO(qF`3Bn2l=$mp0*w*({GIZ;0ZPPb%6w1 z-A~}WiMQY4$Ks*e4ObtxS`Daw8*R)JrnGy4XQJ-S{wZ%X3cm^VH^xl~N7g^GvuArv z1!$@t9V1X^o~@$h6sjux$BtHzngEXw zS@r!u+K7cIW$Z!IQck9@Pf&+#DMAqTzUlgv4C(bq(l0Yr z0Vey(A{5K@tM@N7{jf#Ty89r%StUnSwoGgPljD+e{O_RiWP>$2Z-_yLQgBX5UIB@; zA2iphw_nY`Ib6`Yv1MBl^y#*I+coL>I$d*0sFJ~6w}m|@=`wt-UM_&UZyGqXh2mKi zT&DZf3kgaM>2+rSz9AJG!5sYYD|G$I<$#GMVNF|BDTMMkTp462uKlDEmxXVwTlNGWZM0vHGhxg)C99*IOR-tyj=Mlod71Hx5ZuXhX26eLP+?l8Cy@nXy@8@s5V7FTZ1b<;2 z%yuqE1?|^~0Ox_juc-%k&#ml1hSkK(pI@|W>w#{pd&!Jo-IR7*#q?0jLAR1Qc_}cqxF3^#o*YBBe5Tnx1*UW*yxIFi zS(A607k%J{B}fn7G{O=S`)u0zSH`{AJ`IyA;NDT;N1okON~T|q z!t<DjQrx%Zh_IMb@yDeW9B(jDc(~Y*w_BC%(3Ykig zzacSR#AdAT@Xb$;tNxIL#A1zG`)ghHyWy`+9FwE zg^QWj9FkPWy!`5V*9x~^ zqA5=}hTpq~n&yc5qLFu^}V^4?1)QEPnVtX`?SUAr|Hw5b{o4VZR3J)Xkpo2IWe^Y3mkG9y|toYLSy?l0!xdb$P3yfE!I+ zIj+jDO+_Md`FnZ!AtDEN2MyX+_=bXRjXk#yJ z)Zl?f6gKm!SMn@qJskgYdM84SKph@;;#Df_HpAJx8F(ZKfeBBicU(-mA7jR#oO5qn z8;OKbNX8EhY0z)9K7>x>KdXLp?T80jTMniDu31Zol0Z@8m6lJ#m~n-x_mug9ErUEp z84lNICdlWJpf3Oh=g81@5~s%JpNL=>6Er0qK;TkVSUJ06X5rz8z|!*3VgY^!-_S`e)m+`EgJa1M^UWslzE7jP@6o z+I3|GEyrugY4(slykey4i2Z_Gu~ngd_toY4*nM(qb?Uy^^bg3FVp%4*GK3`!ZdeGk zOeq&`g!%^sI_O-nhfkIP%DcyOSG#lJD7jCY#}xG0PS8AX(PO52cgshbL)Ix^25fF{ z*#3NX?&u08jT&vr>9mAJ9Xfy641b({Z%NY1xFZ4&&o+eN{I1Vn5gxjYJkFt2*?>>d zYG_>%D}b3?VIA>t!B5VcmzQrsWSk;ff+vwx-YVZGvk=}?Bh$v;kdYet_q5-%z|Nzx z7w!ALExH$Jsq@KE(4-JJ`3Gy26zY@)h09VwSjHr$%h~&{#;Z5Pj_oHk07V-4=Q3tx zXRV7xnoMjl{k{>j$JhBr95epoH6+yser{M_5>+m{LI2uh4hmY0r_I&&IjZ7)k`pGF zPibc~A^!nr2^D*xvRjbcZ*w6fyP-woi}-|sqZOsDtl;286sq%Gc&T@S+`rT15~dg* z9jSAMc(G}(koAI_#5IgP)fx|RtIRAL+Pf4->o{7o?!Porbl71FFK`1dFb zbW`-BeWfkGd`(w%*Vat1`SIG~K_Y&Qh!!VmrthPamP`@n4F9OV{N^asZLowejeBU{ zYk|LOttNZ!6zI87)3SDWm0k*6A`95&QzkoqSa2!Hz+Mb+t76SVVhnwW5DZGG^K4#z zqk^jaf*EM8DP{`gQuGH>U&5Id;p4PM9d`m~ZE}+^Dt1I=ht^(-vO!1K@TI`ce`8j^ zf%vrKP%n-g@%kpR790>vV^#FC~W{rUXN{@)Ly5 zo6{zVT5Am@9Gzm&!^WCMAvw!Lx^N$p^XFOS=*Ac%`fkVVI(n$`<~ng)Oc`|aBRR4? zLhnv}{l_X_)&49m@D791Lf`7BrSDqH5zh880t#eRzL$=ha*;+c! zd^|;lHR-vC7_Th}!xcL)z~9jvuBvnUV`dru^M22TiyTO1x{o;Cyp$MKV}Fknpv->j z-8p=56&eRKF6nXao3jM>BSc|CnEH5Bw1wlYa9^IH~O@{9idmX3n~NNt_H6^s%2oi>!dJ;F=h^{GuqA6ZgFX z$|8D$5!~jYMyYYBXa7^=av^#yiv@G(p*&ALX;=kNG|aSY z^;$juBUg?cp>}F91_P5)wF-vNCm1dquq=M9p;9=0Ibnn9hswWzD>kmdqm2ZSN4Z!X zM8VO$1}#_Y9g0k+i`C;d&-jY=(gIR?R5H(^Wk~No-V=k1*WB_m`DgsT z_r0_QJVPguXo}Pv*fB9S8w_>#)~l{s5;tWLy&doG?Cj#pN6;Xm|GCd7CPP$ ze`?dAFJCy;9R>&Om|9Ku?u!hzmFYm>TW#B4=_#V&clVlTIJlakI6^VZhxD4kl{ zOU;gT7XORA3Y41~qy~b+i40sdMBh2Oiis$-M_4o!zucLZr;fw1ic{5YO2C3+?1!O~ z!qg85u7qDqYQU!*FWVWD)~R0m90)_9RbKzoH*&uK3SCADN6@>&*;Qz&}*)s@Da`rMO4|kx9uPHO!=;MtW5@sOBRB^gaLw!jOYj1NT1cQIG zPK=}$@cOCBb#5rLP8~`YnIiB7h4y=a4t3AWu263A$_B&luPzErK~fB3{b8h1Cf|3d z;ga~U`V%cE2yCYX9H)(zsgostI&4jcGZ2KPYXO1>U4}icJ87m1UI3fke{!X|`)1cI zM`z(DT}OlM(fVMT?J`}}`7lX*R#hg4ye^A?%aM- zGl!%7IRqjE?Z&=3I$%iD#-50$HB7@@{4k@BQsJ%q%Nj~xxjf?;@K-#%ENFWz0SKf* z_-9b#XfU)}b!>FqC4SM9e=B4D0`=cP zi@=uU@@K4HL$f$lC;8XT2#g8VBm|u%uh(Kw^=xJ9eJSm6TSZN(f%RNVfj9CEik1GH z82v>xR1l>0^M_8aO($G#!+#c2DBTdK59_l+1{R+UcM^ipXl_+^@Qw2{1*pUi6MbFT zu%=&9z1Ll1c1G>FGW8tyi1*Fb@;wjQMITWT|B?vjiv#CoBcTne4qIs7b0|3+CyMj0HykL|u0*USfI+?4Dn`aDXJo85 zw%JkohfMl+t0pUQraJU9=O@C#k2^u^)11CmvB2Lv@jIy@s#6wfz>A1NPy}@;$yn!V zCd=0LtH>6pqn4Ipl=&)!%NOJr@lSy-_=j7hJn$3mUliJ8ZcW8oxBj6DBuR|qUKls{ z^0~C~jk`jO%o&&aar6@xlLl`rrF~OdQZqeCLXh!SAV5OQd-Bmtft)e{WMw%gK7*8l z9t%DkWi_bXqFhHZZl0(bv5R))rQysrxw1I=p^M|loN6H zXQBG|tUfN#TVYgM!#ZAp^?goB0xj(*d z{V2nxrKgpo1)0=ri+H|01_$&|EFPM?KpxJy-_Ewl0-RGi{G65hNhaZSpgWEUtJjvN zjfVy4Dqu}mEeQg3#Q*UTuMXSaNBpl3_tKa}`H7D7F+GHFb9~;f=rV&5CPJ43LM~g0 zM@X}n7lY6K{73M-)^*Lp8w0;r$*LsGLy}0B^@zHPp*e*kaBZ2?N_SXru+HJK64`l+ z+!BM7*X?*xf6h_}4zzw}9<#d)br2wYF0H~g#=P`9pq6Etb+;pT=2+{0b@^$wH2oc% zX2+ZkyKUO}*RIHA^P`|Il{|wjK($Po?m6igH81;3C4ZG`JAutrfWb(ca32S-yQ#Ue zJ)O!^vc#)f!JqNV-yLFY;!d+ld^`E~E@9*dQ|xsKVt8GyRN_D$>66f5aa3~RMQu?% z0u%lS4h@@qX!&Uzvwd{qoJemr1(iDjRX+sg@xe7WE%8(Ak(6LZyh+g5wUGR+{o`IkP@bNnT+4zhK@i-4SgxAg}0 zob{3Zs1o59Xp(y z^B?R6b`0A&fvM1rsUsM1@G0qEut^F!O*6p7r}*dABd}X z>UAQS6JkFCxjZ$wIFl42+IpR~`q%yEbBBvkF=*Ijg1+J0sD#kNE$7ukn8N<*+9m2r zS(xqTl1l|#MIciFTG{{oxIkd+B}%Cch5+%#&}cs>lu<*Bar1E(Z$s5YZI^OCbyZdK?xm zun!n$qlpfuc1!(iKjn$vo)7I{#OzF8!B`y$x~Tl3WP}0Gk->bcDgDDt^GC#u_;!ZN`u;_LF4oA>9~cJ&dnKckZKfnO*5`Uy+AXeBZXF1E@$vDQ9o z(5!wB;rhj3IzCK3N55m5r{P|^#>wCP{E=>IBh}rnbb|Hz<+w(L^qURYj;RTZm=M!( zeCHu6ysyH7d4BXC$6;Pxm=LL?__T6;8g&vWLRZQDs@LE@v(pM}*TXOO3hjFpjBL-& zYR2!ny)0iPFFs=)u^j{IH@8R5y>)2KCb`+4DHU7y+H$KUh?HfY&RRq@EqZ3Ug|Phm z$iQj;gffj&&8@I;H`=@H{Lcg0PO>bNPMSGX?0?k8e$g#RctxmUTmBs?zrHsgJ%2ks zlQSc`9T+t&nbe$euUN88JLyjWHtDr?AEvUeets8YcuHp$&yb_6eMhcU#kFY z52n$NR0<{$Z4hyO(P3wpKNNV>UTsPbUHV-B7Qma=7Wc$`QEyQKj% z^?VX%?ycb|MPYAzwa;sTg1N@z@oRD7!;>!TR37`FB)OocwDe*`iKEkA8|-idn5vTvb5^?Mu6aquK2~nN4A5?|;RcJv642~JoVdZ4 z>m{9+1P%fE?$`HC&aH=6GCgiXV}>(!SaLh0@onIq^9P`X;%xhD_$e!fj5IQlrBAdT zq632EsiKN_DLUl14YJy7{4>|&G4P?QuBjUfDAN<<3}w6gRDi)+`{kDC{O7e4c|==< zah&R+J^RN~pwA*ud3NWNi~=wZ+uBN+K3`tHMFput;RiTYKyFco&O~*}H-481|23>x zxVd^Za=UgtbUlE0z?OPDrIBxcoSD~}#PevO3t1FJzKrx_0{RZ}OQ_ZSPnXpHzPWlO z@rmb17`8#=lu2IPX>b6#Ij@T5keez8E1e(|_Ur7JKw8dCO|I5aog}$M}3hAJ3iRr)e`2(=P(^mq5x9zrx~avG;wl$mxnuQ~^mQa#N0xu5Z0{WmB5cBkqb z|2ZqKzmTmNteDn}ftwm^oDrS36Nta9Q8;SFvx$X2blc7BkZ!Emq2L56F4%P}S!ynb zgz(`S;*)D?YXE0^cF4aG{ykun{V@15C6-avXf|#ihKdS&#uliKD7BngRI~|7AXn0g z_RqZIWIe)cM~BAm#i~AAGJiUZx7C_lUUvU}X_=+b=k#}o$8}dgLp-7L zFEfl^*wh!K4an*GSJH%?ih9TiKJ}Uz87*yeMPxqenu|zqYH-f>%3q&ej<203B|9v! zHUvXedM~fqSC1!_8|o({cA`XtXf2CP$Z zX_g^JB*sc@OVsycxIbS2qgZvH4UQ(8kx_$}!JAhzc%Pg7-wiykrS;rq;g$U9S#w9< z$YNHTer4l*w0H7`@lSM{ks@=sw>A?HLi*1cvu2P2S}~NFUl0-C?kjIXX|WLi-(vpq5bp_Q_DUEJP|GG^m)@=8Z0g z68c|Y6eB}+zHIZyYIiS}jqsOWDfGh0^9paX_GHX;D5}o=r7C)lM#+B7dAAenjBqxW z9j4#(qRaAE%6_4R6`NQcAE7T8p2)}D@^d3#DCP}D{ljpKzf|XCYSEk7?WeEo|9J&Z(DTM6l(|@+nmmX#B@fki2iX(+tYxrE{ zu#gPTW8C$3ICrH;Qu$ad}8iRKT=TXTpm)YRPP)?hXD)> z-=vp*6s0&l-(gJLMYLktlqR$uo;WXN+|(u%c0bT%2;`qCFPr%YG0g!$MF%f>hd2_CAk<*Yk%4Ug^2?|7`&2({Bz^w zHXqYl_Bv6wBhZ&DD*4a!H+3!C5p)F--LMg?CItn`HwzFA#5_L0{>iBnj_kX6Nj#aQ zifzpNKP-Ss^vlY3$3wYnZTAfmft@r&tJfd}v4>@fk)8~ur=wlPOYfc^Fz2 zkJMptT*5+71ourmz0bQd5WPO^)^$~tzh)FM1otwV3%Bv7o4TtET#Ec87F@b_VdhB` zhaJ1X=qti^E&+B3q1|CQ(f!+f)lz}atnm9^!{_yxnz8CTLx{&L?^DeCP8e>%8Wrf} zbi(SRhRDreR)MTK(ccJJ)SP-X)T-tFjcQs?-*Nn^lbK&_EKkE|&o)#WRTxM0zD17H zVk!T^)(nc>KFsd5KDX_uQ9!^+R)p~iL?TvS^gr=9vwbu+FRftdr4(Lzx z5PPh7JQ6cVprBYQ^Wp}^;@EQA06!z5%*$kbJO(Nwkr?(G*g{z9RcSA!Vfy>NAOGC} zX<{rfXLbb=232$!i;sVa`<^H88L50owNm~4 zGi2f!RXg8b;bn~;3;Wo8UG!b<&QJinO_qOO%`xYG#FQJSwqw}ya14pzbJyH#-{gLu z;pdNbz{dpkKzpvS$*~XgS>-C0x~ES`bjeL6;qL=aHo8`k-8HhG^)$kv9o*$Fa@Pc>ua(Hs}$c> zjE$l9@jr<2uA22oqH`rO>cSxU{<0VBZmxIL97ide*G0Wfy4J`S6o@I!krnTWtJnz9 z;|FALD4?w;0A98IfP#D}bJj#og!#4m@Z*Sa?V+!!%$uG{?uQWh2C+ z0HFi@5+-G<(_YJYT_Ac5_gkV~fBwNMAWZJGyAFdKE3Ht0f;?o6;fWAo3uC9Zp0N#- zp+rzvaNPnm%B?ZYQkS_(P93I;6uYjqS#&Il?wF!L3IZnthL^~ESEM;s@(ry!h+GAh zO{@CYx7KS@%-I$!k>mL>zj)e>f)Jr7$<)^`mdHQ2az>-=MC@8r>cO z=iBkkRWCSXIq5{RIx0?3JQTi;o-Gl~UqwOPXC_^OC@bEv;qoy0=niRzxfiH*2528W zoeH(kJ&~pja{|l`JkFX_6JJ&hD5x8n|K9gBLBER_ce~)BI()hQ5X5`L1YB3L+9>^R zXQBpeu#A2=A}Tsp^~cV>*LUQSDsQrHY=h=YqK02b=%x8PK~K%N(@?8 z$m{{UW+0pJCwMjlud!{=^cc~V^H;N=^zA_SGg*O6jX(czcQ(Z%Q3Oib1l@e*0@4wP zb=0SJ_z(7>I?FmCC<4s6|J#G5E)o_QJdFlfOZBh6bMAx<>@Zzj{H|UtD6jF?mb?Ai z;s8yU}u270E33b72gR8tG(i!VwYsLhUNm91*Q8ursucG zs`CIcex5((p-N>AhI|1$xbzgzXqqMuQS8@* zk4?qdFBos%P)y@r8Lx2@vznIdiNh|Vn~%oT;QVDRU0ms2W&Sd1c6xI|$#bzxH*Z~y z8b}P@0s30Lp9@kb1>Vim6HTqQl|1gNrT^J$=LA|P5Ummg>=ZxVr2Fb?NOe;FtX?-F z?B45i(@Y?D#fxOjU#7Zi8*RLB%kXL465?~~Ke8wtXow)KcOL>ct08Ati$9G@bC+{I zmkwOCfnOSi&k^;G=Y)`lMBL4p(q@_Mqt0<+f0wc9^7Nx2ood{>zWB||bCwo(f7mD& z#_zD-ki5^#Iz1Z=X%J2QeFB`^1|`%4j~IRGS#Xg82b4GeT=NS2a6b2keI|W=qaN@t z;{03m&y|uRAWJEM%2fOkkmY8c;J+oOG@60`wvVpm{RpmZ4F%!sV=go})0(=+Rc`)p zYJ}$p+6vEycdu#xJPI*s}XlMk#)?63Qk=%Vj-qSvP z$?wN~mx=zTV`x|1?jj6Ek5>rlb)HBP3_}zeQYj4i!cxUB-9`f=tL-!!qDjGj2&4rX zR>M|cPM#4I6HJ*SV6W-fW*;cF%eMr~<3?og9mUegH~piQxm(YS#n zheby9x5EseT!G&Kh0Ol>H8k0*{Fr?TGlYwqEAp4(ry~~&_7LxcX?9m{f!)NnX;Jyt z8N;t>XHjFH?#2G0h>(Gs7;sawA>QI0>cr!9} zuYt_I)y70;efabpE-N8=#^}ew<6j{jXd8T0f`3zIfBm<#Gl>aXj<>kLu26d^)ba8Q zb1-~C7MG)x(uO;bAW(^=O2FTXgwwWz7pzJ?!aUC@4cniEttv$1ju+;PheLRA;0|Yp zkt=T#NCW}Y)^X%ofeKBNk+ShnRkiI}>NLmlQgnNPHGH_=b6<3#~|uZqSJpAf|ynWKQ1ls6|#1ef$3a%>WKgm_}Dg z0RbH%M(jd`oYKlhhXNlo9l;Cq-YBJrXy-7cX=z z0Tgm=wS;3W!H}rAvxPAGoKRD>9-+4l9nr%ezu_hv^>OF?1M0M#ko7!d@OL&(y5YH;Yqu1b+cxm4Co)#wY11`}`1(@#2}#v-Pg$UH7`>iE&YZ z#9HcVAXiBBUFrBJ!se4U(cU-0%WrEXg6&i+3M z!2kIY6*BSp_-T`HIJMrxz4;T|i6bnc&aNAbLhN2z_f3oEord4)3KCB_SsG#U?v2XC zN0QGy^RpQf(ZMtryFMZ$eWc>Y=(yIa7>FD^&pQ$eUR9=U`Xz`zeA3K^Rf2asyBmF@ zuddKJ ztU)bw?&m#jSpua5FK5P`h*!OgVG|!aAa)9Pj{4BU9cn4e`J>tI%Om`gnrEX}M_FZ@ zO5C7EsKiO}#}3t|0f*dlt*{d1{q2rGQLyp*wsT?nkes=id;Ui9s#7Y4Zno zNqbezf9X1hy1z^mIG~I0ThKe_&$GBqCA`jO0+s+6KGh|41gLV+1yam<&il4F>5;Vt zQbR~QGAJr{z>4I2H*?w75mH19F;9$t;(WoiQ!pZkpNJ^D3hMcA@(bL|2`6Y#A6pj3 zNlhcSFnhqDGK6YmqaFgChzo@_Fu`ASn?QTa(Ym_Bt|#CaC{Q4ComIA%7$IBzLxH-6 z*EFjlR+@a$tIkegBoPfu(PAQE>z?ezBq035omz8)>KZNMxU2|jyKR~#c`Q{w9yC`m z`M$KM+a!Yj^*7P32O2?YzF6GmtKP4k>T@5Nm2%wU(qJKFld>KoG2ij*6c|3ZCzFA@ zrzUQD!1WcYD>Y{$BdNBg-frfjG4wOKeZ($SR=yPCZxJ~YBX&JS36E?l$S16xw_;ze z=ep#?56C=SPRX_RfPr*JnI#8(-ZzU54?Zj|5rVxCAU=|7bv;W6$CX)UZ5+)31~r#6 z)Qaiq0@s%mQgVy2)MQ0;SPz5MGwg@K;;u(ZNak{j^Mq^`5G`62LG!gE6U{^px`2Gd z2KuefxA)BlJ}+t}^Wyy-$x{-^{Z$pHMKeQSt*HwP*-z^}i1ctJ?g;m=Oy&OgzggS= zt|g8!xjLzX`FHWz^mSo%t*;^0A0)HzE7fPa=@2=My_m1;$mx~vgklQN(LN|oo-#+H zIs-E`1Ku8oqtF@elm}^OY^R6~pk$_yh=qj4Rd}rBH+hgjV(gr#r-6$S4r^*}{OEE< zyfzW{fdi=@;4^3}lF=WLX^9c$1OXRsxZVW5jDy6jeD^-6BwV{rzE(=$f_8a<`LI2yB{d90;W21I>dZlfB`2)=kZOxHOvkZo4 zmwJWwl^wBYrbbv`NMGo|=mL9y^;WyJ3;B-(_UtaW zNTLrSI%x`g_@3PLX4^mWL)(Pr3<-mrYfReP*Ypp`;jb>5TeyM0ZB5PafNLpJI^8H` z>_^5bsmzurur%9kP{6yeDj$8(lmLt=dEki^2?*wT_XdH?p9ePQP=#s( zhQb(?f8<+(*p1NIbKabEyNRNq7ed0K-iN-5JscfejZtFV4Tf0c+gW9T`aUX) z>Qlo-_nvjPoIPRmFghA&&sWTB)(m9wCsw@zdXb=GLY;Jc4y*&SXG=@Hl|191tjqg4 zg0Z{qHD1m8f!rSI`W9uw^W8sJU^`5kh+vKGz!3b}&LD!7(TeB-O03c!)bOrJM|41Q zj>JM&7l5i`YH8)AncyZ_H`>ITA}aQE&!8!#`KQ!+h8 zFH@i-P@Vo5D!>?IlT^H=wq&yMXQpHrXSemSPYy=TYds@^vrs#c zgij?@Z-{7Sl-pGnm)en(2BE(Mc2r?7|JCdUBqRK&8Xs)LI=g z2Uf@0usZDk9sHl(DDjJPkCx4~hf`#S%vbn%+_gB{p@TMNt)#Sl;AK*^S;Di5jV8X7 zlHFR{Ki>U|%;(}=r=0A({`0S$Qo<3yt^Q)rPo`KHda)Yyq@l1=2E8*nvt$Y%^`gHC zydC`s!d!kD@tqr@#H6>ymPf$d^=KayuI7tof`v=G`^HF>`)0x{NQ{!CMk?z!WZt{1 z#r4hse}|TSYcdo%0i<70zkKhrX;qa;@&W=OksA$C@Ddmn>lcib&%adlc$v~k5y8KU z5G$J0cajB^_b$#+oj2Q@XUYgf2Hbqb@`idFYDv%lek0@-1hW|`dOob0V~Gm1*@)Am zEZ3Fv)rFgTty!J-#%O*JxT!+(_D5Ao!_n-UYP)l4%ywxj4m zzH@x~GSYR45RaxnCvMpK{cwHUVFJ*N z^qpwG&!KkV1wXUrmyqDHCaH%O3`=Kb%4UL?f@GPsHxs~0WPg*MzjN#Jr z1+{sAs6d!xz2Qnl2!2(7QO01jMF%VLz#*Hfh^pX%JQD_@4|F1xEE2~d5OY)b$8JV@ zhrp#|CZStZ@Dr*&SYv%x7vwK}fqz_3T+>|pgC4M!R88%R|0!;hzvbqP*Ceu*Vtm*NN${-{a0+~U#>)!UmxUWvcG%ih4w9O$zPyEEf;kF)! z=O1HqEWY+;7|b4zQE-vQ{03d?XSu96&YU78U$b;^p;;fhFqH+Gp(#$BpWwn*?+5n) zUIm?&H=V+|o|;2_Cl@Ia=m^(Si4^CB)8)mNl?54{C!}W9Z>1aCo0WN)@6vmU58-+VP+9V zeqihBo)9)b-!*GjAPBx-BtOYL2ta2rAQ+_Wr;)u|@Vv1p$C$BT#{z*;yjMZBea`p} zS&eaoqBnja??Z%C&uCq)|5kElEVpeEto6exutn?ZpN-e7kv&bmAYC#K3tZm3Uo{O; zl9v$4mXJ9yn6|N8;Mlbl!eIwYj~9!wp9up$oLt$oIqa@F#UC2;3fVqaEuznaIoJZl z(B8C3{Uk2tp5tJlcq6Nsl!SPzTcUyfT=FNaU~e@WfCVTt?U!QqEk1L>dO+`xO?F{m zl^MBRqU3$OxHnFsZaC2$ z_F(IS7%|RC*~>Ap{sYvFMC!ke64oiK-Fz}$ms|b|SjW?0CUOW2BsZ;Gv#u&rz6{wp zd1m6w01qYaXvIZTS@TM~GVGc^8BFdz>DKtDL-*dU_URb;ut&U@M-*d)uIgxVTnkT7lB1m>dz2gM@OFYG%E zmH_A_U?c*O6iA%-31$hD$$zwpLI7HzwtX3kArhRl=&u*UFyI)p=$(fYxT7$mowdtq z+XO$#QBpgZ9v~=JES12XNJr-EX*x(8H^HdHHex*|3T&fEAL7n0oga*Yv@m;1jml%+;JFO$g=dZa-}fk*S8;h-b~WB10qAo(&v4@jYGfhvDv?szNJ1vuGB5@ zDX?^HM-R5tA1=G1WRSiBtQ@QZou=3r1Z~Acje3a+3aa49M^uC=VG#2pt8Kt|s6njH zRxsd)c=z?Zf8wbkhfhFRA4c|ANJS?d^Ba^9VW=3Eo~c;2%x9a`A|guj zr0(wk-s;9L?+IyUra4l9gFs1vY~(1TW@Dcuf^JTPq`HSSZd2mTzTRAk-NU#n0*T1k?^0j!de<5i8uR+gUMk%4bI1TC0dr zJ^Z39Sy4uXWuofe3Z6iLi=ZRiTkcx(2!rJQ8s2gLY%F2IXol4tYO8iO5kNqK;Nd}c zIqeZQ-7RNaAsGK^i+_9|f+oj^8rQh@^*zZum%fX3K?9LYQ$kt(@H`*%<$az%M4@-&tY zvA+@0#kr`;u4|~U&{L{ED{S_%rCaNdVbpM_(*%9zR~jGW(_6JJp-j7%urW|L2A2{S ze$f-1g2He5YIGz(SNDrq3XX9%`A!gHB&T>c(aw}`vGt7(0Q!Oq(NrEuvQ!DE1sWvG43sk<5LX#mo>IdxY4-had=C2SD9+i%lZIp zyfs~QU0G25pW3_()S_{qHqVVM7OR2*-sSU!1m=BSW_`9K(=5SXz;U}lI|w0JAz;WB zPIS7$h^0^V&=lbpR&jV}hpfB^dDKl09owe)=LMW5R^fA!DWgK^2IFO>wN2;m7WTtE z9bUYsEt`!gny4Zy2|_r07w8^%`#dmtv~+}evl81>NVK}XEIyL=0uPsEi_0yj+n95Y z0{+oF>U-neQ^_vmYF;su*XYSQ(H$-iqpbmpWZ8_kCc!x-LvfC}C4}VG#XHb;j|l(R z+yNuPEt2%G>#4T6K%T`)tw{Ooxw)WV9h^JgjS%gV+{;%#$L-+YLb2Ar>rOtP$M)w?{EM#Ompvo*|g= zurpy;f!M}Wi%GI_K0EE^4rGSxU^N%O5*Y@ zFk-j6S^TSZBa60cv)@Eg>4HedS_cy@Tx3QnFx|~~^7KHAIaN2|VCPMp%3bJ=875L) zev<2j$6H+#Kle}=8dUD#DwRPVaMfWtT)*l9-zhT7;>`j#zyj;$sMdR8GI`pG$22(0 zIB<0EVm_^tl&g4)y8G_LN7e&(s;`#goV-fr*J!ue-0TzrPQS-oBOehzo9?8I5neV| zFKV?UP6{-5Q0V<~g!)`f@QS<&F=Fv|{iRiZWEtyr{NE_;2U z6k3HBGmIox;Iw>nD!@lAYexShzViBf<{KSiRy9}W{Iq?sko<>fFibQDec47Wb?87Z z3|1PiXC@^HvNFTy?9weK*8FLQxj^+(RFhxszN6)FEy)1A*skr{z0%VlAqU9B%u52a z3ww6I0lD_ddD{<~2EH}dez@MUdZ`UpqhzEu@g$pXzp;?C&vuxjNwA#(gtR{%#AvrT* zy=TQjt`)A}@`a+1 z# z)f&GVPHk5zr%U}WkPalZ7dyZn} zFTWfO&?&^~ZQXGs9L~O*NZ*1JWuGPk2P^J(=T6EYt)&i1fCdat#)`<1^vIvLUl-;) z_TuK79vTT&JF_HZTCs0S?;?Ugok507Wg}k z_28)@E2gv#XO$xR&`ZV(2r7B%AM6(Yc`7C-aP#P7%_lImE}ljlToGK7h^;>jg-uvZ z9Q^DYdTy=*LCSR%AnwAkORb&VJog77>)ezl@iV}EX7MImZ>X;0EO+POx=`7Zd;%p))j>nZn}`NF z8jeXEQd1c?HDZaDzutDRgqj;J%(Vk9bE{h7yTAZyYT>%#Vu}8TL*t{=bC8!N%v+Rn zL^7HQN{Gfk-Infx38@$I4B!mHQwtXg$3AfZ*b~8@EInVfV}MUQ%`yAdGP5u2e=$LH zu&Sz{t_>H4*GXZt;Ni^u3^X?-+yC{50}#0XK3n>M`&+%kBCOq?a>Gs%n)b)`Q}B4OpkGQ4=^2$jt`@XGqMsqdWJB}~Q~ zT$g?n0A0G<>1Z|)jLaPydE9Nm}{E%#C=}P z_Yd6xkY(Q+6(5KSQO6cCt@bAv8#6CYD#}-`W z$bXi_BN!=hS7{>Byyjk*^^*aG9g4-gT+T}lT{%9bpa3T*6y5hW69dyjQJ*|ra+t-G zjU51Ut7yVKb_mQlWPqS(BJObTlH>Fsb7jnAA3Luq1p7b)VthnmK8Cf5y53ZvvYYzg zYCRYc5qAuiX}2kV9AQc&WoGK>ev^L^^@>|~4;72RE#Pj2x(A1+^NH;8K{`FQHIq{a zfa}~OAB@S7LJo=Jy=sBrp+Eq*ki-orAeMv@N!-axxnY+p~Ore%hGEVxrpCZz?{ihnE$e5Vlq9?Lw=itxbkWmM2SRnP&; zhq@Z?3Fm4%1vI55NmrYCq@yblxvtld$tTT2M#J46q3y2|0{uXUIFFewl#d2_V8Ef& zhB9dB74%Fv_x;ATLl-9g${2rXY~ zwrGnv)&bG@t{dad$PwK;WR}V{H{o*PZ5d8!BqL`OK~n$(p4Ys zZlZfD+HY&%$%nB&N{+1mxyI|?Ya6gnwP1}d_lVO0{Zp7H)i1{QHr;iKaWb{m?B66F zx7vY_%IRRJX8TyzZDBF+!{<*Sbf{AT(n^to;Yi_)1g`33SUEpwHLI_~VQJs=K$Dm* z?dbyWJ-1Knh_w%aA$c21M2`h}&trC_tp!%?h;I1g^0`Vo>OY|8J12m{)t}mvpuAA* zk{Q+IUGB}B)t@_gsLsTldLClt8vkO%JKOS};`zN7{qd>tld@>a6vN-qr1KH!a+Fvr zH<~;Lw%Ym1#8WshxY1&_$Sgr#`mZ=rO(Q?K05zHSaXzu&K8D1Fyyw`W)<{sw2_MlE zh53zg4^ObDgGhMMBohd5>Yn0npqGTejOb^UsA_5rUp|hld~zKU(-Ra0a*c|^SK|{q zs|0)@pT;y&3nDC2nUm^131@pXRTFc1M%aNu{ne{10J%n5IPN%p9kO81(>?`K#5-x}47M(*zd?Is5Y0l$I!%*&!t=f>p?bS7!mSZbol?~9oxK(J z^R$<)r)2UD_?@#Xfjk8SSO}Y?KD?@Tu3+^KAI;J^=qHFYqLk&U7pi@{Z{LgRv+v{u~>Uk7RwE|vTv<&uNH*n+ObjSc#x;(YQt97r}jk*A{|kJ|FZq@sJZWPJV` z7a~E5x)O-}gi^o(INrAP6qltH_p)-xDZ7ldqC0?zKYFt>Kyp&k>1GHYRLb7OLgL(U zF;bca4LB34xe9$pp8GJ9&n)?(fC3A$vr!JNXYVx7|FiTZ5cXYV<#H*|^6$(l+2S1k zyYM*FLK_U?CNLF;GehR+s zMg0@r)gm2QE@Rmqb^%^x#3an;ily%B9Pa`r<53V_i+LV$J9gvZZK*)2UC_MfjP` z%JRe~B&Bs#+c}~S2^X^CY^P;An0ETN7jR7;K*wQ$*r3WBq;3=-6>F{EF5+c3?G?S`op ztFg@`2TY5}to&R1vTz-e`Fz-k7pT&FPFSoTKc zwJ2TM^cGRW;Btsohl|IdFJbuTkM}KQVkt$G!`cxZYP_DDCq6fm*M(Su-d!Nr@0*00~YKz`mB!+=I#ts|~|Hck~BwJ+Bw|H(uZE2^F+ol%&r$EU^*|wy_ zNYw3Zl3&oV?kaO8Y?uU6xv>Sc&1kJX=TX2nS)MOzLog!@XCuCWQNEi|KFPdnFSSUz+Pf z;;MP1F4WJ4p>D_Be5`L#K5>#{jz!0@JQ;tx$lp7oMZ4R z0)ymT%iDj}&VV^&!&L~w&4Z~A4yVX7pQx_u`FUS6!z=ty7T8WwF4;l7VTqC#cXH8S z8`2FGziBpt`xYz7y|JMgY7%-D_B~4%#vsrWPf7@k%mn|Y@ zB1o#=L2_f;04~m)o#IXR+(hohAfw25Y`zJcOqjpF>vLAj@uv~Bd9!R5$sgY9d6zC| z6e=v!e>IK7Jr-M@r!b0nd|pv=b5s)QGnxai4G9VHr{E-h;2LfcI`LX&53!8AdzI|? z!XGzINB<=?%DU!ZVC!?7(BUPG(=r!gr_gYCbCY5jej@!IPQ~rT>01uK5mQBiHe6AK zJ`>erM)pjhE_gg^bk8T2YNAO0jVUUM<8^&*b~Z9GoAL`gt4Ii$O&wgX{zvR{Q-5=r zUq&q%am!YYlvI1Zwx%~>%jaG8hV@?tkKeHx4lH;mfzqmV=f?lsvy0*AzRFaY4rw=f zNL45XZ{>;isAKB!yOVEM(l;~mC>9OzgbScap2L_;Al@XGpZ8)MxK<{yB+nz)S^Yc1 zbG+%GranW)bNOy)<)PxxM<1vD05B8}fm zLt3WBOn71<5r&&GlPGbrrvx?!qthbSqI?6(5ll#lTCrisf|)uK7Kx%K)1O8=FS}PE z0`x)(CbN~e-%vXq7P#m~UK?&1Kj!c0LLOww++kX*`=N-Or%oU}6& znJtCCwDgzxo8PFroy)5qJ%r<9QjMb*(DumtbjdcaS2eHad!YDG{duvZTLTZH*%kz-OvJ0M-$ zR7kfVMYou~gLQc=j(9okj%1X5VU=4o4Ut(hPLot*Oz{Kb4iyF<^CCSm?2m{wU)lC_ zGtLK9E^*@U2@>X`i}Q3g4tg9lfWFa!X*hJWhcH-qS_+k)NMGTXAciB>&L%sk(*%v9 z;LyKhi0Phntk>Iw9NJ~pvG*ucL}I!7jnYQ3?|Vb!MCOaPR$qb>Ltv6(k64_I>Qw?3 z`YGvnfyC-sub8*kM6hQK7Q*Ul-JhQCtdpP#W^9AcJgkX>mYjy1D(*@YMHgknvW(C{wjZY!#8yHMSKL#9vlSRLU+T!s3srnMs|cCqivb)a z-KCpW>*ygzL}mk=aJyY^e|k2Ii8a<5ihN&=F;MjDbQAIMKKg`jF68Lwsm9OFnYh{%AxofqM#2)cDuw*e=Y7>zVQ;2vXU zN9xy&mUmw~t1&OVq9*~58Xub#IR+r}#}lY*G`~uI!Ikz#r)6rsv;*TZ@J!aim4Nl@Itf$;ya*>kJel}3E@>l53*;SC?H&i03M zoYTU^TK&aQI$=qW3#IV824t=4P zBkv^Onbra^KRU5N$D9b$7yn7MFHas#xj)#mfvHJwrA0Hf*lXZBdKKD zP=>=gD<-Hvyn06+>-GD2HNu$PN^|%3iaw5|g_KhFZwJxGF56$4`ozn;y4p39w3NP_ zn)4DK_<891wfUb#KWOjFsVERel>_N~?Rb-@M3|uUBEW>aa9d5JhQN|MOC0qYz~O?_ zr5%raPlE&$h4^S>Y@hB_NkEWyT%Qf8lwQFQbxL0jNrU;Q1wYw6U20u3fvHk^uNh;H z1Y4{K&QwG*sEg9G3_o2(NNVGjuk^1nA5NG|zHEJ#0W0mgNPxkV6w^FhQuEXOWF}~- zgy^?^T^IGiU1y1&xEUCR!4g!%1A)!@;WH>VyHOugjG&|{+Y9B@-wCAVg$9SPOrLHr z=oK#8wZ)b_<>Nz6EGfd=69aO*;czP0=aKo?TCJW%V2}ITQAmTE6;qPsRS#R=h&_14 z*Z*8mxPT*ep?nIB<`z^3CIg<<9$-C6x;idoqI3u6SnCL8#fu@Xv&IXo zv+i^ioAX-LGYQv8VyExjUEyqXSxWyX^$5DvLwH;ooGoSFVb$o~>MYD%Y;)F-v^%M< z1kB9CP;W{}E38uT%7clJ@o3a-f;OK|J&YORI1K1utSbUAEvua!##33Qv7M{aO!Ftv zsiu@n9QgaX(T$e%;e*0uNQ zkzJvR08Oet(nBkL?80ufCfh7Yx65xgct^=#{hLLXuC#7LM{NigrSVya1hw}K-b z^+|rB>4F@(+nas9z+PJLy?w6zf;->DZ>HAto}S!r2+k&FMqA3=zvrPj8qw{*&hQ?Q zWMwrP5aS>Q`o*J{t{JR*>!?RV-(p7?*G^?s9xg*9YC=rr>%5ih7p+a#J*ZB+dbk{Z1L++)?Jy5HK~0+#gW?ax)CX=? z0YYUQ3&(gKU9^-lVPQHcMwynCqhv&D9{~4Dq;uc;rf%d185;lyItT{iGu+tg6>J{d za+LSqI13N$*lIQn78ruJB?_}lVkHnOx8b5E@ZPqW`V;OXT*fg^TKaGka3eW=$j>LM z*D!j{M+1X_#=Ibl?;v8!yAKVuC&r#4H7>5x;KVvjWRm?$JZ$eN>kYq0@ejoR3?z!t zi8ayMA%vxE95Q$-U)210F93?UCa8T+rxpiEPN8X8o4%WMrapw2=m46TBMpMcUHC7L zzPftWK3uo;XSU-v8>qC;X!^&*2!ncP;B6`4t=%6nrETb`?;y$BrS?yA&y2{$0Twie z`)TJ~N2y|~AkcWI*HjN!E{=frhN5Z*5V!w(IIo9vUm^1%!$1mMR4GTt1=EIk!UTQb zV6n+G2!_(uG-yW!IoC5^xFf6W!tlZztW27vCL;3G(Dh2z!>Rno>h!+=Lb&D^iyavr zQ_!;tS8faOIz+Mt+*{9C7K2r+)ZS2Is`o4^B#Xbtt{ymm8!NIKa2}iY)nzza;?#V# zxo^~+W%?dt(<`P#ZH6gpn;B5L>e-1B&YM6PJDT8m3GsQ_(74Mck?M759gcTG%iD73 zgV4vE532}FqymG}ZhoAlm>m}uDg0s zP-hQEe)y%36+_fN!vT$QbY!+mdiK31JcY@3ubgsqNXO~yN4BZoN31a}J*=22bW(d< z;f1q8B^%}UftSmvI9Oa3Reh$cIy%yJC>KdBgztMkT4~~Waby}ysm=@6T&{>|k0|EU zIHpuYX@+^oy|AOk()n@Ywy-U|hlfXQF$(0*T6|j%y~miO_Q{P4z`UK=Vjwnt0s3_b zCC&NJ^|~)0#M0$^o&7{S75$NAsQdIyG1Dx3#}!{xbg|J=4#1|)^7=Eho;n{&qU3^- zj5&QVx(I=JvkWNm(S)WHB!z9JxyoQH>@Jm5_ zE$W&ZUyt%hgjTA8_y>^*P|Q=)#e4Qru~d^YXOi@rv}`VUGNBNd;l5N%E|LcH3z%Q) z^jgWx=krY(=&ZXsyfYPAIWMNjJ}5E&f3g4qWsP<~8>3|6XiYK=2unk$LA_2sa9d!z_m$W{R`Y@I#$+% z%0nbOxCooGuvCe67xN_2e7JNY(hEFOuL%G+J!-@uuPv6`sI%)1V%K;HaQ0LgQgb&b zDxJJMCj?LV&p;Lw-3R*RVl=asi@2IKBz-61xJH>OQwQAKP4r0(A%8!sH{u=-8XBvV zRdNZ{9XP6PN;}g)FCRl93@P~{G)H&4%SOWq zK9C3hKIaW4v`%*-8rw=Gq=3dl&9+p>7av1ovI_c50b$HIRxN$_76U+Vr4+rC--wxev_2G-7{ z@C#!g=0`CJ_C;n%;Y%~kp|HAA)4T)RPVNI{-k&hh#xQ1l)$fO0R1=_&AfC}*-ID(G zIX`pWm51uDY3m!ss+B_d(Cbc1~TU$!?{+<0tklSsUF+j!sM2E0JI zLh~tmi4$q|ZEMhRy!p_!c>?dgVwM*P9U=KgI4w2P8j7_z#t%GV`n$H_I7!YAe?&xv zJFe8TU^JEvBezHFGckb|(Y&7%+MW9dxhaA0(c#hDe-jp=e(NMJ1eo%$AGV2@eOg!W z{7eDE!;e$Q4L4X!ltu-_ZQO&VrAG1580nO^`c-nK5P@~WxHvBFRuR{Pz%XW7)BrfH z3UF*d8uh*+l*LSIX@RVIi|AI7bE$IPBovF_Mugpj&YU9mbVP^t z#*6)#nv;bGT$lRQVvrSbqaM)XwBgLRfPy}!La&&)GmJtQC!vOo3Ezb&*)m}#36r^v zv#YxA#py)~-LcqhR@^EcNa)ZaiLh;sPhyMe{U#-F=W)N8Xy$4G2ei9+xl9Yw<9VY5 z8nCSn~1+eBmNh4n677T>c}N1Xwn4vAR#H|I@>-=`2a8O~o&Rmr z`)?wr7c)MBwYrPIiQ0lsZs0eN(EQ66j5svv58sZtugR%G%;nL#UVKn~Nu5w8;T1>y z$V}-Mh(8Uo((L`jCBP0REAKkMl6|-EF$tB9Rg&Hg@J7Q`aw;NLq+wP-W1J6S{x z2(d8KqV8F)Z}K}CSyG(5?lGWxT9ee1_P)=rRNut#Oo13=S$lva2+HsB+N(FUTqd+6 z+H0p`SS63OoKT+3^DXNoFiAIeBMu(!<3{$9K*!lifO)-eHS+WKqn;gIOyMYxufoC_>woT1Goz8P zCQjU_t6{Uwp~Rb+EEP+`D1NgaF=by2KNIA=eqym~QCE^B@4w}~)F^=o*+{*9fw=Cg z;fBq3{QaR&!@QXO@v*d&3zU&4k-PriBK*@w5W9BJ+g2kWE{>L^Bc_oS=qk-!mE}5L z9A_e|V(^2*e6+w!OejoZx$C9VDWSQ~NAozR6t>firlIH*!qxXJesfW$FM>P*N}P0; z?^MSaykftgJ&~7P{OL2=jBR$jz7AGKQ8pF6`XG(s770C(5_0=^%4lE<*=sgC?%VIO z+}&9?8H-)l+7+TWJ!ziwWT#%DpCex|@!b4=4xs764bnx$A8uqOCDn_McM3PMFO0ew zlHkH!tn3u!-R2k5*5op#yW|(uM|DbcIGH%wa5q984I~(nD0=N+i5xsI9rS}w&;QEl zZG4#y{c?^<2r|GBw?-;96v>Io{5++bFywi0>y~UDwFN)Vy0#5?{CR!fA>r?W-Kt5- zIqu2^7yEo*jbt95LcNf9Eb<)UmLO(&pb@7_I*_3Q+$?fGd?m8pluh^)e>=g%vFXoe zBAR|1YU+RxSB7(X0>tNd-AveBS7&JvMFB{zfaR@Xi}t@@qyH9ZRu8Rt;M-3q-{NbN z8UfHo`Vl6laCRdAQ^km}oizVfF-AS3NVvP;C)&`R%&d1+a+gbkvgXxzSM)XRxX zH9E(^2*&JIP)0JZn!nhh%2hzo5#RBUxF1v21l+xO{3@>L=Hos;5r95f&4F-!Y{^Jt z5uJ37#^CUg>^$5WE@5j;k5_V|aVAu=b%3?l^5$5YOlG?UCgky-7n-004T+k_9fZCD zQ>d+wFCuZax@;0A#BjJluNa1v!O2^TkRC^dY7fSsG`I2-+N++9fL`BqVzwfma~b(g z4!+K>fw0z}?QQjrA->vl00?KjQ(pVVuTJoid@}?4g{26#WsJ#;aNIFHJ$- z$;)az5ZH)p5Lu0)95JlhG7nLNPU}Qy%Ut)1OS%7BTFw743~EvU=tzDrISst7!O+LR z$^A3_;7Ze}Q{X{t{oy6#8=y(D=Q{EIGq)23vEnyqNhv&T9W*(@86hw?V3v6H6ou+- zOJTsy`>|f!n3ARIupbRZY)&MR%-=ern&oR++{=pBlMa|g(LuglMHF=lXuLh=F%Iv; z0N$Pp>}mu~vCoTDT73(*W12BbgOKQ$w$HW_mk5<;zfhDdQpH`b@-@#EPJy0Lzj^YK zbe4okR8fV~WiC|EBR9~Z)-`=8wig$1e32N*X2AYl4sq8eArIY~SDZuA-loge?SkAm zg~c9=Q&qU$pd!w!DS@Zo=)9jBdwNgukGzghQ{~!S&2{Ku?XvJ0m1;%h8AATdBTuo=DW0+qfMhe@KG+NUMHAy%VD}T_AfB)m2u=H={9e+>1-Zaq*9EgL%D#yE z&wd%IWZUc< z0=Aw-#foHh4p#Lh2)2H0`uuNd%4^S$X+snc2y))XrWRZp4De#;R+w!%)4K4+evHAI zEHD&HhjY;78Xh(d<%jgtYMaFc^2%i@P?rJlbl+xg!ZL8UimGdyogK`UeF*u27T{ALEBaJjif~&v*l}gMAH&3D&iC7)cn5NM znyrwsSM=#gRJCm(r3s0h0(a+PmgMiAn2^f@wq2r+JNCd+?7C2vl&wRNj_rUU$ShQH z=wj`wUrNyw^q*a||Lv6={8_``uK{ZGhI-RH0MrQIo|RQrc8=NI#TTRI^F!N+H#yXf z=n@7C!C1)5ZGy7EV3;hJM8KF3X2EFgqvAIX?(rGkzcr7f7qxg^!`gSg$1MLKXty!@ zPKk={Y}!uSQ1L3n#vph5Omg~}o=y6^`qu5)AZ~MM1_1O->iw~se{&pWx7c>y2pd`k z>>L#-x)`?Z#nD8b2_c8wj&{~#L~e3gu3J#)q{P@i7}xV`#DrT^j{xFI{nzsrLtwWy zZkvjmPA^ZtFh|*auvU7V43M07$7vGyPLDM<8co64%1)texo&cgb3tnmdy5fyBAA@z zcr^Vdlv;OSH0?Hp?sB*xdH6q>e>0QIGpP1hjDSW1a5ABB-xDBlYH3YS2-HYj{P1LT z(cOQrYetlErLES%|6^I-563i0k(@nF3Qn`J7bRk|7tpThQB{(FMVgR&C{%t+qw6&bHpVGR>~rssCn;{NDixrxcy(N3Uc-1UP8; zue~G~&Cujk5TRPfCO8LNCRyV>s814?jew=n$&bpxF5; z0|`!J|Lc>>8sMoj0=c-!Hye`c`&y6_ENFD~q(^2h^z-0pZ&6{Af@yh9$isbp>aU6RSd2R`8-e^v+ z69i2+L6*Mn-48o%4wy80J4Z!)D#PcV%3(BxJy2;rgZn{a>E)_!Pg6Z`HIM}}*-l|j zBg)|_)lA&lf73K+ANJC*zBNw&=%r(}#Q5Jx}8Eir8~(ZoKsr%x7E-v3{8t zjh{C8KV~8x$A{ami&nZp-O3ARn@t&Y)QF+tET$j zK5wDYZm=kh!?+l3V9(n;eN?Ojxa5_LMmTgX^sVlj@|M$B@4y7zHBIpPgATXOQs2(=UL}RS?gC2I7jLiwb0ACB&6T|<@I0*_ z_pC%j@T2Tk5oR{uED|@5X+R#?hi6On4p%4Q zpQ=q+;j9x8WVqDM${)r)*xa9r7fzk63pQn=X!(cTgCMmcxvZe45 zBFPn$A9a?W$UNq}VFWv7=cd^+aG>@F)T{jo5dv|d#5thyPPB1^XlCi(cB23Dn%em&!QMdAliJvQ+G?(PRn+={xLa|s$aC3O3wg4cU?ycyY>%~z^Ln;QSKvC z6vW_ss*yDa3c6n^Z@O5{rjBXBMACKUQiy2A68$g2-ZCu8uH4_8_j%uc&+YFyj`?$5d!PGUYwfiJ?niTo z<(!KKG?xw%>N@b$397};SCbu$Nq>|@f!xgE036}ym)6cDhLdCq7`5V;HB3S2TnIK; zgi|3yYXM$;Pl1J;VQQsx-sDG-K+kunPEk@JDm7G`aIWE_GKS@6LEBhW+alDa!twMu z+fSf?nFA0(E^Uo$@Puz{#!icf7LM!Wl5LPwmW4*~E-`OrA}iZ1v@1}q z7^{}__}a_ib`ju+8e-VB>oGV`+t_ilTlsfQYXd!V55+0 zrCIqUCuA@RS>Y;>O@xApy(2qT|QsAeNH z39ohmG1n{GA+WLn~b}Y*V?7M(B%4k7yd!q_q;O3ZQ#n&fx z;3CuY@Z5h?VlXZC*MN{8^%RPV%J;~^7eV-3Pfm_LI}ViSCP|s8bhi#|o}=1Z)pQO$ zYo&{aNC=j2FhYlTG}zwV$GjQ~JRYnjOR_97Sd&GaZM?B`NjC?Gah3~K-&$YC+Ke?i z>)evF9>0}bKs##l%#k=I-ogHFqKW_c?In(1B=+{lKBP#vnq`Cqjxjb6#p&B@bezV| z@`!M}lX5zJi9+=tX#5Wsc4yGH*UH$rW(PLU1{QR+p9#x|i#*0o+S3He5--$!0E^=o zk7@VoAHO}%)ppx`v47ig;Dws%LJkrN&3Yy!KQ(bpuG8_m183QwZ1J3qclUe~yFv`b zhUCYgoqs(ZhgK@!ht$_79Hr_(Y2K-Q3*vrZDD{d)@AEL?Z?`)U$aYkw_6Q1}Nnr9f ziY^=6dOUk`Qo!Z$=8GkCOO$E>8z_f$1szPMz3ou1+S}9Pk#Vto?**QdG)*fy1C9fN zKgMU+SHLEeS#5GysBA%Ah*sK0OwK)((f_21)Diy^Jc5N``0mMQx@s%wuO^&JFv*!8 z$jY4;$XksMRX6A)H=#EuxFbSCDuxZ`VDZ7V=9@o?iaAC+mo%cpFp{Qxlo! zWJ)W3lKgaRvxMJZ{uIb5?gzB9ivwG#vuDO!E28J;Q{D@A*LTzuavF4HwjT)3R2FrQ zDQoNA2Q<&s-A6BBd+!LP2`CwQBxi;;gpYMQ7X7m)g5en3n=jj>8~s5m&og=1Nu093 zHW+)&1X4BoBc6Mj0fe=}tDlq`v^E6{yT2A1=h+84>g`yD+_#`u4N&8;Jg_p#=fJ$Q z-v^dHU(C~7V+A`nt`@z4mzZ1n&>nSxRQoF2GiKsruv7R;RbaL9*7Ex&?XHDgpAuXp^-uga>lIpjXLxr=^^@#x&k_J4inYcer@1v2 z8L9(=&D22YJZVB=ThfTn?wUsg+j(ZdA)X=AF81WGl2Vc?X|%hb)tSLS8)Dj-aj}MT z>5~M0CwTbL_skVqb_$vNMS4lmEqeU5tAayyo3stqrF(7K{4nMJl|}y_s|?>RCEq8_ zZkAa??aO8l^WIK$?ncHHW}M}K@6^P}Uppehk8q@?crFGx4Y=Chp!Sso3A-!{jnph%0(xAbObnKft4zRW6o-G6E;r?Lkq7N(52>yjD16(= zdjvG=ghI+LtDi1+ZpY)+ z(R0mE2lCE8mCg27y0J?%#YSh6=a_!U_8 zb|#CBy6-%=+*+>sV>nUx6h%Q-3~lj?*WY23YqcSojqKMx3;iGY{So|*q7FR&XjR8d zYy_`=W9!l$msRWeIK*8^E~jfg4+0SSBZrS%@1~02YOT`Ecee4Vjt%E4FNu;1m^uk; zigh*baXlun#ay`*Jt6znNps{&C>soeD2?q&r?jKb1z#&SqhGY?S7`iAz{C7y>x8`1 z^}a50y&encaH`?kJ&6B--@GL;J05|(LjRRSA_AaAE7;X#ID`TFlIw#+1x4*j_wk^z zVl0#cV53qaTpfNU@Vy8$){KTKtO(}zHvbA#kkKJ#3PbWw?fZ^bl<^}X%gb&7Ow;6T zVIc&blY+OKyVl+<^UZIY{|FbdG4j*Eo~8RGdoMz?F!P-D8?`7DvcMqwBuCoWlmt*t zWtQ;g&#m%tE#g&z#oT^ji=bVze?5i&`sGhyM7 zum%5o;m^v>(Nk{~C>y5cXT@%{hZRUOX*gES(NoQoBqAS~PGJHdU_kv-*D!sXnv!xm zaDLd_P*v`$0|xHEzqBqv*p? zq3`=ig~gLvn}01uAJ*0MQ`}Lk#6A3AyTqI9fGs&%>Q4_68cgl-*(!7Jo^GCVr4Aha z01Sxy=oQ5QuH|3vyFOsUKL(srw)CcK{d$86d^TTxAvDDYL4z9Fh?C*A1?$fNj`Fn9 zV)H$LQ5RQC5tLS;CZ6X}`MAsAB7KNI^*iGBX-94)CKO$7SJpnwK@;QsSmOG9tXS(6 zBCLLw&}}Ne!qMG!1v({D&(;`)jaM8`Q3~x2VyVO>H}lw2$b!BXJ@@)_R;Vy*e4smw z=i$kKF2=TbU8Zi(g;mD+qQ$fGU(WWsWW zf>EkMxpD4#Q={v{K{U-u-<&^5}63%rQhU$5d#6SbzgYmODzBsr)a-yw9tDv}v)`l-8m%UZ zog8eRHu>uYHUBwR=t6yEelxuoS}_gfxq6l;=>`OFRHI|~q!$1Q4}OTQV9AqzO0kS8 zV^Tuu@TuyztYOFO@}}CBeYC&EoM8g^uok;R{wynj4@F}&`8nRbs$eO^m3of6xG>~K z6UHsHwA8+pCg`dC6rqnakP6Z01Ih5nfnkA1vDJl zHacFZbfo<=(o~i{B5Oe2t+kZw+ivLijnwd zxLnwkeTGytvcCVH_y>M4rRtV63ZmmQhNSjk`@0?$zA z^`Nv&ahSg>*p}z#!$bva;+e_WX%gSxWdjr#?7y1ab?#w%K zwS>y*O_1rqqdp^=#bsb5Ye2q~264`J#PEdCK1$2xsM<4Dc%t(+dI1r~mJt;x9o- zRv9I3xXtJOM8#W~fpFSS(ytoR{@nr3(}ry{oOSp3V1 zSK9CwO#{%p?e>Ar2BAGg*33)(JQUqqpgV8-?RIZ&I@&v$>#?p!e16ofM*Y-3 zzq{i5u-0Zh5Eot{mYwjz^Wr+u1C0~C$??MF8G1fB$8&kPTEk~WhTV%x`i_Vue7SzLbcNQK-iPfISiY`Uj#3vaTS`Vv%i;S5 zL-t&D9cyNRFkr7Q@K1g^b7TUF)WPxtEB5-zr~XqO3Gqp z{7sl)&P!*>8$-%{?PX#mw&&G54wiac(3E|gS1>$7{+4rf#Z<3iPVSUN%JsF`k|NN0 zA0XIa%5%+s&o7`V`Xmh9!W17A)t)n_`aG0;<@wrI7Y)Nbr$1zfTW+Nb#wp%Pv}MSP zR_Q(PD~8XJ;blB~4!mkq44l~O%qVPVK-KF0v=ACm;O|`M)5no-{w}G;=_3--J(SJk z@daB4_RNWkfrwdx_c7V8SSTCf@`3hiYV6t%P30t~sq-ph=#hNM&A8=vWYmGn{b>SM zOQt%4e}o&HU}=thEX2=VwCg(5vD2x7EH*>b67PD=^!dimOV^HJKfeFAS9LpdCyCbJ zT`eN}kHS^Muo1%8UVoV@X?E61w}xApk%PN&w0OPLVJn1D(m-3!T<9esb*E*UA}ucb zZLm2tP@EBfIN$KC6e1%3M99^j;9owTN4z!iYbtW$1Gn@NI_vRp8ka)F``!UYypZxT z@(f7dP6NuXr$B_PW(0XPvaEtlGwnLD=_l4vVGF_Y>}}D@RjlIOZawmiyp(>F)-tHV zWSnwua=J}FM9c3iBtoDUOS2hRN}=)d8dF=i0KXLnFvV=k9OlUlHUms+F3z`V&~}ek zjZIJGHuwe=Dog#bm%|ARV~)9cl_5m9d%jxLb67(=|71+P*g3@*3re4lOSM;Nw2HuH z(W|^a9ZTYCr%Ff)A8f0LV?pc2UQm`-l-#2K#G_+C>+npStnk?_IkPa&4gm-G+{|$T zoM5;N>r3||eQPYmQL)I@GW}i%OzCu}{f1(_yhQpd(zU3H+0wjgFl<`PD_&*uf>Ai- z<6T*%$90Tb6dX1Xu=`psKCgpJ60a42B2T*j-`WFV1(Z4pPzeHu|{VjX+JRM%< zTUH8d>%`R$dlQkWkjHpG;Fw&^6KZ)w66+cW3co(Yiro7RkZNk;uvwx+* zmdWRb+I|&W?-~>8H^sj0V9t%dewhk4l*b-zelVhC{cMXDl}`X+zc&i?9)5Aig1*#i z*j}kJN)S}Q0RjrfVW$IxHOS>4d#X70F5S|VYYJ%Y+VF{Z3%Ud8YmV~ z>+etRkguvI15X=T+}?ZLrlrNv0)Vo?`sd3AST5CD1{3FWpq#SCSiNwZj3VWRvYZP{ z6a_*CT6X}f`ZbzQN6y#$MJr~{X$=Gkub9`3EjGZ-##sn z37v9h;oC*@VbjUoKEG?;t>rg3Y>>wt7v+-srN|;rhKPyJ_$K#4a%3XiJKuSAg?5S< zmmBM2oh6@(6x)GRaM%^3DBBZ6^sfdtLsZB`7^j?f3&7N&d(%`JuddI;4(-8v^@ZFl zcN5bSXA?dDq`hJ8Y*O^gSHuR-mJRGCbQ!&7n-ShUoTwtoLVGHVcRF-FAC(2bB-?En zf5*LRnICuQl4H+ot2ahnC6^57LjpO#ez+Dq5wL~2ADLyL^Ui_dO$%(&I;FdS&wkAd zii#yDLzx+&f+Y(3C}kz%8&(I5d`o}zt(vDUH(mmkchHzOa4tUt&BLyQ^sz*Yi1S7k zec@_1lvl=A!|(iE&&ALvCf7ji)U%=QZ>Q*Gvz;QshhvkR@Kx&6QlKBW{C*`UNo=K$ zWB`U^>u-f0&9KuK(1P&rMa8v*ZszgCvQizk?VeR17d_KlCCf9OHSqD&{!W8lc`sbr z)IsWw{XWu};IQo|rwI9@ho?F^SOmzDpXvAyVCX-MS?HtBU~3~B##GgcMt)^F-{Or# zv8})~(f+hKSpi0q{L4ST75=bdAtgx;a=%W`S!Zim zvmAki?FyeZ!&2ja>G%aY7dSfO3+9dl18HF9!}@VKM;>fwO#MRBNt_Aeeq_P{NY}75 zMFX0HzzQM>VPCLoG;-Wko`*YrIBKACov&o{I8b7IG|&kyV;!s->W!Ibd`cRT`m482 z4d!;he;ACmEH7UHBv_A0)74{YD~!hAJEp}>+JZJbC(Hpr<65A2SL(s5mF^Vv0$~wV zPfom0tRHHIuJlW9x$gA^^*IdpeH=7?43gTp$I$b=De3!~^9}aff?a|D`WqCuazg?3 zI5g2G9yGRw7m^*i9$FE>_GbLE2V5+ssM68&5YSS`6}!j&fWo@u@B2C-(XLfZ*Yzx_ z5^J(~F90`NaRQ>_QN0I!}fu{ceX^_;|qTANq}16bf=RNrunYXUbm80+L05w&S$4Pi^Vy>|S8#L5G`D!o)IX1A zgDDf9-Y9bbUhc2OJ!OjLlAZ{DpV?+z=ls@IhNZO%j~%b=27hxulGqGj7DB5`(m@p3 z$8QgK1Q!0_lL=;C54i>gxBNLgf#vrQDow{4x*CFz*(Jy){hBERI|a3yiZi80~uKzontP;_B^*{fcf4cb3hPq&0x217pO4nZo? z2pOz?tE2>Hxi)aLP`^^dQ^!f}i^-D1BJNu&+n>aG#l=5T1s!ZViOBRIGMt=ML7E$J zk5eti;oJkwP{3?#f*u7;CG>x|w#@Wgq`^Jy$H7h3Ve^dV3@h~Q%> zhbG=u=mdn~%^fZrzwsFk$Wa{JnPLCAzS>U+DId~lDL-7gBtA^>nyF)n1_HA6yCCxP z=D5gKj>7f$hyvIIq}0eS=zv3n zf)z`SimsO39g0GnZ{@IlComPrAmuq_KqZA)nrZf^U!aDdTV(p+Qf zBF;n2NJ;Y{ichBXlycI)>z-e-U({lzo=?e_W!5Ga#!YE@s~YBRHzRF07e(-t6a|{N zrM2uu(j~+IC572b@rRxtMRMzKi&?O1ov@}6j)b8Gj9D+%bSatR2F8epEXi*2%JC~sT?1N;* zXRm#7khQOHT|_1U&*5c``=J%{1_4)nl^YF5XIsfWv&uJhx%Q4MZBnF70HlUsTe_6p z-x74^5Mg`#@gBH-*x^hIQc(LTZQ2j{o`Xl2Xy35vKx8&4&t!#9Kgbi`->7v(TkigU-1jw zC|2}zUHQ0(y~`Y(Y-jmXjGQia-TurFPhVGz*JhtCIf!DAsYKzp@b4KP}T3n?f zYT}FLd6v{iHPj{dnoZQ3Pr0f88m5=S)P1DLS-~Li`njg zTa13&Olg8=A%NLvg_%IaAH5BtMI47;b@#~2k90PP&XqIFU=r7dfjjr*X^*#-`3}IHb&srarE$c0c{1!Xt=p3n#Zh&}=^^sgss|oA z>pmZMzqz?dj|DZdPH-O1MD28!9mEsD4`c3(=K(s(g_? zB-S3&yoN!*^JTwBywlG>8Gf^2lTD5u?1iHLY8bo0iXHsNO_^22fnJI!8z;n-H3n_U zFQ$UFC|h5DeXtA~DT^=i!9D-As_ar-s5pJQINj*0l z@=q}BJ0bz~8EM6Y($u2Zeu%gr2)_&n@ z&T3N@ivrX+G;>_9Y7go+j&#{pPNklzG0t(+FPD<4rbgpkDmQDz8Gx?KrgjVo-?kp94KL5C2K_2{bu({(<*B-O=5RZOISFz`;9T3*i zSvb1i#B68&t{u09E6uC%`vmh*fT7G%?Qrzc-J2-Ou#z@>!PK9Jdmh1>R$oSL;JM|2 z5!s(x3=iAr=ictWmZgd7G>VPaN0;-NEbdpq>i-6|P`oLJQ*vo(mc>3r$G8XQ73;u~E{#G(u3+K@bv`+4W1obLVX*kdv%Ej9AM69CA zYXg3nGW5^C+g2WQx0tERQQrEC@YU5d9+X}*y3*QJ|ITtp)9T`10`#zJ;$?VcaTD|% zsPMZH@&_JORW`CndFh-uRh$Hl_HC^f%-byIg8+4w(PgA@qNJNxk01~i zm|+erP8V(^?SE6Txex?3R4=#jdXUp5;6As4~yty=FAy#s+6^_uaLu;dz?VuMvJ5md=q^3G_&oF?)}|g zWMMjZPoyCoOH8{9hx)5k=M;DkPCe$QefXU+R-y`fm=fT#DV8m&KkRLMG#s#0}JI zUE?8%XE44ef4gp}J%-|-XQ^R26bc_T6>5sc`k4N~@#t*;8YQ&pIr1V8cP`~3pO z)LD9faFf%SaNp=&y?_l;-&m(5mL)3-_j^bnuF1rv_a*(f#wiZp=uZEryVR=PFv17@ z4g>tD$97$6Es~*Ky(;?+&|K~JZJ2hnY@_8`bo^o|o^IIqXrS>Kz*_Kxw2XJ8irwo( z2|bmh@rTl$hn060l8DO_5Betf9t_`0TH;Jg&P*H#2EH^73s;q7TgiXNz%l|jY+k*V zkO^rvNbQ>sXn9->fTUfaWLRmvcTY3auzuHE0`TnUNe`C6BnJEMA2s;BOAjXfQ`WKM z2kS&MZ-3tZ!PJge^y3nIZH96{zIDoA1t}J2ZJ;J~h>q%ouc#tQds{Ft?YPik z`Ci`ECoC@dc#w!ikJEmp;}f}Aq(*fs3%Mv^4nE7ssKTT@K;9|8jvTOZBN_eH#VKzy zl_ngZKPhN3ZqYM~t7^6;$3t4RsWAF_TTaH$IV<9W(u!-@qiHzfA2z?aQnt7jA`!Cq zC87mvk5xZ=Q{&&Mtr3sZ>Ts~EaC`YT^O0+FMM41VaM{p3L6j_l-~J-i9q!o3N}p|W ze~hgezSXVpS>-b{=G+OjuE57evhvt^r;)1X3u#3wiO!NV*3dwOy7rC0^jh9%438MAQWC(c z_Koxq)~fD#*9|WvsY|nJL*Sf?NP*{l&Ty;A-nPSbCA&A?;zbgljkxdf@o#{wJ;Ftw zqOUpQ-~@Z}^DaBab_Y*aFsDG7L6k~Q!szVlErU?`vMM9oibd^hWt)Z^%ehvs`8lFy zVn^mJ`Y=;J+(mUYy4EJ)56JIq@ASKqQ2abfH8oOhBis#^tlm+5dWx&=hbQp6+9q@# z7Q>ci7ewEtB>ZDr{TFIYF)GTkLbuvC3x+{f7uc|tlfqFL@xY*A4Go&`=(kQQLC&=j z@HE{+#qNRXn{J_6HCuf5b|Mlu z7V8Kpy;?xX5=mBX!2W}6M|HY{0dUm@lr%^(>q5P_SX{d_oqEL=rorg zz{=~8D$LAavh3#NHtEjR7clnL;$YWV8R0A6HAf3KScIA-3oJr6ck>dN?gR}E68rRiYKM6U0zbImrVs3s+mW9~6U7an9 zCY_ykLigwIo#xWIAY1fvoP%*D1-+4a76+}AJ32?H{FlWG;_%Ugf9{3Ax>yNX6Eun@ z7Tiu4ot7!cxQ_X`=07>xV*|DoXu(Vai9eCnqCZxDv5Y(oQCtml7uY6SN1jVgdQkDx z%O7SSvndo-?RBd{5a&ropYgG+Uj9f2^0p_sMd zoXk%v!8M;#C97)XP?2tHu(}^F11GH7zz55xm-`)s26;);WqcgWxZwFYCD6s1ZcVUp zuPoGxJfhKnKdLDQ9i$PlcKm&_Ci|YRT|2id+;h##KqeH#WuxG7BDBt*?_Q?9-_x~~ z=4&}Dfnu$gQ6zqojz{O#`U$m{bm(&ZVTW8<4>;Jkbk&&%y_h%BHjenM-h*T#F!}Pu zV)XfnJ9?gtq=~2Nw+r4BM2;9ZR2m4smODv)4lTfFZQF6>c4oXa_NrRV{rlmig-q~Z zfc+|cdx7A*r~nKGZx88L#`4xUo-GiNa&F*o*r#{M|9rDZRo6kGF^*adsAJdy`XGpL zF5kf`zg{R8Cegtg>--Z+^ymlb7x3PQSf~PEosgim_Pw^YZ*Y5ieA{f5@e_zr{bnOq zX>F5A(?uJlRdT%ru)6M(Y{=}CgvEhQ-oFq^pO?oVb}3Eh@6C;|HSR$P@az3=+c~FbuH70~1fRsjAxsN=|L%q4#MXkB|i0k@r z<<|6pQSxAOFU=Q*?vvtW7$zBe&>lQ{l`y$XhALN&x|I&3)}}L7UkHG=Gds#z9Lzk0 z!3+0KO{X|X1q}%cyL<@l6u7Dv`{=1y&z++)r{X~dSlM*OB&t^JL$ukp%W3keoxDDB zq)diiF`|34T$(=wfgjfK8x~=7z1NXqe9?YnIo_v4?FkQ4kAE9IPI?@tK<4PO8z*I% z->ooxA~Psq!LVHI6^DB^*=iUr_%+3Qb`H^1X@*4>8o#lz3L?lm0C1(#n6{;oXY}|s ze5!xmV{2Ij8JuXy)8dv`8riqw_`HI4aZNzBXSO}(%1~~`D7VGTb}MY$$PNA3h^M0b zFjVoCbkdWIO8Ri8$tP92?fxSE$>%4YB`Rk<&{RNR2iUq7$^>DzK6mR|YN&QP9qI2w z>o6s_^f!A?B~Vj3UGeOTJ3E=`imu&_i1|O?Og7dSsI4(i)*)TcW_SbxCLDrod$8f9 zT9MaaSl!Q>Exg12f2TpstGIKT{d6Q@juq%6=%jaALCxqcb?L6g8>Fx$V%poS&p z0H_T>ZqMcxiQNj+Zb!*XPh$+zO9?k}<1dv5Hp{Xg%eYuBAF~_lg?vf+B`;wKtmZPj zeq96M1bKkg-c$$BEJ3EWDgjXd)eBb26(oRX2-5u}j(1VHq!n8!weVvd$Kw7>7A#AVXUiqR^%5&6+u*8m#M))*(2 zI0C>|Ct6rU%TWVY(xt(kg)l{FwaL+LX^-)FT((9&Iny8;u zgfZFA19phk^nUb$Rt&oO92_R**WBktb@>?NTqQpL*!VrB+Pk1=OBK;DdHUcpjjbe{ zS@`k8R(jLH79}_v7JB)M6&Y+(xiEdcC}{X`#OdpwBV7)({j^2LQ-u8zr|VOz%xkz1 z_EC;xsXMFRc?~zzK6$NMtGS1G%{_Sn*(wpnvY1mAIW4|skgIbEATiB%kBC1IT&Zp( z4VlxfS>1OW{5MUpibDd>RQSMu{pncKTz9$Pr3U-OZkQNfiw=fluel*vMz|wrQ7z)V z>YKtXlP8M)b*MbsqUUOW-V6Yf%R+_R6u@JIdqxvmDiWu~&D!U`LyX0i2(+!f7jhYT)oZXpjs33mIu4FDMT%d*RU~8Fb9~NT}CyWQ;=w0*P^ZpsbRs#x{pTE11dc$s> zbF08kE>T^ZivBnDGi(iGcjzt!dF1Uo)!c+zoMXtJ-Dq41|Q4kdQ%n*uEMuYKvet?aj| z;ZZ!PPjfz&OmH~~PwU-w_DAmOY~E@4m2_3`!58ANO;?-G@E3yLSmI)wK}4WYNPMUs zfQid-!R>woGF9)3c=KHNu{sSu4Q??5q3M}Cn$?wA?079Um@mM#MV}$?U@I!zowN}K z*Rg+;IM-1uDQPgjh7YD>9hh>DvzeDWV_s99v;nzzJ#7TS$xdOh=z%~cxcQc#2Kuu5 zKCp1~wM#c-{U9rAp8_08jBB9w8#-V^=r){VUJVwYK_ILs7#CTOsJG|`WtJd8C+G%4 z+z#!65s$-URECtZ6})lI#Iv_zWQ6sOEAd+#=olO&`0E+RBnEWm2|pA+D6rT7142@UlJ)D>pqLZ!FjfBY>!X>_NpCm_dJa63IwJN}$nx~b2Okml_!WUXE zeLO*&5=-)=QuD@2ymCP)`}J8xM&T5D@{{=s#bg9X?Dr2}@0l@_LL4r0`h_osgs=fe zUPC$d50j7}-A_O%+~H@W()TpK6?HEhu~pG?c2=PcOOUF%Q6kCY`Bl6 zjZvih7+zR^e=FPCX6jydU^i~v*SR-@sATdCy!zA_*a&}nBojYMLxOApVCvF-JZtD! zmP6Ei`x&Wu)SZcdcqKWXdXO=wKCW(E_B@`FgZdu-ZXL$9Ty%l3_LxaU(og|>E%A4; z%hb$DxmCbf%wIJe!@{`lKP#mKShWwwjZC|s!{C93)r{R(lYQ(!^76fREOLNOPnJXS zO`BI3Lp1uh*$61Yj@#9ixpZw2;(AB^@bhPZ0OTY=l2s9(wkIGY62JW3L^>dDpME*^ zyk6{hCT(E$n5n%&rCo%JM>n=vwfD%+;<`)s#Mb=ad9!E4y%-*Y`k6#O(8SHOP zA?G7Fd=I{soIb_1zrNcEhfFJOUv8?adR+@~Z5bFhlP^+Fx|#||9veM@vjv;_FP9#F z*uM-5fAUgWcOlMN%fJFB!rSjIAHuoFwqmH|2${5s7Cn&H7x7aZjO}ifCKnu@ z45wRLlnU62E&sX+Y!nf%;FRj4a$_O9sVq<~?Mwj~SkA}Wc1zc$?0R)ePnldM`=0ud zN#s-*VX4{(p+-U$&(+&E4%8_d6XI zz~-e?|3nv1hQZ|vFWx93MrVUZW4h`%b&J7n;Nz+{!ycBS1rE!OvNe{L`%bFt=yZk9 z9x}R2Jw}%9kwFtUW%&3?CQ9FZa(pDm&E&JT*AD(+dck(3 z+Lw`vYZig>VcQy1({-(oT4k;G3bmEVmBw%15G_cCM`XH0vbb^gJya>q)E=bP(Ee@6 zfD>2;`{YQT|0%lL4unpdmFzqzDy}!Up@W^Yk}Kuh#ZMpurG`Tkd^l<9N~+;%_Ws2} zNA!69zeQxa>^`4UcBJ?*B$cSCqV=DQp9r6~tDd3xu;^~b@0Q*gq8-p2Va@5z>qoYt z5it6dK|c%c3)G)je8kGpO#_ov(ET*A*z9(XvOVtPAF9EUifs>uXW6BZ)-I`q5)Vd} zWD{LxzN*n}ZwIt`2!9Eh(x@@P>m4h8aZ~`v(_gEF@4i5@y$@aC*&lSX5f>B1tTm=I zY@B?+aJ}1lsND-P2l6Ykd0litQ=B{RzTu|b{M3w29?fxC&UkB5l_0`e*Xi_Gu;pw3 zsv8i!5Le&P|4iw{^CkscQ^W7-^4+44r`<cdz`u)p>_TP>7HJ4%L7btrEI#bnOn* z7Tm%|(V9WkZ7vbN{vz}(+D`L@0Hwal<|lbsD3ML^Appb-0qNC3FD&!!v4GYqmL)FN zm^*mM>_zhPs;P=)kmG6;OQiH@+O{h^>1o5jQE}>&z((>^0#&y6VNx3#^Lw|MzW$ZU zq>eB}c9=m_Lu+86i_QElz4;e<$xGno`q~k^4ounmw?}jCRMqRYK+ou~drR+${CW;$ z2j}p!y)H&h#FJLe*;s2Q>A7XXU~$b9|GeVorE!)o%_Pw?TI@Gr>=X0&a}Q0TJISs2 z_+`e5c=DuG_dGu?!bB525QXNZwfkPf+<-7;nuZ~P**0w`M<2|IPi-IF5%THQ_mXL8 z8I+r9DMRt`8+?#k8V4O)iDSF>FuHx&%Z+EkpCljwS3OJJOKkG(FWV_1XWie>NHly2 zqr}fgjF?w62LlavmV&234Xeq^3)g1lV30hEpy-MtN!u;G*O-yWgc+eqod)^c*RXrE%}{Bd9$ zh3kDeYJ~2WdRUQzwFY2*GcUw`Y2o{(T?=E%kwmn^iemeeAypj!yJm6@bZBZEmU4VP z@w5rDKW#k%2@(r2HWcfN$t=oQkwo*0O`+58`!UJs4)01aqPmENTi6n>jtc`L`$otH z%_V)Td{oBvG*N(tpOpGszNiO~<^(W5=2@=`v&_2`c9E^iUWx(*FK?$( zmwtY}R#^1xzutr*DBSJ1XeiDswenu>j*wiF*cMf!EZT^;>ADIjD{f`}3z2E1$p&~i z4Emzj?qIaH)wfFrIDns?pS++omAMaJ_bkNr6&^>aiHcd4zk%}9iKTYX!mX`6QW8-C zS{6t5x;lRhFcCAEct|*oTvDn0=5q z-g+dmmvO7jQjfhthO1Z&Db(X!dhZh-O3b^9b62w35`2vD)C#QUC}JI?)EM86?T+?4 z|Jxp%=nj*uM|Gia9h7Er=O915bzTqs?cya}06!}cP1Qho>;8B`0+YjvYxqmE(-hM# zAVHwgpe=lJ^>WH-9c-fY{_QwcXZLoCQp4x0@bQ*Fm=kJsL-F(v(aFasZ^1+I(8E+VcWsnA5jS0EyI zU`bnG$EUTsjqy1SQ7!4Q_3)f<*aA+iLJ_sCvzG(RA;^3sR~SLusLbFW-QQQDXBx>Z zgWCUIT+>I9VV9KDuBIPP`u`YvtFSn?U|ST2AdPDpcL?qh+=9D1!L4x!?gWPfcXxMp zcXxMpx63}~-ut%J+TS;yx?j5gnt#=(8dam@qmA6ws$X0te;}$e-|AQ(+AcY9ty26z zb>f-kezI`3gcD!dmxktA@raFCTffisN%fzZfpgOs@dracw=bta7@Op5uE*C;ScR*6 zXI^!#M{)L1NHRq9ZKp-43lLR^cR;~@E~aO836U`Wn!vQwG-=T-C4RT~E4|n&A+Qa^2}cHASJn z1_K@BRS$-zi$k$si8%K2kjJ&D?w62!#OV@|s}RjO?@j%4mV; z1$7a>@qw5pfGZNrVIFWEBkpx=YJazb{h6ZiJ_dW~(i5yv<|tAv5O9)=I=Qx?Wv%Ud z4%ASvzTL@|OI{>(khdMYoBp2HjUUDV)*1+PMTP%~eb&%{s}AJ?hh#}O1#i^5-y~X5 zgLiBEew`W=kRAceFiMlIf&*oLL6Yjc_@eNVT$^EMApevK(VzNNn?=9|X-m*n`S7(g zujy7=i#MkSl*{=HLh(+|)rJO2>jb`@&$*A^h3wlm;q42N4lw?h@4*TdIvjKSlN8(- z#iw1Qws6V|*?1kjf~al_yI(zZKardw-Xdno zV)nYsYRJ|!2{M_bf$#|{YuO>N)GE;55nPt|uNj!Y0_U&Qub12fiCS6Ou{J$3jhcUs z3;3PVWY1O2e`yjtvUsbCfm{imN26N4uB2kpVlI75MYXiR(JgzDMsG>#1toNlPdRui zonO4#+Pl0dPs418?nv%p_UfxE#H02}NyXUA*H$`<6sKZ&#UVf9gWVGlf00gUrR+v9E;qn?de(2w4r}eJ%^8#sypw*eqk!5I#|Ey z4vmeDLzQ*t*fsMM>XKMjo0jyr;%wTU9^lXEEnawyOxAqwls{F zJ|`iGh@l@1>?bTE$FYxX$aI=WD3x{$;a!n&6T zdN@%;SF)QPj~&2%5v@Bg$lOMB|Ni9jKd*wYppxzWIG?0IMZ!_zK{VFQxY?ngH7w!c z{({l;8#ehH05#T0u#z7*(cF02rHE+fv&EBT@G*)z6KcdNR^%`_Lb!NBc~c0tnS+lk z{-PK@0jw!v!sFLUJ-V&F=V`+X8cXfrcX1tuQigMw>MNC)xeeT5GrCws z;#f*7-PjJ$1%G^C|6VB_iD7c=m4G%j?*Y*lWkeT4RnSJTKa}2Ol}W>sEY9&PgGVpQ zC}}-PV+?;_%jx^!}Jo#%fOnMA{V@@sr;twh+8Kz=0d2PbDEhPuR$jl~GbCN-~D^ z%8>^sv@$y;Nqn43;th;T9J5{mfje?Ox&tJ%w*BjfWGu z@f!g*g>b-PoAb{O-p3-s6~TgwwjbselIJgfI>-v-PDhaaP;%Zl5UO7*(6Nba(X!*D zu<)8Hv$4>sWqdJ4upC1y=n?6gApy%NUoz;3c?uwjU%VoBuyu>!%ny7?{zn>5AgzNi8=lx=N^zgVGo-B`T*XV(_1E?R9n+^gAE2 zz!Li0%}Vvyk{Vd;KPkLn@Z{zT_#(%WD&uSf?Z($gCL??4Gy)m5J}weq7NRctr(Gqx z?A_|!d2wBW&Ns_Yiw$+p<7vO{lUDE%KtNeSK0vUUjEUDN%L>~|`y6)kfc-~|`)-c+ zj_t_a#79V}8~t}iy5H_@*akQhlcCHg@YiO*1Wq)vh+s@Q+0Q|Q6|_e()HbQXF!Qh> z9!KTktR$Y3cq}4fbfJR2;Z4zP0W}5m(0F$VoB7K6KmU-3rtpdL_-DTXvX?a)%rUy zwmWMyGw3<4^^Ze*^+MMVXS@&#yeVC24YLY&bJFsO_9~P<$7XDx2l_zwrmwLKLpJEL zMRV;!8Utjt;80I;RPBE=4G7z!71Psc2|^Tou2nL}%6;fZaE1edgT?UtC8|1Wx_R~~ zO;zku?5+LW>zWh#;DUkgKBCroRi8ec`1ZcT_EbcfO3P^(=>GDlY#Q-fyP1YO;M+KS zG5f%$eQdEa9!r>KfQZ5NTifJZF7G7oI??cGLF-Nn#7!^dkkySYQK%l+72^L3hgens z=AA_V3F(X)(=A9Uike07b;jukLNReN{!xJsrZgnj`U{QQfgEBcBLomwjzUb9r&uu( zBTi70z~PA)>JwH&Dd4!74=I1ty{Al#xjV?mP_Zf#^}PN`@pGgxgYo6j6D_c&`lHOz zwo0b@2QXD(pn9>we5(m8pWlA4^M-+w<8nW=aoetVRe1?;y!7s+xF;tJc< zPfZ}x4g-YWBMiIVmIS#8b+`2pY}x{!!&6JHS~gcbR6IDKF@|f8xa}Il^DH52fQ~2t z>uV%{`1JEJhlVly!)hy8Ku22h4X0{Gq`_*5nA-xV*CV-q(BcFY&Bf?D?kRW1DF0(q z)c9~wUVAHOcr1ffH`SpshN!K+8C*Xs;v#OI?hrWI)kWFylKOg&-9eHTiAk@q=92C; zo(kmo)$saC_?{}wq3tJT`}V8&zXIK7*G(H0STu@q>PbJqKnQ6&9Sb5rpcw9DvAbfK zI0@dXFbakg*LgmV_x{ga;Y$1+2g9?q!ss*k3t*<_rSRitLEl` zY{0PJU9+eNj#Z!ZSK5!J#d-!~gQHDGa8EL`vIoSTWa<>b6{ron9$Mt3oI2BzH^TAX zhINpgWIUp3JX?Om{ndg}5V1pqJO&z-;)CI6rb}wT*P`W5Mt-g;>=+o1Lt(CLwx{#; zH-kIKgfwxy#M~4c!<-dWl2O)?D1M5ZZ-%2jW>2%jhE&IfaV7TPy&QKmmx&LS(Flb6 zY%*T6Fyex38Ra_c;VmsP)pgD88SM-_F}~9zk49H=urEFqptvhWL}AD17nd#BYdSf^ zYfWr|VWQ~gWXyPd`&!4#7Z}-xdD?`#D<3fVJpkP}U4TGCG6q3Pa>3B1Ks@r%x1L9b*f8|1PJua)ce(-wA7!1slYWexhjr8B~&=cCvP(kRX=Xp`{ z^28Bn!oL%_)>=b9BvCTw7cTT_m=SNr8GXL2;+}-RpC+K(7RFv1$Uzrd}3Q!aZZO^Nqeu4aUKo^iGvzrAze)9dG+~^EmXk?tU)ND1;*i1xLyk?Uk^4H+OP1Ro*UAKDISvKJhchjZ&8!X zDH!bCj8)p7?`%gJGCaig4@`bu_%-0x=>Gc8GVA|&1@$+Hx^*KHdefY~F4y#9H}{-? zfti2>%k|0AI^xxmH{WIt{+4S--Hz!!W263t&P?{xME%>ya`#&K*aN74(tqqozFFn< zWNuITDCc^pPm3Y9P^0+^G<>mvHQDWty;mz|9Bp zlpk=t!O<(#q~n$^`@2s8!rjZngb)8BnrO?)ch{q(w&m-O#vh1~$+wxcXAD_4`#nqq zqgm@l7sho8IKUQ&Mm5$m6&&GRquFaDno+~TDyS7nP69`67KeKJcYpukH!pna3dXZP zHq>i>U2{qbWNCzHsHT=^ce@)~2+T?F%j=YJus*CR*|HG(KbgzVA^I`_c`C^&ZWtI&crVtd ztCq5O5%BJ5UpvM8TI0#M?mkLm(D6d94>d$ecn+;Ht&U>~VWP9|s)Zyl!R;isph_;;hMTU7y>$;$w^|{xsheL+?JaJ}8<6abX$I_}eOmBH5O^Dl?t+Ti3tY@MC#hJamxz3zCcv zWISGCF&;AbOjhy{+|1i%9{exug_u`~1SAWYj-^5Y&%70yX!XRY?0*QUC_`0|CxjpDH;$_Jz=!7kAhFHP-=TQT&Z|F5haLW4BZ2M?T8_Z9Fa1 z_;1#wSp;w&n%Z!1onN+f#-auURaim`CUJ6f#wE!gmT;wjkuXKx#ZB zPGc2c?0Z>=`(Y4dlt=wUYed$_g4(46@3vbW^}*t<1iNk4S4>k3fRxUvWN? zw&oPdt<76dU%-a80gV$rZz5zFg z8IsXQpUY#@8P6~wd|1PmBn?R24rjppdl186m?I7d6X_&BW=nufv|_538oM|kfc+DD zsToe*xX28oE7}LDfU%d}dRE&F1&3Qd3uZMhH;n&HHw6ek9O$`O44g_L% zZP%7c-njCNv2t|nZCC3fF1WazqN5OxkgJM_Fhx8lmaOMpGUguo?EX4G#sz)K484LW zc3Kl?-C#2!E_A^tlj#D4$UCVZVFjUz2xP2(v-^l{p}b7i%Vuh-m*XWAVa94Egi&ks zfI&}~OC0E%Ta0pr$Z&k`P`ged_+!eLd-(OG_`1j_Y9y(sB8S;mh2m4E9GzG}bY9c3 z@e1Eo?~yWp9t8D=D2nTo<9%Zma&vwrtx+?9gVnmXsU9!t^YjTBo%!ULSbu|KSO&&B z?of0Rm_oD>EvjWW`26iZl;fQi*inBrok(GS7iW}szN#5ZP2@bvb25Oz`9lD`YgH;u zr{C?$@7c;<`#AudLZ^L$+UWPhK%1!NYTc);xC7(3c!<6~J(UP1f8U}anYAFbUDp2L zi9GCRA$%pWpHlT5hMN1Z-F9`GrUa=@P%wNExg`{Rf|Go-Cbx|f@D+u z)JOau$fy>#lJllu))lPai0#OMk=aMw+UbFzz;=}zBt=+ASad;V&cx$CyZBJr+aCfQ zaAC2NNj{_?9201w-TEMZ5sP)*99iOSbuj`k$?i}zzAb_udhoDEqv7T>eRgKun-s-% z;gZqaf~z;vN4UYdO#q4gI0&B=)_3}F3XoYdd&>iUj8T}l@uHPU+he$91_}q(evG!Q zTQ9TaUYd`Y%4e}_!^N~zu1QlQU1mQmJXb^W4=0kf#r*F(Htc)DBG5ehL9EltR+Y}N z!Dq>``!3il_4+xmyw!d_5M$5fcqhO=@dRW(ckeiW0ae|-Pj^}nDb>APs*f*N=6jfc zfY!*5`WIYQ{Gqk=%zV4Ten`80Q@)VbZH5@pC=xN7aAf8HSZ;8)RRt4&6f;p!o`tls^Eh7x^ znkuB)m-*(cMb%I3E-5(b=gA!Y5Ln2YQTqLTjbo-E;sdOaW1Pbgwq9t0lo2k8;!s*D9!(5IK2mDzGn9z<bKX!PWfj-c$qZgjo8@wAAf$RGn zlyVlMt?_l3L)RX}*z3T51t;Xtt;pH@%1xC1vc4Q#ncu}EznlmumC?P!3JaU16p|Xj z1|}J#d^ch?6Xt)svAD(60V$O*F3G|&-ko;TB?vHbpg9K(KC~t+#X-OFCoUyX3|fBM zc)Xvv-X5tKPO8DaK79*TMcV}G=MHTIs+3zq%=RmlHAb>Unc8qu?p zVLhETxI9y+FDk}mpx6jYH=w2Z<&ZGOd)8Z<&+h&sj0zN&d{#6GMKh37jPf`?@Q%=9{0HK`V{drkK zQI@Zr=^W3nVtVLl7xhFL|9ti&xWcD_#J-y0ENlCn>bH+Pc78m8qA5CZ1|+YKKCV5h zJ@#dAUmK1Nt)kjggd)4?0*#bI?3tONpVj_u#=QK@Vq);$62NnRXtGHxa7;X@h-hNl z1devFEKKtBuoS~3tej9xNKDC3gFVntlu|hW#F$ysN4Tfo8UY|Th%C7;97jOj;ESNQ zu)a37@0tUYH-wS)Pw2Z3VGBRG_!7Z55##Yl^AfqW zmI#!_GMRU;G?wlbE@)b_h};zedRTu7b-%-Qh5*Vy6f)6_G91$bz5Sb?ULtO zc`M$N(x2GABStYV&?6`0r<)@>mZ=o8CNveP+u&fo8x5^0CI0kYq6(S9mXRFij_7bWiIuy9-nGOocE;6!kFBlxF7!3lEXkYE$; z_-BJkabAAK#`o~;ckv^d6Q`~gs_njuXE2_L54Ld3r3T;3yq)R0u8EoJ%!?^iSMw!u zmg7eafVZo~aZ{X1`z~1<)|2RKxg!~fRx#jt9(3=ts*OM=!573f>E|GV(ln^-=5d%_ zDTLw-PhQs@=;a7K&E$C1Io`IxH#dtorfe*ke3*pz5m$~8WnOIHw_j5u!lN_`*|VMm z#C(zrOm@PJcdp%vS|d8LNOzOJ={PwYXwkYb{LbxHBq^>#D!3^*c)sp`icD?i6^NXU z1&YFL2S?5S{75w|S#6$msq60)2!;>h$xc%B1{<+~j9sU5^>U zeo5MF1ZKh;gmLQ2cX$#1yY-B~>$VU1ju>oAzS!RR{LghoK0i^E_Pfl}V;}aWg=F^D zQUViIfW4iC;Q2e>`}xb`)X!VX1r)JstZzMYY>gXeRB_T7TO`4g=)^3@ex0Qhb*Qzc zsj>pm;QECWBDV;hCjVposj(o;7gQRAOkbu{yZT8kPKXe=@RDkyk}u)DxqKo-}!w#r{6@g-4m{f8_JvHthu8J7M$pYUD;Gkt*-n=wSf$Rk$r}?XRZ? z{IpSRA6cuZSo;UN>*xc!7au*|G_Qd8@g+u&dF%a!)&c$#w;J0J&Nq@P&i1{<z(0ecponw~U(@`G~%C%)xfSbXnNCwR1SG6~t~BsKXN-pAo4hxQsq6U@?J zwe20FKJ~#J4x4)+9-$i36&CCYAx!3ot}fS)El8)47=1Zd6rz7@!sLd)dBwh3(ibh5 z`a#t*0O$_`XdbYCdAr=XITR=G_pJ949oO5dX%MRUiTDjs-Wdn`=f``SPn>CNm*N3= z3#Q#ir;a$|fPm_8s&F>LOYoaB`#|o_n5P(sl%W08p4>hjcAk-v0mYc3c$ojO%!w1b ziKVi5S3i?Nryi02W$u7ucXTiP)s(*_L*eJfOe`0y_c0}34K(6%l1d5Y(<~)FMaT0= zhjFp4=UdvRk`LC$nzRAc^qtd#;bKCxo!5IBg7+_$o0q9lCUb=*S$S9RyzC-h0y6l^~dpcL@EhixJ9kZ_!9w=UnEw=$s4h zQ`zX<{COM$y{A052V6ffN%O34aV-|pJP9sc8V8w9MoI09#R`eUFy4(T+o{B!mNRfl zM-@SxvyeJDtOxorU^Iiln9u`4;K?)vRg`YfhRQfTjN6O0WnC}F!&D!Vt4NBu`=w3q zA4~2xL{il!trxrLwk2gP&F;{Y>97=*Swd6D+KMdZs zK!X4-jQ1V@tzSJ5AU60{hDLCNk1!!Zmvp>PZNC{y1Ym@nweF~l4O4H!&vgkm)NAfr{26+h_xouODohx& zE12f1sRdKADJzO&yR#jU_hp{tbrmI@aaFRTh3UR^i>t4k*TBw#j`B=+{Moc|{`xnK zO9==vfqy*iLSP?>rA{0DeS-Yu3*T4%)B&c*P=Ug7s~e2oFhilOxBEKRbEfTD?}scM z#K`Vi*SBqsI*}dcKx)9*zJ;O!oSrP0;HkEIFlS*orIWr}p{*MGPEfUyNjwK&NCDr$ zz=291>0y`1x?;U%1xmaPS)Md*kS*mw)(3 z9MR!(cz!YbwKtmWaeT|u-_JSCVbj$8RB&No0A`1xxl}iHoHj>TTz+et*4PxrUBq$4 zW3rDpPqr}~X&`KMbv1VlPxu9BkFpzMHGg}=f;wHZjPwP#CmiJ>;%zI_!@=hIbP`KS zUdWEg%96`@_p7gN9MzOYHD{g*7}>_hC-#bM5{F~?lVjHEmgtvhB7^P8M`rf7%#fCr zS+cpi_4=(Vj`H*b=s(@HcxQW|ss8aq z*+TLJtJ+wd{=J_YtgBF)gA^b|04U{ZxueRy#oz3T5It$h3k>ohzU_lj?l$z09diN1xwLjwZ0tRLN{4$x)AwR?#ZDL z-lf^*chHsz{NJM&K!1a0 zhe#KQp6zor&b(hz=75|hJ&y}i^@h(owV-N?R<&?+v?IxzVI;HO277P)t?;+BB-^>P z%DTSweW5fTY+cVE@%@O2S*UJd!E-u=-I?qW3f_GbY^%)HUm*_i?vx{)oK5+Tp4v`l zKIDv+pE#~YywdVjW#07bpu5iRr%$B2wEpGg`24B`wUeK!;4#>tg)5ywP}zlH>A|g? z2OQZ?+?0c3_B;yzH_uZ*ZFaL=92jqIf>nAA6v1u5By8<|dI45AU#!vTENd71-(iLfP8=Yo<@H zw6Gk*d;fk>XN!G+^M?r9Vu{t$JV)<~g&JieEE5zp{mb=(_Y>RHY-_~HZ~x6qh{IR} zj_t{*0zil0?x5s=ByA`s@C6SX`o~`7%*Idwfz{BEG3^8uU-NrOMM74|D21g2k3|C& zwYDEkw|TPin^yHvI2Qw5+L%_YVRwMdf)~gAsI7N-^Wq$KN1Ks;iE_{weuTuA2##CY zLxY)(tm?E5Wf$f6w;Aq8IR2N2gNl~@xspNhFy8WI*|gO)1J|`5u798;eE#|9lYl#Y z{4|2_MX+PEQu#RKzY=tg?HohtbTKKw{M6M}se<=LgCnpBnaxI4F0xXYpi1?}flW}G zyUg`=P4nO5n`1Lg1_!SiVF<_%l_ExXY9bdXq6pY9#>DP%-g=(lQWmmTp@p;@z=E}- zL>dls{5j_sdH<={O4cWV#iKYQ#?XWp5PBL6Va=x+AlZMy!4 zoNH)`RKFLkL-M3hC>9|`EHJFJ5)SLU%R!V|>ukXW|6plc(4)d;mGc&?raB(q`4~dz z6f^tpmSG_^#O&?J$#!O_XJDn;Z&U8+DFYOvHNiMl#ZTe<1dK8o=fFbG$E0BH>)zlF zy@JK5@RrNR#G;ZEsoi^N)secY+HB8U)VEt~_$;}HDrmdQUu>}GH51z0{*uu%2jv0p zi4G{#^4PQmxiHB_DKfT76iqN}QP$3TeKDMcKs2*y+lV~QyWpZ9Dg)Daq0dMh_Ox^* z<>-H9dEu%fFztjQ)MEZM9VAPb`m~>Ak4&1eYo-&26WGXcH3V>|BMF zN`mHHRFD$aliVNDOn32-*KJijwzL?A7(Rw=2OStZw3^$}Z=XG%xThoDrx`w2me%Pp zCZjxe{i|ywzP{Y*qNue%R0P@o`Uu0S}ScUg%)ifeQe| z6qg{4P0eN>&p${rdBsJl`u=X{TSScwlQP zEjL80{AKzl_oQBNP;GX+zSbSKX@sh{*FFD-c=cjp=~yF*1rOzBJncvOcZZ zNT0f>Uem-JV5U0u@u=ZzzrUo%PPTX$t83WZy_YpmP-s}Q%%QC7&8@%2#8B&#TT3YZ z3n#OTB+hrG6W>ye!+e*qZ~ZYF<*E4I_irJlFV+h zJkAh|h%O5#(@e@IqE9LNKnQ$O;+bcAJDt<2bAm{vIf-!6vOOGJO zmeDYQap`#1_;hyoPikKI5PuyzblNb|^&bSPBr9f-TfgCT%gkKne|c9m5ow_U!*!7G zkc9%YZ_5WYf1z#bZ&}QqTS|;PwPZ9*`)4~K+}&*FbK_qX0`S<=IE1WuWq2&^*YV*< z7|yqKo8>DALpr*0BWw+k%D>xU1P^9+5MOw;^36*_9O!@KB@FLLw{3G^CoI*B~-Ey!?3LE5^AbkzAn2p<+RV5 zfmq6p2`mnfJ8XZvlOJT|iTg86>-5WQaq@C3o zmxuH2&}VA1X{h$%SthMrr`;@u5Lq-o3lH&!Sw-J^_RsxPxIP5=RgY*8il}4H@U;|O zQYxIjj^uZYpuCKN7jZa@cor^+-{jJISws!^Hz;%{hz|D6XWxEn0LEgBRP74Is5<7L z%)+1`e(Yh-=VmC=RKV+7el4U4iB=&3+7d%;uM@>Ve-!_Wqld=<90TX^K{OQ3c-B9@ zr74N_VY)NwtthhqMaW1TbikCMr_>PHiOR_axm?`B63u6A8&P+;Z;X* z4Dg2U&=8N~t5wg={akRg3P-?MDs4DCrei>UMpI#P7V1xJtqDIOOaVCVQ99TX#C@8@ z%TacfL-+xaBKwcJU*W@?rGJ|dCBq9&m3`)(UHs`dVbAaaQL3}GDmDt4)qx_@QQGE7 z4S5%|33<2n#DH$DLTY(E0h*}^k0!}Lb3<|I?@r-{Xfkxo)^n7-?9s=IaW;B=f}Go8 zSQ3b8c8`ly_Afa`U=KH78?vn7!XJ|E3%|>24I`#4G;KjUm_O*L6T2e{H*f&~4Uf_Ug>~QT;43P0M}tPqh**meg`OsO!z`FS#0h zE=5tq0!~`&t4K@I2}H_RtJ$NHBgIo%ehp8`3ba#S`!nH>?(=R`!a05g9ZcWYo^dzl zZEs%o85UY)vKzo~>~iRpXh&2TRGnb3v?vT#v1p|u{P~w`5$p?<@2c-ke?_YLK4GKy zIyN(j?{^ZgtA1xdG3$Znv#u@&W(-&-?fsLG>^&6Jh6-KO2{hOUefJi6Y0N#%`4*wSXTHr) z4y2l%@3be^bJ%CCnZ~@LV}cP`OEl!8+v#HX#ARj;36iJxWM4!5#V&grs#;q$5rHpB zBC5IyrwbS7802%qoSYv(#W*ah@WwA=zrYkOZm-Q5ODLQ&bzQ32dfH5TyR`k1ISYL? zkGPqPmkSN$4>mU~;@WGIpUx&U)yhwq;l{gi(?7Lx?|>QJZOAcgk3x>716cb?K{c1> z)%?!?{=`mseCL19*qqcVdL^rgZ&|O!*_Zr7>tg^~ZL9|XRb@gr&-0VxjMVz@;_XSD zduj`<{|ZJy>ca&tg2_()g+zunR5@MDtadcgGxv**&AcV#@!#hp#Bl#A`-!|0o&{A{ z^8tO6FE=)YGoo@n98?rMuX(w zH!0@NU+u1zMY#h1mZE<(3~^8F7$%@i*}Pq~ymKs|j5c&6_}bUy=0O|&`=1lHO$jtJ zdis$Z!Y5zSCKsDu+)&|Cbj1kNH1&loG`Q-=$lVHCRDuz_){)CFQVu{7JlJ6wl+9gv z4p-?ify{4tDCH}aer^O($QTIGVxoz7X zif4^+nf1f5S!|(RuE3m&`;#WMJ=b>6XguQ1XSIOHf*9(zg2SYZW*tL-1IWFMAfwU;4 zA%mZC=CQVkhuDC+$gWm;v{+=%7L<<<*}iwzJYrjWX+sI-;B_XWaYU1qCbOZvS`i?C zU)LTtWQko|Tf@EYS}^W-AYF5mB*(vFr4Ca0NA&uCQ7z{z)a(^@ugdd}I>cfJc`emY zZ!BMoBLcFn*hez0IgIc8RzTS|u3f)hY{#5C(V zbGOQ#=S!ahievad_37;mzwGZ#F zq4*1EU&FAJZ@j7`IF@$7ZM5 z3I|}vNW*Y&?h57Qc@W73`7hb>|M(l8xYM! zX;DS`QmjfTQr}jyxL|2347i@bMDQ(_605Tfvn+Fz`g0Bh7>YlIyf}H@+IRCf3u(|q znD`5OE48)?{1y#!NFk-7rXT8K+l}*QH2n#^e@&$t71josi^{;>*L}IN3U-h;La5Z} z1(Ma@0JKkQ(@bKwgZrn{b&G!8k_!Er(G=F4S`XX6b2?gE&?I1Nw0FkcCR*+^%wJ$T zARz#K=CCpk==Gs(Nqi9oR{@;5O#MXl0WH@}C;09tpeUUhs2`)a>0CAeBiBt}7nMs? zGjaPU3nxrzJl1XD*I8U#k)GWSXqY|!R89E*KVNi!peQ?3my|H$Hod%K?2nB%3L+}&Fgd443 z4xT0an=OB+!t+djd(oTd$yP}CBnNMNq?0!p<)IMCMD*Al5cSqH>NI^B;bc}>pM33o z=A1mxlx?#wCAfRBpj%M-wNC?xALlRQXuNMF!!BmcJ8tzT+Z(C6XEG1aL`EOoE~d}t zhRw+>PAfcVROL(_W<6Cdl8zJ2ag$C(htm3<7EPtCs&R3tTd~kWr@Z*`v4I> z`n}{qUQ@|-=ANpEN7+100Iwzdx7c#EJM1zZ9@(p)z1QWG>8f!5K{uMM?@=M>=Cz_q z^kb(MZd}|n=?r)s(u^e6aTZ zinF_9&U;4bcwPuCh=FY0Lmb?h>QsSLu?1|sMuh6NAa(-i>cH)m&VLnA@^Z7va}d@0 zUPj%&C@T|gTN^)Gvxm%KCOAl2aT|ie9t4rM_I*n`@VY_QTjTZmkGKax_f^z)b!_03 zRVwm>@mLNIk4ye9wVXHbk1ik{zG;?;S8VLcEE837~t48GaA7*Pb?p(Lwp62 zpO_X3JibpgdrY3JL7n@eMnqU8#-}pJl1zDLB;UODRl2&&WSrrxVHjwiX9oN!1a{x+ zj_y)bv7~p!hI#%L%#npj_6I{7Dpo(#e<=|d}OU@<6EWB)k`a-4t zxv)IS1D^XCPsW5EHY;8RNA18G#Sj?Q0p?TL^NsOBV{sT8vQX=uh^Heuy%~GI(N*0# zS!zr?iq4oGnWn3fcq7XUPvN^uT5VRjke}K6c@=Wcn#Zg=>Pc55!@~K_+anX;!pj!d zcVcfFR^r9=@NI!Beoa=f$P_Y2hH?INl7>W2`v;UumM8E7h-vIXFH~f-}nG$@znZvBtJ%p z`+i-RtlBi?bZvHCa4BJodp)f9W;&}$C0O5P4unj$r?+{JhYon3e3Qy-_y>zJn3BXe4Ijhj5j!$~M{U>qglhYosg!Gsv>z!dveND{V6l^D5t(2F_NMhhJXftyb>xe3d1QO!55k`H#@8+P} zl-fpNp^m-JD%N!>OEkh1a-To{N(TLzN_giv-QE zM*mKSJDKF~2qwxWdYd@NoNK|*QAi-P61#OePKU<@lZW7dp3j<}k!<*}Sw&mevqlDw zlcYofOP+kTvLVadJT5VBh2f+QGA0%Rb0wiYx zm%XpQo@0?u&nn-(tN3_f-utn)i&rT?mYyH?4SSKCVzUuyar1-uhqPU!};>6k|XAAtZ^B5%}HLYE#?Y^4UbG^{3|9v}y&cly zg9!C89}*YDUu^nAJ%2kb6m4ovrGeybKY{`<7Yv0|d;ROGZ6)U`!maL~l2!3O)Cw#d zSQ%aC|6B2aUqBPt`q`TmakLEU{pgO1F0tP0=T17E4u-oO_l2jv!<>Rzn{{4-jpOd8 z5)W8oj^D|Pa=|_%>BXM~Q;xnu&{Om+1NuXwt)$8l`qW?WFmP~0+CO1| z37)U@bY)@kEsFGf$5mMn_BV-U!nS2HsjV zO{%Zi6dw;Sb=zaaTSL~_X}$ZVwnpM_F?us3AmOQBk4e{|^tVd&<^cympEfhcOxrRx z{-7ByD_J;OWfZD3I=SerK1}96b|%z+IZSv0+}$6paYCd8j~wy!kJkHlxg;UJNOBY(W{5R% zoZ;#6FjE~ew&IT7km7Mxw_D7Dl4xZZljBFJlFq7B*u0OeAX~7*qL~>)ZFzi#gBOxW zSvmPR>fZiIlGF`r()-3#{@!msPG7LYlmo9?HLTeTEiv45W7YUhk$J4I1#(I^IpLNNwukEP2nPJhI!<1_H`>KCh z)BCu!TUG;h%E)5RgmZChPv++>YY!SN3=f7aeE$(_&kR=>4;S+PJ!p`*z3$QZ)o^kT zKSE$FbarBm4g!y7zZ%a=xa9Odn^}VDj9~}67tD~t?~VdNYzP~mx$d+gtQIm`24g5n zOos8DQ*cBkGst^>u#DVEox{=eq6mt6$B}{Tz5@Ov&sdDq25B|Ccc|aao5)jP{(0aq z?5`4htSHg?Zywvu7_icQ))&=ujaUDEMgO-`h>0F1%5i?wFBVYXXesue5nki-%Kq`hZuJz|1eHt%wu)wU}&*(G2a*{lkvGI;iqUBNH3#P2-9F{4(D5t#;)RS+Gr7& zUZPs*%tR8dItDX&cd+*R_Qh>0Z%_L8(rgGN@#WPx@in0ZkUuTr-O+YDD{mjnCfBmI zYqErQo+A2F4wbjUjfIuIDR^r>(B2>UcWz5p3`B{b6eqd@;(`qI}{0_c0CT zU6h1`6%r;Pr#lV0I!et^5#t!g&dkTVihu)-TE)@C0&xR*bFbzjiIKBy6hInc%z;g| zKRQO2VN3QEx>0UZ_C=m$A@u#8ho9RUO+E0IBA{nOoy%}zYEb<=;Y12qb7d`>`z!JJ z!gZ6vsW51mcBFM1!-B+7gE*&%_wTPN1pZ>$Q<2+Mr%%V9N;z>?JDUP^VL*yL_ukLs z12ny`chZr@X;N&cT0)Ov?wxKMue|1$#?7jq0E7XGkmTPELJ1h&=Z6>aPtWhko6TQ` zS>)6d$T+&oYaUA5Q9v|xIcT;6*w`zvvQ|O znkcyCdR=ftEkBY_tNN$X2J=^So%Qz2Ug0o{jL_X7%$JEaRXY+>@B}n{yB5gFGRDLX z9M9Xh4m1XoPw!MyXB(5vT>ju?ko(A4(q2#ekrp!iKt4|o+{BVoA^2?z%rj&E?eHbyEu zJM!**J+{*OXPXpy7312lK$!S5XLsxqS}|W9PzdlV#qhy>qn$B7Sa)0d zed4~l*l0+Vrgt+3E~oM`6RZ7mNMm&`^$4c|b@AqU^L7A9Ii~3PWBS~)g=2@87=F5L zEVqp_(ur`(HHB9mog`tNCb9C0BX)aaZ2;jSV@Y&MiMbg~9IjxxOqp1y7u8#0zgl8` zen!aw6a4MvfiUNCWApKdL02rsG)t*p555Y^fe*dM%Sd7^>CIR3Z9m`;hu?@1>ikxsgCubE|4mOw(B=0ppM4KE>wAlVI;|EFbTCp`HU!^!UyAm7JuT zrDr{x#nho+A_qu&BuYpw^d8tV71UCm5eoXSn|Xbwm9jzfqWhtw&iwdh+1IQ)?yEb$ z6KecF223+K$8B+BBJSKs=NuvDmPnfPZk#~Vyt3(G1QVS%Zx%I}t~lidB|h8b*>JL;0-jOE$EW^-M?gY1pW3t5h*GDS@) zi~+&@hb+ce)ozDJS^X!-0j9WMty)wu-(EM7>xblKCgyWFvnKoRl^itm)--KgU?a62?K8c z9R5FKwk)b>iK|NW&%}P{)r#bs>*$|#$d7r0J(=JPH(DCRGdd40l2foWQ!ZAs0= zk9bpFA$~;n;X&3YC9cCTb^_u)8Zqu*A+5mKS@Hgp<}zs6ZNvP3^5A9!$ARY1w}wv< zHqX5k!+jak1#<>HsE|A3WvvR2!2{x%WY}VB?g4M_wXXg1$qXeQN#4wlGI&z5#2wo=diC3-vdllrOZ_idJku#BIx52q~JH|T-o01hW`3= zFS`PU6cWwT6u6lN8l=d)VvsEzgf zD!MxL(8YVqfL`g#J~jCh_oJzFmEQKMMK|G|B3)h8s&$HXFTpap9og=cR9I9k!R;9M z(Mkx=iam*(MZ3Uk1)b&vmK)1NBIE3O5*)I(k853ks*$lucoBE;cd9~ zdfWkl6JLEL=5(fn#chLseLACIBdZ|&gWUA`v zNQVQDs2=~x{zDmb zR}hsZyjVV$UG+U_yUxNXAK?%0hqx_g$?hlfCvG!)i|}GAgP-_iatUmGPrrZsGg$ML z{cpu~*is?7-@S9CQ5k^i$@P>!c5Jt7r6cGrB0=>ly3G6$NV2Ft#Y9^MTn8BmOH&2x(wG_pAOiD;7{(9#qAsttQPzS_6&AcudskL%1tH zv!>JVIdQa(vpvbG{p12YWdHrasvi0wx(DW~8Tv*mFhG-e`6Lg!$zU)R(i-MEN+QQt zy!L4cTeGS@EZ*9!Y$*2Dr+?{0d(MiPd9H68lU}HBddPR0jm2iJKk{VdSEwB8eZ!Mq zBeU|OslZxNSjUFAcUval)%7>-Fc{z{kX`?U?QOJ0}1d> ztdX_WQW2*7xy3rwujwkfo|6d8{+`=%R16Q0_Zd0wjhY{QHIX*=Z8<@ChW7g+;grC3 z!HZQ4gVUJBHX7s^!NPWVBEIUyee~LcV&m09b%Mj247ACPmZf|su~ZU|Ng zoWS56CkOPMYK9uOU@vMYs5?S_YT+`3CukzZ^RL2ZeE=qDbGoT0j4K#Xm{e)oI0Cck z#%F8G1jBUz<+&4Z-e17O`cCDTkt&oeCdj6XTet1-?Rj61XO+&u52S zht|*bR(Z4J;*zAffMFq7$I2Tt_m^;$;M8wupL~j%aXqi$pyKsMQW&dn3+17<=Gg%C z8Be%acZR!xcIk?M4FQ9il9>qu8gE5e&3n>eipq+|jRv zkR9~xoOy5TySx(@`kFDOpdVj2zVN2Odj!p1c+0P-rIVRQU55}hzXQYu?10F9<`x#Zs&RE{dkq9@DGigRd1Z^5cO*iNf7)oLGIz;|n0PJpW z83nb43?746*I&i?;3LAy_yHofqatyipJUeoI(69(fz8$Rbu^#7@2oEtEJRpEF8wlo zw0uu%JHei6)M)Nq?m}xzo+GDrcOCwKYfUP5X&xEH{wcu zQ1ds|Bc8K7^$SK6!+N`it6cfx1YoN|_^!K-hgJ)t-`7+J0$JM`*Vt{%1rSR|dp0vodL|`A%{>q5GUaHpp9(IiXQN zff^Ffd7$Ky|19;)w>*zCt0i`=rbW>FJseV2eNk6GYR%ec$eUKkvfL9b9t=^ldJ1VB_-`5CL`4G@(Id1%SKj%p?#FqEfO1P=} z`mV2`&q46hRGPB!3bImP_5PAw{-JcMnxl$TqwO@wG=81|beeo;MP>$?L(-D^WgSXs zBlk4XWFXyYz}sHehz;QO_bOr08hc+>s5)SJ*!F5(&16Z2_O#>&fV+1ua1fS1?Rn(` zD?pAP{%_p!|CC#T*eY%*yfnc0B6bqyig?}W5ZDimrEfoeW(6&)TI(pXjjn^3(o7CO zrIvI{6c7#4ZcYprVz&?&<~Vll^78kc?wG%%ZX#8~(;<1rU>VBh-objT*SxAw6;itu z;as5*Lu0j>6v;-+!-MP#7W70rjQ1D+cFAJZPah_(gN}VL4#8zZfki;=hH%~QS}AeR z39I?+ya91Fx*mpFi25tAGt{a%&Z#updSSL6eb>re==jg?hj-~b#^o>7&rr!uM}5>k z=6Eq-_}$Zql#|ryo?cnafTDreW(w$5j~JbB)8|2YhQuHpN+9uWCN&V^-Bf8@RVPam z-&wd4CUkxHK9<|-SqaN@wzP%BG~^rcxE6uP`I@V8#rjtTW@EwwBQ%-KNFLg-t?9M9 z*H(fzdy&E%Jizt-P3MQW`FvQ7cku#T>i8eFanoos!ZHy*;CpcB<1f;~3(gV)qAyW` zld{_#E!K*{E?!6`#@AV;u2D0)sTXnEo~ai@`cW?!psdjWGH5#IlJY`BO={Z*odURu z;m8&ywBIGAm5Wz~oUI?2=Up7j>w?K!mB|b&_P_`0edQ<6PBYFkeEl2cYYzquiqD?% z23aZh+p&I7YTd(wbJi(a1^9x(6-a>~fdNPHnrynkg`{F8sNE(4BqT|IR zUy;6RNMIH$RYQ!5?LwUkczT)6iqF*rp-XWHrRUCPj2F%pdOjX5-|RbM*4jSx#O1h} zqGz8y2own}FGhcEnU86h6g~OG&)sY3#-b^=B5+z2{Lv>-w6H1x!ieJ9^k(>JY zM}GM_vtJ*G^6{7{kWw9dC#$1o z6TkY`wvsLxFjk}N9*?IYbPp%#-w`0WKlN@(?6uK3ZD?~e-2@jg8ISlno5%}aOtkdE zJl)p4E872GEr592`jT;{UHz07o-ykEs}Ko8eS*ByT5N+~ghf*1j6Gs5Ko2VGRFA=E zU8mpymjO!gfT3;%tQ-yagZC@y|h&0$YtCqXm_+ehc9@p{b=vg2a_ zdRx)1Jwv&Jqj$S?M~%UGB! zx(>pTx_+2y{Ka^Kf8h99i0_RiW)JJ+VYumJiVyaBG4mnV=~$U#u-@X33=}@m(?|Yd z$`TxhAW7z{pfO0<{3YJAvQQwT88X2o`0yru6h?%?!Rbwad2VBf>EUnlZ91js3(|a7 zX?#=s)YyS`f)RE^ILyI;=b-NJNap5HU#dYXA~|-0O$&6f8TMn_mJeDzvX z9zWhY1Qx#@XkigMpsWL|VH{18! z|8%(Qt?KGa0gf&LbjY2T;qA*fw39)IZf;Y+#(N>+QcZ4DYaGRW|IPKEqEMn2PFf-( zVxG9sPsR{!**L2PHrH*xegaoM0Ci#{g=s&}C=~D9W;gS`+Y6;gH-+^Su7T->1y3sz zgCce3`RxVGsuYnXTvAk#&NVhNb<>%)63}r>t>C%~TuR%VF|Y3iH3kuEp=?FL2xM!4ttlhlao^SpPb?aYc8fXn{=m zjURHztL(;hoC&gPP=iW-yf~W?ijZn;HcWwbO&ba$M4*~3v})rQDb{z+AqoB~IpupF z>2d!N!|Kz69E;&1$+13n=1Py)-#5MAne7;8Hk46#R~3lk6*r4d(!yY)&Zdc4TyeI! zc(&#wbraYju&nRS$?A4iwc+2wgpfxA>1EVef_vepFvR1y6TCMBD_lt`!`SAfdaHGi z>AP!CeGy7tN|y4QsJDyAZ@2md&&tM$IKjd3NK^;1;%HzlnBRNzgLiQ=^H7W~C62=M zHTS4Go@75M1O^)GQfM71o(gen647y0RppLgU@u1@A8T0AcYJ6Uk(reKQuaeW)uszf z2S>f&4(E$r&I5ye5g%W8eDI&;A-d=t=Z#P9?AP#S{3W__#SmQbqB5e1fO>7HH(0 zY^t%CywJWO#yOI&vHi6pIztvS5VAhV;yLB{FpFuSg~MJFr{NIEkYHI6W1Mc+}( zk~4J}MzHI2#4%ZohAc{o(>wY1EI99c0#<#tLY`_KFlP|Hx$bg)M)QINfP@hlZi_?% zGR&dw@$!?$C-a3-4mMP{sW8VwOE5e}Sq2CtY-ai)D^>z-GsAIz&|y`~U>|Brw3ab9kx*kEd%#wQs#70HL>ivEmbcJiJ0??71G->b$P53CnQiuQyJ^CU#`Gpwi2aB z&wl-)0D+n_Wb%qefZI*Da!x4fYg$-FvmE@2gE6ryMOMv@Ph%`r=C}OGDQOj9XXY7O zUte^cNb;TK;6f{@OS#R=6JWXc6vY5v`s^5b*df)0dZ5rBl^Q&}JxiF_=f~xXC_m+s z)1AQ~x4lp03?FMC^Re(ec5pu@C|i2ES5^5}4>q8x;X}3cks_;uo4@)@^rd;fa^1|= zqA}JJvy0B;;}5;3Lk4aN7}JWSUwnsn&k2KHbTObWh)ou9$P+r%?bJ4JdE+gLwj6-* zqkgK4-$cLS>e+>=5~)yQs<*vm*eC>VB5&*s39Fj08AR2QlA7M$o~o(m-OJNTqZcae z7t^6=vcbYNzHcNla%4!MDdVvV$mj44<4D-?lGcN2D`fbH;jJe~owP`TI((xbAwoWT z8^5;&(?)Pt1pSPdUhpUmUr<+S0ctnM0Ey;g^x`D@C z0E8K3O#%5=)H(zrj2oHbXGu&gmi)6dB{71|7Ts^BY!Yu%QzLPd*S^n+MkHuoJ5Q&w z+s@B%<(N&Kn&Iz%p{xuK!an!3QSZ=LqOqS(jwAT@%xY>IGlt$fw(;nl- zl)WMXUZ`wjjlQE2I`-lN6m@_5*v5f=Y5N{77FJ+>i8g`uWslHh(x-1RM2St5Ff3Cl zTAmX3h%uJYeG#EuHaN5HEq^2safgJde)cWW3RsMxT!oOQz=sTYjwC1*E<~s0_Go8U zDV)KsE4_!xZ5D3adDd@(P5%uu<&RqwIZ(3L@Bg*TC`JdMa5PyqIi3q4->wJ{b1Kui z{=CE`FnyV@B7?{cC-X_j8TkL^4}g_jdmyXeJ$|GtSra(Y_MtA~Gy3yxwuh6RFPJ3Q z=)=ZVZT4l5v{VC4PuBX9k)1P zJ%x^>oxn}}xZ~o+hg7rYh>2_oEc9tN{!^e?Lo`;cEWK26GIe%!Zw{}hs5 z0Ahe6u#cH&eoXAhijp)xX2$kiaa(AZ!SR^F%??kevww0NphmMe4fCv(k_vQz9A>uA zS#th2c6(--ly*I}g^&pI1^*x2NruhFeFg7`{)&!`2V(`JQ@WT}w{h!gnngdN%#v^+ z^uak7z`ncsMsE0!RM(jiya!8*aq5ey#SwE-6kC>b+zbT8AVDEMgzQx{pIj6!M=Bcv zhBg)#B6`Ivy9P)hsQ&Sx8okqdS%3n${ruVD)q6#uzOEPSSrieFfRIZu?qpbW^CPkV zZ#1vN77WOO<;rMjBCyIwF!YMC)<`m4G0sX14!ar^*xfzuB~hrNYUQwA}G zPanS1qWoN6^xAEqjY!WZ&$-tV6tVvn^HM%ti&}^?ro@QWkQf05MDop_YR2qyayJ;e z99CvnK@^w3A@Zn{z-B9IAb%9&(<^L1ow?y;A>_kB`le*>x;|V_aFKB|0_q@LE{hx= z7$0pQjjx2cGvqnN_nxIfh%&O2AF$iFnwg!&#}|naVExi^^s&BGZ)nqFX%afcNKe~= zmmS?J;DFv6f$A|a6k(H4)npG&*;qJx!f$a$3eXRxn^&KOd^c=rrtE&DmPhR9P&)ge zr;D#Jpy#)QI{#qlRYC6$*?po>mUO2j!26H`*Tn`N|59W?h$TmSlvWWe{fNrvlYbwP3-~Z&exp*en4CTvvilzZnv2QoNfz> zFh4PG31vk&e=|%gjlnk`E6^zr%tvSu&xeA=6lzXKYwTU1sXSL)n)X$i8O!DQv_J$Qc#)6&n>(q_E;{tL+g}EKBm3;hY`l_C+(a(8CDD%sp}OJq>9h9 zRCe1NkJHY9pKJ81ScMD8aV{*{ClFh&)QTH3o?!P!+_8uGgbxlt*U|Q{Hn0P5Oa)gu zRjeTWNkd*Vs0JUvU5-;AlggF9x10ng-HZe03;6p#5WPfG0%0Q3F#k!M*!>bG`=#D^ zG7R#rxtg-9&;m8g>=ui!c$kF{Cr4k3(-}w{diIT;dsp}!k#X?#c#sHe=e)3IqKL0w zp|BWYYIA82e(diNTdq6A+$sL{TJ?HFm0vyK)6_>(R1G_|^T`_64hZ+Ch=_-}PQi~w zFgkp}47J+t+LaQGmw5KNR+%(3LCZS4xRR)@Z#ke2OT7_6q+z89|Ch_-eC!NdtfpoSB+Grn zEWhZe{$mmxvd$fMB<_Jnug=_0*Jnn^TVg($LDrJf-iqd!#?EW(-J?;yOw+j6GMCRT zbY{FD;@QyMnG7>l7 zTg&sy9m9pYim)TE$M?<1~Eql)RQ?FUsecX+&|UBP4}#aEven{QcKr61cbE8;3WEtD9y zBUa}EY-aZT{?;63`WOYQAN_s}yznM>VI2r@?ny;8y`hB{!D@XeF7f5ZduDhf#*InM*;ay9K8=fN{D9OUZz zjDMtqo*_8msY}TjrxVtFKlp^ENxUqJ73?h7!(a^(-^X95^*UdDFj5PMn!^adY+M=6 z!-~tsbWT(g>LIIBGG-fIgkE-n5mH_us#3b>D_jFJT`=D}nyb0m5jNlN{K;jUp;u*;D&`R}x`}?9bp=;R3EFcx3+N`1C?)o+yE+D}hn0 zyg-3Q{ZG=+$F%H-UB+SrJ}URUoX>K^zh&Dr0EzV$BXbL}i`RZ;&Fb5!Xz6`y)2S?I zI&f=I#_1<3mymr+;Lq$+J{4TohX{;m8bYQQ9y+=+5@&uN6d?6BihXx6fo-r~1_z88 zs&J#Q>>8uRC63lhwbV%gL)eRC;MV%^HV26w`0HV|sG$4mA^v>?^A4RHAwacGCtVZy zOrYu!N5s&8Jh}#n&XgRp`t5($e*K2L^YdH2#FcoVu-4^Ndg?&?d%2{K8hEVov`&xu z{)f3UwfX$D1ue};l%O5Bq;6(IBfrOI;A)dVr44Le8$Ouv)qChSjvzP$nGQ(S)6kdr zLgQxL6~&TYQiKjNv@eJmh$y$U+9!W;VQr63s^gywQwbgl;tzaM9h0$4h2eky5f084 zDGV>GTY`FIxqRr}hEHa*rTGZWYfOdRD=_L7I8|1_|Muv%fA6KgjYlbDXHzXnLZtVORFKes zSvR&qWR*+!Krk)Ir)&CMI@uVchFRS>cM@^TKhz z)#Ffl#^&C_+Ej*|K;TWZ=&%?67e_e8C40ZJTcjh-~m_5P|Qr)7xU0Z>EsbA-; zf7qrcbj?afDgsTH?fRM)JCU5Kn(C;>eGjEjvfl5Ww0O!)cU+A7;@o}~XVCoYb0p&< zOD5=vJ=SIL1#f1&55jkn=l~P|@c~j))jcMJ4G5)Z>%KGyDH983oi-5|{#^T8wiXA! z9?30D=s%(cKDsVi^+ecwLr-f&CcBd_ICnNZM7!nae&M@6ldN%w(*z)WhcAcQicM$7 zB5}DEMN%qI4Un^{+gW9;MaP{#)|*p47x8#}0?3lbpV5X0w!$>I9|HtWkMe;~wngT) ze5$cm3T>fVgKNzhhq4`~C z`!Iw%V!hyEDEz>b2y=uNVMD!uQ{vaTCtW8sOPl(4)gM;is$4O2a2=NJ>GXenP;&tc z5of#pdadu|v05eIgO1x1c-Afcaq21?M>mK#a?R&oYpxal#>t~A(XP>a^r~HvMAa@= z6oh2JxbD%S<7w%lrx z0y3QR{RwX{<6l896Ozmi0(GmhHkQL8_&f@9LP*dj=hgDJr#6F$dQL4% zvG&5xwwz~Cu_XfJgoo~Hp6%*d@d;asjqK!}M;qBbf944dl@btoyY?@3ATON+USzKj z16p$5gvW=OsAWgvl^(NY*xFub%e@ox*wvXO>;e&j3VajF#8kU|MM`V^z{DD#fWO@$ z^0)m$jic-*!Hj`(PD*3Kz^m8DZboalz(8B))(yj;s~Xp(#npP&wVUO@!qUxy55YT2 z!XvL9#Mi(16g~X%k9{fJ#!#*E5KnqNd&)eFZIB&_WQ4Wc%@!&vJDr}%qJ1XM!(G`Y zF11~3p1@6IIk;4W95l(N!Uknb$t7EGNdxwhcn>^R7n{M34%nUpZ}G;UxKoKx;Q!SCYxcfps|KJz9u|7v>Hu3_t{2vaY4qpa@#~oveXD#t?$DIZs{=?zUx*x`g;l<&N6Foxl*996FQN&MAK=y zfZH;wfbhSHr_P?@Igu+Q^0+_YwvKK%06zvtA^45qPkNh;c)VqGw^Ywv7(dl#qqOJB z2gyy8SvxM0^&7v+WlqV}U=5SXLhnyby~N3UuSo08Z+b~~*M2BJABwOUkUzJ7|F$8w zaN@dsD{=|GyJY4&aW z1Z)a~?2fg6%;sz@^9D>IuNzzbK7H*>ORMM1AYyXe9c$~jgPJ#mA2=@8d1iparl0Fhf?8wO7*)l^9@;Ay~~3g*yAzxa$0z`_3=Iv`m3d^ zTRwQOST6xEN`5jJZ`<-^=_bu-bIZ!J)H(em9opfssnc>`Z9dBOhV!`c%r;eYz@qeD!h@pO0`YLQ%o!6r>O zUbDVoo!CN6)i7wJyRVr-PA>mkt-?j|Gv;@?J7H_0_swL9{m@Q@t8=o`zFL{(go^L1 z{rvmeBvD;CN8<@EV*N=`Wy|=4H=WMOPkYS?YnAhFP+|DM%P-2Fa=+zp`lMHOaA>}| z{vBX%dEU*O1KrdA9;hEc047+ z%p5tngk_;W_#YgdY2PYoDtGxtJJ{f~L_zZt=B3=12yo zDDiRZ`tx+b@1J;Sf{yws4stas;!75XJOg^CUY0TiDQabNV;$0c5oGl4ebPv2Bs|{K z7GvoNth_$dWi*{`r)dp=>sH0Pa8Xacy`PQFEuPoL-1$E~R{Y~gN!KVr(L5s2P&|RbjSTYefH%!S*=9kaW=LR3W(6o(^%Z3J>eP0$I>q+pCz*{opQ5!dS1x#ad`ERwOqs;`Nn-79 zB2T~?&8+8^{7uUb_5Stket{VUBO~klw!GrqZyB=IIfDe@3~tk6eG3i(-ek38US_=m zrEspySF-O~^HL@=hzAaprb-(|9>t7dFq2=|qnok8qc?Yo6wGQKNQ!C%toD#@k`lR>M3gNjhJu*WTBScg!2XTyms;4)T&U(%QJsh#>(6 zvR^}O!N21gmg!r z?6yCaT`wbD8w;Mx+)qt;*$N*X_IOVTPJP^VQ!nj?W5WL2>{n({yOK};vSPGrRZ8dX z+FkKPdU(1cAbF5-P^po!`k!C-|Js6y@z;+FO&lgq@0|G8$fbY%B;xiya9{dcm9v$C zv9cRADZCFkJXqP~pZHRJr<4J<=*Pp|8q|nI#b2zQ*jiqgQhfH!5vNTW`4xN8`1-Qk zF}MwBaW33B5Pz1O!E- zNKdFzMSAZ61f?dF&|B!ChZ2Yoz`NY%`_6HX^X>cpj&b?L2w|1^&UemdKJ%Gt?fmT3 zMVPC}(=cmnboL|uFMgR8f=c1EW+&+>c=WoD2YTswak)GV6u)P%|5bTFE%S`UV zitPvpJa~YN8M(l1YN7(?>>3*zceh`@q@P`Uz%n^Ek6IYzetldSl~e#RH_V0@z1UZXSkvtKa+cxkY0v?(#ap6ygLtohZhRz`wO$m>jUYemYWV= z>6-~?MfSQfe5@>aU_zMpn*9Oi8^(A8h9+1-A)+alQ*yI}(z35C**H~CX(?eypNhY^xKoP!Dt*BYABnOTjKt3664BaUZ=-@c$=+074Q?>TT z&&58$ho`xmuBQ3JH5g=U&GVj0N299CXG?N5ft0vA?*93K{b8-DVeU;uCl@K(pn+F5 z8k5~3Hc|wIe0+S)H(yUe;64324336|*l*wMxtW<^d3Hf6^3k8h(JbW#;CCAGpMWI$ zNf*={lPPoGOr#QpHN+U-<>O4NsTUz-N`vNo-OLGmy+cuT|67n>eMzwgLfVD5b83nO zj9FVtesxoI)0d2j^Jvj>`TBoG9)NM)XSIia<)LMEjN=P)7VPLqnWd|TTpSw)uyDxD zk%P?l7Gf@C90yN~O!F{ghGZRw;W{yLcgxF6*-5EiPCmtLOWTH75_4A6n$J5nI`2QO zKl|b{CslGzzP7T`+E6={%;n~l$5Er zl?(*Axqz(H*ZD(<{XZq7zdpMb46YmJM0(6~6!m4?wTTkWUkUKPp&{)YrDHmR!H{v+ zLfg$fJgTj%ysa!Y^z0)Cx?zWuQ$E?o<1S}v$Gh^jYSwkC@O(q2b)*|^uC)(D6ERcF zF@Z=h5Gm*(5kV=@&W>nE!SBrmFT|Zge#iS@49n8oO^hE@vta%r?A^{6?gTDBE-#l? zmMU^^&#P1%(Sgt8Rc!azGXgG+4n;%jHn zKU-rcmPqa0&8W7c9An`}(4#O=4+~nH^jJ>Kb%>k1NW?zbTu;S(s9)Ufo5V0H{9>lY zw>SW#_SO_O%a&rlwK7F-`UxX{TJe$b*_ufg3u3T8VYNTGFIt>e2n0k{q+x7+K!Aa>%6bKcV!?M9{;e>{9#l+<^7&|Fs$u#Dpu2ONXYy8&el!_PHxxT zjq<`SWZxeZqyjq`3S2rQ&hctKW{y)-Cg`XiO4^=kM#Vs->>kt05=QVd9L#q087dNj zK*mN_t)ITT%I&7_z5P?wr$K)!%zO%WGnS7^}kXQtkyXeti!NB~W z5YmHRK)hR66vzMyZYtgHyw&!C_1zJ~<)4G)*(E01V&(l~|2Y}NwGS;lik~vHuKE$+ zK16;D2{HE5`?;_XW8y!%`p%EQe7EO5;go^F&SjXwwaqW&P`j+~O0gaJy@>pYrfPlW z9cJcHwu#+V(rp=C|4CGKSspx@uD00Ymq*`+y3~XGPf&41t^VR;>jy5Ft!IQq)7_S~ z4kCAu1tl>D6$tFw0AhJW=3KR|rYoo<-7CT$OOjNNnPqZL&UTZPM2G5)YG++z-Yz7~ zoKyNoz|WT6zHN~39i_*Sf}~hkK5KrWaqkY?it4nQLJD5SU~E*6sak%R>oPu2l(Dv9 z|D<4}x}w3hB104Zz|r1*)BaQTMt)~R)8N)L9_gW+3SQ>Ynu1Y^Xl6UiP7cqj&asN& zC)Y5{hIS4D5B$!$Hr{VB%T{SzA*HTD#=c-q;hQ0FzE1xhNyJz9HBi*PzaLK5Mt2Ut% zeHph&z>rx|K2FBqH-Sjwu#5nyrA1yD2PLb(U1>E&s&WjpqK8AC17QfMy--|GE$|&w7!UvtC6L=R}9D)=v#X7wXbtv5_mVNFfx&#dXHTv(i zjAdM`L2+|(WOJV~FU!iI5&FJAjHj|mo*rs!XgS%l)S1cCVoNA0hnL6V=iGR=uM7ke zUXcpzf^JMPa~O>n6cl&+dVqS@KXvXnHbo1ZEm96ExY3pvYLK1z&gp6 zb9(WEWe$TWW6y9wAlZBX^2mD!Sff=yp9wbX@fBi+TTyVIjS3M2r`J+zZDe~>x)b%wBB4HM-=qDb+fb@ z|5#Mz)a^XX+iiA!%s~4hgG;wiBmT9by5ru#K;xl1c9Ptn**T27YQ%4tTd$bpRNhh5R|c!`2p$h`h75x=)q zY1xab{glTIt#y9Uyv;$C@cS}erzI=U3qtjQ{T(f&$8M&yOG5DZAPQ>kQS69n>?RTt z==pZy5xu(SoZ&I&J9i-^(9v<>w=(7OPs;-i;u7pjtn-AD{Ja=M{VF~iZ2M_B+zzde z=6#TSlCWugexWxN*Q&;;iT5#Rb}J4zS?!*m+TQEAo`29D=%c0wLenIQXtHg;C%Q|eJCLqVV9fXdg^lvIIx`)CmV`So7mRemN$ z#+{0b6ldYPtY@bt@oBWo10%eRNE^fyTPjJX-?U>puZDbp69ahN+{r29Ys?GG}{{{34X!EC^r`R-&oJCD+`1N{sK2!u>{h zuX&o#)&S_P_^J?x@>7CU<((&#Z>QZJRjU<9daKjJ0glAbfy?pK{@r>k()m@Bj&t*g z654Jcb{ex_iSku36*UMmBsrSg*Ih%yEBckC(ck#nd)UCoue^sOGXG;m{a>I)0jRI| zxRb*VL8Lnnkbk&Xuo842{gsunej#12KwvPp)6S}5hVMhg?c~*o-eCW2qnuTR&WG?s z{06HmXFbY0A@K%1l+DDPe@B0a{OTIyXx0nQcd*Wmn_O zqr9pG5j71uP3X3oO?`&2!VUK(BQ;<{?VqJ}#YsHz$ZtfXhop3q18s<6wFF(8(7@W3 zb%0pwr-tF*_X;yJ)(j#J7-hzI`{aKmupUbYe%WFcbqzUN7=CBHvWKwNuVsA#KVy>2 z&qc|b{744FcX~v}+_a2fjzzuvt#`KT90ZTmW+sQ(cTRHfWve=hM!aa_bzvC{ZyadE zn?31dN5&^KW>_V_>nK0F6ls>ew6y1<^ho4@=rYmLUDu808SAliIXU2CL{>vdY7$#q zJHh-RtimSI!;7I8wL|D5J0K}u+IFaJ@Fg7xZCfAO*) zVrH?DJbJUpmt0i}-TslOxcH?f^DpWs|GOl~CMBRNa9p*ct?BF8n>vVR%sv#0W=@wt z-IotOYhaOWXy-z;Jv4%DBBlp~DDCaqi<&jVLb)^r4+f?@;%%x%YFcbxFS6|)hu?C_ zfIch8m)a6AtLxu9_&l>~axtMT<4{fa@K&LTvJa-hICqe z8~mk;&^L+GEkyKKCwwlHgD&3W&mC!InXaGbrqw~Rl_tMov4A0|#V$7=OMLYizfJ9+ zuwLVvQC;9!R5#3N`<3H$njzE*-ZUnB&YD@feAJfA+xFsYIKv%DN!`5sn3P{!baH3> zXw}Ll%}JnchE)ocBr~x$Hg%YE{KlbfDJj!K|6p3yM9EUm%PR$LKRs|FJ83MgQSvx5 zvi70u7zjXK4fX`5q$Q=LW3V6(k79Bv(^}2N@ayX$modJmk$5a_8Vg|(81*NB!}&m~ zPMivp{^cD|T_XkcQ=*H^!_4GqWvSD7Nel-Z-s;AE_tZgz?tko$AySIIPkSCYDyqmY z{yB5|L^Ssbq2K&NR2pCi9PJ$na$_WF>RjaVjEaM3dLBlItv48l8-X z1HHXT=awUR*`GD12M+7DJ$o5-QZ<@8>VDi;a1~U#4^sR`2$d~QKDn?^!F%$?D7yE9 zI&5KKR~Ems2$k}Ed`6W9Rv8py@SZOKiw3*DGt%hMJ>&G)D^g|2x2QfdIG{aLRF?%~ zM`t*W$^P5^|NBgSJlP!`mHFU>IFi_af>^wj9bdsGzF(m7CCtnIaCE!z%Z%f(@a;eNr1$5_o$;9-Ydrj`v7a_2NHsUia!0nj(pBt`w-- zJu48cuZfh83=QMb%J>8(yz&I!#l(>B@_qL$P#~m>OtBCarzBbfc~kj; zXqh$EIvL$&$f=a_UK9KSl?epib0dXn((WXX+DW96k)f!8s1lVxQr7uC^-^0SXD1>> zT*okM+%YE%c8|I~)ugBY3V({d_e=Mx$Al4&q!9`C=(0;}<8ModG5w>>!1FhIu=)M3 zkeF;#`G#9fO^xS1+8f(%^Yp4F4^}cYfx-Cki)!*-W(~MU=7PEXZT;gb|9r696t>*M zmHAT^7H>uAfXaWM<y56e z|IuETiD*<52*c(VvKbJ2cv@S+sT)K&6nTyY!)t3G$zT}I>HP0-v&5O*=AqT>S){kQ zxtU%lb2qMvARWi;Z=3L@dwRRB9zrTd%CY~tY;Cq3WW ztM30FM>e@qW7=XK8g|+!dF{?9X7MP6>F3#OzVR=}=f=imyw}}buRapenE@9|OH19m z&}vPxyp|Ag^Y{bleyY1e{O!JjMZlT_P^pzHfBI;Y@WW=qO2*L*1)5i1((1eNKb+9> z$JwQo2GbhGLuoRvP)lh47i!FN7nQW!Ow8M??_yh@{9jfBcEeU(J7FypyyR2WSyjy; z+x$g996a$jTbaO)v$HH0q!A0mdwOR_M^Wx(&ZLxxf3FMH9~?8nd7epd>3_$(1)eK6 zZc-(dl#~=V5NDW4wI1I8@y~wuFVMqtB~Qa(xTYp%6UM;65TB5c@u52^@G;wKVn#;B zyrQBhXR|uhT-=E;jknR1Akh*>yDzJN>Nu?cSUqKP;48t?I|qk{OSOjM$XFs^?ec10 z4t}=wTJybh={|w#lSePadItwF%KnHlKjL&8O{tKsPP>2PM8xaAT|V1QaDNI@ek1VY zbm)6x8C_0^_3uX~0RfTGtJ_w$iI(mUjf`xQQ-L6t0xj}YyWR8+=m^R#p;2GTjEI--`>NJ7Kzwxo1jPkQ+$d9Bb!W9#hf z78VxSu*_|sH_CKh&6;0YLpUVh*LqgRz z336BGALLc2(unw)^WC_2bFaBnG~xt#soDR>v4KN^^XOF|3)#4+Vjqc(*lW@u3L|83 zkCH^+eV~lay@U>A7BYBgOhkabmleJg+N*X=L7`zdB;oH4zh0zp*R)U-IE=W*kn}7v z?8gBa+qL3L^syF67O)pI3oY9Dxezt{lGS@%u*mW2G%N-PgZ|((M+8LzU_Qw(TNbSH3c=#97NfDMj zyX^S4|K53j?Idi0LQZg&8f1a(a>u+(w1GH5C0Sf4O%%;x95(KmPj+c}y~=3%x!=f-#R?$!ccE(>FI&)Ix)_KDzshe={ryd$lAGy zfq_I{hYkAfn2Z*808eIp`sB9yz(SR1>0Wjg*wD!69y>b+=GTV=n&L>#xQq<$=xEKQ za~fKwpo08@kc&VEJG&&T=S5xq+I3L*pRSr+PmINTITV`;xX?zMaTuzLT<=#nFM2W~ zTu1u1#s3c!{1Q*Q)2iF5mG%F`{Qvmdy|POk9bK|wjtqbL!heW}ZggxCxAe92{yWP4 zhwlJqP*}Qhjag&XvNQS5IP!11|M5T9^e=~o{r?O6-;479E%?7}@-P4Sf4~R#zI!>j z8BXof7Ur!zchYcBSEmKov4E6AF;USzfC+T}sS^G|&m``8XTQ~U>a*u!j_>mdY1wyZ z;uua#6^xA1(&hZMVs1-wcgJSMMf_00<3-$`KoXOv=**OqG{nXprVLRaWV|4+-LzvYMA4 zOvNpop89cf^T)@>(?FzrTdZuX!v|PS?9|nxm_@cma26&us&1dpc}4A4Tzj#2c7dM6 zyqmVd!5&9}yQwEkWoR9@XP?id%eq_HQQdwA5C`zF>v`d8{R_IuLO$2c;R`=%&}guaEG)xh6^J~^9fWs_2!Gl~N} zs6Rb9+J2dMZCbr@{o&rHPW9&6@clEn(X@%1$bs zU&5O$zXd{Q(}0(ThKADcLU82o4$;;7Js8pG>ttA;%amW*n*5?0uf37u{hkdbTw$V~ zMXoZf%wU45VM+R8DAGkC=D>G_g^L##u_Ol$sWcohqqh_U)4je1( zw=~tCq!T>&Knx_2)%E>^s-WeYQ+1a0HFYGg`SYKosNDk*DDLUuBjV9Nr| z@qrG}M!Tu__h}7*=-a**`4h7Oy}rJU!a0tq(HHi{HDwo7+w;w4fSE>1n=$M#Hri_{ z*TF2!$9 z5A|)Cp>q~_|FY*=nNobnLQT%7iI*1H0>tBGGv*=|&3GsS!$#Hi>C+|Z1h?(z+?4{0?TiU+M0La%uUMY7xs&} zGQChf=RsK?Z$F#`k=c!x%**6wrC^;KzIB8>77s_O@?@ib&}z4_)p{LyQFwT<%y}WKaTEf8@ofo{l);oc&{Cx> zrpxF2hi}OzVIWvf8R};e{1!F{k|FO?w?0=^S@R~)X3046X`1h!>7DI55n!)ATJ2Wz zB}LNhwe?!qXT=Xr-!nKpTAw`Fx0McDOLv^BR9_p+u!iB3%GhoWViW`2oP-WCC+A>N zzs4p=`Gwarwo!r%PlutV>F-CeRMTiPzquy?Guy4(X7f9t;5RP^;s zxlVRz>?HDa(oxlk4+0q8!xrk~eR>8s8$GP*6vNlCVL1rhl zgm4Ea-_MhjwC%jhga@m?`-~B(?r)}RJS#|8Yis5*14*rG8!Nm3$=iD%(DE0GQUXu z3A>qG+PG-DW66z(AIW2E%`588m13DR75p0|eU3UEi%$Y-thcm$PsYUIr1K!(5h8|* zZ7vj-LMlOF()`)+G!terZ5i7S2jo9I6ay)@3hg#}e=7)Q(>;{L(mdD(6N$!nzp3n; z(3>T&({=Bxkgpm0uW$J@VuDY7&h}BVE@5vmEx#gLU#~+U$h)N<_T6p_VI=U>%$xg+ z%O}7-%kO>{M)wDf2~l7rOD`^|68a2C2)B#jSt7c6ylryp(;xH@z8gv6Pudh6I0_yX zNX^Uo%-?j|S~?2+QeMW#X|=Y1XAK1LkJj~=dvKj)6g$@i#@p3BYhZ0nkbCZr3)9|y zN}Uljm{I@1^oy%Sq+j7-UgWZdygipA>s!C;)+_H9X`BK-nZDa6%Zm?5E*QzqKw{8G z>NlseO<=uEUO~Y%EtT2d>xYec#|)ksrFZ$mcTbMq9ryRTVaFF^FVTzLh#<0ORG&)iJ@a~8V=>J zT*#y*fmwD{colA|8ilGy&E{dsV`V(ebj)=kYtlE2c6LEITZ5n&W zgKwpUoW@?9p?tp=@%aD&)U^2xK&*Szd!nOmY^+fnm5R?Tz`B0GjpkmHTw54OkerW&n%)Bv~r?@Tw8|rD7T9xZ32}rGtMxfSzh-}uc#+sPp3trG+bytbJVxbpRan8Vt-^lrC57A$51-t zhOEl>h9*~zRb4DXkDC{2tS1Dhj3UhX=d_sR#}NevwubKE<9LEogkXOQ^_$>>Z&o$R zGx2YNe#bdZxagnEh049lS*o{@X_c%Zwe7GB59)JUUY zAv@^NSaVG6DjH2fQnnCGDQqh*HC><$x#33Hsrf6+a91a7Jo)Ij{HEI-PPVvj1oITV z>bWgK-cKuD>nc|SLb=gQMZSf;X3`^PAC(vE$+ktTXKLg*o&1DMwqoD?D{Gs8?Gt=g zYf4HKh7i8Hfurs13y%o5zkK}~HLvRZ3eQ*~;rES-T7ZeM5+lT?FE#Z-cgVXpF<|`p zV9g5Mm$m={{<E&+de+r9Zp>!`nZTT=)LS$8Z8T_C4Ot>>hUPo{_sf$GCuGLdV)!@RGP&|+wq z76bG59o4mkjswNr{ox>z+*-Q?LgW#fJr6e}_<|>!{s#4JDGMJnWI^-kT;!T(yA_o^ za^UU$Gqq7WZZkgJ0!@*dYRF4WmB?}WU1r0aa`k%2qje;5-ZImdV>Hqqko&%w zy8owRyblXWX7e&sPqkMEkll)L|3;Les=aH=A36EGevbL}e!QhE_hCw?fQFJ$j2%$exBr08 z%7_1`N@ebyfrKZ?m*Aa2hhROls58EC=zdWrQC8u_!*eNCGX}y|~ zYjW?z-6t?yJ&ig_+g@9NP@qH#wPPCW(s9@AQin7z!`GTT#^80q>O1~>8)&neJJjPd zSdXE1Spm+}sMBgVc!= z6TV-k@mIXpaG4a9XRt0#|fOQa!Ku zYOpkpa`WtBW1ul}!T+8j=4qb2_P9&l2}7)c8M!T2JD@O`N-+}j(HrG`1>hGcp7Uid z&Q$jSlqid({Q22~98Bxr*##X$FzDm#n{x{WX}`Qyx6QpzmSQ147bbRTcBuJlUiv%N zt#7tU{H(qv4!kRy6M3l7W_IPu!4r(J%yTzmL}qzne2xOA{d$mX!2;Ar-c#AE<<_iE zKCYlK{kTNIXHAf>aei=g?F>n?TGMsY^}_oID@}`5vtYHUqxzEwD$~fG^;Gn!l+u9| zm3Z3fj;2hsXNL~9f+7G7 z&qU$?qZ3{LY3EmY+-qZ+2&i>7kl6+21+_IblWz0bibEQiT#bz^5Y2hQ1xWZKqSNs9 z)#HV%T}N;4WVqUJ$()Hl)npWs>T$+Kif#XR#)0;);|FPTTOWOHw@&fQQ-mY1L){C_ z%=aHKZgw6Pb1hg=+H+0-GLIZ*GIZdJ+nlE0_La!xRamorqxx-$H?M+@W^7h5zfR-< z^@vdy`mnc?diDW(XJXAzc$0^gJ%e2!s=j}WhPmdN>KQ8M7*`BPXnc>39_9qsE6!X; zW~~YYAJ&YHk2NPec8RBudnY9)&`k06>K;Uj=w3(GkHg<~3}tovdv~UZFPqNIR0vg+ zQFoBcl+lsh4OsDtskZWCqoF9@Oe_}a_syEho~@#($&e=|BW^~IH(}AbQ>OXBb0=ZE zsY$DnXqxKGX8nAUo2618P{r;BL`SnsMFe@^JoVw2t<*xH#|0MpTjo*(MQjT@X)66E z$!S4lH3B(PB89!bVIJfQvm^lj1J0(ThCS`!*gydBt`7cMc~=lnUKpGELm}cbwCh0n z3w=YzO>Uz}2!fp;ki&J+q848d8hKm!DA?wklOiEHP~zr~BL#CjLiS)BirZm~NW9TI z<$-o;t9bbmyv={UP0J_J@2_PqW}pvunU4~PN)B_p&d0&R&s16LB`7OIRA%H?`ddcw z0{Y?7yBOxA;FH*H>v3$+^XJcbU%q^K+`0n*m}9N38D~>Yhk5*@^P6>)BS!Lu6yI#- zsZ3VUiT`b_`;|Yw|7B+j7Na_nKmFB@KYOk*z5MEn8);I$p&wbLrMA_U0ep<#+L(HH z6qDb%tA~okT3hwN`^#g0ej~hc-~=fR#;?-m&_Arvrj=E}&HJV`e>|oDpagqev7SE4 zt=qGKCenm&(39x-GoYuYUY5c2z{P~0A=BOKR9d3LYF`ewR9s~_Ave>8CS@&8dI^#% zTt%-+hSBt6%v;euf`ip1%`_y}8m2k$=!F-26h6&$l zc|(r1Rm>}BZJtLD4I)uND_aK#4_x|&m$~54J}E2JUQD6KkJPd9r*Klop()5h>#%aS zuROiaFN|+D+Rf%%hBH|~rVW6u_r@!&9sUFp*){~-+UfDYj8XlAUO`l4>jX@3g1H>P z5rF`X==%`yEyBNyzNCe88AuV2syB$Q7m6m=Q+H-VWVBnrk!NQCgEvU7ubUBs$pjL@ zp+Zl0Up3ynFSu@N8~NQx6I3o|KhrE9*Sn=5(1Rl)l?BV&_^s#SaHGwR>j5&#K*h(y z(~?1%;(anQ?=gY*;TQ=`$9=N8RX%n-y?p_wN%~AU1;1eGpd4Q~crTU;<>MVR+jWf8 z)r(uOeN1bh=+@&#TwB1PelvZ6g)i%O#lCpb~^Vcj;~QDHWx= zRG_v|aVMwSd-T09UhAVsS|1aw+F=-iEz;F^^p{KE^2Ti>qr_(vK~C9M0d<%osqV10Cf z&PgHgu1`KjJyMW=13S`Uv(yQbtPd};Dfh{nb0UaP)QIevKm|;I8p?w3vPKtEMslNg zNI|hj6#|J##~H`GPJnQLiLsoS#2z+NSt@D1$J+)z@yX>!bqgL}mR)LnZeuDuGV==y zA@TWERDMo0UJ`$?a5P6`)h2(l49n$cE?Gd@w}kywZ@Xv=k+;zsecAQfFn%8o&!wg* za6kZj4x-{w@zzVWoqFJ;sR^&E`5ZWMVdg`n;^b7?{6)NPE5_;)IOy&IB~_*z(ydZ( zWS3fufpl7QX2rrQTH46eIfkjE{D|mcmfj=b2$Lr0)|mPLOp_gL(<8DhMM&8B%2$@;}%_c&n;&66AI_fc_|vNf5U&z8>dz9rr*12O9w)`z>*(F1{h`9+#xA% zQozk~-3ctUuf>%nUXVWzd|N6k=bEnQ`89FWos=Xu4XDIF2VA+(cpL1u(jJU=6sdk1Ok|RtvRLcf?B0PKY z5O~AU0{Hl3$|ETUoUBs%-Kv+lOajE^aK3h#@XB3y-y6UV3o3ZYB7t@Ii`cr$>9}Ff z)SJa2$A#OXlo~!`oF7@Y5bkRld-&sp?7_#%vnSsu6AumUGpqN-BGcIVsEHBL@k@_; zo!a&hQ*Icvg^iqz;d<_IcXrlbt#1lNMl3p#;~Id==PeP9@vpj5OaGj51eX!l@XSXQ z%p@zmBb>Xs;gkN3wL!los=3S%Kx9WpwcpEdCB?FBYDn$$&4&I)&-H8K>b+CxK>Csn zt8&oZ%fRjuiDc6YQ6-Y{e*pMb`e+j5?~1oG;)Z%Aw|=41FR)?`xY>4R_C!tEAZ6b> z^(%WhL=tY|iJK?fbzKjEEdqr3^f0ne>qeV^T`|kMUB#aoZkTZd8po$mM_A2dG%zBq-us5CJy=GT(~k{I*!5 zy4{ZrQOdDdiJO}WN(;bfUdCD)y6Ah0C3;s2nQ>j(hj`xe=%humI%8PouIo!;dw#D( ztu3uKgLIt!QAK`NR4k+9>|AWv>bu!R6+$3IcydMGGF_wE3Lzn}(aN&uwktjBR3`9g zBZgaL+IstZuT)@7%Y|>Hbj4NPYNE>I%2F^nx1Or-EuP!^_z_#t@hC%L5NohrqY^xL z6qV>9l8M43OoF17;1c_=-_MC=&z)=C7E;=~}wc$s;Y@zk5W=8pw&} zAq+J0WvEGCpTPvI$25&&qv)ndD4E%4e?-^i_G@}C0F$wA5qE)k(`I^3Qc6a)dobgqfJM@iBn*p zE(Pc<{YIFM59vjqP-qt#HxvKaueasp4&ZKRi#d5tvg&nqh}oFNWMJize23_vpsq@@EG)UdRn0ZQ$c5CNeQo{-6`!n%pha)sY5lpF0}759 z?C4aU%`QKV{v4id2_N22@utEVSozfA-FZB z{yi33lZ`eU!OZ}p%vUsqWS=7`hPCAmbI6ay-(@^4*PgTCzlRX94zwe?_n{qizAP|X z)fjw`i}?-Z=B7mvcxN{uk6U2MwE8KWc`Nvv-3jNH!j6=_uM}&XQ899_Yk5y6b{l1v z(6@HvkEc*=G!H4B5k4{0mhUDCwCxqBz$i4HXq-okhFg7z*oEBo;Y*2kKoplvb2+pn zKEwnvyNSt)A2IVViHR=BXuQt~3354Y`}GquXS#^GDF5Xg*7})ZGa9;o3dAp(ul*i> z2747zFyX55ig=s<#RoBmfiKq&Ix(^|-cvwDxEbXTMV=t}C4OoxX035zAQ5QawiHK_glY|70F76e;|AuHmwFe4J)#BR zDo{j9sqLusXY8v0MqgX9$e0U6>k+`&UU|>e{vJ=5M$OoY8D{y8_FO;q(Gdx0)zq!g zpKE%&hREDyfeRAozf|a4G>jRPM5PEAEhUded38!)ECCgf3PGpl5D2l&@Z8AA0VVGR zQ__9*C`=X+Xu=*roj}?al)~HS9W**9w6Vs2^?cxk!+d{jdY&H06&j=k6}C&4Iap{V zB4J6q@rG9zf}cqJa4RV(=~zk9MbN;f9N=J#UDP6jb~0baTlW=ojVPrF$WPS?kBm-28)9qk>*q1K5`xh1{}7JAMmxkcAMlK{);i*IDY;5&$&=;}Fhl&28gyPV7?@=9 zc(xo$u9v7<+35OM@e+V!umq#e4x4j*ic%m~c_hGK8Q(>)`0F-Z$Kdl@mw+GDr&>?S9-D{$A8V6*%K!iX literal 0 HcmV?d00001 diff --git a/ALCF/notes/assets/cooldown-val-loss.png b/ALCF/notes/assets/cooldown-val-loss.png new file mode 100644 index 0000000000000000000000000000000000000000..a81af67ebad08896b6f9bf451e9b979d0cada2e7 GIT binary patch literal 50028 zcmeFacT`hb)GvxCVn;uMrRd>Ae#X1(BvA1ZhDKQL6Ns1f+uy zdWX;%h#6X0aO*#|GBDzHwy*};$;c!?&er;d zW=8g`21drFmXhqaA{0BT>0?QDbwNd5MQa%&6Vp4cwnoaXN-Bn~W`?4V*+EiAB%H;7 z4lIo9^;w-Q%q{K2oF&=+>Q@YS-g?c$&idCU_GXgo8jAN>WvpzCScSNGxp~>8j<8DD zK7Jzh@Rsa9KL-AiWH+(5w-)2!adL9vcH-x@vNh)66BQNZ;l0jt{W=%$1(%(RrMsFuo238LClI-k2JL^B4+nYZ5*}J9PKg$OQ!L!xE z!^h3b^XunE&Zc{OzSZ*c^Q|9?X*@Ktvod$s8eWo}Pe6c&@2{@@YSa85+I}{vTABj$ zz5dtFwgwQpZ|ZDhu6fJU!pPF@uWm(!1^(5l@k=Y8fH3dRW>q^QTNx`0;I}}h%7(T^ zKqn?v_Ut@cH0Sw;=m7V4cGw5}u_FLRmH^jn0rswg2ni%beYzGeAlD;>g_(qBB3R^Z=vp?%qw!<72j{TU3~I{^ZRMll`nzP4KGFaE4<>UUL10fagTb#a%Nu! zTlt4~#b2t~B8L=gUNs*aqe?wPJ9nD3-8*0H0#y)8rYLtLv#W?{J}!+>MOR$>^O7F)^C9drxF7y z$PGT+IeqR+bJf-BMr<}RD@6;VWf_K5Srx@!*xbLeinA$%a!$W}2ATi`P)t$Z74a(* zV_*AJJ*e>_F7}b?w20<~y8-I+`7zn7yh?&Jc_CvJ$GV2S;)?HIv^r?>yqMS1*<630 zPSjP8Vim2EA+Zb8>9vEtRbMts@m_aIC~iEXIP>kvGb}6RD{iIO zbY5^;pjS@ieF0GnbQQucyHL-BrI!TZlc$gBGw^kv-k=eG6yNBMz_Thpg#Uj(kQmEO5<+V>MH_6`mEjWg9ckNF-3rN}0|Nxj3A$s!sM zCEdUlU9C7G)$G0LS1iY;!qj#n@II%#M6Il-$$^MxN;k3^=}k^)J?EC=Y&vHz-2Ey> z8Xd3R#Jwci9lTHBxS_%4@cosQ`5FRYuZ;|@n@g$rs6Gcq_^) z%bGI(dD+}8ooDEh@!`6;tHm@{<={cs{Dw0N@q_evy3^B#`wuYdpJwyt4Y_~s3+ps1 zBYTt&`l0Zk;M+ToPpA6D%xNmHjh!4n2|Fcx3V%A7THKGyk8t7iUHD`0!o}MmtDndp zUNPNoDvYJWp8guS6qX_@#8&i1ruysS!v*0wH~Mc&-9z5;yh)U9`lRu&_)60mHX|V; z3M27^m$qKbifhtsaazhAUk|r3PO?9}nB(Od;p)47Pf@X6p$}z@wVV(j0*I2d{c_azHAk(eHNmrL5}YngICnNKMsB=Z8`b$ zTH}{?lh*uC^Usw&YB!56eO+RAr68~p$Pwxxai$41Z6jt49VzV+DK5HM5dV*(DQ8nd z({mKeVC|x*iQN}0IADjl&*!YaNmVK&BC@qIs2k;x|Y{9hGTr{yi~mDe3)Rqb7l1B&t=&wl!eP1_D=-1 z9c`q>hGJ*hijwep@C23=crxl9>Mkj5pXDRtlV-dvkFrWlz56r73_U83RE7mg@~(uj zJvA|=GZv|`j?^#GN4?E>@a1i*iFec8cs;|kjp@#j zRfFP+)zt9os7|CVq_C^V zuUDw3&)PU2Y!4agA0yV3waAP$4Kxj#4t|&|Xj&B0_GK#5tCX6$J$GbDYoKj%xuK`5 zsAS_Ao<-_}RDo2HdyLYz7}khy9^Yjk{17Bg4Z{rzgA`)$p1Q=(M0e6T66Lb=suqjtyR4&;m5 zL{#Ij{)=y~zNzYWxX4QEOAqRxou!xJ*;IMqddvNmYPt8FqC1+8>u%W_td<8Eon z-i?chyCJ~is<_lgHx*9L2dX%2XlRgZ$WkR*RUar$jeZ5EL2(itZa(_{`bCeQkl*X_ z+zYM`wI1p=M~=jfwAJW_RDvUfvF@*LycS6i4SgT+zVLhw)OK#fA*86Kx8+_-nqo%G zb+g)nuLqSEKQn)jg;&BId86ONL?Jl|=RpyWG8}rq$o$6ywKc^Q1^VT_?>XO@-=>@7 zOx*o!H8%MpQ4^_Ymg$;F$l}a&tI)9tT{TkDXb6iEE@&zcx1(Hmx41GEve>dnC7*q< z;GJ?yU?MuA|4I%l>|W6mXS9ZJJeY>9hfQuod8D2~Y)%HHs~^8AYlCV`|5i9R^W)P8 zN@LUdcR!xOEri{$(;TpI_u3%j4pO+|P`4)1Ua> zeY^E`fxjo0<5fM~&5ogw3a*Q?7mvy_ds43}dVH1rnzFQh8h%k5S%EqW?a63$!1^S> z>s}b4-iwE-zlZ8v(paf;YQU&^3V)NZbE!HKd^S@3wE8_YJ?&9cl4DN$WLz6nhk{N^ zeppVf`3z+0bE>%dpjw8JqyxghL*3}H(c(e>ZT1*-Ur})9j1O|-xY0JWF z(MNV=H1S-c+Wk--BC)c)F*Q(7P4H~btDYlylKnB+gdbGXER&9O)0Z*xriqV@)h%>p zzKWp>3w3>p)LP@0U$4jXnHKTP9`0`G9ccXKz+-2%CiM-Rf$JhJ5jsxXVl)x2mg<~~ z9}Q-p;yV#4$R3!*uo00NSs0-lAx`DXVre#SD!LRh3>Mc`)J_hbVAiyG@0vX_QIZj7 zNT^Dvasxk(cJ35Pomr{5%aI=STC&YXY-p(qF5@ZX;$t_~sfo!-9WO>R5nmGITU>${ z17~Jud_5gV%lCcdV|BKF-P&!C>3ei$QG^rcEPa3STK!5MLYG-rZkDpo3VM<>up(1hHXb|@ zEaeE3_$-;`Y_RZY_G?{Bd(4{{fAl5Llqa_b@nhbSQz}Ln`ORg1m}MU{8?QJy<$1+T zbZ$ITeLdZgIJ%;)4OJi^I@Z!b#aG;tNx3Ub8+{+}H8{9?_}bg0jD?Vy#4z+JIPTa8 z)v-Hk*NFp(WC!2NMJkXHdk>RwQIH*!5Pry!+~Dg?@gUli^3*C=!@-B~8Z((diPd_i z-l~Hg(=(5reczt1FA0ztn2=p?-q+gZ@aY76^dLIA_T~#?Z=9i&(TMRgsIrQ2bNVK8 zA9+F&16?2SC~uTEvlK=|n2b!C?Cz}_D$e_+`<#lIR11XXI3!OV!oIxl;^ZO3-J2m# z$<9)oI&kK}H%{B-oM*b_&tFbWO{q%1c70Gld)t*N`kD0UdpCIwe+;>K^Cs&jl$%9o zS=ab!LaI9eZ_qM0?;g>fJ@Vm_o`zr|(03?$n}e&EnaR_SLYTGyYRaZ!;f(83adAJ%^QAsUX_j;fuyKK6<%z`Bi zh68=oa|9<#Nl_%cMxni_nkPC&a`=rmRc+ABR_}igbNRLOc-)z|6aNeeyfnKvkd&P6 z*JDXiZ<}LrCT*yTug2KpC2f`kJHo24eaY&Za~qh{bBo9t<)J11hLooa5+;pW)b4PP z2DriC_m^FG24s`$bDToiiS3n0zdayE^OY{=mvJ*obg4OvMgy4<%JvhDY~r~x|I7%O znzYS<1DJ{rM?-hMKcn+#=!Ku=OT)^5hLvBY&wBsbzEY=*F&V+XNe0ZH;_Mx>7a-F2 zqnZxlu$a}{Q#)&Ay{`a#%r|~p_2P^CYWr<4!7=I4fs zWK>R6ZT9wZ$UHSj&0SN|q9A%dq;Y|1{k)car+aU(F0 zifOa7d0wo`!PK}Z?zYX^S@nvX;mX^#vz=cJ&2zDz`JT3Z5t%fUlf~&FlGgL^rLWI& z4ndrH=>ZwOFR@S^KVKamEDLMUHSp&W;_1jIIQCRdv~7CdJh$i*X&E#>Tm|Ta6bo7k zZ|?G=UYO-9*4V#ZIJnQLN048)%gJ+MP;Tg*JP&8!qPc0*xWS;=QNyhI{yBNWncOMv zCNuBjYC;bjXoIJi(e-PmD$_zuiE0T1~sLOCCpL z&Ma`I2cX`6@2?G8XK}6ez3MxPuJEU=GJ-+qp@be6KEtseNp{$QHgyQs%Mf%$>adkR ziZuLuwXy?wj%BmNi)($fi`QnT+&n+pi8*=()fY{Y!t-QU4}Lcf$D?qFjKP6YzS+X5 z;l#vE_IJ09fM{-o!p+OL=9pAE!aG?kyXD4_`4+rsp^n06yJ?VBfrI;+HO6Blh#I^qn|yPflEqboxY&X6 z^JQ`QbX{ddbOYj4K9c~?wh%Daa9`%Esw#5AaEW8=jaTG(Dy)bYXhmQGnIr{PG0$xl z!P5-xZe|geTEv1Y23bS}?5w&wq;|ADn*os&?4ZwzLnC#^xD%KS?WV#tD!?*nJP#1N`L!l2Ne`Kfdi zvvP!c9$)<8QDh|T+D&ZDEGWuBX>(2vly(FWiI21;?VsOhA>Fq`j`zC_+d)is(Fal><@ztR9@(e*4+!#f#{UVo5Yg5&G(wC-9otEJ7nLq!T=A+YML;5w1lO!C{8??(OxjlJlEw=^k0CUMptl2pU>lTTL^chE((mzmQsJ_#M0@G656|SlW*+ zc0Z!Ug^?Qsmde(zEm~2@0`53U4kNQP_3rg0tgc@B;|^WNpe3G@5ti81Mk;ku-E(^> z!rMVFN*42O!9s^AH9cFSbV>ZHQg2GtX!&K&%xp%*)WT#ju^mI}Tt#usVIc_YMZV4d zK#n)<1l$K4}sb7Vq2o--yHjeS zCF~w7*{`dW?4ok4%eri#5Od`fu?r2Q$9~1shQ-uM;jN(l=^mq?tBM64pXYKjXTFu4 zXaX!%t^O>W-x!(UuCP99h{>{=)?K(O5APLrv5j(AVL9)C{;b`S)1KGk7}<#$$>|g7 zxuvx}62Om6x_+-i*SZum93sck#$9TmWw)uGNGvoh7$q<{bs~!&SGQ|hsd^5r6PsJ6 z5~Dq3U6}Cq#+NRvTtTg>MtO{3xrm)_$0WJzn(f_yRa9De5|V9AbWNd4O;tlEPpH6nyBmmG>JIVJJFF2X_ssUxM3o zuU{-G`RPj80N?mB{NU|v-*^tp`m!0izF43+h<*9GB@uKsM^EL0SRhYoO`pBIywB3K zCgHKA7<3aB%$eRt_yA_n)`J&d1$B!Y1Nq{eUSmir!ZzhqQp|NQ9<&k!N;qzIVKt`n z!fJz^2QFpqi8wYP6ePO#sdKom$k;fy+t#1%{t-(Q#I?Dg-Eo=^9>C*1%dZ(5tyj5a zr^`S3LbGcxODe+K5950Bz^VGS7V>rIPq14;ea$RFp2RHpt%$1HFg>XPi!0fr?hf!5 z*L5+(fkk)%>=b#N+}UO<<-_*jUX`pbT-`*o2SCsbe++&OOZ9eS+e)2K^x z^L52MDp-4@)j8KhJM5}Bhi8^fRb`p*)9${XX2=VW!iNJj56=B#m}08|176pc5O&7C2-EQHcC{%{`;%#01m}2*|9LRr8?O3xjcXdil|8RYm|dO` zfKEzAYcGsF6j0nyT@miFZ>hCx8=vw|g2#Y4VB8*tcO#R6te69OAHRTUwLfo?{l~;y z{_<>~D3&_@+CL~px{0DWq_1A4qUT;+bZfUIYPcbZBe*Cm>}Z!8TGGtHLRH(fVBig# zOw#)2T-u0h(1#A0fs`S$NDjBR`mXm}N2(ROMIKIZPxT~PnHOTKBche9iw6geYe$(8 zDW832^05%e#_GF3D+2^s+`(=OH5$3*nr#StpD`|?E>J}*fg`)+y0Bs%#ZkS$D|mI3 zfS@E(zxzf7515O`w7J;BsLB)@vTy&oxHkycyjz8z0ecIXh2;Yre& zd)q6FDIwPD*{+p%Q-wm5I*XgfyjIhSW=~_WW9{Zwq!q#bq|_zyIvilq=I6nT)El$- z{(PYC-2CFT=CfR4DNkftmJxEe?3Sl`(NKIp!Y#Eu*25KV*#sqWre~Yf6%G2Gl1!-j)M_uL2NtExiPA=FsE#dh-z;SKep*qV7| z(juwy6br$5bGex0r20Nn#en9NXHIug z2cB6uS-vQ@P3zKQBMN)$Lc3Ns=M)TI@Kx|RPKyU-?upUzc=)9cp#BvQ5HM)@tcN5j z_`x@q-iyq|$_&90Xw9TVJzN*%lY@dchQ_rejAFz-<$lr=mi z?oY$@Iz)+RBWf>^xb(c-72M}|g$tH>DQ_vWf!6p(wPl7K04Q_Y4=bdf62apjh?G02 zZ<~)wL|QHC#dH*ZExuJFYE1aJ8P`I?A%7f8#YzjYHDx5!fGcqKpuAu4zVEW#f24y6 zM(!iuWRi)`l0aU-E|Ba*{YJdMf8!(T8iugi66^&$3fA7*pCHGd&f zekcpRrc%$h@U+-$NE^KNdHIJnb?=x1PL{=(u-suaf0l;ZeJYrRO9|O^f69;G3NEk` zakk!^%|rPqd#(?6dKS$*J@Fx|Ml?P9mXhmJ@t2)Bxl2W>_30$|>Ts8_>^khwm1F5S zeQN8H#OZM)T~3U2dK-jcTVG1VIayadjR}3x|HLefbL4;n1lv!qwv$p6z#aWgn2~9> zzWH?VEJAixrjZg7@6A$Ce%DET8IzjLyb2+11lu%d8kh6IMZ}tIT$@IKctG&0;g1Q` zbTH!cl*$%84ug&V14p5xIJm}GBB+mKUSg&@KYey;8aVF`kLofLzGK#EIV zL>1O)q=&B;bKDER|6MWs-OiG)B^y7WkE}l z+*9Rqc4m3z4xkRPHTREftHT3|6$})W)Ql^(yScU+B;J3hgp))0jqa3Ruf6920bd2g zJzM^cMRMWQL~nteskg?%UL!L73sK)wmHBU z?nbwu&g47xI&h`+Sip3(vQtyo2yA1Nm<=U=s3tEEf;eLnL^+GHCv`fS&n>W!I^WI* zHebJ7pXbZ$n(q=~xE1bwL&Af$NEykO62S(j!d|_t!v9?)oCYW*O6amO0Hjaj!UocTs@skSsibwv+!WdgK|6JI3y?;7%u{?jQ z+lH&5&8#!F;~Gn_lDqNnW%pLb!@X+Am~srUhUK#HAESH}>JWOl_Xh<`Y$IG3Vtr z^XXh)IK zG0`+QH;mixT-k|Ku{ zO~Cix3sUSb2iJuFZUtV|i7)0@>9z|_TEOHC2iY`If=dO`76X+ntcKJr+fZ z=0!8PqmAzOOzWqCkn~IQ#f{DKCYIn+3|3K0>W-c+lj&K4xlrro8XF>p=T)?0H(zeA zZ5^GS7L|KLRDn%fTSxi%U@5JFPn#cNdM(#ax69uHhhv;u>PPLMn=SC+$!5W#8KjxH z>N(@qCy|xkU#7$Kvc;}6ZwmHpr8hdWE9=A~S2A`|H**a~4stm5t(BB%=DE({6wPR_+Y{-VDUT!c(v?fr8-%@j8&^aHKfGZTYoJLIU4IxT-}@ z=*{-A9`6K)h!*t5iB3>E0I#Q)5Fs2@%?a{#Ri=I+*{yfY^6KR*)bz?ED2hY5xYled zSsJeSW^PO))DsHWKzTiuJ^5A(?~}UI0iUA~z4F+9K-^)ycD|~jF9D<`aJ&x2rA>qH z7cSUT7Q7$gJ>m>ta+BtTO3Olg{#07}Q}9QZ)PZ8tL;)0s`|y;eveu>1B~Vl-Q)0Br z@Jm`fM%u~>)tP!KBt#1&AsJ|Semj1>ock<6>9r&%Y5R>pfV8XWXusPG0BvF<-71q} zqLilOpm=c(Ao-J`Y!~o35Q|+{B#m3WZGUH3pIg^pc&|?9SaO%(%Us30!;0W%#RkN3gxv0&xs7@#< zBZvEISJ=~yUSp6h-l=%(_PN8`um!LlrFd~!T;;aO$N@hh5zHw4uAId6y0noxttZ~F z+4Mt5MNjU;Jg?fyFQG(66M-Z)fmf(A*4Ur+ayzx+$8~7Tx__thsw`Kb1fRoGRc)>Q zxXeKMEM=p-HaLV3%+uk%?$VuYn#vsig^BVwBV8hKVVp~ErLrtx;cPN;u)EgAlH?+A(JjVcAPniuPa;z8zOEQDs&+?})U2h2VrQ%6H z8r@#4$B>&HP+t}-aoRLU=rW>_KeYm3Y8q6SYjF5YHuvf+7v8idD^1;2&yAKlisYn` z+L6^m0uew+*YrdI@pBrt9)16Gw=4g=F#Rcyjg^{Tv-zxz8syOo$vSPPcXBZr{N#0E zT>c&_{mBxm4dTI?{K9-!w|F)?aOuXRmFK>5710GB$_sK_R@6XiUR-o;q z&fgzNZvYsg2i+b0`|EAm#4-W-ikWBAG{1JaGdtEsPaqk-lI6&;O^n|s0gPShB+gny z@%!uT-)F-C>JU=;{I*Bi&b9q*#JX<)ls-{ydE@uj+Y_DPU|qEYnK`jXkJE=2jGT>o{i|Enft zLZ`P`rM;gVN~n4{;~|gNmOmh6Q?74TBRntoyT9532JHv;f!))OXHAIjfnYANt~Okr z-o3G;o0WA{kKsv^(=Jv3SeciC^j!3V&g}%s@11z%00;yj^dk=UUmE=N=EJiA@%Uf% z2K)w8%)FlsxLJhX#(3=xZ>|G2UsnGY6#6Sj0v270@8j~PAzkw*PI??x4IdKA$zLkNN#RH|;$_kh+6HLcb5n0~134AnenrZejM>5h-~KGU2xYCI zk@f`(*<&m;=Y{&eV7X?ff06lLjDnZi;m4(TNqiriiz~C`M2{?f)=CoFm-Mz@HgOb4 zC&@RT=BT?j^~&iL6>~nsMKd&y@00vqq*Y-(iezz~Xu5vG)O4oPqoZl2)@Fl=diaAe z9lYwUwEkzxZK`moMeq+>` z4vYZOod9GM4W^N9VR#r1t$QAQ+sj=Qe^rWG5!yHz5aWh?;=(PB9+60|=<-QP&&)!_GNwim|lhqRKKku+?d z&L=>qhvrRN42hc*ql>zQ_I~;@2P?dAObj{X3+8XsW_#7nkv@#~YYaV}mclg{{ZQ-I z65c~~m&zX@dpZ`Vs=+1=rs)c7jBHW~)x)89Bu zcH}uR5;A0xZ{nsOn;LiUiOa%;JN7||FX=OYty7UPdZV6N;+sB^Ry_Eg$HV#=J>24{ z3+MLm3D2a>C~Cp(_@}4ct@7%_nh@G~g&j`7K02Yk$Ic4*8-cV2NP+Ruhj;qpElFb> z+(iaDE`A}cQd0?{^F1vMru_--#dFJfVUU8Hr5CH8+%Ca-^12POL>^7|3#OZPuOdRA zZA-i`0#oj}w{;>1_i*)s9QRKk-SCqi`t&P*Xm)ubqWi6+=A#=>0>+juhu^z;MCPHR zg?Y?eM?=-m6H=bKEw^>B6F$v^s?($-xvg!YmBcbs32KQg6tp_u#QK&m?g|;aoTLFh z;WeE4CV&+3HO(C zAMb-)W&{MCFv zz`8dVA{+QG}$#vZxxvCRPT1y3h=du|ucQYg`yOry+i17ke zwmmNR;9mV-+Umb$z?A--QnXm*!YJpE)i~sgFf3*>MszA+A*3hkUR|V)YekOD)C_g4 z!_stt9wV<~I*nkxhl^gXV^#UqW?Nv36^~`XOaqOF!0#q=M>o6#FTRnSkGAA>?|ifR zgOA03k?3i&_H}m3l92_QXjpY(l(+a{x$D&OtMgRiB83h!TBfy9UZX;Cf)3Z{g)CBwbUPxATu{Htll-o&gb@gNjH z$yo`sdX`AdI$FcD2DmlJT;BOUzvI^(nxDB1lw{w3xKFu5_un2y*a77K zvWaz9m&AJi3;?hnJh{oVd-EHA0KC)6dhZPO+k^B200rOvaM#66<^~Y94&CuR z5;nI_0SOzUORalkE(-vVY^N`9=(ok}4D@HwJEYio_`~U6P~_iY^GCe@YeE0$^nXms zA49fdw09oCpq@-zgjEsg_qi@rE0$&=4on$+o@ zr5GpUdm@!tUJekhl(dqTM2W&e*aVoBx^to^(s<+F)ikf6L#@YK{?A3@6?++xjD()%QqN9=Tyd(;XMB?0(PRy`aYxU+=gLe2~iMj zcHzIu{MVa$3ax7n*UeUeHQ=~i_itBRU*NltaN{tr>S4)QJ#XUG47Ke?y$KW=|XM3G>mVF zEOl#d+AEi4a_RnZIX)k_*eEp%5!-7t@+xrXC|p$RA#e0eT($$Y`C=}e;PoKVP{w>l z8I_+!t`;FjvE}*|;H0)5vsF*3a4?uuhJbI$f%zL$!Bj{B>LijR1lF!buS9tW} z0+O_YEPj9R3O)Q7?2i7Y7FA%?k)}KJ<-C-RXVc?Uf47t(OLN5N7p+9Cwp^TE5iEwC zg)!1XjC~SOc%3Uf6CG0&8nfbQC>lx&iF$C(EmAEoo%%m&`d5gwxpCz0DisT-TFPbl z>Fl!HCWKp_VAA!ZGp^}Rd#oF^>)=1aMa@+EqB-`ElDAwA3nenaVWML1nM?bD;?&Dn z>_gXu;!G~i)Mbl3(}<_H3Q{jaLOGIvEw?9ulU-VE4l}ky;J`@zEWdfR&+l6?f34X8 zl}ko<6tAJ$H3+ftItP#xHED8cm$x(6E&dQUL7kkhcJv+la4w5uB^fAC~f8WU@&KJ zYk^%N9kQvBQC9H4?xanbAs-#s66Y*jSE?J{Is1$VoP?G)1d{2RZrz%hkMAl(5*4C6 zR|+yPNVm#(4?y>D&VvJMo{X?FbJv`d>gCQ>UVbM2-HrD{`Z4P z;C9hzdFKDE=;W!rv9(neMIgj2_zSd1i<{@EPHU`Z>|WF$^J+ix9=v{Zf&ffxVfM~!g%!VX4A__Z>#Pz+s2>#c=z&>c8@0^Ft2iCgN54(IM~2E zKYyMoupwGIrM$U&aLAY=x^ohlI^$eQL7SbM?D^kVb?eRCiNo(spmpJOSy%nB`eEEPme1cdzMMbdcic4%B03ZDFOqK?;A|~m#)H@!RWY`1I$yZdP3kO`ZpV}PXUG@c#(lqY4 ztd7~PyaI#EE3qWaE7qH@>ea-q-hA)2$9;W3oY6}Qf>*#6(`_| zk)b&%D60_0{4wEZ)b2IBedNg}t7)2;fh`pQ*AUsZli7&|3xHP@(zc@44tPp`6JtL$ zGe4P}9_q_PNSlc_yq6Mbp&ndxWj4y$Dh1^kEHW)aS@~+PEMoapzEiEY&AgGi(F`SY z0P$8YW$((X98iaEVKu~eo&OA4R=B2}D?xN5-rmyRay9LME#Fu^-7J5N+51)jd^&aou_!DwwZd+N=M4RI#vTv<$$8W zFpbQ|rq?VoeO4JH``2-$EF!F0vNKlEd2;^=zW%ZYq<(VMQv960%n*N-`?ZFRjf&C4 z!t^2=_fPO3Ch_uxQ7bO(%zjtNq3W^|L6*8jxs8#+q|l^Xm&Gx4CLY~F0isRQS>G|512Ski&@+Qr4tb=fEF2uxTH57aC&yflNUpCr11wrpjYBWE7_BINAEOSLZN7 zxxaWhEPjX+vTko)+|*cmQ5F8xV=i`|W>1mM1PWk|24FKl7hdfxG@V=x5UYP!IQq|n%fYSQn ziHa+-zwOuma(=)MC&nAG`+lNWKHxH%KvAotJufc<6q?R55fL}*1i4+$-ijw zM@arf5MU?mA0hb{P5ubUzXZ-C2(TPqOaXL7~JzUw>!zAO*WO4 z=D>j)XOdW9r!5cPttNPa9UX73Na0*<5MAPCs2aV)HF_vvxL{ezGO^t%6|V^&cc5A{ zDo~>ac1lvymHju^*?RNlGq6#z9{Dw5_n1Ed*vuGNLbnGp6##&%>55|8?p-(D0#0}z zsN~%Vz<+!A1l-`Nl(i@5y##>XOT0e2BW6}bU`Y`k%7VZ3xC7G#fDM|$UVIAwRpqTW zrbiATTc#tEcJDgY4>+}cI6(hDBLCM=ca(txWlu0Fio18M2y81=d~|I0^C8j^z>)+u zK0n>PYZLay{>QyPiRBe`dbO1{ET?rjKb)Fl4h#oPm3w@!+BpyJtbL$&o{-Bcw#U@~ zDqFj2LzTXG@6OnM8sG?8&-aN7yYue;S%AW-Vjq9Gd-&H6fx!=8td7Qq?9lwT2WeHH zSrz+2?AILr&qVfs1;8K_SAfzU8*EiinWW!iLm&_tp2MFt3Z7b|=pil@;%EGojnS`L zl%mgODSbFKQ|7Clpt9Ev-j@_tPr36zGOo@R;CwUmhsQisc#2o=5f(rBQ4ax0saGve z;LTPjs$UoNON23S--`{}TI0`yI0qq43a`KHXJR^UIQhONP%F;@a|*Ivi>l;DBRc7_ z_Sgh`;NWSxW^tr?njQmYy#dkG(xRO{vq`ZyTsf8Q*U<2lFGVQR4Xj=~H&kas@(z}z z@qycPPkW`6((NwNCru6=)12#TX`#k!5cdZ*ZgeO2GvyWI*Ifsq)rB2@Oyn>0(RS>| z#dPzRBq~Ze(>hC4?V-|#md?^;4!AYdrB@U#Gfd??cCy829YsWPb=7fk?Lm)6yyP+~ z0k}lyntMDlIlVX~!(tD)I>hK-^5y_W*9~7eRoDNeL(-+WTm-nfrONhr(C()u_K_d= zVmpjfPuFD0k#p>TQRFlhS!k?Wl|Q-%DMTE%zVArU?fBw8_5&L@V}9NZJKDX>V9+ru zNvQPrg^$0fsrMr z>$$tg1Tctywh1`CySB16Dm=RL9EU(1$<2zB4V-)>$3oJS*qoq#c@A$s6KD*FoYA~Y zd`s2GJRaqtrXRG&Qdk?+x4M~~os~TfavqS&$aaL`f6zzU42sj1i!P3phd0F=5_}Zr zJE_5ynWLQX$u!TP_!v?EXx$qvD_2!rnIW{hjy)^+NG@~Q9nZyn6p>!I`7VEXjKRRU zr&KNoOb6m_h_RD(U+lq`0}@Un?lco5$Gl-$oGygkLn9AF9A;FAh^!YmF&uJ9(L8GT z`64+?7)jO0DX?Wa6ZzQQDK*j#?tHyS>(SqIwKJT?A4`WtXrR;F+G&V5op zP12U2aA-WLW^u^0B>0@yv!!t!mv`Hkb8p#8i_RSI0j^>d_0S4n6926CW9VYtkTe+l z@f&h~q>$BD=#LJvX`ivrXH{gf04b@? z26gYQtg+bu=n(j@e~;E$061KAlC#*IiTAYtM<4NH-yW?-fFlnrivRWU{9lH-%mv(| zuQ*1wyOsNr3Cydu@4+4g-1jsm-C>vPc2@cOgNi0FuhH2rKmTXKZ|(PhtK8&6M1Ql; zKbMenjdeB6)~R)mOHmR4awgO;{;t0}zg7k2rCuQZboY%2Ud;eGn;)Op-A#C11?H99 z%k=Z+$lo|BZ3>X{|2{TLyM_8I|ITGbLI_#R$YLAg$U)HySRQ6ibKbM3murrDo*aIjE6pP4vdCj zSCmF~SpD}y2^o123)1Y_1DOLs*LYVxTQR%5qd#!y0^2|k?n3vcJG)CaT?%l8!?%ZH zyKf$eMsivq$5{zbBX z$bR1iV8^KTC+Xcuboo4B9^#F-cV~x+9-!fCG*{mI&ckhmsRd%kI+G9byQ_F8Euf^x z+eh};QgI!i{&lUf-47Q^j{^E1YW0?UcL|IQ0L<2(CI8gAKV1H|4E_=9e^KO*F8@@Q zzhT86^Yt%^{IS^of;3)#oYe0W`Qw-W4V3;sk$+L-4;1;A<30I*Dio1+!_*{qr1(b{ zIpUd~7D}3HlyAD&Mb7eM8DfxxjET^S4B$Wab!NNB2hdB3iHRpxU=!9dc6 z#YDr5%0AcIUUt7o$K&>E9BafUVKUL5G`5);G1QLN z!*p8hl}$f(0U${KVzM}rBY;I6wc#4OzdtR5l-5@h{j#Uac4NbGlGfGOu&b{R<=Nvl z9C$X_psccf9s&k28UESAJL$tcio*viS_Kmft_N>p~_qgov5aZRxLr8r4O`O(>E~GM$!ET~B zkFW@&V5i0zrHUnes95@!HkB`dpHBB=I=C739Q>D6*iPio-HAQr?&L6IF%3H92g?8ex)Y z*Zn@655Z3p<)m~hUD9ns19r3lBplkKU31cuM7K`qm zWtkrM)~(G2TGM6;g059M8Pqfj4<)G6R5b(FAQhUWy^`la@~+rBROBJ;tOy0eM;oGP zX-Sh8VcwD`m))a(Yuh7?$B=kO;RBhCZtYk24aNc*Cg?=33z>K8pcY-X)Yx>feYG3q zAuhF9N>zTF_L>Q!odCo6r%vynLxw1ND4q)9zsw+1(XVwJ#X}DMlcIAFYY+|+HH12zj&d0YI%NR z7J3;-?rSMT@ggzB*I68KUDSBuK&k75WD<>iFe4v(Cay-$xxlU?a#`8PAv=$X$4<*` zcL5o=Kw|>jcwXjPr6{b^^WNb5s?Gd-uqQ8ZMIHhOm~cC;(sOAZUSZW5Q|BR2*PvT8 zTn8b@n)hZW5Or{1kA-+X55Q!c2Yar!@)%Uq=EckqhT#&b&=aeDEGYg=>)yQzC?`1X zPjZD?unr=2v9EppxKY}rY zC95hc%<{+Z>qZX=*8f4 z9XNeE)F^CRPvcB-NAcMR3=xT!UHSe)2DX*cPlO9`66fKlQ(gEJy)3Di^@=Q;<+Ctm zF^3ZLUv(JkO^t(?oPiWW(cX+bB#HdMJxWZxn3%!R#Dk``9N`rlI;H|4o>>+9UNa`0 z0@HQfv$s>jdbaqpZx9pzrP;#8pw%hV#yyV)We{mJ)*FY~%t8xG;+keP@~sD(m?uT{ zSl$it1ZIbKZP`{BqU%JoMKGgi9J8Krk5O1oj`0iQy6msYLC&244W@yQJf?Cja=v+x zP6hMrVYig@yqZh#5e`!fT8Tfx_15_~x=Po)ALQaXyF@w$g@7uIGEm7(x;IVj&Jt8B zwd;(cmdKX}#E-_3rm*7&U(Qb$f-HVSN}4?}1WeXeO>rE`A{OHzHN4cdJjkV;SJ7_K zLmAOb=eAfcY=&O#S0PCN71C;EsKH*xUg{5YyFSf(K-zFZvTVsc@ZVWxQ$e+lV{39E zQ-`a9N!P-zdBh`&;te0`w2j;j!Gl&!HsPcV;dyluPCjEgF!W_hv?P8&o*lJ3<2DQw zsq~ek=Ki1dzC0Sr|L?n`MI|95YuQqkWZy+(S7{NUlBFyYVlZZy%9b^MTf42AbdcWSEDmCSh1BVBC zm9bgv_=mPT4(%lOrDPRSOJc>j3cQ?&ncHHdQNEMjeb;Y3Z+zMr1%h%94wD|umD4Vc zp%DS&psc3}V}w43kl7QnYSFVN<^__->QJ|t&bcl0vhJRBLMNyAPM zzzrwZT{?zRb|yBA)J2;GBb4#?g-lBxRbidc*@{njE7~gns{wDh%_WatHXMB$+dcyh zus(423n_d*HiT310*q1!Zg3VA`Z(BA2_?0x^XN%A^tQz)Ke`IJZwg6hg5;E)lhX*X z3+nqHA^D%f*i1cL4MyUD3IiUJH8DxCm_^$GP^_+Xo6^LF<(4O6&K}l5?6T(&9^ETg z7N_VcT-(o2=qszY5w+`?GbvUFSl0P?Y#zHz-GeX37Ar%4D z2yQca;m~V#yMe~9fLt$)w87N&ul}8e>suFy=mq`GIsenuQK59^jMp?9%I)l`oC2g$XXM#gsIWYssA_mZ@b^t`zulYv7W*|IRcZkJ z&O&J#0VPJy@Nf11;T`<>jyfRK&#${Qm*WLcGyTu*+B@@h;i*7c+~(1%-y{6z2N(d* zOrT#OW9Ma1xdEj5CB$~IZZrU9#^{jLwzJVqJ`aEhai@3IZ}7e4-M$`s%lmg%{$HHl z28YbmefuPWw60#c_2)%TRyTo*mvc89?`|JvZym zQQ32|{kTkyCSByF1v>i^If*al}pnwFe-gw4S~{o)+x z*qFt<2KwKTzRv@t>A@>((sfO@Z7o)&C4W|e%oF4wcM@EuvtnGDf^4{yzRU_X?-F@O zTu(M%*cjpoZ3yp*rhA=|oh`gMVnA`b?i?!dWbO5BM>JG9h?Y&77{6Wbdr5aFWFX{5&WKD6N-9Liz|C ziy66OTLpZ7;QyM!f`I|x#ia|ol_i>AU*I^<3<|OY@T^YD@#*5lY+OH^qGIKWVIt}Q ztlw=z@@J2%WI;P_pq&Ze)j0K2hSJ{n3|l8ik&IAni`&h^-`JydglgjjP7zf;3IX!& z#aJ*}HYa*uM@oy=Vz*e@XcO4{KHV3{Q}jNB)XI?Q&>H)0dCJzi0B&S08a+D?n+d@= zS~L(6o$r^hRFaLeOi=l-`C0vW6WbJUAs?hH!)JTEy-m{N8p(@6xJrP15(Vvwb{wrg zKGAIvhhG*&fY`lOBBLcs?EiQG|9pU3&;zJ~sVRgFD?0#8;~a)9R}_-A35^iY3&OV~ zcC9+>mOC6WqCRv2*}yS)EfGk)BOSqOFN`NK1lKQuc%xJ7yOd4}^R&9vgI!r%kwL8D z{AMNvLd>G~&X|-q1QG~@$pk4rD_vaBL2M#pZQmJpcbMjfmAK-WGkEKkgn{lBse)*; zoWak6^`UGmqxGDWYdOm#C$8kzB%>@OVvqCw zm+E;xCYjToxH$*iAF8-}nLqqv#pTj0#u~fxo;YXSFXkuA^>G2Dis0OhzS7ZJF0#t~ z=hZLj#Vn-=n-{uULRcj~@^C`Q){6^i2B-A%GnwlFx<_iG>SCzpBtZr{TX@v@_&*oy zk0OF2se*?{Bj7|XdDmHgrebeW6kBd!PoWhdk>xiErTD7FwI57T6T|*!e>C!HabtpZ zLJdr;4%;|PM?N3bIlt7RN;o!H#@+G(qoriEJZXe_XsAH=8j)|e*3@YBWeo$_6yKbr zQ(^-+WQXxg6FbXNrFEgz%?M0d7;(Ma2Ye(_N+*iw#+SGDX)Zg-h27`QsjAarGh zeHY4$un2(L9(GRrgG(8bkm&_tGXB9@5&TkCrq4FZ7B@{m>J%sNK%v*awGGp9z&?wZW=bWVR?}hyGIcSVhi;2cD3J0CYMr zm!|af!n#sBjM5~I6roJ1d5DHc$PfOD^Hbj}w(WTS{hG`hs*IPo8yp;B9?S37olA*0 zFY$tFVUI)k1q+vHkzJr}Bd z*!a#J#pG5==DsesYF6j~9p(K4FNtRw1m;^b{6ZJHeJhY@ZEu&ss<3S_hUlytg4j%y zY`7*9>UTbk$&v?wBtKy_jN-9gcExl!Ep($u3`EF^4W3hzF0UYVAR3^PZiKSRO2wSA zrGoXG*$M;{_EV*O+T!Bf++Y9==qR;rnY5Eb@51Wj4x=d|{xZ{F!ws&%gaGnn6e|Cm zk{gLM9ePRjZDy&|Lieb0AfL|Qd)yWw=B}yZR7(3a0tW!rZqqF2`qf2?QEA*~;TE~k zqM9$?MJvv8LPkDxSCqME_QvjjJ|)`8m)MdU;nf@y~31S;?zWw;DyK5Jbg|ax~2z2c1Fn=za zL+D?SwC!rxQ1;{#4)?r4!Uq+o$04)zi<^PK8fT@y#wwOL*~>a+pqmolVN&BNQY7-1BvB zsLRad9=X@$Zdc7h8*W%(me{NHCxQBphit*?bV$68yjL~3akHn=+Ccc#z$;}O1>yy=IcOf{PM$5#Hh3|an`W$dP8u`U58 zzFC;v=UJavfR>fE?6?b?XFpY$H`Z^A zVV4uyy2bH`vPld4>!kcOegNbDJ^;n&80~_H=m7Kvlhk057d!hCP5@9N@^BY}CwvyT z-8?|5{K)ob>92-s1i+=TSG%y=i(UaR4qxx&xt+BHH^8gF?hW|A7iDjS|L&@-O}@9K zw;y$ty)*m|-A=uyw6-hrJuUqwQ2N)yo>|&%uK$mA*aj_~byzgd=4WcPfxe9uWSRd~ z_RUYcVM`SL61{<2eA>ml{CUnl;{$Gc=JmI@$J4>`)RSMSG@+IoD?u0Qyh zs(4tLx5R}oDI0J*tN#K`I$7GwysPJS>Mb2rDjvobPbmc0OtnorJFY6}l^3a5$Zw@H z%1S<%$g&mTK$8m#3n?cM#%2#j>w|-Be^Czk1V^ZFQ|3q*1a-4(04`q3!w$wn^{j3{uW(DacSHw9yT07jyXiW+#<01hAw zxMLIftm1pMsn)!f8*`@CXL3OqI?lq>CoYCj2pjWs-yd5C(f}`;Eg%z2vdToTdH?=O zo~~<2jLK#OTI4$jRTnqow960lL>k2RS`huU;{RV;5CL8(l|{^U9r!o2Z37^{%pU36 zyST&usa>WnWtEns-z(px-4jp_QN^)W$n`sv*uyBcTfF~JjN<>LPOnJde#I_jdPBS? zfi#`4OXU2jNS)>7?Gy2Pdr%)q(1aJHfbYZ)k*>Nnf_LtmP?dZzgnzk3Uu2b2B0u1- z0{Q!vlibfzsY9feTc#O35${FTIJeu?zp75=ED)YE*39boYA0uATE9p35*bck-j5=k zO(WI;@0A&*+MGZ8rL8wqGrc$)T4#V76f|FT# zEPkZO-+K@mDA8ug9$7VAJYzxG+_>hF_3}vi`5_SOwW*;vFzOKfszGv^v-2sSBT%Ud z=^Jz{2YLG-VSX;hZy;Cr9&@d)`WHM0wjO99WxV2K$^tL<9<1c5`6iS%8+?1Q*vs@Z zkoM$LS~Eeh@@NFf#LC_$X>kvMHnmzGM{fygPO}=1%iM~^(iopL5Pz1*eypMD@uC?7 z5>{W;fR_zZ}{P3*rt!oe!jM#8wav}ZTQY*xK&H zk5dLin!qe@X*I%tNh}KSMy+MH>)gF*Hq`X<;42sd!bUf;qx^Y5hoDIyDF`GwVg1P30GeA`!2d3| zs0j<)OwKHo=JXo*usRiCKs;8__mmR7S=geWm{UH|CPPdp6DZ&vjC@~}XY4c(Z!_1# z-)fKy88iWVh~_8`JT{pN!Ofb#M$1ARp1bM?=UZqH<7W=Q@6az)fFkwFL(=F^pI?Td zJEFOoKd`HD)9m-cN7zK13yo0OltuxmP$ekw8gOC3x1J}Tcz{KA@kj}0ycmWV-&g^8 zN=FHq!n#G3wyLsUqF}bV0KX^?KdXCpXwEl$$Kw(J23|FqMSmvjdaEhfW zP9!P{$YtsUh^6@qK;PSygaLc%m}Gm#Tjs_CVY7?;Jx=&wT)<3DS?aPoy9zfGAd%tJ zTe;BaD)Wt%=xN--w-yBE*U|@rqo~m6V*H6pV9nG94=^21tODaIm6O~FsF{P0aZf8I zK7xU3QReztQJ%0HTUG5h`Wt&8PzNsxWOU>h8WhxH!vSQOW|rF`*H30U#&P38-XLn0r}ipr*xK5|HduC;du-4V>vy z=T{aKS#$$=XB87#>&MKnAdT{go4ToeEXx#ic5?zpDTDQYDYLE z2Nv;}h`sQNOZrPlRSUOnZipiF1JwoQgFTDB0fztU--g&`g;-w z?ky-fYgo@M@~0h2pX#lUcJ12yG~}xnZwoqP#saZLR1jUd4lDrp!vRi!;{X_p55Es!D~cs50xyeF{jz z1&gXb0+*|qfl;J&$ft1qa$E(j9}SgNK;rtHIWefosdo}g+B&v=zdEGPP!QqaLaghc zuZn@3b#z))5b8;MWhsBrc`53gVwa}FT!~Tw7lj(5)zFH3@26g}=NTF*H(IXXG#z6i zA~@Jyx-m2O@87>pKCLn#sGv0zC`u9CIPmg?#|Z=F#hEZERZ35O_8o^6I!%x{YN`w4 zY^br4Vxz|tXy!|eU8L(j^(FDcI~nDLNNPKQebsDbN%ai^yoLuCeCF;@k{w^qaU^O? zCdI^rrKdlJo|EAjAGf@jrGDuKJ3N@$?p)Q(z`Ksr?C*Tu7zL(rz8CL7MZWK>WZfR* zQbV56Tp2@gC50c0M&3p!E8>356f=-hu!!);i%!@|w7F*0%+^Seszst_s5imWL9EQF z#80_o{X1h5ja-r_qF69N>Wc6MCw2%H3dQ-;cnA2>vPV#lT=7;?=KST};>M_wE`4Tx zoPL%HE`RH#3~X$naA~+9bG#jmj(5O{c=E~i+*_LDq4SXo!a>4RB&b2waY3uDnhPB) z)R!-}bzIRu_8Oj^nc3X@fuxaJ1a{Q|gVFcXx%;Pn{0QY&Z)$C|XJun+kH^yRc%wNQ zxiqV&sJ_e#8}Y_X9^G%Zf%)=5J4eDy@+Z6huO zk=E_HgAu%SgurkX(pt#Bz(C@BDRx#pnD+cqrqX4M7Gv7 z>(b+suB`nsrA+CrQ9r{yO|UOgrJVEyvgqiuRYY2V#o{FmkZog7FZ(U zr}9z5_17&>7-%LZ=5XwRmui=DHY>`XRm^Ot$P+V72A&8?JLzRc6e!^ zqZ4^$E;L^F>YE1;Ps?%RZ)lT;>q~4j-KyS7gHe&Gi4G?lJ5KRKTpbWBzg{>B!YZyr zTXd@c|Mb0`8tg#t&53UYuUN^Xi}e)RLawUujNdLjI;a`(qO-s4f_w>xiM)~L?x&Ka z4*HQ$Qwr`xM19uemXuWYp%v@nI_pU!5VXt&v_G7`7-<~L{KN3g0h9i~z&o*?-r~LP zZf;-w$5Y#+oqQ}H#{rS9dUsXoeNs1xR@*m~bwSdU#^LpitkXBno>aN844=>ve%e0p z(~|S~IXa=&BUyJH**;{I*kkOdm!djd@wfwaHnzC882QqT)MqVDv!nUECy!ArujMEI zaCS@FR|m??JvrwbxDSdn;gt>PqT%KqPknbGAtB-N8QPnh+`{&seNPF-+e&F0;w4#G zX=2WtJmL@Etev2xT0B`P=6rN;z{ty({vg#i<2jZOy4>Epwyz%eXt?KAme$z$HMqKa zrWm|qmQ~BCa?}f@IvkLjl_ey2zEK>S`rwQoFZ@0IkPDY&5qq}<^-!tK>%g&!@3i%o zPsv?u{l%AwP7p_-$fVi=EuLByDRw_qnfIqd*`AL1CAQtY`l|4aw>_|5?M=jb*=VUw zq0+8IT6nwx#=&yeN2*M3pPZhtN4bPlaZHLypWI_xd%adUhhykq6n%QNoY$glVJ*JC zSxM0~;<5O9$Fm<}D4e1{*%&hxIPl{WG!|O>>P`%i!*KGqhMRbymt*2Cn_PP6=GLvi zdUZOYppC&HFkt-A;{6;zeB2rB6Jw54>c+E|#l_pNx(-==M5cV}ADf3fh*rphT+^I+ zbLf*Mm@LzG{u)(U0j?TzmV`gRuy7eTL87@Q>;swh`wy~3h2hzzau$Z1^HZPX7bIfO z!ZS`Lqs@;B6N)2M?mfwuc$%H`kXHLD{QE?d6z%<A zMl(e5*9wqzX_0$#V}9u<=@HJ(m% z!&T(MrR0NQp_OthB zOZMbuM@>jn=Ly+Gv{yXU%J6K{x0Wc$Pu4j6jc!g!csE0!H;;!!=Ws{W*GIKI{FmVJ zAkLxFXJ7G5#|KlBb%BG5xGU{AamXCH>XSL%ENvFuWe$-Y2{6#tk3-5RDSKIdZOG@L z>DC4$>x@jp=$t*NO`iXy+UURRI~jbZIHm2BiWCZveKic&%5*G})$fYb511smFXflr z>*|o6)=xqmdws{(_M(%sUNozoOLb!6u+7>GULH_1K{chQ&gjB>hxU(!9;K$zm=!U; z-AHxh`-5+4b7ujQm{gpa5y+_0qI?y6wfC9YuMTMY^MYu`4PWEW;KX@$}i_?Q%!pssh_*RL{!ASy9Wg_~2|9u-%;}V-ga=Q~CzK$0*5O71KS{s6~-Qyp@bd zPT|eF3M^^#ksu6IeW*}WMDxLOsZ!gOdTsOO;E1#E2fzt`d1Dv?b-Cgq-#-a;$=gzzPWo^9H zX!FZF?=s#s@sPNJ9z1sqV1{3RRMdwxnN?8Ue^0x>LU(A8WgqaTrLKQ9NADr8Q4ml;5Rqm>q)Q2ezEM!BfbL_|dW?CE1o zA|e_+BBBc;7s(0#`1y=pCL$tvX!}ABtf#61v~+glGk@u90pjy^bRo185lPB;yO>+r zgTU+-AZuGEsoQ8M>^8gYOR3wsBB}zaF7hB7+o!&6AT3|D7nZ*Emf|mO%Scm7dIJdz z96?}nc5g=qCwHK?)a~E+0twG&fAimF|BVD}FLhf_^*Ot|vm1zAluv+9;I=d+yQJGo zE1>3Mg+E><{F1tD0|vVQ`T4!Py!gC?_?+FW`31$r#rXvu@jrUROQ7I&_i+N7d-FQE z-#KIPJCDa8cS|>07qG3f6Z;uYa|>q=u+;6_gm(5nn1gMt{^suF{>StQg5W=E;TPl+ z;6G0e^0xgCWP*|8>@HbY!wdw!AwEf+r?PN@W+}0c9p#Rv`5#;3l8(VQPf&bB}cdk`P6mafkZFi8Hyt5?`rOE(aKiH$S( zHvgH-`Tr0)VfFa`T0X*$zZ3xEL|AnHAn+3O%jw3~7d%9HbW?%POs?eo!Og;;14T3S)oBap{Uu_$9fHNS>0}$tR>ep(U>$ z)Lj42@wNfgo%!R~Lhgw^<9J6&<0WG7gl;ncv(%ps=|6GruXjgJ5x<)0h*#Quu1d6X zkF26@Kz1xU<^@9mMScq#(R|Q7Q}%c_3ya%xG6GZhrp9LSDsDk->t634G0-yVcNf?Xdf@4Nzj|}Diw=PB5QgT~v#DZsG9` zsq`o_lVgmWsH3A*&hMU;6?f}3?SuRX+5R`;7gXMHe%t-&Bk!N`fsLNHoTEMQQ)B)& zj^v*zuJ4A(muQRW8Mm&nj|JE1+@+0Vs}Sc)W(7#1t5uldh1z8nd&mQxqaZp9_uXD# z6Fh6)0kWgS89mOgqtjL6utESREdCI9caj=u&(TAIW?1oh*u@pW-SzEnfHX!)FST zPf4QK@8Cle{q@KM-b(7l;!e}ptY8ASzCET944q@r-!oUX5DPicuMg_ zOg$bb^UzdCVW*E}l}SqGuyCEqoLO-4+A)K~yVMpnv*j|KqDy6EmU=>Ckt`ZNKVCny zap|CjGXPf9`9_$UteU6vLH8Hc_SK}+*t7%4GWE&2Lq19v0}b!@ck`LK&PH-rY~Lm6 za^+#Jy@f#fp5*_%p>bB1Dr%*FH@w!k* z_~~VQxc)`V$cN)AfPzIj0Q>bu!Ug6VQB+$*((FW}!7f**ugOv-P&UEet{IVAy{D_A~+3c%N8E(tbe=~Y1s2N$L@a04CQv)4s>uz4k2qrTWgjC(%dQB$`~d7i67}^$9CwE>d}pndXz1S za(@veF&b5fd;xiqHCSXrT=l|$*ipPV(({vN)PYX$=ehir;VX7L1Nmb%BX!x^uhl*q z4vFs-?A`Vy!?EK?nslQxZF9blF4*^v7mZ04`50H4hJ9Wvx?UVxTCHLqKPFzBH+k2Q zd&A7}v*-H*(T(UyTBaLqDJ`IjH^P_@OnFSxz{f21ENNATbV z4WkW;U}(aP;n44vOM^ZA^D=R)SU95|k}xRw72J2Iy8}S2sqXUk**u zbJ&W~3t>vIXQw5^PG;6u>;;C+DqC!WXUc(=0iBecaS>7i2XP#BHr6++#kyRQ&7tP7 zkL4QqA4hD02cKn`SmwIEa*fR{8KGfVjI zZNu_QgR_J4wsVP_^@F=W!w{A>lTPWiCtH+z2D786{eDCn^y~4fLpEs|>3V6Xf0|lZ z8hcV%K&!l|kZCho2gxTBXWD=~3^c~);{8u=o>CAm5x2f#AyyZwNx9YU)E_3BC#ygE z`76N4c~f$8o1N!@a6V7|9X$y#!McY>Chzp0>O2j2YMSr2k}@!F{%hI0GHvs59|cL` z(#UbfO(tpnlNZ1GKK6gC-5&fD`c(gA&ttH~QM(8o`mw&kv&^JSIbnWZ)xGH(YY9w( zG9A|}EiDQy*}BBL`XVIg;P28IVBC0*U(8xVe?`0(eIMFhbJthXK+||Ac_CwA^t*9% zr+N}#)jw1&R4hk4HYO>i;Z8NgZEL|J8ag~RtTbGrTAudEzPrAFTx&O%HBlkGGu=}l z^+Q@pGdJ#zOp<9E8a@kh=v|?6AzLGZzaeg|Ze{&gYE!-PEZ2Dn)tjf^tZ!f8TY;m}SNc0TPMEJg1Q@6sw)-K+nkH5`!#PnzMXj*nA$|-4Wooy3YB8lasLdwMC&kBb`OdD;QZ285>GO&VKl7E=$tNRE>cawgJs+AFpN#)p=-|Dl zaF0rvHINRg8c?86P_&1=mVVE$xdV0`f+!#HSbdw5-t#L679$a>8v`-AuXoVn)sNH; z1e8g-`*cx8T~F4%rmLi5Vz>zV;#ob0${eK~S1}r{i>t12*f3qoEtb%o(CantBOS{g7^1QMB0J&|O_tIL|(P*Op2ybXGehbwZ$=VOarcK*I`8O|3 z4o}SvlzH&GJ7c8F)XUKm_&wY>&Es1(65phMY-KG*G1Cgt#ERUGsAN7)sz`1~(n^w` z4PkS#-?kOsi=J1PFjO@xj9OvUca8C_T3Gp7o@t5e%IWe`f0^q2L$r9~;QKSq(v(oC zQCHy4y@_=BKxv=1?n^)Pk(I^Ejc^wH8=Ug6Pt_ct~?A)D3SpGTLwde%#4O9XjJ7?>)cT*O<)je0>J%;y1_Anpy!( z^5WiPnkmHe9n>}%vl=UIQcOETuAI&u$hWjDM=eB2d&W!VN|kt9>|}2i^bC)seMk#~ z-wWYXvq^{@j_7)E-k?>b%><^P84|kSV0SMx<*VsK_A`2U$yWYej zqaN8b>5Js>)b3w?wGKvGN`n@xUqQ59SPzw+uuhZYd||#Zji(Yw31*c>iUEj-WQm?V zmV4n%ygp6V_);53v?b4b?HUxECwy^zlkG{4@<+9go8(vJ?moV``aJtCeXzon)&3kP zLtJEzegpmx0Bmcbb4DT(8Fvr-PKw-jr`%y)-s$3pDM;~of6QXMY2FuMyOEGp9IZl5 zlr&Z>0FPnR_A9=|F8hjzn1uA=6~cq87p00o%aQHR_nb=60e(ZmtJ%@uAbWH~(Oue9Imkcj@fq-?^+jA>Mz!KVWi#trz|W_16{D=lO*TJTmx& zXyKiR+My7Avz*6&_f2@ciC^FugN^vUaPZJ$BvuE_n9Hn`!8%EuUvc+O_Gg2nBSC#u zofVJ$YuING4Lk%&&?C;jNAf%AKI(HUQBMNzLPxxb<88mZ|1sZ7y^_OwI44o%{eewv z&A0gTJo~8fLqwe)Q2ZgB-T?-q)1F5I1lXJFnq`l-IPA#zn0qi#46b_H&7@#$Jb_V z(17FooJ%j|huG0>RP(1((fuWBb^#sA{E#mpk+G)!0M)+-BwKn3oXcuJS>`|YCVn`C zEx_pE$}3{J>_qD%j({~ox&wn2vK3?Cw)^(RoJ;o{#)EOw3 z#X8B)nw9)_pRyV0gs~h{QlLD4)0saf)I~NTe|&U7Q}VpH_Jo2I_6J-QBd}%HjF=92 z>MGKud}`&-4HZGz{uQ91$enlN1}Dm8BW*+;fxGlj=A7!_v}L8LVAA@BPyT+5D5#l> zu-<_5FwLvyhX$h*CWzx0Pm%TTP;xYbtItS(u1iGpSr-*p>}jCOOGhHyLE*TPSv z@$c7-C@$_ zb_2U?M%2ZHH9HK@h>go6KI|hb`U|Oq31_v`B}EPgu$cERp#_0jI!-lI+}bQ!*k+!MGzau zhtk5uHJ5-aurdD=JnPQk*nUdNKYN<~gJ#;bIPIGdgcjqs_TC;!er`!fFNT}2o9gzA z&nZBxkdn85qe;%uet1rZ7sCTcFbxJHbDe#3;X?gl810^a?f?o>th;fs4APM82=##E zJonD088YLe*xjax!8Xrp=UcF^WFaMm`^$c77fe~6oS)j)8<$>;k5UrHS;k~VpI6ND ztAC!=f2MJ~&;GTdk2T@vKWxQ$LH}bS|Dl6Es*7a~GtX)M8`}RwBb$7Qz@(BQx9&et z{5PYL-6bL6Ne0AR`0u9wzjrlYMPPzh58~qc|HY;86;aO;_2Pe!+5g`4|0J$TBrp;6 zGYeJz#|HWDUjEBJ2=fzG*+*XduRs33Jp8LF=l9?r%J{FU{BK+2Usd^6RsI`mO`!Q# zRsL0#e^urG(x0Dm-~T$Ge;v?&4)6b=SO2QYzpC=Ds{GFr);U}He`@+SR{1wp`8QVi zhfn!GZJB=%|LcJMbwK|*p#Q5a^H1V`9nilH=wAo)51;bS4f20z`qu&d>wx}y2Q*)+ z?{obuOCF#wTRm>74mp}nr(D$c{`_VLXkV)T!zxE-cGLVYn@$Z_UuF68{e2|}cK*9< zifd=VE$BlthpWnNEdDa_t)ocecg5Ef8G`Mx8dgXARx@&(vD#U?Dg}t{sC=o^H z&L9JRB!RV2^|*;;yoj;;Ms?2RX6U!utMRZOj=e(6J^hl0rOP>a>4CW}+&G<(XCJE?qU6rD%}s3Jr9}3F&|I^FqYzrDlW%+9 zIn!T+?ru0kBsS{1He)@gra8AphSjY%%60QfVR&G1#g39~8oGBnfH7L=yzbn*SoOu^ zs3l*+_L#-1#Aq?&0`1dExTqnh30CWPcz9eW`IPExGY%9IQfYp%$c{@_Y^;N5rH|N+ z1NtBm;c@6Y7Nlf)aUwS%Grow8)kN#J@ia&H!?*{ex}BXAZ2k|X!MnU~F4xSsg$2@? zVOpv9$Naq*3>`Fx;bc8eQ_LM_t<^}LsGhcX|WVDtU0 zspzt4QLam6JPCS#3u%Cl1a>ky^9FeYkz5?v>xx!a3FB#eo(_BC$>*m&7g>V$Yoi9_ zu8LGFcn@wh^`P?h_H zAEhp?HZL0*m`cUpkiC>0hH@wr|3-d3W4)&lWPUhh$TeXR#+TH|?A+MhK=ylkh6ssO znL|XD(b-PPh$gwC7^wn}xS=8L`>wy{I@NE>`lr>k@d~n;I>xS*D%T1w#%Sf%FDDgF zJ3qsmFAMmf5biDAgzHCKf3I(qcu!Tt<~gWorZwI)P~X`di|XM=IiZ!A5%ud*i}`qj z){rqwtqF~0q{jQL`)>BH*aoS~;Dn{HKP+Q=)C5{yRSmqRdNFoR3^$lbH(5oC+>)_V zIuCwdjG#D5LN1vrulH;UW?&M`M{wqxPrtYo+cej~xtbC}flczosQ|cC>?ax%|8n7e zvH5?nCd}WMOigTcq17SA?-Uhp8!@W8w&iJ+;`)}pQfbQT6c;<68Ciu=57g907gB8) z9i%oVqVubH#fB}& zTtVG2JtQaloDT#Wm@Rw*NCY@wzFy3g-29O>fEIF43BgZOB^Y zO#E-Y{4V}}9Ycq>w>0{a#vc3CInnbl=o?0Z{52-=EcJ`wx5O*dW8^vhwxWA+U(5W^ z#YVo9wzbYc**N#qh;logL z(|9dqj&Re%3_POngiku|Quz1Ft(Ht55%>PwGAmubRYNr_r#V}d-2!^M$TS<)D*QGM zUtQT(AT$;VBuMBXNqQjG_N-Lt;aaiYJ96aVj34Z}e(VnBXBdA{H@D2E{`JbfxZEVW zFS$EwsPE?6gNke(*eba6F;9Si-HW^^bnykl{qWtzD9Pl2nUAL2uwcxN{+Y%3UH$`3 zBpS6xJ3ah{2J6iSP$|=8X5h=G@n{j0{h+P*O-d+tK-_5bSv8aL1ne}>9;Jv0QxKOf z*`+c2`idCShTYoiY9&D8*q{)L8XTr4Jy{rxtu_sHyRL{;T#suyx&Xgjs9HZLm7 zZ?nG!dbGj15uBymHwICRhcw{^Bd1WlPeznw@JI8$@%G%N6O z*EaG6#as)Kj2;JpoqKhH1-e~QoyW0)&j_~cR`f7XG7owh2ljTA_{&aBs%0c1EBj{F zhvAaZnQClYq8h^HQdpwF(ef0!HxQXo$_4eMwtV^9W8N)uE1if~% zZ&&j)=#g^;SEOcnI8~%2pvR?`yb@VlOH7DNm z*`OIcDA6)f{NbXZPseplT_9ve%0X|jLH=!()7?}go4W&cVcz+*7Ni`(XZz{deJKMYQ6nt`WPzy_qGeIfZ4o5)JtdeXyl@cvDZmRLutz?2HK=FCHrhdgE zX3`tN!`b!AX%aYxw*4W3n=lY-2Ivr!e-*spF`mbn=KV+tGoqC#+#{SxXoCB3(+RCM@Pl~6q}M3(vX%<*RFRR4!{yh41Y))t|#G0v$O6DRx=!>C!` zW;7y|r6$>SSxC6oA~sf6DVWi;sN!@VF!MAo%~JssfGWsSLpWAy+e(p$mcr&b!TqNA zq_jY6qNbEM$_Kj9;BH{Gn)7hE{TFae^IUH7uuxk_-f~(1A2sA)LbGdeyNgYx@6m3r z?ZJe#vJB??!_{MDo~t2VT-5L#=#GqT4q*p~N#T#^R>OLEW{7(M$VS_Z>M@mF>vTGG zN!R!thy9k($Y646B?*UG)x}4&GCMZF)7MnPEM8OksF7RL^+Hx~+s*oAB}0d1+a$k| z`iEVfR-Z@F3SW@4>*dEK*N3~(y>F)mcBT)PIm94W1X%W0M08GVJt2s=P+)@}jO#cJ zKcBvwr}!v%Wf+2i(e0Dj{r=)pj^4-HRU?Y zQ`ws`I=%c1;Rvl}GU<8#M;#v_7RiY9$el#GUEqf8{;WknN*aLE@uRw|-zFPP{BZli zi}YGCs9U>m8Hj;(v4|3kx{&ShNnlh>PX3+7nCg1%JZ?OTqkj~GD1Y5o65Boz9c~8+ z8T-nxgB1Ts*W-<$uR&T%fY>-oo%X_#=G=;a?G6T2857XrknhRcriBVfP8eVF-KMS9 zupD325iJYZ;r%gV3EVZ~ILDfnoQWfD)4YHMj z`0u;vb_aelR}9x6T268A{}@=)f*pD zV=srkQ;(ZKV?%^x2*=!^%5 z9!J7f?d8fy)&l3p0abmk=8W`_bC~hG-Q^wz z2!z#pt7YhrXQsFvS?o=51p-FsRa)si7u7W9 zW{g#X2p$w+Wt|$JhIJi#Rn=C57;$Wjh|an&%VeFVEtI76;$qsj*K+g79DUG!QDy%# z33QdbYRE_u$xG(UmIeGVTmJgZ$yTDC!qSFZ0fPD$vcOqzOez;SffY0*mn;(V^$yej#thd#wOB%6myACHgVaC=si zln&O*T*nmC@<}<@6kTndcT*dSMIUrkylT0Tz^MsVFqULI(L-aN;P#H_)Rp*~^$;53 zsI)_Cmh^ekEP58IUE(+m&QO~EMf%j(J6|86)F{||y$d3ZXBg8py24t#y1A_SCDG0h z_{|o+lFR*Cg#QPxRvVc{EkpY!?Lqr(EI_H1=Oc1az@@sR359MSg%GqV0OONnCOb~9 z-yL6my=OF1&|8C4Pay;@gI}%Mi-)D@Ce(Aj*3D`Wz)$-u8`~n0CqUlX^+%}>mAo36 z3j$@zw&8$oARGJ>UoY@8TPmh2AVzD-9{iL)59Js2>U-Uxce?A*N5$*SrmkvJe&Bu? zDXK|-fi*GTlX68}AO|vmY?akzR_@I*%Q)*t>XPfHn+aeI8s1U7LJS@C5}6MdBMNkj0ICQai3XtGIs}&?&tJ)&da*W`pGL$)$T_9 zxHK0z_4mB|ymmC|qLcrf?(oH_}zQF)r>u+?2Trd=3?zakx&k2TjnHhM0CSx{f zo)kqbUo~T@q}WWWrtX~0U8ElmZLzi_LiiRG?jwDD!krAkCz3}4Z9^Lqv!bH~Mp2h{ z?Mv!KS-`WzTqL)cb6mpZAaG2N=@227(>W~r!FuWh#YKX(MywAI097H?dC9`l#oIN4m&kH}Je)j@GkKc*RCznY`t6)zDYnE^Cezw}|L?JjTaYFSWr z_!VkiL&wShAn#hyxJD>>Y*S#9`t)o!YBy73Yp2L^nZeN6Fv<0^?;)gATvg) ztHZ{)CirEw6jEv0aP+AMwZZ4$cb8dPRG}+^p5?}xh}m^D8;g4i$C8iMdi_TRyT(Xw zFI5KY7S|31Qk>L`Jn%!}4)cm@Cra+uJ%adyd!-JJ=#=EbF&k3lx2_j>TWJhNxq)MT zGAU+ZtPGuuhix;kjXF|joj{7h9-Gv^51!DA?dYF48ZiN!73_lDYwoq_w&L39xWJn2 z`&kq2i&cXj3{|O#aWEyy=+jsw$5m;dLp|nRZcVQIE@V*x^p!CIf7!!5o zu&qbvpuFY6z*E7=jU@Zxtm-ot8pk6(V}m?ys~wRg@`#vY_nG; zJ(Ul-zM~Mb*L`p?DWc__=On0hs24E!;Kzd}sWQ!{{IDIREcu5B==}84d%IJu2xnMq z-(9~aT(k8J@}9RCkfRE=Bl)z-$iccpUlzdbaz{}@p7jOoCqOj%^y%?B6cPfL0LE;V zFVDPb$yDMzmiZ$bm6c&!DFJGU81^1`&&pBc5st>LRU3B)@fD+@h}>hh;9qr?#|_s? zn+|-rZn#wz8J7d=DjHM-UF1{^|eJO9D0ThvwcmA6hCZFt~q}2z)F5GcLDADTs^Ak&_TSI^e8+W zH$*G3G|phAwGWHDiU%FnA}bz$UGImkn*i$Ez>BLJxYS}@MHXN);A+{_gSrZ=$?bdq z9>Fq=tVQiMRLo36lp8b|y8bHv7v|r}V&h9}-Y<5({-|Mb3_-cJuant-N;q9!_vD35 zRmAQ$WBo*DoF6xpro6q(YGmSov4*U>BaG3XGffYr^kP&0d19#^os34BYdj zYenvgo?%JjWTbir*gMY&&_(C{p#0vBs72p9b)CITm8$x=F^U$4LKRv~@pkB{ z9w$F=gXKK3qW_arV&hlo*lLA&IkMBnSx!}l>MfXfrkHL=k+ytfy zlO)$priXGF?tS8l(~((&gI)yWMoo+Q<_rEXdXJNYk5yUV!F%y> z5Imey6|@XjR0;n+!+!!yb@};Sq;id(<^hwW5@;RzbePiRq+I9!mvz$%N9CJ82N<)RPdIoP_Up(<7xekE$Mz zx#;j;+^s>Nzp)trmtI3oid`H?J@E4pUt>q9LRxSHoo4{jwG)*o*=lkkLG)|af!+Yb}0u*pu~a^GJ! zi8uCL6-E?QT4mKroA{CsiY2NOY_m1Uy&L?`p3<>-I8v+-n;!9 z8x_DGRW8q|bTi%U_P?dlSw-0RZPY7b&Bs36WFg7imT9I4VRL=?d4~ax;$Oz`I@Wye z6uD>jJ4%uJ;mkn&0gc(v1s(O_;lLnJvO}W=;hbANBs{E=Dt&sgJG6c^2ztvBk%k@L#iv;;z4ZtFER(177(w%wKQ?$d_eSYL zUrbb3fA;d}uCs0C7Y*B<`%+JZS{}>HRALW9Fcl;9CurLFg ztQ-*(d+lUOII)ALJpkFm{bH55Y9?ak@jBh;OFsi13t#5jhRAD?=5+SPai)RY`c$~$ zP4VceT^htV_3Mp(f`#x_!weS`NUm9tb~{xKzP)!C)ANX$Yl*AxojW0z z$&EJ&toQd`6|(wedSRLQ`rrAhJ>8(qFG3Oh+$aWSYgm0Iodq}}{2&#En@?YD4ddfW z!M29+=dM)%TPRPHTW+H+?!pg|OmKeK&u)6y^p5Rt(KIMA2NHvfxu5J%(dt{tjUhH% zPO~r3Ox6Y8#cj1kDP4%rLCTXd#$C8PZ(_NXl#jT4Q&?cyy$b7_5?#k5BvU|TO?1xAE4hOYNmNeaIw{2ohe0N}5snN}mz2e%Lev%H9 zOi#%Oq{qi=Q3zG8Va=Oovo_^lsXCO)eA+m`>W8_0%151KSQ0Hsi0@W&lbyMw&^z_p zOdgZGm7X(Mk>a9Hr0zA-${6o++RpI25^iSv_|vz`+5@W0S6@_Rf``GZHp145s;#)9 zc4{J}B@8K?V-&g9(jrO*CpLvV4A6QRl+wlG_rj-oM|>JQtQ}|h znP3&^`E35m>t!gCT73F|3VP$zPyWQ%IklSE z+G?f($BxEQe%2ssDu!-H3mgG8@M;PP6`j9vg**N2WfMsj$ok+L;_#UObQ90l-KxFP zA{y}5K-ja<{a7*6bVOq&h!7L*Q4-AR~P69rX44L~rb9P=GT3einEO9ZDQp zX&n@&$Qt+*xRnvC6<%VfHb}gGRJ1$O44WxzEZAd`X)nx6iqd%g<$_?=&&t8KYk6#? zjTUX>Gbr}v{KqA*P2iigj_wqF#qMY|u@Yemr*J1=!yX}pnlszE<0>C=nTy{OVLf$f z9&#&&oN!cMym|Enu#bSjssfM|6}+%3)awXms7FpFT`GSR9(UEY*aMI+80gggj`dPm zxV4|%m^mOl^MVt2HbRYfkv{7|an(c)QCk$xMbGF6 z(J$afeQXlAz2O|oNZ_O5<}qLnu~YZ)&-#vUu`uu!OwjFR zlIvS0yU;0Pw1OSq9*lYsE$h#jWnOtpfEmh|0BKgpyI#{J@gXZpvTZmhu(YiFTybjAt!F386)!qc)l3g0)I%HUd}&5_axW@92oy629DFukXTcRn7W;? zyv&T_ajfmrv2EP$EWE{VA#FX83>5*ud#|RS^7J)Z@@slJ4PSIiY- z0I|~Pg(!Nx9}{n4FXr0CSmgE+Oi4e(HAfd$+EK2+JBk$T{-Hq2-aXyap8#ACY#Juq zW9@^y$7pJBv;FQ4S|xOX`A(FZnPZC3p-}{UT?tp~%C~>`s|M+_C9Pfuy;9a&!5z(M z`W>?6i}b+n|we?wj7Lthzeqyh_YuXsAb{yTVwGE~oQzk$MDVl<9 zyWv7`HANO5J>Agh9m^hgwa`AQekn=Fv3ijqM=tRv3~^+bq`qA%)*Ry8gwp{MjKe3w zPaZZiCZAX>!T3{}id%B6EIDa-!yyg*^*evL{$kk?4U1nwSE3EnYG3XbC=kb~OYeh0 zW@$Uy8t@Xe;u9&u)Wd`lT4qn@#gWhQ6uNOx`_Hhw`i=g>=cA5?+dQJZSWwsT8~;!* zC5`bJw@TGDnxiiS7@_A3M!-hOZkPHXx#B;a9fwRWwSP-zw5ImO(97g|Av~5_ZL$_| z<<8@&UL{QTmUnte^Dpfk1DS~{fK9jq#Ax!vwL#)X$PY#nSE;lw^aLJYmfwWh%>8i4 z5h4@-^ao7jrwD3-vR8zEOsIwD&~?t9W^!zW536yJ>c&+jE=#z_lRWNOTZE^2O$u%# zh-m8Qf-M3TV|94gbTB+oZw)5~RH>G&Swbz_r~N}|#xb+qC(UKXj@8<`$x1?YDIS}q zln5nmUf&sB9L-qV5W~5&jtRRJ2?4XoqgqTM)H@8nod|!wui0ByK3!utzWxP(FfT;e zPGg|B5_4pD8g}{#LxF_5`IG}`w$%25)|=xr2)BV$iH*7dj=D!M8K4{?MeAyo za)O^Rh>RQ(WL5MqU7O(b!o$eRuh-OCvW5)TY5$Y9N@JvRY~w0 z?GI$yOsgEn*4Fjhf1Zk-l@z_1?N;cDt@WzG8%j5Gr!S`*kGIYV2HRyFQo7Fit`%u} zr)}dRh@HP^`NIz<-J=>>VlkF?u2ce>lBh`$u_F^e?I8q8S9p!+&D@8@@!FkR=SL_SMo_lfuih?a6e^&x6lt65*nG7hLsoP2n;d6L3 zLGgEl&iXF3$80SUwYjPFyo_y3$jQ{WDdy5$)f&n@vfhmAhf*SLn^pK)wL(lW>3bm# zoxLlKKVqx4enIxW)01`{pFVsV^`cKeb7p)34n^1Jxvf})1=gBddM;+%XUq(?23tAB zWV+S^pWFdu3C~}y?&`|_D)lj~Pa&pQZLPFcZ9lgaOTJcONcZ6ERLWrW{T-dbexo6V zg=gyB^}n}<(1Ww2 zHQ&ohteuQ1($wO8Yt2sgusY+~cvzsRpk%vcqG3v&cl0~?E)l@V@ z`DSd3=b!5pv`MFD`(}|Uu?xKdK(tN*&s|7$kDgO&ymO@o35q84xSo zrEwZ|cwssbtCfz}^rzK!wus4!66Z7a86AyX9!*@%hC5ov+#*~66O_jIfad#;-Eprc zgpR_mlfW;RQhzeM%dFu6=6Zd#wC*SJO&hEK*5|6Rn}v)u#n@=JJef2%J9EQuH{b1h z@j2Q$`dqk|i?LeaRpJd*?WfeL1mq`%aF0%L_h){q)L&{Rt&ZvMeC_nGXh-@|JH*nV zcrHValZkpb0y2bFS8iGqI-7@b*)wO#9GSh?H=nLB!A-w^TP<#t-qdFQWZp+Oe~CGB z#&>OFD8C28q-cO?G~8!d-|Afd`!1pZMY6Z!1-raSeG5M{ZW40Hm5Hm2YWz(v)@`t} z^0YZC-*nXia9+pB&drIpT}o2kdJ6cWz6M=T4sKbXU`N?_CfFrUEGHEEC#nVTVQ}bC z(sEB!`+CM%y-G*#U0RtFo6w2-x3~6-i%N`~Ukr_-dd)C|bb?5{fy-wy`V2Sd z6t=9U-^dsi;pe;v{pcH~5n=+$xxThx_q~Rz@{o)zP!IBW!Iv>1Q%&)ZR$7C>anRa)^e@I${SL$=`u6D$mc9)78 znfUdIK!qPT5FlM@{z^ zS#1|zVdj?v>z}TCmGp*O+~{gGU$5DO3L(wXuso3Tr)bc!pnA#x5Y8!Stkt zP5k8atizGW;)Z3bwb5E#972oor$P*&m=8IkZQkWa7Yrvr@ zq3SA|y;F`M*!JCq{a<)wg-y5`2WH>3EpRoq;0KNQbu)@=i2_} zXW!iSVoMwAUVhU1`^XJUWQm{|{H*GLKhRvAOyYaMTiVmjiD# zZ5$xDyE=Xuv4?)TpOi1g{EKyegxp)cZXl8AHkx~w0OTMOwT3v;86;XdpWQ)`>Z!?0 zgHyXpr)h3F3kzL{Mdv}05izCFr?&*mWZS&hG4q|^`a(3-{e-V;M}pO!tb{Q0EJrm% zGOcfIr(iYkRZb*Nx3+lX)N5y<&Oi;~J#}_pUh_%u9NGbfU3L${0nZZK< zsxIEzWBQ_RM1!;S5T~|WI!mh1eUGikokD_dZLr?=nJ%A+t>9@svRG{SCQ-MJcOf_{86A8u=C7u0FWO+j%kS6oLcvRwP^A1>%JVFR_e@YvQ@Udoc$U{)RnJR{Yd&p4=TMKC0;u~@ zzeaqLu8nx=}$-#Lc~ka`-f3<>OT%c>%TRoOynr&?`H{FYU*Hr@!E`M#loy!6B` zif@Yh0mvUWLo$2ayYGUTtgY4)fz^HMW@0UsB2D)Z~ki;XDFbiaPl%P(Co}6q4Qcr45+& zI_AntOVcq9CZqoUpuRC--8Cz@`BqV9Iewl)I#-5cWvL_ zi~Cku)3HMhn2FIbbWg295>tp?E=eI|exSmB4!X{brKh~BD4bZ@U$JWQ*HCVH>%tSW zk^CfNMy*w)!P4-MGXLnD^I-FqpJ+F8C1N@F@GT8%3XHmk<>FpD0d1E_51ix@5z37V zyYA6w=nUoEAmor%0mm%Q6Kt#fRs2!z$>U{bI?~KRxOT}|k|Kw2*h{Cnd`qlsi6*OU zWks|1heqnn2t3mrv_m~6ceT+Su4Z4_WYsoknofYotEOor^3G~oQawNcoApad5NX`V zZ+O*ipM*2WFOY+T8-B8@pOevda)4U&f3-N<5~ho;nF;zzD6bk@$E{>dq96CkO}zC* zraA`{;@5e9*O@UWC*rT2b8+uUkj+toTQ?WwljHSDxPrtD)f53#o5q`63s9!>X!`f8 zQoIve997wR?@9skbmp3EAuoV`>QH3dBz{aE(pCj;KF%{ewSo~6fUCI;aT=|!lxFKiO&o_I_s{%jO&-^c^eK3P*caQoGCH=z~dBp z^urz7F?}m8>KVV@LB(8dKIq-lMCz~H)h6s|-#I8wO9tLcOs-lA%ufey_v|5%EGeQ# z6P>Fg!)p9Sgv%x4R11r$EX)l}P3t+5tKE<|Nb-!&ax!scX0Mitw2y&Ia{WE>@a9=I z2zQeylW=~`>Epu9Qi9^yjs|px{GoUPHB{~G*MwAL0)pno5YDAV{|{|%0Tsu#tqTW9 zf&>qc;Dlho-Q9v~a19>Z-GW1K3-0a`9GWD!yGwAV@kX1!*mBO^`@TEIeec|Vj2cwc z=xVxZttsF9=A5f~ew9F$U3`rBa8%X3=YF_*=U3hVKy%4qq%9$*i_cjN-+c4Uy;*kQ zvWQq4nMuvvPf5x=;``(0)FvOwlPU(5V-6YzDvA|(s`(o(CM^Lk57?_OD;`PCkJRQD zK3IoACXb<^_LFXWwuKUo){LPI#dPtR4)(J_mV7>+YL9t?1*`FdhbC?v+tt8g_xNu4 zn&DiJ`x}nkR!~~{N0S71YVYGppdBRNCyM`4z4vIYGyUq;$rnw`{VK8Rd)zB#at0FF5^3bH>Qjr?hFCSzk4@0NDA}n>i+a>5@ z;TOk>d#OFL2oG1$W7T@bV{NA}y0N2WJ2VKN#%&XKTGwKz?lQ$8YVADbNcKV`pwOMgL=`03a3P)NFJ39k>b5ced@+$UPbUCymHibL z(^`~=iBm((u{kWpr4wJLeR#AYxlAgpr1|eg{iV}cc~T_^)RwZ((m;1?_xHpPv_=?* zt{#Wi1HZQh83w&9_J9vEvD)APH;Ql{1z^$OhFq@jfSX%tGW0?s*KQB)OYaK!L3fWJ z{J!Q_KLhBlMrk^<#`%cPyIM=rt_yqI7m=dK5_}Zce%FxK0@i{4={2fl^?}7wr2*z3 zso!b^yY>gT#=-m^z>zdBLh$`PMFnQuiy44;=X5C@|JQUL#YKjC&w~3Y-y{$zb98}b zbUo;FJa<9kS3Q%y4M$n5~vuJ1jhJbCPNQ%zSg ziVrF}4+c0?8nM+x^i*Dtc*35C-nl(lCx?`kfPE611V(RqMuEXcDv69+c`l#JAdTmB6^G%m3Zks>D z-bb=7wu3;!*VqZ~N|0im6n>R7nqLIkKGP$So9aDg@GD!!P;TuGXY? z0mi$O0H4fZTAImY<8|RZ^!?^0)S-|CbCbSOM464LqmRcb2c&~;iO-eD6^JD z3>9-MM|>GtKKipPXEX`kJ$HCA`PL^}uFa}mgIkSqUj}Np&J;}KEb9Jn|54({5c|^G zAr5eB?q-N{^&a<0n?~^L>~ez75{5ok!Ntnh^7#X5ku}T(Zh$Q{aV%Tu>GbXk1!*j`Cmp_Qa>L%k>MrB%|hbBx^>HXQL65epur3Il9D(RljN%}3jqOk$FpZvpyC2k%2)ob&(2#yl!U0IGL zZG@|GK-cllo2=d%@&Gg8E}nIXVlkpU^8xT)jH-x}srjobYP!j9&?^+f!p4msnp;g= z26M&|UZNm88fWff-}_H;;mOo3u?6WSDN2+RZrjO)w)5-qyvaGOQ@ckk;9^T8e?Pdw zvNElmw2p%hgx?G&M|^akU{rbN$OpB;D2H0z?(dI1VY1Oy!O2EEEjO+Bn=}1j{alM# zxj$Pd1hY7vZVdwK?(dFU2J}ZN!N1~C(Uv>4igm9N3aNiKrWU&ImTD4mWcm=gjtoy z=O*dJF+(gPFXSNt$NB~?FI2hKzp%Wl@9KXWH~Ea)>uObktta9TpT6lpZ}*bth?Y?_ zbqy!gN@;d_Ov;I}7btou=OM9@aC7WwXDSBJ8NryJeY1HXU7G#ARYN`yx0h9w)6f9? z-q(+ei=LX+bH>Vp^KUN7I=r6~dGPpmiDPRfEuMY zes(>(cbTJR=;8g^KnvR?&+aJ=-t&I3Xl*yv^i#LmL0$(9Q@zu#WF1g{6CR45JOi9 zs(CjsWUyJv0&Nt5n9IQA?9Odx&{z&sJ~H;7U~dx#@GhUdcoi*BR=0@j)8KQ`h5K6^ z7Voq(3NX}>I61Y4_%>h42x_+QN|RuvgHd0nO=}x|JC7?ns(B1p9@{(YWnAcA08J4o z2b7tXrzuv|q_9l6inT02XM~A#oSf#tlUKFI(A)k}ybS0$%ZMZQLs0VhD&f+-bspME zbIzajLf!6uM-?nnaM1IKNVth%-u;7=3GXX^^8H~GMZzl_%J$=NX}0aCA$(o$%>>6g zL`rOgXGSok&kr6_=~48RD|HSUo)b+PEEuomELAZBmX4~QQ={dR=|rrwH^_DeVBQnu zc(0=Zt#mZpfW&ADj6G#$tDVkmabd=h)MeyhUdJOp1>~VG`^Qd>OvCyp8BnOOYk9g` zqIVeiE+&;5jP;GLyGvFoWu|{ENfA7Z4=@x~z!%GFVS5t5p@6QZwKk<^?hnF!r1x+p zf1{)yA%b7QQ&sJ1&hSkg7c^-yao)_wUJ0pX01G=k7Ut6IPa!W|d(_ih{GbBba*xVi z+5kNt@1}^~;d86+NG&)&pF5w^(P;u50kx?Yhc3%r4I-saw@L{`8=FD({0)Sd^FXzb zpDxDE-NK?Sbw)T~ceB7qGW=pz!2@6Y@)U0_bUmh7`-!{;?SRDWq;o~#1Ho_K`{P9M)*L@9lnk-i0vEa~Xg&f&BsQsPn>k&MsJ(sCrz@ zr$OL!$8dK&N`OAW1xTMthch&r$8*9E%_?@6b*|HVw{oYSCPPoYHx-E{GK7JkhfUBjGn@cC~M$3!N( z<9167bYnyJ&(P^!_O}FX*~e$sZIC&RHfT3~pvW8bqbw0r=_$~y;blM299(Z}>6WkS zyU8}Ay#V$BC4qe$GZi2B9y2{oou?-Y)-AF?Wp;CHqLy>=c^i2mRT}2S+RA~8;7{L@ zM(77MJ*^ux&T26lL(PYX0X2}8E0APRs@tBDUNF^tEc=^vfid)gx zPSZL&B4tkYzL~!{8sb<{!g74=Z5QDMd9YfYSJfo51)4~(pUt`#C9U|^ZB=}7DE0=O z_L4>A#gKts>z8_p-&rlw?$3-SYfuer%W&w}InT^O8$b`Wtu05UdUsk%4O#4e`g;Uc zI`rC(?Fl}}a#G0xS(29X9O`q{-Qks5onqW>uNU}0J1&pNGZ$;$1K&v7AK-Zq5dK*x7Jq|DhD*dOe52j$=UyGdm z?9hzn3bmDVuc7Vb!&$pk4Z&QT$EXL>4qt&X}5Q`j&OI$-xPSN_oI zbJWT-z-Z-&+SfCc;Jgy*)+2I>RhcqMPT=!9h+X%NW55 zh7(aO=<4XE#RYtM&A9LL0O8p_|tp^*kk!w?=YzCkk1FjDoAUiWtU5#Pph?!wO_h=8QBnijUJ| z=%+XEV^4?VM|dCvTW<78czAfB;lAzm->>->LbgTe9p}G!Y;19jfR15-Kwu|zk{7xo za_~A?QTRNLxOV%an(jL3>|$qt-yOL|d-hJ)a2yo79Fv&mzV!B^KJ^3 zs-l!s*kmGRjP0S`G6y79NlT3q#a{rw{jdR)t1QQ#EO&NV6PuDE%&~Ik)f|ZEzqtyq ziv=#rZb^;XKZ}(IsL1$q%+7kp49{T}>XH`Hc)l*(G=T0jbY;;kPFG&LySwK*LOoCV zfp;re%w!0%%z_ZTp_@s1#Q}0$? zLI^18*zjHuUF6&c*!0qZrmvm)1juG8G(3e%#`D%<|xX?z%>LN!D?!GRaR*$WUQPp8UCG5%^KZQL><&IG8{9!3pOs zns?6&_cz8qD;D|Vs&KEp)kUzP!Bm!pAN`D99(a&5e0AsuW>&t=usC;tr^N30v26@aiy(+gbKY>QnP14cWF66F~@#>pi#hV!r zIA2+QPB3mxWI`R4#Z(5D=w`VqMUcbA@bwYz8?5FS zD^;M%02>Z>)M}|-=L&10HHxyFCuW$_D%XWs?98jh6EbN`dXUl2aq4z(LYdce`aCd- z(FQ;FDgEH0>t5gIwwEQ(;{~nGa2?cW1u{0+X&Zu@${89UUF_rI;{g$0I#B{U9rYxj z(Fe`@75Sk1d!i`IWq#bX{;tN4FyJJ;u5(xJclQhLd8@x`|4Um1Niq;cf0!Tz;WluI z*Y-75ow_@Nj~I?o)G9yX`!3&agIr&l+t?I7C-QtR#be#K7J#LGYN`WSh7gDP@s3YU z(h6MQZ;z;L`U&9P5-kMmL>=Udxe@tLkgxQr>IU~*JJd=L`zW~*?Wc1jo;|AGc8*xM zc&kd&=DK`QP0gprxz`)I`$ddAlg(7n8QO)>|58l#W%k`hIyN=p*it$(q!d^RGy@ zMm`Kngo`=nszY(N1P`Uf-NfB!>n(o7y=1hH+TKkJYCXPR-jr@I>ue6 z3-wKH$ppkd1Jruop{pAq|z(buC; z96uNLxd*lE)<`p4YG?ya>lE*tOX2PB+Wm*$>VU^U*lzclE=xY&Fhs+`#dwBwRdE9l z+b+APCTiD6W9}=y7YosKv~v2pTkaTuB)&kq>fE z8xWxehrQ;E3DL}a?>D06d9B>%mgzD)=lJ0DP;gqT>~GDU?&d55IpJm7dFJ_&ApCwH zb$tZVC^*Bz+E9mXv!M1%_saWXKo31rzA8Jf@#_KVZOEFMjvd!fGM|TNvkUA>?~VWj z8a~W*YYC=Ji;8&y8Fs%t1f94M!hcEu+1^esfJbFl?unKzd@|OK1TQGw*^K>o>exq) z8!^x4v%RX;0M$Nb=lZm|3ag!U<3-WD)gx#yW8F|lT&-Q_nHaME9pkTJ$_y{Z@5Xl` zYv>smCvl?Nw=a6?&Hl1Q;pm0fY*}C^tv|{t?*eN5p&1qK@LfEVFiejUvI1dPLpA-3 za_8l%<$YtA!ta%$^7OCL`RmsD&0bI_j2~dq1syTfea5+OE*sjpesJ=*mvtLTe7N4Z zMmZZA^w{9@c|bnFchw6bluu>y;=Vs*@F~#pz7tvN_5*c@Mc$Ei!aX!4sg`R?&ObrJ zxsq;2+M*`50s1e>B-_*S~Do^aeUE6b!O$ zTEf6eP4->(?sqZ4J4*F1s2~9dcFsk+SI=;I?9c%5lc7wGnVIAJg7kVjvRz?U5-OVz z4~FX@8ym>(y`nCCJf~hQC6PyfoO;_QQ8w$$v*)^tNA3IaE(+FisMxi*Y9ff+Rk+pQ zdkEj>bnrd?6Foy~wFjsRC_59n6Wr-CwkfmDM+F$}*0UdXe>L0)ROT#C*OauCWTKc; z_nRwq*C(L80|U9K?OtT}Cy@&h*lMTVjf9X&52&1)ws)}}ekAfPUTO*~3bGq1`d1#) z`YOA_iy5zktL_}AFf>o{hr>XUu4`gN>l(99*AFS;()$6c-WKp(?ns1V6an9!i;QVn zeN2Y;8q?*$r3z>SzR9bPFN}#S`EV0y<*^dSo8T=&ToCu{;&i_NR_#s=S(R)alyFi z{@{7G6T6lZ-QAU0>VUxr;XGidV_dDNL`TUaZ|@FdMSlVq&Of$I1lxuEM)Ghkc2_BM zSZQ4*B}i@dY@k*{b9xFbvb~7jb>2_6db&#D!xP)B!gH{~+14<&C6;yg83c6Cx4q5I z=~8iaS1UiZJ6I1vb={s$MHm(W04)hO!rQ|IAJ_K;{jE(bT~v|3%L*y_FEnX179j&K z5&vV7lC;B9jbYvBbC)w(oB2=Qm~l_0R1_5@MGgd@B$hXI=0E&3-J0~*DdIw%50^#m zV1&`IxVJX9FnwOBYO5STVA4SK*Zyt&{h#)B4oE)4hP8=mzU943=530;g5&a8{TanR zMkfIJ(7S~uz3jdIAyZh%=c3DTcb#3v_HFE%74BnBCi*@=$X zu6WO6J_rL4J*Vp9VksFe(KCE!@SL>3-HCjt!>S)hrH3Qf+{&UbG7`0BL5GNgVDK5i z4!96lF$T%u()yv>#@oL!SUz7w)~bBe6bQ&<=@@bY&v4;4mzZbmY&}A4-go-I{X!c~ ztO{;Re*OBpYSe0ldCl1j=Lb)o6)@C!SYBG$w(S4QZ979` zQHu|4ap0qvF}WRDoW>x%R+YZD;P(s(5q`NVFTNG@x-h<#eN){n``Pd`?em{vynY#( z*JenPr06?_>>rJdu{|m@#oX)7M<$C8+zuv-`MbhNS@ElO)DQIBe)93P{a}%|o;ZuJ zER=KNJKOlNq%Uk?k;(`*w2WIhNV7CpUKFvYx^N>t9NgNvi=3K7NPSMH*{Dz@?Y~QG z{DX3IY)tau);~3H1bG?B2lrP=X+_@UrSm5X3$og|H>~`utfKz@{*#M~GPjq+B*8yd z$9BUudnWWAz}8W9W_+v59M`vKckXJ{5&1aG3=DT%K8#?wFYJ`V#i2}VnV>rkf_6nV zAL7qG(5)InMuRu}(w|@B_5*Ia5upRME8+JbfpHA}jgtURiq7oRl$dFT1OJJKE)kL2 zvA0kME^1UY;4cb6ClxG-*VHFG@t*CA)5|R`nfeZAE8$mQkvqQqwVXF=_`}cekB6YA7b*zTwIXU(mf-c_wMEoS)ZcXJ?;d@%{Zh zoDx?%%*ldHD@8gP>?6)Nk$U26)>=Wk_4GH?!HDPQvf`0xeLrO-%+$`ck6+owyT>v) zHZS`7GXYxGu>-&1Tx4>-Mg{W+P|3!at#@isS#`bqK@0omP5%};J%TGiPjoYD^O(*A zPjQ+^0sfs(uR8PEw-p-+!@Or-vWQ+}xc5J6zK6I(C%C0!z(LUCiVV&Yqp;I&0Hi}T zPlN~eZz~DRf*A=T>&BNL1N|3FGWxWaHZSDyMbx67Kx(MMXxwTofbtd zmf-fq=7>xrQkN^2=p~O!C3*R%ue@7i5TpLCRD{1d%_p!}(t_R(Aje%hAwH4xS&Uql zKV~&V02W=ndnM(g_~RECv*GdZ@Qd5Ch46l!8l$UAG5&^r|8|G3D66beJw(za<4%jp z%2;<^d)@4s5*bguOXz6=#pU*V;_XamT@1tij?8oa4Zn0(plNfPm6bKG_c$&#cBeLq zgL)hX?-5h~b@Ow5!-#Xsk-@1?zlwFW!7{QU6#My(uyfZ@i5$zc|L5uEK;XnJ)Iq>% zrCK?It>6lDbj?Gj(P}1RmBsw`De0I+v`nz7f-R&l+z!S0!yS_lzpx4_)0z+*uQRqyqL4iLx+3@bL07>vP=R3E4g<7A0yJ%)<$H$ zOnG>C=(x||a~NaEif6tXOHO3Y`Ojy1d_Ho-FI~_N_5@}$&WNrLsHdyJ#EN3Z#vO=r zhtACxY7L5NYDp4J-z-!;($@bTn+lS6BTNF`z(G2kH5$I*vH=Y&=Pl#ol-ml5Y$uU3~uD?=cmTcE$9dFF>&F>}m zi>k7GnQNH?SNJrw7QUMH6P}MtSGu;U5AA#HcjNCgu0|BdX}2Wy_xHQjydh=aP>XRM zMfk_$T*0@`Hh(D5W;I5#>!J9!u&>KMfgLz8jU@UF98@5hvEpeqeQaXvQI=aqpK<8g z_%EXqgW%}sI9is>l1I&;S1%Lx9806c8Ks=dfX(Z!uTbeEaaz?KNJ0h*P>;!rmQIZm zf*#7xm9jHG+GPT4B8V8`&drC`QDLJQvRuK6k#0z-RV2bedg%? z@H1*>3CPzpH@klV?~s&Ey?a3otZQ(Ek_QN^)ZF$QmuaS?m~z$kNzroLo2P#@Uw=7% zF9R4Aw!&-;Nm#;7?CacjikgUqY`)8aW0smUX2hB}tEVF=75LNKfEDlV=&T zi`v;KR&^=<=C-@jeaFo5YDaiQuCRE(`(R48GgU;P#`o|N|5V3+4d_=+;FjbFJkDSI zA13yXJL^Cwnne!ONKV%O)9d~1mH}o^h=h)fjdUwi@^8GtU*_0XkQQB`>jw#w_s z@~KMuQ`D*mC)#1W4Rqyn-&$+^B0xmcPa1OSkNPv(EiiW_zJ!S zB>1&tvl;$>q|X%59T{HFcdwIlDT9A9?awG%^`QRrl$iXbda-~58|9D)|T zbc9e+vV@`Cl!Yfk`cENw1qUc<&l9u{|B=j-Xd)7n6cw22c}4^U1qF}}XK+Pq4JS)> zDO0e>CsJ^$DobQ~w?{mCCaa?86hB?ABj%n?U$*--h|h5Ef?Kmy-@1dnrnVM#iwx_Z zW6~WC*wuA0y!(I5>i^wo2PX!j7VY_Xsg2}X{{>|QCa+3=3j+;x4te<2L@L4ycnFMY=K`CHsi_UV`APlrrx-EdRj|gWZ~jA_4|xGoI3c;< za)^beRvfz%@S#VdAa_+&mLdrv<(-m(nN?FhgyGqX&*$8B=kSAb;tncW-626apDl6~N6ODzXVHgEGSqIiKB!XkN)41cy-_W$FqxfP!P8Dsb6*|yV#99PwIz@46h!loR)yb zf-`zCYu<>Cuq2+LXrTqToI*?Osd?>IG({V2j#u`hR6PKk%8jW%!Qb~-m0VRK>rTdG zJPfnhF`M{SaoAO3hH|TbQ2-*@X0jt!Kka&*hwBuY`Zc=ty$~N zB}@N``o{}nU4M#(=L2Z$Ucy#~gshz((rU7l$jAEEqjKZqHGhgGxoe>2{1;8M1297x z=cUla-O0Az_)*(!arogcvS;aX68T+Xj??Ca@5L*>awHmxO<~O6?V3)G&v$>(O!_b& zmKZo_#_zVI`HngTv+6pf+1-8b`r#QZ*+}a_4sy|#Avhs7*tX{_MVq-4%p9p9X>!?c z^fjz9as@vky5f}*xeMgbmYY%u{3D|mMorAjry)3Bid;6bf3m9cFuqw*6SkK`O3>pN zP{SDqUSXd*CtN9@DRD97Z&ce{fxm2Io$Z&HnIh6tIf8G9RjKzOD|%;I5BLtDF|9 zkMLoe)p**4Cfra8Zya>Hvy_^KOIb?ooqg2_V_VI>Y>8PJdP=MLdPH5G(FP<`={}o~ z14jA&`D6oFsK$jo`h3!ViTIO)6qQ$>%lpiRNgdZO)ub(<%?;4VK@=Xd-XBK>K_anC z317%dsm~gTnfy%kF-`tlDHW+KOjX`(pioVb&0x zRp-Oy0gtGKDxrK~y1UY>E}bj|q!v0|k2($%+hiuuaNe4kTTrs#N5!?tobTc;X!tze zUEuIc7@z3Aq|fa!vi*%j{Ht6VU?P;6Ak-?VC3;9$l~GkC&X&k@ zkF01ue&w{m z7j{!9pV#1d#|=qwwJ8vMB$y+qfGRPRR` z42~kW^#$_ESG?%)l%vaL1;G#MVsL{kSuSQa%oF%g?9o#jd04{)3-rdVhkisnbIFYi zU-cWFtI>b0SLl}Quq4cB$+XESai57QIwHu)fZ_3~%1z>)`*wVIMiC=U*xoyCWH&CL zYn>0*vnY@OEH-nJp5X6T)h)Ba?}ItSaK;dFVESfMp(5D8s>#KLlbFm;Lcl9!dEpis zqDsroa#%bk(g3ojygx`wrRE{`5KUIVswY~K_#X6ga%|>EafvGc42wpB!Ob+) zE~Z2LFh*DQ^48#?!_um!=*&M;;Zm|8j^8KnE7s+hIeyk~RJ+LZG#T0aE9Q%dU#l*? zOrug<7T_0DHIACI1#crsz=~=07+$164DE~HOjYsO7O-e2!FYQesIp}ifJl8TZ-}i8 zBrJJ5v}1$pY<6CWA&vkWLjs52IXqAqQH7UBD7f#;=x0abW(i($IZ(El#nG-jn;}8o!Bl`E;^p0;~su@=v zAHpb}2NkoMQF#gJD@O855X&@BVZr)Uu+#9J2TS1{I3Oyire{cms~Ri1*GHCIUXFVV zTIky(B@m{*&zH>kfKM4XdefUz0HY@0o*RZ!_Tr8!COztcNYd7lT08fy&@^SoEN=h# zWZ*>XZk_S2$}n$&D(lzehCcd|qT11l5~1ecP_hTek7LnqLptm?1ITof(D)@8Mg~3 zJ5}1dy4oI*TI>HjDBNs2*Ho9~O3&ufFdiJtKC$JmRrIYe!5ea#?zj*rP{=Ccy27;P zsyBPX%;wBg8H#`<+@bQ&{zi{j?2NMz$Os>enb4^!K7b5)R95s)yfnHrvqjYY5$=W_$=HOi|x2 z->@_YekaLtv+$V^@7^K+<Jz(glUh0xe%{M#D!Du}RE*q4Z4y+}zm zh_D6*E1TfgH4IePoD)`Bw z%WpuE0?a4f_?22HFcBxnl}By$m7G zGK6)Bte#0}F!)|@hjHkuoB;P(OMFryO`3eFium2g$A)p-1U&_wGxnz4V@k)%oQZDM z4szr0BGyXLPq{h8_5qaA)L0l@i06k*E1ONRUW=v`*RNkUHQICL4Yft8G_&qquMx9$ zwu)UIc}3gL1T$2zvRT@P(@-WTAI-OXy-}n*H~HF{=;$V0ba=Sm_Z_}9@n*^Ba8|lG z?dh5=Z={IoOaW@jxo1Bw6qm^DF$sep1K4_f{@x4uC(cDa?7b(y?@zwM@vsu9v7I^q z67JSLHVt<$YNmzp=WwE8D zDu{kxU^c?Kw0@O=@8FQXaAES`&%9qv;}!^91E*9@Dbi&NdP@@ZM5Sto#`5?pMdZPA za_`Dh!syeaI(@#KAcwO;P*tjk|s5Du#t8`0|h;-x82mb9V07-(_u=KkF$Z z$(-&QIS@_?;disuYWRwM8UrykTt;$x-{;#cVC0LWGF!KB|JwWK+~eh}5E8fh*Yv9d z@(ZmkYJ*6-u8RORK3U|w+2V2|T9a}k`{aykzuCv7G@h-Q#3&^dE%t1%Vn2v7pK|?D zEXYJVN9!FwS?~mTWV_?0SZt%x@-@e_&H!(8b3i3AlTD%_N)kQD*IL`B?OZKGkAIp% zQQ$qSc83}G6Q&gYfc!bY2*~EifI=Jt9RidsCcN!mu!n1q5hra%a`r-4(gE|-Xdv_(Qg=`NwA;06{)&a+bx=9faNG;)8hl4BM zEj^$ATlN@WDE$-=1eY+8%I)6KAW?jHhduQ5NJ}Pb%*@Q)jf;@eF{F`6OKU}tT#PJG zF&70WzVK3t;bJM2wgdMbB(v=FR^xr2*gW#OH9suDUKZ=MWS0iQ<;Og5H}!b+5PAIO zlJVdL!@KvXNwW7-GK^iKt?4}TR6briGqQ}zd*wR?%U})U9~1p)+^2+;MCOs$_Rp&2 z`A^Bu%q^x=KOLH9iwv%B9Q=6Q9`=T=AF=o8m+rN=`^aha){HXQ=TGI6{3UdA>aJJL zwWjr+;hJcsYD02U(%y{c$y{ovB(;2v{ZhVE$`N101iS`yvT`Y3`6I@alI=SZN#n7~aZ#Gj#a4 z)~Mt_7HYBWhj|UvdUd>hT!{urJJFBfzQva;1|5=kbgz+AZwOnfcivKusI*9i1!FQz zwe`TKRN!^zV8N-#zA=meUO+>dnlLDv`Gsn~u8Hrcf|k+PKKs?9`=BPW>@=yB)}=h7+{A!{?K0#QQMSz-<uaMkXR`8%_rz*T$I!@ zZ&%E-xK}skz82M(UvUqb5V6Hwv#Nm z!xX!FRB zPIEhtDiwUvDgqb;L<~~E=Qx`+;sR`mZQpg}bDl|oR9gqY14+vzyH|-A1P6JLe@8`0 z(zcxTC&asH29DETe-g!ona!>e`PudcxIlW?hi^6;Iv0=%AR=x1{z-pABp%j98bsj# zZhWjIIv5?DL2)+Q8B`F_fX!zv#Q!iQVeal$yn?!+OLKNcc}-U+Dz5BMIyo3G{4k84 zCmiz6bb(+1B30>P$gw^gn@>?qiu1u7eo@GpTX1RQsMvOfJaSw8YUjbnZ@6fmWJX2xm{LgJR>Ti4SRPZHme-+2EeSR`- zP&0v!#X)ZV@+A#daM8Z>jJ4X0wuqojiYGLOF}{b5TC;|Eb-RO!!rfTno1#cFFWl?1 zxKLz~TyubauS>`aWQSy5<$P36!dJminh^O$K-A!Q-sUe3r-U* zaGI|_8v6(y7zise<>7fDzwWw=H-flmxqKy(?y83+ES=r1T3W@dNMw;OhW_xKyHH^9 z*P%o$nBY?|%n_#;B&g#oB9WF3GgNMJJ`1A-i|35q7c9 z*yK_sGak`SDvAjL*4eV}?T5cYt+(^n#2r^BC_pW4mBMvh$NaXO4rXlG4@KcDJF<|o z>LvB?fz!}zh;hjBi>))y>!YI7*?O_|LjfLLAq+5!v)QZK(!WWo!_qPL?_zrhuLXpX zl+?4?8N7Gt-%&3lC{A}DKN5$&6G0n$|HPRJkogMf-vqhI+LiEkiFW>64k$s~%L4^@ zBIqzcdXr&YJ3f9y=97>e{2u++=YHxq z!ydJ1`9zUC!gJ({M}AqkMV6a`$xRsz#r#|g@UG$Z54`z5et1$x{-Xxz34+W^zeWu8 zdZ#a*eca;;r-uOUNtK%Hy%1Oa`xyM zn9AdeGZ9pSFPx6HgdoG2sk+wLmlVL_^%pIvlP_`Ii9n>ma@-)_7ed-2N?i!zCHQ3P zhis$KWMif-pap-kJ*tPqaUZlBQ#=Pww*{IdeB{A*m-le=HZJBVvpKpHp z%)~H^QQ~Uo+S#KHF(B*Q;SbMBlq%1;8`ZtR=-a**m(r)kRk*=x)y<_W*OX=tj&PT? zu1>A2IZ50);AbUOUb}xsUFDdv1S_53eeR*Hmh5hDR` z0M1{Aoi5=WC~)A;W?SQgL=gXOahvA#fe%r(n^&D)HX!=H3R&|4PCqx7zJF`^V~jgW zG!yuR+zLznN2Tc3RczPtC*0eD<(V9UtsX@v6R>0BgPw;QeVgBjXjm>p9_EQ_eHy!% zyAi2#8xZ8+P*YTwXC_MNaY>lhuzdgCmH4=9Lhce>3AGxxBg1VKo7=rCQngZDs?bJz zgk6%jNt)Yb@q`;+M5|?A2WEMub&f^+Jn+BJE3a%|$>_eG5W(D>k^DNTDIOHbmmr-B zHE`u1b>q`;?kLiZPfp`)e$_>FW={>NqjrAdUd5U$^SP0kh0MPoF;T%;6~`lKDZj=7 zypuI%D)@lM)DMY40U z)~kv6Q=;$6!@h`78_`N)^Qt4ER1vFk3)p0bfesToQ*+tBz~*8yR)2Fr^+07Uh@jU_ z@Tekqg<}#qAF2kkT=;4pZwP3bHIs5<#+C;{kuZxFo-%r^3vv&Qo6*0hm1(2aXA#f) zuc9cbWY`)jL*FCP0%^GORr(~#*#k$E^u|$8L#t4&G&x%bKTBgHGo^@#gH%NPNUK^{ z>KYqURHBQ(f((P@E*=kU9SO!x3?jM2nSLJ)IhsfH0ik2~QeHQDs5>mGato_?QrWpI{2(Y0tg4XhQ2Ei1k`RL<|fIAOB4h((_En1~v zoXytCn=n~|2ut9*X3gQ8%?`@2#gSm8Fnu3c?@n`o^XzjA$7)PmIq3E(yPMmveEZ%^{!Y(Dp5a@4^D@WPdHCUZgAc!8 zW!p8;;gU~?dootsaE0}}%P+?JGl8h_qy=ZhDHR!Cw?4ID@*^|&|567>k>a~RI42qf ze~%0=$`E}gS~w9QB{tG;D-g37PW~`q_E}F!5&gBL-s#ELp}~PHHBP5{Zg-yvoD@EX zXXQKw^KuD+y0zFt8lFfdFNYf^?Gp2riExpxByH|NcZV{Y(Do>X)*d(J(dgeaokQI)n zS4z%<6*He)IDC^7IW}ow62v1q8Hd5x*NVCKB6t7rqQ*j+KK}n>?=8cs+`5Hf1q75- z5JXy}1nF)irAxY`ySo-k1VmD4knS$&SRhDu=R%~ryWV^0v-buc-{(E&JO94x{919( zd&cN7=9qJYvvQi+Zxg`(;6yF0xqTgWf+@?WS znX37U?~uyOYJAXqWV)b5dqR95YEFMN{(C#D|@j%p{Bfn`B zS3#UI8{Tb5$z*yCuOxUsJVB%6BW<^Od;h4|N&w_l%MS&Wt<7F8g=x2;x&wN^H_jy8 zyXD}qgWsVE>ZKwRVncq42#84k;vVtd2aZ_-LRhJQRSNvvRsD({F(|;J0z|95$sZCg z==im?a_QobU_JxdCqEE$kp@hG0`FhjJ8I^mK+I%vja@F)kC01lo-#vwEE%hbuuhux z_T|v|X)b*#kT~OmPE?T56e0SKpA}1vD;}L9!9FO2yPw0MNo`Rv)w$WL>ZlWZJ1iVJ z$}YfmSfAr~>@l3zg<3reDo9zhP}a91%PA zV}$!{&Q4utubGTLO<#o*8divkfjl@24CWm!A~4wMj3ARJuT85aQ(q^YPM{J>)wv?x z$5mzQk8*0$c`G4U|MA(CjhF8_`fA}iDP8sa3KkxPj8sM{lE??{C!7?}%ul4Mi)8~$ z%LHX6rIKxCk`CH-sY>Aw#d=5jAI>C-2qK>9lgk9AbDYuWPk1VSUU&I!M-2E|EGrRsuGHF~S?ucO)`;5b7KOX&b!$ILC z-lrc1nF|rrP(uc^XVX>_MWW)+qa=rY5E4bM85Ux^NTTcy@z-2uba7ru_K;`ZXJ^AE zb{4Ui#%j`W5@TJRS^LQ#rQH4Oc>?{5B2cBEBQv?l`3xDSK}yHdX|a%s)p@uM-j%Jq z=Z^H@^H9l`1pK%frp!k2Yipy$6~zSl7E>}B_A3u`KcrhDfct>`5+B0H6A$N}p>c;J zI~sE+qu>8&>S}T}j=m*aZNCSYw;=*XMS}4C{u&7bhdM=Mud4xaZTuw4iD`Y`3+9G> z2BG{1yBH3zi_wHAZC%7Cy&HoSS3KYNAxaP}{0Q=r&P=DFuM34Qds4iZ9nAz-Chu9yaf!(VO+Tm$kk8Z&YZIqXzN)Ga0~yo}D(HK}NRsi&yGF4y z(ueDAj}p&(qmtzSayBMF%M7adN~RWOMxyw8L!34V`TLS>5q4UnF>BLyC@h}jyQsn< zfc=NZVq)prh)hf9&6e$EkyBF;{>V_jsO33_WXyvf7$khn5kDCAIT1T}-#jR@Fn#a| z(e#W`L&s&)f&7b}0(ZNx-w$Up8y1G!<2Gxy4=pAd-c^|{e(QTf{G_xiE*&C0&t02g z5V%pjLtcP=)}4@+Ih-kz(p^}N{K{b%9*eH1_A*f8)L5AwP0hE$9196QS+9i8Us{de zvlwSMXms=zGfrB9xhLG#O;ES9MXMqL!p!cQ+O+CIjj zw2-9_e!YmOemn7K3Sv;q!(UCLQ(Zq3KVo@Ur%2L!_?|m|Gl*bF9gp?nHKh^yp0;tBJA1kk8l7%ObXp&1|pfSijdy;1$QTeUU~ZvKFdrtCjcS` zaiUaJfb<586Fc5U7*q5(V2K0%nx`?~*1;pT!#3T4CMU)+Nh|&zWd7E`aA@uiyep3w!WVvXC=Hs{L8nNL~xyRfU z9bqp6ev5(ni#0$>p0pTSZw1*IcE2x{3n3`bs5E9m!r`^GN%3}>nSF8jX|+wsBh&ei zwYs9~1qL`>9eK6JVI~{VTZ`e-nqOT|xr2Ql5EY>xII|QU9}0I_X};3sY$LwMg}9L_ z{f+Kk=2!^D<8*{bNhiY`r@K}|)M^6;YodgrRCnYvG9=(6B-W1a>RtpfGAoPX*Z**9 zr2{#nFvg>h*Kp^tJZg4e9^1lM%m|j#;d~r0pxPBTThvfe#S`mUDZuA{=;BLtqeL+6^DW({zevCHNSe5N|;;nKJL7(bc7`G zYGEzJUh|SEBC1&3pfq z|ND260yIZDCbj|J&spamS^H}biJ;l-rUg2&D|C`@b`>}Vg$aThHmve=tF6X z{Vq+QoJXv{q(M&`i$pkYk8s$-F}8MS9L=RcFTdB@$CC2C5rGg5F4)+JWo z6G)a>99Gn#myPe=op9UO2lZBj4lh=Z&}l^p?{$VBEYskcu{i|NaS&T{wNZ93RI`&V zmsWnWuOHiwoj$>%-gK?+c=DXPZGa7pk1(%eei#F^2}Bl+qLx_!r)1Jg;4UB&B9fw=EnwaXtFmbe@bd{C{>IP!)eq%zK*REf{_D;_ z%1tD_8q&Wf+=ErfG3&+Yl#B=#=6c$!1MZvRKZXk~FM!4;#@s)(eCZxF8IB#rP+FB9 zvF>>FGW@4I;!r1$_>rh#LABw24HFXL6tr-lbxa9!nq27P(;)F(Arni6kAsVcJWFT| z6BeZEAreIy-9s)$eja?XgY?6!I6id85zo&2$9#p8dDJt%S6V*+9I`xfQ6%Yjq3sh6|+pW6->tWbBNdw7%HC+lMOCOt&6B(>AZ zs7X-aAJXIB`+T;v6r1g&QdwoH^(?-ahk3|Ql`_z}t@B`B_{~7P{kj)j#-QnOHhNhO|QYsXtT5wjzP| zPsU&73@#>J<`5B+YrLxSn{vTJnSu&kC?PX7b@8Ztc#*hp$VZhCbYN`hSYbt_2o+pZ zX~oSwSn5Qu^mDX7ix)Ag^RJ%Be|h=k#6(|=sOEc*l@V=fde25~fcb9X1&!n@#NKWRSgs!z%Obo%`A+t{Sp6kO%&g-)z+m35g(m2#!% zX3R^w=d+hd1A6?eWt%?o^*#=jX3Pu#TL>yt&j35~S+ZC( z@_OGpRZ~f9gr@+W$+6$R0|KnZ?;B(_U|21t1F-I21{LqdamYX44>KPjF6w=w#^!d$ke58p#RL?${lw|%!4h9`b=r8=8D9hh_#}uJ)Yfvh;$0*!ewmILMt*7mFsGb zDF6Aq+XEGLx1dml=1Z^@@9Jb{(y(Rvg#*$%+pyBizAvu~5`x-#)}@b}d};CesYpZR zJ;5vw6gJ!dxak&5&cn_>hVRf`cenHD>>{DDLERhUTfElEGJ~q zcf8vaWU%&rDB&nnT7f2J6&f+Tjpd+{eYIUGpTTA}?~xzwkBt2E5n$BaE&qN`3fD2~ zB?(p@9*hFM0~upxbrGMRugj8~h=A0K1XDpl3@Pg8(pDmas1zO5xC%D~6Bi{^J?%N{m;Qv_r- z#&Fe$-20^Ev8Gv~rl}k84Q1-ds}t0!%r+z%a@`FNiBvI$kKWAMCub0D;pSm z4`fUZKFpXE{@4~0ZexO|OiS$D40->%Y37|B4?oH_fOlN_YJP6$MBsA z>N+DxYlKt0u00xND`9+*%M%?u>J}*7>r3o+pOCYVm}UOLihg=x%e2&{Q6F~#-P(d~aT(9=M6Iec z?%J8UE(C&PMTU#w16i6t@mcpv9@eAsVB_ylDU7ghi@UYY`Mmb&^h;uwH!Ch*XoJ3U z?=LxBzS0Qy$Mit>qFO2zrtlQC%Y9!V@#{Bmhf&*vJZX48GdQi|J9PO4Uur@;C?jL3 zH>JFykW~Gs?MzaO&1K?5KB7Rzirluc9==&9vn-?CpFCJ-9|I8O-6NcSHE_E?)$&B%@J!G0F^13$-9GL7*8d1|cw zu`L+M(OCZ^#o_ndL}2Iry)fLWmB$%31P6(ax)-9%RE!$@kuqu6xXsc@R?`EXyPD^9 zMtX+4i0k7rC;Nd#8AuZE*&z!2tQLt3zX8073gOoFrk{FrRHt4uugVB3OJB;#UR%-IbFtv+@y?&`3(Euv0X|9I@e{GLWYAg&-i z!9y~%bEM?T8Y4T)OCcMq|aV z@ODXF%iT<`#tkm1D&i~-ZPtKBXfb=+$B-53xlF$I6CtVS6F941enNxQ#7cmtFUpqDRO{M+$E1=>H9gV6ax@it${|AL>B6+uH zXS+_b=lZ2gjkIfX4N=oW{ko(yHEd|@j~}l)_4QxW?a!Q=5fos{KM>x@?|-TMl`dNOY;V zY%VXt=i$)3(7fjC_xxev;G`S$jkd6~>854I+q{?0@Njn9JL^F@ScK*|Nga`a*yK2HntNF+>~n8@#hG!^ zsqpIbj+eE}$L7r!%*xQ0rjbj8>vgSJXIOO3_Mvt~el8rFq!SjQ^Kp^JOd2<8C4E62j&;>USzX_d?`RhA#*m;8I+<<^kvr0DvPOwmaDjq}!POxau z0M>66oj*$sJQ6zcU_85())OgwyBpU9fZve=ckhH9Zt&Aq6~wYT8Ge1~-{KEBn>9Su zopdxYekCdORX9}l`;QJYI!Ls=NP{P=h(-OLk;A(rs zf0mBo5Kd^F)6j4k7si#%PCQNYsETLv*v%$EJ}6l8;48}ur{h_Ye9^WN@3&WswG5pR zcIYK~qw70cHvjDDxI@#oh;-%Nq#>pXTyB@by4Spld+$5RfUPrXnbh4i!_I)fz|v0J z8H>pns=?u;FNGk(Nq`KR?aN*W`WQvXj_4vc3R8#^O~x#5 zeB@V>2Zd;wj;sxk(iLcUoIV9It7lU>xFIzm1Wi!PGV=@2dBvTy{wm&wj!Z8W0*;1( z&RjILZR+HNjWjidP?WxvJ2#I`rI2>(;sYmE%u+gBE222I4z!raV(LAhPvvu`%&3UZ ztcb1B$bvRc`Rh_b4IkpAJ?XVvO=E{x`_HB|laR4%%M0vv2FkCn*Dq&8@vRWeS|~gt z3FtC7sCTd=N`IsQC#8y$au3saz0_!Bns-b5xywwPwar3r?**v=`5De4|HVSUeoGH~ z`sorqcL<$^Yq>q0h^6bg-%uxlOY*z3f-7#U>FUBp+0~RhE4Ykmr);y%E#6JCdGe}C zCbtV>F@BR69@P(VaeT==7@lSM?q}gC zbHH;tw{7&p;%W$F%7Aj3A8WDe1bjLd2Qqecz*^J84|)z9_iK6WAH)b8p#oBxO<^g` zPqH<*fe$_Z$a5@U_d5~I_U5M)&m!PZw7w=xb@dGN5u`wbf zxNUMK&jrxeJL@8PY5lez(h}h{h31tx4It&-9KKU7}*%^;Kkm>bnQA8tj z(nU8|Scz^G=RU9x*09AZPm#Y8CM}} z_zMs=@Yv|$f&38Xdy0mowV7w@p^aV4TKC>1FkFJsNf~3{KE)IoWuD>1UcA9K2JgG< z$~{sB30VF>w5UrVUDAeCgNj>IaMZFZSwgzQFac}!o2-83zKz@tk8R18Z#k5WNmcU- z>O!oxm~Yz?KE8i3SihjG(cZa`W;mE3NLM zo;|9z6g5YU7d}XBj0wif86PQfqu1*qx=^0(Nz)-Zq8Udo_2&&kwrL9Hg;}$OSycA8 z&{W5a?womvY417=7#`!?p%kPX-^wn z^UkJ-nAKtfoR+ML3T;fYj7-L4R4l8E^wYyH)USnvY~X@lARQ2|r9Y*3C;o>l-`9jLbDrq4OuwTL7W0T%_SM4fSZf#q6rJMBI-&IgTqT?( z6Y7Aw&i7lQ1kAMYU$ZDHXsOXj#VNK3RW#-|l3*Eb=JN}n1T21%!&PbEuV;Q~1xDt0 zZr&79Zn4g+)!pnb@hW|9>A$m-eat_i!ZIr_siPps3yMk8dn+fuCW6Ws`C$GhNzmqNtOo@*weY9axs)q{~lE7#KjDj?5)z?6nv8A#VDmQ0KM z`LpIQtz{3e1?>K^1^QvOfC?o*7{%!RJokWX!1>|Usq3Qg!AF=}Y+$-yc&iKV+Pw_& z94|-zx(fif^Mqs(xebjIxbDUBghA(bSK~b=K!JLe?SFV;0(;_m@G2A zpOO~jC49f6D5-5gEt{d=gf*r@YB*BWH#Da%rV^bX==i$;6H6cF)hE>#o=|K zwj|d~y&jP~QjA?SQHar8q;5%Kf}HO%%VZ2iygpf!_`9Icf9ng0)9u@lRak5Pt(Sgp z{D=kU#M@CQ@UPNzKsl4Y3v>LJzqDxzco}RqbHR$C{Fh()y)mMcfY(A2;r%(%|4A>l zxR^RI+{NgBRqOayjgkM>tiUg@Fm*VMMVp=fN4-qlK~Pzg&VvWaMqvKF*#GK3I(n*r(}fUk>|MQTlppQ(^+-HM%+u#Y}dL!R?@k z8yxB@+2F7oO32VeevRoGLQyKw33uU*kN-rupOpYUHSt&cMbakV@d_xq+trLVN*S&; za{l+_jzgLN_nG=0_eMFL?2e(I)Y5Gg)|0N%EY~UbAtutC0-x%``_O!277~`uNA+ zJYTO{ox`%Dl(5#Q`3>teTRu*g724?eO`=?OW7}c;cx=$!Du!Y`b&k_Ez1fB>5Hkz^ zF^T(ex0SDa*@JS=h2Z4)T=3r=!lu};3IJR4LQe}g~i`H2qF80Ut>{vP$Gixn=%}e7Vxe*udu}x6yg#2&kqxQl=91uTZaFs z=>8MmcY#svV&+~SnItq9K!b)0>pykw{t@a8x9%Yq5j1$~iosrCh=!9ZnlC0kPhzg@ z=!t*R3b{NP5*XF6_8dDNF0}7Z7Y*@kI^qe{)C6pYm$4Bfo~~NM<>h*?j|dC5cIaJ6~xo?pJQBcK+OLI zo4;PB6$VC=|1xx)#BYHkVDylkbn#CywSNfakqB%w=b^wD*mRYIzc1GvbsyEtRy%qb zgpAIQ8~Jn;FmtOLC3;Y>pc8M(DKOsDFU!xrbA71@dVT}0eINkX7wp4NZnph3OK*Bv z!gJEcq(3REra_dxjsSgT8eQ6)(M3~W^9>2Aa z@%S+}J62=m*JtW8QGbzKVl(&lz~Tm)!IuDF9SA2NrLQEe&xoVWYfW}EIg?*v=r2(H z0|J1_WC8t!zWR}o4xskH_X1{)Y_(^g^UOXj$Gko~)(|3;$@;_Ivk9ZUzD8-c^fEKS zEJYminyH7fs{DP>t@As%lVUgU?(x8y@bTNVKE8O#*z1^Yc!HqeSBWwxES5a-vx#MuL#mHxQxUke|hiK>?vcnI!C-_78W5Dw>W7=#LWho zGE1(JcA4|l8VQgH%$5{l@wM=tN{}x zJTv1f9^@S2+vGHjr4}SY${aIR#ViRrX+51#HfTD<%-P_NJ8(577k`&w0~V>H5AaTKS2D}}>x=$t^^qh1{oYl{`3Y@j6h)_Xj@y8o zNbTKJ#nd~jT9P)WkNVnNZhz0=jxiNhg*q;(gHP^uV*GR~yC4g>x z!N=9|^A&jVBA_hZVyoTyCDKw<7z%OSp{w!w^dxmc?T-(#<3)U`-~K6I{}UG+`7m~Q zkbM)t9@qe0MJ1*FNJ=a)c3G6g1Fk6t*rmLo#jF55JqVw>iQ`a*yP;&fi(sJNZ7Y#} z=J*?Z7rpv#GEU;YHDNXObB7xYqLU67Cd$u@L4Sg#uN+|bFc7ioQ2>(!zrIiPn&!<# zH@nlR<>n+V^z2|pI%w`VfZneOw1XSJJJ5pmW{x%S)iChZ4cOw40eEB-U;P_&urC4- z0FQ?f^4x!fZJ2;VG@@DF&u2PicQUq)Z8=;=HIp5dWjPKjK=_B{-gNAqvm=1z;RaY9 zbAaV>ZfV>-!qqf&eo*MQEGkBSb3sKK0EKo&l>SvP_SXz)0YSNJJG3r%bHYDKyMX}( z(EN+3M(eOy$W}Hq2NTYAUG4GvR$LujjomHb*&PL8rGSzZ?DMsz7~Ph6{-+F7u03nf#IM{hMwz83h3KX7sF;>$A~m0HhfD@ogra zKMi~OjKteOo0Ih?MbABfiOJc#t22I|k#%#9@+F_XF%U^!-T5bmVvEU2%;=w1;orSu zQvt$Jz+CeNSSbtyl)L3|0YvH$_1_kS4Z z|NZC#;FGAPmDm5bhW@QF7>o%xHkf{mB)>iQZ-EH>1khtR+4;Y*`F{^YIwY`IL-QL3 zn}FRA>2w*Qw@{y>ZWODcb2v=7)>IO8$@FQKBd14m56_W z7AQ*>fdkolV8#~Py$w`s^95=b`P%hz(wM72Ed8H<&bhpVUd+}V9XYntdaOk%;73+| z8E`Z-G-|5Z+CU@rbmmZe7Ez?vT3%$)X-^f!8{q8k_wt4<{kL+1eQ6Mbg#y}}Oy&PT z?*GU#)Y68d&1!m{-7$z_INHouXcIgSM&T4%cYt(y(sj4dac2|%0Vrz zrwid+jl~dY%KnB?_`8;RWDbU$OoC94__oDNL*}!>>!0$L6+n=DZ!)C)}WI%Gt8fnGX@7G=7tK zj9Z}Xe(U8vzgZ*5VK3M1;B0Y9l+*qmH@737nX@JOUF?m2Q?sUsPkHU&tU-lsY>PLMDM%O-ihJ6mFJDq zM}Ep$Ze%<)OW(HrK)Xw``22f=Rb{11K2cvQF8ArVRcUA zEOs2MjrwRGR~7omXr1~O8FnjjoONmbq+dRV6Y}_R;c+}-mvujy*cFcr+L-Cah>96U zi7}6s9_hZtOnS%{EN2cz#{QML9^xtd)s3zY)k3P=MGfnPkk4*P$K!<{a`*LfyQZOe z{mSL9d)yoAQ^onEYo%^l_pItKER~q!*FV!(EDbx77=^FcnVObt935<9Re(s>p=G7a zk%W`wzPWq{?b|a~e%kl&C+hoY9=l;5ul)2bbLsm`%eX*MDeE1DkOM~`tJvH6{8%)X zpMmU8=8aAoU`7P*&$4p^@jS9{ zKJKTKpTt$iyd%MPI6oM_s$L=KNAIz-I8-2~^CLg8ra*z{P9@LY`d&7d+>0ZQ`paSN zmu2xAC1(%sa9{2nRxW8xA1rmG9Bu9{l<%d}Lmyb4olDKmTn(^uy-1ke=zU(OT|qbH zx`$?$I~$>_ov2;@Q@7-jS%3eYY(08`;d4gh`jqEnzjXElR~uc1=RmS@snL(ypcO%_ zpaOfxR*`J-2#*|cC%461&bU$~c1Xc=ujp=-u z9@PXN!bcQL$PaCgB;Z9zI-5yu68$=T#6Q;Kb>8zVP&(>Q93NN>c>mRgvE0 zNRi&zi1bcpVJ-Bo=Ssm|mmU2(r8S{fN0rrULY;EQiyHiZMBe@(Q7H0luh6rMdFHzb z4x7akgZ5n8$JD(}92j*asengk+-YjJ2WXPz3Y{*8%_yPx8%M5+j<$6p=v^U9YD66S zZ;IiE@c$i!x8ca>foR7(M)BS20g``PC&MzFA39+I@sFwv^AMD}6!TEKLhnvs3x8~Q zhSO*x!HjON&W#69hfkH3WR#+bG7q@axe!x#p_?wNceR<>n zxJId-@48_Txj!N9r+fl~aXF?Ugl(p)9Z-UurmM^LCNYTpk$8HnIJ&=TjI>ZG7*^G8 z_qI&Wy*hMC%e`_~8=)OrJ1y9lJ`?74C@f5izk}3Qw)2x`xX)O?$$JlfI_Cn~ceH?# z)eM6@rUZ8sdM?cDpF6#!#qY+97?wuV^mINpc3!GGI^#nBbJ!9JH@uTjKYqg{{kLW4 z9NejfM;$HTeJBvaTGH&ej*iMJ zrM@H3?LOg;R?sRATqe`I>0;x#23*<43(;u)I_qfNM^eXTslhMS#E1ij&?5Hz!sBcb z)VUguCkl!aN9nmx461WG^0VvF;5gfTD>Z&I8vQRQgXsXsYd*sHk9<#3^cc?Wc8J`} zS`Cxp;bRHxM<=thWPl`>z!>57Ja;w*&Us|u4k|k}7?nUa4>kq5@@xv#-JUfW9J!}f z17wpsqd$Iu!HWZRITAW7g{cwimY$9FXH^AGy{x0bIOERwO6vMj!vIP2@q?+xF6!#e zCO{#Nlk3J@B5*zw@jTSVgSswFTSGa&A+{b_Evf*eS?JbCe^s zJS8tej-uPJ^kItR`+jHFI=)k5M^6Q&Ndw44%od3rX?4zd0OYcUf!png-dxbA2ZebI z^s^lEObQzQ2z`yW);K+`X{mfRZ4E#7-{NSVrpRJ4(w)2S#(aSFS*rmU%OscoZQ}~_ z5dt{yxZ}m6NXF~-NfJHW^ljVL2jhgLTpW%}a|iv`we zO9flNB7OwnpAqX+$;L&9bFw8nn2q}2^SL?Xb-_$$CBg48O9>cOb$gbEqv-?LKScSR z+Os1L2O_)U1dnFA$3$olB1yLw>(!i(ycIAisVZa|=(RZO0wdQ;6v0MN7rxr^t^AYe zjEO{Y&}ON2ymqhKn?oOdU)509(_kI8da2yU?F7SilCKNQc3?Q|#g>}us( zg&Ns!Zsa{cZ>!!ap}QtaVVVpaf%4@gm-j-4`tXfy8AQP(ZtG(1Bt|oPIm8x`)ba6f zRK>2G#Md|*PTPHc$=u6~xmt>j>tj17Qtg-3qamWVBY{1hQ7Jn6iZIl`6=m*rzNHaj z_j&*AiyGqPJP{AjoKb!Cd{?5K1={48AI}dDQZWW9Fqct1)?%5+f zBI0&5Z*(Epm`FfX)Ua+X2(r8I*zcEYtA?)TJgz>Ao3yfYbGj0(fn!`McY-)o3=+yd z9tNN_ebN1V%VjowI|{Tn-D4Wwy`$}ULGxy?#hId~syg);YMju~f!$5t(Xjy7tKb)y z(a>6Ijf-r@ZT%55x@D$QxT3Wyi)s65*0(gJwc+_1SenR-3|#lly|mpfx=$AJe3Wa-owlN0S16C$JAtffhjZ03-3uV{+zj+F$+oVp zuAVca+8{koPP1e;t#+HE+<1Km%z{5|;y>v;SvTfCz|y|QZ`-5VZc0`MHEki5Q7m{0 zcUQ#qkV4d}Aqhppz29ecSAs5qc+nm8cg}`QhYQlHG(6EVBZk1ddj{QHn0tj%EKE9SA_~IGWEqc8o7h9bBg=tiuCL zo-8BWUcRHR=jEOJF74;L@;C*{6NNQ(`uJOQ5PAX*>nsIeMN59HW*NG#>1yUvFz?g} z6=v47Jj@-^@-J9PnB3FNbl#X?o&6dTkDEJC=va;KaTb5{)#?ID?4y=HLVa0G?@p)T zv9K`xx<$9@W2Fh?i5+m~o>SOU21&TY`tuU&$>)hE03&r19PTf6&jwFgq2PXYvRriZ zK~Q?e@Ns7yJRZB1puAcgeUZE1>}BvpCRJ4fr&Iokw*k$^6E{XX6}5Q#cj?R)CT~q| z+Am-pAQ-xsoFny{T1nc1Q^m-q^V#o4_SjOn0az-r@1ljZDT4~wfya^|I^U>Nwtb?s zR))CsmJY2CV`_lH82!uf;3X_v586C!NI~h=U@z|svqB$p?`Dh@$kKi9&V~=#1{piqlpThIw zjg1m|UnzVx)b%km^W!@dl!Fz0-6L^L{KcfvFIZz3B>Pcv8b3S;TC~gsQ&(P4YX$!PAIjek`ctvpeJ>;^} z9kF$@Lk_12=K4zaH+mEMZkL zVhkC+1C7k|jmh$NH2e;#hPg6)7v|26y%O;jaui*e2_B>xF5l9#t4Fo$UONK?Uz=vh zM4Rd`cpjuQ)Z`Y_@<>`uU?+ff7PjNJpq#|CCnxHbsB*G~BWrWGT77qY{Yw<pdKeU8I)kB#^ztcp~??WU(& zSLV;R1ZutoTYgzHFe>U|575_O*{P!1`rfmjjl=tjiSMb|Zz4}%7T7IDp&b1^SXM*; zRuY@h52aI;?mWDa8}c1_>hPqu%=Y)PgY#cY|v-DsIU?7GT`@jD}PO$ zX{Bw>f`VE^`}&7oC%3o3{%y`3@u;WIRDLHhTU>yB!`-6(J%^W8Gilg^eP?P@)xd7D z_PUhR5xbM()j|)o6?|VpFH%?~kqFp6lyL^w8dWw;rwOGstAon#vMl#wKYunhsy^56 zt>!Mb*g!B#xFTD6C%({(wrF=@G0jD|dAu|E6|kW*mM-n^DJUtv(w|R_e$AwVJcUIb zs1wYLE~?9BV{`clJr~Wc@{Dl0Jj)cUu$=`jr;?D)>&6556SCxO;u^e&nVMh*BKR#J z$p%O*%Um119}RBiMA&;7@NUt5(i}ix80mmlH$NMv-I_HX0-Ul#VAyJ?VQ zsw}lXJ%NXhDO2JGIr=a^$d2u(%vanVuY$b z_Y~BqtG%n^8Q&-^&*vJ>jV^CYO>L7`XN5$%5-p5+^8A2s2PFZg?bFYr=#^8z)u(UI z$k|0+L~38H5^i|)DLEh2)y>8qLP~ph$bhgQZ<-ak%%+E_2l>Xi)72%e46A-NS)b(8 z7i~I}B74(cK>|w{hlF*QKfm;PyLlRUMky?NqnY>+(3WN=dP}7^2Qj~gH>n=FQf8Tq4!~;Gii$=6^7ZJL-24u`- zWaJF{CpR$&PDgoKD)9Q5C6-Ec3NOrG&0*++}o%J#M2Gg70m0A^vS_{M!r|NWFVI$kjIiP=izQ4%BmZyOCN7GT34 z85g!pp*m$qS-&z&iaOvFuvYTkma@+9;o7L=;3Z0|2D#qXOk5@dz&%qhv9P}p+wP@n|K|g#dP)`Q8(l$oY~ww+PcC{w7Xu!`Z?4TOlugs4FZN;5|*rvRg8y z?&|9CT^@f;sXV!rf9Oj6W#VeBHl@PR$JlK@G#86K%JGn+xGLA3(Sa){-uz7Q0!y<@ zean6Xuc@6+#+>uDr*4s5I%VsGNtPzg>o~W74r#i?M;rcZ(vtJ@?p%{mYEY9yeRzw^ zu#3;M8+lO^vZ@O;j+@cBi#J>c1Vl*^7eZV#`s#%J-+8R$j#_sZf1jg8f#NewvwuaN zz%bW_!1O?&(Sf;)A{q`M2S9y5&|>yn=o#h%JX&?mqm381Rm5_HH$eb4fC(%WC1%If z3Sg(Lu5d=m)apJQG51!j^#v<`_50YI8!rg-!>QhmR5cI|kpTTh9I|&x;IOC0__IVi zD}V9_E5IK+ttZ_h0~pjQ-8(_1GsEJihu>UFV`+sIuF_g-U_-MnEo4|Bk zE&yYa@PlnVrOaX)w(r*NujbA|k%;l?!T2{)Wm^Q2Vr;@*bn?>aSy7q zTfL?MIJO}3;^&O)?@&Cj@htPDPYmFghyW<|Km>d{vi;Hyy@)o#1D)na?EunEo>a;4 zqB%_ekj=jfCl|WlfpNEl-~`^{?X{(&#tSHwcPLqhb@ma8!NYR9{_X+%6(P1;B&X1p z^J6ykYr$V0>W=CBv-S!3?wwsEm8nfEHRmTDqKz zTegwQivC5JCeFTgQ_Tb%~Q3y{HwQCdth?LcsJDeCz=S&`XNuyrpyhF_Gn+w+kZ z*!p;P0zD5!3j-t6uiJ&Uz(4N9;<5WAI3LUI==J~RK|ewS1H~b-#K~i_O&%HQ2L7Zo z5^b=)S2}9siED3}R(_h5HZ4;F(^F=Z@>M&*&IMj(bM~dA9u0~Vxto;2pbJMN#i@%l zWFJXy-DFgVzLe-@nFC(ZcID(bUY^Ugy>pF0TUPzuiO(y4h(Au8ug_3?6k?5O}*u^L;fO?`r(aQv3MIWHIxpwIZUM6 zp7R87nU!7r8-?T*2%O%R&E+Pl!%}eo#}(9NF#9TNgUHw-soF}0E@zAiP8IW}Zs_9{ zx$p5#PRv*0a)y1gqU)Q?AiZ z6u*bx`vHF>0B3YY7Fp;;q?}>+C~%;K4GLVf%P6G1uIaJxu-NuyAf6ozgqCsn{tsJM z0TpHU^kL~xLIDvF1f{#1wde+6>3ti{N_gnPYdIlB@1;Nurem28W*q-J@T7GQ^ zm?#Ftk5t2eJ9XP?TDy0;DQq%RTRO;!YmO9YaDl6QO>5(9l+x1&BXW!TOMU@(vL55s zA9lvf;PuC|MkQLZ2k%Le-EJTSs80(R7LRHU|3m;Qj89PIh#l34H4QJxz&{)KNz6F6ehkx1m>dySWl@W7zBeDg3W5AlUc90X`E{zd179YUq8A&X>)*&g zTB?c%lmF7Od)QxT5UZSPi9 z@dEGdnZF>X8pDP01%xlI;b#V(5Pdy^H_z{$-2LBJ#9sjkx7s(XF$Pu6a0)r=Nkou2 zNnc+Gm2XadPB}MRY=^LwxH4|oBZLx=XxHAadK&2uz2j18S!4~iM+9!zjik7rnThbD zY~2q|(@=X5jG8C>>hhudTar+Zs@nX5ObZIH@?qQ7|ruKKaz`hLm*Vstf)X??A@vAH-ZC}eGp<5D| zz|5`=COe%6y)U~Uj2!9;ts`g5HDpDm%Y=MSD}%HF9j?PNkU?@3EN7kvk4Ttq0Tub2 z6p6R66pS{nU?ntRN@r4VwG2+tQmh?8ATda{#ay$p2~MSFpsGmtwDsss_tfG4 zFvF@$M_rnKnJhg6DGYE@>$QpG%ey)~D@1|ZI^6#u1z7K1eZ(@BidfYqm4P{?ZnV)0 z7ks3xH=7ZaP~LW8x}X$MwXi5oD%d;rT6n24Y0gYB4dW9=30(O)LhF}bvZ=)F@7{K%BU@~iBK zvRv95vvI2@%M~umMP@$+&top#B{xKvBcZ%s`>X2CIz*dW|7f?f46Ga{ z#>y4DAl~|rzQIztf#j;hY!(rQWkK_48~gn}$(jy)beF#!im zLCY!0itVOpr{d^uHPLc+#2;NJu*9qcK#Rc;820Dy`*gGj^~jlUkcSrbC%s|nV6!My zt;TRNlrpY2p7u7UnUdM|8HOgEsn*JXkaI?H@aehD2Xj7@N1U1))UzMqWq{ z+ZMAB4B3|&Zg};0FE1Vtv^2DC)}aPDb`(^@F2 zG0I0WaVdmZYG{Yerix3*$yn@PgiLC)OI*wlgw4hQVGhv{Zj8hIg#uZSwZZLX zd=WG50hSBQe50kI%F)XQY@?9-vZC5LaJusvGPP`H10OYCFPD~O*--E&u>#gJ-m3&e z3bMR|-w`F<&3Y1-xoF~&6bTx-h>MLSmYEH6SY=cD6SwXx4`g$K^0A9v$OhRrXVB(I z?B177WU zO9X?;fY;#~7FgN67lbv7|84DF2p6f<0__lj#;(EA#^{vEwwq?Sc`>$clbdY`U=`2L z6|4H2hW-URW1G0d00#{S%ha2d;P;r@Ek>kT*B9ne7}?Ido^~0T*dh=|rmlN$t4SuiId z*(A=etKNhEC2oYNC_B8%Ax|mqwmVQ<8jEiFi2mc$JIIBQIr4sVal~WD$O&or1Bqs_bIVOYguQufeBZ6fd0ia^Ny|ApSB26sgHXo7D5T1``kch8x@-Li?4!>4AnI9iUGX- zemk9s9SJQh_pL%&?l0KLdjA91^2PqMEq*;-KuOL78#49kz-?Uwzm5`>s#=$Z1#+kQ znX-1Ic#58O3#O~RTsoPFWR#kE7+MNIDPhxcCljAME0~>*zVK_RawMTMFm%Ysze`yn zMe>~%9KS>J`ObL!ozG-$%UeHM)|FP2Faj#5mC;+kD4TTIigghpIywLidw4Ow@qLWmGlF04 z3DHe1Px9Q)Gzq!T(BN{5haC0_H7fGpeSw%MO=%!EqyJ%uJ8MTQ^+1I3X=}WskI-xN z4|6SAcXktW#84g+KU*ZQ=-jG1@b2b65UWV6s`l}>diP*iW3>9JwVx&AB<%DX)*EOB zSmtElPukAJ%9QC+g5w9yq)NE>P^RO8UM$j9a>B4uR89E= zx9gH;j1d6|ANR5}xFF8;p8+mU5@9+3TrTohA3RCoWlk1Y{5pS>VurZRgOpJl4l#r+ zC>Vh}UnP6c{Yj6nMzloSw}cvxU0lELpD_plFs8)f4pp-j#~my(qLY2}2#F}POCqx{Q!Vx<2$NJ4A(m19-nj|$%1J-{@4YzP-s>uxYS zQ*AUA3-JkERfD9)4@~3hHHUq3oYZo&iK7zveHS0wD=!Gd==25;DTJ+t_& z&dPHp-LfPGSxvl?J4-Ja;@r~E7HMf+BZI}>_l~0NbUORf00zGX!Vqk`-A;lsT9gRW zf_$L-*GV1YDljCWW8s=tr+s($^DXlGA4twc^4{Poz z>4ts$<<3aiaoV#0&ff(C4u!O@JD!&QIzv9-E6`|2 z2~1RNo8Z?&_*&xH`!T^4YE#iS^f~6a5pL4ule?sWbKrt3%SMQ|VKBi%lw_+*22Bol8hQEA;m zO|kQb=H1}!(R#q)s=hv*Hp98`-Xt2-{rBw?g(5x?3142A3Q(5ddIoo1$^FAO$eY0P z5*lJ0DrS8a>R$L{8mqZ!Iri&VR$sK(WJr+MPL4s9#vgYgJB&q6O~3qY0E&;e`pfdp z;Sx@B6yN_4?DRmY)kWt8Q%9_sA5YI}`AIhSx0b=I?tDyc>B4LfLe3o@bM*8RiLj8M z9NM_pj}_-nSnuFFaAt-%8Y@stYxBE})m*$VGd!Rf$u8X;mQ8 zB5&O_w4p5+eu9N)AhesP68PG8zO-(GU?=#4!i$fibztoMR<17b-0#-vGjtM64KZTK zb6d<>Z}j(@2S{nNdI+~q%$x-E*@XNCO)J1z3b@34OV=(Tx%l;7yNdT@YX?k+dqs_( zIlc zyd|ZadtJp=fWU|k(q5ESY6K?ibp=qE(Z53>{*wgSBr;DJMxxDmimEa2TXAfJN zB(_#TeDfzKOY&!-?^5hep_+b*J^KA(!%smB|Eup2U^2?OllYeAZWck?O=pgljf^nw z1;#@!gVaMM{s!&`QK28#p>Si|RCZ6R!cbi#T1M;%ahK- z3&frW@x|o79*7Yr8+aaQ>*{SOozv$?FJbTi`JXswnB%^|bX3qw>8=-p1!x^nFhb!_ z|7*YaucLikf*t8yP0yL~3mVp{_PVDlc7T^)Vs%yj;RR^vTb~2v*u<{o#LHiOi?t0f zu@1#f+zOX?t}u{97#52{GDMdUJ@PE3UKEL+xW}%y&xtPONAo4-v{{dM!Q#dLN_*zN zBQN2Y(WGd8`#~OYm9=v(?*Pv~x@8D4?5Budbv zuV~`k+3ob)dT8Ugx1;+4)Zw@ls*dRD+YeSLn%pKxT=eKV#;;{-r zxUFqR$c~rHE6pc$%e9qgg`blas{+vB%d2=r0UtSzVi9?Crs@rVmY}KXtwB1_YqT2| z90v%)`_?47+hQ;10M~$*WI+AYQoMzwA%D%US_w7MVV63PhO>U=`swoNLO-T0kYBoo zkY`Eu%T1y{@qg42eh9&pH)d*w$A|eyR(V)`w6^!o=irZz*dQq@TWHMAmuS{nt~FHu zFeD`CJSVcF&GGqB3Yl9cVeuxNf~msU6DdG`58>DWNxTaC@xC?-uetG^CoimIpO^nW@>wIR{OV*P4)zY2 z&-|$r{#Iz@Q`GSh!TTr{1syd;S>J27$JRXumPW)xd_Au<%5KCx75z6<;~T-pDtLlq zpy{B*LNVh3ZcUFfrKwI)7VJB?M2_H*V*Nm{mfCLU1;y{cmJm#r2+IxeOfwM%HQOHJ zlC6Xvi8(rgSBGDVZxU>2=xBRa_c`OE{;a!=2<9m_xYS%Ab^WBAs{n`MRydAbcBo&d z;TNHOq--u+w>Iu-=J}(blE5M(_>`hiIV7EfeHx)_F_xu#)b%OfzvkERCEGs6G#t`w z=%9#qb<%ly+U^lr5p2j0=3=h{3%k4%o&dxjL39*-&?UR`raA+9{l|t89J(Q9gMhZ!4C$3z3e;9nd}6|kLYOFh6P{5 zGnnxHvGr|P$fM{xq1&lg4fwbE45HO}*y}_on?)LtE5&H;X%tI}} zMhEZosDwmPBnKsMHPR~7Lok@Nv$Xr!dw1T3<*fl9yl(81%OgN-XtNy`NWGr2i*oiq zif`^Mr^JIHM<;)K$kD$+{yMnja6tu-paN4Ry6U*D^3m($ZC}q91pm97noHxqqtf4JXC5vbNSKvlCtS?Q_|& z=gF>FZCLoOn+sHPDo!x*Pe8h?_Al4Jbkq6mDTFRAaCnO?4|HuG6KrQS#;V@chJs29 zRN7+;7RAP_)07_htbz2QAe-R44yj1oyE9NBs137{Gjs_F_@99zF3M+*Sl+$KT2>7J zBMYuKkQXj2SiJlFa7gA`S0$bSwV&$de}kkTb|-TRCWm}s# zBraFp!$ea_i{H*&nvR3;esX->xh3{=cVU25x$r2HH6H^ffO1-a)KZamZ_xr>@D#fq zR5&e#i#-kr3F#NtuqgHAJ44EGQN<;CY{vpTaABrF}8 zPC-V<5u^P2>oL0hiTI0hxK758?=dMj6%F^P<6F?;+q*7AFPlWKV5}sK`=%GGM*f9_bNdqP|PQ0r_9kF6V6}mzqTAowB zjl+ZHhP^hdO zDm6X^d&AAPKTO_>c}ota}{_#wz@kd+`zB9j>0` zcv@^EPowj+GpzhSJ3MhdY{^ITuB)|tV0LO|NR z@U{ZU&Tz>9n3Sk*kCkNGIs=#NM8X&1Q(uzVFl1IIkYGd7dH9oC;RdPpPd*SlNSC=^ zW+Z|h|52w&_Q)@pp#!*1v|N=wb;rf$fenPJte|=JTqR>T_7~X$mKYZi(M|7k z6l(DH zL&T3A?uMPJGVZxMPC)8Wz zHT!%hR1gJ*KNnSMFB`zp5(r(m!K|vZkCmbYwwdR&%GU%6_<~r(^~-r@bm{^&f5qeE zu@aHKM7gEzeJ=J}8XS(adQgU*H4Be1IF&(z@&L`Nh`QGNg1@FC^|q`qZuiAJ5fZF5 zE7+*uz@gp)4%*Tz+Fx*8s^){I$qkPy8J2|OlLReg2uT-V@+R}TE`1+LeIA=bs{ z@#;6c*BxGW3ItiQv+dh59)Ql)G2fZc@U{lJ|J-ljsRb4;;il2lXnqlK%txB2wY4$2 zE54h*{5>*dPWesqZMc<{lOPPAjGQsLtXmkvNZ}liu8&>&F=2H3v3)5uK-n-JL~eOZ zvA?CGU)l%mjf!lT$orkVEeyM;6(s z+k#~)P6Ocj8ZUoTF#uIe3^p~=7uN2#_x}W(qc(cm`5BjTl0;0c*cFWR=MRR2-^T9%O?h)q>Nr8fXNtH9S*pa&*M1Ogb0)ogRy6IWPJ;(yO)==L zY!H5^j5dL{Y_cs$dGRgUhjEqbohb&-trf$qzSz!e)GA-m9zDakrgM`2c8BX)-#5i2 ze7x~!`s>fOP`Rj=o>NgR?+*^Pv)eIR&G4&;6k6`Y@F(s@+Z9g&*#VEs$W1pGfhlz< zB>dx6DJhc0Cr2#|1tU~li+k3!F*OlCTtkJf|H9Z*R)nRvLM~qS}!H??tsx>xZk$? zOnJRY85iHxfXI#iqLXPvBF*fJP^%9vO2SNv!|I@A6pxs+s=sdsM9IywD#9P5?e_Th z^0+}H{fO86P7%^HZwK#OS_w{&WwB$>h`12Fvhg)e#b3`e+*L|5>7ExKeJ40*;ybS( zPJg~}giOr$)?m!T>sln4KKGUDGT60*l4#Fc@#B#4591&7VO>_?n=i<#)w9-1&)cv_ z85PRs%0JExkle7CHSixvDxC41(~%dmwn~Uk~ zVxN;DXf_T-t7rbB!^;Q1EkHHT|k79m=4j6Fn_Dy(1P@K7_4He=QAkLqW zA`8~%Y1trZ9L4U!-!BbC69_2K;t}?9_6{+)>k81g{T6amVBs+(KcBbSrppjyxDu-A zS(Rgsnn3>gD)dpmV%Jgbg94_&Hn`XjZn|$Ay#%<6-262T1$9o-#pwfOG`#J@M+6Zr z#HSt(h?NgzHkC&8$s)dZJC+}OV!dS1*D0Gb7;}L6G28@wR0ES1f9qzztlwxDDPooe zR`&XXKx>GzAWTw|J;xxev14z?#i}8q$zS9JDwXQ-GB(oGD(W2tUZCxEr)pMr&sfkM zF8f>4EBZQ9X)Cz*j9QCc8MjInYPGcdBRzMX)7-?+@FiIolZq(DhtndV%m!eBVl-#~8H%Ts z6`EPeElk+{KX&1QuuwjSPxK`2O7HZGq;FSyHzC{H$QMM#cKg;1mvQVINffPXFFT3V zY90#x8kn~(E5C0>$Q+-S?gjOmFc`Co%%o|@w$Aiz`6m1t_l%wqk|aORDJM7GKaY^7 zB*cZRw}pBaERP2mNO~ zFkljQnS=?iEpEC}LGcj<6vfJ4TB5@d?_GDT?Xj4_?5kjL8wtb&uiKTb&cK6IuS$B` z5bUS+;L7r>t2I1@mwHuScjl1XeKqgDmpx!sFmv$S_qUe+C*u$danOENPG}>`kdrC7 zdIt-;k%;u%W=B3>!_@q?^-y8QR$?W?S#NB^?|>@>TyKm%4Mu2e%7IupWgZ;~Vk7e? z6bPB7m{&}vOdoq(qQA#Ls;mTBT7S_szeA9ljqz` zYm~~+6{wxkDo*ZGe!%!RD6cvBV+XliNq)PVlM_EP>AyYAP z?;ypiCXp2`yyl6IID#-vCi*Jo-l0rhWqul!5UUl4{^*;Rme_vyDmyf;{oW5st(8YV zR0P>ZS~5j{b`QOI_73DZsaTU~{XGdtavM)Xq2g^*fk5t+v6T{mYNx&D^>OB>zhI+< zcuC70ihD~>MUPwMx+_`U-5J@_AAn*Ztf&<4zk%MonnfVIv|(jT`0gd@8BgfU4}#DF zVc%v$HsEG+lbs_^ERqc!yR2^Bum*W=?2QP$S#;S{k&h3vlHOJSBq_ap*2JJ`?*cZW zL;7}{@Ias|K?QFuO%J~ga+&?I(atU>@#qHr{(;4pU{JQ=oEFK32T{G}#S!16Efi9w zml<}^w8C(pT=)~L`0zpp27L9GqWr?Zkb#l3`dYQ3MRI>eD9GDVt0r6&5b6kogh z4)#`~{*%Td_H4nz8)!Sp(VJTRt=<=P&)w*FPUe%k>-m2Kv5~~?%EnO>lgEFI1)(Bd ze&k=dJX|y2JOnJL_3&o@nK<(<7}Z#JQ0Tkc#-=S9WZCW(dC!0|)UxtnJ&W$MGsqM` zhO5mn=i@2}+n;LsdOma?>8=Rplhl)(+fG>#p`CQDI;ZM) z7D{w^0rbsSeMLZ32pFB?3Lgn_gHXk(679qKR^dLsD4B)}3C7oF+A&7AA)1eO~^ROk`j^4j3iB#z;2EJQHX zr7QHW^i9QA=O(}J4iQq~y-=Mu@tBILV{$H8`pH) z1cz`vl(W?B6PdbW?VrT4pGoL|N?zdhYXnN$KDGtHwGq-|KgoIEzUD$YHQkbJdC#`} zjNvjxZo}^`+ym8%a2eTd!L9#Pdh$OM0B5N(-X}Lz$%-$dL;f+B*+v zjbN(z746>+-ST_%qa9DF&P$aAX9?3*%Pic2`VEeVN&~be8kT?UHEvZKcVr&IH=Rgmc4_Wul3fz8?3011f&U-6u0MAs$*_b(*;>lm$ zR`Tj8k_E4vF&%Qg)S|yo+cWS+^M{y>2Ww&?r-KgK#xx&ckNa$-nF~HTI=B-GY%YKf z0_qE^WR2CRPkrZ%7GV_gYkJuOUEJg!Yv24hHWkmof0w|M(Kmr7!O7CgeKShQhgLH` ztGF|ROp+*cd*)kjA@22{=z;E%llkaVM3`ufX_ zbZ)P-0-!vgegf)U4vfY2DBpZ1v3l;QuA{52l131{{#VE-*jF>8&E-*5OGDl;?WF%| z^t2nJEhEPmq!{g8RX7MF)v$PtDM&NsFa2hl<7++o_4JIlrB&m~2R~)NA%6PdUTRvr zdUa#o!xAhN6TR~a3LF@=1C}!F32t9Pn8sB>SVgZR)~!2f)p&!BA&OOM$?Fe)Q-e9z zqy8ninu%xoQlfNfM4?AsJM=dbNRe&2$$^GAVAIf*%!5hS=}kz*kj0_tjK%KRj1#p; zG%w2ltDoo<1W~aLbv)P`m)Y|CW=1IQW&8 zcT`j9iCwE?nqy=rXk+PRhUC5zx}eSfaWjD0>yY#U0!e;p{IXW_-cNT`iCHWJ0OQ($ zp8BqRt8_Qw-Td+m*}6ua>CHj_DW7pYo?Dd=55aJ3)#4u>m0ITUxSmBhEQkbl!;Lnj zpBJ!k+o?TgJ_^vzujv=Q;Ax)5VQGDN?q8qpqbYJGANfrhyG_zzq`I?_3h6W&5*SR zdAGT#47mw@hi-c<_sd0bVI6j%iE@Tst4>{?k7zr?bgzDW03A?2Dd>Zfggmmu_G>;{ zgp}m*)Mpk6zty_*q|>Sgifa7uhjORtOnCt4{A|xoKMA3j;bt1p0ssM2%Ef zd^J-Pw2mdp;o!#kq!-sY{!!!9a{saOpSsXw?1lR}x)E5Au50|#!Zx@D4O5r;GP(3; zW>=)n+P4I-?;&=ah52QE2=_o60q{;PDH@Vui&b!@srXq7hdvCY7@<)NBm&FzvkWzl zA;$s!fxs++P{;e*d~SeW~rCAH)y`&BG**iV`IM;*a}+Q2rl9 z5Zd4o7WbEGf-Gi+P`c7De|fyV!+D`c-H$u^i^iEE3KY$Qq+}dH{6er%A$SI`3_&HD zcsObrXC?}avSF@bz_1W3_~znwMe?iQPyRHp%InPE+}%dM>=wjLKE3msPBD$|`{9_x zEqs-p6ix9QqGHwR#<+BMtOP8THqRk1P(qrjAx#) z5o8ZtL5SJ??(RvArxPpz0Y(rj>W-47jlIR@wa6yH7ftG2knyOrk}5bVj;?flZ^O9% zsHfqq52@lEcp!j zgQV{8y};gQjft1T-|sb)ebGp5CtysC(l_(HYBoqHwbQ~#iyaj*lcvZ*-e+7tM2hUh zlTW~o&dt!WI_!*@DO9?&LzUQX&uHdjlxR0z*Jj!&aO!pqG?xTQVY4{gQlktc`EexC zRrQwiLg)LbJ4F-i5j1X(>Cp>%;;}JA8GT%QX*ZjomSV0V=-&mb;Rzxeac1E+uR^%} zerwW#Pm1G;+J&-XBgxj9!`2RjD2X@rhY)ve+y#O9YJ7sq>{@b&ckBTFq+;kQE(9B- ziaHBa=1UFP+zC_aN=4EhLt~=Nd3IcmlQH+spA9F_G=Gp651v$2hKQ6^Uhwrv?=?Tm zFfoV8D8#YNE=7B>srItE46H&ZI1Nb`ho-PC_`>rtR#H%CX3%bb={x3 za0^dZ;n%7tdzQo|!Hhy~dtUAD+j_Po(c0K?4m(fx|M@UNo_d$@_fec`=UY{LzZGcF z;j#XZuuJR$-H9qXAlm_t(|DawGorcZ$|I)|wO^1N!OA!h($+jzgXRl0!=YxCm*En% z#p|;93Hw#qRZ#IudJyjDTuHA7f-tL(M^CP~rjv-s6CB>b> zMi8M!CqFaagcG5aEW48{?YY~kwd&UN^)N80aWnq#hul%{ds)1r`X8r!Ybi^+1V;4evxcm zPv@=UHzeiSpil&Bp9(6UEk8k8?xW4kc`gXyD=z+;0b0eqDfJ_26W+>H<2f^gKMSg_ zri(l0ky*jqaFGcUy|(L}YxSMX_ZJ+P(&BH%Yfb7K@p}BF4=TPM>D8M!x`;Elg09!l z%ug{sFIwmnBKqN4g$nzT<)7l95(;U)7vWM#g8c4W?fnCJxG}cYzH{91VtV;WsGy9{ z4|gG|0h;hT;Y=MbK8BLYQn@`2JyrcW8NfH1Kih>fFsx7#=P~kS^Z|t9h-vDBefu~= zq0R8e=s~!ahFmvO5$VySB|`rSC4_x&^UjSiACEL(83-$KSfvEB{mPfdn8#FfT-fJ+Q+ zqM453;-inw7Ak9Qj6ul@Rjp4eSWnMQ?iz=eFF7^}xobGk{a&4@Eow}=eIKFPhBJr*YS{LBKu1G_BDn*-(tm?U8d+mcmbZ$giB--lA?009h@QKZ7Y zBZQ@g02H;GkEXuel5kzEqnGm4Z_%NWIK>spAgsf{mCE zEH9d@?~1B7drX=j3odMb%Ok}9?=m~!0J7Z{kFCY#STo5i!i>gD;e-IcU+HP~%#YHy zVvo@|&j4|lLwF80(p*64x=`~Qr#k$?!B+=*xcfCLseaZLvPk3=Bdd-Wsp0%Pz*VvZ zM`^-xGFRX1;PHn7{Ui-Zr5BG;wL7bI#6YNX! zgBNT2%Ux%8-;~_hn?A(L?9LZ+PzL>v^DC42uU!9$251N{!;4^)c(vlqo}(_Sm^q&6 z8Of`aE%OXjl(k8B)=XK;<+)%r~WlFvC`1%&QKj7Nv$ z=H86un3;W=VkBn&vOvC~ohMP0eQwds3BGnEg!Tq9)kMGL(dzf>3(s#`%thF5EfTMs zN@1s}q5K*6FpxVFz&J0^39~4dy)To0!u*e?zzqWICJ$NB-zZ>L>-n(L z?Nu}_n>$F66<8|YxmL)6PUkcyzPX4_9*(;1Qy`EO;?A4J(jukmvcBurO07bnMy{p+I&T+;0XAY5bb^M9vSUT8V5V&|*{ zoGJU^Az&S@Q!cy;PhX00IUSpS<_8D`&8%PR){w1_H2EcQWH&HJlVVx-)+#9z!mW~A zl%=+MLeZgJ`H*v}weG5|J85G?=pW~K#TO~q5MAc5y|X&#x{Nk!e`;7|wT@)%htjHI z%4crBn*#+hDCEdO8^oTWoKjs^+on6)*9?Ommff$RsOZ{K${N69k!HxDg>xXS6LZbR zXvdaBHCmKrfQ@3tlQb$Pf8vb#u$y3Q(j*c3rqaB>XHq9$X6|Y?1m-T>N7HLwTc2pw zEoZxjkAB%@wM*?ZXM9~2`VbT}q8eiqj%rU?b;xXx7x!G5xj^a1#j3%|7PDG+(8RIV zm+08)6dx}glXx0c)0%JtZy4dLm zsqkO6=JIVO^Nv-M#d6)Chlvg+Pf*1kSz%dB`m(9-j!E84EQI`p;nFgd!j!}=kEIDW zA=Jxrh*7*?VOnlB;5G~u&~vZTBx3~POL3*Ct6IwM3HfAE>g4k3eKVAjzk@yI!70C} zDwdzk`;USCTjk}TQTEMq;wn7gz3i-Te3^s`Ip_pyrj)L7trhyHk(BIQ`*?|ui_dZc zH!lz8&_&7lh1E8zysx$vdGxkFvF^CO*;^{AyC&zqlr7j;ynbDsIVy4|jH!d~(YFoC zg=KbO-P6slpHHBzqDo(qP(_I!%~Z&z%dW-Y$tWlDd9twm6J`*s7oF6*t_4|MkhKa% zm%10Jz{|a~uBbByK%;g%ct|BqwnQ4XRJm_??;0dwNwmep7SWIl6E7N?>2sSmDD+$co!uS z{ii-hSW(7>bKbv@QPk0zfeMDkL$J*Yvn1ud(7G5{-8DYR%%*g(4}6@2ppVgCvC%`Q zN!8Hbs`U_34#>RqB+wQ&_@E!FBiCTgdUBw@Lq;7@9Dk>;bU-916PJgFu9kMom3GS_ zTwy!ba*=NLGS{5@|JnnXw(GF%uOh(4<)jv`ax4OoBSvax_3pzWTFq+v4S9jGyZc^r zv{%}#N_`Kk*<${Z_Fn(kYOxnZ5#V)`WA0zWf}(yy%3L=D-PaIN)%6EvdNdCgs&b$RUY4vaZ5HdStw_H=!uaz1d~`T)t6Gp+r}-z_?=S8%aX7fo ziH7H_Mqh-EcSxW=g_tF89j4|a9)(aRn7;~x0A;|ahruxo?;Lj=xm#cBg<_4x6fs@D?rblfT1r9n_^0I~!rq ztlnZar9;0W13(ri_6dGySX)_IIaHviIzBDQ%a>xGfJAph#G9f0h3VOFy6I2m3N!uj z8m}BCYkc&Xqtqq*chj`H|EtXbDb1RlH8Z}W(N?_?`?2%c?xIS|440X2*_#M(_L7ps zQJYtMf`39|QGLCdyWJS`tKQjO!AsjH7~b+=&wL=i}^$lGrtzmCHl6HgYA z^*3XN(udTz(RSYMO^~Z@xJ;8=@B(VMQKP7%o7&xr&NKs1c|_>a-ScXRZBJk@pqYIT z+udm(O=;atj(K9+T&Pg*7)Nb*6GdA5&H<>@wyJiLTD#!Ep8D(WDR}*B5`qi+mYiLe zCKXCfS@$2>F*8yFkE(%>`E3x2rJemCTydwYXS?W@yKr;esd?pRS)1j6a~wT6X^QHJ_~ zd+Iv}ckd|PqbUWss2w@!m8$ooqKC_i=l0U&_970C*N$s8ceW%8f4L%fN#vQo0YY0* zJczf0!6O$-MTiOckLOYIo5em3>`J7ljZZbHZ^`mIr}Egco)KpS{e=0vk)@Wt#8Djg z2U3X`KtyxI6J1*d112qNx38DHLROp?v1uPO)$|b}Cf#~M{~;kxsWstP6y!JMMB;+5 ztVNn>cI+|NLTh}LiL#;X&=2Ehqf`~hUouIap=+TTk}y)Tj*gZWZ3Njpiu%_|>2<5` zr-!|z+I}U&6W4uMKWxr;RJ15@Ow*5=*9@oLe2mc)VNS z>on7vaCPrHjcIJMR7Kn4D5{MgU#w@+g;Jl#U)^|dyeYLlc|11R)^Me;`ZInmQ%+kT zg@!nr?8gUyIl?D~wK#@{l_SgkWL*Jdcn?WPer9(+C3A_Xd3#{;SK%zD3~2&r(t}Za zrWj?80)kr=Y1lK|1mGb7p(MJp{8C3wtHpU|G*H?jt(v-4N1;_q26vg3yGxX9P8%2I z9l%(@Lm2fI?8x!Phdnnx;+`n?`MGpE=B$=Ya%8&mRsp&Z(-Zw++|-I&_wh z+93H1-lpW?5qr4m=x2mauwPc`Fj{OXYQKvo5Jb|G+{pdqi;&BH^v3#rJ|9x2oUib+ zWs~A11F4lU?XU3zeg0~%%}vrmwnpFazC?LrAN-MfR~aSnMO}NObF|yV>|AK~;2mfs zKOSGq;!O<}@CoP#!nn29D{JX)YvXiTU@jbBn9~jTj|g4`o*hSdqT+`+nwq(&Vpg4* z{;-nBrYz+L-}kj4-y%VnFHgWmrc@O8H$*FCt9*940={}uGX48QnwrqtSZ)A~W`+p8 z^M##=y6yE)L1F$lsm|ZNJJTHT@rfObW&cfgOckI$-#VBvn%vDiY&2ImfA{MY+Z?ar z{s_%fBM02!L1J& zWmG>po(@H&tc-miyJt6<`;)V4+*-;0&l3EX3+jShG?pT^U(ruo#nd}7$XZnIdlNK( zGAlrf)oX29VT4qno#<4Gds&Z(hW|eP<<+q=ov|J#ac&>?mDZHRxFqD52iPUMVg47p zMXD#$OU3!JyLySA`daI?US*AShgQtq^}Zx)r*}c;#|N!m%M+a# zf=&kfF}!zgq8=$)1^8)lo7)Jro?HX&7I0|J$Cl|@S+@#;j9Q%|DM^XMpX1_(y?^=O zcVuK6T(f7c7*dW8i^}ZJ@1oXH!dEFu%SB({-=#wGU+>_6_&?WT{-u{7i*$w3O z^XwG$4j)Ee3%=PuON0Ja8aCb)S$}3LR7SP^WQBpGloau{%j8Dw#vHF{&CAK1g9u)5 zq+4#PZ*kQ--2o>RJWZ?TEoOQHtYUwlzt|g65$`Uqpnd;%xL{)Ar+!Bb+}Mf;7fR8qrdgc0L#g$QZor56!0mdT&*?;N=IBH4HWV)d5^A5FN&a%GW!6(UE zCnw9xCMUVVp8a1gl2!udvuD>CMoaJC@;9q28*w>a=zP0MX@-0f)|tG`6VGLdVtjmFO@^N7OhQ;!Cz74Q2SJa2qyYtJW-Mt$LQFh}6bkz+ugOWV>N z2h?4D0_&PEKRLHK5#qGY|1s)Y{>_LijYmTumRg~kc z37U!x%d@eO66dp<&CCczLYMXnSR1rE33hrf0f+D-ol) za`_f~f|8v!(4cAGKoBFyh_f5|*;Ak;nd4)U!S&$PaUa=ul62l#Y=?NEks=2QAEt)WYGhc-&K~+0-H0*Hqj>gCs?j{8nMOq4ThpQk zo@-G2zCaJ>4Srx*9@41vf2Qf6Tab#(*u0`=QYY(cOC#<-o%WHMW(IV4l`XToeVI+7 z)JN)^gH8uA)FKHC=XxyQPa=tBW;o+|~&Un>d8+S$8)&5x6+k^o7H_4yLAl&*Zf9)3jHQvDr`23wq3;e(& zG^3%)8!b$5vdcNScn-_7=@~8s@g95zxdoVs zvKJSGh!#SA`pnF3EcW5Tn!jt&E-!kN(b}6JQ0w30s-tPigkboAi*vHHZPN)eY?!tV z?CkAJENo;xu0|OU5qcFdWGZ|0Gfef0jMp3E_3k#E^-n#GtbnUxpE=cE1*b`NdafYH zKwNUI#d~B`Brv;kEt=zpSzoKo!jX|HU!X*RTk{X+7Ecn(V&iA>mM$)%&VIRpp6^G3 zw^Bw%_VO0X2b=yLCkKmyuy(ptO6%+xG&W8bll`N{wv6I=#H9_;X}rdQKv- z%$l&NU(rflo4X-l5rxqrwYCkL?}$cE4a2p!t1=ScsNl7o01B4gJcyCIn^ds>&Z6`QCB#WlY+Wq8 zmYk&0RO4m!#o{v~j-xXsp@*tiuDHDY6F+4WHJ`dZ!Z*N1f!781f!M~l9)Ea|oXKo6)79(ok` zerv3%(k*4IelQx=R=7o2$USnEhPSw=!cH%qmw%6mV#vCJdmvofAhs`)mwfO?)35b4Yj`zsv}q>d`lod)#O5$2}>daCqPESr7mE;p+{O)`lZu8#{p~bu}KM0PH;th@$V_ui0HK z&!P7-uc8X>xgh-KShV)Yxc?x|Ypd?y($2_xfUaTx8F;Swr!Zfk`MNGy0F5Ua}Lx$PBLcv>xOaASI zb#GZ&G;`s)Lix%fTZRYC|5^C?#3Gu0?#K0h_05wt5%LSnM>cSTC+bkz+47rcSv~?9 zIi!5Z>pa*K0ZEXa(A0gh25ef@<&zr!=@)H`qPfb z#|QFLuP;A##2<=BynM7zGZ{0v*X7tAVQ=(YZ`<>p_VI4dv1DR-pZ_;Nz1pGX`}$ti z{i=>#&W`z!>1llYe{6kqT+?m)woE!z6eI)`ks6?+^c1BLM39C_cXv+|gOZj;0qI6! zvR+_SnZw*q$(@U~E zyM=l2b{jF9CgCzWo0+Jr`7pz*s#U^?Dbsl*&~Wr#q_&FR2DOR{-30GtmP*$)LWR{9 zQ0beRyQr6I3`&SPn2Yq6556~_uye`ZTC5_f%TBgXa56K|`g$jEskikbQ*gL7#qfD( zP}*COv33^Q|3bOZduPgm#@MgebyJ}o!Na7rFoQhnXi}+0q>HMSC*37O$P)s ztg%R&g&iko+^m>cAN=RKp>JP4HV__=jdDkuCrWJo96e^6s;#0bhpqEzrJ3n{Te}|2 z#W7F~JM|CW|NLx5r7xcnH#NuU!?Wk}Rea|SCxqAS{jdx^8=>?B5mCmH;pU~vPUJjF zL2P@Nc#Sv9cVkocZHsfzVL5f49uh7)v8mGF6i<4+`vF9&%ratXf57IdMsQTI%En}c zD9!Z#+40Y$wpr$I$#?T46?@YkEJq7!Rvw?tt?w0W*0;|I)b4iV2T*mtA8Ti5>Fj)T zFe!fg?c4PGh@={iUQlhz9XFd_Y#~!*ua5-znE-0BBl)Wnqs1JUD_h%+Yf-b)d9>C0 zCgs3xo{QI6tkRv4URz^zU%-ksGUe6YVs_H|_^?8cHbmUP7j9?$fPj%~$wA_lA)nvwyTz)XdfQ7ax|;`VCwfn^3a&nE5WOYTKL@EjWY=@Rz8BGrZuFbLX=~N-(D@O$i z4xZ*HX(*m3aQ>9lrBsTa`Z@#U>d!qnANx~-=N7PAM3x=6>RF2$+8Rnkckd_PM|{`i z3#spqM{QcYRJE^AfB}4Cl)5f&gLCHUE37e#w7^$bQB7|nWzTqGO6=SJ78USlLSqtPjSyiM%Wz6wa{P^u5Rm=FXO)Xt|Mf|>igoQT3cV+!OTTEjY{>M#x79SQ zJRbRcoau+RyRaAcEPcFvWx%8Wn)aYuhV^YL1#o0dvH4&D~g9mM&{;$Hj?7W zL^Yt+^tgkDN>-Oz6{u^H#gloh-Jr!h&%ZJ*B^O}qS06p{L$93^#KUDXtl?-+(7MLS zMP5wg?o0=@d}VRW_Z!jD>CV{hZPhPbCbc}~>#G*MrqDND4{H@b--9Lxoq z2a)gxN7EfY8mqfDTPd1n2`ap2OHaazg1NMjVbyt{GMqS8^Lbj-V{ffJscj;h>$Wqe z_94{g{i<3`T@eqr$jybB`wPyfK^=opuQ}YkK?h9iFsxc|J>W{=U!(v)o*)QIH8ja(BWQpY}c@kdSMpdk1ot2io zYObbN**Md4=sHYE(_@J3lFL|X8FOg#NZTUl?f-;4AX_8`i`oX8xda?&@?&h!Q)WL3 zgvlK_?=&fZF8=d>QzH;U1hj%BdiGSbKHFV3Eg+IhnLpb;p?@og7#!EV>pW~XxrtK` zKcK&(65d#(q6E|AaqHjww2fgPBs<}7&%|p+6KR|2T#?kxJI58!8aX3`U3C#X=Dp`_ z?qX7BJ4N+VmtO5EW=fIrXD31guM)^l9QI;HxbZ;Q?-jB%v9o=iGqQt53U!1`;E>$y z80QzxHa-_u7}y|GGrlvJG>AfqY2@b`iIgZ-@e1jMo9nGA{(b zpP|603{`vPl0N_xm;AjQEM`GvSdfw`o}sRH2XAnI?(s@slbY@jEAzM*{%~S`G!@i} z{>adkadWpX^?!IE=td07agbJ{s@-|C&4h_!gBJYFpAVGtrZhXfBP>uek?g&L6j6Rh ziCtHI{)KBKzo?6}5SVNE8eu;$Tzzyf3sR9*4`%UJ`Z(dEKubmmzw@#hYN=0jeRf`e zt^?l4o4m81!amG;rw3hac?o01^-<|!V(dw}>YcXcs# znfdE6Het#Y0I6J70AtIQv3PS8LXD-fV1r5vJLpP$C*dFrSV&4KFf}!k9%>}V;^N`} ztns6On6RKg?Yr6>6|IGr3w(WwCWdx)GGUn!FC^H>n-`M?21Z7%r%`k)uV}X1vP4dyO}J3H6oC-)|P z?7r1KHWrViR{G3-DHNubx{vL_r&PGLrz-1f^rp1^y5<ENy|QN`JfNPiLr`#8syoG{i{GFPt|6dJ{l7F4FiZ^6spw6p+e(1W zya4ynb#@?nPH#$~72Osn6FZ#CDWQphx>$~7{TtRYU0uiJeL&PZ(!)vbwUyc4-agN4 z6C2Z1W;FqwI-bv$EjDcaW+c>-{q(}-Gh^Z9^^F#A^&WAndVXF?OJlOz&WwLsfs0Ah z5V)76)8c*94ep?95?sQMsNW$tRaq|tEb9q$bav3htiEs5AZ5pjnrE?7xQ@uni-bLx z;L{0>Y&6ayxLs1S$vO}6?P~<32-m)#>+(h{Zs!3M5k4sPele#x?a**D#I2Sx zPOxTtcfTdIq^zt38Ok>zpAeELZ^+~RDQK;G)5mpG-F3O>*u`hxqjyMafvVLpdb%Vwzh?_jJtYj$1{U_Ea;>snF^;=j2`r~(wS%M&jzakEiLov4vjsX zgM@DJeA{YiZ}mXE>LYm5+yRZUv?g$Tp{C0f=fwH-qtl&7V}q{kSF=r;W)~om8?_;i zwA2=OHCaB_r7qX~f)m1u36As6c1Db8!m?)%^iJnqXYV{lW0ugyBZ303GDyW z8deJ;Y2Et@QLOroj=5h3-6|TsX(~?={xHT>JUWpziuHrPag>{w(77^{Y3BbMutR;H z;WE->TY@BPb#;~JbTCvk|2mzLpNqbE3o)A}0kqW#J;1%`i{*wQamPRxKxwK|nDU6l zag{n<4cMA1eJB^zsI8HiG}s;p5cGZ$IYVEe>l%Aw9A0ewH5Kobtc!8EVvQ=0?Sb!f zL&S$(DT|Y>jWG&XPe#3IF*|6b6z4kH-FW(`N$+qi78EzcJB@$F>%DIEIXxwtcLAd` z)3SCt+>TPcP$QSeZjmtm-Y;CKgn%J!5!3@x=dDRFjm$-~iMKo94yfp9QYWcjIt|Wq zEia^$&d%3r2W9o!8?TwZp02p}etXV|ct_Kqydr&bz_ioQp2<|L;@#h@6?2NgOGX^* zr_KJ#Ajz*=mLbVtDc>u=qj4Glpbd^qTdXyYM-->dU^V*(Lq`4Y?>7G0YP9VE)EJ(E z-s2%Qt)J(l9qxZ^HT?9Dam+9+u0pfKi3l4a>E9h$6e+r@ib4JWT$m!(#1tP-8wCx( zs9!!*WmVCAlz=Ax&lBCeL||rN!S32H^Dzd;#Em_2{fHM6-qS&goOLG` zr}^y9y5U?U>tCT(P8u70Ts*6M#dEPf*;RD4xmq}6M@R1SF5W0DeOzkXpuLb?E70ID zJV9POs6Cyam+O5$?F};8JARxUTgJm{8t)*W_M=Y&*ageB0M&a8y)6RWniWq9OKXh8 zu}mD7Y5OtsZG3fc5gF?>b$3f8PxO74%QN%IeXkqjj*|^R3H6g$X|auS`Z|GqJ=TvK z=pOpxc|5$oTOgO_p?Xc8dSrl-!^K|~dfTibLI-W8{#kL%yB|P(R4xRbNrB#iXy2HV zMU><%Ii7F!*OAG*M}0{^!Z*Jq_WSk!ax=#%SopE(+9RHk5Nf>52Nb7TdZy>H(E6zP zp_bomRNhZ8W1|h58o%X0p1U`F!meRMA-g$76;I4MHdQxh#Q_dL6pN?9&?odb-?HH$ z>YetTo7AyBvSQ317jF(-GANrAjNw2EJJ(ucMWv(+a0LsBo!%9|o=^YWSxL6RzW7Db zSasEH{wBIv3=1b(8=hj1R?LS!^*m~=Dxs-YutfL%o(&V10?1DwGY@y|<#!)+KTWI7 zW>%&Pe5{dtlzM68W|-A=n=_X<{4zlZbZjheEmE>#Op77IxDmP=JFs4wN%o%=MlbU= zq%k_He@X(wj)M6SZ13z|y%)tLvaN$iMRR5v@&-FhV@^ zCj65{V&3KWpPNd}@fWf}gInqF(MYgG@x7v9=2$6%$Bf^$&a;|4e}OmgQTER7)z-{XZ;W{(li*u-?Dtlt0t>wj$p{=rw>W}ei)p+;xK7{KMPrX8PtgJR zwRB;8-Zyp6nS7N6QeT?#>REPS$A~{RY^H8EC2@y3%H>t(Nv97_mCdhcYHNon2Mo@w z+!+H4Wzk2?vhsVSI)1=wvH*s)WI^xrfcmQ8>i zVAy~^Ql!-sJ#z(RXn8TB^Z+l+7;?Yy(>P}RyUQU|k0~$R_K0n{+eYnzy=}>d(~ENh zDg&>GReV{hB{RJ-QTVZ;+F?)k^cVn~y=%u4K!aLzLT986)KDYzla9m_Vj^Q==Ff#; zDQFEA0RfW$=kL44-f;mUyL*`kw)5xDA9z&L`P5vGc1HIM`AD|+=J!r&!f*MdP0?RJ z0c|o$9RDGgZ`89%PL*zE^R;J#_PZBZm1m{TLYqZ_I(PnZ*+jXL-RCH(NUvGcdB(glB2`p{sf1CN)_7d-iH3u#dz zAU`9Uo?mt{OO(7|>V^nN(qT7+*Vr7aweGAPybgVAzvwVVsH?U1^}~6O;&k#isoJ%6 zmD6d_NuP~$-&Hod7k*x|5A_yz7r71)A0t{Wo8z1b3FhxwPpd1Lu4 z!DI+Z)sj{x3Ed$VML8Ajt$EY4qn8c4_2O`iq7LnJeJ)N+l1v1cE5VQq7}8i8efi## zCx-s#Z{G^*#3=1M)(tQqm?Bxs3v)+xNXn3FP9;A9SI%wmgzUd0E}*u+k`d=wplzX5u8`?%XC@=axyWFkSYaslOr~2# z@GaW00-sBXrDMEsxyR1i%`&aD$k0)|Kbr@rHZuJ8P=jp<DnhA5+i?Bs=eyZibut##MB;@11^^ItX2bTT<=XS_-@s=9;Ndu@56^l9jX zKruC)aMiKsm%jAktp=Y%wPrrDCi<4KO%AT{;fX7ui+gL1+6EL=ww6pw#Zf;k*z=RS ziEo|7)t=ROF51*Gi|_X+eobt-qoz8#AM(-dwB@TxscFZLWT23swo$U_GkQh=n(#+b zt>=KPmPD+30Eqwr&dxz~_Oi09x>2kh)P|HI04JebuJ@M8Qb8Ll_GKh{tw}lbD5q@Z zScm_BzJCWc$r=-9#`-wKDpmOUmIDS-r*{~doZNuG@iundeCTKzuOm}Z|voH*3hSqJ9ACgT?zN_aBgKM zZp7Wb{-gNbf`#%@^%SQ;PqUR4|C^XJuuFOs&W!Zi?^kC*nWjRgVc|oSi#^k7s@oUZ zz3yq)ZY*F0ZSDVj;5-8ff|sfzVvMI3$OOz+3|O54Y}cir#mHGu%=Pk<4LFE18%HNMis&W^Qg z_1H`?;hFZ&xt7$MGxND?R%ChY{P65P%c)6b6DgKoXR3}KP}g~QP(Zz7UrQZ3CPA13VP?MP#GB*c>_gVZXE=c)y@O?3#Q zy-^l8*-43sElE6^Jm+zpu%ZC6>z&@&AXl5-*~A(1n+M7@XMXf1XY)4v&h5VSQzyAL zEuTnMa~oT0H1;;nCT+gapDpVd`9V9TAEJ_~r)u}-xxN9#4z@i_vgb`CwMS!9QQ}8! zn}fXGJj19yqLq)KgU;3yROF~6CoLY42(kiN+NEgm;pv^~ zHZ7YjjwNmAMof{NqadT{>5DEGQg=*f%c4XoI>-cj*>!E{2?V+rnB%vX$Mx+~ zLMOEAidrWL2UCG_x8TIaPsiuy2L*dMyn=;YRZe_D5WP1~D3;%LgyMeTvE=l9pghh;t$a0=;S7cT01ffXUBp?sPa zV;yM!>WPO-n~moYjlQnQMEjl`jX+Ktt{T8ukGRWLitA&uoYrrie0o3Rf4>dkTH#Gb zLh@~>=8(y4J(7IXBlqR{h;Qd;lTB{^=*jmF_YNAw_w(oC0Ory1se+DZJ)Gev9*jpg zooWsBJ93QPM?$SYd;_|U`AN(Pz_+_#JJT&WsW_e-=cEKeKr=t*nl0{(XyPu*g?ts0 z*dLzV-)@dvd|LbDHIx2i2ye2+yDur{&!5p zKO+D=qLzywtPV#}Ms&t$Kx;v29m)MKOyUI=sk+30#h-JA^&rDapz>{hQ?@Has0t+X zuB8`+OK{wYwOapPwCbSp&)SpRdUw$wEMKTHubNT+IG4t>Ah9<=^788k?Ndn^DHdLm zQONJ#oiR!h(|NuJx{GQ)^A=6_?@+nzFTDw*IctYEe`d6Jpo9B(Y6Ad7i|-*q=oG(g zB#g;T8W?}&J$q;nAi41E4gs#8zH!f5Kh!bB#;(FKD70dSEKojb;V8dyugS>UdoshM z801)&D?9A+27jK73e<8g)M^U!9S(m#8S!yI5p5aj%Nj2(d?n`BtF>0xb$gM76FxZl zlyOoKw@uQMZ5tGEpK8G!wH_S48Z|ODXPQ0QAwq2})M~6J_S8&a50aD$B9HeoQ9PE- z4=+b>bPv#lrp+mz9;#OXz0+8cE3u9AsD!L%%mzLN(~pJu97DB!!Bzfd-=<)|B~)u8 zlm`x)(W;2^9(&)`3q80`w#^bT{6cS3>DqW9$A7KZ0RH=MS0bxW>GautFedXx_-tw} zL$u&Sj;AIJOZksCr}2O7c`g1Dw|baYH7p+DDnp56ZGj@mu{w`K#|1^ibac!C3RU${lCRu!1q$mbWWp~#Ri z7+t79_qnw3HzdWIy!%9>4@c+FN_)FZ41%1ZFE+ZQ-`||{6jQ>zP{7rG0lJk~eeqC^ z9{m}2f;Ex;Mbnk0MnSZIV#FnNi#GiOQy_$(Y#bTCx$>GNx_Ks`wU8MPYr=(g<>M9W z&ol*G>dVK3UvocYps^xVitx6(rHZ-R{gnr3Zu6vZK1W)N@(u+2yQ_5B$S zlhtwCc1!t}*OZFYIXO8ZC$u%mNs|55sIktc7C^_si7@AiS`UUwS6o9}3KA!DFV}>; z)TO+bTB)h{!Wn!S;6hS+S<`xyZl}^>xRoVYb)6^GP4{hAmhydsvMFp|z^Fi!I~tLV zU!Zc=Tx^M#TV1ovRK#3#+fe*Mca?5yMt-KY#qjRbhrkU#h9T;dKg7d7#UvIWxZY+T zQI4)ko8CY)#6qK@^L4}Dv}E3*W|H_DrDc@C3ceY3@H5sGRafgQ_bTUtR#dhO`R(K% zn4`roZGgpdxAR#1C~E7L+4E|u;5ZmVcwtAsC;u&7F24E(BYd>mh=H(<5xoK(*|vyT zZtxeK5w;#m&U7lp>bk<<3LE=u=my%X!_+gK?<7XWyi8SYy0-(-DLvOVtekV$+FJdD z5^knd@^L(it(tTC_-aNkTlxP*wVR{$dn94j=rrpXv@ZHM(80T795jmb(!ZoG*)Fi~ zYw*w#%1KosKkoG^SFMdYX}kwW%LS*Qh>3HQznoq-46+OfUm=VXP#Rj6^~1kRm?}@6 zI>w;*8biPXxOA)fbM*{;kSeuw9N?@1>KyRgBlB)W_1e!u?nMz2;xXY2l4gQ}0!hdx zf&wUk+5%4Sxwd^a9Ch~yAiCvY)wD-rbgFOrPo4YEJ)$H0q44&1#eg>p9T6KQ*Xk`# z?x#rE&kkDsSL>_w3Jb4BU%ReZ(|u=WXBO2)_Evc2c&sS?C0^EFsOFczk3MYv7p2pbgKGd$9DQ5BR%H`{m@k$T8%#8}k=U8l6-ee} zF&*5V%S74O?r@asCL~xT5W=8^8MGeu7rNzBwpr?w12^S!Xt1K5lI;Gk*;K!V)xmr| zyYHeVrl$=68=>cqXh~uj01qV$D z@qOo7k#h=lmD9Z>eivV; zILx@pypN)tie>xRJDS;#m~zB_3-#X{wyo>=Wt>o~V2TfH=S2Dv|ysEtuL=^lo>gs}Ee9nse| z(cx8Vi6ZWE1t>2MtdetJ($s!MhTH=(@7rHXd=EvjG(m4m9Dlf2UCgb3MR8;3xXfGWg6Gn?V4TO{{p%)8Gb8T5VIwkP+U^@(vhc#@wc{J{C&JMkOjj$xq$Ml3NQ^iBsiY4d*1aE9zo1C78TR3iRfHvHz;0)pz5bAEp~6hGuB|By5pkTl{N&(DDMl%(ZDdSY&x-dlu1bY774pHzq_8S#D z*%G3E&(p7WDjev2T0xdy#O%{62;QD3SJ~!dczQ7pfP=a2D_H{;91w_$Di5YWoL)Zq+Aq=@=teWkvqYVRC zx_x;#|4O@%5$W|r>7QktCSCTQk(K2cU9m+Mba=gA^$#g!4>U{Kl$DiL_Eburp>O1K z$ij2UT@GnnIbnik@LPP>gvOaors7)V#w&!lPyi9WtX6z!feVg@6^Ew3R=|5xmS5!1 z%(;A%RQ>rZm1CxynIc4}Rht9flp?g>5Z5K8BM^Mw`VpzXX!8HPm>FPrO9* zWD$&bdo5$)!kK~VIiZdA03m#d@a2E2_P22s76;C?rImZ%sJ%)q++pD~!nU}uz|_<8 z_>!adJ*&2=`oxL0oT@F$8{eIGvaxkLqo2Kz5mZSK*ZNAEsc&_@0xsgbL@5ErLRk) zD0Kcd-m62SpqKa7ym|*7{`OQsoZVs;xOIY*l%@g{Ow%?x zDdN>lOL}T$F)vM;dSYr$y=Prsw+E75Mf+UDEpllyv}PpY77bGaI<7FM9Cpf3Yd;gw zoQ=)%5(44ndI`sqhQsSSJmwD5v*ZMVTiR_GT=~26b~&Xc3LmsTUNFALkP#ypvuhms z(U()}Q=H$2{|`}Jy{=+gxlz^j7YK$2pH{4(>rP;RT3+kicFBJWcWaRPi0F4V9pZh(o}iyxrHco$nWc%f?~l?i^AL4pw~(@8ay- zqfuqB|63ElIQnxCPBfK=n`Hpa67!}2qgnC+4VP7-{Orev@hUGh*4clrL1o>RVTW(E z`e-7K`U1CxCYN*+6+DyVQUQ*7SZE7d8zxckO~N6>QKw+rL#Z!U3P57N8jbk7R?h_q zV2T()x02(M=?U*}B=;nepWHi#Xu2@$q<&&Oz7`qN2oQnxIFhfTD6h8boY{gzA^stU z_}G0b{nKLW9;OgLN?E0ZvzBfOq}J(c4gUPibp8D#m?Z`H0fo0^HT^*K4Xt$f4aP)oQdinKmk$Tf5Iy`mRy4M~h95% zfdh5j4YZ_ILOd=yK5POLmCRBsCrV_N=kD{enQ&<lptErN12Bgs_rnLcTRS*~Q zvDHr?YYB~_^SVl|R67M^AbvJ~(fHpxv~nHpQKlPm%j`VbRVMe~kkUIE7gp_$TmdZl z(IuU9pI;XPy~5cnV1r`Xv}2CU)^%@aOv+9IB7xLxlO|1`e>%8o$x2c6#6GLXM@p;X zsQa}NFCkp3KCmBx{RroR#?L>4w7j`AnMhaIW|gaUlUFl61PEMms^6WkIk4or>^ zY0S_hO+@ zK_Ju0yLx5IC#*ccDqJD{@FQ(^Q(0XlyWon+?~_+VS`qw@7Qml-pAqi6{sbZPdBvUw z$VAxx+)}MLoBohW)FEb9p^$}M>R6rW?$hdC6KqY*vZL8q`_)bzi;KVa2y<6*@Ui3R zc}mJS@SeujM}J6m7t?UYZHv5R?*XDmBs99h;^3FaClD?p#Hmk{7GFwlF3aDI z2sMTaX2pO@_Ey!_B{CYl2S-uyn`x(t=jgrxc#=HXz0Q3CU)PFDh$}PAY zTa1-}WsA0%aj`rH$NW*ad}P^SF{dlHDbb+32Vc+iDJ^Hr51jh!Za`&&*+timqqWxA zvkI;dvSU3z_d3WTyhC@`@-xy(W30TncU!uMH>znj#(@`gL%jXx<SwB_rKbyr8CZ zrZIDm{!*jd32SEok7wZu&u>cmV#D;a^-nob~XU}+pG5T$L;J5unLBqys zyV<7dT4|N}OH(HoOapzjlwLR)H`p4#wQhfXS8fE6?a}&G+}q|JsYp*s|6^F40(=NA+mudpFVe8U@kv%qUbaF9 zv=Z3^ur(fon*ax1Vt3WE_8Iwy|J9y^&EilaRp$5SybcCkbi>7gtX|~V5!u;;&gvtC zrKRD`xPf+tHf6no%9P`YvRdYR@3UrE+q8k+mCaV>p?SrcQxXkzZF}+eni#wH7JGMd zg7HXlkHt#usoLD}OR`9zS|YXlsR%WnAB;V0Ro^E!QGI_!CtfzPa#fh5FARFYV~?^1Qo13}iHc%sPB=l5|8c%ebvlA6I<~mV5~@A@Q=SE_b(&NMq#N z4}=qwv#yoU7&2lan}0b0i~&xiyYi&CkKgIKHu0Wi9#MJCt*JXW3JcK0n85IlIK1%!?;m z-yJaf9Fl10Sk$ekJ~S{uW0y_&u-F6APVDAK`KXdq)P9tdWquZ^EoZSw@|ijoe)etn zt6}bhE$yES5cULn&1Tmi42pcy2iQO@#_ZH-8}{T9>_FpWEe3!aREQ#aII?e zA)&-=ReyD6gfb!`ZDp007>S98Ov8LnL`y!yLcKyi-XaO=LnpQf3fGxsZzE3>4~_lN ztWjDp1Y!-TECUsvr*Usa0=1s3WL;R2&!mF(#W-G(8XD#`u8b^tLX@6b;ZBhIp; zru_?v`^Ov%5Dkl6$*A!&557843>e&Ep=UOmOB?2-QokR{Z8^M|f2^~E<6Gq-*HbT2QcM)df)B!~){$^U3N|{c0rr6TiaoL0 z#nh{DCKbYueACV+5~3d-?GzchA?8nIj&8JP8Y)wHr2)`Vqrh57xqDUzj?kJF%kM)a zi52aH!=L_vrc~VfSv;+Q^=sHhG2c;(tZ3QV`mb+#p#|%s^KK4{!x>?BxO9w$VN~HH z-UZ_ZG;_tpYL~n&cwD@V<4jg2NW0MifQd69c*<~ zLH*8mlHK(=SggIFCRV|@x=OPZOqGF(p-gH^2us3POg?-Ox3wi>{+7&`nByJ?A|ibp zjYQFT;+l4WEc@9nQql91es~8dO}B1sjag!R3g!h{4sOsek9$#z!En4vCb>L8mD5}O z=lU$(d!{;k6+^!B z!5A#W)W+$a?*5R~^{BW;^BGynZ-N_}L3JCD-thB%KV|1L&Fsp{e@~!hMAtP}*a|Fs z($|Lju9GzpX>4)U@^I-LMus92qdOb8rh$ZUyDwYswbK?pJKO|7y0&1kF!oIo{n0yg zeN<0ifVHmor>OrJC0)&&I!s!mrCt@;i^#@hs7^xs*c7_HZ>Y}IWS9(f0&oSlmW z6bZH-IXRy?7T+DA+H$=jR-HE|FU~LXxPaBH@U;g0I>oONVvPP6-GWEcZ_*C|r_hZ@ zdSkb_t?0~DaTd2e=00jo>q4`)PhLSnzF`U*0-cZ^eU`s|ZnUs3a;p*Ncc`~DTU+h%<~(9Xbd`cbF5ykyE{=6w97S>QZ07nfe= z@cp^AFy%G$^o^cR9dVjUe&!llc6?46BDdYiVr@je~+?0xvVKN z9;>Haq9$WvlAdC1YYPjL86&w6{8Z)FnZ96&NW@t+z86h{+W{qvJf0xlUAwBoh#l5q zhT;W+80^fLXT)m;4VBtCm7vqr@s+`ze48se4jKSa94kHLf%&acI4qj_zV|Ru{r>>@glRIkxCL#J5amKq8#>0mU9kvZapddd|V$rr4hm9!=Yz* z9!m|;^PT_Thw*neo;%^8ITafut5`^TRh5yxk#*CucwMLd1IMhJVrby}FY~6!5UDcc)@oRMC3vXSDN}65 zZjrFd|GQ;6-9u2vMBlz?B6UQBHCc5(X<1YML(SxVhlh)wl~a=_MVMkBQIUe7+R*j1Ya&M{e4FNUa)RUgc6QACnb)ovoi&RW#U3^;FT7sHs33F#>me* z#nuLljGs@{%xORGnp|r}XM8`j;Cr3dw!Vfo1Y1v?On*$t({9{q&59R3yH{b%{fMHi zNRs!@vVLKBCw(*GnK3(YfnxxU^v!sBv#Z#NEH#%ghw}?-vf0MPsN)!-#U0Dej0B>` z>a#3YTu0*X*gwTAj5~oTGKMo7T#f9_#sKF+OoB0e!}fRtOKLsM0TnC++D{b2#SIki zO@K&bve*b&K-XC7nCLptwvlrp1?kf06=lY_r>)*si1*6Yy`2=r|GD!2cs~XuKi^oP zkN%|8ZqjzihAxl&Ta^u#xczQZMn@S{u%L`GOT7oW+S?BXRBMuQw@3PBj|*6*?$Dsk zQV=ABUg)O5EO}8CGIsOX)~T17S{#i7^!k zLZ$WK&Vx??f>Q*W|Q;x&z7!UPFK1WV@f54?@t-?T>^;94^LkWN-%FjCQF zI~k)tjXD0&y4s5R-e1E!zr;yl`$UNa>}eYz=y=(CoQ)TSo9Kg0_Em&5C$57kb6>Vb zAt?QdPHgs3Eqxlm4tw){4Rovn{Med5t&+zNnLF#Emd!RpN7gKBTDA`Dy({Tx3JqWN z=6)DpB{B=dxqaD`@P7{};8G%gLJN z2~V;8&VoX7-eQs%`zz9o@Kc0Xh1*_fk$81t)I=p2yVI{L`0E}ry<-CdHO8UZ=b<{* zFqX*NOzpEZbbtgxS!oE9-|TZ<>84Aj2<7}YvIBVT+{BpCPs!$39==95LNkzLm;m`B zr|oeJmfOx*oYGa!ggC}J<%=KSlvgJv^8m#alrjXeO2_*qnTpvPv7%nK7Y*#HKKAgy zmCF?yy8X!i=UYy^Ak2UBuNTu)SOx^@;f_S{XIm^EJ7)$BNKtWl1O&4W4CWu^V)H(F zloc~>Z!jvGR3V0$G14zFy_U%=z$5o*i0Iy?OEdEvlFIR5xc)qoqT9wXUa;5Gh1O=$ zjj0@ZYT9?FlVTd`n(bwk2D=ssA!yO$Uv?4ZiW8QH6{R^sPyuGIUNQu&KPe1n2cbl` zL}!M}$8*SCui*NjDldrEF_U3Jc_)ZgtGIK%h+nGM|BD;iMk9$#OC5*bl%q}@Ch+PB zz%wD<5fKxYwbUC_+oQIWy>C6mZL@58XZW_=!R<2tWrV22%Ex0O_E`0^9hc%^1Vr`W zhxBZ<)C+p7ZOezapxn6Ly0iks)0H~!|Bd(l@xjj-E@as}ad&s$GPVeL&WyTJO2ST) z`e`}ta}c0klla{LkJjA_xwgK(wY?l3ELd|?|4>7&67Ngx(domO7>+~iaS`Ou@yY4h zNS~~Diib1%d}A6@OhTO63d`TS{z9&fqmDbgBP!vAmah?D?z*w32>nuV!v4Yu$9Wv( zK0L!D%ayD1%lB&&y*D&+IW1|y|-GP_Ici%G`}Vn^s5H^ zbVHy~*yd_Jc6ZToq<{PkdYummU$E5`-Bq%8avFGIoM!4{w;3sAD_HCqfoC_DbMi zKtyvLOJOKt}Xx>-3RhtyFg5xfP zjP#4@+aUptOeg}Sfx}q2kuhfe5c{wayz{{d~(qOGvtT_yEv8@ma26a%y#9tDJ; zL|!LgPBtGwyy#BquuW90_iE{Qjn=+SbxqB6;H7F(&cd4$--YT3U)vG(f0de@Etgp< z=bJWj@6S=gkAtzrI1*O?5J@=|eEC>fV(+cAU3|>%l=nF+-1Wh8H={D*1v+1M!u;$W zrJ)r||qt!v})P$(zsf2=1M zs(U~VrxH+e;SX0AqQhl{Uc2=x&U&F($KICAp?kwK+~6z8MV2eCX(B(fj6>unbcOln z^ldRDeF6v)lzzq_4;8Lbq@y$Z`>eJzf<@~7TVMF>x;@IWvD^*%XcTA__7x}k@{9{g zj7uW&W8@ z6BDRi?d?!(z8L(B>e`5ERU9U&04)PrQO|Gf4>C_{bab>LQIuj4A^8*U;#@ZFb{+zw z5R9C8+Y>s`*x0cSrkM!IXZ*QwUv5l;IK7PF)e!Ak3~R>^**DDz^kz%C7g=?g;lFEb zSeIC?(3+5hLOrpNN?3GnpVd6I_bA7qLw+XmN#LQBc3hx4&2ASA_2jSB`t@!)$rok);OsC@PSNC10?@K59N8<=^HKo3Gup+2e4i7Jk>PKa zgW{+Nmmk5i@L#vyoB|W=7t-rha@y4H)ufzOq<4M;tk;y9a(xg4_x9v@`+7y6Ytz3` zim_$r!>2@0j-zg_k-J8_Ch)bneIOzCY>n*sqqQ3mm~Z*Q0w;URlIW-Q!^lp!G=)Qn z?TL}$;741xgsbZ&*u`LaV+Mg~sW>)4G|tkQbbFwsGjviV_&3#U=L{ZD$U!4yJ=N(uzYmJK7e#pG7N6VPVO^9bZwy z%?5&iE`p(Sho#v7iiN9tY`KcdFbxPt&UT)YOCiOI!X-K%S-W5I=ivV~Imz`95|V;r zP`_putd1+&e7mFf)47?&WMo3ubZKWY*oI`RaggMbwyQv)-PSy^SE|XF-Nf+@b-CFyydrmG}G5d@;nAUkU zQRt7kL--P}=Qi2#9QkYEwkg+R8!5staSIkUzc>xmgUEcMYV~RFU`2|z^|@Tn11%ottWcHm~mB<7vtZ7 zMGcN$37fw#4|5)q{S%#9SPoDE~Ty2!%XJ*Trn-lT28%Or5k_zl` zX;-vI#%8q}v)HatV@js|H-CWeg(;-?IlHh>eD-Oi5PUQ7qRg#i&N}CuCJ5M7Cs;y>lv&k(KOmvSqJ~tdl~>$jS(1C!CBp=8=>f zD|_#~x9s2R9NpdB-_P$Kw?{X}`~4cvd0p32K&N7q@o}3gwIfmBX(P`+596q3N;^wM zIn&wI7KGK!Qf~4lKua{#v(>$y-LM2zaci8iQTLa!Rt|(#^EM4rAPP1BJ42_2FC~oR zbuTPn2lgL3ehey#|8 zrIWka?sqbJm{MrzvFDOXxmXL--jF`-20#kg+|?`|kg$#0?pR`fwVP0fIv7*Q;x(#y zPwTPNNY*b}A)lLX?vEbe*x z8`~;+>9GFlXb}Fn>4f6!$24PejrJfRTrjgW(YU}M1+)9BYGw5ds;X)yda*Q9qbQMg zXp3IZ;Tn6dwm@#dueZV9GmG;&`4_`m!;5^xFT6huG2#A%*BEJ+Cs7S5^4ln-4WRv+kaeh2)oI}W$hzaX~kjNM7uj?z=hr9v&P^HD|y234m0d zX`>ko%|pvK2Na|Qz;K(|wf}3Q&%j}X2m^CO!MSt{W=c;HQ^9gxYJB9m(52ANFQvIT z^zrJ#Zg?2`_lWx{ir14K{+;&V8Z(+#j!Wy;*;Gut5cx>07wS+x><4q8@;)zj+R^lX zfMVrk9?h{@&h@6dO3;UA@H*a3gyW%u%*{h{OUEdCtMODyLCjNelD50fu1azR@SmRC zeB$K8uC(-MVuxml1)yTM5XaGo#dOa?^1;Vuw=awB+ z&P%~7r=Q1pc`fAKHB`~8{u zVB#+xqGaz!N$EkM4z1#tU0Mwn!5eQq3vr@|-aZe&yafTV)ZU@L>rPIpgOP`e$QNei>KN29AY6}>C(H{)o5N8WMn>Dx z*fh(n6_&817HB$mY{~Gdq1^nzt*pO7$J2vx2G^sE-aiyYiakr5PdWGvyt_1pHt`_JD0ND{}~3fLq4iA^c~_`iJ*sLDAe3(Epz3d(V>+Ji|I=jxl|JvJ z3G!>h*p#8m7!0QpMs-L`z5$@Nzl%edX$tvEH1?wYlx>YQqeNhDZpB=k4ejp3rlJIS z{z?Y2zMP3EqnPFdP9dY>aPT+s#^ZsX&pJpzKyvd@4j7LeNA5^mq6IE$5`auYx*QMx_6AuJ8rQkL5J9`tc-j-2UU` zS^uAUK44ZQ`QV9R2;CU=qT$=OMW!A))~f!QU;fJHtH5B{4jRWElA$n5U4UI!+KgD}(DA!S+&nt$<x6^!LK#_0q{-)>$v6!sWlkry?r^be1ljACtVgCKS`=ioV+w#m)*4x>@^4_m|W;nSfUV;dr7_#iGYEeaJ$y$}X^Jx8Tuy z7ubt?iUJKhOyw~43{LkrZDjdiF|CWSeK;3iNZ*h45p$7ZX}#)?H-ifOOtsD;urf&5?YuY+*>o&nzJ8;G#3WFv+ZzN@Nwgp z9ot~@26ODm#im{0_n?lrDz!J}t~57@&eFMQ%Iae++E{DA`A6$iFEZZWj!?o!XO?D$ zgy-j~->$Zo6ES=snsP4c(NJ{IX({Qi-4VKj@)R`@l&cdzjT~~+(#o>s1M$l^@G_QoM~?Ogs;w z&5?eVExRs1A9>*IEZ^e@C>x0$_KZ|tV2`1ivv&UhL)I7^qxAH@b~mTLkM=ClQVjpU zoj^uqfY=P2f6RZ{gt`wnLr=Ia<=NOc`#shJn~pVqm@lNtIAwsi0ytcvKOFFNFc-*% zQu)k3o|77<5PIXjDL-7mJ8IPK-aKO>yCp(MCzxg9MC1&VkrQHH$TGpZnwmaQCxtQ4 zUPQ@W>V}v5BzhZCTe~-7w&(=lN0kz^EHyS3pQhB`LX9mWCjFh7|heMG>g4+~t{rhVC@#9H4@*L$b4+SozjLK-GM zCdl<4Tt&$;lXdjpu*KMgK8U=^_Kz{gB!Na^hfAq06e|NLk3 zPXgfhN?yv*4A)@m!VD6Ru~6H=7%<)DsP{nj#j>9lmfmZktj{}AWWZ%rU_+*gS z);R4>^(8RpLfEiT9@E&SJlS{pL~<3Z&W7q>oovQ ziumG+iRW@4Y%;<*T5mFmWfI)F)X5BSo>HO*cSYZaetMCx>-vOCp09w1!`ztINmy{% zVzi>Ix5XSs!FU#WnvedyFKVS2nnbh{_BKFCad-M4-1+>~2gffMNnY{)x%D6oxE8JO zedLug5@0)+kY~_u5`Gl@s(ZEBKy!+&d;So`1y$J`pza|Qv^U0+5=@hryx=44t|sAv z*2<;UQw#hqhS~NCBCEvx1Ky~6UEAKS7;Lz#AphM8?mfagO;`!^G=9j=s5@biwg`So zF*s@&cq-N9E&L~&g>4{?3C;pRTB~hv*>ysI9_())d}o0$(9_TT9NIKkiK`Rx*=TQ_ zqBQJbSBTu-+k<`Xb}`^HuUfVwU}G_FN->1H=Dd@e-9!748Vqz3bHK1)2;!haRY*bl z2M+X&7HD&XX1c<}deC<&2oFQ6x*f$wPeGc~pX8)i3pu+vply9pVq=E`alN}#cejT^ zlmxK34-(pzz`bXKT@+}m_G?|9#j@f2Patf@S7K^#go9U(7~1yzn7v3cQ8M*Sk2U!Bl|!SSpv8|mT;7rwNrl<}dQBgHrb(4^2(N&Wp?Ra^E^KV9^qL5St2wOm3UDdaRi-; z_cKH?FaahbsTh3goF8L?z8_D*pr7gSZdsAjBp*cdo5jV(#kK_PHG|^sEqeh>W3hDt zNy1aRezO5BCo|cLt}KtwTp4=CF&wuC0 z{p)Q{&m$b4ILBQ`eA9|f;8>hV~y(w{ujuIio?Cm#$Y@D zYa$>%F1{g8h9`-~p4TRCzGa91j^VVYr)PIDLnis@I^8Fm`(Y-FY#BX#xUwyp#t~q< zq0qUQ=`f(Lqa#q=lO%Opg~1Gdj+mZU<#M%ah0}J;QfrQ`<}<0n$KO6jIUQfs)6HLT zn^+l^{2wEAxD@%>GCJ1GbFfdSnr- zh75yO!J6tb5dBMY^G*d<46TFga%0$Sl#-mzq#X5Yx!Bt{4B9e;axqD~zIVOqAlKEu zLzK~ClUVAtutjsV##b_d=%RIR+AY<#BFCabE1vbb9Hx1 zEKJhdl5>b3Wpfe?NSa-2DKf!ku3wnaY!++?7I%La?!FNuA2*AD5}m~2flNXy-|A@R z8yv?t6cKITw87BKPPH3^JFncN8pJ2s9Zh8<54W_ka&7!GU6@h@s$=?FW?|<3eHbk` z%)7Z!{%Vma*PRu7UU^V>GE(QL>i_e1*%+9O`?j>Fm*5(lCFnw<{wgJKSDy@PJ@r$h zo(SEmH{lGpXAR&(YlXh!HF1zYRRR~FJ8W%sl&E@7%! z#p0hPpR9*@ujjG6Rwh%96?{3N3gXYCna!!p=+=*6>RY*Ht;5SIc@0|{tU+4C*W+Ay zr|hieKOC$}`5P0aNd1n1{#bpWLkr>7?I&?-Jq1?fp~eHwWwcb)A$~b+<8v)tev0^K zaS3o0P(;tUeXSB>t$!@O17odd|9;-V_?0kQmv5KaiwAlr z)Z~7QO1ubo77ESe3#v-7 z)^M!ZGB8F(q@$v0H=Q~&{Pq^^M>ojsDv&oAsWB$_`YPq#w&2R*k{fMFveA>SJ&m8i z0O(Y9vC*4a`t6@X{%1E|VZva7-?jKUGK4*;n73oCK-zRH1!i!^f8u!(xETnmo6V}v z+PLcDLb-R=X3oyOdXKAvKS?Br@iC}7s2$1S$JTSy5oh0*p+VfjBFF8e8q8^F>mdyn zIcOUGnsA^yO?bQApv>H%IB<#5mHJuk4O#H304;N{3Y|Te`A|7^277imaa*f|=B#{D z&L{;bT;{rv*ndBiL3GvbGF3k>OMCA)KzHgGa=U2SRhklR>*-cssTjrilz)&?CE^zT z7f}(rooRdEHAPNnZa^>}jl+O@fk9$!ZhGO>RSZV%kBLp>i!8#^u|VVM(0qm^51oG> zF~Pk_UNm|yakLIoHS&id@LX4+2e6AHCqJ$A=JC#gBGZwKq;#NRnt`|Zc>lvbqJ!k+ z$zH>vhmU#Pb|$ef^?XKH+mp8l4U2W`AF_G_disj3RVDYuN2?`moR8{3Z3gV(&ayH` z^ygy`R?fP&r#Oeiaq5Kf?hHKRrs{Otcc*X%<0xkwptfNoaXj0HW&;20;P305&x8`? z%y*(oz_b@gJ>kbSM*WqRvQFi@FvTp(+5fn~7Ain*vnU!#Krg|gLEMQED&>cQ2LS+QH806)FWC8(#Dr|oZk z+k}5xIObhutcyb_jDVWE>~Z7z|9LNhUpzq5a`gSyX(}KDXW<3?(b1~zl66N+72@}8 zb%)E|t`=vCUT-z9XzPis+-9VIAk%|5XHt>+Ru6leW{AYja-Xl4Y_CC1+8y>+o3K~! z2>WIv^9Mt`5*mLD)uOai=5iJjCp*f!#c9+rWrh-e|AP4@?g^)DpW>outV4wYp+5Tg zC?ENX*curLQ9hRkxSbsAWL&g;-PfJ;d%o!JdZ=h_3e7DpsVDM%CAk zbc@?6^aQdhFqY^j&5-Ko%D0aJ3=+il)8GMrn@KeT;!h&x6sh#%YY;9FpYBL8vGOk`tMVp-?fro?g+ZlcRmnsnwbzX{5WKB4Gi5u~bzt&yG|w z`fYvpyB@tu|J>zu7)EK}pCb<%H6ZPet4LPHs=bdjzB~I``P$`Ra7H*VErx#g_TjUX&a0x>2{O(w8+j)*VhDMEz z=L*Tqa?@l#BK(4AN-Xc6^hl?k-PP+W83hv=?Mwl1)@?(nB=2?q?HP49c{|qnonbk< zYRAivI*Rs|)y3nbwpxy|)o_OkMa9K;fTFi5+TnOWChBC;@4#+vf$t=l)^B2MrY50Z z9JveQERTL<_V2p_`$LW6hAe4RPI`&WKubPAJF0X~*6ISBrw2+8F9X!_6o42nT^)}8 z8&r|z3F)yBIbcb0tEC=kx^3n+1>v4_a!ar*`@CbyfxVwXx{O(p_%J{(<#$5Y#M{6^ z6B-|GlEbT@nTz+Y@X)|=W6v<61WqnL0$2?_1Jv;WFPSz~hVA;ipfuPtz%HUahTCMtif!`XXP=<_fz2!AvB)91EiDPD)_ z%^EWSYVoHFvc3#|1-D?SpvAb2uzqgOR_xGuHBorI1*iZ|4zvvJa5nwBl=m;)>k|mq zuWyWLg3Z&w_8iBMQ1p<{e=klwHHFuma^M=ggzT;)|DHd4f{Otd zw5<#)22Dms3zQ+!{w@H3;0?^Y{kRAlD$@8ic~PbnIM9kxx8)ecQ{O^Z4(gjH=dDH< zlb%`(vwZ0=ABN9dh$}v)lGcGz$iLh9?lCKbL&^Lz$q=3y6f81|689tH(}RPFA<%(x zMZd$YJD^W26t^~etsYrf5KB8-_2JQ9omb0p^kL4A;fSG}TU>BiGSJa=?%Go)Qv&rQ zVPHjG0ltQ>0oy)77U~=oUy?2dgYU@Goh`!o<@pOUP-+TH)bcw-AYg=%gPAY*x3MBF zX~G}~8d`nRSbDXL*6puGiI5Sad4Z0NpP3Tit6miF^w@+;qcjSI-lKe-PtV7B@*4)A zx&8_?C#g_eE13Uk*j5t8fcAh?MNnruNm>0JV*tG`(zHOgtG*$`Ji%RHFrH?XhMxJS zYj~Z2R^lfY%ejgeL1Pb76-vva^F10WP^NZB(33p)Q@{gFFfv+>`R*dkE|KQ?t*=K_| zdP@R=bl6k=Aa!u&%Jx&4L>Ew6JP-da*M*X{=@Gq7Se`$%ue0C(+(gnCM;p`i1(VzS0aA9|zX-laZ%8N;EFSKQXu(RUQt zdUxFd{8b^pLNYM4sQjKQ***BdZuYgZ#_q-y!*G5tpOty5|AXLchv96KSK+57cb>K> z)l{)*>s}~>qEk=$qNu}o!P(0FN`*ME__3 z96s{T^m#Qq6#p6dKKggTYiOt$J~l>HB6kEw>wTc6jhm#Zjy;})H0 zoXikg;wTMz9%kLtx=t=g3;CPGgqBqAgY>0XB{bK@@r#J5f$bz9O;GOoqy!zle8>6M zr+|%lRcRK&_*Ih4w}?NjGptDvQu8N?JO6b?6Kf$}QCQj7M1$O)lkKg(P9*3XD$?Qo zb3cGG2N5V6&d;xCsYuk8=ft?+l_gD?vH~4RltA`3Rq%6+D9{bOk(olc=6_XEs9O0< zrqYnC>AA*!2p{aq4l8>X)2slVm7cxQFvE^UE>yAXqy3{)u1O4wVhIwcbVD4Fl@&4I z3-}BHC}%B4l-0^pDpe|IHU&K?<9bkibKc>JZ005<*OuE3r=Xs4rC(v1(1&-Vp=BKM z)Jds^r85Us=I<|)_PvKn82<9<`T3K{lzew8;`o>9^kCsJv|zZLr{7lVqd$64bhXj% zWaZ-Y#TFm(-4%!RkF=O_Id)C3Uw(o4?3ps*TUfm?QZ+cHkQ%nM@Jjzbxgrf$d#_C; znaNL|OzX0=A7Dt%ilzNjrhQJ^@6s-G6OZJM%Y`q%4h3mkuhAGf=A8)#lZNB`jaC#k z*pD^b_pvmsIwR#y500l6M1hN?G8x$Cv z!9R-$HO0fd9tYrqq<`ig{<=)Y~OZ}itOXlb7s`TS{f0webUW^F@eRz1tc*oIm2 zGPOl{s@kSi-wEjF8{lSjB;LM^*JOUHrF!6TKS>-bv%GFFOUMb$8$p0ybwu}|TMm9c zrN3bc>}vt4Gu7SVllN|E!q|m3QQh~)+Jgl>R=>$IvFF;e*~7yRB)V29C_9zh5*~8T zSIGs&?JOI}DIeL_mxa$0z^@a}-G#rNJPyab)IZ6?+S~Ki?AJ9YZERQ?yOy}*{gOE;lyS`4$zN$+52)F` z`&%aJ^!AbR;s9x%MKAu-l${3m!qOnuu;Fkx3!k6!LjcX|H8*$b{uJ?7KVzS0i^}Im zqGacn_ilpE_ksUv=2@Xd+f7~kF>ThQeQNVaufa7EIF-83SudTZQ9`~EHQe0`tix{! z?G#c;Gl=UP%40P|+X>s-z7mTzVm|-Th#6H$W;#KO*~WBS*Lc{5Ec4HtP8TrV5f~^U z+!RmG#-9sAyV)RcDyc!4g6?E5+aWr3n==Xig%nX6e@QN&-j?+ADQAX)TwGaok? z8nGbmmW~KC=c0RZ=tKlm~LQliLV6C+$0Dfi4%Y0th>{M*EYRBZt04r z_?3zhN?+~Wp(ldsINA*xv}uA6>0?&KhQKOu(;b$aw5vZkAwUdpE!1%Gw)2fPe*L_H zyF_rk)nl04&6Ci|jBJ^1flV{!X}k-|E0GcQ6f1^r^ctInoVD>xq%TfW^ttSks4 z%*^oIqsTgeRSjND|GC*)-d0u%%C~#X>zZkamA`qe9}BLI+)gSkxjdr=Byq+dt+D2Y zdIt0jrl<>?ilq_#ldz??y-nxdiCuWl+y#kPwsr|S*~i84zHnLpu5pmd__v0Kv%5r} z>p$>a2MZE7GrG-hvM>?VeKC@SlpxE$wEXc(W#!Xdt{?5qwJ%^s3crCBr`AUZKK2 zUUg6zr2X_}|5Kp8zI@PWejucZV&z(&fe|2@jmd|psiHO7p@DDjf@bK>Ckw=y8#$KKIrl!OX@4aHxapVtF|B{6$iiEq`!gKuZxl5kKxKXCNR8>z~VC&-v`Gs!rnzM+fD z%!Q!@f##rp<4vYFVpt$kLI~Q$7^pZG@S0CAuReS|&xPnTfjuPJdh*TOwKB0`W1!rR z`cfXPidEvSn*@R9eI%yGGworvm~DZbLlIpqifT=_AVRL@5)H$a#(@@rm_^_+;L zDBs1bEYaQ70c5qAK3crmSl#o^eVO+D(gcj%YEv&b?OAUC@J+C>p9b`p~*(-R4e(Dulihv8@25NAdI$?u2?~kis z$KW57!5L8p;uY#Oo5Ml3uGi7>r+)}Sjgtw7PH+Ek`(dWPS)FL?E4?gzsjRw2S21st zoH!6GPjdd0R|69u#IFI75c>VL3v{V&un#vTR|U%Tonr*aBkhC*=zKc8>XuK86|<+&n_^qdVB;9LyMORYKz z-%1oca$p$9TKWt2#8~u6otapdLJU9)!+VO0T{H|*ox~&Jx<%R?E|RpD25pC{oOX2y zp8J1_dO{cU&hFQSdqe4FqrX#ro`yp$i=CSSeqml z8CO`Ll+TQrlEe*;Kgk2;0pp?Oqt}9LRKhT`L;BK_#Cm%~^fr(sh;kk0W5Jo*^glezjYF45s?*J6u#zdyIKhqW@LFv?)g*WW zY4g2x9#JBk^?`1D*}eQe zlq>szHC1UCZOEE>?&z#fN0jr(HT{h)UhnAZ^_NfIFUs947l;Z*XQ%im z)@(IdwzLHzUm(qSCCITTvC|R|;}qnbpF7YuVK$wNHZf-$l&lzRcK%7Wq4Z@9>rE0_ zZ%7m=rT8s`9c~S8Hj(na`7J&3+k2jYbo9h++|Uq6`&ftO~yhxuEX6Ap&fS9jh(_-Ad{8Zy~i zvnW!PoHlorbLGE~%-A=G$0zb_1KTK#V+g>(LbZ2hu^Bz70Z=J3cLGEFI$TEIUPlR( z$2adr(T#L6IT-7Fvm`3fiwt0F#{IB-9zsa6S5FFBKuO3nRBXMVTf zVc6>|0$d~5?9QXT*%Nf3!nbBUYdqc7=2mn8U44`u#p&siB9{Qs!bMO`AW^PdJ4 z#HqsL9Z3O6ZJi%270&e*_^;xPyp2+P7Bn+>uSH{J&DP|{8cl=D4o7?d#xiQv(f83U zULWs=;SAER|E|QN7~t?4ql9wpMp^M%M0tX`)Xp4!z{Ju7QgCKPI(Sxuei<>KcQPH` zQhJJ@g5JO-5sd8EezzBogDO2CeYlY5?w_e-@`1D^{GG^clsCuJ9-=5s?;?+W);mkS zy5ZcXXFQgPSHiAfU~Oth4A9%l!Paloz|)}w+jX;gjyg`4-spR2ggpMKaqQZ&vm6oM zFz`Kn8!vpx8Y6XawPL8Emg(G+3NtlZpYRZizFrLKOvf;1uKdH*h87#O%Rp{7wXP6X zN6SZAjIj&C;ZoOi_CQYbgm_!_iqxM#6NG9dCD4S-umJI02cKEtz5UvgP;W=`^`n?h zL6?ou=XvFA8A1L|j- zYBv(d%}r^~?tBs9?Xigij}egw!Nh^>c#41#H2RTpOT5d|+biLXXud&!L`1|&0H4|l zhY6;5>v&z9_;`x*%Db$1f+>-RZhf~7it={AA9>G|)~{u9FS1&IhAKsu*&jNGmvS_L zFugEaBZI-<3*!Oq08TU@iI^%3maaR&Z`2H~&C>k* zAHb(C(Q{tZ#+RV@X+XENqrnWH={^3z_oMb$`=oMjbhSz5D7BWe(tWvaXm@6C^@r%m z#_D0odp@i$2__&pf1&YbeFR5wiHMD6`(^FFRJ8R&yY?DsFRm#vxE*C|#TT(gM& zR@>d_1%@gt4=#9O^#ur?`5)~XtW>*@e*1O>I5wxy z_b#;BFPDkj4#zs6wfm^X3ipT1YaP{2HobA7p2e(LkL^e-w4?o*A~ZDP#3bUc2`9DK z#adnIcwERG%l(zCBeV$tsB9)U^*gtjQwX~A&rq~Q2V%}47m?I%Q>3XL2GCz(+b}Nu zm~pfes2+G3+JYue-tw2nbSsm=WqI{${3wAPD>8*`+?anK^Y?dY36OXDTGJbBy=I=~ zZk>yQd{SKYqK`h8W==62)$=LXQ_L{b99szbh>Kz_9P}Qbi61@q9w9F0H(sywpIY8y zDl4VTjJ#r{`7|+<##hfxKL8E{kh~6ZtP_Mo3Fi_I9D!s*1t!Wj+)YW*x3FAg>ANZn zcTV7N(G>_2Xe&ijc;e{{2v)vu3lGB}DcM`yGDGenMSQ^T55E zE&lfZy9_iKg{l$0COOZgm*16IT@KnsJU4#*Y&oKmsbB_jwyc9ir|faK286A-R`|%r zr)i~lKeqL=`13r}XTzah$FNhOdP+_zy{LSiRMLKOapWElX zDTdqpXMj|dD}+|c^~uFA{hH^pA?@3U@?Q4rUG8!%}h5`)57Cvwk*)e z0Vl^4YZ~_JNB3UQDc@+*x+&I5bjiiS{lcl52c|x0{-eFL$D7`o{)~l(vmh_wEXI@2 zT!*zyT#b0KpyeMZ?!=5)iyTN>0!|#=u!6aV1Be(GZFmxWjS{R&h<8Zh;=XmSyOyX+ zS&WY6`liB!4Rt`6(-BoCusACGvVjn>RsMD9U&zPltxe;=T!IGneG{oTC2jR*c7VFM zCrv!_D&dqzOPX_c3{|5}-CwFC?rLGLQeaSB@ONV~#}E{7((QTQPSTqR-8A2t^Zx;^J;bQ>(Q;OdJKK*u zoz5K}xAP1}dP2B2ubduMO&-UEujJvTyhatw2_)1av^W0kaSDU+!4xC)vlmhmQ~HMh zt%>?+jMonyl(V@TPcPnYw9NH1n@o94=1TMVvwAKO>bFcTUDEkHQ@U%TH4R~V1+Y+C zwl)?nTSPYiK(o-28c9;yoH$Qn5$w+Z6+j^F2vpZHXx|vAq8pUrWBJ?}XLn;h! zP03&hHTPoW(u!n_=%ExV(gyXumN@RX%)&ibn}3muubT?TfyO=855s0YB#5w}g}D3O zU>w~)cl@^l78&1f7UxF4xhvZ$zu^H1;J4B6}e-B2nL_Sm(6@0mYS`lpda(kAM!6l?K}eI zg)ZfJlDPHvNwSp{U8v?*VFm~1HF|-k*d*Ws#To(Pvb4UP15ba=vtNo%x=+(sdERK+ zLCjO{M3~qK@lZY19rS$W6*C@%ueICs{%RK?KW2VJ8WGOZHSd+D+ng2P%3_Sl8QheU zjwj#QM||BvcAs5{?fD=k@SgXYRm;jkpO?-3j_>^>u0Y?{1?>hHzr=_k1KQ@ownCkA z_1`Ve8lHANPs?_g1saQhcBI zqm`aAa~rDV{EubLg}eikP!xQQi}^xg?>2h@#qFu?UZZ#!Fp9OQ_Ne&enwFLlkSU36 z^V2(C=Q|pj@y#=ouXKs!%S$dZ1LPR*9ZT@t|ZmNK;h9m_|Q?ZR%Efga4RWsv6ysh=>|r5Yls6G0%YS{4G~^^vB2 zizhd>t}dLmI#cbYtFL`hbIt;BpS}@V24r5&TiUyg^r|wCsIb3XLB=Dpu|osT(dd-A$o5w3*Q$pdS-H_;Q}@A)hxvh2zHjLO$!njNTfcoYg*T!C#m@ zz8bOA%>je`Iv@qw=u$x?Hd=CVsdj2KSZ zS<^mg0(hBvFr2VPG%r-$UEjLH?dQtCP$7Z(*N`kP+}#Yk5}{c%YO&nU5wr$vc4 zu*aHZUz7Xj*~>;^`Pnk7gfif# zSkV$&v4BV07V&nO#;4rXOQ-sa!YgHf{BT_Ns}}H3Ki@Zhq$iLo?>Cv6i{3NtT=H=jTJ^CiNBgQ zHYn{8z{Ucc|Hns>J!46GYS!FYTt{t5 zF0Y{wDBVJKtM@C8%L0l}EcMP#SoXyLyD*|>o@YgoMEDs`m6bR9v~Trd8K(g2azb|x zAB{2=z}CrPC>TS{b>GL?IC`j73b56CE4jNRY+5>|8JtBW(BD!HVY?q*Yhn-#^$_q;DVY=CL#|3UG6G=lcHf<1d%!fDVq+w3dhXic$Y$BMdp8 zr%fuQFcZQfL!n=B_9u0n&)lT^x*@4eH136DlFkHsjS?sc*?RPM0&Jyj!sJjdCUX(2S0bq>|D}$vjm7uYZ zkxqA7<0tsVpy4O8K;mt&Qg^=-eYJ_UCuqe;`P=74VRZT zuZj{lBWZ95{jHe5Tlh+gikhUx{jYC1o%g2ve3(lM*y`#r4g9ecS_Y5L)3ne((i7qt zru5J(3B`ch7`GsSdj})?g}=skj*%WRAm0HGKTZm^S*#%1IDM#^hxM#9VP>W3w_jEI zIG&nU=4doO%=%nF#F%AwoLDd|ww3-m>94%Vs8AsIY18;&si>gz@Z@<5{3j&%XCoO2 zFA-l#)2(2+Tdd@y&9yv|f7$(Awe_j5}V_LpU{zTAwi>lOhDeVTa}8g_t9h&pG6j_0$`N9*2Gn zoDO`T5{EW0*BVekd0#LW)Sb!m8s=!o9|9QEPMCl0ekiO0Vhd2S10*R3g5)EPkBA6WO{- z%}whY7#HaH<8 zLJyurgwXWE9VoIk6+t=MdTaQuH5d1Iy1ThGt$J@R7P%Z~E&39jnxdRI)K0$s)XH$> z^h7Kd@;cR|llrZ=h?~x0+TVyy14zvKctVoown)ZD)PpY4uW6l)IP%$loAD4|DYlyZ z%UvQIE+UM{6Q*FNge5QqWJ-umgK#?Tox}Y$zPF;m(3k!SqxMSNDgn)pw(AMZ*0zKn za{ubke)SM@3pntamtr4Zin>?MG5aMA`&7>$XkBj6id%8y$hV#3N*mucCpX4b$;=Y& zSVzU4d-B8vCyrxxMl4&7%KXG|e|L;WVECb`X^#6nNfDZe3O}lQNI$Bak?F{H;Wv3u z8XbEurTooq+vh_U=kV-k-wS}?}n04UEkDCFz@7P3$Xcns>Kt6lF zWTlNGSf9IR>oi0{JMvnE;y7VzL1@r4t8p=mr&7PTC1M?e!+C*gD+(hBVtMi50Tw&h z{u~0%3o?g0jys8#kL`X^}&=k{~_cwpa zR+KTlv49+)1C49Lz&bj9s;}g8CGLml&>sHK4KkQ_`xG` z@QEhcfay`zgjpRvV(~sGVC@+jU1V2e7+`q!1Kzr_9m(*~zembPQpMq!jrOxp9vplu z-|p74Xy`roQoj4++ry(?i7elcIpRT{QAX-z0X-YmbD`$N^riQn?<^k=p5$vE;`8ia z3ndX7=uN|R-6$TDJLRMYrD-2B#M%*ykwKaF>`~GD|iE`^@#EbHW8mn1G$V&OedKc(kb zgozOKJNhVVU&@@M2of)3jZ5+hhR!)q?`if$N_0TCuRPovYJ)$bJZw^n^YsKv;226V z=s0KcLSb<(R+Zb3T@NqEw_Pb8FYu!q1x+HFR(io;Fvr6oiz zl=|`+QlhJCuTkJ|ot=3HP4U&>k)qvZ{Ck)hucY83OomEb%T~rJm{qln>c9v)&6KHk zSKkRFU5IDE^wDGknNg!>nukrghUZjq?n%utHvGaZClWeU4p*7v*Av!q88PM@(^nSD zDe}A3qVUXPb~uLA2}rOG;|^-~P~Tw1Ovx0;L4;B?RA-^V^CprmO=EKn~XZI*C)e&yy)&i{H{6uRgvvp04cTY4qQF01%A{ zcjH;br!cMyU!ijxt<1qJbWG6E1gA?)dR)X!f%#vWsBtCL^}h&)aIaC~j#vM1&=VJb z0z``owQ}!(B;s|X=MNXV&hUk`gOSK(2uC-&=MIOMT@-Dl0@$qNpssIYVUOO{D@wMk zFBHIddZhyg!tLJUBV>*W3> zjQ5Y`Vhf~XLAQd1oLR1ues2XVoA`TRvvz0*cD`qsM+H--DZy7y7CR0PM>rZ;I1`y$sHv&UbH6HQwWRoeTU6jz-K zj;|%TbD=Nwy#;+uJINsz{U$1K7GT|f(TmMggwop{8`2SO4a?Qf=*EsjG3zFlV)A6I z2WQqs+n6$)yEcWJVyHzJ+KWzLg@2Nn{gkYGxZhU#16Sqo4Nx3@rn*VCkQ;y7_lxL1 zA;cfxj=Kx`id94de+QQ()H!ZraS!caRb)bMBst$qM?2C6N`S_^;{u!C{{^{&D`?DD zvXofx0p=+79A+@&Vgk)4eMTgN05oPKdZKPNF-%apCS>UA>j?zs#6rFY1YRZ8tafq3 z$R2174XX+B6UyHbQ-yrw==ns*8t_g-hG}Cu_90t##hC+a$NPzbY-|P9`DVcR3%7H} zD#j(jMafVg26UJpa#-{vAs4DRO{h5J(L}rdo`=%?gc^Q&=1IFRxFG@t&CND1*t=Uh zxQ?4^V)C}Pw>Mdy&P`Z(o!^HGXg?|(mJb)1c(b4KqU0e{f@-!DQ^?-;&q@MSG0bLD zw;Q(ZsmhhMWvGa=cPLU- zSj$M9_QJW<1Y6YG3eUT%S76G4&v_iYnl~tITIZGFZOiOvg0+Y>x8$r4l0nK>vWfQz zOHClWJ{E=g$)taEzKHWZ;uonL5^hJS^fTjS`;9+)3(9x*oVVr##r36s+5XsPD( z4*#ZxssdWodP!Tn(W-oL?bnvFkv|u^o*q`a_O-mq;vRoXN_zC6KweHfJ*tKyqpHK9 z*HJV4twf2`vZFF-WygHFK#bR=UHXGibhk=&3N5*Qr7ZK#RJb2TAK8n^p`uzC-=>Pc zgnD>0#^W{OUvk1{^80AP_VD+zG5pNg6*tP4U&r15Jp&y17c<{nSYGa0N>*T$SC><} z_>j92aS|ShfPCxxwROcSP-27mq(D6tF8e@K-5FJUhnZ`a#nmNTX?eC;g4i1>OYBV0 zZ9MoGCWHAk_uUICeRucRleLK>nKuAox_f$91dA<0vRxK!Z4NrmR_-Jx8_oh4Q|20{ zwm?X`Tg}c3Z+|lz$>S$2INXRw-zs)rUhWeKDW!4ac)tt!G&-N$>MFaWOg-cr&w6Gg^gyyjHk3g!`^2 z?l6btwga;?f*<+3IC_{A=Rf8=ewucSBM#6UXRlOq8P`;f9XXDh+%*&njj-b>B-wi& zO9A80e?E)SA`6=H+I~D^QPJm^o8M34lBE9YDEkD&!$#L;26#dFAJd69)$d%8= zHS7ua!-AxXv*S zu@?w?Y!$M(Ykw~b&7UjZZ9j&e&Jum=Hh#0e;BB`1i$HX1NY4Gf)k0Bba|^Qz4@eB zQmxmV&c9rR5)xI%vjPmTTHgKiq)|{V-+V=LXNE}`HwWLE)7Qaun8Wf9{{KG8ShuOF z^B&3J*IpD6jNIVk;}nOjL2`}$=9=h1rB0{cPfqGWX)Aw{)1mX-y4Yppm-Or4XycLH z)&wm=9*>dp-gDlg$>vfuzp4o}<(qSIG{-6*$||Jx(y!={wB0qNXE3H@0iwND@W`o-FKg&_NGuD7?@=h;yguAG2NB@g6Az(cO60ZO#e{NoQH zes6~_eibI>Ck)=(1eK6M*zzh4M#JJjlL5UMUYrq7yD!jye<6`|tx*Jiyz0nl1f?pZ zI&ZL&o01P53NSr0*IFJm$ONcogzoIvO$-sb@a3>k?!2Cw35UJoko1pz7@c0|KH*ww zZ>!(am}54~7sXQf+E|H*XtzUyUM`k!eYZh`HJ^Ofv~8I_P@?$|FSL-U&qUm`K!ztu zn!WS>ttkZ$9{y8lkc}T!i*t;l{#>U(cF#v2Pl*2NN*6k$uUA&Pf5NA*(@yna2!X38 zTsKLV6P~9`tj3`xpxg{&ELYXsHU9N~GQcMHs&$l2;w2+57xNBUfcuR6kfY9q#WZJr zzQyNRsk}Qnl~N_vcYgm5e9;+7G$}ld1u2P?Mumtyhy)PT3Q3vUt~>WNsY8)!h{{=i6Ig^ndFYV;2`Y-y+TK22 zZyg%BnmhP4?~zkW>q=Dd_pc2kI_|)SXLibPe`sraWSFV0b6)1K#xU&fUd6v=QmAN$ z!t%XmZwu%O6A2AT;{l(Q9m8$aM>EkqN0Hfe`tS-#S^GpRcQD=9pt~x2^;R##e7|5c zDV8THf-_g$(y=ZezCEr=iP1~H&MV=69A@xsN*Yp-Z}!`R%e#w<=duu3YjZFOhho|n zuM;M#IDXiK>i)f)<`DJ*?W(Y;y@MyDrtJ2}H9^JvgvqP_2}6fxQptpyhLbNVf%@RK zFbTvoapLYOmZx1CB&BlIPN3&*&|ceD;2cwD^6E_Vs+-?{9~wb;gDNX4N57jVX0y5H z?weS$5?1uYHV2bj#de=RINl9tSYcZ#E0xjF2{_NGcB-n|ctKKJSio&+sLWwR*<8|@xEd_q>4GM-c39z_e0H`p17^HeaPs#Sk88FG zwX>=K64gx?xiy;f8Ou?24qjB1|CnmzQo&>}1p2?J3Lsrj@Srl76apnW0)m3Gnfv1d z7_!lVZ2J?=0U4xZo&U+@>A+g%{5ZHdqXl7324YiEQpjNy%(_UiHz;fsBkI07lh3MV z*^PlJL(bP6D|qj>#dd^3IK8L|A+xaM`ObRcw%OXOjxZHMGE?g2W! zv-37;aM2CN(g{vSyC*aul0G8rJt_k#NgTsY4Z9%=Qsmj#c5~>D`YQR|VMg9XJJ7b_ zf`t0sg_e_Y;eE-5NZDiCTWrNeIXMhcQr>wH6l5!G=@cQC-h<*Khyoy%=dTo!>O*bN zA07=CxGbdlI@xW*?b>N5*Qj+uxR4YgJs_Edo!{prYXzwauCJ_e>7*YBEZpmJ8zO)sGG`m-51 z+Pq4djgj(J%9|`z3_ogijUFm8x2j%lCddqQhAp1W~ zp8zH1RukWl%UH*R+UkJtbnD0dPRA#6{=$y9T7Zkb>^xZ z_n7o{#<~{0Vs7fFuccf-%Z%Qcm~Sb0x_T4MPbL30nP+LZtA;xP%jqQK|ED(M0lkeh z8TaXAF4MEmv#JRxPd`daw@GjjwUOJLb?cX|uT|t(>-{^e+L?lej_&6;zr7Y?@#cKy znIGi)^&n>~2UGn^_D}}gLy>RcnI9dQByy!widwL))ZKg5 z=hVy!fSj6_JQc3W`Yv!GN9w@STq#E>T`1IyWc1-@E1FT8?cKS0-DTvU=~+6$fMt>t z8$lOl?Pyuv12@{5kTma*y}|`jxme4nG?tJ;1^w>U664ojn!crM-qM)bbm{yMzM7}8 zdTXgHu$j43-IL`wF}#Lvs&X##TL7AavnL7)6y@2IM@lNkuALI|5xK})yL;>4XQ=2_ zg}b}@^44kStF6X{B2Ycximk2@diG}i#aU#?Hg69HM_&9Por2d5ONwEdP&&4GjTpIB z)josx1i2`^@yg6Md_sT~qnIXQi&{n8d3)d4kLu4Udf%p|*WXX8_Tz}cCU z-w0{GK&N|DDX>*TkGK!cm7vom=R-*d>z$VE+Fy;IhKmcIUC>8BN|fwm`Xv{l1#_+%%)x&Yk=+Qv<%T+Qtv29*huS13n^{uYeXg&K; z@(F^^D*RkmFBJ9ub6S}|DnR|Ibi^f%2bOq|k0%p+GYolkA!~YxuF~)}F^@!n6=2lE z#T75oYHHJ_ny#ARj!_~DT`F=Rhn6p<$>;la zE!hw__`Ai^*BF$|mD9;BUzqPMxWgUkv-0-xE-XVqa&@+=wG?bB`z(y~1nYT32JGBI zs8NpKyf>tqq$AQSBK_{#d$-f1{zYjG10n;}KE!UEgm8W{IhXM^2BkSi<6I&kqd1? zY3aRtE$6{^-dEF4v`mQ~u$I{RqI{~q!=GSu3;sp!+Qe?8HcTbKx8m+J1EL0a(Iz`G zq&P~V)s(Q3oSf$cO<|T*Bq?e;&s@E<}1e0uJaT!-P_vgvyxo6i!ik_c%+PVv# zc_{R$6b^6J_sIq@eXg#0G71FzFD}U{qt$x+qtf#Y0rM^1Y)E4nn~Vv#excR=ep-3f zt{-?76cg$u9F0RXU+e0@XxPh4407@97iuzP(=^s9uPwkyp8v}r|Gsa#L?B)ppSskP z^YFO0F?ojVdoyr4xdW^DcZmskUKm(9hq~@X@>ujT`TN;!e#ZKNEW76rq0yp2$T`#u zIzEw%+;pg_;hkcxYqcpQN)8exRh*Q)FX(FZ^ggSf8rf+H-53YD?@13sQHM@|Jx0F* zthV#$#lEVT&&1pHJmBJ$yg=8_z&|`LbE;Ulq3wQ;w_@wa-b-UFKHjB2uB$73SlG5! zgw1xe&!k}`pS-9If8Yj?il{X<^ara0eS{=zhn|0eATVnp^cv2}oC*VD%7{Yi5~p~Z zId7UsEZl0iI;PdV**v1ea$QHATwLQHI?W&FLlT+Iyz<51@_V&R zWPMU^z@L2hmts$k&Dsyzf5ANpd<&z}}$VT0(j*m?aJKE`8rn{L+Ntf)GcEY!Zra*XJ zx*sE?5n@u9z&%Vy>c3EP``7hQ)Pkl|!w9gj9?Gx**VC;|abBb2q=wL$*{biEA-P>I z(6Ajy;cjaia@ve1Zn>of#YbHh3q7&RX^E+7yOrD5p4vIm-#*c-O%v-lE>zC7zW$FV z`1{}$-kfA(F8q?HbqVq+R2!3($LNItvBHE52^SD`iBkpD24Frc`2!x|Kp`s=Co$QQ zp&#yNQ&L;R6Yd(?F1sJ2gxurp5_&HnFhx2pYVuj08djcS+MpySs`C3gxy>;?EG`A% zsLC?jl(MA{t8b0`*pjV6Jutief^V44wF-HKEn}U?tVRasixLr$#r02I z*Ya4H@C?jeJ}6uMC5rux`es)eK)TE>o}tD&kG5ZK6FiCJV@EP5P~hbrEIzYRBHwAd zhV&|Ln##~x!Q|v%<@DK*s$$#~qXbWaQ_(d?`lI9K)CId;3<1Oh+x%Tnp7i&;eYtAC zPb*j{jb}gsPemZ|Vk!+Z&KQ%&AHPM&*}-%=p#GK3mh>6x%&_dgj3zEFP|ccfI}dX? z#uf>tJU@B!W`5y8DK}4y;Mr7&Ja4T&q`2*}3_^er^`nMd+ftVQQJD5Dkt^5qn36|g zV6}qTBRiJnngSJ20rlnZVsO0zN$i^<;FP6{B+tICnY)5`=O0f7FTDQxAl~a{DdFpb zpx7#^AjKM=+Vib&L)t*KtNyQd&_86>z5rSpv98C*TUF<==fy6|RVPrT=SihcMl*sa z`A%R(uMG@jV@?;@TA6I7=gi-1h79D-p1M}#t;3zs?Uia0DAlgwD=$wRo(0f^baN^; z0A;&H%F`#S?W{#mnZg8(Rmgs~@VRO>{HzY2V`E(xaI-*YCxpSXx)2!P=8(@br1T?D z^~^+?SvTPgM*r?py>%6#AU;$?%e-ZYyP9C2V(z4~QRWX^p2yML%`lP@+HPE|r%2ww z--C+$SIT!}U`?O+=<*#EdMm35bU1oInIo1Q-7cOyTFAd3wG%^$D3~RZ$5?+e@pad} z@;vK{?xp;T67O?;I0qV}?$>rU$-~9^3W>pwIjv#xVameFTm$83bKBx=-_)>L>@=>` z8;^&x1w7P6-;il0@>m*aV?rYUg0LghaxN>PLz@uR`dPX26K9*LUiD`ss9$(31DdY6 zpK$U!o!B{?*9Q`8HU9GI&!rTymXwVyrCKu7(*N<$t=WyUszLY!m4s}eqiAz@*|G)X z_8(aU$bH1^eCRM$K>F=4v=r7PCwypX;Xrj98@gV$Z!u&$&~7HEnmrLLa^Bu!v)J6X zl1t30)l|g!pwuP)*bAHDv$^K3$2H5FeJ1n$tJZI38zYCS9LEKh?1L(r{7ZHGiq(m? zxjS9XJ2SRljb0eMkSh;26Ir_Z($Cl2PD`yzg(UlHoAn0P7SPLf!2m49XY%L6t|TO* zaW|yJk_%QPc|eiMHYF&nyF*Umd?-WDm+V5|N_1kX-s9E0Vzv4?VF^kpYD&* zC%4bC*{MrHn!I}qXMTz9`RJu}v6=tl6l5~w==X^2f1PweG%MIXT*gwPa0B{IKYu(W zNX!@v0HBs339zRLpBO{HKP%l+?LJ^x$cu}3W;zgAq5TS;GO*)%0#d@6VKv=hnMKdA zr2M@6u!s*@=|$GC=#uhl;7knin9OPT1!u9ww_hAOx~~@}7K>L+jzkXc?D!T(xJHF{ z6<0)yE)O)VpO0HMQJ&6MeFnJ{;?G8V`-eL2k`%|C)Z;JV4LQ!$>u2qj&9N@aWj2S+ z-2Ap{ul8bx+TpJIO}v$7iPf9P;uT4S12rjRsnAroz3nObvHNNQ<(G;w$8hDyS0fUz zMT1AP`R0mId{{^>Q?FWFvB>?{gKm|>Lq`iEp^6jL;#uCaGm1=Kzq#vtImzN~C{g)r(z@S3R;Yge3@ru^c(o4UfAus*(Od6v zyB3hxeZaYgG_@f(e2d*fjL|lp?93!j2+2$h1{Ei6ycKnN`yv8VvaRR_3_g-@ey=Bg zuQ+q~#W)x6Ic|pAaMJE&x2Y$nV&LMXc+|^@9F}D0h=gU%V5&MeI5fKMuYqclQhxUS!-p9uode&L$yx;|FqiY9 zz{2lhR0UF~rmG-cYyLp#s#bXz>7D>1lFmLgDV#1GxEe5L3Zp4^p%{6&)_N&QSW?{} zT&2!q(P!CcHPuGK?`~Fk0ADeeKnaedJ zaXoNtE$daEK0o!GWobC4B8Ssz(vt-Fp=qoaS+*jgnR@ayffEPAp`&Jz{-fW8Sb4|X z1&b_i*b|Zo-HLmi9#XXZ>ly{oZnhQ?XPmG!R*M7w?!_=k2B}GJHgg2!NrQamnpDT= z7w2m%6OE5XYQ$>sQm9(?*&s9ZslWw2&Wm@KpnlFv6=CBA`QX=z&`7l5g@iLj?|eNT zSLujQx3JZ&_`JSds#;I|zg`edMl>)PriOAImxo&o^D3b0v`1AjNO_+`)5(EC;PHBu zoYVT~?uZlByT3PpAJyLaN3!tdTT&(dg~|`z>z!-L-yPSr`P~^~3vCP>RI;?I7(tSG zJ8eVl$z*rPtWTa2`{$$e@e1cpy#~af7p($(u)Yvx>p_l{+9*WFiSN@5HdU1}C6jBY znkezdx0p%*E6kC5*5A(iioWEEW%D!6g`ykJ;UP;ClEn7;;aQ&TT=|+m|ZTDi$-AncROp-^>A_`-4^&VXP@aYuGs9&7}Oz{RzPap2Xom`;Dk;qtT zLlwR9L(e!TVV;nvR!b`-#qGI2c6LN3`dWa?vc-abAT1|<;Ve;5ZwRcHZPSs&1u zPzTnLASc6aa{?u`yNK_l%MjA;-x5&Dw--RN3fAmg$WsA9Xx#f(4g4vR z5&5_#WU2aiZHScbY_!e?&VlQR7Al_I%$!{ne!4##Y1tIWJArT<7a_`!xB|9t@(LfL zW#)!%lCQwK01Iy+F*GS&(tHVq86#WH{1+yILq?{koKV%57%^;>ugC0s5=9F!!;WIF zXY%NrkE$W1ce?8%KLVBSHwqV!b_Xr?U`r+8$xI^UIMIHkF+_$g4h*PlX=*u)}W0umSKuD&`TU`#9MKn6kc3g zY9}r+1=P4#DaBe z$P_jR{=Z2jH`{VFJ^oNEQ+k6muuYo;M8KTr48vF-yb-r@PMNMg+i)xz_{>7pGcw>C zKtk;t$^Sv&xb5g-c264(tg`;;k3fUF!nzlk^%Sq_addQ>g#@ST!>xLc-41<8^_ses z&;)tyo9cI?hr{IjRVX%+xd-(?y?uAS=d zqwvHo--60|KoI=6|D^cg5x8pkYT6}AJ;Ivt)|x2dX}*)GDq(`Q$P6EDa8ONNltB7Q zrTSG)&W+T;=8uxPi<+%e(i5yIBQK~vXDYn;=EV!OT6>t-ojO@*KqqbZ3EL*JM!fBz zYCV$q7>4E_7{aGl?#{lPf1$P68H3gjK>*wsN`2z&$+ekrZDp-X%zaJc1rb7HnKOo` z<|gT~^T0F1C)Wg`eO2bgJ{6%z(UY@1_}X1o(J_~3s_>Vx^vBO!x(6OnD0uOA8?!Rb z=Q^9)Mr8l2MaFa3fd6WDS64jwZNIt#a~@B$CMEWk7ZE5eRJUkxs6G>0eZm&jXZ}*7 z&@ILu^~sMt2rH(Tguu=6>i#l$&z@I|kTw*3tp-1C^W4LTZR9+hVNELC z@^P?A6|vV^kvEQkhCQ!P6IAF8%!|Lvx3q}d?P(q~asiAM%zP$_s?l-v+aHcU6cX}Q zJdZ|^TEx+iO55^{kG~#jZbJVz&Hv}g`qez-mP}pH7A`L{{4ttdE`&yu{|{n5*MgIe z6qyjz6ERisJvPn3gVF-_W}GW#Gfs{jN6i4j0eu^7qWn$zWb?;O7GYFG z%+-J-4=^Dm=z(c!+}P|b1no^>tZQZ)H-VPi!If99L8pu&+!yo~zp8|%7_W0(vRI{& z72WeVx7yNTov6>I#KVGo+2QT+cb97_YT~M!y%Ql_uw&;wn<-O4#U z>nVwTG5?e>{yf|DlwuESW3h^|T+YKl0#+I|d*?th_btuISN92i z<^@V%ajFeBLp~^gLdq)BbJ-0&frH_0ix&o1KXRONwbT-GcUb%rERP*>hv(2pt_br! zMxaQsB`1bIC>%QEpuI09wQ9L{$ms@JYjjW>Dnz})bNR^TI6U@T#n^3;(N77~&`~YM z$_vsqn`mee{?v>&ry6Fdwl*=8;J0xw%Y<=i%)&EJZ&h^z#xJ5AedVp-80qb&1w~Oz zpa;+8?Z6T2|8qK-t94YCmSrq2At^%wTtMcjuN;5&Foy{~u6=4Tu{3pDS9g#$Y!d2q z)B6c5O-f1%U@j7|e>5pejDp!WPqoieb}?WjEc zV*Fn7ub#j(C2}H+dgvl3MUk+AdqoeDZHb{q)cy5>+6lelGSfq<-@+=yw?5j%&wtpD z50#;@vI;n>*xPZDWF5nGvmAM%m}Zm$N0H+7O5Ek#G;vz&wd`hCWI8#e`l%n8&VQTi z|B$URF%G_IGpi5(t`SOtqndG(rtKg16caF#{d}lLopS5;Dn}eq(_UELzk%=h2?*$J zzXW&+L6aR;NFffn{Wg`d+eXts?9uHPBo8X<XTO9>zTaf1{amK=ke9_XZ1x;JT?(q!RKO+rg6!> z3m#PL<@(_H2~OS(FnbN{fTezq1c%i}pW!QZzCh}peb%P&w>SSA#P&;9mvHH})jw+Z zZC9Q{!7gIL82%AoN~7sC)XPz7tAF)^O5T?@ToTg(KnN7c+BiVo!dvtQ!&8=D$@|;z z-~Rd>S<@}?GJNyv^f2mM{w#@VL`HYpOA@16=3GoYEGAz0N5gH8wapj{s|Ir~Zo^&m zYD7FQnLpRoXA)B6!GWUVtej6em7d4j4@+2RX$4b@kPiT$c>%A`EYqunec-7k3?138 zhu6B+M!Xviv&&ag(Fc8sw92Kn46XW+%KW_l$5uc?kUK~#5pkLKi$frHFp2qP zWEtL$CYVhR^g?4`$1X&SUZpCX|K_)OriSh45mc@9~l+iR3pOUtH)`5s6% zg@tI0zFiDqgE?EgEjN;M=X7*C(_pL4d;PvyI$dVaT~-}XhKq=DDk`6X%`<2DIdj26 z!oA{h(xQo8rtNZ+^u+Mcg~G$Q(-V7tA^diSwol8+tIZBE0)dcEAGjYHHNp6GqET9& ze^^^l0)vO9uJrsTin=&~O3EM=lxEX?=knSW-T*B*<}ZvdMtdU>={cIv*x1bnYeeLr z&qP1hQ2f4eRU=bnfzk>)2%w;(czj|TdJK9^X-}}E9&lb4yB}Y@&Y+1%<8d>2U?1xd zj>=uuauc?AJ5e#yzAKZuZa&ER)B3)A9#UA6E8JthIfjyl1hP$Z*(7p{XwX_N7I5Fi zGJ*6-AaQIZGJjZtc5OTp`z$purbOwDLtTBe^{+sJ{cD&GvdUnHVZgWpHyVUxpQ1)-VhB#P zY5FphC&3AXBj64IKIBaof?r0xAO_q3F0j-A~9J-ZPLlz}3xFKDb-HR{}eVQNoD1L6^caYjLoeXGl zuJ-A6`%6{4<8IcEv!c9tny-_H-+*H~VxvO>v`ji8v9E&qxuaoml5!^4OvBjzte~uu zb?kf7NdX$z|Efg%Mil56=8k3K(zp=$h4LvlprvT<^3-BL=5v%TUvO+;ob7QCmNa`Ib=(96e(8d_~YN@at-HU87Cvy4h)_G&|1^f~;Er z`(nD|p5Uu5i$19JL@SzqUemnhFWtQLjoAU}way$eR@i%n6-!ZzEH@!@hmE55SkW19 z>aSkxj!c|oj%dtDGf$?pdHKOkCX@uzkG^w&5aH^$_y|GITD%4rr#NiyeM-Hb^_ogW2QYd~=%HKcgt zvG^bd$PhPMf#OG@=ak06t*k>OvyOY3pw%L@ru(4u#Wsl&kB1$dRvz5JSvAh=p-MPM0>sD1hD_ZZeggPFA#L^q%yukNOYPdLL!|CqDLZnhn??|OE zO^O{hH#}~@NYcCQZ;^k`{lJ#XARUoqWT$Ii#xKA7^X^&{IqqKO15q%q;cSM+u! zo1TAD>=ikm`dri95vZkF6Y`S!u|=XfF1nvWXCByYu|^_$Tk&dF!5#%4>O`-AK4(5; z87;04gRI3=^XI81+HWc(^NrVA>6++$Iwv(O&saAp1WM9im2aacQpr0qZlBr&Zq4Pjy%PTG^Zb7DxN#7(WEb9 z_XAoEW$nxNWHp17gIWOOPmqY)>s0L`aqoac@fUBK^RuqZ;t`t@!6H6eCET*beRJ&fw z3=$1k96Fyf{IWIGQ>-$A_K)rns8YY{q*65PR{Z) zo~W50X%SOYe-1Mev@USYNE$hcbGw`!sTc%a{e%lx^g1vRbqG4m-~HFe3nPK_!SQ_d ze4bnQv0$r-;bFiKW5z)$(sX#(P!?8Y5wF?I&CQ}VqTz3HH{SbsXjS98D4-(lolc27 z@M2$ODKs-4zR8#RidmUR>U@gtsK}jid6fl*S6L{eS)L=ATz9kUTTdc=aA3Fn@Ui;O z6`)^5f@zk0K6m`VAc|P`!W~Gdst(e*0;I-Q>2+D=E`U4TmhX65fG2_CKheuS4wNYfb7nPCdVWii z+{PU(wEeI$TB&8K@8qB&8!eO^lKZzA>Uj)QC)*A6lRz`*su8|RxLYa3vU(R5E^(`J zdcRxYzNuo{`BSa!K{QA7>?B6>dpUGyAUO*tga)`(l^IQ8%cH5F0p>ANb@#{5`IxWz z-=x^7zOU{&pVw*WnmE;ZqjP#(F%eXJ{)-Y}t)#CEaANZ-qpM8%e2B&gs`p~J;!XGc zSeA98?sLQDk9XPz73z2vT}XI&cQ z=xS}_yS( z5hzm@t)?1IUA)@HYSjo7atk1|Y3}mQ`I--a`+sRWMC2jq5ziefEENNG(*h*pMzDwP00{xXm7 zXpU~ic&lWU0po2zcf3hX42r%1cP6F?VW~pqcO9sA(|Pr*+~oUH-EHwW>~+gZx6yXLVZ_#5GL#5onbAHyuEB3%40q>-U< zZfKO}TSP-8nc~YENax*2PXp9vaxs7i6Q}n2U}F8TMQOhYwO0*3g0QQpxvUUVjgGmT zG49O|CQ8~r8UDZS5fl@xK9S^h!OzPHGa&3!h5OZ#&cL15yN;wwh|4Git-%NUjcVcw zpfooF*Rm`!zp{9AQ^H6xAF0&wPPppC95!{#h>IM)*^|%Y9aiYTM=&`jYRn~ciRr=~ zf~xpqB;4S?8_iCtC3j*LBX>1iwp5>D9-D0b{MPRlf}}yhc<6{rhQIc+^KJ5F8?W~&RPv=g ztS{^%S5&p44j#CJo7lxab6w|*O1g^YJp}IG10A$yvxAAa=qRUt+)YP~Q1)sY4vlyC zh`>pLyR^#TY$TA#nP+wEznSoV0BBeYxQ|8rebPu>258)EK!D8AQDXOS=eJ*r0>TIf zJ{K9R%Y(%r8aL7SibV4TKAHWtitw?LK=hY7Cd06DA)c(=aA`W)5J_Ww(~3X{N6dkO z{rV@0GQyvCHF08m3LWCy0er@LgsZd8V5z3LpCz_|h!?HuhXoJ%Zbwxx;gh-Sj-z3m zx7)csL#QeFVJS6jBq#a#b;R!>HxTX7I4#lwL4u>+a-V|2TFiN;CF`3yk~FGL|k&R zrqbucQlLO^>*YUO8^E2INT2vMg15HqUp|W^LZ|r5lJi(?6@oCP9tl~}yQe)C_tYFM zeNcktCT4&_P28g0_sO%6wXiA>Rspe~kZyTOcs(Hw5K*VOwbt7(v_=A9sX|xUY6gn# z-4FD>Gi#dO?|kX)pdJ>TwW!8{nE%XTn-q%CNBKD%T)1YgWg6y|YWbck2&0F(O_K%` zUmocAwBlam>F?#L+IJCWYDJtBwvc@K_e!YwPPf`|o1(nll)-zClsXHVcPLxJ|VA!Ifz6uG4 zKg-CW`eo1uDT1)vzFG}&h%(pVDsmyV3c-NfHLdc2o6!Z|>nuPs*!UJH5B=eUCDov= zUO(P?g4xs>W2$k}cJd6yYpFAMiY4;O#b;d&ZNL2JA&)<_n0D|153h0$s>XM_>;uTC zK3((#zx&D8R^jzCe7Z0JFPtz}E8Z9Kqdk8;Z*d#Gne> zZl^dbWV5)c`l*>wR@n{MDhtQccp(C*@p=pl9&TL+ZP7Y2rh=+>6nm@zz z1%6@UsfhrHa}Sn%_-zeeO5Y@|)Ft9wEn?x5)hajCwoIO4Hn)#-n9Ag{+NXJ{a5(wY z^3!Iix@93|WfoD_?f|g3$%T`kyO_{KZVSz+rm~y?QBe zA9n)-NH)-p6q$9s=UBu-87;Bb{13|K0L!QYOD2Yv=9Dlfa&p2?<*8c|Re#N`J}9+X zZ2pwTn$&IlEXlLx>1JZdfSG#0I)$LTN=2Vao?q95$`9h;B{rS5?;@u9E3!N-eFe1- zfCbn?^C#)OWs|R}at9W!PDW^$YcIS%>`6y`gSWcdH^)xSuwLcOsvIYE3H9~ph#ZZd zhNjLp@^+1H_-?oQv5;mgQ_RrJ{S23q8kQkj<+8?VvB@hlXvX=`ikge`Be=Xh4v&gm zys$X#b2lyOh9M7%DBjz#j&E{q_NFl3ssmx*Eiw=&SS?t9o+yIlgSrscs!=LM^<8$* z#k6Gb5D*&L2vvt4rR3`f{(M$S`ITAyb(xAcMZ@m^^B=SM+#YzN>9UeuzfYeZXe`{v z7N~SO6^rxz+jxRn2F(tvQRq{EhnSSUF@+o(opn&6EKKkUKL8p~hlrEOhK^=Vk;BsR z!**v-pnFdW4c}7j>7TX+Rl^WHQ0`F+o=%o6gY%MPu9{fTNCM>su7Ez@dv|^2t}^RI zr&j|m3RjT__>9LluDlzglA4&(^L#W{IX1jRT6H>Nvq&sAuIRQAaPD1H zMZ(*SSm3*@Ns{a0W0^Yat7tLwpur$1!&a(oq$M*%Ro)1%oIr~GN4L(`FERX5gkNH8 zDTzhboXowTst+>rt}%j8uO2a%ze16MlzQe$f(op<>pM;eey+%SAa^SPNwADin%tPK zwdsIAGT+f1nBHokENUkRC}`q7tOxDh-4h9hd2Z{k^uviF{eI%-C{4%us{Go1{iNzq zLQn>zn&VcH%dZ}@L0{rO-v6b$h5{$LIT_@H7cHH1;WvV&??_(%#>V#qchOK0cd`FM zr93LQgKQ~s`risz#@%cux)p@gY#c0PWdYTu%4PMV%Q2ffgUD$e+w9(J_I%}d=@7il zt#x)`L6KPUW`*1N5)rn{A@O+{t{r70%%|aMUxC=6TwPkxgpkw2NG*G1Y}@O|in{Ts z5lB&j%gcwOC8DOM_>eaG3NICe;>%eub!7Gn_H?eXA{HKPEmOm?@phHsMY_=+=>Bce z8it`+_8$7-HD0m|v2>!ec6LN#C{p_@Z-a22-M{M+@G4pwoyb6`z{s0cw~{C%r2|hmMwW0b!6H)> zpS8M!cb=#PpZ=A9k%*0~;h4eWdrsq{jny);CzyB42|2nF`bqB%S!CGf5?Z6Y=Sj^W zdT*m5b=nWqm|DKMXE1BY(fAmK<+qt9x~3d;rMO{lJg&(|H!h0~@Gh&V(x$Qy6lCA$ zzRu36_I}I^_nwefVsLnkjmiBmdnzR{m@#VjP=Snp)XJI;M)V2dv%}ntzX437*Ax z+3yfsfUIcrLZNan#+RB$%XjUhf-Y8Kl{%-EVDNZwo(G%mevIS@@|C1Ympp*jeAsl=H*nTzTqSji$5vJ)>PqkUlc2LMtb` z)$=$4(NECYOwvO3!{Tk?ks<3W#+UGh94!%eqVcgGNz}sD`l5#4Poi(6H(Ll>_JfJ_X@)3z^1f1NF|cr@elV-#MO? zJdfsA{9N6Mg-&x7{QO568*~yHv^-0!RtIY_&{%P{A~_FzcgLz7hg}n_pswcFEzI;UEbzTV&X;h??mAO}Mz-Up7-C z9x9*N9Ze8oep;vMPTNtnlXkN5E_8-?EupwoJT%|hX#~#)&%#V$bCM<-^xb3jZZ`o+ zS_zT!^%S@V$wLwa63FhrwcWeaVjjWfG`2l8K2rkbf%-*Wrb_+fk)iFzLMUSseVXn^ z*|UY`o9l-@Xv_To8#KzEGu)v9t#2qgqPM@zdv?fkS3fT=A}-KCxAc;uWSQYu#Zs%! z(rQ1{H!zLV=N#z5RZ+;#w>geI-TopThAQ?_$|F!cSua5e38fKGP_*>^2ppU%?RQ0y zogEQ!aS5C>u}6NZaQOL@b1_YLj>`s4@3%N@G$E~*Zx$?0#XQ%wJ~rKTw^ztsG(09* z8~_CcKtTl_pn@@$)D-jbQB)k~{{{n+R{-f=@*`611ZgV5J;LiaNN=(ChYevz_#%G& z$=j0R?CdT%7OAKcT8`!!#&7YpO8cZeyrY+*!xAV>S5a#l;N!CgmF!bVYBiI}cLZVx z!yWcYu-U2EE;Q$@x}xe%|EogiW0IbN6Z}r2Z6$7F+?BM&4Gpv3j~6iI$Mf`3d^9}O zw-ar_G;2)5W+yF$8O~YXM%;QMd_DAW{A7nL&*zbc zpUshg-h?@-Dq)WGvSm3|1^zpJQeoVf)D_J;M_(>XFia-gI&sVfJxCc9)uN8e6Snk< z{^jnu2m3coMV6NQV0vR8`v-rPpB~@MRPgqD{!J(S*S9g!8wR)&W>V>9LpbV{>$yq^ zf9INloxg}!muRk^EZam2B$58!!4y9a5?HDeXgW2|_@_pqAqR`osmd>Ij^GnkKbi6| zf)IJANkWYE$;Gg4XvV!_XW*Rp`4~Rg;zD4}DMk0xPnAXLicrh9su=Ur(RWf?OQe>~ znKyaoBxpS0@t_A@BdBgJUfGV*Y>sX_Ov7@Yl?D5A%$r%m{;VUq`ERyeI2_z2 z!%5d|(0|>3slzoItvk%P{I;f-EK=UtL!YlQZ=!=aI)mHLceZKj;66loC7&~d(vnZ* z$j~ZrI<3IX#*S;`iIZEUks=9oHN#go4_{hypU2!wIY><5iiFm;iLL1TsGpZR7w*3d za#ECl({72!wRvPqXme;cJ(fYs-pRwn=Mr!xR2*@1JWo6#MimCa3CEM!>bC-(NwSIH z$ri%|e&hZh69KX>#K5s!{vl0&`A)AK=m&5j(a4_Q4@PM@I_v>**_T(T80lGd^F(IJ z3dMB-7iuU-f$_Ax?*{e9MxvCGJH>z&q{~2W)YYP$x>#d{OTA%OXLa_oxOVQ#^fEJ3oySr1|-Q8V^ySuwPEpEl#-Q7cY z)82bO_WQnng@mNLyR>cg-VPjBs1=HMGpRWp3_4Sv&W2(J|oi3M+e$tRBXm?<_{uV}As zP;*kpIrvkY16>?f+ObVer)m_?!~;lo`W>a#@DNV!;Vr;F;%0RKE%^8CNs_5bVRCPO zT#?ybnO@hZ1L2lzNA&@ZE4(#h3@rVh85$|IBzy{#A&P}cUIVngMel9zUWcqsppQ?bbWO?Ix$-Ir3 z)a)%-T|}ty^lY}@d}kqa+HP+JIwdou zVTZez4yWR;;>xcB$hO}APrFu?`AJRR<8QP41qbN~USVssST6fG%_f25Z>eM_`84Zf+jUjZ#J(rI1FDRun*^Nl= zC!WC=?^l_5ryZla;S&>Z$p|du35Y`X@cddTYSiO(xiqt|q&evNxYDDbiaKTg%nvdJnB)D_(YfSe64+n|%)nr=ub-fv(LeoYCv`Hmd3E zy9ohv_66os;6;b0Z%UUp(L;w&wmiZIL=8wl(H;$A^$DVzGPnt=?NJUPnepG-h5vr4 zsam{UtyQ5$d2|Qvm%Aia!!aym3Bx*Y#f90mQXXgXS+OGc>y__|du>IIAQTCWszb7OfD9Xn3#jfkyqQ8C!TLr^*mNH;$m?E?*J_99)vD-JOkKFnr%CpzL$gxny z8{R^K_9xSy@r+H8eK)V03~Ct(I2*|8m}^yx22jqve+L);)xACIqnszfqAomarWM-9 zy|ye5=GPh0twy)!rAoF3S1BtfdzG2?^E|Q3cU231;V zW4R1F1IvJUcjG2ht{yjS?L-#ZCt3&EOK8Ce^=Vg{U5Alph zM}Omfu!s5kFJ39A)T_G$>L0ybkIt_{@2i1I%+2yz-K-4ld{0kqb)s_j#O>!V8w23* z;Mq!U^7XGwjh7K_#n7Ne5pyD1O7TI&-zb??|6e1|)F=l>C-L;4CyL!Po#$F$e^M7^ z*2_-Z)k82rg240rNTybsK^d;@;yPygcYaUIP1;bw&nDoo{MDZd(-WJz!wDD-)+@va zadr35G}`lsL2;%Ye;T&Xzi2A*qZvSeBGzhg8^h@?iDDV37@mo&94!FPH#X`jCV-bI z)xjPVpCkgxFfWXmZ+Uc-*NH`#Te~&ZpYG9Yz0=txpvY!D@WB(Ri=dgnS8%Un4Q%1# z0M@0U{%sPyq|7tRaU*`47x%mhC2B0fK-6C2?C^#Gw*bG{NonIm*Eg7f63=haR;`5;H$%ffJIgf_?mV+e>?+?~jwe1-rO1&#O z5$};(){72UGLj?qy0X;9fC4d~jc+Bh-a4P;b)KsK-=pvY4(hda<+DH?vY-}1wjV{2 zabj1*^MVLBNz~^@0&3rfr{w?=X3`p9`wSw8DZQSTRvjWo? z*s54MVeY{i1en|5i&x9~+lvG7XUwMLWmvjqUkYOD#Ki5dG+#lmS_ zgm^GEy-%w$3|?vY&U0`61$F>U7aQJj;Vpo%fKm01^s1@_^hmllcK|tFq5=9gc*~+Q z)d*{pXt$h`(#yhK+vh1nOmy$py#ehUMbB?Vlo{jw_{kHo1?jlDmX(pLjc#rHZiax~ z6Lj=BpXr=fb3h^+zg24H{3x`voC5^$g9&sK6LJ+9XTin?u0@?WWF2kPupA4Kqwt zT*WKnEy#U~tMN92`&H}SdMYuk1BzN+*4|g-48beuW$khIJk@;d5y$!S6i<0%Hh2&4 zxUNq@T&=fAfupAG52}mM4qnHw{}<2oLrUGxC(wY!i%%XU)Dpqu?GPr|B`-yBGf-x02`HS8_1D3aQ`A7yQ!gI zVEr)V9~fz#>Y&`l%5yh?b*giqn;2+R`AINpRwbu>SdN?R z)-YzptH(_r#+FpHfY>)2{g0q>0U_x>ZhEvjy-rZ^YL#Bm6F_X||y~ERP`(Bw+Q? zw;K01im0fRE21|<#*!MFz-BtGTl``SC8f=sbo_9Xf-n^mT@^o0%$IG-#Z0^$6jp(_ ztG{V?TIh)lL%*UKi=KqJt1*GQJvdu=Iy_0>!7uffd?eYt6-hV#|8apqx-&(9~%&n=H62N6&13D zpwE_Sw9-)irMS2HD7Q>j5WX3u{bsJg{-cwLPQ!c5ZKgL zn;?I@LQ-!>_i;7%=%ulH3%E{M0xklS0tFd}sLQ|_!5hO4?CVuaUFU1=2J=oYp4&|K z{dfrBJsc@zPA}jIt0aylI&V3r4pkVbs#@&w!-TQ{esFgLcUQwwBQ3b;X~e$Eqt7h+ zsO(P1V>bk>ZEMD%N#lMR*;IdOv?hzq`!cek$t^msvu1WYAf17#VYH~6R>YQdz5c|~ zk{i{xam;6B%%yC>7iVQ-NV$T2(gEgpI;=kGETd_wrtLd`yViVg=DTMctXzz>y-AuX zmEFP&(8pB4nY2#6VZ+%Sx=Xo<18 zkkK_lDRc(>5%Fe3FaY0HBVy+mC(oZj14`f9EUcIwWJ@?*!S#%o!~NzVi8qDg&{@Ez zM&9Tupo?Usjzu(YXZH?n`9_TUEiR09GiuT$_6x@r46dD0P@Sobpt-SLC;Ysa11*Ke zAd{w;Pr+%FdP|Vn|Bk`G@pVH02Nun9)bi@CFYAE;xIs4@*^g;UwXOut^)$mnQ~&AF4}W{3>~Y9%_j;a;gHu^blZbRDPW~uu?EZX$Al<$K zGOmlU^eH`$$uaJf_@NS@5aD!ZghE{+FPrz9A~X89p}|9Y%=$v5i+%7cwc&&oM}u7h zIQRM-(qo|jttw26DtG#jWXGUN9((tOYF65Uh@kq&TBYK-E->}E{Zv9lcz5q~d67`O zYSdkkW!vp?7OZ4A`k0%-pyZjrcGJ5{6ANHau39y6xq)=-s2NWhgxZ@mzZ1E=Ry#5- zU)(preeJh@%-XzAhgHSllUckBTIB!*-^@o@xMD*ja#EKlK1z-BJr&6k!t%Xz;i zNFIV6D;g)5Z9NUe^g6Bk*(Z^Mn21e5z+6NZag)g7ZY3|R{SIc({kiPcWk*o4^QAX_ zh0|WMG4jy2;<{;xLt+T1co~5uG2Jjyy)6oI%^APFO5{nu8fWZQp^dEua`xfw5Sy?p zJ6@S_>$04#5#}sxA#*C<_9h}G8lf1~+%5d33%-W?=&$4x<*)Jn!3wmL)tiiIh8rn}eSXR4-#s(`qS082J8;Ed>qeuGiHYhI+` z;P`!B5TmHSPjn?igj3`09yEAV$Cq+2CQYz9z8HO;j%PyxL&(nj`)i-H4tZ$rwFYxz zLeZ%cnOjpBAvfPVu2n8GXR9yIXgLUz(`y#TLnE;YhFGh-pYm6YVoz4mnG9UHka3ulFwY!kD_5pj-du|EHY{OR9$k1UICAwqS#`*|kb zIIeZqzP3Z|YtP`a>*dt$iJ7x=lEvcJEsX2G7nU!-TWQcwV>Wg~eFV$V{--E>_b_-! z;}(Gm2%elJRAKfnNIRU=jI~lGw^sYgpz7r8=nm#NSgem(QsuF+6r7i2ef{_JVAxU@ zX_Q_(?^olv0C@02XBloEM`}XfePPs3<%`kM%2?cRVo6@?5<$}DyugPu|6K!hk z-H0{h0e#lp}cCV@KcZTZJJp* zAp!g&RYWZe&V5wizMS9!wpn3od2ia4uZ&dJO*d~uyme;P&#OH|Jhuj`YR9ZoXZ*Q2 z6w}+%tExQ~(SM2qYETT|ZWd9}hqWBF zyl%fuuU{&H#}V$+dVx~#M2i}rsQ+f-n27-H z6BHDFr^tjJcoUS|-beLkn6GJ;;QetJUdEH(NL@t_*Po#_!asg@Y^Ox#_+Q9eT*o{8 z;<$Up)wbjNd^XOI!Qm~I37VAo>CKnBltUN60UoDs92emM`trgAogjh!>b~bqgF9KO z;rcw?w9bL^Z~c-|ifqUeZI8zg;I;Ry^==lB;8h&-g8;gYL#c{IT^+%<2ZF2H?wc*& z)Zs3I{AhPS9m~V7k(&d8G9J}0`15aJ6z3%DTe$Gsm)ny1rVyZR2)3>jzs^w(aSX)gs54I6^oG-9g}4+jxMv`yt2C! zf8|!X`(f-&hn|>CfKjY&5=sj&9#*G}VXFIe*#J;kWGX6L9o45%of4LvH=F2zZ`Hgm zrS>mSAI+*^B2aw~^IcoA%~zf(h22vMEI$|#R@(Mxj%oJGJVlB#Xg7FZmJiJ7tcpx&Ve=UPdXq~ZF_>Z z5@vGg`5+E@*`grmp~J1Nw~TV$Zi`1qKM|+ib2Fab?60Zv(fV7#j7WZcFwT`7y<&9V z|K-L_S{Y%s$`*b$(R|r9szY>o&@*CCeRB+aJ?Vf=HkfxLccKm+Zz*!OgMxVz!-ob% z6-LY(T1zRehvuAMA)#(v4Rq~jEPrW#0-jdVdLFNmRU}ANlq(g&$s?kX#=e0d-iLtk z59Q+G%1}^pR2d)IpDEKgUwhV83O+nL#7`GMT6@2r>3;sI-BEh~1iIh7SVeq<~qRoRCe+ zBVFq74_M z0!JE4<#HCNvJnkBF~9HzJ-UM^Binz1tA}pLD6NSk1YwZ+ zAqNseOGLwl3Z@QusYU8yUue8bY()+X??dC=5zjH8a*mluJ5tb}kBBVZd5Rl|9#+Fz zWhqCYE0L$TI4q6h@2PDoQwUY!yAm)cXy8whe4~U@6=PbUiyc;;U7F|FwV2MAIP;@P zzM2$`B;}K@Qgq$)qX9K^S&hjlzWzMR7)|*3K)__nv{D--eWk``lSZ_|i}fC7i2Z}M zY;}2d7|~*aH3)WjS~aX*dA827UMUdgu;k-M41lj_VrqiEzItftdvmi+yNV{0*s?Xx z7g~W9dLwU{@5>G1?x6rIQ|aWib|V?h_jD7zeJY3;5?r9InRso*k$EAigjSrMDd8X) zI3+2--llFE>iKlu1o7e&sS@bI0+-sAR_ zb!#m2xb@uV%G3LS718_D{qxnsy)FU1&g-s@Y;Qb&mB6>rAeRysOP#`4Re>)Qb)Gvi zZRW+y3JK2B*_#fowj|Kk`qOXvVtQblvX+Ae#!08FJ1h^2Bi6mg%5x(Z5351%B-bsn z)c|zHzaOt|6(R40cI%fOd69r?@ZqwdA>O!P&3;QP}jf!;n92wcbyjh}6Q% zw;`ukV{t5-bx+@g5o}+zvT)rp_0E_%qPwG5x8(Y9t;H!E9l-I_z5DZCCN?mRr4cph zM;=HkTfG(4ui!RC35S{j`}dLh`(*v+i*7LLLY&MIel)DtWq1~Zd>v_c|JH@d`UCDm zS0@nTVVlPr54=}3r>za$YjfWl=_h|zMz5nG8GPEyyhTJ>q{!*QpVRvLJNY*@4!99e z&b3=dNhJ`dHUDKX7Wlr4Ezl1=b@V;ALn(y;MPnlSiEZzF8<`!6pd+J0GzC>hFeT-2 z6VQ96jxXiqzbui8Cs$S)W9mE%i^x}72%~&B3ShLqW+Kj9IUDTp|FybEg*gkB)zlJK zW3Cz(q7Z^Nl;)VdS$u>hOU$ex2|VJCzjsxtrD~`nNsB$tFLV$xG*@E;7Hct^Q-+^0 zqYP)F7LSjlmZu_ty~kmz_RL_L6oiCoyGM=k znN*e3T^TXN$! zOuSdGn7uxAfnEo;R1cbJk6p)L4Z%N-^Izxck3+6LMwM+sT)zFm1nY7JBvVyo06oZ1 zqtqjK<2m;idK)0GdLx6^OC`T|q9iMmu~154L9@4=^z!|w92vR&w@W{H4+Zo5^sR;l zYO?t$SddsPwsx*YAo#9d&?O#hofnycy6c@i_dD&bQ|`zR@)vFQXB;?hCrG}WP9>N} za#C2)LM>UsB?|U}kMXSN7yYaWmE||XzU~1d;ZM+NF(HYq!=@b622(H!eYrnBqCpgp zShCMfu8}7si=nPdx10~=ycFieUu~-zj)=jB2rk4A$ara!M#sLtx!q!Azi28Db4$+9 z32 zmtoW1kO|(uSiNf5juKG|Ei~~=%t`pDjpT=blTur0M1RVVm50iUL%%HDm{?e4o0*;l zYqEGH{PeD|t~WX#oVDAtsI|Fto|zW!A>mdW$D6aB;Oq}>d8oN~1pfLHT0!B?YtYDE zJDb8SqC35H8>rYXs=T>c*rAhjJUe@cQ#oDlTRF#_-Pn+`xWC1k5u9|u z8|z#5wADQK$?CqFC4b#G=f9W=eB-gH?^)foj^CsD719&m|JU9i3xz7(fVhppt=Ivj z+JI07OzqrxZ@;e9XH*2W(73lq4P9nd7}TOO9i~C~E6RJtarYXJVA!6^2%vo;6*jQl zTcX*!Z(K1Z^Y2oFx3zJZjl?v_!j;0rOwzqXTmc;;iRgzYyDr-=xCej>r z^zYW90ccXAE%!AC|KJjQ{%+}%`4TFo)11Q#**_k5;WH#1>GiF*(+7T?H2J~XkhCe+#iMV*~*vYry&;)eB&YXo<54ftK$ zH9w`_jyC{2HtJuVcPow`_k?f`PW%g+AEO9k5!Zj|;?ZH0QFmNZZ@a4`J!yM-W~j2v zsOxat>HUQ*{xbsn2@Afbyt)&pfJ5Y08Q_b;#Z!v2nTy8}1%bc&8T)(d4tsoU_if8; zAwvThu4~9qdO+{K*IiNf>6-m?`Pera(Da%?%&zj!J*>ZRk~@Mga>px?%S%NQQG4#$ zFe@9Qf0#Ia^UF?X2NMxK?LU@WJ%WPu2|H-QRdrqRB2m;M4ADT~& z7g~Z?b?JLEklQ~%)S!!hQuBc919_l$Zba<*c^N9gwmAwwzu7k^1C6W?%|NE@%X{QS z^dpbkny{Lks~|s;-BV_tE75aJq=*xL2^FbaVa^nv1d|U>ZYM0=u{9_!i=RSyab`c1 zhg`6}p)Rqp3!BlGQj!h3?VN7vv15g_e~1!Q)sdl(8ZpR@fj419-{F*xSdkK*PyKvJ zC%JPdx}RBHWG{_O^8x8XDA9cOlhlmHZsVr&x-Tj~H9cJsE|CogxXo49Z(Q*^=`^F8iqn^OEAGV3 z);^0Wo7l}=>y7*Yk5)w*TVVAQHBI&F_~9lqSbJ^)&jtQn;{7-DhSk0jK)Ws5`)8h& zir}(cO!5i;ha1;nzWB-*M-Ki2K>=c);yylP93s zhS;7fB3FaGJHmCoIG95lX8GZ$67+`Vn zlI<@HV5I+%c;oDE(#Op>~GfF{CyzZ&ARd-(1gxBPT8-L}- zmIg;#OtN-W%lC0Q=A7P3xR8!-5^v_RbnFX+QpW>J*W{!YE>u4Ey!k9 zDoS!gXc0NaFHE$tk=v6$f;X5~M0Zm5b^Hos;m%kc(Uz8$KKL0zlWn7kt{q!gTi547 z(JFon3q5qOX9%7$31*-&HSZj+r@krf8H3}K$j~0U&>ar7~iWQ{H97Qt`I#_R8 z8KLA#Y^{tS?g&V63e76>U$~Onm#+4RW8 z_Z#QqeAb&^#U8FSc<7upBy^I8yC;~5RE>y}i0o^i41o(!XLV~Dg4KP2pI4g9&}M~N zSlwulL@NWb^Q?~(4iL(ip(LclrZcE{heA_FQUWcNbU$3rM$={JitDt9s(#qh9AKjI zU~#LbRGp*eYo#ex$+OM6ApX)evy&O4oyzaGD=bp&_1fEhxvSk9iL$~hKI|%$pe{&9 zwnc2prhMQ}l!6|*LPx4LJe-(N>cBv%BMmm6Hzd_6sM-72s|T(Lk$)`g9tZMFn;y|C zDjpDHfITmoxUR+8zL zixP+I^QkfqvlyEa+foh{-O!3fuluO&-jK~2JUmHB>1wpb;W29TR~jfcd;A>Q@s5$T&)qoN^r84if(@6 z(ej9 z0*qVQmDs;GpdAnCs1W2cZ2efKwFBf`-v;90nWY5*HL>mHhh?MEpu*8-_czx)nBTu$ z#fP7+9_r)#DbKwBmdmg2-JlFn5S(XQ9ozVjO4n@2JOEWoaj0|*xO25@?DZzj`-#FX zhD=LUZUU8!wf=Z5rjq|%7aDcS-cfF1w}$FQH40#MB9qZN`gIPGuMKwQT`oXz3U8ve%o(o=uWOk0FP0X%`6yX<%s`g1y zd%u4Yd6z1Z4Li2Y0q!ff6I`5HUR&D}YOndyVa$&dbpH+Uiiv*ZlpK&Lm?BI@;(0pe z=69mT5(11YLYM^|5Fg$i^Q)swI2fbqpwJ3+tQVni39)@T9Zd4erux8Q=p8MhRqbS> zsN@`;Gpwi(IF?&VzTt#z=49s~-jpX#`4Sfy;y+#d^|xx^k9DqFM32|tiJ{qsn-~FD zTtlXYySB69-BY+kC-I!xClaPngRzWNTwj(HRW-{>M2cY2D-x<-({kiW!#~BY=9K_R z0x)WUgMoVZ;K)Fw+|tgG31I>dq@0x^_qB@9EuphQGx){@s)!{>Z=fIvePlaj^UQVA zh@LGSV^UVB&W~FC5$(gM%b5r7XGlPus*IpqARS3J&EG{%ix>_SMA9c_G+!_aYP+jNR_s#Srae>irbx|JB`|0o z2v%#_u6*??prP0D`0DkZK!mSse$RzOO{+&9g{FaMru6_$I7ja#gsGKYB&{r-Io*m- zPE)nKfi6g(Q4St`$x->9x-TudOp{rxe@FT#fi84Z%Fs!?MsfTv%2Zp3SLM9GmDRNd zq|KGVY4w|LgoJAM=f>hClTi8flt~4APpT$edJYe6A4RHxfy~z{SvHpm+`Nth@L(!bLu)~a5-T5%V+)sxJ~pf%qr$ny zLi=YGfe=8=Q9eF2Z{7M-;rmn=4pHoq3)fNlUGy1x4$eS}t3Mb67gIR&7E1GCfKDy> zOpr@C$_Bm%-1nAM{dm|RrN&WDIArHw*LBEp6hIBb?q)j*OAgIXs7~O!6tQj*9tX5|qB#$x3Xv^i#}YQf$@ z3qJrLX;s|RT2X1J?55b27j0=6cr%t_*C$Tt;3I(@E?WaY$kbc`moy0|x#KVQVtD6! z@k%g~J8k$bLq;cPT-g}!u4p35=}XW&Mo{A(63WU>nbMEn=}40~9Ns#)>S%*SDk-#j z&>HT|j@;?=)XWkG6^SzGe#A7oQl|U}Fk6G@gQhT8cjP}h|JZXs>s{6W)y!J3wm$gH zXbUgCX!JThtAf{KRvY5VMp`M>5m{|NZf=H`*1wvn7(mPSYKJ%MC?toAX<)TP-z`;% zMYy9}XpZ$g)%@y+Y#C))>tM75YPi&`#D65Ssg>-o zB9?&Fv7_#SQytddxWK{x&6zal|wD?Zu>w z#w+=Y{{Tzpmx(cOy?EB#I|()_y7b&LR$r-P)S&JXL_N>^GA@SDB0eF>_dTT4Hky|Kop1R$mPa z2=(OGU!Kkyg_ym4TNZy~p1)trNx*)dS{@+5+Xca0PoP%KINJLl5;mlu3 zzu?lGj#8{h3()XSQ`JqF=iR`0j#!60wk}s*nQA|paxW64Bky*JgDe0CTBgxgFJ$)B^{u^q576rl-q#0l92gms$soh?~$YSD<+$TrEr71b` z&tX^d6X}8H9{0}|qX_gp;8OS7q^hVND_g{*-&a1UnO5SHFInFt3pen86U(Q#EXdVGvsoocf-cDjTLCaVKbPHyo{F`p{@D-f))eTG95^ zQDSm1LFt`VZu@KuoKtA=t@_+7?5EqNdud!Te{dOrJDzM+Max&b{m~qx%L;q%R17s8 zuCTW+6;apz@O7t9GS}c6{=0O!W6x66QvNp>#D>>5rLLtkyb05 z>B||piD{db)K}((yy=~9jDK>5Z%`h2zBofuo66qr zRYpX?9nyF^M(SM}OE$qP*<_GBSJOW}y9tsN(vBHz2Ctie+SeEeRn}0z@j+poo`TV^ z*&!S)dfCI0_tgM_qGBS)&aSB z6V0x)qr+~HT(iS2&M*Gjl0!9CSD?u>>rP$0MLS9kbaX~)so->wX3sxWy7pj{0~2f5 zCkpyvB%uhQ@8?8VE~Wn%OE+zF_f{^cWrI*%M@+J(Czr<8ur%EMa|P$8!pVI3prvfA zl=WlioS9;JJK7zDhSvD}2+>nft(Ya0mS5wo92@{B4vsEJYDjs!48=XC799Pl#0Y|j9(M&Q8#(Uh5YYn)G*yIe^EH*_J3h#avPyN6N{ZYK{y@3J7SV98j zje{doZXc;USq5xtAtz2t({PHW%l z9cC)`<1E=j!8&KI(5hymd@Xe`njCVrn1s!_!ygM*ZUyJg#Ph#Tb+-5%lYnZpF{QK`oj(ANHaOU!?j|kHFF|4 zb7ot)DFRCFh188jJQF39(e~}i-@y7Q$ZC?qXdQpy`~Wg%J${Ys|19=sXF>j6BuEOa zvZx-iPmNLf8`4nGkl{%BM3*DJTf?TJd&Y*1id;>Ph0afRRm+pQYa1KAciJGCS0sF3 z!)wkLm^ML?Dbg;Hes=x04p zwUtUGuwCF(T;rIZ(llo;lO-r3AW`ODs^2+TQ+t&OlSLYA3peinaOQd+ZR{m%8B|m{ zJ?Z98Yb^f4SqxzXP)SZTg`(WLPXsLvr@L3mRZ)mPd@6rN$ch!tB=sXJQ5{xZ?pAmP zjNB5hN~&qhT*v&$DRcNwV7DjnH% zu1>B@-|1khv56nN#yt5y$lo!fuK?ZPIQ$G7@;gPe@88ue%9^iQF8;Zf|Et#6hzIeU z;HI%QHV4J&bn_X;t@GoFH3&rYP}Bt)k-}W)xVSq0K7SR&GdGI^53=^(yi;= zC0rdb^RcbXW|QkLZ}s=Tmwn&F6j?z}zw`TARA1;N9RAej5e}^sS|Ur$Pg)CaDis;- z6)cy=iRS5GpPQ}*8~b)%KjCRlp`dn&K2jrMhz<$$vjo~>Oo-=B)-Of;q39VbM{{I# zOu;c)06*5g+tmRo77kW5neeX_hdd_o8RwSDyOQ~E;OG96^JD+YsHmm3T)a;7T<@{b z3RJOh7rH&nL|)~K8Kn=s8 zrzL+7uq&|;oXK~!3LG09%j2&=A8LPYBwl$xAUaVR7Wu)ZkMYejlb_+2uU&MAdxdu2 zBlhNZ+cJCH6~um-PLRH@RNP-id}9| z(@uMfkCrY6bo`G!1%5g11gO>c&HxEO-1S^ns0W=dbnV=it#sn7y6I>jjQ$1fZ2S*J zD3Ji(K4UsCkaVWiy2i1!8RS7GXun3}1htBT7gb(X=asLK+;&SVey(|5Iq*swO1fwh z#zF<$dX*-vf0#|wr;2HOKqw;iA{Z`wy~}O2;GVDVSaDkC&S<2Yw@dxY)cxOBYUia1 zxUWqfvrYVgvhMF7cuL9dz!Y9{c!YtH2S`@C%*Sk?rM0rhz#Wti$VWUp|7-pwWB$-!*1r9xE%fBBsUZMc*ihWOx~tJrwEs zHJjp1lO&tgp3N)%28+%gC5cj3g<5e&hBp>}rIi0ygG?JN1aKO(Rx{}-a1n#d1x4Qu zg!!rW?&0MXSadNozzx%s!5)P$87@OiBO#G)?MjXlt16rb&^Z=Qc+?dizPMZP}#@`F&I`Zdt z0|p80LPtkOUNr@cBNOYghQpJ3G^*tqE2#Bo%}!=37}J<5m+~sFkYw_HgZY_gvluEx zxDR{X&U8E8p2Dj>Kf-_FwXrHX?JhAvV}>Qr?`#4SXhiwPx=YaKgcLkGRzG(IkP;4= z06t7?3z!Y}h08~#054$i{ZJLhRzfhup_|!VwM@6{I1(OzZc)d?z$NN+KC$KTGKwCh zXlcHA;-b~aku|B$YlT7^8w`m!?!TT649o8gDK-_U%y%R9m4D6cFd>N&46`st=4Hq$ zE$)|dj9y9@raly`&m%#V=chq!c5w8Z$7m>FFO77S0|{HYc~zqF_ea3j;ixtG*<%wuRRqJ{T+dV5O!X^Y#fyUY*1657Z97%0yp&myXKhkk5FogPDI`k;u?cri;n z2;f(vd2K=!sdCk=N1$zlu6C1ElI?3niKm~dPNCP9ix(2|IKOF%U11-j)djwzz(uYY zo+EKoG;3Bs5_Ogd>$sq1B&cdkivLv;->&dZe{6z(ivjG1(rG9L2X( zGU4Gk#wQ8VF?U_lJ_vshfYmMHSj1Z3@sgeg!U0oSs|!(tz-4gUaIBWg`g=?p*nKJKS`Ot!PK+^J$0OE6}6R)sL)iVy!irsI?L`yL;Hl~liifn%#B z?Vp=}m4Lvu53W$qsGb!t4(fxM7T*kBLYwmUFg$Gp{Oh0Bxn#SPJofZsccs%I>Y}5) zt)MD1C3szWJd83r68$hiPT1U6B}CIP3d>R8;xZ(B4+9!~&w|BA=~chdO8;VADh*QX zen1C``twWVN@a{w(`H`NqhG1I?E2zY;G8IHdfJJ1WV7at;1IKwoc`< zc|3O{)qY2CnDjaoCD6d#?e)sj#P#TKRQP5N##2H1xthlp?uTAuFC2D`as)JIi|!TF zf2v&7BdCPX607{1aOeO>S=K*p@Au_SzCl?D`WoBHt;%V3+mL(_iqP0x@IA*f@H3Q? z_sV~KyjTsG58vUfCya%5B$3x7H>RV^eQMOro(oz!G>fH(!UJRI+jCOs;hd8tGNk*3bc!XzvU9mBP(gxS>@Bgw z`o%!1C86(v{I&ssvb%QK>sSa>BI9}RhKwVGL~&aN_IClvF2sSPIA!oA6)c)%3UnF? zmV!E=259f9_i_p4qzCeXk$Nukf;UTyvy#Hi&o_`Fq9_f)OCrRs#j{Jxx&guPe))IQ z%O^nM71w8~mOnVD~5YKlJ_EUSHcF%Z9nKE4=5Sk8r{D2sO zh(7{XS6A&978|5f^C2+Isf1YNef!=3JT*4#15*|%Y4h?J$K8)}uruHhuN)ooyZ3F| z9PTUIh~s9~mYmSc3fgY^WL-}gxQ}ZZV{@;VQ;Z7u8q7W zT+>7F^2eMH%oFsHL@iawZag9Tpsi^=(S~h-%lVB0q&QtyM{d{xYP2t&NTR`2yx)W) z1tOO|>L-3_6j4fmOafw-;B5S?U%&gk+4Gdxgt*`r$k_Y6{focNjQ~Lb9nQzB z48`~_D(fy7R_-Tpk!~?IyZLdn@RtXYxcBNRQ!6)bF1A9n-QEUy-)Ij_OVVk%VIpZV zjeom4QX0{enwTpYH^48&>^DChu{}O7{E7x>Bl9B@qA4QY_@q?zI@eKNH^&Lx0P}o- zLxhV&*Xtoptg4qnufdw?Af&>_mwkXSL@9r`6CXOvRc6-hZ&Q+p_hN3JlqK9PS& zt^aK}Y<7h^Y3Zu`F24(&$25J|^?qQ(!e>Zaj<>zsvZI8YOy;Z1=fUwjc#GW4^Lp<6 zDGGn>fw(QrnZ4QgpRd&2Mz9az?Q++AiC-?m3+fKJRPJ|i`o-u zq4pekaQl@a=*GdhXTW<0wfWVG^wC6kXH32%KYBy(8_zArNeXd&ioBynEmK#TC^_kh`I_^MU z1>#MuMpb*~f>UIusD^b4pJqdUjXy2Bb3_R?8Dso^biH$M{sxTQjk3Yhv3r zC$?=n9czM#ZQHhOd*bBwIp;n2?!Dh%sjjZ-u2etkx7V}R+H0>|W0oB#Bd4K2O}lvC zEav21&u|saPVH*Pg?z#&jGWB$IY3j&R@#ltvl0J{HM(L9N_0@1-F(JxTU?12+ol>k z1&BZuVNk#tf89{U_k8kK$jVGtym>EGl*nQfZewdo(~dXW&DxeKF_ekXt?xquNAbBX zB!qfzJ^*_NPLL+Vff7U<9br|Z@Eve5M8Z7s>UOu~N_ znz$YK)M1v_>lo`_U^KsTU~3322_^KG!2kbkptXMT`z~6Anv%)x{PlgA_ZjW6+~<07 zOVwvT;)*Hz8R`)t`{UbVl+Vp(m(9n*2W0kZ-=OVwW(&D&-G3b>MUY37Nux#ihJ_t@ z4J#XKrKTJHsp?MM>%V^wP5&2qOYESTI4cIw(mZ{K_pJWz1$E?gM5A&31yy`qQZH>h zN0;Zt%J+^IW8bFvxJ=}LdskEMiWlWKDx(z1VNT5D{hZodIJ{1X*W6Yq3a|(he=!^L z+vE3;TD=pbw;_e%Ybv~HH*Ufg%4WuDvt=xxF&Tyl5q&9EU-`Bz?u7MLjvawjtr`bL zv8q5mCl@_8&l9#)ghjzPA|%x84G!QJ)CX`2-wP}GNS&Mn+xLC^MDq&v&j713f!={h?%geCouNiOmJdVuZYIs!FVp{{wW8h zgdnWlJkMtFwT1nT9yW*6x0%N=qAiYEZKnQgHH_ zxT8#;`^k7RfVCr^GsZC;BEFxtZDEd=rRiEhQhOS*_+HFPEH?cub2G>j$AYqg`dd@P zgu8;98oQ5*k;Pz&v)2%yBJ$^#bBWRv*R(d(P^}I~B2^PCf@ec916lP+@!p@}25vhN zc_#bIrav>ZtoDQnPa3>UlXp~jw|$62@6SG+!C-E}yqpM0BDD^7f;vhMCEcovPgbWM zfGIay)jEPR&@5Zx%W25*o(iw=DXIZQD8l7CHGzF&uG!%PTH>$ACldyrgN}ALt6;i~ z|7`i3aez@8zg@^ecbq)jWTm4br2Kg$wi3$uHe3@=CHRO1Bd5lkMblshiH(WmdIv)^ z#kQwdmC}*skC9bnt&zuWXCeuFs2jB++mOinS1)T`GSN?&t;10z(r>a+;FpmFKqu8G zvlZ)wt=b772052L>V%h#8#mvD}y3&P*c->E>?{<|79tPfRk8T&m0GL zK2)qEE|FRt2sjgVvJ=ltxUpCUF4ffkI(QcRODU;>sH{Q zvv2QGqD`1}W=D%K$E*1enGtADm`Ih{%RekC{-zvcLu5DnFvfM+WA3=y#K|&05V`r; z3KwF}vqYhk%2eCehHYR;B2zC6=Cq*?sE+p*M@!yKduWq5C`-Z*QKt|$VJ<7|dLC$> z7%*#Z#~Zv{?V>@I1~!C67)}L@1hjA%I(5e1yW8dX=UQ&Gm>cAc>4Gm--2+-5LVI(s zmXhiR7K2!g750bUyxD6e5ndqL=$Cf=60~c@v@Gt{#ISunM9q(Z4O6Scus-cLsIND4 zU3vz{7wWE{v{J_UmXy%l;HPRr>CMTysN!NY(w~03-r9DlMW*&he{yt~C?)1>?jTDI zarn4l!xHW`KIIB6KFqS7QzUFRVA^g8p^Cr|p@0B1xqnZrEPYTxvVEm{7b3PpdjGg<0!LZM z5PwM2j3wf~-n)_09=aHOy*%d8dZhkXnvBt)cm1syQRSNsQhg2!)Ha+*nsa9A=jK8H}$}n7it~Th8`Tbj@^ehm6*sz6? zjOPwRCkf%t(Eu7~J7U8k7?bcyi5t{tFmDH67N?+Cahji>i*d_;5sIXNA;Bjr@>lu$hkwpd-_ zIMWu;%iRp7%yLVK6lv}RE}2ZM|CG+#iSRy8cNXCDMwNHf+S!8VedY=pPJk9su<@zs zl(Mf7@0r}leQxoizz*W|SyI>E$MKlzsyH=yi|Eh?cy z=c#qAB$Ojke=h@Usmdhd-_>)d)U5fz?Cz~+UTqDI!R-qr%X*D_CLtMnuHk~g!?{~v zaSTsy;_42L5;b?@ip>sid7YoMj@!<7Zq`jxu&+^e*J#xF_%y3%h1_kJ(&$$FQ=k3| zw?77b&P9DcT58_DIw1UwxULG(oZ38`72*P7Oj{=Z{r*uyIqIaXmA$!!@rXI9;c&5o zI!((CD;$SEW`h1HW2~WkrmZb00yZ-F{QCA9iBJOBLnuG#x7eWM!~OSjnzxEpMYDoAuIv5;xxl64X}D;sc14hye|eWdXyLv<3xun|a| zfEu)lsNrRimi7ugEU~mbUisw;Ky&Sh>kSWC8EH%RIu%l%ma^^drk}`4y8SPfb3&uW z#W>Cq4+ADDI?ci6Vdh@WMv%B4?ZYAoOE`s?#~S{R7|8th75#{7PTLSAF-cg^1v%l9 z2S)uAr}a$RjL(p*o5+AT5(G*T*f3Wy33ZYAYLr4 zno{w)F$Ko?wvfB25o-NTF7F$L+O(5AGM{VPRHs~&=L3^8fIqBETa?k)!NCZkSiJO@ z$Lv;sJ3C?Z!3ayUA8TO%nLj2b(hxf3Lz7${5fh?Z*yP%;B%nzC%lZ60norndaq%$s z!+LX9ac}!CtAiCJn2klqyKIcw4$3g&>m%@rR^E9@GofWG2wa`JMnyk>cNGgj_$j^J2wNn;(VcdX11FmnKWeug~j6dsH3K>rsF z|ApBS64+cP^`%(y8r0vdPGD101UQg3&r3=b(yIc!wZ zJrc14TfPx7wDD?r;g-*A-=?Jzsng?bLtTqYVP;tHsQ?ONSr@b&Vpqz8I#vb72mc?sS~9KLlTh_)p}wpT5~3L#_s`?*XDjFUI_TW3vthX1R-Q?=n(D_%tw zQ`pZJ-ltj#&(^W7_!7r6TTOGU87at8>DFDmt`S>clU};mWqm7dDu2;f0slI~jD9pi=W` zhZ^E6vf&1Hm*fXzU{M7jDRqriS9r*`yA-DeV$7C~f0s+?U*Yx0_`fnw}nF!%j0jwyCRk#gmsG=Fh+9qpaTR{VuH2 zPWeJw&OgWbVkk{~BxYY)oRi}X4Kz1e-O#64N=7{$uSp8x<3`h~z=?4?}9iE%sr zCR40sF3dZ>fh--zO#xWPHiPNYB-CBDt&@jh-C@7u&CSigII%Eqi;kojF&5jmgUKKn z=7|!1Jt0>9nUpe($|`J02zp|~*J`&y5os-y>g76xzw2069N~1=(ciSdc_O4k*uFyH ziVX75S4||{3zrjKU@1=Q$%-{wf0}RwG^vhWxNceqR0$c&t&i+81`;BXC&AXF>vW=b z*DN+?L?Em0H||8U&}c(;%85PF%7DY;1Q4zOa?gi7R=)R$wI z;MC;M;^y^8p&)CNgS?ZW4+lVS7`};>$@HIJV|1MIsLXrB^%lp&QB3RsEcANM1CXN& z(xCWBB0enH&D0z1vb8Lzjl(ij2Rx>_l_na*xhN?H^6ZX=HHU#qBZkz-zwYm2BhS@I zvYo?qncIKTdL!i@3n;_~m<42r_!drJTw}&rwS$e?-ytU^F%#l_Y2mEJqb%}I4aMAh z7HnLHi|fAMCK}krtsa2u)b~sZfO6^)-s{Weu}17krVQzh&Y;+txqd_{y|IMer0Oe# zwXZ&-nmKOVL=yJ}(}9&9lXuLa{5ft4@>V)yQz7Z|93oBl2`oZ6rIq}Cac;KkD122{ zv?uSZSO5vlm}X(mV_>JT+rLDJehp7)o<;PPl;@n|UP~n|Acd$vI_mSORIxYKH<^o> zf?y9Bg(_-VYSt0qTvXPf+WjL!rI~}nkP|P_$d}Zs0z(dK#8+`RrniA~HMuW@GAAO$ zMB4pP^pwG{mTJI!fXF}c0je;Ny2{BD^=I2-{wIs;8A8^rFNYD2>PzN9^L!4L8)kHY zDS1*S^+17Vgb!%Fe-|t= z;9%E#<+otLv>-1Dt=7&?3x>FG!)wwi!SpTK9J7-rQmPQj;|u5NZgHPAF|fz_8>V`# zNGYkBzL($cEoZHf!x9VL;}CctMd5LNte7;amX%p*?VB}Dn$_;sD*i&Po+*Y!JvILU zbl$b1_NaldCC83Ob(R`g-Z}-9vMMzsjuuOI*~GI^To!Rs6Z_|i7!mg~e}}IJNaLZ* z_ag__Bwpol4DPDjQ^{B2dEq*hJ^4$l*b_b}MMlj1f(56nm5OOpeu~!nv=dLcH*#Wb zg&h|7wiZ83Y*6<8`<~EDB~m+5vLN_tK`EA;4+evif*8LlcV7LRb7q*|Tufb9n5APT zhb3n81y;kDu=QahL8!FMxWC*QYDWSSldaop+7N5Bkb9e}8Eo(=bY`vrn$PX(w|;Ar z{8x^MhYgcSC^IYt`u=bs6@n^wl0}oZ3A!{elAEw?hj{jnQn8U)D}6Jlj*q)q+sRB9 zUz~5rue*;=XIC^3HEvJE!Zx6=HSE40MPeM_i0n@1z{^zef$*q z!?%Un$zCYv`97VGZiMYX%D}}d&zE~9+nqZKwGMlc!(9HJ;`&X|DgHK)(kC(mM!l+ zs6B-z0E(%RI}XCXC+MC;DL0;GSf;%2vCDDbD^UZzU!;DC+X-L#lf#yjQZDHN=a`iZ z*6G{Bt(Z3=w%A1J$|jG+bbDS=(SC2ot;yNz81O z)8$7gz~Mp8<*QIBBK^~dB;r?|xfxL<69XzlJ^>4=Qdex(i#Ok7ZuWeO5kD}UR)j3Hc{Swr2geqp!bYH;aXRMNWA-7r3_uy{|X zJSbIGRGi$`%xNt}!5}MzQB3YV*y)2&Rl1k( zFAIbQxe?eZ_q9Vp>vGi8SWt;RfzSBwZFGy2aB9Jb2vVVJZa5h?xn&W+Ps|K~H#JA( z9~JG{K(kVHJxc3wShP}=L=fn7Zm|?GPuLwY>)gbD z4y~~lCvkR-t)h@Ba{j_XE%c^>8`earSyGW-M8yN0O^u9QD|VJ3osBggS3-d#Y)r*0 zZ;0T`bXZagL#Q*)I+$eZd6jJJS*R zJf!SykrvRtgL-}tXIrGnYFhaTLlk(fh`^<}hyWw;QSslNDnzAcXn&E{QmUnyC)6=S zW=?bKh=R)2x}Ag%EW=T*g3p|=sKcoc9I>eUa2pMw5wx$D10g=td(EBx!uzY^#zDmW z^GB3%8F5p*47~G@qa=~5yMA{=vIO#BJX`=1S5O~!spP`X4VZFV5pmOnHXQ>DI%bDb z0NFtKWM?k_*GjvZyIep~a*=bd!5*UP^YyK6C%D}6$Q({ANNKpfe!u6zuax*leB)~# z1upUrJJL#jA|{xESPhOtwo8oR6t|e~uSaD17l%9`%bV}IdLPLhpzWRaXjfA^O84mV zM?}0U@LBTrI!wNlQS(8Pa?|X6@@WW^Pq_mD>NrdkIR1L&RV>n z$304@D0IGehegyi_EIW;-UQh+u+lg6ZUGH4TWiAC^dmYS5V&S_LC#>f=gxl}aHD>n zhw05!eq&`kn3=y0_HbhPf~HVBwp$-)##NdtU`t`-MYgx%@(EkV2^;YIAXreLq2Ydr z?JZcc=`;snwPm~W$_Qrn-b4BLCa0~Aq)sj?%`%hmXEoHLBh2Bwwc0lc^*}Q40yGC(KBCoH54Z|p#H5TS>U@Z=E{Sn#P)ucx{0W_E%|l#$%nlw)pDf0vYstVq zWojnaSr3m}MN2<|0b~2LQm@fytt7wJ>4P_h&u^9r3G0L+1QT~DaFfMDDo=(Nc0`fT z4IIA$22WXN+X;Kisnp&?Y}Px_}f{+spyIFw8r~apbOcXXdkrj<}(BVUn(> zc!aXQW)60kGKW@|r&U^Yx==9!fG(OIkhb83nvDx&e5BebKy0Fp?&FqK1FW|ejPb^nH)GW|qNfhW}Any{H8 zr0=JRh?5Yv*Igc-ik3he+CFB!#IvP~lM5(CRK=hQqi}zDAZ5@wj#v_*HEChv)|~o+{liL44I08+9;nXGXI&ZJ+8%WC}FZTZn8P6M)^u(lbCR)oX%ig==8 z9u-VJWM@Q14!kD!Sveg$f72>jb@GEq1 zbH#U4ydJ3Uyq<6A(-kZ4g9X2DOROfK;{f9>+Q8OT=Pu1nRlzeJSrj!`#e>cx;=VY$ zdw%CdizCfdUAT2qsvxCm!PzRNs9K;3S3Wu&c*&7m_s!91Jr3yQ?_{7XQZYkJx}6m% zT4QNbT~`jw;Q|-khAl}fIkYc^yF~lpmJRW$x8Z>!CN)>dPh@cVkrdmNqDBvy*~1jt zDZ7S?QIxini=I;+8*mp6P}}qvT#xO(RRUO7%e;E$+Zms9x=CDS1#Db=Ou(wh)^n`e zZUzanySRK}1li?oT@zym42@^?_e+mY&wTrFAQ zcCT{fY;r)y*`!@o7dfG2SBDvK{Ndxrw$Sw9`>n+WT6l3B7~)-qx8W+{weh~8dbq9n zF+c1ys6QA*WUJQ&6B6u6-)_l+kky>#r|Vf}v)k=#7SS!^X_LEb-}cl>KCV2>$$h|8 zDbi00nFRm7zTg{!@^_Q>|I|3UtoeLDdTOvHKx6#=?&vO&K*GY`Q4vvW3p~@)KDN6S zJmp?M(x&AJcWdfTt6^izro(g#kM$SSW6(hB^8A@T@IP}x)huLPo_})YlVu|Q_lpG& z?fT^+=#GzfkMo{9(B|Uc1$DfDdRc($p=O9biPOPDabS$Ww31-nGpn?E+=xbM1#0oj z$Vl4xxz(0PMnk%mxZ%;+(NqZN_SJy*Dr;}9C?{`{OjXV|H|rl$T4%FV*XG;?fZSm^VJn7~N9SSW4u{0TpKowxKELniS^)rCl*y48? zK!zjHfh#SzJtoF1PtDoP;RWU+dIL(?ROuZ z;3RBU!C_Xc1jvY60<&7Ft=eT!#SaGqOYm$caj2Y@*d#d zw>PrQc#`ZhJ&&sh)COu_EC^Ikw6_NIo899e)t!3G4L;4NE4kiTJnZGK#K?p<0#k|{ zICX2+gbJU|b=&=la@iwlJH?~%`vP7C(ib4$|cA{8zw3OHqgxpI?G)E`$3L__@l7r3q zJTCinqPGK_J*!2}j02ZlnQC+nP#Y8gOn!MPb51}4&_p-a9#YpAC+i0HwDG^zDH-rH zL_Va4KFbU(`x)&=)7n1Iz4(v-_=PoJxdS7E7L}E9CbHaCksD6WSZtL_rqUT?JD7qV zx-#6Dh6$cB|7sq$M*^+nr7jxYd}sU1*GqqDG))|kV6yeSTZfT{T*hFJtBg_-No7Sv zT{VDI+1FQc*fT|4BkCOcBZSW8CcwHaH5M2ycR@#pl!=6`b&sd0(X5dyn$s6ff-NX= zLwGVa@v1dokfSku#{VTQ1K3}S_WiB;md=BV`yaHf34w`qD>Rm(Bvdu6rER(5C{r(5 zR09(wp86gv*OM)Q0u4jR2$-huPThG5%)gQ-Fue^JbR4(ki7D>aa2O% zQYP#dX%Z0*8QGWt!gxxqSVXQCn`|WmV(lU|_BIfIW31obmrUQ_c#Goj@1w!|8fts- zArO8)^u{(Uhi{1Noo9UuG%$i3W@vdk3gI;(7?ly5M95cJ8G|m`BTxU@NUmumP$gB5 zPc~EAZ&FUt%?JoaeV?^*>~oJEtdQ;k*aaNdg--H16meRZmfOa?F^SLl&n~o(Tz5?I z7$H)x%Cl6esam0&x2Y@cU%n!Ad`7Y1rYphEB=_UhUvLqMsILj`#SH7xtX$-VR^3$Q zjsU;c;>=KSn)t1xU`AxBRXX6LAm2NuB_T{TVXqnIybs(u^7uqGrAa@mqI79n`p~q2 zE~KTqhHu&;Kdsa#F!6QNJQsrdahf>Zn0Q0oLDF=!yLf$NrRKybz`_Y__6ex~oLjXh zO#=vH($PXJt6eHqJ{I$+K#vc@sW-(K--o6!;US+Kkt#=5c`LhY&aaL6xa0zGnN8Zn zQHG6xI$CUQu_{lJ8~Vi7@;wb*VUo$_m$?pz8IrHO516hCgf9{j<+-|cXDJHvrbTN$ zYv5V>^){5*uIIw8yN&PQM&1v+0{e6wAQh%MV3$D*wsb+=VI?fLhZC__N=b`4p{Y(- z+5H*6#3YTj%_)>4Y5^a`b9A}WUC4WdV-A{HwhNHSFQ5_hX^`qIZP<@$<NH^WvJo6IE{#13PK&y9c;+Mb2UP65IdW_<5UZd3-r1z--6J`5Xu24D8Yi;-pg6tVN@cij z??1@9j7|#B{yf4b5LTak3ixs#37kx|f^E>>-L%3Aj>P&uZ?*rtS`GVzLfwd>9sQ5# z{fq%j>`jDv2178rR{d23rpGuvC8LaDa>#gm#&om&(>EpFCT3P^J9Su=4%Y5d%8Kg=)Arf^u3Vx;-JM@=1?YhOZa5*|uFv--SABzjtQqJC- zancXVkVuIu4UOs;M?(xyg8D)gXG;YB#>r9s#$&`0=3GF7MVifRG<*JP4t6x86*neqdZXL4>%~^?6jXfy| zq*GZSrkK!X1#M;fps94es75sy+7R+Hnj%kRG_jsN2f1M0mm@D5pnoxm3UcWt0X!k*R!+1%>V{s#*t-f*FZ3NC%schAcROa{i;=U8s@nCDT}FmiCat-QW??5 zo|s9dpo$4j&N|@EBTb#Cg6qY3-=;mtr zNt#jhXt(^#n+xT?#uV9Mp=tDZ+$8@zOz_$|shoQsCzBDRR3tmgRvIi}34(q%-tNP3 zDeCt+5fE->Tw4H#S^ivg#ZdRI5tl7xPs>xv7NR6YoX>21H}9Hu9_zY&1OjVDH2 z4@CdRMJqjXJvE!^Ocuv3n5!yWO}Rp`u>$Sx1OG&UH+&ei7LDcGjvoIE4qPw+Z#g_E zn~onarpwt3Yz;ZyFTx2eCUEh29gx8$wryUw&|EhDR_+USyLNx@(5~O9a*XV0hh z&Ww1bF~h>^NQwzL@(oWb|EucnEuY0YRI+c`YQy3M zz0M6+(J=l?82<;6KmYtO4*p*6e)znf^{O!J9~{-X(DkO^8%Ymro+be?s?D&LXin!0ru1q z9^9m~c`w0Ry?CBtF3U-6qx)c9x3NZ<-1S-TW-^?|vUMnXs&UCN?M2ocH8yIrB?U3q zoO%=%!ZKEZJLYupEXG*4*nJ%J!@@~*a;gQ*vyU;jJVw`MDtuNgv5=HlR>2C?0Q_7e z#GsY^>0Suq2Mu?7_XcOn0@h+SLeLUxc`P(ZE*9}=+GJ6cS<*6Jp@|=3SyIBp^6jz; zL?k!oUM`#9FII`vg{Fu2Z8bBF#K&c%lFNNe3{L$D3HEmZt-Wv+Iaq6tzu9)>!^ES@ zK0_Cwr+&;_enKK&?LBA{xNQ3YT<~!!E}5#`jkWps*)~5CpnuNb^a_X{F zac3-19Jc8&k0E{WQ^G&qlo)dzkTK=zD)3_d?B+CK+#Ox&LcYmHbzLeaBgI#*7G&Rd z7NLdl7Ce&u)f5yUx@ZWm%l0Vq0qy9u2IV*VmCPP-#*9ei3ZBhXL`oqImu}-Au-iHT zTlT(!N3dJ*p}^DpZd>EJ>?Ky1LFoYjTlG-%US@4+*g&XiJ^v~C+hSsEFVqA)Af?8J zV1?0q(;RbB5NU-aH%pP|L?L$HzuHb<0SHF%ZbqN&(O8d{O@Zx^4^>4{tL0-(s^j(W z$URnYKS%4m)~@#rq>rXBMtqvU2^R*{<0wEd*Ry7+}<}X8cbrPbtYbKbvpAs z$x);t^EJ`Qp&#WN>(*Q38Kp&ini;x?%PZBGvY>I4LF>F08Z;K5e>n!1cuv8DhE(-4YOg4{!@nir#p;gZ#c&^+O%&EL%P5$kRHk{|W z!)=jTw+SkBx4W*o%g5QD@uvS_o&VHD&1m0ejNdc}lOX?6OT9`0`Xw7}DYD7!=vCS7 zYN955g;fPR1A7hU+WgvO?6&)hClNQ+d{=g7XDe`hYDz4_Up z5l13^oKOMMD71!yc_jK8OTDj(@pE_=J-jqu&kZ;b-fXUY>?L?&eMvPpg+v(e&5 z;RpH})tBNiJQi{9^*B2YPnz$D0xAofWpC?^SUMYO;3H^E6*5(kA+uM8yQ-R7caRmlSnEs>wBpnrAb@*0 z#ODoyvSQz_LQys$N?O{^5r5QmQWvi&2zA%i0Q~{JEQwK%%%mx99>7?dgDNe?L3GTa z=PAljw<#vXOdF|rsoF?~1t#fJ0g51W0%i@%;_Ig0;qPXHZ2Xk^EpG3BW0>$DPiz=) z-elPYzp>T+jKtOhV-55Dutj|44VhfgIejS^azV%U>VDRAPSCbvg7U-r*g5)`+(`d4 zY?tYhrE9$4rRwBzAJhTD3=g^m;=e_rkl3#nLbA$99c6s3vEVwW!)0)miBrjDa(2(T zz>uXLL&hC`eX(SrnafNh(IAzODy8ckTIa*fcNyk+Bj2~^i z!# zft}&AU^((XZ?UEL9xxitM^$g@>mAEzHSwC*HtR=}9AapEfb5jJL<@ey!F>cuip#%> z;+r?h&{H52K>ztfJ}_1ykOL)CSmzM8X=N@5@lwu6C(IZ@f_XFQIYc#$FwQ4N+GO&p zt=Y#nXZR%RWKG+Qz0vwr8!{2g!2<0}RU(=tTaft^MsE`U$8~O!=xPHN$|#Ci1|Cgk z!?}8WcjaelVYse#jmli%~rfC5<`a2Xmy9bT(;fh z#%GAv9P(+e^${k^luSnV=kdzljA>2jcH{|t^`5Jaxfj3zH!~3EAgCG}ZGovsBC^`1 zHPB5B=_vaVXBzXfOF;{zPEU<8e%_4gQ+r_GOJ zyxv9c+X&l>MQ)rs5^p{r|k`v$PEP25zXEE^%Nuom z=!GdL2fU^G)sXm$@w7oWlMDgva`=oqHPv0!yJ>)(ZA!=tby}arak;K%zFXP(*vFo- zRb%u0S)X>{(xic4dwSg3xTj@d^dcl)NrdQ6*R##u#?HIJs~MlC2haGLY?(xYX3M5e zQq5Nqy$3JcOmZ2BFF%dTb(kYyDwlqn4ObB(usqb`XqtOPB;O9onhB3p7211?SOU~7W7p67n(%Pn)o&&BbRKU|QSA>* z(@WRi%hhy6r{G)dDsH#OYt_lOC9SQ45t-Q4Y8b}JFj$%EN(rjkIEt(x@z4nW|8QAF z0ad3{T}RO>nfdSg6)b)~h$ADC$}u)B*ZWmqO7x9>$N_&6I-M-bi(9S)#aUZB1OFI> z->Dvv2L@)Svcdib{SHjGx4$^a{r>~~bU$B!ZS{l_d~lbLT{*A?`MuPBTTagHI_N-Q z#}e|e?52htgY1(zAf&(1p~{kM%|g|*^}76^ZMkCOA!Tj38Y9Gvgs=!Q-vfD1+rOF7v@Gu_p{T>W^sdaE>nb`nw zu|iqzSON>1&rZUfz!yFVnLSs=7~42~`P)2u*M1%ymHb9xYNo}vps0h@RJndE-6vVS z(Fp(Qv9QHN?psx8e97FB5}xAU9)nugnY2`~S^*wfB=R9?MXCM*f@-PB zO>I328o4lGo0&)a+Po5}4w%|D#ozrXiVDJq$ zg>~}7Rz%wuhce~Tf#nrPC9HBH`4a0K<&R>-3ggFFOLqXuL10J!Bl(X5Yzu&5Z*^%G|XI9Tg?l(*g$nB*x|k+>aHSolP~(isj0e}US0X*8K? zlhw?jMcpd+JbJ-wbUinqb7kr9SVx&?)!v3AvcuFOM}YJHE{^}Jl#7{s5hPz0+wtb` z|EY3SVUSlcOd5^Omunyd6YC)~pAIwa)jD93>2|TGg~@z1%D>p)2)e|e`!%cmn*F9G z6YAr5c6w}03%?j%(Z8=SDyjLOLaZeVs+}pX=pa9#EGOy6sob65JZ7^Bwe#rt+JUpa zda;{v9ZBozsY7bO488W~=&kD@vbS63#Y=WJyR%Sp--txCA?8;w-95vOj=<^M09~3{ z3kj_kt zaL=MtF|&P+ui%=mCE6gG^cLoASc+)#7(+d6c2y4>{Y8bpEW~s1?LGsX(#iGRrUSG- z$W77BP{=rlLoFO;GQ0+531iKzRb4L?R`jO|+a`vbahOpcX zSu9RJcZAf#TpwF5-L_`k`ezn4=hpZQH0Lfl?m2LZh*QB*J`p1YTS%?mHw~pxt8tnd zM8@Z-)H=Zo7wYw31Np^JP6pa}_#Q`ZL^KJfuE88R`P*Z4L~nht5GK(Uh&vcKLEF4g(i&UkDGEDY{E-LBADh5bP% zHN*MpY2P~%t_!G8|KgWhD@eBYY(hyqwL|t#L2hE5)%NF@ox&F0uNdQ#A%i;V!G@Kz zo4G<2=T!*Nr<+heXc#a?(+M2Ea9)wTNPq+G+cQD|KQRm~VZQ_Z zt%rYo&!42$wj460^SV`qM!x+YmfMv?(4}QlAvMYTLbYq~W20R`@+%H*kFd`(9`rJ( z-hn|gaC$Zge8c^ay=A8EJwTvZDG<1Id}FNgybAnaOx@I*|07R?HX9fdW2og>2Kf0e zhqir#Oq5uNn61=BU{#g0hJmpC)~_cb!`iPf+H!eT8^h2E)FG;+0!>K940cBA7w9C* zI~i|smh{W-ufAK6y(dzuT^m>Kdvk%1SS+3*3A1?W0>?1J$etL_gh{x0e=oo62%)Fo zC&w&~G-lA1w%xB8`AgI%rUjmPAQF z)Kv>a`2Nf^4oWqQ+95U|P(C}Rzo9=-IB8rai=Irq85fLac3x~6gM8QsD5eQ@*3t;g#;;l841&fOy!_2^+Z2PC<* zU9RnDgn;sZo*gR$PDX(7wLKcK%$~BxEeqdMqnrz<_P1UVd1^WF+^!nx2KmGcd?J6& z4Z+Btn5TYPp6LD@?O7%LceZ^-MfPry_vC|?=QxGqks9pl(t8krxue7IMf3f3`}kl; zQ`vSwpsriO?J13pQ1%l@?1*kCzq8zsN{En$`0KOmu_sIY_}8XkaE?@3nCT{_3tWGm zV>35oiXR!&8xef^zG1`^;Zlhcen=&kwJm2sk8o``m7w(Q8MI*;Rgt>Sbr@M<4YsqV90*&)`IGI-+3O_7B~BTJ@psPjq}HmfKZyp~$K#qL_= zcBYkg>QejZ{Lk8HjR>;K#bKoWnrR=88$xbc*T7Pn35ZVe_rzNaLwzmWjS*6oKE9^q1{DB-r)1HiSsK?OE=- z!Oy%yhgh1quEUMzY^D}aZTjR43($zdsLf%Pgq4dWSkO`T=6m;06*P6P?0HJM{y#~? ze;yKl*YhmC(|P;F4l8rNe@O9CK3F7tQsemsxFtb5Lu3?ecbwrRIgEdO9or_)7dKRa z=qPwW+T;*L(@8V&@$zxf?OY!(q`kYycTw1+urSai14B|qc5}6KHZC?<&{0zH6@vfK zdfQz{YMNs@^(MRaB`q2zmk|YVjHbP*u(2VEC;QSZ%0ntU{c~ef2r@_bc*r;#>qk6jc2ebPhBCQ2;D)bM9C*qV zDk?_&+tJ!JX5B^})l&`!Gsr*dbTYoHz@``yU(O6YiN6|u}w-!#*frqhaL)TyncKlY{GT`d#w9r z@3{5|a;1P+dNO>HYT#96aGqrht|MdqDJ?ZKM2x)8l<9mG_6w-Gqt^dJ*IP!#p)SFq z!CevtcL>1-1_|z-06~Hc?hqV;dvJ%~?jg9ly9IZ58G^gZoxS(j=j6S2-|t}!ti|fC z>Zud<3Wp<8NFNwtNnz(}F(S>qZ5+u$;>yOO(JE)72~4@~w9d8VUnu4ndeB5`lv zNO}Ri@$OHy3il7~&tZwlO&?E&M(-q&u0lrhhFmfPzP{FSEybPI&qf~J(^HSp;(iaf zWL**zv%S9T{I!5XS22CU5w?=AqI&>aPS`UiFIVaE!F;s&SjjdBsDLJ*aa zp%_p4xItZu47ZM)^JeW>T(a9hB5y_I1-OvRbUADnLNXc(cN-^W4 zO62Qm&_Zu)n*p(3n>FDHCENC-TW^^CLIPxX-lSF|NJ~-|koK!MM_4fPr2?LDO~DWV zV>xc2-5wF4ULF>2>u`zc%IAC}p}k@(vrV6d=6xm$D_GqhfEhPj89Sqg$!SU|joFyj zQMH#q$qx>kq&eenbS~)c9jD~>R8H4}uv4=T@zEj!spXFXmkU9D)UNP9&e^yPT$qPf zzVr=b^C}C7eU~=caDWgvX!Df*h)*&Ah$+5-aEED%&&{*P($8OcdRkYy;B~oRNz0OU z^CHD>0hkx%a23M^>e1PnLhDdP`cbFLgXaiRw@P=Ht8%f-5h>;DI#9|8pDxg+Dt@+K zw}0Ep@KcCfM(U-uSCC|-LJRAc_bshgTQC)pzD>s_-v`&1_MUk2_@2*;H`WGo5?PKC zry4!Wh1vwIIsZ#ipO6aCQjq_*tymJl4qF8>S(>0- z2v}n+*SRlO*sG?VP}@Jt|E6*f@~wkk5Y{|~C+;4rQ^A5kN=N7YbKGaVk;%X=E|I)|b| z_dOrD64sJc!aI{$;frb&)EV0|+o*$_x#L0X{g!-Y`Q#`>goT3nr`bDBC#ntXrV0?F)SFO7sX&#= z@B~o!ZvpOS+X3L0$@7Y_Z9+dr4_e+2=a7pGKj*B0B27kL3s)l-2vn!8tb;E2dZQ~B zrhAewW4C!}Ht;obrk3|nhy3y;6;cgdBUde#FvAOL}Us_GQc<3zkHWXGg zo_?qyP4@j1v-~F%k7m^-)eEqSq^j{MJD3wBo>b;RKnjv4+ zv!C0@z+;YTpCv(V93sCKI(ZodXISczP@Az$)2b}P((W&w~ajyca z9Q@$Ey4^hmor zVl{|im1w?ltna(S;5exstc*Y}-E>+=W47vaxlqDIPjB-jY(%SUCss^7X6^e>QNGGn zIlyW-uxgD8wQAZ5yXd&;DgD^YUV3gCLpVEj#YUFECoDMA2v zpXymu)!cP7oK$A>4k5`(GXG*1AvV$)h5v!^ySJ3DaXHkol(@>3NFqA+AKkJaiOUNn zz9trApUzv-6MT$P&}O~JaO_)uw8U+r4%nb}@nX2xs}6!TxfNdwWV&A_DY9N(({20oS1o7fik;RZGe| zxP8U{1sj}zK8jaTwgbrw&{jjgj*gI=cYLFB%U7{>W)n})d_YQY5bkI38}?Q}J&ij| z0R;4&zG0_u6q=_)U{Q6Q?WrZG@v-Q&t?qd9IScb_nLt@FcqHgj4_>Q0eCGEz3z7B%ShdQqMQIQD37uqUB8|RV6_T%xp9PUQ zcjZ23M8XsZr;^U1I7D4|-XqTykWImehlVx~HyRg=PtxC>TvYVzw+eU2#5n3TbWCXZ_ zA}L046iqbbbj<$1{i*#{BF%SSEomyk9cy*6dc~tlM+;}mKX5Mtz*YFc;B?cd{8r+( z_VdG88G%?OV;_ka+DG1Yo1uL48u-SCSbr)m0A+HVf zj@~tRjpTN)#SaS8X0VT7d}f@QnaFN=|8pLD)ckBQq4sy*fvG^}ufmwCrO;eeo6U22 zmeJ9-;*{8NsXJ*6^Hc0<{%M-9YTelrc!k~38!NVR6H^*0zh+|ry)C~@g7J>N!Q=cJa+XKtq(Wm^JCmxr3Y*p&*I0cxN7)r zjWu>~bw?m5(JFG=6w8bYg|~Z`0rYocRxRCJkH$uNOZMtw1~t~opMs3R{DyvPp{p7r zBW_53MQFf_+bS5tRW}!a5r8iRKB0gkJ+41aBuzb-0an>ld@}<>qMHy2wU4KhKGSf4 z@EZwxRNSe~BzI{Ut7FCj--$oFj4XCBYAvmOt^&CsBo{TZjp!CD`NfhdXguWgOoyah z>Fm#N$;!w9?)fu8t}QZu``hF!q5fmTeIDT8uKT$QnjKAXwN8!5#T}@;8(eRbHUNBCc#o2WRFYujP7T%lkL2gUGwrn1{ zPkU#U?SY#;!xnVlvHhE=^TMh1$J3xV2AM?KtU9V98d6e}f^4$tUm+^B*XeZP<$1Of zJUH^2@N*sC1)ayCr@V3UF@$8Mhh_~lOZUq}NvZJW8WJ!ifI-i{y2&KK`Yy9vLFh@iz0H5vAb)X_rn`Mi%-8#pPj0YuDiYnLLb>j=xV|n&&Z3xy7peP8j7rI;V(oFb zWM=_YGq0EOYLNntk+Ds9s3 zrat`L^8WfE9~I;tUOq|ydRH0g{$`)*g5h2oJ174%gL-hg;fvL;%KoE30X`Kj0!p#1 ztRNilxq%E5$iDK&zYXfIuNaa1hQd(WvD}f&p37Q+uLkA`h4;we9spYl^fM=Rvo)gy zBp#s17MS5CxmZ&BM5OYt68XWYuk2xqBgqtZjOxGA#xK1eS^lVoqgKV6?3ahfo2zK$ ztgWfe=^X&Msmo-Sjr(1Z3h^~f?`n4~Kj}Ec;6lKHAXH#08EARIMRnsq!29gYa4$YZyxhgacz|_E=`Q-v zz>Nys{O-3#6{7z1krV{za*T5ep>3@AC)rx8w*XZ0&WY?QV8cwKp4d$f{sXzNb8nR2 z#Pj6O^5m&dw53}M5NiM7`lZ@StjH0KN%T8r;dknN0nbGhr7YRjy;{4uH2@!kW^T28 z+zWwPbTZ_|w{%NA2b7V%Uupc8K#}$iza8H(9m|(Q;^yO@lU>yQ;QZFb>hM|XgFqQ+ z-HHhsjBUgK-SpH_yyEGvptC9lVdjyLGi`i4${rrKDmHsL70Shkun%ISoBZ<3VL2b$ zFt)sS=|hd!37^_}A+_048}j|+-y%tKr7xGg)7N9;JqxSlgJbKj(E1OB?d*|Llco0n zyebO4n~XFLf=6s?4>iAx9vx$v$?O-}K2n(M7JNZhkg$V0#=B+;cao;gN4hzRhDg5c9)Zavf^-g0w|n?^B_^OO)$6 z=Q{ZA(gr*8Ly%0hE`=XyX2&_+e4el5eiM-&VZGs^}T7kNjY|cRR0d1J~ z?l0PPVx0@up*D5DF8XNsXYE=m)iS+}l*9dcpKgyKlZDPoE9AvAQCoD|HK{@7iV!8O zLa_(R+zRj74iazo$ZcYpnS=WI9m%~T`J^_Ma~E{~ixsa{$JqVYODSE`_GdF^Klo!& zpFGwUS4ocq&Yv~`TgO><55oa9d$WxRc3>W`sB1iE4jp$2!U>1X`}Y54{@MDWjO&7} zPf@aV|E)+%RoLM|)3xI>4#$%=F{BNM;Nx4U$Ay$0-4jr%fza$^iJwOA9CcCwz^^!0 zbhw-Fo!R`&IK{5P_K!UKPvuvbA)n?HscwqTs=xDKM_ii$fHkJqinI-dZVkEG0Nn)h z4J6kQ?Rm>#n>>arys3m!<^8PoJ2j|nfx#0+YZ@skJc5z{n#l*suWtc<&ugPNYGi86j`lRaoy%k{%=bNZsN$2erI@AJ$J z??6d0)TW~;o43d>lc8S=c);h8E*^qgUaQc^Cr5S1P*S-sW`bfqJT7+mhH;deWmG8D zCqq{x+I5+`(%vL9bD_KaCjD)?;!nmcCga9a>r|rkpbJcrT9fx>i z$#m@Z$iU(~)D*e6QZyoiDLT|`F=ne{S{b3t-#!yj&u^BWirZmkWO1c05a>W5bFs9u z-D(KgM|=T>M_G|&p?wstmMpPTC(eu_!&93eg5l1uAI&|zwyt%OKu*57+0C(r9JUT0 zTs(V(BT0Qj8n|~alJ)m)<4U$!uZLD0vEyE5>-jm{rCwQro%yp-Z$}*kG5e_=QdQ!Q zX3MfP0)54NaWil4vVzN^fj_*9cufwb!4WS*{6K3yg)|8SHh@!*>s{ghEDccqoPbvD!E(VJ`9T z+>Ecj;!(o=$9?PIpzIJVck=p5+z>SEP5X)bW5;>5h08$mJAvjq2G8^AB?>(kv$389 z^p?w!g;bjgE?aS={#HF`bfJT5tnRsNAJNHC`;#*Qu`a+iZ#S0sH8`c6u_%#UtJGqi zgFm(`8{YU|Y3>iS`pa%Cql7tVx76)MEHwX@Rf|KH(CMLNN(Ca$EvwhLxWBBf(A9Xr z9t$~eu^`DZ-FkY(Qlo38wVOgPnPluO97>E};oF!K&Y%-Du=H(;@t6N`X$sg>pI4xw zuS6slx6kU0sAVc8{c}_1YYWnP3qR#m9o0j`J&7I4r}rU*W_BkM6@@crI`yT?$h8*+ zJ9vX>cyn)De2m{4-~meWFiXibbcAR*r#DOLtkLSFb5nCw8f*J1@0)qWeQ*caTKti6v4nJL;uCRH74I7=Rf}@`l4sYBsJk_;Mvl%y z8L+3R<~@w%ejKG~mEIS}+tknWmG2A^kN{~{Q-a8!u7%WM#-7h?e3)pNzYgUpr4>&y zA_LP&i)8dz34EuJ5I2_dGG~f{#5-E{s*{8H9NQbfLJdg))bcMJX)inSUB5vdM5|A= zW4<<$Mksr`14uUAgo;(!IJ!BAGc=WB9y(s_%D;V-U9TV$(j!PZ4r|rK${HV|ctLOi zIrC9AVyJTyujP-TMhN}21~dc`CAda#u^~LVWAwk7pDT9S{6EdepX&=m94zSY43jxrdjiq{Wm5gidQb>%J7@jD3P>Lw&kv&|SQ z9-11hHoIp^e4;BOCgSLwQ#86|4|1IAauIRJ+D7}}f~Hdvb^{~mDM#ezl8qA(ZC=u> zSi0;^WQw9xE$Y6pOQS*ueV(|cRH@ucdov8+4g8; z;>?{g=Vjl9{%SFD?v4r)S~>juvCiZd+uKKGo ziz@gKhy&M<^dm>Xj>-BhSB7<6{t?hgoI>|f#7yBa-S}c!zi}0fBQ41>On*iXuzHNG zTc7|^H|hyIKgoDSmzaGzyQES5Yj&c16>IxDysV`e`e>ym1u>q4h3gNBQ2hz|lmNoC ze(&efk1h3uH|$JU?N=o8SXWsox?w^QM?_)`z-uwi7QJj?%mhgz2U#d=;=S$AkrVyT zaBPhZGe{#cm#<*NV~(d|y!xqYlgrdvw%@u+wH1bs--k-Z@OA=vxoN{?k>FE5@!O7` zc;~2xfJxJ`ZT;*s?Wpe+0hmop@Cj$ri0}zc(HLSC);>FUo08fz|5un~#LjAm5?j;c zz7PMIlGk~0pcbEhKbb2sndgfb3-9|M!HmWWhwnSIfJaVtK52@2gH3IoMoCuww)clT z8%db|>UdE9!ZsHL9`-&5l`&De)LS)@>ZnIUqp{XH9rNrWUUgKOr7%z3sRRBF7&onOT>ftz(5)D8oOpSY2}c8e}P zYWzv%dHlxfbbXU+u+)3U;>3aePSWy)kLm^m=*|2&C^c?0|5Gbw5!`luaHEKGs%N~HB0tc)jJ0*{3w0BjIT zmb!WhOT z;6!+iu##uXx&mh}fb>ZM6@MZOHNfGh3%49+e+FCJ_Aze(7M^<7-MfnN+CmLWMut{A z`mHj*f~2XKA@b!a%vZzj@y1t=iRhRK(K_iWo%()z>%o7GzAZ!i)@H*<*H%HZ>OoV6PR@f-)yEJ&0qcCM@P$JcBoLMjdl-eazy8c$)`mdDR~@Ko-A)r zkZ!Z0Kp4m1Y(d*D*cYXDUPZXpxq$AGSqh#@(TiI+C{0ay{6Cy!<>u#=3LBzA%kX}ZskM_x#69V zj|Dpyr777if3w>J3cV|V$Amp@$C$r!G_lwf8>WW@p}^MWp0RvOm&PfxU~elT3Glt? zMXJ&KoF_r^32Ntt)QFt_*vkkY4z=gTJpH?0^DqeFc2RT3S6Do=dR5TRyp-Jy^73wc z?L)&oJa^lKkZ@@beqi^j5aYiH`lp6&@s8rz(RHMkj`}aT|39Co(vGjppxmt4V+SJM zGqJO8bY`S41=M6%UyyhANd|T_4BS&OyJ>#>>0jmW2V(mWfBfcIXUDaiu!sV*!b2tb zG~mo}Z$ndLcvTl_GxIZ^E}}7^J#k%(GM9BA)rj&V9R+F)ScpTU0GDr$mqa8v42q^s zW$aa794#FcIC+FQnH_Dbmsxibr2J415Tc{{Bt_)C;j-cLp zr9f)3g4owY7Hy-k)^C%dv}Jg2XK16j9bBcBf*mW()RwFxDJ2;+1yk1q|69<)PQCi4 z8}{;^9$|vGqBX-365|;adN>AGHIXyPa&Epq?2s)Qt4ZT5hRmR1#lpy|&DI=7E40#- z$@fIbbg=Vz%E&rGcCViA8+Co#AI|V8zK(&2nh|%*l;Y?O`tnwla+rw$?=G;-8{*Fi zV=)px3XYl!#bQI^u<{jf?=E~Ip6lR(QWWC@zgJWN6EGb5gr@@EbzBz!-LP z)OdpnUm&aafpokvK!78XG>y-Lq8O!JRd6KDvcp`A#jkP(6gD?$!kQT_XI)0+!-!|!dwMVuPmR(bE?2P1&S#7tMZBXn7|$e+L2xgq zZEY9cegf?e?&B}pL0Ks%!p4fNqxEYKEI*bvza@nk7k7v^ET~PKz1Q~JcFus)l>~4% z-AwBVR*TVRfLuE`e$MR(N8Eo#MM!XYszHrSU!nv#^|M)Q+gGRb%!+vA?|wU~yqeQ$&}y+ZflVhrfm`#saQ+k!MhL!L1T>9@WvwWEuPGE$-Ylg2XA%BecTnVWIf1`yw{5>^DnR_BUQ1HH?Hhy?e`9rUa`@J7b@aVr ziK55oPClJyi1SZl3_4AIDRWZY>@lG|*p0`LgwbX{L3CPX^7lS?4PRwq#c3~;9yz-i zSnHsIGmhdqGJ-rfchFH}Ll$+vpj?IBTV}fP6`u4BpJp)X9Gl=f&daaInHG*84o^7v z02(2=9*!+=J6+ z9$7pNK;^;&+UFW5J%`%&BSDS0*HMv4_oLQxn0sPA##{vN)1+5jZ=OVxIW;)+FcXHY zE-XOLQKz}}nco7$$$!M+63nwA6N+{Is#@im6@EPhN!pJ_3#-~6C6o{`H)BH*}NIlP;uaFB9T6B+5OGO(>Mx-zPe2P!x9B7{jBZ3S(U!>do{ z``XAl`WUH`wFwj;14+~IVxc#2nfV5WuOv=B=bC)0QroFeP?qRgh1++=ylxAM=g-eC zAqC=Nkfh*pI&B}jcSO#mvJOzT9E^h9Cd_NP<4*n{FS40^{&j6-S-smKp75J|j>3D{Bi-ftBTJ?#DV01Z($BqlbaH`F#3PTK0WnCb zO<&WMxMd};3TM-p-EUalyUP@Y=Vz#fx>3s{Mw`2r>Idh`!l$p7R9vl^BmAQ}+Aso{ zTAw#BoWjDB4EpgM7i@naxW$I1(hVJwAvJI*z`6f;g5cd)UwD?eJo`?PJ^hVBmMXB< zoHy5{yzln8{-b22_Rt?uC51#B-YC7udUEbQoQ@VV z+HPk~Ly!eCuDK-FkL}KOKH^+RH1LB2AFNAuiPehu3WV&dbb6G+i1))s)(sl1kz}yj z`xg{w3$Xvdzdsr13^Mpd;=V{l9{2TM6X`!6WeLF0GE1-WwJ8l6IkZEQQ{dmGHh8`K?X{eQsjl&KrkPU!FGEVo%0;i3Dek=;X>aY+Bx_ z$wMc__1xZa$AgOyw6EYsw-w+t*uqndJrkGw6ZaRud8L!P!KP@Mlc#=H3tWA&1j*4a zf+-jMw;=KNyIWDEF3o9p=5O1^5wAG}W#NCJQERHqNbCK++PCpCEz8tc?=S1Ev}%SR zCA7gI*dHP=ACUkBtQ{I0e#?SikcdXjOEB@eIJty}1kZ$AEV>^?54xJwFT?VZSuiZVveSud`ZiWrhAZ70(?zty)!SwYpF zoy#9lMA9TkzZ>@%-&MZ&;e7J#3`sE9oVHTdR1hIx$bjUa-`^Ce1&6aa8P6_KvU%n? zZ%PrFUk~X8ZueY8%WpSv#$B2#RJt1DY^EjktiH1+8NKB+-s}U<2rTPtTeS>19}YP3 zn7}VZUt1HV`u5agc4DM*8vCfS3OJDaYU%=M#;;G_Pww*v`_0BP0=DV#1(s9>vE}t+ z&A|{-_3!|SbC5TjDv6s1suSFlmsFD9j{p7ufq6r$IR@>7He1ATT)7t?rahAuvHEJb z2zalz%_OX8AbBS_nb;`Rgf7d4U0YJOljg#A-B(_~yNt3@_OEp6n+eD}dt?l&IoC+6 z2EEF8XqeFyjMjTX}(s6>4a^40n(P3^#`L@b>;G^^TBD6KDGI6l->u(HX5&fgMZ zKreLzM*u%mqNVtHtN7_XFPK*Wo`H)5sIG%29f3~}2T{p3HCEJ3BI5BVPZ`{Vy!N1< zsTMe4_Y9@ej>&4u8Z%X$=F#vgHx+{42}S)vhEV-xRKH&jWBy80eTsXG%r2lbl>>85VZOf*&4)>{!^Fa_ zd)Z(s>-eAG86NCJe>0`a{NtgfX!K&&l#+ruKmIast95m2kNg5^)-KqXCp3clG+VU# zL)UfvNj#L~X0kTT!l_$h*0)^m^=3Sv*Mv_@lF#%xhH-_40%H`0l6RTDQ% zo~kPJg@^VuK@y7+7H0!~Z@g2p`VagHZ1nQ0<-Mx|_r8nX&uj}M%Vz@+y@yO8j6%ig zaWnm441CiJOVv$@=o#;Lib6*CmsY$DK89Y2Yph)Av4k@2uYh53*abnz+SqJwhD}otv zHAviUhFI>l1@U?&s>(^%kmdibDU?%sR%+cTQE-h4`qs8!rKHQS` z>dMv(2BG#L5akSnYnJ3!6yu-wGGvl!S&8orgX4I@iNxM1`Y4OHU`b1|cSTdqJ3e@x z_C@w*`MV^Zk|YlKVOw#l&PuD$$z>DFQ9R4!AfNXfptzj1k9Bz8aha|QJzsykMRK?{ zvI{6#yfGKs9=sU}n`$+}2Fq!a_sq8%Hpbd7no+djn(EfO|trt?R;-wqimBP^H!kz-vLNqi`aV;=YqSn3G`C zXa%J$2>VMxpLX{B_OGha%CfQ)4~wC%DB)?dtOR1#Gkx{HNRfS$b`U z0L2|FBrg~_Cw|oO>#SSX_*3%}G8f0QPLr~Ll6`{TM77ylhNM3)ikiQS7|59i6kSm! z9bAa`+Dpban`1X1t%&7ra@Vpg&T^EjBI$e_v@QAy9)7$IX}cj+L^G9v;wl0QtZRrR{sfDHI7v8}tYDOmaMwjrcBz3V!7~DnM=|*hkjJIiXpl1&}cI704*emdwgp)9s1lU^VZE3Uy8AlHdAfM)K@}h-i#(?=DaP@iVIa49PF5k zk5*Td$RkDM*G~D%;4HL#{DAuESynGlj&?>G{wHvllO6V!Ku;+1%&U?GC(UIwa@DsYQyoI+( zsx9drH=Dskhn|oGc!W4X$~eAMxSIPj&q~zwNRP8#enFe%RK2a9-1;FlpOT1FwOU%P z!qs5?c)1uiz*3yQ{yY>Tshf;^S3Q$KWWa<>67unEb$jpFC(`=rId*ve(8hh!{`97D ztgA+<_4=F+ZxF;8;4O*@7&{x^ zd}!kh_qvT$5Ji#_Y(O{Mm`bAqK*6u-@_mz(v?rX0>Xg@I$C32%`J~8g|NIKej%q;X z>%_7i#vFFTaoe)kp`wn@#VkI)L%ASOC%bMFy7%M3Ntf=^1kXTLcOEMXk__E`nO9>B zj!7b-o|&p6eaX8CM|NF5rTH6JR8+>jxWOPd1#P|g`kPneR*TLN);4`Ae)j^ZVauy0 zB1qEhdENS>$8e^Np_t*D%Lc|ca-KehEz>ZvYe^UaS<3Fap?ft2wRu+fil>Xr;>UNv z^b3Lot2PNrIpcB0HMxE?ru;>5rFxkbbS;m^zkw}idxFVg+2)cKbBGsk`vUPcdjUSA z-|eD=5#pP|uB7x$8frT{DyJ~V)P%|bb7f^;dtojDC|sOAV0+01Ci7dsN%UIn0`1dC zLL<9aOXSg>-KUAtJvW!+O)i^>v{P2yy( zAbNG%??BC=$Mw`|;<8xOefVRtIq0sRL+GV$+sChRV->yEO@1am#{7+5P|fNx(SsHK zz^URb^*Etk1!l+m(g zK{bEq)EyYqSt09-JiFQlqpYuUS?77|vRF%JDu-xhyWtZtr7SkS6uX6=$;S1;+sj@$ zaqP3twohz*7x~cV@7BUHb}$j3RtoP;Y~{|hQ`6Y%A95Y|-I8gig-rXjKgyKJcrGRA ziireOq$F0UwDhwJ)b*n^R3}u@fgj#C?3C^)c>CV#lsRGe>bz2lEHK;0$^-8UEf6E1 z2fyJ9LVXl^_v%?X3ori2H>4)0x9_bFv`D$1v1PEgHcJ4pRRO>0v(Z9!0NyOW$XyDzEYGe;=|^i_Ev$zqM$H zEW5qSUoTs`Yrpbfx;|eivMoxs)aR}55qN1OegpaRMGu=wo1tG>Os_Exz3VWJ^j9Oy zZ5zZp5GNbc;VxFi$5>2ArJ~sQm(GM)z5k}N;W|RX`aI(AIW_26u+AEpb{%kX9QCVo z>v#y&-jjos)DGBIC@$GI94&1%I!wh`J%vi^{STvxmDB6=K=7IE!$y#vSs8MkC`B+|YHPx#D$)SOO>BTyoUSv(TIX?-To^K^@AsKQq++Vtll8ZZ$k9^>)U zmffv@6$TNgd4Qoov#)rMD3yn3-HnD>M5kAp`|umV;nkCuq1QQc!m(cti{q?e5ZAr8 zL}AnZcZ!G$>l>TLcX3t*LN#@n2QHa)XjhVWl4kGL0<5Ds9U5!1{fsj=L)En<{BayH zItO(eb(k>W8W2@j-|BEO`PmCwU8W8>eNde~Fs1ZH|Jl9YI*FE67PlN|e}zu1ba|y&f|4@Kduw_Y7u@H2&z{fO*SOtM*SuRg(fNiUnIln!t4XL4P3e@<^?{hCGC;q7^M?oMnx@T2lTq0#asSj#`F_S08P^ zE39wQSQI%J;u{<>-RWpJb>gPjlEheGK+T1!a(Bkdtw@bia&9(fs23pm(JT<<%iI(v zpUonzDJRFQl9Rv|Z-ciG;9MeUr2J?V`kt|$4G9%3!qeV>B-O9arA?PCkzx3Es8bfj zS0vPN>4dul(F+Dtf{5(TMc`emJGhSLdW+O}4#aEVWLY?CyHnbz(MGybeirx*f_U#& z!cpBOcxZ$&aH5lv;Hy)2!4Z3f%Q||vk+W!*THs&_FEdwYUDB*2Wg~QdlE%pX>~3sU z#!P(Vxd&nk4B^+`UIhXB6f{@3+5vJXT23eodzdaXcA&!}0gSal@7IF%klU0#RdRxU zNl-B9dZc9;FObhsWa`HOCpS8X!`_9LHyeH5w?=x#$5)hVnEAS+WcTeiJE`@s^&1@@ zzhTd)C#h`hLdpt~Wlq$_Q1s*>K^r@Ud%S*JF)jU_>9Ut z63QG6G@Q;0GYgA#EjW3Nknr;tkV>x2*y+pZ`kk|?BklMg52w?&K{S*-vf{481NGkY z@4e}w{r!BjINu|G+t=@3i7;&HI=0RKU z)Jj)z`0Xeir=-wC4|YqGuy>bErv?=*a@FKA3XcpXDh=*#4_1r24!$;JU?31J`5^6s z{a4O_r`i6=lHjF5CzsAl1#h)!J3&Yu^7fhiL{E6H01PScNRslcTx^I;7>l}SHeE*7 z$*i%&tDR4)fOBR|_K!2G^6B9Fgzk+txu+gcgk*p@&80mm+{ zc5RQ2u$a@ZKeyv~)ZrM^hM4nuS<3gx3-pqTdA!7wZmB3xiK$rt_FW&!iQ7}hbu}*UlYx)tTj_c(`Qv>=4WoH2 za#;SY`X9DWU8`SnzO8U50)`$Sx{#fyGZ#PJ#xF$GqJ9M#{&1jb=iORgx)+82iP zsK_zvlY~?AJx9h`eOuj=pis$Ml^am|7`+7IruLwmRftbVOUGw! zO^`vW+D%ZUqo)jlhLe~LVD6%G5H>}S2xtC=b;+Yor5%vFM>nu7#tBfrhwkC9x6VTD zoocEUwkm^N3goxFc0pvLq=lc5P((w?#P?tZ?B>MklJ#17Bej>l{pg5)`&jIq&i_0v zMR7F)soj@ds-Gad-0blixfQ+E0oy%dQ_+bRQ%wpWUDKOYkw__8lEx4M_L;6l`fO!J z;*bEHIsR1LM4=)>PVm0^@;BB5Eg;yn$CxIt<6uws% zpIYwL`Ma>|xA9?#OK7|YuGVr%fHF)vm)c!8I|n!HGa79?Ro*w!CZaog#AtFHn_t{P zPu3*4tDbqwJvA5#YDNBPr_Pt8GH(tor@I%0CXa$B0iv1v$X&gWScQbdLSgf3!wwjj zUeCH66I1OdaQ64N-i!9l-|NlvLr9b8_q(PRM9JN%RMntL*pQ#Jr}v#OcG@qB=?lYd zSId(^5o5WuL{5kNYk_+N2R8k+g<#krXu1sGr-OWXgAzXEDfS(Lkf8YFtWv5fah3}m z5j!n<2rMxlJ#fVB^2)+gt6I|I*R#L#E*SGeRHoC3{i^ekPR#b7s+veiea6ylObT_| zmY4HM|Ej|dRcdG&Me`_J(Hk0;Y`{fvael}is>MiXOGiXcy`%Agx2mnbYh|+N z(Io5uvlrfH$HUwT6tUH>Urh|3fctn{840>$Z2?Dux=*T4PU-)V{{9SB|J~72gF5yZ z4LId>3u*u8DhkMli>!!B?CdOjHqT;GWTZhleCK!$lGNA{duXESop%8e#uH!VOTs?< zZ!6PDwD2ng6F>M9O?(A1u$%^iKQqhMdyg{UNl3E&y zDxl~JS1(5=aY`c0@P1^u%{m``EkPh(g(&u!1cIHh> z1^Nzc_(%ZcgK$jLWP`j3iW0+ENauzs+~ByKa_=M?W}gq9mu@zmUae5#bN7e9lM(wN z9VD1r)5Vi2&@JfGFO(X61S-cJ*jej1PS5;GZ_xTC)i{RWBSk(JvpzZaz62OYkyt29 z8I+^VujUFkuD)=;p{6?va}!A!>|`uQ7vU|wF}^lXG#es4L?XVHip2BPOC##)imjNgdQ>&J?G>VQz@3!_4yfC)~>g?2vd_+r9v; z2Sn@@s1SC@0~x_DqSExdXe(#dUs)02`ZoCK3wIHeTx~Y2$Do0RRo9s#v#nlqrZ3tM z6LX^Ik3+Ro2M|cUi`UD}9(UlkBcDFJ$bHl~=Utbb>%S`*Swdl-NHZ*pCrFSM98f#6 zPh&#hNL%t?;dT32WQDine8WQqYGqy8(P_bCf+( zor@?WqLe${5BbK~9jbWX>%iuX;lqVpm)$oPBTpPB+|Vs%jB4dZox3b8D}^cnZ!0V% z0{?k$_D7HSn|UllFZAFi5c{SR{ZCq0W%S8(-4khe85hxu}?S2UQzKhjyZL~lxSeur*F;`TTSS(z8L}EayLoEA4N5!H9zd`riOIb zbl{xTZFOuY;E?}WC8qh9O52=#mDgUVN9{gkG|(ee=N-J|v|6HPg@4Yeh6rw%=Q`v; zPp_GrLB%tpva0$P-K-nI_PUc+1T|dPxbR2Ax3*@5z|WsN2*(PP9P3Y^L8=%r>kqlN z7J_5}TAu;@RZU`)NK0TuTGWL@eWV^>pO7mv-0y;U`O+ATLad#(V%ta*Ak$QtS%9D-TDcxDcxSZ_^Q@J^NcV8Dm-H)0IUBKG z=QX~%`EStE3P1nDi?+{+5Ib5qKRHre!PRv1kE)macF#8nJFlim9l^tWHB`Si8$ZMD zvb~8Yga{uiOLzhmMHT;=m5eZ>)$iy(ED6DW?C+=8xOcV9PMV6xM+h&`TA_GnL?yeh zPSiARJ8pER>%rd3`Lkkc9B!oHt1ND2RWL3f>RhClFBGyu@b2^GBad!LchssPf)DUy z{6PrW^kfx=F}}4gC-h)jbW1Coq=ER|s&jJ;t>c*X&lXv5VSuYcX-W2nukZ4iCrv)8 z_~DfM^WS|JX1)0Uv1Q*k(!uB9MH+DizWCiRxMt-{BS%#BJ=Mq0H4eg z4wWt)R1R+m&oU00&pk!(IWT<7oK#-A_p^<_x7C}~r$x9{G3YBIPs!gM={0*Qw@cfR zzRVb3KgF<458YR{vO;&$d6jWOvI+kBbN)X_FTU>fnx(v}L5E4{Z%6WP{twOAM8Qi2 z1+ANAdH}AER*j43quE|yjJ)(aC$A=zlc@gq9$Or2TV9fNhpLxd2@MO<4B~X|+Vy(u z4l^^ZAA6$3SRURiBgxFKQpW?k zT&f{#7m-p*vymVexeg-Xpz$kzw^mFDv(B5);K_HCC^P!z=8MVfZ})<#r0|7)S2hH# z!8SVwKcSTYI*)5Q<7oimW(SRjj=`W641F}}9xV-k|JfBpar`y52RUFhc*=5Nf0>;_4L_Zr}s z`<;c|^3ZwYg7bgo6dQd&M8immIW)mcNQZ_w-+R*tIg+hl+y~Z}rz`s&b$CMk{B+N6Bu6?4W1kvdB{ ze;%EeNM`>jRF+Y-(Ry7lcl=Rc%-FF1j2%KizL*~aeozsLauM|R%IXb*Q{7?L2UR*^ zZ3J}HX@(0~Tf*nQiRcY7SJR~BmuC)kO$mX06C*W{Q_gNIsB%W-m+FqBGv}j>yE$@Z zRwm2@X=MWMhJOiNbQ=$3gX04>v>!Gqy`Q{ACk?)-q~VACCK`3*IO%S`YGz@bhv5@E zvyfIFB9Cx0Iaf>?EFLWhZTQ?mB4WQXz5ijWkt{6YOOP#Zc~o!f4W}?uE_Q;{w8Zcb z;f04D;ToBJ_go&DEu~kX`e@({8hw^?&+5{eZ0aW<=gorLYAM99zwF4dU}tyUnq3Q6 zke&i8(o}!{)9U}Q_0>UbuG`l|Qrv@-;)UYYLXqI^F2RFCad(GO+}+*XU0S52xD|Kz z;{N6I+;eX4H@`pMWb(W-nP=_2_S&``XV{uO-{4^SUUz|3RtC0OQvP?HcXc1)pi3gS zFadycTIn;Z`J(C7=`VkfZt~z=KMIpg{LhO=j71dQbTT8xA2;XyS)&c{9W!p+Ajbi| zW%F~*dU_28HjE%g))ldZO+Z!qPkY8o<61l+NzYDJz_LF&&Sd3`HA?J)(5t!PF_1Nh z4Sy(s1ulmQq2{a<*ot3)tqsG7f#`KZXXoj(d5^w*1fmb89Q%gGV90TN4DmKTLPaytsZ_iEmm(9Zr$a}L&(sQRyVqf! z4X|%^*%T!cAq!?6F^$0vY}%a9Sk3SF3srM0M>{itkqrSU^lqHzjW@Y@IL!pE4Ytdi>AilG7?kFCCO;*0!6otE53xK} zL1&BFbPQ9E|4lTNaiWbMUA7C4X`CYe7OYd@4{zG>?x#N3V!YD*9KiL?h8l8~^6DXb z@rNA9g7S1~=VdqPMAnf;O$my|8bHyaJhYY_aj|Xy!TA`3hxa(Y=Y)vNf%R=US_8tB zq@86^p3rTtgPw@Df@Vn}GKKqp`l}`)m^+jqMwjpA-TzQVYdP5Q&t{U*5}_!dlxC&# z#^emGl?7@WIRiGjKh7wZ>!#uaH{x%V>n?1M_7L&2lT$Iv81e2vL2RDpX4FpLSO*=- z*=b%GHJeK7u@=hex`-wi#i~y{p!uv3)6KmkiUOd|p)Acg&7wp2_}QNjme!1+^?UO< zqX&&f^P7(;ydY1=##aZ1&N6amnNKPlohkxLcS+xAbSI@s7i7L`?z$W%7`yJM>Hz!g zNAr(Wkhc^)FD=(N5*{Uza6XpMiD)4*oII@JujW19pU}C@MVb9LT+X3d#&-}v;0Avu zY5S!YX1bsW)0ac|o~QG=hDpLss8-t?)P5d1zhcPD*sG!jjC=dqZIbCRbBDdVI2Vzz zf#DaVc~RdHZRXOX95lSzYX^#fL0S5A&K826cz@i|mZHAEwefn^IZdmu$)kP+0LK^@ z|Agn?j@5;lAObL0SX{Mk_WnS><$P1qTE;4|^VRf-zL}sw)pf%o*JAB#7F$T|LlE4c z#5$hg_SOr^EhefgI~{U|YNH&5YiN*}75$;zQQD28y2Oa1*_>M;-^>QlWC>45UoDF? zgB~v8*0L)31WkRNiql?=bSJ43H#iPw-Dzj3CPNia!+J^d3hG%SxWKKp$AKwS3>?%N znq{W}d9Dcu%a?{ltWrqwN)7d59VBF=57`4Gg_&}lVFX|C8`HUTi2 zrqD==I>+we=hZqbUkA??zFqZH!FS3$uK$wKa{e=laFc$W7v_}c*V+<2bvBskL*6F3 zQ>J7OH-p-Rw@!GL^|n5UOE}9I#cyj%NEo7_#4P3^Q6*niKGFi3m8MF59`%8l-ZEy? zSeYqR$ZUrVdp;W_duoyw?l8c+*iSK!`U@`AT!<3bkGDP!=}7y4-;tXaJY^GVjXZ|0 zY1TD5uBrsG|&p+CY6#-w=9?lbJ4N zeK%riy!&$8z~xZhv%fS!75-9`YH zfqx?zO6OcL;3;yV6}a^Vg_)|D!IAW^((FU1NDKhA|7l-PgI{D)McuT|^&n{FT|yRg z5aZwm&l|tuvF2bQUscn5py0S09?2-P_JO6 zf2fu50FgCKdm2GRO5#fEOL13TeN(ZEWU`^W@Z?Wf|3cazN9PTHa|WecDO_n z?Ri~a9~^&s4>ET#bV(AbU16=kvO!)uWj5` zUc1@E5F$uPB^CcQ-Em-pmB>~6mn+N~`WhzQPa+c+Qy1r2pDy%CdH%^df){N+VNUN= zVU`{*aQGj8yjVK-ya#Q_ukt7ge#0Xt`@BjMl@_oy;Yrf@oFMHuk7d!~43>ajTR06? zC?&AQVs%VqV8SPh?3EbWjHx&5(9atUkwI}~e`CX?{;9hhI`mYnlX2gVn04d+Gkihm zP_ih#G+o;vd*{$Z@gO}ux8K)XaXk%Rb=^{@IjxJ=wsm|9XUafgO)m}Tf5_@NC@sn{ z;u~fG%()OBV`bi(?BZ8a%0l5>*6v3!|K{h&&Oz}3hLIeX^3|PKx&a5Pjl@kUMqaWkUO3Yat^vkGUI)upBVC4{DMmBEpKi^y z1jKKTy?6jVcRt5{&u7nVXe7@^x@HIU)b=Q?KInOfD&cb>L?a;JeQvXSjvLPVuhCGk z3S}1}2`gPiiHMr*QGjcmWlPGf+bBEdK);vk`4B-?S-*cbWb>ojO~1!%_6ekq3&K2L zS;YHxse~>JDe))0En}J$9kC(yIifPmIq5B6r2+oGi*kt1nO7$U7W+0AeRy3)|U$QRby0!3qY7LIOR2!lo)=+9U8Zxlb4B{n%P zx4QnA2Tok1V8hKY%PNc1c4Vp zHozEqV5BO2GmMLYBGAokJP7@HQ1CdB90*=E!$KzXdP{CtC#q%ftMx~fK)0<)eLtxH z){@vMCLnPm zfK}G%D0|hiT|Qcb5c*j)i%Hs*rtdzmbD@lOD8`4LcFL5~%$&4%la02_=lih5G&6Ct zDO8Hxt{9=sdBS9A1EtUt((ua*NXG*qU-Njk#t65bLYNt~ywBH*i zOm2q4DGKR?QF)DwC00{N#7o|WYIppq@|hPeHn4$JGBC*rYd_@Z@t61~v&)~%_W%QZ8T>E)vKvR+fgh;{9x?SmBdd$?vLF~Np` zWCzV&uIh=*4EM<)aowPXtY_^N^@kCj*5z8EEMgGv{XVn))+NWq!vp?b*DhNF$?_^p zI@M`$k@>IN+dV9jeKg1PdC`0&PSbE`bQ5wngip?~-Hu;9az2yM{e^VTe@$m(hup2b z>G2A`e-XuYzZL#!ye=g9(el%~u3wF3wFOQnAJ)D_f`jwa)ej+iPMw1?RpqXRwYI*{ zjg$XBT>ipN6AjG$?6wy01xDJxh=Q(Hq+GueQn3)(6Ezg1W_*{PYR7b%2Ems`!LdpC zFX9xE1iEcLPxO)(`4(CGrLDqqfn%i$v=r_l2L113nDU2r+a;M|d37?Zn>pBqQ-{08N_*CuFSQ*7#4+i6wXJBq%guYSmVX3ijGgaV=aC{K?Vq7 z?@yjuJ?bWB3d3mS!{Y3LawZ+)QO6|72QL-F^`2evG}%6>IcFjS@6$#gB|6j-K^>;c9? zkisF&z|U`}^-id}-*nGoH)X|@_US`j;-1$D3-w`gt7wj6$kmu*_a$_;XNWOXc$}tM z2oWaeBt%x+6)KwDJZBHu5A&?tc6*d7SCb)G?*%sb%9XoTba28LVVO3N9m8vX&lI)W z;(S*EcCIpjS!B*ObAZX93fnz37o8(XP1rm&LY3&Cw1?(H|3q&*YsuN}cb!G3eI!#m&YQ)DnW@U}`vOo9kobc;0egXaE zRQyaR$HJd?Y{cXisefDhIvfeU8JL>;BM)#{7nNrTDwl@InQlm)9jv@}BgVr7@6H^~ zE7vuPt4HyT2D_0ju@EP`EOsJZmklHzkRk0rU%Ga>J(1p%0E{TMh8l8&&j;M;t+x-q zBJ9!sPprs5;t)?W?%w)sef$>%eh1RNcx`bg5d=`IkS|YNpsK((;eUV6%FbxA<@G!tW-c#Ir1W}gmMgc&k9gF3ng<8Om&6|_u}VbjCj!RonOltq|I??Kwh;8HGR9h z6Qht0EpE|auJW|Ljyw{`)#X*-$X{vv%uTkKWNo~946+G^G%2f?PD@8d9t8ROmyS6X zHP;WT6nQD$M=1>Ukp)(I04e!!3;})G1`q29BS}BUS_b&%;DV3CdvvVir~_ykwFGLC zaqWN^iWf+cLB_zmgt4vIMSA)|Uq+^40O_MEMn3#CH(k~Gx zYgxXu?pNGt&hbz__>(?_bTqGRKYsOYp9Vq_PP@AqVx6R>&C_OZq}H@v#7TxEPT$*} z4Lh}i_aO7lUwNs=p4e~eH7_40E$2{XDvy?x-b%?TQHh8kEjnmIGThDjis#`lBWW#t z7VO(fq_{TWf7Wi|X7vhWj3QBDtCqZyDK|7RRe5DOk{LzFq|(7@ycyG%+$Vziot|ES z$3>2rs0F{9Biomn${S|ka^GV+$!@I|Kx1kFm{+pqGy01LSt7w;tCzMb_(=bB$oq|% zVKPjQYFz9nP5@*|4xq>O{=7jt+U?8jMBec3mtiH-+AkRL-i1{roc#AAD#Js0AK%#t zq4^27eaxb7w{O47tS^MbQ8`MXCk83OY`<%`emHQ_IwKqfRPd-%R>$k^8cz3U{m(7` z_W@#oI916vm+Ea>_L~y^{YA40>{;?;pD~}+Lp3L_At(7~H0|`}R*qN_pZk1pM5mAU zA@&&QpZHs}gn!`6Iv#OhAMxa%3#*I-@?3O#RW3z|=5geA#|Gxqc8@vva@gfEuq2!i zR~qk9LJIiRH}LXvK}#J7C*_ILvUH-Nz~i+31b2Ld`@8I>Hk@{vR0kA}nw)gz`r#Ht zIcGFzr=ifKZH*?+8_uncr#h}=(VP?lj}y6W;uG`944Q99_WVyZ{`>VrbhL~i20eU0 zd8s?5LHVkl*j3NZCeH4MT)qLB zuX%GP9B|>oYRLTszuIhP5Fo>ILu9A?DFB%nr7gV3g<3dc3 z8Q@~O-)&E3Hz|E0(e`{$0RR$X3F3dc8t{D%3y8hKWuPDUs{PlDkOBwmkAS@NmG%+c zL-7v8+=D`OdlBAX*dZlSP z#jEY-vD2EO;|9MJ!vM?fFHp!u+6h8JvLra2z5~Hh_$~3_z2DOYF6{j>kOR>P_!HbL z&|HH)9FAVix0wbxF8>{*V8CD}Is6?pbPtPWZz;geXYSb(Pd&K2)!$11%sQllr6DV8 z@I>DASDo7+J)(dDzKXCESVBe;BVJCQE$!$9>UG`X0mpBJ!oTgAC#QAQHfe-zxn=$FmSGt( z0>s7W+}>eGtCta?b2z$<*mRmEHJAO}<0p&d!VgoM36t85x%-6XRN7w}sjy!T_5J!- zYHni@M7{D{ePpz!gcQPIvrzZ?nBP-@zf6=T`Qa(FvkGUhB~5x^)P?zmtH#cnG#5HToXO2)hc7#k2KSFW8Ovt0%s~P$E6RabRLb{zdr~oPc$bU>7;bWisqgG|ND-C+fhj)e#mR zE!J%n8*q^G`Q6tu!no$-p9M1-V+3LY@-{QKsH+p@Y26bk1NPIXsZ8a^v1TtYxN~AB z3*_yqe{3XAkQ8yM842Bo>7p(3Ci-2p>f%4bG}LnI%MtxG!ey~l(?|l0dwJH@ z$BdLB$;16Y**7MVoc^G7p|5;WRGgN&J8VNmibj(O#IMc|xWSH}>gQf=V1)N&W zq`Fwp_V7)sa;~(>b`7(3{g$zI9$9z~ZjSMpmo>>gK1tme&OZbK+xOUrH*mWn_^q#k zorl6px|$G{9ZuG?OxiDB?q4rp7ezQ$Q~GT+c#G{Cv&!?{A^tU*&%L!(VT$Nw)BuXr zH%iObg#H60?#zcPb!GdGl7>?024p*`}f2`J=jR&h;>`uPCUz{J+&?jc@Vt2I}>ZwN1=(8Fc`lGvRx|)WbJT-PI ztg>i*>mz3uzN06`yP%5Zlv(tnB-9rpTeQ3wGrmsy!Q1>n*z6E$@t^C^?;0wtHX#U`czHV|;F#;qqpzJ8d?2wzwa_|`Ll?WGAjBrqJ5 zt;G4#K^H;6mu4Wf-c_?R-VKP~buokeLxvzrTwO7ntU3ai(QDYk%Z$t$Fi1s%q)NLI z;-EobO&TAzJCM2$Jl_ocRA$2d9+}~>P`F5msPygRh=_Zx`vbA|=0jLGIbNoOe9u?U zqoLZFz!N$f7Yx@{(#piFn0ATrdf+ySB$y$3vP`xrP&OWp2Ko3TFfo{AiLW5cRqI%M z&ACC%Zs6_Vk+TaXw!=B!U!gpN9LmK4dB4cH0PJgorw7Z<`vE1(4{SMJm*zX*V~Gy?e0tsiVzV+v@Bim#nh>%5lc&7~FNQ zgFG*Kmh15=*Y%j|p0xlM0S}7`7MA*@7T+*BrGSR@&W~`ta<)@@6tmXcFiMG^X+JpN zTo@~uZEx$kUEf>bx8#yl=ie>QZUvAl{}3Bci|$s6!~2v-MNWcr-6{)Hkk+mA20(TsDVsxb$yg$`#&r~+#!2q5t z^$+5%z5G{*Gg5VfinaTY6qdlx@g_*Xiq?KPV@L%fy7H46z2wSU9c@O(DlN9W!yjHxawdjN`)DZG7 zyGmJBzPj`79!nZ5;t8%w?hMFPbWhZhSLbi=KyN0ED_Ux0HsMH`FC}t)ItMQjv&hvz zNisuiy=I3A5zQZk+Z9+M%!QjFfLVCLQ}}n<+UYLZiRc zW?s$hH?4?57iz#Kdbjf#08i^`9ABwR)xhiYH9D~+vZeL1lEN!qlq5ek&4avQ0EtdS z3=3L^qg@FrHPe8^w^_Z_8}Y6z39}HqXN73P!M?ruzG00xNfMnDcF&iadv48iJXxxT z2Qp6aVaVm#Z+5^RkTga5PLMZJi(`=EhjxiHZ&_VqDah%9vf%DoD`flGK;tW>oQC^= zo8Q`&01j66v%LL3>`S_Pme_Zy=&pPARq-pMZH_ROd*%&W4bW9;Y?H1~DaQ+F0N-S? z%KS70pU%~)9XEn>44->LQPS23%3>t?l5{RpgmMf>Wovjpv$s1(*?GDnC3H5=?38Pt zkn2F)X@L@wxBI3l!RrB13LPOJn8{R)Os_jA;YMrxqca-*$JwXiJtY7&>wk`K zUPEGF=IB&d{Y1)u@t10m=I^WiJ`l8~fcaqqWchwy8^e>;q^)G{R^h%*Nnoe>V&XP%c)eZiYT)C}8@#R%fE}{qkBXL{DqyzJ(?Ut?n;j`he6U>ATxiptw8iQI zAH4F!y@v=)v1$?aWT_#8_#yW$PYcHGoJT9l4*-h?GcA z5^ycqwDcRMXvbEpk(31CpqdgIGBX>6@{dIX&9$wQ`Rg^|T?pMC75Rf7WxU-Y{Yu@r zqQ%pfW|s3|TnSk`TgyhsS?@Po0_%`3zAf9VNoX&0?0p84%WLWxo$K9YDWtH!8ME^Q zn*`7y(+!zqkdoA&EI9N~)I(Dk)8C@^UkTvl-BZ}EhfmfdjAXP_!vvr?_SEqxH-kWSS=eSp7 zJEfIKy)x;!y=@w~KhrFQ*>C^~^q8sHrWA8#`GBF9bvE?IvR0lF$^uQ!3_5 zscX?jvehr&_x=!VW;;kHRWTg_a*gIpmYrpU^?sWQ$d>7V#{ulr5l<-yeK-t4DWm=Q zI^zW{dn;A>MAS+K@6HpUl&F3zh&x$Eq(&FmcTmW(9514G zeL|^WzE9MnE0yLjI&7;khMYI(Yi!XnFADW&J4%qyk-tV6_^lx`jU4kly`Fg%AQ6vh z`ELf;JuF8S@NE6JEwQGAoyjeLkiFpOa{+^~opMf#*X-7m^{Mv!zPFPbyiZ3&!Nd}Q z_;cxBzUy-P+WB0A?XrB-v>-aJO+|e<^qs~k>c? zaP5KHk>|t>)~$bj^Ms@=EJ0MlO%LtVb7-i4Db?RRS{``DaMgr>-%cXCASCqNhwc|J zTs_^1S}`AtBRNcxhkq{~_DhBwOi*>A(1!m@xuI8}lsW)by5I|(R~`(f&-W4G+<92! zL!ee!DaUIkjngCwaxF6U;*MW*C-PkbDjcP96j+?nzvYW0J%JV9X0ObqJ2s=_tM zHj=1`F22#Iv-Ac{%ultbyY}TU||ZrF7X*u7j+Ntw60&5OF-^G>`&nRD}!8xLV4V1X*Xxz7cZXaw~>vVf9_>G?QR8U9*A*+-|`2+mBRw!IX z?;6FvFLI2rHu=L!17kKW(PwegQ;%Guo$O#RcHbN8IRnkbaxM)?KRBFy`paF6dBUr# zt(0yb5MO`(+kJcp^$oSYDv`hHIqZu3vX8X$oHol;o;&ypwmxTYVnL4YlIRkV=Q4$_ zIq;fG8UN^h{X(0hGl*g+>gi0O!yugrle75^D&viR(U$6c=M8Rp97k+-?VEkil(Efq zV5ezWICFu$4_6F}r{U}d5UfNI!eG2+EXyTj>$bd}8in((Z17)S_Lqw}5b^~i&H~ZoQ$QF$zgJ|Zl3m1_+XOTt*W()IR~MUyEf@p_p<3>(pMNs zD0`y2$A`sw(WC!3%H|M6M<*_Pl0^~WjU4EQwsF|W(xYDhjwL7i5@ufcXGI4-XA^e`pOkAKojAdN|P_(F%;YS#q6wfs47&;x(W?E!<&%A#H&w1i_ zhh2tG(Xngx3uYgmmm=3v81rm+IwI+v@W{_ma3m<<%Y}7ZPR}Ra8(mQ zFvO(l%Xk%Mr7<{4;GJ+Er%U4m5YVE->P-D?fraQKs-Mk&0tByD0j%IE6a>(rqxM;q zNw>&?CG@c8!WVAoI5p-U?Ocxy?QYi3Tp)Nl_7s9 zb0hG1EKde3obR1ePNsco+&E`>6Yf9&NTsaI-vul7#5shL?!1$_Jii(NohJ`%3mEgz zWFC?r3(H5HJr~U25yjC?6`TV#moe1~ZSB~&QeK|kO$;9N^}tEq4BwP*IO0v$J38^u z%tYiQSHC?hDY#d2ZYlj7p|t5a50aWSB@v<*CG*N5P#%K$IoCD}-G1!_;dkv=_-Z-P zi9D}KG!M&lx=!Fg6Rm>$MWH3U+n(t?!dz_}t!g-_)!x)6x8z#ig zo2JM{X!TU$Pi^g^nMw9Z0nxCT5mRwNp0U{E;KH!ZpVNa?H=bj6_gzEB3GW#Apc+JW z38L8ztk8c1;$|nr=+E|ioPd;BhI=Q&!=O1H8LvHxP=+=#7J^nMU+pysN&f38)B9DM zr&;C5SZePMME=JSAkQh~^Y+0D<(<{B5tjDT)W`lb={f*sNbdAe$Blo+YqHk$hFLdk z9fr1z47AH-Rf!N->F9LECAvD@mpjgplYgiReX8n89FRX*V-bc zFhXq_m1jod>mb7RTJesa@oG4@kakUYKdKB=WdHZ#Cg-cCdH!JZNo14;ra0>@nqi!C zKZL%p2zV~ra-DiWyxOY-Z6GV~oAspqIJ7WFe=E6TaFnj;*%sS5+k>iwNOMiL<+^Wn zO&ilJgoRWfcyNhTYYdMqgAmtozYNK;A6tEbi!stV)Z`+EMGn<|2s0qCvc{Pdz6nH( zos?>73=$P?{~1$#FSsK_Q?0x~aw;`jK)myb=yA7n&!SZrPE8+O_W^UrwKj)!8W?O~@?YUYzp;I+p)#2B4Bfm?wP+U`T7n1p;G>1nviG&SS z`mr+))369ZVSui()6;Z>2WHiIMOExfpSvLhXW#qWzDGP-)MXs_C-s>vHCmHa6ZLoM z6DlQxVZRN4$lSBMx{RiENAUFT=nbp)?d2`E))-f%O?pqi79HY?Ysh$CG$3(0Og#qi zsgW!b!XM8adFl$5ba5Vg^UsF=;5KN09Wn=kKM-2O4oTHqartR^;kG|-{}KNGL+AeT zIH|z`8CwX$lY+-ce=xqvK!JyMVFmZ@?nKkLVWO^&yR71=uGLnq1Y*H2`zQwy(GyX^ z8oQN%THMH|Fn|2;H;T06x^0jc^5kNs6I~7oxmSgggy^Sy8Q<1T%q-m9`)9}Geh&vC zvoRt_qxA&?W39)3m|I>EM}Apl+M#Ie^-t5w%>!Th&YvQ}Ht!=Xjte<=(eQe3!{m83+fRpT8-aVR_NHV zs1q%Bf4z%n#wEI(@vGM_e|%P93c7K}$9{0OpC0iw0;OMpb|WqzjgTAXlQZW|1@qQn z(kKu`b1RI#QD97}t}*ezJecx3m0$&aUsmpm`0rmS*BA?8$Yy78^L4}-y0|eccGO@R zQdvHZyaDK-;56?xjh|uw^qbf?)MhB6nF-2g?ROy|v)b$#>RaxT0TKrK7|85>h(`{g zWfc|&f&=o35J~>j!>XUKQl)Q9WIxpPvO6AVi;9a89vr5}HTdtaJxuaiqA+}M&7jo@ zCu4HGdNV+7=>x2)VG?jO!K#BpGElZ=N{G}Rfh#ajmk=vq1%h85QHpODuLK;2Q);sF zfuB@&c5V-i(C4+%QKD?16S8eO>tKUC)AB0!p{Yy~%Vcf)JW=5nCGmDl*{4dV1=X@V z+AcnTq8=V(HRfR*`%U_Uj9F9}d=5a>QVY>Ya55{?P>&MYwZF~Ho7I(Ty#PXDTc}XK zX0y6VZsvB(W%fMD@0ZDcAGK;;_Fha}@^4|wk`czadQtOfk^8s6R)+X>W%A{>>%6)) z=^yL7XTQp7`8%ZhwK(_2zF66=K5546H}MC4l8G;spAd5BO>PAoy9ELAn5&foaZ~hO&`i2<47xfzb5- z2h^%NCL%oS>ZSX|z5l`Z@}=3A;7DgRvsmzml%CCNOronRDxT^ujz*}FguUOORQjgb zg9Se8Ax_`*xQn2?q?LvX>@K4|kQ$!6DwvbZu?Iv_>KxRuJH?pmWs@3;sv7Udih*ks zkWSEkL%-Xp4ooMj8%SUOvF>>UE2-!5LI;@fh!hz(emO#Wl`jW0wd&bEs?c@C>`NtA z#w>1a<_$egP|*dlT#2d(JnAn=(TXEMzl;#P1 z0QjMeIj6&+h{zO@sI(yPC&lvJH+}i9rl|Mp@bML65miyKiN42p@&K0AsI|DFT#1N~ z;fBTM`t0@Rl9MzXt5Ku7W3{m|h7YE}{t7mW-ze+f=hoHF|4OD%T~EVwtN}?IWB!o6 zJhTE7fhedVWp~KV-3jrm*xYq1AT$IX@O}kZl^Ww}A3k!5d)>af$a_!Q5_8;hrsP`J zaHP;m+LvOaz@;!?p72YX>7mfJwqk^0mS)e`jv-MElw9(d{)3s444b9$%W-B#buv=- zA>CCA_&`qTQSrMZVstG|p7Iz!T@X31>@&)BLNEIL%{vdDM*(1>CR=ke%i5quA~8cx zYcZYnp!rS9d!{KADvXEx2bi0^AbuF=f~o`v@{~kX9us?XK;`r4iok(5Kk6C?BO6dL zB;pYNcU_i%117}J*g+fN81VbL(<6$Yzs(D9i(V)i8kIFT|!w5^3>ckVBDih}# z7Y`iD>H25V@LzqnI}vA9FR1AO+ZN$pYU39{DM0bvcjXo?S=C3}PJs~UL6 zd?C>?2!r>Q;BcV>=YSQ@SFE=|?rlnfss!`&6K8F;846RM`~PGo|9X*M1(m>a*U6rv zo@H|bZ`<9Bs9(W4#bKNFa1T}yeQpAOEpuU;R%~7lE@idH1rG0Nuumg#~OQ7IO< z@$MtMt4bENzjYM2Noc*NdhNli0jJ|NK;u!)&sfy{>Ac>RynV&HCDOmrj$>^deuu3$6DZ`jS!IkPG$B_Pt80@xa%F(U+*n+lLuLViq07G6vgtI2L?Y~KKJ`%q5VG2_bG zp^6E2QZ*GyXZvxSXD!Gw{V6)yeiu&Hg zd_CatpPfr8vr7Bbki0%<6>vNXqKVR?5o#(yPA$ra1r@0>S#uBEb@Y3pNJ0}dCtrgY zpk-PY6bN_Y9lx5_zURa*E!hv2ybJf2n+`rI#Z_R4F}M5r5pQa7sratKDXhe#+WGs) z3AEl8zvBH}BdWi@vV4vAlhT?xz3xp{qR?s88}FMtC4!ORfwd&&A3{tP@ruH7eg%03 zf7|Ru0eF?+8}u%})*m{*bM7pDXq1@C#6eTk_l~CSE%OC$99W(+=rJR{;Vy0t5H6qd zJ%YU0^E~exL|yk@|EPEw9&hU!b?%0o!Dq0{ETxh5tD1g!ZFhEaJ$}Qr1FmnQUA>?^ zLFkAO_L}=}$NG4OE8$OF7kbP7QDHOT%k`40Ow$(e{vaG55(Id*JAXFiE!nEVwvnf!sHxMp=a5Finm;0GNp z?pu2j&Wy|6gS>tsJ-MCi&vJQHJp6gHSP16-tE~e^;*Vk2D5c*0#{1} zZwRwst6|#tt4S@>I+hIQW#!fct55YK;G(?O8|M*WA93|6y0JsikQw^YoW(J2{(Hm! zhf(}QhLyYQR=e>1Px|pU3c6q7NPNu*v#QuvhIS*2tBOM;*QM;6;{)&vhHBEkK{uQy z#gXe3%a(oniuA81_=oQML3e*pd_cz5(lw7JV`f=zM?g-_mB!Y%5@U1CoyE9z&OSvE zHM})1Vo|iL3Xk+OR^~A-h}#`iMKwcyNt;E5S@5{K^vO3eFr~ITYD5c*Z?%2L0f$}F zt>UP3)j%Vfh~ZmDECq}2I{SD74j@sbiO@Dw9i?vMp05>ZT^aJp#i!sLV72> zH_Mw5DT?VgA?6`G!XS!P;lfel9ERzGM&Fr_^2?PO8lry21^bXa@f-up0Sf>UYTxJ7 zTu~|MoazY!vDs2x?TF@yQhEbjLu?Jrn2~xRrW3*2H{Z_;D62*ARd{0Ggr28r%Pqkd zO|7h&!SfuAc)KLS&MhhUZ{lQ(UG-RhUbbkXE3|z`tw(gD)9GD&Q8wW(ff35&A(lRh z1xo{-hQ~5`!O!gY?cNrqbkuReQn==TrIn{uwCEf8Cxg|tI)H1jYQXcUxYy|uFIF>J z#_Q*5C0}7G;dYut{D$}oCR%JXP*Mb83$C^vF&lqU0F8A6MfG~QgL4fcs~$6I|h6*wZ&q;~ri{A$jNIwREmJL@>G#NTCTUfzglQ`_AsBK-F&iRt< zJ3ZNyy997D_Ci4?@kuj@5Vs1!7|KXtJPkGb2qb&lf-CQakZN1?e+kB)8s?%s{EggW zV%sJUdn;`*V?EqR6M{z@fIxifSqRDk3jjOK3mqVv=#EuvqM`9g+8m1%_FAZeYYM3L zS(F1`cD*Xbp3Zc4A8nfasCXKFRgiq~U{_vMOLk(SysS?(00G-?5h4J7Ps&~z8r)zrL@I_ zWxdePe=i9hdVS6-;V%kw&y}eu{Ei3wWlqywj<4@nb0G;kgvca-o-=r{W(w-Dmq5#v zh!ts43k;J>N)mN+m1(P&M^EI94KPMr;!$x#El{u{d1+52vD!XqYgvCY7y1QD)up&T zvcjsX+y1#j()A&Z_papB>WckNp$XsnImB6~**w7lSH{Bg=Jppv2p{>K1?==^?{kNS z=MAx?cHhrQn_YoTBZLmcnn_`B1xk`a$OFfcw~jT{U3AqV>O<2}d1X6h z4V>8v#vKaZ6%|6EUT>?To<)&SsS!89B=u6GzHz#oB`YI%Z~0Aic-VJ6s+fXp1oq7* znL(yLkKfdDn>xF6#B_Oh#qgN?Z$4m&2g8`ZM=>;(Ilm%vj$^^{=wcZC{se`i41?oi zhy1e~siWj5rCK8L>CtG^Qn7F=(<2{Ly0_bN8;;rF_|#pig$>584(TiT)Iv+LBV9@t zG|vuSnb6X1ImAv%{oh)p|1qqyK>|n2L$d`M)`9;r6$osaIN9+!xUnR%;u z(da@gHbRf3B7R_~o4n4~p!soU*YkL&XAzeYk2m*3cFb*#73qN%^u3jhDS4CSmRO30lj6X0c`tKK4TE7k@(YzHHwznA55VbNSm;(?Qjc zcEu{9>)&lU)o$?fnMeM`LysDEEiaVy2m_`!(2#pRRE5cEoWJi{K&Wk5-OC#~?VUaJ zh%P+5!E@O^Wx8KZXnhyK+~9U#=^WK0>Rt`9KoXJqwlHR=n z56?W2#SiU3j8!#_cGdLbiaMMgtdi3+&420IU;qE^pjyO^uJu+&HrgUz6@lx@kL=~e zHD{1#Ums67iIX0sns%{M>l>`h6;Ya3L=B>rl!22=^sH3`o!-K1Zwq76AbFLaV|YjU ztW+=!ogFiI3kJFh6h0D>7myR)3RcgXX+^-Wo!s2F*jhjMeZ)2;(h#8i25gq6HIpO{Y_Wrn7)#mUw>WQ+#L%8q%wVhg` z0)0XMkFK|jihAMRh6M=$0cq)$6a=KDL_%7+hLrB^7ErpoySs))nxVUf?(P`s8Bg7w z^MBtjvlg&me$2jhUAw4G6YN+lidu%oQjdnwj3wW!K6VBY?Ovi$=x!5oqWI5?0F~$b zYwmpY_wPe1TDbUOZkwCAPDOXPuOol6;L!hcQ5{YZyE!sABLKFnI23x#@Y(Qc*Y{X| z%}+tl{c3_PLv+Jgj!=a|B}V9vl|4?UlEW79qe`I4|HLgmN!vszdQ>5zZR>ffifZ<1 zepPcN&V%#$ZE!8~z$q1cya2XOr1AKSMerMT?3=0J_JVV*$g^DjF_gmPBfrV~6LcOk zh-Q~f+Bb!S9Wkdi^8U+rJ~MG#90VlTS3r1fssW0zD7v%kd#a>ni*jCWDE)-3-Is#3Ft8T0Vr>@v*0%GfRCXa4beT~hhd zZ)i{m*Lyg` zE8gu=L}HefzE-V>JHv&*CgS|hPLN~%C3s@45yr?|yzaOQAel{(wr#~l#tzGOJu?PT zIKe0jxl##e|E0SqzYc7QBkFL;q5ehg52V(I;TCg zF=PWDe3ZE^D$cJB2F`~@0N%jSOSgUzB5@b%w~I;L*56)b3L`tyh|uyngf-$BtS?eBrk$ z&R*E+M?sy!U{CSPg2<2d-)p{AjJc`%5Rr1SLM*7bk@=_3JPRV;%osUM zLrjqgi$qA%a6UO-#mxk~eW%BxP(*S*Uz@dGL{%|au^dDvlj}NfBV;~)zZgJ;D9Q6Q zz>7c6^^s08jEgpa%Eq`$Q**!D<#n&wPBBeINeOesg+8~#A-`C%{YmJuyQD0LLWN~? zd-6lhHvhiBUBAot$y1t>XW$QqST;P}!x=BM+zTX@#Ops~EG?EXW(1a=u{nD6rvcN% zgSGdHg?vhX9LE86Y^vb>IvAWwk=~dao_?dqd=biXY-i_ud|EmZ3Le*-CcJN9Vv1uX zEp7?aEDtN?nxO-#$}qlXs(?#BPn4iuUI8V{dAKt$vDf$i(~9*QCNh{{_VKnF@o(UM zGoO~9V39gHSf}cSDOwN^@#2UWXBg^MC@QFIHp4w3c3gHxM1%QY_=No_7rqgLnQ1cU zri?IR#&Q}*%iw>TMom5+hs8sJ!pUX4ZyWkMA9g5if&{)2qG5}0 z(Z^sdXa1-=zq_;|q9V=kfwrUwf#;TeZ{YRo3NU#m?N#$QnNS{eSM@;LxeSIfpKelY zdR>}v>8r4DsCY{#^RmzP{h6Gpl7+|-;8i3svm(uM^1E?ek=6E(Ae!Im(iAh4b?By6 z_U(gsb0;F4Yo0fXp{|l~yGQXB;!xq{a$|dGmsSVHZkyI`yejg4iQp#9pL2Qv9sS7I zh{IFklh74Vm%QSs zN=e~D)qdIy&ycGC(1pIdBJwYq9se?E^X^LoKsI%zQJzWgabr6}UehLK{g22_kPTyW zwxcSM3Gv$hm#9W#;9#BGq9FF9eas)ThRd=hRxXx9g>2Lw+Z+_oHX9xGc9##~!(GZv z*j-7|1Kp1$@k)i4kfFuo&=I|yZDLN}_A1_0d*@WmlrATQ)fUs|na(M&lJxTZ%$#yj z)xRO_clv7RlibTry6kO0oX@jq?29y zfhIRv`NeTL}ZY9V4oZsUU&XOaF9J!RPEXl>#0S z2^^>5$UkW_G?{(7gN`0|;%LH>`on%GJxo4h;sP)u@Ep3#kyLlb?Z?{iZXct0w{gC3 zL+=)%>}Co;giqneR76f!V@3rPUi+`MT)~cIZ0M82gklrSX{2b7$9bNFz1 ztjxQl!guHPoE^0d@{_UoYFtwqQku5Xd)InLg$JsR=H_BMa%WMgw%H5fhL4+U`*EQt z3|73a9Z93_hF%UI*gRGu+20vnp)a-&v*KI(Pjd0p8mX@2e=PT@oB);n)~a4_k==?EldoB{T^2mRY_ zEq;W{a-tz0UA~btdLa$)*);g(Nui7bM(S6KTx9tC-ih5t&BtJW^Z2}PF8WiNEu5%% zp!BUKW$)6rw^`Zu-S!v|X%Tfy|8Lj?jUrl~;SkL3KEbMhS*m43TG76awll*oaR}7# zk0FtCq{b9c*}Qo>3NCq(?Ts}k5_mT{#PSI@&Jga4IGt8vLGZK50k_hHm8*9(O%2WJ zm#(9_r>_cW_rIav8FsgO-X;3SzOfA4NlF;=n>kj`T4#GWIq7X&U80e9WhY5F&%w`m zR9~gG$;Qs6tx%Mj6vBSuz?A66)n|x>>uTmehkUp@JzxI9M5rf``Ex}AcJ_A{XN-9I7uTN(dP7Tkfly{$@9WpVNKKi*P<#v831kacqkraEqFZGlcm& zwmT-mWV5_pNfQpGV@$n1+;tN?UfVEX5rtcMl?dCKv;54FIxKxMAYK(kXgIiQxhbro z7UNo?cwlsimWp$%blg;Iu}BE#XC9f&D`LK5qN?&l-=7+7be_4=C`gR!G6;HrnMQ%e z;z#!7fTK8Gv`IV}$e z0plbQc67GnH09&aRHLf0#hW37794vvYU)CgtD7M>59k^(Z=+pnc9TVE9g~f&w$jE? zV>Y`?^?*Mq4yZp+NL_9kn&dn}`BPHwX?!*idB`2TG^Drbox|VVjl`3kgY1WZ8tkj9 z;r5*z@$sAP=T^*umQ$WSL!|>^MAv1HWEVBu+OI`mK~d+*a*S1*cYsqf>k8;xhca@! zQ$@`&YrXr;;y+p7zcKR+iA%-$9p-P0@!}`AwaV9U&u6wS$kxbwHV^Kb6BM1iY)uJ< zQ+SW9%dpwf`Q1nyEeUa=lMKBu{4=qEnyn^~F5*9Zn$H$xMf0H5mSt=4Prg44nK=c_ zq$E7Li0iaGvD@FmNLV#R!&f>l{B+WhBB*^ zVaDG^;Zy!Vpc}}PxNKI_>*Ro*0~=Z@Dm{uk5OZHu9{j28V?{qT2J$zqh4IDTyL}R1OUviS(XI5R zP)%>1oNDllVQJ%n>1db`Q7HF*(t~A!%QM>2F?B^~Y~;y!YvRtKS!E%P$p_*UaKuo= zp1Ztu@Eg&%OSH(z#N-*$B1r!xWeC47WK?_-3seoZFrU2PEF!72;#}wFRb2{EiS7L( zOPqCwtwhuiGCOU*zcvL!ww^4IgKnNqZYXPyLX(%Px6D?dL6l$H<9T#Njm~oRkj&N! zA#K(D4keZ+-*pZ_ZFh0jZ+hR^ilk^BAcy%ws&&u%dfkib*%X6vl*xT^5~D+ zV8#!mQ9X^n`)|8n9E`PB_crtvm1O8x;XHk1$fzlh`%#zyHS^!Z#g*xvc)tX#W>w;2 zv$;yjHNrfh(!r%E5sf!xPO4=VVRsQMB zw*%3MiEMIsi;bR_4ay>N!9K@HdPrKSQILyJPwcr}FI4PRq{l`@1bu{uw+WEcg%N2^ zX75#Zuf0E7*z^lLI#a%*I`nlQK-5)$DQ1245-Mrdfs` zTd6F8hfM&evc{~0k#d|bSsF8h5gUY{tMEaRrth$7UlSn+k;GdPN70KOI%e^MmPZvA z8(Xz8Wg;56lXgjJuSwmJFvd53WeA7#epik8D+u|1s^DI#_YqAsX^f0Zu3S&X2==BH zK-hSH+kgh~B)(FCL*$ zDrtl6xFB(Ra?`D7tIyms+QzL|7Vt*F>&^6DE?2Ipr(R8KI`(Zk=Fy=pZ2}jZT7Ba4 zx^vR@KgVM#ThQL$coy{8>zIwS9?h!nEnoIO-JsU&a1`~u_K}+hGrgIdIj zck6f!tmK}yQx%qsOiCVr?+tE#3D}`NsN9k7BXwJiB|gvRk+J{3?!2vE-FaI;+(oQ^ zfUo0+wCHeWL8wj%U5*K#d)LgLAj_;MTTW9}!iWpSZdbw*T#|p z*^J`gWp%A|wDiAHki_=!_76F-BZ5`TNF7YqZVW&vHezN6-6z@1kCP!+&xu#|Pc#Bm zS^#cbuoK$&c5jbE?=sHmA6rt7`Z}Q#+$D@0uLF(aGrtmsmkQdHDVNT*U_Pd z;LRD%DEPuZ#%}Gioq!Zt+zvfAdKXc-P$6+_On|k`@gZ>pW|>GwdLZ{QE8KY9g5)cZ z8#@lq=!0@%Hc%pq%}uCYp2$bpc7kOgI}|aMZaLoAN4qfHe?1z@7|av?R{wbNJ#DK9 zt!eSAPMcUi0;xTPB}cg0qAw_SC4lC%4N(^ffj2g}ZnL}9Y^9^_RaBANm2GY*+2(a;%qlnr8i9R?lgO#iHIpRk;@mlX?XFfcJ z76pD<>ef4S^48c6&h0q^##+R^w`KlkrrEj{&No|k=iXjmFr!bt)k)LC4gXF2u<$}) zV~6n3WxaLCQSi@f(>qN52X1Nv@TQP!Z>gCUgSk6rexGu5h^62WxAw*Pgdw5A(@b zcvaRmbQpX(mFE7=_NWRmV|ct31jz=(LtPFtydDLtZz?{$q##?y~q5 zhghn?GFCJjF!Q0O{F8q)$iuE$_tlDjzMNS0~}!*f6=H|`3&Lsr1{wb z0YUiH{2XkJsuCIw9;KG_D>F%*U9)h@E zcCg;b;{4@X@+V^bS5sL*yN|omiFf>?nJA3x*zAxBf!m0o^l!0`wU{r>!`pN9;OR*e zbkroggL+F_dbXE}W`fgGoj>`t!`r7)y?t$nuAACPhn@rNH+4lL=C5Ehu8UEi>&>X0 zLSb-Dwe$66td0l-tfHy%_~_GJs7bfmJ%>5TjKHecRax+`b|<=KaL(9~kfhXx?@saO zn_`u)q1n3S*sAnTor;Epz13vEQR*wL<(56iz>0XGx;BuyAhxp9iI`cRuYgDNxM9@h zUA^7wO+@qbnBIz(4!H@;jZ$%RzCTnrLD#;^J5t~6G>Q)AJJXu2)yI~g#s1*9Ms@yTd3xU}5X|HHq)S)p3Ao4ZKVmKZ=0+Beklg zdI0H4+^5WI;#N>80bueL$n1&J!<2b~zi!$qyEYnH5D1AFTDwLKSU&c9e&@Joefhiq54WAjdv1qlF*X%+2gF7bidNU6F04v7}De?2F z7VLLdX}$9AJm?V1pC((KGWosMp>cH)%`*$|uv?}f>4@Rw@lp#$V0+X&x9$7qZI?@1 z_g*S0gSYglv5Uw#DtP21K>*yywQwP7UX3A@{u%!5!nF5JF{t*n@%vl)|Jdo@2_UJJ z0i=_1MwQ_Z$^S(y^|U^5FPN_Eetcp?M#5H`4(@p5Z#X_`(y=$LcmvvX8kx%Z^7Eki zFkm-HHySfOE>+-abYYGA-we<%^_r1rH$#4?*_rx*P+t3B_O=}9=jh7!yO!0dI*gVX z>$_->+i2~v=}!hOA=Sk~64~ZDe(mkh+8s7%bBCgk4nAMXw^@#L195O4%VV13t6q2Z zdnT1U29xjKysm$67`OP|61{NCh%OWI^PTQsB0V6r@fw@hgg67wNuYy!IrIiE=x|11 zYwbFVDE%q=d6y=NND-9HIng9ej7Iga)=!s@{CBcW?*&1guexKKMh&W75+E_k3cI;sMgJh-Qm>R;9>)mkoe7Nk3}&YR+fbkDxaX+-sOsIq3o8bx|}aotYC4-tBe zzt@XbeXDb~^U589MH-;kbn3!$I*i}I#qT0rqjhsSVsDe{OIjQ5af8q0e&PtgPL%5M z`4lzs;kH1_t5Wq_yr*l9rG}W+ZLP58?akdX3V@gndS|GSHWG4pNwDm}{J3FIuzhiJ z)z7A6vsoMBmA#hI*9$G}ZPep^Ut?*aIf=I4M9rlnET%JIt9q(7qhmG@;7A{7YX z$dexX_Z`;Cgye3*Ni#_k5VMvDIPCqF(Rh0Z;3jzE!$CUwz^a z==rzWXi#OlEt`uS_}`XzMVUM!(+qI$ahQ?yY)|y1?4sk+%P>(6`xjKU_8JYYwU+5i zG1|xJ0cZ(bVB{&0S+Ou*PWeyAX7Fke1jO%gYtUNTo@ZSHgOxaKLjbfg_kDh!8!T=H zy277llXHE0uzG*=NvxMw`qVhvfM7_BsRWyt*79XGcQ$5$$29r0!&7uXu-FUuNWIZh zGURu>ch9YlwYK%-+bkfoxMN^#tepgCJCOJ8XkwSk{1V=?hlLIioy4Cb;(hD)M+u3; zLNc~DMCO&-O>HzUS{|m^UK^=Vw=Iv+JdiP?Fm}Vdb`LXPcP3Z)Hd$a-QX6Z7jP>kh z$4=G{5lqrehS!bGcm9ceSEan=3%G5Z7}H2bW@xATsKKKA2$ycg26{vN9S>^(O*9rY zP;Nj-uYbAOMN0OK?v{Rdv)C{!AeGK}+3$L1cjJW%5ts%1k!a*kELGwl_gpJe9bgMI zTs|3!QwDzdXKNgx2-)qYz(-8NddOBQ3yh@L7FY%>Z!(h6H_pJf-Qa8p-+FmlsS(ua z>epgafBvK!%s_2ls|hIYi%Z>yQUzhm@8A5Yo$yy;AWCXFR~YDT2K-%&`=6=$C1L0! z^RRZT*@zwTKjadWfbte3Tj9hlLjL?p>{k&zWtqCb@xd6nJVJ{-cy&9o(z)Ro|}F4`rDEKNrEE!V0Ur zDKX2Dwfxy$>~MEchksqVpbjK+48Lc z5!X&OR%dZIr+yqI^9m^L2iSDuSO~(Hh1q|^zsd~G4f5^1r0}uZ9ew!x;;ZD2j0p-l zbWa0tJLB|cVYqu0Voj6>(*ea!E=Qm4`!JM<^>gkQnRn8Q#oldGuihaRIQP?iRHC=p z5j{1KD`M7fGzxkkc@N(m`M?4&UzpS9sZstWQrzt+H$YdN0M@Q;z{=1=Dg5F{uZ$`H z7+)!_)}{Tnm5Esm;$}UC!Fq>uVW0gvbGZ6v>yoAMkJrJ2%g*~;8z?eHv!did6kU8x z20j%4P#T~Ih>TpM?e_5Y?i)}{?=z;f{J~r;_;SlLb9*dn@wqK2cMN;U z?!%qi13TE|FqiwLyA7YYvw+(rn0m3k{)$?OnY7g7RE$k$H^^b~-piQ+oX{?RDgg@; zC1O@@&`kQ2c10v!M0)@`KC%15={&?8F#I93c2pDi zQ5U_&ak}+oA<%VBvxVG9VXC~tvY?$dh!Om)*fyN>MjUT+e&ksY1)V#s^@ZN;0ts7J z5Peb}>bHaRl&qhvs`2HN&ycW{AZG?IHY6HnhylNFn4a%^yG3V?6$T?N=}Mme z(<|*|d;NB&9IJnlkk85A!ONC!(M!5Eu%@1va#-*2h&YkN-^4fm@N@bJ4XI@2YO zZ2L!M`ZO9i9=WS~p;U5dUM(uf8=uR`x zlI*-tVHWyq8;^IscyQ;2*1g)1&ubcR>sPMnJ7G1~cKi5r0CusaoJn1|CAEa567s%Q zE4fpSm5BTLSy?Uc@M`RgdhW=wY2j>c#@_r`#=NU*@Uar3y1KHmx{`$ji|uIl=yc5S zx(u;PycyZSB*YaLMb;}v4}#gIf&+k7S}bOH_}%M7!oBan5bpEyolSgP0%%e z##dNouiI+&SqhxDy5XAjoe=AMmF2|9$bR#@Qwcrmf{>`gOb$>40%d31gFJ8DGf;xV zP|K{#9+q{b>E0{BXE1QTzhp?k-YRXfqI!cUyr03oAcMXTe$FTcHx`V)8a6Z?wJSj1 zW>akdFacK!cJo^NKTj7VyIdVDnk3rVyQ(j>o>-Pyx{o)`I_?dsabH5_7Efyz6Bu7q zq@>fd+rDktCH(#EIKsS(hu=Q3BUel!cP#96?~Wh}X<2yZ9!VM@O$uzSp{cQ{yCzMY z`j&WaT@e~Al*8_q30NWzEYLn@KIZ~o-kOIF7u_EN3U#L@T{Z$**w@mBm*9%Eor>F) z!v{o7Vw#To+R!iQi;XMSOs))l=|lB(9XxfsI3~n|M***N@Ud0s-KI_Z$VnBXcu zxy(Q&E91`dWe_kNDqC0CHu9ypL^4U|S{Jj;&ucC(6pjXZh-n&BO4WU5o_qV`CM)=O z*;IR$4l&z2RJ%M5Y8;!U%^qXXk5+*TLbi8s#4hEeAL^w+Zbf`Q&v-?kqQ)G$(s4D7 zYttn8#9LBr5Fk8v%oq$X(_{5ocTx$wbQ}4}Vy-a98H)MiQ|Qp(evS#r)~y=B+k=E~ zyfBkm$XI+*YBC7Wn%9%YP3L5dWf9R~i`9}S1rPmD3*Pw@5yP0YcqKare^@j;N-cQl zQ7@+VaQk#zTCrMwYkkyMFMU-!Lg~Mmzq}Xw17lj-A$jVNLKg(DwSYB>dc zk9ilsyEawKG7xhL*!tr84q-x2?1S66Y#3))+vyc=Po9+2hSd=py*96#3^YU1KPc-j z_B?BD6#{8)_3M~y>HG^Obrpg)enV@^fP1?o(+aMZVX27Rv3N;?!RC4Sr%2qwc@yH| zVR>V;oQ0Qj47DRN4%u5@a8e}H61|6QO|c)E>QihN0ts?0IS6&|REOXh83N^-qujY>?B~O{?;K(G|TFcUE zqx^n{p9ruLWUo{|@LJdII=^o`BLwAVO>l4LmgN!>SZr-=QAbK`4Nnn{`L`AoDLy;H zoI`)^bAO$s0Kd*L(%F!<`7tlGd#g?TVF!dSRN$&1aAB$Gu?-n0WIOBnvD|0J+{$QG ze2{4<3AW6}iw9jSmxl7=_k{n_U^yFD%S-&M+wl6&I^}$-Of0w{IJM!FABu|&GiHM6 z_onslap5v#q`O2A)hsV+@2qoz7E=N+2LZq=93a@os4EBP$FeMTDaSbab?If6E zyT?10En_d1*l_=4JD{fICTIBgFSGH2tOC|2$#fsJs0VAs0WZ`48EE**ajxFYs9$T} zx3SYa^%Gt&m1RC81;7_yz*x+}Kpe4+SfEg-Xp??GWG-11|SR$^W!Ne<7~| z1HSoCiC(bMK4QjnknU!w!kIuZB6udZ%;8ddnfE!94GxmHUEtaJbuA;wvl@c)%+VoO_tdH+Da$Uy=b~Q=yK)pqMqXz;?)D`>Y>ow=igfPoXUrz zTMWewDlQoBnH(A640z<{Jc&(kbb!{KCeo#shMMR8=##*lt*rcwVo^qJ>Hh5-#3Sx+Y2;)8Wc*OdG|E6Cl+pE$dN80iKz70_OxlW z*^ACcA1-i7=yDMs4sv=t4Ll=#n&4zz;JLA9!_H@WdOAr(b2^>{7Vy0TZ-oC(r+Z<52f9%7h~@4-8||brOGz@9rv9)7mI;3 zm}m3;b3yvX=yTPL*Rn5}xQBnfjDvi$cmLt7-+RU%%*ZU>~Jd)wc{;YF+~3x#nWp!WgsD5|ojsh5V`D`H_B9y-)rt~uAMW@3iMxeS|K zlUgv9XF|HL63yen{B@vEHCG<8mj{;|61@66^}~z&rkX?1ott69;cE}(tieNTPlK{# zA7ws`VufuMfT++hTRlnXDrRX`2;*4o^Q{;8@YiatC}QAxk0=k8Z(iqU6=y5Yx|_k* zMRMK!mJU%nEsQ?s#Q2pL=iqx7FaGwHcXDHkWs`N>;{l!>+X(B!JNaZ)Faj4zBWnHvOST}yEBYL>xp>UcyjrEk{%+&CZI&tAvwoR%Dn3R5 zkgZ_}5%kOTWH?-`8DVaNq_VG^B+gJ&wVm%hI;t@=>o20By_=n+&BACTvu5`TCFr0z&ZW3?e#pKQ?#RwX$cLK=?Y;-? z8}qGM(TBiuf-165dK%(CEUzM`oXwpxA`%?JxnZefP_rcKNALq`34?kK? zuI(RB@|N~ynQbqeLsMMem?m!P3lL$~3%pslvKSg!gSkAl5ersWA|Bvh5njKS@u<7^ zh7>5{6{IX&OC9CxVGixZwkWbVsBZ~=pR;D_;j_CnGfPU{_Brk zsS3bMNSs0d7pCPgQh8L7;#yO?JjvL+BmgMtTC|tK!kCht4>Z)-Hh2huo#~ zh9ErOS+2s8g){!C)uPZmyjyB$?u`R(Pi_z}^wxFC=f16Xiz643%no4C;Zdodc@WK0 ztf(4z;qvxeyVUa{;iB`T_2za&TmoJa59Py$2Cwj%y7&EFk3F1h`y*DVUYuC=h5MYQ z&?T99+7{$~n^OZeB+w7Cq8ncEW6wrigVepb>RRV?o1{FR{#p_S(t~Gk1A)YoMGZc@ zz*VnqV&Xw;y;6XvoeQKl`mEb)f=u@^(U>9bn)I^2sg0ys>LJuK-2I>>S71VopnO6` zSh^)d^n5)fM7l4^i=i8U>D9~APGBXG68sfOmS>#Rn9rZ@|6qwWyFCE8VzI?65ZPFj zhfB~cEF`ozKT2z0^k|`4;iE$bE32K!U%xdg1xrhH$Uz&{ENb>Vk{*Le)L)MVAy=e?sHW!>-KB?bTHUag#mleT z6xn?FkMrY*^EM-IAJ-osXs9%7H&&4pL{!=f$tz5t)3d=(RdFzSUge&8+g|c>;)sug zzNG4)$&X~N%#!vf>GSvcIRdEzl@6ha&kpf*O!R&R>mT!^o!Y5Aadh#eH+tQ`AC*Bx zsky!5nVvJm{YU0rJpxL2$9$@k0&cn=Qwk6Fg5%018?R{S9nYk;#5nD~@4AJ5t!^+B zD5Iac7G#QaRJfAOmej&{hqCiB+94QDb$D=iD`{q0al05D#3{=4X}mHZNb@*)X% zrY6nAz#9BS%HUM1$;5Ic1^uCs>}?hT|W41=;@>`Y87r3bK=#upbg+AwAg9+aWalBaP^H^%UVRm zRrJHsct>}?!Pji&ohZQnfsv}P4=Rq1%LJ+zw*RAAbIw$>_|uifQ-Vnj9-_eIX*i=N ztk^&K6wZqsLXUtl%`fQI1sWY{S9zB88U#Q0kP_DQfqr8<_Ldq}aDGDg=?N8+BH3fq z&l&R-1G);4K`PZ0tgZ9r07^m|XXwgAFR~SB1m-JV>8BE~y?-FY|M{BZ&X? zcsdNo&U;nq>+W!>mfiKs97axlvY7hQjBItH{y6mKCPR@th4ach1(hc;dUqr^I*H8* zufm7(OzFwnps-cuoXIOzDv*nP&kV=L!v+(V1Hs_KNAH%)uo72vnAF`}L zw?!(6)s&-vY}heYsJBqdp^h(~l4`f`j{1q7Lu>a4IYc9WDB{mi{N(&_sVmD(>qU*tFd*oY;1x1ei`K@)*t+!sS)O#Ft zn!9`gZ^Pa5W%(a-hc&^2wHodNrF`rvK`1-A~hQncRSJUZ5goKeO2B-;g!1E zTF8=oNQtGhJIriys3H_bQmav0Y;T>*(n>!uG}u_yBs8P}p1nKO7(ySm{lQi#OLX3= zR5x%pInuJ0<^*YbU(00l{k<1PeckUe` zd6FNAWU8Jf^GZ2pJloS<$pWFr(9lq^_qt*s6YHg@z}9&P)}T+;3@bFH>U}+9t3^oo z6KJR+dXD}Rq+}U^YSiJ@Bor|8?hp0jNECb!)1b0fQ5aK`{@^Ln6?QAeXMf z|8Qx!aG}M`LNW5>y3Rk_!I6oh*nG@$$LHSW0){+W!?CY_BfX&HoeQK=_>7;R3UQiQ zt>phoO&pxiRB4evwJrZw-1OV$tQh9)fKzJQ<1j@z2EDU$IjfL(j()H}-5W;KxYx>s z@psHZ`$kS98gPR&1s_bT^#!2d$KGCbvH`KC#Tzvt3N9cMv)kQmV(g^M1{OiHfwSRm zN#Mhm*bdu5R(7Qi#%zVK{0@S-UqV)7h@J7ndpEo6MkFaB-=m!e{U;RehowJEAHLeQ zplq*-TF$#9NLDeZm`pSr_~K4)mzyvCW@;K@E||-b*XS=OlfO*usen819IEO}kASOW zyHLB(8G;Fy#-q`t1HWqi+tvyqwHKpyCh#b|W`K#? ze>;5nhfkG2g0m`(?{VYp`f}pbU`guXARo{=L;0iMxp0CL@ZeA%uzAy~VSWy|6wr=4 z&&NY=uTHH&qHj!UwiWlNW5rADuyxL|^`x2aoPH+BH(iWep-JJF>xcL?!3%X(^grsH zaR9ri#;x6HA^-vP8hG-*QBvZ+-QXit>tq;X|0{4bSAtnb$ZJ%3QDLr0(JsGSg01;t z=Y2CJJU-*H>o(wa3QM-XDc+S;S|;72(>Tc`H8C|dH_gxrPDWiub9T&217Bo0CN~(Z zEG66kuaSb)djQb6RhHRAN6Ku{7H_6VUIxuV7}LX#P%en1Dm5t`@$uj0(( zl-32`MG&Vmtl)ZaqkXsofw8B%~yqegpX+GV$9c(ZmaFQGJ~m;%O5o8V#X1QAN#yBi|6rk;({W!(XVM+Fy-xXS z7|%a-_(NT89U?-Yoixw<0q4Xvn(|G6LjG3npNxbYIjqkE`}kS6#hS(0oap;yUIyXi z!$_dyZOKxdXmZKgk=w#XyW7V)HNJZRR|9kBCFQm`h9#%I$BYys*=>jYZ~sb*opd5Xk05qPbwlTTbHW`J`ODY0gC| z{C9szHU^MUS0_5JuHJRnc$NIhtf)AbM-F0z`+$9yfn4<-&kjK{wbWL}nAYxQ4o^GE zD9w$f-ip6BrTa|YD*b-1KYueb{Ym0iAj!(*E<&L&U6Qb6=TIk+{Te2YXJ@>CrhR;L z^}orqPbg(Zbs~pnUJK*%tD=Kgt=R6W3j3O^=Zs)_{~ROo*R!h*;(`x5w$|isxx70V zMpNS;%}|uXOW-3g25ujhf_3HD6G=DJsY`L$p(+E0V0+o zG#Y71oSW)!>J1}dB^@w7UTBgm!mH9vb6J1;x$xPjUiXd|DVz3!aWScQz4m1;sCLi` zN`b4f@RzG45f!f&afg9O`P?rt-S z9HUiNTh?WVdW6KpkBh4Tm?Hx`bqr)rkCMb^0*^DqXUA=eX#=m#HLK#s z1^UZ=g^|{T!g915gQr$-9Y^C@58)Y<7G-@hI}|fQdguuEqJLUOCyf&Gx^19RD9l=y zE_C&{ALvuva@H4CMLS$ps5ulZc+dLX@>?aV=&w-8r<8I<(QA@(X@SY19%lqW0MeHs z#m`qoE$9vt?|xLh~ybP(+huS z{#obxbk?4cP_2qvH^xa4OJ6zbel*(k_3Cp~rL$&8yCf{5-3RmkOUtsqn4H;)FD+`B zPL_6ZRQ#OJunhFNKARS3oIg`zX_A^_Jm=f2SBw(aH65rQA=O3kiAB25BI$#>A@*8! z%Pr29X1dghqe=Iaff7Y*#J^ahdG+LAN64TZ^T6g~=3r%hnu``Z6G^OGDm!5q|BVlC zI7q>zW{aa=(_+u{VV~is+8F7c=#R-UB3{5G^-L&Ans~ z+g_SdA&)rGWb^GSO7^pz;OUa%k(9KkYZBF^?dIAoWAxeqjptU|OlEZ4J81QqG9PdA|V5qsvGk+ zmoJ{%t-Ub$8mB&VI*Rv~M=2@uo>91$t)efN6`hh+>HqdXxBq;v|8%0d8btEFwQ|Qa zc-T`-0}u`DA9&F!Ef%@JR;o6i zA#{KLAK$>yEGQ_#JzU^jId2*dIU;}J%r)Kvv~8a3T2Mb*ZaP@r)R^Yh zI_^QOqTdvJfFPM4D+>k0LpFx$nZJ74FDBp48S=b1g9@Q}!@H-q8l7tb^}+?>)}jY3 z$z2`92;2|SY~G=NU_5pH=n2^=XA{^8b7)_I(v2bZb@8og+Sd0yXjA>^9+%;pVgk2nZ$=dm zYxmN(z3zexntQ@Vwqs@QACEtaX>d>$d}$lgl7w%ytZ{j#^@oKRVvw_|ky)hIe}K{l zQ7LtXkTib6$QxBgtgpLDJ^{%H87xz`e!d>485>Amxc&~#w1%kZnYI{Z)Nc&%!C=L2+UnO+Kd>jr0UYJA3_ zpXQYRsinwy`rH@U-a@a>QI5*$d3=Fa-5W2kY+b6=Ur@U4{_%JA(C&)z&DUYgo{KuR z4ncl;@++S1y;7@GyBUE44UNJcQX;$vqEONuBUj1-yQN`>%r=`t zuMz@Cl`)w^&-IOVxvcBmPm$fD9jbO#nmz<5&?>m$2Rkm4?E(go|M7DfM6gp!HG_jc zZd#B)&r_p@@2@EemGd`e<=3Iq&00uq&Ku*y>PZ(1i*7cQI8#|ZSVNz?wi7j%ok(p% zUxw|iExV4`DVR1Q`9y}}+Z{wnL!Zfq#S-7wZp_z4!)6vH1788}udSB&q0!#`n2`WA zwAxkj5)IH#JU;hwVe0r)%hbsu>JL>iEsSmhvVrn_~vQZ zJSoZId|0-&fDFXV{owuxQ`a~qOkZtK(7^z{JMeL>li?8RKAf3Shh@BsN%MG1;dS8? zP&HCkS~kRiklJlK_Dr&z-7oiuEHwc$P}Pq;G;FF`@vjH?&5p|n!6s9*S6!cZOGz78 z9V`R0n@XD5XDXhl&?{`#UZ`m_0o&{u_mhXZ|0Ix9MSSMx)e}QD(q+n zO`|BFARs9rAg}@HE-C5m+R_cu9dZDX?h;VCyBm@2ZV>6tO~<$J9Mmi4e)sp^^1)}X zImdX%J95rdVcXfi`+jq#PN&1#Mo9dUfq^4~Uij$4(Atv(!II93;!UYDqH-;{SP;73 zUx$atuNiR^E^YMcqt%^;Pcg^X-BU7vFhXt4@=$tN*j>v}Pgk5+PA*Iq+e%FhhhfKe zs?71q`J0V-V%5_I*xDa6F{YBx_U$Bj$M#*TeF(a$@I$d;qFhgwI$>SWs%u2vRhVvz z2PIuORRo~6r0%?f@Rg%pX#~5R&TU1pV;ELVbRj#-H|BNfvN00$NkC%9QueEAaE4(; ztQ7Pn3{j`Nh=Ib;qV$Sq*uvW<4+=k4Tm`8~j#4a8+?x8&+(Vk~1`PYMQd%gba`40v zCF)|lI$2O%CTk@hG1YG;D57ZlXC`P05}1HM<+*&cd#9c>JZ&5-D2381t;&H*vufkS zY(Kx+Z~b(x4&-~h6DeoV8yBx!Zg1Yl9ie@lID@u(4wMLJ7lSzDnR`kp;+HR6j+R^e zUUO$MVET7_Ttj&7Y8{4B7$7c4iW%H-*ln86!~DZ}FlTs-ZA^Sb*HQenbUh411s1d< z1r{6@mXog+EEzRq=f@d9Z7?oj!k3fU{!^J|v;!%T?fep=9!-VCB$2czuR*A(iCOPO zgt1(z!>(!;vo{3VU4g8>xX>-_9H@xenQO6cEXiJKc3t$i76|dM^DpVQ%Np5H zOxsy`xe?tz|FHrLntn^Ec!zaY`aOB>2NiUSe&qSDwK>Tk&&m+1u+5mya;_S-rLNbQ zPmiu`HIu_np_)62Vpa*eXfLN`a9uxi;aqR|nk=^UEiyRv_GT8n>qFt6StG9_eCI`N#z<@}X=8C+ zEp${41BnygZGu47KMr<9GEs(CT95Y)kjNJ>(I9K+T2|8;FRBt@d_qq{jZ>ON=az)|#F*ihMN5Ju4B!%^ zn_wiPAsjPE1eE;zKbA~L{-`!j6K7}Rk#C#Zx$5C!>O{s#2r7 z6&ifPQ%vr}*M;}T8$K5^Wo>>a#hFc;?TY1FC&y-CS@MFXbY2WVxph?S4;Y!0uYK-k zWemrQ6^_T=iHALmb-*ErGF(?H&sJQIRls4Obb>fwtb2!{CZCp6K=if0pcUcmIb zfN1C-IuU-apO+ZvPJE0d^-6owO!ZIX^^283TA%8eaKS1{Tjd`Y-p}N3z zN^z4Nnnuuih|qA%kskca7sc1WJ1RAEcCMkL!T`A{gKM0QLoX)ul5MbOIuw_?lBf>2~1wb z*C)JLM!Fj_&~I7ul5~dlHgTMa&>cC2=rHaK_`OJK^wBcHuBRs!@<|OUR}@Rn*e|lu zyA>~359W;T@x1g4d)dlf96Xhp*9cw5^(!%O7f11MxXttG?#qF|0Na7N`P9AQL&v-a zG3;JTC#kl3^+cgMe;MF zd*R>cbDhe~XhVvqL@6eh4~=`vYrOCP%7#~ju(BU4mqUcI9!Bl3ISCwXPZ}ASZ2r}J z>LHOL*-##NuW-1{CPWWNf@ixUl}$Ct9>CmqrA}}xYT-xqc5@>*<$OJau6B8O8%2q` zrtHfQ~4-`ePfztpAYm)WW$!}wg?11GV>6x1gKV?fQbZrFswKoedQhm;_=>)USOYIjt%F3d}X8P$j^ zLwUWvZY^i(e7>YVmq94a?RmY@SMznDTyt({HM~`7!osqwTpRU9L~$@chpXo}LFFdi zj+Y^{QA%N)9#XY29x9ON{AN!hqHSZ%{T?#pK1sdBkvoAG?pF%eII4~n-j|0xdL?I+^U8 zJU&DruAe2EMj_x6mhIf3Wt~cC?;M4lFps*`T{ z>lQHLoVgr9b;nkzELe7oc{aJ|znw9(v&+|=%Hax;%phgwl0fLnfnWUD7`g>9vxi5o zw#W}Bqg?MXR5Txb8O@`j`gdI8uhkKXKPc9x1UvD zE22;aIp}ZyR?B%}&*l9j-uU`b*kpv?F88r7yT2Zrm|CU+Crz;4z38^+dxeK849e+P z9pE)W*s7MO|7*=H}oL5p!Pv3`*^7P<(^m)2b{n;&RE9dh{!%Ls#T%cu_(9}L% zLG+pccW`mocRJ_uXv@={sm8Pvo3TMhX`X8@O-~PElQ^H}JrHFm5d)$I_7h^nz|AG= z^vzhoI2rG|F75|NO|hYUhj2m)^~ib!{_ghVc@zxMQfp0;d0vkCcC6pe&_}1g@Y3|F zXRKdao*ULA{o%f9dvn0eNtgTuKw(JULBVoqZzFfd_S#TgGLkK%JY^Or6t!T3R>)bp z)V^~5%Ax;$8b5@=nuS2^ zXnMWbwyKaT*^@3XTSI_H>lecAsw$H{@F_5a5~0kQh`T=g>S+J;R5?HB zE^Yn&1~AnFR(t^SOUFr7KQOP*2#l00)NRSy*A}flX}i)>ZgpdmKO7gzpLS+^@$B?f zh8Yqq4-;h{H@j)h90;Q=SQ!=ssZC@{XrL{`JJ`HuuT10Y*nQUcri%~Y&&9E2NRf?b zo;A6suG)xu!ybz|5z*ga*KdXyApveiR1Rw@Hf`WgfBKKMM|nSG6={AUrB@~xuTcw{ zb2|8tulfvSWwL^n#5v7;6XImOlB-LCS%JXS+f9QHjuj^(jTFVG6Xy&KSv@R`@d7T% ziaIzDGZ+RWFkqq6E{l){6ISa?2{P4iWNW=IRA+1u`;_n2nc4*Ez0Y6e8 z{9R|%wsWM`jqS{72N8E^6qRLy=JEvg((3HAC%Q^b%EsAW3B~k4ZTVtfO$HmMk0)Tv z$`ap>{mWKVj=)L^Ju}yvEQTG{TD}8weGYv8@%7AqRD6qVp#0d0ZmBdIqR5=waMe|i zzX_-MGwa6ySz6A`D&T9>QRpmL!Gc_su)(pC8PqxroG)Ma9?^0qvCO&&>l+v>wdx2{ zudL(at7mn&fXK>Z6DAT9d%5SWgC0gyK^}Ad)gnaGGA8_u5Yic#+~Uwh&tV$xSw$e^NCrnh7mwXPsUd({9_66)DH` zJUo;lR`p&gh?A8eT!S+{zRTdt#}7gbJBXW2t=>CYmi_4u zftREFBWU?$%l;0_hY=A}2t|O|H^~N{?w>4a^Y;z2)!LTpti>?#j!rkuFOEu-AWr?M zh6p(iskNoymC4&bG_v1Y1x2wtr%jn@1W?)<1sku}REy$k+PM;kDi?17W6}cY#j%tR zaB;CgIzZ+olb$?L{&vKkK9(`t@s2eAfm(U4$57Q_^L9B11^w^YPI(x~1~D9-DVnA! zoDaPw-@jsDJ7#J<52bN(Nl|dBm8?4Mt4?D#)L01&Xj2H$3r6Cc#yenaptZ*KG^CRf@N_GJ5yu;->cff_e zfjPT7;W%47$~*1YMV{cIM$CHz*X{Zw6lV%Akz3=!v##97M0!AGPJgY;!ZZ|0OsaoS zaVK>ljY0WC^VPY@h6mE^4sIRlBV{aT7|@t^tp888Swd$mw69oAihPo^GYJba(8ttT zjgxPR?9hEBn-zgnTw0N!2^@$?Na}m&s&A)=mYy341b&P7Y1Dtl3x5mJR1un-WyNB& zBFgc{gbIjQ+LiT1p4K|1MGjN*CvH+^PFnuCwb0cp!FN-bYc+!?Tb%M#z+iv0nR13! zObEFfJve8s9RpHkUM0~prap@w6K>6Ay_{`3wSBKbE#KA=*y-D9g?@|1L2i7igr+cG zUFFR2e=^aZwDl<$h?$}yll*^@RV2kDBvq`ueEo$uRCeY2ZdzkAp% zzI-MWvTVZ_1@u)4PPab8-T}c6RF1;|Ik5?P_dANe_h%Uiv%X#71Dv~9-Cj<{7r(Sq z_qnc8LX-1LZex*MN=Pz}B&isg^0^qYfOK*cM3X}eFvlolj?XPc&)|=Cr_v_+ced-g zU^r>1V0mMMAQRDTVg|^zC%H}a{sx6Mxu!2m{~RxX@?=6MY_#ezT)TCykH#xuuCghs zMBr64O=1(M@WDnsVC*t=N&>0O&FBg16|7>g5K^%^j%Rcf`dv}oGRgbKp0=P4iLqyq zXVQlL-7!XgpAM5&C3Fx|j=TOOr#@n?mQrPJpnJl>MQ{IR(#|wR!R2gYD%IH{&}eM}{kHz^EQ#iSNfSsNf;k0o|5YqnbjfK%?-*xR<(&(!1_RWdM?koG-e(JKtVgpHZTLl{|nmF;2}NjUzQG>5D!6VPk}aF}x%B#zv?6%vqBJg-l}N zEqW*Cu4s;Ly8qspRE3DfxC!>rw9PAmlX!o6+?kVDcF5PDXIY`(#<;ZeuO=HwX(jdS z&qwNw{UJ+xT~V$FLCq$YQ>H}^qln>@JRXzT4WfThSldiOH0u}hu&ZqM%z88${?S|V}{UYye0 zpM65J{&Mfe132)?7Re7R;C?R7Z=yQB6s%zVI0{uqb!ok?6WL)^62TFqTB0o3^L*)R z76i3;Gt;6X3*vd8BC9_(^3NEy?bnY{!o%?T_{U$3IOSJ^Ov* z(yny!1o8EI1$0*d=*=lzyeSft$}a58=Nj3M<(V_b9pNxYU!1pqOs=zNwR+JHMHUk|i_7jY_DxcD_ zWP%9T9TenZVyOwhY;>`}oHd6eS-5tu9k3&CRxX13rRvk4~_ZSw+_$GZ?FlldkzQ2<}5b4@NBRXVvF#AsZ*Yj zl9%5GtzbC*boxK#^v@p(m*8Cz+>9=*D&(J)8(NGqJ+tWg|P|i?uD-fAI9?+jU zNpdpe)pWI6;3+%`N?|Im;&Qa=RIG@?&_F&S9IdB2ce2-isCmAl((y0~#e7D+>o~Cf z&#*w*4=QO;DOphMR)ydP4{CSnSg6^YllsmM$KFO}?IWUc69bF-4O|Eo1j3`>sGd2}lX?oI8Ym?dKEr-)5%rb(hVt0$iomWrf>q4}?1 z5x0?(Ekdj;W|82S6pLe7uhe}Zvu zR+hHMeVlY#HF)bpy)XBnOOCL9@5HkUV-s~xjN{8wyR}KPzN3>y_lD^YMt5AZ^Rgq))}6YZp0(ciEQb$p%B^J7E6a9xLDOV+phzgK30PUzM&HQorah5K zzCoz*!PvQFw38T;yNhdMdQPEGT@J)VCDztaLr4s&>s}&}P4t*j?Jcw4r4Y+(8vif6 zk(~f-bTtTYyNX%K>$WiZeQ<+#;?pwxoI@o19Q&OlvvBZubvg%6cm!hoTbbl=KhdOD z9eXo&mB6PF&Aaw8wERVhW6Ov0mL!E;YtqhV2e`ZHixze>Ay&SSsBW8D2Lv(Sp~>D= zFQr=;{$dh$=nWRs4R_c4_qQfUSjY$W(O98(Vefq}o)Ha-LXxBbRsqfzB8=T}6Ykp0 zt=EikMjU0k#@t293ETw!EsOh&9^!r7qPC3|+}>t4ukd8Te2R~+wLsO+^;;CV^w`kw zUi8Wr_5UcojtVyBy@SME7s`<(J`HVua0Kq`mqeRovEiJM1};xRbZ-PdZS_-=ZthxU!`wcfjx;#(eR6ne6o z4b}BUb}0w#GDS>IPG^3;yB_6>dbR)-43&XF9^l{|_x@ta99pF+f3)eC)rK?7M$b-m zKk|O~zw|YJ+RER*{{;Qs%*+eqBtvxHsro23XKqTbx#|}sJK|UzW!f9%TX~`9SK2jZ zF440e^xuKZ5Sd@M;0r70x4%m+QO;LdlZOUU2#ZatHwyY%h{|zd>_P})|B~j(CqrR_ zsP;78h~Mst*Bj}iRx^^?s+FF-U@Qd&@mR5xrS-5u(!+DW1^X~pkA~##5yG6RB!H~=T#L6P9Jzn%#i0xjsA-VjIh)g{a zO2sw*LtyxyrKg`HM)~^FnWc0sH8=Ko%ct&Rm1l7A0c4AvsAs9$y+Tg2PsRQ906 zO4neIX#_>lKy@0cNmwUJ?uHF=L;cTpSN=9vmFj&8MlxPe_b1f|N{a>t^A_ zKaCD8+Uqv5NNgaqJ~*J~;Ha54H~iyh-(7N3lW!ba43DC<6YD|#oD35#71-YH1sfJ8 zhQX!j7s-;d`Elo3z~9V}mfKDcNIlvW8y-_`v#OflF8j%ydWcOmDdetZTU0BOZ47PRq>@4oM=@0P4zwe(zHtpXn( zrk9&3u|5xzON{%F{V=u*H6Vbz7#n2UhUM(UaM|E^mC(I0{Sk7%Bv$mlmUw<4{mJ9Z z>(~B&0S64}t!aq8=IWnYN9%vU&u_1;;hrJek(pq5(1p%}9io&-K%JX^a;dS!%~se` zjd|HQH8jW*r^TR|BU1U%eC?i^7wEk8%wMy2KCfgrL+9}g*IA;0jg)u)^dG!c@b?Nq z#~h82TRj6`4E+XiFijcZi$*-!j&iNB*AIdhR{?RTs-@+T##dPhA2Z)Sfh|Ai!H^rs z1d1SC)P4f=V~4UkdR3N1!o+eD84*JAh0D6Bhv;HjY_{z7Gz&TA9z)zTM6{ zuv+e^P_&^wIK49q^ye-cm!IqcQ+ylIHEGN5nWTGTw#FMCb^y(t=LxTI6)-NH=Cn+s zgA$Gq=ec(S?bEKmw05*-h0ZyDQpMSr#@xw}l<3%Wahn;pvc#}QVo(n16K@~>c&v2d51i=SDHB^Mh`7mqOZ3zazG|FUQP;*DH3q!8CWoMzCOMZ z!s7xziQp$`-OBbK(unZrA_9i=M1fl&1liKH{M)}(&~&=&?2j?~+m)h3ng{c9pGQxT z4uf4S%{il@5U-Y##e-T#-#p2&_p8aTKzK$Db8Ez8}>1MhC% z3Z{(3H|_L+#&;AoFLYb4lh@)mM-=25@cYUNh6C9#)K4dX5nb~FTQht0o_Wt}I}x;^ zvOcIPRc3olEh-ueC=(i~Y*3VmM)f*}S@bF{6&iT01i@oHeDL7$R27`^q)6tURgzEY zFryaYolHpi{`o$RlkD9MxQ5{sJ>{8qKFGjq9CE#>xp73ccR=(qbJrYA-RoFUy` zm1XAtZM~zImWpg?d$V@;XC~yIXrI{#Ak^E6Xh*MnASj6FbR8S(;5wdSpsTl~Bt~^D z$&mG_T4dek*6%|*UCa?I7ND6J>cp{Owo#T`#oejc+`#v3!@0=X7!-s74Z$fJL+(~b|FGckL71C+ zex?*jf$T5htH8AO;|6+KuWL(@b)IQrrkx`0)X~AhZjPO*5}EXju>D0T`B4cuHppg- z{8;58zdIP2n2{+frsoJDc)GGe)$JJgqVm!dWhDeiOYp{*wMyRFjo7zy@SifR)XO*{ zh`l}iI-eb8L;uV)J_yTswLsG%*J;rZuzD#4Ac|4my+@Y!~7sSv6e zG*)n*eCAZprSa=l0{kZ;>%OCk`i|FaYKu|8JsWSikGrV$H~W04toz|pst=rMuEWVw z;PVviD@dlz*DPtc$(tXaJC!sWxHej_GwZQQ?XZ3XY*_kErwO>s38*ODPyBV4lbXi2 z-s2fci9lp*pJR1*fo#}O@tB;oAs*J}b-_duW45 zpqDU1^O<9d?}O9J0l9GclPxi{jd+dYBI@IJ&#qj2)$RSchL#t2)}zCvf@vz^i_17l zXJ~7tazJD5Ydx8VfnPzkIaNd0><6GV~V@_Bsa|$Y^Gf_?}&Y29xR}VSAm}s@xl4pv}Tjrv=8od$l|yY zmzOdl-K1F0YaoS7WioVR0UoZy!!&T#!P2ni-n>pV%w( zzNvt}vS60`=}akAKINonplnXmtv+(qoJ(#iIQq65`8YN)Hk84l*0xL|cePfo0Xw{H zI9%&)H1ZR{f4`})2>zlbJM9K>;|8em!wm5bF3Lh~)3jN!x`L924DMALlhv2yMO?cq zSZTn8^=hu93e$F(s&V}0T_K6A*r6S3Wko1wUTFC!YWZUviq9Jd)eMZngQj3X_zzG? z${BTm(WXC0wUe0FLr-PA>cvk7%1h^o8;l=mnP6MgG2OdUK)*9to*)^Wm#^$y0xmM9 zrzf?xwx(yp2-O{|AIAg|#@Fy~>K(7}Z>Ijc?Kl%(j6#an$>v<-Im5}LMxuoNgv6gx zm7aj<^$^3^F^TR^kPt*WzqFXp3P<>LL*EvkY}|ALR* zrta5GzDtWH1!C+8y=7Y4LfJd1fAw9MR`nSD&r}ANm(W42j?ukdY#K5O7UVh;yqg(%9&jbfo!h`!{38QRT}C zV{;C2ZhPb5^bGOHy^PP)Ol#QL$r2r!&_%I0Jazt;;(_oayBzU`U}ardm2GO#r5<** zOP!H2$*F93<8#2q0*yZ}g7t)OARv@|W81RnJdJYNNS6bND0aPwyzsJBvM$tsB$RJE zwdJaS{1KE}$w4b@AE&F0px^unbWeJuU5dJZli#-esj_jP8Ie?3&4i=U2BE8kL46%1 z9tStA==1C#(5Q_7ia1%gO4uN|tWr+M<^%fAFtmZPz})ugN#)0ro(EJ;FvgJ0kCi$U zDJGh?+0$<_{SQL}diSgfLM=X0r6MQI)Pe>~2dx0e55l?=l4zW2)gzsQ;FjdH=JabW zyi4RL!aZ@0S>1LtE^<`C_u|1U?U=(JG(5t>69tD z(FFZeyR;}+&7fv0q;CTwy2~twg%UVIY-tL1?9m&7b*ys&@d+cn>cT%(9yDm*&@9k)yM^V)=JwUm@mIVz<)NY_vD+L6ZT#e<1Z>R z@;sAnw|n^{8$Gzj#v^U6!NFtcT;5Vo!nSO9bQ~50oNHhhirm__%R93yTm_T#1$F{8 zW>M|uK#iG=(&ExGeERCVr(CSTN?V@bTBaed9*uIFT_3q{SjTC9Bun90oE*H6T{fYZ z^;*pbC^3RI6HMLP>F`_f_S{E`sOPAwUHHEKOATu3pK5aCxHIV$(}#>i$mC+j(tVsF zn+6qR)7ih#I&MbcT7;BGZ4!gcB+6J5Bt_C>tB&>rI7_%VSaQBl4wYvFt$Uq`q@F1k zp#`_;QL^W~W6ANy?;S~SKNT?=iwi<|&6)n9M?L2|E_>lV#9n_f=(V~!QyTt< zB#Hmh4*x~>5rU0dA%F?^70l4;G!PUczsWZcLa29?`SQ#tG%m%qd|@f`q|snf&yJmx z7E5>Be3VPqWkayQWu- zS0iSN>eWpS?9vCfuzYJo2mPwl<84fdD|+q@J<3Pjl=?OjjZkvv63W+3 zd_cVq+|s?6c+RoZFF`crK;RDMbvyt1KCgZn!~R)G0W^=J6{O^*asM$cA*{&A1S+lB z1UW*YjTn>Qk{5g@owz6O19V!b7F@wker>tyiDivH`5XJHX|&xb24w^j<0YVCS3c{- z#8h6bE||sIN!&q}fPRVQbPRvL;ryez)2fA@&8k&?yf$uq$HiV~y?;wwL~hB--ofo% z02@5<8aDBG?L2Fxx;iMowEyol96bDH}XA9mFk!GFD5e8W0hwR$AJ_gjy7~+8^6b(bQl9g>WxPl z6BL7n?$*R8#huz7VId6=W9L6$!1|RyD^o)UPbrjARdl7nfb#tHs^>jjYy8VmM`_T& zSy&~a@G6*xc2hc1UGrG99iOqBfo|~>+UeM}7~g{pb3*pSocAcGi{;TJbe2X!SBG-f zD#WDNL z8vjvTa_oY&QgKd4L9!&A$PVP$+Nrz>`T7CJQi=NrwS?Ib;_THbF4od_+=1etK9@a? zd{oiQ*6Q?&uQ`unX_YIOl1c`*&ejWu0QLr1+eAy>pun+Rf(l!@?#h5(aF&>Y!W$30y_VNWef1{>hf9{3JmC+E{=e57#^HOkL2BVm3?zRepPaP^X? zSxIs-h&JQZap+Tj>I@3AI|}&&3t)UWAa`NB=JWk4hy}B5@-X)OLl+=6sAtePJaJMO z9dTZ=b}!R-M-;pimZhsoAk7!m>}nl#X0<(?_72pe#~l^crWq8IlkZsb(qC`RkPku3 zqTMRcMjifhdiW{3@ccWQt4C_TjQ*de!ulIaC}2_g`#{LeqAG=;kI|r`+~hpc%yDGzTJw*6Y23i0t;!+sD$s?We6|V z=Sj)5t(Fjh6F`fvRfMO;HGvJ=zpv8QE%{!H*vHVMMi3eO5`zXN0Sm3$nhyK7!APqQ23C@_u+3$?)!&!mlTi0lJHo0VXuFlmPh0*II<7#CLP2awsfN_@(+Ru5;;9Vw zm3i=2pXD)oaC(mICQ`e>wUuucSq)ou>tXY?lgzTg0qUpA>xmlF7i>WcQoCA?L$e&E z-sDhUV4vTEfkTFJqWX~fqEm(;j;$)zvBuunp2Nc6xX!i0*;*$x)NYYtK#uVf`TV^w zC=mx?hff12TV`ccT3A|&1`7Iming})WzZ+V8QY?`p^zN12$jcLE^%JzhWtLwuPgir z#q`!5^L?rbKB8igFk}Pri5KsvaGmo7))~h~utAQ(?PItTwCtZ!x|H8R!4X11-{XaX zdF1iWBUlx_duWm!$M?5CH9mzG0wUvf@z}^vf{G)DSn7oY!dIAUtxB4*TicuLR8CFu z@o4!EH8`^-E#9K$jH2h2GNH(o6(m)13~u9xydw|%5c@{syrx4{m)ttYJI-AuC!2C7 zQM*`ym0}h7{XsHt!NAs)Ic@AgZuUYE%zv)vx5fQ>6c6?oFiO5a6w7i~M1^b`j;a4u zV-%z6t|eS$k12On4C=Yc8h>dZg1~j*Md`FPw;|u7QN{VX;G#CF{`A>OX1P>bCw@+` zVcQ$X*RqX4jna26Xr?w?`;RuV%7ty}n! z%z^R2ul2s<_8fo#C47MD62JykL5kp1%xjJCy|uqH*n@;{%)ZZRyLf^B_p0Xd8tWd>z7m{elI7_@InoDR)dD>^YwVl~Z5cdLS z{{<|F&W7QgH~fD9=VlF!j1Rt1xW@57F{1t^6K3e}j#TG&9eA_bg~A+983u=Jf+JZb z%?5Oq%$y3U*x2M{6kkEdI~8c?{)Jcem%vW?3nHzM!I^bzS?cNOuiXOieDALIH-z7UzZ0IdGRzCaFuJ^2&gGCBtm>vFynRMAa~WRu7Lx8#4GBe) zU@#C&Z#Qc&>o-kWN3O{09%iiCFf4VKFR^}z3nHV6T*~wU;a0~JGd(qBhmPue7^V9j z&;Q@vo~{H;01s^hkN&34))!#9c-y|0!El)0YXQZ_G&HCngj^0BZmU+1{_ zpB+tX%D_=kN=buog;C}`I!cK`_he;2S0^jSn~x?)rD_CxM6etSDk-QB^%liTI9_}8 zY_~f}s|$nsX<NFmAEux%RldDlY`7fmuwvGkc2);^)?Qw1V4FR$-z_v%Os7 zzq;)Ffzj;Q+?d496$OrECL0)uSsqi>Xr$IgK`c6s%O;t5(ja@1tu^ukr!7LCk7Ua^ z(?kbYw(lG6KINI#V=XozGdz@ExcbKiVmP58_x3tW5SjRImje7BLQd3?zaw`o0;dE%_F7?U!GIC(BYUe+)!P;U-NEs%SATEVUU z%EY{DjL~*AMRIb_szm=>-n`N*%}#c9h63m<;DXt{ZD;g)8h*-Yipq1!A(CQrP9O02 z98Zat`;!BH$|_w3N@{KBeQ^5kNTUH4WqRQ);fSa2+ObiQBC4uVV>(uB=kPxc$a3}= zH++Pz2Fw(VpM|Z)Gs1oO?B9bW*!G!6A4j%`YV|1WeAj)5vClJ>y!z5c{Q_@$U(QLX zv?yeXcV-I`YJE}XW)-nX>~)O1PX$AJ3T4jy9ER##ZJKcvbwD~Jx|+u{C2?!($)lzO zJn$O(kgye@VXCYV(X!JBM-jr$W&d~7iGTK>m=S;vj*>ZQg_%*DiAK!YI)*ShBum*} z>zdadbf}abrmIV!C%N`ZI+vFjmrITKz%SXUO21KJvcDL7G*T{buBE`yRwx^*^qH-? zzJ#v02i)-ibnvQu3Rea7GZnM8=UtO+w0W=^ESa>O^XmMkzXa6zp5zg&R~nv!5p%2n zI4zQX5Mp4iplDJNnAjBNMJ16eNkc$oUyCf-HCWKNTBqp~`Y;lL#hsG98aq!K+Bi5S zBRo49uA5-(8K7cKB8->(+I2miD$s5ik*?Hf|jqaEf=3fNcF4{=@`}x=t;4~*WH9hnkY0KzpWwd zSHs#Y7qT}VEU>6`7_$sk=Xq0@(^ajTtA⁡7(_I**|+L(=#)f8=vhZ1YA;;l*X3) zIVV(6kvhEGgkISxl1A>h%8&-_-~6leeZofB7Eg>{UdA()<`He9EXupZjg`&FSnL$w z^gjN2tmvt_R}&#T^*%ion=>ixd9KlSdn{TqjfYpepJ($QOj-*Fb&*WXeJ+M2DlT9Z zR}}QnyZYw*#KG(}yTNlIIa@cwRV1d1zJpM+pe7<5`iabjK4D^+v~2aY2eS3=NoOk8 z9rr5*$&(CnN*kuY$@7d}?b>T8>vpznIoU7?kyl<^i<4s#(@*R#?=w?gc9S&G-6w)Ard(+cgyaNF~$@^?!kc$(GWW zWRJo2sX;#Z>Fvc$nDI^YIAEifI3#Uvpzkiua}A?W?(Fa^#r8L8(;+E$&zpu7Wl#W} z4ZHYzh28JD1WL>+OkhDJ^bAG^>4Bw&S1wsW#Gy^HG);`Qh%vQX^EvklAHu&{#M|!Y z$z4N(uSOnrHVigyP|-~jmp7f&A2By^_vm)xffp?A^oa2uHDC_w(|#V@BT&y9Qq4s~ zSvSmEK5O{NfLFKSkX2%stXLI+-IiYY!{uJ*?t~lCL%*Z1s10X`Ex5oJ%U}F%gnzoD ze@LV;;@;B$gw7dN0mtLW5ShaVt(Zxx)~4>CL)DXl%+Myrom#7}R)e#%tm^GvO;;ot zrP3aoc=5HVUONOe0LKik9mYbh_Tq48S>|^10NJWAPC^}B%9Mg=XPLy zVty<>!8s?Aa6mm=(#Q*Y06VJ%;kyqa9nDt4j77y-m9CBDbz$@y?EyIoY$W!-5H}18zzqz0Z{>4~C(m zw(?PMEPc;?kQHq`NR_bneM*#Z9zLy(eNKns^_$rKeiX0vIK%Z4u|Pro-MxXeVep`` zlp?S}V|glR_V4xMXVQBBn14;8WX#KaB_~Xen~#@8x)S2n_ECow?<;Wg)u$$}8VcY_ z>3K?FKLwi;Ut)>lxL%pKDMET>AH)7Z!RzV^tWek5-MB3p9MdA{-4CI;-rVapz6^2* zLbDs6W(3NJVbkAN$VVKA`Q@*YRxHN!pEwOz7r{Z_Q^$q5y!0At#citj9g@5O zyh1{%-I%$r{*w=4PGQnP4=~*70@Rai<#rNLdf;m|d^1lgqABdHAEyWq)J({`CNSrut~Wj@n63tSX`Yvo%jWPrO0FYET=Rr1 z9A~!2V`NEb8m<~AZ**dYm;kht<&oTFS!i_d7%Job9axtkKT?fPiBbO#oUY?4IF83%b21Zl_T}?Xd2FuT2qZkKp zO|s)XihoG#oaV_MyhFsl%oS13881$^Lwm$o!+}%Cw%WkUc{ua}1_YZ9y*{@>j@A@J zB*TQrHL7m!Q(g0gN!i71a+IJ~dyU%mHSUZgIs{c{!@m>*-4c7E_W`>F`IF&^O!oU! zK-y9E{N+oa18%Z*@YUJ#7j)hu5jN5rdy@W#d45|*Z&!~ND_22TpBNCDWwr&A9242O z_X};ezUiz>2?c7`S@9-3;7Sv%V_wSeMPDoEf<3PGiW7 zihga2_DtZ~vXsOox?S6`FL2^{bx@l^rw+Zc6)Cbt;5?LAJhLUkBC^5GL(-TNdv&aL z;s4@xpH?)E3x5u-q4;?3H^cKBzGoSdHb?vJ9)`CUX`37#N)2V>AjRlHq{_iczh25X z?v{nWeWtPV+ZBu%4*~A>mv}RJHG^Vx95f-M?sV8f_Z3S4@0>|8eP1PPgtE+F*Sj5? zU7)$`lsJWSN<*rgF}}3v`-(@e<^7=U59=7%AVen}cBK}g&%75bIJn}N;Nsku;Q1gN&r5~01^yI>%-l0KYZ_JsC9L!_gGhjUvO#O?~U9FOXW!S0ld?_EUx z*Yp5Tz{&G;S!>G@?#3|*Xyt27B+iSUA;Ph76X7=^r)XmsyibnyR-e6@A(2vw-jBBf zPl#bj9^e~4J;R`bHa*wXLWt~>NGZ6^R2}N|h-rkRxz~v& zP)~M%nGiD!o40c48JV_Nlect6@FKwcg5sEBz2J@(@W}nb!$a?5qDEP(M%v$n@`GU; z#UBCZ%EV8eSR((XULCZTIAs+@PJGcF@}1bA$d)x5N+qT)z8w9bRov|Xir z3EzLiBRZNTjd(^_Mvg$bs2P|@X_I}NpSmH3L`sJdcHu9a0PFqD7VJaFGZQb6b4~5x zexGqxg4INZyk`u<(A;$DLX3vXgIQ7AjZYob}^FcAhFJ!b5o)?+qc&K%~?ZT;?l z|B?D9v~S#%7rvrdT5xhltGc;|WsaF!69}F4X&&@rOWC~=OP%kQR1SBw6bDQivgH(} zCe}mgRooVKj9-vUB~P+oQ&kjnrzX&;P}vjM4diu=ct&@WYb55N6fPdH`s>;90qmsm zWM7y(1KCj+h7C9{0ip#)q5fQ%cr_FZh&li=qfov3=3zT<- zlD~tH(hhX;XEculKF_bUX+|r$I?g6@I<^VhIG{li^$g9{Sj1wevD*89qTFERqw;pM zAA2jHpf1VIhlEWydmcr@BsQjriEy^&gu*FXTNhRH54`+~V4*zykbp8Cky=`~Q$C)vTS1KuFv7DB$8tb=G9}fgWxQx$pP{tW*bpB zQL90bgzs;?arckUUaX>fCUTJ=2g8{fOZsJzuig)-QPge7eMTI-$_ue|V$1-;{=Sm` zBG#8~&?cV)k&-|exscQ9$jS|$%u|-=G}Gv6#8mLYe*D2|enu&upOS#qyckadh|F1n zk0eF@?KvgD*7$TP)WTWB#WkGviknX=<>w{{w^)=*h}-sFwGW|dKGU)<%%=I5!V6&t zFaV^S-uk%gH{}Aw3%(&P;AW4wHWvS^nf!L}S^!O(GY3I@dm(E_%NXa?)M0;!q?Gzc zly+p~^RyktBhwK@9LEFbq_@an&QXBfn@t+~J2nc`31CBh{I=^~?hxEGbq~=Q- z!sVn+8m!PM#{So;U76ly^E)?Mmi-+9f~ zH+(G%blq8(I4xb;qwn#vH9%ZlN*XaP&v4=w^o$$@NgL$kS{(FWF?i{icj_|A=<%hk z*ZOQ*`bfeW6v{`0D4`VMxhEASvxDx&n}i`S(41)*P>&UU7f+st47pcY2W!2Wnk_=t z04%5u)8K*ym;AMi?U$49yb21b#w|2w+eoUly7%DQpAJ6Ie>l)sq`7{cDrY=6hcDrc zJ?s~YX9d5yB~%en7bdm);bBB#>lvH1iQT!-~>8`yT zeb!t3iZHC6GvUmR{{L`=qVjwP1e>%&sUNdhSn`wEsgMKjr<=;ZH3Q7B(!ZuGW3AYo z@r%~3t&qv`7h+yc5`Zh_tmL->=JPqUxH)B2jjcliF)~R!afSw4U#RlyffG}jD;2Cz zqy?xq_ul(I73Ngc*Z2*<|F?pY$@RS?h&FG1$M*6!aO7u%biCq4;CvBu9y(c@@xEHC zY#^<`Gi=1-L+pTzpKs=58hou7y*BNe)9!KN5BWt%+BuSJ)eD%aMs0(ym|~RnkRgB8UBf>&?f#gJAe*#{*s>=r>k52Vdg=x6HK8$+XWC zZG!T?{j}BY7TV#&#P4uWvwu6i|LezC#$P&7a+gw6&bGF>*~$xH@rdFc1B+}j^R91$ zH^d<7L_^579z|bUGe|*IZNknA!DF8NMjVr%RR(oy`K>`XirYOwb{5>a6lT=omM3!K z)=G`P*?%G}(B57XEd`y1xAr{y7wLGUurA`DivNazW#XR{2&K@P*CPJCj8RH{t!adV zLjyPf?zm`y`qA>AX!bv?Cn?3>;L%H4;*87soRZKKVNKz^ed=(rM zwV^AX^^!IGssuQ3yt}vf&KpTPa&@3g-ZL|Z?HcEb*zyY#9RX^@n? zd7XZF?Ux!NtXywKjG?&f3pDzUF0Opr(snLcnwZhXC4>0zN*DM*%s}(#i<4+uC|ct( z8TGTbhy4u~-Z=_&9+t=e9^SOvU2IG@&6$(>Bk^XCYmblO{p4H zI_%eP-j;=GWfoJ(cYQy8IJIeo6>HX7fikww{abkV{CB9B`W?cw z)i{_I%M~9?V}3_>_Y^^pw)UynH~O~(_R$ytZp*@A!3O6x{yf_sm612m6XXK<_oUCY zM4AVa-mnGmcD5#;+SUXT@f0%|?oVjT!j64NT_fsc+YFWKMyPLv4^R-uI#@_m@!p{; z!0e*9BB0C%K840W;@mADaW*!PI48akW899EFo*_K11kCcwb6jFd}yLeXr}In-fdA2 zx%C+m7HSyE@>6JibH3pz+W&>N;(6-W4o~^ig=H@n%3{2*S4=qT_~+o$xIVMQZ#_Wb zr(y^k*i?^K=Oq_`tvT}$ymOe@ns)Tfmo@s%Rs@R_ zUeT+1`80u+{$EzBKvD31Qy?7o%kPaA(vP|b7_T)3=IO$FH*F(9YSxs49TF(gH>!~P zb6!!@6l%Sr?oy9H@p|l0TB=Mr1=CGmNsVVW=tgq9vg!Ws@%g6Yy%RHU#j92J*>^Ak z{NJ}M2`$WR$6}m1-vn(?GldF{c$`j=eqn3jTk9C4tnzLjW4b3iJef$nu<^(pm8}fU z1OQ!>*H5I|=sGz5+Qt8Q=l3J?)mYWXnj|jd5Ohzjq4_&C^^`6onB|vbeX7wOG%$ZD|AP*K?(QLq zAT6RXNbL6QA8<{kLGUhmbqZu`M)La=P^S3xowt9_DWbT}cP5)#?-&U!Nj3#EiJ}}7 zl31SO_X4ea>2_{90sF`oUCW6Kr$@5as|sY_D3GbToL#soVbOqttX(W=l)m{xC^9-n zBd!6Nn$UNkg-b)JTT|*A9M2bCi+!O99K%$2Y`6KNBX7;b9wL*nc9^tLjJ5Oo`T5^! z@eQp0_Hy4BB$2TrN=`4}eUIGHxfGhLq()~ihHqmYhkfj9w#6L%(5Z&S-Pnc&N~OBv zCf8k~ktO?|`JQDfy&-&UPlm);cHv_tUMF~{>J$76WSN4i(uT|h#bbZlJXW~F%6DZ^ z&AlYXL0j1=LK4D(nO#mA1p1MQ-5>fXemG^i@=Y_3@q~_3G%C7S#pL=E5W!N0%Dnx@ z+Z?B~)og24rA`WlO>_(qgb%taX15Y;r7y#bzQ-};PZ3doxwagm^BwqlO0kmnEU%Pc zXR@2-_{8$@T&TbA&u`!7_m4@wzHM4YHD1jrlj|HLddv98W3yE1r6iFw-+cyg37Wq1Dz#*s{MK(vNbQ@!72*S%0-hHb=k14&BIk z&t(MjT;lH6MRhZnpU;g6)GrN=n!IpA^~gpf44uDXzYixP1PX-|uJsYMi8p z|0V7h%HLb3cI?lSSkBlI=>nBdhT(J2JRNhN#&o%a?x(S*+E89eR7)q$lq^tH!j&@R zHwzqF8n1p`PQrwC}aJI zVfO6tW3TVyn76ucE5{$w;e$n&>_9x&+eI+v)Rxxi*;pMoEQcpM?8_FIxo@pdQFfNN zGojkQRMSZU@S+y}j6v%+&Z{v5xooc~rIa(`D;6yxJ#R>&Xas=+5BU)Eld^|f>f&(} zXR%&*#I8))_ORDXd$HDlJb;m%oq|QG!1!D-IX5Fr$of(ucS1TLo77k#)5(j-v_b_* zG~&Vo)<$MnkS3rk;7$?xp|aT$m1Z|W;$!Xh00l*(Z071op8yBjoEJS^4kx znEFXsw0m)KJH>m2a@Zcf*M^f3=ufm(v#*@Uw&o(|IhL= zZGv!_9PN$w2kxD{D^CnUNN&QaUcp*uj<9`NPqa-zpGcIkqI)AiO>@L^jAJ4@QVGk* zy|09Ra6|=8>g*y6R6y$$dpec$!;_?^qn%{z- zdP7GDgLbjGCNn!dQ+_}^whFi)0O$S?c~2I(d`(js)qy@vHCC)^er@qs!KQdymUo*W zu09$1f*t1O&p^&w4cZImjN$;J#6X7D9V3>15EG5%F8A`=5FJzgd^`0%6;VZVR9!)ub)TAzf219}T>rSWo_GF zp{w;=45m`YTlH^Iruc2r?ZmwD&G9(S?IT1BW)1Ii`LWsTU>4asB>|hVO8$&jeb!>3IZU%gedMxlQKWy{q5tJ$lDFSE zHd*QxP$k+>uKBnJop3)t0XP>8o4l^pG}+P;p&Y^=WU4pcmsgG2+Mdyji_IQ2Xo3}m zHtTrHrdAQqp3$kKjyebVPA0uUZ#1me|Cg)mxA&&QY;!M0{P((SuFUrXh=RIW|Jyk+ z?5rejF64tHJEX~E#fqohn+RJ))$;N*;Jg)K=5hvb!z5y&+L&$O(;A<}E+U+-hEH1+ zCC@u=(K?lP)Eh1EbK1rER2g+FKS^b6I!UJ^3g z_MA;OyUFDUv%}EIpC*VFc#H^a1)#6mxjFw7q~xYmcK0=Kryi1e8cT+qZ;-iSpkQ@) ziLx8Zfl0InRT8xN`0l0AX#Flxg!OL+idDupGyA$}ZMZ3v^>EORdoD##-WIs&exEBZ z&@#J{u@5&G_Urcq`M|_WK>t87e$&sh08R!8$()-?rww!i9MVon`*FcWHNy;Vgc5v| zg*}84AaF@SZl=>-ZPH!ANNF`SuP%J3%-m~S=gXTnYD)3I;&}*15zI{(ZbAR<^FJwM zI5qG}pIaAdhWp<{<5#NX%R3Q44ES*Gkt2IHPW2f9#=*yVeca(FZ z{7V*r^D=%(g_nCx&uU-%9klJ7U)T*VCZDZ6Ek(l}R&9HZfwXivimIYLCxOd2Gri%9mq7@qZfzw|i$L7@+#X)TT!6ZuJ6v!t1cb`rR{y7A}mx z(?Ooruj>KU6j-H7!)tAzt9)f6a7Bor{7C8hbh{-zua~DOuCw!1)Vx&zcTK6ms=)8E z;Xijn;o0yeYaUoKJ&pgAMJyh=#o`JQ@OdyV<_5YF$uA-51`{syVy^r2!HtqN?m=Z; z`*Pr)YBj?|(aY~oaRMC5CdcgRzj?`F-TztwOtKu9MjLuKSUDNnpy(#Y3XRmT z7Uff>BZQzQdZ3F79EYSu{-x}!!0?+gWs;{d?+#)uhI{6FrP-uk}LqIOVJ(G|&bo zd?z*;hZst$*%Cj(w;=?M$GK(wnPcYzT1)Q#5d`5^K>rr$V~X+Z;l$sfy*Py5orm@BSSM>prJEU7ZJ<{Ps1tB`E+zP+ejh5GrjT7$iu#y(vJ-Y^WT zt!3bwD~FBt)t?$tRqalv2o|;wQjAX;A8k2CZ~;1L1jH0P52*JXTs0dCL3d=rc|A;e zq9g`#_L`d-HC?YY@PB@+%uOtU~|q?d}+U<>u2TKE1pwjkYsZ zz|F48x{+g0LAG+QKKz+RcVu8R2X?$db{rh3D5?0q9N0jaehmqDd1pE)N79J0o z7R8m?L>=%9uTudm9Hr-;)cKLfU|GG}B@CM8e3u@TP@`d^n=W~yiL>ljiblJqaevz1 zp75So+uFkj8||-P{Wpm+ObXBUqgj}UiGMeZDmeUD$&<}+hBHe&_t;gpRFp5AmqCN! zUAe>4rF@rK8WAvW!wuY?W=W(%o3;&{kwv}QJ-HIl1*VbN>z3}#!iW41WWb_+Up(}f zN;7QDhZACV5#M}aD}q6V)yq^X7g0|7Fu9>YzkMbN3JhhnOH2~NOcxY0$8fc3kv2y#1i`X@Z{d3#dm$>-U zXT3UuD=Jy@l?~Zh!*Y?KG6_sUq|)lP7GLPM=o2cvmdY(+YBU7gJu@s)Ta}UvNA>4% zgZA#ijtCOuyXk(%~&d85?Q;{#%|t8)oy%&dyGDk!;MMKpH1WK~hQ+UcU(JS*i@c~#WrG? z&7@AK(X79#h*&a!&KOB4Ha)ApOSKpruD8J0kx8He?@m~eSRLIu`J!y5mp2`GbgX`QHBrLi;6C*p0w@Ng#n{omB~7hJbh*s3O@Y%<}KTdM?kG zPuo(qZg&^|`6dB`fao>^Zg73&-9{iq19={DcjA#_p!@K>c2R{{+gW2^>7bv<0E^!x zwU%E&s{sB)P3ED)1@=9&?C3o$f!#z~r*+BY7Pcv=y?%M-6k)g$bls^s!U5mqPkQX* zON6fD?#ETY%fzpgg@V5ho6wZ#&^C z?K}M5E3%Ee)Y++V5+3gdxBKj1i&|B#qU)A;GVOeCF$wNeV~99Wa7pkX9nM7$939=O zUpVdS+iw0e!r7!xsOodTk#37g%#hX`E$grFew_9;eQo~k-vUJ#ZI&z+>pl(Zj&D}d z?OiCM`0mLB797pcn0&%j&W?;bKWEe8vgKHA+ET05tYcLi=v(`{@{)-cwA4nmzf9Op zP_wYHPONT-i+`(ZZvLD>w>oN=^1>?dw4(G>3XMp>A&sZGxwx<}vbu(~IIl6tBtCxp zA{19Cr!x=k+SdaImP*1K*!ypHV2Dp?85rY6fsOi%*w~wWmTEQk!}AlUQoyqE*oF1j zov~D5;}4?I9@n7SI#&tJu+8TzcMngF9^8PN@ARl?>9a)9c8co<=mfx-5t)3#va;$y z2WTk%DzppK*cC&GjFg4KJ^{)C;F^{|ximfo)jF382PsFMUm~cXE6xJK%GdB+xEy@t z>}&w}!zp0nYrG-VpOHqHvIB8K=N@r-k$~v?^u&C56C@Sj$`K!rBVJ+IdS&UA85~FQ zrx$>{TT{E_QemCs)ruFm_v)2bW$#rmXN9H^jx^VlYaVmYryLkl&u1rvrP!;`D^=>_ zY+N8GaAE7zZ7=G&0(w4wHa)c{j?t1XCDIa#%aDh>-ZCg6S`iu9B&3r&l|914ND^DZ zv32nLS&e^Sq%uf2_ne;&HO$!5SjqT?etAG8?r6w)w+(vM_f7g@vI719(jB1<3z+fx z>gT#6RwmKo$VrVG9dtw zyQSV0yuH13R;JukTPY$>RX%ztG*_i7}$~IO>RrHMi1c7*5{^$E)t!cEMik|M2W?RLmZkYYB&GW5;J6{4>4AGpo4_(BvtUi5R^cSAU-N3oFq3g24i^8-p3I=bj#Z z39^qO_T2O?Ny^Veu2x3hPh3o4Kr!NhW3JFjD=h!DF27tIO&}0#VaEdANhbkGCHC-} zSnVA-OQ!GhHQpSo*Q9Y4iaGadtUzYPk3-WAE^f)@24aKYosOEb^686B2invIS_eKL zoL%=3l(sg(fAzE!WP(_C8#Zls_m&77Kob09Ry#p8Iz!4QBAJEvKDwE*`jtZ<>OcU( zDeBOI2o<<`6NAl>j~Y){h$V=mvRI)>8$D)c#td%8U=q_)LDE{x^ZGB;AXD102st8c zH&w;KV2g$HY##9?$zkhr_rlmQig+n0SAX80>aBY&!4=-=RPyOqS0RPF1VA4J^xmOh zei1E9q%YByp>2;!8Fqq8SL~rry=PRnlzcxh0Un_QlOjY=doREMR{04C5|BDOp0H>e z?P{*Kb?esmhS@$u&o|#oretBW-Ysi~!~U^9Gq}1@2hpJVEwgrj+#1Z0yGX^<)o7|` z!2=rV*?!kbNZUUaZWXEf%kq~gk++xp>wOT8E7{+cm78tq8SI9hFy>R>G=yo&gWU*!`r-XxNe+t z#%vw<4<9E>ZN)PDwvO%y-9)@735T0+>cgSik{}`o*cvLDzYbBWu&kZyB5OPiZ&*ph z*wBX?c?bBD7>=dx?ojxkb_m7txuv2;gU@%Yl>p(@3=fi8+yLt#BVXQYB#xcMz%h z>3aM?@kVC=v12}o>LQA@!HzPAbLEy)DO|HwRqM^@1NTQ#0qNJ9gUM`5Hj%rn-r*0& zeLmUnvDVR71m-HHpD*tpX{zISzcv{rNmnjQiA^MX9Y^==^3;gF2-=FUKSp0^dF0~9 z=ca2Ohv#<$dhT6Uix5?2g?oaX2i%e4VrBL_rnyIm3X?$SU?^kpM%p7Rniqcr2XWcU zm5sb)U*s=Q(#Kq+>-G@DX)aHmRiHg{a+ned^42%#y=rUB@clw5nPGz$c=*DO!x7`> zX4&RZx~`G}p{dHIC&aY7+u8@#1&K`IW>~qPZaDkZlS^jCN*IJ2bzF6`635EV7Mw-e z4%mbav?D611=2wwRR|YLF_gV>tUx!S^e69QgKwEFPU8ctJ4D7q4?lEWANNZ zYygENotx8R){~fGAhxal8)7+Eu2_5js~?&Y+>U%eL>L23Y-P7exXKEmBhlTxovQKA zBr1t$h#&A(Y_VN^tLJg6t_|%m9cG2smA~#JY{&mP<=Sz>lyfwtbFRxrn(*!5Jc||C zl!O(xPZbA?er#q-0+TLbDrhec4MP0Gt{&Ld7Zl-PrE0t>}(YSu-JMt z>cad~LBYe8olK%p{RPUlRV9V&U{|L<=CTiN+ih z8spbG-uGnYx)Ak}j8(3hJ=U>%O``+C5Vo#l^BeFP@oB?w=F7*c`@_&m(M$R5A0iLd zsbSznH%L5d>&ec@2U+w5@BLtLry8$@Mg+s%)j(&hgWZSBu~qexO$lox5L>uyTZh z->Pb$kl#S#+&1Lx!+pb;*s<(QdrFNe)Rizi=P?qd+jxRRddGLOD|{uk*|aKke1(rQ z<_{Gq<71AHWS+nP0A=z{_^-JqIO~tnG4me;yfL<(yCCF{gl7ufc;b1lcv>%dp68Ny zX}>rzqGP=CZd{FF?L9N%(YY2#gMA=rhRB(N%D|1F8&Dsws{*s6(qUe{7Y6VQk_qBw z>BIn1UD@K`Y)XT2?{r3DMM3tzf*OzE;%-hbaV3p8uTO*1+N7}uP=>7goBuV`p(Y>l zlD6lH%LDm7TZA;e5MqA-Dz5&1yN_bR~7QUYt zsJ*TM)bxpsHD686&yM8*zD=~F;Y zy$DxHPJeBR6E|vO>kPmESVFE2ugJ%qc&;v|ZcXK>S2miiTh=We<)WxVpPPzR^XXBs z6o)k|xy})F1C}lUcUL|MC2ON4k7s~8>d)~@YlCuxIs-)GOHHL|YjN`1ehGZODX1&n z-wiSEFRn7BGg%XAVf`Qy&yK%AiAbTRG)dBS0$ zy|MQ5l(^*>-f94LX5yrKD}?isXT2O=(1#+aehvn?2KlAJIF=s??NmFP}oX{ z(C)t-x*!W3DCTs&lo>4;{!DTYx&!6f~F$|afQo2oyC)< zq0^7c$L-j7eK>55{&**-CSShxL|*{o6Kkrh(<8sOS#iI>z*U{j^~wdeeI&K%mrX0& z;%DDwR8yJ2mueJ^P$7F@o5fAcGgQ+OuwHAT&*O3pS=L8tSuKaT1Z{Gdxb`nkgN*C1 zXsf&j*xqVNDC!k@O>xKAPooCQu+mS5{Nd187&bW_d4i-a{dJ@tvZZ=2pZ zamyn`L05q}4%fSq=B{dL@J@B-m>IM~9Rmu>eFohaP`$?cH6D!xFN`L_pbl-kESwo_@_Jid+I%3In!PzpaEdrFbo0JFc`b<*fh z9?on0m3%Xu>|?JjTOaFXXIINtw6_F*?f|`YU%c2JORl+cc;#RvIkS^jrjFsHOGVkI z1GbUfUJ3K`(!%H?ccqqfMudhXP+B57C66iZ{xpfSPvzSVfzu-L(WeR&2&>Kbn&* z$Xn$h!Z$Op>CvY6oqEA>d`x6mdk(;9`#cb?Ibs6a2g{gFcShd@bMvWbjVp**&N60) zHt|i22Pz3V9=ZzUv&(s`Q?Z{P0(8m76x_&wH&~VTWa*Q1mb;en!2+`s?PyXQp~l8~ zGQRnZcT%ZKwEt%3kwk;go9ys@Ko50nw;r|unq1?Po&|bc5tRIrGUg}b`~J{7GbsDp zYsEjPCPe@?Vo}oxXb3aop@6nAror+1CgPRNhj!bwABaxwQPx!%csinP)=Ba!Gb zD#YrVdSa_MSn_$Hcshn?rj2n!Jn1m<(td?5|8dSBq3~SnK`lJOzU{|0f+X5mB-OJ*zd+hCcA6pc2+n zcLwi-bH?}%>_;4;S^Nsx_6SbwoJeJSruohMPK&{>N|vbTni>a61@DZlADZN-C|!6A zhLsHP5J4zGg())HFOduOxTTq4w{P*%<-;4@0JZ=wSxkm+jUJb2SE=RK&%eC;zO+j6 zVCD9a#m6>FUFm|G=3qP6hq%!dZMp4&W5)H}9X)7gTXyoRq!%A5T|+21`{!idHSvbn zDMaCrl?xzLp;NEn(HoEI-#6>`nKFudKi_Br zQb!~;i+YQBY9Pt1wC|a~U3Qu-zO?Er6!l+^j(H&8G@biH_$3`p)kWc?mqvzQQDc!W zy!Yt3zfX|ArSgDJh#x~azQ%LYv}TF9lACOouZ;A?yP9dKZ*W!i#4&@VF=lMbRPce= z56t?Anrq)wL<@v)+e?)VY2##{b`-afW|VMc{Q#xmsW=Z22M#lTQ|Teq_#c~bz`3hu$I*Jy!Y<( zi9zKY6*9$OdYQ)@H{N37)YbQQ8bcr8^X&DOOx&JuV4CeX9DA+9-x_Uals80r*bewZ zw}iUA;hyHO97WrlA1eZ+-cu}(R?W9Lh+(Eutz6VG;~Uq88XnT(WK3& zm7S1EUb`lnQgLm}%V8+zLQpY`^i8XFd`&ulA5Ju#YKato+M5LG|ARP^@ zj9Y$aW#|gPVKzV|{-?~vpa=Jfay!!&%q()Qbcz3XE4Y3jwy{6I#37-z$i4h+XQIwE zcb+1WET#5nMhhv?;MGx;pJMAqP*T*_nh=M9VTnTRxBmS3;UPc|!;xMOOdTb{hsE9F zJ?zdps<(BPxJNFjY)w)gS}87tKdTj(C&Eg71RcYm(R9=ZlBQZGbDYGU)HVfgM|Eg% zNAa(_b1|btW!}Bj{#npkMp4})Zxj&DQ0(QsSp1IKb7qsg%U?2_rRwymHgFrN)wifjL!uBcxjPBiK}_fE?~OT% zH%W({Mfcp10(`EDeg zXlBu+XEWR#FDp0mE4U_xZU)9E(T$2_S{23sAsLrbP>Tl=bC)r_TIFzfh4ZIu<+m1K zl7nCAaq#`m;1geugcvYIsslUQpUvYZNHMPzFf`gFsVwuVUl6A?s~BNMheCB$di|@n zM!RAQM<@RfHE9Iks;P_dTI+cqF{DTY7Nxy}^7LdG+%!|c>WZ5_Om_c3p6EP{4i7=T z9K@}GI_$miFT%pc2njpVk#q9O0S%((@V9}2F=Xf4O}?S6k{`SjROg^{KJOEF9+oz( zN{ZkoYj|^@G@G_yY)Z{1u}D=>&ig9otb+9Z?g)cNkpQDr#@e#z<)5OlWcsKQ7<*{{ zl1-gfUc+GQ=RDdCF4WBrw0mTRSvr&sOpJ4#mGJaY(>u_!_Do)etdb7MjhL^~8`G)e zZKHr87<;fbM;XoiiSb~r{5967ng8|Xy+`FnGb(ZMZpiwBW~EI#PFvQeo1YT`@&f7@ z0-J|^wyifANo)$~ZGsh+JbGC7M|p*X;*0|(b(H8BTQV^Y6RA3s@GyUm=pQXqEf=Vp z(jj4UW#O4V4A#y^_A#~Fg#*+nXSD5cq6TY~V>O1teMco#TD3zL&qK-?9#f_G&=qi~ zYCJa7h6YPkS6JDI=zsJhgRqlIJdb|5Y&nlUqmh>yJo7(4myvypS`q+LHlxZ*JBIAE z>`u67-$9sJSaM=@_W2J)E%9KJ6Ig}2=kq6DWsLJh?WwfatzvbmSy+p#j0~pS%CT`a z#a--4bz2QMQq8(<3D%EhtrEd=f<8q7VhbO!ZiQ<&x0NS2>H4*D9`7`_XbQwf8kEAb z!{B@vZMjQSH!rcr;Yty=?kXTwydPVAOMNgj4{rE* z2Vrq|@~@{8$FC|}D>*?rjAI`YsT-HrdNM%MxJM81?Z)GPh@ivGia^pbSl;2Mkp z#6a;Oxwye||YaqiO_!J|$*Mek}e8ViDttX1#Y-qJlptj}nX zRqHfk6YWjfqdYhTe>qDSx0@n%Na|?@T$1j++MC;J5_MYscGBkVWNDE|C}fS3_8A!T za3X}u|6M?kpfb`0T3By9{ik1FhYZ%cv3q=Fs()oegvESC<;@hh+Xnj?yLsX^?7hSPlljBhN`J$=b+vKBrr|X}I zPM3EdPVef|)abQirq8nV(v_=ZZ&t)CCZKalM9mEXVa+zC@s~Du-9)kw`Q%cgq7_lT7A(JKLR3K4YV(UcB4J zWEx-Mo~d=2aYz9)P_jIa_u~5(AY}9+A_t);)nj?Z@#)vA~}%owUD%5D+2j->VI^2 zzC@y-%1l>Xq4VpBim^hL1MraLR(OVtwFxk-cAT{iQBhi@N}q1v#D7Ht_IL94KmJ&m z(<@E%)(LgMOnpp{pt?DaRN4|I_2Q-+YCd_eIL0QmN>Z^9CqMYe822AQQeIlKyTv~N z8N1%Ry1JTZs2AyDSHo44bcR;y8GR4a(Eq0)1e_zOD1rCRY8ic))wFVmoVtK^JtPPM zu|Mq@{-mLF4iNb4v>BVY$;kEJh`JM^9`+Kv*%VaG}eU3Q>j$zaDXQli(0dGm}Mc2m?I6K+rFFmtxwteb~W zHQAx!bH$(m0pj@wnNoltT#78nP@8GWL;fJHD*m2t2IxJ ztu}ts+jr7<5dQB=D64?8w*1d*6gOY(TcIKx^8BYC+a6|N(~Ac=La#~T#okz6W)s=& zc>p_|d1veVMv;4reDGOX*4(~XAT3}cm=8>o=?Zo1qtVn7L&`HeuH>tZu|5ai9ij%CAl~A3 zC|(p5-Fl{xv=7*2ebi`7l{oN0A*Zw)z-B2dz`dQP3^~Sd;1q%3ZjIF~DGM@yn`P!3 zwv`v&I0G&E7v#4~l;dV=5ci*jRo5@!el@$8M+*LV$rZYN1-?fDx!)3k>`$KQR#?VitaZkv zy^cNb`n6Kd*3wLNIo;pPr{K|CEox_)p(8_?@tLhjkE&fWS7|I&+S|=OUKax2_!M5NGI^mn!+VkNa7LE}kR95|g;^Xd$&=_zpkMF3#;)lL60+)_=ZmL}~t zk83+@G?~jbmufcX*Hf?7iK5Y%L@T1XeR~)t5!OcYeqMw62THrGQ=QYntQ@AO%#A*S zWEm38R+sq2EeQKK34WADMZ$xEWpO&(o`z};cUdVbAIiG5)sdCiyke8TCM~R#^?{~6 zg5_u6ZZNn!JMrOy#koNoMs>r--5-;<5zh;^9zJWy!TNY+PeQnx3t7R#7#@}$%eOM+ zxWKrG8L|yg^A{K`LQeQ6;+~?nQ$^MoWKFJ0)`m^a&!Sd7?1!q((MrKV0%VF&~ z`Fi|tm{YUk8Rdf4A_@=a5&=~I_==s4L)y<2Rm3FZb3nJBHe5*f{Qn&c*7P%Q*UVRvn(D<0Xzaz0X_Vde7#sPiR_=BJDVP>j8pNE60It<(sJ`n>1OPUJ z($|&>(i1s0y>91m%U;h+4O8=RmV>o$n9?AQBE+8QW_Po%{n+fXs*2}t(b-bQCKO(p z9zog6G26RTett;)H6x(hz&T9@atE;(N;Rc@XAK*N47Z!?$?{Xs@U!yQXJu!nj3((B z$tO#V4s3{H&>jK(AOAb`iZmKoZ>3Rmg(E${jFx$<+opX@E14T)AtFu_<|=PJ?sj>| zePgskJTWTL{^*~nMO3kSamj1RgQZu^*O zg)(E{>k_ZiY){bSw*#)5VLlmKIkqwCDO;~>5u)2?FXwXG+b8OZYBI(zw%?_X--?4wawRiSmq0MtE`ic2-le&% zosnc3xaK`jPK;fV1b_7+`-U!*!qFpQy46FiS>6LhdLOmeT0XXx#3K5RJ>%jupHfiS zpWX*{Pp;&-;>iE9T81Wtg8$~8DB9hJw6cu?$dRH*UrKPx+cR@xdt{Mtkcivg{pRRe z%}8th?aii!f9-h^hVckE@4cl#dCv{zKPE}!(7#fDnvkA;9Ed9(K^PB67JI5??>%uh z&*C|YPQ&(+p`^sJY^B4QqM&47%kpbB#A}0DlAWJJoP3_YGSE>wd(-{WN|h`0liq*l zihLyfMVMVZz_x$(kXrgwpJ(z@1D?<)0^bZCkerg(Zuw;%UihNiyr!l*xFX@BP?g zJlOlZPCU=^oYy%V#|2~4KoItf*`Tu0qq?}0!0n8zAujIT25B4dGN~I{rKKZ%5611r z!{IgkBi!~m96PO^X-;d0sa)2{%8W$-K850;v@P|Wpc&Ld=_#U%vJJH1%C&N!@s(58 z^^T!|(&K8pMLM6t^86kg{bp;x@1@Nay0WxeGhY*C!}ry!S5D^Bw*`vHrk!n4fFv3; zxYmw$T%P;em?VLT#lJnUR|=5Uuen_ZLYiLrzyrv(zzoRc*&Z_yspJ zN+kPph9Nr>PKw(RJ$XITn(SIxcbY@e-@HHb62a_p7mxU`Oc6D_Q|F+UG8D9CpgMcs zAf)hEdOt3s#cCr5H{|G>XF8r0U9f0%jbK@$MS*!ugn{PIsZ(+MMUTu;g7mu&tIPqr7j4}HZAkPA=4&GBG*)f*sk2s%8o12@A${%? z&?LCar;7RcYdU+wK3#J2rcT3#HVKuFwUY(t{g`1%U-JLon%I9?6C^s6Wwyl6s>|nW z9?}4m^iLCTs^)I-iM=>M4MVR&l=8~UzuIS4pY%g>3u)wx`soYEbZ%3|<@~+4h}0W< z_HoVmP>hbs{5{got^Q#&5Q4YnPP*cT?piI;XGDi0-NlRUZ%93kPEMgN34fQ zBU45DS!SKGr$*U8mIUximlPbfgDJh19fV|EXk#*bbpr=hW`LYW9;MB!_ui*tH!$Hm zHz%jZ5RKa`V<9j3JNrUYa(=P-K)uJ<_`~TXiE2qEjukX}$bo^Y(3R%b_V)dibCdlO ziBngxvIUyu@dkhb;wGsw*V?$~=3R(P@OH)8^0uQKtpN~G2we|<+Jmo1DDI@_%V?j+ zWb$&H`cCE1^H(;AvbhU|8BX6X8yYGnqlqoTd^&8Y!x~D}gszwg{joEOTII^B_h+RG zPl=X19IKlal(5PmD>DDA(+U^Hvgt@RMZE{mi>>WSU4io>2+EJh^+L24FU`}0eCPZ%rjWr=!ERhA^AjBb!HiaGPt0EYuhmXluT z?WGny0pk! zyu#VsyUo^q8tYOv+-x>XTR)h4ESr3t8Pz$3FdChg(dCgGeIEdLZ9{>U9de4jn&&lX zl}-12APz=8u&BsbZ#~-FnnH22&js+6mj4jfJQ4WVjiea=T!6>wkZW^ibE><2)^1{8 zLno)oxBj=2AYdX5`4w##?2_tqs9VUhYio5m_~|YD3RhTm>Z)blqnpArTuLve5X4@W zliWI5ejBV2O58bc3rv%CT65{)*u?z9Nny{sgJe(Y;%GF+2i!w%>)_LsulOR@OAerr z{LyBuai&*j_tG8YQ^NAwmY@Zo^o1_};-Lv7Y*l_ytf|N{VMW-Ic($WD<+jhDK$;p-QWxs>h_+~-)?+?g-c5u5EE52gqkl~;GO}Fp4UPVnZHMh2YWv<~CV3un=n0@K6 zN+JzsOX7yux+>m8+IE`X>7IzUP@GMX@wV=`qiB;t2HClxwnI3jmx3pPIE20HcXQCa z1qmE`&Iy6ZxNX0ygMY(y#~(aQ`RqH+$Zn#SG@dO@zh5(s^U>b5xFV0Yf+j0*N@P6z z;M2qePlIC6e%-b*(#ocVYqYj{4hYKF%5ueNPmJNC^_qd3dUu#5Qk!(@wO#F2Bz~3X z7}E<>OX7|vk-nPXJSc4Of_lPA6Vn|RnMLaA>!2s}Emoi1@$n0O#hZ+jjDwZZgoBY$ z6ifa%*{E%z6Bl>=NlEviQw&|nnr6bsot|%k->kHz+`;L(yi(e{90B7RHs!l>$zvBC zG#5Y$Fx5X!nDj;PJRM5-FvT6w*}PK|eGy8_`cw&f`;EuXds9!pzp`d@K_Z1gMbu|J z;~v%82Ga0zLNC1|zpHF&4y|!xfE(P^?y7a>03J41rcjSORexDw$ib@6hP>&&T4R1i z=exB)Qf9#DjMgnTgt$#RETLr1zb8aF18a5OZ%96q@Y~qkjkN~2eev!e21q5|X4#Qd zQhIol;1yU8BN99mcBx%V6SvBt+%84A{6Ytm))PdCCi@?X)MUfS;TpQjz;V0 ztIrSA!UhW90)`u(ZoPk_$b_g@4s`aMfJ-W$HHN(ImT1boMEFo9DPu?3{ae_cua@A|&r`=sG3b@ehW4abAtCJ}a**%h1QW!E)v zcqE!a$8I0eYvtl$rXc97Lcp3UqD=G@(qc$eq?mejF{TIr|GW&Wpxt~6L^0M@CyUyg zM(1KLMAuu@Y~k0Dr_+opI(|7H`DJjL4ckKVhr%WW;~s%R#zdB&Fldh$-!#(K!mMNk zuKiG#ddwG&aVXDI>ex6Tnd3FS1h)s3RM5pRX$kj`O zG^ggJPbK_2KR!L{;=1&_@|rQb%2sLf1s17ux#R0kd9p)qQ%jFbnPx5i7mR;v1U;5F z%64?dRN06cv0ik3KneB7)0?Fo3410Ejwxt9S9Dl;1Tv!;Paoae6x#RJw_Tbu1lPHFVyfu;ekIu-Tqi5i2BvzPieXT{IX z$hDwnn=~*tIC)QWi_%Wfm%Y-FvqNawN>syQgQDl2sg{|;vs@Fqf?NZN6@)MM)aWRO z*Oi>`_Obqh4yvIYdWkpoy>*`P;It{T<<#r>TYM3-w-Y>Sm4PdLpA+{H0SM18ct+>q z(HR+3D;qXum5R?9IkZ41?>ztQvp6>0-h*IMP@VRPH|VrRxHv*;@v6+~;?f-?lAGkw zSKrj7EhDNkn{-bV4Sw3>I?M?a*zyJw+HCi)C+bo0euC2tOYIgV&$e$r^Cstb=-zh5 zH@18xaL_CBi3*nIn8a~r#3F_8-RL{Z*r2ZOt2=)O-aBU4?vE$@Ewgahz%~i<(x#?K z$r=ZOIWNg_nC5v~=EPR_b>Ib|9U!+p?i`7AWGb&S$Pb$c)~6R_*4{j}jU@~pALR>E z6aSefl01>e>@9gYkX3p^+-yrXuQ z$X;EUj?N%w)dhp&Y zUtSpi(2ZxMMgxuy7pCCSmR?Hq!%x^wPZ{nsJp_=-{@!YLSCa`0oBdcfI_q| zb(PCp;xOhV4l6${b*X~&c~3wx9{(sGuR*Ynxj@1ufq<5{5+j#Z({y^msM-p>Wb}Rd zF75Fjx$KsQQw?EGs27VPkF^s62MeWZJ67W#9e-|?JE_T@XT{atpDY4P4SVBS_l$~& zp~fx7PMynEv!5iATpi!$$ckTet73gfHg$C&$)~=;GNkOfEL4 zoETwwkf&Pm@MOIz?%#)>FinIOQ|qtCpPScKRYkj=g)m-fm-7dw)mUpL z5hg5`$*R-8w?P+X86-qBYhrHN^sKW-}D%3WxHvgAC&S>YJN1+;Jrp&BoS)$<{CGPg2PqmqZ=@$nC)t6Qsc(*iQE_wFvi`;&4I#;gdu)IH+Au5Y-$>DWMVu7l z4sUj-0862X4SV)0?WV7mvn~x^npJ(X^4|V#=Zjt-G0vT%>os5!y}C@&C}iD@mCQJs z8P}lB2OlJvCAEPrW2a;D&uS7+h8c3kDw^l<{m6+F_qEymFVs7G%q6*mpmLaT3hSP{ zqF-}-Di9k&y!Z+`<)en|qAsA>yne%#b>?W^g`M8&snU${Np2TR8uo|;urnEU8L-@3 zqJhsER`#6mSZtRBbin0MD@C86nYEKC(xX}7^a~IlN9rlXd3&AST^FwOVyoG0uIo(X zpaX~|6FU<#h}&Os^XFP^N=@Ma1V4zd?Ep|*eia87S!SM2s!F2!nrc*hm5S#hHgn>| zUBiLLMokAF6SSXwvJr0=V?~qDRV&__Txug)&Otd=C4ck0E?bgI2(d79{Yrnl`m890`G)h z$8jLu-!gIQ@J}iiWS-X)7{;yD8Pu(jDYF(D;#o8J7&t7naY!bW75kst=&qM#si;G3 zOIV6?iSd^3wd)q75Qo%SJP2s}jA9x>lR~^aqur2_4-G|+e)MD+y6lJ?i@B?3ENmN$ zmkmAFN|$7IAMnUwftWT(cuk_Ghyk-d@Nc4}C!~3WQQclj2ZlX^&WG44flPdm-Ty8g|ZjLp}YXSvjN z&uMSilhk|h)x+l?wpVmbJ8PbB7d3E^`~+Tu`0EI~Q#-B3^IR2L@rJs(lr3hIy*J!1 zM2R#{L^fU86HF)YB*-x3Y?~3S^tg(PqwZ@?M(y*T$1r2Nc8^QUC4wMfsYM~}82PbL z>BZ}Ha==@)hDo1OC7FP+I_e$EWtXW}-;u+R7>x*v7p1dtugI^fWbDah3>|%iA>W2^%-$>b>@h%WL5^D1gA27c-5R+cOWUOWTiJLIxry z|ItYETAZomP*<(XD_@60IINpS))V0*iuXc@K6dXeK=4PsmX|?>iEXFE;o% zi>_rt$;6UE1aT~G3zK@O4lo%5>$Lh-DV7hWs7+oB~A z^SC3L%TcA>266sTrf(wd!+F6CZD5h`tz()l0Whqit$zs#bhjRkmtzapmN1`zpn{Hb zx9kyY8C1N<=A)*1FOy9Cs+2~)NXgNzY(6^nvJs<^Jze=AGhI`*D4yCw#n241AD}J4 zN;DjCop0ZGlt?1wS}1wkB4db(6xS|>5_uCe<*mpOyo9wY$gqF>)e!I&(||+a8pxJ zxp#-yJ6gW<$r5TYAkDLlPw<8} zKsG~;fV-JRS)(VKhN7sW8a^NH^M=LC)M2z!%=@293Th4|t*^v;t8{d9B3dh7n0Oc6+NjAKmhp*s(Nmlo#vtQsh+%{c90JTTY<8xbBvx*yZOZj%2G|be_^YI( z^SWFwIi+Ms+KnKp4|QA8K?m~-y0$LVmvb}NyfQOd$FFbI-n0qQS_8yxKLQi(`UjV;fLR|t35=~l_58T|1 zjYgld@7!p&vtEgI30yt2mT)&d^nC4a@PV)u*wJrx^HJTC9JaF7x)83qZ`wDx+7G($ z2mOZyH`_MC$x_z__bWcv*16uewo{@p18zSk=`ll2dQhk#x7r3GZUbT$adgtINYZ52+0Dz^_1j2Gc)vcH zYN`eDqnM(WxCA|2z#te^5~B+(xa0%6WWavsILp}Lsfz*L3qMEAhjV z?gTyK&1U$d`qBBbtUJ8PqH1l2QA!FSGuwfdM%KC^SI48B}eT@p8?FJM#j9 zkIzlUM?;kVNDaTp=9SkW1?is#{P+y?k4`T~d1rz4oSVFMHeX3xjnGh*sij)I6^6vY z5ofT`R#fW6jZ}i)vjD14hFc6k;DV?4=lEwx{5FeC{W=}qfHn(u36SUbkaOMPXIhwv z(=hR-=<(-EM~fn9Au$Ps`Xk3An4h#@z(8}zgAB&(0@UWpwpO&t;tcTy?;=}3O5M=R z)KQO-{pxVL z*{ow`PQ&!wZ!EKMhoP_bYwJn%E*IvydM^sJ*LXbQ+`mHjvNEV*_4Nj%#-_yoUSUqo z00P}g%@Ew&jOQT4KghFi6lh|cem_c}R_Xc>fu?1%xA>+?{*0WfsJwhM+1s#dv|Fny zI@CIr+oFt`@GCJ(KYy^8f&%WV1!I^jwTDHXa zMq1RFAsh{{2M`&0d&eBtN8@CFHNVQ6oyD_&x0{shJ>pc2L?+3gr46Ahn_^`xEqcW! zQDx}7oQGaRc1qEi@uk7CBA~WHzw(h-+~MR{7WcvRxCCpnBk$~u`7_Q9+WSZHv{Z=+ zlt3I);v}<}7ojwO{#SL>XJ6CFKSq?gzbR&}9P`y2C+fPsnu(k8zDBm8kAGA%cx58Q zc)B=iP(+Y5mzNqE8AbBtVCQky>Rw${EulN*8CIAyL|pv*I(ix*B-NfJyrrAwd23Lt zctEI?p7Tp}adG$L7cmKmp6w(F^T`&;A=r?+Vc40(OW26H@Xv{Z4T7YvET9w7Wd9w{ z+8^i*0HQSM+~8LJ=ti*1Ji8s@JkQgx*!bvL(rVTBG#1Fx)Wy|1Hi54RJV8xeEG!1f zR%;z!UsggX29N`cSF+6{44*O?ufIorOyN{1vRfmuD5f)w<>X?#YG;v<;!1X0cFtI^ zt~HhzOp_STjD#)%QS@#J5)8X?ChzHU$>&R=M{j1yVyZ?aHc>uhE%6BnuiiR;1M(%y z>ONmc-~|w!`MR5UwG|tV9ZOFd$0a5uozc-tjJ&nOr}vYZnV#TISY0j~eNg+=q_r?Q zniz0kJl^5Vl7h||`{bp9yp5%`I&Y7nbfETlUk6CGh$rXc92jhhutKPcTXXqlQ+6#qs<;y z1mEO5<-fIc`4O*Xa$+h49WUvS)t%xV+m6mh$sh*26_DOSd+H`Xo8+EsT@Cu=6&{O5 zYt-zj-2}eO&ssdRo}ls^J?#13?>0(Iqa=X~bdH!DE^9~4#CpM0xdYmKyFFTGtTLY` zCQ=wIGT$L$Rd^YM9_c*?OJIK`a?N5WrQ5<~9`xE4mG{tZXkab92MIP$~;+oOUx){46HBW7r(}Mr`fZqXJ@;oq>u7 zQ~$or*G)UmQOb$eS4Uzqz?^HBFOUhC^I#HV+kU%>ieJnHG*85CZjy(`C@TC8605GgC5il%;lD9%+U0gEeRA=}s^md})l%qLWrwx~6gEDYH z)&)r>{Y4lMbafVy6)FLJ*N@B!tYXPQA9jIcXhtuVr4G$57HY$HzcI^Pxyc{Kpa!x| zH9A<1 zvzldEc+i>cB*h)ux5^wF+)0-ssjNLDulN)(qXb;_xHL!+4r~UJ75oFTnRsz(Lx}E# zk7c>C?{rwM9Toi;68Uie4A`1`>1q`iST5PA9H8nkw+W1;bX19(`7RQV$6_8 z`jskg5D1^>a4mG|gV`5Y<>zu4+n#TGp>RwmF3mZ)`dF-#oPkWTlcF2_aj){QEA9c_ zc(^{&L3l(L^qS*uJ8%+0)`Uq!(>$!W<0rQvWZQU=683Ls`P7L0qj_p-bYdsZQy%wF z&cU(#IAVYPO|%o);9_hW-*n4$wtDI_>I4&_`au5F$X|Ll(?HB#QcNh$0p3 zTIW{F;r_T#SjpV24pMPb0K#Cij+oRRb2igJ2?O>p8g*wS+sqe#C7a4RiKg6oH zp0yMqQk57V^uBt;2v_<|#9@RSyiWMtBq)E(xWy|$@@!5Y#zr$q za(voINpiN)w`B6Ma?&I%L@kkfLgns1tO!$(EGaewa*5@Bxb__t?)zk0yyP*C__z-1 z9FD#hg#-;hy7)-#O~pjqPX{d2MRwtUvTCZ|7C3Kz-jocp3BJHm;U8w~ehapQH?qbn zaCbZ9rR%El-PMA)#Q1ni2O%E>!C|v$2DL$Xaq|=n#9}Jaj@hcqA4gGRskbW58YlYj zfm8dp$Vgh%n>C+AhWh_NGEDrO*a1IxjooyKIux0CD~CPkD9dg@!;!aX2@ILs{p#bj zvL4=XnNS^<5Dq!maNY&a%48&zAP^spl5g~DNhwv&?`xJFC6m2Gp3mwUSRORn8YcRC zO-OxQ4c4I-bQr=Bn{$0RE9@-s^hHq!(GabLX?9uJ0?j4NEWAzHy9fKKeA&1>KYtRr z!UpU}NIsCvTbQd)oWFW^V|w5+4v&3;wGDzwCk32-{{TGpW14T1B;b zE9-TMe>~ec_Io`_wUYrlyJ>A6diGg7%4{dC8skL!%_la9Uii&2hC>;2_7lJ6=za)D zK*PDQwu8Ato2s;guPU)ZRDMXb-HTsQ#)ZCrDKZruEo-BgIP)wno~Q9!-B)SG;{izYb1tAx z`en=|3c~B#Twpz`nQhQJR+3ianv+mVZlZ;W# zvG3Lx>n*8!lg}B11~UNAp+LMo~7ieIi{r@^EhjXC#x-$|X=2lzu?&2ZQU$Q=IzK0wmqd$l%jWJCA=m}%2`p(TcE*=iBkxR2Q zIPIa!Z`S%Rm1T0oenTJdxUE*c?PR)h%u2Crf(dFkzUAj@@F6~iFCqSs&>HdFb0+5B zf5jAv3yuR@ZT!R)Fn;>+-ksn~N1gE`peCD!TjWd2DDxRa)tPAzy<%1c8MiRQm^)?6 z@g?xb5R=RKo|Nf%5~JF`Vztm(8naSc?Mf0=QBo?DyoxBaJnM>B&B1lIGJX|oX=A18 zu2$j;H{_F$nd8dZJ{|CIhB5c8og;&R&T)j4%O>Wt8yUoV{GJE$DWWz7TLEmO;R?Y` zH8$Z_pZ?<)|9+~(H8h$lMbMbHGaDbmJhaoZ zIPp^lKE+3S*y`0to-u~4ZnkhrVspp&ahB!NG;hpLf9Fw<|F-2*bZ(xf!h-{uyeC@S?t`nAavir}o z3magY>4jZ5k?NU3Pu$$Am5jr1|9)3x-C*dzmjn#nt$uTu_N(&^9*AV#viWkEAz7te zZqQCrXGugz2DHL6m_-0Kj0 zN*p{qHQE8cwry#>{^4bZT>sR6uzxVw!$2g-JoUGyW_pXQXIk|%k=u}It{(Xbf={fb z<6cu+Tbsgjq7oRGNfE^r|N5}Ta$iBvsXvb(@m=X}7Z%)0=oqjiHFMI(TU6rW>Lf3b zkS6U}Js$}5>x*!`X#rFGI~as>1v^ZYf~Q0vN+KcStJ+j#diXKl9+Ex?#?-#R!V)md z4r!yWi9>^>+*>5o7o}eTy$1tTB7N%up6N{+yIn0@d7sf=2rE6>yC|?@DQw+w^w;DY zZyLtjfD*WHfvs1f{slW2Nm^)ORwz_?56oEnHgSw&z`obZ2T|0lpOGpdE_FUWKvEw~ zr)`%LZmUuIw~S_FJm{062T$E~C6nZ$u8~(|xO65>sqEJ$|599(kaf9fN6tMaZVM}M z*i>IyRRq}U71{==d{I&QHhtBnDdxZZF8#5uA!a6-v4Iz=vCzat@bCT@+1mswh}Hht zlJd&H+Q*Oroe1f}IPmuFV`jXMDrCQHIyf6gq_x@rerZ0J?a61n4h|>?`OpK}kEf18 z2XgRdsQqL8u0#VFQk`dDJIs4^+*R(=jLo5ivbrml3po~pce{C$1`fuAM%!#<$_{T= zZ%fYbz8}}lf8f64YowJgbd(&+ZQ>vpENh8d3((gvFi0T-?K-%qH6*Fw~p-D)c;fAuuL%y)8mv7 zl?MNWiO9Z_7n&+6v0N{C4^zV0T{4B3E!*l>oj_DJVt=3GYi%Zj3Pd{0L%=|l=>6{# zkL2Dv=;kS3>K_9?#Be|Mm|oxFiHS*Q&zy8ZOg%E*5O8_Qmj!n5KO^fK5|r}$qf$e< zC&vWTEVmr)jTI-ioT*-r(mo%1jcIBH$2drCJ+0AfPzfRIpNOB(c@VWwASk3FQO!Cq z_Vh=4IMH;j>RKupD-Z}=-5XujeqrL^`P9f*m&LDLSGr9X+&Qw53E^3IuwaUr0!b$o zs7gaa!vX+|JC;n(zBDbjnF|TW{v`INS72HSyMdJ$$W%q}=cl*tt-D9Qv>V>hc%z4X za`q*ts=H;P+z6s=7Fex&v2)F_;8~}A{e3vEuudbdomnG>b)56S@?!BKBrh=%<&#B4 zaGQ1McDXTVfd=+k?cPw*em?8La@MtCb91vLG!5DRaaE0V62}lPjp)|FuMk*1>1jED zsbn^p{@Vun+mqhF9woCC7*5=d<#Jcgh5sU|Be}RMMWm~$joDt7C*4}}?T9u3=C`#y zuMenJtF>VJ4}Rak4!o7erZRfI50;Y9+HBBOft_hOOuk>%?iv82N*10ZQd9Y{6kffp zjaEy_gKOJu_6r513j-aC_qVAxgy@5Ghf;UxPHZD&*IcGl%KnEHZkU9eLsgyRp# zKq_lafs5DR$k6qz@|R)#YX{+s!6&r?Pu_bHdrc&*NIc`2>FG`EvoZkJnYXQ!nexPe zjsR8m^z==I36XJN7mx|t#T<&T+@q0yBTII)@u`9L&Wn(H<&XH{>hw&lAr+Jnz!Qf1 zwdzbR>il&N{HYdO8T9%f1 zMlJ!Z0b%a<#(#eoV+f}_pTwk$I2qD7-|HndPNnE2bLz$CH<@n9ztRBRD9{u&Oa9U> zshvt#-ns`t_SJ)~ZrM3Fqmp}W;XEa?bwFNA+U;S6Sn6= z*=hxd4^#Z1>@Ua*FgHs55DiXA#Oz7Jbg0CukQ;xEhd=^ehCN=q{<@55xZg={r?c`eGBZe@dCBcQ4`EnfT@|~zzoU99fEyFiZgQvhSf3C#s>bm zObP*J0ye8WWiz#gElBNt0$2i+z9mdffb+1@9PX&MnvL>19Bjo9ezBH38jNY>(z3VTTAlu&Bj|as6uK*|x2EyO z;kV)(U~+=BZDYH6ZRVXYBTKI*#d*p0qoSDjZ<#W+z~#%S_)>@2>Zed7h(krEFB|tKfW_LtwmB;FdI#HGZCZ97OVhS5#wV2j9wF>7%E$`hb z_}uvvTF_z5gNtdb?%54;s=+Mguo$|U-KyG~#$^$nfn%6D@*_US=|_V+Wn-@uCoYz6 z@j(%ct?WTOm5=`P6ilSlSc;XAkzD_!P6SM!_OKPWTxe=z2E_`Vj%Z4A;-b&x!Rj~i zy*D`qQ2x-KM@gBt_K2{qan}Yjgjc@hcPVHaYFiGs&_-Dz;A!5i20>Lz|!i32GP)i)DQr14PsGs0ovC>jNkaVwz-Dg*vF_Mf`uL*RV399fp8QOHdxm0fA$ zv2pxG=R8c80JT_$geR2`o<3V)p-PO0A78fv&fB}mLZ4qxAG2wd#$$}dA*G<;Uf)nZ zjhCeXLcRoPFO%jpvpAk4-Zw_ zDmP-5ke=hB;`QijI-HzBdg$0NI?^};6_$yteudsd(Z!l-nKuWuwoCak8;sXyisE;aS=MDnCMNVA48f(zg5TZ{vQ)?$i;VSr<|q*+oB4)W>U1*9Yk$kgfLtI~8BJwX z&+^b{`h;WbYFQgC@s`9Wn4BB)xwl7*8h_Q^uBOquiVzw#)o(^iI8dIkED*b%on6r( z>$sx1YlMydYt5zYnc`$Fot%5?-`?K>C&N_)TFlO$vH7i?Q&eBSdTwK5Bj1o_$q@Js z*zV`&*9o@K)3aT3a@xs}2`(5KlD`bZX>aQ^btFt+@btAH%^)&6F0Vo#$b*N3(jkm?3^6##E)_Y82ls~ymk+DD2&N77^Mf-DgFk!4^ZA)ybL)Q`Ym32dpt z(U+jY*Z>vRF{wLKFBJZ01#OhHQ_)PgyKyl8PVctyMjzhrS?g1#r`QH5DcS;U>d|TD zN*ag~2g^K7M>&rE368kI%}cCAL`e?$zmam(sOm9Vk1AI3FxDdz#5Zn>j%Sf04k|FXu8lri zvZkexySKkK%e?!4OikV>yNm2mO2R5}_R)*zqgo*q4UO2{U4}P!*5`&I>Bu)(44N7T z|9XShMohk{sVrKWafFNwea$g(E1m#we)Q`Z!*co{k#k=EMcJ@ADzFu7ji-8#hUNh} z^yh*r5d~F@JFfoJ#FTI|T2=5Lx~E!!Qb?iF3ef z${avThm|-Kiya4P)}IUzOb8YoFk@$2qlkz;onLU%HLX#$0lh7%yG%9Ej8!_~vDY1;35t$nO{kuU=J)3U4(rfaS z55M!3zbG>t7TKL=vc|Dy)3<-h{c++qu=udb;~=s=8x5GAb3{0_ULvt?Xo%s>nU67( zs3<#km-V}-N`5fm|AhV^SsYw9I@HOPBvoaSike!4p(`tDgY4qQC*j}UBdx$Lcq+s( z`A#R5SOyH%Vd3afFJx8Q%=6uER*ILv!|~dl#XxAyxNYthF3ZLDvuzd)m>lOGC1wA! zjLD;uV*_dy^p_6UF{V9@2Qwh+1QHwU-2(Jm!Q4i4{3r+qHJ^y+NC1dJv zyPrOqqVR5Yy|L!SpR!k$4?a1GCi8}hRVz3F-{cPbr+j}s1>brc zLN#~c!#gqscXndoQ_mnj(|Nr+e);_MzkTuMrG)DQi3MUdV8T*$6@ZL5*y`f|#BE1E z#2nPg;=11Z#&<0Sl=$|-|JNlb=sVR#W`#@%Gaf!j^Gw}j$;&bi4GDG(EYvggbt*ZF z{5W5`OA{m8FHP=q)*{o&q!g1857je%xK&~1ApLC(u$|xs1X^`+rR6sl45jL5OKRR5 z$XKnT;QJr@3SPfOYEB(+NLD|+*crncpP9+lwT1_#FbJNV#Y}_x44qYen6lLweu!|u z+dF4T^z$n&bK+m4Dy`1X&sSAZd2UH^AiFEP_ElX^ui=Zg?t#$gMRRTUfw!80@=hjwAJcW{e*->gmS6_dk^yw(DCIjkQd^ONqv!F-ATBx($?HIN4)bq>x_Ypb_RCAuTLS}XZGHVuuIOIks5`-! zQgl&;OSP9s-x0B?sj0Y#S2{#pIoKUhtdNyuH-=C@Z~gG@FEf&D#7WO!7n68#kkCFw zc5-wi!xh0bf+spx|Gps zzhc&d%^LZ;*gq&vSAdK0Z?+pj$p>?3sJ$jNJ}xR1Np7OiCU+k;_`%>_#~L@+W*n) zu0~w6bd%#o&J$J4Sg2K4y)vfjOBk7;VI+mk?MeIM#pRA4hWAOMd+Z@^J?QG9wPp8f zCza8npVm<3mdqks+KmAqviZf>n7XXMzui@Tnc00SENpJ2vm3YJtw0<#r$-7OTZ7mE zYq3>J~{qZcee4f$ZkP>K^uQgn~u)kV$ zVQrc*L2mf36?S5?rDUsCVRP%`yBL2|_g$%wn|tICV;6i3E2|REfY+`%l`!{1&?FzO zZYNl4W<6|JtcD{TchU>$Ch3{x=jS<1*lYzgWNO%=0Kpq(0oxCJ*P5tSS2k+2E$|>C zSt#lGT|j$eq({?n!Isrih4q{FPv>vA{4c=3BEr}-)BOnc%pgGT+its2v9_uux?5HR z#uBHYHkl-rHy7P1DfKdkcJaRlCj|*#P`OF~&(T>o)$rIzZw*yF!vjscl}b8|^JaoR zhW=3zB1U+LTqQMM!U7tm`jvhL0Ou#9ejUDm?dkf@;SY6`4$-R0|uVw|5QcfY*jWX6v{ho_GV<@(}0dvZG90r+9JD<#cdFk?L6J?M7IIEKWt z;J#ZmOu*6@gmxhSb&_8^iz!-gUnGevTClbuSC*CO+&y&1-nrlMSABu>CU!onf;NT| z9LlIj#MHi3jDTU7O3*oXiZq~d{9DA@Ox%UyjbN-`Wl0!5Rd6g$I=yLhEjLCJZUp;~ z>sPK3*_nA&(OvwIQTZW2zP@(5QeBWNk*%ZJ&MxEK)D@K? zTF<(-1_HDbFxjRWrq^`CM`W=XbCV5vT=K~~d6b9Gpt;x3>)^X;*4DaCdxQp4!Ie-$ zt*WO+hLEtRnu|-;%*zk)35m+Zcz5IeAZ;)|=RP@8hhR5xxXl~eN2?oOOJs2*X`>Dd zv?UOLK3Q%hcZ1>O`~E8F30a?CD&r!15&Wmj_j^+xVz^w+>{26;VmRNKw!tMfYW_Od+01{tXLYMbZ@j5hiMJUA|_LXs{* z00W@}IF>GG@1EAlsqfx>PZNjz89p4NBIG3=iu;(cXYcE0TMS;1eM+z{brq^7lJ@+| zxn&Gt&cx`8RvkUPBsxj|2|pmFg;c$i8)f?LcjG00D30V7X~r-`4=HvF4P|iYJh#CS z>YDdOkEXJbQp@ybYpG=-7DUChDw{@WxT@H6?=<;{-w>vZGNn}sl+iWMuek7dr>aD3WYdaD?W*iyqWfAY`-bq)!Gx^%rOcJ_w5_7Td71;i;z0TF~>mgR) zo>(Z?GaWTj=ewq8)w(rdO3H}rsLzW174-GtluG9XiI9~5($$+m< zK_HeE6vJNAZH!-&<-Ya!Iuj>whZG7_oWUG-@F}HhW=sZR<8*;pHTqYu#gxo3TblSo zu8%}tfL7~xw2ky+z61?J_Nt8=N30<2+dsQP)(-Ko34Ct3F0AoMNi_-oPmbs>I4rM&A!Bljc3}1CB>KBG7@1If- z*pM2AE2ykbz9C-O6CWlFE=52}C9=#fLdul)=IWMfd;2v57`PdLcdQi+_ zNJ+ShgM&N7+0H4UVG-SXY{<@*94kwJ;nAmz1db9uGhR%zH^W?rU-?j&nXpySq-s5wZjEacc)DL$AeSF7Rh~o< z6PF)B#q#hb`kibtZu+o8XS(pruhOdkdxsd`8nMc^d%b4{%PD5x7_7Z+|C_S>l{qo3 z;Sy9vHpzXLi?e0CYa<682@%RMIX_*@9G5u&E(elQ^YzTu>{~v1I^?dlbcO>=m&{nW z`oDo|X=D3@ggdDtiTDIaeVXX8GP}2(WFzu9J?W_d%&6l!z+K#DH^Mu!l{gFIK;IoU zhE=>n4u--^VIZxIdydl!k8^rFPhb# zSsNB3c8)&i*pvooz}}jY(_|N6WenABR`Jb$_68@S)7WY>p7 z7f2rFwwQqW*MQW+c-BK<#cfG3LoLodvGs_J?Q1{gRp`YJocW>5tIo2{)@L}s#QJ$< zW4JkU?${>kyD{5+e4A{`*TpQGrd|DuU*(TA<}Wc@;HqgKI5?)FXJPsZ3h&*Z zQLqnq;=~Dpf@4yORr=P79Y@6FboM5gUmiS7TG@VB^_)iauX^JMA1MTF^Zl#-1~r0e{!IJOD-y`Rb_qhF4+&OW+TCZ&ha*M-s_{JU5iVvqwv~)uZym)b?gvrV>PQ=$7SctELj>g7t)VJP$}g*hAu^I)_vs3E27)MCO^H83+GD0IJu z5tZKN1iXBDBltP_67uZJXw46Pg2yuLx!#O-L+Kyf!r{h(w0WEUKla`;s>ycS8dj=; zhz${t;-gYTid5++3eu%_f+C$zq!WsWf(n9)bOI>Sd+$L&iX?!cg&vR+AcO!R2_%&7 zX7A^D&OZBl$M?SHjQ7X)HwKd}_R86Og{_ygS)IIG5!_5ice2vuVl zdZQOKp;bYHEIr{HUiMPNb8*Z^IfKPfxU$8weVrBiBT}`+_?tRe@gl3nKy_@U@|TR5 zX)FU_;*=xz%*8=0-N5Ti4OzqTjDd{>p2E_PV#nKA^*3|ip_2aofdlJB?eBYgETlA< z$_L|I|H>u2DSJT(0{hk@#{LOd9$BBnpv~6QE|Tw7JN@g844jyiqN*<(5Zu==kryz& z2sWt*rdNa&9>mI?g2t2B?3wrBm1XkP2Lo#~$<^$p~E@#fA^}^-gw||EeaIVy; zS(s^nnSCJ#aEN@gY1EvS_+`{{AR&o=7_+ioJuEkl*Yefkw6gMUdTm(si*S0IPN2Pv zy>B+I%MUuRL6|cCY?t*fqy})eV`{^m;y56Gt5nka^O5B1QCa<&tNr)6niR8 z-B>TY^FiyiT^-wgW@TReiNo1HJ_fz_d1igkFMTB(BiE0{ow1F>Mz`~~{PQtX6;Hjw zz;5x7K1&?$$jaO2dI`Dav#fnlf2~@q$jQu@_|gTXUSm|N6DVT*R^?skp$|`u>I$z;_rMf9_g@v9YnmYBj!#iTR^x=1~d;P+=avO6z{l zjGHzzHN9$lUQ6o&fy#DC{j;goh>xYK>w4z!yBIZU7(EsFrpLz;NSRjj3xsxjEHxiL zyi#hS?Ej!8zHrETWqkJwt! z?)-dl;y;4nfgzu@U*gBx*MkHyRkgI%$C&vuWgv@}1jXX}Wn=~O$H9f$aT zZjQexy5Gm3Yk|;U;QGH@h=2R^|K{zfdw}NV+@?9(f4(&Sfk6JNFTaiWq?KcI3;!kR z{(k8H&0`#Y?*91(OeW;M1=LUN(sz{y-_M!#{C~XCSwqj?zPOoPrTuR^?7#C?yahP@ zJE`LA{})!_!ed}X=@bjr|5Zit-+8FxbK{H{kWp_cBOd-&zl;C(dPmS*0@A>fthbu~ z0qKSYWDu}}VsKYib%p))^=@3j8E zus0s50z}>agxLT29{!&Y`>&oK{}W>W=h5|#wf+Bdh#6k=${N2(w|`*;hfS!)6n(3# z)C5S0+;0t!w6x#d5E1DwC)x*nVW`$0m98C{9^wyblryw87JU7h{!_;1OYknel&a%~ z(bQ_pO))Vue&g8WkrAZQD2ZbtDoR}rWMUsW1{4WP2O(Nj;de-YvgrBRj)f+mz0|xY z4CKd>B|SyegqK%Ug0wy*7#+L2XzIVf+nGn`I0EhZy?~@-Xk^5ln^7sMYk}!~YIx7D zX2r^9JR1ZP6Ti~E5;N`BdmD+D-V^0?ve2=;r9MU0_lADBirsu)4;uH-r_o#q zbh9c*Dqd{rI2-x>@H+~<^|7yvtQ#IO?|^!`v;be+W=n3cjh8}>j@}f5wP;H5>L#)+ zRXyJuI%MBhair@wC=nWWO92#)eK6W+>kTuI-)lb`Y_R>Sx6mwHqMAlrV^Wi2|1F{3 zL7!H~r|zeIbR5WImMga`?Efn5@g33!2Sb*2QhkP=u3Zw)PvdKpo7ieIT5m4^atTSc zf#4E@hffx~Z=ORCel|?D-?23Z`u2AjU(Eb#LlY2%kIW;G?q;8UQgRW-D&YRM=#pu9{f?AB;yYu;K)mn@yh2z3s#}agV3Ax(}B>%^tpyW zZH+kS2f*^n2pH`yS+F9P_lH~m8hjCGMYu6a*VlD_==<0aara``Aa&&D#ibZlG<)4w za1g;}f-(jP!;o9UcZD^XBu>>f)#=$z?dM&U?#V7~?X1;hc~LAVmgt5mO9I-ln`gckDrBCi3Rs7uB1N{iO+KURUd%}>D2#GYQB zoQluoC#IQOh#qU1C_z`Sg{7@CSKDMu+E@TpQ;^rn9iSIP0vT0x$gtJ@Dd_H?8M}@> ziCxD1^8sW0BP*+BxF#)v9g%fiwyfE5pGDBWbY0rO#z^4kU|8S*?Fs2pX^hdqVeIF9 z>7%`bwuA)%z4xD?(&6u1GB#JPlUbYBy{2ruSG%g5@DU#d%`6*dP{rt(L%e5r#3uyV zJUoQ)=&T=X=*26itZ+?#{EQfFKCN)`vFn<=f3aEV2P`7Uw?JxSudlY{?Jpfo2?vQb zffT#w`2jOZfc=yJ&|n-|SSZBbAU7(HfBhWBxzX^08YE^qX!h29gA(!&v8K#2p5PlR2?cuA;7DRJSjOW82J9ws7PdoajPY=xWM zzW#lvn}fE^ek}~@L%6d0qcY}CTz5`4!V3jm3+-g)vQ500SKU!o=7B16&hJi6rM_Xy zvrUuor|Uq@-n7mtbROZ3gI?#NLO9ob9+ncQvu&a1_3Y|0pOpifYS-KnX+J_|?uShM zv&nY|Pw&5={7Sid<`T`nU3L*~PJatoNDtpqQV8H)jlN`al11;AYFqtm5D4(F;~)Ou zRbEq9m)iX*bZpS1*sNPe+}hXSkLudE0lhuq@Zis$sQ}CC9eb1S7?2@=QdWBMRowm+ zR!!P4nY1|=1{7k#*?%fvgJE;>FE0StCx-H?ioV9AG=noqqK&QNi|voOxhcQ3ySH)JA*~D1I_YzN%9!`~c$is#o-))S3?EF%0labGWM35G2_}YSKv?ykpL$w`5vQ>Z?mzOXY%N+eRvMSsU zLFa>;Y}HRnnbX1Z!INb@?y!9YEP)f!yu+spO>f_@IQfl{mHXp$4gD0m9MNI~=Xc2J z%|o+KVUBVA!-Iz|k3Uo=X{&TvisI%?xgz35eP?4^$kiYvk?8k6`ucc2hFbZs-!%@o zIr|Qd&(2tO@2QuhL{md@Th$KNWCv4_7H1*Gvsf4wW26C63V>-{zLN9h%EU&h* zCM*jW31)*v78>zevJgW>JlojcG%{RVY3x7+OjDH(L_t-)A6rxJwMC;n6428zQtJB)JDL$BHxQ;6C>rl&sq7YxHWRav%kb$)hB$=;8K#OzHyl~I959{j zeZW$p1iQHZnu6qLm6T_8-%|eH0pb+$^~w< z96ZsEhw^>bFDV*5oKrn!1^d*VUfh`t=uwlRQj*b z3I+JklJBJrz)gnBgKhcd8`Xv^>DLRK>a>e-scy~+($cc5K`icK_D5p|PP2hjT52v4 z(p8oF$xXiE+RZxElG$<}554t4ZU{S-W;(k2@Xps)z$R1WqnR&JC_Qd8Ib5qZJt$CC z3*g6{@i}Vm93$)s`4+5w(`;DvstY)ca=OLo7kxT_Z^0|NW6Qd-HOFXHoV$FIm~#DK zj4-j|ViYyP)I<1rP132mfEi#YTh+dYI`5QB{96wB;tfIhGa!H6yNGn5f!c-y4GShe zb$5q#$3|@}8VxoD&jF6TQP$oQ&xOTxBfSR9Oqw};{rzn}U=DXag^<%hpJHQoQWz;S zz(_mw19`i|YJg*obqxII%!j6vypSDd-GHX!S$!1I1bo-72w3ZFz(xmKu#rr9Mj=6X z)oad7LNcI9X?z_!HS-#gM7CY$a+MKJmNm)&;E8lkp(!@3tOmL9?Q|mlxyl(SXGI}A z?5MIP?cEPr{_n(Fl4~~Vl)1r?-Kq<15bsgD37q3IEnWGcwSBqj5*JLk{Drb4-QgXV zgr&2RmfFaM!xvAupAZ0xO{*Yh+BufxtG;Az*=AGJ3uL|Sbtm0Mey{Dibkxr?Mg|pA zbe)Iudn~l@i3n|4G1}eSd$q^it`2!Q9V-P=gk?m{%j6xOl8E}I_%=#^wc}?P^??SK z%S}@lI{Ii(S4imRa&_eh(`{jiduaLX4Nby95V00eZqsZNudJ=SXI1e+eb52G(f}dR zbo}V|Tl8B7LAN11dPz7L{^RtL>1JV^X-WHw$&|zTK}9jv8I>C3hd{^fDgIsLY_}I& z4nyx+cKxc$P@5dk%@N=z?c7x0Je^HmmyMd*7?bVqjm_GqVy}f+C8rpq(#U}d8~yLw z<#)UcDZB#WmusBhSFL$CkWnfT9zfBE<}phV8_v_=B!%RDi3diDOPXRXq~(XrmbWe6 zIFv%f=!=9y19rxRe8K`d)5^=GCO- z`@}G~!l>;s!XOsw;g8YH%@9Oy8W&=(FL-Jnn49OA#oK~EE8IJcCX9OZ+?*YRR0ad% zntZ6ky~iz_eCO%v(%llsD$_jMWxEswWgm($5_5vrlWMFi0lmHIM_6bJ(SmI)Y1~i{ zhL?lGnh25>EnbU!nyqAoz}*y3P}Sq7r(Xe=-M6wQQoZJQ7k=G>D_bX41gpMP3Y`-p zV>8lwmN5yff3`M9Y@T{myirpIyB9`@$uFVki%cA=9J*qbt*)cpLRZv>Pk;1(TmU_v zn}Q*ll~`ZSM9lI#_l_thleiVx${GnWOg4zf*g<1}loKPp9C71fmjt43*k z3|5vKAq4{Z&XH(!RC?*ZGZNyK0KE{GwEx}T$jI;^ZCi#5YU3d9UK>U*a9~e|Q6T$c zmpmmU?^xDyD|kHv817Zd#0!40Cw%c$P_I<%2wz^m6*}qIp+@X-bz6D!vX*4ujCZPQ zFrpbtO4qSR>X=#h88;8X-<2R$cx2gZEWB{K#JWQWqk@ z60jz&baqlRlCD+5r3OP!wbCdIJRB@H&r9o>6c($dd7iPiddI?*#+DdC(Z!Z;aab3^ zv7MBUJmXV)Fu+4qNhn~t$yTbMH?=1J@xi()r)=4GrWPND8VD>dM=V&A18s<$8u?SFc>2NJf0~FkOS5KMvyJ{c&Y5 z!(w=;?GQCs%BRheH_CUP#h94*^rvD5MtH_kVbdFM(AoapTHmZ5nyKWI#->Wn$x;$B zS>7+tA{sf~NP5e^srcbr>ypRh&{1d_VDd#Iw`UP)rRz0xM!v&X&C7#yQx$(QrFpu$*S!>`l>@Q(S^KKc{M1N}uf0@^DSjXm zUpshW=~z5_|9pz`%hrs+2SCG%j?*7bw0+MD+7{B@+ER5(ti{-M7)6hLH-iZ}cE`vx zlW})4C3GfWrS}dt*h(RjLv+>WYU{=?Z1K}!j;Od3k2tw#=EKG=ZMa#(cJ$;dDE9@q~pSF%~AHVx)J(uYR<)G;xb#WSj`I{6zWebqi`S{cUrdl5hT}rGK z77F;1KXQdG#D^bp6|0Skc*2B*lf;Mx+7-!50sG9aid(fCl#P%D3b98yNmn4};xO;z zU*Hix7oPdmP!E?`rF83OG7vEma|a=k~dALbTOn%f4{YN!^z!yT^C!+s8{e6R8su z5PNbLtbxnhej^5Q-Cx$x-`?N!V%n1%4A)+-r~2pvT!>BfiD;BXwP>uPIN$-u3zNKx6&v)qX5(##jk@ zg@Y&Gb0bt6d3M z;6-tMmJ|Ue%|E4&6u$VCC#h|zdlw&vcqrAq-|^bGhsj1>76$FDo2_UyWKd6RXe+nI znK-A!?-^3h)oU3-ma4_vIO)5(YfsA#G>z?SK>H?xR)>}Z3uN5O^K>)M(T5d*8@uX2 z6O+Y=P*9b7F7vxz{)jf8Oy)Ag&VIE3&)X_XNOTDB+oc~kdu2GXF4117hg3A#53l5@ z&^-oV_Vm863%0C{{*;lI>}+JE%*GGV>z?-yWQ@1a3#r4e@S}UW=`!GTd}Nee^(F!F zgubj$C-d%r|L5n*1p8UDO4U2%(uu!iV|g;iJRK63>Z(QPX`F=Dbbf*`S0r5@)M@UO zFwyrNgh1;WnS{CjX3qOXAIC!S*>Oq%Dx{ z4uR91E^uVa69KeM-1x(5iJl@LbNu06J0WZGyw)rg z?k(B1pX?@gr}}e$!E?8A*Y+0ov&vk@)%MR3FFeVe_$@@g7KLXpW1y|==^&XuR@;wZ zK_PZC%K^qC7mQd@G;>&6e7n((M7RWkW@x&a#ul+vHB|>PqKg3w6V*;Ots@fp6D2q< z8StpxQw!K1Y`06+N#LnE-HbTaBD3s_L6l2+as6!NUv=+aAv9eL0Y!G2ZC7WorNuFA zYD&nkWw;soHX*X*53fYk$~EI_N<^g`=~{9A&-5Z!crS-UNW}dqyHVBp77C6OMXgPURz=@0HdhI~;UON9uu@ zCnM|QjXffSc6x$tGzn*)jK`KMPd5jIuS;TDMG}2|Av2U^n7lE1oucSr*U-QA z8{SA@-?tk>B^IL^2S?La3(j|5$Tzus4z#a$s8r5EGhay}R?ZKPd>|J~5v*~eYqea~ zR$p5)Qf^)`uO|wO&jlwE>M}hLTdI}mzYgnGtFg$9-Nu7&8^a*xO}qZ zvfoG^%~k%6ZWCLSc47AHeXnCzqL0j5+Rvyn&k*Jvj<$Kva~K7I_W=@#GCh zZcA#HEj`l%x!}{QTRU2YTp6cgfM}Ak>uZpyyEIuX^a8sN}~J8PIjDhdLq%n zAy3bsVl&ka%kd!lUUu#P5J+5cC)NI{?g5A7J8!)2(poKY?cgBU`GI)$04tTWy43AH zTQ!p@<`Z{*Q~b)QOT~I@&w1gfnfn{u4Cx5F#8a{jFl#7BZb5H{bOK- zX~$wr%NI2U8LDUtVMD{<9K0S6>l9?Rho!P|RaZK>=P@LT+8A6p&fu*A!mjn%UDao z?PNHgFy0zf%p$FspA%Pp=2dhUMddjgorN*V5}nHb016?pppbdxbNB_RE+2#&%TN3rs@}WJeqoO*FTS)r~>> z`xA-@Nejv6fq$-d*CzL@?8s^E{1Mv2dDta*kvaTy{re}!1<#0Wwl`EB{mID&{MI*@ z!6~#5fF<6PG2XVy>|0}(ZDlo2mk)YL70FqjzeFcN|J=sy){I?WBv-{G}H4 z?HLN;%07f?YAsf_QLo!_0u4_Pi1TTCYcKHZnJ#n9;+{O%N7qcrpXX2Z@7;~#t574- zZB32ACczKBFCC!kyC105Q*Pn(yEE;gpeeMEteo_Mxi!GthcQ>(e;f@O*^nXqGp;_W zJ@#wIWjwviQc!;!K6*_F#-9DQ~{iz{~s{Z%H* zt*k1QxaFtSPlm^8mMAWrD0eK8`gJuFvedsr_PaLNR{m>Xkk=!U*A$W)-C&eEKpL3v zD%qC`W~sNp%eqfLwGBOpSD%RMe7?SWy4s_2nPOlzt%i2J*g@(`G5+;z$p1CY#O9y| z20mS%W0Y^K8b)b|A{80sN7*4SwukKzDGn5zzGd8zB0<_kE_K-CH+9*yT)h0pWV_6< zD(XG^BTsU5KuN-m5>GuTM&^5!A9l=f`&H$}LawlBkFFzWL*bm?uWzy8?C#yZOf=p! zjb+lgsHG@$mN$H;;eMFxt-xVM^R)mkB04KMhAa=>3GyxjwLa&(cD0*sb^yR`Jy~(M0>CkbwJRXw{{#YTsP=g}m5^ zfN-pgPYz3k`@q-eJzL65x_YH1RNh~lby~JDw7ld7Ix{?a zb*kB{vsW<+=!##@8|qiHaW^wK@oSRE)vwNV=H?MBG&uF}fRbY6k+xJ-)cBTj4}FY- zd~WT!8Xl;@CN4|&lo4wBEh$I=(4=Z*=raCro?Ue1If*E#L-ofY@JE{h)%F;k;W{u^ z^_)V$!FkM3x?lopu3>4!dy+{jDqX5p0?pAPBto9~f} z?pzA!tF*B+ZvaMOXZHckSB)e?e5k?YRIakNvFSVHB*BDFM>^#bV_YcP+@}xOm~5Q@ z`w%jo;99;dO~xhODSS-un)+S8_ilBvHs)YJ-hsBwET!M7E`2XxK>8xsD!*(aJuw3_{PPZ%Tiy+FQ^E-0%BYI|O%t| zZOCSY6W>qaH)fjCQkCIP=`{tf0JP1)VEN?lcA!64gQsP1XLaVom(8NLT2jAp0V)#F zth|2OZAS_L>18&CC*R(*gj%@=COl^CEMhdFSGc8W(5~LMUQjjbs57-2ub(Q=My@bO ziVA@@QrtgrEBIEVN{FSSz~6r-gT!>xZ6lA9{!ow0h6&c^jVt~_q_(^6sdTWQg(6Wa95dlzOO6Bi^v zY_ZrF5BKwStI~$$S>d%m3i=IG-QyG10vv?`2zfPTuh*>{ZTft-@h0`hxWVjbs@Bcz zva(lsVN9dJa(+BS`oZL|sulxY;oE_9*`=U{VX*-Rk#bv3Rfv{Iq9=uu;5LXWc44 zB=}bz52u0YT#d_x$>JIpiKGSY4iuuvQUqt~Q)iYEE&c8oVq{Z}L%tg4H%$(l-jU}u zFK29x@b zUjSRUlG2p=t{#iDZs2SJm`Ax-B;x5Gy+mhEgCK8BUXvCu&@U`p)Ftrmrqb|42G#Q%Ls2 zlAxl3kFco#*kYy=Cd$|DE6YT_dCA#k(bO^RS}i^+j-9MoXbd(Nu|ADs{~&G@D_MVV`tN(#VD z9^E4$m$PL0F`&GcVbX>@V(330S@XY~!i=(WUkaOa$8vh8PlZ1*I8ipOY`yau+5hs0 zsNW`4KchHx=4#-3_+;Izb zHl;hrGjpI3)Fv~#{|=nqoQ9t!h$J`OInzCco_6zW!O@)5&y>eamhpxieXIfZ)y+1L zv|39kwicL4ne7ZP7m!Wur9|TrJ=s4YI>E}%7vD_S@QRlpn%G6HrF!C^UPm=KK>xce zll$hwDT!ylqzxd6oZ(&TRqZKE87TC_r;wY|BWA5`n3l+-d8}3g_=(jeU7wDI+ez$T z89w^r=;1&C-E^=VMX~yZh=5O%Rxo-R$F`hktT1kR9Pg_%K&m#8vAMwFH<$G``i50^>r@{?5i_SV156*AFtXwB&`;dHrNa2zSSG7t>;6 z*u_#s>V|JA5xk3`0=J(`E#oDlvYgJc2Zu`dgv}=B1F;S#G*p1l$An`!pD zW0Q6b`B2(Q#u2^g8Ed((b4J~YvIJJ+N5t@C)dK~9T5?#H<^s} zcg%0DRrfpGtmw5pqwBCE;W+;5o@3v)Er;6BJrAAYvC@A1@SO%My>aUjcNP1qx#tREt*~72{85@9?%yO<7+Bc`!ALo8~8sfBk<0hn>kp|!BmI>#a zxW!Kkt&%z1i_a22+9ZTa_g%?n{4qz4GSa?3k977C` zj3a$N=ijd?mkaZO%J%KVd(VIGRuVa|`^FO>zxlA4C}Y)UAg*i<$CO@~e2K!A%|850 zu+uo>OI~xYcnc?5VU=o*&CgHt9=0blWru7-n1^Y8F3$!h%#_=2oP0|E3`RKoT5Rgq zGo5bJEyw%4kjtb|Ydir549Cu%SU#Qqa}@4ew`4r!R20-kJ+KlLE+px)cJb5GBsB*a zO$v{mcCbOBzfN+}pBG51b{f>+y*YZ5-lmN_FW}KsnWHiO0_K8Le0@f0K#@o$>;~{k zseymlIi@;G_MTZmqxR#&_t^tz^-sQHY8nXyeHu$2vBgc-_BeEo>|b1u9E)T;+dE;q zB*8jWY^wxmAQ`~*C0Of&h?DQrM#A)`9rE;3oi9(#1=73SO0jxq;RCqGxJK7#_#q`m zZNW2j!|qd+@b~DNpmbzg1;C>D`5y`s&nHdR4M61|CQ$tare#zH^cT^^140dkUGc+S z4Q&%iIWCggL(e7hi$9Y^khY^uLAfI=b;t8<1(Q^P8>7yFil5wsC{FEAbS#c8*%e0b z$hc3g$&Vs;*<@689Ug=xX_KRl&h`GagOIc#7klFCFelx>q|s4>7rABJLU!g3l#!nt zkndBLY=Zd&VuQWv)YbfsyMG~azUE_abUB!(?6w!om#M6a>fQ3kV}yJnR@t-wI@sFj z6&zJ;WH#F};KG+?;-xBh%v}$%ldQszj!mqFEdJ!E?PJ)syZ_vyYLd zF`eXag`-rsyVmQ~m6Q7P2Enf9RU7Xo#5)+iMambN7aQqw(EuUNqwfs8&+l`>HyCnA19%=$%&sSNx(Q zn_7s-f@jxM`miSV+WHprB<}KW-txGXUhfhAQkd-qX_JSKwX+hpgkbRb);&wIHB8b2$p@l#;c-dB=wemTk_K zQPBscWS1vY^psQxXtc#5P{U3G4{|!vFClw5xdVF~%dDLw-xi&zy5~)Be_>(xPyF-* zdaAnhhs9!%b~2c+`enPAP18DzmZt`1Vy|t4Xr7uiNO(tc88a;Y2&#_+iZat~(rA zlX~+XnQk97$+_Hhrp0iLVd~f?=8`s_JNe+RFsI4`VP4{iwl@u5zH1F? zb?pz9RSccHLGBAb^^ED#^HQM_cxs!9WO6YXChFGkfl?BE!enRaz%JB@zX)EZm3|Dg z&v!;Q3<7i4p+V9l%2RUzpW6QSrSO{m5reAZ$(qPLH5uZwWg^{&s zx`C$sS$9Uc*y9)^E#GA?+{iIVha{Kke(sf`mh^0XiR6ykxJlKUYBdei6+p>VEN@}j zpM(3$W@LVzQaAGyKaJBt()0d!w+F@&jotW@Y8K!YFafb%Ph<~o-29jhhsH$iM$rUHi1SBY!!7SNJmv)-iBUC?o0#UCZ~)Cz1VQ6pV~-yeZU}I%JI!IG_ZI zh@ISsPC{+jxt%NUKdE$am-vZoIU5*hdQb=prCxEc-x#U1uf+#jtn(g%5OP&>hyeWE)Ycz^8S@NhhI`xgtm!L6mRm{3!P8}$OmK~gzXmJg4 zoLY}uqc?nb@`KZ0rtk!8qrR1H_YX9YtNzE~R6|gD7J}%zmOs_v0Fi=yrUmlo+bp*x z2Nrtt7i*hrWxg9ejC%6RWE!Ca$;%scdd+CQ@#b=>J_DRvX3h-{GoBlLUtO-fKcBr0 zjaS)Lz4W=xy(_Keb`nwx%P0|$SbsYR!OnZX%(OPeSpC0A27P-Wo50j<)@kTTDEb&=1Ij@n*30O z&wHGtN5Q+BNave;z)}=K5FiWHu|gvc0{$#o*QKwlz;%-_HZVm1n?BsWU(IU z3mEhpL7d1FVZpW4T`w%iRM~5NjqK@5;i@~GO3f^8-EL0|Ll0y0YZX=(IB9%a(`EQ) zuEScjEcI$6IcUT7!lopY%Tnq^CTvHGg1U^)%3M0eB~_}y&FEho>eH_KZMfKX^kEl*N27QxL2O z?t4%_z67i`e4p;^ncA_r&e-X1IS->zxm+_At)0_v5fVniGOSh+C z^75u%PNCnemf7JENOW!JgK?@(#B3|%Va{2{Rt>8rYyy%&{ILwc(ukz1wMX$izDUC) zXb#kH<({=gf2R8q-Ud6y-+7){QTT$i7tUXxKiXc;Va>SMt@joL+O)mACAlxJg-hNO zDCtHWfs_a@nXB|uVv#p5WFID@09j%+U4m-L;@%c zr{KRAANVj^kmgy%ZKmtYg%3^=i}v77nX&@?gko(pucy&{v)+(V<&wb%2i< zpNsv28a;I;E?2NmUB+RkAtT0=Up(}5k?+O`GPY)d%_OZj;AZG*rhH;>pus)2zQqsv zs|LJwg)VHr?glSb{FqpKDXzgCdB3j?S11++PLH!G1GL?T&Yw;mP#$!8(=y%X1k=7_ zpPDT-zVs|f*iOuPz%{+KE4dE$oYB09d^=0X9+lZDN{FfA{lTX6p-5FW@t3^Eq~(=* zAM&i=a2Y9}y)~fOw)Z&>ax)#{S{}T!avpc+&o!QpDABx$C@FA{OLf@Hj*t?RJU|8jvTRThEV4ET<-o@6MTQF&Q;j=)3?~pWE<6%s_VtirUGOm4#eEc z%SN2^NTTzky{1n!I%tK`BrFC#2GEf`o@TpgV7YnUCFSWMoQX(UG6ZTzJudzjGVS8+ zOFP5k1xm#>Ac`)Nh%I?DZdu1ytzOMPZ`;6zR;YRpQRf9uy`p*FIyREg_9SmAmf z;3W(TC^Wsmdp=}d>Wo>`G3&B$w=(gRam^S)x7PM0D0O{WJ7j=G5Mh#}=8*SGNL{2)*$j@9Elsy}5Y<>4Lu2%3>2 zn)Y{!dmdcOTP-ZvBa&UE>waBs`TG77&H7pLU?A;yy3^PPKApX1u>6IdY`h4&)P#Jp{5{S>B8shP z8nLQzme-ZXDlmK?oSC;lHmw$fRz~hHZ?p!}dwSGM0nR1*#k><5E@Jjj!XN z`j4s@m#`8hWOVb_X>J`8AY8IwzfSL?($gfQeF)+nzUm*Ipc zIIo!?31gs73ob2e4C_2|^%uIJ^Gc!^%ytbZX8C*WvPLMJ@&t&oZDLnEl zqCsOS10^7LVd>iUgJ0Ws&$`po)c%o;7J-|VRkkN;SYsAKsXKSn7+7yKRHfcJaoA4> zt>Anqe9DlVVwcV|j%b*t&P{{E8AZwnrvs$t~#Sq>!&w3XDW|s-mRT z-j)Cf8Dnq|H2(>r;|BByZq8Pg_~CS7a`BAPB$B@rrNm3}TP57&#RORvFCA{$7ze9#B`@^U1L#DlvcYFIOISa6$x z0coh6h&~CVjfkZN`X7JFfJO_RyOz6lwUpdC!F6m^P?P1|&K~-C9u3S2`o# z?_c-$C+d6^PB@DVkcGMF0!bWh9*;(p3U%YVIfnT)Vn#g|T#O8%*B)NyDA0`1QUv9L z%YGL1>HS%;RJ~634w(Cy(JGsx7j0}>YLgYHGM3fs+_V{@ffoW!>#icBi4FVLShu&3 z=y?ijObHe9vCe zCs<7##a29)nx7Sl%%~sSGd$<~PIXEBv$R8bYP}=}Z}wD_V{Tju`JUNT5FSuFqMS&hl>h6kq#y0;Ibvr^Cpha&8c zAloB_KWwalO1%XDeyA<)Sxz;rMqw}Hzo**GmdZ9FF@qp+&%>>majrK_-zmBW@|2wo zmnatQtMvur-U49N60xRhpY@USp;@WENGDJ_cQfh6Kwi3rGQiQ8FUgL60v2_Wm|$vxkPe8O?=|?MxqtOgn7A)W}El;(tmLgP?QXQ z3!0dH*1Hb`K0gXF(~>FS5+XIR!>203`BPPw=@J&8yAPvM9Tl>+l3#Vl@wmtb^x@Og zbb01DCswtTj^q6!Q=iA5WkuBldLZ6wYH>BYlN6udIU4a?M8PZzU$Mi;hoA-|6f97+ zb|vzq>L=AZSmUA4+ie?3O#v&YLARh&1@_zMmQ4i$A^cDlP(O?i{_wWS(+dHGERjh( zzPD5XX7H_lEteG3;vj4agm)d65?6s!+{oz@x!yFpvIC`sJ_lpUj1B$={PiiBN=@m! zDtk7=h4#;1UKsVd$+@=^)3w3js${Rx>bYhh9z{X>c6N*2Ba(qSlXo>%?uec|yA&6# zUI4rJY}Hwd!b@!X)?&d9J5s&%F@>KgLC;~}8&;EbCOp28K$0WsU*@XzCdd%57~5+J z4*Zw#%om)EevRl@8=rJ&iP0Xt${-!;7`_Pzk-Spmr&Mz|NME^gy&`-_54MvB6+QiN zvsg#d$vMr7-KkkWSr42Q>^3Mhjvs7w*q;b8DDJOT^Ce8 zvmUJd>mkXHOD5F}oc&eB0x}2Nqjwlo2>W{Ny9D}75~Y2!HxljLD!xMVmS)GJe#*BK zkSI*mfoBd2lPz&aQ{j|f;B=m%kd;Hv=0QE6cyy>2cU5j#v}!tJ;Wo81q@q{>*sK!G zq;=`T9&NxOc;?}i{sPsphgs6kP{Vf}evHH;5YKz2<(2_y!1H!sjw06u0sR>d{z;#$Fax86wmaEu<>a=2tsMtzojL%kfx8#cOaVWItuyy44-Vlqj1oSMU>q zcu>Mvacpwacu_R>rGkQs_^pNRri*XyPZr-GT=!@!EZuBpZbUFnfDKK@;+6M?zzKRs z;nX3w<*5MlleVm$CYgKeq39~G$=Z-Al$r0iI5+c7VyYM^V`pSbncD4VE^orCW?591 zm6>#+W5EPqn|vH|ENo5l877iE*L~PWN;2Iynigcz?~v4~GZY1sZK2-Jviw#%r9OtB zwUnhP+ne0UuFmAX+SVdRA2Q?j@MDn#+o_e1p8%YDcQpEFR|owbQh z{S8|Mzx8ueC1uB*Qz(~?sn>9d5jclJUw@;qF$<2 zmP<>cJX1DinmGMBzA}nB>bR|pe!OnkOp_P_)0?txCMGHU*{~h4Pp9RL--)t7H9>}$ z42No;&onpqcV8OLv62laS3?bi=BLyTwsx%?-pr-Yz1wTi21K<(p(W270?b3DvzJH( zg($XwN#M8zZ+@HXNwy6gzJ2J}$niu$(9CI~$tiGuS2Lh|Z-(`9su5}LG@H_R+uT$@ z*M_f;ZPx$O-gU+`opftN5Cj2t1$hxEyNCz@L_mskam5aZG^v82H;I6Np}J-j1r;eF z(nXpCA@p8U5D1-6La$On3na88ci6q_&v>uz-F?5@FUwaZ|Cw{#r3z8BSKsrDDv1&ZlGOZ=^b*BW3Wa?_OYu*dv-9ZFDL6CkxEaE1XnIXXEz7$koZPRQaI(P8zdG||4{dm zg2q)1)cX2!#8SBg5lMU1yTg=0+f?T9m*@#(=9uTxWqy{K3Q?O`Udy#As9jS@5x$(` z^QKZzTl^`*dbK>dH2R1-6>a|j7F7y!D`@*+UH)X%)=36w{4YSUE(#Ow}f?oOEu>yp_lMbNjIBN}qOKu)DIobS)FOw^a<>rco-* z@*EYw*zl)2nG!{o~OpRqP!SKL!-wJ=Z-cB z1q9ye@9rvbm7G#qk48#uI}~}*2!Wfrd5T%hA1e9?QPknU^?BYFBRo6mWY!XONq1sR zm}(m_C^SDl^J&JA4@CACp^CGF;-TA%@3@$6LpEbIJVu75VcW}E>J8=aZEv(<%B%wZ zO3@2=rBke=9^CPYKaaiJ+=6HB#*}5=v@KEAxP}ZXLYTP1{8sR@R!<#!(ece^at^de zq&eiSwd?$=@p;x|P!+Q}Fw=)hyN}OaXb|POVflz)sa$N_<51<~RX@`7U?pm3emky^ zpOW}|OjSO`njtvuN4*9%_2p9|0-!h$7e6#590Fcba2QCGtfp&hCCm<_!Ri~WFDRmD zxz&_hw-zQmGQ1VVE~OJLCN1Yi6i1)tnNVy=9H#YoD-n@h zJM)Q@4E=cQ=YgOzk=HlLfh+dX_2vHcfrb#h@Li^pq1jMh9@~$jss$mOhf$A#@qPP; z6CWxK)aI8@R_{F3f0lLwj1=}P!XFPahE^7b>>iCwMRP-V`b`JUx* z%)hK)iX|#U&MR%ZJP3in^()C*(MQtL`Gm|pxU`}LGtT99(|)WyE=fwNoZj$RUokXr zu@O+)j@RuV*Rr3asmI-=f|zCBZw|-yvK|9j)RJ>(tGQGPjgUUyTU;zOS;)r(Dw5i& z(^gR-xb8w{UODIKwUFss<1cHrI8d&-$IJOYBqs>TVs6VjX^F-;CSjYV{gU6E64nV9 z!0;)bG%q|JZvuI?yrLr{GZ&v`{`lIYhWT14{^?a+(;)NW(|(_gw29LDkaOtig+qf#=8LNOOWSffPbXI zvwCw{I>D0~q!TS6aCzfwYBj-J6z|7k+12($@bq9$?#CkIm1&MzL?{b~yvF_U=M{`M z*A+FOo%d2nHuEF_E!0Fzx(ljF^9*ih<5PpV8fn|=G3k2%&vN5dg5l{nAtjHDD$652L=iuO;JsTzlVRaJ%zK!ebgg^hyzBr0 zlbKM`Y6zSm9u?gka>SSP+6kO%!*%Y@QP>N5$NU64yiT zF5@WE#J;(?AHXv93m5eVPiX*?@v(n;8H?QTuXBvR*>tJ+$2aLkT(~>7tr{PphShYQXq|`M%RIgqOQV0`|d5 z25^8X&3aoRa{a8H%tsnaz5D^Xdr32`cHk94Ay?5kTXGLHJhw}uyx))0V~M)gZU#NH zh>Q*ZbXgErMJ9)`l=(Z(z28ZJhwHcPo8>DAi;)I_w5?|!^7kW-(vFXww!w`lm^QCW zDE!uEx6(ztSth3FI(KY)J1561PBuy^J1{)AItTCf(Lbv)W-#rtn99h0-n(y!-B(e? zd}=lRd0lFnBNG8)yyNgtR? zu~B|`AG>}qglgm^va~t{Dax6PEUW2nB&p#2!a6K@@a9^hPB5H%-y~yY13oepfOm%H zEwbOz>vMbZMc1(VMT9;m_l`zi(aaVncG~oU8U*LbW@?{y%}g@eMk+l`pgzH~5tJPR zNeM)PG?m%k!y{sdq+an2+Z1FjIuN{i5WrG3e6QIq!I0Z5 zsgHcECz3{r_g@IbixDcTeuCZ!Eax+oI6<$fx?*L%xr_INq5wJI;#K`Z`<|+JmzaXWT-Y+J$o6EOxZr zBF{Sm|9I?-n!f$m7RT#+ukCB(1Fc#p2Ca$n5l0+VSOl!Z%S~{ALP9*6(5w(P($!XV z8(db88)YzwY6Ps7+Z7Uyt(Edx!K!GnpkEOJ9y$;zkg$olVWd#T&!EFi_dc94}ttSy$5WXnw>6qv+aK(*LSR zAORxM4RSy{&8O(c9S;iZ+F?VQ2{2K{GwogLexDa5J0KtMP8()r?$4Kuav@D!r(w3a=eqzE0Y#%~ z&_12G71mTqT#^A#NRYkbQR^Fk6`$xTcBx@-ZG^TMYL{|#?W~qCO7ouJfCT=*{Vz#L z|6W$Y_EpBveVp%tqikJ_n#N2O0p&aSO)f{q4ZCJJ`_?1Ol6rB{Ku`{%#aVw8eD%}P zNqIoW^TJ44=GQSBbH3wih$@xM%xFD%f7vp1AxZ*OwQ1J5A1kGF$wQATciI`?L`e@8R!D7=b&$Y<>DxCy9VZmM^{Hmm0$s56d~@ zT%Qcv^E(Wc6>?Nni^knvc`?-?vlm1zL9$)-OIN<6iv3&i7QEA>1;DTH^PhW`V_vl9 zbfy2gECFAv=;|ud`YAH~$w&P!;ufPVS^)4I=I=9p7s9Bsz(b8GI{XJBwPMkZB(`^soEgDEjEG|uK%DgpLH&f*d5HWh;rY@`FY03 zn_9upuWPz^VvV|u11tOo0ot&z0VL|W*k25l=H#FQ<4rFIH9P7u>T z>=OuLcM$9=dp#)TqXa;-`if}ILBzJ~csgp&w=%?3@Y+bLuo1hp${2J@a)ZRfEQmUc8~=HSEp|$^k-P@&rbe>u?g?KEs^IO z{>#1Z=EeQr8*55|x6P*PI!O4re>X#aX>~9J%Y*E%T}t=zYA{Rd$5`uINN|GV8k0Ha@U-`OC(X!Rd}eNn~#iLKvt`47PUUjZg8RkPX@1#uKw z-A8}aCQUJCZJOuAk#nY;InC(jQvO;}TpQpzjvVnRK(f>Q&cQd>N!>Vm8t%LqSAUpZ zw5Nv}RgIl{id<0+7r}je#6!MsPW2O9!OP}Sst8C#4&*+h4$ex>Sr$FTLcg_ZiZj*0 z3YU+fx#yk2p%1qYE}FsKqP*z4Vx2t9+SXGCD;b1E&v1{aKiD=(Fl#pYLhtKe2gV6qU_AfvGVE&|YZkH{i$~xhW4UcesG~T`Sns6<5j%m;oIxzxRxj+;yJPX$`8X| zk`(+7Z7B$tgyt=bH=RI9I3@szLXlP~{X1+a$Uz^$xWm@7in^xfKv=`^8eX#*B=6Yj zsGx7iDj%u|0_Io5_)YqAoMW<@ZzA6}_&U5eFhQr5;7)cA!6r(3XRj>oo1Fu=N=?2F l+OF2D^m9

    Example: + - Explicit command: -```bash -#[🐍 aurora_frameworks-2025.2.0](👻 Megatron-DeepSpeed-aurora_frameworks-2025.2.0) -#[/f/A/A/E/A/c/r/Megatron-DeepSpeed][🌱 main][✅] [⏱️ 26m13s] -#[11/10/25 @ 10:19:03][x4417c6s4b0n0] -; LR_DECAY_STYLE=constant \ - LR=0.0002 \ - OPT=ipex.fusedlamb \ - OVERRIDE_CKPT_OPT_PARAM=1 \ - TRAIN_ITERS=137650 \ - GRAD_ACC_STEPS=2 \ - LOAD=test_rollback \ - DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ - bash train_alcf.sh \ - --lr_constant_plus_cooldown \ - --lr_constant_plus_cooldown_frac=0.9 \ - --min-lr=2e-5 \ - --override-opt_param-scheduler + ```bash + LR_DECAY_STYLE=constant \ + OPT=ipex.fusedlamb \ + OVERRIDE_CKPT_OPT_PARAM=1 \ + TRAIN_ITERS=137650 \ + GRAD_ACC_STEPS=2 \ + LOAD=test_rollback \ + DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + bash train_alcf.sh \ + --override-opt_param-scheduler \ + --min-lr=2e-5 \ + --lr_constant_plus_cooldown \ + --lr_constant_plus_cooldown_frac=0.9 + ``` -[2025-11-10-095114][I][] Detected PBS scheduler environment. -[2025-11-10-095114][I][] running [ezpz_setup_env]... -[2025-11-10-095114][I][] [PYTHON] -[2025-11-10-095114][I][] - Found both conda_prefix and virtual_env in environment. -[2025-11-10-095114][I][] - Using conda from: /opt/aurora/25.190.0/frameworks/aurora_frameworks-2025.2.0 -[2025-11-10-095114][I][] - Using venv from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlamb- -256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0 -[2025-11-10-095114][I][] - Using python from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlam -b-256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0/bin/python3 -[2025-11-10-095114][I][] [JOB] -[2025-11-10-095114][I][] - Setting up env for foremans -[2025-11-10-095114][I][] - Detected pbs scheduler -[2025-11-10-095114][I][] - Machine: aurora -[2025-11-10-095114][I][] - Hostname: x4417c6s4b0n0 -[2025-11-10-095116][I][] - PBS_JOBID=8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov - to calculate: - - num_hosts: 256 - - num_cores_per_host: 208 - - num_cpus_per_host: 104 - - num_gpus_per_host: 12 - - depth: 8 - - num_gpus: 3072 -[2025-11-10-095116][I][] [HOSTS] - ezpz_print_hosts -[2025-11-10-095116][I][] - Detected PBS Scheduler -[2025-11-10-095116][I][] [HOSTS] -[2025-11-10-095116][I][] - HOSTFILE=/var/spool/pbs/aux/8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov -[2025-11-10-095116][I][] - NHOSTS=256 -[2025-11-10-095116][I][] - HOSTS: -[2025-11-10-095116][I][] - [host:0] - x4417c6s4b0n0.hsn.cm.aurora.alcf.anl.gov -# [...clipped...] -[2025-11-10 09:56:45,204] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 -[2025-11-10 09:56:45,302] [INFO] [utils.py:781:see_memory_usage] Before Building Model -[2025-11-10 09:56:45,303] [INFO] [utils.py:782:see_memory_usage] MA 0.0 GB Max_MA 0.0 GB CA 0.0 GB Max_CA 0 GB -[2025-11-10 09:56:45,303] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.83 GB, percent = 6.5% -[2025-11-10 09:56:45,304] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 -[2025-11-10 09:56:45,320815][I][Megatron-DeepSpeed/pretrain_gpt_alcf:151:model_provider] -------------------------------------------------------------------------------- -[2025-11-10 09:56:45,321819][I][Megatron-DeepSpeed/pretrain_gpt_alcf:152:model_provider] Number of parameters in model: 1986054144 -[2025-11-10 09:56:45,322546][I][Megatron-DeepSpeed/pretrain_gpt_alcf:153:model_provider] -------------------------------------------------------------------------------- -[2025-11-10 09:56:45,484] [INFO] [utils.py:781:see_memory_usage] After Building Model -[2025-11-10 09:56:45,485] [INFO] [utils.py:782:see_memory_usage] MA 3.71 GB Max_MA 3.71 GB CA 3.72 GB Max_CA 4 GB -[2025-11-10 09:56:45,485] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.84 GB, percent = 6.5% -[2025-11-10 09:56:45,485921][I][Megatron-DeepSpeed/pretrain_gpt_alcf:161:model_provider] Patching tensorboard from checkpoints/AuroraGPT-2B-ws3072-ds-stage1-nl12-hs2048-mb1-seq8192-gb6144-sp1-pp1-tp1-bf16-optipex.fusedlamb-lr0.0002-lwf0.05_ntok0B_flash/tensorboard -wandb: WARNING The get_url method is deprecated and will be removed in a future release. Please use `run.url` instead. -[2025-11-10 09:56:51,975268][I][Megatron-DeepSpeed/pretrain_gpt_alcf:168:model_provider] Updating WandB run.config: [sandy-darkness-4309](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2cmpsosr) - > number of parameters on (tensor, pipeline) model parallel rank (0, 0)=1986054144 -[2025-11-10 09:56:51,979967][I][megatron/optimizer_param_scheduler:89:__init__] > learning rate decay style: constant -[2025-11-10 09:56:51,980855][I][megatron/training:725:setup_model_and_optimizer] DeepSpeed is enabled. -[2025-11-10 09:56:51,981610][I][megatron/training:780:setup_model_and_optimizer] Did NOT catch: ('args.data_efficiency_curriculum_learning' and 'build_train_valid_test_datasets_provider is not None') -[2025-11-10 09:56:51,982375][I][megatron/training:789:setup_model_and_optimizer] Calling 'deepspeed.initialize'... -[2025-11-10 09:56:51,983047][I][megatron/training:790:setup_model_and_optimizer] Wrapped with: profiler= -[2025-11-10 09:56:51,983] [INFO] [logging.py:107:log_dist] [Rank 0] DeepSpeed info: version=0.17.5, git-hash=unknown, git-branch=unknown + -
    Example: -# [...clipped...] + ```bash + #[🐍 aurora_frameworks-2025.2.0](👻 Megatron-DeepSpeed-aurora_frameworks-2025.2.0) + #[/f/A/A/E/A/c/r/Megatron-DeepSpeed][🌱 main][✅] [⏱️ 26m13s] + #[11/10/25 @ 10:19:03][x4417c6s4b0n0] + ; LR_DECAY_STYLE=constant \ + LR=0.0002 \ + OPT=ipex.fusedlamb \ + OVERRIDE_CKPT_OPT_PARAM=1 \ + TRAIN_ITERS=137650 \ + GRAD_ACC_STEPS=2 \ + LOAD=test_rollback \ + DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + bash train_alcf.sh \ + --lr_constant_plus_cooldown \ + --lr_constant_plus_cooldown_frac=0.9 \ + --min-lr=2e-5 \ + --override-opt_param-scheduler + [2025-11-10-095114][I][] Detected PBS scheduler environment. + [2025-11-10-095114][I][] running [ezpz_setup_env]... + [2025-11-10-095114][I][] [PYTHON] + [2025-11-10-095114][I][] - Found both conda_prefix and virtual_env in environment. + [2025-11-10-095114][I][] - Using conda from: /opt/aurora/25.190.0/frameworks/aurora_frameworks-2025.2.0 + [2025-11-10-095114][I][] - Using venv from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlamb- + 256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0 + [2025-11-10-095114][I][] - Using python from: /lus/flare/projects/AuroraGPT/AuroraGPT-v1/Experiments/AuroraGPT-2B/cooldown-experiments/run-pt25-ipex-fusedlam + b-256-nodes/Megatron-DeepSpeed/venvs/aurora/Megatron-DeepSpeed-aurora_frameworks-2025.2.0/bin/python3 + [2025-11-10-095114][I][] [JOB] + [2025-11-10-095114][I][] - Setting up env for foremans + [2025-11-10-095114][I][] - Detected pbs scheduler + [2025-11-10-095114][I][] - Machine: aurora + [2025-11-10-095114][I][] - Hostname: x4417c6s4b0n0 + [2025-11-10-095116][I][] - PBS_JOBID=8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + to calculate: + - num_hosts: 256 + - num_cores_per_host: 208 + - num_cpus_per_host: 104 + - num_gpus_per_host: 12 + - depth: 8 + - num_gpus: 3072 + [2025-11-10-095116][I][] [HOSTS] - ezpz_print_hosts + [2025-11-10-095116][I][] - Detected PBS Scheduler + [2025-11-10-095116][I][] [HOSTS] + [2025-11-10-095116][I][] - HOSTFILE=/var/spool/pbs/aux/8140578.aurora-pbs-0001.hostmgmt.cm.aurora.alcf.anl.gov + [2025-11-10-095116][I][] - NHOSTS=256 + [2025-11-10-095116][I][] - HOSTS: + [2025-11-10-095116][I][] - [host:0] - x4417c6s4b0n0.hsn.cm.aurora.alcf.anl.gov + # [...clipped...] + [2025-11-10 09:56:45,204] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 + [2025-11-10 09:56:45,302] [INFO] [utils.py:781:see_memory_usage] Before Building Model + [2025-11-10 09:56:45,303] [INFO] [utils.py:782:see_memory_usage] MA 0.0 GB Max_MA 0.0 GB CA 0.0 GB Max_CA 0 GB + [2025-11-10 09:56:45,303] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.83 GB, percent = 6.5% + [2025-11-10 09:56:45,304] [INFO] [config.py:684:__init__] Config mesh_device None world_size = 3072 + [2025-11-10 09:56:45,320815][I][Megatron-DeepSpeed/pretrain_gpt_alcf:151:model_provider] -------------------------------------------------------------------------------- + [2025-11-10 09:56:45,321819][I][Megatron-DeepSpeed/pretrain_gpt_alcf:152:model_provider] Number of parameters in model: 1986054144 + [2025-11-10 09:56:45,322546][I][Megatron-DeepSpeed/pretrain_gpt_alcf:153:model_provider] -------------------------------------------------------------------------------- + [2025-11-10 09:56:45,484] [INFO] [utils.py:781:see_memory_usage] After Building Model + [2025-11-10 09:56:45,485] [INFO] [utils.py:782:see_memory_usage] MA 3.71 GB Max_MA 3.71 GB CA 3.72 GB Max_CA 4 GB + [2025-11-10 09:56:45,485] [INFO] [utils.py:789:see_memory_usage] CPU Virtual Memory: used = 73.84 GB, percent = 6.5% + [2025-11-10 09:56:45,485921][I][Megatron-DeepSpeed/pretrain_gpt_alcf:161:model_provider] Patching tensorboard from checkpoints/AuroraGPT-2B-ws3072-ds-stage1-nl12-hs2048-mb1-seq8192-gb6144-sp1-pp1-tp1-bf16-optipex.fusedlamb-lr0.0002-lwf0.05_ntok0B_flash/tensorboard + wandb: WARNING The get_url method is deprecated and will be removed in a future release. Please use `run.url` instead. + [2025-11-10 09:56:51,975268][I][Megatron-DeepSpeed/pretrain_gpt_alcf:168:model_provider] Updating WandB run.config: [sandy-darkness-4309](https://wandb.ai/aurora_gpt/AuroraGPT/runs/2cmpsosr) + > number of parameters on (tensor, pipeline) model parallel rank (0, 0)=1986054144 + [2025-11-10 09:56:51,979967][I][megatron/optimizer_param_scheduler:89:__init__] > learning rate decay style: constant + [2025-11-10 09:56:51,980855][I][megatron/training:725:setup_model_and_optimizer] DeepSpeed is enabled. + [2025-11-10 09:56:51,981610][I][megatron/training:780:setup_model_and_optimizer] Did NOT catch: ('args.data_efficiency_curriculum_learning' and 'build_train_valid_test_datasets_provider is not None') + [2025-11-10 09:56:51,982375][I][megatron/training:789:setup_model_and_optimizer] Calling 'deepspeed.initialize'... + [2025-11-10 09:56:51,983047][I][megatron/training:790:setup_model_and_optimizer] Wrapped with: profiler= + [2025-11-10 09:56:51,983] [INFO] [logging.py:107:log_dist] [Rank 0] DeepSpeed info: version=0.17.5, git-hash=unknown, git-branch=unknown + # [...clipped...] + [2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 715.45 | bwd_microstep: 2402.91 | bwd_inner_microstep: 1223.71 | bwd_allreduce_microstep: 1178.40 | step_microstep: 160.78 + [2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 715.40 | bwd: 2402.86 | bwd_inner: 1223.70 | bwd_allreduce: 1178.40 | step: 160.79 + [2025-11-10 09:43:36,161992][I][megatron/training_log:402:training_log] iteration= 136004/ 136100 | consumed_samples= 835608576 | consumed_tokens=6845305454592 | elapsed_time_per_iteration_ms=3415.3 | learning_rate=2.0127e-05 | global_batch_size= 6144 | lm loss=12.316531 | loss_scale=1.0 | grad_norm=1020.352 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1798.971 | tokens_per_gpu_per_second_tgs=4797.255 | TFLOPs=53.66 | + (min, max) time across ranks (ms): + forward-backward ...............................: (3198.27, 3199.53) + optimizer ......................................: (158.53, 166.36) + [2025-11-10 09:43:39,572] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 167.02 | optimizer_gradients: 0.46 | optimizer_step: 1.13 + [2025-11-10 09:43:39,573] [INFO] [logging.py:107:log_dist] [Rank 0] step=136005, skipped=0, lr=[2.012568679065123e-05, 2.012568679065123e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:43:39,573] [INFO] [timer.py:264:stop] epoch=0/micro_step=5/global_step=5, RunningAvgSamplesPerSec=2443.9516881034883, CurrSamplesPerSec=2553.443452275532, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.96 | bwd_microstep: 2398.60 | bwd_inner_microstep: 1222.63 | bwd_allreduce_microstep: 1175.22 | step_microstep: 173.52 + [2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.90 | bwd: 2398.57 | bwd_inner: 1222.62 | bwd_allreduce: 1175.27 | step: 173.53 + [2025-11-10 09:43:39,582716][I][megatron/training_log:402:training_log] iteration= 136005/ 136100 | consumed_samples= 835614720 | consumed_tokens=6845355786240 | elapsed_time_per_iteration_ms=3420.2 | learning_rate=2.01257e-05 | global_batch_size= 6144 | lm loss=11.967899 | loss_scale=1.0 | grad_norm=735.534 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1796.410 | tokens_per_gpu_per_second_tgs=4790.426 | TFLOPs=53.59 | + (min, max) time across ranks (ms): + forward-backward ...............................: (3194.05, 3195.63) + optimizer ......................................: (170.97, 175.03) + [2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 157.82 | optimizer_gradients: 0.47 | optimizer_step: 1.12 + [2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] step=136006, skipped=0, lr=[2.0124363314442905e-05, 2.0124363314442905e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:43:46,437] [INFO] [timer.py:264:stop] epoch=0/micro_step=6/global_step=6, RunningAvgSamplesPerSec=2472.2740699287133, CurrSamplesPerSec=2561.3206693919196, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.84 | bwd_microstep: 2347.61 | bwd_inner_microstep: 1165.11 | bwd_allreduce_microstep: 1181.70 | step_microstep: 164.54 + [2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.77 | bwd: 2347.57 | bwd_inner: 1165.11 | bwd_allreduce: 1181.73 | step: 164.55 + [2025-11-10 09:43:46,446173][I][megatron/training_log:402:training_log] iteration= 136006/ 136100 | consumed_samples= 835620864 | consumed_tokens=6845406117888 | elapsed_time_per_iteration_ms=6863.2 | learning_rate=2.01244e-05 | global_batch_size= 6144 | lm loss=11.744588 | loss_scale=1.0 | grad_norm=527.276 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=895.207 | tokens_per_gpu_per_second_tgs=2387.220 | TFLOPs=26.70 | + (min, max) time across ranks (ms): + forward-backward ...............................: (6639.44, 6640.87) + optimizer ......................................: (161.40, 165.76) + [2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 150.38 | optimizer_gradients: 0.47 | optimizer_step: 1.12 + [2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] step=136007, skipped=0, lr=[2.012303984797229e-05, 2.012303984797229e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:43:49,840] [INFO] [timer.py:264:stop] epoch=0/micro_step=7/global_step=7, RunningAvgSamplesPerSec=2492.441497638933, CurrSamplesPerSec=2576.511402462105, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.13 | bwd_microstep: 2397.86 | bwd_inner_microstep: 1224.18 | bwd_allreduce_microstep: 1172.96 | step_microstep: 156.83 + [2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.07 | bwd: 2397.83 | bwd_inner: 1224.17 | bwd_allreduce: 1172.99 | step: 156.84 + [2025-11-10 09:43:49,849423][I][megatron/training_log:402:training_log] iteration= 136007/ 136100 | consumed_samples= 835627008 | consumed_tokens=6845456449536 | elapsed_time_per_iteration_ms=3402.7 | learning_rate=2.0123e-05 | global_batch_size= 6144 | lm loss=11.613136 | loss_scale=1.0 | grad_norm=579.721 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1805.618 | tokens_per_gpu_per_second_tgs=4814.980 | TFLOPs=53.86 | + (min, max) time across ranks (ms): + forward-backward ...............................: (3191.89, 3193.07) + optimizer ......................................: (154.73, 158.25) + [2025-11-10 09:43:55,173] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 144.15 | optimizer_gradients: 0.47 | optimizer_step: 1.13 + [2025-11-10 09:43:55,174] [INFO] [logging.py:107:log_dist] [Rank 0] step=136008, skipped=0, lr=[2.0121716391239166e-05, 2.0121716391239166e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:43:55,175] [INFO] [timer.py:264:stop] epoch=0/micro_step=8/global_step=8, RunningAvgSamplesPerSec=2214.823701527157, CurrSamplesPerSec=1422.5676763766571, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:43:55,175] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.49 | bwd_microstep: 2325.32 | bwd_inner_microstep: 1170.29 | bwd_allreduce_microstep: 1154.19 | step_microstep: 150.72 + [2025-11-10 09:43:55,176] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.43 | bwd: 2325.28 | bwd_inner: 1170.29 | bwd_allreduce: 1154.22 | step: 150.73 + [2025-11-10 09:43:55,183901][I][megatron/training_log:402:training_log] iteration= 136008/ 136100 | consumed_samples= 835633152 | consumed_tokens=6845506781184 | elapsed_time_per_iteration_ms=5334.1 | learning_rate=2.01217e-05 | global_batch_size= 6144 | lm loss=11.405020 | loss_scale=1.0 | grad_norm=230.690 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1151.839 | tokens_per_gpu_per_second_tgs=3071.570 | TFLOPs=34.36 | + (min, max) time across ranks (ms): + forward-backward ...............................: (5129.49, 5130.83) + optimizer ......................................: (147.99, 152.14) + [2025-11-10 09:43:58,569] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 138.00 | optimizer_gradients: 0.47 | optimizer_step: 1.12 + [2025-11-10 09:43:58,570] [INFO] [logging.py:107:log_dist] [Rank 0] step=136009, skipped=0, lr=[2.0120392944243313e-05, 2.0120392944243313e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:43:58,570] [INFO] [timer.py:264:stop] epoch=0/micro_step=9/global_step=9, RunningAvgSamplesPerSec=2261.5249756425674, CurrSamplesPerSec=2589.0806546406557, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 716.27 | bwd_microstep: 2402.82 | bwd_inner_microstep: 1226.62 | bwd_allreduce_microstep: 1175.49 | step_microstep: 144.53 + [2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 716.21 | bwd: 2402.79 | bwd_inner: 1226.63 | bwd_allreduce: 1175.53 | step: 144.54 + [2025-11-10 09:43:58,579283][I][megatron/training_log:402:training_log] iteration= 136009/ 136100 | consumed_samples= 835639296 | consumed_tokens=6845557112832 | elapsed_time_per_iteration_ms=3394.8 | learning_rate=2.01204e-05 | global_batch_size= 6144 | lm loss=11.290359 | loss_scale=1.0 | grad_norm=326.170 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1809.814 | tokens_per_gpu_per_second_tgs=4826.171 | TFLOPs=53.99 | + (min, max) time across ranks (ms): + forward-backward ...............................: (3194.54, 3195.97) + optimizer ......................................: (141.84, 145.75) + [2025-11-10 09:44:01,970] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 137.27 | optimizer_gradients: 0.48 | optimizer_step: 1.13 + [2025-11-10 09:44:01,971] [INFO] [logging.py:107:log_dist] [Rank 0] step=136010, skipped=0, lr=[2.011906950698453e-05, 2.011906950698453e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:44:01,972] [INFO] [timer.py:264:stop] epoch=0/micro_step=10/global_step=10, RunningAvgSamplesPerSec=2296.9740702454246, CurrSamplesPerSec=2580.0686696658236, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:44:01,972] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.33 | bwd_microstep: 2403.22 | bwd_inner_microstep: 1220.85 | bwd_allreduce_microstep: 1181.64 | step_microstep: 143.81 + [2025-11-10 09:44:01,973] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.27 | bwd: 2403.19 | bwd_inner: 1220.86 | bwd_allreduce: 1181.69 | step: 143.82 + [2025-11-10 09:44:01,981208][I][megatron/training_log:402:training_log] iteration= 136010/ 136100 | consumed_samples= 835645440 | consumed_tokens=6845607444480 | elapsed_time_per_iteration_ms=3401.5 | learning_rate=2.01191e-05 | global_batch_size= 6144 | lm loss=11.142616 | loss_scale=1.0 | grad_norm=184.646 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1806.288 | tokens_per_gpu_per_second_tgs=4816.769 | TFLOPs=53.88 | + (min, max) time across ranks (ms): + forward-backward ...............................: (3196.98, 3200.27) + optimizer ......................................: (141.42, 145.17) + [2025-11-10 09:44:05,471] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 131.87 | optimizer_gradients: 0.47 | optimizer_step: 1.14 + [2025-11-10 09:44:05,472] [INFO] [logging.py:107:log_dist] [Rank 0] step=136011, skipped=0, lr=[2.0117746079462598e-05, 2.0117746079462598e-05], mom=[(0.9, 0.95), (0.9, 0.95)] + [2025-11-10 09:44:05,473] [INFO] [timer.py:264:stop] epoch=0/micro_step=11/global_step=11, RunningAvgSamplesPerSec=2327.316936098493, CurrSamplesPerSec=2602.3285686919153, MemAllocated=3.71GB, MaxMemAllocated=39.0GB + [2025-11-10 09:44:05,473] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 719.44 | bwd_microstep: 2365.65 | bwd_inner_microstep: 1195.86 | bwd_allreduce_microstep: 1169.15 | step_microstep: 138.29 + [2025-11-10 09:44:05,474] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 719.38 | bwd: 2365.62 | bwd_inner: 1195.82 | bwd_allreduce: 1169.19 | step: 138.30 + [2025-11-10 12:29:56,532794][I][megatron/training:1534:evaluate] Evaluating iter 19/20 + [2025-11-10 12:29:57,331870][I][megatron/training:1534:evaluate] Evaluating iter 20/20 + [2025-11-10 12:29:58,133928][I][megatron/training:1692:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- + [2025-11-10 12:29:58,134910][I][megatron/training:1693:evaluate_and_print_results] validation loss at iteration 137650 on 122880-sample draw from validation set | lm loss value=2.697780lm loss PPL=14.846734 + [2025-11-10 12:29:58,136014][I][megatron/training:1694:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- + Comm. Op Message Size Count Total Latency(ms) Avg Latency(ms) tput_avg (Gbps) busbw_avg (Gbps) + broadcast + 4.0 KB 25 980.95 0.33 0.11 0.11 + 8.0 MB 12 214.59 12.05 8.72 8.72 + 12.0 MB 12 354.42 12.34 10.63 10.63 + 43.0 MB 12 346.20 23.60 16.63 16.63 + 86.0 MB 12 492.66 37.23 20.63 20.63 + 1000.0 MB 2 62863.52 31431.76 11.92 11.92 + all_reduce + 4.0 B 3300 1828.19 0.53 0.00 0.00 + 20.0 B 68 238.60 3.36 0.00 0.00 + 100.0 KB 1650 108812.85 26.15 0.53 0.53 + 45.6 MB 1650 59143.33 31.96 24.15 24.14 + 45.62 MB 1650 72884.32 42.17 18.16 18.15 + 874.0 MB 1650 -259904.95 246.25 59.55 59.53 + 914.0 MB 1650 -296460.05 230.41 66.55 66.53 + 954.38 MB 1650 -223315.44 268.82 59.56 59.54 + 954.4 MB 1650 497077.13 285.12 56.16 56.14 + all_gather_into_tensor + 36.0 B 1650 61948.81 37.49 0.02 0.02 + 1.23 MB 1650 -527532.78 95.50 332.86 332.75 + barrier + 0B 68 6248.29 14.28 0.00 0.00 + log_summary_barrier + 0B 1 498.65 498.65 0.00 0.00 + wandb.run.name: volcanic-blaze-4312 + wandb.run.url: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu + wandb: updating run metadata + wandb: uploading config.yaml + wandb: + wandb: Run summary: + wandb: learning-rate/iteration 137650 + wandb: learning-rate/learning-rate 2e-05 + wandb: lm-loss-training/consumed_train_tokens 6928151347200 + wandb: lm-loss-training/iteration 137650 + wandb: lm-loss-training/lm loss 2.69441 + wandb: loss/grad_norm 3.39173 + wandb: loss/iteration 137650 + wandb: loss/lm loss 2.69441 + wandb: loss/lm loss_avg 2.69441 + wandb: loss/loss_scale 1 + wandb: +50 ... + wandb: + wandb: View run volcanic-blaze-4312 at: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu + wandb: View project at: https://wandb.ai/aurora_gpt/AuroraGPT + wandb: Synced 8 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s) + wandb: Find logs at: ./wandb/run-20251110_102334-7bjj8vgu/logs + Application 5d196584 resources: utime=21540179s stime=2039451s maxrss=46631076KB inblock=18009722136 oublock=2015065920 minflt=16872908673 majflt=453960025 nvcsw=11096789454 nivcsw=10717725 + [2025-11-10 12:30:19,247084][I][ezpz/launch:402:launch] Execution finished with 0. + [2025-11-10 12:30:19,247825][I][ezpz/launch:403:launch] Executing finished in 7795.27 seconds. + [2025-11-10 12:30:19,248203][I][ezpz/launch:404:launch] Took 7795.28 seconds to run. Exiting. + took: 2h 11m 1s + ``` -[2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 715.45 | bwd_microstep: 2402.91 | bwd_inner_microstep: 1223.71 | bwd_allreduce_microstep: 1178.40 | step_microstep: 160.78 -[2025-11-10 09:43:36,149] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 715.40 | bwd: 2402.86 | bwd_inner: 1223.70 | bwd_allreduce: 1178.40 | step: 160.79 -[2025-11-10 09:43:36,161992][I][megatron/training_log:402:training_log] iteration= 136004/ 136100 | consumed_samples= 835608576 | consumed_tokens=6845305454592 | elapsed_time_per_iteration_ms=3415.3 | learning_rate=2.0127e-05 | global_batch_size= 6144 | lm loss=12.316531 | loss_scale=1.0 | grad_norm=1020.352 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1798.971 | tokens_per_gpu_per_second_tgs=4797.255 | TFLOPs=53.66 | -(min, max) time across ranks (ms): - forward-backward ...............................: (3198.27, 3199.53) - optimizer ......................................: (158.53, 166.36) -[2025-11-10 09:43:39,572] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 167.02 | optimizer_gradients: 0.46 | optimizer_step: 1.13 -[2025-11-10 09:43:39,573] [INFO] [logging.py:107:log_dist] [Rank 0] step=136005, skipped=0, lr=[2.012568679065123e-05, 2.012568679065123e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:43:39,573] [INFO] [timer.py:264:stop] epoch=0/micro_step=5/global_step=5, RunningAvgSamplesPerSec=2443.9516881034883, CurrSamplesPerSec=2553.443452275532, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.96 | bwd_microstep: 2398.60 | bwd_inner_microstep: 1222.63 | bwd_allreduce_microstep: 1175.22 | step_microstep: 173.52 -[2025-11-10 09:43:39,574] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.90 | bwd: 2398.57 | bwd_inner: 1222.62 | bwd_allreduce: 1175.27 | step: 173.53 -[2025-11-10 09:43:39,582716][I][megatron/training_log:402:training_log] iteration= 136005/ 136100 | consumed_samples= 835614720 | consumed_tokens=6845355786240 | elapsed_time_per_iteration_ms=3420.2 | learning_rate=2.01257e-05 | global_batch_size= 6144 | lm loss=11.967899 | loss_scale=1.0 | grad_norm=735.534 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1796.410 | tokens_per_gpu_per_second_tgs=4790.426 | TFLOPs=53.59 | -(min, max) time across ranks (ms): - forward-backward ...............................: (3194.05, 3195.63) - optimizer ......................................: (170.97, 175.03) -[2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 157.82 | optimizer_gradients: 0.47 | optimizer_step: 1.12 -[2025-11-10 09:43:46,436] [INFO] [logging.py:107:log_dist] [Rank 0] step=136006, skipped=0, lr=[2.0124363314442905e-05, 2.0124363314442905e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:43:46,437] [INFO] [timer.py:264:stop] epoch=0/micro_step=6/global_step=6, RunningAvgSamplesPerSec=2472.2740699287133, CurrSamplesPerSec=2561.3206693919196, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.84 | bwd_microstep: 2347.61 | bwd_inner_microstep: 1165.11 | bwd_allreduce_microstep: 1181.70 | step_microstep: 164.54 -[2025-11-10 09:43:46,438] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.77 | bwd: 2347.57 | bwd_inner: 1165.11 | bwd_allreduce: 1181.73 | step: 164.55 -[2025-11-10 09:43:46,446173][I][megatron/training_log:402:training_log] iteration= 136006/ 136100 | consumed_samples= 835620864 | consumed_tokens=6845406117888 | elapsed_time_per_iteration_ms=6863.2 | learning_rate=2.01244e-05 | global_batch_size= 6144 | lm loss=11.744588 | loss_scale=1.0 | grad_norm=527.276 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=895.207 | tokens_per_gpu_per_second_tgs=2387.220 | TFLOPs=26.70 | -(min, max) time across ranks (ms): - forward-backward ...............................: (6639.44, 6640.87) - optimizer ......................................: (161.40, 165.76) -[2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 150.38 | optimizer_gradients: 0.47 | optimizer_step: 1.12 -[2025-11-10 09:43:49,839] [INFO] [logging.py:107:log_dist] [Rank 0] step=136007, skipped=0, lr=[2.012303984797229e-05, 2.012303984797229e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:43:49,840] [INFO] [timer.py:264:stop] epoch=0/micro_step=7/global_step=7, RunningAvgSamplesPerSec=2492.441497638933, CurrSamplesPerSec=2576.511402462105, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.13 | bwd_microstep: 2397.86 | bwd_inner_microstep: 1224.18 | bwd_allreduce_microstep: 1172.96 | step_microstep: 156.83 -[2025-11-10 09:43:49,841] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.07 | bwd: 2397.83 | bwd_inner: 1224.17 | bwd_allreduce: 1172.99 | step: 156.84 -[2025-11-10 09:43:49,849423][I][megatron/training_log:402:training_log] iteration= 136007/ 136100 | consumed_samples= 835627008 | consumed_tokens=6845456449536 | elapsed_time_per_iteration_ms=3402.7 | learning_rate=2.0123e-05 | global_batch_size= 6144 | lm loss=11.613136 | loss_scale=1.0 | grad_norm=579.721 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1805.618 | tokens_per_gpu_per_second_tgs=4814.980 | TFLOPs=53.86 | -(min, max) time across ranks (ms): - forward-backward ...............................: (3191.89, 3193.07) - optimizer ......................................: (154.73, 158.25) -[2025-11-10 09:43:55,173] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 144.15 | optimizer_gradients: 0.47 | optimizer_step: 1.13 -[2025-11-10 09:43:55,174] [INFO] [logging.py:107:log_dist] [Rank 0] step=136008, skipped=0, lr=[2.0121716391239166e-05, 2.0121716391239166e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:43:55,175] [INFO] [timer.py:264:stop] epoch=0/micro_step=8/global_step=8, RunningAvgSamplesPerSec=2214.823701527157, CurrSamplesPerSec=1422.5676763766571, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:43:55,175] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 714.49 | bwd_microstep: 2325.32 | bwd_inner_microstep: 1170.29 | bwd_allreduce_microstep: 1154.19 | step_microstep: 150.72 -[2025-11-10 09:43:55,176] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 714.43 | bwd: 2325.28 | bwd_inner: 1170.29 | bwd_allreduce: 1154.22 | step: 150.73 -[2025-11-10 09:43:55,183901][I][megatron/training_log:402:training_log] iteration= 136008/ 136100 | consumed_samples= 835633152 | consumed_tokens=6845506781184 | elapsed_time_per_iteration_ms=5334.1 | learning_rate=2.01217e-05 | global_batch_size= 6144 | lm loss=11.405020 | loss_scale=1.0 | grad_norm=230.690 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1151.839 | tokens_per_gpu_per_second_tgs=3071.570 | TFLOPs=34.36 | -(min, max) time across ranks (ms): - forward-backward ...............................: (5129.49, 5130.83) - optimizer ......................................: (147.99, 152.14) -[2025-11-10 09:43:58,569] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 138.00 | optimizer_gradients: 0.47 | optimizer_step: 1.12 -[2025-11-10 09:43:58,570] [INFO] [logging.py:107:log_dist] [Rank 0] step=136009, skipped=0, lr=[2.0120392944243313e-05, 2.0120392944243313e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:43:58,570] [INFO] [timer.py:264:stop] epoch=0/micro_step=9/global_step=9, RunningAvgSamplesPerSec=2261.5249756425674, CurrSamplesPerSec=2589.0806546406557, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 716.27 | bwd_microstep: 2402.82 | bwd_inner_microstep: 1226.62 | bwd_allreduce_microstep: 1175.49 | step_microstep: 144.53 -[2025-11-10 09:43:58,571] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 716.21 | bwd: 2402.79 | bwd_inner: 1226.63 | bwd_allreduce: 1175.53 | step: 144.54 -[2025-11-10 09:43:58,579283][I][megatron/training_log:402:training_log] iteration= 136009/ 136100 | consumed_samples= 835639296 | consumed_tokens=6845557112832 | elapsed_time_per_iteration_ms=3394.8 | learning_rate=2.01204e-05 | global_batch_size= 6144 | lm loss=11.290359 | loss_scale=1.0 | grad_norm=326.170 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1809.814 | tokens_per_gpu_per_second_tgs=4826.171 | TFLOPs=53.99 | -(min, max) time across ranks (ms): - forward-backward ...............................: (3194.54, 3195.97) - optimizer ......................................: (141.84, 145.75) -[2025-11-10 09:44:01,970] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 137.27 | optimizer_gradients: 0.48 | optimizer_step: 1.13 -[2025-11-10 09:44:01,971] [INFO] [logging.py:107:log_dist] [Rank 0] step=136010, skipped=0, lr=[2.011906950698453e-05, 2.011906950698453e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:44:01,972] [INFO] [timer.py:264:stop] epoch=0/micro_step=10/global_step=10, RunningAvgSamplesPerSec=2296.9740702454246, CurrSamplesPerSec=2580.0686696658236, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:44:01,972] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 713.33 | bwd_microstep: 2403.22 | bwd_inner_microstep: 1220.85 | bwd_allreduce_microstep: 1181.64 | step_microstep: 143.81 -[2025-11-10 09:44:01,973] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 713.27 | bwd: 2403.19 | bwd_inner: 1220.86 | bwd_allreduce: 1181.69 | step: 143.82 -[2025-11-10 09:44:01,981208][I][megatron/training_log:402:training_log] iteration= 136010/ 136100 | consumed_samples= 835645440 | consumed_tokens=6845607444480 | elapsed_time_per_iteration_ms=3401.5 | learning_rate=2.01191e-05 | global_batch_size= 6144 | lm loss=11.142616 | loss_scale=1.0 | grad_norm=184.646 | actual_seqlen= 8192 | number_of_skipped_iterations= 0 | number_of_nan_iterations= 0 | samples_per_second=1806.288 | tokens_per_gpu_per_second_tgs=4816.769 | TFLOPs=53.88 | -(min, max) time across ranks (ms): - forward-backward ...............................: (3196.98, 3200.27) - optimizer ......................................: (141.42, 145.17) -[2025-11-10 09:44:05,471] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | optimizer_allgather: 131.87 | optimizer_gradients: 0.47 | optimizer_step: 1.14 -[2025-11-10 09:44:05,472] [INFO] [logging.py:107:log_dist] [Rank 0] step=136011, skipped=0, lr=[2.0117746079462598e-05, 2.0117746079462598e-05], mom=[(0.9, 0.95), (0.9, 0.95)] -[2025-11-10 09:44:05,473] [INFO] [timer.py:264:stop] epoch=0/micro_step=11/global_step=11, RunningAvgSamplesPerSec=2327.316936098493, CurrSamplesPerSec=2602.3285686919153, MemAllocated=3.71GB, MaxMemAllocated=39.0GB -[2025-11-10 09:44:05,473] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd_microstep: 719.44 | bwd_microstep: 2365.65 | bwd_inner_microstep: 1195.86 | bwd_allreduce_microstep: 1169.15 | step_microstep: 138.29 -[2025-11-10 09:44:05,474] [INFO] [logging.py:107:log_dist] [Rank 0] time (ms) | fwd: 719.38 | bwd: 2365.62 | bwd_inner: 1195.82 | bwd_allreduce: 1169.19 | step: 138.30 +
    -[2025-11-10 12:29:56,532794][I][megatron/training:1534:evaluate] Evaluating iter 19/20 -[2025-11-10 12:29:57,331870][I][megatron/training:1534:evaluate] Evaluating iter 20/20 -[2025-11-10 12:29:58,133928][I][megatron/training:1692:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- -[2025-11-10 12:29:58,134910][I][megatron/training:1693:evaluate_and_print_results] validation loss at iteration 137650 on 122880-sample draw from validation set | lm loss value=2.697780lm loss PPL=14.846734 -[2025-11-10 12:29:58,136014][I][megatron/training:1694:evaluate_and_print_results] ----------------------------------------------------------------------------------------------------------------------------- -Comm. Op Message Size Count Total Latency(ms) Avg Latency(ms) tput_avg (Gbps) busbw_avg (Gbps) -broadcast - 4.0 KB 25 980.95 0.33 0.11 0.11 - 8.0 MB 12 214.59 12.05 8.72 8.72 - 12.0 MB 12 354.42 12.34 10.63 10.63 - 43.0 MB 12 346.20 23.60 16.63 16.63 - 86.0 MB 12 492.66 37.23 20.63 20.63 - 1000.0 MB 2 62863.52 31431.76 11.92 11.92 -all_reduce - 4.0 B 3300 1828.19 0.53 0.00 0.00 - 20.0 B 68 238.60 3.36 0.00 0.00 - 100.0 KB 1650 108812.85 26.15 0.53 0.53 - 45.6 MB 1650 59143.33 31.96 24.15 24.14 - 45.62 MB 1650 72884.32 42.17 18.16 18.15 - 874.0 MB 1650 -259904.95 246.25 59.55 59.53 - 914.0 MB 1650 -296460.05 230.41 66.55 66.53 - 954.38 MB 1650 -223315.44 268.82 59.56 59.54 - 954.4 MB 1650 497077.13 285.12 56.16 56.14 -all_gather_into_tensor - 36.0 B 1650 61948.81 37.49 0.02 0.02 - 1.23 MB 1650 -527532.78 95.50 332.86 332.75 -barrier - 0B 68 6248.29 14.28 0.00 0.00 -log_summary_barrier - 0B 1 498.65 498.65 0.00 0.00 -wandb.run.name: volcanic-blaze-4312 -wandb.run.url: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu -wandb: updating run metadata -wandb: uploading config.yaml -wandb: -wandb: Run summary: -wandb: learning-rate/iteration 137650 -wandb: learning-rate/learning-rate 2e-05 -wandb: lm-loss-training/consumed_train_tokens 6928151347200 -wandb: lm-loss-training/iteration 137650 -wandb: lm-loss-training/lm loss 2.69441 -wandb: loss/grad_norm 3.39173 -wandb: loss/iteration 137650 -wandb: loss/lm loss 2.69441 -wandb: loss/lm loss_avg 2.69441 -wandb: loss/loss_scale 1 -wandb: +50 ... -wandb: -wandb: View run volcanic-blaze-4312 at: https://wandb.ai/aurora_gpt/AuroraGPT/runs/7bjj8vgu -wandb: View project at: https://wandb.ai/aurora_gpt/AuroraGPT -wandb: Synced 8 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s) -wandb: Find logs at: ./wandb/run-20251110_102334-7bjj8vgu/logs -Application 5d196584 resources: utime=21540179s stime=2039451s maxrss=46631076KB inblock=18009722136 oublock=2015065920 minflt=16872908673 majflt=453960025 nvcsw=11096789454 nivcsw=10717725 -[2025-11-10 12:30:19,247084][I][ezpz/launch:402:launch] Execution finished with 0. -[2025-11-10 12:30:19,247825][I][ezpz/launch:403:launch] Executing finished in 7795.27 seconds. -[2025-11-10 12:30:19,248203][I][ezpz/launch:404:launch] Took 7795.28 seconds to run. Exiting. -took: 2h 11m 1s + +## ♻️ Convert to Universal (Optional) + +```bash +TORCH_FORCE_NO_WEIGHTS_ONLY_LOAD=1 python3 ALCF/ds_to_universal.py \ + --input_folder test_rollback/global_step136000 \ + --output_folder test_rollback/global_step136000_universal ``` -

    E#I9&vk*eGIcH{IcU!jOR2haV0q15Qv{P4lPe5D*(^$f~H{Kap- z!CfzcwPJUb-O#2Q2f!jf(w5bF%X>~Ji{bmX*1jnuD#3&jKAhY2c{U)#Q=zOzH|2(Y zD0|?Zx+{F9Qke9GVD)GxpgOO#k-w=EFz-o#Nn9Eif{S9DUp-r>h8E2YSX(+RP44u> zk$y8(69Ol*1;+zI$;%0k=Xu_pEXAFKJ`x3q>7%_?d%ZZ_s2p~bd+bEG-ks{_JF8i~ z@X+Yz-BsTvJ;#M@{sxXHJ{g>5K%JH_)vq0X9XYiYU%;sKX}cU*?IlD3|9s%F-_p$1 z$w)Esm-*-J>b>e%HQr(P3Q^M6} zl&PPz>IuGS@7b0sVp)jvFt~7l$90*5c%EKH8%#scJ2VuJ3cvRip1v)vGI`Y`5YQOL zKhxOLg+DS_r)_Qa(w__6pdB`Qn=?F9$xSb1h{jo!4_GmrIiMclBc(iid$Da<{Nh^? z#+;}4N4(TC1DTHgKH~I;_IevrW6!xA`-_Md?oAbKhcWFHRa3F#yxDY5`zG*OJGnpW>SNiOH^}qZ#*KUBM{?qNflX;_NpVHOm|h0~x+Jsa zk;^%8uHI?-W}xf%Z?%OO#y?cVrmN0nI@0}kriSwV-T?BJX!bNof0QbcmL<-MTn&pY zjKIoc@8muUh-`qdjs>FhGH9W|n}-nXKXwpQCx24pMD*Rr#k^l16I(Iqu#MnjV_qBl+gCOulN7G^NIL~zmCG=QT_eeL$)gu- zGim1PWR4qimHZraDd>TajJe1{26vgd6=+twl3!@7Qx<6zYm-trXgS;I^d!JU!F(Q} zY(Qu=8d4Uts97h>yXkNn^8CYZw+d=%2NCKkKZia^IX)@rw(I`=9x?-b#>`%U4EY)X zCDDMN>f)W3I&+8bGhENI-7iPJ?GSj+(*}*B(A*%HHW!gQZTAS{@xU;(jnj{#%W*Di z+_(!@3XTd$(i!_b3HuHMnBeV`nnFT1h7_G#4IO8-Ztrg>)sjpZKD&G4;+KhKuWM_| z8!X=^Jq*scZam?MjMkM)SplV|XMRbdndGUBh-#+)sFsT`b;Ytr0%rFd&l>~t8ND1h z9&Un0Qnr?a`K^)*IAE!ks7r-QflihdqgN3V;@gdUhZ!{gXnp!& z{nuJZ(?V<%wdIhmKs3L)1^uLu0)ekIo9a_y)Wdf!6h{wbmCfU^$B%Lb>DEP>xo~d% z+hAl$U9NZezksA`{kNt_NV~54h^Np81Qnv0$~QttiHN%>2HxXbP(OYQXqMQ=N(b)MEfNjrkJUKyH7f0?K?uemT?z!?>~Pi<37_eDQes&1kzb>9MI3X6?8%E`*jIT+|R zY(i+*y#@93Q&K(SO;2?lbaSMSd(tr=6*b;MQ^1ZHPW~q(`K? zp(A6L+>4Btt*8K^ao@5SKUX=R61oB{z*k5%*=31yEmhQwAqMOoPm`!e(8THxEZbC; z#Pcyn@%q9mp$GD*sju_NEza`vjPc;m-Uu@*aNR1{i35(?Ayq$SFQ0CWZ&w$kB+@k0 z^4U*zby~3GZNzb4~?f!QRD$ zqX+Dg{x$16y)!&KpqKqf%?1X8wL6p_SyR(Y)KyjrcBUE*(=ny?rl*`qeV4Ap^9;=G zdrsZuPze*K%axDI@agc{U1{NBpOsL_9GC~#3E0pzdH6U6UF1M!Y?k}Ait{qvEsbHy zDp?0Cv0T`kp|3)!@J`8+&DNC&~K{$>G zYp#&XViJa47D}V#w99beETB%!fv8mrZpnD8t0iFECX2-9t{U6u`;fo85r>bC zi#j=OEsHgGrmMT5Nvj=m6%$C<$igmbU{iuiDJl63|LqX8JFYFZ@{-ojW1*HIuH%YW zf+BECt+u(qrvwjf1;#~m+ChhNbd>Y+7IQI8Jc(+`Y=;H&K$DH_ zOsV5aqSvWCw{lov%ZFY-y{ku|yWY}ZBjd!>!hb&4!nX1x&~^~Iy8UU7e}}YM+IC1* ze_@IC!;El>>*75(eeF}R-h5=SS}HgCCc_J&b|Yca6mv<@_%%<>b^!eFQw6Si^wVb{ zF#`eC$Y@bCa}PR;e`sK~WhqYaLN?_^y?RmWK6_RUJk+Y|{C?2J87t|uM|g1gU!5+< zA%VgKw@S8(l`=Gj3q(#SD~QuNG5cCHWkV&i{rH^>SAWvGFbexVXc?>9bRyU2YtP4# z)i;0NPNWBnq5jfnnW#jCp-}wF9ZKbn3hPwS0ZDzvjjpNEK};#@^7uKBC>lr+vN(m6t&DJfm!{>x2~TMg>Y5yZ%Wiw2LKpI1m$}`ewc@0RF0}U-v!pc`@Qo3hXqxa_rOtFd}TL0Igx{is|=` zSDb9gTX?Oh0Q1kE#(0}OnmXex2yE^kn%VmBB)W?>f>}rG`YYRTL5HXs<+49ykn#wr z8~8Dvpy&OmaXjrCM)7N0Uo5E&!H?CTQaEBuBuMUmMkx;-SQcv*-F8vWZRQH}L~h%#QM5+O#iBb>$_<*uoCiY~9xzHzQ)qtHNAAK9-Ig-L;LlVk_X_5zeE zhgFwe$KZe4qI245>4V`GBWrJF`}}u-{G7b~_8%{R(~9s3&m`a7KK^8&kVc>1W7kKT z6b~$d>(0Ty;Vy5mt$m?I9)xFao7T-SPq?TFeL4{MS@D>iViqtFn?_LZ{vpSmJ%{A> z9<3M%pWK`g?;w5)=p)P7q9&7?Vn{6#y#A65aN zv3_Z(s<*GTr7`*zbT0?U&o8v8-XPh$4zQbltBz0FQV?q0p(MeWl8M%0A=U;f8IYmd zU(CDN<%n02@zU|VY0sq1q!TU4y2_A2|1hkuC0OIs*|y3v{YZsvy{(qA?CG!+9PM+7 zeT^&WYj0@||K)T#-TBy2A=nwa{X$(SK4FuXjzzY$Xh4RsYosqZ5Lr3(McJiREn{GD z)xy+LUAZ}VzS2ic*SR`RE>**hmB+zgKC&5Fpi%X+^wIaIfoG;A=o+imnwp>*A`~$w z0BPa}v|}iFD@&ujGYVGHYV0CTiQNC`Wgtcfs+~A@9hF-;k8@Ro9tB$g7oVx0- z&~#C`12s|qIH3!rKo8^eYVCEe5jKi~@B1@*VBj>iSK4V^Jdi{y#K&o1`ii;Mn6&@ZQqU3sKG+4cr# z#%?`kYNJ#EQN)qO>Xny@aDS6W%&MXj+nU%<+-EAE9?)9!&qhNHha!&oVESdSk$1Up z!-0~--<1Zp&Q>S}2Cry+drLf7tLen4K2Ak13r<`j5CWP`jqSRJ=-^}Yz6xyYUw2Y*MeidTevr%R@tTg%UZ)ZPJnnY2sG8+6M z3MU%!U1WC2|GL(W-~J2BaJD5%Sukyo-q7ZLD9;{G;?8iL?qZ_OJ!K)stYO&-YQSm! zD;4bW_ZMUKj0hHzzRgjAgi)4vQLSYB|FaNSlPWs1yC)?!<~yy+-KpH^J#sgr#{V&U zL{`=(W&`|s0|)w&qIK2$2I6l6S+$hAx;pEDiF`z#88Sy7*mhR#EI#-@Y%ggk2u2qa zlod*n3a?ppUM$ckff|XJXaVTiX+i5KyFO}+24kM$-}3pQV{+nM2J^UqJn6`4yUlmQ zbeDkj?ThQ}fY&yiMKoH+vj3V*oAmrtB>`|_R=8?3aB!xIKE;G6ICweCXokJTQ^M^^mS1f9Y38BoapErquLH9NU>ct z^y-um4$~dYvAP_h&R(m<4JA3dHh3H-{B<@zM5v@ofxOY?j62;s3J0C(CAxLwZ>X7o zD@Qvta{S@?AK^S#8c1jy+3uo`E-#aB9*S4L`09T|t`M#9>{YO+?omF2{XFnUqg4Q8 zSv$YeIkjW03p1VW;Ci_xL4UhKa*w-&`mZ1-#iUYCP~S%eH&&2mFs0LH(jPEIyZxzz zZ*WbHX|1%gANngBKfT0G!zd=GT&-{PRV%e?{D!g;c4X|{^f@pzW$-Y!=+UvRCuvm! zag;oJ3KNM8;(3Q3lshED&a%y@iOYe1czVl)?9$deUbunGp&ktaMGa>2b7a@tc@a5w zZPEk`$ww1_6>I`!mq?HjXfuDcw&4Ws%|Neppu!=pvn-jpLXEQ7JX8oeNPj~5ch2N5 z6$shAsV3Pkm(W%lhFrEMp6?^GD{K+~in3&irz2dBlJcz1H!nWPGZ<1(!YWq}=?2P< zkE|+U`{XIT6Lo=GiiM$7D{w;r{ zM9usO;To(%Nh65c#L*?Eji?ObK`;4u(QC1u$kK08o>r^Q?;jChi<>hZIDS7Vx+y!p zP$1(Pc1n{s<-Ut%a$^KRH~IbR6-dY z_#_atyFh|*yrRS`mp>X3qTW&?=hvH&>sFz>1%XRDEoye}R(j_89xc5Y{%BUEkw_#Wfms}f*~aITTtaHG`a{UzORFjblK z7xub9a)lU!9M?RqZDhtVoFv72MUeMIX+Jk35#l;vOfR2$hfyzom+=wuajMx2lD=t~ zQ0>gu7t^vXi-qUm7Y?I_wr&S+wq@ zljIAG^K0Ag@LT9=bsP^SZZ0w2ufFvlu1S#i2^=*q$CUiIjOw7>sQff%#!WWhox%M58Q*_|C{eBjW5NPqoq6o~VYX=H0Au{ryx;iSlj(xdDH}4t zMIdK=S~RTM{PBaS)39NNO;OXQr&^`S;82sAl)?*Ef2A{xC7?+LOgcUN#yO$m@q_%B zsX~j|jSRv$XZuvf!P{B%p;Bs*)s|lRM{KeNaW3aq%K>Ti?I@Qm8$2QO;2}?N-N?C$ zlbGtlWZqUi3HnOmyJFFQb#@(S@q#T|CJq+AF%Evlmt4QaRL8egu{P(i7X!2s8w2u< z-RmA^WT)A&8|mP>q#5DpAFx2Hd)WI#rJx;Gx1S=ayrxo6d*`m-yyQvIidg?JyV1#a zs)CtOU#e5`!Te7E8XO_Iw{Lk`yM8Z9h65s6l2T`BZLO1~q>1jlm_DZSQ)nnB75A23 zHz>_p)ny|hJ~Y$U>6s45;{KKut}zvs^VRG*QJF^yL^m2+&2{puJMaZD8MU2uA3t)T z#7g;HI`RwJdyW}0LQTC4I_Gxb$cda^FTu)pkVJ&$#6%(M7nz+G+Rb;zOZ3tcSNytH z{CZ=DYGmDHWzMh3(0$O#Zx)INrWAuqRo<4OhFrIO{LE^@2+2e0f~DB@`bV68U)Yme z{#N@Qm{;#mC7*Y&J>M6)H|pz?QO?yO;e&fYFch&JyE{kSA0c1BF722s3e!@so(cDiwIiyozh?1FtOg#oJu7{sg13( zL&8Iu-DK;-QDW-pc)Y#EOFFHv@8z*!lbar<40TxK-sJSK9m>yUI>yH>Q89DwPMvng zshLFm5y1u2;JDY^?tMAIiXWi{@a+%q-=Ax7gIJXZ3D)c8QAn;pJluz0A2TFbWsl_! z$n)U77xT4&~)JZdU<#Mt}g#)Y~$zPCPd4zv@ z-)x*{b*H8<&IKP+q{wPli5R*1>XLz3r~BGfh6&~QxPT~TRg^z1O>x-`^nCO|L>WFV zWirG47SHfaruFA)tD8Vy5XN=5lar%A{8`$_CJ?jx;G~^$$my2ZxGJEQ=My^|O;c7Sr9ZjCrCDMVp91Co z269@&*LFNI##;j*v-0^}-AXMy2h0y;^XD5;^Of(z)ED=%F!2HDbs96q)!-O==YPcW zbYn4b)Ni}y@w*;eR?oY+`(dP4%-%k}bkg*i2|iPIz0cLiG7y7xt#Hv9cH%TMLpC{O zAtN^JVh!wZ1xS|%b{w^Ls=s$#3~{B2#{DF7fAeTs!4a=2!FOFkP&B~4IO;oh=Yw0oTl3I_y07Dj7Z zVEvb$JaqfdjQH1= zO4&t@n#ZB!YHb^1z3~ndX?`?sStEm@vJZ2LP7}g%l>d)RW7n;y<{0mLS(0g}A8oR< zu(GDyK4JG4Zvw;B(=1c&2VfrWxw(GS_--_<3=piIE)4CnrgGMEDX(Toud5NN+GW(g zY5!;J;CqZ}))Sz=zr&a!Ct@OIkiXdJosss;y3aw{HbrtV@q% z_|uV_Gy;OdOLko9joR}03#_?vu=z%5w+e}rof%q(_2&&j;~_6%2JU804iw~@TV;_U z-gB2BUp0rxv8{1s&BkzMrjB(Z&z0NRRwX0eF^dxyTg=-tw*M6*202ut<;ES@`a2{_ zwm#uwf)y=Zv`{6RIB!OkaWn(n zGWJhI6sj$a*5VTn zKnHi{QqjVV(*Npxr4at!qWnqey|yfLyWr+FsIu2Y9;*%e4eYYbnGa+yhJ z@o2AleZsezKAYIN3qrrLw{;ELN#cI7b&6#nP%na<%6Nu7gG_yeWB&q3P@u{U!= z6^DZ1Hz$dI-!yh8Gk#E7NdJl8`u;48{O|~bF9=0^kiWh5H~i$U;<}D*c9BM~bcK6O zV=ZD|mR(hZJhHBGqqxw{hC3gNXxuMEc{|!zM>#3PV)CP+T4N4~q$jnO1&H7eXgUvU zzgR}hQ=nOhOaJ>K@ySwkk$8%B_#aKQbdr+i%*Wr~(P5Q+!cwYUn5~E&>$2yURni?k{{dgON41f<#@v_DaYSC!I;%GJ z>_aWHEk190_efV=enaapOrh!5J{;UqKoS!E_YDTGK%~u@@f~$6l%7&pYZR(>E)p+L zFOeOF30*%bw9H!T?^4PqztX0bxko@&bw~gvaB`gJ$@L1Qu1nj@6<*l)?|$TT^e||0 z^}7~`eoi&6lS`(SFnWk}_4$~|tH`3&ZI1J>Y#g{Xd=leh*JdBFg*ETr{ribKWq@d^PoH4gVWmdL37E z|BP?l^~V>Uzo001+c0rDjdSc2!zrI!bv3B;e!p1Ol(+V9O2xF?q3xLGA#l$R@J~5X zB6V@A8lrVl;7a#9ijC*i;SQBjSkua`RdO_xKyRrSlnSiQ^IzAH<{n-AXe1|bfimm= zFHxTu5U8AFKqV5Nm0~&tfJ0&Bw&LyS{q14e!7Fr&a?l@ zA&k#d_yW@Z$)l)hySPVon4VZRaE^n+w(KKKaA7U}wpkDKrrxrtVrA*?>nuo~ludc( z!6Cb6CF-9slxSzp4LC_W>1YwZtHv{Ul2};c+U)gzcMW)*$jgy05-te@Qp4R~RiO`v zE)qjc9%%ghC~6coc3qj9$k>w-9*-}2IyL5OOy3obvX~m>SHntuosSTJHEtKiB>?;?At^0+Z&VO(V_f>yZbK zTZwnebt~{_;{H1~4s-~kVY)hP#)o9SEk)3}OY{S_Pb|L++h<(MZ6YL*{{Hi=8ZT{5 z54-TM(?1&}2P#&+xhEk?ccq3!Vw`$DG}PE-$h=DYZM{k~3Vo@9)vkRjUpZiZZ78a# zn0O_=Dn?qDs<{7KGOjxb!((g5Dv_VkFcwmkraa^m-N=fD{dZ~vgX*s;IPLvdz)_%u z=2_$GGOvg36?oN9XZ>^^|4?H8Q+dYwd}pet`~7o}MESI&G=;vj$PbQWKG(ng3jWVZ-Kd)+%zIb11tsJuYLn}zuZhbb zvPTCd%!mR+Zo_{x&_F24SrGZwD)+tkd8;qd%)(YLSIJX-phS!{nz{QBZCB15a6G3p zeH9wPf-LA170{L|bLUfV(kh{i4>A_oInTccvvzXX{)>DVd;gGoUKQOBL;m z@VXu)RwigM)e%!eV|wv_&lm7E8-Om0d$niO6D#r^d3aN7EaR87aU*ZM>^Qi5oT|;0 zUIBZP)eSt~iz%{7Ey@78n!Wf0>tw4Ux)?CbMYHB2)$cUmt?`m)>L+cjXhVy)_sW4d zogZXfNK-xM5<<&bltbZQyG`Vtoc65>iB;)@M={DA{#jw#DY2t@9t{%qSGIoNI6cAs zT@^x?c@bh%KLVvvlB+lf@@-)Vz2E)uu*&9W)yS*DLtW$#u$(V*H1xgzAVhxM*AO8Z3suq;;5}|$p;RJ{uKXVl%3#mzh(4Nm(~Yl&gy3YsHR>K1$y9^ zx^Om452Hrq{ZMsS8EtU*bicHYJT75ZX=|73 zYk=?U|LzErhaqz==Qv$8v@*QV*sNdrwl1{fKVOK#zy2wN?RCU&#yHnKlDxv7u_)A6 zXxiq}X>7gxwI&;BsQdY9b*OuYC%5&(_h9ktFr}4dy-@k}>j$w&Oxg+h3 z_azP%uS}G1mB&r&Ip1#&@VFEhXx%1byoj(+F_SacaeyLm_6|r%rADt^tCu43cYbs{ z5o0K`E$3UhIdT<$H8*%Dd*aj8&X$eFXv8^5e#viHd2}ce7sm zD)&A!CPtG$C_F6j(^*|>(oCss$S#V0$Je)cv*Jy|eM1HNYqRVBadp*UQEp#Xa1;;) z1OY*sq0=Cg?gphBL{fU_1}R~rhLR2mX{3gs5fFw{8W}eh8+H0SeUkH7QLaW|V*bO~LTfc<6F~IN>SrY5x>-At!$EA;+S@LP{)#?>gd3yhx z0zs|++?m=i|FHQBM|q3{G~cWFWO|w(lX_08suZK~;u^#{;u^xYi%To8fa5gg#SEkU zvLg9q;}z?4p+1qb$MVlp*PsK2OpBj=OvhLwKdPsLtb0B`PC)l$Hp~guz~Ru&sHhO} z6PKkre&4fE^Vx4ad}<`T^=_;R5T+Z%>7HOs=e6@e#0~;DjkDOopHt614|1##6TGi}uZJ&oAvjN3% zR|+&D8MuE(qAfkIS{sKlTalAgW($6ORZ%LtUhX&8l(OxiPLYR3CNXseDeFeBM8QV) zkIgL3)>r1U7!bR@2i=mp-m7*Y;_)O{r3J8p0IinJ91VFi=OE=Sb*$ijcuVl?`ST(c#KV38b!hN zn*efS1?Vr^NqRY2yCxeJ+`XNop_h{+#Ex9LVOXhW2Pq&ty?eR|0Gi1Y;9T-TH?V9jG`IMZ)X@Ij&Wm(qRqckHCtf5xzXiy)|X)(pmwpaVPl z0?xo%m7Y?&GgI2ltNx+RouD^OmtgLRw^R8zfOwnYnPX)^_8UBDe)%XEO6nf#VC@3` zrYgy%fp9&VFH~%)F69HA(cU*D8 z`!feTl4wzX*PtR?>{b4x6Xx9&m=*>)C;duTn612U(DA-hgi4D`>o*j3S9HdB%(vMeXF2d0q&W06rbz>XOmPP)%ts2By9aA07cj4< zkfKwEgI|xhEzP-g8JTsbnWr7t{zKhvV5rP8(o$rg(;u7>VWUTXT^|T7GY66_SDsEI z&q_%)e&~AV>}YAS1jgb}cMTr&#>*3Q73DF5Q5*eRFOCynjvhFmeMP1dFn%*nY!ggR ze1ZU(cmu(yBj|NHOkz}^n>f~ay&=I60A4amWoTAf#fXT9;fJ2 zxl^}_eN47r=4Z4}1{Rc}D5@kkn5sQbBn7FKKz>d0&==Trjji9f`+5o+RpqdxG)7{e z^84z%Iz999>iE^#`jS~hK!oe3MSg2Ds*92bOZ>LMGcJlsgVOC0;KhBID65~=|6%I?-;y_~v(}{4~dKeiu2t_NUZMRY?h>3%b97dg*Aq9i7&*YVmo)|t8 z)kVXDW=o2`Y_!|^1qgk|#x$)OcZ-?6Sw+3isBlrMI#+9(Cw(O)V7vtMo5zlXf35~j z6sne03YsH(BbdyYm-9NeG|avlCPfbW+-z&$ON4dM|Hc4!+x!gO^BK zi9Y{%D9fufNrfsYJ}EMds#|Ib>WQ#p!|*vdu;CEt2~C+lIgxVo@X%Ph!wXLyZBW10 zt2NWW-sEvoH*VK=$f~ce558!o`7UrwG! zV5zSyaynOR8CkG>Fkl@Ogeo5lGzArci)T5H z=gwcWm)PH9nNR|83TR1nhM|>G$aYlDhf(5d0AGwYXPN&c_qrnc`3hct8Bai`jEqDE zy09w+&g7&6OVngrP|O%aTKGD1;v+I>d4e?Vc-TbJ{=w011Rz1Zx81WD5C&~i-KNg^ zEI8xF#WQ8YOgav$D%dp#CSnikAG<$ld3_V8_u@qd;7o7rYvypAnq#~>OrCeR$|dPO zQ(C@1n&dxP`}sW5o~LldxAE2#acRMml#}fjB7qEXKKAx-enCNPab>rjxv6}Wa8e_` zH=0wt68VwUJp3vfC|L<-30{Fo0oUrc1=Ai&pWTn9^7(XOMutWV9_#BiqyywG6FFIw8*^!Qw>4qI>e5joI$$A)d2+cbfz#;f3nOQl)QsWl60kL{K{gwyXxkd4_ z)~E$|Nj`G#`ZnazIXI56xw_5hqV@=k))wUyypE>a!Z(KO zb}aQY7suMFQ2K`aW2m>cF038~q*c>7czPtqx%CJdw$sErPF9JpK#Q<_A@0N{u}gQb z1-%#_*AvznQ8p*74>*XQn99utDgX~*ABJk;1Xg;G%bDW+7cY?@)Ac9FpYlsg1}Uj@ zTD%|po7Q*+4%>9CgC2IeEco)6$EnP=?>f1Y{-p%a(>!M}ATJZK@iK)bdF}L}^Vg>* zb{P*>ah&Vpi4o4?Wh#ZvtsmW_IwY?thlcr?BGm1tqQPC=hNnjFpBkoA^-->zb%-)e zeVP!n_nB5J(mp^kPJR1-xt_@(P?k!c{;)B0;M_ct)>z_;{^uuO=_U`6hsNA%kVEN)7gI=5I~PaNMyPC)>fnn5 z+q-j5vnk$Zj*W4F+E{^asB(FC<}*nC=SJ|vrp4vyR*{sqLo z+Dtxzx=tY%?#A#0x|5I3_^*XNV`}Cy#r+YYWzw`9 zzd>f*ekMeBD8tEW#^ASjQC#y=6>URGYLyyEG6|VY9dC(v+}SsFU(Rb_IHEeA0Jq;% z+w=eA72~@BG%%H8@_j}OA4S#FpNg2RObD;wXL-;!k{f(!#BtwjkHIs&M?tz0#`i&x z#O8=^qLh$Wm&ZoU4*AYGk$e(SG2^9J7+$>y>}f!c6q?bfJNyds7Bc+g#kFug4f@DGiWNY?J;Ilh^Kte_$u^u9PfI?OYK0&LDCL&6<%GeN2?A5= zYW|RmuZ66L_l5>K=Z3j3PGK&{=lekA)dv$S?KWAsLx*5j&byndvdg+FORs4_zPTXo znHVe(YBL`2&x(MkPz2>TD6clZkrO8BR*GPcWw^Q{7YgGqojr1)03ED^LhFLBvC)_v zqgn`A=RP{i|ES;82#4ACZWhN^nBbgtkL`C%10<0GL?B9bbGE#m0&`0q>?XWuNZtGe z4`%Oa$JzN_c2lnk2OiQPbiq??2m0^fxMHgEdCgN+9`48q3k)bVw1vDuT5JGZQFik% zSehPA76MI?jeP;s6i*4F`*QI6_$4fuWY7$5N=MFC9|Y^FCy$80r${T?>JXRj_DXKS zMqgjc341?CsjjT7RBLc@epbcUTwBLJou2wJxW)Oa%cfF;bMBMnS4yHu)z*P{6V4|} zvasJo@NWPbSHxrl#$l-+m4m6F!Lcj>#xW})Z^-+kBv>P+U>$qxhIF6vtm*MUbmWSHk8Msy;4(8s7D`kWXP$Wgc|ftA_hQ*aU*%*U>*5pPr(pp zlFO)>SHx+;c{=FHlIGIeuEMx)GwPfuosJj0YX=RlDg2l${+pEI{7*NCPB&YbPCyt! zNC6#Jz?cS*@>QsrGQN24?dBfj=UZDs;Ox_MKB{Rceddl1 z4%`nGxys7I_n^=Mg9b(U#Q!XCoF9`AXet;q4JGdg-epa9gBr5R23_xkL z+|KfkPsR+Hf>AR~9Md%_`Gh5bvZXAIVwR@`P41B1!|#&2v(4TPB|>JtZ(wr|{sc$f zkNz&-?6>8n2ZSxAjbEz!U}D@`YWN=aUd^?fxUbu#*^rO#G9?K!Mt}qSf{ngP&a7R36aI^c4!$f zWuTN;8-SmheUx*&bN(Yn2PU%z5@3lz4bXgZxA|{;J;IiqQhdg@Glya$e2Tj$)#fA!Q(b~;XV}xAugGI1$U*y$^Dyy%m9rMSvcRX@FBCjp- zzb0~c|D=0iP3CF$&ZwDxvTq}Ys%RK{YQRvH5sx>hC6u}$BMfb%R1GUOvUMeiX{Z91 z8P>-RRRU*c(?y?^{09>!`1sDhoN>hC^%U=YocebyEvD!8{STvO{ zJ+W3Vi0|o*Z+I&)`w3sT{M%>Gd+8|*C+f<(2GB(T6}ZOw@;3L1OcR%mqRzmH#2%kCI$$q1CHGe`y#R`9DZ3QB1^Lpi8UfSZ3<=|lsJJ-LvcEnKMS$>3!kSO4TzI z5z7y9daiGKUl?g2ZJ6>%Rl(7xy2@-oqz)RJd$M}U8IR*=>+mYtjObXFQws&>*i}gP zyNpg`Igk^+u&!!`YpvVZ<4YaH4h`w;b?+@c{IimIq(b9XKo=u;3g+X#(3hNXH@?w@ z)0I8&NW^1p=dGXdltp%2R7!rwjPl?)C{jBq}_}*Ae2=im^Dr_2*HXQ$CN80Ov#OM&0Y9Bk85OYzwv6b%m zwFLh_7LfnAc>-pYJBpESxV6|KIN-AboaSFTg9TE3f* zQuK7Qh)YubHU@UDROS{%M9)((Cu2FPpJ|d0;5Ykzo*3JbCZMUhJU5o;Jgkc}x)$W` zQnDlI@6;60Y;O+bcU*2xf3G_$L5O;85PbS<&ChLfnpNklsrr{)@`|G9b zgMAX^3yKA#y5TP|seQK%S0PfPXriFCa@lC>sIU3lrB7Y-#G>%<2yt(P?G~M<&eiNf zDE~auYWi7rq>?IktWhQ15w05vWAu?gH5>1(LVS#SD^Jo`ACKRqI}v^SQXgmGciuk| z1D?qwIJlX+%Z<7_uNSPJWAh}Dx#ipWW{-P5y7e!1|Q*L0+kVA(v@E;&wZJJxA|EVj=n z2DQ22x-Cs`Qaq#?)%fU)Gw|J#F!ErxpCgV|7~D_Wjv3sv_5!hX%2M3rR}dV8N-C`V!d2vd*TD zkhW|*jk}enz?U|OFX-R@s`7bDdCc1Ap3d=SmZISf5p}4ZJCDEuI*wqtNFc%JIkLEV zj;Drz0OMY3Lr)+L%!{M@(1r7M#Dw?xU;iDTg98>uZkTV`wjfn$C+=5wDVIAe6}MXx z&ZHYK>8Llns!HY{Lb64?WayZZS4tmaB&Byl>@H}{jO&Vg$|&Ez;PbexNQYPxm$^7@?7Ixs_1jE{N|oa&DK260(rlx+f&EzI|HC0{CEty=cp^Qh!6 zWzNja==h<6Y+FvPV`uw&hJ@kR|E~GvO-yr!bnIE);-MQuda#t}X`>hh&3iXq1+id* zvXDOf#k6UUvV$G7oL*6ZT!22s4TT4QLBkxA@$o_B2tSbEQ~92se;jsQ_FYngtSA&?-H6l zR3hxd|6-i_&f1XXUY8_fD^VZn#t6k%N+HlRvbW z-^bo0NQ;jj4xF{?IN6UANN8v1F{jP0ffLD2m3$PmD~ln`{%XB&0ervZ*GRpPmC1GZ z*T#r)UfrFfx=3A+H{-YPL(HZ=s;yivy&0f?Bas*p1fl*dVEUPXlKaeG)#(=$zfAW# zP}-rsI+y9#BQ;{EbRq1T@!Lw`F}8Zcqx1{^@F-lfh^I=O|9L!vhuOWSO62YTsa(wo z%Y7@i1Sa#lrL=1+SKFd$O(zsz4xiq}Y$bQf^t6^8>;2|JSkmB813v=) z-*_-q!$$zP~HG2C(ft~2bnGdSOrkm}tiV)8}IbJhZ+-%jqo%Mq3c2EKB zo$H>x0KrSWCpO;Q&e^; zyjK-7oe`>yQES5&HL>}L2WBR%?kX2D3#*xDn9|ySvbu-@ow#Zf)yVoYA$Ct zlgF2gP=PqXVc0A9grN93`hewKQ4@C_Om?Dp%Y`W_X3n;10n8;=Y*8Bi?E~AepUI0Y z{;KZhG907pK@)X(d|dEY17Yj?N!51|Z1D*f-_b<$Q^W{pj~~_;A$M?-5#|;JC2`cu z^NK|0{e$tDNiN@GCO>@}`xzw~t*|IOd>ird?oU2%={$H-+?#pmdHzLJ3~a$0n2>7N zU78XV+XQqgmq)Q-Id3N*KNt(hi%OI&SI}rjc$sx)D^zwHl$%l!N)s*Iy+&0Q?Dm@c z4?p3Dj9tHet;M5q?Pq?{Qa8ZZcInqsv z_vlNneQ{iXZP*WcJ$_?GbOu&qeG(z~Ue&M|CZ;BmoIdc8^v8Ae<&N2w*Iwo5l&5MX z@fHeU>F(!AfvNQvbG-Vo7MJZQ*y$>Sy zQ|%j5ShWXnnh>@u&DM)}LAMe*aac~aOMYaVkc63n#qXY?Ia6kh>2fUUF876Ls&4xJ z_@EXA6Ksu%z>TfasjP$YaQ%e=AJRVf?SyXV>C?Nkk!$mJx!8S|e24ox5@RV66h#xz zlan>aD$C@=9b1=&_QZb2fm_5S30{s{+IbRA@HvCJRG`LUM8&Yc|8Q*p!WcqNuBqkO z^U5DjvKm&4COf-F;oo51H>X1og%2fJ^Tp!c)n1mq6Q(*e(O|LB7@GjZ1sZNHj4PtW zW|LQJ=!;mki|-^FBKEWjOJC?I_QvW6%FAvHhdR<~!vwT?>xK<}+>{2p;fG9x`;+HI z2?O66jS%rUuRZ=Ha`A~KS?lp;Q+}|>(V=TeE1SR@(!|7$kr85FPbK^ zlZW;a6FVx5;Y7lB{YZz;Ct`Li5Y}-#+QQ_}%VeZD@ZoF_k@PU;np_2t;YennRRbGj zOHB!vZjoeBvm@C~o_LQ8YCt8mWiU^^rV%1jkPSWEq!bY!aod+?y{iXJ3V?Px7^rj9G2dz?C!@dhD!R^p(0w|1AO|kV$MrEy_tw&WjQO$+&dfnXcc$D{0(%)c-@Q4trVdEG}})M}qOU$LlYm8xSf* zQwin^YrW))K_T>T0QpVNLwn$Y$~v}QIfg{oq_4oKze9|rSjKFP3RglENYCc?%gw=# za2Sv9#B0mcWDcg(&@(IO=t1tvshn?-eWiS|`z)s6wJ);B zSQvSB3@&poPb88>m5g_^C#%KSeq1yyfFQ<)<89Kxo!39djv*5V4#8Nhh}f~F{@l0I zN&8q}g@Fz^KZQqkQj%%?kXBG&$zFfpQ3jgF`{e^G9#EvsPXDZgW3y_L= z9jMKafAap-Z&zA1V0d>^`+k7op#&j2vQdPRO$%RPmsp6h!r(au&O(Li%V?)dHKOiD zzlM0yXv}}m5r~mYY0$eHRO}sC!FPm_YB7TPV$5Yf)QYrTy_?;LV~zNBTcYB%m`lH( zXX)~?s)XP=@V`!k%VtJ7LHJqvx*mEt-wF^Ly7bo2bHOk5@l!lB-;tngz<|LwujRnn zEHtCcn8Q`1yUC%anSj0-5a)jmI~%Z8{<)YaX#H$R;J-$j+eFdFX;Ac!k?=E9#DgJ6 zL`AE8cY7gFh7r+)v_9A=QA`@lqRwBT`j8aTel6xN(qx)5C3-!P*e!n{zRycgQC{&<->om@Dy%6TqqQk2p26`&>w7}QhKP8~Wz-601 zro#cC6L$bORxuawH%(0V&vbpT*24L7on)^VLGO{T797@Rc+T}J=th0+o`r7i@wFd% zKk_l(qucUZ`U|lsC@B~#r6F)wnQKE~MGV)UxA(0~OL5m3@&=>_6{A1{|CLq=#vfgr zu;{BU!o8(ShG9?4$<~D8asVi%PGXdHMp-fX+9z536r-A<&6w3=`fxef+H*aVxyDWP z1eQ@l+gkne$EUJU+TpFp1Eclkysdg|3U=h!>KE1VGq8huxqX^vA9cbnbdD{_o&I84 zLu1^3kmMRvx|TYC8XO1xD+>PYi8Tj1(&NoH0_=BiYRH5`Z@uOZ4)Uy%Tl?TbHzF5n z#jmN?MkiTl~gErTpam>D*?eYfteu=aclPda^WDJTpCGXKO19i5crRu-Q zImdL89?ZhGCsCvGID~^uWH$h)?qp=J4x_R?DJNg=D3dpSDd)%e_q@t~flNb|ly-`# zDM>adY_Ob9^WGhP0nco^nTa3RevU7`X=c*#{-xPc!z0j&F54-luQ&t$Js^4>VH604 zHF0D;v-egn(3jc|yECVmZHIsYUjYR62j&mQ#J7#Q@LtGL4P6hXe5MFK@8#gtG*xyQ=MVo?#`Pw(Jv>5H5?^2yI%Oiac5NZOfyY&kWI%4*%6rdCFB>6mEh{`^CnP#=cGo;;5jXs9LLGP}ME z#w3^p!=GxHV%jBCr{AkGJ;O)YXzUD=oMg1>>@sbxVoS2^YQGybri>qELg??-mMxf| z8Ek;rksC?uc1#YI325Y88jv}7u9)q(I3E8KLqIr{i(?Y4mP>f_+rDEx>rvC{B8mbhvdZC`V763)=QAb<65IrfXq-8 zJYIGE@nQ8Wr!%pa>!(%olS4&Y$jm!9cx=+2!K3A?QHuO*TANP_S#ff!y4h}>ZYfMaY@Co5t8R$>AZga8cjFBl`9M~GliGz{~LlK>l(x0CCDQCY( z9$svEI!u9iB-Ntw-HMmmhvc; zRS}#`*7|E2lSh+@gX}ak%%+Tv7Ir*;@yBZ^>)U{Xd2UjQALW6YE&mAi>5;n?5iL6HhW|GcE8u zLE!*aqr5Qa7mDS!PnhZohwzKvO@}->py}{~lnp*pV-{1Bssfq%vGJW|O1aeVF(B%# z6)=N|7!LJ@4k~LEioZT6nw(v(I3t$|2 z;unt%1hr%EPrtXq8VHeR71x!b!(3Thi?+hyUxZcE&;Y6MZ`lBWt(Q)h zy6DnKm{KZnt-)%BF_eJ1WP`UACQM$J6yBDOx-@vSx^fxs8v{5Ua}mPk7m z7Dn`sAhG1F(qGOj;ebS0A!T~nu`;2v12wv=bHAE$>yy!7NN)oy%fJ!*Z>s<|v&SH3 z4ORyGky5Q-b1kQ`VG}2L;zqX12+Q2d%1H#PvhwtUsl_#=Z&Fj81nw?_Z2kRQp<14c zHs@77x((hs5f%enXB{0LF6=!`hbxE<7ADIOzDmH1Bx)>=Ic9$0GuTvWl5z;VAZ@iF ze<0vuv!OPFGb^+rv1GwmswExww$jkoNnigcnuqPD1nX&OX%qJ7NZJ5=5%-xFhWU%;xkUXd)Zb;OeYI`i z?i!)de;NzM^)Y$7;_OFXbFb}~Ti1^d^5-;Lj}`97wxO)5s>aFQ#L;rP%41HJl#&eM@W`-UmA4JQ?MEEu9OKC~C2(d+k@t!@Ow9DO)w{$DM2s=bq zJhR`Z3LmZEOCL6Avr6G{OR9Bb0od|c8AhB{k&bZNRNB;Qvae&e535mhzSiG32?cs5 zy5|ok8tDUMqFlq<+5!dgDkk7poA!S7$0PQ>PahzgJx(gdC%x3yH#Dz|c~c0g|Io$8 zEio6x7fx$6B+sTpe6kzf7`?HhrhIUb2A%tP_-UVIi)=DVQM=q41XT4FY1V+qtpL0Y zgLPKl@#DuOx2|=**7}xw@7qwPU$Msz;*1nBUyiERxWRv%@V1=Bw&{>=E z!Q`j6U0$hvObHr1BKInb1(8^oL^uss)AggzC-EaP>C!%Gob+gQ9_mD-Sf#r2k^X&e zuj6vN1t&;Kz-MMWZoAgO;&og7uO`Q6m0jPH;KiHy;s@qn8v4F;$agsg_m{bw9D7joUUCVif<&#hw5&!yE zzwEn>Z$!uKsG_RmAF(E{yid-btWpZ2W)Qyh|;`A$K!r5wquB8yJbrhgED=2 z-8nJ^y_-!dKqFon%5%lDj-dA@D{bbQ;Zt9Z9`?lOa2a0Qe4oHd;4feEh0ZeqdH;6S2y2q8x z;8%gTwQR1Ia|j>`(0^MWPfn40xI;u~zLP3nTGOR3&NJmnK})FziYqq+y_o2gR>;5_ z-?K$DLEjD=yqdWcde3 zB=MZ=l{p|P<}Ad4pZ5=v@z5C0iO=9urEyc;P713heq>tOMP76}hxzM05D+gt+pOZ{ z;DFv_y!^p7AvMoDoGXA8P3_SHPQXQL>)Tk{OLw(hcAD>`5NzHSQba&VY|@tNQDai~aCw~D42K^`5g?1m@=xR9V1K9S>ojB>2Ks7} z8Xp`Vv?KlCTlGN(`_g4kqPAj1Azy$o@uoUNZblg&KNw#aIfZmkVYC~mPlyA|bf4)1 zd}mZ@R@g}ViV3Zs@M|GC3>V`0YT373EBRfD+06>E$LG7JJ_C_VvIWuMbBFG+26SxE z0)vf@pZAK*&f|+PySh;a4{$o4?M(7@tnt~e$KnRD#|{AR9-FCkc|O<2x}-c(($GjCZIdH_wB5W6r0r~+*h|!P+HQ-v zC0NgPiZU!%RUT0WEg_f(Jf(az)J-j;z@`8uM!a6rcY;p-1t-+F_Yn4hJ_Fj2`!oV{d zofi=Cno5{_uKl=txM&{1LzP(~C&Zix;&}H2_Z$`_)1mXnjaQf|NWdE+R3gVd=t z!Mm%S`?=kcECNip%>-bW#hhkC#&3zw3j6Z}G9_mDG5a)ia$YYJqWD`1$Bb;!8lE>y z#?-oU<8Ag#@Gsq~H>&-aIn9$a0D zT{aPw11VCc)?eUo9(f$z**2*TvH)tCB!PuqA3sHlLpSa zhnvOE=84myFcY4!H5f@h0W3cO9ysoNEU-qIsgh+2HrWL;lEB06dcVP28Omfa&1=e; zOQ&25TME_WMWjd|aPDv(rmyzL?AkI1sPqD_v-w@<{8&l2eBb1~qp=g^d*#*PYP~-q z)LhGK*yL&8{Yhk|;qIhz{u_mxaS>~q{mMH==kvhzsDbmvF^+`%104j~ZkSp2Ygkcv z#*DsQ3mK>r2K-U9&zb_C>uBR@+!Ym0m;3;(B4Ob-OLS|9{?a_w?q#|cJD%M2ID)t_ zjXQqP8sl8Zl1-;xio%Z=^GyTsl@|%<^l{Jid|@Osd*lP1$3iSBy-BGnSJTdO^gs3q&fc(b+VLwaeR##ciZ=NiU6ah`0ONk$EZAnQP!!y#;NG+C@#3;5v zGtzH`J430j)|Wc$ea`ec&bEXr2X2qt1OEqi`mT%dk)`ek$4$*S@$|xEoYe|Jt!J1Z zx&Y=E9rMON_{=}3>z>Lna4AJGM%9?e|F&K!GG_ATtd+EH@Y{QS?Nmu(BUbuu%dn}b z$UiWr*dNd4LX&BcaYjA^G5YF}fo3f$?oZL{i7!b^(W6YiW2Vi9saw1EMz3X$+V@~B zi{~Wja&c?1FtC`xcXud*H0o+UgCvUTwcJxuNxMJhxU@L{ZAPZvkhlY+nD1}>qhI=y z*=@!^ZbA`5!@>Sd#=94Z!}y5(=GWy*laH?q`xsQVxt`a*c9{rh83tAP zTR8Rmk%gXNA@3lXPuDBwdPup}D^vHW7l(ss>HfqfO?g1=2~3*GB!KH)Pj<;uYbibD zyP(wxQX|)(lkXVoQT8OoO7=m15Vh8o2nRYH!||esj$*R`B1&S!uUZ% z>D>L%aM3ykiv<%Y71!GL`5SAx+hHiykox#Kp((xYHv_^)a+{=UCZV)&XHX9$q(&%#t7Ixt{<$$Y>*z; zEtv)3=0U*5fa}C%9FLjE0*sR1zYpjdfQ8$DPl5^8q}j4uQzTl}aN9G9C+ zB5b7sW3-eUb0=vtmw?9trH;7&ZwY100P2M`*h3xh0`IW`BRxAo-Y_o(NXG`3UnUu} zttK6$Xe5uyi!(X4)agtE-qB!Y>Icqfh{%E5qn70=oI(42Q4D#|cf$#9`&GLRNp)0- zq$GgLb(ady#Z-NY(6^j4m`R5dUm**h^>jrD(v;I-`fZ=haN*$}+VWPUMhOCwO_zsT zewz9Q@!2Qg;ZncV@xQn6V?;~IrlzTm;e_}BLt$A9xpj^FeXj|SJaqfn7z(1M#p5eq zET~i>ug{PpcP*ES&Prpx+CGedm1(oTYu^f?`*@ zjoh0@8niw_>vUB^cKG$zKUDnboWYD$1i4c3EC=Ub zJ#Oz%rlHvo-=YpGzF2_L@zH zQFl`_VaAQ@=;^~A;HTq9worRt9fhZDbSIyWW%wC}FKQnN*IF2K-a+byX6(EWChNXF zRG^5^rU#tI7c^>`Xu}6%UykJLnjRBUFIuZ>gx85;j0O;J1*{YT}J(_CCOca&lmww!snUCCH8)*#}kFrSNluHjAgBr9`B{1t%y{c){KELv3Ajd$=`6z`SDJiRqEwNi&w!HSFPW{$I#j?V9kbb{`$li*p)W+qCjboAX4u2^I^FkXH3qIdA| z5!!9@arCfS8nkO|G?Nwoo0|!GV=}rkCH%^>xt%3ADyW^HmBEE2<#a29y>qU-gegFO zJwf!b90hRGk6B?nX{#rSEX1JEV1H`7(6HL}nQ#R9``5IrY8LAUc_FsgQ_ef$kG`F6 z)d}vtettc{P`_Yqq30_^T%T~pMa@{zk|x(2h}#Rxp)J?FHZPEf98bFu!P%Nm&U>;0 zWt)Ct?cH#~r}L;N*L)=l<|3u@@F&ib|9h?M>Hcv6IexDtUWk;(6{HE&FbuK_f1`o= zaj(To5TxyE-37*ut1~S=n3lmB{d|8kuHNzb9h^)rmA{{SlM%N%t?qtbPprVpj?(d% z$0qyYu(Z*zM$&c$<|^qdqW(Ax({JYK*{WC1EqL_2`)0r`LST7*@Nw5Bq{`2H?IX_U zqdZ0*51zMjz8)TCBHP+9`&9w4#{XR%Hzwl?>@8si?MSbMmJs&{6pX^DzTU}Cn*}Yd zQNHLbu2F1jw=l7l;*)8EjPCydErwHR8Zt7xq-ASl_rOvH@sh&zjL%txvK^Zsic*ziWv{ z=&e}1tDWXSl?cA?RABfYiWbpkHU)1PfnHaVkMiX3f+O`7(#=jrwTl8IN{0xs=@3L& zU@rlc+ij-%o(!CLiiqXK)e&x9fxq^Trd8nf#tECfVCjcq(o z566J|_^_7R|7o4SCBtZ@Zhy%gc`zTM6)(-q<9(E@ufTX8^!1BjZV+_mt)+SzT*&Xm z`!xf%d2Ytv>&l~##?BpW2|`S_mb^6p-u!Ur`{2q`Y*&|2Lz6m0qJ-7J`von|P>pAg zcgA&?>%UbWWYKiab<5knkawIJ8^D#QYn6P2*qUg&+0Phx)Y`WdAgB^lS2VPh|MQmmZb*T;UcWyx-JLmv+C6lv z;Qd+?$-f%BGz5%4^(Q~+K@@gptx;j<@HcqcK$}6$HTPqwb^@F4eq-b+YnQ$^##KM!Y(FMjVhE;V(wtY0cE|V5$EaEL~*4q_3*H!^C^N= zf?26Ee6B$iet(-2Mw4a+sb1>7$VaXPS!E|{S9q7C+qLWBc=&Uq+!GTP_ZNNZYlVBu3 z<*R4A(X*pgLR>1^A-J!Dr2x{OLj8nlm6s9&hk&qKf;+w&svcOmz)(NxOd81GVSa0M zzVdp&dX^W{g9zr`ZnB|jw5f?!F(c>(J_;)l1D&Wvc*fcBIyxVi)mW1ZbNZTUM+K zRRT_!UyMajWA*GU=)DoTJBkk{dF8Bn&kVT|A|;orz&21Q@@TV%xD<`|d#nN}%k+U0UBG!6yZ%gd7K-A(^B_IqPc-u3N z5Mp;VPycGdSqRiTtQ%qhzoirTswDS`Pe`l51@t@YB()`9R=+P@18AUf$z zUhdbt$}Rr(>22m|&>Oz*>gU1<=(Xd|q&ka(w(?1VjZ#u?-mxH*nJ^PN(!O^Xm^#Wm z0-hh?2HEV)6yM(|)lgSD0o(HfUhkh1JR15UE+aWp?lTMcw})<)$kM^|DB-F2X&i_s zr>sdm-x&su0JgjJ1mhq~tWXF-_2Maiyb+|mN6CF3x9n5UHj3J#cV8UpmMlPW?b7*+ zLwrmUC1_O0K+|7#Ntlen`+NXGKJej$c7}kRTSCh@Q(Vp+rYg)B2ldn>?e( zM~AP~e#}YChm7a*aH|1eI_tvr5f2~nu)VI|qF_Ayl9 z9415Z&lJ%|U=rG9^>*w=McwYZT3ChH;(HZHT#yc}$no*C{->(`)Rzm!2pt1llI6RJ ztTfkSYEKEvjkxm>vb(OXP}&kj8IFM!DkzHLm_h651MkGpqVsA_(2FFOS~%4J(3DPR zFY>)M5i1E-gB@}@1g4lM^DPz=z4f8`J%ehWUl7$c)umg?`3B1xF3HGD{BH+&7>z4L zf$;kr*V-p9ob!~pwcdCor63L9THG`=C30o2?&en4<`n+O)M9Ti4^;5n^IW_WaJ~D+ zoS38D;7!fqfSnMiZn>JRVjIp%*3=hXwknPOu2(zSYu2Ut<;f{uD~~Q>x;^!qMdLL* zbOf)j>&GsaKpcwXdg3j=G7g1ZoYCf#pcNUv3wTGF$XUbk#|+o6-k{PKj4&1VM6?~l zRj*}}el70*N}PZQI^%A0?e|YLe|`{p*Awqf0cV8ammmW_D`#+KzqZCih;5U2A_z9g ztnTEks6$pXm5q*G{f2%Awg6`Lv-hQ8Uk;J%eF8#c5lpl-IzM27`ggOf2_piNiIz{v zblHmFR7a)J2YN#ro?<@FiF^9+5@Oj9qM3}KuV2$PC(~pB&;Gp>``Yv~2^bxQ>ui+Kk+{7*={*lW&6eqhn>o zwrbfrc+^$BryfxDO(l9Z?lx#EKR=&@8+ya(vrXivh%cq+v#Htt!qEQyOy0!5kr`7B zdvg9>81r|iEmG-So6?p{Yba{V!wV@#y8(en)4(56%m_kB=1N5KCei;sigasz5K99B zHx}2W0~4D_N_lNOhs->x1-d0MC5%?asut`-GhQJf+IW?>f_f$zy%)cUQz7o8g(Jr{G?h9x4m+ol`P|L<{r zYYV>S$ck8h9_Jm8KodDtpn_ELAa|OEKhkuLc)Su0ct;zb{6&Kx^qony1Ow z3A{XjU6#!xE%)y(JCN$QHEZ)@jPuTGP6U#r8=mctXw`~h7s!?aRWP< zAP@9_)`@a&^?wxQCHWj{$v2s~<1cMKHdQ|xRhBUl;TC1L2RpKy=}Mm@K|$+Q4aM#t z={KL;G}qp+vq@<1UeQ!?q5_?GnVR;!S}X^wjoRB9C#*Ve2ZhDFQSUT!2e;dPXWzik zjldU>b8y?WKW=R3!8ed}GANOmazVczbi>oxiPHTeIE+kRj;eW9%2^2(PF- zvO5U+!HmD5>$*JZPy1U+zD1eJ_5jLhBi02X&6h=Tl9Y8{5r5o9865tpcT(%oGA6BT zKYRmSqBXK>m4}H&1AFsnZ#bhBkOT6LpS^QXZ3&eJ9DZUJ=|GXhBRUnX@_QC; zx@T5UOF#O{j%lfT?-O!Cb7FjUY5(=rRq1%ehnLz3FS={i_Yn0Kpao=%(@_t7Zy)O^@3 z>n8CF_vrExN&trV+Gjt%U3AH53%drDv-6sXKs|2n9g+sU)Ey!YlA~IkiB7&RLP9a& zt;@@Q;_uE>VcIPWbHV}+c>n$WbZHvG=Af6NX0lA{E$h;bgfoZ+(0r+zsiO@8|7C`} zK{zxKA9no1C7bVJ%H~t)+3DQ>M8^Oc_dcV_Z=d+ENMKszKZAwz$qS5Jm`gzsP-(}^ z7+>vQkPLnnhA~K%1`A6=Qrcybv%oc`E85K+LzwV~eEb&Ur z>!1$@Ad{ehkFsGq4=u3nCELr<3Py(5D`@i#8J5P++}hYie|9?jUJPcs=)Y!P%x_nP zWGp*TG36ILgvmVe)f{SGP&Nora&|rGTm`=+hXcZluo6(#G(E$zY``F(@G8v{IuA58 zl@rDPO%jlKboBlqZ_7YxX3~5+0?|~{n8YhVPdzl3i4muVQgu<2>?zWDo4CEdQ2J-N zbf>=6m=3TPt{z{qh%o%u*yq%vI#z+cxYv(qyl%#V;_}@brk5nP>BO6-N!paibrLH8 zRg`UcfziNZ5;xgdBSzd(D}R%gAK?G)xF=}1**@4{dOUAuhn6?jfepP+n8a8`??@h} z@6A+$ro0ikzk255a21%uHi2*yzrc=9`~kXmJYLnbrwm~!{C&oZn%saMFx~kMh0v{k zR>SJ&fS;y=ul7C*s7>2u6CYDhF~T^1nbNtjJp#__oejf>b1^pPfa&u46xNpwNC4F} zbFSIeMMuMr!JehE&_X4TTFsj&JI5I$(n~jwJ2<$2Fv% zefxfZl0-#qBc7QN7$Z1NV(&_P2CUWn+^@!m8SKY>L3ekLw|qUDkJlW?-?Hw13g5%+ z!Z!lKpq7GnTCByTc82e_*d{*S%;2I>@-%o<`%oJmX%}vq-MvS?(@;S6EKi;HV-Mfy z188zv3Hcw>#n(R?<>G)!wvQXZ@QTLPE&Q^GLE#lli`%GSuF=r}>ejEBQNw=Pi)x zs8a!u7t$XQgn^j}H07I1@p`9N2NF5yd(8K)f#_N++t^3w?yU3Ud-|HOZiT18^9bYZlu?ap`YP> zViD1Z!RvKG;sI=VO%L6fjE4vaJr}@6?$ATM<=t-cJ~wl(aNhj&P&VIimK)rbQQkau zkzG9JQOCJ%MP$5oc|K+~U&fnH6+)ARhkCg#=W3~tlYS{l@M%f@h(|z!)@`&pG8~vh z2XT2!WpxAqnM3aQ!oot{*pCT9pjaXg^UB>F!Eh+0PF%BP$)~fvn(?`%xmmy#g$(~{ z&g5N`gZ5>U1lv=kyh4rB$0_WS(|?26U$pqFU4Vl{2ObTk2ZwoZWDSNdMI5M^e#j4% z7jK=4!y`~#w$ygXbgh@-<^7t{A=)zZ7qUl-EwrQF3y`RX|MflW-!08to zumr=P776OghiF!tNeP^6onKwruIzbpfT4*yd%V}6(GdzI_Zzy-;REGbz!5|+{ zi_>RD2NxfWe)C$X4U29P##mtFI4@$FT;o&i3)Ad{({PL`jT-41;0dfG>U>9#o>3A) zz@-=`U%}P%TKAXHY)-UWhlee(3N&Fi|A$kVi3R!`6(E0fY}@nAY3HCOh5=3{de}px zBS~Rbn9D)nn2f=BO%pHlyrd<$N&?zVjC?JTWAsP^9(j`ul~<#M!xuLZvTdRk-|Crc z#8e|38bJ;->+Hp@JQ6!3Fq|EGuxj>>mx5f=3fB_V9S|=$FZgM`Sm%c__zdXF7?pkN z80;(C&1>Q{2$y-2S!%T5o?0x$P*6Hn*4BfdzkpbKEmw&^aTgpOeE-q|l(0|cwTY!? zhV2FvjD(zy;0b57)nvx_>y4-Mn+Z*`H3Y|xS4?7TD_QxPPp{Lk!u~>){`yBM`H@xe z;(b#AXa+}QO3jE(-NyJ@))pRb338I&ot{rvcZ0sLXV9ZMfMaBxQzp-L_yu?Yeb|OPB=m zbKJ62rf+n+HvtgX!B(9Dal)vU%77GS4@hx&NfB4TK~Tr2wvZRw}7{2#*&RI0XE`?kJ`k_>yzzXwD;8d^Bl>x4dCz=1xskvT1f1+;yI z8I8Ah5$ZHCMC>HT8OmQ{u1g)sOM7bs0fTNm6KzRGG`t{YMxZm?`>p%(@^T&Zk$TgEQut~$aYWLIWS?*Ono7+GCIY>W0Mcyxjqw$x;#IcT%CVX7& zZUk}aOqN128>lv`mSvF%a=eqpzC)EURdQ17;|zfQ^pbRy-evs>G7m$g>0^FVr4oZE$L7A5)V)Aa&>!}GCkhTjYY{GNJO_Z>$rW+^f*YEbs+5m zbf_=emO-v$_uTwVo!RrE1V+@E?;R4_Ej4M0!XhH%4kBDdf{i7KFDq;!+{z;ciLidp z&QjJWJGM({#yNC(lu|E#LQXbATFY*{i?bS64qlcK0zMKEtgx?^*IwK5qq85Rj;<{5 z&&=iYAFg6nTO{bS(a0SB)yl3FJD?CwfFBKj*+(!OT5mT_vm&G&t$ zp=ojAU46R{18Gx0*#Vf_Gw%-c^j?&;gw?w?ySWO=gkQEYd|zn0$4sy&``+Ly$K347P3^g+I5UH~zM`4i>l3TrJVmrTW1xVwuQTOH7c5`$a)J$T0m&b_Q}>;bxmB^w)LlDcqR=`a#dZWF#rmq!Q*0i6?t z&f#G`wi|jj^tQlIbu6&DCXCIj_tP%N^H@_va&g#9Ha z`h8@+)V6KGUnTRJA@2BijZX}Smd!rGR_F7*+xa~o?>7*SccDFftMHL~Q*rt0_u^!U zBzb}O2J=rDJa%h|bFL-clK?2l25Mw|a;waiJIi;RF%w9SsQ7_vYY#D6);a(*_LhsVc@@{sNJ z@rMM{;6HYD&xMO*@V8DYtWNzn1p1sKc{5L20fx}*AOw)EkQ7EySGo5_;Nn51{~6xRP9$wvib zAU(r4M-EVzi8rgpTvwWn_R=d?j(5z$InVB`vmeiZ$?fCfqrqM0TBjC9W-3hCS;f58OT#0K7Ip3B?_EcX7$*2+8;sB6W5i z0bP>hKRF7mP41ML&9$^liVx=37b?)N(;wE9e)H!VOB7Wcul^Y?QoK&Qmm5p5tEHUi zn_>;PWC1A!1T@9)kCr$?nrtT z130RE&Moetkc5BmhUQl?UGRsMm8h-rtwxK-JG}_kCc|6_6F5c7TZdXV6YO-s>?6K{ zI>>Cf&>2GCVaYWKsYKxuLg9M+=g2)JZVZH$YQ>r#~oJ zpPd)N{yi{z2ws}Yi7L~Boj|_7&96;^CK2p~tO*qvXgL#h>iT(zk?0H9`;M%LgYpIK zJUhQ$rr(nfX)pz;(o?VlaO|eUi;uRQh^Y?G?9a1k|ZJD~dw&GnT#9BS`B~VUU ze_$b@#0y%WLNE_9gxOv-a{~a0Bg}l!UUF%c^Gv!{da|+*w;n$MaQ?zg)?|D&Jvru7 za>x3E#HmI%s{!By!?ls&-t0g>GqKv{*;T^^*EhVMU`5UWfier zV9~dpptS+h_oI^q$;a|im*qkuilw4n4=p!c4}U7#!PHXc*DF4ncMr$+w*6uT>84)3 zMjm@w=8F6^q}EvNBb4hY+8G1^B`om!a^GPy7ouIGR=x9cy+oWT}y@fUau{=(++voG6G&@N2pNwW7zD zIK+{PD0`NOM@IgqTL!nFR8oGu6nnOa&GfE$xXx)hM;oSTs>!(T$X|7 zg07F*-&EC(EMJ%^w7cm$8{XTDdpsN;;F+Pwr+kNAouu!VnB?x={aEc`kxV+A3I${k zwN(WvOksXx3^O^HPyiVIB3Tv|yr&}i(7%)`%xe{~pFN{5bv77Kcl<8rr|HQ%hrS*G zv4rNtQ?afa25K4WI{^+Umfm_mE83J{j7MB=PAnxWZj|3#lh*$Jn&BJ=%C)$U2n`m+ zz;^X4IYY3LTl`n3y+sS4ac!2a1z2YWH=TvwE~WIcX)8#7Uet--^msVV!sVLX&-Yy% zql2aizkivmq$lWNDuM>do_gZX5vq+Dz4<3+iNBj(XMJqq%=L*ds?6%%%11jlTdCUP zUYe(j6%v%|p$b27X`a<6lW3BopFK7>J{-*RyJVCL4VuP3$pf$0RYd$D@Mwo-qFsZ& zWcdzWpJkyV$W{J!7q;8`Q{J31pEyGS4#{W)H(}CGp&xxR$1jdWyv*u_b-nWlAd5B% zlI1-0=PzuL$Q;mW5rgd1CQw+D6xM22lQmS3m6JX}Or2alLzdS93I^gUu` zL-KqA{YBFz%(AiG*OK$g(&J?d<5eDtF36bycd8^xt4$WlpGi`!YAHQY+DyJ1_;wiq zw7T+BX>U_a$48}SBQE}}5YnZ^(Q!w9OZbe$jV7{%10gNI{M7zg;jHTxr`}!ejGukz z=<_MM18sn6P|!qt%lm@cMKZBz5oByBRFKa)hk+=PsoSq4h86VRF!@Wvm|L*#8 zAlBSwW91)(ErLk9zPOnVdJfqV?TLV@8hijO#xTvmH=7I0aD(yI6{g08JaV8X(~jqJ z&5r6GY%Bl}^|Dz_K32RR8204pX9;I-NqNJdksnF)p;#it(&h@c_`jKuu+nY`Nw|gU zbQUYWifihH&1Q@Fan^3kY}us-yOm6{Mk{{3&ilGoZ7Ld|BMNC#8`>+)zIJs>P%)H`6NU#5e7ApzB};$JTGvtVxQ<8LNFEG$di-2pOg0`L z73ny7V4uv^^G6PNf-n;pU@U=4wD+^_99Kc zn$~3jf%C0=<_|cDl*UDzAOqx0v=rK86AXtvKGB$&zhvk4smI#!6(H*n(-{iyzF`JJXT zG{i~kbvVF)mY}d(ux!CVcy+&HAI|$g>s&jK6&!x2fk04+EYvx7XIRRb$>cQ(&E&-y zsAuz5a627rke3GC`_zQ<@x59PO3vfQrfnjRRZ2fW?`J#+?pUeE}v2p^l0vV}(Y#3bR|E@@>e2H9L{D!R?V-G~#ea#)OH~skC z7sBukbu~-FkS5Fv4?{^uav}Mw>B zUTw)cE*W4sE27e)g4z2JID7NZ&7SnlR3M#Wa*w=a6Pm!?#dB=jqAEI|3j`O-E=p^5 z>2&aZ5e$VZRp3GXKB*8;JEa=a{uisi_u+<|nJre=LCe zH)ia9Sit=Rjozu|*@De;ZYwrAkLFA@yH;~rR=C`~G*c1&z1UZ8aD#4(+u8+N-pT4t_cz$gf!b3zAzW#vbv*ARSuLD4g6ZfbSM%dC^zJ0HI(+I(M14|0*OJG zrQgMwpOrMScLNOau+d%cw$`Tzvw@>;fqzcye9nyQ>r_`hz4+-1Q7NO__srwmU<2vo zG9^Kqw?`K)KPs?n6_-}yWc9`0-RWUF6*q%IT9;eD7|xuq1B-Mvj?a=)bTQt)v?ORS z{jKxsaLStAYwp}|1~*Xx$j`*KtW&NAgT-XL2P>pRY0+yZ7KhC1+!qsSC6$w$Gg}W- z{|w^(o{if0NX+S-wBE74ANiQRLsfMF0J`+?fvDfH2uYQx4@4$aR&+!KE@H0W=!E}T z+SctSuvLM>K+#XXYnnomJo!~KzP~@_h4sXFwU6q%%w^S|I#Gi>%nLq@jH$z zC|&pX=rHOeBSfRxJRKkTlRlmdT)1GR#-X7CT!W>un^8@_vzlxYz~`F0{oYjUE{@t< z$Gwln5O{p(JF& z<1r(@L#v6e10^(~Lv+6KSbr)t8vTPBaCDH#uPOcyV zR%4nn>kHC1Ja1i^PVmjC&mbsYWwDd9EmKl8p1V0I!bZAoUVL@w=3ViBQYav!_n9wBXn_8)%O|V^eb_>B)VCJ3} z1KY};7`f$QpmdEcwqc5xr$$anhV--T(R)b56WxZ`Jq2)FJq@F7G zAMYw4!(V+@HSAf!L2mXsDhjp84ZWL>dg?LT*GdmF^jCeWk-B?ziYarXlgTGNCVp_# zce*E=ks)E~D=FL@aVtDxBtH@6)DF-hM}RINOE2bg$+f!?5ONQ)o*S$66a8#_|b3v&*7>j@PM!P%yhJKWc%;z2^;=JqlrdUXIJbuYHP1h`QwcT6W)=E!l_6wgd z%cZ1VKbpg(*%`;F(}9y4;=OBMIwdPT3qf#)CqDmFS7sZDyJXe1(%3!p=hSp~9yMh@ zlLv!40JSzeFO_I`HE+LnHu24h$r8Q9xw6mr ztD;AUupO$#nnQY==~g1G@8re;ZdAgey?wd5d0U+0e8i1t=Cy=vhV$^HOi)Q88gPHj+s&Y>MZWIR&5`;`4mT$6r_`I3`bbCzU5({_Z=*@CFO6AK zLugi>$j;>W9^wPapD=PE$Gs%oTod0~JR(roM#50cLu%rn-!y&wEAZZjC~eZeh_4fI_Wq%(kMHJGv^{a{X!96C?ZJyb5PbfbO<0nOvz@yJ zCz{r`Wd3Q=G{Q^?#)Qp&*yr&n(+KX9j41}m)@I#l0Zku`KAu%Xv?CxYN-rviV?6(w zqVIZUgYR5cWiXTOY`IMOR)Ts{ut>t7S!IkE#LG8^=9;kLiniF-jWLhk_81k;&lc|O zoAIvltL9Osc;Q!$H5MPe06(GxAtt*Tdi&xTAIv*CW@a(O6XHTRd-Ndvjpd*(v(+Ys zemOqJ$Om7k+|IG#o#^9<-J{KEjo*ERRY2~G#*(~#oIY|VO|1bD{p)%`GTR1rJF3i0 zT-q^@2aVG!Ke%B;MEgAiefsvHc!AIDjh0<;g?$&v`)1`|HKMkY&fh`_@?F`@PdtaI=h8@XcnmqJevffKQGw0HpKDmGm%e8*OQ?^$bT{r2DY5^m^s6&%(z-Uv$EuMm4|xw-j?7k+fq zy${XsO3ObHyLH^$(4YBZqTt3tX8S`z6t+n(E$p_s=m6|D84I=*N`yfYFx0_bA14~D z_Zb%V?|6fyx6pptesj=X<=00|yb8YpYwEl>`x}p|{ESSthxhQBsVHgwGZ(XeIsD+4 z<<&6I*9Fj59XpRzwGB|~-|W{z=bgHPQyF^N5zo<1yYZgjQf zEcxs_%=}3Wjqtp$I;Mste{|`N$kiIg*7*)S|LB);{m#0y*_6!AEvZMVxfzK$d3ufjndU$!2yDwWN`+JTFlu1MSTm3T!AL^;2z>_vqZ zg8Wwej+cup9bS?gQ{4S(wMhmpZ@m;^TkpeO@v0WRVx-V~suPcJYP^KMLrWBM3@g z5AEkEK!09*vps!qVm+Do<2(gSjqyqs+Y&c$Y5gWF#9b6{`QF9EfGD_oTPw;+x9**y z_bktT853`v*=2`LP!8TG&nqE&Wj%>VDSwcMv0lgTsZcy|geQeHJ6+Cd5y|vc|4VvgRyroO%Jxz}yhH#X% zPe0rqrsdr` z&bQb<)6ed2w@Nl1;dqZ#v6wJ&&BihKk)2P8yU9bBF=iWNm;<3Nxo2D8X;xi5QudfA z1GAN$fkim-{Jfib4ngzI)JlP>N*^1#Q=yFe&8~=vO?nkei%l2Ni38@J3)=&I>vaB_ zz3lUfn@_PKz!v+8e5-Ut>jrZY9yOf3f1+Nz8|WQ>eaW+i+TJ*O#j_26>9^(KKz-)$ zxKX>?4)10C5EJo0(wrM^0` zDz+y&0=yR`E6uCXiii-bt#1OD+RZInB-`&Ki^~dO*yaduy^D>XNY2>hn?|e|1MW{e zt>mgo1UEU-V%(COSU>_4C@elX$6(;57fgy&g>4TG)J3Tz-$(k3{2YlOnTIq}d+L8E zg1any*)KapLR`P<)qw&dj5udGub)^g#sywV=^pwNhV}Nj2^lD4$9{i8jy(aSV@Vynl5+E_P4A%#8mktTPrXSg$|+ zEv-1_#+O{_Lw4d095-~Kmv#%9>1QjV$1WqoEvc)6p2yQHvy&Q$tBNk$gDoEC+rvZY z*ICC)=X;-Dl%+y{7@elCZkok&e>K;i3|7wtadvT&z8xaM97%nytNVRTkY`vD+iLB+ zJb)lWF$1?V70>zb8}W>|Hi{8$#sb+_^mq)+7q<&^6kp#k3eW_8h}CBjg?ggKyFMqn zmNu7RPU2eFt+75;m586Gl-m%Ru36Sg`lWGCUrkD?otB)XiY6NW^VMtnq~9KFBg6XmPB6fKtdkWC7-nMhJ-=zdS3w`8JQwa~^P!1|ExMGg<-USyXU>uZkIe3|G&omgnZKKFPlODUtx=3qXf>*zp*c zom+l15ZB47Wo&$Oo+k(%tVHi33KoH*yx9gbJjqe)813SW4(f2^W76#%O$>j`)bFQ;G#g3 zGZnFr)$E|(cDBzlu{cThldjwGVX1+Bqb=)d=ngs1ld9Z_i=ej-b9J z=OvoPae4F6oG`Rb`}A*@za{J~53jFoI`v2+>tEPI=Zw2ewNX^6(RTq)W0&^xR|$7= zGv4myfB%B}8IQkR<34%dv_I_=OgoD?>QDP=60xmY4v5v~0X1CVR}HDtuXV(t`(>yqVYn=(PKYq?< zNyLhzh~6E|ihKHKY_Hrok?LqC*zgM}CuqENRY19k`Fxkenn6o{jkS`Gd9v(2UePSe z`U>*=k6mTAdP^Y>|0_q*67n-tY&VL+jmzCrQC}Gp=hG1JjN;5Z)u)$3N?rxMb7an7 z>0gpn)lH}$mwLGD^>_%mAU5~4plVq7bzy!;6_I%uW#VZ%p-C`?&Spb%REL1AVL@uv z$S0y=Inh_LEBUHjN=CR8(zqWt=fA2n<7BfwF$+x0Op613YMq7Azg$6z&G2~sy>7^^ z`WKe9<<*73j+4LL?}Lk+cOe_SSx%JC6|>?YbzK37+$NnU!~H= zc$}57F0vb8EB{e&9H{dUz5Z2hFc1Dr$KLONY3hfnrN_cYk(6YG!p*?wwKq5(?+431 z3e?~xP-OBSYSIX{!7p!5y}>_Us`tIhot>I{T$Wj_w4h@0HZE07P#-VKu7gWAhNcK) z&M}&x6pG!d=j~>1<3Djh>7xp+YA$hjgMqrbY+koRJCy&HdD2i*-yoe;kGCjb|F#Lb z;vw_K9izH1Qf@>WRRC2oa@}ZO(6nh@$#(11x@(PF^DK4jd2B;dFT1?R9^tfQN>Rch zzvo`WW@ApPlg&%pO|CkW=%@(O_;j+5K~7XM89}kKthNVzD#w|rt4&KGKS+oeKSjuz z;=)>E)hdDM#7PiFurcY(Qk;A8M8MpPzjmtQ>Py5^kl}Tn2ZBeQJ?hUsAiC^3DWiVGRK-u z58KeuY7) zFOT%(V0%g=p&og8V+q}PIHu#f5Sj#xu^Z=Daj6gyEo=~UXwb#T!m!KCVV7j;Q&?3o zPrp~R_@k%Qz?v5efkX;C|wD%|CosA-73c0kt{ZW~jy{1GPCTJ@_4D(sOTDGN%+KfOt3+#f1?ysAbi@_BiErUIwNU3*% z8Htx0+o$?V%#hUkNP-k->C#hricVJPfLVo4BCN*=Ra9hYFggMw&;|7TLez? zJhB|AV6^wCThHjV-f7S8zcbm<u+dTR@;(*k1tC1V@pn}rry_AUfk+=#Ynf;coi$h5H1^i=w=y8U)y`2cI(zN_&$?SJ! zOKDtvRzLZS!YwsR>Ot2wcsP2)Hr172dSc?GLcpI+?WIsIHy=+{+48NvakJ@DwgL9x zx*?&Z05b9WnYUSQMABSmVj7048)DREf1I8j|15jVL5=9qI(*nPfU%D|WOELQl?HID zqdZGk@*AF@41{_-sS`|5PBxJ@7zQQy}!S!pkNeG zoO-{K(|@;csA!fjBOMg1q1gQ0T~0UQx+5!b%7n8!g<4{-5Q!#U^r4=~|3zj%w&Sd> zpjIN{91Xr}r0jbgdhj_B;WMHirZOHPtUXOm?MF|NM8Rl=YMiBg>s)7!3OU%TZM4#e zMB$csH@W^o1l@+CAlyw5W!Q5xr{u+6;5i|UheiPObB^56i>iwtrBGWh#YK~LhTkvQ zv}~^$p9P1gP`iG;^2nqS2z1p1PlsZ9R1KFkD9cogbSH+8{q9~@5^*$sGQ`s15<#t~*S5&Xf&E9N8~_ z+_$oIQLH&a?S(CQ(X>!mjjlA2gMmYMyM?~AwUL+QlvxQ9s zM)yZVfge?Emm9%$W{)p27>-*WbB9Jo>F^KdDPu7XE@~fsp?v42IxHm@JnLG^EWET0 zopsT)q!AUN3k!{YAx&|g`v=L~J#*leR$Q64!F<>JHU84#Vfou$MfwhUv1dw-SF5y7 z90=;a9zNXA%gTN61Mi2T4tgy3%|1=xXq z@(~N%!~I@P({_CEw`Ruu13Fbn%=PuV^2Ar%HONIo17+$oxhIgHEl2|-e9C^*WtFHC zz1fdrC9yC4+tqC!jd3`F3hjfBeO}~~k`-u$Q;x*CKjqkRVphf7tT_CI3=uG;3r2s@ zDRht`2P37loS)>xHqu%QP|!R;VI5;5he{`4d&Ep!@r= zSAHLwTz0x(III=^O3FH<2bgnzet?4uOuH-b=xq={UoqLpUw0nEaS-(Cr`ahDW~VS0Gn=wwikPKxOX1o# zHedTaw*u*x$zG1X7JE>?SvfRieZmyX zWhkcAn#oMZStg!I80%Rr_v`3a_!N)tovb=?u#53NFt-sqdXGFeiD#+xBgBA#9) zGP6#ozO>8ZBGs~QAyVRt#-KThM_tnEJ>N1U_Kd~Fef%1c%!+}ba3VcolplUibMD+| zTB|F4JrNdY{m}g5<-~LBwg(4_>k6=Y(?ShqZynZnZrWybE1fuc^zkgN zB+)foJWpH8I&u4ZOj9(WHz5fY;!G}Dr3t{uiMr=cj95Bo#%Be2#8))5{1m zvQt4%X!hMYvQZ*hGKa!Pn6Q|-ry+9w?aX%P615OYIl@PlEXB8DmIsWQ7%*rU^Yihs zS10mLH$>g^S_Yu~g6rl(`}Ag%NFnO7oV%x&{hKMD#GUB1?qYOqzq>$~6S{W7hCCNg zb|)TgTJcOuhA}+Jb+hf-&#BA^6+Hs9Bx9G z+?$@5&}TSjuxAGi{h0~Dd&#?P-{dEKS|*Q5(SX(e%h|59w)F&czfK0CRY^7YnO3cB ztFpJerK_KEtmwIlC4VpyWZKuMWhCcS-6LIQ31p~xRc40@*#!KHGRNEZ$QgM{pt66* z?*^rvwz7A>#DA&=>jYM-{o|Hi=8-^M~GqW@2l>S}Ep5;^C*Wjw~%-g0eSFE!n_bAp|wkT0VVcWz1i2>)y$F zn99)h+u|kr;!btab11?cS*Gb+z0q{mle~A5{bSl=c=_M5or*0UdnDwHr0s*Y`Okw2%JU8>M8PESx<<&%{3@9Nqm|DDN1Jk_9E` z9`|;G?B!1Dg{N4$KYpA2U@>^^bC^tR&J}tBWAb0kRk8ewcSS# zi5p1QtirSQ%vE)bk~qla&3eb;cTxB`iZQFDvv}=zjGzrHYk*0 z;Z8iRQ*EK7$2w;`6_(%8(?hw5U!wZ4mZyk4+R;mn5OcNgykqoAgjX zZnR01RtB4} zv^I))C3^qdGyx;(@l_hvdtM$Ps+q`lp+?yAY!C4Zv$!+@9hN?&*^bw`R8$sa*wWk0 z>?|>uw%mQM@zWz6svCnQh|?Ei^^DAmg3$^zt+{>0~wPB&VH7Y%WS}J9yMM zl)DwZ7x&}{nczpR!#jvRqkx2Cfw{6p1dS}IuwP-j*6eKX!j<zUeV#TMj zClkSSxm$I|puPR0A4?aPTCEx)nIo*tU3kEkeY7T&&~q4KHafP~barA1^n5-%u^E|( z^b40cHtfTYg*(_$r?dGcANqA(80r{>-A_4NZK!9ZMGCYGVTMuGO$L>Eqw@&T3(~YE z>fE8<+tGgiWae&W-Xy*+XD}19Kc7Z4`%v8J?$jK7FRyB4HOAV0y%K5}QKKtGk$NHT zg??KelxDs5I!Gmf=NWfkJccWE^6=I1wEifRlGp<&K;FVZm}Geh` zP9oJeDmf7(&-ZwE+}$7ZqEgh(e(IT{1sA{jeH2aDQ2OI(b?b%@@^2L?A=9*jJ|7T; zl3PKX5>{NB=x-ASdQSfBY^r1V!&=$PocW?zAOCclh%BN}*Mbn<$(JMnAt7(SS;W*b z3dB@pfAHQ>92+s?=N|##N;p0?J}o6BhL?dJGy}J5oF0zpw6QlnOE3y#~3f6~&XjYA~cJTQQ)Iq<%bY<(*C6nm0 zkeeAC<4_)$n?b5*!!Ro1qV4?`@nNouslfWW>DL7WOLqC1Y5uz=Aza001|Q73ml08& zg3L8sT*9^3yQ?f5FflPv+uBu|g?CRh*)~6e!MaHwG046%f-u@gq<|yKv(F=rmgu)egCf}T z6vMPTHRh%UDSX*Acw+a8D~J<9_;}2tyr{o{!0770guNhYCSTb^$fAlF?#ToC>Le?& z>?ann+JD4g0iMc>d9Ui5?HKxUZD{>P@GuW^DaT^Axu=C0spSVEQtf7K?c(r`4^sg` z2L+TxMA1|^5kIP9kx!n7y_t*kxX@P^C}NjZv>T0_9mQ6*P2F;(M{d(OjBLM5c6)H?$gKE)^3cG5VUhD14r1rW z$4p>Kd+dG#X;RWeOXL`>cs(!GW~`Q%FI~XB^~*cIxMYG3>U*z}A83^y8q3t*Wl34%;K+c z5PaDBiqcq?z;|DKM}A`L4f&hT{F%J| zT#x+9H-_Y|z3S^NvyVD8UzGpp+pF@%$7#tR*yl;LorO7>`tG+S=}JFT;91Bf)jL>5 z9~(rX(_kF;RMqTHJ|q5>o`x$t{%eTfJ&F~45Hb>vgF!wjQH*PPMlSdTK)R@vweh!} z$TbagI~Ra!Yw*AUt5r>ak8(JvopuLKJH;gD9*<)Okh~6%ytW-Y5{Yo~8l zkg=-)ZZco@0?7&f>BfN49J*twYu3toP!@w`G`;!t*~3@{eV!Dw&Exnm2hZaK^YlM5 z)DpLhX@cC;BPWYGXj{;)q7*_az_+d3iuaWZ=IZ>WmidAS7V5J_p9PD&hZ=4NG+wvk zjBvVlyrwch<5uMlo`$j>wqCf;xi5ab1=I8Q*vN>CjSd6gb(q=I)SjDP0Gs^im^8-e zpz1S$fTHY(5X7ejy`AF3XMw;XXF6#46iU%(HaCwaum_qy`p0{VQ2`^UbC#auzWLecu=W!_Zg*BtfVFreTt|hNU8L z$mk6|C$YhsUiH;>9!hoX1Ml}yti-AaObPJTS~Z@OLhrDYQoiFlQ7K8XboTCJ2@amj za&wTp2mR!cCK}EJD!L%6GTJ0iUjqHp?qc3n^?{cL-DOhn$7vB9|f@;3(MXTd7h#%jvCiOCZt|~JE?p+>Q)U+QBg!O z8o`U2>tr$`X}!yBcT9VwK7D)#ZLm**y1DsjfV<8(oXRd0ofhqDQON{#c&AQ(GE`y( zG1AJwSQ-r5Ye%R{?#pEIz7r%@9~Sq-nn}*vACzcmDBm&s*j=JyI7ug2Zh=*L=iUSJ zca#88C+7~+gf(!SH9(M&z@i1Q(*(&C3n+MQg4bH^q6<3N8e|Wq=Wop>cl4(@h-zb=f7!CYPT?yhb*e{O7F!8&*6>M-`T9^-!4z%saktU6P@VpE(+NIdOKpr}fC-da}q(kAL zMizNci+qx-aFmZmS(=~z=GWxPIW#spn)AJ3(;U zKTpgjw3Ar_bOKeH%=e~z5aSz!7S=#m15bYqq~Ih?sp-9yJ`89FoWU(i9|!9z9?lAK zZX}U_hCU>Qa_!_+a8zvIy2Qr5>>Q7O>*X{+^7o~>sdPuZy$4%~19V8Len36ZjeB@O zs@=YWI}Rd15}?{yvw9vfiyMk9&MuW3TRx6%V_|ts<{=-bZ4K|lop;~Yo4oY|zIoln z!`CApC5w^sZEFXR^W;Ggojt3X(iLVJZCYxjH=dEno)xL(7Nj!$9WXra%jiH{{>Ik_ z<>y{&1j?DmM!F*Z&%fG`-~6ow`Dfo>mq&{}DMp{;X<09pYjStCD1Y*&x8;*R{1?*W z>M55|sgY|}LH-tb4T2IyFb zmiRU~R?kHw|D%Mg{@$Eyeh+Nugu4%mdI=0^wBM(}Y}>%^xfLDLleJ;H z{W0ye<8~Xk6W^oS@;b9N#%8LZW^=&1C)JI!86U&eNwX!Lv&E^Y-Mer@swyz|?gX80 z;r-KO(cnF_IaxBE`OW8{c5s}*=DDBc=L(Xuf~IHZ<@f*aJMvHdyZ>4K+3)^=`CEkt zBebvvPE!pKRJ6b(pZj#WHJjY9a@Cik9gE!7(XQZ~7)^lcNmUq=S{=v7Ro9QV!{1G15z z)WjLHC_uwR&X?DS0Q8FW`2F0>+L|oiyVp=FW%QqO(~tda9I?loW8(ParJzk**Uy2U zvg0-8=5@D$fCp!i1~cM*Es4NmH3=h_o>D6_2*#O=mgIcbiVP>6eqJs8Sjvz8Q0m1E z4=DeYZw?9d%7ex{e)z$T{NsPVDF5TPSL8kCrTW?zrE3tgDqLxZIDov~Vr7>u?~xSbNoRt7SR~tcPbj~V4MFnG?*!hX zXvuX7V-ZOpjzo3XIUes^E#dA%x+=!p+4cF(^L|QGnzloY%!PwkRWIS{>8$}rmwe6Z z5D~aLHx9kACQRdk0l58!oh;*K>2z(G;GOE@V%X$(&w?KueGDZ_G^5=Ch}FkNC8mT0 z%U~NQ2K=PH7eZGo57-GXN{~;wdK-+LNgP z&Jwx*;E`N7e-6(ImBroeJVxdY?>9*sG)`-<7je7@St=KWVn@*^g?Q7qT2`d`21G4?n-&) zyOMnN-;$C3xcu5r51W9VZ{NzuTc2fQZq@G-GT%I$0(WK#a(OH%&t7gEQv}c3$?sfd zOST^Wxm=2qnD>*N9MU!ondG_E1$lPkFJ#bl=vCp^NL{@q*Z#)elFZ#ZQtFCHGFvfk zND1;$8_+HvpSapK-=}?N1dgv*u5tEj4dbE8=IYa4^rxbHRkRK1HAq`n5SY%eB;K^T z4x*Cr{fW;Y$)*)+E+*+3YgTMAq#dWg~#HqDuNQ79Sl}^!V>^j>W(`K>0Tw@g) zCOO&{2JNh4Ph9cTOBC&46V>&cvuXW@f?f9g>PkUzO956|>D*8o914ZaeLY^AMKAj6wISJw#c?FF`KxfLGf z=#u++*u3F9?s?!I7G#nW?6U)ic8Bs>2-aElzLD42M|l{)QwJPf zTv-7NAAwJi?_rJ33|*gw;qLZC`kF zK+_sojUz3~o3j*ZaH_+GLz$D)}0w(##w{FY*hmXw60-htZ_BB8~ za5~TEa5+lppx{|KL!ecs$$;c|e?vR@HL#8;vA6Etlala`PprUVk02PB31uzD;}zbDNuz-BzHXf5;hIL*Gu31Pp4~VvyNTF-Q}I zF1eOnsb!|6vivp}y@#82PSCxU--HZ+f5sFXq+=ie42&o+9QU*aK)7BwTlXttqR9Jb*)EjFh7jeqaZ=X`uu)<4l^=078mUy2U>EdVf(EF zzxV34nluVIU}<#!QWqkVugMGm@&dqZ2PAKlIdY*ic_ge(7t2@*n^D zMfnH6JTA{qDm`YTBCGuFY(cUGzyI$Yyw8flCcd)9k4ql&j!11>yoq~102H0lNxyKY_ zlB+$~UZc#B3zF9q@i_|eRPW6j&O!Va&uu-8jz#tGu^s$OPySjRm_cdEZ^OJRLZw%M z=^T}DCJ)W|ro?t-ONrk`keorH`Nes1rk<=PcT(!~{@q{u)rP+j1ZWE_tbx;70}S4& zgA!0$LIc$Lx*aC9-w+VE-3ug7B^rTr0;6@89isEr>tuuL&TV%$Bvo1rw1Rr&1ils* zp+^oN%-RKbXvs|O2b0$}Yy`1((2);j)YSt(djhpIdgki-oPe8$ zZ_a}GII=Q(1J#s#umcSi(~(-~ygaNo)}LlO&tvd&6!96#nA)sa_jIUOhfPp#Jd4={ z`o1{YY)CfrGf}UI7P;NO*K&YN@}ys82Jbwq z+zLTo69iOGCEu$cXLvuIXngKCJAzKX<1)4tx^3ZeZQHOs0{!i0CdhG>G0gYyHIRq1 zgYgn{K;gsZ8s_kM*FYW2%PxH9dRKj*r~nrEdiMo+E+*O&m&Cbnhx_jX0)n=kdrfAnu$kbm$i z=VSo&Wo;`{m3woC>sN625UuXWaD5Il4fB%TJ_7yh%(8Ude?!J0qe82puDZ_+%JmoH zl7H_%m+Jc44fn5#?^7%!E#_RMSTXR_?RXw3Lq#Te zZHKwybBVtg$aIrm`KI}8{z1D}%_42_bkwDPVFzDxSmX~JFGL_4VjNWvpN_(aNmw!- z^V0!3z5H?OES6-qoe52HKC)kW`3ntyy_nJ;T37?8wFVry75}40CV4B6yZ~@Ve>hSM z+}Q%Ww)K@D@X6k*Yt+Zqpsn4mmjwQF8G>wao!0DAWym1-=mLUvYVh&mg>%Ak3=a)8 z$SZ7u5v>7Kf*mICrt1+L?C*!92V_p%{MmT!#e^?XZbi-0BDidi*X7>tIvo<)nc;x( z{=lYHekB6Ysyt3`O#gngI!8o&y`*Eq zFIpf2$YC{$K4H2wC|TOxl+wLVCAsVX#vPpgsZLU)lYM;)8pCd%kb2}`7{^D?OKi|D z`VI&Ki+Apr4b&qq%uIVbqX5Yh@M}c$P#Jxj`XPWw0!IXk9S8||V@=B)a8J-o8a381 zAi0xCFsWgEQ;)nSmVwmDoFt;IV+YS_B<6d+YFPl)kN;nxPI=H$&|-dmG9_Ps5&HJ7 zZDlHkO>Tc5WSL3t$imEBFx=s2uN0(^Ib6F^fy_{BYg2BnI`_|%39R_YB}lo@JP0OUsT%?@qv8X!nsRo^v&;#G{Lt-&=W)tFB$ zuVbuP&z=0bW!q7A9dzbe(YR=dSktz+!Ef2*1Oaq@22Fe9cljNLIgVkSW@jG2m&_D6 z557U!!7drNaLIGn5;?DBYEmnO&Ky%o{0Glg@&1(evtb1{q`G}H4uznSg2fUV&W9gC zTt`c^1k)bJZ#*`kg*6b?z;V|AndET*i<~|}W8liFohuYLxdxEAtE&qv@jd211~tip z%;vx;%ufP;8b~MTX&J`B^gTTQu~h_VC%B{A;YDIft@QTv8lZj%0T$ybsG|YN$trNo z(JA+j9WWp{rCyx6VX~k^h&|RW30|9lqPxM7N%4=j>Grz}GRZp)+)1!bPXkI}T9*yP z*=ZThjg86(9Sm;W@i~~CUpO=q*E|bq5aR&&e$!7NJfZHFF0TpDsTmAv=WbncGClqK zs**Nc<+aUYHUjnYtsRy{Zu4>2w|CG_=C6U%ZaCXGi^CdcMleliC%XpN-e=%k)%W^jUr|_k)<6$5Yx~xr z#p+saC?;#E3zFG!Sk9$bLJI2J6oIqTXgAvf&g_y*a^<`yg0K&eJkm8PvGK0}sJ-sh zU@Z+c`MS?s_I7l*Qbkb8l49dl(Vr{3ea147*MjOl*7A@W+-WVv5W{`$MbN(<)i>y{ z7*A!gEI|MD*~fm_7-(*wU*f~(BsOGFYO}ANfyVgayjf3Sre{AuS8Gvfn4gPE4Vu>0 zVyK@K>hi!wkjOv1acJz-i*~>&|BS6b@-l!b2C-b;EK48s$R~Riq*ka)d8sPJ0whR! zu0vnF-^-|$o(53ku@gZ|3Wt$T7Q@Ao&7J!Khr#%H`oaNjjZj!qH>gB_Om=}cAuK_S)_5Yyez~gwvO)002M$Nkl0l+P**X%19>oTXC3%Tb=h}()sS(11RhD8q|)QIa%}&g>9(_Q z|1?zP(Od(@U+qBhn))Uk=7d48yA?1})zDR-fn%gmgN+y($RMV55?uB!ve`e_I}Y5=@hp|unNf!P$!VANx0 zLwocX^izZU7ZIS3!4hdB!;MUGg5p&Lo)R?EzXt)TGhn3#l^rHI2Z27AYchBL(I7g@ z%(Axv`1Nsntp@{u@?LbJNnVDg9KHM29NlmKe%S4MEg6V%b$uPINvHEax3DDFu1q#C z%=gM8xW9;wxc1s6;jFQP4W{*iccQZ!G$Y5Nn7Mkt1<)-BnaxtFr6Ni3K}x14xV43Q zPSBIiAxiCV$zYm#y44jh>G|wfR=Ymee%JY~@g7(vemV_+4Qzj4sQWfP>jcLgki2sz zz`jrP6dyovlJ8frt!hzs<`^&L)YC0eDd4S}Qbh6J>>(t?%lTrig z^N(fy&V5NZuc|5bTCR8$~So@e%e>q*HCgRfC(`8 z_rcRnif`HE)t!eXc^z!J>gxN@EPocjt$zd{CFbc+u@Ul(8&AQV-m6 zlN|$mf_@6NUsfVW}mdbbOpnSOIgsur@D!?#3!u^%dk51Cu-sxpv$0>cV$2fxc21&v*yIN_0^Hiy!hC`t)vmnJ-&TAU(}&Xp5u2V;E7T*W z^+IZ8q8qan)@rizLp%itm|2c($ttRIcmzQop|%aeUWKvQ_bZWY<Q%J0sj~ZuYIQ`;KE7g5<|ht6`pD4K%3%%8D$-R=ma`y#vXWadZ#^z0#m2c`DVl zClFx*cdD-o^U$&x!Qzl1O_>4pLo`6`q1+u=z26 zYGo53rNf~e7*v8cS~uQW_gfWx&(+Haa&`Cq10zAgdfwXEmX)=2qtUGyvU_Hv5*RRA zjf8eF$)Ov{dmA)IiUpD??m#!V-!RD)%g+hk*$5(kdi$>2!PG4aRtNLYzpqR}iyY6+ zQ5fZotkZ`1QQk;P1MTM=EIXe8YWQ2{Q@dwjKY5>ow2Ue6$t_GvhC}b!>(?-2M0M7i zTmT(dWd7Pt&^F|jIBSLNUnj;OjhuaA-f;}Hbh{8dp6AkS(EuE771is`!Ev)67JD%N z@OSeXSXhGABidmy_xE}pn&;5@)*>90w!P=<1b#cTq)z8Kg(;t=8YqB4o=4C+>r$4yNy--86mbNd9{(LX8|0E$rpz!~h8FzBsEogrow{e-ds$sJ&a z_!rat^c zwm$qwF1TkidgQ2rGNRaKs=YEGt7cCE-^cD!W1nk31!NCf_gl%u*{g3uI7)w$3A8uU zBxit4O0C2r6&Xvd%egrE0o%|&&)``_9!V5C37<-FQ-dQ}smK1MM6Z7+eZTSRGWN=s zy~|-h&*vw)J&=4W>*$jI!~sC^bS5tg^N(e+vL*TZDAORE_=DTtxel~a$VpuV;f$bR zi}T*XB1t$zfWpt!Fu{Z5>4wsj5P$*cG#!9h8IG0c($$!l=lIqj*R z`i!pOIV{i5qwZWOAJ%O@kO=tk>J|LJO{oF5I<%R4*N7o?!X)SOQmr>0cj&V0k%9g? z?U6Lx9o9ft1Ba;r%8F>Vvym)uqx+B7-v_q|eMKu(VVL9qI%vEu%Cq$JAb^qbBo@J- zu^_lTzg?R8;TP0318N5x4R+r_med8;)3+SS75{wA^0a`S-A6!suX|?G>SY2S^2rHm zll%H}bC4y;11LJEIhy27`8hm^;~ctdl*!??pap|-2KV(mUk_n5%sJMVue zx9;5Q3?RRJSw_hsKcf3+Ho-A!KgvWj1K1O}B~T$Sm)~2)wRiz(w{x>zbfwJ2N1;Vv^HAM>e_baOL3GU_M%t zzg>RK_VNAUwG$YA@aVC8^4V?q@w*?$pS}5({PFkxRNnjWW8;KrcI;`~Mh6J%lruuO zJZ@v)SL^7Yld7`EGjZJI3G+Xq8c1(TWXES7XQD~jiFL{4XL{ss|9e;ESN{H_^p0JS zTzSMhVYCDg5r1di*#A*XatvkhzeW%`t0A(<s63FthEc?xkX)W5Cv!%%qm)g^wv>Wbz@y(j~+?? zK0)Oa9ONkk)GVwlAGjDX^gZHAL%4=GL$$$iiBk5auM1$R|4eV;JM-wRA4zF$#ycOC z7V2;sZDsl;TONcix!a!uNK^&5mPS1t#5hI(&hD85IC(#tu9!6ZN12vHu$$Al8iIC~ zz#>n}LsR{Q0+Es}we* zxb_(W`_-jz7j4Q?Ud9p~6t2yQk^sO*vg=+7g5)xQ85G^uCH2yOAj#+czC;JENEBWB z2%RFZIGr2-JIhUGahe1t;2(nI6^!qK7QwR_MKI4!uC1P5lJeAVOC7Nh&ZOZvh$7CG zwOTzU+ogV)fg@Skw3u^NO`_@&_5_l_r?kVaw9)_}$1zyVT*num-{!IB*S6IzUeyq2xp} z@P;gE3ve^YB7-DqP_tP9QebiGGz1lF9?jB|6>C2*+)qZJ2Jo95Lm-=JEJM_$xo{rL zF9k?)J3(@?lgsev(roZfY6Mfdfh=hn*`)+%S$++&oAT?gr%CPnLqJ+fn-J92o@ABu zK&y_vx$g`h3!Z>IC%LvVqx-FxU1Tu_8~0Yn2De|ncoCrFtOqVLKq?&z1poHBN5Q=K z3=p`s4hK3@bUiaY^EJS^J4?Xk!R-X?TbYpxQu`eyIrQDY1pDCQPYtN3S-v_S)(t`U z0R*q&GcY-E*=XA9G>2K|^^ldGSOvOPO=JkJe`<`rKHb*<7Lu~>x zKsbBhEEt~yexuTGit#&g+{X5o_m4&=4d3PVPtkFgD9ryTY9J2Bua-{9V zSR*T~h8c$Sf)6U%ePsj&D=sgXpGBmpKz{BhtQX&taejv|OKAl`(a&A9Jnh!Gv#rY_ z;AS7%x;k{l4{9+ua#IEW+ZpLb{GOpPNPYB4Yy`nOhaP(Xat+ch(Fap9pn%6HK=BH| z^tH`#DIp>%gLqc+4gi$FH5Cy6y&oil{Te=@7=t){aUR(JltpltAv?6 zoFf3Q3t$qDbW@Lf(E-WVRH8Pku>wdm50+$m>YfY|*m8~9m{>dif9?o68vHz1k4M*U6EIYeFau$sMd3tqO zMlN4zT~ag}lzOU{pvRwGlvwxq)@e`a9s_zqXWv>KAbGjszt43*ym#;^+$;z-bN!is=VP6dP~X?u_KpJZu1r*rLK`haCXc3K3!glMW7 zk`qknglVS_A~?yv&`y613;-n83}u3eTGyL3KrnzFHVYyXu+`vqt6Fj-Lfl3SiTE70 z1<9fDsv9W+NOqc2t8+L^a{nAQ-Gy_IS_y*Wg8<1DliURB#6V#O1Aw3&Ij7F2z^u_g zH|s%{&z_6R0yG%4bHT`m=&z(se1LTmyi?`J>Ham&>uS?l5){;f4dd8QUhn-BgU*!m z=f;f&FPW1xva-MU;tO(ZauV9kop<61B5H?%0q+BBZ9ZQF#Mg~e!=K93gO6bF1EsHn zU^AZye%uID@}4me=ORFT&K%Iwc$x0%#Dt8GjUZlCKfrM~M_S0Z=yM!q-RhDPRHKni z&idFdNN$+qE?e2Kn~|SC+;3gNPj1}-19@AXxpGCWV0Hi{h)#mB-M9lL-|c(%4Pb4{ z%$3? z`Ki6mZOgR+_WY-J?#eP4janAU?hh{6w(*+JigBg@<-8bcAeTG-!3Cci4fC0uiF88yCJX8(;t^C@Gp*|qkPoR<=$GF7g?B*kxN%jM@1ay zDo`VuNzVR?>a*-8%V^&ufZ7)UkiRL#$6$_)1H3La2fbS5OKtI5ZnKd|4hL}=blM#R zZp>r>M0P8gE_{+}lv^q82EgNpCy{hPw;I;#0L>3;8(?QIz~S2qt=bF*_$AEF>jWg< z+1Qe$$1~D}c89@44{A}!1|$>+mh2mp-t&_HGrFNkK5jsAn|Bw${PSQX57~DEz!3Gw zOBf5R&YYK24FE4XAXOK$FQFYSpwER4qRpb?B}cFdBR9u&(rq<98TTycE}-w*cuMM# zPp>$$AN!0R`Kr`#t1BD+_#yGaa}xixUzLTqN3t<>N4jTcB`<*)R7u7NE_qZIo>h-}cG+vJ_^KunPGFJYw0ooK!Eb03^TM=IG|T z_${2hKaJk-wyRp_5B(Ec$Q)zVxvNt)2UjaxiuQto-RBtn^dpMK;*P83B*W>7W zYG(oDnIKo{8Q!^i5pt=})j!D~IdP6)%r84UW!u!iBuAsOTaz3$hAN98#)GyBSozuY zDW~Ir9R$g9{m!^Kf?(j?m#$;$Th1Y9yCP}(;_6&Z-g+H<<<0;5pG$mvTn2vmm-c-D zhDrY5v3Fbq|tPh1To034G=%NLz&yjd_n zC^uq(aDpur1q?CmDgyYVcba!XbOjJNwwmG0dl`|PGwJ3XdjkL|nF5}3b!ELRfY{6d zG}Movo6!ATt~O3pWLY#wzX8y6X=%j-r6f>kjSb-1+9nt$)G+pY+SHI{`I`Y-O8~Jf z@X34#V6S32ES+6PVFYFPfws1Rfzl$%PDey5BT%M2oOyspE2z(d%7=7gYBE{e79{6$ z@#FVCfL{4SnE=y#3<2{x0ecBF7652@h9+wr*Hf|qq1+vuRiNc=xVa8D{W@4wJKJsQ z6(21Cm2;5exq0mwBdNn}MRZ_Rn=r_1`lDJ_pkJ%m{AOa9}Q(b z!_F)S2WaOypW|$%`Fo|yzPY(=o_9)?5Okirbje7{baLNxK2LI;QW}pj)tO}u0^$Un zmzUQi0#=J{ubP!@Wo51LnL{G#jL#zoZ2``?b~cBp^_*=&qZ4syX|?4sd?(oUSc{>p z+}TbWpjUk zLO;rJCvcWeNj6siQpf$iRr?$o_r?HZSBrf{Z?w8omE6v%Y|K7HehQdrJfz`Q9L)3R z_F5xI4wiZq5<2OvBKl_~837T(nvq+9{780lUh)eIcm({RXlZp_2Hp2Pg6~lQ?eXH` ziaqo|%ayx#W&PGIc?L%pFu>n<^hgFTPwbf&beqnFz@WA53?fT zu^uUJbju39SE&RD1H*U+;&{Hnu$+x)9Kro%0Pc72dDDlqS_Xyr5xVbY{uWuaql+z-iwR&0!1DzSLM-sNxGs4 zlG?Q(-2nY_tGqtvi7h6exjfV_W0#(hi~RfA&q`@yS~B1NJ?VL4+MF?l;PsVTb29t! zdjNeVy;2M$sxs6Yku6B25MY~K&dFbXyeL=412j|bh{lx-^vl*~Kanetu_!M%ZBHHD zh|ELO{CIuItA!en_iYft*nTsQNUxGx zK|ZxK9A}jEY9aaIW%QJ<)G>xB-@Qjxl4SS$%|yLKXWxh1n%jrqwjP7R7BUTMD4d- zvBew3v8$M7tOFo&9|L*vkD%7G*PoFKV4SxDT{ppG*0UFSz~mjl^F}tewloH;1V$|= z&wb;#ue`VIu7OeXtzD7WfC$SbBl|R+i)CAMa-_~>FqYp zN~dKYFsvONRj_7yA>+dP)6U2~=Xb@<2(kB#vo4;yaUG2MQSY4(=0mxWHPq?y+8S6* zeR5^;vh?8H?!;!nkmJ4OY=r$9S=>_q4;U!6mA5_u#zsHhw|1Z?L3ZshVmVAbVvaW3 zI$fW-jGTQyz}W_*<9au~jVu#--(_}nBT zgnL7))*yY-jq4sdzU!V_T9mRED7tQ(ZLRL*NvZp@Ue3%WoL`i2VlXIJ9dZX<-AU*+ zBM@u_vnxlP_G|PU7@qw}DelA&;L~~LBKeH;*2*p4=d)0^wOWtV(B-frkQBb_JtG*j zCVCHa`Ppq~UFRUhhCWX<1D)h(S-Sh+IM-l8o17^JkS~D+tT6gW9J3&k0E2;FV6K|< zvwW@*Bxl-G%so$+^F7@GVQ2?|D|_5|^a=FEQn<>)zV9FlOA(snv?u^Vg0`LHnV*}H z^~^T3qw%Q1AzKD$f8o*v_bNJDZz0KX~-Bqiedf7)B z9SA1%es>pYkg_8vxXLoP;PA~gV3k!ur; zWk#PW51sL~&;E%--Fm34&Pn8xAIa#y`R{w@uf{6!{A5~we8=h6WdSa(WK#0l3r?Hi zC2X?7p+_;lCCRNBDNdtzR(AUBHAvS~;27w_bd`hDV{~T?L;%v#9jl_RlSMt2YIF>N zXpsfT8R^4VZd}R(7aBnF2)?B$jFVcOH-WZt{iEB`_wjvM#eII^7yfm7ngcI$V0UlN z&qyE1KRHM_BB0?xt+26*ah>w&sEo)mPoV*5b!|)v&aV`=;9vteNfuUQ^h@pg)++nc zv5Lm9glA{f9dDBLT!Wvc2zjwO-mO9xq*tKBKd|0w7|{_n`go2J4_5{u!xY(LEb=s0 zQUuX^BT=>^&bOUo)v+;*<>45o)Jk@Z@3_+#u&0nrHafy$n5{Gle59|B&!5wGXM2k>Zyb}w zASJWkYg4D$wTj23t%V>ts)r!tevJ%E8%)X`OHh(JxS@qL(5eQ80Vd4OF9JArJ_){3Tj}H_e}@iwCVP<{MzeZHh}`~PCfAS$?fqg08sgN z>HK+l;klb~etfJcSWbqxmOUYGOn@_3uLJ?Lb~b@QO0ev84Rlb%sn6{bdTkNqMS$a9 zcqPt&==cZ~~@WDTyrd$S*F) zm1~pUWpuKV$%ECm{8;Q4q>6jDif`zL8x@IeVD?j@NB;U3QUJ*d0Lgcc*r`Vz8R?P2 z#u)k{39$1q!_f3~o%^4tmLXfz+4Tg=i!UgoYklB}{dj&S+tp14-`xb>O>O+H+G8?> ziY#n6W$Q!G&dDAG&|4=+?q6GKctm<$e91e;=2ZY;Cjb01()-pUvnh)9Aa-wAVwnX< z0$nxR#)sqb!qtigl5c|%es8WMzxY+iHU`RN^R>%OCi5Z|b zX@e=VN*OhAr4y{v&Z0U zizsHdX`A?dK4l&IEVbc1o@Ey!4u+`i9{w z=QqMuJZc@WLiU$`;~9ySkZ|Q0zB?pI2J>XUKS+K=E3qq`m(MY*_)pd`WFBT2)<9ST zjWs|pQtN63gLfLK5B1fcj&U;px-nPhmjMKcV7XQVSY^QlyUi{+P050~1QDI!myI1? zo7OJLG}Q7H=fH*~7}ctLWU+MNzM+0X=AnHS4UAf{E>sfVip zEnasJ{IX>}$dX{2X1ns9=<_wplc13egv!P~3dhG`jo>nM$gf|$BKIFYmZ^u2z}Ubj zRv8(y1XOt+h5(KRO_z)N=tziQKoc~p-)*;(i5%3KxBK+659dv&K^FJ%`oHwj7v#n> z*V;0Ubw04OMF(F83B6Mn>F4q2jB*1hG!~h zlz-BUb#|t#qo)tdXvR74bjdNphWCx_oHd^7?t_O$Ud3*+%ZqS6eeny=8^(XLwuSG9 zaT=~TKs*6wg8aPaodk8Y=lMLBbV%@C8c8NR=bab>*HRkG8lO+=D6rd(a!DP0J_FYH zeDN6%?PS+LZF@s%TMk2vS5m|`Bmzfw3f+Oq2AnXE=SW@zqa1^?C}wcvOcV&GeTw5f zj-PqC9Y9O?7Fjd$QVLM8+dExW@_@Oa%i3E8gAYt{4z%mV4g*zI;mj-nAV=F*$MM$c z^(6XD-3Z=?a#rEw=pO=S)G-sd6Tn&>Ade2-Sw}mnHOONam&f7QW?!&XKhvJC)p7;> zrpIr+Ee;G1_Bx$<9ez`jNZaf7xyqPjR9{(cb()sjtC;yv0o&GE1%1O}65lg4WBkU| zj7{P%Q32E=ulIFJer4QjufX|Gh$8NW%baAu$Rxgn2Qk7Ps9w)__dpF2;BRi#=^JfRn|!t? zU%aj;kGpcmj5f?FT%Ua;S2r@y%5Llj$KgatE_L|K2l9?X>%KS3vN-h;=^d2Bl@~?s zI|(yz6g^ld$WOk|h4$=RgIrD&vVidxvX})tb1ML@y4;x+UE6qrzWqJ<5PgkT|6QM5 zA1Kb=KRLbvDI%WZ#w%b<7je(|u9UzcuK_gXqg4h-oWN`gS{-Du z(l%xSOnQf&$Q5@|y(v2`u>l{^fjfEqY+qboxk+z$E!d=IOFJO>3Z%oBn`30058Zgq z{xr^qDgyMzGful%LtSR)9|0u)DKE`QTdCt+D@{+EHnWEBXcdm97-raaWSE&(nSWC% zD}J2;ypgV#zk=ruy5NQ+gqeRZ%B|Ek;juZS1puxZ;~mkIo~V4-($BQ z=d$rDahzug^A2kutO3X%3;@veUwZ0jTj>>6580p^94711i}t)n{W)N<-^dXZ(s|j# z$;Kl%V)qd^(#&a_F1R${F+gvE>a=D7JM}TcLy(bC;MO*Rc3K8K(-B0spca3p02tb9 ztpGIC3_|X=W&(Gr0iqLV)4y(;?x4#quxB+eMo^|5NKRmkT525d&=LuN_F!Tkm60j` zDuOco!q0xou(Nr++Q?|tV4>c}XF!*ope(1<28|97?z00?o>SkOV0!!8N#_HYeZ9T% zjjz0BBt`fBY?{nanYCCn7jGc*W@ zmfH--jdOrPr%jBy{d77sGwO!(-URD{pgP&<&5UhjMa5M};=QxMf*EvHOF;4ZZ16{$ z7o8NWCvC%MlfLo#Yx2|I_?iixtJ?vcj_Y(vQ4Y)6Neke~c8%==?c&;+E z$}`Vi^*8U}LAf&z(CeYMZHNIpCwzzZ7C`d5vlf~4k~OSQznoGl)#V3Lnrql)eng?Q zo>*(%ciElV-%`Y^gv|x8-L0lH92-NIM&*T9zluJeQ<()kLvyQT+021J({~Z*z3ZP0YhAgbbp7B>`RKp-$Izp1f6WaCsvl8|S-Q3`Q-lsEXcs?z z>o3&a&Q}SNZ?7)<2Y^K$QNxo6#zpab^J}vQMezJasc{bnTX_YJ`_)Hw*P)gLv?HJW z!SBeY|Jgq}R6fUja2;cb%BHVHM2VF$K=K0WrH&V^0?#q!d$_%%2uTFVD~L^T(#?@B z_OHUWKWv+H;7-0V1|)Zr+O~_^6{LbDv;xQKGJ;Dw?--p95irL`uifxZ0-{)L!0)@` zL{dWV+!JkRM@opHK_g&ZhSX@MR>zDf1Cp!RK4g-oZa(XPG!*2T^NUhTkixiHMk$l zF9<}FrQB(K2J>q7n{4mbU-_c^{J-`ykg35GDEvnw&=~}BZ9XM*hV{qDDVtx%HXWD! z0Js@gmPUrS1su(W7XQ!*LM?oL;6_oVpZMBWE(oP6ILzZ#RSrELIX7#@Trn)YN z&7YubD@H!s0sY+bKAV{bR=nzd=r2@gC?gHd!EuZ-~8(9^6ZUgc1`iK z`RFABF86f%ta#S{ak8rQxsPI<4j{H&Wd1WCX0yP|!Sk>lgMfHCoo$fG2_AEp-*l?z z`$k73oizI0I!ylIv7b*3BoH_{seqv(#8e6qXCP*n+7wk>=g z=O0h&O}o}NA-NeJ7x`hWywOH+FDI|AP_JN2@WtK zhZQ=P=*#m0h;tsvhi$(!KtTbz2BX#7@$d=k^t3SU1|2Y_e#U_ z_d?5Hq8obTP$biOG)SI- zlV@t7Bm~P_*-jRLrkB>_N_DzmCWC4JU=bS3T|*KBc_Wf?+A!+fr@-vaLzX0HHT0*-kkAWSstAz#t@q@! z|ME8xVE1uwciZ1JJdYLhxB6Ti@d{$MWIp(y?J=!(R?&aRL*LvQC$q$Q{I)rQHarIT zotB0+td@rp%bj9Uf#WoN?{>@N``J;&4AR24zbgX>;+S4u+Rvd+HZ8#Jss_99ewBd{ z&<{+2mw>rh0w^y*YRNMY$!Z{;opfUxWR2|){~QAt3&|DFuw8@X+>XZ!2k_J99fXaL zYE4gL=^V+r%lG?Z0kaij{2Jnnc?N*FzR}tD&O7EQTQN-XnHi^41k5`H$pMg8m;O?U zGjJr^<`d6a>c(}<3PHW$hQx+O{Hw(cs{~{@{hgh!X%ZLjo*gGP; zM@yb$3Bq%QHE_miz|r>hUp-1dP^*_#1nWct+g+G~+$rE{z<2)(18||-6lA8ZvSw`d zLr1@r7cY2gs2|Sq>?O5gj~fcG$?bM7DOnJ6L0SGbz%=r`Jm;~B1SV!kTC(QhVmX8cpvh*z=NZ_ayo1B28c8E}CxfRQu zpr~d3lM!y$_bMkHKUXF$$xnX$D@Ge#*V|sl9{2vBj10BHHHf>{`5my^_8gqlK>)m2 z9)?k^*yeODT|9pdb@WyFnVWpU(D zeC80_7QUm_wp_^-``a$p5j0en!)`bE;Tdjbl+e-6D}v%ELlBVCV7PnJI4 ziQ}*WWHtC5)5iPW%+aFLgzHat4a5L)^UqqXHVQJ=Ir~$}uH;HL7^7X?y_j|mu*r^d zF-9t903LyLd}fXPA1l#ikOhLze&1<7^-^Nr1~VFaTtK2$z=U-$f$QZR{51y6YL>47 z$;l)K^U+R)XFS9#G=`~1PRnO2vbog=l9P2_1K3wDZ5wHqO?-=19?e2pqVX(cv@eBC zpKac_xBahGwDZ#+ek$F8JacHPH!zc^o#g=@Zq56^a}Q=MjqJgSZI)jQOzWfzNOFG> z5;Wf?LhS=y7 z6tM?kX+Uy?QjvI}9rP-xx5~$@lQD^5lJ~4SOmZb{(la(L=U;xs%P+WM`j3cbw0-@u zzY9`8)mzh$5ctGwBUpA}Bq28y2)&JG?EZY~Z&(@(?8OHQGPVjZ$Oa6?Yg2c@YGzzR zfYtKtPyO?DuolT@C9*W_rK}`kl7&1@A9SQ|T=e@sX~@t#SU^k{%$SMxpKp*{DdGFM zwX_=yj{-bTu5NmH>y;#)tDaBq$-;mA-^$#_@9AA_*Na(5xIFOL>2y{OO)mrN-oJzb z0rw*MZl0k&bx>J49ys%-mDMXCAZ;%}p2caqB4`gH^UbffRUXZ@XO|j1+(XR}iGIPFx(-QwDHZfYg#_ zoV^Uv!bvtZJyoy$q4Ujo4$L)SpO#zUo|p#1@!PV{wSPE*Z(}Ose=q8+w~>GpDV()2 zv_qbuuA`JPI>|e&}j`2Tv73jC%VV7YnZC1tax0oOPXH&#BzKthey6Kzk6@$reEVH?ySn67&B2@( zoswqWPBQa#!%fh*Q>Vjz%VpC$Oz)V-sI+YI;0J|(uT=IWg!R@mhFSlq*Qd6 z@OiP0D1yWcm}=|i!ntus6^+aMqQh<(LqIbP=u%tzA^xmUhJ>>j+L^;5_tRw1=_UX^ zf^|0+epF`49m)%gWGtG+RJ{j=>Doq0CS0`DWsWt?uf z?_}At{pD|KbXp7}i0XcfZ5W+ZODkOlV5YN*_koP~U>)u!-zSll-v!of5r8v|;69&q zIxbro-x=0p@L5d5Im372IE*tR7-U)-XEN~p^W77AZ@B)n*FYITH_JQFEVj25vJoku zgY}xvR;o~P1$x6h=Tg!;5SMKLi2Ul4MNZpZUJy`c<^1(8I2+E~zpM?XA7o~f1=v>m zsg6xZkt|}Gw#8@3BuC6n0OXD?c_h)gO)%$y+)6FazZwVJFwe&lI~+hgmHv#1}CF_~Xok?pB_-nH30E1W@upoQ(M z0syL{v{xCAFb%e0b#`7ZPbvVnT_)HU!F9$Ux-HTmo=CjCcVw#=^|@BF@TS zuUwdS0By3+@Sgm|_vNL(^>w0Ww)0g33*qXj;f1)kGv$Kqu!kC>BfmZ)P ztcFx~CKaC&&x-|om z$NLcUn2d6_4aL2$KEz6Tw+VLsh1og=Gxue?;K&yVGV>eplZMb?^Is{he?js9!9371-98}1li## zn_a5}KrIl*odlU}UOF)WG-_yDBZIipK$8Y=#K)T;UpuK4o43&&S8Q=_JERIM>(g#a z0!-ANWCQZ!8I`QLd9ItLu-z zoCOT^c^r`3ztNb zFq{$~?c^TF>eo8Tlrw4;c=9szzWJvCa-N$KF=mE@%M7HJKLhRR=iZjlE`uOBw=b`> zXOjn|RtVNp%R6XRuikDaL)|iv?KUo1Cj?%tQ=tn?Zq7I%U~P?T>mXx!9l!(YteF+) zkibAJXgohL= zOdTREDJ5X5p1pdz^#j@Ft|jYid@J0m*Sp zE&WJJ4}VW055Fs^^%*}CcP)lKaa=JicO_`EAqhfA&-&`F6cISzA^{1OV4`Rdy7>Wc zJ`0(OH9Ms}ctDfi49Zu91Cl{Hjgh$=O-i*4^!xv3?@fa&$*%LT^Y&N!UftE(^fEKu zGd&wI*f9hkK%i+dL@5|n(6U5^91ewT1w#&5v=jhPuX1X$8=FOX#H*em&_dDPH zj&D^kZA&E(EBcQ>f7}BhxQlw^2_Ts<0%pl!=xGDEg;jmI6-u$FHE=(TJ?NzaX9MV< zx!_z_wPen4PxD$Qz~%ENvvTotn>=@}-GSzxe4;~Mc&bxA|6G@R;iAaJfu#+4$)}R}F`L@(%f9TRtm;8xGvWR8mwHrm0EOfQ;> z=;ycsC(ZQOw4BX<5CGGZVQBP$!T(c_OT5=B1O}j7$GdR*`!5C8vfWA?PJxV(c}aBj zN%rh>t{nTil5*}?*#BWr(_0AkogheVx>_}iFN%?RS_A`n`u$76>8l_r#GijwF8ta* zk*UXYYiq}>$ zSq$aY@yKca2yYQjKo03d=#b)jpFrGEVmad!km%nd>eXPyzi?b2X7U0&aMvt5h zLS>TE3lT>kU6qI8kgJYaEOpFY+N<%oGwz(VW-vV56Wtyi4}>WXW-aWM9!&llawW+1 zotJ;G;opiW@HgLnZ^Qo00ICfqINWep0uOizbc3y@j7MH7>pnqiGASMK!=5WF>m)ze zZQYp>+_YdNr=_<7)~G*g->0sQ&Xx@6R%fOD^avu;6HK$`YL%AT$h4%UvHh@dI~j{A z;lSTCY6x4l0o#hG*N|*Sd+qzJNA6t-%HDzRVe{x1{`PCG39(!UKxA34PO8KV5KYp3 zTO~u(-^+_b{(J$rWm=`Lww^lj~ll~;XIgg*TH<~yf}68#I9N7JeKA_a#Ln}4z^6a zR6LF29{>PA07*naR1e-?$07&#q{jAxAb7N#bw8JsREbKh=zg|=Y@7Gf$Rby!yUmw@ zr?+xa*|hh2o$uQR7@SPa`~(BI+Z1{kLCQ+=A#JGwc4@3DE$xG8NkGe&#+V}v-dUL| zNjvn1Q+=3zj)84p)bGvheUKL9^H8pM%uF=w-qkyhJnXCN_LdsZ1*gn}wK`s9FL-D3 zcALhn*U`r;Lzaagt-&U%xF#E9j%n*4OVki!Com9X>tvaNbzEC`SK23Dk>Nsf&`zdJ z;hWKhzTj@HfEg~8sTmn;p^2Qt_an27zBo0=+rjX)RzX>nz)s((dwjdy#`l|Uo)$QuwVuO^S8ZT7Z1P4Zf^COLJ__u@$KKytqx z;c5s{C3b?62Y1wZg)%+$++8?Nz%XA3$1o|(Z0maTvDODgwziKPm!m)Zg?Ww)*f*ZaMkrOs8|bH3 z=|wVDMP19?_5ea|ZeTnpaOU?taz>u~2merJpMS!n#(sQz>ep^GZEN)-j&o(8(`_vU zpg)ZOo@4L*$ZppuDCh6TOapsI@BE_2nx* z!`JTE!FnF==FLC)1L<4v7_~fvGeD{!McjV%l^x`807(GSqthu>jPsYj^U{VBP`5p9 z+NcKfRVwv97pSwUj$>3WO0QtfLU=sxs{t9OMm4TW6mBZi6d>&@C z6<`xhjxSd?fS+vVT!yo~o}YonI2q-BnW6np4Z~qOFyI#s(3?yzX1;`F%7U+9th0Ax zAI5)pekQGE>2p_9lRU~m-}^P}?8y5rx55GMi+}m@mRI-M8yg$Q7e4i|O~IWlGdtXQ zSOO1N33MZ9i)|1X%#NDmo*sGV&D@RD3UAR#X|R{;+h$C14&)UKAy7@gbu-{p^U~$9 zqr(8cyPbrEZrg52pqy-Zf*Wc*V9Q3OK?n-kk0b#S9T@0w4mc0Kssz?~jDG1f)x9s< z=%5I`#^Qjw%{-U{m=v(%wNSn$dT;lBMa?-dAgY~WM&MSj!KSoKasq<3-r00qXMH1s zz6~s}DBu`P`!N0PLS3^gVKT`HvO-{H%iCzBNA4ZqNmUT2YgJBB&2oawDr2+d3C+$P z1Zid6R%r{qKdk$7-b9T6I)T}!Ly4UUvdIUaw@tv<8o^w$oY`dfTcCuvIkonJdi|tzx1Hxr6 zz)GbdcYiy{ar8r=D_mRk>ZQjfN1F%6XaW)g z#aITi8D9NMpdKb70OpTV8Ettl99-Hd$LMp!?tw|31i+n&1Pye-JC`W8vNZ$KNwSn7 zwF0>f3zFA>T2$!N2)lfc4*-Lj+VZ|NpdVq6sTVO&Ul04Qnl1zKNIQY1UE0FxM?MwK%0;cEQ-LuCGX8Zc>ytUIWNehGuoZEJx;5NKBL< zqY%#mBnJ?|;GG0{$RtmJSx#Z8vXFi4fnngCLcIhtE8t89LF)16yCWe zvBv)@)m1R7m%HUl6C$64yOQmC&TYuh-g@vxKlruLw$qE^N%iM3Z>m6K0DAKkU-?>NA@H2QLCHMB#; z3?vTInEm0;@~c-$h%v!}|0zutEMvIlIHb3ljNkmR@V=Y&ZSkXy=T(7oqw5E6ZAe9j zL;k&YMLzMuhJ6RUApu81I|3o65F9yqXHv%Bd{s_-8| zKgcl>5F~f)lV4aUI?u6(?X-BjD%~BB54q+2X1Ou=WbesG{EW>4=W1YB6l1fB>!=}c z^}^U?890e)Fb|ozpq-Jx%}#P)2G6XHahq%qM>B_Qv)T-H)A`}m_2^;I?X|NyD7@tXD8sG+g2X_KOEKh=slq*s5f_H|45#7!uL1)&_RyMOq08i?b)3KuF zvB;Z&blpcMnvFj|FqL&aYEg!{ci-rs;$63rD+I+^wzNeI%vhFc1JDI$@JxHYDEQ8_ z1l(;Ir@&N?N*s|DZ`t_E=Df_KqrJlgDZN)%MU)AoMjq)hAsf7|6Eu; zQDeO>wg9b}Zo6ZndF|vk+xk|sCwSe~8MBkT?pGQ*3+RX;z`s*tTHfnViX{<@H3a}j zEgu6@wyj4xPh=%M6t;PlB{0dsCU1w_O1!g2)*Jp{83!iIRY_q``x9yG6RFgSD**IU z?g81%;C!E0LEueb$rkV?SXPM>#0{;EpmP=^uR&U&T=KyT zFv@Gh2>pPHZY=JP$5XsrXrQhkcF(SCWl3U_{pqXGqAZJsikwD2ue01deO>e4ZfS3kMSic{X~!a813~R}R4b#Ox7w*L5j;kKcf-9&Ie}R+YHN^nxWRJi>(fmVM?bU8FtWKL1; zSsQUwo~)^NwOCTt0N9q(;rO9}{3=UxEzc=nu*V?9WGx1#Q4;O`ZmlpgEw}#g_gfyu z0H_(ro!oD$V{Bi>ES*k2aBmgQXW?7l3NHigc)hraf8i#)flyHCcADg@r=C%61#K!( zT{S#!$0SEghlkbJ_Tc}(f>|t>B|;glWV|jz{a)Q+Nl;cr)9^SY(J7%V5vxjXGHmlX zv!jHVCbV?~qy7)SaYKIh*Kf%0{pL-1`D-&$hXY=%PT9b*>-zKIT0)7S9F+mXH=(+% zE;&JRFv(}arw{|ilRkGgynnRUWRhF|44+yR@t5w!qIgBJMc_{6PYlcwA){old%1?4 z92>S|J4w9{X~-)3SO4-K??X<9Ssa$Y!&?FbW*9(D*~tXKsP}C_ZI4;)_v&|Wt5Hqz z3OXmzz?TH$E!a!erhZ^{aEd)8ae+Y%E1L}-0kkO4PBthTXf*;P`Vnl}W*iBWq}dI~ zula1Z%x1h`kD6{XxfOy#mch+{eFx1V=QRg)#RVb-6N4XS)+;iv`P{a8R%9@@(s^eY5D>N?tt}VFB*%Hw zb}P5S<0#3*V4rGasBR{8u>?)o%|JP_nh8WRfUX6e36fiI+S4O%%^+r-aX>kkYh--$ zzCA|(m@(+Mt9gv?@V@zuvrXzuwSxs7G6)G)6KLlEb`evI?R7BFCuf-~EQY{>2a-1f zVkH1%mPyX@M4eMSzmu#&d&7=Zv{QDH!8%bc_iD_8fIPM1X>Iw7sm(MAn;3CHesx4S3njG-(JpCQYYAa4$AEjo@`>F+mX zw8I~qZ4hO<^H8>Hn>naG|p7uwG+sv)Yx`5HaRJ`M(?_3dhX04yPBo7 zUEUAk{$!~RkUMmKmug_mB?cwg3s4oHQUaiO4UF;z02IotB%tXVZy%6+mFVH5yQKtx z#e~V@LA=icyCwqUR?m|F`5u|%HN3A$yz^ZsPith7*Vgb1Ae+H|n+ETM9yxW(c}W%| zuL=Mi3BO;=yH*GL*~kT~lv0xG`kWm3)X&SpbEerYr(q=B7W=&`T-x;@SUhm zPDxkyW8VH9{O!a$IEe42S`mFv{u}Vw;VTPEQUZ8-`oiOyBv|Gk&oDPO?)clmMy#{} zjA+{ht38f3!OEbN0B)}vaGQBE*k#GdrwoH#oWy%j4)n;y>5&hiU2X=ztF^Qg>zI9! zh13mXND_E_bvP1MP)#x%!mTOuu_$Oz&8wmX{&81NUk9M#iEyQ4FFK}%KIXp{bGx=gqz}U)@4>OlhkKe zR@0aT*MBNZfHe)^Iqn0$YXI$UFT~}Mq2o|^@PH;>cgjd^6xLQ`>e>~_pl;C0XctS} z?Q){4`PtUtcqk(fs;tktbd&=s+`cKt44CeA_|m#Bx2H z{p#Lb&bF+C_4>eXXdh3&njOU_)PL6SL2|xUoRu)pGb8JjV*ttHO+-5PdLWI$266QB zd+Xk;6JBZqJIVt+aQq+^z-9EGuX^7Do@f5nx{ScV1h=KTE62v39(U^x)gSi35rn#C z#!T+|@qU&krna0f&Hg)?zZ%gD zTKp!MMIcP|zl-Rgvk|mLrdu>MY=_r=Zipf_0R&u*Ye-gIE#-v>-2mM-F<`Sbp}JSvR%=$xod) z?!X$$yw?04=4DreP0kq=Q6_El{+z6cDNnoK3*Om0aMk@DquP58>JZv3GoOGjWjs{N zn;iqnGEr*68i9U-&XoD^EC%_IfhNg}nYjf3cOfH*pt*~K-~%nJM`UnYX%c%5u2bG8 zZ2-S0Yn*BFU{zU7aykdJ&P64fR#w)WECmCOmVK!e);V5BKL+_0Xk}U3e0eT9g{*Tb zLnb+#AW`GY2s+=Q)(>selgCHgteEJ&-TwA;$-xBl2g<9cTW-bp(v*4YM?aVfiELIrn%-}#DmZ9@w4Y4i?rk>r)_3>D0 zoU7cB&5PGv#lHv5)&R)m@a$>)UQkngC&sd+Q|PVtF3I;^eZx7=cXHglreREi$?2I5 z@7BHMyWR0IG+}XVt~dgz4MRTR1b|rdU-|%68P;WWsRGtHoSCU3sGCCogT35ic>K&~ zzino`C_nt>oZP(B0@T^>lkG7-*QX;Yt#%I}`6A*oOx(JwIR(oG7~U>;XL)PM6(_es zwmCj`79@xJ)JcK_8DgbcNlB>~wiijXj~P(?;vEDFPszeU*jELX?0{lDz6b~88Upg| z)Mfy1-}uqnV1j!`bCbhb&(%=jtbySW+(V5s?*p^Kacn>cE~0VO}bBomRT?a_4l zJ5GrO6O^yi$jF+|*%Mj~0A&e|BT0d<(gC2n3xRBhz@~-NN-Tw-ohbzEj0BaVE_uNL z>tUJY;6&wp@5VShutNmCUb`4_v^~hp8~P=KPwou#v>7LdhNSHy zKV>&4qqPUw=KwZS_fk#TDl5>nkq zg8-Kx%(~!>>uVms$NIN?<64kU9etP*!SIO1;hfK$JRWSVqpwt*mtH~+$IYd|`py1oFA+$#eIhNJC66_E<=Au#N&p#NQj10pb_X`tO` zJgiJ|v^`{!2ZnouFUT{1-sfjiMZ?Xa>_7!x#( zjvyaqteIn#=VtDZaSM!29}}}6*#3a8cv34ZzTp6sD^uaS-tso^4%U|E0C@XlgGC$| zM*7U@=H&P&B)Y<~;tc=llINIfPPp@5N)Y7{44i7QpJ*-?K(97=b)zZw|94p5_CN2# z91lz2ummMayP@skz=}6eBaSaaK7?&_Ge$XUW*rqTj zHjVATX0tnN8SuNp`mOEE)XWTk^4I0k)eizmBAb4{)>$`i+#1Eqgt3;AHv4s+Xg*8R zW6=G?;McEujWhzrwS&o+O-MYASQ!0?Enq-$O0Lvb5I_++zuE3M0>WW5MFW#`ZGKKx z=cm2j`}DU62Ky3%{}sUYd~l70XWw0nYzdU_!SZyNZUPiH9LNbK`4*BZp4B^r--;Hm{#kH%!`RdRKmq$9Jt=*GF zsZ-xQ4T(Wu(20RC4=1_u!&idiqAT?h;>66{0Bha@m?-Vg*4HhkA{xM%R7{Q!q#(Nj z^b^L}u?y=-xw#5vy?uVD<4e9gP@Vm&@eu58OO<4x_I;Ur@mD4P`oEOYJHIKVEB}uq zXTB~+210KkXD+1YHt!q6@a<17g!{9dNjZ8k1j#dS>U53I%H%Du=(n*d^2(VYwrqmP zBwwv}%W5K&oWTuC0`W?sRxPZ{>evU-&F*&$j^p@UZ`M%D1ZE>;R&M^mmnGq7#c<5v zaa17N_4t>L2wt7Q8gw9BzZLwxA1e&VT0=kA8fC)zKDa9jqc;(IwhX6Rc%KBx%WxJK zP=71<9tMW(buh`R4QMc_A#uakB-c$3+fqSpUN_$&5AQ&$R09^0yX|f#0k79^Kfzuy ziUE?FF)WJ7VJ1pU!@O!q<-%`w9OS%?)8`!qQdkeS^SB!N5HXB>0z*BGEr;KrVEc!# zQb&J#;dj3*>#u~4pG_|VW7YEAv5mWNKU;$20J<5pvtpdn%p}%=Y{Z8X?+M!p7J6;T zpY^rz%-e2Z`7;h*V8~zbp;1f*XkOGm zw>r#}`LN|x!n3yykrZ?S4}x_cy{-rGcQ*#dR-f8|ZF z!7`6-Z^z)=N|NlRY{(=hz^mQ1RcNxZ{#&E|wFTqI*giILRMzmG6G$a!O*Srpa>{=Y zxMpBMvcA+ zuLJ9O&y<=WFiMA*W0IpSu*Q3;nCHV|tiwpt@*T0yo!gHN56khRM_PjCdfaZ8bWV_M z&)FO_g596Fb%yBv?UrOXYY^3FTDE80Z=W^O z@Lt&dsqy)eNzU(ynq#6Ht!dr9mn9uPcGiTG1jBbsLGudQHr6-m_}a^P_UjmaV>x?3 z%Y$+&w!Ahw58qwujlr*ERyM|5hhaw~9b)s|ZW#dVZ(Z3(*MkW*=C zK*h-#ZcuUs<6X-nuQyU4mq*I5kwkkAKoXZro$(Z!w5$(owF4~gHyZQttf$Fc_3@<` z+Jm@1OQYqrL^fvd4s~G)x>|BzP&%G`T9O3LeejWa#-P#M${5^`d9*SH0j$r@wA$6Y zjn4_~lmf-vSp!I3Dq-e?Nm)f-w~PRWF*3;!yfXo35cSAu9+UhiF#GUcxV=(Kfk_Tf z-GHoQ32zGe^Pm=nW($4V@Y_O=T*nG6l5e}IQGmYf+#6EMqph9~lULDyDj(T~Z#_+?%o-fpZ~$dM*oj&a^{m?j_e>^mo`G>0AL#^h-kSwg0-JXE`>)8c zvlpU=H~p>w;4LgJhN4O+cE==#d{!n-@bR`wAHzEyH&Qa1NgYh|(li*ty1^{tX!ld| zVc*c~YFee<_rie4|k6IY_NIklcF>ujX~d!z)329+XoIbP)^2EiM9x zG+=27pa<{mepmtQ40Rard`r-yoSvS8goM`y5e&6#OV*9~xkbsX{^Ye$bRJlT!5ZX^ zW@hH33XQ;4=7LKaz|YMr06Ut`n)Q$%BkKdd^VX3yk3j>$bTzUbmX}tg;_IO5_MI&0 zBr@2j177()*lV7^)LFI@YV2rG%7pSGSqubndGG$l5#^kKJ^R+w0ax~z+lG5sz!dVW zjDTO=2Q-+D-#S}Rej}Eq3cP1dxu8irKRT}- z1mnH4j_S9G&Jwl{n*}4%Yvy}n?^CyPISE!d_n*LY;iE@}_syxp=YI>@_GJC1aw*}= zgt7rB?KSN8`pU#OnBgHPTS@{rg+r`SFUuU6Nku$og0J29^7>UUKMRLk2Yz=zs})2YiLFW9UEBGE3 z!2(VJkRaQ46#&xs_ur7}o0nvmA17E2(kOKRjf3rc*S(FY3Y?-#^LQp3ufrW={^U_M z>sPJ=bobf{#%k<><2d!~v$6`ulUfSj@GAO`t@12jwpCZeA+5*7@!k~RWNLMsrs1+$ zfRxzn>k{{HQGt&DGa=CD=SswKxsr@cBKT$%@)`cV1^}@Xu6zTj zs7Y>_$cREJMKF7}yDod`r5wf{IsXE&gMIT=IC!U~!f&%pztwsS?9bHlYm!~O1P2#E z^04+F=u;BatlOEXoOxLK03Pi3XOJDd%P4t48kPFTXp*gzjrtpwfP`9wRwW_w-U-V$aC40U?-kz1;$18dSH^|tUj z*Ov40oo`%|(brDO@YFT=WxxJXUxG6BiJma6o|DTMw5t<6kc#l7ZE~0$F$)I#$P>*H zHD&a-R^NV8I<-9DrAmMtnt{B}&}p(xy<8LMOC_(~3Z9icmt8NQkz7Z8Ll({mR5map zXr{C`1Pn1|DK=8-n}=%aYY5acz{iGKrhe0X{0Q8sVC_%%8H{0m_|2e?+Bo|hiH z+fhqn%-Z+vbs4>UNfK~M=$k^2oDw)iu$-&u=J&#}%+WTIY2LzmC;A6kT4=9;7feDHDF<1$ zM4}=CgJ9?1@%A%W4P})>4O>L{EX=Jn6#=#1i>bk|{^5FJYn2W4wCSMx@q_jH@@1)_ zeaT_WUfJf^HsH)H-yUswWp#{KC*Qd&u@N+EQG=vSdiECDjc^C+-#UV%7o+Dy+9aH~ zfuY}D0`{?+Z%uG4RycE#@v_vA9lk?_mibj`Xot-+=Xh=cW2+E2_i~{_q5SXuon-#e zKhkV>wB#85$dRE<({B7)Z<_RQ$95$^HXf&nAKGCF?4$%JFTm;TDp5cH&H_3EVCoGF zfZ>3KYp2#V)9D!7XolP&X#iz?WKLjk$$>l@BM)ouAH(CO!CuuuEIb|WxNkj8>9WW{pdy|6V~U4w3$TkMe;u-}!jnuI1? zFPOS2-$3AyWvS)qw4L$HdhN6%&Flb{lv}w84>AMA(WvdrnE@IwkN}IVI$ap33?MKf zgBkuLTHj<3z=Smy^skDKG-pus^mMs*BU&C&*$A?_J#ZYe8+`q5GQ@ej)o-UH3Mxck z-C!7R6%;c_xXH>SKnwzc1kBk^aF#_k&PmN~vdSq-q4kp>GQqT?_~7%pD667%!v&Q@ zGr~@&1K<7g4S}J>^sF1(83PRR-DG{I-Q4isme=(X~_r>;ETF*-c>>~z)*XU~wSuGgY@>T(Lxap}C^ z^&JOTa{7@IZZ?bV->)SBW@@bSxv>4Jf)Ue>ca-haPL1I3jq9UMcU==FpWLC>uZW$N2syy z?a899aX_T{_53DwcD4s?UG(_-{kt^XEXPGB8fW?N``XGGtk=hOMxEBW%`W+zuaDlA zwR}!4UN|q5M!FZ{b6UCsy;pyLLdPK`oubeCUe2-Au?@VpE5@l;02Q{HY)6}xmHs1r z07>8sPgiAqDkk|Ew571vS_97-QXPYxQoNm#b{`~9fO&}_^S)g>oq%z?vujIr;8ORm zOtvbomis`(uC1p6!Ie1rMjdFsyZpAehQhR5tOhK)cwvDc{f() z;CO~(2nTcKw3S;~oWrwK%rcU}K6(7?BeL??IVpedWw#^Cx{vFL=E-(02B%zH zgP9VWB%)6p2Ur>o(G0#p(c|1s^Ji@u&d8-1(bm5P%*zGiYsQ7#5pf`lZS6;-BXdfs zvp5gFRdEEAZ0kEE9V^f!NAOg&B*3#|w%sww<9d_WNv3QMy!P8&m*&G|BMGK!hIxBH zIRoQ$qYpkX091CD*2s-Yeflb%J3vXiv$bUp)Fr;rb>yg={L~9*YcSht6#)75MH#(P zmdBqSz_~WolU<(LPu>S6H-kok+>&@t3h!(Ll^X~OoP4xj`iGB8Zn-Lj)dlEkLk{wj z0CfS*5{OPf55DFHvMXbjMfJv2IrW7Io4f`&l*v`JqkNd&d8%o&Wy2>oE&l=>fiGRg z`Fz=j8ssx_X*dZ4FB&Y3=#AaQC8_4uWT1G{ZR4L9z-&svY-Vv(?*#TM*ww%{BepoV z;W}~%96CSeBjL;!=uZ*|cLKuw?IUmwX5qXl!ij)es)JxN_raTT;nQE-?6k60Sd&?d zDFZ_>5F}r%V7!LNzr1cb&Iov?Ysn4QQ!Xq@*#gM8o_G(+p7kGA2dG~gAC-Ymwg$9| zQIm`#NKT{Virl>;13&ZE-0y?@Q^B`= zW^TeY9Xti%8??w+2~J9L+yfG+vbK?ZJqRFd=jUU?

    • )f1z{#5AT1?vSfuQxDQ*Yb3g>?JNb4$jB$kdFV z4>cP5r4@QNHzBy-Nrfkq&peuii+^-yoBS4VOq~qqFQtTMMDLj#?U#)u)Ev6)k*xdI z_99kY>G1u>U)Go$?un0*(E2vyZZ%Qo3-emTqnDD!9NF?9^M^Yw40kpr`o_PDoJk>; zp}RKgoBc?fhv6MsD8SGJ6*6_cM5&r2EDcQuBrUdqP)5E*LOHRchYtin7NgJF?W@DHhIF#cVl zLOun+CyDQj``#eg6=%0knl|f408kuEOCPn5{%q%}=tljp6!uQzM(*t$O+N6j^K3wE zr%6qoGQ@n?2n?spz^;x9&uRFqx zunQ_?nFl5&=H)fQlV!H)HCig|m%)3q#wWJIC2s8uJXId6PGiij+7*-4ez*^H0_MWF zL{U89E2{XEekn}PR`9m8%xvv-Xu$X=k{&iZB1MHK+~h0 z-FS(+*vz;WsOZ*I0l65^SuRE=L1f`M9nfyK}2qvm!<5;N9VstSVuUp1XtQ(#U%e`)BPFd?Q@0uI(=rR zso(P=N39XsO0PZaCg!|-DpeR3R~DRV58_SS6kbP(6Bww(DKGCV+Xz2c?i%xrQOVL= z>wcfXY+>&)QfANEj>=6E3fP=?+)U12K%(+Gi~?2#$J1Wcv9PFn`HfdvcH6IvJbq%7 zCx9rwk_Va{rM|ZBn04GkaW>i(vfG2UW*Et_wW$N<>QM7$HMF4}Jdxh@avREANlw2ARPAI3$@b2V2E zRx>iP#2dgT4{B~5b{CsI`t|2B^vTIMFyLTeUcP%p|Cl_`N!}iCJ?ZwGxi`S)K)p~N z|EIm)`cGZ;#RY2*pZM#=c*~zN$0wab<-Wp@K2}beL+iQrh!wnt{MQ_oJoXKUnc~!M z-3w8^_apZiQt!&?*dakOR=T71BV((0uPa9vzEcW6XIPAs;GN|{iOm9r4hh0yAM(aI zwPv1q^wQ{PbqXkhl&h2`Vs7s~P1t-2sFxWGOGIkatY|| zJLmm@=e3i*kwbgWkL`IMdCv8w&E1HfLcV8$s#aPQmr|Q=Z^wGD5|^qw|1^90vs|~WFwO=VTEp-ajfG%!7DC)3(T!$6a?%# zqofw{z#R3it%PBaGV2Rp_3xclP1UG}Q;cJnEnq51i8~hdygrck<|vFP*t{w=(xE_% z>5_e9V!t;ji6l$JSE#l!#?#ua-6g|&Rlo0IkHly3!0xf&%7OJyR^ZW$D9I^p-}>M4 zeYJs)BVaICeqP>Te-adG6dRgbRmBeE_i2S4Mhkyl^(DGsjW1-z@I86)v5ssf?Z<}G z87?SWoh<`T8lM&B712U@#kA(xWKAte-zB;ofF#!ftJJ-ptD85^kAA!hOy2Uh*?cuy5{J_f5YW*PzA6J}u#o|o(vO@vKwcNc0ij=EkXav^%14F9HWK^W&N&L|fz?3D*Xn`xs2T(@3dZesTO3g+c9jEvwcz zUq*&`Sc)NOc6@VW#YMH~pnF$J7ao4J9Y!0z958q5FV6ZQ_M+=BvqeD`lRJLX0Yzu) zKeh*SxrV+B#q)6WT7zq>ra8xCN2?cPH^(^m1^BSLL#6Vtuwip;m6qnoEMT=yd}|GI zJ!DAtlIDTB;W7+clbzE*FVqdJ-6i3H8Om8`gacb8!zH^yGhgDBQ97HO17fIYDY#`Gm2BRqcvQF4`Kc%<_^XB*>v!GI; zTRxR7{NbE3mh|n$-bs~Kz&g}3K{T!I<7g1@9vgIsxvV;Irw^F?h2lEtRy{3<NCEIbDboR4M(OG52*JTR_R z&qWdLGmUT?6>Kp^7gP~1ob_t=Pn;S`LcR`fXVv6EDbOMrRrk2w0 zquSYZpHu4UFxOnHR(Hu$pz0MxjZ0Q7%ZcZ915ZkMaa)I429ovm=ZF-fp7;$ z5ifz8j@3F<{|pZQ$5H61(!omUUdr6TR?iEg)De#^0yiw7OIrQ6jLB-c+ zMQ5FBfq|Sb2Wt%}yyce!pePAS)WMH7dU*|MrF2eG7r*5j%pcJqbe+t-z^@XK@6js# zcZGQYTyR`l2Kq)S3xUI=r13i3^06H6%@N1nHp9?e8^0K^>A#hcZG>lc%^-Ys=)m?l z#2Qu?t(`qiMvT)DQ=Jrdb%fnbrN{C7dc1l2;>LsD13g8a&k6&c`lbKV*6|ru2QLVzu@mT0s%d=H**;_N>Z|7ullbY z5dAmsuh8HB<|#9xZ;#}-*qow=#F^7t97VbqeC&8NmaDV+7GZX}VmkDJY9nHj4{~)l zVQuB_8GF}QM?tz%@y3+7eM2-ZBY4GTI-y@q`=fp>T91n8;(f68rgY(TY3UG0un{cG zs5azx7}vzg`M=6yQBy#0zib8ykR6VA;}`3lK$n;~uqxY~a1`n{+9VGz@c3U`_&*;D zo$)g{Gdv3w(aMplbfkY-T|$CPHN{C2$!4OIT*CQjvNJUkOGhxyg^)MEZJph@?APEfq)2ssDXdUY@ewgY0|s^uOt~W)<G`8QP=`4%GE!(@XsCiuTmxvs3SmD`;522H3JK!=7ELfJjmmFP8Wmw_P}k-C;)kUeMZ2Vn;Uke5iU>1PfHLIK2Jz(-4>nxkGuX8XRXyv zVW4!5t(X10O#mA6g`P)vfNH2Ord3q>Sg4H3g&&bGNlF;q;@AGqsd4|of%H4ro8=Y1 z(=9o#+}f5gZ6sB^C4vZnoJPOst@xMBa9(tqUy6lug@T#AHoiVds#Q1%6|1KDs0ux5|h0|WD54g(d-qnAC> zTw8}qfJWq|yI#(1^x(WWj&@zybcEzw=}uY2<>q;b2e~OA*6(C*fIbv}>8B>TF^$Oy`Z{x92 zR|2$7f*T7c*bBGtTjJbCWL^(!#)_J$t9}}7;wM=%f*xzu*>NO?9wfon@-OcXfpk+|TeVznM&)`nv1<)B%^TNEfGk zEAwSEghC&KEN3a~1e#NA84GP%8sI*NuY!3aq2`FWPO*w%otPaE7S0U1(k)$bpm!)U zG`Bf}ly{D5HUb)x#1HW)Q(ypqnUxG7G&>YJ3OWMb^PW3lx+^%%^oTR1>0)$s2}eKp zv_*CB+E45YYjP=&k06s1pAl}=QHhKPVIeL8A?7;PdXnfn3byYUa3{P4G|@ z3mZJZ+A+>nK>QQ=*{zn6+PrD&ASg@~t$Z6^i~BbW6T`~?Vc%_3C)J`~m?l9(;qpbtml0#x*zXupelbP{+ zA)-+>RG)*D?XO094`>1uEF(gli$PD7HXSE07I%v6Dx+R87yql?powjo3u{xTLy+0y z4ye0DB+E?VEy7;8!I7MI?lrFiV4L72mNeUU*BWNNZ@LEG0A7{5-&_L7- z^w1a=EceE+0)9_Pc&iVsY;bG8>MM`}ZXPQ=c6a2|Tnwj{j&aWZ^pTV=-^tZ&*&p2i zM;W~`!?b=9L4Y51jK9sw7`pNmz#g)#{~wP4w6?lbb^pyFC|_@BdBXg(hZURbq2jO8 z-BkaD$nV&^@{sJsA z={Y!r1wWF9Z&95O70!m&PNrxezv`s9TH=NNXZzxxLk7rgT#vTPai!>cMZh25Ceu0_ zxgPeiP-%tH{LjjWU=aP(;3mjp^1YI9F#Fuwu}|zp(|4W0cU#1rJDGtNVaD-yr_lu1 zD5<}sDIf-+MY{JjDa>l`GBQo6OQGA}KP4H09t}L7V>o?I33K4#-Av31?cenL5;iVQ zteS-342FulJY;@~{{o9Q<-DPoMRd35N2*JlsNUJ;AQ{vC>w^M)i|#ASZY&D()ox7HUm&pI!0mms{5&sn~=yYhg7FzlrM9?7#d_aa+6D=EP zR8#!S=WU0%>s3v_P}(0HE0`1X0w-@|A2CiPg3!S#YbSx;Dsr zKfYzB{#EpsncC0mRgcUK)}c@QM$(1Ot1RGWmt-&TLi|KDi7v?Z5;iI6J0eKA1#bRS zj_;|{KgFMaDG=b7fZB0&5|WeV`t2A0Y!}kO4fK?lD_vP}7Vm&*Te^$QMoNgzhLCi? z-|m<*>EHp?)$$Rc?kP9-@ljB-M)oM>2nFl+>u)L&fz7g2k)kr;oR4lp zQ&oP+_5Pgt+Q&ebLg@wWk02TBmsf9>=-m!K`XT*-LVD1@kk+$sV2jA#l+x}ho(1QK z%GKh09NmXsJyu5id_L#re}LqtyncgDyt3}6o=VTah;V*GMoGDOwDTpBjHUB8aoU@; zw`B8W0X|(GBi#uT<3!yI*!XlWF8%w@9Xc7FWh{~Emw){C^r9n{ZdNDchavq?tJT`Q zw>0#B{xGFR#s`df32rBx%@;em)^i!>y3D1(eEe`V=Gj-eO%`CO)xNSK{r3kZGnPJK zreE_fX|=lG_T-csO3>mU;=d4k73=%2dEM!j4AkS`B7;89{u(ka(3-F{mV0 zRDMEI-1szUXT17EXwAuH>Eq{{r_aFp_n!rgKuqqh;}u5K!#}?`z0NmK9x>(UL%k-& z%;HHS!SLq<=q3pUu^+EOLFPWix?T#~Z%!rP6puT{l{MRz`@kI&XJzT->2ttA5?Yf+ zkJXO$eh!w?RPs-@Lf8EoyYt@Y28ZFQd9Zfr!3ON3%DHQf z=pGN4Vm7ukHbr`I_NV+lkN_9R+{8vu@Hrnv5VTs5sqf@!sBCRuN(GjtY6(7CL&Zp) zaC;b^SCk~fkWX!F!5BgT$(wH=JT@Zyko@iuG`#cik*Q8QTEEBW`^ibR--q}_a8}ki zWdR-%Kg7wlQyI){N#Dm36(hl4|P>Ja&X7Q8yN-kxVI}_8DH?GetM>ce`x9$DP1B`JhCt7!dxV8Si2^^eC${P zhAuwQwZ=Oy@EVHG1NMBi;*HYzLzQN7#k)s7@>fpB9D|?P^St7)C-M0CW&N-bH!s&J zy7nj)mYjGOQju60?|b?6@CQ8Rr;482FAHmbz(P^D?o9^b=Fj!N+;=r=B#Qm~`&M#}B|)DPU7`mLCX5!cp6#&j7~Z1H}% zOcNZ%xbAD0qT!IZnx$zvk0t_e2hPcA6%@2Bd8gRFOBF1;Dwst91-T*PA19MHRP9M= zwqr>=IQQgBfm)}E(zjr9s-XWyoj?R2@;yGgDkrHI!9={y}^%9nyo3nkiHZdB%nx|ju z4)rH~C&PTJcLmPa`KC=+^{;{Jv53^U(<{?^TL){Kb_z_JH*AaeK@~Tcq#D*0kpO@g4tohH+j;!$>is1esT|k{_2;M=F z3f%(dt-X$pVDR1`xVQL2Wp0k6{cPrkE}tjxvE3f)z9S{V)q>4C=kN54+$kUGfz^T* zL#N?1MP9=G#?0eG&&iDvWF6ua9;k2YjUGMV;O=i5pU{2X5wa=I{*LlH8qUvJ`##!# zE)}kGG;b%zVoIs3par5-syrq}CB$&gWyLE-Y7f6!t5#?*@~V^J9osiq*z;p5B_Y`SX`ut&%|jrl>|rr`ul` zzY%@@k@r(qz+|E>F-j5#hJ(Qku+<)6HnbB=f$9#U|e$0T4r=jRwubh`D zg6{GCrJ3Nzk8QYF@bu&f46q&UP!rj!Pa@atJ@$P6h!F$*u?n&r@@N?rdM|z zA8pdvf*fdm6vHHDISWz>26=>-&RpFr!7XO(=oDK9<8p~ZLG0OMSs!eq^H1tE*6;xdmt)CF?0>lp#r@%+LL&BVKDOdCgGlKc)JNm-h;B6{mF^1HCX4)m}qA7X9cM`lDShrF!`zxk$Wy4}PBNY1oG2q)kdOcim8lL)ro;vtNEdD**YZd$VF##|{Y zoozaZF$>JUiSzM#OSrYpq!SM$i&aBVWjT4^W9{qfjh>yFu;(Z*ekd_HYw-(=Auo4t zhL+^*kZ}>^kdxY;STJfZ*j@0APV*|)#K(~;qKcoH&!{dAhLNBl%2nXXzZCC3D#-Tu zKDl773j3qHCyryJ-w`1KyM1=hXCsw}`P=U2X&p~E_ReE~%yV6T1jal&yvYLwIGKB3 zMHpp;xA~}*o}W(qYm>EKT|+YIBiM|siD)$zOO9!E4Gs`khZvkC$9d(Q-xAWK zJJgo7%Wg!5z#Hpy$+OQ7{Y9NImcgYqb_?*%J$ExV z^;X;Ekr|2@ro9}G6oS2NpWR-w%Fl#?XgPK7{X}M#!6-jsM$zz?<2TgaC-+Q7chmMP z7>kl&+LvvTcbqkn`feEneZ9h2Xi9Y6YNP;?fb1Fv&H*_`ZR$YR@89I$hd^X(kW_wP z9fJqeS~fzi0rLrBcRQv2WE8-lAEtj8Xm{saqUl;KUoFqK{dGL_rUv~Wuhnc*J|MZJ?jt=Bl zPOqSSN4{eue5J@H*9LfR^cZ)GU~%MQ-t(qq1!SS1*?>b@&iUTU zb76p5>5uQP5Lh7Ji(uuGYFr<56kH9#tMCSaJ3q0Z_)5&jVav0eVabP(f14lIEb!TdAr{(C ze!1h85A>BNbCCtuBe7D<2H-!RCtQ}!6W&P<0;lXH4wS*qTNhLfP$=ad*3Tuz19se` z2}gW|`aJCCsSp;9KJIWq0+2ak1`}K6*-EBZof+3Fo<1~y$?|ohnpwCAuzXK6Mp`EcKxUIQ%kE150o4^e z#L|g=(hiN~G_fUk)s-X{H@-IC3_hx zZfS~ay}gfpaVih7)p;NbW|Hr_Cz;GZvJdS}BlvGDz)aW&j6DztAu2VHZ~73(ahDG= zy#XcJeuG+pgH5qQMjAsTKuqa^OI>mtUclL;LC8 z#^DWtd^SRX;C$HvLkxD~L&SR@PI!LpXoASv3vnb0?k9^(ktaoT?nQgcXpmhH{>B}w zAW+)nhV~2Jbh8y|4RZ5CP&;+_se+ugf)b3+$|8>VO69w8zg%GWX#766C!aSgT_rwf z;V5dmfybrA-U4PONB{0lXzT29hVv>hKv1-UiD<`$=+gxW?Urr=Q@szoXw{}|dX|lk zl&C>aWl;_+d>H|5IMshYz=*)Xq{9i{y-A7@OCLGai+1&}`jKZ*-5&=0^TVZjxZkf& z-Aw1vY9ExY#{tnv*m$Z}R)+M3lqTG7GKGTnO5vM%2Nre_A4&Q`|F{)EAVH+o)?m!W zm&fs0VR{(E*4lRs;I?#PfnFV*-bZf2-5vTMf5X_Sv>m=d&dQ=EE&&AAs#)QdH=Hdy zJNgzaz(%Zlx-3@;g^Pf=vQhcTmmU3e`6TqqVOwPcxf@R|Uia{8`?3bHaiuNeS<-Wc zi-n_p%0hbkvhB~$r88aT!=NZXL^c@ojm#qBc2Z+EUN317IaP%G#>-QB@h28J;VikM z?&#lQgDNh(a?56i)cpfUN^~%_mi~dJnMP^{8AwInyVoWT?ZMoouR{Zib`|Mj-u|wK z6L%z5R2U`9cF^GI>{_QB^@Ur4?`mlLGAe%ca2wUu`Nh8ZhCM+4ThL1Z&Ha-gKE(dI z@~dHvZv9fFld{B^SfzZ!kSDTLPeQLSE2qBd zZ)rooL=Lw7KIope5g`8hpm!trXLnAYK)1ls3rzzgN~`DyX0qriaBzM>ZeI)3&)d-l z0E!jcAM<63$29irelT3V41$W?;c`PSOSEXZA-w%E5aVvN!#Y_CDFV$2EqbWXmDjI< z?OQUk-UL%e4GlaH<@^LVx(py4=^{yO%rO)XMe4k*vA@fI5f=ANUav3>LC`?M$u5Z07)v-&5T8no zIa*2{WxRgVYN>WSf;c&7Bu8M0YDW1Y2PO~PMZ?Jl({n%uky8YxA^}23i{=|ZCX`Mb zk^}gM+QE2*ver(q>a@4k@~KKT?if)3xAMkg`wi8Ewn}A~nFs^REx{e?``W^;fiZ^k z>(WuKvcbxG7T`REG^n8OUJK~9V9x_h(enbwzoQG7_H^5Gn)Bbt-hX@_NcWg3%P>GU z=Iby(G@7Q@o3A8DHRa3q07<;8)txtqTY6#oW>`dq_$FK|H)zr zc*jAt6>fE+ZG-J%GpKAIJ4z%HwhxTGcy$3>_SNXhgVxJZzhAKI>kHyo0IXY6ttj&4 z{5Bl5!9phGeWXcc7&%%Kriu@AbEIC4ajBZ%6Fi?_ciP3F+QBR@$Qe>(ef&r1uHR^h zWSQMxo_7A71vI!l&*m}#xnsmFtcZDwydDPHi81TW_4;#WDa+Du?9kPHcYIzhv{KWd z=qy*f_{3^b&1mwT$&&WF%2FDVamakCItd>_|5AZTijUGUztDMM%A`l(;Dc8)gs0&i zvz@womR8x_(I4jFQK84@cu6_-NY#K(JXD+eFa+(K#ssa7IGV`0NUi9yq0|_Lrs(2GiIP?Xt01ed@9QW$;e! zI9tE%e}8YLZSNTIxL1o}G?rMj7Gh`ZKvyOwYJZ7`3JN?BdP_R|GVEx0(p!M=`O-OJ zAkvQQW;K;PX(mZy^|lV=FTE%ZJmGp=-NH+-m7mK)E)&4FkJl=>V?I~EMvmKwap_Ww zhQpBFS~^0I;?#iy!dOL|A5^2F@1agh4z@Bj`6YLi#9?d^Ob_ef?lL4nAlV{#rKd`^ zaS_QmZ-Irg|KNqV_R$8{RT0;pi=ra#H`N!aYXMg(T!XHUC! z>h1#7v5DegJFAb~F)V;jnrWz+CbH0c?)990R`>B2f!TV7A&}oYF2@D|{$)9!tin1l zOTL?{&ZtD@&|U2Fa(uu>E)A;fT5W9OnOp2F{ukC+LyDh2dGJZab%ukGgR?+6&zP(o zBg>qE_;szv>JN`Q`5SvgTs+ASxpV~{3FeQc0+NtbjmBY>0wj7ep{)c@mbHMy-@n!! zAVFl-UFal5o=1~#)a#!VfM^%uURYZBEfNt%(W}9iG8L%Xc9+F?kz%G)D}N zDtqOYo^h|Kj#0mG0a-Eucx?F_%^x!{y^brdpuqJ0&igUo4AYkPU>|}!cl#Jd|E(c4 zAnr7xuAwd*7Cm4a$S^Yh#PG*YF9l+eTK7@=gE1Cg$ukK%wrJn+@?AY!sx<<146qF& zv9Cmdjx?&ZV)G?OO78)ozOp3|cvahJHg)alVWcO=oE6YyFCgU+_X9^666?Vzc8_NQloz(CfK8 zt8_q#>xyu9Ev)=b3HnTw5iJYDb@GDEf%^DO(p=I{m+a4D6%c{?z}`@IkI^tTJL?Xm79 za~EoHgy#Jki}=AM)nCEolef)22GrpH>!1GlB;tb|;XPI%RF>KdgLr-fb&XdkMxEvf z#)&Kh2RHVqr=4fxc(XU1i@5e`2QgRkMo{ZlIj^P2$RiNOMotKK%#Ozq(oij4=*wl{ z&R&{#W8WxMqS_SsW?dV}MTa=-c0LV9TLk_!APU4#W*-eT zr|#IE(n%`c(vs0=`w}adh&eH5K|bJ$Ad7DPJ_bu4RP#8ww{ZUS%#&!s>fJUKK0)3o zoT5muph~X#OJH;U{_U$dpJFUfn#E}NzD7|u3dr9ry4C(rRUbT=JyJJ6=b|oQUuJ(? zFe2x_XEsFsh0x?1$x1rq?Njl~=;Dy)Tg=9J1h2r`qr#l$PaQDf+B@Coep3vD$l*oS zlh~F{sv(xIKsji(Zyo5Y_{o(hm;%+NvRsud5q2_*-}NYa>0sbh^%u7*b>?<6WwVDO zl#xrS#Yic@ESB^!iM@BU`s}$m3szgvN$rjhIwu}4(Iu(J=Y}n_jx%1q`Su=}5@iQ? z)D-hFIj4X5LEa}y6Zf>I0*U=SwJSFX{Y!P$%crG;+%_|het<~YhP_93MgGfg(gSeW zke__LYZnfGZ$1ABGTzq>N31!XM;7rcoNzsA#4Qz9sxUWFQx`I21Uj?5H(EmLTSc`!Tsy6s9$wB2p2i} z6?rmysyW^fTqZRsd#z(%EOeF2h%s-?5a zLnN1Wy{)18WgdilLSk%d#81|5(VPqQIxFjTC<@wO1#991zChXJJ!$Bov~dySp<}!q za12?v*n z|tRh>z;`1~78lZ?G)iN1091m0aaC08I9Fmw2D4aziSM%x)i735hR|&kFAS1kb zm`4^40Vy%dj^Qe*!87;q?O^#CPu5uZBLy%W!iA;;uXoEEhACGjB;N!|AZ0AB>y zbJaE19&_AIRBE_uI{0VX$<-Hikb-L9bwz3_hAfQtIC8^;AILO=p&*ndmu?E~JB~!8 z!N8lHkp2RGQn8!KDc!&0u3bU=*Aq>mi3Glx$icpE*l(a~bFKpUqDrnw3-B09A58xv ze_Qi3MZmLxZDS-bH%v`NRV_9Zj@19Y0;d|>&%!Iwye<+dBXl>0fOfh`);r2Fw>$a7-#ip!iUCWutyER6*m5KiQcuqZUQ5GsKi}EZ zO^aQB%QxU>c7gY^dbnC~<;p=SKxX%2ViC}Ab+cc0<52P2-Gt9{o^5*YeT6&s>_ zmF#ZV@j?5r@V>*{I)44=AT310$729@%Fh0bwj{BXjAvO;%5x$kU5)M+`4Xzo zBWjQEtYTb9__835OdYIqY$6Iz!)=NHeAk_`y#3nC6ZVa0uodd}H%?ZH;8(;^2Ey~i z1*gfSWr|)YnJ1UCxDn%W|2Yew*83AwoYm-7;ekkRMvOl!0xVnl7S+=EWkS>-2!tfs z!JjZWcuUJ6E{aDahw6EgG14s+h9t;bU!;)BUZdh2sAd^iEOmgy3EbLBNfeoqQl! zn_wNxY2e@hnn2qnK7$3Zkj1NOp54Bk;FuX=DJ(w#K*TzE~_Eexc4 zN>R6B$SR}L`6)4eQ&3B*9dK*u-4P9q+LI{-=aVVMqN^gP`|Y_8baWt&&(^9_l9PI9 zZ;Lw7qzhx)?>P>q?tUUEIZfIBEPCMryR)BWZW#P%h4z1RQsGv)6BBzu z01qempsgj4rT%htnp zj|y4p;p(6I9`odmv)N)8&}duZq*w75%L7N;b+>#+d|Y?7n2wAXO}$FXKMyG!Nawb< zEhX96&3a=7vf{LfQ``dHmC~Z-x`1jxh0R1~aVh4vThVH7qt+jqX1akB-I{rgEfL=U z(g)kbsL@O}2=t9ez476|Iwr)ZX7A)y^-k=62h&X)xH}>|)u>$cMB43C z*Q|`{$RA!aVPP32&lFpp&Nw;KKor><7BlDH`=6`#^OIry`o`W*W4Wc*9)&cIN5sT? zvzyk#2e>lbsqK~PixE=nd!=NT$CHd$%5V@uJ~z>}DxdH5&>J1;48H;T#cFjeg~Q&w zy*7n7pQo2GA;5W(?AM>yyNF(}*6Hj*w=}-;D!Qb`u0fFi#QLe~udfX)HO&FL4Zd|? zFghB=A9VV@1zAC=K&G7R5L*j5=B8$#i!s?Sr7H%u^Ujg}^}oTXj*?t&Vlb9O`uM5T z$&cL+3Pt553@1}Q($N9O)U~V=$7QwL9&2L*xPeQrrE1fzm4sdom3zn7UtFaB)TTo7 zd5{XKhKsYn)$)5rC0@HXeJ&(3ZhK-KaV=3KQ{+;D@1IF8Q0BnR z`60zCAkhQ7ulv-d>%Qa6nY6SdOG|Q^$CTE6XxVoL)%>UJY8ol#%GzGvOXrHN071q+N_D-=0k%=+~W9ffQ{4|z~ zWGjqoZyzh5Y0KkwY_}426sJ^}@9b(xE~wof{c=2cau9e3yt37Uv)ONcB0$&J>vz&& zMD1=a-wz8hd#{T1w0oS0c88qhosVjy_Y+*dD&5zPCz~0LCTS9%wvL)Y!#Ur>^&`~2 zZeSZl>98yO56;vn;Q$nPYn66(X8TI~vkX>8llHcPqpvGUUuFz(t}%6QjdlOfmnG9B z(Z;%RsuGvE9_I@GbeuC~2mtC1myaX)B>;1x^A3DPQLOHci8?kNv$6aAL+x=Yf~M<(bc?z@CKVs&+&t$R09kQfm!qMndRs5L83ewKG{uBbp@Ci=|orpmYjIGxX5i-O?=$0}M6vz;{vK-~aVppKF%uE*5LJ_qoqKXPz z%M0?-xhpgQHBbiFEwzLhgT_+AEwaDq$@@G$UOS-|L?^+QXS$u4NnYz%#Z=DjUdJ=k zLO!{C?H8fwO~r1nL;g$RlV#9Zrja@D_J!$Mj>{>~397E{V%jD_nxJ7OkxmJI6As1K z&X3S!?1^F-v9&~as*1RHy-@#g$UeB35{U0iJC@wPym;+*NTR)?F_F_rWJ9tv0eN-# z#+WTLNb>-xxbgMA()ttp52B`rv2f(4UvSe$EZMx8^9cg#}mm9NO7) z5CONkUJBHa5?&WlCa{o#-1Qu@YIOH9k2lXk9eV+U6Gp+~W^FFR*w)5#%+I)(ni}>% zdgI*K1!up^#H93d(ce&7Zve}4C7!w?={(eP^kjcwq&wh&PC6uV?!@JCR`8Il9>uf- z{I|cihay`x^*lxF@q2A?$TIWN9&Lk$EDKzSRCPg=5d(g~H5TF=*DWf6Oq`mG-{>-i*j)s4GPK6e@3o|y={w&~|IX7u*si92N-DBRmLSjwI=f(+m6%t!?k85#APFb%p z`RvrDy`@NZ{&QmgM>+}Q0A$43M?W|`iYLBjp(}^`iE~o3qSaBciE*B)g=blOo;&*x zt0>F@q0AcT9^`i-*xMDi(+&~^Xs^WhWLiN&lVG93QJWaTzrH&;022gj%&p(q_ZmKN zK)134m;-Mn^L#|os~{vO$PPNHH915w;)Xb-hYw@ z%uj5=sXuz_@x7o6RGL@Q9sZ`V*ua;zBIYBvc^qEK-TJNENRG#3BgWyKkU)%iTbI4m z#yiIBEpw6rPqt+0Ux7iOWsCCH3BbMYOlc)u{Ci>X5?<<=^Dz{7SjhlIZKC;&lelR# z4---Sw4iipf^Zt3fJIHVeg>Ty05Z5F)$>)$1g$G{A;ILrVlI=OF<~P+#Z++Y%KH|O z6-5lQL0hR@P1Fk5wCfno8IE1iOHO~S#Z?b^E&FPnZFRL`#*%yHaYI^L)9Y^bxSv`| z!_#mk6r*|9%I-kx`gcNV-J}r(BkvDa`uk0~<%K)G{Wp7OKsc}K7 zol=ADUeWCz`1keYR(Tm8Ka7sr|Bag0;=IvLJ+#Eyi)r7lV@=lE?5XZtW7T;M2q*4k zS*+wKvBIh!7f~y5M$q4Ua1~FJWid%Rpo=0L0t!a!uYqJ~?Ue1DL(b@9M9b))A`KQ; zWsd^`+ChngMu!-0r(9QZC28P~w1S(+ZJ;hWw*)5AZAvM_Uf z z1I-W28vb4HP2M9GJsrO*{sJ9%H%r9UkRAk&hhN$P5+Aa+m#1TeMAW}-VFOu~a1agd zrHy0396Ea9_o>W}V;Y7_yDra21Ld+L)$DAPeKx!9%yiTgsIwDRmGr*hH5ONPj8^@Z zsPy2Ri*LzCQf20SWn}YEK)3B&Aoh$Z+QHKVpjp2=5SH`Apt^dNYjIs1Lb0RUen1Rg zBgrO0Uk&2Gaps_7*GjM-HrFQ%3R~WDotWsuXypdgy(F)-(G}e#c~+M7Pn5mS;6XMQOZtp&Ts!!8AhtmQGB3xKS~$lv52` zjFTq7k00AMSn!_#Pw~I2EwUf}J<@Iz| zP@!yiua^2;LqZ~KZ}oCIDT-*;t9qID+rL-n8m8~#295yS8dx;x3a>3M4fNFV^>yyE z)##AO5miv(Foi_Xn+WA1_l3?z9gm;w0EWpP+qlU0JJ}DrX>43dmciJ@aEFNT>FS=a zkXazZu?Npz#0oG4($tupIWW7=ND>o@{}Mf$ z@I`&rj=P9&L*CWCfTmn@dZWH?D7B8XBZDVKl^YIieGf*8vbia`K1=aty9ua1dvZXN zmL&E}xhT#|1X=&fHc%lrh%ZK!35sLox6N=OJRvrJ4}0@s9dzD)ISuF&&*JBf=S5CJ zy_Wks3(ik;Bg*I8<#R^j&PQ!7HH$uQ{}*Op1I$x{u={S)d-%dvZK+AiaZ_2@y-@_e zW3^N9CWh+E(;=X?94G8J4EteO9IiqfhKCmW;n`#NjUzAjlf}<(TP7w}Xgn}!iX$V# zlwS<7k_EX{tz3S)S-8W$;uY}ViQgWdeRAB~Vd6`1bkR2x{My4xz8v^Br7N_%Us_7I zDzWyO*}Ohi3j$ex2L)eOucc8Yh|`8Zq70;&7(Tp*Y__<2YF=(inH!}UiUdr6M_xsp zo0sOrI1gPEbyh~2k93-AX1^g)A!+}!(Zr&$ZnVmDzb6?n*S1V)e#xzITqrC5nh>+D z!8zs8h)m(NOzbyPrh|1h${`$IWEdC{Z!vzQu4n9fdCkZ3_FJC!mx>UtyH!edoUQfL zP$}wP*8`A~O(VHAC&5pEkf&Ln#BX#}>z~yD^w3lGaCrMA<#BHHz-21_c9-xv}lKd{;`vJGMTA(FsJyy}YE+UX!nMreDot^mP)D-(i z@V~i^n>k_@^b5%$dohc}_(f!7q)Qjgl6|29EEb4jmB(FaCe)>43e~I@oEDj4gLzsy z?An%Z{fD=)XGVFGU0jU;AYsv0$kEd)Efl#S#rc$vJw9GjVIF2oy)oP#!ckMU@or}R%RqP^g7KAENdMw?F?~ZDI9XgGw za?sxRR*;g*7H+%6iwO>u>%#vXMiVp|xNtRSQq(t2MUvm-%P^E)9#J7RsO|{li=|Im zR$l4_<&2vaDQ7<{m+4c%=$@2fWueB#!jWtM>Z?=PNv>U0J0|qMzAx~loX$H} zIi1VN2=~2OLR_QW*RSi2cQeiplH#u8`Me-7`eEmK5(ZG;Oe8Xdm6;fM{5m`D9}#Jj z;5y$fM0%@CNKcGU47h$FoM1hm@hz4w_>5$aGODL|1?c)oZ@62B#g;Z_dTVDg=+#GF zTu^(CnO*dsXD8X<8=+2(CcsaAU)7&&*-(gw1vd*YZjpSrx*SBXx3f!ZwrL{fkuw2i zYHL?OGbS0s?J!>npYIqKEI3GSA}4Kv5Z-i zIhP>qs!0ZY^X6Fyl34gRdH+#Hk4t?Ri5aEcjMf&4l{07bk&aD#M47q3zK43M~A zI_4xn?d|mIjGu4AO=TtH&M8JT0Ch}@n&sXn`C*jWAW-QAOSpJ_1?zw#A#5=%HOYW( zKCll^;}H|bk?_6i#k2|K(g)b{D`xpm_6Cj`lJPWlObm`5zzgS z*4$Jc9LIL15uw|;xJskZeTs+^`rc6-qO zVl0M+8}Y@9_P!{v=C>0Ynj&*+4m}GCd#*@sls*v`NP?fZ8jBg9MRv01~m=WdR$-*ef*?$_UhGe6J@5QAHM@C+|QVw}^vYsvNr=Xw*rq|7~8Pf(jf%C70N zEDHbtQx70|V^lAr4kb(H(At;R{(usJjD~}gh??ET2(5u+H5R|ExakvmCf%7HYPf^~ zg@lhKGdqW_jPQ$s4TD!faqBL|h=bL2!X`I;0$+vDN{rPq8gQlB*F^hAdnwql;L;~( z$Yd_EejX7hObj{asacy0M5JS4DLmd;TK}%nyxAjIpHbDehAvrW^v^x@(wBZp8XKyA+R)m6BD@z5y^ zE%hH>znRW#cbAPq*&Da-Zgaixa6ZIw@qWD4=s!=)P0YM(6@UGf_WO(J zK?G1mj$~+6mQP;hv2Or&UX(0JVg^ECJ~2M=Yphvs&jX^*q~=!3EUmtDxy;uFJ!m_b^F&PWM`AjtB8gJaQUCBZvvco03( z2UA1~ib_jd4g#Qh005X+LFXJ%xR~2yl;?dr=18orNSghe-ZOO4Wg{r6T2~|@I$!(i z(VT}0t0rP&(*L~Ra_KSxc`@H!i@4ABcS6m4EM_wLVqWocXty9CsW#y9m~25HtPTCe1Gxzy>tS5aqfI}_Ph?ziRoqZd15_k z4DI*c6G>FZqAT)liC*$VkAJO{T7K`OR=i@gU#4_)Xy&=yrT0j0Ud@udFc+8oK@z>? zp7;;t+N~(4d_nJvAbHhN>ZvL0Y}LSTyPyB48+oyx z<=+Q`70=jGJgx%bJT_^*$#SoaTz9IlaFwm6irTlRr5{LZp$TxNi6oBa`OO_8%tKXU zd)7xBW@3n1Q{yLU59}oY^jT&5nyWGFA=6;#D|tqEN>8rPku7|{S8V?!9YEc(1?N|Fs7LXd>P-5I;h( zZ1$4X?!5)^DP2r2-4peL_Ev+=%8nJBb#e2Wf6Dt`P{CAu-s{q|{*lYJ$WHZ!u^Ot} zlUsjE<|S>?CtsfMLs5}AWyH4`ZLX0Kt)!IgFm2tFZERIaQJg+wndfKFSgBy?|D)WG zMOY6YUj5a7IB(#|x0L!tDLdHMU%Ep04Zs&241T!t1dAT>9i-CECEjss4EH+xhU=?5@uC z@v$zt+^b;mVsX>`vKP@V82q1l_7{;PRTo6Czc)cvj!y0~jY}Lk6ybQW{s;y-(C53; z%+&&IOm}#o^Pre+S>#D9N4%;2`C5$J87(GV6|nO{7lk>aUMFc z2?KIPqXju`Mhb9vUxKhpmP3*{f5G3q@ysWB6G=_5HUG1_vo>QkTg&G?f+zi#qXnJy z$H8+Enf;{8i;C7l0dn#R9p#=sL%YXjJz(gFI;g0yTfU)}gfbl)zYYc8BYl@DOUCf8 zxff5=`4e(;Mj)zviSc1QPfa>Bqxh?{5s`zIYVantX}=6dXKUbP4x@VWueEmq|LFwR z631w8tRV-n>l;6?HwlP}K7zx~6E{-}2c3L0Wa@3&t?uT!5YaKx(LF&=J6-fmqHZZ& zX-fbLiH$zA0e`|sZ9#9(1@}(b;MK4vpiAj7qgj{&anlhY@6ng*41!cEQK@W|r(H2x zRie<9_V#PvNecBwquJBZ60E|2@qlQ15r2Z-Xps_N>0&Yy2sDcEISFTFGxt$@MeVk0 z#9IiPw-}=@lG=UnUef3h>?p{~ceJ%dxuel8kZlKOZ@PrUQrB{jJ>pOVtgCzEj66G@ zLyvj6%#DSdHWgf+7Eq(^h5S8kslWbO(jIb}+S2fjK7=j%h*zohlpuUhe zt`|m@I%%hOYJ${_w$ELQZf2=5I_>R;nS8cd166KWnPTsS^I4!qSZg<|!7)D$Z#`YZ z2Qr{MXW#1nqzx>yT~DX|KmE}v`knI2HeHd|@}vW&x))?#Dl+NC&!Bs9CZ-u@LKFyn z90V)$UL27o%v`T=I~|6d7ZMuW*|w6QllKWwktrx4^f#ll#%p<4^7P33S)j&qI9$O(LakOeRrju2cr0(tKlgmAhaln z#T1!mq7@btFrPug?1JQGp`_ z1JhlrMh+ITHWw)8u+6<392g)S+xA}J?cl4U&=Sf9nyo@#$5<)>)fAQ2ihp8tMGbA# zNn2M!nel|vvn(>;uMbphTD9Bh7GZiK|6 zjY}$2w<+zzo}AiXPHvbBbr<(fSOHVxq8Wi!-Mhy1CNys7;Z|3hTA!&(wJ2w;QhMlHgsH=^p)m4`qkFRCPDaxM`yJRwI z!c(jTCzL{|d#W~#j7#x5wh8+P_8r2caL8?LJ@CaY4haoCX{-nIS5YrPA3u-%9|!`D ze7IS@$T2#JT5__{B#86oiB~TI>THUozNfxm3ZY6`E!Hn1q=|bA;6&`qY1#1|eak-) zQ|tq#!h8;~*!^<`%g!N$LfZizMaD)7S)pP=pV_TX*8C3|U50+oRQdcQ$wJU}>Lhs4 zetTQ#OszvA0eHFK(G0No5U9fy#9(8y3(eo%&jTow2Q1_7Vcojhk+HDXm0sfY(b%*W zowf-wU3`bRtD$AmmXR2~<=MT`*S50|{r zY@pT8dZvy^R*0#*))Kkdk=ov$Fhj=CQ(oWm&r|su3hK>cRlEl^X#z#DG~vnt;~G0V zn{;#anac-gh})V%O4fv)!(@iA>ljx$G}a(bfr%S-v9=Vc<$3lJt?Bq4Jd>NNwt9Xt zAu{1yQzj$=w5%)`^)2JU-Xn`;&Ob3kAkZ5% zC;NTL*QUOkCrW08-Ix<{S~?z&Quq6&=PrXH{ij`J0HQvYEe16`E|~eCT2dt^l0mmC zsbmt}KNcxvNN4T5G9&RSNBE4|t4VzlNkq9SI*ec4u#V?=FD5SGOF{i*$o}N7Y?-IE zGpC)W7Jj*^dr_~Pr?x!QitW0*jG#p5G5sn1e?Rb?Qkb3M59Edo1J{XINe7^pgN8-; zCEYL!17ceylmT~)5QR2}T1hSRt=ho(bS~T=e9V9^{3}3O2FpcQJT={PA!WnCbp~TdLRqSH6)8 zKE|xVsiEO9_IdY1fFa$lv`}9Fg!VV9z1hMx+0Kl{wr0QgHg7i_|6=r@On`D??yQpB z)xDes;43V1BbMPvn+^hw(5|LPOQg_qSyPz?&t2;Nx10ydCDYJ1VPU{znul`Ody}`6 z^0yV`uj^WMD6g*40#fEWVen3;#i)T!>%JuuTSP*4SjZ+UDc8UKO(A z(WLXm;+h&}A}W!um+Qz&j=iB$7OspVMdYj-4T2&*OeVNP_mrUiaoHD+zuU_99j}7X z@h#mdTX8B;Pi8`Ndl$TwfqX2vx8=_JXQAV1UdI5Pfs=_7tGsGrdWhw(ATKAyUe1Ox z(FTaZ@Y1$|pF0!)zg?Wyn5jcLesZN5fR==IFCD)OlD&drX3k@&SS?f;?Jjjy$Xq#; zPreRs#)RkJDd8ICAl}mcSwRftoc5#IT3~L-tOT3fa(N%xTSZ6LnW}BNf~w-G33D!m zOlg1UB-G18JeL9LFtZO_j-^Kd|M)TuX#2!n)M`F1F_Fx+AgxHG?A^XYPrhcII_&R# zV{>y@^L++eMX7Q&V%7Q*ncE?2qk)6_&fwU0H)8Ab$+;YZ!^H%LZe&o{zE+FWy*5Sp&d@se%1x`+zQi^cwn2KM;?PvO`?7qLKlDaF6Ze>pXgZHh zh05-I!+EX}!ZdGUl6a_C=IVo(qS#y5RQ1w5;^#@@iH$VL%Tfip2fS6lsFrGJ>6w+H zrtT6sW-ZBmir%id6UEE8^KP$w;l0HqclIPNyLMzlkUB@V?nR_-Sgb>%SJFk0<=^T1 z`eWM?aO!nD+u=vrRh-D#wW;7I9#ti*1vQTHj;pU>LVLSW_@jVnh6pc1>Vpci34>;1 zU`TIq_mzl3S{gYds5Q)pYh9$>uYbLywzzED)VOqvY^6_@YeKMFA#>kquB9cWh1f?{ z$CQ17K7+1Lb+!QNh2O~ltlhSX*&CMF?=4fF)ks>@XNR-Y7{L4kg4ouK#y&=R6u0s{^FF}|a5NQ3dKNjv=8}FD zn0;EF>7*NN``JvZN3LN5co}$=3eOuJ;m7ib++>=^Ech?_xdr4fJbD+(@5Qtq^uU1us z7-sZ~Q@dXy7ymHbH$wpFh@?A*74L>?FZ<~m?&3)zp?*@b^o(%(@R^VSOT^fUn!=6N z?@2ywrvvN$i%V&MMF^3iPN&tXvAGO)8kHR_a-#|xUKRd+OOmo1(`{IYwQ#hrD6rc( zewb^mM3Lw4)HST#K0O!dccJHsWC8kLkrW*DOLAeN(uoQG@eC|f*K_8JSsJ*=VPh?c z>=qvL`Z?B@l=Kyh=1w2`C@d_j&>;S*AhrLh__<@++3slI@vI~!Li;+z7RFEBm`|9Q znVDeGY)d`BA;@;{7XE~fbKV&RR4hSOvm ztjFP|R^M}tO@nErxTi~(fi`@_@L_)vsK8_V%rIa6R)`C$ z6l-*yUz-!kR$~I60$lr4l9s0?6B{z4WX54>*X0efDyZWqRS^`;^x&-)EU6~Z^t5ia z_OhY#q2u|i-QR|AoDTteGo4iD7;-%fSeYA7Jc_B8^hTqByyRe!YDqIgMv!1@f`wtZ ze*C*4CSJ>9tg_#&C$VG%=HVC|tw6;d7o+OMO@!2^@Og{1315j+NP30j{nh!aC z*V)~kfcF0>uK@W(>Y}tOsus|%Q1TPQ(3Gep9dXf1@fPqSH$kb%jjF3jSCoBQ_L!&J z_;7BWOVZr|sMcFqsjuXn9WUFo-FsGq^O2Sh%d1`_np#0j)XLv-T3D?rtx_VAl`y{; zCiYAJ)&2PeCW)^SRik_7d$5xeeZe!23^3#cKV~g@#$8KCn*h7VJmXDwPY!IgjI7)L z!I&JcPEk*;BwU-fIY+_nvyU8}dN z8sw%>oVvY$reGmC*>sDK%PVzklax40z1tzwm}^!KEry@oYoGmS(Xd)sWVY&Fy;@A= zWTKX6NKQfSbOx&@lkV;9bKk^xXlw9RDAQqBdgT_?z)g<-Tp^JA@UagiUVJNT(iiXm z)M*)+Ola@XR^#5D4I{5WlY_(NE}6$MPWskrfMSJ`Le`uy!LHlCs(NT#ZTo&>t_4LL zHeZLO&??|QflTFkD!PI@Gub<&C0Y9_VaeKxk6O zk?6TpE)~J2yb+6jVZX6IoAo!=_DLa=a;{L2&zu%hYbK|pKmkgk69xU%wZ>d0n;ZEf zG*vMw(x=07xRBNG>n2fFP3ucP}hJ|GUmjoQao$+(F*Vi{UTP0teY~l`@ z_&x$e;{MOzpe-qO0-~)f79-YivVT-4uD1*ePI@DqPS>K6KpNX8$4+x_>V)_sxR3mw zh>Ui{;1 z>I-leq6{`3`w6UCokG`HwuhGv+XX98i_?095rdBMdT-$HI(xyzVf7!_97%bq#%8IL zrKbUTv}Xze+$%V6$opV}Qu?XoL_G2XQ3ZR}EZr%~(U9n5+N`vFHM#36|Ai|+#d?V; z9-)7ER9XOiXQ!XfAm!lXq3wAlT+aan2Y6I&#?0Em-bR%I&JAN35zy3-vl~S?vEc>Q z&#G^L-5y#-KK+%V}i%NVyERWF!X6jkFHaIDX* zIHGzM!GUyVAl-q+*dM`gDY>hxt*rnqpy86!$?oRgC6py*+5IpD>9^$0lfXEFi=a>5 zEU`$A)?jO36Wm<0Z}U9Oqli({hYkdRgrOFa~;~>KvJ9DpFoyL?V;%r67eQabem$?nUkkX+CvZ54RbwSrnn>7snPt z9T&uC(9y>$d+vJtaiq%MDym;4EXZe3^2yAJ4CbP+tVXrONzoK z_Pq70iOACSzX!oMbFUOfWsR9CPrSZcc9p###?KkaX`^^(Me75PdA)Y8yUTn6+-$6< z^y2u{zY?qUy!jf(TdR~5uhoE2(@kjnn~&Ze3zIf(0%tt3M8B1xBw6La2*O!=p89~) z@7QZAggH_4VfvkovW}Zq7qS74`|9MtE&jrR@AeKBD>ovTj7x~D56x)Un5!$w?#i+M z?j;64H=@KVYQ_aTH?flhObX1E+nT6VbH53GOl&!YrmYV6|Bd1kOC9P)_bG(VLec(O zF-hS>UVGWB1-Yl>&(#xH^Xi$9H5YL;`=b;?X3vrOl4h}zW25}%dc&@S?D8#6Bh%Nr zmRc%WoV^<0&GvDzrDkxUNlY7IReg_gig=w^5q^*F7n&kbiB2rh4r-UZ%e=CdS)K)n zz}~mqb9g&(84Wn}XkC9>Ujrit^A3!3NdkqAq2XdN=a^vX)@nJ>veb2d(rUj-EmU4+ z|DZ=GCKhb znZVZu=b}fg%IWF1G_{H}n1#K@%4-^kv8w3x%CYTF%r&j#vo~x`a6L?|gncq8v6QpZ z(q07lI}3aO#H_XwSd#_Ki}>2BaTd$mfcx3ie68LNHemF7=4B~CHOKN+Vl>AxPkjlF zwCZRjgWIFp#^#`KFX_^7 zFBazo5X}11)rj6xfd*Ogg1TAuE~*W(e`7~`E?Ok|r+-wcD(PVJj^{$rV3I!ah zuVHUTmC>tGv0!yz)|n5iFX0Qe+S`@>ggpikv(b)Tn0s>1k*l*2tQ{KPZnXf+suL~qhjz?v9e$kI%r#?*8Ige47Q>^KeJN6XMRxEcO!u#uw)LIxYMP7& zC*r|r5o7}zw1{$3U8VSuw?Ht3UOv%c?RHR$p}`DLYELq_GReYlay_%e z!@F55*uao_bmCmJ%j`RCpo8lqO+_ngULo5JgN$vb=}v+8-E?etL@#kQ320_U^;$C& zHhNe0A(F>x#ua{gQ<(f?{CqQk4WpwyJr3{&98+XH;EHaw=c^9r?8)XgwuHA0!Awv1 zQIfnYqz& zoLhwOd-`yQA5QAd4O#RbUK;{bdqobcStBxxiB82O6GqqsW+E=uGDF#K+*Hp*`>kxs zOP*$F{N>`G%(mDJ|jN-(0s@9Ac*>*^M~f2lA!TB#Khb`Uw=-!lTIEYks5sWRk1W*d~IR$ z2Bi9sjfJ>_@a&e?TxA6T;7{*e)6udAA^}%a98Ej)pZqjg__lnBHuLv&tP=cq<1U`} zH9R>X4BmFjSlUs{`-{lMv^lc3psOF>B=%X)3Sjqy`bTZAuKi>J=M*lk)KZY29+OcV z+^sI7_3>7qnRgfhJDZHCd~40+r(S+!GU9Wv!;B$`rULn$*OV20d5yS>yVWDZ)|ZPC zu{U{@45PBXX@re&``Uj#J|gsmaAtcV7VeGcxcH!ASp-?+ePuG$4XdiATghJg&HQg; z6Cn&6>g7)!HI4hnsA|1xu6bPUWM^AlOnZk^hfl(2Xo6MpvFV^`_2(!L=^z%^W*NIc znGwR3yNOnbjDu!SC60_*oefjl!wXc9&Ls@z`%ndpGQsh2&JVbR7Zl`U?;KM-^L);4 z?eX02`D=Vl8&CU8=?>`Ourzf2ME2fIEUDwlX_G{Ll_xplYmcWCHMfLT2g5TD;JP)k z-GLwMY@{cSs6{f;O}o2zJM%449{zoiKAn%|4=XAZN2hh_HH5&<&{}I9#uQ`bNEZ2! zoL*$jJ&9Op?fGJ9@}VX zK{%vKI#4B79wz-MJ?zmfQhsXKb<5AKtsEj=Uo$W>l2H$Wz^zY?{Qa*#Mg8LifCmQS z!7}$v`vwR$S03oy$xVxmHFJjY1_$vzg>_1RY=(7gW-4WoHJ6V94V9}MS1+#DtrlrK+w*zKgdxMW;s4pXF-f*Pic?|DTdfY*H) zXvuzi05dCeZ^gul|Fa;RdYF4N(|2v2d+55{eQ>Kn=&|Q>ybcehevdrq55pBlknDL8 zJnx3IyKfUiM-sqHkvNltzm}9U7i*e|i4sKiT{bE4X->rfjiUPOgS3jiihFz#AHD5O zMTR%}Aoo*9Zz`&4D=_u?nW*|1X)lN+f*IzsM`9z-woqMqqZwcWSWX$rK!EF6(k*k$ z|If`@RMb0?Y27s)RwsBR^@Kf-L#s+`OAB&&CX;~3((omKPc-Kj6jSi~2h_-^u%Z}Gv(tX+55XQ)bToIDb@i<5byUwn|gewXiR z^Z)BpMLU1^u1uij_*IWNScC@6Cz5@k6qQW8PvT=wj9(_9CZtvN^I=Q-F_mAa|eEF9>Vof?K`)iTeE|2qwq;XJTC)m2KHC`2N^Ex zs%ZV-Nn&Ebppf=KL;_>F z6i-4TbRO9S&PZ!a^*R}I=DlI3kgKI~HnHMn~Z&+k~=eE ze6&ipER#4PrWshLLhRmw2eCXN zxVr718GZ6R*7DGYgO=pxe;&OMtd||ob#C~W3xwQ3EM4(ckE0BUnb)M;8rTnCR4rfjgQJ1YY)D>8p}asnb&tothPxUb1z~ z)?qVyb?bQ#GlS-PSWfxm$j+|>`)B7hVBJ2pxoGv`$H*%w53E1xzf7;i{?2oiJ=T1# zwd9I4Z1BSbbU(==MiBh#sTTZk4@C>K>cwTdSy-8CUhY4&)8VBHET9i(y%?sQlqFwy zXzo-pkuU0*6q@+TgPv9uEI{A)J&!nQbUz$s;SVIn(Ei@y)Z9da(L#uxR`%WHpjXVr zspfBUW}Vz}okq%K8SWIv5hT^>IG-2o{Mhe=2`rfugIggkeje*5L5@H|4BMYfUth-y z`aC0Z4KPuAVj@8K(}U06%tc%HDA4<~mI!2!vVk*)n<3m3OUyJLeE84{XOd>A(@-*7 z>9I0j(&Ay%8esf*@0NV_;Ut(k0{idi#94>+jGf$o^Q>gZ3-vBOfYN=)h8v!Bp_aUj ztg^|%Hsd<;xjP_*z>V`8WfM64Du1^+d%fy%{_71n$A$4bLQZ@4`$2Km8X_d)9n&_ITukpS8x5g5 z^N-l}&&N`HgB-?8pHY46lJ#rn(;ohJ2rCduS50u{Vu|&OtKA zE@ur2S|44;+{HUhbUZB74FnsDOiN@uPby4C_2}b$b35rUwl*&N*~7ew-(9+iLzyT% zB>Qw`b=g$O&Q#Svt!@-Z%jtY<{Z&Kk4u@H18?;RSM7Dn?wFW(Wo&sjd%~Hm?gP_f! zDo!9Ps{<2Xt>}Fl&jTFoXGagUO?*=D9=~#rF|~U{H22uLwxW0=xi!(rkXwEP8X3V2 zd)1S*I;aaI(B@qv3@b5a}%GcY$R`OB392UUv;KN|TI=1b+e8e<$tDmkrC#ylgjY z;T4h5OgxDhsKy?Tp#lU$XVY~H`@JSC@MnvtjwW(a4xZKGZ$4K418LRx9b<5(1huYc zk6?jdXWd)NKHlQV;U*-{&9C3#6HB(iJkn@5b@SR`qW3XI9gI%Zr;*>nI_tmKb$Ki# zoQm9DFP*c8u`;4lK_`k21&*WaYFX56vbGA`cX>}w)phs}%dI;Smg+BRy(9yd;_TMY zL2Pjef1Jxa7r=+F@QVP2lC*x?&L>vOJGxC(NE>#{%fvuHO8`>`-} z?!Zsi0;u1Tbdp#&F21HR7^MP{-GTS{eQyI7t68!cigHlkZcf|ce*A>kIB_g!$#qtv zHum}V*7@gnT(aTK#)LYf%%s1VW&Y10sINA}HB9v(2;~@g#0F2^X_X~2^Dy*HeA7FC z+J8zKy@xrY6ZR3D=pG{Ql~nds*YX}S#9r8C>_aP{pdk6K^mx3KTY=vxF66D#;C==- zd9RZ*8GGikJ@}RXkM4saz0z=Q*ivlbw+wZinjTFj=f+@%d+S>^m*jT(%dO+k$#J>Y zeyzDJ`Q^>ujGi_Gs26;-9aVX2{t1~!;CMVBzE`L>8MlO>kwMr8K2 zbxgF*^y|N2!__e$K*>qc2a~V_8EL9_LL<3hvC_z~QbcB1^|y+A2rBp1o#MUk1QzBb zHIPt!ihq(+gD2+Qm)}+2Jk}@Xm2^5Q2)HXQb_MkSO}t9=SrISMtnW2EOU3z%=Jo>{ z&&$wkt^5H$2h+7>vLyBIqvf-%p;V$!najP)v+Jj;mj~2O^bv;)ywo%DmB4H-au5~o z@|O92GKc3+xZ_U4`@6AqyKvnBqPgSBDkWiY3&waAl(yzn^7WwogGp1J6wZDh&v(lY zJaz_oPtl-LDs(FHoXqj?AaJE?uR1h;HVt$#>or^K72%v!A07pYoYaC(Fj1$MCFlPR zQ{gZ4cHerldzkpo9>a}C-rAXaT((OKy?Z9ssI(7>?o4?fwSiSbg-aQG^qO@cgFYy6 zpVg!ml_ho%_w9PY%6+TW&M#C6KC$I>t=P*PWQ{%IeNF5Y*z= zdt)wW0grXT)C|ewf-Dk-DcY@7(h6YwvsHfMRK4$bZ$0GUiYxY|><{LtXmClm9By-b zWaq_xGKsIzn`5-xqX2-KE+*6iuLcw@XQl6%)_&h@UMN zUdipXrHiT4o^ksJv+u9rZHmp~f=;vV7Ju{F=0P{loe&!8tWP^+uIZk`yt;i{!w*(8 zkwXQZ2$*ohuzSUv{VOb%=W6Q?7uznLK_QfaX*Wb&&mZ>|zu5FT{!veVr}UGOYD@m; zL7yA-7d)`3Y36!8y1*+S6nEq&JX0~pIq(JI$ZKb@%wY67f!*mwv}opDw@V^>Txk?& zb83OzVzAg=Ne>^7ce7Li{rj0?Ol-Ipbg#em*$YMF>iS`1!~>A0J-?#5XPtP`iG=0Q zZZcYqX3b&9K@uHtt}*&;RTGuxc;3b{&&2IkMR+FgynD+)YIEHG**8u zq+RVAlzOID6?8f$jKBs0;xxZ11`V5_tRB|n0TUx7}duF*Fr;m zsjW!rHkLJX`Blib!(+#!7{{{Tusi6EO_x{3emCvEYn}o}X;}i1>i5^)^cORLOH9aB zX4{g#=v@vf`2m0NL;BryOp)7o^p9PIPVm~$KT@*lCTHX95Bo|#d(Q`FX$0Py$*tv;*?}iulo#>9H>25^% z3Od;`?$Bb!vD)0{qa+c_n8fy>&=l$JoCX{W+f1~wD}^fK2wtF;G!+mj96Rsd9+92i zzSVIWz$o8()(i_SDdpK#R(GiumW_>xo`rZwY2$(VX$2w*=GO5=na#{fsT{Y-%S;Jw ze)DvAwV zoIT{`BOtAr?I2YtsvH<#6*V%9PL!)zEh#RdJKW36m5pxO8+na{^i_3>Y$TtEPb)(g z1hFQ^&KUAsn<)NJXzM|apD{)iTES!X1%q(0?%ZpZH&fG{0{uS_t91n{E^(Is&7i8z z9q;*Rfn+H+4@*-v6CI2lvv|uEkAtU5Y?`ENchAyksi%VCqt#5e$h4pMoK}&$p=Yb? z$cEIUfy`cRK+La$=Ea8~8G;fTn^RkA_&z=HzIy+WWBC(Y>oYGg8fk>URc=HT{X+YBLI64S9oS5hZZM6_GuFjg6a~ zMLhP0H`B!h3s!a63h~974cWyu-!-qzv^X8b0zj+JRhYNuuS1P}eF=Id!)p7hbB2|| zLJ0Xq05>EF*ZDVZKPWxfopyWSniOq*vt0?Ct2iP2^PgY$Wdy>?_4RN7jmvFC#R+tg zT>VLVO9|RFeK&0wd2}1Ky*=SBBtZV-i9ta-G2bQT+AJ)Xb{~27^@7^1QwO zrA_}SqNSe%9U-E-b-0q#AH8ZTJ$8dqt+q$Ghh*|)a?KJeW{DyYk-iT{VNnM9E87{e z&6}Q7wD*YN#SDt%&^&#}Dyj}W-aS+P5TQVY_E~1BAc1Lxru}r=kvTHzy~D|&OQU=2 zM83t>w)#AE$|kyN?_DE=K8 zv%<(SZ08)J7ql=si0d{Ok<-b>FGWAyudC0UMLVdKx#@JiK!%DZ^V$IJms>H98Z;vHEF|<)IiK~Gytm9|Evv;=<^|4NMKa>2rmRhpTAnA7HPqpn zieIH+8scl~tn@hS3HQju9qB6@Ne+(3xd2us8?NiT!=zFB`IiUlAMcPqN&VS#GjgK7?l{ze`-3= zbPWC=8YuyOFflJ@hD)W$#S=X!$DYc9uc5b7&XSim`yRJKol0VLhIGjRyxBT7Gn2tt z!JcGjkX~N318syg|WdPf=grLZ()eFk+;0mL;dGZ(5V%>BZBT{YxBA-eDi&bF%=+48L157 z4iH#P^Pr#aZcvQ3ZCtcHHS3;J_VPWfeyjL6y`VNMP;W;`t}nJnUkMCzy0&9;q=s?t z9(_J)t_R)=_dKbJ;BQJ^p3M52v6uGk5gc`u;DbNN;QNP^ zQF_B7+vKYzT<|1Pr0I|!-1p|P2D-4t&EG+htO0yIaW05 zxY@b{f2r1Qsv1nh?;#wTO8+cl`e8#E09WmYiVIx`@Hiz;%i2CoC0C^-x!ch@HGk0s zwJgfo%98xIDHRkz$XcZQ;*a8&f$~&|Hk#|6$9t8L)=GgW>>}BlVhfgm`!$-d=U=>9mXi{+iSj?KOm-C3 z3r@H+Rux|{^I2^l!dzO)!e8N=G~~bbTYssc=%WSB)*?NMC3jyxJ$+1T>)EK4kP=ck z8|?lIK$Pwypd}rQbjZq%xeBN&I)Gfla?d@IvNkdjd9@MDrq0ogQ)!7)Z_1qU{CpWD zD~7pL3r8ldxvq{@&*0?F@)KUWPfJ!wj@_DMhA4YjYEHiy6*m4dWi%^RB+Pb{*l}3Z z>&ME~$Sba|x^$+b-g$pVJ=RDSM;`58!SyT9yu{LU2uLn=;k=KHWcROVN!u$KccimN zu~QS5-PK2WCfIN*5|;XF{XwU`4b3igaHOS`_02gVD|JB=QVA>&b~w7AbR--ti;q-R zsq}Rc?S+22-mj+dQ8nA7H4`;Vn>8egpTjG{$h$%9F77554W!4T13QT3;z^!9k@q~13!HWRxgb$Gk^Y^?H)WR<$fgj(ufR=F=wi&GtX+p-j5T|D2R6& zzHRVIGuL^y_o{BTM%2uRUq=qk35iYI)L>+w+F!A6wFrK~a&Y1pFzMn9PhnJ)*EaQf z5*v6nRMyuoSaO{GI@_&s7@Tf{60&>Fo728n%Drs$u58jR7edK{p7qn4Hq^91k~`15 zb(>|V)&OMTo*`K!Bl|?O-o-T-*Qki)fvZY#c!OM(Hiha0Yj)VFFM2G|*q2tJ>)Ef9 zN7$xWZ-cW_6Vx@<7T3A1iz#%WVxD45N-2Ie92=|$DE}s6|G_$KukQs`e!rxDpzs8f zm7BnjTsXi@a~Y0=)Z!)Wdw@e47>pR|vrOlW=kh1dP`|V}XyLx=TRm!=%6h$V!bqEz z1X9-UzuB=hwJ^fVbxfwf3IXc!ZGm}?0xMCfp2`A2pPDu(T_L}cVzKE{s4)p)?s8X1 zSeHdi6NMfgPy6P?b7pmXPBR$iM1A#Y9St+gIQ@m5@8;>W*2AKY5>nD`~W)>O?O+Ve12Qpd1zP0g(y65mHC^4DTVxgU9U4A zaT*yskV}@;q`_8=CYS6e_%vTlB(mh+RAA`X!C_Q1y&Oav1vW zH2E`G{=y53594UN+#&Fcyl>#shp}lU6Fa|Z+Zawzu>c;A$CJ|eujzY>3yZ%jV2w;a zYyL=?HGqdgU$|?<=2t018F%W}|0;dERH}QX#~%E#qtLQwP@&i#L_vBaV`Aa%E1JWF z(*rsh91y##c!^jFw8Pov>e&&dHAqk;6ovmtqyW>W=Y`fIV(6v01 zbZC|D#^dwXSnGcyreDv%8vjQsb(|Mh%P3kH55=dFqY30G0@`vI4W)~V4s}Uc%5BXp zp%@8Zd2<~Qc|<9{wGtp6HjSA^5=b&d_el$@r4IE)mig8tQwZC%?sFvRaIgbvtL?VZ z#Hz?wjCjcuppzU+zaO;IULYE-Jv)`gb52OTJ|<&eULz38G+d>TTKbM%PoKwb>qCM+ zJ9w(pW@m#0Mb;z;uK>S;FNcObT~JyLIQ;>_t)(S@FPwHMwHZe~*5A^-{As-=mf62T1m}!wG>MQsmneDWrDPB`X`l5HqPb8wF zkX|NgWko3SVs5UrqYy5Y1|?dkRJUv#i=sZv5buA@o~tw-d% zMT3o@mbL^f)K+1b=-v7(sNQMskzY>#8@rmlh0Ll`hu81s-omUFn%fx5ObGqQaCh+e z?EOOs_(k}&pIRCPim$%`6jTLWDEu7Af$;yn$j{DoBiIXSG;(DXZJNF8vjWcE`Vzkq zhaefDpzi7uVFW2m*D{$x`dx$GA;E#38VAH1W7AwiVwg3T@39NEqQP(7nkt6v#3ow# zM6&tke|%r=EOk&ctUA_Fr#QMbBi3~sX_oMg9DDQ#+;7Y^Y*iO7=AM;Th324cF)j|j z=CA`%bLFBt#NjnA4cFn!QLOQQq<#P^vcd<6-#=N} zmOm;b%gs6!3Fp7qe9qUbS3)h*@cfnt5niapA9P<#`FLm0Wf>F7`b-c!WnC7>hMS^T zr*CPX%vG@^8^)6~H}*bd=+M~Pv~o1HHbywiX0{g59fj(~H&0PP^Mulcb=#4-*W-UmIgadrQ|-+{nbq>K+4F-aAQ1RS#plnFM#?C zy6YO%bH*-sezA$g=XTR>?eX}5kNxeA30ZGtL7w!T#L8Curb(MT`f*83c6Mb@^%>~YlULq3O-!w1xa}O6HEr>D%r~|f*`>V564^7lO!fB4a#{H{ zWUkghyh&ipxW>j>iro~;BIA&N0cKsX&P(}Ngyk149pdG$Cae8$Y;|bhxrOK1%Myg5 z=yEQwIevTR)RTi_KAW@B&>XdKT8RJ<(MnhIJeWUDKX62TJDm;Gcvp zJ@rc6WW^bAT7!N6Ou8ZI#VQbIs0T(u&mgc-nS2pjd8+)(KOU1Tn0axHTMrYxhk5Kq=yiBgiuZA zl7;6We!pVo(e9{Hbu?m#?dzSWkrN?*%`=tp=wFRw94kD3sy}EzB~}urjiz&M)&o+k zsaloy4NY60M1n!4G)i@qR|{|I_SF>X?+_lo5BDs`0`INxnQVDpDbk1T87gn^V9QaK z8pL!c$Q_qwCf8<+Zo-h6&Zp_~nRxwVm4N5oBSd(Q2xZzdW>p(HniCI6dKyI z-EaD!Hcc!m~p|bn|?#0n+H8!sF6b2jHDp+$silRw&eAf6rxG?(+*|S z^U$F`pXh6PCSYFVew+d(gTpd@{L`5C8;$&GKHptDjN5Ab4tZg!|HS(tPX#x9ady6v zm?-_E7HIVqGiC*Qdo;UNwFNUAZiS_d;jZ4dJf-(WDt#z4n1NfU@w3`au{?h1Qij_m zzRyfih25@-b0$}Dr0MB-eCO8GSbGO^W-ecW=tt5OHy>rcGI3ZaYB)Exd7qM@EqySf z6t^lJ;Fq=Ed`Ci|sv3-ZO@@R%I@0#(&728RH2Ic8>pBZU<^-|97?z z3yb$WgycwuU|fAy-V;?0d!gJCq+a_`f!XCGN=3?$%I0PXE-UTR&~`mBcy=6T<_h|> zv-11&e^MO)JhBGD86`I(Xvzr zsYMy4y$5*5r39{eGy*&5m|AN60+No`Jx}igLf?tN?fKn@bwcT`%K514QB=H5tJ}TD zKU4(IOBm5Vn%LHiZHwpm;$1w}(BPodSpJi)>haC!SRJ_x&HUMbW${h;AO{MAa;~e$ zeV9>|`{k0I9?v30%iR^c2PSQ6qO$wS#ZL(R?$JYfLe$|PoutgwE<`NmZI_I^InX-# z_2xl-KC>}N!Z4~gaYMQGH`k>m4Uj`ZZH2a{Tv{8wo7hZ>SL7s0=W9E$9QCH0B+$?^ z*PXdY>Du>woz(e47; zCt+4Tg7nh?Xos+MD(fnt8|G%016l)qnvO{Dk_r zQFwl2!KAB4w^70!7keb3&7e$sMcR2`&+n71Rps^l*Un_3-(KikXJCaaxE}c5?%$p3 zjKwQ>cPHH!^~7xPwmxOv$^Aq>;C7Bs#VksXdAsTp;#DVIczXVFsY*M=U}1WmJRc}1 zR*EY49BldK9S&y2^BzMZl|2j$CPLQ0u0MSI-)wE`0TuvjoCIEIn6VRt40jp^w%1X1 zJZZr3zTC7lh9s$S!HLgT%@{#dTGX~{D>mAXN(-My*AIg|9UC*Sx5}yfu6E5!Z5F z{nr;LxNvn+_qn#7^yv}t`;^sY#rocdEib|>t|IR)*6Agi-cDLAsO&~urm1SzOSOC* z%rgufG(X}(Yrfhzj6h+xPXMf!`HQyM_?X$ZKO6v8$1S&T6snUGp*wi%Q;W9)I+^!@ zO={H^S7#i}QiP$23oh5OC*MY%M2o}U@H?)N!v?Kjd=t6meDZ6=eXD0e78GjMWZI?N z(y1kYW9aFED(Ew@4_IPU^&`vnOE=oQ`~Or^FD(Sab)XNz%cwlc5)8F8ihjL4Z)S3J zV6=VGt16W%(NO%!_G97?8#7^AO^K1Sv}SUQ%G(p`Y!UHGexWTXL1>~V{CgB9SQL}V zM{G!Z)gaZ*n!o)NUoaP?$3pdNymE|?scbV;(sEvpFYx~iTz`QTZy3qD;kh7G zBK=%#x++zI2_*~?B43!nl{j5QX6lAtjPkQYKcHX1B^3+f|g|`x@t82gK7+)KbqP&qM5Rl20#aHFk=p$^uc4C$ExZM%y zrla|?S167~@FC6tqg0;qU+mRqW;iBP*&>j=5z__+)nq4%DGsYd z^<@|e{J1NJUGCfEKi&{QB=~Z*(s_)MnZ-T9Pqw8m5L`%zmkLc2uY=lTkLHA!D1tS# zzmMW$E2h4ERuhxK=Sx9sr*k*qdHI626;d13d+i%^JUiIset9~%L60h8a7%}8O7t*4 z)%G;zhnJ)$w6=_~gnU}rAw0(i4t7voqH}$Ow*k2!vvaD`ObNl~e?$0IE(;nNdUSZV z>0iNoV`iB4#Fwe;mL2-4E3t=5%Oco#d zP*0CI-RUhjyqq@NJ z+s#6jK|GMxnodq=RdOq$efT)FlS;ISgN53_oPzNDmmCwYU==rFQN@@=oq`EgMRAdp z9CtD^Zcu;MvV|7tGy!mIB+bi2o&X!hLmsKUhck;PUZ@qtVLYGi4g5Y+upT@4L_Bmj z`*=O-n*$R*=+sR7f)Gn+|Ee>mz|7X6=(`MLz4eb^^*ep;ND0ol*<%*0s917jfHz5d zgx&aZu}awo!*7cp>aQAW1w22|)WQy=|5Gd*NYesTYH&%eAx-EoONNK8f@uWaGUs@U zPP-z-yMa$9-J?~QA%_Tj#SCH%4wn*sKd70mPhA zddnpK{oR?RbMB23AtGtnI%5`%Z2X|a5&`Pr5;qr@hB52U$eaX>fqv_cn9WG}sF~@8 zmNo_pg`0Z~v8qs1TRRsOBL-#$<@cK~xx^s_pB zJY&J};nT~Dq}|c@NGfVp9MXLkqr4px!?V~hY2Dt? z^OTtm9!y9k6VCOIA_-^_ofTwyus;C?6YRt1K@<|6HqK2_=P_z811cHdR{MK(RU}>^ zDbnMakk!=fa=8;=r|0)+66F@F4xB10tw*Y-I5G^iB@cbU{44SP4zYXQO_cYTx3!-v z=5C9AO|SNuxX_H!0-c|+W8e{fQm>T)ggU2tWgv{1#~E0tVh)&vu^olIn^lQV9OHqd zo)Ytx45V|NhU#2->Oq>?5_#-B&G&-9di|=nrdF|Q%XaoRQdsVQdf+Z_JGdd-e;Ye@tzK z!+qI&H1U+D?C6{z7dz-_$&FhuaETq^ZExnsYG70(+5Wb+?o4tHRI)HMt8!b5yRyXQ zw$xuhy+I=5fV>`x9z zj#x!1wQxAoq?R7io!WEk#jkA927yrT*FO*Z!_odc?oUrE(+aGy_W?&9v3APtb626ljlObyEuj=>95Mz(m zd-gdl8rVM0hGgKtOvqWFTG8pSJ2eG4^#zaP82m5Fo%{!_=fe-%>{fA9p)joCoZ3M| zYgwEmFTgR94H#Xtpv8XMI5>^n;xfHCK^wALo#b&epayUZH^Bqg%DppR!=j_JzW360 zwTo71FhgE)Ij&**<>Ua);4C~Qt*jEyO$TnfE%F}H{r8fGakUku;g9U-^?}s%4Xk(I z8ONXLv32x~>CQi9@lswOcj@2bi5u|%w=NV=IVMT5x$L0&fj(pS{tFOdVfj>sJbvti z!TWhFthzrD0VXEaX$p4ol;)c7fQ}G*w-P_>exk#}4B07>zHnDHL2PaLTx40~dm$wU zfSOio%&>QjE_Wc)HG7*9&^{-+san&RD)Lq0rKRmfj3q_q%X)sZ33pTckmye@p`$4AI8qw^_; zb(bH!^0LYPlGOd$OebfZDTxfT7ELb#4BC!S)5;Xa41e_LuwtuPCcg^DR?}4~1PP+_ z)o7Z5edTX_GKC0NBJOi$4CCAs2jNYDX{(j9W>)$OqRP~rM|E~`8f7NGc>K<_-U|<@ z58owrcpuHr%a~7TY<53}PevZ!+FeI1u-NMw%>KL>4*ohJ>F^b&dHI|wQ7N#T3hxEH%y*ojsFZVyJSWL>@%)@2nPb?U&a)fSPHN41NRB94SQC{p|#4O{)9K^g8jX+nbL!W@Qu19ERfYhM+E_y z%0CAAKNayFf|w6rydgVeo>uG;ByJt#SQBSNOH+~0=Tn`o!^U(`$0Jxj_M^1flA*xv zsQQ$-zS*;llRMcI4{*}=vGqe-U>)&Qn)$;soe#1{O~~pmzq&Oti7R7PN?x`lWw5Yd zqxm^ujAa+Z>oc3B%sYgO#R`Ih5qsS3iuvlqHz$lTXEDa{|DsFY=|8+^)ZR-qh1Dw5 z@ffMt7*sK~M~|JRr1cq`IY5*LfjOdIu&UGuL8Q|3SutBNoy%A7b?~F9d)`_YVb>6K zwsqu_bzx`iq?4{Mrd8#B=GoN_i!oADfW7Rc3W5^R%xHkhv5PBfDz^&p$kTtmp?5eL zu@4kaC-nKLwzEQ7jc2K34@IFFbK2*_kCV%3%R$=eai@7IYbb128 zG6rGm-7A1rq>%)oJ3|RR*CWt3D&YKI7i7Ggl9y`&Ml%_NHneVd}XGjIS@-S!gW^gWX+61b?4WpbmM% zsJQtt4hWTNRmdwypF|v)o?-=`Sd{t9^ApW?faB@Ch4Ik+@a+sokqfoba@>Uoc z@HO5c*4pN(8sYPG&^{XsB(H6T8@^qQ`1!k#iD!26gkM#%#`aToB545Vlw#!BEUMkj zQa2>yBlKe2Y8<`)t>JVAw<=9Z&AisWlX^EazL&3KNcqnw0>*YWi+Go64qC5!K+yNRvgDqME`z+hn`aT#LqX+Z5N z#t*0c|DWl1UU)aDzS`4JwE3Q&pJs~+$l$z6s<~?Pl8muy#dWdvB5t&#>PUByF4)r( z6_j)-^-@dyv`RuPdwUtp3vKo>^f!+ znjO9Jd$cOah2p{{U&`Qg@f=O|tuv0rd_2SITN3{T-4~MBe}efB6a-dSlb1T8v-!zEdDqZ_&{=&%LSKWN*4!F8JN; zce#tmBX?dY=bR2auD7{nhvl~1@N3n9_44EfncT4;dByT!(Z+pwW9N1xO2;34v8DY{e531YWD-Q&9#80d(iuW~!k32R1{{sp5 z%Oipx)q0QYnYkZncqvvz=XCT$Ph`w>G`t6qgc64YsbO@U$;GkLE4MM!>+hc&~w-O}Cp{h+E@J*Ez+{^B~xcFda=BS`lD0RVt3d?@Wi0 za>KradKQS}=iApyh|K4^$D7A^oQ!c?He2GJZyDYzFTM2*Eg1#I$9Qnc)~~g8d{*_p z6y++BIA(L1-leEMz(e?*#Q~u1}o#LK%2$zYhwBL7*iD9l0htx z3QT>c*p4OpUx(@GGx*!Ipur~W_d<}(_WF~Tvimp05~BXk#nrd@Eb ze^286onDAbHQ@K=alMYkgS^)Ix6+(gdy5lti0w)=NJ!s+oTM5(GQyZnR{GU-yHeEk zN_$Fh1^0`nC)cx=o z@VcL%1dEPkYaxS^-0K?)J@~IKxIL@oxdCWKz~|=gBV;!`JM)cW(Ia$!eNFVtR#QhR zd)@symS`C8sjM73pK4v7ZDBLhE%sT4h;7pDlap+eB-FhyDX@&{@uU<+PynIOQ~wvnR(e7+Fl#@8{nWlVj%MnZs0^OWmhKKDW;MEZH8PRJ03^$0)JX*~Cw{Bzz zKY~cqbko5lX-)JU8TA?6>ht-AlpH`P$R6^bvpk-Vn|q0g!MxqE40lzLYL#cY8t8C! zp`D`L@RHChVaKpZw7+*KqUZDcV4;E;X2pZ4a5e~{SjBXFRr9LzJ^?=Gr@#PQ)~W{urVux0L1Gqd!!|V@p|)HZg|K=e3hvvSNJwtS zK(6$e_fA5Hh%s=XZvd!hPZ0H(`N%ezk|cn|%pb3^(8|2LHV%o|*;^+CFY;5DNI#&I zD^Ti*8DLWd;NaHu4j~cgWcoQtA!~@X(gnGa2bdk$lBT=glnbZL#~JH+#wW|A7LnI} zeyx7%ezVvS1k>Wh*rh+*{{~>NQ{mu(O_T}H%b>r@%FB9$ovr;h)=$muWC!ZjInN!W)hb)8ZGAou6Nz7f5lq>TDc|2(*zf6u zEtmNQJ^0D8ea7~_-{o_%HK`kQ1EbpMR?i-HiRX8-3)V$)v%j*{&zkY+1o*@)zUjNUWa zBWFp2(aN&;jE(-vPQ!O6=Q-b^S7;5&r>pUvkZdhyQ=C3|$DwVb8to_Oy!NgT+cpjr zVO$*!1_kUi>oZqZa{i!sW!&<>HsubX7j%6ql0{=m`qdV**YluNiFC=pj>7v`eP zNq?W~;TPW4wLBcT$&wCr$#bz457Y|Al&IejZW=kf!4eTi4X7&ctI{z=B6^|psxuGlUu@+xP<$>K`^FS4!a=P$h#dc+vtT&-<8clst? zaa~%e+39*W5BV;37*abS!RyJU_1X#GYkYmfh+Cm+)8c*q#jmO_(z!^ZRA=!&(wlk>i6<5`Uv}fZ?{XB3~ z-7t9aV1hcF(2Irce3F3BeedJ-0XtxONLWkGN{>NWLfh@M{GZ36iE5ZTv+10z73Vz4 zi~b1KPsfz+*X-)@JtpW9bYqnKk^vi~cfNu`W45M40VH}B>$^0k$B?^=e1Ga$Jen1z z`4wpVDZm8V+)Ice-z)TJSv0nc(0Y4(0}64GEDdw8vx_XCHDwF;k6ospJ7p{CA5)KK z!Zjv~&WP5J7?r%Cy`RqDn;z1+m5=+9p-w$yZ(hR)7&SNbBN&LRz5K8Gc_e5HKA?I?arC4tXev}(^iv zh`XlOn>@B>_SnvubRl`E`>Z~Vvxl(=5j;(BO8#|RsuTYupTo^lcVOTxSLb?7b+aP% z3Y$PpZ9}56kf~Hv^8z|CIU#Oj^btNfh=4UI*iR?B3PE}FMKu(nlBZB4m#H3(=Of&2 zN~iHRYRQyr9q^9PY<%ex%aBJI)km?+Ri~I(DEzEvc|7`6IRmbB47q&U@6`cmX7I1n zGSR2HCFl(e5H%-rWo^ZtjTWhQIb70#?gC>l=b-!h5{7=&v$AQ1mrYgtFVsKoI`a$d zwf9Kkq`WqfUS+>lKmKPEloMv&QIe@1w=!El1zh?G)8?TmtvWzp{!=2gFx7aUr%g5! zXzY|QYyIJtdcrj;Bjl_>#?N>e+kpr%bM+vOgZ4_da$RPUKuZ`PWUBxyFwp^W- z4EM#R=1!SM;eBt2GBdL`{OCi!XCYmGp*~+9jOY&M>4iqbJDlL4v%qt)YT*9qBmbTN8*3iGqziC37F2K+t4F@BD>HS? z@-)$kq=WTAZ=iN%zomi!gcY97BR(=+X#IT9X+jAnfAV^KiF*L!WC&psxzkzKWoPMg|rZVQ)4OQEYx0=x>Ii;t}C{r!8jBflldGI5DoH z8?22coYizFE+lxdIKVx;eUjYo8*2=Fsci96!m_DTp3&wq>^s((BvUVCyp*kco5K0O zX8#_Dv??t<-hijZMVP?^&_?6%)VKnkn&4D0iKM|7_N z*#ERF&#suNMyBa;=I!&;#LR1SB1Og$Wa1|~gY_>3WWNs^-#Pd+y*=PPkhKDW@z=HI_K>0>&Av5=!8iMY;T zRJa5p+D7jL3xoPpH)YC20J-yeAqEovTbJkW29`^>kyBZMm(BA2DesoEY!9&wczIm)VR~=#vrpjN1JaoTF-g z(A#IYwPNOtA3;$DPOWnR4F5A&-KpLcAfGqU=CzVzhQYmT8S0b2s?sQJUE(rBxRar8 zkFT->5m3nFxn$kRtUogKB>xIb{$T6b7~sTCNl2B|<=K)&i*7_j&>J$aYD$ZTcS+j* z@Ec=`h0aO`9owKtBi2|1`m3#qN7RmcTpp(?!>{K=DiLbaww3_f_yR%*tL;i_HYNF* zda^n4!@g5zK9(0GRwU8!*w)QWL9yu3G+;I>yy3vMxfM{#*G|7`XCO9%^Qrmu{{9Jt zKrdB9uF;gtw{w$YsS-TjF-d$x`wC8}UeIA?g_>x^(V?-v`Tf56;4VKhxA1JSp_R+| zvY2t)s3Ss-Ci*{a()N^@4qWZdO9yn{ho8p`<9X|7II-8z8QgWf9bBgFfUZKhPgtmu z*yZRam<>!+Iy&OodbeI+!Q&8l?^ShvFLNze*}uuaVV^NrgFmGuedY~CqLe6>^ooyVxb$VAXfn2X-U(#v z3YF^GQfU5HQg;9{d-sq6qh>Fp!SxCE1*LJ3D{+72i)t)RC;B(3*Jr$bMwMx&TxeIY zLnii(w$vhvfzNzovcwyXquEd3G6^n)AM2|4eysbxbgFY49#@vxYxjz7r{C0U$tDx9 zXqj9yj=Y>0Mm2xFKj!#jmDpdR#&3Oc5!qO>(Y@pW?aKg!8-9TRjPhPoI+PCGv9FOb zyn*zFRcNfveAny!&mUeylbNLW)gIHYomgv8vWb3uf8^3guggGC z{cUGyQ=v)07NiNRSng~|qnbWr81F`J3tk0nXc|`0%i$b>Amh#LeiE_Ki=A8Vcmln4puvMGBhw zZ<%Dd*BVr>zP$NVzeyPDi`La-pCmAQeZ%#GcR;74SV)(TlL%?mRS9)qczQlZ2Mz=!LV}FG84?05(BHY}=X0CzJ=2W$m*kYCgmB=a z@yrvvw-g0R;E+e!OKuK*ThytO_Kx%DAI8Qox!n%#1+i19cGJ$~lBH8V$hS{u7X|M9 zk<6YZaPY-~K$P^ukF_ZdIPK`)tOvmTZVP|v<_DL&(m^I~@iB~m4N%e|Vz>-Gj300! zu&4vFArF)WRwp8+IemWPI%ATwW}Sio9{|KDVx{|?rW_o`lo>oJ|8V3O za`?x=-{dfH+LMPL=j{*2Sbz~Iqf@Pu9ibGDTUF%VoN9i>7xAUc5ePo~U?V=056-!o zWnx5&WAJC8`o0g8Omf>qQGv#t90<1jtX);xS_WMeLf(Q{=`rtl?f&No%5-muh^H)i zIEnsO$hFGAV^qvYJTW*GY2DW&F(s31NurkFg$7K3{Q}~uC?RQ_bxvjFUCV17^4A;( z*g?S51f&EVX%xKsV#579Mi7W0L)OMH+Ha#NN0U)@+B_t4Rn^3qLIUK`qH01Bp2e!8 zA$zFyIVY!6V8*a1%)rLv_@29=a{$UVgga2H2} zskOyC$8;2TwKlxG+WeS7o=QjDHx~V;lP-mNB%Jr)VHk4fvCbI|326&XtWohFXhS7r zR0)kzbwcq3_}yRF%M;?co5BQzBLtA-He#hx%AtC)UmR^7{*!YKD+cdV#{pU0vlJ z^d*-)Q{sL+`+y*(X+OPiR1`hW=vBZ*RQ+EM)IVABECvvLUHe|BR$v@l!+@N+;N!;| z1bXrR==$ogs@iW`fi2ywbW2KiE8WeerMtUBLXeVfknY%YcXxM6cQ@Rve&0Fwch9+h zE}zE-SZls>jxpz$;}uqss?C$x9;OK!;B>oPh*LSR1*=CWjIY(drdIh({-`GwA^q}c zJXsR_DW_S2WSkwb5_BYn22=(Ofu;>zd=r-<+^H4A`l=1-0-id;uVZpy)}U!&|GIGw6OQyt`T!x=hT-50I2Yh;?@e_i>}N z+StDH3^IlLJ4G>p?C`yU8|kS=A2#OckXqTKACG$ zdA8qwHjn(2#MnxBaz7k#1=v(_F>_)LV2rinJFJv7(i4Ea`Xso6ekD(JfnTuZB4c#^ z3`}BWDt?l!y?>;DApgxNDZ>P7%%QQtt5fcD1q1%Rv;WK(1#tk_k#F~EY(}Q0tHriQ zFWv`7t}SNL+LXH3u~&VEhj>>MBC6$hUcWP{Q8c~BP_FA%8+1)OJ0|~~Vh0c0Dbs}v zF3UlmrE6N=LFpX%sc+Dd$qCAH0p@Kx4yDE%#hpX`XM*{gX5N53lk{ldg3GN7CK55u z9-F}fF;gG(ZBW2Aj;u3Q%{7w7+WdN;Ofo{Br)7jN#`+*l#mW65-p{>nCVIx3?b7zG8ZG+L!fx!b3yaRk33yf;QL zE)b@15A4%f8SBF9*->TEVE}n?D3rH@iaLPv6EX4q!|6y$Kr&EgS z4Y^Y^u~ANI&S?x_HBogyEtF>sEJ{n(pos&X711l6yNe)(Y1U&Dx2y^fAjd=sVDRP6 z8-cgGX2WH$K^I<``P-lUs~S4wkW0vW;%1shJ$Xg5;B1YsM$60twdnmJP{MjwjAZ7E zeXyTfRWDU83@?9`UO>vf%&hd6$~V?8+Sqw`>ztT}Dhg-^rW~cm^t}xfObcAQ;8EU^ zs5T!=Uz3RXN9E=Bg>?P)@dG!@r%};1YjE#u>1}u6H-~fzAXqc`7(}Y|Uydl&U_Bid zsA{hmc|(okE>>>_j3>gN(~#JTJsmWqHyUYBk_0m~A<>(=lfUCqGjHq3^2j@NLLS1! znGd;^+ZLU1GJ^ANyCGV^z^T`kT<(F3<_o**{{;h`DH*ZsrgF!Ii-`Dce=k}0++xIz zIU~0h?Dia*qQX`rUQd2|BXh4^)Y%{$&cE3{DNYIDW^DMBzc!ToRcMOAn>+ZPXWe>H z;VL~=Vc~;PbavS+d}8Jz#(KRt1_T;t_I?$Esdzc;UpB=U0obhk(C}V=R=83SeCy}Y z=;8}d`5K2_TeLV)Wri`FAmkV+yJo#D1Gm@%u{!eQ;-+&?w#smg)Ty+@r2fs&|#Ej{H z&sQH`$s06YL?cYXCTRYl5Ay_m^AJQ{J9u{7j-&0{TW$e)dH$GZk4S2{%G{qebC#?Y4XM=7XzTLgK2M@1*AjxZy!^F+P9(*# z97Gw}sC7n)KSETs>GdISd=N*gYTLim;nOtG%y7^cyI~Q2EejVl92zuE8a#Y*30{)l z3P}|wu9yQHJUvo_uTRjA`{n)l@9Tf#XGhKmc8*J|$#UI*E-Rs2kSKZvC2;NnGZZO= z>+9Pj(1l8BTnb=>6k*Q)wY9et6NFRK^j9Bf<<>&1>&hdF(ZgU1doBVb+OqXp_B?gk ziwnFhC$6x7ypUG6vQx9;E|Cjeq%f2->I6FrZ)4r#8B6#VP4#ozw<-(KTFdKHoer0x#2()c3Q{n5vRtR(IBfiL7_(};{KD8#zYnK zAG__7L@NWQHn~AZ2VhAgu&?e%6>xmSKrD!**in7xh6|)muo1N)UGTs8(l`=>Euml- z=UX@cPuS-MS<1_^Nh)8X*@hso%I6H|v{q9VH)(j#V1x7ehe(YiB!?FRSPm(hDr&Jn zRr9O~L&uK$t&tz8R2M{4G1R!-(4K{PNA*?dA@$Y*BgVgN4j!1!z*i*JM2#vX`B1r} zsB%nJHzbpZ0b{W158dolxQdJnAbPBah<@FxICqv*`O1F<*AD`6v)Gkyf@wYiGtX-J zCaNJoo-dt$HG6(&RE$I>OvOux_SKzZyQ)IpYurjn)egaAc)Dn}oEql+Gr@o9@pvG4 z87E1f6i+QsLDph20+gS4pK2_~zp-S60gc{>LrSYh;=Kc#mo<^J->S^$!_kpNlIGOr zH8$^hfY|-?XPp^Jj+Zq`n}5rm8NC5TDoXPw=u1~S0x?^2deyN1;Al(Y^&8H;yL-l# zi?~Du4gEqv8!rK&F+$-pQ^9u|NFRNAspbQ3$5Yj6Nc;3>QLJ|hJ7&0EBsxs{eM(azAfN$R+B&_ zc^K_G)^}1A?w8}~4P!EkU0FPKxc!!srCW=vCpfI3$nTe35p?%q?}{opZRTS_U})(~ODkFOo`{nG=Db-- zp0W9x2Ga9M%13KOm#6v59;D`_#`pY&33~s+BwP;2Ng_-a)o)ky)Rm2rY1oTlf-2PL z%Ccy5iO)hNL!DY!uy7Wb?{m)ns`D!7A?yG)c3(KD z1A$DH^#X1Fdg;P)+W==4n{OIe+`{LGtGZhQehQ1C*n) zm0iA8sBaAh?CRS6)-w~R^yH1JD`O}pEC>q=L1LQ@o49iq$QO2b z30}xE@bPUy5SqyDJ9ON>pnye(KE<(GA{75=)nybk)a9?PvlBUu^hGo8n-OI?Q|`hR zueOgO4hb>fYr6#MS(M-y(o)Xm^QWtAz07nM$Tp92rqO|~Ul#8OWZx%X_5>ud72+en zJvNGnaAt_Z`q`n`jyw0w>2P%0@Mr4f)y(5Aplw`=WB<`fW>5xF44%n3;5)5di70 zB4789*`GI?G_QKvbou6G4NHg76RC@?CRvQ!pK*}dzVlrTyxyor@;^r>4aeWJBE%0B zRx3t!m*7qY=h)zt%4ifZL=@uU=I7+9Z?(>VxN|TM`qq4C;LIyn%p$q@l@|g~Q6c}n z#y`Kp-v_^Qj3|Y)EfX3KUVxu{;a9)DjyFc5h}L=)OfF@j)}{N}l@|G@Wf-G~;9@sH zJ+fDlZvg_?8q2h*div2o51<}@bhfC~cxi|n91B?ZZJ%-e&?w;JV3G2E1bbv`gpVnc zZ*ixhoVUQl${>KdMKDUO=R6)8rvOm5KJ|U%_1Gutrmj^kG|=K*UhQn#>#g#`(8ml( zG`dezntk_n8Bh4sX_N|W4kh*#Z4vO~?JC;sUgpv>kOS z-lz?rz|%y$HTCCk?&!3x9LM3E*QelCx9nXmN0UxOEqc@@y z{onU^Lz0P0FiC7G^Q!NP>i_PUVS(zbsZLIT(q5r%OPDSZ8*3BD9v3z4KVGwNB;Rmn z_&D`hRUJMJ4$`uAaX+V=2NNN{B1XQyi;q+{)P3v z`x(#g6T#KY%MBF|Yn^gt!um1|Qv-lAa_1sd^rP0tsU??3TSBM(fM94o(ay^-1PRF3 zfzpk0$SdqeU&lK*E+ zeGhJ5?;k%uQo11pAWLT|(;*^iuwS*OsRUwd+S<;ZIg)|LAGp-+1go;E8H9eqS?gi) zY${#8?@~0<}% z=rZ|^_hv&4V_9i{xOxEX{cIQOse=rwiRFXIZ#cO%Ac+?2&&6cIQ;#o#qiA8sykR{h z_p%xIRCdNH>UXeT3M|f1i;jY(st4eDO0*<#AhpUOY4lZJb1HEcj+Yh(xCpe*VkvjOnF z9mR9vF;Qr3Kb8v8Ws@aWU^s6dTYnet6Lxs%R1`mDzi}xlbl|WEuq4k&F|&hg(W{_E zs16lrJ(&++4 z$Pr%~&K?5D%c~42R{TS>r@FT01iNo{+^ZjH|c+q;r|mtCK7^lJEK2Z@$mv;HinZ2L-V?*nbR5s!uwr} zjYza`MPR>(an*|sZC_i$-hmg{h?TzJ>8lbzZdb#)fmIWzpk&+C?GMEAPtpPJyUi-?Zut(Z_z2b5Zdc@N zToN2%-){DAT;>8YD*er3pm)D-eT6#DpY8A!g}wMG&loXb-uVT6bDS=?#we=BXfiup z>Qrcc(frdg{CVDIL2wNdHr!w5m}gG0m#~*@^HQpL0o;!1GdMR6Wl^S` zdvf4%1}jIC%+1K4B6%MTyGtL~0+wn>7?*p|08SO2{cEGTTDZ3_L*d`hBe!wign? zA�_fg<%I)~5uUG@`B0qt&6eIW32S?K+he&W8$u~Ys&oAUTdsLHI zR&6fGxD_-(ozpZ`vTp=6Kuj1D2@89FL_@We8Qa}>w?2u7f@86iGaKn#@Tlrs9>KzG6(3!qoRrnpycJdQG5P^ znMJCNdj0p^eJ5s=tRK8=b`0nZIW~-!^K~kbl^h=AR+W+Y#oS@& zC{J!Yd76zmV|L6dq3i$n3fC<0jF&}bZ7&8}{q3)^w*q?;A6uK) z?7fn>xBxU1^&ox>@|NA&@arHEpL$xjyn94@$cebcDj#O;WzC40BrFe=3u-3X6HPvr zxR5o0gNYX<`4dOp5l@nqvfmZ%|3WDL^^h;7cLw)z6+tV84$1FFl?Ft5KZ=c?rEfCZ zUsLmDcT9_>%cnY`Ii1gM@dHzF!|i&TaWNaQe5sm+vOf=PAN=P$ta^}W^BIP3HxdUaJZFG_T~gK=0l8V@#^8hxWyK5Pl#6 zQWXw)S`L19=pa*1$%t_Z_mcJz6}Rx&;-O7PrHhL#_+pTSz0CDC{$cpo&PP)OU1Ufm>BC`GiT4 za=RV}bcZy3*pR<*hTjy$IuEEw0;31IyMzIK4Sa0*J{#Cgm6bFj^R%D0eb9orrJcb0 z{Vo51(X0*!tw7LWl6xoMg)6wDNW2%$G~Di5mB zr;s&SFOVoW**?z;Q!{xc`5m7_AoM>*TU3YshO4%%uobvu*O$^o@lt$D8RdL_lgI@;fdQ-lOnHKYTvau$xk)Cw(%j-K6L-^;2BF?I;0ue#je zKI{79CI8fKY^Wj19ZUiKXJl|j0)9mQ+^uw0%pb=1kg)!zWE{G$f9Cl4E6A)DDkC(| zzrHF$FE$WZV;*;biYGSVe^~Ucu}t)GvV26cW zNsEp^%Y9b<>1vYzv0;nzJ+pRWz2tpt%2Ae!`*);45GXc-_JIiE0&=xwW-o-3%w#W~c>bdWIu?ew8id*W$u)fk9>F%I7?-)Zm5LPmL2k zVbd=L_Qabpu(0*epG)Wr!48l6_-$RCt4xKFmfc6AgTneKj~O>-jN|rq$p%zfP4DW} zGq8yG*II6uI8Q&(T!Uz$@YQ~^BVxcP5`?V2Qu)k>UllqDWAa2E9d5Q_Uvlf5)P zWuRN*j8s@+6Ax4bh6Mk^McVlRjKpJ?>7KRMqFl8npcFIh?zU00Eh$;SL}z28FSd&f zP8wViEq&l*a*;VD30W6to|yA7{oN`Jx3hbX>wZJ3-hT2{gmqVS^pS`#I9&Hh)~S$X zy`BrbbuBLt`S{3l^8SP|h%8E%M*dupn!&fmOn&PzW!f!UXccoMyY~&HCuu&7BhqL| z#*-xR$<_gDt#k;mZg2J!#B=kdvogc={1TVQ0cx~qPA%Z;qt{gK8cC%Ci07fc{gI%C(p{7G*3}@e=t~(6+&kFOS=G5|_ z7VwQEVeQ6$QerCtI~!8aX=?-Xk|cX{t;vSgwrY?LZX)90>^xi&PWWAMZFCvp_7+m+ z%*u3bZ+%o>lPIuAC)W=-gNxkw5P9A|GvQjGIiObUq<)Ud+e!tRlYX}5ZALJexh5ls z(FSQD0L9a~AvhKlMo$z+`O`meEzU_aFNO=sFJuuMxYQO-PExNOcLpkR(7`7kDi&O$Q5VWTYAkLgCXyNydhwBY-^ zTXROMe%gbr;{o6U&WQo($77KxyzcRWL=LiB{>9NJBSVWDl0x`Zo+05WhclYT0z<-S zaw%OR@K@&F!~>=!o)ZV_mkGMIu${clCUNr#zNel^oqaor8j{L;ALYD9Z&woNFd<t)EQ8k_7~tc5%_6`*`BNDUryhtL$fIBxkF==ri%J zK0Nb3uV>7fd5nbY4s?SjA+hVD50;ywi=Pxv>$=mhw(gWQ%IN#3VopEKrN~i&BGovr z035_4(3%@rPFkZSP#i7Iy4Hz5{;kx1UeJF(FpCLxmn88}OsNq&o2aPlWpF|13+%qGmyBmC@ouM@_q=A(bzY@x@`?Z7;5 zJiIC7(MO8mKUj4^RaE3sV5+ISx6uTruZnFi>roM+a=K7G>$1y=c8tfrbz z@SV=F!p7s)@@)!`v#-_EQC#qtXR~(H89?a+eN=n>b(|exI0x~cP`C#$l$LcW(3<7B zo|%T@^SDL8c9(QK<5Eq$lL!5La1j6Awb04+zBFFIbK&<*h@DtTt>QKZCFbKv<` z{ZMqz^8(r8Mnh`q2gQGq(QW=kf)Q?VhMDk#3K`)A&7bh5hwPy>N`4(=R^e9+M;UmY zIOcFJJYSKis-@_#K{OVo+=)kMZbXTRWhrjuj*`kYo`WOzCbrmO$@Wk+QZw=o)foE9%J&V@Z>rfqg3cti8QTV$nztlJ0mb`p zV#tUp_22LOB4ZRA1eG;HB8zT#Vy7Z!#ZLwI5>IamqwKxIom2aowNNtl5uJBz*|lCQ zvMd{%!*;$iY%h>tEO0nsTO)HBi8>|csiB-?8jof5-FX09WN~&MIFkZOWt@#<%2;6Z zWh|t)t-Y9N=ko#}N5ZJ_+KuY!oh8%*kpnCThMGDOZY5 z9eqbJ3ghZAeY?yl7jh&8u5i+5u;9W*(DI@j-j zicUrkfH_Eo=Cf?}2c6=TpvQxlfY;>2v0YwxXGI!J7%<1Td>gPx8<*V#I>7DsKDq9mgKh7QcJizy0)lN#kXCYeU+Ce}jDo(`JzaZ|SPm}4#q4sL#dRqlP%QmT zTkZFhz8$WsAI=lPY7w3!j^y;l-LT3^z}Lv1 znmV`@dD(q4&(O`$qA-2@MRw@oUj3LRbMRW|!TmNm`+GRJ{fG5%rAtp%D0@acX{~;h zLS#9$)Monq7`G~WicFq*d(R*%F)nhEgV)aFEsV^JX)ZRbhBmplhD8L65NjCl)0}ym zof+?1fMk&jO}+mQ_M-Kcr;oog=6~0_e;*9}+U0M(u|vs?64@asz?vDB>}w||s8ndt zpMqzmLgO)r+ylD^^P&uEB|5Zpy!0IkZ1z&-E{%ras3Pns^voNLLmekl*Y3w~CaOBM zjaAZYu-at7;N5TfgBx@zF&&e(dWTbpWAtH!5W{0>c0H4g6SUX%MrF1Ue&o|3oqOjJ=UiG*1n0@z~cpugH8 z!JL#xp!-gh1boaha?y_+NwZ}tUNeeooq40n5dir(#oK02_d9b4s={hDIHE!M{(#U% zA?!$FlaBh~O^KL}1PwzneV;^?@B0nQA-)7nm^#6o%IDiEz~@#tckb`cgn z)WF>&+f!){#<);7cjlD$<Yc(#DQcu0yBqY-rD zl*Z0p_f0ngo2Ln>(d8AwN%-8NQ2pnw^>XID`?Fa}`hk7^&nQh;4V6poRDi5dalC&@n; z>eHVWLc_H+PHGhA+L5)GIo8Fw7OZ0n;R)P+h{Q)sbqLO2bJ?QSrm_R#%~aMqoUJ~w zE{ll~w%H%)5w$_Q@>n?&)@1+>1UOEi|8)l31c?F^H;#|zvO>e{37;_Cwp;( z3I<4&+CDM&>wJYyd3VhKJDW(Q+w%(jyUnM9Ud(^J0HoTGvl26ILt8oONApBo!cXO^ zsQ$(oC%GR@I+HRm$yQU?fozoAPA*umQL#D`RVZh7ep!*JV?^WAu z%Xb6j;Kst={Gb=*`zjdq)Ju-QWb?J34alG*-HwA5Gy(};2UAL}{*>3|iK){;1&9#w z6-gOzX!beI1s53`t;mk+gc6sMPqO{JHz6Hu!yU3KQgY%dIf>v9#1D8wg%p9_d|_>W zXGiEA#NT^jr1a14N(%ATzac5Xy!xyLi$;sC$76%MIoxL<=jNl!IWD*MZ8_6dxYUA? z-Tf_HtsH^ZUs_rZ*Ny{O?WTAehlQSVXHM+_ih?p^N^|IdO8@%Ohvgldnk3;gl?^oe7?WP@i_dt6UJ zYD;9cPN`yLq9HLs+0x)7x(e7E?xq*JD(2?l{IP}dWVuPM$<{Y~^I?XTl_PCu zqnY7_slZpruKsXMl)@UmJW)NLa=DlMdEA{-Kn7WGPo7(xWo%c(kl(i0k%5^rp`c?0 z8qxHS!I}Dp9~4KRw^OW57Xx5TVg*#WaX#SX%wC4>=O zyIhB(KhuwYz~Lz4R-)qtiR!E3kUGGW^_ie)QPvURFX6Cl@yb~0;lG)qLUwGGsp8}) z3zq)d*MRe*O;fy4z#2`|Dy9vauS?c=i;4fPA&rpg= z_)QyqYKsZy(|3N8%O^IiSh}{`9{&w>ix)H1jY019;ZUPO;}tqJkwNw^fV)L07+sJ3 zv!@rR>TKyGT;jAhlp_I4dlcYUP9%hU0XZCaC~iw*OqO%=;rLD~gE1hu)x^j0>C&f0 zI2Ui=4o3FP^1}u@+pAMdYVC59(F=LdTdhX*oH1{&z_RsJdrt92htp@|=O@jSjrSMK z)BqZhI2~{Hd+D1Ah1ZnZKH{=e%r|6%NlMBZrEm3Bb&kHzfO%=O;%G@~_#i0aRKB#I z?=*yk8uV}R{f}@njPwlwf?Q0x@p1AOCore&We7 zKh@7MVt!M$hZ}o(L;MSmsh$lXP@sp|6;+Amnb1r$K?^~;SzFiBJ&~>hxjW|}!`<){ zI%9;eEb`c2QsyW{Bt>?Nl2(iTq^J^i&J6g%)Z21ed5qHLM^{okA!MwVECb}pI+^U@ zZHo3xh&WD!nU4-0of97)+2uL(Uwp_)!zLdllVDh6Dt{N<@nBYEu4XmttDw zdvWl#YC!AW%||>iB1EcUAIlS9WcV(PpV%GMv`lM^L2mDBi_^_Pm8u2{;6)X zV724TFZqFhq9vhoE|`81hSV1pr8{hmDVrk!$GQ6}~9!@>1?ETzpkQ||{S ztdkO!ID5{a?BK|`r{}!6Z+S>vnJ6dAS-lmsv=K1_>(J|xo4Cc*?8Laz_(px&JG1Pm zRw?1Gb+WP)I7k!5S%?#6q(xSI?t*;v>x7hAA!n>DGRY~NmbB~08ft{i~YF*i5uObB|%V8cEWI1(EJL-tc%_iUp?+@MTm1%D0A z<>-1l8wcWEPDAZdqC0|$3o#n2)7`lu+yr%F!z`SM)dhzuerFyqOI7b}*#>*}b6(v0#jJz3Yd$2G)`ABcjK5U=wiQIt1tlq}MIu}Y*VhOH zFy{M542{KoGi1q1X~NXmo)AIznwY$%j2eU^F=`}?$dIz`f;74wsEYYr2y2b0;|GYz z3X?`$rDpME3mze=zTVq@`wG77r;jhRcRFw7}E?KS$+cZLP*j=O_wr^|qRybk+>*vBr z{dB(v$wWTv?_<7JBJ2sgTw^wYs+dF;lj1U{=aTWZxO~Jz@_s>yZMg=kG6#{TyUc%c z465m}rak=s4T?ks-OFT=YVb`zLcT>*R#ic2nDC#V@PZmG`IIe17ebtEX~tUt#EUcb zDu?QSVj+?L@XHHES!U{K>vG)z1G!1zEYqIM(LAkqra+61f5yti2&~ZlR!qag`h3aJ z_UZLLLpT5QIpja@Xlel}ah?Cf)n*_zI1ucfh#SeBcu#a%m(?T#GQW5qe2HMU}FopN)V zYTB9_*;)QtZk%gH=2>$w61nG7Pit%;VcPoQ;_QD2l8+V1E>XKq12RoO8(`Pj15Qc? zc4-qmE?yr=<@sM>b!}6=1U*i^5p(GeoayQ}$C*p}H#maZXND4r;UW%PYeQ1Osx;Z37>Jt$t1-aL?$*&)Eov;Kek zQvZ=T{fz~giEa^DyUH)wj}~r zO{2a!pa()R6{id-RYG^1N{^$C46$ZQaVbTr@{H4l@zo%Q5d>U#aMp-mqmR*dFIhy$ z7CG4KuiOd3q}nFiLV$-_v#TUa9ynx8&?(hepGXLVXLaU?QGRxsAvwQgYwrju&J8*Q zt7{(ZKE%69!fBHl(|&zh0~{Xc^V2-x>=H== zlD}^V1w+NKj%|A(YjcKg^4>L$f7v=NG}@gN*Eew99~_N7lJ78wVH} z2A(DPzyj~&r)~8=1v>z}-q8;0_BoexAI(Os+v|Ck|CvU+okd8aB6*dH2b7nTt3Bm|aF$2ba%;PV(rf`x+^ z;$bdqa$)`#WqwGiPcaHlqH9=awdbF(oYf{&EIpaDd&e)cSrcoZiBJ6Y~t zyZb4E0RS!nuJ8Tuf4r&+aLHZj0<)gT}Wy+LK=K35*=aXuZj}VZg)4%21^@S_@v#qLdcMRFN&@cUevh3DxW^-N{%SR9(2)?ns zfsyAuf*N024BJAM(LyWik6lcJw}>I{F_S4Q29@Jsl}Qw3c(Q7Fz?rCK!n`An4UM_V zLlI*4nz*vDnR9*AJqkSEXcE3etulTtsvLyp0+Z?E%FyFq&<_nQ$lT57(M82^{Zkj! z_Sam=x((83C6LPwmv>JDl*_F$l0%ei_^12p)3fEed5e>h%_;s?(ed`y;W()ifG|oY z&Z^Pm6xLnKNIH>>*@A3=tDYz_wM73UhVBwf}Be9&Ax(hFurD zGo<(fHqF3bb06I-`GBY@@guKF{~&EqktCg|(cHTT<6+#THWT|gwP7BIld6`$?)nkq{)zioTj3iY-n17R(SjteH8RNm~!7G9>IF__Rsll)Z&pv?8Z%eU?q} zekcteD4mXtYgSa;#~@1nX0g&$pT=P$14>lMjniP94eqJI&Wsr2)3 z5#~59iqPS(aJuBs`>&-PyWuNeYR11@XTCK%$vk(*2p`Tuo}e>TR@cg59%U))@Mc3y zRrJ7SrZpB+_a$B6p**&HZ+M)nd=dnuey{vBKj1K3={>5gwM>micJQIloX0eW$1kDb zgAId{+XvKp_I^HJmg2W+s#~&?6WKT_i=RoMMpiP1Y6KLvJCLbeNSyoH2uA~;qM4ea zB53QM7(V~R_dGUI`93gONmu);qZQopb1FGT4MY$sTaj-4)Yr3X_PAlcu;FIx9sS|n zY{|%~!KRV*r~KxjW6c3qr+i(uXK3suT^uN92ZL{lZHc zw*jB(8+;pPcGBNL9x7xKh5T1o`xA@bB#884lfPtCQ)bc4EBt%#^KiU2JT|Hx$>Ht` z_O7k!rODNJ+e1^v$t2#|68dKhLEjk1QU=qj&Tzpf%T2cnuo8BqPd(TS5o?Q!)I_yR zfC@$1;(J#b)@{Cdbgb4H*LlWjAF**yd%E8gCfg;_a4g5)^9Ux3|Ir|CPr?Y%0E{*h zEBfwJ%KIm(fVHCpte%zY6J3H-*rgZ6M#AHm&du0iq{eBn#kuT5`m@a$?})gr3RX_m zmo-{>(A(^lD`j(sg{M29d!7&am(GYLdi-V(wK!I+Bh^`#x-0Ko6Bx=SwOaTqbX3>W z$06EZ_(AZ)glQjO4?AN~@Le zs;^2$MRF+EYhLR4&{YHW=bGl&(?5aOKlQJh%-8oyS#%EEAvWE1h|%RsoPRdcng>oJ zq=__4&5aa*3Qm1PtCDb$njt!DOsM4=uKBIbYt)py{K{82f5BXu*{0W~u1IgdT5N5N zzZFB;ZN+Wu<+qv|rr&MkB{A-4W`fUWQ8S^A`pu>EgyHDu8OA3&dA;@)O!g)N1K@xS zMc3|Df>(nlB_LPtkYx=j%wSB-znM=q^vxoEi}-NpjfAtU?-P1B?h_RpT6$30k~J9< zZ6uDN(seZpEp|-Ts|2hc6GliG&ni6!{qB_c>{X@hGneU5VuW7D!K#&LLI>xibeZ(~OPZFO8l#6*ANQk=*48Oky)JbH>>2qeKAz0I<%n|@$*XGP zVq+&IHfwO*TV7XWZTH@hWehlZZ~?k=S`1>vSzr&A^&INMQYUxo>m;U%^&s-^&&+tO z@iQ!ur^ADRit@%N1}I!VJ@jA0|M!+Vco-1WzCaVg93TX5KMRmNDoPrRn4Jp;Uzhtf zG&(A72ct?kr8|P zP@ho*Iu#X&Y05!g#C3RQaBhow(?e^h+lby0XgikUnSaV-+{==0*%`oRI=Y(UNwVii z#tFOYz+i+7?BAz+S&(y3BsqXe_N9VwIIJpN@jEP&tzUW(;Eta<`t3=IA-Mw^PXH4nqm4MNYN!- zH~6sFAPv-R;5%EcniR@o>!e>t2dq;FfiVRyaccY`3Fo_k8i)RYLM66#Y{V6*vuAHJ zwF!`G2TX8_1nZBXNp-O^POy&r%}=2t*h7o%^k>5Jw!pU~_OFm#@}uyO)6f#oaN~bHappHK zL*@KVht1D?6+$f1yq2w!Y;IlX4_a2ZOZ5qlI6rte5_IELv^zltHrUwxP# z4cOI@`ea!CFj_KT+`4BA592`zFg!4XU0>jzF=^+%5;9R84zYx7fP=CURzfQ{Og||e z7HXZ!+hmTCgomI+y(-=eh#WuLI-)u{TO0Tn97_P~d+VrJhTZv0^^a$ zlV7eq3V*FEFucr5?xx_z7PBfL{@7{~f2m!4ADI|;Tly>D`62fB#C0%@c?woE>$^!^ z)t?Qs|9}I39`rl`pV2t^bB5Yqj>bex^Y6t#o2|#T&A1x+8vAG`<9(?cm&e`*cPypjWf|6a~-`PZCsERNq82& zDbm)IaD*;?1FSC#E%gkvq~Aw86zVrWvW7Y2-}b>Suk|L?a zP&u6tc1T;6ZErO+@L3_OAz^2mATc}@*5l0+DkmJ=s11k^TD!&~-JuJD=_m>uL$_$j zNb!h#SqXoIq0tRiw^(2fSo?%hK1%Iu4?S9f4E>q(l5P|!CDGj#ltDaS-WM;lqapc} z3bF2oU~%`16CZ5D!VCMA0ve9L;%eXW|0p4ex^+qYQr{oIU3zD_W99@*2vD}9 zaYN)=m|JdDY8Jnq%IR^QpV+>6C%JXBi)u)|2<-GIT=k9Yq5E8I8f1kpjtW7;?qHmE ztY&BWItv(g(Mo>bR7ZCuOEe(PxyE%G82 zaTa%o_3r_(P$ji!wPOAR5~YySsaQx*X|$S_1uESn+)s4B9&srn%e=e3Ok=GyH_ER| zQtJpbEE4oNH3QI+LIVuG(n|fpjxwj2{33&ICfrw}fYHzbOCN~HEmLk6YYq#UyEI>? z#HdUJT4$6E_1l(A@7%p{=NEc&+`mkWHl?g~c7&NW@KmurOfx zevGO91?@}Fd4?44K83?hxHfX?^;fdzS-4ri^O=}OOA>qOYHv+^I8CZ&MWZd92dbsI zjGnd+*Wm}P%hvrLhnF83*h4xzii0S(K2;>_$7R4wT;I&*t_#BO{AJcYkx=>OrQFeZ z&OfDgqGJda3xKg~o310S{@c8ht#IvZqrJ=>p{Ap;8FzDeEu2K|#@^u!?a(QXd$5nB0!OEZ96%#^LlEfPJz}lHOji**&8JcANuP#In zokovt8pN};-0$z(i+ic}^0S3PLX!{m(JGHshyvq!*VVnW$|?3w~>|U zW8;keOSs-_*Hj+|`{s6g{MTexpYvY16wJE*quFwXk;Kg&XDSq30$YNt8Ilnq48#i8gauDb8f-ESVptPvDksmn%27T- zUm1Mdox7jnc>RkBvpC8QtMitK7!H{RH$mYrYzdQozLO8Hkom@G9E|dlXivaR{)|(Y zz8vq;{2Y9`dHW2KXFr8AWr|${1%I+M*V5|ahdW8^sjY5az004oU{^K4mg^B4m#6&2 zp`Q9tKV_{Es+0i}r)4$}i5xnSif=&=E@63@8v(0_Ytf}>vGLx;+)R`XF*J86IL6hd zR1pyktl6EtJn$%IJ>S@q9qt~J#b|7@qK@)Y5u06OAPf6J2yDoGmwTkZO; z-%FI*LZx^u-)SQ&c$x7yx3`QCrNia-)16v>8Bayz&193-&~do7r=B}=Gj<5?v#}=6 z_nM(}=<*c!pW4fek72;4?TKfoL>IA7Zu4%vHkV;oW+DF!qD-KyJr;h(?qA-R*3L&| zeqqy4qmBC46G|fmXC88ik0&_;4b?N;U%_35CngWnRy-4+gVSm!+%1ub@0zu_8TkY< zG4nJE<-wA$#Ajap&=_WBRHC{BlGb(Kx+I3qrDoA5KsP*)B}BA)wnzDx7K$p}KhU>H zda9)45bWz8Aa%R7itk_8+ogsv1oERkmmO96=t18bk%8^}r+;s!bU}FADQg<#A997x z{a*Gp^Xa5(ITmqgh4UzBbM;!HJcYgrlZJ{#RSndT0zG|!+R!VQPOnOQ=FG&ft`t|! z3Z_)mvc>@*Q^a>QO3+gGCz*}f=nugHz>5JcTL%jnY2|;d_-iCcJiR70n)5h7nG~w5 zz+gQVjQsTt+(H9R*V9NF7n(S)|^=+?mn8gM0rW75H`nZiX zI?p{XJ-_^B$*-srMJ6cNwRDd=0M@Bd#C5jEx9j;)4tGcRT+~2O1L)DHbPdlZ-|4lD z{0Z}LnmM9B0M7awQK>%~x+Hc~#i`a0G=nKvlPnB8y7pZ5v-S_9dp$Y0wRkD>Sz}-8 z6<%?I_v>jEH$7+OS#`DuI&0`l;gxfsKV?}{L4pZVtkt;^D&!zLwdTS;frLttAjwf$ zXd{XFy!&RM?8u9$um`%1gq@OTvmztGY_LDu%Kb1Ws!*szX)lB%_My{q4Q2-zW|}th zDGgOD_-#=cULBmHgXXw0zOgKJOAT7mx;-cuAzK$MSB6@8ctiMV3gGg%0G!V{cwb#z(nE0u> zrWd?b!aA}BihsIR2B?Dix8uLhq?vtyo)f_jFhq^@+Z*t-Q6SeAETXo=}b}mE4yjYYgZsQtpep3nu&>^ z$Sy!?ek!1Z@I+BAPi@8y2BZ4lt|=XS(cx%CdxOXgz{T-S$XBSnK9kqK=GQkse zbgf}IGkQ8?HnfN0bjZ|@p@|OSJDurwyyzyU*yKsy9G3w)YnZjUzM$CIUrcR8eX~$!~tWs3bY^aP1u zhVI`|(KpYk>1~+iJBC|ta;T`wP5|Z=v|(>+nh&~MlpcA1#Eh)rE*+fb8cS+cgr61C z@E}x_|2a9IsIsz%{r3Cu7`=TGeraMXN}osDy<3L6b<~PtKX+5sNMtp{6l!+1i8FgU z2{A3<>A7~kdQ3(B#BM23NJ};~6J1--;BQbrOe`()2b=q#&>&wPD4y(Of z0x>yYrBY?Cx&+Hyy)864ztPoeru>6z|EGojpEI;Z3Mfn59JSGuhyB%4Z}FgmtDIXw~-?k0bPE zoc*Fm2McHukX#;`^~9s!t=%o|)}Fg+0L%d&l5;(KG68uyt(-Pq%6ymr%4!6hvoND0 za$v7dW}$p;h!Hw+5)p?40VLm|LjBYASlwo#>PuYcAJ>6-8UpXG#AE0NvctR}5N0XZ znHe<5eM&O8*<>>45N)>MF20K6er3b4i*H_k==Y@7?k&hT|Qa(+6Ll|x@v{3LWhkD)w zZTOQvQ~>NW(FU}MuG8gO27C#$B%l?_#6Y;odret${R`45|Ck#@H=4sJQ_Vtw%+p>R zvyDr6q_@PAcGXHC$xUM`m*5X)hCwk}DZ_N;M>#2b?tb>!oo7f{$~5o44X1)XT|0w2 zLk&KBa5(!1#fwsGC_GB_e0E+3?WmTnp39(xPwD7G$|3M|IzuHB58rw1OnY*-?)KXM z0tH|mJdI^9jT;}PU~v@>N;wIpsed0>zkO;0vEV>&NuA-DE0des)E|5y`W8UIN!@B4 z;vfpu@Hdlhx%2sh(Qyp{?e~;7!Cj*8F3*qN_i?s1C@ZI2+zqcU5zrci3r z$PVt*AT?Sq*dT@k$9*7n+q5Rru99%}tA?t!dD2_(QK7GQsg}n?eB*qXwR@9-?h6zxp07&msun46Zceni3#p+A{pNY7EE$N1%`8gqI&@ zB8L?~wuYZjB|6L^-}C_ufiYXRbK1@DGh!iyjt9eFS=o5O3%Q~!*s=&1qzM@7Zye!u z@|O%mk?Zd&TT32%u5V~8TC`QUg)}O~OVfshWlsiF*N}2dU?6;>pIcd@C|?&j7+&kB zA73Sm=N}f+4yv{K#^h+yk1}kM&D2+YdjBVImOLCO!J9?SWivh6Gkdnm%H-kG5g=9{ zKHPRI+9(57|nx6X|$ZH`~l_Aq5y|TrouHq>W%Z9PRZ1DAZLk--t{vApO}}L-dYIY z27(dD_f7Qls;`VEnkExDX_#qbIMLl8)GZc!v$QLI%_4R(YrpYf8obP25g4QCrI6=~W$(|De_G zPgmfJ12sm;ZfNM(DFGhOZ*3RbxONnOv4k7}ICiCTkJ0NzQhItF2}W|0mvq2C`WP{w z5!5xk#9!T-~?oGw0npAeEi{ZRQ8Ko>*8G=Q(Rm$L-~vx+7emg%$2CtfnM(?8f`us>@Y#^wf3z=yzMkrsZ}}@d}i~ zdIRI3TZ8kgGN(f_RjX4=dcRIMWaOn~ zEgd0a`?;eBcLOL=MaEt@a{s<7MOQ%7t%`!#5sn0gH)KiKD2T}U7JXSf0l zAUmj-eD<+)-VVcWfgg{Xq~;Qms^0SP3HnIlZLnMWbj?eJ?IxtofV#!rPEZpTu)8~2 z5r`tX9zf1&ur8P;f$A4@W%~(a)6*_3}h2lk164rP#}2AA8I zKLbP`4TlA1X-(ntUmUDU*JL5mCMA5;A{pCLy8J8?8j7%#zT4{0oBEKA=U1F)a8^7^ z#Ro_g-pGE&i8;bbzZyh|&y?sQNREKqX&o{Jw{K!FU>V0S>dAZw=n{Bmkons!TZekw z@KrH(^$bf)N<1VbxFcB19>hP%!{-y{VvnUbLQ*iUB&I0Y-c>L8wM&WN!k7-u#G41Y zcnZb4oG%%2fhp7i84w4S(2J$A)q;1R zi|=ang?ZT_S-4$R-p*l`8N6Q8%_p^yZcFq3X375n%;Zo4OzLRl6e#oa;zU8_P2c>roMes}%q6}g zSb=2jxPWejjO{VL&j-#;@i%vT3BV5gemzj#AOUJld*m68J0Wim_vIILx}}Z{NO~oh z$gi$r^LhzyghCE};^yX|nPpNV6_6h>I`5iBm?(RJMij4a*yFFb_99YnsMZ)D8Q5*( z2MjuiffqpxYF8qXu6>5fz2z4)9MhR?Rby{)tj9KdgkpSuqTYTG_=a_3DhYW;3Os4y zpIew_=?o>`>l!#XL`KARZYSE?!sa9J{0WA&o|_6Sg*ga`lSfT5Rc{g18tOh$4Xa!lO`?nipIXvyz*lZ3;iH(*ZXd zu*42=g(T^8h1a~1VM(}@9;feqm1>(U z*Y6ZP_v5@Z^;x0t%{0bT4aLpeb;}zM8Zoro?&fm)+v&=WT$>kaex(PHUPFRkXqSFd z;nE`MA;`MFo0CO(P5KazIe8o24Pb<`8UV&F)QJ8UK?lf?0p3RP)_D zgW*mB2Z(136dfIzvJZfBHru$6`0ZsA! zeJX)XUmInB=LnboDisyzyQWZ2YgW>?!yIujrlTL_ehSQRM*gv|(;t(Jg?gV6yn8T+ z46+XXKcf17epsd=QM_0d5>{1jK@W=jI+CKkC||>R*;Athg@wd5{6&e%cwxAWp=5rU z=j1!9p+EkNZ49FyTd8wS{zRo7d3^&eaqiSA={T;*y(r=NPi)z}9-}SW4$Sr&!}?`P zmC-woj1>pfi7vMnD8#<_M0x8XUYDt;i!5bfLJb3Qey0FetQGCeoc(iquLO+Dh6cIg zw`zGCo0GY&>Uy+t!BVB8Ylxck3FP?h;;|UrjIe{|I>-k}(6UzvgDboP=O2W4^BJNN zup=YcN=v~Q5Mx4Vp6Lo~G*Qr1C>8K-ccz*P9qy@#37aY%gx-a%O}83`FFb-DNrr__ zvPnlXp~C1F%|9=E%U26gVDRxkwfte*&Kn`87KQaTo_<=VBD2tkYJ_}9GM1T>6dY@` zSA?VO;I>jJ>~GkTlC0VSTc>s`x`*dF6+LQ1lp^c}tK8EL6S^k;v!EyC5u*#SpRaEG zxVOBw~`9vCXfiBq8u z(`25P0^XWI1~)~rUJRV56&M#{Q+XH6kf6wNV z5g_tUuyh_%)l0lRxGI<*B3(CZrzZS#$#=Hm!oqj3%(m6w)yz4)o@xdAnrxVYk?t?s0)T;G9H7vXWh6X|GHkeE+V+(jT)2q3i zB@!x^2e28Nf+vrQ?_Et||3ADcbhA{74=rG$;BJwdI*wqVsAYOp;6b}w9u_S1J;#Lt z@cPQ0yn*l9mE8$g2P(!K0TzZn2^yL0pCev->CBoiaY#<_HE9wRex<{q&u=pX`2-p- zJjrqwwZwCXx&4<9qi4NIXg(2D&Rx7kG+&7iDWxXA` zw{*~#o^(WCBnkJg-X3#qcWw>M@5>ZV4%Q+X;O9&XGy}}-UN z0I6j7GZB^-;9{gZ0i#ELa|df$>$H$YyaMK6t zSjX^n`FN`TVWZ&bOgDg?o*im&%P9M@YAA7$omdJBZ^JnHp@@q8Tes~&Bl7XvmPfi~ z2bh*riX#udt6%rU)69M4jo>df*V?OQ&mXYKVeBV!`tT~>NLj9+cyu4@N0WBM`Sjsv zi3Pk)sTHFz@Z5xa=BwR)3c=9)3!g}!*hIl5 z%7$hza3qvDezL~7A^N3IZ)&q=_dbBS!hsGYK`y3HfZk~$%=haia~=o$n?*&B)jt|0ygtl zx(a`|CgJ_d-%ms&t4b%Yxw1^@F7EJPA@1%CHr&41>38bf4RL zQ-PiV$n=hn=xD36chVR2T7zYHX2bmWTt(3_9Xcp?4Z+*YkL&g~j2=ag1lV7d+Zoq0 zul*iG)JmQlC?GR2HBdFGFsL|WOthIYA1BUq@$aVq%2qT8ZOWKB9rOAi zm5O3gB|5TIof6To;4(g_@}tvtPdJj<^Kw?!ktpdq9r?a8Mc%?wt)mdH zO$fYhb-3OEEV{HJC)&l|#;#G<2^8$~#1>a0Wb;hf-CBrIe;RFPPE7#^MK!0k>-h`z zv7XZmi9j-58XdQA_FXLWt+I88uD#pG$m!e>cIyns$_w|2&xD~RXq!wN7~k~Xo^Xoa zU!^jaY~A+Q)HN&nZo--R`q!hFe?(YBS@o=N4F@?>B7!TD6Z|CM@3d0-XbM`xXBh~HhnyQyTXtM?U{L4hYZRND+*D+1Au=4N|*4~Tgx`3Y3 zXl1+HqT!2Sf0$EQKH(wVNBA2!_ygWrp(HHoh z)gQBYLIeU@yT>{9+CQZ^Mj#mo`}1-ZASbifBRqHc`Nk)0RZt;0__sKF_6}}5<^Y%4 zEK{HOzIx-G#Ys#u5g0wi2#S<-F=}wSTK@^C`Ykcq+xcstYci(dPl5 z4Dv*H0C51x3UHYXms3FzB1OSMVV+Om%M4Or5{ZAy_r&g3+%5@(78IEO!x@uVuc5`IH^wFGi2@G#{)|c@6RM09`VxtSZJdTx8&U8)=^gD zo6wOhF3d6fZHWR2(Es{~{3^Ch$~g9uPaKK|{pY`6$^v3VFXPxyE2zjdrOqq1Zdh;8 z>M!H}{fVhLO;mGx44y&E`}L_)KReY=BXYDR>V4Jk>0_ZN@?AOX6fK$d;4m)xgf-f1 zb*cVnQI*WnY$t5*7(F%w*2JHnu2_%=G$me*i5H7^$ryA<+jz`A)zel%|4(xK3%dO~ zQ)d5q8_l{;u>vOXjA?3>VfZl(uGR?G}XF3$b8rU$%jT3@T z9_Xu0gYt`jIvb@>Te22ObJqJ3tDI!Z+@f5<%UuBSn6rtbxR~MfE+6Yp@i?-YqB7QX zjP?xj;HwQ+r8t}!^^LDiR)h@*RHdqxirh;RAl(RIcX$0-5;4^nz5Q(5CVb_ z%|_op*S~aCiI^1q;1M`LUMqzj*c!5|W zWi7pUL?i_%gd_RUMmOwXGqK!Gzcp&yTy=MhdMI%q{+|7|gJ=0>dN7oUOjJ0LsqcKa zGb+WJBHTBDP71y+JZAEK?qdHRFMvMYPX>RVe}D{0*RYF5Vv-Cp?2WGU`OMh^NkQH4 z&w>$9@#iO4;0l)L`w>>_3?P#bO zR^bUP+{p4Abrwi0TBz-+^4qSYsG9}&5*Q^qZoy!&r3 zdjew(VNM-Vf-gZKRFhu^2qe-_UrMsl905BC|5m^KPtozOoC2w0iwhsd666s$f}y_V z_jx9y16*FO{0nKW(ZmYYb#tGt`94E|3@Y^KpmP?yW8a+FzWTM~pC-|n?!|S-p>Fzy zyAiJu%93($(3MV0_ohkJ@A0Wy%LIORK`YIQDTjtc#cC3h?3uKDy0qyIN}fLNuSE{lRoDGbrSByU6Hx+n?WIT1*ak|yeEcPS)q&$Gw z1CPT1vOT?l&Z4>_9l^J#%8E^wB7{4Ofc05QooSs#-k&`=!4I%;^MC3qh`qjDa zgPO>b7~z4AGmi1y4C-1|BdS!3Rpjr{T7H`_!sEN~11KSXFdkpW! zlYC_78wTUq>DeBKzw zyC$(g|MFApV)ZAW#X??lfJa(MZ&>H!kMJMN(kLA;p|NJrh!a~MdW-QQhWKc;#v`g2>1JTe2Xkc?!vhk5`nXH~y)+iR5HE zr}TN9?W44DZ3mQ#`&|i_&B$?D4lYa=K7|^d&P0E#dM*6otKXZB!F)e<%6NbEBH`nj zc7G!pa9dmD(eZoB3s9Rrbk2?vzl6iFvp!*j+m3!q3#*oX8cM;6zSMyI$R~P4MTQT1 znQiPiI{I=U%k{rN+kfsQgS|ZwvSC;NU`g8dI}=N{7X(^MR8;cVtI6 z)7p5xn^K6_nu$lpp$@T#)$~$CVl8B8U$S#=`%^o38ay9Qf+bp`CML?zSPtBzx4a(o z*;r^&P6o`z#j2FzBpMkYT)z$Tl>*s`ulF^vA0}E?cSW@9#EeFMXmr}6WnVW7*gjU2 zJwTjj`Uz*}o?rTUY+{c<53_f4Y8!=fW{Gl7YZrb&;~$#@Jsxs;oMP~f)?`3;Ih3=+!4^C{g-^9qa8G=FPlRxQ z?Gv?sP1wVrDhxztB1VuBF6D+5%5gL_4)^bDZhJyA3n$y25#H{SK4J~N@9n73Q9W>X z`n*KfwId!p)u|Yx(`Wu>!>3x)`jYtOOKhz1B>(iW{C$>2rshWABK+srP=3vQe(A%} zG7cDfAUMks4honU3D_?nm2%?8<722D;qMyPD4Egt+unnraBtI(7}{Aj31T9qnNJcZ zL!_ce4c$oo+5yXj-&+bFDg;aE9>%oZe5*~@#zu5nOH1iaUd&6#*%I*|Ivm;A$j*Jv zTNM=|O}}9B>>>bl^190Kj5NMdc0#vJXm(Y7VOsdJeG^y2r-4=*=%IG%WPTQevPeOO zC>}sahqfqqo>Y>1UHwurkidpw5YV-XQHFT%@@B}?fq|76p7{em7q7&ov_)@U0$e$( z6BBsAlyJ5PYKloN_Px^;$4g?OL5qVc8}TC~&wj>x>cK?zHN^yK%^VcCmWz)1)r!QmmT(A zZ7-XW+zAfP+3^)cS&S{EB%lu4QFEV7{K8ismor)b(r(bLtA*87_|kAEB%d-M{DjEy zry=>jURray7VKo)_AXkX$awV|RWkCxiIKd=~$>CD{3!6-S^>wQmxtKy# zlm(h1jI|uW6SUtLlqf(vr$Gz+9OB&avUd0cwo2g&*_DWWhyQRz4|Kn`vc9k@bOxC3 zp`?*!>7=wPJmo{@>AY+Q?3@ai$V*^rIbL7%h*N7yG&}4KSKuIdKaaEFjiYC+z#?s<>}Oe|y9gzDpW}lm+W6x`W??Cr z$D3o@7X|)-5JSz;!;tbKnt-`oB9K=0#Y(a6j-0bpvbGE!S)h<-InVVVhDJx z7YCBNo{%uom%@RD!f;rR`uv_U>0C|F9#6d*@#Ol7Z@o>_~?%PY;=m7g#7A%w|os1Vcv6tQ9$P(&UB=Iqozw#9T+ZrW5qTV=H8y% z)5V{l6S>h2fR9?oJ%tQ(ZM)y+zb%gT3A%2!d)+8=_y5u%05G@fE!!UR8Y&IMG{juu zMgYYR!r0dUb{p!kU!gZXahq?6pJ2?|l4->EAIlj#TlW@QUk5+gl1G2a#=X8^RP`^# z>@9T!2&;!~lJpBsGVc8&S1|Ud38GsR<~N?iyM3rEpfTvSC3#>w*h=o&S;r zbz%lGP3is>>VRY{Y2-qyM^%jYj3KBq+ljA-R94IlmUsN&mc6xBm(`zU9F$)kzAbna znoT0(-o8giskCp;IR7R;pdV~j1GQ-b^FtC>VAV@I)f+ZA{7-CQD8*` z8NxlK@quo%!0hAk6QMB+b#VLu@StL3v4E0{wU3EZn}H|nhyi={L;LFTSQl%CEtg*w z;{u}YF&|S!>4o5&L3CO6U&n*U|MlYbCBU?)eO+3A1{1oWSOkW}MnDYDy~U_8aS2-L zixiF9?QStBE&adk>M@0*>^dDSpCS=8@1WobvBEnsuCduHo36-6<(VRpIg>K=<4-Y9#{B0VPc^{-YRx9Z4;m33tMG$;NJhoLFKBLm(JHD5f{+DY;-tP ziz`q!tl@H`khIQP?-i*+r0W!teZb1Mzss`>fZsD3)PcS8uUthTkwJAU(W5foYLiS4 zNjK$~={BLZy3TkAg?D|k*>j%5jl$cSQu?jPytU%T3dX&hvLGS2Y7YgVWQ)CK9C-wb z6`XK}-el$>%FQL`_t~g0`WYw11KKs4OMBg8NPe!z19Y(;WGU6Qn(6Bzn^*Ayj*T2W z0wzt%8!=7`R$5SgQa~tw)-pR-HBcyOivlevLG>?>QucE+K+>w+AA`1r$znIM$XBJA zbS5;~uMY50?zhxFGCGD}XfOxEaKJtcJkE|9=te$RW*_DZ!(WmW?sK&&JwPQaTlP%o zo===v{xf@!3$@0O_>%T5tg4rHq7ttxqA<3yjR&ZqP`Np}wul?zL!@5;h}>>=W71eC>(n69q2`Yk_An<*;)jV~@G|8LO^H*^_T z%PH!E;o&siAw5v$k$YO~XE5qd?Tq-M*K4(`-^^dBRTBTT;Sx?`|vxyT28UX3+K~AhKa&*x8R(t_Q-aFt~wGWe~{XUJ0qKf zuW?vnQT(^`e}a38oezuXbHP=moa(+>2O}J+n(n(Yt7Uk@e#2DM;)bfZeLcuT=Ja*R zV(@5MF|sj^X**5eW|u(Ur*XxAiv4?`h*0TlD5m=Y%*u(OP||JwQfu1T`S0fS-9u?M zOWsZQ4RnhIHIUL^w$*SZ@oT7sp)Y?-5ML%k0`EVc0O-yA??hYDh32D8X&(cZhi?+$?w^ zZdZ_9taM<&cxJcyq_v-e(d)%xV}DAsrari0m9Oi{R<;3VW@Qyza^sOra3|F`Lt8W3 zu3x!^Ibs&X9O3FhakD&Jir|S_sCqR+w-=sdnUmHOs8Qg=aotyV_!vmM zc1u(I8@?$(p{|j=9q;oUx=5EUq$nHkPy<^AQV-rwS`U_<(AfbU-*=YwtY*j6tMi9GijDdo8&Mpie-)+*V7*N0z)W z7#t4RGt#m;ssX0OCBo4FTuF71w6gFSk&9F@f3&Rp@>_^cVqlv6%oWl%vgRILJqG3E z3`FN{-3m~tQ&yO)r#%TVq);tYMKLM}tYKa9Sl>_JEBE#LeHtu2`nZHt%T_f? zSvU2K&&JjFsC!$bU?%5kO$ByrWN)>CiOERSjb0Zh$cMy2zmXn^9HA_P@#B(cqj)?qR)~k6pV*^hl46lmQu$58%0lMKpy?{1BbYx) z@gDOp;Qil;h6Zh)1OZXI!IZzpqpL!a%F6o7ONDXyjq3a`o|U~D?nL*iG#5Jbvq*r( z0!4^eU=?S}{XlD-KwKMKhTsQR`D+|oB^i>HW4SE@5by|JRu&KYEIQiq1CbOn z^R{veS612PtQL*|D)M~0jhg31Q~gl}PVvZq3$(}%aU|{)`UvF;WVe&KQtV@qnJqgH?}&kk zHBmPq$C0^Pb+G1}fZjsR=#~(UQZPSFd`dR5GkyPNH&MsnL%y}SV*bN1aQvl8nj{&DMk;JoKW@!a19fYFgN1+Dc!0j+sZj70?k1ksu;Qkr_HL8e0U3R!hHV|ZW< zQz73JaqO+1hSIr_ce!9gjrWDVXfEd2UkY)}SWIDk#05Js^eQjnqh^z!^4Ehd(N#NK zo`PeL4NLZTF%q(maYTgPqdu?>_xv0h*1d3?g|&65rUa>?q&wY68A~Y{fbY7VogPYb zYFY|Yx^L&UF-Q39vC+$!|Cr?La%PxOxV+geUi-0+uK%>aMemKMV`T^cSYW>IM`K z!{ZSXN=U_DJ`|`G8ySV}Hl78P#G_!EVif&DT{YKVXsQiPe38V@6#AOW zFj4mAJ(6$xPc?R`6~N&?{F|EUFF?&Z*m;qx%gs)Af6_(l_ODrMpsMCZkRuqMvG};> zNN=LIjXt;^)4uib*dO++OQn8z`ySWY>!a3DkJm3UM%*kc!2N1(T3|VSg5zGeIqIvD zx+))qq6n+K?f%aGRX96+4}--JV{5FWEWGh|G0jvZLf!QjFtOURZS{&ysmN{b%cpjS zS5mBx{?4p5)Qti)M$aefIZYXd`IQ%!$ucG;B|F#p_NDG{b4AbJs|Eonr!g}7mVaK0 z*6Hk8O?le!CEA|Z*6`z|KMC%xk7CX3uH10xB@&V!;kMLx1dfa%p4~NC4}d2RlMhTx zyvJB0Rw^}M?&-slJY44Pc79AV64Pno<^-@tzY zMz_my3+?i!b!KB=cy9<#1CM;grYFTHeoO$ea?2rLRCU4(fY9*tpEza7BB%zv(B8tHnH#ljs*r*~^E0J%1$0* zzg?{Y=g=3hwt?sD>2d2y+}v_p>oNnv9^fWQ|Q})RajLRg@lO$@8TKdi?L#c zy>rE5H0~T-XPb0;{+Vr5(G8{07zTWt3;(H@0V9I8* zbp4`io^{`)!`Zxg%asF@N-1oW?~-fWbvOj9d;rE^0WBdaL*Q7_h*{IfT?fuacul(C z2hlvf%?0&68M7{LL2y8c8fac<1vD6wr-?78AYc6_UI8aqa6}L}C;Y-LmbgP;embDl zJz1rl@3kLZpG)FLZdVL_<2K%BsTWuCm=kYQGree03AC?X2QL82*_th{J;Yo3}IZX8f{gNn#r&}yo4;JqLM~$ z@96-$rimGk`MDf!6a*YglB?_A9TJb03wwOY8@aiiIMf3KR?3B44B0N-2?N(V9u+=% z%WRb#T_`~1qd~crenE&Uh8>ni!Qxn7Kr->EPixuN_BYSwr}Yl=n$Z8!+xc5Bq++2Z z-*F)_g|2FWhiI!?Xy8vNAk~@!SMWp>@To~{Bmz2=yKsdSyuOv`XSkG0A9Oxt9?pNf z6QjLd#c@=%U~jPLB7a>a6<^v=PAO*U4#Ob4+5xCq@zhe;XI@=<9RI=d{L}^C5p}yU zkRD${2d0mK#wHjw@qw_5TR5n?I2N0Vg0K5n8=SPtMY5(#I8h=IKDsEADJ*4z(ADZ? z8(%~G!Xls1=ByiP%D6UUn^5QaBPm>c11KK;GVftj9N9TSnx_YQKeLNhYcEr1}r!|@DV4dN!8qm=E$Fo;IlEV?W}3 z;6>j|Tx~X&p3DwV`#s{L*_XMuKA|xI)YA+IEbW?_LOb%h9?8=yt_Wa#C~*I)7} zvhp-@-Q2ZeXY^{Wf?+ukf(YWK4=lq0{fUseKQX4U7N@_~m*j*xOH<|$?RHm#5L*(N zTpjY083{bSv~{D>z|LO|0fKuSxa+7_FT)?(Pr64uL#Zy2g)Wzf%#>X+cE2@SM0{=b z?n^(-74N3VN(JwTj;Re|36})E5&bTdQ<{BO_?K1fa1EJqVL$bz;H!rpC!q z9F^#M`Qhk)rT$lgsOk%J(>SBMV!(i}{h#XxI>?cn^on=4lwXSD$~R~{3g=24`sNCn^Qf3T1>_=I}okIe~zGaWUpVmuC zrFvvy@zO*ExW(a`??Dr7l#f?TtcxpOU1ki3JYA8zdW1aa%48N0QW6vfr|$@b@_&_b zL;vs4Sv){rmE>OC!oMGG%;=J$~+ryq$!8=;!v!Ecu+x& zzJN_Wa#K-=FcuyUN9zhjAuK~`Hyeh+h$E=kON1xv%l5v+8*D>w7Wav+UIW_Ue$hSd zY|Zl|6GZlP-l2}K%=mK9qLj7&Q*^N7*^KY)in<{~sV?8H;O7TV7oJM{OqLG&G}{V; zf3mehe6cS|MFJpvZHzTmU;7P)uaVDRIl=w)DAD>lSO)_vnc0vfq0IQbWd?l~P6wk+kOiR^9iWC14}oi`|F7kg1adIt zMzANC^GYx)UG{GVkKCHQPe?ObgDLaTMjBiUoGD6@d6zF~4S;oz-RM%5M?qx zhd*K08Op*5KF65ige7!dlza!zzt*I*_pUpH{d%uNRyeu#i1)RV-uccspWid*&ff95VH$=1kFRjYFJ5mxPrc?e zL=Juqw6$mw&GpNG?Jr-pR($q&OrC8|_^j0DhOdjLx<5<(^=ZXzJMp#w?b=S{r_kfa zQaGi>hX9s0VOjr5P3dt?(ttYyg@y!PL?#q(>h6)1G5@rp0_T+x?`LOw$QZ0QOiisq z;$#^^a^v%WhFneWGi-`b(ndOctwI^9)xwo z$4jRous(VI4zn54aU?-w*Ms~8 z)BXP^*{1ipuG|0libCKU&Vd{oSXaYiLK}Nj-UC*8=$!Bscv~pSloHHYBU0`t^anBts~Y>nt$)&xV0kgQGeokeBs zog6zg%SpoczGn|9J#Snf0x{9QBg0zx-UI|D{!ReIe2Y7*Iy7Vf=p=lrB-4ZK5I2wG zXL-3(oNiTaTxekq6eb&)%{COOtQ1Rmp@bUH!%{~_@BM>i>gqTm8Tkl;={U{>c}Fxh2KI&xI=#-9PmsXek>?XR<2lYO9CAm z0GN}A;vR;8Fu^i0EPRJfMQu9$xf1TtWkE=#1rtkRPbFv3qq=eM+A4<~hK2c1l^b_& zNU&>I8_to=e|yjf+Sf}fZ5@`zvCm;ThrYiEN}@Q6B5Ye*L6(|%)4yQGB!$I4ter{&QdEJ` z8ud-s?96*#O!K*kNe9WMhIS9BNz39kG1LKzREwpN5gt4(K_29_{1+64SlO2>5uqrJ zwWv2604g(Fs{J@I8HW)0L*u8#SRKVTWFb^W`+u;|)%5$o7PRc4$vqaBIJU;Y_OE*! zeb`m$xnpvdmA;K`#Dyz+KeOJieft@;bsqPz#>&#|_N>{^y6PK2Lz zVak-va3)=VOW=Z+mm`?m#&VTt&`cpk7^;<;E7>kKOlf80x17QRme_Bi+h+KJ*TkDd zD}&fd@~~Jtm)^iN4`ye!?cAqwc;)N=_T@h^*?-bluXO`I;Y>|V#|+TC!s4*VMLlSn zOVE}i5SHf_>+8eO=7;Y*_@G8oaY*a2{Cq#JVDZA9@^b8n8OR0~xL1%<|FW=yzfjAH zl%5BhN+QD&7eL{1*v+jrI}i<8;tscP3NKd88>NPS6aX*?*;IZgAKr~Z@r4a_cYSaPwN2Wl5KYZ_%)Kb-&o;rB)N$xUs{`Q% z4pr*B!$!oOePD}z?wC&G>yAOL3+=@Nrss@e65*bZKoC9E(r$c(HjiSgEl0mV~G4EH#87Qqbtm zBys9{BJ|#^zkSnvOfOvkJ^t1>gRqAS8v2c}9LWXKDSMgKwxt8#dj6{UH)}N?ALtv6 zu5kj${*9~q*x}3bZIuo}gm+oC8m>cuf;k3!crD|ebb~1+l-7Th*m9> zRvRu(7x|fcO?e#l-b|{Wut@UkMprb}_En~TB_t(8 zr-G9ilzsiwrdkpZtV!6w$t+sdf%#)6q~W>O-)x>1=YUb*(OgajCvD@;V71N!H*Ce= zmhtxEC*a0*;|FUwIkeb&5fTG=g^^6Ol?TZsrtA+=C3(9TnM8sf_Rr9vnhbE<&F%jP z;fj}(tfr{=+&T;{VO0CscAetWXwcx52ZdZ8^yuGaB8!@OIvayJhzLLzZM@muLKvA5?sE)0@njuBTZ2gFf5fkHH7u^GBz!@x_bW#HveOXk znO^9u;c2V}T1X^?qh9TX?c$qRXpxX(vR^KzB>=_*CWIB8N+kvb`K7dwCdw{fgjCXJ zQ(^-r;@qdojFoW8f5V+yo0sQ^@RuXt*cHhJDYF(#w!Dt$u6*37o{Z6#YCVUR_vSkF zgs@NlxxS{y*%86Qw3rV-tq!3Wu~8j$9py}t7BtG?A*Ie(shp#V578IIKxUND3m>N%yddo#P!S%d^7%6I2=AU)9Q2JA>G5f%->>{1LjKgW@z; zvUiJU7j#P6qJw+(D~+uarBIo?p-mJ5^ShWFm)Q%qVP@hU8Y+|-FFsYIc_on>E%-&! zPh05vC#u&FTcDl#(?4O4aZeD!2$J6HY`i8V${~Bw$kI@7&Xi7-mM-3NUZj>aCu-2} z`iR5byxO4{YpITU!tOyM77=5*Ua+VT00nX%0qlUxPG3ElpTgiJ^n4jt_DKC9wXGkb zB1xjZXz7$wbc%Z8yw3FRG4Q~aYV;9wKkQokhzrUtD@HZ; zsEpQcGLkpAu3p@_zi7mGOM{ZjkT87mU5Z}201L^2KSxth!zh@l;66Q%6BV*Ns@k82 zi(~aWY~xjOfoZKKSb0~$SaUCHk3%%ndkdxCC*+($dw)^8(LooGNb7l#1DL5I?^?qr z@PcL;0#3mw{zvfHD+0)Xjq&O>eS2Ctl(e*S7untx_k83H_% zC(+CR6l(J>;7Vk4n9#r2K%ZAcHZ=iIU}%l8N7oFA%31nY$efZ0#vonCS^oTG7c(}Y zJ?cpeor1o)l&m?2W!%zOJRqXqC9Xg{*?xbVqcR2H;G@8?0#`X%X6lUs3R*A1;%jpQ|{+I#~vWWu#0=)3#Ck8bDg6s!q%%xaG8)vkLf`hdch6 z`_$?Kz%&b=oO3ES;4>H@L^htIQ5GoyQ`WLyed%Or-mR2H0G*UG9vB#6H1F={dKeeD z9av|A0n0WFUg-dG56KLHE(HI)!so7+sXJP53Wyy1edmM??g3ssIv z6YNlXC@hGtuQhBObH>u=nS$?`YZ?8}s*2z<7xejBYNSwm$;?<25WC98G6X@Jbf*8y z%YNZ=M3^@nJ{Dzr@v?KUju)4yd`}HcpTL)_EPg>*xX1jc;%qllh^exO z8F^{J&P-S4#e&2}YP8W?a)|SiE)j*UOgJI6*@9J8mVB|;nU{9OMEiCTXRm+f%3Hkr zq=SLSc|j>!4;_{`eP&$hIv}WYuH%T^$PMUZUDBrD@^$gd_qmzZCTt-PkO0N3HI1_i zK}Iuw;#64xs%Rgk_`|CJVb?+}mThexL78uQ7$CnC+oR64N<^|FZce&SE{a4l&y6k~ zdVwdFjrCE4exe`Aw{|!7ad+qRe4B->bn~(VCZik`3OEI+xW8KDJ~>|s-(7?Mm8!|H zg4v0%J^DUp=@(*Ved|tZ%;R)9c0IenveD^nc0T8+{o}&oY*sR11K@Ob4{&Basz~8_ zV>N`BhcIWydhPO<~8(j{z^-|sWDv#W^g7rvLYi3(<(5M=gvJWBpd{q>!=#Zqj3C7jm z00Xt0(`k?F4nhJ14JT27yp(lX8vFQ{KJl3`z>0eEbcuI=V}KC`?I(7E6zuKKB0cK= z#z&?Cfd#`{VZII@C0n7NR6B-6e@S;6+6?!3kmjG@&D7gEr`3b3J(~J4XLLo*Ppt*K zJVvC(# zAydd>EFM+*=)h4*58Caocyq7a{Baqd;!r=s3RYSauUA8_+}WVvx?8+`HGh42{|yH~ zGu}hTFOd-SwxJ%U1sUuW?<3R2-OHG;$13Yf;<$pCGLkZyzqBSuI zVWZg(EebedI9b2HAlw*93t$FtH}^fBlNk15!QmB=H{`*Du9(lO@s>nUo#J8+fDyDFgSeFxwS{D414=p7q;LI@4`cF{_qYpU9C_fG-Uj z(#cMcw(rC(gx(+&k=vlS|Gf7|FM}c{lrt#0z`zcJC+PVOspzt&Q7T7zAF_!RVNQVW zwkRs}$Jj#$*CA8i5}iB!s8zNc$Y(ut_wq$`?+3taV7&Z@HFV4S7a-YY*;!YN;_|v< z(=J9N*hx>TLz0u0`Wd&)cX+EUz>S6)ePK{ITQE%XeX!IVv{w<6Ye6*$=TtK~4umM? zO~s9(R`hzXZ4=wBtOKGXsaUukVHmglWUA4Tm=A-Gq68+H>jq=PPX6rWzVux*ynm!e z9+DwWNb=@nh_NmTVw?U5DMfEjVt!koF!3^_I8=b2SCE6#{AXkuB8p2`x>e~Z>!b>a zG2z#N%wso-C#QOCX*6TC9n^5TaDVt@<4cVng#OUm#mSE1j2HuAK4~00Lc))hvZ)IF zu}uDhW6Y+O)~rU&lDMH8P1XG{3!3m^JSHwZ7FeZw$cmz6_*pjtZF*)A23eG3grvUgh&AW3(#dmo#-e=jcw%X5)iJ?~qiTT+&r zjcc>1hk{8MpRB^amBpEr&tbq&lE#3JVt*|5MS)0qkWEPu{Dzsi{tkIa)xheR~=j408?9~q?(ZgpKl8&ScU!WGv|VP*5<~E zjKCeymM<5=5uE<6VFqtb+^5}R`$df$g_8?C_Mp zO4b@e*re>pYiNqN`x}e6QO(&{l9@IMUKc^@RrAQvRC|C0mV9W#-WGwtT{~7S#RPUl zksbk@|67EbAD12CkR^;`=!0OTs9Uhka|V?=g8FUI7CdB)#gad1V4>tw306tf798GX z98o>$RGXpajHastRIuEs2*~FKs2lCl*Ub02m5lprYEQO7%EZ@BSNen$-xN%KjREJN z_{s-5$ExKF0ikhOazt##Q%Moi%ypx&GgG?oj$J(~BBMn7R$IKvM5)K9en9pvMp*4= zmy-ebGn@^F^4HDkOrE&MsQxQvSQ}(p#^>HYJ`j9!YmJccUfaY7x=}e!-K1J4nB$q~JK#%8@t}2;}iugbUUbw~+H0$@{ zybGCffY0{9oiF}5H&X?`I^j#aV&`ATV)UZn3wFw4IB3oNd6=yp>0mk-@sn##=bAg- zNIifuA=X6gn8-pwmd4KGbK3LAFRx6I)??PoL0b8DW88)QN?CuRIbH7pF4G)rm-Tam z51T0oKW_e^QD#sN^+?o4o|b|9n@jkco3XcpHKt=@Ar#%qJ4@cA`^YYt53=i$i6kvCOTbw6%WM|le1*^7 z0_MNIEmgoyOkh0vPVX2Z@_Lx??!(=g*}L!&O?q=H*DL8@T0xpKQ2B7G_nvq3LJ}G^ z^zz9A38FU;yaHAFc(%HavfB#Omo6>ShK=w#32mCBJHgg|6#vst`JZkOzXBMd*w0Is z&^-fasY%;q^w7bMBwpjudkcv`RxH5-eM8LTT#T9=a-OY6V0u=et({KgXZFil>34TV zne`EOZ!{$3PbB3Za~HlhMU{>YyTngyxlSYM3!v3IX{z*BwPoUpG?PAw@4FE z8A|321fvx{%65c@$RTFRMXXWwn%&qJQX; zQe>DxMrO~H5!F{2$piAC`h?=?l?>6P^E$EFI16cd>Q-1bV+MM#ZcJ45FiwF=mokbf zE94L77*+RB54!?&(u>6h5pk06oW=5}5$c?0u&CPVx5{F+jCA`oIR#0>J*=O)Ifjrc zal_;|9!UVju=Ii7kB=hLNUJOuImWRRu6P6lc$Ue(_6-n}veLKi8KWyq4SrwW+NoPt z$Rk8KZ~JX^jO6+kYBZOV915ycM8)IQ1qK#+xsh7(MA zPKh*$e;HyR0>1Tzunt{vZtYXNb9wwL3e(rB;0 zN756W%HUVe>&UJ3H6V-Q%8Ur<)q>@y6mDvY$cFT^cLYYX0%w@?huvo%o>^25?p0)P z4$Z_+;`2>;;Pd@-FIteB+I2Ip#+-))bpI(SGi((jV>Nsw9Q>~qVKWKBj|Eq6j z$KlJL%;A{R&N?`x`B^-B?$fVPnl;lF9=lBiDnifx%Kv!Cr-}BKlDF?RXi|9IQ$i}8 zFVC<2R!sP$R3%lG+!+y*YkYNn8upW4;`;w@?Nk8Vd>Ab$A|lga#qgFL;7ol6To^Z@ zA0Kh)OjlqvWF4JXqYt`!>e1{NDSTUP-PA#l*0ybR?6W&ZeR;VAa6h?rt5vbW$-61EOcR@Rl-}!6lWv-`J?jk)V&meh_lc6RNgPhlZH10>st>j+Ly}#Uqqm%8#!Lst-O;M7|73 z3+zBe%2HitirzxrF0Y~$M)7OHK_`pNAT<92#jo9sQ? z=w;856jo5Ung0rP=67~}7`b^R2ThQr{5W8$WAowH?|na?FD;jrF0Q0u?xC=HK0AOf z&z#zdF}6b4TSCTS>Cz+G7U&Hheu@B%qhwGwZQf3vh}7@=C;3|A`FEL;Yy0i?uDw~q z8I*n!BLpY^o&~UTTQDrbVYgKSXbxt#EiYbID+00) zvwa8UDUz|q)6$DKT1R6^M`ji`LV<42JlT=i-Qm^K<%S!uIb97v2@)sFHuUgSHh#o+ z;cB7uDTxU2*|06Dt#_ozgoGV)(O)p(D5>BR9-QE(ejEF+o=Rm>@-OaS#V}Nx2@7}V zG2j6?djTb0{ES3$Z4b#5oM3(^t-h<{@1z(C+KdcNk5?h0I?m%@e9O23VB8AeJ=R5U!wC*p zeR??_J`Y!WLQ$BZXyD6CJv`Rt_EZ|A`}Ts;D&&Y0u(c$pz40jpG*l>wlhRVj(~?xo zc~^(5gC2IiNnasK<{g%f!+I-Z9+aZ5C+%qHrqt==gPz4RSI1{L@25az1rs_SeLP=B zu6t*LK5VRl4-^?L!kxekAFmw`FIW1Fa07%3Nl^R*`eMCfg=WX#xww>h{i_#Kv9;gA zL2IbDjDW3~#3h#@jq`df_H7Dy>y~gnPGEF-gI%-NM8%nRf}QA*AIZfo_2nub%`Mx0 zc6+xZ7N7E1JGzBa9x$|Cxc`Q;6=81yBU8Tv`P zZ3%z7coA7)_P1GshAa-!htO+(p0!ttjSfAEGtp%6iXgIEalF-?E#}RyQGG^x{KaHq zPPkQZabS9_jY`nJmk6uTl+trz%=2q2-$$vEJFkJSp>9@rEpQLKi?A^GpsQ^4Mwr&l z1hE@ZC3RC}<>TQMwIKXeHLHMU6JUI7)4#V|?{YTF_0Sy*S9M#^dfStTW<2Hn zb39o+UmIFD7Q^nJ+tO~sA8gT_8*LHahVojyvs1!wW3LO{qG+%X2~f7YP;%_|I=Gi-vIu|-q9)1DaEafMa00?oC5I^!u>){=G1FP91D#r z>b(g>7O#r-w8no9OGH@dqVGR{Ce1|Lsn93&KKAfc?HFJmU>esLe~-5LY6SjBc#7+g zCzkHA^f&dCs7rAV3BHq0r1Ec_Ivfr~&t}>6_G`JX@Yy^JxY0AaN8OYBUi6?=M$ME^ zH^q5851FdAjehA+Bbu7uvcTIkXs73?+Pq|nrIx10 zABjDj z5^aPfDmD3La+{|<_P<$D*=zd(We!ExRv)iyziT}ll)*)r-jqT%PJ}$LW)DP6)mGFm z2J#B7hOTG(eGwfc#{#h}_}!hAic4G8xikVUcENk~$L6w0lR&M`GlF@}jXBg5H_QjR z=Gt!R(~}D#g3BYSNdVu5Mx;P;&GqlNMH^RxX8H3ic}jb2rWif?-&-0H)TIbOpUXA& zbA37{?e6&kvA(*KT8Yq0^nPNxI_K2XY>&)WmPCuK0&h6dYg2Kpn?P_^)|%sG_pJ<= zFk*X;y^1o&Eg3f^c3F*kN=2*0Wv8!C`AD-PNX;4-B}ib(E|IXp&jD7okB^;n>Toz_ zGPZ}Mwk)P&&}cRfdmaAfst>}Tgz4>9T+@A30`+}V({LDh?;ltQ_*Mk)nRKyO^R29W z+mTHDnu{HhUlzCAN9n=8F5HQiqkDntDxOMyW@f42U}&4Zkf>=Xo`k{p5>^7O8qb#SjIk*(BMu=>UxUE;vi2o7Z&>}+>(>=#$_gmU)LuO{o z%?2(&Jbozs1wQCV$SFU>b9<4p#Ih5Cqi<#PTlz|idxe0II(Nq`JAbuG-{cGb?@vBN zy0vjp6;I!>q2~#rX+QJR7T3Lxx2V0c^|X z?c-*Insru|5NwDm&6z{h(5viF$3v3O#*4UhS%N|C2^YORABcbfH2DEpt_1A@f&xq4 z$*VC&n@2^t#)&t+;8u<1)cx;5=6m$7mrTo zE^Yf_Rm1XA0^^Q+g6R+Z&NK(V`(09P{qTFch*6Ze^IE@OwJx`F*=|LbbpQE3Wtbv8=P=^(Rrnu{Vox(Kn z*~(v~%ER(Bh+NtSRwd?%;#E7u+=7aX1&kvLCmA&+Jt9>KVSXtxQHG`j$H2@GJkdN zfYBH@C)4(Yb#JckpW~JzzKe}*0v@Hu%$cI z*D#!YIgKj#c>q~8@(xdgTDtK-FI+U2Z1}Y@sYgOhw*m=Ael-(z(>dBCrtwLFzBXEQ z*{V!nK|@>097*l3_WEBsTcl-0=qU|1yHvWi2PBOwC|S`G(R8>90!wHj-r+7yl>JT@ z1hm*J_n)Zph!=r+y}oh!Tm|}A@7Nk;Y6AGmXgQqBaL4q-bYGtrgDM`FwAyGu>H3P)4A@9gk>6*Uj*pGPnf7ycSn#acm z>gRR$l+q#@I*Oe(Kwup!w1gEOn}r5aJ-Nj3)X*(&yeLU_W>CvOE9Y#2>e0aU<_57j z;JtnN+nz3AJTh#0uD)T+7~A+E>TBx@PUTY9Gs=C^IzLj36mZ}}2%zjCSXL!fzDAxf z_3|N!vy7?6lc9YdN6n29h{gyf*gWP@8pTvKG4`r^gYO0Lz@%xMSiiQM`D zJpH~6fKZ>?o{YRRp70l%RW0o(L<}io$g4SnG)vV1N-vh4G{Oy=oJm*@9@e$uAV%f^ zLuMCafmzdv^kbBBcltpK#^ma5^V1Qif^i!+s~^8DdOvgFC+m0sTDdTKIM@di>&LzQ z!WQrvC(3=ztKaif{rIPlNY4Yp)DI&k4N35_$?AS`49z^O{DYCxJa57I+dHg~x!1*j zqI4$I<>5u`^c}XfkjQT@&xNn>Yn(05n$_2ks;;)d4_oeNz>-!Ti%U7n`!!}wJtig6 zQP(Ma@m6&?CGbeQk2@>{CPb@Uw7s-D^M!(J8;&?%U@^iRx>B5+z`oCVLey-+uN_Lf zaQ_oxOTBgFnIQjn5Xe381Mu)?RoBw$?yFrdIO+JnF74E$pWB#)AOahk$Db}My#Tqetx*4CfcAWDZ8@KkC0AGh=$<*14 zP^@wTu`Evv@*@z4Z2a1zFp2)Ixv^u3Le3O4fR_bKVSRNnyRtfL#cHof`zcH66Y z!Uc^7Oe_PdHezNHh^xWGTl@>q=*@#MCjsw$@gd0E(EE@k^-2mr+)mxrvotCRK>i1a zLQxhc%a(tHbVt-Ptc!ZE)o1}jN<4+|`qeK~|GWEjW`N?N!A}fYh6CA%)tOLo3sbG- z3&C>cq7cbn+xz32&nSyYqs~vvE)1M(5zH#Qb@=@*Zz!SB!#byS;i{%j#2Oq}<<8!( za^SUdP1+3AiOAu*TXZ4=qnY~9#$`L~5wo5uFCk1yio3Bcd8TIpm0B1b00%+$RgT*& zP=NGw#@Yi5tepXN<`LIO%_p&BOf`d+$u;2}h?5_0vWG~CFwz|M0DMTE?DM`J)Nd{cfit8eko#MdC%XEa@V1L^006j6t6Z8S6-z&?)apUclkx zNy)yDI9t<=nt;!pP-03_o`CUq82zL-KKuuxwHAiFJtoqf&WBC77DzU}>YdOn6oyg$ zF^hoQgW*zZ^vz1LYh&3Drf~rK`lIFkcuQ=(fAKC2BnOw0@qrV3Kop058ZezZu$6s_ zNTj$^&tAZd3jZflZ-+k!^fZRlV%jK<9-8pu-kAXOpp0hfYGVDJDpd;CrW&-b`Mx51-ru@LX7P@&QjDsa z(bb3e!JN)M;k662QS9L@UTQ#N=XrD7e0u5$-S?n}2N|?VCt0uNC$fv`j zUG#%&X8pWAZF>V@j^~X<$Pt^EV81e8hgrHj*WPuNf>h7I;SPu9>uR*9ngg>)dGuW8X#I^Nm1hlHqDQ$(8R334P(T zW&RO3%(4sYKB@r`lHX9)(?PbjodkgHiRzbVKPk?GLL30QSCp{(6_>(-mq)y3L_mmS z$YzDnHA&*dX@IoWepBUWjvE$h$)L6RwMmLEpRyZB)5iJfz)UuAviUfB;#@)*UU~&E zLVH-Pr z-u%wyzt?TKgj^FOF&a@EmFgYtfevUw{dZ*9dSt@KXh+V1qP8}uJotxwunxwFQ-b}@ zo^SR>;EJYx?n^|0J_Q|Mk~*^YVg7R(;O?@NLD<;J&qoXlUL5~*e3p~2 z4?Q)_XMga3zR7`uyucl=#dpm+gqb<6Cj6Qqn9GYO$gmB;)L6iih@Xf(ce2+RdNQJ9 zl;e26#P5A}Nd}S{HoQLkITXZIe+HGpbEh|t8@Z)qsmt$j@U>u40A2c0SyUekd|1A- zY^0GTrw^yGo@SZc`l!LIO$sVrz6!$@n=YG?Y55y4;z+bXm49>zH{?v?HMGEHz=JJj zg3_e5K0Sk{j;s`Oc@<6avD;d_v%8|Ds5Ai*rqw)2!DmNMX>_GSKe#c$QL-%jN<&E? z6Ya3DG-WUDQ>e?e+Z}4t)ZZ8M=hOH|6NsMPitOC=h`1@IgH&GC-RC!*l7fSZ{v_CIjJ{LQ{_Sl`}c~SroOt^Hp%Crzo(%9oP%Ss`|h^L!2e=tLf~rcTI-5wVHuWRSEukWVxGob7Cvnb;Ie7x4rnf zi83?kD{dUkr#P#X84%lKXPVs^5q}7UFQ1|RP~MycIx*93B)@6#U$-#U^iMJ)0B>S> zRBBrK_+N|a#SgF2dg3zmzWTQ~_}^^=hy^Bk7HyxRR%$;QEcsxq*o=<~qdON(NQa-= z#C%ma-xR~tOl%^_Z&rG=2}Jvm1iQbFaCA)H%VOS+4r!c`x31B zT&JBBF^yA(b}MW3oFfknYqa*DU2ol-M@$&l@yq9}wuG+o?<`LOb9f)FN87fuU2(0Z zWMvd+nOHpF!wau035^4U&jW6-jo)}<8wE95tpDOtpn?2+6}E@vA#%njbfx4U+Shs! z%Ee3G?$?ChShYnGccNMZ>e^mNqmet!#pj(vD&86@JTKCH$!35#w)GvqTZd-vCxv-q z#+T<@?l-Pd-#enEtdC9%cFpzAIiPOeo@0@7y4)ZyR%&utPiipUI$QOW>K(ULUgh`0 z-b^P{H%4cE7ZA6=W2c19p)Up=0>WPYsMk3JEP~nNn1J@7%uugz05BQ%sbR1s3ZsDs zs!xcT3*w;yADMxiYXD|08#=DkpK#YX^yS!E@XLYtnnm$tAO{UN1u<2Mh=OmxO3x~8 zK12#0aT)bCu>sZMNdhx}Rymg)wF8VjNb0@iqYyO<8RG+nOxZsCW`~#E!vwGp3OWAf zcLaJVu+l7t)PsI{S=W|PtS?e`gYoHs(9kb=__}_w8AlML7z}-3ubJQ5-O1m*7pVB! zvz|aXlfSVpF>KY$#D(cy2qJ-`Tki>8+Mk=ZE2S&RD9u9Gdjr&FAmPhW1bWB8RfMKA zXdqGe@uPL=&teYlsv7W!6)a>i-}y#eLWoN-C8f7-E)6R)5(6g#ShJ&SOSW7WVyk}a zobuz&E_nTZiHUN~$9H-B{kdoAHW0;2kYX|cabA{kJsLkQ93@utP(O#`LuK_hdP;xU z93BExU=;3aEJAwOBZLR%`bV>mEmKaHINUL9^FFrG zF#ii@RL25i$hN@?a z@l+Yj#r#2IsoJg^wr}t<|EXo95sE|o^F1U*r(#){*26?;5IEWPC&=YwX8(RKb=?$# zZRyD!W+}-nICglX%8?)2;J~{+e)Wr}8o4iVk%Orv`_i{G=7Hf%u+HV>ecvrjJwiA@hxO9%2r|K-O4FPvm1^)&EN=Gz$qV1Cw(pTO+1%Rz1U(G zqM4-1L~f@Pc%x`$^Fks*O0*)hW`v@OyY^-vxTOm#~ar3 zM`HtZYye#`cf86!nD6(2e4Nev4xw&~5XS~JB{ zv2KSmqv|Ei)PQeJ?nl+z*{vQ?x?0P^Ggw6YMWhKXBztVL}G_>5%nxUeNNq+w; z0s;5^j;76;4v87kIT^TUEHb{|Y`s)D<>?qA#qAc%8_YrB?S`RmK|((X{x&xd&UeH+)8Wgll>sq#WRuP*Q~X(BU;5Oyv?WpZ zv#Z~R5BaP{E=*1m{>=0NHUy;v|S*e#O zq!Q53d9un0^Q{n~>&IB7fw(xsDcdzD9z9(TW^?4xywY_lNAvXDeJw z&jyI{g0M1=S`>ZGsdq(i@Bk!O+=B6-}PuQC+Qo$d8*FV|9GyctM$QtGK!m0k0yghmQ zlMaZT4268>Ef6_g4CHnBAn+3*Dnua1lI%D>9pB1U^XE+;%oG;}J?y1Oy!OZsb=>t; zpW1U?>>EtpUth~sQRn3DdylLk%IHXloj(xN=bgczA;+xMjx1er8F1iD0`n{j{A!~w zhYNV51J~?~KOKxz$^mrZ^6}!UFU+(+Z@zY)mT;yMb}aopg+p3s$e=BOq_x^9rau4>0D-30o^-;(=wmkIM6T=2mGjGL1wpGO*fzaL)*r}ztYc&)>|7$pAFgxh^FuJgd3gGD_mXYm<1L=P z4+($Wo5HFHE%5l#;#xIXw=h#NvBbYbIAbFwW4s)z2khGyT6w+6iZch~A4&vU|@qB9phr6`o z`y+}+x?Waqe}JvLYy-9gMe}YZIfrUn<#qZ2h|-9&zL>(zCR?Bf zvBtGDe^qqZRp^P2&m4=&>ix)qoWE_8+Gt}CmT@0=nKhWGVgG$xo|>XD8b_3i58J(? zxO@wH+)u}wT3=J70!mwmel z^lp~czyRa$np$K>CU0JcSj;xg1KHYS*AZJs;@|9|lQ_njXl-xVT{vHY!BMp+l8g3I z_aXi3E$Y5aj;4o^qlpsp;uRDF`VK)NqCC0+TX;hbz)CPZbhWQf=b?$F5Y${D?+w$c z+*mP}mC11Kz}q#572xL(Mi!TNvmi*Z9!jIpflO4V!Z%S|eAisaZN&i+BDaw60uiAX zt~}Z)rGgjVOv9er78wtj1HAP!f%-a1WDC!?DU$_;>S6-r^wG^G+6=r?@BHTk)Yd{U z1Xe3=c>=IHIq#YAg!1!LcEaMU4r0jrqPLNSLj~=6ZO?*b#Vs{gD`)SbYefZ~4t#CV zlnxHuii)KSQ<^#nRQgsOZO(XJjnXk>k3JN7AV4kOO#4Ea1!`g7BlPunR>;86;Ni$j&(j(oy3^w|+evo5 zE+AkSx#;ai<+cDspj$*1?2H2l(U#E!l5={82r<4dh8uQ|^o0QPu3KL%FWK}nT9pFizmaGG} zzbO$v1eaP{n@3`R!ecizp9}Y;D#aQ3e__2r-hZ)N>`X%9if|mjcun4i7hk-_Gnshd z)q?5cOrwaYvaXLOIGNl^i&s@G5Wtxt9Q)qGX+pxU^vwX);O>?QxkYtoEw;tl%BxcX zI%QP+IF7u`$`-kK>FpyNi+glHAm5 zSJ|v_e_^r>I;^opeJ6|Ev}2~1XyYA*%2N94^b9)Dz0`Mrs`aoJ=yA%)OIy*V%imExvn* zBU$g=5n*4HNIxIESoVna*w=kDNU@&c*Swo}%4Nba`w`W&wzFupI2a6np23oz9X?HC z{bP+^bK5*V3H3?z9ElX#A-`8-k4yI<=ni^Kp-uPwEdBST{V#04|4OL{4Y@NohDizw zc3xw1C)||dvyf!|r%xP0e6oenq@S=(z5jKfRdu|B9;%3`7$LKPfRg!7tPIo?jl1~d z_Tp7m0+hN_6`o>jUP%}=V4hxcwRGZ7bDK4*ShoYO>v`b_REy_d&Ft5ITg~*nBN*F` z(T3Q|ETp>Uc0G!wy(`WH;CfwZIXo|06G%D$2v}7qY$-d{t!H^^XseynYvqhatc!}F!UuNr zrVHu&IgLVEdkp|P&>h*HJW8ziy}l;oC>OAF{H`qsIIYFxM7sQOs-9JVn#yKK*eds&%Yz0Ga^{Fr8DF$4e7Wo=;`~ zsDESOma#Jk zLTB?(dMC4Af8YHMfv|FJmiw)`BS=JIKypwmfc-c`a+yr>2`73QMMtZD%PT*i+ z90%SAU0SmkPkeDAo@ zf6){DE4N)#9g^L-W%*tQvDYj2N0jmVj`n|?yc{%aX(bD*Yh*Q!%cbA9Ii-4x4xxHp z-^lMHR=A#Wnx0A!1orRrN%0FNUVAv!9YeWcKA?m=ExSs5lZ4Fc3v0ep7b^Xk_5Qba zc$wwOfK1DZ0-7dbVZ|xXvfV|qHDyRBS zc>F)vwRg^x0sFm69;k@cdesFM-^gccCOfCn`vURt@@-;09bP7W&C8&(X`=0Hw57Gg zv3^!?t-O@8Wh&BwW8^q;>=-qo+5#VWAk7%=xtsJH%?RQTj38h{9yzY6>a~KMwxSxd z`kL)i)H>&}CIwpe_%h)0u5`V$JR#qOMV1W@^Rbm4zLc=_^u9b?-TA&gd!BYqJo#R@ zaxD&%(b_LVFGIJ2yW~+LaBgO)%b(%81e6}xg8=0@EJK2v0*n`R+(?I~yc?NRFpg1G z@Q1tHpr2LEMRuo2~K2v3ZwPG#Xn-5{{<8RD5VkWx72~zp4 z)WKu;>SDdhb=8U5#qllaYzzQ&A`7Zmc!~+M9k@YBs?P`YQDLxTGmM`EZxM|^fx~4L zHI|92rj!gb)x6b!BGuG6w3mA1zltf!HTGqmz-M*y@-2x!H&;|L{kigdybMwUAafT_ zZF>9{cVs8KA{Hpt;8-Jtc=CIqP3p0p3*)c#;Z|=Dlnn2_HA0}Z9A;ukorznuzW+D1 zwG_ZF=s7-fMA70V)FSuSZxSeFKNNR&wB9tVnRuuAfV1pjcvcppRdd?*DL8{Txifs@ zM2NkCZp%c2U+}NLj?&6SAsVYo(tUDTL4Zm+)uA#;nKryOm{iGhLl%l9F)H}Dj}kVI z6mSm$PC(Sl?u28nMYP7f&AQe}gbyKU3kyV;nPH84&KnCphh;(N9{?q8eZ}yxyd%@b zw7vk@#xjYu;DF`~k#f?Wg0zFpy+QM>vyVUetPR0ZufLGH+;_kt!ge_x-B)ht=PoGd zRU5=^pgZEx%kABIdnYkud;i!;RC5qxZSd9V%gKb}AjdSla%xqX|Ls>{f|UTnf(Uz; zLEglxmahG$;^>yLe{EIpoe0YJ(jY~Gl&KNw!9jC7voDe%@8TezbqPS#j!Gj-#CyPt ziFE9TG{8cE(heX_<*NQ=4pw;#hb|GJobO5-_a~Gty=e}TM@RCvyJ9;o%fQSa3k5@M zcSCVAH+y~G%LvV#twdXhCs9q}*zjskJi1X^+~;ytb8|jaRh-&46MH`pKzO}G4F8Y5 z@T5+cPk^cg6E;GHJ3XyxW_|c;Y(wiFuz@CEu3yzF^zCdNM?;~Q(F<|Wic4s zdg5a@CT7^)FqNO-{a0(*GWNiCc*;3qw+Jd7KcmD}o`S`a14O^73vrIn9Ou7Mv9LW_ z2Q(RoI@p&4vw_yZhH0gOfUS?(xsDsab$EB$+G+b=OvbII=7Qp25%4DP1^{p2Li!?{ z!#O)QO)~j!Ya$isy0d*DB}$6hNRMBP-Cvk~=H=XN>#f*szk;vD(Ozxp zTu)su+7cCzN+-y2p!s}JX3yKW{OV(i<++V*f>_mO`Qr==xuK%Jru zmsyF8R>5Zz{IoCD(@vh8L!$&`+2S$quhK-&NEIyZm5@V~mCdflJ+G!ue>LX(1mETS zlj5QT+O(c*((|}iX(&#L)PB#kka{|>cUl^^3|unPb#A_ zYNqeE!RPjIgb*=IS;;rly^rNd8VpZ2pe&i`~WUyyeR!4nahvEMnIIoaD!7g~5;YBOnb`(XO&#z&H#X z+#Jgjzn}p9OpNdvE>p;2@s|h+c+=6x$E8n6V?cCdOKP-}44ejj%~ zP1?}px7>~*Y@9?@WY<((?L-ygN~b!8J^5x`FZu_2pjbD<@vU7|hlw5outh&1?xN_r zxun7ExbZ_*Ssdb?);1DXWQ3Q!xIEK}_(xb2GRL2@D+c3Z`6#)bI8XR2168Ef>SUZ% zF$pST9DF1{7OR6$yodQWFq7=b-+E0LJwDh+OL8vJ8oW>R^zOdbgXk;O8xTTI;*1Wr z&{FA~e(KCk0%Tl-o0_r=Y!7kmcy)G{*&V)qdw-xwZb0_tEU%N+b?|lYr;cbS*P@wo zO1t+ZuFv}zMAO8O09%AR!r{vY0VDtMBk_7!bI_T9Y)Yb^uX56Gk}p7gR}&7Q^skh& z;V(GvJc?pfZw>3(6#(eN-DA1MR{D{p%L>=xQhPLuHaOm&WRKY(Nptz1*9!O^Yov{F!7^ChIkqdfFz% zD3+>0*LXXuG|}pVhHm#)h>XXZedyt~@t128si3lMFGNVrHMC^E>8iiI>P%qFAF61` zM!ejChJdPJ<}1LYRkOFRd7)%(D&~5+&*@P9>z!9y|J~OTgCSXGYf#Vr(Lf{rL}Cfq zl#SZs!U(2lS~>xXwT%F5wD>+!dsj>f$c%xl`iZ#XXk+YBFl;xt-|OK<`0;FK+VI)% z?&}q>F3Je;qX8-)u6>Z~#McJtetxv$)X&t4&~yg;9SJ6mW` z@EN{%iebzD`q%zrsLRlvEI7ic34b;+6J}?8jXpq+F>6vit+~;)2i~o`roRL-UqRQr zP{JAdp3d}M{@zHnS||#;o2`6vQd@rx$3vbpD55z{fsg-OR4er7L#$pm|GuF!z}&$lW?xzr$qkf8pK-jo9b%GXcLd3?v?_9>;c;YD?`Efjm~Izz9^7Iq+zL z8Sd7t&}bMbxQ8x{BF%a#*p&QDCW+-)zUHhQnP;L{9C>wTQLd8;y7PrJocBSU|Exny z8<0I0|&jYb%(08b3Dx7eRwb|ax{k99MtpEqL|J|?MDZ=MSot&`Z-}!yWf5gQ61xo=8C%qd-D~VLM_#jrj!1aibQ*D19S3dfkbX96i zl6qw8vn!TqRZD>O$rWm%wqwV#f=jzc#LMm;=l~IUP_t3nJ^wJP)Xx^p3QZu4}Nim+>Dgm^Q2Z~UkRn-&e?RHpkeOm?Gvro-5?gHtNuyUBY7VHAw zapjq0QQ#>_@6GUyrugMJ`E?0ic+AyiWFZ@6^EG7>4>AuiOik=MB1zO9>94VKJPjI` zRbF51@p|YgSH-5UgmWSs8)q6muJoHzDL5WZ(*wWgqX2^Zqg<$dm#w#r9;Qh^IHmK@ z2Y`4j#`C?S6RB+m`W@80j+u#+-P2$z;5B;>ejPI2XT-nlmaKPG2#s~cxWh}su0ajV z;%}dWz!6~MHaWJ|M@zeJ?FriNC!J!ZKTxEM?KDR3GYHID2!$FVNV0L4=kRK;oLKLB z5wF!&IC?Blc|Wr}QIHIs}FYqd^dmo z9UyXSDy&mDE^r9n#hCgPm3)X{KR|{_)|~hphwQTebv2MSqXpMrXMb{_tN_{8Cdn=; zGVEcfK;SBzo$a>zhw6z(SLr*`Z?|i|q~$G6zXz;$(3cq9Z~?cwxG0y%Yy_U(#`CkQ zq~a%Dp1m>gzAvFJJE<|JX$QXxNs=#OglLwUt|-Re6GejgY+YeRY}c2R6pllS2Ng$( zX-GFXv(uRn-uLgV*Ubmzo*GaMu;b$qR2v(i1g*n2VEhLW*tr7u*8W2UYn&2Qdv!{v zL{0N4f1pfcRB86d4d_qXY`)JhgoNTqAK6XFi(7|dY}S1{(x<9sfp&{F;qoTF zh6Nlqg_cw8b8mFEYf#3;I=6r_SH=a7rV-v4(3U(8Zo$}pK&=zVUI}&*-U5^l63X4OP$p8$)2Y1H5+T3{ z%!tL(-DPi3K=f6Sk1%1w1iV)LwVPK#masv@sb7&pZi;i{f^{y%t`U{|K0G6YpnI-t zfAnNE%Cp~vfvR`^_E%21`VW1rY9gn=al3lqu^fUE!+UPj zl)KUk{E^A;o=-zH>N!EwVTW0fcQQIyD8>Qnpy>IVS}*FaVds^}6hd9abh`=u9jt5d z(3_%0kj2S)>}#&zc^9}#1zTZg=scD44aLlpt^BY=99(~;06&rG}|H6aj*nGK* z*!ET=8UNDP4gj}9GWf72*A$}Wvy;=@)wNs4r?j+av8e)UiXMxvpQ$~;3@hiAtt%XN zO#X)U(eUqQ;OfMC^3Lf}oi$pWAjDN<=P{sy=-9#qYX-*L- zeFT|9k1j9gv)SSt_C1gweaGptdMFRnA7)sTByS=C8={ONIg zaOn=>sx&;&T7L;F6y=l*gs$bId!EVYt!-Fx8($SOd_C$M8Ww(gO)^7TuOImRXWcg` zXPED)+|dY~M?k4`{wL@>D|PiIJ_ix5gb8u81(W$qNTg4nWsrm$z3W91{iJOLYO%9> zxqUfizq97Y1{cbFkWkA~{=_c;^DnN?Rl$jd0mH?vl>;Hxf2ZEcTff@!cZ#$5gR6$& zEl&c=!k=2&QwZ|c?w|1}^{TZ8n*zbvOB;TH05>+(79 zM~Blr0UJw8`udfaihw#+nA)?GM;k#uznNQhZg-R&lhc0{UU&6ZHvhU8-j@PF8Sbhrw74RKhZzkwd^puS>&-o;w zdX!Hwn}bj`J4k!gl5TAt34B*@{(WWt(?$K~-KMfvS47e~nKB(b^5mn=&0V!$*hIP|kve|J)uv~mm7z^7T65c!^FO{*P>iwsea9MN6iIcog0^zGcOrH z5KB09Af!KcMFT)JN`$7p5dO8$M<;Wv@1XLomOhl2M>!~w(`d3pmyB~15D-d5VteMtpEcp|Zs32bua1WL>@h~2G35Grv9YZSGSEFfbz?{~s= zl-FQb8mVj{mC#9jJCDeD_i}@})3Iy1_mr&>rM`JChY{#`%s4?aJT+d{Y{$xdPp$xA zS)oSTDM8!0!$@%DoMxP+4malqH9-2^$%$fRSkx%~q?*cVi-T^5k3?!+>j7|pv9S&` zhs;OXg)Ehte#A1rP5+4U!7vCa^P$cL_&akW;mFdDQLTcFVd2wgODZgiHyjP>f;{bt zI`g&!8I)de$_UQ_HSN3>F^ef3?}~XhguL1a*o7J$2X{;4En=0b%)*3^DEqa~p8ei_ zuyK1g-~N_MD?N}~H`Q(2`bs?}DzYE&hcExdJt3O40*z}gLZG#x4EYuw+6PQmYvQrg zaWG`tNJ`Ap%}I3crV%mM*88eZwBks;ED>t_dF4A>1LAvh*mwvO^TlJI1+W1_?&aC^e!InmtvrzoW0-e_i!u7L5kw!?mgYSx86F zQK?QxT^l`Tm#prRJtT8o5O_82U`hfJUyj-w5B_Q1%*M~4v8qN7e!Ljm;TFF5>~)%2 z=q&sPW+SJOe8FO`041d%c@8n^-Wb%?5frd}@nuCxFGe;C49k#W$>N!GSV}NoSvp># z2fOO`(5=VWxo}tC?B#Wp>0ypVZ)M~6FjMYmH>_mntFZp<4w{JamIaJoh=|VM)EvU4 zMxX5L=Tiy!6*{$jrfe8Wuz@fA0Bm5?pH3N3^Z5O$-*#x!%wW$EN&;HUj~ z;9e<)6^J$qF0$pKH!t2KcALbN>|{4^tAJL4>>bEko|_TxZFk%6W%beH^l2Hk8OM3a zsF^Imm?sc2fT746HwICTUqiqz?;54-uQtDO`cCgo%C{#EuHE5yk1-YN(|!-m^tke5 z!=yhy6~L|^^wiiRYVS;DAS^_x1Gmuo`IpW5VMyhnt57hr45zgeU@87PPM~-2nt#=) zJC07Q?hQAMfOYlYF*I7_z>BHmGtspB8VPdkchgjCPDezs^TG5D7M+dd0 z@V>XC)06HD?5B@zY4Xpu4;c zKs2A3!{b?;H12nd5Yt$l+9|qLx%UADcHy`j`e8(M$yL=xfY$c4nw0rv->4|4vi)R+ z)JxJr=phd~PjNFPRXis{6(T}!Y9yT|o_})vlqwB{e?oL?z!Qvdj2jRiHn$Cn@TbhE zoZ5H<8S_KoV*xKORd5|Lm2z=iHC^|7#uo+dSj>JUK$`ST%(20jRgF4t58Uo<6+{K% zCBYYxe1wJF^CBDlbtBNA(6`tUFPzJ{H7u$VPUx(mgGp<#?YXL1NqFe%6i( zf=3SVNW>T}la&d9PRIPs^MG}Wl8)1bi}Ljvz*UdxFXN|uDP7yN1fgfxoa!r^W!?^! z>$mO{Jx>2h#WFQdPlhOx{=?Z?><@>YYaa8#pfVfvZ(k*6O?w(4R@zU6<3-t+FkcA8 zyPSDIHO3%ajl`N74j6kbr#44m-CD?@9ew7H?c&mDtyQn*(05cXm7Vf&9Dtay1M24y zW&?8oa-T+RZlIan=R)icKwaM5_&Km;BK5229Y2(fYoqL;#!5bKazv6PNugH}OY;j+ z**O}mDJNhX4w z!XCQ2lcQ3cx7vQ9nY#a>2Ge_!`W{m zoe7T5rDsUMFZOqx@Nn=?E3wr6(`?xD&#aa>cgBYvP&P*>*G5JXk+Fqx{R{T}>RE1r z4A7|RN6x2!(~+XXlKEnEcZn69kofoZBJr#zxFVx)T}tM9idH(nO!Y|)*2yo8ySb8g z>&?hqu|-K3W=0$s%c0*`YRXZZ1Vw`Dd^zPMG{mww$$lLq+4@;3LD-eVp62?1L1ZlYSdqS?G?Ty>O0X6_Y#45 zn+Ir;Veg&A)fe+i^|3%kuH0-M&wd$8dPeHbxLZci>n#y}lmXW&6nWt#CogxN{ryr4 zihG{cu*7mQ9rB@e;%@Y6^>S4%Ri0~~cgZewp!(Z*iZMF8Ufd~|i|t;|6Bf^2Uz=oi zqaeUgs@FbSJqe|p(F%7n(`&)Q9682ht(Wb}zl}dy9(w3%SzR6dEqV}^2mO<>Spj`M z+oN&iou19dmvX@zK_@>j#7D7w^qb$Axc;~Q#OUyA(0_8nP<$QaB|8MjF(669JSon9 z5+ltz+UY-B6V$KqqM>Z-A6n*qf7%ZfbaJvSrxw0nu3^;cxO=(e*R1VLSA+V@LJDTv z#jT=(?ZbFJd=AVNBJKlDoxJ4I z{KFXk|6fr{;l0VJLhHj?^!^dNz&8+#Gdy;vw3+(hzwsPMid6+n6kG8!)SFzdAM3Q0 z;|pZV?1Ogoj4DviPj(#rbQE6)u1+YZpL$gN33uIxp@p#$af)#a;`Gdc7}ko@vJ@0I zLNJ4*0guH!nlyz!1>C;89N&`&zy&X>(eVFWJB;zq1rm^4k*W0riNfHz2WpNcUqK

    • From 6b36b3edc79e08c9d1e5f668845e6ea3b6afe17a Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 18 Nov 2025 11:26:00 -0600 Subject: [PATCH 517/590] fix: Update default `ROPE_THETA` in `ALCF/helpers.sh` --- ALCF/helpers.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 5020bcd19da..687ed2e53e6 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -231,7 +231,7 @@ setup_run_cmd() { "--num-attention-heads=${HEADS}" "--optimizer=${OPT}" "--pipeline-model-parallel-size=${PP}" - "--rotary-position-embeddings-theta=${ROPE_THETA:-5000000}" + "--rotary-position-embeddings-theta=${ROPE_THETA:-50000}" "--save=${SAVE:-${CKPT_DIR}}" "--seq-length=${SEQ}" "--split=${TRAIN_SPLIT:-990},${VAL_SPLIT:-10},${TEST_SPLIT:-0}" From c39584d155c7492481b9619d68df52c769c4ac70 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 10:14:25 -0600 Subject: [PATCH 518/590] Update cpt.md --- ALCF/notes/cpt.md | 107 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 107 insertions(+) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 2c66586ae89..767c4322017 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,4 +1,111 @@ # CPT +CPT is tthe process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed. +This document serves as a strategy cookbook for our current runs. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. Here we focus mainly on the learning rate schedules and the data mixing strategy. +We are not exploring alternative approaches like **model merging** which might be worth considering especially when we train on the math and code dataset. +Here, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. So stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. + +## AuroraGPT V1 +Here we are using infinite schedulers where we warmed up the LR up to LR_max then kept it constant before cooling it down. The advantage of using this is to avoid the need of rewarming the LR when doing continual pretraining. + +#### Stage 1 to stage 2 +##### First strategy to try + **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then + 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset, this function handles the data mixing. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do + ```bash +python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt +``` +This means you are replaying 10% of dolma and doing CPT with lucid. +For convenience, here is a copy of the ***mix_datasets.py*** script +```bash +#!/usr/bin/env python3 +import argparse +import sys + +def parse_args(): + p = argparse.ArgumentParser( + description="Mix multiple file-lists, normalize internal weights, and apply global file weights." + ) + p.add_argument( + '--inputs', + nargs='+', + required=True, + help="Pairs of file_list and global_weight, e.g.: --inputs 0.3 f1.txt 0.7 f2.tx" + ) + return p.parse_args() + + +def main(): + args = parse_args() + inp = args.inputs + if len(inp) % 2 != 0: + sys.exit("Error: --inputs must be an even number of arguments (file weight pairs).") + + # Group into (file_path, global_weight) + pairs = [] + for i in range(0, len(inp), 2): + file_path = inp[i+1] + try: + gw = float(inp[i]) + except ValueError: + sys.exit(f"Error: global weight must be a number, got '{inp[i]}'") + if gw <= 0: + sys.exit(f"Error: global weight must be positive, got {gw}") + pairs.append((gw, file_path)) + + # Compute sum of all global weights (if normalization across files is desired) + sum_global = sum(gw for gw, _ in pairs) + + for gw, file_path in pairs: + # Normalized file-level fraction (optional across all files) + file_fraction = gw / sum_global + + # Read entries and sum file-local weights + entries = [] + file_sum = 0.0 + try: + with open(file_path, 'r') as f: + for line in f: + line = line.strip() + if not line or line.startswith('#'): + continue + parts = line.split() + if len(parts) < 3: + sys.exit(f"Error: each line must have prefix weight corpus, got: '{line}'") + prefix = parts[1] + try: + w = float(parts[0]) + except ValueError: + sys.exit(f"Error: weight must be numeric, got '{parts[1]}' in file {file_path}") + corpus = parts[2] + entries.append((prefix, w, corpus)) + file_sum += w + except FileNotFoundError: + sys.exit(f"Error: cannot open file '{file_path}'") + + if file_sum <= 0: + sys.exit(f"Error: sum of weights in file '{file_path}' is non-positive: {file_sum}") + + # Print header only once + # Compute and print normalized weights + for prefix, w, corpus in entries: + new_w = (w / file_sum) * file_fraction + print(f"{new_w:.6f} {prefix} {corpus}") + +if __name__ == '__main__': + main() +``` +##### If loss is not recovering, one can: + - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. + - If still no luck, I'd: + a. take an earlier not converged checkpoint + b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) + c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + +My guess is that since the distribution shift is not too strong between phase I and phase II data, you will not need to experiment with step 5. +#### Stage 2 to stage 3 + + + ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: 1. First, if running on resources different than in base pretraining i.e smaller num of gpus, we need to train from an **universal checkpoint**. If you don't have the universal checkpoint, you can follow [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. From f74d237b407c49f0a89d239db51ba47547002ad5 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 11:20:51 -0600 Subject: [PATCH 519/590] Update cpt.md --- ALCF/notes/cpt.md | 102 ++++++++-------------------------------------- 1 file changed, 17 insertions(+), 85 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 767c4322017..ac916ae419e 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,20 +1,21 @@ # CPT -CPT is tthe process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed. +CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed. This document serves as a strategy cookbook for our current runs. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. Here we focus mainly on the learning rate schedules and the data mixing strategy. We are not exploring alternative approaches like **model merging** which might be worth considering especially when we train on the math and code dataset. Here, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. So stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. ## AuroraGPT V1 -Here we are using infinite schedulers where we warmed up the LR up to LR_max then kept it constant before cooling it down. The advantage of using this is to avoid the need of rewarming the LR when doing continual pretraining. +Here we are using infinite schedulers where we warmed up the LR up to LR_max then kept it constant before cooling it down. The advantage of using this is to avoid the need of rewarming the LR when doing continual pretraining. +Since we have several training stages, one need to keep a buffer $B$ that contains previous training datasets. Recall at stage $i$, we have the pretraining dataset $D_0$, and previous CPT datasets $D_1,\cdots, D_{i-1}$. The buffer B will contain samples from $D_1,\cdots, D_{i-1}$. -#### Stage 1 to stage 2 +#### Stage 1 to stage 2 (weak distribution shift) ##### First strategy to try **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then - 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset, this function handles the data mixing. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do + 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. You need to mix data from the pretraining set $D_0$ and the current CPT set $D_1$. Always start with a small percent of the pretraing dataset i.e 1% then increase to 25-30%, **5%** is a common suitable choice. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -This means you are replaying 10% of dolma and doing CPT with lucid. +This means you are replaying 10% of dolma and doing CPT with lucid. You should also start building the buffer $B$ in prevision of the next stages. For convenience, here is a copy of the ***mix_datasets.py*** script ```bash #!/usr/bin/env python3 @@ -102,9 +103,19 @@ if __name__ == '__main__': c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) My guess is that since the distribution shift is not too strong between phase I and phase II data, you will not need to experiment with step 5. -#### Stage 2 to stage 3 +#### Stage 2 to stage 3 (shift to math/code datasets) +1. Start by trying the above strategy i.e mixing from the previous stage 1 training set then do the same. +2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage +3. If all fail: + - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. + - If still no luck, I'd: + a. take an earlier not converged checkpoint + b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) + c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) +#### Stage 3 to stage 4 (shift to reasoning tracex) +Try the same strategies as above. ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: @@ -113,85 +124,6 @@ This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -For convenience, here is a copy of the ***mix_datasets.py*** script -```bash -#!/usr/bin/env python3 -import argparse -import sys - -def parse_args(): - p = argparse.ArgumentParser( - description="Mix multiple file-lists, normalize internal weights, and apply global file weights." - ) - p.add_argument( - '--inputs', - nargs='+', - required=True, - help="Pairs of file_list and global_weight, e.g.: --inputs 0.3 f1.txt 0.7 f2.tx" - ) - return p.parse_args() - - -def main(): - args = parse_args() - inp = args.inputs - if len(inp) % 2 != 0: - sys.exit("Error: --inputs must be an even number of arguments (file weight pairs).") - - # Group into (file_path, global_weight) - pairs = [] - for i in range(0, len(inp), 2): - file_path = inp[i+1] - try: - gw = float(inp[i]) - except ValueError: - sys.exit(f"Error: global weight must be a number, got '{inp[i]}'") - if gw <= 0: - sys.exit(f"Error: global weight must be positive, got {gw}") - pairs.append((gw, file_path)) - - # Compute sum of all global weights (if normalization across files is desired) - sum_global = sum(gw for gw, _ in pairs) - - for gw, file_path in pairs: - # Normalized file-level fraction (optional across all files) - file_fraction = gw / sum_global - - # Read entries and sum file-local weights - entries = [] - file_sum = 0.0 - try: - with open(file_path, 'r') as f: - for line in f: - line = line.strip() - if not line or line.startswith('#'): - continue - parts = line.split() - if len(parts) < 3: - sys.exit(f"Error: each line must have prefix weight corpus, got: '{line}'") - prefix = parts[1] - try: - w = float(parts[0]) - except ValueError: - sys.exit(f"Error: weight must be numeric, got '{parts[1]}' in file {file_path}") - corpus = parts[2] - entries.append((prefix, w, corpus)) - file_sum += w - except FileNotFoundError: - sys.exit(f"Error: cannot open file '{file_path}'") - - if file_sum <= 0: - sys.exit(f"Error: sum of weights in file '{file_path}' is non-positive: {file_sum}") - - # Print header only once - # Compute and print normalized weights - for prefix, w, corpus in entries: - new_w = (w / file_sum) * file_fraction - print(f"{new_w:.6f} {prefix} {corpus}") - -if __name__ == '__main__': - main() -``` 3. Then, we can run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): ```bash From 1f76b5a34f080f6ffe1d730d66d73eda9c5ffbd7 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 11:36:45 -0600 Subject: [PATCH 520/590] Update cpt.md --- ALCF/notes/cpt.md | 58 ++++++++++++++++++++++------------------------- 1 file changed, 27 insertions(+), 31 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index ac916ae419e..a096a8357df 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -95,6 +95,21 @@ def main(): if __name__ == '__main__': main() ``` +2. Run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): + +```bash +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune +``` +Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. by just setting a different value for LR For example, we tested rewarming to LR/2 i.e **LR=0.0001** and 2LR as well. +Here the following options options/flags should be: +```bash +DATA_FILE_LIST=path/to/your/tokenized/data +LOAD=path/to/your/universal/checkpoint +SAVE=path/to/where/you/want/to/save/checkpoints +--universal-checkpoint to load a universal checkpoint (not needed if checkpoint not universal) +``` +Note that you might need to convert your checkpoints following [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. + ##### If loss is not recovering, one can: - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. - If still no luck, I'd: @@ -102,20 +117,23 @@ if __name__ == '__main__': b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) -My guess is that since the distribution shift is not too strong between phase I and phase II data, you will not need to experiment with step 5. +My guess is that since the distribution shift is not too strong between stage I and stage II data, you will not need to experiment with the latter. #### Stage 2 to stage 3 (shift to math/code datasets) -1. Start by trying the above strategy i.e mixing from the previous stage 1 training set then do the same. +1. Start by trying the above strategy i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. 2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage 3. If all fail: - - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. - - If still no luck, I'd: - a. take an earlier not converged checkpoint - b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) - c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + a. Do the following + i. take a checkpoint before convergence **i.e before cooldown** + ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) + iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + If that does not work, + b. take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. #### Stage 3 to stage 4 (shift to reasoning tracex) -Try the same strategies as above. +Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3a.) + + ## Legacy agpt-7b checkpoints This is for doing CPT on the initial agpt-7B checkpoint where a cosine scheduler was used from `lr=0.0002` to 0. Here, the CPT stratregy followed is the [replay+rewarm one](https://arxiv.org/pdf/2403.08763) where we replay a small amount of data from the initial pretraining dataset and mix it with the cpt one. The steps are as follows: @@ -137,32 +155,10 @@ LOAD=path/to/your/universal/checkpoint SAVE=path/to/where/you/want/to/save/checkpoints --universal-checkpoint to load a universal checkpoint (not needed if checkpoint not universal) ``` -## New agpt runs (phase 1 -> phase 2: weak distribution shift) -For the new runs, we are using a constant LR with cooldowns. The advantage of using a constant LR is to forego the need of rewarming. Furthermore, with the cooldown, one can train a model to convergence at any point of training without committing to a token budget. - -To do CPT here, -1. Convert checkpoint to an universal checkpoint (if running for example on smaller num of gpus), **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** -2. Mix the datasets as above, I would try different mixing weights here to experiment with -3. set `export LR_WARMUP_FRAC=0.0` in order to not rewarm -4. Run -```bash -DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_DECAY_STYLE=constant bash train_alcf.sh --universal-checkpoint --finetune --lr_constant_plus_cooldown -``` -5. If loss curve not recovering after a while or loss is diverging, one can: - - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value, data mixing strategy etc as we did with the legacy model. - - If still no luck, I'd: - a. take an earlier not converged checkpoint - b. Continue training with the base dataset with a cosine scheduler decaying to **LR_max/100** or cooldown to **LR_max/100**. (I would experiment with both if resources allow) - c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) - -My guess is that since the distribution shift is not too strong between phase I and phase II data, you will not need to experiment with step 5. - -## New agpt runs (phase 2 -> phase 3 or phase 3 -> phase 4: strong distribution shift) -I'd try steps 1-4 above and put more weight on phase 2 (then phase 3) data when mixing in the new datasets. If no luck, step 5 should work here. -## To do list +## Worth exploring Follow and implement the [recipe](https://arxiv.org/pdf/2407.07263v1) where the new dataset is incrementally introduced. This might be advantageous for example when the new dataset is QAs as opposed to pure text. Here: 1. If base model LR was decayed to 0, one might need to rewarm it before following the recipe 2. Constant/infinite LR schedule was used, one might experiment with the recipe as is. From 0f23c947a8a81961f6b93d5c7cacfda2d7bd7191 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 11:37:58 -0600 Subject: [PATCH 521/590] Update cpt.md --- ALCF/notes/cpt.md | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index a096a8357df..3da769dfb68 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -123,12 +123,13 @@ My guess is that since the distribution shift is not too strong between stage I 1. Start by trying the above strategy i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. 2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage 3. If all fail: - a. Do the following - i. take a checkpoint before convergence **i.e before cooldown** - ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) - iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) - If that does not work, - b. take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. +a. Do the following + i. take a checkpoint before convergence **i.e before cooldown** + ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) + iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + +If that does not work, +b. take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. #### Stage 3 to stage 4 (shift to reasoning tracex) Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3a.) From 04f5b1e67064d8a11d63da1e4b5a77b8e5497706 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 11:54:05 -0600 Subject: [PATCH 522/590] Update cpt.md --- ALCF/notes/cpt.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 3da769dfb68..5b33e04f1aa 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -124,10 +124,10 @@ My guess is that since the distribution shift is not too strong between stage I 2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage 3. If all fail: a. Do the following - i. take a checkpoint before convergence **i.e before cooldown** - ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) - iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) - +i. take a checkpoint before convergence **i.e before cooldown** +ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) +iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + If that does not work, b. take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. From c10a6f2800c146c8544a5b8b9357362fab6ee555 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 11:57:16 -0600 Subject: [PATCH 523/590] Update cpt.md --- ALCF/notes/cpt.md | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 5b33e04f1aa..cf1321c4893 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -122,17 +122,15 @@ My guess is that since the distribution shift is not too strong between stage I #### Stage 2 to stage 3 (shift to math/code datasets) 1. Start by trying the above strategy i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. 2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage -3. If all fail: -a. Do the following -i. take a checkpoint before convergence **i.e before cooldown** -ii. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) -iii. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) +3. If all fail, do the following + - take a checkpoint before convergence **i.e before cooldown** + - Continue training with the base dataset with (a cosine scheduler decaying to **LR_max/100**) or (cooldown to **LR_max/100**). (I would experiment with both if resources allow) + - Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) -If that does not work, -b. take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. +**If that does not work**,take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. #### Stage 3 to stage 4 (shift to reasoning tracex) -Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3a.) +Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3) From 1e55c76d792d029f24a65063b4ce82c5b7c794ef Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 13:36:10 -0600 Subject: [PATCH 524/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index cf1321c4893..e86ac9281fe 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -127,7 +127,7 @@ My guess is that since the distribution shift is not too strong between stage I - Continue training with the base dataset with (a cosine scheduler decaying to **LR_max/100**) or (cooldown to **LR_max/100**). (I would experiment with both if resources allow) - Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) -**If that does not work**,take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. +4. **If that does not work**,take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. #### Stage 3 to stage 4 (shift to reasoning tracex) Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3) From 7abf24d4d1d9cd6a37e89305159ba5e59360d725 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 13:38:18 -0600 Subject: [PATCH 525/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index e86ac9281fe..04b43c0e74b 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -2,7 +2,7 @@ CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed. This document serves as a strategy cookbook for our current runs. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. Here we focus mainly on the learning rate schedules and the data mixing strategy. We are not exploring alternative approaches like **model merging** which might be worth considering especially when we train on the math and code dataset. -Here, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. So stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. +Here, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1, \cdots, N$. So stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. ## AuroraGPT V1 Here we are using infinite schedulers where we warmed up the LR up to LR_max then kept it constant before cooling it down. The advantage of using this is to avoid the need of rewarming the LR when doing continual pretraining. From b73f02d6a1076050e7d6495df3c635644e83bdf6 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 15:50:50 -0600 Subject: [PATCH 526/590] cpt data mixing image --- ALCF/notes/assets/CPT_data_mixing.png | Bin 0 -> 169092 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/CPT_data_mixing.png diff --git a/ALCF/notes/assets/CPT_data_mixing.png b/ALCF/notes/assets/CPT_data_mixing.png new file mode 100644 index 0000000000000000000000000000000000000000..3258af13d46f2c5536ab52b2bac99408d267e11b GIT binary patch literal 169092 zcmeFZg;&(y_67`y3ZjIhq<{!W3re?igVHG_-Q6G|NJ=*%Al=>FEg>=FNHa7;zh}PQ z-&*f_*S+@-c-LKPU;)FL`FzgVXYXe}&$H)?qP)bDN5qeikdU58NxoA?Lb|VqgoK=h z_5l3F`J~Gn{GwqdDyk?YD*9Z}-p0ht(ijPeF4`%oS*BATyC>t*k4&*wTwg@7g&C?P zwYsaF727G_cSkbG85aCFrTOx+urlD=^7GA1AHRwAZY4*&_Y1G;>O!AYK~Ev?tO{Li z7b1fq7xse-B=Oj%tkF%FFt-+8C=-SR5M76US9ZPsJ#qM-Tc=BRQcS z?CK9Y-&z)RyhCPo%9S!G`>EB=AVPgn$e3L6zouxYL5}Yfz%m!ziEl>Qvu|u9%C9` zn>S2gt+|ZVrA*}HkY0hG(U6dX&5%&RPsrel7OZ>fmb2>k1qjU{eQ3W z?^ox4>+-*K`G23{|DB_MpJo1kyDpobP+%=rCA^oL*^2`Vr3>Rw#Xn*~sP8gy(O*cd zTDRUf*FxNP>RYd&ENSYfPz}^~;7K7=o@C!k(P9tJ#=V*pk3UJK$t^1$u;z^j!CQE4 z>E!)IQGJf3Px7J4sWz=Q(@&ng?z%Z|y>nXo${s;t({#9SzT^YDy@B~OMhaZgk!F|| zyT$a3QJ=VfG7cO^7Vcv&fdPepcd=o=245feAqW*k- zds^xf=n`pT|CO!BcFF5JZB6TD)aQ21dCrWMbFXSL_u}?A<8Bw?wvmP%S*PcHk#qGY z#bzO1%X$l3a5{%B0k6;HWX8f1&V}W4?@P1uPAtJXblTK0ugk4GjTgln$uHU*#B~`! z^)ixKuHQ&^FwO9)A#oUD6KxD}Zo9i#y1QQ=%9w7Jo3!wD-trsdvyK4u|ZakIGXCH{!0@Cu*-3Q2=XASmmuV5~YbZFNC2W|J*C3U3!RU(Dh z(0&TO|GJ_0S-a_UT1%KPP~QAtx!t;9D;MJiqUn}B%4feW=`#FgrmnQ*a=`2Cw`ukF z9c@BfqHIay?8<}HFqQ@_o8~u;%&Se~E6dQH1{7mVgWHP@7us~!k29|f&gX4f^~WVN zlsKA?1{kbQ#;;6^QYDht6|#8IJrBDH_lqh=9zB{ldiolvLr9ABr|by#Mk=fN$KH)( z3+CEROy1eip_;zXUJ4&RIhzH$Ub*w?Z-U4VdYD__AA>t?0tt10!Xwr*FcyVT8&Z&* zZVy&EJD;n#NzrC_+Sgv~HzrUxQ;UDKJ%@PMxTl_}fmmtE{F^U#j$W5`{jtAQ}Z6xG^gDBL4-(PG@%%S$jgf)lVr0A`;N5;-u+FN2Dcx-VY2joBDh36j-UEH`WiQRT|XS};@kDB z8;0YTXc&gSc`7)k_Qu*rVMyR=*9_uXi+@DpOA0aiCKq$@CxzwkNoed1_$3||9@l_& z!isFOk_>PEbK)WTz$4K7XuGgfHK(He0m)~0%XkLgVHfU)wbS!1d^K&Q7?#qYC^}2mG*wf8>$l91ib*rH?Y7+!UPi|ri53x#9F$6bK z9Y$p5Fzx0J1&#zTHTqVl`dhyLddRj`Tr)S2<^pkB!M5TNJV;D8F(KN(|C^aA5KB|T zM{jNp*?$21%h!%_b8g)E{vw`BM2e3{Uy&F^8k?D1w{I<6~$?Ou4+nRxCA zLX>f(ve67!buY?%55n$YMmF#B3B+CqOg5S%856j)AwM)qcdkZN;D+MW?$ylm*zj+b zG#yEEZvFVC8a8)u$TwjL{}VbJjk+nDz%bL)?KWI-eU#Apc*D@A>H#;X#a{9i{%@4 z$J`p;uj?5DefII$kk^0@rw8nNUklaLLpITcz2*J;Ip>w50z2d$yKKS!T-JoGH)r!J z>Yiz5f5YWcLg*&z_Hbu#O3Vyb9Wt&&X5UA-4|a|`r+LlPOS7sh&2{yQj5oXfNVN=8 z_4|Iin@uJZK{IBap9^%(AH^`jcFQu~kt+7QrfToNBN(_*nLBi!|0d7VZx9}46M2M{ ze?7;!NAFe}`mwD~Q(-gqVZNGJ&#NzdoT=^{V0l|_1+0Z*ai@#fRx z`o{VKp-b0)I1Q4??aEcf3$+ofbq69()~NU2)VAIf8w$TtnXL$E7r|h{BPd{n z{?WU;oQYyqLAo&P<{dG>so5?l42labXEwg?OdzO_^L%K`u$!JCfWl?LuKj)5mA&`f z%{dD_uGE#7)P7Ej? ztoCMUCU;#NbYd;uC)8eU!uqBU9jtnbSH@6W6+^{*;F~EGL}G3RE$4l)%sj zoapx+Vm?-FJm{r-6P6uq6Sz;SkBG72l0UXC0@%PH`XvMd!q(C%czYO0gVEhJ z1MxavT=5W-3WYMo>oA^u%xaogbERG)hS4q!XGaPU$E;qR-+E5KuEC|$YN7sab16TY zv*2YM$W65$h;nIs8?J7=0cySrmu7Eh8Z5RQyTwZ+`=}?Tgtj$gh63W8A+1GBarTuV zZH5gTkh7;kQWA7(Eftak_GmL~eJn7nx7TbR$j;>%r{8(qj7w02D(`+<&~R zKi5IbpBt1jmzGIgOYeRm)kP9c~1?uSSE_Ov3HIa&a!$ii(=K z%S3i@LS!cNk*RHzll%heO#T>3P6;~hH631{da!LNAZM_`i9dMpnRDav1D2)2fY(IpY!Be-gW}zQ+Hk9C|(Y!=7G#2`h*O;`7P9n8H!6C~XD52pvPp z^}e2zrnVk)BS~~;cvUrVoR?V;!|rp!;>VC+?^RQR9E>E8Xh8FvCy`wSTVv|R0{ z2w;@T)^yH={zGUONoL|+iT)Dv9%K7;dNR6p+rp_Yqs|x_WqhNC;Z}PZTvRm3RI9Ct zHA#Ndf{1POT>5#w>U0TFD`$a_=$mw}A?lOV4vsyLrj&Fw?yh(PQLHpu@6;*-~*>M!c(juwYH0uO;+FcRjIvCWY)tj>s65t=(HqaekqcbSo zCqdzLLMN?dqcp3S3|SXEyfv2D?=}AedCGP;-I4!sZvPy<7MZujT(|7M<%ISg><}IhWEmD->w{M&9Ke(QG(xT!e9G)A)0yADEjd2SCY%7Cz)8sMzI|9Gd z6BlwLA^8pVoE$DhX`Yp26D}@s!8xRl?zVyD_D}Z(N!tEel-;BB^wkkgOsQ8Jb zt3`QLR_D1`avz=uznf^PI8;r~%^_0UKejYnQJEB{FG#3zKuvImwvPao!9DDIp7wUY zyy>y;IHV51Fx+jkZ^M?)#$K|*t)yil=B1naoYr+choK32>5$TA+^IRcGVF>q4&C@_ zKiqPdB!zm|qc)2DI$&VOZ-04EvddkE@p}1O%`YX_DJ%W0q{K91W;c|A^E4U&H;687 zm==8gc$rjSrXzW?cM}}ErZMeP?dq)M(>Fjo(`g%bG}ZTs1b34Nx&z8#yuR2*-m|<= z332PyAc6a$WnE-`2Ax5HN8iipK&W-J(DQ`^W1p&Q;T8#&iC0^chm>JQVq3Lb{Bh=5 zY&qz_*mBa2gfa}sqC&>9L`B!q`_#YMlZUk;(5^Ox)-6+c*Hg3PU3k71&DQ0X{no{s z8Etm{;P0VVIA!k-CqZ1H)p7lj22IhmsXN(47tg4O<%f1|9yV0GF}93oYKb|TFB!^d zxf@>fIla}BZm7vtmSZa${9hi*(VbdW4#&6PLR>zfhu9A~J-{VeSiX75bXzoOL+E~v zv{es|Tqh>yO>tc7Qm%G2>NCDCapCGsms|DuQ+RF+ERi~c@vupSJGtXfF#$_>E*AgE zSXnjv@cg|LgoC;@8EWPB$%y1Q3Ti>!g(rb}DW9nUWsnwIkI{Qqww?=dQhpW8MN#W? z{n4G~p5|jboKG0o34#xrtb{@x{^=zmj=k*^F@wh}Q)p#=o4SS6mZzFNhkXxYev}9L z&7Lj=KiTY_Lpl|ood7;4Ym)at`@>1UYtZ_!G6g&TA8QX1>r<)Gi$mMcv@8K$@H zMfrM+OfE)H5{pKJhw+0Hj%AFun1>heboe(mK?_n-oS9D_IGU5x#e zYC$=_FZLT@9Re+LP5zAh2qSIwBkQ8yIvF9|#7+ z`MP@<3w2j{qn97*ywbi{qda;$qwSF6ZG0o77$rfQYAbS}q;uNwh&DM|Z}_OhmY3`) zG7cRQ@-d02h8w^tR<*0sMjU&YTyPD+!_nuRyGUVM7hCX%sX%nN>uJCsjd$IRkW9Y_ zu3g&b8@GWK6C(;YdaUxVJ$c0x3-*YP`ZxeS;_`DaTkU4i(km*DG?z)7yQy3GXTS(( zunR7x(f`)vTbz&`F6D}b8qI>eeJr<5=3e(VC&B7>LB6nKuCB{KxNFXh$xiDL$1jRk zwBMJS#g}RK)Z1lRCgphp1Igf@nn!B}`UR&L@xis8!KJy9*&%Cab_?-@_(d!3n> zkEh<%>wm7b6Tx@X7bC_rz$-;i``fDM#PV#;{ID=bDyhl%W5IL7LremwNSO!2Y!qo4 zKLLfxrZx)Wk--x*>6Jx!E}KQoY9Yw>t0qDSq4^XS^a&BcNZ37TwO1Eh>YeQo-AgX1 zy39QrY2(WL;d#_gJ1P2DpqnzjoiuWCP#N=F z;!ML^<3*gG6KuN&lN(m^iZ1o{YuiQ1O$D#0YSObx;Z&D#T)fiSC2x;@Un6Q`CD4x` z8t@);M@^p)fd;)&D!eG+c@+hIr`Y@ThNAKAFD_N?H0HLko}pWLM7M3GDSsk9DeB^d zH66eG;aaT9%sBNC`l(6Nz1pIJZ~vy?epqdsn(lAVd3w-HmuS%IRSwxSS7FbZTu3Yc zu7R;gf0Qoe+=YO7$zhaVLdkzdMr|WZtnKQQsejBcl3x|{ox&1i7WBk))gl*?7H)ZX zGD?E{IWN9|OKx0u@sw*)_`%rvMo{ccGs`r@3*GUu|CCRJOqHu0N^|e*Xj{e9VgOXBLLz= zkwb-`dW2L^dCCRfQ5XXL%z=G~HGubQ)==l-x%q``m!=S65gY!sh(f(|dX?^6>SnHj zKh5kPlG+L>`p+Tb1wjP!3v>0q6yNP8i`N2j=ew41+=o;>%R#m$cYxiZPij$v(AjV- zI`;#HIw^nmkPae9Ia-&j$rbwd= zwFg*BHyns3= zGKQoPjdhT!j?>i9Bu#6%Wq81_`by$?_Fkeep4Q?=rNQOoK>NX&qIKK&iqWHnHWZ9H z4CcD$LmZiyVQ#NU@-_P^1p1^guTD`H0DvMjQOR76U`j}oTOo{Qw4+xlwnXeJ5UIwv z{OSbTK8V4vR+1kW#ygB~_p%*F+)BFL6?tO6&5bs`kpK(c?O44OhWaSqlgY;MSpWK? zmYiV5tR{5}QAQEL^{X{u$_mi5qXwcqidl47TJ!>}5<&OH{h|pk#@itUhR|H;c1D50 z(F^>*NzvnBPCZ3ZFTkj6mAeY^&m!~EH6)^Z(PNL5hOeBCGzCzQ!s;6DNoO)nQ@1(F zF^;aW8uek<0?w}yAcg5C6aswjchz$Nt+_bDA_)bQVKCPv-4&>1L83Tje)8q_@k%X< z%7e0_J7b?=a4tH26D=3Y6vkm7UzNA2T++M_|MqrTi9j*pGEPM1sC>!$>b}p!V~Azd zxOfCC$*gb+Ma3%}Fd@r%$>Nx|--apP%ygYco-_2Pm_QSt`8W`QS_r{@d8w%pVJaS| zk4MyPZK`~dVN`e;VKf-8;Xx$UX!kPcw3T`chMzHrGC�$K1h!1e98>X1zZRbNDQC zfN$z|_9Q<0&-vm4$?S!C(mVwMa#c35Ijpwb!*cX4?^Mjtao@^u9th{%uiNrw4B$>Q zM)c@C=f!%Nui=!vibxo0Mdb%G;7D#$^9=pUdh}PT_>Hq!b&I`^Z`fLO!ara--+hzE zwJ+e%Z`vbL~B@4U(9C^TOizlC}*Bx!Lo9$S#+s9R({> zM^L7>Pi^`Y?TgPDVqA^Q`e}5{l3g)EVqBdmY8C`By`zI->WUmU<3tVcods9pTp8+O zjv>-74pmCmUUal%v?TNx;u~n8&+hN#eWlSYkJ+uB(M30=AEhqDyA;EJS{fq%0FyK^ z7s@O3MX!x^p_Abou8muAI`wccxjS8IVO%aukp>dO)*QiiVTKIeLp62DX+}q8uKXc=5Oque{l$u`JHNbC=-N^3;VLFl`>v7XUKJt1&&K3r1Te8z# z#k9MR;(TO0aRIOX$$o%7?}ctY_5?X-?pKTTCwH3dIr)2=uTK)Auwa9>u}x8}qFDSK zW`DHS#38ko-Bo}E%{59gjz@P}@rqDeix`wkZ>o4c>~IlbbN%_GdG*bqz~xg_g46wn z9c)GfOBsRo;^25cyZDfB{Pn!zM zklW}ABK@}h_Hu^-GK}>!oNDK1X@23&vW9zv`vS`AUIzGt!$R*B1j~xZD0wk~=_sYH z1$Ux~lsQ4|CfY0h+|}X~Mazae8PS=VL85gCDoU^h2B#>vuXig)Ush2V{)sy?BXqxC zgfI24M*SK@u;Wu|e1DRHbd>6E<(`Ci9HC z@e};gMl@FbI@WfDUWKMOaaT7pE_a#F2d2HB);%)5;~(!x*!jQ4Tk(GyJQ9pf92){{ zk8FD3E$yJVuFkh#2^zG>??qzy$|ajw0qqZ=#YVVC(!_K;a@?J4t;!z#R?Q!zSMv^X zE{;?#{U;oO%%BkI=EEqgVNOO|V_W6{$6x96s-xJb%Y#xtu^%aVXFslR*uA26qHNJ2 z%&VVL)uI6h#jtLqE`{o=Jl+6RNu`6TI9{-x*fWI6-IS#Rp>8BL4gbp@YG z6(U7*RZLU*D{*>X!DX)>FTI20ll6|mcaOrNV;}q4S;e0U$J@^~;u6u^m0wqzwr8gM z|KmUq5-s1W=)AYIm)k%y}LqiR`8av3#|o?`lv^=xaxBf zbm!czb&4B^B^=8TgJ8r>;d33V?F#yPLZ8SO!obeo1PX)N4(ZrVg++Nb6TIXlch z2W)EK>q=i$#zIE<9H!g!=0J@^=52iZEYcQA4(VU!N4b0-`7fsQAkcGB2qj?@`b$H` z5iG%LY$IG~{`;3>DO8RbcwA??N_vws%i`f#eC81;v=Up|e2&J` z+682<3LxApiU_N>M@9I#*kNBcbme4M24Po>hr|DPJ(3L5SOvQed+f6#Xk+E3lSP$R>w5EsK#JO&?>Ky6!w zyN^4V8aVph9(bkd8f9(wRJ*`>+Nz_WhjqE;7#HE^xDJOlO(5j6fA^r{L8jLWyZ!tq zT~I5FoS`Q_FTLSKtt`~A+8tx3s4`;DoZ;Wg;-DPsC7pxhB!VesN0b7K)N zV38MfJ@%k~fL64871tdFB$O%tqAt{dtg6o0kvl~_O3Dx1q`0bSt$?@g4E@6Rr+o<(R~8%3qR6DRhuGr1f{>}Zf7_?*aDv#< zqL6SOeUz|LuhS_t&1mjmPN60T-+~unnVsm5iCEjLDu>@@d6S!_+AV;lg~aFpNLEWL zGz=zZG{IpP+OZ6X9s5LxxC|PWf7W7(mQ{0A^yX72CCVwfCjgoeVl_QPhL1UKP9_`2 zB@zsryv)IKf05Gv_xo5e>JQJoY1%^j|KiK4iu0Aloe$OHB>Z>5@;8DuRe^KMdFR`Q z`oHYSzx_TZ4Pv8ab!PZ4dG>GrryG2Jk2MX5 zky-(bIGKNW6#mWIWg>3hiottP`@j5;7P{ocT(N>G6tEco_P8K+gaMqG|1HsfUm`!Y zId_-8OZ`Z4x%KX*v=OKg3xh1JyMWY8LHSeWqj8C_Y#?Z;1BnZG+KVp;f5qagVdUP+ zM12%aAdNR>-0Yg(G%5*~{&Ps${O&CRm&3j|g^pp=eJ|Mo)ks|UW_IKPa2hOvmnSiL zAGx+!u&Od6%YXXI8h}~5KEwO}?4#f6Q&*rlFMtkxtpzwuR!x+-cW{xn*Ua+LrkSdL zDQ98L=Dl-N>O{l{Q5+RWLnkPHD)QXjWL_uZS)EH5&^_pDOmJushL*j$wTR}g!p4Mt zA~-@@4ImfQ0rq9i3-Dcm11Yb+F(E&=teO3Y;WEV5TUXvg%rM{eH>ZRMwW4VQD}r@~ z*BN>b-Ho&)Az^ykJ6yT-8bGwTfU4$Fd?<%Q-th@CM^7&z9{Uh+^`AKXm}(a$5OTHI z+p<2TVAk&OH@qNuS$z>Z20F#g2w21t`iw_ZC~4A23^PDh8fw1WE;Pe=3HUjUpLrY& zhQYsSGJI~`G!ucw)wT!{^FCoKT+CgT_ivWiqqe<1?ES|FF!Yt0P+bzZA?o%9qxMrV z{3yKP?TCigY%buz6Z_7fCk}#l7o(GgbNvxQ^RcDi=`0UGFm*m`23CXugS18Vy~J2@ z_suL;F9U>mMa7r*<=Vg}wWnO5o-APU?qUy!Z`^?;biiP$QdEoo5NN;e_mJGyW4}Jn z1CX=#B92D|m$E!zIR~dRdW~1GRzW_?6B7lwoP} zy^7q2w$xWLH#7&Jn$!ni3C!v^|2C$iXVhl_(#7q|dApfx$J}GyLB`5DVA;uCJ1Tu_ zD(XwW_*g*h?QbLNbhkhKFO_nrP`qBSe(VZieL_}i1R|x9cjYrK0A|H6yyPg{2yQ9} zbzLckP)UI^q;I>VaerH$XY0pzQ{Ln2MPM`%<^*;bhd}Nn}WDqhu@c_#@ zN}BnU@h!6ME-(>I?*SPwA|LXIYD^vl-slE$_@X}sGZCKh;x?=*?Q58`4Np@TV?{`3 z!qj8-A;phYYE?~tqB8*1JTeR_59Jo%rx%6+t>(5MZWbyeimKS~-IInCI6SONo6jCy zUCu0`CjRLgx>sab=DPpp{`&eig?(1Z_5sUC{f;Npo7G_0-%`UoQt*W;Ldunz@? zjv}ax%csO;AWI%lsNwLmB)7kAb@gfBNo|L%gQDhe_k= z{x-N32Q91W56tJ}{0Z+#Bsp}~-=$*_BjwV8sH(@wH1J!Ed#=Y>Z@{eG;dq9{yIU5; zQdFiW{t7`_2DgFTrLdMY#R^jkggp0+bXs`AO61)6g!4o^>w`2R@agFMK+<4N!wDug;GocYNf4mxtja;;Rpq~#^K*IWu6(pvu% z#7rfW@Dmr};!2Wc(`wgk+z-^*Yr{{8vVUV#gi>g=3~X4azK@FlHsw*7mLUG@X?JZ%==gx#Wj?pQ$9q5ao0;EbM z$D?_FC;VDFc6R}{BRMYX`y1* zsS=w%a6lmaX}P_$F!lCHuCGyDU?1V?EN<8?z?4xds(-uasDhC4>ww3NrKpI<6K-1z zTvuEdi3MUx8(`ThUMg~M^8wvFP>e+})`WXZJUH(D`;yk%%d+o>fHK`9(DOVjTb{Z( zUxK{|sPwLkP5;}qvj6MCIUSbF?WSl8#Kn8BL^46Z&mv;r>S)t39PCFe_N;Z|o`{k; z69+p22Z|BuRMCX?BdP#6G+0>b!SUHJ}x`OwGS8yfe(=XMe~F-LyoEr@FbKr z9`=wdcHt_dPFW!~*cXlCU8bacqF*VQ%L3cF$5*zBlpi&Kp5(=ARMWN#oNPlX8Abn zs;Rh+cqrz3bs~}~f#CH)00}_Anx3YM@~EL=ZH)YD!uitN@3^&CFlSa$ezs04SL3zE z9jU@*fkAT_0rm9_jO`viCRzqWf!kHSRT%mZ&rXp+0b7aBj{|Eq&&3j3)QZ?njlN~Q z>s7*}*TmfRMnnq@;7%tUmjGy9UZvUb$l)B5*FmW}YezE%8et)c8cst(8&JnKf$5Pn z0je)^KB3&-kS0WZlP12i%eZZL`laC}{Hoj&;h4=<;JAgaZKy8SXtca85CrQHgsDe8 z@Oc4^Zlwk%QQsfEgEv#$XP(^8dLaI72OYKN&z|BD!Es+vTacl7Rfyd+tg zPn0M3jb8+)cqZ-Y`}#i~X>iDp`V%WTh?eP@Coobuxm3}yDNdz`jk1l4zb@=3`L;;m zm)^Zx&2FodqiWS$N8p$OM&2|e0Xc*xL_)$P{LM6)fKIV2oR5D!O4PrUpCG>DprdO$N|~p$2@4MI)2Rt5@`07qj0VOGm$;VcawBC6PNw^ z4W1FUH~AkJ{DNgSGr)JK zt3Oo}L00Y9LbzS&#C*;hm$=VL8@6B7j_~Yo;g!%C&H!s(S)_8}1IcP3!!H%2c3n98 zAkJK5dza@@tQ+Xni@u`)b0Dp<8XEjdPW0i)q(ZP6$q$)JMo9PLhesbnu^TsjsVMCA zZJ=No-^@y=RWPPFZ0EUIVN3>Q(Tw%#T805oFR6l_NZZH@D}lSQT+* z3E_P!2ybU~QP&N|g-E%yesX2MzR@-aEtvRMA{?I114AUs^F$BQcIba8`jEd0v^y-Jl zc79&a62f}B%DeGJ4fDtOAwf3fi10%)ya>0#TqA^qx+&uq2PQA%DLp9*H_^=jbpd7C zKVIDJxVZjU75OCi_{<3rFW3vSkR3u@dTwjMx&$@i+0inDyo9^tD%k`zha{ z*FXn2M)d9ZD1&Y%|6Zb6NsvYJ$97DbmKEHiZHYqMQeU}v5bD?)FrPb*5Yc*9c{N=( zV|O_p7ta9ZA5smv+!IUdj?LfzQJ}j9euNV=`ng|Z?e$IG! zkKgPwAM^moZmF&`mhLRhmsG*Ux^xNSE$oC`7h)Kl{KRpc669`1!2RVRb^Hha*O!d6 z=-aaZGs+SBL8gS~!8eSb74;CN6-s&X*NLB-MYtAb7zQSrK1uv7*{7(t7o$E8++-VP zNB?ZPJTO1~Lc9m#P&Q34ty0bi;X0|_98_e{I;cDrxhrT|GTb29!o@kq@_~2Ko!%|o zp4GNZGPp6-|A-b)YTPyyEM#3j_q~THd0d#bi&kaT&U7b?@b-rYsXf&<&%b!xQ%m1| z-3#oK{MNugCRr@T-3#M#HO@XJ+t$B$M$sGjd{Q$7{p;%MxL=&!(#!w|pZe3p!PGTJ#pzP3T{EY85-@W0)QJE z=&XU{4DR2Q`X*f|Q{>}2`+8+qnt!eQ3A~R{{8fYR1klFyh-m~j8VgNRE1(m%H^iUN z1XN}gae%b_F4G0=*Ea5(xh1n7Bi6BWm6%CcA5WjnjJGCR)%;H7Ohm7|UVZbbAV(`- z;eEcq{?43uL1bYxYrM?I#GKyk_HFVf{(1bVg&Cumn+8P z!7*s?6g;`$s|?&@ZCJ-&LR(w7#`>OjJ^BLotiI`|%~%Kr2q>#lFTK@#hQmI!N?j5| z9x_nyhsMjQ@0aX*TJR1HK}ff<8Y`gJX2G010pKlv*=PRIY`^{N0&(VxVV>pbKx9!* zw;B5_s%NQh+;V6Utkt3vIHo`UB8co0_1)T-nAp=diY;#^SwgSgg+p4$T?-;fXr?p{ z=l&eyQv`949{Lzi=$dA!SYOE?Q5Y6&*U#vBoO06Zu%HA|E$3CC`uOy5U7z46KEPIV zUmg8LAiI)CuUCHiBGGsQpc@?TdBIf;QeIWFu(*Uk0Gv2><9TCxiQ|2jvdt++vb)<^ zVwiG`dli#=R)D}A?(?<^i$Xm8&`)cBz9wF8F6gP+k5JTfu>a2kqC4;4y@IcL#)WY< z-=P3oZ>xXx%a($RPE+%RtN|@d>O8~xSqg4#W~_WcG;d;~*|}D@H|b6oX|9eqzXK42 zv?>f3o@x72Pg1mAPL=wTo|SxgEvNRyi2#}4I*ZxlEYcs%5!X-C>$~_BZ}_j@Q;FSP zRIAI8X#$G%x9HUPmz3Y+-fDFJAl}1Ox|=VOc-pbyuIInK)}pwx^yV=Sb#O5Z()#O_ z-r`#*1ItSmHVV{9HGKC-B!=5@4aMx1^K`G;1fr~A$9MqJ_!tGm+@6``d|v{{*WW&x zzFa-$=m~b@0jM$!fUBHmsu~k9h$e`A09>QqQPD-J!{(Wcx`7jnC9|E~w@YAF0=n6k zRYeA%R~=`oMid|}xSDHi+wT*i5~476eV{+kX#t3G|7eg!K>>@5hl49;32@~dZ2J<_ zf%m&@Un=CK@m1b|Qp|(*yw9b@sc7sY)OW8pJPBcJTlkh=RU|ut#u4@h!KSQu1Fc+V zR`r!fveX`jtnWs*!*VH3j&~cD%J3YCqy-H@iQO7^Xeo_!SpI<}uvwB}W< zlagsVBSj#0d86;4nAlj>E;?_~R1t1ED8fEG3P1T$7a{f;J($c{eG_+l6k!SJWP_b?UFh zSl7`w{9OZhyOmcZlsj!wM{qIu4%)gRO9BnOlf6PjJqcDd5t<+A61F5EuJ;!shmEJX-{ zufg;fZA*4W8B*A}O>8X7Z3UpygYB_L8B%eV5~Z#=Qp-m95p2-ss=itqwELzfj4uL2 zU2OJ}U3Zp0ciXCO4lxkKJb63&oRs#i#V)o8v_)TFu`hLzaY%nTJV`BSKAYnxz4^Tb zyr8j-vENdUcpqd{9CyC>sQ&3Dy`hZ}UaY_dFk% z@U7P1-hJ5OUu0;2ZB$N?`Y7yq1#1@Uc`RRnyS*z2n2JP z#JU1Tv=ojbErlTN_wB7;CwiP6`D=M2u{SY5F)Ta80I4u`w&sytobw!I?*(GTPcPqv zpScj{R}R*H*v7@_B?ua>5hY9zgr&0uT6IA9pyb|D8>KjL zC|Br*iFK}l_lekCCNsb97YS__d*a#@n*^iq4swPDA64Ywq%1@qO$7!8H53hPaW74sY_q+n zm@SRH7@-y|K$&2wgBJg2f}Hj)FVW9WsUl!UgFjM>JLB>^aYlfhzCk`bvDP_6A@8c? zsq8tShEcR*>ojJ`12_YgiY!~;T50sM#GMF-jU5zTh;OSrnsMJ|&!W*F?2Z)0OtXQN zUXTD2?o(%ZK+tf;Zn#dK_W)N$)Q_QbGX{_hVAhCvTjGcSNmP z0LRJFjh|4D5T>=16D@#)@wFS{p-0#ZyK0rs|9ut!SN^g}?7i@_FCkj#3RAs`TSEum z9#nN-M5;-52gm)@$^QfO1j!wD7-RQX5gURlz&JTV)V%1LF5g`= z-2>EJG7Cc}&ESkDd5JPUVAsF40RndOWs*@0_U}^J#aXe?DSXsjw=4WSo!xH*43F|e zLWCaTr~^h>)cP5AJy0)r={5$;eb7=t{TN}XAc|4V>vzPtuJ#2V0t=@?!tIW>V zdRNkEdju%JTv=@xO?k`FFu|ly%|6(Umjb*!eHU@%2m+sw#~<+V(uBd=L(23s)oAc3 z$wdSyjUZaE!@X2)e>Ie(jFh8cog%9CK7p>QK`#|D;GLOKz-&a=6`&Hf1#aPnoR+g? z2iBk^EZQ50*iT_ce#|n$cPtC4{1tbRUV^4MeeI~g)w<`*Dvg2`9E|cdYXRdHS%B}O zW-0xG{M<)jWfl5nO}=I}6N>rsa!AcXK&B-$q2U94UzQgiAed`-wHkcnQPq9WyLZ6Q z%=T=a9%7(j!?AYt<)A-67&XbF4_SfUvk@~s?zu+WPh(8;3IPYn-Yo;30prOMXx};3 ziL-c*huC>llT|mFQNYy|dw?EjF)zcL_)*a6Apl`^huc|;A3ny3qw)kG9VHj2&pJea7;z<_2Q zXu7Euo*3@}Bg>y3zE(Kb1ds=USAew)1>?@=9Xe(%{6K)V4yDiiWWemcp{m?+Hd3i7 z1;TtAyKM$Am6mHIfFoo*Qlo8rUI!{T!hmfiWUPuNg=j552i2NnTy=H3v2%BDBUkWl^g(O}wj>Bf0cj^}V> zu`+PgY$~ESUY6GXAwIv{DdjDCk%CTV3wjjQUP-F>Nlb^`eW*7uLTL62%mbhI6ZeEcQMb<+!0x`zbP8_f_{25Gn4Tcl3QEB^I6oP@f zmZU-a+6R7s<)s=T)fXlxilt3+oFFb|$Pn>nUnGCl0*KG|5kIp3E~FLh0T}IOAj&C& z6L_VXbb^3}st*wb1RKX|i0^+$zOPOruqcn2mPl}HP^f2}@s(=Xt20uEA)IRwg#53WNUPN( zEr-gJj>K$41b0Uak@Fo7ttk?`GLk?_z?K#4Hbx4=OD;^`-Or3gq-2oz{n|hShDQMV zh?6(WxwL>#+7=ayI%54asKG$DQr{?!1nCN3Q13d{<(}os>Yoe*J5@bJ9)jVJGY%bt z2#UR4;f;Fm59e=zexNg_X945eUQ{o`=w>hgNblRK)BATf0v+!Y`Ojvlf84XEwBB*) zkVYg#N)+|1k9NW-!AZ!66um6~Jk>NNp|wBu*-tn3fetQ9L|up8@&J zc#1yj`Y818XQGT@pl)Jh@q*+fwE*OyNenS-vuA+Q>twwbm@;uDERmyViBjQdwS2*P z3I~(b!6X2ZE7o}f1tn|E?}AU70_EjA8UX=DBH+pv0J)kUH$kF5pRcs(T0zAQLm`vH zXu{>9+0fB1j3Pt6esTjQ_gy-Jv9K_lScg$=QFAqjTo}^{v@C&Db9-t&NxqXqgCL~c zt)5_KVE>|*K5O|Imu;q@IomCYk~_l*0WzL1$Lm(m6+V6Tm2tVzw7m%oTDI0^u4`*< z^WzoUv|N}R^bWj#x6T?yJ5KPiohPxQ z304PR_0&#$sPaOzZme;4z|;F=?|FJawGToF;7mKuYVM;7`>f%5A~z2dmq`k0n~H%U z7&@2@wgE#ytj}A(xOfBT2>Dx6=%L=YK$JQCrXYZBGU1p>guaqPuc<#(J8#`M%n#IU z;jK_7hT8{b3+#JNz?S(qO(4P$ebDYzBA7gDflOrW^{K}G4M4R~RQLcI z)bX>PZ2Fr*7{W9rd=Ywd|F=az)-H8GzDE*qGx+R?KtZSL)!M3hFQZp!h%ueTUf%nQp8Op%mt)M2yKU#g9>EJxu1Dx|e;#)%!xEn17OP_3#jeeH zAciw9a^;%l%nPR?iYj$vmcekMJz{2Plb00OssmXz!1`_nVUlelWKey}wps}<*^sMu z@Y&wy2B5gOAjUMK`iZf(=GVkl2CBT~giiu~CPWARsy%T~^h`k^?Wqz}({cK69xQlG zNoS~4{F#j7lOk(kw6OG2x(mBnx%vs9KncHM9hCCwgb|kg*bCOGL%#PQ_HoDrZHHpd z^~90VA?2hmP$ zt|COwPqNx>(73G0$e1)kIzuHBeVXu6g;(#cy_65h+}5LoDYcS5N2pRGsa(%2SK1(G z+%#9lF%9&TAjk(b9jLl&V9$g2T=kysiGG4)Y_ZD-mzx&&P=pTlLgx&|h)MJucAjC5 z$-Bfk4LBr9!*1wpwX zPTKbeErNKyMA?}qYx8--!@hK?2Xrh%Cz&NZROzr7f!9`V=Xy`V02ndV=y-eRyR|m9 z@tOC?p9jD2A+YswCokB0SI%o!^qdg1L53H0YLdLQ>I&2C*?oDN zWWWt{rqBaks;wOPbBQucG_a1^+-0|dl)e9{}w);;A;M&-Xvp~nGW1)SaW+;OK_jhXH?xKH1$ zct~nEjSaC~ah;oZm}9*Oz_`h3?_o8Qm0NX1z+Vl+TrH$Rk>i^K1*&fBiF?BVP=?M1 z@>hIYiS)bRQ*66c5DOOjI-fZZn3jn((|Jdp-T({_aq-jX`h@+ld*$3ZzxL= z&B_78?V=|?D6H{JuJxmpTVrdy$T?az0_l-?6a7jT>vo7zed+d7HQ67|OZ) zoz`^#$RoxT3WYl&OV)dd=6Hbz)hw9neE^OkIxu>O*TD8fw2E~%2upIKOHjo-Lt+l2 zVpCK=@@TXuxYCquV(Nyk!?d7NICQ2pNJ}yr{QCPH>=iTTZ3q>rv+0Vz=iU5Jo|qEV z99hQuz1!9rk#iokfO62P?nUTriog`qCzi`*Xg3RLHhCNP&OcAHjnC7v`+RW|p{Jen z1@`V$42gGJ8ZB~)E1-QR-6~}KJ?}LF+X>`^GskS(T_)mGXf@h$C$h2@k8LIfElbW% z^_>h>T66SAObyLebaH4ACBCPO_mI9kD6ATIxsf8|i~e$>1%D@HMPTJsg1N=gist~F zh|b^tiZH?Yy(I(%?OOO36}_;Bzc*vVR_;T}Gu|9s*9bz`NyiJcKS~nOag0T`Om!hJ zCts~~#FF>jkxj&sr}&cG--IN(f)qW6u$H%~lkadZ$#1ARC^+g=&X5e5b-49%rea^6 ziYE(AYynE)s`sPwgv`71m1(9oiH*hn(E&KEhA?OLTRW~G6Y8}_#t1SQ@ZlAs~Vp?P? z7k^%;?n3a_|0OQ_6VN^FfyTm@uny0lrU5&?kIP8IH>p;=#4u^bK~|cGVw-c+ zv+}gA!Ee;I&*=}Cg|IQ6z$sM0YPU+5^N*hzif;k$ll$r>Z0N<^g9r<38%3KW3yxQV zb}oK!KK@BjY`x=o^1f$}M91V_F83f^?l*HZc34-GiVK@mS#7}&id!~Ex6``&+d2p) zf)`n^FmNZxfC1GWtkl-%#L>i$X<7txeq76x`gs+F(=^U9RlzD;-yQgi%udXSez8Xm zp87c;3Zc$Aeo@c0qB zo$&b(sl!r_C5Sp#3f(V#!cJ!+rd!wCY z3bUuE+~j@f9qWJKE%GXCD~$S9im9+X#BX6k>m3m#2LJZ5WB9QH`^J9om53VyNgn7< zmD!G*ADbl9$S*)GaW}~oluIsx;?21LgH}P)A&9@iEwY3G&wwO-fw6>5$Ah%T^`(1l z9Vl)>`{`pRN6^^hPTV@}ITi*YA0Og=V^{9{32b=NiNmnSvUauPB9+B1zU~T*0GHmV z!f0(DPQ2W)Gz1HAkap;3{GY%uzs^y5Ol(2edIXeH^siDxLAmg`fJ(%sSz1SW^k2KA z;G9$Q-W-&o;v(_9J7OYPfS2U51_B+$jG|6FiacYfONL4A;}s>v_yk;_9NIs@b2vSr zYIX2@2^Fij@LQGhikaw#NLLJQ?zwuyv5j>Gcuo!3=Zfy6%_hM2tS@E)w<5HCSAKT& z@f*us(6mY$3QiV35;e(FfK|Qt@_&RZ$t3;Og5mCnM3Ho3e{8QeJl&H@)h7xoi#|wo z*jldmYg_bdw%HJ%`95t@1yXJN5$R|N8I>Rky_~e^?y&F@Jr!p0#@*wC!*DY0m1)WJ zMkkRPW5}oI1IgCEa5=-QvZ3je8(CdvvDw(@LtL(Zeh+ZYyAD>d-cV1GP((^ovHp%; zC$TSkuBXy>P7Ztr@bAfuCKdTHiQ?Q>lr2;6K%9xq(LT&=E|*w$jtyz=k8Dqn%?1gp zwqOxIc+;8{tmUM3rW;_s$zm|DUgK#!04S{R64Xh)BL0k5gspEbd9x{}EhC zLxB7z-Ko^=FWvnc5dl+Ozx1lLz4OPoQ!6SQ#Jm}Al_u**C%KA#sDvY&IH)yhg5SO+ zB!2@Yr=#D=I&X2TH+;$#j>(0DjgE6-NGRSsm&aCPO3NYF>%B_Mc!}u`$}FFY9ypE_ z4jmQI?X+S81+oaO&wux~wKm3?r#im4*rEo;{FyWokL%pi@z?&`Z@9&&K6Qb@v}xgr zOTzix;sfl$C0|F6BI{(WP+JK;by5=ys)X>qwtq5xZoD9!Q+xly^R&cIhHV^o7jnluOu!l zE)BJp{jR9gGwXEe1Au<#(45(52f3OP~t+Y|jkUJqygP@=^O7LrRPr z&|drrb(CRFFoMp>htWP&>48xEhj0lVF{~%q%~ddn5q!ouc8LSgEAt_3pMbkzH`x*m>o!@QZ13)&&gx#wDk|NC0)%VAyBNy-9!(nH>%**AVE@@PD+ zsN2XNiJIdU;EE1T&LiNBW|W~BmP%;LvVKZpZNAk%3e`$Yt|f~TRZnKcR&z2V*hRH#b`;Q;`_7UPAl2Nm~HJ4xKbYic!eDO=w}&P0It;`HKh>a$cHr|D2jvITT5 zOjkX(F-r>>)jnf+VXM#Cf$Zx@51I)umB^{?9Zqv^p~D3?ZRCY@VcV4t1XcZzH!!tT z{*GAYrZ%42#;NF&IwM2{6AkTwOd6w(iX;-!8AiLK>OPnQ1K-eEFpWfTIVy4KO0thU zdloqAYC&Y95N6JYb9@Sbqst;dvrnhu+>EZR$ly`v&S?qjo@efv%s?=1?EPY;3=*;8 zUBB55s_)eDn>5e@D1CuTQpnvfgf$7Itw#k=ak&V+{LvvUqLgi)`UiayF&veyl*jO@ zWHE}-&<;RI#C>S2YrHtXk8^yG+LJ-Td`G|r(ay~joYRXq%Qp6okf(iEth*e1+`Pxx z)BVj9Ynjp)lI$&C+V}m1PNPJ!Bj(-vHg1I{g5sW-9YzDo%sVC74c7P`4))kPVxjZN zG5poil7x6AS3y{@aQy7Qn&0SU$?c^!KCHx>rs(sXWBtN^&0{baNO3E)+90FKGMM_j z;|tJ{U%Wy^&e2(?O@}YfiAo1lWhs*G{Nq(9R|k}_K5fPJux8>_)ya8*x$Xgi?(v(( z>biWZ1*We-*kUoSld20st5s@`=H9wHHBise)|qj)ub;<$_BBWKr-c5xtQUUO9p?R6 z)NBtcjZ9cE{4<>&`P$-Tgv9%mU5Ab0#`h}^pU&ZPC5RrAM)9+QKi?prc{JYwl8 z&hL{|pFp!JAiw12J&ocOkv(-&bF|c<2 zqM{wMh3j{#>X3g5=n2y#4|QoWjAcEVWV{IZ7)C{aU9&=_c=hP>6}a~sRA-Ufb-;$@ zaN%9y_;j?hygQJ}M!l}ji6cEQ68n#XC!u1B zlZusY&-J7mRBzgYU4z+_`hOSoA)}^FfHZj@n7(qq+!H zNPqkXLq`)yxgG!)!7uZ(vS}xsleXWp^V>TjowAS51$ku5YSM6#OWcZtZCW)mH5#Tp zS(g8?4fb}qMlRq%Sk^pE1u;hP%AB*U4Bb&VY4 zTVJp^Pi?4{Rv75G=1lT+K?VfS-x#B~5&Y$uY1_%BxgsNvp@qDsiudJk3){nAAs&>h z+Do`^yhYx>Y*T~8WRxpzSUtaWxE?ajAY0rB*O~6M784D$T9Q;;otLN^?pn4%Ec2~* z-YRB#y|um@SEZq)RdK#kF?UP9C}sG-X@f`LEhhgz!$jg{JY|y)3Yk2I3m!F@kD%=@ z;wvn5DSp>sZYyWmiS(!;dv{qM2RaHG|(FO$Ksind?-4wnE|jI31H8T-mXgac z1$iST-l4NW#8rwkGT>T(Oe)fQxNtNAG;a!%f83{Er zXACZ%)Tr!suFO!B_K>^tPqYMe-1438l}*^vg2d|p)e|X@n-$%euT6*RM`x5o`L1|V zAHjvkU`<(NGYfra-n+mxv|~nK)1B)6lifjRC_+s5Z-*oc#rx-hE(Yyy#0B#^e+q!M zu-Lrz@YYj6{xY3yKo$*dk}}U83CKCn(!8sqiD)*BK*@F5*3aDxE%<#8Tc3(x96Ymx&BIdDZ5P!x+SU zK7iaGiv!_;oL3@u zQ4BMBXfzhQ?26yRo=hIIs6mp= zRiYu^Btxm?lRx$P2 zokd^zo+5KASW){>b`r%&P(u2%fsgmhW&}XOX?TiK2UyH`l?(xi9}-J^3i_}u5@mRR zxrCi-t=1m`6Y*}M1M0y&I;~qfVB=b74FB%Wk35MO>)s823Ays`)nCl)KNTg9)0h?w5ms@Qwo=POjjSv!C<6r59PA9T-u$eB})MYfKPMfzCXgCCeS@v-K24yT5|x zA3(#C3`96|(omP$Pugw#LDzc6z-C)g7ae}`Eer+|s1mtDEL7Dg>(ttR_zG32o3P{R z#Ycl-Lx3RsH_1qO-TP+#2)@75`q39(HSyQ^h!IXQ6bqAxO6wCXGJ%dH3?lQ(ngurQ zQ&+-CDmWT7>m$;8n*%Y?QTJPW)bTHCk3F*F+!=yNmX;W~ym)srcYUZ@3rqx|FdF+n z)KE!_6_)jb;MFsx@G0=}edq@XCSU|jvR_8jA%3#H7Ws-&-UVqk!@=-%WZG4I_+Od1 zq~B?}HkOv$1iS6GpXNY+o;>1x=r)jBU8O45QT;FfYU5KbZ=vGg~Z4X z#3iAsZL}P5)78?}f31#NNq*o0;8kbbBln@~V{4#?NW(%NrxKilKZKxk7IhOh-;$$!ug7Vz z?x|u86z9ZF1fmx;CL-AK%D2PV7*lRwcCfWX-XMXFj8*lSBvCGmz}1*!ccKOXWXhDO z5F!W^PO}DV``=Tt6XoSN1RBb`lSU{?B5Z;WhC6$!4n*HS0zw*Mgr7Z|@YjSspQvF^ zq!o<+bM?E%802UdbcpS2aFeAY-SCvKmj`Dm(T8gaCNhn_Q%y8_mh?f<(trDwM&aev zk;+t4&GNi@V$M2`70v(bbX^P`4G1}I%{iLzA?eGtCXp!3o-44`H z4{d{2=sOw%uBqWwhc{S$>$y+fd?ho!oze{JvJ1i-LnhTb zm536rLP@{cA5WU&rT=}%)VZg1I{X_gcu)dSW1s>Pp~bff(CC#MS#-Slbe)ZI2DfU> zg}^cM)6JmdZ>&zPn3O7(IUI<^l@Q7CIu7IbxbLDBMHCWwGRN9A--4s@7< z7^U?SdA(nE@$(m8yS}5dNmOeMoid7K;z1;P4ZvXgp6;N}rpC_H9@`!y5r{;nb3e#0 z3;tiUI>M76%9MXM)RmlS&}`h#A@K%x@SzRE<(yoV`>jF9=qtVMLckr)mH&@xntX{rb?h6zL&Xw5dF7?I@=-~?P1)bUSxAY5z!>#8A zVxhV31?@*khG3vxPs=f85wwN2aZfmBDXsxY0_WkAj`;HaLCu8@46pH*WwQ($UVI2* z6~(bUf%cuQ`8g{PcTxKDJyW1O~8SCJx?W&OEh!%pA|G654Q$9wF z$P&CJjJ|ycWS2-5eq50aBceuFo>Ou%ef1t*W8J;`wut%=kj@ZMIw|u{oqt|q zyX@l@vk5MI(Yyfe&mOdb|L!%jYiK%9=w_OL;um$Zorn3Y@<;W~n(+M!r(slHGz+`L zN7f(jQN4z^FO((bcu0t(K3zN|z zs>X3N`mTBg@6vnIq_{jci7&_Bs)A96|OGsdxBVB@VN+W@kojAJ9YF zCVce{ZtA zv`lsL+z_laP_nq#zE%G|Q0@p8XS15sHm88*nArK29O&R$WW&Idy<4)#){UzKp$R&( z>*{^sjLHn=bgv`_lMOBExI-7lBlot#D~L*OvLx>ZBZbX~xEz12Y-n9O40;gFFw(LI z?md-?R?410FN0Ou7GxK-k_rD;k?|k{qaCL$J48Ix+E*m}Fa`a6^q>p4D1&q0r z;E2ZmEUPh5!q$E%S{=Hi+^SG}#Ghh3z!up>%kwIA$s`ngv~r$?TJLcKg_Zxo1nS-?o>Yx; z6^(LD$92$RqxI5tTz!8U&ZCENhX+^v56cCz(aL@g_?rlD5f#Q>0n+qsUfM0WUZ6US zXnS38P@Vv0$TaPAnwK z8Y==KD=POqmevwZPmEnH;W$t9~LMUiw|Eb}y>&>I3Mjo#t&@3Z_a|VRoVE#XoB*24`E@;x}U2)5Ye0Kv~7Y zXDBKf0`$CEbrDEd*8o2bx8!!2Xg$LlU`g3pqC4pvzf$?;5YdRvd{`|{s=#jml94w3 z45Ju1Z2(@Q!D!M_=-$Kqk4WEQKgzlGGhPW};D8hE7H1YR+)tAq1E7sBSdw`nu(KRQ z7N@$qqtTX+bpAlRPnt=p;|*ubr-ozpS1O75caWr@40K-PiAEnT)p5QqhK-gQep6yx zUA6aC$l?KRS=*l7Sc)P(#g{Cffu~BFkcCL)=#tnK=cv`D$+k7;yJ*c9#*s!;h6We% z&R$y~sL19s+r?OgCBbzX|6t2vw&RoV9EYn|SG53ECNQjGkiE>!%sx06E}3O3ZP$3U zSft>Ni436xQD=2#;*@2{IVQPHg||W2EShDX#u}Vw>9? zH6iV7xNL}uNiwqB9|f>kfhFwivS~8=wqw%j`pq9g^;fvT8S61A5*&csEt32HNY-j= z4vmq6-M%fa=hQ4*Iv=8=OY>{u_mH<=R9_ZUErCY#GzFdEs1crbta+FtfIKxg_C(zt z>)6!N;jkK}h0$(Xxmaf?ma-;RX;>hrE#d4oKI1YH8K1Z58wORNzP~VjWXLt{{8l;4 z{COoJIrLA!@9S2pEGx~lqxqCoF;U8P?Co>Geg3Co(w;|QF(D=Q4cjLD?MHye?y7*2 zJ1X1)2k%htN6~wuQD`=1`@JOlyw9JKp_krO9o29?fP?v^o#(GLbN~#n@bGuxro(9` zJHI7g&0RPbP!hU1_GE8&^=c(sp~zNsUku}3bgZ`VD$kT-_4Ng~y>%C+8wD)rXwQuz zA9S=tcTwrwb!O09_0(V(r}G*e{rk)y_$l(Bu=EmOYQ5=jd;_cX8mCw&Y8{VP()D64 zaj6c8NYxCe7I~`8&D+;;PN8}_^{)VXSln3S6F1%B(v=-&Gr;rs-_w#$UAM3zgjW>E50vb0Smb>nhuQEYgeI(s)2qHKm=WPySI&49QkmO ztLM`L?bp5U&|w_FNmUnREw~O77-^%3d6deux~=67B)IF&(8g*QwLK*3wRxYNkd(6N zGE#|->NMAW_o?U6p6fS7%;whLm>kW-Dj*)R_3_1LFtzO=np zt?1e?Yp@-i@o}tdt?sO*4ZSe$v~AOVM*yUF{?i50WPx;Nv?>0r>x`3L;WLA*gv&rt z{W1`#r&{^wY0w&YLw?*e^R;++_oyijd$G~x{?^33{_>_}T9@deHqPzcyBUip_r*35 z^a~5<#*>C1QO@j334+^;*CN6S5Yx<|Kxce)UAc*m9oEp{(MCom(?Oyw?dO6wxMq-?*y@rb*$pQkeKkKJ8&@YyWJhq$8 zBco);xBqS@Tr9NN47D?Yjdzb-g&HtI+TM;doOM7rz4crH1$ctZ@lHblX@$T)`g@xW zGp-7P0WH4P(c3%)qC%?%rs7m%^RDm!nk%M8)b>hd>jXrJlC%!+P!2^Psqan3uVXLV zAfv*3+BxHOvr%Fr3!<@ecPqOdC`+Mk5E%G_FHRO3-{K1{K^yg?AwbG64bWU3mw+ap z41(Bu%xRZ*fLYt*ww(5PRlj%46vVz-a3 zp3NaXxarHz3qRlBR4kkFY}E! zzUD6F)4F9b)jr?7mwr5mEmr|sIS_+jMIk8)UD&%*)-G!|=Wu2#pL>g^PDJH?JsWmw9s6EnW{4%$9usSS{L4EEF-k7K+%;6?h|;QHC+R#3INkCK);t2-~4B& zp;bkF*TK}yi3+y*%$-QsZNxVcPlFq;Bkgn1xu!5*9)B|5BKwHM1V1%eM;_@C<1U#^)U;Y~ zGCRg*@q2RX$xuLSqU(elqx0Ri`ko@(@R-rN+t=D#O1}oe>GfNg@h?nYc(Q#j5sL#i0McASxnr*dnpfvXg`to3(YV4O`8j zGAsP;NXgxIH(6LTWqd<&MUu5)+$M#uz~WXCn!QTz65U?;5!XSh{3Vr&4j$4uw_zk+ z&^asivME+s37~xk%cnlB*8o4}ZSV|M5$Y$Md&h#5`1K#+DKM<<`}v~pb%tG51ANHW zHA}>*iHEB$hXhp=~G!O$(_g;gig0LP2@3(Kl_{5o$X9CGt(yCc$ zT@O6LOCb5xEC|S9o&>a>P94GWRr#i*Tltx}DO>r&!Sm^ur|RDq$<+#kg`;fS2hA2{ zmmVcaCplc6_pXjeY66JY_zub&a3eEeoFc$3S8aRJ9D_HDFLNuKd?2?`tlk*!Bk;7j z=w(_P{eWlY5x`i#w90Gi9`sOHzgn-}8iO_5fR3GSZ7d;*^T0|(JQXOQG-Z3liQXm` z2n1B1^Nms&^4UJnnucp#@4lN@SyB39$mWx1#!oGq)=e3n1hGJh>^!%fLyT&M3Mvmj zOv=Wlgc_KHtE{c+-eKF7`w`chfb@n_=VnAwxUX|-IZs=R(?CRLDTUUC#pM?03#>Fa z3m`V3$Zt=CHNI$I6hrOE;Cd59t~;@PIBkhigG7@=hjZYoytY{#apE-p0aFI+?hUCk z^ZvqYCP0WmD?P?9d~zMG?e1VOu!8P8qkH;4cWBhj8!%B>ASM#NkhHT2{|Jow{TRG< zsP}ZFx8U}yx<*x%Cq}UAd9ZtW-v~`ob9yS+sRK4E>GV9DkS3uKNr2$ZGgf0Ce3d!($24DsISmm@57afJ>~mlVN$> zR%HhNO2Kp%TLKbQatk0RCOxDj*;Xlv2Qs#a95jV!cQp`6Hn<>Qkcj)lSQUB2$aZI( zVd-#?KAW9%YX?^8k;zO5z@36lGPu(Tx3KjGh0t66Z zL?YN*&$_ZjdSWbD_g)1m!crwcvvIi3V~ni1bw9qoc1`wt!(GbwS7bC|f&T^5yO*jq zY*@VPBEVp0DN-GR^IT`DxL%@rg{Xag(*=w_oG<#4e(QY>^-_1!w*%U9G<+Y5eRZDB zn@K*7K4>5qlD1wV5+q!zz+KQs5h;Ir@J9pdKhIq>81H?&3#Xaj96%=Wu5MNA`=KQq z%t*h5?T5M7B%VjU68VT_@jwSU;>s4_=!2~>d+J8dOfzk;g%9EDW0CWYx=;(5w3yb_ z9FX~SDKr^nU{q{*3Q0+G2`j=*OsdIyg@Gzy!NPmnOlq6%%+U(KKZDk%t9+bee1`fk zFC{~cm}E76F%dsh+E6t5M1>plp+2Cy41=oHOzU)y*eP9wj-CR1i~~uj=8T+XrSx@= zIUPQS=Eb$5W@)s~zT+RLGdo;(mh}5*ZbO3gd1hh9?x4W0JXMT_ZdXZl)G~C~k8j>+ z3n5o_Yn*wD*x3r8pUbm>hh)cP^C8r>t@e@9+n79G-MEh_*C)z~3Vk^kzF-aJrZeat zbj zgL@xsTWT8}-NU~?zULYjgC(NoJW1~Wmpzw}t=3pR-{wuF>6v|rQ-N;MSwqdN)#o^c zXULzs&0zTk!cI-hQN0%YMV`A)GHD;RFkc^%i>_W#U|ETaZm0^+$qQuHS{$Ak{XbegT~XmG4n);zqJTTKbhO z5-2oe4zQc0Kb!5PIMoIoOHv(p?ezIb9XT6oFJ4r+|Fd#9Bh(=cy~1~g1-a!u+o**u z^fi7~UKihjd4TsDC$$M~BWW4O7OpbXXUni^aT{T}xEl4NX^W! zNG`r-s0d?Ru~#8#C*g{Z?v|z)H4OPxce-1Ck$bg{i~-wc!M`$l z6rI0#^z?^{mVhU&C-hxPDeI?aC@X{G2xl#LT3^h?MmjaEzbsRcF*RZ<4i*0M$_#4& zUeGM95h)Wk%{eTek!dX>h{})nV>7`NX?$3=T*E~-gGu*D0`XT-AlkjlD23sJ-xof+ zpAD37?J}k@!J!uHm`sZ*q{mYW-NBi$#m?dwbCQrbmkhP|=-I9&Z+>C)2Is?4gAY3d zIkvFtBzEFNw|cm0R9$D2lRBp>pR}YolvhieUm^)Yw+BR`Ss6SQxV6UTDLjCdGYw#b z@DBrVTC!^*fED^3h$YArqkES>+;dfede6+URbY^(s}Qa5{O6y0w!HT+^d{P0YsSPt zaG%Zx_5pHY0<-b|y8ucrD4Uq^2M=88xGJE^TIAAT4pqiYkImtg`uka< zKg^!kl+f~h%#m=;=KkM26A$m=U-RuXnJ+h&&t(ey%VTMYQUyPgw|LJL0tm!S=rsYqr8Uxzm+e6|2*%^gO*8`9lESSC;4S1e#WbyIwg zpo`j;{d)S{dHc5LBWY+L)$^V#VJQ{o{@VAUw?L;dH%XOq+O?1^foZ>tyQOZY0$*$j zNN6#9S|q0TT=-HEY*}FV5K|(nYMVKy%kZ-qTXETN#s0NqlAU_1_^S6Nv3E1=B6|^u zEeb@rv0%>GvOtO*i3=4&Z*Qjh*`4YyYRPyy*Xa9~@bdxV{l~I|D(&i`<|kH7^Ccwh z;n0v60DI@*+lAWP8IXBH8MBt-YY>W|4`8Nb?(Z-4UM z{r1b0s^F*<$bk*$RHGpu)0-4d4XQz5YIM@#`L$GcmnhxeK4bnGO-GmxdXb^q(F8x1z}x$CCVc=`{g<&vbk|6OD`kb8~&0krh? zn{*Qs#rdUrk@W?D2^9inDDT=z{AXJ)uXmKLk;pV|+fLcqyOvwNc=A|j6u(sdVT}>< z2{HkC6b5@}W$*nW(-ZNXa`6|wb2bN$NRhFmOj9kuk~?BnfrTL6k*10}bjS8^wx8;H zTbqNtm~7RQY}UwxscAHw=D7kX7^GL^s!T!F5DA?jn_0f245mF8+kB8KDu%SsaMx)+ zUCgi;dl4m%S?Vtt3y&qHDyG{0SIio{$vI{J*~E$&Hu;B4I7mmxt>W{%orkaSY9CFz zgS!XH9|y%38YZ_~;n-rnulFCPsrEP}3iwF8*9G8e=;pHpFOsF_7x=PQO(+FnA$+M` ziRdq?;wfXykH3z}IsVvzs_n(~Vajgzuu*fCM$I}3+$e2;u=tLinR~{4$#Uv3W-ZWJ z&?pG#Bs|y_r%#Fsujvg`XX!zix#OP_TTc8* zFOsjb3u_TcaUO+YX8QDh;tMMzG)g!6_Fz4eHvjqUMG+!KC?f7 zFB17Hlx;T6v$BEhg3uxB&wTVJsEH5-hD;e&gXvjp5M8uDely^YoUUsb!`Y_*c!LW> zK$Ecq!-pv7YF2vn7Yjm-W?wXX&X`o*1QO&Juod){J*+&k-W**+z{(>K+oHiP1&TO^ zmt_0kp6uc#>VS%YP1To4)V*$h|6=L^W<;jT=-KK0Fk`Gf7n8=`pP?GMlrBf#Cfk~B zKGlBNxi}oHWEZ;bZ9>VKw(llpAnOkk=KPVyobg{=nu9}4T$sFX3D0=QC8QH5b(MAo zM3HSX=QMST*PbRLc;B_Er`?pJv>sbvYLt6?XE&x3O4^wYw}gW=3-6=KkQ~ad7#$5v zh*h6%7d{l$9U9RZ#mM5wsXRRjv|wwA5V=3~$%;LUlUkaRKVOQyI#$kN0nHucjAu*q z=lx&ZwL`84O38q+aD|seMM~3P15fe#<80l9RtAIt{ZMsd#2PF1j<%$+hD}kMHv)V6&4AId!-ur}ruqof#SErNVI+ZK6)0`S zPCM{e7NhH_Ru-))At)@7x1~8$i+xXTXUz0raEUu%mXFM)lvrxXZ=C8%IS*uS+ih$! zjM)53wQLZ+489-#IrCnwC0XzF9;>%GuH%pdK0u2S0c0h=HVdk@`@qG%-78{b81Wk}N$$GLSD zi6?zY=HRQx=j5%vT1b$2u*$!He2!t2rN-k}bjQ^R5t6E-2KxTuXKB#`ev&p`8Xhe_ zFA^cuM&TmLg~U9Yy-D@#i^Kz{*bA3pAJW|Q6>tV<=J@ z*L!t=F>Wjna%OEDR1{^w10v@e%zp3JeY5%BKDOf#Uz-ubXk3^(*JVBv>n|CmnGV)k zmiGg<-zJ4x_fH9=UKeRogrq?qh&2VSmNl0xLzF{^hw~tpTN&w&bW|tGpL5Q->93JI zFk1s^m+)TpCl1R}d~St4;mBv$`@hY4=6L`gbp!lO{4VK8#vXBN8p0x%0_n$hUe8Or zqA7t`0FD0~@X+5OO9PWcYoG#_(H@5K&Gd3tkq#}SrYJoqoe>uA=S#p#rya8ua40Hi z+~d*`31j=51va-}NUhXh3d%OK_|zv+11tWn9en%#L(m!2vLcc` z9Q46NfjgC)7T-Cc^`a|`OQ~Sk*k^88Txp5wFPu=^ZQEd+!(#QN$Aq7L>ruZ>jfor< z-pncs3H0Rz7-ukVMlE@9h#n2g*&$%Z#`*IA24gZYX`+pqRj>I{*D5 z&sZzlSFb~5EC}c{MB!wcQ{20E!{An-x(a22$e1p}@6p)MpHM1_-_2>#w z?2`!(jYS)5X_a#jwxnQ)S}LPt4K4~lSkXS)82-`Jb$?v4m^Gs)3YbKUZr9XAj?& zVqKUs`6$w*rO={th&HjM6LAbM#A&8+eIT#SuK-KlmT_-Tqm1CsK#Sp|Yt(GI+AsT& z`%sbGlhO5V}WTg*4G8)m@9^!~wJ1je~w|nUWTO)>44=XP0C*b+y zfTlwxbG(3+IOMVOd7r%yzQG(XGXK*!N}#hNjh7(^NhC3Xn-q$Q?62gj?c(iT#qKG3 zAzPHlgJo9JJ*aFQK{W@U>! zRizZK>Uei%$j4=dV1Bjf`t0R<@&kHWC=E=)03vx3adw+RX91?n~td?|Z z$UVNvtIzW}&-3M5BeK}$f5&23@?PWR*$2F@@?@5gHW_7^GbEg^<6^wFa@oIE^7N6< z0Xce>iO=F9(R|(aTFYj3REyJ>hxrnVHI*)kGqUec_v+6vYx{G)a5zLl|FD=%m%H9= zi9a26)LA)U+d(tIHHx#_+uf;+ZOnSuZQQ3J$P12~Pt?(Q%b^;FveNv#;>3c840)Qr zD1Y%G=*f9GfJ}8fE59E5AzL@tUq{`xH_X-2$YR&N=lQ~*wj#CjBu17LMVxeFQZ!p(j9p1h z)}}S2S1gj+@92w?z`%Yc61S0u@>H`LDNK?YbNL=3E~#51G?p7eXP5Qe=L-hKaz{I`4^g8* zqFYi;4lvzhrc?BPw^xGUZPxreA`SkJrn8KS>U+2_As|u;A|Nr8gdp9mAn}tBkdkId z>8_y#L|VEL5G048JEXh28-{LXVBo#{*Lv4{gSBSexpU^+bN1Q$*)h@gvmy@7Wy+1V z@1Vf`mWj%$)7*#*Mlp|PD}I=2K%!9WMkDc+NyYJZoi;Iu_a_Jcw&?r$N!s^4KFsy* zgx))|28*S3_^@VKqbQ|+QE~A)o1r|I>O?=9hiSR5=?W!{qAz=vIR{v(&1TpFW=JXA zk?v6MrA85>>qFaA?wJ=KrgN0j+On)j38-z7A@5e1>3ikw<$TmuEY-6+0>bcGy5EPdH5Dpx@%tOlG;l=0 zI}FQ&e4&c!!GiNl3QZ4)W6$)wKVZA`_o#u`&Hj=c$1@*0c8q5;bF`82Dgn7_#ZP^` zHY3fF6pYD#0db$0`GGNY} zs1Xxs*TZnY9}TT(2_r_ZaSF;a3HIntWHoua{zv-UoOAX+YAZwrTT!Kix3BBSVaALo ze>=iU-Mno;rTB<6iY4fw?mWNoZ0SYIC%<{UmQ|7AD7_EvK|)?X~ESU7LoVE_XQ&ChA7}V z&1gnWLV2VVeZO28Q2Z8E)!alZvsKWac&VHX96^Fx_H0?-o2xr16qxtgQ@K`54DY?- z|Cy)hhyRd(j*`Yc2@cI&Wr<6+uOcB>hO7rdQ676zWFwOyUf8&xyB#u655eqjQg^GV zy~kdUVjvl?<%xq61j?25(~(hpeSw!gA=iU8+vF4718d2w^+|)FBjzH~l_T%xlmVm2 z&9-%j>H6H8)IccYD=p<+*%b3hs@B63bM>@+zpK-i(8D_ijg5)KI(iXq*{G zJte`5UEriY+G(vqvSvF9o?}pqSS+2Zsq`sM2Q{O@TW`V@v)x+y_+!0Z%!u^{jSxKd zH@mi8M7$Ob|HQl7@KgB9+id)7v8WZk=5KFK%E_APGyTQV1y`e4*{pN!O>|$j&&;}# z{xRtGhYh!PuvI?d@ohvBJ>5@vDCAyze)Y1Wj}0~TBhn1#_BnNA;#4B5#i`ShpBi)f3(sU@GR+xE~fPp^$e&DR!gTn z8#XdwAJ)m)9~oENy^Kjs&oA+B!QN-nVZx!1G}^0Rg>HZxK9(D+6*tCq+qn)xrnvY| zF*v3n`E9DY_`P~0hD39iE~-@+)-LkDip`I2fF{J za3pWq?>gSj8usA0GPYGfM%%J7U;XOa_V{K@E8N_+|Mmc5yADj|0P+Sk6byC0GCT31iAjh`H(1$=?Qs)OQ~3~d@h#H$w)hI>^kjM|9;?p zRMNfted_y{rf5sMB-*+qHl6SGmFWr`jCQF$L-z&MM%wdd^9@v})3AV8VZ-7e(nC4z zY+k`~oq%^)q+OzJyFDaM68<;W$w}e^1ufGG_H96Idxsk_`gGJofJDC#&q@EcUY(n$ zG<(%|bEa`$cv{L`=ESG$Cn(>}`hkolC_#0xaJIM6Qw9LdCq1bjU9Gin8%=b1K zJw^nal^!yuZQ);_bpQ4mdj^09$5s%7OjVUmA~elNYbq&@Bc7wt+EM(8!?5u zZ+7TCYnICzv%lf}Z|SHvmn~g}wP)5~m(ePem5c@FZ*Ka=BXm{-NEbQP&qVBWLhL42 zK`U@1U`o)+0P34k9tI6#FmZ*&4vdNRhvt4$XJGs!u>0(=HC?(RHHm5%p4r#dHInWo zFkevWqck7!9CohIpL`H7p0U_qOq(9-QgH1COUm!kz#0znDCAeAK7{ z<=LZE!8`n3_1ul_`gV=TPiSNB(e`j&_WZS_+=4&=lw_piPk|(Aul{Dh@ZRZ>`#7D3 ze=&Rd9Oh{!li8^q<_Q;25EXn{79k1@-qv1y+T>7<)g)L~p1|%o-#ydBz>Wc0i=ZGu zdErvFwXYe+p%?#+NU8p{|90Tj(1w*zRhs7VD!RbCqZ$_ok2}WbbzQb);Frq~Zmi}A zwg3^d+fOu#qY3sqk(5T2=H4to@<+019_YiX5Zkrb-{$urZvlwDTc?-qe%J0(cz{7e z;@y8^CaZ1MhPU%@V5y{eIol&qBLZ+yLzm5bO0M1yMh*qR+|kHyF&=|jh@nUbSH9nD z`#_NXPU(%&0h6=!j2v>i#zpn|hs>ZfWlx^+3BQjupGLRWyv(kAmHLNrC`bq!IThK! zUN$nTm`{{Q7`0Hr^xs5|G8+GjooKr$Pby646T=m^e@R6TsO1zl3!4+5tV=%@iQZ5N zc;?J?xjC%qk?kTuu!rUAmfBhg_)KeXJfP2PH5QTSIlOlT=!qP8M0gTfJAr4@qz!fu zbCI$n&g8#zdj4T7DiPX@yPrQlO;^$r+I69&+s08MKHDLE<$GX4?yrdKGt=Li=6&^@ zS$t1nDv0KSmgsif_Zq%X{`*6q^b^D1asD*zrf0q61t0zRV>;q*(;KpO;ZpyEkGcQ! zx@vHuHHO&#%O4?93|dCiySB)>EN57&iM`pd+7X1TYO86&3!RL|jz2VH4cZ9)w3Etp zBS}r}m0hA!sYn-4%=CX`kUq-SlFxWT-H2TFQmxAhJ(kj2`n=t{9l`zVpx1Q!j>Z`N zzRt5tPxc*CbW|MKXDC$-pq2Os$rVFD)@d4Gs<$Tg)Lhm@&aSdGVyy}{lsvdbRGBqd zy5hD~WIr|M=R6QH7-Gg5kDAfIW&VAsO|#%+GIH<6IUlECIa>w6H~Pl#?pUQ7Dawo0-CHRi+y*cLGuf5=Gmxn@qHfQYrI{8 z%O2n(hPxhYO7?Hz%9mias>({i{yf7*HGw&*I(SiWh2*kvD0RrV>Y+|+xVjd!Rxu*- z1Wvv5$#?!ts*gob6tGxo7>Ip(bq&OdC-LWRgZ*Q_Fc4u&~~VjsMvP{UsJrM zK*)cAz*fv^hs{vpcnDAy|*DpAc16q3gmGwnSmiNJ1u_WoK}+naO+)M67jBEL=R|6~rBeR_b_- z&G&VNueqbHwE{hDM}j6irCd(xYUd#?S_`e{<_Z<3DQ55xtsE0ZI!*S~ji zEAT$mstAq6ad$e%xAFUV(Z1uFD4XFgZfq@6@n^ zI?2cU;IAY$GBtaF8rptVk+U27tMxiHDy~lLgdaE+F}gj1e9WA)1Xp+tm0>=y320b^ z_hd{D-FKM_yMM^4uDnt72b&7?*Q;l!Q020XqMJi-Cu8xO#zDW@>>7bWu&$132b6&| zzh#|24&qZHd*7GmjLcTb%t0CT2)LcNNEBVb&Z%29C>h0i)pORQq`;1GnB)`;#go2S8odK##$f`>P8ZB1dWpf>Uekm&NOVoAG?C$2SJ8i}(p z4_r^P1M*HU6A7P3{fXmYa<9U&QSYg*!1BvWvrn!Vh^LNbN8aR@1_T0Kvpm-wT~AEG zZSvtCg{d%mV;NUo{yxh%tXuA`vHkbtA6JM_;ciX)b2FN{hNmY}lYaU!SErkyLsg{bBHxO|Qug9TBqM@qlCY2^oS_ZF+}!U0lT zs(Zv4Cd2N#WGz@M$zi&Sg9(Y(!0e2#tax@rrwOQ00gZNgCWZP@7ELP@tD(@482=~3 zKqJ@uUZR&3KK0kqX2WMyeUp^$D|VJ*?3M4~?bOUdmw;_wF3v9IMxf&d6Ps-2c{9sM zFw3^waD`)G#aW)!dE+h~IS{+XH8&Q?X0qT&{Wxdt-=%ZbK4oKp4x zj-`EI915$SHQ^%S(2=(R6png=4GpO%r8%SBwq;qbQJnr0)Vwj7%s7?fWL9!kp%S}2 zJ?Qqo2mJUA=2*h2#u6MirkDJj) z6N4XIJ*!9_W%TIWdXxfo-rlqm$|ETwH3qL6yXz8$_>SWF!xMnrsI%fM$8=ZIrILvy zR@-Z7v!5IvIK#01_-M(c<%zz~X!My+NDC}3R)5aVd4gEBzBxb}s;tbBR0yf6zn!)& znvlFGftTspm*^KSg|gpdJ>1<((u72v;gvynMf(+W=6THyPP1Vmt2f3KuLEE*CaU;l zI4rrD$z-DD^_gaukR~k>K!^0gYE@8WqLO|vDKEy0MHf>WK1n!TcSJQL%H8WSZKo$g z4Riaz7*Omm{I(BlcoP0RNPB&76;mn_!!5C)>NV|;1|}@Sg22Kh-oMIDdKqwkL&#PM zoRT>`b9jF`8h72hH_x|x$!$Rc=&zK$|NhMDC@}Ex1d`yjwR>z>!-hC;O{bkEy{6(q z0~JI^1%@zjL_XcP{O6SKr3uH}Kh$g-CdS`hgUARiNh%1AksB$N(Y8Yi0yI6qty`1F zYe^QS^qI2rx00sVr(e)rAz$*bi}=(vKDV7VxWy$jMuz znR!hGD>E4sXeE5@bQ-aICZ|)shMlE1Eu&6s1q=!FaFPPnaoIHRYWL>SJLeZqO97M@ zY>Fgkk3_Op=DDWj(7H-sNJIiZkLL1e;v`F`#1aR+BJjDsofmI+8!?^H>~JH zvw@k_&XOu=WyLfAV>YWU#;O6dnB@=0?qyh``d)f$vL%Fr)y7tw2E93)#WI-bve@j) zdGH+ZHpnEO&vXnu!kE|I4+*Yk?mH#!kucjOxmfqlZISU>=N}BWB7^o_&SS*BVmkSd zZuvQ%F!HOw_=G0a*0V?|UUMqJDAxUh(5)(qzyhi!18Q#r-4mc~DzUXa^6-u2aT*Kz znFk5hC}Et1Sgh=yiA84t|JFUfL^4q$XS5m;@KC{)KCJf_HLNhx{-43s0^7N8rM~+j z{AQU~Qpj=iJKz4c#$fH_;PznVH+-xyaaE@O0K_ zn_pEc1|5agOGe3F`;3DsdEH;cB&-3-S+)m{73cf*}ybz6BA|TnT?-O$Rf;xx-paRoKO)1+;rrtQ4woFw0EspH^bFl zZG)nT&jK1sruGf48}$AhcP%q)2xS#ziS~jK!~A_c;C{_YJ0%|N6w9wQVx{5ZM>Ewl zdOL1aN8S0ou0GmCp=^~*c6cX4TJQbvhi*)frFyTK|s8y z5(!wxM>nwClJC16iO*4G+`o9XfJd)cD%l%$S~(18Q^UwV6OH>aodUO>CHKEkkf{mV zhF8i3H8ggudo}Rfhnq0Pf~!bOWTfvtr!Y-*$s|$AU*-A@Qk&HP0XXsDv{_KIx%hLi zCyCHiFb)MHgce2bfH5lQ&vRh~TUZ^8I0NLISA>>=^fmR!C)2(snEVg-#}5*dprX_J zpiUcEV-jh}HqE>J)_d#Ae`GD?786#XK1n!K(evHTBTolYxg8;P{*a$vI^mnCap`Jb z+ZS12IE$%i$7kdW?v9P_(grba0Sjh#RUK^=-qX`Fz;=aNoRr>FiBP$D1VZ z8!=ibHbwH8I8*K99q-GJ3|K=-Sk8I2@!BbJI$b>77<)kJU}lX!#hb{E2C_z&FW~Ca zcm#^KV1e5{GGcK{vcEMYAHKCcv)7T}Dr)Bl2 zc}9j=I>}Z#VbdVrB2y2x@ruM{P>RIC3$ITg`CdGIGe-iZ!k!5czPuo-rh{e7*VP%D z8RQCq;T~`y&>)dEEUjea5^y6YhXp$T5n&7J_bQXrTqHMwEvJ#B&`(yM_LHq6)HKHz zWg`}qVkHL1c{gDlN%@A)viOV+IF+E7j}fxwGH!$2>%XoIa#H=1rC?%oPLRt5-+)a~ zbp`!R!`pW}nq|pI!o7T=*}VA&$J>k76E!$=3v8cKtZYp_N+b56sb2adY18?@f>(=* z*G+y_D=^{aX}gy#*I|_CQ60SX;kNZ$xW<|-#BL2>*FlxVPCIQva|xIpCJQ90k+;DA z7vm+sN$9rR2$P_-4*~xI2ysS9e(lpBOj44)1kUl#$^JE8vvAi0H`>e-v#=3|Bt&!bEUf88JRO4ZUP~E&lEe*q>72 z6ldNN)k^&}@TE+B;?B_R;u%npVsmWla?#VfDLmq-e6RS#c5TY!tOwXL5o`|fNs%u9 zEqD1*4<%$))?+0x(d`#|)x$GlMH1!GZ9n?)^&H*m_b(}`_BuAAWu7;mRaA#B!oB** z@wQo+pQ?7d5kuX`F~q1Uo5GmZW@zthL;zmfnH&tO8-? zz;f~?t7wPnHjzh%PY=eU+J*HrhyQ>bigpNb?S$qvwrF`$Jkby;Mf8t9$m$8q+t60Lj7=0}~vU+@@N%=jB z&p%bPhOIJ|^@4r%!a?in&X~ndoASS`P^8T*lR|e)!h_3WYtB?1e>D)^f%6_4SDW27m+Em_SxXY7}cNE1q&ee?RTd}j(u9lbNoR0A@TXbB4(axI|e zqx??&M%?Fkiji7C!YZK9WAF0W99`dp2)U)_Q11RC3gcGIws)U|U%7 zouZza+A&DFwR;}y+Z~Ao?7Yaki8@y)A6qRo&!ErwUlX2e6g?g#U6zN6+wcjdZ|DPN z@B#;X|69ZcK$Wm4)rt^zXFA}wC<3`6$XZTf;eQ*}E%#k4SVHrMpx+RVq!eV?>eYev7^rIcS)_m10| z{w9im@)h4}h!2n+l_>EYx^3V}{k|@nJt&nHr`eA-LZ_0kwYdS?GMTT%%2scykI|E| zOy~P*MD`Lkc)k;E2=vR2T*v>e%~Ft!?_lc*)DA>uSm;oDE9-KHL%Pv+3<>_Q+-QSG%XP2k0`z{OUVo0HFDDH^L*j2Pi+7*{`SEq@l&ETGkG1;zO7uX zx>n12i2Ma6d-jyhrqKt91MU3G;8rh#nHA#;ycUt|dI?Y0P#C7t^g<%H9Cb39 z|H=GJi`#tQUXyIdEx+l`65@z4e=; z&;rg(uH}?3l;%F?v!|A~FFsHod*J+Squ)VWq0ZP_$W?;LingJ^nP3!mvM05(P(aP) z6>owkaQ}UPT@G`n#Fq}cHq3!Nxxwfn!xD}1UU@q9u>AgCinFK&v9X50g0v5B!DLJ^ zn11?mYiX5(YW2wiuJvlud~JUI+`Xw!U+3cYg3npy4C*C+E7^$dwzTcF9Ue80WgRWD zjHS!(p;v7UNh>fl{cEMSWjuREwCG&??H02mBpyI)HePV-6#mYhJLoDfj17WbsQVqH zKJAbHsD6Kwb-$MtGMo)q!laLv(CkK41BU~lM|yQlmqziKAY&w;WG4~8Y73Nqcn+j? zLN?S_Lu+rydBR20pCxy7f@#i^;VB~!b31I*Bs%x!?265w+z-;AeW&%=Fl;9#tAqWe zYB7TXU-98REzQtt?yuhW)VLmz-m6O>)dZMK5B`G9nQ)}1rI*(!MbXQJ-S{r@`edy zT1c1x-8wzoJ5{)OIIriZ$*ewNQ(3G3rbIl={S|v{O%er(ejUS30lYnWCbSE>8&#Vvs{wxUlmq5xbCRN-C>dCKmY6=&4Wt-Zo6tmr)2Vb+L2}-$N8OY&9 zS#fp0T{h>}zh{kvRUOvaUgG2^(MHo)iMGL~CMhz~w=$0%n=>vi&+?=Ggw}T}0*>Z1 zH^$+O)SQ%ClcOBre9)cq3|sq*_dl_w%881N&^`dItTpbwGg6;lIovzbKeK!B)~w+D zp~KfD7D5k;gQlY$bGxR~Za5II>}^RNBA9jMYHucMz3C#g_n_m1aGm|Eq4_EY2o4=j z-7E_8?)v#Sok~=@k2da7ojzMw6SP-opB-%z}N)QE>AAMRZBsxzh@W6rn4S6L5;tf{rVndjV^ zB7h0Qn=zz_Q=@}ZCNq%7J&>-3Mv*dZXM9)$Hq%e~?yln>UeMnrgsK2lyz;YL`Bov# z`*ls10yZ22<016vm;R9}kl=1*PYbQ-dP&JrWPic(!n5(6Pfu2I=`S%iuYo{zbM2^{ zY;VM=;D4UN%ei*i*h>*_h-*ki`)`VacGSv-GO?0?Hq; z3Z$I%fZjJ{oq|{ub!X^o{FVJhTLxP^M5l8$V8wjPHKVkKSlxkH3 zMt7Wztgy6Ml3VTT0j2Ap+sbf$9L+VqUlh;^Ie#}eE>hpAgJ7!o;X$Pb4g<=njvtWF z2l%JsU@`-3pG^kP9+o~us8a9Ga<;Zn3*{#(-YxtIB*_^_CL1s7A44*Th(bwE4$^1m zZ*CaAA5?|wuFxhbUJQi34j=m8=)k58>iC|My2E$QfyQm1{1tJ(2!O+3QN^{5sc2M^ zw9;Hpu;`xR7?q?a!+*>##Ua!~ZT9^ktQJ%(&I{Vi=4*FA@U_}P+mZ}Qac+=nH5f(a zAIwM?LCc1%I(SX@c(Ee20lry^+%)q38}uN)E9O3qu_Utqux{7zOkuD&hf0qyOHBGa zpaRY3x+E=u#s;FTxBCY;%8pw0hvC41&2gm0X0Q!;Blf*4E87t$uIb~eQ5loKa8yLQ z*6?ovBVB0R7J1nxO87kXxC#iCJHBfzfDIm{>p#dkdr@0zs5FXczsihDiLb(-O_U$i zCvc@mF`6g>m>sQ&E3KuCuCW6UmO*{zxh9X%XJ_7B_wgbF^|26c%rsP$R82K&+x&)m zQFwqB{J5of!%twvoD-e9X#JDb8gCIuVWhm=et}x%T7aAnu}Ngr0_>S>x0Cmh5v2tx zw&m1wV$9D{6iB{rO<*FNB~##0_)pQ!P^uV}f(6yBC#x4$MFfp$SnQO75a+Y@+CTMz zEy<2Wl2KXal+`7~jn1Q__%6+?DJ3e{y-Ca|>?(ZMv{Q2WYXrIYdmj*An>+0C-A%lC zkYIq1kZ=d=|K8j*M4k5fkJgXA?25!R48`p3)4qN@RrcDTT&a5;Jje*%;lImgiM6Km zDG?;;3_`0I7zHMN3#&#UeM;SEK5t^>rj6(l;~pbrsy?UThE-lACqUjsJr>zcD0=Yu zu|cmH3lGolrY+g>RlaEI)Lv-iO_??E7kain_6@>d9Zf4(>suc;+6Qv2fB$eF&Pgi4 z;f7ZcsN!(W8~qv$34D|HVG;7Z1O6+2F9qi1DI&y@@r|WZvd>2g2W8a-0SyXZXsvlTG6U%iRcTJ&rTar~D|%l~I&0yo_*D3&ygy6i8@ z3P&39K+4@9;oziy3vw(x2S3`o1zq&QQOr?0X;05K(pS=TQFA1QF+>!(v&Lp`+}NGP z-kxN_!>)&%{zHCWma?vD7cV5XQNpQezD+cndY7~~t;01=o8(>1sw<{3ZgW&~feAuG z$wHHRepR&_xv`uR*lY%%w$>`{ej!!+`@LW33607UTJmNSG}CiShOa`1AloSxnt+4D z0(?2UiAbj{1g(0nT-DT?ffU6B0zC!ptmX|aY}|7Oj%$f0bz`SxPjPBZRuj-{@UMJM z>#jAt^8^LEDKgKeYc&l~?bqiwovn1Bt?wBY2Wr?6U(MMj7MY~!eDStHl%OhNEuera;OiR?LHSqM7sL?Or(X0A6CtB9XtKP9_$>awCYAd%1B6`8Z zC`a6uVMi&r<+|x04;5aVCN}YZVqtBjRJp;qbXHBDBiz{ULPlJ-sA2AZA{n~zL#X8Nw;fNZRRSud>yAStcp4EkMh|ub@Yp48tI%_qf#J2tZ%~4Gb!^D#>C)G+- zobj)0h`qwoN41D6Fj7-3CvWpOtfZ@RYjSBCpXDEof8s2qxvuLeC3UIvN#m{p0T5On z1xVi~1GVNI@#DQY4dlT(2tm60#DyG(sXM~+2At^#)DKypLvE9gX`^||rB?x5S=o)U z=mrD40xXlZSmx`7QSPbAWfpRthJy&L8yBqAVpk!mUTPoXSSM2?5J>*s6carwwxNl5 zcVoWqiGG3A{Jubbm%zo9c8{uJO%ksq-uEhY3a(|@&O1tB0^vc^AfS)rxas+% zRuPy=@LiT{vWDGPAub$3BhhkFBv>VEYjbH$7RAwcU{VR|dMw?%e~Esd+ya)Y-lE#i zN~@s$YEPl)&tLFmmHBDOz}Y7rz>n$<$MGNzaO`E`Fo9+`?*BU8s7&{? z-)%_VgIxg{|E3m87NFhKX1eyK-pk^TC3eB2Xl4PTcrj2OE4BxdQHB0x$d#na3Y@_1@6Wdnifa&0jQDFpIWmn?A-S1H*6&lq}fCE zIplwXluau`sLO>h7PyC=7_tt5F3X4SvWWT?S*AK zYXp%aO1`gH9T0U<6Ic_q--8HN&zNKIJk}B(i8=%A2n5q)bh*a;djJc&?|ndSl*rrT zwhT14M2-?1oJUm>NNOQ9fYvi_4WQZa!b{J%sz69mT_tBpDpZKcSj%LIu_Ww@oCF=>p|VGi2Vw;Uu!~tqvH@~`K5;xI8O{+)m{vw$f1{G!~4y$qE_6jd=KQkNH>56>E0cAApM?XnXoLaIX?6+Xz?!Zb!%^*QcIG5ESy zv5;@rh2gD51fcTfP;i+2O*dtcF2@lhF5b7OW0hFp|5*Skit{P7|Hy3jxh}r#jsKZ? zjBO4!0(W|dtT00S05+@`j*A&2#HPs~+oa?i|W9&u?_dya>b$MQTWL8|^x}`%&nTXrk&|kCd<}F^n&)9ByIygg7!pJ~7T*}2;i7D*s z7HH^3nBcBhRZiBdHkv2-=W}xWXr?J(DAl8hIox%M?KmBf2WB~z3~tSDOop@=d*dLO z`bUu?^mLLZ|5e}#?j3q5`I^pETv){#ZT(&SXMmKi*z!2O0)d*WcLX_ zRJq4Xi2QnTV7yk+EMw>GhXchU^ojwZ59<17eZ@(k4Fm#IetwJkB<(>BlG7oIfs&aI zZe7bywLQ8&w&LsGkjnXCS<3xZP@E*^0&X_d0UL_OqP`Rw5o6!#I}qKQb=fSJHUf@l zOm-OzqP|`ixG-k=db7R%&0tGjg+#01xVam&@eOk+COE$8<^A8q~0ZG{Dh` zon(MAG%LqLh`#~f53)}t$TtVf*pKGHGohUse?L+SiZX`nbTHpAftVBQC4LyC|K+{V zrTzu7ZC7BQ@c#o^c?aIXh7OhHgDzz!hRIhjP4iG6Wc!P4#Z3-`d~*!w5~;{X&fvWz zwpBzwX>7U9K4>yZE#t!_wnTz4Tr62sm?`!}P3Bn*`pfsPh{|x=u4v6pdnQjY^PdYGNV-kKvU8$;RyT@CF9z|{WvUh=q ziaTym;R!O#piuXUQt@rM01i1>6ovJrs!?|fcmS!;-n1-U!*B|b{>|Fy0?6xn^U(baM_ zkybONR}h*}Ry>R7cQ>4%rzg1ZiVG9cgeH(`H`{%}sb(7&ij>$KU%KoUE%q#M(tQfz zqLvsOC3N=mP*=fhv8~rzYwL7Gqs$Q>+g8$JRuA1RXqm0Qs=Fi`5w*RKchgm-7h-}P z_p{7j30zV%Xr9*R^R;+!flnMt3*d!H4^0`7_L z1XCn&9|eTix%#-j;UuK>*tR=Ty%%%~?L5K&W(gFZJlHQ*f3u76$~(r#2@*uHU`xFc zv|CA{hLrca-Mgke;w`tn(8SMRGxr|M6MjPz2xx6;Jt7tFl~Jaf$<*XoI1Y3@>c?sd zit3Ve33t6uj;q4jwE#L5+h-vl{NhVBdaqRE(^In%oy&P-u|E^--670&yW(JldQK#x zmERdHrUZc@U!)1fQF*|2RAsx)huktaV{>`4zw(|EY7c2e(R?Ah^Q7PdMtLNOAQeUqhck{D1I@5Lj)z7M2|Rw2uW%(f2)vtcFI!JnpiuJ$ zlZtgqEf{k`=(($vn(GhW7amvuhHtE>YBZ|S(2K*wn<|Ry9*Bb1R*-DfDJ$kSU}oh(O8VmM#Y^CC#_s@6W#@1xdSzu zMFI;pmEZb)y*YnV;Ci?f4EaVKfJy)3o*Ns!(RNR?E6~xZ=pM+2lHTrUJ^Xu|t+!rR zWiwedxaAwGJnwzL=u5>^;k;+jOn~A`mIWQw9*SKB{Mo|mdTihZY{lYl4MXmPg2Zx& zHesr9&&IG9jvey zx2k=LucyxH4J_ghMHpBk3gQn(SGq`EvlK)tA;2>-z$YFWm07Ypxdb{!sK55e`WPp8 zPF_<_p$Y_lBdi$NMudOF5jW9Yx-I}6X@)^J^TV}{Ma;V+ywQ`uT z-bmPGptWO&bpBnq5LwShJs{9)F)2VUn*s1D!F|vV7e>NdR@NqU6R_hr6j!{&x;NG> zJ`GunCQBx~ukN#?yU2b;>TSxQRanFs`KN%HQwrnr4Z|Zt>3fft);Y40qFj=y@zcSW zhnxP&cyK;)SV_hgsHDbaS0CfGE2{p>Fx3;B^wA=%4I25(NM=>&ZmrKdBtPZMbeVIB zBF_pr=t2dbjH$lW`cr*;L!31kEQUV-qu>{5@&nT4Hx_BLE&m2OS!1^{1{Ws*++))Il z8IFm$&_1R}g1K|1FU{jC4gowPDT0&(xWSkym`LW+yo%s^3i8xRiA3M(ADH@}baxgq zzeW1?!Hjqy&U+@FZw!gEI%0qK4iD!U&h|h+PVM>D?7H>wKMyr;i9J->mwKiDT7ME` zzEmKDfk>K0gRHrCMPxV1YU#v0qfBgW-5h>RI6!-IDB9>>5;hcf_l3)WdB}I zh5#0ke*srcQq)_w-#2>>MS~Bw4mMQ<_8OUTI$x<9f?GfNUI1Eb!Pq$@{kS*U?j1IZ zfCsmH{RPY+Tc657>5B>Z;jd4+g()X>*3pM?@Mua@dnt|k3@-^8ksijz$BJ|w`kf|1 zDIOvYt~n|Pc11VNo=SIR9{S9(+C8h~{>Heo^ip&t+&Kq+ZA6(bk);FbK z9f$GWRn4FOrmruuJ~W_yUf9l&?p1%$FqxKnxGSq0tuK=8o2R9^hI)+S3-up{GnEgcG9?ucdEM+-P4^^l<(!V5L^i zfs5+L8Ra19fJW+i+xVYP|I-Fm?*DN5SUtkKIexr6H6af=A7mWHe0=K*V-iN43_l#Jk9!M9d)|W@S;Q3`H5OCsTx#_<_qoOqWwKZvM?}PU-jmj3o z_8EROW`C!sM?{4DU{=knNjMx(!KxM7->srKVdON44U=w-XgjZF978Nz0o2YHfGL9d z@5Hs_O+DB8W1Q;?V2FCq1%fv!UmZ^P=kfm*EfDsfG$X;>TUfFk0S$^`Zc0^Cn!1-y zSvfN6)fXfsZ>BpB3uC@etC#l{Y7;uag@Q5g7auphoB&z+>UrZ2*7sToMBzTJ zNugYiwZ@OJBJ+%9Ai~Ad0zA#m%VL1~v+{<${3^DZwqBv^dl|*$jo24zc6$V1#Y!ABaJ130z zl%Cw&yv0HQz-OcIg1(o|B)F%dRu}#qLcT)Cl9hE#&oU z(d|IrgqiD->r8bXS$GisC9}yH4JRQ-ya7i%X5cuR>2oJ)Hz`IPP@h*{+o$(~W zdeF@lr}`W{WqrVTUI0Xf_(n23jKT=ahg|?ayJ>_$(yFyHkBqr$>tv%L3a&Lpl6msI zsWN@ZhRIaxkwj>8>VQW1Vufj}MRdnpP%GfdL5i|0{bJw-Jf7)&lp;Aqj)F;~RA3A_ zqNT73ZawAq^V{YuzGW2Ol!J}mR!_m>>v^#T@Lm%31^|cegwwk$j5cN3(!pRlDcS! zSRwC{QD)g$*+fP-_UPDK$DYS_9LGN6clms8zyIAh=W@NS>p32e`y(O8qKAQ>K1*{o zG&${l9dLbG>bW^BC|}0}Vq0l3K9z1M`?mc)*A8}YAIxOzwa;T<+EAehjg`v76MI6% z#~R=uQ4*y*FSR6y$n%-U3upMi{~@GrW`Ewd0*shS|6!^Q5?h&n(lH0IGjQHv9RI{7 z`paur4fgA=&?yi5)tO#;7m!%6(0eK$&7N0b9sUcnj8m=cOW~c(AL+Ht@UdEiXEx(z zuReTEd6Cr&zNhZ^jp9c;=w!CD9e!D>m(rp0PHbaT(A<6hMFyJ}R$l!N6a~Wi9z&Vk zfYmncfs8(JrKJGNA~Q@(FX*L_W?tW+&@Zh0Rhs-|K?Qd^ zjfKPrx|IE^yx*4JA6;DbL*5eLs??om3sKwpyjT}KgvtWh-GT*YR&mKkY}?cs{7x_1 zcFRO@cYIS0ZoWNLbxV(r5AHh=$ezE0D5Y1{b7uk8HN$@H``tOb96hwtMI|i*=|eE$ z9o#KS^vi9p&|xnM?vnY3);;v!RIs5eB^Ma=xps&56;rnFMZlJ8sUhnx{#PP&b^^7;V^dDoBFvX?LNujEe3rorEwFOyGkwGJm9RZ3$gV? z!K;UR65U2i%b#oc@p;mETiT-1UM&3Nq@}?rXdQWo#T_f<$qAEGi%z>UnCYkNaHEj>iFXb##{CbbU@;Ah=`079Wrpi?!!9L1d zCzv-nIb)QEg(sS5j?eyeoR4>hZzuhe8G@h~zI8r$3&3y#)Ld5z0d9l8Fg3V$#(NLhK|-+Hv`tS~4u&FL8zET{_xIL2m|+1IE9i)mb<8Lq z`LzH{V^a!7dOL7tN)OqqbH^$`MN3Gyy`j zNa4#l_kLV_r-DKC)pA-Q*i%4cEA*-%NLrRzl0!&fZO@2tM{BjMbS=bXNgSxY)V*Jd zQoN;*qnh6(FrBO6m}Jl|KBRl>c5!DNKKLT{TD>DNzQ99`yEMI<-=*OTZHv@VtG%QH z5K9t3x@DwVZfbkN^Hgj7x>62gt@Za7zm)V3CePckmJFY1X)Rf^z1rUc@6!v-_ak$g z7N?C@OAA6QS)41*j^9pv$^)$S0M`o_@^5Py^T;C(tLIjT{mw026GgWeo|6mQe|Hiw z^-;Z&9Ywt42UJFtREP#R7Toe9^EvQW)f~S>o)xBz=~qBwd=(^Dya93V^9vurs^89# z#r0OMk>h4EF(ZSenIk@!*1s%BKvan5rO@?tPF8;*Zq!FLG29RFmfScu-dp^yI<+zI zh3khNIK_W+e5y;=Vt@r0;fH(LtDSM2mWY2t0G<_ZNO z7H~epL*!DF9A16<6#%3JNWpf)RVLUExC(HBNnxyMo^AQJS+_xONqV9N=nm?Ou_Zz0 z@8P6P=8oin1yPtqnxQ(33iz0N)9R12e>`JO^B&P*Ktkj+r=$B@`0Yh2uAx0$m!w`~ zWgJnY#y`INVUThM?59D~!|=~;G%+tl4t;*H5mKb+ZTx;CP|9mhW)Lis>;(;t@S@;I zWs|Le1C!1Wy7h*YEwo07Ii0bU5^Y=3l8|#_BI+4mh&`iC=>9?ouKU?7dv*dUfiF&2 zBv^{(i)YXwEA1mikqpf_k1wLA%utW4%tpp35s1Tz{&MB7Uu6reo@y2!RUYy6&}5uH z4Qmh)M!{lXm#eb@&VaAalwIWNFPx*NViqq=&NOjb(q~iBKGMExU_g1#QU$mf1|WuN z^?&p;Iuw)iLA2WQG7cYN+UsPpNDXK1{pPhI16) z88YdvXC)Wi=jqc0ne#(_7g1b4jZX&+} z=OYK0az)qvzri41AD}!w4!V`X3*HS`n?wg6Rb~EeII!2U8hf-FIy0v4O}+bKuIHpDk1d2z?j; zKBOk8zNH7~#@;Jt@hm`&*lY@;SC@K19&*7tllGOA&z}FU^7incNDWmP>au68xMoXH zVkI0Km`(DC4Y2$TwBColJ?AE2ZYdY-gN5a19#6O5`!iZ*Qhpk~Xs^ww;do-P{A ztjCS!;N+oQQBKKCn@RYAKv?mf{so4rUEa0=(?9)Vm*Jdu|EQSPeyq!ChyCJ|Z@whKLj-&gILxIE) zzsxaCDQBZ~sg40Wj%f>c6>jrLjz<&%H{fyxprK_KdZ`ry_(BvsC0k+*q3UC%dbwJm z>uOWlPvht?-kf$@$6s%_Q13l1-4o>B>hWT3fnV+lqe4A>^To?*qB5rO7fm-9cZ@*k zVvPrwINDvpWTDr;*)GJzePmUW#AWw+wS@TBC4dHs+bvF+_I}5%2ws(pWo^5_%&|cP zT92z||E}k@UseM*uXHCWa;16M5fyW9Z#Gq*UZ{p)w!y|HS+ zXUSqMXL;>3OZZHu(qYO9-4U5QMhVPwjP?RT5FBxj_l|i93`!{+1ccUr(YQ-qo zXwZy_Wd655Oo(HJNbY|B&*zKVbW&{tjc}$%-N-{~8)IrtXoh->VF(vN^~-;dZT0jI zSM(&4q+2@qK6#)-3Ra{yA6MIVtA+Y{v`A$wYUZzqC!_14x@&sIeLr*K(qco!B*Dx* zUb_pq8sDC-wO{vs@zha8zNWVS5kw*H4&!g= zh01}(>@+tr-UFSbUGoeyTsAWC4$YAQ2KvH;<|El88R!gUaZEe?bP6U#s{O$g#z-ok z48hdXf%NJ@TR&vk3*;}QLo(5G;`l9xNkOEO=;ZU-+2@Ev2HcKCXj90%hR4Raswo$F0d4qJs@jR(M zFuY9u^5w`<(TpEsm;`Uyv-Z;WY2B$cF8<1Iomf_RyEJNd*Hxe?CaT{bK8py`;`flAo(ngKxP_=sVa`_DFNI^GunyM%-#WL{m z(Z*5ID?Vz<72C{bmZ*uiKVZK&SeB7m*5vOW4r&u^ccFO87o;4j9Z6s=@tr(^N;(-_ zda25h0EzYpnO#-&umOrf>lCI>oZX+4qB3hg2S`41KuGd&%7+{&peY7KC2um{W55C+ zh-%}qRM$Vwu|vHBO+S()zD0Pbk=_q;y`q|@-0E?=Xop{9{D{lKb5p2b26>CwGXDLmw`Z*}uE1oCD)06t}DZzdhRK&zZDk1x}-Q#w#-vF{?HD=+VcLuj1Bz%6`Uu6o1r7_C(Ud!7Y9 zAqwC7qV4LeqyM))CD*QNv zGGd87J0-mwQJmA#rhx4bR;^Sb6Ap@-Pg9!<=g#MjLEA4k-Z~6K^axs{v2fPE6lhi5 z$m8w)Ow}EZnV1G9m^Pc~gNK|9tEq|ZF78=`F|*Z9Ua`^8ji2EB{)?F=LHwBydr1i$ zDrs+{B`<Ym|&kSnY{@H!|k1RH=)UYVLFy!>FrB;9uT$8eli8&!tPM)T}P_S^%$0}$n zhtK$$nwrU!&6>UkkdT7HAlQXOSE=@X7bk@dbH@%}Lx=pHoJfr9Wr%MV7*t&dq53s| z03ra&tk2Pt%(#(NbNh}^)22;hHkauCVSgr(N54a#mLa{Q0$s$~ttrl4;-(l|d>39A z%*FzqRsc6dI+P9W12?qTDrWViF^JIi=Dhv6{G$HiP9Gl(CLR1<;yj^RL3(eDv|ewZ zD;Fx?FUMy3xKPXI4>2Rd67w)4b8uI1?TYpwnDKW)3|BjvIOUqYru&60Uu+J zT}W8k`NKR(Hx*9JT=gHM3-iZb0}u7T7m{%&ttmh;SVe6Ud8>lbhP7VMNb-ZeCdVpN z^`?ptZ;UR|R0JT)|LYW#`0z22Nm@;KxlG<>Lk;50!OP`9%j^C0Hzp(Cb<-n9xPP1G zbsRM{FoY%ErER|aiZwR;}!GA){)w$d)$MFV;A^rCXtPGow)EAQ6C`RON>Ki!9&2hLGn_$f)(5&dt! zAM^W*$7vU&G%N@&HX%nk`5eXyZEPK=0n`e>>=iua`Op@yo+q0r3_&(il(t?`eWOGDxBLQ{@leH9 znIpfxf?#%)cBM4_GW;{M2m`7F^O`5e?a&LD=VzMaUW)IhTpi%FO{;UVZG4!?n-zTS7!xsYV9w{e=at|0A(Z%~d_B}%IXV>j z!1Jlqy#B|ucEf#V7W)q-}5u3#)75y&_m4j!CZokA8}DOgdbveUZE-6tII8!zjeZ6&&2Zs zRp5N9HJZD*m%^4qeqGX@lP*_({eJ$vF?4Dw-EXcQ4TN5JRZw#kyRGP~C!b+_0H(w} z9?oC@TR-ISR;EUP=HsTz?YAL2ijFn%eHY;L>j>TjY{*^<*l*;Ff++u z`lc3)@s@g9uAVq5m)y55wpZw7qyekH{X@EEk1t($CGc>K>ynf95Cuv= z*894^7e#>`249zEOsQMvyuQl0;&J^b*RG;t`CtKbywIw_cih~#xX7f`oXtG@^7m}h zjK%Y|(>+C}ZYr|Mu^xE7(Iw>XO96!>7iH21)%6t8ooIh@eR(%7i6vR{Nut=Fi zMP)iTu=bRniAAqUeU%$A-|FrCsiOd`q}I%{q?Tefazqvm!umM4A0sQ1Dg@TTt>SG$YmS`3nUK-F0%%P;6O|qS(VScK;ir zLr(wARIvRKoALJ9Ea|9rzP*|u)V%sHEo|!V=o|T!pG{fKomAo`usaNpL30=wc6X;s z#tzE4(A##}%)MEw=RjwDise&HJEIn8nDMstP6W`2IvuayEDDR($+ zah!BP?U?>R>6#BT;#&S@onJ<8HP}gRVOavlr+D#js#upA$6zubsnLj zs-oUI`@j0&5o?9^>oIE?zwvtL>c|rq;@mo%=5GL2I5w31_$-_3#Hn=hGF%dd$cS1t z@1Nu+mj{>Zk%6p_=2u+GE7w z=w^{nx@k!qz=XA3@8$P$@ZXe)pZbMtn>K(XcI zPb{z#QluO_iA+AWAvf=42A$;Wk_mNv>7>cnhG5b#q41ZvbylqX`D4WgZ2ljXeN_?^jn>)UQHy3rorF(2!h-b(3lboj zwfhKt`;no1`qLOUy7P>fxY^)+_TZaM7>^Np$J@o4rAgSwQqIRVY%Gl<*gkn#09H?Y_m25H-Bq0=G*gE*VmgRGhKvl7JxUbe{tHM`f>GDX_wh`rZ!jucC_>dHofXGW6Pfo$Qv$ zw3u!uAp1qvX(Ce@{y{6_*!dRhOaWs4&4$pI{KpZzQg6)*{Boy_u;Bdi>gvbe@xqzy zk?qB-gMK$2>>WARhrMJjXXfOTtmu`zifmz-cba^-$nYs$c1L+BZ!pc`O#S4F=xFGI zYJDx?pLvhY?GR?m5l6x!Vf+(tDXK+q04&(rI!{US-a>&xE0V9=&@a&$JrkJ4w;S>AYgu>T|FaPTvw+k&$yMopb$wn!JB4D8qrk zBw?Tr<`51nURj`K&G_o`EIbVEhV}bE=^tsMicD9GXBs&ACK)Gpa0{=&g1@;ot(?!{SM~3Q1W}migU@K5krL?(Efg+r2<(d#hDfTB;{pz z`NeAYI9FK3`PkT>UN>CmO#CLr;t;(Gs{|;7Ri09Q5;aJ`dy;VIthRj!U! zzxxxMWuu3()4_LQVwpF*zJ;$Sn!~$(`$7ty4cr<$eJLPo`YC)mXr?P*0M6VN{F}pr zfIqm%4C9q=!Ansb@67z=aMm9|T+!ETz`HmDyU+8ns0`9RKe-Zg_Dk{@#MigBPNCec zhUL%rE_IFfdZe$KB|;V^VW*?|pcN})YunN?VwbhE+_p&jF}8fEZAT$(fAZ|!qRGkG zMWE^i_BXfajhhjJ=b3d_0)d5B6SEu(LB=%vtLP#}|4sOaB&x%HfHE^OH8ip%T#MhZ1edsFRVj{3nedhA1X>;>>~VUm_P z^!y#*@Q?KvolN9i{q?&uuGSaO-Ukx09RU`LA=Omdj8yV~aFo$0C{#^=0tb*`uG_fsNlAWh`O@|`MO$o}E(CxhVP62@{9wG%RFwFvnQOdIUT zd9tnzG4Gc5N}x31pt^`NJWj)a3o^>mXxZ`SzSit~1IJ>BTd4{#uDr^m-nt=Gi-?MM zgA7CM3Q_Z6Nv^0>GZf}qrTPq@rkS$BC{1nvX*T@N6}3R{Gr*y%LTBUmxDU_vkNS8+ zvpPmIC%3nXa15PW7AB3v2A4K0v66?PQMjVMuME89=8iiRb?sYZ;DC`X`ksxPU_A!l z0p7IhQ$Fo+5F)!GyD97GFf4Li!b(XeB}w6PRC|ZH(}DAH@?D-r{U56GO&j{cU;1MX zo(g;Swmf?LP*D1NDUICVs8BUkb#puK?hrAOjO8y{%k9kJZFf>t2YA|HWc zo>C;BAE_;8S){k8P;vE2oz_#BEOq2SAN*zrO7o2)&EahK7qVFbosLfjgOP7sR_Si|II#PKZ6&cV zI6~w1L=C3r(K)SL9KnPz@@%5aZMXxrASBcRnD-g=`1E)VOS-pE(t6J{QrfgYehFPE zHw7W?>@V^!?fVJ$L6tUg|9p8Z7NfN)ohg}m*B}zZj&{qrcJmb|-H|ZVchKSZ_UdF= zWJm4IzMbQwyF5I7MXJb1)lhsU0Tb!kHga^HR{tjmzs!H*;KOw$$p?%x08b{veD1G* zS=;1|ml;9R^j?jY@D6*86vJ({4`y&7zR@?!9kr+ZRo3w&IW60-bJB4-!S)%a5w4tB z5q9WHXYaaVHP4603t=o}H!FM@o%e@Ad|!pV9+cL=$`1a?W6aiNCxnZ)h2|bW0@;`% z6b1By)y?_ET4Vn>sJVGa6&FP>vn|!(MZJ5LAl3V6&dET!PRW%aI_e>l<~z^1see`d zSlDhJt>LmGnl%j3yev0NrDusD_cjxHUC`W%?d(CxfD&h0wDPtMqrsx^=htWVkzuY< zx4x2g;avkAoo>#}diS17fE2#_%)XyoP@Lz;R8Un@w%b)_dH3`fXuR`5Cmma1JgVE_ zrd>lT@+VX@G>l>_G;QU-giw~(-CC3X{280N@p)JjFq&wV-@YC zxFJW!^mY z=yaHmDMw%AZEc}V_t4ydB*YFDHu&wQm@cp`6xmyhO7 zAaGFbapvl%>SqMwfS!W{D9?h|oGu!i>cDMK|9WRHN$EcAAIfFN^$5^6ZZVvon zZm+#~S*L(#+Hw$=2irt{30;rO6lQ`DlMddQzUTbmT{3nh_dfCy_+pUGZId$_l&O8~*t z#olcbL_aL#^d+WyvWKdz)RsnuYGmUEeSz1y!Uov)-h~ryOg8H50ALb(m9BGoZRV3; zR+HDo-LvLtBp-RO-y236s(6Su$+SKm4@)%}N^tC5D9-$|kD=c=jyUX2(4AVjicfAHSf z@0|w@8>8BYGvyfuj7ks*LzDEsj5lSgej zwS8W-ea7aNxpRO)+D(`zD3-kyau!}LoQ?soN|m>FjolQ>8+?m)%Gi+MQpA?Nu^%7s!kt~Xtkvw6?8&|Vppj1CFc_B5zSpeQjgZ^<<;LUsB(c2)+!5#|Y1)MJH2fx0VHuz3~<21sjf zmCzf^>=M<9c6jhd3wD0jh4=ggt{+oPRRZ~$u(I|~*K|Y-P?2(!fHP0;l4BEbhaS|J zSt)uDU%h4aTx|Zmn?g*i99PsK|AN`tV>rMt02^4uPyT9?j_nS^`sfww+Z!xTD2`0s z6}~s512zEj&j29`yQC!v(hq4R^7Oz`@CZFsyjAy~6`xWUtF59QE@I6gUWYvT<-t4h zsjINiD|}z@=csuN+?!=~@tFyA1*+tC434;v57{{#^49O#L$}|$`^)I?sohOPj{WkxdwL;MUVfdeJ^X~$%#6%O4 z7ut*AZ%%Zo_mup5FSw}oIm98hi*+V!D2wDA^!nM_nwJbiTiBn@Q+IcnDh)>do01i} zl0!#cb_9XJ2Nn@iHFaqLpTVs;4prkzSPgCONL#(`U$@rzyrbF zjf+ffFD8k7h-_ZE!1$Owz?>#^vH@_0HV2o4eYS#i5owN#8*8`y>U*s=(>e~Sm*p~# znu3t=XM5b(i6Hr%ZuzqVE@#a!xwPYhw}z|J;-^1toP^WPc9W35jlesK3^UHI76@{m z%-}YV3&1gE#r}cl#~O}B<$7qRbR9Fc57V{P5jcVilxGVY#!Fp|-Qag}J(SZQUCNG@ z)yah7A!H1B`ywV3b#)iujziflJ2$F+J=?%!e2(d}aZM;YatAMd=%_>=wZj0&&at+swopTf0G?Lt%G=LVR551bYQ{TCLr zR4gv~?Ed>{z8e8fIVIVV(gm6M{A{LPxUOwt+sb|9YS)akP6C!wWiH<&MWKio+oAH4B7-9xu-3EIiC*??*JA z{+xd?q>exM(0JX-{PU{1+=Br&Cv6-`u3Q2;cMx?p1d92mO6j8O*&Yx7V=xS?rHk#K!Qxjfsj=V25(l5v#d-^w&1*%%a5rWzd^U#C7TQW zYrm^PS-Q`8)%7Mhp9$4XiNqa57q7=v>ZMk_?)lSGS3+7mn2XX48U+QQcp${~V21;U z%e}?V^=&+jVPFQ{PGY7TFg%JLmf{5ZgJz z3nVhK_7OKrfTkjEe?#u2yx^`DUFbC)b=h!PZ?Zb)Q`8=Bx|#XFDmA-6TD1op@);E{ zI+Ux+Yp5EuWH?dNs08VH;3wKLYa+zB67l1w}upmelKBVNVaXc}|d_N7t{nzn(kZyDMaFXwFhI*0W&JlRbwDe^Rf z;7AYqW}zy;GuMwvaFb7yoTh^m8g0RKzv+?2^PGGDp53jPfpt#!QEHKhL!(n z+D&7}u0RStk3aC9lv=#!;X8I|hn?)jymsuVB)h#R*&5-!sR=9oCAq zHQUzrc#Atbc7huv*pmw{W9e zhgVa^xg+~qP{b|}(O3I*yzSg+y_yk_ z%Zcdn(tSHtJK`VS*wy1`H-x=@z;c`>b0P6Hy3Tl4Z8N#jICBAS6KLN2pIm?}S|xd| z0S-}kf*i#!eoack{iADVm{vOVMOS%#z~C&fHcHNTN@&K3X_p>e_VtG{xl6NH7uL7z z2STg*v*wUh6JOY~dPQ#reyxU~_`UR%;kD_dSJ&;pCjh5VdsZ?KSmDiGOi2C-S~xg) z?A_w2U3GCDPx|({2LmCWs^BbEAe&m}H+aW`Rg``!yjVG7cHoz+U}apVJ)>_C#J=_L z&pg38BQc0U`GB@Q8WM!i4tWlYDKjqLa~DiU{U0X-}o zAxs_*HL`XveQhs+-q%!y$vPX>Iwu_ikS-r%PHF9TZLAp;$r?|PvlE*28mvJXIt<-v zhy{+|eU9YKm|H&Qc{XQxzg5`~)&5sq$+R3%OM}KIK$iUMe-&42?*3wig%(h4Q7iq% za9PTJRd5!=*QvHYp}n&FKFV39l)anyr!ZL>D^)CZ=T}Heois%u|614YI-QIizofg2 z@=hhYPW0iO5g(Q`O*a%3zCtaWev33EBN&(7t|=@@iQCO7pznZW8TAuQl^4WLhOJMB zA^Rh_F$A+vPswMLrV;`FdjB}8Y?;jh8ZV39Z_@a!yVn0YMO~cO%?tgU5`8ltH6uY* zPFwJnvMGSu8`mteuUR~HnOCEH=w7_$jxVmR7R+qfz|3j|p2 zzcE~8_N!~3GRS3y|0D?cWwSNvhFN=%Aj@2XKab$R3h3I;?D*>Jor@H~`~F_C|JJ%> z1y=RuP;@6eX~!F{W)|=g3Zr}NqB!W5@SgF}B2^1(TD!CVPq{rfDU#%#i7b_Hk~#CX zIYBS6WeP21(&RG3<~}YLG!ldfg4=T!}l3l`^IvYA7WzWAW6AQ@r%s&Q!n?^o4j!D4X44n;MvNYSNN+#^(I>YT+jOMOa#GP-A!cME4%NuiF3&+Kig5~8?RuQb#S`r}n7iC4 z$Pe}Z-`1_HOYMx)I-mTT(5}FSY&9V3**rn}>W4cn@TMUjpS+1tDShzedoHu-4czSRH6V$~6AEUWa<|tL6rV3GXq(5K;VgK@ zBb5nofR*>p3Lrll+tH<`%shsln?;#^P;>gp3>PNq+&tkh*tonaFT)oM{9K_>D&1T9 z@B+$>y}!y1$B;U$9K6Wk%(g$z&AWS#4l(x;OzI<{elph+*@hpm+ksfe05OsX5|fhH z$tTQPV0S5q+7Fi@H+lsbh9O|ik<3?Z98FCbbyb!+3i*?d!;{(G3;JKvlBxUqFx$q> zc|q_qFLGhv4wct1A5b{2`VqHmnxV2z2mQKH+A0YaDyN%0A`Nmyyy1%Zuv~F}dH+um zuHUf$;3_v@cFTG)aEUc@9SBX6D0OZ=YAytC=9aVsQ&lR|4IT`AcJDgy{d&zD&L5bf zV|SmbDF|qftnkmXJDuAQAlU`4jRL2nikg(8_4FYHxc)CJF4xWn5NCT|V0HWk6fxdg z^frO=4k2W`_l;%MrsFKC<3I=PuYu{%!31N4GIY-;%j@ZEM$4P*xBD+G0fs@>RpF`@zs-rW zjZEixR7sPQu;obwW@%h%=dJ@Z=oG?aqg*rm-tjuNa-IyNaAY^k9hTL0V-g;nhTxV; z)f!cq#?%9sdIM~{goPhE`Ix^wQMKEa&5j);7|d2n`K%P@{F!=jjhP5Gf`Inu4Hy=% z7s8HXz&}}NeMy&Y)3C}TYa3>BU5ncRQ(G2jb2=^xelh=~8Kqn(nAZ1+R#cFu7Rqa6 z?Y;P`o`+8iRLvNDLd6K|@7&pF(4;KL24)d6tNk%u2VrU%Nt$@J&6^`YNJZKuD zk4~5jwmW3_4})#T9)0dccn@$creAD1%G*p`ejp1gR}^)ERS-p*y_oS^_@x{UbNa|Z z-R1N{+>CQQZ2K8!_MH{$h`{J}WTNL!hCl8brl|C~6WD2Ej&=6Gx_t$h(H}doYo+`& z)-s|T=9mflEl>XKP3@N3c(aLl`n`9WKeT-UTQ#&CCI-fVXqX$7VXYPx$p}c$X4t=c zW&-1VJCn`M)<;SnsLaNsVV2#KG&u_4kH-)S3A;7y9bWJc6^>iaQVM1Z41#d+#<~&x z1!gIHkIjsN$PvStu}7#K2!tVMvBQOSBww7a`!3n})OTkaAU;~{6}{$m1zf47JeUB~ zwa3lTnf_GYy`A!qn>omh+WkBOGMg$8W!pd|&65KF7x0eiRshV3z;e&eJrh^JOc;7H z<>g$z!T~weM;*`3pOwxRDK|3iSJ_YUuS~m%8u=Ib+xbuthD285wE_(t0~rEGYNbUI z43L6Mi|FcG^SyqCEIy1(Bf-)h$2KpfHH(L$<_9HC%YN`%Vp*ZV5t<G{MOqwr{=j-wp`UhZnUT}1f0Sj+H_PhutAGL+Zx4jw+3-gq9Vnfnwb&D-iPDO(Z!~=(COJ--^3(#E|9p-L)iBU3Cn!d^>`)h z@MFugcdLp@myZ(tM&3@=7GiD4TfsX63d^5&4aj(dq<>G zML-T~1(1DH2Qcy-`;4Kd*5cK?oer5Zd<`V0@1{Sh#EA3N2z)gA6euQ~y$?v!ynO-7YKmo)rjSv1FJI-+!mCxu?Pv zH_gyzO#i8NxLV!^FXMlbdDLer0m7>n<1KMEE>Em&rWUJGLS92J0Ld?Vnfo1qV%ACqM$tEp(=@B_joz zj+gQc7ks4i@?KT;bwu_{a=ee;aw+8aPT_A#OIGyxsF@tS29*=}#wUO$rq7?I`#h0P zUuPdgWspT6kU*{`Lj&(8ff0_B%>WjEw&}Frt;}3L=0i+Qv~;#MDtCJ`f%HYSEm`hx zq1Z-`nNy}Io|G4ayW}ppJAJrF9h^|kvUywe&@gG}E?A+>bWj6w8eUU6KNx%hv9(Q< ziX&a7U1U2mJ^AFnh%fF>lWh3c?>a%8ohMrH#?LC0*+5yxZpFEuHF_@lAfU4fUZdD+ zjS&anz3Amfr(LWl z!?Fd7X=4tLjoQsHt<4W|OL>^)TQnU^fLEA#gV#ASyUr~ zEYAun{=kLG@(j2$UYV`B+zJ3l4OPmk2}YZ_Z&emu&S%F215v?&kN%DS3m!B~9%5`7 z(l9T}OZE%F4fUkTtppJFa%Zw*f4R7zv*{a2tHDc#+N$y`#+*+3slGp8Esp#POi1U( zcV+1*32*q@)G}G7sJ-kWF3$rs0w7TLYnUE$-97BD0f}fLR*x*}M_*SbD(rMr)m4`_ zHkUWXzy3-BXx7aGcST~xDHDTvZR44Em&Z5!plRMJ0*OsIiih|WppK*PC_b|d@StDLWYGY$oO9?#<-vc1aA zE{Ma?0(=g>RF%cVQD6rQmu!q=7YR5sYAF1IOni4g7z=<3Nf)xHfEc?ohX45axPnT_ z8@h}8T?-R%`Q=|&U?68asnJLAuLW>A@q^yIsQ|qF^G>~)N>~{;Rujk}=Bo{%kgf5r zzs$0GJR-L89^xjGAuVu&-`y8UjPv9P|kmf)xTo&JXe4SAm^gE{lv|^-l!7d>@zPlRr zZ!kH(PM!3i9IpS{MSEI0urhNZc-3e^ZRC?S_aNR_mRmcgvHJtdG(4QCx@k%!yx2^Q zznkQhZRXT!Dzu;W=SebV6jXe)v@X}ssQ0F#ck9Km!sGNIihK4gGP2KT;rV5oYr*V9 z_Q7?bSiC6wj$7NLS(;eSF&bCElTpZRj!n$$1+>k9Af%t>^ort%sXYaB@ji@phbTPs z!G0jP4epGV8;R-l#P8bkpHJ-q$yDCsLBt%$S82po= z->cijjz*fzzrzihLh%PHt*ES-G^I`sxq`(=TA&MFSMgyEBJcxm^N`RcsjS%by$#Wm z2;^O8@bZ=gPtdiDTUIB!o_B|221x*`c>>rZxs@M_Di}+NKF-8klr5pLHn|}B;!MYp zgRn@mM0nM*dovg~H_2ybv7j}5aT{$3W^7Y1U=M4!e^s&RFVB#e0d+njDo3m}z$pZp zt+PHlF&;2VGe6s>P{Je9T;W-pRT_%taL#+CJt%iGHj!#QIZBQ%hs(*?Qji_%D0e-~ z=W4s_lO;wWKxftUj!~3NSNkh1Wm{twtdBlYN5o*$*j+f?;1#hJU@R~13;4*oi~rG_+%#qO*wlGJ zi&?)8aWICoBqvd28xs%QAhGYpa*D9y;z^(&GdJ~Vq|Ao<=HFPCbLaJY172h*xI9^S z=RDuOwHA3;T5KX+b@PJg(jCciKDW=GTdr$xp7=o@bPxcvkv0z$9Ya}HpVBZ_6XktaCu?}g zqb7!3jhW;sEU%3hK2C(4+fYvdA5ku1x}I25s+t|CK%4~r=6U9^A%A0tZ-kh?eaBe>9zSSX2Ms#_13c5y=q( z3QFhb77%GE>5`HL>29fk(hZW*F&ajL(jB8gnvv4=oPD3`_b(UMU^_dX^N#y|-BQ+F z71(yv`6uFy+HS5JU6uddvWt9z|6@C{$wtZC{u7)ic|9$6o!IZc88yb|x~>opeC{S@;R z2oDpD?i-^GKr#IjXuD)~BFZiJORJ_6FOT9M?UX*>3$b0fhknyO&$sq9i8!+(yQUC$ zPF7fO6G42Kmp+pv6RhmGTaf&Lo6HL2LUs$Vii3LzH9L|}=mIr^hWnY9xyzaSO0P3wnAn#k3= z(><#%%Reawu!ssEVDgGCzrSGZ-a(zv2DS&{6Emx!qn=yhY1gzuVPUnbm-dfZJ41xA zn>HZ0aW?HO2OI6*d+>ol`r8{Yrx>lv5Eq}waCQ<5w}|#k@1fM-n}YXu*ve>;&Ww}i zlrWTs=DMJ7W{7=s2{j&aLRqi zl-8rRrKJDizA|_i?P4nS0}2^+vflRFmiCrICTi;5mQ<&z6wB=Is!Osd)f0^5wa=}l z?ES83b$@E^Y%=HF zGIVrve6)1fopP-_x*J9-a@I69R6E896NrOq-?!Xca>aToYKB#Zb)CIcAEJ~h^_hJ< zJ{dzBmPiMWo-RZ0fV%buxAub~`r>gIf0;m0*_I(%IuX(WRaz5Jt*|nd4d$UqXVSs0 z%~SkiWB6?GOxsBWR)RT#cZ>J1RA2i=g#6?BsS!ypHoU4a2{+|x81JL~ii!ZN5nsqI z<4QA3jD1X>3fsskJ&-k6TiEJ+0Zm`%^uzo2SD#N*a8`>~S>1~hsPoxPq92MNCktQC z3n(5wIWFM3A57&)w>*dGqSChd14r|LAwUPxPIsRW^s~-sMy{Jz%+w=a&Y;K=AMem} z@=Da*J5iM%9021UcZ5Fkxp(`e_fnMMzn>H{4<#zOxZhU7r2DV<)}sAz6tn;3GfiJS zx`5+U)0RhpyEwaw*jKMa7-T z{?DxKW0AD2Ij>!aYQqvF1;^+%<+GnB_qKCpt%+Ophv_7HtHH>8x_Iua@z#{A$sX_& z;swsnvO(C;-6MdEFzr_Aut6A7Rf`7bZ?1mD3tePxuS+baY6>nK&0z=wir&3ObLn2p zm;0$7yIgS3j+H^I?C;dwKrzt8O=*vGC7cxsIhJ&DLO_DqQaix7iHsI3Cp(j>F*6C6Q~ppLP{;bB@!2(P{X zauuAY_i;>4N^sQwv8{sEpFgFY409(YR2KipZvMCDy3%kbk@u|7KxPXm?oBp(dHv(B;Kp!wi|GON7cNk^@VR??5X_b=mfskJwDP z)IPafM&3AW*AJFsk>XDskkGb$?+_I%+4`@ntd>qM-kz(mn_WrXFA3wZ z?@Lx1CbnC%9a4|4W?1P~M);*UPP%>tZ{H*exB_!n@z@(cR*#477=F0c4L4keqO&9D zVTwJm=;;$~_9BGq!j25Vxlw zV;6EqI1b_KeEu*843D5;t)Ryk$BO%TXea-Th(+EJ{%BEb;6EOtF0wHmzx5jthLbz7 zRU?ldf=%%@eda~7){~=X6ExTQd;Kc`Be{{mpnbI5BC9ep~%nzXYm)GtqxL z=$na1J^quY20z>Yy1>lvxL>pN1L4Y{Ke`gyS^~7@r+L1Z_x8__3_?mdv$5w_i)K=&`}b_ts(YybuvuP zZ#A~+ZvNV@cMi}LiH9<=Sax}i&>EkSSINt`L~+s{LhAtIcAMF46<5|pbBqg_*>m=> z`oruWC$45NV%#fknAFu#$S`ck$`&>yJs@wIgW$P&VM1;A?4c^F-H>8vaCllpU@gQ; zH?k(UiQYWoLT|sVed)??`JusS59{V+;eJ_pc>(CJvIXbEXBAX(O?Wn{CT$S{RIp`^ zO4$32X{uFR6%Q#u3&z^NwT*jLIIPuY;vuyDjbZO*2EG%@;bWpn!`=T!;5Pg5%)i2g zC9RaVqgJpjk&oyGIAE#t%;~mJTzVwN(dcriNm^)JSbL9m<;`+9JBhRzsq%h~^URPC zb8b8ax4Jyi>JQPC(?=wl3*8 z@YJSPeY{U6N_|FwqYXOo(wx0tP}Y20fZ1*5d0ny*1Q9(T<)XpMb>RhO|7it-46GVj zYmj1X{9yIXaC-oB_P(0VPm6&Bba zM)~?3Fy2`#R^*l4+?r9-`8(@Kv{i9VXV;+kK^Bg-t+PKc{!2}t=$}^ei}+_YPM%;J zadX8cK-06Zy~=%WNBEh*zdj^zHEw0Sf#4{La4XgABUo$3!;$DYWT5bqU9dclbvv6G z$bKLC`m-XJ)A$pbM7#`HBpU}2YaZ;V!FZlmb4*F1G*ics+Hxm>Rd^dhh%scLIllok?a!sezer;G}09*bBv~&A%Zh zv8&ca5c=*&=h3fHCaPm3a$GtECiRvLB9CBllHr=0PwBTv(}ovd>k(!ya+H-RZJH?E z?DQ2|Jk|19Tr8MNS>O{P=XmMBA8PfuhO@{-n5VYF1TT_vht<5KH9Ev zlV}g@Fn1uJuZ*T8W*G9T?zbLwIKR5~zp9Z5@$x^>*riVvD8kpSl7Mx4G0uCrUg#3# z9<<4s(oF+uSjw97re%stA$|G4Lq8#&*QxVTuKl<}eQ z5JX5+%!Ha?a~*j8WseXL&{m@ABz$dmp1CAN5`XhkN0TgEruNrU@Y{45v0Njre`3u2 z2G84)9VB%ix*DAbL&8V4^*}uGNHdF7ilyl_Eu}__=AoMAAE=Q4r`6-|%xypKOqk7J(tK<|bwL z{CTpBv0z#&X|iSQg?e8?(^2Vyefn6#SN(s-TGxi+qHBQwqulljCci?uH>E54z z-Kf2WNQ-xAhV+w+bfZ0Vr+ahmhSHCn4R|O@pd3b<%itB+XT5@*;~wH@N6zSjM8S$E z>JWbIn@?9HcJS(5ExL&??QzZuvw~}Bl!xKhtb4~xq|jN@qRRS{aPhk}R~`fd?e*{C z6iRX;#~k(QdtFbBZ1-#0hdu{So2GwN6Fq%SAn{m7y*?*bau3G^`g*c@3Tm498+fNc z6~$N~F}rTLYwJ-FQ^!%!`HJi?N>pfIW^^n@_!`Sj>`J;>n#$p;mGah>^s2T*rz?Ph zTk0dQxw0n^?O=};iCbL<`@&;hdt}FWBqLv&>GWyo&8PXVTejdA>h`V>bDp}nHR{A$ z*)PJQ9(?w!HvP?RWo#DL%m!r*@GH-d1#U>G^xo{npiJ()>}D%`-Eqyh7OOsQuG~?- z?4|om_6o@g3kIBIGewS$%0$Ayzc-n}`L9U}jA7QcMZj>gIFaEHIjz}KBMNJ3NI$Yz zN$0owM+Qdo!bSE#iqW+$$4yTHw??K&U^}MmK(eE`Rh9Q}Qi-+^+liEtKg25g% zF^~IzJn4_|@DJbjaxz+|$96s$1_u2y6~v`K zsky!TZe8~J7Fs=_q#fn84ej@c79;?QCU{XfOV87if4q>@dH>|026qmL4mKc~DW50o z#*UJxhi|cTwS3WfCrbFnSC|oJDl`J+t{Laz%Lf}G?7~p*GjmkK?dL;jjR`@S zymD#UxW^mFp20vTUsJD_YyLS^M>|8z@d#8}_=wc)dw7%<)Cnv=P0cUP5_C2Y8{=}t z?{(oh?bXV=axvm%y+yQg(dMSOJIk_}U3m1KkLJMQQU)7gN%cah!fAZT?-tRZ4FEy= z=4UK!eiH6{3ui^=T>43UoJ_8)*E&Z-gU?G9f$~l zqzbv1U%u{vSk5<=S(px;lo3Zcmaq5DEmmX>MAQk>n*!SBTSRaRrmS#<#u8)xJUj;bLc`@xz_w-vPFTsD`zn;n6O6fEF`n4#` zl{4O*q7;6@V&~b$HSN-2#4#+gu5RzAmp#Qb8k}c+4Y!(kw{@!?e45)7$WUBhI-M8> zZg3tt>+TiPEq`SG0uw6v_RwoTDk5K{y22h*p#~wyKYEdFn^zNOQ0!$}P{nIFnE&(3 z-vmi31GZnLA+29N3J!ePt|D~1py~{1cTy$IKA&q@l1=N?!<4g$WXs~qy13gG_b=69 z)%T+8=QXr`lI_$-v7fY}X4X=FMMV%Q!Am}1?wrT1BmXbW$hI#ga;$V>xq4PCXC6kS zyT{Zkdn`7IV2QOB+n=?Zc?8-mm5RFAax1UV=8P7EH`LhLE)1K8pqpx-hSq{t+f z-wWO_EX0&5a{gmD{fP~23^*OcZt5O%7#qEWG$Ct^Qi%&asrQ$bzmX z4|GV3Zwgp{vSb$o3Or2-=?sq=vmtbO3hlEROd7P^lgD~Iv17!JBxwJDz4a>c{vTy> zEB|^dMCoU!5B~?@*zI5aBKK6|Y*G!Lq-OLhvfVYPEq$PIn5~C4+m#covFRHhURb=+ zMgXb1V&$te!Ez-_gRRNr{596@YD zF8tIxJjtn=O~}oFkA1Xf4U^EL1AWG7<8OMS zIX1j^Cdh-tMR6nG?dvnyjVj{TY@aFx%zF=VJXM>D@<&=)H(o^&XYD*xxVM%1mWT0E zN_VH1|N8IEcxzhq5#0bTULO9*pr$XdA6D2OFp2fw`fKjDeSDW?Y#xOeD*3)Xf#34I zLuoX+XhA-GPk9()zdEsuolTct=Dh}=p1~HkGJ}*~bRR&){kDoJ8te#IZcPHmD+BNxdc96 z$8`&LrMk(FYK0Zm2P?dZ5)RIVVm@8HJA+<=)!d|K&l-NXwTTo`tRy+$a6H924(BGS z(9MPyE8Go?w?q1p@R$Y)d)759@Lm_hto3={C5Dv_mRr%B=dw2LZN7mrMePQ*@fuqB ztnOmJu=^xnvmJEW1+t1>C4heE_sx48zALKJAr(27Dy~^Qh&-;m1%NXd{R;aVzj!}K zCqp}|9w6w+y(EE}+95jzf|SikJ-wEI0{fC#RDV5iyE!hnK6jzlnky00MEl?xF%AR1 zJvOv&0%ImdHazuUbgeG za)hynWV;)s?YzoJOn3J)cRs8kyt+VSMw}=ctekd{Aw*TM!a(J9 zBj;d5JJPb3$bV_0tyny*&Ttb={CP%@?aCo~hChmX%5b9!%HHvp+udPhfaVW(R(fwxf?YB~Xji9_t1`sT@MhNnp! z2$q&@UKAwvpV?x0LsGR06v+us!>7LPDh*}NuO`sjhbO<#t)S_}-o;@Vm9B`h?{ zh0}QL6yuCsCN9lcmT&FPVX)8$TEPxmQ5*jSxXBT{tT^|iB-ng&vhzcMef}ETP3hQpSm+nHrYt%pM_-zRv3u}0G*x34>w~yhj)ON2trq?6(x>13=dw*usgeyy+ zO%-F1q4oT9kYi+Ri%V)q@|pd*xTmIr*NhS2!PFM(2N-ANYjsewuR-GG90C-cV_gyGG>=w zU8OSAnykz4G*K$8e4rW3AXQ2UlOROm&Tb7=AXH+E|70wN)D+zHPcKM}7)~1L=4kM*_isRJ z4UxqeK26k8d9%dAj<6l<-f?_S2SjG4m<51Oy%P%_N0=YNZ+|9g^mhW>e8<=*o3W9t z+?Zw4jYb;R1i#v6Yb+09z}3R$B{J zF-HU&Fbino;$&IdpWc2@h)v;BwSJlaNnQPL!^W=l996oYAqF-1djhVEsQr#9y&P7c zS+z{nT%rLKa&l(QD_=#CpP0E;r{FXYQa}^W;pSY5dR03jRLUZ?=XEgL>mQFMCyV5; zO%Gyh;$HS*sftqa{c>P+7BheTi34ux))6YINF)$;K;Kk@@{0^RVQHzt%GA{|%hRG@ zTzO9~Y%a4^tX0_9SQ@|aU}Mo`6xckpyJ`%n9hxSW(gcbiWo%?^J2KP%{)D9>CNLl! zf;SdJ^HwA4b3)5rtAW@R)Rt3(gF)3S>8$79ilYb&3V-vWPNJ}{@g@n{Yx4^1f4_`@ z)pBSgVt#1*tS0EwGWX-aZ-d`2&+utu<)s!vM=(@1HoW<+%Idu)rNGUOGuxKeK;P%h zQAK5ef&H$y|D1>bh6KbFAWrRncaj+uW%zhA>poP6noQD^Q)1Pp#Y4u9(0#}G6r|lC z^fcT7gcoI8nfBSZ#00*CgL= ztRUeK*y*WizbG%6r!_lm#KXbXXUD?wbI$6uMfjBOt$S$hi;lWAl&x5J#}TmN<^R+) zMEp(|lw;0gPo{IFx=r0AcX$6jkNY zhpV8dDx#BewjK?gaOFvT7!>I@SA2rrN3y}%fA~w7Vez|3?Js@$Pt%xiXZ7JlCuOnWX&HIDo$ z29=ksk;bS-5aDvow#Bf;CPgz{PFrUsyjp44zDdVxHH;XHip@7V?l-$d+QXQ5;*Ww- zvS@S87-cTwD>dkBvq%fN&7AHrs1;%7Pzu*ze_N|#EL}zwUWq40W`}rzMboN||MZ9uiPJH_ADi2bEI;-7 zm9DTHxm#T^XI=?UVISgE5eFbG`?5=~_)!}4wyfLa_NP-6bg+lMP$~NYGZyc!{(w_a zJru?+=VFqkByKZpn#K4>YovjZTgB-*mG})i6(0>TY(kmQCv%F*nml%UvGob+7M4t= z*YE?jFli|cavR9}bdFJPZDydgO;+l3-;y#{|Ej_3 z(w|TBlQA$*cxhLeWR}qLw>0DTrHM^0mh{TUr&}NLvhf6-Y<=UEhiw?U8Cl@ET_T|R zU&m$68}~(^4p9YBQsFbvdonrxf*j9mmVw?vZ zMM;g=S#{q|VzAK^lCYD#*DR-n|29YcBp&>jM)=g%xaL#%EdAf&`kA_O)Qi4=ynOiI z`0TsCi8FU_5Kk$Ix_xG4|3*|meqMxNhb`yoJJ}}3sx&=1LgP}=4&}YdyMaoL2iG1V zxv4tmEP{v6oE@-6Vg4bx&bkWM@hDXsP17eJQ^?<8ulQ!R!J1GusNJ0( zFiR}ebQ#Ku;d7CeyT_Un#L~~%nTSKij=AQUYzsaWL~7sx^{^1R>OX1s_YlYrke#T}8AG9-Pdwa4_=@%zy>_7>9IToF=3)zH{5nMMr+E zbmZ=;UmUg`Z2;!quPGQeM5#9jVPE&g7!zW+&FMgdLw{KJJgpKUH#Q2>Voc{DGx9k} zlU3z(%mXbIQ{L8ps{!>WZv*@yV2yAb*gWS>K{&Pz5!_1og75#V<^aE8*@7t|h68KQ zIuRfEm-)>)DsGVzvS!p9U|e4b&$k)e{~_k~)4Kb4PjjPbK&DAYs5*y7C!P@KRGA+2 z!qR%V3Rk#egcB%gRu)bFGrIuD5Hu{?kYgAHqgwVe+Ya=04aYU$6JuIt@zJJ`|4M zvQcVZ%XXpt#pJ5HI(bPl*)yT*kb}S~Y>XP$iM0qb9=`?F8)lpNRq?DQ|Dt7(+^y#i zTNxu65-U~y6I7VW=mD6#IIWBk0Xq-};Rz(kC!l9-y2&I!hdfF=EtvGRFN+!yc)CyY zwrdWltvqjBUVPF52zhR1+YKEk+@8Gh0L%3Vuet>wOgb#sYE7ZN|8IBzKDlvDV}idQ zy8tY1&hooQKp)}|%REQkHf(3FPNIQh>VK~7?`lrcd?7Ieer$X;Y+lf3?igdEpO?ny z0X!Q;hg|9-)GM2;ZUJ%Kr#6(duuGJ)&W-_Gqii76N^{z+o~`{jltUwC_BzMU-ETVP z06A1Ad2a+5cZgeb9NcHz-b`C?uDTrpzb>k26RM?!)qP(w zf4~nP`}4D{ci5tPIn;*#J0oo&?c$UuRp1gU>80Z3f9OyC z*^?2(%mSDCb%Fx(Y|Yi>Ax-9{^qrBPBx~FiONeHEKA7dcebr=-tg!hx3=_^@03p9%F(>|% z-ba9D__FVJWIy@)RjexP|B6ci*|5D!t`0CO0jb;;F72&u^+e0NfnEd1-)9)C>6}T9 zQ|5sO#O5Ez`X`V5n>dY6*94|Y4bMv?^GurF3)h2rCzS5&0-jwR6h?@3=Je7^vS$Cu zv7O839_r#$SPjAin+bvh2hnymY6K{`(aW%~h8V0E7w|#<=@X|r0;*$_R5!AmhQo&J zDPuWW4r-Xcy;lE?x$HY2U~+5Ieqm#io;OGz{ra025EI{ml&{<~6Fw%tK*v~BHWw-v z?6}~pa_-VbqZ>LwCRw$e_Al9b=oVYFZ{vhHEuC5>s zbAQ<0t+x5Ctgu#Qg-hl>RGyUC_UR?k!;~~{K|A~#u)BDiS2b+XduK0j#eXq|U2!y8(29PQlJ4g$BW+PHVxs+;|)#Mt@b`4@T z7_i$XbqblDM8lJjD`UfPaE?z`+~-k9K0?k{4x@QSb^zyC#uu7(Ku~VqM&itm^MIs_ zP09RNXsM&6)Y%BB#^^iPwKn3dy{g@LjJLwy%Rg{iw?3~yt$Wh%%@o)oGaQT(Fz#Q^ zf6imUm?U{SK=jI&gbEmuNekOkzKJJ+)V6Q&RsIUn7MU#wj;boIiui!ne-{~n2Ey@G zB|QIw_Ken|-k4~w-+}^_+XG-NeBywRE28~-+p7O)AQ6r&Y;9H>j|(wJt$$@5w{%h8 zV9=~CH7J#*GsV!FR_+x^0`xMdE_RXB%4tMaX{C+vJ>w5cPq!cgwpr$jrY)JAO@u>w zY9HwGUdFj{{fJpv!iSfh{#aZ6Gn3yfGt`o5D$D^&B|vQf$IQUbH~`qnp~!~8kKlS+ zB|<3T*AlL#7%%@Du#{|p-!!nN;CU-I{4^wGdb%LBFu78)T`d6hn!EG;!+Ph2PDnWlyG_jP9}NleZv*tS)x{+Q;>u~8)k zHcr4&Xj5#rTn5Um;>yqaYkqMSeLOfEC5;(SBRYr~kRi;gv4Ss!8Wer)JEwn7L+$zZ zd)`k$sO(u1HePb?LGfH9wJ1>3(UK5^6Fszna_0&wUEjcC^PFF}@{>BfnS8;idT3kx zl|y&Y^(k+t%FydUQ~$~t1zKujS1#3|^=UW9O7A5b*Y8!aX=FT>c{RJD%uLj>n^WCl z9}k!1ww{?YSZgSc#ze-sYJ^lIZrxWm+MdQnew^4T&Vx`U#%tx7D!6Ju3Eee&X->Y$ zP1Ph12Zd&W5X>R$?@fNwBJKAdyidc;gMPBs&(N;iZSkv=e-YOax?9YduBzG>{tspf zG7H$s#0#0*GXI<5Hc2E$#nncg783=veIAi^?@`@aQbTDMFi&q2SvBEOO^LbY7GIEyHy zfr0;-)b1i@YNN0YE*e~tWjfG?F-C|aI}}m%sy(<6B+0Sso%W8Tqo`&Ch)=r9SDZ0a zYo4^RpWUehxOy|XfY(4RWL$Jp0+c)Gl#c&d6RRl6jkUJooZmg3SNi)JTqX?n*pDlV zG`@MV^5lsA_UbQ__`L7GJlgUi%waa_eYc?n8N)>+n&gA=zF56!Os^suWo%JlYC{C0 zc^lT0xiA^rew|qQhru^g#F+eBCU98Y*aZ%+_F8F#Z_c-5VyZhNbQcV`9yq-7uq@@A zQyk_Ha+d^SkUf_`BtPACU6S1Uq?gPcGL+`w1B}E}D1waMfRe37S&RzTn8|vk zUq3|)C8Qg@j1)`pArRDx{fPj-@z)%eDx>FdF~Baf$7@*KS04x}x0=$}PF3P^HB-nC zRZ2I-fL?L>4P^qMzI-4?EqkxYbS=8FylQZ-DHH55)B0ZJ5#(_U2IZXo*J>}Tl#fC9W26j166w|)nvA;c9FZ^zB$YT=UbOBPh9zlv0C3)#@y z(4pyE5AB%_M|Y=^6)6U{-;w|JMb#nX@Q-2A>u!2tq=z7izeBBupi-jP-__wc9aQ(* ziw3X7g!Tp~l+Dq+_ThWKKt$SLWz3Z94SfCvQT7s+97(kdih~g^XRruRaV~)+c|qZ8 z4|5KIctT=Afz(h7t-*Fuf^BI}_pm^dT1U$(z5E6?RvdC`k@gMhqhgR;RH0oh5zH=< z#07o&HKHKdP?ygS*KE*fv>Ao9kiTSD2~Uch%eO<-Nmd=VaIM z;ZQ??JG|qNbyu|nc?lU~P>ZM*>X0^VC6#76s^`2Hc#uWBb-VxU56rsCaL z18WC^4zLzbx1G=UhFQ^cq&_A8nIo|H4V#9lz=8pdEz<1}!>MX}tjTca5f{~1`1aFc zF>9@t;lZA3M&9zO(Zic%IQW@Xd*ip`z0 z$W8|NTBzvhohUKF>A4VLBcQW2Xks$m<#_U1z_s%;duj;NlefIab`Us@YtWd0;&+6C z(alF!q*q$TI%Fn6!!}Nr_yjv%G~L=sIt>*)zP|(L>&pN7O!FF9mnjFGm=uziWptuU2r!vAxFFVFRHGFF(NG&A21ZX|= zg_h~l5KoL6$Y4c{#LP*7^DiBe$e|w)Bvg-;9&66M&k`48heE`vY9fej|QkS?Xo&E#!UxMSFUcWN+;33TMxYyr8CpBOEe!AhP zJGwgh97)TqPqS)KyZTTIAvGw?6@~~vPjSM*coUPKdS2dw2d$>77Is_GL$O6xQ8llo z3f-N;dY`j(p(c@`hoF$wxQSYuR~Ix#cg>1no;>Z3X-r|^?0kvxn@gulhw(|z{-7Xp zI$j%n+^~JspYVv^viEI9NisLFqZ8|SxfekHS0#Kl#6$M$N?HdSe+Uzuot)ERVb$|? zVa2fSlpnA?{w(q@DayPYD&Qu*E3WSAjnxmhATVL2<3qf{6YHlBjC$8NcS1@QSQb-F z75};4UdLp$bMV7_uHk=EI}rLz-ONG)Q#6MnVMcLJIPwBorije@UhnWK7luNF@D;bd7U0$Eca))KL`|=+8Ex%3?ST zS8+#vZiCy6CONV#XdHA58D{3r*0v2){qVo<23(&a{d~KCh`!gpRG63 zm50iLn+dv4@nk3iNMn%)`{k!s)A(W8p?xTlxp*eafCFz9^9X?2ssWXbuZ=z?HJ$6A z#Z9IU4l&rvLEo_Dweqz!nhR?E52*h=f0_=j*e%+;)o=JeD<;|lvggP4vSU9S4lS}* z!%0`X0W(A@-tL-Dl}cN>UcYL+uSr0_cWdIb*I@Zdrx=^)&16z=pFJfB?+Kgpv5wSS z_T8WCga00{d2Jh~&+0!W6HM(Qc^9s~Kt6vHf-q}`rOxO%m0#MZ!!dt&I6&mnu-VY@ zMLZ~q1*S}pcfY4e2f0}uJJ~zkONcseO`lUZM!YSuwWf|L_l@q?Ih&IJIJfL||J?XO9-jQG zfWy!Lpjb^`q5C4@dPch&YWpfuWnR#2MKEQ3a=`T4;>n&DMKVVeZ;_M&_Zc0_@+BcF z%WtIgY)>7L?iCD^kV;aTl>cIN4~b$u-fb6};5}_ofF3;++Q})&l|%X5DT`y%kITTD z1EeKi{_2`W5(bu#W35A=AU{kyJ|BI&8)YGGd%Qm~YFFrfm+E^)Qt2Tx{p2&#RxgL2 zQ{7UbVic=!X(E;1MF$G#J#ZL{Si3*pveEXdMZ@YYp(tW|8h~kCBZ|Qvs(7r`rDcqcmV!QqQLUq( zZayE#IbFy@_DzBmMjsawR%W>l%Y?KOpUx(roFuJ;*Zix!YfqV#cO%LU z#J$yw8V5&C?_~-ylSenSNRGuv3F(PU=9Bst>Q3XN3&_oPlT_8f+~=?#EN zboiXx(l@+;P{G*dcx9nqPZIScHBM8*RmkJ1e|!4)Q^2gNKYWUXIa)K#nQN9GYqs*z zJ{*VdV+H;Rl_?)ay9mO0wwUp2X<`|+8ie=d4iBIL7EBFMGR4)6n@mVrb*xSo61f|^Dzd+YWHFiY)4kVa~OHMQ+v zO<58%LT^4W6aD61!-`GtB&7~>@)&bQwSaYOl?zXPqYSG5P^D-ppVe{Oi3V)Qt5TdPoOUhCTya3?=A# zIWGP{E@_)SGD7qF7031Hq;}DiP1fe8DThf@zlr!P4YF%EC2A>dga-&MwSy~H(#mQw z044k7c(sev6%zWT$ugK9=GZa9c=iJ0lD7i6r*J7J;9$-ItJ~i%5|ro%XjS222eO(( zM0~>yr@qt)fR*U$kqp_+Xq_wXRV~%T)T*>qIv)&*UBH%YvhDgIZnqCL;qLq7t)OFN z^)153V?yC%*?GqTgo(=_Rfk0hgU3~w7;?H+$^I{&-@huScH9LBeAIN2`9)}`urt7( z?z34NNfw4dTkfE2KGKcJ`hyK+4XDft&8*m!KnH(tif|XwcZLxjiPf4!o^N$%<25Fy4NmK z0>;6~qG}TaAZU9pKg3yXNTfRtLKy@wFnsLNF6u7!n!t*xRV6vu^K?aZg^!q(NXj{X z%ySETh|pf~mm1^8`^!oo7rG{51KCKruiI5U^FxoP2>QCkFlQ@vvy&&h0zHuyI20HQ zroCG&F$iZ$i9@R|+1$`_14+OJ^Kn}u4&EK%DVF#hUHVn>G)^+UB&T$Ty0@60*%9+e zF~TuBb_0-9)wHIZa>8~-ScoQ=GF4&F7NJX?1$UF5*p4w+DOL8 zLlyZQ!RQkV)(mi@7B|UDD}QKN++=r^w3q(D2qhR6Ub!Agbl>0#l-cf#^*u0Ssfa;W zW~9K5##0`C_eNQLXbF3(0WSO-6XIyviaT~k%0z3?5bYX57w~e4++4pqeFZR$ZPxv- ztf#~(lRIX(OG^zz(l1@@f8!HKzC~r$JLpIYCD1OsIwmU&$}p^D3d}jaDq{82RE>FogPdp{Sp&rIJxn{Y^DRKPe&O-NF zVX4yddqdhu%M%^l(g}*eRL^l*+#hJ*wtQO4$>~D!g=+W>>cUvr5Wm)RKM0IYx@gsJ zC>BHQ!g)7*B^SqhK%vB3{||i!g1sK0{A0|Ev9*rELLbUy&hTv)!zc3>M%9kUTn37f z=>5OR{V*dqV*z@XmdG3~r^`q13u)}T)GGKd%0_|=O>oz0oCW3GwjqWIC=^VZn&i0N zgH_|l^rpgvO=Wws?)_1$kNQ36slTtEx2O1x{-BKLx|Q z1FS>UYYyEOR-EEG{=zo7n$+P;ZKS#mRI+O(Y`pVsu-(OhE|Y>BoZ<{kP@?>b*uDWT zrcLwD2tMW(dp#|{MW zFh_-|rM{Z2c`Hp3H?#bflKwoCK0_~xB`B6s8^O->10P9~0r{m(OgIs6y7 zdpp3u{HmB0#U`CO>alJjA#+8ox>S&v7@|h>GbVEB$4!}{;mici3pz==>Jr{R()d8D*~*6j!(Dm$G_Q`_G6zAwo0_4$>8YuewC675Z`gQl0sQ|rhRb4 z_(cP9Fxyyflq7Gpi7l_;mTyDZO0OCsl1J|dyn#7NO$=;4B`!_1$7GqW_myQb($*Rw z^>srs=t4RNYYq*z=B6UdXk}_d2^E_~_ap=>YJw!3%(@nN;i3m7HlrBFYuta98Ah_{ zqnBA8ko=u;Iij?(0&O`DR?2Q0C%y6`c+_iEc9F7c5~U`v7Y+vzR`eevM36alnsy+p zyLdCVZ`#z)ScB9i4=G^MILJn5dOMk`7ZlofLPKuys$o{@kmFV4G^bUuoxPCe;`$Ux zc$g%y8%(R1TcoCtj!LrG6qLEG=_3ogDbB#q8N_Rg>YDdzQa=5#C@*WFkWPt_r#X)jBB`Bv~K=CQj7ZLCAC{+=Ln(7w=eN>uk-5exRk zLc8v*#R;d)S2DL&IR6P8K9hHZ$Fy(3w}y1Ei_&Bqg_zNr2NVVjnDc&8CCb^rM21pN zUT$D!k#d5`3c{HQ-=sUZDrH&*w{u|){AnJDDktL)LX?GSxl(2Q)3s`I@?f&b{fVJr zZ3Jco)88Na;A3sspVcY;bmJ4)a?oCZh^BKRu@ypO{c|QRFJcI^D4)9nnJT3-DN6!r z29PSjw9Q5ix!@=zijG4%JKQwlgPAH4CI{Uwyi8BxOD+Vd)8i_ank~i7QqQl_H;f@i z6OG2*@fOLM!sd?nOUgMOlu^h^W1q_$Udy}PGoJDlzng6kos|@eDz7k5dSk$zeE6%V9$0ot zx@D040Qwij_o^wL*kfNPN4#-po|V%hh=r2FN!LuyATH7q!{#TI+g-mbyqFJvdpibdSSwnc*ByyV4S{Gq>!2`BK(AWZ)C&Qs zmyaj0i>i$A2+wTMDqetzhctwt%I%+E29bK03juMP9+HSC+fPJK1R#XEYx^^>8uvXv zS_eDLaFxGoaxN!N&|b@(S`t9c(o}6WFse6(s1KluYz|rOzV5ooYxlb*9fjW*X0Hlk z-k33|Gq61qAzECkK7j4bZeiJhYU!lnY@#jDde%oV$YVzeK9)bL9QMaid5XRz329Ad znuqd+JD)tcNpoof!euIh+$>Dy7?$v#J+Z|i{f^D!E6o*v=$4mS;9pEAcOXBSOa6aU z3o5!V{-@H9uLLg_UG{IT)d15<8bwKduXfq+36YE3ZfbmlHOp=amMQz#O>*0(i{{$C z`jNMkwXVcd%{~FUCdNP|VonoYR7d)y-z0%%MrzXL2(;O{U1FLyMeL{i?Ejd0%dj}2 zrE3((0E4?b1b25&NYEg`-QC?Cg1b8eNP@e&y99!}ySv@yob!IqP5&h`J-c^Tty;Bq zt)HxFqb&0sOCk$kX7h~tfVo7BExCuej8b$nTwHRuv6e#_5$SVynNJjUNvbXDHS)P+ zhUX9*_Kb%jK}Ogv?Pw{c!+YwhJ5L=1fSvVicle5Y`HC8hm}&r|_0-3Y8nG|c(;?1r z2>yZ_KDBfS<%9{P7%0}A0oVKFz0Uq|w?<@#2IaNCTiNfIdLYP(KMQKco=2N-%*x_t zzZ{#+`N$w(y>@+k+EKAmu5m0ZeX3Sy6GXH+EikQS&wE;Y^p@l5{j`&|=PS^S+ohgc%7==GHo36QAyiSkvF8M@ncvT1UDJVD$!aM; ztyjltSrB=f6P%}n5V-El#W-e;ym$uMVe+WHFw4vcW_3H%Z=`mZ z3u-IVN8Ga2b}I?Ji%#qxI+qWO8Xz!(;l}>kO_*D<#k}z=A>+6)`a^)q7Tq%M&g`CK zt(J8NUzY)+HIh}&+6yg{1LqwfsA7wSzo}85v6fd0*)nfz!#w@>sAx46AK;Z+eEa*& z)PmQRd`qQsey)G9_eeY*@&kP!WP~^VcMi${YY+%6dX0xoI#yzE24}x%KP)$AeRc|_ z8wXz$v<{2OcvSz5YNBh67{&~{1-BC_0VYe*s8$nB3l7?>tNAC02L=l=HOn2f1F6Bj zT~u$IYaScy+M`6YL~cdHWo#MWnp02>=)KLdys-Hboi}7PbFd+6=(b4D22!Xt?l&Ke zm56sRNl2}n5j7WTHNVtYGl5d4AY`gjzw5ef1m6f-!*&`)c$s5%yw4uBXF36DX^?I% z*0fDsJ7?gADJ6?uLc;lwLHPu~CP=LpZBojzX2ggwH)~O*KF=AbDQ6Kb%%-MelioZC z>fC+LsY4;RNZWzw62+kN2^kCwRP42&6OMeduTvm>rfv|SE>g_34k*@ItNWr5fq@+5 zpUyYHY23p=TF`2DiNlC5v^uD2SdBb;6GB#<+T>i%#5u_PcB(p?S240LrdPJaIzE>tj6uo2BXbPXA;=oV4OR4%yM89e@F z27zR@Km{{w&{n@Rrq!rC8WpKE?zA2EMFp^Qo~Y0r3-;+FUj4OPBt!>wlU^(h{T~k- zgzE6z`u42=Veg-WV_HrY&~@VjMo!WB;BXEXNs<)Rm*2;Y&seeiZU*5!fiL8OO(RM5rr2#8@8DtGPLu z^Ic$CrmN4Jvs=}v{abN&?t6uM59qvv=ny&;s`Q&4QbH0nyG78z|6E$8f;AsoM_?je zo`-$4uD=KUDIj+v!h-b3+8QbW>s$KVD@_TNK7VfnTYOZLr(`r@Yvif37Vej(&k& zwX8A+B-TY7ox42aM|)R*2aa&VxSj&YWqu#D)yUx3;PuS>W@x@se%V-vH_Xk{DmlqD zGh}j?HC1ZM4-=+?r*~da8l4#3Eug;a)Eu7EpMKTXIsW=oq+XA=yY zDt7Yv_FBguT2Cp|Vijqm1LAq*>(eB6g}jQB$5HkPLM)`dwrmu%jGj&vfDcZynHeuz zd=vKEgG^P96F7&16l0$D++TrLL;ZYwBe3M|GNCi>kX`(y^zB>U>iDXssMRkuntU*} zyokqHoA@tCvDd35qxhXI>91n(XVTRft+sb3OAWTJE-ogfF0TLH0&JvZ=dd!Eq*P6c zjx3s-w0|GS8P28|;7pj8Zs*aLOzXq*L=+&s@MNd0fAeXc<;ucn24+{Z#%yVBAMtDZ zCytZOwx0=8$g~R9IBeHh-D(fZ5GSJHM~Xxh2;9o3b~f|)eC>e{RYf)JL(6?loRw?r z`=VeyUO`ej+HN>;ara$Um~kW5D=O$~V@!}=`7g;i>nQyIdUTdxw;a_S`kWve5o#yp zBPL3DcFF?_WU2@pwsAuTc+*-W52M!U{gf#zGg=PBy}y0&GJ$zJlahKnS1_36=apaa z;agKXS?7Pz`ZKse;sjD~2rRnApZyApcG6ChxD@1TpSVw?r_sFwm&fJ?iIURjA~ba6 zsxyelCxFyzo9#N}>m5g8F6&6EV=rQ=`$G{lvDtWbdqcZ6pNGXnADUggg^OTq9yBxS z@cFXoJO*4WkLjv*!cGwDm1p1BZguB9`szb(j^A@l9GzflkiOpHSvg$E>@ZK6+#b{T*fy<^iyGW8?f6A|DJfmimL80$68oX02F`cK&oM|&w4 zj7v)1Cz@R-sDlWg`TB7cUK#XM!MoO0oM}uVkQr=_+c4fiVtN;beXl?~4gq|mUH$elZXEeo-85^*D5q_Vy;au^jQI(hO$PymN!n}v5UKX1>a(bEp5TRkJBATiuDlnM zO=1Nfy&^;1ZlWuqBMKppJhAt+!bY$C>^lDva}nZhuPy%#0mVvUlvClXenGTSMiq5< z8xshAg$8{JR4oFrd!mjyo6K^3NSmc%i`{FfOPY zx1SX4`TcLcf^&au#!6dtvDstpfBqQM?d>PxCsIm^ss*2$VI_mEKUSBr?> z&^KSQStgVo**uZX{;pD%Xl|T)yQ%sgQEiB@;kTBB%y$k|rSTnXKO0rr6?HKqGbOy% zyziQ@=&FK5TfH)nd&Ae`D89gJas<$8qn=aqNvV#7cW`+OhClrxF5QxIVtPgY^p;N? zOa1`{u`{|M(|tGX)aY=ZR@_RI6hdn@y4;gr>665$L4tQRICl!r*3r~Uq99`s1= zk5zb%#vc$FX^hWE18yP8*1#`(|0@u<7Y3?7G!%(ocRPAM(crEE1~?ZZ1J5pDeIB0O zvr&CKnJ2DqqHS_Xmq4#;u5D#eN#rk$@`^6_vXX{FbDBI*&p?l!!EMXrF{x_ryXqa2 z)4tFcAErd7QR@^2` zy~pAz|6P4Zs0Y7JXdEq48a~z$w>fRsp_jY0oS23PCiC^4&#;;f?HBz7S*wwDvB7>~ zui63SyR_K}<@t$7f%7@uw`ny|{r+I!MURR9j@vL5eZcO&kXChm$P%?q)v*M3bI^Bq$5|j)|O6}ENzWtGSl%>~s zjgU~r7*OD#)PqXvePaxG`He^)rf386S(7$r+BPwCvmz%lPHdMtVcyO-M+hzoF(9q=i zC5nIQqSu(Oz4RIZtCpMAqaU6HCaJs(>Ut-R%9;RBk)8Tw=_xg&LXZSbOx9=dyCzd- zyLB(XqlC}la){j<`H6r)JfGr|3{Zsbb3rYJXq~ji4TPCJbQx0xUts$zX)TxJv)-dg zJOZQ{yXn5b?YyBrtmTk(zd#b9`i~FBtSP*1=j*v{V;D2})bRAW1wDoH2DTk9F7AL- zld{x{kt;S71saUW2VxOvH|Ps*WaHD_PZ~8184dCUnu9jd)78c!F77Rc1JMqP=EXz0 zua`K#H5*QuCZ03OG#iEm@0KY@yw3dt8R>Hxh4(TYpN)7!VuWAaOk)nGrJ0j}Mpg%J z(Z~U0AX1`$q@*M-CEUgPcDILsH0%4LTmpr3yJ+yYNY(QZ;=` zb}3mE)IAfny}CJ_1R%<5aoH^IBew)(8~AcECqzfu17|Y47kuAdR;2Kre002r(+FY) zErrOm-^1wpH6=#M$*DY_gpQqjf^2HFWnkernJXHA{guThRPaN4^2}HQ)>YOr$R!&8!DQ4FO_8zvX%I6Ze|MR@~SwHrSFJC zJd=8?8&;@*@w8v%+fT7Ddq&=W*#$f#jQ+f|HxRo)lo0Ufjx?NAc^)^js{GI*a@_)4 zZskq8>7}j}9j_h_faO=imlj3vGvToKqZ=>o*dg(Mas!0_W`Ky4RQUDT+8UhJyw%ny zv?CG$SEA#D;9MDb`S;)PpU*bMzoao_2zfV02p|3`0BIK1Soz29wou8|Ek*- zl6tFK)%DTb1g(ZbE1UE<0|R?Iut{7(QSv|jaH`))?0K*oeXe#$p0#ne*T)uO25e`b zKu{nldXUA;FESulG+{LI@D$jtwHsU~X2E?Eyh*Hl+zlbHiFLo)J%NF#Cx|BES9!fI z7H%?~=QrJ9aEHbrMg4zQEI0VwwCKIx$M-rUrG|`XOpc*$-1`H~Vd^tZlX#~3RQNLH z0W$A4Djv5@rrjW3bPbLCB9m^r0GoMZj5rvyV;M<6Nds-L46$+b{xJJsAWiJk+>LYn zCZlN~63cl^?0EfH)Eky7(f^q^mDw?B);npM0{8hH2YAA|O$$rFV^WJ>hy7n-I);Rb z!sm{ln zE=Z_eX?MH3J0O!3WAg*BM(@`!8)!Y|__c>)0!T%KZ-VQ79|*t7|1^-s!3%??0VZoW zyn@1)GHjZM3Qb+%G4iLVz9IGZ#{`}Zk)jg_7Y)w=0cJSv3o&h*#|_=pNKTn}FAS2w zm9mM4neTDff0k0do)maVJ^&|+UMXr!#^Qlcp;9@a(eCBGaXpiw<1s{(-&2nd@^S>f z&BuM_jW9eKoBBe+IHfYO$bYY&yCF`&A2|)eAFVT+%KvYkA(f{h?SxNnSoM&ee{>t7 zOaVg?QNi&Hr6zGq!XT$yue9B*9?8H}dO|isB4rD>SLYZt*yVP$_!# z28aPK7?e{WvsNPn$zMVEv%&ix&X(|Fm>;t|09kUyVjhSk%|rn?UYB7`XR9l;7OVtD z$SY{{W=*j6O^RW1%oL=|mUae^zKUtOA6~D_3eTYkD(oXM) zPl=x~b{7eMVP5pTJzgmM$tlkHzVh%;keC%Tdc`tno0ugu5Int+f%VNUCx^#qi&jj5 zNODWtq&S35AkQzUEMO*IRJ9JDwt}un{Wq5E=N)v#_uY`_1hz{hE7C_@l`rQqpYBzL zTl90NG@Zz{DYdz2J+2;EwjiSF#qAL9iIUL7f39wQ})(=D(+A=m_t}YkPQ&K-3al^+A<9G3bU8R3+Ou zSfT{3bHhM2ClcJAh^G85=10wkn5f^Eby`iw6E*Pl3@wvn!wyHG&F^;#;a_jP#Mehx zqW>=LZBiyDYfFW*oticvZdwNA-m4`7m!F@7Abv zvXQ0#%@6O4xdT1qduNxuFf5)Ov))yQgN^G}JwN@3eUE>+qhh%>E6taH2RwfQ=nWW% zvZA5}gEk@j*6MUTub|_7qYRq?OYA1XBTF#Qr9ICYxeJvQ1(SsOzZgRVHNW@zwDxw2 zSvDbfmx=M1TVNG6K(zuSfn8*QZ3w2(-pW*i-O)5TT!y0v+I(@2UYP&I0T6hb1S*MM z72AjmyGReVUcV;&`t$;>m`(l8^l|+wHaF-}BLKxhu^A^O%x$zWmPC|yVWQTGKNN`H%tS20>sZxrDC~prbPhQWY!_*&sXZ( z?35}CM_MDH1Et!30S*|Ih6j|2*|w9cnb{<%meL7De~{lV~JSJ6F+`%qycpr`!qayHNzrMR4AL zA1ghQ%(xjGIuIV4`$?9x7u8&1o;0V?thAgd5yEBhKfiMVI5#z@b?tBi0>>&P6I@)gYUI3 z^5`~Xxcc-1)43rDejiEn``zk3Xgtf3wB!CKmB?3a6eeKQf#XAHQ}|M#sxnL^k*+u`fmkY8l}wP>;KlI1pJzlMvb*%P3M8W9%e7mYJ@VYzZvJUznM0AU zdY%bB+|21p%Vvpx{%`Wwgo7%rPvbC;*So@kZTq}#NRG(ue!0!CK5*9qaipBD*(AmBWia6 z>3>ekB}fGI>ro?s=2Hm-UR&wBq(cx1?9)av0nEB7fChP80Rd}ZVBCZZpJ+d_* z#nD?Y)jJpYPF8hl+96;w493S?X4tArP6Ws&ge}&ZD;dqo$?xcT#O)BXxu6ar^#YUP zg)`eq$FulbD7Se18A>C^;Mgeng}9dorN^LjwxOlSWYzsT0bezy_eiF#%qlSvwhYvN zDOmI>g8&v`WNtkdQ-=`$We7=%x^D;RhphvL`oj+bLnoDN{d&h@_Rmm%6oD~NI>vEY z{u_LKI%D4IkBV@kSpIL&G7={}-?UN-lK1Oe6EjXw@nil-qXgtH*IOF5&CW6zTnU>;r2GDZ;dQ#?3=fyUOb=q~>8b+|U zuA?G!?`pQbA(K?yIlSn(;gZV3klKwfHZ&>;{iX|7B@`28s$OdxJXb0r)2sbEfE00M zZu=2wH|-5sr1ox@LqvRML;w^GfNpm{82^!$c8FF7^2gLV8%idG-#a4xh05}JaHSl< zzA$iK`G9@aA-;R7GlLR52*3uAKQIqP{XNpt>b84f3!9iq*;R##$hTuF5ZohncC#QP zmc1HB24Y3}2yB>*_D$m`REgJrMg$yYz?uYKEIkQ~=Gl8lyjG(|OMXNUSEHc*@C565 z_(8%Vr%RE%KXY`bxLS;K_;TE`KI*6`y|=U}-7AW#@Mo1q6jT~%)b-ey_s3=YdpNKB z9biHSiF|wus&IH{O=Yp~k2y&z?4iyXY=u|*d9L9n^4lMR%6eX(NXzR##9)&kIi4RY zbn54iTVG{mo$WFGoirP4obQAR$YipUQOqeiM78WLkRUt8o^MCtnCGcFlLemzH=rB+ zuR74yCc%XWs5E@w%FA%^hJ6$OMckvFoBy2BH>*pMjj0OcBkK&k#sD^Phx*1I9i37>JCq%Al zp|@wp1t8NG0;y54&O*8Cykg%=q?0@fXgZt2(m*zB@>JQ zoT1p{U~In9T<3JMIE%NrN8H+Zk-ZTd0TzaYVh!-yaEcV_7%}DLIRqk7;!$~vP0Ic9 zu$hxf3Qo7|?+}Rl^}o7w2(}W9Td@H{3DhI4?pK-X1GLpdd`_r{=YhIbTbpA0A6z4; zQY`#}hx<4|4h6PVH23Ti0$U)VM=-K*0~-uJVT20&8Y1csK~N-9s?4XbT%2=5nXhE@ z6c|;;>EbGy^;QQ~Uk(4R8IfPE!?P(IFR>?*iVMZy@Jv!?dmPeolUaC=^|$jTv7T>b zk}C0BY;=}&`x~k-2^j<~yTn15RGa^$!sB!L1sI{yvzd2;7BZ`K(3MmENE(M8ZXW^} z2B?l(`AK0!E5w4~#`wEP;gI9sS*=n(%0YPA>XH=E4~pL(5n35V@p6=Jt_$b;D z(%^6XMHLl4JXDl*z$Qeg{Hj(Xr&wP+)DBDd(VJwS9)PQUn(oH;dSbrjoe=^bxIQv) zfpL+qL=iI1aG+Yba88;8>5<7WpYT zZ4RajFCUL8S~hDGaf1!aokLJr4|CH(Ezc|^OQSA zD<7k5iBLxm;o>Jw@bK#se+d8rE09w2NuJsWw1tPW)R?yH0E#U>z_;iPT7H-dAvY^UMj7tj(5I0z~?E$#DAP? zFek_1{=>)_KXVz#1}l5AnpPBE|HYIcVFHIa?kC^PwmBAz=gr}cTug9UYV01p-b?fr z8!ee|X;7-5XT2UtDL|1>{+xNBH~FYzhZ4;)2*E7h`np%%vQzWzQF>spZkxn9W2h<0 zLQ)Z020<@dZpEL{BO!|ta3@Ay@PZp4zPD}3Ct?_@B`q8=luc|lfSp-9(VQO zfDmIY7!ipHgp@QQ8sfjKYI@l}h=(A?0mjZN-lvn|D9-%Ija5GS?Mb0RpWavp9HkB0 z1?d@InLD1zM?nwOJ17bV)X?7ZMHK`0a>9FyA^UT7J2j_hqRy zn5#dvZ?Lon`$KMl?G{{bm%ZN9?L*RLekm14)7t{+z<4VE^XZjO&Up`4bl1SeXm-XK zabh*eXk^uNjIz)BNgLV{?O$>_WSvrSRFDfhA$ZNx`;k6ly zOz^7M{59`+4Ik;bQtid7av(IQM^k1q*6yaz2ca;6% zOWvcz7(D0SXCsrS6ACk*)tJ!QsbOTLDumAj^bkVR>SH02SlbKUnj=*=7C#cDXX9$v z(48vy1`}&R2l8$CtsZ)gztyWUTe&>*1b(t_f_tN83H`lc`vbO!@6*o6d%i7;bS{oC zqxLRfRv^}QJYPyb7<59ISA@^0B3n@lS}0F#YhvMiPrW?f?9@owz6p%qITZc`GoG_& z;*sNf>i%0n{NigC5*L_2DD|6`9ROK^A>52`+&>|wTX$|0lMz}1c(9?Ayi~Jc9O;y? z`#zd^yTGlGiRUq4c3?W))*aMg1i&Jkrs6DxR%Y{-h$$L7txreUoo;Dl~9Ds+DQIG zt;R$9n{kBkD7AJcz#KSwWl$-OtLT3e%a_|!+0=JZ+d6#+%4YvSJwz2-{39rz&mn%jA$s-H` zbR^DM7)ejyMw>qgj!$4;)a5KpWUy}E=ExLZ?gQ1b{f}va<~ql{2p%HSC3ElS83RiM#Y4i#o#!J$Neqjxfp@6W0!8 zj8xnQhC-waVGN9rJTRO%dk&MdmoeF^N%=G91(6 z>qH;Q?|hgT%nkNl2j64ZL8n5i%7bxVH#Kn=EqP>?2jeJK#_75x92v`2uDm1Y&D7x^ zi8cN6I=p$wqWM={VYa7!%@GTtbkv{-)MvIpBL#o>zSm?iAD<%rl7BRY-wH8 z=%~^5Yjm4dnEc^3PJ$c z;T{{t7AYrXpBEB?TFzHp30Ynkb1uv;8bUu_{+_KIXD;Y`0H>&Kdc|Zh z^|ilDqXZTkF+SlN??M_egwFO6icr`Gz#BzEVuSl1Dg=%hvlp0HALEg1+HvN1ixi{$ z|7HPnx7S$?r;Rz)y9^MYiw#ST&|z(lQr&D{EJ!a8mk52ivv*S#cz zux5QO~lXcOyix1u_6#JWe{zzPT2|`5u zyNsb$g(d2_6q?+L zn#|s6%|5Q$ff&ZV6ggXK4G1D&X$s6DMLc&#KzEK9D54MNsLhSLW8%lphA?^8IkdW zy0MH+Bo$hB>&k!jz;HVM{I;=GdJ2PpgXbfK2r6AUHDjD zPGhf@$?Wuv!ntC6TQY|Hmlyr{17r^c!CIP3@5Tn@=ZkcbLA$R-H{TK-o_&&XTE!Nv zz(GkB3zdVegM4J|&YxfgwVb}T?8p8zSCdqoP5j%og^w*bE95Gh@zoj&T#_g~&1jsF z#p=TcK@X`0jjHu5MvBhdiPT~yejWe`;}-X8_&&q~B7;0cXB|5Hd71O7Zq}wU`{1{x z6Bl2ro+gQ#{+eDr&KKTpDKyydb0r9{pRm>>yH|nfeSeU4ecPI%VG3S-Xn`-4p#$Ax z&HWEqT-V~C73rIY!-!no!-O?$*Az#pdEfAiHkV*c?Tc#Jh^;?JJ|`w(QkN>z=g`d3 z|7pUZ>KfgS@Z^8|w$tmNm~b-Ltm*$jELAHQCV|?`U~SjB8`3nthqZSX8PYCv#Et*n zcuU%h<)SfN_RrUst)rc6^-N)HZj-=L!jAq~9MAhnJu_zH9FL`W?(1S6+fn&e2uKvD zeiUyF6aD^#NzQ#djJ(GF=KZhev5r?X7{5!kn!Jp6`Z6+Q5`ll8+4JJKT(&a&U3gGK zbD%^Zo;Si1!#Z{Q{dUtc?{hzL9@sSS-d`}{l)XLBstfMWtLhq&yHyT}`IGET{(Sfm z-E=t%&O;)SUO1U^!E0HL=_TVcM0Z z+jPE_T$o9BkODz(u+J4KyUCa+8%qe}*LH?TeU8Lz6|3*&PwIbM3)w2NK0%>lH4~`k z^pXrujklCPgkrMp`!^`j55uP<$WDiBeT%)Z&v;ms3K?w5i&-K0WVFJkb$iNYW&DVC zfg@+CD3f&~`-l$Y*#hArH3X-Br0BZsF%5~4x4jE`X}26&Tsv6SO|q&rHY*!{HzMH} z?i{vbE_bb^&KhX2BvcXPVyV{*7>Lm?kC&;Gzf*0tY{Ob}XmoKi+U8_$mr;=^`-#Ai zu<`hP6FTj#SRdLQrksuK9R^&K60B@8a(VE_4hC5@oMbl9(b=(cM z^1khKgl$SZSRkdp(H%cTweDJfXf@-N*zIcTQ%@DGf^#7~TCB5N;R~D@!W6NqZ461ojXfce(tj zG8`}|tygGn)k6q0m6I9=;*9#_XF9_1wQ5=P1UktwjXJH;DgCFS7>cwCXl4;oT~rn= zHWETE0f}|?$5zdvw4a<2-6qgq)4V^dMBtcy&LhD?Dq1DH)ej&xF4Scm-$*-i$Vc%> zM}3!h{6O{}e8m)0mavlZW-J!)TK$TjkQ8Rh&}7gDZ+e8~6ZvHL5`xH9N1gVgH+}am zk+E9mK))m;YPk^!AI>^R?i`7*T{zh>?w!9NL4|OQ+S?>}J!Qg(Gw1SUGo~L_PT;H- zZt-kBHmySV9Wvsb$%ijoKdmMlfPAlCfUv9ixPFH04`2xbZkuBmC4j44ZgYZ;914Rx zrfq3$-@|ZR-N=p@AN0drA^`~81fZRz6@o%?ifLn&mtO4=7^&4J8C??N2W#N?S)rRR zgyQldV7B`6)qD4tNW5y?X073JGv$+!?aZ$@jE^#ID@sqtCys%tK379Wp8uHcTMclYG_;hr7cpaMZ(i!FI4=JT_ zpuXalJ7Llzg4c%MBx`$gVQz72w{Qm$K-TF&G=batLSkl=+|`fPj84{*wwc85nfhSX z;29AUoTI=M1Vvyr>DsveC5cYiK_BJ{TNnQk`^I7_KlnqVtwkPoUUtyo zT*cWvtB*a;qq3-$&RPM=Rw?rpvD2uLu@O0DIUt#Sy)P)R2AUE{VI@l2DM=YSlg-LE z)&=$o_pH^`KAVBi^uaq2D=l|}kI7RK^dHuC0W~Q$TzM{l(_vz|7`@!snsdyM+xz)> zlBiI}Z`QWG3~HWoFGLIsPuWJk5>XhFKdNBOYTv&MQGJs3c*}pL!T`RlY%rY+=qM%1 z2X7!Y4UKf~?s30;-3}XtHQKIr>z?-(uLQu~+BQJIBm!zaJ{a+>8-VSvw_c;_X(ymU z8Pi}m`O}cRP3Mt_+r&cyn zH~b&sb{t3$0shxL2;#oS_`3gmILIflBFNQVeG9fvbD)+T1`khP(S&N!tEL;fL0Y1! zd5DI@`ya^n=&ZCgf1F28UrI{%w~4u-@?(I{9pHgvM=yf#=%cfCWgZ_qHepGe%!K~W zN2nGBd952UT>XF#5C9h2y#Valil}BSpp+sCmh7ZeDjq~>rr0RhvD^vYb^bWCpbQ(@ zWx8J#QpK9vWlx*&^DTb1T9Jfktv!&#c6HQy`;P25z^sI3HpAA%@k}u&Nx|qb5G~?4t31b9i4g=MVMiZD_b^Q{AL7w*G8C zU9p>t4&L!^1W&{IbFX>T4AL9Fgrca=KR^2jRCj#dHM}0##N%Epp9GM;I&c)6qH938 zLPr}}@LGPM+Bw4O37{(Ou7fXqqPP_2hCUrjC{P|VRaXyB!kJ$_=FhS@O*!mr$!`D9 z3587?6lH(Lqyly7G7)Uy31u*Vn7Q!$&+?5v#|=xV4usxNdzC#Xr z54M$Ih|Vo~gUE{Ri8r^?D`ph$iPKl*z;&x;a=YDqN<}Tg8rzOzUh_7Rm z7^P^fvD^K(_*)8~fta)|?DC-wgoDAI^amtE!&$DY2WG6^nR`40cBs*q5I-uXqD*v zp!jm9)Bkd_)%cI)3hf@(ABvmW6(U=gXLX|%AJgNIvdjC@LNrDHC@rq(nW3W^w1IIc^)-egjbaksKBhs%WA|c<8!5#7ZJXVH=aH>drn!uMrl%HU{^OXtp9Jh$g!`cEF=1qd?7RpYJdawi4n zD*Hxo|vlLc9!g31(wVpI2-XEjT zna)n=^!L_70ztra&qW)=a*HLo)Jnyk#m^h8KMXk@Q8|H&Poc3+UsZ%v{!uRCXgTGz zUS3+V8BIZ@33()E{;jPiR4*x-srarl{oS5b8Q9!aR%M7f^>Y9AEr?WRfF=h;iIqh+iFE`v$b_g5b#yMa<>gtjfP5f;?R!IZhG+w?Q~RDTNcuOUn&ZT%YeRm@2re z7Y8M&vpxZIAtnVM1AzZxv~+^#sS6cKihK!h*WgFlsoS}diGWX zVXU;LStd!T?y=3jh+ul^FCDm;X#ReKJ%{yD~~0HCtNW6|?N#AE9{5&EUwQh$BAvNWZiSdK^@g?u-& zLTl+H!(!0;)!nQg{lR_~=0WLYak|}tQX4vhbmaRKr`Q;b_iqlj$8;n`>zwrIa&3N2 z-pSE$?AeCui`dE}UW-jsJkrV3s%qq&e79QC^#87LuXRMCWem6Vxzylu2i?pAZLN%weMP0y49&@>9|TGgMZ zZ0E`MN@QkAOq_ogRA0IfS;{G>Kd>i-g)pd>z^dCix>s3Xn;JWCug2(7b*q>#OM+p* z+i+0psoLiTRALu3S!}w_2Y=j*3+YQV8b%-FRH9!<`6c1t^d;E2eMZTvfgrIm8zL0U zgfyQjR~GcWW;iT9ob@A;t8C_ z!ilLbykuF7ziA-tLU4rSP`~+%9x`dx%d|fp($BS*)*Iwjzr8+<sXuf#gcM{#-?}n_PV3K3=yz7C%G+gz#wYl< z9q7oF!%zCcsP|4aAR03$af?yvMrsH#x`C)u7O>)wmHJn?GRy+m%u*xZSia^%(jS#%Z~T9*!v@o2 z9BfmE^gn13jIX?1G2SCS1mqESyDdlRs}WOAJ>f?+dz9%q!?|5MJRmEE8m)l9a5iJjD87} zR6D%zHS^Wop~-MEMS8c*Oyf?Sg#(Z9CAn=3RQwZg#T&Z~r%(J|S!{ z#%HD4Si0%7*>c%472%NIFzvgXS~?}rlBc9fN=OHKI@5PJOj7A|^s*%Vi859UlQ4fu*ZlLDM#syp zv&yqt$UeMh^~M#l+A@E{W$DJ#rin{p=oU*++bQ8Jb%x^eTcbtoCEY)u`x%uZB%M8l zReZB}LfvL@QaoG|Rn0@GTr`$OPvMMxYN(vA<9JdRT}3K0mFERD#*KS!FC`f2q}SpI zs8MYyLpKq-kfE6*rrqb5aKl(JF2&KhAPrTuos;OSsfXPJ6D>~z9@tE81(au$TSz&0nNbidRL= z`;Vq79*DH_Z-{A@bF${au;7_kra6IK7YIg0SE2l0ndV4cc;ME+H~H3k@!g}t{f8^; zL%M~->EDI~eNkM~fZEQq1Ql=xfez0luxTWdX0>&u7N+p&#T!pFZmn!bdu|o&e2giC zDgIF-%6H#dB0l});ZMqvNkSgxIz(JSG@r%XPs-!yX7LDnL^icNeONouzh&R91g=?g zbIo!a4TSXd%y90`_ZX!@x%b28M1@~JPPE;JA9uJ#)bc-Q;WZc~FlaTYn2ZvTuX(>^ z5iYl7XnC$Uk4tdbQIn92uYW9YX2H9Nmn*8uic6|JH>_uSA9@|o30p?VXl>?&$0w6& zN~4!1ap;p*%RLCGhvr3)MV~2GY*D@czZY^-MMb#Ck@>g!$t( zRux#m*z34w_Z#tA6Y{sfNwXloDbxZIArh+bl%xT<4BBtNAQ%qeeLLQ29h5Xa+dzi3 zSUN}S=mZjp*N-pCVIEtu+`R3>o9K_^-V=lmX{FDA0JIUBa&Y+k1ke!<=e#|e7Q5|e z@8jY;#|lIW17_szj!mzRN6u`{tqEbfjGt7p0moJR={#=&^Od3`z($WO`mW<@oc~k+ z39}tgdY3$58&@|g?U7=r(?fTVqE@~do&`MhZaDgnsLO5XchcxJWy)-ClgGjH?Rq^& z5>&@aAkf@JV-u;`r?lPqF=*5+&n_8}7eHU@KXMvc>S6Y`ws3223Ld(|FyXQKm{vyX zfr!XQ^Hd$@iU#N5$BZ@GxD-2<(5lN^s=wQ;s-(M~Q8+I;#XD)gRJ*Op;PF0t{w|`| zU#c}f!8Am*Ohq5A<8|B}1iDkqfD+YjNiKT2#Bp!j=JVmZmKW2SW?qD4wLpu99z(yQ zba_ryzg%fgZk(fQUbz2j#4h9ZpOOj(SU(nm9}iH-;_CTJOln9Ez=ySTu@~o)Z~JeP z*0VB+dx$$+J*cQV!dM0h-%0X+|4LImzZ>4|yV~k*(}$~R znMP?6TyBlUwoJO_hXb>+z6T2xQggq~{{mZxRRWEt#XjUWgte!URro_iA&f8(amAxX zsiCPHkGk5YzCigpE0geHprs(Pbg@&vBrOd6F-cp1+KOZD*{woVAqAz;4v&ay5e}JA zqxP%Bi;#p#5!LW0zICDE6_cW4M~~5$lOn~IF`wyeKZ>6oFEG^#s@3C}!$C7VoIQWL zFGMbT^ZeBFZa3DW0QF*k05E7(4l*Jijw}XFIx)X=CQml<^s+o@=(4-PB``%?N%LdF zIUA2Z#mov6+1)+O3CZMYbFYjS&QHm zj7)$8bBE@Z3ZYi1=l}hWy277^sf4aN%YB2BNOu-($P6~`z@LRhEm(^*0uB!UP#ipn zQ7(W(VNTxJ;+Gz`$7!E^UxcvAD*D71#POg&25^u$(Q5N4NmV_3x+w942{TeIGq#jS zEjMwDb?IMci-^8c@o5<6M^`JMbZhXKrX!MYeLs*SGGD5XItn~j?hic|F`8{y^U*x* zc=KtHhvVxz6;}CaP?9Dd;us&6Yu{dw8Ce39pY+z;O``(_7dy z@B*xQ8Y?YEC7}lrbI4i2buiA-=hyA9tAmjL2;rJUAiQc>KQc4IY~SdDEp6Wg}kG`8)D(b%>*(L49EZSN15`Ep@hYn{iw zANgkOnBQxE7orqfXH)?d=3rXqLrzM9yN^AY!kzSEcio?7nLrDKKMM3_&u-jo?7Gr; zR8+X^6W$(Hr_zUuge5>f0BonuZY03?v}&NK98_3jU3`Do&B+nMfqWdBS>LBdgKqSF z2)dWdM>lZpzPmg7bd~sB!4~6U-gNeu|7#I2kZQI;8QIcl#F6cO7suqp+#WPHB~S^f zEV;;IxCuzBj|S!u5A^n+=OZ2XiwsBLH#y-^6sqo^K!t~newp$D=o1j)vpeYY-l^}% z*duWO&E9K@{R3h}xlmB|xADxgwgLA`+@l1zanwdPY@3Eyu|iQ_MYG(b44F{O9jKW5 zdcvAug@C8`i<>`9k6~uFlWzC@baIjpq;kpl@0Ta%qv@KzO9xXiesA?8u-v%)o_;lH z%=Uciz6Xg{tDEJV0hH3&a+0fTO1FuPh=5OH2JbsfBeNU8sjNm`*f6e}>t5#S+{t0& zz7BXFhuVlp|8e^2j-7G+SBU66+4lrV4`3V5%~bbxz5%4yxq^Xp3iO)+^s9m~jYzuJ zztUq>|NbPlfWrK!oV9NpOMdx?Wj_@;$IUm!oliz1Xu~4RlYKn4tLk!!n)ociJnY%|lgxrsrCMs{7|}u}}!1L|vL>v`kCwVG*|)-ebZ5^2|bq z!HLI{dKZW^@2I(f6Mp!@A`oz(Z9r~^gP#?ygP+}cqAt-$v~0XLqOF>a1v{@HgADhbs!=-a3cL{1*xP4+51Xboiu8gJS*`*n!}@{0qxu zAMzkc5vw~}PsH#@Jaj+PTK?X%q{BQWk3Y|1$**(jH=8X8r(;Lw7cBu^38%jfuHO6s zEsXN(R+m@pu*y3T3tMTXm+O!iTI4slD8I)L50KdLw+!hQq1q&Jj$c1it4XQ29z5^ z%MatA%SZ9ZNSxf8{#eKVs59fY0vet1MS9rp`zsan=OJT zjp3#mzrgt4|Jc(2sJ$_!_HkOqs~LZipZ?LV7*&U^qQmtS>I%G{Z|7)UR)ZSk1=TwC|h{J@b_<8cKG zMbgswdNXE6$qfI~4+o;UuF5290+w}Sh;+#aIMmi42jzb9*CjQ4`t>&UTA|_M6-}P3 z|CBDP*@rPX6)NRsKT3YdD74UJz+}PI7jmx-X7fh+EiOzk&Q7@P;lAr4LEC{k%;g1G zx~zoJhhxu06OYv7oB&}_Rng5`(KKqiI=b*E96N`M6)CN{CO58VR<0E!o_P_!LMW)o z4ASay)Vn|zXe=`OsEknml3!Q3$yM^QY^=ezV?CWSf-7N8xSRWUlPlv#%;zZCoZ1S% zk7@)%WBN0$C?x91FOMB=I|fFx)G*3KCrXF2adWY$ipRX{~&P6%NCT8gMXyg4xap0SiL`-kwYBNBW61GS59*O8~ENi5r@XGTJ

    4-G%PuH^UenF!NiYk1P>G_H#-T!_zbo(^1^>w1LI3S)KNkMKIp#^b7;g4T=fAkk-I*nrm7nbAtv5+V2=^;wdZL--o$FN}K1QOeF(}V{xJ* zaR3e&`z`bEPKnTQVQM_K860#2$%Jvj9^p)*;=Hc>= z{`jp~1Nxyi)hd}1^`&8Z|w> z+HQPGO23;iAE43{;^|3!>8i69LEd?r2a?qeJTS7Oa-e{-lLe*o+9IJrB%;Mw4Kc z#P0f`8k^=|#IkR$C-vq+=F4O`Jk~X@O;DG3JN@!4nvEH4w(T5z4b4gADhe35y<;RH zU<2hJKewxa{Rb-D&fHq<7BOm&fDUHoZ{A{he22IYryHU#LrhNB{ciTI0Y)NcGU9|? zw2OueN&iT`7lcXlP-G)|vS1mD;plth82p|8M)SF+mTh`daa#XN`7b7i z_+m5jdE*OCKa{omSh$CC(V(&-Q$O;oAGPknQjVp_%5Ey`HemnqgMg zdnXe1mKH)<(IE$3n-GfyH=>yLpVE6za2~Im{bTV&^TYkA1fk2W2sel2ydf59HFMRAI7ep zFqg(4P9-}(Ivvth=mP#FNYf@M(M)g$5l6X%>ZS`2C;&CiRpU)^_1<38Yo=f-{GS~efG$hj|DJJi9Xl`eah@v-& z*%{i37@FAeUu0FcIuXjbd_b^9=`jN>a@)w^@bd0{v_aSGAOaq6a(1Y^+vwCxgG1ig zDF!X3>-f1v}BthUN&3hH{jVlP_U7n}G{u<4c)OS;2<7G59C}iL-1)epL&ea21Oo9D%;I1$HnpqmnO-#`f6Rg(MWGA#xj*rKEEl zPyf^I$k?o3y#NJc>b}u&IOJVzYlq~y^UQVJ>MPmzyB27m_Qm0Aw^>z}>HuBmT~amh8}OD{zP~1*p@`RpV1I8Mj$?Imd`2n% z+mq=a9#P50v|-_3kn&xH2LG~5vb9_>SRo0moU{d=uOdn9Yq@&zU(`&Q%0c%1`8ma8 z$f3{FjM&YP=%RzV{`f`*NTIWu<5Ui{T}=+# zlc4&Wyr%dy$$;2#0K*9ainlX<0LS*ocQdcpo(ix3NpEToK;E>sOgJ6uvvaB2!#p_< zqGW2T6bPStIAGKUrG>|ffEmCb`yZydCORTY#Wn)pO+&f`Cel-S2~zDl3BUsMWfbc0 z4bt;5fTQWmebd>M7m6zCzf%rvVhX=Qq6}GDl3=!&cf2P@LK*|IG#=xx7gV9SBOGg) zEXEOB4X1cN_*I`~Il>7)P5_I0ajY{LOnULqqReE1#;pyAZXp~`0R660^a}>_$X!p# zCcN>fw6Nzr1^ps!cyYhpFZNR+YD|$7jG;b8zqc;hFWtm8AW{+cVMB;I2e%nCbOYP# ziNm8s|kk8CF$$I;SjBZe0 zJBvn_TloyxC67CIcOvKzPSK(}E>ure+qf6ts~j_vbO2%=L0Lg<|jtbdlmFY2xcfA z7tZoadEL9-upaK*!DX0j?xribBkltE#sx2*j?Z|X37Dy3xb|{ba4tnd5eWpTgjIWl zAgsMsQwu9bkf8SXKpgsJ;K8m=D+t=>Q|?v-8uZrBp=8#fTNA;@nX^N@w|OuP`l)-j z=GXH1v0{g1Z~oLDtX@e~2Qn$qz#;EVlTSlhJSPeygQeYYIHW&U$UaqN-oEfU-o@&f z6#VeWxG|>0wM6tx4FAezN*>~wg5$YbfUz4taP7V8HFO!qy&;z-AZ_e~e>@Tx$ucDq zyDOwZwWF7_YWkK{&r<{S0@;7-48;(gp3yo(TH#azuYS4lIhKuAf`@(?I?gq&90(KO=Cd|Y5BkNeJ=@Q&=t7#?YK;nj3) z_>4P2bPw$5MyO}yH;vkW@KP17I}1;#&)%;mhjeQO@ZF~r zc)xRZKO!_L7?X^c6>-Y%bucUua+^SPlU>l3%VDXKa+53?%6v4JTVX$QSy_#(uC+sE zurlOO!M$-;YaQQc*Sonw>OO=M z&5PLiLhM+6p>ty9mM7n zxO&l1Q1fW+9UMj9fk!Ue7Ca7V906<(X8&r1i!6Y1MGsF{OQ{a}A!nNe{ci)WDBb)- zO=ea+OQw%;H&e#kjC#XS+Q9kRU#>&88gZ!4(f6tZzRvdEoK&AtNl?GgNFrz& z946km@47$fj}gfreis=U(e>t44nW$%Imn&r)FRNgrjdZD21dc!-D*_ZQX!y6bqO}Sh2Xf7k}+Z1!WgG1f#~1YC0mH>CUKM z>>fMbO7XTkik9Ch3&M>KD`^e1@Wz;=sQ4=k;CyuZ09}#=;bfVt-|&iF4ZYB>&h>r`MacsxxGXMXYPs$QZRx zY^t0b8c7FV@`KFA16ZaesIxKLfMLP`ct&Y{ouO0T(z&yf7-RmEivh2C;qAtDA)$Iv%Lx4Q8hZ&}I zz*&10{Z-p5_s!9UvJjdBJR5)Gy-}~C=1T|r>l6dXC+h=>f0U@? zkd=l5wsxj}dk3tyB=gNFO=cqS#L9`s>l=;=c>|y3=zR) zQ7Y{hgb;t1&k>(XTK|-dL{u+N0vy3wrLLq@eqc~E7lGd0oGMdc_d-sJ@Du~1(a{-W z^V*J$lG7KQP~RHp2M~R=%P!K*b^-GhBcW+uJi3vv@FOWiLjlPRpNef^e7NO4$82|8 zUS{YGA(z;2?*Gk2EkPZpP0?mK=;|)UG>V-|b&U5OLz3ORc}XQ{;(+){Xc)g-y$A@P zY2j`3Pf}o8GQ1hL%%8g!IQ>dSeITX3)NL07C=ge+E6?u3H`)t%51~LN?2&2mL7k@m zB`cio#KE{-Bu$4y?$_I8NsF^iN}k@s&IVLF>DPNBT@&m>@MqYakT}3P;sGJ3(Po*g z^{5~+tAfykJP%9PNL)m!Ct|DwI5A1Y4ztDgj@H@4``Ly|_;WXfp8hWRX5?U}0l+omHC&-OvI(kV_{QP84 zZ1yOml5@!el$@RsPdLJ=2>a)uE(cn}Fc@3mRW0`E^e%Ec_cnQs*tg@v9C}*@p*kBO z!-W2=(T@DaO|wQLEZ^t8CbbaiMNc9ht9ra##iw{~R8)LE|I%ou0~_#tlfoYJY`obX za#oHB6@w?(#32MR)5Wy*s2MUcSn2!os;LwcSM zNV+Zlzt9_A_VAC-Fc>#i$%Y-L{9}u;$lIU%1;zY)j@o`n`hFLDvbg)Q**9}j*JXGq zu_5I}z2NBMjcdZPRP+AhSwZ^hh0&BL+IxIu?9m_nQMV$-h|PK4o)3}B=?}+=Ta=p3 zf;K2WZt)u@8dC+9gYEL;q><3bn{87+8>sQDS(6TkJPGVdsaXuJ#48?SeV2e=Cx&{>X zFF-ixXCZjsw5>&mIi}RF6Ec`5>aGz;V}axLm}179VpQsbI2!hl{B06Whnl+zk{(fp zDf)eKX!MW0TcRG1o0pAm^{(e2NTLi%pUo-9Fb~MadCnIImdD?T5(<4c{fX@J-!4vl z^U;&FqY9#$iV0-`?|WI!cr~c4KGvXQR##bHGC?vC6Zl*rfa4z)g(0X#O`z{*aC>p1 z|LLSgC9#W1F^!1PTd~r^!e`b9-=>*MgO}@U>TZ@CcR3loK&RE38&=H< zj6VlWC0k*tzH~rLUVooS4w7nZSUDsrCUWpbXbfw?_Ta1m{8-WZ)TwL}(v61T*5_>>2%QxGbsUaK4T415vq$04Rz!Qu zzF~Xx=D~axgh5j+P(|C+($RbdRO_FgH3Z-!mdYnhMEI&F%$iP)hdZ4Le~?9IHpO)t z_(l_YH{{6vjd6<%!3n`EGCIJkhe7en@qW~3}F#lQ7-Q=?99)SDle9}<#~(K8osi_kuaH29e8hhonu zwKl(-`T*zuT*-;*wQ8n=ewaz=IkKA$aimw>HI4;ypUx%?#h_`7sJ$7&Arw!gxLTXe z)F`<+x#U3t6BWjX++J%Q{&zSSOq+^k>t5Kq zXY=+&_x3t+pu}U}+q`_TGw*b^akC(9LK8!7=!AnQN4G5lk9`;x_rRdSKzsB0*hQ{IXbxwhC+lJ zPl&!e>3<~u{m?3k*!2A?KVPG)+Ik&wx|e1gvszg_cNh(yspOpx1<}f8>+rw``8p2efF1sABjQ1{`!QOwl~(v&goZe+)ZT3kHzi2U$!88745F| zF)~77EZ?cI&rpzBmdNYDntI>g6U$BtIWs?%BJth>Amw*}A`7=gM65+_DK-Tm1ALPi zNF)i$3#hb7b*|X1cH(x*82&d*LDhI0arN8wVs(OPA>3#Zl>J!SK5}tRihnXEf>i~Z zi5y%hU3~&pFu^K&iDLFiQYnR0jEfFEk0DHEFoM^A6R`4M<4wV?_^Uhq(-7;G{b*(z zs^d$O>ue1`{n+QHtEF0dfXzk<(`~A4;Qk{B=b369N^t?T?{!0BpBctFzQg5{e%awa;+hDsYxn~z$m9W z8T?CCg%$U|wD;C2sY#6drWdoebuRnBP#mfIlYHs~fFqBr=PMsRig$;|G89^&-+qx2 z7FNZHm`p6&O4Bj;QafiLKGoO$IDmxo%QtAgw4W7@AMsPDhi{A~TdCXhQ|H9m;VKgs zqhqrY2brosleU63gANZzz{3B)l;t~mD(>;p%YN##Km9ZA`NQ$1%okll5pSbyeVD0` zMCxnjbN=i6Xmx1tbs0CkI>dvCrBxO%6_DE<&v>dKd#cKuoMGE0`&)%_S%#SAC-!8V z=cYHgvm~(mz>n)RA{qNa&w_Se@=u^|UYIE*N@7OPf05Z|h@g*XUE;`CEJ}QRi@gB` zqzLM5=edgi2WnWSXYQ_OM;nrH&3-yr>xgKu(UhW7laKpmJ!i5pJ5dMRz}0H4ssG+= zQNSTU!R+DBTTcJzv#J6Bypdl;0}>GlQ$~?AZ5-=madgQ#J24UaN4#G2{~ORkWI-x> z&O5F0dG&{iy(bndNE$@BMI|FmrkHB+nYF6&N+_f9R=bL~oGZxMK7+@XYh9G5&NKnI z8lql!V+5!n8+Fd8VDY!(=4+8aTAqeTHdb0<-R0n(F_5q3lksj)r$1~5Le`CELh$uD zc+6nZk!@!F7;~>w%0A)2T$k4ML0Ig+)3pYt^t!r?eVx}^k#xQ-BCT$x@QW{%hbE?A z^vLY3YRHSse>J|Z7jV@Au^_cd$jS4lkWlF@dm}Lapc2D49s4o=Ek}M+;u;Lm?~bn0 zG0X9>DXa5V7%Nz+m3P=kE2MhLsU(vy$Kmb5fml1y<0n?>&G#jQx@-mn9ayB&r}$18 zxQRv3zz}5&Wc^OAfvAuW*y5q+o`NB3fFM@+GR+a%;!4?T(zog~B58%Wxo`o(#+uO| zPt%`=ntieiM7%wwV4M5v*W`v|H3dnj&&P|(va_HXU!2%eZeBKn7bkPZt1A!Fn7XOOgY79G$v&ka7bJC$`w3YIa}xW zO-nt$9{46zN4V5kI;q$M(*7)`-^HGmHdVh|Z$9?FpxDqjSKWRVo~#*R^mf3JkG{_m zZo@~m2b#`0Dm4mo5Bw^e4G)xl&9D3YRQ(jKdUKh-#1C~!P6UIITz~y4$#oX060kVg zYkZdbqLRz;LlKdJ-cGB#H!3X}MLW0QYJOmhF)-G5(Hv<(BDv7cP1|`M{?lnrB)ZU( zC~P^uZ)iPtCp3y>I%S&Q5mI*?TY6pFHjfH()Xr^TX$>Y zhEq7_nl!1>f%MbHbZQSC)2VpH39LcBC(Yj)+B4ltZr6s)qIy2MPxea6zKJ8RQxYlB zZuIYIK{Z`6J~ZvPS&2AP5D(+L9vs$PrWCQ~z|2)&nKE$iKcRH19^#$11^nbu&F0}w zVZ$?X@C?C92G59{`%ze?Lk9sk{wefSg9F(ha1%r06+ac(b zJk2F!#d@HT=TWarv$XR>gBKT&TuJw(>In=dqs7ss29JO5UwrF@WpJo)w}R)ErDF;05w$|Q#d(dpMJ#4_oaHpQ+?U$g==}Is`BRHkG;uG zpH`ItE2~oOAl{Ujlb?MO)$j#}dEo9#YJBiQeRw~J zeRqejmTRoiE3!BxQWT!Wn`kIxTIEV(ICSp zhuYttx1$rkdZO?7U-xK!srcE0ip$bNhUIolgM3-bb_mJ5X-HUSthuW@x9&%^skFD% zVd!OdiraQw{Ct7`1_5XK+xCk~s@k4S#|c|^Yp05VgI0>7W-t;Ya1EL3nM#E8%rAGx z6u0wK#H(zfXhbJwI6o0oand4K$uPmrpvWf=cc1lW=>SaZre64U%4VB&lk9vc>GWSTqMUKNT{h4r$enKWd_t6g2v3 za`K7({ZwNV%)X`cS*N_@wlJ_Ha6xZgE3s?^euk8|AkQWQ(L2r00RbDs$XfvXEAowN zuADJ*?@*r~-P>ZcR)OwtQ1T+Yslc|czVpX2yu1Kc)HpY)X923J8K{3r}LF+N{ofbMlDRlea z7ABJS6MOmE6P!m#>o*TFw_`YT;5u#hP^(Nz-lC4k0&ofbE8x1OmK4OY&C-i%FzNTD}BrS5UyQl&f zD}2)pW-T26gJOB|;ys5LLPV_kD5!oVJziJJK5pyH<*ETTYk74yAk(^ERBSUmGXIjR z_FZijRXpiCS|I!6K06XX6FDZC@v3N7l;s)Q%LQcm;6X~?U!Bj(01$l_q54IL=vzW~ zCvLFMP`Yt3a%8Pm0*uihL=P z!#~A8#bDm#e2~Sk0+~C{yh?V1EBx0qc=8|%T>=NnAvfGIGI~n%ADFxWdnHd znx4N2!8y|>eopdVVTErlMXsz^M=QGpqgJ={8PR=!t3Os4h#-XxPiBaxPp~+Ch7l*U zf!ZB3&;y@ehuC_AW9y0y1{(JVi+WDzC1#tu?kgAc8Ooi6l>4+Nzz`At`pUxh{&Z0+ zaHngLP-83w5{(y`Ee}KOUF7Yl|1}i@R<6TWyWW7+v7q1cJqCjIc^y|hjAPl6 zL>=qFz0Tos-||k3F!p88Hw#|Rc@sJ#Vlm_V-gab-iIm=9FP#3Je>MW?2rq zwQW2`<@9H%eu5Y}YAJv_vi#l^lw3$3Uu0t;Wg$B}$jBthH&J40i^!CiyZ*AF|INg1 zASIyoojWllJz&!u{f;N|f?m{)13?4E7lE76=VvcF@ZqI5t^_gl`>M`ZYbntp$Sxsd z^OG^s=BQ?*!3&S#JB_gi;D}|hkQjmYXs_3`mIXQ3V?T>-t93lCkvOBs_YeIrmX>_O z)+qGCQk)ddJaw1SGpJ7Tw+E{PwYGdXeKaA4IIXK{r{*b!mlCb77EU*D)9Ieoo3j#~ z&0-qNPW*yx^GZj@)s`*DHuGHjVa*m$7Z}S zD6(fi5uLfZ1aPBLSlb{dA;y;RMi}f4AW>ZnM?hvtA^=Y?3$Oh z=Nz&d{R;`pw1=4oHI5L(X}OhJ%Z##{7z^B^IM?I?8MaG~g8u8)1&o3ELM6XfHFQu< zr1a9fioXkA4`62YWfnLH`1y=D96|Pu_P+_<+!_2YboiCD^k0gHF<_JFtC<3Pil=_x z@kea*7$~Lz^fhCNbqOkDOl`>|H$DcE-BXk*U2^+CT*6Y~(CAn`0wu}SVL76IpiqT~ zI8h`*6%O~w>zcGN4{RFo|A0`C*rJ)LJiGvu$JY9{8AGrv@JL)pn(v)o3NjnCGV45w zjvDJi%a%YEk%{VEt9Cj+A1YRd3_O?4+*}q`;G|8zoeRVfncD(3Zt847rpV09&70sS zHp}*G9lQklE5t%Xm@C0No2C^0`aAaAXZhOV02L(X{4DhYCYL!yKAC0$yN2%2W2aOfTPCL~Uw}XrSm|zXP0fiR-Kz;a1J9TK6X9tc1Nx_;k__!{NX3v-c;UW6Zb$mpz zJ$zeB#<8}EhKaP8L(0g}aFzMuV0o~?BhkChsdfG$)`4E}b5hyd*c~IBeX7#azx|8p zwtqgeXY#+=-YM>d9P4|-^XpV#oucKKST~3v~9M(0T>6Ac8Y$0fcz4^?qpE6A3 z3U2GYDS|MT+Fa7U{E6v;w4ChfWSYzM7=gYk;oZ;Mn z&hLGmpoFm)yZ#Z_zJJ;+Tw_hzF}n6w1Hi=cGY#3K9L_GrgGR*t@D#>f`mG!R`LrFy(D zsbde=xHex+mcw~3z!w$l8f{vKz%d_cLqKGrT{dZQ`pi(4G$#Q9t>wYAi$c6eHjGY~ zK-7DFeSei@Jt~M$UpJFGdci3{<`sFX*%m7QcZ-L0x=-uDtyTI+nvq(PXWrAlwLBYT z`%pxe0#~Pz#hGJ8yF&Lb+>!Z1d5-C<&mvpHkMmv8lWuMB0KL9r!Puc$(N?NReAe;r zsF8jl2+qM2X%apCB81!`pX&ZfZToB00!`w{&qescH%Ev03FXzaerk&rg}vRRcI!@o z`Ag0|rWsFKb1+x)3c)(;zyZP_l%feMzvpHKdi#&9kEC{UT|J?n!T2a*o~j}9pI^5> zQE#?*79shL&z@|s{$rbtPWre)|CkO-W-G(l$Kt+De4uLpuvfei+uGq6T$UxWD4hRF zi2r?~!?UmcFz~oK5J-=Dzw0<3#P_l^ECf)5Fu_e<2cetFVdZk;LZ6OtUOQD6R3_`2 z8YkcqKMs6-Dg};>bdt`^7oGjf|DCRoz7?pYAJw(GCv|^}BbsIk`f6<4R~71zpmpt( zGF&O~J{K!FR61p99^qi!I?OO@n(Y6HnNQ|}-Jv``Q2w}kkNGHZ(fX=6iOe0n3c874!yZUD}(e`wFV%?e}tKIjJ-+uZS*U5!8cCvt?NZ^A`Z_K&{ zW8lSd_5Qj-c1aE4wB-T;zeK#{G%_5dKW_B!J$T!sSZXPt&UGAz?LibLc8PpMe{_22 zNECh%ad$3$e~#hTNZM6=QGdR?t)>qptZmW06L5Fcw-omD3Q&6F7LE<(g{CJ2z5DoB6fLs(X4R)-u2j7eN*V(cs5PtOp0_jz5n>$= z_62jYNLrD;5wsoe9&d_Y|MKbcyc!+g#vU1Zs(*Uo?J|jZd-VNyjD}m0>Enxyk1#HR zUUwWuXGTjqTvlK)pUiT@Zl_qe+JQ9Z2!o(pLpK%G%i8{cen9@pGKeHG**>ScqT||| zI$35_Ur*?D+t<=LXq+)fL2GN1k^1$9hW(ZCm93%nC0V^{-G8!K9X?&oM3KU7HU<6Y zP4KAQ@?%H&gmZ|7X}xJx18)fC$l6U-tD>|<7^1OWnWS=VuZ7;j{rFU0{;k}jV)Hn0 zDC~Wu5oo*lRu;#~@^qTho%rx``>RjU8mC&`nmE<0aCGj~=5+aM$fX6Y0YTx$P}WVO zjsQfQK0l}7C5IWw#8$;vd=gNj^ik-IJYl8ivJx2$9Ia|+E`-2Lxl17V$ zkP&c^dA4drJ6C$m4v3`rww0LsYO=b*L^#NGx}Y!p=6H+7{A0e*;K&!#%6G}gZQ}BA zY!3wuCU?loJ7U{^*E7vzR`Y@(GGe}9cT8sI#>h6>IRr5=x&L?b8Q$?S5hyNHSVdAz0 z!TI3L%C8g6Ag_n?m2SZN;O1W!z3r^F{Yc+xB}#$jm!dPUMcxZ>cf3pVn4h!3_9DOJ zc&Nyb&E@03W&0aGt{kVIubIFO$;P#SF}b(CCuGt3wRxpaiqbW%h3RP)=X)YNBrzm8BxT>Noh`}8qD9l)AAGChJz(E3k7~;Plj@bn z^fV@Sn##vh68W6~Cbp50&NixHkv1BqW9#$z8aJ$$Hz_PC)E#k1UM~v-$lnT2_}A_r zf-4qnk8uaYLGP`Tz)(y>E}n;gVb7A|5E40PYkF|i530||5KeLHVeq~#C^?yDlYrUM z-t$A&9k;&-fgUr&i8gd1KvsFUgRgp!@4nhN$2lfA#_790J;wou**o~Wof>Rb zB+q9jYnMksJ;R4SIjZLBo)7?{#R^4&A9TdV8OBc0KX(=X*G2rbD%t+OtQM! zXJe!3RF>U#b^?d#2CCDeS2SIdX4z?%jn$mFbC>AS;*ji@1vuEH-Ge%}DpMfzW~_PJ zVcpJ-73!z9rU`DfVZs<4+LD=Z=7T2G8~L<%Z0>uq5VrPSrlzf&jknHnA1;66$09po z>LTLH4Ig$2qn38czeiJ7ub9wHi{CbfpNEe6f{`UAB{ATU-ktFnJ_et^xq9X zi3SQ1+*19U&X_iGS@iN?Z%#!ktnCXU;LBoh3^r2#FS6UjzFgD~y$~n`SA;>hlM0S1 z&!rq+zejoCKlMDNUyu#dD4o^6sCx5sjD|^JXl-vScbgAQ^m%DHQ97XYp_&5Xgf_-W zR6}yS!gjz7G2}4<=p@G+m+cB)zr8uv#Dkwc&s>B|V|tMj*f(b=c(D;0_mWR;La#kN zBIS~@p|x`=fJLB}@|B&8u^K@VpQxC_vaot;|i#BiW< z5qjaLc1cEA-u|`pS}03c$dzTM5?7th3!=dMXF3jIXlMPsNk8%TgIYf8CpJUbZOL!i zr6TZT05&sQPH2JaO$&Fr^cxTJ#`(=&r)GB>MPNm$OCUG*v&*_6f098LYUnAs@-3N& z8E$oi#=f&laAAxe``c)>F`V|JY1E-<4YRxB;?(SL2fWn?70$Cl_%SO57W87w>AJJ~ zn|`XQ)W5{ypT^?8sLV0)90{*WKz=ri{db4nFr6-PGXzeFsWvf$hO4kYa09R{JzjnW zxG`u^GoOqq>Wt`~nbujTs*r_gRy04a0Q{9}T^bn42Kb*Lo)1IvFi&$l+6(5xb#?5G z+Jf;6{qz~_7FqS`fANi4%N;vTqierq7I!F3>rLGo*|%wivsA6i`N`lfK-Bh1AiAkn zG#}k}CLYlW8UfZs_paOQ^G|m;-A{w4O8aPOL2kaKRkzuz6T@DwD`VcTj+1}nrdh_K zp+-q2_16AHx+&>NSW;&B80M8`tyntTyIuwSy^F5$f&^Z;;B2fTZn=G5>g)*e@TFE%q%>sGy}>!FNZf<|N_n)u9Y;p6Ko}-TMlUjX30gy>}gR zaODo^eJK)}E|01-K>3Pz&(JP11c3v7WZsm{wZYib&^}@DCH`@zwe2$ zvWmW<@_E*!0=@V%RKhpd$hRV**IWKK1uen8-v|2b6}rwal;crt$VL|7_Ho?(=3=ea zDU|!o+|0D@YmG&E)0SkDVvLj#?B_JcTX2Va=h+6JDb%ZBVTLC@_tn8)22XuBD;)@W ziEbLa>k2w;{&pF!d)cpFsh;*8ha8APp=kyu6o%N3ykaV9mdvkz0Y%OuU6sqfF7b(! zueLhKeO%*?fwPCTygcliyInV>bh5u(`LX3S>S-sqI}`C|vYAbP_j`a1F+P)CJ-^11 z>}7wxUxx>v@~prLU*Nl}<>E{Gp5N?LRoE~fV-k~R`NVRUZ0-)a*@Mq90EF8&nKs}d z?ylOIfkZ{VvTG-2b=jz`IHxBfZ8Bd^ywp9Gx=I)3_LC~Q7Xb1)D00zZw5VsI!~Q3r z>k7mV3gD+(&v4PmL}%duwO&_fvNt0z>F!HE2gWAtQg^Kr9tk`P*WExmG+u|wRsXvs z6!qNZlI-kTo*cQUwgxK@MWZ;m%)o_=jHzKdEt33G5T^pwSPU28<+oh*S5Bil0&kXA zE@yQgzbs?`K*}QJ4C28{YS~w!2J`) zesq&H&#AQAu3C{dDXIb#VkTUM35Vh?wBwSS+1dp%&qISqqf}l@`Lc-F%cfU7IMZSbXRh%R-bdo+Y&1WjzAEA70rf;sfMbblZv9VG)kQOJ zO=TPVsE=Ej?{R!#^jT~OnRO@PM)Ow++K$1GlYJ_DNy!+m|8kq8wRifBXufFZ*mX{z z?I##)YEM27X73-z<^V3N4pFPf(h;mDYPAO!gmgNqwQ4GWmmk>bgmBGQ?}xs`iwuqk zZ%kJCVWtwrw%K@p7@#~#&tBUuLB&e^ivTwU-yl3*F6tWCOq%Rg0e zhC@~E*Rxsqc)*MHkw*!22p0>FaK3kNoMJ7Zj_|Xi&aCsejed-`)Ti^sbG+mKXgUvf zHs806S5Z_^rPOYzO|YKvK`_KX?3YFDkQz4umotCiSl6RWX_#Ekjo`#X;J zPss7)dY=0_@ALee^;#EC@bAx^?`IpQ=_D98xi&d09ZS6P7s%i=iHpx2@pC0y!?ktW zllMhhXA1jn+TuMXIh)waxgXDCA1$|4&QxhAsd!1I9WPMn{)@ElT}}el)n?!FZzCG2 z0qDr5KPNCs3TD_(Q4Mexe+(7zF-LnW|2iuWPv z)3$vvNeidQr&eUA=3-|V{Iw&wS{l!h!Otzl?$O?y=5_nro|}U84;|p?aWJIULAA#3 z(b_9AJS8!a*`9lTx^e2O@ZGBFVImleHby8l*(o86P%*Of>$$3j1c; zWAsEGb;GfvI#3X#4m;B^oDrFd_Ie^NlyiirI?Ubcu)q@Qd1DA;cC8$P)ip z_1ASTqWq?GGb!`lWg=xr9{q|s4m>eFg3oCxGMm@0G?xZhf6E4>tO0%7NcPHI1gBP~ z_FG;44O__&IcjsOjAgP{k-!9kM{C%Pos#)yzx3nc?x2g50{k3H9L2$G!j!e@&+1Kc}?U9q_)z)YQ; zEN|Nf=jS^k{@mQ$iRC(@8@?O|bL0A-Ajc<5P~oK`{YsuJy<{_y6z;0#JaTH$`K$L2 za=!H&=Y2c00`vr@;~X;dSj<*x^ZuY9sE84pkGErS6f#2VSUJE z^kULPx=2QHX!+>+^BYT4)PsUi{>krg5F9P3Dgr{J!%Mq)-^)V2fYP^~ss?ngGX%eu z5wC{(ZBSe|E1$(@yJ#=CZ-bytOS8az3@(hG`{PP$^w;y5dpEz#&-8u+hlm^Qql%co zcyxu$t$CUVa_+F(b$0VA!l+*DaR!SLk=aykEa&psl1UI9VH_Jxiu5<%oW4S8tEDYZa=paXdZqDVa&)R`P{bSY3VqU&6GB}jXr9y= z9kyx7;FJ!zCyda6>|U?zL!B*yJc5CYgx}qYmMuF+e&61d|C(D=EK|~}G$SzbsBzxY zxJpC2$>kF;b{Z{sT{sSSpqx8>A!*=Yp-R=FMQ%)8k@oOw(Ou8z3LlC+)|RaWnO_sXgl*s1Mg=CEKSirsOL5lw<=ST zhleS9aB*=*Me(J*FqD=g{Ny*jI8HA;AF{i>^%-?a8j`R0$0?#6zag)yt2Z^4!DjVb zc=x<@!}$rJV~Xyczh!~MuW2PQ?eTn?InSO-2|x!JRtP5L0=0xHlK%4@$-G$U9GM?$ z8MdURGH$C`76l&<^`S=M6^${iZy%{m$^4aD_UO3u8&rKNPxdq}XNtA6GcvIl-}sDO zHT>ah3ODMfe7)1|R-$N!-N&^ogC@J%%e;Nqb(d&u`_Ygij0Ie+n=;K!(4)k$obvZD zCORcW8`4;E(^b8yKfd!*H|r!ut}Gdm>u6taQksFIXx4?rBfzkU>N!eqQf*_6HkOi3 zeop91NdYqFQ^k)XY*+Ko<5dS8mKf1|qg2vil<$ zF|d@9@kr=l&e5Q0#k{|cQ>Lc4W!7g>R-^%`C2rjE+{S6i(SL>LJt+4Ukr$*;}o9VI-=BdH z?)L#Im|;F7!0)i$8T3G+{ZRNjX{$HfXFE9>&|1Q20lG75V>XFlH+t=9$6^xuXj@j! zHCXCFZ#tGoy!Kv&H%{E*pXnXKbP**?38U()!iF(hL6}yr*~33vk#$ovC9-UHJR0q) z9lLUSEW0o{%QS`g-}82v!fEROFrKjk=H!$CGYM_TBr-srKGF2ayoL zUrz0r+Nze87_l1Qy*AUmHsY+-GL7s7x8Rub8dd1F>&Ej09^4gV!nzZ|A12=Jns`m= zsA0-T2Lz-|@_FfUtsj%?PVN)0G@Elxeam^JhLq7~5dCA~-J6NL@z?O}jb6GPjW5h6 zqES&?I_nSY=QgR`I~aS1qiNXju-<=rW~Lb-bd+K=#!;NvRWfqB^26;9g>{M|@U!xQ z^}wIH2L7b-eHu6EW*1GdTlS_IOH-R9Trz(x2|3=FxbCeNcu~et`UJsk4ji-ESmL6t ztIQj=x!SuNexXWt$a-2^cWEW`BwnlAos0ua1?k=3=*KE+ttN0>KkFz@8IOSbsDLZ$2b z5ixex()Hrb8{4GJ&VD`})X*C5Pc(7W4{a@w@Yy?cLyHkF>dx9oyIT9~z!PB=c8SUz z`IRFbcSmkp_{1vA#?l|FruLb){ zqW@Lg)j0_#OdPF163(@%{wp61&|^y-OEUwm6KUAmnflCB@B2i6h4z1bsQ@)r%<7XZ zUF4EP$dQT_a0UpRYjwFC$~I-T-rZghSd;AdT`6DX7`kkgTN0$`_3uyhgGmj$=g*I#QS=x$JgcOV;DqSefN8PN&~No0~hhCnCAL< zs9!}PUp(l|zUe1ooVqx!54f6La^g%BPS>yMv@^i`YuZZ920`@1b|wX&M`wf~!#cG- zC%smP5MR=jU1`uKh^jWIgbEXUtXA_E>V{6%$z6UX?r?bp2 zdNlOgX#*Q2*bYgGgYrLTL@MMS$Pz^tU~MR_`Tf);9+HB;bSz|LF{?SzP~Zn++n^_gBdC1ju+;uuU=xpyeXyY}KfUlmFq z%QO)^ku?#0=S{Ooa{EBOYsuzpcVrlai(C$>8SqZkyXusxbBoa$f&$|HL?1r>v;S}^ zHQqJZ^?hgOmm{6zN0lY$X(~k0i4&jR-T4uE8Ml#(Y_%9FKV6<-FlWEL^2sWXOEFv1 z3Nb3YT8C~g78Dcqh-_DZzv9;X!>?TX#&{65dlMKqSnr&)E^K22@TgU7tqIXCo;20Ebl!WZz)1vQjzhBuZ$=#^N`KEm!Y-{NI z*lf;m(@z3`sAz}>uPJicJ^OM7qjCQkr6R@A?fZ^#>Sa(si_^M^aay+})Ai5RFHHr; z1~enZ>YcwwL_HNitmW%_goYk^3?I*U#w)E4KaqNGP{7Q`MJtFjj-eDl{n3wMwMVs6 zTgy+L1btVWXDJCl`mNeOZsb^gL+$;3y=Vl-9Db5@(GZ=2Z3SF*L-9d0i$)h)Rac!- z>-yv_DrQsa%dCzV5a+8P?W(K8--!Ew)6H0n=ngI8aj(8C#}BV2^~AX%4+L4NDqDlt zqiV+xcSv7!;ERH02pINK{Yqg`mg(iCN<<+Cw?4$F>WClZBu-hqxG%l0@Rp5M!@GAA zmbXNmT)s8R%OqtES*Sbun#NyPl{5=+9RxkEjZ9B z7p$TW!F&?P8v8V0W!A9Aprh3~vI=ky?&%Mg?Zz8?i$A5a%IIA*)^_Ox?VK35_n4I# z(5~6PJMYYUHi&>Nzc#I`PN~FN4lYq^T#`*E#tSp0^10M500h#vcCipH0-(&4g50D|lr}U2 z>!2+Ijr@5$7{ZC*>>&ZMekY2iqM#xQ+^O*hZZMdgKQG}a&hjR2rklgnJ4cCoDycn7|2=wNUZ z{%vNqYnD_M68G-X#{0)A(WakwE%<5HfkoL6)Vmg6nOb7gRfcxOV@(d?xr{CNR@v{} z-GqX4E-~2J`jq$|k$X;ir8;b7`^RU!>*9pAp4Tr!S``}K7>sJu&$1`Y4R zTUp&Zg=9yRqT6GGg**-`p0uG5j!nwCXiQjhi$ldL(EU7Uk6u636&vN$fR5XX3s^gA zTRRw8AtI2PB%jMF+P%#UK3XAQTOe(~)y|+H#4AZz!jnkEbL4PV4!zq(woLnm^xc3K zM5OC#Q18c6{DaG+Z(n-2Vr(G54d~p{OOLu9F4Oz2Z+hQ;f_De{d6+$3TdlsOh+%)( zPOx*(!`p=!&bQzFsnd}$ImWFEM~8)!6rgk(gz_ zc&StAi+L;Sr92s178mkgYuiLlE~Ihj5VEbHSKb9PI)4`9j^3L*`O3;x^h3a+@(OjL zlxVN=?++0LOtA4K+7}Glh^^_t3m}gbZS`}M5Gqe8$mpde5}kn znqitxcsSR8TqT+8RYyDT*=V<5Eqlj|?fePd$0vZ8&L6`H$LeAQo5RilTz1?|#7%@t zgY@pI^&fjb8@x>O@ObwEibGw-Qx)<=#hSZc*LM6KgF)26`0R~7gr^%`_p7&!JEc@E zg0UBEg#-lN#nJ^1!-LMsUJ+|d-fO{!^WU(@z6t{YqbDx!UlG4uvQYb~|3UNN$45KV z3tw57ka0!x{^*9fL#cQUKZdKDy2CqyR68TbxjhR5ofzIuR1@OLU4yqv>ZA(_R>wH) zM*xtE^;I_yy{(^=31Bt#G0T<0Kdwiw5p9u1OtT%2PQ#B9e~vN9E2iqLoB8Op1L{{D z=tJt#ZC|kCQJu=SNDe65j_LFI4oKW|_)Qof9(8!=`?Pt8x`rKAQk*@1xh#)E2QOEj zY;p>DEo@%{0X zt{h5IbfME>GSO|uRj~FO>*u_|q}rr9o=+%$?~Ot){+eg}1l`j9?DXnsz`}}Ur|7!t zL{i>1saqeb@)agSwpkE`=V_7^n1`py*h3vNbJ3{@xa`+}65Q=(8!ofYh8v@%tCQG0 z$pRg_8TclO`{>|5ev3SrS}zI%C6dAqCnH8w2Do#2XvVjkcfwF1Uf#&5?_m-TV*sCaQf;-?YbQM4UFsxlT(F;|%Op*S_VNuY zFFvONr`VErjOyp~t_C)}G}~%W)0aB*Ce!^To23#N=`mRrCmj&?vsILSJn^q%vo^)i z=R;60KmL!eB-T&1DB>EUR4K71=OOyQse-Zx@PO>Pi(JK&%q(B|H1Rg*_;UjHnL^^N z)bzl+F032#))!+gWNY{-Hl3a!DZg$_3~t4Icvj}S#V7XH>T7`h1?>7sG;$D+@ASXF-#8^58o>9=?jCLSfcPU!a{(s}jv2#lS}BTdE&Y9x<@ zSQ2;SY%EUIrDmwWU@0Y&)CaP#XC4pO4m{XF*PlQ>N*RuyV_X8>uwC|ud~ZEn zYEmOYPwN_c+fsb*{0_eB;WcmZ(V&@9oaRf?z6#<7QDrYCJ{zmdzHv}_n?ODAXQ0fU zjRqV1wkz*J|9}T;NYwyQA2b}co0S#Q_o=U98_$dkjWw>`yO6*+MV8kzYMt^nA`3mR zpH9V7GN`^!8vREq7d8IVRHoa@<+@Q5GS2vbvq`Jjz6(B;^1>OYFIBiv5Bp!Dp@x-n z0NJA5MIE9$6vjH}-@vGS+{a%tD1%Awx=l}E^Iwz5VyCkWW?@vNoeYtx2qf8I&*~1l zqAi~*)IljvN;Be+@K3%Om?YeM-P?tV-g>e0@<_jw;PNYDA&-k9U7Mg7V#C*;0N$tF zyM5_*k;AMJSrn7N$Nc3fb&|%3+l4h;5K#^8@Kf zxPz9@4%4l{!U+u@Cxy*e8F5}{G==xc9~uD|DFK6KqyG*IjPLFi7ZK&&X)1N?ob6~c zE*3B0$r12iwLxXsBx7&`DaFt!UD+6QH=F-`ucC;x!1)F4fT8X5?XY>nfR4Jm`=+3W zXhVM!@R;d~fBdZ58ES7%=F7{Q_-H+vxd(i2I#0$-e*$}bpJ_B7w;r2s;>NFsG%GTA ze#)nd`Tm0K``Z_VuPZJu8=T|R@=I9SpeirW6L%W(UBmJN>|I09gzy+W%b6zNM(7$z z_9ZIyE9fobuK2H)QTmTmc`OeMXe2!5UI8nxcaKi{N^WIxJfWD;-YIH#~zxIT3a({1fl zPJCIM7(N;1YR2WW_B>ca4DCVl!G*-h*sk`kW>Zps1p|@q3x@ZbB9B-CnuJ@;>gBGp?i$jyMbEdj>xquFSC14bI@dwj?zM|Z2v{!d(pFs9&4-3jZBKgoNG%$OX3SSf;XVw>fWk%QP87V2VG3gM+<Tz{?DN&cO8jcTPIa#0Kw1k) z7m5%{MN2PUanJ{XUvwY?cun1R{x#n%wVI-<2U)+ktY<$Dmz&CQ+Z`T>#%X z9Rx!v;wp3e`!tvkIYx%r?o)yhr`@> zanUbX{6W&ljn|AKtDEd&+@=^PTaCfWtDRR@lxaUBPyP`G2cC!oPV-o(Z4|s>9GJ-b zi%(GH{H?tK2(o4dbggZO3^}6;mxz1G8!jrAy6<=5?@C6>sYB)=>cj6B-s9L94XTqt zJQi+isOJRishZA`W_m#?PjSmV4YyZdM<&LN%h#G#@$5_{kh|<`Ct2olWsKh&3emzU zdE{*FYDL0(mIA;`THPf>S5jox2j81y7Aw|&cW0*ep4twJ4|m%?t*btjxWQDoZ&jdQd_ef~6PLjL=6?hALVHZ!3tnDnyo~*SMOtfJ(a^z-7Rv8kLM*M} z=Z#GyHdOn`!s6tFqibJvJ-X=ZOUv7@tKiL)G+~5Bj0R`p<3&ZEtqqsv})tcCyA0$A~rjOvUU;I4R$j6BoqciPPgWTkKS%w=%u- z4sKU`fRey^yw=$LUwnYLhZhGG=1OD1;fpQ8af~(LSoY6E;f?+hX%nDB%7HE`2zn)vC8<@ed^BR@FJO&bMkAw@?XAYmmu&b zq5Z|&MV=YAq`J+`tt2A$8^!ur&CWHluG$C3IvO3_(w=ihDW}nY-68BGXDT@Ry%e74 zOr(FYnlNDkWdIwE<`4^)-S}zbDe)cV5dj%nCefqRZ10}brT(U+qfY#}f|cvcI@hKZ zFK*pJPE)m{vAf-I*euz7Xq~~zY->AQG|=?cVk2h3YE~+qNAK>kp~#_eg#%L<*L)*G z9kEr*7?`otkZZQgs6D>@yK+v!xBC^=vh}Y_ahnfY;e$f>jQg2zN5839=4}|d#$~#c zNjGjDIbb>4A=$pvYS!))=bNf!8Iv#aklrmwT>R3Su5ZvYOB5i zr6PM@)9ZDuP}aB=+709&X3#pb)#=Qu2}8J*GSPn>_au&#*6K^qPZ)Ma`A&sa+dOz1 zROvmpz)e-{C+=_|{EEt*|6HX@>STje`h>2Sh-jk8@ndCV>HWm|+BzLQ5G1ZTPrSEHif_1>n=-8@im%{3EMTO=A(-Fc9+8-aN+HeR=4RS1cFYlZW|1& zS6X6-B<)|lVwVL{6jEKGayN5>RzwZCZzvfcFeCWXC0jp6Vihyi7%k6OoDK0cl4pCs z6mU9uXrU?ii&Y>ufJ0mrr9KZ13u7_oNg`I{@W<+HZC52FBt5Ib2FmU(re{rWJ5^YA zU6SpZU-|6SXQ?S%N3(ul1r0*ka~3NULqE~plkBLhle z@TmdP6VY-f#)Hxxx9WRUS?&GCca8#AT zKHQ}jtAz<3$%+E^|IY$wMba5w>{IYz4+?X?$MNRa`zL9Ju#=~ePq zSvL_8^D(*ypGQG?7}2AyWxZ7i1PnLhnpb1&b^V2AaC5* zRaQ7*%_<6(v&PJ&>-jWLS4fz2{RW85E%%QD8; zxXIlX^%@`Gy$_a~B=l;a32~S)_YVJaZ`CMBEY~m=9$=#r#be(*=%%tbr6yU$d#8`q zkmxbM+^A#F>Dl6=`D9NYaTXf%e$ns@t0iZi9$VC<1??1NH&Vm>AnuqRmS(&NKvqwL zX|ElVxg_YR@Zn(D*BO1@kB&Cm_izCR5tEC|-y2b^(e^&<$pj<^OBLnowBZWW+y9tE zcCp!?z-50MqO(w@o0sr?#G>JG7FRcu*StAm1qr#>YYy{Fd}I`F>_#EF9!9Z zj(|A=8SJ_3Aou3DhW>9gub|^&rF)P8pJ2%+d+pCH!zW5KGd=@rY>F01Hb13cY5}>d z@JZ?b!Bt9DbB<#NckLz~$3MfbS7eoUG<%C_e=z;p1vLY6J0O14U+D-L(RdhMGDvT6 zWk>!X=&8ETF=_lNi~&=})!{Us3#W6g5}g0J(D^tff(hbXe;j6s_X&Bdj49Xz|Cd0h zTDm==KOxbPt%@a;V4uJQB^6MX<3P1jMEL?SFuL1(e>J$=qo zSV*QndiX*GQzq@{qv-O_Yrl9u`{McvU~GS}C~`{XPH^*rWI*necy`%o7f$&g6`0Q4 zl`tfS;RkeFtTDW_WbN`pL?8X+2g`R2$$NjV@rv@YVz!H##y{oXb?+yU zCe_$STe^l*@Li(zH-5aqN)njh2|Uz+mt4dGdu6=pC*g7lll`h5UlrGd`XhkH4Y~1f z7$>pFCA2A3`t**fh!}G;;$1iVHE5-^;(Mki%!bhPMnz;j>^<23{%0}b zeYtmgHYLtCM~8k_@ltDVMTn&bB6dHH2#7Sm=e|xa-t?u7&?3c@XQ?b5u51omryp6? zzwU}ZqqSSV=#tFwe=2GA!1}41P_y$R;|rtStFeulMhlzXkMsHr?p5@;!8t)W3KQhj zB7D2de=xIFT<~zDRF8(YH)|Gn{Y}DJ`f5ir3ss{a+XuU<*7k9hV8$Gqeo!S+< z)Rcdm;GIQIL~hoThwjxSx)I5L=RCg1uGn7q7k@WxFId*$^Gh?PX4dCfV2sCTd3dXe zwnokWMM_c_Hklr;d_?Rox&_hsGjy-wD^kq=NNoJk)JXUf7}Ah-p(@_Gp7+KxZu1r= z^_pam#(bBZjXFyepbF%yROL6;ZQx9*cCq{T{OT~^c}JV{nXus|Z94QP|1}BPuVF)y z-aCAj?J%ZgB5FdwbCYhaStXM->+Lz#J6JB+D@Ksg-=aWg`Qu~HRKA3DKZ)dFE4*+W z1%Mb|zIJelh_2}t-P?!*Tn`*NnnH!37LOIEWK*%Bwxzs5PrX!D&-Cl0O0X@=%)Yoc z`Mt(``!XsW#_eok?ED2@&dG(c1rm~M^{$4(JzebytGAh;qZ~4G0(*v`ZkLDGg+ z+o{o6E9AX(QxAW*oAUP7HmbJf&S5eIZy$9=fPKiM&m}dei(?RLiM3)*ggct8tv!09Vm2lHmn$#U4%!^aAZqf*1D~k zyBv43H&=hKS%Lp&tBb9|3tRmakUzOQc6 z(go|1T@@$>^^r6-uyo&-M3>J<6S$l`KhCCweZ1u`i@HiB8t ze^+~6h5xeaDCpqf1K&A0wP2>QlW9CH*6S(>zoSEpuH@m`;LdoqoqiZz7-BND^R&=w ztDLBy79O!)8SHgrf2Fn$?n_@)D$H_Ccg2a1Nlxgs6GYG%5)`*b#9?B``b}R%JtxK@8?JN^r zfYFo_?N^rES4Jf>ipQjk$F#8%(bNiiR;i%bE!_%N89j_RE$q##>(o7PH`=lTXx}Go zb&xXspI>pJI#@_3h%ExJy)r-ScWG#GY*X;(MSn{k<#*@6>!tJ&JN1C`1)HW#D-N@j zH-Ju2*TT@IEwC|==dO3Hz9$zJEl?GX10iTv85h+suHz54|IF-0Zb=D2FIz1B<)XN4 z=gIKs;z9eka1<($>O*Z<8j zq4~m<$VYd%Lbhk21<)8;!_{)o9LOCXaOkJ9@Wo4~QFnjHch*9!EEb*od@xyurpbS} z&TaG}o_o&3-$a7!zX`FheA%Ij_Y2eofFw=*Ogk~tr;PL3*&v;etm4P>dfevR)0M~1CVOf$e zLdURz_m#hYH}ziF_yY>?Yw2DTwrRw=6+lj`MPE{l~fG`j01G#0Lp0_@x> z2>TiEE9%WArzhqM7^HU ziajY66C%B@9I<*3#Zk;bm^!1X1B>w0d*!(sW)0dDxL+zY+l?k#!5 z{Y+OIIx~bNyBe( z`h6_=tdonj_PxjQ#x|Nj^&hqiCBSFIl5v_)x)z%mgPS^UjeTc zT`plwwMK?`uBvG0{jF14mkw<`l+)#I8e*#2 zzf3i~-N5I_J!=WN)NcubI}-e+g<5#?aF`;?RTnq)x~_EUzX&Bv%qAU?`A&q=y3aZ7 zVS}Dyb<@IZui2OYwgSTw?D`pdOpUZfraKx}_zLe9yP`GboJ%0NOj=6v>s&}|HF>e4 zkY3;<`zeXC2h@zJJc;}&@xzekTtM!KD2l z0}l~nqbjj|BhNfnw*{~C1;m6LspC>qt9M)xSl4W^@7#x&Y_I1spF{7+#H#zhkld_k z=DKpJQ~S%mdo%LkXS8p!aD9qxiT+7C-CR3$JredLStB14eT2lyQv+ex$FfIABBa}e zrExUXerwWS%lB5G;}qgNh;fA)m{00ZZ|BzT01VBH>GC&no&)eupN;%dw|Yyj{WxeH zm?PEopTsJ}*3V2>9LjjVUR#G<;)ttazSIsf(_{eJHa6TABD1-#f!oimU;&J}Y&yjA zL=;>MRjo4!&$kW>l}$RnQP6Wf`l>UJYk)rGBc?x17BU6uG30JP zHXk2bDDFCxsD|imWh3LHH9QnCSt4M~O`EvB$a@<$Gh}r2;mol38cP zR?57C))aF?P?gl#`ghKy5|77kAsB?~y{5J5QakYan~g}jqI=sHy6&17@-$&{kGfa< zs?5f0t_M#erm@Iu)7N>GzbB(-0%1 z1n^|^8kcYwUyDd-JKyJ?fR#@My?TE9`X$mPGfW!l=-3Bd0@+N7VxB7v-L! zjDH!zXp3$VPXzu%jFlkJpv|ah3pYA??`C%icG=fvZE=o_CMwxcj)#cnN zh8`p1cpjkka^6PLG3 zo!aM^;_mC>Uc8S$z1vI)Aw>%K==b;uq5`44=98f0`6YQITE z%wj}Rt&&sJt2aHGT9tNG-$Kd#wPk)7NYV>WB$>&|CRvKwbtU~;C^>brt8uFPGf$R z!G4?l%v$YvGh4k|^IPrS>}k)#2sFQFHEVTm#8cK(ScbEJMa$L}%d71!qSs7~bPUo) zdTpcKKHUKBN#gC({z@lA7Nl+ltWl*Anp!)&W9zj?ne!Wa-GAU~7v(2kh?B)ouijat z1$rhYlX+XGwJe1~w>)b-Fh8Id&do*3H;MJone9xSBIiy6pD+DqYsF0dWWZ4Oy;qlU z1@@hPo~7Z};rmFa!BKJGiY8V0bIi@Q^eR=&=x`lw1NIpAdH?)`kyq%=MoRT7@o$^9 z4=+7jq0UrXlzd~f`AHV2Juex=%OIjueoiv{y_~|$u1~Uo<`@?%F1$`tkne}I7R;Zz z+kc}lr1gwj4tB_)QkDopIL zi7KA^%L#MlFoh^{GGcAJVQFQ(uD4XKW0DukjeAWqy7gsd0_C70&vgX#Y;uQz74$ES z4yLtW+2{JE&cMpdXs#hSedAk>5{f4iSQ^@#U+OqVoqCFE^+Edkd-z)qqZ1-e-73xr zqR=FtMK1FtSeZ1Q!^n;t(^aY=QkwJLm@IejW_uP=dv?lc)walIt)B6TY7Td0d%U9~ z0}X?HSu9SW?p(tltjJk<^nR{*l!Ki4y!Tr}=7n}n7@qeye5 zmG4f8dm&IRUR1L4`9bBrfuc6_Mm)E<_Ay4g@uvZsT9#o7Z)ZNO0WZ(D;VRE(g2h~{ zYEgt^yJS6Nak~`D?I%n`Z-&jxyJ<&1lV-6+v6=4)dEzt$mkBm;RRL7TIMfa5>vijY z;>6%{pY1I?u4~w2S36Vxz*jY#;VP;7(ou+Z(XryM0n>P_9O`YRz3^IJL>V=~tAI`SRPCZfqL z@x>e!_nGjO54P2UlHGYU#I_S{eYrkyXzBdi(f2;j&nEZ(8_}<&k0ut5+uGT=fmBSe zqmOlUE%hz@a`7^$)F&(v-_zmuclXuMI%#gE3U~K*WK=XCmyuqJWgBODYGs*fLxEuu z+<(*=TB>KFsb0xHS)Uys?=xz8qp)QD4bWT6O=0lo^ExdHB@|Rwm-*};E3h|CV!PHa%wcOmGB<1CAlj68`d0A8^>>WEHvm^uyF;I-xB2W8CEg8Et7V9*m~2# zN8z|Pd$=uP_{HoGYoI*gXS29Z{X^MiAua^yqx27Ll_A)5fQwlvsjPyWGp$#NL%W?F zK)p$Rj$FODK(uGy6*y>D;*%|Vm7{Mmn>Y!^vP_;kCc;**UR2h*Bs!|3%|Nh<*pdBU z3SF7SOj)Ai{sW4fC-rV3)lP|V;pHbAg!VrP`)LcAf~(41!nL!KG<-1Rgq$RAKU^}o zbSExWPT+N7yVM$QQUE3&i~c$<>GrCkA|kc;pP40@0(ZkpJ<^nrerP7*sz$w{2KJ1T zsw4m6f3Kfrc}#s!Yq7?s#+|d7>hik|7dNjBExqkCTGfU4pU1=Erfrhs!0^uV=t|g~ zks7bj9K>x9{oD!;0;eVhg6%Q0Buy!Ndi}klXZjDhhblf5IzKdbdlziW4YW6mZw$?c znDb+Evl{a2Wjx2!Y_fJQr1b)Ac>h_uL z;4AwkQw|AW%Pujt>Vk!YvRE^#S0MoBR_vkgNlBpogwMmh2I^9`UJ*) zceYj3U>W-%l9&TU=*{zm!;r{_%AU|ZDS}7HU@E7p!joq#hx|AwpFr{vjlC@ow{glT zoawwKAbo0L9nxry!cjU7MY^PxH7~C1Ran zr?BS+l@{doEW_`Wpa{0hJBniH;)XG{>ipp@HAa%w^>FiVbufe^W|g)(|CrRD`#X5G zeFfH z(+}F4w@~SADM`p>7bDEp*$GUq(_1w$ITI8gvF4nR%&`7ms$SFvC#m=O{b2pDMnZDF zOSsUI< zcTLVE+Go{j4&|>wKX_wYslqIdV;p@qXg`{QQ&K(6Q?&}FpN1; zzP}i7r4GKkE-A%1#OvjocCu4k-4o2?0e#wj?pf>AKebqmG(pVXtwcab{&!mL)5OI# z>72*$C2%sp6m=UfKctCcDQzi|5XDD?YMUE@K94Y{oQn3*q*4SQV}qs7ekho8dLQxQ z&@J*8bmXcvUhVGOadgyMLfdQWJ``q^r_djr6`<#9D z+Iz2ct@i~h;QKrL-kVh?^Bm-8Q5R{yM+~IO^hfs9a;!f;-DMC3mmhlzxD?vahcaj( z)PBBexiC{Q_6Mj7KJTX_PC2U|zT)pp@_6=0o|A(7UXJP*WM~})tm~()y2$9YTvy(& zXu>agq+6_$Q;HHNhUU7FIloy`85ALp&qDyBVFnTc8C<5ufb(=)m;F2wR~!LUVkXu0 z4Ngu1oNXHLu#d&U<8A9@k5LonneV=%<-P1!cuH){%^%#_n9DE9qe)HKniAFP^gEwp z%lfy1vOZCK4wbkQx*BQ!v(Gog`yD9@{2~_?^Wu5-wv)inWi8hy{izpKlN+PZq4|0fR1L zLG}i2$)nT%&0LKa#5ah@`9qN;#tOq?CucaknEvvWn$UV=9~NTxZMj@1HkvzkPR$RA zmE7@qz{b6U;b2vO@ZDL)?DK3@Nc?AXBheHu8?OgCQTr9r+G7P5!OV()e;Qz_-w2ZM zs64McBgC~i*P4n{UB6>mt2PYs_E_9J%69lGpsG9V2uWl!@maFG-`#P#UFlpgP7H(3 zSM8-v>O9%--s*UujxF-DaBAyC7dTty#WX<^^IGzH;e?aHmvFV8b{9wgHI z&fWerQF|Va{bAMV&;98h@`VBkO(G~a7`f%V&ROeA@V!f&c|X3Ctd6diN&NBOGYQr! zahnZGIHyrp*L4wh%28EYCt_)>E%{EOoPJM^_D`%*n2Ps{w2uJgqwJ?f%6stREeChi zRj?EN-MFiPQFVJcSKyEFDWws0{OmZ5m+X9o2fJ@TFS>3Eo^S(hR=JY6{WtL)Tv9wv zzsu7vame(>ws@mbsnGend?b`uzFntlc+qrSEc}!&A4&+lX^o(Nd$^F4Xv@ z>is%u9daA_v&<>s0&ll{yp*(D_MOPM(y=#i*QUj!VNy3nh55FQy$ZzbjUdI$2&>8> zSJpt%Q$S1Ma%Em4p)QK7HZg#pG3HP@Oq$|Z@=IUb@6F~Uf3x0|adRm=e)JAR(lQ*H zpvX!afBz8}o1wk6GhBz5(WRqIy{lzH_8vYNXzck@q@w*mIxNk4j-c;F{reISzns*Y zNywL2SEYuD`^y!(5u498odHP9TWCc&3!uE3&CuZZmwMCVZ0g09XGjCZ>+p)UavN1$ z`_;L<>ph&a12-r;evsu(PUp9$+@O7A2WGKT^g=}%7yZ$dqZqM9+Qdi86tQfuYBTDV zXLD{-z@?ajLydff{_g`BwQV9J0+(~lKovB%C4ch`-q%sIHxJ9FR5=7Q3Ur@ zZy6o=6jyOOd|<@MDov)$(oUuq%P_mGxBRls0FZihq(+wdTkKQF z#LM62(QJnC#QZNP#lT>v<`Na7XMgij8I&Rwma!IETQwsxfoGhjfxJr;oAAnfF!IJV zwc`(9FowFEaUZwrWjm;$?qQ=LHn1hF`zxW; zY;UjqFsoRz#jwu`S4N|Cv^MI+c}BZ3y5v{Go^=;Yi=C`5YMjs2>M1ejn*)Bm1%H;k z-5z_pE@#3Te`-Ei51x#4JzXyBbd_Nqd!_*R-Va|$d{0SoI=>O~8BAP`ptX{nchlsB zkN9lO48(3S=m0Nda#H5Jt}J4cR)F7Q{>HQV>#jWDkI*}3G8X)~AlMD2D3jg>Z?aW2 z9yF>4mSmlX&YaPCOd|)rqC;3dy|St^WO`3;M*Fgvw@!~nIz>Kh=oZfwe#-IpAFXF+ z{rPb{gYiZu=tjq(e-uv}uzngWtI!swC|vdhpJGKjThQ-|4=fdy8Htd}jZ*0)djAv!};W3aK5tMw%Ha!lNlv&}&h(*|$e?*za;QZH)u z>>O1w60@+PakbXaUZlT_M5Pkj;RV+x`b%9pNSc zJGNP-ndW@|y&(9!Cj5Kk)6piYcmRM7UzqQ+WR{UQJ)hf$pPJm@@{p+9-ZC6F^J(@G z=0DEOXB4v|E3N$xrEb~U#`iO6LcZ<3`Oe6ZWugTVCZVT0# zS9yQAe8uv#;dvrR+}I5;x==D&%{aozOe)(r+NW`V)|Ox3C4?~kMQeK>Ni!2QJ8^&x zf&A>HpuISrGS!x{=2>u|t(l>71U?+x{mUJ0)nWB$^~zr)RD$A?hvYaojbz6aa2fCz z2w_xhzQ zYApAaGNq&SL%?PO*ZLdAEV$Ymc!F1ZfgZg9BYmO?y}~JaTmtPV|H`+(^3~SF+sAHf ziK%OG?;7m0n~lM1EQ0X+sVY=t?rcOywr6~wsBbdyloA)Of?jNyx{21`qVD_6PrN|ekEd;_VPN4u=D7KgSp^mUhdgvw6hR-H- zkn5uxMubxjF7aRH{nS;Vr=C)$N3)MNe$L{`&Nl>GYxn)Gk7nGmiO&N#|8GXnGCwI&LGihfjEZwR`__>$~p z14|7g5kgiQM&HF#JtvM%s*;l>i4@&H`a|O0#TXk;zCF;-(Q_}e3 zxMN=C$!KqsQop=#&Xvbcq$?HK$(p#0OT3N-e$&-55HF``S!?+N#p9C^qIM`Bp(tb6 zh1w>()9Z>Q0Ph z2%)%ohTIx<2%WrpS){og`c-Lc<=qz%xQ}ybHw8dgWOV@Lqk!|2G`2wN7tK0iWOs43 zU+8UK(R?J*~{(~K`sFk}t%L}!#Ik{o-zw_=h%4)QDt7debkr$@i zaBd!(?{-K#9Qx9@ifih#jSo@~$jGnX(i1f&$1O>bkmJ0{da%NfR=$D5G9sa5?)5#F zmtb%J-CD*bGrn4xOGTSRzY#iP`T>vE^<)TRTPOHIswSML7u6H{0*n^HNcu7( zWUmd3z0E}Tt^G}%E@QoSna8*2$3q5Ag!nT|TL(MCYL#AZ_BErvCIOfT{aYds9@f0p z275u9fOHWl77pj`iBlAnA##Lb?p3m9B`Z>%;i%%Mt-zS15SKVP^7=;3j1SJCuEEX2 z>;d4j>yYr_Gykx*T9lF3%hJ3SO87S=^ajg~FYKOvM+Hmdp8lt1j8Lhi9(6QAtvs6N zJ}m9K87fHeFsmuA_&v=2C{6F2irs|o{DSCJoXqyB!XGDL_8|(LsCmD%RN?lXtFo1y?t5*TM{MGlC_B8;N zvp${%5k-xi*D-VF7#fIlL~C$stUKy{osm4^{F$rS^T4^OeCFIDZe%ve#dgdW<`S{P z@?5zL6xvAS|DOn*od%78{(!2=WDp;JQI+YqROY7aadu#B+USw%8bCOoCKGk@{rctt zLbO$6xVjZSp#1ZGNB5aJSA|+_{}SvdVJ+9WDF7emZs&Ni)(lqJp1@Bd+7hY?yeZ50 z{@?xI9nBsupS=1Xl}>B$o#Tm=c)6|qYymi~pE~9CVAy}uEw(;egRqS2DM_x(o#}of zQTVAi5+Jo<#))THKX-~Xz-FY^G8$WSgsY&XZ|L23b4ntFSQ0vmneAsjaEN|5wvudn z!!rCKfp>;$i>jWOIa3uA!tPKaKHGR7>kqMz!{gZ_5uoc zKO{iXkK&8l5z3s?k~Ga8>;cQgfqB?C_w2i^$wYNr%}#Pqw2XxnbsCXY`Hb5^ z|1;k1XPaUU9Ti` zsPp88*PYzQLYaSB8?Vk;nv#zhdFwnH2vmQbXwfm|a4!1pcBX51`rGD2+mesbNceTt zuHG}oGh){h&f5~FB8$DzQ~eJbtolSOsqTAd?(hxRcOY>n_j5DFYn{~Cx3r+4s6*n3 zLt5`+*HzY;y+3@-**M2tUQ2fX-BT3e!qRqeg=jpYay_4Zk!*#&a%@mA9)=vpec%7@*IQNoWfu)5?fhxST zyE(a$tb$CEtL}XtFR9H;p@D$dT`J~I92|etl(<-*h&=9M?sQi?Ki{qaM@A1v?$$8-fs=fQ7bH}lr3ym7;=CHrOR^R{ zNcd7~)N-I!W!#~C%CP$A`LOmN-*sSoX5>}m1Sup{RngxH){6W3GAZ;oLH^j4p_Jwr znFgj;klGv|LH!R(#Gm`eKNFE!4?wHH+oj&RNw@4PJF5W1l^lchO88)zeeJx(-if$H z>>XXB<0BUAJ39J%z157j1ET6%TJ%VdgB*s=!VP&Mz2@zr{d$|(9st+2R*9gv$}-w& z-OM*}=!PzXR<<3ZSoSw>lBaXC{|tsnFs6Q5Rl?}G-UA3{Fs{P^@b5C$pa8UhUSgAR zcJnE;@@?kWewiDQ`M)~(Qu$I(S}3cz!H>xaED7p;%F_NXM@(38Go=%rz@8=#ANnA8t+Ry!M4Eiy*gib;mDAU%b#o9DBeKQRNBlF zFoG3knKYax!D_72J!**NnI7S=+sKP@o7j+XFHsyJ64u0Db&laS;(Z>lO&WqaGQz49 zTNNL2#kD3*Rv$%Rjvs9<+>H|s<>0?yejzN(=IhzIvC?tSMK<}8&}yQ#fB?N`;=_c` z4(%{@07+DDh7Yo(=;ODKzpg__z>3K-QLeqwx%-QRVKw;ywCoc788`FGx3h$=1?wH` zXpZ+h?@h}5MOS6dj}1xMKsPI%k~tn;KRqs8jJAka(72{3H82Xp1SC>6D_+G8`F|ZZ zgPt_4(^D8zC8FqSs)x8YRNN+)*xkRs_?-zgBwf(`WUL!7k4*ux@Oj@J@~&?fRBaN% zY1n*zjo=kr^LE`Av*H2+fH)iHSCMI1T$!BSy@Wy-*}3hf>Xe-(Or{0g?XleT8%m<` z^4);Uoxze=j=$XJAhoCUBQHo-{Fq*%_9^Z2FdE0eQrL5lt>4e_b+2yUn?t)Tf7JJk ztyVPyaw4ZWQE$C7yL-u_ph3tCI6+SH9|Dhb<=|$B#&WAR{Xy_G=OoGB}7?!Tf0lNhA>_pI^27USDD3^VBsi%ZiE{pgW z0v$?2A*UDK_u&eiWKoMkxu$>7Hn5b^D6SIv`sMwvt4*JnXrdxI*rx!+$!^f1xUs+K z-KI3HgmG(65aKOo;|p==ptBav^tPtIPt|r^k3L zDb2G&;6(B6PJBw`c^2)@WtC@uMf!=70YN^&)2J+(cCxvawv|LkzQo<@De)ibDBXqZ z>gLh2uacI0ZSTy??h|taxV|i_mU51>`%fUiZ5Y=aL<}z8atFK=#M0c^*suNm6aa`- zW3ms7Nil!D#I#@l@1q_NSc3&bQR%?3s%R1~I&ppMH^5DDIpziDF|e7q2%XNt?V}{5 zgDM7%tXPfeOEv0Snjo5-XwRBa*oZ&V$iOTA{OJOD1FkguNmQo}-rgsVRicMGO_8JF zDuOFNKn~h*HnvdgbHjMz$x_3h@7;mJ!b74Ko&|oL?v|DaML-K+;_6EsS=T0*$avxu zqfoxa2IXch5;EU#ses-67rm?;+L~|_spN4Ubwl_*6u%&N7Z4}@L8AU4W&G0*PSaz4 z7Yh_WLTB-S0}e^GnaNiY2ZO}EKS@)b8m_0^S&k#d)lBCk>hg7Tx1Ro5TPlS6{0?8z z!^n~dXnouy4yvBBn}@jAIaY{zG}8AD=$48Z*6npU;rnY+>?LL;AcH=x)0(htaD+}c zrnK~wMluS%^|7G^FS$&pOw%*~vbZ^v&-o!-PRvsQy69rUJ(5O8N+aY@?_;;H{CwUY zEAJ|9Z@(v~{kI$v9ZXAM@2 zw^pt+Y65#%%Ir6D2;5H{1KMtD=;RY114~yui8<|86Jn-KOglV=nVds;JN1~(ll8>a zN#qj2+}_Se&OI9bH86MEp9uFHGq&>BtWd(~RM39K#L0e@0tS5u^IeQ4s_2!G^KSkf zxXZ>_RS!@aIhGVWt^ca+;Bd()WWcZ|Fiz<3hDW|HWe~PuPVWuVGX-21#nU=%hZ6+yV=R04%)Y$8B`DVecqV=ZEvodr(^|)b<#<BTBZ#4af_1>yQGtXLQH4gmZc9ktlN+QOS$*h~kN84FP zL(#SZ9jj;Qio?MI+koB4!Nwu4wkNYI$BaGD+iQCNnE1Qf#S_H`j00(&%Y1j9;4GTvFiweq{tTaSt-4^=-~9rGcxo1e~sn!!HrjZwzgb-BBe zOI%wiax~Q4#?fR#WkX~`id9kFWofJX7ghZ~*zNF zIF{jb^@#bJM*BQsvAu60j}|nzK)p|!jSa3(Tmx5{CHh

    #){T#i@r&j(ni-e~Dm( z$TgYfEL4E3_=Fs5gD$xd^qqwz@nF`TaP4JH&o+8tZ5!^hw{=?>EMjVXGBTgB~ zI$^e-RQW0BuI;(DBhb_KZd1pgNpnH>VePS^`LkG=u;&nw(~?D-gq{flhml3fjy!_! z1O6yi`wTV_DSjnHE=*?707yZ>BV`{nXmXBDMfo4C*ji}yxuVG(C%&QAJqgyNq~vP$ zdhN8J-+l6FmVG@<3RlI~lk}>+6xxbv_yh4$#_I?Bj4pDpW~Qj6?F)Mh2Bv{Lnf=$W z8Z)bM;RgTP!F;|pC%Aj)B$bZ=75WVd=U+wXzJZW;hJ(4jCUYdkhM5ctWmVo|04VN^ zt~)shF*q4WvF1*Hu~6JGs9IaeT-A6h{+L7!3Kw_j3rcam@@0qb(`D3|WP_SD|=IUnK50LMH zrJrlBwn0!KvM+FYImAxocPnPhM0B#PaBtuyTQ_ViI(du7HkT^QU;XzG~Tf#!z! zxRqOT`4U=!$zm+mz)a(r--IzdZE;ccJb^lXXz;mNx8z&@4*#A;ESC!rF zlw0ABLDFXQk+!u-nRrn?@Pk*$TEOwHVv_y-R9;SP-cay0vkT_E3^HT|15Azs`)&HC zwP=z_e$yRzR&F(4fHd{x zF{AqAGqcf)5rY*i>NVw;|AvzyOO zTf!=82Mbq`%J_en%B3^Gmi3dsnR@yxd)bHi`K5q-*XNg(gVUt9m0c)VHFh~C*Mmbb zq?P!cwFrtOrR&%CJVbG%CPG}Dmt|F*)W$tf(hKvBI6nOPvc@7W%e@2Ti{|W+@Bh#A zc8{^Q<0}kAo9o(XLpBs7}S=E*-LHDgUS?)~@`(i_qtk>#aNNA70vm>%W z&HaKMxAhL`UREp||Ca>-&m_+R(yt9L@zA17gtFsT9Os*A5;BK|m|T42qvU18pcSJq z)C-Ru>>gK8(45dQ_DR{(iKJ^(?pRev;DoMqoOc z=CF^1(qV;5(csyp)u0uk<8bGEsV$3Ar$S2O*Q&kQsWl^8-zojw`*q)kYa3f@oAWMQ zIKFuX>Ye$)*P)w;G45SX>LF<&WF#=v{UU;gUmBdZ;@vh{Sui3<=kRxRy>^ZCta?}6 zx6SQrB7S|{rn7mM;^=&&%Hd{twxVG+T+YLKs22kW4rO<8+r*!c$#<}x=;$Us#{~8) z8&Mb@z);&-1CFyJ@m0ngCcn6*tv&QDsB% zFNGh0jJYrYP*1&lW}|@D_4p``#1S3P+j#@uBHl%sS$d50?`JWn$Vk8Wo_>JERx)~c z-Nm<`Put>{W)h(~eaz>{-l+87%D*RM?k73YJ1;$|c2Dm`kaN>{%gU> zrpr1@$;Lg;6v&u3cDWt?y3IZ!pnEcDc`L8h!+?h0Kinv2)x)&uZH|hP;*S??;LGVRrPWY^T0H_2?sd%@&U-sV(pEWOFT`o&|+-LCm&j@ z7=jcqVi{l&cjE)_r|?~(AO8Z3y}^M;otM+Q7505u#QJA6r`6Ck^zxMDyjh89m@NoC zx4uMummOtUMO$Pb88h+*Cwjme5K6l{7ve{B`jrm|p)i$U= zSX{z6=ZJZI71E@Nf3z<#&D)hWbGg;H3@{Rdk3xoZjBI8RhFbQB7XO#~Z)jrZxBJQ3 z3@=jP0f2y|*R7)oTJkqjO?b)XV|a#ThJuxTq%4vNZdr`aZ@M#W628A#GQSoQGch;b zbgb?_7oGG)7G{IUktYjcM3f}P2W{Bi{LjH0D_0G z2Xc~FJ^BMc=FDiH8B;UGc)mUlvj*7}1q$v~i*;GHv~`;I(>+Fsy)C1#dL*O9XPs>M z&q2sLov`bfL`?HkGF{*(!nu((wn8|XlH7Cdyn8g1{X_U_9LFD1+4_yI)x0eCA9%=L zg|j&iM1M=55#l-)G^{a8;fA_ktr8hPop9j@-ZUC}9ry)yZ%3kLfqa&+Ndf%`y~^@e zWz?xB>`rrR)WL)8s2*LM>|-&3(%!}2%u~(QvBEw*$(n60B6G<84O^wr)Y1cLx%z3q zJGj8WKM6}^>5!FDEIhR3g}MU#BKUk1^z@c?OyV{q%Kn?Ab$XC7DXhom2LGEru)w{{rYf2tUWu3cNdREUrAP+vomiuTM86BmxN8?{MeAKhHe7A6zi5%l9;LSX^CiG#2t-3Yr*k z_}vwJc}A*sw7GjCrr&fSq^}&5TOsT^bOZh4toD|O&b)uv92K`yb}xTn@nA5P`xHIk z?ZcohB|76%0ex${R7>WO!trjdw8+Ah9YVw`w&Hd#xHpP8A}zNC72V-mmjxZKtE@9n zaruM_2R8EWyf#EmZed?D0sK3XGEs*9RSv9%mOKZYy)1?>Dg-BKdsvocn!{m$hsYaLS~C zj~$*74iwuO++bZP>^^8hO3Cat#wyYtvAQLX!J$|C7j9IGtdU{MEq2?SIHh+j@1yzgD+L=S zv?Uy{L3O|bQ-JsU5JBS2&*Ncn@9I&bt9s<_XW8C#xAzje)=q)@ma;}1Q)|(q?5D=H z$_H;|Gv!z4ABdi9$9JxO*G`tGA?r+g8ER=prs%AzmJH$-&r2%Ld4Ezl4I(D#r(2jm z#0ev*e7)qWaZjP^I8e#xu?F{_-rs65E(GxNb>M2hM!D#qR`pJe3+vwE?ZI?Mv8@Y6 zl3)*;y>oP?#V^Q=Pz@yJAy@tMXVZRr{X<*VgV?%}AK!3^gIn|OCmAD@F&O`c>VXPi zN-VES2xP{iCqgmg7}$f2G5k3*v6nS4Eo;F@E2=%cYSAN!QIkeq|6RjrHzfM!ncu zX$hgxAfD8gP6tP^V`rFkd|m{@1~FRfC3&Fu#=vrgUetu=OpxIb!QU|e9XEsBDmRV! z(Qcq_A5UyUj{23M`eA@xmKJA*IL{Pl^kjpk-1 ztKd*4n!+YQQa$oj4W1ty$CoC!g3) zA+tivSw6mGJDpYLN;hJ+*tN1dt0%)3S&~?n6=_z4o_J5TNJ|;68Sgdkf7E<5 z4zYve)1JJ^vO~+Qch(H3@b39{y%j-8HF{0NfUjV+%C^frL&nBRmB;0untPy;3$`-A zBZvetL{sSyrGs(yj7WH(Z<@makqQeNbgYB0`|mm1|D8Jw;`ER)|{`KIjt*w#Xx zu!{}e%kVTs@3Z}#aXUGL{%y%cDKQnHv%QL$jGU1)t&YBg*+m2y9YTt>KsmEtK|;%6 z=})`|)w`Lm2?eiwri9(9f?PooQF8s9vO_@NW?)c5{tk4QCl{;>TiW6M0ctZskQ?97j`D!C7oe)*#9t58=`O)?}*kN@^MD zNhf)^pl?SPrwV*c6(yuP-I?G}kNY3=ELnj?i7}*8;N-n%)Ifn>=8iCnm{oyArcS9T zZ+?mR47;1xMwK^$A`9Xq6D9q5TXAfjooE-rb0q_j&Sl*3dJESK zX^C5ZKy14FD6!n>`1{7&R_#3ixu-C$lszz@TD69I5sh2|bY2Pz<-d^wV(rt+esghF zeNkbq;!_ER@XF~`Zes3o?P5dm{+g@!L+)W)iOU&Iom3M2j}u;n&n+Yz3i8BnOo@huZ zw5KE(CM8PxNMp}`b;KDGVb^?qR=-BYMwlO-@~D zQjxX52^I2fEvb^c%8gSzZhU|%i6tu0H+g5J?O?Vfih*qFNhB(Dl4p4DR}I=l`YU{s z(1F|wH%<+4(<0Ut*qaxVA}-#F2x2F2q6zF%w*XMi2U!;#J;{E2zK_Hdq{6qo`W0Uy#BeDc?`RuZksABhfC}N#%*E!8IKFK8gG(rKR-m(1TFVkGr0Zy0G+KX9?o? zi>KsXAY&H*R5WC(wJ7jipP&?2Y3fKxw)eHknUyZ>YOqM8GaKhfpz;f;SYryFU&o%|7}8{OkY99B5>}v;JFm~4)_1p3UsKAoBMb16DGE}GHi5zXr2Wzs7CK+_U-$!su1b6Q z9_da{UeM5V6zg@cH1!V7Ik1a2{}uv2<_3*7E{6npQ1sv?^CbrKb0z#j6+A7TL9QzN?=(m9QR?a2xeaIuAUJxL zx`?YoI{vms$m%YoW-#imx$0QZ_2$={px^EBc|ouLMB-Y^32&?7t%`57T`78sdpJ_@a$GY=kKc6zo(>@vzt*tM5u^+oP^4S3IJ9l= zNw%Ujv@`5O-Te27Z1AEJH*mel;r}H{Q+net=D3fLqVRAt{JR@Qz2KV3`Jf!ar*28; z9{bUlH)&ol`PWD;zokldI5I0jgwAbrn|IT3Iy&F(Auj|leO7@D8IC&qR-cLe1?L{3 zXRGUAPmAmwXmVI`oRoMd=QL{NM^w)=yKIG$7Rh176Ni{0dRhUOip)IOzsQ7?vac6& z`zA6ej4e&9V?j2Ee(pfitKD&YNwixejZnxx@J(J3C_bdIw19u+HKOo)Nlh6+^9yNz z)Pyi^^nwXsHF^vP!UZ?UzG?OSqb`w10<>#`Lv2`thUfgyxHQ-ntkQ@p6)OmkccJh_I5@*jzhVz2i<$~1) zuMqwG<<`qA??u#4MhU((%nDR6p|Yqqtev|$1vy9yL}#)WT~lMDRsSUao0xBIFb~yj zsF^9FYqle5-*T7V{^tlGi}sO8W>jrgH1sdD8ixI1=7s&h3uLXnB50A)Uh0o`RQtyE zjZH(BEZU(4x@duEK_{}8Ie)hvwC>66GwBcfT(Qy8|w%L-zHZ6h@azQ@u}_Kt~sP-?QT2DfC!pYwrqs*Ek`zKqxIjHWFk;8zYk;kc#}8~FPrBBTKk zas~q|C7|?o=W1hG7h#d~T-(5DY?rlf;5`l;CrJGp#E;v%&O7lf3Do<1&`pJ~Fujh$ z0Tgv^!2@}O;>Fx}?M>K;v7vfFPuYp}O{NkxaICuNIT~%x{IYBr&tJc8JxG^{vlHtb z`8C7tuNOzAf!(%$e`_}O0W`cgyl3wXeF_@roMW2yk!=Avou0-M*0i@fkwO6!zVfJ=p-A~QWGNqg42RNfeQ zBZ06n2L7Ngri~>#Ck5i<^oyp~(>S&B?i;&vO_KT{bo#;V+Z$!!^)P*!1})!wsZ9+O zJ=VTP%C^B#W&h|Ce5<8ynzQPOf~uRVxrtMDr%MLN+tv1YtlKlIG-8HLZYM!ZkoC?; z%I2t80>sn}FT==Ua!PuiVTq@bd||OUmGi?{?x6RZw=^*f7|xsz%2ZGjG=~o4%H8~p zDyY8d+)KE5r{Q?XFjY(szb@z`$F}o#(HAu0Q}aa*zD`pLffa;j5wyv0A#QUnE-&vI z_)dbPD?XTRfcuB|?y(t#DJee@-gZBfT}4TlA=ng^*U?bkrhS~|OmgI6*MQGq?n<(0 z)X2t*fp$GbE=WC(9)s@IUrMe*^=WDX@mi_-YSUiC__4R2-@N?;Z1{Qo z0}$3UzeM~?(q|@XpBiTN&@KrX0WdUa2b!ECrDgW{G?yMn^O&6`7b#7Djse~p&vA()JHDz`O(uYY2YD(Fy|&Z!c0v1#7v`fi%t z6t-a4O>8xaXh-!gfYGrbt#b@ZkhCDnw3RD|qba}Kw7bFYT0p~AqU+};(UF7`_L>SO zNvGwt3{xXfF9OMMpHeZ8Owu@_0SUJOfdmRMq_1ufH|AK61dydOd2Ln~di2X$%ILcC zFsx=WSx(3;p3{op7y-s|Kbjv8JP%`_OrVb)18Qh-VL>qeN-nrOzg5>KSUM8s?kC0U z(=mamF;e|?(H?r;jxqlp+Xb!rs(9p8HYxcLz08s9S!;<~=N+S;Gw;Bgp9wo+j%n`3 zp1o6{)MQ0v6maQmrJOa1@9H;OSu7nu8NSpU`!bqG_I z>yOVyN%GV_F=x3S9;E`NT9GGQB#m8@*-%AFaVy+my|Xv93NiII@S)3oO10^KT7xpR z)e0&l1PoV9x3yPEGH_C`{V|_zu@+KEEk4q@F@VNun=2NTc`_Wzif^x+%7IIX&WE(N zuAYq)@)t&$LYEHzw;M>7M#=)7Np;6fInBDH#Nr9T+(6o6@*e-$bii?vqxU}tS4S%Z zs@{CoL94X8AEM4`2K6729z}T`)~Z{-3g*t;OrR5UsNina$u6&!7FxH-iZZgS5Oc3d z>VGN&KNa6>WIJfrZqqFbU-^g`*7mc&fQffPePevYq#XWr%ZmPBB3+pIM5c(K3$(8q zXB67LulhU|7Tkhb7=1IKy#PfW=iU*>N!~QYj>-{2rV){Ny4B8$cqfeP&vZ9MaDeu5 zE0!HlsetPOS>gEf1)9acjxGLvaB>3kH>?wmj_GX{ratwM}?z3l;75R(=% z3)CAVsW6UWEp%yb8_gSvoz-zB<6KWn}9G4aoZY* z2UED#({RgNq%7rzat=84=zN*}6r4>mBD)%n#T zb6Zof($X(_Z_wu1aS!Jf6Lnej5v{kYJ2B`Q`T95r95Ci(;aI#qUgBd5fS5L+e2D1Y z)1xGx7&L7nRHKh|prjD-r4#e~>nssLWk5T$Qf*qm3A%{kyf=yA7W@RMH1a#fTpt$h zlLT4ay_}$n+)mLk#!aBM8mj4@ADkRTmqmxRS~=S*K;1`d2)>5V;5$3zzkr7w;3->cRpOO!iiI>=p{DvTT<|-V#4pO^FF&bz1L8E(n?;n=hI^ zBY!xxE=xJ#8>x`1L{cJ9uwJ<3W=d8y2GrHNC-=`YkKjN7uUmX~mC=!yU0RSm1Bc=E z0);W7%3U1>_We?AyY%i#8qx*xEv@S$@t5;&F6#kzRC@#;L3e6G^>3*A+@oh>%Iu>* zEocEZ|LveCPa#pL-f_a|*j5*`HV~g8*Ymg12zfoUZlXwx2u0Kjbls~`ValvQ)zLSKJPh?b;_+uNwahTb%asPL>$db?Sa*<c+Y6l@i&(rSHsSa0uZU!t zP?c`>({tTRoC>eeMi;Qe+R`y_ zT{E6o&bHjty!%aR)PUuhudp$~?(KGhp!Zid28bMSFK+>l8Cg$rS>pK!5Ha%1>_~1O zQ5uN7{k54N0Z!8KT9JIQ{}4bUn1Oln*kn*CDV3$$Z58uewZwAKNJlRDf5Rku>m$m6 zI#)b~dO9Ac=lfw)xX(=(r&+~Qe)`6{%06ij(>u7&>4}TRoIyb=i(He7&QZi0Z0xi{ z37B~pw$A-6^W+F2c}8qBPc*pCkS?SJ^ zl(ZE1xNlfHc-Dxp_>NOO*I+}&Jdv_&cAsps1XG0B ztMfuf@CJ7FXcI8d506IpDeFBWQmMllUz5u-J5B9L-I4%rb2%)v)&Vt6k(4}>*y=Uz zjr~-~Kr?*ZmSC;De|vQZB6DoAA$XX*l6x8VmVK};f=nlR_bXm*(A&0*$Hrdkv;mMb zrC2IHljDG7)L!IBUuQxP{CXl?P4uBLB?i4kho#S<`AS6 zufGEa*^UQ|B2-A>eK{!OzE>uHsWm0#vU$P0_QL(B4Z1to(tEbIz%&fmS?T??|1$%VXKjxh!JA$Mjq+e5u**`PfLFUXMbJk>b!sPC|N)WLXx26mFoWGIduqvSO6b~u<@ zi^)5U-WEB=2@Z}#D`S%?tLf^UP*y6xg^9RtL)sHMuWpE2W9NOy{u~p|zD{AzAhS~K zshkgWH}tg82kno?dcLX5cPK-Dr0x$_c|76~eMjWZgWc_e5Wq`NliztGHdej!{tLVI z^0PH34~d?CQD(Oy*IN_iK?ULh z)BR3OO$Bc7rHRK{85R3B$u*V}L_X^UZth!ZUItw^XSJ=Y?xrFwd-xD5y#K;}Dvnk6 zccpn+;g4~2qe*BOPkMszTyog1EskkhmKP=?BdIM8-1Z2NxR9+Cygx6QUws{ocV$6> zr9GF;^=X>gjLhh^yq3)^{AK&gj@pbn=#hbCh)yD(J^I$NTD#$@#Puu3E^fGE@0Q+F zxrwuko^OB6{@+SMiQ*c1Q-k$&{*IuLPLMo{DZ!{m3WS0BoV=^*r7^O-nmv|Z5V+k> zaui-P<%JRUQS67`#Ii({vQT#| zWMSC4*M8{%=9xg$#yb7%XTFo!~QTd5Q_;)mmd0VAl|L32Sg4{sn#0-BzFSZu+wRptw6=^wI1)OvGE6uSzUo zJPlu2+WsBP5siQ}zi=)q^F2X>i;t=W6Ko&yktM5`kr{-~#w6k+HH{GAuxAtDdupL< z78VA*L4g9$N$x3tx8RW(H)#W-T!;aUjD1o8IL)7`y**J|852f5Q1&A$MKKrJUzRH7jXy zqIOx$$G5_*`ebBXr$cp&yeNR_{pkPvtwZPRWr(j}%P1yH@D9|K372vl_t9}(K&{WAtxoWiIU%;sFOJLpOUYwc_gD-AWyWI($6EO}<0 zjDyYHAQoYA8hxp<|EpfI9EFsTWo>KX=vdnKFL0Jk(B`MoWG({X+1*(M}B$taq&G#3+E?N){voj zJ0ldnxD}QBuHk>)k}ZB?{#Tq@jK7rfrA>~3dVNAf*yVs$KII?bJNiM#msGeUIEv=C z@gf;IQUAYk#rYs;^fSTA!x|D{rN7P}G)B_Y)}_A@V)+6}bs6-5;G#c&j4lm$=2S^| zxjl(rTIreXB@Z>~Y2>0!=L4^%@O3F1vkN=|kAZTLd1Z#wyZaj0}H99KXibH6e{}N+|ZJQGoL}}&os?pTV&`@TzCszDgX6WT= z3yMsHNllT>JWR<(yv)v~GgJN`%rh9>5vJZQt7}PcPpBv@?Spik(V!M{zE!l)Lvxl# z_etT8Vruv_q|?~5QSPc6AL^F}?WySSPnxnO`i7ZVp1aQ_rLt9S-e^ugfoshy`Bv91 zdMOd_uI*CNI*B!X!eO1Goo|B*!Y9qm;^BslBO{A8tpIh>!L0sA+ESN6^t{cJk<8#6 zS1-JWDbCQ!IXdksdK({}P3uDJ?7w1gh}iRJ;YG))Z3axI)8HC)^Ovy(KERYNUh_rn zgT4gwr<2vNrtgWr&+1X>8&i1><>bsW{gm<9 z`Bpuc2fP@quFGwT>9-wZQh- z#ysvgLh?3@zF@|BE%_;L-|jCahj)cek4BfHO8L~~x{hCB4HcqEVq=sBgV(6cZ8Ll? zxw0H~;=|Z`a}om-`S_IB$u(yM*x7NiR33On6^;xej;= z=#AA`|CZ?NKOWC%`Q!;*I9*jAQndE7#}GYAk-xtUL~6`$f^5I0-sA{E;@O-3eLX(| z%=ZS_)0_e_7KXv7<6rch!v?Wbl3wU+Iqdqadr#My-pZerySi^Mq3D&U`rB#e_LU@Z zwwdwUdDJ3&zoyvKBS6r!O7i+1`D%>f>H|m@lEarWcDKMYx9!(gf9(BCf_Ucfc!}}+ zsP>QAXyk2Zi`6+lh`H(523KLb*L|P{~xEJRH8TuzvsDO%^8Y4~mS9B(Ey+DN!iTZ7Y z#r#LmA{&8PPc`$EVxiX6*0xs#?bi~winJXxe-CngDQ&n;J*HP}w$`c*F9d`S`DSES z#}eLiyv!6~^hMZyox&I7tKp4PsDP&}lHkwomk-rwzB!P#`Ss#r@b2-GUliVVPpkSV z*G*x!k1&a=|6Xmzc2);KbAFh_tQjCF!D@G3p06v8;tN?S|nMcNH z%tU(GPxj-%%oID+YK2d{%>AqC4&x1O!l1a_ZBs{p4KY~M>~E*>o_E(+W0fEA)!f>X zsX3eBeU;xG4njPWC9kf-xi(uV_RSp67iNAzO>4(&M?N`7z*?jj{<|DcYl~Lrs6tP7 z#kd{}-x-ZzbZV zp|Mnc!HedGyc)DPLR$al z9n)aUm1vHqAPr)TVYF=c-Gj1BH;5;Y+!X{AoY>8qPHe$d8VtSlvcO)eFHe9$h71x< zziTmTTXtIJD*88^IvhXu%O$5^>SZh+ZaTktzwmDFTZ5QnOgO{oF8J(3J4|y z^n>6~L$Ce@_qQ8aiZzu|Y^)VA%o6o*GivvRQMNY8ccicDx1DnpZnPCd6ythHF>RHU z7!{Tq=7BizXD?w5sdo7W#1knyl+KQn6F*xHV(9-H?E9y>4t2@jEn1p{`D!np%{$jJ zirUTqP7DWN2wET{r+}v1aB0O0rpN~yC&tQB_ofP}9N=QcSl6h+b+Xv6_{|LA0l0s& zVkRJv8L7N4qGnPQ_x!?3@uq(o!c#Y;XC(^V5!8FDt0wSnx&er^dv32p0nI#0!mavl zQ(j^w-FfJKi^wVggX#+o>#y<=cz@M!`+^&i1{0q6{1EA-6~QA?+x$!^&zl$)t)11> zeJ@uXwaA=~lKl8UfjBQ}7Z32hybg{@KD}7r`}>W{3ODpZ2roYoq}0*GKrN>bso2Nc z>)9JW8oM(_XJZ-O^Ig(lKafOL^SF=8ZgCVWzbEFcX2iU(7}D=ChGjR&eS0ynum5-5 z%7yo^TlltKIX%VHLuuwq@Tg5pWH*d0c+YdSTp~pB9K?#V5h5vO`2-j9TE+R(Q@eh8 z!S6~Yc&_yFT`rVN#kX9(%OV@Z$HZx(F<3=$Y~aO01)|tQs;cPG_{A!2(Rt>c18xh|Fd~@hiKLLNdHEOjUXyWj7BqT|d`6yawy`E^)D=8|V z1ua(13&pEk?uv!_Zbh9~T)bZISL#`aI1xRZqnSkv<4H%sHv|GTE~_fqn~mm}5{6sGs@A*EeTEjjl9(K*wI539e;x9l zvQ8`BRmX@W&qt5yo?kj~%SVW}*Ocv*N>M<*^W0B^*rm9sE~^ezjr;XJ9N#!9PM63R?Um!TYq`6#+55X; zgUL<71^$NCu#n=QOYZ~SM#9n09A91CRcbwNfKMg|fY!WBKeQc_HTA55xRqU8MLsne zL+-`qHs+y69Hr+rrvFY!tSs3<;~K zJk?ey}p$mNE76{4FF zSnBO`B;_Znm}AsB>)J+N-6q8hv}1qdx$6@BYR=1JK?b7rUJn;tTFLq!WJy|0R|{>wpmM!4$CJh3V|^3BD;6gVSJpU=#XggC zijq7jx$Hog0Y0&`Xq&6IO&L+ePBuvA?1;l4T&WpES`-PtVzWL2i|4Qmt(ewOP;5JqbFd*ux>}5le-c?}RHSE`v z6r}ndY#ZyC!Ufao$7@O-Yaa)870eIM&}`-SL_71OtV_xP{3U4s-bej1tkH=1+|P$b z#)+i@pNO*t^h3d0)Tc!21u(Ht$rzC}GW}$Lshr~_l`$GoHm~hjKcF7dN=Lg~_2=Ra zz0=&>;auk>HP$5ULaMTOCCGA;%~*~f$2^qZ$wK{=%8GfJdI_{?z; zJ3+}>Ug%S@jiwR67cjFkr0Te&eK!^>k9t(Np=IUXTe9FeNRgUV){gHTd^L`JYD!8y z5;_qh)_$+bU(0-`ASr1h)d7odO7)^0Z%7V~lf>owq$QhtNM=1EQAG}oGkpr$$gtvn z=fp3;R-@F@%yN;V3-`NT(1(=0bE%iPwO#aa1zG8A?=jT#;B-7rf1j(Mq^*Eq|1u=` zkgcmv6>6DD3pt7SY~CJ77)d9q((``FTw(e+=GSTX#EI~*!PBsEuG2334xsI=--uIzd9|2o5>Y^cOT5@NvAdBYZDEgC@i(0g072aJi90wV8Qx+?peKA$)6=E7 z8G`S|>KID9OT{eyF(KJGeEtIL)JNYHDrg2M*Kx+fqYA%=UGEuJ&O64u;d306qhc!h zUllx3U^uMAbNiJxX_Kp#CkW>0Q~KVS-$i0bi68VG@=|*Ghboi8l?c&o@QNLL+`}7aSMUb6uPmUFX zhzwIp7O&e^;+q?r7MvJ?H7&m@-%Up}%F~eeTaR)h%ZsWyOnnahX)l;fv7BUxE5L2` zg)bfgz#oALATYMle^0HXmkauSx$Ql427s-DqP2ORsD47?9|%*7&&%pk+O@9^%N*Vp zCwueOBIqi7cYPl9#X&aOGgYO}>P)mfzk zm5-k-gs~a5{G6dHk$$Js?NMe7T1wtOtLK^5^y4+=fV^?((sPp9fV5rC`{Ulg;sTHg zRz(fZ{iq2siFilftx21eMpN%aGca8q7$+2RtuiX2F+U^%Wpw98Jjw$MYm07AA+5CiAKQsL`U5hEB%-N5L zhHw%8?pJQfw>+e4Ko3v+5vCHMX}yndwCdwM>U0}Og*b~rsl?Nkx^X4+d=`MA{8hBj z=q4^ha)YI@oLIyobvsP{b;$41oq$SvCXm+-zNQo&d}T7=H?hsC0`uM!dOEof=GW(Y zi+;kqBY!HKn{_Dmp18C_-w0O!CD&wW@7<8p)dYO#(2lhXfDPYB+$s0&mNbhqXk6y( z3%p&JJ^hOOQlK+PzosZUo-Ol|+Uy(onErGw|Fp0fUkTD;?kjnCH9qESBXepY=S9yr1Uv38CA-0Blyr`LI&i0dx`AuF&Xx?df*Ro7S)s z_phTw<|z#3^-@UFC=`B(Ek9*VLxx-(wl0mFNpIP~!PQ)(2V zz`wJ4;z*;{jvvsqWIx$9?@u5SBEBB&$yJVm-}cP!B@-6!k2r5zn#^6!?%jfaVyfWo zD5lA`34(7-oFL7WBWGYI+r*Ok$q)6$Z;QSHvS6=S#I!E7fSYtug76`8ukz0zDG(zg z;plgN*W=0WI{Rd1*k_}yTLo-!9XPYa@q0Cw`ITHL=$42SbssHQLY^&FOIxR(1;>Hn>F9{W24R{ z>@@bq|J=aUAZ&?{w#@fe1Az{r`s&KQbdF@sfFrloc@G%Kk}peUO~u=Hr-YJy z0KpmR75j$vS|N{*X2V{Y_34z6GdTnVqzfp|pr8HoHI~D=rr*Xa0*V|UcY~z=cH=@b8duYgr)yV=g}P{ztI1FL+tViUp`adH?NZBKO(N zeB&&Z1HjCgCD;#^X&YZ#v=>`1sb6(=1fBVPJv>`1E#50JDhy@)I9vW}I$A<;pJp`U zml#lGDH~}H{XHja(@_(zc?al|w3$H6XcBZ)f&cj`P9+Y`gU9OsDe7_wd10)L>OWKu znKt~n$en39S$7h*eXrrLIPRf+g-?V;OnSZBEulW^xW+#NXvjB}bgUiePuv~tWjmN! z5BLEx2MHiMbgDQ~9PIS@R`Km0xetOIP6Z6KD!^i+ok_`q>d@WGTgE7M=FK;lVb6Mr zXH7T*c>d#c68+k73QgrW=baf{mq{0qn~!3nh!*^e!yGUKlR2ykZfc2gJSMUUzuT*P zW^Gbuf9spUorn2cRGmuWdiI7LWZv8zK+S$x{Qb@La`22z&YrCVYT4?WH%M0^Yl}~8 zVmo~qrS1Fhzg4O!gEm0if%$^hiz^b`lCURadrD)!);_3QXs!o(c%YRvy zYf))aPo=I0S)bb-Ew`WD9QsKd{Spe9G-ni7 zqI}!52_36>Z05BZ%Qf}mh5kRuB$If>C={g;M-rwr0bGf)U;V8^IXH=X@71Vn=aa;9 z2dv$Tmg^3+pLS@u`mds=q7k55x}D;HahaC*tKZ#iKwp<8eWZmRBQ4tU&xw6)O+^*` zrn)309wY7FzNet954r6El;*kz_2*8z6ScL?!1`EmKjnz#)*b%~XT*i)P%I`-V`zS! zJ2kdE-Kn9kqz)Z=c?OSc-eSHFj`-c49L>&!Q*rIs8002yyqeye@OokM zbP17(<#bBCAnxKd&ZE(m?E+$VQ0qnKhOM`#eSl4_njJX9GL~JTrjt)X*j^weTWcRV z&5p6&YOk;)Y3HdB+TF{{mcJ|&>{?a(Uj_LC44tgHZ(TPIetFZ!Qsw9F0>|+jJ#3pm zDES(eD-Y+KOD_9(cZKs0N}0(!fN~dEO~uq0?l;N}X*zoLR`?ocS=q6%=BwD|anqEZ zczQ_#LV6#kKS&h^>tFXT^WrGNweq7PQMtJB@v|t-Byr$>C^ z$lQuEAi+8)r$&+Bb3khny_XxcepKx+ec+mb#QR%kHl}LeT5o`zwaX%}aagEB&=>Vh zYT>?G9#dtjW&^e$_{L)Wy&0RaQYnWg`v!p8$k0eL^OOW}p^&6>R>#9;m?)^CQtyY% zZMnH5F!MU7Fgg1ZiT@f()q)W|Prd;*mOp9j=Xvp&25FO-_f`Ybk9|iP&GoPDUnzVb zRMC1U|J-f?_(>sIcxLxxLgiF4F=M=z*M>enk2izlRN{Jh(E~LNc=E3*=3KrPfid}u zb55eRTQ1~JoslU7I1?_#tcDrwiq}UgorNQO3ymJ7fO29;96o2w&CSgL!bL}u z`36UrR0vFBq`0EnKIzJ`KiCYQiY^b-{$5D@y!+1ZVAJ?~o(A(Y?zq_(%+$Rf&f9iG zKuW2s<}bfY{ACEwEL?s$s@SNOsAZPo;WavtFFiR9bq=cryLf>n4Ams*$H z_vdVkY8*)#(j0bjoB?MLW5zQ_m-B-~C8q5eD^$aM+XJ!b7ZwN5f$RK=f2>#Rnu;{T zt!ito1Ic)+xi!(JTmb{O6GOmUg18;6?|tM=FQuZFuxJ3%h;85YdhF*V5wnNZR`6zb zI(`N(cN8*^#=8l~jE5$~v0VLUV`I~<|EJ6%XoWfv54SicoRFcn=A6jUNa!>}1W!s0d48PS=>F2F#DQ7aTcM29lG>c-d1l9k=rSKJ!C zv+l_~4R$DO`;}{0!@8kEN|QONMC=mDX$@{4gdjVR9=lV5vi}_kaTJDP&}3y`o>K+I z`6_PT(N4l?1$X0W9nSTKq+<7hLlfeMyT(1`mak3WOMe(NmfA}NV%*Z-;ecHb0dSbo z6UJI0gsZ>MxoY;${%uoB=|{!o%TaF52$|6|2dpBJFe_X&Uk1QIqn@}5L>2wZzXDZG z+QqI-=7fOTjmFf zBbp!0^l)}-Jspy$Sggo1ZWx6@b!;Pdy3!yL$WE z3BdkBp5bnbZNt`Z>JwPPr&s*eR6$HhITYDp$d%9rk3KFdqPh)eUCLfy!DjnMyM7%l z`1>!**N@oc?vdMAMRfA!N^Kue88mTvCk{uw8h;9!oao?hNE0EbF06meyeU7YE27Qr z$M-T37#G!vqed!+!4=91whv()kvu6lih4PWS zY|M4rUA~K;=9?+retn;4BwjDiDz6x~PX?VC^ly?bOhUs-Ibj(+T~ ztR;p~wj+f&$Vw-B$|&E$yw)bP0{`VCD4SCh$bQn;$Ot1E;$;9Uc)g-1`0oAcx#Vrv z3Q5`cZ&f;aqL6VuYC`A>Tr7ivZN5YT*?0mq7R(r76N2#`WiRpWVkSikzuhxGAR9!f z9VDReyz+q*TJE~{q~Qlg?9LMKJ7tN-0`p7fj3ktI5aD97&=nEq%|7I=!!Kq&|W zY$O80yZ4GT3YquD1`JueGCa|Rqm|1oZse2aT(R>HLL*<0Um zNkYeMbSWS1pqHt7`hp(y(SMD-q@(zfFYl#@J49k|(yc@oaG{$an>-eX1V-a_d^E|< zgCwF>s#HvWR5~mO(1R$DwvEWX&pRaOCADAU;3AMqq8`B$nTls;rHl(G$2UJ9{#kb# zT;lM9tu`6G`PJW*pL>PawE#WdZmgr_N*vY7mnWVd)IMV?DE+y9A_b3$d)U7b8jvm)`tNb3`7b8DND++P7p#bY1q`Q8)ISTTnlzVDhQV^zWRiF;i5w96p(|`>?^Zcyj1Qs=z-|3_N zIi&e${5=SdPwLr))y`hR_Gnb?dlW~&HjSMJxxMkgC?)SG9s5srQo<#d5jnkAuPm3Y z{+#dSgn=I?S!A6{I-tqq)V@&-vASr1K@8r#@g4d1wOTy!@52IC*;{&Rig*7HmTLBV zQ_=PmL~A~NuQwDdR@W+-2|FK1Ba9^5HSL>LB1UWzqup>dQlC_-g1sJ!$)PS?cYhb@ zo48J7!=EW(3Eo6`jHV06t4eKKI2OBux|%$IKyC2;sLFjexiEy;qfEZ{!jEtcJ0^k+ zC?2ZjKa!hjL|j|OZa$kwLkS{X;z*wt!B_`9d9*SZTqpi5bDA%%{quu^XQ7MlNW$+# zB9sj|AXqz-7YCi&Isi*4nW>dhs&1@N2le5%&FBW2o~lj9hmrHfN&dZO%++X)u{RPa z7FSyR_b-C`3dfWlra@Pyb%sU1;OPm}0`e}mlKn1H{YYvq(ABS_35LOh2=uM-H@Nut9Evf(bN;2S*P$!-yv=J1jJ03@E4u5f*P<&U-&Ue(un?>NW zQ0u%D$Z)vHnC~c@#_kglH~978toE0a$6s_3I_?RJ!chC~sv~U$_C&b^_i} zl{FVQ>zd61gZ4D8$yqHoe*9!`=3yt|jY%A%%+IHNi;W)Bh8}i$0XsjVDkWH`Y1c{1 z$!P27g`?6;VwuiWpeliV>z^qhRUB&Xf+TUil0VB(Wuc^?--emKQgET+e|}8ZgfkLP z-?~^F0pZK5sQOMECmG}VH#N0OE$BH7_CDa_{Uqs4>q|VE^;8kQ{WQ6FL@Y*T-Yx^Y z}>tVDy$;r8O6ZTk=bE%f?j%slqsm_+0?ysOj0Vz!YRf@EgyB>+Js*5L$@S zB*32OTJ|jgU>?hp@0NWL2f_^2yu{a5nLJn;Vkzo119 literal 0 HcmV?d00001 From c5ba8d8a8f85a126203213948bc1bdca9871d70c Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 15:53:32 -0600 Subject: [PATCH 527/590] Update cpt.md --- ALCF/notes/cpt.md | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 04b43c0e74b..25b1b913ef3 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,12 +1,18 @@ # CPT -CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed. -This document serves as a strategy cookbook for our current runs. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. Here we focus mainly on the learning rate schedules and the data mixing strategy. -We are not exploring alternative approaches like **model merging** which might be worth considering especially when we train on the math and code dataset. -Here, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1, \cdots, N$. So stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. - -## AuroraGPT V1 -Here we are using infinite schedulers where we warmed up the LR up to LR_max then kept it constant before cooling it down. The advantage of using this is to avoid the need of rewarming the LR when doing continual pretraining. -Since we have several training stages, one need to keep a buffer $B$ that contains previous training datasets. Recall at stage $i$, we have the pretraining dataset $D_0$, and previous CPT datasets $D_1,\cdots, D_{i-1}$. The buffer B will contain samples from $D_1,\cdots, D_{i-1}$. +This document serves as a strategy cookbook for our current runs. CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed while avoiding catastrophic forgetting. +In this document, we focus on three CPT approaches: a **data centric strategy**, an **optimization (LR) strategy**, and a third one mixing both. This means that the following are **fixed** across all runs: +- Model architecture +- Sequence length +- Optimizer (although it might be interesting testing how changing optimizers across stages affect training) +- Evaluation tasks + +In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. + +## AuroraGPT V1 (Stages 1 to 4) +For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that was used. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. +## Data centric strategy ## +The main thing to figure out here is the data mixing strategy. We need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) +![data mixing](./assets/CPT_data_mixing.png) #### Stage 1 to stage 2 (weak distribution shift) ##### First strategy to try From 0a5ac67cfa932c244ce92531187c1d9ff42e7330 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 16:04:07 -0600 Subject: [PATCH 528/590] Update cpt.md --- ALCF/notes/cpt.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 25b1b913ef3..67d4f409209 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -4,7 +4,7 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, - Model architecture - Sequence length - Optimizer (although it might be interesting testing how changing optimizers across stages affect training) -- Evaluation tasks +- Evaluation/validation tasks (need to be consistent across stages) In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. @@ -13,6 +13,7 @@ For these runs, we have 4 stages of training with the first stage producing the ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. We need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) ![data mixing](./assets/CPT_data_mixing.png) +Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. #### Stage 1 to stage 2 (weak distribution shift) ##### First strategy to try @@ -21,7 +22,7 @@ The main thing to figure out here is the data mixing strategy. We need to sample ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -This means you are replaying 10% of dolma and doing CPT with lucid. You should also start building the buffer $B$ in prevision of the next stages. +This means you are replaying 10% of dolma and doing CPT with lucid. For convenience, here is a copy of the ***mix_datasets.py*** script ```bash #!/usr/bin/env python3 @@ -101,12 +102,11 @@ def main(): if __name__ == '__main__': main() ``` -2. Run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): - +2. Start building the buffer $B$ in prevision of the next stages. +3. Run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): ```bash -DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR=0.0002 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune +DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune ``` -Here, we are rewarming to the original learning but you can rewarm to any LR you seem fit. by just setting a different value for LR For example, we tested rewarming to LR/2 i.e **LR=0.0001** and 2LR as well. Here the following options options/flags should be: ```bash DATA_FILE_LIST=path/to/your/tokenized/data From e7ad38286cf3f3c8118ab1afef2508ed266ced27 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 16:12:34 -0600 Subject: [PATCH 529/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 67d4f409209..ec769fb8767 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -11,7 +11,7 @@ In all that follows, we suppose the base model was trained on dataset $D_0$ and ## AuroraGPT V1 (Stages 1 to 4) For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that was used. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. ## Data centric strategy ## -The main thing to figure out here is the data mixing strategy. We need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) +The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) ![data mixing](./assets/CPT_data_mixing.png) Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. From 307f78fee1b12e052209c80e6457281a1c636f64 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 16:17:45 -0600 Subject: [PATCH 530/590] Update cpt.md --- ALCF/notes/cpt.md | 1 - 1 file changed, 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index ec769fb8767..f49e8a27251 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -22,7 +22,6 @@ Note that you add data to the buffer B after the current step to be used for the ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -This means you are replaying 10% of dolma and doing CPT with lucid. For convenience, here is a copy of the ***mix_datasets.py*** script ```bash #!/usr/bin/env python3 From 749c5d37b0ffb572b94428310dbf2442ab3c1dd4 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Mon, 26 Jan 2026 17:14:22 -0600 Subject: [PATCH 531/590] Update cpt.md --- ALCF/notes/cpt.md | 41 ++++++++++++++++++----------------------- 1 file changed, 18 insertions(+), 23 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index f49e8a27251..294ec098e0d 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -11,14 +11,18 @@ In all that follows, we suppose the base model was trained on dataset $D_0$ and ## AuroraGPT V1 (Stages 1 to 4) For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that was used. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. ## Data centric strategy ## -The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) +The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. +See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) ![data mixing](./assets/CPT_data_mixing.png) Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. #### Stage 1 to stage 2 (weak distribution shift) ##### First strategy to try - **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then - 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. You need to mix data from the pretraining set $D_0$ and the current CPT set $D_1$. Always start with a small percent of the pretraing dataset i.e 1% then increase to 25-30%, **5%** is a common suitable choice. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do +**YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** +Just naively continue training with D_1, no replay data. This might work here because the datasets are similar. +##### Second strategy + **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then, + 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. You need to mix data from the pretraining set $D_0$ and the current CPT set $D_1$. Always start with a small percent of the pretraing dataset i.e $\alpha_0 = 0.1$ and $\alpha_D = 0.9$ then increase $\alpha_0$ to 25-30%, **$alpha_0=0.05** is a common suitable choice. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` @@ -102,32 +106,23 @@ if __name__ == '__main__': main() ``` 2. Start building the buffer $B$ in prevision of the next stages. -3. Run the following cpt command from the Megatron-deepspeed folder (you can modify GRAD_ACC_STEPS according to the batch size you want to do CPT with): -```bash -DATA_FILE_LIST=./ALCF/data-lists/aurora/mix_lucid_papers_dolma.txt LOAD=/flare/AuroraGPT/AuroraGPT-v0/checkpoint-copies/checkpoints/ws768_ds_stage1_nl32_hs4096_mb1_seq4096_gb3072_sp1_pp1_tp1_bf16_optadamw_lr_lwf_flash TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_WARMUP_FRACTION=0.01 bash train_alcf.sh --universal-checkpoint --finetune -``` -Here the following options options/flags should be: -```bash -DATA_FILE_LIST=path/to/your/tokenized/data -LOAD=path/to/your/universal/checkpoint -SAVE=path/to/where/you/want/to/save/checkpoints ---universal-checkpoint to load a universal checkpoint (not needed if checkpoint not universal) -``` -Note that you might need to convert your checkpoints following [the instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) here. +3. Load your checkpoint and run CPT with the --finetube flag. +Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md). -##### If loss is not recovering, one can: - - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. - - If still no luck, I'd: +##### If loss is not recovering after increasing the pretraining data weight, one can switch to a LR centric approach: + - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value. + - If still no luck, one should: a. take an earlier not converged checkpoint b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) - c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset.This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) + c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e $\alpha_0 > 0$ as you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) -My guess is that since the distribution shift is not too strong between stage I and stage II data, you will not need to experiment with the latter. +My guess is that since the distribution shift is not too strong between stage I and stage II data, the naive approach or the second one might work and you will not need to experiment with the latter. #### Stage 2 to stage 3 (shift to math/code datasets) -1. Start by trying the above strategy i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. -2. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage -3. If all fail, do the following +1. You can try the naive approach but it might not work here. Then, +2. Try the second strategy above i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. +3. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage +4. If all fail, do the following - take a checkpoint before convergence **i.e before cooldown** - Continue training with the base dataset with (a cosine scheduler decaying to **LR_max/100**) or (cooldown to **LR_max/100**). (I would experiment with both if resources allow) - Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) From 16fdd8816aa3915bb3bf8f2be375afde74d0b583 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 11:35:05 -0600 Subject: [PATCH 532/590] Update cpt.md --- ALCF/notes/cpt.md | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 294ec098e0d..075d3837b46 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,15 +1,24 @@ # CPT -This document serves as a strategy cookbook for our current runs. CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed while avoiding catastrophic forgetting. +This document serves as a strategy cookbook for our current runs. +CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed while avoiding catastrophic forgetting. In this document, we focus on three CPT approaches: a **data centric strategy**, an **optimization (LR) strategy**, and a third one mixing both. This means that the following are **fixed** across all runs: - Model architecture - Sequence length - Optimizer (although it might be interesting testing how changing optimizers across stages affect training) -- Evaluation/validation tasks (need to be consistent across stages) +- Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. ## AuroraGPT V1 (Stages 1 to 4) -For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that was used. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. +For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that we are using. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. + +| Stage | Dataset Symbol | Size | Source / Path | Notes | +|------:|----------------|----------------------|---------------|-------| +| 0 | D₀ | 4T | Olmo-mix | Pretraining dataset | +| 1 | D₁ | 2T | Dolmino and fineweb Edu | CPT stage 1 | +| 2 | D₂ | 1.5T | Olmo-Mix, Open Alex, and proof pile II | CPT stage 2 | +| 3 | D₃ | 0.5T |Olmo-Mix, OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | CPT stage 3 | + ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) From 790e1f062e0069086d8d7b6f7dc443a9f94231b0 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 13:26:10 -0600 Subject: [PATCH 533/590] Update cpt.md --- ALCF/notes/cpt.md | 69 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 52 insertions(+), 17 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 075d3837b46..eec417ee566 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -16,8 +16,8 @@ For these runs, we have 4 stages of training with the first stage producing the |------:|----------------|----------------------|---------------|-------| | 0 | D₀ | 4T | Olmo-mix | Pretraining dataset | | 1 | D₁ | 2T | Dolmino and fineweb Edu | CPT stage 1 | -| 2 | D₂ | 1.5T | Olmo-Mix, Open Alex, and proof pile II | CPT stage 2 | -| 3 | D₃ | 0.5T |Olmo-Mix, OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | CPT stage 3 | +| 2 | D₂ | 1.5T |Open Alex, and proof pile II | CPT stage 2 | +| 3 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | CPT stage 3 | ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. @@ -26,12 +26,31 @@ See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. #### Stage 1 to stage 2 (weak distribution shift) -##### First strategy to try -**YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** -Just naively continue training with D_1, no replay data. This might work here because the datasets are similar. -##### Second strategy - **YOU NEED TO USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then, - 1. ***Replay the pretraining dataset*** To that aim, use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. You need to mix data from the pretraining set $D_0$ and the current CPT set $D_1$. Always start with a small percent of the pretraing dataset i.e $\alpha_0 = 0.1$ and $\alpha_D = 0.9$ then increase $\alpha_0$ to 25-30%, **$alpha_0=0.05** is a common suitable choice. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do +##### Strategy 1: No replay +**Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** +Just naively continue training with D_1, no replay data. +- Continue training using only the current dataset \( D_1 \) +- No replay from \( D_0 \) or buffer data +This may be sufficient under weak distribution shift but there is potential risks of forgetting + +##### Strategy 2: Replay from pretraining dataset + **Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then, + 1. ***Replay the pretraining dataset*** + We mix data from: +- the pretraining dataset \( D_0 \), +- the current CPT dataset \( D_1 \). + +No buffer data is used at this stage, \( \alpha_B=0 \). + +**Initial mixing weights** +- Start conservatively: + - \( \alpha_0 = 0.05 \)–\(0.10 \) + - \( \alpha_D = 1 - \alpha_0 \) + +> In practice, \( \alpha_0 = 0.05 \) is often a safe starting point. +> Increase up to 25–30% only if forgetting is observed. +**Dataset construction** + Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` @@ -114,18 +133,34 @@ def main(): if __name__ == '__main__': main() ``` -2. Start building the buffer $B$ in prevision of the next stages. -3. Load your checkpoint and run CPT with the --finetube flag. -Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md). +2. **Start building the buffer $B$** in prevision of the next stages. +3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. +Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. ##### If loss is not recovering after increasing the pretraining data weight, one can switch to a LR centric approach: - - take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value. - - If still no luck, one should: - a. take an earlier not converged checkpoint - b. Continue training with the base dataset with (i) a cosine scheduler decaying to **LR_max/100** or (ii) cooldown to **LR_max/100**. (I would experiment with both if resources allow) - c. Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e $\alpha_0 > 0$ as you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) +If increasing the replay weight from the pretraining dataset does not lead to loss recovery, switch to a **LR-centric approach**. -My guess is that since the distribution shift is not too strong between stage I and stage II data, the naive approach or the second one might work and you will not need to experiment with the latter. +- Start from a **converged checkpoint** (i.e., *after* LR cooldown). +- Experiment with **rewarming the LR** to a higher value and continuing training. + +If this still does not lead to recovery, proceed as follows: + +a. Take an **earlier, not fully converged checkpoint**. +b. Continue training on the **base dataset only** using one of: + - (i) a cosine scheduler decaying to **LR_max / 100**, or + - (ii) a cooldown schedule down to **LR_max / 100**. + *(If resources allow, experiment with both.)* +c. Introduce the new dataset at **LR = LR_max / 5**. + When introducing the new dataset, use a **mixed dataset** with + \( \alpha_0 > 0 \); the new dataset should **not** be used exclusively. + +This procedure follows the recipe described in +[this work](https://arxiv.org/pdf/2407.07263v1). + +**Expectation.** +Given that the distribution shift between Stage I and Stage II data is relatively weak, +the naive continuation strategy or the replay-based strategy is likely to suffice, +and the more aggressive LR-centric procedure may not be necessary. #### Stage 2 to stage 3 (shift to math/code datasets) 1. You can try the naive approach but it might not work here. Then, From 87731fff19cfd4fe5b7b4478bcdc815c972ed5e0 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 14:00:02 -0600 Subject: [PATCH 534/590] Update cpt.md --- ALCF/notes/cpt.md | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index eec417ee566..9a2086b22e6 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -28,26 +28,26 @@ Note that you add data to the buffer B after the current step to be used for the #### Stage 1 to stage 2 (weak distribution shift) ##### Strategy 1: No replay **Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** -Just naively continue training with D_1, no replay data. -- Continue training using only the current dataset \( D_1 \) -- No replay from \( D_0 \) or buffer data +Just naively continue training with $D_1$, no replay data. +- Continue training using only the current dataset $D_1$ +- No replay from $D_0$ or buffer data This may be sufficient under weak distribution shift but there is potential risks of forgetting ##### Strategy 2: Replay from pretraining dataset **Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then, 1. ***Replay the pretraining dataset*** We mix data from: -- the pretraining dataset \( D_0 \), -- the current CPT dataset \( D_1 \). +- the pretraining dataset $D_0$, +- the current CPT dataset $D_1$. -No buffer data is used at this stage, \( \alpha_B=0 \). +No buffer data is used at this stage, $\alpha_B=0$. **Initial mixing weights** - Start conservatively: - - \( \alpha_0 = 0.05 \)–\(0.10 \) - - \( \alpha_D = 1 - \alpha_0 \) + - $\alpha_0 = 0.05$–$0.10$ + - $\alpha_D = 1 - \alpha_0$ -> In practice, \( \alpha_0 = 0.05 \) is often a safe starting point. +> In practice, $\alpha_0 = 0.05$ is often a safe starting point. > Increase up to 25–30% only if forgetting is observed. **Dataset construction** Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do @@ -137,7 +137,7 @@ if __name__ == '__main__': 3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. -##### If loss is not recovering after increasing the pretraining data weight, one can switch to a LR centric approach: +##### Strategy 3: If loss is not recovering after increasing the pretraining data weight, one can switch to a LR centric approach: If increasing the replay weight from the pretraining dataset does not lead to loss recovery, switch to a **LR-centric approach**. - Start from a **converged checkpoint** (i.e., *after* LR cooldown). @@ -152,10 +152,7 @@ b. Continue training on the **base dataset only** using one of: *(If resources allow, experiment with both.)* c. Introduce the new dataset at **LR = LR_max / 5**. When introducing the new dataset, use a **mixed dataset** with - \( \alpha_0 > 0 \); the new dataset should **not** be used exclusively. - -This procedure follows the recipe described in -[this work](https://arxiv.org/pdf/2407.07263v1). +$\alpha_0 > 0$; the new dataset should **not** be used exclusively. This procedure follows the recipe described in [this work](https://arxiv.org/pdf/2407.07263v1). Here I would give more weight to $D_0$ and vary $\alpha_0$ from 0.7 to 0.9. **Expectation.** Given that the distribution shift between Stage I and Stage II data is relatively weak, @@ -163,8 +160,12 @@ the naive continuation strategy or the replay-based strategy is likely to suffic and the more aggressive LR-centric procedure may not be necessary. #### Stage 2 to stage 3 (shift to math/code datasets) -1. You can try the naive approach but it might not work here. Then, -2. Try the second strategy above i.e mixing from the previous stage 1 training set (the one obtained after mixing) then follow the same steps. + + + + +1. You can try the naive approach but it might not work here, +2. Try the second strategy above i.e mixing from the final dataset used in stage 1 then follow the same steps. 3. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage 4. If all fail, do the following - take a checkpoint before convergence **i.e before cooldown** From 71d3ec96c632c8a870278305f6343ea9a5c3d5eb Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 14:30:06 -0600 Subject: [PATCH 535/590] Update cpt.md --- ALCF/notes/cpt.md | 77 ++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 63 insertions(+), 14 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 9a2086b22e6..0d19ad33b21 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -5,6 +5,7 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, - Model architecture - Sequence length - Optimizer (although it might be interesting testing how changing optimizers across stages affect training) +- All hyperparameters besides the LR - Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. @@ -160,23 +161,72 @@ the naive continuation strategy or the replay-based strategy is likely to suffic and the more aggressive LR-centric procedure may not be necessary. #### Stage 2 to stage 3 (shift to math/code datasets) +##### Naive strategy +You can try the naive approach but it might not work here, stop early if loss does not recover. +##### Strategy 2 +Mix in the final dataset used in Stage 1. +1. Construct a mixed dataset containing the final dataset used in stage 2 and $D_3$. +2. Follow the same procedure as in the previous mixing strategy. At this point, the model has seen 6T tokens and $D_3$ contains 1.5T. Here, give $D_3$ less weight. +##### Strategy 3 +If the loss is not recovering, sample from $D_0$, $D_2$ (not the final mix after stage 1), and the buffer $B$. +Start with the following candidate weights (some exploration may be required): + - **Mix A:** + - `D0`: 0.33 + - `D2`: 0.0.33 + - `B`: 0.34 + +- **Mix B:** + - `D0`: 0.05 + - `D2`: 0.48 + - `B`: 0.47 +- **Mix C:** (this is called IIDifying the dataset) + - `D0`: 0.00 + - `D2`: 0.10 + - `B`: 0.90 +Notes: +- Even a small weight on `D0` can help stabilize optimization. +- The buffer should contain representative or difficult samples from earlier stages. +- **Important:** Add samples from `D2` to the buffer at the end of this stage for use in the next training stage. +##### Strategy 4 — Reset + controlled reintroduction (if all else fails) +If all previous strategies fail, apply the following procedure: + +- Take a checkpoint **before convergence** (i.e., **before cooldown**). +- Continue training on the **base dataset** using one of the following: + - a cosine scheduler decaying to **`LR_max / 100`**, or + - a cooldown to **`LR_max / 100`**. + *(If resources allow, experiment with both.)* +- Introduce the new dataset at **`LR = LR_max / 5`**. +- When introducing the new dataset, **do not train on it exclusively**; always use a **mixed dataset**. + +This follows the general recipe described in +[https://arxiv.org/pdf/2407.07263v1](https://arxiv.org/pdf/2407.07263v1) + +##### Strategy 5 — Rewarm from a converged checkpoint (last resort) + +If Strategy 4 does not work: + +- Take a **converged checkpoint** (i.e., **after cooldown**). +- Experiment with: + - rewarming the learning rate to a different max value, and + - adjusting the data-mixing strategy by **increasing the weight of pretraining data**. -1. You can try the naive approach but it might not work here, -2. Try the second strategy above i.e mixing from the final dataset used in stage 1 then follow the same steps. -3. If loss is not recovering, use the buffer. Mix from $D_0$, the current dataset $D_2$ and the buffer B. You should try weights 0.05 for D_0, 0.48 for D_1, and 0.47 for B then 0, 0.1, 0.9. Some explorations might be needed here. Do not forget to add data from D_2 to the buffer for the next training stage -4. If all fail, do the following - - take a checkpoint before convergence **i.e before cooldown** - - Continue training with the base dataset with (a cosine scheduler decaying to **LR_max/100**) or (cooldown to **LR_max/100**). (I would experiment with both if resources allow) - - Introduce the new dataset at **LR=LR_max/5**. When introducing the new dataset, you use a mixed one i.e you should not exclusively use the new dataset. This is basically the recipe here [recipe](https://arxiv.org/pdf/2407.07263v1) - -4. **If that does not work**,take a converged checkpoint **i.e after cooling it down** and experiment with rewarming the LR to a different value and the data mixing strategy by increasing the pretraining data weight. #### Stage 3 to stage 4 (shift to reasoning tracex) -Try the same strategies as above. My guess here is you will (need the buffer and balance the weights across the 3 data sources) OR (need to do step 3) +At this point, we only have ~6% of training left and one should start the final decay. +1. Try +**Mix A:** + - `D0`: 0.33 + - `D_3`: 0.33 + - `B`: 0.34 +**Mix B:** + - `D0`: 0.5 + - `D_3`: 0.25 + - `B`: 0.25 +2. Cooldown/decay the LR to convergence ## Legacy agpt-7b checkpoints @@ -202,7 +252,6 @@ SAVE=path/to/where/you/want/to/save/checkpoints -## Worth exploring -Follow and implement the [recipe](https://arxiv.org/pdf/2407.07263v1) where the new dataset is incrementally introduced. This might be advantageous for example when the new dataset is QAs as opposed to pure text. Here: -1. If base model LR was decayed to 0, one might need to rewarm it before following the recipe -2. Constant/infinite LR schedule was used, one might experiment with the recipe as is. +## Things to keep in mind +- If new dataset is considerably smaller that previous ones, one need to put more weight on previous data. +- One can reduce/increase the batch size by a factor k but need to reduce/increase LR by a factor $\sqrt(k)$ or $k$. From b03749e84798f35d9ea72cb96deacc754ef57f6d Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 15:55:41 -0600 Subject: [PATCH 536/590] Update cpt.md --- ALCF/notes/cpt.md | 32 ++++++-------------------------- 1 file changed, 6 insertions(+), 26 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 0d19ad33b21..d0af666e291 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -138,27 +138,6 @@ if __name__ == '__main__': 3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. -##### Strategy 3: If loss is not recovering after increasing the pretraining data weight, one can switch to a LR centric approach: -If increasing the replay weight from the pretraining dataset does not lead to loss recovery, switch to a **LR-centric approach**. - -- Start from a **converged checkpoint** (i.e., *after* LR cooldown). -- Experiment with **rewarming the LR** to a higher value and continuing training. - -If this still does not lead to recovery, proceed as follows: - -a. Take an **earlier, not fully converged checkpoint**. -b. Continue training on the **base dataset only** using one of: - - (i) a cosine scheduler decaying to **LR_max / 100**, or - - (ii) a cooldown schedule down to **LR_max / 100**. - *(If resources allow, experiment with both.)* -c. Introduce the new dataset at **LR = LR_max / 5**. - When introducing the new dataset, use a **mixed dataset** with -$\alpha_0 > 0$; the new dataset should **not** be used exclusively. This procedure follows the recipe described in [this work](https://arxiv.org/pdf/2407.07263v1). Here I would give more weight to $D_0$ and vary $\alpha_0$ from 0.7 to 0.9. - -**Expectation.** -Given that the distribution shift between Stage I and Stage II data is relatively weak, -the naive continuation strategy or the replay-based strategy is likely to suffice, -and the more aggressive LR-centric procedure may not be necessary. #### Stage 2 to stage 3 (shift to math/code datasets) ##### Naive strategy @@ -199,7 +178,7 @@ If all previous strategies fail, apply the following procedure: - a cooldown to **`LR_max / 100`**. *(If resources allow, experiment with both.)* - Introduce the new dataset at **`LR = LR_max / 5`**. -- When introducing the new dataset, **do not train on it exclusively**; always use a **mixed dataset**. +- When introducing the new dataset, **do not train on it exclusively**; always use a **mixed dataset**. Here try $\alpha_0=0.8 - 0.6$ This follows the general recipe described in [https://arxiv.org/pdf/2407.07263v1](https://arxiv.org/pdf/2407.07263v1) @@ -217,16 +196,17 @@ If Strategy 4 does not work: #### Stage 3 to stage 4 (shift to reasoning tracex) At this point, we only have ~6% of training left and one should start the final decay. 1. Try -**Mix A:** + **Mix A:** - `D0`: 0.33 - `D_3`: 0.33 - `B`: 0.34 - -**Mix B:** + **Mix B:** - `D0`: 0.5 - `D_3`: 0.25 - `B`: 0.25 -2. Cooldown/decay the LR to convergence +2. Cooldown/decay the LR to convergence. + +### Summary ## Legacy agpt-7b checkpoints From 86d8ac6d3cc2843bbc80044e6d8243310c70620e Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 27 Jan 2026 16:31:00 -0600 Subject: [PATCH 537/590] chore: Update `tools/gen_cooldown/gen_cooldown_sweep.sh` --- tools/cooldown_generator/gen_cooldown_sweep.sh | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/tools/cooldown_generator/gen_cooldown_sweep.sh b/tools/cooldown_generator/gen_cooldown_sweep.sh index 0aaf5075cdc..7066f976954 100755 --- a/tools/cooldown_generator/gen_cooldown_sweep.sh +++ b/tools/cooldown_generator/gen_cooldown_sweep.sh @@ -60,7 +60,15 @@ while [[ $# -gt 0 ]]; do esac done -[[ -n "${COOL_R}" ]] || die "--cool-R (cooldown steps) is required" +# if [[ -n "${COOL_R}" ]] && [[ -n ${COOLDOWN_PCT} ]]; then +# die "--cool-R (cooldown steps) or --cooldown-percent required" +# fi + +# if [[ -n "${COOL_R}" && -n "${COOLDOWN_PCT}" ]]; then +# die "--cool-R (cooldown steps) or --cooldown-percent required" +# fi + +# [[ -n "${COOL_R}" ]] || die "--cool-R (cooldown steps) is required" [[ -n "${LOAD_PATH}" ]] || die "--load path is required" [[ -n "${PHASE1_LIST}" ]] || die "--phase1-list is required (IDs 1..4)" [[ -n "${PHASE2_LIST}" ]] || die "--phase2-list is required (IDs 5..7)" @@ -68,7 +76,7 @@ done mkdir -p "${EMIT_DIR}" # 1) Build the checkpoint table (1..7T + rollback) -${PYTHON} build_checkpoints_from_tokens.py \ +${PYTHON} tools/cooldown_generator/build_checkpoints_from_tokens.py \ --ttokens "${T_TOTAL}" \ --tokens-per-step "${TOKENS_PER_STEP}" \ --cooldown-percent "${COOLDOWN_PCT}" \ @@ -89,7 +97,7 @@ tail -n +2 "${EMIT_DIR}/checkpoints.tsv" | while IFS=$'\t' read -r id smod srb; # exact-T OUT_EX="${EMIT_DIR}/cooldown_id${id}_exact.sh" - ${PYTHON} make_cooldown_cmds.py \ + ${PYTHON} tools/cooldown_generator/make_cooldown_cmds.py \ --load "${LOAD_PATH}" \ --data-file-list "${DATA_LIST}" \ --train-script "${TRAIN_SCRIPT}" \ @@ -104,7 +112,7 @@ tail -n +2 "${EMIT_DIR}/checkpoints.tsv" | while IFS=$'\t' read -r id smod srb; # rollback (only if positive) if (( srb > 0 )); then OUT_RB="${EMIT_DIR}/cooldown_id${id}_rollback.sh" - ${PYTHON} make_cooldown_cmds.py \ + ${PYTHON} tools/cooldown_generator/make_cooldown_cmds.py \ --load "${LOAD_PATH}" \ --data-file-list "${DATA_LIST}" \ --train-script "${TRAIN_SCRIPT}" \ From 248cf22c029da95178f28518b2d8b0faacc272e6 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 27 Jan 2026 16:31:31 -0600 Subject: [PATCH 538/590] chore: Update `tools/cooldown_generator/make_cooldown_cmds.py` --- .../cooldown_generator/make_cooldown_cmds.py | 33 ++++++++++++++----- 1 file changed, 24 insertions(+), 9 deletions(-) diff --git a/tools/cooldown_generator/make_cooldown_cmds.py b/tools/cooldown_generator/make_cooldown_cmds.py index 79756595795..e7516a1e72e 100755 --- a/tools/cooldown_generator/make_cooldown_cmds.py +++ b/tools/cooldown_generator/make_cooldown_cmds.py @@ -73,7 +73,9 @@ def main(): p.add_argument("--min-lr", type=float, default=2e-5) p.add_argument("--no-override-ckpt-opt", action="store_true") p.add_argument("--extra-args", default="") - p.add_argument("--emit-sh", type=Path, default=None) + p.add_argument("--emit-sh", action="store_true", default=None) + p.add_argument("--split-by-id", action="store_true") + p.add_argument("--include-header", default=None, type=str) p.add_argument("--checkpoint-iters", "-S", type=int, nargs="+") p.add_argument("--cooldown-steps", "-R", type=int, nargs="+") @@ -96,9 +98,14 @@ def main(): for R in args.cooldown_steps] lines = [] - header = "# Auto-generated cooldown commands\nset -euo pipefail\n\n" - if args.emit_sh: - lines.append(header) + # header = "# Auto-generated cooldown commands\nset -euo pipefail\n\n" + if args.include_header: + if (hfp :=Path(args.include_header)).is_file(): + with hfp.open("r") as f: + lines.extend(f.readlines()) + else: + lines.extend("\n".join(args.include_header.split("\n"))) + # if args.emit_sh: for rec in records: cid, S, R = rec["id"], rec["S"], rec["R"] @@ -118,13 +125,21 @@ def main(): extra_args=args.extra_args.strip(), ) block = f"{tag}\n{cmd}\n" - print(block) if args.emit_sh: - lines.append(block + "\n") + outfile = f"cooldown_{cid}.sh" + with open(outfile, "w") as f: + f.write("".join(lines)) + f.writelines(block + "\n") + else: + print(block + "\n") + # lines.append(block + "\n") - if args.emit_sh: - args.emit_sh.write_text("\n".join(lines)) - print(f"# Wrote script to: {args.emit_sh}") + # if args.emit_sh: + # for rec in records: + # cid, S, R = rec["id"], rec["S"], rec["R"] + # + # args.emit_sh.write_text("\n".join(lines)) + # print(f"# Wrote script to: {args.emit_sh}") if __name__ == "__main__": main() From ab22dc17828100fd0091d8ec0b76b255e95c1fd5 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:44:36 -0600 Subject: [PATCH 539/590] Update cpt.md --- ALCF/notes/cpt.md | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index d0af666e291..3d3fb51a823 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -8,7 +8,7 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, - All hyperparameters besides the LR - Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) -In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. +In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. ## AuroraGPT V1 (Stages 1 to 4) For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that we are using. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. @@ -138,12 +138,12 @@ if __name__ == '__main__': 3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. - +At the end of this stage, we have ***$D^{CPT}_1$***. #### Stage 2 to stage 3 (shift to math/code datasets) ##### Naive strategy You can try the naive approach but it might not work here, stop early if loss does not recover. ##### Strategy 2 -Mix in the final dataset used in Stage 1. +Mix in the final dataset $D^{CPT}_1$ used in Stage 1. 1. Construct a mixed dataset containing the final dataset used in stage 2 and $D_3$. 2. Follow the same procedure as in the previous mixing strategy. At this point, the model has seen 6T tokens and $D_3$ contains 1.5T. Here, give $D_3$ less weight. ##### Strategy 3 @@ -174,9 +174,9 @@ If all previous strategies fail, apply the following procedure: - Take a checkpoint **before convergence** (i.e., **before cooldown**). - Continue training on the **base dataset** using one of the following: - - a cosine scheduler decaying to **`LR_max / 100`**, or - - a cooldown to **`LR_max / 100`**. - *(If resources allow, experiment with both.)* + - a cosine scheduler decaying to **`LR_3 = LR_max / N`**, or + - a cooldown to **`LR_max / N`**. + *(If resources allow, experiment with both. Try N=10, 50,100)* - Introduce the new dataset at **`LR = LR_max / 5`**. - When introducing the new dataset, **do not train on it exclusively**; always use a **mixed dataset**. Here try $\alpha_0=0.8 - 0.6$ @@ -192,9 +192,10 @@ If Strategy 4 does not work: - rewarming the learning rate to a different max value, and - adjusting the data-mixing strategy by **increasing the weight of pretraining data**. - +At the end of this stage, we have ***$D^{CPT}_2$***. #### Stage 3 to stage 4 (shift to reasoning tracex) At this point, we only have ~6% of training left and one should start the final decay. +***If we didn't use Strategy 4 above:*** 1. Try **Mix A:** - `D0`: 0.33 @@ -206,6 +207,8 @@ At this point, we only have ~6% of training left and one should start the final - `B`: 0.25 2. Cooldown/decay the LR to convergence. +***If we did use Strategy 4 above:*** +We should keep decaying with $D^{CPT}_2$ until $LR_3/100$ then introduce the new mix at $LR_3/5$ ### Summary From 747b7fe9075b22ee58d6f3daf843ced1ea849b9f Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:45:34 -0600 Subject: [PATCH 540/590] Create readme.md --- ALCF/notes/assets/cpt_images/readme.md | 1 + 1 file changed, 1 insertion(+) create mode 100644 ALCF/notes/assets/cpt_images/readme.md diff --git a/ALCF/notes/assets/cpt_images/readme.md b/ALCF/notes/assets/cpt_images/readme.md new file mode 100644 index 00000000000..5cc98683a5a --- /dev/null +++ b/ALCF/notes/assets/cpt_images/readme.md @@ -0,0 +1 @@ +Pictures used in cpt documentation file From c832c0424def1e6720c75abca44d7189fabf1cf8 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:46:09 -0600 Subject: [PATCH 541/590] Add files via upload --- .../assets/cpt_images/CPT_data_mixing.png | Bin 0 -> 169092 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/cpt_images/CPT_data_mixing.png diff --git a/ALCF/notes/assets/cpt_images/CPT_data_mixing.png b/ALCF/notes/assets/cpt_images/CPT_data_mixing.png new file mode 100644 index 0000000000000000000000000000000000000000..3258af13d46f2c5536ab52b2bac99408d267e11b GIT binary patch literal 169092 zcmeFZg;&(y_67`y3ZjIhq<{!W3re?igVHG_-Q6G|NJ=*%Al=>FEg>=FNHa7;zh}PQ z-&*f_*S+@-c-LKPU;)FL`FzgVXYXe}&$H)?qP)bDN5qeikdU58NxoA?Lb|VqgoK=h z_5l3F`J~Gn{GwqdDyk?YD*9Z}-p0ht(ijPeF4`%oS*BATyC>t*k4&*wTwg@7g&C?P zwYsaF727G_cSkbG85aCFrTOx+urlD=^7GA1AHRwAZY4*&_Y1G;>O!AYK~Ev?tO{Li z7b1fq7xse-B=Oj%tkF%FFt-+8C=-SR5M76US9ZPsJ#qM-Tc=BRQcS z?CK9Y-&z)RyhCPo%9S!G`>EB=AVPgn$e3L6zouxYL5}Yfz%m!ziEl>Qvu|u9%C9` zn>S2gt+|ZVrA*}HkY0hG(U6dX&5%&RPsrel7OZ>fmb2>k1qjU{eQ3W z?^ox4>+-*K`G23{|DB_MpJo1kyDpobP+%=rCA^oL*^2`Vr3>Rw#Xn*~sP8gy(O*cd zTDRUf*FxNP>RYd&ENSYfPz}^~;7K7=o@C!k(P9tJ#=V*pk3UJK$t^1$u;z^j!CQE4 z>E!)IQGJf3Px7J4sWz=Q(@&ng?z%Z|y>nXo${s;t({#9SzT^YDy@B~OMhaZgk!F|| zyT$a3QJ=VfG7cO^7Vcv&fdPepcd=o=245feAqW*k- zds^xf=n`pT|CO!BcFF5JZB6TD)aQ21dCrWMbFXSL_u}?A<8Bw?wvmP%S*PcHk#qGY z#bzO1%X$l3a5{%B0k6;HWX8f1&V}W4?@P1uPAtJXblTK0ugk4GjTgln$uHU*#B~`! z^)ixKuHQ&^FwO9)A#oUD6KxD}Zo9i#y1QQ=%9w7Jo3!wD-trsdvyK4u|ZakIGXCH{!0@Cu*-3Q2=XASmmuV5~YbZFNC2W|J*C3U3!RU(Dh z(0&TO|GJ_0S-a_UT1%KPP~QAtx!t;9D;MJiqUn}B%4feW=`#FgrmnQ*a=`2Cw`ukF z9c@BfqHIay?8<}HFqQ@_o8~u;%&Se~E6dQH1{7mVgWHP@7us~!k29|f&gX4f^~WVN zlsKA?1{kbQ#;;6^QYDht6|#8IJrBDH_lqh=9zB{ldiolvLr9ABr|by#Mk=fN$KH)( z3+CEROy1eip_;zXUJ4&RIhzH$Ub*w?Z-U4VdYD__AA>t?0tt10!Xwr*FcyVT8&Z&* zZVy&EJD;n#NzrC_+Sgv~HzrUxQ;UDKJ%@PMxTl_}fmmtE{F^U#j$W5`{jtAQ}Z6xG^gDBL4-(PG@%%S$jgf)lVr0A`;N5;-u+FN2Dcx-VY2joBDh36j-UEH`WiQRT|XS};@kDB z8;0YTXc&gSc`7)k_Qu*rVMyR=*9_uXi+@DpOA0aiCKq$@CxzwkNoed1_$3||9@l_& z!isFOk_>PEbK)WTz$4K7XuGgfHK(He0m)~0%XkLgVHfU)wbS!1d^K&Q7?#qYC^}2mG*wf8>$l91ib*rH?Y7+!UPi|ri53x#9F$6bK z9Y$p5Fzx0J1&#zTHTqVl`dhyLddRj`Tr)S2<^pkB!M5TNJV;D8F(KN(|C^aA5KB|T zM{jNp*?$21%h!%_b8g)E{vw`BM2e3{Uy&F^8k?D1w{I<6~$?Ou4+nRxCA zLX>f(ve67!buY?%55n$YMmF#B3B+CqOg5S%856j)AwM)qcdkZN;D+MW?$ylm*zj+b zG#yEEZvFVC8a8)u$TwjL{}VbJjk+nDz%bL)?KWI-eU#Apc*D@A>H#;X#a{9i{%@4 z$J`p;uj?5DefII$kk^0@rw8nNUklaLLpITcz2*J;Ip>w50z2d$yKKS!T-JoGH)r!J z>Yiz5f5YWcLg*&z_Hbu#O3Vyb9Wt&&X5UA-4|a|`r+LlPOS7sh&2{yQj5oXfNVN=8 z_4|Iin@uJZK{IBap9^%(AH^`jcFQu~kt+7QrfToNBN(_*nLBi!|0d7VZx9}46M2M{ ze?7;!NAFe}`mwD~Q(-gqVZNGJ&#NzdoT=^{V0l|_1+0Z*ai@#fRx z`o{VKp-b0)I1Q4??aEcf3$+ofbq69()~NU2)VAIf8w$TtnXL$E7r|h{BPd{n z{?WU;oQYyqLAo&P<{dG>so5?l42labXEwg?OdzO_^L%K`u$!JCfWl?LuKj)5mA&`f z%{dD_uGE#7)P7Ej? ztoCMUCU;#NbYd;uC)8eU!uqBU9jtnbSH@6W6+^{*;F~EGL}G3RE$4l)%sj zoapx+Vm?-FJm{r-6P6uq6Sz;SkBG72l0UXC0@%PH`XvMd!q(C%czYO0gVEhJ z1MxavT=5W-3WYMo>oA^u%xaogbERG)hS4q!XGaPU$E;qR-+E5KuEC|$YN7sab16TY zv*2YM$W65$h;nIs8?J7=0cySrmu7Eh8Z5RQyTwZ+`=}?Tgtj$gh63W8A+1GBarTuV zZH5gTkh7;kQWA7(Eftak_GmL~eJn7nx7TbR$j;>%r{8(qj7w02D(`+<&~R zKi5IbpBt1jmzGIgOYeRm)kP9c~1?uSSE_Ov3HIa&a!$ii(=K z%S3i@LS!cNk*RHzll%heO#T>3P6;~hH631{da!LNAZM_`i9dMpnRDav1D2)2fY(IpY!Be-gW}zQ+Hk9C|(Y!=7G#2`h*O;`7P9n8H!6C~XD52pvPp z^}e2zrnVk)BS~~;cvUrVoR?V;!|rp!;>VC+?^RQR9E>E8Xh8FvCy`wSTVv|R0{ z2w;@T)^yH={zGUONoL|+iT)Dv9%K7;dNR6p+rp_Yqs|x_WqhNC;Z}PZTvRm3RI9Ct zHA#Ndf{1POT>5#w>U0TFD`$a_=$mw}A?lOV4vsyLrj&Fw?yh(PQLHpu@6;*-~*>M!c(juwYH0uO;+FcRjIvCWY)tj>s65t=(HqaekqcbSo zCqdzLLMN?dqcp3S3|SXEyfv2D?=}AedCGP;-I4!sZvPy<7MZujT(|7M<%ISg><}IhWEmD->w{M&9Ke(QG(xT!e9G)A)0yADEjd2SCY%7Cz)8sMzI|9Gd z6BlwLA^8pVoE$DhX`Yp26D}@s!8xRl?zVyD_D}Z(N!tEel-;BB^wkkgOsQ8Jb zt3`QLR_D1`avz=uznf^PI8;r~%^_0UKejYnQJEB{FG#3zKuvImwvPao!9DDIp7wUY zyy>y;IHV51Fx+jkZ^M?)#$K|*t)yil=B1naoYr+choK32>5$TA+^IRcGVF>q4&C@_ zKiqPdB!zm|qc)2DI$&VOZ-04EvddkE@p}1O%`YX_DJ%W0q{K91W;c|A^E4U&H;687 zm==8gc$rjSrXzW?cM}}ErZMeP?dq)M(>Fjo(`g%bG}ZTs1b34Nx&z8#yuR2*-m|<= z332PyAc6a$WnE-`2Ax5HN8iipK&W-J(DQ`^W1p&Q;T8#&iC0^chm>JQVq3Lb{Bh=5 zY&qz_*mBa2gfa}sqC&>9L`B!q`_#YMlZUk;(5^Ox)-6+c*Hg3PU3k71&DQ0X{no{s z8Etm{;P0VVIA!k-CqZ1H)p7lj22IhmsXN(47tg4O<%f1|9yV0GF}93oYKb|TFB!^d zxf@>fIla}BZm7vtmSZa${9hi*(VbdW4#&6PLR>zfhu9A~J-{VeSiX75bXzoOL+E~v zv{es|Tqh>yO>tc7Qm%G2>NCDCapCGsms|DuQ+RF+ERi~c@vupSJGtXfF#$_>E*AgE zSXnjv@cg|LgoC;@8EWPB$%y1Q3Ti>!g(rb}DW9nUWsnwIkI{Qqww?=dQhpW8MN#W? z{n4G~p5|jboKG0o34#xrtb{@x{^=zmj=k*^F@wh}Q)p#=o4SS6mZzFNhkXxYev}9L z&7Lj=KiTY_Lpl|ood7;4Ym)at`@>1UYtZ_!G6g&TA8QX1>r<)Gi$mMcv@8K$@H zMfrM+OfE)H5{pKJhw+0Hj%AFun1>heboe(mK?_n-oS9D_IGU5x#e zYC$=_FZLT@9Re+LP5zAh2qSIwBkQ8yIvF9|#7+ z`MP@<3w2j{qn97*ywbi{qda;$qwSF6ZG0o77$rfQYAbS}q;uNwh&DM|Z}_OhmY3`) zG7cRQ@-d02h8w^tR<*0sMjU&YTyPD+!_nuRyGUVM7hCX%sX%nN>uJCsjd$IRkW9Y_ zu3g&b8@GWK6C(;YdaUxVJ$c0x3-*YP`ZxeS;_`DaTkU4i(km*DG?z)7yQy3GXTS(( zunR7x(f`)vTbz&`F6D}b8qI>eeJr<5=3e(VC&B7>LB6nKuCB{KxNFXh$xiDL$1jRk zwBMJS#g}RK)Z1lRCgphp1Igf@nn!B}`UR&L@xis8!KJy9*&%Cab_?-@_(d!3n> zkEh<%>wm7b6Tx@X7bC_rz$-;i``fDM#PV#;{ID=bDyhl%W5IL7LremwNSO!2Y!qo4 zKLLfxrZx)Wk--x*>6Jx!E}KQoY9Yw>t0qDSq4^XS^a&BcNZ37TwO1Eh>YeQo-AgX1 zy39QrY2(WL;d#_gJ1P2DpqnzjoiuWCP#N=F z;!ML^<3*gG6KuN&lN(m^iZ1o{YuiQ1O$D#0YSObx;Z&D#T)fiSC2x;@Un6Q`CD4x` z8t@);M@^p)fd;)&D!eG+c@+hIr`Y@ThNAKAFD_N?H0HLko}pWLM7M3GDSsk9DeB^d zH66eG;aaT9%sBNC`l(6Nz1pIJZ~vy?epqdsn(lAVd3w-HmuS%IRSwxSS7FbZTu3Yc zu7R;gf0Qoe+=YO7$zhaVLdkzdMr|WZtnKQQsejBcl3x|{ox&1i7WBk))gl*?7H)ZX zGD?E{IWN9|OKx0u@sw*)_`%rvMo{ccGs`r@3*GUu|CCRJOqHu0N^|e*Xj{e9VgOXBLLz= zkwb-`dW2L^dCCRfQ5XXL%z=G~HGubQ)==l-x%q``m!=S65gY!sh(f(|dX?^6>SnHj zKh5kPlG+L>`p+Tb1wjP!3v>0q6yNP8i`N2j=ew41+=o;>%R#m$cYxiZPij$v(AjV- zI`;#HIw^nmkPae9Ia-&j$rbwd= zwFg*BHyns3= zGKQoPjdhT!j?>i9Bu#6%Wq81_`by$?_Fkeep4Q?=rNQOoK>NX&qIKK&iqWHnHWZ9H z4CcD$LmZiyVQ#NU@-_P^1p1^guTD`H0DvMjQOR76U`j}oTOo{Qw4+xlwnXeJ5UIwv z{OSbTK8V4vR+1kW#ygB~_p%*F+)BFL6?tO6&5bs`kpK(c?O44OhWaSqlgY;MSpWK? zmYiV5tR{5}QAQEL^{X{u$_mi5qXwcqidl47TJ!>}5<&OH{h|pk#@itUhR|H;c1D50 z(F^>*NzvnBPCZ3ZFTkj6mAeY^&m!~EH6)^Z(PNL5hOeBCGzCzQ!s;6DNoO)nQ@1(F zF^;aW8uek<0?w}yAcg5C6aswjchz$Nt+_bDA_)bQVKCPv-4&>1L83Tje)8q_@k%X< z%7e0_J7b?=a4tH26D=3Y6vkm7UzNA2T++M_|MqrTi9j*pGEPM1sC>!$>b}p!V~Azd zxOfCC$*gb+Ma3%}Fd@r%$>Nx|--apP%ygYco-_2Pm_QSt`8W`QS_r{@d8w%pVJaS| zk4MyPZK`~dVN`e;VKf-8;Xx$UX!kPcw3T`chMzHrGC�$K1h!1e98>X1zZRbNDQC zfN$z|_9Q<0&-vm4$?S!C(mVwMa#c35Ijpwb!*cX4?^Mjtao@^u9th{%uiNrw4B$>Q zM)c@C=f!%Nui=!vibxo0Mdb%G;7D#$^9=pUdh}PT_>Hq!b&I`^Z`fLO!ara--+hzE zwJ+e%Z`vbL~B@4U(9C^TOizlC}*Bx!Lo9$S#+s9R({> zM^L7>Pi^`Y?TgPDVqA^Q`e}5{l3g)EVqBdmY8C`By`zI->WUmU<3tVcods9pTp8+O zjv>-74pmCmUUal%v?TNx;u~n8&+hN#eWlSYkJ+uB(M30=AEhqDyA;EJS{fq%0FyK^ z7s@O3MX!x^p_Abou8muAI`wccxjS8IVO%aukp>dO)*QiiVTKIeLp62DX+}q8uKXc=5Oque{l$u`JHNbC=-N^3;VLFl`>v7XUKJt1&&K3r1Te8z# z#k9MR;(TO0aRIOX$$o%7?}ctY_5?X-?pKTTCwH3dIr)2=uTK)Auwa9>u}x8}qFDSK zW`DHS#38ko-Bo}E%{59gjz@P}@rqDeix`wkZ>o4c>~IlbbN%_GdG*bqz~xg_g46wn z9c)GfOBsRo;^25cyZDfB{Pn!zM zklW}ABK@}h_Hu^-GK}>!oNDK1X@23&vW9zv`vS`AUIzGt!$R*B1j~xZD0wk~=_sYH z1$Ux~lsQ4|CfY0h+|}X~Mazae8PS=VL85gCDoU^h2B#>vuXig)Ush2V{)sy?BXqxC zgfI24M*SK@u;Wu|e1DRHbd>6E<(`Ci9HC z@e};gMl@FbI@WfDUWKMOaaT7pE_a#F2d2HB);%)5;~(!x*!jQ4Tk(GyJQ9pf92){{ zk8FD3E$yJVuFkh#2^zG>??qzy$|ajw0qqZ=#YVVC(!_K;a@?J4t;!z#R?Q!zSMv^X zE{;?#{U;oO%%BkI=EEqgVNOO|V_W6{$6x96s-xJb%Y#xtu^%aVXFslR*uA26qHNJ2 z%&VVL)uI6h#jtLqE`{o=Jl+6RNu`6TI9{-x*fWI6-IS#Rp>8BL4gbp@YG z6(U7*RZLU*D{*>X!DX)>FTI20ll6|mcaOrNV;}q4S;e0U$J@^~;u6u^m0wqzwr8gM z|KmUq5-s1W=)AYIm)k%y}LqiR`8av3#|o?`lv^=xaxBf zbm!czb&4B^B^=8TgJ8r>;d33V?F#yPLZ8SO!obeo1PX)N4(ZrVg++Nb6TIXlch z2W)EK>q=i$#zIE<9H!g!=0J@^=52iZEYcQA4(VU!N4b0-`7fsQAkcGB2qj?@`b$H` z5iG%LY$IG~{`;3>DO8RbcwA??N_vws%i`f#eC81;v=Up|e2&J` z+682<3LxApiU_N>M@9I#*kNBcbme4M24Po>hr|DPJ(3L5SOvQed+f6#Xk+E3lSP$R>w5EsK#JO&?>Ky6!w zyN^4V8aVph9(bkd8f9(wRJ*`>+Nz_WhjqE;7#HE^xDJOlO(5j6fA^r{L8jLWyZ!tq zT~I5FoS`Q_FTLSKtt`~A+8tx3s4`;DoZ;Wg;-DPsC7pxhB!VesN0b7K)N zV38MfJ@%k~fL64871tdFB$O%tqAt{dtg6o0kvl~_O3Dx1q`0bSt$?@g4E@6Rr+o<(R~8%3qR6DRhuGr1f{>}Zf7_?*aDv#< zqL6SOeUz|LuhS_t&1mjmPN60T-+~unnVsm5iCEjLDu>@@d6S!_+AV;lg~aFpNLEWL zGz=zZG{IpP+OZ6X9s5LxxC|PWf7W7(mQ{0A^yX72CCVwfCjgoeVl_QPhL1UKP9_`2 zB@zsryv)IKf05Gv_xo5e>JQJoY1%^j|KiK4iu0Aloe$OHB>Z>5@;8DuRe^KMdFR`Q z`oHYSzx_TZ4Pv8ab!PZ4dG>GrryG2Jk2MX5 zky-(bIGKNW6#mWIWg>3hiottP`@j5;7P{ocT(N>G6tEco_P8K+gaMqG|1HsfUm`!Y zId_-8OZ`Z4x%KX*v=OKg3xh1JyMWY8LHSeWqj8C_Y#?Z;1BnZG+KVp;f5qagVdUP+ zM12%aAdNR>-0Yg(G%5*~{&Ps${O&CRm&3j|g^pp=eJ|Mo)ks|UW_IKPa2hOvmnSiL zAGx+!u&Od6%YXXI8h}~5KEwO}?4#f6Q&*rlFMtkxtpzwuR!x+-cW{xn*Ua+LrkSdL zDQ98L=Dl-N>O{l{Q5+RWLnkPHD)QXjWL_uZS)EH5&^_pDOmJushL*j$wTR}g!p4Mt zA~-@@4ImfQ0rq9i3-Dcm11Yb+F(E&=teO3Y;WEV5TUXvg%rM{eH>ZRMwW4VQD}r@~ z*BN>b-Ho&)Az^ykJ6yT-8bGwTfU4$Fd?<%Q-th@CM^7&z9{Uh+^`AKXm}(a$5OTHI z+p<2TVAk&OH@qNuS$z>Z20F#g2w21t`iw_ZC~4A23^PDh8fw1WE;Pe=3HUjUpLrY& zhQYsSGJI~`G!ucw)wT!{^FCoKT+CgT_ivWiqqe<1?ES|FF!Yt0P+bzZA?o%9qxMrV z{3yKP?TCigY%buz6Z_7fCk}#l7o(GgbNvxQ^RcDi=`0UGFm*m`23CXugS18Vy~J2@ z_suL;F9U>mMa7r*<=Vg}wWnO5o-APU?qUy!Z`^?;biiP$QdEoo5NN;e_mJGyW4}Jn z1CX=#B92D|m$E!zIR~dRdW~1GRzW_?6B7lwoP} zy^7q2w$xWLH#7&Jn$!ni3C!v^|2C$iXVhl_(#7q|dApfx$J}GyLB`5DVA;uCJ1Tu_ zD(XwW_*g*h?QbLNbhkhKFO_nrP`qBSe(VZieL_}i1R|x9cjYrK0A|H6yyPg{2yQ9} zbzLckP)UI^q;I>VaerH$XY0pzQ{Ln2MPM`%<^*;bhd}Nn}WDqhu@c_#@ zN}BnU@h!6ME-(>I?*SPwA|LXIYD^vl-slE$_@X}sGZCKh;x?=*?Q58`4Np@TV?{`3 z!qj8-A;phYYE?~tqB8*1JTeR_59Jo%rx%6+t>(5MZWbyeimKS~-IInCI6SONo6jCy zUCu0`CjRLgx>sab=DPpp{`&eig?(1Z_5sUC{f;Npo7G_0-%`UoQt*W;Ldunz@? zjv}ax%csO;AWI%lsNwLmB)7kAb@gfBNo|L%gQDhe_k= z{x-N32Q91W56tJ}{0Z+#Bsp}~-=$*_BjwV8sH(@wH1J!Ed#=Y>Z@{eG;dq9{yIU5; zQdFiW{t7`_2DgFTrLdMY#R^jkggp0+bXs`AO61)6g!4o^>w`2R@agFMK+<4N!wDug;GocYNf4mxtja;;Rpq~#^K*IWu6(pvu% z#7rfW@Dmr};!2Wc(`wgk+z-^*Yr{{8vVUV#gi>g=3~X4azK@FlHsw*7mLUG@X?JZ%==gx#Wj?pQ$9q5ao0;EbM z$D?_FC;VDFc6R}{BRMYX`y1* zsS=w%a6lmaX}P_$F!lCHuCGyDU?1V?EN<8?z?4xds(-uasDhC4>ww3NrKpI<6K-1z zTvuEdi3MUx8(`ThUMg~M^8wvFP>e+})`WXZJUH(D`;yk%%d+o>fHK`9(DOVjTb{Z( zUxK{|sPwLkP5;}qvj6MCIUSbF?WSl8#Kn8BL^46Z&mv;r>S)t39PCFe_N;Z|o`{k; z69+p22Z|BuRMCX?BdP#6G+0>b!SUHJ}x`OwGS8yfe(=XMe~F-LyoEr@FbKr z9`=wdcHt_dPFW!~*cXlCU8bacqF*VQ%L3cF$5*zBlpi&Kp5(=ARMWN#oNPlX8Abn zs;Rh+cqrz3bs~}~f#CH)00}_Anx3YM@~EL=ZH)YD!uitN@3^&CFlSa$ezs04SL3zE z9jU@*fkAT_0rm9_jO`viCRzqWf!kHSRT%mZ&rXp+0b7aBj{|Eq&&3j3)QZ?njlN~Q z>s7*}*TmfRMnnq@;7%tUmjGy9UZvUb$l)B5*FmW}YezE%8et)c8cst(8&JnKf$5Pn z0je)^KB3&-kS0WZlP12i%eZZL`laC}{Hoj&;h4=<;JAgaZKy8SXtca85CrQHgsDe8 z@Oc4^Zlwk%QQsfEgEv#$XP(^8dLaI72OYKN&z|BD!Es+vTacl7Rfyd+tg zPn0M3jb8+)cqZ-Y`}#i~X>iDp`V%WTh?eP@Coobuxm3}yDNdz`jk1l4zb@=3`L;;m zm)^Zx&2FodqiWS$N8p$OM&2|e0Xc*xL_)$P{LM6)fKIV2oR5D!O4PrUpCG>DprdO$N|~p$2@4MI)2Rt5@`07qj0VOGm$;VcawBC6PNw^ z4W1FUH~AkJ{DNgSGr)JK zt3Oo}L00Y9LbzS&#C*;hm$=VL8@6B7j_~Yo;g!%C&H!s(S)_8}1IcP3!!H%2c3n98 zAkJK5dza@@tQ+Xni@u`)b0Dp<8XEjdPW0i)q(ZP6$q$)JMo9PLhesbnu^TsjsVMCA zZJ=No-^@y=RWPPFZ0EUIVN3>Q(Tw%#T805oFR6l_NZZH@D}lSQT+* z3E_P!2ybU~QP&N|g-E%yesX2MzR@-aEtvRMA{?I114AUs^F$BQcIba8`jEd0v^y-Jl zc79&a62f}B%DeGJ4fDtOAwf3fi10%)ya>0#TqA^qx+&uq2PQA%DLp9*H_^=jbpd7C zKVIDJxVZjU75OCi_{<3rFW3vSkR3u@dTwjMx&$@i+0inDyo9^tD%k`zha{ z*FXn2M)d9ZD1&Y%|6Zb6NsvYJ$97DbmKEHiZHYqMQeU}v5bD?)FrPb*5Yc*9c{N=( zV|O_p7ta9ZA5smv+!IUdj?LfzQJ}j9euNV=`ng|Z?e$IG! zkKgPwAM^moZmF&`mhLRhmsG*Ux^xNSE$oC`7h)Kl{KRpc669`1!2RVRb^Hha*O!d6 z=-aaZGs+SBL8gS~!8eSb74;CN6-s&X*NLB-MYtAb7zQSrK1uv7*{7(t7o$E8++-VP zNB?ZPJTO1~Lc9m#P&Q34ty0bi;X0|_98_e{I;cDrxhrT|GTb29!o@kq@_~2Ko!%|o zp4GNZGPp6-|A-b)YTPyyEM#3j_q~THd0d#bi&kaT&U7b?@b-rYsXf&<&%b!xQ%m1| z-3#oK{MNugCRr@T-3#M#HO@XJ+t$B$M$sGjd{Q$7{p;%MxL=&!(#!w|pZe3p!PGTJ#pzP3T{EY85-@W0)QJE z=&XU{4DR2Q`X*f|Q{>}2`+8+qnt!eQ3A~R{{8fYR1klFyh-m~j8VgNRE1(m%H^iUN z1XN}gae%b_F4G0=*Ea5(xh1n7Bi6BWm6%CcA5WjnjJGCR)%;H7Ohm7|UVZbbAV(`- z;eEcq{?43uL1bYxYrM?I#GKyk_HFVf{(1bVg&Cumn+8P z!7*s?6g;`$s|?&@ZCJ-&LR(w7#`>OjJ^BLotiI`|%~%Kr2q>#lFTK@#hQmI!N?j5| z9x_nyhsMjQ@0aX*TJR1HK}ff<8Y`gJX2G010pKlv*=PRIY`^{N0&(VxVV>pbKx9!* zw;B5_s%NQh+;V6Utkt3vIHo`UB8co0_1)T-nAp=diY;#^SwgSgg+p4$T?-;fXr?p{ z=l&eyQv`949{Lzi=$dA!SYOE?Q5Y6&*U#vBoO06Zu%HA|E$3CC`uOy5U7z46KEPIV zUmg8LAiI)CuUCHiBGGsQpc@?TdBIf;QeIWFu(*Uk0Gv2><9TCxiQ|2jvdt++vb)<^ zVwiG`dli#=R)D}A?(?<^i$Xm8&`)cBz9wF8F6gP+k5JTfu>a2kqC4;4y@IcL#)WY< z-=P3oZ>xXx%a($RPE+%RtN|@d>O8~xSqg4#W~_WcG;d;~*|}D@H|b6oX|9eqzXK42 zv?>f3o@x72Pg1mAPL=wTo|SxgEvNRyi2#}4I*ZxlEYcs%5!X-C>$~_BZ}_j@Q;FSP zRIAI8X#$G%x9HUPmz3Y+-fDFJAl}1Ox|=VOc-pbyuIInK)}pwx^yV=Sb#O5Z()#O_ z-r`#*1ItSmHVV{9HGKC-B!=5@4aMx1^K`G;1fr~A$9MqJ_!tGm+@6``d|v{{*WW&x zzFa-$=m~b@0jM$!fUBHmsu~k9h$e`A09>QqQPD-J!{(Wcx`7jnC9|E~w@YAF0=n6k zRYeA%R~=`oMid|}xSDHi+wT*i5~476eV{+kX#t3G|7eg!K>>@5hl49;32@~dZ2J<_ zf%m&@Un=CK@m1b|Qp|(*yw9b@sc7sY)OW8pJPBcJTlkh=RU|ut#u4@h!KSQu1Fc+V zR`r!fveX`jtnWs*!*VH3j&~cD%J3YCqy-H@iQO7^Xeo_!SpI<}uvwB}W< zlagsVBSj#0d86;4nAlj>E;?_~R1t1ED8fEG3P1T$7a{f;J($c{eG_+l6k!SJWP_b?UFh zSl7`w{9OZhyOmcZlsj!wM{qIu4%)gRO9BnOlf6PjJqcDd5t<+A61F5EuJ;!shmEJX-{ zufg;fZA*4W8B*A}O>8X7Z3UpygYB_L8B%eV5~Z#=Qp-m95p2-ss=itqwELzfj4uL2 zU2OJ}U3Zp0ciXCO4lxkKJb63&oRs#i#V)o8v_)TFu`hLzaY%nTJV`BSKAYnxz4^Tb zyr8j-vENdUcpqd{9CyC>sQ&3Dy`hZ}UaY_dFk% z@U7P1-hJ5OUu0;2ZB$N?`Y7yq1#1@Uc`RRnyS*z2n2JP z#JU1Tv=ojbErlTN_wB7;CwiP6`D=M2u{SY5F)Ta80I4u`w&sytobw!I?*(GTPcPqv zpScj{R}R*H*v7@_B?ua>5hY9zgr&0uT6IA9pyb|D8>KjL zC|Br*iFK}l_lekCCNsb97YS__d*a#@n*^iq4swPDA64Ywq%1@qO$7!8H53hPaW74sY_q+n zm@SRH7@-y|K$&2wgBJg2f}Hj)FVW9WsUl!UgFjM>JLB>^aYlfhzCk`bvDP_6A@8c? zsq8tShEcR*>ojJ`12_YgiY!~;T50sM#GMF-jU5zTh;OSrnsMJ|&!W*F?2Z)0OtXQN zUXTD2?o(%ZK+tf;Zn#dK_W)N$)Q_QbGX{_hVAhCvTjGcSNmP z0LRJFjh|4D5T>=16D@#)@wFS{p-0#ZyK0rs|9ut!SN^g}?7i@_FCkj#3RAs`TSEum z9#nN-M5;-52gm)@$^QfO1j!wD7-RQX5gURlz&JTV)V%1LF5g`= z-2>EJG7Cc}&ESkDd5JPUVAsF40RndOWs*@0_U}^J#aXe?DSXsjw=4WSo!xH*43F|e zLWCaTr~^h>)cP5AJy0)r={5$;eb7=t{TN}XAc|4V>vzPtuJ#2V0t=@?!tIW>V zdRNkEdju%JTv=@xO?k`FFu|ly%|6(Umjb*!eHU@%2m+sw#~<+V(uBd=L(23s)oAc3 z$wdSyjUZaE!@X2)e>Ie(jFh8cog%9CK7p>QK`#|D;GLOKz-&a=6`&Hf1#aPnoR+g? z2iBk^EZQ50*iT_ce#|n$cPtC4{1tbRUV^4MeeI~g)w<`*Dvg2`9E|cdYXRdHS%B}O zW-0xG{M<)jWfl5nO}=I}6N>rsa!AcXK&B-$q2U94UzQgiAed`-wHkcnQPq9WyLZ6Q z%=T=a9%7(j!?AYt<)A-67&XbF4_SfUvk@~s?zu+WPh(8;3IPYn-Yo;30prOMXx};3 ziL-c*huC>llT|mFQNYy|dw?EjF)zcL_)*a6Apl`^huc|;A3ny3qw)kG9VHj2&pJea7;z<_2Q zXu7Euo*3@}Bg>y3zE(Kb1ds=USAew)1>?@=9Xe(%{6K)V4yDiiWWemcp{m?+Hd3i7 z1;TtAyKM$Am6mHIfFoo*Qlo8rUI!{T!hmfiWUPuNg=j552i2NnTy=H3v2%BDBUkWl^g(O}wj>Bf0cj^}V> zu`+PgY$~ESUY6GXAwIv{DdjDCk%CTV3wjjQUP-F>Nlb^`eW*7uLTL62%mbhI6ZeEcQMb<+!0x`zbP8_f_{25Gn4Tcl3QEB^I6oP@f zmZU-a+6R7s<)s=T)fXlxilt3+oFFb|$Pn>nUnGCl0*KG|5kIp3E~FLh0T}IOAj&C& z6L_VXbb^3}st*wb1RKX|i0^+$zOPOruqcn2mPl}HP^f2}@s(=Xt20uEA)IRwg#53WNUPN( zEr-gJj>K$41b0Uak@Fo7ttk?`GLk?_z?K#4Hbx4=OD;^`-Or3gq-2oz{n|hShDQMV zh?6(WxwL>#+7=ayI%54asKG$DQr{?!1nCN3Q13d{<(}os>Yoe*J5@bJ9)jVJGY%bt z2#UR4;f;Fm59e=zexNg_X945eUQ{o`=w>hgNblRK)BATf0v+!Y`Ojvlf84XEwBB*) zkVYg#N)+|1k9NW-!AZ!66um6~Jk>NNp|wBu*-tn3fetQ9L|up8@&J zc#1yj`Y818XQGT@pl)Jh@q*+fwE*OyNenS-vuA+Q>twwbm@;uDERmyViBjQdwS2*P z3I~(b!6X2ZE7o}f1tn|E?}AU70_EjA8UX=DBH+pv0J)kUH$kF5pRcs(T0zAQLm`vH zXu{>9+0fB1j3Pt6esTjQ_gy-Jv9K_lScg$=QFAqjTo}^{v@C&Db9-t&NxqXqgCL~c zt)5_KVE>|*K5O|Imu;q@IomCYk~_l*0WzL1$Lm(m6+V6Tm2tVzw7m%oTDI0^u4`*< z^WzoUv|N}R^bWj#x6T?yJ5KPiohPxQ z304PR_0&#$sPaOzZme;4z|;F=?|FJawGToF;7mKuYVM;7`>f%5A~z2dmq`k0n~H%U z7&@2@wgE#ytj}A(xOfBT2>Dx6=%L=YK$JQCrXYZBGU1p>guaqPuc<#(J8#`M%n#IU z;jK_7hT8{b3+#JNz?S(qO(4P$ebDYzBA7gDflOrW^{K}G4M4R~RQLcI z)bX>PZ2Fr*7{W9rd=Ywd|F=az)-H8GzDE*qGx+R?KtZSL)!M3hFQZp!h%ueTUf%nQp8Op%mt)M2yKU#g9>EJxu1Dx|e;#)%!xEn17OP_3#jeeH zAciw9a^;%l%nPR?iYj$vmcekMJz{2Plb00OssmXz!1`_nVUlelWKey}wps}<*^sMu z@Y&wy2B5gOAjUMK`iZf(=GVkl2CBT~giiu~CPWARsy%T~^h`k^?Wqz}({cK69xQlG zNoS~4{F#j7lOk(kw6OG2x(mBnx%vs9KncHM9hCCwgb|kg*bCOGL%#PQ_HoDrZHHpd z^~90VA?2hmP$ zt|COwPqNx>(73G0$e1)kIzuHBeVXu6g;(#cy_65h+}5LoDYcS5N2pRGsa(%2SK1(G z+%#9lF%9&TAjk(b9jLl&V9$g2T=kysiGG4)Y_ZD-mzx&&P=pTlLgx&|h)MJucAjC5 z$-Bfk4LBr9!*1wpwX zPTKbeErNKyMA?}qYx8--!@hK?2Xrh%Cz&NZROzr7f!9`V=Xy`V02ndV=y-eRyR|m9 z@tOC?p9jD2A+YswCokB0SI%o!^qdg1L53H0YLdLQ>I&2C*?oDN zWWWt{rqBaks;wOPbBQucG_a1^+-0|dl)e9{}w);;A;M&-Xvp~nGW1)SaW+;OK_jhXH?xKH1$ zct~nEjSaC~ah;oZm}9*Oz_`h3?_o8Qm0NX1z+Vl+TrH$Rk>i^K1*&fBiF?BVP=?M1 z@>hIYiS)bRQ*66c5DOOjI-fZZn3jn((|Jdp-T({_aq-jX`h@+ld*$3ZzxL= z&B_78?V=|?D6H{JuJxmpTVrdy$T?az0_l-?6a7jT>vo7zed+d7HQ67|OZ) zoz`^#$RoxT3WYl&OV)dd=6Hbz)hw9neE^OkIxu>O*TD8fw2E~%2upIKOHjo-Lt+l2 zVpCK=@@TXuxYCquV(Nyk!?d7NICQ2pNJ}yr{QCPH>=iTTZ3q>rv+0Vz=iU5Jo|qEV z99hQuz1!9rk#iokfO62P?nUTriog`qCzi`*Xg3RLHhCNP&OcAHjnC7v`+RW|p{Jen z1@`V$42gGJ8ZB~)E1-QR-6~}KJ?}LF+X>`^GskS(T_)mGXf@h$C$h2@k8LIfElbW% z^_>h>T66SAObyLebaH4ACBCPO_mI9kD6ATIxsf8|i~e$>1%D@HMPTJsg1N=gist~F zh|b^tiZH?Yy(I(%?OOO36}_;Bzc*vVR_;T}Gu|9s*9bz`NyiJcKS~nOag0T`Om!hJ zCts~~#FF>jkxj&sr}&cG--IN(f)qW6u$H%~lkadZ$#1ARC^+g=&X5e5b-49%rea^6 ziYE(AYynE)s`sPwgv`71m1(9oiH*hn(E&KEhA?OLTRW~G6Y8}_#t1SQ@ZlAs~Vp?P? z7k^%;?n3a_|0OQ_6VN^FfyTm@uny0lrU5&?kIP8IH>p;=#4u^bK~|cGVw-c+ zv+}gA!Ee;I&*=}Cg|IQ6z$sM0YPU+5^N*hzif;k$ll$r>Z0N<^g9r<38%3KW3yxQV zb}oK!KK@BjY`x=o^1f$}M91V_F83f^?l*HZc34-GiVK@mS#7}&id!~Ex6``&+d2p) zf)`n^FmNZxfC1GWtkl-%#L>i$X<7txeq76x`gs+F(=^U9RlzD;-yQgi%udXSez8Xm zp87c;3Zc$Aeo@c0qB zo$&b(sl!r_C5Sp#3f(V#!cJ!+rd!wCY z3bUuE+~j@f9qWJKE%GXCD~$S9im9+X#BX6k>m3m#2LJZ5WB9QH`^J9om53VyNgn7< zmD!G*ADbl9$S*)GaW}~oluIsx;?21LgH}P)A&9@iEwY3G&wwO-fw6>5$Ah%T^`(1l z9Vl)>`{`pRN6^^hPTV@}ITi*YA0Og=V^{9{32b=NiNmnSvUauPB9+B1zU~T*0GHmV z!f0(DPQ2W)Gz1HAkap;3{GY%uzs^y5Ol(2edIXeH^siDxLAmg`fJ(%sSz1SW^k2KA z;G9$Q-W-&o;v(_9J7OYPfS2U51_B+$jG|6FiacYfONL4A;}s>v_yk;_9NIs@b2vSr zYIX2@2^Fij@LQGhikaw#NLLJQ?zwuyv5j>Gcuo!3=Zfy6%_hM2tS@E)w<5HCSAKT& z@f*us(6mY$3QiV35;e(FfK|Qt@_&RZ$t3;Og5mCnM3Ho3e{8QeJl&H@)h7xoi#|wo z*jldmYg_bdw%HJ%`95t@1yXJN5$R|N8I>Rky_~e^?y&F@Jr!p0#@*wC!*DY0m1)WJ zMkkRPW5}oI1IgCEa5=-QvZ3je8(CdvvDw(@LtL(Zeh+ZYyAD>d-cV1GP((^ovHp%; zC$TSkuBXy>P7Ztr@bAfuCKdTHiQ?Q>lr2;6K%9xq(LT&=E|*w$jtyz=k8Dqn%?1gp zwqOxIc+;8{tmUM3rW;_s$zm|DUgK#!04S{R64Xh)BL0k5gspEbd9x{}EhC zLxB7z-Ko^=FWvnc5dl+Ozx1lLz4OPoQ!6SQ#Jm}Al_u**C%KA#sDvY&IH)yhg5SO+ zB!2@Yr=#D=I&X2TH+;$#j>(0DjgE6-NGRSsm&aCPO3NYF>%B_Mc!}u`$}FFY9ypE_ z4jmQI?X+S81+oaO&wux~wKm3?r#im4*rEo;{FyWokL%pi@z?&`Z@9&&K6Qb@v}xgr zOTzix;sfl$C0|F6BI{(WP+JK;by5=ys)X>qwtq5xZoD9!Q+xly^R&cIhHV^o7jnluOu!l zE)BJp{jR9gGwXEe1Au<#(45(52f3OP~t+Y|jkUJqygP@=^O7LrRPr z&|drrb(CRFFoMp>htWP&>48xEhj0lVF{~%q%~ddn5q!ouc8LSgEAt_3pMbkzH`x*m>o!@QZ13)&&gx#wDk|NC0)%VAyBNy-9!(nH>%**AVE@@PD+ zsN2XNiJIdU;EE1T&LiNBW|W~BmP%;LvVKZpZNAk%3e`$Yt|f~TRZnKcR&z2V*hRH#b`;Q;`_7UPAl2Nm~HJ4xKbYic!eDO=w}&P0It;`HKh>a$cHr|D2jvITT5 zOjkX(F-r>>)jnf+VXM#Cf$Zx@51I)umB^{?9Zqv^p~D3?ZRCY@VcV4t1XcZzH!!tT z{*GAYrZ%42#;NF&IwM2{6AkTwOd6w(iX;-!8AiLK>OPnQ1K-eEFpWfTIVy4KO0thU zdloqAYC&Y95N6JYb9@Sbqst;dvrnhu+>EZR$ly`v&S?qjo@efv%s?=1?EPY;3=*;8 zUBB55s_)eDn>5e@D1CuTQpnvfgf$7Itw#k=ak&V+{LvvUqLgi)`UiayF&veyl*jO@ zWHE}-&<;RI#C>S2YrHtXk8^yG+LJ-Td`G|r(ay~joYRXq%Qp6okf(iEth*e1+`Pxx z)BVj9Ynjp)lI$&C+V}m1PNPJ!Bj(-vHg1I{g5sW-9YzDo%sVC74c7P`4))kPVxjZN zG5poil7x6AS3y{@aQy7Qn&0SU$?c^!KCHx>rs(sXWBtN^&0{baNO3E)+90FKGMM_j z;|tJ{U%Wy^&e2(?O@}YfiAo1lWhs*G{Nq(9R|k}_K5fPJux8>_)ya8*x$Xgi?(v(( z>biWZ1*We-*kUoSld20st5s@`=H9wHHBise)|qj)ub;<$_BBWKr-c5xtQUUO9p?R6 z)NBtcjZ9cE{4<>&`P$-Tgv9%mU5Ab0#`h}^pU&ZPC5RrAM)9+QKi?prc{JYwl8 z&hL{|pFp!JAiw12J&ocOkv(-&bF|c<2 zqM{wMh3j{#>X3g5=n2y#4|QoWjAcEVWV{IZ7)C{aU9&=_c=hP>6}a~sRA-Ufb-;$@ zaN%9y_;j?hygQJ}M!l}ji6cEQ68n#XC!u1B zlZusY&-J7mRBzgYU4z+_`hOSoA)}^FfHZj@n7(qq+!H zNPqkXLq`)yxgG!)!7uZ(vS}xsleXWp^V>TjowAS51$ku5YSM6#OWcZtZCW)mH5#Tp zS(g8?4fb}qMlRq%Sk^pE1u;hP%AB*U4Bb&VY4 zTVJp^Pi?4{Rv75G=1lT+K?VfS-x#B~5&Y$uY1_%BxgsNvp@qDsiudJk3){nAAs&>h z+Do`^yhYx>Y*T~8WRxpzSUtaWxE?ajAY0rB*O~6M784D$T9Q;;otLN^?pn4%Ec2~* z-YRB#y|um@SEZq)RdK#kF?UP9C}sG-X@f`LEhhgz!$jg{JY|y)3Yk2I3m!F@kD%=@ z;wvn5DSp>sZYyWmiS(!;dv{qM2RaHG|(FO$Ksind?-4wnE|jI31H8T-mXgac z1$iST-l4NW#8rwkGT>T(Oe)fQxNtNAG;a!%f83{Er zXACZ%)Tr!suFO!B_K>^tPqYMe-1438l}*^vg2d|p)e|X@n-$%euT6*RM`x5o`L1|V zAHjvkU`<(NGYfra-n+mxv|~nK)1B)6lifjRC_+s5Z-*oc#rx-hE(Yyy#0B#^e+q!M zu-Lrz@YYj6{xY3yKo$*dk}}U83CKCn(!8sqiD)*BK*@F5*3aDxE%<#8Tc3(x96Ymx&BIdDZ5P!x+SU zK7iaGiv!_;oL3@u zQ4BMBXfzhQ?26yRo=hIIs6mp= zRiYu^Btxm?lRx$P2 zokd^zo+5KASW){>b`r%&P(u2%fsgmhW&}XOX?TiK2UyH`l?(xi9}-J^3i_}u5@mRR zxrCi-t=1m`6Y*}M1M0y&I;~qfVB=b74FB%Wk35MO>)s823Ays`)nCl)KNTg9)0h?w5ms@Qwo=POjjSv!C<6r59PA9T-u$eB})MYfKPMfzCXgCCeS@v-K24yT5|x zA3(#C3`96|(omP$Pugw#LDzc6z-C)g7ae}`Eer+|s1mtDEL7Dg>(ttR_zG32o3P{R z#Ycl-Lx3RsH_1qO-TP+#2)@75`q39(HSyQ^h!IXQ6bqAxO6wCXGJ%dH3?lQ(ngurQ zQ&+-CDmWT7>m$;8n*%Y?QTJPW)bTHCk3F*F+!=yNmX;W~ym)srcYUZ@3rqx|FdF+n z)KE!_6_)jb;MFsx@G0=}edq@XCSU|jvR_8jA%3#H7Ws-&-UVqk!@=-%WZG4I_+Od1 zq~B?}HkOv$1iS6GpXNY+o;>1x=r)jBU8O45QT;FfYU5KbZ=vGg~Z4X z#3iAsZL}P5)78?}f31#NNq*o0;8kbbBln@~V{4#?NW(%NrxKilKZKxk7IhOh-;$$!ug7Vz z?x|u86z9ZF1fmx;CL-AK%D2PV7*lRwcCfWX-XMXFj8*lSBvCGmz}1*!ccKOXWXhDO z5F!W^PO}DV``=Tt6XoSN1RBb`lSU{?B5Z;WhC6$!4n*HS0zw*Mgr7Z|@YjSspQvF^ zq!o<+bM?E%802UdbcpS2aFeAY-SCvKmj`Dm(T8gaCNhn_Q%y8_mh?f<(trDwM&aev zk;+t4&GNi@V$M2`70v(bbX^P`4G1}I%{iLzA?eGtCXp!3o-44`H z4{d{2=sOw%uBqWwhc{S$>$y+fd?ho!oze{JvJ1i-LnhTb zm536rLP@{cA5WU&rT=}%)VZg1I{X_gcu)dSW1s>Pp~bff(CC#MS#-Slbe)ZI2DfU> zg}^cM)6JmdZ>&zPn3O7(IUI<^l@Q7CIu7IbxbLDBMHCWwGRN9A--4s@7< z7^U?SdA(nE@$(m8yS}5dNmOeMoid7K;z1;P4ZvXgp6;N}rpC_H9@`!y5r{;nb3e#0 z3;tiUI>M76%9MXM)RmlS&}`h#A@K%x@SzRE<(yoV`>jF9=qtVMLckr)mH&@xntX{rb?h6zL&Xw5dF7?I@=-~?P1)bUSxAY5z!>#8A zVxhV31?@*khG3vxPs=f85wwN2aZfmBDXsxY0_WkAj`;HaLCu8@46pH*WwQ($UVI2* z6~(bUf%cuQ`8g{PcTxKDJyW1O~8SCJx?W&OEh!%pA|G654Q$9wF z$P&CJjJ|ycWS2-5eq50aBceuFo>Ou%ef1t*W8J;`wut%=kj@ZMIw|u{oqt|q zyX@l@vk5MI(Yyfe&mOdb|L!%jYiK%9=w_OL;um$Zorn3Y@<;W~n(+M!r(slHGz+`L zN7f(jQN4z^FO((bcu0t(K3zN|z zs>X3N`mTBg@6vnIq_{jci7&_Bs)A96|OGsdxBVB@VN+W@kojAJ9YF zCVce{ZtA zv`lsL+z_laP_nq#zE%G|Q0@p8XS15sHm88*nArK29O&R$WW&Idy<4)#){UzKp$R&( z>*{^sjLHn=bgv`_lMOBExI-7lBlot#D~L*OvLx>ZBZbX~xEz12Y-n9O40;gFFw(LI z?md-?R?410FN0Ou7GxK-k_rD;k?|k{qaCL$J48Ix+E*m}Fa`a6^q>p4D1&q0r z;E2ZmEUPh5!q$E%S{=Hi+^SG}#Ghh3z!up>%kwIA$s`ngv~r$?TJLcKg_Zxo1nS-?o>Yx; z6^(LD$92$RqxI5tTz!8U&ZCENhX+^v56cCz(aL@g_?rlD5f#Q>0n+qsUfM0WUZ6US zXnS38P@Vv0$TaPAnwK z8Y==KD=POqmevwZPmEnH;W$t9~LMUiw|Eb}y>&>I3Mjo#t&@3Z_a|VRoVE#XoB*24`E@;x}U2)5Ye0Kv~7Y zXDBKf0`$CEbrDEd*8o2bx8!!2Xg$LlU`g3pqC4pvzf$?;5YdRvd{`|{s=#jml94w3 z45Ju1Z2(@Q!D!M_=-$Kqk4WEQKgzlGGhPW};D8hE7H1YR+)tAq1E7sBSdw`nu(KRQ z7N@$qqtTX+bpAlRPnt=p;|*ubr-ozpS1O75caWr@40K-PiAEnT)p5QqhK-gQep6yx zUA6aC$l?KRS=*l7Sc)P(#g{Cffu~BFkcCL)=#tnK=cv`D$+k7;yJ*c9#*s!;h6We% z&R$y~sL19s+r?OgCBbzX|6t2vw&RoV9EYn|SG53ECNQjGkiE>!%sx06E}3O3ZP$3U zSft>Ni436xQD=2#;*@2{IVQPHg||W2EShDX#u}Vw>9? zH6iV7xNL}uNiwqB9|f>kfhFwivS~8=wqw%j`pq9g^;fvT8S61A5*&csEt32HNY-j= z4vmq6-M%fa=hQ4*Iv=8=OY>{u_mH<=R9_ZUErCY#GzFdEs1crbta+FtfIKxg_C(zt z>)6!N;jkK}h0$(Xxmaf?ma-;RX;>hrE#d4oKI1YH8K1Z58wORNzP~VjWXLt{{8l;4 z{COoJIrLA!@9S2pEGx~lqxqCoF;U8P?Co>Geg3Co(w;|QF(D=Q4cjLD?MHye?y7*2 zJ1X1)2k%htN6~wuQD`=1`@JOlyw9JKp_krO9o29?fP?v^o#(GLbN~#n@bGuxro(9` zJHI7g&0RPbP!hU1_GE8&^=c(sp~zNsUku}3bgZ`VD$kT-_4Ng~y>%C+8wD)rXwQuz zA9S=tcTwrwb!O09_0(V(r}G*e{rk)y_$l(Bu=EmOYQ5=jd;_cX8mCw&Y8{VP()D64 zaj6c8NYxCe7I~`8&D+;;PN8}_^{)VXSln3S6F1%B(v=-&Gr;rs-_w#$UAM3zgjW>E50vb0Smb>nhuQEYgeI(s)2qHKm=WPySI&49QkmO ztLM`L?bp5U&|w_FNmUnREw~O77-^%3d6deux~=67B)IF&(8g*QwLK*3wRxYNkd(6N zGE#|->NMAW_o?U6p6fS7%;whLm>kW-Dj*)R_3_1LFtzO=np zt?1e?Yp@-i@o}tdt?sO*4ZSe$v~AOVM*yUF{?i50WPx;Nv?>0r>x`3L;WLA*gv&rt z{W1`#r&{^wY0w&YLw?*e^R;++_oyijd$G~x{?^33{_>_}T9@deHqPzcyBUip_r*35 z^a~5<#*>C1QO@j334+^;*CN6S5Yx<|Kxce)UAc*m9oEp{(MCom(?Oyw?dO6wxMq-?*y@rb*$pQkeKkKJ8&@YyWJhq$8 zBco);xBqS@Tr9NN47D?Yjdzb-g&HtI+TM;doOM7rz4crH1$ctZ@lHblX@$T)`g@xW zGp-7P0WH4P(c3%)qC%?%rs7m%^RDm!nk%M8)b>hd>jXrJlC%!+P!2^Psqan3uVXLV zAfv*3+BxHOvr%Fr3!<@ecPqOdC`+Mk5E%G_FHRO3-{K1{K^yg?AwbG64bWU3mw+ap z41(Bu%xRZ*fLYt*ww(5PRlj%46vVz-a3 zp3NaXxarHz3qRlBR4kkFY}E! zzUD6F)4F9b)jr?7mwr5mEmr|sIS_+jMIk8)UD&%*)-G!|=Wu2#pL>g^PDJH?JsWmw9s6EnW{4%$9usSS{L4EEF-k7K+%;6?h|;QHC+R#3INkCK);t2-~4B& zp;bkF*TK}yi3+y*%$-QsZNxVcPlFq;Bkgn1xu!5*9)B|5BKwHM1V1%eM;_@C<1U#^)U;Y~ zGCRg*@q2RX$xuLSqU(elqx0Ri`ko@(@R-rN+t=D#O1}oe>GfNg@h?nYc(Q#j5sL#i0McASxnr*dnpfvXg`to3(YV4O`8j zGAsP;NXgxIH(6LTWqd<&MUu5)+$M#uz~WXCn!QTz65U?;5!XSh{3Vr&4j$4uw_zk+ z&^asivME+s37~xk%cnlB*8o4}ZSV|M5$Y$Md&h#5`1K#+DKM<<`}v~pb%tG51ANHW zHA}>*iHEB$hXhp=~G!O$(_g;gig0LP2@3(Kl_{5o$X9CGt(yCc$ zT@O6LOCb5xEC|S9o&>a>P94GWRr#i*Tltx}DO>r&!Sm^ur|RDq$<+#kg`;fS2hA2{ zmmVcaCplc6_pXjeY66JY_zub&a3eEeoFc$3S8aRJ9D_HDFLNuKd?2?`tlk*!Bk;7j z=w(_P{eWlY5x`i#w90Gi9`sOHzgn-}8iO_5fR3GSZ7d;*^T0|(JQXOQG-Z3liQXm` z2n1B1^Nms&^4UJnnucp#@4lN@SyB39$mWx1#!oGq)=e3n1hGJh>^!%fLyT&M3Mvmj zOv=Wlgc_KHtE{c+-eKF7`w`chfb@n_=VnAwxUX|-IZs=R(?CRLDTUUC#pM?03#>Fa z3m`V3$Zt=CHNI$I6hrOE;Cd59t~;@PIBkhigG7@=hjZYoytY{#apE-p0aFI+?hUCk z^ZvqYCP0WmD?P?9d~zMG?e1VOu!8P8qkH;4cWBhj8!%B>ASM#NkhHT2{|Jow{TRG< zsP}ZFx8U}yx<*x%Cq}UAd9ZtW-v~`ob9yS+sRK4E>GV9DkS3uKNr2$ZGgf0Ce3d!($24DsISmm@57afJ>~mlVN$> zR%HhNO2Kp%TLKbQatk0RCOxDj*;Xlv2Qs#a95jV!cQp`6Hn<>Qkcj)lSQUB2$aZI( zVd-#?KAW9%YX?^8k;zO5z@36lGPu(Tx3KjGh0t66Z zL?YN*&$_ZjdSWbD_g)1m!crwcvvIi3V~ni1bw9qoc1`wt!(GbwS7bC|f&T^5yO*jq zY*@VPBEVp0DN-GR^IT`DxL%@rg{Xag(*=w_oG<#4e(QY>^-_1!w*%U9G<+Y5eRZDB zn@K*7K4>5qlD1wV5+q!zz+KQs5h;Ir@J9pdKhIq>81H?&3#Xaj96%=Wu5MNA`=KQq z%t*h5?T5M7B%VjU68VT_@jwSU;>s4_=!2~>d+J8dOfzk;g%9EDW0CWYx=;(5w3yb_ z9FX~SDKr^nU{q{*3Q0+G2`j=*OsdIyg@Gzy!NPmnOlq6%%+U(KKZDk%t9+bee1`fk zFC{~cm}E76F%dsh+E6t5M1>plp+2Cy41=oHOzU)y*eP9wj-CR1i~~uj=8T+XrSx@= zIUPQS=Eb$5W@)s~zT+RLGdo;(mh}5*ZbO3gd1hh9?x4W0JXMT_ZdXZl)G~C~k8j>+ z3n5o_Yn*wD*x3r8pUbm>hh)cP^C8r>t@e@9+n79G-MEh_*C)z~3Vk^kzF-aJrZeat zbj zgL@xsTWT8}-NU~?zULYjgC(NoJW1~Wmpzw}t=3pR-{wuF>6v|rQ-N;MSwqdN)#o^c zXULzs&0zTk!cI-hQN0%YMV`A)GHD;RFkc^%i>_W#U|ETaZm0^+$qQuHS{$Ak{XbegT~XmG4n);zqJTTKbhO z5-2oe4zQc0Kb!5PIMoIoOHv(p?ezIb9XT6oFJ4r+|Fd#9Bh(=cy~1~g1-a!u+o**u z^fi7~UKihjd4TsDC$$M~BWW4O7OpbXXUni^aT{T}xEl4NX^W! zNG`r-s0d?Ru~#8#C*g{Z?v|z)H4OPxce-1Ck$bg{i~-wc!M`$l z6rI0#^z?^{mVhU&C-hxPDeI?aC@X{G2xl#LT3^h?MmjaEzbsRcF*RZ<4i*0M$_#4& zUeGM95h)Wk%{eTek!dX>h{})nV>7`NX?$3=T*E~-gGu*D0`XT-AlkjlD23sJ-xof+ zpAD37?J}k@!J!uHm`sZ*q{mYW-NBi$#m?dwbCQrbmkhP|=-I9&Z+>C)2Is?4gAY3d zIkvFtBzEFNw|cm0R9$D2lRBp>pR}YolvhieUm^)Yw+BR`Ss6SQxV6UTDLjCdGYw#b z@DBrVTC!^*fED^3h$YArqkES>+;dfede6+URbY^(s}Qa5{O6y0w!HT+^d{P0YsSPt zaG%Zx_5pHY0<-b|y8ucrD4Uq^2M=88xGJE^TIAAT4pqiYkImtg`uka< zKg^!kl+f~h%#m=;=KkM26A$m=U-RuXnJ+h&&t(ey%VTMYQUyPgw|LJL0tm!S=rsYqr8Uxzm+e6|2*%^gO*8`9lESSC;4S1e#WbyIwg zpo`j;{d)S{dHc5LBWY+L)$^V#VJQ{o{@VAUw?L;dH%XOq+O?1^foZ>tyQOZY0$*$j zNN6#9S|q0TT=-HEY*}FV5K|(nYMVKy%kZ-qTXETN#s0NqlAU_1_^S6Nv3E1=B6|^u zEeb@rv0%>GvOtO*i3=4&Z*Qjh*`4YyYRPyy*Xa9~@bdxV{l~I|D(&i`<|kH7^Ccwh z;n0v60DI@*+lAWP8IXBH8MBt-YY>W|4`8Nb?(Z-4UM z{r1b0s^F*<$bk*$RHGpu)0-4d4XQz5YIM@#`L$GcmnhxeK4bnGO-GmxdXb^q(F8x1z}x$CCVc=`{g<&vbk|6OD`kb8~&0krh? zn{*Qs#rdUrk@W?D2^9inDDT=z{AXJ)uXmKLk;pV|+fLcqyOvwNc=A|j6u(sdVT}>< z2{HkC6b5@}W$*nW(-ZNXa`6|wb2bN$NRhFmOj9kuk~?BnfrTL6k*10}bjS8^wx8;H zTbqNtm~7RQY}UwxscAHw=D7kX7^GL^s!T!F5DA?jn_0f245mF8+kB8KDu%SsaMx)+ zUCgi;dl4m%S?Vtt3y&qHDyG{0SIio{$vI{J*~E$&Hu;B4I7mmxt>W{%orkaSY9CFz zgS!XH9|y%38YZ_~;n-rnulFCPsrEP}3iwF8*9G8e=;pHpFOsF_7x=PQO(+FnA$+M` ziRdq?;wfXykH3z}IsVvzs_n(~Vajgzuu*fCM$I}3+$e2;u=tLinR~{4$#Uv3W-ZWJ z&?pG#Bs|y_r%#Fsujvg`XX!zix#OP_TTc8* zFOsjb3u_TcaUO+YX8QDh;tMMzG)g!6_Fz4eHvjqUMG+!KC?f7 zFB17Hlx;T6v$BEhg3uxB&wTVJsEH5-hD;e&gXvjp5M8uDely^YoUUsb!`Y_*c!LW> zK$Ecq!-pv7YF2vn7Yjm-W?wXX&X`o*1QO&Juod){J*+&k-W**+z{(>K+oHiP1&TO^ zmt_0kp6uc#>VS%YP1To4)V*$h|6=L^W<;jT=-KK0Fk`Gf7n8=`pP?GMlrBf#Cfk~B zKGlBNxi}oHWEZ;bZ9>VKw(llpAnOkk=KPVyobg{=nu9}4T$sFX3D0=QC8QH5b(MAo zM3HSX=QMST*PbRLc;B_Er`?pJv>sbvYLt6?XE&x3O4^wYw}gW=3-6=KkQ~ad7#$5v zh*h6%7d{l$9U9RZ#mM5wsXRRjv|wwA5V=3~$%;LUlUkaRKVOQyI#$kN0nHucjAu*q z=lx&ZwL`84O38q+aD|seMM~3P15fe#<80l9RtAIt{ZMsd#2PF1j<%$+hD}kMHv)V6&4AId!-ur}ruqof#SErNVI+ZK6)0`S zPCM{e7NhH_Ru-))At)@7x1~8$i+xXTXUz0raEUu%mXFM)lvrxXZ=C8%IS*uS+ih$! zjM)53wQLZ+489-#IrCnwC0XzF9;>%GuH%pdK0u2S0c0h=HVdk@`@qG%-78{b81Wk}N$$GLSD zi6?zY=HRQx=j5%vT1b$2u*$!He2!t2rN-k}bjQ^R5t6E-2KxTuXKB#`ev&p`8Xhe_ zFA^cuM&TmLg~U9Yy-D@#i^Kz{*bA3pAJW|Q6>tV<=J@ z*L!t=F>Wjna%OEDR1{^w10v@e%zp3JeY5%BKDOf#Uz-ubXk3^(*JVBv>n|CmnGV)k zmiGg<-zJ4x_fH9=UKeRogrq?qh&2VSmNl0xLzF{^hw~tpTN&w&bW|tGpL5Q->93JI zFk1s^m+)TpCl1R}d~St4;mBv$`@hY4=6L`gbp!lO{4VK8#vXBN8p0x%0_n$hUe8Or zqA7t`0FD0~@X+5OO9PWcYoG#_(H@5K&Gd3tkq#}SrYJoqoe>uA=S#p#rya8ua40Hi z+~d*`31j=51va-}NUhXh3d%OK_|zv+11tWn9en%#L(m!2vLcc` z9Q46NfjgC)7T-Cc^`a|`OQ~Sk*k^88Txp5wFPu=^ZQEd+!(#QN$Aq7L>ruZ>jfor< z-pncs3H0Rz7-ukVMlE@9h#n2g*&$%Z#`*IA24gZYX`+pqRj>I{*D5 z&sZzlSFb~5EC}c{MB!wcQ{20E!{An-x(a22$e1p}@6p)MpHM1_-_2>#w z?2`!(jYS)5X_a#jwxnQ)S}LPt4K4~lSkXS)82-`Jb$?v4m^Gs)3YbKUZr9XAj?& zVqKUs`6$w*rO={th&HjM6LAbM#A&8+eIT#SuK-KlmT_-Tqm1CsK#Sp|Yt(GI+AsT& z`%sbGlhO5V}WTg*4G8)m@9^!~wJ1je~w|nUWTO)>44=XP0C*b+y zfTlwxbG(3+IOMVOd7r%yzQG(XGXK*!N}#hNjh7(^NhC3Xn-q$Q?62gj?c(iT#qKG3 zAzPHlgJo9JJ*aFQK{W@U>! zRizZK>Uei%$j4=dV1Bjf`t0R<@&kHWC=E=)03vx3adw+RX91?n~td?|Z z$UVNvtIzW}&-3M5BeK}$f5&23@?PWR*$2F@@?@5gHW_7^GbEg^<6^wFa@oIE^7N6< z0Xce>iO=F9(R|(aTFYj3REyJ>hxrnVHI*)kGqUec_v+6vYx{G)a5zLl|FD=%m%H9= zi9a26)LA)U+d(tIHHx#_+uf;+ZOnSuZQQ3J$P12~Pt?(Q%b^;FveNv#;>3c840)Qr zD1Y%G=*f9GfJ}8fE59E5AzL@tUq{`xH_X-2$YR&N=lQ~*wj#CjBu17LMVxeFQZ!p(j9p1h z)}}S2S1gj+@92w?z`%Yc61S0u@>H`LDNK?YbNL=3E~#51G?p7eXP5Qe=L-hKaz{I`4^g8* zqFYi;4lvzhrc?BPw^xGUZPxreA`SkJrn8KS>U+2_As|u;A|Nr8gdp9mAn}tBkdkId z>8_y#L|VEL5G048JEXh28-{LXVBo#{*Lv4{gSBSexpU^+bN1Q$*)h@gvmy@7Wy+1V z@1Vf`mWj%$)7*#*Mlp|PD}I=2K%!9WMkDc+NyYJZoi;Iu_a_Jcw&?r$N!s^4KFsy* zgx))|28*S3_^@VKqbQ|+QE~A)o1r|I>O?=9hiSR5=?W!{qAz=vIR{v(&1TpFW=JXA zk?v6MrA85>>qFaA?wJ=KrgN0j+On)j38-z7A@5e1>3ikw<$TmuEY-6+0>bcGy5EPdH5Dpx@%tOlG;l=0 zI}FQ&e4&c!!GiNl3QZ4)W6$)wKVZA`_o#u`&Hj=c$1@*0c8q5;bF`82Dgn7_#ZP^` zHY3fF6pYD#0db$0`GGNY} zs1Xxs*TZnY9}TT(2_r_ZaSF;a3HIntWHoua{zv-UoOAX+YAZwrTT!Kix3BBSVaALo ze>=iU-Mno;rTB<6iY4fw?mWNoZ0SYIC%<{UmQ|7AD7_EvK|)?X~ESU7LoVE_XQ&ChA7}V z&1gnWLV2VVeZO28Q2Z8E)!alZvsKWac&VHX96^Fx_H0?-o2xr16qxtgQ@K`54DY?- z|Cy)hhyRd(j*`Yc2@cI&Wr<6+uOcB>hO7rdQ676zWFwOyUf8&xyB#u655eqjQg^GV zy~kdUVjvl?<%xq61j?25(~(hpeSw!gA=iU8+vF4718d2w^+|)FBjzH~l_T%xlmVm2 z&9-%j>H6H8)IccYD=p<+*%b3hs@B63bM>@+zpK-i(8D_ijg5)KI(iXq*{G zJte`5UEriY+G(vqvSvF9o?}pqSS+2Zsq`sM2Q{O@TW`V@v)x+y_+!0Z%!u^{jSxKd zH@mi8M7$Ob|HQl7@KgB9+id)7v8WZk=5KFK%E_APGyTQV1y`e4*{pN!O>|$j&&;}# z{xRtGhYh!PuvI?d@ohvBJ>5@vDCAyze)Y1Wj}0~TBhn1#_BnNA;#4B5#i`ShpBi)f3(sU@GR+xE~fPp^$e&DR!gTn z8#XdwAJ)m)9~oENy^Kjs&oA+B!QN-nVZx!1G}^0Rg>HZxK9(D+6*tCq+qn)xrnvY| zF*v3n`E9DY_`P~0hD39iE~-@+)-LkDip`I2fF{J za3pWq?>gSj8usA0GPYGfM%%J7U;XOa_V{K@E8N_+|Mmc5yADj|0P+Sk6byC0GCT31iAjh`H(1$=?Qs)OQ~3~d@h#H$w)hI>^kjM|9;?p zRMNfted_y{rf5sMB-*+qHl6SGmFWr`jCQF$L-z&MM%wdd^9@v})3AV8VZ-7e(nC4z zY+k`~oq%^)q+OzJyFDaM68<;W$w}e^1ufGG_H96Idxsk_`gGJofJDC#&q@EcUY(n$ zG<(%|bEa`$cv{L`=ESG$Cn(>}`hkolC_#0xaJIM6Qw9LdCq1bjU9Gin8%=b1K zJw^nal^!yuZQ);_bpQ4mdj^09$5s%7OjVUmA~elNYbq&@Bc7wt+EM(8!?5u zZ+7TCYnICzv%lf}Z|SHvmn~g}wP)5~m(ePem5c@FZ*Ka=BXm{-NEbQP&qVBWLhL42 zK`U@1U`o)+0P34k9tI6#FmZ*&4vdNRhvt4$XJGs!u>0(=HC?(RHHm5%p4r#dHInWo zFkevWqck7!9CohIpL`H7p0U_qOq(9-QgH1COUm!kz#0znDCAeAK7{ z<=LZE!8`n3_1ul_`gV=TPiSNB(e`j&_WZS_+=4&=lw_piPk|(Aul{Dh@ZRZ>`#7D3 ze=&Rd9Oh{!li8^q<_Q;25EXn{79k1@-qv1y+T>7<)g)L~p1|%o-#ydBz>Wc0i=ZGu zdErvFwXYe+p%?#+NU8p{|90Tj(1w*zRhs7VD!RbCqZ$_ok2}WbbzQb);Frq~Zmi}A zwg3^d+fOu#qY3sqk(5T2=H4to@<+019_YiX5Zkrb-{$urZvlwDTc?-qe%J0(cz{7e z;@y8^CaZ1MhPU%@V5y{eIol&qBLZ+yLzm5bO0M1yMh*qR+|kHyF&=|jh@nUbSH9nD z`#_NXPU(%&0h6=!j2v>i#zpn|hs>ZfWlx^+3BQjupGLRWyv(kAmHLNrC`bq!IThK! zUN$nTm`{{Q7`0Hr^xs5|G8+GjooKr$Pby646T=m^e@R6TsO1zl3!4+5tV=%@iQZ5N zc;?J?xjC%qk?kTuu!rUAmfBhg_)KeXJfP2PH5QTSIlOlT=!qP8M0gTfJAr4@qz!fu zbCI$n&g8#zdj4T7DiPX@yPrQlO;^$r+I69&+s08MKHDLE<$GX4?yrdKGt=Li=6&^@ zS$t1nDv0KSmgsif_Zq%X{`*6q^b^D1asD*zrf0q61t0zRV>;q*(;KpO;ZpyEkGcQ! zx@vHuHHO&#%O4?93|dCiySB)>EN57&iM`pd+7X1TYO86&3!RL|jz2VH4cZ9)w3Etp zBS}r}m0hA!sYn-4%=CX`kUq-SlFxWT-H2TFQmxAhJ(kj2`n=t{9l`zVpx1Q!j>Z`N zzRt5tPxc*CbW|MKXDC$-pq2Os$rVFD)@d4Gs<$Tg)Lhm@&aSdGVyy}{lsvdbRGBqd zy5hD~WIr|M=R6QH7-Gg5kDAfIW&VAsO|#%+GIH<6IUlECIa>w6H~Pl#?pUQ7Dawo0-CHRi+y*cLGuf5=Gmxn@qHfQYrI{8 z%O2n(hPxhYO7?Hz%9mias>({i{yf7*HGw&*I(SiWh2*kvD0RrV>Y+|+xVjd!Rxu*- z1Wvv5$#?!ts*gob6tGxo7>Ip(bq&OdC-LWRgZ*Q_Fc4u&~~VjsMvP{UsJrM zK*)cAz*fv^hs{vpcnDAy|*DpAc16q3gmGwnSmiNJ1u_WoK}+naO+)M67jBEL=R|6~rBeR_b_- z&G&VNueqbHwE{hDM}j6irCd(xYUd#?S_`e{<_Z<3DQ55xtsE0ZI!*S~ji zEAT$mstAq6ad$e%xAFUV(Z1uFD4XFgZfq@6@n^ zI?2cU;IAY$GBtaF8rptVk+U27tMxiHDy~lLgdaE+F}gj1e9WA)1Xp+tm0>=y320b^ z_hd{D-FKM_yMM^4uDnt72b&7?*Q;l!Q020XqMJi-Cu8xO#zDW@>>7bWu&$132b6&| zzh#|24&qZHd*7GmjLcTb%t0CT2)LcNNEBVb&Z%29C>h0i)pORQq`;1GnB)`;#go2S8odK##$f`>P8ZB1dWpf>Uekm&NOVoAG?C$2SJ8i}(p z4_r^P1M*HU6A7P3{fXmYa<9U&QSYg*!1BvWvrn!Vh^LNbN8aR@1_T0Kvpm-wT~AEG zZSvtCg{d%mV;NUo{yxh%tXuA`vHkbtA6JM_;ciX)b2FN{hNmY}lYaU!SErkyLsg{bBHxO|Qug9TBqM@qlCY2^oS_ZF+}!U0lT zs(Zv4Cd2N#WGz@M$zi&Sg9(Y(!0e2#tax@rrwOQ00gZNgCWZP@7ELP@tD(@482=~3 zKqJ@uUZR&3KK0kqX2WMyeUp^$D|VJ*?3M4~?bOUdmw;_wF3v9IMxf&d6Ps-2c{9sM zFw3^waD`)G#aW)!dE+h~IS{+XH8&Q?X0qT&{Wxdt-=%ZbK4oKp4x zj-`EI915$SHQ^%S(2=(R6png=4GpO%r8%SBwq;qbQJnr0)Vwj7%s7?fWL9!kp%S}2 zJ?Qqo2mJUA=2*h2#u6MirkDJj) z6N4XIJ*!9_W%TIWdXxfo-rlqm$|ETwH3qL6yXz8$_>SWF!xMnrsI%fM$8=ZIrILvy zR@-Z7v!5IvIK#01_-M(c<%zz~X!My+NDC}3R)5aVd4gEBzBxb}s;tbBR0yf6zn!)& znvlFGftTspm*^KSg|gpdJ>1<((u72v;gvynMf(+W=6THyPP1Vmt2f3KuLEE*CaU;l zI4rrD$z-DD^_gaukR~k>K!^0gYE@8WqLO|vDKEy0MHf>WK1n!TcSJQL%H8WSZKo$g z4Riaz7*Omm{I(BlcoP0RNPB&76;mn_!!5C)>NV|;1|}@Sg22Kh-oMIDdKqwkL&#PM zoRT>`b9jF`8h72hH_x|x$!$Rc=&zK$|NhMDC@}Ex1d`yjwR>z>!-hC;O{bkEy{6(q z0~JI^1%@zjL_XcP{O6SKr3uH}Kh$g-CdS`hgUARiNh%1AksB$N(Y8Yi0yI6qty`1F zYe^QS^qI2rx00sVr(e)rAz$*bi}=(vKDV7VxWy$jMuz znR!hGD>E4sXeE5@bQ-aICZ|)shMlE1Eu&6s1q=!FaFPPnaoIHRYWL>SJLeZqO97M@ zY>Fgkk3_Op=DDWj(7H-sNJIiZkLL1e;v`F`#1aR+BJjDsofmI+8!?^H>~JH zvw@k_&XOu=WyLfAV>YWU#;O6dnB@=0?qyh``d)f$vL%Fr)y7tw2E93)#WI-bve@j) zdGH+ZHpnEO&vXnu!kE|I4+*Yk?mH#!kucjOxmfqlZISU>=N}BWB7^o_&SS*BVmkSd zZuvQ%F!HOw_=G0a*0V?|UUMqJDAxUh(5)(qzyhi!18Q#r-4mc~DzUXa^6-u2aT*Kz znFk5hC}Et1Sgh=yiA84t|JFUfL^4q$XS5m;@KC{)KCJf_HLNhx{-43s0^7N8rM~+j z{AQU~Qpj=iJKz4c#$fH_;PznVH+-xyaaE@O0K_ zn_pEc1|5agOGe3F`;3DsdEH;cB&-3-S+)m{73cf*}ybz6BA|TnT?-O$Rf;xx-paRoKO)1+;rrtQ4woFw0EspH^bFl zZG)nT&jK1sruGf48}$AhcP%q)2xS#ziS~jK!~A_c;C{_YJ0%|N6w9wQVx{5ZM>Ewl zdOL1aN8S0ou0GmCp=^~*c6cX4TJQbvhi*)frFyTK|s8y z5(!wxM>nwClJC16iO*4G+`o9XfJd)cD%l%$S~(18Q^UwV6OH>aodUO>CHKEkkf{mV zhF8i3H8ggudo}Rfhnq0Pf~!bOWTfvtr!Y-*$s|$AU*-A@Qk&HP0XXsDv{_KIx%hLi zCyCHiFb)MHgce2bfH5lQ&vRh~TUZ^8I0NLISA>>=^fmR!C)2(snEVg-#}5*dprX_J zpiUcEV-jh}HqE>J)_d#Ae`GD?786#XK1n!K(evHTBTolYxg8;P{*a$vI^mnCap`Jb z+ZS12IE$%i$7kdW?v9P_(grba0Sjh#RUK^=-qX`Fz;=aNoRr>FiBP$D1VZ z8!=ibHbwH8I8*K99q-GJ3|K=-Sk8I2@!BbJI$b>77<)kJU}lX!#hb{E2C_z&FW~Ca zcm#^KV1e5{GGcK{vcEMYAHKCcv)7T}Dr)Bl2 zc}9j=I>}Z#VbdVrB2y2x@ruM{P>RIC3$ITg`CdGIGe-iZ!k!5czPuo-rh{e7*VP%D z8RQCq;T~`y&>)dEEUjea5^y6YhXp$T5n&7J_bQXrTqHMwEvJ#B&`(yM_LHq6)HKHz zWg`}qVkHL1c{gDlN%@A)viOV+IF+E7j}fxwGH!$2>%XoIa#H=1rC?%oPLRt5-+)a~ zbp`!R!`pW}nq|pI!o7T=*}VA&$J>k76E!$=3v8cKtZYp_N+b56sb2adY18?@f>(=* z*G+y_D=^{aX}gy#*I|_CQ60SX;kNZ$xW<|-#BL2>*FlxVPCIQva|xIpCJQ90k+;DA z7vm+sN$9rR2$P_-4*~xI2ysS9e(lpBOj44)1kUl#$^JE8vvAi0H`>e-v#=3|Bt&!bEUf88JRO4ZUP~E&lEe*q>72 z6ldNN)k^&}@TE+B;?B_R;u%npVsmWla?#VfDLmq-e6RS#c5TY!tOwXL5o`|fNs%u9 zEqD1*4<%$))?+0x(d`#|)x$GlMH1!GZ9n?)^&H*m_b(}`_BuAAWu7;mRaA#B!oB** z@wQo+pQ?7d5kuX`F~q1Uo5GmZW@zthL;zmfnH&tO8-? zz;f~?t7wPnHjzh%PY=eU+J*HrhyQ>bigpNb?S$qvwrF`$Jkby;Mf8t9$m$8q+t60Lj7=0}~vU+@@N%=jB z&p%bPhOIJ|^@4r%!a?in&X~ndoASS`P^8T*lR|e)!h_3WYtB?1e>D)^f%6_4SDW27m+Em_SxXY7}cNE1q&ee?RTd}j(u9lbNoR0A@TXbB4(axI|e zqx??&M%?Fkiji7C!YZK9WAF0W99`dp2)U)_Q11RC3gcGIws)U|U%7 zouZza+A&DFwR;}y+Z~Ao?7Yaki8@y)A6qRo&!ErwUlX2e6g?g#U6zN6+wcjdZ|DPN z@B#;X|69ZcK$Wm4)rt^zXFA}wC<3`6$XZTf;eQ*}E%#k4SVHrMpx+RVq!eV?>eYev7^rIcS)_m10| z{w9im@)h4}h!2n+l_>EYx^3V}{k|@nJt&nHr`eA-LZ_0kwYdS?GMTT%%2scykI|E| zOy~P*MD`Lkc)k;E2=vR2T*v>e%~Ft!?_lc*)DA>uSm;oDE9-KHL%Pv+3<>_Q+-QSG%XP2k0`z{OUVo0HFDDH^L*j2Pi+7*{`SEq@l&ETGkG1;zO7uX zx>n12i2Ma6d-jyhrqKt91MU3G;8rh#nHA#;ycUt|dI?Y0P#C7t^g<%H9Cb39 z|H=GJi`#tQUXyIdEx+l`65@z4e=; z&;rg(uH}?3l;%F?v!|A~FFsHod*J+Squ)VWq0ZP_$W?;LingJ^nP3!mvM05(P(aP) z6>owkaQ}UPT@G`n#Fq}cHq3!Nxxwfn!xD}1UU@q9u>AgCinFK&v9X50g0v5B!DLJ^ zn11?mYiX5(YW2wiuJvlud~JUI+`Xw!U+3cYg3npy4C*C+E7^$dwzTcF9Ue80WgRWD zjHS!(p;v7UNh>fl{cEMSWjuREwCG&??H02mBpyI)HePV-6#mYhJLoDfj17WbsQVqH zKJAbHsD6Kwb-$MtGMo)q!laLv(CkK41BU~lM|yQlmqziKAY&w;WG4~8Y73Nqcn+j? zLN?S_Lu+rydBR20pCxy7f@#i^;VB~!b31I*Bs%x!?265w+z-;AeW&%=Fl;9#tAqWe zYB7TXU-98REzQtt?yuhW)VLmz-m6O>)dZMK5B`G9nQ)}1rI*(!MbXQJ-S{r@`edy zT1c1x-8wzoJ5{)OIIriZ$*ewNQ(3G3rbIl={S|v{O%er(ejUS30lYnWCbSE>8&#Vvs{wxUlmq5xbCRN-C>dCKmY6=&4Wt-Zo6tmr)2Vb+L2}-$N8OY&9 zS#fp0T{h>}zh{kvRUOvaUgG2^(MHo)iMGL~CMhz~w=$0%n=>vi&+?=Ggw}T}0*>Z1 zH^$+O)SQ%ClcOBre9)cq3|sq*_dl_w%881N&^`dItTpbwGg6;lIovzbKeK!B)~w+D zp~KfD7D5k;gQlY$bGxR~Za5II>}^RNBA9jMYHucMz3C#g_n_m1aGm|Eq4_EY2o4=j z-7E_8?)v#Sok~=@k2da7ojzMw6SP-opB-%z}N)QE>AAMRZBsxzh@W6rn4S6L5;tf{rVndjV^ zB7h0Qn=zz_Q=@}ZCNq%7J&>-3Mv*dZXM9)$Hq%e~?yln>UeMnrgsK2lyz;YL`Bov# z`*ls10yZ22<016vm;R9}kl=1*PYbQ-dP&JrWPic(!n5(6Pfu2I=`S%iuYo{zbM2^{ zY;VM=;D4UN%ei*i*h>*_h-*ki`)`VacGSv-GO?0?Hq; z3Z$I%fZjJ{oq|{ub!X^o{FVJhTLxP^M5l8$V8wjPHKVkKSlxkH3 zMt7Wztgy6Ml3VTT0j2Ap+sbf$9L+VqUlh;^Ie#}eE>hpAgJ7!o;X$Pb4g<=njvtWF z2l%JsU@`-3pG^kP9+o~us8a9Ga<;Zn3*{#(-YxtIB*_^_CL1s7A44*Th(bwE4$^1m zZ*CaAA5?|wuFxhbUJQi34j=m8=)k58>iC|My2E$QfyQm1{1tJ(2!O+3QN^{5sc2M^ zw9;Hpu;`xR7?q?a!+*>##Ua!~ZT9^ktQJ%(&I{Vi=4*FA@U_}P+mZ}Qac+=nH5f(a zAIwM?LCc1%I(SX@c(Ee20lry^+%)q38}uN)E9O3qu_Utqux{7zOkuD&hf0qyOHBGa zpaRY3x+E=u#s;FTxBCY;%8pw0hvC41&2gm0X0Q!;Blf*4E87t$uIb~eQ5loKa8yLQ z*6?ovBVB0R7J1nxO87kXxC#iCJHBfzfDIm{>p#dkdr@0zs5FXczsihDiLb(-O_U$i zCvc@mF`6g>m>sQ&E3KuCuCW6UmO*{zxh9X%XJ_7B_wgbF^|26c%rsP$R82K&+x&)m zQFwqB{J5of!%twvoD-e9X#JDb8gCIuVWhm=et}x%T7aAnu}Ngr0_>S>x0Cmh5v2tx zw&m1wV$9D{6iB{rO<*FNB~##0_)pQ!P^uV}f(6yBC#x4$MFfp$SnQO75a+Y@+CTMz zEy<2Wl2KXal+`7~jn1Q__%6+?DJ3e{y-Ca|>?(ZMv{Q2WYXrIYdmj*An>+0C-A%lC zkYIq1kZ=d=|K8j*M4k5fkJgXA?25!R48`p3)4qN@RrcDTT&a5;Jje*%;lImgiM6Km zDG?;;3_`0I7zHMN3#&#UeM;SEK5t^>rj6(l;~pbrsy?UThE-lACqUjsJr>zcD0=Yu zu|cmH3lGolrY+g>RlaEI)Lv-iO_??E7kain_6@>d9Zf4(>suc;+6Qv2fB$eF&Pgi4 z;f7ZcsN!(W8~qv$34D|HVG;7Z1O6+2F9qi1DI&y@@r|WZvd>2g2W8a-0SyXZXsvlTG6U%iRcTJ&rTar~D|%l~I&0yo_*D3&ygy6i8@ z3P&39K+4@9;oziy3vw(x2S3`o1zq&QQOr?0X;05K(pS=TQFA1QF+>!(v&Lp`+}NGP z-kxN_!>)&%{zHCWma?vD7cV5XQNpQezD+cndY7~~t;01=o8(>1sw<{3ZgW&~feAuG z$wHHRepR&_xv`uR*lY%%w$>`{ej!!+`@LW33607UTJmNSG}CiShOa`1AloSxnt+4D z0(?2UiAbj{1g(0nT-DT?ffU6B0zC!ptmX|aY}|7Oj%$f0bz`SxPjPBZRuj-{@UMJM z>#jAt^8^LEDKgKeYc&l~?bqiwovn1Bt?wBY2Wr?6U(MMj7MY~!eDStHl%OhNEuera;OiR?LHSqM7sL?Or(X0A6CtB9XtKP9_$>awCYAd%1B6`8Z zC`a6uVMi&r<+|x04;5aVCN}YZVqtBjRJp;qbXHBDBiz{ULPlJ-sA2AZA{n~zL#X8Nw;fNZRRSud>yAStcp4EkMh|ub@Yp48tI%_qf#J2tZ%~4Gb!^D#>C)G+- zobj)0h`qwoN41D6Fj7-3CvWpOtfZ@RYjSBCpXDEof8s2qxvuLeC3UIvN#m{p0T5On z1xVi~1GVNI@#DQY4dlT(2tm60#DyG(sXM~+2At^#)DKypLvE9gX`^||rB?x5S=o)U z=mrD40xXlZSmx`7QSPbAWfpRthJy&L8yBqAVpk!mUTPoXSSM2?5J>*s6carwwxNl5 zcVoWqiGG3A{Jubbm%zo9c8{uJO%ksq-uEhY3a(|@&O1tB0^vc^AfS)rxas+% zRuPy=@LiT{vWDGPAub$3BhhkFBv>VEYjbH$7RAwcU{VR|dMw?%e~Esd+ya)Y-lE#i zN~@s$YEPl)&tLFmmHBDOz}Y7rz>n$<$MGNzaO`E`Fo9+`?*BU8s7&{? z-)%_VgIxg{|E3m87NFhKX1eyK-pk^TC3eB2Xl4PTcrj2OE4BxdQHB0x$d#na3Y@_1@6Wdnifa&0jQDFpIWmn?A-S1H*6&lq}fCE zIplwXluau`sLO>h7PyC=7_tt5F3X4SvWWT?S*AK zYXp%aO1`gH9T0U<6Ic_q--8HN&zNKIJk}B(i8=%A2n5q)bh*a;djJc&?|ndSl*rrT zwhT14M2-?1oJUm>NNOQ9fYvi_4WQZa!b{J%sz69mT_tBpDpZKcSj%LIu_Ww@oCF=>p|VGi2Vw;Uu!~tqvH@~`K5;xI8O{+)m{vw$f1{G!~4y$qE_6jd=KQkNH>56>E0cAApM?XnXoLaIX?6+Xz?!Zb!%^*QcIG5ESy zv5;@rh2gD51fcTfP;i+2O*dtcF2@lhF5b7OW0hFp|5*Skit{P7|Hy3jxh}r#jsKZ? zjBO4!0(W|dtT00S05+@`j*A&2#HPs~+oa?i|W9&u?_dya>b$MQTWL8|^x}`%&nTXrk&|kCd<}F^n&)9ByIygg7!pJ~7T*}2;i7D*s z7HH^3nBcBhRZiBdHkv2-=W}xWXr?J(DAl8hIox%M?KmBf2WB~z3~tSDOop@=d*dLO z`bUu?^mLLZ|5e}#?j3q5`I^pETv){#ZT(&SXMmKi*z!2O0)d*WcLX_ zRJq4Xi2QnTV7yk+EMw>GhXchU^ojwZ59<17eZ@(k4Fm#IetwJkB<(>BlG7oIfs&aI zZe7bywLQ8&w&LsGkjnXCS<3xZP@E*^0&X_d0UL_OqP`Rw5o6!#I}qKQb=fSJHUf@l zOm-OzqP|`ixG-k=db7R%&0tGjg+#01xVam&@eOk+COE$8<^A8q~0ZG{Dh` zon(MAG%LqLh`#~f53)}t$TtVf*pKGHGohUse?L+SiZX`nbTHpAftVBQC4LyC|K+{V zrTzu7ZC7BQ@c#o^c?aIXh7OhHgDzz!hRIhjP4iG6Wc!P4#Z3-`d~*!w5~;{X&fvWz zwpBzwX>7U9K4>yZE#t!_wnTz4Tr62sm?`!}P3Bn*`pfsPh{|x=u4v6pdnQjY^PdYGNV-kKvU8$;RyT@CF9z|{WvUh=q ziaTym;R!O#piuXUQt@rM01i1>6ovJrs!?|fcmS!;-n1-U!*B|b{>|Fy0?6xn^U(baM_ zkybONR}h*}Ry>R7cQ>4%rzg1ZiVG9cgeH(`H`{%}sb(7&ij>$KU%KoUE%q#M(tQfz zqLvsOC3N=mP*=fhv8~rzYwL7Gqs$Q>+g8$JRuA1RXqm0Qs=Fi`5w*RKchgm-7h-}P z_p{7j30zV%Xr9*R^R;+!flnMt3*d!H4^0`7_L z1XCn&9|eTix%#-j;UuK>*tR=Ty%%%~?L5K&W(gFZJlHQ*f3u76$~(r#2@*uHU`xFc zv|CA{hLrca-Mgke;w`tn(8SMRGxr|M6MjPz2xx6;Jt7tFl~Jaf$<*XoI1Y3@>c?sd zit3Ve33t6uj;q4jwE#L5+h-vl{NhVBdaqRE(^In%oy&P-u|E^--670&yW(JldQK#x zmERdHrUZc@U!)1fQF*|2RAsx)huktaV{>`4zw(|EY7c2e(R?Ah^Q7PdMtLNOAQeUqhck{D1I@5Lj)z7M2|Rw2uW%(f2)vtcFI!JnpiuJ$ zlZtgqEf{k`=(($vn(GhW7amvuhHtE>YBZ|S(2K*wn<|Ry9*Bb1R*-DfDJ$kSU}oh(O8VmM#Y^CC#_s@6W#@1xdSzu zMFI;pmEZb)y*YnV;Ci?f4EaVKfJy)3o*Ns!(RNR?E6~xZ=pM+2lHTrUJ^Xu|t+!rR zWiwedxaAwGJnwzL=u5>^;k;+jOn~A`mIWQw9*SKB{Mo|mdTihZY{lYl4MXmPg2Zx& zHesr9&&IG9jvey zx2k=LucyxH4J_ghMHpBk3gQn(SGq`EvlK)tA;2>-z$YFWm07Ypxdb{!sK55e`WPp8 zPF_<_p$Y_lBdi$NMudOF5jW9Yx-I}6X@)^J^TV}{Ma;V+ywQ`uT z-bmPGptWO&bpBnq5LwShJs{9)F)2VUn*s1D!F|vV7e>NdR@NqU6R_hr6j!{&x;NG> zJ`GunCQBx~ukN#?yU2b;>TSxQRanFs`KN%HQwrnr4Z|Zt>3fft);Y40qFj=y@zcSW zhnxP&cyK;)SV_hgsHDbaS0CfGE2{p>Fx3;B^wA=%4I25(NM=>&ZmrKdBtPZMbeVIB zBF_pr=t2dbjH$lW`cr*;L!31kEQUV-qu>{5@&nT4Hx_BLE&m2OS!1^{1{Ws*++))Il z8IFm$&_1R}g1K|1FU{jC4gowPDT0&(xWSkym`LW+yo%s^3i8xRiA3M(ADH@}baxgq zzeW1?!Hjqy&U+@FZw!gEI%0qK4iD!U&h|h+PVM>D?7H>wKMyr;i9J->mwKiDT7ME` zzEmKDfk>K0gRHrCMPxV1YU#v0qfBgW-5h>RI6!-IDB9>>5;hcf_l3)WdB}I zh5#0ke*srcQq)_w-#2>>MS~Bw4mMQ<_8OUTI$x<9f?GfNUI1Eb!Pq$@{kS*U?j1IZ zfCsmH{RPY+Tc657>5B>Z;jd4+g()X>*3pM?@Mua@dnt|k3@-^8ksijz$BJ|w`kf|1 zDIOvYt~n|Pc11VNo=SIR9{S9(+C8h~{>Heo^ip&t+&Kq+ZA6(bk);FbK z9f$GWRn4FOrmruuJ~W_yUf9l&?p1%$FqxKnxGSq0tuK=8o2R9^hI)+S3-up{GnEgcG9?ucdEM+-P4^^l<(!V5L^i zfs5+L8Ra19fJW+i+xVYP|I-Fm?*DN5SUtkKIexr6H6af=A7mWHe0=K*V-iN43_l#Jk9!M9d)|W@S;Q3`H5OCsTx#_<_qoOqWwKZvM?}PU-jmj3o z_8EROW`C!sM?{4DU{=knNjMx(!KxM7->srKVdON44U=w-XgjZF978Nz0o2YHfGL9d z@5Hs_O+DB8W1Q;?V2FCq1%fv!UmZ^P=kfm*EfDsfG$X;>TUfFk0S$^`Zc0^Cn!1-y zSvfN6)fXfsZ>BpB3uC@etC#l{Y7;uag@Q5g7auphoB&z+>UrZ2*7sToMBzTJ zNugYiwZ@OJBJ+%9Ai~Ad0zA#m%VL1~v+{<${3^DZwqBv^dl|*$jo24zc6$V1#Y!ABaJ130z zl%Cw&yv0HQz-OcIg1(o|B)F%dRu}#qLcT)Cl9hE#&oU z(d|IrgqiD->r8bXS$GisC9}yH4JRQ-ya7i%X5cuR>2oJ)Hz`IPP@h*{+o$(~W zdeF@lr}`W{WqrVTUI0Xf_(n23jKT=ahg|?ayJ>_$(yFyHkBqr$>tv%L3a&Lpl6msI zsWN@ZhRIaxkwj>8>VQW1Vufj}MRdnpP%GfdL5i|0{bJw-Jf7)&lp;Aqj)F;~RA3A_ zqNT73ZawAq^V{YuzGW2Ol!J}mR!_m>>v^#T@Lm%31^|cegwwk$j5cN3(!pRlDcS! zSRwC{QD)g$*+fP-_UPDK$DYS_9LGN6clms8zyIAh=W@NS>p32e`y(O8qKAQ>K1*{o zG&${l9dLbG>bW^BC|}0}Vq0l3K9z1M`?mc)*A8}YAIxOzwa;T<+EAehjg`v76MI6% z#~R=uQ4*y*FSR6y$n%-U3upMi{~@GrW`Ewd0*shS|6!^Q5?h&n(lH0IGjQHv9RI{7 z`paur4fgA=&?yi5)tO#;7m!%6(0eK$&7N0b9sUcnj8m=cOW~c(AL+Ht@UdEiXEx(z zuReTEd6Cr&zNhZ^jp9c;=w!CD9e!D>m(rp0PHbaT(A<6hMFyJ}R$l!N6a~Wi9z&Vk zfYmncfs8(JrKJGNA~Q@(FX*L_W?tW+&@Zh0Rhs-|K?Qd^ zjfKPrx|IE^yx*4JA6;DbL*5eLs??om3sKwpyjT}KgvtWh-GT*YR&mKkY}?cs{7x_1 zcFRO@cYIS0ZoWNLbxV(r5AHh=$ezE0D5Y1{b7uk8HN$@H``tOb96hwtMI|i*=|eE$ z9o#KS^vi9p&|xnM?vnY3);;v!RIs5eB^Ma=xps&56;rnFMZlJ8sUhnx{#PP&b^^7;V^dDoBFvX?LNujEe3rorEwFOyGkwGJm9RZ3$gV? z!K;UR65U2i%b#oc@p;mETiT-1UM&3Nq@}?rXdQWo#T_f<$qAEGi%z>UnCYkNaHEj>iFXb##{CbbU@;Ah=`079Wrpi?!!9L1d zCzv-nIb)QEg(sS5j?eyeoR4>hZzuhe8G@h~zI8r$3&3y#)Ld5z0d9l8Fg3V$#(NLhK|-+Hv`tS~4u&FL8zET{_xIL2m|+1IE9i)mb<8Lq z`LzH{V^a!7dOL7tN)OqqbH^$`MN3Gyy`j zNa4#l_kLV_r-DKC)pA-Q*i%4cEA*-%NLrRzl0!&fZO@2tM{BjMbS=bXNgSxY)V*Jd zQoN;*qnh6(FrBO6m}Jl|KBRl>c5!DNKKLT{TD>DNzQ99`yEMI<-=*OTZHv@VtG%QH z5K9t3x@DwVZfbkN^Hgj7x>62gt@Za7zm)V3CePckmJFY1X)Rf^z1rUc@6!v-_ak$g z7N?C@OAA6QS)41*j^9pv$^)$S0M`o_@^5Py^T;C(tLIjT{mw026GgWeo|6mQe|Hiw z^-;Z&9Ywt42UJFtREP#R7Toe9^EvQW)f~S>o)xBz=~qBwd=(^Dya93V^9vurs^89# z#r0OMk>h4EF(ZSenIk@!*1s%BKvan5rO@?tPF8;*Zq!FLG29RFmfScu-dp^yI<+zI zh3khNIK_W+e5y;=Vt@r0;fH(LtDSM2mWY2t0G<_ZNO z7H~epL*!DF9A16<6#%3JNWpf)RVLUExC(HBNnxyMo^AQJS+_xONqV9N=nm?Ou_Zz0 z@8P6P=8oin1yPtqnxQ(33iz0N)9R12e>`JO^B&P*Ktkj+r=$B@`0Yh2uAx0$m!w`~ zWgJnY#y`INVUThM?59D~!|=~;G%+tl4t;*H5mKb+ZTx;CP|9mhW)Lis>;(;t@S@;I zWs|Le1C!1Wy7h*YEwo07Ii0bU5^Y=3l8|#_BI+4mh&`iC=>9?ouKU?7dv*dUfiF&2 zBv^{(i)YXwEA1mikqpf_k1wLA%utW4%tpp35s1Tz{&MB7Uu6reo@y2!RUYy6&}5uH z4Qmh)M!{lXm#eb@&VaAalwIWNFPx*NViqq=&NOjb(q~iBKGMExU_g1#QU$mf1|WuN z^?&p;Iuw)iLA2WQG7cYN+UsPpNDXK1{pPhI16) z88YdvXC)Wi=jqc0ne#(_7g1b4jZX&+} z=OYK0az)qvzri41AD}!w4!V`X3*HS`n?wg6Rb~EeII!2U8hf-FIy0v4O}+bKuIHpDk1d2z?j; zKBOk8zNH7~#@;Jt@hm`&*lY@;SC@K19&*7tllGOA&z}FU^7incNDWmP>au68xMoXH zVkI0Km`(DC4Y2$TwBColJ?AE2ZYdY-gN5a19#6O5`!iZ*Qhpk~Xs^ww;do-P{A ztjCS!;N+oQQBKKCn@RYAKv?mf{so4rUEa0=(?9)Vm*Jdu|EQSPeyq!ChyCJ|Z@whKLj-&gILxIE) zzsxaCDQBZ~sg40Wj%f>c6>jrLjz<&%H{fyxprK_KdZ`ry_(BvsC0k+*q3UC%dbwJm z>uOWlPvht?-kf$@$6s%_Q13l1-4o>B>hWT3fnV+lqe4A>^To?*qB5rO7fm-9cZ@*k zVvPrwINDvpWTDr;*)GJzePmUW#AWw+wS@TBC4dHs+bvF+_I}5%2ws(pWo^5_%&|cP zT92z||E}k@UseM*uXHCWa;16M5fyW9Z#Gq*UZ{p)w!y|HS+ zXUSqMXL;>3OZZHu(qYO9-4U5QMhVPwjP?RT5FBxj_l|i93`!{+1ccUr(YQ-qo zXwZy_Wd655Oo(HJNbY|B&*zKVbW&{tjc}$%-N-{~8)IrtXoh->VF(vN^~-;dZT0jI zSM(&4q+2@qK6#)-3Ra{yA6MIVtA+Y{v`A$wYUZzqC!_14x@&sIeLr*K(qco!B*Dx* zUb_pq8sDC-wO{vs@zha8zNWVS5kw*H4&!g= zh01}(>@+tr-UFSbUGoeyTsAWC4$YAQ2KvH;<|El88R!gUaZEe?bP6U#s{O$g#z-ok z48hdXf%NJ@TR&vk3*;}QLo(5G;`l9xNkOEO=;ZU-+2@Ev2HcKCXj90%hR4Raswo$F0d4qJs@jR(M zFuY9u^5w`<(TpEsm;`Uyv-Z;WY2B$cF8<1Iomf_RyEJNd*Hxe?CaT{bK8py`;`flAo(ngKxP_=sVa`_DFNI^GunyM%-#WL{m z(Z*5ID?Vz<72C{bmZ*uiKVZK&SeB7m*5vOW4r&u^ccFO87o;4j9Z6s=@tr(^N;(-_ zda25h0EzYpnO#-&umOrf>lCI>oZX+4qB3hg2S`41KuGd&%7+{&peY7KC2um{W55C+ zh-%}qRM$Vwu|vHBO+S()zD0Pbk=_q;y`q|@-0E?=Xop{9{D{lKb5p2b26>CwGXDLmw`Z*}uE1oCD)06t}DZzdhRK&zZDk1x}-Q#w#-vF{?HD=+VcLuj1Bz%6`Uu6o1r7_C(Ud!7Y9 zAqwC7qV4LeqyM))CD*QNv zGGd87J0-mwQJmA#rhx4bR;^Sb6Ap@-Pg9!<=g#MjLEA4k-Z~6K^axs{v2fPE6lhi5 z$m8w)Ow}EZnV1G9m^Pc~gNK|9tEq|ZF78=`F|*Z9Ua`^8ji2EB{)?F=LHwBydr1i$ zDrs+{B`<Ym|&kSnY{@H!|k1RH=)UYVLFy!>FrB;9uT$8eli8&!tPM)T}P_S^%$0}$n zhtK$$nwrU!&6>UkkdT7HAlQXOSE=@X7bk@dbH@%}Lx=pHoJfr9Wr%MV7*t&dq53s| z03ra&tk2Pt%(#(NbNh}^)22;hHkauCVSgr(N54a#mLa{Q0$s$~ttrl4;-(l|d>39A z%*FzqRsc6dI+P9W12?qTDrWViF^JIi=Dhv6{G$HiP9Gl(CLR1<;yj^RL3(eDv|ewZ zD;Fx?FUMy3xKPXI4>2Rd67w)4b8uI1?TYpwnDKW)3|BjvIOUqYru&60Uu+J zT}W8k`NKR(Hx*9JT=gHM3-iZb0}u7T7m{%&ttmh;SVe6Ud8>lbhP7VMNb-ZeCdVpN z^`?ptZ;UR|R0JT)|LYW#`0z22Nm@;KxlG<>Lk;50!OP`9%j^C0Hzp(Cb<-n9xPP1G zbsRM{FoY%ErER|aiZwR;}!GA){)w$d)$MFV;A^rCXtPGow)EAQ6C`RON>Ki!9&2hLGn_$f)(5&dt! zAM^W*$7vU&G%N@&HX%nk`5eXyZEPK=0n`e>>=iua`Op@yo+q0r3_&(il(t?`eWOGDxBLQ{@leH9 znIpfxf?#%)cBM4_GW;{M2m`7F^O`5e?a&LD=VzMaUW)IhTpi%FO{;UVZG4!?n-zTS7!xsYV9w{e=at|0A(Z%~d_B}%IXV>j z!1Jlqy#B|ucEf#V7W)q-}5u3#)75y&_m4j!CZokA8}DOgdbveUZE-6tII8!zjeZ6&&2Zs zRp5N9HJZD*m%^4qeqGX@lP*_({eJ$vF?4Dw-EXcQ4TN5JRZw#kyRGP~C!b+_0H(w} z9?oC@TR-ISR;EUP=HsTz?YAL2ijFn%eHY;L>j>TjY{*^<*l*;Ff++u z`lc3)@s@g9uAVq5m)y55wpZw7qyekH{X@EEk1t($CGc>K>ynf95Cuv= z*894^7e#>`249zEOsQMvyuQl0;&J^b*RG;t`CtKbywIw_cih~#xX7f`oXtG@^7m}h zjK%Y|(>+C}ZYr|Mu^xE7(Iw>XO96!>7iH21)%6t8ooIh@eR(%7i6vR{Nut=Fi zMP)iTu=bRniAAqUeU%$A-|FrCsiOd`q}I%{q?Tefazqvm!umM4A0sQ1Dg@TTt>SG$YmS`3nUK-F0%%P;6O|qS(VScK;ir zLr(wARIvRKoALJ9Ea|9rzP*|u)V%sHEo|!V=o|T!pG{fKomAo`usaNpL30=wc6X;s z#tzE4(A##}%)MEw=RjwDise&HJEIn8nDMstP6W`2IvuayEDDR($+ zah!BP?U?>R>6#BT;#&S@onJ<8HP}gRVOavlr+D#js#upA$6zubsnLj zs-oUI`@j0&5o?9^>oIE?zwvtL>c|rq;@mo%=5GL2I5w31_$-_3#Hn=hGF%dd$cS1t z@1Nu+mj{>Zk%6p_=2u+GE7w z=w^{nx@k!qz=XA3@8$P$@ZXe)pZbMtn>K(XcI zPb{z#QluO_iA+AWAvf=42A$;Wk_mNv>7>cnhG5b#q41ZvbylqX`D4WgZ2ljXeN_?^jn>)UQHy3rorF(2!h-b(3lboj zwfhKt`;no1`qLOUy7P>fxY^)+_TZaM7>^Np$J@o4rAgSwQqIRVY%Gl<*gkn#09H?Y_m25H-Bq0=G*gE*VmgRGhKvl7JxUbe{tHM`f>GDX_wh`rZ!jucC_>dHofXGW6Pfo$Qv$ zw3u!uAp1qvX(Ce@{y{6_*!dRhOaWs4&4$pI{KpZzQg6)*{Boy_u;Bdi>gvbe@xqzy zk?qB-gMK$2>>WARhrMJjXXfOTtmu`zifmz-cba^-$nYs$c1L+BZ!pc`O#S4F=xFGI zYJDx?pLvhY?GR?m5l6x!Vf+(tDXK+q04&(rI!{US-a>&xE0V9=&@a&$JrkJ4w;S>AYgu>T|FaPTvw+k&$yMopb$wn!JB4D8qrk zBw?Tr<`51nURj`K&G_o`EIbVEhV}bE=^tsMicD9GXBs&ACK)Gpa0{=&g1@;ot(?!{SM~3Q1W}migU@K5krL?(Efg+r2<(d#hDfTB;{pz z`NeAYI9FK3`PkT>UN>CmO#CLr;t;(Gs{|;7Ri09Q5;aJ`dy;VIthRj!U! zzxxxMWuu3()4_LQVwpF*zJ;$Sn!~$(`$7ty4cr<$eJLPo`YC)mXr?P*0M6VN{F}pr zfIqm%4C9q=!Ansb@67z=aMm9|T+!ETz`HmDyU+8ns0`9RKe-Zg_Dk{@#MigBPNCec zhUL%rE_IFfdZe$KB|;V^VW*?|pcN})YunN?VwbhE+_p&jF}8fEZAT$(fAZ|!qRGkG zMWE^i_BXfajhhjJ=b3d_0)d5B6SEu(LB=%vtLP#}|4sOaB&x%HfHE^OH8ip%T#MhZ1edsFRVj{3nedhA1X>;>>~VUm_P z^!y#*@Q?KvolN9i{q?&uuGSaO-Ukx09RU`LA=Omdj8yV~aFo$0C{#^=0tb*`uG_fsNlAWh`O@|`MO$o}E(CxhVP62@{9wG%RFwFvnQOdIUT zd9tnzG4Gc5N}x31pt^`NJWj)a3o^>mXxZ`SzSit~1IJ>BTd4{#uDr^m-nt=Gi-?MM zgA7CM3Q_Z6Nv^0>GZf}qrTPq@rkS$BC{1nvX*T@N6}3R{Gr*y%LTBUmxDU_vkNS8+ zvpPmIC%3nXa15PW7AB3v2A4K0v66?PQMjVMuME89=8iiRb?sYZ;DC`X`ksxPU_A!l z0p7IhQ$Fo+5F)!GyD97GFf4Li!b(XeB}w6PRC|ZH(}DAH@?D-r{U56GO&j{cU;1MX zo(g;Swmf?LP*D1NDUICVs8BUkb#puK?hrAOjO8y{%k9kJZFf>t2YA|HWc zo>C;BAE_;8S){k8P;vE2oz_#BEOq2SAN*zrO7o2)&EahK7qVFbosLfjgOP7sR_Si|II#PKZ6&cV zI6~w1L=C3r(K)SL9KnPz@@%5aZMXxrASBcRnD-g=`1E)VOS-pE(t6J{QrfgYehFPE zHw7W?>@V^!?fVJ$L6tUg|9p8Z7NfN)ohg}m*B}zZj&{qrcJmb|-H|ZVchKSZ_UdF= zWJm4IzMbQwyF5I7MXJb1)lhsU0Tb!kHga^HR{tjmzs!H*;KOw$$p?%x08b{veD1G* zS=;1|ml;9R^j?jY@D6*86vJ({4`y&7zR@?!9kr+ZRo3w&IW60-bJB4-!S)%a5w4tB z5q9WHXYaaVHP4603t=o}H!FM@o%e@Ad|!pV9+cL=$`1a?W6aiNCxnZ)h2|bW0@;`% z6b1By)y?_ET4Vn>sJVGa6&FP>vn|!(MZJ5LAl3V6&dET!PRW%aI_e>l<~z^1see`d zSlDhJt>LmGnl%j3yev0NrDusD_cjxHUC`W%?d(CxfD&h0wDPtMqrsx^=htWVkzuY< zx4x2g;avkAoo>#}diS17fE2#_%)XyoP@Lz;R8Un@w%b)_dH3`fXuR`5Cmma1JgVE_ zrd>lT@+VX@G>l>_G;QU-giw~(-CC3X{280N@p)JjFq&wV-@YC zxFJW!^mY z=yaHmDMw%AZEc}V_t4ydB*YFDHu&wQm@cp`6xmyhO7 zAaGFbapvl%>SqMwfS!W{D9?h|oGu!i>cDMK|9WRHN$EcAAIfFN^$5^6ZZVvon zZm+#~S*L(#+Hw$=2irt{30;rO6lQ`DlMddQzUTbmT{3nh_dfCy_+pUGZId$_l&O8~*t z#olcbL_aL#^d+WyvWKdz)RsnuYGmUEeSz1y!Uov)-h~ryOg8H50ALb(m9BGoZRV3; zR+HDo-LvLtBp-RO-y236s(6Su$+SKm4@)%}N^tC5D9-$|kD=c=jyUX2(4AVjicfAHSf z@0|w@8>8BYGvyfuj7ks*LzDEsj5lSgej zwS8W-ea7aNxpRO)+D(`zD3-kyau!}LoQ?soN|m>FjolQ>8+?m)%Gi+MQpA?Nu^%7s!kt~Xtkvw6?8&|Vppj1CFc_B5zSpeQjgZ^<<;LUsB(c2)+!5#|Y1)MJH2fx0VHuz3~<21sjf zmCzf^>=M<9c6jhd3wD0jh4=ggt{+oPRRZ~$u(I|~*K|Y-P?2(!fHP0;l4BEbhaS|J zSt)uDU%h4aTx|Zmn?g*i99PsK|AN`tV>rMt02^4uPyT9?j_nS^`sfww+Z!xTD2`0s z6}~s512zEj&j29`yQC!v(hq4R^7Oz`@CZFsyjAy~6`xWUtF59QE@I6gUWYvT<-t4h zsjINiD|}z@=csuN+?!=~@tFyA1*+tC434;v57{{#^49O#L$}|$`^)I?sohOPj{WkxdwL;MUVfdeJ^X~$%#6%O4 z7ut*AZ%%Zo_mup5FSw}oIm98hi*+V!D2wDA^!nM_nwJbiTiBn@Q+IcnDh)>do01i} zl0!#cb_9XJ2Nn@iHFaqLpTVs;4prkzSPgCONL#(`U$@rzyrbF zjf+ffFD8k7h-_ZE!1$Owz?>#^vH@_0HV2o4eYS#i5owN#8*8`y>U*s=(>e~Sm*p~# znu3t=XM5b(i6Hr%ZuzqVE@#a!xwPYhw}z|J;-^1toP^WPc9W35jlesK3^UHI76@{m z%-}YV3&1gE#r}cl#~O}B<$7qRbR9Fc57V{P5jcVilxGVY#!Fp|-Qag}J(SZQUCNG@ z)yah7A!H1B`ywV3b#)iujziflJ2$F+J=?%!e2(d}aZM;YatAMd=%_>=wZj0&&at+swopTf0G?Lt%G=LVR551bYQ{TCLr zR4gv~?Ed>{z8e8fIVIVV(gm6M{A{LPxUOwt+sb|9YS)akP6C!wWiH<&MWKio+oAH4B7-9xu-3EIiC*??*JA z{+xd?q>exM(0JX-{PU{1+=Br&Cv6-`u3Q2;cMx?p1d92mO6j8O*&Yx7V=xS?rHk#K!Qxjfsj=V25(l5v#d-^w&1*%%a5rWzd^U#C7TQW zYrm^PS-Q`8)%7Mhp9$4XiNqa57q7=v>ZMk_?)lSGS3+7mn2XX48U+QQcp${~V21;U z%e}?V^=&+jVPFQ{PGY7TFg%JLmf{5ZgJz z3nVhK_7OKrfTkjEe?#u2yx^`DUFbC)b=h!PZ?Zb)Q`8=Bx|#XFDmA-6TD1op@);E{ zI+Ux+Yp5EuWH?dNs08VH;3wKLYa+zB67l1w}upmelKBVNVaXc}|d_N7t{nzn(kZyDMaFXwFhI*0W&JlRbwDe^Rf z;7AYqW}zy;GuMwvaFb7yoTh^m8g0RKzv+?2^PGGDp53jPfpt#!QEHKhL!(n z+D&7}u0RStk3aC9lv=#!;X8I|hn?)jymsuVB)h#R*&5-!sR=9oCAq zHQUzrc#Atbc7huv*pmw{W9e zhgVa^xg+~qP{b|}(O3I*yzSg+y_yk_ z%Zcdn(tSHtJK`VS*wy1`H-x=@z;c`>b0P6Hy3Tl4Z8N#jICBAS6KLN2pIm?}S|xd| z0S-}kf*i#!eoack{iADVm{vOVMOS%#z~C&fHcHNTN@&K3X_p>e_VtG{xl6NH7uL7z z2STg*v*wUh6JOY~dPQ#reyxU~_`UR%;kD_dSJ&;pCjh5VdsZ?KSmDiGOi2C-S~xg) z?A_w2U3GCDPx|({2LmCWs^BbEAe&m}H+aW`Rg``!yjVG7cHoz+U}apVJ)>_C#J=_L z&pg38BQc0U`GB@Q8WM!i4tWlYDKjqLa~DiU{U0X-}o zAxs_*HL`XveQhs+-q%!y$vPX>Iwu_ikS-r%PHF9TZLAp;$r?|PvlE*28mvJXIt<-v zhy{+|eU9YKm|H&Qc{XQxzg5`~)&5sq$+R3%OM}KIK$iUMe-&42?*3wig%(h4Q7iq% za9PTJRd5!=*QvHYp}n&FKFV39l)anyr!ZL>D^)CZ=T}Heois%u|614YI-QIizofg2 z@=hhYPW0iO5g(Q`O*a%3zCtaWev33EBN&(7t|=@@iQCO7pznZW8TAuQl^4WLhOJMB zA^Rh_F$A+vPswMLrV;`FdjB}8Y?;jh8ZV39Z_@a!yVn0YMO~cO%?tgU5`8ltH6uY* zPFwJnvMGSu8`mteuUR~HnOCEH=w7_$jxVmR7R+qfz|3j|p2 zzcE~8_N!~3GRS3y|0D?cWwSNvhFN=%Aj@2XKab$R3h3I;?D*>Jor@H~`~F_C|JJ%> z1y=RuP;@6eX~!F{W)|=g3Zr}NqB!W5@SgF}B2^1(TD!CVPq{rfDU#%#i7b_Hk~#CX zIYBS6WeP21(&RG3<~}YLG!ldfg4=T!}l3l`^IvYA7WzWAW6AQ@r%s&Q!n?^o4j!D4X44n;MvNYSNN+#^(I>YT+jOMOa#GP-A!cME4%NuiF3&+Kig5~8?RuQb#S`r}n7iC4 z$Pe}Z-`1_HOYMx)I-mTT(5}FSY&9V3**rn}>W4cn@TMUjpS+1tDShzedoHu-4czSRH6V$~6AEUWa<|tL6rV3GXq(5K;VgK@ zBb5nofR*>p3Lrll+tH<`%shsln?;#^P;>gp3>PNq+&tkh*tonaFT)oM{9K_>D&1T9 z@B+$>y}!y1$B;U$9K6Wk%(g$z&AWS#4l(x;OzI<{elph+*@hpm+ksfe05OsX5|fhH z$tTQPV0S5q+7Fi@H+lsbh9O|ik<3?Z98FCbbyb!+3i*?d!;{(G3;JKvlBxUqFx$q> zc|q_qFLGhv4wct1A5b{2`VqHmnxV2z2mQKH+A0YaDyN%0A`Nmyyy1%Zuv~F}dH+um zuHUf$;3_v@cFTG)aEUc@9SBX6D0OZ=YAytC=9aVsQ&lR|4IT`AcJDgy{d&zD&L5bf zV|SmbDF|qftnkmXJDuAQAlU`4jRL2nikg(8_4FYHxc)CJF4xWn5NCT|V0HWk6fxdg z^frO=4k2W`_l;%MrsFKC<3I=PuYu{%!31N4GIY-;%j@ZEM$4P*xBD+G0fs@>RpF`@zs-rW zjZEixR7sPQu;obwW@%h%=dJ@Z=oG?aqg*rm-tjuNa-IyNaAY^k9hTL0V-g;nhTxV; z)f!cq#?%9sdIM~{goPhE`Ix^wQMKEa&5j);7|d2n`K%P@{F!=jjhP5Gf`Inu4Hy=% z7s8HXz&}}NeMy&Y)3C}TYa3>BU5ncRQ(G2jb2=^xelh=~8Kqn(nAZ1+R#cFu7Rqa6 z?Y;P`o`+8iRLvNDLd6K|@7&pF(4;KL24)d6tNk%u2VrU%Nt$@J&6^`YNJZKuD zk4~5jwmW3_4})#T9)0dccn@$creAD1%G*p`ejp1gR}^)ERS-p*y_oS^_@x{UbNa|Z z-R1N{+>CQQZ2K8!_MH{$h`{J}WTNL!hCl8brl|C~6WD2Ej&=6Gx_t$h(H}doYo+`& z)-s|T=9mflEl>XKP3@N3c(aLl`n`9WKeT-UTQ#&CCI-fVXqX$7VXYPx$p}c$X4t=c zW&-1VJCn`M)<;SnsLaNsVV2#KG&u_4kH-)S3A;7y9bWJc6^>iaQVM1Z41#d+#<~&x z1!gIHkIjsN$PvStu}7#K2!tVMvBQOSBww7a`!3n})OTkaAU;~{6}{$m1zf47JeUB~ zwa3lTnf_GYy`A!qn>omh+WkBOGMg$8W!pd|&65KF7x0eiRshV3z;e&eJrh^JOc;7H z<>g$z!T~weM;*`3pOwxRDK|3iSJ_YUuS~m%8u=Ib+xbuthD285wE_(t0~rEGYNbUI z43L6Mi|FcG^SyqCEIy1(Bf-)h$2KpfHH(L$<_9HC%YN`%Vp*ZV5t<G{MOqwr{=j-wp`UhZnUT}1f0Sj+H_PhutAGL+Zx4jw+3-gq9Vnfnwb&D-iPDO(Z!~=(COJ--^3(#E|9p-L)iBU3Cn!d^>`)h z@MFugcdLp@myZ(tM&3@=7GiD4TfsX63d^5&4aj(dq<>G zML-T~1(1DH2Qcy-`;4Kd*5cK?oer5Zd<`V0@1{Sh#EA3N2z)gA6euQ~y$?v!ynO-7YKmo)rjSv1FJI-+!mCxu?Pv zH_gyzO#i8NxLV!^FXMlbdDLer0m7>n<1KMEE>Em&rWUJGLS92J0Ld?Vnfo1qV%ACqM$tEp(=@B_joz zj+gQc7ks4i@?KT;bwu_{a=ee;aw+8aPT_A#OIGyxsF@tS29*=}#wUO$rq7?I`#h0P zUuPdgWspT6kU*{`Lj&(8ff0_B%>WjEw&}Frt;}3L=0i+Qv~;#MDtCJ`f%HYSEm`hx zq1Z-`nNy}Io|G4ayW}ppJAJrF9h^|kvUywe&@gG}E?A+>bWj6w8eUU6KNx%hv9(Q< ziX&a7U1U2mJ^AFnh%fF>lWh3c?>a%8ohMrH#?LC0*+5yxZpFEuHF_@lAfU4fUZdD+ zjS&anz3Amfr(LWl z!?Fd7X=4tLjoQsHt<4W|OL>^)TQnU^fLEA#gV#ASyUr~ zEYAun{=kLG@(j2$UYV`B+zJ3l4OPmk2}YZ_Z&emu&S%F215v?&kN%DS3m!B~9%5`7 z(l9T}OZE%F4fUkTtppJFa%Zw*f4R7zv*{a2tHDc#+N$y`#+*+3slGp8Esp#POi1U( zcV+1*32*q@)G}G7sJ-kWF3$rs0w7TLYnUE$-97BD0f}fLR*x*}M_*SbD(rMr)m4`_ zHkUWXzy3-BXx7aGcST~xDHDTvZR44Em&Z5!plRMJ0*OsIiih|WppK*PC_b|d@StDLWYGY$oO9?#<-vc1aA zE{Ma?0(=g>RF%cVQD6rQmu!q=7YR5sYAF1IOni4g7z=<3Nf)xHfEc?ohX45axPnT_ z8@h}8T?-R%`Q=|&U?68asnJLAuLW>A@q^yIsQ|qF^G>~)N>~{;Rujk}=Bo{%kgf5r zzs$0GJR-L89^xjGAuVu&-`y8UjPv9P|kmf)xTo&JXe4SAm^gE{lv|^-l!7d>@zPlRr zZ!kH(PM!3i9IpS{MSEI0urhNZc-3e^ZRC?S_aNR_mRmcgvHJtdG(4QCx@k%!yx2^Q zznkQhZRXT!Dzu;W=SebV6jXe)v@X}ssQ0F#ck9Km!sGNIihK4gGP2KT;rV5oYr*V9 z_Q7?bSiC6wj$7NLS(;eSF&bCElTpZRj!n$$1+>k9Af%t>^ort%sXYaB@ji@phbTPs z!G0jP4epGV8;R-l#P8bkpHJ-q$yDCsLBt%$S82po= z->cijjz*fzzrzihLh%PHt*ES-G^I`sxq`(=TA&MFSMgyEBJcxm^N`RcsjS%by$#Wm z2;^O8@bZ=gPtdiDTUIB!o_B|221x*`c>>rZxs@M_Di}+NKF-8klr5pLHn|}B;!MYp zgRn@mM0nM*dovg~H_2ybv7j}5aT{$3W^7Y1U=M4!e^s&RFVB#e0d+njDo3m}z$pZp zt+PHlF&;2VGe6s>P{Je9T;W-pRT_%taL#+CJt%iGHj!#QIZBQ%hs(*?Qji_%D0e-~ z=W4s_lO;wWKxftUj!~3NSNkh1Wm{twtdBlYN5o*$*j+f?;1#hJU@R~13;4*oi~rG_+%#qO*wlGJ zi&?)8aWICoBqvd28xs%QAhGYpa*D9y;z^(&GdJ~Vq|Ao<=HFPCbLaJY172h*xI9^S z=RDuOwHA3;T5KX+b@PJg(jCciKDW=GTdr$xp7=o@bPxcvkv0z$9Ya}HpVBZ_6XktaCu?}g zqb7!3jhW;sEU%3hK2C(4+fYvdA5ku1x}I25s+t|CK%4~r=6U9^A%A0tZ-kh?eaBe>9zSSX2Ms#_13c5y=q( z3QFhb77%GE>5`HL>29fk(hZW*F&ajL(jB8gnvv4=oPD3`_b(UMU^_dX^N#y|-BQ+F z71(yv`6uFy+HS5JU6uddvWt9z|6@C{$wtZC{u7)ic|9$6o!IZc88yb|x~>opeC{S@;R z2oDpD?i-^GKr#IjXuD)~BFZiJORJ_6FOT9M?UX*>3$b0fhknyO&$sq9i8!+(yQUC$ zPF7fO6G42Kmp+pv6RhmGTaf&Lo6HL2LUs$Vii3LzH9L|}=mIr^hWnY9xyzaSO0P3wnAn#k3= z(><#%%Reawu!ssEVDgGCzrSGZ-a(zv2DS&{6Emx!qn=yhY1gzuVPUnbm-dfZJ41xA zn>HZ0aW?HO2OI6*d+>ol`r8{Yrx>lv5Eq}waCQ<5w}|#k@1fM-n}YXu*ve>;&Ww}i zlrWTs=DMJ7W{7=s2{j&aLRqi zl-8rRrKJDizA|_i?P4nS0}2^+vflRFmiCrICTi;5mQ<&z6wB=Is!Osd)f0^5wa=}l z?ES83b$@E^Y%=HF zGIVrve6)1fopP-_x*J9-a@I69R6E896NrOq-?!Xca>aToYKB#Zb)CIcAEJ~h^_hJ< zJ{dzBmPiMWo-RZ0fV%buxAub~`r>gIf0;m0*_I(%IuX(WRaz5Jt*|nd4d$UqXVSs0 z%~SkiWB6?GOxsBWR)RT#cZ>J1RA2i=g#6?BsS!ypHoU4a2{+|x81JL~ii!ZN5nsqI z<4QA3jD1X>3fsskJ&-k6TiEJ+0Zm`%^uzo2SD#N*a8`>~S>1~hsPoxPq92MNCktQC z3n(5wIWFM3A57&)w>*dGqSChd14r|LAwUPxPIsRW^s~-sMy{Jz%+w=a&Y;K=AMem} z@=Da*J5iM%9021UcZ5Fkxp(`e_fnMMzn>H{4<#zOxZhU7r2DV<)}sAz6tn;3GfiJS zx`5+U)0RhpyEwaw*jKMa7-T z{?DxKW0AD2Ij>!aYQqvF1;^+%<+GnB_qKCpt%+Ophv_7HtHH>8x_Iua@z#{A$sX_& z;swsnvO(C;-6MdEFzr_Aut6A7Rf`7bZ?1mD3tePxuS+baY6>nK&0z=wir&3ObLn2p zm;0$7yIgS3j+H^I?C;dwKrzt8O=*vGC7cxsIhJ&DLO_DqQaix7iHsI3Cp(j>F*6C6Q~ppLP{;bB@!2(P{X zauuAY_i;>4N^sQwv8{sEpFgFY409(YR2KipZvMCDy3%kbk@u|7KxPXm?oBp(dHv(B;Kp!wi|GON7cNk^@VR??5X_b=mfskJwDP z)IPafM&3AW*AJFsk>XDskkGb$?+_I%+4`@ntd>qM-kz(mn_WrXFA3wZ z?@Lx1CbnC%9a4|4W?1P~M);*UPP%>tZ{H*exB_!n@z@(cR*#477=F0c4L4keqO&9D zVTwJm=;;$~_9BGq!j25Vxlw zV;6EqI1b_KeEu*843D5;t)Ryk$BO%TXea-Th(+EJ{%BEb;6EOtF0wHmzx5jthLbz7 zRU?ldf=%%@eda~7){~=X6ExTQd;Kc`Be{{mpnbI5BC9ep~%nzXYm)GtqxL z=$na1J^quY20z>Yy1>lvxL>pN1L4Y{Ke`gyS^~7@r+L1Z_x8__3_?mdv$5w_i)K=&`}b_ts(YybuvuP zZ#A~+ZvNV@cMi}LiH9<=Sax}i&>EkSSINt`L~+s{LhAtIcAMF46<5|pbBqg_*>m=> z`oruWC$45NV%#fknAFu#$S`ck$`&>yJs@wIgW$P&VM1;A?4c^F-H>8vaCllpU@gQ; zH?k(UiQYWoLT|sVed)??`JusS59{V+;eJ_pc>(CJvIXbEXBAX(O?Wn{CT$S{RIp`^ zO4$32X{uFR6%Q#u3&z^NwT*jLIIPuY;vuyDjbZO*2EG%@;bWpn!`=T!;5Pg5%)i2g zC9RaVqgJpjk&oyGIAE#t%;~mJTzVwN(dcriNm^)JSbL9m<;`+9JBhRzsq%h~^URPC zb8b8ax4Jyi>JQPC(?=wl3*8 z@YJSPeY{U6N_|FwqYXOo(wx0tP}Y20fZ1*5d0ny*1Q9(T<)XpMb>RhO|7it-46GVj zYmj1X{9yIXaC-oB_P(0VPm6&Bba zM)~?3Fy2`#R^*l4+?r9-`8(@Kv{i9VXV;+kK^Bg-t+PKc{!2}t=$}^ei}+_YPM%;J zadX8cK-06Zy~=%WNBEh*zdj^zHEw0Sf#4{La4XgABUo$3!;$DYWT5bqU9dclbvv6G z$bKLC`m-XJ)A$pbM7#`HBpU}2YaZ;V!FZlmb4*F1G*ics+Hxm>Rd^dhh%scLIllok?a!sezer;G}09*bBv~&A%Zh zv8&ca5c=*&=h3fHCaPm3a$GtECiRvLB9CBllHr=0PwBTv(}ovd>k(!ya+H-RZJH?E z?DQ2|Jk|19Tr8MNS>O{P=XmMBA8PfuhO@{-n5VYF1TT_vht<5KH9Ev zlV}g@Fn1uJuZ*T8W*G9T?zbLwIKR5~zp9Z5@$x^>*riVvD8kpSl7Mx4G0uCrUg#3# z9<<4s(oF+uSjw97re%stA$|G4Lq8#&*QxVTuKl<}eQ z5JX5+%!Ha?a~*j8WseXL&{m@ABz$dmp1CAN5`XhkN0TgEruNrU@Y{45v0Njre`3u2 z2G84)9VB%ix*DAbL&8V4^*}uGNHdF7ilyl_Eu}__=AoMAAE=Q4r`6-|%xypKOqk7J(tK<|bwL z{CTpBv0z#&X|iSQg?e8?(^2Vyefn6#SN(s-TGxi+qHBQwqulljCci?uH>E54z z-Kf2WNQ-xAhV+w+bfZ0Vr+ahmhSHCn4R|O@pd3b<%itB+XT5@*;~wH@N6zSjM8S$E z>JWbIn@?9HcJS(5ExL&??QzZuvw~}Bl!xKhtb4~xq|jN@qRRS{aPhk}R~`fd?e*{C z6iRX;#~k(QdtFbBZ1-#0hdu{So2GwN6Fq%SAn{m7y*?*bau3G^`g*c@3Tm498+fNc z6~$N~F}rTLYwJ-FQ^!%!`HJi?N>pfIW^^n@_!`Sj>`J;>n#$p;mGah>^s2T*rz?Ph zTk0dQxw0n^?O=};iCbL<`@&;hdt}FWBqLv&>GWyo&8PXVTejdA>h`V>bDp}nHR{A$ z*)PJQ9(?w!HvP?RWo#DL%m!r*@GH-d1#U>G^xo{npiJ()>}D%`-Eqyh7OOsQuG~?- z?4|om_6o@g3kIBIGewS$%0$Ayzc-n}`L9U}jA7QcMZj>gIFaEHIjz}KBMNJ3NI$Yz zN$0owM+Qdo!bSE#iqW+$$4yTHw??K&U^}MmK(eE`Rh9Q}Qi-+^+liEtKg25g% zF^~IzJn4_|@DJbjaxz+|$96s$1_u2y6~v`K zsky!TZe8~J7Fs=_q#fn84ej@c79;?QCU{XfOV87if4q>@dH>|026qmL4mKc~DW50o z#*UJxhi|cTwS3WfCrbFnSC|oJDl`J+t{Laz%Lf}G?7~p*GjmkK?dL;jjR`@S zymD#UxW^mFp20vTUsJD_YyLS^M>|8z@d#8}_=wc)dw7%<)Cnv=P0cUP5_C2Y8{=}t z?{(oh?bXV=axvm%y+yQg(dMSOJIk_}U3m1KkLJMQQU)7gN%cah!fAZT?-tRZ4FEy= z=4UK!eiH6{3ui^=T>43UoJ_8)*E&Z-gU?G9f$~l zqzbv1U%u{vSk5<=S(px;lo3Zcmaq5DEmmX>MAQk>n*!SBTSRaRrmS#<#u8)xJUj;bLc`@xz_w-vPFTsD`zn;n6O6fEF`n4#` zl{4O*q7;6@V&~b$HSN-2#4#+gu5RzAmp#Qb8k}c+4Y!(kw{@!?e45)7$WUBhI-M8> zZg3tt>+TiPEq`SG0uw6v_RwoTDk5K{y22h*p#~wyKYEdFn^zNOQ0!$}P{nIFnE&(3 z-vmi31GZnLA+29N3J!ePt|D~1py~{1cTy$IKA&q@l1=N?!<4g$WXs~qy13gG_b=69 z)%T+8=QXr`lI_$-v7fY}X4X=FMMV%Q!Am}1?wrT1BmXbW$hI#ga;$V>xq4PCXC6kS zyT{Zkdn`7IV2QOB+n=?Zc?8-mm5RFAax1UV=8P7EH`LhLE)1K8pqpx-hSq{t+f z-wWO_EX0&5a{gmD{fP~23^*OcZt5O%7#qEWG$Ct^Qi%&asrQ$bzmX z4|GV3Zwgp{vSb$o3Or2-=?sq=vmtbO3hlEROd7P^lgD~Iv17!JBxwJDz4a>c{vTy> zEB|^dMCoU!5B~?@*zI5aBKK6|Y*G!Lq-OLhvfVYPEq$PIn5~C4+m#covFRHhURb=+ zMgXb1V&$te!Ez-_gRRNr{596@YD zF8tIxJjtn=O~}oFkA1Xf4U^EL1AWG7<8OMS zIX1j^Cdh-tMR6nG?dvnyjVj{TY@aFx%zF=VJXM>D@<&=)H(o^&XYD*xxVM%1mWT0E zN_VH1|N8IEcxzhq5#0bTULO9*pr$XdA6D2OFp2fw`fKjDeSDW?Y#xOeD*3)Xf#34I zLuoX+XhA-GPk9()zdEsuolTct=Dh}=p1~HkGJ}*~bRR&){kDoJ8te#IZcPHmD+BNxdc96 z$8`&LrMk(FYK0Zm2P?dZ5)RIVVm@8HJA+<=)!d|K&l-NXwTTo`tRy+$a6H924(BGS z(9MPyE8Go?w?q1p@R$Y)d)759@Lm_hto3={C5Dv_mRr%B=dw2LZN7mrMePQ*@fuqB ztnOmJu=^xnvmJEW1+t1>C4heE_sx48zALKJAr(27Dy~^Qh&-;m1%NXd{R;aVzj!}K zCqp}|9w6w+y(EE}+95jzf|SikJ-wEI0{fC#RDV5iyE!hnK6jzlnky00MEl?xF%AR1 zJvOv&0%ImdHazuUbgeG za)hynWV;)s?YzoJOn3J)cRs8kyt+VSMw}=ctekd{Aw*TM!a(J9 zBj;d5JJPb3$bV_0tyny*&Ttb={CP%@?aCo~hChmX%5b9!%HHvp+udPhfaVW(R(fwxf?YB~Xji9_t1`sT@MhNnp! z2$q&@UKAwvpV?x0LsGR06v+us!>7LPDh*}NuO`sjhbO<#t)S_}-o;@Vm9B`h?{ zh0}QL6yuCsCN9lcmT&FPVX)8$TEPxmQ5*jSxXBT{tT^|iB-ng&vhzcMef}ETP3hQpSm+nHrYt%pM_-zRv3u}0G*x34>w~yhj)ON2trq?6(x>13=dw*usgeyy+ zO%-F1q4oT9kYi+Ri%V)q@|pd*xTmIr*NhS2!PFM(2N-ANYjsewuR-GG90C-cV_gyGG>=w zU8OSAnykz4G*K$8e4rW3AXQ2UlOROm&Tb7=AXH+E|70wN)D+zHPcKM}7)~1L=4kM*_isRJ z4UxqeK26k8d9%dAj<6l<-f?_S2SjG4m<51Oy%P%_N0=YNZ+|9g^mhW>e8<=*o3W9t z+?Zw4jYb;R1i#v6Yb+09z}3R$B{J zF-HU&Fbino;$&IdpWc2@h)v;BwSJlaNnQPL!^W=l996oYAqF-1djhVEsQr#9y&P7c zS+z{nT%rLKa&l(QD_=#CpP0E;r{FXYQa}^W;pSY5dR03jRLUZ?=XEgL>mQFMCyV5; zO%Gyh;$HS*sftqa{c>P+7BheTi34ux))6YINF)$;K;Kk@@{0^RVQHzt%GA{|%hRG@ zTzO9~Y%a4^tX0_9SQ@|aU}Mo`6xckpyJ`%n9hxSW(gcbiWo%?^J2KP%{)D9>CNLl! zf;SdJ^HwA4b3)5rtAW@R)Rt3(gF)3S>8$79ilYb&3V-vWPNJ}{@g@n{Yx4^1f4_`@ z)pBSgVt#1*tS0EwGWX-aZ-d`2&+utu<)s!vM=(@1HoW<+%Idu)rNGUOGuxKeK;P%h zQAK5ef&H$y|D1>bh6KbFAWrRncaj+uW%zhA>poP6noQD^Q)1Pp#Y4u9(0#}G6r|lC z^fcT7gcoI8nfBSZ#00*CgL= ztRUeK*y*WizbG%6r!_lm#KXbXXUD?wbI$6uMfjBOt$S$hi;lWAl&x5J#}TmN<^R+) zMEp(|lw;0gPo{IFx=r0AcX$6jkNY zhpV8dDx#BewjK?gaOFvT7!>I@SA2rrN3y}%fA~w7Vez|3?Js@$Pt%xiXZ7JlCuOnWX&HIDo$ z29=ksk;bS-5aDvow#Bf;CPgz{PFrUsyjp44zDdVxHH;XHip@7V?l-$d+QXQ5;*Ww- zvS@S87-cTwD>dkBvq%fN&7AHrs1;%7Pzu*ze_N|#EL}zwUWq40W`}rzMboN||MZ9uiPJH_ADi2bEI;-7 zm9DTHxm#T^XI=?UVISgE5eFbG`?5=~_)!}4wyfLa_NP-6bg+lMP$~NYGZyc!{(w_a zJru?+=VFqkByKZpn#K4>YovjZTgB-*mG})i6(0>TY(kmQCv%F*nml%UvGob+7M4t= z*YE?jFli|cavR9}bdFJPZDydgO;+l3-;y#{|Ej_3 z(w|TBlQA$*cxhLeWR}qLw>0DTrHM^0mh{TUr&}NLvhf6-Y<=UEhiw?U8Cl@ET_T|R zU&m$68}~(^4p9YBQsFbvdonrxf*j9mmVw?vZ zMM;g=S#{q|VzAK^lCYD#*DR-n|29YcBp&>jM)=g%xaL#%EdAf&`kA_O)Qi4=ynOiI z`0TsCi8FU_5Kk$Ix_xG4|3*|meqMxNhb`yoJJ}}3sx&=1LgP}=4&}YdyMaoL2iG1V zxv4tmEP{v6oE@-6Vg4bx&bkWM@hDXsP17eJQ^?<8ulQ!R!J1GusNJ0( zFiR}ebQ#Ku;d7CeyT_Un#L~~%nTSKij=AQUYzsaWL~7sx^{^1R>OX1s_YlYrke#T}8AG9-Pdwa4_=@%zy>_7>9IToF=3)zH{5nMMr+E zbmZ=;UmUg`Z2;!quPGQeM5#9jVPE&g7!zW+&FMgdLw{KJJgpKUH#Q2>Voc{DGx9k} zlU3z(%mXbIQ{L8ps{!>WZv*@yV2yAb*gWS>K{&Pz5!_1og75#V<^aE8*@7t|h68KQ zIuRfEm-)>)DsGVzvS!p9U|e4b&$k)e{~_k~)4Kb4PjjPbK&DAYs5*y7C!P@KRGA+2 z!qR%V3Rk#egcB%gRu)bFGrIuD5Hu{?kYgAHqgwVe+Ya=04aYU$6JuIt@zJJ`|4M zvQcVZ%XXpt#pJ5HI(bPl*)yT*kb}S~Y>XP$iM0qb9=`?F8)lpNRq?DQ|Dt7(+^y#i zTNxu65-U~y6I7VW=mD6#IIWBk0Xq-};Rz(kC!l9-y2&I!hdfF=EtvGRFN+!yc)CyY zwrdWltvqjBUVPF52zhR1+YKEk+@8Gh0L%3Vuet>wOgb#sYE7ZN|8IBzKDlvDV}idQ zy8tY1&hooQKp)}|%REQkHf(3FPNIQh>VK~7?`lrcd?7Ieer$X;Y+lf3?igdEpO?ny z0X!Q;hg|9-)GM2;ZUJ%Kr#6(duuGJ)&W-_Gqii76N^{z+o~`{jltUwC_BzMU-ETVP z06A1Ad2a+5cZgeb9NcHz-b`C?uDTrpzb>k26RM?!)qP(w zf4~nP`}4D{ci5tPIn;*#J0oo&?c$UuRp1gU>80Z3f9OyC z*^?2(%mSDCb%Fx(Y|Yi>Ax-9{^qrBPBx~FiONeHEKA7dcebr=-tg!hx3=_^@03p9%F(>|% z-ba9D__FVJWIy@)RjexP|B6ci*|5D!t`0CO0jb;;F72&u^+e0NfnEd1-)9)C>6}T9 zQ|5sO#O5Ez`X`V5n>dY6*94|Y4bMv?^GurF3)h2rCzS5&0-jwR6h?@3=Je7^vS$Cu zv7O839_r#$SPjAin+bvh2hnymY6K{`(aW%~h8V0E7w|#<=@X|r0;*$_R5!AmhQo&J zDPuWW4r-Xcy;lE?x$HY2U~+5Ieqm#io;OGz{ra025EI{ml&{<~6Fw%tK*v~BHWw-v z?6}~pa_-VbqZ>LwCRw$e_Al9b=oVYFZ{vhHEuC5>s zbAQ<0t+x5Ctgu#Qg-hl>RGyUC_UR?k!;~~{K|A~#u)BDiS2b+XduK0j#eXq|U2!y8(29PQlJ4g$BW+PHVxs+;|)#Mt@b`4@T z7_i$XbqblDM8lJjD`UfPaE?z`+~-k9K0?k{4x@QSb^zyC#uu7(Ku~VqM&itm^MIs_ zP09RNXsM&6)Y%BB#^^iPwKn3dy{g@LjJLwy%Rg{iw?3~yt$Wh%%@o)oGaQT(Fz#Q^ zf6imUm?U{SK=jI&gbEmuNekOkzKJJ+)V6Q&RsIUn7MU#wj;boIiui!ne-{~n2Ey@G zB|QIw_Ken|-k4~w-+}^_+XG-NeBywRE28~-+p7O)AQ6r&Y;9H>j|(wJt$$@5w{%h8 zV9=~CH7J#*GsV!FR_+x^0`xMdE_RXB%4tMaX{C+vJ>w5cPq!cgwpr$jrY)JAO@u>w zY9HwGUdFj{{fJpv!iSfh{#aZ6Gn3yfGt`o5D$D^&B|vQf$IQUbH~`qnp~!~8kKlS+ zB|<3T*AlL#7%%@Du#{|p-!!nN;CU-I{4^wGdb%LBFu78)T`d6hn!EG;!+Ph2PDnWlyG_jP9}NleZv*tS)x{+Q;>u~8)k zHcr4&Xj5#rTn5Um;>yqaYkqMSeLOfEC5;(SBRYr~kRi;gv4Ss!8Wer)JEwn7L+$zZ zd)`k$sO(u1HePb?LGfH9wJ1>3(UK5^6Fszna_0&wUEjcC^PFF}@{>BfnS8;idT3kx zl|y&Y^(k+t%FydUQ~$~t1zKujS1#3|^=UW9O7A5b*Y8!aX=FT>c{RJD%uLj>n^WCl z9}k!1ww{?YSZgSc#ze-sYJ^lIZrxWm+MdQnew^4T&Vx`U#%tx7D!6Ju3Eee&X->Y$ zP1Ph12Zd&W5X>R$?@fNwBJKAdyidc;gMPBs&(N;iZSkv=e-YOax?9YduBzG>{tspf zG7H$s#0#0*GXI<5Hc2E$#nncg783=veIAi^?@`@aQbTDMFi&q2SvBEOO^LbY7GIEyHy zfr0;-)b1i@YNN0YE*e~tWjfG?F-C|aI}}m%sy(<6B+0Sso%W8Tqo`&Ch)=r9SDZ0a zYo4^RpWUehxOy|XfY(4RWL$Jp0+c)Gl#c&d6RRl6jkUJooZmg3SNi)JTqX?n*pDlV zG`@MV^5lsA_UbQ__`L7GJlgUi%waa_eYc?n8N)>+n&gA=zF56!Os^suWo%JlYC{C0 zc^lT0xiA^rew|qQhru^g#F+eBCU98Y*aZ%+_F8F#Z_c-5VyZhNbQcV`9yq-7uq@@A zQyk_Ha+d^SkUf_`BtPACU6S1Uq?gPcGL+`w1B}E}D1waMfRe37S&RzTn8|vk zUq3|)C8Qg@j1)`pArRDx{fPj-@z)%eDx>FdF~Baf$7@*KS04x}x0=$}PF3P^HB-nC zRZ2I-fL?L>4P^qMzI-4?EqkxYbS=8FylQZ-DHH55)B0ZJ5#(_U2IZXo*J>}Tl#fC9W26j166w|)nvA;c9FZ^zB$YT=UbOBPh9zlv0C3)#@y z(4pyE5AB%_M|Y=^6)6U{-;w|JMb#nX@Q-2A>u!2tq=z7izeBBupi-jP-__wc9aQ(* ziw3X7g!Tp~l+Dq+_ThWKKt$SLWz3Z94SfCvQT7s+97(kdih~g^XRruRaV~)+c|qZ8 z4|5KIctT=Afz(h7t-*Fuf^BI}_pm^dT1U$(z5E6?RvdC`k@gMhqhgR;RH0oh5zH=< z#07o&HKHKdP?ygS*KE*fv>Ao9kiTSD2~Uch%eO<-Nmd=VaIM z;ZQ??JG|qNbyu|nc?lU~P>ZM*>X0^VC6#76s^`2Hc#uWBb-VxU56rsCaL z18WC^4zLzbx1G=UhFQ^cq&_A8nIo|H4V#9lz=8pdEz<1}!>MX}tjTca5f{~1`1aFc zF>9@t;lZA3M&9zO(Zic%IQW@Xd*ip`z0 z$W8|NTBzvhohUKF>A4VLBcQW2Xks$m<#_U1z_s%;duj;NlefIab`Us@YtWd0;&+6C z(alF!q*q$TI%Fn6!!}Nr_yjv%G~L=sIt>*)zP|(L>&pN7O!FF9mnjFGm=uziWptuU2r!vAxFFVFRHGFF(NG&A21ZX|= zg_h~l5KoL6$Y4c{#LP*7^DiBe$e|w)Bvg-;9&66M&k`48heE`vY9fej|QkS?Xo&E#!UxMSFUcWN+;33TMxYyr8CpBOEe!AhP zJGwgh97)TqPqS)KyZTTIAvGw?6@~~vPjSM*coUPKdS2dw2d$>77Is_GL$O6xQ8llo z3f-N;dY`j(p(c@`hoF$wxQSYuR~Ix#cg>1no;>Z3X-r|^?0kvxn@gulhw(|z{-7Xp zI$j%n+^~JspYVv^viEI9NisLFqZ8|SxfekHS0#Kl#6$M$N?HdSe+Uzuot)ERVb$|? zVa2fSlpnA?{w(q@DayPYD&Qu*E3WSAjnxmhATVL2<3qf{6YHlBjC$8NcS1@QSQb-F z75};4UdLp$bMV7_uHk=EI}rLz-ONG)Q#6MnVMcLJIPwBorije@UhnWK7luNF@D;bd7U0$Eca))KL`|=+8Ex%3?ST zS8+#vZiCy6CONV#XdHA58D{3r*0v2){qVo<23(&a{d~KCh`!gpRG63 zm50iLn+dv4@nk3iNMn%)`{k!s)A(W8p?xTlxp*eafCFz9^9X?2ssWXbuZ=z?HJ$6A z#Z9IU4l&rvLEo_Dweqz!nhR?E52*h=f0_=j*e%+;)o=JeD<;|lvggP4vSU9S4lS}* z!%0`X0W(A@-tL-Dl}cN>UcYL+uSr0_cWdIb*I@Zdrx=^)&16z=pFJfB?+Kgpv5wSS z_T8WCga00{d2Jh~&+0!W6HM(Qc^9s~Kt6vHf-q}`rOxO%m0#MZ!!dt&I6&mnu-VY@ zMLZ~q1*S}pcfY4e2f0}uJJ~zkONcseO`lUZM!YSuwWf|L_l@q?Ih&IJIJfL||J?XO9-jQG zfWy!Lpjb^`q5C4@dPch&YWpfuWnR#2MKEQ3a=`T4;>n&DMKVVeZ;_M&_Zc0_@+BcF z%WtIgY)>7L?iCD^kV;aTl>cIN4~b$u-fb6};5}_ofF3;++Q})&l|%X5DT`y%kITTD z1EeKi{_2`W5(bu#W35A=AU{kyJ|BI&8)YGGd%Qm~YFFrfm+E^)Qt2Tx{p2&#RxgL2 zQ{7UbVic=!X(E;1MF$G#J#ZL{Si3*pveEXdMZ@YYp(tW|8h~kCBZ|Qvs(7r`rDcqcmV!QqQLUq( zZayE#IbFy@_DzBmMjsawR%W>l%Y?KOpUx(roFuJ;*Zix!YfqV#cO%LU z#J$yw8V5&C?_~-ylSenSNRGuv3F(PU=9Bst>Q3XN3&_oPlT_8f+~=?#EN zboiXx(l@+;P{G*dcx9nqPZIScHBM8*RmkJ1e|!4)Q^2gNKYWUXIa)K#nQN9GYqs*z zJ{*VdV+H;Rl_?)ay9mO0wwUp2X<`|+8ie=d4iBIL7EBFMGR4)6n@mVrb*xSo61f|^Dzd+YWHFiY)4kVa~OHMQ+v zO<58%LT^4W6aD61!-`GtB&7~>@)&bQwSaYOl?zXPqYSG5P^D-ppVe{Oi3V)Qt5TdPoOUhCTya3?=A# zIWGP{E@_)SGD7qF7031Hq;}DiP1fe8DThf@zlr!P4YF%EC2A>dga-&MwSy~H(#mQw z044k7c(sev6%zWT$ugK9=GZa9c=iJ0lD7i6r*J7J;9$-ItJ~i%5|ro%XjS222eO(( zM0~>yr@qt)fR*U$kqp_+Xq_wXRV~%T)T*>qIv)&*UBH%YvhDgIZnqCL;qLq7t)OFN z^)153V?yC%*?GqTgo(=_Rfk0hgU3~w7;?H+$^I{&-@huScH9LBeAIN2`9)}`urt7( z?z34NNfw4dTkfE2KGKcJ`hyK+4XDft&8*m!KnH(tif|XwcZLxjiPf4!o^N$%<25Fy4NmK z0>;6~qG}TaAZU9pKg3yXNTfRtLKy@wFnsLNF6u7!n!t*xRV6vu^K?aZg^!q(NXj{X z%ySETh|pf~mm1^8`^!oo7rG{51KCKruiI5U^FxoP2>QCkFlQ@vvy&&h0zHuyI20HQ zroCG&F$iZ$i9@R|+1$`_14+OJ^Kn}u4&EK%DVF#hUHVn>G)^+UB&T$Ty0@60*%9+e zF~TuBb_0-9)wHIZa>8~-ScoQ=GF4&F7NJX?1$UF5*p4w+DOL8 zLlyZQ!RQkV)(mi@7B|UDD}QKN++=r^w3q(D2qhR6Ub!Agbl>0#l-cf#^*u0Ssfa;W zW~9K5##0`C_eNQLXbF3(0WSO-6XIyviaT~k%0z3?5bYX57w~e4++4pqeFZR$ZPxv- ztf#~(lRIX(OG^zz(l1@@f8!HKzC~r$JLpIYCD1OsIwmU&$}p^D3d}jaDq{82RE>FogPdp{Sp&rIJxn{Y^DRKPe&O-NF zVX4yddqdhu%M%^l(g}*eRL^l*+#hJ*wtQO4$>~D!g=+W>>cUvr5Wm)RKM0IYx@gsJ zC>BHQ!g)7*B^SqhK%vB3{||i!g1sK0{A0|Ev9*rELLbUy&hTv)!zc3>M%9kUTn37f z=>5OR{V*dqV*z@XmdG3~r^`q13u)}T)GGKd%0_|=O>oz0oCW3GwjqWIC=^VZn&i0N zgH_|l^rpgvO=Wws?)_1$kNQ36slTtEx2O1x{-BKLx|Q z1FS>UYYyEOR-EEG{=zo7n$+P;ZKS#mRI+O(Y`pVsu-(OhE|Y>BoZ<{kP@?>b*uDWT zrcLwD2tMW(dp#|{MW zFh_-|rM{Z2c`Hp3H?#bflKwoCK0_~xB`B6s8^O->10P9~0r{m(OgIs6y7 zdpp3u{HmB0#U`CO>alJjA#+8ox>S&v7@|h>GbVEB$4!}{;mici3pz==>Jr{R()d8D*~*6j!(Dm$G_Q`_G6zAwo0_4$>8YuewC675Z`gQl0sQ|rhRb4 z_(cP9Fxyyflq7Gpi7l_;mTyDZO0OCsl1J|dyn#7NO$=;4B`!_1$7GqW_myQb($*Rw z^>srs=t4RNYYq*z=B6UdXk}_d2^E_~_ap=>YJw!3%(@nN;i3m7HlrBFYuta98Ah_{ zqnBA8ko=u;Iij?(0&O`DR?2Q0C%y6`c+_iEc9F7c5~U`v7Y+vzR`eevM36alnsy+p zyLdCVZ`#z)ScB9i4=G^MILJn5dOMk`7ZlofLPKuys$o{@kmFV4G^bUuoxPCe;`$Ux zc$g%y8%(R1TcoCtj!LrG6qLEG=_3ogDbB#q8N_Rg>YDdzQa=5#C@*WFkWPt_r#X)jBB`Bv~K=CQj7ZLCAC{+=Ln(7w=eN>uk-5exRk zLc8v*#R;d)S2DL&IR6P8K9hHZ$Fy(3w}y1Ei_&Bqg_zNr2NVVjnDc&8CCb^rM21pN zUT$D!k#d5`3c{HQ-=sUZDrH&*w{u|){AnJDDktL)LX?GSxl(2Q)3s`I@?f&b{fVJr zZ3Jco)88Na;A3sspVcY;bmJ4)a?oCZh^BKRu@ypO{c|QRFJcI^D4)9nnJT3-DN6!r z29PSjw9Q5ix!@=zijG4%JKQwlgPAH4CI{Uwyi8BxOD+Vd)8i_ank~i7QqQl_H;f@i z6OG2*@fOLM!sd?nOUgMOlu^h^W1q_$Udy}PGoJDlzng6kos|@eDz7k5dSk$zeE6%V9$0ot zx@D040Qwij_o^wL*kfNPN4#-po|V%hh=r2FN!LuyATH7q!{#TI+g-mbyqFJvdpibdSSwnc*ByyV4S{Gq>!2`BK(AWZ)C&Qs zmyaj0i>i$A2+wTMDqetzhctwt%I%+E29bK03juMP9+HSC+fPJK1R#XEYx^^>8uvXv zS_eDLaFxGoaxN!N&|b@(S`t9c(o}6WFse6(s1KluYz|rOzV5ooYxlb*9fjW*X0Hlk z-k33|Gq61qAzECkK7j4bZeiJhYU!lnY@#jDde%oV$YVzeK9)bL9QMaid5XRz329Ad znuqd+JD)tcNpoof!euIh+$>Dy7?$v#J+Z|i{f^D!E6o*v=$4mS;9pEAcOXBSOa6aU z3o5!V{-@H9uLLg_UG{IT)d15<8bwKduXfq+36YE3ZfbmlHOp=amMQz#O>*0(i{{$C z`jNMkwXVcd%{~FUCdNP|VonoYR7d)y-z0%%MrzXL2(;O{U1FLyMeL{i?Ejd0%dj}2 zrE3((0E4?b1b25&NYEg`-QC?Cg1b8eNP@e&y99!}ySv@yob!IqP5&h`J-c^Tty;Bq zt)HxFqb&0sOCk$kX7h~tfVo7BExCuej8b$nTwHRuv6e#_5$SVynNJjUNvbXDHS)P+ zhUX9*_Kb%jK}Ogv?Pw{c!+YwhJ5L=1fSvVicle5Y`HC8hm}&r|_0-3Y8nG|c(;?1r z2>yZ_KDBfS<%9{P7%0}A0oVKFz0Uq|w?<@#2IaNCTiNfIdLYP(KMQKco=2N-%*x_t zzZ{#+`N$w(y>@+k+EKAmu5m0ZeX3Sy6GXH+EikQS&wE;Y^p@l5{j`&|=PS^S+ohgc%7==GHo36QAyiSkvF8M@ncvT1UDJVD$!aM; ztyjltSrB=f6P%}n5V-El#W-e;ym$uMVe+WHFw4vcW_3H%Z=`mZ z3u-IVN8Ga2b}I?Ji%#qxI+qWO8Xz!(;l}>kO_*D<#k}z=A>+6)`a^)q7Tq%M&g`CK zt(J8NUzY)+HIh}&+6yg{1LqwfsA7wSzo}85v6fd0*)nfz!#w@>sAx46AK;Z+eEa*& z)PmQRd`qQsey)G9_eeY*@&kP!WP~^VcMi${YY+%6dX0xoI#yzE24}x%KP)$AeRc|_ z8wXz$v<{2OcvSz5YNBh67{&~{1-BC_0VYe*s8$nB3l7?>tNAC02L=l=HOn2f1F6Bj zT~u$IYaScy+M`6YL~cdHWo#MWnp02>=)KLdys-Hboi}7PbFd+6=(b4D22!Xt?l&Ke zm56sRNl2}n5j7WTHNVtYGl5d4AY`gjzw5ef1m6f-!*&`)c$s5%yw4uBXF36DX^?I% z*0fDsJ7?gADJ6?uLc;lwLHPu~CP=LpZBojzX2ggwH)~O*KF=AbDQ6Kb%%-MelioZC z>fC+LsY4;RNZWzw62+kN2^kCwRP42&6OMeduTvm>rfv|SE>g_34k*@ItNWr5fq@+5 zpUyYHY23p=TF`2DiNlC5v^uD2SdBb;6GB#<+T>i%#5u_PcB(p?S240LrdPJaIzE>tj6uo2BXbPXA;=oV4OR4%yM89e@F z27zR@Km{{w&{n@Rrq!rC8WpKE?zA2EMFp^Qo~Y0r3-;+FUj4OPBt!>wlU^(h{T~k- zgzE6z`u42=Veg-WV_HrY&~@VjMo!WB;BXEXNs<)Rm*2;Y&seeiZU*5!fiL8OO(RM5rr2#8@8DtGPLu z^Ic$CrmN4Jvs=}v{abN&?t6uM59qvv=ny&;s`Q&4QbH0nyG78z|6E$8f;AsoM_?je zo`-$4uD=KUDIj+v!h-b3+8QbW>s$KVD@_TNK7VfnTYOZLr(`r@Yvif37Vej(&k& zwX8A+B-TY7ox42aM|)R*2aa&VxSj&YWqu#D)yUx3;PuS>W@x@se%V-vH_Xk{DmlqD zGh}j?HC1ZM4-=+?r*~da8l4#3Eug;a)Eu7EpMKTXIsW=oq+XA=yY zDt7Yv_FBguT2Cp|Vijqm1LAq*>(eB6g}jQB$5HkPLM)`dwrmu%jGj&vfDcZynHeuz zd=vKEgG^P96F7&16l0$D++TrLL;ZYwBe3M|GNCi>kX`(y^zB>U>iDXssMRkuntU*} zyokqHoA@tCvDd35qxhXI>91n(XVTRft+sb3OAWTJE-ogfF0TLH0&JvZ=dd!Eq*P6c zjx3s-w0|GS8P28|;7pj8Zs*aLOzXq*L=+&s@MNd0fAeXc<;ucn24+{Z#%yVBAMtDZ zCytZOwx0=8$g~R9IBeHh-D(fZ5GSJHM~Xxh2;9o3b~f|)eC>e{RYf)JL(6?loRw?r z`=VeyUO`ej+HN>;ara$Um~kW5D=O$~V@!}=`7g;i>nQyIdUTdxw;a_S`kWve5o#yp zBPL3DcFF?_WU2@pwsAuTc+*-W52M!U{gf#zGg=PBy}y0&GJ$zJlahKnS1_36=apaa z;agKXS?7Pz`ZKse;sjD~2rRnApZyApcG6ChxD@1TpSVw?r_sFwm&fJ?iIURjA~ba6 zsxyelCxFyzo9#N}>m5g8F6&6EV=rQ=`$G{lvDtWbdqcZ6pNGXnADUggg^OTq9yBxS z@cFXoJO*4WkLjv*!cGwDm1p1BZguB9`szb(j^A@l9GzflkiOpHSvg$E>@ZK6+#b{T*fy<^iyGW8?f6A|DJfmimL80$68oX02F`cK&oM|&w4 zj7v)1Cz@R-sDlWg`TB7cUK#XM!MoO0oM}uVkQr=_+c4fiVtN;beXl?~4gq|mUH$elZXEeo-85^*D5q_Vy;au^jQI(hO$PymN!n}v5UKX1>a(bEp5TRkJBATiuDlnM zO=1Nfy&^;1ZlWuqBMKppJhAt+!bY$C>^lDva}nZhuPy%#0mVvUlvClXenGTSMiq5< z8xshAg$8{JR4oFrd!mjyo6K^3NSmc%i`{FfOPY zx1SX4`TcLcf^&au#!6dtvDstpfBqQM?d>PxCsIm^ss*2$VI_mEKUSBr?> z&^KSQStgVo**uZX{;pD%Xl|T)yQ%sgQEiB@;kTBB%y$k|rSTnXKO0rr6?HKqGbOy% zyziQ@=&FK5TfH)nd&Ae`D89gJas<$8qn=aqNvV#7cW`+OhClrxF5QxIVtPgY^p;N? zOa1`{u`{|M(|tGX)aY=ZR@_RI6hdn@y4;gr>665$L4tQRICl!r*3r~Uq99`s1= zk5zb%#vc$FX^hWE18yP8*1#`(|0@u<7Y3?7G!%(ocRPAM(crEE1~?ZZ1J5pDeIB0O zvr&CKnJ2DqqHS_Xmq4#;u5D#eN#rk$@`^6_vXX{FbDBI*&p?l!!EMXrF{x_ryXqa2 z)4tFcAErd7QR@^2` zy~pAz|6P4Zs0Y7JXdEq48a~z$w>fRsp_jY0oS23PCiC^4&#;;f?HBz7S*wwDvB7>~ zui63SyR_K}<@t$7f%7@uw`ny|{r+I!MURR9j@vL5eZcO&kXChm$P%?q)v*M3bI^Bq$5|j)|O6}ENzWtGSl%>~s zjgU~r7*OD#)PqXvePaxG`He^)rf386S(7$r+BPwCvmz%lPHdMtVcyO-M+hzoF(9q=i zC5nIQqSu(Oz4RIZtCpMAqaU6HCaJs(>Ut-R%9;RBk)8Tw=_xg&LXZSbOx9=dyCzd- zyLB(XqlC}la){j<`H6r)JfGr|3{Zsbb3rYJXq~ji4TPCJbQx0xUts$zX)TxJv)-dg zJOZQ{yXn5b?YyBrtmTk(zd#b9`i~FBtSP*1=j*v{V;D2})bRAW1wDoH2DTk9F7AL- zld{x{kt;S71saUW2VxOvH|Ps*WaHD_PZ~8184dCUnu9jd)78c!F77Rc1JMqP=EXz0 zua`K#H5*QuCZ03OG#iEm@0KY@yw3dt8R>Hxh4(TYpN)7!VuWAaOk)nGrJ0j}Mpg%J z(Z~U0AX1`$q@*M-CEUgPcDILsH0%4LTmpr3yJ+yYNY(QZ;=` zb}3mE)IAfny}CJ_1R%<5aoH^IBew)(8~AcECqzfu17|Y47kuAdR;2Kre002r(+FY) zErrOm-^1wpH6=#M$*DY_gpQqjf^2HFWnkernJXHA{guThRPaN4^2}HQ)>YOr$R!&8!DQ4FO_8zvX%I6Ze|MR@~SwHrSFJC zJd=8?8&;@*@w8v%+fT7Ddq&=W*#$f#jQ+f|HxRo)lo0Ufjx?NAc^)^js{GI*a@_)4 zZskq8>7}j}9j_h_faO=imlj3vGvToKqZ=>o*dg(Mas!0_W`Ky4RQUDT+8UhJyw%ny zv?CG$SEA#D;9MDb`S;)PpU*bMzoao_2zfV02p|3`0BIK1Soz29wou8|Ek*- zl6tFK)%DTb1g(ZbE1UE<0|R?Iut{7(QSv|jaH`))?0K*oeXe#$p0#ne*T)uO25e`b zKu{nldXUA;FESulG+{LI@D$jtwHsU~X2E?Eyh*Hl+zlbHiFLo)J%NF#Cx|BES9!fI z7H%?~=QrJ9aEHbrMg4zQEI0VwwCKIx$M-rUrG|`XOpc*$-1`H~Vd^tZlX#~3RQNLH z0W$A4Djv5@rrjW3bPbLCB9m^r0GoMZj5rvyV;M<6Nds-L46$+b{xJJsAWiJk+>LYn zCZlN~63cl^?0EfH)Eky7(f^q^mDw?B);npM0{8hH2YAA|O$$rFV^WJ>hy7n-I);Rb z!sm{ln zE=Z_eX?MH3J0O!3WAg*BM(@`!8)!Y|__c>)0!T%KZ-VQ79|*t7|1^-s!3%??0VZoW zyn@1)GHjZM3Qb+%G4iLVz9IGZ#{`}Zk)jg_7Y)w=0cJSv3o&h*#|_=pNKTn}FAS2w zm9mM4neTDff0k0do)maVJ^&|+UMXr!#^Qlcp;9@a(eCBGaXpiw<1s{(-&2nd@^S>f z&BuM_jW9eKoBBe+IHfYO$bYY&yCF`&A2|)eAFVT+%KvYkA(f{h?SxNnSoM&ee{>t7 zOaVg?QNi&Hr6zGq!XT$yue9B*9?8H}dO|isB4rD>SLYZt*yVP$_!# z28aPK7?e{WvsNPn$zMVEv%&ix&X(|Fm>;t|09kUyVjhSk%|rn?UYB7`XR9l;7OVtD z$SY{{W=*j6O^RW1%oL=|mUae^zKUtOA6~D_3eTYkD(oXM) zPl=x~b{7eMVP5pTJzgmM$tlkHzVh%;keC%Tdc`tno0ugu5Int+f%VNUCx^#qi&jj5 zNODWtq&S35AkQzUEMO*IRJ9JDwt}un{Wq5E=N)v#_uY`_1hz{hE7C_@l`rQqpYBzL zTl90NG@Zz{DYdz2J+2;EwjiSF#qAL9iIUL7f39wQ})(=D(+A=m_t}YkPQ&K-3al^+A<9G3bU8R3+Ou zSfT{3bHhM2ClcJAh^G85=10wkn5f^Eby`iw6E*Pl3@wvn!wyHG&F^;#;a_jP#Mehx zqW>=LZBiyDYfFW*oticvZdwNA-m4`7m!F@7Abv zvXQ0#%@6O4xdT1qduNxuFf5)Ov))yQgN^G}JwN@3eUE>+qhh%>E6taH2RwfQ=nWW% zvZA5}gEk@j*6MUTub|_7qYRq?OYA1XBTF#Qr9ICYxeJvQ1(SsOzZgRVHNW@zwDxw2 zSvDbfmx=M1TVNG6K(zuSfn8*QZ3w2(-pW*i-O)5TT!y0v+I(@2UYP&I0T6hb1S*MM z72AjmyGReVUcV;&`t$;>m`(l8^l|+wHaF-}BLKxhu^A^O%x$zWmPC|yVWQTGKNN`H%tS20>sZxrDC~prbPhQWY!_*&sXZ( z?35}CM_MDH1Et!30S*|Ih6j|2*|w9cnb{<%meL7De~{lV~JSJ6F+`%qycpr`!qayHNzrMR4AL zA1ghQ%(xjGIuIV4`$?9x7u8&1o;0V?thAgd5yEBhKfiMVI5#z@b?tBi0>>&P6I@)gYUI3 z^5`~Xxcc-1)43rDejiEn``zk3Xgtf3wB!CKmB?3a6eeKQf#XAHQ}|M#sxnL^k*+u`fmkY8l}wP>;KlI1pJzlMvb*%P3M8W9%e7mYJ@VYzZvJUznM0AU zdY%bB+|21p%Vvpx{%`Wwgo7%rPvbC;*So@kZTq}#NRG(ue!0!CK5*9qaipBD*(AmBWia6 z>3>ekB}fGI>ro?s=2Hm-UR&wBq(cx1?9)av0nEB7fChP80Rd}ZVBCZZpJ+d_* z#nD?Y)jJpYPF8hl+96;w493S?X4tArP6Ws&ge}&ZD;dqo$?xcT#O)BXxu6ar^#YUP zg)`eq$FulbD7Se18A>C^;Mgeng}9dorN^LjwxOlSWYzsT0bezy_eiF#%qlSvwhYvN zDOmI>g8&v`WNtkdQ-=`$We7=%x^D;RhphvL`oj+bLnoDN{d&h@_Rmm%6oD~NI>vEY z{u_LKI%D4IkBV@kSpIL&G7={}-?UN-lK1Oe6EjXw@nil-qXgtH*IOF5&CW6zTnU>;r2GDZ;dQ#?3=fyUOb=q~>8b+|U zuA?G!?`pQbA(K?yIlSn(;gZV3klKwfHZ&>;{iX|7B@`28s$OdxJXb0r)2sbEfE00M zZu=2wH|-5sr1ox@LqvRML;w^GfNpm{82^!$c8FF7^2gLV8%idG-#a4xh05}JaHSl< zzA$iK`G9@aA-;R7GlLR52*3uAKQIqP{XNpt>b84f3!9iq*;R##$hTuF5ZohncC#QP zmc1HB24Y3}2yB>*_D$m`REgJrMg$yYz?uYKEIkQ~=Gl8lyjG(|OMXNUSEHc*@C565 z_(8%Vr%RE%KXY`bxLS;K_;TE`KI*6`y|=U}-7AW#@Mo1q6jT~%)b-ey_s3=YdpNKB z9biHSiF|wus&IH{O=Yp~k2y&z?4iyXY=u|*d9L9n^4lMR%6eX(NXzR##9)&kIi4RY zbn54iTVG{mo$WFGoirP4obQAR$YipUQOqeiM78WLkRUt8o^MCtnCGcFlLemzH=rB+ zuR74yCc%XWs5E@w%FA%^hJ6$OMckvFoBy2BH>*pMjj0OcBkK&k#sD^Phx*1I9i37>JCq%Al zp|@wp1t8NG0;y54&O*8Cykg%=q?0@fXgZt2(m*zB@>JQ zoT1p{U~In9T<3JMIE%NrN8H+Zk-ZTd0TzaYVh!-yaEcV_7%}DLIRqk7;!$~vP0Ic9 zu$hxf3Qo7|?+}Rl^}o7w2(}W9Td@H{3DhI4?pK-X1GLpdd`_r{=YhIbTbpA0A6z4; zQY`#}hx<4|4h6PVH23Ti0$U)VM=-K*0~-uJVT20&8Y1csK~N-9s?4XbT%2=5nXhE@ z6c|;;>EbGy^;QQ~Uk(4R8IfPE!?P(IFR>?*iVMZy@Jv!?dmPeolUaC=^|$jTv7T>b zk}C0BY;=}&`x~k-2^j<~yTn15RGa^$!sB!L1sI{yvzd2;7BZ`K(3MmENE(M8ZXW^} z2B?l(`AK0!E5w4~#`wEP;gI9sS*=n(%0YPA>XH=E4~pL(5n35V@p6=Jt_$b;D z(%^6XMHLl4JXDl*z$Qeg{Hj(Xr&wP+)DBDd(VJwS9)PQUn(oH;dSbrjoe=^bxIQv) zfpL+qL=iI1aG+Yba88;8>5<7WpYT zZ4RajFCUL8S~hDGaf1!aokLJr4|CH(Ezc|^OQSA zD<7k5iBLxm;o>Jw@bK#se+d8rE09w2NuJsWw1tPW)R?yH0E#U>z_;iPT7H-dAvY^UMj7tj(5I0z~?E$#DAP? zFek_1{=>)_KXVz#1}l5AnpPBE|HYIcVFHIa?kC^PwmBAz=gr}cTug9UYV01p-b?fr z8!ee|X;7-5XT2UtDL|1>{+xNBH~FYzhZ4;)2*E7h`np%%vQzWzQF>spZkxn9W2h<0 zLQ)Z020<@dZpEL{BO!|ta3@Ay@PZp4zPD}3Ct?_@B`q8=luc|lfSp-9(VQO zfDmIY7!ipHgp@QQ8sfjKYI@l}h=(A?0mjZN-lvn|D9-%Ija5GS?Mb0RpWavp9HkB0 z1?d@InLD1zM?nwOJ17bV)X?7ZMHK`0a>9FyA^UT7J2j_hqRy zn5#dvZ?Lon`$KMl?G{{bm%ZN9?L*RLekm14)7t{+z<4VE^XZjO&Up`4bl1SeXm-XK zabh*eXk^uNjIz)BNgLV{?O$>_WSvrSRFDfhA$ZNx`;k6ly zOz^7M{59`+4Ik;bQtid7av(IQM^k1q*6yaz2ca;6% zOWvcz7(D0SXCsrS6ACk*)tJ!QsbOTLDumAj^bkVR>SH02SlbKUnj=*=7C#cDXX9$v z(48vy1`}&R2l8$CtsZ)gztyWUTe&>*1b(t_f_tN83H`lc`vbO!@6*o6d%i7;bS{oC zqxLRfRv^}QJYPyb7<59ISA@^0B3n@lS}0F#YhvMiPrW?f?9@owz6p%qITZc`GoG_& z;*sNf>i%0n{NigC5*L_2DD|6`9ROK^A>52`+&>|wTX$|0lMz}1c(9?Ayi~Jc9O;y? z`#zd^yTGlGiRUq4c3?W))*aMg1i&Jkrs6DxR%Y{-h$$L7txreUoo;Dl~9Ds+DQIG zt;R$9n{kBkD7AJcz#KSwWl$-OtLT3e%a_|!+0=JZ+d6#+%4YvSJwz2-{39rz&mn%jA$s-H` zbR^DM7)ejyMw>qgj!$4;)a5KpWUy}E=ExLZ?gQ1b{f}va<~ql{2p%HSC3ElS83RiM#Y4i#o#!J$Neqjxfp@6W0!8 zj8xnQhC-waVGN9rJTRO%dk&MdmoeF^N%=G91(6 z>qH;Q?|hgT%nkNl2j64ZL8n5i%7bxVH#Kn=EqP>?2jeJK#_75x92v`2uDm1Y&D7x^ zi8cN6I=p$wqWM={VYa7!%@GTtbkv{-)MvIpBL#o>zSm?iAD<%rl7BRY-wH8 z=%~^5Yjm4dnEc^3PJ$c z;T{{t7AYrXpBEB?TFzHp30Ynkb1uv;8bUu_{+_KIXD;Y`0H>&Kdc|Zh z^|ilDqXZTkF+SlN??M_egwFO6icr`Gz#BzEVuSl1Dg=%hvlp0HALEg1+HvN1ixi{$ z|7HPnx7S$?r;Rz)y9^MYiw#ST&|z(lQr&D{EJ!a8mk52ivv*S#cz zux5QO~lXcOyix1u_6#JWe{zzPT2|`5u zyNsb$g(d2_6q?+L zn#|s6%|5Q$ff&ZV6ggXK4G1D&X$s6DMLc&#KzEK9D54MNsLhSLW8%lphA?^8IkdW zy0MH+Bo$hB>&k!jz;HVM{I;=GdJ2PpgXbfK2r6AUHDjD zPGhf@$?Wuv!ntC6TQY|Hmlyr{17r^c!CIP3@5Tn@=ZkcbLA$R-H{TK-o_&&XTE!Nv zz(GkB3zdVegM4J|&YxfgwVb}T?8p8zSCdqoP5j%og^w*bE95Gh@zoj&T#_g~&1jsF z#p=TcK@X`0jjHu5MvBhdiPT~yejWe`;}-X8_&&q~B7;0cXB|5Hd71O7Zq}wU`{1{x z6Bl2ro+gQ#{+eDr&KKTpDKyydb0r9{pRm>>yH|nfeSeU4ecPI%VG3S-Xn`-4p#$Ax z&HWEqT-V~C73rIY!-!no!-O?$*Az#pdEfAiHkV*c?Tc#Jh^;?JJ|`w(QkN>z=g`d3 z|7pUZ>KfgS@Z^8|w$tmNm~b-Ltm*$jELAHQCV|?`U~SjB8`3nthqZSX8PYCv#Et*n zcuU%h<)SfN_RrUst)rc6^-N)HZj-=L!jAq~9MAhnJu_zH9FL`W?(1S6+fn&e2uKvD zeiUyF6aD^#NzQ#djJ(GF=KZhev5r?X7{5!kn!Jp6`Z6+Q5`ll8+4JJKT(&a&U3gGK zbD%^Zo;Si1!#Z{Q{dUtc?{hzL9@sSS-d`}{l)XLBstfMWtLhq&yHyT}`IGET{(Sfm z-E=t%&O;)SUO1U^!E0HL=_TVcM0Z z+jPE_T$o9BkODz(u+J4KyUCa+8%qe}*LH?TeU8Lz6|3*&PwIbM3)w2NK0%>lH4~`k z^pXrujklCPgkrMp`!^`j55uP<$WDiBeT%)Z&v;ms3K?w5i&-K0WVFJkb$iNYW&DVC zfg@+CD3f&~`-l$Y*#hArH3X-Br0BZsF%5~4x4jE`X}26&Tsv6SO|q&rHY*!{HzMH} z?i{vbE_bb^&KhX2BvcXPVyV{*7>Lm?kC&;Gzf*0tY{Ob}XmoKi+U8_$mr;=^`-#Ai zu<`hP6FTj#SRdLQrksuK9R^&K60B@8a(VE_4hC5@oMbl9(b=(cM z^1khKgl$SZSRkdp(H%cTweDJfXf@-N*zIcTQ%@DGf^#7~TCB5N;R~D@!W6NqZ461ojXfce(tj zG8`}|tygGn)k6q0m6I9=;*9#_XF9_1wQ5=P1UktwjXJH;DgCFS7>cwCXl4;oT~rn= zHWETE0f}|?$5zdvw4a<2-6qgq)4V^dMBtcy&LhD?Dq1DH)ej&xF4Scm-$*-i$Vc%> zM}3!h{6O{}e8m)0mavlZW-J!)TK$TjkQ8Rh&}7gDZ+e8~6ZvHL5`xH9N1gVgH+}am zk+E9mK))m;YPk^!AI>^R?i`7*T{zh>?w!9NL4|OQ+S?>}J!Qg(Gw1SUGo~L_PT;H- zZt-kBHmySV9Wvsb$%ijoKdmMlfPAlCfUv9ixPFH04`2xbZkuBmC4j44ZgYZ;914Rx zrfq3$-@|ZR-N=p@AN0drA^`~81fZRz6@o%?ifLn&mtO4=7^&4J8C??N2W#N?S)rRR zgyQldV7B`6)qD4tNW5y?X073JGv$+!?aZ$@jE^#ID@sqtCys%tK379Wp8uHcTMclYG_;hr7cpaMZ(i!FI4=JT_ zpuXalJ7Llzg4c%MBx`$gVQz72w{Qm$K-TF&G=batLSkl=+|`fPj84{*wwc85nfhSX z;29AUoTI=M1Vvyr>DsveC5cYiK_BJ{TNnQk`^I7_KlnqVtwkPoUUtyo zT*cWvtB*a;qq3-$&RPM=Rw?rpvD2uLu@O0DIUt#Sy)P)R2AUE{VI@l2DM=YSlg-LE z)&=$o_pH^`KAVBi^uaq2D=l|}kI7RK^dHuC0W~Q$TzM{l(_vz|7`@!snsdyM+xz)> zlBiI}Z`QWG3~HWoFGLIsPuWJk5>XhFKdNBOYTv&MQGJs3c*}pL!T`RlY%rY+=qM%1 z2X7!Y4UKf~?s30;-3}XtHQKIr>z?-(uLQu~+BQJIBm!zaJ{a+>8-VSvw_c;_X(ymU z8Pi}m`O}cRP3Mt_+r&cyn zH~b&sb{t3$0shxL2;#oS_`3gmILIflBFNQVeG9fvbD)+T1`khP(S&N!tEL;fL0Y1! zd5DI@`ya^n=&ZCgf1F28UrI{%w~4u-@?(I{9pHgvM=yf#=%cfCWgZ_qHepGe%!K~W zN2nGBd952UT>XF#5C9h2y#Valil}BSpp+sCmh7ZeDjq~>rr0RhvD^vYb^bWCpbQ(@ zWx8J#QpK9vWlx*&^DTb1T9Jfktv!&#c6HQy`;P25z^sI3HpAA%@k}u&Nx|qb5G~?4t31b9i4g=MVMiZD_b^Q{AL7w*G8C zU9p>t4&L!^1W&{IbFX>T4AL9Fgrca=KR^2jRCj#dHM}0##N%Epp9GM;I&c)6qH938 zLPr}}@LGPM+Bw4O37{(Ou7fXqqPP_2hCUrjC{P|VRaXyB!kJ$_=FhS@O*!mr$!`D9 z3587?6lH(Lqyly7G7)Uy31u*Vn7Q!$&+?5v#|=xV4usxNdzC#Xr z54M$Ih|Vo~gUE{Ri8r^?D`ph$iPKl*z;&x;a=YDqN<}Tg8rzOzUh_7Rm z7^P^fvD^K(_*)8~fta)|?DC-wgoDAI^amtE!&$DY2WG6^nR`40cBs*q5I-uXqD*v zp!jm9)Bkd_)%cI)3hf@(ABvmW6(U=gXLX|%AJgNIvdjC@LNrDHC@rq(nW3W^w1IIc^)-egjbaksKBhs%WA|c<8!5#7ZJXVH=aH>drn!uMrl%HU{^OXtp9Jh$g!`cEF=1qd?7RpYJdawi4n zD*Hxo|vlLc9!g31(wVpI2-XEjT zna)n=^!L_70ztra&qW)=a*HLo)Jnyk#m^h8KMXk@Q8|H&Poc3+UsZ%v{!uRCXgTGz zUS3+V8BIZ@33()E{;jPiR4*x-srarl{oS5b8Q9!aR%M7f^>Y9AEr?WRfF=h;iIqh+iFE`v$b_g5b#yMa<>gtjfP5f;?R!IZhG+w?Q~RDTNcuOUn&ZT%YeRm@2re z7Y8M&vpxZIAtnVM1AzZxv~+^#sS6cKihK!h*WgFlsoS}diGWX zVXU;LStd!T?y=3jh+ul^FCDm;X#ReKJ%{yD~~0HCtNW6|?N#AE9{5&EUwQh$BAvNWZiSdK^@g?u-& zLTl+H!(!0;)!nQg{lR_~=0WLYak|}tQX4vhbmaRKr`Q;b_iqlj$8;n`>zwrIa&3N2 z-pSE$?AeCui`dE}UW-jsJkrV3s%qq&e79QC^#87LuXRMCWem6Vxzylu2i?pAZLN%weMP0y49&@>9|TGgMZ zZ0E`MN@QkAOq_ogRA0IfS;{G>Kd>i-g)pd>z^dCix>s3Xn;JWCug2(7b*q>#OM+p* z+i+0psoLiTRALu3S!}w_2Y=j*3+YQV8b%-FRH9!<`6c1t^d;E2eMZTvfgrIm8zL0U zgfyQjR~GcWW;iT9ob@A;t8C_ z!ilLbykuF7ziA-tLU4rSP`~+%9x`dx%d|fp($BS*)*Iwjzr8+<sXuf#gcM{#-?}n_PV3K3=yz7C%G+gz#wYl< z9q7oF!%zCcsP|4aAR03$af?yvMrsH#x`C)u7O>)wmHJn?GRy+m%u*xZSia^%(jS#%Z~T9*!v@o2 z9BfmE^gn13jIX?1G2SCS1mqESyDdlRs}WOAJ>f?+dz9%q!?|5MJRmEE8m)l9a5iJjD87} zR6D%zHS^Wop~-MEMS8c*Oyf?Sg#(Z9CAn=3RQwZg#T&Z~r%(J|S!{ z#%HD4Si0%7*>c%472%NIFzvgXS~?}rlBc9fN=OHKI@5PJOj7A|^s*%Vi859UlQ4fu*ZlLDM#syp zv&yqt$UeMh^~M#l+A@E{W$DJ#rin{p=oU*++bQ8Jb%x^eTcbtoCEY)u`x%uZB%M8l zReZB}LfvL@QaoG|Rn0@GTr`$OPvMMxYN(vA<9JdRT}3K0mFERD#*KS!FC`f2q}SpI zs8MYyLpKq-kfE6*rrqb5aKl(JF2&KhAPrTuos;OSsfXPJ6D>~z9@tE81(au$TSz&0nNbidRL= z`;Vq79*DH_Z-{A@bF${au;7_kra6IK7YIg0SE2l0ndV4cc;ME+H~H3k@!g}t{f8^; zL%M~->EDI~eNkM~fZEQq1Ql=xfez0luxTWdX0>&u7N+p&#T!pFZmn!bdu|o&e2giC zDgIF-%6H#dB0l});ZMqvNkSgxIz(JSG@r%XPs-!yX7LDnL^icNeONouzh&R91g=?g zbIo!a4TSXd%y90`_ZX!@x%b28M1@~JPPE;JA9uJ#)bc-Q;WZc~FlaTYn2ZvTuX(>^ z5iYl7XnC$Uk4tdbQIn92uYW9YX2H9Nmn*8uic6|JH>_uSA9@|o30p?VXl>?&$0w6& zN~4!1ap;p*%RLCGhvr3)MV~2GY*D@czZY^-MMb#Ck@>g!$t( zRux#m*z34w_Z#tA6Y{sfNwXloDbxZIArh+bl%xT<4BBtNAQ%qeeLLQ29h5Xa+dzi3 zSUN}S=mZjp*N-pCVIEtu+`R3>o9K_^-V=lmX{FDA0JIUBa&Y+k1ke!<=e#|e7Q5|e z@8jY;#|lIW17_szj!mzRN6u`{tqEbfjGt7p0moJR={#=&^Od3`z($WO`mW<@oc~k+ z39}tgdY3$58&@|g?U7=r(?fTVqE@~do&`MhZaDgnsLO5XchcxJWy)-ClgGjH?Rq^& z5>&@aAkf@JV-u;`r?lPqF=*5+&n_8}7eHU@KXMvc>S6Y`ws3223Ld(|FyXQKm{vyX zfr!XQ^Hd$@iU#N5$BZ@GxD-2<(5lN^s=wQ;s-(M~Q8+I;#XD)gRJ*Op;PF0t{w|`| zU#c}f!8Am*Ohq5A<8|B}1iDkqfD+YjNiKT2#Bp!j=JVmZmKW2SW?qD4wLpu99z(yQ zba_ryzg%fgZk(fQUbz2j#4h9ZpOOj(SU(nm9}iH-;_CTJOln9Ez=ySTu@~o)Z~JeP z*0VB+dx$$+J*cQV!dM0h-%0X+|4LImzZ>4|yV~k*(}$~R znMP?6TyBlUwoJO_hXb>+z6T2xQggq~{{mZxRRWEt#XjUWgte!URro_iA&f8(amAxX zsiCPHkGk5YzCigpE0geHprs(Pbg@&vBrOd6F-cp1+KOZD*{woVAqAz;4v&ay5e}JA zqxP%Bi;#p#5!LW0zICDE6_cW4M~~5$lOn~IF`wyeKZ>6oFEG^#s@3C}!$C7VoIQWL zFGMbT^ZeBFZa3DW0QF*k05E7(4l*Jijw}XFIx)X=CQml<^s+o@=(4-PB``%?N%LdF zIUA2Z#mov6+1)+O3CZMYbFYjS&QHm zj7)$8bBE@Z3ZYi1=l}hWy277^sf4aN%YB2BNOu-($P6~`z@LRhEm(^*0uB!UP#ipn zQ7(W(VNTxJ;+Gz`$7!E^UxcvAD*D71#POg&25^u$(Q5N4NmV_3x+w942{TeIGq#jS zEjMwDb?IMci-^8c@o5<6M^`JMbZhXKrX!MYeLs*SGGD5XItn~j?hic|F`8{y^U*x* zc=KtHhvVxz6;}CaP?9Dd;us&6Yu{dw8Ce39pY+z;O``(_7dy z@B*xQ8Y?YEC7}lrbI4i2buiA-=hyA9tAmjL2;rJUAiQc>KQc4IY~SdDEp6Wg}kG`8)D(b%>*(L49EZSN15`Ep@hYn{iw zANgkOnBQxE7orqfXH)?d=3rXqLrzM9yN^AY!kzSEcio?7nLrDKKMM3_&u-jo?7Gr; zR8+X^6W$(Hr_zUuge5>f0BonuZY03?v}&NK98_3jU3`Do&B+nMfqWdBS>LBdgKqSF z2)dWdM>lZpzPmg7bd~sB!4~6U-gNeu|7#I2kZQI;8QIcl#F6cO7suqp+#WPHB~S^f zEV;;IxCuzBj|S!u5A^n+=OZ2XiwsBLH#y-^6sqo^K!t~newp$D=o1j)vpeYY-l^}% z*duWO&E9K@{R3h}xlmB|xADxgwgLA`+@l1zanwdPY@3Eyu|iQ_MYG(b44F{O9jKW5 zdcvAug@C8`i<>`9k6~uFlWzC@baIjpq;kpl@0Ta%qv@KzO9xXiesA?8u-v%)o_;lH z%=Uciz6Xg{tDEJV0hH3&a+0fTO1FuPh=5OH2JbsfBeNU8sjNm`*f6e}>t5#S+{t0& zz7BXFhuVlp|8e^2j-7G+SBU66+4lrV4`3V5%~bbxz5%4yxq^Xp3iO)+^s9m~jYzuJ zztUq>|NbPlfWrK!oV9NpOMdx?Wj_@;$IUm!oliz1Xu~4RlYKn4tLk!!n)ociJnY%|lgxrsrCMs{7|}u}}!1L|vL>v`kCwVG*|)-ebZ5^2|bq z!HLI{dKZW^@2I(f6Mp!@A`oz(Z9r~^gP#?ygP+}cqAt-$v~0XLqOF>a1v{@HgADhbs!=-a3cL{1*xP4+51Xboiu8gJS*`*n!}@{0qxu zAMzkc5vw~}PsH#@Jaj+PTK?X%q{BQWk3Y|1$**(jH=8X8r(;Lw7cBu^38%jfuHO6s zEsXN(R+m@pu*y3T3tMTXm+O!iTI4slD8I)L50KdLw+!hQq1q&Jj$c1it4XQ29z5^ z%MatA%SZ9ZNSxf8{#eKVs59fY0vet1MS9rp`zsan=OJT zjp3#mzrgt4|Jc(2sJ$_!_HkOqs~LZipZ?LV7*&U^qQmtS>I%G{Z|7)UR)ZSk1=TwC|h{J@b_<8cKG zMbgswdNXE6$qfI~4+o;UuF5290+w}Sh;+#aIMmi42jzb9*CjQ4`t>&UTA|_M6-}P3 z|CBDP*@rPX6)NRsKT3YdD74UJz+}PI7jmx-X7fh+EiOzk&Q7@P;lAr4LEC{k%;g1G zx~zoJhhxu06OYv7oB&}_Rng5`(KKqiI=b*E96N`M6)CN{CO58VR<0E!o_P_!LMW)o z4ASay)Vn|zXe=`OsEknml3!Q3$yM^QY^=ezV?CWSf-7N8xSRWUlPlv#%;zZCoZ1S% zk7@)%WBN0$C?x91FOMB=I|fFx)G*3KCrXF2adWY$ipRX{~&P6%NCT8gMXyg4xap0SiL`-kwYBNBW61GS59*O8~ENi5r@XGTJ

    4-G%PuH^UenF!NiYk1P>G_H#-T!_zbo(^1^>w1LI3S)KNkMKIp#^b7;g4T=fAkk-I*nrm7nbAtv5+V2=^;wdZL--o$FN}K1QOeF(}V{xJ* zaR3e&`z`bEPKnTQVQM_K860#2$%Jvj9^p)*;=Hc>= z{`jp~1Nxyi)hd}1^`&8Z|w> z+HQPGO23;iAE43{;^|3!>8i69LEd?r2a?qeJTS7Oa-e{-lLe*o+9IJrB%;Mw4Kc z#P0f`8k^=|#IkR$C-vq+=F4O`Jk~X@O;DG3JN@!4nvEH4w(T5z4b4gADhe35y<;RH zU<2hJKewxa{Rb-D&fHq<7BOm&fDUHoZ{A{he22IYryHU#LrhNB{ciTI0Y)NcGU9|? zw2OueN&iT`7lcXlP-G)|vS1mD;plth82p|8M)SF+mTh`daa#XN`7b7i z_+m5jdE*OCKa{omSh$CC(V(&-Q$O;oAGPknQjVp_%5Ey`HemnqgMg zdnXe1mKH)<(IE$3n-GfyH=>yLpVE6za2~Im{bTV&^TYkA1fk2W2sel2ydf59HFMRAI7ep zFqg(4P9-}(Ivvth=mP#FNYf@M(M)g$5l6X%>ZS`2C;&CiRpU)^_1<38Yo=f-{GS~efG$hj|DJJi9Xl`eah@v-& z*%{i37@FAeUu0FcIuXjbd_b^9=`jN>a@)w^@bd0{v_aSGAOaq6a(1Y^+vwCxgG1ig zDF!X3>-f1v}BthUN&3hH{jVlP_U7n}G{u<4c)OS;2<7G59C}iL-1)epL&ea21Oo9D%;I1$HnpqmnO-#`f6Rg(MWGA#xj*rKEEl zPyf^I$k?o3y#NJc>b}u&IOJVzYlq~y^UQVJ>MPmzyB27m_Qm0Aw^>z}>HuBmT~amh8}OD{zP~1*p@`RpV1I8Mj$?Imd`2n% z+mq=a9#P50v|-_3kn&xH2LG~5vb9_>SRo0moU{d=uOdn9Yq@&zU(`&Q%0c%1`8ma8 z$f3{FjM&YP=%RzV{`f`*NTIWu<5Ui{T}=+# zlc4&Wyr%dy$$;2#0K*9ainlX<0LS*ocQdcpo(ix3NpEToK;E>sOgJ6uvvaB2!#p_< zqGW2T6bPStIAGKUrG>|ffEmCb`yZydCORTY#Wn)pO+&f`Cel-S2~zDl3BUsMWfbc0 z4bt;5fTQWmebd>M7m6zCzf%rvVhX=Qq6}GDl3=!&cf2P@LK*|IG#=xx7gV9SBOGg) zEXEOB4X1cN_*I`~Il>7)P5_I0ajY{LOnULqqReE1#;pyAZXp~`0R660^a}>_$X!p# zCcN>fw6Nzr1^ps!cyYhpFZNR+YD|$7jG;b8zqc;hFWtm8AW{+cVMB;I2e%nCbOYP# ziNm8s|kk8CF$$I;SjBZe0 zJBvn_TloyxC67CIcOvKzPSK(}E>ure+qf6ts~j_vbO2%=L0Lg<|jtbdlmFY2xcfA z7tZoadEL9-upaK*!DX0j?xribBkltE#sx2*j?Z|X37Dy3xb|{ba4tnd5eWpTgjIWl zAgsMsQwu9bkf8SXKpgsJ;K8m=D+t=>Q|?v-8uZrBp=8#fTNA;@nX^N@w|OuP`l)-j z=GXH1v0{g1Z~oLDtX@e~2Qn$qz#;EVlTSlhJSPeygQeYYIHW&U$UaqN-oEfU-o@&f z6#VeWxG|>0wM6tx4FAezN*>~wg5$YbfUz4taP7V8HFO!qy&;z-AZ_e~e>@Tx$ucDq zyDOwZwWF7_YWkK{&r<{S0@;7-48;(gp3yo(TH#azuYS4lIhKuAf`@(?I?gq&90(KO=Cd|Y5BkNeJ=@Q&=t7#?YK;nj3) z_>4P2bPw$5MyO}yH;vkW@KP17I}1;#&)%;mhjeQO@ZF~r zc)xRZKO!_L7?X^c6>-Y%bucUua+^SPlU>l3%VDXKa+53?%6v4JTVX$QSy_#(uC+sE zurlOO!M$-;YaQQc*Sonw>OO=M z&5PLiLhM+6p>ty9mM7n zxO&l1Q1fW+9UMj9fk!Ue7Ca7V906<(X8&r1i!6Y1MGsF{OQ{a}A!nNe{ci)WDBb)- zO=ea+OQw%;H&e#kjC#XS+Q9kRU#>&88gZ!4(f6tZzRvdEoK&AtNl?GgNFrz& z946km@47$fj}gfreis=U(e>t44nW$%Imn&r)FRNgrjdZD21dc!-D*_ZQX!y6bqO}Sh2Xf7k}+Z1!WgG1f#~1YC0mH>CUKM z>>fMbO7XTkik9Ch3&M>KD`^e1@Wz;=sQ4=k;CyuZ09}#=;bfVt-|&iF4ZYB>&h>r`MacsxxGXMXYPs$QZRx zY^t0b8c7FV@`KFA16ZaesIxKLfMLP`ct&Y{ouO0T(z&yf7-RmEivh2C;qAtDA)$Iv%Lx4Q8hZ&}I zz*&10{Z-p5_s!9UvJjdBJR5)Gy-}~C=1T|r>l6dXC+h=>f0U@? zkd=l5wsxj}dk3tyB=gNFO=cqS#L9`s>l=;=c>|y3=zR) zQ7Y{hgb;t1&k>(XTK|-dL{u+N0vy3wrLLq@eqc~E7lGd0oGMdc_d-sJ@Du~1(a{-W z^V*J$lG7KQP~RHp2M~R=%P!K*b^-GhBcW+uJi3vv@FOWiLjlPRpNef^e7NO4$82|8 zUS{YGA(z;2?*Gk2EkPZpP0?mK=;|)UG>V-|b&U5OLz3ORc}XQ{;(+){Xc)g-y$A@P zY2j`3Pf}o8GQ1hL%%8g!IQ>dSeITX3)NL07C=ge+E6?u3H`)t%51~LN?2&2mL7k@m zB`cio#KE{-Bu$4y?$_I8NsF^iN}k@s&IVLF>DPNBT@&m>@MqYakT}3P;sGJ3(Po*g z^{5~+tAfykJP%9PNL)m!Ct|DwI5A1Y4ztDgj@H@4``Ly|_;WXfp8hWRX5?U}0l+omHC&-OvI(kV_{QP84 zZ1yOml5@!el$@RsPdLJ=2>a)uE(cn}Fc@3mRW0`E^e%Ec_cnQs*tg@v9C}*@p*kBO z!-W2=(T@DaO|wQLEZ^t8CbbaiMNc9ht9ra##iw{~R8)LE|I%ou0~_#tlfoYJY`obX za#oHB6@w?(#32MR)5Wy*s2MUcSn2!os;LwcSM zNV+Zlzt9_A_VAC-Fc>#i$%Y-L{9}u;$lIU%1;zY)j@o`n`hFLDvbg)Q**9}j*JXGq zu_5I}z2NBMjcdZPRP+AhSwZ^hh0&BL+IxIu?9m_nQMV$-h|PK4o)3}B=?}+=Ta=p3 zf;K2WZt)u@8dC+9gYEL;q><3bn{87+8>sQDS(6TkJPGVdsaXuJ#48?SeV2e=Cx&{>X zFF-ixXCZjsw5>&mIi}RF6Ec`5>aGz;V}axLm}179VpQsbI2!hl{B06Whnl+zk{(fp zDf)eKX!MW0TcRG1o0pAm^{(e2NTLi%pUo-9Fb~MadCnIImdD?T5(<4c{fX@J-!4vl z^U;&FqY9#$iV0-`?|WI!cr~c4KGvXQR##bHGC?vC6Zl*rfa4z)g(0X#O`z{*aC>p1 z|LLSgC9#W1F^!1PTd~r^!e`b9-=>*MgO}@U>TZ@CcR3loK&RE38&=H< zj6VlWC0k*tzH~rLUVooS4w7nZSUDsrCUWpbXbfw?_Ta1m{8-WZ)TwL}(v61T*5_>>2%QxGbsUaK4T415vq$04Rz!Qu zzF~Xx=D~axgh5j+P(|C+($RbdRO_FgH3Z-!mdYnhMEI&F%$iP)hdZ4Le~?9IHpO)t z_(l_YH{{6vjd6<%!3n`EGCIJkhe7en@qW~3}F#lQ7-Q=?99)SDle9}<#~(K8osi_kuaH29e8hhonu zwKl(-`T*zuT*-;*wQ8n=ewaz=IkKA$aimw>HI4;ypUx%?#h_`7sJ$7&Arw!gxLTXe z)F`<+x#U3t6BWjX++J%Q{&zSSOq+^k>t5Kq zXY=+&_x3t+pu}U}+q`_TGw*b^akC(9LK8!7=!AnQN4G5lk9`;x_rRdSKzsB0*hQ{IXbxwhC+lJ zPl&!e>3<~u{m?3k*!2A?KVPG)+Ik&wx|e1gvszg_cNh(yspOpx1<}f8>+rw``8p2efF1sABjQ1{`!QOwl~(v&goZe+)ZT3kHzi2U$!88745F| zF)~77EZ?cI&rpzBmdNYDntI>g6U$BtIWs?%BJth>Amw*}A`7=gM65+_DK-Tm1ALPi zNF)i$3#hb7b*|X1cH(x*82&d*LDhI0arN8wVs(OPA>3#Zl>J!SK5}tRihnXEf>i~Z zi5y%hU3~&pFu^K&iDLFiQYnR0jEfFEk0DHEFoM^A6R`4M<4wV?_^Uhq(-7;G{b*(z zs^d$O>ue1`{n+QHtEF0dfXzk<(`~A4;Qk{B=b369N^t?T?{!0BpBctFzQg5{e%awa;+hDsYxn~z$m9W z8T?CCg%$U|wD;C2sY#6drWdoebuRnBP#mfIlYHs~fFqBr=PMsRig$;|G89^&-+qx2 z7FNZHm`p6&O4Bj;QafiLKGoO$IDmxo%QtAgw4W7@AMsPDhi{A~TdCXhQ|H9m;VKgs zqhqrY2brosleU63gANZzz{3B)l;t~mD(>;p%YN##Km9ZA`NQ$1%okll5pSbyeVD0` zMCxnjbN=i6Xmx1tbs0CkI>dvCrBxO%6_DE<&v>dKd#cKuoMGE0`&)%_S%#SAC-!8V z=cYHgvm~(mz>n)RA{qNa&w_Se@=u^|UYIE*N@7OPf05Z|h@g*XUE;`CEJ}QRi@gB` zqzLM5=edgi2WnWSXYQ_OM;nrH&3-yr>xgKu(UhW7laKpmJ!i5pJ5dMRz}0H4ssG+= zQNSTU!R+DBTTcJzv#J6Bypdl;0}>GlQ$~?AZ5-=madgQ#J24UaN4#G2{~ORkWI-x> z&O5F0dG&{iy(bndNE$@BMI|FmrkHB+nYF6&N+_f9R=bL~oGZxMK7+@XYh9G5&NKnI z8lql!V+5!n8+Fd8VDY!(=4+8aTAqeTHdb0<-R0n(F_5q3lksj)r$1~5Le`CELh$uD zc+6nZk!@!F7;~>w%0A)2T$k4ML0Ig+)3pYt^t!r?eVx}^k#xQ-BCT$x@QW{%hbE?A z^vLY3YRHSse>J|Z7jV@Au^_cd$jS4lkWlF@dm}Lapc2D49s4o=Ek}M+;u;Lm?~bn0 zG0X9>DXa5V7%Nz+m3P=kE2MhLsU(vy$Kmb5fml1y<0n?>&G#jQx@-mn9ayB&r}$18 zxQRv3zz}5&Wc^OAfvAuW*y5q+o`NB3fFM@+GR+a%;!4?T(zog~B58%Wxo`o(#+uO| zPt%`=ntieiM7%wwV4M5v*W`v|H3dnj&&P|(va_HXU!2%eZeBKn7bkPZt1A!Fn7XOOgY79G$v&ka7bJC$`w3YIa}xW zO-nt$9{46zN4V5kI;q$M(*7)`-^HGmHdVh|Z$9?FpxDqjSKWRVo~#*R^mf3JkG{_m zZo@~m2b#`0Dm4mo5Bw^e4G)xl&9D3YRQ(jKdUKh-#1C~!P6UIITz~y4$#oX060kVg zYkZdbqLRz;LlKdJ-cGB#H!3X}MLW0QYJOmhF)-G5(Hv<(BDv7cP1|`M{?lnrB)ZU( zC~P^uZ)iPtCp3y>I%S&Q5mI*?TY6pFHjfH()Xr^TX$>Y zhEq7_nl!1>f%MbHbZQSC)2VpH39LcBC(Yj)+B4ltZr6s)qIy2MPxea6zKJ8RQxYlB zZuIYIK{Z`6J~ZvPS&2AP5D(+L9vs$PrWCQ~z|2)&nKE$iKcRH19^#$11^nbu&F0}w zVZ$?X@C?C92G59{`%ze?Lk9sk{wefSg9F(ha1%r06+ac(b zJk2F!#d@HT=TWarv$XR>gBKT&TuJw(>In=dqs7ss29JO5UwrF@WpJo)w}R)ErDF;05w$|Q#d(dpMJ#4_oaHpQ+?U$g==}Is`BRHkG;uG zpH`ItE2~oOAl{Ujlb?MO)$j#}dEo9#YJBiQeRw~J zeRqejmTRoiE3!BxQWT!Wn`kIxTIEV(ICSp zhuYttx1$rkdZO?7U-xK!srcE0ip$bNhUIolgM3-bb_mJ5X-HUSthuW@x9&%^skFD% zVd!OdiraQw{Ct7`1_5XK+xCk~s@k4S#|c|^Yp05VgI0>7W-t;Ya1EL3nM#E8%rAGx z6u0wK#H(zfXhbJwI6o0oand4K$uPmrpvWf=cc1lW=>SaZre64U%4VB&lk9vc>GWSTqMUKNT{h4r$enKWd_t6g2v3 za`K7({ZwNV%)X`cS*N_@wlJ_Ha6xZgE3s?^euk8|AkQWQ(L2r00RbDs$XfvXEAowN zuADJ*?@*r~-P>ZcR)OwtQ1T+Yslc|czVpX2yu1Kc)HpY)X923J8K{3r}LF+N{ofbMlDRlea z7ABJS6MOmE6P!m#>o*TFw_`YT;5u#hP^(Nz-lC4k0&ofbE8x1OmK4OY&C-i%FzNTD}BrS5UyQl&f zD}2)pW-T26gJOB|;ys5LLPV_kD5!oVJziJJK5pyH<*ETTYk74yAk(^ERBSUmGXIjR z_FZijRXpiCS|I!6K06XX6FDZC@v3N7l;s)Q%LQcm;6X~?U!Bj(01$l_q54IL=vzW~ zCvLFMP`Yt3a%8Pm0*uihL=P z!#~A8#bDm#e2~Sk0+~C{yh?V1EBx0qc=8|%T>=NnAvfGIGI~n%ADFxWdnHd znx4N2!8y|>eopdVVTErlMXsz^M=QGpqgJ={8PR=!t3Os4h#-XxPiBaxPp~+Ch7l*U zf!ZB3&;y@ehuC_AW9y0y1{(JVi+WDzC1#tu?kgAc8Ooi6l>4+Nzz`At`pUxh{&Z0+ zaHngLP-83w5{(y`Ee}KOUF7Yl|1}i@R<6TWyWW7+v7q1cJqCjIc^y|hjAPl6 zL>=qFz0Tos-||k3F!p88Hw#|Rc@sJ#Vlm_V-gab-iIm=9FP#3Je>MW?2rq zwQW2`<@9H%eu5Y}YAJv_vi#l^lw3$3Uu0t;Wg$B}$jBthH&J40i^!CiyZ*AF|INg1 zASIyoojWllJz&!u{f;N|f?m{)13?4E7lE76=VvcF@ZqI5t^_gl`>M`ZYbntp$Sxsd z^OG^s=BQ?*!3&S#JB_gi;D}|hkQjmYXs_3`mIXQ3V?T>-t93lCkvOBs_YeIrmX>_O z)+qGCQk)ddJaw1SGpJ7Tw+E{PwYGdXeKaA4IIXK{r{*b!mlCb77EU*D)9Ieoo3j#~ z&0-qNPW*yx^GZj@)s`*DHuGHjVa*m$7Z}S zD6(fi5uLfZ1aPBLSlb{dA;y;RMi}f4AW>ZnM?hvtA^=Y?3$Oh z=Nz&d{R;`pw1=4oHI5L(X}OhJ%Z##{7z^B^IM?I?8MaG~g8u8)1&o3ELM6XfHFQu< zr1a9fioXkA4`62YWfnLH`1y=D96|Pu_P+_<+!_2YboiCD^k0gHF<_JFtC<3Pil=_x z@kea*7$~Lz^fhCNbqOkDOl`>|H$DcE-BXk*U2^+CT*6Y~(CAn`0wu}SVL76IpiqT~ zI8h`*6%O~w>zcGN4{RFo|A0`C*rJ)LJiGvu$JY9{8AGrv@JL)pn(v)o3NjnCGV45w zjvDJi%a%YEk%{VEt9Cj+A1YRd3_O?4+*}q`;G|8zoeRVfncD(3Zt847rpV09&70sS zHp}*G9lQklE5t%Xm@C0No2C^0`aAaAXZhOV02L(X{4DhYCYL!yKAC0$yN2%2W2aOfTPCL~Uw}XrSm|zXP0fiR-Kz;a1J9TK6X9tc1Nx_;k__!{NX3v-c;UW6Zb$mpz zJ$zeB#<8}EhKaP8L(0g}aFzMuV0o~?BhkChsdfG$)`4E}b5hyd*c~IBeX7#azx|8p zwtqgeXY#+=-YM>d9P4|-^XpV#oucKKST~3v~9M(0T>6Ac8Y$0fcz4^?qpE6A3 z3U2GYDS|MT+Fa7U{E6v;w4ChfWSYzM7=gYk;oZ;Mn z&hLGmpoFm)yZ#Z_zJJ;+Tw_hzF}n6w1Hi=cGY#3K9L_GrgGR*t@D#>f`mG!R`LrFy(D zsbde=xHex+mcw~3z!w$l8f{vKz%d_cLqKGrT{dZQ`pi(4G$#Q9t>wYAi$c6eHjGY~ zK-7DFeSei@Jt~M$UpJFGdci3{<`sFX*%m7QcZ-L0x=-uDtyTI+nvq(PXWrAlwLBYT z`%pxe0#~Pz#hGJ8yF&Lb+>!Z1d5-C<&mvpHkMmv8lWuMB0KL9r!Puc$(N?NReAe;r zsF8jl2+qM2X%apCB81!`pX&ZfZToB00!`w{&qescH%Ev03FXzaerk&rg}vRRcI!@o z`Ag0|rWsFKb1+x)3c)(;zyZP_l%feMzvpHKdi#&9kEC{UT|J?n!T2a*o~j}9pI^5> zQE#?*79shL&z@|s{$rbtPWre)|CkO-W-G(l$Kt+De4uLpuvfei+uGq6T$UxWD4hRF zi2r?~!?UmcFz~oK5J-=Dzw0<3#P_l^ECf)5Fu_e<2cetFVdZk;LZ6OtUOQD6R3_`2 z8YkcqKMs6-Dg};>bdt`^7oGjf|DCRoz7?pYAJw(GCv|^}BbsIk`f6<4R~71zpmpt( zGF&O~J{K!FR61p99^qi!I?OO@n(Y6HnNQ|}-Jv``Q2w}kkNGHZ(fX=6iOe0n3c874!yZUD}(e`wFV%?e}tKIjJ-+uZS*U5!8cCvt?NZ^A`Z_K&{ zW8lSd_5Qj-c1aE4wB-T;zeK#{G%_5dKW_B!J$T!sSZXPt&UGAz?LibLc8PpMe{_22 zNECh%ad$3$e~#hTNZM6=QGdR?t)>qptZmW06L5Fcw-omD3Q&6F7LE<(g{CJ2z5DoB6fLs(X4R)-u2j7eN*V(cs5PtOp0_jz5n>$= z_62jYNLrD;5wsoe9&d_Y|MKbcyc!+g#vU1Zs(*Uo?J|jZd-VNyjD}m0>Enxyk1#HR zUUwWuXGTjqTvlK)pUiT@Zl_qe+JQ9Z2!o(pLpK%G%i8{cen9@pGKeHG**>ScqT||| zI$35_Ur*?D+t<=LXq+)fL2GN1k^1$9hW(ZCm93%nC0V^{-G8!K9X?&oM3KU7HU<6Y zP4KAQ@?%H&gmZ|7X}xJx18)fC$l6U-tD>|<7^1OWnWS=VuZ7;j{rFU0{;k}jV)Hn0 zDC~Wu5oo*lRu;#~@^qTho%rx``>RjU8mC&`nmE<0aCGj~=5+aM$fX6Y0YTx$P}WVO zjsQfQK0l}7C5IWw#8$;vd=gNj^ik-IJYl8ivJx2$9Ia|+E`-2Lxl17V$ zkP&c^dA4drJ6C$m4v3`rww0LsYO=b*L^#NGx}Y!p=6H+7{A0e*;K&!#%6G}gZQ}BA zY!3wuCU?loJ7U{^*E7vzR`Y@(GGe}9cT8sI#>h6>IRr5=x&L?b8Q$?S5hyNHSVdAz0 z!TI3L%C8g6Ag_n?m2SZN;O1W!z3r^F{Yc+xB}#$jm!dPUMcxZ>cf3pVn4h!3_9DOJ zc&Nyb&E@03W&0aGt{kVIubIFO$;P#SF}b(CCuGt3wRxpaiqbW%h3RP)=X)YNBrzm8BxT>Noh`}8qD9l)AAGChJz(E3k7~;Plj@bn z^fV@Sn##vh68W6~Cbp50&NixHkv1BqW9#$z8aJ$$Hz_PC)E#k1UM~v-$lnT2_}A_r zf-4qnk8uaYLGP`Tz)(y>E}n;gVb7A|5E40PYkF|i530||5KeLHVeq~#C^?yDlYrUM z-t$A&9k;&-fgUr&i8gd1KvsFUgRgp!@4nhN$2lfA#_790J;wou**o~Wof>Rb zB+q9jYnMksJ;R4SIjZLBo)7?{#R^4&A9TdV8OBc0KX(=X*G2rbD%t+OtQM! zXJe!3RF>U#b^?d#2CCDeS2SIdX4z?%jn$mFbC>AS;*ji@1vuEH-Ge%}DpMfzW~_PJ zVcpJ-73!z9rU`DfVZs<4+LD=Z=7T2G8~L<%Z0>uq5VrPSrlzf&jknHnA1;66$09po z>LTLH4Ig$2qn38czeiJ7ub9wHi{CbfpNEe6f{`UAB{ATU-ktFnJ_et^xq9X zi3SQ1+*19U&X_iGS@iN?Z%#!ktnCXU;LBoh3^r2#FS6UjzFgD~y$~n`SA;>hlM0S1 z&!rq+zejoCKlMDNUyu#dD4o^6sCx5sjD|^JXl-vScbgAQ^m%DHQ97XYp_&5Xgf_-W zR6}yS!gjz7G2}4<=p@G+m+cB)zr8uv#Dkwc&s>B|V|tMj*f(b=c(D;0_mWR;La#kN zBIS~@p|x`=fJLB}@|B&8u^K@VpQxC_vaot;|i#BiW< z5qjaLc1cEA-u|`pS}03c$dzTM5?7th3!=dMXF3jIXlMPsNk8%TgIYf8CpJUbZOL!i zr6TZT05&sQPH2JaO$&Fr^cxTJ#`(=&r)GB>MPNm$OCUG*v&*_6f098LYUnAs@-3N& z8E$oi#=f&laAAxe``c)>F`V|JY1E-<4YRxB;?(SL2fWn?70$Cl_%SO57W87w>AJJ~ zn|`XQ)W5{ypT^?8sLV0)90{*WKz=ri{db4nFr6-PGXzeFsWvf$hO4kYa09R{JzjnW zxG`u^GoOqq>Wt`~nbujTs*r_gRy04a0Q{9}T^bn42Kb*Lo)1IvFi&$l+6(5xb#?5G z+Jf;6{qz~_7FqS`fANi4%N;vTqierq7I!F3>rLGo*|%wivsA6i`N`lfK-Bh1AiAkn zG#}k}CLYlW8UfZs_paOQ^G|m;-A{w4O8aPOL2kaKRkzuz6T@DwD`VcTj+1}nrdh_K zp+-q2_16AHx+&>NSW;&B80M8`tyntTyIuwSy^F5$f&^Z;;B2fTZn=G5>g)*e@TFE%q%>sGy}>!FNZf<|N_n)u9Y;p6Ko}-TMlUjX30gy>}gR zaODo^eJK)}E|01-K>3Pz&(JP11c3v7WZsm{wZYib&^}@DCH`@zwe2$ zvWmW<@_E*!0=@V%RKhpd$hRV**IWKK1uen8-v|2b6}rwal;crt$VL|7_Ho?(=3=ea zDU|!o+|0D@YmG&E)0SkDVvLj#?B_JcTX2Va=h+6JDb%ZBVTLC@_tn8)22XuBD;)@W ziEbLa>k2w;{&pF!d)cpFsh;*8ha8APp=kyu6o%N3ykaV9mdvkz0Y%OuU6sqfF7b(! zueLhKeO%*?fwPCTygcliyInV>bh5u(`LX3S>S-sqI}`C|vYAbP_j`a1F+P)CJ-^11 z>}7wxUxx>v@~prLU*Nl}<>E{Gp5N?LRoE~fV-k~R`NVRUZ0-)a*@Mq90EF8&nKs}d z?ylOIfkZ{VvTG-2b=jz`IHxBfZ8Bd^ywp9Gx=I)3_LC~Q7Xb1)D00zZw5VsI!~Q3r z>k7mV3gD+(&v4PmL}%duwO&_fvNt0z>F!HE2gWAtQg^Kr9tk`P*WExmG+u|wRsXvs z6!qNZlI-kTo*cQUwgxK@MWZ;m%)o_=jHzKdEt33G5T^pwSPU28<+oh*S5Bil0&kXA zE@yQgzbs?`K*}QJ4C28{YS~w!2J`) zesq&H&#AQAu3C{dDXIb#VkTUM35Vh?wBwSS+1dp%&qISqqf}l@`Lc-F%cfU7IMZSbXRh%R-bdo+Y&1WjzAEA70rf;sfMbblZv9VG)kQOJ zO=TPVsE=Ej?{R!#^jT~OnRO@PM)Ow++K$1GlYJ_DNy!+m|8kq8wRifBXufFZ*mX{z z?I##)YEM27X73-z<^V3N4pFPf(h;mDYPAO!gmgNqwQ4GWmmk>bgmBGQ?}xs`iwuqk zZ%kJCVWtwrw%K@p7@#~#&tBUuLB&e^ivTwU-yl3*F6tWCOq%Rg0e zhC@~E*Rxsqc)*MHkw*!22p0>FaK3kNoMJ7Zj_|Xi&aCsejed-`)Ti^sbG+mKXgUvf zHs806S5Z_^rPOYzO|YKvK`_KX?3YFDkQz4umotCiSl6RWX_#Ekjo`#X;J zPss7)dY=0_@ALee^;#EC@bAx^?`IpQ=_D98xi&d09ZS6P7s%i=iHpx2@pC0y!?ktW zllMhhXA1jn+TuMXIh)waxgXDCA1$|4&QxhAsd!1I9WPMn{)@ElT}}el)n?!FZzCG2 z0qDr5KPNCs3TD_(Q4Mexe+(7zF-LnW|2iuWPv z)3$vvNeidQr&eUA=3-|V{Iw&wS{l!h!Otzl?$O?y=5_nro|}U84;|p?aWJIULAA#3 z(b_9AJS8!a*`9lTx^e2O@ZGBFVImleHby8l*(o86P%*Of>$$3j1c; zWAsEGb;GfvI#3X#4m;B^oDrFd_Ie^NlyiirI?Ubcu)q@Qd1DA;cC8$P)ip z_1ASTqWq?GGb!`lWg=xr9{q|s4m>eFg3oCxGMm@0G?xZhf6E4>tO0%7NcPHI1gBP~ z_FG;44O__&IcjsOjAgP{k-!9kM{C%Pos#)yzx3nc?x2g50{k3H9L2$G!j!e@&+1Kc}?U9q_)z)YQ; zEN|Nf=jS^k{@mQ$iRC(@8@?O|bL0A-Ajc<5P~oK`{YsuJy<{_y6z;0#JaTH$`K$L2 za=!H&=Y2c00`vr@;~X;dSj<*x^ZuY9sE84pkGErS6f#2VSUJE z^kULPx=2QHX!+>+^BYT4)PsUi{>krg5F9P3Dgr{J!%Mq)-^)V2fYP^~ss?ngGX%eu z5wC{(ZBSe|E1$(@yJ#=CZ-bytOS8az3@(hG`{PP$^w;y5dpEz#&-8u+hlm^Qql%co zcyxu$t$CUVa_+F(b$0VA!l+*DaR!SLk=aykEa&psl1UI9VH_Jxiu5<%oW4S8tEDYZa=paXdZqDVa&)R`P{bSY3VqU&6GB}jXr9y= z9kyx7;FJ!zCyda6>|U?zL!B*yJc5CYgx}qYmMuF+e&61d|C(D=EK|~}G$SzbsBzxY zxJpC2$>kF;b{Z{sT{sSSpqx8>A!*=Yp-R=FMQ%)8k@oOw(Ou8z3LlC+)|RaWnO_sXgl*s1Mg=CEKSirsOL5lw<=ST zhleS9aB*=*Me(J*FqD=g{Ny*jI8HA;AF{i>^%-?a8j`R0$0?#6zag)yt2Z^4!DjVb zc=x<@!}$rJV~Xyczh!~MuW2PQ?eTn?InSO-2|x!JRtP5L0=0xHlK%4@$-G$U9GM?$ z8MdURGH$C`76l&<^`S=M6^${iZy%{m$^4aD_UO3u8&rKNPxdq}XNtA6GcvIl-}sDO zHT>ah3ODMfe7)1|R-$N!-N&^ogC@J%%e;Nqb(d&u`_Ygij0Ie+n=;K!(4)k$obvZD zCORcW8`4;E(^b8yKfd!*H|r!ut}Gdm>u6taQksFIXx4?rBfzkU>N!eqQf*_6HkOi3 zeop91NdYqFQ^k)XY*+Ko<5dS8mKf1|qg2vil<$ zF|d@9@kr=l&e5Q0#k{|cQ>Lc4W!7g>R-^%`C2rjE+{S6i(SL>LJt+4Ukr$*;}o9VI-=BdH z?)L#Im|;F7!0)i$8T3G+{ZRNjX{$HfXFE9>&|1Q20lG75V>XFlH+t=9$6^xuXj@j! zHCXCFZ#tGoy!Kv&H%{E*pXnXKbP**?38U()!iF(hL6}yr*~33vk#$ovC9-UHJR0q) z9lLUSEW0o{%QS`g-}82v!fEROFrKjk=H!$CGYM_TBr-srKGF2ayoL zUrz0r+Nze87_l1Qy*AUmHsY+-GL7s7x8Rub8dd1F>&Ej09^4gV!nzZ|A12=Jns`m= zsA0-T2Lz-|@_FfUtsj%?PVN)0G@Elxeam^JhLq7~5dCA~-J6NL@z?O}jb6GPjW5h6 zqES&?I_nSY=QgR`I~aS1qiNXju-<=rW~Lb-bd+K=#!;NvRWfqB^26;9g>{M|@U!xQ z^}wIH2L7b-eHu6EW*1GdTlS_IOH-R9Trz(x2|3=FxbCeNcu~et`UJsk4ji-ESmL6t ztIQj=x!SuNexXWt$a-2^cWEW`BwnlAos0ua1?k=3=*KE+ttN0>KkFz@8IOSbsDLZ$2b z5ixex()Hrb8{4GJ&VD`})X*C5Pc(7W4{a@w@Yy?cLyHkF>dx9oyIT9~z!PB=c8SUz z`IRFbcSmkp_{1vA#?l|FruLb){ zqW@Lg)j0_#OdPF163(@%{wp61&|^y-OEUwm6KUAmnflCB@B2i6h4z1bsQ@)r%<7XZ zUF4EP$dQT_a0UpRYjwFC$~I-T-rZghSd;AdT`6DX7`kkgTN0$`_3uyhgGmj$=g*I#QS=x$JgcOV;DqSefN8PN&~No0~hhCnCAL< zs9!}PUp(l|zUe1ooVqx!54f6La^g%BPS>yMv@^i`YuZZ920`@1b|wX&M`wf~!#cG- zC%smP5MR=jU1`uKh^jWIgbEXUtXA_E>V{6%$z6UX?r?bp2 zdNlOgX#*Q2*bYgGgYrLTL@MMS$Pz^tU~MR_`Tf);9+HB;bSz|LF{?SzP~Zn++n^_gBdC1ju+;uuU=xpyeXyY}KfUlmFq z%QO)^ku?#0=S{Ooa{EBOYsuzpcVrlai(C$>8SqZkyXusxbBoa$f&$|HL?1r>v;S}^ zHQqJZ^?hgOmm{6zN0lY$X(~k0i4&jR-T4uE8Ml#(Y_%9FKV6<-FlWEL^2sWXOEFv1 z3Nb3YT8C~g78Dcqh-_DZzv9;X!>?TX#&{65dlMKqSnr&)E^K22@TgU7tqIXCo;20Ebl!WZz)1vQjzhBuZ$=#^N`KEm!Y-{NI z*lf;m(@z3`sAz}>uPJicJ^OM7qjCQkr6R@A?fZ^#>Sa(si_^M^aay+})Ai5RFHHr; z1~enZ>YcwwL_HNitmW%_goYk^3?I*U#w)E4KaqNGP{7Q`MJtFjj-eDl{n3wMwMVs6 zTgy+L1btVWXDJCl`mNeOZsb^gL+$;3y=Vl-9Db5@(GZ=2Z3SF*L-9d0i$)h)Rac!- z>-yv_DrQsa%dCzV5a+8P?W(K8--!Ew)6H0n=ngI8aj(8C#}BV2^~AX%4+L4NDqDlt zqiV+xcSv7!;ERH02pINK{Yqg`mg(iCN<<+Cw?4$F>WClZBu-hqxG%l0@Rp5M!@GAA zmbXNmT)s8R%OqtES*Sbun#NyPl{5=+9RxkEjZ9B z7p$TW!F&?P8v8V0W!A9Aprh3~vI=ky?&%Mg?Zz8?i$A5a%IIA*)^_Ox?VK35_n4I# z(5~6PJMYYUHi&>Nzc#I`PN~FN4lYq^T#`*E#tSp0^10M500h#vcCipH0-(&4g50D|lr}U2 z>!2+Ijr@5$7{ZC*>>&ZMekY2iqM#xQ+^O*hZZMdgKQG}a&hjR2rklgnJ4cCoDycn7|2=wNUZ z{%vNqYnD_M68G-X#{0)A(WakwE%<5HfkoL6)Vmg6nOb7gRfcxOV@(d?xr{CNR@v{} z-GqX4E-~2J`jq$|k$X;ir8;b7`^RU!>*9pAp4Tr!S``}K7>sJu&$1`Y4R zTUp&Zg=9yRqT6GGg**-`p0uG5j!nwCXiQjhi$ldL(EU7Uk6u636&vN$fR5XX3s^gA zTRRw8AtI2PB%jMF+P%#UK3XAQTOe(~)y|+H#4AZz!jnkEbL4PV4!zq(woLnm^xc3K zM5OC#Q18c6{DaG+Z(n-2Vr(G54d~p{OOLu9F4Oz2Z+hQ;f_De{d6+$3TdlsOh+%)( zPOx*(!`p=!&bQzFsnd}$ImWFEM~8)!6rgk(gz_ zc&StAi+L;Sr92s178mkgYuiLlE~Ihj5VEbHSKb9PI)4`9j^3L*`O3;x^h3a+@(OjL zlxVN=?++0LOtA4K+7}Glh^^_t3m}gbZS`}M5Gqe8$mpde5}kn znqitxcsSR8TqT+8RYyDT*=V<5Eqlj|?fePd$0vZ8&L6`H$LeAQo5RilTz1?|#7%@t zgY@pI^&fjb8@x>O@ObwEibGw-Qx)<=#hSZc*LM6KgF)26`0R~7gr^%`_p7&!JEc@E zg0UBEg#-lN#nJ^1!-LMsUJ+|d-fO{!^WU(@z6t{YqbDx!UlG4uvQYb~|3UNN$45KV z3tw57ka0!x{^*9fL#cQUKZdKDy2CqyR68TbxjhR5ofzIuR1@OLU4yqv>ZA(_R>wH) zM*xtE^;I_yy{(^=31Bt#G0T<0Kdwiw5p9u1OtT%2PQ#B9e~vN9E2iqLoB8Op1L{{D z=tJt#ZC|kCQJu=SNDe65j_LFI4oKW|_)Qof9(8!=`?Pt8x`rKAQk*@1xh#)E2QOEj zY;p>DEo@%{0X zt{h5IbfME>GSO|uRj~FO>*u_|q}rr9o=+%$?~Ot){+eg}1l`j9?DXnsz`}}Ur|7!t zL{i>1saqeb@)agSwpkE`=V_7^n1`py*h3vNbJ3{@xa`+}65Q=(8!ofYh8v@%tCQG0 z$pRg_8TclO`{>|5ev3SrS}zI%C6dAqCnH8w2Do#2XvVjkcfwF1Uf#&5?_m-TV*sCaQf;-?YbQM4UFsxlT(F;|%Op*S_VNuY zFFvONr`VErjOyp~t_C)}G}~%W)0aB*Ce!^To23#N=`mRrCmj&?vsILSJn^q%vo^)i z=R;60KmL!eB-T&1DB>EUR4K71=OOyQse-Zx@PO>Pi(JK&%q(B|H1Rg*_;UjHnL^^N z)bzl+F032#))!+gWNY{-Hl3a!DZg$_3~t4Icvj}S#V7XH>T7`h1?>7sG;$D+@ASXF-#8^58o>9=?jCLSfcPU!a{(s}jv2#lS}BTdE&Y9x<@ zSQ2;SY%EUIrDmwWU@0Y&)CaP#XC4pO4m{XF*PlQ>N*RuyV_X8>uwC|ud~ZEn zYEmOYPwN_c+fsb*{0_eB;WcmZ(V&@9oaRf?z6#<7QDrYCJ{zmdzHv}_n?ODAXQ0fU zjRqV1wkz*J|9}T;NYwyQA2b}co0S#Q_o=U98_$dkjWw>`yO6*+MV8kzYMt^nA`3mR zpH9V7GN`^!8vREq7d8IVRHoa@<+@Q5GS2vbvq`Jjz6(B;^1>OYFIBiv5Bp!Dp@x-n z0NJA5MIE9$6vjH}-@vGS+{a%tD1%Awx=l}E^Iwz5VyCkWW?@vNoeYtx2qf8I&*~1l zqAi~*)IljvN;Be+@K3%Om?YeM-P?tV-g>e0@<_jw;PNYDA&-k9U7Mg7V#C*;0N$tF zyM5_*k;AMJSrn7N$Nc3fb&|%3+l4h;5K#^8@Kf zxPz9@4%4l{!U+u@Cxy*e8F5}{G==xc9~uD|DFK6KqyG*IjPLFi7ZK&&X)1N?ob6~c zE*3B0$r12iwLxXsBx7&`DaFt!UD+6QH=F-`ucC;x!1)F4fT8X5?XY>nfR4Jm`=+3W zXhVM!@R;d~fBdZ58ES7%=F7{Q_-H+vxd(i2I#0$-e*$}bpJ_B7w;r2s;>NFsG%GTA ze#)nd`Tm0K``Z_VuPZJu8=T|R@=I9SpeirW6L%W(UBmJN>|I09gzy+W%b6zNM(7$z z_9ZIyE9fobuK2H)QTmTmc`OeMXe2!5UI8nxcaKi{N^WIxJfWD;-YIH#~zxIT3a({1fl zPJCIM7(N;1YR2WW_B>ca4DCVl!G*-h*sk`kW>Zps1p|@q3x@ZbB9B-CnuJ@;>gBGp?i$jyMbEdj>xquFSC14bI@dwj?zM|Z2v{!d(pFs9&4-3jZBKgoNG%$OX3SSf;XVw>fWk%QP87V2VG3gM+<Tz{?DN&cO8jcTPIa#0Kw1k) z7m5%{MN2PUanJ{XUvwY?cun1R{x#n%wVI-<2U)+ktY<$Dmz&CQ+Z`T>#%X z9Rx!v;wp3e`!tvkIYx%r?o)yhr`@> zanUbX{6W&ljn|AKtDEd&+@=^PTaCfWtDRR@lxaUBPyP`G2cC!oPV-o(Z4|s>9GJ-b zi%(GH{H?tK2(o4dbggZO3^}6;mxz1G8!jrAy6<=5?@C6>sYB)=>cj6B-s9L94XTqt zJQi+isOJRishZA`W_m#?PjSmV4YyZdM<&LN%h#G#@$5_{kh|<`Ct2olWsKh&3emzU zdE{*FYDL0(mIA;`THPf>S5jox2j81y7Aw|&cW0*ep4twJ4|m%?t*btjxWQDoZ&jdQd_ef~6PLjL=6?hALVHZ!3tnDnyo~*SMOtfJ(a^z-7Rv8kLM*M} z=Z#GyHdOn`!s6tFqibJvJ-X=ZOUv7@tKiL)G+~5Bj0R`p<3&ZEtqqsv})tcCyA0$A~rjOvUU;I4R$j6BoqciPPgWTkKS%w=%u- z4sKU`fRey^yw=$LUwnYLhZhGG=1OD1;fpQ8af~(LSoY6E;f?+hX%nDB%7HE`2zn)vC8<@ed^BR@FJO&bMkAw@?XAYmmu&b zq5Z|&MV=YAq`J+`tt2A$8^!ur&CWHluG$C3IvO3_(w=ihDW}nY-68BGXDT@Ry%e74 zOr(FYnlNDkWdIwE<`4^)-S}zbDe)cV5dj%nCefqRZ10}brT(U+qfY#}f|cvcI@hKZ zFK*pJPE)m{vAf-I*euz7Xq~~zY->AQG|=?cVk2h3YE~+qNAK>kp~#_eg#%L<*L)*G z9kEr*7?`otkZZQgs6D>@yK+v!xBC^=vh}Y_ahnfY;e$f>jQg2zN5839=4}|d#$~#c zNjGjDIbb>4A=$pvYS!))=bNf!8Iv#aklrmwT>R3Su5ZvYOB5i zr6PM@)9ZDuP}aB=+709&X3#pb)#=Qu2}8J*GSPn>_au&#*6K^qPZ)Ma`A&sa+dOz1 zROvmpz)e-{C+=_|{EEt*|6HX@>STje`h>2Sh-jk8@ndCV>HWm|+BzLQ5G1ZTPrSEHif_1>n=-8@im%{3EMTO=A(-Fc9+8-aN+HeR=4RS1cFYlZW|1& zS6X6-B<)|lVwVL{6jEKGayN5>RzwZCZzvfcFeCWXC0jp6Vihyi7%k6OoDK0cl4pCs z6mU9uXrU?ii&Y>ufJ0mrr9KZ13u7_oNg`I{@W<+HZC52FBt5Ib2FmU(re{rWJ5^YA zU6SpZU-|6SXQ?S%N3(ul1r0*ka~3NULqE~plkBLhle z@TmdP6VY-f#)Hxxx9WRUS?&GCca8#AT zKHQ}jtAz<3$%+E^|IY$wMba5w>{IYz4+?X?$MNRa`zL9Ju#=~ePq zSvL_8^D(*ypGQG?7}2AyWxZ7i1PnLhnpb1&b^V2AaC5* zRaQ7*%_<6(v&PJ&>-jWLS4fz2{RW85E%%QD8; zxXIlX^%@`Gy$_a~B=l;a32~S)_YVJaZ`CMBEY~m=9$=#r#be(*=%%tbr6yU$d#8`q zkmxbM+^A#F>Dl6=`D9NYaTXf%e$ns@t0iZi9$VC<1??1NH&Vm>AnuqRmS(&NKvqwL zX|ElVxg_YR@Zn(D*BO1@kB&Cm_izCR5tEC|-y2b^(e^&<$pj<^OBLnowBZWW+y9tE zcCp!?z-50MqO(w@o0sr?#G>JG7FRcu*StAm1qr#>YYy{Fd}I`F>_#EF9!9Z zj(|A=8SJ_3Aou3DhW>9gub|^&rF)P8pJ2%+d+pCH!zW5KGd=@rY>F01Hb13cY5}>d z@JZ?b!Bt9DbB<#NckLz~$3MfbS7eoUG<%C_e=z;p1vLY6J0O14U+D-L(RdhMGDvT6 zWk>!X=&8ETF=_lNi~&=})!{Us3#W6g5}g0J(D^tff(hbXe;j6s_X&Bdj49Xz|Cd0h zTDm==KOxbPt%@a;V4uJQB^6MX<3P1jMEL?SFuL1(e>J$=qo zSV*QndiX*GQzq@{qv-O_Yrl9u`{McvU~GS}C~`{XPH^*rWI*necy`%o7f$&g6`0Q4 zl`tfS;RkeFtTDW_WbN`pL?8X+2g`R2$$NjV@rv@YVz!H##y{oXb?+yU zCe_$STe^l*@Li(zH-5aqN)njh2|Uz+mt4dGdu6=pC*g7lll`h5UlrGd`XhkH4Y~1f z7$>pFCA2A3`t**fh!}G;;$1iVHE5-^;(Mki%!bhPMnz;j>^<23{%0}b zeYtmgHYLtCM~8k_@ltDVMTn&bB6dHH2#7Sm=e|xa-t?u7&?3c@XQ?b5u51omryp6? zzwU}ZqqSSV=#tFwe=2GA!1}41P_y$R;|rtStFeulMhlzXkMsHr?p5@;!8t)W3KQhj zB7D2de=xIFT<~zDRF8(YH)|Gn{Y}DJ`f5ir3ss{a+XuU<*7k9hV8$Gqeo!S+< z)Rcdm;GIQIL~hoThwjxSx)I5L=RCg1uGn7q7k@WxFId*$^Gh?PX4dCfV2sCTd3dXe zwnokWMM_c_Hklr;d_?Rox&_hsGjy-wD^kq=NNoJk)JXUf7}Ah-p(@_Gp7+KxZu1r= z^_pam#(bBZjXFyepbF%yROL6;ZQx9*cCq{T{OT~^c}JV{nXus|Z94QP|1}BPuVF)y z-aCAj?J%ZgB5FdwbCYhaStXM->+Lz#J6JB+D@Ksg-=aWg`Qu~HRKA3DKZ)dFE4*+W z1%Mb|zIJelh_2}t-P?!*Tn`*NnnH!37LOIEWK*%Bwxzs5PrX!D&-Cl0O0X@=%)Yoc z`Mt(``!XsW#_eok?ED2@&dG(c1rm~M^{$4(JzebytGAh;qZ~4G0(*v`ZkLDGg+ z+o{o6E9AX(QxAW*oAUP7HmbJf&S5eIZy$9=fPKiM&m}dei(?RLiM3)*ggct8tv!09Vm2lHmn$#U4%!^aAZqf*1D~k zyBv43H&=hKS%Lp&tBb9|3tRmakUzOQc6 z(go|1T@@$>^^r6-uyo&-M3>J<6S$l`KhCCweZ1u`i@HiB8t ze^+~6h5xeaDCpqf1K&A0wP2>QlW9CH*6S(>zoSEpuH@m`;LdoqoqiZz7-BND^R&=w ztDLBy79O!)8SHgrf2Fn$?n_@)D$H_Ccg2a1Nlxgs6GYG%5)`*b#9?B``b}R%JtxK@8?JN^r zfYFo_?N^rES4Jf>ipQjk$F#8%(bNiiR;i%bE!_%N89j_RE$q##>(o7PH`=lTXx}Go zb&xXspI>pJI#@_3h%ExJy)r-ScWG#GY*X;(MSn{k<#*@6>!tJ&JN1C`1)HW#D-N@j zH-Ju2*TT@IEwC|==dO3Hz9$zJEl?GX10iTv85h+suHz54|IF-0Zb=D2FIz1B<)XN4 z=gIKs;z9eka1<($>O*Z<8j zq4~m<$VYd%Lbhk21<)8;!_{)o9LOCXaOkJ9@Wo4~QFnjHch*9!EEb*od@xyurpbS} z&TaG}o_o&3-$a7!zX`FheA%Ij_Y2eofFw=*Ogk~tr;PL3*&v;etm4P>dfevR)0M~1CVOf$e zLdURz_m#hYH}ziF_yY>?Yw2DTwrRw=6+lj`MPE{l~fG`j01G#0Lp0_@x> z2>TiEE9%WArzhqM7^HU ziajY66C%B@9I<*3#Zk;bm^!1X1B>w0d*!(sW)0dDxL+zY+l?k#!5 z{Y+OIIx~bNyBe( z`h6_=tdonj_PxjQ#x|Nj^&hqiCBSFIl5v_)x)z%mgPS^UjeTc zT`plwwMK?`uBvG0{jF14mkw<`l+)#I8e*#2 zzf3i~-N5I_J!=WN)NcubI}-e+g<5#?aF`;?RTnq)x~_EUzX&Bv%qAU?`A&q=y3aZ7 zVS}Dyb<@IZui2OYwgSTw?D`pdOpUZfraKx}_zLe9yP`GboJ%0NOj=6v>s&}|HF>e4 zkY3;<`zeXC2h@zJJc;}&@xzekTtM!KD2l z0}l~nqbjj|BhNfnw*{~C1;m6LspC>qt9M)xSl4W^@7#x&Y_I1spF{7+#H#zhkld_k z=DKpJQ~S%mdo%LkXS8p!aD9qxiT+7C-CR3$JredLStB14eT2lyQv+ex$FfIABBa}e zrExUXerwWS%lB5G;}qgNh;fA)m{00ZZ|BzT01VBH>GC&no&)eupN;%dw|Yyj{WxeH zm?PEopTsJ}*3V2>9LjjVUR#G<;)ttazSIsf(_{eJHa6TABD1-#f!oimU;&J}Y&yjA zL=;>MRjo4!&$kW>l}$RnQP6Wf`l>UJYk)rGBc?x17BU6uG30JP zHXk2bDDFCxsD|imWh3LHH9QnCSt4M~O`EvB$a@<$Gh}r2;mol38cP zR?57C))aF?P?gl#`ghKy5|77kAsB?~y{5J5QakYan~g}jqI=sHy6&17@-$&{kGfa< zs?5f0t_M#erm@Iu)7N>GzbB(-0%1 z1n^|^8kcYwUyDd-JKyJ?fR#@My?TE9`X$mPGfW!l=-3Bd0@+N7VxB7v-L! zjDH!zXp3$VPXzu%jFlkJpv|ah3pYA??`C%icG=fvZE=o_CMwxcj)#cnN zh8`p1cpjkka^6PLG3 zo!aM^;_mC>Uc8S$z1vI)Aw>%K==b;uq5`44=98f0`6YQITE z%wj}Rt&&sJt2aHGT9tNG-$Kd#wPk)7NYV>WB$>&|CRvKwbtU~;C^>brt8uFPGf$R z!G4?l%v$YvGh4k|^IPrS>}k)#2sFQFHEVTm#8cK(ScbEJMa$L}%d71!qSs7~bPUo) zdTpcKKHUKBN#gC({z@lA7Nl+ltWl*Anp!)&W9zj?ne!Wa-GAU~7v(2kh?B)ouijat z1$rhYlX+XGwJe1~w>)b-Fh8Id&do*3H;MJone9xSBIiy6pD+DqYsF0dWWZ4Oy;qlU z1@@hPo~7Z};rmFa!BKJGiY8V0bIi@Q^eR=&=x`lw1NIpAdH?)`kyq%=MoRT7@o$^9 z4=+7jq0UrXlzd~f`AHV2Juex=%OIjueoiv{y_~|$u1~Uo<`@?%F1$`tkne}I7R;Zz z+kc}lr1gwj4tB_)QkDopIL zi7KA^%L#MlFoh^{GGcAJVQFQ(uD4XKW0DukjeAWqy7gsd0_C70&vgX#Y;uQz74$ES z4yLtW+2{JE&cMpdXs#hSedAk>5{f4iSQ^@#U+OqVoqCFE^+Edkd-z)qqZ1-e-73xr zqR=FtMK1FtSeZ1Q!^n;t(^aY=QkwJLm@IejW_uP=dv?lc)walIt)B6TY7Td0d%U9~ z0}X?HSu9SW?p(tltjJk<^nR{*l!Ki4y!Tr}=7n}n7@qeye5 zmG4f8dm&IRUR1L4`9bBrfuc6_Mm)E<_Ay4g@uvZsT9#o7Z)ZNO0WZ(D;VRE(g2h~{ zYEgt^yJS6Nak~`D?I%n`Z-&jxyJ<&1lV-6+v6=4)dEzt$mkBm;RRL7TIMfa5>vijY z;>6%{pY1I?u4~w2S36Vxz*jY#;VP;7(ou+Z(XryM0n>P_9O`YRz3^IJL>V=~tAI`SRPCZfqL z@x>e!_nGjO54P2UlHGYU#I_S{eYrkyXzBdi(f2;j&nEZ(8_}<&k0ut5+uGT=fmBSe zqmOlUE%hz@a`7^$)F&(v-_zmuclXuMI%#gE3U~K*WK=XCmyuqJWgBODYGs*fLxEuu z+<(*=TB>KFsb0xHS)Uys?=xz8qp)QD4bWT6O=0lo^ExdHB@|Rwm-*};E3h|CV!PHa%wcOmGB<1CAlj68`d0A8^>>WEHvm^uyF;I-xB2W8CEg8Et7V9*m~2# zN8z|Pd$=uP_{HoGYoI*gXS29Z{X^MiAua^yqx27Ll_A)5fQwlvsjPyWGp$#NL%W?F zK)p$Rj$FODK(uGy6*y>D;*%|Vm7{Mmn>Y!^vP_;kCc;**UR2h*Bs!|3%|Nh<*pdBU z3SF7SOj)Ai{sW4fC-rV3)lP|V;pHbAg!VrP`)LcAf~(41!nL!KG<-1Rgq$RAKU^}o zbSExWPT+N7yVM$QQUE3&i~c$<>GrCkA|kc;pP40@0(ZkpJ<^nrerP7*sz$w{2KJ1T zsw4m6f3Kfrc}#s!Yq7?s#+|d7>hik|7dNjBExqkCTGfU4pU1=Erfrhs!0^uV=t|g~ zks7bj9K>x9{oD!;0;eVhg6%Q0Buy!Ndi}klXZjDhhblf5IzKdbdlziW4YW6mZw$?c znDb+Evl{a2Wjx2!Y_fJQr1b)Ac>h_uL z;4AwkQw|AW%Pujt>Vk!YvRE^#S0MoBR_vkgNlBpogwMmh2I^9`UJ*) zceYj3U>W-%l9&TU=*{zm!;r{_%AU|ZDS}7HU@E7p!joq#hx|AwpFr{vjlC@ow{glT zoawwKAbo0L9nxry!cjU7MY^PxH7~C1Ran zr?BS+l@{doEW_`Wpa{0hJBniH;)XG{>ipp@HAa%w^>FiVbufe^W|g)(|CrRD`#X5G zeFfH z(+}F4w@~SADM`p>7bDEp*$GUq(_1w$ITI8gvF4nR%&`7ms$SFvC#m=O{b2pDMnZDF zOSsUI< zcTLVE+Go{j4&|>wKX_wYslqIdV;p@qXg`{QQ&K(6Q?&}FpN1; zzP}i7r4GKkE-A%1#OvjocCu4k-4o2?0e#wj?pf>AKebqmG(pVXtwcab{&!mL)5OI# z>72*$C2%sp6m=UfKctCcDQzi|5XDD?YMUE@K94Y{oQn3*q*4SQV}qs7ekho8dLQxQ z&@J*8bmXcvUhVGOadgyMLfdQWJ``q^r_djr6`<#9D z+Iz2ct@i~h;QKrL-kVh?^Bm-8Q5R{yM+~IO^hfs9a;!f;-DMC3mmhlzxD?vahcaj( z)PBBexiC{Q_6Mj7KJTX_PC2U|zT)pp@_6=0o|A(7UXJP*WM~})tm~()y2$9YTvy(& zXu>agq+6_$Q;HHNhUU7FIloy`85ALp&qDyBVFnTc8C<5ufb(=)m;F2wR~!LUVkXu0 z4Ngu1oNXHLu#d&U<8A9@k5LonneV=%<-P1!cuH){%^%#_n9DE9qe)HKniAFP^gEwp z%lfy1vOZCK4wbkQx*BQ!v(Gog`yD9@{2~_?^Wu5-wv)inWi8hy{izpKlN+PZq4|0fR1L zLG}i2$)nT%&0LKa#5ah@`9qN;#tOq?CucaknEvvWn$UV=9~NTxZMj@1HkvzkPR$RA zmE7@qz{b6U;b2vO@ZDL)?DK3@Nc?AXBheHu8?OgCQTr9r+G7P5!OV()e;Qz_-w2ZM zs64McBgC~i*P4n{UB6>mt2PYs_E_9J%69lGpsG9V2uWl!@maFG-`#P#UFlpgP7H(3 zSM8-v>O9%--s*UujxF-DaBAyC7dTty#WX<^^IGzH;e?aHmvFV8b{9wgHI z&fWerQF|Va{bAMV&;98h@`VBkO(G~a7`f%V&ROeA@V!f&c|X3Ctd6diN&NBOGYQr! zahnZGIHyrp*L4wh%28EYCt_)>E%{EOoPJM^_D`%*n2Ps{w2uJgqwJ?f%6stREeChi zRj?EN-MFiPQFVJcSKyEFDWws0{OmZ5m+X9o2fJ@TFS>3Eo^S(hR=JY6{WtL)Tv9wv zzsu7vame(>ws@mbsnGend?b`uzFntlc+qrSEc}!&A4&+lX^o(Nd$^F4Xv@ z>is%u9daA_v&<>s0&ll{yp*(D_MOPM(y=#i*QUj!VNy3nh55FQy$ZzbjUdI$2&>8> zSJpt%Q$S1Ma%Em4p)QK7HZg#pG3HP@Oq$|Z@=IUb@6F~Uf3x0|adRm=e)JAR(lQ*H zpvX!afBz8}o1wk6GhBz5(WRqIy{lzH_8vYNXzck@q@w*mIxNk4j-c;F{reISzns*Y zNywL2SEYuD`^y!(5u498odHP9TWCc&3!uE3&CuZZmwMCVZ0g09XGjCZ>+p)UavN1$ z`_;L<>ph&a12-r;evsu(PUp9$+@O7A2WGKT^g=}%7yZ$dqZqM9+Qdi86tQfuYBTDV zXLD{-z@?ajLydff{_g`BwQV9J0+(~lKovB%C4ch`-q%sIHxJ9FR5=7Q3Ur@ zZy6o=6jyOOd|<@MDov)$(oUuq%P_mGxBRls0FZihq(+wdTkKQF z#LM62(QJnC#QZNP#lT>v<`Na7XMgij8I&Rwma!IETQwsxfoGhjfxJr;oAAnfF!IJV zwc`(9FowFEaUZwrWjm;$?qQ=LHn1hF`zxW; zY;UjqFsoRz#jwu`S4N|Cv^MI+c}BZ3y5v{Go^=;Yi=C`5YMjs2>M1ejn*)Bm1%H;k z-5z_pE@#3Te`-Ei51x#4JzXyBbd_Nqd!_*R-Va|$d{0SoI=>O~8BAP`ptX{nchlsB zkN9lO48(3S=m0Nda#H5Jt}J4cR)F7Q{>HQV>#jWDkI*}3G8X)~AlMD2D3jg>Z?aW2 z9yF>4mSmlX&YaPCOd|)rqC;3dy|St^WO`3;M*Fgvw@!~nIz>Kh=oZfwe#-IpAFXF+ z{rPb{gYiZu=tjq(e-uv}uzngWtI!swC|vdhpJGKjThQ-|4=fdy8Htd}jZ*0)djAv!};W3aK5tMw%Ha!lNlv&}&h(*|$e?*za;QZH)u z>>O1w60@+PakbXaUZlT_M5Pkj;RV+x`b%9pNSc zJGNP-ndW@|y&(9!Cj5Kk)6piYcmRM7UzqQ+WR{UQJ)hf$pPJm@@{p+9-ZC6F^J(@G z=0DEOXB4v|E3N$xrEb~U#`iO6LcZ<3`Oe6ZWugTVCZVT0# zS9yQAe8uv#;dvrR+}I5;x==D&%{aozOe)(r+NW`V)|Ox3C4?~kMQeK>Ni!2QJ8^&x zf&A>HpuISrGS!x{=2>u|t(l>71U?+x{mUJ0)nWB$^~zr)RD$A?hvYaojbz6aa2fCz z2w_xhzQ zYApAaGNq&SL%?PO*ZLdAEV$Ymc!F1ZfgZg9BYmO?y}~JaTmtPV|H`+(^3~SF+sAHf ziK%OG?;7m0n~lM1EQ0X+sVY=t?rcOywr6~wsBbdyloA)Of?jNyx{21`qVD_6PrN|ekEd;_VPN4u=D7KgSp^mUhdgvw6hR-H- zkn5uxMubxjF7aRH{nS;Vr=C)$N3)MNe$L{`&Nl>GYxn)Gk7nGmiO&N#|8GXnGCwI&LGihfjEZwR`__>$~p z14|7g5kgiQM&HF#JtvM%s*;l>i4@&H`a|O0#TXk;zCF;-(Q_}e3 zxMN=C$!KqsQop=#&Xvbcq$?HK$(p#0OT3N-e$&-55HF``S!?+N#p9C^qIM`Bp(tb6 zh1w>()9Z>Q0Ph z2%)%ohTIx<2%WrpS){og`c-Lc<=qz%xQ}ybHw8dgWOV@Lqk!|2G`2wN7tK0iWOs43 zU+8UK(R?J*~{(~K`sFk}t%L}!#Ik{o-zw_=h%4)QDt7debkr$@i zaBd!(?{-K#9Qx9@ifih#jSo@~$jGnX(i1f&$1O>bkmJ0{da%NfR=$D5G9sa5?)5#F zmtb%J-CD*bGrn4xOGTSRzY#iP`T>vE^<)TRTPOHIswSML7u6H{0*n^HNcu7( zWUmd3z0E}Tt^G}%E@QoSna8*2$3q5Ag!nT|TL(MCYL#AZ_BErvCIOfT{aYds9@f0p z275u9fOHWl77pj`iBlAnA##Lb?p3m9B`Z>%;i%%Mt-zS15SKVP^7=;3j1SJCuEEX2 z>;d4j>yYr_Gykx*T9lF3%hJ3SO87S=^ajg~FYKOvM+Hmdp8lt1j8Lhi9(6QAtvs6N zJ}m9K87fHeFsmuA_&v=2C{6F2irs|o{DSCJoXqyB!XGDL_8|(LsCmD%RN?lXtFo1y?t5*TM{MGlC_B8;N zvp${%5k-xi*D-VF7#fIlL~C$stUKy{osm4^{F$rS^T4^OeCFIDZe%ve#dgdW<`S{P z@?5zL6xvAS|DOn*od%78{(!2=WDp;JQI+YqROY7aadu#B+USw%8bCOoCKGk@{rctt zLbO$6xVjZSp#1ZGNB5aJSA|+_{}SvdVJ+9WDF7emZs&Ni)(lqJp1@Bd+7hY?yeZ50 z{@?xI9nBsupS=1Xl}>B$o#Tm=c)6|qYymi~pE~9CVAy}uEw(;egRqS2DM_x(o#}of zQTVAi5+Jo<#))THKX-~Xz-FY^G8$WSgsY&XZ|L23b4ntFSQ0vmneAsjaEN|5wvudn z!!rCKfp>;$i>jWOIa3uA!tPKaKHGR7>kqMz!{gZ_5uoc zKO{iXkK&8l5z3s?k~Ga8>;cQgfqB?C_w2i^$wYNr%}#Pqw2XxnbsCXY`Hb5^ z|1;k1XPaUU9Ti` zsPp88*PYzQLYaSB8?Vk;nv#zhdFwnH2vmQbXwfm|a4!1pcBX51`rGD2+mesbNceTt zuHG}oGh){h&f5~FB8$DzQ~eJbtolSOsqTAd?(hxRcOY>n_j5DFYn{~Cx3r+4s6*n3 zLt5`+*HzY;y+3@-**M2tUQ2fX-BT3e!qRqeg=jpYay_4Zk!*#&a%@mA9)=vpec%7@*IQNoWfu)5?fhxST zyE(a$tb$CEtL}XtFR9H;p@D$dT`J~I92|etl(<-*h&=9M?sQi?Ki{qaM@A1v?$$8-fs=fQ7bH}lr3ym7;=CHrOR^R{ zNcd7~)N-I!W!#~C%CP$A`LOmN-*sSoX5>}m1Sup{RngxH){6W3GAZ;oLH^j4p_Jwr znFgj;klGv|LH!R(#Gm`eKNFE!4?wHH+oj&RNw@4PJF5W1l^lchO88)zeeJx(-if$H z>>XXB<0BUAJ39J%z157j1ET6%TJ%VdgB*s=!VP&Mz2@zr{d$|(9st+2R*9gv$}-w& z-OM*}=!PzXR<<3ZSoSw>lBaXC{|tsnFs6Q5Rl?}G-UA3{Fs{P^@b5C$pa8UhUSgAR zcJnE;@@?kWewiDQ`M)~(Qu$I(S}3cz!H>xaED7p;%F_NXM@(38Go=%rz@8=#ANnA8t+Ry!M4Eiy*gib;mDAU%b#o9DBeKQRNBlF zFoG3knKYax!D_72J!**NnI7S=+sKP@o7j+XFHsyJ64u0Db&laS;(Z>lO&WqaGQz49 zTNNL2#kD3*Rv$%Rjvs9<+>H|s<>0?yejzN(=IhzIvC?tSMK<}8&}yQ#fB?N`;=_c` z4(%{@07+DDh7Yo(=;ODKzpg__z>3K-QLeqwx%-QRVKw;ywCoc788`FGx3h$=1?wH` zXpZ+h?@h}5MOS6dj}1xMKsPI%k~tn;KRqs8jJAka(72{3H82Xp1SC>6D_+G8`F|ZZ zgPt_4(^D8zC8FqSs)x8YRNN+)*xkRs_?-zgBwf(`WUL!7k4*ux@Oj@J@~&?fRBaN% zY1n*zjo=kr^LE`Av*H2+fH)iHSCMI1T$!BSy@Wy-*}3hf>Xe-(Or{0g?XleT8%m<` z^4);Uoxze=j=$XJAhoCUBQHo-{Fq*%_9^Z2FdE0eQrL5lt>4e_b+2yUn?t)Tf7JJk ztyVPyaw4ZWQE$C7yL-u_ph3tCI6+SH9|Dhb<=|$B#&WAR{Xy_G=OoGB}7?!Tf0lNhA>_pI^27USDD3^VBsi%ZiE{pgW z0v$?2A*UDK_u&eiWKoMkxu$>7Hn5b^D6SIv`sMwvt4*JnXrdxI*rx!+$!^f1xUs+K z-KI3HgmG(65aKOo;|p==ptBav^tPtIPt|r^k3L zDb2G&;6(B6PJBw`c^2)@WtC@uMf!=70YN^&)2J+(cCxvawv|LkzQo<@De)ibDBXqZ z>gLh2uacI0ZSTy??h|taxV|i_mU51>`%fUiZ5Y=aL<}z8atFK=#M0c^*suNm6aa`- zW3ms7Nil!D#I#@l@1q_NSc3&bQR%?3s%R1~I&ppMH^5DDIpziDF|e7q2%XNt?V}{5 zgDM7%tXPfeOEv0Snjo5-XwRBa*oZ&V$iOTA{OJOD1FkguNmQo}-rgsVRicMGO_8JF zDuOFNKn~h*HnvdgbHjMz$x_3h@7;mJ!b74Ko&|oL?v|DaML-K+;_6EsS=T0*$avxu zqfoxa2IXch5;EU#ses-67rm?;+L~|_spN4Ubwl_*6u%&N7Z4}@L8AU4W&G0*PSaz4 z7Yh_WLTB-S0}e^GnaNiY2ZO}EKS@)b8m_0^S&k#d)lBCk>hg7Tx1Ro5TPlS6{0?8z z!^n~dXnouy4yvBBn}@jAIaY{zG}8AD=$48Z*6npU;rnY+>?LL;AcH=x)0(htaD+}c zrnK~wMluS%^|7G^FS$&pOw%*~vbZ^v&-o!-PRvsQy69rUJ(5O8N+aY@?_;;H{CwUY zEAJ|9Z@(v~{kI$v9ZXAM@2 zw^pt+Y65#%%Ir6D2;5H{1KMtD=;RY114~yui8<|86Jn-KOglV=nVds;JN1~(ll8>a zN#qj2+}_Se&OI9bH86MEp9uFHGq&>BtWd(~RM39K#L0e@0tS5u^IeQ4s_2!G^KSkf zxXZ>_RS!@aIhGVWt^ca+;Bd()WWcZ|Fiz<3hDW|HWe~PuPVWuVGX-21#nU=%hZ6+yV=R04%)Y$8B`DVecqV=ZEvodr(^|)b<#<BTBZ#4af_1>yQGtXLQH4gmZc9ktlN+QOS$*h~kN84FP zL(#SZ9jj;Qio?MI+koB4!Nwu4wkNYI$BaGD+iQCNnE1Qf#S_H`j00(&%Y1j9;4GTvFiweq{tTaSt-4^=-~9rGcxo1e~sn!!HrjZwzgb-BBe zOI%wiax~Q4#?fR#WkX~`id9kFWofJX7ghZ~*zNF zIF{jb^@#bJM*BQsvAu60j}|nzK)p|!jSa3(Tmx5{CHh

    #){T#i@r&j(ni-e~Dm( z$TgYfEL4E3_=Fs5gD$xd^qqwz@nF`TaP4JH&o+8tZ5!^hw{=?>EMjVXGBTgB~ zI$^e-RQW0BuI;(DBhb_KZd1pgNpnH>VePS^`LkG=u;&nw(~?D-gq{flhml3fjy!_! z1O6yi`wTV_DSjnHE=*?707yZ>BV`{nXmXBDMfo4C*ji}yxuVG(C%&QAJqgyNq~vP$ zdhN8J-+l6FmVG@<3RlI~lk}>+6xxbv_yh4$#_I?Bj4pDpW~Qj6?F)Mh2Bv{Lnf=$W z8Z)bM;RgTP!F;|pC%Aj)B$bZ=75WVd=U+wXzJZW;hJ(4jCUYdkhM5ctWmVo|04VN^ zt~)shF*q4WvF1*Hu~6JGs9IaeT-A6h{+L7!3Kw_j3rcam@@0qb(`D3|WP_SD|=IUnK50LMH zrJrlBwn0!KvM+FYImAxocPnPhM0B#PaBtuyTQ_ViI(du7HkT^QU;XzG~Tf#!z! zxRqOT`4U=!$zm+mz)a(r--IzdZE;ccJb^lXXz;mNx8z&@4*#A;ESC!rF zlw0ABLDFXQk+!u-nRrn?@Pk*$TEOwHVv_y-R9;SP-cay0vkT_E3^HT|15Azs`)&HC zwP=z_e$yRzR&F(4fHd{x zF{AqAGqcf)5rY*i>NVw;|AvzyOO zTf!=82Mbq`%J_en%B3^Gmi3dsnR@yxd)bHi`K5q-*XNg(gVUt9m0c)VHFh~C*Mmbb zq?P!cwFrtOrR&%CJVbG%CPG}Dmt|F*)W$tf(hKvBI6nOPvc@7W%e@2Ti{|W+@Bh#A zc8{^Q<0}kAo9o(XLpBs7}S=E*-LHDgUS?)~@`(i_qtk>#aNNA70vm>%W z&HaKMxAhL`UREp||Ca>-&m_+R(yt9L@zA17gtFsT9Os*A5;BK|m|T42qvU18pcSJq z)C-Ru>>gK8(45dQ_DR{(iKJ^(?pRev;DoMqoOc z=CF^1(qV;5(csyp)u0uk<8bGEsV$3Ar$S2O*Q&kQsWl^8-zojw`*q)kYa3f@oAWMQ zIKFuX>Ye$)*P)w;G45SX>LF<&WF#=v{UU;gUmBdZ;@vh{Sui3<=kRxRy>^ZCta?}6 zx6SQrB7S|{rn7mM;^=&&%Hd{twxVG+T+YLKs22kW4rO<8+r*!c$#<}x=;$Us#{~8) z8&Mb@z);&-1CFyJ@m0ngCcn6*tv&QDsB% zFNGh0jJYrYP*1&lW}|@D_4p``#1S3P+j#@uBHl%sS$d50?`JWn$Vk8Wo_>JERx)~c z-Nm<`Put>{W)h(~eaz>{-l+87%D*RM?k73YJ1;$|c2Dm`kaN>{%gU> zrpr1@$;Lg;6v&u3cDWt?y3IZ!pnEcDc`L8h!+?h0Kinv2)x)&uZH|hP;*S??;LGVRrPWY^T0H_2?sd%@&U-sV(pEWOFT`o&|+-LCm&j@ z7=jcqVi{l&cjE)_r|?~(AO8Z3y}^M;otM+Q7505u#QJA6r`6Ck^zxMDyjh89m@NoC zx4uMummOtUMO$Pb88h+*Cwjme5K6l{7ve{B`jrm|p)i$U= zSX{z6=ZJZI71E@Nf3z<#&D)hWbGg;H3@{Rdk3xoZjBI8RhFbQB7XO#~Z)jrZxBJQ3 z3@=jP0f2y|*R7)oTJkqjO?b)XV|a#ThJuxTq%4vNZdr`aZ@M#W628A#GQSoQGch;b zbgb?_7oGG)7G{IUktYjcM3f}P2W{Bi{LjH0D_0G z2Xc~FJ^BMc=FDiH8B;UGc)mUlvj*7}1q$v~i*;GHv~`;I(>+Fsy)C1#dL*O9XPs>M z&q2sLov`bfL`?HkGF{*(!nu((wn8|XlH7Cdyn8g1{X_U_9LFD1+4_yI)x0eCA9%=L zg|j&iM1M=55#l-)G^{a8;fA_ktr8hPop9j@-ZUC}9ry)yZ%3kLfqa&+Ndf%`y~^@e zWz?xB>`rrR)WL)8s2*LM>|-&3(%!}2%u~(QvBEw*$(n60B6G<84O^wr)Y1cLx%z3q zJGj8WKM6}^>5!FDEIhR3g}MU#BKUk1^z@c?OyV{q%Kn?Ab$XC7DXhom2LGEru)w{{rYf2tUWu3cNdREUrAP+vomiuTM86BmxN8?{MeAKhHe7A6zi5%l9;LSX^CiG#2t-3Yr*k z_}vwJc}A*sw7GjCrr&fSq^}&5TOsT^bOZh4toD|O&b)uv92K`yb}xTn@nA5P`xHIk z?ZcohB|76%0ex${R7>WO!trjdw8+Ah9YVw`w&Hd#xHpP8A}zNC72V-mmjxZKtE@9n zaruM_2R8EWyf#EmZed?D0sK3XGEs*9RSv9%mOKZYy)1?>Dg-BKdsvocn!{m$hsYaLS~C zj~$*74iwuO++bZP>^^8hO3Cat#wyYtvAQLX!J$|C7j9IGtdU{MEq2?SIHh+j@1yzgD+L=S zv?Uy{L3O|bQ-JsU5JBS2&*Ncn@9I&bt9s<_XW8C#xAzje)=q)@ma;}1Q)|(q?5D=H z$_H;|Gv!z4ABdi9$9JxO*G`tGA?r+g8ER=prs%AzmJH$-&r2%Ld4Ezl4I(D#r(2jm z#0ev*e7)qWaZjP^I8e#xu?F{_-rs65E(GxNb>M2hM!D#qR`pJe3+vwE?ZI?Mv8@Y6 zl3)*;y>oP?#V^Q=Pz@yJAy@tMXVZRr{X<*VgV?%}AK!3^gIn|OCmAD@F&O`c>VXPi zN-VES2xP{iCqgmg7}$f2G5k3*v6nS4Eo;F@E2=%cYSAN!QIkeq|6RjrHzfM!ncu zX$hgxAfD8gP6tP^V`rFkd|m{@1~FRfC3&Fu#=vrgUetu=OpxIb!QU|e9XEsBDmRV! z(Qcq_A5UyUj{23M`eA@xmKJA*IL{Pl^kjpk-1 ztKd*4n!+YQQa$oj4W1ty$CoC!g3) zA+tivSw6mGJDpYLN;hJ+*tN1dt0%)3S&~?n6=_z4o_J5TNJ|;68Sgdkf7E<5 z4zYve)1JJ^vO~+Qch(H3@b39{y%j-8HF{0NfUjV+%C^frL&nBRmB;0untPy;3$`-A zBZvetL{sSyrGs(yj7WH(Z<@makqQeNbgYB0`|mm1|D8Jw;`ER)|{`KIjt*w#Xx zu!{}e%kVTs@3Z}#aXUGL{%y%cDKQnHv%QL$jGU1)t&YBg*+m2y9YTt>KsmEtK|;%6 z=})`|)w`Lm2?eiwri9(9f?PooQF8s9vO_@NW?)c5{tk4QCl{;>TiW6M0ctZskQ?97j`D!C7oe)*#9t58=`O)?}*kN@^MD zNhf)^pl?SPrwV*c6(yuP-I?G}kNY3=ELnj?i7}*8;N-n%)Ifn>=8iCnm{oyArcS9T zZ+?mR47;1xMwK^$A`9Xq6D9q5TXAfjooE-rb0q_j&Sl*3dJESK zX^C5ZKy14FD6!n>`1{7&R_#3ixu-C$lszz@TD69I5sh2|bY2Pz<-d^wV(rt+esghF zeNkbq;!_ER@XF~`Zes3o?P5dm{+g@!L+)W)iOU&Iom3M2j}u;n&n+Yz3i8BnOo@huZ zw5KE(CM8PxNMp}`b;KDGVb^?qR=-BYMwlO-@~D zQjxX52^I2fEvb^c%8gSzZhU|%i6tu0H+g5J?O?Vfih*qFNhB(Dl4p4DR}I=l`YU{s z(1F|wH%<+4(<0Ut*qaxVA}-#F2x2F2q6zF%w*XMi2U!;#J;{E2zK_Hdq{6qo`W0Uy#BeDc?`RuZksABhfC}N#%*E!8IKFK8gG(rKR-m(1TFVkGr0Zy0G+KX9?o? zi>KsXAY&H*R5WC(wJ7jipP&?2Y3fKxw)eHknUyZ>YOqM8GaKhfpz;f;SYryFU&o%|7}8{OkY99B5>}v;JFm~4)_1p3UsKAoBMb16DGE}GHi5zXr2Wzs7CK+_U-$!su1b6Q z9_da{UeM5V6zg@cH1!V7Ik1a2{}uv2<_3*7E{6npQ1sv?^CbrKb0z#j6+A7TL9QzN?=(m9QR?a2xeaIuAUJxL zx`?YoI{vms$m%YoW-#imx$0QZ_2$={px^EBc|ouLMB-Y^32&?7t%`57T`78sdpJ_@a$GY=kKc6zo(>@vzt*tM5u^+oP^4S3IJ9l= zNw%Ujv@`5O-Te27Z1AEJH*mel;r}H{Q+net=D3fLqVRAt{JR@Qz2KV3`Jf!ar*28; z9{bUlH)&ol`PWD;zokldI5I0jgwAbrn|IT3Iy&F(Auj|leO7@D8IC&qR-cLe1?L{3 zXRGUAPmAmwXmVI`oRoMd=QL{NM^w)=yKIG$7Rh176Ni{0dRhUOip)IOzsQ7?vac6& z`zA6ej4e&9V?j2Ee(pfitKD&YNwixejZnxx@J(J3C_bdIw19u+HKOo)Nlh6+^9yNz z)Pyi^^nwXsHF^vP!UZ?UzG?OSqb`w10<>#`Lv2`thUfgyxHQ-ntkQ@p6)OmkccJh_I5@*jzhVz2i<$~1) zuMqwG<<`qA??u#4MhU((%nDR6p|Yqqtev|$1vy9yL}#)WT~lMDRsSUao0xBIFb~yj zsF^9FYqle5-*T7V{^tlGi}sO8W>jrgH1sdD8ixI1=7s&h3uLXnB50A)Uh0o`RQtyE zjZH(BEZU(4x@duEK_{}8Ie)hvwC>66GwBcfT(Qy8|w%L-zHZ6h@azQ@u}_Kt~sP-?QT2DfC!pYwrqs*Ek`zKqxIjHWFk;8zYk;kc#}8~FPrBBTKk zas~q|C7|?o=W1hG7h#d~T-(5DY?rlf;5`l;CrJGp#E;v%&O7lf3Do<1&`pJ~Fujh$ z0Tgv^!2@}O;>Fx}?M>K;v7vfFPuYp}O{NkxaICuNIT~%x{IYBr&tJc8JxG^{vlHtb z`8C7tuNOzAf!(%$e`_}O0W`cgyl3wXeF_@roMW2yk!=Avou0-M*0i@fkwO6!zVfJ=p-A~QWGNqg42RNfeQ zBZ06n2L7Ngri~>#Ck5i<^oyp~(>S&B?i;&vO_KT{bo#;V+Z$!!^)P*!1})!wsZ9+O zJ=VTP%C^B#W&h|Ce5<8ynzQPOf~uRVxrtMDr%MLN+tv1YtlKlIG-8HLZYM!ZkoC?; z%I2t80>sn}FT==Ua!PuiVTq@bd||OUmGi?{?x6RZw=^*f7|xsz%2ZGjG=~o4%H8~p zDyY8d+)KE5r{Q?XFjY(szb@z`$F}o#(HAu0Q}aa*zD`pLffa;j5wyv0A#QUnE-&vI z_)dbPD?XTRfcuB|?y(t#DJee@-gZBfT}4TlA=ng^*U?bkrhS~|OmgI6*MQGq?n<(0 z)X2t*fp$GbE=WC(9)s@IUrMe*^=WDX@mi_-YSUiC__4R2-@N?;Z1{Qo z0}$3UzeM~?(q|@XpBiTN&@KrX0WdUa2b!ECrDgW{G?yMn^O&6`7b#7Djse~p&vA()JHDz`O(uYY2YD(Fy|&Z!c0v1#7v`fi%t z6t-a4O>8xaXh-!gfYGrbt#b@ZkhCDnw3RD|qba}Kw7bFYT0p~AqU+};(UF7`_L>SO zNvGwt3{xXfF9OMMpHeZ8Owu@_0SUJOfdmRMq_1ufH|AK61dydOd2Ln~di2X$%ILcC zFsx=WSx(3;p3{op7y-s|Kbjv8JP%`_OrVb)18Qh-VL>qeN-nrOzg5>KSUM8s?kC0U z(=mamF;e|?(H?r;jxqlp+Xb!rs(9p8HYxcLz08s9S!;<~=N+S;Gw;Bgp9wo+j%n`3 zp1o6{)MQ0v6maQmrJOa1@9H;OSu7nu8NSpU`!bqG_I z>yOVyN%GV_F=x3S9;E`NT9GGQB#m8@*-%AFaVy+my|Xv93NiII@S)3oO10^KT7xpR z)e0&l1PoV9x3yPEGH_C`{V|_zu@+KEEk4q@F@VNun=2NTc`_Wzif^x+%7IIX&WE(N zuAYq)@)t&$LYEHzw;M>7M#=)7Np;6fInBDH#Nr9T+(6o6@*e-$bii?vqxU}tS4S%Z zs@{CoL94X8AEM4`2K6729z}T`)~Z{-3g*t;OrR5UsNina$u6&!7FxH-iZZgS5Oc3d z>VGN&KNa6>WIJfrZqqFbU-^g`*7mc&fQffPePevYq#XWr%ZmPBB3+pIM5c(K3$(8q zXB67LulhU|7Tkhb7=1IKy#PfW=iU*>N!~QYj>-{2rV){Ny4B8$cqfeP&vZ9MaDeu5 zE0!HlsetPOS>gEf1)9acjxGLvaB>3kH>?wmj_GX{ratwM}?z3l;75R(=% z3)CAVsW6UWEp%yb8_gSvoz-zB<6KWn}9G4aoZY* z2UED#({RgNq%7rzat=84=zN*}6r4>mBD)%n#T zb6Zof($X(_Z_wu1aS!Jf6Lnej5v{kYJ2B`Q`T95r95Ci(;aI#qUgBd5fS5L+e2D1Y z)1xGx7&L7nRHKh|prjD-r4#e~>nssLWk5T$Qf*qm3A%{kyf=yA7W@RMH1a#fTpt$h zlLT4ay_}$n+)mLk#!aBM8mj4@ADkRTmqmxRS~=S*K;1`d2)>5V;5$3zzkr7w;3->cRpOO!iiI>=p{DvTT<|-V#4pO^FF&bz1L8E(n?;n=hI^ zBY!xxE=xJ#8>x`1L{cJ9uwJ<3W=d8y2GrHNC-=`YkKjN7uUmX~mC=!yU0RSm1Bc=E z0);W7%3U1>_We?AyY%i#8qx*xEv@S$@t5;&F6#kzRC@#;L3e6G^>3*A+@oh>%Iu>* zEocEZ|LveCPa#pL-f_a|*j5*`HV~g8*Ymg12zfoUZlXwx2u0Kjbls~`ValvQ)zLSKJPh?b;_+uNwahTb%asPL>$db?Sa*<c+Y6l@i&(rSHsSa0uZU!t zP?c`>({tTRoC>eeMi;Qe+R`y_ zT{E6o&bHjty!%aR)PUuhudp$~?(KGhp!Zid28bMSFK+>l8Cg$rS>pK!5Ha%1>_~1O zQ5uN7{k54N0Z!8KT9JIQ{}4bUn1Oln*kn*CDV3$$Z58uewZwAKNJlRDf5Rku>m$m6 zI#)b~dO9Ac=lfw)xX(=(r&+~Qe)`6{%06ij(>u7&>4}TRoIyb=i(He7&QZi0Z0xi{ z37B~pw$A-6^W+F2c}8qBPc*pCkS?SJ^ zl(ZE1xNlfHc-Dxp_>NOO*I+}&Jdv_&cAsps1XG0B ztMfuf@CJ7FXcI8d506IpDeFBWQmMllUz5u-J5B9L-I4%rb2%)v)&Vt6k(4}>*y=Uz zjr~-~Kr?*ZmSC;De|vQZB6DoAA$XX*l6x8VmVK};f=nlR_bXm*(A&0*$Hrdkv;mMb zrC2IHljDG7)L!IBUuQxP{CXl?P4uBLB?i4kho#S<`AS6 zufGEa*^UQ|B2-A>eK{!OzE>uHsWm0#vU$P0_QL(B4Z1to(tEbIz%&fmS?T??|1$%VXKjxh!JA$Mjq+e5u**`PfLFUXMbJk>b!sPC|N)WLXx26mFoWGIduqvSO6b~u<@ zi^)5U-WEB=2@Z}#D`S%?tLf^UP*y6xg^9RtL)sHMuWpE2W9NOy{u~p|zD{AzAhS~K zshkgWH}tg82kno?dcLX5cPK-Dr0x$_c|76~eMjWZgWc_e5Wq`NliztGHdej!{tLVI z^0PH34~d?CQD(Oy*IN_iK?ULh z)BR3OO$Bc7rHRK{85R3B$u*V}L_X^UZth!ZUItw^XSJ=Y?xrFwd-xD5y#K;}Dvnk6 zccpn+;g4~2qe*BOPkMszTyog1EskkhmKP=?BdIM8-1Z2NxR9+Cygx6QUws{ocV$6> zr9GF;^=X>gjLhh^yq3)^{AK&gj@pbn=#hbCh)yD(J^I$NTD#$@#Puu3E^fGE@0Q+F zxrwuko^OB6{@+SMiQ*c1Q-k$&{*IuLPLMo{DZ!{m3WS0BoV=^*r7^O-nmv|Z5V+k> zaui-P<%JRUQS67`#Ii({vQT#| zWMSC4*M8{%=9xg$#yb7%XTFo!~QTd5Q_;)mmd0VAl|L32Sg4{sn#0-BzFSZu+wRptw6=^wI1)OvGE6uSzUo zJPlu2+WsBP5siQ}zi=)q^F2X>i;t=W6Ko&yktM5`kr{-~#w6k+HH{GAuxAtDdupL< z78VA*L4g9$N$x3tx8RW(H)#W-T!;aUjD1o8IL)7`y**J|852f5Q1&A$MKKrJUzRH7jXy zqIOx$$G5_*`ebBXr$cp&yeNR_{pkPvtwZPRWr(j}%P1yH@D9|K372vl_t9}(K&{WAtxoWiIU%;sFOJLpOUYwc_gD-AWyWI($6EO}<0 zjDyYHAQoYA8hxp<|EpfI9EFsTWo>KX=vdnKFL0Jk(B`MoWG({X+1*(M}B$taq&G#3+E?N){voj zJ0ldnxD}QBuHk>)k}ZB?{#Tq@jK7rfrA>~3dVNAf*yVs$KII?bJNiM#msGeUIEv=C z@gf;IQUAYk#rYs;^fSTA!x|D{rN7P}G)B_Y)}_A@V)+6}bs6-5;G#c&j4lm$=2S^| zxjl(rTIreXB@Z>~Y2>0!=L4^%@O3F1vkN=|kAZTLd1Z#wyZaj0}H99KXibH6e{}N+|ZJQGoL}}&os?pTV&`@TzCszDgX6WT= z3yMsHNllT>JWR<(yv)v~GgJN`%rh9>5vJZQt7}PcPpBv@?Spik(V!M{zE!l)Lvxl# z_etT8Vruv_q|?~5QSPc6AL^F}?WySSPnxnO`i7ZVp1aQ_rLt9S-e^ugfoshy`Bv91 zdMOd_uI*CNI*B!X!eO1Goo|B*!Y9qm;^BslBO{A8tpIh>!L0sA+ESN6^t{cJk<8#6 zS1-JWDbCQ!IXdksdK({}P3uDJ?7w1gh}iRJ;YG))Z3axI)8HC)^Ovy(KERYNUh_rn zgT4gwr<2vNrtgWr&+1X>8&i1><>bsW{gm<9 z`Bpuc2fP@quFGwT>9-wZQh- z#ysvgLh?3@zF@|BE%_;L-|jCahj)cek4BfHO8L~~x{hCB4HcqEVq=sBgV(6cZ8Ll? zxw0H~;=|Z`a}om-`S_IB$u(yM*x7NiR33On6^;xej;= z=#AA`|CZ?NKOWC%`Q!;*I9*jAQndE7#}GYAk-xtUL~6`$f^5I0-sA{E;@O-3eLX(| z%=ZS_)0_e_7KXv7<6rch!v?Wbl3wU+Iqdqadr#My-pZerySi^Mq3D&U`rB#e_LU@Z zwwdwUdDJ3&zoyvKBS6r!O7i+1`D%>f>H|m@lEarWcDKMYx9!(gf9(BCf_Ucfc!}}+ zsP>QAXyk2Zi`6+lh`H(523KLb*L|P{~xEJRH8TuzvsDO%^8Y4~mS9B(Ey+DN!iTZ7Y z#r#LmA{&8PPc`$EVxiX6*0xs#?bi~winJXxe-CngDQ&n;J*HP}w$`c*F9d`S`DSES z#}eLiyv!6~^hMZyox&I7tKp4PsDP&}lHkwomk-rwzB!P#`Ss#r@b2-GUliVVPpkSV z*G*x!k1&a=|6Xmzc2);KbAFh_tQjCF!D@G3p06v8;tN?S|nMcNH z%tU(GPxj-%%oID+YK2d{%>AqC4&x1O!l1a_ZBs{p4KY~M>~E*>o_E(+W0fEA)!f>X zsX3eBeU;xG4njPWC9kf-xi(uV_RSp67iNAzO>4(&M?N`7z*?jj{<|DcYl~Lrs6tP7 z#kd{}-x-ZzbZV zp|Mnc!HedGyc)DPLR$al z9n)aUm1vHqAPr)TVYF=c-Gj1BH;5;Y+!X{AoY>8qPHe$d8VtSlvcO)eFHe9$h71x< zziTmTTXtIJD*88^IvhXu%O$5^>SZh+ZaTktzwmDFTZ5QnOgO{oF8J(3J4|y z^n>6~L$Ce@_qQ8aiZzu|Y^)VA%o6o*GivvRQMNY8ccicDx1DnpZnPCd6ythHF>RHU z7!{Tq=7BizXD?w5sdo7W#1knyl+KQn6F*xHV(9-H?E9y>4t2@jEn1p{`D!np%{$jJ zirUTqP7DWN2wET{r+}v1aB0O0rpN~yC&tQB_ofP}9N=QcSl6h+b+Xv6_{|LA0l0s& zVkRJv8L7N4qGnPQ_x!?3@uq(o!c#Y;XC(^V5!8FDt0wSnx&er^dv32p0nI#0!mavl zQ(j^w-FfJKi^wVggX#+o>#y<=cz@M!`+^&i1{0q6{1EA-6~QA?+x$!^&zl$)t)11> zeJ@uXwaA=~lKl8UfjBQ}7Z32hybg{@KD}7r`}>W{3ODpZ2roYoq}0*GKrN>bso2Nc z>)9JW8oM(_XJZ-O^Ig(lKafOL^SF=8ZgCVWzbEFcX2iU(7}D=ChGjR&eS0ynum5-5 z%7yo^TlltKIX%VHLuuwq@Tg5pWH*d0c+YdSTp~pB9K?#V5h5vO`2-j9TE+R(Q@eh8 z!S6~Yc&_yFT`rVN#kX9(%OV@Z$HZx(F<3=$Y~aO01)|tQs;cPG_{A!2(Rt>c18xh|Fd~@hiKLLNdHEOjUXyWj7BqT|d`6yawy`E^)D=8|V z1ua(13&pEk?uv!_Zbh9~T)bZISL#`aI1xRZqnSkv<4H%sHv|GTE~_fqn~mm}5{6sGs@A*EeTEjjl9(K*wI539e;x9l zvQ8`BRmX@W&qt5yo?kj~%SVW}*Ocv*N>M<*^W0B^*rm9sE~^ezjr;XJ9N#!9PM63R?Um!TYq`6#+55X; zgUL<71^$NCu#n=QOYZ~SM#9n09A91CRcbwNfKMg|fY!WBKeQc_HTA55xRqU8MLsne zL+-`qHs+y69Hr+rrvFY!tSs3<;~K zJk?ey}p$mNE76{4FF zSnBO`B;_Znm}AsB>)J+N-6q8hv}1qdx$6@BYR=1JK?b7rUJn;tTFLq!WJy|0R|{>wpmM!4$CJh3V|^3BD;6gVSJpU=#XggC zijq7jx$Hog0Y0&`Xq&6IO&L+ePBuvA?1;l4T&WpES`-PtVzWL2i|4Qmt(ewOP;5JqbFd*ux>}5le-c?}RHSE`v z6r}ndY#ZyC!Ufao$7@O-Yaa)870eIM&}`-SL_71OtV_xP{3U4s-bej1tkH=1+|P$b z#)+i@pNO*t^h3d0)Tc!21u(Ht$rzC}GW}$Lshr~_l`$GoHm~hjKcF7dN=Lg~_2=Ra zz0=&>;auk>HP$5ULaMTOCCGA;%~*~f$2^qZ$wK{=%8GfJdI_{?z; zJ3+}>Ug%S@jiwR67cjFkr0Te&eK!^>k9t(Np=IUXTe9FeNRgUV){gHTd^L`JYD!8y z5;_qh)_$+bU(0-`ASr1h)d7odO7)^0Z%7V~lf>owq$QhtNM=1EQAG}oGkpr$$gtvn z=fp3;R-@F@%yN;V3-`NT(1(=0bE%iPwO#aa1zG8A?=jT#;B-7rf1j(Mq^*Eq|1u=` zkgcmv6>6DD3pt7SY~CJ77)d9q((``FTw(e+=GSTX#EI~*!PBsEuG2334xsI=--uIzd9|2o5>Y^cOT5@NvAdBYZDEgC@i(0g072aJi90wV8Qx+?peKA$)6=E7 z8G`S|>KID9OT{eyF(KJGeEtIL)JNYHDrg2M*Kx+fqYA%=UGEuJ&O64u;d306qhc!h zUllx3U^uMAbNiJxX_Kp#CkW>0Q~KVS-$i0bi68VG@=|*Ghboi8l?c&o@QNLL+`}7aSMUb6uPmUFX zhzwIp7O&e^;+q?r7MvJ?H7&m@-%Up}%F~eeTaR)h%ZsWyOnnahX)l;fv7BUxE5L2` zg)bfgz#oALATYMle^0HXmkauSx$Ql427s-DqP2ORsD47?9|%*7&&%pk+O@9^%N*Vp zCwueOBIqi7cYPl9#X&aOGgYO}>P)mfzk zm5-k-gs~a5{G6dHk$$Js?NMe7T1wtOtLK^5^y4+=fV^?((sPp9fV5rC`{Ulg;sTHg zRz(fZ{iq2siFilftx21eMpN%aGca8q7$+2RtuiX2F+U^%Wpw98Jjw$MYm07AA+5CiAKQsL`U5hEB%-N5L zhHw%8?pJQfw>+e4Ko3v+5vCHMX}yndwCdwM>U0}Og*b~rsl?Nkx^X4+d=`MA{8hBj z=q4^ha)YI@oLIyobvsP{b;$41oq$SvCXm+-zNQo&d}T7=H?hsC0`uM!dOEof=GW(Y zi+;kqBY!HKn{_Dmp18C_-w0O!CD&wW@7<8p)dYO#(2lhXfDPYB+$s0&mNbhqXk6y( z3%p&JJ^hOOQlK+PzosZUo-Ol|+Uy(onErGw|Fp0fUkTD;?kjnCH9qESBXepY=S9yr1Uv38CA-0Blyr`LI&i0dx`AuF&Xx?df*Ro7S)s z_phTw<|z#3^-@UFC=`B(Ek9*VLxx-(wl0mFNpIP~!PQ)(2V zz`wJ4;z*;{jvvsqWIx$9?@u5SBEBB&$yJVm-}cP!B@-6!k2r5zn#^6!?%jfaVyfWo zD5lA`34(7-oFL7WBWGYI+r*Ok$q)6$Z;QSHvS6=S#I!E7fSYtug76`8ukz0zDG(zg z;plgN*W=0WI{Rd1*k_}yTLo-!9XPYa@q0Cw`ITHL=$42SbssHQLY^&FOIxR(1;>Hn>F9{W24R{ z>@@bq|J=aUAZ&?{w#@fe1Az{r`s&KQbdF@sfFrloc@G%Kk}peUO~u=Hr-YJy z0KpmR75j$vS|N{*X2V{Y_34z6GdTnVqzfp|pr8HoHI~D=rr*Xa0*V|UcY~z=cH=@b8duYgr)yV=g}P{ztI1FL+tViUp`adH?NZBKO(N zeB&&Z1HjCgCD;#^X&YZ#v=>`1sb6(=1fBVPJv>`1E#50JDhy@)I9vW}I$A<;pJp`U zml#lGDH~}H{XHja(@_(zc?al|w3$H6XcBZ)f&cj`P9+Y`gU9OsDe7_wd10)L>OWKu znKt~n$en39S$7h*eXrrLIPRf+g-?V;OnSZBEulW^xW+#NXvjB}bgUiePuv~tWjmN! z5BLEx2MHiMbgDQ~9PIS@R`Km0xetOIP6Z6KD!^i+ok_`q>d@WGTgE7M=FK;lVb6Mr zXH7T*c>d#c68+k73QgrW=baf{mq{0qn~!3nh!*^e!yGUKlR2ykZfc2gJSMUUzuT*P zW^Gbuf9spUorn2cRGmuWdiI7LWZv8zK+S$x{Qb@La`22z&YrCVYT4?WH%M0^Yl}~8 zVmo~qrS1Fhzg4O!gEm0if%$^hiz^b`lCURadrD)!);_3QXs!o(c%YRvy zYf))aPo=I0S)bb-Ew`WD9QsKd{Spe9G-ni7 zqI}!52_36>Z05BZ%Qf}mh5kRuB$If>C={g;M-rwr0bGf)U;V8^IXH=X@71Vn=aa;9 z2dv$Tmg^3+pLS@u`mds=q7k55x}D;HahaC*tKZ#iKwp<8eWZmRBQ4tU&xw6)O+^*` zrn)309wY7FzNet954r6El;*kz_2*8z6ScL?!1`EmKjnz#)*b%~XT*i)P%I`-V`zS! zJ2kdE-Kn9kqz)Z=c?OSc-eSHFj`-c49L>&!Q*rIs8002yyqeye@OokM zbP17(<#bBCAnxKd&ZE(m?E+$VQ0qnKhOM`#eSl4_njJX9GL~JTrjt)X*j^weTWcRV z&5p6&YOk;)Y3HdB+TF{{mcJ|&>{?a(Uj_LC44tgHZ(TPIetFZ!Qsw9F0>|+jJ#3pm zDES(eD-Y+KOD_9(cZKs0N}0(!fN~dEO~uq0?l;N}X*zoLR`?ocS=q6%=BwD|anqEZ zczQ_#LV6#kKS&h^>tFXT^WrGNweq7PQMtJB@v|t-Byr$>C^ z$lQuEAi+8)r$&+Bb3khny_XxcepKx+ec+mb#QR%kHl}LeT5o`zwaX%}aagEB&=>Vh zYT>?G9#dtjW&^e$_{L)Wy&0RaQYnWg`v!p8$k0eL^OOW}p^&6>R>#9;m?)^CQtyY% zZMnH5F!MU7Fgg1ZiT@f()q)W|Prd;*mOp9j=Xvp&25FO-_f`Ybk9|iP&GoPDUnzVb zRMC1U|J-f?_(>sIcxLxxLgiF4F=M=z*M>enk2izlRN{Jh(E~LNc=E3*=3KrPfid}u zb55eRTQ1~JoslU7I1?_#tcDrwiq}UgorNQO3ymJ7fO29;96o2w&CSgL!bL}u z`36UrR0vFBq`0EnKIzJ`KiCYQiY^b-{$5D@y!+1ZVAJ?~o(A(Y?zq_(%+$Rf&f9iG zKuW2s<}bfY{ACEwEL?s$s@SNOsAZPo;WavtFFiR9bq=cryLf>n4Ams*$H z_vdVkY8*)#(j0bjoB?MLW5zQ_m-B-~C8q5eD^$aM+XJ!b7ZwN5f$RK=f2>#Rnu;{T zt!ito1Ic)+xi!(JTmb{O6GOmUg18;6?|tM=FQuZFuxJ3%h;85YdhF*V5wnNZR`6zb zI(`N(cN8*^#=8l~jE5$~v0VLUV`I~<|EJ6%XoWfv54SicoRFcn=A6jUNa!>}1W!s0d48PS=>F2F#DQ7aTcM29lG>c-d1l9k=rSKJ!C zv+l_~4R$DO`;}{0!@8kEN|QONMC=mDX$@{4gdjVR9=lV5vi}_kaTJDP&}3y`o>K+I z`6_PT(N4l?1$X0W9nSTKq+<7hLlfeMyT(1`mak3WOMe(NmfA}NV%*Z-;ecHb0dSbo z6UJI0gsZ>MxoY;${%uoB=|{!o%TaF52$|6|2dpBJFe_X&Uk1QIqn@}5L>2wZzXDZG z+QqI-=7fOTjmFf zBbp!0^l)}-Jspy$Sggo1ZWx6@b!;Pdy3!yL$WE z3BdkBp5bnbZNt`Z>JwPPr&s*eR6$HhITYDp$d%9rk3KFdqPh)eUCLfy!DjnMyM7%l z`1>!**N@oc?vdMAMRfA!N^Kue88mTvCk{uw8h;9!oao?hNE0EbF06meyeU7YE27Qr z$M-T37#G!vqed!+!4=91whv()kvu6lih4PWS zY|M4rUA~K;=9?+retn;4BwjDiDz6x~PX?VC^ly?bOhUs-Ibj(+T~ ztR;p~wj+f&$Vw-B$|&E$yw)bP0{`VCD4SCh$bQn;$Ot1E;$;9Uc)g-1`0oAcx#Vrv z3Q5`cZ&f;aqL6VuYC`A>Tr7ivZN5YT*?0mq7R(r76N2#`WiRpWVkSikzuhxGAR9!f z9VDReyz+q*TJE~{q~Qlg?9LMKJ7tN-0`p7fj3ktI5aD97&=nEq%|7I=!!Kq&|W zY$O80yZ4GT3YquD1`JueGCa|Rqm|1oZse2aT(R>HLL*<0Um zNkYeMbSWS1pqHt7`hp(y(SMD-q@(zfFYl#@J49k|(yc@oaG{$an>-eX1V-a_d^E|< zgCwF>s#HvWR5~mO(1R$DwvEWX&pRaOCADAU;3AMqq8`B$nTls;rHl(G$2UJ9{#kb# zT;lM9tu`6G`PJW*pL>PawE#WdZmgr_N*vY7mnWVd)IMV?DE+y9A_b3$d)U7b8jvm)`tNb3`7b8DND++P7p#bY1q`Q8)ISTTnlzVDhQV^zWRiF;i5w96p(|`>?^Zcyj1Qs=z-|3_N zIi&e${5=SdPwLr))y`hR_Gnb?dlW~&HjSMJxxMkgC?)SG9s5srQo<#d5jnkAuPm3Y z{+#dSgn=I?S!A6{I-tqq)V@&-vASr1K@8r#@g4d1wOTy!@52IC*;{&Rig*7HmTLBV zQ_=PmL~A~NuQwDdR@W+-2|FK1Ba9^5HSL>LB1UWzqup>dQlC_-g1sJ!$)PS?cYhb@ zo48J7!=EW(3Eo6`jHV06t4eKKI2OBux|%$IKyC2;sLFjexiEy;qfEZ{!jEtcJ0^k+ zC?2ZjKa!hjL|j|OZa$kwLkS{X;z*wt!B_`9d9*SZTqpi5bDA%%{quu^XQ7MlNW$+# zB9sj|AXqz-7YCi&Isi*4nW>dhs&1@N2le5%&FBW2o~lj9hmrHfN&dZO%++X)u{RPa z7FSyR_b-C`3dfWlra@Pyb%sU1;OPm}0`e}mlKn1H{YYvq(ABS_35LOh2=uM-H@Nut9Evf(bN;2S*P$!-yv=J1jJ03@E4u5f*P<&U-&Ue(un?>NW zQ0u%D$Z)vHnC~c@#_kglH~978toE0a$6s_3I_?RJ!chC~sv~U$_C&b^_i} zl{FVQ>zd61gZ4D8$yqHoe*9!`=3yt|jY%A%%+IHNi;W)Bh8}i$0XsjVDkWH`Y1c{1 z$!P27g`?6;VwuiWpeliV>z^qhRUB&Xf+TUil0VB(Wuc^?--emKQgET+e|}8ZgfkLP z-?~^F0pZK5sQOMECmG}VH#N0OE$BH7_CDa_{Uqs4>q|VE^;8kQ{WQ6FL@Y*T-Yx^Y z}>tVDy$;r8O6ZTk=bE%f?j%slqsm_+0?ysOj0Vz!YRf@EgyB>+Js*5L$@S zB*32OTJ|jgU>?hp@0NWL2f_^2yu{a5nLJn;Vkzo119 literal 0 HcmV?d00001 From 5b930b875a152c800cbf8a9677049acf9e374a62 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:46:46 -0600 Subject: [PATCH 542/590] Add files via upload --- .../cpt_images/stages_training_initial-1.png | Bin 0 -> 96734 bytes .../strategy3_cpt_stage2tostage3_decay-1.png | Bin 0 -> 104582 bytes ...rategy3_cpt_stage3tostage4ifprevdecay-1.png | Bin 0 -> 122381 bytes .../strategy_cpt_stage1tostage2-1.png | Bin 0 -> 107135 bytes .../strategy_cpt_stage2tostage3-1.png | Bin 0 -> 117945 bytes .../strategy_cpt_stage3tostage4-1.png | Bin 0 -> 125954 bytes 6 files changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/cpt_images/stages_training_initial-1.png create mode 100644 ALCF/notes/assets/cpt_images/strategy3_cpt_stage2tostage3_decay-1.png create mode 100644 ALCF/notes/assets/cpt_images/strategy3_cpt_stage3tostage4ifprevdecay-1.png create mode 100644 ALCF/notes/assets/cpt_images/strategy_cpt_stage1tostage2-1.png create mode 100644 ALCF/notes/assets/cpt_images/strategy_cpt_stage2tostage3-1.png create mode 100644 ALCF/notes/assets/cpt_images/strategy_cpt_stage3tostage4-1.png diff --git a/ALCF/notes/assets/cpt_images/stages_training_initial-1.png b/ALCF/notes/assets/cpt_images/stages_training_initial-1.png new file mode 100644 index 0000000000000000000000000000000000000000..b83de36cdd7c3a99dba8518dd5785f31a4e0bb4f GIT binary patch literal 96734 zcmeEuc|6qX`*({-Dq0SeG+HR3sm7#ilS;_Wh$vEwjF4nsrd?9W5*f?LZd6m*vdm1O zB*qexEHgsJK7(P#?9YeJ=~UnIeSXis&-0u=yb_d-UiLW_SF#bLYy- z%lF4L5W5-8gG}Y9x{*tJXP7e+6BBqmJ~lQMhr?kon5kKI@95;f#I%I}8x<9mqeqXj z=Q!%>>fzzx&pOiH^_739|F}N9#>2zoYei3Ych8|ihuYJd^b?!Vb%U2LU!Iudei>!$ z-MjZ|U!S~&DFgyxzb`a5H@C)odRtr9+|uld{Sy9~TKMew$G%D0=-?;vm%%BHYIOa{ zlP8PFV<{;qMMXvFc{tF`_x5Mqa}S~KKe#u?7ciz6Sy@?X#~ie++=lyxipvR(&vhI+}_rG1)H8Gju9a(dU{=o6e82Qef zJ2&IfD|((885xa*z+PG9Dn-^!KfT@7)|T9OyUgjC^OJgRPh;c#D(*mc!0m*Q>!vIo zhxMtt=Viyo!8a2ljK$uicTx%p+LL*6%uTNM`UZz4ZkeyMiKy_%nqIU>agp&ceR$B~ zaY1z8W?OWgRHVY_qc3kRoml%Ci#Yw6Tbz6NXtQxrUX-bgq?+{89`egXS{(gsB2OCFl?2>x z`pjPKC2&ab`tya;JGzm3U&(Dq0x3@I2HWQ3_Z71_FvesrxiFhL@6l_JhE*4lJ z`NasjN=#E(Ee37_NmwT~0h+zTRx*b-0b_vV#%SYx6blK37B8>uase z!ZnScQfqrgl+X4xoWK?k9wP6G733NLQtGsWFe2~6+o}m@*#jAVPG$ntw`+#C=|fLk zg+qK+8fM6?OMpW%_O985-T-`rpa6(ko`-ZeI1qO$<0jmuca`;biQCV;ibG<-fGG5V z_dZ~4#v(WJT>9^9mP$;rSJcNApp=ZD2G@}ZiwY1I?nO~{%fB1j0Vc-~FOrV*- z@qG*iw#x(rRgwa1x#{NimgsWe_$>bta_{OtH=OEQ|9^6NmwNoqxoFncR*PR@k=uz2 zsp8TvR{VI>M=Y)}(m0vZ4Z$LuFfx>a8K9KM{ z`}}>_}l$Cnl7E^LqSPMos+5#=FQPN5qa8K$nS;SMC4|UlMM!BmcPML*(uE zXM@pCl#S-Sz>X!rk8PIx#}f-n;3TgvlU%d>5Au2$IOO4BC{)SvhfdrW^tg>yyaw3$ z!{y&s|NV*cM_z2WNl;!cyW^N59By;il5}_pxjo_V4|QFcn70wR@EV7=eP8|eC+J{^%Sxi&wyYjM<{tN;FlvRJhKu;=x& z`u_^!$-aMuu>|n^f9lS^!uVGh-#ya5KI2~p`foJ(H?;nXGv>31e?iyZ81?^iVVon_ zGH);@B9}3c@Zh+1kbp0vNM4CQu+@I$kzO4!skPf= zc4Ji(^AU>s>aZM&9j0onHwZr)W=X;khJ$&#zn&->LbYq`?qE=N%g)8bL^Sek!E^Z? z)}d;-&(IdIOPct!6^jjlMA>wynnka~Iv~9$S-SQbkngBmLyE8UD+C^jb4ScZmpQ$j zhz+w1DW22Eax_A}WyKvAj{_yh&3%NO>l_c-(zk6<|`h5cK62OOQR?3sKNlP>hymZs%P~(h|kM z%>X;Fchw7>aF8jGwUAdEi2?aXt{8TsQDXrzmy=76p>k=JJ`pSBKf$JrsNCUrKB`kn zPZAb3bMnm}2~*EB_@@(am)rQ?3PD5;t*%V(BZ~XX(pp3Szgp(Wc^=VLl5kAEh1ei_ z6)785KCm3f*gVcIBvT1LCsXIlZ6Q*tV6l>W_>Ys%lZj(%ipR$`yuy_5e& zmnF>>5k0ots(6@}2I4H^mhF@Z{=kVUEe<8~KY(-W;UnHJUuU4D$|bYV%Qnyt0vVOd zrnA9Zy~n>Lc8_hyGrQR_Ls0p+9VTmFu~ytVKk_}0uQZo5*7*`f)z9kIq)*uq?y&Ku zO!b+rNvC|oOs-Sl4diXoTs{JPoXOizfEdOvNMHFE>2u(uWw_g2yXeN>c*l1S&cv)Q z$Q9|A_4ah@O-kmj-;e4K##|vxpYJq*O-e^hX`l>p(GDjt=t7$29@kmMaVReXU6W@7 zJ=*qChZOOqhKtK#pP1q(8>A9d20WyLQ~M+;hU zJ9M<*SWdeKO-@M3&$YMKoj+!2)8(=^TFTbGy?%4~&AZ4I%jZuy4;c{pt=fku7%1CA z=9@E>3Oz~%IwvicB8Y0zd& z_2}n=xtsn$AsWXg%X~Bx!X|RiIdZN;F3nq}`Zu^`w%(o&x;7;+h5C)%1JNrOtEHyF z{wYJ@lXsB;i_v2ez)A7U=7pNH<~N}tB3TEMc*Aup$(WD=S|A!sKtwKelYc8vrd!<2 zt61$0PWR~Ds6%d_niwza%&o(x0tC`yNkE`$|3RP>VPeF(3S#)o$%F5#EXen67@_nt z_&W?ceGk7&j(ue|F}182TA!(P3Cn@?JQPl~N?83`9+t1PYt7fCJkt|(0Kk5 z{ba9*X>#C$;}+!hu&G&1_Bqj!K)>$Z=!Q)9yu-UOhg48jd}kcmIn0{yZc|~0uu1+> zX>2;&#x#E$;DC$(=H)+kdlDL-*Y|2L*-Oa!ECHT5xI{OeaClUnUEQ0>d-ourbMdgj z`+|)ppF2S*B+yirC^UzEm%md~p>s%yJpRs|0NPGoDbQT-zsG;^zeH^E4vQjBQ^e|F zKBG_yz4;XBKy;)`8!Q9ROYn5ZwB(kGmIAI~c0?v@^0Q$?yng`iS%N7I+xb}^Wo_mK z3v_IuR1#KL1Yl_r6W>1JcWW8~BH_#S>T#aoFOp86k`I1JJoPsj)}F>9N})O@ z_CjPMnEu5@2tbS%2w?;1VT188dLAN7$A=f!w6*u_dIiUOyan%fHp@i|Y|Ow<@!sz? z6x)whtp!de?YrV0{aJ|nlCYdAy9#s7naC%GF>YASnT!N4C(#EG)QRt$TgYKb37_dU zo%!bX5j<@ad8<3|j0~t&(M;L+ zowN@}1?TlXuT;4ZVA|?9!RF@0yY^{{?gEEfJ?&m89P=vfm`D-&j5(~9Zd99>gTsQ0 zH8}d4i**|N7;2Z>d2-T61E$qfm^>J7(QUyEooKwg71BEQICK2bDbwVk^v;4fz1>72 zGY{0u9WBJ^aU*#a5q?J-BX*_P6PRXv9gBHID0``67WxmNtaom1GCmKz>UqQvyYQaH zo`f8hVa%@v7L&4lTgyv2Te z2hv-oS!pFQ|5=Ag&_(ZgPWdwsCRYDLnBXfCWe0aU3+70Hz5F?{4^^t6+^q_S2z4x{ z#o#z6XCJ!Wz7f*11jT>vwSujES*wk2j$>mv?+g{n7$;ONus*@o50Q6}*K1<|ASAwa z-ZNit`UOH9vL$`1?12{Tg|t;S6}=e3CR?2l6lO9Nmt7fz_!9mA?{@C@nRp0kIOyyP zjuPRRS4_Oi-GDV$=BsQ3OqqQ?sh7^>uTfMzq_+&sH)dARBtHKRL+ z^Ye4B0&b_a2a9+$sd16UqOw~Zn=!=scprdGPJ=UzY@03r>IDEeGvS85uyRhpXpbo@DZ~S$}uEe^$XX<_#74o1y;t z#QO#LnaTVlzwb%FT`}^mc}3o!W&vsc%_;r%#Dc0r^?x?k1+=cR_RlKO3muZ>KfH|? zP;UFHjx-NfI>dg4D+?ho*U0kzya&7gc3R`8*>AbP{1Zy#Fs$L&bBhMg{`GYKr5M^^ zJbtpYEG7wfW3<=FBszOpAB)IKdO%Xu+9t@LCQsY zbshtfqwlp;v4*y9zr%<5qdv{uw*&3ON1H>HQh*wc{x9XLpTUvCx5h>iyqz=u4hBSn_X6{q;?=fzI~PC5FYPC3atZ^G4Y$xiLdnaMjg zUwi()mw*2>M?=N6r~nS}mCK#c#9ZLYW0PIJ!gS})OwB%>7Wog=vLirR!0NnsfXweN z9IDv`q}w;T%m6U3ci%j|fvo>s^RFF4E2RLjT(mqY7qt9um9S4&{-@=)WuuMblwlEo z%5|D+2oC>`>np7V4YyW|wdLhNFO_7hT^+G&9>yI6ELG}v1yy>xHZ9k4&8`FZ{-E%M zMu4)LJhyp$h@4M0(P=+r%&E1bJX~dWJpAptB_u7>!~CQ5+( zY`{J}m+-5I6ry(a3I$52PLBqJ%E z|Dxkujfs{mGuXY5Au`*(TRoQm(vq<1MRV0QlE;_2(=pq~D{a0>-kMwfbn{JXzokBpHdnrbId-w;i`SDbkRG3KjWs_%%-%dE?gt!IG|U)C$>x zve(=`<~yb}X}B{9NWFUdeZx6r{eqHS6^{3Wr?%U&Y;@k$K9}#5%z{JB`GFnwv$4sv z<7~EOM2JX$A&`{zne6GE<0YZkA~t(`nb4eN1N4L3u6_9>eNTol_xPSz=LlZg7)~^E zd~X9-YUiv^SO-|0;O}!7h0o3eYVBvR^?UY6>PD8M8K105Jg=1WHN!5ZGV&$*mWa6*gzm_f$fZGs)89}HzKYs za29$${9JM%x$#qdptI51i6?UrNrwrHo|Xaxho*@mPPxVGd}T+RvEh+D;(S)v2}j<& zoNQt>ybASdXn18Y(Vwi83JRzyEd_Po$oKlg9lU|8lhVp301X>6TYwn;=9hP@MD95_ z6Lp|KLFH-#A9-f#r!*RT#` zSCsl{J9*~NK<_#>EwJ~#(+C~m*zm_$SqnGphB$MKU4*@k4Z+4folV)r4Wmk3vT*83~X{#PSWcmP&q;v+azsrS5~_!d8RcxdJq0U z;;Km#_J^dc#P#Rx1T`Sq%WRaVYiw(DkD_-ZNDxiEbIw#Ge2`x}^L{sKm$=p7k;pT$ z*h6YBgF|!y;8MD9JoK%95qiM!g!OOU(_QomXjM=dMB5u64P=%>E4bxpi zZI5dsZx6|ao!B_te9f99t$n4V5Yg!|P;xR9M1GpM)1+{oSU7xl(8D|VCmt*+*p5_d zJYXU;J(gF906L)V2G`5Grn(@sLkNnmCqje*wn!H4n*lEpG>tozN({$8kQn2xAwP{) zYj}1CSaFfZd`6ZF{VVW4EVncBi2#>FvN6xmy4GeGQ$%->pmIQ_EiWd|OF{aB#0e-l z-7?&s>sEKZL~sZB`F7$myQ6?#zEnG{@#JT`TH8}dTj%aeXR$F6;2mit+TN+OhN7#i)O(D`r7f5%Rm1^93^IT^6KCEKs#i71U>NSy`xHHfO}mlr zsT-aYjby!p3|Yh>Z%bYs97#F*OGk>CMZa~x{ra!|41$%pOr!@bnYxv)*u^C_^tuWcT96Imy1RWsn14}B+bV}A-s zhx>he_!P(U0lev|$+@lqbZ1WJ3d}`D*@zSHAB$AwQW2x!V!H!?mxJb>ydD+j#H&2-l`RtTppyV$y8*Hb1 zOZ+5EQPw1uLY{f)`YqhK0_5Kk%(51IrBY0wg6{5Itlqbh zNF;aX9=FJ~1}hvR?B2oB_TRbNaru3}5wv-cMu!D}idjJVxaIfsahKTGs=VtHsh@Pa zpP-ZV2=bJk2H47S*V4C@;fE^}D@V(UhUD~{tx4PJ?m6|-0Y?N&RSSZVPiUnV^6T2f ztvGkN>N4bULw70$pvWNKqb}&DsC(_0Z%IbGqYKXn)5JvXaos)Y-0#hx0k3}LxH-s| zbx#fQw(;bnQ(2RN?KbL+aAKyMmeZNGn7q}QxwEsU)v%rGq@@2lMP`+cAdt21Dr;Ml zj(~HL38RBfHqpA8@4nbEGC0vyS6IbXY^;qsINBsjojH|RNI;D$Vx6prgHO_|KJJPy;ed0mRkby6OD@W3W)1WX_>VZ%^7kf0@;z zLC&R_?%}B&@;1`9)bY8QAdNeBWi`9b89@yP4uBW{@(KT5Cer$WyZxS{ldHprU5afX z9Ttm6ji8XE@k@G`gC|xN=^ol{L6R%lgm--iZv!id<#3jKxk{x=uxaUlJAJ$aV1A{3 zhB@b`wr%eB7+gqcnQa?J6GD)2Qz18|hmD?ZVX5C$iy!CHr}Pav8^TU-KE?ADZe*i# zY`JA_@-+W(D?)0Q&;)x*-a+>CDj+?0(>fm~Nys1DWCGi)5ePU&P!X*RQy1IMqpqtF zpU$>Z4yNj27?UIU2q`tj$z)L5^PZxPwe9z-r}-O%RO)I>+W4+5@l;Dgb#be@YKb$- z@+}fd&=?$`k7w$pH3UCH{@SWS1nq(FRm@hhDQY^A{mH_Mq?%^Oj6%H{n@o+s*FSCLDdx{jm0lk9oI4805oE1%hM`tfw}x*Y}akI3g_9;y4@To#Lbu*l1s zdIY=C;r^gUQ_9Z~wA_xgqx&q~>s2^!vS4y<6XW0rex_!weq5*|`aFJ+HD0HAIn)c> z?1o-oYX*k+0-1zTb;hlp;BK6%b?fsIJHg>l8C!7h9l%14zCl+e0bV;XWj=U;75}mj zpjWXx!Do8UQ#wV{y0spejXw2;6EV^`E`f`&8l5Y4Z=ov#EOUI@7!KSHh; zxzV_4QYa6@PqmGq94`&If`p|H1MprheS*9Jj66u->i8#&rjz5 z(_%lRi-#FIs++y$bDm|1dB(rzz?tDrW>1n2L%;WknxEyGk#-vcEY1SsE+_GOZkPW* zvKoIsv5+1*ulr#nKvCdV0_XpIis%1+VYr-vVmcI4@_{;Ga~|cQ70Fv!=iGb+;le@& z&At4i`}QSJc-P}w-nf|Al!qh;2QALLib5XP*K1$5Vcyy3s96+!_!;`y+meXkbDvB= z$cJX_07;wVG#Y0HFN)&4*LtrAn84QsCamDzU#PP-7pdo)8HL0eLO*xt4y^_VBu}xX z%)QeKLd2Ud40j?QcZXQ!~q4TzV`%Ggt?mu2tVMJk>%fp$mbc3_A^YR^O-CIad0miMfHqaP-(15wpKU}YA5 z=)W{xSH~d}4Iqh%bj_oR@cUhoM!?ZM`2%3GvVV5Yd-Kue2biScd0m6AJLNQR&)t8ZhY)ME7W zyXR*>{wqOBgOJ-b?i5P)8bAjUFA+`WLEW2ajeS3Zy7?BBUy18~9XQas^B0T%voJ=7 zv7DMb41vpRW@IZ!Z62&U8$6fQdM59gr{`@ahfbCeCHj#?j zN5r$I3Sl?h1YD=)Zp<=)r)Z{`tAeFYdy6>2P?uOXOsEmE-&l`c^2Mpam%;68!=#Cj z$8z4KPA$c5T=WE}+5G(%e%YZBHkdc2ch~l7noQo~YH72ZQ-y+e#ty_Uw z)KJ)g`j}_~gTQ{d3$bASsU`fW!lQZ5@)2+DAdz}c0CidzNBXr8b1xHJrgY^NCmjD( zuH#8`2&-efeg{V<^eeu5RKLm4OR3Y?)O$I4Z!wM4TZz z(dNc!LaN5BbXA~8SeGMG^*PTr;gn1i2&thI+$2tG_BK*nBR`k6{O>Bw&vl&I<=Et$ zc(!2b@W^#u^HB0d?nLLi17jBW(LV~#1f;8N?0D|2J^h==?|I3jq$Bwksvz>NT?I+c~ z+VQY+d)kvscHKqik4IFPTfZ?-`^Fi`FqU~P4%LLR2U>}ahn+AMI8sBQ{$)D^TN{i^ z;)hij!$BL!D*=w2wDOxs(?+elYdi`bR{U=PCxGf`%Kg`jrs#& zy*?oCE$<(3530|m;fDDe$t#Zm1755CyTZsBCs`3f zkS%is(t0KZhyZKS#Ye&*OpMfh)BQ?(f7|epQ$@gD!=c*Ltq9a2^txP9~>y=Lb2W6`F;?I5;Gi03nN~ zf@Sx8P=G@$0Y`D>zv#wsE7CW2j03HVM@&rE?~Kev`(ihOiU$IlE@HZOX{)q$Pnjno zttz?jB4%~M#OY!|#jq3N#ZiT;=2(@me1v}HT)a_?X@;;bW}UrCVd~m>uweDQo9o5` z&Qvglwsm*D)pkIL{QMAvR*+lCjgQNq_> z9!3z03t9^iVI_(HmDj)bNin`R*U=~!Z3VOvloU;VR1-9^$YHH%N_s-)y1HGWId|e! zSQ1E^xf2di1P;LN??3m8alW@S3#73?5kjpmh>>F{r?8v?`=KE(A0J5bG58@w>0-f9 zz%KNJ>#36@pfsmZp;Tbq$~w`--rvTr5HQcOElF=A$?Z4we+>Ek(m-7Y=-ozGZ*$}1 z*oC*r$Wm#kkEK@)?dVfJEnBq}3^Uy2jKQXm-aNk-o*89!m}JBiy&M)Kobd#`SQ=Xp zd>t4Y{fEN;J_1tNgBBpf##guj6-Fo|EesdV+^zCQ zcsj!?dE;PcQhQQ9H4xk^JyCB>diz{<=4e1M7bjqSyLZy6A{r^a83=UC{tk2tiTj`J z2wKI+ileo?V->GoH>*a+Z@g8w?2iJ3ayr{YlRbUuu_anTZvDiEf+@GiOn3P{puN*4 z6c)bE{%NAdSp(?OU4T-H{}&H;f!rAId>nohhYntntX`UmEDfw!FHl=X(%5hOO3Ie% zvokY;WLznM5Zo#eW zeDD|+DN(eHkhji@CXLAD_(#Bc?U;?v`e8g%rf>q{Ecdy~){=BpjNIPy_f)|zim|~7 zTJ)?JOa(vRn@$*2V7v=rlT4vUv7As*_EYP*BDeL9eGpvO)Oj!8z=HhVH^Otii!-kQ zbKuo&Jn+4X^9@I|4D_<7%R#uvlxIL|tQup)K^~}$B`xG0tnc|dKy7s=I}n0I9d3$SiuUUB;hfVqgLe69;Jxo6;?CFU8S9L4iHR z;d;(a=wwa9)OkLe_&LAQp*{6zS%}viYgm!-o z->UO<={pOQ*Ul8S=ycc6tGwd@`G|d~aq$F)Ib`vu4IQ}ZU~hgd;wNnSi^-yq%EoBI z29w-F7Z?^1#OULAP?8Xcrw>a}R%={&{$in7^hWd+tDXmgY zwf4)%DwSxRIu;Le*Z{OXJ;?fJG`bvB&po% zgC5s^ndJ`CT$soH>1KfiRr-H^5c|z(ph7B~?!t%SbjL#LU-zkj!`sxSx@MVw1eS!# zB!N%Qg}+lh?CWS5&22;GxzAuSU9&?D9m<}D(z~vF~UR+d{;Cc=hvF2UUz1%kBiIJlSO+P zOekN7&7UXtZegTJVDJjKvQNA-H#4*w#Pe8^mWhH8Sp_;%Sp&8dLB9{IsrWL@*ATl#+p+*T7brIq^)lf z@^F4F9ciy_wugUJ(}Akk=dG-uMy`qr8XAZh6Ex~3JwUE#sD(oWd|~rC&`t5Lz&z51 z@5c0a0!T$6q(bS4au%y3v{(DS@76xH-ubFvJ5AAJdha9FGF)Y+f}ybAnL(ZEBzbnw z1=EV`rpjCKTumoD?~Q;#EEDk@7*vs-vq=?DQugz#+{BU~B?67E6p~*3yJ+c9@ayKR zA>5hD06fp;5w#U);I8lB=MkJj&NPF>!_PEpR!aC(kahgX8E4ZDd_#jY4g1q#oWrKZ zAT&z_Q%f_ew8uzg3~?^g)$j5R8F4sdn-zZr683ox597a_vVB!uuyt>D0nHh&ND)Mj z9Y=4@jnsrdu2hC*JZ1a6%Tl=U z0c3&G4GX%vU(RPH5O|FdvsUT5P?F?HjrRz~rtN(O%y)9e(3CW>!4k8kY2kPZrh)cN zy!AG4af$3acLhChw$#Q9HP@6D=3I9VrJSgAO$`pw?VvyGUBQo*J_gCT2upWOrjUup zV|Mr;B(n~v4;|S6>*Ym+kidGCv&PWy+9Zz`jz&Xqy{z0#tbuX-1E-h}vRb%s3m8xh zT_?Ndqv?3hTl?N(hl*Ag1EJ$G!2d~%p7sKxQwPi8Hm?qq>lc@2qR;(l6UpXK>9o?* ze1oX0vBRSVw0=gG(|Ths-ys#FK|pjxUjaa?!iTz$xZf9mDl%F8c5*f`m2sZ{}E=Uf8QuSx`l zT}=_byWc|Dr)2*J>Pw%jd*rd*!1f|NUC6lf$AdZg(C}A>Ug{qMsMWXMk-{6(Jalu_ z{WCH3p|Rx~KoyjeqyTI%2~X3%+$3N(x$tTHy%(1~eg|Rm_utqXAn7uNoik^x9tyLl zsz^^qhJe+M6?@hv%fRmGXgSq*umyGa(D$Lrz`$2;rSGjX7L1A)2MnOmtI<{@iz2h3 z;HR^*8~yS(gx?DMVrR2=r+e@%q*iB`ZB6oN1L(Ug-*HzFfLCxr}o!OX3OPhmtXRibm0s82{t4NO*={q^^DK) zdXG>Udi@aPV3(vTJ5GSaaowB(3|n~DsD2W)a$AB)UfJxjN=*GpTxAmH4=2*<*UhUn zM`xRcd~Oen;%Mh*JU?{!YTB2gI+n7(+~R+mvjZJ}`x52-u!G}P5kFQrb=u#MS!qQH zo(mW5e6x;SvsCr^VC12`lqfiJXjkh!uNyo+v%`j27Ne>_)}!<1>l`ARuJv)8a?z)l z4aXw6HJ{VUbA~9Dxc3ETX#uVJ#~r#JjrlmEv_(%Gm(OOt7LJrOBqwt{7eDF&dK;-1 zAzK@%U3gbwuSYI=h>})m9Ff>m$q7m4C`ERYo2MVSfo!FU;APr<=Ld0B z8{lok1{JRe9Te+RX%lZqJU9*F5mWTqU~1|N`{l{`Qpg8)n>b+z>iMI)(!D8fLQ%r4 zjSuUAtg6Rq;kuZ;Z3T$wmcmBLt}A1s)&8Qp#Dj94j_FBSL~Zv`Mr|GCphU%-Vq58h zcQ+w2<%X5FUJT?6g7BZyDsZ`W^_P6J(oI;PJyQGdR7(AGo%LIco;%ORh3^yL zCBk7Q>zO5KTD!(OJCOg9Cu9Z8RYfSsW%2AV&0n~N0(I#floNozx?JBjBU$J!GqLO8 z4xyj3&H6D3y_fP>l^Im1kQsN8D)M8N;ZmI4*qhW>_S*%HZ8^0rhVQVVtd|EB9zAW7 zQ-2qE_rRYAp-twS>@k(58^{-!+&997HglpwO&7ZOk*pBhndI0(pI|Dnc;N92a0#Gv zO#wm#Z>zNLeqE6?j%mIf*hwiLfE)>&o>LRYf1RCZ0@F#69gyvfTf$ilV4^FIJDAiQ z!BL8@X$qtrr1#A5ob@>GYATD+sAHm!2cK09C-J@}g?4yzuX%_Tmp!3~*5_dI@O!rV z*SRN;bPI13v)3Exuo%D?iMrGdYt1|JMV)}~wnZrg9@DOGF}Tt`aw05MzLvRe%XPy` zO(>O}i!;N;xY1w&d(3s6S?@-;4RhU}2A_tc)!ozfFXv9;>_P$*?abKCrErK8d3xe= zB($`mCkg*rF$QSwV*9;$t(09I~+hVW&;4q5O;a=E>x$*ua z6~7G_y89D&bQk1m!{3jci>Tyv&OTS3F@lC~+u$_(=vlaTR%2&{oRveaT}h#^BB51u z0=!GmqDiiz+k9j-XxlQALGF91PsIgIJ&6#xN>H8g8D+Vel(F?q-220=M*zAk2#*gE zoDT8~IQzQWlib^1ydTxUHn1Ky%q$LU=?Jb=3lQ@Mm27z8nB>`M3y0Luh36xWp||Yr zS0<~i=VeTk~;w=X0w&v&P~x!<=)nE-Q=)5bh>YYV^WQCFn@n+}eN zcWFIh#9MoN`B-A$?N4%;jL+HuYPioahu&0+j46`!v>nzYsfl3iH+MWkqVDUv@|-$m zBzb!d*qD`#^q5HiH0}QRzJs^W7G6X!jb9I-P>P;`p0{5&$!Za*4R!!oO+QY0#53 zq(x1}JxfD!@isS*2;9onlVQC?rPZi`p?;nMj%jjIW;A^oL?^wN*)i3gWkK)NN_KR7 zp3hUO9%mxQ@`GOd^Tf_6G|NQaR4Uh7cX1{mw}Gw+r8yhxWmYeT3-Vol`MSo zE{d69*4)ktYqLOW(gkT_pJR4^+z6l-a?~8r!ZrYEI0Pb?f;;C5kP>$}FLkX-J95x= zsMe-9-?(8t>rxoavz0WIC0+D##uVBehmO$%)sAA)1@9XC5AIiSG6Xh0SP`1cheqrn z^iEZTvlCMvqqa4;%&Mx@2due2eOh$4zGwUt)%D<&@x7x2Q-q;WqE_i~)W)zC%;kK;VsvqKfD z!j%OIsiU`-HKd#^-n}v(acl|sV$IidgC%}($Q9uM=hC<3jh}$FxjW&WQ&ogR_f_p3 zLWVJ8$)y~10;ljcoTOKPC>8bl-`CzOn}D%@@$4?AO<|~FLn1pLVHGY* zO+zB;UzqOOViMMSzj&rX&4VlvJbW_u=oiXZ6|xj{ar*^+`m(qCuZV>V@EcT*LEkL} zE^-4}Hf^p0mov@-?sRLj-X%V_eC>zhOLviH65E2?f^S%ZUEXRpP-d_*;febnWTLxP z(-SxP92POWpJ>(OgkqcHv_Ap7pB?B*&-Drp+~(x4Y}b@|purPGbl?^pJ~eFwFBqMB zG!-u0AHs9XQDTKsM$fQwBRiXN%XA6aLxQ;J6rg`sWz(GZ-_y1~6465@01rylj7GQ` zJ6c zGkXc@n`w*tPZiK+3x(VDUI$VYP7Hc?KF4y<{rNVn{&}V2RA^K_Y$Hox!dZ%r zdBn9BUc2#-!kq$*`n!Wsgk?bGw_?6G_U~yhpu;gNV78w;j->4i1s`ZVIXOnJW33t1 zo2d*cyTX(fTlF#EI^VPE1z#8}dR|t~3R`XHH?Bza2=VxYu`byRn~RPba<~ z!?e};T3o4uIM$~D!(YkB+bcC9+yIplR5)Jvl2cS%8-X*K_A@LGl;i`96w4%!@OOXNDXeY*&1g+Ud@Bo-|G*XC<&A%mw zNE#?d7dnBqs;M6kJzQm0P|KsphtW&*YV>eRePeT!0hcw!h8b^C$}^-k8RPxby~y@)UDR#Z%7wQj=T_e44vuv^Kht~qgJy@^{pkUep_S8zMdDrxgl|g2$YR@WaKP4vA!=hsi{bLarmK*wl74A+ngr7Taqg zaKQNMX+>eI%&f|Qdj{c}za9>dsC6253_wbEcD~&enB|mUO>)v=@7$?paQyMav*1~$ zi-w*0Y6kR9Ps!FUEk=mH?gUal!Zmg;>(pg;>`3^2y^q4@&L(;YqVJiGqk%U@g7Sc| z)W-3)&+$mEwA`Mv?e#l?+jIjvu6N7qlEuGlSq+Ts9scDqL?AI!j!JqO9mw1saC+%# z21otdl=Ac^0=k(We@0u9JI^8)i81?YG|8lykkY%_U7astWJcAkCQ^{xXPnkjx}FLB zgECWU)tT8RPl!31g2-&*ng>0cA!Au8U zW`egR^%BV8l03ZV@h;h}`70rRX#_l*YeV{$tf{&)1w#qtal6^o3)HiEIUOU4j^?MH z4X4qabjL5Yn~8>lbaoqWNm$|gTg?nt2d$7xY!(Rt*)Cy0JIx9ryX z2dO)?4HBwX`-xRZsR}ORs8!=heA>}KB3?;PUc?qnP)iz#&Ag$mvF3L6D3mxALUuk? zkE#fq;q>h9J6N57Qao{qIvSyU z>qIn`?>O<`;_=s%S?|Q&Dr5kq;<`1y!UvnIAlDu-wzmKgq6^B7Ud3z{S!yLauda{T z+kL><92(z_k*eQ***B|FchaJx2$$=}Z)(z68D78oFzVb}FM#ugfA0YLo{4|_=jLX$Pyq!@f}NIB=iUqA?~H22w`p^-7=Kj zg|;R&NDCb3uWcSvWBqP7j|Ogjol-o3*PVIK=&L;9J%1OE_rfI#Kiv983#h`ZJEo*$ z>@CTtBFr!o&g^7U!z;QI(he}sJ=N=g#VeG5Vj^S4hC?x>xm-dl?^Z#as^Yt)`x`_A#&zt92RbMHLHo{EIs zmHcB#nixQOjPq;r_@_&&Y@Fci8igB|y5CgB+`)yeoUnP_QKn3c)zQ+$JGPpM?yv(z z?f9eX_HDYc3vqQB&^$(74PbWQisnDBX?+uxjb8eO@uF%m)e0DE6!)cg+N~-aA%1wE zKZBYwYDO4VBw3SceXN?F%Z<`|d_A5nD|3-C2vjd#3fvX)i<6x1C^Z#ahrU@ky!mD{ zv}pS~i6BpZuyM3!hFw|X10^r+()e>xtS21tuc9a^~p4GPZ<7mS{ytij3@GX6##(Ek!DgvJ6=(B2-Eg$?{zH7|itP`+J_h zo`1dEbI!TWwY;zO^pqcJsF1CRDo^lEv46nZ6Wo=IRSo-J$BP}%NwrP5^gLIy*E73A z=DLMa_O!=V)zb0~e5BzPOfPZ$Xz6=NoPCoo6R2-8t(O@CGBI0QuG=J_);?M@)^TY~ zM^J}DWzmyQrX$sfi?smdYYaX-=?XhEsU>YWQ-e+g(wg7Hl6zn3xDV8*uz%)E(oogL9-hKGb!Nk%2 zLAL_(78U;27Vp~pjA!(azW$~8c8@kPjM#jl?tiTLDrF$yfRN73eB>ARtBC!ZD<~xb zW}t|Hm&rxA|KS{(6djpn|G&=)o`6~r6fV9(z$}$zZIlv9M5C|JBo~vy`yU|r$cX4L zKkfh2PBov7haIM*yJ+)x%SS2C{K#{2h+ShWC4AueiGdI63A?t_SjuDx(y_TO59ep| zSRyyqu%9`4osJ-*!CpP{ZN4(#`0DQ@gv?&F>Og^1^XDu^P$N@bH9R4ueMJd4yeDmv|4-!UFa(PR*joy@EO&!e>}q6 zKK0+N-L`HO`FI>+tN@m@%@RYDXFf!pIDAD(WMw*bZ@*i?lOvO>L^Yvf zu72B&c*F>gG$1$+S$-j7^W2E2=x=gf*n8WGAQ_SG-@d+15{bvzPfc<+w1^*Ig5hvp zlqxL+lAnYuCvLXu$_3<^izAO=r~v$5SFlao+jA}_*l-!el!3Pjq>5kP3_?@hiKCI1 zFu0Io=SDx0n*YYd{icrR9m~Q*YjW=S;~z}bRj2Kn*)A%Huj;5~!P(ETO2=ecpCtQ$ z3=084PrB)vOTXOWzRgc$dShp*9oplMb?y3mf^C&uaBbICoivF8yL67h0oicYz68a} z76q}Phk@0lU*=CKp~eFr?_rDyacM_pcx#F-gshSU-?Vu4r<@$Oj{adhJi<5q;mV`- zeUE>zzTQ%kt6x+&zD+#NkrCRK(8RawE=@_D}f`M(}?fD~68H{Hw-XOn$zeRlZI?GXjbT9?Gt2rT1uUUi9q}|lhbxjmLL+@ba_cUx% ztrCjo3?C$uwGJ8(n-v*2=o-ia_i-7* zA`3K4MN4qo`c8FLM&W#~nL zJAeMCGSMutbH6*t4dzM~N1+@;%2lWdSu2^$L{Ab%+99@(+adPEmD&6b+Q+CDgl;in zltQ<#BRC0p4S{B!`oeF)unChg@CvU|^AR$2h5S>frJrIMN<3U<&L-w(3W$q|-C1eN zI)V;Tcqa}6Az(WD!mr^nEIuH4*afkHX9p<5Qx=MvqeGZt_!%#K-vo~~z{Z$@uY4nL zLp>inyMT}m3|JNP86*qC05wc~tp>a=u{jsg%nSE{yt_cD3;SzpHyWpjTOg2{zI;MBV;7kU?BM|!fF32<)f2j3P*17rkM{9&e^igrSPp+}3 zdoM^HOwpF!1@=IW|MA1o9NEr1%unn9yxsRC1w;EHnfGpD)&U0uT{R?5oC$OrF78U} zZ%K6g?#>|amgA@y=Tr7e-;Mg33iME{e2%`_*?f8dj9OaEhi9df`kHEj)Ff&!r`a0* zY%IBS-A5V)a0v!6RBpfvFFNAH_qC1D>(L%3hB9sH!tY}nW_4N~S9370T*Wq|110aj_;?}@uUNG| zs3bkaaScIc|AbL~pRgfuQ_GF{J?|ChGD21mWORB?WRL6GT<|dLg{K!nXAWkUhM+y1 zdaGyr0vNT$gN~t0GbUH>JInZLL(rW{jS%>qT|Zj#dl`Vfrz5I9`JowKae};YIql?k z$XJ`rqoXOOPt=G?y82Tx3X}VKLMonhO;8*(-OIFm9e7II8uWGfM}!S^&>n4b`}@b? z*R@ffhQ@yw2<#)#=@T9I6nq@@ua@zrVMr*~li`d1HZrv)JeNeD0TfUx+zv1BGdEh@BiF2jV8-N&gz<$Z{NR@k@@gNf7F_kTBmI`mw`OqTcud|%YxA3J57is z++jVXR*_!Sgh6%A%PTCm+1fIPF+b4@;huYP8_T^h+f(Jcz=orDSJ-|?3aKv`D9jr? zc(=VfykHJ3(GmW7uA;KI+pFtB>ref0)_sio-^sqITx-H)XJ*e9z##X7(9!SIHSy(q zar?y93)zv%jG^@GS_~Njy@^4}w{?#btFRuK?k(SgxR0W|`>M;%v6kcI0(1WqJmdG7 zOWF!qymec&nV%@ha?ClV#Ot<1?z?_tO4bUMq&tFZJJ<4B-sN4=TX_b7367)p=9SXo zY~<(vtj(3no|c~DZD5#qA$56!=LVc2sk=L(nxQiHWWLQ~7j}k}hql3&=|AZL4{tQB zz*x0?xqD3B;eB4@$KViUImL&2H#yB!1}PsqWX~Q>b~S+K-6-N-lQqb+wejQb7?EFFn9ZzC*s=lo^7-&!`oKdN z-w`42#CJpKE8hN51?Oy?*XQK=e2%>n=ktfJku6}C$X>I6ZOl)Uk7}S$yN;J^gRwm< z`%|#U`ym)j{Ty~UiQguEUT-`68b&%T{cQs5^R0rjZ64vq3?B@j5$VatKel$O4+Wi= znl``b3UA9CE4+z$cCEMf^cVi;5t|EER<34LRHMV_p`tIDq1t29ygu(G_uHaM*mfSc zp3Myjy)RjDueHg+y1+foxQrlPHQ0<0&SY(C62`SQ@Jx=|&lN$Kb)g&BqxJINp3B*_ z--q3YV1C2lg!fU@6+Qr6$X93&@lK7L5@~#3wFc@ZOZvHuIC+vbv=R7tSXWvqr>F5u0AT@@WM@8!^-)=KYU4?iQ#G5v&sZ3D{ zw`m`GBdL`qed~7p3xtQ8lWP3GZou1D52j&LFJZ6(5MD4T!zja4k*&xu1_U#svye-Q zbyE#I(=*~_p|hT8x9}P;RJaT-_IsvAIpp->--OZ{b3|G9?jU`U+6Z-#%8UD0-uX=m z7sCXb>$#?@|r$>W)*y=b$R zX$NF_U_UHF2+CUgI}*Llhq}pjjNY`mwM|J+`G8np!VQFBoZ-dDfUhSXa5iS zz`I*y=}ExE=w>}^9>;B4zRb<+!&g3a=S=d4Y_${o*?m|LVlvRL;)h@Z6?m0uVG;OJ zFDnG|D`^cbV+vtNGHkLy+T{Tq2Gs`m9j&!^-Jlp|9)2g9m&E8jvgNop!Do|;l640R zR@UQ$UtGF4BrmBgjhb8P|l#0+d9UW zvPJ&7xQtirhZra~IEeNTH6h;dmrR~qq655qM0I}#hQ!SgO6_zhodUod4>x5E_Rsbv zao4bKK^$<$s}Ofu!ol`Kpw(nLq6T-8*n+S|#Qr-&`N=Xf7Xd&3YI=Bq=r zSUfP@-UX%$@i?~UXr&U$1!R9b;C}pAoRtY*NqOs#yzYnIouj)f&58Iyjclw)R^^9n zW_=VLv5kZlJtnkx8{z&?t8oi+TZsr$sF+8uRSGQsr;Rm&MZpi*)`~pfgoO1LE`)2o zfYhmkdVQAxmjOM7_MokRAh--0GqQREMfKI-o89)-hn>5*-#Dqru4c-Wf;zj^7WJ9m z59iY{utP2eD^NoUZl$_!kP5Q?8PP$-@~f{&&P#YudLTHxVpagFQn z8N$0r>w_$eh@Rpm#JWl&1<7EceH4L|OwQVzyTq%2k2B~& zVlLQA%HvYOpuH;b_MnAi|70pDEu9wf8cQw*d5wg}z5{m!-=LtMim?@o$tX9#bsUlz zIHtZfIDL41jW4+~gOu}+EM5Zx%^Fhzb6^0W`y6#7xee`+H9;@^vrL*! zYvZbgTO%6{idP3x0(2Duy220V)SX^EopY&$bDPS|wyHZGc2aN85v4P8W zJwo_>{#$W>ADb9)J>?LP!(3+ym5fN#iMi)(cSt}>t}_k}@{8omIfc&2Z9|V60F5oF zPeLZ~(v{cRYM>%(Ge3l4B$@pPCZ|B=L=^I% zZC^vPL4lH9!3{DQJe%xPA+7w72}emxuS-B>%Wh2 zWRG6Q4;sf(m;9h>0H2L3kP{v1pJbNZpaIy^0IHcdw$;It_RQ zvGZiRwB}Z1ZJfn(4Lb%fH$e+CxI+KQo16b$5|F^ zCL~R%?rAltwfH{R20KJ6;s=K_Ngwl*Pn=E09wsZ6DjNN~XkYh}wEJ5y4|J)2HaLqO zPbDpjh6FYn!OGMrJ&Oa*FIr8-`O6oCB>sW(w?dLntzxBrzvn8(n750jgcd3--%XAg zDpv3BV=d{A5KqCm!i9KkTH9GVTBRDMvJOwMu?1MQCZYi92W3 zpk|WF8B(RiZh|>e8<+X%aQxueY^-QJ^+zCdG~``Xj)|$atreSkc2Ys9Aj%NsJC?a zI~u4prw0+=tzCy%12O#&%S6tmkCe7sUupb=sIcA7Y_#SUnDUK zN?iapUkk5FJoI(7J^OiY+YQV*B#;TJN z?|shx+;mk`L>5Cm>OGSHZAo1M)gv;287}GOW)FGg7Gn58O^$O;? zl0SaXJsbPkKUeF8t#X|ocqIP;lNm>A2^xf5yTKR1XrWrtrcNk#{qa|j2o{p2=r1Zf zspFOA=UBBiK|J_ra4SsO4VpG2MxXxU^ZEX3p*I?+um}GVr;MXKw;u@n`s0%2-KuXu z+?w3@YP%EM1YDVSi$X0M?TME{Y`2YII>|8{Sh^UDt7|=e(4d<2hE2$N3Lr*ywTSFD zh5S%1n>}aRB8$P@$D))H;#AwurHg)9}}AVR}^1O0^i- zOIh0CT4YK-xB`|0A2w?-+&8XnnN^i{xp5i#OK#gzsPb8J|4vdx4TGn93RVEVc(@_S z-?sdlfIDMM#%jv($x3M1a6Ped6znTpfwF|*9!N8ur3)FG;RcmjAOQgn?lQv$sG@}7 zmC__X9~n}AxCy$*pxfv{;Uh!h)1F^eFLT;-t#>oVe5Y1Ri>F}T!STx~BH&%`$c59F z-y2&S5J%z~8`(c=qNKc2@zu?f%%fkj%(X=zWQc%xnz&^6K;3C;xKT4ixbI*_>mANm zcMSar+939U!tk^ZIjV{)$&!uKO2){h1esc{ZeXzCixlHT4O#H6U!oRbj&KOYp6imy zc|YHIgIC;50c+%N>!NSDi~Y6cmQSBA)NQIvp9ZW(l1NApYP*BkTmR|MBhvgwCdQb- zjZ{oV@j`UG?40B~O+CI|6nsN}6>~-DIDW8K^EL)(kt-;fyd`=iwUJ%c&&BXBhd7e$ zMe+7U6`4*hsC4YG@=0t?{p|en!}$jtC~QU(tA3WGi66Ym=lUQOOZ<_TTSB^x$>1)) zoWAgUX(y? z&9u~kFrLQ2aZ8ceIKb)@Ai99n7vsj^qUleAB`3}lZ6-O#OnT4)iOFTUnl3k zFvd(Raw(M`mhp&^T^WJ4%2_j%>H0r`=a-Lw5PFdnxQ^bB?S~-*mjo~8jX-3D5YpKE z1EY!JuOs+8b3}Dqw9QrUuNGR+i~}3YcxIgvo*% zH+ZoK^uSI;4=i{&&wr3ZI2;g)dgyWfS>CbAoK6_e*x>OX#sn6s%_KSV-V0)`VYe4; z-LY5;u#F0KE?wzZ@qO%lY~Nv&N-}bnVa+pbZCptlHSCN3O%~NfRYf0?R&v^bnBXtp zds>L(ylL`wPIe;NL&k)-8OWfqNCr685_U7{nR>r-?oN3nzMdOK$&e5B^FB7R&!_r) zTap=Lj4%c8k%Kmi93ys=M-}274icn|IzVrY&XxzczJ|o`sWW9>u#)-WO0XZ%Dv%{~0hBc&`1h-&V7 z9WOB&DAsF-E?o$=acriD1JvU#MX$hS{V1t#`c2>7adGcHv?v|?|D+x1n2f^bqU~ne z>Hfi|Ozibz44xFOgE*+4;55_cwzMJzv0$OY|8RngdZg!v28O&x`x5!wiVIg(k`0Q5 zAyM7^Zs4zC_56d5R9bFP_{?vX@y$0@7Gy5pDzpr(N#8gBYF3R+OlJY_UpeCYLB?D5 zvl#nUmv9I>g){McHbI)+9Kr?!?oX2(Y3T8`Td-^@HoPyf@DAw~M1Hq0SybOjtfBi~ zJP4+ltV?cT`073#IyBBa3eiAq7-q5{>NLt^?1`75PC+tuhdxlk5{zV#g7h*O$VQbC z?kL(emI;V>{lxh+$i^=01`;rdN6N#C&%NFeY_oO)#jH?=qVTTAd^wVA*hrhU5V%&s z;DSG0@R7^Ck}R0p8P>6A@5tWmVxTLywn&p<&@FoQZI+_lKq)zsQ z(N|XZ4SrI~zsyMG-7M(=_hCN`)aza+9k*FCBgM z`9Vvk)vtU&*Q{K)PihYl*z!;yI6@4h?FgMJZ($Ps(QJ78{rmqFd6fioNtO{fF2vyb zRwQk$Vc$Wyg1?F6&-ShKfx_cwM=DDmy|h&(72K;7g1bXvl@cp}Pi|89gzQQ#>evSw za-mYm@7}i45NWCyy9e!YmkX3R$3LuRSA#nt1hbBrv@U4Gh=Z$y z37iHWF-P$))2HtI>S@9COj}c1Ah7Xw#FD@r>*@w!gwDAY)!sjQUqaAQR7K_z6G4D< zBh%06VhiRho$vlUGOhlPpjP~mVb3-vXjN809ZsG`dnoqGKSVvh4H}b+LIi4`H{+4} z@GpCwF-Wi@M|!ys#I(G0&vRu>voi(jLgmM@V@pl8EXw^R{=U4+XMzGr z4qVUV@e2ZGSQJH@)A|NlrP1|c``IsjDa)8NF)<*%)Ngzryc-g6X=bR>K+cO?!9qtV zAJW9OICi*-q(60!oj4K=AoA9);+!lvlhj=Q$j37)wlT(RqsZ7>8U_2okO+GCGwhJi zvE!?wrRe~;F2&e#+~7V5kiR{P37=)d-o1CePZgzl6?0kXCw|cP9)>S-wV&=XI=YHUCx9Bu47 z3+FFAK|02D`rAo5EeUvX6bTzqM>aDaEt~|Q_0mZOdvkV@j&*Onye3#k^B+zqt9q)R z++KXf8+{ixoE+TD>m6a3_3fw7Mit~V2tp~0K!Y8U&I3h}@a35Ums}IbfFV)XQLpO} z>`H1o>6rODeySCy^J6+*PotzmB}E>bgN)}@%q1nDE)r;tzw`<~u?I!+X~R6qedh`{ zWMaE~Q(tIOPC~-lTSSblk;$9~13QrGG5DFI3rUpNYZV(fDX*6+2%FXEt z+UPV+k`a9mlOcGoCV^$B4!sy~5u}PuqosI4h*KxkMYDV}YF=&_VtJDI$k-T3xs8@6 zym3Ta%rgC^UGNLOsX*5SM_dfVC-9y$sNH6orA)Ez5b6FRUz1y;(z(yzyoJF`J>7UGaa^OUx6lEHw+QAqK~#BN$ix$m(R1d|voBw|@1fTgCpFZSg<|+KJNCw+Jv3e) zG&3MF*y{BVcI%=V9L~J6Sv`GbBZ#*6QcQZ8mU~IZA|seAWH`Tg=k;1^sG>w`q|=!2 z(hzl?FSOwP!aRB(Ti3mqr6p%YDRE~4cSNnNUrJTm)kH^t1Pf#Zi&s}JoF~7P83a(5 zTH7t!_-!Dy=Rcv)MkjG<$lbZ-sQ$t9cJhBw*>2WHVcGW;RLLp=w*u*gvB#VK9TRh! zyYcoyoDQA>K$;6tjt^X2)-$lwu!~TGk0sD{SqtUeZ(St<*A89+xmkM_inB};jYK%barnnBS)QQ1lIV?e zyLC`4Kln{9*-t?LY1O|%8}Krwdt*^yF5LBf3gzN4F|SL{9Nfl2rAx@teSA>dl>H{6%yy|M`KT9m+}gCahSu%j06-e<2pIg zTC#09y;ijC1KQ)a`|SxiKF|2n7=2}6*Nf6cVK_;M7k4;ulX%W`LO=#lGJjbLBd~eB zhf#G`F~LeDka3-GFm`$${d~I_cHO^qj4rB_Bh9+H^w$L1;|{t=rmdwk1B3j^XvXS3 zNEuRyNUO`>NcXW$YkUb%&?8yzJKctMA#lx`lMboDiC8Hoq6xbxKS7c~x1oN8tY16P z<7x-nEid2(^X1i0s?10pos5|%c{a3^AK_s&@9gw(?uk{qEe_Me)>d7YD8H_0L`14u zV%J=`Y7?@F!g6FSxFiloVlnG|%}?)135K}C3U~W029Dq}7qq{>_85?f6}7A`NduJi zj2^)ct|RQC@HJ~+!{+R7tkWwuh(%GnY?=}n^PAscn&X`?%sLhlLlEqW1?^D*Wt@H2 z{-gM((Bt{{u_8b$hy)Fq>Ak~(+(f2oCFeP_`fV_w(;6qcq&6!*B$cR;Zn(I*b^!!<>Y3x#9S7^>+JIlLr))Yk~ZyM(sz{Xp;6Cw zy#3ZJdna{0lqn3dv6J_AA{>C6Q}mwV2!_hNr@+|uQir$}(J;pj=)F^ShvYut)7C7U zy&W#|jH)rDc7;gYoOc<5h0yz#9Iylq+zJj=xmzGZYczjAX1*75ffvr9>b zIcl2eE-l0=7BUJ;VikqA_cI&)_1NKJ)AEO~V8fW#E`G(WcegGtNsccqxhW^f8Hh_s z&iW0U<*8UZ!FwO*(J*Uh4J15V#aOgK94YP{x#nT{fPMt?Y5Iz)bei+kgXY`cvO!FT z=bs`Bi>1{!xhsEOZOv&xwI0l)ZG5j#n+qJm@;s9SkyhfOUZ5@pK}-GcM@jWVUE?Mm0zE=WM02VTWHh~@xmpFP z$Y`?AOzh?apDF2Sp^Cyxk$A#=tWH$ov!Cc#%;=G#3e+od8Y%<^K9OrAK2g;e81(=h zcbE{{YWDd=avanJ4oQ*v{4EE8w8kEGoUV*9rxzK7?AKAsC^>c%q9u0CvV(y|#OxH_ zp8cC}v+u!I>MHq1D2^trUH=H)SZ8q=P8Uu*=L-O$VG=-g@_BWQ+QLw&=`xEjh{ ztE4AV5?yc*pA~px^bq^fkeD-6csr+UZr?sOu_lW3q^@hiyTZ%kdVQ5|b@g>Vr06EX z)&?O2;O4x3{Ka2iEQj~Oi46^8`j(7+qLLsNM1UgoZHXU0;|7oBsG_d0{GGFa_u;sa z7)|JU(uyG1QS&>aE|pn%&`H~=?Vjdh7l)cY@$7qvRyW%{jyE88XZJY%cc}?w6JFTyNgv@GO=N}jPLp)ddZ*xWKD?SDZ(k+vuO-n2#H&hr0~mrP zYSQgUwBIS5M45-}%M*jt4&G7(bpy!UQ0bX0RI97AMKkAS63^=1SJfp1z?cK4Z{_V_ zsRQ-Cm)i0zlU-N!=HC@Ppw4DP`b*Q*u(|n#m;q zcbE_tvzJx=SYMiIxRn91se?Eto27}0>{rZeE4{kvn$6lR3p$%zPP5EVzpJprCvHy_ zQ&_{8tmC?#3Qa=#bR|X0(?Z4Dm}lyr9;}~mRsq{UX2DQH2+Cdh^YWMe7N$O^##tmw z%?9dfSA?Lf-at_m8iMNe?Fu-1cl~P6PkSgdp|Q=-f71NzLt07zi<3c`DcZyHT+w_Z ztjMn!sPlEbR2>ACCH_rIIg<7{t*A(l6>uYi5?_+gw=tt`>PCJ8r=>SB#+;`JF-?e_ z1++@jw-0T<&O}AVpw6nI)YYWcxMlAob%%PZSD`+RJ_qcv6`{-jk-DW1IYT#%6m0lD zNC8Vh=eyg~$bj&0OVY^?`w23A$Dqk={eK}i+#pXa;UUg}n}D-dNzOcr+Kd#Gc?BX3PTc9{?@X_oR~OEU3rFAJl0FHBpy4O-VFgfV%POE}qI?1Kq%K+;#eAejK_Uoso1jGifj(g^&W6=^l*R_jE zJuLFCwZM2Fw>JVG9cPZ z>wU7=t%=ftpamy8eXx)XMSc6v$X zAcyU$zGi5BrN4m|cfl=81f%QTUK9AxxRcAr?dVy?n2Qu1YjiK!_7iwU6{Jrw7?W{r z%wVIa2+g6)72@0TGtD4m=1on*lA8mReyqzZL)XfiY?Vcy@+D`bheI%CLk+R6i=csu zP5a&R=+1f#fOGpOOhSViQ{YSWvAo5TX%Ng6at6a~6*mP7ouDS)vQ`iviOuik`^DUD z(Ruw2+T+L>kFf`uhD6T_#dJvNp(u;cVn%(y%dYQUPNyWg-G?lmO=9yMbTsDKpY5er zb-x1g=A(Q?9ilbA$%wZ<%f=hjYr+33f9)REVnE-|ToihZb-;;?Yk!vikxgs^#g`!i zh-J(kg;5CAV<cNCx|8ni^^ld^qFnorY#r zfFg@$qs@`8O~*DAV0wM;f8vI>^#t!`dJ5fvrvqCsxh(#z+`;KDC>0zFwZ4nqarkyp zMIqmSf{L}EMHqKZDmKkMKGNG5g6s9_m3vaLf_l7Yqah@$Op6*X8xm)ywhuXR(`@qL zhFJkpry5Ey!wWzOUI&dXTKlr)QN(V9dKPA)Ppy-Y z@^I>0-Lg8`PU^mH1Vs;($?!6=q_U4iyMZzxZnQ2?z31)Lt)N&=QpAazPgw#4?ogbe zFym$iuN=ZAkw*>-x%nuOx49IZN}U0DtyPM zPzX6Crhj@hvLrG5VBh!hBF7Lxv>|arGLdLWIEspV)-4MqRG!o_M6*T?ca!p?(qD#^ zsX=h26+ReK1K1pU*5HegH)3pJ0+bJF0c9~%lh++ZwZ}{Z3-2a$&JYbWA<_s@0VLkr}c{R5mhlUf&VL7 zVvn`HD$OT*W9M!h;!C6p7P6-hoa|!52(^Wd<0cd0eL$o-R(ly?mtZ7*F* zj)pezR@YUV?_=wcHa&IeC;*mK)Cm?EqMwv9Atp5Vs;&kpTk)ecH#HdT(JfIX0)bs} zT^McW68Kk|sD|oDsh^E?2#ozjvN^a6 zHUkkYn;G=@cbL1_*2u`eriT6Czih>Z1)T*MkL!obSM_I-)zz1X@9u13$`n?hh2*IT z7^{%TgtC%6(}Vg@wn*9C;M}U(`PKrr?hhb5^b+he9dGNUUW=ArkHBa1hh6s!Z)O33U-{gq zJao*6SUDAmt2b5K7G$%Q4Cxl5qRX*O5Grdr|J%eH7DM@f<73{+RR+wUJ$7{(7kSx9 z@Fwytb0AH5t)MQdU%u0Vd-;Y>0PPKsx_Rk!&_~%y=5$Ig){C%+p=j?`4<=q#?LKvK zqr^D!AWZ^BI6W*mX=r)R1xCcD(8H4R)j$b4{}$B-O`a%Rnj|)>#y4=JoodTpAbrO$ z88vqmDi!-FLPkbt%N8HROg^GiBKI@RY9z?ai>jwxRLVa>R^fjKz#+Z31}xkg(XBX~ z{py>U=Qbl8lW>z&Z6#RXMWT}HWR|t)IYOLRJ>j<*+#vvSrNmC-8wnWK-caFRFVkoa zM%~glf>b@IR!3QiUh~zMg}h|a*hH@F7YJ=Bv^Tvx8e0QL3#1&4x;D`K=IKm`i<+Lj z6w7pYA(jGE%(;b`dsBI}h+jFsmQn??kUL#ULP2QZ_t{NOxnG8benJmTz_XU}tMFm7 zBg!rO%&N)DPRUzZ)=)}CX<=j&jOrrRnW5AV`5g}9g&_6QF>Ib`BpmTbi*GTxr}VbU zf7=s5$aSIIU(gGz@2zQ3g1KV<@bRZ_$!%P+;J50~qm8Bzt51U^H3?kSdrudSvtkOK z8lyd=t;T4M!4J|9%Tzy?Qz*c zXTu-CS^6u>F)&){gZ_G9=$HGM`;V{AS6~y~u}yh4-RSXciKMVJA_lsd{?~m2#M=@U z;G>m?uY?yn1rz%=-+(=WhN_Zb2jleipUeet05U+GEl!v`oc-79ndj+&lI75tb&M|r z%#$<4JkcnJH1HehVwoP1x(9;6`MlJr0~_=ZF(F1b{ZdlPZ(VF2{7aQX?lknZQrWqE z%_}!qeBTG^=xSR<?npF0?F00`P-93n+0x4jDnd|5C=|x!CIH7?C?rfrOUQ(> zw)OCR-E`UF2OH3c@6wJR&x1FA16LbF`T#Dr2tFFHLRb_?tRz-almLoRsOOUcp z>OwS5nHpv+B}Jg{b3)pLXs~hH9^dbwnDvdAIV%}^qzzWl?dy%Zt?;2YVpJ^mmE$_ynUg!;B@88R20=F z#YxBZBs~6n`Uo3i%wviqk!y37%_e*}d{&uwJdS+<4?WWYYumt{?cymOkPIvU!wH}o zkG7kWz&!x1UVf+Ulx)b$PE}=vs8WqN@ttVB7D~&=_hWS;#A#*!jXDg$Mb8?1=}s^u z9Wx+eSL*%jL%#ZRxx14)3dG_JHC2{i`B|t}mVHnnsjp+3svGh6Jbuv7&w|L^AQ%9- zgZC8i`wLLshM3`)Ot*L7n=DK&^o*=d9I1=EgUMJk_Pd-zUEWwMIDHo-`!8J&U8hXw z77nPZnH_@R4|XJ;pF1vF{CW&*!b#$(I#Yojmp+tv*vY>Z>b@Xy(z{qpNt@C>1QC4` z;I?5doo;RExvmcl7C zEw#Z%WI$UOkGI8G8GArq41*jgj9-+LJv4C~ASN5dD9~ypL8#5pV<~!|O4y~_8nS^? z6si%O`8kRU5Bx|s*&tTJPg#h%k{M76+(3K*=E1KC0v5{Ei5D|{$KYIl+_kG8NR7PH zU0#mXrG^WL=KjitBVk5f-d;Z1<|IQ~mya9kwwsMm+PMGn-vl;V(MIVVF`@3}m7h8z$(Xlz(|q+-L)q zsgRvbwd2_htULSdygy(f7CR zWOF@y6Kd!Sm0C2`XPG?&AzxxVzJH1I$09$C5B=d2G{Yz^T-IOI(?-Si3!ZaJ!S)?V z;0!7-BtlA+@5T%ln^+22y_p=nza`P5v;hW#j6A&ZBB$J}h&+C1J&Yd$o^q|L##-Vs zqzy#WY_cJ}-GE0;lk(X9y5uD=1g)jEs@3r&GeNumVAfxD)U`gSOL%CFZHyP(cx0J+ z)a}INWipzs%vj_IRc7~wV%i$KAH^!@5`OPoKisayiWCG3P(_56<2t5WTNh zQk*v+cqCN1tE0+?*Z+aHu}&l@spJBmi}e>Q&_Q z`+91_rHfOIvDS*fvQhhi_3s_dMP?_`KBaDVxN>~ zKJbv8dpgWW;t3Q^#-6y2gd4~rvP8Y(U^2fwXcFI=Gr$U&grI%I;V?Lp)|!Xm1)ugq z7--Eui}#UM*t4JOH=~l)=K05c-6*n@kq2W$wOeqeD+L#t&p77$SVm}2D{Ce z?&D7!g0UK2qKU5>(%{BwDg5pp0sUuNa2b0bOFG>IQ`-Bqe&>Xbb(uSH0yJN(gzbvG z5d{T21RuLFw8xCbWlX35QlGmJf~NHDGO_b=zlqK{Klz|U-S>D75OVXk+v0vV0CPnQ zwddVf53%b-J3$kcY=ZW&JoPRf0KAVZKq^IhcUezyE+ioSq+L>4+U8$njj7P1d*k(o zsXC-SuBkUMOth_s31kK$M{Fjb(8QBrvSjOhC;Le$d44dRONdCo?pdQoG~S%x0*d{3^!E%mF~mec7)K zeDctLEzU7<{@zA!sK`;J=qz#*`Lh29ZE$H*#qE)?6HgAoT_7`6e!+6eelo#zmHWZn z45sY#uV5v#mmQCI{b6oONj$qK73)KZguY(DCT>FT}OmLGC z@yClz`o!JBFgrdJ(GIj8;1D#?H^eeEGkt~<$OO+!{|-U$QnNM`pL_NJ`po$d4f~ID z4pf>hjJ|vXIBy>$gkP}wo(hMpZWwJ9_x*_W2$b#Id~)D458$q$Mg2lsdI5WVVKBk; zsHCXlaR|y_0qlQ77Zr;>KYxK&aLLFm%!Q%3L=)my|H{H=z@o@>75THs6dx_yY)Cws z>F*DtRPXrMMDsdm{4<}y?-%g4!$dY969G-2OJ5LJb1fBK{y1Y0#53AjP? z%Y(mvM?rw(yxvVq(`UBVDMW$*3C#c-nF6Ub@(+~+zO!^&g4z6%Qq~)0z8??Pr#PD7}MtKB^3BsTN zDgY+F_8&Zb%E4GXCodKIvmjwkDDoaA7rLLv%(k&DeM2ttzq7PUE9km_dhK{$M}Oe5 z$gdOPKJXUZ@#A5}9&Td5WZx-7p=oZ})>OGtULINjHjp-njvQgUfyKB!YdL=^FuOp| z>kCz!p!#MaMRkqzzQp)>Ov1~SOug5 z#1Y|bp`&{a?FmnOkq1+dy`1Ia;0Ia9?(DL>0Oz^mkNRvUJFCTqf)Ki~M?AU+8+Q>C zDu`Hibq4w*p1jVj;gZ|B#msW zP?W+z5-De}>3k7qO|#f#Di>$joWm&pFSYS}_2sWZp%yb0>p%ADcQl+XY)p*rTQNM5 zYW_-jv7wA=v;%a|JAk7%f2+4m>oZjD`8#I@zJSltAiX|fX2K;xCIZ5}hpCe5wQ{vEs1C*aN|X5ZI{YqX z6k2T_DDtV&Y0XeO56fq);~}^^xy-&_2OWVm{nI~Na~Y~Ls5y#y zJ^PCIu9t;YmX_c=AX9w!_K>ut`j|F2NBRiBz>tk(uj3yHouwMeBYDXDt73Zp^tp{r zsS8N`IyC=70Ts?(>TMLC@{2qAkm zUXE4^9uGySoRNu8Ok(3z2Q_%^Wnk*fgV@PuwbyOdZli+x%htKlkoeKSZ~8O0rLX&O z&<$Z<#KgRdA)@T}Bp1xH@;0+ol)e5X>9A(uq;x2X|J!h#7Zioi*6oYOB~djqN@U%#(4fto3f>w(o>~IRQKkaL0Paw_b!dBO z(qLAZ*~`n%DlZndd0|vhda6%1zBq~>49Lbh)%VX8Ad*4ZCC%tLv7^8;(HK_3o0qx} zIdwdFUQR;|+S7Aj+6H`h_DKE_Hfnc#DR2X9{X#80Gqh>xJ1%5Og*eB-x#34}1xKhn zq2)x;Py&+T&VE>yZQ2fw1NL!a?Fn^FRBWF_**ne)t026zcslBdKt?_AUB-p6Q?st} z%~?)ROrm4saH2r)YnBgNLQYnR^r#~WLQF<4_J%l2&iQ_`pPIig*&Fa6SpXK%2}+PA z(`p&`PC&Ou`6H75dT&bfPza`c_CXSx|D3p2nfjM?ZD*3~W=y+;#v1A(x_VycM$Y@_ zSwBI>4~N6}VH`587MCFenCua72MoGP+ihXTZ=6z+0BL)p$KUe>_h}M*&cpDuUvGK7 zu4{YC@A91ftx(k5fbumbtBR@jU)^8JpE<-QrspiI`y!7EA;pu;?st}M*-tCq%WDTF&;#i9QQD6Bx4Ri4i{!3q!oBQEwn*(ood-JdcM3O=Hu77 zmZz^!{H-Ogg<~?~q$qQgmylfMee4U@Y^+z(Q{IKh8?cL(?8R(nPjS!Z>e@_gHb4p~ebz2aRUS95RSUYC56 z{$ixX;@agN6Su`$fIkhU6F+>xP_VXln2qcNd64sO@04lB9T8({!zDQuPnrvlVNHqG z#y&mz%>R&E(XeM0XgI$UnIB4?H=O146{mH`w7qfbQ|9NvHXq7IIHBJS;YZ8#cO*++ zA9+w-$QUUv*t4j@3V#=gcRp^+JTdf6HzH9_EcZk|m%RJ#hGHu!54&ZaT@l^kFQ&Is z-`1fFrB4{1Rfyf=P?*{LhjBz1rj$KB;F)GLy)xK3#n zR^yy~nn>NzBH51Lk$dN=W^7Q3wmyB*?leQ7pdl86G_e<236eJL!?e(WkYp(Y2HVzP z%6QLdoNQk=(!EE*lxQ*c8zEC8LMGbz+F8V_^I%?;@YL)YSymSl7z2=7saMly(ZU4Cc~dKaU41f^XOjp{9GrQ)VVo3Bi=do$hlRMrLwDDebGFX$#wk^h!h3 zgNOq000pv=G_mOkkt~V~Ad1qUykCm6Wxp4DYhf1&Q1apvm7vE!Dcy>dv-edL)mA#0 z=Bxca*k4gE7V_@Tad;D};gzCRtSbMfr?I_Z18ZTFHf-lxYBg2Pnn4cKE$0*Ybg3kmXz0pY_ z_kDYbNe5_kS7EvBtDwd>ef9=x^O#7UaSeOgAiZ?OUB>9kN1CIi!J5WaVW9;q~wkG+zr<|Yed9pj?=^BMk z!Zm5Q@gF}g)p+omr`gG>*$V01v*N5XVS9UjJ9|B6?HM01OUy?G`7aMd16hF9 zrOj?vj>!&9sUQ262eou}Z0fLWzm1#1 zRIbd(Nb<`~D1S6lc@d3T*9ZdU6<_E2jciejt!rwo-Fwyw?hp77RwGA6fLZbhInk~@#GF6Mk zWn46f7vASdh;xN%)=KM|i#h8|=)tC|nh8ZEJf1CSzNS$9?#KVZ^;QjJQ8T_lSDLeJviDdvp)e=6ga1ZWl?IPQg^8OBb9RiY`Q{D4sCR zzdEweI=^)1xQQA<@z+EJ8c zm*NmF3MxYwa1uq&Q=t66vES18UX6UgZZo1($Ig(h-h=74WbQP?`3EG?$7!7Hy@E}m zKUQ7Ve$Ls6f8|HOI+1aJb-rW`Q_uMaH3DHg2U&}}6B74LI>937ZlCE!XyaoBP~z7u zwQ2`+zS_K!V0FvRkQusBWzligSQo=;?gPXx%<=6{G6P_YYaoBz#op@!aep;9-%Et< ztMvZ!u|FfWg2`Up8fX47A4HMgTI{q-+B!Bh`mE4M$&Uwa+dLgMpubjl*>lKv0)u7^=4k4+ zoI>Xdzp+c(G5Y9h|CXeZpYEuJgWrzsjjnamF?|x(9a7wmTGI4)jzJWl;jD0)a%b}ad}`Z@v#yS9dQtyqX_s?LG6JlyQFXh{NTV8&uGqZ zDoQwmixTP3MObxNr-slM%~H-BwxfCG*jIRU_jiN%JH=P><9$8to$)QvALC4keGnj} z58XYg-R&)tXWy!}x zcahe#hw@5D=mxs$l^VBvd+y_LECA$YekJGS%*PC#3n~iSW8|R;_h~?~xF2D;r*l9> zeOd7&azl_U#LIrHwe5EnwHP_$%>wA1Ke^DfOE79WYkpH{miMT_)R*3Y;_hD)I+MYN zQAT-KXQe!6UD&e$5f{4zY-bIL{bs1oa*NI(b37(Drg1pl<)n6JD z&QnSEj&WS07YbN?=o-DTq^vV;km-(pV$_z}7ZH~mg9qZSn;7WmganXFBKtMh=e@x+ zY$l0~zt6n;I3H@n9r^n;VK4wGIP~gUQ{oBSSi<;mf{%QF@9jT`!Z-~I<0b-{t`3+( zl_`OH7D-$0_|}=&Yv>Oc$KjBy7LOz;`vdbRJSCNofktf%>?QVftIs`8DVY3F_v7F$ z@GEGWwx;oegAGj@b|!qodq@!U^UWwKsuh4i9fA9?51dHaU5=ZxM;vO3Z+$;!P8ZiG z8^$Z8u67VWfS|*_K)}S);O3`O1ks%E#6L=oGImg@BjzVDHwbqzFcnO2(z^2|&ZLI$ z@o6l_ft(#PUc2YFD!h>Kp0^z~3jeV>Mg#-1>dDEqC5{OIzkXw#N>0y@4wd4IyFmfa z*O?!Gb`$>0^X-J6Q#HF-%WktfepAhdGwxiL)t<$?Y4~#UX925{h`!h$J%LOAjM>e> z%k&Kr0zEL!^a#A%0QvB4m`4`gxdY*m^z&`K7C9$z%fC`f=sF#Mn%Owm=;i4!yLoxKeV)$*plcw<8P01ITP$&rsXC z&II-U(Dvo=P_|#(g`_BwLJ<|ICrchpwAhl6q$K;+MvXN@S(8dph>$5u5;0i{*|Mc5 z8vBs!*|TTg-*eAmX8b(w``7#V_^0P_yS8(k^F8N$&$;+y^-5|2bjH^*8rr#wE`D#S z?5ujtdT6W(xx(V)E4&+@3L$gaSEW1)sDG$oGe=<|nU_5*5_$Fuj8Ku68BW zP(AkEz@ufTA-5s*mW4>~Fb{a3#ph9AYcf(UzO?(f@tJIVCY4UnlbS{r1snkoqULjB%c5 zxzq1WF3?BnlxPMjC$(hX9snKk*l}&?=DbINwDmu*G?iX<_iQKiP5>i3y(}U&-!$z| z8&UfD_&d+h1@VmXIwfO)Z~ycZA@f+_!JH^-$bv2YYL5HUaxEbn_9zPN27P1kr;(r z1~(QUh7HYkEmf`Kv!G9vntgDvDTkn#_{a`ww!@CTqpUI^3t^@#$w_@ zpiq^%dd)Uag21j;Un?4~I|8jf)^N}ygEASjezs>5o_m7zLE_ONRimNamqgt92JHCS zd%HDLw)g1xL9rZxZB9u4>Vy+|qQ`GHduux9?A#e0)zYBn2!M-jjbj!QODI|ticoJT zp5I!ap-+n5E~8|+N=Akx3gu{2jz_2caZp4x2xru(n0(V5g*GV8Z%u{D#NFY(vLTg3 zv`h|nz!x$_QBe}b*P0X+QOq2Ty}J%*0#Ut~^SojGAH84KSfPC+x$13MaMK-P=5{}x zxaMGgLqR`_eau!#143*2Qtf|s0aI3h6xgQi*)b0=dtWn8rp|?v(LoUR+l5e7U`V8* z$q3J<^q;Kj7EK9V?{WwGKDnXz4Cnwx9~Sp@LkM({ioI*t^6nv!h5V@#mP;fvDWI#s zcc~v=IbHyck+(-#&2NxZM2X3xu;CDPdWaTg)U#UM3`p;ZKBK&ns7q8RLpoFFXY* zFZ+s2p5KpQW|we(+-{nW-&RDG9-G>?=fLTB-MWaAOun{gjYjT>r8Adagi6`QY;m5swk-tWLpBO@Y! z^4UPu2sWqP8pud}dBY8kf|Ya>1mY28Tft6^?94IO9?LxiXSlD7TmR>QI8#HR(@sSKOgVMemnd65KCNl*)t=L)Ui3A|Rkk`h| zctcISzH)#Pp9@$8=?bHv_>k|Uond>b1!E zZQ2U*GRR1^8@F88qqvQ(8*dTVY9#3aLtYuxQgh$BbBQiMO9q2^p4M`VGHyxGD|0wO z%+bF-&wvS#) z(wt^CfOnS_??RvfhU)|-bdOQZH$Q%z<16RWSHr84*?%0K+n=c*_^1qyP1M3$lF58MtKBQZCBK!E`yfR^|BmWQs9+o5{d;p}`F1a}A=R2yP&$2HUW+-@kET*td&(h_Dm z+xH%A)k8BHa5Ws`8{;4&|B1wuC1}ni{$CM4DA#EVrD)7uCCj_{`OvKda`@}wb(I{? zs-07WGC-QAQBwHD<-iGdzn}CcC+LC(vB@C_SG(*Z=$CBDi}drpfB-?-upq4JilmI_ zH8bO(gm<91UcD?>vjIB3ja*I8E4~mj1XbQ!Nydw|x#Smkjk*|te+91($(ytyuIALQ z5&(d{gCyyvu&l znR(Q8zmWU5Sv5nLFxoRh?$aGTyPhzior|Ldyga!(rKe%7HK2-eaMZ zupuY?1367u+ah%AIv;N6iBht>14q$y4hY!; zryTb|mBKZulqKcxt^_K1V#%j z0O5l64DFlEq4_~qJlM%C-*pYj?tp-5;PmllG?w*~q$of}sNM^iYp8gE1*hnOfZ2`1 zo-r6_q;t{*r}z9CxJ#EvM046Kgt$pObdr{DsigUCE%IL31#jR&oFSe7Z9iQT2qkTl zH?E=irc)xrIF17#4tYbCrO_FG-3b@|>J=##xrZ6VU_qkg3&&WZ&{FU8%$2K5!$`(ZB?`OR#_VMO4BCGuSG zfjv>Dl1eJ!62=P~KDzqm_cxH`Ka6Q>Jqk2%sbh5&zfStCc{0LwSMb(~P6pQj{ISL- zAI~IOXuBA&#SxXCN&M~3&THOUIeu#?C7!gDEO8yUUK@uDTk0CHQ2CAGF|*s}2~Q=-^*a%+(nYvn6HQ4}hO@99R3xG_Bw$WxQpy1Qtj= z`u$d@;1lZ|z(X8+n{P_yGx%eOYQJEPF?jyrxgMDfOH&(E6bWv$3$$i<-)y7?P^C@k zg(|l&zW?Xf*O#E+`wj_x(N>y&0h#aB{5Rn>T46V7mT)oDp|wY>v-o9cFbHJ!6K&@Z zMm5yYGyXa8liH9}J9VI}=Qv!~^mb;w0j(c!c4tXvM+@AM8{IS|IbD`Z=_APvo8o&O z=RG+LLFG$Ugxj+M83|TQtS*P^2JRHo%uF!V-w?(hd+eug&a0X2wKiV&hX^&xi%MAl zIZkyX>94-3le-{X4Tko|qE|rY99TI2_QF8KaWG7jPiY$BirXM9*MG_Bj(pW2MGKMv zqhrDoYE36E6*)m$bD5Q;^&nJUJa!97;?d6XAQkMueNhEQ?a_XtA3A_T$h^K~9WCjh z)Uh|&=k+P~5K_sU3B(x2Le83JSM~NjxJN0PBPxKUxVb#Yoq!Cxk|-OQ3&7x-T_r*; z%+QI?^g%nk#f(l0(!M-jZR{=o35u?KF;DvwUCWSvwaCK>RXERA{--0P{tO zHPk!PjOgg05b694%h#B+GzL6}9)fzH)jG8o3>LLIp=xK3Gt%R6L(dkum-UK(PzH+v z?P9A^a-G0Qj3M<2qKzqCeyC}e;K3- z@J9I-Jo`ToQe4Ak=CBE-*)Y<~3+T);1<)B`o-{r;3^tma`qWlm^(~Yk5GM&u;b-5- zdLH`63}R6GR>Yg0@Nnymmpmd`w+z@wI4c0Q=cr@jkshOO&*9qq2T5)xQpYDV_Io6F ztGFBVTr!i&R^{TypANP<;m(-zN%Jt&pDFJowvwaiX1-MbkY-@#)WQ1niPVQ9(AY+% zu-@=9e_U5Vf-;J8e;y&4QB4LDv=9_Vy|TagROZK5ijN`fGs0aq$j+#Ys(vz+6gj@0 ze)7(if~q-ch)iYA2+ri*6pgl>JtRo34C3PHJRXY|zzhGnOY&!F^2*{6%YhoxKfdHT zXu!Zt@aU6y>|Z8dWYnGix*7d#yG;`E47JXMLfY z5V=psl2N;;fs4NEmr-&Fo(&?DpYkJb&w(pklk@lnpFB#D%$cLD9}+(iA9~D*$1?WQ zxH6>)B#S?^8-R};4C-6t)?8pv-MgZfqySf8^VKBH$?YWG6}bP^RhgU6Ql3X=(OWHn zeM)#jYguVJOtq?WQR~n@%zh-5eFq*j0V+d~nG60J8vx}y|11^2EQJ;xoC2>sg*WZY zp)~&+AOxZXx^9c$-+kbS&>HRBkFSQeOG%25;2comkuY9|z>x~atK0D$is&Z{aQ-1t zt`GGk|9KHIzV=5LV)AlYUn)qmIP-A z5=Xg=n(?~7&O*lp?_Fsp>CE8m+Z(@{f5lnw{hshTp8REnXC7m+#3A^*h0pE%hiICi zgohLi+P0i$e#aDimel2fCiM9LVoKXbf=fy^$}17QfD}M?NXX?yaOGKeD~zVi{qf*p zPkLegz>bX(aWt*%Jj?>L^s$n020-WG`|DtFExJRfi6fmhl!BZZeS~65x#^ z&+|6t+#5k$K=LJ^tsPTQ2!dH4H(Xuj4%;C9IfJXEu@5&? zI<$|x7;ub0{#4Ep-@Zt|@&Awnm^Q!q06<;Un|*cv93gjTf%}QHZkosnIj{gWl51t) z>5ovlJ`0;~5rmfnkx+yUlv`&;W3~tmbl!lR#!4^7;-Bn<2%1E*m{sQqnn?nkMi+mE z_i7CEf+NZBG8hl}q>OBKUQeS+vYti8V^jY=jvEww3=bd*J|J8)V)JyZ;tQY0)~$Wl zy_M1BVwf028tdqQOWVYC2)ssYWRm8XGg%p(M!_qGR_L43chTO$E7U;1jG@i89x#6? zEji0iqg3T*zoU7jpk)6(v6z8SKtl0_e`-qHw%`Bj&cf8wUiU1DqGnE}i3~VF`6r6!mfh@(w#SP})nn z49rA}Oal}$6#Ro2u+8Kg9F^Y?h8`4ymIvDZ9B$<#hBJ^91d`;K%;r2>05Z-`l5twa z34lTerLT11Wf2?cGl(9T*~nfo##!ClA_Jy<&yodNdJu<89wH@nr54&ZI(11pX*YOe zQO|OsQmk)o0OyK*WmVC9C*0yo{B(wQ3P1p)AzD7FbBI*Md(V z^(x+qo6|2xT04bs=rqQ6uChlSKwu%odT22yaBS-z`W9XjL91K_abdY7u~-XhY`=9V zQTxU-J=Q3u;K-G=#&JlOF4`Jn`A?D>&cJ6M&{_w!XDj*pqqm-=LCQ-g3Cb2aVP}HV z-c|yHIFeoOZU(tgX=n+X`N44WMg~=rl@aJk=&99CFRpbcmGa@D=IGxuV_xMYR%)Ou zJV?rdw)o{S7?H%{ye4(3?_j|C$E+IIK4Kdf?v-C9fDu_C&-%{fg&^_HkHPCDb+4;G zCGw^unM&rLf-7+!xe~N#3Si}*F7=8W@|*7?sX^c@^e&W+v`AktI2-BAz_Yc+6!am~ z&=Fq?%YN5(Ix$@Td1ICjN}o;GMSDOF#mv4t8q@o_C*8SkB$(2mvmZLje+B<|ySwNF zqxZ5W;Fn+%QmWAj>SMj=(I-T$;|i8x|3Zq{fY)HU_FvHFk5^V#a5(c-H{_M{YQKSY z8ihWi)tBvwicA)FN31s`UaxeG5-iZH^maJxH&)uUq!~ z4s38BJZY6ON!2WK5eSdlkDJGe?x2iE%k372dKrr^#h2dsxeCJ*_DGx=4hAu*MJ*w( zoyY6pn&hu6v8TXEYwg&h`6P6wc-zs{<_anOuh;f`JoM)`m~r71ap*~;jfnBE z`C7f7Z((188jpQSa)+l0Dy#Jpec+{z4=ck^4d@hk@M_;)eAUT!Z@V70cKx=5gD8}$ zv3QcRzxcWwJOV)7e{ZSZ3KSsiX{f}n(ieIGeJNqUO^pVh5{$tQ1swH3kCTl-^$O{d z(BzSJ$0K!TeX~>_0oR31{YwaIF|JOX9t4Jkw7zA1lAa!l6e#yCTPWR4FH=_u4dLu=Zt#rGFYE~ZU>l(jTBPfst{ zDa>%ZB2mOB?cpqjA<_h!%^I3CI{2@4k7RYu1M zgmW|B7qzvyqt`Yc95S6p@RrZ?q91jC9_uA;! zsi8t00!;#qOS_f3R04in5)@WCpN@M`0+20@gU4wsy zKlvF0^RBqifLq~M?w{!~JnlkqD&gLln{QR%Q-yBvE;e@G_<8i^I^hqS)&=-4O_MTS zXrk$G(+=Meth%I1B5^o>Y_CJdtuR-+GpYi~Od0~p$2F?Avj<>=*F>?>sj4LN2VnL+ zGRwjYx;!XuY#fb>?i$U=dF&|NfzJKt6`R26kP<0J=TnSP_P-Y$BPk$&(p`T-MeXyj z$Gzx1*bxKm6w}G4+ZVs%q{U95W&9dSZ`COzcB$*5bDhp}v8@x1+Cn{3stxT$8y#sE zpW)#Yi2H84*L2;nR`i3ixkT5NQ~cT1qEGYL*Rh*CLzZ|lGKG3eGZSNuiuQ;|PJY@O zH+{@Xtlf|`6H{i0Z63I$bGFUv-U-P~s%R&L^01vJUI%cX=qL?vjU1Y4$&BtQJgnP9yLDy!0CP9eu10&m!BPP*LZYnPa91Cx6j- zr?t~JyuF!?*pHf;IDTX*6uPZQf9&cSb*y0x9^LBHFH>%MkU@+l|aS*Op&6Ry@ zf~a+cYvG{Z6^$C=^1>dO${5)Cl*5fJzcX>E`R^TUuQG0O4LfX&0eNeZXa%W0F~-@v zWO82T)Y2MLE^ocpuWzRhC;E&eMuhu(3eEn2bd1@U0?F*xjaav$H7dku<{08OH|Q8I z2D_SD!OeU~zL~bj1><7~4lt1yK%HkVcSMEl?CC3R{4D`jQzA0wdHR?Gs)Ys3=r|7T zrVpv8MInmd%frE~U-XpGGm4;!#(N6!bJaWU=KjU#yPSMhm;kx zI%IOnf(nuTdIQ@!r%{-;o@d@%L$BZjRW4;fUHk-P24d}$w>E1S-#yKA1- zoK_iYL%V%cdvg<{^1_cGIdyG`#t(108VnqNy%%36Oh>*Th_F&__EcxQ;58UC)j+-PnV%icm)=y@nXUpb-l}(<$q1Lbz0U+!Q^B-;6@f>-y z=SN4yZn~9g(D^(jUyEEst4Hs-2(LkF1mb&2@h+B-5FWC9q3#=iQ9n6xx4EJk|E?Y? zPme7QJ2R-R6F$yPQqKe}OVJ)PHmJhQ`tW=1>n&B(f)&J7Bh#_HJZ?huj$P@nNMZ6K zmmOm!=4jG8zS*DzvviYF#C?@E#6{FrO6D%~zZT0b?q^>o{4e=LmKj+X)H5$^c8#?A zM{d%fvDn@-*nM%KirV?0bdD30PBdJ|y>TjBGXDNGS5PyyD6uBK+0yQXU#jE*)yy=1 zWq0pzspoY7Lc8ML|4_5D{K?Udx*I`rosdn8*^iRVX*(IQO%N4vs$Vdnr5bO4Y@F}@ zF3<%=)VhFXXwSO5cZ0UZZZdT`0ZVdQmYP^TwC`opWd{)2P_oc&=%F#WAP=qT&IT1T zl2&7JpxNWT7*)+k8>-D7F`Kh{r3@NDn`EmA8ez%gRz_fN|K?KV6V1d60TdCUb3TWi zv_~QG=g52&R{M#hf+{}I+W1@pTo|7g{Y*(VuhsY?81YfQQO%zm1Keg6+)kdg03t-= zdN-^?(U{xUvdCp}A8e6vWOpQ;NeFh@?rT=r?VACvmwgdXaeie`0Ry&Xi=gu6I`>~>dKz|f4HCsm+SUNa_p!zWuAFM(L%{${^gqa2GimQjHJ+uVA#bug zEf`Gi#;?L2I}%O(%lp^X&)-CJ`#+?O<*IDDtHZ3KmWSBN9j>Qvt)3z%S!p|q2SkKd z0r$?F&q=?2vNIE|%0bdqAsH<(brR-U*jC|LJ^`T^voltEi5O@M4ExL6?d1b|UN#xl z39{>a*~Kof6PYP&J+*|>GGrh3AkCpY^2RqjKUD@ZMc$PF< z0-&MT=o1l%U2mirT<1Ka_Z0lor5Ndjo$i)>x9NI>@8~c6$pKJAKBR56Pv>7YXqKl$ zGibDTmy8&kGB9WAo@kfnIOp#Ue3rj!FX@VW!-JKjQq6#ZTKWsP`vNcw*yo(EqcL`` z;~lO4E*opPImpHoYe{rXm8?sjjeeVpQU5&tS!=KjeWXrmD8jK-8WYX1_1q(pHP8S{ zGA3`bZ{5}t8WMt)N!ZNx$ILSe!usbm0z1$CVDAC*0GOm%cC>4Cea=l|I7(5{=70jB z!UQMt03`Y{JC#%IYBpBa&T7Z-E0aC!It;?3(->d2$UeZbveVX?O+o+>h1#4h?svS2 zD&(CtmG5g=Zwlk(lkX%G02eA7)Q4s)2yXJ9{v&2JwGXc3zEq=MSxtsGL*&yn<*hrE zO-9~sCaGvzuyq0^JN#?M-Gxq+uD#rBgHq`rwhmNv09jQdN|yqsqpY<@ zzvOa`11zPUx0)H_qv;H{v6Zvm!}#s4ZY_312A8@3GUHr zckPskLI+d?s_!xB>NJ6Rm~6OZ#M#!L=mcKCu5Hw+mv#bGp{lb3c7trB%VU2NrC@gO zmxT30Op|sJTpoelq~ltq(g2h)qWq_C2Sb4y|cJcst%Y@{m%Grt${!2!B6SgmdDHw{5mfOS}B-xeUjeXKTGBU z97I4QmEmv%>+1aP0GP?gYl4Y(BP$D2#xhubr02s?Cljz6ZU136767+edS0Jv+9)|c z1;>6%OUxQK2-t2&zEUE?+z2<8OyiL(Ift;@5==5uo*oh6L*(tofJ=NC$d}4 zDAM^nBxzOgC~&#A24LKt!d;gBYlEV9W}2DO2XlGhho_vRw#1yNDsHm;`FQK~P!*G> zA*91uhI%uV)yxbkLSkEdw1{qV1s-p2YU3W^dcF*?aGEPyI37^pbrWo3L9yiyBsUhM zkazd|3j>_+$GgGGpet7ZyP1d+n6R%n!p-~?$);klz4ol4E_!M>DuhjpCQv{btWIoT zI^H=p&qx#kQ!m4hAq>8ZsQg9xk2xZcc3pdyItf4XTg2&<`9! zo{R*3j#D<-^q#yY2$rKL(x+ZnZj1VMk38zpy3jisG)B2|JR`i%(fNEKS?^^~*8mRc zXE3JET_#PMtb@DIZs~8juzC|}hY`o~Ea_S=bDC{$pwwmR_3V7MyIX+YA{-ZzVI-+t z{l2nA+O^Cx))ownIJH8cT&ce|z;xb_R}b$xOVawk(GU3gRbNc1=Uta`5MxC~HauUEUS$!`9yKtDEt>c_n#nIC9LN`cG zPgVvdqQittbM{vhGr3fpVUF;8a)VI+vDgO<(v18bJB1m~EP9V%H1f+8|uX zjS(Gzj97WMuzyZy4^KXu8c z^c%h7=$@H~`pJ$QKrSc9$c5Ig`1T#&c-y-Ebo-Ec#&j`Qyy6>9q2^Z&al zU+rM4FO!c6i7ZeF73>a!{)FFRgV4-rk{`sWU48ghv-C|4JqOM_L$H=7$krKgy(k{k zOGK;&rPCKMFWe14I(TA3GPP>)EV=cP%ubUhSmzIO2qr75k*+FcnKTCiZ1!bdpE<7t zGgi9j^lJaj8z}dm8ly@Hw+^uDTp`F!c$btVenM1SQ%vSU z>YyFz%OX6Bqb0NJyz_)eCL+Tcq>Wt$v^qKg5<;BHk7|Ke4b{3ItX9vJ#M>PJ!$c0@ zEmIGYu)^>Q&T|ccOCCSBw#2P@R`acO>Pacq-(A5PJ@qCU#7if@_ZAUA&F_Lc@6@pe zMW21nZZ08u&$|5+D*S!ug;X)~@FS1`57Ed#LFsv7EN)j4<5=s9&Vx_fLBKD5 zm05Si5GRstJlqf{lfL*oUDf1i07)Yec6G_~c^ZM8{uPYbf3&gSa@4j8K84;liGep! zP27zU_7_G^KLmg#v@$^J=t778Y`?1&eo$D4s2#q+Mf3GokeT(J3jxR*M;dS$r1r37%l;qaeQ*W^#>54MVf{Stp> z^yf$zpvxbvxw$ANQFfhZvb+(n0`S{+^_uepZm&KcJ0Xst1arg&U-OBSfPzYD5l*w z{9r<@-DVJh!_*p=qOENqVj}yKy>*|2c`2R{oIC>RBi8lhd`a%BJ22us@`waDK?zjP zK@=racXnTT3<);DcTo8~s&bz9isZsK_>w`8dajg>^#x=+8gf-KLG~iqljUzm9~qnu z++%~YntO_v`BVR4<^hm9vtvdQiXCf7u~Zng)UkF8Zc5J7hE*V0hm){Q65Qpi<~HKl zx52ulH(U_)=rbYScrqLyTG~X(?6%2~22k4iC&X^uKLb}UGwSx;+s$hg)(L<94@&`? z<;gek17oee!_u4-RN+lD)tSANiP;YqbT0xIUy&~8aYC43DF*ZneAi zLzacT^RF-5A^8Gr=m+`}@)Hx4wYMD#)pbCAb;vWNm0xgrT2Cj=t7;JE*^6FCe$(Zw zf|IF`X-@#Vxd-u}m-%cRfSa0nv>f>M)05(Vu%)faI95ODq7MzOB>;x2$jhS*9Kx3m z23uQ220I+#W;@<9`r8+l*aekybBDP81t3hm*vll>z);Q6VZhPv=1Y%hw*7hdKvm#S^T!?Bc-o&n!iN|{F#hyENdnlHXcH}HQAez0}Hm^A)-nO<;KwV+LOMxjRQ&H;c)y z6J{s-L$onRTL+ZqW#`($Ad$hYHn_|OB0}cl;m`Tj%*pT{1RW$o(m~5MH#gB#$qc8| ze29cmZ6F8Gt=Bbsd5GP=sXq{+z?+sF6xzV*?QBd)8E-t_VYC9#fWfD-mT}YOa+1vJ zm-LUd@^WNrhB&NDc{72tw-10ZdDwkFuO|yrJYlaF;oFJcIMoB@#0D`Z1UNyt!p*RS z^*cCI7o_S`mIf?0e%G72g!dLOd^=FwPlzJWnuT##`y@zdVojOVXdI|$<!VKm1v((W2FLI?wv=(=26A? z#v7(mQ{sC;mITNLOF)TK*4GgQSlWKSZ02h(xNi#fMlNRv*>p*)e@x{u+7co@5k=qNsCWA-uQZ1siZ5!k%GeIB zH{+z3?An8bL<<{$N@Ah~$?s4Q=A0qUq;yBd$lEdu5n$c{BPfxvc%&@R@CPV8TXMdZ zHsG_%QCs5J$2VOi;*(>LD0Ep>=Hvby@bfQe4`XSxo9D4r05GFg-qkhW?mC#^;Ii~W z3}kxMOU}iB0rZ22fj%x3Tmc_X0OkrI^1Mu8HX7nuhbBMr+$xrq2IInaXY{x62Sc1{ ztYcAw$9upQXlf)(5#7~^rd~X^PV9~)ijWE1oC$b?Yba$uC-gA!F(trl4i;cQ9Y9aAH-nXH0@fb;4G@T{&Hemyp*rCGm02FZE7B1bt1 z@jHr#fo$r~z}7KGp*%eKBnQx0j?(iI*|zcsn9(AEnJj%a>ev&E>N)%Z0@l2C#?r5& zP1O%N?MwzeH2ck|W?=Q-)c^>5$!3-o5bVu|B<;cn%wfJ8uhFEd1^K-MdUKcgbcfpx zm&*;nM%$1r60LdQK+pq7!HCEp*6dkL|D%&)Qi)xTz!TWU!vj!aD=F|qD|SFsqn zVvuz^O126nnV2@o+5BydsTdE2t$iy8$T`6Im+cq5nRF&2#R!g*lUsz9N4q*)%Wkf? zwFgvJ8nviW+|zUSkeG+YXsM`U?fz5fiCwQdS$_9E5({D*8<~WtXu^~|T`&+otfr@|oBqZQ zVe5{r_7_QDg|$c1L{RSY4yK;{AH%cZl5`LPJFYu7otg6}Ha=~3>qhj!Pp+}f!H~Wu zQUtzc0GXgk+ivc;bKfSxrSZ!l$)dlVX9v|@EsJ}2-#{(8T{Xv9H|qQCjjN}Byw&qg zC;EuvG1tk-)?w=nBM- zC@zS89Wi&CQA;{Pl380+81r-1yzQ33O*JWqH3%aK!?GytdRwsa*{faWg&gaoPoW*U zquxA#sph8n#kZWP!|av;$WF!w+bU?X_OLqziPH5GC2I#0ok7~m6je8Pe1u@p*s2BG zxN71sgaBx=W=nv5ZwZ*2Uq{>}9%YkIkOOKlb%-VT?~!cuMG&NiWIKN>vW%S&Wq`9O&D}5);$^KZLbTYahPa^#@$na> zA^i?BNDlQsoCFw^#aCAMV^>MAff~EE%Px-^JPU8!^kB!MhF&vZDpiqmA1y%vR>^Ya zWf?|x$IIs@NfB2E#FKZI@YHfpa1{oPdyG1I@C-}N_pT4bgbWS5V9Iqq%SmE<#9_JT zk{|ya@k%oK@S@iowB`s|YtrV1&S}y+Wn5LoX=gWHOb6k6;vM{zK?4wE4(P*UKZSOi z-2(7Nc0HGQ<`)4H-EOMF^d)5$T7gd4zD2}mT{J|4gf4eB@K1x}ek92eA&NkSZO?&Q z{2NkI6aH%jiQ5~U*p=v-5;;CpkMcndhDZb=(ZDFi`ZwJ1t&HtMqI?x39lr|PyfZ#=$z~JJoYX<&Aaq!Q=BrRSX$GhBoX_%6M`1Sj#a4ID& z#nA(NiW#(T(lL8$5{9~ixLDZfKdTpBWvUl#fLnu!tmSi|Ffje{hEFlfsF z{htAaL$@7vyCNwaIr72MCPmiUqZDFOy7UL%tHO<;Q}kh{vxj^ycHU*q(9-`7XhMCG zCfw!$kcJX#+4}>}^4a?1!Krah$IvGbI^Cmp96l8bE6W?=<1#IB?BWzm{zk*ES&Av0 ziv-z8g~w;YP$Uza1^Sp}a6jLN{aZ6Y6WY+kKq-L;IzwEbSa7XD{cq8fa~cw*CsdX1 z5C}Fia1rRXT?0^QwG^gVirv0e*0$>o6GRoa+z6>hxKvSs0&`^>w!aQDX{&-{SqtFI7J5NtAU7^-FBBm4_VRIi4N%EUlytzXJY zlcwmFDnOg{*WKpp&y(W__9(}JOy64d<*TJ7Se z81Rj>G;S(8+%g<2CX$T*f@krPN{FFmku9UG-3PLqWVUk?_VQq=1So9w$`(ry1oxkb zU49dNp11Z;h3k+A0wqli63j7Ww= zq6$5pA+bb$--%ZndJ4AdqkSj0K?;EoViX`?IJbbBA9uHX1BT9}V1%-{4Cgu^7uz(c zA(BMUc?7(^$*;GB;zHM;{DF+|MR!=n-S^p2# zW!FK?%aem%1VR|a1t>t`kgB+#9pqaLE3_Xx23%bUW;8#9pdRTiC3C1Cbv8!mj2TX3 zuojevDC_*y&B_L9vHOdF_p-fmRRoYSKdz9>Qw0Aa$wGgW*Rk)Uidu-*!iRNV5>2nQ z+(OUAM#nt4cBdTXrN08__AW#&d*4lCX9WgHq5gh|@@m+1miMFYBYKj|IwIuAC@SO{ zgjB9)WBQr|?-JEI8;~U&)E|pw*W=eC=@T*mXgTu(b|twl$*({Rf(cDe&Jb~rNU6|HhE{B)!5tguzJb$qqC~f;Zo)ti(U*{xmV8O zMqPt@;3Yji(U7=Jo)|=UXo$0(xHy*U9MRJ6JgNti4Iqo!GIwDR&8QFVf_75o<)f$2 zzYdNy3c@;v09vGHpK5bgp`U!PVDPlOHX0<)1w`{fvdY>J&V@V57`6IHCoHza6Ud26!r zYJGawN7${sGjBybTel6<7^=aOSh~$!XoyZYeI%%#u*xP?uv)AM-&79vB0*Q`vw;FKU%5%kw z5lv4{^e+p6csn8%7h|3MZrL}@4vGjNur#I zIpftbnRXys?-hb|$rQNdSQj?Occ4DKSk|r%&bZ$6m01|WF6~jzrgiPE-T&NQ6!B2} zLR#W7th*YJxo=as)u&R+3X;WdM-Tp#$VhVemZ&)l>Y9%1iqfj}vw%%A#aTw!obgi3Cc+FHp zUWOg0v;d$`9RNzdGrOm{P<_yPBaojZJ`*tEsM*S%z);8o|Cz72@$l>DJt~ld;?f?C zbzPOSZR{P!V+I|wBnq?v$09j!?_#B!&N-}lPZWh!+SNH)w_g~>_JGhWEpr*hMYeV= zTartk{TmVzLK5h#k1fnv`d>hGg@oz<{%kUtS*@GttI?5E}xbc9E74GoO&YjU<%MQHueEVRh*Mq$}-f z=YT%H;GHK3#{p@!QtOtdi|eTpf{55r90wNg291uYManky!n6o%d>d4{h4!eWL(pf$ zX&_Qfi0GE<_~o^$5BNTtvUmu&_vF?(C22{mb2N9DwN%S)$NVj2|ox0w0fWeGPqo zlC-)bb}@Jt$56vJ!*G#Lko6UxN${*| zft4?YRB#lxMw?C*Y-`GO;!-4LV#YxWaIE( zQh61Z4!zTGefZ$Drf@R#MM}>h%gBZAZdDD_80~ISiFX;kzyIC!G*40zR<%0GLUzv1 z;mL21?Wg}iqY{^baPalDX-A1J7dKs7<)T?b+u%Xn&@^T}eh)MIALh9odLDA@?cZZxCpz$|m(C#%WKKr8{M^URCQ zWPnpm8nDB6GIpTlg^DYKH|h{uB%4Z_uP&X^B4EUHp#jxr{aL_C%HO3+QWykd zPBm%|NGN!vusKZLted133kPqFxN%7`SHX_TQOhUQj#l7V=(-cdj2A9#gnWnKDQmZJib6RSjM#Oh8c^$R?>fXT84ZKqOd5 zFo?SAoS7<_{n1CS3wPuP`Hm238Yp-b+~EG)w3~%qU!n-*7ZMCF&v5Q93@tE(_X0jZ z%9xS@qr{6~l9}xewZ|T%5fjZjZ7!)7W#tClcI8LTk*p9wsZ%*YT*=tm)3{v8&X9Jg z4TPmBw0=4Cxz63xi0>QSI^j$IwU6M&MhrlFKp(;wOkZ~}IMgM`aP}oRzq>GeX&Kn0 zx)(p8{=wF?cFu;o6T}c$>})^NT!+Woe%y@C#ZZcqmL(JD9e|J3-86*_`$pc!mq$-0 zEQT#x+HS}RnS;6Dgt~&-%N2>H8CPFu?k_&KgJ8&nJchIJZZZq;qn)eg0ZCVoHh&q% zfJG0@c*ePBYH}}_>R^TUfpR@RW@MwUqX0ksSCJT(J-yS^ciT9v=J?fNL1+6<@!Q7~ z&s5$F`^^dYLmCZAvF!Aa>{L49@@S^{J7U?7uDf@pyE&mOWbGJ0Zh5lrxvVCP1)NqR z{X{4sFbjm#K9o~6a1z0{;gxi*<4%%+LZ&l zO$^@PX-$8JL{A*-J zcv?Tx);hg&k=sK!z8a$03t7)0chh z4p}!82#J(e33l@a=06MvDaTE?&lCv20*-uN;MEIH7LiIfj-yQX^d~N?o6%$G-RYq? zOrsMif=yY9TwRs| zkK&Bu!R&AL9E$mST7USYeYl!;u%a7s7-U|>G7i8i#GT{(`VK|GqGM_#A zi+|BSf0rc*hb(h?Y`~LDk^Xgg-hI(J=Jj5>L#w-sH1gbjiES9m1Q(rjIuqs>4<~T_71@R(d5YyBObSIlgdt6-mI4X`QFR<-P|O1^*N5! zFY+Co^>FD5#8PKPvQI`E^SG>>*ago4u9=TmoD(BV&qKsH{%#wtRWvSiE4j9o=sQx* zv<+m=QD1?8A4FH1P5sQyYvf!z6#_wCXZc-Ya$VE@3XuX)?HZ2Ezw0(r*^%!MC^1IT zbAGRIIJJ%5GVB?17kLHShrB|5H%}lzuiHPo1WtbRy2&^5Oo8o(wT?SWU++S)gT;_z zizk{vLH%0fbRc!O7pYQzwf~*Tm%UR=kJ@dDX?tJtpw2gOtZ#cF$Wv5K$s62Y3^dn`o z@5c8m-xJd(A4Di5i^PR$w#~RrbxLJzbK6)j?XMX=T4@#=CcAb*>6#-v+G%?%nb%xt z@g&2U{>saJ`h}$i8iD;vhmNsHf_^HlxDc*#rAOsfJKAX{%2oLio?by*XIvgVW3o5* zS{%2_;-z}b{ucb*GCtO?hdd0kUwmaV1NJ6Lmm?D+(B_Hd-e}nAQ8XMl45@}r^d4j1 z3Li#$zOli8Xj7Nd_X;5$9H<@)xh=7{@5Tqd9$Kx0c00TyJmO=w0;{5!X1Iacj&K=e zT-$c<2SpX*R{iLWLu2_0pPA!&y-G&&&yE%1Km9Hc|1t$6mWsg#k_%|FhxaQIr3Kk` zG*qKEri9#QO*ED2Doo}V8%x|)5op~%yg^am-#LL}+#aIZfwBT_85koL<^U`=A4j1b z3S8{kmSO%293Ic|-%AVD#_DCcw5oTY*)t8)0{pQm?C|Txp=HBSoQW}BHDp1 z_d*7MhVk(+SIe#~iktX7hEtlp|K7~Wx^t!#&1l5sZ6vT$TVi@6OUv04y9(7m2VRu7 zMcqb=-mdf>DrB-B$9Ke`ueqWMRkKsNN{;S`_TD8&S0??YsKs`KEy_UccnAz@i&FW_ zs%`hImqpC!;$Ekx_gFpiI?;b$3`1NrKQ!DqOp`^<0HPxeZ$+bFzt;F_n{P=78XV_d!^ui;@+OAbZ;oi@{DgIB~2;ttQXCAb}o zEX8-!Rj-AA6|_8fV1Vm*K3sTq?9%#4zc~zx_>4&cdcEV0RSM>EtVfQRFPNP4-o>_= z$=o_27v#X}+t;<)=XY0B)62P{Hggn8ayDt)R3XN7 z|MQXbqr$#XOkWrrUS3}?X5Hj7gJ#dAZ1pye9d(7Q=HuaOi`iZq;20KEhY-MMwPXI@ z(}Ln(k~|%nUgY~;g{11sF{Agx+AcPz`D=38#%en(h%yr?K8F;=uN91fKm34ILi9WS zbiDYNT#USp#S5*%PlYH}HN`xpmr@!MJeM3*GBL;gg`CwUCV>qbWDRiVwwm6m zWLBv35pX$s+k-WbH!f*Uo^PSKoRXQ^?8drw*#~b0locxX)3pF`@apqoFA!KR!t6G8q`RG5px8)rN_tw%0b|fsin$Hf>cmlhqT)#840{f@;~IjoxTC z^!MP_6?RP9J|_l}#xaNYdEeQ|=9UuO zRpS4qYgjdEQ_uAc2kE5c#a$FCUDUlqQ9zoRM^Ck(XT+0Y;53tmFX~jB1t+LmOjY$U ztKGlb`(DJyHnZ6n3|SYRR&_3c%q_7|&QUUz*LTUmUXI@hASRimXicLXHY$7Z)#f zyDyY4&KA2bj=KL~8M#r-^}8(7T(E7+qdh6yrkd=V=4FSuOt}MGC9R&FIGxYTeu1^* zhGTcZ&i6x#?D&fTQ+$g&e2d+(i&O3klfGAC*8VCh=U71f?ZUkXl-%s}@D9^MIZish zCpC?Ie?K!D99*Q6we#~kR&{y0R9~2V_fWIjtle+i7h?^KTgkDjf^@t1>B2V;UolcE z-?vYC)2#xAXmMkkk?~4JIffIF)^q_F*WERYC*U9CpRCRP1bPed3*?`fJ{Nu5>iViKe$htR5<}T%fDHS=5>hi3;`^mqy zdeamue`GFZakXNN66NdFvg=;X)wzR3ooOgWO5x*j#C25#x2Zy)WC=Sk-CIrojZ>=*1@!X@13@G7sjs9>ib$ znq;nBSEs%9tB}bA@2*Fgr`dP!d*GS&{s(qiluP=tjMbq>{9I2|TyTzFO&<@XJ6w5g z7Wh0_xqr*i_Km)!Dfk9#ngzT3fpmwPK670@xK&qnb(AuXax1^~8W<{aZM$wV8K~5N z-5h%Gl$R%b#@-7>w^)o^7Ylhu zPv`vO<%`Sge9VHItP*wkCQK}=7ntbh*YWB;H_91w%I%-tY~3B-y;#7|6+DO;KlP^h zm~{W#p~prSUWe#OOb$*9JMx@Yj=HOK=us6qCsbymUCWJbM#&fBPx`HUI|_1FS6Vdl z+W-5fQErF(5NiMSu)H_&Ho=#!)t0ZH-xO*f*VNbL|IMZGd)mB~eQKjf&EKiFIqBlJ z&pw&7$JAHbVxL@kf0jM%&B?jW4BehN#?y&mhi?3+eSxpPiX9shZd(G`PDK>SC z;?+mBndJXTd|_)n%`GkGEv27{^Oy86ND`WUz>_m)h*!n2q(_|K6q}s`6bc zOO>9;l*R$IC^yH>yG%tE%4-6+QZ_y`O=-^hu+D3@&y${mwK`40!>3VsVOr8owgVjp zG9oH6_lINu&Ev6Cs#JH$xoxM@Gg-aR;WW6}nY7fr0+>^;g~Ty3aiQ;_W^QtQ*Y<-S_N^Q4lwOR{X^4Afudi zn_iRJ?ua+K9Jy5%ZSA$OylidFEPvyl^9B7~DBhJ9=@E3|tJ9cS6k=7bh9yWe>P4b`nv#+$NxR#Rclk;`|F4%Z~Jbf^yKg0Gt5b?gOV{xRy!4rx(|d|*v(uP%{nwF#rsu&ox=>)Ty2#Z z%4YVpddea-E32pf`#0~IqTl#|eQwrb-gAd@lAaA|1qviKMT~hm-Hzxr@0}nz(PA(do#-S=lu?5*TB1hu=v@$^ z3`XzW=q;nS-1*-7_+3TFW*E(yz)^lwmQ%+w$jUgnN+o`r} zb=hx@Q^}a15IG4hx0s`&bzRp=b2s7r!w!geX&|^g)aN`gtvrWH{<;*S!P5;H!00R=smw{cyF8Ll$of|OkgP~*h>@l_%6+g$9=Q|W@Tc3-n(H;5dterUh{ zv&2Lk!yKxuO(uGos%A?Q>326#MIuES)9kg$3jIR=A1%PP+}?(PpTD!~v(6&CD-A6g za;H8-s#W9A`P@#@w|nQo>uK+~*MH+QwmAYyCbJ`I&6<@(;RE$_*&2&+QxBJ!a|$WF z#kbvepk~j9Dl>1`w`Ur*dkq#c8W)?d`03XUJRrCmDNK6O-a^BuD)&;04tixcd^^XJ zC}!NSSt=)+m;IEU??|sZd=8fLqY%4<-K8 z1lDerAp7n2-QJ9I%@=>$0D|4o;y!OpKD=54u#VmGFzbmQhdM|%kc8MjW?_+A&rKxa zPP<~C?BXC%5(s_#%i3&Y8+h8B!v+JqG?GSEad>H{&85tcm z(zCtAbhXi^l(xR!+e?_3c8exjOP;kGB@x1q-=h2WjYY|=d#8mdB&3VYqj{M_$2g+) z3KB~=HSpvLFmfys)RDz}9`e|?!mzS`25E5R>9SyzaBE=UJvj_}Dd(x3b9V19MHAT; z*BO}_cLf&PW9W3Lkca&G>G0D5yowh*dTmm$>(ui}uCV}CYpWCLl53heEYQ{;5i5!r2zM!MBgwye;`g=S}UGvu{6)d$+gkWoH%} z^!Sk%|LuD;PB3b|MR&=*hfl?Z^AOM|$grH2HKa$V;qD-EV8xwHmy$7DdRE!>@bu98 zY?vK$&9&%9UUtLGs->Ci?n{g0<=&EGcMM}@diKlbvdWGW+CTiEP+v0nM=uV!`h(Ax zDR9G9!7msudf3oT%wpE%qOB~9t z=}Eg5(SD31*ZeY~gt`j+^K`r5>WrFUJ*D4QMZp4#Fd{-Q5lpuEdRizwoxmY$3u$(z zuIi2xYXA5576#)|pK6#|0!|QkhP$VvkvZvjTb>kP9z3DR5z5ON+6cX*i||= zVags#;Fsaky4c8O(qp@a)t`pd|$Y-E7kG@)}tLdb}Hkk;7t2H z*zTo8VJkgc1Ni}9WbqO}D3Zee=U8O2f4`_6R5b5Vfzki>d!n|?L&Xz_28+1C#$CHC zBE;d7$`@>HNAJGi+n3UTG}vR<#tCqmJI5}bCQ#_XHI10bz^^f2y@h&+?mYMAh0CM` z=F7lW>OAJ=L~DhsSwg0FhI7k^XADi7d!J$T=6CZm>)C7`l)OKvrk4>v)G^t|Mo>b0 zkK^@TWwP2b4PNh3@Ch2Fy5`^aga3Z#{h;(G>qGbFrAi`yq@WMPs!T~all7pu#Jxtp zzX`Rju9aJF*62ZoY#B<2sXP`<(T`BEG3~7L_r(7&r^0<*^EdG)`)B9k>BC!vHk?Y><;Nllj8%v_?SAyJqsZEI6a{?=xL(X#OyyW`F` z-PV4b5nV(4E{wYfOgU%Hhj(J33$E@~r>!kHcTtI7`w@nEc~%SOXfzScH=U7g~YYX-Wc#46#qPq2Kop+6;oWpz?tRny=gG zC12fl&Emh4;Lw9Lu6(|%MK#zYfdjyFOEdyQyL)HJ3sNbS$rGuizwYnL-kCpv#|8%g zu%0@kWN7hoE)1$sZc~7a$nbbCpF?=ejPbA-^;%?Qb%T<9N5ikMU<`pmcjaOglrtNl zsJkFu8PyGEeHjRE%KOn`tL62zKZkqkQ;&ch?N!|w&>=N#u}(i1p|%fY2IW?yFZG64 z^%N+nE*U*Mrd$K8DY>FeB)WWl-5b`GBp7vhc~`ELVl9a*>AVm>XlakB(Xa;|?%1AR z@CeQ9v2~QF)}_?>O5zHa5?>l^7!`#-yi*$S{=+R-B@k1k5@!V= zfaU=*|MZ8m?g<>^_fw>NsT6ubbZWoU-*xQS?qL_j@$7zwSy(mxM*D(}FJF7==JB&s zqC#Nb)M8J@d0n{Cet`2|v>RXRMFRh@#ENUvA6gerep#dGo{{7|m>FhGzLTR?L}b$b zv|vzX<*I1i>UOr_iuYjfF+NIwO+TW1(cY_5+R()Fx2YMz#mmZA>o7@pIRtd)Q&J8|ud={*k zU<`wdH^eE*7ggD`;(h_zHa#KarJ+;%#GbqTytIzCZO zy;`F=PU!)E^0d3W7j*dRnM;a;*SpU-B;R81Fd42QY!Vb>&F&oE9N08y@@g%G`Tsc8 z<`i8gT8&IEjGF|EN2P<`N6tMjehhy^cZ>v8e%)noc%qdK2{n$Rq(rdue1(_T`gI89 z2MS)FP_LRUm-KOvsENhZmZ1FMuRs0}XMGx@u#NReX8QM8pNx{$)b&dtWMF_jk)~^> zNvXI`eemP^F~Qwbr>I9U_)7LGQ_}v%?>tH2+XnJRAcalWv1Li6C$hpS%LLFvZ9b9x zo$rr~YYVE7*-SRznlQ!u?#c?~cO;)1?2tfRY9_~HJ5tuGc7W_L;Nrl?)1LuGBR?$q z$hLOeo!zQuDq36#io%-734&_xOIhUWM_Rk{rM6N!H+q`y5NghmPJvUzwfdp}6YW2&mgT@$!I zyCpa4CeazLoA0jFaV?^uUIB4lR~LJIK4jY(mdiq=cC9(xk_b$P75za|fjM+pgySS7 z&{qco-gs>_CcNvj6=16z-BExCi2q@4@>^R1E@RIg_;QPuUB zm;~I8G_|qr1@O5im#|)?#6Y!2=?|4#uMH? z%j(({N`n<|)d`f+xG+ZFPZbpXGT2;wYqv^A+wL8XF&Y_g*oW0cvkSCCQgr8ot!#MY zLCfs@2~P2AJ^{DxgZrjZw?<{=WSoU@!J9L4-?EJ{4-Y@5dl-50NuSWFd@Q-=h@aHb zr8&IZx;gl={&XY0k^XN4AG_>V->;rajdLbfh~YvI!@1#hZehY?FQ*;brKR)7Tyl2{ z?QCxGRZ3MzGHhG@#?f52l^gjtCP_)!M;J2cVqUfes`a!3?saeGirKlJynYaOk3-Yj z>I%3l;NI3xS5iU`)73e^q3%G*Vn`%`8#47LLNJnrnXO}k)!334IR30V`O zAlMV|RrYIPq+gm7tOIlIsV+%mP+;AW>U%K~?&P44DcU0=-@RAceOjV(@rWzx!)~q^dHwA40lgAB zuT3xT3HNM#UWHr(^}8myr$&T8jNn-#20}+PZ@V6T+v7E%>xr}r{D>ftcrb07FuyEY3E;j*5&3 zS#2xG*F%ibocHs;{3SkivH4@oL=#I6O@4{c`x%3yeSyW|Gbhn{Qa+^POv7t*|`~ZE%>)_j?4^ig4@dr-z0T--VRUjasCg>+2~i6 z{Z%OyiXe$#;Tq@)y%+K0w>(ton#838UP=J!K0O)bcjLdN>imSF`93EsICesZP$&D| zJ=r{D9z(7=CU4(<`YC{OIdvxiiCVk5ARqq1qVrROqYa1t_UU$&VAZ{+`A=vEY)1W6 z@($k#_Qc)@i`3I;vmK=Ys>kt2B4wRg z*H%DLZ9vf;*B>czLBbvHF-F#LkDS@)C1dgrja>r=Ok3lD>6r~RN80C5>gI8d1SO@Z z;rsM}c}@rCus4n^WG4=+mP|A3Mb26&!GFitPRY%_o&`I$Z?_u>fkFc)qOBcqlZ<$# z;#gSg>xNG@AlrxX$2XTRt)qUO-qJw6yXJjAnbJ1g!KaNgdUE?YYdXVyce&Z{+JMk?9R7{e-oCdJwpa97414Y>nH7ld6%ENlPg?8V5FRw z7?RvTh0MZe5Qz-*U1C#zGpMRVb}=%BDDL9qJ(T`>E#k`2=jnB*TuJ~*CX#MgNxAic zHjQXER?x39xZvnGIo%}avadYKX5yAoWX2F8#aQwKWEFKX8-c~8f8ih3O$8J-(*Fn( zPkv1K09`E|+WL!ebK~9*RO&X>$!om@p4ct#>jV6Ir~^*k`hzAu;C#)izyg6FOCv&1OZuSW;js3fpS+0{;)vO{c-l_c`sa4BNG=KXEU;WY?yn zrdZQm=sNuDWtnW=yiMNV*T5?lF4iPld*{kEz=VH-O1YpICYJkAR2mnj>?`N=~3y8Rt)6llC``e>veBR7)X+AJuD|&KMl^ z{B0j~Gp2k~YA)m>NU2uC9)g;l$UB$tGhF0aAI#Cy0_8ug$ZhmKOw#QVv>8wft%{e-jH@ZNO37FsJ_S=NGnff>^gsmMo7np zLjlmnF6XMG%UV8tFHchBd{A4#063=;HS$x-_uT-@5k($XwhIVpS~3k6-IJH(P|oErq=y;vR&?(-)^A$AuCfUV!D-4)zDh8PN;>Q|mh=7N0681O@rTSAG+y#& zH|S-A0zPUzB`hODr+9bXaqSP9`Gb|J?pDP}^ixqEMbg5r=lfMWlOnx$MyGvf#XV@R z42^vpqXbVHWOvd%uc#dPld-@NsyP`Ffvr>g8()yj9?0L$IM~dGS5`J8wS!P^^ch=z zKqDCEYnz9p5)YrWwN4M)hS_qs^ z?ACHn48>jOB~%;VrM+-}Z?du!a`El*fsQncG}$6E>vmv%Vk&bXo$XK&8C{?9(6;?| zcMO?!cDkq4ZIQ{oGJ5i8W4>G{@WBbn>zfQi<}M7JJHCIUqMWE0Zt%I;{IwXa{{Z*2 z;Mr@9E!78V_nA0!)_U5tXNPV|>xW5?TqbJRzhWUPx87x$Rd9iqz{_YmldI(rkkjsP zxcyZW^5OcKEb|3OS1?wkJ0JxofP?b=!*W+h>QMEg94uZ=hvCEhh}SiZFZeCBH)=lp zwOSo(&6Ux_5o^gGNbO*r;}PM$yPjrF)zja}(&?FH$Fg~W`qUpaLiXNF{C zNENZtKDoV}MicceXx}*IcBufB=$rTS!NP)BVz+u)l%cmC&>^44Ze{Nsu{WgWJJ)TK zq9W^65_AzXb+(ZD)%3OEV%e`^PE=xxKRfnUK8w4ZF2(@yp2x@RJF|Jz_~4@6+!3>f zsk^66%c8PKmmciBZEd^nf(knIwj^%Sgl3qUPTg#XgXjj!5bA-Lqxm6jYy<~_dm&`E zj#j5VE31iMn^NjDRc_bE;?O5{+3MUv2=!a;5G{|(wfq+iWtprtBXa{!1CZ>MJ|-qa zYvS`Z`iUINgzBp;^+DwKKWJRN95@rQCkQ-F<~G#f$W+$yxJ%mM0wd7DRyL+D{ut!F zlp6uFWJ{KZtBefWSZ7nbInRd8V(8`cW-}WmOY^fpd*DsYH-|k|&#PD~@9bLhOT=y) zc4zHK$Lr=JnPrnwfN8hg4~nIbws#{$>SQO}?*Y^9bFgKr@-S9~gOxkuSN+S#LiK!# zRD6LuQE?&-N9A(xhC(kSOaR;?@>AudCub*^K}l7*g=iyQCF z*~lz0R-I%Wd=m2CI2qFCWiV5r-W=~iA-&>VIXV@ER}HxK%5e)B$6X0&E_{YN9wPU3kC5Co*lh)tLo;aMAvxvGsmG zKPo*VQApR*ujM9eOKB$ib)Y_@*%z#Yr*l&*l{Z;=nnjXVrtyTNzE4&=1z)Rb{{15J z!O0&d5B1IK@VW{q0w`UINTmwM0LajAlZi;Bt1uc8!!2HTx8aD%FOLKm1hS8|voFvp zlT~a9reXu_pb06Cb@AD;y}gJ1h=Hn0wfwhVGBOI$RGA5g(ZB9%42XA2qP?fLLrokDDdyU@4Dqe0X-%SwoS1nUuW-mRYYDh2Aiwl_6v z;kv%cHXTaTEcI7_?P%%X_v9i2`qcOJ2yKcZjgLcl%-;OW07CIDf?^(YH^mz6IeeJ& zQe6wi`sE$7X}t?0J++$pgh}qBa@WWmMi(n^x|aAb=e+mLcZQ&xY)K_?Sh0Sjw1Nfj z%|S;q-PKlCe?o{t)hW3|Xrk4Og4;v*P}|_}Pqhk>sG;hsw`L1sR!B^MeO{Cw1SZy^3h3rlGA%-6E@hatx-PEm)-Dz`_Q0ELYX&f`uOV{%?% zg!MO4p?W)?Ti<7*o!=-jd$DC)?0`Td?%#eaas|Afwnb#A0}`CC3W{YSH()D{3AEf`PGa zfp7wgPiwfsGYh|~nj<8hJvgrL;`?>Sr$8-m3A`wU8Z~&lPa=1riqH(xp@l??E&s5%x)f?UfhVRXGp z+S0-t5RRDfS4n`AcKO`21%-Phc7ln)S57GcD=GE3ZR83q=+K0b_NCY~()^X7fc}IS zHYtHc5QU^OKHpCd0Vs;~h3JLux9bng(YB!qs)%jAn>wU3^$I%$`uH5e|Le@EX5^cq zL$9{7)|*p%5_~wINeuV2bE1CdNzBSvlVbm->CL#-K^#@o=41Avo#e%LjGLE*@A|&2 zhTp+vE~rYsylK&ZyH96x=J#7laD4fKJq&a;PSqD$T(lLKKaTU zZ+L`zB0npg2ZgCq%RZ8>P}$&ll`>kruLe?K*@W8= z%(cE@jImx9xuJy73vaw-`waW}n^$M|)UMOHb2`Yw+Bspe9F4ngTd@{5^QZ+M1s={? zFHC8jAGPgvW+sV6ml!VE0c1E~gT@=AVk*Zf`s7xK&%MgY>Jfc%2@Gj{wr91{k%=$L z+Rn6MlY>#eY`W4*h6w}KdBDuVg7yz00isZWM!HWXlM9MhkJW)(oOl;OW zweQTB^HqxwuJ3Jfcdhe_{*?NVsTBbQU?Ix;3(3(b+u$76Ohj4Vi1rK7tCPNpm?L45 zRL4rnTXN7{(k9|lLWe*KjIBt&P-s|CuQ|p?1@;2i$USpTns=D`}Q;QO*7i)X~+2 zQk@yU#C?O!>!e;}I^NtayDY$~PfCzqd3VW@c(W;@ZT1^G`Axa--Z@7M;~j0NqPZwu z@`ym@vzu?O0Q1w)3#OD$N4%6e^eQFaf#(v<%_cT9lShp#$f>E@D+~8T*7$+=ya~kC zj>S})cy%ZWrEBkr5|Me~TK?;INww8PSc2CDo37!2=Jj!oc{x3_fY{;JLitw{otSgK9;1~4xKXF6)HfwEyW>mX6THB+M~fJNB~h*S zeqwlJSYiT)!EEA%(Q=L<3QyXZb^L@KMht}+g2kR*Q-H+AX?9@&%$OqK7vZ-2yvS?D zOT%Gyj({?(YiebO>qn2MJk*us-(YawC~p1;rsr<9o>ciCEghpOri9QozCjk*e}!`Y zFJ4fyg^&O4Z=V++AM5zHiZEB%`nJ+6C%0}{BG|97hQ{m(!#z3)8qj#LQ~#oMz4y$r zY|88&RP_SgV@s^wq+;ar!IYPOg08V;pmpPbPLf$hnu^HxsV#wTIfiATYP{JMM#se` z35sJYLocOWmHv?{O0l{d@MV}c$hB2`F2ftu@UAeeTczh+Iid+l33es8^GdUY4R=g+4l(s6#yS~j;);~G$HX<)T#B ziBxG`6f-x!b!z>%}a&IkWJ7uc%c$#S_yQE4N(4%9ad!{m{y!0VDa{BU;XeMEHVoIU|qq~IcEA~Pp+@!CBr-JwS7hN{mF-v2w$8?8F z&%zeoQ2h7R3=cKHteNC-D}i*No0r8o$`RY(RVMgW)5lIeZ6UneReKuh>nDtEQU;1L z-l4R%fHSgaqzXU#uX@BB&Ty)Ku*6v(CF8w=20cRQZ5LoKo)uhX zAlccHl*J@$QZk#YB*hk4zP!hC=VfPWK%fKPJpLQ3>h73C=S%r3AKPV~J#S(58a*}& zFX_dNZTxV8z#H1b(uxLWyUSJQqTw9w*#Aml!~C!rg>^H`snxvl3VkgT(Us$t%gZtv zlxZW4ja86rsoCYXxh=gm@q;1N>UrDF_hgiV2<(Yojl%ndV^P4K)3( z+RP_(gC}!l8$YWO^49g(xXO~D9yLYh5*WH+JJBW}+};lPDdtDFP<9xup}DZ&F0oIi z_A&sqN&BZ%+oj_M z63RPZ2>df`7r<#MJ8`Ak1zpsl8IR1iUie1@$}n|o_&m6++PkUA<|-F#CGDodRs5af zAB`>+qje#Y4Ob;Yl+5q2g#jYBlErD(#8Do9=mMz-%l$qMByerTIsRa2tO5r9C_p&9 zA3_v(`aL_OedhSq0vd7hXk@1+$Cx~QYdXHJ*nsh#=|A&cWy1JYFwsu1LEeYSE$Bx?wl%e9`*AXs=OJknzo=MO0H>>+dJp%s3*fp6pmqJG?S*;0*qNS{bWO}5ka`ka;t|k3pQOoh_9tcaF=dGnp%avo0w%10GmzNXP z_e%dL9!_CZNL%?^W#B5UlD%XM4O)R^3ZqlKWO<^3bdAUPO6uu1lvNP8EoAcJkSIvk z08Dz$Bl2d{ae`tuAEYV>`S|QcC*q$8i-LXWsbX6;+0vazE8nvVTPZ(deD}V-*QB zZA#6YrgUx%q@%#3Ot>SkH+WwATcRzDy39iLcsAl-F`7aNOQLYT9O1jhrlhw;nZAN; zHhDY(>5Q>|{{#S~DIyg23I3!fG|iIoQMK>mQ;{`8qjfrLX(`NLV%qY0u|&k$T7;9c z%so^CUd7haic)h*V9aVQ`uz@)(zP8|gjC-Ak&yX(=7ai9djM;QV(n(#(!%8CXRS3N zWAeO-XqaeL=A6(RTBKJ;kM1AnDgBRk=windEyj=rQSpf;#Fy``fkd>%$82l( zW90e!!12FkS?6=ile~*D+JAiht;&-7eMh@#GHk{qPR66jm+JDbk`>(3eqNaH;~4b4 z!$2zwfxq>`4x4bYid1Bv8h=OKD~~2CP3BOP(c8uqlYhVF zDN!ds^Q>Jo!7r|aO!jKNe{S;yVRHw#Zd@|1U5stT#bya}Fgra&pQR>K#}^@8c_jTC zzr^uXvQ2t?lFl$@Nu$SRNov+U_^GCypVi~FD0RQA27ez{umFjwNjd*lRYM{HtA4FF z_KM4jW8o&lPI1X9DCK_!adAUML6@kP+j6FGfsyDDf~oSRzzi*o^$74^&TgYa?HXX z$3+TfxF!unjmZO&D}gM{mN#qw#=bXMss!KRYT6TFA(Zv{Sw(@>d{mm`vj-1rYr%I- z6&Z9p*7(0nWnS&iCiYNH>rO3wtKkmR&|bz`^3GmVBU4~dBS^?kh7sCj)5Zw}4EG$j z*%4j3E10WcsvCa`pP}s=iLV0{`Fr`wUI^b)JH@7#4g9=G{-$mGDbIQv4c{%bZ#zjU zBiQ#NZ;?LYw-(X=z$7I^rVgKAO>`uB)MneqyVjHAmH_3H+>qY0`X>N8Nq2II%>bC~FcYU2mqTg4t_<`*`55iv4BJ9DyG@q!Jr_Bu&YYePTzzsr44Nbv zK!DaAG~!A^xVtv1j3A9#^h<$qN87wnBwaOztct$S@>Yl~g-ScaArFkCWfot$pX60z zOj}2;{N|coDLAhp*q%$>r4{znf(Xi%{J^3vdz&R#Yv%G5AjeD2E-yl%xSa!8apb64AzzpdEmW&FQ3QMdQ#w zGSsx2U}?Ko=i@XIRJ6vmZ^3+pdCrgj>;nHV=p+0A87#)BNXM3d>iA?i;4Ao*AVKrg z&P(A5+_OloD7I7HWDPJqn|>{3k-0k9iNN$g2DU>El6gXin4MCpy*_xlOey`n%woIMs+amjH(=k6uEGT5xs2$YijZ!3AN?v* z^&UW*mv$g0|F4O#q9?M-J>SATlRB{UA%60mm39CK`2DvMC@Kkd{_6^Bt`dm$@wAS( zxd81&c<{c*V?|PZkStBogo-Cl>D+}9`)S9XTM-gt)5V*v;Jwf+cta)~c5W!-rAJSW zy_T%K&gCh*pfBVryta@~&Z|1ng!PN?i13~CJj>YrXPCjP4OQu!uW(sXGNPadjg2XD zLc+|nE-CX#ASNLFgMvnvO>R481Uz!#oyHx~2ZU+vzgPOtcfyt87hYjSA(2)=XWB7I z3pfnn&St-XJwDaz_@}(gNjN+)-+T=u7Nv5SC4L*OCn|uYE4fgD#X>kiSA$pc`WN2| z7dg>fPP{LQ?CZ4$ZYuhrG-Sjv?Rr%h?Gx3+;rxYfCG*qLxqqePT^_v0@O{?g2-|8xl|EaH9eHPv@`BVe>c1H*bV$U?L)zKxyDx7G=kv;%<-H_| zF=GQ>D>TC8w~R!FLxz7}-YYTNeGAN1ViivHv(?Z&o8mbPUep?%CiN=#>J134Git3u zZdhvLuUsM4KFFU1ab&aSzN>7kn&CtgcL+etBh-ZvTdZ020UqEVa3gGcknRpapyS%C-?n<{}ec7}6!3R`eOsJ{L)AX=BNTeclK*n+MM z4G9oUWqdqCo8~?F`i~&Rzr3UnCffK!bigikR^e@gV=mNFxPaE5_zcfNTWC(tUb2C^ zKynwV6Bk`gEBfVN7^?@tE$Y5iy<0vBpqC#KI&IagC22feZq?8(q`r+rhDevZC@=y} zJiJ(H%GMY9#%rq<1R%X-GP`wH_k*$ymZ){BX%jt(RsE8fAGk-qG(DkHdW~%?^w2~d zj_&!2e*bBpYnMQ_U@7Lv_*PGeO5m$n&KvJ(@f?ax55TkFGq!t{|HQwHm~YA43xQ3~ zHQ`v)u8rT)uDRc{)8->up0FI6cUK2IMP(|Hu|Q$3G?-Ih>c+okr^@9Ij2ikiY)rme){g;n=V+dd(1t&t z;hGauK1_pz`luT8Sg?UJFgz(sxfA{^>)V8}m*Dp-hZk1B!hBGT!EsLDtn1FfyZPs~0LxriJY)%bh;^}(Oq-=`~x_F(Zc z2jL9H3FW#WWJ z=X)}^+q%YF>67tM(91TwfCF|>H*gg^*N%HyP5UBUI|?VV3{lqT;IM!ToX22P*ZbF| zE5)*^0XK0qoUx{Uc+yB1%ixID5j6?$r}Lj)m{P(Du*Y2|QXownys){`n*pW}+RPDx zArl2vY%0gkX&^MRum$y!QE(;FrtYmh?!seYFf2%S3m06@BK%R=Wo$7<+9b=V%9r+~ zjm(fqwoOvJTpr*p&x?xP>&+=KsUKCq!Z91oC?vy9`ega)wEE9U9N$>PF|s3TxBAP@ z$8;b|#q+~9hRgl~Z((g5m4XEvn~KLWvkzvrgA)FG#Q*Z-(T`lTg!H2mInUhU(cLm! zXn&9<_Vvy4q|;d9`TJ67!WmE?kfj6aZ-OJXNdFLbG1pyvzur>H0qV;S_$?Iy~ zm)oqy)2|O>1D$Z$9EW7?JM^(y&bR5D*_?|UL1-YNqPIWoFg04*Tq#T)$W%mDih_md z(D(kjw07atfv*u*X;DOTWV1Yqvv_+C$kCmQcd0taS~5v2-I`$Y{J0Stb2!lfV*z&G zlC-O0V3ek%TLz=`7hD` zax+@i$0uXWJ!^N;Mj$AfcqtycK>-Scw=U5Z<^V4bpNZ5`6)jwKOG2bq?Vi32r(g*47} zjPnu~O_|;mPtm8-jnM7r8d-^c)|6}RQ*8r_15HR1=XeJ$or5=pr9x%EkoB31A}y9&@to;NJ}?J zOTP#2N1x}v*8BPW@UD9y>dcwBu50(+zrD{1QB#q-Nk~PAgM)KZLH?Nr4h|s<2M2HB z##P`Q`jhfa;2#1mMz;s1)()AOnYXvMPft#dkB?SXRu&c(TwPsfW@b!GOm2F8{dM@u z%E~ISbuqPbVSoSN^XJd8v9X6IXZxo+cy6Cpc2CximO&s88ynk?x(Q`v<+-h+gVnvs zT`UTP+8kXIjzhnB^Cpuk?vqMxp=8?r(fQ-Yk1-j2y1KeYC+95hJEo?l>n+P#T3R9^ zA|^^ll9H0DjF9cGKaY%z{5(7zMGlFGi1hdO*Vfijy>2x#GwWMCSV1p*U;CAnm8F_C z=-}XRa&|G_GPAHh^{RH7I_Kh>M_twT&8V+)g_xh89jp4{J9Bswe;#j0Nl87=A3+Gh zzfS(_iD+4xTf$GdnA@G~m^(PyIl@W0$nIHkDW5Q^JNEJM$;-=o_Uu_hYOQG5MOPR4 zIbu*kLgMiF%({5AX68ar&CIv!Vzh85YT#n>=ecdk==S{2z$;$&OYmhee_Ur@j`IiI# za^PPM{C~}XR@R>*7p?!}fplJTwEn=W&3L8xWBPnM8J0MGUOqM#u-&?a6n#B?x;TBF zy$bqu7NFycf1$a(9k3fEuH%P)ahw?1_e1vWBwfJpwNPn_aIr-c4#$ux1C z@6VR<074E_#QD#FTjy-}T{C?F1$wVnIkKnE3|sGB`+Vk^vGt*awLj;}jST3J;7$G? ztLec$4U5@lbGOhhq;C>j@Csw*z_dljW3(7z)GK&=_M+>;=UR3-wD@^$f%koTiT(J2 zt#iAY^qM293}`v?8)5`bKAY^f=XXg`yssIdkyiuWzvjtra=%y$(DJ*dwFNsLk(i$O z5puD!RQNe)RlR(tE8iJ&3d#>b3VK>dkz@nk7rcUkVs9?jz;RK`DCljiD0mISIq@DUmkAg&f0XVfeUym#YZdZi>QZLB@zFh*n z{R3EC1>bO+P2)Ea;g_Gk+`9#!mss{iSvE$rF7ReBo;Jhr8~VHdvj_T)BRB~lQd2HN zSLD>2Os(9Ez!U%Y%}djJ^tb+e?vl<~nBsmIQoj+Ya8<|;jAVEasb6*_gBiul*k6g; z2xENh1zrKT*4Y4eZniW*#}vF<=Kx6kzts3$a%2U`hWEM&e(%%w|2fL#7K6GjOlq-S z;G}N=JHBPE`TC2fk?O|YSq_&-9NX&ptYd`7!^!&b+^==|QqTyaKtQUA&7Hzxe>C80 zIuzMhhNRyE>OGb9)*=oo1qYW#`S3REk~~T zxF2GMLOUv?;8eI~=f$COJn1g~zVpt8(8#evU8I3Z3z9xD*5R<$U&_LDMhduocQ*(L zd`%h=uPaL-zByy76cUQm4=q{}8tPKtZ7$sUG{d58ilfT~*mk;B1Ux?Ax-PP*8J}M3 zjP?z^mM>+m&N;0bB_0LxeA4Gd)ok~Z44y39##I}~D^aF`Fn?lLARM!an@{-)F1i2N zM-sx$cnnivSkXR@mO!uCc&No*8khtc2NM@^+YyX(Fw4YkbmmE}A;kepq8T!98;{3Q zFO^7*6hqdM-^6nT%o4fBej$;;oKFCYl-eKxHu(L29TJ%o2}(d~e!o(j{Prp1)eA{t zqL{Hy;na5AUbC=+EOMuIX2&5IWa6W!Q24*@n%H5XB5wEFgG%SG98awuju=IKV-GGa$6^|Jhmca?PKg9Pxnl z|Jkb8z*-^?mlhKjC~>v!|9v@JT~Z8ZOle0qU>Cm$VvEQ9y{^XX%Vk73jsBclO%tg@ z)o(3v9h}S1{|bf0Rf_&gbZQ`y;8MXQzW}EXGw0ml`FqLY_qqtOYnNxbMWl=Kn+d=R zOkq5}D^>p`t-KjFZeKKV0BiVX>%1SpS!&=KfM6ui4=KYM!v9_ZU_!PIN-yc}n#}MD_K$VfEm5mN9cvo7N73fc`@rbcsvr zfr8_AE~;5b2ymf@7~EVQiu1%N8%O_556^(~5Gs>jTeI$zDf#7b<(=kOP7?S=ur!*s-fPTv zi}J@pH8jMA-C$@4l`>2^FPht6LfJh8!V3$Nb0SVuRoKdaDtwEvd;2#>zIP4Ccd;1=Cl>z1K(&M)%{Iaa zW21VNF89gP4fas3+Tt6J`fT|qt<#3RIpbn;g^)y(?n;IG`;b5>zp!54kqj8f>I z>w21Ygg9Hpl=sDicFp2z2w@4VT>82>e==b$aUgfrzX@Y+FE$F_d_@6mHnu?DXNbR* z$dpl)^VGZ>H`^f5X4L~utM2Tl`J@V%k#Bw?_|q%6^z9^RV*S55 z<9?|#cQGQQ?~$!%4BR(JxY&HXi|%j2B=Eu5dZSx?^KOs$xY3hCyy^6cL$#mZP~R;j z-`^p_tgpigt?QEJdHk5hxAl+pDVj)NL}oIV=)D@ytjXbtm6 zLyCXOk~qa1fIV?YqGevxjYWbUa2ug(A5c-98N%9&YQp7aUv0jhsomNd)x6Hz%Qag5 z=qrbNIkE|U-Fo7ts}zEFh}nyx9v?=IWT}k^{Hx;>uK^u@lut>>%cqlNAJ>Mptpjbc zMQ^V; z$U;R`p3|+lkQ-|_S$WGCL+U@~rK~A2hD4*nS8tv$?d?yRN1~bR*=&Gicry}rTGQ5) zC4a*nq!{8r0%^2OU)F*_awcB_uQ8rc6Y^+%6zuV7;xOffu= zh2wG^CM`b;t(&k=GYT|_nemE0@3Nl|taptg5*8*&@+8C3gJv&GGa z_IAC$Rs0C1GA=K-M&0bC^+6sNwwTBu)eabP`6aS>4a`CUrY|tHPNe8FzG(}28rU47cKWi9yYhK3WR1O3#q=_JBYGx# zcv^FV9$?i+CXAcuUJr$Q0-QA2D^avki_*Mi5FiHKQ5C*x&8Jt5Aq z(r-uJ2s$i!n*&VM&Vt?#9;!bP!WbqjXdV{7H`CX3k=An>+Ta%zj7q&(mlx7lbpV=- zq)3)GM)%yOzjZd0&nCR8d!XXFuAbd&#podw)D#2{VwXQ20Q9%0$^AN@x*^h#*A;7S z`iS&`#Ka{VsDg_?{>lRx7rPW0w3^D~17(->H0iXEWb3;H`re=|Wxz-apFswgiAzk4 z``JJizO{J{!&uCH;ZFZl#i17<r6-Oo}KZ+qR){BOBXbzID`H6U>URrrod3nJ|;9 z;4q!I(?S4~`Mtl3&0!{ad_AXz7i9$@0>7kYyX!cuHbC; zDz2&Lv2RC|;S1SN1{1S!n0Od`WasrKJ32deLKo98_pq-?mNq)3l-iMSs0sS91$uHm ztFs0!RDgJu?&W3^(JDWdph`5D`OJ{HZ)yn}*RW5sL+3fVNVMfbkw2Q}$VC0(zpB8C z*)xXN1*_zL)d?2NNeY6`R9cRdoJfBM{}`>TNuI0FPfaM@XFh!m$@JeYJ>zMK+VLX^C+!wOT>_6K)|y=SslY!hmeC11Hlxe+P5 z{p-X?(>4&^msK>mw=L<*c|GX$E9;6IHt-3i2)9QiG6nU5Q8{L+D-8PM4KZhBI(`i!P{EE-2(`SmDoU(R zZBCL5L^_h2PGL2_xE2KfyLz5vIKJdFKjcR87|+Ajr52{=h(JG%1yY4>8sMMnx(gguIK-Bdy;^w;3uyZl8P}h?@?5cRE6#{;sYB`flr^wa|aU8UR1K461>gzD_Y!NauT6Jg`l%+%I1VY<;FsEGk zvA__`kxHhHo z&vdhuV{R-BVToO-G8y^v+w0)McgeDIJF-CcTR}JNmfMQB4R04()ZR6wE$>1_aFW?u z|44#hBaqh25NB6p=K<5zi>J>DNLgL4>XxescG@xl5|e}^DR9}~5}%A!$bS12wMtuZ zJmy!%EUcjmGQrsGC8%{2VP|DXdqFf5*P={5#wTxQ&~<(>D!t-x$K(cwY;^oGFNIe< z`2zWP$Q>B-wSw>hmE7Qt?QiFcR_%7C@ghZ^y zwA?Yz1&scR?$DEoMeCfPI=E818N;#N54O$mxYX;VF?^$0&KcQCyI(RG9qb`@4xTZ7 z-(HgJhnPZMR@uv%B59tn2p2wgH#OB;W3GrfJ{cUT8HU1$YYu&Gf%&c&PN z!M+#7@%=eZ1?!DGw|l5!s~)1lzW(aiPQ{0)))&j&$x{&HLy5g-*`d!h%xxC0?|U`R zXzT%W;Y5v90)62sk^9j!$bECU?db4$%5+3Z2Uvk_3oUp{^ppNPndviRo9zC?kU-5P zF&N(B=?Dv9GdUx)qx5yW+&Y=ft|}L>n+DqRn|q#7K10n`6|wFx$-`TkQXXy|cBA^4 zlQr;60c6Bj9pev81x*{<8pgO|5?BH|cam-vPQJe1@6JkWFkHbDBKhpW(;+_|)J%BE z=$^pN3X#*b@*U;ZdyCoY=5izSY>ia)Y_cMjN-!Zw$Yj#G<_F(QpTK-KBOG;EiX@qb z)P$Ao-cFO|HnXc?{V!Z+@Akb_T(z+`j$w0WZdcT(*>zN}tQ1w3K!$(6AY#5q-v(xFNgNMUX%VLf5 zqO+E+%K}3=XKkZ8X;|q~8#zNC%P>nH7VprNrsVBRB2!#2QDPi1$Dc?miB^z?Y#=Fp z0#?ls-@hEgQJ-^=*DOK#)@k=pH7gBhn=4lKPnGXqeIA*7e=ByWrH16=9cS-RtIfq< z_U8{hIH|2fRrOV$YdjD-e-a3Xg|@YK3Eqm0Zj!INuO5qnzv${L2J+P_;GjoWoe5pG zop0L|jUAy7Z4tqT%P6Z_hRg`s>U+x>X?ufukqRKScV*Tg)Dw(A$C*kB+z)fM;WQ23 z<#xlvaB7?Ilsx-y)^-!@r#qyk)00{S@7@KM=d(SY8bJ}YcADLXEbnJi2nE$gvhg#; z;QRo;3->}HYy@uI#q3dS(GDq}w3k`3)ym46JLiaCln?mQ<&27zm^@fF9>Ur|1TzMh z2y>^8q7(&j>Q)p3*DQ zH)1Obfsl%ke$u+C#kz^;eah6GZLasho;|kh{H2O1WaY!rw-!aUp*fPtmx(`LdK9rT zbb?=Z#tBn;Rc}wF>^dfmq*k)y!gkqxl2?|n-m)IgXtZOouwqyfJVpr7yqpsA(m2dd&4VtcUVk?B z!{Y~9ZHlW|8^=hFOuF&`F za~gQA|Lh!2i6L7K%QD#B@*klULn8d1X86&2BZwBdoDxs&ua;&tQqp#3L0z>7 z65`N);P?lOnpzsvWDp504f9^0liD#oM@&dHO3(6wAaqpC<-tZC&K0XOAp`u-DURl-~&{uRUM;73)s~X*sroHW8 zJnFu5$KsJmKG=F+GIii(xGMz;%A~djO(%~dq23|Sr|f?vHgB^@Lqh`8W0#F$i$Y$i?WIE(+1nORQn^m-@nhp*bNEeDa|o^%w=Msr`eV{*a~8F~a? z8x&cdunnQKE(^zfG6x%8kgw}gN$>*ET_sZ%4^6W#n9AacJT7Z5%PY*c9=ZDwvrz}} z8_G7Me!L{oL$WKl@5wI{QOA89h0yM8oCP}h7S~{PYte^Uzmo$s7c$u4EnghALeIk( z)y#a#i0uvQXR_qrh52S19x0JKvV0plyma5k8H4yI68sXO?w!xaBA@gPe1&cdet+qg z(InS?M_(72{xTik&a-wOrCGp zlxsEA#eKhR12hU`6p%4smhN?>s0n#@gd;FRBZ)qE8`n7ZyBRdKV6N1GP6c)re&Q#!>@6-MsnnTVwiI5U(BJzX8C=30wgDhPi{sb$lMX zlTd3hL1XT9{l)| zyZ$aY6z%?A3<|m@X#USWk`R~&0QAz>VT^jFgcQfb+IL}TvQeCk-kD5x^}Pw5k#z~7 z?~Q+tVh|IV>3z_mAmuB8Cf<9_KGuH%Lqwvczp{*ec;Qg(2=*8`eCNu%bRSHiGnKUc zB4hCb3QPie_D^y7&qNY1lXL)7%+BgA9*yebePV-+IrSJpS#+a;odx7MFpaS?_Hdl= z&(VN`T>~%hsrgyN1;MYtBMffpmhX-4NfqnZ+yIKXTqQJaFaTlxCv5_hgDpM)2u7r^ zO*+gqUYp9G2b0OY)ky0XyK@h$CwHFgVQtxN3Lk6zSA?XPIQE9)F4hS6%wU}vf~yAZ zH)U3@ia;Usl|kNzE1^*YF!GWWuf99~sXH&{E#iFCfn5l}9xwY_urycLfLX5Uq9Vwx znOBc*;lirof@%E#egQuBG8g~~t9TFKd|U=X06r!+^f;NQS>K&6*d=9HPT!^xIU&lT zV&hjr#wUpu^9nV8{C{YdBLe%$4aqSKzsr?)`-uQn#l-Bs_gkHKN)}mrxOcvOV2-uc zSAS9h{n7xJT!(Gi$;#*UB9`n)h$;kn2WG6hzZvY1sFj<)?t5|Y`!IoUjr~gu^#cH~ zwS2P9cW_yne3FpboH|33k9{{fdK6%=!Orh%Y1wUgJl8z5>hBu}U;2Qjbl&lwuLH9= z6*X`%7~;h)15NMFOyvws-hz4#ziPMN!sd67vV+llb~6(Dnz>JcJE^YM;_tRcGXT)f zrmI;(FlP&2+&xqKp0_S8IR92(P;Io`I{CAofYmswMauL01{ks_vi_n3Szt5?8*n%3 zKhs8RZqxuXq3W8aq9fsAIbPsq7m>Fcf`JZ^T14z<;W@7B~a78Cy zq2L1F=dC3RWedWbz)>V6#`S8eFX}pIFzZrGS$qHO4p0GeGSFCL0Y52Gat&N`Qtu>$ z6ScTKv(Bm|n*kM{rnepgtPSWDEwW-ZmECy_Ym`Uefe?>aiR&{I%!weu|iW1FBPhR zmy#7J|94&}^HR|J`E0u|9Y2>!;BQaz4Cvcgm*tp${5_z`t9XEkA>{zSsh%HhCBYq1 z-Ro5KThf4=RRQMU>hqso5%susthN<(IW9=Eir4-%`0w}?hVRm+W9#9(8BtOeZ;*mC zr$VZ7xR){jf{Dka(aUD6f%=h67lXP0%8YEf!Ls%1k0O@O07(f67E3%*WiF_%JgqGW zS;Q)v-m!(e`5O?E02&sbsDmcN3$av{l|6oBz=1)D*OFs1{xmNz7Qljn&oI0>`Zf*S zQ!*^H;O3!R(1}B7+S&R5Uo36q?cg6z|J&Jm0@yGoKKL*p+hX^JV?ir=AZBr({)xAg z3Qe&HM&v`eP(u_nvnIE??g8LUIO?GORp5Q5^z2a*_y=Y%*ABeRCnpxq-eV@M# zs)y7Hn3N{}-E1qbJMl(ytWG8~ zV12@tY1TQQKn&BJ?fbr^Xr;Ehyet4?6SL}KJ>X3~h!Yr9uVGsnx(-8e+)A_MoKLX{ zkbI|zVu2qh>uWMV%JAD|OCsF4XS+5ZM_mp|mJ@*2BA3nJzUJ}|5(&hQ zm%aDm#mfiEUorZmgDrkVqX%-rH+p|Lmc`7$WRJ%6`15?~`^VZ*O+4&4;9_I}*K@05 zxnZ8F)CcIc0Ryr9_~BPm4yF({x{XkXJU%LaPQXug8LdS>X-X$Q|fQY z8H9uB$%0yqZne5uBmJ3+$LsDd?Hq?CAEXo1Q`$8+=2Jd+Sx!$`o5^3Iq>8p2VSL_5 zNg74qs4ut04X<~Fieew?mjhW34kp|fQhV|7!IPDxi<;qGau1ewa3z-HZaI$RlltvL z(sg?~6b-k-Cii#KJ2q=;Yc`U8Fa5kMj(%#D2@8)J>FY-s)VXcRI@Td1shPi)>Ag4P zs&@1+J>cAG7kv~vcl6ok5e1koab)E1DtbGW&U`fGc%^WdogU(P7XBkD(Ipo7#vUng{F>@ZPv2+Jrlr zCQB8iGO5&s^S?U!#2iF<71CsW**qWGzk?2&dZ`B_8MygLHSlu7P;Kg_8^c3I;7I2L z+6XK4A>!15mQZKDv${#%l4NiF{Fil{^OlKMQ@cC3+6s7m7zZnA)@tFF_8l=ydg}K3 zQA>Rp1-0K7zSB4uM|m(23>(!na2wvY4-h+$Ue0_IY=b&{B?;-P0hdak^*D8yW&diR zB$^phM0C`8#tnIEW7JWndr0NoQv(miTMOoI{AeakDckcS&jNkJ@}8Y=7Ghgx&PbU$ z;B^)B_t~wD&~06dWo5?2?#lr;?^wfX4C0sr?U&6`Rf6EF2jJm81cj>WyKPzzH@h`X z5#bMqZ73%iaO~Kk|FxuuxgbA~--ogS&kuFgrJvocJa>(|uo$v-8zrSJ*XgK81F79mqkC2ntD_`;0S(Ms+V;CZ|mk7I9Y8- zEu}B|MgFeI@#7y}8F`$nPM9p^M>?9Io9PB2)Fr#o#l7{;U@-L-*=pH}!!6I3Kw}qh zRCcD2tLi(00#n;ErI3-@jiofC{s@MHdhm%~RYkokJLMZ9!xO_`K}fpNN|d!x%e_?A zUgG&GW8Rx?e7wqEN(KZ#@A#PNKJH<8`z{2Z_)-+gdEd{8tj=f;OS}O-o9hI{U!z!f z1#i7qzq-{YOrL1#$;@WM^SYjE_(HUdx|1V6$?EG^&L0m9MDx)%@V4f?{9div;9Ij@ z)uX?Nr7Xlla+#rTJ}}p14qC3v*GbP!R#39%G<`xAn?6#3F=*h68nUdzRDWw;0SY0~ z;pCzj{DDXZtf7ggA5}V_H?C)O%tD^ChAcp3@&-7h$&0sH-A+XHaBVzZ&v){AsE(E4 z?IG{&GO%m*t=)318O@Adv_(g80!y!h>egVzBTM*?Ai} zOCMHQCr%iC7~)nAGpd+KP*JIl*4$i*2ZgY^s6rtmA!mUix17~UH$45o*!(^6mG%|d zl2YqPid2=`^R@6nrwaA*x!xesP6hQL;VkQ@zLco$K40RwX~G}ue91cJ#@n(M*J3_? z9bCKx6Dsvg|BpO=SEZK_BX3lg+^9%c9Wvl13}FO`l~ri|p3H6ZY-N!xEE(JJ7CHqO z4#>KPGG~B^G&>IsP;o^Kl%%Vvwj20xQb+O4&zF^))X&sN2?ZEhdQ&g&?haT8bMR8;y@j9GG>EJ(=ZAHBgA| z0S|DzwJvWKszQW5ak}DL?t>xLx2fF&@d(*UuH=iAKpbF6VMQQpTY*TMsxE5BoTb)`OKyr@q@7EE$f*5^#prUwNy z2s|0VRy2y+(F|>a#Igu>jUgUl>cO(Ngqb=S9E-R4CwEcqoQXAJoKs8o*eA zG-l5jT^Fdw2(CoHp8Gz!LP3;Zpp27<6FAkNK{fJk8#fT{l~1Y6!Ckii9V+^W5B9r; z#j)f$UV9xv>63D_f5_;;uq7&vAgTn_)E8fw0+0_v3CJ4E-p%dGk(39pki^59xCaloryWW(Zfx!?lh722Q zNib@tsS0ujyPU}KP26yesAnE4%?X3YJ=JY`f`}3ALMQX6`<{+eC5&8cF(2oHRtx_l5#bU< zV4N|ZXiL;vD?yUU1FTjH?|WALqWeBW6)iUX(L4t0Ft`KP+QU_uEV{YSZ50cdbS5&R z%D6)N#R~Bh<57@X19Dq`^>Wz6#jS2=$V(8de_t20qX5$PyN0xlfDbm@Hv#Ffm?y%mcHw1d*LeuzfI39~ux^=ovC=cq z%Q2U^Dhx8TcEi<0F)BucNDuj^$at04U6#6+r;wsE*Z&xXFYImx>3lvZl&(vhqTsI- z>bh|PBmANtf z&d7&GQbxV!_kvrJ5=f7~aLa)&da5knzUd=w|@~e#ItNtnmS~6E6e$$#pp)^)1oxc6Y3HTPBf< z#ga&<8YpE_K=Av7h(O%+i1C)0Z|950!zQ8=u@i?(j*?T@YjH^FC`;it@O_e)&_A;P zFKzCwJ?(8WOcdezPGdfn9RZHjswye6D^~J9{;6afGlvx~eXyy$%XbTZaBibU&C671 z)YNp^pnur0yh1RYHCRv;z2#_UxfPn^$vc+q;Im`&X!nIAMD?By%YQq6K*68S#(DeA z@TjYYd%^3QFrdM+Am^pxUBL>MMZGj05hZlPdRXyPMteJhH#(PHLPCvzdNfo5vfIr- zq~XOQ7`k^TlfSYfUp7#=96)w|3naX*```uhh?Q@hLah1L0sI3iV%+AydX^Zrn)F>@ zb29gda07<~Y4RgKG5emV%`zV&5o%%Y_-KuRAmARpxF@#C07UY8m}Jzb{#L!CwrTxt zMaKG2AiM;3pZ1I`jK?I`#(YF?Tcdh3zd4C07@j8sG4>ho?VtG2pQ6}OTKn{v5P-JX z@xdp@Kl#+d;g5HJ{y^* zTp7u0MF|dBI{@Q#0Rcj~7tB=oOa~#^EK((JFzVw721cj7xC+v+I6#YV;C=>F_t<;x zYklaJOd&q_mbr#ktr*3K2POHuA7znB=-;&i;rf`0NDF+Ddk-6X;rQT^+Pd$ylfDlw zdcsihuH(0cGz#PjxYUiyEA&|-njB~Brq|$31UZ5a{J`U|5o=3J-EvJ5_W&NchWcL$ zMZh#}o>K!OD7tC9Xag^G<80o%TmNLEePLb+wWNCqrx6FL@_6iJCcaaId0c1?zM^|W z|0_`IQ(99qJS42lHvhecs37=iVkvU;O=#v1r z*dYuclBpKHIVvXVPphpZ8@B!GWb7^q8fAKapk4BxiVC<_7U=%0!Qk6-NxdKRPk{hi zAv`1wCx(wPG{?~)#>Pf@>*VCDrLg4pepPJqz64du6g33`Zi*QX9Mo7@S^2i9=?&1S zZ$URQsnQW5FDAo~qa*RsS$zLna{{A#({lS-lNOExFB{^JM@527Tl!RY_uI*mQqYnQ zO)HdihiLYo>$wOO$NmT)mkE(XpMMTzSJPxs3{B890$vDfL>65keFLA_s78)EION$y z^nR-@m;|?DvDLSPEqT$!>T4Ln96LV301B}T$9bn zqwcrc>Y_Mrc+%Dqery7F=u=CV?i0}{;_#QfqioLKhdA5GrVPvCn2B!jyHi~44;qbo z2B3_@fmz3}s&jndje;B==W+(?Rsg4YgB)FX_#&W$P?-=myyRfi#8T~w#1@)?VEU1_evl4Ms%@xx0uJ#CZtgf*J{Z^$V?>*NgBiP>%Y~h@yO3$BK z*zx!s-23#1i)WM1#9t@VcjA}uCv^kg>qjfyH;ON35`mEfIj?C~CdxN$Rm!DVUz<8i zd1=Nl0fS4bG73rGSwp#Xqf^Pmpt|Dr-8MZbEkwG_72(56k$j(hPI^j5%xmDhrPpxt zM{Q~!__N?KUj@^%pAwK#JyDimrXV2ws4;&h@{~a}n^o?uZt_YYg%V|jwZ5I(2?5v} zzqp}izV;r){F7m`J2kgLw83~nHb(P71{4PdRRewU zdnoShbT(SKV$+98ri-yxbobf+k%o39Z4Q4E7_uP#|p)s@e z%#m^>Q<*3Fc;G%Ha>ev9_F@wce6|&||B&qC4NTGW69Fr|p_rM588KkWB>Ifs;Vn_T zCliEj@qU6vFNS8lJL1|S2VM&{%HpbDLUFuX)p8I)i8onjLDMSwT$x2YJYh`Ai0XJ` zX!e{NbxG?A+gxAX!GhoQf*TeNT9+LXeN4EK+?-5Z;vVNZ0NgMg=dbW=H>#YNGbC#6 zrG0>AprAuBvXHe3xIDkO;yE~}0J1b-aS8nnzM8Ul@nB4k?Fm7?wQcjsprgN=o4fE1 z|I)tU2j-vDP;~>c3myX+B})7TU-5Ph%=Lm`YW^7{HW>^h&xF>7anl>c6c z9RqMrauO)@`DB$B%c+OxQDYuOq$GqWXbR?>+6mv83lqS)lCzuN{n0@zhIrx%tj@vw=b?)bX9or9n3Fu_4Zg1>Bm`GeQP7G3Sj>_B5D>(ctSbFqglqRLK zVEXsN5RB0NO1D;I^7HXK%0+!ad~Sw##USC2E8R7GY&ZZ z3U}Y})Z1^Cp?yq>vox}%iG#fnKVxL~hZJwA-h`{ZlgW%OE?}h|deQA*iq`I23FiAS z?GbF{Vrw^%?(r2OXfr$YNxh7!2kW||@vHuf;CZ|@;(!Ss=sv4UK&Gv^Pmi}AO8>4B z>seW!Q3_vfdQ-3XxXV;SY*$)2*1HXhv5y{qO{p<2r(LZ{ek;hfmswy|X(7jpdSY?E zu!SdLlQ88m@a&0`*KHm(iz$zC2y8TW8^uo>tjCUrQHXRNU(b_`C3miU5)`7Pr?)ok zq*k~^GcxiGDEmS;Rv94Jh3xWK3|?Ssvlk}l`Q8=-bwXtv7~DUu*jQRykw)1Gtoh5T ze>=vFLz5);DPsxW8gh~-`_40joCSOJ;<6+I<>dEauAH$KyD;SPqR4D2T97DR|L0AM z{Cc6lu?x`V`#PAr(w7EbvT|Rp0ph_V+{0a}ghU=*QjE=%&|IT5Bw}TUZ!P9H^T0B` zHC?4?+|zq+obN2cYmR*SV>@aQd4z4=Nh?KZ!lyK~NU$m{K1^^eDYHZYBUx>cvtSrU)5ZhP{m z2b)q}dIe0eanG^2?i0zNJrZPd(F!?7CV=)eI|E`}R1`yUPxjL}9cW?jNc=m8aFZLM zvraAZAXjSDC%V1U^UqZ*wZ-2X^YSv=hF?p2J+(*d>XA#lo!WHyxRUrgsKwiBDK0Mg zfKq?y*9&GEGR^3=(+3r$JU-{qLw~ za5HjZ+~yYG9a%pLUpCv%u~BmL%E}^p`0oX1oDI&N=1)qZqa^@?`)5eOL9>Vqap~fU zBs07D?>cvOF-{7CP|(@#0=p%r?5Den<&-(q$&KxM=RFB&CbKu7?l99H4ktaAUc$GF zd*NrW4sNPeyBj+@ryC3nyf)vKz{%s|yW?YXyE~_+rw6A%CYj4C;O%7y^hOY5Z?K-L zDc);TDDtg{jSVE95=HTvXQp7Rb}9o@N7s0+fd;0-x3vEenh88EVNcykLiK{iO?T<% z)a3*a@Xky;e=@w{Equ?_d-?7Q@PPj}9|CLN4qN5|ZI!?uO9Zwlk)9lZ&3TbxsjxxtXE zu}aV?9o=(1@)$u?&C@j)Q&fSX<{t^43wdYYMy=X(-${1E`hF1TfjwI6f$Bf{=xU)v zWOR=RhNb*DY~tY*yGXZ^O|-WA%h`Mf3jH|t;FT?Q#2?Q*ygMyEp)+4+!rAd5Nj!&FxBTu z=Woi)=KB27hp~H`3&B6vr+$(`=FL`XwimAxwfby?ZjZ zpanB^!ur);IZ=`h@}>$b#Wvq}4HGfm8nS+4t7z#h8xpZ^H4$`(vQvaw&w11|gV*Krns4x7PnDv+Sv!N# z<_XaQ_x-b0QVyXjy*9xmebXOIqBbuY6zL<}2IiP93<6lL7!}K9bn8!FJduUOmLHU1 z7J|QhJeBdZ7K--1>08%F&00dga$YDHCZY?n=T5vIT20UwI?g(Num{|VBNq&X`Xm;1 zdxkb}h0sWdt7h4DZ5N%A7@oa+=LDAfg5B#%e0)dRAMf{InqSA;$mW|euL&d}+1kd7 zp=9uZi$jMTdn$ch&5f#)JI2B3tT}BrTUWP_7;`H*UY6-O4${Ss=Ax?~sT)R0jJ%wF zudWGS%$5;<$9YWdch`xMdsJv~Gkpcfb}oO|u>EiK4Jot_6)m1!^>)*pZlUvvQPaFGlnyWb zE@fq+>C!(Bv-fc~{Po(}A(ElD$Hi-OZPR*DA!hj?T^b!{II=RU_B@;ugUBarr|`wN z%7v1A{8iqDRaP5HsZiNpAqVw$&O3+&z4J_pI~VNO^%K*5x?=>nsS;`VHe?WCgdW~@ zUG{ONg9>=3WF|R+bl(2HsY()3U@C#m08)};*s+JgaG2Wj7`370JHfh@4(7bJ!ZECC zutKlpvG2KP#j06BItb?e>!Kx=7N&s3&z)+qEH|**HvPL!a^14Ser%ZV$n18~lKs`( zk7(uT&K3);Lv7X0l;j9n!}o2klq2I5tUY|2?9yjn@;K#>4ity;Z1AO}nN83SiE#0r z3k;56)|UP}Q)d6tbeRvihM8qw1NS%`L6uc|;utMCzkKwn=oAP9Sv=4;@6OCF#Sz8C z*=G#i(Ym<%MT=vwb!%3!fBlT+j0uv@COCq2R#+5P{lW>4V2x)V&RMP8S*A)$Nu=k-#D>>1=fie8Ek{sSOWA=G>e7j5h+f_( z?40=A&W!BiGYrQ6RiHSsGcnEyjJ9w7xUIy~mlKmWeN49$Q=!JvPq}4Voh2{|yXLKV z)%!|g5+`UX#|gY4qSR?Kp9+skX6nlcrz0r&Sdg7s6@T;)W9!V(=p^Tlm}~<{_d9&d zlmMRXbys)TS8jKcNfLcW60*e~bYdTT(EBN4RW_r2T9~n)@^$Zxy-cjv(y~Ns2)ulC zzcoR>s^tG+@6E%Z?*9MrQYdT6l0A_nYnY_$LiVl5p0ZAs>^n&)L|MYvCc-e5vhPy~ zS%xskUSaGe#*$^s?+o|n{=C1x>-+il_s{QhUH5g}*L|DwI?r>?^E}UUeLM?JUV6N6 zrQE|iL@6FO(8dyRJ{?`Dfzip)$;>h9FToO8V4{XngEK}!_o*1z=Z)7cIG}akX0txb z)@1!bX+CwIX>IUUPl;!}JkqzUKwtLwAwV=_}uK=%#a>6=rn1G-5i%cj3NhnM+srk_I^V zvL`z?)thmRu)`);HuWOy-BgaqTa7OSE(o)hZdqPwb6LsyW>)lKz5FD|%27gVaEF|R zI#3T9)Cy;;ER1S!oBcQ{Q};qd9%gd592gp_XZZJ=4{@cseyzjKQg^0Ms83TsC!mxl zdIoL$CC_%g{Zx;$vP|jP+mtw_(Jk_a>ed1RNqPci_R-|%!bf$baWd(n-XAyRU2xE) zw7>-E)#%#A@p>55D~2~Uq%FH#%~z4>oKv5jSs3$o*!39+n62t5KVnhrTg;TODtdWu z4t37)2zCh_6GObs?R#J-WssP@Ly!y?jmXiw-vt7L_Bxj8#*N6f>)avzW73l=Dln6s z1qpENJ#pUQ3upN|v^%b^{tP|U;LNPw@zKJ}eIWR@`0sIc$^NFv-W%EJ$?Y{;=V8i5 zzKuon1@q876X)9XsxYjJzq60;Rw<3AmU!Ud`W;aKL58V{oey(%g}?yqV7+kYdnQ z;W*Sgv^gJ%lg#G&aWQpRQ7_C>*2yy-?vRgIRJwrCvC3#wjS48sxg zW5_=PbT)~Ti2CAlX@NIi#|N!**NWyDGX`F^5gi;Hz^p~JyJ+-$l6!V-kiY!Z+V7xt z*{{a;eZ82Rcmt3O3sF=drj_=+tnNk*F6VE@S-`7tc_&Kcl}~k_)p0bsn#BKFpjimQ zyDxjQE=3FmiZ3`XB{f1eGc{{*=ypv?(uVjF%1MPe)6_s50=LJ zrputM6`!(3Zgc4JvDU+Gme!4ga-qd9o7kL(%Ii4iWKW0g`wq#O2&8CE{=z^$%Qhij z@zUysYbxnp6Xoo_eK(2KVx|%1>z`0q?gK$z+*}pBQB_j(oY8SX-nM@X(J*LZ(kwTK zghhAXY}Tk@tQV^@Nb;geoXmDf(54+p5H;(jS=+;gI7lHTGMhVvXP^`XBl)@lvJ(>M z)$JSPjj%Dz;p{wF#kQ?Fc`JD{soF+aNfJy#RG(XRP>IS@2XZ&-9g9mYa}w`ON&4tO ze!Fc{5m4z19;Ed)mhnBBI{mO@goRq@>&zY$y1zyk^iPn!ZPBBdaFxDrCtY zS2XJAJN^hZ@--JPDT9RhXqy__imYE_`eY7w7|uv8(X}}tuVME5cP`dTz98M>j$Zfd zP_hyGc;4-<&<1;)2&nVFW>e7rJ($bq{uND`b&U0^cqUGXf&Qv>jrvY@62V<_IZ^10 z5n)T~yZ$a-3tjp@_lWx?Y_vwS3*1&a#fk4<(>cLimsdUy1ztCcPxt^Zm>oMT~1 zjGKjz>=pDECl{{4HD@xYNhFN*NARqbnyNLeqaw!!Lk8W>dXQg+w z85Vv&F*ys9J7v|9h4hdzoic^QeV7iFi_3cwhkyU+Io#n=?Tfu}q~zb(VmbGZi=jM8 ztClg@V~7j2h1(is1uqkKvCHnYkZv>046XCeQ z#tC&Y?5`Zcj1yY$pi%`x<1dl__~A7*JA`=WnN1eidz}*c_FiwX|DGxybY!gPvoOaBfi%Bdj-!iI5_yeIm zdT@0owWYSA>)RowP5~Vuw0J|zRF-xjpJmI-?}R~R#n`0PV--7V>IQ1gr6AcdY)Ksq zU({zyamV?E!mrMGDuYTW2q!GENU}AX16eP^?>Qm|%dJmu!g%QF0 z1dx6`hhTlNn;NHFKKo3V5`q@=3VYB3gE?p!vtG!Ngs_Hl0VL8b`mduU%n*VN_dUkv zcpye1!u=_30<+(NDW&f{H&_J+q?MY?+FiK$yOxS>*0ZIJgZb^e$h~j*3m-La3JSu6 zJJ-J6L2yA{<(b%1OSV2M_JR~<<-~$=2Jnney%N~`c|?h&p+)7uGLi#B)APdXomRTt zR1barOy5x5z)J-}Hq1NWyrBYOYyerZK5VrIckto;KIsPxysNzR%nL|XmNO<)q5aE5L~GcWPFS2T{DvafZ<32J>^J4euHZsg>0HPD7~M;cpvbQKvGG-jGnPT z82x4>Yqur^kHByL%lsg5VG!c?soCQ^aZrumbE@^(MP@#I>V`zWjEfI1k~o0l_gwqC zsQiC(mt~HWaxG;<6H0`3q4f)Qyp-I(>ALMlnbZK)5z)!G*2??A3qYJ6sad)JkTMuc zf_??~mY?i2{ulG{f8jELo(5n(N&NBuAEZ%+)?LH@$_4mez}Ejyux1&xOY5nqN&K~&Gd}Dv0XH6=j>9tq>{s2hzW6>r`2F*6cG;Jh^xBUGrgW6R& zT(sGAkclFW{W{(97aexU!?BV*B1E#eeH+w&OjrI}`{D5vHh%%qs#reN60$f@e9$N= zj7%YPhHu#{dwoB%K~y9T;9biKPlI_z2gy}u-Tp0WX&}SM@ElMrbgtgM}lLuUDm6z_u-qp8m8!7 zGiGMY$|CcKmmRoYE4#To$^!oi>FiHKajN8dUxf?zPcJs%0b#~*<^Zm=Or9YSsH1KL=gm_ZLz@x-D`(d<} z=rT9qh}&ckO2dSz0IDhwTLsLO?GhP9s%!E=IEU(K3E6`_8<7@lFf6aMzd80WY%?RFlNWdTiWSDfgH zi*OSgK0niAEqD!IR-urS#7P+aoW)arQJQvc^wO&bf_=8{&fD-T!(Q!rjn?}Goo=+h ze0zNs)FdKf176>OV=pbexvn41j7Wb9LX`p*?`#2I0__+^;q7|c)UopWa>EFq0G_xg zap)wtHjp27xaJ;{5Vk)V##Up|Ogzj!$Uu2J?{{!dThug3hY_1qejX$!P7S^u3fc=@ z90t5dO<_=5N5ko(A+nFkdPVDdCP&Pw1&Z}K6pGxI*q9UO7 zrnq{)4*TW)0PIvG7QCzDH+uP{g)kZ2Vah(ky zW;KNoXY%izz5d32D{%$0iA-G!A)@t>sX9xzFw1HWF!InZvFISAd=r}Y!zC0V5}--H zj!soW}(*+u&3BrUh|doWjCplkW?5Fs>wCB|-NGJl6GQ9M#2lHXL^HEd^f$Md;6pYTh8b9=}i z@6MxFs_w)g%^ZZJ10p~TZSWc$pkDcmO9Q(&PVh-Kqra?;+qm8M9uua73R3ziz}bQW zjUskR=V*WPn-1Pt?nAYp_*QnTVV_aEnq%E4hHj&SPWP5Z@5Om{RC2)+=WE2Ni`fOg z4)UwA!|<+?C0rc5qj%*l+?o9rhCbG|ANj3x{^)isU& z;P>yzZ#T8?>y-?;I_DEj;>#{w4Fh$%7B6B(K`Cw?Z#^H??7IS;k21u8o19Sp2^%Y` z5R0*C%JC}>Q)uV%!-A^5Vf3#ny^+-US0~!cf-5G~-bx}h7N0%E_W#SnnPS0w`ohxO#QnA)za7hD2uX~Wadht*u+H&qf(f+ z#lwc1+UA-BTElV$abbr(hlwl0R$3(kPY>s05=ZifY)}VA6`{mKg3R=HQ!jt=Yw&Wc zM4N&@N0)|fB)X2(DuVTqP}K>+Q(+#m<~c2O^x$nD|gX-4{bw<%lgxy!n~1~Iq9!O@%iwA z6ASv|8)5gx`}6ze^l?|Vb{fQ53er*}e`o}Hj0wV~%g3??IZP$)#u;*Iytk4KdzLTMm+}^jS`z+cOx*hL9Sh*z^p$V95FlYGTfyhy84kR}gGXodI{~w2{X{CCeS` z@z5v-Yu(j6)=msv{%lWMJH+x6V&)sEcoqE(5$wqmpQ>VJ_YV z^)*SvVv|L`0l7n@L!sgOtb+`j={|S&jPRAQr83^g(-2yptAC-V8e}!(wJp)P?n>z0 za2OhPPBM**)4v5518&{(Cg8TPuZs@uS5Mpy(zUK{R7*W$|aOa_G=ziWS z0fKtifENbe;LUDYIU{QS{R|unr%F^n0=0|g*^moNrPltwLACakzGvDYByAAj1V`K( zS*JZxIDgnGr;2Hg*GtEUAFl5ZYn##!(*_y{gZo1{_LufjrzedEk`&UWd%HhzR(_j> zjgh&*2FNK=)X2?6p=@GzBRmfUvJMxzbL{LIYTkA~;dXv3Gt;db3Vm{DSGm6(qzUC; zNsjVE-#vG?5vpZUP+nk_V!8({aO1Xjc?z>f1h(K;$ROQJXFzuku%kCI-tPgT^*l3# ziH7{wVXiNBY5I^5Ya9{26MA44wQ=Wr0_xC(%dqVG&P)Pw`aROx|XUeP*4HSRj&&%$zesO}ACRwI2?Ll-dTN(kk zI;i?13^rWg-JhqMR1_5;yq}?|W84 zxQEi$=M|;O{5R+%7Z2pGYn^|XH~yVQZtTt)B<6STr5-Ixx->I8T75#di||` zygR_6xJQhh_l*0^o$`l~8=W)yk!B_%-Wbax@6L1pl(zHx@5bg|Qr{HfLl(j}Y ztb8{|OVF0hsiGpts_c-iaoYaD?~F^9=m8;sVrBK40% zpD`!L&S99#A9J!_Igh`Fm{;9&;+4p+ZSk0FU6wgc+!ik}$q}B47J@sBQstlyj>;7P zIaokWEl_o{lp9`!<9GW|Z-X)%5~ojKM0U>KN~r$_+v6F`d_Tw3bPztV1VToIVP)21hch_h z1AYS=f?l?_8<)CR<1}$IUu5GVe*f&|u0+wwf_sK;Y&5!2uA~aIV{-o^73$n2(q;I? z1EMX!WKik-vTyR_e;xYoqg`e_Ic-DS+-Vz{CM+i4qZn*L@#sF>T>AW;i&lvAY=UA} zyFN~L;>Es*8uSIV1j6Zjnq2M!+CPs*AGjBGesUPRTzRAO${zHsJ~+i)wwzkfH|if= zH^=8Z0P3jqNzPBu6riRg&V!JeWn{Y z`GIJqQaSGOir&0rrC})Fz(JY-1t;Lu>>}v04dt225QE)YR{$GOVpR3va zzOby|*}Gqj+^BsMZu+XdJe>~#ivtNZq5ED{!en34s9ilNUwSw+OkaBV^(DVgQ<=qf zSjF4hG9Huhg4t^+mspt@-v;Jjot449V5nDu9sX@i1to;JS)qTJ%kBThkY#OPn{>tx zt+Tl|4+%}@rkZyQO?6KNGwdv8nDt~ld%&1z1=)8UE)wLow9#>D`u(usyHoVbD5Tj< zUZBernI_1^+6CnpQM+c_AAf`X zg6LSx@Z|cyI=I{&K%)QW`asoBty<@eO!dR|oz(o7*AynT;@r{e!jF5)5`y>5hW{PIJNxC716#hE-Z8^#AF zTUYdcOC}16eP8j-z*_m$6Mf6OrwcsZxFXbR)~zi6FVYf(6b`+Xp;w2@vc25-l_F^V z=G*9g&^%Y@dBX|sF`C}9+13_&{>n!q){f{{x^t^>gX6}Z|A$L-kPE7fi)5$+893VA z6|S`3UbtgaG;knO9xwce=2fNsl{igF^N#^v%YEAhimf{ShuczK=0Fp9c7*jkMz=e9 z`(BOi>mbb&(hs$hx>~+ov%>DJo!{GB9M0>3#1L{y>L<|qt6>{e@CsSLYZkxosYC+< zudNN!oEmek+Oca@f4cg9E)Kp68D^#|wtUv5{}lIgP#<}C{x4neF5Th6H7l9`!~0*O zna^M8IaK?8v91664LOQ7!Mi(OI};_?6}u+eUp|7^FO%iA0+#3nl63KWPU9(>laS>d z?ujg0=b}#&JrkDQS@U)pTV!_+Ipdz5zGvfNER_;#uyywClY@E``?px9Yf)YPx1u^u zvrWQmzuu)f`z8z9r{>8MFGa{xGt|BM3YY(?@|BbqC>#H8hKYI#(|0JiZ*YPf{G+ZL zUVp2W3xQ1k1LsvHzts^M^VESix0$bjTqX7eKRbIL-<^%B&WXp2eNCf2jRIPBQ;95- z0@|6QOIN_DU;F1GKHV_H?`9byt!23xl}-W&P*G6_H(`!QXtkgWxYlph;^pt?FA?5# z@l@WW$?In~+wwJ4dSI`X)T%sqW2zQct;LfB8wz~-w+P}y>or#xr7^HjJAASzx87gr_$niH~PZb zMeq7sYDo(ZDzu2#hZ~<(KkSF?M90oe6_ad(3E>q>to2RkR3iksB|IN-)TmQp3yYlw zPT&sWE{gbf1V!8lBMuXHI{bGl&G0reR!pS_TUE2SNRH*V!7sBqR58_ceFYq_KRy~rc1>r^9NcWg~=_|?cSKcT`)9?`GnxvHHpk~q&aj5GtaCOtIKGpNb$DFZ?g){9 zBJPWap?ftbq1Y+WXvQ;8I!z}M3R{;tdYAIUc{#f)bznlg4<#BVvpM-*Kc{j z5J3`r-X_!M;oBsNrcOXImORKvO!t71v2Gu`&OIfaFwvxKgQjxF)xnrz`rc_9fh9L( zu0Bzq0;h7^M)|u|_^r($k;SNS-5R)q%vA=*CPqPa2yPdZ<}Obtaqc_|jWKrnG~-!` z%iJIAf*?qD@#kO{DgykgvbYai%gVlD;WIgbkgW+TgUT=aHxOp;{&F<%k(Q7e8jUZn zt<1ShqnJil8QwZcC!90>TdPwawE&VinZhWQ+oqK-S~nS#p0@p~**qY1rZuq5RQblf z?YwYx&qdq6>$1ghUz83be^o2ls14sk5zb;n@G8s6GKf)4p6!GjvMYHGf4{ot24s3k z5$dczI{l&qT~VKW1YJWf&BwA9UuDX3IA_@`iRsJm`amL@`isk)S5xqfEPNY{6ygi+ z=eq=$7pNOJ7?bKCgGfF5AbXN3IS#g^FOSI#f~iUm&GhmEygNZ@{J^Gu)s%ADkKH}tG*R<4h1?K zIE0(f8taRo=cgy8AJ4^K42<7Kju0Xv$#^Tk{cVIo5; zQkD#1Ru4LH?tuoNDKc==`>gVRxOn9JcngcU^v>a-@k$q{%dk(xw(;J{iH2T7875w9 zz1kS%j74S(tiB zw~}J;ku3mKYzCU`Zi=Twos=&>d9`p`jEB@yM7yr!TLqN7OAY;CN(xfQRe}vRKQeIg zACRK-CQH!6L=|+GPAK@FJ-<#rYqtXXAO$JKX2^X!(yWscN$Sl|Zh<@XPbhY>B#Kep z2F-I8Z5xG9>O|~zAzQH7STr@Mk0!hz*`H+PGeWrc%di(L(L7U;Qqli$un8fU18IHn z@KvrUy{NLGCZR zOklg+za9cimk`MP92#;hb=|PCn)Hc7QqLLyAZSkpC`rl17FwU$!w4HxHn4-do2+e) zdXb}_nxh-*_2EZ~v2lzH7yvpQAM2~{q)Txf&+Bx!&56`)g{gt}uY`bmCyfGc5%p2!$_P38fK z=97+Q%_W-ySzG0WyAvbz+$1D4(A1I*3;=mMQ8odz3h!ppX&NlZgGb93gRstMc@ zbI_%VF0%1dJWk9sk_OaNYGG)y{~~tNT443H$ai1>otgukbIaXi8XV$ha#3b_899`c{AU-_gDBX6VsWY1GqE!PU4uI#0-M-wFCS-6GYTW|RfW5}+ zffxq^h15nCZMY~lX zg%x^o((0qW7YpeYHA|e~xE*d}h<{1^0JoXKW($C=`R~yjQApM`4r=H4XUKXZWHq#% z#M_o8JWq}erSD+6Gr@iS{}^MA*?hb#36Rql4NjE%-bI$qPf@x2e7y(lN*AfVK@H|~ z@o+vjfg;NXY#)ltG2P8_Kjv97(zQ&{pWkChdAUP~^UP$so9}ksAD3T~?NiQ6&%aDY`rQh0yuRY?;c;y= zsKNj3*A)5R&^6I)(RN#H_QyD87cJ6zI{U@8^b&d3ZS<;*940d$JFNYfHAhL832KOs zH?^7`rNb{TE31s?`$kB(EJCy0AZ)*bxHyOHEAsqL2ToQsXgBV^rHogud)FFkH>wk8 zDdxtBbS;)c3Wzt2z0j1rffyS8o$!A3udnO}+gpZi-PDOBMyIUr(ab<9r6XjCpUNB8 z=D2bS$;sjT)s4xtUp+=S}h6R_Wi*&1wzQdyYLf zH~pcSsI>+k#Y5fnn*pE}3;<1DCP^Bmyn>PH!(PH&QL9;XYST)w_oF|?9YXKMY=1=uQ=YPb<(dARq zMKLV7#i}C?w3<>4T$TFR!&~^9Y6YxkBOkNKw57-GLB+m#MZCush-~@r^e-q;tK_q5 zy|J}gc+%!Qll&n??ECzd^>{)fThD7su8=@Bh!m(7v?EzN^15d`@alXI%8@!(OGZKa%UBt;Zu<3V~{KOfuIQ39)52XBNee{Y_PdH)S)I z=Y-l9-$wRcVvO)2wHV4nKCk`Fc##cyD47*D@K5^NbrK{o;5i>E2p+jirh?yxioMK1 zWC^J%qo1_Op9;BFBV(U>q9DIj>gDoE*OM?a_3CcV^Ud8$ZpoG_TFd?vw^WkYgv{cQ zMH#<}?!OnSRnnh{AH)DvbS?yN$OMT=4q?ZJKks2ZOKlB} zC5L#h@6z%L)F!phHNuws`SHO~53>#xt*f+5KzBF0d3Zi)LIoXymfJlakct2#x;+O; zqN|}zr_OAq>uXnTZ|p;}F>JmcUaO&rvq)|UY`Zne>?@+mP6^#h}k-w1tx}#)Kf&4uWh^|ApLbG6}u6ns+kdH61KPCo*S@ZdJH|; z&r01Eup1OO0?adxAtxPU4D9m-ddNIiEuuW*>YwaW8&)7td27hjs{?E@wO z;y;!L9*Ag@;|MrY&ShJl}diZA$yLtHP75tfZgxMIMB;YGG|iCfmoc z*Kn-bM~>COXDTN(Wq1l`==}5#-k7-I|9EQNPaS=k%zDL?MOw5mk7Dy@Opjev#wL6_1EDx z(m{Qj^>67QDV%dJmp)QL2fLASn%~eriUw_IK%5s7vY<^m`_g$SdHC8`0s-Yh>NS0My8rfV;P*>&9TZVS2J_X%2 zKd+6BW`B5>1%M4Nd@6BWTUl$laMksqpAnaVnT7Rzzz{c!JL~WlAv>EVb_w$A83F?A zu<@6-&8B}$PBI{k9bS^Sxmn~s)jiUFIG(0}j0ZoKJCy|r31?XsjMSdrd)P}$4=L1A z)pcFcE6_>R*s6=x$>v;cIGkq5xfV##;LJMmfYwiUalobJN`|lV%8%lEUITZ?-kZvs z8Urm@x~QPwJ_`z?>&3?@pQq-9+_)u?0}L8<3Tu8X630{a4Q1{Yr$9-P8M9_|^9jTQ zBCL+KuznS?9ExNM`t*d8K6`Tr)0-y1-yzXdSN;`rd^;mv*8-~$m>8n)IBPSFPiPWA z2zIZxWJnnQcK+1=+2~?mS%HE%f|*u#XOMEMTKAetp44b{z4ODXyl-B1#v6@Lv6Bk| z(FN?tUO9K4GH~*mt77@b4PD4D_D~&#u?lM6tTvN$9QS z8YHo0LMl!FEnx#LsKSmR9^GnENV3F$O-OgFLAmdHMfJ!zkDePUzADb&q-2W!bDy$I ze8&koV47sB+-BJ|`?#Rej{2E?_)+Jf6G2pmHvgr)a71&nq%3^q<>G zl^;EaE7VHcI!O&XYU-rVtl5Nb7~p{cr08lkaTI+^KGJ}PavgpudOupKe0$HA=$GMU z#Jm7?<5+PQ-CS}uV(qN`YNs)2S^RZTEgW>{3SLz-tliQl#jD@@rd~cu#LIQ~n_^?H zSXOYdS;@3Absx@6Dw5tTryQm>giB*+4p=7&&GiQ;ju&iyOmKui=+0h;J9Yi%%}mYk zGKap78c-BwP}!#)ZCA=9NDDC~>^?0t*F$zneJ5S-WHKr+9?+V`*16jXU2arQ4^l!u zr8IsU5dOvwLX{i^lL^K&-J0GQ{B<-DBmF-Ur5@h>=3ZklJet2oddrQ(B_zw$!z&D= zno=30`tK!jb?lI!0W*}Mk6`nSeuL|}gW^AL3p0sY za}14G;Y#>$R<>62JyJ4Z5zEDN(>x_JMEcqeUVB*#3%&yE?qA1mf>h?tE&0il5M=!o zO1Q(lFJ2SpnFOgJ5g3}a1XGL9qnv9rY(ZPa;1hFUJ}h!$mNeJB@@1XTpa|Y~P)gQYDp7-9(UlG!SU$ z&6!bHxw(v5N&pKfLsha|e5346cp?C4(?@4`ptcH0E*jDtM-G=)u*Hv=rNXO{JG1W` zLLM)_?|D>!$NGuW7w7;qcP6-!CW3Y1k0Ji^+_2*7V;)W`9Qu%?vGjQh>u5om0DAwZ0R6HPRZ>K{iRBk*ogzJ-QfmX>y(J-l;8d=$6HhD^7lOaTXNJqx1 zlS7Mc^(3y$IJ`cR1`3nNc*mCcc(kcCrM=&U1HoMQvv^zqA=)^UJ(W7_n>YC5HM4)} zUAiy^AUM27PT|^wz@1arHo&_zvX1?;5R!Wwob!U>=@6{7M=firahC5C4w% zqOwt6Qt=Vt!f}vU>oiFveFghWpRFn}g8j@MN;@^UXI@&9sQWWA^5Ok<=DgTi?pmM6 z-)vpI_U@DzHkxwumuM~}`s`<`qsN>{ugj#h+}CAu(&tNw_`ki~*9%ESN}!zL9N3}~ zHc0D8=&3#A^&wwWBVBN`WY_I4bh$$Y>Ieb4g0nRG>_yi%w%ScTtUY|$VRoxqMv|F@ z#R?fDxLmjvXK&TPm?bI(Z5XX=t2Te+Wq2KOH|jGI8BkOEA$8r_YQwh4#&xaiW?XAv z2fVyMfC3pQXmB1nrMD&SE>+5}PyO3j*^}oH-%}U}aBcwm{n(A%_vF;1p9vc}Wh1fDYiep7Fka*HxB-sU+ECKSD5eRK3*Ws$p)_f!~H$ zaFO`U>vyKaD5AR^8WV()b5-g0D4=x3mlidE%h&+#Ro(?_sGh`nkscx7jU5i&0{5sRPWKow#dl~|ho8z@1f zymN7ar%|?d)~)=ULQNj!B@1A`(niBxRh~+p9u(>BkJC!e?4CXiQU}s_X)M6>UQ^3> zsUhEI*N^%02k9Li0s;hYh*w^h$s=E9Eagm}EwyF9h;nsTX^bRVHSiCf-<7Vng&Q__ z!#@4FG3UQ0v^?-m>hgVO_teEVB7SZ|P$ycPvLYm!f#pAjnBnhX+zo`b0 zL!vKBQL+}3&A}c1{!m4y7p^KW99w28B;o3b2~Okc9WDF4=4jLWYwtdP5oE>5UJcfp zKNl$UEw|t)e2d>TW>LMhDok5yr@=pAct%GOKVo;{b$(~5`Ao5Y-pv(_0e40 zp(cD}cx1%>;@+F|8o?Q8%=}`XLlwL0+L<&h?Z&Me785q_&jzc~_AohnvBVCZL)tm= z%R*P7VhsV1$h)*zHQ$=QJ}{`f?a|qA8AwLmFmv>_eil$SK=O)~B1x_!zz*|d9L54q z6DB^qkDb%Ho)zV6s=3ZClmGeV3{oP8wh8uyWe5dhX z!qvj|*oD!?0_GFeK>tjcXXvS3wVUqa)QAczw;wFUuH)mQP9JR>NP)S0!TC8scg zGXjNaatm5wK|h|w^B)WqhzVZ1H{q2q4m-UH-8u-_6MQGg3k&J4{!uL%#$xL~5T~xV)lM`sm2$TjZ zjS}WuB_RZB(}n|CcHrdp?PHzcGD&`<HjpVF8%1fW3t<>l*O99Me?D*N+$@#MA$o8a=RSL8Byhp=_o<;?ukW$;E z1to6}Wvl%F&{lD;*jg|Ia_=_iwU+-C`-wEy;f?m9a-$Tr?YQIh8j*CQp*(DjqV8l? z(CAno+;!Euo?-3-({3zV4Ra)C7X~)i0?u@Rz4CB!ri$Se)d1siqO`+T!PuVh^f+a| zj*{|)z=M)p;Pt!XTn=DUF+G-dT|4%rI~** z$PQV{)DKAultX+AlgQuxEjM*j?hTLT@NyLb)UqePb%B3+K$ZLqhICH^PPL^9Vq?{! z%kfGdl2!^%_QT#7vLQCBHAJ)E=D<8zX}Nn$vy;G}v~7m(d;LoZx5_zPgD<*jzIk2r z((bl?g@~qm@0PLOtv8N$+xeZ(3pDeK{_)VzVMTMvpRt9ns`c3d z(wXTu6FG$mxIs<7YQIcmxd$a4i$w;~o3Lbv|78r^71pvPk1ZnoYK=AZQy0+w>DNnm zPW7QH%JxS%p-XuoMzph%2bR2bonop1f8JZT`%n!iOKmF`CIqwGmGnfgimi5UuG8k5 zUjT7ezpj=cTS^+o``@3EB?g~n5+nz3br2|`_Yds5|B@uvpflk?iPMK4TDdS zK_3}~laSpg|2MDftaVv2UKep6g%`hBf6;03AJo63kBVy@6h6+>7H%QA8)vHTk7$La8pZ`Oua z^$Y;f{S(2nFtK)y6q{zKXX z>yvMp+|E4KW-mVjh8UQh9`(oK{aR0LX=C>f(4oabaz2n%VZoy zZ&8w?$-p>9l*FS6u+MXt*S+k`Q~DGMI7{?k-k4U%pOw9Dq7ki(nN|93#@%_pZsC0MnYsVC-gC%;MHo!lG zK1AOFyCS&_5+a85H}=O9y8l8u@p8fj6woRB&3zrhz^_pU@Fk!@Fp&(zyr(ekpfW5* zXJ4Z^#bX|b5VT~O@J7jfu!H#e1X%g4V5MWV^EEJwNV~ukR+E=bC%<~@?iyi*6l@JG zV;_jCBfA2WBMeZE^d4P2mIUUc2U~*-g4!;|)zej8PJQA>GM&$h30IVulRp2^)O0dr z=og8>K*%7H^bmoc=LF)Tdz#%8!|p_PHRnu%KaKYJsKOXb;ioad3;Lwg4a(zu;A9oj zE5aKN(+82ny?3fxRtIi?R*}Qv%N<=zf+r!WsLH;qR8ED<$Gdcgv`c&j_-w$hQ*`6Q zq{Y(pd|LCk3qIYnN^6_h=#>|7j0&d3LlPtaTD(>P>RW=kbr;u+BQ18SoXse%iOe33 zjTp9&=aVKa1s_+E~KfQssU|@@w%X=w?2^Q6)cx?GS9*IJsqx_yVjI)B*#t zYYB7)ocm)tq1}tGNyBehUS8qSAlFjTMOlFCfvnrdXa`zn%K#glN3bevTcXqT7Xj=u zv&rAkZ05Ul@9pl_gN?^JW+P>H>IMb=u>jA4^~RqDIaFX@mhOl975@yHOaMNU(PfD0 z!8HT0+RZS=zK~(fe;m$$KsfdVykMS5r+JjCqLg%xvf$x_M+l{};B_%IM*F}_K)1T1 zb7cgbTfn~j*_iw~oEh}pnd(ihdmxE7Hz8`Fyl^mLapx-TfQA(B0}c`XPrxk#jx~aP zk!L}*MVoY&UW?1p$sWe22iqB0@f_;*juejti z!6eA_|8Vu@@lbc~|9Bh9+9LZ>QOK5MNJ1!5*_tA|l992c!PwmhA>5Xdu@eTNNQ^Zp z3WFGA&o=gvEMtj&=ivT)-;dw@->uiYUgwqxb)?3WF9XjeNE`t`y{MaP|t3Jk=$AcO2PXA82efJcs}Rnt~BA9d-_|F_gE|TQL3%9Cy4Ez{p_9isxy-VnD>< z1b79{NzyH*dGDi@|MH41wH|xzN7mt)O@5yy&%}5Dym^iOzV#DUaNkcDVkn@oUe+#q zJb=;xbnD7JGJB~lV&Gj2_SmN{tfr1%y9@g#hroGtnlV)A{tK7eteBF&p0+dv15kiF zG~jymRf*g~_L{&ZQ8q2#+#FFWa07;C@!fuHIR`l?m9}WEwY<|$dqiSHsywS5q6ovl z3kVME^!L33A~P!e-)~L6OXjs(7glq>ZD-$L&)M-$isV2S!}>@)!`6Iby&{|DC4PEV z5UerWjy=aWOu_Gos=ar}teu{wK_I9`j3K;sOFGBo>`G%V>bMIRl8vf33t*@bRdnQo z&=vm-eHrhGySI^McW#5F0WQ_v+x&=)xwtY_{XUg2{qO*mxRhp~=-9*!Hkl|vf1tqv z{BUU%Srmiw>@qsFyZI!@XZSSKug|lORaVk4;zQQY%~Xk^S_3a=m(b_Xf_AcLLdI|9 zMBuvneLI^6g_5lS8Gu@E+7UdAX!BM5C4>880o~|`7vnD?59efnT20dZI1^}Rjm49uhx^`ad8jR3A_t=jen#_+~!f`9fC*{o{8;{vN zeen%atSTC3uvENB>@$Y5FJaggZW`d!)N7l!T5^)}7rReZm1zn+*3Gu>$&Fl{8&dM% z@|I=xK!8P?g7fHYmLNWMj9I&IO=+US`qHE5~C!wZ?Yyl zPkH&=BSsjsOWO9k_zGo<$Blj&VRXII8zr2@em=bhxn?(gXkWqym7 z19OFf1}Jh5=h`yJE+5XRe~*v%7Et4A@0nBOU<32*)g zOmv996mV5=L_mWN6FadCS8Uj4EZ#-VJDJ<^nVMM z+{An2b4TooshAxj<|)o;%bbcFWc)Rk{lSTo&ag9<_W%7Lbu;;!-;{;`SPWp+*N|Bc zr=xyVOy*T3&eYuH(Ma=Z`UjW7@kZwv*aum($QS1|piQ@z_mJBWXKF-4C3!ujCpS&e zBHC`@f!I*F^!pZOA-qsmdxs84ho~2sgVXD%Bx0=#4zfmnt#<^vgbk<`G0-<4~0=MkUT@ zBk*Huxh?uMc!3!8fdKUet1v5~-X-&<;TEVUgqfnOxiPStOfc001KdSEm%9n<&J55< zJyR&~n?s6*9snq_hHEEPS4XS}{18X92j*0`0phRMnN(}zO}yfdLW};aslh{|(gR-r z_S26v8dzv-y?dq5(B9Cb2YJ9DJ%IG-PNoO{AL3fzoOJrUo*3G9B9km>=EJy4Sv;id z^>p&&_EG>!dA{l5R9=(Q4lbhK;`AKvO}L<6$Xeort>FuZ#RS?0yZ4iM-K|z!Z540F z)=pDDmt5yp_s{DIqqAwnb$n>hZdC1^99!Bsf%dEqhgGHvn}j91zlcgKTv&-4p<9xu z8i!X#47B;&JkKi3hOpSI4Xm8K9=Y4&_!>XBT$uO3c9#E&q39Tv2(mxkIj92?0>vB% zxGslUY#YhPlt`&@060HDHA5siT>l9ousPRRwXHJ0g*FnZ^n(^xA71dJXAQ%}Z_(Wv zAbju2`2tdBAgqxQ0KTPQj5pFn+;;6*smZ@o>mUf$Wbftw7IXO|pxwfzN}v;Q4DE*1 zhPn@SAG>Yx9L#h1HJ#LRdM4y_1Ov1OD3rv{T%t5#SBV#t0yn)4)!@UU=o|WV5JoLd zVoY^UM&wZ5xZO=<;entNx|(NSEtgq^k0SyneY=+H-3(LNc4-Zr0kEV5${U5U!Mmb& z4PW?HUI%~crEFLZzdElchz{08Rnf-hm?_0e@5V{dFk#ChMH*P$9MVd*|7;U_0;tFx(|>o@zM!;TYX<4w$Y<~S_iHp+IA z99_x6d$=_N$i13Oh^?Gj;^0i78ow+XUs~ z)IWUsAO|_I4e6jxXutUNF?_nL79^;2;*r~Xe%KD2m=RI$;FmPn(wshkgThRhor{jM z@W8bUvNzh8Na2GLe8-Y*2a1dQEFE=p`)2~3OjL;UA$Uzfny>IU2tz7wJN+~J)g{AU^|2BHpDmqoRt4viPM^r*N}?>6P#-fSbGX61Y)%M4(NJ@qF;Y-=p$Up z$D7Z_Gx1lV@gBWJ@imm+3VZ%+7u=>m2(BMneW|*iLJb_8gpa=!eMS<<1&^rc!8Ms7YIZJ= z&|~)Dj;qniq$U&=e?$Y@-x;J`2RT-8sm=*T%JVBWhbmapbn_&x0wACd%+D%-cv8*p zRQBbc4M!W&z?q@CT%{T%RsGNCFGWxp%KJGXmH{bY2j@Z@n-ZW}Hk~S;Y&-viYJLQp zX8#hrBBA=+kCc6oA54ZMe-_%Eg-L2U;9>2swSu9{D_z%a?|^)T=U%=m+G4d^?i{JH zR!KtO7utX6ekHF<+}-sL>0uMtNFH>sliwmPFWY>WQf$C~iVK8=$ot}1v^4Mf9{>Tg z=acB6BNc!4?}GmCu*OA9gczmvS9DlzWC-vGAr#@2A7Km~2jdhJ;9J4oGu01WWOraXpo~ z`MzIA39JxU!-&0K)QJ{ZkVplN`77HC;ozTo|8}W2Mc+-kNMiK+<+rUoy?SET4$Io1 z_AwA=!F!bbX@#uYD}MQ?({ol~X2d~1+7E<6p-mcSubd!3hgknAvCG8VuIpA0o7ylH zMoi{5M^BGTo!2{s-niRg+N0aN^EC4OVIUeYhxBC_k!)%`KEF0=CSV9@n4RT_rlJ3+ z-qLU0qWwPEvjpGWz0&W*6^+o9v9_=px9~skB^@g~Kd z492(;hitCG)h~6N?iCi*;{Y`apnR=-RttLKz}7y$v!3IK*&-#D`;-;*w=x;<=l|3O zBVyt{!8~F)j&_++OB7|Vj55(_(XH6F)@xl1)VjCJEv3Y< zkolNddiavuu(SQ=PYTl`>VM_RpMR9MsUWqdrs)Bj5Tma`Wo+(RHn&3$qqE1H=VDmF z(vzx2@VIQ%E-}PG0K;R~SVw^Kn2g z1pCc85RxGyl-PNm8eJ3CV?_tYZ-vkd&TQ)wcDOG9mjPOG;z0bS7Jx`(27Bgx25I3B ziWow!W){E|4J_GMc*zW}vU8A|K_Hv23C^{c69z{$sa zX2%}6S^e>A6&DS?I?>P^o03bhc}%x!(0_s@)tNw7`;e+Oi%BLY!EJi!=^yQZpl1!j zmF-}Rji~&zARPgisR}X^K7(vhTS9KJ$zoPzC%x%u##NT&CwY0f^ink5i-{$(3Eos0C`&+ZoGoVPO$5_3miX4#HP=uvtp zF6rrB@z(80dC549wV`ASr5j!P#s2Z>epn_l`P?)AdAGs1OQ!Xb+3Dw|aFo4U&8=tO z6AxTVC%XZ?H944i2#7hLhbDF|_(?>dP)hWLm9$NxSQ)B+$x5)AQRauL{*Ham0S;~Z z)=MP^`p**B)^d3gBiqCY2Y~r|Q_$Sh=?Q9gHb^@KNP_p6HK)da{r}Iq z>s@S!d|U;8xK+!&?5`tjOFobn7JDimBd}Pk&6c295d)fyf~NKN#;Yp?UASiq0zmYZ zfG=j9;+|P$UQq?~A71pO4pj`V=VT)H;xJ6$3krF7&eus#a~roJR4pzSc13<*bHkPr z3E%0ziJ?;no9aWAf=(|kNqPdrVZO)tpH((a?j_Boea?QYNl6bszCC5Z*e!I`|C3%P zf96`+rY)HN0jv>`8%iIHPFk$x$sdP$^PzqTC2Awd^TM#<{>!$c$kKQehdHaDNn4u$ znES;_vhV4aQL$nY&n!X-|1qxQ;zjm!jqup#)<|a*ku8T$HNolZ0Aau<|ApZq!|4jt5{+sx+gNGkR6e1Dgb3~jhw7QArLt~bEUtVR?FnvpR zvi14PgGc`5u&bk>3fvw%>5qQtVfK=}sztYqOM0r>+528R?ee2T6~9Ilsi);09wLPE zIbKz;r&Oe+`96yBct3H>wo3kJlxw5)$+@p}R_Ap_mp{;=SF_pQHL3#n1w!p7h|3qC zpFW*0GV!zq@M+r>_L~L210FBl*mjuhyu+|j|L3|{h)y%!VNyN&MAinH{(8#XH_J1X z_q?CW+tJ0`*2kzfmWragsb+V2BbC0E9OldNno7YSK(Pb@YY7axo>(5!)Mn&~>_&XS zc+yhFQ=z&(nKT;p%6eDd{{!y}lm7In6DxH+_4Wgo0wTKtq1pclzdn%q;9+5ib1H!v zV8$6R^x*Y}VMFAV+l-9i& z1bZgiFT_ZjQUtI0U|biT$%4Cm@H_C4Vi0m9Bj%hk!93M|8Y9A^jP9NwBeRUPt=HYp z5{C~uTyau}KqkHJr2nrTmnI$cxCe`v%;b+7{JNWH+V!6Alf2g&@<)Ihi!3dR%+!0e z9CWL(8T5*~+IEsY=mbmVW|;f|v->2%vrk1PY*zyH7+YPrP0lDPUp}b;T#-V;*o}kU zhr&lqnhvMx%8tzGlO!bWvpEM^>NB>Kn+;{X+!AtBI{fgW=e*b@qTu^RgkFhe-7{rc zde*17`*}jTlGmW;38Eh&@K!zr5lQLEcjV-nRP9{qXBpf5*tPg-%?&WFyldczd~EE) zv@&OegIef2C#$K$RE+`K+aPNGOw(}SpJly2&0o7%#T?uaZGwg4mK)8%o6XH(ufm;= z_a1%Ia{P%|((T5Ix#tMS~o+LFKxptZ3g=0*kvw3>^o6}d`i5Z^B z@4V$kYVC?!U9t-0W+E2n3L5fGJF;^&)sVY0vDbP{TP@uKqr3QF|aaGKg-hElqgw|x?h-rWVxFgoNonHM%=Q{p7ae4z!iGg zjH0=QWp0QJ3dZWhj-U$C@8G{3U%i>a7B`8;mB*Y@(@vaMtZTC_|lp>cAa#M42z z7`rOjhx398PmN;PFq$8WeT4Pr-s5Z*R(pENhLTsUoIx+~Ma!m^o)!RsU=75LAd3Xr zOm!pR6bG;nx*GiE*R+kD(L|b@i28VQUX`xSr*6I;1210A6Ok%{)k7phnZ5!%fxm#^ zjRwc{F;N^&fh4Uk&b35xG*h?26}OF2y_2G71m%F&Lsr#=*+>(p*F3 zIc_FG2GV>p0flonds;K?|6IJE_+my>K(B<@Q(ssEnZs|^q8->)uSzB!*iJ*RH#kJTMtd@y&NrCrdft_m@D z1CcMJ%W(X-8Kz7G3{5?}dqAUM6|CPX8E;^J*s-9zo>Anvq~~v!3k_=8#C{zt^v4aJ zzg7iII1UmcpL?zQq$gMM4wS#II;;cveBW!t==4(mM(>pGcy}>ez~hL(n~3fG>tcnH}Siz@JN8u z+4p>80m6WqshqrdDaWP_3?2r(^4l!9j$9JfV?-BjsJMH1#qA|7Za|~_HN3ZDob*`7 zi*IHhhb&^l9e$8{zYiD9ii0^)h2$sF+7*-N)~5IBt#gcr&T}_hHj7hlIAL~PzOVD8 z#7oDQt*1VyBMN2RO$=qw4^kav(d+{Xg#}_o1lLX9#lJ0o)p_&2($XulX?o>5 zNR(z@L)jf?6{g+swx~>UD*N%pT5(<5I!aT4%RxV}a-trmpEGdivo$9Q zf9NqpYWiqufx&0@jw7YHW(A@LT__b5HV-6^xLz_uW;ef|ygjUd{@nU{xgd@&B@8$6 zQZHBgtdu}a9PMi6@{e_-)>}*6i95G>dM6(=q&iWBeObMDy1xL05On=2rwmP)Dp5!o zUE}d190nSIi4_WMkMA19{-|o&WrCv8-6eXb0Bh02*~AT-ZPAQx_~h}rD00pBYZlt7 zH4Wr%q5CYGx)^l?i&i2A2YH=^l6|;z7+Si4Q7QirANjrtP!fOhkrkK}ZIRu?Do;{K z3J5PH)(;M1)v}nwbzduzcM6b>I=UHa&wJ%*lSf8zjMpqe`xjPXpROd2Xv$Q;SO_439bJthI@$CH-b+$~lIW^6V~e zE&yxwGbdV_C21&1MK7THY{fGJ=k68hKT%TbwwfxItg!(Fq9L18XLJ$6Mn=cgwnz+r+glogcV&oWWyicE8Uj@L?Jo#AfpI|~05+TJP-ufeJ6S@d6) z4y^7h%)ab#`_!G6nSq}z80*73o=7&6)!ge=y5bHlBrGZPUFtl26=;)VlX4+rK8iJ( zy5a+avYfKr)ITyz2kO7YV8$L+wZDZVv_YtUYVXX0Ovnk zEWHA3w`!lih7zl6wQE081#26^+o}?IK$9}p*tDsOT|E@Ym!kl1w`$y;<5)fa?r89X zjnL+tzg5;!V}H#44HX^}UJQHs?YPiQgK7@c!g9yZ(CxgXoI`CCiMf=sw)S!(Clgw~ z{=OgcPGfs=h;sY<6JgHlg{Z5Q+KE+7i^$wqCDpfrJk<+yE(CKQYY+yMs6C(IT5#nw z$PLd&(BnLZR_YLEpqEvcz0-c5#}&XCL2Y`x(%t2PGd5rtCec8ulA&$9M^4{&pmw4z zlcmC8X&E~s6JC5U{m+(ow8vp)%Cp438P2cV+gc9af11*ol!;jILMS zlnLptc3ofw3K^_K$NJ-QF*}h!>R0&ZKXrVRRJHgKgK%0|ehd2l?L;i|pa-I_N&ID3 zl#%Lr!QvCIqopC{7v}s?MEu;(EWWlQq&fWy8BzWRg3`!7evTG;B|Sx8zUXu#Gd0iR z*kb;!Q%VBBW1#7ihx&>t#IjF}pi}1h9rulbfV+}wjw+C|K8VWA>}Oro;58_a(Bt!) z?f3_V(?w`ta9NWVQSY_wqKQp)@k^b)M7~;%8!1ly2iedg9}_)iX15RRCVui*qlUkl zo(y8RFL6CZ`SLD5K7{5K`1?PE!H1A+~d$y5cy0>b`7UeJU%%QapF#kY0&s z!ZTR|D7<9ME1Smpm5%IMM63r=d;k821>gjFkjDcUj->qQNy@-fZr*~#Z^1NFx$~)5 zG;Up~aA_eeSzBWX{apQKV8e2|yaFwL z-DAdIwTxUz__oXn`(k)iSU$VV2Y0cA0gXxm=ZBOp?nWQ1VZemvAZYNhS-LeiY64$h zBLDQ%>OEn-L%_$lj+xPp;)iK^5V4Vko2JEcwr<7>w<-Dlu<{v}f=C$%th?X~U%&kx zhpAo0@15%-c&?0~5vbp<;NE%{z3&!jrQ(jfWpM9l)s zrmex%-Af~%D5^-USSp^Xc|^U;nTJ}j)a4^6wxNRdfhSxbk}IJh9V?=jGUX^X_M zoYSdfu$G(wBicE3Pa$OJ7=AiRYmSh5ahJ>(C_cf<-709R8}Fy_E<}U-0BfSAuiJ%a zs}lBuneeq+7Y#WN=%+c+gk1sA?N^^GtCiXIqI0!y`v4XBw}qd>1zWS=rR)1i7EOW0 zQBmAM-y(yuDryCcEbv`o=yT5UI$VibPu=XmZ(xeX!U7(TZ{UqK#2)U~4H8mJPsV*( zrPzw1WkF5u4(-4bKHQF-a3@AN)eDw?$Mf=vtyd0gcov}Q%VY^A-Gh|X1^B{k`uk1* zt?>^o9ZYPz)291nH>7(5>WN4?wBshrnG{dGUq_c^(X`>8LN#rs6%+rb0?KPTx#H|G zxp~_xJxr|v${gZ!$!6?NnO=YViRTPI@RctT_eP~o}oW-S7g6|%OW*5sBAXgxA$ zgjx8ShHA%mt(M4&-tSPBNOyfF8Ks*@Q3xsRi)0$eLv)W#m?1w8>;sp2uR%#2V0nt+ zWtE3^K{`ToJC`W#IK1;L0D8VYe@`ZL)p%lZ==ke&@18k;f5^S~iXlPZ)rG`xuA~L& z%D~52&JPKQA9leNfsSk`6(uRn$`RsheY^v!VvEvuTl+GHb;a(v6rGFWSm`e>4l3RT3Nk+ z3R&c-1V;%0=ZOS)=T#c2qN%oX_O$?^?AZ&AntX!62@C0xwfl*seX8AE16XETOH<;l znaD&)q{TC*)rugIEC9;=bb#^_M*E9SumVIb`jxeGe9v;5i zrX+>d)==i+Gv220)-(6%z8eT143sWJx%YwkuEBXwiz@p~)(FPZ$I+Icg9-f=BFX*p zX0OGoX0d3oBLNaDm5i z8+t5b0k{#-q7t%ck@(a|f@-tX!Ng#JC?*sasihnj7So#`bN0>skfSu|mJ{E0o7c60H62JtkEvmCyN&RYa{4agsjdUdb4_Gwizj4$W}UdiLu8nLR_1h zHo)iR@uLs*mabY}Nc7-glc0Z(qN52i>6q<2t8a+8Z-=qA_V6oO=X@EXFK$x^55|T4 z=t7Xlg$sazyh(HKI-;X?10%ck2REqCN%^vs5@6L6el!JhCg@V;vCHBbX*ap4T5?OU zQb<@yIdK)12_4THDE3hc60u8kJbL>{27?g8H;pQoa9RBF@rFJ<<8uhKRvzwQ@luld z9~S)q;8Fwf9N8&gH>qNhiv7-IWOB7^VKyq3i|vln*F%(qmYtfQO|0B%$=1azDdqh} zS{HrFOW(eC{~)x)-xX=bq7$Ev!z;i{7}=EV5L5QYkDyd_g`aJigEY6bSspXrGvvi~ zFndj~ww3on4_^$jG7#z=Wqg7MvuBjqvQ>i#niT1P0Os+H+GgYK=cEfB-LC~NHgPKR zJ)2Bm7u&YQAiaH=u7R*P*@E*aa~k|YHtMlohfK*{iu_0eKx;OFF)AP)*s{duq&m|3 z33dX>UM_XPPix?EOtt-r-LQw5IyJSM0$b%{Up^_r=5~MOoEE57EqjVcRMLxY0nrFm z?Z2Lu)kkS)0Wzkj+Cwtiv*5{8zR2*ILev!yix@@IAH8dSRKk3kQuAbZK^{RZue_@N zrc<4FbFlraDpAAX+Zw9~Db8(S?rrLodRYOF4v(d{#5yfZc%=fmT(t4hT+@-mb$R=_^d16%9rhfOQX}?ni&2neWUq1+i6a7X2CFRc zt8Vh>h|gW`G1l37$bVhL6H6}he|;LH_Tu}i48a>)DszJ=F*`J1@Eu%id%!6>?^Q!r5vsc7z4@qNZF!WDy&y1y!y8Abm@q4%f<%9@gP zs#j5*IXAJjQ^RouR=F1xGJV=Ku+$?lF05mCeX0kSLV-MoFxVXla0GGiufh}(YvE?L z)cn!hQHdQKC@lluHWu7VPH$!G(?V0BMJE_8V~RDq@;) zJ2z6I?aT$*gSfvTgy3?Dy#y-!?XQWC!CIV970>a$9_@b_Z#cDN>vbwynPK^4K{qB# zu1S9DTqp3-X*1?GY?&X^@SFg?gL2Z1xR44xi`>C1VW_Y!rlZPnI5I`2hspT&_}9MQ zpGJEr-TL#VDET7$`hHwi=>3(-_WtxeTUcXt>>OgJCTED9LVsRkR_>vs%vaOYcQsXy zR;TTg+=v=ohRIJvtisW-UVp~*<8Azbzzp(yBdCw}?%O0lkqR1EaRYkt2$2|Pri}(F z{g^ABA1~8A=fqtRQ5n?dNSUXIbJ}%`)jq;|jH1JMJVP*&a%kax-!p&t zgd@fiq{v)Lqm-dSrC0ntZ*)^VUOuXdhdm`FX+b{@z@6T+e~YVKi>RG@7j0nYqP@uH zeA-sst4T(gujsAM%CwkVrRtYli!-Q`zJL3JA}Sw0{|Q>{gDZ74HLM)00N{21nbNI7 zwp6j*)8T=qQ>7>KW7Q{^Kxf9n7$*9+>czG`GUdt8^2g;W0T+5`P3ZI--t>Z(WMH+g zEoURwepk~!fz>Kj!8m`Fb1sxt(6CS0yv6JwSmaNMPRT1dyOLwcP4xGhup z*S7f|{{gu#au>p9>kY(l0i*3X1Y1QW0#p{VXd|_ve~W~3plJDIX1An0ob+;v^)Rdi zdo0Ok2)bS^3~ua(?(->I#tcO*z-r*k z))s5YSNOPcvxAMEF$`znb(eTcWeXAlSVtX@E5`u3XqCl$OQmb8Uc_V|Lblw3WFKoC z1=6n02(PM#&xVY^j%Ai}D$DyD-vbZmby2oy772t+d=QPG(XF1}jCO$nV(Ik~Q;rqz zqRKtfGjE*L{AsItmcg63g#Rdje}MjyOFoC6?ItwqcHlRPhBD=oUQEcXTdFbDHed^= zRh*^Q0alKq0IXbd3A>qqhR6w!9npe(@kg&@yDxQNin3Mb6beYGuy*$ z-~3eFJzJdz-fI=u2vU3sLvw4VGGptXL$(^G{&jjXx9-Yr-GZsx4bTiDy$DD<3*zEz zz6jE8!wjE2#JK(V2K>GAqtWFJeai~~?z`mIsZ6mcJhFq(LxxEJca+RgfvdBsr2*aD z4+cWQ@W-4cJ`=rg3Pdta`F6M2?0GrNfUP41KoyZbhB7_5HE`$;bWEo$Sl;M9J8&?A z%;KxGbXNtp%YbL0|J!&BF`Uh(<1zb0MSi~u6dD^n2WL-YfL%_v{Zw|%g}{oO@0`&styccyU8!IaiBA$GY8B|PeW~lS=yapi6Z zq5K3WcO|2JDHbcFTh;ca0js9Qtt5SKlwFz9Dt1A?yl}cO|B77Atnb4 zybcaBNSc}+Yai%SQtH3Q_H*EwbpB-UOzXAN1Y&h{5aL?Ji$Q;8$Kxg>Q`16TNm6~# z7-<YqL!l4?%Z`E4;&cd2yTE_+pam6q|<9cq|7B9sKo{mU@)Y$_qCkG8=qy3W|$Y zpm)#bT9y{?K11ESDatYPa^YZdWN$@ot@^--#mu<#Dt6wA`QKx!tzjN3-YI>YVLJ ziI5ew_q@z)Oj%qlUwIUB82* z*)wLZ-#fGTYwW&I{oM7lPNN<0JACVP1cB4cccjQI-v|_6a z<9o^R+f`lLvV8QqqibDw)tW*lgD{4@K%6`k&w-q%=WRQ*me>ZJ|QMBB2*9D zc2LAGi7eXm1dZy!@ug*xb^-(+0Ne}EQd^0`lF<{m@-Z!*B-ymf+!gr8oWtRzJ{7%m z5!f69@THEsGH%jm%?0=oC6f*zo9{ z7N0X~o-@a@hQMdAD0Lc%G3Fdl0O%DPfD?Jq`|2Fak29)4OhU9546~j|EUm zImYdq%e>2M$J5DkpDJGY+i-TY98t4exc7J}7;}9t10BFnF5Ss7*@B!jvlTJZKZ@{l#=nC-7 zuT5B3)x!E}nNrO@_^WW>3-W#$BP8bqh(Tf`B1#sHiXNcODuqF@;_uxMXEVZxD# z`jtEE>_@))VX$&GF-w)WWp!HGh)(AkojmW@oHe$xk>vF!JCh|rTN z4E6_teZ4T)=d3~8#QJo5%M1_CpK`pyl8v?I!OmZl5yOkzu{mB5!*i2v@O_S{ZG60D z{%oq*T*coxtx2@T`qNKgrC*Y6CANM^(m+mm)~60$56;B$lpR8!vk}V)a z=C~{b>yyAH-2k6pp6kn)`pjpN2bGgUz;(MW_U-x0#srOJd%XPq?AI4hoS*y2c_Ywk zD2DL?i}I0AwxPunj*r4{IgSEbLqyJ|kRQ*G(~p-kd);OHX!dY$U|&f~=i}-66>2vt z^q`9eWEI>ZAR*^dUGK^Cze2QU1=<6zG{&aN4jSH0?|48V3c3g%$*@JRv#@lRjfN}> zVl%3l%D#FhC_4JoTWL8X4t}#4=3TpbS+6W*+^XmH<|EsD(bd6=N5MH(FNJgffeAsw+9ufIk z>yLewS(~;h(LkX6;It*D%IWK#*bjU%l;m*3AdmDXapTihT<`0J7tFfJcBZ^V#TFU) zSTBRLFH8M-df-9r11<8cm~A$awU{8*$`E-&8n8el0%W5o?nCvYg7s}jU1c;bB|^K` zJ$=NyAo}@5l^2fDv*#*)B%<~a82PBdIW|wwGDWDzXWF@(9|^m!NyN$yEw`R!W;md7 z=Xv*QDqi-NW~XmqV*7=-`@ghp4x>Gdi!Ys~XZ#1GD_N=*8i#=717=(4PH~((I#^{b znYx+$s8^GC<%*abd#qw1`}ymyo?m2lyIfo%5Nh?ZUcZ%wOAo2U<`+0Rs zSbA5NsKNUsrAhSiW!NCi_L~izYhwfdjg|9D$VLtr2)|5o7IJ;6Xa4A|k=@lfwnNgl zFBda*-F+`uc3UwF(^~i*gUE>*3q@qz9qO~av60!1lm$d~V(0gS zxW|Z+VaEOwcyQsSb@LT4lAE1xZob=<$vd?MV<5rphPG@^jyD@R{;=lE!nX+ahvUyP zHCNa&jh?YRQP<^GL$6-Q35Z`5=#1xXT-_Gf5?jjtEUS1{6##Mk%Oe_J8dy)Y7AigE zX}PjvCDGrI(A4c2xZ9D``mESU5u}Fif|Aga_r3l=ak^EgfE`}Vg)M5qwReZtz6In1 zCQ32dM5XzTZjy&6bzolL4(WNyU(C5|vhjQMD-uF?=v?e#J<7{L#h&5I0fEk_%sTn> z=X~-SubYSboF>t=bcZ5J}SESLR z#T9J}rfwm1|NVLxqOjveX`EQq5GAJ2AlRp=`IXc~a4qwz`6ao8IC z)X*J$>)|*>^uQBh`ppAR{$JAGbk6-x`NjC?v%TJ$F+K%m+MLxxcn3%OxtmV9jUgSK z9RwV4GGkvKtFcP-+lj!y6>BY%e*27}lFZjzDhmJAJ%T&xP0{W3%C}J6rB26=#*Doy-vH|f142Nzsw_QW zeEh>?h2#%~sG}_@U;kOvuu||KMe*pE5eG2eSWcRBi)l7C)7j-t5)v^+rdsF1s5;Dh zg?$|RKm)PlkA72zFeDG8@G?|CsYDtD_JS0{6eOkB}O3sK#&$2y6~%vIjS z+sRTFuGHjnj^**I_LTKi-~!@SN~#C1soVuk%^Y1N2$SMk%SG##+2gtCTo*Dc)%|gx z^^cGypu-ZxhHbvsvm;?~gu*K5hTgDHpnxGnRR$!?T~8=XsQaysi9E*FlW|{l)Ko4p zRPNeEpA}i50t-$=s#?-1$opegr(bkVymnA7r6b+?R@ZP<_`*JPHXrk%Q7sy(Zi`B) z7cbMR!d>)EBOf9#?(YR^1yg?M`M0?q2$O3wtP+@0z05CXBQjIkR42Q5K0T@Ez4(i3 zt6Bi!$ljk?n6)eg@#8B*i$5P?Cbv@dVosiri~s0j`cwBi&9qjhI5FHmh8I~v%gqUD ziGCjo^i+r%Nz1-0^r$mJzA1?ppK$Cc@rWKE`K{6yrj+$ED?z^9>%g)5A{loaTJ`UH zs>#FCc#+Jyp5=C%(mtIwZ1zYR`MysN~k%Dl%nCFZ!wih-} z5iL3JBI28lwmhhXWJ_a@+5fNrX()qZ9~v$EYhcb$g>!VaE5)x`qi;8969jv*faL|c zH0CA5R}XQ7scI)QJaNWSA6Upw-A~I-KqFfv5LtLd6TME6y3KLnb|CC6K#LsDEsxqg zVEpbL_ao{_-kL}y-Ez~TXtH*~(1a5DBEFaW@Cs3G^sU{klOR$R*!CwlCZ}w33$`-_ zCekz%;aXRSKYxq+J#0@7Q5xb5QzP=R%w7Y93goWkLOKJ`q}30Y!sx}k9#qWQ_-;sO z6k6UV{BfzZtNNJ2o)E8paQh2U&*`A0;FAf$s6*H52{ks4S$0nD5aJq;Re*-&lho1F zimtJ7L3l-^+`rIiu}-nWk~mswrDQCRmnjI{fwUq$6D-u=1SN~aa&Kz>5alpZ1$uqd#}#-?dg!V=gmFkG z+$m6Q1P*g!kzjM$;Bm`Tjv-iVU~tYdd!U1Qi-TRD$V2SdaQg zBgVpWZe#w)Ue|@gfWMQDg1QE6b0l`MdZT9SJb1|-;sE?Wa*MyPC}(qfe3v|nkUU2~ zi@H^-B7wi8flhM@Gdzjr6}15=$fCDOqnGc*zkUIpgM9*aRz#1cmE5A-MPfZ|mtona z?3uTSWAw}6pm-^8v&F^-+XWia?^dno?;|B~kls*LBE0BGjO=YQ;tO&sU$VYze56h^ z+3dHZr~22bOLHjDY+NI5NSYt?EJEh2pZ~d|Y~pq*t{#Atzi&^aO3NUQn*V2Yc3w}$ z^wPz=E_Ab<$7O;`u?Go{;7U{^3Ri%xf$kwpZ3d&(9zJ%bllPIv>*9vKd_qgXrIl^1 zi&2*U0xS_IaV5m$4#Md#S_y_odY73a{77u=je65cFGEF=sU7JhOV3LhA$A#~r`*i} z39}}f&85amLzZT0TdXT*zrB7}RYf3(&Gp?g|J@b^;?c{JY_QR=#cq?EN%OyxdleJw zFu<6^V<_WIA>fcml@_dP$l-W4HsvG-PS?!tA3e#SC;`>bbNOTmX6e>HQUl*hZ?SO) z#EO9Vyy~HD+220}y)d5t@kS#TFu4q7|2u{910bL}>3i^70rz&)ry(6!k_>Eh8OZDk zJ5?ui>BBy~gS-!F4K~Zqq9x?eqQMtzqU=gbmmN?a^c5qMTE+dp4wB4mmD|8b+s9`T zu0^+kp(CO3kY3fz{#_5?eR0}6`K=&-D|N3+zb_QI0U&IttuMWDaZHUod_k@5^Gm#l zB6o<~pq6+*>qTy71g|)S-XQym>}7l?izjO%LiG5PpK7l?2C1rmbvn(ss@Suy)Z#sc zo%lU|z5Tni$59a8p>q^X8;@M#9jyM|CH|oHI~QWMBmc=a@)=p9{+reul(}+B;sf7q zW}g$T=^u(o{-*niq{i*{!ylqHno1tYW+_lTnH4bdu@OHo~vlqi<{p-hf&L?9c4$g8fJSH zBcJatKIoglOINyLsTG`)GkT<~8^w#hVc3ed9{OAdg6&9l2JUOZ?ULx4Rjzg$W$vGQ z(B=cMur?OdwXaI1@Jb=Nh$kjK6?8LAr7z=`y$mZ*<_8Ff10kpUz=~NIrS#< z<4^kKP+a=(MQM!RKTA+np|K)6Ewha93bQI?s~L(-GF{Y%`Fs(V=?$A;q2dwlXbLT> zgo7Mfh|ji(C#j=@PmX|-GVjJlqBRvIsy$#sXe9d~cqBpqw)E>$LfOHjJ-2sfB^j@(J6^05fDJcEk*ia52%70b^jaP$nTE zLI5io(crJMMS(TdyZOH>sh|QmwQkF?;`OTA1Mp{{r*;r=N7RXBzWqI=j^&?j;du`nhQ&v3 z@RhG@xiAoP;yEjE06eKrJ~vifS{mzm?Q5)i&DvI9*mMUc0q1Ji<`BB+;pXKcwCw*z)`O-mP}(`$Po7-c9=&uJooIhNqpU>A3{ z#GUe5FfQQ&x8p+(VciQ(ueI)6Fgp`8a?*oQ@3Qx!I^r^zIJyztY-z-InKdktlh!cE zG1-VK8Jb{Zi(G&}0B5+4ZkP>Ta{p(HZuba}KwSWi@aHP>ud>ui_LqnUe%1;bSV5+r zsJj*&X2P=p|C+SEh!c9%^QTV5x1kBgaPFte7wxF`qHA0d6h)KeJ290<2B>kyW;N$v zSE3De*4ZG)5ZQVEs%8TY*=bc$46Br5L9%E9XwkBXJ7;Gkn^flW*GCX$FIMc?Z*a=; zft>H6f$`wq_gOw>*gR^2*zp+@x@Z6NqN$dlwU59HcJT6SiNkVXdTF{~YQp1G&zKGE zA7e%%;LX6v;et98R2JV(`AkYQs74HGXA%!DTXuS>{d+NfJ5$i?$)>~-Z|?}%u6gmy z<=ok0$fkX(Yw43YQvMaD;Hts)L_pSo6DsZTRqpy-hL*R1p4?th3w-s$ddVnpaQedG z>cPXydf8a)3AUhs3}Y%NR5ALVkX49t%10!hKS% zbX@nZ>X%qv2eY=h_w@mzm9g)?ER=IB2=))UM-@Vyo68UoafWq@0RY6V5SRXY{K|> za9NM_Ay3*{K=xyfzCP_xc)$nLKea8)``97mCnXMozk`_Pj(;5+f~%XqQ?{MkD0P%P z#NP*i*9y_-S-JI4?w=>@B_tQ@bfgQSd0iz+&U|4>zQueR-+Mh$YoMFR>B>@0rLmnBKQKQM{ze^!6wD$#SsmLFLrQC4XL#2!(=&mKwb)pwX6kp;x3 z8|qj|rgE{Rbah$PHGq$-bFMhWIX+fchfb&JJv{KOr<* z9qE6yH&5*5*_z;C>6}8-E;;=r9);gI18wx0&ma;chYepXL~&3Ka~58oY_jg@l)e}SCh%~^N}Q??YfG@7v(V;0 zV`|J^ZZG>mM|&;!bF}7(gxZEh1J4Zi9sK^Z=(D|p%K^>Ds++dY&rQDv3yAGA3xfs# z_!e!+w#!-MyaN$@>8QD>;t$i|;lxNMg^-Yt@ArOp2Gs|y5LXJY;o8*HtP`Qn=6fN}5L}?_EuxLs#ihj|pb% z;cPv=**d>6Ob^9~>%A!5-N-~;bW3Po?hI*p8SM3a^e*6?T^Im(-@vNPI8ZB8TQ~=B6NjDL;wPaZLr; zuWRvLUlF_%xRg6>-X6k&e4`Ucq|CgRDKHjLn&?xX)wtvuS|MhkOj;K*oS37&U#F@&)-oZQ^Ao zY$IH&3^{M$?+Y~PgzHj@ z*TDbB)tkpd*@ge(rBEa#5``9#6osKkh?E#hVv_Bt7#V8`*^@RQq0Jy;X^g#+eS0b? zGAM)WDvh0LEZKk8ncMUEd_TYcdg;FJIp;dp`o5Mkn)K%f=Fr%$g0klLGmN@+vTgoL z^t7Uc&>Ar(&RIwxoTJA!aLp~pb0SE|>%lOaQPYNmOB!p?%n)vEp3{-3h@yX0V}~yd zjS)fuqv`g69pb9uuA6*lSbaZs35@MPdn z8_=)Pxv*?25Q_LIEJW-_ChTwK*Zo$}EB%3SfhyJm%do#Ox(BQV(v~hx^{)e~Q1!Fd z8$KNo70y{L`mFW1@KCl7%pH+msC~auam%3@>5zast0XjfE!-*4 zedkwF&l_7`r=xYDb~k{~dCmD+f49gs@{_u!4Cv{WlWz`i5?D(;-NloyKYO@*U3Yie zhPUG9dQCQLJh^7QsL6&kYpcvT_wLX#I(PZ>+SMnorcCz^_bj>2ap-Gj;s5f|CR1)dnZ`6+J_Y-r4iYa;gLmUxg}?YnEQ1^q3N zTkH9!I#l5-S^HC{#a&mJ*7k-uTel!0Dbqy#=gC3?D{$qKyNaNampa1DsLHId0-u;* zs{Cy3ThShcSP8z(N$`J6C#k))-I~Gc;U>DnMrkv0@7Nrif&3=O!)(RTE$qB&bzPA{ zcU~d;--jjElj@a3xkF`nJaDm)rKvr)Mblx$0-m3;mDMnKNFlazvI_wyBEsT_#)V`W zSCT}%Wxg3aIS(ZH-u$q*7YsfS=W;G}=DTjqC#AdW8j9RClszOXPS}>W?{uAWXS`_q z9l|E4ppg0TZ)PZg<>H4cx9~o3@qJ5HX)@~g+WLY+qLRp|s!2B!#uW}J3%l$=(E)|J z4jIcHWs<0T$(jIWmQ6oOk-7X@l~%DCJ#Fc|r7yW-E2i7ckyq}v4M$JR_(~2Mnh=L| zTcn-)_1TF)S;Z*xGRgbRb@AYGkX((v?&tpZSA5t@P>0zmBigL=6M}S))$78%<}RxVvA|*J&+ET7=@YW zmp1p&PEM2FE`Bxe#Z+E*pn^(w$Nb09QQ5|$r2AR@vJvlFP8SMb8iu`77=hIxGJH26hK$*9XaTpUQJLf$ zwPq|x!_zP3|JjB$g*a9l8@#)adGa+cfGrDzXR0LV#coH_2CGrz?Gu#-^rP?;1IuL7vyp z$lLjk(2Q@~*8=ZP*tbzP4J&{GT%NU$ZudtestvJ5oA@#JsZ3*v3*(5p_sgy!-@NT? z2p*UQrq)acle-h+M7%HU8)CB;A=psxl{%u!$<@=@c^_9f`dM&0fb7EKmj8nARq!*E zt~ixw56t)#3Zso0YTj_NDdHGSrWW?C56lW0oSlMofJX-PVHBpC7Jj_p&Rt87PhH(; zGF~Z_N8p+d&+9NUv8wCcRPu0ujEJTRd;O;XoKB!nf19Vc(q*NOC?Xwo#4uUMTDDP+ z;-p8f`rd2Uu zRSgNpbO_nD%jaxtU332xBg6tfYyW-IVFGly`i{iREfL0wv)b6YQG!uteS+vze}>4` zjnikfc;pR(y-b=)FPFi$yLzx*l2>RPpT7ppac7hPsG*<7*=0-W<`3S6nW7KxDoVUg z)|;oKXzQhZ!HaJCdAlO+Oi7Vfa+uSm;;VrNqU=geA|$?NBZ>Tx>qBhJaN|`Vo?;g7 ze1JAOW8r##S8rwSG&17xP=aX-I=b<-IR3!HKQ6}BU@030lOklHqvko$=e#h`E znW2_-7nP98)|Zm1ZAs#ANueqRKC@`>e46Qj&BRtf0HHsDHn5cQ?=>3^%oLu+MbfI`$>R%zr-(LhTZNim1apPEPb=(~; zo>Hbh#J}JWW7u6h#j1cNt74(^^sSX1$d*0P)q5ztlCxb0&wAsyE&6URDhJD`X0uL( zB10)LXXoY4LUWfp$c0~folmb!2|YYs`^ET#t1#*kihmm@1P`Mlw-w!O z@@zF)2mRJ?^S(P>qsiOrn(fepm{k{Yq&#pvtt(1C&Q|}nvQx0s5c-VhUD~@=E+8Ed zaO~nh-DwVQ3^QGaCdu`Pk>sG8xqe=41+H0eB53$k@$c0oUeTH>PXsHQu#8sYnh#uE zz!WBtO&C60M6ri%jY4dUrLie~C9Mr-{@`C=AUxFVB_1;>;qP^Q3KkVF!s!8L$|O}3 z7Lyg`SG9Ua)j0W`;%d~Q3*XTbr1`;i>4{@Rt>G>Qe$uITpOYkSjz~p>N{Wk??A25e zmcHz=B(|3ACurmnC4HZDl6#K885$%XmyUoa(~lM#Hexxa66$w;g&5dy8GR%+-gn9R zpGRd{5j?KA_QiJWNyo8H%}WHmPSmP!gZxSou|g2du_DleyB0L&?GMod<|TR|ZAWg~ zYjvI3o|ByhpC9ZhqKS-*8Ot{-AH~3PE;c)Fz?DA_in@McxG>LULOeYl`QGVgvZfO* z`OzG6Iypnm$$eeGx>GRe)MhA2DJ*Fydh{&C4tGtHx}G;Kh5on(&6b?SCFa2}%Whah z7}r3GaauyNgW=?nhM5sJX=3eM46D=I*>!48%;YCe8TlGcHa|G!rr;%h9HEPTgQy#Q zwh)_LmW9T!_v|8xDhkAA#}JHP85UbS_k1 zsTws`hz(n6cRjYLd8d*_yF2QS=_IcB-%I z-2JU;pvap8KIvuUQ}iN=!zGD2{rNb(RM)X!@-UkZ6i_dze=g0jOv1Gxx^2kmU!)(M z!_FT%d>Tg4J{b z+f6ik__kwA=n^mR$6DOly@l&Nt=1UgI-sub26su z#@GEPCm#cMRAtV%Y8Q1p`!qL_v-+c5!7Q^S6^>f%yiAT`2D~z;>)v3C!>dd^LC1W9 ztRRfoxhId%^fd3IGri7Cp(6aNy2*?V&|n{dYkl--%6>*!`ZXL&Q6ff$0~6fch(J9w z3$b4V4hrf99qGWv#EwU|dC$o-ZYGtS#~DSp?Yq13Dp#Y~2}j6ua5eT!t|)O`tc!4u zu9;W8)a>`1r5>@77HdfNZA~w@MIN&w$Jo7v<$(9{0CoW~beKdAiTk+0x3b_^8f>5k zCrI^R0XIC(({-0kTG)o1c`?Sr%Zk_QSIo|dYpU>cj!Ybgt@Aw=U=U}GiggPt><^Xa zBZk{*?%|Z~`TcS|W>_90J3cOkD|g$-h%X+}XY*Tlh+KTrW8em6v|8EAqb=q~`HZlk zYE1SCi{a>Bb8Bl&#ayzLM^sMFBO_p%WiyT&^fG3pg^`8V${-U$p@y^7xjM;Lqbo`= zd);iju}YJn0%(#p!+>v{XL(+SkfT0F+}v4|hXSPx+H0sF|8%2!?$vhO>Y$vR+5>v; z4~I4j-qyCfHhT*OcU_U1 zQKJnWb!Tqk0n5KBmrB}|twv-2>tJ!l^%D}ll@1$#^cFiL?+cu!@G&MjWjbbtZ}NbL zgrb;>Z?V@7nwk_cXDyD2;pb7?xMHBox4Nf1uY<|e5xvihNG{tkPA~^fSdEJiDTB$U z#6N^Y-_cfNb)WvNa{MiHX%$YRI5KV~mD$Mlr~u|h?U{n4`-@?#kEqUwRI2DR4d5>Z zRH>+YDVkepJ_SFOR!ZxsBSOMr!m8NSb;sy!@}EJ`>v`e~3tK19jqGve8!AS0g8zaa zan*#iQ)UhuwnRnE4;1QgdMCFELrmSO{966XcvH^r5QTb4CjF=7&v>IJzErU!^x|y|)qW z-cZZDp-`xoehVEB=+{l?=PA_jW~J^7wHuFXtoH40?yAy5(X&X}Gj~Hv9vJGqwDaJe z3dzZM4D@xzM$!i5P|HdQIkvkLV;J36U(V_R;ls^L4yB6$WtbUiggLv^1OsvcM#6pU zQV+TSi7fBBq2C}-HeF1c?0Nhfn%pVU!nr6#1)7BK~|dRGgm&E|3;M%Fozhi&IhWqfjcgltVg#xDP!0*;qp<@%GE9|{yn zSFslQS6FR180pr}z2wS8r6FyBMRcS2rkvA~4&VK7OGQ6K)>a?J{1`me04+xnpTC#0 zZph3X`3g7rHW9WL5Ll60-z-SJq~AEZ8Oa~pbO44)p-xYFeG=3AWF@(|uGmKfIX!fz zH9cMwO=pPIb`?SKx7ekw0fFx%0k%M=PX!l|pBIq>EoWx-D%?Ek(M03KPkF!x4GMIl z$bxsn$la$Uvxrw3o~)z%fN()Z-)@B%XoA6OIL>Zv*R&M%kZ6{>((pW@o5ezJ>3#GL zu{q%1vXlF6sI3}UN6GyDr6h!mUO!{HVPx}}p}J(*5Tb9M!rkgMjFC0eB3dj*LJ~9A zi2lWSSdD$c_}F!r9Hr0(#pdtTwbV`9*&`sf;QfL4Z6|`iLMV_wQ(1GXi4A}(Z(rm% z{$f3RJGx^-v0gulcJUe9$Syl$(U|I|tZsB|Er2}=$djQEaisfQMVks0W`dTQhjpSC z*1=iCWSgW)ab8GaM*M)QqaVb3VQ)NRlQ~Nl&w3x>A6?_zv7Ju$ua7(+RG{L$!3OmSBo$N#8rFg$QWO}5b@ zIH%SCsTo3y&;W^Nr)=Y7_gw#>!wRvn{OP4Qq-?5Igxx#_E2FBZ`wj0aE1RptjP$Er zh-kk_Ey+}o%cai@0-5^L35F~|SaF^Y-?xtSf_Edx()D>iTa396n>CR2`?UYV9mM>K%2l`8h^{Js1f>;CK_t9@WJAXAK>T z4nzt;uT|ztSUJan^7u8JygJ1`l^0Vm&%!#j%K_8!0T`v)n|Jz{|? zml*Bop!6m^H{p@f&NmXTQv@1;#j-RVI7>t$hlBrf;X6B0l^4C$cpvhv~Pk6Ff5f& z+U-(%u$=7RS`;ezj`Zr_?yaV^e=mXe^qBMIzLR*Za(k`WYm;v+_Wu*&6o}7mKpwnl zrq#I0H$!8^nip(?Ih^8&8A80)Du$B7SfHoqF(dgRDKPjmfuRTT)gnX5tvLxxr~U*- zBlk=T-==~jSn+!I7F>DUz}^WRzoe5(NUKEgyor+(ycA6L~L2w37{ za(6u0-zLkfj=@JVjJJ!Ea!P2$N261nG2{R6_&3m)0B5sEqO6(RFmi02*5_UF6+qUXssFAm!L&+@ha*56y#?7W4b zc*%nhbbKl;zF6Ub9e8vho#J74WMbI2pQ7A+s}q0MaXg4Xsn)3K~$f1!M> zbC*<`2&W@Jal}9QLtSZGcyS&Q)~v+`v`d(#z$!YNZ;`9RxLWH{U+^HGeUtq*fjpk7 zV{{GI)zdMhI*%|jC&h+2i24AwbnecGHCUATJoZrvIm@Ahy#nF4$ccf6ZRj@%3ka#r zE)~U;ie(BlG|QxU??j&DyCar6k)r*0;4r-%ohLuhFxmOKb!)@x{ClKD>9G*-}K95`b z4fQi432{0m|%PX3RYCD)Ir_4>6UZ zuTX{aRlE=3*G6yMC+0~(Ctn*OlOrh-2FDA#o#~|z7x$XvRilFvih|$>*=H*G#+$Y+ z)+{6MQ}EJK_+|Wv=_RlrUmxAt^keS^Ze00RZk)r>YP@Q<&$0~>!o;vzhu$`~$H`{j z`(c+03Fv5E?&XorPtn}QP+GEL%-<&Di3MMb|0E9EM7k$E{hu7-kzDHZ`_Rg?Zb#<3 zpo60APty?7)@8f}FaK@``I%FT^jbI7WtA(kkL{xp4In%T2+KNVbXEQ;kSE;8I9|jw z_z=ypU|TM;=49Q&i_exLCR8<~MK7W`GP~%}%^JJ6d_?SY@e6O-d&2hhB;T=BedafI z2^BoWm~R7{+~E;q8#~)5o#|ts^5}n8zmGkL*0O6%a3tF<4fd6#Nicg3`p_b{-U^T7 zghBfq-J|uB*CmUERY4&-%5KTkL2newkrq!W`IV8QW9%VZutA%a`NzW8pD`~l`cb`j z2U@+s(<>+ItJ?Puv>&I#qk65Vah@HeHx*>PEGIKR`8L}O1VuZ>B4;v|th4t%2Y|j3 zR*bSMa)#d@6ocv%(R{M|Aksb8a2;Com*n!sn1KhE!c)I@UQ;+7&BJq(r%H?dQS-f+ z4c)jjb@4k}w|s9t;|F?cs|}tt)4et#`H|LeqQGE0a&Pn|M^jDBw{tGEW2S9E@JB&^}KT$ z3HLSZS0t>wYHZ4Uw?Y+hcgHdOk1q4_ zhP?!a-)uIpi%2xp#N?uOI~H4aOp?e$K(T4V{x5AiKx;0;L^Et$yuvrrm({T@GZ!=e=GYt30G^hZmH|X8=iq~sIga|P!in=7 zxRDm#Z%H|kSEv6yYl(^yAT$zY3zydoGSjJf!)yzQP%a3&Z@Iyfcr{(qiE+fzo2yYi zy8B7Ay%t2WLR2JKo2vk|;2f3XeI?eLW@{+%sHRW8gs0CC9J2Ov?`2ImFHit{VO@~< z-1%q=c9iQ@Vn>+(IABv3Q75y}?WpHxBeXh^a~n@D>2B zzbIjK=cH7gXO~Z8)rPQ@xKi3tt8=!&TWEv#@m33iq|F8i3T&p7d55{?0Q%|+*Wcq% zyV?ETKpsM8qS)zgDPw|{LdnIE;MrV#VwsSMar1eMn`KD8w%|p?X4V(xBSWj2`u0SZ zOrrp9sni?pyvix<9J0b{DAu8AMUcueX}adExUrI1=aNt_)qMi!D3VKcH{Z(hT8C{4 zbEy9^1M^z)0OWW7i8c+3vzbHf2P7|vSPNr3CLufc(3~t zgBV+|)9eRCvS=mcJ$X7o_Xx5Px~wAG^8q9LymCZGc?kFh=7uS1`+;_8gqJV|ErCRvBK3C^7%`xGGV)iMm%>H$`pXc>4yvJ}d zGq)aMW9)_H*tit_W2@M)hMuc^qUaZvH^GBbSIJ`5nSGd1Dnq_AGznnwOIH%0)@og{ zF^7xCs&Wwx`Fw84tC~@SNaU>_G7qzonw&*#v8e|m*nx~0H)So}HEb)`12F?UO2K-Z2f5(kdHba>**;ohh32iW;c%`Xo_%-C4C;Sm?#hsz-X5*NyYL;JG z4|oT#Pk_1(^)hJD<>0`3GHy4jeeFRwLEH}B9?w)wea6*90i$Ssa3)XzA5`S5077|% zk*nvv)&2eQO>Zua3vv001MiR15&uco_dSVeBkFlhiLrgWhoh+EK#QJZ(iE<?WtVb=c$Bh&Xth*;^}ew+V#YGcG{@u^xt$?uQ#Qy-ZjYp1v-L*oNy6?X>8t zjhnU=>09LU1eDVUQJ~SL{J~vImh&8iT721rMeVYX&%V++jqx64oeel}^z72dW?RoJ zMttT!jQj-=r5KG>!0z!i;DSOCv#J!^(W};I3!+wsQt+?^X0+2QMQ#@<@qT3IK@p<6$iP{?NiKF$ zIhpvFt7|^ANrbZ>7D=3Me+ym){6tjRyKLn;9C2zNZ?tiW_RHo|gH@qk)la_cS%s{e zs}>g|eQ@O(gwH!BVmtoQH-p0=Qd!_mG=N_u+U@V*q=}d|PZ_Sl@xk+mt4W8rW7?u! z?YgWRGX}`NcZnnTcXYBhUQ9cfNgF&~y<;)(e$+puX$P*~C&aA0YQcvBw;#<0>rnm6 z8g;l`N3hTdsf8lXTdWJRjZiJp70ztNi6>*Rx(Jm8;E z0I2`iH{|BHPaG2}E=;Jd&>RgU{qLRiP$yOad;u4A-uMm(veF6q$vvI^Tf_cdVCFQj zPx9do?EEE3zJ)-py2xvvXVIdL{)@;(PGLraJ{;ahB)cSF{-OcKm69Nagt7Tt17@F; zc`Fj#Wlz=vGN8taJ-S7Al}}*}^z2$2Ejo<-9waE0FwL`q55u;cL;g90d&WUGVSRdh zF+=$Gu9)|qIyA79VJw#F2u-FQ)adFOV?41AuR807zUH!FU0Ss%1{u#Ao$?eVBXszSwc9`cX1#5zOXnP*eSo#3RuDRAGdB zm#k%TLc;vIO2@xX0BdtTd-!j%DJMCZ*$VDIO9dut?p3M0uMTm385t3+*%;Qw>^GN; z_FKG=Z>E9hN%wQE7}X4>g0G-P^AWf>kQBmk030@OqD0nX;J~vZBxZ&(t!~LFpNaZ3 z)V7PrIp)BI=I0xZ(}$Tdo`3hVxkW)76BDCFpWo zXLF6eIlDN?Nz}vcR1#6HTD6wae4KrT|7jUc&@z!9_fFK+SSuEy!!3Bbxx&p7$mO#Y z7TR>*6GMI#FgY>R!Z%n(7`E2A;o|rZbzgqH71M44+|qrR7F$V@Zu{GhnOr~YwPmR5a zE#c|f;YEG-At@2hAm~@hs*g9o_f`NCn zZ-f)}HVWdp*2*asCJ;Z*t0H$8aG^EGUAu;BItH-T_bK89diCXGq?Hf9{^i2WQ>9Cf z*T z)EM5nUqmZzJ}yO7x4;7MZ)cuAR;8D`1^XCv)QlXtgsmqQ6=7^jD4JFZa_&i`9ixk5r zNJV*Oa7uWlX(~ck7*CF^KL4>q7P$$qEL4yS%Q}wt_$B|nrLNAOrg#>)HCtZ3iWWaZIk3JR_mstd6oG=B;bHy|XnTtRLyiUr+MN_Dos*b^qNis9p8qJw5lYtCr(v zu#Kxf{}@b7xR>KIX**Nei*!e3#%i}x0&J(dBx;A;QO`s|mmnscGEMB*h+)u`={Tg* znlj!9HxRs6tt50rQ5%OEd-rN?< z5WIAk+-=k48 z>lQ1@%bKYfXTtw9<*u2JNrkVoh98BH0VQUN=q(%j)ocfRWH)3x&6;oyuJiV&Bv0hL zV7FwFrZ%Gv%#cyV?kGYR#&i!+AvqisOr}||@}8GbJyN{5S08!GC0TFpbnR)}fYUhf zmh|jaCaa_-b1Wa81)Q=83$oIn&}BGwjT3X>?`QZAizjA(Uu@+RO=bp-&6fT7Ugiq9 z2+W!;2aA5;5UQGG73Z*ls4RY2|MpvG7m-dO!z^pN&-16@bW%N^L|}`uWGE2_-L5(| z)-CL%Icn0Gjc*IZ(2@w&o67!nYJK@Fhw(-MraE&?Ortt?LBJamX9Sv`e(l#=3m5df z8|1p|9rV*!n6fFrVB7|%X7p12r-0#eMLnwh?aimw1M~mXZ8BAJaXpfss`nSH$p+kb z{GIZ7Gv6Whie%>4AIO9FwJxD_uU2_?OQ3JLb=gDsE!E2{PNyk5P^`t0_)>}rOTYuS z1-bwOt_1sUo#6Q*D>AjU>-{MaM)>Yc(&c**t;!*FS8O8n&11c_zz_ywdjig^fp+N z{~^d&r5<|F`bvcpXL+h>ZLT4D9^&XoK{t$5MYdpThbY?pTXj^u9dXg|PHkYhCOuHb zsA3SUXPg)>G_dsLQx8V{RczIfXAA1B>;2lHIbzOiq0$rK1y zzP#WiVrf07u+pMeBoRk*w7dg`q9az~=A_g`(x{rdKK*Y*N zYtsNv@XhF-zE-Q_yzQI+s|8q!zEIez8W*=OiW5 z$e1aedm8TXkXzc)a{5Rc48h=hS#Qk$l`ZZ;=1}libt6 zTwvFpgpnP|+dKeK{g9bn5K;`u=MZuLi^Rh^k@1-8XV8SB;|AKyUW-2|w7{i=kC{cY|DY2OvU7oIDHhd(nJ=>m?|HC;1$-?*6k%iBg@5I>T zChfouRhGoe>Dmw^+S*Yv{Aq(KZRSR3XtF0w6S8@QmQaWEUh5tg8;Steo)cclgQ~VG zQCor@<0@0XPU1#hY0+Xz~M@5U3&9J-?J~7kcob-p+LYJFhUPa#QHf!+?{!p{)lH%&D=PzLe77djNVY$L2R#Vl~IM1hT$6T zHlv<*Yo{u~gG3gi2iX6k0_S9)v_#XyYli)uMHO9-AD+C`clI>C^mmZiftg8D2kY3k&qSA$9LGak)rGh>ZfzbqF*{l*e;{wryN zE@;o~X03ljA(E@K+80bH>|*hP+J!Ww zFf1tDhY7N9Fqx<fU#{#k^ZT5S9T&eC#n zx916GTlK*3i9&A1?-m#SIbYbpg&4C)L0Tjk0)csa#_=}@IA%Ny#d3D6TNG-##qEl~ zxQq|GzkC2!^zb#7lpqa?vvcDeN@7aSxMBAV#bOFW9s-$CI4b0|{us(4Ag8(*Wov`p=o@nr*K{+0m)n(xvpbv|63or+jEWo~F zGt*~3!k|om!=cv4^w&-%DPvo4qO_qr?Op7g4Id!%a>?%V%4iQcQ7!>`&2O<&YkrE= zrL%rgs9xIM+KZt$qoaXi_%wdfl&-^-!RP3ApcJ>YNlXQE+@w%R1yq$oYEcC&p=aw> zs0#86)$D8D$~lw_6?<1+JbRZJJLYy!^`Yx*wrS$YNom&vja35_>UosgzzNCQudqID zQIc_iKAHDS=^=#}E*D~I6U#A)FfVDHrBl=Qm%Y%f@7M@EGwc{XZ%FRkgQeY-6%SqA zwxq%xf5+_q3}mBsI#D!ts&E;Rs>Ybg0s#8fu9Zh$GIK-3|AG#^8XLlxOVU(lOt@~F zt4%B0jT(QEi^bNk!neuYZI7FThuoQayPoqNU3Hzjpt<9Z0U;QonK3aIdFJuhMbs?? zRhn0jc+h6v8aK_)49G>9RbIOiv-oAZv7+_Bj>*71#!$x~vRBbzaPxA(^YWqdK?1!L z#7KH@{5D*d^=VsbH$M=Js&?r>-c5qyPYHII$PIdwp!j1lKt3et^Fzy0^O{PE&_JLG z|Ked;;Su+q7K@x_^{X@Lw6^Ehtrj_iXm`j#J87|Uj;|~?lycVWo9%uFYT^T%Km9M; z;BOt7{Z=QPcV}3`bnN3?&lY+_ntzz&(wHE;=rA_%(9>9P^i(h~GduY_y3U6FM39ij z_!X3X?)J0dlRl~Z41>~7yBE4;jUkcwk(n!V2g_6Ei{}F0Sp=_#tjjgWRRisO3+-ae z7)!;aanlTS94$7ZOtM>;a|mEbFX>}T3uM<~wx~fRZ8etKvfW7%HKmhd=QJFr?tw55 zMSBGa(n84H+`>51`@Q#3ww81hd4nh<#c|3vWTvN!HJ2tOb}l@*5X48K?vc}*)`JuB z`6*Xis$cvMmR(g)KspcF#S9oq+ypcIkb;WmMeD>E1|q5_EBJXBT6k6~F7q{^gRM=L z3nBfJpqw~ETRb84>p^fww(odlU=a*P<*-2KdE8G9Ol$9Y2f_6Xvb+$;?=7_!2W|0) z)D}f3b=T8>8(>OV4*@YUZqKchFyG0Jt>-GvGA06BZXTJ9Hg8=+qCVd@IAsK%7FoNi zpyJ_`Ha5RLjKWmg1;CrFOTSTnw!{}3Wx7FG>+SW?;~^YSIn)xdvmo%n6@*EF>REpN zib+%7ScV0QX{wsAixzu7C$-%OXik*8`?XSIWb?zR@prCTtD$?wVson3?XDX(b}v#W z+<*N|A$Vp=CDd>He@=IN=Xsn)nw1nrz1okMg0Ou7NxSAD`vJw)^ojcv;@(%$mu!XL zm?XL!3L|CQHtVeMK34Q;#hP#HNdhBlP&WL@mbdC7$%I{RER<|LWFI{^K_Tur6x4H{ zBOkESfmDR2C3<`hMc1Usl}5;dn(4iuATZ&d`kKT$Z_ zQf5>Z28S{3QA6qHVNj@rpy`hDyj%6z56KNMsxt>-Qe)P0KU_qOwOJk$|Hb5-CVkJY zUt|0Cz=&FA*w*2J3o@f|D2|6dJGP->ema3?4R(+2_}O}-cuu~}Mu1^pZgkpjE&2&z zEqYMU7dc6`(s_jePqr%oo72dcCtLPJi)nM?nyBl)jF)rZz@MCqHxJyc!T)@AGiEEz z*epB4W8CDw@XF>rdDx-3>l|SmQW||fz6+;o@{yRXwehhveQu}al`zGoh!J)b$+vsx zag$#AyU9Gs1&9QPzG`u)6DVA`T^l=%1jI#X9)1{U4`8I_qor%EbTG?0?D3Mk9U?(b z&EQCzVszSL<(^j(W=Gk)Eu+YGl~|q>YE4&y&H{PHcI|DbP&WOrUSR+|?ep!c@$AOV zwp9u0jb?*wtDw4q|LS+7UTXX(gpz&@iA)Pt#UJ}Z;HN^KD+`}YVM&!G+N&%x)1Pc% zli!qh8>Mz_=&g`C?X@-gm=YUTm7IT_oPLp8WWoyA++JCp&qh`>-DohA57*FFd6kVZn9kXX@F}wA2{66YB#HItSBO;_j9+5 zG3376wdnJa6Jc2M`3UBRw1z<5>AUNnMo96oYaQ65Kv&_D(&OOstlfq=x^i=4iVMm?^F?~AC# zR~(JLX$#>Fjw{M`7FdX>;dJmhVbOAy69aP|CM^3ao=MbnzT8-yy3NLl0b|?m*vJBh zx}K%SCh8qY!Bs9te!DQ@C@_u`rLWJO88>Au%GDQAh}#|=>i0yOeWupzLzG#F!y-8^ zjHP9@AsR9hvVqo~fR8ZWZ29{UWNrkaP9%VTVwXc5I)XJqDG*HYZ=7Jv?f1YE{A-Ti2k*{*Ay5GJPz3PYazx+eUqDUbb%`6s*1NHjGlY*FmwMKAbRsznt~YW%&gZ{0D@Z#=H&eOtPEWW+z!FoX z_Hx+sr{Q_CJ695mZSQq)nH9OCEONztD*aDq$Cbm~PomB0sksyVrgl?k%{po>aYp z+t9}zfwq;2zV2vr6<~dh7)8sGL4NC0n>&-Aw_cDjVxW@aTLa%FN|N3+C$3MdZ7)8L zZTBFy{0oeEU(0fr#=63KRb%5ku^YXetx2%>?ludlZ<$SZUKNa?@^8pyW23Bcx8Nd# zV*`nPpR?||KpwT0c;9DBSWU6u+DxC8u;l-@dj?fH^gqc>?6FmRz@=Ov{n;cu2=x2hRq`f0X^gz&I^E7eo zN;i*d8yF*;J`YZ~3EhR#c<3ITd-f13e~uMpGeNEdpnzee-``Bp6kWUkwlQi9G4Ra* zl0-mP5L4TThmC?!4KRYJTqjpsW)tZ#kuDJ%;O;vueucqw4imsV%E;2{p}qfA`muBS z6`TyE?vs2@-TCi8eHe+{t@BNW50_WUk!%VNze~;gg8@h#Z`dxJ8wW1m#Te)0nx7}8 zN^29*cW>4=-ww^lxCb%!rx>I;8_bGgDX|%4T-`X&CA{~Ba4!1zegK8@ z?-BA;GY=n-LO`#r?Q0rMQ#uyRUTQtwPcY_wl?e?#} zZl^F9))}KxVF)?MHRG&e(B;h@zo5jQ&+mc0Je&YCKa5=IYr0%+fbPhCw8PR-e4?g&7EH6VMIx z{l~_p9bDik)#Zk4@n1oz-Vlm{z>SPDcN3Pr+t_M02vggFX#-Q5Ov}~MGnL-uv+X6e zfXJ~CJXndk$_>9jnU>elIpf$}gRxlgxEn7n!s5!@nNQ@IOr!k>T5-zS9xR-8J&ddm zmMG@yN{7-?X-t7=W7NShCh9XF%%=X_0z4sR>JBRZ-{W*em=)oB<{b8@^M8*jb!T^E zo-1V5;AQ34)7p0Z-5w{tw-+j6hpXc0wy&?Uq)ks@ub_f;pglyMIhkh_6+N1g-8gYE z=Xp>2=&qF=%F9EQM`+e zlYNiQh+rIw!wze+NnT5C@#ODH;(|1-%|NWeFM+2B0;6PR;bz0dR9B2P%S|+H8*%is z$a;p-a7|6FO2j!nlGJhGRMr(ipnUgb64BM~jPC#0t{$Mdh{^Jrvw}GPTh2D{&XMIQ z>pE8||9`R%J(47n+Qi4ZNuv&h^`g13gVKJU)7KnfyCQ9E|2XBJ*o|cjt1+Kp>v8T3n)WFK(ow2#9Lvc~Q1J$)Nt zw1T$#l#?;#2KngZa|kh%ymNN=XTi-u^W~TtD#e6+Kd9y_ z-@IO`KAU-YCies$5CcPU4~}4f?Y(F?8Q_`SA{@Y%?Eit(&*Z7es7Rjldg6_s1T^5^PFXj=RV?gKse)?jWqKa?6*a>)H z?Gy)T>fX_hY2G{!p#ZY|3Rh$6uzZAmk`>0)FEFkih7hcYjFMccGu~?y$bi%kfzpYe55(>oPmqiU=Daw-ZBK>f%yXJ zkR$N*{yyue1Z`YD#_GBM_jy)r`huc**#FNw@iHFT=H0_7N=s;%&H5YXC3Bu|J07e6 z<-*$-Bi5_EjhjFzQpuYzZ0RV%md?__l})TQSV?;PytkLW#~7M6#xA=5$#c+;#GF zQ(t@KdMs-5V*B1qHAtlVYOn|$mT-NS08mTmp?Z8EF}v{7qF z`Jms{{$npvs@q{+KhI6nt$PjE`p$({RX;B~6XWa}%+4+v=Z*Q(3rZRV4Y_cp8zvK^ zMUt7>&?T)d$Sy&{i(tU!I3#HkX2=?Tg=dppZTZ>StJHYA4!v~KhfJTIclX%0H&OSY zAuuqE7sgnDf*nJAW&*OPM5cqYOnu^>7@(o#zdwXYGv$HKQmu{Eg&@^P^rxW%F`w#aTia~ zihW^(c*|ZJx=jlolLQzm9Ji^<{nlx{PueNv&a*>(lNDaPdwF}ia!NhEI(v71K(7A$ zJNE|*lKJht3A+8J*MOkL|Upx1OZo2bG|}2Ifsbg?mGxtAw+OV$kQp z889(v5nEburzw5SM>?*Er^`M5MN0LZoXgb<>6R>Qmm7+l4#M{L}?pQ}*>kPG|W zsK@O-fn}XSJqR7pZO&=&w&V6kp1vsL!fEgR8}|?>w>aIPMUNRy6^M*MO}h|^%S?V!I^CwQWfh4_^_SZfo`;~`2b()SzQo)Fg-grOu^+HoYj{!^E@t@GkLRmN z)lH#!-(@#If$E%O{?QviuIp>#Yk00AbuHW2{O|?GLb)H#jC<$>(Wum}S=(4KI50tS zC;EPb=R_t9G82Ir3D!c;=~u>1kdk=1XI7L5{>lNpP4ra62L6 z{lIlZa(S>q@p26y%$9l6n4nfWFilr~1~jhb+F}@a69ITt~!2F1+jZDX7s$# z2bfFqcg_3Q--5r;*rbffI_omoO{%<@(ym7BR0#7l#Rsq3LD*{Rk$Gnr-njYewp}gH zY8yHXeVbCg0pm`}pbrMsu_4+v3-&fY1Vq_tw4IXOaiuU}Y9%eU;O)5l!U|m13$k>1 zW}~?)dJw9epxqAp7mjl?jvkcH!nOLf;6li=oY!_|z&2dnZ)HE%QM~jB%doscqaIOw zkK;T_YH3SR@c8rm#Ex9Jc%}30^yuE+6s7}|wVNBg1MJyug|+8n=(B<1U9QGv!2{RD zlZgp`$T~#amLhGmtMlcxp_S5R?G%CZNGhwTS)~0^}q0!3$aVv@NJm> zMk2^VY6Jdl|QFUKZqM|h5=X(z9p@DFXPN=bnnbsH%a-bIVMkep~TNC1V@rY&IXHv zD>}zzUyxJTqMQ^O$U&0YD;=u0QUa{r%8B{J?hD-PEDm+wa&aN7l%q8>y->qxo-FOy z0QUR28uWj?fJ=_df5t2c;4AI1QOvbi1{8{A1p!q0nmx?>=R;!?V*P^E!iJYjSB8>)6q-_xl8AJY3RTSd;0b7M2~*D$i)3U+&CUq4Q_1WZwQP~I&} zP?XIuv1pV%$_0{d0496IRaorrtuF~EtFZl+%Gy2un6TO-HqG(p5;b_M- zUO;7_-^ryOtxZ_C(lU1L197o$ET-aQKPQQrGVEvA?2B{bc;f$S+&Z!ubKG>7>lk25 z>%m@V1QmC^R_`u3d{Xw@I$GPsb-vtf7P-tvmUn`t{!+29@H5-1T?^Miqc(;Yu z>PL%DIk{|Deqk-{N#~lenY`EH4@K`{yAQ|M3AjPanxSMv-%@f`ad#y!_6ow^qKlk2 z$(dtKUs_In(DU>{%~+t6(G_Bt#7&x*2GS+WSh)9O{RGd_{V}mh) zot)rVt6vp|&f^q?^{pjaYMw89_&>2)V(7;+ZTvSgT{>#F_v$Z+C5Jr{tHHSCZF>BT zOoSuY_3eC6>YoRU)YX{1*JS%eh?|sj@Gd*8#aMe{jNZXHD{F187&-~uG`9Ldrqzj> ztAWcM=o0_4L8w_n>x(b89V(l1fySFQ%`HdA&aO;iCT!JCYTek0%|v=-`#*2cjq_B% zHq-Bi-s9c6=S9)4wus)Nte0LP2LDqysP#&cTJsJhUotH zjdK*L#X`Sa?k@}0cl~5B6t`jC16oyEalwzo3 z8=7H(Nj1PMi*v>h3g-st-Gd_;qutJQ5;|-3cYwmIyVI_CdU1J912%OE3Fuh_IHgsz z*&L~yWH~8xK}I|Le(GtHe=fh_%WRulGsYf(K>DHcTz9H}kInEiB9_So?y9bG4qs%v zV$M#c>1>OG%r<^u&cN^Tn}|5~D_tH!SYFj08+C!s^FlIphV{~SoqFEn!KmiV6k>zc z0NtUe=*8GNN@D!v{B2C6+x|zSVdw|5GBW?-K&bP6z@fL@?DoGM#{2(q_ttMweO>r4 zN=hoJq)0c2Lr4h-2uOE>$N)o!bO;Ej2#EC1Al(cM-O>)-DKYfWU49Sxd7jVn#vkxr z?{&=&#LPKo_FjAKweNMWdl_&;xc}eY3}mI6MXE1P)&L38t65(bfFWE;4}@~MqHbwn z@Jy4FrN6|gioemr?f$tBus+?lA`~u+v!aRo>uw$HNuck%4sF=lj%(!c{DSD;_5WAy z`1TWa0OY$?7YC|8z1=)mGOAOTmv|^ZzmIR+{yzQOf1e(ZA__mhg2jOx>cTxSASbCG z<~<@zC+}&%`KlsR>WA^R-2Z&gC8FSgJ`ulE{umC{PNaXFzeR<7sGkDAxIZdta z^A|M!@^=PrEy^Iz`ir~v^2E>Ix^_dm0>1<tWo&6K|RSos7AjI3GI3_(6nzR?Frd zkhA_wn66PXO-2 z-@OnT2nTZa{oDs*W#=|khZ~LnTMTEtam95dzQI{ZDZ-P}>Ov%-9mM zsk5|eJoX}Bz|q-#T?|~s^?8Rqc)0axJ8fFDYt!NHaQ{vJ-unObKgC&)X0YEiyskR@ zl6LeXp0YBZ+vL zetW3#2JrNEROeMeQh(-?t^Lo^{4-GxAe6i5Awa?6M9RAjphT9nok@x8qX)>rS!qBC zsrbh)TIm1Ri#?Y{U=>Xb-5y5BCOro}dFcSWeFM3%3X$u-$Jf8az&~H%KB!7yee(xu z0JT3}t$bk#xcIS}oKFCLzU@Ko6iuz~`wwdOuMu?RSrdxOy8VU~c{+~-px*;(Qn-8; za_&ku+uCt9f-mm3rrh88;HZ=0>#G($^Q>|__fOIUxLx)-P_ET#1e_op#H;Lm3jn1u z0IIU9#^tQ7$F~i4-MbpGyZ6r`{#%o9yNK&vD47Gfp2N-bIb*=!bolxmO5O`r(AGUW zcK7{$K9%qD7q)->uhR#5-FX5wr;|3lT`!3cz+67k#`STc3*e5^Pf2b2r-bh>c6PgG zxdLfqpX*h(%_a`(Rllmn5~@*K6`p?k?DwZGo%@g*$jrbz`}M75* zh-ca$^v++GEEixv0ouJR)4SgXh#Gyz?W=hF2Y>>c_;?xMkeMqm81`(8ldE9&ffl8q znL^c7cWZ6B@)LKvz_%)1`Cfk$+_(3r3IPssr^rk%>7&;Met=PeM4eElPW48YI?B`|~pdlj>xmTa(4F!-BqxzlCBCPh;R>c16U4Z8D;d<9PFnMH8b$J4? z)jiT+RSwH<0er`y0g6jKEK#QLpIW!uMxnRhLIK+=V2wafA%I-T@2btnAb=$Su%WdL zF$rg}(*TB4iG5G1;PV(;FI29;5RpIJ;B_2k0H7gu8vuT2)vGpJ@7g!8{&F*0vu|;g z@9d-pAQW;P-tC@VX|r`6Lli)5>>ref@=sCC?HE9C0SIri6&W0%dQuRY*e(OVg{(G!=$r(dn)raM23J_a-%|WU=AZE5e{S;<08qvY_4p49 zvv1KEJp9JR*Zx5IVeUYWg@Ebr6j}T`MWi5?>Z&47r^_l&6GDiq}&KEFA-<<>npuG?)`f`#KwpUUNN!9B{`lP8QXPv zg8%uH)HguyDg3ICR@$KhKrbk`gyYBA_t*e#7yTdA93T>(`a8h4yZr|>jUrDI5>uD0 zopf@X-n92n3pSxwQUPi9n1(7(I!^x93uycWDsMxb$JV2xA?rX0Qg9F0@KDkCI?uFQ z>H)5#!!pUppE74jCB47r(%b8=+@{J~!GUCyEIMgaZrkZ#Y3s9063i!;+H-*5&!>=k zhV6f)+Cszzr_hax*}T(p1AvYDl_eb~Vco?@-sxQs6Gg^0)OH%G=X1 zfbjgomi%MLt%cUFcV^;99GFcJ%?t$39aUigygsvucTer-Y#!XC!>INDQ~?4i@BjKr z_B?Cxg|Uj&j`$1n#B{KY^Wv8?T#shM(QOr7CGGUv9Vlk#&&K@c9@sz{7uf12wj8Jd zWTcC=-?q4=H5ibmXWkmpXBj(x^fxX%`S;k%H2TcMfu{i77Ci3Y^6t5VDlCH4XEgur zJcT<&K(tCq&dA)~V6`{okFngkyYInJW_Rs&=y`Gd^RS>7e`2;Q+~nH4TI1X{85Nba zf1(zy|1OXG3;e zi4D5cyj-w?3+{{{jIPlP@)v-p;C41Gz<+_LTcG6o=Ge1U4IuRS%`M~QrON*V+o4Y5 zF~z4}ese}|QjAfk2%zbj)C5AAIYzashnmja$6AU)If5xhHMD1iTcm-vyw8o6M6sa# z1MLM$nbK>$Y(?G(?w_iukdgzNamao1n;>H$Jj%UkVWTvD@SX7jRo0p-6~pdG*ECvE z^Ao2k0Lx(~>Cwd`^b?YVEBl?u$MpGC@4%9L>kggaksEMMqz@&XZNI=x3B`g3M`L%k z9&ML5q6yXEo*M-GCQlobtJJjJp)*kb=YP=}!2ep^&69j74cQlFrajXX%bYa175xbt zV#PU3$;maiThnX=&%Q|1Bt~LPVt_6+PUpcT^WUgWO}7-&b}DxBgctlr4Y#6C%;2}( zhoD0nW#UI0Jb--1#m&#EafgWm``>CXmE-1QRNI?#-fZ+;_-#B=9sBbI;xjq*yW*CQ z1~cPw>V97#nzKz=tF{X?_jpHV|t(t%Xy$7bfHnOIL^NoBh!npgFsF=XV>R z7D?M9lAFd#b2?(<4crSnd3teyK|5p*omgxQIp3w3%CQ~MfK1-&z!q*5TL^Avvh#=z zSeZ*>qi9}oM%LaGV_1_v(|mB!j$a^7eynWg;3gx4TtNO@!Bs^YpdSAz_fela#t1*6 z^d#WJ-o6FDEyCIgd(i40ZC%=?HhF{GBQI<}UDLhU2-eZC0o^>m_P_rHyt=6|NE-k} zx+MmcknQ=^Wf<;CxTly}-UA+IAM(dL{COBi+hqg*I-Yg{l2Sqb+;=eU2$Tk*0S6Py zXYE?Y00s#FZ#u`y8y%LuatNSIcl-1kegmMF} z#l%(`$dLhO0!7+d4nW?>{^H%t|9<+`n8)ru1t7v4&wJ?WjT_;O*^xW*qvAv>Jhq&a zz|pw5<+2;O1qj5kOk6u}{`<&PN_bA7%SeDH_M>o$zcfcc>KgC@iDCRiqds)_lUemw zy8t?1Fd^1k#he68BHkY4#QRnz@pjxH(2K}hX}S_{$oP9) zI@h8Wy6OnKV;k7xk^c_)tSa-S0i@q)fp!tSxIVjRVhb+ITcGXT*1a@Bp zsN5EgFQuEf&;CDd`2)DMHU6Iu7A%<5+t3|#4MQr!Q#YoO6W^(h_Ptc`i>k9Lje z0lE}>4`?U-8P~SuvF-X>)ZGoOF}HHD*tnJ7n2@Vo!=q|L_5Fdd2W$T z%sQGEyGfWaP>LlWQTGS31z|TP7g`(8IUMIqmnJ5j-?53)2j95nx1u7mle_QI-unhs7I&jE%=pX~GIw@Mx5gTZeTJ5^Zl~`yJptQY#E3lg zGTgPP*t$f!Sa(%Q^t!y$7x+_mfgnI^aACk+GQ!il)>$o%)+haT0U#TH@=8 zese1{2=ntz=S~dH`lUG&!e~tHMP3}a?_D*h2!~D)7Jms1y*?M2++8rWva6PGH|}}o zRp2}upOwGS^wyx?)zLD`XErye&reXc{aWC|P*2-DRa`?JsbBs!ESw^w^{??)-TLP!zlu^g&{a=}QFR7_Y0m zELTOt{O9#+jQtmZFR9Im6SA{VDM7E5uQ-4cS{)mxf*cIfD^KA zBa_G~e=?o_zWMft7mAn-)Z&P^Cr)?&y~_XP9sjxX|9Ae_poiIcLP0wac^V3F7)+WT z5|nFa+kL-DKueP*vGW>PpvYTi;)&4lPu%u)_N;+LX|wdOh#%yCHAZt#~QM#rXp z?(!t3@oY~#wANCq_u5i=#@SMKs8oj<`ntKe>_lUJ=tJd8TdJK?Xi(x(E||e>ZD-kV zQAjTrIJC(V?X?T9?R5~j@b+$(dpi6)MN55(D%xd7Jv7R1D5=1wHf3{wx8PEf>0x1G zX}KL9zTCU_87GF~k^!oB9NrvkIC`G%omwl6Jgup5t}!&cY4l8n^a;RuuV=t8pywHR z!I8Ti@^p9i#Q&l+$5m_1?cB|++M`AzTGoE;=+eyTe9ziEbG81v#KSA&WU5kS*I&cU z{v0i*p``x7v->&M1cW-|p@pA+DH&ZHFJBtgHMg9V6x$YrjKy+Q?>82;DtspPz)OO| zx2j9bE*6dC(<|(_zfaYy(dsl|H%q<)O->a}N`iv4t_|-sqp;-LQ$f zGh~QsyV{q)vBMfJvHE3EZ@6?h*|@^IKZ_fB&bS278}3xI`U{#2NMM{=6Zk#%O7kYe zY|=5X7Sz)=*}_EaFKQ@xi+XWW85!R+o$F%1D2U!z^37iso;57&wKU7ll!U;A%Juqx z!0TNe7(5h7y|MHDFdb46c3E8KdMQ7p&FMF6Q8(1%yx24#vn8j+KG%5crhI+r11%~+ z&9}6>&rtfMi5FB1?KrP&3YNUDKGf_6<6Zc>2}365Px7yGtbR{@6rW-$khd>?2!_?t zD0#M%7an;<_{Hvt?0PcQH>jqpPu6~pi@h7%rcjs8w2*P|ym+JQ)Ianz4Usch@kKv7 zD9wxCE2|dM;BK=YwYIcrm7!6?>4{+RaMP?-EW#lFRhi+~v1^q1akg|=Uj_!pkih6n z+R0cs=B(1deJna#85l9Ca^V%39|&KQRWxU=i3q`#`|6ZkBCJZy24~8)DqMRjq2-c6 zt2Va4b=z{$=4m7McoHH1(7-RZeP(x`yiy=RWBUo^Qng&k&-<>ZBQ@33`WMJ2@z|=d@v5Y ztC9QW=g8cgBIrBEK*@2xLtgpRtZp)C7ZFhE>{r~WK1o5lu=jl~_4O`rzn5&IV#PDdk`@}tGe*gp{3bqMG~q+Y?L|ZR zpirxU5xbK9BAg7`(Z@XM1f(Rv0T&pjCvS+zKhNhmd-Qr2W&FB?jGuq zk{LAL#~Wkp!Cg=(G6?R?%51{%=A_j(Vhn2}`Ei3rIMjS?YTJ`l(ksODBp&9Xk+nNmSqKExRItLId;^6*)w+e+(<-%x>TkXy84$< zWnWxn=sae~L+7;H*uWE^=Gog?ILEz753mQzy*7sGj#@+VRs(cZA zACkpccOajK@@VBh6K#kc2xTwsaj`yLMkrr$$CML{aaV_u7}H#k?72kt1-FmBuvz1+ zrxtzU(8vG#*XN6?m=sz!iC1f9&QotvWt6(U+iLeV*Q^OXLx&iy6*MW8@LqEC-Lx1M z>x|s6T7#@&57tgrM0`)Sua(vX9Su&M2w3~i-wPIY9CF70OA(rj5Bt(X+9~T5J8ZR3cw2blHK0#p1^j2a{e4M}!401Wb4+sL~~JSBRyE zxpMI*;&0_XSuu`78)E58Sr`wu*b2pr>KVmvCGRmfGdgQD%%gmfrnTU%+5F>^AA70R zif4UtjeW!YGZ#x9O#S?!lCRUnKFF)^C`;j`1?MJ(y+MP)=<%op@;q8{quiy`*}BV< zq}geQa2KWoYsP#%u^r>y^9&zaWUm`X5^a}uCj(3FCHPmx{S+-`%LQlNAQ|NysqZj_ zY{tu^;03&>P6xguC8rBn4-$=$;H8qZ-3GMQ#sT`V8>MEdgfaeejmi`po;4oaQC8FX zDq4s zL2U@)IG4}TsQp^czrW$sm#utbYO8ig!T6#Q&&f`c==*Zdnt}2ViB4&+)^Bh~DoH3zmtmFHm{4dYzCUo}3SJ z@&Za4l~o>Pf1;5de>qHPCZylvM-+1c=TTvZmU3^h%pHkO=Ui!M;EIyRPtCdi8as!6 z!P|_w6EGU3F0@^2IM1ucpZc zJ(MOtMD}DLM{TseO9_!lBUFC5DBQS(+DW`(8LvD|GqKL!+*QK!wY011UV&5- z=2!Oe@?ryZ=*LoxOVe#TD=RzZ-`5mnPV+}7OIOn=>uWU#JB^XCO8a zC$!9hC#f13;jQ7Z2x2*dhSWlwWRZASsV2*Ii`ki?R{yKjbt|vKj3cT-;+HYB#y$4j zqV})FIcbTQqxBj!8&J!^>2PmglyzD?2ZFbEQNy4xb(R$Ml?d_(bj%`b-}7#^@7E7b z+{m@?`bP}n$yRc?!AN19aA-SJjPmASe8Fj(lzUNHL~YmXrna-Nwv@2+EsnLgX9VQK zi3dU(US;JSKf*)z94Pbf6}y{zKsoW@XNn7%sY3>&;JXeh2 zNZ*g#p)_Ar%8&YHzA}YIeG-qdiS}5g$!Bt=y36#uDOt>S{%}j~d1QlRD>fZ&@Z;{0 zsB!F#d%0J`3=;|)misGA3jAm$!YYLG@9(CfM5IS&$DA-hARB8YLo2q0aRr6wj+`9O z8y$6LgRHlU<>HiSBh zvz|6bUgToS9D1!@{_wUR5!1#X14qD;uWbN*AMVjT^!*=W#*wu43K&l2x3 zt!2onVo|QSn!DGRm9Y}ZKlw}dvE489ze%m%`Mt-+dsZL>hfs$>N9@Ajk5ZcA-`=e1 zp4N(;86295mHC8GA^rpxbB+#*{>1}TOz6&J(&=s+~SCkyH5|s)6bd(H7)YVo#zaA{b z93w)PU$q^Xo=+ejCgX1R+Rf!*Sw5=q|DOG5b4#6(#=+lNBVBLy;|d?^Q8W$fkEWZ~ zvQ6w{92pYA1gh+9D&aaKOBU->dug@Ay`=()Lw#)-tAwLoAi*Mx0D4mGClV?EbH^uh z<4$i^sV!xPV`^b%@`DJy%PT;MPKkSLy&TunrvADka+V+EY-aS})7DeIJ)fnPD+G zdNQMWPwZ(Ym;~l9|;vLQ}RNp;|}BLD(N6J4B+~&#kQ)?0kR+ixT~@1mpg!J$Y$qgsztLy5vs^8 zZ+Zf&@%2ep9{+ps#q^xc`(bPE*Vg`Te%6 zOnvQCN83)D&+6rwnZ;sxdic2lu?jK2=;o5X&(BW!>P9WrDn$UKe6d7*t^{(3@;b%% z;SMr{ZnO5`x0FU1krn4W!bOi|dCMA=yZ3PKGiT$N2lo~5VbFmT!=t@IUDY&*JgP)w z^TfDQck-oGTM*lZUP@{^-Qo+=I{Wlb`BI2p zKb(OB$a9o^*t3~){gS?3#_l^Jsyek^a&}409A{re4W>4_brq0BA0jqAK!s{9f*D1W zj2c~qi8Gws4mctWvp3msE%gZdpqF*$y7i6)?zOITZ>~3lvf*R>G0+|6>$nN8pf&QG zCqo%sZt-CPv)@-Wwx>VZbCGPUen9nXkZDzA*6)ogr{Fh2Q)LkQ_RdyLsG-@_=!>t& z?nvHux}udNuNpk*b?J$-GGv@lzQsK?e!q>vpw7QO0E$g&BmMrcqAfB10ejirR0B5d z-RX#Tw*JZ*mKsS6)zyZgzVI;8(1BzA54P5aWRQ?|DtUUZ?tvP6keLLF>_x<`b>SIm zKWdf8lpfd@uem&Zq7vxv>Hpr%e9SWq}i-M78FzrUX%OkT>iJmV6An-8gS~j&e-+N;boe`=uN*w0SNn@d; z`JMZ9iVm(iL0ug^AEgmqp!Pei9j)>cz?@Z=P^T5JH^T&lK%j@C3Wzy`ay+!Q_ntg0 zQZ}?Ja%GSmBjaVzcPaz^ zed*a_3m_Neym+aWc^*fSr2aTY{n;{~*6Gvf4U9)^FZwCE#D9~)`zF3S@-jyWDn_aG z_UODpJAVKkkrqSdp;en<)H&bC(zK0QsS*S?MkUzBMiUEFeByDijU=K*yQW^nW zHjlOR;}TkwzWz{DZ&Aj=kN2Hql2XRt3WH|w&Crg&b7OMH6Zm30^=59_iX1%@xfe0S zL+q_HP{_t&Bcz>&s_E_egnXow2QW!a zP}v6Kx5*4y_1*yNl`+!5V8wv#r^P)#4y+Ttf+Jo}IJl<8ja4@ZQ`U}|&VYg)X)VNL(oC*Nm9L*au1_n9{dTWiDc z+jdE=3pdQ8rqDs3cFAEb%b+yogMhiD!e<&?=DhAW0@zls(+6a+*utK$=UO5>j0 z8C3*zao@RDE>)dZWsiIWpaiq{G$QvF8dK69!I|YaVq1@_1lp5=rK%R?2G&oEfaW5p z9AS$Op~|`5_*@jkVDJo^Ws4weJx`q2$OmUBA3l(RxxV*bl$)a*`7`&DzMG zl*4cJ`5&&{^vX0srQ^f>rYkO*V^&4+Db0Cc$<#w=<=9g5|f__gp z(k6NW##2b8TZ{%DN@H%4=?mezegBk&6sk#iJG@Qlzn%H^bCAG|@mzG(bCLRlZy0_d zn+I_=aUKRo1O`}Rk?oa!7_lv#H)qBVnj>&72n0^Id{SHFq#>K_lR4AsS!D)l`u8cH zUl!9nInrSI$|m1nhq;n-Puewd!)Z8Q1A0z#;kLb$5tH-S+bx})I+IKxhKj#7dKFud zA$R)oJATRuk#o9iz?M75S!MhWB1p{0MOKU5HYD>smM z&Lh}cXCc)J7&3=PutCk%>(JH z6+#j!-z>WG;qk+LJ;i_!lYYDN(5%xeqf51QalZ6J9-skGkJ(zf3cO0hdR@&#(;!ye z{5xUB{Sxuf`8&Zq>Sjk>d>CQS`d9~IiDAZXZHD35Gib1k^9}jO#hu;rh${^-#dw;S zCj{p}Q1svqag#sQVdfDp(tWVsA-yZoaUec9$KwNp(GcyMw>R##4Y3DXX0TgM&+`X1 zs8l8GDcvx~b~>X8!rB1;`AHTVLxKrUKDm~WO# zN=4!vs&&bq=|#U)Bsz`5Q1|{q zH@tU!u&m0fgy>lib)b!-*<`y=?@Yx-8tJW#eF%wgBeNKscYmzZFCHE{<5ittM5oe| zP2o1L|440T^teyuMw}Dr(8-A$Jx{FbH@-fT<@5P;`dKEI?%17!z;5)>_N&J_f}dL3 z%g`{JG5W#DYYRl};rAQm5sdi(Xo4yIk*9+G@s?KI)HGq$RPiCvlTt=t84=IO7j{F zV`o2I>j@zKPzc{Q9H9~;mzY%b@~V|)@hZeDN77hfW^|M>l#ufEWC4?tHTfnR_2jL? zRP2i!M*i^cT3ch4d`yChZgxRm>SmZ^xlsq_*hSk6R27X&AXX=H^=uvTPJ?gBb+OIYs_%aA-(^t!N9nrRlTI0k=lu! zR(#{k>9qUvBhFBH0{Oc$-H`?U!benssOjhw#hxm=oYYD@?-;rhrhqGyTVI~_!mhnE zB?Wnqn*b;HjlbPlFtWwns<3`w)P_Bv_w6zQOJ%o!A>Z@qwrzN>>CJlqNrZU5B8|qK zqBYrQnXpK0>UvJe=p2o2=|vV6m={fu)+lj``PjHT)nav!ihHm=)uu|bsYbmyyB z<)kMUcLc|T@=9*Fhjf>wE{XX49!jb#XWgv-G97$)zH+)!pIDip@}cD$XU7Lxp+F;3 z@aZ$NiHk2+o!$IybY7@O1wXf={XP5Dz9n8-PbJ{#LFrZo)^pjN^OIyMs@7_ZeB^7s1w*1AeAd3^q8I z@CVcr0q-E>!>$_}FP#Wg>UcEXJ zkxs!AiN(@y(Yb*7*m7ewlvbQ|9u=HBcRgfB)gt;U<*3rV5h|cxk%IFq2P;?TR8(m1 zjFHV}^>ry)R4p{OBk-7a=G>*v1rNPl-dQRraTBGaY-DeC z=;3ju>%yjC5caDcang%9`ADbVPp9xM4sTeJjYg1MJ@jSN{DH%aqd3=2q;m1pugn`i zM-Iw?D%;`{vCB&z$dR7e*UxO>X?TyD6xRq>_wgb>HKLDWfh~|H4C+^D2#b^f>`pbV=PlWM9hueoEgnDh8G1hrzL3 zsMhO_-2y^Mb4^zJO}A?xAE_EO>lpn>AqvLTzqX1jO3XJ?pHvuw-9Dz0Z_SnN7tliEJ{-7{2aSE!w~+Cz7c!ZGfZdNe|Fxoj80|pVR*jZWU;VT z<+j=9@S#ahQyhUbvISqCf`yjMoFQs813XQ*g5$Rpe{`kc>R7J9mfZY1*H|jO-6C;< zZ4Qygll5xGOtDfAwnZPrA3bX!$H;!Dz5DP4r=KzDM;PyZs##EFOHTgv|AcF)Px&>t6=kfJMCYW)$RWd&LMOTsmb+})4tXB+6+ZIE7KcA|> zfWp(s=^QAPjlPSYqQ>a~hWKog=C)(wXAUhmV6zEjI|ATugxz|qg!kYhc8x8J=|j^uO3)4uMtQLomLRpPI^Y}Ax+ z>mq41VJY_{pb##VR^*4w1zAi6)vLfM$#FyD`@hDyyn7ld-8B*@+W1+8Dqw;8r4oPr z>Mk?`?#2z8)!}hA?#Wln$(u|Ft7cn ztOj?0RE-0c-el@rCfHu$*@s-m3`2H{1e4K7(<<>~JaP*_1q6Thh7fD zT-)xPWKx(a%v5b853S;!)+Z#H|KmBoxL!r{qbScOa6VjoGyjrN2HPC{yK26S^NFgk zUn+VJJ#dniL(}V(kvARC+@@i#U!Hpgb59}A^QXL`7^t_;nI4Z-=e|E}sCE*rF0WC| ztnKRPX#9rfbm?>Iykq8s5Ij5D@!9j(H?wz^);RPjojRLe%G{lg;a_%gSI6{tT^`Ca zH$J$(WNamT$^OFncp;B#8W|#?#v*p-Du(5Rd_rmh1@%KWadog_VQo=9(>loBJE34G zHMihnzZck{4-GRepsTI)fr4*@)Ze>YzCGl;7jcz@pvgyzr}FM_I;Z)Se|U+r($@y~S-c3}?9J z%dDYhVgnlUmXacajnNP&{YtOmn}OXAlur|oo+?sZIhjenXRz6aD5o6_^55!|S6vEhQt{S%wvU&3K1F8Ltyz>>Gtxq6Pq-J0`-_!Q4aZ)==WOCZs z;^=uM$)vVRoCWqB{eJOXu{^(Ip6{TP0#}}ZRtZjS44Q!iQdW>`C%g+npyB?Tv4s8J ziBZ7$et`2{o0SH4JnHq!f*rqM&ms{x%b2B0`n_RRRNN<`B=8Ai$O+l z@g|9RgCJb4%G%2<&_tUtfzC=(;#*IN{uNY{N@ozNQGfneP>4^|qzR>q>Q4Xe6>J)tZ^Aztjwj#z!y_Ne}+44 zNGYn@Pnc`-G22AdE)_pxXCVKi2Ad}E5fx3s!=h8wMW9wp)WOlT#&~FRb8I8@-Bwy! z!rDY+=Yl)beC%E7b7Y7vuY{;-X-bhoU^VOeyZ?WF?ctK{h2$`UO(cccp8OA$@I~?6 zf!vW1Dw6)SNH2+B4UyhL-X(a;snmy}4wm-FjSYG?r%uf?Y##8ElZps5HI+8p+xe(r zs&NBhH1LH+Ma5+G%OiuN)YMA<5mxXs%lVA|5)xY!ReyO1CvokBYkjUz>lR1p%y)`Y z`gVg=%+qMLe$1xg#v|;Ng)LQRXzaqqH=9=lUEwY~#{-bV5lk|(8YE~&PEG@aqy&g5 zDqTM(jRh_ZI-o{B!D+7Q>>ajxQp*33F>mRU#vH%&Mt|Oj7p&B)vATYMN)t`BxiTd( zMx5wWOtmeJB!@wxr?xGm_mUP6t1CHCSXmh8(n7j?vI<|XVw7wNLKzsOS*W22ZZMc(5*2aF?@o?hI zNSEed4dS_>&5d+L(A91ddWQXLzz(tjErp4`>{$aLXUZO|x7Pf$FC3Tz4q3g4cN#~LiwaoJovdiu5%)l(}h z^0m?5vyQL)ciH*%-f1rqw(rKQ(?ln3XH08w_dq9|NUXn@n)YVC!H_kZxqmuw_No|f zwI6+S#$U)c;UHp7+fN6vO>HP@w68LIICbo?b0ps}Et+x|T`j}!o$C@hwY})7TNyBC zoix`wYyL~sHO`I5KDj?-=i-H6vwHz7Re(kD6z1~0TIh&6X7fOZb$zm}eKlHHzt)FT zRn}O}B^O{&==BiL0@1Evq6oKiy4cu^n3&P{Qo+VZ4+DRV_w&9M!pj2eI1ebYdHa>YpMBVIAf)069YYL>t$QCMqs}SM;VIt1FZo z3m8rd=^x1ExB*4J(A87FP&&KG$cB{NI#-n+2}W#8yTA8Oa<4HJE*dABRo6nc6b-kwVR7#pru$C}bj1ahN2c1Q+Y~as z8qqqG-Su^2aBW(3&GeUhK4YS_V&?!RqTo?H)k=mxP%*OB4m4mT3zztJ-`J}lD>Ftc zCmSYCg|=wu=m+>6G0Iuyugf(DL#1kI2;dYI9qq9!tpi2KRQjbk>TJFI zR(vI8i5Ti$phWaEX`{xGf6F*7fH4oak#dtve1@7o}=dd(AHf=PP=vDu4(& ztv!Mt0V^Jt!wt@u;z)g4z6ox|Ed=)B)lq4ku#x#b%OqC)fJbZO}9M>%(u zDn!gfs%$PLABDL!D|!@d0PC)Xi#7dbUE)X$gMEO?h)B3P3o_~@75W5BEv+s$iK+D@ z6HaMOM?(cJJQ_?t@p;hD&}}s!s+KW5FIRr|k*ma>45c={RqyZqR?l0_HOy(eGjsV! z{Yf7Y9?7x@6|m6Ulj2lIc-DXHMfjuz8@KXQU5|&)FV-_;$krY`o3CsV3iK(J;CPG8CauQ3}!4@50> zKBPsg1x~DM(AV?XU=OCSre;&aJDiBbO*Vqm7FRwam1_?-JI3&@BIWAdR=d_dx<74i z?|sAAZK(*&{*(*WMS4(k#QxH2--Mc5H$u%M-}wl=?L_(X*RBP$j8Mj9#hgVbQ&dAr z!0<}LIcmI{&i00&<|H#5_b_N*q$K6L2I!)pk`mkT%Tk&8rXGJg>{Knp7^g!h<42yI z#N^{nKZymny?U)zp#p_*zsbL~E*_q^AB>eeQ3~8}FJr{8p%ND!u9^BwG(~^VOnHLN z0i>k+lA zbAh?piXBW-!}IiXi`)@5AkFCv1S4+6_0qS4bXr(N$n;Fw*P79~+ z^v!1+zm16ZOF#o5%2EsqboS*lI0|ts`GocWV=(kE41*q6V8O79bX;IDFM)zTtii68 zMzNwI?qU7wg&A?J$^?7-Ib?9*5{*T*He_@PpO!rx;r}h@f zo(s%$0nwC#d!krX3@|`IPbn7*B{8+(-UfhAiPPV!Xs?n1_OLj&Whi|^zakW_#ql#2 zsVnl{`23lWfd{#*Z6!T#kz;%cPTlIj7-^Fj+eDw~@t1EONz?ZLtu6mZ?0ip~I!phQ zk^b3=yEGeGhV#Y0(!>gv#_-i(O**h=YGawJ7GJ%}IA87bi6NRbJ==g5EZ;pda(tRX z$x33e`YxGNz>dDW?@R9GqSp)HMw`@8o^V@q(EX2z1>nXsLZQ#C;+OHQ(B?YrUDqJu z@!ILtN08}+xa@dG2mA2|tWp5z?@$A*Ikx;@cxt@Rlhhc}bpS90M#$E~-gFNORdM4t zKw<0+cdJyyrsTi+>5YF5Zp+qqA6|t93RO(hD+RxZo8OymKZ<^5PflUkCo?2zQ3C!5 zy)i_5mK&4Mi|9J2NDo{kIV*h*j4x*0F{C%L!LUcc7_mE(x(bOiACGT5q!P zCKuExc=~_)8lJeCtB`3&nJ1NE*Cc@7ZcI7w>Fx6p5 zNNVqmG+)!$NxvltYjDF*v$#3hHdHE}8Nxm4bO9AxmN(RNd60jNyt;GL9?XY_Xl;7} zEB*LtkvN@KjvzhMScxD9+Zgdnq+@I}RVFxZt!#UY^O=m2JDP8GjMkW;k|(H-V^9WZ z?(D`Lwiy&}i9LT^3x%dVYLF@g?{Nz{U8O%NTO^$t7QT}F*SAp=z33>+|3B@W>08q2 z8poT=qeoM#Nt_l#XEIVm$C6xdH}z68HBHl$MxB#LiX{ZDU^-;wNK=Lu5IT{Nxu#{{ zGFT3p8sy}zxa0yAiaUmbI=}PcO#KVzy5@QHyuP37S?>G(e!kDR{oXm^?KIIUxc4un_r6*woIC-tb^>=kL>q704lB= zxJL(Q%b%l7bl{oDuqy{r^@Kg8wtmnq>KI$~Q}&HvWO8+jEIaGuKHE`}jvQw3*ct?9 zA%*KpP0z;j#flO)oaZ>s^_%jCloci3)Hn*WQe$*w;wpc@TkX9NILWkdDR3Cf;7v=z zXdyI&AhA?8%K#o@U zB|O@N907-Hl*6!|PdxUz8gOB&S9@>ONIMPfa-4;M;~$kaN@kC61Wy1gRBDqeXy27i z@#P4U+HJo`nhh#5CwqJ9^O|Ux(p5X`mvKo8=Q_Tf@=^eR{e-> z4OOC!m*uAxuIi&-)oTN4&94u!EOQIV!q|p8v5GKPi)>@OaJ+m44f1)>2Sc9szZq

    jg9c9~q-+rTa^DiKx_?aC)K_MAcr5l^(kyE8{Qzi0UKExBdyfcQ3h(|3_lDAt$f)ug zbnMd3lcGVu7R<{`S$FwF_A`~lHZ($X6&hW>_X;KGN^f`XT0S3(xzC3)kfvr{pI^}P zHM;>jAHS27-HR2$2~1xlP-+@B5VVS4Q1usuT+ z$tqZHZ6`l;CU4u4&4nh~CR|iNuAiEup1!Q6lp_;>YYTmZdjonmil1m)uoNhJqV9+Nz2`v)qO`aw|oB(K|7_KO3J zK^NJNQYPQJBTB>%EOh z7YgZ6e<;1M&2{;*-%SR6cs_T6iKyo6Dp+KadQ5!Ml8m_rpC7Y%es-(i-3X6o=bZWW z%*8(3-Px?$I)v7ACL*JHdhw9RfBm6E@hwZbhW-;LKgDr4$1Jc)Bw;W|P8gdHD;{*f zmGPb?UR)e$CXjac()6RQ|6+yqv4+fB(Q@Hn19LQ4x5M!Xg0M$q`OPTxRmfwpXlj}L za8Nv8x-$n5(k`DW4Y2+Nj@|s2Uw3Qsx-Dk&qB(fE8rxfy(rc;M0}mYY#pq zd~kp$Jwxq_ZeqZOvwC-&=NVjFBxyw#gnsW1F$> zWZxV6Hs5Qg=kxjc`~lye-nS{-%r%eeT<39?`?;U<5DZaMq@}(>O+rFKtMvGhCJ6~O zjD&=I_zW5F8;*U{3h<4};qh~45)$@$;(t=RPFHCX64DxosR` zxP!;z84UwSoT}D4miG?`tE;Q)>+2(n+ee3o`v?1j>s|Bn^Sisd)6>(2hK3UplP)eU zG%xFtTV`r&YZDr$H#aveEiJdVw~vkpF)=ZRTL)^{{M-9STjLx1t9$(X{A2U0yYt)R zt~IH=G5FrKqy0m9d3g~Lk$L>y-dNXq*UDA{a?wqEYIEdZer)}d=T28u(?rG2>_&Nc zIUc*3-&~*GKgBF6>RL?5?~wOifMw zwYImibuc_Uj6@yj>4P! zza8xD*FM^B46gt9(URA__h!U(=+Y;k*f>R12bymVaq>h7Xt-4RnJp=)sb zHG0aUbmaH$%65NOMP7Nu1fj67(B04Z^AI7pgOE^P|5Z2Zx_ydI4K%rr5Q`&MCp;7_Dj;G!%)-V-dD5X4TbqPd8{nkz*#@; ziL11;6o0%8ZdPj*0{ggpe)l^XGadekD_S&+38ejb@m^_M+_LUqrSPXkv#_KGqXZ^1lKOkk4{l9|!2fCB|7Q*O!5DW! zWuX6wK)JPvur+^8$knIg>OKdwHQ2?PEi}RYU^7yZqgM3j*QyssUF6Xqp5Y*I_Rx2# zxd!ij+IK^^Lvd}4(2XZOXjGR1t~;YJwu6>g-@kdVsZOY0CH0zO84c(f^Qd7snu}vP zxATa2;~4u=|2u2jzH4KKo0Wt+FP8-mldE_+PRkq>Sy1Wv?gkps>dO3nMY~WBR7{~C@AIAB=B4qMISL%(JkxRb1MhzB02`?%z1nznXQV??<=oK^y!<&Y zytmRe#<-j#q7Ts@1mbe$EH4e!zR~HN3j!wb*b? zCEPHZA`LqCAjbHN*|#Hd2S{}A_EJXA%i-m4x$&VM8V67SMLGNyJLpiSXdLF@bG3N? z3S(`LR|7igsx3%Lop>SeaqYAqpVjXaX^2(ojhem0vBImCuXE=|Mw@#!ycpUiRvbYr zfLfUI)o(O8Q$ut!?zk!J@foeqi=xvO5y*IAXxRlrq% zn0vE=w0wHlN*1-bi z2WYp7YQ|x;91cF#&Pb1$KIS!1?ttU3j-c6do8MB;0|Fc?==619k#hK`uk-qgR^W#X z*=;%nR@p}CaoD{&K>LX8Tl`Zn3X7MNdmO%*lRM%t$D=X7h}>Y{CAQ$Lax?*4DfGtk zf<9(K2UaC^kcl9aksnsgUV8CO07?3=dh!c=}m{Z*QVhpl|Z? zjV}&+B|*thx7|JysN>$*(V74)K~=&hl;=T{liRBWm9}buDpAA4J+WE0ozJVrj4x{# zzENiC7`A!yXc=uBv3>L~Uzq(kbZx}n$^Xm!qEn*=7#%MelLnnzmmKR)wSR^@lLjriiA9$Nt1q>>-e2L&LSin_xMh7c*1InjKqJ#Byp#%l5 zq;Uc3FFaf}E9*HJPZq7&A6Y$`FnE#y=fR!#CCpHA#s%hzMeCcf7Z9ZhI}N#jI}MTp zQ}rGBZUYu}R$=wvS5=Z=)kgFIVcF<#++csh?(kQFJl5XlywtGXSYJ;~WV{nczR|bM z(Xy;Z0{Eb@SHh?*3HipRX~v8hbj?nqpidG{068OfK1l(BFh>X53td?cpR9SsXbX4& z6)eSA2OqG?ZJ~h0B-QNtg@ea*)Rqnf6)^NJJt{RJQ_p7bm+gY1mxJyq!3vBnGI|KN zL*Bt_qD}Wrpp^V+vd;lEvz!}|14D?uqvc;p7^^CWr}%dX`pmsY$c|u=#_kHbp2P9O zuj6QZjZ633{D%7XR?1K}l@n`}7rGvjHy^m*(G2$)(`WSHEI9tqr@z{GfS4px830eo zKsyVL!+w{hT^+ZTBSp^NV){C&6^-7_Evw$2S@m5@9y@bm^xLktIi_qhlq(=ZCM;o; z+IB-IscNHJPmMnm>YG7cz0vJCMyOl$B?OfDxN244XFRE5cwt9Ew*`*E7W>aa@N8rS zGGT5*XbDY-@%vLi+`4mK7A(^>Qgv4rcr0A3#3|gy`LKEpCVsE%o7|VH-8|+6u-`Qo zQJ_Xudsu{fM%Mjxwj3Ek)~A}w=E^cYFFJqgtUeAgvm@vD@Fw91VUA+b!RuIUE$Iuw z+ytroEOV9gFcb$XJYEIWo$YG8{p8=E;w{ z)Eo3Z++=sWbu>_8m)7IGGGx<<8}M$W9B!E5ITKjhbDH=F@x({qByUS{%x1k1CcP7` z>KN$|{zSfD#@#HJ<8*#ai-uL2-K>g?myW1&Hssy$oP>$Pge3frH`1x%+X7nmyP`|SFjr%jR7}hz8vlIYd(1sd- zy(6=%+IRU9cl|DSmZt0`3gT%SiNsoGgehqfMC`}W%((2F5S1Z+C$CmeVU!rd8LE8=$Ty( z7o&a7Du1F(qGYiWR^8u@16|599c)P%qp{cg(AaLBM0V&Ek*$ggUSpbOp~n4cf7TVXi| zBwZ?^dvMqGYy`sPc81D+9man>oJYKKxK{K)rAYmiQvz)xIfdqcMUOzF#oCb__Z5p;9)$cKzZ)7E3`DF%^ zv}zuvr017X&G?wI$BV}*b=Uj(j?eHjY&m@O6SuL#vg(K5QrPp_1{nLrJL~PSHc{np zj>kq@Nvr(`z>0&Ve}?6e5P^=q2r(3;0YAMtr4LR;@)*f7rR*QN`&i|c`p!kfCAxze zE5Gd;(j+uF9c-bCtJDUcV%hw)7DvgMGYO+u{%Mc2AaU7QJ$Py<{3-cS1dv++@PZ7^ z2OI>EJ75mhLb#{V}AZT7_Kp}c-!p<`-y8uxg{+rS*?!Y$ZkjY)VV_UYP z0uJSL1^LF=q$C9!fX#cib$f9DNc#AgKO*BqUcvv$yNb-QKTDn9lfnM0Y%H$u2Hb|+ z&)~Uxes~WxteixJ*~UL%XW-+uy5teYwxR zjh}YE5#y~_=?LPi5gJhvxm*DsNI2K^6s36UyfWYG+=16~?Evt|lsM5y2Q37}Q}-in zT5mp6vhOKg^g)c2R%}()2i*<2jq#w#IJm00KG=$Oue(tC*LXGe#G!DVqHk396naml(VEn7-(U9AKOer9ji@3OyajM67ek& z^PH!40C}g)P}Vz&j)iax-eXEk7m5*)1DjWwoz7hS;g9ciwhFQ0>$hb#SeLXv`qJm^ zrSg6~fvX}cH)o$49=$FxMK*JRJn&y@PW#82$6*w@Y<5GCR?=s<16kVd9G}WR%=;br zx%4~gECd`$wQ#sUCj?)g)SGXrls+@sGx8|YK4KgOd%j|3tD$$pS0gRvl zLKh_l%Y{}JZ3Ci8V_l5QQs(_`^VyHOD2~KB3XVn7`dsLmrlgJ&^-T>R8EFGshs8UB zYW5F*-q5SrPys9D_2{qjXUO{g3BCP~=X_IfH8YLh$TbMJQJ#OqPg$FQ9U0zP$=bf| zCW73Jh%k(;?v;GU`y;ayE{>cjEZ(*^R_F+u?=I<8P4b2H#(y(CgxbL1?r8+xpBEVf zo2Fn*Z8KPxvkPVNT({10zt=g|d;!%zIf@*}fDqw34@+w-^yHN9zz$`Vgt4e*tK0$2 z&#DwQ>g)@;3YupNS9hZtaeiKSsr;x&BU$aH9#!gl!s&7tn=2~Zkf+D72|#9AljO%9 z34}a)9pRm~_o^99@wdjj@yTO^@%yP7?2i@0e!oJ4z$ zRL725(D`PzcFCBneM@f!7`!WfxG7Uf<7ZkeQ2P=B^Rht2#2`)wM)LJ^l?&fkKU$SC%093j* z;x7GOV>C$3G5$J^wel`AB(3Jewq5^g+fo^bIgVDBwFhBdhpEGyc~)O_C*Flo!y?iI zpQ`|v`A@?5@{hu1VG5Bpgc?sdjONB8#P4GW*V^Mx1lPiE_F1yqC<{ECs=w%~S7fPd z63~}*fM~Z@!yt?4F5_ob0h|`NFjefy@3zkZEI=VM3DfkuO-g~=h=d;nFd&}~p|(C< z3L{3NK4~Oy=@e{F)Y$>3UwAvG{YYRb+Sl_|zY!s*9C#3>{3bo0%a#b{6w>m!kk2_B zlX20+H$ml32d5mvl7i>hi4S}4DwBpE9q>cM`We%hYsWGlYn1B`oK+YHYOR@bHZL}R z0B_!_eVnFuu1l(xSk|+g0}3=M0MKv-FtX4*A-dFYRdVXs`U`+yNlY9M>w;;B`E`JV zlT;^WE+UtC9M+0{@y(R(xP)_l6n5Q{3NVwC>X8lzE$0gkDLH}C@-!^|>aF51#Glep z38?Xa!bgk4Fnf$?nc=Co-D-O=LC%L+2esF>@G0S4?)=zepfWu!A}pc^02EZZQ{N*w z|Lj@AE>L%01xlPp?#DIXCRGqojNaGcjPpc9K~Ok#T$!dJB&iQ5IrYK*39aK8lnkW)LTmX#m&~p-iax@0G<*UhxFCL_SsDf#fnhd)uOSxXf3=!g zNZc&v=I0+FvU-4)?E-Wjc5K(@rfiSBo;!NchN8&73~-1n*-{*`YxTvmoWxoofZ}v! zpV7J(=a>CgTEPC}Vy&;>HVA_zfSfk6cQQ^T^IqcZoDIjr9i#y;Y<~mdr{%aRmudD}{^=w%2eKx?Z{-fxLeqUfuZvc;fDk$zG-T$kv zSt5}Ho&NaH?+DNeaJ$#U%>lRwZ0+Aq?NOpCBu*WdZ2(t6w6-*av=PxUW+PDduMGk; z&p$Ra%K%`AmU8&(0al{@zCOA9m}9tb2~e^#ol>Y)Mn1xqlk}KJfM}&C0G?_K_(Vz1 zX&OEt$)HHP*l;%h!vo~V=@13d4W?Br=+q$~`G1Tx+mC3h1D(bjLsoV5lpNzQHI>B+ zB^)PWqSCv(Kl{CQPs7t~P@Lw3prY#LFg3gFu^AeR7^CXVy2V0OEBmFH??4bgUX+D- zB%77&*jPTDGN^C2&UP-E#TZq&;3Z6M7Zo&$0@Y)J?*Y^xOKXJ+Q17vjr$ITg*{2Lr zhwX+fMr3b?+A zaciRNSb8k<4XMx~KI8^%9mS6}} zSP3%e7%16&TXs6q6OM@G!idyp_g~51T%CeJWwqv35l-3W9hC(5;fLg5?f&lr8#~WF zBpLkieE_KgSY7hT0b!-F$L8^!y+h9&^!;hEbk(!)$}uA!UsynM2kleG>*Pd#i(E;A zKFsi}JG@(UgNQDOaNxQc?kRqI_4$6c|BR>o;kdxzW$*sC;zvGX(KFyifU}_x$=*Km zVNq)0#0&o1(dy)mkULJv7&CqFJW^}1MwHf+ckC40nbucJS_}wWfO*+KTvC4@CalmA zqc0BN$KJi)kfrx$yq7ImfcO9(J36A=kWnWSalQZ~tv^;w)WG?c!{aAQ|MP?Y+N`lMylngM}4+5d-VsXHkX>CZ$o9^LtDTxkSm0T~a;Xoa9dc$gKU+ z2h+;e_^wJ|{cA*P4r8#sgnX-I+XcsXwmpr1?DUvrVHhWJfY65lT@qT?->j9@&^FZg znk##nir!NAySk-IDp4e2IR~sh$5;;6qLr6h+vC$3y@pP5|6hY*zjrL{`7l|`t66D| z*`O0n?4N}3a*vjyL(V<5u=1?*j3dY!*)DB;>^uLYjz=Cq?sq&+h+iM{Q3&OB40oXbhu#Kld>f$iBA+|(A@z;6&^%Kdy8u6m%KM~fC%dg zKY#noUBYC-XW{W_&>|#GsOo>hJ&(Hog%J2ixa2+?%NG`ZxV2ku#@>3w-Nz%EWo}v) z#r&|+k*PsT^+gR@AZ}Z22}}KPoD+6Acn^>UC<9_EofF;q#FI;L9!YOay(tRV6k_D6 z!(Z|g$d)AuG=;WnIS<7aabKJjvQ;#zh`N)*Wz(3|n7Z6ws;ZA+EwsB?osG&Chst5Z zwX;+UiF-p#m?!xyX51_T;dsjwySBk%O>s|HzwfEbPjm>poRh$@CHWwKq=Jv-nb(r; zhGdwDrUt5f)F3oO17#aA<&ckE$>Zsnkg6p~|oLs&D%2(JK_zXs$AA?gazD0Y!T3MMS={6+8PV z1FFJv8zC$F27ULM=@u@?i2)Y4VCmLVE7>kxrboeNsu0%u6}0_y!?}PF2KPo}F=e}m z%TUT;mM*t;OA^x{k@aAoIvgNnc7woN5e^5|Q z87{Ou9PE!~=v--4c*xfo=gwyR^(K08T|K6oFPRj?7Z+0x2eQ>a{R;pko>RL6Yl?g* zkLe9m%}IY21{ar({uKpVvGGlYfHC3x-pUgTu6FWXHmt;w&-q`G;Ys_UdW1S;6m zLY|v@Pyh5R{(w~=|NpA%5os{xg6QnfZIl`twrehhw3XkCTN(`J#s+qO6&z_^EMhD& zxe$9js1lwlHcSpO03|B=Mg1FdiQM@Q@9P@E=bq`SDqQCchNqv;x{fH_&<1!<`Br(+ zYr0t5G6@aJ3{7Mrqou_IsaF3N&I6yH0nqe;e5++V5t?S6v}k}wWCnmReF4t5U zHPIp2&)m8IqJ2!w0e4SLZ~?ip46!|C?p&_s3=Ne^_AjO%x(m@VvMh1wp%3;*DQ6M(HwMyUG{lxsDp$9v5#->p)p0J>^sECM7V!kTGJmkX`Bbt$;_+K9V? zB1X-quw>i#mKrJOR{Z-@@W1r~(0$Ry9qiN%(X*G>JTp794S*>@B=h~k*TcG)t65bd z9uTaHaHJ6I&~rf)ZYW1CUmj^Fk2whC!EUY(`P(R$S9yQ?oK|5u)|N7pYfFdA1M5<6 z0jvK3fa=^ZpMzr6RmCoPoMaJ_*#&IDPA;B!uU9js+mJ;HIbZrwTMewI5ibu$=Wf{V zooUZNFdG{dFVi_HW0!BGko~{nwzI*i2}Q{%RWjq|k<8sR3yPRXnTVp&`{yoy`|!I9 z6)D5+K+!nOm@WNsNR6Ida#1G(2(3Q)V1_*F(Yws)W4&&_TojeJ(h$*BlV&SF9!#2E z!MQ-bq8HRk8Gug4F3({*wT5$B=?$=R&wWZ*FZXSE&%drOhpBjopiwAD4lVHHImVuA&nEf9YFxu>lTjk3he z01jpzJTfN$cH3=gRkH$HIu!f!5@_m){}&Cvt*K5%3y4-wusav3{($EZctBDZ=>@Oa z6N8{Uz&u^3laidO8m<79xomc?((^tAq3f_QJ_zgPKH$-10C%XKPD8wYqG4qcA+o55 zfk04mkE2T@VSyf(?0vNvU=*uoty-(=x8I7%c*HLCB#I3na|T`LI0t`zfzp96Y4G?f zPb3XkYDR}Mpp%XiCNj8TSF?`xt<#xbhQ#hEg0jS22z7ItI>pgn-kgd#sn4f?5Y?WU zyKDSvx9fMN#lp0={<&;^D{&d8g+kdfq>cl>h9)~m&uOl$ondRN7zekwyTDuTqrHKx z2UIf;yTV}fGBfibR9rgi#w0?ppKWR>`2JXu?6mGLj?;Qu+$m1x>NyA-!gSaTbHTU zXO@3ajC6S_9xQ`Ws@yv~!-c6_i?T^IOCVEPdZOj>QKv2D4JAPoJ8*EQc>*$A#u#x6q*NW5qy|aFp3a3!Y^yzE`=Wkmjii1gx!Q_ z5tj}#AQWE{n;w9@pX|qL38(s4Kj{D-t{ncvH#_ziA|4w=MQ)<|8nq8f z7FC&TQ5^1i(^V~7s5wg7lsyYA`nH+Q0-i});|Wwl@WD04iKS2;^2On%56^KT$v^2w zlz4>Lb+LJ_Mj`C;)?3(nHLqIJptCzW_oa9R9=b<^9jhyNE#xq~J)iOvfVB7r{5uE1 z)%3v4<0W#1o!BLL@W6^WbFk*o;BApM@%=5b7~=^I5=iA>|9qX(ci$)jq&~N`bT=;2 zX6ue42%C}F6=~z&!qQ;D4#K1`(qkrkH=@`T-=qXYaE9YGYXi)bO1(tJc^Ur2>4L37@A z;7Q+=80@cm{X43i4{SRqFWsHwWSA2U{8XRY!n(Zp)_J{pq{f*vquVykLU z8S&ig94c$pbz<0p$x(fw2JOORnM`H6UrFhnHW@A@$8`Wrd$q6^mxb2NEkLP~Sb zbeG$p#Vv}5*l6K0!)ovdI5vw0#+wLz*}EZn3KSz4!pz8Y(*3N^x6GZdX@^x0!eI0+ zj*Q{i$he`E8zg=1PQR7zk>#53U4KL>PUZ$*D{kTD>HR|E>dr8TsK_khOE6E)air(e z5oWwBBz}WOZg#1<{&IWcesOb@+u3fZL?N@cKNV5**EQK(SW9mlCfVH@=WKpNUfH97 zUbJyFO$G#-t>8RcagoT^Hr}WiQ4{SC?!?KOtLG}a@=RTCOGBVlP>~1L?0MQwnJlg&sE@A^~A7UxjQ0$aK{@_(+CYRTS)3C38ZFgRj zKaCfKpLhyfkxMB|dUAxCL6;^@wK6%Dbtts{Lx)Ysii$^?;(qt}rEC3S15=ru?b6Yj z8P83gI+FxMG}ztnjk|5m;THX1@?m7m1~GSx!v+xvcO`gXrC_#lPSwEu5CyhYksT3o z=yPF_Pt5x%<`C?M{GhJO`9|R@_w48wSaoSpX;9h>%}+-E)N{_G5H0H}8IQD>R}7}w zer)6tH$~E`BC}@8C8Tz0I#R?YAebMNz2PM}LRY;?bZa|)DrocM_=nLiDTD6E&2_tm z#0Nek-};%&s%NGW7`f!uRec(yRr+2?r~n}#ea6iCY(ezZea#f`c8@3PI(X;pP-S6>8+0p$m@Duwm3{r!n8AtU+C!watU7L zeG7-h-d<@sZ4>0+9?3KMd^0r-x(V0Ju>Yrk_$Z1Kk-b!^n+{3cBJ1c)XEDNO()3W# zf0RfDcOl*vo4b99$d4A9(bmxBb8NpW^H_m#TX?Y+%CByUznENV#f9MwaU5>RXGa4AV zqk_)i9k@39m1ibi|LH)>Ym=XEdmBn_x``*7>P@Qor-Z;I&=ow-_dg2^^nVD|Yfg46 z8ydJ5H22}-K$Q;qH0){att^l%P@L(5D`G$5`U>0!R_t6WjxV(8d`JG#F72icfq;)?iVS$KxeNt?HqByEN$xo@TN83*IK!q>q;7K8N zzS-`FR-00&swCzr1hN#vKl205+kb2F?n%vC@YM3L*s{U4z%OigyB?pc~oJqGTb0orkG7vfASL^q}Vx5*| zP7o|z67&43m*vG>lgi=tiC?{?4FwghZp~6i<`NruFE9PvA}jbxRtkWT*bR7kG68UK zJU#;<+WZiV%Z0hMKENyX!f)lu-G0|sI#|SQnG2;@x~Y0^j;g}2%b$z-FM~`)yZjmy zdoeWib6O^MeeXke<03P?<*h>@S{?idZ21VZvXE#VlWdK>pKtS7hKi>%`;0!B`ILds zjUOUQmh>VN02~^ zi@vh1MQaWcvo}6XSq?9feqmcGVQpyDOU;!1UQ?Nqp&ZB|2-_T8FRfo4#jI(cOrx3<7CP-xVD_x&oXwZMJTTtkbR z$$X0UH&88?^yE!?exlJiVLFKHbBX~tpDSYGXSfu#X+dG4Jh-`$QTc*q+;F@O#?4zp z3qP-1S7X`fdJ`Q#Q1mG=@e}d`|1$f}a`?KTO8_IQ)(4j#Ps)7Kc2A_5^GNRYURb~J zO8_~a76usR1yHzJTn2*25hwEs$LhKtj1PccP2n(vQ(IY-5wI+knFDC&wws*@v;v3iOGZvJq`)yS-tI{N8h+{7*FP9+#2;}xO!<~ zOAcc^04e-N$0J_E_y{X7Id$iz)uN_E)!|n2`fbb=eK4Qv@5=y{78B|ROfVhy+;B8k z&W+5T1(D%1(&D#Tn7RcX3O$-Ap#%jO?{@94$|%c%^y9P4TMAaME=b7py`{JrYL|*( z=1^PoetvOP7xkms(AyVMXD61<_)$*5JuA6`;}g12M6>|<`E4TENt@4E=VmCe1R{1H z&k?&UogR+emO>L-Ze}&|Rre@o#(uumOLl z>1S?}8p2HEpj_%BV6y2ZKn@7ID}A|7c>?{-iQP>q*xm#1tk@*LHU_YDDc((^taMfv z(hEP?Ua8LD73V%O@36IB=u;za_rRPJ0m+ zah`nYn-)-YyT*F%UU2sUCQ@aA)*zCZ^0YMHU)qY;KvTa8+X{4q#FUYP(j%`2ux*Ye zeNrVd1H=)~69OY@_-_kJjpB)o!CyYo zhljFwbpoRfYAXm&&`2h1%-ZT>2D6})T0pPscnJ4|&>CNrVcIN*FIX;|hV>pflp=St z3BU0&xa!<%5^4D}=lq^YeFl^b3?3e9A87f0OA3TXW#%Av@fxWDrZX}Cay5Aa=XLW` z-Q*pl!R0-E9HR73T|gvOZ*uR`D=_N_yNO%`cr{fis~T5oI~VfRi>NQ97l5&RBJ+Oy#7S@A z1(}KC^uk!}S`Hh3_*aHLB>|?tOeL_VP!t4Dcr@aNXnDHCms#5(QC)r-w~=*luyPr& zcHuLIv{FVBg&w26WP!z=MbpoLIcwmEL^w8u{_0DU0XzfH|5Ve%@Y@`Zvr=XqyyO#il4PmEXNn_NJqJiFx_o^}oXib;#NtoW&2q z?qz+9Nk?~ZA0bsH?jBs|91x9)Vi`L{{;36kK6L$s___#xX(aP$Yb_7Arh8bEh+}*7oGl6mfI^<=g(5 zUMHSu5ca~*3{`%i`kC@`Sz7JY;|Jx=4SX|&x2MxR9)i`3C@85n) zMkPQB9?uwHT1i6v&;O|{{%L;469+vmNSZ+Hh*a|He?}VAFaFccawdC8E6#Azv-{i3 zrdJ>`LlmUvgEXUvu1p-_K6ad&h<(CH=<(=9-kCz~CLl-s9nXgv;q-krj!+h>GStaP z(16X(+~Ias?ewUoA@}1&%oEcP0x>>h81itV=Zs8 zSEneCZR1ePeTX_bocwS!H}$DDlHCmd?Xat@Y8r~VVP5*icg&ZO0}?GcPlHq<_mCyu z`&7m-BntUA24b!Q!klwP>>kV|hNlhSZnOn(SC^``4#{_IY1m;S^`ljpM0hly83s^l zy`!B>?zW+h`k!#6%Yb#ySHhQ{B=_nMSOV#)vX2apPd^!zY$<%YU(0$p`qFg+{CQ;Z zE4C+I)Wb@!RACI!Eubs-RCjppZB-P3F5WalrGM4W5Sh;^*ksM|2z}7c#i$)96N*Bl z+GiJE&Oo$fXzV6xN_&Xx1#gW_zDb9AT5BM}3;50e^9;m0Is?O*KzcbjdNQyG3yeRM z-rE`2U-f03$Xe%>IhxBX!^i|=0K7>2Xarv;r&-u#>|?gL^{#-(RH^;@1&IW16EG?{ zHg31@GKe!8Z248WcLN|j=3z~ra%woD=K+i*ky4qc6$x4+ptm=5CQTIef}dc=i32uHI?XYsoPxti^`?cd04ZvR>@S`W*!FZVT=1}VL?r3h{Z-0vRmGKrt{e(c1`GI$pId@5??O~Hup^3g(jsaL!e=uCoZqc)ReYS{2e9_V z3OLJM<@s4R!7rir?^?|~?G1qkIf>*pRDUqU(dc02muQBHm-b3|?4~+Dm%uF?A*j6f zzmk{%*8&EvUNjOt(-GucFjL(5qB~cg6K1LZC`c;cx(@bB+FgFsnf~Ii544zgklwu_ zM%WcC{BzBaOM4o0@zHUa9Y1fuE_EBI!e2#Qc?UOXPMdgPj$~$PS~@Ma{Y<6(-qS2f zkE3YgSL`Rrkl5pO1bOO%1N@lVx<5%e*W200*N)E)r9Y36lhyK0$x`QOd2fpQi~<^E zs#bNE+_2`e<4uV`+wQXqD64c(Xz&GEG5v#x=Tnh$pa27@@a2dvCh{>+jUN(+#FKKd zK#w_rdLyLTG9(y?*}!mt_33VB9>tzQE>#1j`uaTmqrF4Bu^o;0I zbXCz;XUqLM3kVABU7l%vv#~7lriry_Zp^MXXVw|3R!7zqiqb~|N;n%7idT{8@@@2L z>pH7jW@hK!)>FlH6AMsZo4%fiDWa70=${G_7!w1y|BHiyzOjpn zDO7HeZ}39wnfWrJb4*V@HSb=ex+q7fY7P-TqsL5lS8}QHBU!6pQLFEr&!ausv8|;} zHrCNDfi2w3xiM)_&o-QZb?-|&(ldhVd5nfY-4i{W*H^|S$xTnr?+WF;LmEJuQ zlP@Q_xj_G$+$=#Q9qM*rNEhcS=Det=|4|0F#BF11me%b7Z?%((Ij0{6A2}^@Goeu^ zNQCo1d*g!9UAP-Hbq@j4+09C&V=1@*uLL{PVSka73ecSSC&H4AeMLJ!2$W`WV<28Yi3s# zqE_4j3I>Xtzgy>WwFQ358!3nD9t6f02BBV-mqM9)T#6&xZ7p0uO`Q#m?Pfo|Q zv8MAgv3{M+UUd0Qi~}zF;I%TLKT)p!B+B6npzn(jYr_GGQW}MQJt6@xaj>9FMA_Yd zE1U7>3h$+c0BytMWFcmc%QSUd67vQ!uvV{hM8*eI3A(c2s}$2q@X;C`7;kN2ggo`z zQmbF_Kbg_pjOl|{&WKFwk!etd+Vfc3-L$5iy+Wo*t)wBGjAf(RPP~_;TWO+(Vi9Oi ztu!^&8$5K?g~N<{a5$Ev-n2}=AC0DB)Yh%= z{1&yMgMIDlD6avSR78ReYK(u@0=E2OtP7-*FPU4}R`X{YHu+#|ZDzU}aKmNs?2o0U z(KNr>bi5i1z*V2%9 zjJS2Ha#GY_55mmTa%PABYV#dy)n}Iw}OmDu=>CA;S5 zvfv7eyTtK``N0VH3`FZPfqSH6b8Cpc+t)4TI!?8W`-8RB=BwgRe;UqPyMz(Fn=qKW z%|lHFx+W*4-r>PHMpdBY7C3nNhVv@K-qk$!&2L6fc(SEZ*ZCwz22)&EFT7iv2L_i6 zDG*X(9?ZAU>Cy<1IC{v$VR=w-bDjz|c%BU3lxj220MmV^T^_!dlS^?Mb_LA;i9+Cc z!j|VkDcb4ne&LJsQ&L9@yLrn00c6rh9>0~Y0LD`?x^A3fnXC)uLl(Seq~CqHX5?pA zip$4rl(t-4l^=L|)aP6DaA8FxSyFac%4`yvqei?}5Q1?4)Ko3K#o`OsT#W#BnWJ}f} zKWiPM#a0Waf>E$7zJtz#6P84nVNF?&QbU3aur}Q{SfTXpOfWdFQ@*Gu(yFe!CO7fk z?^uCu1YE+u)zA&tH#|wcJS&_@5pR;j^B_`_e;d5bm@I0)V zt!pOUDNC*-9F${v8+l!D3O2*nDY5)Z#yY<3%(`-%P-C*&(!`|2lT1p^$A57yrbml4bk)Ee#aGOJzB zH0y5s;VL3XzjyxvNyAf>BQbkv4l2-cowh3R=o~0*rl44l4z#}yr-Lo@^%w}3@D!LL zH``e9uZE>L-_XI@7cVNwf!pSOqTS>y+9YzIkk=gSgu=~0`dF7A6j=nGrXFc&HCUzD6hSNKFHP7ilmoYCB zN>eJW2Vqyf2v<4gp~}A%p^C6{L&Y*lX{LU5=G8l99m;_WhLzC#b)Lns(~8e=GK=oW3g#ae#%s$@%=M|bQ2;hV!7*h-piSIdQh0_=P%*b0@it^3?*hi zAtXll6s%O$k-l4#7mg>O%O2=BTjXP(>kU2GQ{MU6l|ygcgv=RD&1t%a%n3BoG_{@h z%BMuPH*d(#3;o=886$BG9DOHarc5wcbr$Cns}U>``nFo;hdwa$Zk+Y%Z5GeSxjZG| z*ESTggGJf7)k}^}Gj0+n%X>esR6ock9`bqt_4yB7^D}dX)4Ty!BytVtXHl-P&pSmo zyBtOh;>)IdId~Kll7E0NZ^UZGoEcRP<&K26fue|KU7x;#pL*6}w$7(9@{ZDVfT zd?=-DV1B1pZU{-c&3j;5!~~l*35Eu2B3$r8}9xr`Yk;N)zMd{a-iL~Hc)YH(T>HWbL)bso@dz{2D?nLUO>**5 zDKdSYo^2vs4uDS-v{)ZKszoBjfIl`M^=4#w?rj{}7Z@f`fZYQpy(SG+leK!^?72Z;PxCi4?{cLCKCHnF-7TqFH=m zaF`?J0yqRc!vsuTohz89PM++$#si~s17Y@uKnZGkXfcY*M{$Sau%>#vnCJ_>a2Lpjhb4TA3lPr}G4@6oc3mRc{NQpC-p(z~vg&N?Wp86I(anu0 z9q~I;OnkExpuW;fr9o?i46O%aRD7_Ixxyu7h?cQ~t?Txh!`B!l-{s|zhLNv!o!#|% znz|OAo&u8)aPi}zh(6sru?%5!?qygZaqFQ=Yr4s%JU*g>%F2sw5jr>PzLiTlBc1M9 zLs=|*xdLp8Jj)ki)6nIZmo`VEZe4pa0*nx5oEq7E`XT?3 zpO_O!kB_e-RmO<@TtD?iAO|;?3pkqRIG`8KcYLHyh+i0Oscisjpz=J-@fVk$cHApF z%$46oB+d|L62)*LkGD(PA|h!7#1q~sy<2X)uGJmp$Y_2iA7tb+W5fQ4^OC%hJb8J= z@;6(C1@@=*DMebqA&a6mR@;UUeN30z!i*aP3ud>FUX*LP`KF@(LiVlH!qVu0bm4~9 zE4r_f zm*&nlN56Z)mw(I4beK6_evmx;X80gc%V{Zz>#o-CxWj-YALpSbWlWjSa%PSn(54txU_@m#K_)~==#1yHyHl_$OJ4Cc_ey*YY?GL1 z+jiZ4nY7A=mvTwIB0zX>C52}`2aB*k8ifeR#F|GQ@!sT|%=d9S=w+>J&j_nlscFFBmDr`>7!)!;V$`1Ly!LXB!2kT z(k$YYER7t7EUApUm?=KGM^|6_m(SkMdC~_0NRITcsw;j-9zDgyv&yB}nyaoCGZ73n zo#_fe8uyG+GoSu{e0_CTl-={bf`|f&N+Y3C(h`f(CCyTcAT3HRCEXyPC@n2Z*8);X zHzEQOOD-v?bV@h-JBx38-tQm3ORj6ZlxLrFX6DQ}XXc)J^c?w0ix!2yo8)?4L|fCr z=O^h{&rEDd_d@fR4&u98o$qYE>Z{9w;~lH0d9N}He4cnij}*0$eQ zN}CUoL4DZ?dpMCLfl^es#Q zgi|rPPc}(>$9d*1Oo{w%tg@(~?bFj+SyVBoVY}jW)O*k^a;|2vxvU#( zCA4jF$J8g682D78?xI=@z>;s^ap-x4vpo$*T$h~CTwRdT)j{*-CVcK=ti+i^+9nzq zJG4C+s~B+%q$HNAHF>3`;P{>*dS!}0}v^sjfsca63LbzJwXgA7c=AFEE=Gn;;w z3PnhlJgL;RnkI>U8oWy~k-VLbQ}0K%IDdRKehPE&3!h%xQ#{~JpZ2is_2|URC*%8E z)-tX(TQ+TY%zCX>y2T>!@~=Co)?<3}4p!CKuq4&%cNzn#>iWb1+)Ulk=@dnNDf5-B zmgM?04*83*O`&TgG(EAwZ}4`lXBr6XO~$k&P3No2F|NY`|8X*`u)jCVJWOS z8$8i{I!}X^x9HiCdvZ{lDmNydr*lfziD%Uu+J3HP^c0lrJuNaz-C|QR6dY-Mj@F`d zF=YYekmNOpSObr20*~}7=1Mj?>dYV>4Jz4#ZJC2DV{ja&;apbIC`X=xyqc&e8wU%o z4o97tYY|khy;g0lCb~-#!t%>tVJ&U6VD35Gcw^WvD3%t-Zh_oi?G>5+RoJf1w`P*V@HY%E8={Xwh{Vn!TsQ4>K z!d}5oSs{KSfz);H!!I;FvC!n7n$f?KABuGsAnrWbVh=yeD`8D0Y-+?x zTF6LKT@Q*!Z%q5+3QV0Vg{Y48?mRL(*0Jx3lv$SY+h_->UoF`SH_re}TL#MXBdeq~ z&BXDMZq}{c*e~JzROx2?@1qLdv-;ae4~~tXmFMphR)h+-CMrj0KS%SU>4MU?9iDxJ z@nn6{W7)_VI5}!ei*p)3Vb(q9}HxKGwiBKXg%9SY_ZNG=Bl}nD^ zwm#$IW8KNcua2(@TxYg4*PJkU1ZVz1tz8<+VP?KFt;$ln!6_Jkh>%9zD9y`OqqSqjxw=0$qZokf``sic0d_yq~hE#V!==trb7wcIQ-Q}1<#?w`#8Zx0PS4mw!?|G0H!6*-g z+DcHgQca7=Mw99*mS79JfUkj5zHAST*t`OEIWS4d?SAp+<~87X(!5>?Je#9RFypF> zXwX+}(`ZtUOAyl-6&9WiC9Z}*@kLGalJaz zUBHv^C&9N4bc=MquxVaI`hi#*))4i{`%Xc&SW*Jb<$;YAd6E~NP!dvdhRnV+vl3o@ z9g<78evT>UoI*>(HtA5dN#eTjRE4{rBJwRStiLJ9KYTPc)3ByN^6;h8ri$xqTJ~sh ziiT!W;R3*xGrU3Z!cPlq1~nHT#ets`%2m z4_1R^kN}Ax@Tq-)137#NLAl=Qup)kAv+C56-ZE(q37a6-iJ!?|CLVK1)Cj-z1Sh4$ zI=3T^)t57>-FB~AabxWuHJnm>x`Yy~1>3JrO+(HUo0z-EQX3%JY2<@_xYiPOJ*E={ z^6#ZMN>p>s`3(g>=rb7}H(v{k;`^aBa8a|eH|o%d1Rt{dLt(mK!Ug>pu-S#2a3iM* zPkE82mAgO7_CNfXx}D>A6=pQs5_8dxFax6c<;FUUSk$qxx!NAG_Hb~S?rJTS9Uong zzV@tV+U*tIiqo$Y*Q{E8J7-PD;zZiqN;t?iWB8Gw(JB^PN#&40a#8SG4hw5F=S1eUIRj}#ZSNo{Z8D$J@+cn@SN>$np;KG8 zeeU&k?hAa=^Xgt4=fpGsckm4t8wol1{AGcL^?_Uw@>EN=kU?*=ZpllDJy<$2B(I&o z%|5oG&gCi9eI%V^RlSL12{o?jklY0;ffwyFdI0Yor`#uYR)|!71u27tCUauox zTi1UBoz~!M#giA4`IYrh^!)0ZhWv&H%n`kCWx$K|fjJnVLrr2JA&LBoJtQ_TYfl)_ z&31!{tlq%j9gk)QVe_ZKYqf<%G1PO*ODtluh~BN_TYS}I(hp0E=6!~hl~_#?oXS7i zcaluXnD1)v$)s+5w(Ha)avuLS9v8(T^63#B+50haU@YQ(&rMrhFScME^x&N8j>rm{;F zxAvh;pI-JV%YY3Fws7wwyQ4Y)m#E5evaOMutZL;HSJg&J)}T!_lVU1=FNSpNqo&x8 zwf7j})l`Q!i5>NVX*|La=DxnL%O1lzwJhz`6v7FIh2v74^L!IO3qQDozRHe?y#}wy z;M+NS?Y&l(UmkI2mNtJZrXo7VlDk9Ht}-#pr+l%PTuP~_y^e~m~Zwmr!~8yhs`hZ z)Q09N$|lCOqkcVPF$;N68tJ&)y`o#G=9(;=+C$;Is9o>e;ics9aYv%EX2l4TdpLzz z6UQuU2y!9@UL$3bx$s&>TlM){@&a^5e+42C=U5X}tCB%#q9WYX zV%@ITxpeh1Sg6s1Yu!QFx9uxf`F|R^8kSaDuM{~hY#$cL6nt&xy52wLi^{o>@`;6P zaZ-KT<|1?^Y*stMM{Xcl0_vN4D7rltQi^3+Oq+MYooXNXu?m8o7`N*^mmvEz=(77% z#}N1HPrB6CM?)Lzn`6S6;L)gBTWc7U@U{1Y;^`79;Q3tlq`C(=xJK_JOIs0q;xaC_ zUf+G^2v$41V54nw5y!f4K8N7My{!+=ZvA=T-Tz^pZW>DgxGk$9l9uEx|UrJ5*KZPU`xjBl0BNN_?}fUyNkv zonQoJ_z@xy(%)&e?3jiWNAnjU-=Ok*H%4i(;I6&kXO>)~A8@r{1~mt3VSBp*oH=Jg9sJP&_`Q2utO^Ul1i%olF_4DP=#z#uWGYez8@sDBj@T9KJYF*vF@ zs^QhV)ceT@*QuEsb-Y?Y^+I6T&vu0O&Bl0~hlhQ$CTm+tp{P))u~XD~-GmYEiCHTP z?R!;Xd&l0OAN3tZY6A@L)^KE0>;93KLHolezlgA=$N-RNpRj6%T&4=%%fcQ5o5-B+ z7Bey<%Q+1+y8u1xJ91d|ciXomFcwF?qg%SIO(M0`NnWh%iXrB|v>#vJqKk$mQ57z$ z+zuCFH_@Vc=nTP6N3J}AQ_bI}YcrqVm9>2x&ncSd6fw#fwm_8_7MD)4!#_gV2kM{i zyoinK8s`1jUOP@~BIMCP6KhSX_B9L-A0l(hXu@AZZor{6)id#IAG24nV$n9W?R~*_ zedWq6y&u8=N(T-PQoqd`u+rswsuf2vGLGKo7n|Z&DzZyh+k5eW!(n}=wk+4OuKp3^ zIh1vC@vvh%6N-y4?_pldE1LgxidyyM+1WLuZyj8PRZ9`gmV^?}v5oC`~=KK`nSa3f%1PYTgf0gFA%v}Q;SAjJqo(M(vaPl4Ffs4w>$Fe}H8Y4i4Q5haj z)ZfyOuRs)kjlETFiO$%X1yL!rXKkr7V9QdJ+x0k?RWN!#rFOeU_$F7k00>e?_xNxh zT~_1N$$1BnP6#_TR+vdc(xYS2i%LEn+D*L>3L-G^wg0~OaVbKX#0t5O4~}X%uH~O| zELk7W7a{9%Pm<7y`NVmc&SR?5XA2dBnjaiJiCS_kmj{zmw{$be%@n=7-c!G{mj1>9n6B)&O+`7?4OcAQjRpvr z3Zfo^pzGeS0$piu7vX^TK56F0SSkiqwg!_1Cdc{6+jUOsU#`GcddvCcKBa7t#+xjixBQLn8B8+l|lSz4ohP+?X%-E1qU(>{KQ<9WRU5$p2B9^eI-@0oK(YN%yPTf1*?H?cOKC3jQ1y6$kmd z#s~1aflsjlB=0jT66M~mEW@}`JrHu1ROyxm@QW~Qokjla00gKCDO!7|0VcB8xwqN( zif+-S`nFe+N119hiRhZB3~Hl(26j;m#`S7g#j}PWYNFvxk++kGKr{VIQWZjr!2NdB zbZ@So!!xI0iqHt%>335hTFQ3eRZsy>tAd40Ps6KlGs2XY3iYZl@>}O#2M^Bc4|YNm zuc`TQ5gac26&a}zB(hLLnkkatVFww>%L`wgX)Z*5`vv#ec*6&2t2H+F0|>h7wX#pe z0|>ez^97MdlNp{~b-LNyWs8m<@ge0=Q^IvdD~5@!#=YYO44sI zGLOS`bMrI{U*|FWG^g$YjH(9EdapRc&MH#Q_F zt^%C+{9@$|Jt8ceQ5e4bOy1UX>%Cn;_~34unf4dnUj~SjjWIK={GntO$;YmJ--`3| zj5_lBZ=XKaPQz!Z#+i5pfXENAG*ca>U(U0&lcBF?J&9Ba-2@FxqR0jk^Y9_fC{e*u zw~@Ue$6bC-V|Y$lsJE=segvXW+yCX^+LD2js3(A3P=wNn1UH1k%sm?7mzV)>cLidg z{^!UKO>D+(1D>*JYc&*rkYuOchh0u+ytj}WZJ9Uv0{)zl9{@b-AH}{K@MXd7c*-6y zf51I82HzvrD4@b}X!!Pt9(Jg>S3<0?g;Al(gb7FcpVtT&k@K;kd041=zy-9m6sQo+ z3VuVdGdt6z{Nx$iZzu!ophT;4brpFbiyHOYkI55cm}yyjAIJ~FgC14;W#jK*zoGW8RSw?6dH#UkI` z-!qNHmF4`l(7yKNuc69+@I=%2lw�uzzl`F8Zku@5~SrIdowwNU9LLQRB~gKA+`N z|8=FjhrNUkVZIm>4*=E%Y~_ofS~P(ZTQ7)QQ++kaoGDBFdUy(og6lr;Vv;|J{HX~^ zGwZ#;&~*I6ABzz6M{G&pL8w+OG>V=d*PXwGT0c)hJh>5qtS5qlPzSF6%?W=AXbm1z z_#pN5DVAjYoP}-WZx%Kz_Dca^~fbxA6DM9R%A#6wZ(?;^v(K(9`mJv%O{;Nvs zO3f3ud^PyNF@Lwa-d+bY37NZoUis~1=x`}Vi# z;j2P`D))b&$e}0IKjG z%=E48e`RyEQ3Y-Le`MovPLV_EpZq==FgyJVeeR+8XPhjTFh~zz6SmOUP(9Owz1Vh@p@@)<0zY6^IM}fv8S05SwM}c+! zDiEjFZaL??m1$zqeod{~vk2!pf6Q;+??|sNDIV=Yn?s;$(UWv1%~G|IsfuSRpW_8u>2psD9LlasGILboT2}%bfDZ=aNP*; zO;iy-vlrTUvEJ5veY(weQx;0xTs5Noi^?dqIC-yp!L6~l@m9}f5Y8Z$w}+*sc8c!r z_|7Hnrp2y^#u;VbwXTN_dyN>_qHNFdf+Xq!y5OutXod+alc4p4&g*}9YJYzU^vOhD zFdwJjHY(|KL%j2KPX`6IbY?3g7%08|@qX7EY*8AB`?5+lDLpuv(D-rKli(bk$k({D zVmD-Eg}wUCUqnfrQ@MffF8=qsi2kIR^6xuq_$VkPp#BXTJb@tbjKzB!DCc(X{ccpC5^|=W}!{nizfMzN=>sS@F7) zSR#oUV-m9zru+#HoI2Xa?4+L6YpH{=gK`!?csf3VNvnL6>&~kV(}!*nkg)S5heW*= zWs>GQqh2A?E*)YiAqO=*=VOonW4K}gDx-`@K-3Tk;Qio2nxc%oNLxJ>fZg5$fJGiMu9<))B4-g8wcZm|%#t{q>=N)7U`Mt6OrcjkY3@xJvA*St!B^Wh z$uN@yb>cr`s4n*8{!l}0KToNMM^`AHk*^xiYU!gI9y&f@mq2}t$x3?a*HDCX21Q_b z5)myQm$2Ul&SxFoXDf)Jjgppr&hIc{Co4tc{x}U$r_tol(94NzbLl?VkN`Mg{o|3%d<@i;d}RZbq4* z)%G0gGP~_{+AH*m0k8BOt9oj8VP2il;W6#rjP6%C%6+-NH-2ryT~c?_3BPTme~457CQQe_r<;LzwtS&UO#PEH}&06(jh;HE!TYzK>K88XTyjnchuXU;g2WujnYTUN;?{N)Zm`2&MbBkCrrB0?M6hsSm}g-s?GA+`#-U756y(++ED! zeK^zHlTC0uy+SO0dejlWFMi^cSI_e9V5TWDZDt`=ybklbQ+31FVJ<{f{N%`DBcHqB z0L+FMj~|a14-y}D-6Wb=Txaa~_W*YcRD8)xCK`OHXt z@f+rHruilodNH)9$oTO|NO8l#R`anDz1O|R$20MljoyI%B7%!_A1!Eebt2!BlCZ1I z!UHdUa(7S<;1h2rI9U;Y>msuMNJ4p=1ij;r_?%zeFLtn*db$p@;QWI>m~{MjHiWxz z#7JXR@21*`22|djZGWGhfIiU3=O7R5>C0v9hMDoZk@h=_1K3DI2J*`?#ePfT$*jCO zY`J~Im#gZqCEm!l=u=hM)?5zpK2v874liVlI-@wd~&U5O4OX>TBU_ z`O(@U12|y|a1cJ)z=?+ql78H@sYk~as(roBWUp^F?#1U!awe7;FkF|Fsh!j6=b)9X|-F> z{>IdBMXZMOS@pBbgBaq*8eh!xVo0%g{NO&lVtyz0YjJT>ftZfps$wUK6;WgB#2ZR7 zRb^jK6qnRF)h>3t#fhQ!@xYwrT&Bv7^6oJ?lqy~IK5c0VUX=({8ZEw)r&%?H(f!~s z$3D8--*MJ(kYbj-DqHdjQ9gxCGh4frT1E?~CnGPhS`P`2^bu}&40*Nh=cfk1Haz3L z_~QT8_GW*%oUp;*)&f|&aS)hQ9UQd*5;#HART{RJXKSgaa}Q(r$Y}8n#ngw@$IN}U z23h*6w(?cQCood_&}@yh@$GDMj`)exkk>~Aqr{`%79o^40*1@MN50-nqqP#NIah$0rMs>5j$CCzc|R95`z(Q;m`n)J=BAC4az^IWuIHC*k=EZ(^Nt~}S0L)9@U&CII}y$3FE%xemq06v`; z{B-0bLbqiz95HDW)&IH@Z}5mL{{FZ)JcZRhrrd0E8Gg1RE>56ZjP~$pCd3DP{AM5O zI2SF}HqKVte0R+;j6NBU!gy=4iL4kodH44l4`w@y-|)q<#|kyZY`*)}t@-KXq?5a` z#0Jt>g;~-riIw;$30)##EZ_K=ewljz^;1(um0i{irQ{l?M^T5|<&RJ$+S{h1SS*jM`Y#Y zn8dlHYRsjWsP~^J29vnLgi1m-+{^AwAHCUAj_>1l)SnWedw#&+m9P?4Gf?VQth#+gJbcd zw=T!@zB_FZ&R$BE_gT~KmyeF!gIOzE8c+qMmko;CI75HteIsF7*Tp?p5u<&n{cXdo z#qSk^*4#@Hm;1F~RJ-YUE)=QT>6Dp2&DWKf>ShOiN=3`8LYW3s(bA_HY`x;@!%nIN zP^xDNOQ5=X=y&3gquRiDQTf?Lh*j+VV{)-?gPPD~Z@me^&IadwL-&=b!5uEYecaz0 z@S~3szPl5PmjUG|BcR9M_{K^~?>JZNvs0Oz9g9jmTWoYOIh3bXR*XpH-p{Vl|5z>l z8O<`A8&LK`;qq_{)ml?#z}>4M1mqow0cH3#T@~?VdcBcUWU#5TOtRlmI`FoZLQRX% zQVTt`WIFi7F$FJgPR0TSyIlVyZ0bJa0o(vODq%^2F2DoxA?=N4iepwCo&gOu_amf8KyFN7S);{;9~8~muBWYmV%B_szU54;=+N|Vl@uB z)lFR~MtkNEgOv+B#GZ}hS10zd$FQ@9A~|=e9I|}7l1^@__T9)bAZHmUKy&oPShIAJ z6J(Ov)4>ci1BGHN?a4JVSJ|13^m8Y|I4o@k6y(R?Lq{nWKR(4#bF?VC*KBcq1yL7n zekOO8S-W3>XkhD&e&v_%eviG^Mi1H?iQ|jo670QY^wlr!kcfzlwu9i%e25KitNi`^PgxWj;_pJ?$(JJAQuBZs`h+A*7fBU){0l?Y5FdhZ+g z>crsAq}ISO-~%tcShdkuh%C01R9MU`1lLkjE`E)%G_Lnozu>p$w z1YPOXdTxv|H9dyjR7IblOrEsHn3R?`DMo9PL!+!MEe1yNz;0EC z^XGK$=L$jzB#{o$!=gtMJzxB({%QZh^-O_0Y5mb$jp@!DXM`JQ+6yFLZv^WVjBM@*! z{=F<)(=@(G^sB6kDv5fZd1aKMhu5E)OCWP!`Zjl5YL?RM(K?h)A#R1=x+>s^mCu-bze8h zaY}`Cpi`8!zH>PlV%iJZ3%|orDxGz5HMu%Fzz;uuoXi2ZgA#nKCo(||a_VBr;XR4- z27>s)I3SbF-6S9Q@?lKnPQ=tnT>p4q&)MefKoV2=u41wHhw^N{3Rl;>ZO?>xzpz}$ z?KQvCcOip=T~?MUL6z`5k}=V`gC`G(oNzEL3$spF%d7Xv0M)rB>Ss%gow{+PcyKd1 zdS{MT2g`VFe2?+*#NBawbGF9l z-0LImOvkE<)fGo`5v)@^h#^yai{opizJPb7u#}2L){3D(Q;pvpg%q-KDO6Aw415Yn zBuCQa_$=!thgDpLSecUp{ROFxU4MzaQ?Hr^VzsA-?K+|Pd$V@9j#K{ghF&pdu19MO zzwx+FoO=7W#H1|Or^A;Ta4*B8pp`p(bb%1aoSQZgd8XZUrYUx=TzP{uBpdbz}E61t6+ss4?dj`WvZ7 zAm$(Gj$h+ou)uw`6+t{Ub<(H0f9-*f59TaB(a8H~hN2^4-ZaL@+r!Av$yO5fQ0+Ru zPv&%*pDI*Mh(Fb=yr56{=x%gch7RwX(pY4c&OCqOcOioc%!*6Nq%8+uO|}ZVW|}!Y z5W62%>={7pJGUlEhiU!_Sv0=aYs+HHmNKHd zJ@mv*3^KtXPhDiTcc|WcD_18Vm!(c#A4=aR>e=V_hv^u|{w4?+2+MiKr#zCVB!gSC) z9L~w)ISPP+CBdJuDQktgl=}a#U_<}D4bpki3)#HK5xzU9E4y0X2FH!Yd?~9D>`JLq znBR^sbQOmfTKaZUIjB|2K2VXRDbz?|XccD6nAgV`2;-QT7RIwd3MTO2&&-5>%VH5hAe|Idz%d)lz22h`fz z5+AmemQuKK-AO;Ejf8!D9UK-MQ&Nu1-@K1u)Dctm=dx%g2<{_CJULSki`F(PXN2-j zTbkMQDaOW$ILnrN#~>=9cpiJ3>TSSo-+nyj*2CH@>hIZN?RFJhkSZkwS*2#fmsudF**;ahZ;_+1T;CfR zU_`kI-d^a>fb1xJxt$9{pg?0nLo99>oA7oif`Ys%bGv<3 zDndrX#PCtyiJjK6kaSFIIRD@T`Aw!$lVMaBeY9$OA9FjKQX=n*JDOt3hh{R?mI$~T z=lm65u42o;1+J*=K8^?Giq)UsKQ0i=ZM6H0FO?wh^7U)C3Fzopab?{>3Yp`+o$N}N zZYhVLX}?Z%lSG!SCG2NZMtLiJEN2$Kl`C5#b$94qZs(@BLH3-y0&sJw^bp|YAn!cB zh5$ErbE1@j_8&J#-}Jo(fsBYSmcVt|lAGz~&*8t3mnXN?c=RFg4&KT07rZ=7oX?j7 zJKjW1+{ou(PeedbLqf$@c@s+3LM1gO!iplDQ1|dgeCL}P{Sc!$a-2$OW}GD{vr^|_ z+Eu2%!?fy6#BjEuQQrhr++b#=d$P^G9exA>a#vZ{fJfzV0JJRpSIhaN%V~+CbEZ+~5q@|_DJP(1% z^0PaiMj|^yFJz5IuYWZe6j#jCZMdj>T9_HZA#q`GBKn&G9en&;^RVC0-_@YV(|2)U^XQ(j}D)3ag3OTR&I~-4{ndO_FJO2YzSI$J%6EGa)oU zdq2h`+Qiy|P;)vSHi{5k5nx$`C@dW=HWZz*(>2Z$tEBuY2 zfeD^M=XTM-YFP)n9>5|UJy6Ft7`3mGlAd3|tj%lDFRP<|e%Mde^D z*pNF41&XanPWkfUxS_1b#Ye@~UM7|Ae>#EVn*#sn&pL3X{qWbBwuw|9WwkI&@j>kZ z9{8a~emTl!QT1Vqv5-)I^*~Cv>B}PpV|BqkwX}4(CP}5pK)g&R=s;4<{Uplx99EnV zGvfK;yYdOqzRc>5bfX6M^$>MO!$sv(nS;TpQOT}p!>X)lW~bDtpG1|L2J2k%jQe7n zCC^*O-@r?0IGKpYD7g2Ykwgtp5qibpXT~b}`oG2rG;>B@fZgr$fA-q1h9;J2)W*_9 zSfN}-p;;+`he>87le0_)Y5iQRtH>$jk&!+EK*xv6_Xsj(+yL@~h3Hbw%bMLT z_1c7`uGgH~A2E^_VWu#o%q4{^ci z1aqs4Cw$KF*SAafb#h~}-_K)?YvVU&ug`Z!6`LEo9fjrbn5H^nv0-BIm)l>kt5xPx z6^Wqmc*xcN*j3t>2_wuQ{b@XAtMTk4b^D2`YTb;XTaD=6KtaYefF3Q)Pzrjkfw$}e zd071LPKLvkKdKWYI62A`$DGQEW>vpr`6!`EPYO$H9wY*NRM15NF8rL>CveqDS!$xs zRUM`_eKO~Jwy;vqKyS45)X&FOm$=dA7~5-a)z#*`SOt!AEy8?*|6Yc}jX!UgBduDw z1_>Cj3q5_1Zy^mfje38q^4X+108TUPg}S1>09Z6!@JxMR-;PU-?>AK_ z!VmPdCtCC~{Bo5MRfUyt=lUmvwjEh6s3=V{!=1zf#PAa!5C+)#+%AHAROtm$d4xS4 z+y$b|1yD&7R(Lj_!EO^hd|ipOY0&xt{HT>CwFr6p<8ZN?XHg$6t@gl=mbP#!SemW2 zV+(E5X@8ji_MeInTOz;%n7dS4h*U*8NtcVGvkJjJgZDwAzl&Y1H2+Y9ZwPb>bgNeb zpbrA4G0;Jf&TWGnaJljonW^V;5gEp^V6)NKyk#0-<9C6i8acD|7Sod@frjef4|1Fm zD`>GF@Cw`FoX;M#Xbv|mD0>h<7-=9*l~w%^1}GH22PcOoaH?7`ZI>}8iJ$8sA)cY z>_CqxyVt`Ab_(b?P<~i_u0OTWCiH~-Nff7JYfHQ{%vLf&5EONyr_4fDK`j z(3MzO_#-5-fSfm35f2^d#7c7wc4Wx22HqGM}VVEt@J2A_*VV-*LKRFepDe^j2>=) zX54uaZG$l&`b)bwOoP_<)us7}d`|#I0jug7X^A1r9Uinc5ZNdjpo1v_&<+CDWjj2N%MiQaT^fSsuP&NP*ql0F zJ@tpg+jwwiNHY=K77+Fr34rE$KnEa+*wl{axJ2*IA`zviedEt%F|ELU@&NJVMprK*HfZZ?OAfQlp?+b@y zZxKD@NAlhSJLlh${fM`=k^#jxhQf;^>Wk(f`}HxyV8fUq#&U25AT-mI+f`r-@$6fiK?1_ny(07jYySGUtxQvpfzT=_k7L9E!hg~EKvxk%};tiMmlm0gtQR;7=@4poq$Q%QJ(W<^6SSc+>l*NAFNx(n$G%*Kny z4Sl)4Pji-~_P$f}L`zG#iKAY|=s2=aCVP}K+AF{;7pwq$1Nws7N`68HsD(;evaTy4|Y=Vz;(#sw>akc%*iStCtL z+i@F^LBt0ek$(ei()8ZzH$8UMhfb2uQb8uv-*YF^{;IzBI9@1nX-_dQ1hDGs=_@CM z89?u%X!CAuxLh}7_LG_GdD5Prr*2tu8aRSXUz#h@WFy;qtA4l~P5qI-GhMTkFv_-A z!KRk^z9n5{?Ldb4ZJcMdEyhQ2Fve{N2y3Pb^gkwEWBM=H6}!kV`Ev8)pDPDPLqXC zK#GaAvl0inDdhf8z0905>-Th0bA&n83>(M~3KX$s*yc*3y?x}f$#dSGi*+CC|2$++ zO#6_puE&~EWcZu+Ax+<`xr`e>FxCoRg%dcpzPmqLWtzr&g$c5EzuY&%Sk}0^ip}6q zlaL~smyLHqU@|WNp`lx6KEdcB^()C+47{(jCtGssfB?M&UpH`8FFp9fd&{NG4Zwz3 zI1`RoFH@|ucYCQ_!p2M*873a zrNKd>RH5!KgJSiMLIT`k*d9>aj#UwB`D(%=2~7xE>%8;Ga9nqQW9+xXN*JKr3jO^v zuKm7_a{bA8*V zo_y|?k+}(@j(61HZUXHic7a%G%~}C@dTp%MJg>uxrdAx^VOKurfeuQ@fd>H2BWzgv zxXwOPm9(5uLlFfMVMX&%dUqWwr;1JfEPQXm^M$`1_u^Voe9G0lzg8)b=X_2pG(%Dg+e3a)|5kQS&)Ta34U1~` zi2FSp?mk(8%btdqjG{AZv$KGV_{bKEj+OWX7vM>R@WA)*!@<-X64Aq@=$d+!? zV1!VSr|=DmGJIPY8qFCUnB+tl?991&6&`mZh$Hj9T26lTYQti|)DSaG{aNLI#Q?{jgNa>+oPxcaGhKzE?^1 zmk(pg7VkUJS3Eb2iG_MDk0lo->G*@LwaJ}+qj1vPzCo|`4+uil@S`1zENrV@&F_c!TSw;UR?_yRceP9m z>Iqjol0;QS)iLE)W-z=Dm(=r*uy0rY=#}-xQZ?rG|;Mt!BI06)g#^w!xkgx#8=zVJqBr;Rp94S9xdk%sdWD z(7!zm=_tqpsPiE;i1l}0Rw&BZvMjK9DK*$S5m;;1H{gc!#1NP^jQvYQ=DVpUzclJv z!`*9DwnQ6TUiFuymV4_$5*7Ld6mg%5!yyNyF8SnpgJzWzp&6MngsFSfA1 zf4tz$HgFAQGjAk+ULrs*!UHRlwdNP>c7^4^Yc4_l$hI-J+Py7kM74V_WCVAhd`Kvz z<~qL1kL2*u48qq z8=zKHriFZQwM#eR9hs>nGwtFK+I7ofcj1%T#+#sFYjjh%2(gRPwCt+zLjpVW5s4~A z18)7Z#xd!Ma%sYn(D=t-8*Hm}82Q+Um5kqyt%c=SpEhiB1K?p?@v+AoBkk?)GN^`) z@kNU7`HI|Oh<+l+oVe8Qz#Mf32bBebaCSk;feO894(EL>EM3*vA`i+V>>;AHh@J^% zRoCL(YoqPWnQ;kGxm0lT^nOuqSs@M3IzX@=PkR0qRz#CkE1GFW%g5VT$RZ$h&?5t% z1I78(Tj-z>?c!KBFLY!!Dcn|Ru6oC`_-!C|UkhhHejTrqn(zMWBT3cK^*;lyyaTp1 z5TRX#7gfCoOlo{W+;+`-GENFg>+NXIu#_QI)*H!t0}$y(a#8ZetEI_mZ+|@`g&apr zV|x=gB%i~nLA55AfilCM6oyAyQACTmU4sB3oS)y8T=Vn;*b7t~@Jemcv2(^=kgL?H z9HD>vhD+zZLC=p)|D~Fv#P{{vo$Na51Q7O7Sfdnn!q24-%e5AZe_|r4a{Yp|2|gXD|oUqhNQVYgLU?8 zSM^-4AEo?>ggba*p|^Huuc+45-Idc9E>q78TL{273w{b^ga)#zCFUm_zVSNMpBgx? zYyM}>=2#&PG%|m*dre2xV4Qj?uAKsO%BhiGw)EOvbTELP-6e$MJsO@B3fxb?Lo0=Nb2XKldKWW%T(t zV0Lwt?=OCBWjW*uPbt9ui%iHJ;k*(;bIzaDWyc3$z!0%3W;`FduZz8AJiyxY;Fi3r zG7&li*!iV5zXtHoLH@vWRru}sPd26AeBdlhC==hnOI9XJ0cZZ-`=&->=Gk|mM=DAH z3&fcTi%{%j$y)%iye~HqYJTf$8AyNqAB<^mcHCHF?k)lY3xhF+D3$on0wicpbfyoK(Ri=6Rx|ODd3SOy1<^m3d6YaF2bS1PngmOwW*`-!g|No4hY z#}6T1(*dpuUPGmeM98Jm>x*BL5Id?%UPIjv1-aid=M;N`3BGmBo%#|`Nr6JEtQdld zx&x2$8~tmmP(BE1BF7a<_ZlKCrC0aK!4m+hk=hGO-rCn*l)eqxxaUojlqm(YaAA;< z;jH$9Q>NgQEcZj-_?%+^1kjyH_{V{0q}3?EZ;Jz&kndkmUc@B8Z$xnTVbX6T?dmlY z;~THoC6xj$F8PSj1WZ)HgM?4uh?*Cs(SbFVK%^XfoegDn(-~Kka-!MCWw7MKopjD# zwNXsxaQ|RYjng3k&xh^nSJkf1$~NkY=sW7jRa{!V#cIU*c#@PVH( z7+`0i8n0e-lBPr1lOTOP)Pk^fvuM&L4B}O9k~w%HxN**=X%-%xB%tm5k!g@#n7?|c z=$ui9fVhs9qG@U_{y|%PDW9p51R2}K05ylX5_%6@{Bb1U$1eLnr0vOm%Mf4N?L+HA z1s7RT7Q{ zNnOugCERkzw2#k(2!7>m6;`B9>lG(}z>25CGnp2Mr^*t^OO6kwn0vsr0P8;u`}s8J zmU*{59iwE(%#HH#GFiJ?Fmstfle4ZgX{_TJ%HZuu11xQgvWe_Fc+)BtW6uZHDeU0e z;5{@zo(PciI@S01PFj|STsEp+5v6q%S*6}Qee0OLx?X3oO5h&`XQ=)t4U;$k^(U@~ zNyhu~(B69m6>{)f@L(^&2g`~~npjIs>Y^<2=373`YWpqv_A?0ImX_AOB~p1AurXLkh%dNb zjrNB08oj)7XEjU6cp%y&+KB8XBrGEO(=Ovd%Tw9 zvmj4^`&%(a2);@8%R^JXi{NWO7k&%RS6okfZ!81ZC4-e7 zJ>y9*80hA3+9PjA>z71O-k<}Id`|^;!uMsZhWIDB%NF?f2=~oe&yGQ=)(=s9139fC z!P;s9eqyPCibfr#pVk`t?*jyyE|MtS%9Mv>7`)#+W?2^+;$M9|7@CF}#Bs~> zD9RIutDQU#(HskuI%5i}_eN7v=j13!ggNk3RiS*;1`&11@3Oz7N7k+77FQJG=e>+Y zzJNl`@#Uc1T^%f&vFF#~(3(0)9X_xKc&FzeGyR)C;?dz|RejNWgp=z*XU(&0?Y(pA z=E9TLSno^eo|<4)vy6U#TiaM0ns%v=A5ny_Gnqd)6fAho*mM2qfWc;GlRT(M_&*3a z$g*?ma<=xE#@h7?X{-nxu$DPcW4`ROc1P22wqBXeK@hK zMi*S7%)5)#XmfH(9l{n&a>7Ez;DpX5J1S+Wf68b3tQ={M_WKV*FvNCBSm2s#=JASZ z4CBDwW@sp7anSscf^nIM=iiBAIh#v=rlw>?-qBi&^mwaR0Sd(u`ZU6+5P9&CAtKqQ z|H}o)Wt=H)$|S!39XLzn+CRvfK)eQifB7+lBj9UAhwe=;cO)h0;qDDdH2ikPK~gB_ zBQ*B>BJu(3c?zxm4?rZS2O=XNGTz?`^|`%kVmkCy+Fg?z90%+9WE;r<47;t)3iLKWAQ_KqM-BJV7(o>WC;%#O z-DCOwd+cjc`Th^wzOw~5S4{}bYVw{2ALwW(`SPZQ0uQQAaleHoTL)QwY#|;8fNoS+ zF2!o`TjV?XO!#^+2|wCXPBz6*JiirUxrQ!3?hS(Zvh(&E&8rDb$D!;4W1D3ms{uab zO?=h3RKcc7NR}_PBOtf3u&#*958$<6#3D~Ikdf&6y9I+9DUUFDkGO8=&X~pA6u(XTUX`PNHjP!c!ZOSMD1(!t(5*#zLOYb;ps_n2bme zfTNaFIwYt$E(m8XEFn2^*`XM}w|ePY-QFMoqUnL5($fZpe9(>6=ixc4_t$3W!Qz0g zr9i$0591=4U-J)|#y@cI?Ow%zcf6E(a%rQey33Eu4{jt&Zpevtxg?i9BhN2YpaT0I zaO5A8?PLCY;-EV{5;OdYHqz3t6wZ2hy@$z^`3oONt;PBE6d1?!)}g|fn?7ZOA6wB56VW;LN3i83FRvg_vCm~ZP_#JmPvk@E48#1 z6H7jIBbwk3$?q?lV94OPdTSdaUQy{Ep29sm`qY^Rx0EcN$BrKdY;&9`-jx&b(hZCpLT8W-A}k zl_f9I=qL zUq(jSvg9s7@9byitN%TUkJwhD*99H@!Oi7J1LH~A*>SrEZP$5U?Y<^-TLu&XcSmCU*9*{cN`m`ox ze8HUBX}if2q>)z=;JaZW`rf0bWYwuHrM*Vw4}HCs>Nz}%lAMkweZ@!f8-QzRiR%=* z(nymZHtvhBJ0$WTOiGEV+fMr%a<<~6+V5)sZX?F7={VVnhf`Hb*3kqO@FDMAp z>za4${UGi8iE05F3b^|ZJroo=7UIv0e#)gouWdhbD420W6ia$V z7A^82oY?Rzxu=&^46yf%%PNxmP5b(0j-LKm`1FsX(<#hA11nT@wU}6#BWY+W8;`7} z>AU>n7Dg(A%etoon{dU>b&Y9lciV-e9u6BB&!BgP&{FIyr6p(%Ew%pGo~6S0Mh*=- z(3S9usqDCfjE~O}eM)z~;1d^x&NkQ{0$j_oON@{GUMdmK&QLAalE)zw+`1lo9b27O zw5hgxW~sou8#T($6Lv%Cv!C22Zb8Yjz4|X=yww)^jLImz2=BOaB2AfGxv8ZFf+Vvz zM0A)=j?T+s=mn93R3Eb)WRVAIBT`VgNTv)9rJ&CTsY2d^yIS`H7Op~t_Ope{;e|u2aSYUTO7AHA&JnjfwqW@8pd!Z;qhGw7WJ+A6dSJN+*AQ-g-RY zI6VjTL$)1EcBi{)S??2FTcHh}nI%6p-O~xrxSVR3U9*`OwFdpoX}pozLVHl0XD$bq ztI_9A4pOB{-9mpG>I-QB>=LA|kAYHPyc0Q^t=gim_d45!7EVQTd7f^B_t;mJ&6K*;yI$aU(qT`(=^nwh=(kx=G{E!KCwZKPxtt|N>y^sp% z(z9=?srD`0u6{L>VCncpml(tKJp~2-No106-PQXG<`QD9l;YblTG=zq#oj}ezJn?& z1V`;FgQ5DzTO|Z_E=}scN{~qatEw4t)zf&Y^M`p6N9^#RIO#a#H!|>MJYx}Uml%lHH*5*_9K39u>GcotA@bGxY zu(Cl~iBW#ZDpGA=yu-&eIH!Xm$1J5p@VYbUh~vowNn?aEPF~YU2Vq!d9RpOMe=j+l zA>s1{j*?dr*Ppy{qLGy#irH&W%;sVV$STMAxQweZssWqT{v_nI6}~YatdEVJ(&Jj; zHa~Ukf~V;HHtg+L_u4keO}$L1%-|RAqD@kLxYbidIrB$$oE5OShJ}AY(P87Y*GxCz z&6qi8;)!PAcs+z4J$~FPOB+bhqrrv3RSuWI4@0bZu@h4No< z=I@lI6>|OiC4q~(;L;wJygmei5B=CCnx_4CVLv-0Ac{EsxT6 zHv2{kSG9>6law%(wu{bVb}3xG5kSR1m}O3+LDT3z*~;S6sVJ)P>ZR7J%I3vAJJLMV zpyt5~+(n*QzqT%2k{R3JzIr8XBQ~2-XGK!kc20^!GK4uT(8k#M>Z|Jy$<6*Y&~4;f z6EIF^$FRKUPZ6`WFCF+87W}UUBq`u5wN#T;g9gZY}T_@^z zN$)$t9f@oyv5_X{;quAy_j3rld$8y^+R0N_J9R9tO+`zgB0L~Csizua$8yCYCw`gm zMg3w0);jq4ljDGWF+uhPuG_Hh`D<_Hn~E~1lv49b+CQ}U!Qi&B*+0LP!p~4%u5FD}!pw|i0 zCny}&HugAvh2Q2C1J;iHmMWVud=nkUX_%xlEkrY-ZTbCV-@{EFDD$b%q*bXR4Q~6^ zQ=7KV(g>zZClxiipWb)e1&c+KF05-ls1>}VfE(KO!Bq8@)al*UYsZ*ZzkZ^NAg^=O zs}!m&(A9hO;*>HVTOsQH4r46Z*i-myW;*S45vYuVAlRX#r9n?hte^bRm6BUC^WEA~ zG`#v*!Kvi0V7_VZ*~S2nD0Kf+&$qQAAEbG*u4sCq4i!g6@U zbW3pp{oXvqnm6^bm_VkWRoe}cdw<^D@z*2a^FbnP{-!q?PcIPFC1$9Ir-Wcq)>JPW zYeZeqyNvHJZBA<|&W@1Q);u)RLFx|3s02J}f-r0Bn$UOj!v5QKDkipGkoanfp>l~+z}XHXO#Et5wl-Q!T|A3@ z1N&6Kdx)DX3XV55#(IEwg3+c+G8a**sIx#8Prw5I0;~G0+S5O2IqadLH}=~fb%RAE zr+!!uzJ`3jztcf0(1G2!tlPV3Ecm?DN&GD6OvR7+QHi}+O;G+(7osA~cXqa4Ms(`l zUu}h;dO;rCl6?%@!cX}f&pF(tC4g?y_*GG%gi|_4z66^?U&3R|zRLl=+0L7`NvHkG z%2!@;Jbe!6(=joptfr-l1(P6OM`CV%N`)a@z`pY!`wr3ZY;*Q*bSwJzz##dq)n(ce zVEIjpUh0bCg-3OsD3EWoP=@hO;F8gX7bMNFTP`^JX;?Vv)ZE}K{0+L~iaMAXz)*)_ zM1fn8uhqYN!KnXW99%%;&H<|%+C0$X2wscBrsPh zpVyZH{2aN0AFo?B(3!B{JKM$I2{pe2lq!9{0Hg|6>2D72yWbqTPdrqh`?wmTwG_n0 zQe)A_A86HwwZ8~%MpG+iBA!BFu!^?y)%7fm;({;&x&<%%R||wNl1(fLtSz?92j68N zUVigk>geXxd0VvG3c8FvEa~RwB9dJ-6slS`fRZ4*u#K+;_A8dqeT$w<#vmvHBz8*2 z6!}*e?6Vz$MnMbY9Xini7tih8<}JOZ`wN@7|B`|~3AiUevmK((9~!Lxc8#;qaw%No z5c+4XxdF&2XAm03NDg`|3+`Gr&CmiE^H)dlWZt3x^VBPW`>{szKyo5@;oOy-ld$6% z3a9K%Nnte3S6Ek-yEO}ekbe6nZ1CoBL)C6zk2ya#Rost^)WgvCGmy)Dtd{ zCGBX0ZgwJo1_|CH9jBNU+;T`iT#V5rUJ8bA05usw+ELE1;|C|y$Wc}kV)#L35AW1u zOzvo}7o={13+WO#x|~PZ=2nm{{EOA zsltODG##6Cv8T8``6O>*M!P^ox4vdanW}s+s*yr&UKt*vV~vta$bvWL4DlHh!DmpW zOZACO3lA81nIhBA0uz?&2e-8rWDHTPX7-XdLJE&qK@_i|YYx80gw zWB9^wER1VYg6WfP@u#cdtA@#Af?Hz!`c4NKaa$_;uY9!}K@YDpiMJ1UbqYZGwmUDi z*7C5pM%uam=fsF=Lc{Z!{>?ud_2kHwPTewrA~ub7ODUlLy;HW=_^qp$cF_C5l%wOp ziom^>BuLpOJwlE~SUuxSObCXuidLOf>VMX2(b;@$u2CU(F`Y5PPUI>&QD{oU*x!X+ z=H2dXD(vyB$TYp)tcwf5AR9BO-sq!|83bkb!ug#<4OSSG1q|g7l3<;hvF*SEqrZNZ zk1_gQNbzhzyY;502{jDE7~d-OF${=DJ;JOPba5{$x;>gH_izOb?oeZP=ug6 z3Kuw?2NG@KIns3}c19_mDSmx0j#py!oWbbexaz7#9*?^yZJd%rNwS;$3lka0DB6rk z3(oQYh}C%XD&RI5Kd~Ocs}y2*KxZ>J1haf|kIX9sGf^=SofbV)-6I7;@s+W(zCPEC z2GXa}1`Xd{t{bakX0CHA&R+Tc)A5Dpj9amJcVTf4m`Izb-?y<;Xp&e5Lj98V`h|m4 zaknd^?|Ix&Lw}|!$<4n(`!iD`1e$F9gZBf+6mWIJ{ok)H{6ln`NgsL?!Z6|`wR8tg zUnNTNN9bHXYi_($Ea4sL*PgOd-&=R0JQX$?rk598eqz5w>$#MB)g!;*RX<@hm5Yd; z=ObK`Z*fJeefq!fF$#-%D6?AUJGDvBVu?N zw|#%hRrb#bGv^q6ePXG(&1OA8%6jW9)cC}t(O-9NQAQ9RrNWcHV39G#JM9ks8D1u%Ztw?@Ic1ekP6bs=$MR{O`{zIY($BY^{ZCnFyihu!*dL23 z=Rng#R)?@&kQj%&+@VTguml3XjX|Z?g?uj(Mfm$ehlC<-O%UWQfAjeIq{QI7f0)I_ z`B-Gu3|c+mr_mje%9E7N+DclH(F#0c|0ZUCNsvywsF>68(lASwGOqo80iH>`Eq=rj zAI}%S`T@8lPyw5*juSUgTDm9jUc#pcMLJk@5Vvq1@!Y-m2PgSUgox^ZBUCq0Ga_HD z>`=}%yORogzn7ng>EzBo|2*?bPEpn6hve>Tr4of8(h&sH+f&r zkj37>$G4JOAqU>F-2S)DC*{vW2f);~8>LJTxe(R%6V+b+IpddzCi)J{{l7+od}Tk$ z=v8c`>MN~&*{?6;-2Hq?n1Zfd&Xx`> zRpjkh;9B($3uF3Ur_%N~mQ`h$#-%=VVB}t+zq!tXQN4Gqtoe_3o#Z=#JTJX*#UGc|q@?A}BZ&UgZ4``gIQF^hN>_S;v*}|D zMe*}RUh{Ad9c}M>2I>2^q*ywy)i?~}5XJsFo@VN0uL>X~B{MZ0{BCx(|7R($3A=M_#pYRYPT=r z<1ct}mj>J4eoZ>@(|YCs$!heSGviu>vopv}iBKjTPm0TBk-Yq1%U%2_jf02B=Cz@_ zA8>G9NW1R&Om=%D7L6|`CpR)5q10U?FiYHAx!EypHFMhS&MT3a^^P1alGao2p3aD_ z(fq4*Oqc9z&P>4gRkgiJS4qd|cLO&^LAE~3%r8<6;sc4GP-pz=BU{4(lQ;mggE=rx zEND{kab2ag$&V!Kes9T;wn{^TnrdQ`xu=nBRpY9BecbNIiguZbt0ZP;zoW*~=4Wi& zuP+7(q5Douyub3Y^afa3%hv?gpv(KPkN5yAp#FkW#Nn2g9V`ZS16?0^>gvmOSMDY9 z@}fzq+(jp*te#-r;dq>pS;MGv23USnu3riUlSX~3uX=hyHr0y_xS2FdNeO&pvG6ukbK3McJLy7-LuY;MOHa63=8~tWW+q7YtP$3gBkpk-YXUmhRT4 z$EK_=1?06d#hk*S;QGqLB#8EP@ZsYbGtRb6ZhM2UH{ zo9BRYP-92iIm>j)4wPT z7J$lp{oy{o{3jbPaW^X>GRwS&RvEpjs^1cTqIlY8*hiekH*QVg!ctvY#cH+l-00G` z_HSrzLc^Q)?^zWKq}l7gk5+b`C$$i)v`4$Hv$3s})LbKs`qVR~m`5b4=~YR0{(0fy zhCIu=*ZT-_P;m?A2Y0V6?Xqap-F72d4|ULJWd?+@DNVWPwPiLg_s;2l?;LZ(?&2Nt zo$=fmfpe=a`9(DeBHtzIURyJdqKz(IRua<)q5v$y`Fg{oy+PDu8IeV-Q9>4R8BNMW znc>C!ds5X(G)E6F@MT`aSU*^o4tp-%z#Q}oNGYdZ(F$5Z0S&hlOc7|`Ta~>@RJ9vm zoK-o~F83{AYySPE`qB7WUC%GpDknL1L2RWIOq(yq-Y`ZW{xN?=m?4RPGy+`fhgq*z z&~_afll)JnSb^1eGBCGy@Sgo-+cxb)C2PzBTY3z7F}zL-02 zhcC-`S+ev;&S7WX{1H{4gcKE(OrjGS3)(-Kila&U-^yNaE$H2#i*iHmqb^*DEtU^0 zAqZCGO%L?rETB0HJUFBGoCP{_fou9a>MF9f9(_mccsZIrGMvC1(Np7Rp*f)f1o7 zgV*7u;VPNb>iRCbyf)=Wf6__Ypr>6L+1G?N3UmRSV#x0vMENfn zm5qNG6!_cLXN0&23JZ?qk7(lWr!uv#nJwD=n&ZHxlo^4sTzKDfSSRpa!sKf1ad2X? zBv43W^(XJYTEYgH6O7Q^^an^{mY>^D|)a zYKh(MI>}+)3UGFX*k#(eeO}%0Co?X)f0tdDJ6|&0#c9x*q|g%tJM(w5e;L*&-8hOp zr@=~#5gmNJPiMi~S~Iyy#Z00aX`>{9V)CZD@sqG=XQ3M#D1gp7~Sn#rO$b38;jqRK3ym6heVWYN!^W$gxBU@`{I~Xk}!Of5oiVPT9D6FyFPF7 z&`hp2_Bpg$JF1<&WG_+}%z@+=6*n@Z%HAlGF_@a5pw&JYkEOf!P#;%-Pg$>)U)#$7nUXRxHJC4Nml5a^|PO zaW9qK2cpfrmc(}qb2D-6nS`MI;FzrdexzJ6JYsuSWTq^yPBO@a__M-#%LKIiJyc*0 z;463wOwiAZjOA|ce=|!dKP-VJ`GPUExtL{8AxPqxd zSiRsGXZ~_bAj|lnnj4N0|h$lt>2^%#?Q4HG_RDb z0$bm^pMj$Xg62sQ)rh@3K$KP$Z?CFwCrlOuv*no$~fDqN%XgS>xwRhOSmjN$X1%Z zoWnX4M`OW1CZ$F#vJ{XNs0Z)T?_MtoJ2v4kipp4bEOm?bn_HjRAkJ9k1sB*Y@vq{^;pQ?|oEnx^^Dc&w|~*QH!^K z#WjlD^>yEnh~5vz(D>ypauOi0TWxP7n(AXO8*f%Jhbc1|9ZndSOK&2CU`T(z%;Llym47&IfB1T{Qm5TuRDWi08%p_rlJ5Lm&B&F}olu5Cow@}i@B_H%S6ea80g9&PfF{2X{1~X{6htt&Z_-4fQN%};!Mr|;62&?VM3%19MVUB_{Gu(HwdKODKtK(U|p3b%aOM6d@qOa8z z+qq5A_?Ej#Z)Vasq}HSd== zaOwi`#=7FwRH=dMeioiBzjtLjhE`OsW9fM+Vs&QSePV5K(_tYQ2a#2Wr95Ad&3hvd za|pz1f-JZP;3Y#R^CZ% z$;91nL%>4!y+8n`&A>eZtrF1 zKWlmx4Ge;+?`zY#@TJ@-3_XxoVEKf&3c2tLXWC2<#bzo{Y$;Ii|g+%{id~r z8D3f{xQmGK&vUq6BiyPrISg~6p<}^m?!p%fB&rJfIVbi~ecb!3z|JgQ58k*aGvtGb z2>~4i@(+LbUEzgF6fk3L;fvV-#qRD%Mj=k*LO<<7pYlN^K0U96ZB=0H%j`rkev9RY zA7qc0lE{ldvbH)3Hfee7QX9T*@}xGqA!~-?NSrfLqhl{A(f`NntPz9OJ5<-)@g=e>D;!E@26;+MWe8 z4U&!B?6-EWRKV2zHgp~ z<|*DjXrE5RT{;NsKbIOJ{y&auJ`|1tgv#rN-Hiknp)9X~)@XIbx=I4vVeF zRq6%F@l^6SCV3#lC>+0a5~jkj_1St(xHaWnOEZAjfKaR-tk?vg z)`Fledjvv$`%**KnTJq?p7*N*R;ZTcer_BR7*SO1!yN^8{`XXTgdx{pko-pWdklzACESLo3vD{~L7PJK+P*M3@315(+q)ZC? zsla~l7K*PdE2PtBKaxT>mPmoN2qt+h2?%bA?(dsc&Io_MW?Sa@E#Ks33XY@#PAcRDDmTVZZjaj zx?IH+Oo#x#nbp&D5qUYMIyL|EBr8fsK5z_0E#_juN~c>RbgI#!N`2_V~%bR4d`kkC(W=hqB~w{WEujC z7&)$A|AO5(1guBL=KJ?zOT18+gS2D|Xp~5T?l;e_o4Tp>`UmW10g6WcE?MJ1q&o*?D#zYCsD#WraF}$ltO#7(E?0OOX!+s93#gwwq*GR71nLu2-tRlJ8}a*c2xwNU zgAVi~hSs-%vz}VgwV7(O}(a+jJ4)1*RXGYD_{q&$fqffSl+==2BE zk5i@X_gC{>kAh_Z@TF0~s3qIatvrrckKy&F@GT#dQ1^rRoKG!qvcL_yoHd+9q*po^ z3$Q0ghU7T0|KvB6JqeFf==jtFll7q$)U>n@;s;TU_eFjb00IE?U3Ey`^k_SDUlDT~ zWL+w`mlI zeab3`tY*rP))6CXJ&>-w8t&&1wG;RMt{<`z(6^piL{NiFVi2Vk-+zUK3$%X{Z9%U^ z2f7di!h=JB5Q=F++W&kWy2+?a`$q}#RJ+gK3Zqf{A_Yae;}00t;`lil1)i`i90S-@ zr-!eGNfCEq*#5kIc3oyT9B+2cf-xIc38AU7tR0ZH+^b@If@mN%9DO`-jN9v*9+EW# z-p4D-InJtda_d1>z0c?l9VCaS3_G_WQ2QnK*BKchq0&pL_haF>#>i!3@ArYbSwjxk_VQ90%L} z0O5B4>8UIX(>2VRck4ma@R^R@AU9`Udv|w7n>uD7i(A^-2K^$@*K=m+Zx<=xi~?-j z-Q8#06tAE2RS5#*HR_twkHLP(1_e&@zN@=emEB;Tef&Navmn<*J$D^Asj!6C8H9c? zgOqVC#v)|$Tu=GH5C+V}nfb13C^DL;uLlfcUgbO<=JmYwyua3ui$-?3ZngQrXIk5W zPH#lED;(AF%|Hme1`udjWvC5w@4oP}@;0I8Tgk>>OaH8O{W~v2?yt;~Nzn3f4UjTF;7fG&2-#&N(4?F3O z)@d)4hkOC>QBVg7_^2XoE?_>5?8B~meb_}EeZ;%pwW4(sJ51Mr$dcwA+Y-$3QxYu6 zeOZ305p8?V!5n09{7Fe~RyF zcdL3Ik5i7@Z3Y#TZwU19Uwyv6%d~qp7J<@gBV-5VPJ1<{g~WoHfVcME1d=j5G2KPRl7!OMX0jcFgZ zx4-rDw3kS`=;r63kDbbq3MHF2_`_pRth+OK=Vtr! zqet#ZJQ|`pu|R>{p6Ug*tTOR*TBcpI{qDS41V-IG_f}r_zp1rcB8*yQn4c|I-`MXx z3h&q`q(|r2SI67u8m5(%^RLZ-prGCfwy_t**^i4)Mw&N&^chu+`6_wms-t+n5^vFU z#k`Sjmi*+ikK0C!t#25YCHPaWt-AjN`~+yTYb-!4rpn@-SZy8qUd3)wxmCK?_Dd9e zb|OU%ij7N(r43Opc>DO!U-P~5F25#UvfEAZTgiQw*KDQewZZD~5+Uam^mIwQQ-)|b zX5+L{hTdY&+ET{aWwhpKfBD4fs-YU9E{ZZn&NlD%`}Vk%wqv~#6~6!(429xNjlgk0 z-k0aMMTr|kD)2p|0yZ?M!uDMJ@SfR=95x!krb^L?;FCX2&eHRay$j7y8F@P%A;5{9-{)ev8j+YMMBbQo$Qz#kTdN(jykpevPkv>LoWdo6X+_mf z_!^6C+})p_G^;Ll(hUdJ9e8??Rld)aw@uB+`D2$+#H5>(czQS`GRdJ^T`f|b!fU@e zS5}fi@T+Kn+L`a;T0_*yO@l0#PeZft6iE1ZgRW|qd`Q$vdLY9z{TJuC*2T-XCc`#= z>9lK>U#0mOq)nwhRg~2JEDCp?CWpZuFR6N8E8|9rkn)`JtSjbQqEjJ@LQoO!Bg||< z+W7#z{+OAWMA<7_Kio01V-fQqSWcDXJ2&GZ=jCy?EmYHPk#Kn(CLf0$xIg}13Q)pR zxz-@fmuIMp9bU0auwKG3s%P4bC&_3|C}uRaN@^Yx`{K=elYD?c=PO!Vm=p?P9PhMk zxb$oc9>lFo{dLlXiIl)%*L7x4?MT=G@rT94UQO#AMq! zi)=etJJyYl=+B~c{JbmG}Old4JJ^VNHinI71LfqbC?x*Ad&%+;(;S%XZ8?Ow-Byc17%TKCZJ&w1H9sd~J76O@7#R20x>dSw z|ABaRY{7>+C3bGO2f}l6Z>89w3JR3aRhaBm_7r0*id^u$)^YC~FNT#bC3+oA2*7q^ z8?T^0{RykXJinUk&9k~|tMhIrzJgm9Q>b_)xn(4wY^7W}Jkq?f?JlM*`2>>nD(oc4 z!lra~j_4t`y)}Q&hN<=+PY<#fG3t`MLXGXE>bw8clNUaK=PMOM?Gj^7cJOe0u zyc`9t0)`A6&E_y>VGrUPUCTY5Q$6YK+$#PTK1TYU`GNTd%x}<~{CIZQr0dq*yv_3u z&2Q_NwGUialc=a;$Xnt`d-#OL;=L+m*b857Io=JcXhFW5+_6YTjI7nfwYP$y{8cYV zQf{h)TDTW)#fB*~3Fs0~4B!2d1KOBCV^x4o9eW03sbp_ZXc41|lNtj>L=FR<4qZv> zJfO9+!B4O@;1y(U6}+0pw$SWrepOaVQtuhnPv;kOKDif}URGtOAZN`_p?w~!XH{a% zZisn2Z@772LaQETr>^`dUhiys1(^U36T5o;#kA6gN%}oad7hcon3(h3Jx`CHtpQE~ z@-<`ED9)o8^YWoWl1T{6rQMsuJx@QGv7IcL*dV&y$T&pxFNIM%Dj5uaK(6xp*JfmB zks$|%P~$^-bwYRw1V%h7d{o=OPB43^>9u!@i%J<6?5O7Fq~5o6?o0)~A~7EQ^>sQ+ zPV#dny89G}j#i8vozH|;(hpCpoh4nO{UKr=0O#RKbPnmEF!%WHI3Bs2gHLipE{|nq z$fiB-{1twixBX3`jB!|oo%Kc^7VDJ$E1m1XvYYYV^p&;jx14R$GULxIN-2Dl8*I!s zzc^@ZGn&SdhBJz5o_d;85HL0(J2|L`RJQlP+NtIkGRc*d3EB#*<*J;q9kCc+RMV&; zMd}N_OkpJ9@<5i#xbtbRrLtgE@{Cmi|AY7q~q(R>|q z<_H=kT;jc$MB2HQpc-VGLYhx1#)%a%sX3*sjbwGTyWH`!0LmP_8JT5Q4rFXOw9L^Bt;Y#dq^JpBthRM1ltuKOZmQbzy>6H)I{g367NF2dv!Re)sC6Z#Kz!Tm~>m}WpojaZXZlOXHe|JpPXXQS#bDhDxMq%xYXcGzY@CI|esaEgig=4tMh~%~osedKY z&TpoEQ^e%2Q+;@l3vxbxWHLgvY=9j!7~2qLuXHLr9#@Ug0Pn`vHQE>U003sk`$IJ% z64UfT`a)-OTv_rT@hq7O1ZHGQ^j(2pYest!KkH@whs>9IXh$?`hLVX&P|%^Zcphoa zm=@mDyh}NQ%zLqK-Jq*4^=f@h-Rf6M@o$|}8jVLs&lYUYzeYPEJ@a0GhbZYX;q&bg zL6Zg;TMX6T7ZGyoL86zKg=Togm+DT?V{5bE}uFsI3qhaNGmZ~BwX&uSEbTTbDe(BYeTF|jW~`a2mmK= zfK6X@Z%d62vg;|x)Y0blz&(S&DV7Zv8!nCY9|F<>FgbCU7%w@HU$cH~!m4rqemB|i z{LX5Ojir>+&>grwq6Iz4wkZHK1;C9{4!Q9K5QMI&KrPQSTqOs%?6`C4n&TVDuv}iX zYz5e)a}qJO2{ERbW#_p5LLXQ0{f=inHyoE5<2(1pFpJ@3>!8-ocrF|Pq|azSl(2WXMuqTogOf;oF~m?o+E#6eXn^y7>u2HVb@ zt;Uc+qU(@#Br<`uLz`O#Hx;ogCxBuvvN;WJPI=a87zKe^YT+H&BFD7DLIi8wAy`>~ zYhZ>ojlGBzdPh1!PgO#wR%1Z<{ltaMEB}iDR%8gXFU$U|Td#-j;Uh7{E>$TH^+G3B zS|a5lnwCpvm2NYUaOr_tY&QFIzm{R$ISShnH)2DI3rqJt4lT0p@Z$qtFN!pmU1zg0 z@8$x6k8qw*D22+%`nd5XC}(FRThIM#M!eG>K+m{j70^Yt%5ZQ`qVySmIA)bvxyAD=XtVNFmxGWzpoUf6xHh%0Z4j#jj&0#tiW{HA*68#}vN3`g=i z6Lpd1!PI9(U+Rs6PN8s`7M&7?oP@hafguK1fiMDKpFVS4AZnAQ#v`BUi9hD}ao9 ziJ0ly@}%VBXPn3p(=wV{1Ei5&kD`skS%QP zS?U4S`x<67f56m2Ha?J0Bx)P5b~0Dc?&e(mofu-7cU4vk(+avWwrX>qc~Ok3`}gSL zDax0Ao}|(klN%Htc|{ryQ*0+aO2+#miTr;R@>9T!noO(c{x;|6YCfs_U@gYftGoYP zMZf#WMu^fq66e;dpjOZR=ItFF~&A7(**63Q_H%3So zwiQ>0c7OZ?Zt#Ennj|1{zT|Vms!`|9bL^9%Qailghpvq{&-`u1Fbq_$=o!ZMOP7Fe;0afzK~W`({NYjdMPDu1??&s`zMFg05E zUVc)2c2(1^%EJrj!&w}>=*rQw3hkSh&O}M+`RNT>?kyB`k)w_BI60C31_2Dg8rVAV zcDbW03m5n=ZdV>>lxx#BVQ%}1dbIu>YpNXgZceFR>%HcvIuerdNNb&c2DeY5Pn8qY z^fQnyA9-!+DF^v<)Ej2LkVpG&qA#ZGtcuak7jA%Nc)F;F?QD2Cz{!2VEX#&QL>||d z>K9s8taS&9oONVKCD^_$6-?N6sYWi$g}XEUKC9oqEuZxNarNHuRKM>ZcoCIZW)!ls z-{EBMm2q(Fz2_maS4d@MBqYZs$8oIettez1Bj=Evkr`z~_V4v_`urZh@9*Do?)$#3 z`e$<>YQx*A{j2^wC`Iq_&IwZ8rTG*0iS0GQ}f-gb+4z#Hhm=(Twp$8 z`?SYE=L2CQDOz=O%!?##J4HQoAka!~Y8j=-6Yj;n;TG)q`4YU@nr0X9)&L!*0cC|7 zndM$T{w1_v#VU&$3-qJbFCUGr6iVOG{xFh8uL=|ndT=y^K2p|b2g_7?Xhm=HGrG5( zsgY?5eTWHQa#1&Kz}lQ&2mSFc07?by)lS)eNPm6d=U%=AlGLvPF51mRL>;sZeQxtx zPHQ~|A;-h@c{FDtHs4krDFAZS*1PyMscy6>Fzu1bp!oemnoAf&`UvBiKx&=Tb=AwL z42uT?21VeR$&1gP*w)UKXV`mpei68+v6-o0T~i{kk+MIsDZJ;0hxYCiu!%llsi__`Q;^rA(*QAGg`5CSUyeqdx^r7L366gnFu zn|i0)(&cgC?EunAMxa?0d_1k_D&@uPCusb*9*@V=;MSyB zHOdLjCDL)tSj0!+y>0}7J@AbQ9bj_y*V(_({~p7gUKFXIFVmC?)B|O)cnaCOp#Fj9~V3X0fj%o-&u(Uf<4j zU-Sgr@}gVM8>pd;1yIu{zuUAMcruD_9Nhx@NV3?m(v7uSADv3X=t-^nY}SkCu1PH@ zbn~Oq=6hzQr$T_o&!s+6VKeWB?Vlw%y?`EM_-1X;aAX9qnd_5$p_~VvO_#_AR5AlC zbT;n57UUc{IZ5I7@IrHXp_ss%N+O_QPA6;5a2BI$GCc_)WMU${t=e8Nu(ssCI}%?i zY|_HCE(1sN{LN@gLD)8=ZNaQcm|OC=V3wFJ^3CMgw1p5WttKwF<3Gkaq~)Fz#x=5! zcqY;b7Ml?v9(S9zrg4QiU>pcI%<&5^l)pToxy8N!&p7C15QE+GET%@HCiJ2P@{kvA zoO6CoT5OZ@KzdZot_ljScrABeXVb{c&2J+-A+B`{jzNYBq7P(5GNG=Qv01N&e>w$| z!~=a7Q9$`irsFbDk^sYOFqjymhF3U5d!3gJh!nKcw^UV4-CQZCX=6Hmo4+?dG3y_w z`)0cLhNF~5Y<2JQl-rs(Lbx%+!tBYdnZf<_`S~sQq|7}l_E4Is()%@Bv7=d+A;l#R zjuD9^Kb?6>=Xkwl0|GdvS@3jp4=)xNa9bT(W-HF^ykj( znjh{^K(AtSsyZed#R~N*p|velg40M9_qix0tQKluRJbwbF*<@4`)a(>R-@z9~*&Be6;;u-@Q(OZ@_Ab z0Cl@USL6%DVtZKV2dsTTFgA5|`A}PdFbPcUu3FtKyQ2i2y$6I_`f>3zEX;W6@C2nJ zD}UMNAkCZ9*po)?vUf>Zp_6g9Oy8#cZ*bhlMJ+#yz`hxg;UqrZ^UbsRA66rU|EW{p zqZwD{y2qgsu}6eXylPN}SD~re3P5|hH}Dyc?<9^maggm+K4woUt(l#NwjTh(0z9{G z0S~`I26^6N{|#YS=ycECJ@G3~j131cYd-$)#Soh2)+q3yl)U9m8Q=xYk>p?&!ey7_ zL73lb&-c$xM+up*WUCRf&N9Q#ZPs#*vrO*&P&Y_G6h?FN0?kMxm5lYb$^cF+x9w_U zzI1gQ-)f};Tdvs^wr+xMKgb6t;WL=;Zq0sojxZ&K6SoSWKKm)YVbHhH0!joOq4Q>L z&r07@%R5tO zEy}gH_qacq?SsWl4*8CHazh(wPBL0H`&oSnl}o2RqhE z3fKONyXYp$1DYydBuluLGF%xF@hol)wtd_P_)YwKW*E4^^9>$X!D*gt;+V@{bt4ji zNQ+188bHVL696#k_u`_X6QpFm1HJ>l`kMgkxu^nKw6}(^+)3P$0l?J&kq8~wXeOb8 z0bbLZh=)J2Sv>1NO>CAiRWf@9R1akp7^uHI+wlG5zR#+uH1`d={^fpI-OHD*{TilI zm%_Hu?md3!w<=gF-hKPY^{;F0vRWB@HeKX7uzZ0DZTP+ zYNmcm($qYQ_i7`56NS{~V7=b9s+5ExFw=;rvpK^6l!K^I1E2Z*K4oXHoB^Olihz7d zR07=)D&tuujk2rBBP@^>JFW^MTj1f3Ev7tXH%|XgMy)`GS@~TuGm?;&>%HV5R&C8~ zRU_+}CDgRE%zavd={KCNylw8;ba~{3t`N=GYn%*LjD?ZHRr|K@tAVu6dh7a})|>@6 zGGqhn1dt&qFqB%Mg*s**9OToH&`eGSma7fn`umyLt+4L5VI^H9)p`^%gO$HqsejR62cEF-4?HKZZvuK2 zvl_j^zI>zoWPoq|+XANnW|?E913YJfocJwff1aDHCLzkRv%mdyyDc|NqnCv){p>HD zc)-3}Mb60m=k!htf}rWXA_-H46Y0dP`V;LAT-R`gfOeW6$$%GA>tN4=d#H@Gw4xwg zn>)rLt(z|rS{k@g{`ryUix?4jL$Jre+tvUSgwtJ4~M_x%~Xxvkh=xUFy1Y?Zyrmq2tc9%S?|5p%tOP! zrAf-;u`mx>kbl%*%JnG(Ek^ZJ_V2-E2RnR9d~jh-?*o(Q$xnWyDU^pL zB?zO|2&~Pt1r4P?z5C&C{9j13BdYPC?0=1Nnu?*ZCS(b z!hgt06`|##Np3%AXp;2Ap40?pv11Ah9;vT6q!h97^9P%dD?&Ry`3_*%u_5dm0W(eiODsLi*xm3w9eeSwdr=IW-a814IPaJZTIm=WAG;He z@Gh*K)NnLkDe(=VUId}{>2X$GNi{Yw89rlDrFVM@sP`Oni0o!h-fTM5Q7yMY9*q>ADk?-(4(=c>5wbYCm% zqI&B{!10PuVZl8zBElj0<86wsOT90%UEcAp6`wkJstmE5w4hr`TbgGc$oftZR}7ME zDft=x*{nb(Ric|UDeX%Tf-KKy>zE%4Ww|m^spA&p1|z?le)!3|a;xMvbKUt)O=b-J zBf^7Kc2;#<0eoHmdPr_`+9Yqh8kPf21%cVbAe1p135bQJEl#0~+3Sxac@>~; z+qrgqDPM~u`5;OT)qXlO^u*ijmA9^g z&D=P&0utg)?rT|7D|9%hn28gGi7n)wyNdi`vD&H&yj`n_b$^xbsVxjCLF^*X6XtfG zpSEX`ZpiiFPw*l`NHgUg(~^nYr~mAw`eh` zkyT9@6ZQ5`l~eCI`cP2~6I7$^a}(HD8A`)ezy;ey6e3N=Gbz3xY#*sPq7FUtOuR-5 zvrm)w?AdM--gQpl`Td*TmgG|QYq^)Zx*o*M939qoDAJBOy>$1NODVjBu5XrIG1O`F zWuQquW9xz2ZQq6Xj<&Sfh5*}U93eRjo}*p=W*K2Oj8fe4e8_!i>W?vkk1;H0?x z1z@$9kaf&eOWi52?5r8!&p5(`+D)ZM8}}Yk-d*`yc9QfwYxSKVkJ}q1Xy`*uJDgEF z*^(;pp0cJNOp5Oh%vdUvR9(yOl3$M1clgeGxMu5 zIT@{%Oo8Fvq1Ca<8LR;3E2ZGXPuwatqibo?^0|&0l_ZDTYF@8R4?g0Nn`GnacAs2i zkBHs+VZ+szkGxbkspTRe`<2+7*e=Q@8@^Qyq42ri~?gn;LCAK>%!5 zpreNI{39y^OF;o!@vz~y>TE!A>2r*}EbPKWMqd#jCzacg-YazH<&WN93N7 z2D)gn$;+6-KU)r-c1hfOiLf!C-S*Vr_1itnq%WZn$exbMC8}wb_?UIKPVdz~qB8V- zwc_N$o$jzGw9`+ncCYK$Ja`yRN`ZOI?LKg`DX3UIWjv3`j>arVO@$vSL1X<8!l?Ot zgR9~7nK{2}o{AL(pRuWruAxyKfXKLiT1Rp1C6@sk}{O(dZ z!t@6f&Moz}^kWbdn*@-}Cu}%5_p#4B0f`~sOeMHyyT2BHSg;NAv8(nWFr>cslN>_6 zRCzgC8AE*QFp3k_h7KAZzp98WJC9sz)DOS&r|Yq`Vb>oO{o|P|I%O;kp!G=zg*ikQ z-_yS^kvzEK%NM`;!yEh9;CFi6J8G80A8$2;pD)gRR2a_f@w%CZl)v-ge3$-G!OMzPF+a5kA3{Q z&(RX-LZ@H-eP4N);?P`%P&DEmivgZLn5UkYgaVePBG~B354qQq?@u@6m}R1}{hqOY z_dGwtk@9kIjg`U)JTAZf{wW;Q?mZN}JZgO|FmiK>OcYqg8l`_v1)KPp7tkWxDu$eU#|)mOs2 zs_*(tla-qre9QUjb-CN#QhPntWtsh$W}dkSLHp37U7>t8bF^Dmzeb-_+)&h#ZcgxC z3!>YAP5x(bETSs&FOn0Wbgh*4L0r$IfvX~302P%0X_&o&J(<6~k;@NP=qq$eE7!c+ zc!zaOoAgNMdqGIxT;Mv%o2_2B|1%Evx!Yk6IIQ9c{I~|9E!+oq3Y_eJK1tClvBFJ{ z7(X#NvBkOy{9725Y+zxUgFRx7i79#)g#~4(XWHHWiUmBQM6IFB;vQ$)qn?*DUpNTQ z>MLUcH$w>RE3I&%`+g0GXt{L=$osRxq;q2u{PS++YrM~`bWTn=~(v+ zleg8qw_!EC)*t_9wg(&f7*uk$OIJ>D;Ca<`!7Cor#2?*Ub41p-R8I`TL;$trR5y4n z2I0r}s)61#4{o?yX|IFT@`zvmp!Slja_I{G{?$QW;7J0Hqi?oJ)(YPgo~eI^;z3y7 zh(T-}f5~>?2YBf^J9;!U`Z9bhoBaP}rh6D^NU6u%Bfqg|ry42GuIxlfmi+3egE}XU zwG==O3?O_V(YMw$E_QC>f^&sFcZY{~;D29zp%u3b^`*Ca5xtd`9vcb({mnvz0Tl^E zXd_61EIDL!Ze8r(86`j`;zk40?l zMjI7EOMqp7;6@t!f2bm~>Q_4CcxzJN8o+f^cFo`QM)mS_=w-{cF!5ls_9d(T^#Yil z7yh$R?7*pafr)n1xn&tvPi)1wfR&t1O*(FwVF(OM5dnW3nG@IjzYBhg7^Lu}i9@zf zyPpv-f!iej+og;Je|hbeSOz+ydPc7MayL!#%nV13W!#ltT}A`91Dx$4_{SJUOv~Jp zvLI17gu5;ux!oUr2AI-t3eQ(z-6|HLjNX4%fbX2spY8#n-_FL0_4ZXMxe(1;YYGc< zo@c+TD(Gm=8=id9Lz&^Q#z8Ti<#_Y%18yMjo|(m%6pn=}RA)o{m;Fpdg6tseko;w# z59fQ-XUA(SUE+s(xgX}A)zMl*9&Z}+nXU{~p6f+yA0CIt6^;`2)ahX3`>W#{g6L32 zxF@L78_;matHCsL7CqVb>`J&3k1!flrzY6#v;xtW7Wnx2yke`ZL@T> z;5Og2%5M*VO91R$S0rffY*9S%zUy=jBt{XV#+TX%=STj>oDqThb+aMom(HC zh?X?ye(cNLv-izCkH39_I^6grcUnQvd*)a4810krmjnKwvrnrq|Ke)%&q-tiO;M$R z#P~B4To?#x&~27ybnm1Z79wM{9}~UsmUha!I^*{8zt_8zzaHi(o#rmgDPXUJG^0v* zvL-{fe^Ege8}947)Doc$AOVRGL!aj;hTicmhE-pGAo9tI2pe_EE8G`709Y};U#bSR z3`u9iQ4DQR2KyQT-6Hgw5~dR7kb^g192W-c1|%W0t&o4>_&pq( zn!Vg-jpR^)c#O_TsuyRTKWOz?l})FP0ZerhP61g(ib2@z4U37a_MZpe87sI%1JT}#{97D_;MynDA1;3{cMj6t z-;;2DvKG)=eCxxSR6>bT_%!|UXl}Sd+&MfB$cX1P)1KV3e2){u zL*;7N?n=1iP5GZEJbu>oZr+A{=cw;6QD4K)3WhZ#_V-tD7aXycgfl0oZpZ(Bx5+>S z9pvkb{ruwqEtk4Yw(1c!l%aq<(J`rUKtz$PltoY!_|j;TLhefdwX3_Uo{1}U#jg}4 zj6??BNUM=zZ6ekUFxl+odVhhazT-Pr9liqBK*49XxE>sFNp-~Iwp_AK<7~cFSrBd zl0iza-K|d$Xy?Puq>vFKC$m?3I`4G#C8co&De#j+eMOVk~sH>k?dC~>r z{L$VySRfvzn%;F2!7#$B9*#3AG5RS4dvuLw%= z4WJfBn6|IOOlSl-M2y-kh6t?(!N@W}j@f$e;!Pjy!u0XONoHPJ2?s4<$h;Xd06Cv| ze?E0RO8eLF_0=&)%`;yDG^HGT!JllXwx?g?Kk{tbcOD79Qox{>!6yae>!DKRjZRf8 z_BVJ!BzndzFtnh#=&()-lz1PQPmh>c&B5^6jewT{64)21B~Gu-kwyaN4A~1~KlA6L ztx^$1;i|C5b$6d~uYOg9zB6eSn#CCNOJOSGi6!lW{!oy2H{B@(3`f-WmT;wZ7 zE{|$mebr$OSPH5BgDO^=ulG%XJKWWoDPMj9<5q(P{R4C)L2hGIpo>8t=V3N%@)Foo zCI0aAl!zQftS=W)1WmXx>M$CE5dV`|pp7%gIRE;4G_{(0>ndC}4dGNH-AE0=jeoib zzzi_3Hpy>|^nd3~GY$V)7z#XZ^reHomFiqEF+XS!1TyB2b<4lbPkxMtmu&g4u3T+G zBU|_TD~E;7AB$rjmE@dV?Fc*Ss@3_zur7Cc%yRzMQc&}7)N`=RSM2ekKVq({lVYzs zbV`h)2+*|jBNDsYLTpu_EdwwN{AK?KCz;dm?e)zI;z>d*&_CsPyvtGn2R?QQq@(0N z^gP08DqXiM4x@3e)uE%aVdtAk=l9Oy-zAp{Qx!O+5!?Ufvxz_(VCY>pqgNIv&wmXv zvb@3D``=(_DaHWNzZ|VbIRkm_(C>X6UeJHMVjU)kyW8s~*FqEQs^m8ZlMo`mWqhf? z$>NP;E3`)55RDAdVaF`lmOQhZ`2;Qk=nw;_LtH>=d!+T$!TT&{_SG#NJ~fq7**a@- znMsjhTs9PRy&bDs!cL?HMeu!BREBH`a_>zT3EuVYfT8%TJNC(zPhjeaQS(U50SQe1 zahX$Vi8~-6N5rFd z)b=t{m{u?B)1E%=f(#lHxG_%f#*l@`Z#r&pRpHPDpFkSubDtpoaS8~xe-VYL_=l4b!x$=H z?VNiwMmxf(w%YCsHZ@1R9XCt@3?=RP?BVIT7E<^_IK*vR1yB*(zh~Sic z%60WGL3*5Juj$&gp^;usT~}mfgs-OQXDWnw<@7EM?pBmH)R#%DQ`)Kcv2sVfJuO4JncjMOZ1-1&l9~sG=xXI5Xe1@% zAXLs19xnV(yTHH}NLFTmFdqo*_K`(nQkC>@65BfLQ%QOfAu+906DS#m-{S8tqEEaO zx2F7WRqKks{TSpf?p0;}{t}IWOF2Ir8reYDC;!dIq5su;MVqLj`Z()LrnK;!{`_Jp z2x%n4L9?8*H2glPK_QScvrcW?l(~*KWwv;Sh$2aRQoWoWb0TIuVqqzUtz~91LCTjC z=)Wjfc4DhC*qOclbWYD>!;USwNonzc9KqFel1TThC*JhE&P#37O(jwp9?Z=J+ z&onH$A-mbnj(M7!+@x3_u^m1=Ek& z_seBD*lUOyB5OvXOdHU#%s<_E5clVud_;sYKDIthKY;Sx(G{txb&TlrV6Aengj^vG zY!(^5`B4dDV`7f;)%TSeCg*rnrIU1c^<}ed(Fok>6)H$% zBd(${JWETpj^FafqhAmeXwh{eyBJ+VYN`dTlFXuE7bVV1HwlVyrj)OyF;f9wHuftr zG-><|MP&_Z#LGO(xFmsB?%Wqrx(p${U^dIcs3R|t=pR2MyF}X z#iPqh-j@TsWGF_*%03F=M>;OVFM5m;sL%%|P5CCsL9vV_HhJ`0=%$#` zbV)a8r$Ub&Z+gN)KU@$&?TGh-W+>Oc2!zA01{Pect6y{*AQqKq*VK#%54+&R6-{-D z0ny)(Km_M+n0hj&ArwMyN&3=Mjv3BtsE_>NaTlxHzU7&9Q}iQ&AqpWI?HmgGk%0HF zqIg7w6ry6ov$u$n-Y^_`H{On?jNtA2raLDlMweAWAUw}bY zhWdHd((~5Y&ZN$R0^P9Jg?^6{IiD|SpQ4dt5fmQ~`((Zmi6_@T1+JfnJVd-1F+bD! zfnH8gVuJk2xGD1wn*FD&58x!9irqNI**26*-{F88F6m87H)F+4Bwy|(1u_S0E%$C< zh2YHq;@KDnsx8BxEgFt~Xe6dgSXlk-gE6-9UexbTaT(!^HW9o)Y}0)8pe_XFnD~Eq z1E6JX?3UjiGPlK`+e+@{xWys-N)g0PZ`^hV=5bV{f4^;9W#F5v^_D=5dVgdC~*| zHHfg%cUUB=d$|RX8dSiKWPnh%@E9(&rA%-9=>CkSg$#2}hns0jv>d#o=Hfh~D$uGe zYY~_2N3JMJ}6RW~gbO|2%*P-TynJMf1Sf$jC`PVpTL8ZTGD+Z8L4R zs+ZyWY6SVSYi>zP%6C1>koT|Bt4^e)34BMNoDMt z)kyrlidoWc?~SJzl0YI^&juCsQ{PE6EEKoCOd}+fUl9)ChF>93g8H!WDQP6*NP!pt z#j~TwLDageMg3|6g;2B@cxX%adU-&lsi*O0=JKCz(>!lO{k=8R1idL89$bi{zOdU^ z-BFWYxXi4jg_JLh**iBaHL;)BLv)>%xoIe(G7C81W*yw!TtnQlF542xZ_Vvj?9IDY z2wmSsxdsPaw;*F?D-BPYcrG8BJ`lUxdtYwjsA4w{$FEz3K2-iSw8{xr{j1_#hHR{; z3!(MI`(muFek;iU*`yfyp0?i=M@NpnLgYImb~3toZJD3GOmjpYkA@qpths^?96^qq z(cgt6-P{*3eS*@mf@ftdjke0ui&a7|T-e$wme7z*X&Cy1jw)64+cLpi~_j zzr&OdfK(|c9uQA2DMU6NM?AaE=NQNo&^q@6DzCIEAM_8uNb9&$9R;>a9~h1cg8XT4&Hc#R-xe{i=Ss0zi_@ATtM`VSD! z1bzk6|G81FE!_6Ea3`v0{}u3`PMe;TzsMTo0dv=1i-fD1Ed2v{xG(O3EN?kUr5x9> z6&5H$w`mHTO=MCmuBwB8wiF>1oFNzQF-&&xhsx^sZ1XafrX8~Wy0-VY12PXV{;MY09`X$ z<;Tv%k3P>uY6Dz|dPSmxbB+Lx!~{tSDA@fD7XHD#-(Ldu0%rIxY~@$LxyG4Q$M;iR zBtasmDv5Q*-6Q%8F+t@YS=3hNz>R@i>k^pNP(fb8^jpQp5IGLsn?aaz%SbVBHQIJw zGNUH&I{xg)o}+iJ2$F)o0XjW2E?D4n4L%#Xx}m5e=x;fO#0F&pOP1}@E^MlRv^*)I zKR8@{!sJEndy&=IfJ-W4ZUC_HpAPd^kmEQ+m5UBY-$*+wo}3P4fU_76)pl2Qkp#1t z0xl97`~@mNZ7Vhk8(#Mc-AtS#aq9&o<^934mGvt5F7$GHBI0H<*o9z7E%cvTKmk`V z(e(nSTkyDwO4ravr(X?$*>2UsZ16Ec!&WgYK^huC`KL3?E~?vbRQ+GsLK2irDhSYl z6VJSc%8<57@}rZ4?Z)LVe||eHeXr?#Vox`hT__aZnVlJ;UxY>;LX|MSsn7pqMKEf* zyTH#s!Sq+%a1oLpKJx-{P%TM1_`+vXY z1&Sx26ko?a*}-tTap|ySy*0yLq6Sxp%TToIq6UJ^dmQyW-Ge@w;?{eAc9iw`bxY7Z z?W-!GqMfZ5T!qryyHH+J0!TtOd)wu>fKZKsW#AQ}vYn#yYf|$i(Hucq5|%yEY0n%Y zUIU2r8mA1C;{g#~GKugpt!5swajH0EN)t(6m?}kp5rXZa^Ov1$r-uzORf=r(Up)Od zlf@~Yv`D)rf{P@qG`GYaw%m03hdRh#L`4`qAat32xY4}qWu2Pv>VphB-0D77w1K+i z$k@e+D)#=yXN>}REEYQY`IN*E!^8v;+he)$XBiY8KprnjzRJKNJf9QWtt^ElSbn^? z9L}zaZM8#5Bwr)zofe?k84Ex_X#`KK1(OgC5hRAts*3d8u6yCS|9Amb^8htqgmlFM zT9ES6h|=}WCXgM%^O4D<@bs&ZoQHO|s3R30B#OQ88nXdbz=Vdr8vuYvh)MIF08@a5 zwc9}2F)XaG&LIOt&IS`#JDL2{ju)wDzi7^3u}hO-v5)<7og97ZboH;)^_MH#5qn15 zm$3RaFn%ie-?NAOy7@qhj&!43Je3v-0QKLYV>`7j%q&J2m9gVj&Y7Yn#O6xylARUg zP8Q$5+FZf?jC=eT2eAg~sf<<(FMUczMp|zxbKEg`y?i7ZIaEFooccdN40d4@Yslbn zEJu=Ty$(nJY3STD%LKS&hC0l*iG4dOCZ6(;U*OSVnTSBB9re#92@2~>t6N);Zv?bx zL>OgrGQ}4zKxl!8d>_Bo9e7;Gnc&{y={;`pKmp+HHS&=`IS-dwJt>)1G-C-ri5gQa zQkX4Y%0T{o{7O8gTFyx4^>Mm+o-j~72AUx@XMcylQszx0;%&h#rclyKe z$EFr7D=;*K1X8rCl}mW=gf$uR;!6_ugH6v+Z)x@bBa&Ew<16P6X2J6bJ*b@sA4yF9 z7ib;gPsk5v)S;;Jzt&=H4=ag57?VQMsiq!CWDl7NFVDp)6LSw({#g4Gq-jL57?ia# z24FZy_|Y4pm-`I>zz9f}@;vti#f~uQL5Sb^VDHklB)0U67@#PMQ2W;RZl9LE25PjE z=?V!fy3e#`%jX<1ifcjxol&D}4}T!P*{QxfFTn>8m>;^4_D4&Y4m(Q*X? z%=p8_;@v}f9x||Gi18s;hz>s!UmtO7+RZYF;h+q4AHIz1=9;bi4%6dIH$^Cg5Om&F zHt41_pqXXbe~bk2O`D^hn`ubR%Q#VGkU>s2 z#yCfJlY?h#G9RK5w0X+_!oZ+PLoU}Kobxc=Ueo5&&%1;<33*OtyDz>^h6*}6uCk?P zOlW^olDM6t#N}d|$7Z;@1*8`2fBx6O(8q>foCeaRZsr;P`gUQmo0~F5L$~JD+xY1n zh!6Ur|*@i>`Y}Ut$ao3zVt@v;hE72?!;P3SzzknKg$N{^$sfr@r!!TZOc=-nk zD0MYC{fdRFvmJB5ftLpNOA#wAE&JC7`rD24-|jD85oP+Y*>zz16It#>XT6Z|S3oxz zeP|Cm=499|le@44vClxKNd*y7gI4+cOo1e^Tf&s<8^Pm=I2nqhKm4g0DHjy-d0Clt z-Zcx>g$d-Xm3KrhQ-U~jtx6`qHHXc_1&xbE_~bDam5AS$*W<_7Ni^7CH9!I7hA^Cj zjuSq}F_?U{NwgVPX{8?)sNY71dWd`C_wF?DWM>*B20Oh-@#>^B@J$G1z}%OY-}(0h z+Z0N|b5W+O@NNkseE`VZ9-^rq+NDpKicNbsQgkM^$nxKA>2Enx z1C89+?DX&)&qVtF1VUY3zYA)rvLl5h11CE$QJ%I2ojed4v|mJDPMkm)EDRSs@xJ9R zuih93{EZ@}yhu4;_v5*#k`m^L(X*`SH}y0FKb&unIg`Y0W*vERJ*FIFB@Th^FocS8 z84PgJ>e^(p)Ja&*a9!KKS@l;2= z-M@i95gZhsJKm$@AD>i;F)i_Ya!Q|hz`fHzq9{*oXmau z9%Gz!;Zo}!y1@4PBq(ze#raN0=9{sRGyj4lVSJF}2()yM)pA{T%U6>$OhW-Kx+8>0 zv+xd~do$B2-&1-1I}W$#8`~Wkw#ALVJQPYdH+`R<@bbp}n}ee9(t}Lj5!WpYOJ7^d zulpVkIo^?wFrrS3Uuu6P%&pfwug?=$|3XPe%KpT>1XLrPTJM=DW7p}l$bYB~AKr8d zV@1zSkV0u$fL-e|%ZyXJ2}C=iPiyY>+%Gz`)uzgI#FQ)AO^|gu zGbPja!NcUK*S8!W%1M}w^m=~zhPZWnI}2VuMJPHm*su9YiBMK36wM{M+?MLqMQeOB zw{>O~OSbEsD@;0#0CfZ!O16KJ_YLJN$=!4NjBFF$0Fv@T*1ZJTOJ0+ z8aSF;s<9PrD=hw;13{mJL*M^LJpQ)N8K!Lon?#+|u23}1bHbaiE^hs&lO?Ic(LKXA z-6TwhQpTP1My6ovjEY5ZL z$C9f^lDO9ai3R&a$Mgp($PD4VVyh=BZ+Et=<_FWNAy;(cirfx96aOd4qzOU zcAx$+I#v^Ca~=J?&`-Nph0u;aaq%NHD1aj<{S0^AJ}2Qd1BSxi3@ys_-Tt3s%ig6Y z%7Z|6*gUFf4I_Wh4QG`95^wkH#9>fT5%p#(e!Cn4)y(g5QSZQ{bx-a~Ka>00F7_VVzU?>Y!cziX$L?kq zC>i6Y;5EU>$>PYjzQ)X*4DaTT6r1JyBxvJujk^ibPzXC1-|$*Ym_#!jUMB8Q;7t-Q zkOvdE<%hRz)bsiMyaA+7)Fciaz{hky7Cc8V9%Y#HIvX>FxlQ|(@fa&L&}i%6QePtZ z%u-7UFn0p)u7y3afBPuaSWP$~-sh=c0KGlVFU;yP$kufwyRSm*Vfch7_J7w0fI6q& zR<>6I<>JO-!Q1=YYISQ!A^!ZuPqavxQd?Y-n=r_z0J`JE|$!~#JBT>~IJ%Y3KB z_D5=;b^hz)2i#M)$5?b6Esy!Uo+)1Hgk8J z@Z4-Iq9yPf8r1*kk@2GGhbkxc{{P#ut$qo8WZ9v;xk&@7p@GL?9sFHg1BjUBQ~3C( zD`Mpwsx^HwC$gsI6Zq&QuFPaf43+A38#(#7ib%y$uki7au#ywnq4>Pe@BfNoVyFB1 z;%;*;eA*OA2~r~cJX2~PK>`Rjn(U=A=F5laM5$yoQ^ywv8^DxBt@}5!fb2~l%tZ_f zmy_c(D*EJKbqJI}CDdic$Hj?{GRqt{19m;vrIB{k&FspDj2y3MX z=*9wGJw)i3uJ?}IW(nAx1tf9=0A2>fUS*7sNKak`;S5zz|t_H3q-WLr!MEX!TuXmY_D4qyBe^+#qBCWMBAjt|k~O8rtU3 z3nZ3I<;i}0uo~daVPdbGmClnkYo@3aLDo+GJp;E-fkTS^noKwrDSOles9f#Lai^#c z5AaG&fN#qJdeBB_O-N7scO|Fp6#yTQ7Rl{(xF2*+12&p7(!O-*=5&wStb`kqWA;`g zks#2Edv*!OW52P^02Z4kJtI$QkOxl$qWPp;wQLFe!SwJZ)J*|M)Q7lUS+4%+ z1!1*(C#$$b(>e=BWrzF5Pd@~N%pPf7?x!zTrV%PE86^Nv^6)hP=9lwB8vBCnD?kQU zh36V1AL&(4ig+7&W(*V2#{hT_@QN@Chz$*PF$OpG3*KMR$$rCERMja+r;}Uuor^ej z4qUEVCf6KuWT}3I8Ltxsy7)A8Qg?Sqg~bgi7Iwuf15FW&AXmho+tWx0bGfTm(}!4V z?~Apk84@^028ftEp=-3j?FBuyF7vsiwf+2J0ca}TteR58Nk+#Xa;CI#|9KDK4J3uj z8f^B=97cEFRL}@7Ti6ZqJ5l7gJ|p7Qg#`Fot405znIB&H7pXDg&Ef;BXwKg(IquQj zjr;_tBR7yT--3vZ1hr>;iySZ9<|I6FsvVu;3yHYJI{?zO5j^*~0hKOKk5@YtkWC!$ zu_%OMJaNmg^@}?Hv}^!@vBZjY_w3qn-RJ)7WZU_orcldqPC}{cQAEEX#vnHl!JX_3 z_Iv{Xj0qma0eq=w&W1k8utPSsy^IOA+gE_RpZtwR75SgXG8G^4NMXIFh?QVDj1W{qjWA%Pri(e4)}8dIhL4htzd_ z>bvHDd$kM~2yGY26sPYtV!h9O%ET}=EZ$TXLnLSj^w(?NANXbCWL$e25S-pgCdW9i ziz5~RDMctw9~V&m(bx|QF|^mj`A&ypt?UQcTQ>U&mb3O|J!nW`Gm76bT)DcHOsMJb zpxySSWd;0fxaZ>?m3EUGV9#!%aL;5ms+YnSjX;sW4Iy1A?e>85u#AaYq@<0c8A9YZ zC^$MWwDm&2lhubKT13^{+|$vBO5nzUwD|mC-UW0X+xHaza>XwX`Pb|1e%11b_M=Dq zAfuPx&Dgh_i>K#R_nRuzr9BnjYPJC2)b!adgQbjrcEboSF3eaOp+At+`7fb`4@`j$ zFGHBfhg1{A9*GCkGpp1@&XF1$Wapw21i=2X4TzWf2e%w^E)p~@4LrJ9MhfW)&Lcg2 zhtw7?Sf*9MocOaP-%7V!@y@*B%0^If{9NdVyZ%J z&z_Bg`1W_Hj>TYs$0g`PJ17wUZ$Jf?{*`v0;mzv?P!hB@R(suBm%`18(QeWQIO^1N z@O&Iz{Sg2iniCN(aY-RfqTs)SecUhb8V$6F4SAl;BX)6v|A8_f^n;EV^%cC>cGy_7 zVYij~V?6b*@zk@+Ll>jNNAi}^d804t5@~s)p9~x$`hk8aooVVCbR!?hzdl|AJoQ%q z%(Z!L5%$a0ugyyt!>H@y%|-upwsN2h)fJ*x@!%&`rQKHjem)glqXAA@E5sf$8p>%` zvLcGN&DZhqU2>ii`v2kTE5o8}-nRurqy#BxP)X@nQW0qdq@__>mX=OM0qKxly4Gds zmJ*RtUfWFcb47fe;mK}gI_qj_r7Q5nrp5&&vQnc8@^io3N;b}Cw?HB zwfkDqTJMAVbTk|>BT{=qJwvPy5OWJCqr;%$!hrV6(Ux#1`}c^00E$ICLrbT;#BikR?SA(b3WlOEE!=`^cXN?Wq)sO6%p^&U4Bn7B+wwytNj2 z+M6VTn~CSR#VQ6ra}rAhNr@FfK5mVIO%ywX`UQ_~-jE)VmUc{SELYK9?6{H8s$hH$ zwvs(FPgH{;Fs&ChpA6Uq(Hz3cF6b#b%Vhc|ZauhPoUMmiIIQehfAQYk0Sf03rIE`J zaO|RzT2yR|*&am0lsb)j^+whhH6;EJJnWEdvD~UN(n)P^mLj((f-(cb(8HIzDyYvD z!5=Nz{DeuV|E&H=o9f>F8u4Gea=8%n)U=)7)U-8O1@!g~AlT88J{A=;kdQ;G@+@_zbtquZOdaYdjQ-?bscX1b9RJ(COUKiIAA&y02^Ij%;K*xDo`0(Rb zy;b(aDA+#GqSm_ct>_+lZ$600#qM3gO2&CPKCa>1_RN*&_tV^60kR6@7NAP`LqLS} zLEDBB_YuBa%zriw5}=~sy9Xz4#VkMOMkP}yqbZb8T2k^(=A%O+cJ;FM+8fG(k{eWB zg4ve$xGvSn%qd+Yhs1x}_%!!Pi*YS?M39Yj=qG09B1^mRLEy^H!5k(gZeU*ci9}oW z%#hvH+K#Nokp$B*J6`%FWDU3%wkhbBZy2JCVmxj{(-5TqATsWT5Zj{|Kvn_Hn14Df zp|9~2g&&8nYb6>`fLQctb)bErNa*TX_658}EFf{?p72RwqAdn(6}AotE6F zuM-IzCKXMt@e}JeABL#&Q6%={{8F+eEaZkJpq4JmN5pxxvFuO2g2=L3Q*ph#ut`t* z>_L!#-CR)5*Csjnl;#mk{Uj4tyP?DK2ytESCGU}h_=Pr=2|5L;%@7BQnPRM`S0W^1 zI0NRPJ{W3WRrARoO@gfX-Eesu1+zwC!iiEum*Z4lH9^FBXJPJ_KBpX8yZt%GHbIPt zo~8LiElScOZ$jwCu>bIOMEBtKe43W8ydI+u8RQ{be3$C7<|u?W zKi_;pa81ais)7?1o`xX%{?Fnrm(*WJf3z;v@-FiCAh#pl~pAp1C>(f zb4cReq(IimPF>yiYy?vzj8J9m++vO)zeJOmG~07h^=lA`Z zTQgt4TDY|gEH;pLe;jvYezI+QKVO+ysI-%;Jdgno0RM1iv*zPNxtBEYR9M|!?j3I~ z8^7yi$J`bb8eSp+5&}rtu&lJgTkcP=^7 z1J6C2XO5$TM6`I7KpsucBkWx5iyZi%WQZjFE|vU%2j>>+o+*3H>d9%A!(u3U{om7b zxh9EXM4f_zrC$>u;83}mZEt6n4OTOUK7|Fr={K<@{p2h9ug`lr@Z?e|->UQ)?ihYq z+XzX1A&PAM@z*#2&4@MHhD6Cym% z%a7jcgXYOg3|y{sAH1GdgnE{EUu9(=UzNcBnlE{(F~slF=fwBjQLK4K`B;Vrh>^8$ zc|#PS7%yc{!?fUkPEb12m$9Y9-aCag_E9c2tZY*f5rl!m#x=}|u1a6>iXVO-DDjh` zfJDfQn3Y7Cem~H-_jiGmU&SVPeGYzpI4*aHM2n#=()Io$wYD4&JI)Wke>WXconvif zno+Wr-XqhIdQX@ExcAIXJe~Ji!&I}X|erQ5+(HFh&pnOQSJAFQ({ zFXWEtL6)H7mi7JeEYD$EJ|%BSD}LLG75Y_Qb@vfot9@?Yci(|FbXGfP>L}K9%{S8H zW*E2Fe(Z;pJ65^x-slfOmnXTf#`qL;m5slqRD55$(Ur^ ze$c?>j54|gbLPAC7`-E5?;;0&bgF=gT=9)c?qIU$oL7vu{)$soD@mB^2$^jSM;7pu z3=uhp!)1(fe|+v(m+Uvu-b3pQb9>0z8mRZ}1gB0b$V?AmTrlAkrXRj~%I4pZA+GlI z;7XomQ)U-=a&Fd@iK7`~0uB-bk)nsfZ1*k4uj7(&I<~XbVIqezhb_T67DH2=%cEeW zx(1g<4F^@3O+@+hJeE-7BHD3DsTo|VfnYYqheil2GU@~->AcN@(VM-3?7#i;XNa#2 zn;0xsR%;FG8Jk<*bj(fp#>uNt zM(-s4MtA-YgfDzjA&l~lv1}0khK`y0^6b3VpH2sdI@@aZ( zMqSb;D5I;y>2M$;km`JRU#U1Xqsfa#Q9>$2OK{1MVRkd8{jqbVJ}MQ;JK17*t=xwG zx`Z3|vzlz$^igEhR(&n@OPJ*Y2`uAy|LZ)&o-I?~3s~R7hkI~}Zph%-u2Dh7rlqBk z5))3pyyEBB9VQ+5-biTDo>l_OLiY)+kF}_5uW;NP0cm|S~Y6r4#hd=Gyf4{cBJ(* zmRA}VLfik}_bB~NH&LrUuy6480uvJr`}8=OHrpvu-HBN>+Iqu&`0p}jdEhQeTki%l z?_gA%IjxZHg9$obX6S)J9B=mj{#pcVxy0CcsK(qu3|OBjI_@RWjaZ*4c0wpgB+=1} z0XJI&Lye-Tkpaijr#229H(VM#fcIsTkQd#MspET1jw+~hRc(sT7M`8b8aXo_x6qXD zbJJ@T1Hgk{>KGEuweeN5tGGj88sONb%q0YwB_Gz^xejf*0flX~Wz}2=vqA9G)PG4} zab+s9Q3I^?3S>aZzHCK&dQ8#VUHqrxYA$^*VST6VlR4JSUv1V^P!GVJBX@&}M4KC0 z8sei@SJqgcnR0|cwl?dKm_*x37xsTqyyj;=kb)p3$9~0jA$CI@ zXND9XHqp_37B$N!9w>wT34GaI_^WCC67Hb^S|hCElmM~|lYhG*<18cHC1OPm&Oo7Q zt1FY)G3*HvzOAc{%AA8fxzLJpZ)bouCIM*77n}8O-5gh4g|OMoj@g!B2m{}1XsD*M zt@*t;=Q0tWIqkE`k;)zXsA6HuI`c-7X3dQ+uhwDRwW;)!=yPC=WiU)(0&JBbo>IkZ z;wo+`@Ufm)G5q%rjTD`lpmu|s7H3u77;DBiG~|FR^vi9%bjOOjc7i$*=k&hd*8hzj zSpRusL74oFNnJ2d6thNTb84fatkG^n0<5EYpwHlsD*n8{rM0%|pyZYh3A&>GBK`_? zIKjOB7R|&12-rxXq|xrW`uyDpbsQv1PwH&y95z5HHe}qf&E^f}ZjBSRV%TjR;pp6~ z$YdnZSR52NeKEcN^Z4E(c0EBkg9rE235Xakv^!T3Us1*)U;ZXdHASTUhXr_Fd{Iyr zuqmd3m>Kfbl1xmngh5bmz~{Gpi7=(#l5~4>Y>T@(XODO5MqnX#OaSic6idwAxlqZQ zC4Pv9Tv4BmaONU#LLb8zw+c0sY!*>-Re(#?iNV7k$i5nOfZeYrQG~bdr$s+Cl z0k~@C zAz~K08qRZIos0KBV-K%**Hbqos6Ai@vO(7J`|T-l&_M9WGw_HDH5499d-)wq{WI7h zz=Vb8TDrE=!OK9QTY1tj>ceB)E zM7Dg_MF+1B){a{}BE9k89qnIlUeqmXQC@>S2#2j20>9((&fEUFa2a&rFT4!q98!A= z$BgBIz9$5e>p-MuKt~BY+>t?(7xrDd;zkCPA!$6TzAM-_NqRTK`v26LV(XGvu#W1P zH*n#C!qi4n{+ygzrm7S<+(k?B_^n;S)U6SYKfh@-o|9uL2aqimGp{lqCD_+%M)%zD z<;iZmfAOCWQ{*hFUJ(%SqZRJBhw@4dVt0GRydHiG*Y#)(8#r>9_vRkihYvh1kZ9;y zBvj&|iBP!~RESc{uNgKo7T;;oBFLsN;?sJUbIso}a>;jG5PWHH(Uteb(>U%dkFy_I z{>*ghYeDs`+sZYE(O254R$R%RE&;}I=~*I4+l;K%v8R!?tuu?GpRu^a4)l`rS`$@Y zSk~+`PdMM(&~eV#5ogna24o97txKf(YjujrSwpB%9v`Q>NJ0}X+-%{cQlAO4B7@XN z^vIZeqO;Z1Z%8$XrgU6;ma6o%*6-24tz-dziP7#V^RVA*JlG#R(g$2nHJ7&u2q>1y zN$Z0p%7Wb>3AKSHEsO#T2k&i}2NH%8l02j8n_pImkB25qP}@q@Lk{qpJzCCFXvcz= z7M*vadL4ZPb@eCE^HBJe_-P6?oTuK#6$~lWwY@?X`eEOW_DO0XzVvD*h>rKqYKh5z zD(2ti!f2T{O7dD5d6DbG2S_e>+&3fOf_}cLy6)pcblW|S7#Mbl8370`MmTX>a1|F^ zNP$#rJ|qOeXx6dvkN%;vvFVUd3`|$t7)`{uX@ZLVcvnlKxp`jY+Vw+kBZ-YxHrP}5 zRN{L`7?A1M2w}|EKC@Mx0jz@tL%s=+maQaCGv8h3B$l0j-H2{c)uSDM$}_*=EHPT? zrPJKkH=IERh&9?1-kVzw{fd`{wXs`u7o2B*e{bMJ22k?Bnli1f!3h&%RhB}J`G_j{_iLForM1(2@rey zEBbb0+EV^*@6;C~YA@#^j6ne=y!e2jO+m5Bqh%rpHG$ z7e7A^*1^e~=>6Z1jcqIwxQa{lLMcnabEjq4m$GqXcw zcfeE&jEba93$*om(U4P6jL$t8#;m0xWwSH{XOtX&?=yxy*K3Aje=KrjX4*O~3x^ox zdxYDgricK)wGSh(Zz;TDO@XV$;FHh~$UnaqUmu3v$%6&}nz((1I*TS<-|l4I#SXWJ z%Zu4|Vfsjna>&-e-8cq3p7sBJjeqmP12Eseq+OW66TPh zV|nL?-`>qWhiVrz7KZ&A|pLQc0nKu_MAU=>j8n#-MuC@y9aw9Kl_8|q<~YPlA; zzF(I(5YWT;n1O0g=DSB)gT{aE-6%KI z79$lh$eSO&`)+8uBB|fUMG>|2!mtnp*`O-hVoA4pvW5Xks*&caeo z69pjZPKFH$3;nEeAzcZ!w}0P9Xe&_P(vX!qM817E5EHKHoc*D)qQd0XkEJOina~Al z@>E?Q!NEWtQ{QuDGumieGF+bpAa{E{nw-v6xjpN;trU+Mgh6LkRGW7#k~NxNnpXaM zNl4$n!@3`*qc>ag82A7k@@;ABTWW~y-vjy09_A3@KHOG>lF5k_#Of#e|*o9MFsV2qUaVE)@d4@WeZYQ zfB@BeuB3_AsVrSvq&v`Y%+4Y0bH$oQ>*=nW*s$CAGg%6XS@HuWJ@Rvgi^wd=O96y3ek;B zmsEgOC1d!@sdw0SBq*4zicZ?#+lnc=XXF}lpkS90V(Cl(YRE2MCB1<-DW`zK^F{Cg z+(iA`lV*4Lzr4qwErWf}0Io0VLW!0%2yxqW{52%~1UnA1zDF*VkHU0~#?jj8lRguRTAhGd9Yk`dEi8LdCoyN`K?A-gA(6nf+CRP2^S*ctcR=lAdU zQ>(y>ZAKVE^m=eTfTl=f^zv`xBuM96Y$F@`IxqD53UN_M!jWD1;t)rmCQ9 zvbTnen!!M1kSS!UsWvjJ%1be zV6oc@0=Ng5a1RK98bhgq6M?HutC)ZxP=>O<)VDxE9*%_lVRnM+*svi#GOg!?B^;iH zN-TujLSd#haWe*JBJC7f$~wY*@;O6QuV4?ED80X@T|aaRYZanKLYaBxpb8c8$#s`; zEfAH6BW9#ccM;1V!iE@4 zfHK4zLXe@}FLtKT`NM=)BZ9vO-bg&$kx5a2Ed@G4mYZ#Q5Vl zN<2E}wgo9d$gNPF(X7$8+wD#k?Z#5=Q7z9AQ z;%xcOW0x=F_=&|s+hO|u`J~Edm?dvu%0=|e%^(`vHVt&#*qgZ4yVxOgep`j!#(eX| z*_3{d6Zc292`}F6&F?b_*`bR9Np;a*CKKV81r)r*i(KCBT zo_=z|LKWK=9zSJuKgsB4KW3Yo)ezY4Jw18yYbO&*aR*&7=Fwj1BSZ~{z8$?nZpIQX zyNjn_UOj6}B4&_CtYB#<&~WVO17}WV;fpcpF6<|Ch?r#^E1~25n&Wt$fLupUX{0?? zgoJAlzb-sv?`uqA*ONGq${>~e3?>{t5<8voCMG3NAMacxafa+-X*BM?>_vD{*5M}? zPlmdRwkpXSJxECBf{pJ3f%W&f!gG2srI+fEx>U%2r)*^`Oz9?R>rJyD#u3Rq zjIkO~ZhS>>)+A+(Vib^$`tdg5ZJ$Br&2cOs55U3#aOYZO9IS7!yFsdnqV1bc_0?F9 zLPnbxhJkZSdVrWQFm-lQ) zRMB>Fahn>zPDBbV8{}{$)B;xy;Iqd~po=R&X^~j6v;E8*Yfe%@t-B5|X1j*j8125? z>@mwqewOKlO_%GtpD~ z{45#HpzLvbsYwpX&~nqvJZ?dqT(92eVzg}Z0AVi@qX~E^W4khaMeF~?rvd&?e&qCpHq1 zTfLH#EUV~SOHUTjqtEkD-LM^Uft;=!b~xGas^PbIq-2O5zP@qBLqg0P10ZZL`8upW zso60itJeRObbjV z04DABXPd(>_Z`)}eZw=duyFto;&W6-keSWQBu{KGA2RIo(7Ycy32gvw4WyRF15Xd6 z4hu7RT+ostMv(sDf0wRLEMFvYv>VbZhiMRow>+mZ&>A8r`ge(Ipc`%mUSct&0yK~W zZ#(Vr0yhaPC7K4ukwytTgS`)~(s4{e+4mSQFRP={!Z&OcII@QbrKR@|rWdLWC&DGo zvi%ujkw*lG0Jq!kh)ulOc*umME{v#H0|PMt!mmL(MH(e8thl9C#ld^5h=GB`23RI` zovM{=LO_WyR;ehdG1cVyi`w|!29q+RsQ|POU`O=BeYW&@bd$(RuP76jJPCw>{d0TL>~2e&zsUP^abNV1b&1`=$1RJiVg{*)Y6zHD@?j=n_S#$_ zg?|-e>w2tUFD*TG>}Pe9+8~cL6YYeFp0$d^OQFBo?~;fHpuf?(I2ec(Hdi zlQfiBg7k_=e42_)a}-02Au6^~sO$=+pNF@f?wXsctM%}qtispFI`ZMQJ_^?T4)?!` zXWdWiT6wqP z#87#DDPQhG82I!dfbG0yYq~si-Iof*_lQWWO($n8*j(#ATwjqZ)CJ@K$> z@(Ou+OfTr(lpM^DTr;9oY`xZ0x02apC6n1;?ZqPM=>E#+^0n)I>2y69ml5TNmb=pW zUmp9iO4BtN?%7`cxo2olXbIl^q%BJ$^^48rdqKOjF zyD6?%+B2eN4yI4*+wCcgE#Syw4;-sN?7@5%Y|&c~8_~|3?23NbK}gFT!fK1*2}}~v z&$(3pJt%?Nu#XMNDvqmZvT#_U8?GfGvQi%NwXN)$-i>~ueArG2AAR0@F7co#z={w; zehCT-;6!}R5zBB0HFyH2@AuUeRxL7rvz(jPGe69F^mpi2I>9Ec*hZ?MF`vPT6hXLKQl;PBKh6m1p{$Rn_JTeDa6 zwoiP)HKrq+dSU%QlYLp%3*mr$gi)6ARha?@gpOn?7^Oq}HIy8+aV{ zDuHU0PSKE{BCTMbq*&rsr#6rL=&iwq9;;+zS=)q`NMF7>E~S2g3+1aXK^fYnv_@_x zs;N42LSfVw{qoH&mNoB#`FgXgS@uGXDrPJhdb30Yb(i;Axb8fLQC=){;?M3zxqr1pi$(0Wmsd7=kg?0jSb)|BXkWI<<+Sr_FY3?t1MM*dJ_p0XDM0F zu`I4TNT^Kg@%Y(#zkaVcTHMNlaBBPBv%%pyS|Ud9wO+1mHGxbJ(*Fd>lb-3_5dxcO zk-gjEF}gX*@2&T)!p0)a$(FUE{8BhU51Ze!>02LjdCUqlCT`vhvVVz~Npgp;+*f|Z zpYkDOt=}e#bKB|~01xlyGAEU=FOq$owO|I3xo0jy1Q@2v-M^!8LNB61c+~LH)&rxbmY{bKF7i{3g-|29< zYXCOLmm{rlINnm9!BiE+f-6sGKRvp7W%R;N3M(X3&hMxndSjlpZmsID;L%;@$Go~; zQ)x^hdXT?j)1G&Q)2$$@k(MDAaC*)k`(IW}BAe9wV=CtWs4iu2wsy-sMNqc_%KiWa;8#w^%O9`* z-QY$#tkbBydCFhxid?fQ4Ie&<`+P;m{$ER&vW(h|p`XKD+rgskEa~!;wTij`cbus3 z&F5)&PsSD8j=)1;tC_Ln^C$`aan`4m=2I>g0kzVV^$$gUCwg+G2U zOfOCxnaKV@%fmSxJL7CXynC1O1Flu80=lf&h{E_1P-mvB33`12MWUOp3gpFtX$16d6ui-zNF9G}A)do1-|x-AR4(CU*}MFw zM<}9Q&Z{<;&xA-h-lgl6A@RG`uj9zJ-%PKGpM#|St?;T%4!^u79)W~|F$e4exC6NK zr5m#yh>)k7G(uI#a05&`6zq7 z4#WENKMmHy@mpW+cf$6{qYwXv5-lE$mN4tKy@dbB#kT&yv`|BO6BN=O=&zaOjytVX z$MKBr-U4WDOHx;RajCSW8bD*Wcrv-yC6tGZ?&Y~PPx}WN@8ePSr1q46NPsO`-+aZ9 z(5&P|Y?S}haC%n1ix;{4dJj-3M=C11u9)d%h6@9wX9SUDA&NREfd9rA2Sjsi~1 zZlS`!ku{OH(pjoSON_%OS*fDwju8@?Uv5??aVPgKz{J#Hf_QEQfQv-6tg&N{wqVO+>n*me2#5y?5nN~S6XVGz!xx=qk#O9;@M z^D#0QmJ)Hn^qme{*J=>wUZZ*^XvaI^rFiFYQWMe%bJc@E%GjZfTBPt9^1g@QT|<+B z_;`Ce3WgZi6RW_hK?=1#2^U$ZeMpH@BFSiF#GAx38E&s>Q@^^p)AI9q$Fnw;Jo)%0 zu;KPlugiZmo7yMh>Uw$){`*#KPh$>09(&kcU+e&^b5z%}7$dt*X~*+#I-6CHziez< zrA%&g6{Zj^qQqj&`|YJgfGXl2#TbF$3LYag|Ayz(3G5zmrnn~k0qGLY|`?=$B_ zKu=o`(Xelw%t=>qKH)k3lmrg^N;<^mh{}K3}vqfJIDI$VY%k_$i6TyW+ZZcKH z4Td?kxNlp=&0{qmsExc$c6QwCEnru;yJD<=Hst>_`~#(_@@TvVp+=b=9bOEE)YcLPYF42!zG zSu`%p3`EZvUP`@Jvhv;BbRejjYk`ar5*Uq3XSmg#ayJS8r}p>+KbRX(GAuOyqa=f~ zujRr?MC6+V1k>?{+I*EOZ7>0jATuq_9`L+D#h;sRraIm~(1_{6^9f3E)K4p(W^(9% zSOCBdvbtbef4DUN$k}*COs`&C-iVg0TQM3o!6>*>Lty0f(69VRNM`if4*D`4GI0l& za~ofQs6ubOanb!1>UrqS?aHTDPa|N&;K1Wdwb8Sq(v&}p1>PWpsCBXG$ycKpgDaDoZF&(D>R^Ih5S!Q4zdOWj?-aTWkB&+Db1G&v~lfhW4HQW&Yl+u^Vvv(=;@y0w=T#R5m^%Oxs;_ zcQh$0AMXkj%jB|6=*my|4fgg-tK9PI(<(T(yWwG{U zfcVHQnfNX7r#{{OMHH-YOslL7CRCC-Folx)WeibIYx7(;$J6*=sPPF!5d8h<_=cHf8=f>R4>KL!`P2|QU1HL+K|hi)hQ z+9UYDk)RV)Wq#|{w<4ev_;#Q_&TUPAYC`Ml5&tA1=&$DQ%<4rBJ_k1I<&VG6JLM9~#wvkkG5Yy56P23?;}%Ist9=U(#IH@o6^xIHE#J`aN&O1VaXajj%jt*D6V7 z*aG5ykvu0H(mjx(SSr@{M^pkA?7~ui>=>C^LMX7 z5ifUy9fBFdw+V1e`*sDy(O0f_uy67AUY0a+K@(OhTt6ie51U%Zag=K{=4#P72x1Q0 zP$;_eld+xk&!*sqn18e)Tc$H#a-YQue9mVxw0m3tXQUKnNnK#9tk8BG|q_`#OM@uS@DP}uwzX+SAq#R0_`aq}vyaVX{ObvvVpq=sj^0QmA zfCv_l;xK+`76hf#p7qz$EB+h<5_JZINAPWT#O!JpC#tBRA1=(c>SGg16=iX)mL(F#@eG!tirgS> z&~iM%Fc8`a7hc#c5DhIfM7_`?+}(lHJ$pETzV+tyF9=2qR{Cx@RJ64KOyA!$Vh9DcY#age|FnArXr z;mVn()273rgaS|?**Gq0EPEpT_vaJKdTu=fJDO-}DfEc&9jcD=q;=6fnY3J*==L6y zsqb|<6RZ=r-c--K4!yv+yj1*Vo7Xj<6VeL^-dVF3at4YZjM+n9dD1KbY8#3DEdmb% zoqsrSW;4bx6`6i_q_X%Z?ol2Tb^o{Q+81%7-A{V?ZY6UqGz>r<45)Y|#`S&07ZUN4 z==>69^Ovu{nLWQm*vX6~3dZ5k_cQv7*V=;@RKtd|D$mYcBD20Owra{-z*?Rq@b7Vy zRRn`Q&>DNuV?kV%GVr>uT*wYAQA^dI97{=-6m0Ja)Z*&!`|pY$PCw{iT{t@(Mf z4~VrjogZ|Km6E9In$k014`NZQD>!}|;yBOlsU^(u?~5hbZ1>~)rTOVb5beF*E56+} zUs}qc1eTzmn&En@y+lT3{~gX*l;U9iR=3e~4i}x%IjUzw(LE}8J$E)=i1zv3$=2`7 z`S7*QBlXCQJB3ttg-cQVA^Q7@OP6dLv5h6a4y7$|JjWms_UYkGW9l2?hx7`C&?MmRJyd7Db=ubCXwT$UbUXz5x%e3v6m1T{><+`Np+@&u1=4VyVypl#e z?mv`)qPQ7*kDsyYpg-|v?^J2#1AAr0i@L}{p4RrXOQhc|>UQmpCfbeACnEH$4EvUz z!x-KNO-lnEbIMCivD-aiSI$djCEYsw{j|*A1oJhHy!7q)01s860^m5M<0Z#`PzGN6 z-#2lC+>rOqWRA$xTR?em#9-(~yPPp$^Iu=Kc}Vvyz3Ye~^=J|e_25(weo#XwXJh4A zEGqKAfrnQ?E634uw~4vyn-yl*r~Vlndipq$#7|fY6zlcio6>C-kP2UsqXL}I3M_*E z-ZtXM;o;6bV$)e9pCqNMh|@|tIPX8QKn0SGs6=yB({#t$CN6lPt&g{~FumgzEe%5W zA#-F1$@z&MOsO=<^_A_3I`AIA{~5TjE5f|TiLE985al%P00U~S+cF(Ctg}1<6Y^X%(R?6@9+3>`zb`a?babke<^3#aCwXjbCE%r zI6cMcEvFae!haPrgQg+np`ja@&k&2?0&yVE>`zqOkc*0puEsS`;qszvY7Y(vici16DIa%zkX(O z0w7j!xxu%$YPW9SR^<%53mV$Y5LD3dkon9U{L0gC$p%VgZNFSkiNpmr#|n&W;KYk^ z;*XGzKYPPR#oksP_wLufd6*))s?0>|@qP0aTaDxj8)>cl+wx1vHe);A*93aWtID@d z+G7~D0B%S3)yflghwCBtaqDvp$C;^yK*_v#5zahR%HAk7a6e~ft`C3nvsYx0tG9U; ze2!U;$SEd=2Crmx9$0J9=X%2DW;={-w_tIqd_5T@TX|o>was<%UXGdu`&d zX>CC#y)Ug#Wq5SGhFjh2$cW?bx;nEeu6qY<{4_WTY83iX23Yo<{}4lF7seDsaTA_} z{{%Kj#7I!19jdXrNMTQ{5T1Gd4&mi6AAu{v&=|4V3812Vi(p4{W5CskRobLVa_Ph} zFU)CXpqB6&rsm+$!L?$kn%?m?@vT>Qc?$u1$gxWsKO2XCC8XE;YEp z8R#nzyFiP1$Hq2i#TiHt^#`n+1f~tTy54WZd8d#=fk>XYYb~MDt}(xpW+M`>x41g& z$fXVpVy1nU-=*vQ96bqI#ZH09B%R~3mnYg~6o2y@9Zn!5;HREJ$$^h{zXM9|sqL%` zn$G6ON$~Qu(dlfzk<-=JF9Vkjdf;O(fIPoBLnZh!@8;>j>xFhH2FQNu zL?{=6|K{7C8*6Ph+(H*L$o-LUZ2oY|DHDPVh-0uy4jbm=^GCNy& zf{Q^e2~ugSL;N0)4;)QWA}{DN9(*!DwT7#SLfgpXQ^4kbzrv7ceYwCSXAX`J>Q?c_icN$JOb ztH!@uf!=hJ14eLU%L_U-5Uq6e&6ZvF@8_hAF~k-C&3uX(aVAo?8Rp0Bj;pwHNcsXm zC|#ShF!#e;uyjSIzS{y!O`E|uCN{>uMs?51p|C@5^vS0;j(7l>JMR1i7|q|?LY1hV z#y;oG*^qT(4v1d9C$sJf=l6RhUy-mM$3dObZ`yGI4_*l_;03l!PEp|SBy?(( z24BCcg{}VVE;lk)EQ9r7FNzR>&OAr1l%L{i(x6>5CNqh@vAv4&U9}VySOzegt)bgCY z?dim=yj~}du)Jb8!yg(Yg3skY@Il{n;S9hl4A^JlWK5(lL*i?;Vfzi=_ZKzM6^oH^ zUG4#~tH7Qq6R3(8bK7GtJ|@_k*rZfL7BcN1Xn=K0yHTgGL0 zMXxT3{P#9svE^#8kADM;Dt4^!(2`Lcw4@zSEhfnq`I)PaO3Ok3ZR& ze*K{nlc6TnP2i%7II^}E*36}c z8ukm4b>=b`3;}AO^=(wl1oKm;FGV3?QVVxbL;i&xN!ioK^5p5}1M*?-M{F41g;PKR z^2`Z7(WmniJx5V_MgO!=$Y;@^;~_d48o64!F+HiWc~1U2^tWk%Gec@tR?d)pGkL^z z+69~YoQu+1x@EA#y~7htqAOteKxn)f1ptJ=Se))4$lSg(aFVjaYeP#u-g+Y4G4>(Z zMVxjMQ-Gmpcu#Kqj7dWGl5x)9lH;ms0i6B=D-oevV#}CiQ+;DYDnC=-`8NrqkCyDN ziLobL6XBA(_5v#|%;zHi-ubs_);qFik|KV>U#+BWFs6$=Y3m_O0Cxv=%xpVQ6?L5C5X{$bk@K^E*OE}d6#5{Nh+AZ4DW1YaV!$etm*s@d8W=LT zdZF{iWx0fLT#m?%!yPU>LerJy7B_HdxjIk!YqLO5X|5zg;n7uVo^5}rZVOcB5<%Fk z$5Q_={fQj>cV7nj*hItj<4>jx@AdomU zWlz^2JYV%R?D@STIzr2J&CU6pZv!s65PAZJh*uZCZT&b8|MxY9P==RU{o?HL2y7*y zr)0qJk~pHcD?KJe zJX8+9|eDD4e<|6Vie&)eVJ)_{xk05TlFG% zjI5-7Ufd|K6zsJO0cc%!E1Yqv!f@`vA7F?T+(lo>jk8d4RD(PFqSIdV+~Lt+>vbe} z4!dWfO%m8r3*&s?r^2nT5q!}>CFu>iX*#Yhyr8q9#fo`uFB61^$cwFJHqA8G7M~X!^;}*b(WLx1* z2eG${bY6SM;a<3kEBVtvmIgbHRv@5f!P~UNc`?d2Y5eonXnlaUj{tARIgB9u%kOVG zvF|>k+{;`kY$3N1L{HT_SkJW?hvmC+a>&-z)tx4SZSB8wVthhEbmvaiiZa zJe2Znhrfh9XmeFSL7fZA;d4(X(QJ>%2vE903&rnA6tt-zmP0)WR8C@-XkQzWv=W$# zCsV1|vN2;A1IH$q*KeqNsfg~%lJ(J!s>td6%JxLWfqWVq)RX|7W^?}&2VMgOuQsLv z&|it4WB-hfv;13em$AJ?R=^{-zeX0}`E*K3{EIxf_xSm{y4Ke3hgV(#yn@7o3J$kb zbR@q@8}#;T?YZQ`!OLlIwq%RT9ooxLu(?(*-?8h z(-9qD{o&N>&*LZUH7)Y&%Yq*%1&y~o`)yktf3hjJa_O%r)_;Da=`%?hMb}|3HUY_^ z@|j$FH=p!(?djm7HO+|kbHcjEoab6@TnbgQ0R|5UYgLu%G@hl&xZ-On!&x#15T{0* zBnhu$Fc(VK{9@9&?q$#1=%z1*r=E}X)Rh-l`{iH3!-PKY3Q$l;7^i)N`V_z?L1gI? zSiyH_8BxzCC3BP6c;~gBo_w|tm+$ArTCT`b~V6l6(`P8x>JHm+C-8jA2{9LYTp)Md((#ST{^CiT>GlD0&k&d z%1n%@h(q$rqta%5`ExMnTdMh{i^!Dkb-rNOmcCo&b}n|gVx`aYsnRBXp2Q7-nFmQ# z4u{Xf%n$nfwD{X932@#L``RXXih7MHE%w{)y0aGHuqQAb4U$%0UwDC!o94UvTxi}E zmuJ~qI@X7@1TszjnWeNU_t~FN>yo`Lwe-2BJz-SUyP4bZ?5VEbovU{0D5+z4$HY6m zMAQ(-plZ8wIZOwTICr0_<)|I4cD08~M^13{VUwTGxVj1+*WmOs(dM}1<~mj{Y0o))XJ_Ybfddi}>qUY$k)7C4jP z=AhpIqZI4cudWoF)jkVH6gXI|1D{6d_KX}kVeI_&W9rhVVZ*-fU3!C67VCb2Yc(wq zxQ&YR3)1_w&|Vq5G0JoMR2lC#-)F#Bn}q8q&$5@le(%m$Tg$_VCTTYs8}lQ6e!uPD zS5HX-!-1#?uyT1N%^H)4#c|!?<)2E!u(1Ab=`N)i#d-FBP*=?L8&K;nK*2>}R(-Zp zO={$y7-Ma0v5F|gw&`3nKS&oVY%`P<4y|LJ7!e*|mb%NY^ z!er~jyn32U_IXMjsVwlFJnmxhY=705xy0ocOSJh1TbYN(1SH3I<$yI20}9kSj*Gb; zF=^lYjgD>xCm7!Fjd+dN3r|U6i|Fp9E7afwrs47ksn@YvGrNP z+WZiR(eg4WRpV``Qg95kO_DbTH<9Tg~G8g{)xo^ILyD<%2}_&Fa94EfTqU5$9-%2YHFR-59}bK(`R(- zcBy`Y1})+_;!XJIb|`am{&{t!1--PPHbEDyXI!HD^d*fMo(7%mpI^Y5#Hv!cmV5;K%5=;!zWiQvNA-Z z1LGYUxhNODmp>HPTK2`KEpT(y=x{|svSia1122deOaAI@{H2n89DJBD6p|vmRbIQd z=_38Morte)@z#U!AAWsc0c10v#YisI)shU6lXQtMazXeWV_nrBX+e{46gC-x#(rWORLz@AlMv?`96mAor$?@%XFO}OsNu}U8P$gTQE^7Qr&UQm z<%vi$z3<6P($AS197YJdpt1ttWnJoD#ZiO6v$EPVj!Ai-W-4!VD_8=AUs^TX`IML+-i ziHZ`~OX}sd@ZxicQW9`!%!x1BJHfBgvv4(ibO&0DBxNPWRifns2KQTXUDKG8Ev5;@ z@7nry#lTz09bnjzin#WGm%s9WprIm6HZ)81ZEM4KO#nZ;Zj_8G`2lD);`X&qyMP zNq_M_P4m2Yg-$LO0!Fn`h}Y{?7X*Z5?9mTSF7Oj2{ufJ-2a>*zDbbvlEkJNI87F-5 zmK$HnReCI>J)e_XAqRZXU_!Jd!Gv#)_GEm*PZijYPkc(vV!j{YAuUjG-PfdF1>S<% zk-D3bsKo9V_I^w&WK&haRO)ZO75)Fbj~t*pcT^I{K3W{|$XrE4D=fjKNt>*TI5d+U-B`WNfA^R(>)_S6gri0BY!!R5!# zIT4$-0uwTp7=nV0yXX1a)mp%1pcstm$yB9*p$bGKUIJjU#|PgF#HMB8j=R3I*qOr( zEm<#o#$L(F#zwbkWMURi(%WZX%&PxI)s_zLF8MaIHYt4ljBc(`2OBrh<@XIY_jQW6 z0P$bD-gyGdKYB+kpmsVh>Yet~ypi-0OYyW9%QkPx7_dvn#-=DiXX1`-`QAW%F zK^OjIbO#ulKBQLW_5e8QIGp>u+NQy0phD6xUg7vG7GKv~vug$ii!g+nN%pbNB>15o ze?^D?V+ynYyXson0uc?t(P?a#k;XIjx9q>jRWUXS-Cy;^pCR^Ay-q+AHAq#sHtg?i z<7gTg#9iv8pn3@>Y(mQIsYa}=jtmGDqyP%w=mhuVe^2{QWpaCmsdQUrtf+jW^DfMX>1V;dC}ZSd=s4Eb?p=T|JHBD|H`@;WR57_4lo=WfS8>jssEi-%_ z=1)r=82-3VbQwyKh4IPIAK%#lK!7-<{0BMqx7=-|Ab~^E5E#j`wAd)WeRO^AVsO@) z3j5rs7A@ANcI?2!0M&k=E zvqtcJg@*eJ{PlYdpuZKCnP1w<)Sko4Z*Hi)WoGyYl}B{2(U2ROnOvK~-4hs*KT&Q? zc?x2sPyM_31OsmHdZT<}c1HV$@MJW^D9r$odk&9gfB2chHBTF#4=&xMsPuFAR0L&d zZzP8#SCl?k)f=7nbG5|4LK^AB9yDtd!asF;nn=GV*wsH183xcy)hy6$|JewynYGLX zVSgGB3m4!5`OOR7;5s`0AxcbX1E@8*X**G+DvM&*@Hg&+PV>};NrJCBYYA|yRr zu9$C@j8$1WoIP|g-Ur|M0ce+(QwS#+1j(4aAm9Ww!dn6UF>|+pzlWiJ{bCNV7s&U4>wmDHr)^gLai5yOck*J1VVf%EY@DEW z$E`kKEuN4S#$lomiktps_U2T0q87FYioxV}$-@jsbV0IbR9Y+LrN~f)(zCCju6O9U zOSBdGzv?*He@wDwhA-?d&@r8I=NE`B6}q;r3l z%nO1Ct-qKFATazi@ybHw00%>GhuJw}*u>DU1Z)s2@ttPVB*m}fN0sed&_?S|);<=b z7!c~IIgq^+kAzi0##ek5ci0Q>j)krY8_}^S2)jm*&Ut zCqHTfU%nJiSGhCEC=-(H+_>>cGW>U#dl+#3c;war+>u6^%NW*5=hp~;8Euz+vdhpo z^@p@rL?FpcVT*b+At7<7rgU#x=G?%MfVe?euec3LGeaqZK&smr(yWd41KcUy;seVG+$blfIiPH^?adx1&u zpUwki=bdQ>2PzfpWmj+Y7_=D?`zE?ot0u} zJwwXjI{1R1ML{!t0xNGp_mSaKBvoMz{4xL7T&$I>FXz+m_|HKR1muRcc8eK_y6!r#F1t>kYQ|Qu?t7;@23Hvn_Z*JXg(3fs-Zr zHm#HlU?Ca$ zzd85g(>QQ5{^gZqo$1VsxyOg2FdCiE5611GX-$)IcV$fYvRQ4`yIXalplhgi@96_< z7KdGby~oi1u~I9`)k1<2E;iZEx@Pw8RM^eZ;p*kyXLQQK$x;n;R0TMSL<~i9=pI>N7vh{u7qgMWa35s?HY)4o_OA{dd z7gDfN+vZ18bbOeTv+x9c1=gCo0w8dZ%~RuLU@BQpFy7x9k<%TD9XJHL%MJ1{YpIF; z2hZx1lE6=pJ}|@Q^X>_S0Qj^|I|U&jI0U8zl?n&vmU~Ky5&*A!F?Oqpp11p5=&i{h zLEf|#z+=$lP)KYMQJ3{wY*?Om0DZJ}ODyTCx!D?AQ7wnNejJ;_F|${)Y?X zU2zlgo$BCMHc=pRy;4S6*OQ1XI`)s)uriEte-3bCqC!8jluBSne}FtA0vf~(5`z5k zI#6M&S8ER)h6jjH9_z=CMgPfB{?k~XlC(Gf>53SwD>|K8=Fc2FqGe`EQ}2c{Dl?AG z0wCQ60Xw~V%>F-iR8(0|nT;FR7Y=0PNnC1t09JVL>+mWuWA7qjQ%XQOdAhtN?gIlr z1gIP}>LX_sC;iZGpPfhg1bfvKTtPZJP=F?y?0Gyq$p*X_!3aRDu603unee7N3Y=M(59;b{Jal$t&i?R}*-{pT=4j z&{zVOT~X<}E>Y=!9)JMm0Ra!EXHCCp0ZZ%}sk@PaezBu7saU|cd(3T0_oBKIRN4nPs4+=O=8 z1@IyaTZz?d9Chc7FaSD6F5nd+L;633Ji;I7F_}B0{yLzhM+fu|ql7X>IV6bxRDOOs zq`}(ry2HgB}K8gW3%p3qAlGzEGh6C?knB zbZVePPTb0wR#n8ZU;HZx7V_^gi4aH#9tq=2bCj4u(FBsZeDY;=)L7% zMD^dGz@x3IJznW_GFhv?G1V~vbdeumYhKXc0rER*CmX!1PhFaRLiIcbQjUNiMX)_^ z1GEO@S*8i`YYKB4W`>md?c%RYzTrQTTLMwJ@BezqIH3H32k=1Ml@8&;kMpr=keq;X z!3RK1R!RC<9`(fm7x$kAiqnih#w-9p_NpNw2!LU13=P>!+i!ln2|8Jqyvzx$B1wK~ zB>LY*s^l;|G*fo@;#E%2C+N?M07dx(S4)-K-RtPG_J~D}Q0Bi8A<6&zkhCmjol8oS zk5iwE_Ho!xMf=}H9f}fRANGH|5g*HMjkcek`IcQcOICH|VT)zKw5bKGitceAAVvUs zUWkoF^$1T*I9M5a%`qwCld%~H)|qXqzh+pn-Epz0yhkz%@_qucP;>pqJmtql_i{1a zBL#l|e670IvzF;ui^RN;L0Br&UB1$XJxuo9QlouG#ql+Kz!m)iC!&>yfB_;|rxneb z_AmXa4d*{{w%HKTcS;2zZ+NSIgeRTxiOo9-9OUu@bCbn^A#6T1R#2L1VfA@=RV<+$ zZL#LBC}kIv9bI%mGj6l;l7Dbp?dn2Vjz6XIYZsC$C4Na}cO{ij9Yq!%rOQ|!{m{7l za)kqYPt>N3cegcaZ2BR;IwQe9Cs)(=#$ID7=HOD#B3rWcY+=6egk`7x?#Qm5Wz6*< z!AL(W&O-~zHs;t^Jd{~E*nd7o(u- zDABH7X|a)|&ev$qUCTloq5<9zj$$CaM${i@j&h$LOlyC{No%BwO7HtES=M9Yj^7az zvYQr@9u-at!va9}e(`=Kz}sYHWI0cmkL{UofDT@<*MjeBFI=O38=g#q>QOb5E;t)m zcS!P2UNP3sA98V*1q>0HN-1KpGqJygQ?i5s0-YI+WsC1)I^{jPumz<{@3V905FP`1 zeyNb@mk^_Y;RV2^ZhmwZ7CTv5So=OaFt%8g?5quYk~i!@!(E<88dBu3~5$ zQJv`HObqEpe$rIF%3r;S<{KvfA(YBCj4XIL#h4=8Vzg9_m0MS=eT*Ie&|3RIQ@+#<9eMCKmuUw%Il6diIhf^4-Pq)*D# z2Nz*;rp@1`ZRY@_RQ*svfYAd2PKW0nTI6)rkUQZS(i9vgH2)m+-o+-L&1(+I=Vti~ zJ`mi$lC~RH2tXx13z$!kRXhW6><%l}5avzaF3S_<(aa2vAx~$@hK-9NxZ+IY?9bm8%e7Q-8>cDhIm?_>+%qyleff+O3gTu+5 zup3)b>)&up<@C%4jwLU{#<3E;J#3Yvr?K`);BSVaPIhH;K-iq4yo!17$y;IN8%I)E zT$;$Kl*;`^wbt2>n=yovR&PKMM>++245-HvF-o1?znf1AwZq^DzZm3B5P|!A#;vgw zz?kHDm8Tv)1@t*zLUr3F-{{2eIndstZ+Hoe<5i4&Mh8d(HDXz_r#)+6^`S_SUFJ3u zey6+%{jRdFBitn;@9Z8h&qu>d2m>d258!PTu&9(yfO*qz!2Aq3@Jjs%{eq4S#lxZR zvL(9PRbRO$h4^DvZHn|~O=rN%m6{G7koOgX8A!?9K*yRxx5Yrs_cSa7s8QntOemEV z6sRj*)Z%(o5XX+5Q0|r*%W$BLiM(f)6VGoxBU1kCX)jY&EaRN@<55Kh`QdH#N}Ijh z`wn!cAQi*`La4E@u{{;l0gM-;!^<@#Ax!oWSl;G0u%Xr zBKCwL;2da#oC8MgFAM=6sFrSz`T)2}cDwwnvlja|rA8-4FF9+GnC}7^tkW}$blW0T ziM1bx9vA|E2k?A=Jp8b?Krp)Tcu?l1}==H2ZSUpd?`1VY#o8GABurRW#VFJmp2_V zoRp<1Zgl~^#V$kV;sMGXoFAA677)r`A1IB7!wWkV-ULP?b)A9v zy5Gj-)G&Z|=y>rB+sFUC0eDf9a%oc@KR~Ip$JYp2Bsh`!e|`5y5cF{AlmgyRzJ$a{ z{O^U2?=4J(SGc)B05s>^E&7=M?bP4C{eR=5z{0z;=w^{#HfiBS_xf)AMXzFBWnYBU zt7O($)$Cu37XlyQSLCiNwc(yx=5373QxroW| z#BEpXIZ)C#7(xO5ea{o4o%aV!r1m6Cvlrv%dT4#K$uDH|IfX7AlU#E?ISM8m&&S@$)!c+e4L=x{)J~bzrJDrFt&+v{C9v zh7Fs7bKwSqD(mw;%oL6^$q~jGo4|C!Zrmj=Pe(>hPnV`c%H1(k{l$&4v%eNh*Id^| zdPhz#8z*^Nt>3i#scrv+x0Guv(t78;Cckx26Ht)veye52u!FhZDgu|({L!k>7L|Dp zQy)GB3?}}id)A`IjJ)CN*jg$gQrZfa-P|YUGI#Y&r~$rp{?5CH;cS?#S$Ih!nv3}w zhaMB9k>TsL=3u@LGw(c!vCzl*Z>)#8Y=>H|8eEseRoHjSzRPK8OS=a#!?VLwxFIZO zMg|5m;p!{RUKiMIAxE0Hr$6z|AF5b&(WmU2YcUXz`GxyC!xw7s;`Yx4HcMO$oi0`S zUubz`(Cq4u!MBq5=GxKE8roe;hpc1ZnUjcetF*2@ynG&W^eb{&i43mLArbNI z@R;x*Z@rl?+frMWo;y1Y%I5ta7T_icCUeb9=O_xLbGU!-`;#joXPXgbYpUxWSe|G>dpE( z{zaeSK24-In|9;=*g`n`!_6yo1dPb>V|7&61qYGv=_yUY^L#5~@(o==my)^0V{E4( zzA+A)^KaIIDp3AJ*Hdh(vD4XR;<`Dq#Z}k5Y`Id@3ltAg@fjC42Iooj*{)L_V`Mp# zGJoh6qBK@UR$KmbvM~vTE&3Kb|IFB{gd=A>)1-NH@jA{ij(qw0gc^*Gwzk_Nt zs)#SPOX9K0R+3LMU9#z&XZ)0d|H8$O$(!6hTR3v${vsQh8VQaXiPf<7?qIe-)<|fk0m%@M6RXKm*yL$4hA?6@_a3QUMFszmpG8k4@8$(a( zSs%h%YY}|T>6FXA!`@3Q)76mPd6%OxuyXx^YG(Kk{bxSx+(EvaX1A7=JJTU3k@1fL zXBO2)CaFY5@&0p1YUxFi7bgvw2f0%Xm0|Z6*vsE@z*aLe{4~gy8Vw)We2wh3%*&-R zbr2hUJBMD0^V@@0evNUjjGiv|ZWzRltwOLeF>{ZEHgoOos|x>m@jHA!QNLrbjF9ru=hTCX5S2`vGo%FDjmfVfAhAhW!EI? z>?>OuABwvu%*SrkqXI`50bzt0qg5uP+#B>U4)%}+!h*K>pR!@E`2F*#fBb&IS5yv7 z&`uG^LrGQEIXs%f(l9-H-KI*gAjB05Ppsxuc~`4~p-_@_k92*|B>S_X^5||Yciy9} zl!8I8{Le&D7UPw}HUptck#WHy8!q=a$7=$U<%`k4dFkV4O`{o_FO;P@56KSp=L5%y zydWZU3as*(0;k;|C2sk4l(t{tqkgZjOw8wN?hI{?rKXYBXs^vPo1)Iql}^}9b~1hm zQ+Xhvs6jjRl?q$GjnjumnuouscANIUL1jRO4wsmf*E>-3Xft<2SZq7Ix2`D!t?xf*{nop=6NyNczQIuw@xTV0Haqk9C&Rdi=bCR`0yUWjW6X1KpJl^`IKWN8OXIStR zH$*7N#f!o1_Qp7cFnt;Kr2R$ahU_7x{u>sSpyX_>^Ae^8gJ2&v_bgVHB3zcb7u96C z-{T^70#51XTJEPTHKIzRKcN)!v39RUW^GA&aqSz0%e4Jna{eUCANJ9pLoUlk|8q7g=r z-%?iYtaV=6&9{xuNVOZa;ki=m9vU(ruIvxV-NxkYr6xIK(kc~RTMskdx;WODQ}}g^ z*-o8x-zA+PweN>odXF#bd6syvkK7eDSEO-XIxkxZm$AOY!aoRkFX?GOS>^Ht4}DKy zo@Jh6_2ip^fexKUwc|0(yHX7mLMp&e9BPx;@aj>E*{Hp9R zQY{bd2~7yN?b&c=(Q2y27c`3r{&-uS$h?mZ+p{C76|$$SAY@s|vxF)819=G*M$V?i zp`2 zYgu?*I9PXi9T2C^LtSBbFu?ae?%S$n9CQ`PEci-l=dwXPtk*Jh8O3j&uRZU-8FC~Q zd&X`R`jLU+z=yQYaf_L~uXCW)w-nRmVud-+wbQQJ~ZHjPc$WqD1)$He^Q)AeciNsFHOk$ zc_fyHEAJ03GL4?O`VnmAMPK^FlBl2SU3Mgh1&yO=RL?U*s$5D~X05m4lXAQlOG^is zD4h}NJ4K+`2(B*4_0CI1bYVoJ@7v%$3yr+jjz{tDLQOL&<%q7otO;2vLgRWShXqvn z4tErJxXhk7I5UY4Rh@nNGUMzptzQ(1#eCQ=%TwM>ro=tKxsFRwg05c)*=bm@HrVs9N<6}1r=Vu$1&mHt(V`ej(XdwUWbt%YO{juO&mJ8d< z=fG_LhyulZ=4Nd4LkU4gHQ$p`C?=qYH%@Y@52^%s*qyQq>a+=RFR8{w?6$x4y|zrL;^ z+e);uonT^mF&N{`aw|&4hX5m2?s};B{OjTDVUOC;VyH1S2*N71M}t%gCFCoXI*^F| z$_y?Xn62D1SabTg*d~5T9FC$%iOcoFVk9?;gl&BM$Q@TP^E8%7p{I;nT&xxY9W=uw zh+rwH`n*BN*~IM#8)caS7heahtU*X*!g22C8xq=Q1tDcjf!4WLlDk6+{rDYGicdzUv zxnzZl!#f{}lZimpWoy6u~(j_4#a(>MtFYF-(z4SMV0-rN|etT4rg_wVdiKEs& zxa1|?`ZiWs6ss2Gia)sgGv9^+gnBh^ITmX$HBYI=ngd$soH7w>uo>&qn+hFH5@7(C zbEQKPqsxQa@mg6_7M^v)F~rK znysjNq7Ai*1RPc++-7+eYZ-Y?xs#%=g?KeK5=)2O6VE@xy$`r;DCNmKN#TfwlO*M1 zp-gcR7&%8f>GRcRPqVN2dGC(Oli_OlMzOOtEfVT4pATB$6Qy`Y@^E=7{X@l3{Fp{l zEmL2a%gLbiM;S8wYKAQ};zXMt*ruf@!CRTh{fjF?gnE>fP*A*x#wEU^^m|yiCS=!> zE64Gp*DeV^lGwA9b}D4*^hVml%0p4O)%5*-MX@oge#EQ`J)t7`$;ON$1J#e>YAPdV z7nq2^4+0V8&xzeh$V6zVp|%B+tT>&kOLQ3a)Sb34dZ9QTK6*YEEQdzTGUr5He%NGI$Xr*G#YfB1oAe$d@u43H~Mn$s6Q&OR4 zh{|&$kC2S29_Mkhw-0NR8a~u%#jo1U>vvhB(OBL3)hmFUofw-Yj?^sCH!t@cNRm!sX=VfkC>pJi*%9 zoy{@F+|xVA`dBEc<{qx+8dzf}Xy?_OPB4F09l)b~@D6Ig8t6nv9mZ zN#4O?bNuBGHXbc&N+#6}s`1RE*U7jprk*NtD!;o+?YfhR$yRQ*-_^h}h zGl3&+L!=l|$Sa1@25~sIAxLVX&CsZ@D$PsjJL5y&~vvF(g8OA1=5{z#Qs~znn7)aqLsjFONIgUrh(-vv8qK%#@ ztTyX921(@UG3+BY5Vq#ajfT8p|IJfSvCN4}P1{*gCT-TZr;|bYIgK>VG4q<2C@i zGb!v>a$jUd7u={xx_ekJ-fy5{^<3WWc@^gCNan$5dm?8x94~Oa?G57SqEJ9?>V`(F zi~TXl&vL$LcE3re~(&E|(cE;waq>JSl?or~3dt#^yQgwaaVA{>S%giF16(9Da$B84bobWmgp!DG$b$$+=!{0Q3;Z{k$?AbO)JVQ) z&OQ8XPW`>azlkmPK(k#C|egb--GA9as!px10QD~pIhQu8j^x0m=DMeMTK4OMH z*35QxanZ#p0^_)QAi|fAj z$eiC$8H~Ovdj7%K{xl|3Ipj|Q>?S&*e=wwOIwh`^lW5e@8d16a*y3e2ip5|Zt%L)Z z=WC;uxb(Kteg07(Hj;*5&-)%RM;s3s5ERRQwQ5&oDJeM5A(lv)Pv$p#^pO9T`0$ye zWW^9!;e}6Syv-7j{SSZ!93t^-Jfi-mSRNJ?de+gm5F<#2w({ApE_o#&BLAn|kJKlEX0;%rx+pQs>FECv~g9*ZsMmOi|MSk@%+K5 zrx^oBk&T~}#AH8(&V=L9=Yg8}dA?LE*wg-Pfqddho}?PV2k$9sU6B0@M`r(dY;s-_ zYcR~13*MQ#@)_~Rw~~ux-}E$uFno{hHdu+v9dno#bXR`6hN*AfRQq*sX7y0}RnD_y z*wv?|QG!G`%~)0%F2fMgMhAR~H`&oX_)2mepp#w^wJG6sBP{n7!z@U;D?wB+i1nSC zC8(4LNo+8>1?44UqeRAwEJCOb2X#X@2ox%mC7oF^2&;U#y}+Qod>m3R-}Z^HdL@xS zKJOgo+F-Z?&l~!}W!=v-;5NM8v$7_dopZ&T;e*sHPpBZtfJs2fT>H2-~x7eV< zo-aTdRrg?1K2NVa8bNUEsc;9Vr#a|TpI_O!k+>oAimaU+ykReAYbi<6^}zLuT)Uz` zMeGhz57kpQm7EN*@Cl=>a(8-3y&hMAGQn4SpTdZoGr;N$E%n>erW70LHscHE403n1 zCB`m;&S%8S_LFfn^RAZF5fKeWhIiXFEKOt$M?rkZyt5N zH6y^o7fyQCDZKCzfeXFkm96xa`ZsA$ONo(7SxXf(p!`N$$XW+Y(Z;Il$I47zSkmah zG(5YFP- z6$19j5{iGUH_#U9WzU&(SBV;Y<7mma7#DR62-sk}9R?Ho^dMuk>W<)_eIcr%bEJ{` zO}g%Z5;R;PE$S^o9(>=lkLZ<%lAAtD8ZQH+y0 zaTv4b!=_({n^4=8^7l?#j8c&=u?d{5)&+mn!`FiCPjJp;iXo;i2(neU{RtT!hS_8? zg9zlT{rR~YmG4*tn%=k`a*2<`;0tXv{{BVA6+!0aQ$Mb5!da|RDHAy&CuP%qp=OhI zQo@IOd#xlR4ctoiOiO&?tFUM^^ME(#}jg( z9yB=Id8j?w>r)6{P^^FPDQU>`;~!rI9g4{>dehvg&v5)1Ir364YO>7BXG~2SzkW^Y zTIsg>gGVlt;2FIZEwpXP-SH@@gk{|6)#8!QbjxwxxNz`D|_4x z^(|eonL@t!@^KPB!{pONJ8QG1)vwq{3NDW6U$&p)IUSix`V>wR3PBCx2nKv)R9E(z zr=eMlg<^Q~5|KIEVLw|OpVc^rE>4`UQ19a=@$nd{vK|o0@6TNYX5NKv62&d?tUxcB z8oqnPW2jJ@jZzskaO4|EVEAf_2idWV>e!GUl{q}C9yG~$$F!AyyXn=>O}zAx8s zKz97&GK>i(wAPNqSSl!vdqcV<#yP|~iZ1O#9(nhqB`(FE-Miys0&*5V+PnHpLPec# z>k5BU=jC_N0$ zu8~@@EOGPmuD~lI_)!gxfoLX!{DCl1Jed?1*;OU+7EmDfw*{}9Ty&cwAMvsPiOIY* zJ~u?t!9gd>S%P_@TS5Y7X!KOzh+RfxGNbq}h6!_tdh*TG+@6p3&k2f6ZrJBa>^m#J zCnew=XrUu_vA>$ZZdkO0cB*l?j!eJj{zL9K*I~wYvn|v~i?FR37PIR7b=sHz)Uy1p zK`X*ajpHbLON>XPza%&x6+PL3S4|wAH@${)wcF1M^%UCWwO6?7bDQjJZYyb_ zl|OlLACh%~ky7SRP95Kc%5F)E=@x6T>6={N`JRs&cS)L$HWqZ4U1I7~Ku1=8j!QLo zF$t+umS_{`cnn!q3(?v@2qU@H@^>q?*C&aJ7MEuKXcVKlA+`qU9Cz+@j-L-M&a>jn zr!Nmz<-rTJ;guLY>t(;rXT3&PY|al&TfPvtRY}hxipmeML{_XRCv=`%bk2Lw&X0L@ z&Ek)}P-iJX)K>YHjts7WFrX$9XO)jpT6VsMf)5QL_D|?W?klO6Q4jt#+5C$d?#NQT zy7ybtxGmtN%jlwqEwsI)6QX*~(CGhOz&AUZ=ckAWN@n}i@!1NmHfw}faOKk| zNA}x(-^d`W+@bAN&eI$@P{C6Al}q>5@m&{b<8(O+p{D-Y%$O&ZtW&^Ki}}`z7E^h- z=M%E7g=gb`*3}znrTuG=<%W;2Vof@5?)*^NjlptZMPK_HB8Q;;d%Bd#^xWAg0`YNQ z?Cv`>@s%4iJq0cqQ7S%lJas;|O8??btdXiR`>U7*iO5MVv@J8#`8qOVS?==hyQGxI zj!hadqt}piN4>L*u3Umr1cjYz@0FZu9f-ozavPA8!OqE-?Q?KzfoMF)yRXJ=dzYpbo(Vh zT#)Z4tuB8degecD1f_dLimx;b^`;I!hEvjb(%Yz{HFOVDBj(8FPb};?88woKl#=k= zU5GiJZY3r2cfh{`MhN<|R>4@>%W05!D%X zvifseRpOe3s_n#{_b5T&3D*<_F@v?wv>x{>&z=!{mJ}6I3S$bkFpK@kSdlhuT(YX8 zmP6wh0Rs>dyrMYPB1~XcFqHeD4j4OL&?KD9LRLYc8^hblNOF z_>9V!rz9hQ>S41p5|-Os3D*U-BANmg=ELZc<2c=!`{5{6ZSAf60~H;d>rrma#&)Vz z6!>drHqeweu0C{h))9!w2N%1$zF+TcDzOv!6Tq}QR$>RTpYCZ=s%uOOy|GF@TdKX! z%+g^ZL$n#fubAvT<|Di%UX0bt_CYzVT3+PIWB-6_berRP@i>?yXuU?zcpwK z%=^;>8E-(sdR`w=YYRL8KUkzq$}q}yAmn1d?S6h(S_@Gt=rxn0ZfVruj2rCD-%T~fX4&B8Y6eHgTf z#s%Hr)i$1z>er8R!N)QJKk>GljTod*`n+E(Oqtp@R3Zz1Mxz*qvu|krQG|Aw`{ui* zgOHOxOdK}(%sgVRZ|}J*Dvyp*M2h(zS1c!ibm_^^e)AL;e}6DYPTB7^oNKqD6jIc1Th=bHuR32!ePvI$C(fF&S`4nWg%kgy!_mrKXM{cK zWLL{}dN%|s_UFgSIW;ew6S9=j*emi~vB;e*Q6VHY@2#5ghzkr<;UK%3nRUORI%jl@ zO`UK0Z;U~2za=>0A7SGQdmt{%Zj@H#hwFFxL`ZDX{1-n5;PU>$^Zx(!L282>=Sb}4 zLFk+x67hsj{2xs574{U*?!mp&qExrr==|Mzg`?Jy&LrzQ$Ok(^n9%xCzvboR;L;N{ zqlFy$gJ6*Y2S-OoCBhk;!^pCrw+f2$^H2q0yL#xsL9;Ra^eIIj+oAIh*o_F7|IVjL6?~Ssph9V$PR-8= znflh?|EIkx|A(^e+hot$rj$K$Q?ZTo(ZY2Vo-EzP0X-q-`&;+$YjO7w002_^%&Ud>)-k&YTAJ*K;h-Z zmOT8nA_cb{{@!c0(HeocHy!)OGt_ORw@TlIM@)*mFcM`xyZJ<>Cf(9*u{|XXD-xov zOp;>zos3C-$kfkpqms)CZ>K+{7HVOCo;W&c)gm@KtNW)tzY8mNu-6`3p4ZrT=oiKi zum)I{?Cx4#Hdf1lqXc($W4ccTn~8t_FaGJji+w0|$*Bl+Cach`msvIWTdS3wO@zvG zB}~(F?nx;+LJ$n_EsC`P+WOlEdC(+5M?Gn~^8gS}Py>KHC$4k>cm2m!wvDy%S6@D-)pg z`GH3NU3IJSnMyOVdtn4KI7lbxI3NW8wps&a;V)E~Wbs>A;Bbnh;fQ zhk_6%U$~_2KRS+FCXB=ZRfGHc)mKJnGU?cDNCKBn_b3DSXe(#NQ_pXLR$HE&{bwUO z3ypRRCdwrqWW`MkPQ+!&i*%u)4rWP%&Ba$KK8E65m}0APUN#?3m#YjQVt45ES$if; zF~)j#d{AXUOx&KJ;t3!3JTr*WIam6Q$hSqKdkU+0)^v}qJG2qA1I`0Qh|g>L`@BK@ z8$>7++Fks5+^@Z(FLE;r62w5G`9}xmGmYWAY!=3~J7XQcCgOs8?s8Wg`BixA7&+ur zPi0Ai;~DW!j@C--9eUGkD@bdFxfbbgK)RL2WPlj0#GJ;-8pQ z{mQ%QO6Mo?FNPaLd+}PV^~Z+{F=9Q=g~}mF+12scHkxeLJ$=smJXEUG!Z#~iM7iEkCdsq;M-7E~YF5~oTeB#Cw({o=XIP0x=@^#JhDwNw zr+>TPYV`y>FbU?#FqEK9#pup`I^d4}3L5Q*2(TLPOy&sfeIbA*l}>mDvwk!+KKZk^ z_!Oy~PA@wuGx}U}=CS^#!bMeKId8r>qd%rfzlW(HCkv+J{ZENvRRLFJvyWk)xIgsP zZ&j~*>KT^N4X7)1iW+mE3Mqh}pn}uo#~jP|A=9LVJDIPfu#;Yt3k%=l5kZ15_b-lk({SQgU~IebW2tP@m-=7!63-mtKB}q{9}{B-R^Pty8}(do z{!Y_vz!yA>ObU9IB?Hge0h6pPDMqEq2%8HKnXc|`Kgc(poo2TNkpr%(|8K)xEuk24(oM2_O83}5cFN{DAExjp> zxRKcJ=G-LRY#y0C7eB=H$LF^}E%h~*{Cn9he-|l&HHhmuv-O9+*f1X57;CKGJof%V z<^Vff8X>W{5VUY(dr8Dz$d5)Jhnch$CYo}7SGT033633Z_h+@t5NkH7H2`tDHJE}O z5MEB-v|L*vLqh)aLX^mC$6#v3HEMfpAs5judm=p-n#{m7KQKp+hDvg{(D-sB1pUg|Lh1MC7 zq&z0Ez7iT)+QNd5Up?6R;Mys!pub&cV?I1W@FWGz83|w%d-{3hD$V9oWv;}W+Pxg# ze%59ji@MKrD#)}klyLpgxV;4`lhg*?Kzg4mec#y@bK6%~9yynwt-j7$25Ra-Wl^e4 z{+=FqFV@etfG_Ce7ojp22XO_+@#(;vdE7KI&c0Qgyyl1rQ@3vJaX43W7hm##rH7Gu z!O6E@#PM~sVIcA)GqH1v1sgCi*^WyTIA;2e!70{Ji;wW#t=m71g_Q_Dx@j7pbTF_k zPF$;?;@mQ1gMME3zMzW1E`7AQAth?sCz_I$*cnWpFfmZN`N%C9tR1d#72kNQ`8@!kp@&;^Z&$jF&OmYewhej|0HQA9WEXj|>z&&)L@>axZ7z zj-hCnF!4m;bh9Pvf!4Z;!4`8z)(8|0I(RliY=bieUB2B@WfEq`uZV|gw}XjNLGjsvDd#l6_1LUGfjC&ITyX5|y|&M#D10qez&V~or>Mg}ByWXuN6 z0`JXkUkNyBp8n5u`Iks~OO?TRznTdp_S_9Tn^<{IybSk7G9Pn;96|*@|Eitz2$SnZ zY=brBmAAck6GmNAyh#zQIKN!Xj`aiu2=<62B|VT*x$di_IWYtVWV1G29sVqIldnH* z)r)Q(P}jMfxuYH{muBjQc{zaK1$A$cU)tbEPr6%y{4Gm$k)Th=S{@-ZqFHi>QJCcrr zC7ajj^3jJr&t9Y1mG{}3GgYFC8yt5tktjnefUGTluq995-Kz6q{83}WkzOgr!Zbe8xW|Lb$OJN8kSjV&c?X86yDVz(bG5;zuf8Ae+fe=m0TsPsxCQ;WI?^^VPP)-dJ_ z`OI38JlOb2-ahv{Uviq*tdRIrK|l`TOdo3tf*LR{W^7tu)@yYrRrS`^s~jId+{Z{H z%$R95^IT75D*A=(pv5N6TJ@eSg0XBeuk|0ZPl7uKdpsjMqTOgf?jloZ=Wnq;)P!*G zj?(5j|C&obF0IHZ+*(?tOCGN#;SRC!L z%#|j>)CZ*}5IASjmX2oC4_CafO-0R5DEp!>r{+-}?Dv3bd!-rr?8(0TK0^1DcAxmGjzi1PzdiIjVN1V zYbOw}>{b#bts#h$aYsimhseRSNh*QkKB5ehG;aFN@bx$`jA=q63tu0Df)X|MTyhIc z-goFYUgwX{#2W&C)Z!g>OqFxtkOu%#V9sE%TAZ9L^l)wXqZave3Jd}0*z^+&vGYeJ z7UoJ7N;By!dpgf#N6>XmoCka00=NZRnB(aMrt4{U$#fpn3FacuiyznQivaoW1@WRm^_6d^!rlWu=XWyARAhc!1aPgVjw2m)ZL^jV^?}{YRg6~|lN%x} z=tq+{x51S4gIZn4)itf7>r{HDb{ce4p()WRs5*DW(@7SGge^>RNNo6Zc>P1x^y3bc?G?;S8Sixi?0cNEbwS8Egp+ndLvjz|Cl(u%sC*B z*>nG;hTd=UgV(s8z2Btf?vR`TSm^mJ-Xwn)ez*@Gs!pq~a;mUS5oN(JJmCS0wRh!) z!2ntEeP)trnViFivV@1)J{ubio5)2s{Kt&%%iHwwLJCDb#Nwx@Th1=n@S?#WOyKm} z{UYlazpL-+Wh`FfXDDl2;ZDb!_vM*$5TT(3tzw##o05`R`eeib6kGL zjH(e;32&7zwHJ3=+&I=ZA__i1zUy5)t!9KgQ*f#8Q5*c=bk#k-O36Q}{OI`=uPZF5 zREzgCL-Q?4%gI~|+VV+DCv5|VWDV#DNjm87K!&+9-{5tR0O0om1dt)`aZdnVuo@@A&!e*u(0<#iY=7}_m6_4H`N^| zX^wkSdvIqNg@woWih^=Hia7Z5Fo(;MWwGt`o&^>dw2Awh@bmGZk4btO5`3DLNcQ5d z7WbyBe#4vDrbtJ;Lh=T-GhS;9^b1z{^+eQEhTGjZ7);HgZz$k(Q2c_p~9We zw|*DCR|sbDBlpDXfuP;AAq1#irA}o(kn#oafT9C%%01~)ON43%D%YjshRwqxp(DHA z9l17_0S`BNq7Aht_bs*A6c{Exbcaw;EwgY5O@!M(u8%edbgopmmR@EUP~@`9)iwO; zvR!xA!Nr$q3(A7_g@TuKRhF8fuFy+8wu0&jVe7>xF%_;W zb{_Mg{_0C#Pa5lbKK0BmSbEYPKDc*JG#Mc>$Y>#!3ZucYlT4H!nD~Yo$h191Kr0>OR zTdQ+(H0b`liyeNE;@VA$N!;*w-e(8PF};)@L^|KjgWfLfokY)w2pg@)0^tFPXNZ#( zPclS&y7}q!tD(u6A~^lGy8Q+V z{+$N>f9pT%|F@PoNwyA&hFnKjm{0T`o~v3&)u{=|%ErleM&*Csiy7w;+_(Z7i5<^4*Y%}(;8;s?>X6kdNZsI=9lDRzcwRhKGkoM@OZkq{M@&qN1X7baYBfOHal( z{1~3gC5<*WH&^r)EG{iF78rALb9HrfuG+8cMT|;Ab zW@fu@xt^Y$tgP(hEF-J^>(_c%jIN$DcyNM2rE!poQ46W{ORDF_7O&Ye5v=H`r=>|@7{g;l!UyLazSsnvH% zfBX0LCr_SCFvk4+{Nld_)lokefj^~Xt-}pNvjdTlkvT1+1ql(ZMR0k~u;k?AH%-qz zQ9g!+g;fQVVooM5bhmFp;%0KfUo}LE)y9mjTeo-JIUQ}&JL|_*hJ$w26s7aoeK0lV z%DenPWY0NwAMgcz#%0}SZtUevyI*ThM=4lnS(F^jum2{xwnpiG_^t5GetB%dm7_{Y zBZS(c{9t#&o*Kqy$o_UZH!e=zUAJlPy7gO6uG_Hvf8YL34*yev{~5#oJi-4w;{WsE z;IP=ept8S2FL(a$T8T?nCp2`ft}I23MJDsvIhRLCY9$|Qd?=~BU#E9OZDViY>X%cJ z0m_XJlL-x!ha!{#=ZYu?YvEg!;qIIF*rc>;GRoJ5$+vSW_a%sI-s3vAApXqAuNAYE zyDcyQxn|_0>6#W@VT5Y@Tr@mCVLMR~+dBT1IhSVfmE8)5ykMJIHT^g`BxTt$cTdAhdBMtZN5 z$2bpNV;Db0Bd(IgYIKkx$I(NoT&cEOL-{fGCl>=Bv}U=LKWtTgx<-|s_i-kEIf0#(!_uE*KxaatOB3P_zepPF87 z^O+>tgOn4E;n|=^yj!W$5QfEUsv_;=6cB2pAWYn+V++yOBC)bN^bDg4Be1eNW289u zLQh{lZ^?HGIbv|_6tWNeb|L!SCR{qKyplrGGYsAcH#s2V$bR|rA=N1KQrz>WXf2$` z?zJ1e8+`6zg?G#}UfL*o4~ucyfRWwt7`#nCVbfFezC(J%=dT`PLpe zd?J~u*rZhGuwSxY+dg_7OsTSpy-a6LT)N(K7ibiEA6P)Gfq-PhQ}nhi*P&KmL!0|IEQg!?8`|X_d;LX!J_ulQksnFmM3QZ^O!nuK#6;MdVMe_yybrDjm>xhkk0=9w#MhIHO_7a0-+uRB0X$oazJwt2G!sC zc>77@=DkurMyct|Ee~2NNv6jO?~3aIeK~l3*)al(ap^%m+LUGpRA;-8fQB^8gt9{b z7M%-D0cX1xoNvtGsRC7)N$=IHbiE552ky3VyJ{n=H{Hy#J!w-;Qg&T023JZ2nuF&Q z5R(F*`}9*46!eIBqnp2TX$75fh$Rm*&MdiL{_s*6qs0K>%u0Aj0 zsUl49DEC#4YOk7c<3w&%jz%wP@1#s9H9oZF-Ln-a1J8~7B&o|!(Lq6iLyZr&a|rYv z|7g2c+9>gG<3oSJJ+8F#0Xqx?&1%KR8u(HUPw_yTwsI|HE`0iwObZU?b)O6arp4o@ zZC}27gPx(bHuB1To7aO1xAhFUs)@oItBGh1OwohZkFAD4$C*GCrCp$N)K|BGmy5i$LD9__(KuV?HKPbF%Z-9;LzxH~t5_drvdv?|29kF!uYGB`s*;V`}GVB4m?Fa3gN(f zy=-`frA7BzcXFn4aO@~V>vPruGfIEH7pfR_5-GPw5d3y}%Dtby;Cm!i_yTyHw2TON z-Kqb4-CpP!u$8C4e)TUL11_U#6UIJl6U?OP-CihAZ608KCxP;Sg|T$#X3#pAiQZFm zoZepWGcK3_+@y)645_RiZ~nh6hV@C7HUCZOHrg!LvuGB8Y*A7XtHY>wp#4fnAQ`d+K4vVxwWr6JpVe@UFB{4)O_V)mZdhJOFOHW2TD zKgA2I`W9=WFxHPZg4L#Nx(R%l$@zcSGg~qjj&0ow4gLqg{WP#wEGPfy4^8{GJsU=` zBvZQ?oZ{2pp5iAGH&{;Wju>FVhDZKk&uqzP|JVx@b;mylj_v*LzX#4#R_Y)247346 zh5z3uOOYjUWfam)zT@hbeOp&VjMgzFq|%L0JIaLY!==0ZI{y!WpFJK3>hIy-o8RPK zFBWMcP1CGg-tyA~bbp$Fjv#!dZZtW^G%~vaVk~&(hYf*H1f>URVbb6$btsv3|c3sR#^EI7&jrE?M4z3eItG&C)Cjbm|;jQ`~#{3s;J6ebiCB!Ny z)FnXVG2W&5P|YIoD?YY$&L=XwGN{2z>*3-V_hqZbif94Y)!p+JwhH~CGOG2$f;|SPpC@r3jkUizFMWe(xIp?`4@+^Et zGVnmE?Iyh{ey~EGYQOb!u)u>)wq`}4^-uiG%xCGyL$UGr0-XAZ_W{ElqxxO})HK?g zmJGh+`85b?0q^6rH2bEpxg|Dk`8-b~b#x@SCC6E`$YZ>zAeO#7*|T@7Cyhp|kc>2# zB3J0UKm%p?fZM(QI;2&UwocAiLE$M_!QY+O5SipR9NXLOsxa^*VRexQnjy2f%C+pK ze_>VoW}&Y98v0stdx20DBVZ!d8REbU(@it;>nC9Y?_j(y<3DsL&r^DkII?17Xsalf zeS;HlCPRu_6$HoyVQ2S3W8?qym{Rz#TP0OyH{XJ75?o>QsTZ&3w(knr@fUZoZ2#J4!^9!2`-&h$X&5{3eqysqWW*4`I2n##7URA-xzTF#}} zQju0}6SEU%k=vxwr0=$xivQM840V!9#yq<89p9#wDHEOGy!r5T9X zUM6@k$yrH8a?nz&H!+cB+4}^ZO4}|4O{E>)qVVfHTYc+oVtle)3trt7ftz@Tf=D5S z=M-tVk8FPim){0AV$0TxOH~Nj0GbJ<`$y;oId=@GJNLz%GK2$n0{;V zwfE0bqxB63k#v+}aZ%YP((Mi8QH#R6IPR0+L@$o2lHL&vQqvzM3oqKV%xl)=@hz!$ z9vdttJ-({o@%&h-Z6E*H69}_q=tyjXLwqB)n4>B>B8s)I>GDN5?%AD`CMK9L$T zsaSM9NqsQFEBhoef?8OhvLCuDkV_+jGvkwvypHj%^LAG&DKjs>$ z5Cu^vm}AFWl8NGI5lFIKeg#|*d&3AdNWm(e96s{((b_t$*SqF;KMW%SCcJV4Rj;_W z7sC78#;ta?N#%^19HdslX7DYaB3G*Jq}tw(M4np*Hwn9{ef7lx$-TEV zp~<2?l*@)?Z)u^gKgX9a-cF@amtF|Xo|nK_`CwF+c1uH7s59-<*H6)&_tpD1EFIng zvM`%_tc>SsmhIl$8#Y(AUSe#64r~-Ypu2Eqnq{%!H*l$HJCeKCa-_%x1xw7 zZk~l7kQYVo-7by0Qr}{S8F!TZ$P|z1vvVK2W;!O*@U6>uzt6?!iC8q%u^|Zam3wyI z1>`-DLSgRyN3IJJvHclX^Ss`umitW^vk7ObkybL+Wx^7x?|lo)7CG0NotI?n7duKX zsJ@N{b#vFth3s}cV$9yT81O)lT>mBy9q@u8)M|?jRy7bwFm-L+Y^@j?NQskLhUaW>x7);Dt= zH*==idRuq}rbg97^{&T^JA!prN&MDTm*HqVkKyiLMI|VY@3lw7a+v35YMQl-eyZqg z_9{ZF8o%mvtvj>E0F+fG@Q!mtSvSD z;n>DfOH=pCf?0S?4t!)r0FJOf&wXX452`>2Yt3q*{ZIfln-cQ7Lr<}We~=+~?svh? zyF&|msn@8mB{NbrG7jQ2ZsLG#ZZIz^GHWyT8OG10tolHkWX{VZpF~>uj!R^e$%eP* zlt$vIqj|wpU8Bz}8fSozUD^r79p6#sBsaLGro$TkRxkJ$qp3r+Ep{G0@YNpDwUX?j z&I|6)(xAml?+bF`W*(N`ZE&E^#b%2Piw7ky=VNzRc(tc)EH&2oy|~Abf&PV!k71# zFxX)sQi(Zn;$k?$ynad5)NyxAof|Hzq$r@jDk|C9ilHLlHWAj`Vn9r=exjbzY{^2F{B$IM#fo3Ka+1-``st>%`<<&P;# z{WDxXxs$LDQ@0jUogT^vHA*V|tapD~S*D=CIU6kow@23ybwiPLyAI+&prY~Td0tL( zcMcRtUOl{Dy1*(~CdKf?^cBzUBd`ykgg%LR%qo_){zFb{7}&?9bt%3mk=}V?dHM#( z7ecQ?w^+rbD}}U5Np&Vq z_>DAFJ|)+)^`*!4XMHJbNt|xAJB09EY^rmxj&l$k|88?O5N0TD9k;~~HC`9pZym7K zHiRSu!sov`03CmH9IP}<%-0C|adZbpcxyjWj`XWmHMb;=44me%e_bJpN1r;)xgMm~ zptuBh1DA#!I0yZgKQgWm+{T0a1|#gMNsnbf?yx2FihMqnp??VeVP4zN<9Q)E@A>3` zgP^SA`6DYhKl@UTI7tOXJA_pVvl1NGhhvh`)58VVURP9-##CtGp2t$t`+|Y>m7ijP zg^yX-PR}2e?eD6*=YkLZEPjthkW2+0MF|g8jQh8_K$)qfw+2rk<*xm|UNs&bL?&IV z!yht^QIXq|!~u8za=&9MNNMl>R^i`Kk%v|&YfF&4tkLPURpy#e6+m9vKM5@^0N8== zcdqq;EPQ5dwY_YO88~;-VU-fN#t2-aNLRj=ae;;^ynVs~lzywZoNBmqcjd$dCPN$~ z#rUDy0QLB92u1m0tU0Hl^lIySoWTTgDfFeGB@q_}nVgD3ZxjEWX>f|*XkjwSzV6WY z1JIxk;LKzG8|ol{qg}X#=Nd9VUa6s8K%U6;ewK?C>$e)K0HCYZZvlkOmPCJ2@a+|5 znXgZbS?@`gc0#F)hUv3HH<7)2Y`)3{#MfEA=sjH+A>hy z9Dz)fCEkSObbj{h6x~kJb2;j`9~>Of{_x)g?(%0?^Zenp%$taKyA(op{R8X|=oE*h zCd}mX>E9(#Hs465ACATpqJ>{UYGtF)k9Pit=GvS@(wV&-c@DaVK6*FJPp1jB2Ut;r zv<43Ba>vmQJfYu^up}D2qeldRx=8mw^m2beYsqp;LhMMcCpF~Iz^aG)SicH%DKl^} zIO$2umO%B=UA&9Eq$NecGBMQm%u!&WuDZa<8rrc2= zr3B3o<2h)F)j&-w8yVOi1bnP{8m;G>iuO*kDx%1x54iUHYH6+Dv&ph)1F;MI&uV*m zNQJ)Ui|UtB2S;))HGDihf;?2Jjwl_Pbxv?{Xl|am1dwwaZK3BKiH)$ZVm`a+q93Nz zf`GMlUQukBa}?fElo})wx4H3i^LjlS7U$>jJeIfM4mRw_+%={ju%e_3eih$ zAOz^t6Ob6ZR{%X86(7-67|>EzoQ);el$CHs!X9;#KIpz~`s~`;@{S~wY`SBLIER7+ z5T<^_>Z1bu^?dKZXK2Gwp}1T=1X0qn6rZNy@toIywK>~ReyiEjzs!D<9P3j}H6yM{~ zZWIP}A75S0ERI$*y{q8hs!+FB4RZAYVR#mzm9w^NfI$`<3Ps={>`jxm_}BpW-gM|+ zg57rzApB$itcJaWwIrY;SDr@%#R^(7_Dp!(eu2#@=8=ze6Lnhs>d5@UOnd_oo$}4; z+D$j#885*!L>mtv45gtpS}E9JDzyz-5H0kk-7kS4widygj~+Vg3RQbJ=A?Q6;5#}% z?QK|nDLL>yHejrC7UNwWRv2sTclv&-LdiqR9})?IskDb!)yWapH9LB8``08K-mJD( zDE2nVK&WKRTdAwN(}0-!A#AzZ4O-@*mVGD9Tx0oDV|K5Lu_OJha+8f4-)uxmu?TN? ztLe^PDhhfS|JvbWSc~;mYuOrYeWWR`9RyTfrV6Zu!MP;)y_=+ z*Gb-pL8rZ2ZdTEO%?w;;?B4UJAM%Ph7v;pgE${_XZ+RyUhEJ5hr6u_JPrmM)!Xig? za5T76Ai|a5JoBN+WYIz_^GzVz=ETmRR73?a7#Wu1J>2BJA9`B92RTyA%4mes(p9A3 zR!bpz$m+A0_jDA-_cjX>LTPtde_uVzT%5KmbM`Lua5yq4FRg()W(amNI^ZOy2v9hf z+Ko_SkVh4w#WvsR<$Ka0ci?aJc{ z3*Rk5lOHu2Btc$TyNEc&r8GKx(z$|LM3m)}7Ej$6O2{UVv3W#yZo zb95sXeWvInzWpH`r=3&xbto5c^Cr3Mnbk>1#*0v~9DKWG7&HFE{c6$u^d7lx(yU6l zy87~3qmW?@67L_cxwtx~vYN77OUj$wO>SfB4tcZp>&S>!Q+BBe{HW_+bjLf*)^#W| z*;7C~dF}WKtQ~k${YgeR5;<_=3S?WL#)^LnssyXL^ld!ekbUT^Rna<4CG@X%+g4fc+5dAfUzcsXO?9fkMQ>417%~;0a8&fzcu%A zyVr&Hg;1cwD}4S5ngMcJ6ZNnUnr2FR#B6z$5yMA4Cos?1ej{w0eq07-;(Gpsh~r3E zD*wGA_8I5^%m%TZC98O})6Lgh$uGmZY2+)EBKq8QxAiMG@Y9VwdA?JbnqL6sGRVgx z;ap2z=9pD1JMTB$$P9P1G%S#@^VWS+Xq~Z(tw9<4tR{niZhpFAoc6>&UU~oaf=2f~ zWcBQzllFQJob5MGE`QSJ&mB>tD)Kf9MHWdKnr$ojm8993hekEAl2+znL)k){Bj!NcR{L(Ry$f8-Ts0BV9Ik-nj1Qui!U?t!3>(MF?t6*l*NYL%a= z3zTF4G~hv)Q;H?reYi>_V-vKj1WCbDp6U@>==N@>_i8!lM%UMMT8@PIq7X>4rPg9J zyiZyJ12F4el1Z8XTG?g*p{SL|Qm}l{R)ptq(B`-Qg){yvY$mEtJtp~~=0-@P%xlrb zm>%R=oZyLXNKVG*9u;X0c>aJ1UJA$Lso)w)^D9yHTc{g;7D4@9JKIg25?Jcoddb&t z7#XK~d3xGRI;>_a^BT)a#^g$hdkXaoIhsy2*q@@lxb?@L1V59OzpT7-7HcXf!tQ`8 zJp)UtF6LOA3vV9EDIN^NE|vzKSX$WY7sE&9x_e1$(?L}bYQl?2^XW3qAj{Id5Y2{>FjWOv%VMYl5FJ z@=4t7g5JtB5r{Q+=q9Iv&P|Q_+IPOL;#SgSxu4jBO+ei0d$Pkqz7#gcIp zVpV!5<0VruuSVB~0ZUI3C8)H~@#H{-8&|65u9koeOof3Jr3SUx^A@A z?dKmGA3}-YiGY2GNj`_nDn_KVs`n~3+*mCN#3p~l-7Wl{mF0I#zb6&`NW0p&54m)} zFIqfR-H|+a-rA|ai3I}ORso*=7Vz{+)(sfJ!e?hU?FJ@lNqqjo9ASuBPxig+Ah9!8 z>9|7~?b60TSFP#cmruWs&CBn>s^(#bALrFMLhW1{c{aBCjf&L!sFB*DGE|KSgb{g( z3AeYmW6%(Cx{M+@<1(WGx^X2j8g0QGNhS`Sf2R;0ZB^{FVoCWtct!m|JbJ<%TBg7- zKL2gH+|N3Oovlf#R<`T|Rr%#vlJt<@b%QULl-|NI zTY=M*+sU^imU<{>QQD#%CTg(iQ+S>@y`OZ2I{Cu#8 zguhq+vzX7AIZr#DB=u=JH453n*Dg*jyx#TQSF-Y<}Ha~Lb|sCu-LCta{+R~ z%*+<6%>n-BVD2sz*86gJvM)14K!#a!a~tYLf8&DY-__Esxk1Zm73N>*Aa5}d7ckxZz1C3LMlLOP+E7+4r|4dQL5=}*W&z<% z&X?)pn3bfeZP`-F?HbuK-MD8^?h&O{qf5*AQF-EUhp;psf}}BD#5KBlKxh>H_itrf3_dcWpwZA)l}K*Y*C#+O15mZP1E=qkZnLx;><3qNl+RwbaU z?+^wy+b`54v`++RPWMsg8tUv+-ptOvpQ6DfBa!dMToA-OSdtTBZE^ZFm#$U3q58|^ z^+(k`g`n)nC=^9j1VlZ#XTkunTaOVw$M(6`k7t3HbH(R2{Nnjk!E@cXZjQBn;5kGgx_g>jCZd)gHRk=YPO{R-mB9qW^iVWn-^bDwOVVR zesNKZG4_6kNF()JzUsVdU@m^O5H{ViI=Wi5cXfpTpF5?%U}|LEcIn()R`@hhoxT~~ zSt`2v^6kXreo=WGwx|J8`9hN(u-&-bvx@D(J)a9L03eG4V%LB<=CU4f0MvnnfQ1JT zioDbDi5$5wddQ$18M-j$`M3jX2|v*lt-F_fb| zhsqqszrARYm$$2b>B<#c{6@oSak2@oW_85HA_tG+i6}ESU;H{TY(>r&Tb-LevhZ2# zv~fv<;!TIHMw5dMq26C?j3iVf7F~EHfEe_Vap6dDVL~eUdNR)4r=+B3oMNVtA8@55 zX(Ad;!lRugqCboBgf3JMUNPolAGYylt1m5ptCXLmA^l|gy$e@Zx#IN;ZpCO7iQ>=% z&r;940C>DeGso-m;olch3bFi2_3KM$`D!LJo^#WYHoHWnr!t{IC66tNEIc6|dj|4LyKQ#c_!(@8{U$~WPEH9e^jx^8p@;-QOU7T+o<}?8w0z6H{2FUtRFTii zRac5jL&{6R)c9+Q`^WeZhK%LVQF>(jwChX)q(#yLp~ALfG1^c3rLN1}jQENloG#1B zZ+|j|OAqLF4)8sojFBrbFCR9lCdyt=w0?&caQz>gbCu2k@S>uL$y7}#7(nD z_10FX9Q$wN7um51R0qH}6^)Rm?To~Zet$D4hk|TtZ_K1c%o~2E<D$vm~=m_$Ky{B={>!pfGQYm_txh?Wm( zd?9Sg_c0|?c!-J(>MIWyiL|m()10!ek?`oq7G&j=_HXjz=XH^C9Y1`HDBJh&mjf;z zU{RtZCo~!p-<|L$YY-Nzc`jNmN?GA%v*q(>ZFM4aSI2?1+^MuF#sZ=cJ$&%fi=Zv- znemZoiL`>_VGN5q6$yn3O;NQ(j~stNtC>f+CHKWa)~t_u`K0eaGF+zqMpmMHNlR&N z=0RPHXcy0%Li^^`qlLbnon8+lRe8qFMwo3py&^RjN*FV5k5SBRgr|jEt877~H9Zt< zV!PD&L|-ls&D8Y!H(#^xfIUACB4P|jlWgB!@V%+0FyVq^d(Q37SFLF0`;SfMAID`Z z=e1?G$s)6I6p0+5`1vU3CAfP&}BCPj3B*@G`3Qa zA)Qr(9_I3IQ)Cs!{%8G~>VsQND?!@H;@7b|1h;OWpew17&_Hu$OmPW9$V|YVs6@JQ z#_5C+_4qjRD`&-|Oq#>~bs{?;-r0qOyy4&P41VrC-R$LdiKVD zIKt_bFHFOue1zPjtk&U>~?Sz&a@)I-A@H_{&fv2| zL$@kyi3vHqYhs53sz+0Qom+new&P7?^Qln#)5x|EkJ)qAm)-UFGQbgb!m~G8yyfZcuXStY_fc33<4C zS#kfjJI3-!bra*_^q5Nr`J*Mc2Ss*OnH2?GoX+&sT{#sn7r@Ag?o{z;)1~hSz_>(~ zG>t73S{HzFE}aL1x$IHrJNPDnmHwi5o09@|VK9>xEY|j;5Ox9ipyDZZ_@fn>%8v;K3O4glezQs_s+-_9{46@> z&ET{evu5}RtmSK6!ZTTV%7RUSh1(%Ugkx&D!-LKIem$Fa-bt7XmA42j7^%3tVqWYA za&bpMyCljrxIl}S35rX(U^TBm#sv7R0J*=G+yeO_QPfTnjUagt23)4|3QKey@WUG2 zb+eJ#*$s94aqv33GreYSyP8O|qxL5jGwf+b+s!sgYH^`LJZc$rE;1^6W*%D{3>dKN zt(tOVfmuO$43o5b0&shha1|kiuRq@m71>mgBc_|v|bfwJ5f|{QW`ji%fA|FfbpB~ z$eqrdC_@C-NU5@ojVbSF; zh%FFGV#!GNk<|@&k<7&~t+p2^+*}wrtaayDfPtmqRP}FYs37JB!g3!b2PF?&AFBR*Wfyxi93cC1fd=R)Sj3U!A>N zMdE(v3ibVl4_4a+P^CZH3&b&{MtP{$%#5?saV59q>$=u6=P&Ed%;-1n($ycn!yxwz z+`!Qwc3t<;Mz8>Po3i@VAtEtW(x`w}-2ZYu{bO=szr>rt%>0OW^qDE#?XPXS1|Kig zFcHM*HZnmm(zpy4dF2e!jGMVC1UIP!UE%*=3BPD3&1$Cl78Y|5X9``0tJ2;Ikna#B zI@#P{y>u^b>Y^6BOP%0$YvrNZq;llv z1;RFrDo^G7JlL*+E40%flZ-l$nttsBf!T)Dh*jQy7eFihfM{O=nEDsH74=M)6y!1H ziAQ456~}a%RJ&Sl@?d1~#{|_zRF{_~^O{*IZywHR3xzzDq}A(Lt4@m#cb?rPuwegRwsZDZ>##Sm8UpNTc}Uz7EcyjMezt0$9*?1up#> z$pkdg8vzyu<`e+Hcl;Mo7!V_VmJa|XHlgs;;b_CiunW+<(VyBehJMa;!^@`J= zefcOUI4QcALv|0K-LRO=(?w_t3!6T?MJ2hMWA+H3u6*NAfv(URwqzt+O+Ap1CQ7jh zNjqt$+QU7OyASFsj`hCQwa6|cl?EoQytzQY5U040kdgp{_hbSa4OE`b$G!+)WaH7J z)cXdj$*1kzi?bp}vGZ$#fnMIV)wy5{Cka7ZTMS&aGr4_*jkYM!bV_rkBhmu_ z=5-Qr3A|kPTzge|7_R{Q1NQmE{eBQbD}5R;*%|=amU7sInHdu-Ny>sbE}YMO<*@oI=F*{Rv6y=HF^M8GMG0n^!JYV^S`ZR zZ&4{g?d_jL5jNPunwaueVZr-)#KXF25JZ(XdeiF675@IN^WP>kx%1m8&~ZHa*>^l9Ce&`K7DEnFsQx(D=d z+2bAnB?X+oKg3x1ur=oU7{f`M>1)+|G+^Q5vl`NQQA5gZ(9pOJZ7F$WB?*WoVDg3C z!Lxl83MRW+Q7l5?9@vK+34nU4A@XRdA|sX?XG?CHeZq$R)9tlOjezBd|7W)XhwgLB5%Bj+c?c$C89Bu&^c(f;Iv>e`u zpzu`xq!Ix_v{T>jI3TPcx&Pb94v_-|WvsXicjQS5Z9a8E7f@?{wf|LWqG;ao{+C|W zTsBOL3g!i~YJVW*Lj=Vrkcl5PGZD`Hf}>HByR*PZhP+B!x%AR7?R$;$(kC#F0hv{h z1+BZunih-RrxI4BF(sFGQX6?|kBia;7;(IO8Sd^vOW3;&+edpcB7n$od}181Etzp%_F&Es zS5!>U%9fmJa$W5pX1@kJFxw}N{_6BNb8Ph8Xpu;Z{H~?UPfa1MmM~MLIJT>M%z1US z74(@N(mPpfFmRMduh{?^d~zhR44Zc7PGqZ}Il*Dnba^@V{VXhmHu=dq@iD|&*h#Fd zT8}8A5(>%j^H(JVzTbvRKmOw>x@9f<)qDY*-JR~_8p5do7Vo|@KfBW`)z5v4qrp7K zm%Ab%hs26pOJh(L@D;lys^Og&*eotd1MI}(;HEj{+jCV6lDspn%y(+$NdwLlKQnr| z13-x?#*ykH?cY~$$p=oglwI!#E1tcyzK{nn>&(FTgqBC}1@`)!O%Q2wHY|dRKIwH)AQX~bL1PGt^WkMWiH0=uZ zMXw4`#O2y+RmlD;)75p@v{ipUTaS3L5sin^8M!otCpr2TTxx3I~j4zOgn<;$scS-De#bm|$QVYE@n9 zsTsob1NgnNR9|oYqCTmpni@Ria37MlWM|jf3+A9_S58*U9#`raawWdU|7*hMxLKHc z*X@)JS6oq|S>L*PZA|EBP1bm~-+ipC)Ui~W>1UG`i$m--r=qT>iWn~_r+&4_LHv%n zd;!jC_G6yJs9TwM(_6BcJd&^$-}u2tL3B!uC;wD9Ws$j*e{DE~vh;TUu0yXz6O;+F znzY_t11%;aQrTu<3H)x7zxAZy!v%^tWmy}Vpg9Yw%6Z# zI@kAsK|6KcGiT;3F4|Z2&h-lmMcTTxiRD^por{I9r=swSOE|ZCN|F`-lLx(hcf#zcAJ=^d|N%NFP6Al5Il?N zT+fsUbq2NBM;xvOybG}RFLlc=!FnGe*d)|?GW0~ieA5%hd9h+)xXERjY6#?h1wrv6d3mihQ^E<(VkeI{Il=>gJQj)xvc$uA7*pjktUAW0Y?IUfMY5u z>WP{bP&B`PDVkcKBV!&5`ZmeG6b+2Z9LmHl2EkqH?Mm{c&~K>tK#fu~!jjeV@B7@a z(>Q%13pD+WGjbxuvt*?3cQvQ;4OR;~WxhutDoiX5T|w3iI^^nLgz)*_BLgTbfVI?x z_Zoe11WkXZlCzeK6FE{~YDBNyV1eqF$o1&-R1jsZW1-?!hPkC-pz&PGiBC5U`d{j~ z@rVAr92+fGVE}R{-GlN6AK(2Vf|&Bw>0AGqzFKnobzEu=6W$E)KINHrZ-nk8%kZ92 zmA~b0hZzP`$m_Kg6>HLjTk57)ipCP#XHWPq^(~yhr5A!x7l2lag9+F3|2`W@E60Og z>JZYCpFCU?8`}_gKBNCm`DlRq2y*ERREvA81#gla&|arzU^sz{(A@cDiuz1qj}J;e z_muUntL&8Q(s37PO<>KB5n$7?TmG_iK-1HmYK2(yktl7_4jU}uMpx5JkxMUh+>RQA zDt%LgJQ4|L=c>?-4J8;#XvKymSRt3}$LDU+9L7bMo7g%+lwG@Uq2EH0ZN%Bt9w;%8 z#NYX4#P7juZy1Lcw9eS2_lTl_6prPF};Q0+rR9}*vF(h(QgH@GJXO^tDftaIQacF5k$Oy}m|ySNiYrfbEC z04|67ut8^ggXp~UqCHbAKac}%avM~H=Brm4rEg&Y2MqL>chw{AhysB55sTLS6Uw@_ zk2U0}$hh-5i#M(80IuqBSGs2oqmN^DT0F|RITUrYjMv$}mk~=p&+Gfexy;n*`SOOP z*KS?6_WML{Boydd@7fNshe9;V9wYoR=m6N7oW?I}g2Lo`JSd$=Iz6=d`9Wm!4T?lQp_a0vIz!xp`bp!rp<8t;h)lj^Vn z%3ZANvH!NSw>aC2qSK=|>Bxl05&Tthnb|p_+eM;exu(L(!RZzs{KvII)S>8_Hsdg_ zab4um`m?=kulEm^HE@k0PdpR~m{ysQm9PkvNZi?W0ml@|(6DrqW z|8P6P;UhfVtWRY<14!0ip_yN|PY8K$#vji&Iclcl_Oa219kchzxA1NWatrw>ncm=U8kGchJgdrw5i0T{AkmZdJZmVe%PzJ`nVB zQch8)rEYvCMRvA49X~(8w*vSnehPoT?c$G{AKo$3{!$NbxSQ-0;lC`U$u5Vptay1> z@A$}zPFG;#k&pm2`a7+5uXD-VadC0T=G&%J9y*vHJ4CstHa$2PaK$*X#mWC8|LTwd zyuisSxkovguCjW;AQEntxf>vSU;Z{>_v>0AaIwVI0t2EyMQ>MJ-z4qHw$F2Wcj-P# z-tbC^nt)SwEmGW)(?*%~&eUrlS!0vZ)LMCqygJ4wWys2z(^j~wt}IJpb<~8D@Auie zA6j)qdLw)HynpXDzM`ApQRWyi&`r1$V3nG_)%+=%2R`%QO8fP%TsWrWM?}8?c zS>>=XX_CCPbs(c}P^1ij@fd86of`6*>Fy`0WQ;~GLFezhbTt6m3slE={eW-yiQI2RO^-4X-TA@RWGybYwcEufF8;k&Fhb8SiFim}Q z5V_i0Jh<@;})qC{?#c80le+o+b z>gyL2N)=*l9ES$*l>S1>zF}2Jv7=Qy!G6;;o6WE9slyl{Ow^qgH?!+6t#MyPIGtY3 zagA}S$a^J(dWx1tNGhK^Dqf_%6kHpZU5Z+^AF(LiQDQ-9^^1GDo3PN=>D-OIxY7+2 zP~9z?rt4tkY4H?YeoY(6i!dC#r5PB|<1Gjlhk3c1$wLZ9#2F#a=Dx`0|GEX-%L8)D zr#(n1%sI~opXhPu?p8n1+uCcN+e;6tC+@$D64?WrkuDe)nn0aBC*18;Sr#~Y*5__r z+O<}{{Sz7y5`8ZSHD4QrV$+7ba=upiFjl^(?0fsYNz!`Su*f?-;5$Qe1~xOp7{BCP z$((muoNMQO=KzpDOK^YJakCjdig&gvm;-P63Zsc&&Gm_cu^(RKC2Z{iy@$p#c3(LX z*+ff>3xL?hkYH$_F4wxbVeHI6?dpv>JmrF1txW@aX26-AjGRVw`De$(LXaADJ@*+} z&LO>M=)@6%^T+85_^O7{>|+1;UMRWFdbIS3WPJZQ$}pAF-~()Cnh~m{z|$omx&wZs z34mrwsh3=!D`_)>R`z{JqXqxbTLm`Oq$bXMwcD~#{|$w8LqGOJDc7_8@~QjS7pYPt z{td-fkV{;>@_BWHnw2;YbUnZ~04Mojrw8u%(5kczqNaMxypBGkaf zJfY4FU=)@Hr#{eUJKnY~S3BY_?hJ5hjJnaD;-RKxX`r&W^ogtb)ndtBzcr3o2dA9a zgvhu@iD~KW)Od46&ZE{gA)NT#J$5{z7UiDrG`R+&Ac*El!u_W8}O90fc0cZW)bxOy6hA} z*s*u37Sq2*w4N)ihNStVuIsw-YPl*JvT~3imX=<&yyoo}IP;~QJgQP!ujbC6BPfkR zgZUxOclZxAQ~a~J5H)0{)dv>15?=;hYq}7HD^-Uhx+y1?haONcehV5CAzWt&0ab~2 z)<%nyz7U-w`AD=>=3gp^}QN?BrFA=@eThNY;;*=M)CH_gy%SBnrDcc_KolRb29NDc|{+#$6BP<7Uig(PqyL(@P%2UF(UY+wE zI4CaZYBTZ3-M;^ay*K}da{vFwPaBo3U6Q4Uk|k*fS=#Mt$k>Hajj}{!Pm2Tk#lN~8RZEsntW^c@( zHv~_M`}UVcSWMQ9XMSqwxN9$Y(Z=w5@YS1FXEp8a-MMjXZ&|ey$B3G$w}`#t&eL2! zWiYo%qhn7_{`gRn$ob=-TSKto`2L4!pFT03K8$w#FwAAY#^qZKVA?6TT>^L1^S29l ztE9cK7K@0gc_`23IUFV8nlC=nN_he4kvBO3y6c5?O269M-P{7zHtUTtvuk8_*3EK` zrQi%>XZt(59~Idva?9CQLTQI-THy7r)VPW5_B${QXR @iuio3ZFY%aC6B$q>(#1 z{r#@x+pATb1tAF!gZ6i^ym!pJGv|p;ovUCcfAcM4Q&t;MG@r7G*>v5!SC8Xki1kZS z%=Fgj@MpHAZXi|7v%%1o#A1D9CUmW1BQ#h}iI5uX(n&K<54I-hH)&-gI28mK z%5za+CAWA(o3w>mk~Yb#Rww8AH_hBnIf4y5&8u2RzQC9CVK+e-KPpJakh4U*^$+j z%7@CWJ~Gpg#wy}c=rIxxK~8(@$o<%3xuYfL@_j1ay<-y5&a^dHn^xn#ty2P?He-D#Nk;0L@rdv)e=cewV zJ~YWTC}JaR?+e^F(HgSYIr+d+O_C!o&|YmYmtnu~);aVPyGtvpq3EaZonhNZH{<2X zS~lI6{uVM9Va!Z^ad-66-mQ*jGxT<>VXx16=wKD!-94fIxLP(>3XX8RKH^!6`8ZHE<3_6Gaq*6Rt4yiAXi3u$BFJ^ zUu~C35`3R^%8KzBBU5IN#c?I9+`LQyI)_|stP?`#Gzm8bVH)$+CCPR z`KX`IdLu8o@Z%>T@_O<``V>?ReK)_Zm2_#(_RffUdc)0a&0KALA(#DUIQFADrM&_o zYeMM9wB{!Zi(bUt%@K&t2zQ^C%7GVA%Ra6=zisY0vnW$Ie>MooHUJm*(4Jv^g&8OQ z%F(U{sD(DNxTq)>3zD+Jb;M_+Yz0;NNVc9`(@;x3xkH$uW0y@5l+Ebt(zxk7XL_L8 zP<5lAO@bi%QPy$hin3kjN|sNcxV zvDNsLU5Iu)`6W4)W$X16ib(u@u6S#-m_v6~)5yq|#`3&%@dT}q?zOBBKTVOp#Q%_e z71(iiI>O3wUul@uY3tVx=4)$xJ7$c3T6^&KWqT0zb;J^MWG5rIQo}A95ch0G*Bs-a z%Y^C|(Uv`87tgi0T(V<) zw0v2Lu@snep13gSxNoR}(SEjhz|r{Uo9~pbX?>H2@nJbN?CBT}bWQf7qjnO!7jFN+ zH4RuCZo4~eU}q^QakayE!uID#uUwzk*Pn&6j!7Ve`rd=h>b~PPhwHUC3XM8hYTk5T zP*D#VJ=RO9_*Ox7o|~EA7r1hd#F%4^TD4Z6Sds6};Cx$#8)aZC?EE!2oY6Pge0DUe z4W~d9)3YdjQDd0Rnk=ku)Ef!tDB5mITOG{^xr7->ahxpi;PA05T_ zJ;npV&YUU1=kruA=)<+OtQn<>#FFUowJ&luwH+z_FuC4xuJ3NI`IQ!~Jq?vy4I)G~ z&bdN^+5q&9{Nq?U&TYccrB2Zy*L0`4liP~Vk6}`vuM^AE^t-0WeVy76{k_ytSZEkX zBR!pt1K5^p`!y^7aX)7izqnz_U^*{!NX?A@f=dyzni=qDHrPbCpFSgSPFJQ4dfzZi ztDk}0m4D+BgJFyu`Z%_A=Hw3R6V(oq4inXWU*>b^gp^R`5lR zjI7p(rSZ}AT&^)@KQ{rzvLOSvVl|LvPN%S0m9jU_8Yt#RT74hpzkyQEju*L~?8L;I z+kNAbbB7&XnWz*sWu9NSF{rTfsz#F{cIr3>TVAEyBj3dAw&v}2Q#{o}r<#TEcTcJA zM)}U$jeWmg(Ui#9qCgy$Rs1X?)1wyqBV};&{MX1i4-#WadN-Ci;C0;&QxwnOPok&Z z)JsMW{$p6WZ2+~ehin!aG}ddNQ@77hG|Ovi&%~ZfqI_TJxR1+{naj{xQECb=xF6!Y zEJ)v_F?sE|r7y)T@=M-w#8BE_No(%p*UiLr2fr-+V0|hkPQr}cw~*}Ax*t_J)6*1q zM9LqRpQ+&-+a{EMxMre|Wvo%|B=j!bMy-!!JQsyxXRQ+QCYMVH?R_MH7OMh%5VsYo zW{7@F$~~7CHOoK2!+oF7ugU5b05GJkP3Kbvd&9eXRx#e^o=qDhbZHF9DB+lm+TOms z1GTsLI0aR=uh&lWKOl`xSmZn0{JJr)c{+v?TRKP_{PFX`BnmI-E!j4@`>01LcX=-cR? z%ZE~K8veYYFe?BBb!Gta=Esiz+}l5Ru}^E~R=Fdmtm|^Wn@;A)V2-t8;%5~TJ z#o!~wOi^o!K#B4Nt9!nhap@K+_p{nJ-K}2@yGGKLG_=Y^DW+)KK{)~d~aekPhr4-)mP|s z_vvH7_lPS4GUyse3ca?L+oIXd{gt~q=3=1z>wME6)-4P>^b-p!gG z#p0kp?u#RvYi@g#`R!A@4we7zK;=}kfU*FzVs@z{!9CgsOXFw0&#kvU*nYi*5OH`_ zi_xb<_S>-xx;|aD0ri$yAy6)o;oD5HV)(dCB-+VxsQ7@P^iGEpSl9N}jG&6wT%0lW zI@n+?JC~owYo8J_DRPE4n`dKNb@^|mzs)yy8NQO0ba0i-hPPj^!G%IuK9f#9mdA#! zuP`|Kp4Nc$jpBI;W#W6~Ld(AD@7^I2o?@AUQ6cE+SwGimk(ZLsRNFG@jP9ET+E342 zJ@lAA*D-OhNkdrC*24DwK$KJEp-h~S5cej#`MkNtuYmeyOK-#%Pi5L*3^Av3zsjCJ zSm$&i#?U6-cAJGQN;(i{afH&s_0Z*;-2BYN3|+aHuXw9;ZotMEd9Qkt4~Yuxbv+)$ zE2-4Bc5vHQZt(B3Ai@C6;dzX*svDCdVXmpYyIiyS^|2JrC@w)d*TBx&2g5(?H{`R? zb%u{N-uTwN@jr(npT3x&<=Iyls}+l)6aVE(Ha%(`seu z(KEOVyK5#1&|WvlLR>hyd)^|$sVLD?u>^7u9_MesJ)u+e4_xm0cDVb3RiS}gl7U>Z zDucj%fm+)b-_|P=K9*ZgS#HmYFMP7sa9GNs%SPv^kPz5Baw^R1!JaH0Z zJRr1wNY&!X_@L>;39&x*-DC#&__LtzPMk`RAoxZ`@=%vDJI)qWm-98>p^NxQz0Ga7 ztSm|-K4J_tQGE_H3TO=n=*kPtcm(!*vVeDB20kHgXH*eaMCcw(?lT+fThJr7F^4SF zCcl}7_!wo`wLE#i(+a&3(5^?Bx3y*WX?6M*qVbAK%riLO!5@^d>?q%yn+~%h!wp0r z94-zbU7y}&qxA!y3*2ebzQ<#jZKtlDZQ7e7ZHVIvaxLC9pOP*s(B_<(lQ$nQ+#=VM z<=9#5bt8H3R9mlH1(7vxBfn+F6#M)e$4gn4bSILNvRmQ;amAIfJ@%FfoD%tu+*r04 zcC;QQKR8bIF`Pe%U7x-NLzU=20W-DYccCrzCpz~QA(U5VYcWACPib)LmTl7OfsNRd z+sOTDx7J^ZAn(+^$e(nA2+`)t;3n40re>Ml^+MT>bP()6Dt+z4_T@gZ!qg^Vd(K#SVRrkoUFR5>jJPwhLM0RHYpu zdwch8xz?WvRNv~zxsy92#tnWvdPlb1V1+u~C#sb?G}OL3!ZHLh0?jLZiWP^uM)$r< zI;aAC?VEAGR@A-(jgyB_vha4_+U=9c&_Zk?3?zSq;Ee|yS?g4+?vTk zqol!cPMm|wwVO%}p7&+$pT$ew)ho1Df4zL`NQU< zb@ta7Wq>ocM6_9e!|%zsCA?|+o7}#}X3Hzk&0L@8dMEG9rKSOJ!}>paEzanJ@Sf{Y zYd@5;Ybg;K3sEPAGky8=mT{_nPEU zKP>b?!py30|AqMshqRfSSYzw&W9)7&E(rvkEpm6&twvr>7p+P+l|YXKW*^BPJ9{yR ziH_BEUzZt*@yn&Uj5feI49n$N04q0DgScf294{MC^nxbl2fI@2(eri6s`H)`iETHj z4YQI^vaYY6_NP62t!%g*(_-J-Smp~2cayyGI>*B<>J3N~86D+fHV&V%iuY;0IXYPU zhLmbb}Ru4`xdM>=LxhYz?O>R%IMb^?nuut$=glA^CG9p%CI zhSPkn%q|prlw-38?h||Zajqupqr*KX>d3McQrCTSc}^LccjM%5e{fB3#hc?gmUr1FGs$l)vkE_s zp>sDjHk$Q{_|I!fw@atT=IpE4h{<#(?R;MV((O?9q?q$`Sqk+Tw^|#Dxm?QEDM9=C zPORDnd~OyZP#!Wo->V`WICd&Qx&ak_C$F%n(FNc7z|PlzgZS-Z7UZBW@<&9ndXM?G zkev`AQ{3%EBe9@FG(!5m(>e!bv1&`zFn1~NFxOG?hQ8-%8SS`of>4^oY*#`&Xp|Wo zy27=O931sJtu1%>Ly&?h6ss)s%A{!z5voV^-kk*svxPx0;>!K&F&kZHVIT=zggl^E3=ZZ9GUGG=I2E<)=ak9-G^-9owlB9+p6 z^n9;oQ^@m2R>~|v4CgJNYlVd8=6u3vxhXF+dD>owx%Q&-!X8;m47>kcV#+Pew+jx`Fk z85&}!B4>ejG)7`U&^U%PN~7r*V_4i{ZLaBw@Ay&krEm=2*p!!+%ev#y<)NxV9wz2F zyGGx7TL`K#zM7rl>iW+2L1ZQcC5%(_b%xcArMVN69< zXn*B=%ii?}@fw%+3{+9i(Ia!0<_GdyEc(d3N`|lMjjUg&*!y=1xoQl*(OO>FD@=lS>TSaI&Lbf`+!=0OG7O2(K? zvc}tpcV4uGl1}R22eB$dmb_>T-krf}`NW0bqQ;Jl7gcceG97WTl*iI_Kud5#X1q=o zE^q(OIW;mXGSvX$4H3Tu?J<}EP&0Gm9p+i*%g{0wJSg^X$3__Y5Dm|qu@jwo9pz%V z0qHoma8qhG{9nfi7|Wi^bZLq!U;x*<%|EBz7`GGfKCszAF+!C#FZ}s-$OpAME)RcO z55GIF&_2}K0F*FTHnW*^KQQM*sj+Tf87wBXZU7ofeQ5)-f`p9oA+4L4&F9!LaCpkC zA`53?amh*=T~iF=Do?0CLDbD%Yeoz7ip*xygD!2jC4VxypIv;1%;_QaTG^i2K;q^O z>?uA-uVKaiDQLOff$+fcD6Q1J!h1|c0!-)Me*QM^GvWC5nrzb#H?gYqJ2pNWxpVKb z{+#c8q#XZTr`(i9VM$5A)AG4#ujx6@70|>O-wN%dd9OByzV5#A1}U^DI-;ac)K$6Z zzv(ea@Zs2AaZA?0?3&KH(y?%vNTJzL#g-EjsyxR|(A`&Kq{~b-c{Ow*x$O=w`+af& zJ^-W|B)U8H{y4Y%8gT5yZ#XT?@JW|Nb@NfC;a| zNaxmL|HAC7oh~tMET`DxY5mB3PA4gb0A?%HR@_{ynP`UFP^(prcv9T4FYq1Kf&v-F zacDw(f()sU*T9O~^*Y~KCsKL06?OFe2~hKorIm4`zKsmB5%6a+pw;0 z#E%#d_g%0wexTf-`YI{TYaK}apq%+9Q*1T&c}R|{#82&jfht!oD9*Ek`~{2zZ-(Au z^!tjaVg*%J&`RwAYlN(v_G6r#V+CDgq=fvjcW`L^46p}R89cSx1CR})a5ViE1UrJ= zasA8{q~WAzm=}7%K|p()DEO_G2JN?hJapV+@DQhxIv5}?!teVDM2!b7AQ|H03zl)o z9V@ebdtxYv*d*4Hz9TU4jXc2A-$78kCo+PuozN7^;#)9!c{v;xUp?b50 z`6waq#kCT|o*1!C* zKVBD9afoO(gcdd}JsQKxvI=g&vmejxMxB(+MyB_`HZ+Sqy*&@g;;8ur?8Fig`k>!t zdNjm>5_0Q$3|Ob_FvSV%_+N@O9>a_#Q<(K(C+yk|ZVJ|}2r6KJ4gPKIK;u$^)dCDb z(-1y3wQ6GlVp%ri1zJgYBM0-NHgWA8h90UlTs+Ue|Ex z*5ZZaq*4j?)VSG{05pR7^U*|WTnI-Slo-GZtd~QJJ=X5z>~Sx8_g-QP)C8XSO(YZA zn@5hm8oLHQ?>#R3LI@J-F1(4J{^doef?VaSPxRKgOf8T2aQofs5%%2znMyHnA7BRp z6W0NAo~RR$7UCb|yeu2P@>G$C$ zAVOxxs^x*f@XHO$TaQ7!(2`}Qp1namrcYYbFz%-xd_1wLpiSlStjNp(!E)7u)F>8V zY+%Vvj-D44Qh;s&Hjz?KfznP@hOfx?`BH+}-7nG*lGCrih<#mlt{3OUAjg0CkwzAP zW@DGVIq20SNa4?v;g^0KRPu>uZn6$>9Xp@RvGXjJNdT^3yFcD^-8(?#6IIw z*c}-UdQe=e~-F*HIP7%5d)%*_urA~z5iN7 zy5ADD-}5S@XD?I;L*F;#qq&Guy*QHO|L$NGH==PNx%te}usJY@|MhH0Ms1P-;M;rf zu>c()^wO{D&R;g00OJ7!IH4z_!$*_TyCGCN@$VD3XvUEE;=i8^K3Fg^Cu2dciL~;D zqx_#+rH=7~fj-fA25b!jq?{`;HZW>SeEFy?hz=gZ7DnG@>)DNL=ijnX+Ny!sS6^oq z{mttOVivd40?xmf*m^4zlYUMK%-Euttltg>Qq~D30h)LP!X5LrQwaeM8vV2tqn2l`r(ddVq4wo+QG$a9 zTMcns|MM+@2PnP+rQIT#dXE(V060^*tg%I%n*{(%2 zqgUhq*{)^b;eXpM_q-TNZoIv0>$tUf6UNcm{nXmGD>hkj<()Gh4zlEuEw{fOA?d`^ zDm!C2Xw1zoc%rk@m+984ng{pirM}M?eIo7b5?a+#R7$twhq!I%M3lJe7pLZ1ga)4e zd(!9B(IfSDo)eC6|KncFV^yn6EH)Px8gzfx7RB|`i6#EamZ3dXZ<{+i1~BUMlPA_D znaPe*Dx1AXJDW2zGIKIB%U7-3pnYiNe5EpRKK*iHT&~Q9`k86>Znt>TA@|Xu`1+ii za!-2Rk;^??SE#{tN9?RSEZ0jP`)6g`+CX%Trq^Zftt;K+tlEdP*J?dGn`Vk?I|(4n zn#azWxOy+ri7SnL-j|tH$8onC5ck<#s*^eXE)z#Tmv>2a;I?xWk$&PJew{v1rJc3o zjf(=&l)x~&CpK)ut#^Cf`&LS|lJ~MkiimV#ca-1us=ThqqEFQS%G#a-zto#iVe5Z< zdDCh>o?8S-w)ot1OfR-No1I~%K|1U%4DjIl&S2dr8$;PF#bX+=k?8)eX!Ul2r07eT zbe^QRg2z!VLCTKpSjz7Ff4Wa%cl7UvKgGE59r@4}Mm<&Igt)|;fr$+{L5cvSo2|^> z5I=Af+baE#aOKtUXNns_AN*sA;(z~!-yWuZZnx7X-g|85l;|g;pc*W7u%6>H?^+e5 ze=s&^w9_>?vuHix7e^txC(G^Cj zVz+sd!%sO4q_{Xe#@TCs`*7avL;&}JT`Tm7>y8fOw!k#k0~Bv^v*~UNQCSNX^a#ur zt28{)=TR&&W*)Zrt<8jNAkJj z>}8Cj#T$&A2thKij+X-RXM9>EYn^vmTJ*UwJlbHxz&qEOryGzGvypQS84cIncCh5B z)Y$C&ht}0~|1cOai#@tIKE0|^&Lvn>@C1*ugavo!#)PjXR)UU65pmB*#Qj(Ewo)SxO zy?7d#^5N~imdokY8$TBLBO~Vv&q};HwWtO7$Fm?m^Xts$J9oFN-I&|V;=#iKd&RG} ztUbtg!WyQ(NIe0YPC2lfbZVWg2x>@1A(nDS)S$4v{W9U;k@$_bEpDIKgCfVMbm2FL zsf9<{wS64b@tAUx;6aHw|NN{0)f3`UDvHFjr(|T)&!e|Yhd9m@b*kSG8M6`R@@~VA zBtq#Rk74^}_~TJ=t#gzNV_v>DsxV~v*~Tx~cS)M@P%i8m4Y_k2bCJoLf0HG<@-daUs_$3O3T^@jl)R zT-ZC2qASwSvFF;F339sW3vLw876yUJA{I8ZhU7x<$MoR{%#X*7Z34&{|~CW6U`cIoZXMU~+GvNtXn#_+HE!{J!CZvL>XMMx@NE&YNs| zSdhxoMv3dTpPJ!*bz;pX1YO#TDwuEFA<@kQkF)vmkLnlecViS2ol*zz-~pdb?~es8%JKn#{(s|j8ujLmVC;UGwM zoum=6D=T!n70OGmH?2?~nJzr#ffiq9D}teNXTCB+pQSCV(y z4;&~-EL?AjlD?bWbAQkfo5WWO`}ZQun=a%*Z@^9PQ3PYKuj! zQJoTjfDLz}N3LYmer|~GoNIk8Hu}OneuzoS_4w3W^P`bnsCz{KE*UHUBxTnXB;d}= z2*`?9N@@{FUshTTB=yYCZ<81WALEoB_EQHe-`fxOG|tF-2px8SOZk=KM%&;gtp!jR z_U;LG02g6p!Kjvn}ejbml8^u1wg(w;?V$J&+JeerG1B;^Dt z5Q!;e*xji6I(HXtxjqI6P3V_#*R$(`GtYKH?{Sj?a!ye5oWscYWaz$JEi05$aqfBk zZ~9oZSL;_LnW^lc-4;L&1Xn?FVkFDfwnA~$>Jy!?x|M*;RlZ~GwT)0F+UgLyFs7g3 z;4MBa4I6H#e{IFW=fMPO>F=zx1gtXWn9WwkQZ{>x5TPbDFfJVj;!=?$TzK9AL;zm8$rH~FAfC)xhBOBls(3X^J zk`NbfOeE{)JtDBB20q-n(mCfFZ{ohWoLwX*}@qqm$%;_n1SOlMV0DG ziC&L5x1G?y)>tfORfisNmC|<@>D4~3Kzsm){0tO^Zw7rKgo8@!l*Ou8Afa(WV){ekQ}LbF)>y4%TnLoC>s}luvu!Da4VMp1lc^D*cwh?Q2pjC9#OyeB>d? zS^YgcC3j5R!9B}^b_BSGXJpx|d`T<>UfPig^!Dgt^ORyJ!vQ}EgP@5iV8kod(2Soc zQt1LNT|lG@J11=M;f!SHl`|#-&v7;=OW&x?XU%Q-)vZv{1zFO}jcKHu2-rLH3dD&E zOlYxxK!u>O+#IM#6ujYP-Mk9=a9!)b3U|WCT)$OpHQ4fxF8e~bgdfc|L;B?v*7msm znF=(wdj}kuXWH>&{wV;YeRnasN3z{eo{2jhWp>aB&KFXy=k60r$@PL}MRV-8NZx0U z0=@wD|2L!H-c1c(FU+#c&(oDioE9 z8Ie28+A?{Nr#qp8v)FaoE$;q)@BwUR@uW&s0B9=aLoSNY#2TK!=83FJnDNL-j)=(j zA#E{XM*~K+0R7_JD>e;M^x>nH@Wz^W$fu)i9y9OKkv0R`EQ^n z0TSce?kQm5M1}C%_f(~joF6CMctc!22qRc6mYGxPu|UizpwaJd6DF;*NDN>W&;n*_ zZ3Vz?Tg0PB$l_5uFk-j=vY=0T`4Bu1mf_^UN-0w?rS~4Pt&C72cGQ2qphHYee>+K1 zAg(PbhgUos2d~%;?X1kAG1PO9rJEY+@L&Wl#TngJKrVmm8h(8}a&bY*bBHOYeuxho z!=5^Tz~p#*IDoppcpHx)fFU_nuSx>Q%TyOFsuZn(*J)Otuhng#m>u;d@1y{B^9r2Y zzSqjc>j}To+os8;o%mi86XOI*#3*wW>kwddZCjQ0uZC~ZyVjpLpk;L$bg>Bc3I~4+ z?cjS3H{NjeCxD*!gPs*?s4mtlGW@&&@kldkXF`h-acccwlOVy9v=iQM;ifcX9D&%%RpZ5Ph@8w zq=FU*f&~Ij3VQZsk(w?3S%(&bG4Kn<2?&L6Sb+ruTLAy@(LN?0apvqZ7+~90Y}k(_BwGcvQF-@MCa9=4hLytqJ6g zBw+cHJ`CROgctISfa4e`^Oi#xSweGT%O*eE3Rpr(WC^J>85NtsSg&X~dffzuF zsVA8+uwSU?Eu3Mov$!%slMum^AnCpM;ed`>P6ylruM~tU%ZxEYJ^|9%4vxeSDTdO6 z?*w_)gO#*z@EeEr1k`1sqU~5_Wa1!O{$9=g4LxXVZ3sChW8nTN?5RU$-~+u5f)88? zd>0_aNr4bW?}k%!H3RqLxDh>2C0mrFSHss%_Grk@nY_@%axK2Q`K_l*F3|HaFpfKO zd0>IGk6hr_z;G@&2{cx?MQyHH*R|?!x;UUg?<)lOQci{mD5$zuE;~gF?5Cz0f~W)I zkrjx$gE;v6k8I3W0ql=3`J^o{R*=eSQqd`!AFF`pTab-=y1i*m{?}D*G!nTvksl<$1oM({ma_Z6KK}Ddes9lGMy#Ek5AEe+W1B4}i{FQamiyM{88!GAB?@AsL2_k=tC3 zZ~q;T4t+tiU=SGWg+xq(A?xSW=}HjDH7pDAeqo+S__U01sCo*YONsbgfk-zk$_@@2 zTu}p(GA>0;fPTZi)XBe9p$+W}su|tksS-hwzr{Xc^iw4uy4`}#NQ@ufic>nD)a)D zdB?D7{wu)gKx-7+kIN1!xH~G&ZT121q=ii%0VtQ60Wl)j=+($+$5rZW@KWy3?h3!$ zg&=wb;{tN{Or3a1@}2eHuy+64NZRzOjk5Ru20T!s7yiNuzNNi?qxyQ(HFWMR47m#(;N3;;c$KYJ3 z0@G+|$dw#ZqrE2BFL8oE@Br%5-K61;EsY5>IW9UumTNkHQJFd!Gj^@)v@ z0}g44ZIDqIwKCjzTMu##81P154A#R;4^2StbHMN@&_xj zevoTu)L~#WTsrM_2$V&y5JPKBLZJ~a`0>EK$p)qTncyefbKqsqgy1G#^MWpGTR+3~ z>ZBMmYdHfgn=x6q?>dO+k^8bj5U`x{t+zr|@~x`1)Bk>wI2a8g5Bf9qWxKKiaEg7F zNwO`Tu#aMB`^a;-Z$!AlZTiFxttlrb=}PFy+C^7Ganwgci_RYs-hS+~Ybo{Fy^e7%kTE_h=X-Z6=U%>nj-$yHo%K zqd`Y|P(x&oeQg%u5-pI+94epjDB5-r5S%rfiviVwA`zDM+k|_o^*dbki&+pH42ZR$ zPWP)|Ws&d=BR%}O#%8d-To%$5x=$eRgYwp20T%P%p%zW)8wR>os5&%0(_vJnF3O?* zhc&9y#Mo3ZS`6=Sh)oLW9ZxqxYD@nsa`}y)dId;>Dasc#w}%7ag?S^I{ z!JcATio}=6D+n342D!DZP$8{+UJMq7NUK(*;*iFz@d| z%JI7X^L02r2ecOY`SvJbWmS`+L@WpUyLnKint!*vh=pKCqMU~Xg$+R!$-#x3%5Wjb zpZ*UEJrsks84E?n{yITW1J-^`03|Z$U8Gi-X(R1)b1R;W8$_7b;Gs=XpxUAG>@Hi; zR7_^>vA1?pIWMH$3nArc#9Am3BSOd*bgK>sC=lO4+7h$;03y5URS%Y;vwA&j45W22 zJaJqZR#)z2+4Rh^Y<;oJj*$V$7t3dTJ}&Om_hm zu(YY5r_Lf)rCbZr?9EAL0;sn1uUdZakpq|@%aW@H5OYfs ztmxe@(*htzDTpRyj`ZCYy(PK2<8pyj{Fr$u@-ry&7Ys> z3R~0e{D>sjM{tHHfyf5V6pD^N(V_&Z90)XDu$>(CLS#9V?eiiy9r~`lRkda1NZyYR zt5`4P=z8gV#?rg~CTrn?U2I|kc+whb)Q<4OP^XwKzBv&xn)7WIcL3`c*im0+7Zaz@ zMD3Nfs0uT#BIi7RJEqrx9WQfwT2bpqat_){B=S$#oq~*Ua+z`Y1zaseUn&_muAQou zxMA_KH{6Ss$q#{A=hovJ+VhBXF=qeVA!&LnM+A{S$DU;7;!q&C)RQ{P4|k_AN;^IH zRHd$)4}r~rc2metsPXHf76*xrV^vpjC=#clw1*~pATwXPWC%=Y@DUr(%HUp_de@9j zG)JimS^oNV{P67rgK$rV*q?ioyw^d-hHTc~EUwiGndk$eS|g8FAffO=tl~)$9E|sJ zHprQebmvhZA{g)@()nMivz}8}SN#y&qtsdn02^7BFv=w%mIcT%g!y2Ok;Yh9#!wjN zMln2jeUcgBuzf=TEC@@@G`iJwoYM?m^1~ARfROwX@))K1-UEIRX~TQc_i7U{I*4~2 zV~V}H3GUqbQjs{(9P%3A74gZ@9LTh9fe~={s}bOY-_pa?G$AHjou@Sa-tl`R6?_JY z;@`M6SWy_h?67|p+9Vlhm~3cDg#6SXGQ00GU~*1r`L7IfFKKN++k~E_nUFscbtGnB z!ic5Wi$_7gA;PegeYKPyPQB(Ht9V;>2`g)rsc`5dlc&%p&7;`$IgpYd`Wi&n2klf4 z#Tc>+Kp)GG9?^k_DwTzupQ8bslKdZL7Rg6cB-{xAm%Jy>v>Q9Fm>r*{ODxvJroPuE zieDLgYkFfhYRIU0CuO#eLpWMJ90f5MpM~{l>D{O`+ZO%W0#E=CJTmZ|-`-*J@-}!N zTbJYeo*)OT@J*Y3Zrc%a7q5uO)Yc@%N_Yq<2+jY<;*k&yMw>@`u^b#8E*7NvRiu%- z?QWEuTAftZ1JczqZFUI$3QfUJFXk(V5km~>)eM}y|9($3Y~B#sgwgcL))MSnMVk-~ z!e;!Wco6P%Wj+gAk{K_=cmH8~A^bp4pYkBi{TVo!_l6l7H+G@u(>8t@;u612^1fpF zzrr*(Ws&HpVzQ>~-bWjHj$vDsLw@S!&9;_5L5kDnjmmnINg9mL2R9$JHK)l2nu<_HS zw(Q!C6Y^XLTt$V{mLDQS6$amizxtw4uy65!G3^3!|Hkq65ic3=Va#U?Wv3Z2Hp2;k zC20W=pb9`h77oI0M9zfdHAQ57RmR9J)1g`NBkQXv&<$4Tee<>ZeJ8MSlKEMqh8OhZ zeHRz_kEAVbO=jqJYgCQ7=M#IAq|s&f)yY0t+c!O*a|~PZ=8wz`%EW+I+>F47MdX}N z=GyYz2q0;>7pN*&=1&*lza){InTk6i65gaZs&vfy`|K$XyMd%b9oV=4>yT8WoAjXn z)gQjiP&o!*(wzx=I=eZAN>23{v$ZgeFZ1$o$WK6I4d7Y`q9WxFoo$GWIf|f1& z2aBb2=+QR4ic(r5vsCCzgKlYYsp4X&e18AkWh-kH_0%*Dj9!{BX4LS7&&+8o(u~h~ zj7yVCw#elIK;r^!*CR%pn$;{VZ%%{{IlTRn1l{Xgu&Yjj_AP!=k8x1H%E}kE9C;R&@v(Cg=lvvOn6xq4ZH(8UQ+cG*l{7CpGs&nF)$AOL|YK63iF-PC+hVwjy8#69245H<7=J1 zs3A{#CIB~-FS5srgDDlj?+>pO9k2?j)_>w)4+xX5_)u;6;c2)w6VV|Pp3s?BFPK7h8{{IC_bp*5}q zP|=QGwxZ$)sHwd?p$k*zsQtulsDKg*6klu0jj93CtxfImFccs3iyp;#Yyc$O>-X5q zfR|aQ1+S=jZ73DyW2!}*s^YBH$~O?#G*&8F0;gGz{>NY zZ6pea0O0gsM@B`czM(X%+s39j-2_&o$u#qu4=E3D zA+wp!@tZdkl7HR$4}bu%$MBgou@;dLU++tJypDzQpv-OGW!xBoONX3!dS*50F>d3g zr6Bq@6nQIfo8dRdwS5YJj#;P_n?{Qn6u6A6L)!%Y^r-~s8Y&DOfHtP4%ZO|mfy|i` zWfGNDbF07+EC`~e(IbinSGY`)d5@ohO58?Tm&}Xmo`x zysF;u%xPv)7AXg#S1?=4TLc0BvS$HHfX~U*dfCU-#!_mY ze{9ow63{I8zzKTDEEycYSeFBl3BYxVSZ+h05CML`lyxE3-^}i2bK_ZlB0RD7AK#V+ zqM1c!Blfs5zHW1+toO(2&OB%oF^Q}xAu)NgazFN2H}^jUPnZVIVn8G?OF{Om&x)o4^xkTr3C+@+#SVxyskp+ zkn--5*Aeh(rFYtQpM@u{Np7#uCK_WwBl0it0TRSLAUKFXG@K4E7KcuekG=uaj?TBs zBjdd09J^5WJprw!F&awMw|U5oPJ(o$Kge z7cv|x=5&7x-?a(&k1zEXV$8}Jmhk%15FUUCVKGXE%^#}==>WesnN)YLdbN8k%su94 z__%EX@CY*&prihafnBG;}(iPn7D7bh#sCaZ@w`90I|2-V%lvVGE&Hvr z?XzMdlT&GEfb`aM3d<;bLyM_SkmEO>h2tLjd&ayRxO_;M{LxTUd&JR@_$>t>P2X3n z1DbGqX~nl9M&Va4AB12Ch;yor<_{Nyq}I(4JW{2zbw&MaQbODPG4h3-aP1ZbBb{Q+O`Z>H3yx(qDMt2)ZZpRhIBuS#?t*JZ9z~r*pWC=xYbz5 zcA_}={A*{L6H70{rEc2S@*=!<0JO0z(71tMDTt69IYkc%Iz}~9YK)q0;CNw!G>7R1 zN7ImyJLjZw@uXVzbm*HwV_FAva2Vakzp1R94XdwZDCohtwKXec* zmrq`(TSB05*kPk~9UyJXh=I1~03MV&eyJcOkrI;I=PtaLgMlWZ*J#%A4|hbRG&>g{ zK|RqkxzNqi>3Hrfek=yNP2E4A5Nxyy6;E6X0Eq`lv+qyvV1E)NceKuar1oxpo@k_y zCV)D(GW1wJzb2_r-U*4Q11>yz>0teVZ)(HN98Ib!1Na7%M#+024=n>%sAc0&t!LWME6ILysuj#;5s;j@ ze?&S!aw1n_dcUX9Png6^y@b7e(s5&DO%m{bN<>AB)uoi@#vT8aal^PrbFt4j7CwEl zR7omK2QXvUVE=7bV-Rivg1jHScdI~ywfC<;1IPiXajDjp^y;WK{(!lek_22fbnupV zZ{U2(E|yYwsLodn_YEM@^Gmx8cs-~DUGU1^K;Z;3!r4jKr;sxst)wrX!!ld!p_>~X zP~-p{&3hTy_RAsD5IAoFlBTBf-*7aM?@nsRRy?v|<%&Lsm+Z^{yRC9yIh^t*0tcJi z5)*)G_T^;XYvc$@v^9ioA}`TPI$Gv)HW z2aw5!z1*SJsBZ|Q5GSy8)oR@!99LZorO5Oo7>|vAp1?)XQ3kFXbYEBJk%V^v{Ut0! zbUi60AKxNy%6ZzQ06_c})H&Y|(?g_>pBMH(t(H~PjKvy2#aNg2-f|%xJe*fUTvwk> zxe;hFTs^YK*Cr|fXr+6ah0j1M6&SqHWpDN?qd}xLDPK?;IztcVlu?j;#xCE-?5<>oZYg8K_WHRF#s8{L0>Pe8i?4NCmrL=3QV%4BYD4Wh- zPY%^Oc@uP;oy8~D<<)0{WN4vO7gS#Eb(oq752W&>|+GKdMK| z;B~_D0mj@%6;joJyloNnSpX)WP7yrm7CU`nJf3u#clB{_Ka}>Dq ztnJz4)CoR%07VjhJ5j%~Z5lH76#8bm+XH7HN89cR`->&STF(khu1IwLbv77 zpcy8{towmbnLj%~g)by-@dW0|$U z?9OdzhTRA}ZyZ}MZ+0ew=k8JXF#7bHF!hq9g+fW_I5)>wy7 z)KK1!RBLd)y&}#jP2WLf(0nc*5L5EhT|ZB5tS0YTCx8 zb`6%-JpWVdoak*EhJG4=Jcc4v_~OB2N&VvmYm8APf_! zX1Ey%&q)Anh3rvmMIa;XI+!{@-G|hg@CC`iE=1S%mR$<27p=oZfhGGwwN@ zVEP<``C2%Wv#B5V`jJib<)il9fEc5kG_|emE@=ylrw0eWpt{gObXJ)-b$_ETt+7b- z8V=nNE-QnBfU<>R(?^>JUP0;pdzydvP1w8;X25wEx)_8&K*%K6Xd%4L5fuaCb9ERy z{DL1s2F5_Z_x`T9(7ui`S7w$y^M3RkWC|x8S5(#ngN+6yIAs8@HA>CGxoKHANCT+Q zJ`%pQbN1{$dm~giy_%s%ebhu9r2Oj9tiy+{DH1!5`gfKk$LSPA+(6wAgB)v7e7RhA z0gnvg%gOOWb=$_SQ1)4)x}T<(wKWBk^p)s61=Z7#l8#W*KWHkX{Yl&tC~f!he3}9- z5wwzr_*CTJW4Rs{q3AUF16TEA&9zosU*$e%?%TW!XJ>~nY)RDy8|6+qo|N-q+xW8A zJ`yb4b7TCACg>A4vM#0IRKyK3^+nqHF+rzUF-}F$C>*;}kFmcTZXfn}z+K`kAF8Hl zOM=>sRBotc8DadxLoH^daJjdjCC44IIvg#o3I>PZ;26q>*Eo$9!L@kF@|{Q-3rQ0V zJCbUpk$Qi`gj5-~st2dH>PKK5%sH-%Z^GEv*=5g12*G+^^Y18QjnC3Sm^~mRsp0@& z_8>-EXutsQ=&P?-PQ8|$KDFV4aV}GbZ6F7?dDk>xc}*Ex8%#bTMScxkn2r7ahDRZK zH^xVxHm9g0w0aiw#e&6UpsbNmG)Id~nqnu%C-IjL2<`1yiWPNy$S`>4R{*UH0ko~- zV4w}(sDqh^HR{r(^lxnq{-m4H;YlqV{f|hDXP`q>(qFQ}>wqBI4@gf7&=EqwVTy#Zj`G6FH*WB|;sU;1 zg>xxU`qdO=6*ww5T~JMT>b0tqgQ$mHE!d-kujB8gb^URr*!>yml0TU6Y@6X9YB7lT z@_so~t#1l=^;k+ksAw)E(@q6B(fn5wv9*@y11;+Z5h*$n@fimU147(kL1T81_Bqg0 z!1yCs&~Q%xU$9d!Zp}i!wCK%v$*x2sp869jUB<`4WmpD%^~}EW<;iTY)MB@rHd~s5 za($4O=R`0jo)-dJ`VeI=r`zw|iiypb0FLAV#*3@S}ff$P;IaeJntAJQ@xu9g> zWYV};l}sF0G8Fe0@4KX0R9^wqTlmIi!<%}_=n*^rwz8Te$~}^I7gUSH&OvIQ2Qa^?$Usi+C;aM_Gr`(_;wq>=<(FWVKQ;83xa5Pat2v-8%30_= zYZrr*pHbeF0Tx5~rOATmX$j0TEZyCgz!SREtph|90u3#(un&cRv;bdZ={9=sqdg3<1)l>J+Dc!POLlx+492Mf}v>F?rVsO zkb;JUqU5C@dAar!!SY^pLD1;xErE$2cQLtn=>;UM|6HrgiDkm;#5CkOZ^TPFjl(yV zhWf|7iyCgE-5D69`s*m)*yjye@pGa+Un0QfOERR%3`fGPqX!$*WsX3fB0ezN|0?&S z={i0^aIYLo(Sef``_w-~9>&~f1SKVZ|2p#~3WPzPVZUS~M`*23-Y+CFDP6FAV<{Au zE)@}oDZtW>@Fpq{`8`R^&IiX{rck*wl-_ zS2J)$ zpE!jo*^~C#7cz5}PYjMeu*kDsh566LG4Xji8|)c&P?&$!9R0>gBZNRGVe^PHTS zXmzT-Y$xc;8lesz@rkVQQ6nA%wnCu)5TgfY@i)!wLMfOiW<*N>6(1Et;C%v6b@;tT zP6u^%^hl>|MbqbA3?d9QV*z}DO2qsb0*dw@0Ku*UG~v3tcVrW+S}n%PO?*rqVSSJ; zcB1`~JWuBK$E%IHRg|zng%a7Dc}x z0uqI2*qaDFwwS7e!Y-GF=arOz9us2JK9Vv^3Gb6UV1RnV1|sa=a3eTw`(t{ICrCh( z1D2;CLDeG*iz!U z9GXC?!%QJl0Fg**_>}l((f^OH?~bRkkKcb%Q4tN1B9)?KWE^`|in8L^E0l34iO7g( zBPx{~v&?gjkz}tBG7gGEcEYhojy-?x`_Oov@Av!r*X!ln_x%~~d0p4LumRwU3D%je zonL$#Jzb>dMTSO_2U`xIxB!mu3yt=hcMG4i4f*%<%JH9*X@jZ46lsI^c(duGeJ#3$*K?a>RcM->_5ZhM-~TK&QTfUMz3GDZN(c1Fg;*J z1A<fs zeD zdhQC2|NT_T1?V86aFV6?3934jS-xLVvSm3sX-)lO9g$j~+DUz>}Xtg%JPB_LRw=x{wZXL`!nI*TAN6c+wN1c9v5W1Rze1G2PtWVj|F z_avoDpPyW~qzzJo+FWhY4sgd{mj&uxe1kW4Sby_hSirR;3!$2&kuh=?yq<~I9q>13 zaD!~c=@SfanTL9q^X1Pi2?!kNtj+)FFOzbxb(m(h@hJ&?P;aIB%p52=A~ikhN;{Ci z{SUR|KfDRM6a*QdG1>A^=u!mSFzIvoA9-FIVOX-BZyl&+wN`{C0R+CkNXmZ%EaB1t zahp9fQQCiT_gxtiFvwD-9++{!hEDQ85<_1O5(icPDR}->1%$0uyd5`8CuAp;3XJ%` zTRs{mvl`iDRAR~YuEt83@=gC<(@)HwYn^Ak5UD*2o zA%p=izr&i}<@EIo-U;RJk+o~qI(~7sY_A6N?ZX3`J6k=xa5-4ON?kEfdv%>N9)NQH zAL=Q5G~mwXXJq5lX2EsBot$BCxM*x^+d6{xqG?{B z#trE7ZFGYr|873C!X9yDxwz%VHa#b?YaTr&@Dor`HI6)<48aD%zd&f=PyzW@TLwP- zr{ntvi<4ys99ZwIRWM+)IyF{(VHA!;@g}57{J$YXfc~iY+KEbiRMp#O`62=L9ix8Q zE_GBot1&5TTI316BBng)uZ8cM6VeB?FhWb=z3_cEz)p z@}>XXb5mnMB8d;!3@tIg%i0tm7wf!#S2DQ~L}j~iVt{oG#gjNj0vego2*Zks(_4du zkd2AfgEaRDOpAFYz`>VA}2T?D(VM8Q+9_fd@UCV=71o2C{}uxK0q8mYJ!}5_ z=tv>-@FRbgM4Fp|A>;7LmucIAwCyacE_raBlY8UO?y|{u@+H ztq-onDza48Y*rM2P{U4#27Sy2%*7BfCcR^0H5hO~#_k$rP_M|AP`vWGsZ+`+A{Kxy~r_(_iCnat;0#3ZBFVqNoBnW!$`-? zg5b2I|2VDx!uTQ~N-^=B7_q;wW2iw*TX^H`eTE+PU6OHDqb&N>BAp=!%^!jVY60Ot z;V_b2){$hQ&i$^hOU9AYfDM`6c@^j<|Np2^RxjPl^5w&-uv$p)*6!5{>}-R4l&*Z_ zle-a-PMI@Va{@r9eB6lE{nb&(UhnNDy1vG3{NQP)B7M^sn6M1D-Z050H0B4NC$g=z zeeMFjMya5mon|P*W#~+5!hUbzHC~qUUPNvKcGTno@H{syc33l$4okD^fHYxxpwYD*T;><h zR5A*@eaiPj$)0Uvs`UG;AUpnz0*0PAQW}JIoBx@D()_4n#6D6X5C-DkDz+Rs2@4*P zjkUHmF6O|kp@$P^!%r&|G(kqT$}m4R@YiJvs|HNPf(2Gl87*5TL(yD!cZ?U=BN-qy zRtNr%KyU!V!Wzs;?t_H}W(k!9Uvm^iSD_WL#omGngFMuC<*y(i9o$7>CYoGO^E@?xI$ zOHc=VSKKiQ!I(XuMBe|;I^bSs6x|^2y=o?lnp|i-yHvZT1<8dVwBDgM!1@D|t!8cuIH zoIzOs%ynI8<0${K9guS(Dznhx{RKa!H37AZMv{z&pp|+8Y_0lFfw|&-=*8T(ur@6K zh*;8j3LdY0*Q-Ixf z(?@K)BKoEwY%cS-FQ=yfwD)fgiicP!nfC`x8OSTzjYCNA0&;ouf5Cn5=*};nXj{YG z?2Q;^F2jpgN{m((kb%eqFe0K(EP06y0AmcW!%)EZFMJZ&;o0YGd0}B$7G&q&&p!Z% z3wW0;tm9yMRQ`pYBzK1UtYmpJauzLCo2_O-?DGNUwFoeCn-HBjN{D*t0*o<166R35 zIh0tC-6o@z?sdVJM*k8ZMOor&u#nud>Zk^V%N)0lvg}rI70&s!Z@V;IE2Zq<(e6#KbPw+Iy+A z##=k9tQLksKtEpFA7VURDcy7)_x(ZMrK!7Z-+ogEaplG7raUBtA?wib#YBg*U0n|+lD;j)*y<{ z;cov>2fB!vr<7TV+QK!eRViME9n^r9(aaSK-~zYo1|v?z6FsN3E(9 z@AWj?7v|kWQH=||MuQs^{`eNUH5y*w;(FGrj*;A@(E5Fo(L57z4$~yGkp<0lcJ}ZS>bl!}Tx@`Q6Ju*u-oytADq?!w!@XJ+#Y;rd=bQdEh zj~qSu5Uc?7XW0)FK#%fha2%-be}iU%{Q42IVk6A!rSk3G?PR;kx%|qEVm@IlMqNQs z__d0_c)9wDg5DGBtpz7#;tz8v_TGV!QblMxy8U@Qq~P8Q27EmK*7uJIe)o%v>dls3 zZS>xIZK%kKn&=LMIisQwB)f(kh@Z@KvGLWg^*_U=Lc@qsS(BxJpe~Ho;jIXQ=zTU8 z#Ni&N7%zM`mAp7GyG`R2a`qaXe~J&Tby#FO1VC->g2TImJg{u>;z#tot~-#sFHo5! zq7b{}faTk>0lqY~R$;Z|3CS~Z%6{y?K-|JoOi6Ie2kN3N!9kggJkrd*mWb=n3wQ;HWmPaO+A z9;+~|kEJT^4#7DoF38>4c;NjdUqKZplOR7;H0Kz)s&&kG zFqS+Xz94h`VE{aWw)$^&Rb-mv8=x|4Tc48~Z>Glkc`VqZIaB@SOme6&o#q0F*~ zrU-jP@!N1GMo4If>2&N# zIB)?3@VHPW->w3G9}ssTK{M&^fEO54@bPT~h7sq+L+tZ+>ZJnGr~Uxaa~@V;81wN- z+pC>Zta2Kyn&=0TKmf0zxHt8mtHT8!1~Hj46d6e%NK_h(7RdHL#X109ajIox1iCB` zXRJezsiuiZ-BxZ@{nD?gFe}HRK`8*7##1i&)PfWUdt(I$ z7pm@-W3^=3aoC<)PAC#v@|$rEE%YFJ&$M5 z!!DTJMDJYE5(mkKLAKp|<;WUbR`AbKLGL9;yqUCAYo@SDINW`qAG1>GP~#3261DrE z$~QaH>;2r%K?V@4Vwe-(4z8x%(Cyr<&`QVyAuGD+`*~Joy3h+#m4i;}mjTI#0SknP zDk-9bXYGiV6H^drN}Sf+)Trq*zYAKRJF=anpvW`8T6i7g?u-h-*zN2bI?ectDh+SC zT&tH#R=y(b9Vh}(wOwKwTS|#bGqP; z?YU-PDO=}GQrKaXlRG&6NOr0)dg;a=mmM-R@|_2Q#@;Sa@=t?6kG4>n5PzZ= zR!?m>$KK_2M%!i+&(_l((A2z>|1RP6X+U#}CfBA>YmTx|b9~$4#CT|0*V6FmxdQ8E zlJVqecz2W`eryjq?Bz;={Hj}^os`IBk(BsoKtC^^HH#`R6_c)6%-|f5WZ_DRvPpnh zOwwOkdQ#!5$74ffnmN^w&Os$$Ar9HhgtJ-PZJer~9Rd3y5$6`2c_(i_Vcyz%Stf$l z`3GlRRM5beT5S?&)d4riE(UZ9;^gZI6;c}Kvj#?ZDm*0U6ZBPS(k~FRsw;bb$m|Bt zkdW&iBOOUU;B#hNq`vxMXi=(>?qdFu%tT1G<6wgzqfDxbMp<&{S8B7vTnvbIw(jCZ z=@es^aEX_WD4w$!vo7HVn9e*PdiTYojYn*F1G(0N@~lNDUwwyM%L3W@Lb%xHH>SJb z@{jRqi_Exq|2*N)WTIY{T)k$x%`@Rb%|dZwNepUMj{#-L(N-ouIkE;MEtUa*!?S+} zj2@1(?m2fnZtq!@e^I41NnyIysX2Hfp0#mL7UKnKtMPtIBAijeP67E5%duT4Ru8)t zl;o#ss4P%7iEo3F{qxMp`K&m$lhD}JXo(}6auD3UR)5ZYB=eIHf?eJA#*wj~b=7!s zD7`S`aI=YRe3$|z_u&4UkTc;V>tuDg)h@5+58%?Ljgo3D>p-G`{HJl^35u?I*3qkF z)m01alarAzic;>^FRi#qK21NQyeFsjJoP;{x8Mn(rmn_b!gYdEv7rG3$>TWL>@N6r zj;-${>)vDeTtsW2#6bmF{5rMcY1q|CKnynt$WaoXtZux>*6A8V=p3kMR zBE$CfkI^b24y&tTWupi70T|Q7`I{-T2o%Q#wYFIY5h>!q5%OQttv5mRbV;?Y#rwMc z$Y+r*!mL6n%XLl{d%M0D(Sj>t?OInZuWB{G{u2goe8hu9g_NUf28=3&0d*!ctAU43 z<~6EP=0;@fIJz;6J%3U!@&OLQJdXp$Yo~WKo%b@yVXmj|Vw?d?JaR-K!?TzlTRjBI z!qORsl-lXt$>}4Id5_gtI}}N!HzqAtI%j@aRXUnvIF|CYk|< zPNsuSXzb6)QK55@p22WTC&gSpfM(zwXk?bY&%;^?M2qW0?#(b`TtTP2p0DW2t0*d5 zBvq@6t5Qm>41bR9Bx+R|leX|SQ;$>7P5NPuVa-o;&%2_?9Z_gc5tj5LTAsmA-191;X2!IL! z`L*7i^&?1*98Y|*h39;@YdDMUW@feS19EnNsmIB4IydkO4EyvFr(QL}~~x2?Yq4 zf6|eIyo}|XQq6M#nhI5WR))~%2^Cag422)iwE?yWwWoYK{F#;+Q0v{l29Q)IL7;#- z^`j;iTlaPA9s=db4S(}4n2DaVF6f<7qF6gVJWl^#u zEdgkl6iDKWG*P9si!5kZzN7{dCY4}_j@3)-^i5_sm6xmm!_EL^{ro|w9+*2P#pmfQ zPz0HiBF{sSNvlDrJ%tdOqMSYdj?xSMmHj78)Ft?A>3I*^`KGVNn?{zN$u(El^br5; zX{n4=(F%-I#?9~ZoXp9l>n5u`2`VtlFFQR+w}D^pQg_R|?j)g6HF>`8A;o}#I4<-5 zw!E9lF)Q7vMT^U>r0w@trt6B9rX4E1I&^_M3~8PX0n$8E_X|Ztx(;%zv{Z^!|0CA9 zO(Fn=24mZ`ZEUla2=~)JB?D_nqJAD2@a;MVh5T%d+$>bV_oZuvMcMoNTV3u_x#_-_ zn@GU&EL~Qa9R%w)>iUY+>IKm4a;{pB*HwIh=(wxjrn>{8 z1+_xACUJbtX{TikLt7&Oozyuf0{}~COJ(!CR>Lq3oye|}+ye4Gp|XL0^$`(ema0~U zsmmTkI@F`Hxif{}CU2teltH-HYnXR9k3LX3tkKy0dCK<{Osx(aVavn+dxs0bLA@mM&>g3y^_y3dSd~BkZZbMo5QdP(S{YZ}QBhw*PQ0>E5DsiY${DvWS6X2UM5kWES@m(%M%48qC>?5_m=s|y#>ETxv7Md$8Xl`ml z$?3uHm;~s|r@ty@jUqmV{`&?j0$1UsjcZ)GWP4IyI!$n(mPx88Ft8E52@pF?>tIJd z#6p*XP>k!e|Ff=4ut@y-Okr(1&vQpR9Efm5NU7vEia3K8S`<~AAFV5sgT*Ns+-oY! zK-~E=RnPnz$@7(|mrUCAIK(BI|<9?0j z_b9ahg5A?D-miWM8XL@awJ5Q!LfX5Xt`|F~lv&$*d4Qy|KENCh9@1}RrFX#7)TcCs zXH7TIb!Gw$h-JGVs9zZsb^-T*x?tGXix%`IeE}iZki3WoAfoKJURPPy^jrVXVd%(L zP$gUtReku0sJKj+Z1H==sJmFD(vLDRAgaiVw#VHChLGl8^?#Vx^=<_A>)LRZXP)J6 zt~ng>(OSFHfKB)oj3xC=E*y-y!Q!5!25yn(w-%5A84y^1m>&a;hma=~hG#%oc2v|oo#t+`rXS+cAi z=E>YgjSXLE#@-z2i=C=2jT4i$;app>zvlz;4xhW#8-01E{mq7I?28HVa0}CuviI z0aU`H?^>y7jnr*7&-T$dnJGi>{W5;-hgu(DHq1WTp$F-QW!{rdB<4Em;eWgic}C_J z#hVb@aE;*}TkUW6UIk-y_2WCYOk2g-p+RZ!vp_O?jnB_c8$aLrX!pv3TW+yyqm@|?2pG#W`ow4);h#ZXcW^#)?wG7y@>{AzsW_M-35Rs z_SYD|(dWol<xa6(kMowU-!_b`XC^2V)qeECOjjnhzQ^)k7Ss|sTZU~k-q&xYE4~BY!6NH*ClkbepMCf@^8;zS!SB?4 zB?Z>jdT$jbe}4_l<-N!3w{5a|hIMF=e<4u2`Vjj6@i7o|)W5w!Q=jOaL!R#MYrXoV zht7ro=1T>xv?yI;?tu&Cm~ENsdN{2lPj{jDjLwl;=D&lqX9$?sAt)*5n?&g%VzRor zngVA*?F8~|F8XX%FWq{k4n`6HYN|y4Z6**|5?eir=bG=lp9>vq7e&x5bz;HIAB6Pk zA6$zCqHwh}3dfGm5u4mVc!Mpyf}-Q84L9B|$I;@EnE+BN5MEO;o@?b1NCJ7va>xJDL$wg#dNYavGzCc@HHpOM_UC8;S2=67s_X*;9s~Z z1vS?#lcwl7+4Gl5`{?bsq-!b6P}za=@*7-@xJtMXptXSd9Ee0s0bPu|G05fs;4F-j z3w18E_GkO+M)s^b9+JnG-FJ9goWKDTfGj-yv<~EiN*Pg$*kTVpsV!fo_ zpa-LBt^NY9BONPSz4Ywh6c^{THd5gF;CB&aqNuC5UKEN^avIvGXRIz@bm?uHt`1LP z-Csnh`#-oa*jG4ea_ZcgN&YX8n-6ryKTZ7WJ$t6=VnMlA%&8biFX?wY;^g`q z=Q$Yf*NrzTNs;ye_)rsy5r3->LjHhg3Bm92MHr?n=t?%KKzW6#KFbzMP6nL1lsp{} z?v|pi)N*}$j_sovn?>ub3hw*O|)*zs;R1TxWIcvPls!G)B zT0S<+(i&+|eD#8!KW8la=z*0#;=ga6qxU)smvr4jM%;0)HPhoadE%)|r7nuvw_tqj z0<)GOX=k!d${L494i__$F4`M8zJC!}`TXon zu7)BhVWZn~owjyRAFUQZs)G>ttZAwv4ZyY(tBXeZl*)#^MT~IzITSkhABLMEpmsLF zF3bj~ubdJ+iqde4(&9Ii(%wd_wk;a^U~@yz*&AiDQM`u^n*qi|a{b~Xk`B!w!q+|{ zVb~5Lyfr8 zoa{-?-+xvW_zIUVQm?m16ay=`M-`fnljWrGLgi;vGE2Ofzp!>*OFl+pzyt z-VsTvDSer*Bk;G+wSkQRYIxol-t_j?d=RcC?z~2khhl1|PvoX&d7G8?Lch8sn_v)T zoHl>5JarlTHC5|3Ap|=bBn-$aPvH|NOZtZMl30;z5(D=2yMMNXzZ`vdfavz!w$lrr z5hOx1I5)HQi)R2$P9+|eLrdq&<(nmaMvfPI_kLd2__}!xYYL?xtIhoKw&n)NprCY2 zTC7#h3#*abrEj%qk*2PjQbg*V$te1>LQq{-dN5QuHsqOqGs8xW&2 zdiTmm0UM|Uu@9DIgzAAM6799Ixy-E~fV&MG6?Hu4FNZybVM*@l)NuO-u)h7u-wq-_ zK|zqZW`h!x7vQ$tALxAFu@I-`j42kwdxYbUps~oKRz98jbmUQ@3@#EgTdYJ~x>VDb@|g2a=aAI|R38 zW83bSJ3#^OxVE^L6Y{WU*wQoMr~0;cHM(5&JIk>2>7aLuCsBtn(;5kiqmeeSV-oqh zV3FMHDG|9y@1J?Xc`=0fhjUs6jvn&z6J_JOS4C*w-K}Eos%NY{z&N-cdYn7p3zeR= zcR9GtoBJ0`Dc|VE1pd=)d+HKKkD@L-W;y?3UTm@nil$p%%jiI-GuUhCXevr@WY%?? z)ZP$T>=aRY1oVK`Cj!h!Y+TmxPNG*!;D%X(JzeeAK`G=%jSMA}X8Eb&3ToZ| zXC@q|!70B!y{pK__EA5UGKHWo)i9aQGm(N1TiSK4^u6picpGE|KHQpazrO)cZ3y~) z`3Y%hnDymX&y}U-@xsNiWpu1qdoy65#JJ}Z@d?{O%-LUmyuC54+1Uc0=rEmz^=8O>;qAEE-FzdeYxpY1dV}k|82GBHgO7R(iYjM8em_ zC5zrC95NZycJ(bHM8$|^xh)+V8zZ%Wa$kP(1*p`H+dR(M6p1PWnI6k-BP4J z!u>`Nl;oLL`BL$#a_7~u8YtLCK3eaeZIS21(L3#n)-V&-lOb@QcQz*y$8@DjNcO9Q zLXSX+xJJI(R7k+oPaK5mVP{oEX#YV-C&=?Vu0mJkxF2*>I5x`x4BPjZYY)`E&f z**ktrimxUKp!^l*JI6wr*E;P|uU%2&CaNq=_LZDkvuGfy+ezGdzU#Fv0z36-3}l67 zqH5)39Do)cPw;y828*jmkVb4xy` zl5tFDF5%Pjq3p&mDj+fO0b)H1j6jrZ*EXPge1d52kE%R55*VoXGWj4Ol>ZMi)B7v? zOfi7|*l^-WUD2@{5ihB|cY4@IC(W>GkG%AJsGsCn=1rW7)I~_C##j4$+6E=t=4G7N zfj{^C5bC88EJEZyXoCpyFlFlmmdjMxk?B@+3xsP)#tXFt&8dYg)1_3e`9^FUCY7Au z{&oxz&}}wbQj%Nwpq1_)3ppgiA8gp%TFFGunAqhOH(#c&oqo(~T5u)Z-7$$U7^31{ z7>`&es-W-Z851HOFjYzJIUauMyroaLce`KrN)sN0Fe(H(JXL* z5X9R1kTmSdt~tthx%X#9x6R9SB>x^`V$p)iP+Pv+(jn`147CZi1?!6PU#$EP-AA$% zvFh;XdO(;rJ>zlfp=}&_RMggJ-z7v7`LDG_+6*Bc(IK--wkghN)SGZ&W?H*I7}B*z zGLI8)YjaZ~HolsH0!S)*)O?HSRqKwu(;Bvq387!3O{ur>wc*cEK7wsxf6|-EpP)F9 z$c1_dK_0TW1_F8?A`GY0$v*GScr=h?E(SJ^ls59!Aa&(gRK_@jr+TGwi@GKk+92C$ zJLB5$Gk$Hp3jea(Jp!9Z)gNRfbVy}Uy6O4L^K%A@N*9T2`{Z>g8d0x!r-Lr|89j`n zCV3_HT_jK7*^jpOI2e2NU;<-GGM|5ht9Bd_5x14z0dSvD&q&EY=!H!{PIq`cw3`vu zZ3xQrT*?1D-q7g^_T43;92SVYp`Bh$Z;1A6=O!n+`ww*s_c59`>#U;_A zMa!1jw}0B|iV873+S4f(>m_7ls84-o~H70QJ zI*QIzq_6^(_V7eXd{@>|YLe`#6VnNd?%Z1Se#au1s$`IQpB#Oaa#k#t2H@zs#Lm!Y zpC>-FtvaBDIUC5YgOtdDPc6rwr47lM&ly3HdcXl}^ll?#h*4E~1lDS{0k8hai>9vL zUf+Y-7T)+-oq(L+E%$t-YN8C=Fh(_PnuG=yRF%*)FolXpEv zJ*|MDAsuBM7z5C6nb`REKoD#<5~(k+X}6_ap!Dlp$|uY(`FlU8NGMuOL5G7?Y{#tYo`3S4;2ux=|fL9!74f= zOJNl~a;~cAmfO5thENhg@=;LH@{rkb&u1ekHw;zjnSO#M8#^n=ec(&WLld`RkVW(C za;;EIc!3>!yR|yYaG$AtLVwKx?LSht*FgKON)AA(E7uh1O1As#<0N(QD`)NEH6dc( z$#m;UyRYPx=Sh9LM7@Or0(TJMG+{wNT`_8X>V*7;}SnwvNI2br=aZkExz6Ln+1Ub@e{?T1fB zfTma{-+=uEq|k3teDcGk&UFb_1g7*AKr&{f1M@ybTfLH(f;~n*Qr<;hY#{OZeQg@C zEanpNeg--j5BOku@2%%hu-7o$H?OesdMn=+{6mHN$4#hQ zxKJS~4Cs&b*B`Mu92OC&jeZY$l6MZ^aO_A_QV>8gzVV}yzuGi80=EO**UIX(hvK)Y348iIBnO`*a zG7Vc6wPaEs6%@92ipc>ZSH%89tCN2qXq(ot>JtM!mxn%kImxX#?PILPq@>U@go^Uq z87Hea5AgU)46dU)<6(?QYuDMVsEjjZknVn^!!E<e|sKzDL}Rd&N@xBIzZHa({oTk5!caKN?=M#hv1?SqCL75DwC3tO$PHv6+`XGmSt zs+`GAs|j12Xw<5?t@l4CYe{_lshnHv@?E!9QGKDgDV)P&Lnka;onYc@dbQy4TTTyN zoLgb$<&xygNdyzn5@QRVTO5ExhoSFzWFgdNOnruRTuq9OqTTUj>65(ow1_XG%xeFo zV<=dgjq>`GJDu@cGELWfW41r+>Y&E#Bq5L*(Jbc`dN4}71ZCyC{*TF-$Dl^nws1y- z6WB2I;KolQ{!36##D>^E_Or~@d!;$5XpuC0Zgyg$1f6saRtcrwFxvEU&67Vf-J+K+ zHZ$rXjwpEaj4z+!^n#b?gCu17su3qU$i7j#Gru(HafkdcIuG89!zWiQhI%*}oUfUH zaP@kKql*iQo~IzQDD{e6&6+w@%tngN9@`>lu;pL-(rCk^)~&aqMRUtrS1I}F1cMHHWkUW5-_f;Z zBmLMZuvU;!&v-6M4hB9#YD6YC>IY-3-dmJRmNl{TgGIUg$D%+a;uJ(wU}|snn(ffM zDz`co6<6<>N-)^ov=nC1d~wa;>=1;ZCZZ;#w{wzjZacS&otn`;+yn>Qp~1P}n_+J2 zpyWeiVT2McqovITY)+~TTq(yqsFRp$Gm~4M;*9cI2r1ew@9d0v7>3EXCZO7+-W~ax zdOP>gcVk#^YK>no=fIwrp(B~L!tv&e=U)2KqzXf(0v`Tp86$H!4^&|M3Iro=sZLq4 z6GF$nx^_?vUz!;-5$h|EAGz(inclh37Dl-n0aVJxF6t{>z}9ovX$3d~Z~>W-3-}!x z^A4}Zols%p0~Y>N+-^x8fJASrjqsRUJ_^Hk-{cxGNz;wR*^n`T1ZBKB`skaz`V z+p01Um3-D_*6~3E#hdo{RZ#h{OaO7)@97_t`X;JHF?|n|eR6V>0sutM|C?R#el`u; zBgYHleh{`aO9_cjP8|rwm;j4)RsSpwyM4I%DH~y33ZGW`U1KPzCMv6434xkJ7K}n! zw1UTng0u94L*>7SAS)JE9ma`KG37bq=HmW~$CbBiRTX=HM>h?|o)V5te2Tup<-l@j zmloTxy|kFUeu5%P6S(hDR6FQ-8VKPizU1$d(Rvn9r~&g~qS z*?5{Hdg~4bb8O)>N#^CLi=&^>me1G{u}W_xFEwNCVjk?)^5x{WPvO!eev1F~^l0q9 zt#g?7Dwj;v&i=@J{m^fAT4t19M?`<;GZqpJ7yqVqP1NSSpW}DfJj9Hh$LGAx)npCZ zLNEN?2Ogw=YT;gr#!ip|hBxB5+pc_a-RYx#ed-}T$NQLRSH^DbCrVb)7?#K6J0=7U zd=CHBjE{r66i`nt{`JpJ#_X=g(Q0=K#5*=6a+6fN6R}N6-!f>-y2^QlQmtq<-(W-5 zuBpa&vlojG28;Pg=CC;5du$`V^5dc+>J6rRdV@~z=q@}r*B4H_)urTVz4rHC1ANcF z!JHR;nzh5`E{2vzizM?wpn*i281KV`);_%fUn$ldjyX&N>TOzQqW7Ofn{Cz!Rv2K? z5&8C#5S^;_+MQ2eqZ#qXJNEn&8H%XWlF3#X*|w;iEH7UFwfJW?N$G8l(?o6G3#;fS zcM-*OdCS7UM(v=@$X9P6N1Ncc;|-q08}iqq-(dC;Xkd5mVBV|29v%5ksO0w9;(V57 z=Rm-jYDSUpy{s|M{jb!xY|;s~+i1(;{)f+SCVZ<^U!PS% zRi4Y(3HbPK0-N4mw3=nMCJM!T`zLxtSXCvt{{S9?kuSy-oaws{4H4pG6yoHfj<@Untq%#bB5fQ7z$cMX0A79z4_O zyjv56Q->$FXMK7*N>gz(!~7IUl%F}2fo2{@E}WW5WI0}B`A-MhUwnLr_Th7EnrgJ2 zO!&uJrn2RtG4D;_&^{YreLE2F=@%A2M?~V;p@D~+&l;WFrHNwO8q2onhBoo~DNSOl z#WVG?Xl%S8YyvZE!tP(2fR8Wk47)~yhx-PfQT!$rD}ACo+^-6@U|XePzWOyh+lU`- z7v667X5S59m!|&QIryj}JA7)nyEPte#jD)>Dvpm2A5D}pEcpcIBIog8;z2lehAxH! zNyaNv;cU0u3pT>WG2$!Z`~P#}@G1wVp5WaMlHPv#!P9gPZ@Z69m~GRfHu0)$g0zy? zF66RrSidr``i~U)|62X7M6C7X9@7(F>D0Mk@jozL{<@_%9{VOEN#{-A0r+O!|9|yD_IGszqR@ zPpvY%GusHd%xr5qha%~6m;zSupg1W>-^#GpLh*HpEN+jc+7q~C`|tmK%QSEAL95Zc zaX=I6pZlwC3*RR#wk;7G?|3j6`EQy1tRhDnGeua3Ea=c`8xJJLKLGoF^)LH=CR$`> z#rbtKy!>(u<{JIQ^juvD5<@iR#%?8e0YMT2Pp{U-nENT1K8?CipM+*<`1~KR z{29w;t5O|WeEiu$c#H1}TEw!F5PDzyAfA7w^fpC{`27+(5gVft z(o~y6Ostt&=7l+o@Gf||x-#D2)D|hUTITvuqq~t7opFli{GESWiW0qGFt2l%eJ7(4 zr*w{Fyi@QeCRwvU*&KFk&C0}M`-tBebwvJNwa0G%a)&|q9EZ)u#;@^wyc>1Wj~$_# zkaELWW%HNU5$>xa{_xvxlgxydtSEB~DbZ`VG_EG@^X?!%c;zB$R-<>z!ke@NUtvuY zJ9Pk{3S|Y-<%)DfBE-5m3!LF|BV9ZnNPi55-j?~TP44+id4FynJJRAY$dDW}8Z(*7Y` zj+wX%PDwU}2|V)JMTdWvNPOHpH*n_584hrMbOI6V3jHWq6p(mZHIM_a!gIUw6KVK4j9yLS#4 z?|A{%=n^%GP$tO))2H!WgRr70K(%EHy!A|fz0DhlU;dHBUH)~uP9dA3aaor@%DXxWHaKJ|j2Dzh=^8)`oQZ)h6GNC!^JP{#oy+4krRML`0WB}aaO$q z>NNpRk~W)Ax(V^6JUUKhDGPf1Qk&qEiC;uqh!$~APL)FPx17Le40i_2`U!;yRr#Y{ zuo{hBw8$>~jS3{b7F~dN-Gh2|D!}SYo!WGBhIHu{SD^TWkxt)-|}Ux zRi<~CxwzmPtNZh^;mmyZL;(u+DCf$I!R>xROSe<2hKVcv#({mWytGhv=_M6s<_tQp zqqlg_#K1TEmpI9>Sh~Z!iMjfG4fpV&ui1XZ!0UIO1&`77x;?MK6ZF-g5E~(9&1erM zOnJGNz&5fCZ*an$|J6Arr*GTU+0>CwOMQ+>(eW#*;%ttwh$7r2KtNg#(X2bTY5^VM zptra8>}Xrw>dMvS$JGfdFNfZ)%(bj6_H`s16`d?`Tu=)CIbIay750H?I9XVSc=moB z^AY3asVLvx*;To?gSN}z^9Mwns-+MseC&1{Cek3~`eu!!tvc!I(#0uzeZtY&Db_k2 zD$@=f7F+SQJ2w4sbAep-2`e6M7nTe1rze}E_P$uKy)neU*f^0ouxmzswUu?gObS0X z)8#P3h9Su4q?>n(+@G7Y4tT^nkt$zZD^cM@_Nb|m9Ug(5ik^Cjw>@XVt}=|%t{5Hl zO||m4wY|AT5*Ht?o9?ckcw1|ByKmW9x+9{yz*X$3#l^R6nG!7GJvH4g@2#~k*%uYe za3SJDB~A<^;ta7_Uh@ry!!SK<4udHvqB`j+-P%D5{=6&IU3-r3&W2Aq8rtfb1Vz{lgte>d87(AQTMdzWS^Lx?p zUO{_VN}sFDqYY(*oiIr=Uta!L9XK^LtGBRA+b^eM!9kc|b#5|UmvL3~$)}K5jM#~M zlbUVC6O(;{LVYVLFX!4mNOCnhk=<0C%D5qP+O7_7;WN!E-{PQEe3jD|-y(MT^weEX zw8+y2jBb!theOtNxG-CM|70_Po+zH8ogoV%uUL4xj=!A~>3Oc+hf_Rrp$wPV%6hTZ zh^iEU(6AlgA%V zVTUE1Wa4jD)9#1M@VXEV_*Kc!8k;xh-FDI9HoV`|w)Dne-%K4%8Dj}xPmJHd{3qU) znj$e&rcd-;knL;i#sAPLs`_bB{zz-Mae`?P9XC_3)bGD+?WBD3_;gGNdBHrEt#HYt zJ=VmG7$WGfj9JbvKHE@YR5btC@78dCSiZ)xX1Zm@i7qGdE!9G=ba+logBhND%?-U9 zZ z!;J@3Bk+xdQFrSXEbpjEC6_lJMa5=Z9~kFxqPfzljUr9&s=R1gQ7bN(FOjv&Ek*bF zH4DFQBgqcr18$Kbb(VfcVl}Y(HVYR>;ZM=&x+9+;&bI$%c>N$J=|!-k&N+XS#sZ6m zyZXaDzNb#k$sL53$>^pdh4=iJ5`7|?C%Et`*!699x7l#sxPHRf$pc?xszGa)ZDOmuHmRO? zB>!9J^DjcUo4mN78C+VcwR%d_<1?yulhKb}8vHyxO2DeR4tI%kw&hR!2&tl*ZZdmc z8gRC8yL>Gr-0gsN`+IX1*mBs4cn)7L^G!b=Ot+VBNW@;fgR)N`Yoqw}enhuNiHeD4JsrdATJp$tx#IKZM!A=P`UzyM#UD}c6!x0Q z_(eYxx7Q#YF+E1td&TEUp<8!Z=$p5#S@D+-6(-8Aef+8TGD|;?^_@6RVn}YejFH|> z$hx~x6D9r6RU5D8K^RkI9m?gRdci{s(e#!SpFTxW@3!K71KYN`7Gl3-3+G{IKb-87 z_tPur8G44h&C8LdL5!K%eBQ{@*8geIiE9;S)H_m{b<&l(gZr1>`cZ_R@{8)a^Gr)p zj{$dg5Z&RUAC*ToacQ7vEVE4hdY_KVv4E97jozo)vTkt6VLt@UP#p57=sMPH? zCOM1U_c&rvxb#UuwR1b95g|7VowXCbY9whp;cj~$I}q|C4fAgPsG~H6XR}gzy7>dI zh|?tdi?T2WMkvaLym)P(iJ;DnBJ8 z_0{zr&-}E%Fys=Kh~?jwI<|imSD}o`1cnc&_8gxZzI@i+lGpOnCG1!A<>H)x?TJjuz6{=p?{rKpCY+~1;qn$B*af=s747I)PeEv*^ za;7W0qcP+P%v^1a5{iefr=367?Jj0tUhw#i^z>#n&TYsAm4mQHylnXzTfd4Hs^{Hn z?{p5$A#{mAfLBcdZ(!4@QUBiJ%QUllSLSlfvIVLxqKx*)<#_ZG+=l&0=6KYLYmV)n zf&J*l4}F$@oeVeL`!R-Z2&%A$?ZyvQlm#GrO!#4V0)0bo&!y?`TFfUp-z$}SPCtQ4 zl3tM$evpr5Iry;BFy*0X5|E(f2m{U|!jtNI1x zO|q}7ol$rbKMvo{fd+zkfw<&9I_X;_O1+M;l4PFa$+x}?Yn>vg{xRHZwo+ibcVJ?C z&LJGOTRh{$1yn=ASBE6+m#wqH3))2qMXm<#abq*jkMHk7rzsptdi8_Qf}b#Enbvf4 zTiMm|$=i>isIhi~WTZ&fA3LE+V(8;*$$(m2KR9?X7WOSSE~?pvks=!2nx~iF^((wJ zHvD)@de&i^{LcIJPkk}(ug)u?OwL5FzSVz`jPZ(*9X#f^89ju~sFs?}viW&XX+LQ* zy2>2pV%xYHz)qu^C0r^DjFqejSS zqwbpE*q^WrGSTUz8{B1^w9J7rPH};Y*3<{6OYGhc5!SCVPOT@3D>CHFb?r^=ePM6C zR)s`Wuu?A>W7U^?W?ff37q}RBy*>Rt>z35c9rpWF5Z=ol!OK0X6^IY%CbYbtc&mxY z)tuZg7h67cqDYB8y5T_JhlG!dx^*AE}JUWvwt+4D&} z)ga~X@cKmZf2TaosIyrU_2kg6m?Kjx8txNA0S2glat}cp*Grd(MCpet zPj`%WHClT5Zwx8?rWcSkBx_w;k>r{)1zFD%51|iekgd8>|u3!|_tcc#LiUVIowHg~Mri6)|V7so3cWP;q&Ows?`kLI$=im2^@zmin> z#CUo$;v;##%0Tp{T#NHkv=26D*}@4Te8COri(bB3js1chl*iruF>ZD<&#W)eon9w+ zFmA!5qv#Xo!5p=)`>Q=s4qxKDm$mnVWcO6NZxYM(oRM^rv(m~gmxSd}gXLL#XmKiZ zsM5BF_uA|hG|^?}{!i1HWyp=iHff{tI#+D- zCPYC(qRqi9-d8!$FWZu1oHS6?(qA+G`b(r%jpPlWPt9-D|z~;;p=r z9*ozpqec488ejfo7$BY)pY~UUchKIH_-*?aK2;nXcT&B}`+)1`Bk(gQDU_7Fh4J$l zr3P*8FJoU++&OAAtnU5H^7dMt8J+b`=!uzLQJYWO-S#BaR^a0JxlxwR8`pxYuCJoU z-($=cz*BNBj|IuE+8bWJc~w*GKF#uU7@ke$*a^rSt?YNsWiisYC6YJJ??m^?80X4B zCC&nppg+||@K^f5BA>LBqR)BxJ@Dkdq7reQZeJwci}~!EHsLJ;;y15U%93oKS%+$vTNvKy81qewrS4$6^~ir^x{pUmLYwD?&J`^sl#y%5 zD2pi`i;;{8Y1&bw^JvN`(@XX6AmnJDmPK;FW0b!gM-xRP+rIp`Xh*NcTj0isw{r4) zuwcW8XFIU9E^z2pcLSy>vccTsPM^?vWGj@97GXTR8tRkS{`l2gtxeF-ts!*l&L`&+ zuADDKGqYaa>iW!LK1A}smAChAc_j}-h_Wnu9F2W<@Fu>yXhu-WNK%LBLeCjwa&C58 z%#Hos;x|_%R#fDyhbtK5lzyVObNxTw-aHVhzkMIKlU<1@l1f^rFbXM4BxQ*iyDVvB zP_mYNZ$X$>jyMBOk)w1jD0N@nEiJ<}RCXMiI){7FSRN`)JkHl2*(=nA#Tc2G6K_dh6mx#IOA zU2)@{-Iv*9o725=5EM4OnlaD|d#IP1EWaP4d`M2IPwCsiO{VWlP$w-Fgq`H0@6R3_ z-VoK%C|R*m>C@JX54boCNx~5G-gt|1?$;0P`|t%rUK#J2CmQQozwZv%sHu~hEn7|S zu*|0?5jPAeA*^=qP3G(x28?W@_Ja_}IBi8~*t_ zS+40T&pz@NK0b_24&~B|a*3)DQ9o+4%hVGk)F2}(uM_f0dm|=VWJcF+$GyzM#$;9P zcBwLb)Gh6)N1|fW4F<@)4E?s6&HAlreCtZ)3SAkX<{Ad!ZuZ?kyN#fmaeK}a@STZ! z1}n+xD6dVFj?=$abZ#{QN(rI|Ok zj(z9uKCN~hr~RVAh`L(mg&dx{o9dHyjB@=Lzeg(GPVBd}LCXe`S|1;Lp!70#S0C}~ zX99)YLgv=DJ<40YxZH4^9QoMqYsjy*5#X7`eElR7rOc=BzSPy2=!|8U{AxAvg7fYE zTLSlPrf+-iQ0%O`{c_s#s;QlqQL{0HXkpf-yM|3^C4$Gu`Wox{c()IIj@tsY)Jz=d zlqTi(%5M&;*UXjeU%!a^z0z_$S~hsYzyZH>9Lp+K-I(g9X*+?nX|xUhp0|$FPmBw5 zJS#}TUJ3_V=?OE%Zu?qW67lxa!{hlir2#9Mr*DaD+dga}K4hBC66Ar~bXI07rsz+m z|F*Citv-JHXT4u4{ddR5<0}ll>7oKfZ{+q%-3BMZa2lyBctx}Fp%EvepG|J0#)-o7 z&*TD7PItK_hqoq(kHtx?H?)lZLjX8gZnB=gFd6AX@(6A!t;@TO0_1#kmYwD$%_lC$ZN2jwKD~eP=nc;tnFlr~nxF_A& zS()N>cjRl%8TWMau9X6UmZ~DgjDtd&_WpNYeT-B|w?b|8omQYIC!hY1ZnZjkCb?k3 zDq|q>UTj^Od=wL?C8co zb`miON~KsRmEN2^>m8?Qc(d~N`E7*X<5A5+=kM?qax%-N>%wvGBl0nCy!CyHIJV~u zjv?`mtH=0P9*u_6<-hs{to*0GOrH5>ccAjmW~;{mk|d(2B@ut!u}|-OV7FEN{=&08 zvUNhgf4si;g@2@C;`<$@D}w?JZgSpl)?KaXONj`<$*hrT{9}k3AwD{knp)`FCb30@ z;>7T#emzbHRpopb<#k!+ND#zpdf?LMv|(0?yG%}6aQ9A)wP(sB58rgpsWo`;5uL34 z9sRunU3>I;?oY+S_kK3#Lae9Pm&@+otTVWtIL1(OY|DWtE3ITqplnxVYO^j8@yGY^T{bDJvs0@gXKb2XYoB*gDqBn z{*2{KjGen*gHS?ls&Gk-D@SL6G+P~TZ2LAOsdGCDwHdnVlSLF(v}ml*0un?YFfKKt zFObRC{invq>Sq6-;twfb*3*gF{^De|@6`BC%Ynu_{a1OzZ(JC5%Ozp8e`E&eBs-nQ z=y(K4gxzW3nVp{cVW-uXIeX`e!l(TbR=gGX$b$)TS1tRCb&Ut|T5PgkV(J~+Ejvwt z%TA+XG+i6i4BVZOsq&+M<1Yo^1n|K-YdBu#@O8VqY^+h?<#OK&U~{*{T{LdsAAR%Z zkC^KCNi;XA^}4W%Ue=xCg?2pw0e`0SIyYL5?kJ6gh^fmwk%R!yvJbc|Bm11A-I8$f z^!9^1`Dyr`WHgFL*zbr}&?eX&IiI~Q79Y!v1jaE9YqJE}Nw|JdCb>zz~Rg_x`bNsujr2gw@ zRD>Bbyd5R1tkXb28lKgaWpY@11>LRpZvESGd6>2*#3~q{J+iNcb(0FJICv%zubz$5 zysq*}*^r3W{^0w9(km`iUKN^yRid2C*gZS zE%2^+iK$zGa5y1f<$2nXbbb@De=h+`6fpRqw8^-~4|!_iTTrR*XHX7*nwo=6RIBeg zrf`#V7{+H|wU1l zJ!8E+5SGUdz64-suDc9n2x(^AY%EJi!5O{8jCtM3#(Q(Q%U}D@ctGoE6qe7~+ks-- z;Pky%VvVzRk~H^wPjjBLHNW&JoVr=O<5kimmEZixj%(#-fU8_e#8awDWSdj7TSu`& z>xYWjnQp(>H5G`wv~YN7@BC1Vs7p#x5r;`>!bfJG=J#%rDy>4l<}1(Zf1NW4!maQv ze>7?`E6miBogL^a)0`x0t56k3l7jTjAH$wPR@YBX#!ZphScCh*5(hOlfEa;`Hw5j~ z027+@!RpKY&YAHo4G;E3<6LKMUBBI8JoBSPuae-P?jT$_ophh(1!$Q4Gc(?DM=lMYe95IB@E6y?U(63qdmT=yDVGO1Qi|ArS*`|m zEF;ydPx9eF-e{w5)sm+$gPn)CABab+$b-b$pHrQC%*1EXW#TTjh2)o23=ahOPi^_L zQ#$}>AdT?#2T1rQK{u}nT*WuJ{ZR+O^aRS+z5v9Z8#w7;+-*nkzH;?8$wKNOWCzAj zYj2dwK-)zljt^YM!~;Wv#h)g;I|v=tHp{0VqbIzEt()hH?0wmXo;x9rXo;0O6N z+Q%7GHM3)67xUrIO6f5i{12+`CFBNP7#amKPxyeoj{K$rg#LAcFD!>Uj184b1Q`d9 zM=BkJV6!O}9!R#MK&_sm;aVcDO~GYOSEYP6#V4Q}B?6g6_hD z={&v1y$T&YoN@r8yV-=(5MhYq0nRPQnmu>pd^NgN2?~T2K!3+1=-4-ws2Y1hDEmo51c(9 zXBX@t_ej{i+g@xhY%M4ON0ujZm-TGXYAEg_m{-fq7zu-v`^LS-yMhZq04v1~W^JRu zW3k+%V@TorchIgMk3YcgaKN?`Xx2{bw7lK36XL==TzRK#(W z#$KNE2s(_c>a$ocErU&apcZ)Exx}<%`)DILr7+|b->^B&?53=|^HRPP0egzvz5jUa zO|~I56^h(NM0iJBb?_KJajqbKV;TdKrXiv-slnxcx!RK37!7mc?Gxcs2c6xskCLy$X&%Hu?m=|hP-3OkZ*z?{WU zO<%6bK@u{ns0s@uo(viXNB-@s7e271yFm{;58HCi8<;Pq`N4CmsHN{joCDlO((2&- z31c>~@hq5JA;l{HOlFFO0w1?XewC{*7@!`i(JG-vLj+-Bw(W~DE zX8?zWM2HFp5YPvFH92B@wi`0(lbKeYdB5TC^^T7!C?PK(lbFO2yV{gID9RJT~GmCKfwh%bGL$yLP9oZkSiiz^aV?J~EwE~BkYOe^uJG-WCblmOvo{)JWLF0tSFzxo_XpO;;YQDJU~-L~HJK0n$^bSeWa04@!1RQ;8!xKtnRA0s zjAc7kkDkqMur$n#2Y;jI<&IsyvlYPkBwKVxE=z)dwLJxQ!2xlM{XbR`426?hvpTWw z6#W?3{N%D2W8@gf;C9PTd~7ofXAmqVDl|G2ZhWa*ZjDMhvUak2qjRx>yoiq_Paph= zuma5Ow*5aExZ$psSmi_!mmr(F{E7g@qtxhyK}l^Udb_^*K{*R_(1}5o*$YyHG?#ut zvo_vn731Kg7+CEcaONZz&YYDFWB71}3Tj1Y1qM~cu{-d1`w$)nz8I>nLWVS7#F?;H9ZuhYcTxGi$^(Yp+D9qN2^f410sWMu!_6 zeRipzM04zA2yW%NnGBrnOWtwZB)2Qg%VG$S;4sDwXv3)EJ?-$q3_!1t`s=XigHB`i zDpSb=F2rxPD=<50@C0T|B|PzJ0={Hq7u>#jcX*FI{Fx$%-DZ=HDe`pu=teqVR;d9q za6WCpnj`5;%e~nq;qoO^2IVy3rlBv>>*#ZLM#63)lkZzbc)0H?w zXCUxiNrUZd?%QTBttre*xVvoG`$v=OTMp@gq2}}|cy3#GYpDMmGZRYQz+KZ%Ke%{k z8!Z*zT!QJfRN_kiLIzGxXhlHWgQJ($orC~&4hc}cC*yk*!i3p z92L|fu0!lkdQjE2f`An+`%nlcBn?i8#j+CuhM&v+Mku|k!`LfF`#f&~vgdPB|9(FW z-Dq|s@zco+xGZ>UY3~UV+Z;Zih4p(F@+}cQLG&ljVJ8459%9j&i9!_4R3W3n(g`8m zt~PMt3PVa=!ay&YkU0?(axSo*Im*sEGVcS#yJEfe#(LC!e0sC3s-<2Dh4&Z~#8Hb* zTxxisnvkgpRd?DHv~s}1E94)(fs_{CBhO!n(?s!vu0Q1bFt0lJ0gJb-!;-MLG^HV@ z3~t}r_}K~cq0$KOV$L~W@di4@Hy;ZVuq6f4^yh8dxH~k z_-4d?j75(HX`o7fbsz3BKK)BpbkmW(O#sMj09pMSqjE4Eh)eFBPos_sl9C)TV@JJf zq&x)$ndZZ)+tI4AoYE81?| zTSp7)k+@evg_@-X&9a60o`FzUEU~463HM;)i@81rq+cK)1E18l#p_X5R4zO})rO1L zCo>K`Yv@yCi23yjaX4}<-ut52f8K_ig8_@fasJ&xy`8lVZ1c`gczdNXa?Z}lu{rUz z&RNLcfZR%>w1uV+3acX8o|)Qoekc-aJHxzK?-Ye~BZ zwb!Hq_TX2bA*L-ToOVYwJxXH*mFL$%*pPg=Pr596>f~84-*MfBvqy8cE}T7TkRBjm z&xxZ_<1Z7t3uN_C?cq)r{p;`IMvst8OC|YM!LgU7K|uw|voZvVE;wT?DP`BtZg~Km zX<#Ob*#GLBZ4e5K|1N!q@!tXt?c={*JsU@|;*b}{ZO{1zW5;}Hfy8i8y%*u0(ZWMzgx*8^)J*K0kzzK%m+1nXvo4|;W` zAJqFxyO#|+DaBIS@c{Xx7*yN{pXPRF&j^&q(G{N(o!a8vGyg2&!H4M=Yke2Ju7fJ zY%>&t@RtA8U_&(|cVWn{3VbPENZ-y}B?mv(x`?Ob%dl=!LFpiEF0}gRV4~Yx{X@-* zUKm+lF{_FDmYE!|VX1b1xfFcT1ktcR4^K}>Lk#3|d-WfF7NpIMfv?AV3EBaZDtK*=O5w_%YQ6eo70) z_}ciK(pVLP*H0)Pf9l6uvZD=qSx7Z3ALD=vry1X&uL$XKC?yGvwm-T-?D0f)nugQe zH)oEb`tIniM(6I)gvUF}La_^h(s!{x1-9+w3S-pd&XnB#a|5|#+CJkVVkKwxkX8+K ztu6*MnYdyx9aE3z(&!FL##SIjYW?f6A3P!ZfUoSCY_mPEYi;H!6lD2|Pg26m9J*!0UK27)c$Lg2;8W~%{!G<;515}rH@ z8P?}mrcQD5*~OCS8txUbpyXOp<77r5@&afT_g_hlTD1fKJDQP#zt8%-z75JAC!U-h z#nKHFi&gK}Qtn?V-}4<_pM(DQ#on5w{cOmIr6b^3Wthr;x1w+weP%XSM8guyoi|PNSkrVZJ*Ae;_zkRo6s-T+yKeL8 z4Vu(;AyA1i5qjoadEM+Q)lFB?YUP6{o@chI`3+ZQLK_P+O5x!j(;j}o6b?5HHLuX6 zQN`N&%M79;yAV4$6~IRk`m=#U%{LI_V26WVjvVxoPYPTKohNyNvcFad4?ew|INvJ5 zqDt` z6K7T|TXcd>`5_oT=DK_S3Nw6z=q(|MxNs%c=P_qbI5;CeXKuL`-*FLV#^)fIDlNbS zohosAzFbSuI_;W>k5mF^0w#LsHIe0Hf<`JCUQ$2MocEm^oYNpkFge}TJ)?p$&b%5I z;mY?7@XVxwn}rD4FocOz)BNGTL54FJdJmvM^Ihvr2r@doDnL&#k`#7Jv(yV-l{v88 z9f-Q1Rg1BCF-TG0coHt*)!y#5WA>{YAOEUA#95~h zT)_yut$ zCa-*!cFPDyP=Fl7_oY=2m6n+U@GNQ+PEjonhrHaNqrkW!-BgMD0n_Wn#k&qDKBeK8yxYFCFRTCjf87BF|uWV!+S-- z7}V#Q*%Ij~{B9zYR|!u`GV?>I({j!1adzlTd$K{6eA zIV`gW_lS{{Mh{go4Ge&Locih@KKRU&Gf(tUTcx$Kd4JxD%8CatqjZX^wb+A*-Mb-w zoLlf9nI{G4=|3`e+$9KUECT3gDbD{QNS=$N^V^KsGM$av3b%vubV7b#CE*AvBn1ES zv5?exYJXKfC_=RUDkR2J`>WspTwlYB$rWdmBSG`ZIp-AEw{(v0bhD12sx0{ekBf{r zT_^q4g6(HQ+kVTcFtiR=bs*9`S$~HVpNW)E)SfnYfz5gN-k9~P%OF3}(KK+#kGrXD zO0<4q`eP_YN7Qh2E;qjft#aK@*$2xovTG(b8Omj@WEy&%7Y5j^=K7?c1`~8yG$?r+ z<+vmB``)NfYPl%92_|=NMxH4*-?u)cceaMi@_0#R`KxPFMiQJ4(4gh{9i+b+9exLv zAb^SlJmr4vRrdEifFi$!Y_NyJh^d%H;F_wQ#oj<9M_I8R#WJD2&L zO2P?X4WVG1DA#QX(F5*&`K{i#2;#>LJR*B9^+YoUFd@ z$tnk7n#va_qH`ZfvD7;%jqD7=R~(|U??6|~HU7>M!4P^7uEv@!e%Q{%7d29pIIIyX zizH}jzqkhh#vKT`u#sZuv&yJ7jL$P)W2hyNj`O8W`nG{+9iQ^VXwO4%c`APOwv|2CXNER({dSArmckbnaNQE>lCvsCd?x1r^;-?Px zAQ?7=c543SB$1z92_3V@ADLL7sT3hYd2;E_CjUevhpqDuy}$xPuk%Wv88%%W+NJJT zo`{P*s|uav^4G9@1a%Y-G;wxT(|HTzzx@HZ=~M(X;bv|*n7EzpAxp`wjlivf!XkWG z)c_?GPi)JcL@Ezt7J@#Euee;CPjX}(vf{4<@M^sfn6sA@Y*Z`+J%gbaY}2aRbbU4q z(xQwAX^N7J@7pO5y0s(P3_v%{P7e3BIJ|@N21&>l$~wi8uJ?Bl%6|UZuf+1qenl9j zeyR6f3(fZnHS+SD4q6YTMvMr>Y6?F8$8%(_P61*Lookek^KRS zVr=LE5eb6CBMjX*%?kjijEKL5_nj#eg5)$;KY0-PlkP(A-`j*j&Hm~0o>0B^Z*?XP zz*3$A#^TM=1UIjN1VOjj*{SIRa{2LF7%{Ru+Ew1Vw{h8V7whnCHavm!6ge@-F0`tB zPPGID1tBSI6QKdw+<=S%yyJ5yO^N9ddd%t2yEa=Mtji%2eFN}r%n3$+%_Q5wt5iK&;Vz>17WHPol z5&*r_NacDh6;BlEkgb&ThvWWap}}tc1wiFPYvSd+O>)rr@IHZ2WS*LXt#bvMkPD1e z$sJ-Z08ugCSiJJzdK26>wcnh(-69)Ox%A!ev6LNPJz$tNGp`U9Q5)2&Ywcm=pOl5% zx*2jScwgqIMZTcZkAa>eNCfRNAz}N}49MSB8+K(u%r!qDS7vWVOzRUUGa8}?Xv_k= z*R+7SoR@@q7ut6%gSlW2^qS74I6DHiPs_Kdsrl`|orloBtiOcZp-TPmZk9YcQbV*F zQ4t-YHeE{s(V1oPM_I6}U|lhSXZn_PFjKqL5lR?2uq;PG*f*?il5R7tVxCvHO-WnJ zgB=B9*57=9umuZ;?!RmS=Up&c2&qh@M4h5V88LN~G05Bb$7{Q<{B`QC=TNirWj6Qc z@2NUJDVp98#5$oh`~069-PssMieXH#QQoYM8ga<4_`x6uMuX$~oEv?RoepZBE}Zn0 zz#&Foe4b4X&w%|D^*fM=znzP9cHdL!r%I{4O8FWelZwA>_*Hc$FYIzj+AfFSx{QT{ z$L4&~e9-RehYaT3{bFynwW6Jj|T$)ZEy^m^tUMENHFVIjhDQaWy< zbGJ0&#CJ#%`Ts}BYNHD25gJw6Z{V)=Hi#bTy2lPQ})v_a-6P!rPMEF@+wKmCx%J2+l>dVzls`nqrIRQ)6G+9C5~ox~ zf=9AXf{wlUV3m*BW#Gx5J}31mo9xmB@H_f&2cR6=lk;T_2c)3HH>Z=U zy%1|amcM;qghd_l_`=kZ|Ekz_uz1$S+jJwWqIytuz=U0?b6KxrNEVq=aHGfMzG&_t zE2X@si!T--NT`2cMd2I$7O(7-cCO-Z2aWdUr`%*Q8foAhO^3aF8IDiBKog_2I2Z+pQwR~n32|w{ZO4^f z^D@ulWMI?OD47#D7B3jIlMnM(Hf_BTyN)2yI|^||cyK<>2m{-04UQGmrC%T94>;!n zbdz-t|9w{jl_k_MCS=U_6|#jSjaq>7@izbk9_eC94ODMMD zU4{e1EeY)4;-$I4Z30*yF#BAQ%1FYsjk*^Hht9JCaMuNP369&cYMGVZ0UOh*+7mIe z0yI6ab8@}@ua><=9|Mc@%x|O??OO#G53lc5u|7s#p^(*oDyLsD7?|e`G>;MO*T}RS z>Sje#I6eDn@%KxUh*RM(E|4qf3H@PI)?Hb%2$J!j3Pnz93Gn$(T7zG|Ny)`Z9a?~l zaN&(TBvZ@R-&RK#TO2~FL#HMDD}*&F7c@UV&Q>4HscQJDb*~fW2}m$@b_4oN(8-`$ z5a5go0BTwslL)>l^$CnVBsEnG7&~`Q+IaXrZSqCj8=5bz1FWZT;SiZF=%akjh(TNj z3#k150raxCKF_!l;G{WMpI-N13I&sD~j;O3kai<9U>Z!N^v&BGTYrc(X8iNhH zX>r3YpQOc@flDw8?SAl-r(+k0sl3ALKvTBze};A&>|=hyN^%HX(%`l@dJd^>gD}8D zsZl6UAxv;VE?!F)(isXzJXO#I_2On0^9Hu%zXrbsOyWXnW618+k4gBCnOiM}1@v?% zdq9PE3g!G^j&_4JH>*oSW}VqyfG4IJ`TnZQsJx>Q0GUmc%mPkobVB6o6w;K6?^ARO z45YHYRapfV*r}^+zhUlAB$YYj8+2Ovhm$b~1|)H2lZtJNoASB&!99-9jq)}n%TC3n!fG8cEpvhj#WIAhENH-bI6?aJ=Q({L^)&OSLF7O0OJamh8ZJs*Sw z!(HY7b95Q!1CnVmudavOb{&o!#d+#Wi-0{R_UwDxsV`WU7k^0`OHmUdZ3JS-U$tO_ z;Hik11+hL9cr4<XfBRAKPU&O(Et-rGW5Ye-MF#;C)3bo3RLQ;xfGvF!9HBOOXe! z6Mxg5LQO0Q-*Y_$-2 z%;5OV{O$z`_SV^Bv0e1&=x?CmE}N-kdksVOQ^>`R*oECI_6OX&U;5ndL) z$Z$|uUsP1_2|7#l-SyCQ8OcF{)+zzSb%JQBD@hK16_Gz&9*RL`yaie=O(LN>3E)H^ zZY@N}7P7M&nm5&dy9w=Ptavgr2!~A^uI2_fQGV_nb`pBp-2VoTi5PNs zMjYeWhv$(LK{YFJQaCHMNeb_6Edy&kN?L>H-509|T7E)|$MbKYo`l`C;k@dEqpZEo zJpTx5Xl?Fa{#kDID_P^A(+H*WpTVxQcxRn~|0NTC-@73a#{zH&8R-@X8?XQs;=_{; z&<$&_l@{-45gz;>T8Q~tGN&pJn zcPfUkxrgyJs|em^nD>G1eup@J2iirQNvg1UslU{-DswKgdn^!VPrTaCqxs)Dcha)|2e)SWC9+7U8$$e7`0vI z;(hN^!-3gYBA+shVg``7>5hn(A5_`1}X|2+O4^sX-Vn+WpiCek`!+vklX* zAvxg<%xR3?zVYjet^@N33V|_?g^}p)3msYz-HS*5|!frc4PxF2Zcv(cFU>kdmxm~+{)%=xlLr9~h4L!Y(n$=D-CH2i!N27DYw zwcvMp;ET1r5KytWa7#J`5HVnl3*r`+b!_I2qPGJ)pL^H(Dm*@!)nK!gqhhLRLpOv@ z=r#B!jzlb>oF%{j>Kb;M?#A8)aqZ~F;1MSJG<<=&rfgT&$3l%hxcZIdkfvGXDs=?% zW`WT(-3bPN>#LxGyv_xbo-v0wjsu}}WcNH3MV13Tn2$A36 z;DdOv#~`j9@+}_sWbrzfxP<~0GY!c8jV43D-_?BZV*bc;1XF#KxGWa}`@I)tAis(; z?nmB$%srVF7;B6=3czY7&{5*`u*^y=e>J(_bs#x!GPaHnEv)Z zW4=2WmkRriUX0q(Npr|?QuuoZC}#b61_+gQZb{zz%NzL&uKpw&Aeqy?{uP*LX9mC+ zY4}p0xa2*5lPwYh5y9!;B7gcH!KoWVe(ai)aa_)IVI=Kc-09c!D2v`8A-SjT*QovRT>_dU;Wv+2BO7#TC|vB2j(PIUzE_L4f-IBC=*a|&uC*p zkD{_1P=&JTw@QF`eN3y9m)W$CIyVIFI2}wz-lv=%!)DL>_W1U}Kw=u_?hJYE-O&f6 zcJP4fd&#?AjwN|P|Gwn%AR}97trk>lX>?Dp!E}Z?pE6F0E4+_b2X;WA{}F0Tcqui= zuMy7+SO3C`mZQWyAoOn)ua{Msz+M2m%=>+(drM(9t2{jU93wX7%#C;-I01TtOrptQ zJip0Rj)HDK^rN&{^68GTPqv7*AuzVj(gIdF1RA@HU=TF5H4PcN{lW8$246*1Ja-6) z$5Ix4wv3O>sk4(!nes?549_C>TbXHVm*7LdCrkJ0vgxdASTJt0`mgZ4ToPyJY2Uy! zykM|Hv!}3b1AglK1|&`)y*V@s>fgAsDiJqo{s!+jabIEHtwgVb;(;bJQTdR<+a9Yd)li?@JKsX?9{%!vhYD6db7t+eEd$7vsm%nunqy!j09$ASMN!XwaH9jq&* zH#Y8Qj)&ti=Z(Bb>~GQF6OcO|AQ4il#{sgkL*fz-I#}~y8vWc9Xz+1t+zn$aUKMq- z$sN&A29Mm_WZs84{nz0LyYJu-r5UEJd4ReL>lzw$)uE^bcKesWR!% zy=V9Xm@?Hhf5iaUz=-5TWhOUO;0tp5%|wvN^HgMc5K2*Lfu>V_jsJcP3lsklhvYR_ zf$O{}q=<>&3-jbK?=hW1*`g}rxCWW4LjkZRT;L+!2~V2+$QX5L8$*%VxMY5y`u!u6R-ZJl44 zHB$|(h>b)aJ$7`BzQ;pdYtw$?YQz_Mj)uwpLrb25VCiJ>5}#jHYk{JL+GPAHak}z? zbNKzK<3PzW?xvmfe+VGwz~Cryc(u$Mn#=}^$NIrARS~a{I*tT+TPUH@4qgGSiL(s+ zBaXCo4`j@875GQmT_4_QsHKyfub}uTd1(}j*Kr2Zh&cH7m0^*IFDA#|X<_4e^6-VodCE(;h-d znNLjG^W&%ogtoCPgC-?QQ>~V;lL*!Ws#1prc*MtHM>%^-+d_*0QRB zgnpu20`5yX!i19$$JW1?aQK!8{8sh>gOWib1x_wi0XPUsG)i_mot?kIPj1e{wOWg3 z?H?1lh2ZylZ#-?X#c}e_PvizC@$+03Qp^*=1;7=kcL3f!kgSY+zGZQ`zdr%3bL1|` z+59;xrZJdttpLM8R951n`1h=d;9n;4K}4c{o#3zD>}UE!{Ptc~=U1zqivbzhcFy{U zmV^N()v^Hga6bFTQ#n6I-12wNe{ z0N{H6s#XyQ3Ag>^3E^6rfE1M*K!y`o_aD;0>v{;HWkZhfAoS2sk8DdhI4i1v<5npHlHUt)0f1^%YxJ zdrDO*F8kdZygdtxtDXP2zS9N{rcDg82X?6XI;)9(J|+s5i6#8ja{MR_S~NEE;exKY zYpgfJ;*P71e~O*1v;(3NEoe;ta6%R#qAGWYv_I+x9mY^4$cZH=W_|k&+-RKQn(GkG zC^UQif3$4)Ww*P|UPmLIRHT87YEn`GkI|bN?decaq+{|LZFBw=%&Fih4M1z|##b>w zG0iE7Sbvk_70r9J`#;`5bZqPY12)Ex4!`DL_rRe`I$(Y<)TU7?d&hbY9*xFy>fWo5 zaQ@q-u{0FV8)1YsH@XM2n0&n0?(#G7>h$(a(pH=GiS0 z(uC1CtnT0wgb)zj=QM_l1@r}Hj6N(C)fs~TMJ9yL@NTskHD9$UJBC`+BAqn5WF@M< zikv{T%D!%A1j7a5DH?w?AUBvUn`=0v%v~@;^~hw1R^ZcCSR?$SI|+MA!SOO3^p8@T zbglaXz|N}s_JYqNFeO>T{1KYhW9j@6DjO4@14A-0i%&`P(%66mbM%tvtgp0tC=6SPpb$HC+SNSYaL6YQ zZ0CW0zsq@p8E+8y5PV_0kd{5!Y{aolelyiy2iDB%M~+gs8!!0e!cK5ma)=g@;9R-0 z3yd81WmsnPmB0c^jU9m_ZZ>H%diRV*3!qb=xGtu57~be6(0>H}JJP^r4Th|u=;(S4 zIxIoLfu%a5^DD>W>-gc5NWB*kF1$k;KlhMcDCi$IB4WiD$ zIN?Ov5C0V_iMRx9l4;Zns5lYGB^S1oamVu`W^?H;%SM!>AVQxVf5Wd5l=5KktPuKd)ogE>v#}W1)iU ztxQjWwKY1?^1f-S@;j#gVQ+<8g!I5vf#qNo%wshh33i7mJ%~42&Dsja_eDwU(28$} z&;fudvxsbIpX*!lIVW`zE?etM>F^a>9_shAbT+t%5wZLB7!5}rcbaZ=`|-gc_B{a2 zD)``?Cjwxgiz4=*oTszA1rTnyj~8G((o#|ZfBw&dB#P6#<8O~-;D!XckD zqOwo_Aq+u0w^ZnC_#hoMr@Z4f@`~1cxpA(7;a6mW6RCc>yD=|93gOMH)zB1wAl2!8U*l+BLqZ?ypQa^82O%?-w9bVaQ!>}R;u~+cbcn|? zD$=!4Wq1n%STEkpnFye&U6=;6K;6yv&7H%LU5w9lQv_h}DzrmH9!zwVn zlNXtU|3u<4_`qYGjT;P$fn0Z;Ei;a7s?>7>&xG8Ux^wz^>r3C0Au+>T7$?zx#NsLt z@zVvu-koQJpKD&{V~eYCcqkA*_x-U|OB_*fRF-wHgqy&$GZxJj`FpYwT0wy--AAPb zRBKt$&5llkAvb!)=jN>crd-i1KOH5Yc7LqS6}cU z%pPwBKB#&FZWgFjXMW;AYh&GvVox#TLBS2Qbo)W7UQX?W-sg7HQ0&2g{NN#$&wEZR zonoPse`UfIRDeXJ5^-sA!(YBAlWX{+AX;iJD!jUoljgugRo-&Jtk?WZPJVXa^r@!1 zn`K1B#f*1lWsN9px9Ka-HIG>9(}N9G8x%H;7r^vx^YM8?iLdrMQ+Jt5u=Nz3AEi2e z#7EahKivBJB}>UI=1F@*(mv6TPrktpu39~7XW+E=Rd~kf{Sj4`L}N^6wSdx7kwS8a z%1QHD4|M0*J$189nh76n`|1{qujk*eHvP*?qauIU(8`OZ_Y&hKJYB*^~WgQ}lg z-U_=kV%a{MXK0JaE9G~bfaqZbStU#mJ@zec{KkkAm666d=o{vP3*gO%7qTb;&FF8F);U zok@)sJuO`V}dND86YQko*K7gBPp4OW_j6CuWF1uvjOu4*H2`dpW7Wc8b%tbp)1*a%w7n-YdDUGKLmVas1ne- zD=!V>s}gyhGZ3xVLbuq`m3t39aDm<8JGWaJIWY70I|kz;RDfuD_*ISKK9C`_&h%R~ zx4q6scGY1)hUQJo`OX)@#)82m*=RzCG=d>FyM}hN0R4Bs5{*KO%v1d6An)!$r!`t9 znAC>I=A9mF5!7dV&H&+s|2P4kz?oR;_2fJ33T0mZqdmacD;?D%q{&uUC|ut?*t;r{ z!$6F81F!ZkLP;KvzRrmoOP!OwOsHiuac5751@B6OK?wlXX{= z3N*W$lHtQ=XFzc=M{@B_0u;lTwQ#RwSn2tNC8ViCZ@milwdN~vH5u?m%GIbJ%4zE1 z97XQQO6NoZ^|DDgZ<)Z;nN_CGqv_r4>+W@|fxO5Jji1Puf>d2fB*Y^fw@4Q=RRuL| zWU62vBU~w!gSznf4!|>#h`)I6dX+=e>2oGh7m-i>Xq^KsG!{&?YOhV*Ms5bdR$>cs zF(=`b67fg&--T)_n|a}u1(g|%IBa*p#@C`gD)-xsHJeHH+}eTW-jR<1;O7(L#5;sD&IEu*K0yT=)KS_}9C} zjd|dz<6=hZFf*~hZV#mGo={u~I%2%-=4@eZxe~pxjyz||ZU&o^0Rbm_KK*_-`)6|Y z&zTdoJza0}t72o#@AGno@o3e$*;u6Q=eKou@3HmvpF;uC89`r0gf9}-wO9z>A+8TI z5nkI~^PS3z4w#?jyBxZ@W!uW2*q!}{k}Id|y$o8) z{l4IN$Z6cG%g#@hNM}+Vk#0NWF{OXs|BK0fxq6*$xBcXuTjU}camj?SnADW@ z%>8;JN}2xk&hDQbibeW!%Rip&dMDLAaonI{{nneG-+Gp=es8oo=a2mHRIccD^7{)n zlxn~J=p7@(t}*q-Z+m9gU~{@v%;VQ!^Ou$ryg#ip>NzM$F{nmYmeIcUc2uxauRYu>G7GIyHA>(#{cQ`@vYkS zF)HgPujz}a2m>o|jx*6}W%&o#8NG{RD&G?OMYd_(#_5(DSq@K?ZTpCwIx}8+b@KMe zA8pJL$*v8l24v=mZx4M%lt&-Vq}}z^zM%PISme>s^j}d!EyKEJH^{?RS`BBy@2`55 ziZ{$5vToUgo;>eh?A!O}22O@=<&oB_(WCUcjEiyt_uezEE4#kZew3vC+x|}7@3NP* zQH7xg@woSBU#}^%TW9{X+FdWq2*?UskG`1I^zib}=`~yL;D*mgG5>lrp~)k_-xYOT zvbf4Or9^z@PBQ*jKsL{{>r>mSm*1eEQUNSiPh`DKc#8`a_<&g#Gs`)9bXIGKPb;@pS3muQech*sdO1KU%kM zV_T_L&c{NBpMRuQ+aCU5%=gal%%R&|2PNwdq&7e2>}8RJ6=EVWuJ_Rl6BBWsx=uOp zeuIwNRPVq<-7u!mRD9*0i)|KN=J$$c!_9{KGbcjz2CD{@@}mYT9X{y`t-YmXq6eQM zX-If&{OQwrg6|RH)!>!de40&4Sq7b!$B%oH<*!ehNEPsMv}`;P@JXq)_hyfy+J~P* z(AjC8cjMNM*LgqT*}1cD`scT+FTQnc z{C}05XEfVy+{Rm_gDO>{C~DW<)ULf}Vy~iRm7>I^t=$r}#fTY6thT7pgnw;96+zXi zt+7LpARcd?_s{e0ckc5$_c`~g`+Tqab6uBkFD!n!%onW=Ofpa%n8mlJ%&@dm0-15& zHrDzBPQ$}jGiP;EuMKT%s=1W?j3AC&6oe2`)I6izjr|cZ=aZ!9Dq)UUi1mp=AJe(s zbAoU93#}0Atkmazg98oaAm^XXMMAJk*0nyRC-?2%!sjVyx+cuN7;4M=HwRF+g zdhO$9VT^GR6-My^1L9UR@HqjR;=G38eWA&5)!w)q`WP~DjK6)+%>0^2@Z$rH4+&7E+MH2}B8#G_0iAw+8 z%Zigt6L|5zEP#^B0A>;8==`lCf9(g)AyekdlHP>YPKB?6>`8dqx#Z@lB=#eI7dD2a zb^KTVYmG-z-f7B|Sn(aHI3B24uo&rPf%<|GU8#uhqu4vfH7!2jT_%p4?Sq5^#TGq) zZbWR3>Rjm_A)6jP%H!j=ZV$>ovXXQu|EAO^`xX1_VR`5b`-nG=Dmt`ZBH`c4n(DkQ zZ$|R=>kp;7vz)_na@P8uc;)E3^}2bGKaT5p`cXc@lZ@!N*j?TFG;}siS zdlU7OQLUI~_u;>6rfuhd?`MyM$im?s)|?Xphoi`I@jGzR4(H*_3S#Qd6OALGklNjA!JJW&vByuL7fC^5lhGRcxb9p3#v4sDnL+t48r5+!hdhRX?`y&H z4C=9AwYq)7T!G)+XY7aHj!T2&C!mYVy^1bqNIpMRTQ6R-+Ifk#6iZ*_~T0H{gRBo4cCJ}p1St`WmSuRiBOEKHNZHP zx8x=pAh*UN*)kb+_z7D)3WM&%upHT8@RZJ;W16V?6uK{xZDf_DGsH5Goe}i;fNZBb zS^Mkr!NubuaR-1E2ZQ1SC6?Qx2mOgYR@|?-g zg(p|<<%vRVa6-n5)usPDqUHfOwT}pL01V#WpJi(mCrJr{AgN!$+`> zDag;tV0})xYko4&kUpqjU}FF_t{2CsA61B)puvC4Q}why*^?=#ef%rL4!ZKud}hsi z4(#lO>Xp^O2MNDy96NuSn*@chDz1APy1b}$%(hPaON_FfEUL6bkx?P2K;p3sXr(GU z4rYp4CEgrBM4NR;rMw9k?gLq2K@})ryH3=(!(xiSZ#%Ms3KDOA{`zucvTw5ArV2^9 z)P>8Ml}ZEP25rwRIP>BuQBneb_@Ezi_+1kKEznvV31%@9Y$VG!8FEMlW71aYSk`Ko zSBoQKIy1{4Lo%YAa@JbzyqLAHi=_1q8dhZ!I9BD>v)D z_w%|pbOiW63Lt;R@!a;l2V0{fg=#D7E#ywQnyQzd90Mj=p1jjWwwBvwMve$#3zm8t&f$UU%pPn|!QMDJC^HeN$h*`xd_(8nx6o&r79i7lnedgq2M}-X zy(Diojn}c^#B;})9z6axU`rp5DGvhhU^yw=W&g% zW#jyJYXJ>(?+LrdxAdgSxy<(;+9!x;yKuoE*uq&ojp4_D7ma_D%*3l-r`!K0O`x(r zo7T&DR{H&Fq&hBJJ<(ak-Z}**&(f8r0pw9k&+6L z^q9y(v&=k};%dsfx^5ymiGVyI&)~PEHg6@CB`R%j0t&`KC*o4#Pdc_%X&A1Jb2;>Y z8bV^jqhc)?8}|M5TU-_#J1@xh30Bsqi;ZCF^T`vyaQU3Xco)}QmTk7{)E=^9&zVTq zB+{mwrl9Mj0UR3{^ja{k9>Q!JNI!f{cZAQJt1D-syH4#cwTJ>}ta`{c2Vl zj?VJfvDbdONREQJKB-~6nVt&D`2qtAR-r6P@^ zpIJ{2Wn zx>%Y#BzW0W?&ofjdlGI7`=;+EMIf4UP3pf>@d~zg_Q|D)1l+MgM8i?G3hkxO6tw!!(})x z#>QL*NO!+f^g!d~U2Xm#Ix*++v0-(vy2_5f1S!=M2cLYtz4VXsXU0kL1BJhtf6!3z zKm-F{?BN$q=<;4f+mr*00VDX1kx9Q7SXM{!1!+Nmsf!I<9}hg_%eo!YNnd3pzl?_;{!fEi|b4z zMvf2Xbb?%&H?mzxK=?kHfVXVgUCf54pn9%5$T^(c7E?Aj2>D<2{;GH3} zA#beHRkm!ez&qO64$RW=f{X`L_OR;uq(YzaNmTeI(tvioo1Jvti(&$Icmhjbjukga zz;Vc2sWkOCe$p(ik@a&tDM6ev&_$8IZ*^X|an5yI?`^(A zZ@abydn1tX{y%ybYXroTUkW1AuzmfgpE6XSRD+P$&Mug#@b!MT6*d^NrH%1OQ!VD- zI2OUx)W;i_-TXB(ZE!Pr0LrUAa2syF@nwrTWV}lW_;(Lc?pX{N&b?p<8gou6)~3Fwtk-yqeZ%R_B>WtKz&QermTH^ zCh>+Zt?U-_rJbur!_d*>-Rxi*8~Hn4?xSsYLk(%Gw}r$el~&AlzFV^B&IM#GdJ3m! z&RCOr+`d&09&yRl1UBCdNwKK&<(Am#ku-v`2l2R-LsYDq*bH@S&86fP^lU8rADjo7 ztEChgT%rv-1Rt?W&|q`2wCc`@<79ERc7vkSQF-KowY~Kqv~ihB!*ceq$Q3|{h01`K?m?By&Kht zoaBBh=nmJB6ZVDP+Q0933t((P6vZI#b;5?jxvMwF6F02{s-1&Lu5n7*J0Bo?2Gdt5 zmbMc&p?dQyT4L^Ae13BM>ldWy^Jg5M6ET(K(LcXy-axjL|8d119YvUY3A;OkcQJ;} z6XX@OcjwI@p2b_yu=-R)cyYJWE>9jqx+#!mq@eB9=2vV;fm-HzU-MMy7QX&8+~sj8 zs^)1^6fPC+s416n4MKOwj*V{IwLK3;6ljETKU$+W5+F*_RI@z2%WtlF)?XRbo2TXU zB^_y3_EuVkS1MX}_GMF^#m5NbI=1Uwcxc2SrmVdD1ART$M2)+Mst4D6qHgzbD+Z?d zNYGGs%;R>~hO9HHg^5@^!vb~GUl)!2mRI~1tM>gv-TuM)louI|s~pj#SE=s$lByro zB0J8}@#hb!KY%V(Mc$nrSU=Q7AtK;guM|1rDh|5;k1s_A2i6r?QJ0dBD)b-5gj+-b z!xPL`IIdRLvjra#-${v9a*FZI$CpF4(*Zr*v(ulI_^O@=B$O)7<`N?vX@xFdU4Ct? z%OFt%JIHBAPB_jg9VIHg7I?q@v0Qp+e&{IG46ej*bzF(@t-(`W;n!X=T0c^S<00Fd zMO=Y|xm2!f4grgiOfRhxfr-3Bjjd`y6A)OPk`usP!bn&p)K|o>D>aq0-#OH*avp}D z>lZwf7WFXlg=&@~(q~d_*aw81+O%A?kx>`^qzrayN&!T*{kW<|(^Sd7u6IH_aW;)o z)FAJtIkZ(_j=%=F1{x^?_iGY&tEo%;#9v9alufDd*!&Uec>JdKs?a_Rs$@B@2iHIZ z_18}VCN1Pd0&;VWpVwuIWKJ6`J*~Ce#}Qxkf`{1PbsfZCvC@}Au=jMN(z7C5Ivu5>JagQhyYN3| zDvFd)s!aH@_G!<68{2wNq`vy#eFsC>MJF$Pspo_ zNK-&;M6`p6p4 ze5s$Llq4%Dv}i5cR&6($+Me@i-1&v0PzZ1&4)PiD6hh{aI3NG#kDh!dtSu2POv{j| z%gL(x!N}Rx5Xyz2WxiL)|BY+ZlM9KcU862fm{;pCPngqes0+sr({g6nR^(8~5o5#M zNgVc zu*POA_NW6DMNkZ{cyK7Fj&6;^tzf;4VeQnIIT%GxLTN+M%4mZ&w>O1t$r4=igXn&e zt80s?_z^j*V6xzK-);>BNjQI+Lm*m(ov2O>)KFwv8tFD0lF4yddK2D^%$dsXwqnTa zFk^l*y6xc)aG~#1d^hAUTeqpI_@$4kGC53`a$(g3yC}0!+^*V1QYdIRO0;ngU($RN?@y&%EOwDl=^Jz;}Yv7>vjDP zrJJpqIHgQ0H^VS#!q@cq6dME}x$jy@Z9@ZKpvK-Nj2F_&hcXcy%!|IWR%&F33{M&= zC5bcHvIjR57B-2~be>{_;@XiwIWGkLkyGtD`~2CX5PJo0dvJu4p~V_EwEC#sRRKb}K&puRnO@hbEIk{-!1q$Op1SSrvX{3&|NVU>ZNzW` z3*FAa{7EiBDro5Ov z7Sdw?*XAoU;k>WwE}Ml69mi^&8|uVHLzCKUDn#Ge&-riiVh-kxg_r@k=( z0Uu1^Dpm=$(Bc@g2ihlY$8WBZT%olWyMx-bWHuFC#S9p z2Qb;4im7czG$5)&22FZL&3$Ga^7%h?xmSJX0P&u~4bbKvGdR6bwlKdtx%Z4+VPUZX zT(*L5LOsvDnY3{kqyGIf>7%z!PiFvlXD7|Vi(~w*9T}LmO}2= zp@IyE**_M>{P|ID?<;efs{WGln-OHsi=SOMfbfD`r2M{wk-^n`rIYYD=mGNn(N<&) zP27{Odj$l z-&UQ3Rx<_5A6b#$1&KM^YOM_fT@?ilL$s}*O61qqpmnhj{V})X(fX6)81q+8rdkd@ z`nq56M%T`Wd}FkwY-bDsTTIv1g7vFOAgdugFW=?@VzX5-S>|$4!B@mGL09U9ycMu*IBr> za)#Bugxv!?ox~SV!MF*GN@V4BT-R^ng4~yVAJV<=F-K9yeK-*5ZcR0s+4?rF*=pr+ zcO?rk8ShkFNya-ye_EI;O(bp~cX2dJ4xmVne=KHjm)>N_&GXLnWh+VF(5;|r`aR>5 z!zRKdJdM%5Ly0lEQr+sUC%stEI=o~0(Ram10~|TC8gA8{I7F=JTLGeDbQ+pk>21cA zt?$-qLN70R$Ya8md>eT1C^_+W9o576E_%t>&=S*MrA{{C)EewB9;bY%YcS=Ro<|{LBsA4(f9lp}mIw{-d_u(AN{(gwWCv%hb z6aAw;^9j`UD@SJpK4*jdiBT3D(&Nv&u~f*Y#XNZvJ(I_7Oy+I)6&_Vfld3GtWOAe?%A9Z)z89XeV@A* z_%{cd+&wEPf3O|toW-P6Wn;Ob3m46|Y!UaeLW}ZdTr&?T2?g6CwVXj31Y^T}tIEQKdf-hY z+;VcBs+Xf40r+?q5FKll87L}YE>LN8B@%GDObD%=m5y;7Y_L)+Y2nq%|5Tger8ys3 zuHy}JGIOEn;9tW0X4tLfrq2C4(OH&}r6)aBvX9E)o{z2vZqUln6IxIw?%%MQ@ajfu z8fl;m3f-HK!OK|ngZt{XM|et(yo?c>uysv&1|OW^L0Ls-SfOD3eBiVEtFS{XTk_aa zczUYPjhk{w!WS0T%P-4)WELihnI`YTO2*@H2Z67|JQy{cg7Gpfacf)sV-Fc=xBG!I zt%2YO3qp&EOS#@7CQY$WhRfoB-VyV+6>!f}XDLw=B;$g4j6 z5WE&kgVZwyn!q4GUNuXWkt+VQTDWbQS9H&@6kT$F^izEoY}SzZ_k!)Fv;R1BEmDY; zGpmb5$LN_Ey{|riucpy$L_GMu)+ePtptGRA^Bc%ri)`4r8g<|BUv-ZI+c9IvcF z8A~!*?wsXlPs1b^*RCQewqPu?t?G8C(Q8zDe*gYFmOpT0^-Re3(3bmmAqS&9BxzxD zh8DTi8E?;p1-41ae+wL}m;i&p4cOPWFXqRRhNAP3%3@Jqngh)cmZOtp+$k<5n`XL2 zKDRHoeqmEuN7?9XV{rX|ogfDz@49euEq-XssV=0hsL}iM@&2C3-s0p)jTkeZC!GFI zL?)|GB#aL1#yEnM_ceA1>rQn;u8)OqP*1g^A@`zLa72KANt;G=&>M0m)f zlwn!6Lk;dPUx>zOyLlv*mbj{#mIz&xGBA1MH2sexc&9hU=*fsID>PzGti&RA>-1DQ z2A}X-Tp~96OekLkyMd|MJD6e@j`tU;0Wagitac-ZvzN*12Oq*#c$s)OKNgZu;k6N8 zTKKnYv^~c;#w%)8Fux>SI3qxfUqL2IZa7|Qm!`bg_aCwW-U%F6zd*-U(B&96 zB7I*Iu5rgQ7 z#GHaW55oW*P)uzO-IB8+Fd$*^kGj0KNK+!IHGa`a%iAOUBw^zfS}U zWh}IVNw}wqbgwM*P*QrhlKsQ+GPF#yG-zX~TRB7-_BrTm>1;gm!|iQ>1lutDd8l9g zE|MR0Kql^Rq7B(qJ+3gZbAteEjjoRQSpXbY${*rHHpz54zL}E2rgvkDrW&f^B6cCC zD=D(#V6twN^UG$pBQ{KLchjmaIWqArGMzo$gr_L_GQ%GJ#Kz7%9!X_9)!B}d#BfD( zXXQJ;dJ(y#*{wyu%n=ucc1~pWwnp@I^FG#89KPHjd#njCEqkI*Z01~*zG2`5`Av-= zb@vuQYyaauQ3O$}cDxRk2l6G9-P&s)d2~CU+&(JZ3%$&Ea-Dvk{cYmN=cTtLYGAIa z&40X+Q8Mbj-C)ZwN0TRP{}(=yCDFg}+O=zn2HF}x%Ks<0$e8rsC0zqOuN+1m$#AvF OH3J<}?Rw4U@BRnV&z+e7 literal 0 HcmV?d00001 diff --git a/ALCF/notes/assets/cpt_images/strategy_cpt_stage2tostage3-1.png b/ALCF/notes/assets/cpt_images/strategy_cpt_stage2tostage3-1.png new file mode 100644 index 0000000000000000000000000000000000000000..dcaebb109bbde60b5fd38ee2a3078c0a44a61022 GIT binary patch literal 117945 zcmeFZc{r5&{|8KpN;qvwww9x^B{5{@Af*_)?3IwQ@9R(`NocWTgh9qIw(Pqk5ym=p zLy}=EW0^5B!_0G!>YUE^oZt8NT+j3G@4C)hr>pL{@B8!K->>)UJyCkPS{(aN>}O(P z;<$EI!+?otKZJ>C&-h*z;BN$}r3B!A``oXZc`-2wyk~rvU3+|#nV6X0=-n{Vq|@m% z8m)P?_3YWRx;bLNwKFp_OrG`WA}^0vhf}wjlDF z-VQ8`;MY}?x+$BK_4W0Qjg5$ih^3{a-rnAhj*jT)=*qF&rm6DjrPWFN@(g~#z|eq1 zB#n-a77V3HOG__qF1Acn+uGXZsijR;k8y`Ivv@aR(z-}1L=O*-!otFCQh&JxYHI2W z-uqN=j;5?IcsTG9?m_ zU322%;$2-`Q4Le`Jvj8UkE__lukSw#2ndW4M?Kz-9XWEOHT+#MI6I=*<()^RKD?hw zqkSfSnONEg3=A~LlWg{{RZ~;j_?T~DVX;J}eqCSA&CSi0Nq*Hj(=*owsQ_CQsc0tm z+|3)Vb$CmfBlf3sG)%mG^XARm+U8Jhe`ixib7vogv}v8yRoIAx!{Kwp>9vimWb~uZ z@Q~{2>QjEO`{hG1F)^vBsc~^}wY9aCm6cIZQS%Y{Qu?tjS2I_${O1<%O(WG&O{sUi zm#3JRPBLB7P&4vpUfhmKKdK2M$2f4_88}jSzeS!*V!9K4vS;n&o~u#KryuXX&~u+N z_Dk647wTuNURPPlW<9u@3Z~zw{i2cPl{fF4%#$;p;(Iv2GsX++rwLtXf83UNOO5%6 z>i_-se-ijV3H+Y~{!aq`CxQQy!2f>BK84Jwz;ATTw^ok0WGgy4q#V#zb|YkWcVKvX$jFQE_gSHjz1Ti|z13T&4Fe zs&uMGofEXzi?Y(Xg#54+Cx!gb!0)s-^;@30upVKnRj+l(b7jz~+|suyIuXq27p$bE zVm9c7%_yX-jq61M|A?O6`A6DV*HmoL%v=u5q{A+6is6$Neu&LD6n=g>SQ|=(!eh!HCeoHi5BS0#$5$yp= zcona-|Lc92qH5gjPt2fbF+His#KWeBQBf$K_|MEW2B+iL=Bi+)1OgOQTL&<4Y<88q z;P!eC+2*u4EOms{Y;=U5Gef>WM2_z%M7()75Pv8DeijNk%GK_=XH6A-peFe~OcDgV zM`-&LvwR`K^E3P(Ajf!?(Pei~ zwgxLQ3YB~7K1_d~CHmwZD9Bc~_(jexImR<=U4i!|{qs7VYtLI*yeSY7fq3lx`!KbG z0&r0faIf8ki1;=pkO%1K0X->^2Pk3THC3G-mwkNz-06Tv)4+j0?n?Tx6(V5K2LQPl zHZg&+4pqTkgf+kn!Y+aLcrw26alDP0aT`uR06z{7w>R2lDs@%O2L0-A{?}YU2iJf< z^2)*OxuGDQ2ZgUcyQrz{o8W#Ef2-t-Fz{@lHUIm%HaacgB^JzjQs@7$(aB~7-j5-` zDwx_nguAKsSIG#4WmT@#HWm;bl`HsNqQK)HRegrAo|;4DrW*l@1WskJ%hXW&GemnI z1^Bcp9tGYGkK*w+!aoShTV5=3Y-QCN;t&0fbkh~2MO%~?9B|0n|V#9Rym6JUA7YIWw*{oCK9JpdHiWy^g)zvzlFedHeP{TJ^3h|;_MvQ zKbXSvMkC;Ixb}GG?UF1PUYPEEaXS&sp5nk5W9*$p?$ll6r^zvU-3fEm3msq6cWq z?QCfQGcvd5Zd0w%>4#^c6Q#H?4mxv$$Y3U+~x#{B2SVX!AxxW&&QbKUvg z50DJ*!?dHsMn&`tr`j@`7+OCcbc6{cv=5p5@{n2CYq<0>`~RZwhMxLvm2Rv&$;c+?m$+q9w5vBwzOlWcYALE z!GaG6I`?5Oc4IoPMTwc~TB4sFnjstzfcKjks=(ohcqYK}xjD`jBHpWwWHBLyuc@M6 z$G2t9>;`Vcd;9?*n-Q79*yeUV4D0{p?NoI%&j3anX_QLVw$yQq1AL7UNdE11fS|O$ zUK}|H#48RUrrw7coQ`G%ynYJ4sLzCn{~d0xt9`fEGYqD^?+Tnv$p3yJBP8y87>@ea zA@=DR_O?wk%R@m>MOB>{q3@DoxcM)7fFm-Wvq07=PHM7X;tzd4B_k&N^051WRymPo zD;4nHfFK{R5AZbLy%@gncfsD>30G50poFCV@jl$hPkTxK{Wd`StP{W53rKtSe|cl> zI`Q)lSujS{VSoyOHwQEs_T6Y1M#BsdIRQ8*K|N15NXU)-1^LF*1)*2(tpLIOi>4WI zt;baA`~x5!e2;ZIR`DO|V>bmt&?!DRV(rj(N&SnocY*wf6(qw9azmpfX7l%};xs zvnUC-Kl1D-5G!(}=ZGvwD|CX<>3bXy5qycN&II?7_kfv7sObpzUWRm4Yhl@It*Ox2 z7$fl8-Frnu9 zpmx0naYduu^I>M(c4mZP3*S*=YGd)Dq1hY{mNA}^voOcAgZcP3m*`lcS74PMnUnX45cBi<@dft6I?q&OD(<-T}v|rA&G;% z!6aE?_Y`UT4hq{qyKQH&io{kGW8L_=1{7_(${fOB#@B`@=kyGDn0};(zdOyn6QGUb zu@*LXkpPUV+>7BC+j>Mtug;e()3`}gLhe((9M=U!hYg#B$Ms)TwkV&Gj$3`|tWw7o zN8pG<5Bo%zAR=6hH1SUb^dket4L<8)vGFtG@}t7D#8vSd4!(sc z;0v>+OJ%8G%0QDKJmugVwKJe-Ns@jarUWa*bHEYrxqiI#-(w2w3R;V!wIEd6kISyS zX)zpDdyIQf;uUAwWXelFAy(um$JD)eif>Qw$>k7;|9_O z8_3HbbJR1>o+Al}#bLw>I2=fwI~lMaX=Yu1awwVHH;^@Yi>#&-{i+h3;Va%!FSOgdn%Xrf%0%MoIft5#tK?j{81D+B90%W9e{|?7)i{?LU$m43J9j3 z0lx+E+)i4Nj)6r4m3GTZCFz$K@pdXB13G)sgWO%abuDiw3@R#~2VXEXE9dGlz5GG2 z(Irv|bA!;(dIWB-`2j>{#ux#YZt$SI(J@lF|n$xFz)aJ5@l?YKNf@OBe8|KP=7O1)v zw1t7Qnt$?Oj%+CdUu>OivZux9tl8+jxC#mZ!yr#$=Sdw(Af*n0Lw8f@p zDEv&lBRRtGnI_{|cQA}+xie~PRUM$pL4Sa0>PQ(=oXGqyZ<|#Gcq_Ah%qPJS{Om`9 z?wS)<@$OYY-053{u?A5q&atfVz24Ysb`p?8?%NX1J`RhLD7gKzSMPz-h#Y(&DC+`g zcymb2OV99AQcK;h8SdT`b4}xVR#82l;G3f8dWoe5S4`PUh;Lm-3hQq4XZ}j>T9|8} zZt0}NvDa9<0DR2s%ggLzAitEii|no2#v8Bt)zBv;BMg}^MyFMOWR<@szd^lzqg#Z| z+Y933fxq`AN)kRXD=xd2Xjk4 zW>{kvxBc6ou-?1B1s5?!bdPWcQWs$N+GQ~?&DHo*wR<+>E6j~C-Y+4AjiY9Z&oa+J z*DGNI>+~MeYuBl-w$&?}lvTMRbJ&aq(=W!Y6$t?4KppI;%ePxVW0k= zPBjE9pXb4art715CtFSkaKE|hK69J>DZxrL#Oo*s`vyjS+x6v#SsBL88T+& z`kN`yF1%>cxp{y@A-Waqf+M6eHoJaCZ=MGN?}3u5?gLP@F4e+XY`IT|8b;Jup-+3y zma+LH-m){#yus1ye~@~Uc8>ZfM%AWiU~7VJYwb~rgUq)Ubk1GVAT6$zh0>CWBS3t0 z`zvLCby54XP~*tL=~KTc;H~g^x3w}SjUoUMKOL!VxWrOheyLeHr6O*IvaFB;ajC@8 zYN^uiSV&_P22k!$W~-N;9HM~9efueezm-avw=s%$qb~4FPv_kpk?cwtiGdAPk?JoP z+sS@wZB%Tl&+)V>qJBfkIX0nRm>pPQ!cJsP?t*Bqfoy=v{m=4(4I}1nq4#Hf(Tv0k zL{1K(q16MhCq*qIPxG3?Duz=VO~A$A)@{+=Xcz8-L9DE33xAH6Zsr(Ez~EC)HvsoF z@v&)6SJl}g_&+5_^EC@3t;TbCYhTgj%7MHC>?I1vYCf*l+gDhn`c2+KLBBOTdn)W{ zfmO}AIEi3n-m#jmDuR=B0pD zGG`tv-q5-P!m^N#TI%>^dyXw+7ii!o^c`(V)2#a^pN0uPqE{B zNnPlh>dd%-r*c4;eDd!_QUE4nhGHul_j<$w*EZ#o$c_S68Dh$ylFg-_1HCsv>);@ ziALiTq52cWQ@p4%Y?ivr0f&jxo~4+$2~u1&2TeTq+d^XI z=&iFz#|%tt58ir_3LR{9_^MewjR^hCO;(VqJQyR+ zzmj*y(kB}-$7GFH?RW9q2LL-MuX`7V@=*A&;-J~qBBk|(=R>LK)z0A~kvOAeVM2 z{cj!jua9cu1aals`y_45qWsIxHU%&OPko{uF>^aUen8`a)$o- z1U5$U_!He@hHwuVI+VV>t(JOlj#ZiqodjUKy8i*l{QWB|5N>)%{j{^pBN)uIKm6>O zo`i$(P@pgn+W*sq_b{;CpC)`77-fXipO*kBMbw5`JYz;viR7hthV!;t18||^Pq+|J z!7oq~@GTMt!cT_R3$)GRO`RKeYN~^^{|6BE_XGtD9OO*-XBjU>W~PKM>?$yT*w1Vy2)tL+|L|U17*Zjkz;H7GV&b4_EqetYCEVg2-X z`};87SV1|@ulTH^4&o+mWeG|Sv3#4rkcn2j3lu(B18?fRB#8#USMCr*Z?_VFXIsa=sMnD{G#k8ZU-ye2M99CSKq5yx%;7j%#h9Y zu<1ha$UF_jEvugGrw}`SwDFG`Pj8!(L+^?f3{FGMhRW@#$btNNYUpa4I8COsH&~cS z`6R_QwE?#6D_g>0T6!=!hjbQ>*!8F2{sbj8P{_@>ns=|>4$tZ>xNljS9fZ*4D=s+4 zO=gs?tq!#Fkeds|JvKU&6wr?%nS~57<8NivroWv=PNSXkur^O%eF$3>q zppT3`byNf0$Aax>4EDm|-8dWV@UFMh3x*|H$pjyI?X60yo}8gSvcsp4ajU3Tj)9A- z3q#9FTieZzP+~{N#OHD9tJq>>!1~RhOozpN5Sq2el_80_->h;PE=2+ea%saiU3!9= zPRnl$gqzR23rOVJ0IIvw+TS}2N>!-d@Yh!~=eCoh!}iqe*-*6T&i6BSS1bf)bOm+F zI9U^7Ce=jg7cF9991Cw&dC53?b$)wu;g&wp4%%}~L;+ddYL?X+LLKXB+{P9A-j3tU ztKCVF`3jwe!B9QJCKjN4I31mM7;`|lX)KIX$^aC(^-If+ZzS@cb=B*?pdfx+zza1c z7ze|1m(pI!&C4$M`LcZOw;!WKxV>oZQ0fvL@IM_1`$TAmXy}+0MVztE$u$Z1U^4GJ z({@7TYYzbL4j%uU+bE69{f!a!pR0EOgMz2Pq-zIu6wkz&Tz}L!@?et~HP6 zTN!VSCAzO~dV26(of+u->|}}2y@Qfwd`{LJSP+{!dI)ViC>S*qf(|yHBW>I$v}%Sh zaC>`>=T_oLyFo{HbrvdHB)Cl)DlOD2yPEX+YY?BBH|5xibz%;*XEp)h5PR*zED~rOg?||| z{D4XTsr5gY(T$G&11;{BEL1K8+`f`yGdc3BAnXhNn|S9FFNGm=PRe)<|4c)N8~*{2 zU-Ds68#c|aDOh)N$Oe@b3QU|mKnwc_c~v6F+>1N1`aWto{y+dgdJ+x+veFNCpKv<@ z`Kaxf9fxu}H()tz74Lha*M;AqV{&??uRzn$K{v3&?Zg3)o8M}oCVD@{q)=sq%S`%J z73?L)d|u57U>5Ytt$|)gk70FXk6{t~bM=I0xu5nU%>r~?1s_k!%sLmhCHrQSU3ENO z<`8gaHaWFRLHtRRUg^Sq0yCwzFvbV^-I_$}P@ijnsJ52@$uQ@<%^|pBsJMBz-9W+I~n*irDXKE#MiZQhUfbqs8k&@&^8h)>~KT75LcHR3w(!Vu`rX749P2 zcAnn=Q>|^u%M41_y>hf=0aN*cF$y!Yxt~L82VkmKEVs%>gRegp6Zt+kJNHwYOEQN) ziFQbgON0ArlhO{(xxNax?4GWM?i#<+m;zp)TAr?dqlrcdnj}@J=uQb#bh?cBdZq6{ zuhUuAMSENysJFw-K_qLFZnIuG~8 z3p}t+dm9;XW^s3lxx{{|qLQRsKk9=|Gz@!v9O#%WnxBKyq9QSk-_H;V&mnUHXJ7=~ z9lhqp4;L#ekr4&xL}2tfA>@+q{0`MENP$#Y>|#A%Rnj_8b|&iUUHM3EN-RgFOxJ|eXl0n!|7u3jdQ*r?Bz3F zb#HZ3h@c&3vNb5~D**geQ$##LM#jn@3huu%4c*Eq0!H0{+26en)7-o#M<-38(%6Tx zz3hBzVZH`v+qd%y2Ux9BxdiFBr-n|N##@PEk&eFIO0-(k;QLmq)6pc4iVPWs!SsjtPTZi61Hk5shR>}@REke6jO-F$ z)RI?gdIi$6$r+tMxm935mAepP^7{gWGaN?N9Yb%WE*$gqb~Rrx--&Mpl4*qKaM%+Y zpm*ZbRTGf9V8<9cQl?%3-j=Y8ty$kZ6Q_v!>w!LRw<0y0BQ|eVUL4V1ycLWgFhHki z?~E+omX{4Riq((AOawxB;`{@uBC$sGr23NOJ&y*k}|JHv*T85#v^<6O3QHgabgj^0`kCCr|R9C%JkI$<6LvrzI@mGDMJnPnxRFu5TtZGrc&8q zd}vjGl!M$n7d`Cz)^t1H7o*!?=EQ;PjfmJomm!|cw0CS}b-Z-O@KYYjaj9iJXhUFO zt6M%fnbDRpW}TJv-NGIp_xlG>1`2W3sI6?@jY^vxA;*c0F?ew zLHX;v`{KD8dH`?WFviHWLVnk_RI7b#jzIPLW=@`=!4+_+uZ2XW{gfUt^35TZLfTwo z+hs7t5F++T+RytJw?Pl$z zIGFfC^Rv`|=-gw09K@VX959zqMF#ZqEvHv5m^g8QUOw3ByMJiUYoO0Klu@2&c(hn3 zVJ`%j8ANkflnq*s5XQm|nf4nP3%-K1_m`CpKRn+VGY3(2d$97nTbbH_y!VEFY4D)+ za(&79*)(=1Tg!k~h89k!@#&SZ!n}hZKi{aT>NYH5XVRB58jUmZ@R&$oyEC7$k%? zxV3zut+OC@}S)^3uyHsxM%6&=} zc`g%I#5(`gV6;fj1Jo$nHl(CDE9(_pEEX+3IA}RXFmf#DS#PZuV9CHOeA-8l-niRv zDj6s*@hE(MMCeXy5&sCb8(SC(pmP@(WD7(|*+AtUGt5^0d7L1z0UMVW-@YN{JfKV- zp9C`f<3fb&safOo-uC?v`G(*zt2yu$Hk6Un^&po7^(x!Q&Y>d^LUD1`h0Tehrv18k z)uJPgrh?^`=UZ1FZx_4&u5`CcnKLQZ?7%}9;S}wyuJdexSB-4_y_kiaVgPi-e^dgP ze1OCMv|% zUyICiJfJuxcAKA;;`3e5YQmuipGvXhv~aXwzAUAgtLyi=_2i~c4p6~fL5>jah%xFG zSIA4HjCly{)3U#Y)kW%fx8<9ysX_6eV+!V5W}z)$^s93l5z6cJEUr-6bRFHg19MJI zM{BY&?4?K+#Vs^qyK6&c%t`cy^kt~`G7TN7a$b4O(gpEg33B;vbQQ#c_kQOjr7$^=qRdvP@Vk#&X6h=s~RH zJi%;C<(d$rw$9&!CbFa!=wLFCue~SB!Y(aVn}NeJ3{U)v9%4FF|f@rqf6& z-Es>n)!(R-CS6i*B-!0~_AgSv)kg8@=9$z*IhDe;7gmxE4wZG;I;;a*%>m#S+`})* zXJ6Knu}Ss;MF94yyR}%q5xm{mTfgDuv%2AaYy}yx;1$GccxtGluk-8O4J^fZg0J2Z zYGB&0U-+oBRy%Z_U`$tu8w`!kUM|LRjN&h12Zlr(pLd}GJ_@XRFK0V*b=jt(TWj4t zr_S*{D)t7dOJi_1CxqJ~OktkuW+?U91-+o;ox(Ecb?C8RKRXndfr-um%2*nIrH-*aLQ8D-~n&eue0_uY<`s4(Cw+n z&leju@I*HSCn8sBeJCt+TuGyGy(0ZRmgiabSk4zhQ=L*jIiYN8$VbIAxw>)dmf+YV zWh%1Lk4nV`E{~`v*)%|kRyK2nqGwax7H+iL&A)`(Ppu3ZXPvL8DeA!}gqyowjGK7M z{pEdRk&6e#mv+rkhdtXZ``Y1HPbE=>k9PWtRrvG|7P^Y$jz_2EawH&mREIE_X7y|_ zFk-&o;bZf%^u~@}eJ?#LzW~53`}#aP@F+E*KQ8Y$?zVf`glWH8{GRF!W#d%m&F0zB zUWL0 zRfa5*2E4=Dv9_BfWe%{zLCdoX28o2v7TL+H{TXk9ZW3f`*d}L8*ela{+j>Jf6tL^p z^t@2p0_l;c+QR&}MRGcXUO7|IzPI%DtkZWWd{&EIxKEe+tbA>E@YaIiq~OR~pI4#? zq~iqxKIpfY+<{e6U$doane9%gs>MD5V<0xKEL`QFAf(1#t({6$Fm_N2&I|X9J|!CY zB=6=ebhiq9t8|{OOz%8kK6D2mC(AW78ANueZF+Zqb9R%Xn>y7J+nvT$h>+w)-$`w~ z@5XJn#36os;u0A*emL#iB(V_Z8&s9uELd45PoEz&6}N>QT&v$GZ5_$>{+MIsh1|@w zBAR~d96)^wuYP~Vw{&u{Q5mMCOgcXBDp)op2w7F@T0pdAr3%Bp_yd-wDsm)eZ2+8xMmo^rrOeOV$>9e8Z#5If17_&Pj zG<=Oh2n|f7Ds9_5SH72C`QcttcWAIe+Sa^aniN(Pi;r{OcOg??S^(VMe_-}uBX5$$&ZyxI zP-m|xXbeCyI@K5iL_NciAqo&J>tG+}OFLf;Cm=#5hPJlYjLY44?1f7dzfIIBdf2@g zffH(;sf_Cp?1)7<;`@L(5QWp-&GE6`D&C<6K~NgG6lo7MEH`#+&spnHbHRP4 z{jWFM3Iam&g2=SZ*HpclPq0p(c?+Q823jB4XjTR)%Jv<7|C9mHFabg;Luym^>Wl2% zL8)2a+Q#w{CD_(L9o5*-{nG=JGHDr_gBGD67PbAdlD&jWOA5)E2(N=PlD-1)&_v)J zAJ(z7^x3ANjEA=G#87Kss^#{ghW{LKoo(g*L8SiLLNpaFYFna|`i6 zv$b);G>`%r^e+;UNM*##>bt?i0Gh}43v5#mr&`H>vH<01Ri{O&Q5_(#P1<-}vndT; z-xTjoQNv{~tbc2>K?jiKRcg9Y(?>;>Z9I6yBTWS1L0o;_8PZ%mF8r<$x77%VoR^C6Kn<+Y*#^B3dq~A+?7CNkh3{~+*?3P{rD&nO|D9!Di}N~q{XEG{ zTt2_?o<+)bcP%9`fl}{*=qwQJTMV-NUtCrONr?k%$O_SVX~*rx9+*!eTC(oH(LmR3 zKt(L$W%6JOlr=@WrmYtOsWgQ!(dbkCVe-($5p zw`CM%mm{ZxX*ngs^KIQmMYADq5!+Nq*Vw!wP)uE~Hl_YOeau_5~hKUj+9K@eK#uC*XE(2fGp?RjS&9aRtVqMS`8r zyVvza2Jen10C<3&q5j_nme+vZ6W~MdaEbS8EkHs1oz}Y)DvpuWX~g3}gq#zt4X`LD$3o+f z(dsaI7@kIG$npf$v(2OK61JujZ^I8{Hz_08F3 zrHzWv+*px8VUW!EmV&Mc@FY=eW5jE!KBim2hDa2oJ3jAJ@80(mea+JAZ+8-)p2$dS@ zi=&qsSFU6$37VGXT~kUw&i%U;Fxmdk;ofm>4A5$Z9dg|r82ZdxYis4Tm9t7~L&p}E zTPOuj^ja1AV`i{Y1sXJA(@jgUlr!T<+CSaW;n_?NhPZTAJtTH`5+cDWQ0nGdpkgI8 zWHr9{5Em((qn;2#?H9z^D!thLl&4e`j~W@V?%JcuaZENsb8c2%@Eou>#9+Kv?Nlc} zsp`M)Y{m2f;}8bB-W+P7hy1W*o2NtX`#d;q$+jNIHBVMocrzH}yurOJnMi1f$5Kfn zHxNeEch_IW!c>JZmSh9cfTc;=J_5kP0YKaTBvH#%!8B38$W=T4A?#3o>?QK+vSn2p z&)oaC7g{lmwkLsQ4nCW?rb1u>bMBwPL{YCaXwZA0u&Ga3)y-wK5D{vj=akSB!}f=c zg)P8PE0G7tO7ma-jY(b*tIDNahKGFgqrR%#Q{D*lmG-B9uQ#nvJ2BgHtIyQ%E&A8t zJHwU%!aO4Z0DLPx0MZKHm#-8@sXxZL3xvn zO?5UKNkx{SI+?ibkW)Li%)fh1&;Tg?a(f$RFcXioMGijfVvzjRrAHY8*5Hnp6)Hre zh+&EmxbUK^&G{h{E86F9dDrrHV)w#eJ|cu7s}q>A zVzbn1pq*xvwgO(_SJjCqtewG-uUr>7_;u%C30phjI?Zk>#Xf;P2vWM&)fJ+b5}jvx zqoaHt^5g0Du-EyfjBzSkeA1um8)6gO@q1=;uG#2ppC2+qa3E9xEM2#yx!k50?wEub z*@-vKC%5_o+4T&cU$@Y8ajFBvu(l%x6(+MfaU?nbB=msFoZ82nsE(UW?!+l{@io3Q}%5T+@Y^xi(pjlG68H@<%5oaqxAWwUD9xAm`v z+o%w0ojBO+>nz9R?;Tv|w|~IuJBwe8ljxV8Z~rh7+^A0qDl==3o~$of>z=9ThA*-Vmuii_BI1WtKAs?BX%TMEf8o2A_7>XBFDN<<4<$6woW3zBl@8 zXVK!n3GsO*V&v?w^WZ>m-EDQrkWKGw0Ad;E<+bD|gx=xqx02eCDVlvy6hIMPu>73M z{33-Ky?*DA0rJR)fv39$N)1LEVD^~mP;cvr%2b(BH8lTvutQG+FevFHO@`UbD=@CZFD>h!kAesds0V^q8f7Ih*Yx3iS+lRg>%M=T#l$d?;* z&8~2I>q&iV2o2|n)M%ZaxBQe@WZz$CuX)E3El-Sg0_y@MZccWSE3eFWp}jjpO~=mi zuWgx;?t!ONnE9vo1I{Nvh8}oSa$Zy5M?f~z=_EV4wb6`91 z)wBG&z(^u@<)5@0;0uKTrlrF=z)s-v{IB`IRdat`We?0}|71Hc_7sW#7Nvn{962g37#_y7j6Qu!xS?eD}&W5z1xd(~uMr4#Xm z<=>e<_l`1F9a-D!7+Mni$)Wf=0nsIsvGTd6y_0dx$)DG7F?c4*0zbBR{{0$u%xl8I zIo{ax9nM3-G0L@xQ`a+a0$weROV0b=0R-6JD#UOD&PRb0%Y6l3jI4YEOns&t8-8-* z(83`WJ!vUH#|CTdZV$;4)do^bsTb(a=BN{wsDkr`2Id$=%9x~=reCiQwd;3<=r0WB zgO%LxA;a60$yi0GCFn^C+f8PW&>>*UM`vc2-2|ZOCtQqB#85R0M0a^COlFaIWA3)s z_CQ>Hek9st!=}87M`Lp7vS-pC&*HtneAr`cpp}K1_^UR9J}x|k^~y@BNuw)nK}rP| z6gO9UTt$(>JVS~AC&n;WHMts)_O9K3N!yTG0oNMZ^YLj*gPiE%)VJ!A)0nLR7pvjI z z7G6S`oml7rxL#{3o&A%1AsvMhi>2>rp3ckG8;$I~V|$>1;*Uc*U9B3e1&{7Gq*$hvBBfJ^=Ai0+CPGEP{aP==T5?UB)@yzSohJc?z;>?y) z;(=r1L}#8D1#trZP!4=kpcFkC8+0`MCau{SB!j5WRB&jPVM_te@8Sm83`Y8!Y_uQr z>x)P!f^Lo8c5rJLSkG>)$0}YxQYh03*`qf@t6Kxke;fHL-m+k_h`I! z)ncN|qVwch70C;9JvuDTf`w>50Dm3Yk(uZ4!GBLH%rsnF?l(w?^MyVh(XRnjfm!~m zG*LiYW<-WO@CVa{zVBGbZ*24`iP=Vj3rw3+>YLdVd zDwPt*(YCxa*i+zj5>uv(dzoskR5nYE+)9nHWia4P@p(D%I*nNO~mBJ1N zw;KamdHz!?04M6JaGA)Uw;cULv7h&?YN?i%KY0w?d(i4@Tso`Xf!Ak!8$ju*5MYoj8X=#KQ))UPi`{J09lR26ptvBUlbt4WCoK(O&= z z>%UQ>Zct(~0t~0QKDedW%HI5XBl3~O8_PKX;k%r{3|>Q%0{>nRSBg+-Nq;87L!ZqC z!2xQdkjG3rLnj;7-e+g0$J&tJRFjyuNj&Tzn~MoG@qrWDgdT@6Q@4h-cF|AdRqGo8 z^p4^>dK0h2xL}QZ73|vYH_N)faUXIp;JEBMsgOXA)~h!xRx?%515_h zs$r*orhZZeuZj!o);PWw6kT#7>P}b9>v}IYa&Eqi4lktP)}pq!O#|FbUwp<9(QXts zx8W#;3I2q^hliM2`Hl$%jV{u@zFO5I_OQLMCyy^yZ%CXwCUDoMQi+Pn1QveBWxnoy z8y>6{YbL}nu-_R52GmhwmD(@3JNWdx%qO|M#|>WdOnco8$z>wfz!*dd{uMsSoRDET zj|uDf9A)H)+?(SEKrat>=Ar(O7ip-e?apz$qNDCdfa+M;i2B27Yh{W^Zv!yiM?uKs zldJwnvm;b-@rG4gid*oVcCTEUn!X`f>hjpjMiD7sgN&BP`?t0lQXz!$@JsK4BvL!A4rLBxCkp5Af*cwN z+mJQspbfD`J3fq9fzoEz+R!qV8mn0go`4FENS%cwa)Km|&8!zqfQFPe>^<>EP4TGe zObxzgk&NW_%gUJ{r2?M$ZTL&3&4uyB>1qT9mT`ry5Tt>)Br4G`jf8~874Te@S~_mA`TlyyaW)WpG)zz|059Y_{Boi| zdFl9J&}`>X2bF49f~j>0YMI>H+2aIEEE&~qixEQFGn)sZ&Z% zxuk)7d)D+r43Z=B4+aQh@t)V0i%nhL?ZwOr5kjeTqr138ZgE$jP3JLyML*Z@yp-C^ z^D&V(9AiB*cX*Z5>YCF5);&pOst<}fM*;{yqmjHv?v57;1JmAOxas&1{hewyXsXXn z6-=b16a{ZDu_yIQ0p8bAWD`8&S#197K=`t@ThR+Xe0ZO%|a%gxJ8sJ_+; zEl#c)R_S;D{9f5Adu%`va!T?-5cW1V=-8t4>ZeR1HL-D+U2(u~sE;@25Xi=TauynH zJSzfz9FyZ%Ss=@}8su#Q<(3eiFIo}XC={Z%s&B~5# z@M}M|RPcUEvt>Y$^kB^HVxKyscQ%A3Lf(&i;8%rX^ICr8F-=BcLz}U~Mm=|sD!x&p z!`rDX(6=ku)@4IvPKVk1GlRg-X1tbP?6}GZLC7uEzK;l?!92nSN(clv(5%Ke^`1ti z{g~k@T_RYpVZPekYE-Crj6vTJh@SL_WbQ^0r=}+t@kJ^Eet+^`4vH2j+I7cWtX*5B zJ;rP3IBm=2Z=IBnA}{!Zy~r_cHo?7HMQfq0&NR{5EU`!)Us8{G-4~UbP>!425t`_Y zX^dJ?ougX+h4kZrWhsr`X&KlroM0>Tb*j(GBCVq$5K1^kxYg~4#20Qp*wX^;!4cU|b`4a|K~3XNebqQhy z@yG)nZ|01Ft&+2hTK)U44n#5PsXW1s(3=$&&%ll{e$_CoB^oNVMOioLdi4NdjO<@$ z;?P2_Bx!L(zz8vk3vTDi@s^$=Xf|)tGj&miHERpNvl8f;y>nxo1pnG-cfB>Kk87s^H{4rv!mT zelAVFDN8J=-{M7pbCi&QDnZ@qNvY!0u9*&2!h6spbXt>0qH+YiXNzoK(+P?Ib2_@w zM~9spJ^0P+NHrb7xmGL*?c#MFmdCY;R(sh2?x{ za6#4+tZxzjd&7`z8g;ogHYTxsBD7Ar$hzd#=A)4r_r5g421l;Sx?_)vNYn{;2_H(| zqJrP#YX|*)v!`C%Ts9t-PTkVo%|NuV&OcumdWX}$w}9f*aMPN8yYT!HF3Ivdw%*3F zLWbbIJw3N(LC-Do;dL$S?n}G<4j7dH@Q{A?fBt?%lzMIXptal;8?MNwOPyFhC(qm4 zK;07Z+F~PrN$PB0G__dno=m5=7s4p(6Ws#$xI5x)sK$OOUi^mpWG`;-3ElQXAG8Uq zvdWX81=EWPBFcrZUtFEAn5JVvlx$1J5A=kyxfbvP^QkXCyGrinFhohw4_=g)qvHVy z1(~Zad6Ln_ro^Y&jrBosJ~W5a8!BIN&I^v#gMA9K%a`nlVSd_dTIg2G#@^k5KIP2q zfXx<7)_LP2-1u)UfvE@lZ2*vh`&1F{STAih`dz&4Z?e*31_Re8U1RoJ1+IomOk$T; zIx}%p-^xkHxUIng`C`Sdd{>v{DQtwI-KPDS_W0PhIS*T^5YS}@iD5PL0w)NK2X}6s z(tf13bVqiX3yH5B?m(2UT5LCZk;7(hs4&$k+Q)hGUyG)?HLNxzEjQji=@}vPdgmul zs{be+Tyy=gvCUvs13#!Enc&6_@T`h?aTAxuXcaID}k6Nr{i|oH@*S3lcngH`3mD+F^ z4TjU3hSZ4eX04cXm+{ZHFEU$vB$!j-%i zX>J4Cnr~bB7H&8C=wTFHyQ^U&o~4Gll5`+RVsEBpj=T1@h23^|UNkb&=LRg(n_TxI z%Q>veWu`zkw*AI~6LC)_s!2JcDR119aL% z;q{0O3LlzDB?3;A_3?Tguwmup@WFwxXSnSzlq6h7P5T=Mqai%XY#G?}Yh)z$Bq_kXi4WwIg5g2&_ zzv%*T{ua+zXaRo3YKIID?8n73YMSbDe4rJHE9yE zwJGaJOtKb6hA3;YC8xzwP6-)d?8DfyFQJVRlWm6VA|w0BSh7FYZ7FTv*X#K`&tLwT zG57uXT+91w*w9l!*I!x^&w04)dtsd+2l@k~3wnzemm@^{Srq;S_xj7k@b}^JbB{ z%YX4OA62sqPm#}lL3n+)9Q?h|Sb6JQUDSEP^O2?kDgU4`l*s;~zWW}plDH}M{z;W& zUFh3>&QXK08{1xvXa79bboFrE-=kbv)4g$xBVN^F1)fK#QwizYDSqk++nH^;t;gMC zcInd1rd*O}CreERgsmBzrrg$aHBMHz|I2n>maO0vr}w>Kk&`Wpt{J7a3w zKZP0@TrX4jD5VmYqSe(ug=HA4(MvW8QKE55%ZG@CRa$)fpc_GRzxtHe(@pi8kykTP zvq>)+p>>syxjd2BU0{D%kl0gnTf({6HfF-?GwI`?d3-~EujRVF^_V0=E?$#e#}*LI zS<@ve18f0=rk`%N9nSQf>rZ@c+K8!Tr#9?qFqz1erDNQlv|&ZcdTk>m>f=t3!@Dj} zi(S*ZR}_5g>{ZM+YO@v8tX=mw_V0|({50xN_OsN9hq71afF{d{9ccV)C}D3sSHL+_ zao{e=VrLcK!1SBphjYgJs0Qw18;hI99`@%&2wDztwO-tvnFXpE*^_EMXxnfupwUu2 zv#HpD>x!xKsrm!4o#kASuGSNTR^FuQ* zjNdD?VQVZ(Wp%_Dk^VO5z=c0mk`rl=i5+<1tjgETE8lu6fK8rkX|9PBgSmfbu&EDH zAonVkyW`JLGSQ}ot>!}AikpXLP{r5lBQA@zJ6A|O2-@D}(PouNv?vdr&Kh{JJ(n$o zrV=#UpFa0cHc`0kYD9m6-nQX9|L*2#rNVJK7v$zjTOXHF4~i+)mke~RwGi#ZwQ>X> z;Gl;E^Q`{L?{d*xDsxVM)a^@{L0DFe&Q?suuD>ip5aYRjx$*qcbn|}evv%*wBP?_O zWeqCjc#|w5N0LV{Zg-zsj@feu+xdq+Js~+0aIIkGgBVXg(M-fmGlMkMz4fG8=TZ0n zj!xb^m)~&US#&zCzl3qnW8eb4Ie6@Zyrc7r%h$$lz1+~b174)dw{DZ{_m2`%3 zF*QDwJ3(DSV!GH+YFByd)I@-sIE&i=rb|znBJ3k6(z~^4dh*|azw+1Mcv*8a1bpPj z%L=qCwc>vtatEt8wAE@G`6Q@Vc7=S6{HGi0@H zGwq|s4_{U+8m85c4@&UOO;S>o3^n?d+@30`DBgf68td~4I>wb8lfGWFtLo!$Xg+x8 z@gPd!q*Qe4nYPhMA*0Z)d;I^syX!tt6B^tdKo>Kb^f;93JE77@63bfk#%^#pG&o1c zpLI`QSCER0VZN2L?roKkJfr#&2l=TLEuvl1Cc&aZbw<36ed8h%sn+JTiPx)SZXYZ- zq|NbPgfI+M{MB@DrqCYUR|vjM8^31mkc9HZCyUkI#e3b=F?5Q_F-=-i=rG{&tdIWC z=>cP2|Cg5fp;?&u;6<_+A3J-+TZ{R_^g#dYF~)5)>6-@_+}6b&j?;z^xop{Nn_3u8 zvhQ%&28th!R%P5+cAr_d=1qgwHhGHpY8<;mt1U}#aOHrcVXSkhbpf+*Mis`6`|5|l z2SBpiga$BOwDWB12ZDrgCxkYjqI*K)|*mi1YGV=^r1$K`rgP97KK6g@CE z2zKVcgiYJw?)Y*2$wtVKpXV7f1(H9(7)IhjIWLlgOIMgjBfFm%ug0Aehmr;il=qk? zj;lx~m{a6*pDI;jB-Z~!T}UKtLkg3SnP$g%i9L6Wi0a9Y=&v+B%1vB{`k`8wz@o%r zk1FSolN$H-;6o?Pz1G29K@CN)5%Pw&4YH$DP7*YPpXd{jL3Ta(g*q%4zK+7x=IrX=C(kfK+BdLRrz4Gx-YaEg&?YKtnjs0JpWKK|6eUCv?bLLIuerp4W z?{~hPueL&lhC!=q?Ns)Tgu@ANIm&Q3?1Q*coQAtuJ9n^vhb3c7F5cnoJJcRX3hg^> zH(w8f{HPRxcCc9Fl(c#r{OEo7QFw|^d<3O?Ea#pgz39yAe2z6Dapx=){XoWJelj3} z^5@5T9PrH+WTPZ?uDpKMxn#3Gc{6{{2H=LDO@g}Qtsro?e!8AA{G5z-ey_VIca1zm^|9&wg$!lJo-M94~jCPW>`K~kP^@TM5~nooEiuDiAA5Y#w$comP&eaMMWsC|^g9wv zk7>|n`X)y)IR%XNW_oHl`lp6M4Ndw--H8^0>FzTRVvA072f?XGMWRy8Wi@L1{$!*7{>4zoI!QgNt=;N_uDzXaj*WeMrB4((+rOA*)KFcA4oNoA z#>WkvU$_O1^si2K(tc@}^~{2_H8H*7GPA68g31;>y9*JJPf(5$MrQzBn&aWBk!6|o z#DsO6;OIh}=r@ljy*q9p&`}hpgG{H8476-EKfXIMw;j)*_F7TT6E{@*wUpF7>A6qi}v`XQ`_;Um?=KxLHM z<=KageE9(pK0Fo582aVG>~)h~`9ks&=~F|Go!?spH}xbiCfj_S8E3SZOZK4;5fuJYVpWtjAz)tS|HRh(MhuRdBi znHeM;>Al?WF0V09-{SO@7F(7NZdB#3(dBn?Y{Wb(0?XH4*YAq?1($H^1(NR61hAaTF*i8ZlZ)fmk-q+NiX1S>qy?5K zAM1&~u2$`< zJXE@_7_jqcWFsjw7617Znr)13SdI+m{j-OHl#=J1QTFvo3>Uhde|q($X>1##+t~HZ z=IM>Dp)J+t2J+JFtco>Kq~9hwpzMx=&B^Admtq{xO^DQ8iQ4>r8@jewZclOQ#@VFMIVSsm)kZ6dNJ;V8`}X=ZM*mQhS0@Aiker_gF?kN5a0 zk&A7ZRK$r>p|Vr)q6y_T{dFt+WEDO*q|IG>9e$yoy2nbRI=<$ZAUUV&Wo}OQJ3CH{ zQ|um&v_XAxZ*Z?H>5O-43(6|QE_!l2&nF}yAq6|EG&z=xa;?8025vDQ6nua;Oa>|E zI^?raw#bcgQ1`!vxeB!rm3#K$=-NnHtm^EcwnGXwvky8dPS>VBPdur{5)`<4Z_YEb zAWBewH?H;2nsGJFMx|yA%->dOZyo4?YP4e(p5ptBIMzAx^~=dno=sBpFI4I`Qk%`k zc&SX+Q}i2gDYb)W>r+ZSa!*aVt6Qjy!i8f_r~nB*5y}3BJQISBYhScWcIxqzZ02%4 zE`e27j#c``t8Vsh$Cc)G-N2VpGrc^Y#adjT7s{iGrz~-E^VhmsE=E>tE!J(tUkhzfwUFnzZI7drSE^qO`qHnC*OI~d53ibx=mf< ztDARu!%k_N2}YM2`(@=*_;k;(dwjMk0c5 zy2h{FdU-_A=!3G6+UkdlxS&dRPC}xIRf^iX4!w9**IE5a3EtSt68uD6>@By+9e?-#R_8#Qc@GjhTvo*+Bu$?9LHG_Q9(ZIsm3HMvbkv$pXe`x=tDyNNSZ zzy9920~LGfGRIG?tFe!_hix^E-eSkMj^tkS+Ns=mcd(_eZ;K4;*P$6nbZzMyQ5cbZ z_0lUNrcid2oJ-2Bt~Z>64R+;O2Hrk-GZS`tM>D!5vTk`Ostw_SoR2U=ArG58W36pD z^&b9>?BPOJ(=z!YSzO~K@8Pj?47Rf^>obA~UC$;yNbDsFy)Oi{RT%9|Pl+>o%I-O8 zIB~OKF0wGZq#!3rDL2h2I75*y68Gg?p2{5xQNK@0 zy)5=_*xN&*LxgCFlRTP|vu*ZM@pk90_~diu(-guRDSN~5s$SPp_v*qI;;4w^{yU0l zJbVS8Jn?sY^5(vzS-7dmFv!lX|3uQAD(-T1)jT>detwwQjQv%6`GK%wCT00`Y0<8A z%xzQnh`Fyg-PzH)ZjZ!|a*RfZsoHi}_i&mv;#c!$gE()c3{J**(3|^kv!d9jMlk;UR3 z!gB-PO06(`$F5H{vvo)wI^RXq-JL32d?lkYV1yT;61mwkdE2ue?}VOsp~BGWs75LW zN#&Ws&^y}{S|9DqRUL9T%jYglnQiVD{0rk)O1}z>)g1UjW!Ig<%H_Z5qM`5L!>-$h z=~_L#GLb*s^3rwgR@3Zf-hcX^dA8hpUd$~j{rT#fihCW+E5=9yV^4QZlK|d6V!~M; zHYI5pa=aG^*uGCX`O!c=zO??C8E;s@8wEn#=^Ktpgj!5SV1%NzbP@w;JIeP>%cFa@ z&HB>^gZK09c?YS`^IPcKdCBFT4;|)=oeJIL#}D|PAP<^_yn-7@9?D6zv!+s;XC8eN z?<~r1Jp9Zhy;kYDoXXHTTX7q9gTFpCpi;5pAv*UQo(d^mIQg!xMOpT*kh24EQYtZr z+pp@8_t`iI>v@scwqF=v9PJECt*s{XplH-b9oO~HS6gI4X)CXHs_+geXqa!b)W3P< zRgi&Ntr9(d!XnG#!A<#go=VOcusp|G5GE?0m%NKWv&;_{2`F#su9=Pc^3JBEwR^THt3KbKwmLK9VVN-d z5PP2XxIfdmS`7)&(kB$VD{M7f9rwc}`0yR}&Je0#ifSJ?@cBi8;>C+u-4 zHu3uhKBN^?#K*eSJK~I{wi3B_kHIJ;z9%3;1C8UYJABRS?ZOTYBwx_dJAX{A7E>#Y z=Y*#5Edx555Q2!aLrLB9tqN+{ur@dCPE-%z1`YSKob0Qa^r|k(I8JRa z6UE8DDUt8XCgvH(dus%lb=XAymtgO4Mw0FB>f2+UA530EseO5`-!E?0q1SW5NP_3( zKT0$D^ec1Ada6E*X+2_z41rpO8qM$|aR(Aov5#&BS~Jx#tc%!L8^I(yz)&)M)P)Bt+rxiTYp(3SNwygN8s^ez3{7#I*?js8ltc_ z1AuJ*K$PhbhgIU&FU6Rd{1& zN=M$>w+v=&^|@T@)^ol7Y7=?iX(b99%_QE^i)p$(CcoHCX6`1w#_iaogw`WF`AbS( zP+k4nx{^ArvI-z)^(8LA$0pm(m7Vv@^2E!n>RLayd*;f9_5S!Ek?qELN_nfNb(vXB z(tNjLW!?Ah*#@%{I^wqXN=%vArnXYM^wr9PG@R5E2(*{Cn)qy>spsFp&U5jm{QnKD zfQr>lzNv71^JfOd6fgXs>}kWRbX5OiyM`!QryBhW9|3#U2GU55PP6N-P~-#GD6yXm$maRbF$ro3;j|7_>plYfQ_H;Ys6 z*FE~??PddVOYsM32hOSUI|^dPPt^|;t=CuMx@+)4*3rgJBR}>$WqLIwvj)!(X3q(u zuV(Z#ZmtCMGC2>QctZ#VnktcCmx1nSWjRyQ^R36oFYhVI*RECyCQStFD1Y{-X2x7j z>{3<2t+b+T7I!vG?4f|tX@$Led)`5BG7@|Jl6fK9j4(1^+H&>ooQI^mTJ=cNpq%h$ zZr7TMb!nv4C?`E}8yT}4y@?co>=1tF@G6mnM>+Jt8Kq}<1+{qvumt;4kFPu(&bu?{<{5F0-5ALH`!Dbb5-1jkqPV363+TYifS$i z))+0xq#y(0#8ati`S+>3So!33+~Ac;veWj}&64eD1WkSN%fh!77wXBO*3avX)@VQP zQ8iLyf{B#PZl|agsnFXnHb3_oA}yp|FAc9r|9n%oqbcv$wv?^wYE4gKwW!q0-P!I# zPYc85bm?lH7`URg@lv`t+ojfh+hN*@ve@Tq^)_jSZ6WuAGb;iM(6#Q~G2JHDCoVE! z=#QmFlizO|bcl)z&8%&HFJi+lm*G7AteQTR9C$s4?s-A6rm!G2E?ZI;njJ2~iQK^F z;@>er{pvX~(fRE~)slk;pQaH*>~oW#1AA12bh}*>AQ(H}D=8Q2;qyu-#`a}QB;I*x zb_1;xAIsFB%(Rh`>EQf9>RO19QzUI8?e#7G$vWLNg;jZ%>O+*shMh5q8^*GZIgC=D zWEhDFbGei1>KMf6I6-CITs(FC=fv|Fpk^b?$dqAdD?zFzJ*j^9!K+^f5@!qUnFL8A z%3c`!rM`EyW7XuxG>cHPXb;k$n2*R6au=r6-XH+ZJ0WmaF@xq+6w;e zozClY?>x|IQm;GM3GVp~&2Z(qfv4sAd2CyJ-`}|AQE5~WXgx(;#XE32MI!kywpr`A zntA=?#@b!A4h8kN!=Ikv{X?`K-Y3rTer%JX*FR?%QV13R>^BJyM*0E|_VDKiyTg2` zj!)*%apOC;N?zBI#vuMEa^hEN(w6BqasAir_4Oz=C27Ox5Wb$Y7EV=i*NwEk{K4oG zq@oC#!nv zZJJ3pzYmgwDq}TZ{>#BTGm|)RK7XeoIhpxF*Jk{&pgdcy{hZd}??^KC27PL!t3&%= z&BkZOLvhZEOIxUzi?eYvXSJX=k>@UnW-ut}(KyzW>d_Q%rh9#cB_A$^NRMq_xfLxB z-$&mX@Ci{L8sxW)5qMs`+eaKR7v>!l0znaJXCcZApiXQx+H6-f;NmDmxrYoRP-x@uFC4CO#zT$&4h* zOUNED$Z=7+cEfO*eEmgOSMR{r@f$9>Rc{D!^|Jo^n%R(nrpWx{ajtzti%;_<<|aqe zA7&vGMtNtr=-S%jnGIQ}wRe@&3c55UefB;RIwJL&{j=d5>CmI%61EfQeB*`+F;0|A z(4EF6EzxeA^zgv7q_KB#d}hV5JTn!qocup<$sOVRES-N_>GJ5nTuYx$n{g*?aP;Qz(YZe^L|82dQU<}Tam`z&T>*2!I~%J zH4L=(zm_klpUu+!^gXT9;Bel5wXUnAx@9MM_~+uHU&z!ywS>|A%94-+PR38(U>{x4 zqJPC7wY$^yT9F+`gEyo|zP3J%lezjaIPaxv;%by+bGB3d^^Y%aIZRsaKo<=hz7h!q zlj%P$5@W1o*_Tu{Rl3p}Q#qX0xZ8P@tmW`<9taJy8+?`4opNr^_A5>Q znNCsC~h_$1PIuW6cHiJ$`fn6_)vBw-xq2Hu^QKj2#e&*kSw z4>h=bkUKU6?%1h5-y+V&NkZ5YUvub(*`L*d$l|i478$h_L4#SQ4O-q2a<0Si{QfBx zN*esVxf;$Re0-wUn)P$jhWsMl*!7ib$W3Gx@DTS^+x;+q7Eyw4%`~-t8TVjS0xgp^ zH2D`7*=e*{JcgA?r=ZJdS+CFe6Kb#k3XxLSwPAiwGaQ6zp-ffa$KIJA{&fJOaUM-? zGcLIT>5%(a$pHJsU?cJ3eiK0`oHjQ%Z>uof2HSYfEDmygxNQYYpg{J@EbCo@d-vAS z%(z0-rHPxp$=>f4_9*hW=eH?{NWu1xVzID`4NXe~v3UEve+k--zOAfTSeizt_@M>i zM`3`;Y#6whE0lm+)x1ff1 z?Yw(F0~_T4&+Ol8r%m8O&3$btHn7RgCSWN%lr^vy>z(%@G`+icsCvLH??R4?!6gTi zR}r*`OJ8r#ShL_xYFp`z_V4na-f4a5LhMiae_0`WH$uaI7cU|7-}xgQ$Wm!` z-p?dW?Afk1B}3z@56wc+Eu>g#lS-rEzlg$rv0RBNBw2*sjNo+RgDm5w`JDi!=l)$< zi}y}{^IycC$xDrGB94)%Sj%ddISacZY}1%1&t^$BfAtr?Z}@R`FtYgXQJx?`=*~-y z?r+j=Gwi*=FgIB%P>Yf9V)^W+P%rktm6!3$ik8>Ak!zwRk zcDC4V3+HiYY~28%^{RmFo987M?(`(c+Km(BxtVc(_mVVjCiDvUaH@yLvl-?!7%g#R z!Zu_XAo;+p7(#+3qEsW6LmyXfQt&Kd&6Phs7rpFbi^zK=Bx;4VJ~U@f_|H{;_K?|H z*tDVT@*6%0B9rnzpSM-+-0E^4i}s}neR-4FH`eAqUCnTbdE?rTxL(X^WUT}!?0k#Z zHHz@N^urtZqCGGNWg4nSx0M=bt>(wEE91B4~8ny#}o9-9B3p9s_ZrNGS8~7AQ+e33#vm4$)#^8E}Yl~5b zQo=u6*eO>J&#f)_7c8T5`aCk$>ftL_k7w_YsIGpHf%R$e$RSQt&IVfAf_tF6k9l#B zx)OEoFAzgFU?Y%YS&a+2?l)`jDv%uCW%bo}yj7INqh9TCy;>NF0=ZC2XuKJaPZRMGp5*O9dk^St@zFhd zQH6f_&Q;W$`zPvuHT(6+yH~(ECb)k^A4HQS;`oI^1ks9WETQJJMFMS{k8Wit^N*N3 zEMXiuaxKT1a%76MQtd?4*Lb%Z_enjYD^Mneu(xs9cdKsUvxB;t);h75GoCcmARN*Dd&y7S|2?=8T0)_gcoQRN{k%RboT@vqVI<%aco5~i$1>0`Q(B*t5_jMQ z47pYvfl*nYegDqe4?l}Co-wDLPwbL6;L#!A(-gTUc2s4vKUF zu@Bs}4+B-9+wqujdb@#OB;Q78vG?XYN$9I9di~q~3_TyJ?1sE`iBT_L{oVr>93*)C zkzH4&un%%^i6k8J3f%3t#(M=%WFSK<51&_pCEN#9Z)f-Drx z{@1nHb}u|y8Pj5`WafR?K!e$f6kyEXLjiGOu)n{bEq6hIASda$)Q~ldz>mpmNQl7; zD=opxK_teS5_c0(2-t7&t=AQar$)X&Z#A+<5Ac(35}NK{anzf^0KC(NZ~bOSD)HN9 zKXBXKxS|`dmh>lA2y@<37A_T8VD=5IBOq$YTf$xKs#?6OCSzETaTjiKPxe67)}Q9U zM?1hawEUwNl%afPC60MtBO(fKl1%v*ZfmI#?)z;aH#+|zA<0w2T#>=zM7e=0^wTX<{t*MQKm%v({KR-@aZ#TCPd-d$ z*&vCmYnO1+_s=YDV}zw)KX1R}fdBqosQF-D!M_(3Ly|%8k1iAZT>Q7*?15n!mb}>6(s-n2mhn2W;ow~ z4QG!w&F?CnA3?EbhL>3!!0j+2Am*nDIbh^|P+CyEpO5J`FPI#Kov7*Ww}K0YJr>hn zjAQ2qek|$9asvg2|1ws^koJjW(*w@45u7HmBQ8U3QP z647KI{4EDmvB&{MHGT5e&KJtTAa>vORV=g1h{OB_^T8G^Ygk_XTL0INEZ4Z~7hp=) z??mwDf}010S_(XtpU{u^9XzJ-k{%pE3gYI~n`K(} zfBndEKWGIF0rtC&E{7zWw=51|_*p3te@}iKlw^v;Z_k%5eBu7)7t(N4E$!R>)v+zx zY>WJ#@3`T+Ct4=3|JRSeZWOTH=*Ljuc<}=>2=vK+z5_@E^DNovWrp`h!1q5;z@J2E znXUPKY{~+re*sK@t@*!xWVxU%+8RUv7GlNavi6g$ncrA$az6cG0w@y0GW)$`MvzIU ze-fpo&maY?3!Bd&#Dl<|`}>TJWv=4?`jO>=c7X>Fs-9)0+aFS-;Gj~3mXHg zIAXLr=3GA2&Ba$?qzStEzND%X`b+oTxVKDT?qqxN`w9kd-ub>IvVj^@+3WpG%@~KQ*YGhaIx5D zeX^Iu8uiRNgN+zF)_+rz1rK@Roz%O=8I|wBk1(HUA0p5cq?1d$gmH0?GB{C}CS=BQ z@e(YYO;Iq)C;L&v%_-y?7Jlcg@8`<0`aw~(RK6o=T6Gea&D(B>p;Jn-pTjAMbj7z9 zttDVr!*p$4&6S(B=&nTDnfB;RdHYvkG6vR-=i^191n4;RmGrE$CCq4gKiB2CgSH;P zQ9J-eR07^I;x4J-`af4udPnzIL&e5g0*w}sJa_oaO8AT>@)>PwY6J`X?8PBmoEzl| z>W;~d`_{)j;h!DM_~mB_*uM3pRI62wa?5gADKJyJ!@JI7Tfz#s!Mm*c8QhM4tb1-T zLbZ~8J3`$S?0L9&JyS0b|E|)>G|zy_5z(3KQC@92r{GS{dGR@Nj9(Qhe=H2h+hX2E zvVZ#L{A$wc({%Qc%EmP+q2ube?)KeFePU0XsPO!>`noTGNSjsXbIT42qUCWxfflF; z-R97iCNvQ6ZMnjW+Z=HnYMC{l=fJ97c~A2P|a=Y&a@EedViqM%J}ppocJEe8YQ_tGuL9{ z;l2DZTs8*=mOb={7$OJoX7#qFU9fC$7dg~#ow318clOQ+vIi$N*+85&7d?25R#_!ScM{kRqw@k1nhOBI^ zkga~WqUUl3@%W*&amUDK!{h4t2Ize;J<^-}D^c#{pKi0GQsawFISfxbO6I?7y16yb zCEr>{9@k2`KIPVU+v@!Lv13riyv5WL&!eEW@xI4+3bs#v8(o)xW!Eqsw%8vAOd+g} zdQY96><|3ZRS6=Og7uYO6VjINPI>m=eY<4ybOn{X1-6448O=#K+h1eFXh|uD_%n<~W6hYy8AB&s9UvBEfLIh3^BwQw zsDiOj3#3Dn&PUuvU z%`?wK7?3V&YVkDgg^8MXn8He^R=Z|$yVoIt>)Vrw-=kiOvCD%|S&G}yS{60ZNC>vT zhH%mq)OIQHreNRH0jQIDr!2=xxsUI$9Rcc{G?)J z#VOB5&5RmfQ<|?_Wu>p{`S*2}R)9;KNO$YE4R(ecuH0MvMN!*&MFDY4t zG9N76f_nDy8GbFBoLUZv4Us*e~J)-Nk(RD(-)=8}`SurC`r~fBGe;S~Er!zKXiH4ayM8Y;*zo zMJIiVRu~{oXi9rcCD3Pzq=?+wi`N4C$(G{mOkx?MJ?W$)al*h)^OX?dVRT%5tT8gU zR+V__k*p6IThS7U-{b(yS z2%#^~TIv8`;R+UroKLv1K6gHjMCV&Yc(*BB_T~++qlE5;7c+w2BNDJ}5wO4&yKXV) zS_W3#fC_+|moa4f2>qG_M@sXfDB5at#vX>iweUO0-nC`C01S?6J0#1iJ;a)V{Sq67 z$;b_u#TOpJr1q*9!no&xKA?aQS8O@nR`XG@}c4S>b>BNoNGiYc@T)4U> zy}hsq)=qwVn;%c%d%g=Y>T3IN;nT!aAYi|}#CJu7+6y41-4U*a_i5`>CzU~ z62cF)o>pR#r4zbm9T|z-X5S6VH^S^fqu@z`9_suFayu?uf-40e)^Y%U?x0N z4S9To8i|WOcqc{j<^#->H4L8kcaMHN!NsI(7fuIFH(YZj1Gc`Bdu6qqpgtj$zi751 zFvBhD1=TMdOcN_wOA7ehCF%yTF%Sg18IWBP&%feN!M?hwQ9A%HEOrKyF_(!yqA;|3 z4Urp?MS+`s6lY)z8``MPQu0xH9k@fJ@v|+-SjrKX16NTT;X`eD+`yy57w5?b_~Ihd zh-`F*klqLbTJN*g><}x(4=<^{o9K*tI#&<(Ec*SPmqO2l2b?ib(@y@l0h4jzMSPDV zDw@`_3B~bbFcp5n;eQ@!Q3|baOAHqCnbeTfs^yS_*R<1;ZmBP;J)&T=QwlU^9Wg=G zVtF8!?SlkzI+|K3*h$~)!m|5@{tYi!Sfb9-7e^J()haERQKg%GHr5BKjEhuAC_Q9)E zw6o(Jfu;H1AHJ?hj~;rMcONS%x|4~cDjLGpCByfP!B*8P`3uY}!V+_Dd>ELtnaG8b z^mss^dCI7a>{y91@mZ9u9|FhAn-P*KIV#Qx)eTcY=ci}F7d7Mq2((f)l~b_y+>HqdlBgQ5a=rw8UPR!lDhD`s1NXKr3dFpMpGDrfT!Hclw*)u1NAL zR9r3e$$S4KhVWFrJ_&ZVnqVJI4av8eBI25;zWDB13zSCdE2{oJlmeLm;&Ha@2g3@y zyBA50M3Jq+Hy0)Iye56JWD_b?gdZL_@OMJp1a^XFJqQVO|CH6L9Fg~XO0rqlUfcJ-DFp8t!HdPrQ)QcFLNqP-`FbvlxPQ&(}{-yzwRX zV7puuhcL=ZW=LTu=CALzA}QBHGxDlC*&chWDH;GczyYjo%&WuCz6UX5c4bAlL|Zj}I<9VaKQ7lQjHs)Fy&|6%qYsW^ z>Hfb{ZSZPmqO<|bp53ySu0v3#9MKDTF*Eq^C^#X2DZg>WaM{CQ4Gj%W@05+m_k5Sz z=;P!(%O1uNe)B8ltFG?>Jt6=Ii1`r&AUPsAqH|Oh%jZ;;jmn!opMjH6ji)}5YIEp* zucS{VGA-KZ#jyOY0Vl_;u19ZGo7YeMUHR)E;_0<&}52HI2$d4fb@r#A9LM+6t zqoG9IYZ$x1HZ8sG=QY{teug4KehR54>Zn(%LBP*Wu5hD2yb1Pc+3WtO$#(B$6~wG# zhr?AxT>!B0SO%^&XMGpHuZh^Ly0oZ{KAzJ0LBvf4pg zKTDmWuc~R++3@p?{^(_jQn6GjAVMg|dU!Kk4|#2MAAIEvd`si_#2Os4rh=N8()WBK zvf!K|G*1z65<{L{$3*D zpa~gJ^pUEqNNxg2uqoUlI(I5e-l^*ulyQ;+cq;P!>6SQGfTxgekcWNMo7cWm-Ar7M z;?klVaW=7736^W^pCwWbSES{7c0L7hF)00^F9RxyHi0;YA(5H+cLRhq|8C2rxe1Zx zyZ^4l1?dSA8yXo$>P;NeZZc>y^&TS+!}Zn=Y7d3FgQ9Q!y-;xzu%aAQP+f^+-w1LP zVRt}|gI>tn0TKqZxumws<#-J`!*bQ+L@eLV#G0%}OB%3MQzC1`TvT+%Q28Xm>CEci zI_7QYj7CfbXe_LpW_@SP1eZ5rrsxzjAp>*RPcA{|A5AWNXIeC5&-%BM=fbn~ONdZ0D>2hi$|=i5}0 z^(kjyO-kIwVaO3(vwCz13C z58p4bKTqz$UNKLPRu}ceE5c~{(}0zi=yX_P00$LR5=fy=PvrO3S;2H8>}Ysm0rf84 zVzA9X6xu9J-u=eY>_>1=@>ZxH51B7LJ)@JZNjOAtMTfY39JDpsLh1k#JJN-&cp==7_I+g|15L_b+eED#<{Ur!n=Z`lll18&fn3bTZ zmjvwDC29Ck4ngdDaqpL)0uK(HVcRwQ1x3bN+bC-eK;RPm%jbm!Hx@O|Tne#B(jM4~ zJmZdDT%PG=aKKw9_Y8`}1CsjoW4|IPcuHSEoV%0Q$%N|7lej~I`$jLP(?ux~^(K=7x5f+Xr}fO}#Y5}tB{ zItCaL$4(GjnhWsF(Eot7upjjyKG&)yRGjr}L`#I}&?)J)I8LbS16F5Vj+UPkg!QZP zwWGBu!Xh|#^rN{ch2d2*rg|JzPOtlonV1q;DYx)^u^}G4BeslHie&Xui$LUuw(8(e|KavaW%@f zcD2R(CwRTFaCH1%BkEc!YtlarUk5Y- zlj4^YXfa8R@E`!bsxb7)FRZw59k`s{?i6z zMnGU{x@E^b@#u_G=2s%}a;Jk!sD6`wwigg>4Uu#?Jt`PbMoWP!8%iT4g*IFZrp)mu za3k{w;}?{l2s07ymDx0FLBT7#nJ99x@ef}|=3nuG0anUC zc#fLa5qUY&!Az)VnB>2N)0QJGe>{9iJ}!m!VspiIs%s69mwS>w3+5ZDf!gQed4@#h zBsq!P8(DwhL>H(QxKlnAO*^cngJm`#m5l2e)E3-Pgrt$9R}Vh~G`)18zm?=QZivx< zRUQY;_%|}a7iIS0+`p{A*ge1Eg;(5**5a;RUTU#mhXQ;bMDEc=!1$9X1{7~%L>9z}&biAWrhrBEdU|SSHh;4qnpX^kc)2mz=Sdk2tZ8cjy zUKGQDvsRzl8^M5Ld5_SBOgKUF-<&SYhj2s9!J>)5yNj#|3!n(1^TpT@u4^}VO_rFs zU-2!_P)}27Sw18wu2+;#z|w#+^W2*!Z~nGeQ)3dY#UOCp!{*Hv(U?NBA=? zUtff%3znE~da#uh30GiE#BU4)+-of?35kHP`x+Bi)X0kCp6AgS#EN!zvs)HW9MD?E zgqFAjwqf6owgD{52MirpA_*tGqvZP7)kr-Znhj^DF?Da-v@^d1A4Y}DIp6c-QJeuC>ROYge!%+XUI0W>omeL=~iWe7x2`E#_PQ`K}bdJXn z7zszUt4uZNT^63OyuXQQT2} zQ-Q253O4Cij0QzOu<>Oxu+DpSl0(dSlSZ@w$}Z9SzXlmqf;f;w4^>V2bDv^+lP!Dl zUN+qnNT3N8=xirCXCD}PnYDQ{BP#m%ceA)u^Ds|mL$T8+)QAR`Qo(vVv*a%Xp=i3| zq;#E^yi;3xw*--!|Bpt(w}|Krd9Zmglt$B|Ede)b3B7NBpGNr8Sqb9aQK!4TfO}C@ z`F1btlxM-}BxhU+cQZJ2^q-i8Xi{F*O}xBw3qq-r>Ob& zqT($#H4ES^Qk&ABi|45|DhMs#63>i*%+Ds5Mr&<0oPYt!oL%2Ya|2pJc}WF+F&XDE z^e5Xri%@Pshck%pQ0IZfo(GTYl%!`l#+~o3!;kO!uJy}_QB1L; zWb(8nh8kfEB=BmVyhc%NMqGGdC~f+kL~bX}@gaD}c}4rJedzpaWWus1wM9l;RV@XQ z`=_(w`wdY9+Ng5;gO#?t(2n%GIC~*hUJzrw)Yx5BEMK#D7^U)HmRgxbUT?}-7M!)} z>@(@y+;mguSpgdQcSaRdeOFa?ZM|Y;()4{fs{2FmSi7nEX7!0HQQbAgDOepKJAh7j z0+Gieod6NXeAz7gMGBi`Qn2$4jsKx(k#J$(L67{1Y z@Qt?Bs3+qT$ZBPDddNm|OdTV|4@kqOvms`wHqmX=da;B+x3Hgo=X!>4!a8 zISI7Sc6#|4tMF$!_-t+R=gXWF~ztZ)mFDgTo+SehSO?1vqOLIS zsWY>mXe*J9h*OJ}bUB7zWFke0&VmyqF2*-7@vl#*TQOnmE}ls#VpHQqRmOAAm8=+w zHHA3;{P#Hj2YSH=ALc}8J0%;=IDWQ306>rcH9Xw~zC08|oM=&E{0yTrPW(P{M&#EG zMIuU_szkYYdI#p>HwM^_8eD=$hj_S{4aXiR0gnxb-TKR6Z}pt5oN(!(h%v5xSjF}D zoHt2Id>^LmmcV^{NsIlC<{c8b^c95xI={zlK6q~uv?COWs_oSzsCaru)4VKCcYpHY zVtK5yk<$NqDHHUo#Q@s4doWDZoz8pC?<(#JBo!2&#t-OhR3h^j?OyHCJH4V7cM~5L z^Jjq`Rv=HfRK^rLbFoa2c}%1tAskmf^_F@q1RkSj#D;Xw*XvSg8iBZB(f9tu1VK8H z+jF)$YZI!HO_I}O;T606hMySjRT;{t_ z7L;f1;$`C@1&n3TzZ0?;P5dU5cGRRdL0TZ#r{n2FMo+mGd$t#J%Pe3y;~L^bb;~Lg z%XBR~zv3U^f{GWudmBziKRgI+cjq89Q;ISP+{2Hj#1f^>wV9&71JIu^Gf)9hBITEaj<6TCexTJMmBD97y2a+-q8%N3vKav8pz zSZ@s0l<=X}k+eD*{Br%L0IvkN^|0T^kPtGwrs}K8q-J^H7jrqrFM(#|pLo^VZai-m zHbPGN_rb<}lI~dpGC9!@X6RtOE^0I-3;5y9Yq^^_W{*7H3SM>3ccin-tHN)7#=YO{ z+tV+$hnWY)s_zeh3k#^es%>cYk~q=sOQ9Z-e~ci3KU;CIh9KtlOeiiB0$vY1rFvgN zL2Znt1bNIW^qnVS#8v-n#7RVM1M&ovu!Do1um1Ohvr@9C{N%;Rev=Vp0^oG1_GNj% z4G#{lZ0#=K9CPiWCe<Th~aG77`s_`_C|_{KY&>Ye;-JB@2kp*DhJ!*)wcFA;9l{Z-x!C-2+yW- z4;siiah!YxDrhMD+(M-3TtS7_Uhf?|s7n#9u|OdY0q+;^KZP7Q_KyA*ukCeKEJ(}t ze3S?RryHVUS5*wfcmLz)fyggPw5e(UCl@RZztFP)8?uB`u=uQT| zOSX`8aKa1YT2Dc(&)VBeYeP8z<*cEc?fu2^R#Tq{`S`|5yP?cej-Y4oLcqu_o|(R? z-OTI4mDJ}{zoUP)=upVeGUZ&1t3c!)reaTH_g-cS%~1BNIX)fj=3D z*HmFHNdX-VA!6RKg+la=C&|01=I@P$^6Sqash>ZBLwnS{ry1A&o|tj5t;%B_MG2rg zaOw)Lsn@-|qi9|MrxPom@ST+a9)syOJVquYMRRaK8o_HzpCE7a6+d$?83LX86We%# ztej}0zwPYWKYV+U1@Xbdy^Hyfd&(!7=Vo|zNsmDa;KPQLRyNfQsB$PV3g7bKISb0f z{kuQ<2_=HaO}k8e{t57e^o^E0rvoo7{{Im6)d5ZJ-~TEq1|cdUQhHHPLL>&#AOfOvOoiZcz zsF+^dJ_FclsCs;NVb5`J`1_?6$2uaYq_$)KwP6D&4AVfL=wND_hQFKvac+MxNHL4&ol8@lrhq~WeFc2T<{^#2F?c9#)(C|z( zW1mlpD*D_EY37f;+l2^#ur<`@Lc{wgu+)6Nm*Sske0n#^kdr^l?{9~C3ulDY>^E-S zhH_Dp{j91jiUYQh**~$ncZ!5}BFE<{X$(@+fROg%yqzWt@t@bClSu-`tIHQi!G1zu z>kBt7I(m$Cgl*H^-SKaCer?}JjaR7^^Y^E9C<@O9cxkcJAbm5ye*cx$vmhESD!T3l zv@vN5bzJRaxqmPOZ4WU0L8rqd7j&A!Uap=2cf6 zS0VJf91;5tLcf0{_59p*R3x|?%hwEfP0K4{gKPA5xaj#0@a_&1tZjq%7B#m2ieQK& z0dZcOyF-ucKbQT<_mfid&3IyiEb5n?_$Nh!YFI(nbj2LtR@yQ8)eAZn*45xue|@;; zAlrE@dbuQEq4Spfp?dY(tHN>7rhNN`e*fv!`b?#Ub!1XBDz z&)OgHsu%c$T6eZr)5r<+om3{W_zr=Uy{`_V4BMA{T!)M4iI4LBY~8K=FhwOY&vIl% z0}?J+LJViAtJN9Eb03gbG5HgpmeZij2}}12qF#T>;ndmh1QE0f{|ccl3iZFee>x^f zqJYE^BRTY?ZO0S{N1@*G{kH#;tT`G5nr9K+0n@G%yFMP7HYfv*cKKchM7W$HvEWtg zLi?f9aKX?2(IH#i#teG-i?SHNK`|wiy~P%>Y3u^99p`TSk0<^yPB&DEU;&Vz;LEa= zY5~m;TEY9xIuzrFn2*cKTTo>YKUZo!wD_}R4$z`X#L`wue?|)Vgp1arq4Q9#v8>3> z)55Z&%;iIY$b1N@{N@DBhj7e)BFO);>jp@4ujbXo<2}L9d7@Qp;61xp2W<%_DDK{g zaX^Nt|M_PwEld&9976dBe;U8w9Wn`DJ~?edvejZh)W$ZTYP`g5qX6~T`5#Fsl+O)q z{3KhxPhb9S9D21WPK&gp_(QRz-i;o-z{L2Tnfy-x11~J>{mw1Q{qKgbJxWkId5Gav zBpi>VUD}4)8V1B7_4S39A3hZJDYyv-PCf_o)1chNC}+O@+>*Q`F!zjhK!@YM&$Jn! zrGf+ILnzK1PD@hV4mex8Jr3B5N)i*R(;Zld?8n%^g|Xq{mw`mNK0C*XfqY@IkYN3J8*e7kY@aPf8OOzAuz(WcC_X@+q0c;-63izGhJL23As!at&*o4&3 z_-sb_YeS|D8chE-x5+kj`DXimaRH2!2?5mc0m(LcXSKS#dG2ER^@z*eWhZo@ZkhQz z&RD-mkVYloC0&Juy%@}OV9lVQ_2P@3TR41t=``~|XFHn;hyf-1+3Y6cB+Jaa57Rb(h)p;XC#x`j5v_-g`wJ5H|p5K4yAv zhpjG75;^!}SrjK}OfCILXzOh!t}E33gl0#h$gk7W)0YB3yJ!JuF6tCqg>+rd6Iy<= z9qKy}<7Vd^m%llr3xI0M|6xOb^~I{4-u7bo09Fz#wXlhior#|saA0huBz-1`3z0>z z>wUk^%fA*M_)xj8W-X;B6%4N8Aa_+drPXr{)_jmHP5km4&Rp=%Sm*Rt{-XB3ouiRHS46m`KzL~2#}&@TIo3k>SF?NrU-DaR ziYBVf_lEt^b_YcCucT{z=JJ!RfXBHP-oo?+o>k2%-JfVfO*iY^`NlrKic2q)9G=CPtng}XL)5Omi6TF>KL&9L_V>B@ zm%%pUuHz44V36GNsHk=TBEf;yq(xEQ+Ze8PjSgLhiT+`w!L2Bj+&>@mDeC38qy(4a z$3}tY;?kqlqL=XZ1gWg*-9veI)g1&U- z2m}(qWB3(Ftml_<;w1a==zfJM{CsO+7glJ9Es$C?)bDG2?;ZrxE?I(>vF#7L=nHJzRsVT0IIQ}u=1*G=_GQqB)bf_8wfYn9b?gqa~_g{WBKZP6{_+S|A_*rMq zqDnO2%h7zKB#--g=9_@$zNp<}=9c5TreTmQ)32Ya$l0;|S(0xR_dAAeLQ zEAjnl?ev0>w{z7c7y1~N39#PK+45`0)~&Y@G=2*`?<|eG>!qAgRF8u@1)a-&6rNic zWuUACMvy;00Gl!hgpB{ODJ|_t$_POD_c+8OQ^YQTduO_1vNKL&D`Nccpe&_y=FlBL zd4X8*-_>F27f)Sp4IS-HEe^1BQZ)w8sg{*_;|mc9Q&upZGfpv~fc8 zD_wvXqwgvmau39S8aSDlxmR?MG@20t;QcE${QdsxZ7>yQhTVP`u63mNQHqTA5jNnF zvpsW{vKL_`s^|Yz+qMy*S_3EM3@wW99%P5><87EbaD2J75ZVVapXgVpK>Vj-g{^t5 zz7^3RvA0@)0lyhqY%cxY=XGUYi0LRrQcbmx=S&26CRlNUJ^FRbXm{88kf&46LS_%l zj$eWPI^z$%8e0U-D;8z7bkgOSMGco>!!G9~JxJ&A!}O8G@)_yE+QA4T{t#;X&G8%I z^fMQTEqT<<~3x$2pI+Nl%`aW0NR% zK%_x?%k4@%k0)xtesy6CWi9dq@6ND?M-F z|8-cWU0eqg!If11`ig%oexi1IPDm8F$f44(~eW0A_j69Q(y5_-9TJ zi#NZ)TTVX?ngr&B^A>fv0X^id$P(AYY(x1-bmXT3?0Dsmo0buP39+%Yl|L4s>A#>3 zl?9uI)F9L0bAkF`-43@jmZ=3P5Qa&tZTEae`EaagdazJV#x*z%3jfz^ao<|3bQV=9 zObeI;D#Cv|R3n%jxSLq;GKyR`uJ@pe!y3w$dr6I-CE>?Bd4qc8oP93$!PNyr|34<{ zZ$rP&5rNoLh?BE~rsM5ZXYo7Kfzu0E*DWZgtpHW+@BjXvQk-NH%wN=a6dua|Ou^u* z6qG-uNMTb~w190}3T#Nu#c^Pp@i6_i8|y9ocOcWlqs zF|GKEv;RgpI6H^`aCQFXqJVYx()Hn9aV|(RT34-zAp1Lg{;H^*ey%GLenauNww`;? zzo_gFF@?$G79#15dxWvLj>x9M?ZI)=oO=`qY)#MQqcz)E&9_0P&abWhpUncO9#(rK zzIX0Mx_vAJqPG+vA%d6nHmdXMvUP|{vEhYg=0B&6@Z=e%v64__ST#Viem%JXP6-66 z|94lZd~Xry`;P4K*$z9Ojs|(UQ~|ApD7drs4}L3rj)nd2qA{G zrIRXd17NWtkC;gFJ^~^~tXR5gsK`s&0pZR+hkbDE&7oQ8(&EmAWK_#(0Hwa0s!1Jm z@>C7UR%M|?xZ10JN?M#tBS;?gX zwo2l^?6o;E-4v+YK?&A-z|!yG(;Ad5Qzg(c6WL=}ME+qU|L3F&7@z*dJ~{!SS5ATp z+^822gX|+d*#3vu@AnzMfslynaboipg(of#UWhUAo7AD0O9eK?GsTiNBc!S2;{VTj zmy&57ETF%s(4Al5pC%<*^u!l_-28(_e)Iu>X8P4S*6;eX{ribAYCGf0&9(nx)D`cv zNU$rvHcR5Ue58$LJDZ{u$Ts{rlmWN=W5ut(IoZ4r=5;~%dWtX9NPGh0W*K0~2Q`gr ztJ(KkqrfxA{<`V^6MS_OSQjb~OM?Elr={y#24 zE}5*_34ncn-jIvpy+*Zq-0!$3DVkxKZGMc(^B|)E@ZmM%C~ra;R`b< z&}o=5N$j}jMJ#x|yd6QI9Hlk^{&H*o;a6>JH!u40SNNAbGf&uUDFY-~^3z*MF(1oS z30JlQSSI?=1zIzx9_9z(;a>#Iksj2`q{7V{zrXq~MM9`fED5?_HT8ef{Qd<2KnWNx zRXS67uC92XIc)NN9wd0fx-QJ7u2+c#*cujT(JzP1p(6S0_a*yhv0^C~;*zEZ$G`X2=K5Kn@J!i+?oRz<$r(Km=&Sub>@YWG4*Z>0w9)!?9*# zDKAr^fEKUWQ(JK#v4<{zy8db1eg`%EP_|9kDJA(OsBC{^YpT&lmDjV3CTAYTGv30;Zeq*^o{L9}V86p|&540g6JED89iiaw>p2gF zq$e6C$`+u`fX5k`V-e?=5#uq%FSN`aG2y-82Dv&AtiJ54(1j{n{$t7u zNSMH41i59J-X}1MJpASM?k#zH_p>1S`9yPsT+F;vss8y7Q~;{h2K?`vVSfze7^=oJ zms81b4sxjW9a@t!<;nyCs4Ez=gQBZw<*%#19;W=eSl2-k2q*{Nu57)bwEbDxP}vE_ z^ivjui$r6~^2Y&~;9~`vS2@*W*Zs`zoq#ap7~o6hL!;CpaR=UUBL5jreq#t5j4J)H zkvTXcZ{ZU1v)GAE(1Y?HKtlk=n=^vN^-I$hvnylHiZHBYdnMc3K)U!bc#PPj)3%9gAaYIQ2Z>4>_;}g^%&SjQs)N^FUSxD z#$zDJDrokeDWSGRt|LaQPmEl2(mS_Jdk&tfQ+;zwyYu=h@Rs{C8=KR5B5l$ZLVF8e zu&H_)?P9ExmZbfg9q|v3dmNQ_Q4u;v+zrZK*y<74Cpmm7`nH;vlGx8DWV(KPg>#o8 z3IvMXN)*u=JfbGO}MX02L221w4f66hYPGsPNkD!*;1MOI~>ei$kVt z!2_#B3(MCuEj0GLh9#OMfkTfHF`nyg8QYvd?L086OujYz_yIM9Rb?M~sa4hDp`b|E zwMknqw^<66JjgWQyAYDxC%-n8@K=LzsD?j`EJsH`A|$b?@Mwzqg-frPzOW(Do9i1Z zk6A{M!+Pn@i`KTWX=H-XQhkDaMLO7R9$b6|iUZL!4Jk88-ZhO}yj&Ij^)A8dfV!y$ z#j+Ll-W9Yf%q*^qPT}H{>#JYWj#>|3L2n&N1R_Qb|vl z5+Plln!%3UU$ky;e0+XwI~MjrSW&~;02=h~!g!##(yD%?lUZ&-l-h@y{MO%9i(rx+ zm66A9>AMeFf$P~(MIDv3G!kK3x0C^@84bi2J^3f|ci>f3mp27TQ=`gTn6n9xZNo&o z7*Ua#0I#w7HJ9*CL$}tW>2@b>Ag;}uC=-eRE~6UWf+qdvzhGnmqQICLgdTdXD$A=a z8E?27eQ(9VbywHcGXjg5YzE^!6xOt4g~L6grn&E!llGOymxzRKg&^#9A#6-1k`fKn zfVL}q5>T@I4v0@`Q&gy$WK3}+Dyfvp(V&)pmvFZX$jOX1BGWYaV zs>l;r85S`+#U@vH@}`YAzvnPCp9~wiJxr13gjO5T#rNI!rL6w7g2TAnSk(xYJEFpC zCY-U#8noeXZZ;9ULba&pCaM#zND)SP=9;4NGY@8-lGz-<2iQK2p;bL~s-U zD92>)%(3o4?D_JhW&YXfG?(r@IEIIY%|3QT*z95^>bvy$VXkDS#P*yG+`-otaE!w{ zYksrr_qAJwI_mcy8+2;YOw~>#2i^xbyYk4JxQ#@o*AYM80~wa+|wS@6IN+YjuqL9 zOg|PAHz8*kNp1@ki&S0xdMx~9`pkw~m_wVPVK&W~XxhMCSLh$Vy=~{(hLz%^h6N!~ z=jaaPo}ahG6*ky)vq##o1%j7DZ-3xN9?@iS>cgbSp7K+K^rq*xvAHJ`uEG(F_VKF- zqgp*hjUxuI$62Zo`JsVQ_HE9|WIO{xnvFRjw(|RJ-a`R9y0&e)VbGpT)~3wF`fSdr zB_2D$iINNAnO@K-_L39q$8}z~)DI{6EVN!wPc*3rGiIi!mR$|y5Zub0z0LJiSE;6f z?GvQU@MG_QOM+R%Z$qM|xrPAwcl*w^NM&Eb|4$> z|L7Q?2%m82GhfV$(9OX$b5{s z;=XzD4aXJiVh96*!?&y{ieH-$ydd_96$I?Omufg8)-Yu zhwP2#r2kSTu$T`D2o+3_V^YYp#Y>k}QJXzTgOaanl-wpNFNbh03j9i=?m)Z$?zSz6 zaxE&4VjGGt6jmA03gb%~SF=E!JpePd<Ro0R@F7J_km&`>eTgs8LWPprUPX(u zRa!L!Y=hlWXHKVg>u%M9Q z3z38LD9TOw{zAp7>zwM?^5H^V$NEX+xqFzod>!|=fd(XWE_QM$+(TitI@ky%3dd{d)1>byi_ zrEH56U93lRr0e>ccp9xG*K{v@(kdGIAyDuNNQx#O;Pe~g zaac?L>tatEH-9F{vsWae5nUvdxKo#m+arEBG`fVQzIIs`}b!ymo_}w@}rTZqC{W+obLuGoufr0w1YKKL%{GOugrEcVSb)k(i6+ zCx3|O)Tc{Qud34qgc2PjN0wmTs6jwB$4#vIx9eOH`HF*U&!;;Dd{1zH;g?CMBT;WH zc>vWh=P;)>A*_^5(0b{6drX_s7P%`;*QR$>b6bd9)GbVia{Rr-KszJR)c)MFwY4?Q zvpG|XOwq&fs*VS~&$!T|qnflh#1OYR``GP=81T}1S5c;IFJfg~Uiht83>i;?D;3d{M7(!anC z@{L6yQ=y(=jjm@R*HyV-RK|k^-ot8^G0~sTN>t~nCe<$pndsj^w(pM|YJ1&^Gy&%x z8VPN&*z?7^eYk?VDEU>ZK91PFos&;lZe!+pVq>1*BFIf$)%TO`q;rfmdR*0TU1)Fr zcG~azBwS z0b(mCe4sV*Fg}}Hw|oeuJQiEqy`k5@^2Tbvg}^l+LRq1PK1M&=+{2;#AY|p`J0jid zi%K#l%tAeTRv0_hzTtD!SR`3eimRElP7gP{Ny+zSOED!o%llACbucaWmCY!X!QWUyQ|~q>c|Xs$^lVIP^UEI1v$ixCsT&sHk9|#-qj)p!b_dO4wuo62^P*SI#UAEX%qgqM}7cwhiu_!7R!S| z6!BNm2>UHKOPWS!s>&R-n6V^Czhym@;ldTs51Gi`Chz$0@KYu=rGos)+xA-+O?{re zpOF$s<{z*X?}Aa za`!Ev^5XhDMC9fQxUQ`ByKjlrGe(N;wzEZFy8@1JAyA=7OnJQgrs)^;kbGMD&wX`W zj)Ud>!Q8sYhCn@u*rA;jg_pK`39MM1Gz3N2ugG+`1ESHz7pz~k0@FIW@qn}o9Z;_e zo?IaJ?$3QgO}r+4CSFvdtcXO{ynf}oFkZaeCuDz5GH@0BvFG8rI3R1N2Hb7(SVWZ> z2}pk|XD)f-$Q(nl+obE4~tQVdv zd3Y}ZdzM8^Wi4ytP=pe7;2#wkim=J-oZKI#xDL-WRwZOgZfe(X+Ix&5zqZ_8EaTgL z+bWrn+9i!-6Y|v2qZk+M8FoDS+`KL0)7gcZ5RVP3y1vIpJx|jjdM?k+1xdfaNZMU6 zK}fonl42I2unaIrpy-QBrF>2d?@1$J{^*|G2Q++$F84G7OH%zMdNhJXM-W0eC zRa1EjwZ~@-8f}(y13#M798h7b*snULPo;gAE>kY&mOuwO* zuvI+vCBxC;5zDKyvf%z#HYzu$_Dwn~FGV};F*W)^dIP!>?X?P> z9PWsim4-35L;F0g%jKq;e~??Ues{wWe{O~(Fvh9QDZSbzs@&9zs>|J-Y@@of5%4w+ zh=GN6O1YsPv9<+3+5W(^z~0dfb6ALJhDKrDfGPKf9`&8+(}q)!I+8>(nf+AA-O z-EsZM`T%jD*g2eA3K!oZ-mtx`sw&S|s>#N&aMCcu2@wEAZ=7VAXY@TBS2B6mrKC(H zRaL?e-B^G5E+Auf|Axh=WxH=MoJkO6GoKYPN7L-Ri=*PCz-6Vx#U1=;t^u`0T?z-m zE+RinAsxOixW(PWe(Y_ioNYtgP-$BX(%h?$pu|tKy4hYs8i6s{flXWB-tJ1x^!Pl0 zGYV6`c!O>uDB4gHL8=}-#4fGue#aN`D}9n-hTqWc_}nM5+;!*bJC@RRdnCwu%3n9O zvmT4tkhm0b;MP-4CY^Mm9rGzt*$R_SK=-P2_`LM=@UA@FQNE=R7ia+mMyzLWI%@GI z-Y6&)-}Np!ntTbAh95Lk)Yhj04R>9Qn~*8B$!sGI7$-W9Wy}Z?n-2T19@1e~iJO>IgC%yTA zMtF%*i9E0dRlP}sX(n<%4Whv{5ewnMDWI@0{u92Z&V5Hnx9HsEt|OUEnXN3JKB_LQ zT^4pO@l&KsuxeLDNuO!z5e;9>T+C#q!ZBi56AJ-#{cV&(O6Kx~qULV8T9%+HhT^;o zD)nTGcLXc3O?a8*+apBO`vEF%$4Lh5_z1Kr_7N)p0{_ll`B_w-Rw5wjQLDCM^?S1q zRC%q6*TP+CD`OZdWc8|v!KK+p6uppabxmn+mi!BJjOHZmhIva4u85f1jhUg_cufel zwmH~gd4F3OuRF~Bpqz|%7xEr)OL!&$%%S96iJ1kSR>U{?5PyjFTz&Qg^;aAqOZ4#Dk1e4(`rl0NghU8F*Gqgc`x}8BerV%86)g4 zFP2{ItoPY2TifM)U(csB7V_q+D=~b+pJ!u$f(fvrt+=x6*4&}{<)9S_!tx*NxS>o( z?WjVQ)$_f6iGCv_9Ew&XL~!HcZ&&`1{LBLkOnyz$+r5u*hjAewRW?4!Htd)^OXm{i z$S*J6imF$(IIea%j5`W%=Xs&^2rFteSjwA67SVo@(~ILiHov%6V=HdzIS9^FvXAvt ztDIC2Tt4(GPG?1MwNaTUJx4?S_4SZ3ewSWtxr3>xdz88$WY4SG2=EIv4rE^xu+50bag*suihChrWhOY9YY`=)=W@N)sz+JK07TX@52jB9G(?_3wtWO!)7~`80 z-gfehgj-q$0O0uLyG3&I2vs=%VPOEk+5=b%)!M@kyc`mXL*Lm+er1%vv0mDM;$t=W zIn{0Cl!nqyn_QpvLP1%{#14|zf$C*bjn1|Dx;C~FJ^}xx`AWyeuATjmK@K~O3v@3D z_83*MA3{>Lls`>R4`Wg*l`bL5gzOKiv6k&$GBI_DJN&qR;z5ro!3S1nI=82~y{~9> z!gbUass+Ag&8WTmnGsb_zF-f=?JH*zg{1C{z3?;GpM$VJUa&vZDpkEdW`|x7pe;6! zlVcy9-WV>yWQ$O;S@^&Y`SRJhGwc%AW!=>l4G(o+FLe&(?nMnf^I-SY_em=6w|7yf zovWogpHGS;Kj%2*OtG$nb26oQX)D32{6^MPPdx%UYFMk|L-CB zRVovu!ce!F`ujihp4zyKJKf=4g`%@So_xz!gS24z=v*UEJVmikd1%Ujuthj6;}rsz7)$Mkb<)R~N=OBVd5I>m zl+GQp@@iNtOQ{#i>TcZ9e=2fGfq><%?&=^-GnNqwYYN>52`oiEo!rR_%U>+xb0y3JEl7z1=-V*1PjHBM9=kMkC)cgt0Uod&jz3R;xM~ zO)-&>8Il`ukIH8X?7Xyz-`lumyQ%Z%b5~PX;%M%DNzt9Pu{oV*Q`2)qt;$CyaYRjh_|`1TNzf>ry+PB|i! z!V1=;G|%Ogc1+-XaFR{)>|;5YLy3C}cRYH8*`sc$En0R2Nfe+nb`$M;4`~Lo}J;+n^4oLIg*zIJG4!hzVjR)EioXftf}{@SAwa0-G!3 zN6&OX`3`L(gXSbY8hj~=8f$47EG#OKRrQr=^z5dt&Vx?N(w*HX;|FatPNo{+f)O^$ z@f9tkj9{@7O)UF+SO&~*@NPa&cevU4%o;>fEzWM8Pbf9BUQOc^?z)fniteT4+yEK$ z(O#E{)`#HkH~ezNLjRP_y(*Y)PBAqil?zwZ7v#=uPPqnd+beICi~QO%Qs5q4#`T_E5`X3-LwN zy55CHZDxe?^?As!AUd6b(KJo+toOtM+GicTp?LDT zVIwn}R+A8Ivdzi?tx*mYWkN>zBv1;`r?x-kM;%CzDT-@tDX;6FT*OBBV=8f=f50^G zeV&>Kg=ruSGm}F10E50xTwqH^I5WgqaAabua(>7joxfJBTG;8>--S9JS973Kf)8kP z79P&l1#|drRN3L`y$Av>0{b}hYrEH28hMZTM#PcJ{~+Xlf&terI*qv;i?_(DhJ~|# zxubF{1R|&LP%G2AciIvLh2Z5&m-DwGYQ3*uYFGrFilT-e86F41U;fMer{a!3c3HVt z1Bx&zNeGt_94@qof+8=^r0IkE@d}M1?}3oZ&l3l8Y6jgzNH)t1dyL>C0M+|FA2q`u zw1`I5FfBCrqE7nBUd~A0i6qZ!ivgB~j-f>`iRSIyx?o{@vG;=z2;I*xWMpLmIs#5| zY#7wFUU|babj}4+(q3M-4se!IBqSU~ctLTK)en1`nL21Ki*L^T$N4#c2xw;zu_e64 zER`SdzFtfcNhX_0MRk8`pQMXa+zZ8}(=iB->ial%56h{!FJ+~ZpWq{W+@oTBX*?3& zTC`kJ02v~)DB=Xv4}dqvCAL$&Db3aYhm!^Dr)o>J&!1AfngrvszBgq)|o7c0-Fj{VR zXa{zS{j)=7|3(BM-oHS|%=LV-X&H1ka-3n0v>ggR4>O>0&m>KHbAz_A*gdW5oLZ-~ zMTeRQ=o_j*{IbCg+&8ma4KDh7`#|j?US#3P4~CNfj#acxY!CI->*3E&k}E~L5mh@- zkB7I>YipSjHl8-d-vb(NN;{i9JCMuLQ|?W5W49tb__|{2MtnLKEVqh1;Yz|AgP^nc zKUF)-h^93pnJ|XVkt2CNJoGq54GeI(AV!tTyM?hVzPo|==;B~(K~8i<@Yq=yn|m0p zp#(o)BkiWL3U3&!YyI0WQQLc{Ki7mzPdtCBN6wo#XO;(=9>%W`#kOLgdRAoW+!HHo z0P~Y?w!dI;70dl%_8@vzXTb|eUtQMIKAtclW%cGsL@crD-vbG#i@7&gnAufcH*gp_ zN`BC@#E3H9f3fb{jM6;Ln0L(H5y{zo)+{)6PY%~Omh@RpzB8SdBUmomy2@*=F7nXz zR_CKIO!a?QZ`M%%b=5plj4PgZCFLXoN%Xiivcy^S6v1)wbpOW1lC=Ke52QyAh|2Ca zJfr!|f}2s1@2;3!?0D$orDdBC?f6m0lZaL2CAse*>VN)~PEdzL$Ev|R9D`^$U&$#p z&uV#IfXk!X(7@%iVX=bE5<}JT{dhepKpJB@nDA-qSA+B`n_MN1 z7K*NV)!ZTjbBTbZQB9J-9qaoCbuJVhzy!1Js1FdX&t5G3EChYKxo^cr;8ct|sS`p& zg~~=ZZ9y|_gUTyJ2zuWDlM57(*IEU>s_*M#XjWculGN*HGjHimMu8h49TfFCD zl+Q3L0pBSY&#YLf+NZxibF4c}l<{3yb(bx&GaI%3b6B6m<1fA0N_rBei=A|X@JdIQRM3;Km^6C)ys|C5FYV*G@` z(5*UAXJflb%KZ%z#MEktY!Mjus| zWz&bks=*0Wcb64y(d&l>jB-bU|=-oZZGZ^Ur@Wfok;tx`XX z1H##R148n67wX+^u>2+#1SdsJ=7wfYhfPN_hO`n}3lu`dk$GJ+OnXDJGq8 zI+l_NA!7C}F+b1MEiYQQotNfibnKNR3~lss!{m1CBw8`|C}`wh^5kw}*h3r;Mh{YP zEimSMDr~=B$N6NKEA@^C>M5FjMcNINCC~|LJ$vI=Z8GQeGQrKV9JS8xs~b<4>ejy< zb`MHzC_m~H!PDgKK>W1LDS5|PKa3$Ais!^8vv{Vxi2ON{I#!M7m34tX1q)EQhg16= z)zO^@37b!;5f$#)To%qoH!+5yAyXqmT{dl}j=B|yHu5QA+%dPrNEFJ)Wl3+h^78iw zBY5}^up@*wh2d}MM#M5JA@-yYw@ot^t@MsQk@Hvz^7YYPWf|@q_g<=KLZuY1P*O&v zBiEJ+=j2D;l}dL$EOzZ!9Hw16?(4|!oUQJPM4LpBTW|HF`>SB+l)G~3RptQ@WtmWpzv+Y7$32Tf63tgn;#*90QIh1{Cb>Wr1oy%Xd*J`+I zZ}wgHMg&d2JxvHr+E}A!gUrEL>mR7Uws?vlDVap>t%>-u5s`F@p|2%VspjMz?Gd+o zO=5PjE{MMd9K<`?_6FpkvL03Ar|cFt^le4GbGs42vxpj6+<Yp4zx*Ri%p?j`%(*1M4%8R_iSG0`lv znLG_LN;{K6w#`_0MDSc;p+$Z8wKh=3AUCm2uB?QQ7g|;mo_LgWsBY!#^cnV_>*%^e z)7e_3wcC7QRyr(8o2}UTG5Xs$m-lpi=jeyE`=Hr2+JO{pf0|rJtT|}QmnLLN*!@@F z5YI}4&5+ZEP}cKO5wXGAx>z&G5q1174MfhHZz4+)=j=M7MU&zNjVkmzHNMIQX;T=q z8*BrAT&S>3h-+8zHh*2UAQZ2R4l|+yp9zM&UO22&P|S69uy+2n{akay&BKQdUAwFu zOjp5ZOiLE@x$M$<>e`ZhY}MT08N!^LRR|@>Pw;@*^O5C(L`kHTc>EKIu)t-**7IE87 z;<>evEY;_zr^|I$ohHXU;#}p^JSX?XbNY4I?DUrM$BY-n%z4$mS>3?Z%9&bR&FZ}D zdpPWXx?~!ec*!^JXd#O!;froy$6fEp*f)mhX;P_|6*TxS22714blL8>bF@1~-~~ne z>)e6Axr^fYkcYXAv2X2ms#rEbcXtLpnb)HC^(}6-W}&z(o8<@2 zDsyBJF1*M(5yX+%aw_aS16{Y0_l0K4&5K%@^#^^29LAkb6A%}7B3{x;YCZ7oLcQE& zr!O%`i+HU$f1Oc)L%ZSZ`3RY%id~}2J8!&tq&|#_N{JnR_>cxskJM}8*KR12KRf^iJ>(*0(K7-Mm# zu(B3``|4eJPi0LWia_T(`RguDwFddj_7_bF;wCtaHJB^8#}gOOX4dc|DKye1t_V{R zr9ddtGnw@e0pC&Yti_(mMZBt6IU$QkWq@UmWIUTrJ1n7>wyPhP5*v1wa!`LCllSW^;ja9whlQ$ag1RvpBd;&h5PoV9%!Vwr)Bl0p4otL$92SR#|%J{GyCWH&ZJIWJ))&}Z`05>E_ZfOs`}Vmx z3pc9eS?4MIF%OU*NdhuL{XQYS=}limEv3(A4Sur&L&+IQ%aCK@bpiXb_$!L}+MI?Y zTDNw$Q(HtnK9}288pE;7WL16%$M_B95^S3xKSG0OAL9~L(0Ib6|7=PIYs8*_I&^*a z6-OnHAfAq^?bp$#rhah&iURx%L2cM5^&9R%5yq7H#U_}mnGos*t;MTKY5P|)Rk|kj z%rqYG)Zq{{Dpx;n(sHT3`G&*mCSd-2bvBiVd zn^5mOz%!^s41f0wDic`;FC~sCdA@sxk$W*-L^z@)I224P`cAH|4|wIhYP8O%0L?k>9fo#WsBhtW%91%bw)}j; zS$9sYb}4<*DX|AaLW72{vEiP=2pt-TS_*De;+&?Kdes9mb`PdDHHs_WHoe7;PD8eext;QtsYdOsv6zoD{+KBJ* z{RDM)7p*6uLw62-DnQ$4@QOG+JG=BKgk(kxj8SHF@mSunjLCwblt0`TYRQ% zbwqO$knATXkeH6|13ixPdw*E#0O1D6T@MmJKn^(GChB>YFg*Ms6+)QiLO{; zA8sUC9rBNc6M;6+-a*W`q%AsmHV5M?dZERmfBuyb<+?p_x%Z0Crqx-`*-C4km5E^q z#>;!Jr7JILp%aQLyyu_EF-6^zcfR}Be&^Y?2MF6NnK!AFLvT+TZF=PgX%G_Z#5ZiF zry^oWsf=JKJgTM*?vqRXk+pO_UrSiv@kee&L#X2Tjc!Y}dA(S| zm!!aj830tn8|0&Tln9S|jy6b2eyyMO>hE!zhj}fuOmjnrV0nY4 z#O-V>9jxIPf82jmX!y4ExxGheh*n+eVnMokU^7dE$t%0pJd`NDZmIV!t=VI{`xc_t z+6trmLY{KaxQQ*Xazd^!eCnF0LRe+b^%9z=9feo2@iz`3Wamhc0*rrHI&Pn{XAvOuwTox-Qh}N;}Q?!?nFu90#-O zlPe3wMfekDc8#V+IH`~0wk0)LB&=e5yDlkw zvDUvItHzNDZ14Kyav3v`(eE0{_h-xNNOPmxcw)PW$U7SqyH|z$)wbYG> zz`c;rJpcpCH1N~F5)J9yH+wA3e68UUZp^QFm~C>>E}&w3*nRFwex;_1z7NAUo+XU^ zqBE@o|>*ZvF(F`p*>W;z%=6)F@=n6>&q_ z@U35E^V6XU_>5+!uHO(Y^o&wbngHYqoRBNfpbi{^G-^+Y&K~Z%T)v}QP{zXgCWv09 zftH?;M)}P~_TnZ1e75JBDkr>@lI}GxZ_ckweKWVWK3-leaPcrJiU(Zq$mX$o#-(aG zYvhe%GK-33q2oR)5~4cGOP6|f8yVmjrHTZPs=e&f`#hd?mOX7;18``;_KwapDb(2K zkOXnB>6Hzr)|0#ckBb_Vvk+F|3g7dvk zq+;zV8OH+(5i`=XgpfF&aC_WQ+Z2dTve?V=$rtyh zoOC}9qqBIo?K9b-ud1=e-Fg#~1=*y^4zO#dAq*(`nOq)PM75%a=^|rkZsa~8bbH!t z$c@#_eBRmEGaYfY7=2~5sZ#cBarZ&j9u!Zh;440p+OZ(3N1-InGiej~M%ne#+};lx zjn4HFkDIAP@bdP_G>dYQOU6uje!4-vXq%x3@^k_176^Jy|M51oqHS6G`ZMS2SJVX% z^eE1Vdg&Ifr8n2GcLLg{TC_xVyp3<8a9NEjAo+Kiy!91bTGkt($lGwd#;6&xj|Rtj zeA0dpu$xxDe0VA=!pd9|s0QXEgR(o34f$_9`)>y8h=#h_KX09R9DkHiyJ5%H|GgJC zaxnI@oa9U8WUI1BKDpwgepE1*5U0Ai*HBHe&5fPnCpxc$4RHBN;1;$A*+-cYvNICP z>@&HH((~vi|Lc2`!(w?5PJO| zbq*-YEt^awPtg(`G6BU<1oaByKba#MoFQHb_sD^mian!kPb4Yb>`_;j!xH! z)FT`PX07`C_Pj%KoHuJktzyS<@=H$brWRYD(;A~qTipYa-)(T&>5sW1=%!LjGn=Fk zFMNKK1}*nejj+op%K{yqWicYTN%q;EixOVyx7&*jX6yqJk==x%HPO_9dG3rJ-YY=j z{&VoXoUkzf4ZOPlBV(b}4(gc-9>lA0*AU)EFixig59(FK#`* zFKHRPF87aRnCvGCTrV+0EU2%M?@iDaWvZs}YKPinm4E$k&ROnyaReL&aKFpN9VyoK$W;$N|*y8?xpw9ZNv<|O+Ss~N$ow!u|FaXH)J|cKZkTtDz|JIWA#K&l-D3$ zwA`ZY$Zm=e-O9dJp169iq)cM%Vs^U8fIA|04V(UEMEHvQvPuLbG8ak|YV&8;pr%CF7N(Mb=6XFqv`?hW#_6)+y0l;rqJ>HD&f zdxvkuN=(_#|UGYwRm0F>8e|U9b#m_;;h;joLw$fM&917GXlTPS{|k zOZrB5^X_1BYf0Vys)*G`Xc_GT8vYAstMPWXJ!{|fyyn78DlP`iCfi`%cn_~+n-t8C zeeJ~FxFS(p+8HM37R7B-U0l2{+u5-;Y&5fDZq|2@cF3wtB4b5-w9%$?d4)oJxoAhu00W!GoC>EZ|HI*F^|%qn0#?;sW$L3xa9>0c z4>uQ7kqm7QX!*0Yk(NETgf%dBEY5zlU-Q6CAGZ@NTDI7-iq3qAX*b zv1_%jWzGIQQ@L~B@B8!n{ps;Auh;duu5+F1ob!C1=cUH}F(kR`dwyI^?jE;2TJ1ab z5(nNtpg<_sk<_LQFV5D6MzIP^oCurn#k{#nxXIG^0*{)77_W}*@UX3VS$@j1u3OhD z!nE2}ox1{LZPr^(E*jK-PC9{FBnuatcHXUegi#+GjJ+{Qfua;x%a>`3@$ChNxD>X`8!z|0si521_N80nsk8 z^wjI@y@Kf<_i&>|tujvK;H|h`cQn&ZeV=onGs}@z*P)SD-&^|$D{&2v%4C1yQ@i|h z&Y>0~R=fANmHul8s*hRM#ZzB;FBCT`|>hVO5*@W@od;_9RanVxJYUbkm_WYb2 z&*8O>l~WfFgD=Q8@^e#NMTwz`#@gYAvWvt{?5#`8h{WI*@sgv8g|`hggknVAD}Xzw zxFs-rbUU(#iLNEem>C@gtyay#Y5XCangn-0j3p~kO*E{-Hea#*N`MPuaThwXq<3jj zagko522*ST=Jn?7bo$=+rxc26FYidK&#yq#9Kp?;ll)qF-k!%HG~YV zrj|b+k42QU)h-!m(=OksdN!#A%Iy^ssfYXb?%xBqHzDYXGSY9=lV-X<(Y06>F1mhP zvRYW|*x2Wy?;iP~mt_1f?}$1{ycce}G2Cu#dpMsL!oblh7Ov&@_h~ka4mwav?AUm? z$05XbG4$C!d3AiPxR#@?F+o=<0&SYIY9^8=g`68J#HV>L$ezXSMCvHMc}3TYd}i`cz-{jjPx|MN@7HRtKSVK@A=X5- zL4kH)c=boqnDeCY1GxKYtVmmiXiUNH#6J@9C`nF5jRa$1WG5F83GMAUu`#C8)p zxvB>GPtVI@vorQIiyGMlQgN!(W=y)ego%DSE4qivSY0o|uM~rxD_KFBU)I`*6mdqK z^pw0Z=iGe+k7BY{T1&~qmI!D&LJ1mNjJKf0!xUH@cRPec$R{)lXFwZL#KU_%$!|Ks zxn@;?DLoq(`)p{(9vEK6-K$#2S(_LU%TbvFGpUaue{{dJj@In#tYW~UGFVs%p_$2P zRy8ZFv{*}L%HKJGebNwij?~EJoAYAf76zr+G9Ng(jK(Kw7g~H6bqfD%&V!+bvpKVQ zn+ahq-lsBWLpi9)Bzqs1xiMe>fh$2e3}L1=I>i{$97{$vE_ujR_-2e$-x^Z)6ec?F zT8UNHVXWnODzt+WRJK!K^s}7oFBeKg^Q8Bv7TO~M5xggpgYLh9u%_I*O=QPMI;ac)`YqxRA@= zmD(MaOX3v5wi89iw!w_Ut=HQ2h`dK~Wm#T=r0W55DRJev?(#V8u~3)x6N7t2zFm_h zuW4rE3v6tp9n`=cAW<V;_wmY~8 zetjsr4bAcljsbh}C_!JS2&%k3LskYgPbx8qYu}jHT$l2i+05TGn|YiH(N_4L1k1)i z3wUKDPC^K3sjpXKP7sl*Kg5kNttBT1(W%1^D7f9FeNa}c?<8dA`l>TXI9 zHR4lFvd}puI6^;CVuCgZNv{TkG4m6wQ#~op5hXPvs&e&o;b$}>qM9W|UXd6X!KEhP z;ck2o=2e;>j?16(TCMFE37KR62gZEC4~&rqhe_5hj^|6xR6K7Lg$dO23Q!~n<56Qx zF1qaKgvI?I#owFiCLKp7L^~c?3hugT*z}EomMH4{7jfVUZ5Sp+bi2%wnsb$CUcTd= zF5ODi3#Awo#gg!`GR-5gx5>+-TlMq1OY*q9i_@cJWWxC0o-roeJ#o2um+o1Dld{Xw zllWmj+9xozdv71}o8C(lW!a99cmmsjeiRowA`#LQ>5t@y&3e;*TmJ&4Od+Ew3 z5TW;>Op-=+{}5}pl??EnvRqg-4&S9J8i84SHtW(&9ez8bQZxbF5Esb-Y<3W$0*C0= z7Dk8~NH{U~`gaFr2ViEo-x69I$HS-9iL)9bwZ_w@H+~EYNlr#RL(RUMu;nD!%$qG$Re8R2kf_qWm56@8`N22g#B9X1sV4_2)sj6G z;LK*I&ecm|K}()l?Fo>nzY=@fu&EmV=FJCrb&lg&5h*<4=r919-*#bQ47{JB{Jc#) zoG#~Fa3%eC)<-cnZ@ubnkMy9`jD^S#`JSYPl=Y5ZPmtE$@a(2%jb~Z@tjXR!gbtsW zZeZ_Sd_&NheaG!I!kx3xAR+(Qcql%8uI`=B0wd=jIQu1Ohukv&q0bqwU2o+*^%r^> zQ!SIg9F)ylEewN^f$;p)ufd!!Xr^8)k5=|A>mH*4%2}@*|AyN4VUC;xhCpAn6a@lu z*hYcCE^vDPoQ&Jr*V8;V%T~KL>brE0e~HMcyooH@g{V3exRKNM%vb*g zY~P1e#)~BwBZo-E>y<=boGm3XuN1|Rr_4#T#s}jud(1?NDG)}jP~8A6dx?`^t&hN3 zPsEzVax1OPtrnE2yC8aHiYjiiLf4Z7NJXb+qx_kRsZC3jqTFh{1iv7PRW(0Hdt z<$^(v^tG{60rXfCrku4iX%UQF1fX9{+11{_!MAItzgFaUYdKk9^O#1NZ?8JK+&%3g zp17(deXInt`h9#u?;CIPDfD`*Mgo`I>Jp#n_v_!a*0OkDpUk9}y)@8`rsgKX5i&*7xB*1gy|g@ce>J%)9(5HR8WBZwG(2T}!&m?~&*)5WutT#IKdO8#l(z5lK(XtZn-&uT|!F4dkNv%@6xn4#yA$JNr zQ#?D@TweQ%n?1cbKMqHZxKKsSj`Ll6ybvBcMC;j+X9;fIMYr8Q#N2}9suGW#{t9uJ z6S%5p?Vsk+EZkMu=DhtR!oO6^xFk3;zCxh~22 zn6Jmrb7uG!!=n<6rflKUQ|z)lhL2X5&+R<6+Y8YyM0?rWa&J@`qwMS((qvucjrtPI zXxw>1U8xq-YjL{$H*KW;Lc2euIfIE7@gz|o)Q^CDAM#5Xx)1G=rRju*?ncD`gsmrP zho5#{`;Md#Eh~5FI(qt6@p@>Mw`OVoZZ@HXF%OW#;Uemc2~}NsKH6`whvo&@YrJoq z6s738W~PoJvd3z-H&Dx(O|S3x$X2DC2D8GUo%eaNs8@j1SRF~nzmtmh*^tU`6b0*- zQxxV<%#{68%nZ{tBd9IUJ9&Mcu?T0AS)Q1=^+1px%{Av{g+S_`G zBl%$sDwBLI5v{94%deKF%yuFpPTPgxLXfML@tt!T8#k2J4`#Kb<;rb@(=GYIt*P$2kc5?y`JBB&g<3f~ zm&f8_vmyeh2ETtddmA~XsrJ=!uX!NPVAy5n@9He~H#+l?ezB5yM`W_yA3O8{BAwoT z;g%j1qNwr!eke$cTz?p&A_t38kHC!=DNzA{rS(^NWM+N*1(8ZEJ zF}%xiDOm21+4p(&i5jQoRZb29cg?Y{V+aMae#=5*bz$-Ns?$A3SFUFwgcy^r4`z_M zb6%f?aPs=4E=-;ygNBwjU1hAdFl6HX-z>!PKOGtw+8B=UcYH| z4Cu=JUI+nnSZs3o@F7Ii3#N8X$Hl6~_V3|7?^hniFx*`qKb1W-_g8!($tS~G;GK8U zIS8O;Xl&&YIZX&hBq+UHT6;%R9iT^;<-OD{T?K;9mqOj!{f?Vl0PN|K!NV30senZ+ z^Jho@O^t#{mU@}$jEHEUydV4{VeUL(?p-+HXactxBl~TOz;$tnKD7yz883v}cGrMj zeyH>@POP$Hf6)E+ex&>>KW8CVcQu%Fp}wI&l9J;A}!k?Ki@3FZ$apC zptu^4;$@=g5L2DQjpi!noPWPM8(iTOp{f;DcT@%?gQk2;`kKl%V!q+uEBOy3u44ZzC;>7`HV~0i2+PE9vNS{7*O`*t6pG6RR@cjmBPf2_MmC1KTWWlzpt)sUGclSB_Ij>S9IstUfZ;>Ym}; z;9q?|fly#?-&sepo+ml&{^s|rftzn`vBu0>bYN*PSlI?XF=!TF6}u;#^F`oYoMoy! zSB^XbF;b*hOn+8+%!e6kS@7TRTE>U!uK8&WBgv-*MOaH@{L8;?BApwu#5P{1rN+6K zR2&`QLzHrZz?G1h*L;dZLM;D!UCihc_lnVFMaQHZ{nGdsJr-a~H`ecs{3Yg<*`rc# z1Icro21kh=7121R9et*-zqd#M(ptm$&BFCR5w?Z%greS6-qYy9AyRpxWL;|M z!xaTyEh>aL_Kb7@uU31IV?t&v8|&fgwte;7a_fU;yKI;Ry7v^^>A8U- zXuj&kQO3&&jiS@iHGneb@N#zB$_B{xZ}dzq$?zqi9QE&^-+Vp%LSCH}^~XCyKJ*jU zXMRn3{!)BQLUcnJyTq0KNzqWV`!##B*xbu_hKP@hb9z^MoA`)jO(b^Buj|~V9k|IV z_xZ$mUyNp>{_uqFWZ8or2GT7(P911}F*s25QKvyqL)*m&0QV@C>PgeCP+K?IU5dVPb^b=;O4ZIx*JhkdWxp@gWN8 z39n2}Wsg=Yl?tfGI7Y8vvv4J!8eQ_y)n$DKsy3h+`IBo!MRZChQKCEOPsyvN@M=Z8 z;S)z+qm)Ea!AS(nJ6wSD_WlGT3t^;45x)4z(xU-El+c7Wd-~Ae!$M_q`@}O=)-jqk z$}Y<4z{G2=B_333U>irCb45g$k1}`_k%m}kOhEfui)!)PIw1M6{GaFU%4vT>8w$W7 z`)>eu8vwXEPt|ykA}@1}iB<4N-1BN$J@I7et5gFuBJq&a%ojR2>E&Z98!1wJa;p!W z>rcE`o#ilG8wo6PM^-dPbcWVt=sTDZo&nV?SHP5CX5Q&diKN_aJH*;ZF?aZ-ZHRcD zE1h>R&TlES+Obf&`1YION5;#9)y&jAT%vfOl4!>4fx6D~^`>vZ!aO3$7?p_i zw*1&N<)Fj9>pt$OFZrv4xtEijw+n_ysEzwoIs(g{i#xtMY-mOH>hO`zzxXLU;h2nx zvDVvHZRu%Z_Tke=`{3Axf{q$gFgMB8X3%1WXwGtz50AGsk?tsDs)=>-=d#_<-g{t} zW<+j=%OLZ~%!R2VBLN}@5N$Fm9>S=_rRrDdqLPz=Yv!sTSh(M{FlA}IslypgNRvV4 zNxFs<)bU}}oBSRVdbGs+(_7HVO&T`1W=rL5-*Miz?=@GIt_7IWbywvwUCc=kefWc$ zOJyb22@&tgy;hTycKdPybLQBEj%x*hqhdNeq`p}Cb1fD!P5L)4?nSn%`l^}w&R}Ef zVMaRu7RM&P+N8(6p8f33qC{jLL0|Mg|j?_VF2^ zAyrK5;```ygV+(3AlX9cT8<_kes}&h(&~ZfoJ(wcHW+1bMg7+C&{}Jr8%`7+rH9H?bK@52jc910 zGO>`etjGDuCF~o0mp68|>J-h*hWi8P^(*e?^|9TgvX(qZoIPL8)G%$#a~NTAd`gl( z*bR*5l;cWHeq0o*_=@{0`|B%`wWY7IhHo1-R#25FUGYe&>F9%h0??bTJHQpuP$6(3 zN{lpXTDd-e3zzGV5v;N_84`|Gm37VFKyUZ$)1%4!k?T}^NEec-EHYuBtx8p=U@R!F2s_sEz~cePt` z%EYJosa2>*$gO|Zw6W1Pz>F;Eo>>gr`JK48Hp z?M~vNQQX=MLDH{a;b_S_o-V2}47HHY6VYO zd`-uMY^;K#dkPwq1|cR;Pabl3w|mSXcZ$85G=Xov;A6;w=cR+w)N$k7LJ9R$| z?rSW#2m9qcIN%@PfW2H#eZ}D(h8@XxVQ=yj=&QM(O$aBJp79TMUs{YARW(mF0f~fu zs~nKaF8r~_w(&>jUOXzNp5=4%#gri)!2b`JuDHD-_2K~`>%=<-PLv@t;yA!f)LqeO5E!nDtt*v|5*0-@pVChKbpp@EX4CFAMy z=8wlb%o#*-!f{b4gh-*?nD3I>5D_aK{ykp+y&v*z(o^6e zQ|oQ?fLafgU~^qNMA1O_4`O+?su^=JrZTe6?;TGlrs*lQM&gXyImlalss1w*<(=5! z1oe4g-Z_%79knrkZ{}PS&{Hqg$=r^EOSp|cuIKo}gFW^x-K$(b=uL(KoYC8FOXwUCju^+Q@Bc7d(17R$M- z3}&-}Y{u# z_jKvm=#5lj?29p`iQf&p8c+w(l~K`Fr-pBc6(2CASeWmQI_eC43H-7g65e`7rpgx zPAc35MK~7vBvd!eNZo4ww~(}98}sG5SERX=uL;ts21H{9j?OgGA5rkp^P3+H6v*j6 z$2)^kQp65;f+f=Nmf`9{j|3HJeA?QrqdK}t{AgJQv2!{NHD76Sq2G%1ADLBh4%g3F zb4Iiu84m25#$AR>%kqA;r9H-$NO5S--k3T{Aqc(+Sa6#an}2lr9hWDx;fG#BSjUR< zezyou45VyJ-VPCh)urU2j4$@o22famBFHZ;=^w8G(4YSC?L_`@peOB?u=<=QP)RMv zW2`PGIaQYusi^KIBwFxw(fJDGBZhz8y_t`Y_Y_S0Meu0A-$?i7ejk9VJLa2@m6w1P zhI<*^PfUc^EFL1ZoY4RDM=)`Gh@XQ;NeU%uCVX4<+z&J7ax?;&bQqvMKvFPEDYS(U zNG@*x!Cy?h|*h+nQ?*p4nl1K64mhr1Sp&YL9b-+RSi zd7FE#w_&MLuQ$8aYtq0Ic|$E2ds_1iPr;S{`SE7lO>S}ogKFK`oI-IeBRQ*Q6}IL*PdZlDGmdA0|^5I>~v$eu)2Ii&aLj@s4ZWYM_a1Ptcw{ zW_9=Au7QImgD@I2;Y)p|sZGE-a7X|12>3bW*W?M$t^Hz~E)u0*TH)H2!CdJW7DrCk zKT8ww^LR5$yYu#)55WRZIZw83=zfe3=qt%X9L^v6#ioD23kfWwG`zemvvJtC28)7ye@cy)SFEWqscXrWm#p zT$SoGlmi&;X!dJuY9pw-{Jrp7idn6{43DoEGqrDy$&=m`e2nRhzNa)i{+Q81FhzMn z&$Vq3kO5&J1PBBFnsx?cd#w5pv)R3vgZMNi=EWfecxf)L_H67&{M*}i1ds0qY#m0c zyCqZly<$|*p%;CCsp)on@jg7)JIc1Z+XuB~rCOtIkNJPxx&IC(86@|qoZbO4eeCIGHI-%?3{OB%oaVgT|4fEY zZI3?>yY1IU*GVK60`&ar?BgN!#nX$O)p@XVowhXOs)D}4AKwZq@{X4r~z_XvLk#P^Xgmpth?|y`-pH3Yh)zzDaKh)DUPhYSovZu)U0tB3*q-Mcc zy&o4oqnEYp!`V`3;k25Oz3JXmbP%@s$v>gmW-$I5x{+aQA#TAf4=p-+&<9u+vLwF|+-UE|ea z`Efdxa?3vbZx0Xa9?$7fJKQYsXXczxc4ocbB~a>7H2Ab^hRuoFQuR1jpm)euBYQ)e281io+#XHMQdMxy?+oU}F` zeso%twqvK;?&Zv8nvvm)Z}*&GR;Q{@8VV@ijWM1(gfVnXN(Vv&8BFY0}$8l&%<$A4cP2*a6uUmzyMr z43{^*(vBE%i_(2|CWTe(R+e;W!+vKCDVtI|iAC411&7feoX-@OUGWYN48p|kGMli^ z_g-vOHBwowY-)PkqX`iYk12xa6quA_)5UT`+^-U(i;G?9z!>_O`_}p?nGN7om0MnA z4c!tWi1jl2SVeTA%ZE8(_Os8YstqkpKka+p;&L0?UGmKh&|m2nG2dTR1Xt$dRU|dR zeozX7a8pJ%b0bjjyNZP7!d$ryjzZuS>F|ZhiV8<&qBKuBX>KDo$+Lm^f}IxIZg#)@ zKtT%?*Al^0VqEKdJSVH`Q=&rfY|drVljWA~ow){lU(XtrCTR?;_ZLf**M&(k$;yRY zI(NFH93yVKV`yYUQewIEIZ-VM)`y##m?g}AvisqO)&b8LN&F9}7YE-BZ5v9}a@42Y zjp?53C|%?{0?CmPZDhb=LGbV)RzoHD{Rt7=B?1`t>SXN~j> zObrZp5I;_i8s4Ts&$kKQ$46L%x7{mD-%uR_k_VPJ;1vh>@ zT8LXq&fHBwo#$wII`+L}o@$=MsqFpT(fpOP;1;&jaZ3!>oKMsx!NqjAqp~fey`&Qi1W2+jsDh_is|V4|Tt6esUrRQqB6r_ z}hNEzcNU@_J;h^&Blk84xE-Z~J0mx9_#Kpd$+GrE`e?Qs0Zx+q4G*0x2#g zMTpDwZMJGJMz8|pMJ+RwY~Sq1mN*(F52Op{gfknRC%PB!;8{aZn z(8-T{dCj25OUM6ag>_lT;)(Dy<}PRPM8}!EyLsXZD45GUoP^5d30Y;C9Iq^#!lJ($ z4LiRkiAFI}p!FRQnkvM&ct=0mJDg+*prTG!0Y8ib`d+u2ZIQgbA;kF*sc)8IHa>s3 zckx}=LC@Nd{2q=+2dUBXR*WI__4Q;?r^&&5W4dsXeB(=}l{!3KiBnF!Y^(%d`9V7 z_9JX!+z)<8PAfb$h&?K4tn}WMU_V~20p>jb=4}Y&75~F`y-RKOE#$!h%VmbXm*v!& z1sR6}h!N;1P5y(3F*od(d%1DrASXN1cvFN;nv~)6OUr55kkadEFSPW$Je9J#a%oEf zRMP?eG4;27hH_bxxJ|&v#Wzakz*(B{1KyVml|G;L^!}BMw$H|W`|v3316ikc%_$Ld zHE|2y6!XQ4z4H$T6`srw_U%UJ@Y(JuJE#t0PnH z-n3E9CrtNUjcvlr9__X1!oeEr#(eLw5=2|jEtT{+n{peg>$rhuQH1;WlGCb=L4l;# zEAgeEPLdd!3n$EoT*&44G7h8kJx_*4B<>+M6Rqmi3QgwOUA;+@asDZ&Tn703Hfy^t5e1lt31p&J?{qqRt zYoqvPPWFtfDdc3%OJ?l5ZS=5B0?xV-wBUQOzs{7)WZuK^R0RXZ94W8<&MMJ$VIi*t zEwOeurz)jusI01gZNmTTp(9Yb(1N2*8YmKvQ@alxD*X1ATkeK6m;RGS-Q^jmTI-aF zL{_Sk} zDeg(@($^26{2ng3(sxjS*kwhHgl+3J zYWa+Im?h_(w!}C`xFx~vUNzaP1Z=K=I9f?$dp`%7Yq!`f>C*BRV>7ejxFgXcYM1?< zdr`ksca2ZTcJ#MOb62%9dVL`K4hE&R>X-FxwBK4&v4t_tioC12{(|!&7?fDtma-g{ zb?3EimC+az3UjZdUP|Idr*U(ln#`8Eu&2@?Hjxtl_5eMz>t!E!7BcYdi#7C zvwrFHndXVt-S9bXs}?A?Qt0q<(E{Cj%iooy(b}qjW=-aHLuu!&EiACdC<2pgUNOoYqRfoD%xM4&?o# zpm6<)v|4oL+vOO3f>x8Pz%8%mi6@Gpc{(k1mT#Ka3icLStFOQ6R*l&fuXB(|cxrq> z^Mklv;D`alJDtt4ZSpHu{Mk~*V-6Qc@gC`#cD^LyTZzf6gjLJL4@HYmdg+-21j-d- z*v|Li0!M-{L97uq+#-H_Ux7Kqj&J68``2Y{Ez^VKcJM>*K-8}o(|tpE>g^W{`-6lK z?~sYx9`R@NB(2J3=r3wmF5Hm)XoM9nc-WO5oZKb98+GFO~%NfRXY68Czh#hjyWau z(R{JOc5tJ|m8KG!I( z{#_P^(Wog?`P)k`+B*#X&kiXbgQTs!T3ZA|KeAjCnTva#Fs>~7<4k5hfz2?cN3@s2 zun%O%9?B>`e(LB!Cj{-W2W+J=ROduPgAj96%I-ia^ZDa=ojigjam> zTYJzKx%4Y${31`-v5XE|a}CgLd)bhbwL^i(_i{B&by!#D#KnZ?i5Y8y6wb>Ro?^t@ zQk)%j*!Gze;}SB53kTF^GI;gM+J*|p4Sopj85%r8WqYLZwR7PO(+vf;XKqsgI7{3Mq0U$05|3S>Yv0V}~h?aehjv$Aq|s?A|p+_UaF*{}`;YafJdca?Q> z+etl^S&y=nZhn=yLqWH8*V9bCdMzL%#jbFxYxi&&Eq99} zk&T1sT}3b2T`G2cOs#I*RYLlHcrd4(?Tdpx=FEL)U?iHsEP*`N7g~kHo~eVuq>eYF zthCMxHvI<>!unrmUTE`;IAzq~~rVU(^tMV22^y7 zRz275OYEs039ieGZ|KJ4Uu8ZysrRF4Pa5OBP^6CRqsS_aiUT$qpJqz8U%tA8`C3g9 zJzM@5kD9rrfBV8mo1mkka&T(JPOt?yewe&AOIh(;Uri8v$jgKu^eb>@b)^pkq4z>Czzs=6&rXk$q3Wsrj_POvd0gEHtvtIKkM=k+6`7>g4#u<-tyo{ zzECOrNrjvW!c5Uld6zm3oB$mIa%0?YJ_s{#LS;%y{L`K1#eZyzh|2x&_VTyJPvb8v z&)mGwgJHgafAJ+KA?}j(!}w9eal$0BwFEzYzIAn_r!l6PpvC*ms1LG3pg} z+L5tb%K*&Xcww;?+8OO!tBAb$uTM|+S&1>;+*5|x^<{$EE%|FE-}8JO9wNVe{KCM( ziEgd$G?@<cIvHD6-d;_r^Hiwyw%Z z*i~0~0GfYqum`rk6p4t?&UcgW>*&w$}#LRhOKc+J$_}&YPF}nfF3|m{_Ts zi(hu#dg>xrZf{$2bxPL#xmk{2j;s{g<(QXAy3xr}**TIEC;i17E;N@=+FObS=I-gU z#zqBj?HbxWl-NTXO3*-;;=%b4^JZIk20qorvVw_`i! z>`lw;qX#uG;f<2>FPEUIWu}ieAxt0!1$=U|a2Y3G7R1P0?^S#En#n}H5)&mYNNX3b z5lg1L4KSH%L)l7u_Pngi43hzx8IR(LRdP&8!174ZH4|rYlbRe|X$w^4-S}O%mMx;V z#u=%QQ^J2`&eu*_b;revskhBHgol`9jMxt1CGRY?dmleHN200?zVKiu7m z)VW8Mi{B12W7OpAv>aKMmkyb}{`*5G>2!)6o#U+BOv=$eEDvLu70lGx9z~pR74lga z$txS*FPdT6<|I><*`Ag^?_R1ASh6~j*U%)-qF2S^rL@ud{nD4$&O38}RHi_%_FOhE zun4aLp2NJar17x3W0m9RXn#YA>&;+?Ik)&PJ1}2L#?j*xKTG#@^bqf``zDAEYI=OV zfxEWS3?v1!GMOq1PNH#dwjv!=9T=eM03a$^p{BMC8KHNq(bC*wQ1Vu$R&Gz?A;vjO zf8GSe+(2tR$5P`*wj`r8b$ULrC#0Nq0eUho-jzobUD?~YWA6EqZVv`Gg_tjb&J?+R zrE58C*luG+lR#E0n}pWeS{V)vqo3ny&fS^09P!J^15BY^g5g)cqnn)OFYk{@T`tY$ z4pZ#0UDENU_o$-57PBv-T^RU9(PMRCwWZN734sK?rQY>?0%H`i8U*9JMt>#MXA1botd)u_u#44>3FENV)}wDbNe8NJlTz)j}~YuaGx)a0XwC*7f9^ z=!lJkPo~Y;D=&r`GRJ%gtanyUq= zL$bU{oN%OcfBe+6mIm4a8NZ6RnCc`+p}91gKaXMry}JY~RvPYady#zVV9H5|;!{K+ z&#^UC@31J&fHUGdQ-$K%tx=|=6~2*utc_(cLUNao7kt76%_{labP&*ZsB#aAJ7krm zQYp&*H^l&UtXB?rOkL>LcI=PUpLJ44%Ds)wYhVD)u(-{j&* z7<n4mkEx2FF4$aTa-`D^N>H|MHz^)8>UDru+cCNr$CiAAPO8br##+l$j#No@^ zU#ZSsvP!H8`+mec%QiD(bLaGPwx>u&VRIXAQ?*65A$3N7dQ8y>!`O9dD?2rUus37g% z&(WeN{!4A;F68g!rPz|sC4PrDfj(EF)6i`hlg`V)YyN}s^}$L0cZUQSRKD%3?YNll zXYfunng6ILf`RWZW43j^wT%4T5K$}u5Uk;(6@lg7W^;&;utk?V%$c-Dd(t#6xzDCbW()k_a4M!OOC;9DY_=^K#(K%RBa-= zTc-`mZV~G&jYKa5rZUUS*qXkP{9(j3WJ?Oy{p>ERIfA>$&P(0YtpfFaTF7bmvq09usT z!waq$eNTz74Z5evloOiC<21;2Z>fiMd@%hnoM`+WX@;}wedV+u=$ESs+-yH4 zyz)0so=gqc5rXkap&6N8lmN`ogiNn9?W&&?f~EKwL*TJXx&fbOl8#77j_oAow{5Xu zj+S4_+n_ofbsG?2g|OWi!><>|MEr5x>2@hEjKqk2lF-?xwrSN+afot3D=v=W~26#N^Ej%9M-(u&*z*@UZR13jL)YfD~H`T>x zY6|XM;GlzstRK*j#Y9xotS{0h(B6L2uBZn~4vExHeCS^+fmt*!;&t-S9sJOSBh8Jj zo=HBKJfUu->*<2nSnGzxOS1ikCIZqQ#6fccx!;v6d9{_#^sI+BM4p@e*CV1u2#shT z8CfzxFL(ci-MgI55zJ3#4-s7p2%o}eMzYC-r2k|KFd}H2pt4q%m$-Lf!Lk?x_sG9i zRF@F9DJZl=0MX$8`=|XDG3~RS+Fn3wTzPCG>M{mRfA|8^M7G;^Zb4|bnLh%1$nt?KTMHJ3Y|Z_^(MPlkcu->y%+EKkgaALygn9v%{7hNU2`;&+wKx zDG}br+gqKosKAkD9aQHow7ND}rp`T-aP4@bKrM5At5P$NY+xV!98O*4p|#%IXIdW8 z7GoBA$#tzJY^240lH@fUXbu7IoJX#Tu@>q+wS z*C=5DDJwy`nR9f%7qbal@`6{}G5bY&6%74vSZS;MDv%2%eB>>V1SiqkSUC@yeto-n zP)!=cgf_CpYda!=a{PdW0pLdaKQPE2HGGo(RZ7XAePF<@h?vxLQJME-Tx)asOE`yn zF(ysZBld87*K9H=|IL+wA5TJ;c2iPFhOgGF^e<-H2=w6@QKIAY54^c+=nS$dzdQ5) zXW^#Mx85Hc(eVobA)8-*N13P=?`xUo=r6OD4X7JH?BeJs3 zM%QKZJ|@U9N2E;T{i}SoYv6z7H~sHkNHl&mW&I=;gUyv;2#9WjSspWec35!P^Q=)O zk|{%RAysv#P|(B8F5dj$J*XgXuz4|-m3c)mGYs5q9aajvTCZ+wP>u+hNBeRQu zwNt$U&1eHFcTvMjrJ3oVfNFI{*u&{Off(E6T(BV(|KSqEU;yX;=h;mjIj}zi4l{|C z^>5h9Y;KigxM9JLls*P4hDYJ_|GfPlJjS1f5@BM=r=C6Itf%=It>0;VV`1bzCSw*w zoRp*@U-t5UI;^=*z&45>OjwZXg|^V2)IrH)l*8F#OsHf!=^m~-6Ubb@0@W>KEHNuKZ5cPbM!KCt_6gNoZ^(5`|SZ$Wo{-|+|S|1Yl%3Q+DcQQ_i}T=q+39T zhVj41ql^fzVocLM2vI8&uWk}n4?uBhfMU+V>*Mf2jV=^;7e44m-8|+0>0Q$3kRsjD z93so?Oq9Q!OBm2zdeHoxOS_yxxY`a-A;)jjS4t)hVXImAhOk3ASXm z^iR$&2fChVhiGBv?8=fcKorm#$ihPnbom57^QQPC);>ZWG~~W2rg`X*xc0aC$9iJ5 zM3}Rx<05oeH3Sw`Dj~)E<#aTFOP)03=`EJi0PdO_zUcJj2Fv-w{1impWplE2t7POI z%mN=3vi$|6{2aXG$EJSaL zmx)CD&>gy#2n&aswT4yaMBpwFTN+_ED#>R<~&D(U`dvfuo9?nj;_#OkcMd^j1G(@#D{xBPZ%+;U3&KSNXAP z`|oAh3rF6$HjU0@Q`P^UWx1OSwSO9~JC(<@^G@6Hl^8<#8Bw8x9*0NpH?$b_0wf!U z{R_4&rLLr>c5vrTF!-K$-WN+ z`2UQ0DA0R>hV>X9#H6iB_?D`t2HxPHb>x}pdzdELDIs(`IrrQpzPY7o$af^*$=c#S ziLFouSWevbF*jEL4&A##+1KE%o8&7g`EE4O79u7QI_S<&{o*K2=)Y|m09H^UfexI! z^dCK$5I5mQ5tocyc?i-e+Lo) za|LbM8v!JcL$gkF-n8l4h8P}fxSKXSPNIai zgH0#6_1C}n*RSE%H(d3k-&so+6d3csq(3%JDX{l|kXKd8E3G+!)jWz4ZJ7F}{Qb9J zG1>o50C&zT8Y_|ah6%|70QXP9Oc`L_I~WqurRLGV@>%{rJW$sGY^?ZHTaa62u`0M9 zRfo`y+DUCk6AiCZ^G*??hx0bM07a(ikiwb$!3Q=a0m$th6MKkG%Z>9Af;)akIq-^j6uGa-?Nz~ z{cqZP)iw#bkUW{@M=JxizxiFK$X5nfq^@J&nm2bd+2?E_Lz8PZG-L`tIaeZ^O%c_j za}_5o;z$1`Em`gvc>74?@{nS}HpS z*;$8UW^caNkyCQl@AK#HkN&vt9;f&F{TkPJUeD`!S?caw_h3N2bv5KEJqin^_{G#y`dt2vn4ZC^pwiDXEZNcWs^==m%>I_E9XZb9|^4!1KS$~9tSJR`hM+DCr(%r5}pZ<6_7u!l{hzzu@{Hm-g3ChNT&=F;$hc9K)GmaR?N1dljjvtqEOPPUnPg5SxxaIe9D zli-^HfhO{Y`29}WLMq%%t($?dFskwfkaikhH1 zmm+w3&c&_i3>xP^9odZ5BZ3HK&wnEre1D9ga(!Y-Xk^fddb_?7pop&_U^hIWQTlOi zXvAAFJbsxPLs)`8Qz+o(c8kQ8*k7Z3C z=B20(YI$Z3+WRVUy{sg#{EoDUkm7rS>*ryV-|a`Z8IK^Qq+_eB zWONuW0lw8=7ZmdBILHj$L6suqtqJnw@oz0BpiNloQbJuD`XNYhPSqZrr5u$y4_F-4 zUp94pLv}@DJ5|IXP3r~9G44@4B*s6B0a?&>0CVH&lRaxExDLMtH+|qQ`6)i8hGi&* zWl(_TJ~ftK2Mf+}rK6G)GWG3VXS^-;{f7g9EO8W%DegUHeknHTwcf~;_0W(}*>?H< zp4%O_G+yVP7w-lr8u|BA4XY3d`8Cu-eAP=n+CZFSBjnqwAq`O}59RaZJ!WpEkji9i zc>8tz`a?e(GpDAh?NFsiEg|`OSKmJ1uJqY4eNKIZ&VS8$)jT7{vm*o;qDO(Uxn>SF zjKVvrt!wL+`};m!@Juq zvpoRKoL{f7Ve-OScnOe^kwMjaCokEaW@-atZ~8Rj9Z%2O2NEGi(=KY#p#|ICKr#nX zb1WN&DX#Uy0zOca68e&-;!c?&|J-TWIH)y0$N3?P3SA?RP`FOfN|Rn3r^35Z%grO?%N zPTmlC#`tJeazpwh_+-4YLD%cPtZt2J5RY?WToqn}@fn}3O?bjLrC#!d&o?osx~3Up zYy@!ftr|G=k+i;}q$F$P>ce>PyGp`^-2lh^hIT-6yL62#@xk|AH&P3_s{d@2Ez@YAS@#bR72P2WZ{dzZ~+|o;)o@dxE(FA&I3d4mx#xKRU#U(x%# zqo)Y6x?;cZYxOSqt;=R$c3U=>-HkOcFxnp*rUZ%FBi_3q^Z7{D8#ed*q43qg2NN0l zD>tHxf9e<9lc{nPrJ>OFk3cY+xxNsE-4i{;)J`=4!HkRP%j1;Ndw|PiO!(~0KbA)j zwZqZ1Q@;9Uiqya~9p0Vi{`t8{5oC!!Z(ceP!i$9oPOgAbAP3Kele?bM!kZ(e8sq|; zyte|AK4SJ0lo8o$4|{~mR3V&%CyCI4z2D&8m%+Vv+Jq%I0Z#+MKVX{N{BY}2cz`Wi zb@1p*NQ^yaphMXSW5;Q!cT7CCasR^itpyv!ZZ;m?_iX2?SWN}xMSG5pJcEf=@wMAM zBBiR@_u`NzsAb^DG!Tx=W;m1ujmv~N2h5CW4!2IO2jMnoLC3Sv*j`t|No{;C|B`TJ z|2*{Y(aA6`6T-iR?zk;?LIq1OY~X0wh>9yq3xsr=%7#BHSee1_0^!3K`3rR? za3TcS0X2fPbIBAppEV(qidQjkCLFo{!U=Hi8c)Ln6DMxAXnqY@hhHI^l`^j$11P&J zq+Pq+=ci>4OSteJ2Nxu%6&=*4f2!{KQ(3Ki< zZ2oZ=Mf*|vg0J$y_+TkB}kez>@yGTbjgAA z#Oh%ETji8gh|`J*aoR?RV7;}T1i4^wvK}#hJJY(L?Pzk>I{K=Mu}csTu+9iKixj`s zhtB|DYC+tOb{p;cDc7naQZwx3TS?b2Wdy}nC9)WJ;>yl}CTozpa#b^@1EvCpwbrG-ZMdwf~0;g@=P8ww<|xNfn~xTN+$~ zw28+vULXk58sl=_sA|?tg)1H|!QOvZAL%L-P#|zQQxV2&}fAsuE zPgghrfI<8#>Ved5l?q#QuF57bj{t}X;30LZC_T2XTkL?Hp^mQ&bd~nq^wt%^Ps)U9 z4C7W{z7o)U(MM28rUl7XoiV-B>wG9d^j>39$7tiD+hO5;icV?rO8t3y;qri9{Y3`_ zpHVy2nmJ@4ptyxeyxmdx=p&swPiCz(=2E-7IP#6|!$3>3*=ySLT3VORt zFyaB|PzyPq#pj>r4NalRT*LIA>(+8UIvtA3H!5%b`SWnp?lMF^?=f=-a{vA70$6qL z$+IDJT=O=C5O1xqz(BqviY~7C@capG8X_n?g1kL}`mq_wd`N^ayK#~~lAP^1H!p-; zKdUJcF~$Xl&!(J;=*-H&AAJ0gMZf6FIuKF~gY5&Lb}0Skr`yMis2HZlB=E9VVFuGu z=bwPHWsNT*qeZk=@0AK$tGXJ3yw%SPq42k&)YhMM6EkLzo9F4RL@CM^B{Ho){>{&^ zt%y=n^9m@XWf5L~(;#`U9Ohs-uIZGJF25Hsu3-dPyRn8Ief6Hx+Cmuk3qER9AM zRKV4l@Cy*zm&F}Y1^x9ZYuDFi8_@`kndZTzLM8^!t$%bsf0F!T!)p{(^v^LKjR#O9# zw2TxjDSXIGEopgk_8`%^3BG`73l`;V9Ngai7MF=fjy5O_ZnFQrpsO9};16B1ktXu~ zXYYg)(INn%ou{O!wN;@qf~I^mQlw4ZhdMZA!p{hV(D1?9u?2`zL#YzH_&tS{GIxXK(@!;+|3p3Z1*nTa2{BeYOu+fLFyYOI zZ)&%ai%8^t-B7pL2r9rq8H52E6UzO)2Mlby;JU&0YR@$28U*RxRve>LWQ8z|@Wp=* zbu^%GeaNISd<5~^*yN!qpk|bv8X0(_kEQ_+%3mTJw!dNupVQVN20cOIZj7T&&P=QE zcVx~plB3Uvjss(3Z=0se^p=^!aN-{k?#t$Y1wLd5nO4|y*O`(5bf?yUP6|XIqM7(s zeq>nO+xZ53(DpC1?K0+?rtM>&9)7(b&-+}84&nm3Rig`!rxK52+FcGBV&4qEMhH4FrdDvNzCF{tR~D)8+lx9eeYlKROD# z*5RifnO`rv6fZ%6h-eL=o0JLyDVp8P>xb^|TcEGWYwcw<^_+MY0gky3)n7{O`pJ1J zGhNdEMbK$(V9DIw!{1{En<$3gM1zr8Oda$hQ6nSSx-0Jf*iyV(`S-#5Lu?Ezc2%0O z%_VIm-#|U##hXU18HlkV1j`6+}mb76>&%1_WcT*;b|+a^}W7ct0%DNLDPJ&07Wk9f~YaCR^+EWFIk;rZI7{|%~c3{-!yqkb;FVCIQVM$F|A2PpNo zWA={M9D`Un4~UDZ43Hq2{|+_&sQ$sk9S2QYJ(QAfqHDtlU|J(;Wd(9~X3&C|y=`5{ zoU34G3H0ubnHlgSuW4rHlz=?efOeC>>#~q`P6$rV)4%*mei;)UwR@|5%%WUs|zcQ>V7*K&4Lk9Hwoty*1 zh1{Uzx!K2#L7#j?wY{D({CI5vw!~sb9b?AEFgAvTLiqZ|FV4_^1D%O#JoDuebpuSQ zyokMKQq_FIYRgaKvDF8q7=N+DJV>i+CovwAhGdmcl-z{N*XQ!#*K#RQr}Lt~8+!|TZw1H$V&etOc^5nOek1tHWwi5$J)2|Uu_u+<$%eZK9`Yl4v*ut z)4&{VXny=9HM~}IoCVSrK6f1WrbS5If&7+*GP_@e)Pu701mRF^l8^(+wDDynU+9^* zCNHF{D6G&;y}lcss3cz5A;-V`exoZri3H-upzH0(V-+u0xE$ubOhU?4bl`Fh`Ky$9 zxsIHdEq-1BfKZ1Lmh#V9?=BMLN}`rYdM+S+s;per>yl`2qD;L}t(%|y*9$>yw9UDI z1MJAa4e7G=>c~mt7j|RcYN^VJLKiLY1R(PH-;MmW_FkvpEcRAxW9B_WY6DkFQ1aC` zmD|dy3gHa`4GXS&1R-$#7ie=C!}0)M?RwSqSz!{6!J7Et8z{MxA%vw>3@yd+*uD(m zI{pVJ1#7G=6U}9t;n_ZX)yw_Gdf0jrXa&(&%#?;M3dkju4wyc3q(?o-1~_%XJWh zG?AEH-0BMlz_L->OCnkJB&+}--T#X-hKE&A8Q0@p3bq^ak*xzix2+MHeWDe}az@3z z*_lj0euVI65L)asex!xkniRT=gwuH9herE6f-NG5DPF zr&>4*tzm1Ia=%oSo=Nq~KK7dL)R>#U2%4LEkceUvV05krYUG=r&j3gJ(?Do#Yt2VP zFX#4R8Na75u+^wnA5C(-+>#)nhj9E~cm&v0DJPwsddA#TNlCaIv{T1`-I+>rg>Er=?(*<3;&6f+Cuk`P}AJp=V;fps<10s`SrC`M@wk zMMiSwh9W$r(z%tAi

    IKs; zB;t>GGYS%9@8}8FA~B}nmITPFoJ4z0O18TdsRS;4e*kc#-|^X){nyNN~-^^|tM_@N2sYw~&o+lz*q#_;#+fm>F_JRilD&HSkDC!!#&LA|9HB{~zfJrLKwl-deP7>UEv@ZVPMsv{DKt8KsPe>F09v#_k6QW{= zN-UGlnLTy@s446`mP4Yfz{CLQ-Q*a>(Pk>8mEw=b&FWSthK-fEZDK% z^_0yPtPKbRF?fX0vy=iLM~X7oHKivEVKLlOnS(TUmaVe6N~c4S%=B;o1)z?%=sW z{{tEfU^5$_g!nHm92l?B?NfGTW%uiK&akKfv!8>EA&%%C%8V9a@UlYDD-gsufJFY! zh!k2EpoYQcT{yXO>U|Lv8GgLLr#mD`Tc#UCjw??dR8sxIJ1hdTe*w{}_4(NeO-5kV z9#*owL6xw`WH;-}v;OLNKV0fVj~snNIr{u<@mhZRp;o&(c;*90T7R&YGNAiuac)H? zEEBC4o`LfN-oK;rfh)Pt^xM7XDK#_oM+|VW26Jvb_pA_;nT(WuEznW><)0h~!8ZKK z9jbgeZ|AA51kKfJQ_C3L!6H+jsDG?6;h8q{{}Ff)ctHDY47mPxOf`ii$2;W(MQgax z;D6MJon-D%l=57{gj-vvgdZQ$!S^x|^7M@se4UfHwhwm6lTOV&7L7P-LO-ijmt#NY zbt-hI<_$DBP!4`rxf2`e$mkL1Ks2{*Ax!(JtO!ak z-;}*7z#ipEaNLaUA;e!2g6rQhOt0~^24&|>Cu$1(C=*E|b zr`fOUMe*^YCc)fps=iBu!VopSletdE#`*n6Uc&pcl6+k)<`kZfwE7q*sMDCY1(VfH zNGbnrcwDv#H59n+Y6`@}gmQBIqd1^%exv0Vi!**or$Rve_ zKP!)d*DbA`9SoS?eNq3($ZY4J6rMYI+>)^^YM@eou)VID{YXFlh$_{*E5@_E6ZMK{3QN6h-w(#W}yU{k%o`EeH zMv6!Cr7k!|RHGC}>}0GEUi(*4oFC1LKQu*fF?#Akv*;I?fc+2JC?^xbYc6O1#7s!v zmF~uyZR=-u&YA6oW9D}9>tm{Yw7Y4r0Dw4iy!kGBr?_=W)c ze_omFlu+v2vP%2LUK~Mm;Cl>KqifO-&y9ZuzkAe`6RzRGZ#(+3kmVW-ioUw70PP4h z@ZUSIKMgyI%0k~H99I3UWC)Y9mB|;W7`FJNkX+hx2&mJv8_m@E^nf)K{`p${&!JJ| zl*wIyZ}9<+>XqgDMZ#JYZbEU=@FYSx;=g$oWjB&@a>A`Klgn2I#+HIAy8fMnb5%*| zc_^nl51@mC?=HB=iT~CPm^5wC_~*(1aP$(j?&+|*trm1EPxOnkRpTw(271w^V%JVh zUU}H)`a*4rwZv?ad|0S{B?``KIFwo&{<%+qi}---LEwomk&aaLkltdg-h1q>$vA3Y zEQHvHh4}#*u(JO#3L96NQ&f^M-QAS0(v0*c$w81Lr&l7Q`6NU{+noP5^bGF#peVc6 zWYQCIf@M+tN*dJ{dfAq_p?Et4WEWpv zj+a)zPulp?zVP}3i&Ws9NaXdaYlJM6<`wmr1nQ6_&S@UV4}YaY|0x(`qzDen1n8LE zWle}2+0?C(YG`N4PTa_Y2<>C`c0CBpd2<**5IYMDD*!_CmL6 zDrj;cO>kW^C2l)EW(4BL(fu*OL@@Tt_b7!h~trJkmWtTKZ zMQl$2;dj?SYTjR>kDo#Slzm7}qly5w48onDe;!hHIx}$5?y?W0{Tm*+3HKRBQoDeB zOt=X9d1txEgMvR7`p7GY+NLx%zwm~b<2QG$3>s>ssJSDZCC8Y4dd)oYV-ONo&M zgWE1c3q~C<*WCVUM<@YPZ+7F>t@^0?mTc`{KaQ1Hki(~K(VEe*>Bhfr#wNfEiUL+3 zqHfnH@*0#6IACnh*8ZG>VQJB?KJU~3vfJ(rw|NsxfX|Y3!@J_EmHd+C?SW5RzoIT2 z=!x=qaft?#^<_hX{AW%|3KI^SqWL4^p$Cus4qxJ%BZ&rMpC!iP-xE)1;D5?V)Ba6Q_!}qykFA#mtT_?RArIN6E0ju5s~ZDqG+h^0A4Qm2XV`0=p60{292tj- z;hEpnpG}#hv4keCmcO4gvl%$HgZ^0hZw%<>Px@C;tycu_r>gqj2e}Z%4jfd2ht0Ii zTnaGU2Rgms#VJS268>!R-W=n;kT8v_0DCvv#DVvjGq#o?+vMSwg0_7g=63suNwmqs19&KetTU#|{B41aB>?&3u7tj<>1bXF`FN?d z5z|sc=p_4Sk1qT}VzAB+P&5F_XUp7GFE6{IXHyu)Fuz|x#M^Tr8~cg#tqF-ilP!U zeMpPXT(Q35ZHB-F+4|ykcvAGKOb)TV77R4CsU4?4rMQR!=d&n!|Vr(AcqnNNA_yT}$#-u&&Xr0I=qZAjFCNkgjre&m6pRN=%HKt_HmW;gjERk(oR2RwVrqMPLq!uhtEx(*?k}@MbP~e z(V;XJa&6FsYO;7>n&J@Qs9N&ydm_h;>)(4v=r3EDSXt(NwrccEI)8M^fJury0LG2@ zZb8)dpNhur4bHmhE-X##WlV)PMl%rzY)-HlXO8PfYB~tizE! zQI1o3!~JSH;yIm}Lvz|bXY5T%ew-D;yjbQoY6q~4p1>Rq`xA<<=+U3Q%6Ot+#NnQW z=J!y;^CQj&$`J%i`6oNLJ$U^;q9qCrjmac1yn6wqv3mAqWlptov9u9@TViMjr-p_|FFzK*Ho$Wy zsi&@hWmXcXP)%r*T`R1bQJme6PLrQKDhRe=e>Uj0wJTk&AZ2NM;b|v z(0`)89FpT`#1YL6Px;`t6=7rra(49(&RJ>FO(I(gwjguj-n8Dm?6V8>HV?MiJQ$^$ zd`Y2l*#F2YjzZ*_v(-h9$Og_FG{U2Wx~!l9Nc&tfum_Jc93zFMsjCopWZmWF9&#Pb;z6|4$&cQ5?^{+dSm$G*UUPFKi_T<`;T20<1h>NZDt$nLeEa4Y? ziKvlcdfU+y3rOwp4Ud8YI4%Kr4g)dFVLJtLS_|s6Y>mC{OOlfW5mkI`(53LkFCNF~ zwc`RC{CnTvydU|FJP(6_trLZjS9EuJaebA$qwbC3IUB>t7TxOqcq<;CwudWr2?AU= zCTW~LFvsxO9vI;&*hhJaTG0AH(~hqX>8Y1tY z_n>1}tnM%Q5bAGAf9bu_!^6m)#zc0bvT$KncJ!S&scn~BnPlY7?WC27RXU5H>q)<- z6T6$h1t=9`Q}I81Z@ccpR1jIF(2%rFf;ED}9yXX`M#2VXB;%5?b(O zX0~|ELSrW5WYxhBk5PkWcf=2GA&K5Xk48(NL?t?t^BB+U+d;B2wJ@}bpk-tlkxp1( z9jFVYA@L+q+Cna|^>D4~9ZE#a_x#I>&`(MnZT#aB8Im-eh<@!aw~&ykNyK7rla-CS zsBDckosS}izE(8V^PJbe_loHIR_Dvh)($^ke2?y(>FN8^h0otz zvtpn9zU!+oU4rp<=jB#_hn-t4I*$wW>J%@r z>FW=XL0WQ7C5`FrT{TMEiIaNq{zaRYJLA|x-oDa!lQGv7*U2yY(#O)Zyq9Ar8QY*N zA|^EE-m6I_)Y2?ls@qyLv9!?jw(ks$!KxPGxns4nxarDmqk^4LjC`G|8PgTYCZ0|T zhZRg*aSsff2l`Gm1dA0t=gF(-%j_<{^Tn!tN2SlK_DJ8fv(y1mTR)}`Ip$m4&ehqB ztc=@EGn6I16bM!%q3$fr=2=M?q>0-}s#C=;c{Cb99YmQ_6e$w7K$mw1Dl{Bjs-G z(Q+}(7D>kPQS%cxUcLN~#V)bDB=Jux=C)pZ14$H%1~GJE)v5A4qWn(HL%CU= zC*t`#PkzW-KE3bxjLT#uv8ZQcMxIB(s-NydkXhGdJfLSQ&COxMLTP4(#Ob9^ zsqc_qNTfl)kZ3>7QFcXMpfo5%ZWo3@R`BUkaI8}oYDejKh+vk57~j6*dV=~t0+)h` zWPDWa_-IuKz9@Mn>mf2gKYoRQxmkv8*7i=%O@SY8KlF1;*t2(7S(4Mun+>|&Ar&(V zBW7_j@>b$qEo(Li5}|9(oxikPI^cyKw`)y#DKWtxr-3QDC8vjbe_WlbXO!o|xBkJ{ zcscJj8K*i@U9riE#MHCBOs4b9Hd!A#{ST(h9ZG$lpfg_8lgoQ3=t3^VF5d4SgbVjI zos~n=^;L%2YDw3U4b~iwvfXY#LDn$K=ZErY;I})nIMK1-x?TJL~TxtT1IP|5jaamHU(U{2-hIkY~e$WC5wAd#DttIdnTWXyyKtr$FL zezJafPzHIC@olPa?{lS?YEoW5)afNV&m<3_GwnHijAfq3vC8%NWX@>hyQKynRKI1t zDf-NXdG8^M3Re?dx1eRTt8-DKPIOt&`wqkr|K!D<*3P!VS<&U__JxZNBo-)0D(Ctc zM~}2l*0=eEH+9E7VtwMSct83=ao>Z14?2rt%k#FAAyU-o_oF=#HR1_(>Gi58u9>$l zOn2O`!O7U;E?1Q%;8?HI?pB^0W?l?t%iNjW_>n}tv$M0%O7GNrqkcB2_I-*))=O{g zr(+lLJ~H=CiR|ufF(=Hk@A5ZVm3JZM=-EdP>v{xnOBX+OcoA#~r<)1jW z3o&cZ%5Uv=-wK`B8EF5_@-c)Q?_M92e`aniD@M&!sOUPq9Ljfy_l~>VsJ*~luZAP6 z9^FLC7vjCts|D*NPj`DpkIMW?cLzJcihkJI%-^xWkq0%&EnXLIU5 zDi$%pJs1mi=6kLmM|WJX%5fne@@b^uA(5!us;(K6;vF_qYscB}W8o^bgk%>2lOYQ^EgtIP+T5~etx-Bg{<+*2rXNhEF8a#BOd^sAAE zWR1m{`J8=y^vt)$*|Hm&sO$P7m));+9-pTYxFB%qMP$+tX{JKWYM1h(5{^r@8AmMy z^*g$bV(#uabWN1XwQHW3ucoDR$Zp?+Quy%jsfB8HsfemD{-nLa+^t`iV0x~C`L1f( z_7lZHRWZoy(k_4heUuJg&bSz;9KSvI$Z3Y_i`$9ZPbIWd%FgH8{oQ!P&$=I^zvX1s z`JSiQkt5#n!@Z*8absz?iKiN9jt!i$`NPv-i_UKKi#~uZs9Srn>P3R zZMzS2Y6jje8ZD>7X6!%@-O~HtzQ@?Z|?Dhfx1VWR>iZh_A&*#bl_k&xVhNb;N$` z4bhfVG-wyJDJpfWmNE!+3-B0H?GkEuQ1V(prYg8zskYtAsaLXUwnRSh_Vn)UGEdIm z!M;B9kA}H&GSyFDSP_)#pA_= z+_oZkY8%29>wI!?%xbZvF5VQ%=j5lEPt&u1+J746M=gg9G ztZYPs{$8JZA3JdG#VAi$Zf1^f?Kd7-*;3Wm?-=PAw#Rz)YR}jby+ZA^FRLYE`RQb9 zlDNKS>+ZnKHk?wn+2^9W6zij&%E7NLZdre&;|1qC>u#N}`A<)sj7R6lWjJyw%v8U9 ztRQcYo{bm$S`p>(OwajKnf7Tatoozo&Y?+18kN>J(<}8g;wfB4GkyyKG>K(dCUG*N zqw>}+ZlURG6gd?8$?j?%8bzY~=+q-memfQ= z;f%51`N)UI{o7l~(UOv3Mhi?ewZ#K9sbhnZGC@k9EO?s?<*CHcARvw zKkMwD>~2~^&sj9Ck4j-2t=s3DUo+bvhtoIjP8QsUvrj$q@v-qr^dXB20(E1vNYQT2mp8i+-=;h(ZTBlTSZvNfXP|Xa{(+b61w~n8uFxQzNP23Hgpzv<<7R3oV(i(x*t7@ zVxkH({ev@V_MYdOZY4ZaOfKuanv}RyI#)Q@_Ee3>xq$d)6N8dR-Jt@6xXz+ybWq8C zy1~!(5EryLSJbe{a-au>Yw4?C0tcew;?c@wiTl zI#!$3Pe)jG9-VniB!fN_J!Uk+pKvmH_mja>ePdV-dq0&<-m?v}qbNbuh46`PGEcT< zcjx7>qdh&ED`=x{H#IeF6kjB`{J^>$FpQO55bYDrG9T6VJf=67^$=yG_Q=4!;EY(w z{1TVNM>0W~em6f$iwRfaDTy06#@C0oejl+l4n;j%*m5-A@%p5uE$86=!l@L1?hchz z6_w)YCZN=cNE#xyY3rE8ooyt z)?1)2xKdsfX+TSSOTacnEicQ;>}dB#85 zXTf_yEHP@TCMh)9PcH0G@8olNZ?5#mzU}kaek1X|1U)w4^RwQrJlSTXvIu=xLtxDv&6aZhzfyq&=r~lrkzetnHx3W62P8qtLO7 z6aMZ+W1@-1-Q)bzk{%jA-c2c;oIH`4HXLvZ_p!Ip65Gx8q?HWesefX|<2rrxO5aSm z;jqacUI|hiWhY?2^AMvH51O0oWdkv_3 zVzaku1Vr|ot>sW196fX*bG|&7HZ1$e>jo}Smmh6XOjct-?DGlJulQb2=TxJ&D{B}u zzxE{Ksr#OFXjUsOLp#Yju{1um>CpB?p;UqDW+T}}cS}zCT{RzC4!*G7>!1`=U~MmM ze}>i7@Nq$lh$T0(RvTT)yRS_SSMRcEeCVj-Q5E(z0NvE$Ml)h`+ouMtaS3fInmFI+@33VDwq-)XZa*0xi9( zxj$MtPUXctX#bY;rg$>`Z0)|?e%rK?C+B4+1@#OQ?hAE|^mMnrP^#B1NG^LxgGiD@ z?(RnOj-rmu$n9vqf@Qhk_pHoIMc<`FQjPA8UXKCq;KZ0*f0?keoVro|l`y!2P>3I1HPTn~@%)f}+Xmrr>cdTU#ZugG2QGNh z)|OYmq0|AxXQEPb-D1*h|L(_>kKKA^{lo!#IP9iZ62gBe8{Jynpe7#O7VxNxJ3Cxl z_=Yg1=whYlIIsNv*(%QEaIBDflS&ZQ{;r7QH*aKfKI0zwg{$Y;nrs|wYq5$wn48N7 z6|K)pzh!RX8Tc$Z#O!ZWoixffFg7!rfs-4^d3hf-MH#wRiUDI3S%R8!9MWu3;_5WU zec_*FD^cTp-O(M~)?=#U)uLGAtG=2HgEd|wm3fDD7?pfV)m!cxa2RT|oRU|gsco-U z;>nM%8XFT`y10y~vvv3AuiWQ7JnYs}%HlECnm$M|FP~B@hxuXTV!3kkgN(bTWkqv% zW@(A*t+5bm8^6ekH*vyu-C$sl%xw9^0CD|khw7y}7^{=J-OJtQwGbCFT60j|-%n|7 zeSuB1z+nCO-q@hr_(Rf`7An$_F$Rf4fhfA!#(ha;OOelQaj%QCSie7xa4bHzC#;~i)d_?1yLb4BmrAH^2|O=j=GNXWi4b+1kr}(yvs0;0S?@yrotV&` zn~H(1TCa;VckkzEwMXd4Cl*}n4_2Kn*&Fh-N1X-{TvvdK5=Mu1L|tx)6J$%;SAv?I zoN-U*vao%1CM}`hVlFyKySin}zx4$+UPm@pKeFS8|lBt8@kLxix_f5 zBn=IkFGNJA(RCxkG=n)KO^#Blh`(ZT$*cWNj-p=Fyk4^tSx|CF{_&M}5)K6neN)d^ zV^2H^HA-TO7A9kL5qoVX4}z?^@m%(cA0uvA=djpGA}(lFr;y@a#uT)^ZB%0on|~se zkIEJ$y~HDTj8!x(?X^=RN`Z;y`s&vnI@#RJrjRGf)`Q`nlzlr72FIs~z^7T`UI$~z z5zz0QM5(Q#wr?+jZqJ8p_Vp}<7y0+fif4LWAS<;s!>P7IJeZ2KJQ-$3)2&c1bKOd} z%4@n1<#>2ba9e7LQEmvQd%S>r;5%@Pw=BeeMCH#tCx10oZ#})|Q$!R>slEUPP^^9_ z_jq5~aL`@ZqiT_!Ejho@V{og`>}-v;!pa+$WVYyiyfRg6(Qu_w3vcy4%s16^?-wgM zHK9)`b_Ix#sKaHTutLKs7ujEj8viNopLjr^(mcG}-UybsWmS4pVC?Ae)`T?`4y7*J5BcB z)yUE#3nj!yW&Na!6UqEoPVw^U9uhWs`^?*Cj01#oro(n`N3?fx_aRq|t0h$z`u5b8 zf1clSq-twXsMF~$4L1>LQ%=s)pLWrO7?lQGBHH3)X@J{{^p|6Cl`>J`%?T^5Y~b@j zs!)kW9q<_BWfiu0Z>OV(YrGW|o;p7Klp}%IbRn#pT81AryM}Ymqk_X+Khiwt_ z?@Jp@32UjPbKYkKKSrcziIgJrMcHsqfUsY%Zb!>Ns+IHs$+kK)2d>Y&{&2{BX zu8Fx_e>>GcAClA&HBqKehl+9!Pz%xDK^Sl@EE!ym{&ALU$ivqNa3B& zZ_!04iEAG2*QqM@V!p7&_8NYRwCexx={ZxiIZe_>R)NV+t%?Enx*c#umuFDwnjS-Q zdy1{jQPI1|?RA-Zna-&*jMPpppW&zp4RSuLP*h1SnTFlXEw@OACQF=c@#u9Lkwp5Z zH|y~&&JKKI`fNHq~{ezhN>P?Z-6 zA-p-l+eNK1p2-vYKRI31n-kBY(5|YTgP>P!=y~W-bptvM()T;uA}qU-zi=lG+K5kQVX7PAX5JPQcdC{&cAOX#Y7vzPX_9$w zavYK0p;_W|C$7VB@wsr`vnL9g9*fa}74~(v=7e60C`=YDoDyCrPb*d3#+pOn-`1?S z^8JO>l4|M$hcDx?O-VOGDzyeKCL?B+!Y{M*RlbnYth#D!xI5AhN7L|TfO?;|=X9t# z|BI-Onw3w5drLhG!f_(@YS-VW1hQHO*7YX|jWsSsvPra@ZA&M(6nh%8JBY5vZKxlnizttSA*p%MSh9<*d1kQAf{;c znJm=GMCZ%CIWA=KU?ENH1AiH=gl6U~I*ExohOIRZEObXiwah4bXL5*=9e3xFo8U<2 zeS_5{w&cC9xlOGySydA2A&IHn-$tia{6n#5_PD6+I5$67J@crGtMRomH+zt0aFU@B zPoGaEcb+%%LYeUBXWL8ZS@gyp=+uszJxt4)mc;Mhzv3Paca9pQ{nEBKM~#MWns*&(DzfyEDCQpU3!@ZrcOK`S z9lC(|)MFG`ttImKrhiby&BcQOPo;d?>duR*tYqiqMj?)5ld#ZwXDI5Rajr1f?}4Pa z34@GsgQ1SJ^N}ZRFSfqmY=5foJRQxHXB&U}O1A*<_1#~ZbEm(iRQAqEHeYGT|7K$oLwfRw9L^^+?Zc;B3crv=zOSNxn34sPiP3xUwT--J{P{q%|pQCuRnajj83S zMqe{Y7tfoZyz*w!GE;du7`=!v^&Jl8r2>);x6R)M47<7X+nu7RlZk55xY~epL!v5P z>+qfXbXwg8_r=^r@nuQXtWIA_qJX(>)$lW#k7Y5r0hj^$ABVCftFS$-xg&41odxPX zfBr5-89*!QuWFyBPBC*OWG0&prCiAA-ZRWua@wmL$FRKUz?)QJy_mx~w`iMM#jn>r z)^@3XY0LBzhe#di{ElI(g#uR>mg%@iVZ*N+*MuMXow+6G-ddo6VKq-WQy9aH)IXTf z;}!U*)a>ZJLG7NU9hjnAr`8F|;*f*03Aa+>y((Fe{&U7dgLQ5b(k}+{#}iCX&Z>GI zd1*ky#o#4iJ2GS5nk_v(qQ2NAx$to8Hf?e!zkuB^L&@oCAFF}pAIDu<4#o?AJv2GM z6@PxwDaEyk`HH&@P?eYSgkr=DdmQ^ zhhy?mFr)AjMA@abPfi+Gbc%g6lqe3b8gdnhY+`#98L&^(A{g`dyyZ7V+UWxZY=Sd_ z>S|t*)Tylgt!Cs(!}6|?<6%txM!rn;q1PJaiZC}Eqr{R&%K4NXrDUDana(?UzZNqn ze)Svr%5VG>72wt#>x_)?O$pVFIdQUv_lHU*g;4R2p(E#XFmVQUJ%+3a*w!M&oF|hh zt$x%z)WR|*XEK}XnTbMrna_!a9CX<~t;*aYPusBiS%8FZes$Hvw8D1Br=uLaelo&* zxUB`@%uYAM^SY4^ml~!A9!+bxh1@M+mXXKQe=;&6{2>sNds$!rM&vdX| zpVPENQblpj-ZRM}a)en|RL_0rq{+kZx$!_Evs03u&MD8PpK}d-jLXTBvry5wf-;^y zsn>Qh-cc<|a=Bepfqkg@YFz;}jo`F4#=!0a|7Qe!CM#mc(){?hsu@ab*V4>RwXPf~ zU%iV8{1ICG&yN%jy(uYEzu9$4cH~ZkxhzYtg~t5e=gu_K<_=mPI3qKVcMa)in<=Bd zM#Q|~OQ?`)m1ew?VE1)!+uDhePo?H&YRRcgUy)n+)}%x&1btC{N_*9~Ow=FhG}kN}F2tvf z?-q2uNQwYWH#u6^KO7Lwh13n*&XU7}b>lc#$ z*u~r}$xrm-T_hP%OaCFO+pn)_x+SRg<@VKkSiD3vJEeD6JLOC}>5B6bFWq!HV^s3{ z<;mkbko=OWye;WO|L{l(m1R6=X@ft-7FUXcxK*OVRZjW3cYOf1fSA6s1rNeFV?_lk z6B)SvT}qvDH$lp4FR`Xrwf^27ns)At_|>^#_3~plxp-N@TNq0fwZ8J|>FORx>d0y% z#nTI?4H*((iv3_a>`Y0*o>7)@7{RGhJ=>R>b15eS71?pvsc2WSBBBxj@b{E5N5|N zo;Q>4YXUa1@pe%PD@?@ep9r%v&@O|DEiCq~eKGmiPFXt+sgKKyBm%Iu1#!f6@$J1t zAb4GxQQf^+;r(?r8qnFZq_5m=tvi<2YUQfprV|X&qUwI>Ra2fx=uZ2bSgc`#jONQ4T6CxAza({fU1$OPO6C-4l+r1E~)LpIj zi8!N~-L$CUpoTKQ!+b#Q=t&2j6!uf-c8z0Zc)O(8^?)xKJ!iB1=?8ga3Fpb?=1rV> z@b5&;cd|&1QcEniHP3G52kO29f6J%ruUW^dSo(GT_X%_V4rVT8*9>eRN;8L9WFlGuH$9uO6I_^H}7te{dSPx-Z0KI_Q0oOLyj5SeekH7DX;f3C|yn zuJ9J!M}D!dGwvwcpm&+W2wTBL^P8eYEP=rw57wSWQ|}Udce3)`>;Bx9+`}jiq{HaM z4%)3#rk&xy(I2WhHRRQBdb8s0B=e~4(bxHZH-0)(n)%H4-`2jaN}jZOzgLif#%$~j zs{qDWgpWC=0Ki^CD(|hfZADp09zMtQ;zN|YQr!_N&=^r2Eb`8^syip6GB3&{RpLJ+ z5{0;F^yFs&Z_E~I01wsVKUL(9RLfcOw+*p~+%MrJaMh+sH%i6id~tNlG?_EL!@OD$ z%pjpj6WK99+L|{LLogprD2ohfynm{)FvD(;K{$<$nGEV|6KfERsi4^?ohF1Mm9yX> zHc2dXEmxM$uFh|{F*r%~L!;-326HFy0^7^ABC5jNJH-bTwH0n!F0|py=gv60bj>|X z|MiiaQb))-*@`H>pcvl(v(>`l_xg@_CFC0=Ta-dn>=SGRN;udDT_-rwu7aNM7b;_m zwmmXoFYh*lUE}Yr|HC4dtpJHWh50`&(I$ILOu$0+J@hRbR!yM4i%5Q^x+HQD6HKy6 zy_n)6UxTMa+Z8?57?W@>6`R1#_<_%+N6o}i+9pYi?;R8K`zt~VvFU^p0I$(|u>vi7x-JdU>F8Tk zehjYM3SVyF_yjx5ucSgb~ znW%p22V+pw>ze5X1t`8wiu>m{m+;Q;=VN_WX-Tq-%XAkbDrl)qWRYaE5@A4VAtW~i z0jO`;&Ux67sDGdH!m>iz%rpmGdl7-qH;3W?7W+yJlSF@uPxTy5FR)%3ilQMwUb>+! zW4Fa(AXa5*!W(eG)KgMsmeHQTW39QYL-tpgJNPI7i_`lHe`E)*wCOjk?!Oxx?O&?Z zW{z*9SKHomm5|_&Hm5=DJ#f_h?N&X=k2U0be+69Q4L`Fgj!Pa zz`OGMG_tW} zS~urSVet+0a5!*;dV4IsMQR!0IfvC7{yoybh&Q?d6$(a!W)$vsU(!eR2QrIWgDi;@ zV^o!J#0{Ap5Kow{EF+v2@E5a}KhSqOwqf_VJ3vuZg>8|IP(Y{?fzQq{qsIcrA?N#2 zugrTdaK;`%NiHeHmvdRw^TI(hzlheaXk?GDtr)WZ6AiR%z8Otn>9gKU!6bX$2CP)h zds}+b-3X4-Esp{`=S>VnJaj z<#ygi$XcCbKw(FWuc~4??`tVnoX+z?^%_32tqO%%w%^k{y>R4|1(=3CpNB1*!fdSW za}jBC^CR&FoRaE7owEm{O^@I{PB{yOQ_dLWmS}Q_fEc?h1TLSGWZG;-^ufL&ZYNrF zIu6g6CN;YpofFX6_lY)n5y?78s6qM$FEuVLZ=ufKLqWA8misYh>N9iS{1m4toaFd3 zQAR880BYcH6P?cSZ;o!(QvZs4E7kbU2w~pv?DMz%#6DXRAXoV(Pak|=7h1^;zReUm ze+nbv0KL~G{>wakas(rP|2hOS{N$5OvMV4agRG|lH&Yz{pa1{HG3?jgI5-wQ{$TMG TS||UzLv#Gi)vn$)@aca5!R=cq literal 0 HcmV?d00001 diff --git a/ALCF/notes/assets/cpt_images/strategy_cpt_stage3tostage4-1.png b/ALCF/notes/assets/cpt_images/strategy_cpt_stage3tostage4-1.png new file mode 100644 index 0000000000000000000000000000000000000000..fd4d59bf4e054fea9185a1379478c63ebd5fd000 GIT binary patch literal 125954 zcmeEuXH=8h)-H;I1w=(qx@FUw0s)aO3L+p)h)5F@l+b$z3xXg;sTx`YsgWSPgHlvV zs0qDD3xp1#hIUs#w(fJzcfb4N{ybyAa103VJJ(#ZKJ%FdsNYefKE`;AjEsz0>DCQR zGO}YZGP1*cN65iX*fw%!!9PcBZ|ORak+D{hz7JTpI7yR{9avUZxqEYGXXj2T-}cs4 z^}=`X!LhNih2|MD`|>k}FA{jd@wS)e=H@mwHm0Vg#>dC;csvIO$J@7WmzS5z%gb-v zxFH+e%=)4>8#~pFUv6w{Twf=?e*JoNb#-QDrnR;8<;$0az3E?v@<(PC2Jv%a_$f_I z&8t_ha&vRH$2NW&F3IeSBhIXMcXw~^>@YGiZf$Na5jV&kD~Ep$50>;QDJjhpmghI- zU0vPA*Cw`B)*A^e>zf-jH8savzdrfYz4|?Kb!h2INW&SgYHMrjFa7g`&iR3rp--8a z9~y8mF)@OIg2`g17cs5uUkH)>N8fgr!(MF1P zE_Rj_m*^z6yBG8?6-FcF6Y}+t^L;Z@UN=6gC;#RDgGgAPbP!#LcR-IWF840<=XG~w zV4mcwl$d>1ifR4+w9>EInudl(K|x{U>)^y9@q<95k&zLqZEC%@Au%Z_Jw07oT6!_V z_xkngIXO8(LPCC}1LXswgodg|j~=xycK7%96Q`G+KYw0VS7%~k@_nHpt|{n(b-dhN zsI08)&c<2?e!0E9JuknYx3_m=7+YbDOifKqPEMX#!DA+tUj+pC`};q6^2E%{44c^8 zI$l?d!Hq7g8%H2JlUsMTHteJm#uf&f+uIY-x_8LP&XXzKkh|-8V7fdYfw?vk6^9X- zeT#WgZE`B?sf98h2R7`<1&90x{K8>~X9s*w-g z@Lvu5R|EeaG%%s|jJNYMg6!C@uP@Kf-0NoZ`n-Bt^Jnf0XlI4KcjXF)zSlDC9OIK3 ztRA1pa6jjyP3_!ZdL=cKO%f&>@IvgIx=Cm1vevxsq zma+k8e`=B84>nXH!&xy=C$BANZ)p9@Wg_A{|hk&B=x9{1$<6ixEYepubyi=0&YnL_~@6N&P>MoD(UyWn)n;i_7H>7FK zeVesV?Yo@nw}2%|j^pyZST6_l%NRny<6W9OJ2|&lE=QEx9D?{KP9iN2R13LZ$FLL#E$;XA6<3#bR953wXeX zsa@R=W-8RQX>N*^cf9CW5y+JyxZF{FRn?=kD4e_m_2%T|AsAd%l@&3iMrj&zP*goP zQx}^>21&wEs0)iNka6$-b8K)`vm$0Ols~3JSWW3lYnL%p=L6TC}uKtw@BW))xv`0R%~;LzrRu$Y7!> zMQ};k&e+R8VP=RFh3Z2vQA?`hnyN59{vcVKcdk!Ekh%{U?^osl&CREBr!i<@$za2< zZ%=;0D7a-+(MM_1%&V{;vr=$1Us)(9d`4ift6aTjfEumP8QzD}7$9O#;iCGgL>TPO zJ)oKH9o$Q_&*$ zl_P&X%@zV-w}ogQJ&xkI)B<6QP@n^9DuQ2-vW2`jSp;{l{(_~t{SyWQ-x;SH!D}xc zG>AmxrGs~hAgEDhR0sb!@BsbCuMZBn-+k_aK%NPbg|c&5k?hS5BJT_&zhDbl_69;m z03pHmUD8Q*9~&V`dMQQTS8R|Rc%3FtwnNo(WL@ub5k?O6@CrENiQ`^>ANDXt->2p2 z)!;d5k&D8uuU=57;2;uIz9Vszz_VQv7+i}N#yBUT!}Uy)Hu43-TI9i94~FRCsE@nM z7+6sIj-r^;n-0f^x{3^Av>q1Q$p@e9RL=T5 z-|JX=$LfyPSWN~I#gZkNEu=Gr6ye+Sg%=qcn*Yoiwj8_Rm3}hIGx8-9E;Yu)BMl*| zd*FClU7YYY-@yxV$7o&dsB6(WWF8nXD?f_KYkB2&c;GN(+5chk;Q@$UHSm{MVE<&8 zyGjU2<=gLy9G=7Br|Caw)1p$2){odYpp}8ORn<@fN2vxzL6(*t_7ol!aoKMC5V$$# zVHmw;7s(+QPPbD4AMmsN37e!2&rv;~uEl-;!q0sG1{2{s7x@0A!u} z7h!emDV*nMnz_%c=(Z9mPC4cr~Rila)7vX z71%sUbdrCN%n*bIGKltR_4csSd(lSdpw#~=I*1HMC{Zam_Ov}4CQ0>M?t>W2lXtSJ z$AcdnDoHtXsYUy!0s=|OK}Eop>qQUCszx5d)%d6a>jp`99&_L zZ08f8twbKm7($ZTmKZDaV zg3zM6do+@>tMpyAgFu-7Fe+gD;8H-ykm3T8>_#n882NX}?_EB#8_PhnIsiVf1a9R& zzHtdlH3XtZ*`Ylp?MApg?~zqi1{co=l3Lx7rT<-Y5aF1C2Sg+G@*_#AKjbDk!$AnY z2rzDv?Lol>p@4515AP+CeQV*EjRlty0V3Yt@s?yy|1SCG2V5ZfC&?KBqXX{ncij7@ z`yGVfkqlKpQEo?pI8McIUYXL(CJ8pn`*XclUNPZ_s?j@}O+(1>otTp#LjLlm_d7K2 z;*H7JgO6N1G2B%A(-mY{cHp6VB7wKT>>y$+Wu~9*r3}iHh&F6eJ9MsxGzeJXzvryG z5YzR&oknVHtFkj=1l2w=u}wO5(6mWimnwl zSVrCrV^r2hb5*=#N(nZW3r$ga4&DJV zupLoNgRss(eLN*+p{^C)ba2cR#FAYb5?MG4O0Q2n*x32dl2oy*Vj@w>--WKLSl87B z%{IVKUs+u{KSZM_-#EbUy6AWAdgk<^^LSU3L8XUD^ALoFuCMr-^ntS&$sNLKLqoiN zsCCp=tY=HY31}`i6+}JvJ-K>3%vf$WvXTz8pOQhU(=LS%#Pux(1kOJW_4DZn)wDjH zR%Y0Fh-m25; zF4xTiz(>iCKQ*o5zTA35U5n86=K<|Vl_PkU9y8IZpMm_X6xSnH8-fEb%#VNAVBq|@ zqIAn!(nC7IsypHa#nqwWZ9_MPVJp7ym&jt}rg2t8DyIsQqL&g684T7X_fP)t}UoRh0@8r(yJ^qB^Be!v}XxpHF4 zdS(a&%j69gq%_0oc#)jDjyF91&Lmq=muf zC#&kGo3Wea60;L5_D!P~?`i;>WhUFtuj;I%+UZwf(#S`;C65#>k~?;WEmG&cKyrt} z?qs2JUHZPpvEV#i-@!Q$ctk*031G*=%!H0Taf)5ZX-V{~wp2v>oyW=RcFh0Kb*x%atk;ii zm(Xr8E4B;Q3p}8vWgX4B0YhDr5%uym#p$EX)pNNyAVip`!JzR(@qAM{5MGZp9a%zv3bgDF52K*|={tJp?QA7!~3LQLrFkxZ_^%e@Aw?^CuAg+`qXgjI>W zg*If4h^v9vU#86{+*dq4JF}`1VRn&CI}7MKfDF_bzqCCJ%lJv}8Vw3gYEp(JS%~Dp zey??!LuX%)MHwUnFpe>qNx-a~Jp4N;z2-7X@5;&Wa8*EWbw=HugT)UnTH`D447yzx z$WaVgrfE^;9IDoY0!yJHNv=`h0Bhh8kW;0LbwN;qEU{a72TU5E`FdSHAB}m4v?+rL z0P~h|Fk)dW`+mUKvxNWBH|5;Qslwa0?I3z*ZQgKf+|rJ1H}ED$nY|YA1W~>b#JJ?o z2$FS!Bx6R;xSwbFw4v=@J9LU;TM@?Ml@^NTW{sjyErrvy<~?!iSG}Cx%s1QnjfPa( zIh354ql1Te-H$qf!acT*_IPb}6;ya79NfuJGSH!2Shs7}#zu==d~yN?Thv8JfZbPlx29P#eE~Jp(ka@-4TlL69oU-?K~W z_GmcFzzRE~Yt3u1qbw70Cdc-!U`$u@SIMM+@V5mm#%|fj!}bhy)_JTZ-k?oR^M^p( zlib6Qt&ecD#<fbPO`YYd2e%%&=*%3fCwDv*!VYrQUC19* zLD}z&uMv)95Ey(Sc|pr8B-kr!#?JIO%*@BOH&nC~YUs$;FTrEFu6be$vG?Q&u~$g( z@0b`}=kA$2C#@oiwC=kR_dfvBc~3H(ze_#PYNe2>*Z}rEs22Mh`{4Ou$9Znko5x|^ z7kisaO)VC7%vTr2bBv;_-l&SvFE3UKDdI|VGB}|QSRI#_2C&d$FqA+-h?=rKxKvuQ zUqS1$tXmnmK@zEwBULHOCGZNm=Pw_si5NAvKZm!8xAR|rN8Gte|6}<$Kb!*9&u$@xGW$?-o6`NE#ob69Xzqop z-36P;T?S^|D%7p6rx|`4Qr%Cm&lhJ!usM~gWOn60zL{P0b)sD_xkuWaD?XvVO%vF+ zE6L?awE;a>=&1=tYG4|q(k!hD+jrsHsByU~SO3V=RgUkOJfyh)ExYZ|bYT&l_m9vR zYC}=KkQlsDXT53syoI|hWH2H(kI(*KoEEo~@G^Tk(6TY*FSl1gfEHE$DyYSF3naQw z*+Kjf>uYwGlNkmIg9v5?i7B=}H+t?H8FJsfS9a+~!V%z3h<5{s)Y$bx-v#(-H$U7Q%hsGuNBTMnasQj>q|lhx8?+lG5m7H>qfk^?BE3^J^xdtbET2bWSG1& zK4zm`CXVJdZ_?@5Q1PbpRX8kR`Bs9$S&qU`A*o5v59(6DVDzLu_CFX+54O?yRRBH+ zZQ#FZ+2gSG;7VjGt_H(b3^!5tdpy=-0wrq5KiPj#&4P z0?>+66MuIW8pX)>rc67GZ=0b_YzEpv3rDSPt2=~Fw-`5l7%ah_UFLm66X@`sue~&k zkZ;NBJmm{#hdMY>_6i$@)`yYG;y3$_#i8DQ{SpcJA@vX@<}26;o63jKg?GJngG(EV zoj8z}p2I6$JRC2dWAfG~zTo-DlvjatJt=s;c#L8ug}Zlr9Ohi|gaE@t#gc4@8AO*G z>rLtW$fCZ9$CpuSiJ8%RmXx+{Nz|y6%B7}eZx=^&YjsT7PS+~6mO@Y=pTsJ&bO|nQ z(CU6~j`SQhAqZvGaCGezNd2MWWU!1g^br#qP3ggGk3bg#r*{Q`Xg7HA89=qp+%j&h zp%z}jV%F3XpYl8fG3_21AMKrI0Sddw2eDet=ocIo(CbT$QF$v^uVq(DLO*hc(?Z+E z0RBKvLMwJr6)n)7$${?Eg4FTf7`xQ6TxD)P>=%H1@8|2T1RQdVly;4Wn~v_nR}{M` zmsEwID&X#3MgYMXhDDiF`QHT}#s2ZpmS2?+skZo4)BS%abOZVgaf9#QkN!bx#0K3`of9xUw+_n#KRK+6Q zl=MWW7iWPY1BYQJ^cGG@%cP}${g0t)H+#8OTY7q}d_plhudfKJdA~ITNTv3R-}*c? z7u%C|?KY^7CMBRxGAr!2c9F3lBt!CN+v>b$a8nDxV|UGO+PpJLFsLZIhLLo@z2|%k z9|U)pemI{VgciD*s<`$Vd2ZU^7Kv1a)F=QEb=yH+fVLk4fGtwfVy1wQP;^*d9DAH0 zuUgzAtTvD-PSVW3Kv9S58z%l}nazkW_ZG(`7weF}F6?swL_1Tci#<1jO5tD3T!TW8 z;*Y*6t}lV;PxHGe_#^rsy>zLCP9$2;M~$R4s|%`yX}JMNam!l8wM|3{#UFq42ZCT+ zh26Zx8t0!31LaVGttdc2IR6b)0x@(C>Oq0KZs&a~#-jQG{c!beo+Lw;&M`aEA6?=8 zClYmVY~h6CZkbCuJ7yLZCD*QDbp0uvPO~HduULwlhFy$=TCduD@?374?+V8_uq$WNZ{VVu`2+G*>^H} zeRO00Wrvwu-I+4@*VZ$TcfWC!-@<}q{___A4!qL0(ij)s)ctA2O0Kmv!Vh3he*k*_ zu1{Y&XnUbztg0B_tzOZ-#hJ%}(r|Y=Qg`uhcoV}j)+zv~?P6}10cI9XIZ-?OEnY_d zl(}d{(MQHJ>|T%t_jzv6d)~c5(nF0zs%S=t{fOT+RFA#t9`Z$}kA4?VZ~YGRE@Fus z%i3W~W;YS|3M7-G;eFSqhG8#g(t){T7!Ng>lQ4rUQoSxtx3xhXW*)OK zR}kf_H1Hzwn-54~T)gH;C@~$x$@9PBBxyi(yzDhHz4Z&WQD@UzC0pIrVJhLc)d~9f zZsHsY(RatE(F~h?cXsWn)KFI!Klj@^`SBXkQzz1q;&zU&E#a&&1Dj~F{tiuL3Gd4! z^nLmqlMcy$a0mWKD;0aguz>rZeX%)(P>v4e{Alcez_kv)3SN*9;q{SDh(~vdWd+Bi z_K23yvH79{ki~eDD&$$2*WYO>O?HLflntGxlU1c=`QyR+`3e&NryaP)98)qBCk6F! zutko4nM*ddD6lN%+9{pTSfWp_3rzaZI!BaB?W~zQ$ty)U`MBGt0*;uCs+=CnL5l;No9quPMXf@yEfFuzaa5sjH27iGFReZ_VgwuVR4*D z<>Ij*eZ<>gLsCll`Fw#_ySHk*To+N;Aryazfey~;MN?Q7rL%$R6eqz($ zgdO{C6=^(doRidih-xsH-y!leY|_ubItw!FGVsCTx?8@UnCRA>k`>;$$ZhsdaQm3j zq!b_L1k>VoDeeB%7yOe|X_8Z!Vg00fz#;7l>rD6yIHL>+w+BcpA|<7!X*&o2@1EWO&d>cEt#{IolRj`zh+ zZqkmOvG&WeX$)m*p@p~My12|vThF$f3)*`}JJVlldmjM9_Xgv%^nTuEJXNzdg8|d6 z7obdlpN0$|N9ESuu{k8Ats_1uB_LevA+@Ek_0B~;g%nr$DwHb}LbgS`v(?=cJNyje zwaSa&2_;+Mj|ohpb|p5ZQ}6v3x_3`@?t{T!r0LEZ^>}sws*B6*0#sI+X3L&7k)n*Op2 zIp5#KFP2wff}n#Cq^t|EA7+HUrgY>^lZ#h#E?78yQmK>XSMJMRKjByjlkc+0$Y~w7 z*WEqKyEAw11F%B-)E|%kHP22o!!+guTL_2_&26TRxn0o%CiC`2Yh6Tu$jIpA;1Auf z(j9G@KAu{^Eyg`76A>7|#DT)8o0ML?a|{M!4vz^&WmJH{ zfWvM#;d2m_lDkv5G!}?~ymYZ0k9*8n<<%y7Ok-y~xl6tP(;{ob-16%mTpx<+GM93l z>#tRw$=Z2uzl_fmQu8FRNl9DSLAEYm@|m8LvP<{?-JOT|vyQix0~8~hw?Ey{6br^Q zfib&7-W^at?cUHmS@d-MHzJZIP!S)c*YPbSKU6ka2{IVmD^{#Fj&)e_f{4{x8EE7J zXoZ$HOL{^gd>iG~bD?wZ`OIS)(oz8ad$Gv3-|GbVBG)sG^6IL&?I1}gPz-Ea$pzB{ za^g!(+Iutz29(itjq-&>q1Prhm7#1F6=R%&rSQWEcb26NO7(KJyWHVC5|d|rM0?M; zZa8Php2lEq-7j*0lQ3g2lLh&NE#%Y$z%^!k^-x7hF_N*Tp~91}A-{CNlrlM6J9l|Y zy{Siv2|rc{dLtJ@`?1MOUc3**nsD4X(}>>8E(@M+nf~fdJtAy!<>=R5bU(*7rkp~1 zFM09AzMhu^#Wogn35UksKOs#?)6!*_FI_&kl}`)Rx2p|^$*Yy;0f+kx4hIZ;^~6+4 z_=Q1!FNmooy2I36dyPlv%3J%#o6eeUKdjcevffV)VjCGfMgueDN0iT=`$@Tbv;zu` z*>T4(m{+8uAtY&ocO&|;f07QAP!LmP->cRWU!WSeoLky%V@e9U28=CEWeU~r6h^YgLF$9Y!job?&-36bH8Yy00-t1{} z;g$a3u+HS^RFA#ZmS#CK3^NNKj^kLxUvv+jvlD(>`^CwyYg{EO{IZ0 z!3U|z6aBT_GDpFU@J0G&eS^;%9Vc_NWweFoY-PP_?u_vF0tNf55_E!Uja`GXyyDD9 zQ>9B``E++(R7#U4^cjHPXk4w+niqoA-+({V+bui7cM6Vp7*bg+(^{km-lUnz3QG!< z8`Y93f>&;A|6Gl3&zd{18smsQIkg+fZj;m!323=3tyz=A{;+ z+vbxuIH&GNp_Q!WTIy4fcy&VHIs!Lt-$!?RQj5_IT7g5`7tqRV+O}J0x^d&Pu8N!r zr0sOh>ft_eFKEF$F@;KgN14DBt)XQ|v*eHa#@8TYXYs9qPe=7fc43}Gl||nqp>~oR z8rOi0V2C&owcYWSAixb+Fe>3ed#!5POjA1bn=pk#00UNieV_U!ZLlrM%t734SitIT zE|prda^-l89;L2UeFGEW2HIu|>V{A5rEkf=U>j}1Xe**jgX?;|GFjf!+a`CP$ec4| zKeC!)bN>c9-x_;wSS3omT`l8%kW4i2qb;W ({dg^!Z!RZOEowR-Kj>8$#y7urqe zn{30V7s<*;9XKAdIbwISdyBEbjd5}aZD1V3G>NACO(p_cvgHNVj2Jj5V=u z!+FO0!IB8CEP4WlD?Hk@yhGN{=1i!XhK?DqvCM@Hj8>L(j&lTee{LYoYR9lt$E`+( zrnzA>`nys%%%at`IIywS3<;MW6geffJjs+DTcYKEzJ<+XE={^8K9=E743BxVLm3{f zA4R8~{)m)L-}@%ObU~N4oHiMc=$yLO_UioWUB^?pKXt^>4q{ZFvy418IO|{)5v9G{ z9x(;swuwijh+S~8Xdj10>8mzJ%dF8J{T6R~TRRuylC?-xXqvhK4}=@thN7``s{$Wz zY<=I*bqY{))|nGjj_)@H;O+kH1{|Yqpn8%jhQBA9DxDfp;t(il?qL%_vFPmLJ`VSW z3=WMShCp8ZLtpAzTDi_pEpj4>?^T-6o%t6M-wUn3c*t3IkBAL=6Vz&tZt6h2Hk}Qy&*3t4cpq57hw9}% zIv-bkm0O*;iwicLXFDx6Z5<}Oh*P~EAq!Nwd8u`!sGDtXmv4MUG2Mu#wOHn*DgLxz z$S$lCV&Hhm`54C1^U}xn(lUZ*BwjH!`l@EcTL;OH(naI-+5Q!QVKZ+jrymx6Is#x( z|68nahm&=m70- znWM`6s9>J@pXLfTcx|w1h5jcJIHhSu%oeF@Z`rz3b#6Yi6<-ZVs95G03W)ZOGZmxr z_nd8>F7)28%D#PqYZY6@AG@6#Q+>D>HQu?g(%@oS{oHe__j~`d?c-bOODAj}P9L{y zG&5J5gc19hUv1QeCb&y_o^#e1p(fB?Icd=R!$5HtpG8n{+AQ#EL4nChvQ-V7xfzd` z(poOFQ*S!|coVmXjSgjvC_GwBIX$%C2J&wg37%aZdPF?2bTl+|oW^i_GJ^#=x2Wpr z6Ov~s`uGE6H#WcYvpdBG8h%@gG6xu~EIMs33oN909V&T|z6>>G<@IS(P-2ioM_ZSa zZ&l*Ft+Ni!_f-m_U3sTakMuW`tKh6#)*V}hB}iErmoMWvK3$BW4F>{{-A9t@#~9;P zB=nobz9rmT(ZnoCiGjh=n|PB3>>X_f5 zXh+6ti!S;+_xs!*5N$1}b2Nkf+=M6{!S#0Xxs8#+V?PcX2{(ip%f-_gxz9i<(Qy^k z%bt3Sl`G?foxaiLG{v>>Vsm-Z_JHWJEOBLz(@=*hEtGru_S*cCjQ57IHmuy(I%55G z*x07VSo!djsD|+1rBbJNN(Sz>6}@N1zjAkH6|4A^R217~6j$iJP80tifuhDe=^97J zl^Cxj?-o__)^`;zK@^ z9X%@htZLXiS2q46?qg9jr+)I6FVWUSyq;RfY%!5$aC%k9DI$M%$6fmwyx8mD7WpH_2ydWyI+>>&k0eU-bD-Pjwt}X-b8z)h?(dH2zV5c9-Kl65tnQx9i@uH_`DfaXuh%z>rq}yhA#=k1u_$(ti@w!+W#0*5REm1M=rBZi(ynH=H$C?>?foi6>TRZFIWNfmOl9B4*!=-#%XP@qtvk*huCv?YGdE%y@ zAdGU1ZUuQkD0>2>)yzm)tot!0s%;kFQoU8AD)q^3H^dl1+<;b{MjI?(2)5!QFB|eMsMT`~U=!Mo)q;6!x@ii;+3# zCFHRk9UeXy7{I`FIS-vxGMt={mmXIy>DIE{<6il_unw`ri{b)IA~(LKOMelIJ<3o& zIZ?Rr6BBp-@`CgW;D_OzL+(9%KGq9DK4I08x9J@oQsg?9$4)wT3q)3#T3Iz)cfR`U zT0-@>dxmA8*Rh>OFp~O#flOLk6*VA}u2{S{!E+Grwi~+I->3_cSfI@R6mt%CO@4uA z{QNOoe%Xn(4LCH`R@k^HM`d92gGvbf#$`dXACe3A&qcCov?p8_XzRx=S-IAQ3a#=l zpm54UWCq*g+ixB6^qUiOw(%3NZ{L@!{pm4Flm!Kxb`U8X{d{ac-yrH-#jIIk^c?Sx zYk9T1l9kBZgcNL{`i-SCLp6HQy|7*R*GcUnaC zN()|dt`~OvP)IALrlN9YcJz+ZsjVx84c~^J%)w?mGK=Ss+*Yojb*h$M3E*>)_?GNh zUeuGIr?0xKWcoVxD=n<(P1z+WJE`6t+Bxm*zn%q-T`5aa7IzV~D}Ai4u2pkxoi-TPE<^JM){gkv}#@n(Uu9A+EdfB#)p`4p;@ zqyK!uF#Duf@s@ARrsY()_k&;u2<;N>3F=My%Q>#>DH66OrlQ>6{fy;g>evrUk#k9V z6VPIVI3ldTKy4*FITwjy~Oo;X${cd&^fE)wdsd%Wx1x?cJ<{t zh#$f5ePo!e1Z+z$yWjlI`rLi8l&C)X+sFWBN%&EQWW`cNO3drDX=hQFQxBHB&TTO( ziWO6{LQMVrWX24F9p8n>>TyB|HEP8p1~f0k=!B22hZ7H6uYKJ}?o?o;CNOr^$J&mS zHiD1UYw)jLF;`T)WBE?ljMYrQD$h_Iul0)8VJV?f`g+}*7@r>g?}G!XCo9_z?gRwi z`{_n+UldF^v4o|%QiIjMB_z1`5a2VPUrKO-#R5ja-~?Pv3oGB2PCCKjn`%n zJ?~S?r^lJuo+daWt5wT>j1Emi@&q&DV?~W#?5c!>EEOI^^*q{Q9b2+W|p#z|gUj6Sg{x?mv8a zyT_=NjpOHim|79^FdNp&6QA5{!+S(KOG{FA?$(%H^h<5f)#H)rS44LC`lS695DEVR zAqUK#NEt?``>kJ6hYO0)o}=%}v;yL#bRYOkrCVCeK`5ytRv8L91R8g%wl z@HfBjK5OQQo+Zy47Zsj`ZA-2{l5Y`wYE-dMFv^TRUQF0os-_9`iJ>&g-f~BKC0|PB zo4t7SXvx$k*Ug25jwyydmyTpid|g?qI#kQ$4fLs4m_r?r87f>aJbDb3axCcI=A@*B zkn1V72~MX@FY9-iX02EqpemhCKnbuuK1+b5Og%;x8rc+-oAx&9OZ-@oR2z6Uuti3M zId__U)&n-<{@Z5$UKJn_T1~ciB0u^GEyXr+PFebT&jEg!*k9}@vy0zsk>F|te_!oB zJ%XM~5a|S{$=3(y@Ahg2sWqD(TP2>=R}DV!KeGS9D$Qwn*Usn2_rXC(ac=49amGsH z2-1K{J=ayf{j$jHR_rb!QBG-DHv=a9KluB^8R$Q@?xFL{`6042!D(3R=JL_PqJ?2v z&VitXv+GoN>P3gRyH0^L$zM&y+$k!p?n1h4?uZsAPpQ&$S- zsTqB;gtZh|RFX*zi(yoXk)FueGIS!oHEoEx2s@!V+;qXit2^hhQA?$l<5^PE^!MsS zoA)?|pDs$|;w!6R68m9NaJ>RN?@1B7r{_ad=4wWJsQ$eI5slpYP+^y}?h`S4n9f!J z)9y&6vO^*jm6Zp?vECIY+@~&77J7k0MaB8i*ZWp@sNvHTdGzg#oE|VyEJt#GUcXz{ zz2_Vo`%4gj(YFT_?hc+Br7kM6GnsTZl)>f8Jg+*rSKn=^!7^9OzC_~h5w0KG6QG)A z=Z7I6dbT9H+%bT<0Z?gl7gX9NBMp&^M-G%+f#x0s!Hh^NT+=2ad)7$UNpx>Ib8baL z29lP3%J(-4$_EBqJcrJ3g^2HS0f0fCVnd`NiY>d_U37{!XI}DTb&TED78}k!C_FZJ zeBnw7srEJ7#3R}=0A?FVvfE#qHeelUcgk}B(tzmpJrC37p2r*);;SPSal+e<>gFWA z;Z4$1_Fn+_Lzpr`cgCB*DMy{36;Q$Ace{H8-x5ynWCBuB#*sgU7r({~N4=iLM;jBm z^^p3k$VHQWl(y5QzUgH2Sk*(O+{R#{7%2<|$*}y7K>5W)Gik#0VN2<-=25q#j2(mG z(xSUgP|F?IdVPN_6qR@?(OQUaqUs@y@B66TI%-6XGExUHGqm@&Qqo9FqhC~<#1vPU zuC`k(01vPeil!>UXLmdlZpB1leC1yQSXF z5OYHG_+!G`Lq5YOd8B@~I@IP#^t+9kChRB21yss&9AUM1>;Ms_!+1bIl!in#q28ZQ zx2Ou`w!B&{qZJzUEiXLm*0p(^XsNSKCM7%sGeh#^>kt_H;b&P(&%D`oW4Q3>FqrEt zqf~!yTk3X;vbOr{T$ zCS5Qi>4NtcN=Q2ZRf$M^Z&dTNbY}3U#JnMxPQS5ZZE{Kh*c{ZNf<=*-n%DN(@P80o zx^Xairr|deMT3`lmD{k6+jd%7p>QTUNKBx5E{P{;_nWcsH&3$rFk}RGn+K}XqngtI zy|2AMzN9JK(FV=^NXhnz#Or_bn@<6@68~O4(Uy_zoFax*z}dyu9_Rp6P@eo>sjj}= z74V)f^|!&sXg#9=8t?{ql|=>e&I6@K?$cxdclA$h;uZ?5UqR?2%t&Vx**_y`EA9xu zKS_jQ6Aj{N4w{6Y5#AR3UiK>4`PBxO9=e;wE#{cviZT#`MyLE7Ea`j-d&KcDT@ zdMTWjwa;N~9-BZlj28NkWj2D8Vq^9mCekr}gn|7&tOqzYc->uu2nt-}0gznx_V&~V zxI$$sA9hCJuN10;#tm*Z5fPK9OqS@K)WjWaQKJquC2kZh1TseyE(u3%yQ<|}X<(gn z?BTp!T0>(YCbjStO9Oj7Jju$#@r`2$2KxwrCQ!&^BSetkJs*A>4DLsWtkHtU_jhPh zz^Blq5rrn_8^|riZNEb^KK~edFMniY$@k&I!jCtuMaHX6;ZL|{(242uiK^{vx_Fgl z7QuNgp*RRIn1FhZ4xlZr?I(N=b%(o(2qU-Gi}6ux8-+)jp1~Dgmy5xu;A1td1oxDC z)`m=lYPFF{V=s!tbt;sa?H59dDhfdvGx>-dvaziv7bEjD%mqOsd*hDAGjNZARDasx zJ4GXO0Dt^iigi)h_@vDV>t(izv&Nm4-S!&WF0Dxu{F>o(iZ3R1c@$5;J5vu|w-#S( z8sn@eHqN+w74Rq|REp8#g?W5qkEe5{c6gf-cj5;>eqYrACC?qOFT1mxsG3m#*rNIF z4*71tBr_cM&Gv9T6ZX`^E|ubq{csv*bsJ%N>hNj>mBi@Rcw&7ftw4X`l~-fbIU8MO zQV$CVT0Zp&$SlZM!JXRjQf=us;X(bSQxgfH-gD~!K|ed=tWv@X)xrYu8>5$%c4RW? zrgW1>^(_oR6yag|ZR4QuVnwCAgX@JG4<{TwSAnd*t$SUy|8v7kzQSfx2WKy|)ahaS z+=REL_n7X2pfNV0SCSKkik5F;6F8x}krKPuZtH1tcu&O^j-NATlsg;LB(GBF+1Fbo&_ZY$bQ24E<|X^0 zulRv==iFy-ufu9F(g+Yj$WMp0Q;{YMhbey9wF(0FBt$lva_C_~`WsE@dx?^TiL|gl z7xBm$+Z}6uxf_5fVJ&=q@NNEw=Be9E_O^bQv~Y3mH*eNKL=oY`eG+;dov#PAF<@ae@OUQiAw{xno1@{8 zWeb+^FgMe%5z0E|NbUv0ah%C_FjZ>DMko^MY_@_RVX zjOq_o7WY%YqPEYQi=x69LF9Y;{x#n{a71obfe`Xf zY^8LXb!)iH$KYqxj?fOgL2*x7<38duZ#re1g%Zjn(9!62*SplHCk``~`L%eMwEJL= zX@h{IlX-#7Kna=x)-eG>Q8|Qt&5}e~CEpLq;J!3io(ZTIUIUdcT(rR%(>VHUP2M_` z?tu76lMuuB5_I9=r%%L2{@PUT!iHEBPOk7{@abGkaay@x5)LLkZ<1X$UzODCvK5tj zyiGr5+`z?r@bh_$XMg@=>x#q5XWX++BiDtOMT1oX^h=u!#%`n7@3YeaYGwD8Usk~R zpU=0Zk)kZc?lVOY~!51^Vgkm{RUIP!R{5OM>pk_az{E)yl0=ogHsJ03}@UYXPH0*wI$nw^v@EW9 zO-_$eS-X%0B4bBsxjcXVQtevRg7a%Zp>&0Gngq5ejrP!2N_i zTJhR)t>lcxRJu3OJn+VGIdo%^+0GW9)EPlNXCtcCDt)x1$L)`9)2$M0go@eYTSXGM+;T0HMAX(tGnfrfWY5{M*8DE=wQovP~RuWRk+P^vS2 z8WOPE&2{rXwpLN@N#8CI!cTjNa>7F5e0KUBl{{NrGx8QAuV&5V74r3`CQFB*a7e+5 zwNPmJGZ&-YXl=FJEN}S5`s?bswvgVQwnV25#i6f{#s0FM0J^cfPSwA`Y>rz`LPbfw zx2bgAU<8@GHi;G=s_u{tZ!`)?26IL&XL6?rVk>rilBDV%= zye;!l7%vRx(HDX=eXaNCIT~BRv9s2x_E}%L;#)J#L|=vpR2gAx*rw0$%)c@u7Jy{RGUt!c+H+Si;-KM6Yf@a8~j^?o#ED0N)ePffn+i=iN!{s-8UAf! zT57Mga?~0UpH1u6e_qL!iYs)|lK|>EF1+YuNH;k9EGIJU(sSI;8Kd&;s?41tf+J5o zxzEP;OdpRS$B9ihVyhyXQw%4Y*Mg0p{$g~|5>aM*!953C?9)8BUU;^eXQ%{6zEs+e zH}aHXI5;rmu8Tum;21^DYHCJL84$nA1kQ~Jcuoz@T6!0E_tI8=m$?!4+~Bo8AryV# zhNLDB*vP2+N|_6i3nL?mYeNE`LHvJGv#GBtIKpX>l33`Lz4T(>Ieh)T+s|#KmWu2K zuQfyR3rW7;PD9(9d$^}Ib86>A$0pnLOyk#ZGVY$=H4S}qyM9WyC*~ zT-G37D@Uig5!If%&i;JNhD|bBz4W!eBsDO=pVdx}tH&uDHVE@ZPPm9z;bJZdnzY%LKc_4vj-j8ab^m=zHW_v#CP0 zDqf0Rj=5u^?_}KlaA!@gH?a}tyrC=iJbtk%C`xdRrdFmfFpKNl+O0=l&l+{g&k_`p zEZ4%tw8u^|z(*Ak&8~`zkK@%$t8-HCyiV^La<{g64i6E0`DkldpvYnEg4Fh*v&^_} zUZsVpUEjB+1X=7CZ;Ds9zf@v$d6&Of#td3BU5H82VN&&bFyirkf1&7fs3dTr;TvUb z_>w_gj}9h62@)z=I5;fLdV;OeaXA}zWT`A@m+2#Ra7#gaO#Zp37eDX~Y3K1cLD!MF z^gN_lb?6cMfrtLd;&-K*4S`)Ds^ zq+%QP%wJ(+YD`*r-iSKCTHMm0-bj3N!{k}B#m%rUu|b#XYskvxzP>NByI)X8yKs=4 z&26bI>Dk7%oqO$T`|Z?6J0?>|nnUppYUy4^s-bWppH{C_&YaCW#K=RpR`MKADRKX) zcWymQxuJ2=EE|k+~Z|9I^C8v<0}QI%WN{IB|JPH(df=PQVMv~xI97&$-He# zAlgA3#QPYs0R$5yD`-+*$wDHmpZP_9BmMsZpTCqI2QZw`rG^R@zIPSy)Wdr;r~7#O z{9d8_95>WG4-(7C2R>Pi)fe;+_F}fU_S<45`sh`! z&>ARbpQmi0^kS-_VRX{Vde-7L^eK@`54fvgl0E%A;S}Yw9r{IcVR~YOjJr1AvrH&a0vuEC?!$AI}MwS*%b1S8R>!ou%&Ll_?oMzitf|25dD$^Q;Aaf|}Z*hW!2~ZFf zB8wmP@Ux|be26%O6BF%>)X?H+W9|ht-p%!d&BB{q1d}cUj*eKJS6`hi-Fa9(ptoz(Y_x|wekS7orvVeaVm{OxFu@?uwB8%#;cUJ3h`%+xcty(2K$yB z&h#8>E*LxNq_1i2`S{k?3{y4n(-C7YQw5GZs+VcPI$l3E^0KuzC&QI%=8Ao0uT=M> z;_zm5DLgAME=`|XvF@VQ(*mPrBg|I2=hImfvwC}iv3$1wj-TeTfW@f?;ugvtfpNlW z(brD?2tDSm6V$DkQ2u#(OtCLH@W6uc_??%r@CtZbgmE1A|KaS<xsfX#0Nd`+0u9 zf4$6go!5CD=drwxr`lOwoQrer#7nim@8O5&yo!oWP%l^}I;jE!N&W)s)oc z`R3;6i$WJ)OCLS*mDffoidy4x5@h$TJKRY>UY`J;AnMd(?d(&Z&T=|`DDO)86gQ4) zkGD~rz9FBN{zF8jrFF=qOD!j%>$PQbMxYeA$-YTvwYfEkZ{v;xL|8;gICo;03VZlfPTwz$T!tq(8VHDMbal;Vo%Qef4_8WF^K#_>FwCQQArHeq zAalArD4Bd4x@&Ub#%X<(vc}|z)Vd?}TaBD~`fA6=d_;#u3TguT z_1|@~r#A)E#iTQT9nx1B-5ZeFan8M{56^q4KdM$}!yQob~5Gs-p)!#F!cowGSEvZAOH-Y++=<)pva`qzpD; zFk;#(0p+grK>Y``$!wjQ%oA&CGZ;~eeG@ws&u{t8&Oj5-KdvV+^+}60(lSu$3Rl6I z)K9$AS1)LdM~V1hM}F)Zy;P*70D<&KPS=_BStjN8qR4xkYCoA24Wu>eX{EIHXP+G4 zdQ)XTp44VN_HnM%+7BBZv*F}agTOX>u+rm%#FuVaUD=&OHym?fwEMJ6O17{r2RVxd zyzLVmYO)YZ>c`N=I3%#qj;Q9zIgM=@e4XB4GMV&PR48bQvFcF2%ei5EGRT$^_? z-j%d!Uf=i)n-IwWO^6#&@rXmn{AYjj>1BfH)P1Q@r}tz+nVZ;9^WKh zRb|!_Wx99_F1N?gh$psB+!Pkv2b$>ZW1S~aMI$^u_qpGdx;M{ya5gX9Ffs1tqO!PQ zKO$Bs|-9RycHLDWrnehQGL-SqnTs;uX;aTLnx2{w-ECO*t%o}L*z)(7xq$}unIE}O=crkE zSM?hvyT6pq7}>?U~H;Sy>_m&Ib=BaegMxNiy?f!CaU!(sYQXzp*-R+mn^|yyt{Ro%-ZNw zWNI7%5o~eb?D2aC?>Zv9Gb zMB9e}P8Y%VQaA~)&3%7hn}9Vr?_4W!^W&|kJ56*C9CUI_KXNHyV{bvet5J`Sdb*)~ z_>x1BwSXRxoQSrNBTlzEJ!bNFPQKTUA7*6tuToD-I?j|ezK9ai>uWO<^^zTYf6IR? zC0c5lA;qNUz-@=&^{x&q?e{pfJ7Elvj#EfavSJTo+Fbs?@s`W;hYcr}&(gOL_C&Dw z(o~p9gvTbb?GEl?^ADjY7$`e;G*kUzm z7_y_U9R?||Aa}L=UVCv?vpmlc9zdF|s8GPP>7AbG0|q7q{nLRbMDdx&)I@jjwCl87 zc0bPN=e1Kz`DJCgN`e33(e$NuHqC4I8Bm7?bi&@olPk9oU%G3I@c917@+7k* z?(Dw1TH1qHcIxt z(!wo>x4dp17+EWm8)mg<*lkKRCtA?kG_=U``OWo-gUudK%KGum`drz}$gmyR3CL+4 z@I^yeG9NCv*@TjS22{Yum;dDa_-B+NgN;s1)~!U0l!f;G_|4A?rF38SIx996+TT>~ zzq;_!jOF`#wE=wpapk3;9p8^WbTi{^n-~_7a>DC!_7xV1WK2!Or^Q5H=RBG5LmzZ%Tn8!CPEJ`^0p z_tdN9HjqvLXh9#vuE$GA6ddo;ly;Kq!A&+iFw}f2NZ`@9Hc&mR;H(`LmAv;d`*C}d z^})k$>cUTWeEQI=vS86!b;H(`yH~Tu!@v=>*lfzG^1@Qa!_XqyD|?S>#xuqKW8;{q zgd>tov2ZuyeuJYhRDe8!NeQOrQ9*a~nM zx1B}l#jE2YuBjg7mOkv(QZgMicm4FpVUHWl9zyy3-woY$3=ffp&@e^lZR_p){xa7C zh+^7A?|flbj?J!zV337C$dSh~ZPdWV7M2fSrq~kK7ho3+iobYmBA9-g6}}%tUD2#h zY>Vd}k&Q9zJDzd^Zb}2|Z#eRR6qUXsyD}uX95Zdp2FuXylK|%yNtGP_KFaS*csaPp z>)E@@;WC@4936Y}RlnTI5kL=c8<|#Y1+p05xcds=28I2a#elRCPYh#z2s=vK-Rjw{ zM;xrt<*Hw9F-X0HIpKj)vitd#GO%L}75yV+#J#jX#5s04H+P4k^hoPB-ZE=`KM{em z?@RYkclTaL5lp>48;xL*Nbu&Gf=0lZwJ{u@3-HN`Tb{G~;+x$>r4ir7`R$hsPUM?8 zeq&W@xi$T2IO_n7}vTYPj~+PbY$uwj31@U#1jeL*ylQyR~eTE zsl?v8`$LZWuMVQ^%YEFoaTV(r8n>MpD@JFGHbgUQlew|zFW8Hn{*kTlQa^7j#U;in zdSf=KG$yI(YG~1LQ}T^n{93;AVz}fvn23lBFo0%iWOmq{^;$42gLTV!Je`0iX#vv| ziX^!nLY9j-x7|-)w>5O!nrL`T*(-z?HmR=U(zA6WKV#Np@`KxXHZF{0s_y7%^RNkd zV%trnoh z%-oWLjo%1TI;VaeDAGq>dfJc(jlV_)NMcX{t)p?}bv)U>lm6O!NR0)^vOlmm&vZ8# z$&7LMWdBZ!X`IlX`6>5dJU3(@Cs#f7Tu%1I4z-OR*%r4QF^lFb5HGTJeMHlw%eL>Z z!&VkHNvY{%xAwHE@JAN{qo!@bX;B*c=!2WYs62eK38i)(JyV{3LOJjBi3K~)ooMHN zN({LKMJW|WaNpey@433cbw8~}0iA#yscXWKpvc#hr*_0K?kc%gm-ms}dm ztJA-$N($5MCXiLM^hWC#x|Tloi#<87aRw4PjU&iS^0)lP zO0Rb1Mtq|4VWE@E8?f6Su|(_+5~*vPyp1-jWOeQ(&E6`p8@NNxzC#^RjwJZVk`6?F z)t+EunWhM0@VsyWo0*Wzrox-Y+Ed-=rK-kPC z2S4m#cFVbS4!J6qIGN!P)MlqI2=;Nsic(3d-AOq|vs)Z<-tc2;h43i!Su3rBb}tXk z%&MVm$PV5M-siNE9(?hn&=?YNwgD&kB&`i|3)KXE<_MT{eJh!>Z%u;nu3 zr}bRh2pA#ME?&Rmd`#NHosf!u1!I)`^^FZxxIE0PPMGZ?Z5hQmZUA*WnRL4@n<{`6 z)gw3a#ca-2mulZ$QEt^wj{SGAT9RcybDFp1^fpd>5zsxL=#}D!nSJz9T4xYxmWn5D zw0rmEO96g^enA1Y^zMy;;Cb|kj0dqStuG(;y*}oT-PKtza&*dV`Hy|9}OA42QvRvKOoD_-QKk~Ej7w0ckMGtomIJ;-P4r3c) zpLeO-g6qHOIVX%``DRLm(i&8jtA9?mViFSXbZRAvYq%2|tMbGIMitr?Tx?PLH%>(o z^TVR8FP2$fOGT3X+awDky@ZjaGaB1jPMKA6pzTvdQAr$e6}x_jtB1tcvYm*!;M-?; z#3|CQ;py}(tfCDx?}z^zs$1}GTAzsAMIxM zIp^vQw<27!3K+;2;Hi;G=|GAWxsamCs*JJR&D4>CTRZIfLYq> z<9!(&`%)I_@u*R5v&IEws#8}q+j}jEiY(qInMEY3ellJr;lekIrglg~xKdJ>-KVVi zsqEgD(eNfoMp`wPS@TX|?cAy#VkHUN|6!ub!nt2WQn$h zFXeB$zmHFRQZp38ohZ+Nu~jKU^m4FbmvK%o=kuI*aU7^cpYg-j#KW&jb69x>%rCTt ze$zd9Ay1#IK)m@l)^zY8uZ22^F?hkgEo-rwt6-=@nQ20|==rUxhNW97Z^O2fn~v0E z#U?)1pMTQHD5TZmn&z(W64g*DtY>I z^2A|i1)iJb_Rfl#-JR1euKLp&KfNhOIqz+&cJ}$7Rs&eoHhh&#(6*Sn60SWD#XI^& zNg?7LGm_1(ht9tJ+&``;UU_b5e{8<1&tTIq=Fw*Jr3K%zu)qs>m0vJLt=Vy=Ig6~N zqj;gJt8N=k2)ULoCcHR2Ac(HF*jJg>pyOd;JXBLrK_ylc;oO}i6*I9M6c@KFQ^Lnj z&~86eWk~d>t*8hYjv4!r67h}oM#8bU9|M(5+k!aG?Tnmh^ClAyUue@b)XA42P8O@1 z9r!T>zP7g6V4hPga+U%H?L(h>7UV}>4&u&9V)Df&?({a+X&?2LmT_IQ{gCl4M1A<) z5;_Bq&ufhO;VyMU@}0p@_MYzpqt>pLlah3YZhP->y!FXs-*@S*({6f;p;FOVLmzo_ z7CVH>rUrUG{=D0|y^g5Lnz&H-c?ddQHi#M0rv_bOom$qocOP%my>0$!RE^;=|Pa0MX}tddak4v|M3W=X)C#1$N=Q3ikV$HHxBH_ zbQJFkYrkyva6imhaUUHBsH!qOa_6Sm7ChUuxgy1=xm!OEz|9uo&JvwOe}?7iowLn)DlBj9NuBkJo4;2dsVZO#P*1wq z+o;#$q{J>iy`kM>`sb~JT^O;A-)HK3R7k-c$+`jDFPhE{*xeJqEw5B>WnN?=BLOYG zC5~JpKefWVmVTTpqQ=llKbUBE@OiX`Y(aeYfX!F`H6)HWW1D8OW4~TM|FY+#KfeDo z6q{EA{9WvTg`-63b)02{wbHs^}BJE)|k+LcC0U{jPl!2{KE zslrX9$5^kXbDlin@7FtaNayje^NpMr>CF)YZ6VxjkKxkX(hqZKu|nFDMG87$1GUts zMYpHlokI+-RnPuFJC_SOkFqiTjKJg5z^~1Ns*lj_&6xB!|CFs-_9je*j*l#!ebt8^ zg{SMa>NkP-=p&QYy%X+Ww~5(ozm2uNs53x4p?Yn8Huc7hPrLLhx27L)`lcKbEOCVX zO{9O8gIm*_C4oFYe8~-~I47#gs1(02lTNevD4)7DG-i)k%QvyiZS)uINh7bPXkBnI z0x)_V4+4_?!m(N4Y|8v%c?0#5=~fZjAnW9U?+2hNJ(1h}vX;w^d5orK519s9X#(c- zRII#%VKDo{6Z#NhC-I(xx)mjQRK}p2pNO#=7RNm!I+d{DqnqW}%RI;M#$j!$s~ks^ zSk6zpW*EDaqrcFY?e+bEKDT7$(%0;}=fb|{YS*;gcDRl0^;+_D*?drZ55_awv4vwO z+HZM^s2|g@w_5z5mVN3>IfMS8PRsbm{=kA(UVwtB5c`LJ>W3jsv-&vnl0sf0i; zRLB(rA|MHbMAHpWu^Y;YsZ<}ybi2laTKsC2qv)c>%&08gSz(;tWR?3?@g_1I<9hE5 zJskT0^~mn|%4G6gte4Zq;?7xDdudrAgX?*-)J--Q4bCkkNUGZq$PQmpCxlCAyvrLN z%=a)vJ0FheD70uY^=|EdfxopNAqxdqvh=uSm;@A|;X$LNj`{#Y`eCYkf8JxU)VQZR z`j|EoO?Knn7%31r>aAyE&+Oe7x!k0urJy3R_&ytoNk1?5^674%74mIge248<`dsG9 zGR6_@sG6q}gG=5!pt`$2$kg*Pn(9hMWFAfbccp_26%m~c4LaFb?EP*j*6oZVot!8R zyZl4#qN!D4pKTq5{gM0cSRx$T$-!II_##R6=Cl+c@ho{aJ|Km4Jc6r<*+6IypbB6?In$K~E6XR{LxyCihr<+)pr11D9=aT`nRzXsuR`hKlCwZRd0TI?+OYVs7O`nlPMlb=i9M z^-YpMNaZWCu3s7Rx>XGwU-lH;i_jC7z6cq~kcolPs2Mon6%P zvCMj)X!FQa7iKB|`{>w6(E;5Xi@Q|yIwN~{jw!b+L+cT`>>j?k`km?ACEc=yOGQXm zBO6Y8pqAE#zV2G=vD7zl3Pj=~?WbYhN_!vWYZymh_wha_OIo)&guC0l7Rd3fInK$K z6my+($f6^LGAd(gm|?{-C(+&F>5cBGKJw;^ZM;kOD;MV%?sxnB{HJEmH3y_DwA#J6 z_f3ZjQkv}nMfKA4$R=v-DF~@KUgZe@N!<=Wb{Tio3u0iHGYX&;qzGMD4n|+hmBF}m zi$Qz^P-)Pd|Axdx>90N^nhxl<%*bZj3w`IlTYM z9u;EqEn(@i1KpjD1NgiODCridODDx6@cmuiTlx9Tz6#{akm}EXTMT zQ~6=vL)kE)Y4b422@dPlGiY_6YgAAI%vpN1yF8ftW?03eiNl$V#n|#!djbCv=1Jv) z0KT5Vo~0mHvE0Vp-k^_{qd6!paU|t(_gUvF@~+!xRfYL;?FOHxO+0Ak?8<%J;GdyB zEWc;(1HXeza>P37dHx7C@#FS3j^p2uqg9cK>vPYvE{}*T=8HVXIVIke-SbhXUEu@n z%z;KX-^QAd{%CXk3j@a<9H^pL$&$&sM?g+-VPtONJxVcVhR&sJQpB%iJLOn5PuBF{g}9_|1}tYzZj-)V>Cus&XVO3Q;`sPTwfXHB$#b`{(Ls~> zi&C=AHgaKU9-a?nX7HU#yS@d@jI{gpk1Q{|?WuiHkt)n%*t=k2{;ox)ApeR(;5a5l zJu+G(wW?vkXX=yV)aH$K6?RL{XQpMdt5TOaaJ>E%Mx_nV`7npsy&0xHHoR>YsxMIk zGnCr!0|OG1^_Gko*8mvGbPDE9Y$iRhg1Hmj7&2w>gk6HJBuDB8ca@+gi68vuiX^+` z)cZp+q3F#zHJ9IM2nuiazF_FyVMiA9H0Q4~(H4;*PaN^OnKQE#@r}d9Pi)yk!hYb1 z)_kgNXG}g%2r*sa!SuOl!eg>koy9}AOj>|k?Y!tqvC8hMW~dmFqf-9*`E z%UJRqu_WK8E+~ubQA#!P-e(?Rn-y$+NZj09HG2)6^D!ir%-MV>yy9X*FlXe*kB#DK zv%z&bH!8Fq{c$u9ce!r0|dD0P;VBYy*pfcP|fjAfxV7_Y`_jKvunmxK5 zVWK&wjHfP)7pQa!Kgm-CtGF0P_7_iJ(Y49z32o@2Ib-Z~q!{nOo+}R1bjEh0>qo5@ zjL)=cxIo|S(}5m=dDp6kU>TCJAPV4UVIPx~{ zi-)+|VFH($#8l8VVxoR|&`s@Sm`YG@SgM{*a13Wc0q9d>616S1SMX#{O$$8cq0Kw? zV3UClhTK-Q=^O>wo2;B*`ES?N^_J|xHx7M?$LF2UJ8-utQ!wJpz^hJ$>6@N9N%fZQ z>F;$TooDQ&x~Asnas643y;dfnN-3kEqJPOjzy(3Xv2?-jnJ!P@c<0h)25S|Y*rf~2 zJsPR7=C&8*Mf5wnp8d{j)%(^((=VejcckPxMVIZZ3+EzGBYGXjKKHU{O5G^?Y(r>@ z|GpSpBp^K;dUuiYX<8_I#;LE1v|juAw-!*79>l?f2SDU+!mv^9#5TNjez$+bH-A<> zF|=sj6OAQ~OinyNkx=1EIB;QLBFdB@S(x)1WXAN7*hmlk!4o5?$; zs=o5eZ1pT-iS)8P5YbrMP)V0sRAFn=d=FLD zhiHWLC|C|3+wVD}!Y6LT+!W;Z{RXEgkzzx+dy-B5Qsi*IG$FCl zcs+$o4Ad_)-PmqfUjM+rJhbi>-uycU5dChhAV&?rSG6rRWRgOGl zE=6g6l$~abGwExXjT%1B-yVu0!1Ponp!`BQl#nHXT`awYhnW>?4NBE(D?45MVaJzf zZ6~gpM7smNb`M9|_pr?mBzCv4MQ>NlxM}q?qQnC~@fe=x{5OvguD)T*flu~qNRuaV zm5AvD)tfpcMi8=9YWI#Hi`etn$K`UK^L`_JFHoT*=1@&Bz$R$Gb?EtW;Kp z=daD&g;D>{O4~PKV9dxGkx=rwR}&Z8s~Q!g`r3*st%!{Z4mQIJIWMJoYClC(9Z{XJ zmrx%!ka66EOK#CY(R0|9gIVRb;}z=~KeNZiHpty4;C$g;F}mo!qcY~E0xorQ|C zu-}v3RasKVuU=7LGS@MZCRqHQ)$3fKQ9`12eYGVJmfl6(qDC zs-kWsXboJSbiT z+CRS9`}+*mW+<=DLw>VDO?BT@pO_R1l@R$)n!A9mK;qx=;mDNu6_x~qpUDwruLPWc z$^Z9%+mee_;EV+-!wdfo!KPd;pjr!od>$fv2=&71I<&fd z+%}Tbq|CCer~NvFqG46akcF%gtX-EQz86~Rjag2IZ@;7itQ?2 zWirI8F|N*pi5?~TmDXwiEx5o$NgGte%s?WA2g8E*v> zs3`=OtLKgU9Jc3ioauUOtZBsAhSYT^pT^ikZv9DIc<;z240p|6f(D2Ha2j5X%}kZu zg&w=hxf({rfVYdz=1Y@vpbgU^Ic23fFCk=WX&C;w_WCk3AZ7;O@mYk9$?!5~s z(;S-e%oj$N*2}5L?urE#^1hTe%L^L`rnx|K`QGqatm)%80HJpDiEt%<(Ue7xk1V%O zhHYVk=b^vGq@^s2DZrFx@68=1mmm8gLPMFK4Q!M0#PIO&0dEysDjIYwo~p#M!2VLP zQEBO`y}N?VdZYGip|_Kvb<79HHErSKJDT1k+0R;{wJ=*~+{<&eYI(fM@X}1@vZl*x z>!!FJDGz844bR-oPJCZ?`atpA3k( zFDcF5!fUskP@$Mm&J2;}&wo27x3J;!R$}orrddLX4kf44s(U|=z2)5h5Ld)>Z2tyS zt@toOn%$+?$mg!Y%TzHT8jPa>vC27*VObGvuaWhzZ=&oWJ~@Q~0Z`nDRT0~+rhY~A zctk^7qFY(|-R6V#tUI2T-QJmZpZCa-WBb$G4OA4vMmFjD$MXg#0!wOl5;&>hxmRyv z=c+PmieVRoy_tHuhT+SpY2A9nIYELJfgRP}8QG0- zRTPbb#`GNwiCGnj-$xt6Pw#6gdYpirfqO^y6Dhq?!)*t-a#xIjOkvVzqLajk0o8|#&DNL+;p-+ z#!e-ht0f^3+(W8_8%T-PcApZ2d1CzuLgEXsLrYh3VLmD-HGz5 z?qDzKqz9iXmw*0MBuJNgC?269JxZfczQT8O9;krK`E48rij4392%xk#N2!b zhYH!1)wM#a=;pPC5NedmM%Nt-t1OO`^UnXb9h%ZSc}D8DTM*WJ{2TAp#1$)oYFB4_nj_<3~sN%{=wW zh={y?D%4;OLNUC-$D=ULLLs*4r>HPr7)-*xh*TcKK!NbVvlmM%8HXW;_L5ckDM(24 zjW^J$-?@g$K~KRi(o>8xDm50WftrH9TFHSkU~0V?TKfNoV}TZ2pp1JIHd6#G_)qhR ztQk1L=(Mmmoj@7vc#F^+>w(b#Z(($;&=8}j(&0^bZVM;`qrCMqP?!6`tuV9PLSOpN zum+^l3hIs1R<{f2mv_RHk`E0e;eWcPDNL=u4hMWn^K0n`{@3|{dBxxque{l3W1k!y z@-GPA=-&(j(AcLFK6uXsW=$(owhqz$zeMpD|0&$;^TG&~)e38AJzQdM$eTl{+OIA+ zZ)jE*=my_vg`D;lBKd2+?yqKiYVffQ-j%(}u+URk1VX?6*s?!lX2oazkLSRu!Kah& zIuB%%zHP$crS12j_5c6yeZZEts^P#4Qhe|g#RT$qWz7|0!<{oE?nmfv!=!-ObCt>L zA8J}ynFlK1<{y=H`!u{-sscl;{7xAS)j6{ul}1PZ`|;&z0H(RD!Y2 zMG?gB_g=fne;hv89Lg2>-(v;E3GG_HK4_Kg`+s#EMQ$=zq;`XP?$j62!R6IhQ$G?wu?E>G~3@oWAb|8#XEpS0C!*%`l zj>6r74}`n5E4WG6KFx^EIcyjF`)#Zcs|!#{hqv+JkZ~j$45B$UR%ojPnnPgPy#AE1 z$E_Q1{ilV$V#whBk{?DGIicme(?MDMxVrw;f;{iauRwE)TjP$<66sqhq5=x^LHOoX z&-eG%zN5MV!w^6REYWg{~$rgCH%ih(AYZt|APdr-T2ZCBQWUq(VRYU4=4&+pSl#J zacr$-bWOP@02{cK9P1N)HXZ!ycMX?D5o`6g3R(MC&<^p;c4a_lzw5cD);JfqRp#f) z|LV_&4MtQravdwE9sW;D*nb$QRT^mE$N49;{AC;ef0ps@R&pXol9>^1;WumZLjJIL z|7U#zHy1P{P({5h04B4t|1>}AfmMN9Gx;DYl$$3IOdj@Gx~{muSOdW87+Xo}wli{XXv z{lCv)O^f_r&w=8jtO_w{1XFxG2x1AiyKC4DqPk#0^h(FPHgc^+@1F!m;^NTaATsN0D1_qh4@j8ror1P+4fG=nWfwZW zXbq)nT3)<)ch7(mFYEJ4b{8cJRHzuM+O+D2CtglU?GJ6WuXKya9hm6-+>Y;X9T?{) z%;##g#YJa_YXl4XKhP7n@aSGvCehoa{K?DDAA$)Faz;#8TFq zm1(a?)XYX0Wx{jy`PQaS_a=HrOVVPMr3F!jIT9g+f?)f^p2fFxplurXm_lU@!Lsd_4F$>AppeE2met5@W{ z#3b%eu!TNl2CGMSDO-Tgo4%?4)f=xJHg>?aoew>Jt;IZV!M^k`qwIPFG<{Ef$#7hj zIG4$wW$lcb_PqS(v)dNW56L__1^~{LV));24eQTbF!-oNzU^(p%dOzSD6X6Rfia@N zxwJ2}c(I$Wqlmm@^OCrh-CHC+@xsB?>ITw|S< zeR1WHxo@^83kDH_so|Tl8b@AXcakN$5!H0dh$6g`3_*^BP5U1zlbNtPc%DjzyW6hx6kFrxO+-btZZCk# z4lU{_o;%jK=(VSfad?1m?wRw*XrT}-b6XAcmPCdI3@bCml1CkMh>m^)3*F-R&TNf_ zw3!1BH^u3qy91Zq2y=|r^SZgbM!6=%yYutex-M-eEd&}a#GQ#wk;(VHz2Q()QMO7er8E2>6AeRCb`CV*1 zX6_j{WX?$h@^OOcSjMS_dsJnL#F094R9aqknc=NiQ)Su%CGtjY7(Bol9zzVDq%nlj zhKcneZwoZ7pT#EDZr(*O?MFbC4ryK@4+yRR>>e?1TW4EkREB(giUAE&Apw0>QHxsr zG(?~R-M}12dIuB39nWk9H)&i3f+gVNd>W`UmGI`L+Bo_|(`O)hd)WnLEHJHyd~OH< z=AYXp+~3$|x+G?H`SA+{Kq)?4l`9B(keYtpkjU4-`PyPD7iu{=j=Us(qR*II)BN0u zDGf@SzCscy`nQwXuMoQ@h@+u73ka~r4}4k;Jl=d#%NW(H3%KU)m1J-bK6xAL8e|_X z{1H_#lBywCLuZuD_~7b3@k9T|n6A@2uO9JvRSnIY#+5#tcJ$~#$b}(lR6HL6oB{JvF_Yu}cDlE=4DesXi(bNuH5_+_HP~4v}IB<>xagFTW(ai zATptBP;;Vq`oo1XCYM{B)EmJUgR%DaK_V&xc&!Sl+e;aTosiiZ&N7Mc#-n%Lnj6}h z%%k@d#cauVkwzA7;25=M;6RP2DicSGqRv7*5iTEV`Uy4}Hp*wGSjT4IlwQoRAO^Mo zNU_%=bwqZAq?M6hM)pQ1Ja%zdE0tSrdp-Chk{FkYG?Px4z{sfo%w-bR0dq{jC9 z*HE_e3#LKP3vi#cVScxjPB~oL6=ef2oqznGvr~=r|Mh1e$+kSHJi?YF5FzqZ&HIx$ zmW=fPVM9CLt~IX|0R?j1m^@Htwo(&J4HSqqC1%FoeKp@Uuwyghy}GZ)le2g+!<`k^ z6f=^x;8-RgkEn*Ae!E|+x~sz)6Acas>NkF%=^#gg!K?^vatoaY3g6)Nf*^K{RISt= z(quJ&+e$DErdfSEU>>(&o6gUeHI2&a!;G%jLK>~{R0rkLRtW0a#PDsCp7Tdsc+`1) zqDLzLinW{3qJ7jDOxzy|0$N`905zztB$)P!>}cue0-U|9-syX-lqE?(1)pc5t5{@< ziWm4-Hz0u`R1H!9!<7iZ8@`hN+@7bS2RKwz5^Zf1V4XU!ON_8fkZt&NnjR?1X^v4l zfD9yDddVYvjnq~obB371%26Wr$M->C<+6&T9feh@72sXEUJnAyj$#iMV`4svS(?!s z+wD{2^`Rm106joA+i}U^M%2jgLki-&7XO3TBN2@EzIbw8?@q08ti4s(G(UWl;BUPH zb#oLX8$#pzC-jN#i8nX8&aR&eY{~RjTGLk_aV}OT+M7Jl_5>1K-%~IP3RGV3<&yJ< zk|C~s&|OvJUyQx5fg%E+0^bWH7@dK)nkLg%@Nh7`9xkZKpu_n1!NALHDlMO7Ms3yXb?4G!hq#=c1Z?b6 zV_vmJ5Tih?#7a2Zg~&N03IO)7tX%9oAdMNp{!BE!0obMlA^H=x*V4yxXn!BfUzTWC z0~GE4%eE*b4^-14<$e%Mvs?p74Zi-^S?2*T{+&PK;3My@>K)l|3&a9zZNSbWe;EbUPa;ECN4Fd7x=gcoK{D4cP zb{ufb>WU*jDT3-uM;Rxz_cdJo`mC4RF=)~2K*Lzqu)0|wyJ}nVw7Q^(9#)VbE(;mj zg9%0oh&Y|XeoK2fGL|Yw<^N#FQvtFuVS_UCLMyD4$Ng| zLF~Zm=f~pTpw%*t`OJ5GYcD>AV+r|v6l)kRI7R2-5aHiH#~w?8e;@zgKs#Nen15~^ zM#1F2?*|YslUC|j%xqD{1{Yr^Bfnd6NKd0~0{-6`cDRnG!P=bv)7sdj8Ewel_4W=v zd@03)KtNa^RwiFF^I;5}_x9smVl50!^oYz%|LZm(S`qY3KZK{ALWcT?4Wrb~z;1RP zdLiLSWV71C01RumwF8iaf!9zd8vM#025dWsy#9*u)Urb@cMPqMGyU`!gl6C70_F>n z%A{(B5NNN_a~h7+kGR}Gix%a9W3OC0b_Bg%y(6%%i<>4)0*PNS?+afTse3<+J7qAa-P6)24VFK6OjljXi^5!6*ai+zR;0Y8C zWMT`SrH5}_b$I@~%Vq6&XhOh%xieeJ6@2X&jUQNl;=resQR2S$u+763Pbau!NY{(9 zJ>P?j6{>HSA!)pR1E8Hk$)A2(o(d46-@o#_ut|*0j*xBJCUW`GL z#vZijd1LP}NL_*Ig_7An4Ft53ix$Xu+`h~PQcWq45k5_-9`1~eB~Mww41M(yD{yBH zg25I?&KL~Y`7g7r>{|8Z?_%+=5px-*%+GjVjUzv2M5Q%1PV%?r**jSoX@I7WJq9H) z;54qHiIBf`cTS5-OT;1ecA0@-s;tb%&BV-&;XDWaq2oZoQUVR;Jz`oA&k3{_B-wW` zfCFLxX)Fp73>YJVWD6^kq(b7z6@=Pa7VnM+V1y`o9UKa%torSD!!5JgLLZm5GL&A{ z1m}X;sgv63eX5^1z$3^17dBQJe6$MQ&3*Sdp;b8et{jfF$z@v=EK(>MLeg*^sIz zXFg)qXu|j6^E5L0nm+qs4o-|IVoqA-^_#+IM4);7{BIeA34GKHpFnWcOKUpG;B<+I z^kJMvQWGVP1w(AhlP>ruXxk1H1IJm;*pZTw0t2|I?X8zT??#Jmf}8r@dlfyc}h4c3v+1JAg2l*LLnoPS5zwlzK#$zQN)i~Gw*)b-nuXvefuGS6u<`9|NbSC zjRg$-4vm(mWS(@m&h}M3)b$*>k};CUs>H#gIR0B<{t|g=ElR?yddy?9`J>uK{Uh&f zHi6Hz?_U1p3ldVK!l}H2v(qsVXP|h@{}0FOeXR!QT$g?zbqd6EkACV2^^EH2xVo+JKxpBc67W5;OrH47e7IK3M{-2ghP?B? z9X0^9Yp%`LX}=seg)&f>Wu|O8@`r6gOc{05W$5{pgy%uTh(W4=t9hipFXQ^ zbsgv8HFvR*I)nHXIt5Lk{hbn>gOl+qaj}xIHtqOq8#(K@3u zSipu~2k7JC+V}opNUQU(-@x@&O1`Xi1>PnI$RpPNp?0w_!>2#R0#w3UsQ?%!N@lX= z8du_hPi9d+57FBG1j%@7k$=7B_%V`{!(JBerMAF~&9oQ=$JNh8(ngTuTFb(38OPMH ztE$;|vhks_An;rXabzYN5yPg>&vFiCGLK0CDq6unN7;t!1cDWYX-GO0mI0?!?58oY5; zLHv>1ghZuU8h-W}^uis`*7RhUSZvA0ATiJC@r2iWEnXzO#KLkpe+$*G1QmEI4`lbC zemDYBv(SCEMGH7hOnI63=|HCH%~3gA;gLi0w|{@?pBW3^HaxR8B}r2gtWR{5sS*38 z^x*3o{#-0|n8RqK?O9y7jZ#uW+fw7p+upxl|4;O)hR26P+&fb7K&W-vFv$v^4OV8w z1-OH4W@|cFc6%y1LSq+%`oCS6zYqe96G~{N80!T#Ya}Txk=C!ypQ`K>F{&RIJ0w#x z=Ln8A!+&5~w$prb7=Km(Q(QJ`*%t}|Qnl(}?#6hAJ;Q!;GaE{&2;hy7gA-eg5RekH zl_G+Vlwk1=R$m3?1q3ZCj;-@p$Q$Uxvy2D(Z>90QgPqldfUeeTF&%8d+Rh`Y1V97c zt{D9f42~KQ`i-q%IZ|G)?qE9`f*2t&(sy#F7ROGEDY*TS9y)-um?r-Bt-22{@5hJt zE1f!Yb)~4p}CKU8KAfH7*2n+l+4?B?c}z2$JhIwlb=i!1*d(* zOko+})xby|ocy;}my5lC2-$LW>|jcrY#SD;8dgY$t{lgO$D9WZmEDLBmHCEIkp8W> z|MZYyr%gquq^mSj`H8bf=jUht+yq;+sAPB?Ie9XJ9(3WFq`(d$6}rBJuyI`_dX=)82T6 z*qa}DxEe8blGM{;s>KklUUW5}1@^}rW+LQ>u_CzS3eB}D9OOoRH-IT}r*X{O~Pe z+Eh#Q0a5~a6OFxeg$FX3{P)AXzJAm*1Sz5WEhWE;FIL@0(HsucoBK z)VGzK4h@PS0VmPBWjNMU6x^X$fc(z=6L%uFirG6l^+a$!c0R}a8{ZMggoU_e3W_n3 z3dJCPu3>4DBbp+h3zDs%g3?N;%2wGd8R!iYb(A668n%*KzKUF_NyT7r%%8`Sb^9~1 z10B~3u2<1XI@g5yYAo*{ZNb7ETKE6BSwOy|0rD)GeEHH9W?VQIuOnNtSkpB4%BOz& z%7D^9ct|M0o8dOn2%K@jLi$cV0n||Kp@wE zD$e=))a~`_tgemd2HBsZ!YJJRuV=JT{6Z+PSO~I5i3g+YYhLGyJn(t>Nz48*Mljue zD>Z7QGbWboEpN{UqQ16{|2_D`5&UerJ-FQU#SfO%K;Qwbr>OI*v8F9@E(>oK?df0< zYY$lSELT;{yL)&S6{DY|J5cyNci=pvUWC!(f#x+_&&FpZ)N{q-A&YWERz(HkM;?*a zYuI!FB|Y&F4&aNHz5JZO4fkQQ`y!91Uw*}GLsyqOrR<9}_2fB2v>mkY zD`<65Adi9BW-Eox5ImiSDxym2iLL#c?vKjbOG>pYtPcf`jnV#f6lN1nqPaH{7DjyW z8x{SX0D=hFUkC9yT;$O9(xBH>_-L7!sW9ZEi(2-bZHBDSA+zCk?D5M4en-&M!>1Mr ziCo@G=jSY(hg}Sf*f(B={7USzfT2revc#bOhpy|6r~3QmsE;u zx@4t{+>GklGa6DxCCSZ}>zcR7UM*x?`;wh?Z8C4xy}#ELQNPdk&mNE5_d4e_p5t}Q z$}1%Aizk@~#V&8Sl}U)c;Xth4S3kVfLN}UD1ug8hwY9n?^EKoDoqN^c4Gxo9UI#g;%r z8-Vuvp8Dt>)<3$QtxPhJm$mYZC?tLQ_jj!SOoB)@7BC(Xz_oEBb z87+?Z6b>0a{<>vs9SXBPQs9;PW-dfZx>~KXWmB^}RMtFy_aWfxSHf8kuNK_fJt|zoTeNBojJKUaAS&Y&=kLSg20v_{SoOaSoRF zNW9r{p(KG0RWM+){*j%=mqq|lb}ax&OVr~ZJ@OBDh~EXo_Ik+L?z?cupoh|ole}KP zTLR4`_U!E(b-=L_T)PLB^C4&`m#h?YT6SS~jl?_9TD#as+$~Nj?Ywdu!MDSyRsdlS z1xO@0i`!`nME0d$peH1-)2MqueFu8~H{CiKr-I~d)Ab#-?K`f-i_-O-SnKk=D@AsEB9G_Hf7O}O}9Icbqk zhLHmu=otq@ePChda3cNm#YG_58yjOQ#E`2w)*y8sE5oUxUh~$rcI9&~MU|0w?X$!+ zJ=cSeK{&i}WD2i`{(;XNGO8AJn6nqO!h9roYcKbXf5YD`B>q-+YG+oL$cVdaW7aVR z-NhncFG68MKII}Sju(P52A+O_4E@iiw|a}uq>QZm=(itv6X2;S z{aJ7_QRFU9v%HZvga2F_R^)FJ(r?0I&hyRZ`9A(^w95uWdRTuSis1TiHGSo6yL&R! z~>B1T*=)cZ@oyoKM5d-`g5+{_woskwJxB-13Xz|kjlpqhb0#^DWL!y%~aVjd+ z^}Y-|`hB&#TSHAN@eae-gKdO9v#W~wmmA9X{x_`Voq=u8Y4UH&9k4Bu)4 zIbfX+@TJyn!$qMJuxm~&zrKI>oyShlDy)Hk7;m^LCz9yE-6zn)N^M$|1I_w?OfMysr7mp=cRvpW1Y6AQ zic-9T9*c$z^k@AaxH9=R?LsIBWmt`jc58tI3v?(;xa+_9?DZW)O@qM{<@7<33z&WI z=kDJ!&M2R{S*r`JX0z*C5h7H*+y6(^3u7;Vi!6%|asd?z#`D99n-@mo8uD_}%@9F@ z_#hw#7vx3%kb9q>#80 z7~Z_s$)kER7e{wPL&zGvfpOhaY{Pu<`1t6M_;_d@59ocM+@rHrYyV3S!3WfGbfp#P zC!n{B$#6Ar;y{@7`3O7QYi0Q_1`E)&RyBOu+*A($2&M%Rl6l_ru*BdkJW&g(5jVqNuHr{Hzn0MJPZ*x?U~ZdV5?=h z4sEH;|8GlwHz+lkP!KrtT!>4C-EOSr*ZL3^oK-YFGpSye32~$K-;x_xt?S=%H`1}6 zLXyM_>P$=h0HJMHick^S2-?p7s`{XYs>#F3Ra0*r z9;u(w1fS`Zn8|N?W{?mj6u8>bEs{k<|MwpK>>WQ{hXCxt#;tb6AHTypoRT39aSsox zZu;xef8`_+DcIo-cubT{gR%`zZzlOb4~5+hE=(3Sgux|h+-_v`z*mS)EdRYp;o1(p znO&3->d0>T$;!IKlDAyB%$v3$#w`Y3XmGp5qY!ZCJC;%GKTI*~vkT!*>F}bN!2-}# zKk%XMWH>WmiUPV{&O1Gli8Q$)KY7bh$woy8$NnF`u&B^8(IpO42To2>XGJbpe}eeN zsE}c=eBxrQIkd{OOnpr#j+S>RQzsa!+6t1NWZ{W{w<7M z21FpJA1?iJ*C}K*@81tyrX0PdQh8D=(PSOEaPEJE7n>|B9KTZfO3%KoB0k8N-W{}d z?$MaI#S#vT;JwFBCwJUGM{Jo@Lyc=x1Z@T?effus^cUL($TRe@C*;kDz@um!L4^Y; zrJb3${4U3f{JrQVf~M3gFegNv`uAa0L>-uE(Sbe&^erK#Cw;Quu|S4LeCD+c>?0f2 zd0{CrpwRkxrNT`eNWT2#qGp+vUj#^C#H0k8Ja)Z=pD&7hIvv*tb!($8}QdJUu#9bsMP`k z+Gq=%kBOv9v||Z$eb!Ac|EBBsxHxDCq9@ktn?A!&U%B4qM`(f_P8|mYC>FAf1eEhj z=C9%^%OgEYS^#<%J*WWeDn=(_*VBT%jU;p7(SmB9>PguEe; z{I(6jN^*&fcZi#mZ$UQob+_jF?6iP4Q3+D>;6<$oRN=q?E$D+G7z}1XhWBzqOAR1?NO|o+K1I$$U^SS;? zSHHC!fXwkvlK3OYQ>{JsyiV#}gQJ0&cicY7EhSx>Lz4z_Y$T#pvlANNjeipuXV-Q-ZMb9}_{gg98M@_k!#%2%n5CxQTBK$aOgnZ=9)+VS7x0KjFpFw=ql z+&b&c+a;?!&%xydT#-^s8!T1MWDegZY>1;J-i;slA&dO@0`|f;U_17&RgOOQRc> zndr~1s}hIz?pr7`xJP6UEZ_XUEgVZGHUgEME=@0X7SaWJ#q0V5AujTj(RtYE)I)Tpwpc#emT?_O zOVSRT<=N?Qn?ZOLUM!2~s`Zsb?bco;c+upDF1w$-l}>kpu&+EJ3f4$33M$` znYih2C$D(f^lY_4^HEGHD+k>lYqlJto)+}7zdzDm^s6kd9FKTW!Ux!h=1@NS{Ucd@ zq)X%m1>U{**!__S2JY?=&BvA7!Hc7eZIbnY+Zb2tsnPuh(Kpf_Zp zmZk>K8y9<3bILxd2=OV#JD}~j$&z>up(WZ9Dlw2R{>9G1=M^Vh(!z|)YqHA3EUL&p z{!mGgNHjsV!XN)>(Ee2fF6`y9^yvJD>0vc&S2UrHS_q*R@eVhE6q0IyPfG&Cllxz# zekQglR3YoV1&|G{J_KmIF8_nGt4*&p!W^RYHaSA5vPA{c!HgO=#^+n?P8@%tpu zq<`YKNuGrsgbXJnVeqM;-bMIS`}>cN-C%lz^#ss!INQ|-YLTnUiSS?T$Ga}{u!8Ei zy|?Ch*ila&a21rYY(}siyp%`Fas{p3pTC9mvJLX@(NDlggEsKh4BDir|2zaA#miu@ zUiSzq2iNPSrZE$fVdND!o0b%`e~DXQX$`Y(|6)2TWnriZ#G$6V-wzx373pM0GRYuzu5Z6qZ_V zLs(sllERfL+}Z(jYo>Waa!hhUhN03cmGPY`FaHqh4^Yf((3S1A=HyNPAu(KBXp>o3 z!eN+@68kdiNxmlr4$2AWi)-tS)vc3j)$0CREBD zi<#SZ+*ZExjR=Cm;R&(Iwg>daeCap-A@u2mTntc;obsNR2itIp39xE@uk~=1@oY$8 znQ4OX29GS2uK#GIA)uIeR(o9w9k^94GcU#rts48j*6T=XX}qX5_LGa7o1g(B;>N3G zB;#KI&ajvyfpYYhCoN~`eYIt+^6Ji!5-NG<8e&4%ur#X1u6Aq$>{L>pPTv zjd+3qg!}p@xxTw=&Jq(O!s^p5Aa{di=lF8;{|g2A2xRq_O`xGJ`|3_DyxdC(a)aRS zHpDjn9Dbmsk7c+?ne5UB@P6jMrQQm7-=}*0M4?~b+xsr%Yq#@=kHG!sMC;5k!TqNX zDP%n!lq7{&f;wU;4M3RrSC|OnFG5g4d_=2UioIK?GRu%NVAo(&-LO|=S8Z9F{h{+j zDvku`5m92zZy62X?f(|PX=U-#7mCNHRo-Z+Lbg~~fS65gple7Y23P&%F87 zwdds30MP65J$DcmJOwX~ph$b=n50Z!_`Ck2XQTV_Kz`i`vXqN6&Z7ZORR73APCyzq z>}d2O+{*p77uN2PEz2)pX|fHy<~h-gex;hy5x6v1@Wehkfy<3VG}s1dqgKepV84QY zabd^{9lbRyA?2KQ*)T)JG!{Ap*dd+GOv=}cxc+MQOjWqqkH!u3NNf=2cV%1$$Hxy( za58qzE4^QZE~I~bdJy}Q?3Qg7UtY#L+*H9&Pc;aY1_~lEQ(WEOcKy;alW?wRFl|;U zt=Q`{W~seLmzXx${@5wBs{4CG3dNbTLe}``krP{L%hIB)^jlhixAc~_;@!O*>Y%#& z<65G7&_D@vhCxysg|)riDkwwi$nEwOmNe(e-XW_W=X|H)ZML2ttm_FM+o z#n!84reDn!k(P<@Edrlt85uk21#ty6i^X9U^fVex?%FPOn82yffFyancUI zK5Rkk_QAzBBunj;==7-+qI@fwjDhN_Yut_}88T;u_3a$27rI7bT0D3sdT@$CZhJ3MgWw zHfB0nz>_H*CXCv#5IH#BzBVI5%xs-2U2r%``$B-nL(dUS?W0d zR)NC4EK`aWlxa&>+|wYZZ5<_{Ue-!YQ|TE?Hx{Zbl(;)CMtswm4H5H+uClank$;x$5Tn8<)K7SSOKhUS;!cDnxc!#k~7DhSB$JdSmM-J<~37hmFJd&a3WwlHa8d89yIDm=l&CdN+BnS*peoX&J*(5`X^M`J!D(t$gnXqQwsi?k&~6_ii+;?0Fx#-dcekVZNkFeio?L7MtG! z9>j2Bhv2&VSf*;~{+=1lO(-D{Z~Q4Q|MU-C3nZz>aUd;fNxGC9EGX`soZUR0zqIQE zkqre>Src#Fgt~*AHoAHznUv1lwm_=PYxTwj?M&fpsp&3r5f0RoJL~QqPPRLW=E}%# z69mCw!T+FVjzTSTWZrnL*{1QER7th)xj<*X5{ekN{^gf_mx(thzux!uqtlWEeaq+h z_&a0g)>}Vfw7FutHHmXQnoFMctdSN)ZK#e1d|6|uTVC>tLZjS)&Z3*=0k=5RW~z@5 zns)CnREmTXFg+_mC7lXcr$&`4qhJH(U2N{qs4AR5X%Da4{+=H|e~*Q7HhLuV8<-+e zJ@+p9gKw?ydqZBLYhoc%8fjI|*B{Jn*zC;yh2&HKz20_mb#N5uJw=lJa&%%xqy>}I z^+)Q7zxMY$KR)mD_?3w1V6ek%sMBYOkX(@fNjW+C>l)pG0Ms^}0$|PfzdAucjdsWm zYl6a&?z!{;e*r!-cf@`BwpK8`jNWS2d2SlH^XQwVSDuo$xZd34a1Sk*EZABQJR~+? zyE=5{Ozxcuq0$g>M_i|qSEInUCtEeHZNWDj$qf&u#pVF+w`3alhfImQ0t10u0Ia1T z+T24GU*YZtv7#SQe*Gf!v{_5Ra0Tr<_ue?G&!X2Ecg%EPPF7{dUCAk{X4SQe-}{Jb zc4o-5Jh*B@tBpuWLFC7MbOB`^^MHLmLE1Dq%7o6C(FI?iLHRT0k4(4O*&w)~1~RAj zHFJ%w+$T!{j}<;*zhCmSK~S*IZb&zelWgMWg7{*Y)#^4E`Tb}H_R}a!r zzo+NiSBINS6;cmoQu0WYHdP?Xc$bN?Kkm>)+=?Xw3m6;lF5F5}^8;r}^M2khVP-A_u*l3)Mmgi#eL<%?sO&_k4C>>15FUmKoX9xJ9azBCw|eD!a>slOsOxjDbOTwKkBK zqtK!Pd~yVuL+mzN>ggi7*@z6#vI@qz%~_XN=Hy%S)h$P7*O+;`cPnP_`^m|ZExhfr zeo?L}6+zvuYqOf*`7Sj-;>PKFpP|ig7}i|Gx_-;%?h?3Po1GhUQR(xktc@%Th&K5c zloq8Zuizb~XE3|(zSJt#|*TpJ!5e&(G~zfec6DLQQxHhs8R> z>Y34O!pC}ywJ*~z@AsDQ*)f`iG6Ysm+8Nz0l;R%ijF?bPcq z>6(zar?;=Bt7hGt5h{J(=w6d(s(p@r4>h0~2y1Rg7Xy<6%O8aBZzd2NmE%EaTYKB0 z@zc$7DndGf&-yM_`>?p3+1LKod*G5TF@3)5b$R#>+3-6UIqSHf ztXY_wyLJ)iSz){Y@@9yqa=iG1Y|}4KO5%Ax4yqr`_Qj)mYp;K_VHUVMo%M)=h9cXmoqPnF$H#w~ z93Z{ATXh5i+GVHxgr>TDVEX^x`=cjBx~mN)3O7hGNWS4$l*$k+jQ1#!&{L#z%~?=- z=Qt=d%G}#aJB0%jG!j34R>dSw97wS7pLbgW!0C#@=O3K;0SHISl-omruzFrQ>u`wv zqdD-Su8~fy;!ZFQsK=dkr?a5I$yJPO-q5$bFhGVOBc+MEE-vE_E750pW?n5A)b3}&7x6f0#K(d@pCVR*c0fJVpX>7 zSi%33p*tT}dB_t3UA{6?z!gR|UvSLf9j-}PuEcKX`3a5>S{VB4Egd%M&lwG;qWjns z6S{KcG~zb`mAvsB7fFF1+ws20NbsaN0XEMu+rV8GE^q^a)*~MKgr5YhsfCynJRs5-bmVN(lClCj;Z4bKW zRc3MD_)6Aum>mr2E)5J5byBuF+&7%wCeAB<+>Jcg)Mt~Gh@85t<;w}lM9wb8Q^nxr zM*dyko2YS|E1G`17JY%VjNic+l|_Ma=5MFWN77{|VU6o?R+oPRvWo?p$#OfJ22tb3$ozHgFp(Rf(1t_?lv z5aB4%4XG&tuYG#JprO!6j=2N`77MGZVR@`DfVB(rZED!nO?svdPgQ_;Dnzf;vx46( zd7yCc5|f8%UDgm~{jCd6_&N7o>`QabGAl%)4U(F*672k1_H<7U|6o9bC}7)$+$YNT zKj>t!abHk(e;8@vDq#by$s50oup^hg=OWa#AfSdJ*j{3@Xj^c3q=u7I8_Scrzm%HF zC>BK|1>*KLZK)l+{D9#pt7(JJo3V5~i)oHS_3v6ab@c@YkIqW+^n_&2zTRt{;3TKg z@YMM0POT$olNhzy=5#LLoi+gBJq^_*&x-NtbY&xqPPbytQn7`JQ-}9SoS>|I5V#j^ z;EcXA@+2a#t#hE&#W{KV)~w@2CU-rtn>-UOt3p0I+F3i~>a=se$w%p?+2>GgV}#g! zezaAg;ME3&UCx;qRnP8;TAyU1vFSewtKaDw4i6;qBXi4+gT;J4FX--RRLcH~T3SJw zF-lhF3J)GR=HuDssuSu4(}eS<@Y>;KxB)%f%erhmL9g?>x17#z;-1hi)t=74@1Q^D z0FWlPfGCpbTYSzapiTd^U4vjno2j<>Ro*A75zmv?TaFELIH1aA8+qEVy5UC>GUwDh zYt|yGbC37(_ksf}# z4IXo;&HOjG^`Xf^m~S)77}b?HXRqnl_m}>1oc(H{Lr|)4sk3JtOei^v&Z~ zr@dIK<4anA)FjI0z97LT#LSU|4d^(nr^&ke&?as42?d+c0w*rP{@e$Xi~d2U>3)S6 zNJ+61s-1Zu7)kWYt-V9HwVZQ1A{^c$B-V~JCe}=nCdl3IHw5s$vKW|m9w{2D6!ycG zW`~^#^(^^z?41fouaw#+o9)J{0e34|k=_4{mR@AdAQjgEN9h|gqA5Ckfkiv8gx>{$ zxx*94?s{X`bLwkVH>c86>}4}+Uhm0EX~;RFftfjPiJN}vhp38(n!rXzw#Rh;L}K1uT8=ObG=cLN7xFQkjDxhW z#OcGzOu{`j0CjY4B{b}OP-S3tOQ&RF3tk4E_Z2<3vnCnynHYS$Rd2e?$so|SHz%)` zVdnOCOk%M8^!w^N&RRwr`L`faSi#1uT;9OBsd!f z1m3-K1U?#FUO45_L_P$vNaKG|Zy1^Md#D~*`l6fncThCfqHwu5vgO^O!Sl7=x;8jYRX0J4RvbG+acsra$V(` zJ3c5z$leD`G$^@ZU=5apX23f9Ki{qmLYx>69YHxW=7)-D`p*GV-%L35)mOwtQ7Rub zd{TRYbXk`uM0siVc=}m)rqfB|PcIEOQjByQTwZ_F+jE6Q+UWN{(bJ#3CH+>4qN}^T8{~1q&b;?W`OZ87j<3b@ ziZ_4ds5dRH-3MtkK(Ra?5}$drX`ifT`3Jk~9C1CQIyNH3KFIhCZh!8XHm6^=WSVcl zp}cj>qYgXhAwWMckG~hoGp@d>);nu)@h(;9^x?^iGYCr8Pgop9`{pC+_YMb|fW0Uu zbk35d$1ELOaLal`y5cRSCrBvq8#bUmkoR`Q#v&mWIXiZVrkh3vk5d~+(wcoo0o8gK zVM5iz*3|2Y=Z|bcUuan_Emm40A--_}R!Vq_$>45FJU*mWewrcoW2xu-^ReslL+K5p z8s#K@-VXCHpWGZN&UyjNk<9H(Jl3BhaKCVtKO9eum=wjozx#HF)-F1~skDDKdjc7% z05#CDC2ST-zxS?y4j5+6)>}@e{i5mFwWK;xN;V3%4ynOhL;7%(I4G-t^H7i*@O3-W zFnscrTIqZJKo0NZL1tvVWBZwsbAz0wKg^vcs^rc_4RiKVi7!j<$HxwIqw{J^wZdjw zU$AO7 z))YqzKr)#o(250VnmdNuXfyTM!lxz!u2^m_i$5x=0{a?tD_71EGY>z?6n^TamGNA; zgu);99HU3?7tX)BKUL|X62u~BUKT2Hp+CPkS5`* zXf=J4`yIq^`s9yw&H=|BWqHNb^Xod2=4yoIbfYTflApAdxs!w%?g}-<^81H{0Z&LA z{qwZ3z2sDO6AWv{QC@WWuvFs5NKN7W zJdJ6O@Q+Ti%YIqM&M7T2j^CY~eOs+|=d{0s6$Dzm&(L+}%x#_+Y;OI!rV+^w)hW07k*^u?xfdTJ=X(_5u(eetrMm!rUY!jXdj6Pfl6 zAD-;p9-ZGZ{2UdD9JgsIQ{a44RI*j|njwDrg4q4pDb}9pkfCPLo@vH)cEEXlfYV=M z(Q?G-W>gL1eu_B<*Y_@ezNf8kAV|bUf4KYRBkAQ&Fv!{Nbq@_x8ULt@7kv}om)ADa z!I8czsDOW9*fX+#|ugy~K_Z+!1auPXf)FC7VFi`{&*A`J_|0|SI&L!Syd(8cY zmfN6t7M08H_WNc`g&eEMU6H3I4#nfesUGR=W-bmqVHY$Y;*kBIfn_3#CYqSgb+KDy zep?2s)}aMRd01`Rz-Qf^yM2l^8{WD-qPp2Aa{gXQ=knpSzwz}G{2?4Bb6E`-VIVyz zM#oAU*rwIvCyD~xYE=5xJ21`AvfSjei@B1OUcIxEtvbd2&Agz@US8xyiyzvL5AU7E zz$^>mU^4{kUTh4^DcDh7GV<(uc-Ao(WzGspJaiDyf^%!e4F*Oo!9fO|6>g2V)p2>h z_gWPJV$iS66Z{Ho9U`tblZt$>EzAnWl+6>g5c31wpOsz}JC7EBi((A*`Em4GNx^_S`=P)XR=)2{6)69)7By@xgL_mxK*prwo@T@WrsW#pGbU>|5J5t?+v#;<91 z&bwh%^qV<)3@6O8@l@*ckjx$QR5Xz}B9>Qw1?)vjjoXm3(V-!|Lp|=f7i8QYhMR=O zoSvVf*(PIp5+tvOo2*}rkIlTf1`!w+FWSdhW`)=VL!cc_3fv4Gx$rdRs&qk#1x&;8 z(1N&bymlrvcEbm^_Eq0?k~e=Qy|mjF;&6fYj@!s^Yf_0mPsGQwDnd;mO7WsnBjrUv zChX*cSK6xN*1q7q0e$a?L=qUg$vpj z3%X2nHKdrfW%pEHrUX4>bKa?>lF7h)mS_=u$I^)x<;C(HtvAeF&kYm^&km=Sli)m`6Jud-E?e7jzU z=urwM05DswJ^U{A<7X8imwD2!v}spaNa%}$i9w&4sWFRY+n#ZYzxkt9JqlAzxSUW< zCO<*{UM)XiwUonZ5vdN0ACyF?>TfP2CE(xd;;u5XJ%9TVWlFn#V|jwjLGTJ)M16EX?Gd z*ImaToklXLInBzchN)4XGysAIrd{8`AOA3%Y}6O9_Hd8OKq`GnXRTP$CI-Z2<+{+B zKsauXIv10@xIu08h}pDx(>v3Y9;b~}E&1}k(+w5WA@?+cn|f0hb&19HRa(=5ZFA_? z8Bj3?-{fTbFl+BGT$CO*cPm5ajwdL{u$d+Q*rOYG56#f;f^pnbzGpR%W-u(R-=G7(pko%q)yD2LrdeYk_{#OQ!{ zC=7*N*vVQ@pS9;O>mxZJTuN2)C!4$N8_yTZdyCsa}}6$4~@UM*$!0b7i?j!!hXEN z!fb{xBIqEGEk`@>Rlmd=>MOSKrA*+RuNxbpsaL;r_Kw$Eo^y<*M?SME63gvrt4|kd zE#Dks&*I2c;gdc!9F<1SN7%s&C?@6T=qXCpk7q^Jj;A=ZKCH{ifMI*0%>>a8mGMUT zVv>|zTw#n8d^^&w7-p(ea5$Qg?H<-)1A;<3GlGL9Q9jGZ(k~6Y1SBhY=$yQ~G>=^} zjvk5A-i?-epL_zb^I-G7{m4m^Xnc@%rs8XB#2OtcuaQ@B+h9^8(YN8# zT>M_d#~Muwp1d^4;fPUe7QQDRNXOF$UdZ{@&$-B;&=ND>T$1E^9|juSEU9w*JW0r( znPBWY?bycSmu87^^KH30afEc-hdmT6P$4xxc7tGtOyQChZQRJe3X#Io=XBIy`cCUv zfM3Oy-yl6-V0kJf&nw+L_%=90Ez62pqcmwbJSO}kJ_xO0nEPeNu+OW*hXOI z&fmD0D8d`?vs{Nz>iZ`}R3HAj=yE2_WOP+-mE)Hjm}32q*q%q#Z2!v4@Iu+3`m3Ko z1xksXKdbBuIT@MiQX@D2Xp(rJ%jIf0M<6qM=+G)Wu_e|gj{%k!>-O4a*Sk$Q1<-?7 z1EvuO%LEu-~zLac$OFKhM_NfJiF$Jx~-q z(nLY2ncjinq1Vn0Tia{+#_9_ioKH{d6J>Y%tryPAnkpQi@dhvpiik`%wp5g$MGmR! zJ~1({?|&#hGN&UG!tT0VL`~a!BmY_i1q|7FR!uoDqAtq%@8TD39l|{|b;CY^gVc~- z8mj0Gtp_<@PuLUWtxlL&ofWbN_bCrMlOxt>l@-ZzuuoZAAr?O^czoaux4)!(lF3gE zw^!E!c89bm*GX8=QUzRKehNNw{)^hSbp5V&q0%sOz52Pi*O-HXx_842cIZ04&DQBY_Rf9@OMUeYjOhFXApn%gSW<9fejz57; zG$;Ra_~x9uDu=IrbKn<-amni&rHKY;Og|qV%nI;ED6d;(95jNTHWWFgmPJNcg?1P;_+|vBU`@64%MqG6U=k#-Qki z4Pe6AQ6|)T60@#f_Zv#_q>+O^q zJyJBBfMy4~c&HRcV!F`0C#OpVgKOH{de>PsjohIp;FK^uUMcZ4f}G;hoQchJJ2abI zq-LHo@lvihIeyy1z%J{|!?yO=iPz-V*yG#iv14J<_s_cD_q)v32AUfXuV0NKTh(Ph zc~7pYsa@X=!`A`=hwu&Mj?X6?twKV(PtH=uln3_tMAp@piHdpR5)VeGzj8jn0LFFV8i#LFOno=?DkUWz z@9i94<6%~j4wfHZ0CotomT3kRNd&S}OOjT`%Uij9(;m8L?HHL}nwyhb=iY+n(LPbr zX=@x$bSRwWW@DgPe%(mSRjE4uN}0L%#TKm-tJnF-ZBG*xpSpM4_C}DwI~H%^_uEC+ z-gUZ2Qm;BmbT-@?-eyNDcN*(7GQOf|hq6Q2MA!OB$eL)H8XHSG9GmSstQ_i%K8t?w z7Tv#L!Y;Uf5unnbX6LAV|6>vlu$S8tP@wGd-JkR7q5_dc51wazQVV1ScNMe!*o z9_P-W;cvm9QM2#lH3Q;J?17Z+o)T{TwbSEPM`iqdsaYYjhm<1Fg<33U0&!R~?W$9` zIpg)7n`+F7Ec4rZHIvzv9~EBpTR)mz~PTPIyl4-Ihc zU2%U0SmH#j^RP8NJ8NE=*H9%Ts%{#oD9SK~qmy$t3ln7#4Nf(BJ6f}zr+(o*UFE|v zeO2$(v35MR>=+?KZ-8PThyNhTVI>-;`We@Ml{qVUf;bEUUAwf zeTRzTEv0bmx{WG|XeRk9ei9-uKhj z^GxG76bWLNQ>4UO#Kpp~Y~m`4YZ&poP3LxLjE7~2lA^Uet`{eI)m~?!@TRiFZP2Va ze=~gALzDRWqT28UwuUnL$1N;|^Cd)zzCOFyZ9JQBS}0KU zz!iktYp<+Oznl%TA#bD&Id{B&X0GgZsC|{+b$F!#BfPdy?&yUMjZ*Gscow@n0pg)`l7~@FPrX5k7$g? z&c>S=X0o1>MjW?a`m87L(`v8 zMT!Exa;po_@3Vn#JXohW>0ZUPB+Ja!vEwd?Tk%RPlvr<VUWc|@o*SE} zOo`|V*nQMM)IO7G ztatZpcW;d)2WE$Cm~cIi!3--{_9u3(FpXlyxp=d7U# zt}wv9{bWm|zLyrMYk=IGoTKqAOp~C8be{e>XVDZ$j7ePSpZ z<`ndY^@O=8cXy43IajGM3U=)5ML(Hqn`2O_i40_ws?i-cAVJAm=|=14oKYUROO2=f z8e-E6I5x>VR`E8gHgeLy==^|n+=bs8d*c34{wephcpM|EMve69xZ=~QqNKg$J9}B& z4m^xATfIS@WR%8UamBP8-4ye!tX6AJZbt=g(ZiN)@8gQ^_O@q)Z_&0Pk&}8er!IS{ zQ2H14bd_Epx*Vyfc}_xENk=`>8L@Lnfq0{S$I-d_S>F?+X_|4oP9C<*yEUyR(#JI3 z%I-b7+Mia%VD^@Mi&=Rb!+p!DuG+c>WshXT?cy?fACYp>>iR><+;_oUZdh@bjV&G- zOwBvuaOY)xHQQsc3Hu8*{13k)e1BzEe%rb1RWk0jyq0E7l8xP@dZ~3!M^hZ}QBt#t zbxNDrWa*0n<5=v|^75CHw*ot|W9|-J3BQ$fCMju}^*m2HE-C%+&Gc~7C+E?YSQ$k6 zoDy5z>Sp4FKFpIxj@h`Tjn5m-te=oU&f&)!?BQRA@`6jzaiFKWDF2$|idl%k=S ziJIZN^L%yJSq@$XpqDZgE?h?37;DJ%98?RACoW4;!t7RPT!()KnL>vuud`Hr3m0 zw;4}M>w{bdhr_XT!Dpv8zuy#h zY)xG8H%}QvdwlFHj{-?TMq_+jst`Hc*KBrXjfs7n33mA8ba1JU;La}Xk-9ly0Ygni zlYrs#n}pPd6ObsD@h8H&h#q>IKD;^IpDa%*`@Tl6**1ogz0qQCizVyCbAyYn)7E7PGcT4&fHPq^=!JxY?rv8Lbp*5oqijl5Oz6uUQzPdZnl;=<&zj%Imo ztz$H{Y9d(yVW$(-zxwU%3*}t94F(UvzEW0EJe^Py%j=BDZ!9mZpuB}+&ij94kG3$=OXMe z`=;Hq6ZGH=xY?&H3~%sjJO}o;>&N15mS-6vOlM_QYt%%tm)JeFP|@U2QM{0MH%>Gm zoAC~lz1dlk^FTyL5PD5%1-Fz0i9FjJkITP*dtKR24aH9in1n0ej0k)!LE1NN?cNpm zE{?o&`;~FW=3E|UO981EC9{tBCKK&Dy1Z^l(F9yZ-xC=H3$FYYu3nF+*@$Q7lfCMV zhmomXv?;s%S>xslkyss98D+&dvT=9CG{H-*T4Ov*wCUGF{A}7;hMOwbcV+isjT`f6<+ZLNS-k#Dhji|E@=VqWaJ%8oPsP&Eb+o!e`B?2eTd2< zuT`3gpUSGAyf(pQBh20^iMs?{Pif-2Xm{g)Xpv+iZf!i)cC|`b%HY{a%*mP>^|?lj zdQZrn@u9mhVE-(5GQLUAZHvf0O}jk?iXLZ)k82o4FP2-=J|9W@y3=h>Jx?t8mBHzY zq}l_$RyWpaM3@$$YIH}rlQT>L^IymxWpwcyIBR+T04#2a;YXz4tNlz|P(A4%Y1^MB zqiHOS;FvyvhgD+Zc2GlchJl<|ee`bGHDN_;0jtUD^(v$;P3cG5H|XcrysC&59palS zQ7vi2=f){+xPoxNQFV1$g|3L;aOBU8SA@kwo(lF#)n6q|*!fz@BAODVW4b%;Do%5| zQY-QgJ5Dn-o>fDrsu5o+AB)TEwxdn{3PjIOTcz`Ibtc1DsBVg4%S zg5q8`OBeQu#IjcDiT3<0!>M+?S1BWQ`K@C3wwrx&MM3W1T_RcE;3N(X6lqB#EbKhT zIIr)~&|IhC|C|tvK9Iy~#~6K7vu4`-S_3R_SSWi&t1BMHs|xpe5mr{NPWvnQUZ#+lS@w0>V7BAtJ_Wri=SWurAI zGLbf4>P8z9bemVovF+)}r)Cc5bC6o6(+y}n`LuSs{JBX&^~P?3t6~Ak?ZpG?6V@a1 z1h;wCZ2nwN)1pHI>)k_R$?EmKrnp&$E(k2Hp^?vS$sjg)L*=6ktKSPw*v<4`7yRhd z2$=_@>YShm9s=Ph0tuGJJ<o@Z$UD6$sPk5!K zm$mvb!X`+nEXIdGW*tUNgw`2URT)qxVvuQ8pHnVU+5I(33Tk@B)MhRZyHkWcZxY<4<&&MDk~fNmiNTYUfvxcRa3kxh16!bCEZyLb<~sA@mw48V=F^$MEO_3|-Rnqmhb@(Cpmcod z0ulLj9PY<@*&fT+q@MDbk+eA8OV32e6RywizG#~4HHr&TJsXQ1HP6IuPp9b`a=sK- zCak?4py}Tz^%4sc5?!wmsjt*;cQ(arO_a7WUi7dyKqj5RD_)PO@3C*xeerdV zr+WXT(rs!a@AJw;1R+>l#4bqoRW^m4x;87%6v@+v*Co1Wzhx@B+33?Ql{%Oj8N8-( z;r8f>$fOHgEl8ntqz^9JkCAq?T}%>1H;Sg|ovcut_kK3@JVmJQxlp!ltf>s*P|o+` z-%g&wW7$|luRP3PXc2yAD<+ii+`{z&t}a8V-r4cjOy89bU6T#^?xCtT3S;@T)@ zL|qcgq3>Ir5e)}xn^Pt6$F_fniv zy>&p7-TObT2r3|=QX(cuBL+^S8zn|AL%qa*}rK|x8SVIa-eZc0+xK}Kw1%o zw9@VO%txQc_viQflRvn3?sKkl#q0IDt~;gU5RIkW|G5C{~%t)33P`;a%SN9O~(baGafROWCA*7E}5Ux6MFTUPA7uC3^x z3mm5N_;*e&EE|&s4bAY_x-U1F(W|O3e-{F|{Ob#zGlS-|knI=RIE=t?*E|OvOXf{U zoxy65IvkWZf)aUn=az-U$!IzRI3yo2U7EQ(+aY}~oEi$Ts7rVewc_H&>tnqhK_SB- zoSy6JY&PNhrXH`*wc(a>Bl0&SC%tp99>#*BBXjguwBYoncqHPI`1%7&_$B5+D?^ie z;AHmGfQ0o8ItETvj}2E$5@gyud!|PiMV!TaH0n_+zqhMuTH4`&B>FA{h6bGzm%cp) zAc!@FFO3E1XlOc!WMqruRB(F)QgbCiB(!;3BVkAv3O7&7A*MgVTpKjgf^T4E_8vkl z?zE=eW*>@UM*sY5_>Rqa^+=W8s5=L!Bux{=A6wEtxOiiH92FBouO+KAyAu9Z+J{?HD@o|6OZj{rpm!eWJ6*YL&7yLoWY~hrt^hd4xImQs{6!*8aqI(%p-FSw%ea>e*|Du@d>X&TRufjhkgZNcPoNhjXpi z%a(NT*h#aR1-peP%;;JLdY5ono5?%9Ntc0!3Pc5B8b@@kgaLM73Lk0Gf)x}f zkgjXFOsYs(`<3`X?n8ybD-Og)AiEqNk@0vXUXJ`q|P_XOw8K#?2K{&-MHdIgMCnp?{ z&h5fNb!gX)!Rz&z2U194G%V&9>A0xvwF#bc@E_d(>z_L0E%*qZF|L3904jlPTN7z4iNDx6(h zB)8p&SwUa%fQOht6P14-yKkYzS@lQfyd{htWK9f@fH8v|2FRJ(2Iks{I0y^s(J+{( zR=fOW04xJj0zVJG03K#^-$!<)w1ttkvcHhHvmAi8nqU*}he$4VSsokMTEy9iQX7 z@$O<wGWK@)LRIv z-4pS}h7KANs3&UXj{J+n}?|~ z=k9kzhDSznCma>~i;0XB8OJ)Z;;xUA%*-L!OS|1JBAVA`U**RdqMpL-?a59n|s=r2(q_7GpVnO2~3a zdx8EN}_E06>@m$-H6$w$c zPssoHLLj%1PkZ@zXF~r|+twcy%EdQc^YiQ-$Mbl2oyd^%h?_VQ zy$RNYIB<>W2N<-RG?Asnb0_cZTD`sk@COng^wMcj-Qse0~CP|Qd6vm1v4rp$v@tEog*Fu zCN7YR)eGnV#|d!SgA7c10m8qG78$dn3on*RnCvFFcM^`deE7 zdr|$f^b0Ln9B=Ay*NL~BZ+nROJEAl<&5nn4$-$~R1^;?@@M)et65UGA6X5S9JS-NL zY`2e`5s8>}i>jaTTC<0L-IADBJuSUMSRVT_n_Fr({poo;O`%i9Yu$RRxo4xkZt+KZ zeYLCSWJy8?n;7)fE+b3bTNv_aA^mi-4WXhpp*4h7yZ48kZ+U3=APY{2pR&V_-(>mJ zxOy)A5oZbvc!}=6Djg%RQTQnoclyYbIA+^QgRC zEha!9TXvhjshxc)zqcW>UKGsmwufU?UhS|Bs&k))pnC1O658<5a8}nn@X*SVlN#6+s5&x4_fpqr> zFqrS8iNbW|dP+s1j?BY!5bIpI#`LnvKM1y}^n42!758ysH^B3o5j(Oa{8Ej~x67U? zTG58Yy^v3Uku(hBTnig^C(Awe2DdRch5;()y^Q>Hiye>kefX;T@_YdJYQv#IsOqDr zA|O!SB~A_?9(BHtc;546Z87=Qvl|W^aK`7Ota4wGJKFo<;DQ`$_%8;Oc$pG!YuFBv z%i2y_YO6(;3mRC@r3+#>#jK(7$5EPHp2EcnVujXm&)J5Sw-&l>UAKuE@DigU|F0j> zu7z<$RueRRwL`OT7n4IM5ninZz5GKE|0zd3;+Fl`%K(hQVT*WcO$O=W=HF#J1t9CMYIL77Y*~jGY~_J&@0@6}Qmxr#uI7oZIFc8Io?GUC20lx5iKsT0 z-@FPB6|#s6NEZV&^EX>xmgKG5!5dVUBiPkR8LE*CKj1L2n*m07KVCb zZqhP7LDmbMIAQfFJ^Vz8c)fnR!90PZW8SN?0u5WQMlKt)fQVtWI8yngXf6md>TYqL zl(rUe9FGgSC#3_|Y>us6uJlbA6-md21HM^~3i_4YlB)p|6!I}zQM%jj;EV1){$BdSg9y`GO>1waG!q=7!PUY64^;pfiH&|15zb-#jS<>90 zItD+7%g7$rpNVM;92bCTeF48Wj#s#BJJg!BX1mq=5OeJTC~EZr3bzjkb`1#kgS6T` zBAld*!~S63*19C(CAUQu{1IDy34g27YNW zucTi`3+Aolc(`yj2-8uG-LY@%tQB^C1Zp3{1SG=IWb?L!eD_Y(3$<5_BsJWJ<^eTS zo#`W5w~#aL^e#e&Ash|hn`YmlW5n*L!{)Qy7Z!yhYf`622q2Dghxb(tSjMgGEvSUq zy-HfE#}ZkrEkE>PUd4H!&q7un4;WC>CU_i&Zt33n^;z)EOR%xT>TImMu*8_6Q7O+V z(RusI?p=WH)wQDj%eIS6ki@dWX2eYLkg(h5sYcI|P)tQ=9))mb`bIPN>SKkH7HsI9$nd!}d01UtVf@y!w0)4R!#!Ue_yvBl zXmUkN7FjfFO@FHu9I4M=HG7q#4{S(_XjC5^(ZFNU9l;8Uup+yR_A>j z>8xpcg@$poMMC;x1-xn1HFA3&0VrevbK~|-$3}6y1rQqEDnPkZgMd33r}pUL$Z5gD zM*zT4;zFrZf1hSw7}OGQY0AE>02)Zbbc}pv|ITrVuG4upxm5L@QPzXCRD-vOtspJp zQNb>*%P+}0&h+x^ATn5{SIf=J1z=2;@MOl_1N#9dTnXryV_aE=&z`ZZC$j^oN2UwJ z-byV#jww-jEOMvtkSQD&=z1jWW$EAx-j|+x-pgdygwouhp@h%L{xbsC+GRx9j?dlZ zLzR-wev=of<|s58vXdZnSVOSYlHvL*MGf?R&y!CF zQE~f8cPIW%Q~?F7I*A!mWF&-=ZbkTD*!EzZRSosI9@^t1Mm1gu6`d*$O71dmCC zn%k86WG-yI_Nw($TXMW&#RWn*2+-*CTa2~(%|nD)dddB$QrRn_2B;APai}oa*|%z+ zEO_^)e-0Vs)f9~a0HA7Ok*-}3uI^RGL>AW8Uj>JqIx~z|chrzuj30 zH55UZ)&RKWvlyKZQskWFB{vVd?@NGIngp0i{IF%M=w(kfHXl$E-;ov7xetE!9Qn-H zGkM)IkR*js0Z8g0Kr?u7E^J^kcEp=Aou;aY_E>rzRKuL62QIVyFNcEBv`ri-{1XISG4c2+<%8h3%m}sZhEN- zzRJ#93l-c5-T#uupQvweyrVVP-aaQ}g04i_{uyVijNdJOP zrf~X3P-=-i=N6GxkV?AMceG7LqQ)^*Gp7CldC1Gh0F+lJ!BKK@KW2Y$3Ljh0i2Q2r!u;zsd36oM-Deq?tp}f*<(3B8J8jt?gv>g9K zo&k7fE#n9i$k`Cz-<_Ab>LAHCFLYl5s!C$F`5J2S#g)5H_egoIx_5e8q)Ni)5N{^g zQuj~2#ha=^@q@xBcGP5C_M4dmm9zH&Wgidrx+aZBknj6R$SUBDV;KbOSCgVFVSepd z$8`X*EvBE8cz**A0RoTf9AM6heuE*iP3Rb#^lr8vtA-qZv;b2!s}U_}rDH}81Twh| z8otlUcP6@7fIK;0a=ZGm@&2-;_}$a>_AErAC^M}5W+VF1+v^x;Hng z*SPI>)V<4Fhi}deC)r&ij2B8;q;2FA&2m}3jaRC{X6*_%y2w16(qIjYxfwzv7-@yH z;j8qJwWC@8ea{N&5KyQxB4g$cQxvMVO^M$W+eKhwK9jLqp%oCB7MG#1RVCaYg*G|W zdZifn&Y93z*L=81lIw>SR*0v#XVz8HL0Er0qU_W_p<=!#rqFY8xT>nH-;fU5D?2$j zAk{Nhd8qo>g5q6Y!h1_OEm)iPRKG8A1Od0+@8!Gq;J7=TPd@v9X=-GzQrBIWt`*#TE5J&Ph=5tj4sid+z%pXR1Vm9q5)PhcGhR%H{~M9xGW z`O2n>gOKe@ATiAX9_J_ko|w`&_BJ*J8K*m=TbMM!r>#*4*3(mdf+7r-iFiFyt>n7K zD}g(Ljh}F@IJ@@@g$Jcilwa<~xok%nDE89#^g6?aHZfiqFK0`oWbYUbTmY%u0ojHf z=T}JAt$=O%X{!reLCaiq&zq)mqqXW4wmgI73jxr35`^W3Y|JU#=H}+tQ2E}>AIrhs zsZaAEF=5Y1!6wa=uAkepn@^UgUqGvx5|y)tIbV(>k|3#hq+2upZUVt5!O!oilpO)LvDUB87jEXc17gsvzI4`GHAVIbl+;~*Gw1lXc+3qzqaxaZ-hdltcI3a?Zwj=3 zB_n|tFS#y$9&GiwOoTyc8Dxb}NgP&*RhD<{MW?uOTY@teM8h7XV|PX67p$*0^|@@6 zFxACqR%o!0^2<4{tXCmjzqrprTf5Yd)$7BKpJ!{bu@Pi=lYQ<2YoPjHFI(phbVvI< z?#~AV+#J1cqXR7Mfe#Fpt9}JO!{xiT(d- zjQvI?&wb@=a+i!XMDT%~yC!wzMNXkqRCR1k3I2M@ekO=#l`wFi?%60R68oem@24=+ z{9G`GP6;$E9I$d7)xUG~{;OJcH;{vze5?Gr%n5Y6=d6G zC+f0x;2yePo)mkJA+89Q#Mi0ClVuAF+KVqv7ea}6@r9+#v^IJPoNm3cABAu?9fNQw z((8NGdLd^gzmrndYDaawuT>eb$&uP=dhW5##A8fc6p35vrN_EJE*JiHBB=#CkSKd@ z=hsUF%%iR{l!Gxg$H@J68{C|5ln7^yY2(=gl) z<4~=CCHpJJg*OSwa=XTpM|IWrWQUXM21dtvZ8p!dA>sOT^U%%Gn5vVe#2as=xHQjK z_YvNW$~=JvB-}anS{*^W2e@R#$>A{ zd0vnlxdAwuG(jt^iG@ye1$LfvOzXn5A*nWHXV&VA#uUppP6Joou63UzTZB@|V@RJm zl%p)tk${`MV|2Os?MYg;*AAzi_!-Fdnfy){#9@x^k`Z5Su*s`Q= zFR(B`liwAj`$Z?)90MPeFs{tBiDliDT(4S4EQ7=f63K>>BJjbBPa?8fV^2{`O!hSh zz%Oic(_bxHSOsq89s@o?_~s46%ipE|tOaWXUeW#$M1cpE zfIQsn2PyXMQq|W>-!hMSWLt}Z;dVW_%xbXanT?=%9FB>7V45FEam2r#N35O zL6J7!!bB&B9xa%mg~&>pn+hyf3fe3?xdAY{FwazJ%zo1K&jFYcmLNIU6Gf6eA2bju zjDZNR@o~`RreRU->h~{(MUp=F2V~dII#NW!ne}_?wtNlLNceyUpi4QQ6j>FB&GIQC zN=G=1It0@(WNOUGFaRP+{TvVcN0{yT*F^!z<%vg6^YjtO(w0xW2)JS#nZR|9z_gb3 z0z(#}M+JZsBBl89{NGa~|i{fiK__x$BD_v155^Jd>cz@ zsCTX8#)bx2NBlE>=ce2wBLHYP7eT3v;DWpTwQhj2E0j|iJ2&daUY*RFMi8e`y>E`Z zYm1R6P5kY#Y`_Y8-D#>TXT|I1g)a>r2jUr%?|^nmKrRPt&)QVzAG!D?JmrI==DBjf zVgTL$Kqz2K8^-ep=yZ{_5Tszz%-w8XnC6eM%_pJmext2K5p6`;6OfT=!7Oj-LKyPE z0b@m(s_sZKy+KFv4Rbu!R{dV%`7UnK$_<3E1M2}UJnqBQ|IHxubkhYJZEjGHpM6Ql zOep%MFqyac@_3TWg+`LD9HW}yh@IgDDi_Bmkjlu3Kxja^5T}>yaXm9AFpycC)wVY%q~P@sx$7)s&(8PVmtyMm%6`9Jq|(4SRx@g zsOmnN-Z|1ZFiVno{&)}=YXGlh_bzwyTm}{XyIlwZQ*JSH-d;KQn%KL~Ir8!m|p|4bE9ggvu$8~*#~KI3OG(Ywzm!Z z&>0`thXbN+?e_ZJk?END^^Zl{Snw8M-YLe1$o+I)TRYjh zbW+uGqpb|9kh4gDAd4H;*d~dFL?7C$XMuXpHd&d;A#>{aZ6(7i!rk; zr87J{T^|(09;McU%;4qqSTl1e0ydYzL5dbm&4l(md0}_Msoq5_nM(lWs)6>qsqNe_ z+Cwy|a=uvuzOChX)O9XEgS7Lxu_N44X zj0+opMxJnkqhpvX*c=XdTtlYK(esn87)j|&5ki+PP8kwRhqHWhzwF#F6{R$#sIA^` zN}avEijnR_CYZKl<8?^2j*zAf@wkHAZcSTOn^IECw$GSXI=5 zfA7Wu9``EJp&Xx-c~4n`UX7w#eCgb zq)5WCw9NvQYy-nuPAPsnW$U^3D@O!th3rC4x${U(h7)wt4d<^Z!IUBhI7lU>zCGWq zR65GWcKwMe_)-OxasJIA8WKGsRisCfH^6jw~Gkf9Yuqy~#eW0(9J$3W+Rd}~9iFr_enc2hG7TgfU4-ijOrC`;S z8?E5`a|{iuAM?BN)DT25*GOOCqVKf1b8At0RqE(gv?IA?y{8>k#C*YyY!tAtU0-nh z1(_`|2+}{F$WZ_l0_3}p?)8+TwjMxDE-L)j+<^vltR0Q`+p|jWHsSI&guUHbm7Yj0 z$$(TAG~)1lev+o`*P{#44AyV*bpc%|!q^vVbgkUt6>+V!8H*XbCS+%#;?ci$mFJm0 z(_F4TsSNdF_tK_1dVrW@1c?0s6l`i$ntEiI8?BGWR^2((ArCf2ssB~w%fkzLtjx)0 z@wm63*D`|A%P3ioeeBH)Pt(bJXq4$$e?cD}L9mJkUjpZ++7kO`N8XET!x&yAi|oFg zDB7j2$i(`D9E(4P%W0lhR|rAkP;iqgZXE^VYP$E5OgyIaAXWdu@Xek(kv@2Z2kZ_aIhW5y37E1{v%}~q|xBMrurRyZ0Cj<)~ zfQ5@&JDKP9GEUjl))Gz8*{WK!qxkOhDvzq{-Y@?MO+M(S>F4VXlw&>|DKc-u9#p zMlb5s$UKn}8bhyE&g`B9Dl-Kyw%p^&~O@xdy;vEy(dyFC- zp^d#f!#R0u!D7CMP$d$L@E&68*tVTFR)Kv5?2LubbG7qbEF>crrHv6{uUZ0LV(|h{ za0>htvL}(Q-+W-~o~j|*;IffeHr9Fj$Nk~N{qGO z-EIZ3K8_|c1S-jt%+ftZG^MMq;7vxUA^-Ke(=f4HyG0j&L zOkZ9;+XBQ8bBqmURn>q(_x@1amo)kAJQb--EAT(MLt#3vhOb@sgq$3Ob$F z_Xuj~74BmJET&FoS!aCvy-%x&SFk6dRq91fl;g0N3lc0Pr%9tJ3e~)eq-1{w1cdL~j>kDV!Jz?ri(^3s%IgV1ad) zyg4=I${J@Kx824u=C{qZDi;iGPQMzf@22bTaVp!GOWJv#6w+BO*iqm_U7dm5HnK3@ zD0z7Z0094PJ4L&9AmY!yN3CXJa0DK2C-P8XuKTC<$g%UP(?1CgPmenBo*KYVy_lNm zY@xT~G9r;(3qB?KKa#wqO3HUog9$s&&wBy|F}HLCqXYr#@3{*9C^hE2M?~8ih)m~g zCa;g>JWKW8BaUSc%y9U7q5W2NY8HQr)hD;_PNJ$+_v+U7^8CA9a~5B3l9P+Pa{RbD znQuhRpduaxobO^F?R+dId1-ST?Lq)xvhd(ekmZ+Iuzp$W|3ZBE+{*B0RrXB zPL@>=@>`v#44{}7x{P#s8pqzNEm%s(>@pc=NN4zu?&zjo^Y0`t=e-C-OJG*AWC?GZb+*c;KsbYVHY~f2P zE&(!X(l&de-e@M-!WqLX_^FwV`c#X4bp3^#u8qYynI)xh6j64}x5sGrmgn2t5`5FG z=?(Z(>@dUZ)%ie-hPN5agKfPGWDjb8nAHCf1s7w!=3#}EhJ6}%lUpZaDlH{YJvTkQ zNJdrmp7AuL>*^+^@Sn5WR(TxC4k)5i+?CC}wAZ^UhS*#Fi*U5v4$|d=hp8`DC}wOg zq-1fIzrM-#!}T$r)d=e*a{i8KFU=rZh@iT_HLx!p7+ACsN;MWu>5|Mgq zvP$e$Qmy@Ly6G7gjPdBp2uSSR>E)Hxp4_5ysUq#AvYPMzak^$n>Cni6%l zGA7~|>jYA}i^B1198%mJd7c+JeAXW*7#nPJovF>GRt-wbPVIaxuEP~+oX5Yq?5pr< z66NH#6^_Z4`5+O;n_Ig#qQ@$AY0_Nfk3NUc?b}?dVf3|s40I* zLnuu+EP`F7*X4O!1*EJE*{l=!I|dN(m@R>vUCE@7qooaGa^;B&BUe!N?b(p$Mk|YO z<18_WN>tsVaH7_=gqyv|6GgOfe%1h+^1qE-Wjk@2bguZ_`d;&)5PARD60WeYs3rWw zz}g^3G=ub2w%$~g#w}^ny@O{YonDBhshqXtnf}RHntx9@Dmz=X9a)pt)b=F17i0I; zAgWN1;wdOHE*EE+tFo|y7K*Ts!bkrlATFUZC8pxZDHpORZVTgXQ0`EKcjk5u=Pa+- zdD%-o0pW~<-mj0hZRp`)v%TrAABY)KMghWo!`$Dz*9NV7*p+@>Ty*_8s>bY<_H82h zsEtna#7N*o?Z#IbO_56*VJR9vdyOumH*4)!6M6{P6pen#CBxP_c8ed@o(JwnrKtOx z+P!cSa$31nIEW0r>w5vX2)71IUkq-r$>Pypz5D-iWfXz71u5MZ=Csi>8LhBoO}3vM zKhh(m%i2Q!?s+HiYEwv<^;EC9t0%jYQDFsbwado$)rKeFi(Eq9Cer!i>^~GG(s-U} zEH@Q|bbjZp%ke~~(-&6^X~Wp!cZQUrmvi<;;mn*{8>3F+u2INq5t4*&U%&n=c6-S~ zqspZw9n<@^RH;*=qGf{{wsvLejy0eyQwV9gKC`3O{GmQO4mX7^IHNg$gT~T|mRz4pIc8QH3nkAMObTklVVt zR39~D`w&2G^nlI2P3c^CjIF6Sgt8RSf-ybAKEqtAl4wHTQis)Kc$W(Xt9KAE7^6JR ztDDU>@$H&L-wT31&)VCKKl5svTqQAc?HPmx``2ii!u(YOCJwv~&(2yDd#`m^Ol>b% z;G&H8@(-syo1YWeRW3Emo!RDFChGFLs`K5Z(!O7ZWDqKFSe!y`sv4Voh00iCys}l@ zh39D)o*UW@;xam0K0ts>fBrW@Lh0fFogu%d{(R89FluPMq)og&W>GDgfG?WAP$Yb9 zCwBIZ%XBV}i`Kx5?k&0V zG9>%q`zYcFaXb}h%PB2Te+cxyTY65F75!2pSlGL>4$U~LGu135+~rl{x^+Y2S;~!{ zLw>#!*O|K8wo;;wY@Tu-$+i;NzD~NBiht5D3`h?b(fE2#!iPG8#Q|){iu3+-s-#oYaD$1{$fX5GjIn>2zc>KN7`75@WR!}ben6!)EvwEb3kM@VDkA0NMDVhmZ`6C3qKNa zjCU3ezo_uJoU|j@WhYa4;qDrUobv#|3AtA)JzUlcP(O|6*FQIMY{eEGf2vaZ<+%MU*~?EHM3*>^Hv_U@0u<|uVvUP!%IwypS;6lP3O5kH^QpGJTpOCHUmABT=4#sf1ja-X8J9$ zgYCJFcZf;4-E>290fq52o`Jb3>wSV7^Y z1&=<$$dh`d4w}O4szBbwh>GF9W#tp><-)u=XttR!Tz{+Q{p^~p!>4t{3-YZ7hr-&OY?=fbG^dDJ57>f9q_j ze}mXhkkI<&MWQyc6K(0xqM3RueP(0D{ESF#<+0;Qdd7AAhLX+h7nr+~6%;H7lut)ix zWb)L9`|eR9f?ZAGrsGNy*m92T4K@Lg^Dv(yWk1~%IM;Gr7e)$n7A+*dT8vrewvAqj z6>|8h^+A$*S`H#~oz|12Z$pUY|>Q_g4E zz=Nzc@lUS!w=xGrsx2+p{7MB+q_TdQP{|q6#KU=a1(ovYsvueG>>{30)r4XL+8(Xp zK{FE`IJ7YC4;47>0kR!UK7@h8I9=}m6k7LGOyCnaDDA@MOf%r zRsOJ8^MK#A=+zx%R#b{ez4ClY2SAFOgXEkb-|jf-Qwa)=XE=e|6t2!VQ8+L7T_0Tg zzt`zbKmKnjHfv5JPNd_A6umW`sLkrRmEtFLJIT4U?f4{DNvNm(^t1jX)kcvS)ddfDR4SY0m)>0CmE`5FbLT?5(YV_(;I`5b@u>iT^H^-Va zVF|_(@I_cxlk49B|F06UDiSmSw$P4?PTf3e1(X%+>e_-Stou6b*TXPhdS+5+3q36h zyM5Ui1nj4)k|t z9fkBY7=ddwl7kJr0TRXkS{y*h9#(|2UtSgK<6*MDoQydkQFm6;lgYH!IdN4r>|u-j zE0M`i-5DiW$^A}A3xh`zMal)D=%$*-S<75CTtU}_x| zlU0m3=0QAWna;V)mUMf5u`W-hBrI&@^iGXi#6;R+kGQ)BmGK8LdC(N)?3=p@`q8&z zY|5L**g?Z5!=DfU*7-jICMFQ~$%E7FFO}B91WpywUD0Xrw<1G}#hhFWwf=ys~) z`?wk|>j-V;)aZAm)+tGg0CYDGK~ZGAA{BZ;y3HH1ZhZR2Z2uPSCU2spUNH}+6`lU;e$P!ufp?59E@f#s#Qg;pIBBf3go-EJ_=AkU^fq3kMAj)27T#QL!mz>6j9MLaN&>AzhO=2%d*Xhc=tme`cdge-`5#Gg2|f zo8ho?l}K0j>Soi+U^ddhKNSdYNci6khw`8~8<4fjn_x8YO~`7OM@H6cy6pozJvlbEH?dHUhB`3{{Fz*~=W;|wZ zb60fC|0hZD^aD@EodN{43j!`0xZSRkp&3T&M$h~Ib1&NKvLcT|2@b0{$D})=RB43kJB-s zJ`i~YDiyZDxDUEpT?JgzhtHK4>{?V{SxQZcv;Ej0bI?Wb?oUlZF?SF@Z_vU#92lWi z(7)7N#DyyT@?P`a~%UD%-=F+%6KhG=NC5s(?gCS zi79}99$P8;#{CW0KF@R9v1S#R?-yTg6x@^-aQ!}iml_v;t2SjDvySpPl=h==O1$z96 zY^+^}X2zEK+m81Du15c5b^n5Ij~WGlgX`AQgsd#OU!)BqzVCp6Qop*~&-0RMXLX34 zTEvUtN@pGcXcnXh*&ZpJ4}7TBfnBbWQ18rti|!c|xi-o9JO<1)A&^yz*Fr4i zoY}xX3IVu@?=QHC(y#vOtsO*Pl6!TEDbj*p0|3Ckb^!F96)6&AAqT95;W^lGZLs6{ zvi<*bx(jywGO}91E+vas+o*b@IkXrU+wj}Y>W&XH`k| zl~8wYYuS7UC$N+-S?5Q6Q85@W;A~dy!yYpgDck{!CAtG<1uFj=$v|d*(k1^Log38! zaNz4;Y}4Gqm?Wpc1tCr#wmi%XMmLB7VU(JYsnALKeA!nI{{jjBZqt7cw#zO|m}F$L zq`({v+Qq3GBlK0IDT3+eT4U%d%;2a&v(2w`f^yVU&cGUhE+7M=lLjs~d5y%pZDdL$ zs0DRs$&)6~%(ip^K1kQ@i0$g{*HLQzyp$F}??@^Nq<$XIO|XAk;we4U5h&0&XOl08 zLa5h>#3H;MJXoW^4gaO<0#7Ld1L@dmlff_$iy;900J^nA8)$|9YN^i;1TFQHb*z(C z64GT}j_4~MgshJ9Id+Kyti>_`GfGVoBAx`h4=QVWi~n<1%wG_&3nx~#=wuWxfSz#q zJHv@_pcf;rQ1FL+=^ef+0gjYuEx)2W5D6UrGcW?&s9*7E8>8G-SPBxRr}99#9`vU2 zs$k147#d^}2KGY2Oa4@E&A(EeI0(vu8zFzyj2U3w#H#;41>6JeLlMAU{>+vHgOGl=NZ3&9C8Ti+TyKO2+e(srLb?K}%<0fI73w7| z_zwloM*bm?b$A0&)&KXOC-j0?zEQsuN5GtF-gT>IQ0~Y-KCuc<#;U=l=;MYhy}@v@ ze?8;>yYUx)-m4yz$t78}MJ=2%75q6rz>z8%0Su9nwt?(23W%?F=Ho9J?g1tW zF>Kai!%LQ!x9Sks_e+)wSEA)30|)1cY%+4acNz#xcfqAyJ4=+!z%AMS!8H}bzY zN!HrH^hx_KNOFw|8!tjuqf_=z0yfnn!DYU9j7kgkvj$A28BESexpdD?bpu0w;1pbb z`z**f8hu6(*-lTC&j20Lx&ML>JE)50dpDl|wc+dzZ7}}+AoOho5cGR9J3o6Hrcxy) zb(!Eb=hALob>OUJ&>`8A9%PUUNBa6SQv7ENgl>EJX}nrv^~7TjCIazbcR4ihJ$`ev zwl~w?`z8ab^iWsuWHi{|h|hin-(O;ZP=bTwcig26r8XPK^mX{}rEGfV92l@SUU>}3 z>E0FeUWlD4b5|YX+S_1JayXQWsc`ghd6N=(+;f5D*_E)=w3vb-IEZ);OGlCLD7`?F z;c?a>1XekUP3Hg<#Pc-XmCp?_I`u)Or;Jxc8xBBFOT$0TeOAF(e_|JIG zfTDVdC0?r;J9kHa@D!ZFF#c|wOLy4yI)!B$a7GgHm1Qmtn5B2n`?zTyN&NR2QxRZ_~ zx<{=0CHIf&%wt)#nPJScK6+v2vZtopHUv1#uf_c{ItFA|Hbb{d{0+{@P^Nf|oc0~5 z7G7k+W6z$xdGC1h|0C?X1F8PM|I6M)QbbBBSrwO%6;k%Nmt<8!T|29c3Q>t7E!Rw0 z*DA83G?chTT$>VEg(7?X&g*g|?>^t(fBt!0uh(;&@i>q3I6yuSymqe9v+qR8Kxcv# zw5uQ9NKrW^$ex2>+z5 z+aGn~oZ3LFw_UO8&CHyg<*_yH`o-mj)W7y^h|>Q!fWAq86&A=5w3_#SAE{tZ`;uhmtqpo! z!R>5jO_A6mH};vm3jjj#erCUK)MY@`AeF&Rr7Ict9B8E8rXkjRxY6Deo$8x>_M^%C z{!cg1#Gk>g26YEB_+@R2KODK3YSUI;&pfoRPA@iXfBi-QcJ{V@gSZHUb3Cz7WQgL? zjMX%AL0E#ETbg@sonGC8Rr_N4xF=~}Y{=b6K2sX$2v+D;^xEVUE6P+PXh zo6lX$+f`qhf}XvOZN+6D$gXl_jnPfGHX8EHrF!?p9Jh;Nm>l13UW_4UNTGDD3SmsT zP}y1`c`J~%*rhi=KKy`}hgOFuA0x4bb4=syM7rKtva|`q{_gWIx?W%q65u$l|{Jz9G`MbLh>EG7>u7^(ZH9IJW znfXFLnG^KnN{#!(X}7Fjqi*hem>iXED@?@1!Xw2Ab4)(=qs0ANlM~Ti%IM_2y9h;^ zVmk;xy)1^@3pZ@+HJ>?le8}v|u6W?`-8;4b`eTmZNYod$fK*ERDLIopB3)!QHQkHnT z7#|<(fzah^%^aAeGcbvq0V+#^L-0V1bM(9M5GjFgNkcb#Baa(ty7I|NKOQ8`kA|t` z3+X8ACfa|#5ZwLNsh=MIv+!yTo?SL-CMi)N*`DZKk^ z{-E=h9bq8950-*&@~i)-M~_HKX(zs`t2jT6v4|mbpAv+KY7io-{332`>&o6kfBUq- z4f->brn3pm6?I!L-hJlrsXyyZ`EyPge$W=8pLa6h@s-Y?#}MJA18sH-^yX^9`e~ zt@2pYS4Rf=E_Sj3Xi*IG{+q*|UM@&{)XPm1l%r>hAm@bhc~lRPf4cqYA*hoYVZz|V z*FwM3dJ#rk=aZDAG@(Jk{w+Qn7hX zFVgQ`qu0)j8uoQ7M(zBz^r2Ug57q4$=Q0cwRlb&$G`n{P^jpHm3X2ha^d2_zNmD}A zuk7>o302M|PaqGqAf__ou|?o>C3mGthytHD;Z|$kC;cF^B} zO{F^e&yP;QHEq*4J)o@=)vv;IPdnGpx-#L&kX`KM6sfk07Ze20NngDy-t}fY*xq8A zU)f`bl9ROPymF0R^psb|-N}=R4d<6_T=@*xxCNFgf!Y}8_L;cyQj1E~kPwuX+r%Yd zH!!jjc=5Mis{FXI_rP{Z$aTE|f825fFEe_7r+%C_HtvK{cUFgf!9e2IPNopV&Spz3 zcpn_Tmne&7F17d|TwAiqS^AsCfCt|qp)^GW>|Onj%3;~GSmWDnAN5hwTdYs?xPAnw zJBSS7ep#Do@9(|%uESn%2xdNi&mlKoJi+O|=`e|3cuw{9Kg$7D?@_u^+@+lrcXWwn z3m>jjpf_YTstFAB?D^jPa%Y;)9z|!F=#0j4bx)^!|fqPbyF)wbpo!8DwC+l3$d+cR__#&$RdLejUN*JIUDJi-2g z)W;0kKH+VS=vN2~EY{HrGSBP}LBD!r%&uja;y=*lVw_lSJg2zBuV&h|Q^PXi@vmR| z>h}E@9lPdKvcq)pBmG6zMH3A#9W2tg;)|F>fa!4OyBu2?yk#L(-rXG3oluQ-1#MF4 zD8X0YijvLBm;u5*>6h z9LP+%;8lM^SYnL7(!={8?Q)=64Lu{8C}W)9y6#)ig6O!Mk8E^CZ2qU&z+0Wk0lB&$ z=;PShy~D7CLS9sYd*NPkipXPkRbz$*TXxQ<#4huF+9kKWq$(P0B~ol|mCIRS8qUTx zU7fghZTH#NmcP)qQSbL`oQ-avOTn?XkEbBQ4JkIhsJ!sf^PofJ?__1!9*LciI>%ef zcoh;s{{aw9aHXudT1%&=Jv;%?8nAf>3Qy}kHQXtzVuQX~`T0mPM*l|6ePQ#snrnCH zr3zWK@H)E^t5vrw{(PipE+#x)6J2wBuI<^c8|x*oTDN1*;=)b*8t|cktD=RDa^gmU ziLUbAv*v7Q?Zxr;;Q>mv_`6{=zIJwzEkmFEZ zPm}RLA3zybM}O-mbXZr$X=JmThX^ZKiNA}7-%3QzgQ*g*#DNEp=$bPh9U2z zOZ8zbwTr2o{mkwnHFcJE9laVS9#oVkm*Wd?*VqM;Lr29XvflUY6uZ>9{ykb9xmeGR z-$p17R-1=AZMC0dfy!Iym$qgu&m`QK7_pt__PKXmaB`k$`Ce<*rFppIQo4XN&*p;~ zFTX4HtYL45o0s0F^nJ$qfn%n3;$fsUv}U#w>0hK{$B zC~eO0-Lg`pge$3&tl17)jd6=ImM()SbrEu39%?KfAy5pGg02PV7L>8nE3i+&RdBID zuZ0d;LUvdTBP{4&%v1a(&+cdBpYzir9S`wsNLH`nF@=61kKnu3;nJd%MPDr{6WDEs z-u6EgLS2h-CN{j7*f3csDN4Nf{q64?*BW+t4#p_f9KDB+bWw1@yLOx2#^Ag9>f22} zC*16tOLIUE-F?!)aJxnnopd7)82W?(naUxEl|05*+DNjl!voq}VutN4ENpd+;Y zULSOAtn-?xHc{U2u9hwUIzgfgzo8{y7_i)&doz_@(p1Wo?+Ga)SyY<4^TG`lb*4HG zpJT+5HnTO7}+W+;NO=8k%__yiO_e_x&O_mTB!ZVQRNRAzAd)#&C< zwlD|{I}s>4#6iHLPxyMUjNn^!^P-yxqZAnZL#OzJvioLF#)e5>9)9al&n>iVkiuZd zN3hA*T?ZuIpK!}3N=`?Me%4FuD6cuy%DsC!uHMY`@b>L-`!e75q#(iBrtkX15FF4+ z9B?{A;(qv+ZI^gv25+HrHcp<}W&mNCdgTb4>FIS#?wXgEhS*+qI^$f+dNLP)PulM1 zKr;(U?#J2NhtFLgh~S#k^rv?Q`@a*mY%rS^j!iMCH|TC=8T;b);hVMa0*-Op-iN4; zxVPDc-KMM$xx>!*{&E@aldRGC22IFeR8kRuG~KNYf<%D_$oXl*mz0Iyi9lkx3R1*wUBDK>P%wfT`4rH~N4Q^Mwh=M?vM z-8^mnNtiq4;IAedv-=G)?*uXqamVmGeyaf7YUl91fgpX9HNBz5yRxcp8Ii^?y*FC4 z9DkwraiBz~QWNzRlDwd@09qJ7(iMVS~=)%v$0aRRbi z7b`eQ)7>IvB|A_jtc&zL}zjdzFF(I@X`)bOZ%$y0I$8%i%U7Mts< zp@^d{4#FoI><1}ljJHW#6Fi0{WWC&Rto;LvxyGG?r8!~Sa^9X3uK1`|Xa-6y<-21g zN(q*n_4ehvRkzg)G$u4OleYIO8c1t#s;J8g_HEZcTl3!DAs*ayJ~mE+_xn|R%Z7w5 z(OQrqB(?b6*I3WM(Z9)vT;TqOx7ky6>-sn1#^Sq!JLZKp+*Jvpbea# z#|p1yCZ(p@IMocGn4o`KeUu-Pxjy|#$|T_(+8I+p5k7CDp*zr(0u}NOdZq!+`*E+R zFLQpoY%19G?o^!jx4BfqzM7YbXZP%X%Q53B;V0%^(=0!>o!7)YH)gsbt<{(B zvWI`LLE^<&!{~$V3y;u}ku2|X`*TVT*_vIe(kJSne2+#y23Su9JV6dJaKgDyxX)_L zGd8_V1dxagJPD&Q0J#e1SPGwA#8Z{R3$v|S&4Z;ls$*_u^}AIvqR+jA<T zQ{ygQBX2)?_+g4iXLLkd)XQr@XS4BJLhEVwPkS9z|8-h$xWEl(O0;j9YH(%XBf#8r6}WhhiO$wdz)YYOvg&qI=Gcpm*Kyw=GlFN86CF0wbFoZgi27a_l$@F}uj zL&+WV?|aw!7G*^rULVco7m{tIKDPITEgwG8n_70)|@@Wxe#`1 zRD^NxFcCdjY5&5fO1o5mLyg|O$J`OG}PgvKUlyN7h7Y9OML*`_Tjo zJ)S#uvijQMwciOgLMae3A1C9x5M}Tx6n2#}AIRkyuhqU`oa7Yao~fpPaX<0V!Y9sL zneHj9g6+)`!Bm?UCp%(wt*h%T2hS99G`#x#US9yEb>?n%Q{;PlMYY`w%59(+d znA9AKd6~4GAGgfg+}dP4m?@K~x#{=duZq4lk_V6y?zpZtJsX_aGQw3 zZSq!jIGQ=21l5HNpub+Z9%9e0vR?qHu@vVxETDuc)Pzc(a)y#E8m}R9wb)Fyq36g^ ze586q&**#-<+gpM`p{VMF*Ci@c#QQqMM>j4fb=cyjnN}bd%le?7uD)+V2jG~H5k&M z5ctd?@YN{j$xhPx^B&uI*ksejS?u{mbQm{WnVl2wh9g@NEF(jCe(qauNyv*a$-r+J zFQO=U)~^%Km5mZa6Y_!%v~Tl>$5~KK1}2`rRKz-`?q&0U^}dG6yV|+f?Q?(7}Q@W*HHGw1C!>?5C>p>o+lBa@Mg)V>x+KdRzTfP>glW z5?25ci)WgWd)wJmxwMt+z@t#r!zbaCw;88-*!93%LFeiY<0G>D>kWrO?|QKW0o!q0ohtJhz;5J{21t4!1? z-FYyN&Mh7n+xHXYjd=zoESsQWT0t4A5Phv|2~=EL5ODG9E4&XorbdxR!q5+T_qQfrE%%_`Hg60HFW>j--bk;?=kONc7Ggd(U> z`V~|8dPc;hG=cnB)(7S?_o39JtSl>v61MB?}XRwtqOBfeG~WL>XC#`P4NS`Q@(||E+wdFpvy?X5(o3C>oA_#z6C}J zZSG1N*v#>VgaY#Ejn}aYk38`MBk~S%`T2nHbUO0{r?u)6+JO0OA}7{3p-0OtHh1$br2{d-vFuv19DuoQZ1AUJ28 zuT9~hLtwr0{#ahptV6_R0z*XLSt*9#>-hUeWHIUz-U3)TCGk=_n_%%DwCA4lU1&dd za~HTsWYZxALI%Dr(MpidmNYs=T?E8gP$g?LHq+j4(bTZfI4bY5_Xj`});~X0_yn4h z%jOGn`W(+?0vkRg6q;qoX%V^>32V;Wz@mQCw-+JRFtaQL9 zDmw*Ow&Ac0PL=rS(bot!R%(@GAZe5iGOP$8u%V+2OZz zxL^9w(04moO!SB9!7=4Pd4nuYpwIUJ74IrU_J3&pK;sC8?YIQvc-n@Ot{8gHMuL#u zvc*cm<$~!Q9)3>|A+5a?N*SQ7OQyMKza>pWSnA+Ks3=g5d+Z>PXH`TfxOIN~+ojU} zM_7-W^KAE=v{6xMVEKeLaKKpxU1joF-U6!qQy^OneCzP$24j1^%*2}KUKrQPEVU6z z8^X=Fl284V%%DDp%tvrgL76i~@7<3Oeyjw56n*@+!uj%(6EUC(!t1DQ!-JH9xm_!$fI#iRyDF$RxDh>;$7AH$t9%0_ES`T^#naORuMBT8WT26D zeQ7Su59!QZh1K{mGx0Q9+0$6vurPi!22q6pFPTapv2St5g+>?$O;*OSQ@(CKn zOR#@8C%)*RX^cMc!-@um=Co0L+DPF5!swKzY((9yT@W(tee(HU-7ExYuabT$KJ_bh z0)()%YKpUM6f>b%u*7?U*H4`_we3AO`kDf){a*pj>OP+1emeZ=o-nZrrgurt9zRD- zb{EG`c6OQbqFTGfMSg;WoBv-#V3=G0g1Zvz$Nf4N9hLR1TITVe3C$B>U|C3aLlC+a zs3sM*9+W){&uw3&cKkQ%DieFeyMw+eaNAU2kuID9(yHCT9${^dla-#O|ItdX1-S&+ zL()kA7RZuPItWyk#Um_lL2lwKb^=sfuKRXohp4r^F-O*(x#g!>Ee=`x?>W{x06427 z(*+^^eP+rAdB_q^(?4sM-_uUeR;XZ<%CYw@_zup6!~{5DpxrgYpgVr*Qp_{(Et|;= zCBU5{Pp}aDx-_3^Z>G&<5Y_DldE5GH>15AvA1yJugI^?3Oynnc25NH0!`e^;&fY9^ z6KZ7@RpX@NMg!atJ_dJoS7E#$(f)^vFRGF$?MA4Ol`C?L|Dsv0@{oA!#7l~pA{S(g z-(-u}uoN6&)3-q>KAQNJXqQx*k;VssT1~ATbsBLsl#lLb_x#&Y{B?{upbd9v>E^#Y8Z3fN#7HJkd1l zuw&ByX9*gigY9?Yx~yB{+^uZO$6qwHH#-g6Y5$LBA*ftx-Gx7nB)~9}1u#fQ>Y;`Z zH*}E&gepn*2Yvw2HDcnA&rx|t(f47l=e4dcfZehL-fH4yM#PjXvA+jP z!e3Z6?uR*`41bRC4Vc)yGE>heo`VkXx0k!|)3c>tfrr>DrW{vfniJD*4y0MqpadY@ zKGRA9N$R|VvM%=w;Qd&S!%utO1&Tq^UiROu{xp? zSntAkn2mD^xsSi4&!0FnTe)@f$jHp$Oz7(bZ{Z2yid_TZ@p+Vp8)IL#Yvib(qbDak z{@BJtYV)nE6_pI~Fj&)>Y!X-81ekc>iBDlQB&DD>_aZ|mYrKex_s`?Oab|s?)8gLl zoku&x#=xk&hP$~EajZy&Ljbo&${&U**2NCogZ)}c6HWl+=XtDkSaa>a2Ty|}q}qqy zZOMk_b9lN(+8+k5g?gYZf7B&GVVkafpIKut*2^!?DE2-BHMsoMwwSE)45Ir-?SUWD z>!@vF`Az-%0Z7#aPrgCFpeBYK`W$**pumdS%ibAGF;JwWg25E8B#NZI1c4OX*Q6Y# zg3{vmh??+{83t)L%kVteJ1^2}t>6mk{EsX6b%~r`lQ&*i`4Q%QPr_tIG7DBTrFV$4 zX_Rkg*4_zNv#(@B3}+dX?Uu%Mk=~sSgr(Mgr4$T|YvHj^hrST2!Mr2x>5jya!H>(pv7 zycf)ua6wR-Gt6>mx>w&<*&zZ&EUSoBCh!?W%(5yuciL0aFjXtyR4_Z4*ZMdqx35~Z zJ3MvcHclQ7USKMb#phnN_$%`eufPdxX2aV&Z^As7_$?K_k}kne`?!PBQ59*y$YH-g zKJ05&?ro4BDzHx?|7jI#69z6E>BXf$sup<5cbVwH^uv~s-9E{7A(UZw|8V%HNd-ML zE^z~*rQ4U+`#`GAW0~&-bh_h3+L%EbQ5^u3Pp*oj{>&U)$_aQ>h`(QsdAISyZ;?QW zuw$rq4kKJTPN%&_butjYkEB-&Fvn;BjTi8}91x-NHoNCa|5`|s(vhd1ek4!qK43w83RWBe| zJWly<0^JOW-B7CBJ(-;Lr@S$Wlo$y`BNgKVQP|8vIocmW@B=9}_JCpi_qs0QI?M@J zoved~k>f4h@*i;I8*2~UeyAe5lDX1joyi?85EBobkO`}SU8sT##c6G;0vxP_F13Tv zmp|TaQiB7)_%tZv>@K4HA!uzG&fdSC_#E!(ijc}*!2^#faieF4U_WCCQqNtA$aj&% z_-$HFq4xXnO@hZLgA^SRjw1~oC3gbRfF`bjXJC01x1t4?5Sk$NiaRf%1zkVZx{kOt zS|1iW9Vn~)a8{96m6Qzuh+w}FlLLz5S&IGJhoJ&S@^*0I-hc#u?%kr4FB}h7^`rdN zzY~f7m%d()R{ zIkTTduk*b%O?Z}EH_r}w(M!%!is<&x#&_`C+TVcg3?K3bKspD}?iTskdEnV!5)a&3K9cYBR_IG4wL(Q;J3X`=v9j*^Wk}E*4{s8@a;Nn8Cb?~4b?T!2Ces79)we5fT-d+!vR=!p-Pb#!g6zWsZ+)%@@| zuts}9Xw#UC^^)eOzK6mv)nO>a>;a|SSa(!XR9>E39UIjFq}?j&FJ62L60`o0BHHZ8$mlg&H*E;9kaqsyh$k8V3Sx$O|XL826r` z_IlngYbWNm&5LF^o{B`f=FfK96wW@RO)Y+|2)qB4+z0!Qn@Q^HJcggHObJ^*rrBiT znlyW*m3+Un@B`dtzAy@j{rM=Mrp$SPFd}?{19}>%0Qza(0D&zT>H`En2z#sf)_}Zg z1!rWEQL#r1)WPH?K)_lAFEez+Icrf*mIgw3Xj@x0au^uk_Lqs$fMLa{Hrt^Otll-| z`=xv2FfSXav`F|Vj9Cxj5_G}Ns*($L@LTjqa~>TD9QQSV27k~~y`GhHR;2iut;dK} zKxGt+BD>kZcE2(nmkW?#wdG#kdms(_X1)U?`y9o90G1FVnF=L}5;0siO)a zMi?d>0uvpY<_BNqHBGozL&g!U#BBA92?284&lbnsXJLc@Wv$mh=F%V$FSHTngu!_| z{-XB-t|JQuI4s!>8e+OAsLOu4u=Z=9v27)8MZy&12Q~u;JCM{ZXFMWX4X$UYi!T7n zP$PXI4`8EXwIAi6Evy+hzu^Ee{&BW4MrFnkN&p4V1u%G^;dScFT+oD>9}aCT(+8k2 z7#V<`oswkdXvkW0>q0H+7HkbDmXbm76-_kwA0#L`T(~yyr;TJ^vH1$3#stZL>Thch zuM9kK4(aTJ5avJqVF_4_VUY9XybLys;f3%16Y0MMdbNPoiwIp?>uHPM`d-yBs_zFB zZn9t;x$c4N1$4KOCWIOhMb3jZf}MrdP@-5e%%59oo7#eQb?9_q?yPrz70|;gyA8mM zkUuV%pm(@n3D;Sg(F_S3Be2OGi!B$%zYPp0!Ox5e2LrSMBTOnf9*myI=k^+8iP41? zC=W-w3PJd#SMWvW$cIeNwF!V-u*LyC;1ZQ#K_0p{QD;aNONgc5I=WEkM2VDMb7Pgy z^C}r`2-LN5)R%1|0z{DJ`VwKN2|V5mt(fxKm&$#>E%+Q{woYZj4~QNvZW-F~oDX5y z$CFdp66CU%n}eei_I+-JS%?+%-$&o1j)UBe_eIRVG|%3?DzYgE>lzUP#j7l$Y7Cs z3Aj4LNn66}(RF&@j`>%)KWyk3G z2+Q&bK`L8e13i{2EAcQ`8tIFdx{DVQ6FvBOQNptr38a?rVM%}pnZFkGB{y)si-SmW z|7sj!Z47Ib3S9A##nj>GmC*kjzEG;4ZWuDyiT^8kgrdZOOdEagc*eTljo9?&;cA#} zNt!wyk%Uvc=vbypdj*UN^_JyO6dXXGwd!mRVbHx)918QKW{n>6+ur?%9lI0cJ%I*w<^d9dvH(9def#$)z#x3ZtZ4_u^H z+)T9(jxZRAB@|>;eGPi~j+_oi#pTP>JUt(^O+uYv`9Eu;NlT>o;5v;WyIxH%&K z=uk$ICl~N9LuDb|DxPEPiZ-{Pti}=s&2_;Lk-WhY+IgwOz!d)X+aVMO7|R8kjUg1* zccHZDk+Jj06@zBKrD#40H$0OQnR{mX08(GX!`cOZg@DFk5(1hIOT#HnuQ=cSE2fEZASPColPzk#!*Gh;wz~^0#29r*Nf8H(8ouz2iuX`ymxsi0HIoHAchD% zycykX7I+Q{g=H?NV+Bq{riQu%sWBDC+y9Ewa`7hiX4+nO zPEW1}9p)djmplBA8fYt~_xf;sy@(%dM?y68svjT;W@dpt0D>z*4Jz&c51<~~wHwdeo^epo^nHnB2ya0%^S1>?lx3u~;_PxncYz2h-_VLx{ZAN2ZL2!dP)6B##`I?xjuFoMGC)c{$H^uI zjsmID3N55V${4u0G}GXbxTm4^cNSkbSRLQTdXWy3gMZ+{+b*pPA_mv6jmEgb%Qe5p zQs6AZxS*A5y&4xIy$A%fAWC8PVCtA0Air|kCEv#F282)rRouGpoRtag)&=rO{D)o2 z34xBOx<~8B155I1Sza|2he%yTu|?ZI;EJeF2{Q?nn<%| z(3AEn9-zdp7XPcQc^#VI|A6{N(o3;DPCL1mJKA?bV>qbsXf0ehk?K_}1HpU(^&g+S z8Xwg!!JGU>=5wCUKObq?tv-YR{OfA>U5<`Nh9c1ap58{2Z32F`c^z!yCh&_JVI6Ng zBZs}EOEr6{c$L)K;8 z2|zbRtf{?*XfK=VRjcukZzL7*x7%1$VfQ&fyP;tO8zX-QBqe`WFro8N4ah!i(lG-t zsJePx+n=BC`7*3oc&CZypvk!hh*jK2-pcp>L9F-5ZdrUIo#akSm9zY$PC1(3#RWeR4mRZ+oS*?-W1!luumBkgHaxj3TRm$vnx^QX)~SQRb{|!v;nvIOFw1< z{D1!}81Qz-@8YT@6oZz7v`qJJv)c{V`M@TquG)mvFe`M?7gEf#3hEKc6EtZyou_tL zhK;Uhf^{%?Uu47A!`OzyXMab)+5`XNSHLEG^<-p!&Gx=5 z{{LI__{iCxI0cc7z&0I1>m?_MHCt$+Pc43U4AT_5xsOwM>Jge@Jm-qA%1ZtTxEoS# zwATY1s@kQ|Rg^tw*(qK?yT(DWb-U%Bj5C4qTdgQM{GpCbP7_FVYy|t7f|47LAu)uz zvU+P@!<^uyb}2Z9QeT*6)fI?tD(6I*_BMjj{h$9%b}D504%|UO5%~7DTI(8JdtV1g zG8|q(fBqr-K!!|ZAs1%-Iz-#{=e~Ii1j#=^tU?9)VE15TPbmEQqyIo|>yNGBG~@$qKq#jKc@(WpqC;TN zRwx;eEd8|#a2*melx`=`vJlfps-NkimiG}pL{!EUGM~d~zu^$joCfGO7gEKq%R^>K zEUOt@5N)VcBd2(JP_rT-kr%d-^bvbNMs!afVe94^)9knBNf%tU zgb}2Pz&0V$L{I=*H0LJz3~ulzCJoIPft$<-E4$dGPmmWQR^<+y+fE1i{hP_wZA}7l z`VAh(41rcG|L}6G3Hwn$PLbL7lR}5s*hUXkWDu7iRgn>f+LCobhK;4_N~Plmz$DW> zXI^FonSolJCdw9CA(?@1{_I$~#QcS3>_AIP=kQ!&{G}@y|i5yNVsS1cZ=dT^{Hb5i7VGKX&-kvV$(b zC*}7I5C+gK*@#yFnU58(uyU73C8uuq$Bv%kL0D?aPK9xL! z@-%hPQoRXSc}MM!V+^mCWX?#|@4qg$eEc9a7-=OMypP+LTI3}UcVh=RSzCL6d5#m? z^S?Ozr_#>7&46o&Y_^k07?)8_xA?=aFv#tXe)^gNh?HLKPhxTqgC(3dVhRCS-a4A( zKuuH>A91T0l_#)OgCL?J*A13r1VE()iVvTAe}ZUD%PSC@LE^{0DBdL5WyI!+gkNcY z@V6!!t}U!U08|T?a-)X!T=^8JP1WpcL{Yz;^O7C2K z8A>4<#7n(;z^ltbGo?8D2MGC4>M35Ru|2gquviaOSRd|jQ%MV?I6UpH zbm$gmq)v{)hRLnkFcJ$GL55*1l^`6OzVx=l8g8=xEcVIR$+jZ`X zj36Kja$fh?_B=n>2Nm8LWkX@UP0n}()ig9SgEM8wI+fxQZ4tY)X!51j{RMOoQ+p;~88UWMga zBk+IBH4t~o9*R7P^0ayo`3IRkuGKtYr7 zt$gzUGE(yp>!oO!Ie&MrNhe|}$UHh=oB&t}L~QskCqeU0&;Be>Wr!xP_2XWiDOMB) zgo2SX+K}!o+I0l=-WcjHlZjWeZ0v7>=_OZ**RpNkCk)%&DeFb%)`9O>_rWa?c5kF; z;bk#xfeBlRO}o|d!%((c8f$%$T2{ahrYz3ni<*8k^%G0*6%$X%_e=*qcF+yL8nsX* zbi>kdFdxN}0=lqFSFPOYRH(Tb5ntg#J8g@kf#ZmXTfK#@=I#jPXD18=E0MauvsD1q zA33ijE}WIhEHLq?mD}YVW)%@^hH_){C%WT!nUcW*Y@=;k(V2}*AYDhjarwjmoem@! zm)8#BFX&);U2cSWcSn&N?r)wA0X&3e4~($Ao|;rF^lENM8L7wA zk>T!te0Q+s@YO!FoGlnE(#{`;GAE{@FDlhv5Mc(a=C_fu7Q|kWth_T63%nbE)0-L7mhesBf!qWxzZ0+by$UFGa|NnRDB ztVSSCQOo-20Sh_HG|zrIrdjH=T++`f)?!G(Z3e;q(5M13@d zc3;PxZ|yPy9LXASB1E^#`Bjp4-f192g`cYyboG*y0Vvb_6QA^R!qa3bHE{)zxk6IS z@DWiNPxdUIzkNiv$Io>#o*zhT_!fxd&HGlp_}{EkSpM~p`Z4R? zXstir1{GRT-n-6F+j9i)3kRJD&W+B%QUE;gpNhZ8mv5#(hOTo-5)UAu3JHq3lA6Tx zB)2qL1T&2IxmApyzm?sLh+AL5wRbF@rPd~lMj&IVj;T7^Z#aoBd~d$#`SF}M)*W#7 z^#9l0+pdRt7hHVb%2W8#y5&Gy1tl8{|)$d3JoS`(y9{O!NJ`gv;s_i~DH5-=C2 zS>~Iay8oi+p%r*Lm{LmSSN`#KwZ{XY)rT7r4a#4{jqf`?;Kq%r$t;QWDyR{!%+2p3 z563_B-`8=nD`;mQC_w7B;{iyS4>y3-3cEb zGUrfhs@k}Wn_baIl^mzh3|+hO{{|caei7}{&0QN(B57M@>B)hHB$z&(mztsWDr*}} z{<^g1)BP*otGys|0f1?`54x5Q4$Ii({d^X%ocEpuLC-{7#$=3!g;^~eLu&Hdt(p0B(2*k<^IgflQ3%q7Dn z*=q5<=965N!}c!i#p$(_K-=E^=P>;t>q5E#;#;Q<4zTjB+f0f{SQBY5sF6f^@bvj;k(dv$ypM{KmA z)sDnBQPS7G98gc2dRP!C?X6#R23I#uZDz-eP4EE+BCq`R2;5-aePQDp!a$U==)fG>=HbYqA^q)8 z&1h3yljk8!FO>v{wWY7m9&Y>!N#+u6_ym`W^2~|%Q+I}u(d{s(ZYg42%ejG%-v@$m z0TPoYS+?4j$m^Dhvv6yi&J__4X$Yi_g#{=hn_uwBJE4I$N_A;)bi^rrR8-~;jO8gi zH^}o?Hh-9*+k+oabB?PqD4)N{puhmVH&bOrSx8$s4a^^3!R-)^wj@vvAoe1?1R-Vr z_m8+@i9hvG;CSn729DkVYotsMZhIw9UP4bTNn8?=6-vN1`q&cG?C60VU4IyqLDzpv zEBvu=qA*g#UX@zq^ZZ@>%TuOhExpxA>ymuUb0_N5IC(9o`{X-EM zS}T5a=T<25{C)7nNe->RF+1MR!Y{v07!oD_&dj&mR>jQK5U{ZPzQ*`owm?DUT6nH?FLu|SlN7#wE zg0INh1MLoXu*0ssHQyO^4Xr;aY=HTo6A>Qi+hAF!@D+J*n?aqkEDZe+Eby0lq|SSz zyfy`rhJ0qPRudqik}_9#2b?-8Lhz#^IHb+1iu`|tv+x2eFnU_Gl99!@Ij5Y?nE{#` zK}Y*aij7sp{-METY;s|y(%DI9ftv1 zg%$>b($%y$c1fw6gzHGJybTht3e5rMR&*El?V*!Iu+8-ZYPZ1Yc=BUiRX=RF8IE*8P;Z`#{ zsW7hlfUk14=)3=5j(zwpzyL@yg;h~b#lu|VPj>NfiHah#Vp|1zGvMd=0n^`QScr$| z4G^HMZjHJIdwPb|&8su0_~~PXxa0m1oRi9g51`N`aHEL}7vLlb_iXE$RY4F1c$M(Y zyD$=6>C*ugKSwQaSs-*Uh^e_SlFh$^nz_3``F(cT}drEZ{ z3f&3yN`PgvB%@hr@50CkCGuee#|lL4wL<(XSQt@1m31vI>JQ%x8}DecvQ13Jt7}$d~{aHF9_$U*@^;{)bu=hj2Q+PT8;4At}X&RR_L;I{Xsp{GAl8tuU$Yh#JN9r9+c5sVy{iZYWb7 zC7uX|9cjy)vmJ7#1!KkrOiiHRC{qa6wjkZ5;gA#TZnGa~~E*ZPw!XDK+awFfN>8Sj8`xgB-5o(&i0eoId4(x$n@fUYlB?%-L0^eHf zbX?Bo4qf~aRDqw}U}Bh9DFZv_vg*{VWoeZmAD+j<_e=riwlc3jXGfLLpO}*wF%r9X z$pWhU=c|{PGN7&;3e{%6AwRKg7v4cW_c<*2_cnkUD9*9_^;fi9d|xrynqVbek)4Rt zLV>fp5)k~O36U0IhzMRg(7+WvbsdU?B*k@$nMfGf^MWf3Ws;YUBu64^%LYiV1%GDJ zjWn48`C9rJlrb^-#Mf$oXg1UL7|^}=eR<-bqQ1vs*xr3t5LbAH9G0!AC2p35Dwu2s z(RefM_V}f+;!8zM`*HPqs-%EyH>?2(LIel=7dK}66qW@IIf%ZUhpq%$v*?E%ep(EE zyA@sE{k#{B_FEqf!Px4IXmvD-T-Y8c=>LpCS!KyBZ(a1SlwSM6F=ZqR-?PeAtz>e` zcntDzKf~`2H`5E*)Q%JIY40%bds0O z1N&U}$CoIPaKFHBL2c?d+b-d80IoeL`uG`PW0p2QnHpw~nVukB>GZTMt$YZqI4{4u$Nd9^_1Z!C_ z8lW21@&%*49S7f0Y?9!QHz8IQunZT2jKCNS4+k0qD0Tj?eAHWAz5;Z8%jv!PPa9MqhZFykj{M2tV55+UJsUZvu%BGThh46}sj`spTkaWQduZV< zVuQ!QzK5V)jNFJD!$B=^Krk#sA5|7k#T$&f% zD%fk3{stufvOI56uu_L?f#VuVO)5Gaqr!D+WS3Ojrga5l=dvmdTWfaKXDM;LZlfu2 zL=p(((D2uuEfV80Bwi)HrSfVQUUlJCd#QVPoDKDi(xlP@)=Sh&Xq3~S`*MF6g`Y2onbb&f z#If|la03eF<$e2$z5?ibKL3p)iFg||aMwLzdosoQ$?Yl>ekGYqmoo0~)}EW9Df{ZU zetE3L>Zlb8tjj$G7JdqV=5{$O9LPy3pj?Vwl}@%`P{vwr>8g>RSix7X1iowI{#3jP z_?vKDSSs>i8OPC!Vd+brI=D|C}PXDfh&brtYzvS@XB5cO4!toPg~Ks#w#E8?|EUI@j51oc91=Z4~P- zo6w0AzSMN|$J(5dtaXJG?g)f{;(zQd5nz=m%}$Nmh`xvY2B-}ka#khGXh@7tpVWc- zc#!NE)$SiNoThAhT=wFHK2?z@lB~*{bXEUlXugiv$HxR zzm2d^~EPw6=VeJ?Na^Y^7TL=tMm)G{i+r9J`2K*#jY>6AYxm2cFiL1e~u{C zPO11}Az+cBe?m{Ssl!~W&zh-GV*wufvHUMrVE5(r6Lv7~kU{M1WoY8dyV7OwVNojE+GlCOxUjt zKLrXD#}&7>Z-DfMoG`8+n147^h~esh8w@?_68%k=mfZCrs&=|z!yN6Aob`MVp;+<-Xehdt50>cPK?eOL@(5Y#UECQmn4 z9l9`m60@G8P&_Q?x>AFi6{nVbN_r@GS+7-I_8*qx_m3D+)4aLu_%p9UZWV7#C_O(K zpUZOQic-j{#Hi}hyNvhLoYm;rXhFhTh{gdeC1uQx0<+L3V zUx8z&d~~)cn?pe9px*VFx!sE#%8k#(=?-l2^BrFLH5flEpRg*L__+kZkOZ+LFa zApincp&DB9WT$(#pxE4x%daF`_J+yQr58S8z~B7Bg5NTE3`Q@|m73Ol*fM#wN||_f z&mN4#I>d`+OZQkFoSD8E{-S3nA z!AwxfLe2tbE#U13a=UHZQg=3(o4`?Gx)$-pP<2S-@h|UtpA5gN+ZfKwGR8vda}rHZ=M+8l`#0Xpol$`%MAos1nWDZD znHKQ9xAk!4EnT#y(=30ksg&ud-goQ5*xplYeyJPu^IqxVvB0EP`JLlZeS0%h?$TJm1^!oc&+Vs42e1 zHJNRCc0%ZT=5rqBvHNtY+D^u)#QW5mizg3V4F}k6!gU=wXW}4LA`jZI^=cBXQ5OGE zH6F+03nM^}pmWY+B$)6fsT?lahJ+kWeL|}v$V=H}hwh-J*yPvdI7N)4otS$CP`YRj z(sREX(w^dYtbr$(+pJ_ZP)7^lM-$TCQ$D1@gIJUcO4tzfE*pp+*iZtc5ZhC_c8|-q zt3+DHD3O2(<~Lv=o;wjicZQ-PFB$16x(PZnK~UB<@mom8)4L%85#tP;!q5>|w%ysg zbN`*lIJ_O4kel)l@9m(&aB%czOW(y$`;NC=IY|+7?FK)h)7$&Dzcvkbcw`frNL6&U zAQ&^ht8I?~|H%y(2Fqnnr07jU=rc=&HdFz445#`HpLD|SmbG$`%09p9AS7b{8o?W# zwgV+x3$Jg-3ZdKJ>0L(a6^0`j{WKj_(*tcX?KFg-UFbE^7zA_z2A^sc=R{KeZawz4kmvb6rXj@|CC?aCG$UQ_IVq+n*{1 zGEdKd9eEVGiUv9Rb3Y;@NSHk(4NvXl8D@Rrx1H_cDO8`ZJ@3-=ZL1(-~^O zA??`VyfC~aN`mgAJ4gRlTJiQe`v^H&AvQfaNi(O;nN9|zBd!c_sZ@^;GT{WYLe6>} zCWbs(T67aXi${1NT<2W>u&=)T&JXL?|G(niJF2PZX&08FfCxxY1d-4qDn;pp4iY-j zr7Iv+s`RRYQbWLiC=fbG5fJG$7Fy`NLllC98X%zslH9!S@4nx<-~V^5oBy)*KAAZ= zbN1|+{me7`o<^5W!ilEu(kR;g|3y+j-~U@=dA3HaTCMn8`gJDXD^a5YAtmFQ@#g8J(>3(;SqbQqcUmxfI3P{AGr(;>&+A1lG0mSsy>FH|Nj% zBS-6@L%&1mLz3tE&X96SM$vcb2}QaHYeJ>RqM7v-b66qE+y4f4gDJkSDpw8aGYT@V zXLd7`y;j0a1*I^KSkg067|~8C?l6hMV?rU@TzO4FH!1(E;?iV|qcF$}4}>rZB)f-> zWaH*?9!2Y?3DSPFw|G@*%`ZZE6)D$oQ(}@Cp?g*`G}dp#DMuNL6#jH=ib=R$4u98v z*~IZdYlNNg?%E2)8ic6-Yoi0Nxz0hsZbVWR-TxTu3tx)PwUiEnxBjET_e$&1+OpFL z$_ps`KppDHj*i=Jg4xteWK!^0jg$^YOq5{8i!HM%`^^JJN;kOwJk9d|Ponf4x8bwQ zdo;lB&-s|>-}J0cXNJk9w?nSRBvJ&q{vU<}QTwlS!J>%(eI?g_-K=3fb?3kfb(gFT zx%J`o9>AYU4@YwsA!-xnHihMmsDI>0e0@F0c+BcnEr~ zCY5B53Ye@vozAtlS0ZL1(jL2fzm`i&%}PxFRYAJUt#*n2(i>W88ABO%H-CCsO!698 zppt>_-s3;>pbwsw3$)n+HyYihWvN2wsQ-IzP|E9~?BcZhGya?(X99I)o^zG{%&A(v zRPI_6Ty*{VrT@&jNvSD$Np;OgbNcJb6+ACyx|v4oKSNY}WA~|(JY@^*kCIT@UCfoP zN1avXh6avSC{}l>|6GqxPz$^u>&L7qJ3t0yA5o8!abyyLta|&5YW8RSyFXQ}%+EB~ z-(311X+a5U(#lM&?W@tZiL)`JjhLDirWWP@JxaxQiT3r6E06zt{&8`3W!FaO=KneH zKfm$cv;JRx`5Hspn;Lm~2p@Hb-o-rKvhH;x?;XI6NyiK0=~Lj#`L`nQ3xoN~VQ0U= z71}&uVx$Q3ef!DjX_D+gs2Ks@e+)e#uWhYi5!{9`mKUEFY7${!vuPjl0`&vjq0eWbR>sl$)1g}`jWmu`PzP?qD03Ia@|tvSH#A(F$6$H$$KuVz*%k3-EWgL z0_G=+(aM=Wc^0Gx5W!cEM#|QyjwdXr(eD8538Up&D z=O28d!3-Vy8~=j^STlA2VPZ1eUy|e3z~Z@`g!#4ErZ2o9E-+q6M*x;Ae#i4xOyFPi zGzf~&f8MoBRWU?`dmStP0(CrG+XSH-Vjdlz?OQv@iuHt?7=~S>O#TWmERwy~R|z;l zz}nlJjd*UqsZGGFj3m~YvNRuU+qT^}A^+?AUGcUjS3cZpqb|$TBI>20W)m7VHJ5j9 zrH|pfC7u_NrxuBSmk0*|D=u>ycdhi@39t*l@P75>9Ei}9UL8tqlS3n#2Oh6BIuk5U zjLXvGU%8o5KU4^lJ$RA7_P0&+k+z&E1dafQ|0iKKA^~fKv<%4jL8l74|HPP9&XMH_-oaIep7T50~B zT=id0gOP6g{XBDzS?xKnUBsH~Vg^s9pMHC_x>oW$`lZA8`8o7BBIkH^R(2Or>o0$7 zQ^Z=UK=ThOEa2b!kbZzSVznvMoZ0yhySny>ZU{V5KQX))gK7GEdy;lKg;vL>m4p*C z9`VHG!fGREqc zU&mV_uU0X1O^_?n>7Ck`2PxN*T)ZNnF1II}z3fkxfW)geD?qeRkPPu3UWE}a!Llz}?}K!c^(BPYirP%rV1wlU_vKWVK#sh}wf zwO;AJ5H2s{9vud+n+}5y&;OFZ0Z*0ijdb1RJd(qb2hOd*Vdg5>$I}VtwPBF!?fpil z*q4dFwLW;Z{gbOXts!l-gipzQbi(OYZj!DfF%PUr#&GJFX49<~gpwIwu+^TW#R#Kv zC@L8{ME=B8+A-gsjh1CcDJuvf@Tz)|$QuOdMrZ|V zvv{rojw&%eK9c=TXEQFD2*^g5BAE>5d6jgTKI;$(ks1vYjz2b*GcVEr3JN=HS)2yDceIcnMd4YBUIX>DijHq^ zEf+GnJpR_FU};6lzBv+>u$8(pY`r8@Z&uo2_*A>vQw^ zNj0jA+GoeSE|X8SVS5P1i2(&2vO7KiX-HqtII{u5Zz`E9I8 zN%Z?3+X^EJ@edj~T?9W={%bHJy?5h42Q%lfPb->Qo zNH0Y6;V}PC)8yXK>r^1?Rk``H z7@eQ*wM*)<+<%;QpL0d-kByA-LOvnbz`p+YUuz6!UJt50eM+0xNPTPRx5D73J#ptZ zIlXT#N+@+za;KMNgQ6+-c8%sQ`PqY}8_v?4VOYK+?OoP&=2n!8Gp7g;sLGf1Gu^QJ zn~I{@4rF;jfcrDEG8a2#9nodW=mdF}O8K%-CobrIADm^_&Tbfs*t2rom3fqDAE)J= z(~w%b+{Bsl1rT}SAh=cqA;F_DEhvmeWXG!wlr~qQlW&g2qKNPvm1EU#crOG1oybuj zJybtze&lEpwB~IpuWK?a1p^uTI1A}{)97mD^Q?oyo+uT4w2z7ss)Xe8&)ENaty_HD znbpEXnpDZ23)`;Rfp+vB$N0T>pGJCrUsG*C;PTF|yexHm9FNWo-wirDIhTac}Hs^}PDe6P# zrS}K3z+i`q>?>mDa@wlHs?YeASY9R;JwlDqe{*beN<|p@z6K=%l+%SWR4O48EZ>nV zNo|5#)^hY6QPAk%t(m*nutdq|W&e<)yhwtATpx0}D4`(T`VjgCt} zO9xYTcA3pMwL3q7HTypNmc?j{-`guD%~3{mr20lcMX(q(51;AAFo-@}f*tmri24@9 zRO}YQ+ZO(?!c*^Fq&C$*LxyE7S%C`AiI%#m9oy|=gf_hkDzSfrU+P(x z0m7i^#$m_j%(46aRqrIVX#FI!$K~tF6Ajzefa2L%7R?_q;s{phNK}j%j_ay|!ZDTT zCfLGaE#L4kE1@QPWiW#7YAYo%>lrwx>9#aMm*5#Q^TZ8v-UH}e3b_1_C2f`XGr_hn zQes&*>ZINYhd)1mtPBi&eNh2uk6m@VV!R%VUl4aoj7N=w6L3*wl z`QAZ^M0#xYnR0mU(~Y(9?v2a5UD5H}{Py+}=C~nF5O67^`A2)W#bdXA`(OC;agem{ zPGoq-&X0c1;u2h#K#SSe*AIyFU2Z&FIwMD$!*b1$@9JJJzOYh6&lO+ca3w8g6M!Y73mSu2xBwTk#kdYHG7b9t4!F@?FmWa@Qo0=WrbGMHTGK6f&$XOI1Acs=a^nH`aoW6SK${5&itA+mVp5~Iu=lE_7NQ%>HYCi zJ%>}8fG<>&IjRphlE$-;b!&|<_CA$}2g*y!h&3ouZojdfojt#EM>o>vv&P{e-6_!7 z*)gg)W_A%A7C0pXO~3!qzn=B&`|7IFu=c60vKt=7&b5QuLV}OX?u!h+KKm8Q4_&HY z6F|kYc~v2Rtnd2n9(L$d-hzAXkZ@otGe)c8Sp0g_zXP~4B?ffl-Lx`XR{vQ$RpJQY zoylp4*&G{*P`a<>b0u1`Y!AS0^%X?BBoY5`qJjr-mGHh<+Br88N;0{>y%{`qA`e^O zW-??S2~kXk>oyv_pLu^be(gK=^Q}yCse>DjMCK(}${Z(`j^%`#`C*>CUjWo6B0!Hj zf*wGq-Zo=+LwemqT!XXL1Ln?MC9$aZrN8O0nWRkcH!TQvAs`?w;oSyS^?QC=i;m7t zZb#^sDVZh$vagvpbd$YRa_BPgJoLMHTI??Wn&qEo7Vg0>w=KCk4_8Sh%{CuwuqZ1v zgO%{%^KLNXbWtNsz?B7b$$2t}fAOM;mJzXLL;T)EvY+I{+ z!1G>=1CfunpD-g&(yo-)2V9OL3c77213s}AOS@iRqTTgxq^HXd*1aa>D77QohfUb% zrzF)S`jERT;Kxe}HJWV>yQ-lZ0%zuq9jC=$S&vP*^QNjyufEjYtQ$F132hZNRLYEiK7#_4!NUBqh&o+fFW2;7_~;7UmBnKb3Ri)c9g}# z_x&!w0VWU8m_$=Ns<@U+>5k^xNnAV$SqW#OSeW)BMwkTmr76yP)Of-Enk%q}E?>ie zv^|Zjr*Myx7WA!Bh;6`Ux@3NtGf1!#*`1jW$90AtYJOtmXs?IFFQ_NGlDf`*1W zex3jzR*hT9xASg#3%UEOBx9$a&dQgy{uJYkO*W`pd@#ckmOiuiF~HY$lQI8};O9SM z;{VPFG=wG~txB7)?o5xV^FRW3=Pj$vJzWc_qV?yNHEF0lC9qSDgUkDdbOvsk33J%U zby`Or*mENw*~g%vAGFw}^Q#P0L6!LFomW=C)ju+3KDp7%QI-p@bq9UWPyKcnxj90N zc9Bsg+Gd^X=^3CN?HME&M90J~ z(RDYT0Sk^uwh33gL)pt3yiNJwkd6p*p6)vVR>RV(e2=0f>{`377nF0slr$AJR9WX= ze8_vdYW}oR)LC{;H2qh8T6x(KZ6@wf-|2b@C>^kb?taI~-1TkyF`<(9tYbe1cuy!y zsMK?!%JU9m-jnW*iTLUVm0=k(Wq<|pk-l3O;$F}&O=57CpZwv&bfuVi2FQ-;cg{Jh zfVT3uMS?|YhhgCv#amV064q}aA9@*6!a18YmoUZv1I=EY;zEQoAXF3BdHRbqFRZYl$#FltsFoqxv)9x`?U zH?^`oKV&VT;f!t$;pe`1i;#ZDGSrE8=hZG?IFPQSIV4KU74!b9mHFhD{@LP1Wr%$` z9ml2AUzQV3{oXjFBKvBdPe{ClOYDW8|6x2VPCg$LT)t$(b+Kk0!82c&{fkrMu;tc3{f8lGWSK zJ%^&V@z^+JkbRm`&8@TMXN@ zn&SXd3CJv@;j0}C>Mk>U+wqW4h>RGhib4l1_!i-*U>bru9r-pqI#W>Tk zmzqytRE!puE7sP*IBLld6dG_ZFSZl(IdW2cm&5;f34}YCD}{v3ZA-ry;VGW8U0mq5 z!u^1OOyT3Q1uEs<%XPoqhZ?wDHlj8f>o{-13WKU)+3T(q*)#ByOUSRl_Dygy%cy6>|@6RSS=u`*$WEo9TU7v@3DvH5VbLY~vlTjg7KF%L|_m+5WPw%&E8 z@qMO(31sdrt#`=3kY2wevumcP)PO$T7v9=*2~GFi&<->9yE^|PrB^a z?lU4QcE3AZgx*$YhJDx(zqnrFxsX{S{U$mcfn{u!xMxjA-{Qlr*OuaV)7N?oH5jP8 zVDwgzJfJJi7$C_b>`jtU{-x-~IO1BK-5}`F@gU*&v@NLQPefVeU{SUYzbO-2;pCDr zWB;tIeF3VEvv$C{WJ#64O-}i;Jz#%RE2P>Rh)slZ0vD=FN)snX^{m>2sZ9m5{j~kF z$N$|*TKxk)7`Vk*{T4TV_cE)&OU~}*w2U<;J+r^Y=Z2!vN9aP%Ui5oG4%ZDKVSk9R zLvDjt(cSO{h#PCx-FskOR`rFTS~Zzg(sv5jT>xcq|VomxwkSTb`I6 z({_MyL>ZEhj&ROaH_WGF6@}iJB1>eI>8V@j^)yz9d}yLMRz6LH-!J3lwj>pg@Ndyk z2Cp2m_G`f|vpZ)*Nh3rtF{+$6%BV1KI9-BDV!wb0Jf1z&Hp0=v%SDD`mlSOJ*Icb0 z9kQioBc0n#g;MiKx6BN)+3FBx^;MhLoqqKIj!zwps?+q%HUf9nAKS6nHrMf})s;nl ziC&01xCxzaFztqS2QJRWL|qmRLw1?|>E4rI`xZ6wdCJ}6 zcZ`5O8@cNKrrUbyz?3_W*evidcjdz6OStQjBS%v0=Q0USvnO*}MoCI0lVtnFX*K=S zWwmZ7X4Cp>;jNsycBghl0x6|;(DF^Rhmz`G)8t9ihZ`wl5Q)R5erY_`PB&$aI>#m#Y?{#gARa-sQ);n_B-!*JaD7p<7UmB?QzEa^(86r^p z&nIryt+Y0Fl(GohLKnLVeMwbAuk_Rkga6%wSi8aa&-+0iTz~K7U?B-_Q9|}iNiX$x zg|j8PLXQmyU_$@QTJT72xocQta}&9*-WMl@Z25hxmXKBM|5acj(&xUyyV^a4>L}tG zwEdJR?KbJ*=R$6~r{;->`>YqTs> z2!RBA-k3{H9(dC57ao7Q9;xL^<7LU=A0M60`>6J=_gVuEAbQo(T+n4|#E+x^lb?(HdMQIF-?N{o0Xcb1*k^X${mImm)nRYsvUgK{6=mh55v& z=)wSj_vE#CrBQF@ox)4c!zV5IF<)BY^oja>&2_A|iqp3~aWQCnTz$RvYus!wY^{n8 zzlN31nEO)l#5}{@>2&edGBUc+w7YaPbv|rhsw?uH>@RbXin2WFZ;e&N|H0^Gtxf!PqJBw;ZxE_{xVX9*wk%4Q_ox3V(925d`MXcXvT(n!_({9(tyR zc}1nck;?u<3q{u+ST8X1JX)b({Ka@w3nP19a9uNqc#kBo zcemM`^{km?_afn87L@~;*QzaZMbE);pFhYv43dCNw0lAy85(ogMQY^#{nx1Nrdsr! zczkn3E}i>dAXX^9R~`tWrhCr-HU{j+rfoXbL2XJIqaydQ3s=|brlaogtg*L4S( zcVRy(lPjti71MQPJo>xw74|&)a5vx(Vrsj-p0@by47FX;3KAlgj^Tf+AHHF zoT=wbx(JBHM)*wj1dnm<&Ob1+Gw@1ye$bJVs}{0GGhvhf!n(hIY%~mtaoMvSg1~Lu z+qawoe8xf(L&75Ypip1oor59h zNi}}g)b&z2cq#!qG+Xf|ID?yCVT2Lyg^oDtd8n71)e-c!%*BLZw^u6cx%rH2fVQdr z+n5U^qua<>BDccB9dp&U+M;eDiK1I0{>mRF@|O#HvBg?nvHIcXY@Ix`KjiT1r6F+@ zomvvUVf??>X6QAR43h5%@wqs=H^z=n_XF;)2=}&@?Z51xCi3yCtID9adSl8z70YCI zOT0?2#^2AE7icdI&ghQ05oO5y>F_)!7_xB2z+_VrH9%#?m8x`)U`(tTOI~k+3n?FX zxBV0eS4h}FKH-nYJE_c1ur=e@9$|ha*A7|*rbOb+14nnK5(EP^c3n$!cr|L9#1GD!^J4$}zM{ivt@QRHZaMG#;H{yb~t@C1)47h-p;s7ne?vuMZ6xCoY{ zcWAGbevkq8y9I5alDlNK71bSLSW|mwG?r z_M(QoLpKX9Hebc1$!k}bKib<3kJ_G@4M94B!O5LRMG^+o_Jxq%wV8a* zsrn&S^s)TH*=pY+ST0H}ySXZ)?FL60KzMq83<(5j%XJ$1{y?MFJAo9*Y-LNA_DM*o zsoS+CIObMUXDYdonztFC`jq`GVlt<2?Ob>~{eBI&n9iX9^|pHEg^Sh9_WUQ$XpP^r z#48_Y`iTI?f|b+k3m{9BnEJg9U-bDo43!X1z4Lp#>o%y1ox~smk6%wesCFXQM--Waj3Y1l9u*e zdpqXN$o(G&V8Gbux5L#by0HOAsgjICOxa12CMoQNfK>}MQAWBPBX8_MsfrMLaDThq=$iy7^r$xt0%;UNqE}XpF zi29}T{*d(Rv$hxy8{vD`>9`mvClund&nBZ${VT28?T|N!j{mQsBF75-2PhEnM|Me3 zgfplAv>RZe_q=lR9~*<`bktdUXM^ejy~+G~+>(b~tM|l0EIb}9ww^*T^7s@;sn+rl zENSmXCkG084vqk(vu<=|f!3?k^aFy8qNV*Uq65>F3-xF;EewpkU06Nh!Dxlnkl~)h zDl4o>l^%ZEFYkU?cDGL^ZaxcRXomW4;3x&p7j& z?*iE~-&b5QI7(Nty~sDm0kQ082~*p^K(1E!)L`J3ar+6i&0;9}{<-17Nvc-{`)z4y z_J2aAcYW28oVsvA9$^~yDq)ToIlSPk=>4v?f+c@KI)A>Etka?Ihsj7oph{t2?r9CT zr0=-gVsfeMTcttl=YyY0A2SF*m*^^P$y$fZ-*E_Bwy3q;_e?d!8ip@mw@1jt@F4Sy zH@`sJKGKlc;RW%|7ZLaEV)J36R1=`VfKSEv%2g0{GD1_QZdB8HSgrH9!2XSDM8V3J zC>Lyp_^{*$bDZ<^OcrLNlF)t%^&Mz&@cKuG0IylykoQFm!7b)ySDZc~pgkD6hM)CuP63DPF)_cK79SeL#^OQ_ zs4d8G+=bdkb@qF2?vh2Y!XopG-aYajEhG6!r-SBZDF_h{W}@MT#{Rm72E@Zwv9h*V zbRls77|5>NRg~?tElvnFtClw}0`n*@YzL4HgNo1Hj=cKK%_4*1RnqbeyG{#vMhD6D zm9ARSR%XNL&Fd0c^9kobVJdO*sOfM*`M(m=S5Q){f%)5qVf`rehW(kBNpvOl;UNs! zLV4yC2z58}CpTz@!n58cuqt95;%Oj*e6T($a29eBGhpNW!{;+xl0(o8w8H5E) z8XYa#y(FZoDq3#&sw3Bja7HU+hfZpaI71Ud|BJVGBx96pTz$u2Fw}HOeTijQX1@oY z)AG#{XqR}<*AEUtHDo;qMX93fMyZz{g}$X$SY{rLHr3A%@SLrbjj079C0b2&WD3Qv z40p7={lxbh{mV|jqq%BMV9ntUo>awdHN9{6Dzkl$r#;AP)#xk&3Uz3aIuMZ2*o_81 zu`kdr|7$yymkXCa9f1isJ|0Rp7BTv+iD3*WS0?;UpH;~TP3Jy1xpN4Pn2&)* zcZ44&v?BGV=tZr@P~&~p{!Mn4bwu_F`c?9t&YSZf&p6h4PhYb0>P`NGmMH)L z6rEzL)>;J^==-9Lk$6-WMV@w+Sjj#;pxFR-XqBHLcH2wrd2>R)+NX6hd&V(Pxn6tq z3$@dHDv~fhtCF5(d=Xk!=%otspQaqmPKfIsUfj;nXJ|gDj65H)I9rS*85tSuXIFqF zs`gGxh5!IU83Dzq^7rHrJ)H<5jQ62PCT{~b>a6#-KT61|2E3;eGQ*-OSi`gV7rK&r zO}Qlp1LaIZq5@5sli1x5yt9zg)2-<$MR|368v(O5|KT&!a`W*qv@bO(&};SU{2swO z+yW9*PKE1nwjx^+ZYv)Cns;t?5o}1Vl-KmdL^CDvNDAxJiq;g5X5z50Q4tk|HS?%6 z?w|hORxEw7(|sdtqF-G6=fS{v+CX@5zM&oQrPex&T>XCHIu?pYU{`_oV z-4pY&MTB|4WM^G)S3>L-351`9>m;F_3&}Mw3$JL@+|E?W!i*ptG`0QC)MgT!3ScpC zUnfj|bot6`aO8B+nipi;msy27?l~J;LYaFwuA^!9D0Wdw)gMrF+CheSrCc-9kL*E- z`a&9M@b6PkcLyT8L)KkF8n_wVe?&%Is+W)+^q*?(oV(DGjRizwmMhrSFamd;A` zOzH&p9lyEuNpP`(by2DdQM7P6p4G*5jm`Vg6pd>RHO;BJOP^KjL;XiX;6tYJ7o7>t zQ0h|>_oaoF%m0NLEl{KV{(h3W>ZFPrGkjsYc2|q`*>(Qos|&QWumA5uN2mGr$#nH9 XfA@d6GzDCud~_chY1L>vd;7ltzro_B literal 0 HcmV?d00001 From 992fb432348dcfc1402c3a354d5286686b71e065 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:47:21 -0600 Subject: [PATCH 543/590] Delete ALCF/notes/assets/CPT_data_mixing.png --- ALCF/notes/assets/CPT_data_mixing.png | Bin 169092 -> 0 bytes 1 file changed, 0 insertions(+), 0 deletions(-) delete mode 100644 ALCF/notes/assets/CPT_data_mixing.png diff --git a/ALCF/notes/assets/CPT_data_mixing.png b/ALCF/notes/assets/CPT_data_mixing.png deleted file mode 100644 index 3258af13d46f2c5536ab52b2bac99408d267e11b..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 169092 zcmeFZg;&(y_67`y3ZjIhq<{!W3re?igVHG_-Q6G|NJ=*%Al=>FEg>=FNHa7;zh}PQ z-&*f_*S+@-c-LKPU;)FL`FzgVXYXe}&$H)?qP)bDN5qeikdU58NxoA?Lb|VqgoK=h z_5l3F`J~Gn{GwqdDyk?YD*9Z}-p0ht(ijPeF4`%oS*BATyC>t*k4&*wTwg@7g&C?P zwYsaF727G_cSkbG85aCFrTOx+urlD=^7GA1AHRwAZY4*&_Y1G;>O!AYK~Ev?tO{Li z7b1fq7xse-B=Oj%tkF%FFt-+8C=-SR5M76US9ZPsJ#qM-Tc=BRQcS z?CK9Y-&z)RyhCPo%9S!G`>EB=AVPgn$e3L6zouxYL5}Yfz%m!ziEl>Qvu|u9%C9` zn>S2gt+|ZVrA*}HkY0hG(U6dX&5%&RPsrel7OZ>fmb2>k1qjU{eQ3W z?^ox4>+-*K`G23{|DB_MpJo1kyDpobP+%=rCA^oL*^2`Vr3>Rw#Xn*~sP8gy(O*cd zTDRUf*FxNP>RYd&ENSYfPz}^~;7K7=o@C!k(P9tJ#=V*pk3UJK$t^1$u;z^j!CQE4 z>E!)IQGJf3Px7J4sWz=Q(@&ng?z%Z|y>nXo${s;t({#9SzT^YDy@B~OMhaZgk!F|| zyT$a3QJ=VfG7cO^7Vcv&fdPepcd=o=245feAqW*k- zds^xf=n`pT|CO!BcFF5JZB6TD)aQ21dCrWMbFXSL_u}?A<8Bw?wvmP%S*PcHk#qGY z#bzO1%X$l3a5{%B0k6;HWX8f1&V}W4?@P1uPAtJXblTK0ugk4GjTgln$uHU*#B~`! z^)ixKuHQ&^FwO9)A#oUD6KxD}Zo9i#y1QQ=%9w7Jo3!wD-trsdvyK4u|ZakIGXCH{!0@Cu*-3Q2=XASmmuV5~YbZFNC2W|J*C3U3!RU(Dh z(0&TO|GJ_0S-a_UT1%KPP~QAtx!t;9D;MJiqUn}B%4feW=`#FgrmnQ*a=`2Cw`ukF z9c@BfqHIay?8<}HFqQ@_o8~u;%&Se~E6dQH1{7mVgWHP@7us~!k29|f&gX4f^~WVN zlsKA?1{kbQ#;;6^QYDht6|#8IJrBDH_lqh=9zB{ldiolvLr9ABr|by#Mk=fN$KH)( z3+CEROy1eip_;zXUJ4&RIhzH$Ub*w?Z-U4VdYD__AA>t?0tt10!Xwr*FcyVT8&Z&* zZVy&EJD;n#NzrC_+Sgv~HzrUxQ;UDKJ%@PMxTl_}fmmtE{F^U#j$W5`{jtAQ}Z6xG^gDBL4-(PG@%%S$jgf)lVr0A`;N5;-u+FN2Dcx-VY2joBDh36j-UEH`WiQRT|XS};@kDB z8;0YTXc&gSc`7)k_Qu*rVMyR=*9_uXi+@DpOA0aiCKq$@CxzwkNoed1_$3||9@l_& z!isFOk_>PEbK)WTz$4K7XuGgfHK(He0m)~0%XkLgVHfU)wbS!1d^K&Q7?#qYC^}2mG*wf8>$l91ib*rH?Y7+!UPi|ri53x#9F$6bK z9Y$p5Fzx0J1&#zTHTqVl`dhyLddRj`Tr)S2<^pkB!M5TNJV;D8F(KN(|C^aA5KB|T zM{jNp*?$21%h!%_b8g)E{vw`BM2e3{Uy&F^8k?D1w{I<6~$?Ou4+nRxCA zLX>f(ve67!buY?%55n$YMmF#B3B+CqOg5S%856j)AwM)qcdkZN;D+MW?$ylm*zj+b zG#yEEZvFVC8a8)u$TwjL{}VbJjk+nDz%bL)?KWI-eU#Apc*D@A>H#;X#a{9i{%@4 z$J`p;uj?5DefII$kk^0@rw8nNUklaLLpITcz2*J;Ip>w50z2d$yKKS!T-JoGH)r!J z>Yiz5f5YWcLg*&z_Hbu#O3Vyb9Wt&&X5UA-4|a|`r+LlPOS7sh&2{yQj5oXfNVN=8 z_4|Iin@uJZK{IBap9^%(AH^`jcFQu~kt+7QrfToNBN(_*nLBi!|0d7VZx9}46M2M{ ze?7;!NAFe}`mwD~Q(-gqVZNGJ&#NzdoT=^{V0l|_1+0Z*ai@#fRx z`o{VKp-b0)I1Q4??aEcf3$+ofbq69()~NU2)VAIf8w$TtnXL$E7r|h{BPd{n z{?WU;oQYyqLAo&P<{dG>so5?l42labXEwg?OdzO_^L%K`u$!JCfWl?LuKj)5mA&`f z%{dD_uGE#7)P7Ej? ztoCMUCU;#NbYd;uC)8eU!uqBU9jtnbSH@6W6+^{*;F~EGL}G3RE$4l)%sj zoapx+Vm?-FJm{r-6P6uq6Sz;SkBG72l0UXC0@%PH`XvMd!q(C%czYO0gVEhJ z1MxavT=5W-3WYMo>oA^u%xaogbERG)hS4q!XGaPU$E;qR-+E5KuEC|$YN7sab16TY zv*2YM$W65$h;nIs8?J7=0cySrmu7Eh8Z5RQyTwZ+`=}?Tgtj$gh63W8A+1GBarTuV zZH5gTkh7;kQWA7(Eftak_GmL~eJn7nx7TbR$j;>%r{8(qj7w02D(`+<&~R zKi5IbpBt1jmzGIgOYeRm)kP9c~1?uSSE_Ov3HIa&a!$ii(=K z%S3i@LS!cNk*RHzll%heO#T>3P6;~hH631{da!LNAZM_`i9dMpnRDav1D2)2fY(IpY!Be-gW}zQ+Hk9C|(Y!=7G#2`h*O;`7P9n8H!6C~XD52pvPp z^}e2zrnVk)BS~~;cvUrVoR?V;!|rp!;>VC+?^RQR9E>E8Xh8FvCy`wSTVv|R0{ z2w;@T)^yH={zGUONoL|+iT)Dv9%K7;dNR6p+rp_Yqs|x_WqhNC;Z}PZTvRm3RI9Ct zHA#Ndf{1POT>5#w>U0TFD`$a_=$mw}A?lOV4vsyLrj&Fw?yh(PQLHpu@6;*-~*>M!c(juwYH0uO;+FcRjIvCWY)tj>s65t=(HqaekqcbSo zCqdzLLMN?dqcp3S3|SXEyfv2D?=}AedCGP;-I4!sZvPy<7MZujT(|7M<%ISg><}IhWEmD->w{M&9Ke(QG(xT!e9G)A)0yADEjd2SCY%7Cz)8sMzI|9Gd z6BlwLA^8pVoE$DhX`Yp26D}@s!8xRl?zVyD_D}Z(N!tEel-;BB^wkkgOsQ8Jb zt3`QLR_D1`avz=uznf^PI8;r~%^_0UKejYnQJEB{FG#3zKuvImwvPao!9DDIp7wUY zyy>y;IHV51Fx+jkZ^M?)#$K|*t)yil=B1naoYr+choK32>5$TA+^IRcGVF>q4&C@_ zKiqPdB!zm|qc)2DI$&VOZ-04EvddkE@p}1O%`YX_DJ%W0q{K91W;c|A^E4U&H;687 zm==8gc$rjSrXzW?cM}}ErZMeP?dq)M(>Fjo(`g%bG}ZTs1b34Nx&z8#yuR2*-m|<= z332PyAc6a$WnE-`2Ax5HN8iipK&W-J(DQ`^W1p&Q;T8#&iC0^chm>JQVq3Lb{Bh=5 zY&qz_*mBa2gfa}sqC&>9L`B!q`_#YMlZUk;(5^Ox)-6+c*Hg3PU3k71&DQ0X{no{s z8Etm{;P0VVIA!k-CqZ1H)p7lj22IhmsXN(47tg4O<%f1|9yV0GF}93oYKb|TFB!^d zxf@>fIla}BZm7vtmSZa${9hi*(VbdW4#&6PLR>zfhu9A~J-{VeSiX75bXzoOL+E~v zv{es|Tqh>yO>tc7Qm%G2>NCDCapCGsms|DuQ+RF+ERi~c@vupSJGtXfF#$_>E*AgE zSXnjv@cg|LgoC;@8EWPB$%y1Q3Ti>!g(rb}DW9nUWsnwIkI{Qqww?=dQhpW8MN#W? z{n4G~p5|jboKG0o34#xrtb{@x{^=zmj=k*^F@wh}Q)p#=o4SS6mZzFNhkXxYev}9L z&7Lj=KiTY_Lpl|ood7;4Ym)at`@>1UYtZ_!G6g&TA8QX1>r<)Gi$mMcv@8K$@H zMfrM+OfE)H5{pKJhw+0Hj%AFun1>heboe(mK?_n-oS9D_IGU5x#e zYC$=_FZLT@9Re+LP5zAh2qSIwBkQ8yIvF9|#7+ z`MP@<3w2j{qn97*ywbi{qda;$qwSF6ZG0o77$rfQYAbS}q;uNwh&DM|Z}_OhmY3`) zG7cRQ@-d02h8w^tR<*0sMjU&YTyPD+!_nuRyGUVM7hCX%sX%nN>uJCsjd$IRkW9Y_ zu3g&b8@GWK6C(;YdaUxVJ$c0x3-*YP`ZxeS;_`DaTkU4i(km*DG?z)7yQy3GXTS(( zunR7x(f`)vTbz&`F6D}b8qI>eeJr<5=3e(VC&B7>LB6nKuCB{KxNFXh$xiDL$1jRk zwBMJS#g}RK)Z1lRCgphp1Igf@nn!B}`UR&L@xis8!KJy9*&%Cab_?-@_(d!3n> zkEh<%>wm7b6Tx@X7bC_rz$-;i``fDM#PV#;{ID=bDyhl%W5IL7LremwNSO!2Y!qo4 zKLLfxrZx)Wk--x*>6Jx!E}KQoY9Yw>t0qDSq4^XS^a&BcNZ37TwO1Eh>YeQo-AgX1 zy39QrY2(WL;d#_gJ1P2DpqnzjoiuWCP#N=F z;!ML^<3*gG6KuN&lN(m^iZ1o{YuiQ1O$D#0YSObx;Z&D#T)fiSC2x;@Un6Q`CD4x` z8t@);M@^p)fd;)&D!eG+c@+hIr`Y@ThNAKAFD_N?H0HLko}pWLM7M3GDSsk9DeB^d zH66eG;aaT9%sBNC`l(6Nz1pIJZ~vy?epqdsn(lAVd3w-HmuS%IRSwxSS7FbZTu3Yc zu7R;gf0Qoe+=YO7$zhaVLdkzdMr|WZtnKQQsejBcl3x|{ox&1i7WBk))gl*?7H)ZX zGD?E{IWN9|OKx0u@sw*)_`%rvMo{ccGs`r@3*GUu|CCRJOqHu0N^|e*Xj{e9VgOXBLLz= zkwb-`dW2L^dCCRfQ5XXL%z=G~HGubQ)==l-x%q``m!=S65gY!sh(f(|dX?^6>SnHj zKh5kPlG+L>`p+Tb1wjP!3v>0q6yNP8i`N2j=ew41+=o;>%R#m$cYxiZPij$v(AjV- zI`;#HIw^nmkPae9Ia-&j$rbwd= zwFg*BHyns3= zGKQoPjdhT!j?>i9Bu#6%Wq81_`by$?_Fkeep4Q?=rNQOoK>NX&qIKK&iqWHnHWZ9H z4CcD$LmZiyVQ#NU@-_P^1p1^guTD`H0DvMjQOR76U`j}oTOo{Qw4+xlwnXeJ5UIwv z{OSbTK8V4vR+1kW#ygB~_p%*F+)BFL6?tO6&5bs`kpK(c?O44OhWaSqlgY;MSpWK? zmYiV5tR{5}QAQEL^{X{u$_mi5qXwcqidl47TJ!>}5<&OH{h|pk#@itUhR|H;c1D50 z(F^>*NzvnBPCZ3ZFTkj6mAeY^&m!~EH6)^Z(PNL5hOeBCGzCzQ!s;6DNoO)nQ@1(F zF^;aW8uek<0?w}yAcg5C6aswjchz$Nt+_bDA_)bQVKCPv-4&>1L83Tje)8q_@k%X< z%7e0_J7b?=a4tH26D=3Y6vkm7UzNA2T++M_|MqrTi9j*pGEPM1sC>!$>b}p!V~Azd zxOfCC$*gb+Ma3%}Fd@r%$>Nx|--apP%ygYco-_2Pm_QSt`8W`QS_r{@d8w%pVJaS| zk4MyPZK`~dVN`e;VKf-8;Xx$UX!kPcw3T`chMzHrGC�$K1h!1e98>X1zZRbNDQC zfN$z|_9Q<0&-vm4$?S!C(mVwMa#c35Ijpwb!*cX4?^Mjtao@^u9th{%uiNrw4B$>Q zM)c@C=f!%Nui=!vibxo0Mdb%G;7D#$^9=pUdh}PT_>Hq!b&I`^Z`fLO!ara--+hzE zwJ+e%Z`vbL~B@4U(9C^TOizlC}*Bx!Lo9$S#+s9R({> zM^L7>Pi^`Y?TgPDVqA^Q`e}5{l3g)EVqBdmY8C`By`zI->WUmU<3tVcods9pTp8+O zjv>-74pmCmUUal%v?TNx;u~n8&+hN#eWlSYkJ+uB(M30=AEhqDyA;EJS{fq%0FyK^ z7s@O3MX!x^p_Abou8muAI`wccxjS8IVO%aukp>dO)*QiiVTKIeLp62DX+}q8uKXc=5Oque{l$u`JHNbC=-N^3;VLFl`>v7XUKJt1&&K3r1Te8z# z#k9MR;(TO0aRIOX$$o%7?}ctY_5?X-?pKTTCwH3dIr)2=uTK)Auwa9>u}x8}qFDSK zW`DHS#38ko-Bo}E%{59gjz@P}@rqDeix`wkZ>o4c>~IlbbN%_GdG*bqz~xg_g46wn z9c)GfOBsRo;^25cyZDfB{Pn!zM zklW}ABK@}h_Hu^-GK}>!oNDK1X@23&vW9zv`vS`AUIzGt!$R*B1j~xZD0wk~=_sYH z1$Ux~lsQ4|CfY0h+|}X~Mazae8PS=VL85gCDoU^h2B#>vuXig)Ush2V{)sy?BXqxC zgfI24M*SK@u;Wu|e1DRHbd>6E<(`Ci9HC z@e};gMl@FbI@WfDUWKMOaaT7pE_a#F2d2HB);%)5;~(!x*!jQ4Tk(GyJQ9pf92){{ zk8FD3E$yJVuFkh#2^zG>??qzy$|ajw0qqZ=#YVVC(!_K;a@?J4t;!z#R?Q!zSMv^X zE{;?#{U;oO%%BkI=EEqgVNOO|V_W6{$6x96s-xJb%Y#xtu^%aVXFslR*uA26qHNJ2 z%&VVL)uI6h#jtLqE`{o=Jl+6RNu`6TI9{-x*fWI6-IS#Rp>8BL4gbp@YG z6(U7*RZLU*D{*>X!DX)>FTI20ll6|mcaOrNV;}q4S;e0U$J@^~;u6u^m0wqzwr8gM z|KmUq5-s1W=)AYIm)k%y}LqiR`8av3#|o?`lv^=xaxBf zbm!czb&4B^B^=8TgJ8r>;d33V?F#yPLZ8SO!obeo1PX)N4(ZrVg++Nb6TIXlch z2W)EK>q=i$#zIE<9H!g!=0J@^=52iZEYcQA4(VU!N4b0-`7fsQAkcGB2qj?@`b$H` z5iG%LY$IG~{`;3>DO8RbcwA??N_vws%i`f#eC81;v=Up|e2&J` z+682<3LxApiU_N>M@9I#*kNBcbme4M24Po>hr|DPJ(3L5SOvQed+f6#Xk+E3lSP$R>w5EsK#JO&?>Ky6!w zyN^4V8aVph9(bkd8f9(wRJ*`>+Nz_WhjqE;7#HE^xDJOlO(5j6fA^r{L8jLWyZ!tq zT~I5FoS`Q_FTLSKtt`~A+8tx3s4`;DoZ;Wg;-DPsC7pxhB!VesN0b7K)N zV38MfJ@%k~fL64871tdFB$O%tqAt{dtg6o0kvl~_O3Dx1q`0bSt$?@g4E@6Rr+o<(R~8%3qR6DRhuGr1f{>}Zf7_?*aDv#< zqL6SOeUz|LuhS_t&1mjmPN60T-+~unnVsm5iCEjLDu>@@d6S!_+AV;lg~aFpNLEWL zGz=zZG{IpP+OZ6X9s5LxxC|PWf7W7(mQ{0A^yX72CCVwfCjgoeVl_QPhL1UKP9_`2 zB@zsryv)IKf05Gv_xo5e>JQJoY1%^j|KiK4iu0Aloe$OHB>Z>5@;8DuRe^KMdFR`Q z`oHYSzx_TZ4Pv8ab!PZ4dG>GrryG2Jk2MX5 zky-(bIGKNW6#mWIWg>3hiottP`@j5;7P{ocT(N>G6tEco_P8K+gaMqG|1HsfUm`!Y zId_-8OZ`Z4x%KX*v=OKg3xh1JyMWY8LHSeWqj8C_Y#?Z;1BnZG+KVp;f5qagVdUP+ zM12%aAdNR>-0Yg(G%5*~{&Ps${O&CRm&3j|g^pp=eJ|Mo)ks|UW_IKPa2hOvmnSiL zAGx+!u&Od6%YXXI8h}~5KEwO}?4#f6Q&*rlFMtkxtpzwuR!x+-cW{xn*Ua+LrkSdL zDQ98L=Dl-N>O{l{Q5+RWLnkPHD)QXjWL_uZS)EH5&^_pDOmJushL*j$wTR}g!p4Mt zA~-@@4ImfQ0rq9i3-Dcm11Yb+F(E&=teO3Y;WEV5TUXvg%rM{eH>ZRMwW4VQD}r@~ z*BN>b-Ho&)Az^ykJ6yT-8bGwTfU4$Fd?<%Q-th@CM^7&z9{Uh+^`AKXm}(a$5OTHI z+p<2TVAk&OH@qNuS$z>Z20F#g2w21t`iw_ZC~4A23^PDh8fw1WE;Pe=3HUjUpLrY& zhQYsSGJI~`G!ucw)wT!{^FCoKT+CgT_ivWiqqe<1?ES|FF!Yt0P+bzZA?o%9qxMrV z{3yKP?TCigY%buz6Z_7fCk}#l7o(GgbNvxQ^RcDi=`0UGFm*m`23CXugS18Vy~J2@ z_suL;F9U>mMa7r*<=Vg}wWnO5o-APU?qUy!Z`^?;biiP$QdEoo5NN;e_mJGyW4}Jn z1CX=#B92D|m$E!zIR~dRdW~1GRzW_?6B7lwoP} zy^7q2w$xWLH#7&Jn$!ni3C!v^|2C$iXVhl_(#7q|dApfx$J}GyLB`5DVA;uCJ1Tu_ zD(XwW_*g*h?QbLNbhkhKFO_nrP`qBSe(VZieL_}i1R|x9cjYrK0A|H6yyPg{2yQ9} zbzLckP)UI^q;I>VaerH$XY0pzQ{Ln2MPM`%<^*;bhd}Nn}WDqhu@c_#@ zN}BnU@h!6ME-(>I?*SPwA|LXIYD^vl-slE$_@X}sGZCKh;x?=*?Q58`4Np@TV?{`3 z!qj8-A;phYYE?~tqB8*1JTeR_59Jo%rx%6+t>(5MZWbyeimKS~-IInCI6SONo6jCy zUCu0`CjRLgx>sab=DPpp{`&eig?(1Z_5sUC{f;Npo7G_0-%`UoQt*W;Ldunz@? zjv}ax%csO;AWI%lsNwLmB)7kAb@gfBNo|L%gQDhe_k= z{x-N32Q91W56tJ}{0Z+#Bsp}~-=$*_BjwV8sH(@wH1J!Ed#=Y>Z@{eG;dq9{yIU5; zQdFiW{t7`_2DgFTrLdMY#R^jkggp0+bXs`AO61)6g!4o^>w`2R@agFMK+<4N!wDug;GocYNf4mxtja;;Rpq~#^K*IWu6(pvu% z#7rfW@Dmr};!2Wc(`wgk+z-^*Yr{{8vVUV#gi>g=3~X4azK@FlHsw*7mLUG@X?JZ%==gx#Wj?pQ$9q5ao0;EbM z$D?_FC;VDFc6R}{BRMYX`y1* zsS=w%a6lmaX}P_$F!lCHuCGyDU?1V?EN<8?z?4xds(-uasDhC4>ww3NrKpI<6K-1z zTvuEdi3MUx8(`ThUMg~M^8wvFP>e+})`WXZJUH(D`;yk%%d+o>fHK`9(DOVjTb{Z( zUxK{|sPwLkP5;}qvj6MCIUSbF?WSl8#Kn8BL^46Z&mv;r>S)t39PCFe_N;Z|o`{k; z69+p22Z|BuRMCX?BdP#6G+0>b!SUHJ}x`OwGS8yfe(=XMe~F-LyoEr@FbKr z9`=wdcHt_dPFW!~*cXlCU8bacqF*VQ%L3cF$5*zBlpi&Kp5(=ARMWN#oNPlX8Abn zs;Rh+cqrz3bs~}~f#CH)00}_Anx3YM@~EL=ZH)YD!uitN@3^&CFlSa$ezs04SL3zE z9jU@*fkAT_0rm9_jO`viCRzqWf!kHSRT%mZ&rXp+0b7aBj{|Eq&&3j3)QZ?njlN~Q z>s7*}*TmfRMnnq@;7%tUmjGy9UZvUb$l)B5*FmW}YezE%8et)c8cst(8&JnKf$5Pn z0je)^KB3&-kS0WZlP12i%eZZL`laC}{Hoj&;h4=<;JAgaZKy8SXtca85CrQHgsDe8 z@Oc4^Zlwk%QQsfEgEv#$XP(^8dLaI72OYKN&z|BD!Es+vTacl7Rfyd+tg zPn0M3jb8+)cqZ-Y`}#i~X>iDp`V%WTh?eP@Coobuxm3}yDNdz`jk1l4zb@=3`L;;m zm)^Zx&2FodqiWS$N8p$OM&2|e0Xc*xL_)$P{LM6)fKIV2oR5D!O4PrUpCG>DprdO$N|~p$2@4MI)2Rt5@`07qj0VOGm$;VcawBC6PNw^ z4W1FUH~AkJ{DNgSGr)JK zt3Oo}L00Y9LbzS&#C*;hm$=VL8@6B7j_~Yo;g!%C&H!s(S)_8}1IcP3!!H%2c3n98 zAkJK5dza@@tQ+Xni@u`)b0Dp<8XEjdPW0i)q(ZP6$q$)JMo9PLhesbnu^TsjsVMCA zZJ=No-^@y=RWPPFZ0EUIVN3>Q(Tw%#T805oFR6l_NZZH@D}lSQT+* z3E_P!2ybU~QP&N|g-E%yesX2MzR@-aEtvRMA{?I114AUs^F$BQcIba8`jEd0v^y-Jl zc79&a62f}B%DeGJ4fDtOAwf3fi10%)ya>0#TqA^qx+&uq2PQA%DLp9*H_^=jbpd7C zKVIDJxVZjU75OCi_{<3rFW3vSkR3u@dTwjMx&$@i+0inDyo9^tD%k`zha{ z*FXn2M)d9ZD1&Y%|6Zb6NsvYJ$97DbmKEHiZHYqMQeU}v5bD?)FrPb*5Yc*9c{N=( zV|O_p7ta9ZA5smv+!IUdj?LfzQJ}j9euNV=`ng|Z?e$IG! zkKgPwAM^moZmF&`mhLRhmsG*Ux^xNSE$oC`7h)Kl{KRpc669`1!2RVRb^Hha*O!d6 z=-aaZGs+SBL8gS~!8eSb74;CN6-s&X*NLB-MYtAb7zQSrK1uv7*{7(t7o$E8++-VP zNB?ZPJTO1~Lc9m#P&Q34ty0bi;X0|_98_e{I;cDrxhrT|GTb29!o@kq@_~2Ko!%|o zp4GNZGPp6-|A-b)YTPyyEM#3j_q~THd0d#bi&kaT&U7b?@b-rYsXf&<&%b!xQ%m1| z-3#oK{MNugCRr@T-3#M#HO@XJ+t$B$M$sGjd{Q$7{p;%MxL=&!(#!w|pZe3p!PGTJ#pzP3T{EY85-@W0)QJE z=&XU{4DR2Q`X*f|Q{>}2`+8+qnt!eQ3A~R{{8fYR1klFyh-m~j8VgNRE1(m%H^iUN z1XN}gae%b_F4G0=*Ea5(xh1n7Bi6BWm6%CcA5WjnjJGCR)%;H7Ohm7|UVZbbAV(`- z;eEcq{?43uL1bYxYrM?I#GKyk_HFVf{(1bVg&Cumn+8P z!7*s?6g;`$s|?&@ZCJ-&LR(w7#`>OjJ^BLotiI`|%~%Kr2q>#lFTK@#hQmI!N?j5| z9x_nyhsMjQ@0aX*TJR1HK}ff<8Y`gJX2G010pKlv*=PRIY`^{N0&(VxVV>pbKx9!* zw;B5_s%NQh+;V6Utkt3vIHo`UB8co0_1)T-nAp=diY;#^SwgSgg+p4$T?-;fXr?p{ z=l&eyQv`949{Lzi=$dA!SYOE?Q5Y6&*U#vBoO06Zu%HA|E$3CC`uOy5U7z46KEPIV zUmg8LAiI)CuUCHiBGGsQpc@?TdBIf;QeIWFu(*Uk0Gv2><9TCxiQ|2jvdt++vb)<^ zVwiG`dli#=R)D}A?(?<^i$Xm8&`)cBz9wF8F6gP+k5JTfu>a2kqC4;4y@IcL#)WY< z-=P3oZ>xXx%a($RPE+%RtN|@d>O8~xSqg4#W~_WcG;d;~*|}D@H|b6oX|9eqzXK42 zv?>f3o@x72Pg1mAPL=wTo|SxgEvNRyi2#}4I*ZxlEYcs%5!X-C>$~_BZ}_j@Q;FSP zRIAI8X#$G%x9HUPmz3Y+-fDFJAl}1Ox|=VOc-pbyuIInK)}pwx^yV=Sb#O5Z()#O_ z-r`#*1ItSmHVV{9HGKC-B!=5@4aMx1^K`G;1fr~A$9MqJ_!tGm+@6``d|v{{*WW&x zzFa-$=m~b@0jM$!fUBHmsu~k9h$e`A09>QqQPD-J!{(Wcx`7jnC9|E~w@YAF0=n6k zRYeA%R~=`oMid|}xSDHi+wT*i5~476eV{+kX#t3G|7eg!K>>@5hl49;32@~dZ2J<_ zf%m&@Un=CK@m1b|Qp|(*yw9b@sc7sY)OW8pJPBcJTlkh=RU|ut#u4@h!KSQu1Fc+V zR`r!fveX`jtnWs*!*VH3j&~cD%J3YCqy-H@iQO7^Xeo_!SpI<}uvwB}W< zlagsVBSj#0d86;4nAlj>E;?_~R1t1ED8fEG3P1T$7a{f;J($c{eG_+l6k!SJWP_b?UFh zSl7`w{9OZhyOmcZlsj!wM{qIu4%)gRO9BnOlf6PjJqcDd5t<+A61F5EuJ;!shmEJX-{ zufg;fZA*4W8B*A}O>8X7Z3UpygYB_L8B%eV5~Z#=Qp-m95p2-ss=itqwELzfj4uL2 zU2OJ}U3Zp0ciXCO4lxkKJb63&oRs#i#V)o8v_)TFu`hLzaY%nTJV`BSKAYnxz4^Tb zyr8j-vENdUcpqd{9CyC>sQ&3Dy`hZ}UaY_dFk% z@U7P1-hJ5OUu0;2ZB$N?`Y7yq1#1@Uc`RRnyS*z2n2JP z#JU1Tv=ojbErlTN_wB7;CwiP6`D=M2u{SY5F)Ta80I4u`w&sytobw!I?*(GTPcPqv zpScj{R}R*H*v7@_B?ua>5hY9zgr&0uT6IA9pyb|D8>KjL zC|Br*iFK}l_lekCCNsb97YS__d*a#@n*^iq4swPDA64Ywq%1@qO$7!8H53hPaW74sY_q+n zm@SRH7@-y|K$&2wgBJg2f}Hj)FVW9WsUl!UgFjM>JLB>^aYlfhzCk`bvDP_6A@8c? zsq8tShEcR*>ojJ`12_YgiY!~;T50sM#GMF-jU5zTh;OSrnsMJ|&!W*F?2Z)0OtXQN zUXTD2?o(%ZK+tf;Zn#dK_W)N$)Q_QbGX{_hVAhCvTjGcSNmP z0LRJFjh|4D5T>=16D@#)@wFS{p-0#ZyK0rs|9ut!SN^g}?7i@_FCkj#3RAs`TSEum z9#nN-M5;-52gm)@$^QfO1j!wD7-RQX5gURlz&JTV)V%1LF5g`= z-2>EJG7Cc}&ESkDd5JPUVAsF40RndOWs*@0_U}^J#aXe?DSXsjw=4WSo!xH*43F|e zLWCaTr~^h>)cP5AJy0)r={5$;eb7=t{TN}XAc|4V>vzPtuJ#2V0t=@?!tIW>V zdRNkEdju%JTv=@xO?k`FFu|ly%|6(Umjb*!eHU@%2m+sw#~<+V(uBd=L(23s)oAc3 z$wdSyjUZaE!@X2)e>Ie(jFh8cog%9CK7p>QK`#|D;GLOKz-&a=6`&Hf1#aPnoR+g? z2iBk^EZQ50*iT_ce#|n$cPtC4{1tbRUV^4MeeI~g)w<`*Dvg2`9E|cdYXRdHS%B}O zW-0xG{M<)jWfl5nO}=I}6N>rsa!AcXK&B-$q2U94UzQgiAed`-wHkcnQPq9WyLZ6Q z%=T=a9%7(j!?AYt<)A-67&XbF4_SfUvk@~s?zu+WPh(8;3IPYn-Yo;30prOMXx};3 ziL-c*huC>llT|mFQNYy|dw?EjF)zcL_)*a6Apl`^huc|;A3ny3qw)kG9VHj2&pJea7;z<_2Q zXu7Euo*3@}Bg>y3zE(Kb1ds=USAew)1>?@=9Xe(%{6K)V4yDiiWWemcp{m?+Hd3i7 z1;TtAyKM$Am6mHIfFoo*Qlo8rUI!{T!hmfiWUPuNg=j552i2NnTy=H3v2%BDBUkWl^g(O}wj>Bf0cj^}V> zu`+PgY$~ESUY6GXAwIv{DdjDCk%CTV3wjjQUP-F>Nlb^`eW*7uLTL62%mbhI6ZeEcQMb<+!0x`zbP8_f_{25Gn4Tcl3QEB^I6oP@f zmZU-a+6R7s<)s=T)fXlxilt3+oFFb|$Pn>nUnGCl0*KG|5kIp3E~FLh0T}IOAj&C& z6L_VXbb^3}st*wb1RKX|i0^+$zOPOruqcn2mPl}HP^f2}@s(=Xt20uEA)IRwg#53WNUPN( zEr-gJj>K$41b0Uak@Fo7ttk?`GLk?_z?K#4Hbx4=OD;^`-Or3gq-2oz{n|hShDQMV zh?6(WxwL>#+7=ayI%54asKG$DQr{?!1nCN3Q13d{<(}os>Yoe*J5@bJ9)jVJGY%bt z2#UR4;f;Fm59e=zexNg_X945eUQ{o`=w>hgNblRK)BATf0v+!Y`Ojvlf84XEwBB*) zkVYg#N)+|1k9NW-!AZ!66um6~Jk>NNp|wBu*-tn3fetQ9L|up8@&J zc#1yj`Y818XQGT@pl)Jh@q*+fwE*OyNenS-vuA+Q>twwbm@;uDERmyViBjQdwS2*P z3I~(b!6X2ZE7o}f1tn|E?}AU70_EjA8UX=DBH+pv0J)kUH$kF5pRcs(T0zAQLm`vH zXu{>9+0fB1j3Pt6esTjQ_gy-Jv9K_lScg$=QFAqjTo}^{v@C&Db9-t&NxqXqgCL~c zt)5_KVE>|*K5O|Imu;q@IomCYk~_l*0WzL1$Lm(m6+V6Tm2tVzw7m%oTDI0^u4`*< z^WzoUv|N}R^bWj#x6T?yJ5KPiohPxQ z304PR_0&#$sPaOzZme;4z|;F=?|FJawGToF;7mKuYVM;7`>f%5A~z2dmq`k0n~H%U z7&@2@wgE#ytj}A(xOfBT2>Dx6=%L=YK$JQCrXYZBGU1p>guaqPuc<#(J8#`M%n#IU z;jK_7hT8{b3+#JNz?S(qO(4P$ebDYzBA7gDflOrW^{K}G4M4R~RQLcI z)bX>PZ2Fr*7{W9rd=Ywd|F=az)-H8GzDE*qGx+R?KtZSL)!M3hFQZp!h%ueTUf%nQp8Op%mt)M2yKU#g9>EJxu1Dx|e;#)%!xEn17OP_3#jeeH zAciw9a^;%l%nPR?iYj$vmcekMJz{2Plb00OssmXz!1`_nVUlelWKey}wps}<*^sMu z@Y&wy2B5gOAjUMK`iZf(=GVkl2CBT~giiu~CPWARsy%T~^h`k^?Wqz}({cK69xQlG zNoS~4{F#j7lOk(kw6OG2x(mBnx%vs9KncHM9hCCwgb|kg*bCOGL%#PQ_HoDrZHHpd z^~90VA?2hmP$ zt|COwPqNx>(73G0$e1)kIzuHBeVXu6g;(#cy_65h+}5LoDYcS5N2pRGsa(%2SK1(G z+%#9lF%9&TAjk(b9jLl&V9$g2T=kysiGG4)Y_ZD-mzx&&P=pTlLgx&|h)MJucAjC5 z$-Bfk4LBr9!*1wpwX zPTKbeErNKyMA?}qYx8--!@hK?2Xrh%Cz&NZROzr7f!9`V=Xy`V02ndV=y-eRyR|m9 z@tOC?p9jD2A+YswCokB0SI%o!^qdg1L53H0YLdLQ>I&2C*?oDN zWWWt{rqBaks;wOPbBQucG_a1^+-0|dl)e9{}w);;A;M&-Xvp~nGW1)SaW+;OK_jhXH?xKH1$ zct~nEjSaC~ah;oZm}9*Oz_`h3?_o8Qm0NX1z+Vl+TrH$Rk>i^K1*&fBiF?BVP=?M1 z@>hIYiS)bRQ*66c5DOOjI-fZZn3jn((|Jdp-T({_aq-jX`h@+ld*$3ZzxL= z&B_78?V=|?D6H{JuJxmpTVrdy$T?az0_l-?6a7jT>vo7zed+d7HQ67|OZ) zoz`^#$RoxT3WYl&OV)dd=6Hbz)hw9neE^OkIxu>O*TD8fw2E~%2upIKOHjo-Lt+l2 zVpCK=@@TXuxYCquV(Nyk!?d7NICQ2pNJ}yr{QCPH>=iTTZ3q>rv+0Vz=iU5Jo|qEV z99hQuz1!9rk#iokfO62P?nUTriog`qCzi`*Xg3RLHhCNP&OcAHjnC7v`+RW|p{Jen z1@`V$42gGJ8ZB~)E1-QR-6~}KJ?}LF+X>`^GskS(T_)mGXf@h$C$h2@k8LIfElbW% z^_>h>T66SAObyLebaH4ACBCPO_mI9kD6ATIxsf8|i~e$>1%D@HMPTJsg1N=gist~F zh|b^tiZH?Yy(I(%?OOO36}_;Bzc*vVR_;T}Gu|9s*9bz`NyiJcKS~nOag0T`Om!hJ zCts~~#FF>jkxj&sr}&cG--IN(f)qW6u$H%~lkadZ$#1ARC^+g=&X5e5b-49%rea^6 ziYE(AYynE)s`sPwgv`71m1(9oiH*hn(E&KEhA?OLTRW~G6Y8}_#t1SQ@ZlAs~Vp?P? z7k^%;?n3a_|0OQ_6VN^FfyTm@uny0lrU5&?kIP8IH>p;=#4u^bK~|cGVw-c+ zv+}gA!Ee;I&*=}Cg|IQ6z$sM0YPU+5^N*hzif;k$ll$r>Z0N<^g9r<38%3KW3yxQV zb}oK!KK@BjY`x=o^1f$}M91V_F83f^?l*HZc34-GiVK@mS#7}&id!~Ex6``&+d2p) zf)`n^FmNZxfC1GWtkl-%#L>i$X<7txeq76x`gs+F(=^U9RlzD;-yQgi%udXSez8Xm zp87c;3Zc$Aeo@c0qB zo$&b(sl!r_C5Sp#3f(V#!cJ!+rd!wCY z3bUuE+~j@f9qWJKE%GXCD~$S9im9+X#BX6k>m3m#2LJZ5WB9QH`^J9om53VyNgn7< zmD!G*ADbl9$S*)GaW}~oluIsx;?21LgH}P)A&9@iEwY3G&wwO-fw6>5$Ah%T^`(1l z9Vl)>`{`pRN6^^hPTV@}ITi*YA0Og=V^{9{32b=NiNmnSvUauPB9+B1zU~T*0GHmV z!f0(DPQ2W)Gz1HAkap;3{GY%uzs^y5Ol(2edIXeH^siDxLAmg`fJ(%sSz1SW^k2KA z;G9$Q-W-&o;v(_9J7OYPfS2U51_B+$jG|6FiacYfONL4A;}s>v_yk;_9NIs@b2vSr zYIX2@2^Fij@LQGhikaw#NLLJQ?zwuyv5j>Gcuo!3=Zfy6%_hM2tS@E)w<5HCSAKT& z@f*us(6mY$3QiV35;e(FfK|Qt@_&RZ$t3;Og5mCnM3Ho3e{8QeJl&H@)h7xoi#|wo z*jldmYg_bdw%HJ%`95t@1yXJN5$R|N8I>Rky_~e^?y&F@Jr!p0#@*wC!*DY0m1)WJ zMkkRPW5}oI1IgCEa5=-QvZ3je8(CdvvDw(@LtL(Zeh+ZYyAD>d-cV1GP((^ovHp%; zC$TSkuBXy>P7Ztr@bAfuCKdTHiQ?Q>lr2;6K%9xq(LT&=E|*w$jtyz=k8Dqn%?1gp zwqOxIc+;8{tmUM3rW;_s$zm|DUgK#!04S{R64Xh)BL0k5gspEbd9x{}EhC zLxB7z-Ko^=FWvnc5dl+Ozx1lLz4OPoQ!6SQ#Jm}Al_u**C%KA#sDvY&IH)yhg5SO+ zB!2@Yr=#D=I&X2TH+;$#j>(0DjgE6-NGRSsm&aCPO3NYF>%B_Mc!}u`$}FFY9ypE_ z4jmQI?X+S81+oaO&wux~wKm3?r#im4*rEo;{FyWokL%pi@z?&`Z@9&&K6Qb@v}xgr zOTzix;sfl$C0|F6BI{(WP+JK;by5=ys)X>qwtq5xZoD9!Q+xly^R&cIhHV^o7jnluOu!l zE)BJp{jR9gGwXEe1Au<#(45(52f3OP~t+Y|jkUJqygP@=^O7LrRPr z&|drrb(CRFFoMp>htWP&>48xEhj0lVF{~%q%~ddn5q!ouc8LSgEAt_3pMbkzH`x*m>o!@QZ13)&&gx#wDk|NC0)%VAyBNy-9!(nH>%**AVE@@PD+ zsN2XNiJIdU;EE1T&LiNBW|W~BmP%;LvVKZpZNAk%3e`$Yt|f~TRZnKcR&z2V*hRH#b`;Q;`_7UPAl2Nm~HJ4xKbYic!eDO=w}&P0It;`HKh>a$cHr|D2jvITT5 zOjkX(F-r>>)jnf+VXM#Cf$Zx@51I)umB^{?9Zqv^p~D3?ZRCY@VcV4t1XcZzH!!tT z{*GAYrZ%42#;NF&IwM2{6AkTwOd6w(iX;-!8AiLK>OPnQ1K-eEFpWfTIVy4KO0thU zdloqAYC&Y95N6JYb9@Sbqst;dvrnhu+>EZR$ly`v&S?qjo@efv%s?=1?EPY;3=*;8 zUBB55s_)eDn>5e@D1CuTQpnvfgf$7Itw#k=ak&V+{LvvUqLgi)`UiayF&veyl*jO@ zWHE}-&<;RI#C>S2YrHtXk8^yG+LJ-Td`G|r(ay~joYRXq%Qp6okf(iEth*e1+`Pxx z)BVj9Ynjp)lI$&C+V}m1PNPJ!Bj(-vHg1I{g5sW-9YzDo%sVC74c7P`4))kPVxjZN zG5poil7x6AS3y{@aQy7Qn&0SU$?c^!KCHx>rs(sXWBtN^&0{baNO3E)+90FKGMM_j z;|tJ{U%Wy^&e2(?O@}YfiAo1lWhs*G{Nq(9R|k}_K5fPJux8>_)ya8*x$Xgi?(v(( z>biWZ1*We-*kUoSld20st5s@`=H9wHHBise)|qj)ub;<$_BBWKr-c5xtQUUO9p?R6 z)NBtcjZ9cE{4<>&`P$-Tgv9%mU5Ab0#`h}^pU&ZPC5RrAM)9+QKi?prc{JYwl8 z&hL{|pFp!JAiw12J&ocOkv(-&bF|c<2 zqM{wMh3j{#>X3g5=n2y#4|QoWjAcEVWV{IZ7)C{aU9&=_c=hP>6}a~sRA-Ufb-;$@ zaN%9y_;j?hygQJ}M!l}ji6cEQ68n#XC!u1B zlZusY&-J7mRBzgYU4z+_`hOSoA)}^FfHZj@n7(qq+!H zNPqkXLq`)yxgG!)!7uZ(vS}xsleXWp^V>TjowAS51$ku5YSM6#OWcZtZCW)mH5#Tp zS(g8?4fb}qMlRq%Sk^pE1u;hP%AB*U4Bb&VY4 zTVJp^Pi?4{Rv75G=1lT+K?VfS-x#B~5&Y$uY1_%BxgsNvp@qDsiudJk3){nAAs&>h z+Do`^yhYx>Y*T~8WRxpzSUtaWxE?ajAY0rB*O~6M784D$T9Q;;otLN^?pn4%Ec2~* z-YRB#y|um@SEZq)RdK#kF?UP9C}sG-X@f`LEhhgz!$jg{JY|y)3Yk2I3m!F@kD%=@ z;wvn5DSp>sZYyWmiS(!;dv{qM2RaHG|(FO$Ksind?-4wnE|jI31H8T-mXgac z1$iST-l4NW#8rwkGT>T(Oe)fQxNtNAG;a!%f83{Er zXACZ%)Tr!suFO!B_K>^tPqYMe-1438l}*^vg2d|p)e|X@n-$%euT6*RM`x5o`L1|V zAHjvkU`<(NGYfra-n+mxv|~nK)1B)6lifjRC_+s5Z-*oc#rx-hE(Yyy#0B#^e+q!M zu-Lrz@YYj6{xY3yKo$*dk}}U83CKCn(!8sqiD)*BK*@F5*3aDxE%<#8Tc3(x96Ymx&BIdDZ5P!x+SU zK7iaGiv!_;oL3@u zQ4BMBXfzhQ?26yRo=hIIs6mp= zRiYu^Btxm?lRx$P2 zokd^zo+5KASW){>b`r%&P(u2%fsgmhW&}XOX?TiK2UyH`l?(xi9}-J^3i_}u5@mRR zxrCi-t=1m`6Y*}M1M0y&I;~qfVB=b74FB%Wk35MO>)s823Ays`)nCl)KNTg9)0h?w5ms@Qwo=POjjSv!C<6r59PA9T-u$eB})MYfKPMfzCXgCCeS@v-K24yT5|x zA3(#C3`96|(omP$Pugw#LDzc6z-C)g7ae}`Eer+|s1mtDEL7Dg>(ttR_zG32o3P{R z#Ycl-Lx3RsH_1qO-TP+#2)@75`q39(HSyQ^h!IXQ6bqAxO6wCXGJ%dH3?lQ(ngurQ zQ&+-CDmWT7>m$;8n*%Y?QTJPW)bTHCk3F*F+!=yNmX;W~ym)srcYUZ@3rqx|FdF+n z)KE!_6_)jb;MFsx@G0=}edq@XCSU|jvR_8jA%3#H7Ws-&-UVqk!@=-%WZG4I_+Od1 zq~B?}HkOv$1iS6GpXNY+o;>1x=r)jBU8O45QT;FfYU5KbZ=vGg~Z4X z#3iAsZL}P5)78?}f31#NNq*o0;8kbbBln@~V{4#?NW(%NrxKilKZKxk7IhOh-;$$!ug7Vz z?x|u86z9ZF1fmx;CL-AK%D2PV7*lRwcCfWX-XMXFj8*lSBvCGmz}1*!ccKOXWXhDO z5F!W^PO}DV``=Tt6XoSN1RBb`lSU{?B5Z;WhC6$!4n*HS0zw*Mgr7Z|@YjSspQvF^ zq!o<+bM?E%802UdbcpS2aFeAY-SCvKmj`Dm(T8gaCNhn_Q%y8_mh?f<(trDwM&aev zk;+t4&GNi@V$M2`70v(bbX^P`4G1}I%{iLzA?eGtCXp!3o-44`H z4{d{2=sOw%uBqWwhc{S$>$y+fd?ho!oze{JvJ1i-LnhTb zm536rLP@{cA5WU&rT=}%)VZg1I{X_gcu)dSW1s>Pp~bff(CC#MS#-Slbe)ZI2DfU> zg}^cM)6JmdZ>&zPn3O7(IUI<^l@Q7CIu7IbxbLDBMHCWwGRN9A--4s@7< z7^U?SdA(nE@$(m8yS}5dNmOeMoid7K;z1;P4ZvXgp6;N}rpC_H9@`!y5r{;nb3e#0 z3;tiUI>M76%9MXM)RmlS&}`h#A@K%x@SzRE<(yoV`>jF9=qtVMLckr)mH&@xntX{rb?h6zL&Xw5dF7?I@=-~?P1)bUSxAY5z!>#8A zVxhV31?@*khG3vxPs=f85wwN2aZfmBDXsxY0_WkAj`;HaLCu8@46pH*WwQ($UVI2* z6~(bUf%cuQ`8g{PcTxKDJyW1O~8SCJx?W&OEh!%pA|G654Q$9wF z$P&CJjJ|ycWS2-5eq50aBceuFo>Ou%ef1t*W8J;`wut%=kj@ZMIw|u{oqt|q zyX@l@vk5MI(Yyfe&mOdb|L!%jYiK%9=w_OL;um$Zorn3Y@<;W~n(+M!r(slHGz+`L zN7f(jQN4z^FO((bcu0t(K3zN|z zs>X3N`mTBg@6vnIq_{jci7&_Bs)A96|OGsdxBVB@VN+W@kojAJ9YF zCVce{ZtA zv`lsL+z_laP_nq#zE%G|Q0@p8XS15sHm88*nArK29O&R$WW&Idy<4)#){UzKp$R&( z>*{^sjLHn=bgv`_lMOBExI-7lBlot#D~L*OvLx>ZBZbX~xEz12Y-n9O40;gFFw(LI z?md-?R?410FN0Ou7GxK-k_rD;k?|k{qaCL$J48Ix+E*m}Fa`a6^q>p4D1&q0r z;E2ZmEUPh5!q$E%S{=Hi+^SG}#Ghh3z!up>%kwIA$s`ngv~r$?TJLcKg_Zxo1nS-?o>Yx; z6^(LD$92$RqxI5tTz!8U&ZCENhX+^v56cCz(aL@g_?rlD5f#Q>0n+qsUfM0WUZ6US zXnS38P@Vv0$TaPAnwK z8Y==KD=POqmevwZPmEnH;W$t9~LMUiw|Eb}y>&>I3Mjo#t&@3Z_a|VRoVE#XoB*24`E@;x}U2)5Ye0Kv~7Y zXDBKf0`$CEbrDEd*8o2bx8!!2Xg$LlU`g3pqC4pvzf$?;5YdRvd{`|{s=#jml94w3 z45Ju1Z2(@Q!D!M_=-$Kqk4WEQKgzlGGhPW};D8hE7H1YR+)tAq1E7sBSdw`nu(KRQ z7N@$qqtTX+bpAlRPnt=p;|*ubr-ozpS1O75caWr@40K-PiAEnT)p5QqhK-gQep6yx zUA6aC$l?KRS=*l7Sc)P(#g{Cffu~BFkcCL)=#tnK=cv`D$+k7;yJ*c9#*s!;h6We% z&R$y~sL19s+r?OgCBbzX|6t2vw&RoV9EYn|SG53ECNQjGkiE>!%sx06E}3O3ZP$3U zSft>Ni436xQD=2#;*@2{IVQPHg||W2EShDX#u}Vw>9? zH6iV7xNL}uNiwqB9|f>kfhFwivS~8=wqw%j`pq9g^;fvT8S61A5*&csEt32HNY-j= z4vmq6-M%fa=hQ4*Iv=8=OY>{u_mH<=R9_ZUErCY#GzFdEs1crbta+FtfIKxg_C(zt z>)6!N;jkK}h0$(Xxmaf?ma-;RX;>hrE#d4oKI1YH8K1Z58wORNzP~VjWXLt{{8l;4 z{COoJIrLA!@9S2pEGx~lqxqCoF;U8P?Co>Geg3Co(w;|QF(D=Q4cjLD?MHye?y7*2 zJ1X1)2k%htN6~wuQD`=1`@JOlyw9JKp_krO9o29?fP?v^o#(GLbN~#n@bGuxro(9` zJHI7g&0RPbP!hU1_GE8&^=c(sp~zNsUku}3bgZ`VD$kT-_4Ng~y>%C+8wD)rXwQuz zA9S=tcTwrwb!O09_0(V(r}G*e{rk)y_$l(Bu=EmOYQ5=jd;_cX8mCw&Y8{VP()D64 zaj6c8NYxCe7I~`8&D+;;PN8}_^{)VXSln3S6F1%B(v=-&Gr;rs-_w#$UAM3zgjW>E50vb0Smb>nhuQEYgeI(s)2qHKm=WPySI&49QkmO ztLM`L?bp5U&|w_FNmUnREw~O77-^%3d6deux~=67B)IF&(8g*QwLK*3wRxYNkd(6N zGE#|->NMAW_o?U6p6fS7%;whLm>kW-Dj*)R_3_1LFtzO=np zt?1e?Yp@-i@o}tdt?sO*4ZSe$v~AOVM*yUF{?i50WPx;Nv?>0r>x`3L;WLA*gv&rt z{W1`#r&{^wY0w&YLw?*e^R;++_oyijd$G~x{?^33{_>_}T9@deHqPzcyBUip_r*35 z^a~5<#*>C1QO@j334+^;*CN6S5Yx<|Kxce)UAc*m9oEp{(MCom(?Oyw?dO6wxMq-?*y@rb*$pQkeKkKJ8&@YyWJhq$8 zBco);xBqS@Tr9NN47D?Yjdzb-g&HtI+TM;doOM7rz4crH1$ctZ@lHblX@$T)`g@xW zGp-7P0WH4P(c3%)qC%?%rs7m%^RDm!nk%M8)b>hd>jXrJlC%!+P!2^Psqan3uVXLV zAfv*3+BxHOvr%Fr3!<@ecPqOdC`+Mk5E%G_FHRO3-{K1{K^yg?AwbG64bWU3mw+ap z41(Bu%xRZ*fLYt*ww(5PRlj%46vVz-a3 zp3NaXxarHz3qRlBR4kkFY}E! zzUD6F)4F9b)jr?7mwr5mEmr|sIS_+jMIk8)UD&%*)-G!|=Wu2#pL>g^PDJH?JsWmw9s6EnW{4%$9usSS{L4EEF-k7K+%;6?h|;QHC+R#3INkCK);t2-~4B& zp;bkF*TK}yi3+y*%$-QsZNxVcPlFq;Bkgn1xu!5*9)B|5BKwHM1V1%eM;_@C<1U#^)U;Y~ zGCRg*@q2RX$xuLSqU(elqx0Ri`ko@(@R-rN+t=D#O1}oe>GfNg@h?nYc(Q#j5sL#i0McASxnr*dnpfvXg`to3(YV4O`8j zGAsP;NXgxIH(6LTWqd<&MUu5)+$M#uz~WXCn!QTz65U?;5!XSh{3Vr&4j$4uw_zk+ z&^asivME+s37~xk%cnlB*8o4}ZSV|M5$Y$Md&h#5`1K#+DKM<<`}v~pb%tG51ANHW zHA}>*iHEB$hXhp=~G!O$(_g;gig0LP2@3(Kl_{5o$X9CGt(yCc$ zT@O6LOCb5xEC|S9o&>a>P94GWRr#i*Tltx}DO>r&!Sm^ur|RDq$<+#kg`;fS2hA2{ zmmVcaCplc6_pXjeY66JY_zub&a3eEeoFc$3S8aRJ9D_HDFLNuKd?2?`tlk*!Bk;7j z=w(_P{eWlY5x`i#w90Gi9`sOHzgn-}8iO_5fR3GSZ7d;*^T0|(JQXOQG-Z3liQXm` z2n1B1^Nms&^4UJnnucp#@4lN@SyB39$mWx1#!oGq)=e3n1hGJh>^!%fLyT&M3Mvmj zOv=Wlgc_KHtE{c+-eKF7`w`chfb@n_=VnAwxUX|-IZs=R(?CRLDTUUC#pM?03#>Fa z3m`V3$Zt=CHNI$I6hrOE;Cd59t~;@PIBkhigG7@=hjZYoytY{#apE-p0aFI+?hUCk z^ZvqYCP0WmD?P?9d~zMG?e1VOu!8P8qkH;4cWBhj8!%B>ASM#NkhHT2{|Jow{TRG< zsP}ZFx8U}yx<*x%Cq}UAd9ZtW-v~`ob9yS+sRK4E>GV9DkS3uKNr2$ZGgf0Ce3d!($24DsISmm@57afJ>~mlVN$> zR%HhNO2Kp%TLKbQatk0RCOxDj*;Xlv2Qs#a95jV!cQp`6Hn<>Qkcj)lSQUB2$aZI( zVd-#?KAW9%YX?^8k;zO5z@36lGPu(Tx3KjGh0t66Z zL?YN*&$_ZjdSWbD_g)1m!crwcvvIi3V~ni1bw9qoc1`wt!(GbwS7bC|f&T^5yO*jq zY*@VPBEVp0DN-GR^IT`DxL%@rg{Xag(*=w_oG<#4e(QY>^-_1!w*%U9G<+Y5eRZDB zn@K*7K4>5qlD1wV5+q!zz+KQs5h;Ir@J9pdKhIq>81H?&3#Xaj96%=Wu5MNA`=KQq z%t*h5?T5M7B%VjU68VT_@jwSU;>s4_=!2~>d+J8dOfzk;g%9EDW0CWYx=;(5w3yb_ z9FX~SDKr^nU{q{*3Q0+G2`j=*OsdIyg@Gzy!NPmnOlq6%%+U(KKZDk%t9+bee1`fk zFC{~cm}E76F%dsh+E6t5M1>plp+2Cy41=oHOzU)y*eP9wj-CR1i~~uj=8T+XrSx@= zIUPQS=Eb$5W@)s~zT+RLGdo;(mh}5*ZbO3gd1hh9?x4W0JXMT_ZdXZl)G~C~k8j>+ z3n5o_Yn*wD*x3r8pUbm>hh)cP^C8r>t@e@9+n79G-MEh_*C)z~3Vk^kzF-aJrZeat zbj zgL@xsTWT8}-NU~?zULYjgC(NoJW1~Wmpzw}t=3pR-{wuF>6v|rQ-N;MSwqdN)#o^c zXULzs&0zTk!cI-hQN0%YMV`A)GHD;RFkc^%i>_W#U|ETaZm0^+$qQuHS{$Ak{XbegT~XmG4n);zqJTTKbhO z5-2oe4zQc0Kb!5PIMoIoOHv(p?ezIb9XT6oFJ4r+|Fd#9Bh(=cy~1~g1-a!u+o**u z^fi7~UKihjd4TsDC$$M~BWW4O7OpbXXUni^aT{T}xEl4NX^W! zNG`r-s0d?Ru~#8#C*g{Z?v|z)H4OPxce-1Ck$bg{i~-wc!M`$l z6rI0#^z?^{mVhU&C-hxPDeI?aC@X{G2xl#LT3^h?MmjaEzbsRcF*RZ<4i*0M$_#4& zUeGM95h)Wk%{eTek!dX>h{})nV>7`NX?$3=T*E~-gGu*D0`XT-AlkjlD23sJ-xof+ zpAD37?J}k@!J!uHm`sZ*q{mYW-NBi$#m?dwbCQrbmkhP|=-I9&Z+>C)2Is?4gAY3d zIkvFtBzEFNw|cm0R9$D2lRBp>pR}YolvhieUm^)Yw+BR`Ss6SQxV6UTDLjCdGYw#b z@DBrVTC!^*fED^3h$YArqkES>+;dfede6+URbY^(s}Qa5{O6y0w!HT+^d{P0YsSPt zaG%Zx_5pHY0<-b|y8ucrD4Uq^2M=88xGJE^TIAAT4pqiYkImtg`uka< zKg^!kl+f~h%#m=;=KkM26A$m=U-RuXnJ+h&&t(ey%VTMYQUyPgw|LJL0tm!S=rsYqr8Uxzm+e6|2*%^gO*8`9lESSC;4S1e#WbyIwg zpo`j;{d)S{dHc5LBWY+L)$^V#VJQ{o{@VAUw?L;dH%XOq+O?1^foZ>tyQOZY0$*$j zNN6#9S|q0TT=-HEY*}FV5K|(nYMVKy%kZ-qTXETN#s0NqlAU_1_^S6Nv3E1=B6|^u zEeb@rv0%>GvOtO*i3=4&Z*Qjh*`4YyYRPyy*Xa9~@bdxV{l~I|D(&i`<|kH7^Ccwh z;n0v60DI@*+lAWP8IXBH8MBt-YY>W|4`8Nb?(Z-4UM z{r1b0s^F*<$bk*$RHGpu)0-4d4XQz5YIM@#`L$GcmnhxeK4bnGO-GmxdXb^q(F8x1z}x$CCVc=`{g<&vbk|6OD`kb8~&0krh? zn{*Qs#rdUrk@W?D2^9inDDT=z{AXJ)uXmKLk;pV|+fLcqyOvwNc=A|j6u(sdVT}>< z2{HkC6b5@}W$*nW(-ZNXa`6|wb2bN$NRhFmOj9kuk~?BnfrTL6k*10}bjS8^wx8;H zTbqNtm~7RQY}UwxscAHw=D7kX7^GL^s!T!F5DA?jn_0f245mF8+kB8KDu%SsaMx)+ zUCgi;dl4m%S?Vtt3y&qHDyG{0SIio{$vI{J*~E$&Hu;B4I7mmxt>W{%orkaSY9CFz zgS!XH9|y%38YZ_~;n-rnulFCPsrEP}3iwF8*9G8e=;pHpFOsF_7x=PQO(+FnA$+M` ziRdq?;wfXykH3z}IsVvzs_n(~Vajgzuu*fCM$I}3+$e2;u=tLinR~{4$#Uv3W-ZWJ z&?pG#Bs|y_r%#Fsujvg`XX!zix#OP_TTc8* zFOsjb3u_TcaUO+YX8QDh;tMMzG)g!6_Fz4eHvjqUMG+!KC?f7 zFB17Hlx;T6v$BEhg3uxB&wTVJsEH5-hD;e&gXvjp5M8uDely^YoUUsb!`Y_*c!LW> zK$Ecq!-pv7YF2vn7Yjm-W?wXX&X`o*1QO&Juod){J*+&k-W**+z{(>K+oHiP1&TO^ zmt_0kp6uc#>VS%YP1To4)V*$h|6=L^W<;jT=-KK0Fk`Gf7n8=`pP?GMlrBf#Cfk~B zKGlBNxi}oHWEZ;bZ9>VKw(llpAnOkk=KPVyobg{=nu9}4T$sFX3D0=QC8QH5b(MAo zM3HSX=QMST*PbRLc;B_Er`?pJv>sbvYLt6?XE&x3O4^wYw}gW=3-6=KkQ~ad7#$5v zh*h6%7d{l$9U9RZ#mM5wsXRRjv|wwA5V=3~$%;LUlUkaRKVOQyI#$kN0nHucjAu*q z=lx&ZwL`84O38q+aD|seMM~3P15fe#<80l9RtAIt{ZMsd#2PF1j<%$+hD}kMHv)V6&4AId!-ur}ruqof#SErNVI+ZK6)0`S zPCM{e7NhH_Ru-))At)@7x1~8$i+xXTXUz0raEUu%mXFM)lvrxXZ=C8%IS*uS+ih$! zjM)53wQLZ+489-#IrCnwC0XzF9;>%GuH%pdK0u2S0c0h=HVdk@`@qG%-78{b81Wk}N$$GLSD zi6?zY=HRQx=j5%vT1b$2u*$!He2!t2rN-k}bjQ^R5t6E-2KxTuXKB#`ev&p`8Xhe_ zFA^cuM&TmLg~U9Yy-D@#i^Kz{*bA3pAJW|Q6>tV<=J@ z*L!t=F>Wjna%OEDR1{^w10v@e%zp3JeY5%BKDOf#Uz-ubXk3^(*JVBv>n|CmnGV)k zmiGg<-zJ4x_fH9=UKeRogrq?qh&2VSmNl0xLzF{^hw~tpTN&w&bW|tGpL5Q->93JI zFk1s^m+)TpCl1R}d~St4;mBv$`@hY4=6L`gbp!lO{4VK8#vXBN8p0x%0_n$hUe8Or zqA7t`0FD0~@X+5OO9PWcYoG#_(H@5K&Gd3tkq#}SrYJoqoe>uA=S#p#rya8ua40Hi z+~d*`31j=51va-}NUhXh3d%OK_|zv+11tWn9en%#L(m!2vLcc` z9Q46NfjgC)7T-Cc^`a|`OQ~Sk*k^88Txp5wFPu=^ZQEd+!(#QN$Aq7L>ruZ>jfor< z-pncs3H0Rz7-ukVMlE@9h#n2g*&$%Z#`*IA24gZYX`+pqRj>I{*D5 z&sZzlSFb~5EC}c{MB!wcQ{20E!{An-x(a22$e1p}@6p)MpHM1_-_2>#w z?2`!(jYS)5X_a#jwxnQ)S}LPt4K4~lSkXS)82-`Jb$?v4m^Gs)3YbKUZr9XAj?& zVqKUs`6$w*rO={th&HjM6LAbM#A&8+eIT#SuK-KlmT_-Tqm1CsK#Sp|Yt(GI+AsT& z`%sbGlhO5V}WTg*4G8)m@9^!~wJ1je~w|nUWTO)>44=XP0C*b+y zfTlwxbG(3+IOMVOd7r%yzQG(XGXK*!N}#hNjh7(^NhC3Xn-q$Q?62gj?c(iT#qKG3 zAzPHlgJo9JJ*aFQK{W@U>! zRizZK>Uei%$j4=dV1Bjf`t0R<@&kHWC=E=)03vx3adw+RX91?n~td?|Z z$UVNvtIzW}&-3M5BeK}$f5&23@?PWR*$2F@@?@5gHW_7^GbEg^<6^wFa@oIE^7N6< z0Xce>iO=F9(R|(aTFYj3REyJ>hxrnVHI*)kGqUec_v+6vYx{G)a5zLl|FD=%m%H9= zi9a26)LA)U+d(tIHHx#_+uf;+ZOnSuZQQ3J$P12~Pt?(Q%b^;FveNv#;>3c840)Qr zD1Y%G=*f9GfJ}8fE59E5AzL@tUq{`xH_X-2$YR&N=lQ~*wj#CjBu17LMVxeFQZ!p(j9p1h z)}}S2S1gj+@92w?z`%Yc61S0u@>H`LDNK?YbNL=3E~#51G?p7eXP5Qe=L-hKaz{I`4^g8* zqFYi;4lvzhrc?BPw^xGUZPxreA`SkJrn8KS>U+2_As|u;A|Nr8gdp9mAn}tBkdkId z>8_y#L|VEL5G048JEXh28-{LXVBo#{*Lv4{gSBSexpU^+bN1Q$*)h@gvmy@7Wy+1V z@1Vf`mWj%$)7*#*Mlp|PD}I=2K%!9WMkDc+NyYJZoi;Iu_a_Jcw&?r$N!s^4KFsy* zgx))|28*S3_^@VKqbQ|+QE~A)o1r|I>O?=9hiSR5=?W!{qAz=vIR{v(&1TpFW=JXA zk?v6MrA85>>qFaA?wJ=KrgN0j+On)j38-z7A@5e1>3ikw<$TmuEY-6+0>bcGy5EPdH5Dpx@%tOlG;l=0 zI}FQ&e4&c!!GiNl3QZ4)W6$)wKVZA`_o#u`&Hj=c$1@*0c8q5;bF`82Dgn7_#ZP^` zHY3fF6pYD#0db$0`GGNY} zs1Xxs*TZnY9}TT(2_r_ZaSF;a3HIntWHoua{zv-UoOAX+YAZwrTT!Kix3BBSVaALo ze>=iU-Mno;rTB<6iY4fw?mWNoZ0SYIC%<{UmQ|7AD7_EvK|)?X~ESU7LoVE_XQ&ChA7}V z&1gnWLV2VVeZO28Q2Z8E)!alZvsKWac&VHX96^Fx_H0?-o2xr16qxtgQ@K`54DY?- z|Cy)hhyRd(j*`Yc2@cI&Wr<6+uOcB>hO7rdQ676zWFwOyUf8&xyB#u655eqjQg^GV zy~kdUVjvl?<%xq61j?25(~(hpeSw!gA=iU8+vF4718d2w^+|)FBjzH~l_T%xlmVm2 z&9-%j>H6H8)IccYD=p<+*%b3hs@B63bM>@+zpK-i(8D_ijg5)KI(iXq*{G zJte`5UEriY+G(vqvSvF9o?}pqSS+2Zsq`sM2Q{O@TW`V@v)x+y_+!0Z%!u^{jSxKd zH@mi8M7$Ob|HQl7@KgB9+id)7v8WZk=5KFK%E_APGyTQV1y`e4*{pN!O>|$j&&;}# z{xRtGhYh!PuvI?d@ohvBJ>5@vDCAyze)Y1Wj}0~TBhn1#_BnNA;#4B5#i`ShpBi)f3(sU@GR+xE~fPp^$e&DR!gTn z8#XdwAJ)m)9~oENy^Kjs&oA+B!QN-nVZx!1G}^0Rg>HZxK9(D+6*tCq+qn)xrnvY| zF*v3n`E9DY_`P~0hD39iE~-@+)-LkDip`I2fF{J za3pWq?>gSj8usA0GPYGfM%%J7U;XOa_V{K@E8N_+|Mmc5yADj|0P+Sk6byC0GCT31iAjh`H(1$=?Qs)OQ~3~d@h#H$w)hI>^kjM|9;?p zRMNfted_y{rf5sMB-*+qHl6SGmFWr`jCQF$L-z&MM%wdd^9@v})3AV8VZ-7e(nC4z zY+k`~oq%^)q+OzJyFDaM68<;W$w}e^1ufGG_H96Idxsk_`gGJofJDC#&q@EcUY(n$ zG<(%|bEa`$cv{L`=ESG$Cn(>}`hkolC_#0xaJIM6Qw9LdCq1bjU9Gin8%=b1K zJw^nal^!yuZQ);_bpQ4mdj^09$5s%7OjVUmA~elNYbq&@Bc7wt+EM(8!?5u zZ+7TCYnICzv%lf}Z|SHvmn~g}wP)5~m(ePem5c@FZ*Ka=BXm{-NEbQP&qVBWLhL42 zK`U@1U`o)+0P34k9tI6#FmZ*&4vdNRhvt4$XJGs!u>0(=HC?(RHHm5%p4r#dHInWo zFkevWqck7!9CohIpL`H7p0U_qOq(9-QgH1COUm!kz#0znDCAeAK7{ z<=LZE!8`n3_1ul_`gV=TPiSNB(e`j&_WZS_+=4&=lw_piPk|(Aul{Dh@ZRZ>`#7D3 ze=&Rd9Oh{!li8^q<_Q;25EXn{79k1@-qv1y+T>7<)g)L~p1|%o-#ydBz>Wc0i=ZGu zdErvFwXYe+p%?#+NU8p{|90Tj(1w*zRhs7VD!RbCqZ$_ok2}WbbzQb);Frq~Zmi}A zwg3^d+fOu#qY3sqk(5T2=H4to@<+019_YiX5Zkrb-{$urZvlwDTc?-qe%J0(cz{7e z;@y8^CaZ1MhPU%@V5y{eIol&qBLZ+yLzm5bO0M1yMh*qR+|kHyF&=|jh@nUbSH9nD z`#_NXPU(%&0h6=!j2v>i#zpn|hs>ZfWlx^+3BQjupGLRWyv(kAmHLNrC`bq!IThK! zUN$nTm`{{Q7`0Hr^xs5|G8+GjooKr$Pby646T=m^e@R6TsO1zl3!4+5tV=%@iQZ5N zc;?J?xjC%qk?kTuu!rUAmfBhg_)KeXJfP2PH5QTSIlOlT=!qP8M0gTfJAr4@qz!fu zbCI$n&g8#zdj4T7DiPX@yPrQlO;^$r+I69&+s08MKHDLE<$GX4?yrdKGt=Li=6&^@ zS$t1nDv0KSmgsif_Zq%X{`*6q^b^D1asD*zrf0q61t0zRV>;q*(;KpO;ZpyEkGcQ! zx@vHuHHO&#%O4?93|dCiySB)>EN57&iM`pd+7X1TYO86&3!RL|jz2VH4cZ9)w3Etp zBS}r}m0hA!sYn-4%=CX`kUq-SlFxWT-H2TFQmxAhJ(kj2`n=t{9l`zVpx1Q!j>Z`N zzRt5tPxc*CbW|MKXDC$-pq2Os$rVFD)@d4Gs<$Tg)Lhm@&aSdGVyy}{lsvdbRGBqd zy5hD~WIr|M=R6QH7-Gg5kDAfIW&VAsO|#%+GIH<6IUlECIa>w6H~Pl#?pUQ7Dawo0-CHRi+y*cLGuf5=Gmxn@qHfQYrI{8 z%O2n(hPxhYO7?Hz%9mias>({i{yf7*HGw&*I(SiWh2*kvD0RrV>Y+|+xVjd!Rxu*- z1Wvv5$#?!ts*gob6tGxo7>Ip(bq&OdC-LWRgZ*Q_Fc4u&~~VjsMvP{UsJrM zK*)cAz*fv^hs{vpcnDAy|*DpAc16q3gmGwnSmiNJ1u_WoK}+naO+)M67jBEL=R|6~rBeR_b_- z&G&VNueqbHwE{hDM}j6irCd(xYUd#?S_`e{<_Z<3DQ55xtsE0ZI!*S~ji zEAT$mstAq6ad$e%xAFUV(Z1uFD4XFgZfq@6@n^ zI?2cU;IAY$GBtaF8rptVk+U27tMxiHDy~lLgdaE+F}gj1e9WA)1Xp+tm0>=y320b^ z_hd{D-FKM_yMM^4uDnt72b&7?*Q;l!Q020XqMJi-Cu8xO#zDW@>>7bWu&$132b6&| zzh#|24&qZHd*7GmjLcTb%t0CT2)LcNNEBVb&Z%29C>h0i)pORQq`;1GnB)`;#go2S8odK##$f`>P8ZB1dWpf>Uekm&NOVoAG?C$2SJ8i}(p z4_r^P1M*HU6A7P3{fXmYa<9U&QSYg*!1BvWvrn!Vh^LNbN8aR@1_T0Kvpm-wT~AEG zZSvtCg{d%mV;NUo{yxh%tXuA`vHkbtA6JM_;ciX)b2FN{hNmY}lYaU!SErkyLsg{bBHxO|Qug9TBqM@qlCY2^oS_ZF+}!U0lT zs(Zv4Cd2N#WGz@M$zi&Sg9(Y(!0e2#tax@rrwOQ00gZNgCWZP@7ELP@tD(@482=~3 zKqJ@uUZR&3KK0kqX2WMyeUp^$D|VJ*?3M4~?bOUdmw;_wF3v9IMxf&d6Ps-2c{9sM zFw3^waD`)G#aW)!dE+h~IS{+XH8&Q?X0qT&{Wxdt-=%ZbK4oKp4x zj-`EI915$SHQ^%S(2=(R6png=4GpO%r8%SBwq;qbQJnr0)Vwj7%s7?fWL9!kp%S}2 zJ?Qqo2mJUA=2*h2#u6MirkDJj) z6N4XIJ*!9_W%TIWdXxfo-rlqm$|ETwH3qL6yXz8$_>SWF!xMnrsI%fM$8=ZIrILvy zR@-Z7v!5IvIK#01_-M(c<%zz~X!My+NDC}3R)5aVd4gEBzBxb}s;tbBR0yf6zn!)& znvlFGftTspm*^KSg|gpdJ>1<((u72v;gvynMf(+W=6THyPP1Vmt2f3KuLEE*CaU;l zI4rrD$z-DD^_gaukR~k>K!^0gYE@8WqLO|vDKEy0MHf>WK1n!TcSJQL%H8WSZKo$g z4Riaz7*Omm{I(BlcoP0RNPB&76;mn_!!5C)>NV|;1|}@Sg22Kh-oMIDdKqwkL&#PM zoRT>`b9jF`8h72hH_x|x$!$Rc=&zK$|NhMDC@}Ex1d`yjwR>z>!-hC;O{bkEy{6(q z0~JI^1%@zjL_XcP{O6SKr3uH}Kh$g-CdS`hgUARiNh%1AksB$N(Y8Yi0yI6qty`1F zYe^QS^qI2rx00sVr(e)rAz$*bi}=(vKDV7VxWy$jMuz znR!hGD>E4sXeE5@bQ-aICZ|)shMlE1Eu&6s1q=!FaFPPnaoIHRYWL>SJLeZqO97M@ zY>Fgkk3_Op=DDWj(7H-sNJIiZkLL1e;v`F`#1aR+BJjDsofmI+8!?^H>~JH zvw@k_&XOu=WyLfAV>YWU#;O6dnB@=0?qyh``d)f$vL%Fr)y7tw2E93)#WI-bve@j) zdGH+ZHpnEO&vXnu!kE|I4+*Yk?mH#!kucjOxmfqlZISU>=N}BWB7^o_&SS*BVmkSd zZuvQ%F!HOw_=G0a*0V?|UUMqJDAxUh(5)(qzyhi!18Q#r-4mc~DzUXa^6-u2aT*Kz znFk5hC}Et1Sgh=yiA84t|JFUfL^4q$XS5m;@KC{)KCJf_HLNhx{-43s0^7N8rM~+j z{AQU~Qpj=iJKz4c#$fH_;PznVH+-xyaaE@O0K_ zn_pEc1|5agOGe3F`;3DsdEH;cB&-3-S+)m{73cf*}ybz6BA|TnT?-O$Rf;xx-paRoKO)1+;rrtQ4woFw0EspH^bFl zZG)nT&jK1sruGf48}$AhcP%q)2xS#ziS~jK!~A_c;C{_YJ0%|N6w9wQVx{5ZM>Ewl zdOL1aN8S0ou0GmCp=^~*c6cX4TJQbvhi*)frFyTK|s8y z5(!wxM>nwClJC16iO*4G+`o9XfJd)cD%l%$S~(18Q^UwV6OH>aodUO>CHKEkkf{mV zhF8i3H8ggudo}Rfhnq0Pf~!bOWTfvtr!Y-*$s|$AU*-A@Qk&HP0XXsDv{_KIx%hLi zCyCHiFb)MHgce2bfH5lQ&vRh~TUZ^8I0NLISA>>=^fmR!C)2(snEVg-#}5*dprX_J zpiUcEV-jh}HqE>J)_d#Ae`GD?786#XK1n!K(evHTBTolYxg8;P{*a$vI^mnCap`Jb z+ZS12IE$%i$7kdW?v9P_(grba0Sjh#RUK^=-qX`Fz;=aNoRr>FiBP$D1VZ z8!=ibHbwH8I8*K99q-GJ3|K=-Sk8I2@!BbJI$b>77<)kJU}lX!#hb{E2C_z&FW~Ca zcm#^KV1e5{GGcK{vcEMYAHKCcv)7T}Dr)Bl2 zc}9j=I>}Z#VbdVrB2y2x@ruM{P>RIC3$ITg`CdGIGe-iZ!k!5czPuo-rh{e7*VP%D z8RQCq;T~`y&>)dEEUjea5^y6YhXp$T5n&7J_bQXrTqHMwEvJ#B&`(yM_LHq6)HKHz zWg`}qVkHL1c{gDlN%@A)viOV+IF+E7j}fxwGH!$2>%XoIa#H=1rC?%oPLRt5-+)a~ zbp`!R!`pW}nq|pI!o7T=*}VA&$J>k76E!$=3v8cKtZYp_N+b56sb2adY18?@f>(=* z*G+y_D=^{aX}gy#*I|_CQ60SX;kNZ$xW<|-#BL2>*FlxVPCIQva|xIpCJQ90k+;DA z7vm+sN$9rR2$P_-4*~xI2ysS9e(lpBOj44)1kUl#$^JE8vvAi0H`>e-v#=3|Bt&!bEUf88JRO4ZUP~E&lEe*q>72 z6ldNN)k^&}@TE+B;?B_R;u%npVsmWla?#VfDLmq-e6RS#c5TY!tOwXL5o`|fNs%u9 zEqD1*4<%$))?+0x(d`#|)x$GlMH1!GZ9n?)^&H*m_b(}`_BuAAWu7;mRaA#B!oB** z@wQo+pQ?7d5kuX`F~q1Uo5GmZW@zthL;zmfnH&tO8-? zz;f~?t7wPnHjzh%PY=eU+J*HrhyQ>bigpNb?S$qvwrF`$Jkby;Mf8t9$m$8q+t60Lj7=0}~vU+@@N%=jB z&p%bPhOIJ|^@4r%!a?in&X~ndoASS`P^8T*lR|e)!h_3WYtB?1e>D)^f%6_4SDW27m+Em_SxXY7}cNE1q&ee?RTd}j(u9lbNoR0A@TXbB4(axI|e zqx??&M%?Fkiji7C!YZK9WAF0W99`dp2)U)_Q11RC3gcGIws)U|U%7 zouZza+A&DFwR;}y+Z~Ao?7Yaki8@y)A6qRo&!ErwUlX2e6g?g#U6zN6+wcjdZ|DPN z@B#;X|69ZcK$Wm4)rt^zXFA}wC<3`6$XZTf;eQ*}E%#k4SVHrMpx+RVq!eV?>eYev7^rIcS)_m10| z{w9im@)h4}h!2n+l_>EYx^3V}{k|@nJt&nHr`eA-LZ_0kwYdS?GMTT%%2scykI|E| zOy~P*MD`Lkc)k;E2=vR2T*v>e%~Ft!?_lc*)DA>uSm;oDE9-KHL%Pv+3<>_Q+-QSG%XP2k0`z{OUVo0HFDDH^L*j2Pi+7*{`SEq@l&ETGkG1;zO7uX zx>n12i2Ma6d-jyhrqKt91MU3G;8rh#nHA#;ycUt|dI?Y0P#C7t^g<%H9Cb39 z|H=GJi`#tQUXyIdEx+l`65@z4e=; z&;rg(uH}?3l;%F?v!|A~FFsHod*J+Squ)VWq0ZP_$W?;LingJ^nP3!mvM05(P(aP) z6>owkaQ}UPT@G`n#Fq}cHq3!Nxxwfn!xD}1UU@q9u>AgCinFK&v9X50g0v5B!DLJ^ zn11?mYiX5(YW2wiuJvlud~JUI+`Xw!U+3cYg3npy4C*C+E7^$dwzTcF9Ue80WgRWD zjHS!(p;v7UNh>fl{cEMSWjuREwCG&??H02mBpyI)HePV-6#mYhJLoDfj17WbsQVqH zKJAbHsD6Kwb-$MtGMo)q!laLv(CkK41BU~lM|yQlmqziKAY&w;WG4~8Y73Nqcn+j? zLN?S_Lu+rydBR20pCxy7f@#i^;VB~!b31I*Bs%x!?265w+z-;AeW&%=Fl;9#tAqWe zYB7TXU-98REzQtt?yuhW)VLmz-m6O>)dZMK5B`G9nQ)}1rI*(!MbXQJ-S{r@`edy zT1c1x-8wzoJ5{)OIIriZ$*ewNQ(3G3rbIl={S|v{O%er(ejUS30lYnWCbSE>8&#Vvs{wxUlmq5xbCRN-C>dCKmY6=&4Wt-Zo6tmr)2Vb+L2}-$N8OY&9 zS#fp0T{h>}zh{kvRUOvaUgG2^(MHo)iMGL~CMhz~w=$0%n=>vi&+?=Ggw}T}0*>Z1 zH^$+O)SQ%ClcOBre9)cq3|sq*_dl_w%881N&^`dItTpbwGg6;lIovzbKeK!B)~w+D zp~KfD7D5k;gQlY$bGxR~Za5II>}^RNBA9jMYHucMz3C#g_n_m1aGm|Eq4_EY2o4=j z-7E_8?)v#Sok~=@k2da7ojzMw6SP-opB-%z}N)QE>AAMRZBsxzh@W6rn4S6L5;tf{rVndjV^ zB7h0Qn=zz_Q=@}ZCNq%7J&>-3Mv*dZXM9)$Hq%e~?yln>UeMnrgsK2lyz;YL`Bov# z`*ls10yZ22<016vm;R9}kl=1*PYbQ-dP&JrWPic(!n5(6Pfu2I=`S%iuYo{zbM2^{ zY;VM=;D4UN%ei*i*h>*_h-*ki`)`VacGSv-GO?0?Hq; z3Z$I%fZjJ{oq|{ub!X^o{FVJhTLxP^M5l8$V8wjPHKVkKSlxkH3 zMt7Wztgy6Ml3VTT0j2Ap+sbf$9L+VqUlh;^Ie#}eE>hpAgJ7!o;X$Pb4g<=njvtWF z2l%JsU@`-3pG^kP9+o~us8a9Ga<;Zn3*{#(-YxtIB*_^_CL1s7A44*Th(bwE4$^1m zZ*CaAA5?|wuFxhbUJQi34j=m8=)k58>iC|My2E$QfyQm1{1tJ(2!O+3QN^{5sc2M^ zw9;Hpu;`xR7?q?a!+*>##Ua!~ZT9^ktQJ%(&I{Vi=4*FA@U_}P+mZ}Qac+=nH5f(a zAIwM?LCc1%I(SX@c(Ee20lry^+%)q38}uN)E9O3qu_Utqux{7zOkuD&hf0qyOHBGa zpaRY3x+E=u#s;FTxBCY;%8pw0hvC41&2gm0X0Q!;Blf*4E87t$uIb~eQ5loKa8yLQ z*6?ovBVB0R7J1nxO87kXxC#iCJHBfzfDIm{>p#dkdr@0zs5FXczsihDiLb(-O_U$i zCvc@mF`6g>m>sQ&E3KuCuCW6UmO*{zxh9X%XJ_7B_wgbF^|26c%rsP$R82K&+x&)m zQFwqB{J5of!%twvoD-e9X#JDb8gCIuVWhm=et}x%T7aAnu}Ngr0_>S>x0Cmh5v2tx zw&m1wV$9D{6iB{rO<*FNB~##0_)pQ!P^uV}f(6yBC#x4$MFfp$SnQO75a+Y@+CTMz zEy<2Wl2KXal+`7~jn1Q__%6+?DJ3e{y-Ca|>?(ZMv{Q2WYXrIYdmj*An>+0C-A%lC zkYIq1kZ=d=|K8j*M4k5fkJgXA?25!R48`p3)4qN@RrcDTT&a5;Jje*%;lImgiM6Km zDG?;;3_`0I7zHMN3#&#UeM;SEK5t^>rj6(l;~pbrsy?UThE-lACqUjsJr>zcD0=Yu zu|cmH3lGolrY+g>RlaEI)Lv-iO_??E7kain_6@>d9Zf4(>suc;+6Qv2fB$eF&Pgi4 z;f7ZcsN!(W8~qv$34D|HVG;7Z1O6+2F9qi1DI&y@@r|WZvd>2g2W8a-0SyXZXsvlTG6U%iRcTJ&rTar~D|%l~I&0yo_*D3&ygy6i8@ z3P&39K+4@9;oziy3vw(x2S3`o1zq&QQOr?0X;05K(pS=TQFA1QF+>!(v&Lp`+}NGP z-kxN_!>)&%{zHCWma?vD7cV5XQNpQezD+cndY7~~t;01=o8(>1sw<{3ZgW&~feAuG z$wHHRepR&_xv`uR*lY%%w$>`{ej!!+`@LW33607UTJmNSG}CiShOa`1AloSxnt+4D z0(?2UiAbj{1g(0nT-DT?ffU6B0zC!ptmX|aY}|7Oj%$f0bz`SxPjPBZRuj-{@UMJM z>#jAt^8^LEDKgKeYc&l~?bqiwovn1Bt?wBY2Wr?6U(MMj7MY~!eDStHl%OhNEuera;OiR?LHSqM7sL?Or(X0A6CtB9XtKP9_$>awCYAd%1B6`8Z zC`a6uVMi&r<+|x04;5aVCN}YZVqtBjRJp;qbXHBDBiz{ULPlJ-sA2AZA{n~zL#X8Nw;fNZRRSud>yAStcp4EkMh|ub@Yp48tI%_qf#J2tZ%~4Gb!^D#>C)G+- zobj)0h`qwoN41D6Fj7-3CvWpOtfZ@RYjSBCpXDEof8s2qxvuLeC3UIvN#m{p0T5On z1xVi~1GVNI@#DQY4dlT(2tm60#DyG(sXM~+2At^#)DKypLvE9gX`^||rB?x5S=o)U z=mrD40xXlZSmx`7QSPbAWfpRthJy&L8yBqAVpk!mUTPoXSSM2?5J>*s6carwwxNl5 zcVoWqiGG3A{Jubbm%zo9c8{uJO%ksq-uEhY3a(|@&O1tB0^vc^AfS)rxas+% zRuPy=@LiT{vWDGPAub$3BhhkFBv>VEYjbH$7RAwcU{VR|dMw?%e~Esd+ya)Y-lE#i zN~@s$YEPl)&tLFmmHBDOz}Y7rz>n$<$MGNzaO`E`Fo9+`?*BU8s7&{? z-)%_VgIxg{|E3m87NFhKX1eyK-pk^TC3eB2Xl4PTcrj2OE4BxdQHB0x$d#na3Y@_1@6Wdnifa&0jQDFpIWmn?A-S1H*6&lq}fCE zIplwXluau`sLO>h7PyC=7_tt5F3X4SvWWT?S*AK zYXp%aO1`gH9T0U<6Ic_q--8HN&zNKIJk}B(i8=%A2n5q)bh*a;djJc&?|ndSl*rrT zwhT14M2-?1oJUm>NNOQ9fYvi_4WQZa!b{J%sz69mT_tBpDpZKcSj%LIu_Ww@oCF=>p|VGi2Vw;Uu!~tqvH@~`K5;xI8O{+)m{vw$f1{G!~4y$qE_6jd=KQkNH>56>E0cAApM?XnXoLaIX?6+Xz?!Zb!%^*QcIG5ESy zv5;@rh2gD51fcTfP;i+2O*dtcF2@lhF5b7OW0hFp|5*Skit{P7|Hy3jxh}r#jsKZ? zjBO4!0(W|dtT00S05+@`j*A&2#HPs~+oa?i|W9&u?_dya>b$MQTWL8|^x}`%&nTXrk&|kCd<}F^n&)9ByIygg7!pJ~7T*}2;i7D*s z7HH^3nBcBhRZiBdHkv2-=W}xWXr?J(DAl8hIox%M?KmBf2WB~z3~tSDOop@=d*dLO z`bUu?^mLLZ|5e}#?j3q5`I^pETv){#ZT(&SXMmKi*z!2O0)d*WcLX_ zRJq4Xi2QnTV7yk+EMw>GhXchU^ojwZ59<17eZ@(k4Fm#IetwJkB<(>BlG7oIfs&aI zZe7bywLQ8&w&LsGkjnXCS<3xZP@E*^0&X_d0UL_OqP`Rw5o6!#I}qKQb=fSJHUf@l zOm-OzqP|`ixG-k=db7R%&0tGjg+#01xVam&@eOk+COE$8<^A8q~0ZG{Dh` zon(MAG%LqLh`#~f53)}t$TtVf*pKGHGohUse?L+SiZX`nbTHpAftVBQC4LyC|K+{V zrTzu7ZC7BQ@c#o^c?aIXh7OhHgDzz!hRIhjP4iG6Wc!P4#Z3-`d~*!w5~;{X&fvWz zwpBzwX>7U9K4>yZE#t!_wnTz4Tr62sm?`!}P3Bn*`pfsPh{|x=u4v6pdnQjY^PdYGNV-kKvU8$;RyT@CF9z|{WvUh=q ziaTym;R!O#piuXUQt@rM01i1>6ovJrs!?|fcmS!;-n1-U!*B|b{>|Fy0?6xn^U(baM_ zkybONR}h*}Ry>R7cQ>4%rzg1ZiVG9cgeH(`H`{%}sb(7&ij>$KU%KoUE%q#M(tQfz zqLvsOC3N=mP*=fhv8~rzYwL7Gqs$Q>+g8$JRuA1RXqm0Qs=Fi`5w*RKchgm-7h-}P z_p{7j30zV%Xr9*R^R;+!flnMt3*d!H4^0`7_L z1XCn&9|eTix%#-j;UuK>*tR=Ty%%%~?L5K&W(gFZJlHQ*f3u76$~(r#2@*uHU`xFc zv|CA{hLrca-Mgke;w`tn(8SMRGxr|M6MjPz2xx6;Jt7tFl~Jaf$<*XoI1Y3@>c?sd zit3Ve33t6uj;q4jwE#L5+h-vl{NhVBdaqRE(^In%oy&P-u|E^--670&yW(JldQK#x zmERdHrUZc@U!)1fQF*|2RAsx)huktaV{>`4zw(|EY7c2e(R?Ah^Q7PdMtLNOAQeUqhck{D1I@5Lj)z7M2|Rw2uW%(f2)vtcFI!JnpiuJ$ zlZtgqEf{k`=(($vn(GhW7amvuhHtE>YBZ|S(2K*wn<|Ry9*Bb1R*-DfDJ$kSU}oh(O8VmM#Y^CC#_s@6W#@1xdSzu zMFI;pmEZb)y*YnV;Ci?f4EaVKfJy)3o*Ns!(RNR?E6~xZ=pM+2lHTrUJ^Xu|t+!rR zWiwedxaAwGJnwzL=u5>^;k;+jOn~A`mIWQw9*SKB{Mo|mdTihZY{lYl4MXmPg2Zx& zHesr9&&IG9jvey zx2k=LucyxH4J_ghMHpBk3gQn(SGq`EvlK)tA;2>-z$YFWm07Ypxdb{!sK55e`WPp8 zPF_<_p$Y_lBdi$NMudOF5jW9Yx-I}6X@)^J^TV}{Ma;V+ywQ`uT z-bmPGptWO&bpBnq5LwShJs{9)F)2VUn*s1D!F|vV7e>NdR@NqU6R_hr6j!{&x;NG> zJ`GunCQBx~ukN#?yU2b;>TSxQRanFs`KN%HQwrnr4Z|Zt>3fft);Y40qFj=y@zcSW zhnxP&cyK;)SV_hgsHDbaS0CfGE2{p>Fx3;B^wA=%4I25(NM=>&ZmrKdBtPZMbeVIB zBF_pr=t2dbjH$lW`cr*;L!31kEQUV-qu>{5@&nT4Hx_BLE&m2OS!1^{1{Ws*++))Il z8IFm$&_1R}g1K|1FU{jC4gowPDT0&(xWSkym`LW+yo%s^3i8xRiA3M(ADH@}baxgq zzeW1?!Hjqy&U+@FZw!gEI%0qK4iD!U&h|h+PVM>D?7H>wKMyr;i9J->mwKiDT7ME` zzEmKDfk>K0gRHrCMPxV1YU#v0qfBgW-5h>RI6!-IDB9>>5;hcf_l3)WdB}I zh5#0ke*srcQq)_w-#2>>MS~Bw4mMQ<_8OUTI$x<9f?GfNUI1Eb!Pq$@{kS*U?j1IZ zfCsmH{RPY+Tc657>5B>Z;jd4+g()X>*3pM?@Mua@dnt|k3@-^8ksijz$BJ|w`kf|1 zDIOvYt~n|Pc11VNo=SIR9{S9(+C8h~{>Heo^ip&t+&Kq+ZA6(bk);FbK z9f$GWRn4FOrmruuJ~W_yUf9l&?p1%$FqxKnxGSq0tuK=8o2R9^hI)+S3-up{GnEgcG9?ucdEM+-P4^^l<(!V5L^i zfs5+L8Ra19fJW+i+xVYP|I-Fm?*DN5SUtkKIexr6H6af=A7mWHe0=K*V-iN43_l#Jk9!M9d)|W@S;Q3`H5OCsTx#_<_qoOqWwKZvM?}PU-jmj3o z_8EROW`C!sM?{4DU{=knNjMx(!KxM7->srKVdON44U=w-XgjZF978Nz0o2YHfGL9d z@5Hs_O+DB8W1Q;?V2FCq1%fv!UmZ^P=kfm*EfDsfG$X;>TUfFk0S$^`Zc0^Cn!1-y zSvfN6)fXfsZ>BpB3uC@etC#l{Y7;uag@Q5g7auphoB&z+>UrZ2*7sToMBzTJ zNugYiwZ@OJBJ+%9Ai~Ad0zA#m%VL1~v+{<${3^DZwqBv^dl|*$jo24zc6$V1#Y!ABaJ130z zl%Cw&yv0HQz-OcIg1(o|B)F%dRu}#qLcT)Cl9hE#&oU z(d|IrgqiD->r8bXS$GisC9}yH4JRQ-ya7i%X5cuR>2oJ)Hz`IPP@h*{+o$(~W zdeF@lr}`W{WqrVTUI0Xf_(n23jKT=ahg|?ayJ>_$(yFyHkBqr$>tv%L3a&Lpl6msI zsWN@ZhRIaxkwj>8>VQW1Vufj}MRdnpP%GfdL5i|0{bJw-Jf7)&lp;Aqj)F;~RA3A_ zqNT73ZawAq^V{YuzGW2Ol!J}mR!_m>>v^#T@Lm%31^|cegwwk$j5cN3(!pRlDcS! zSRwC{QD)g$*+fP-_UPDK$DYS_9LGN6clms8zyIAh=W@NS>p32e`y(O8qKAQ>K1*{o zG&${l9dLbG>bW^BC|}0}Vq0l3K9z1M`?mc)*A8}YAIxOzwa;T<+EAehjg`v76MI6% z#~R=uQ4*y*FSR6y$n%-U3upMi{~@GrW`Ewd0*shS|6!^Q5?h&n(lH0IGjQHv9RI{7 z`paur4fgA=&?yi5)tO#;7m!%6(0eK$&7N0b9sUcnj8m=cOW~c(AL+Ht@UdEiXEx(z zuReTEd6Cr&zNhZ^jp9c;=w!CD9e!D>m(rp0PHbaT(A<6hMFyJ}R$l!N6a~Wi9z&Vk zfYmncfs8(JrKJGNA~Q@(FX*L_W?tW+&@Zh0Rhs-|K?Qd^ zjfKPrx|IE^yx*4JA6;DbL*5eLs??om3sKwpyjT}KgvtWh-GT*YR&mKkY}?cs{7x_1 zcFRO@cYIS0ZoWNLbxV(r5AHh=$ezE0D5Y1{b7uk8HN$@H``tOb96hwtMI|i*=|eE$ z9o#KS^vi9p&|xnM?vnY3);;v!RIs5eB^Ma=xps&56;rnFMZlJ8sUhnx{#PP&b^^7;V^dDoBFvX?LNujEe3rorEwFOyGkwGJm9RZ3$gV? z!K;UR65U2i%b#oc@p;mETiT-1UM&3Nq@}?rXdQWo#T_f<$qAEGi%z>UnCYkNaHEj>iFXb##{CbbU@;Ah=`079Wrpi?!!9L1d zCzv-nIb)QEg(sS5j?eyeoR4>hZzuhe8G@h~zI8r$3&3y#)Ld5z0d9l8Fg3V$#(NLhK|-+Hv`tS~4u&FL8zET{_xIL2m|+1IE9i)mb<8Lq z`LzH{V^a!7dOL7tN)OqqbH^$`MN3Gyy`j zNa4#l_kLV_r-DKC)pA-Q*i%4cEA*-%NLrRzl0!&fZO@2tM{BjMbS=bXNgSxY)V*Jd zQoN;*qnh6(FrBO6m}Jl|KBRl>c5!DNKKLT{TD>DNzQ99`yEMI<-=*OTZHv@VtG%QH z5K9t3x@DwVZfbkN^Hgj7x>62gt@Za7zm)V3CePckmJFY1X)Rf^z1rUc@6!v-_ak$g z7N?C@OAA6QS)41*j^9pv$^)$S0M`o_@^5Py^T;C(tLIjT{mw026GgWeo|6mQe|Hiw z^-;Z&9Ywt42UJFtREP#R7Toe9^EvQW)f~S>o)xBz=~qBwd=(^Dya93V^9vurs^89# z#r0OMk>h4EF(ZSenIk@!*1s%BKvan5rO@?tPF8;*Zq!FLG29RFmfScu-dp^yI<+zI zh3khNIK_W+e5y;=Vt@r0;fH(LtDSM2mWY2t0G<_ZNO z7H~epL*!DF9A16<6#%3JNWpf)RVLUExC(HBNnxyMo^AQJS+_xONqV9N=nm?Ou_Zz0 z@8P6P=8oin1yPtqnxQ(33iz0N)9R12e>`JO^B&P*Ktkj+r=$B@`0Yh2uAx0$m!w`~ zWgJnY#y`INVUThM?59D~!|=~;G%+tl4t;*H5mKb+ZTx;CP|9mhW)Lis>;(;t@S@;I zWs|Le1C!1Wy7h*YEwo07Ii0bU5^Y=3l8|#_BI+4mh&`iC=>9?ouKU?7dv*dUfiF&2 zBv^{(i)YXwEA1mikqpf_k1wLA%utW4%tpp35s1Tz{&MB7Uu6reo@y2!RUYy6&}5uH z4Qmh)M!{lXm#eb@&VaAalwIWNFPx*NViqq=&NOjb(q~iBKGMExU_g1#QU$mf1|WuN z^?&p;Iuw)iLA2WQG7cYN+UsPpNDXK1{pPhI16) z88YdvXC)Wi=jqc0ne#(_7g1b4jZX&+} z=OYK0az)qvzri41AD}!w4!V`X3*HS`n?wg6Rb~EeII!2U8hf-FIy0v4O}+bKuIHpDk1d2z?j; zKBOk8zNH7~#@;Jt@hm`&*lY@;SC@K19&*7tllGOA&z}FU^7incNDWmP>au68xMoXH zVkI0Km`(DC4Y2$TwBColJ?AE2ZYdY-gN5a19#6O5`!iZ*Qhpk~Xs^ww;do-P{A ztjCS!;N+oQQBKKCn@RYAKv?mf{so4rUEa0=(?9)Vm*Jdu|EQSPeyq!ChyCJ|Z@whKLj-&gILxIE) zzsxaCDQBZ~sg40Wj%f>c6>jrLjz<&%H{fyxprK_KdZ`ry_(BvsC0k+*q3UC%dbwJm z>uOWlPvht?-kf$@$6s%_Q13l1-4o>B>hWT3fnV+lqe4A>^To?*qB5rO7fm-9cZ@*k zVvPrwINDvpWTDr;*)GJzePmUW#AWw+wS@TBC4dHs+bvF+_I}5%2ws(pWo^5_%&|cP zT92z||E}k@UseM*uXHCWa;16M5fyW9Z#Gq*UZ{p)w!y|HS+ zXUSqMXL;>3OZZHu(qYO9-4U5QMhVPwjP?RT5FBxj_l|i93`!{+1ccUr(YQ-qo zXwZy_Wd655Oo(HJNbY|B&*zKVbW&{tjc}$%-N-{~8)IrtXoh->VF(vN^~-;dZT0jI zSM(&4q+2@qK6#)-3Ra{yA6MIVtA+Y{v`A$wYUZzqC!_14x@&sIeLr*K(qco!B*Dx* zUb_pq8sDC-wO{vs@zha8zNWVS5kw*H4&!g= zh01}(>@+tr-UFSbUGoeyTsAWC4$YAQ2KvH;<|El88R!gUaZEe?bP6U#s{O$g#z-ok z48hdXf%NJ@TR&vk3*;}QLo(5G;`l9xNkOEO=;ZU-+2@Ev2HcKCXj90%hR4Raswo$F0d4qJs@jR(M zFuY9u^5w`<(TpEsm;`Uyv-Z;WY2B$cF8<1Iomf_RyEJNd*Hxe?CaT{bK8py`;`flAo(ngKxP_=sVa`_DFNI^GunyM%-#WL{m z(Z*5ID?Vz<72C{bmZ*uiKVZK&SeB7m*5vOW4r&u^ccFO87o;4j9Z6s=@tr(^N;(-_ zda25h0EzYpnO#-&umOrf>lCI>oZX+4qB3hg2S`41KuGd&%7+{&peY7KC2um{W55C+ zh-%}qRM$Vwu|vHBO+S()zD0Pbk=_q;y`q|@-0E?=Xop{9{D{lKb5p2b26>CwGXDLmw`Z*}uE1oCD)06t}DZzdhRK&zZDk1x}-Q#w#-vF{?HD=+VcLuj1Bz%6`Uu6o1r7_C(Ud!7Y9 zAqwC7qV4LeqyM))CD*QNv zGGd87J0-mwQJmA#rhx4bR;^Sb6Ap@-Pg9!<=g#MjLEA4k-Z~6K^axs{v2fPE6lhi5 z$m8w)Ow}EZnV1G9m^Pc~gNK|9tEq|ZF78=`F|*Z9Ua`^8ji2EB{)?F=LHwBydr1i$ zDrs+{B`<Ym|&kSnY{@H!|k1RH=)UYVLFy!>FrB;9uT$8eli8&!tPM)T}P_S^%$0}$n zhtK$$nwrU!&6>UkkdT7HAlQXOSE=@X7bk@dbH@%}Lx=pHoJfr9Wr%MV7*t&dq53s| z03ra&tk2Pt%(#(NbNh}^)22;hHkauCVSgr(N54a#mLa{Q0$s$~ttrl4;-(l|d>39A z%*FzqRsc6dI+P9W12?qTDrWViF^JIi=Dhv6{G$HiP9Gl(CLR1<;yj^RL3(eDv|ewZ zD;Fx?FUMy3xKPXI4>2Rd67w)4b8uI1?TYpwnDKW)3|BjvIOUqYru&60Uu+J zT}W8k`NKR(Hx*9JT=gHM3-iZb0}u7T7m{%&ttmh;SVe6Ud8>lbhP7VMNb-ZeCdVpN z^`?ptZ;UR|R0JT)|LYW#`0z22Nm@;KxlG<>Lk;50!OP`9%j^C0Hzp(Cb<-n9xPP1G zbsRM{FoY%ErER|aiZwR;}!GA){)w$d)$MFV;A^rCXtPGow)EAQ6C`RON>Ki!9&2hLGn_$f)(5&dt! zAM^W*$7vU&G%N@&HX%nk`5eXyZEPK=0n`e>>=iua`Op@yo+q0r3_&(il(t?`eWOGDxBLQ{@leH9 znIpfxf?#%)cBM4_GW;{M2m`7F^O`5e?a&LD=VzMaUW)IhTpi%FO{;UVZG4!?n-zTS7!xsYV9w{e=at|0A(Z%~d_B}%IXV>j z!1Jlqy#B|ucEf#V7W)q-}5u3#)75y&_m4j!CZokA8}DOgdbveUZE-6tII8!zjeZ6&&2Zs zRp5N9HJZD*m%^4qeqGX@lP*_({eJ$vF?4Dw-EXcQ4TN5JRZw#kyRGP~C!b+_0H(w} z9?oC@TR-ISR;EUP=HsTz?YAL2ijFn%eHY;L>j>TjY{*^<*l*;Ff++u z`lc3)@s@g9uAVq5m)y55wpZw7qyekH{X@EEk1t($CGc>K>ynf95Cuv= z*894^7e#>`249zEOsQMvyuQl0;&J^b*RG;t`CtKbywIw_cih~#xX7f`oXtG@^7m}h zjK%Y|(>+C}ZYr|Mu^xE7(Iw>XO96!>7iH21)%6t8ooIh@eR(%7i6vR{Nut=Fi zMP)iTu=bRniAAqUeU%$A-|FrCsiOd`q}I%{q?Tefazqvm!umM4A0sQ1Dg@TTt>SG$YmS`3nUK-F0%%P;6O|qS(VScK;ir zLr(wARIvRKoALJ9Ea|9rzP*|u)V%sHEo|!V=o|T!pG{fKomAo`usaNpL30=wc6X;s z#tzE4(A##}%)MEw=RjwDise&HJEIn8nDMstP6W`2IvuayEDDR($+ zah!BP?U?>R>6#BT;#&S@onJ<8HP}gRVOavlr+D#js#upA$6zubsnLj zs-oUI`@j0&5o?9^>oIE?zwvtL>c|rq;@mo%=5GL2I5w31_$-_3#Hn=hGF%dd$cS1t z@1Nu+mj{>Zk%6p_=2u+GE7w z=w^{nx@k!qz=XA3@8$P$@ZXe)pZbMtn>K(XcI zPb{z#QluO_iA+AWAvf=42A$;Wk_mNv>7>cnhG5b#q41ZvbylqX`D4WgZ2ljXeN_?^jn>)UQHy3rorF(2!h-b(3lboj zwfhKt`;no1`qLOUy7P>fxY^)+_TZaM7>^Np$J@o4rAgSwQqIRVY%Gl<*gkn#09H?Y_m25H-Bq0=G*gE*VmgRGhKvl7JxUbe{tHM`f>GDX_wh`rZ!jucC_>dHofXGW6Pfo$Qv$ zw3u!uAp1qvX(Ce@{y{6_*!dRhOaWs4&4$pI{KpZzQg6)*{Boy_u;Bdi>gvbe@xqzy zk?qB-gMK$2>>WARhrMJjXXfOTtmu`zifmz-cba^-$nYs$c1L+BZ!pc`O#S4F=xFGI zYJDx?pLvhY?GR?m5l6x!Vf+(tDXK+q04&(rI!{US-a>&xE0V9=&@a&$JrkJ4w;S>AYgu>T|FaPTvw+k&$yMopb$wn!JB4D8qrk zBw?Tr<`51nURj`K&G_o`EIbVEhV}bE=^tsMicD9GXBs&ACK)Gpa0{=&g1@;ot(?!{SM~3Q1W}migU@K5krL?(Efg+r2<(d#hDfTB;{pz z`NeAYI9FK3`PkT>UN>CmO#CLr;t;(Gs{|;7Ri09Q5;aJ`dy;VIthRj!U! zzxxxMWuu3()4_LQVwpF*zJ;$Sn!~$(`$7ty4cr<$eJLPo`YC)mXr?P*0M6VN{F}pr zfIqm%4C9q=!Ansb@67z=aMm9|T+!ETz`HmDyU+8ns0`9RKe-Zg_Dk{@#MigBPNCec zhUL%rE_IFfdZe$KB|;V^VW*?|pcN})YunN?VwbhE+_p&jF}8fEZAT$(fAZ|!qRGkG zMWE^i_BXfajhhjJ=b3d_0)d5B6SEu(LB=%vtLP#}|4sOaB&x%HfHE^OH8ip%T#MhZ1edsFRVj{3nedhA1X>;>>~VUm_P z^!y#*@Q?KvolN9i{q?&uuGSaO-Ukx09RU`LA=Omdj8yV~aFo$0C{#^=0tb*`uG_fsNlAWh`O@|`MO$o}E(CxhVP62@{9wG%RFwFvnQOdIUT zd9tnzG4Gc5N}x31pt^`NJWj)a3o^>mXxZ`SzSit~1IJ>BTd4{#uDr^m-nt=Gi-?MM zgA7CM3Q_Z6Nv^0>GZf}qrTPq@rkS$BC{1nvX*T@N6}3R{Gr*y%LTBUmxDU_vkNS8+ zvpPmIC%3nXa15PW7AB3v2A4K0v66?PQMjVMuME89=8iiRb?sYZ;DC`X`ksxPU_A!l z0p7IhQ$Fo+5F)!GyD97GFf4Li!b(XeB}w6PRC|ZH(}DAH@?D-r{U56GO&j{cU;1MX zo(g;Swmf?LP*D1NDUICVs8BUkb#puK?hrAOjO8y{%k9kJZFf>t2YA|HWc zo>C;BAE_;8S){k8P;vE2oz_#BEOq2SAN*zrO7o2)&EahK7qVFbosLfjgOP7sR_Si|II#PKZ6&cV zI6~w1L=C3r(K)SL9KnPz@@%5aZMXxrASBcRnD-g=`1E)VOS-pE(t6J{QrfgYehFPE zHw7W?>@V^!?fVJ$L6tUg|9p8Z7NfN)ohg}m*B}zZj&{qrcJmb|-H|ZVchKSZ_UdF= zWJm4IzMbQwyF5I7MXJb1)lhsU0Tb!kHga^HR{tjmzs!H*;KOw$$p?%x08b{veD1G* zS=;1|ml;9R^j?jY@D6*86vJ({4`y&7zR@?!9kr+ZRo3w&IW60-bJB4-!S)%a5w4tB z5q9WHXYaaVHP4603t=o}H!FM@o%e@Ad|!pV9+cL=$`1a?W6aiNCxnZ)h2|bW0@;`% z6b1By)y?_ET4Vn>sJVGa6&FP>vn|!(MZJ5LAl3V6&dET!PRW%aI_e>l<~z^1see`d zSlDhJt>LmGnl%j3yev0NrDusD_cjxHUC`W%?d(CxfD&h0wDPtMqrsx^=htWVkzuY< zx4x2g;avkAoo>#}diS17fE2#_%)XyoP@Lz;R8Un@w%b)_dH3`fXuR`5Cmma1JgVE_ zrd>lT@+VX@G>l>_G;QU-giw~(-CC3X{280N@p)JjFq&wV-@YC zxFJW!^mY z=yaHmDMw%AZEc}V_t4ydB*YFDHu&wQm@cp`6xmyhO7 zAaGFbapvl%>SqMwfS!W{D9?h|oGu!i>cDMK|9WRHN$EcAAIfFN^$5^6ZZVvon zZm+#~S*L(#+Hw$=2irt{30;rO6lQ`DlMddQzUTbmT{3nh_dfCy_+pUGZId$_l&O8~*t z#olcbL_aL#^d+WyvWKdz)RsnuYGmUEeSz1y!Uov)-h~ryOg8H50ALb(m9BGoZRV3; zR+HDo-LvLtBp-RO-y236s(6Su$+SKm4@)%}N^tC5D9-$|kD=c=jyUX2(4AVjicfAHSf z@0|w@8>8BYGvyfuj7ks*LzDEsj5lSgej zwS8W-ea7aNxpRO)+D(`zD3-kyau!}LoQ?soN|m>FjolQ>8+?m)%Gi+MQpA?Nu^%7s!kt~Xtkvw6?8&|Vppj1CFc_B5zSpeQjgZ^<<;LUsB(c2)+!5#|Y1)MJH2fx0VHuz3~<21sjf zmCzf^>=M<9c6jhd3wD0jh4=ggt{+oPRRZ~$u(I|~*K|Y-P?2(!fHP0;l4BEbhaS|J zSt)uDU%h4aTx|Zmn?g*i99PsK|AN`tV>rMt02^4uPyT9?j_nS^`sfww+Z!xTD2`0s z6}~s512zEj&j29`yQC!v(hq4R^7Oz`@CZFsyjAy~6`xWUtF59QE@I6gUWYvT<-t4h zsjINiD|}z@=csuN+?!=~@tFyA1*+tC434;v57{{#^49O#L$}|$`^)I?sohOPj{WkxdwL;MUVfdeJ^X~$%#6%O4 z7ut*AZ%%Zo_mup5FSw}oIm98hi*+V!D2wDA^!nM_nwJbiTiBn@Q+IcnDh)>do01i} zl0!#cb_9XJ2Nn@iHFaqLpTVs;4prkzSPgCONL#(`U$@rzyrbF zjf+ffFD8k7h-_ZE!1$Owz?>#^vH@_0HV2o4eYS#i5owN#8*8`y>U*s=(>e~Sm*p~# znu3t=XM5b(i6Hr%ZuzqVE@#a!xwPYhw}z|J;-^1toP^WPc9W35jlesK3^UHI76@{m z%-}YV3&1gE#r}cl#~O}B<$7qRbR9Fc57V{P5jcVilxGVY#!Fp|-Qag}J(SZQUCNG@ z)yah7A!H1B`ywV3b#)iujziflJ2$F+J=?%!e2(d}aZM;YatAMd=%_>=wZj0&&at+swopTf0G?Lt%G=LVR551bYQ{TCLr zR4gv~?Ed>{z8e8fIVIVV(gm6M{A{LPxUOwt+sb|9YS)akP6C!wWiH<&MWKio+oAH4B7-9xu-3EIiC*??*JA z{+xd?q>exM(0JX-{PU{1+=Br&Cv6-`u3Q2;cMx?p1d92mO6j8O*&Yx7V=xS?rHk#K!Qxjfsj=V25(l5v#d-^w&1*%%a5rWzd^U#C7TQW zYrm^PS-Q`8)%7Mhp9$4XiNqa57q7=v>ZMk_?)lSGS3+7mn2XX48U+QQcp${~V21;U z%e}?V^=&+jVPFQ{PGY7TFg%JLmf{5ZgJz z3nVhK_7OKrfTkjEe?#u2yx^`DUFbC)b=h!PZ?Zb)Q`8=Bx|#XFDmA-6TD1op@);E{ zI+Ux+Yp5EuWH?dNs08VH;3wKLYa+zB67l1w}upmelKBVNVaXc}|d_N7t{nzn(kZyDMaFXwFhI*0W&JlRbwDe^Rf z;7AYqW}zy;GuMwvaFb7yoTh^m8g0RKzv+?2^PGGDp53jPfpt#!QEHKhL!(n z+D&7}u0RStk3aC9lv=#!;X8I|hn?)jymsuVB)h#R*&5-!sR=9oCAq zHQUzrc#Atbc7huv*pmw{W9e zhgVa^xg+~qP{b|}(O3I*yzSg+y_yk_ z%Zcdn(tSHtJK`VS*wy1`H-x=@z;c`>b0P6Hy3Tl4Z8N#jICBAS6KLN2pIm?}S|xd| z0S-}kf*i#!eoack{iADVm{vOVMOS%#z~C&fHcHNTN@&K3X_p>e_VtG{xl6NH7uL7z z2STg*v*wUh6JOY~dPQ#reyxU~_`UR%;kD_dSJ&;pCjh5VdsZ?KSmDiGOi2C-S~xg) z?A_w2U3GCDPx|({2LmCWs^BbEAe&m}H+aW`Rg``!yjVG7cHoz+U}apVJ)>_C#J=_L z&pg38BQc0U`GB@Q8WM!i4tWlYDKjqLa~DiU{U0X-}o zAxs_*HL`XveQhs+-q%!y$vPX>Iwu_ikS-r%PHF9TZLAp;$r?|PvlE*28mvJXIt<-v zhy{+|eU9YKm|H&Qc{XQxzg5`~)&5sq$+R3%OM}KIK$iUMe-&42?*3wig%(h4Q7iq% za9PTJRd5!=*QvHYp}n&FKFV39l)anyr!ZL>D^)CZ=T}Heois%u|614YI-QIizofg2 z@=hhYPW0iO5g(Q`O*a%3zCtaWev33EBN&(7t|=@@iQCO7pznZW8TAuQl^4WLhOJMB zA^Rh_F$A+vPswMLrV;`FdjB}8Y?;jh8ZV39Z_@a!yVn0YMO~cO%?tgU5`8ltH6uY* zPFwJnvMGSu8`mteuUR~HnOCEH=w7_$jxVmR7R+qfz|3j|p2 zzcE~8_N!~3GRS3y|0D?cWwSNvhFN=%Aj@2XKab$R3h3I;?D*>Jor@H~`~F_C|JJ%> z1y=RuP;@6eX~!F{W)|=g3Zr}NqB!W5@SgF}B2^1(TD!CVPq{rfDU#%#i7b_Hk~#CX zIYBS6WeP21(&RG3<~}YLG!ldfg4=T!}l3l`^IvYA7WzWAW6AQ@r%s&Q!n?^o4j!D4X44n;MvNYSNN+#^(I>YT+jOMOa#GP-A!cME4%NuiF3&+Kig5~8?RuQb#S`r}n7iC4 z$Pe}Z-`1_HOYMx)I-mTT(5}FSY&9V3**rn}>W4cn@TMUjpS+1tDShzedoHu-4czSRH6V$~6AEUWa<|tL6rV3GXq(5K;VgK@ zBb5nofR*>p3Lrll+tH<`%shsln?;#^P;>gp3>PNq+&tkh*tonaFT)oM{9K_>D&1T9 z@B+$>y}!y1$B;U$9K6Wk%(g$z&AWS#4l(x;OzI<{elph+*@hpm+ksfe05OsX5|fhH z$tTQPV0S5q+7Fi@H+lsbh9O|ik<3?Z98FCbbyb!+3i*?d!;{(G3;JKvlBxUqFx$q> zc|q_qFLGhv4wct1A5b{2`VqHmnxV2z2mQKH+A0YaDyN%0A`Nmyyy1%Zuv~F}dH+um zuHUf$;3_v@cFTG)aEUc@9SBX6D0OZ=YAytC=9aVsQ&lR|4IT`AcJDgy{d&zD&L5bf zV|SmbDF|qftnkmXJDuAQAlU`4jRL2nikg(8_4FYHxc)CJF4xWn5NCT|V0HWk6fxdg z^frO=4k2W`_l;%MrsFKC<3I=PuYu{%!31N4GIY-;%j@ZEM$4P*xBD+G0fs@>RpF`@zs-rW zjZEixR7sPQu;obwW@%h%=dJ@Z=oG?aqg*rm-tjuNa-IyNaAY^k9hTL0V-g;nhTxV; z)f!cq#?%9sdIM~{goPhE`Ix^wQMKEa&5j);7|d2n`K%P@{F!=jjhP5Gf`Inu4Hy=% z7s8HXz&}}NeMy&Y)3C}TYa3>BU5ncRQ(G2jb2=^xelh=~8Kqn(nAZ1+R#cFu7Rqa6 z?Y;P`o`+8iRLvNDLd6K|@7&pF(4;KL24)d6tNk%u2VrU%Nt$@J&6^`YNJZKuD zk4~5jwmW3_4})#T9)0dccn@$creAD1%G*p`ejp1gR}^)ERS-p*y_oS^_@x{UbNa|Z z-R1N{+>CQQZ2K8!_MH{$h`{J}WTNL!hCl8brl|C~6WD2Ej&=6Gx_t$h(H}doYo+`& z)-s|T=9mflEl>XKP3@N3c(aLl`n`9WKeT-UTQ#&CCI-fVXqX$7VXYPx$p}c$X4t=c zW&-1VJCn`M)<;SnsLaNsVV2#KG&u_4kH-)S3A;7y9bWJc6^>iaQVM1Z41#d+#<~&x z1!gIHkIjsN$PvStu}7#K2!tVMvBQOSBww7a`!3n})OTkaAU;~{6}{$m1zf47JeUB~ zwa3lTnf_GYy`A!qn>omh+WkBOGMg$8W!pd|&65KF7x0eiRshV3z;e&eJrh^JOc;7H z<>g$z!T~weM;*`3pOwxRDK|3iSJ_YUuS~m%8u=Ib+xbuthD285wE_(t0~rEGYNbUI z43L6Mi|FcG^SyqCEIy1(Bf-)h$2KpfHH(L$<_9HC%YN`%Vp*ZV5t<G{MOqwr{=j-wp`UhZnUT}1f0Sj+H_PhutAGL+Zx4jw+3-gq9Vnfnwb&D-iPDO(Z!~=(COJ--^3(#E|9p-L)iBU3Cn!d^>`)h z@MFugcdLp@myZ(tM&3@=7GiD4TfsX63d^5&4aj(dq<>G zML-T~1(1DH2Qcy-`;4Kd*5cK?oer5Zd<`V0@1{Sh#EA3N2z)gA6euQ~y$?v!ynO-7YKmo)rjSv1FJI-+!mCxu?Pv zH_gyzO#i8NxLV!^FXMlbdDLer0m7>n<1KMEE>Em&rWUJGLS92J0Ld?Vnfo1qV%ACqM$tEp(=@B_joz zj+gQc7ks4i@?KT;bwu_{a=ee;aw+8aPT_A#OIGyxsF@tS29*=}#wUO$rq7?I`#h0P zUuPdgWspT6kU*{`Lj&(8ff0_B%>WjEw&}Frt;}3L=0i+Qv~;#MDtCJ`f%HYSEm`hx zq1Z-`nNy}Io|G4ayW}ppJAJrF9h^|kvUywe&@gG}E?A+>bWj6w8eUU6KNx%hv9(Q< ziX&a7U1U2mJ^AFnh%fF>lWh3c?>a%8ohMrH#?LC0*+5yxZpFEuHF_@lAfU4fUZdD+ zjS&anz3Amfr(LWl z!?Fd7X=4tLjoQsHt<4W|OL>^)TQnU^fLEA#gV#ASyUr~ zEYAun{=kLG@(j2$UYV`B+zJ3l4OPmk2}YZ_Z&emu&S%F215v?&kN%DS3m!B~9%5`7 z(l9T}OZE%F4fUkTtppJFa%Zw*f4R7zv*{a2tHDc#+N$y`#+*+3slGp8Esp#POi1U( zcV+1*32*q@)G}G7sJ-kWF3$rs0w7TLYnUE$-97BD0f}fLR*x*}M_*SbD(rMr)m4`_ zHkUWXzy3-BXx7aGcST~xDHDTvZR44Em&Z5!plRMJ0*OsIiih|WppK*PC_b|d@StDLWYGY$oO9?#<-vc1aA zE{Ma?0(=g>RF%cVQD6rQmu!q=7YR5sYAF1IOni4g7z=<3Nf)xHfEc?ohX45axPnT_ z8@h}8T?-R%`Q=|&U?68asnJLAuLW>A@q^yIsQ|qF^G>~)N>~{;Rujk}=Bo{%kgf5r zzs$0GJR-L89^xjGAuVu&-`y8UjPv9P|kmf)xTo&JXe4SAm^gE{lv|^-l!7d>@zPlRr zZ!kH(PM!3i9IpS{MSEI0urhNZc-3e^ZRC?S_aNR_mRmcgvHJtdG(4QCx@k%!yx2^Q zznkQhZRXT!Dzu;W=SebV6jXe)v@X}ssQ0F#ck9Km!sGNIihK4gGP2KT;rV5oYr*V9 z_Q7?bSiC6wj$7NLS(;eSF&bCElTpZRj!n$$1+>k9Af%t>^ort%sXYaB@ji@phbTPs z!G0jP4epGV8;R-l#P8bkpHJ-q$yDCsLBt%$S82po= z->cijjz*fzzrzihLh%PHt*ES-G^I`sxq`(=TA&MFSMgyEBJcxm^N`RcsjS%by$#Wm z2;^O8@bZ=gPtdiDTUIB!o_B|221x*`c>>rZxs@M_Di}+NKF-8klr5pLHn|}B;!MYp zgRn@mM0nM*dovg~H_2ybv7j}5aT{$3W^7Y1U=M4!e^s&RFVB#e0d+njDo3m}z$pZp zt+PHlF&;2VGe6s>P{Je9T;W-pRT_%taL#+CJt%iGHj!#QIZBQ%hs(*?Qji_%D0e-~ z=W4s_lO;wWKxftUj!~3NSNkh1Wm{twtdBlYN5o*$*j+f?;1#hJU@R~13;4*oi~rG_+%#qO*wlGJ zi&?)8aWICoBqvd28xs%QAhGYpa*D9y;z^(&GdJ~Vq|Ao<=HFPCbLaJY172h*xI9^S z=RDuOwHA3;T5KX+b@PJg(jCciKDW=GTdr$xp7=o@bPxcvkv0z$9Ya}HpVBZ_6XktaCu?}g zqb7!3jhW;sEU%3hK2C(4+fYvdA5ku1x}I25s+t|CK%4~r=6U9^A%A0tZ-kh?eaBe>9zSSX2Ms#_13c5y=q( z3QFhb77%GE>5`HL>29fk(hZW*F&ajL(jB8gnvv4=oPD3`_b(UMU^_dX^N#y|-BQ+F z71(yv`6uFy+HS5JU6uddvWt9z|6@C{$wtZC{u7)ic|9$6o!IZc88yb|x~>opeC{S@;R z2oDpD?i-^GKr#IjXuD)~BFZiJORJ_6FOT9M?UX*>3$b0fhknyO&$sq9i8!+(yQUC$ zPF7fO6G42Kmp+pv6RhmGTaf&Lo6HL2LUs$Vii3LzH9L|}=mIr^hWnY9xyzaSO0P3wnAn#k3= z(><#%%Reawu!ssEVDgGCzrSGZ-a(zv2DS&{6Emx!qn=yhY1gzuVPUnbm-dfZJ41xA zn>HZ0aW?HO2OI6*d+>ol`r8{Yrx>lv5Eq}waCQ<5w}|#k@1fM-n}YXu*ve>;&Ww}i zlrWTs=DMJ7W{7=s2{j&aLRqi zl-8rRrKJDizA|_i?P4nS0}2^+vflRFmiCrICTi;5mQ<&z6wB=Is!Osd)f0^5wa=}l z?ES83b$@E^Y%=HF zGIVrve6)1fopP-_x*J9-a@I69R6E896NrOq-?!Xca>aToYKB#Zb)CIcAEJ~h^_hJ< zJ{dzBmPiMWo-RZ0fV%buxAub~`r>gIf0;m0*_I(%IuX(WRaz5Jt*|nd4d$UqXVSs0 z%~SkiWB6?GOxsBWR)RT#cZ>J1RA2i=g#6?BsS!ypHoU4a2{+|x81JL~ii!ZN5nsqI z<4QA3jD1X>3fsskJ&-k6TiEJ+0Zm`%^uzo2SD#N*a8`>~S>1~hsPoxPq92MNCktQC z3n(5wIWFM3A57&)w>*dGqSChd14r|LAwUPxPIsRW^s~-sMy{Jz%+w=a&Y;K=AMem} z@=Da*J5iM%9021UcZ5Fkxp(`e_fnMMzn>H{4<#zOxZhU7r2DV<)}sAz6tn;3GfiJS zx`5+U)0RhpyEwaw*jKMa7-T z{?DxKW0AD2Ij>!aYQqvF1;^+%<+GnB_qKCpt%+Ophv_7HtHH>8x_Iua@z#{A$sX_& z;swsnvO(C;-6MdEFzr_Aut6A7Rf`7bZ?1mD3tePxuS+baY6>nK&0z=wir&3ObLn2p zm;0$7yIgS3j+H^I?C;dwKrzt8O=*vGC7cxsIhJ&DLO_DqQaix7iHsI3Cp(j>F*6C6Q~ppLP{;bB@!2(P{X zauuAY_i;>4N^sQwv8{sEpFgFY409(YR2KipZvMCDy3%kbk@u|7KxPXm?oBp(dHv(B;Kp!wi|GON7cNk^@VR??5X_b=mfskJwDP z)IPafM&3AW*AJFsk>XDskkGb$?+_I%+4`@ntd>qM-kz(mn_WrXFA3wZ z?@Lx1CbnC%9a4|4W?1P~M);*UPP%>tZ{H*exB_!n@z@(cR*#477=F0c4L4keqO&9D zVTwJm=;;$~_9BGq!j25Vxlw zV;6EqI1b_KeEu*843D5;t)Ryk$BO%TXea-Th(+EJ{%BEb;6EOtF0wHmzx5jthLbz7 zRU?ldf=%%@eda~7){~=X6ExTQd;Kc`Be{{mpnbI5BC9ep~%nzXYm)GtqxL z=$na1J^quY20z>Yy1>lvxL>pN1L4Y{Ke`gyS^~7@r+L1Z_x8__3_?mdv$5w_i)K=&`}b_ts(YybuvuP zZ#A~+ZvNV@cMi}LiH9<=Sax}i&>EkSSINt`L~+s{LhAtIcAMF46<5|pbBqg_*>m=> z`oruWC$45NV%#fknAFu#$S`ck$`&>yJs@wIgW$P&VM1;A?4c^F-H>8vaCllpU@gQ; zH?k(UiQYWoLT|sVed)??`JusS59{V+;eJ_pc>(CJvIXbEXBAX(O?Wn{CT$S{RIp`^ zO4$32X{uFR6%Q#u3&z^NwT*jLIIPuY;vuyDjbZO*2EG%@;bWpn!`=T!;5Pg5%)i2g zC9RaVqgJpjk&oyGIAE#t%;~mJTzVwN(dcriNm^)JSbL9m<;`+9JBhRzsq%h~^URPC zb8b8ax4Jyi>JQPC(?=wl3*8 z@YJSPeY{U6N_|FwqYXOo(wx0tP}Y20fZ1*5d0ny*1Q9(T<)XpMb>RhO|7it-46GVj zYmj1X{9yIXaC-oB_P(0VPm6&Bba zM)~?3Fy2`#R^*l4+?r9-`8(@Kv{i9VXV;+kK^Bg-t+PKc{!2}t=$}^ei}+_YPM%;J zadX8cK-06Zy~=%WNBEh*zdj^zHEw0Sf#4{La4XgABUo$3!;$DYWT5bqU9dclbvv6G z$bKLC`m-XJ)A$pbM7#`HBpU}2YaZ;V!FZlmb4*F1G*ics+Hxm>Rd^dhh%scLIllok?a!sezer;G}09*bBv~&A%Zh zv8&ca5c=*&=h3fHCaPm3a$GtECiRvLB9CBllHr=0PwBTv(}ovd>k(!ya+H-RZJH?E z?DQ2|Jk|19Tr8MNS>O{P=XmMBA8PfuhO@{-n5VYF1TT_vht<5KH9Ev zlV}g@Fn1uJuZ*T8W*G9T?zbLwIKR5~zp9Z5@$x^>*riVvD8kpSl7Mx4G0uCrUg#3# z9<<4s(oF+uSjw97re%stA$|G4Lq8#&*QxVTuKl<}eQ z5JX5+%!Ha?a~*j8WseXL&{m@ABz$dmp1CAN5`XhkN0TgEruNrU@Y{45v0Njre`3u2 z2G84)9VB%ix*DAbL&8V4^*}uGNHdF7ilyl_Eu}__=AoMAAE=Q4r`6-|%xypKOqk7J(tK<|bwL z{CTpBv0z#&X|iSQg?e8?(^2Vyefn6#SN(s-TGxi+qHBQwqulljCci?uH>E54z z-Kf2WNQ-xAhV+w+bfZ0Vr+ahmhSHCn4R|O@pd3b<%itB+XT5@*;~wH@N6zSjM8S$E z>JWbIn@?9HcJS(5ExL&??QzZuvw~}Bl!xKhtb4~xq|jN@qRRS{aPhk}R~`fd?e*{C z6iRX;#~k(QdtFbBZ1-#0hdu{So2GwN6Fq%SAn{m7y*?*bau3G^`g*c@3Tm498+fNc z6~$N~F}rTLYwJ-FQ^!%!`HJi?N>pfIW^^n@_!`Sj>`J;>n#$p;mGah>^s2T*rz?Ph zTk0dQxw0n^?O=};iCbL<`@&;hdt}FWBqLv&>GWyo&8PXVTejdA>h`V>bDp}nHR{A$ z*)PJQ9(?w!HvP?RWo#DL%m!r*@GH-d1#U>G^xo{npiJ()>}D%`-Eqyh7OOsQuG~?- z?4|om_6o@g3kIBIGewS$%0$Ayzc-n}`L9U}jA7QcMZj>gIFaEHIjz}KBMNJ3NI$Yz zN$0owM+Qdo!bSE#iqW+$$4yTHw??K&U^}MmK(eE`Rh9Q}Qi-+^+liEtKg25g% zF^~IzJn4_|@DJbjaxz+|$96s$1_u2y6~v`K zsky!TZe8~J7Fs=_q#fn84ej@c79;?QCU{XfOV87if4q>@dH>|026qmL4mKc~DW50o z#*UJxhi|cTwS3WfCrbFnSC|oJDl`J+t{Laz%Lf}G?7~p*GjmkK?dL;jjR`@S zymD#UxW^mFp20vTUsJD_YyLS^M>|8z@d#8}_=wc)dw7%<)Cnv=P0cUP5_C2Y8{=}t z?{(oh?bXV=axvm%y+yQg(dMSOJIk_}U3m1KkLJMQQU)7gN%cah!fAZT?-tRZ4FEy= z=4UK!eiH6{3ui^=T>43UoJ_8)*E&Z-gU?G9f$~l zqzbv1U%u{vSk5<=S(px;lo3Zcmaq5DEmmX>MAQk>n*!SBTSRaRrmS#<#u8)xJUj;bLc`@xz_w-vPFTsD`zn;n6O6fEF`n4#` zl{4O*q7;6@V&~b$HSN-2#4#+gu5RzAmp#Qb8k}c+4Y!(kw{@!?e45)7$WUBhI-M8> zZg3tt>+TiPEq`SG0uw6v_RwoTDk5K{y22h*p#~wyKYEdFn^zNOQ0!$}P{nIFnE&(3 z-vmi31GZnLA+29N3J!ePt|D~1py~{1cTy$IKA&q@l1=N?!<4g$WXs~qy13gG_b=69 z)%T+8=QXr`lI_$-v7fY}X4X=FMMV%Q!Am}1?wrT1BmXbW$hI#ga;$V>xq4PCXC6kS zyT{Zkdn`7IV2QOB+n=?Zc?8-mm5RFAax1UV=8P7EH`LhLE)1K8pqpx-hSq{t+f z-wWO_EX0&5a{gmD{fP~23^*OcZt5O%7#qEWG$Ct^Qi%&asrQ$bzmX z4|GV3Zwgp{vSb$o3Or2-=?sq=vmtbO3hlEROd7P^lgD~Iv17!JBxwJDz4a>c{vTy> zEB|^dMCoU!5B~?@*zI5aBKK6|Y*G!Lq-OLhvfVYPEq$PIn5~C4+m#covFRHhURb=+ zMgXb1V&$te!Ez-_gRRNr{596@YD zF8tIxJjtn=O~}oFkA1Xf4U^EL1AWG7<8OMS zIX1j^Cdh-tMR6nG?dvnyjVj{TY@aFx%zF=VJXM>D@<&=)H(o^&XYD*xxVM%1mWT0E zN_VH1|N8IEcxzhq5#0bTULO9*pr$XdA6D2OFp2fw`fKjDeSDW?Y#xOeD*3)Xf#34I zLuoX+XhA-GPk9()zdEsuolTct=Dh}=p1~HkGJ}*~bRR&){kDoJ8te#IZcPHmD+BNxdc96 z$8`&LrMk(FYK0Zm2P?dZ5)RIVVm@8HJA+<=)!d|K&l-NXwTTo`tRy+$a6H924(BGS z(9MPyE8Go?w?q1p@R$Y)d)759@Lm_hto3={C5Dv_mRr%B=dw2LZN7mrMePQ*@fuqB ztnOmJu=^xnvmJEW1+t1>C4heE_sx48zALKJAr(27Dy~^Qh&-;m1%NXd{R;aVzj!}K zCqp}|9w6w+y(EE}+95jzf|SikJ-wEI0{fC#RDV5iyE!hnK6jzlnky00MEl?xF%AR1 zJvOv&0%ImdHazuUbgeG za)hynWV;)s?YzoJOn3J)cRs8kyt+VSMw}=ctekd{Aw*TM!a(J9 zBj;d5JJPb3$bV_0tyny*&Ttb={CP%@?aCo~hChmX%5b9!%HHvp+udPhfaVW(R(fwxf?YB~Xji9_t1`sT@MhNnp! z2$q&@UKAwvpV?x0LsGR06v+us!>7LPDh*}NuO`sjhbO<#t)S_}-o;@Vm9B`h?{ zh0}QL6yuCsCN9lcmT&FPVX)8$TEPxmQ5*jSxXBT{tT^|iB-ng&vhzcMef}ETP3hQpSm+nHrYt%pM_-zRv3u}0G*x34>w~yhj)ON2trq?6(x>13=dw*usgeyy+ zO%-F1q4oT9kYi+Ri%V)q@|pd*xTmIr*NhS2!PFM(2N-ANYjsewuR-GG90C-cV_gyGG>=w zU8OSAnykz4G*K$8e4rW3AXQ2UlOROm&Tb7=AXH+E|70wN)D+zHPcKM}7)~1L=4kM*_isRJ z4UxqeK26k8d9%dAj<6l<-f?_S2SjG4m<51Oy%P%_N0=YNZ+|9g^mhW>e8<=*o3W9t z+?Zw4jYb;R1i#v6Yb+09z}3R$B{J zF-HU&Fbino;$&IdpWc2@h)v;BwSJlaNnQPL!^W=l996oYAqF-1djhVEsQr#9y&P7c zS+z{nT%rLKa&l(QD_=#CpP0E;r{FXYQa}^W;pSY5dR03jRLUZ?=XEgL>mQFMCyV5; zO%Gyh;$HS*sftqa{c>P+7BheTi34ux))6YINF)$;K;Kk@@{0^RVQHzt%GA{|%hRG@ zTzO9~Y%a4^tX0_9SQ@|aU}Mo`6xckpyJ`%n9hxSW(gcbiWo%?^J2KP%{)D9>CNLl! zf;SdJ^HwA4b3)5rtAW@R)Rt3(gF)3S>8$79ilYb&3V-vWPNJ}{@g@n{Yx4^1f4_`@ z)pBSgVt#1*tS0EwGWX-aZ-d`2&+utu<)s!vM=(@1HoW<+%Idu)rNGUOGuxKeK;P%h zQAK5ef&H$y|D1>bh6KbFAWrRncaj+uW%zhA>poP6noQD^Q)1Pp#Y4u9(0#}G6r|lC z^fcT7gcoI8nfBSZ#00*CgL= ztRUeK*y*WizbG%6r!_lm#KXbXXUD?wbI$6uMfjBOt$S$hi;lWAl&x5J#}TmN<^R+) zMEp(|lw;0gPo{IFx=r0AcX$6jkNY zhpV8dDx#BewjK?gaOFvT7!>I@SA2rrN3y}%fA~w7Vez|3?Js@$Pt%xiXZ7JlCuOnWX&HIDo$ z29=ksk;bS-5aDvow#Bf;CPgz{PFrUsyjp44zDdVxHH;XHip@7V?l-$d+QXQ5;*Ww- zvS@S87-cTwD>dkBvq%fN&7AHrs1;%7Pzu*ze_N|#EL}zwUWq40W`}rzMboN||MZ9uiPJH_ADi2bEI;-7 zm9DTHxm#T^XI=?UVISgE5eFbG`?5=~_)!}4wyfLa_NP-6bg+lMP$~NYGZyc!{(w_a zJru?+=VFqkByKZpn#K4>YovjZTgB-*mG})i6(0>TY(kmQCv%F*nml%UvGob+7M4t= z*YE?jFli|cavR9}bdFJPZDydgO;+l3-;y#{|Ej_3 z(w|TBlQA$*cxhLeWR}qLw>0DTrHM^0mh{TUr&}NLvhf6-Y<=UEhiw?U8Cl@ET_T|R zU&m$68}~(^4p9YBQsFbvdonrxf*j9mmVw?vZ zMM;g=S#{q|VzAK^lCYD#*DR-n|29YcBp&>jM)=g%xaL#%EdAf&`kA_O)Qi4=ynOiI z`0TsCi8FU_5Kk$Ix_xG4|3*|meqMxNhb`yoJJ}}3sx&=1LgP}=4&}YdyMaoL2iG1V zxv4tmEP{v6oE@-6Vg4bx&bkWM@hDXsP17eJQ^?<8ulQ!R!J1GusNJ0( zFiR}ebQ#Ku;d7CeyT_Un#L~~%nTSKij=AQUYzsaWL~7sx^{^1R>OX1s_YlYrke#T}8AG9-Pdwa4_=@%zy>_7>9IToF=3)zH{5nMMr+E zbmZ=;UmUg`Z2;!quPGQeM5#9jVPE&g7!zW+&FMgdLw{KJJgpKUH#Q2>Voc{DGx9k} zlU3z(%mXbIQ{L8ps{!>WZv*@yV2yAb*gWS>K{&Pz5!_1og75#V<^aE8*@7t|h68KQ zIuRfEm-)>)DsGVzvS!p9U|e4b&$k)e{~_k~)4Kb4PjjPbK&DAYs5*y7C!P@KRGA+2 z!qR%V3Rk#egcB%gRu)bFGrIuD5Hu{?kYgAHqgwVe+Ya=04aYU$6JuIt@zJJ`|4M zvQcVZ%XXpt#pJ5HI(bPl*)yT*kb}S~Y>XP$iM0qb9=`?F8)lpNRq?DQ|Dt7(+^y#i zTNxu65-U~y6I7VW=mD6#IIWBk0Xq-};Rz(kC!l9-y2&I!hdfF=EtvGRFN+!yc)CyY zwrdWltvqjBUVPF52zhR1+YKEk+@8Gh0L%3Vuet>wOgb#sYE7ZN|8IBzKDlvDV}idQ zy8tY1&hooQKp)}|%REQkHf(3FPNIQh>VK~7?`lrcd?7Ieer$X;Y+lf3?igdEpO?ny z0X!Q;hg|9-)GM2;ZUJ%Kr#6(duuGJ)&W-_Gqii76N^{z+o~`{jltUwC_BzMU-ETVP z06A1Ad2a+5cZgeb9NcHz-b`C?uDTrpzb>k26RM?!)qP(w zf4~nP`}4D{ci5tPIn;*#J0oo&?c$UuRp1gU>80Z3f9OyC z*^?2(%mSDCb%Fx(Y|Yi>Ax-9{^qrBPBx~FiONeHEKA7dcebr=-tg!hx3=_^@03p9%F(>|% z-ba9D__FVJWIy@)RjexP|B6ci*|5D!t`0CO0jb;;F72&u^+e0NfnEd1-)9)C>6}T9 zQ|5sO#O5Ez`X`V5n>dY6*94|Y4bMv?^GurF3)h2rCzS5&0-jwR6h?@3=Je7^vS$Cu zv7O839_r#$SPjAin+bvh2hnymY6K{`(aW%~h8V0E7w|#<=@X|r0;*$_R5!AmhQo&J zDPuWW4r-Xcy;lE?x$HY2U~+5Ieqm#io;OGz{ra025EI{ml&{<~6Fw%tK*v~BHWw-v z?6}~pa_-VbqZ>LwCRw$e_Al9b=oVYFZ{vhHEuC5>s zbAQ<0t+x5Ctgu#Qg-hl>RGyUC_UR?k!;~~{K|A~#u)BDiS2b+XduK0j#eXq|U2!y8(29PQlJ4g$BW+PHVxs+;|)#Mt@b`4@T z7_i$XbqblDM8lJjD`UfPaE?z`+~-k9K0?k{4x@QSb^zyC#uu7(Ku~VqM&itm^MIs_ zP09RNXsM&6)Y%BB#^^iPwKn3dy{g@LjJLwy%Rg{iw?3~yt$Wh%%@o)oGaQT(Fz#Q^ zf6imUm?U{SK=jI&gbEmuNekOkzKJJ+)V6Q&RsIUn7MU#wj;boIiui!ne-{~n2Ey@G zB|QIw_Ken|-k4~w-+}^_+XG-NeBywRE28~-+p7O)AQ6r&Y;9H>j|(wJt$$@5w{%h8 zV9=~CH7J#*GsV!FR_+x^0`xMdE_RXB%4tMaX{C+vJ>w5cPq!cgwpr$jrY)JAO@u>w zY9HwGUdFj{{fJpv!iSfh{#aZ6Gn3yfGt`o5D$D^&B|vQf$IQUbH~`qnp~!~8kKlS+ zB|<3T*AlL#7%%@Du#{|p-!!nN;CU-I{4^wGdb%LBFu78)T`d6hn!EG;!+Ph2PDnWlyG_jP9}NleZv*tS)x{+Q;>u~8)k zHcr4&Xj5#rTn5Um;>yqaYkqMSeLOfEC5;(SBRYr~kRi;gv4Ss!8Wer)JEwn7L+$zZ zd)`k$sO(u1HePb?LGfH9wJ1>3(UK5^6Fszna_0&wUEjcC^PFF}@{>BfnS8;idT3kx zl|y&Y^(k+t%FydUQ~$~t1zKujS1#3|^=UW9O7A5b*Y8!aX=FT>c{RJD%uLj>n^WCl z9}k!1ww{?YSZgSc#ze-sYJ^lIZrxWm+MdQnew^4T&Vx`U#%tx7D!6Ju3Eee&X->Y$ zP1Ph12Zd&W5X>R$?@fNwBJKAdyidc;gMPBs&(N;iZSkv=e-YOax?9YduBzG>{tspf zG7H$s#0#0*GXI<5Hc2E$#nncg783=veIAi^?@`@aQbTDMFi&q2SvBEOO^LbY7GIEyHy zfr0;-)b1i@YNN0YE*e~tWjfG?F-C|aI}}m%sy(<6B+0Sso%W8Tqo`&Ch)=r9SDZ0a zYo4^RpWUehxOy|XfY(4RWL$Jp0+c)Gl#c&d6RRl6jkUJooZmg3SNi)JTqX?n*pDlV zG`@MV^5lsA_UbQ__`L7GJlgUi%waa_eYc?n8N)>+n&gA=zF56!Os^suWo%JlYC{C0 zc^lT0xiA^rew|qQhru^g#F+eBCU98Y*aZ%+_F8F#Z_c-5VyZhNbQcV`9yq-7uq@@A zQyk_Ha+d^SkUf_`BtPACU6S1Uq?gPcGL+`w1B}E}D1waMfRe37S&RzTn8|vk zUq3|)C8Qg@j1)`pArRDx{fPj-@z)%eDx>FdF~Baf$7@*KS04x}x0=$}PF3P^HB-nC zRZ2I-fL?L>4P^qMzI-4?EqkxYbS=8FylQZ-DHH55)B0ZJ5#(_U2IZXo*J>}Tl#fC9W26j166w|)nvA;c9FZ^zB$YT=UbOBPh9zlv0C3)#@y z(4pyE5AB%_M|Y=^6)6U{-;w|JMb#nX@Q-2A>u!2tq=z7izeBBupi-jP-__wc9aQ(* ziw3X7g!Tp~l+Dq+_ThWKKt$SLWz3Z94SfCvQT7s+97(kdih~g^XRruRaV~)+c|qZ8 z4|5KIctT=Afz(h7t-*Fuf^BI}_pm^dT1U$(z5E6?RvdC`k@gMhqhgR;RH0oh5zH=< z#07o&HKHKdP?ygS*KE*fv>Ao9kiTSD2~Uch%eO<-Nmd=VaIM z;ZQ??JG|qNbyu|nc?lU~P>ZM*>X0^VC6#76s^`2Hc#uWBb-VxU56rsCaL z18WC^4zLzbx1G=UhFQ^cq&_A8nIo|H4V#9lz=8pdEz<1}!>MX}tjTca5f{~1`1aFc zF>9@t;lZA3M&9zO(Zic%IQW@Xd*ip`z0 z$W8|NTBzvhohUKF>A4VLBcQW2Xks$m<#_U1z_s%;duj;NlefIab`Us@YtWd0;&+6C z(alF!q*q$TI%Fn6!!}Nr_yjv%G~L=sIt>*)zP|(L>&pN7O!FF9mnjFGm=uziWptuU2r!vAxFFVFRHGFF(NG&A21ZX|= zg_h~l5KoL6$Y4c{#LP*7^DiBe$e|w)Bvg-;9&66M&k`48heE`vY9fej|QkS?Xo&E#!UxMSFUcWN+;33TMxYyr8CpBOEe!AhP zJGwgh97)TqPqS)KyZTTIAvGw?6@~~vPjSM*coUPKdS2dw2d$>77Is_GL$O6xQ8llo z3f-N;dY`j(p(c@`hoF$wxQSYuR~Ix#cg>1no;>Z3X-r|^?0kvxn@gulhw(|z{-7Xp zI$j%n+^~JspYVv^viEI9NisLFqZ8|SxfekHS0#Kl#6$M$N?HdSe+Uzuot)ERVb$|? zVa2fSlpnA?{w(q@DayPYD&Qu*E3WSAjnxmhATVL2<3qf{6YHlBjC$8NcS1@QSQb-F z75};4UdLp$bMV7_uHk=EI}rLz-ONG)Q#6MnVMcLJIPwBorije@UhnWK7luNF@D;bd7U0$Eca))KL`|=+8Ex%3?ST zS8+#vZiCy6CONV#XdHA58D{3r*0v2){qVo<23(&a{d~KCh`!gpRG63 zm50iLn+dv4@nk3iNMn%)`{k!s)A(W8p?xTlxp*eafCFz9^9X?2ssWXbuZ=z?HJ$6A z#Z9IU4l&rvLEo_Dweqz!nhR?E52*h=f0_=j*e%+;)o=JeD<;|lvggP4vSU9S4lS}* z!%0`X0W(A@-tL-Dl}cN>UcYL+uSr0_cWdIb*I@Zdrx=^)&16z=pFJfB?+Kgpv5wSS z_T8WCga00{d2Jh~&+0!W6HM(Qc^9s~Kt6vHf-q}`rOxO%m0#MZ!!dt&I6&mnu-VY@ zMLZ~q1*S}pcfY4e2f0}uJJ~zkONcseO`lUZM!YSuwWf|L_l@q?Ih&IJIJfL||J?XO9-jQG zfWy!Lpjb^`q5C4@dPch&YWpfuWnR#2MKEQ3a=`T4;>n&DMKVVeZ;_M&_Zc0_@+BcF z%WtIgY)>7L?iCD^kV;aTl>cIN4~b$u-fb6};5}_ofF3;++Q})&l|%X5DT`y%kITTD z1EeKi{_2`W5(bu#W35A=AU{kyJ|BI&8)YGGd%Qm~YFFrfm+E^)Qt2Tx{p2&#RxgL2 zQ{7UbVic=!X(E;1MF$G#J#ZL{Si3*pveEXdMZ@YYp(tW|8h~kCBZ|Qvs(7r`rDcqcmV!QqQLUq( zZayE#IbFy@_DzBmMjsawR%W>l%Y?KOpUx(roFuJ;*Zix!YfqV#cO%LU z#J$yw8V5&C?_~-ylSenSNRGuv3F(PU=9Bst>Q3XN3&_oPlT_8f+~=?#EN zboiXx(l@+;P{G*dcx9nqPZIScHBM8*RmkJ1e|!4)Q^2gNKYWUXIa)K#nQN9GYqs*z zJ{*VdV+H;Rl_?)ay9mO0wwUp2X<`|+8ie=d4iBIL7EBFMGR4)6n@mVrb*xSo61f|^Dzd+YWHFiY)4kVa~OHMQ+v zO<58%LT^4W6aD61!-`GtB&7~>@)&bQwSaYOl?zXPqYSG5P^D-ppVe{Oi3V)Qt5TdPoOUhCTya3?=A# zIWGP{E@_)SGD7qF7031Hq;}DiP1fe8DThf@zlr!P4YF%EC2A>dga-&MwSy~H(#mQw z044k7c(sev6%zWT$ugK9=GZa9c=iJ0lD7i6r*J7J;9$-ItJ~i%5|ro%XjS222eO(( zM0~>yr@qt)fR*U$kqp_+Xq_wXRV~%T)T*>qIv)&*UBH%YvhDgIZnqCL;qLq7t)OFN z^)153V?yC%*?GqTgo(=_Rfk0hgU3~w7;?H+$^I{&-@huScH9LBeAIN2`9)}`urt7( z?z34NNfw4dTkfE2KGKcJ`hyK+4XDft&8*m!KnH(tif|XwcZLxjiPf4!o^N$%<25Fy4NmK z0>;6~qG}TaAZU9pKg3yXNTfRtLKy@wFnsLNF6u7!n!t*xRV6vu^K?aZg^!q(NXj{X z%ySETh|pf~mm1^8`^!oo7rG{51KCKruiI5U^FxoP2>QCkFlQ@vvy&&h0zHuyI20HQ zroCG&F$iZ$i9@R|+1$`_14+OJ^Kn}u4&EK%DVF#hUHVn>G)^+UB&T$Ty0@60*%9+e zF~TuBb_0-9)wHIZa>8~-ScoQ=GF4&F7NJX?1$UF5*p4w+DOL8 zLlyZQ!RQkV)(mi@7B|UDD}QKN++=r^w3q(D2qhR6Ub!Agbl>0#l-cf#^*u0Ssfa;W zW~9K5##0`C_eNQLXbF3(0WSO-6XIyviaT~k%0z3?5bYX57w~e4++4pqeFZR$ZPxv- ztf#~(lRIX(OG^zz(l1@@f8!HKzC~r$JLpIYCD1OsIwmU&$}p^D3d}jaDq{82RE>FogPdp{Sp&rIJxn{Y^DRKPe&O-NF zVX4yddqdhu%M%^l(g}*eRL^l*+#hJ*wtQO4$>~D!g=+W>>cUvr5Wm)RKM0IYx@gsJ zC>BHQ!g)7*B^SqhK%vB3{||i!g1sK0{A0|Ev9*rELLbUy&hTv)!zc3>M%9kUTn37f z=>5OR{V*dqV*z@XmdG3~r^`q13u)}T)GGKd%0_|=O>oz0oCW3GwjqWIC=^VZn&i0N zgH_|l^rpgvO=Wws?)_1$kNQ36slTtEx2O1x{-BKLx|Q z1FS>UYYyEOR-EEG{=zo7n$+P;ZKS#mRI+O(Y`pVsu-(OhE|Y>BoZ<{kP@?>b*uDWT zrcLwD2tMW(dp#|{MW zFh_-|rM{Z2c`Hp3H?#bflKwoCK0_~xB`B6s8^O->10P9~0r{m(OgIs6y7 zdpp3u{HmB0#U`CO>alJjA#+8ox>S&v7@|h>GbVEB$4!}{;mici3pz==>Jr{R()d8D*~*6j!(Dm$G_Q`_G6zAwo0_4$>8YuewC675Z`gQl0sQ|rhRb4 z_(cP9Fxyyflq7Gpi7l_;mTyDZO0OCsl1J|dyn#7NO$=;4B`!_1$7GqW_myQb($*Rw z^>srs=t4RNYYq*z=B6UdXk}_d2^E_~_ap=>YJw!3%(@nN;i3m7HlrBFYuta98Ah_{ zqnBA8ko=u;Iij?(0&O`DR?2Q0C%y6`c+_iEc9F7c5~U`v7Y+vzR`eevM36alnsy+p zyLdCVZ`#z)ScB9i4=G^MILJn5dOMk`7ZlofLPKuys$o{@kmFV4G^bUuoxPCe;`$Ux zc$g%y8%(R1TcoCtj!LrG6qLEG=_3ogDbB#q8N_Rg>YDdzQa=5#C@*WFkWPt_r#X)jBB`Bv~K=CQj7ZLCAC{+=Ln(7w=eN>uk-5exRk zLc8v*#R;d)S2DL&IR6P8K9hHZ$Fy(3w}y1Ei_&Bqg_zNr2NVVjnDc&8CCb^rM21pN zUT$D!k#d5`3c{HQ-=sUZDrH&*w{u|){AnJDDktL)LX?GSxl(2Q)3s`I@?f&b{fVJr zZ3Jco)88Na;A3sspVcY;bmJ4)a?oCZh^BKRu@ypO{c|QRFJcI^D4)9nnJT3-DN6!r z29PSjw9Q5ix!@=zijG4%JKQwlgPAH4CI{Uwyi8BxOD+Vd)8i_ank~i7QqQl_H;f@i z6OG2*@fOLM!sd?nOUgMOlu^h^W1q_$Udy}PGoJDlzng6kos|@eDz7k5dSk$zeE6%V9$0ot zx@D040Qwij_o^wL*kfNPN4#-po|V%hh=r2FN!LuyATH7q!{#TI+g-mbyqFJvdpibdSSwnc*ByyV4S{Gq>!2`BK(AWZ)C&Qs zmyaj0i>i$A2+wTMDqetzhctwt%I%+E29bK03juMP9+HSC+fPJK1R#XEYx^^>8uvXv zS_eDLaFxGoaxN!N&|b@(S`t9c(o}6WFse6(s1KluYz|rOzV5ooYxlb*9fjW*X0Hlk z-k33|Gq61qAzECkK7j4bZeiJhYU!lnY@#jDde%oV$YVzeK9)bL9QMaid5XRz329Ad znuqd+JD)tcNpoof!euIh+$>Dy7?$v#J+Z|i{f^D!E6o*v=$4mS;9pEAcOXBSOa6aU z3o5!V{-@H9uLLg_UG{IT)d15<8bwKduXfq+36YE3ZfbmlHOp=amMQz#O>*0(i{{$C z`jNMkwXVcd%{~FUCdNP|VonoYR7d)y-z0%%MrzXL2(;O{U1FLyMeL{i?Ejd0%dj}2 zrE3((0E4?b1b25&NYEg`-QC?Cg1b8eNP@e&y99!}ySv@yob!IqP5&h`J-c^Tty;Bq zt)HxFqb&0sOCk$kX7h~tfVo7BExCuej8b$nTwHRuv6e#_5$SVynNJjUNvbXDHS)P+ zhUX9*_Kb%jK}Ogv?Pw{c!+YwhJ5L=1fSvVicle5Y`HC8hm}&r|_0-3Y8nG|c(;?1r z2>yZ_KDBfS<%9{P7%0}A0oVKFz0Uq|w?<@#2IaNCTiNfIdLYP(KMQKco=2N-%*x_t zzZ{#+`N$w(y>@+k+EKAmu5m0ZeX3Sy6GXH+EikQS&wE;Y^p@l5{j`&|=PS^S+ohgc%7==GHo36QAyiSkvF8M@ncvT1UDJVD$!aM; ztyjltSrB=f6P%}n5V-El#W-e;ym$uMVe+WHFw4vcW_3H%Z=`mZ z3u-IVN8Ga2b}I?Ji%#qxI+qWO8Xz!(;l}>kO_*D<#k}z=A>+6)`a^)q7Tq%M&g`CK zt(J8NUzY)+HIh}&+6yg{1LqwfsA7wSzo}85v6fd0*)nfz!#w@>sAx46AK;Z+eEa*& z)PmQRd`qQsey)G9_eeY*@&kP!WP~^VcMi${YY+%6dX0xoI#yzE24}x%KP)$AeRc|_ z8wXz$v<{2OcvSz5YNBh67{&~{1-BC_0VYe*s8$nB3l7?>tNAC02L=l=HOn2f1F6Bj zT~u$IYaScy+M`6YL~cdHWo#MWnp02>=)KLdys-Hboi}7PbFd+6=(b4D22!Xt?l&Ke zm56sRNl2}n5j7WTHNVtYGl5d4AY`gjzw5ef1m6f-!*&`)c$s5%yw4uBXF36DX^?I% z*0fDsJ7?gADJ6?uLc;lwLHPu~CP=LpZBojzX2ggwH)~O*KF=AbDQ6Kb%%-MelioZC z>fC+LsY4;RNZWzw62+kN2^kCwRP42&6OMeduTvm>rfv|SE>g_34k*@ItNWr5fq@+5 zpUyYHY23p=TF`2DiNlC5v^uD2SdBb;6GB#<+T>i%#5u_PcB(p?S240LrdPJaIzE>tj6uo2BXbPXA;=oV4OR4%yM89e@F z27zR@Km{{w&{n@Rrq!rC8WpKE?zA2EMFp^Qo~Y0r3-;+FUj4OPBt!>wlU^(h{T~k- zgzE6z`u42=Veg-WV_HrY&~@VjMo!WB;BXEXNs<)Rm*2;Y&seeiZU*5!fiL8OO(RM5rr2#8@8DtGPLu z^Ic$CrmN4Jvs=}v{abN&?t6uM59qvv=ny&;s`Q&4QbH0nyG78z|6E$8f;AsoM_?je zo`-$4uD=KUDIj+v!h-b3+8QbW>s$KVD@_TNK7VfnTYOZLr(`r@Yvif37Vej(&k& zwX8A+B-TY7ox42aM|)R*2aa&VxSj&YWqu#D)yUx3;PuS>W@x@se%V-vH_Xk{DmlqD zGh}j?HC1ZM4-=+?r*~da8l4#3Eug;a)Eu7EpMKTXIsW=oq+XA=yY zDt7Yv_FBguT2Cp|Vijqm1LAq*>(eB6g}jQB$5HkPLM)`dwrmu%jGj&vfDcZynHeuz zd=vKEgG^P96F7&16l0$D++TrLL;ZYwBe3M|GNCi>kX`(y^zB>U>iDXssMRkuntU*} zyokqHoA@tCvDd35qxhXI>91n(XVTRft+sb3OAWTJE-ogfF0TLH0&JvZ=dd!Eq*P6c zjx3s-w0|GS8P28|;7pj8Zs*aLOzXq*L=+&s@MNd0fAeXc<;ucn24+{Z#%yVBAMtDZ zCytZOwx0=8$g~R9IBeHh-D(fZ5GSJHM~Xxh2;9o3b~f|)eC>e{RYf)JL(6?loRw?r z`=VeyUO`ej+HN>;ara$Um~kW5D=O$~V@!}=`7g;i>nQyIdUTdxw;a_S`kWve5o#yp zBPL3DcFF?_WU2@pwsAuTc+*-W52M!U{gf#zGg=PBy}y0&GJ$zJlahKnS1_36=apaa z;agKXS?7Pz`ZKse;sjD~2rRnApZyApcG6ChxD@1TpSVw?r_sFwm&fJ?iIURjA~ba6 zsxyelCxFyzo9#N}>m5g8F6&6EV=rQ=`$G{lvDtWbdqcZ6pNGXnADUggg^OTq9yBxS z@cFXoJO*4WkLjv*!cGwDm1p1BZguB9`szb(j^A@l9GzflkiOpHSvg$E>@ZK6+#b{T*fy<^iyGW8?f6A|DJfmimL80$68oX02F`cK&oM|&w4 zj7v)1Cz@R-sDlWg`TB7cUK#XM!MoO0oM}uVkQr=_+c4fiVtN;beXl?~4gq|mUH$elZXEeo-85^*D5q_Vy;au^jQI(hO$PymN!n}v5UKX1>a(bEp5TRkJBATiuDlnM zO=1Nfy&^;1ZlWuqBMKppJhAt+!bY$C>^lDva}nZhuPy%#0mVvUlvClXenGTSMiq5< z8xshAg$8{JR4oFrd!mjyo6K^3NSmc%i`{FfOPY zx1SX4`TcLcf^&au#!6dtvDstpfBqQM?d>PxCsIm^ss*2$VI_mEKUSBr?> z&^KSQStgVo**uZX{;pD%Xl|T)yQ%sgQEiB@;kTBB%y$k|rSTnXKO0rr6?HKqGbOy% zyziQ@=&FK5TfH)nd&Ae`D89gJas<$8qn=aqNvV#7cW`+OhClrxF5QxIVtPgY^p;N? zOa1`{u`{|M(|tGX)aY=ZR@_RI6hdn@y4;gr>665$L4tQRICl!r*3r~Uq99`s1= zk5zb%#vc$FX^hWE18yP8*1#`(|0@u<7Y3?7G!%(ocRPAM(crEE1~?ZZ1J5pDeIB0O zvr&CKnJ2DqqHS_Xmq4#;u5D#eN#rk$@`^6_vXX{FbDBI*&p?l!!EMXrF{x_ryXqa2 z)4tFcAErd7QR@^2` zy~pAz|6P4Zs0Y7JXdEq48a~z$w>fRsp_jY0oS23PCiC^4&#;;f?HBz7S*wwDvB7>~ zui63SyR_K}<@t$7f%7@uw`ny|{r+I!MURR9j@vL5eZcO&kXChm$P%?q)v*M3bI^Bq$5|j)|O6}ENzWtGSl%>~s zjgU~r7*OD#)PqXvePaxG`He^)rf386S(7$r+BPwCvmz%lPHdMtVcyO-M+hzoF(9q=i zC5nIQqSu(Oz4RIZtCpMAqaU6HCaJs(>Ut-R%9;RBk)8Tw=_xg&LXZSbOx9=dyCzd- zyLB(XqlC}la){j<`H6r)JfGr|3{Zsbb3rYJXq~ji4TPCJbQx0xUts$zX)TxJv)-dg zJOZQ{yXn5b?YyBrtmTk(zd#b9`i~FBtSP*1=j*v{V;D2})bRAW1wDoH2DTk9F7AL- zld{x{kt;S71saUW2VxOvH|Ps*WaHD_PZ~8184dCUnu9jd)78c!F77Rc1JMqP=EXz0 zua`K#H5*QuCZ03OG#iEm@0KY@yw3dt8R>Hxh4(TYpN)7!VuWAaOk)nGrJ0j}Mpg%J z(Z~U0AX1`$q@*M-CEUgPcDILsH0%4LTmpr3yJ+yYNY(QZ;=` zb}3mE)IAfny}CJ_1R%<5aoH^IBew)(8~AcECqzfu17|Y47kuAdR;2Kre002r(+FY) zErrOm-^1wpH6=#M$*DY_gpQqjf^2HFWnkernJXHA{guThRPaN4^2}HQ)>YOr$R!&8!DQ4FO_8zvX%I6Ze|MR@~SwHrSFJC zJd=8?8&;@*@w8v%+fT7Ddq&=W*#$f#jQ+f|HxRo)lo0Ufjx?NAc^)^js{GI*a@_)4 zZskq8>7}j}9j_h_faO=imlj3vGvToKqZ=>o*dg(Mas!0_W`Ky4RQUDT+8UhJyw%ny zv?CG$SEA#D;9MDb`S;)PpU*bMzoao_2zfV02p|3`0BIK1Soz29wou8|Ek*- zl6tFK)%DTb1g(ZbE1UE<0|R?Iut{7(QSv|jaH`))?0K*oeXe#$p0#ne*T)uO25e`b zKu{nldXUA;FESulG+{LI@D$jtwHsU~X2E?Eyh*Hl+zlbHiFLo)J%NF#Cx|BES9!fI z7H%?~=QrJ9aEHbrMg4zQEI0VwwCKIx$M-rUrG|`XOpc*$-1`H~Vd^tZlX#~3RQNLH z0W$A4Djv5@rrjW3bPbLCB9m^r0GoMZj5rvyV;M<6Nds-L46$+b{xJJsAWiJk+>LYn zCZlN~63cl^?0EfH)Eky7(f^q^mDw?B);npM0{8hH2YAA|O$$rFV^WJ>hy7n-I);Rb z!sm{ln zE=Z_eX?MH3J0O!3WAg*BM(@`!8)!Y|__c>)0!T%KZ-VQ79|*t7|1^-s!3%??0VZoW zyn@1)GHjZM3Qb+%G4iLVz9IGZ#{`}Zk)jg_7Y)w=0cJSv3o&h*#|_=pNKTn}FAS2w zm9mM4neTDff0k0do)maVJ^&|+UMXr!#^Qlcp;9@a(eCBGaXpiw<1s{(-&2nd@^S>f z&BuM_jW9eKoBBe+IHfYO$bYY&yCF`&A2|)eAFVT+%KvYkA(f{h?SxNnSoM&ee{>t7 zOaVg?QNi&Hr6zGq!XT$yue9B*9?8H}dO|isB4rD>SLYZt*yVP$_!# z28aPK7?e{WvsNPn$zMVEv%&ix&X(|Fm>;t|09kUyVjhSk%|rn?UYB7`XR9l;7OVtD z$SY{{W=*j6O^RW1%oL=|mUae^zKUtOA6~D_3eTYkD(oXM) zPl=x~b{7eMVP5pTJzgmM$tlkHzVh%;keC%Tdc`tno0ugu5Int+f%VNUCx^#qi&jj5 zNODWtq&S35AkQzUEMO*IRJ9JDwt}un{Wq5E=N)v#_uY`_1hz{hE7C_@l`rQqpYBzL zTl90NG@Zz{DYdz2J+2;EwjiSF#qAL9iIUL7f39wQ})(=D(+A=m_t}YkPQ&K-3al^+A<9G3bU8R3+Ou zSfT{3bHhM2ClcJAh^G85=10wkn5f^Eby`iw6E*Pl3@wvn!wyHG&F^;#;a_jP#Mehx zqW>=LZBiyDYfFW*oticvZdwNA-m4`7m!F@7Abv zvXQ0#%@6O4xdT1qduNxuFf5)Ov))yQgN^G}JwN@3eUE>+qhh%>E6taH2RwfQ=nWW% zvZA5}gEk@j*6MUTub|_7qYRq?OYA1XBTF#Qr9ICYxeJvQ1(SsOzZgRVHNW@zwDxw2 zSvDbfmx=M1TVNG6K(zuSfn8*QZ3w2(-pW*i-O)5TT!y0v+I(@2UYP&I0T6hb1S*MM z72AjmyGReVUcV;&`t$;>m`(l8^l|+wHaF-}BLKxhu^A^O%x$zWmPC|yVWQTGKNN`H%tS20>sZxrDC~prbPhQWY!_*&sXZ( z?35}CM_MDH1Et!30S*|Ih6j|2*|w9cnb{<%meL7De~{lV~JSJ6F+`%qycpr`!qayHNzrMR4AL zA1ghQ%(xjGIuIV4`$?9x7u8&1o;0V?thAgd5yEBhKfiMVI5#z@b?tBi0>>&P6I@)gYUI3 z^5`~Xxcc-1)43rDejiEn``zk3Xgtf3wB!CKmB?3a6eeKQf#XAHQ}|M#sxnL^k*+u`fmkY8l}wP>;KlI1pJzlMvb*%P3M8W9%e7mYJ@VYzZvJUznM0AU zdY%bB+|21p%Vvpx{%`Wwgo7%rPvbC;*So@kZTq}#NRG(ue!0!CK5*9qaipBD*(AmBWia6 z>3>ekB}fGI>ro?s=2Hm-UR&wBq(cx1?9)av0nEB7fChP80Rd}ZVBCZZpJ+d_* z#nD?Y)jJpYPF8hl+96;w493S?X4tArP6Ws&ge}&ZD;dqo$?xcT#O)BXxu6ar^#YUP zg)`eq$FulbD7Se18A>C^;Mgeng}9dorN^LjwxOlSWYzsT0bezy_eiF#%qlSvwhYvN zDOmI>g8&v`WNtkdQ-=`$We7=%x^D;RhphvL`oj+bLnoDN{d&h@_Rmm%6oD~NI>vEY z{u_LKI%D4IkBV@kSpIL&G7={}-?UN-lK1Oe6EjXw@nil-qXgtH*IOF5&CW6zTnU>;r2GDZ;dQ#?3=fyUOb=q~>8b+|U zuA?G!?`pQbA(K?yIlSn(;gZV3klKwfHZ&>;{iX|7B@`28s$OdxJXb0r)2sbEfE00M zZu=2wH|-5sr1ox@LqvRML;w^GfNpm{82^!$c8FF7^2gLV8%idG-#a4xh05}JaHSl< zzA$iK`G9@aA-;R7GlLR52*3uAKQIqP{XNpt>b84f3!9iq*;R##$hTuF5ZohncC#QP zmc1HB24Y3}2yB>*_D$m`REgJrMg$yYz?uYKEIkQ~=Gl8lyjG(|OMXNUSEHc*@C565 z_(8%Vr%RE%KXY`bxLS;K_;TE`KI*6`y|=U}-7AW#@Mo1q6jT~%)b-ey_s3=YdpNKB z9biHSiF|wus&IH{O=Yp~k2y&z?4iyXY=u|*d9L9n^4lMR%6eX(NXzR##9)&kIi4RY zbn54iTVG{mo$WFGoirP4obQAR$YipUQOqeiM78WLkRUt8o^MCtnCGcFlLemzH=rB+ zuR74yCc%XWs5E@w%FA%^hJ6$OMckvFoBy2BH>*pMjj0OcBkK&k#sD^Phx*1I9i37>JCq%Al zp|@wp1t8NG0;y54&O*8Cykg%=q?0@fXgZt2(m*zB@>JQ zoT1p{U~In9T<3JMIE%NrN8H+Zk-ZTd0TzaYVh!-yaEcV_7%}DLIRqk7;!$~vP0Ic9 zu$hxf3Qo7|?+}Rl^}o7w2(}W9Td@H{3DhI4?pK-X1GLpdd`_r{=YhIbTbpA0A6z4; zQY`#}hx<4|4h6PVH23Ti0$U)VM=-K*0~-uJVT20&8Y1csK~N-9s?4XbT%2=5nXhE@ z6c|;;>EbGy^;QQ~Uk(4R8IfPE!?P(IFR>?*iVMZy@Jv!?dmPeolUaC=^|$jTv7T>b zk}C0BY;=}&`x~k-2^j<~yTn15RGa^$!sB!L1sI{yvzd2;7BZ`K(3MmENE(M8ZXW^} z2B?l(`AK0!E5w4~#`wEP;gI9sS*=n(%0YPA>XH=E4~pL(5n35V@p6=Jt_$b;D z(%^6XMHLl4JXDl*z$Qeg{Hj(Xr&wP+)DBDd(VJwS9)PQUn(oH;dSbrjoe=^bxIQv) zfpL+qL=iI1aG+Yba88;8>5<7WpYT zZ4RajFCUL8S~hDGaf1!aokLJr4|CH(Ezc|^OQSA zD<7k5iBLxm;o>Jw@bK#se+d8rE09w2NuJsWw1tPW)R?yH0E#U>z_;iPT7H-dAvY^UMj7tj(5I0z~?E$#DAP? zFek_1{=>)_KXVz#1}l5AnpPBE|HYIcVFHIa?kC^PwmBAz=gr}cTug9UYV01p-b?fr z8!ee|X;7-5XT2UtDL|1>{+xNBH~FYzhZ4;)2*E7h`np%%vQzWzQF>spZkxn9W2h<0 zLQ)Z020<@dZpEL{BO!|ta3@Ay@PZp4zPD}3Ct?_@B`q8=luc|lfSp-9(VQO zfDmIY7!ipHgp@QQ8sfjKYI@l}h=(A?0mjZN-lvn|D9-%Ija5GS?Mb0RpWavp9HkB0 z1?d@InLD1zM?nwOJ17bV)X?7ZMHK`0a>9FyA^UT7J2j_hqRy zn5#dvZ?Lon`$KMl?G{{bm%ZN9?L*RLekm14)7t{+z<4VE^XZjO&Up`4bl1SeXm-XK zabh*eXk^uNjIz)BNgLV{?O$>_WSvrSRFDfhA$ZNx`;k6ly zOz^7M{59`+4Ik;bQtid7av(IQM^k1q*6yaz2ca;6% zOWvcz7(D0SXCsrS6ACk*)tJ!QsbOTLDumAj^bkVR>SH02SlbKUnj=*=7C#cDXX9$v z(48vy1`}&R2l8$CtsZ)gztyWUTe&>*1b(t_f_tN83H`lc`vbO!@6*o6d%i7;bS{oC zqxLRfRv^}QJYPyb7<59ISA@^0B3n@lS}0F#YhvMiPrW?f?9@owz6p%qITZc`GoG_& z;*sNf>i%0n{NigC5*L_2DD|6`9ROK^A>52`+&>|wTX$|0lMz}1c(9?Ayi~Jc9O;y? z`#zd^yTGlGiRUq4c3?W))*aMg1i&Jkrs6DxR%Y{-h$$L7txreUoo;Dl~9Ds+DQIG zt;R$9n{kBkD7AJcz#KSwWl$-OtLT3e%a_|!+0=JZ+d6#+%4YvSJwz2-{39rz&mn%jA$s-H` zbR^DM7)ejyMw>qgj!$4;)a5KpWUy}E=ExLZ?gQ1b{f}va<~ql{2p%HSC3ElS83RiM#Y4i#o#!J$Neqjxfp@6W0!8 zj8xnQhC-waVGN9rJTRO%dk&MdmoeF^N%=G91(6 z>qH;Q?|hgT%nkNl2j64ZL8n5i%7bxVH#Kn=EqP>?2jeJK#_75x92v`2uDm1Y&D7x^ zi8cN6I=p$wqWM={VYa7!%@GTtbkv{-)MvIpBL#o>zSm?iAD<%rl7BRY-wH8 z=%~^5Yjm4dnEc^3PJ$c z;T{{t7AYrXpBEB?TFzHp30Ynkb1uv;8bUu_{+_KIXD;Y`0H>&Kdc|Zh z^|ilDqXZTkF+SlN??M_egwFO6icr`Gz#BzEVuSl1Dg=%hvlp0HALEg1+HvN1ixi{$ z|7HPnx7S$?r;Rz)y9^MYiw#ST&|z(lQr&D{EJ!a8mk52ivv*S#cz zux5QO~lXcOyix1u_6#JWe{zzPT2|`5u zyNsb$g(d2_6q?+L zn#|s6%|5Q$ff&ZV6ggXK4G1D&X$s6DMLc&#KzEK9D54MNsLhSLW8%lphA?^8IkdW zy0MH+Bo$hB>&k!jz;HVM{I;=GdJ2PpgXbfK2r6AUHDjD zPGhf@$?Wuv!ntC6TQY|Hmlyr{17r^c!CIP3@5Tn@=ZkcbLA$R-H{TK-o_&&XTE!Nv zz(GkB3zdVegM4J|&YxfgwVb}T?8p8zSCdqoP5j%og^w*bE95Gh@zoj&T#_g~&1jsF z#p=TcK@X`0jjHu5MvBhdiPT~yejWe`;}-X8_&&q~B7;0cXB|5Hd71O7Zq}wU`{1{x z6Bl2ro+gQ#{+eDr&KKTpDKyydb0r9{pRm>>yH|nfeSeU4ecPI%VG3S-Xn`-4p#$Ax z&HWEqT-V~C73rIY!-!no!-O?$*Az#pdEfAiHkV*c?Tc#Jh^;?JJ|`w(QkN>z=g`d3 z|7pUZ>KfgS@Z^8|w$tmNm~b-Ltm*$jELAHQCV|?`U~SjB8`3nthqZSX8PYCv#Et*n zcuU%h<)SfN_RrUst)rc6^-N)HZj-=L!jAq~9MAhnJu_zH9FL`W?(1S6+fn&e2uKvD zeiUyF6aD^#NzQ#djJ(GF=KZhev5r?X7{5!kn!Jp6`Z6+Q5`ll8+4JJKT(&a&U3gGK zbD%^Zo;Si1!#Z{Q{dUtc?{hzL9@sSS-d`}{l)XLBstfMWtLhq&yHyT}`IGET{(Sfm z-E=t%&O;)SUO1U^!E0HL=_TVcM0Z z+jPE_T$o9BkODz(u+J4KyUCa+8%qe}*LH?TeU8Lz6|3*&PwIbM3)w2NK0%>lH4~`k z^pXrujklCPgkrMp`!^`j55uP<$WDiBeT%)Z&v;ms3K?w5i&-K0WVFJkb$iNYW&DVC zfg@+CD3f&~`-l$Y*#hArH3X-Br0BZsF%5~4x4jE`X}26&Tsv6SO|q&rHY*!{HzMH} z?i{vbE_bb^&KhX2BvcXPVyV{*7>Lm?kC&;Gzf*0tY{Ob}XmoKi+U8_$mr;=^`-#Ai zu<`hP6FTj#SRdLQrksuK9R^&K60B@8a(VE_4hC5@oMbl9(b=(cM z^1khKgl$SZSRkdp(H%cTweDJfXf@-N*zIcTQ%@DGf^#7~TCB5N;R~D@!W6NqZ461ojXfce(tj zG8`}|tygGn)k6q0m6I9=;*9#_XF9_1wQ5=P1UktwjXJH;DgCFS7>cwCXl4;oT~rn= zHWETE0f}|?$5zdvw4a<2-6qgq)4V^dMBtcy&LhD?Dq1DH)ej&xF4Scm-$*-i$Vc%> zM}3!h{6O{}e8m)0mavlZW-J!)TK$TjkQ8Rh&}7gDZ+e8~6ZvHL5`xH9N1gVgH+}am zk+E9mK))m;YPk^!AI>^R?i`7*T{zh>?w!9NL4|OQ+S?>}J!Qg(Gw1SUGo~L_PT;H- zZt-kBHmySV9Wvsb$%ijoKdmMlfPAlCfUv9ixPFH04`2xbZkuBmC4j44ZgYZ;914Rx zrfq3$-@|ZR-N=p@AN0drA^`~81fZRz6@o%?ifLn&mtO4=7^&4J8C??N2W#N?S)rRR zgyQldV7B`6)qD4tNW5y?X073JGv$+!?aZ$@jE^#ID@sqtCys%tK379Wp8uHcTMclYG_;hr7cpaMZ(i!FI4=JT_ zpuXalJ7Llzg4c%MBx`$gVQz72w{Qm$K-TF&G=batLSkl=+|`fPj84{*wwc85nfhSX z;29AUoTI=M1Vvyr>DsveC5cYiK_BJ{TNnQk`^I7_KlnqVtwkPoUUtyo zT*cWvtB*a;qq3-$&RPM=Rw?rpvD2uLu@O0DIUt#Sy)P)R2AUE{VI@l2DM=YSlg-LE z)&=$o_pH^`KAVBi^uaq2D=l|}kI7RK^dHuC0W~Q$TzM{l(_vz|7`@!snsdyM+xz)> zlBiI}Z`QWG3~HWoFGLIsPuWJk5>XhFKdNBOYTv&MQGJs3c*}pL!T`RlY%rY+=qM%1 z2X7!Y4UKf~?s30;-3}XtHQKIr>z?-(uLQu~+BQJIBm!zaJ{a+>8-VSvw_c;_X(ymU z8Pi}m`O}cRP3Mt_+r&cyn zH~b&sb{t3$0shxL2;#oS_`3gmILIflBFNQVeG9fvbD)+T1`khP(S&N!tEL;fL0Y1! zd5DI@`ya^n=&ZCgf1F28UrI{%w~4u-@?(I{9pHgvM=yf#=%cfCWgZ_qHepGe%!K~W zN2nGBd952UT>XF#5C9h2y#Valil}BSpp+sCmh7ZeDjq~>rr0RhvD^vYb^bWCpbQ(@ zWx8J#QpK9vWlx*&^DTb1T9Jfktv!&#c6HQy`;P25z^sI3HpAA%@k}u&Nx|qb5G~?4t31b9i4g=MVMiZD_b^Q{AL7w*G8C zU9p>t4&L!^1W&{IbFX>T4AL9Fgrca=KR^2jRCj#dHM}0##N%Epp9GM;I&c)6qH938 zLPr}}@LGPM+Bw4O37{(Ou7fXqqPP_2hCUrjC{P|VRaXyB!kJ$_=FhS@O*!mr$!`D9 z3587?6lH(Lqyly7G7)Uy31u*Vn7Q!$&+?5v#|=xV4usxNdzC#Xr z54M$Ih|Vo~gUE{Ri8r^?D`ph$iPKl*z;&x;a=YDqN<}Tg8rzOzUh_7Rm z7^P^fvD^K(_*)8~fta)|?DC-wgoDAI^amtE!&$DY2WG6^nR`40cBs*q5I-uXqD*v zp!jm9)Bkd_)%cI)3hf@(ABvmW6(U=gXLX|%AJgNIvdjC@LNrDHC@rq(nW3W^w1IIc^)-egjbaksKBhs%WA|c<8!5#7ZJXVH=aH>drn!uMrl%HU{^OXtp9Jh$g!`cEF=1qd?7RpYJdawi4n zD*Hxo|vlLc9!g31(wVpI2-XEjT zna)n=^!L_70ztra&qW)=a*HLo)Jnyk#m^h8KMXk@Q8|H&Poc3+UsZ%v{!uRCXgTGz zUS3+V8BIZ@33()E{;jPiR4*x-srarl{oS5b8Q9!aR%M7f^>Y9AEr?WRfF=h;iIqh+iFE`v$b_g5b#yMa<>gtjfP5f;?R!IZhG+w?Q~RDTNcuOUn&ZT%YeRm@2re z7Y8M&vpxZIAtnVM1AzZxv~+^#sS6cKihK!h*WgFlsoS}diGWX zVXU;LStd!T?y=3jh+ul^FCDm;X#ReKJ%{yD~~0HCtNW6|?N#AE9{5&EUwQh$BAvNWZiSdK^@g?u-& zLTl+H!(!0;)!nQg{lR_~=0WLYak|}tQX4vhbmaRKr`Q;b_iqlj$8;n`>zwrIa&3N2 z-pSE$?AeCui`dE}UW-jsJkrV3s%qq&e79QC^#87LuXRMCWem6Vxzylu2i?pAZLN%weMP0y49&@>9|TGgMZ zZ0E`MN@QkAOq_ogRA0IfS;{G>Kd>i-g)pd>z^dCix>s3Xn;JWCug2(7b*q>#OM+p* z+i+0psoLiTRALu3S!}w_2Y=j*3+YQV8b%-FRH9!<`6c1t^d;E2eMZTvfgrIm8zL0U zgfyQjR~GcWW;iT9ob@A;t8C_ z!ilLbykuF7ziA-tLU4rSP`~+%9x`dx%d|fp($BS*)*Iwjzr8+<sXuf#gcM{#-?}n_PV3K3=yz7C%G+gz#wYl< z9q7oF!%zCcsP|4aAR03$af?yvMrsH#x`C)u7O>)wmHJn?GRy+m%u*xZSia^%(jS#%Z~T9*!v@o2 z9BfmE^gn13jIX?1G2SCS1mqESyDdlRs}WOAJ>f?+dz9%q!?|5MJRmEE8m)l9a5iJjD87} zR6D%zHS^Wop~-MEMS8c*Oyf?Sg#(Z9CAn=3RQwZg#T&Z~r%(J|S!{ z#%HD4Si0%7*>c%472%NIFzvgXS~?}rlBc9fN=OHKI@5PJOj7A|^s*%Vi859UlQ4fu*ZlLDM#syp zv&yqt$UeMh^~M#l+A@E{W$DJ#rin{p=oU*++bQ8Jb%x^eTcbtoCEY)u`x%uZB%M8l zReZB}LfvL@QaoG|Rn0@GTr`$OPvMMxYN(vA<9JdRT}3K0mFERD#*KS!FC`f2q}SpI zs8MYyLpKq-kfE6*rrqb5aKl(JF2&KhAPrTuos;OSsfXPJ6D>~z9@tE81(au$TSz&0nNbidRL= z`;Vq79*DH_Z-{A@bF${au;7_kra6IK7YIg0SE2l0ndV4cc;ME+H~H3k@!g}t{f8^; zL%M~->EDI~eNkM~fZEQq1Ql=xfez0luxTWdX0>&u7N+p&#T!pFZmn!bdu|o&e2giC zDgIF-%6H#dB0l});ZMqvNkSgxIz(JSG@r%XPs-!yX7LDnL^icNeONouzh&R91g=?g zbIo!a4TSXd%y90`_ZX!@x%b28M1@~JPPE;JA9uJ#)bc-Q;WZc~FlaTYn2ZvTuX(>^ z5iYl7XnC$Uk4tdbQIn92uYW9YX2H9Nmn*8uic6|JH>_uSA9@|o30p?VXl>?&$0w6& zN~4!1ap;p*%RLCGhvr3)MV~2GY*D@czZY^-MMb#Ck@>g!$t( zRux#m*z34w_Z#tA6Y{sfNwXloDbxZIArh+bl%xT<4BBtNAQ%qeeLLQ29h5Xa+dzi3 zSUN}S=mZjp*N-pCVIEtu+`R3>o9K_^-V=lmX{FDA0JIUBa&Y+k1ke!<=e#|e7Q5|e z@8jY;#|lIW17_szj!mzRN6u`{tqEbfjGt7p0moJR={#=&^Od3`z($WO`mW<@oc~k+ z39}tgdY3$58&@|g?U7=r(?fTVqE@~do&`MhZaDgnsLO5XchcxJWy)-ClgGjH?Rq^& z5>&@aAkf@JV-u;`r?lPqF=*5+&n_8}7eHU@KXMvc>S6Y`ws3223Ld(|FyXQKm{vyX zfr!XQ^Hd$@iU#N5$BZ@GxD-2<(5lN^s=wQ;s-(M~Q8+I;#XD)gRJ*Op;PF0t{w|`| zU#c}f!8Am*Ohq5A<8|B}1iDkqfD+YjNiKT2#Bp!j=JVmZmKW2SW?qD4wLpu99z(yQ zba_ryzg%fgZk(fQUbz2j#4h9ZpOOj(SU(nm9}iH-;_CTJOln9Ez=ySTu@~o)Z~JeP z*0VB+dx$$+J*cQV!dM0h-%0X+|4LImzZ>4|yV~k*(}$~R znMP?6TyBlUwoJO_hXb>+z6T2xQggq~{{mZxRRWEt#XjUWgte!URro_iA&f8(amAxX zsiCPHkGk5YzCigpE0geHprs(Pbg@&vBrOd6F-cp1+KOZD*{woVAqAz;4v&ay5e}JA zqxP%Bi;#p#5!LW0zICDE6_cW4M~~5$lOn~IF`wyeKZ>6oFEG^#s@3C}!$C7VoIQWL zFGMbT^ZeBFZa3DW0QF*k05E7(4l*Jijw}XFIx)X=CQml<^s+o@=(4-PB``%?N%LdF zIUA2Z#mov6+1)+O3CZMYbFYjS&QHm zj7)$8bBE@Z3ZYi1=l}hWy277^sf4aN%YB2BNOu-($P6~`z@LRhEm(^*0uB!UP#ipn zQ7(W(VNTxJ;+Gz`$7!E^UxcvAD*D71#POg&25^u$(Q5N4NmV_3x+w942{TeIGq#jS zEjMwDb?IMci-^8c@o5<6M^`JMbZhXKrX!MYeLs*SGGD5XItn~j?hic|F`8{y^U*x* zc=KtHhvVxz6;}CaP?9Dd;us&6Yu{dw8Ce39pY+z;O``(_7dy z@B*xQ8Y?YEC7}lrbI4i2buiA-=hyA9tAmjL2;rJUAiQc>KQc4IY~SdDEp6Wg}kG`8)D(b%>*(L49EZSN15`Ep@hYn{iw zANgkOnBQxE7orqfXH)?d=3rXqLrzM9yN^AY!kzSEcio?7nLrDKKMM3_&u-jo?7Gr; zR8+X^6W$(Hr_zUuge5>f0BonuZY03?v}&NK98_3jU3`Do&B+nMfqWdBS>LBdgKqSF z2)dWdM>lZpzPmg7bd~sB!4~6U-gNeu|7#I2kZQI;8QIcl#F6cO7suqp+#WPHB~S^f zEV;;IxCuzBj|S!u5A^n+=OZ2XiwsBLH#y-^6sqo^K!t~newp$D=o1j)vpeYY-l^}% z*duWO&E9K@{R3h}xlmB|xADxgwgLA`+@l1zanwdPY@3Eyu|iQ_MYG(b44F{O9jKW5 zdcvAug@C8`i<>`9k6~uFlWzC@baIjpq;kpl@0Ta%qv@KzO9xXiesA?8u-v%)o_;lH z%=Uciz6Xg{tDEJV0hH3&a+0fTO1FuPh=5OH2JbsfBeNU8sjNm`*f6e}>t5#S+{t0& zz7BXFhuVlp|8e^2j-7G+SBU66+4lrV4`3V5%~bbxz5%4yxq^Xp3iO)+^s9m~jYzuJ zztUq>|NbPlfWrK!oV9NpOMdx?Wj_@;$IUm!oliz1Xu~4RlYKn4tLk!!n)ociJnY%|lgxrsrCMs{7|}u}}!1L|vL>v`kCwVG*|)-ebZ5^2|bq z!HLI{dKZW^@2I(f6Mp!@A`oz(Z9r~^gP#?ygP+}cqAt-$v~0XLqOF>a1v{@HgADhbs!=-a3cL{1*xP4+51Xboiu8gJS*`*n!}@{0qxu zAMzkc5vw~}PsH#@Jaj+PTK?X%q{BQWk3Y|1$**(jH=8X8r(;Lw7cBu^38%jfuHO6s zEsXN(R+m@pu*y3T3tMTXm+O!iTI4slD8I)L50KdLw+!hQq1q&Jj$c1it4XQ29z5^ z%MatA%SZ9ZNSxf8{#eKVs59fY0vet1MS9rp`zsan=OJT zjp3#mzrgt4|Jc(2sJ$_!_HkOqs~LZipZ?LV7*&U^qQmtS>I%G{Z|7)UR)ZSk1=TwC|h{J@b_<8cKG zMbgswdNXE6$qfI~4+o;UuF5290+w}Sh;+#aIMmi42jzb9*CjQ4`t>&UTA|_M6-}P3 z|CBDP*@rPX6)NRsKT3YdD74UJz+}PI7jmx-X7fh+EiOzk&Q7@P;lAr4LEC{k%;g1G zx~zoJhhxu06OYv7oB&}_Rng5`(KKqiI=b*E96N`M6)CN{CO58VR<0E!o_P_!LMW)o z4ASay)Vn|zXe=`OsEknml3!Q3$yM^QY^=ezV?CWSf-7N8xSRWUlPlv#%;zZCoZ1S% zk7@)%WBN0$C?x91FOMB=I|fFx)G*3KCrXF2adWY$ipRX{~&P6%NCT8gMXyg4xap0SiL`-kwYBNBW61GS59*O8~ENi5r@XGTJ

    4-G%PuH^UenF!NiYk1P>G_H#-T!_zbo(^1^>w1LI3S)KNkMKIp#^b7;g4T=fAkk-I*nrm7nbAtv5+V2=^;wdZL--o$FN}K1QOeF(}V{xJ* zaR3e&`z`bEPKnTQVQM_K860#2$%Jvj9^p)*;=Hc>= z{`jp~1Nxyi)hd}1^`&8Z|w> z+HQPGO23;iAE43{;^|3!>8i69LEd?r2a?qeJTS7Oa-e{-lLe*o+9IJrB%;Mw4Kc z#P0f`8k^=|#IkR$C-vq+=F4O`Jk~X@O;DG3JN@!4nvEH4w(T5z4b4gADhe35y<;RH zU<2hJKewxa{Rb-D&fHq<7BOm&fDUHoZ{A{he22IYryHU#LrhNB{ciTI0Y)NcGU9|? zw2OueN&iT`7lcXlP-G)|vS1mD;plth82p|8M)SF+mTh`daa#XN`7b7i z_+m5jdE*OCKa{omSh$CC(V(&-Q$O;oAGPknQjVp_%5Ey`HemnqgMg zdnXe1mKH)<(IE$3n-GfyH=>yLpVE6za2~Im{bTV&^TYkA1fk2W2sel2ydf59HFMRAI7ep zFqg(4P9-}(Ivvth=mP#FNYf@M(M)g$5l6X%>ZS`2C;&CiRpU)^_1<38Yo=f-{GS~efG$hj|DJJi9Xl`eah@v-& z*%{i37@FAeUu0FcIuXjbd_b^9=`jN>a@)w^@bd0{v_aSGAOaq6a(1Y^+vwCxgG1ig zDF!X3>-f1v}BthUN&3hH{jVlP_U7n}G{u<4c)OS;2<7G59C}iL-1)epL&ea21Oo9D%;I1$HnpqmnO-#`f6Rg(MWGA#xj*rKEEl zPyf^I$k?o3y#NJc>b}u&IOJVzYlq~y^UQVJ>MPmzyB27m_Qm0Aw^>z}>HuBmT~amh8}OD{zP~1*p@`RpV1I8Mj$?Imd`2n% z+mq=a9#P50v|-_3kn&xH2LG~5vb9_>SRo0moU{d=uOdn9Yq@&zU(`&Q%0c%1`8ma8 z$f3{FjM&YP=%RzV{`f`*NTIWu<5Ui{T}=+# zlc4&Wyr%dy$$;2#0K*9ainlX<0LS*ocQdcpo(ix3NpEToK;E>sOgJ6uvvaB2!#p_< zqGW2T6bPStIAGKUrG>|ffEmCb`yZydCORTY#Wn)pO+&f`Cel-S2~zDl3BUsMWfbc0 z4bt;5fTQWmebd>M7m6zCzf%rvVhX=Qq6}GDl3=!&cf2P@LK*|IG#=xx7gV9SBOGg) zEXEOB4X1cN_*I`~Il>7)P5_I0ajY{LOnULqqReE1#;pyAZXp~`0R660^a}>_$X!p# zCcN>fw6Nzr1^ps!cyYhpFZNR+YD|$7jG;b8zqc;hFWtm8AW{+cVMB;I2e%nCbOYP# ziNm8s|kk8CF$$I;SjBZe0 zJBvn_TloyxC67CIcOvKzPSK(}E>ure+qf6ts~j_vbO2%=L0Lg<|jtbdlmFY2xcfA z7tZoadEL9-upaK*!DX0j?xribBkltE#sx2*j?Z|X37Dy3xb|{ba4tnd5eWpTgjIWl zAgsMsQwu9bkf8SXKpgsJ;K8m=D+t=>Q|?v-8uZrBp=8#fTNA;@nX^N@w|OuP`l)-j z=GXH1v0{g1Z~oLDtX@e~2Qn$qz#;EVlTSlhJSPeygQeYYIHW&U$UaqN-oEfU-o@&f z6#VeWxG|>0wM6tx4FAezN*>~wg5$YbfUz4taP7V8HFO!qy&;z-AZ_e~e>@Tx$ucDq zyDOwZwWF7_YWkK{&r<{S0@;7-48;(gp3yo(TH#azuYS4lIhKuAf`@(?I?gq&90(KO=Cd|Y5BkNeJ=@Q&=t7#?YK;nj3) z_>4P2bPw$5MyO}yH;vkW@KP17I}1;#&)%;mhjeQO@ZF~r zc)xRZKO!_L7?X^c6>-Y%bucUua+^SPlU>l3%VDXKa+53?%6v4JTVX$QSy_#(uC+sE zurlOO!M$-;YaQQc*Sonw>OO=M z&5PLiLhM+6p>ty9mM7n zxO&l1Q1fW+9UMj9fk!Ue7Ca7V906<(X8&r1i!6Y1MGsF{OQ{a}A!nNe{ci)WDBb)- zO=ea+OQw%;H&e#kjC#XS+Q9kRU#>&88gZ!4(f6tZzRvdEoK&AtNl?GgNFrz& z946km@47$fj}gfreis=U(e>t44nW$%Imn&r)FRNgrjdZD21dc!-D*_ZQX!y6bqO}Sh2Xf7k}+Z1!WgG1f#~1YC0mH>CUKM z>>fMbO7XTkik9Ch3&M>KD`^e1@Wz;=sQ4=k;CyuZ09}#=;bfVt-|&iF4ZYB>&h>r`MacsxxGXMXYPs$QZRx zY^t0b8c7FV@`KFA16ZaesIxKLfMLP`ct&Y{ouO0T(z&yf7-RmEivh2C;qAtDA)$Iv%Lx4Q8hZ&}I zz*&10{Z-p5_s!9UvJjdBJR5)Gy-}~C=1T|r>l6dXC+h=>f0U@? zkd=l5wsxj}dk3tyB=gNFO=cqS#L9`s>l=;=c>|y3=zR) zQ7Y{hgb;t1&k>(XTK|-dL{u+N0vy3wrLLq@eqc~E7lGd0oGMdc_d-sJ@Du~1(a{-W z^V*J$lG7KQP~RHp2M~R=%P!K*b^-GhBcW+uJi3vv@FOWiLjlPRpNef^e7NO4$82|8 zUS{YGA(z;2?*Gk2EkPZpP0?mK=;|)UG>V-|b&U5OLz3ORc}XQ{;(+){Xc)g-y$A@P zY2j`3Pf}o8GQ1hL%%8g!IQ>dSeITX3)NL07C=ge+E6?u3H`)t%51~LN?2&2mL7k@m zB`cio#KE{-Bu$4y?$_I8NsF^iN}k@s&IVLF>DPNBT@&m>@MqYakT}3P;sGJ3(Po*g z^{5~+tAfykJP%9PNL)m!Ct|DwI5A1Y4ztDgj@H@4``Ly|_;WXfp8hWRX5?U}0l+omHC&-OvI(kV_{QP84 zZ1yOml5@!el$@RsPdLJ=2>a)uE(cn}Fc@3mRW0`E^e%Ec_cnQs*tg@v9C}*@p*kBO z!-W2=(T@DaO|wQLEZ^t8CbbaiMNc9ht9ra##iw{~R8)LE|I%ou0~_#tlfoYJY`obX za#oHB6@w?(#32MR)5Wy*s2MUcSn2!os;LwcSM zNV+Zlzt9_A_VAC-Fc>#i$%Y-L{9}u;$lIU%1;zY)j@o`n`hFLDvbg)Q**9}j*JXGq zu_5I}z2NBMjcdZPRP+AhSwZ^hh0&BL+IxIu?9m_nQMV$-h|PK4o)3}B=?}+=Ta=p3 zf;K2WZt)u@8dC+9gYEL;q><3bn{87+8>sQDS(6TkJPGVdsaXuJ#48?SeV2e=Cx&{>X zFF-ixXCZjsw5>&mIi}RF6Ec`5>aGz;V}axLm}179VpQsbI2!hl{B06Whnl+zk{(fp zDf)eKX!MW0TcRG1o0pAm^{(e2NTLi%pUo-9Fb~MadCnIImdD?T5(<4c{fX@J-!4vl z^U;&FqY9#$iV0-`?|WI!cr~c4KGvXQR##bHGC?vC6Zl*rfa4z)g(0X#O`z{*aC>p1 z|LLSgC9#W1F^!1PTd~r^!e`b9-=>*MgO}@U>TZ@CcR3loK&RE38&=H< zj6VlWC0k*tzH~rLUVooS4w7nZSUDsrCUWpbXbfw?_Ta1m{8-WZ)TwL}(v61T*5_>>2%QxGbsUaK4T415vq$04Rz!Qu zzF~Xx=D~axgh5j+P(|C+($RbdRO_FgH3Z-!mdYnhMEI&F%$iP)hdZ4Le~?9IHpO)t z_(l_YH{{6vjd6<%!3n`EGCIJkhe7en@qW~3}F#lQ7-Q=?99)SDle9}<#~(K8osi_kuaH29e8hhonu zwKl(-`T*zuT*-;*wQ8n=ewaz=IkKA$aimw>HI4;ypUx%?#h_`7sJ$7&Arw!gxLTXe z)F`<+x#U3t6BWjX++J%Q{&zSSOq+^k>t5Kq zXY=+&_x3t+pu}U}+q`_TGw*b^akC(9LK8!7=!AnQN4G5lk9`;x_rRdSKzsB0*hQ{IXbxwhC+lJ zPl&!e>3<~u{m?3k*!2A?KVPG)+Ik&wx|e1gvszg_cNh(yspOpx1<}f8>+rw``8p2efF1sABjQ1{`!QOwl~(v&goZe+)ZT3kHzi2U$!88745F| zF)~77EZ?cI&rpzBmdNYDntI>g6U$BtIWs?%BJth>Amw*}A`7=gM65+_DK-Tm1ALPi zNF)i$3#hb7b*|X1cH(x*82&d*LDhI0arN8wVs(OPA>3#Zl>J!SK5}tRihnXEf>i~Z zi5y%hU3~&pFu^K&iDLFiQYnR0jEfFEk0DHEFoM^A6R`4M<4wV?_^Uhq(-7;G{b*(z zs^d$O>ue1`{n+QHtEF0dfXzk<(`~A4;Qk{B=b369N^t?T?{!0BpBctFzQg5{e%awa;+hDsYxn~z$m9W z8T?CCg%$U|wD;C2sY#6drWdoebuRnBP#mfIlYHs~fFqBr=PMsRig$;|G89^&-+qx2 z7FNZHm`p6&O4Bj;QafiLKGoO$IDmxo%QtAgw4W7@AMsPDhi{A~TdCXhQ|H9m;VKgs zqhqrY2brosleU63gANZzz{3B)l;t~mD(>;p%YN##Km9ZA`NQ$1%okll5pSbyeVD0` zMCxnjbN=i6Xmx1tbs0CkI>dvCrBxO%6_DE<&v>dKd#cKuoMGE0`&)%_S%#SAC-!8V z=cYHgvm~(mz>n)RA{qNa&w_Se@=u^|UYIE*N@7OPf05Z|h@g*XUE;`CEJ}QRi@gB` zqzLM5=edgi2WnWSXYQ_OM;nrH&3-yr>xgKu(UhW7laKpmJ!i5pJ5dMRz}0H4ssG+= zQNSTU!R+DBTTcJzv#J6Bypdl;0}>GlQ$~?AZ5-=madgQ#J24UaN4#G2{~ORkWI-x> z&O5F0dG&{iy(bndNE$@BMI|FmrkHB+nYF6&N+_f9R=bL~oGZxMK7+@XYh9G5&NKnI z8lql!V+5!n8+Fd8VDY!(=4+8aTAqeTHdb0<-R0n(F_5q3lksj)r$1~5Le`CELh$uD zc+6nZk!@!F7;~>w%0A)2T$k4ML0Ig+)3pYt^t!r?eVx}^k#xQ-BCT$x@QW{%hbE?A z^vLY3YRHSse>J|Z7jV@Au^_cd$jS4lkWlF@dm}Lapc2D49s4o=Ek}M+;u;Lm?~bn0 zG0X9>DXa5V7%Nz+m3P=kE2MhLsU(vy$Kmb5fml1y<0n?>&G#jQx@-mn9ayB&r}$18 zxQRv3zz}5&Wc^OAfvAuW*y5q+o`NB3fFM@+GR+a%;!4?T(zog~B58%Wxo`o(#+uO| zPt%`=ntieiM7%wwV4M5v*W`v|H3dnj&&P|(va_HXU!2%eZeBKn7bkPZt1A!Fn7XOOgY79G$v&ka7bJC$`w3YIa}xW zO-nt$9{46zN4V5kI;q$M(*7)`-^HGmHdVh|Z$9?FpxDqjSKWRVo~#*R^mf3JkG{_m zZo@~m2b#`0Dm4mo5Bw^e4G)xl&9D3YRQ(jKdUKh-#1C~!P6UIITz~y4$#oX060kVg zYkZdbqLRz;LlKdJ-cGB#H!3X}MLW0QYJOmhF)-G5(Hv<(BDv7cP1|`M{?lnrB)ZU( zC~P^uZ)iPtCp3y>I%S&Q5mI*?TY6pFHjfH()Xr^TX$>Y zhEq7_nl!1>f%MbHbZQSC)2VpH39LcBC(Yj)+B4ltZr6s)qIy2MPxea6zKJ8RQxYlB zZuIYIK{Z`6J~ZvPS&2AP5D(+L9vs$PrWCQ~z|2)&nKE$iKcRH19^#$11^nbu&F0}w zVZ$?X@C?C92G59{`%ze?Lk9sk{wefSg9F(ha1%r06+ac(b zJk2F!#d@HT=TWarv$XR>gBKT&TuJw(>In=dqs7ss29JO5UwrF@WpJo)w}R)ErDF;05w$|Q#d(dpMJ#4_oaHpQ+?U$g==}Is`BRHkG;uG zpH`ItE2~oOAl{Ujlb?MO)$j#}dEo9#YJBiQeRw~J zeRqejmTRoiE3!BxQWT!Wn`kIxTIEV(ICSp zhuYttx1$rkdZO?7U-xK!srcE0ip$bNhUIolgM3-bb_mJ5X-HUSthuW@x9&%^skFD% zVd!OdiraQw{Ct7`1_5XK+xCk~s@k4S#|c|^Yp05VgI0>7W-t;Ya1EL3nM#E8%rAGx z6u0wK#H(zfXhbJwI6o0oand4K$uPmrpvWf=cc1lW=>SaZre64U%4VB&lk9vc>GWSTqMUKNT{h4r$enKWd_t6g2v3 za`K7({ZwNV%)X`cS*N_@wlJ_Ha6xZgE3s?^euk8|AkQWQ(L2r00RbDs$XfvXEAowN zuADJ*?@*r~-P>ZcR)OwtQ1T+Yslc|czVpX2yu1Kc)HpY)X923J8K{3r}LF+N{ofbMlDRlea z7ABJS6MOmE6P!m#>o*TFw_`YT;5u#hP^(Nz-lC4k0&ofbE8x1OmK4OY&C-i%FzNTD}BrS5UyQl&f zD}2)pW-T26gJOB|;ys5LLPV_kD5!oVJziJJK5pyH<*ETTYk74yAk(^ERBSUmGXIjR z_FZijRXpiCS|I!6K06XX6FDZC@v3N7l;s)Q%LQcm;6X~?U!Bj(01$l_q54IL=vzW~ zCvLFMP`Yt3a%8Pm0*uihL=P z!#~A8#bDm#e2~Sk0+~C{yh?V1EBx0qc=8|%T>=NnAvfGIGI~n%ADFxWdnHd znx4N2!8y|>eopdVVTErlMXsz^M=QGpqgJ={8PR=!t3Os4h#-XxPiBaxPp~+Ch7l*U zf!ZB3&;y@ehuC_AW9y0y1{(JVi+WDzC1#tu?kgAc8Ooi6l>4+Nzz`At`pUxh{&Z0+ zaHngLP-83w5{(y`Ee}KOUF7Yl|1}i@R<6TWyWW7+v7q1cJqCjIc^y|hjAPl6 zL>=qFz0Tos-||k3F!p88Hw#|Rc@sJ#Vlm_V-gab-iIm=9FP#3Je>MW?2rq zwQW2`<@9H%eu5Y}YAJv_vi#l^lw3$3Uu0t;Wg$B}$jBthH&J40i^!CiyZ*AF|INg1 zASIyoojWllJz&!u{f;N|f?m{)13?4E7lE76=VvcF@ZqI5t^_gl`>M`ZYbntp$Sxsd z^OG^s=BQ?*!3&S#JB_gi;D}|hkQjmYXs_3`mIXQ3V?T>-t93lCkvOBs_YeIrmX>_O z)+qGCQk)ddJaw1SGpJ7Tw+E{PwYGdXeKaA4IIXK{r{*b!mlCb77EU*D)9Ieoo3j#~ z&0-qNPW*yx^GZj@)s`*DHuGHjVa*m$7Z}S zD6(fi5uLfZ1aPBLSlb{dA;y;RMi}f4AW>ZnM?hvtA^=Y?3$Oh z=Nz&d{R;`pw1=4oHI5L(X}OhJ%Z##{7z^B^IM?I?8MaG~g8u8)1&o3ELM6XfHFQu< zr1a9fioXkA4`62YWfnLH`1y=D96|Pu_P+_<+!_2YboiCD^k0gHF<_JFtC<3Pil=_x z@kea*7$~Lz^fhCNbqOkDOl`>|H$DcE-BXk*U2^+CT*6Y~(CAn`0wu}SVL76IpiqT~ zI8h`*6%O~w>zcGN4{RFo|A0`C*rJ)LJiGvu$JY9{8AGrv@JL)pn(v)o3NjnCGV45w zjvDJi%a%YEk%{VEt9Cj+A1YRd3_O?4+*}q`;G|8zoeRVfncD(3Zt847rpV09&70sS zHp}*G9lQklE5t%Xm@C0No2C^0`aAaAXZhOV02L(X{4DhYCYL!yKAC0$yN2%2W2aOfTPCL~Uw}XrSm|zXP0fiR-Kz;a1J9TK6X9tc1Nx_;k__!{NX3v-c;UW6Zb$mpz zJ$zeB#<8}EhKaP8L(0g}aFzMuV0o~?BhkChsdfG$)`4E}b5hyd*c~IBeX7#azx|8p zwtqgeXY#+=-YM>d9P4|-^XpV#oucKKST~3v~9M(0T>6Ac8Y$0fcz4^?qpE6A3 z3U2GYDS|MT+Fa7U{E6v;w4ChfWSYzM7=gYk;oZ;Mn z&hLGmpoFm)yZ#Z_zJJ;+Tw_hzF}n6w1Hi=cGY#3K9L_GrgGR*t@D#>f`mG!R`LrFy(D zsbde=xHex+mcw~3z!w$l8f{vKz%d_cLqKGrT{dZQ`pi(4G$#Q9t>wYAi$c6eHjGY~ zK-7DFeSei@Jt~M$UpJFGdci3{<`sFX*%m7QcZ-L0x=-uDtyTI+nvq(PXWrAlwLBYT z`%pxe0#~Pz#hGJ8yF&Lb+>!Z1d5-C<&mvpHkMmv8lWuMB0KL9r!Puc$(N?NReAe;r zsF8jl2+qM2X%apCB81!`pX&ZfZToB00!`w{&qescH%Ev03FXzaerk&rg}vRRcI!@o z`Ag0|rWsFKb1+x)3c)(;zyZP_l%feMzvpHKdi#&9kEC{UT|J?n!T2a*o~j}9pI^5> zQE#?*79shL&z@|s{$rbtPWre)|CkO-W-G(l$Kt+De4uLpuvfei+uGq6T$UxWD4hRF zi2r?~!?UmcFz~oK5J-=Dzw0<3#P_l^ECf)5Fu_e<2cetFVdZk;LZ6OtUOQD6R3_`2 z8YkcqKMs6-Dg};>bdt`^7oGjf|DCRoz7?pYAJw(GCv|^}BbsIk`f6<4R~71zpmpt( zGF&O~J{K!FR61p99^qi!I?OO@n(Y6HnNQ|}-Jv``Q2w}kkNGHZ(fX=6iOe0n3c874!yZUD}(e`wFV%?e}tKIjJ-+uZS*U5!8cCvt?NZ^A`Z_K&{ zW8lSd_5Qj-c1aE4wB-T;zeK#{G%_5dKW_B!J$T!sSZXPt&UGAz?LibLc8PpMe{_22 zNECh%ad$3$e~#hTNZM6=QGdR?t)>qptZmW06L5Fcw-omD3Q&6F7LE<(g{CJ2z5DoB6fLs(X4R)-u2j7eN*V(cs5PtOp0_jz5n>$= z_62jYNLrD;5wsoe9&d_Y|MKbcyc!+g#vU1Zs(*Uo?J|jZd-VNyjD}m0>Enxyk1#HR zUUwWuXGTjqTvlK)pUiT@Zl_qe+JQ9Z2!o(pLpK%G%i8{cen9@pGKeHG**>ScqT||| zI$35_Ur*?D+t<=LXq+)fL2GN1k^1$9hW(ZCm93%nC0V^{-G8!K9X?&oM3KU7HU<6Y zP4KAQ@?%H&gmZ|7X}xJx18)fC$l6U-tD>|<7^1OWnWS=VuZ7;j{rFU0{;k}jV)Hn0 zDC~Wu5oo*lRu;#~@^qTho%rx``>RjU8mC&`nmE<0aCGj~=5+aM$fX6Y0YTx$P}WVO zjsQfQK0l}7C5IWw#8$;vd=gNj^ik-IJYl8ivJx2$9Ia|+E`-2Lxl17V$ zkP&c^dA4drJ6C$m4v3`rww0LsYO=b*L^#NGx}Y!p=6H+7{A0e*;K&!#%6G}gZQ}BA zY!3wuCU?loJ7U{^*E7vzR`Y@(GGe}9cT8sI#>h6>IRr5=x&L?b8Q$?S5hyNHSVdAz0 z!TI3L%C8g6Ag_n?m2SZN;O1W!z3r^F{Yc+xB}#$jm!dPUMcxZ>cf3pVn4h!3_9DOJ zc&Nyb&E@03W&0aGt{kVIubIFO$;P#SF}b(CCuGt3wRxpaiqbW%h3RP)=X)YNBrzm8BxT>Noh`}8qD9l)AAGChJz(E3k7~;Plj@bn z^fV@Sn##vh68W6~Cbp50&NixHkv1BqW9#$z8aJ$$Hz_PC)E#k1UM~v-$lnT2_}A_r zf-4qnk8uaYLGP`Tz)(y>E}n;gVb7A|5E40PYkF|i530||5KeLHVeq~#C^?yDlYrUM z-t$A&9k;&-fgUr&i8gd1KvsFUgRgp!@4nhN$2lfA#_790J;wou**o~Wof>Rb zB+q9jYnMksJ;R4SIjZLBo)7?{#R^4&A9TdV8OBc0KX(=X*G2rbD%t+OtQM! zXJe!3RF>U#b^?d#2CCDeS2SIdX4z?%jn$mFbC>AS;*ji@1vuEH-Ge%}DpMfzW~_PJ zVcpJ-73!z9rU`DfVZs<4+LD=Z=7T2G8~L<%Z0>uq5VrPSrlzf&jknHnA1;66$09po z>LTLH4Ig$2qn38czeiJ7ub9wHi{CbfpNEe6f{`UAB{ATU-ktFnJ_et^xq9X zi3SQ1+*19U&X_iGS@iN?Z%#!ktnCXU;LBoh3^r2#FS6UjzFgD~y$~n`SA;>hlM0S1 z&!rq+zejoCKlMDNUyu#dD4o^6sCx5sjD|^JXl-vScbgAQ^m%DHQ97XYp_&5Xgf_-W zR6}yS!gjz7G2}4<=p@G+m+cB)zr8uv#Dkwc&s>B|V|tMj*f(b=c(D;0_mWR;La#kN zBIS~@p|x`=fJLB}@|B&8u^K@VpQxC_vaot;|i#BiW< z5qjaLc1cEA-u|`pS}03c$dzTM5?7th3!=dMXF3jIXlMPsNk8%TgIYf8CpJUbZOL!i zr6TZT05&sQPH2JaO$&Fr^cxTJ#`(=&r)GB>MPNm$OCUG*v&*_6f098LYUnAs@-3N& z8E$oi#=f&laAAxe``c)>F`V|JY1E-<4YRxB;?(SL2fWn?70$Cl_%SO57W87w>AJJ~ zn|`XQ)W5{ypT^?8sLV0)90{*WKz=ri{db4nFr6-PGXzeFsWvf$hO4kYa09R{JzjnW zxG`u^GoOqq>Wt`~nbujTs*r_gRy04a0Q{9}T^bn42Kb*Lo)1IvFi&$l+6(5xb#?5G z+Jf;6{qz~_7FqS`fANi4%N;vTqierq7I!F3>rLGo*|%wivsA6i`N`lfK-Bh1AiAkn zG#}k}CLYlW8UfZs_paOQ^G|m;-A{w4O8aPOL2kaKRkzuz6T@DwD`VcTj+1}nrdh_K zp+-q2_16AHx+&>NSW;&B80M8`tyntTyIuwSy^F5$f&^Z;;B2fTZn=G5>g)*e@TFE%q%>sGy}>!FNZf<|N_n)u9Y;p6Ko}-TMlUjX30gy>}gR zaODo^eJK)}E|01-K>3Pz&(JP11c3v7WZsm{wZYib&^}@DCH`@zwe2$ zvWmW<@_E*!0=@V%RKhpd$hRV**IWKK1uen8-v|2b6}rwal;crt$VL|7_Ho?(=3=ea zDU|!o+|0D@YmG&E)0SkDVvLj#?B_JcTX2Va=h+6JDb%ZBVTLC@_tn8)22XuBD;)@W ziEbLa>k2w;{&pF!d)cpFsh;*8ha8APp=kyu6o%N3ykaV9mdvkz0Y%OuU6sqfF7b(! zueLhKeO%*?fwPCTygcliyInV>bh5u(`LX3S>S-sqI}`C|vYAbP_j`a1F+P)CJ-^11 z>}7wxUxx>v@~prLU*Nl}<>E{Gp5N?LRoE~fV-k~R`NVRUZ0-)a*@Mq90EF8&nKs}d z?ylOIfkZ{VvTG-2b=jz`IHxBfZ8Bd^ywp9Gx=I)3_LC~Q7Xb1)D00zZw5VsI!~Q3r z>k7mV3gD+(&v4PmL}%duwO&_fvNt0z>F!HE2gWAtQg^Kr9tk`P*WExmG+u|wRsXvs z6!qNZlI-kTo*cQUwgxK@MWZ;m%)o_=jHzKdEt33G5T^pwSPU28<+oh*S5Bil0&kXA zE@yQgzbs?`K*}QJ4C28{YS~w!2J`) zesq&H&#AQAu3C{dDXIb#VkTUM35Vh?wBwSS+1dp%&qISqqf}l@`Lc-F%cfU7IMZSbXRh%R-bdo+Y&1WjzAEA70rf;sfMbblZv9VG)kQOJ zO=TPVsE=Ej?{R!#^jT~OnRO@PM)Ow++K$1GlYJ_DNy!+m|8kq8wRifBXufFZ*mX{z z?I##)YEM27X73-z<^V3N4pFPf(h;mDYPAO!gmgNqwQ4GWmmk>bgmBGQ?}xs`iwuqk zZ%kJCVWtwrw%K@p7@#~#&tBUuLB&e^ivTwU-yl3*F6tWCOq%Rg0e zhC@~E*Rxsqc)*MHkw*!22p0>FaK3kNoMJ7Zj_|Xi&aCsejed-`)Ti^sbG+mKXgUvf zHs806S5Z_^rPOYzO|YKvK`_KX?3YFDkQz4umotCiSl6RWX_#Ekjo`#X;J zPss7)dY=0_@ALee^;#EC@bAx^?`IpQ=_D98xi&d09ZS6P7s%i=iHpx2@pC0y!?ktW zllMhhXA1jn+TuMXIh)waxgXDCA1$|4&QxhAsd!1I9WPMn{)@ElT}}el)n?!FZzCG2 z0qDr5KPNCs3TD_(Q4Mexe+(7zF-LnW|2iuWPv z)3$vvNeidQr&eUA=3-|V{Iw&wS{l!h!Otzl?$O?y=5_nro|}U84;|p?aWJIULAA#3 z(b_9AJS8!a*`9lTx^e2O@ZGBFVImleHby8l*(o86P%*Of>$$3j1c; zWAsEGb;GfvI#3X#4m;B^oDrFd_Ie^NlyiirI?Ubcu)q@Qd1DA;cC8$P)ip z_1ASTqWq?GGb!`lWg=xr9{q|s4m>eFg3oCxGMm@0G?xZhf6E4>tO0%7NcPHI1gBP~ z_FG;44O__&IcjsOjAgP{k-!9kM{C%Pos#)yzx3nc?x2g50{k3H9L2$G!j!e@&+1Kc}?U9q_)z)YQ; zEN|Nf=jS^k{@mQ$iRC(@8@?O|bL0A-Ajc<5P~oK`{YsuJy<{_y6z;0#JaTH$`K$L2 za=!H&=Y2c00`vr@;~X;dSj<*x^ZuY9sE84pkGErS6f#2VSUJE z^kULPx=2QHX!+>+^BYT4)PsUi{>krg5F9P3Dgr{J!%Mq)-^)V2fYP^~ss?ngGX%eu z5wC{(ZBSe|E1$(@yJ#=CZ-bytOS8az3@(hG`{PP$^w;y5dpEz#&-8u+hlm^Qql%co zcyxu$t$CUVa_+F(b$0VA!l+*DaR!SLk=aykEa&psl1UI9VH_Jxiu5<%oW4S8tEDYZa=paXdZqDVa&)R`P{bSY3VqU&6GB}jXr9y= z9kyx7;FJ!zCyda6>|U?zL!B*yJc5CYgx}qYmMuF+e&61d|C(D=EK|~}G$SzbsBzxY zxJpC2$>kF;b{Z{sT{sSSpqx8>A!*=Yp-R=FMQ%)8k@oOw(Ou8z3LlC+)|RaWnO_sXgl*s1Mg=CEKSirsOL5lw<=ST zhleS9aB*=*Me(J*FqD=g{Ny*jI8HA;AF{i>^%-?a8j`R0$0?#6zag)yt2Z^4!DjVb zc=x<@!}$rJV~Xyczh!~MuW2PQ?eTn?InSO-2|x!JRtP5L0=0xHlK%4@$-G$U9GM?$ z8MdURGH$C`76l&<^`S=M6^${iZy%{m$^4aD_UO3u8&rKNPxdq}XNtA6GcvIl-}sDO zHT>ah3ODMfe7)1|R-$N!-N&^ogC@J%%e;Nqb(d&u`_Ygij0Ie+n=;K!(4)k$obvZD zCORcW8`4;E(^b8yKfd!*H|r!ut}Gdm>u6taQksFIXx4?rBfzkU>N!eqQf*_6HkOi3 zeop91NdYqFQ^k)XY*+Ko<5dS8mKf1|qg2vil<$ zF|d@9@kr=l&e5Q0#k{|cQ>Lc4W!7g>R-^%`C2rjE+{S6i(SL>LJt+4Ukr$*;}o9VI-=BdH z?)L#Im|;F7!0)i$8T3G+{ZRNjX{$HfXFE9>&|1Q20lG75V>XFlH+t=9$6^xuXj@j! zHCXCFZ#tGoy!Kv&H%{E*pXnXKbP**?38U()!iF(hL6}yr*~33vk#$ovC9-UHJR0q) z9lLUSEW0o{%QS`g-}82v!fEROFrKjk=H!$CGYM_TBr-srKGF2ayoL zUrz0r+Nze87_l1Qy*AUmHsY+-GL7s7x8Rub8dd1F>&Ej09^4gV!nzZ|A12=Jns`m= zsA0-T2Lz-|@_FfUtsj%?PVN)0G@Elxeam^JhLq7~5dCA~-J6NL@z?O}jb6GPjW5h6 zqES&?I_nSY=QgR`I~aS1qiNXju-<=rW~Lb-bd+K=#!;NvRWfqB^26;9g>{M|@U!xQ z^}wIH2L7b-eHu6EW*1GdTlS_IOH-R9Trz(x2|3=FxbCeNcu~et`UJsk4ji-ESmL6t ztIQj=x!SuNexXWt$a-2^cWEW`BwnlAos0ua1?k=3=*KE+ttN0>KkFz@8IOSbsDLZ$2b z5ixex()Hrb8{4GJ&VD`})X*C5Pc(7W4{a@w@Yy?cLyHkF>dx9oyIT9~z!PB=c8SUz z`IRFbcSmkp_{1vA#?l|FruLb){ zqW@Lg)j0_#OdPF163(@%{wp61&|^y-OEUwm6KUAmnflCB@B2i6h4z1bsQ@)r%<7XZ zUF4EP$dQT_a0UpRYjwFC$~I-T-rZghSd;AdT`6DX7`kkgTN0$`_3uyhgGmj$=g*I#QS=x$JgcOV;DqSefN8PN&~No0~hhCnCAL< zs9!}PUp(l|zUe1ooVqx!54f6La^g%BPS>yMv@^i`YuZZ920`@1b|wX&M`wf~!#cG- zC%smP5MR=jU1`uKh^jWIgbEXUtXA_E>V{6%$z6UX?r?bp2 zdNlOgX#*Q2*bYgGgYrLTL@MMS$Pz^tU~MR_`Tf);9+HB;bSz|LF{?SzP~Zn++n^_gBdC1ju+;uuU=xpyeXyY}KfUlmFq z%QO)^ku?#0=S{Ooa{EBOYsuzpcVrlai(C$>8SqZkyXusxbBoa$f&$|HL?1r>v;S}^ zHQqJZ^?hgOmm{6zN0lY$X(~k0i4&jR-T4uE8Ml#(Y_%9FKV6<-FlWEL^2sWXOEFv1 z3Nb3YT8C~g78Dcqh-_DZzv9;X!>?TX#&{65dlMKqSnr&)E^K22@TgU7tqIXCo;20Ebl!WZz)1vQjzhBuZ$=#^N`KEm!Y-{NI z*lf;m(@z3`sAz}>uPJicJ^OM7qjCQkr6R@A?fZ^#>Sa(si_^M^aay+})Ai5RFHHr; z1~enZ>YcwwL_HNitmW%_goYk^3?I*U#w)E4KaqNGP{7Q`MJtFjj-eDl{n3wMwMVs6 zTgy+L1btVWXDJCl`mNeOZsb^gL+$;3y=Vl-9Db5@(GZ=2Z3SF*L-9d0i$)h)Rac!- z>-yv_DrQsa%dCzV5a+8P?W(K8--!Ew)6H0n=ngI8aj(8C#}BV2^~AX%4+L4NDqDlt zqiV+xcSv7!;ERH02pINK{Yqg`mg(iCN<<+Cw?4$F>WClZBu-hqxG%l0@Rp5M!@GAA zmbXNmT)s8R%OqtES*Sbun#NyPl{5=+9RxkEjZ9B z7p$TW!F&?P8v8V0W!A9Aprh3~vI=ky?&%Mg?Zz8?i$A5a%IIA*)^_Ox?VK35_n4I# z(5~6PJMYYUHi&>Nzc#I`PN~FN4lYq^T#`*E#tSp0^10M500h#vcCipH0-(&4g50D|lr}U2 z>!2+Ijr@5$7{ZC*>>&ZMekY2iqM#xQ+^O*hZZMdgKQG}a&hjR2rklgnJ4cCoDycn7|2=wNUZ z{%vNqYnD_M68G-X#{0)A(WakwE%<5HfkoL6)Vmg6nOb7gRfcxOV@(d?xr{CNR@v{} z-GqX4E-~2J`jq$|k$X;ir8;b7`^RU!>*9pAp4Tr!S``}K7>sJu&$1`Y4R zTUp&Zg=9yRqT6GGg**-`p0uG5j!nwCXiQjhi$ldL(EU7Uk6u636&vN$fR5XX3s^gA zTRRw8AtI2PB%jMF+P%#UK3XAQTOe(~)y|+H#4AZz!jnkEbL4PV4!zq(woLnm^xc3K zM5OC#Q18c6{DaG+Z(n-2Vr(G54d~p{OOLu9F4Oz2Z+hQ;f_De{d6+$3TdlsOh+%)( zPOx*(!`p=!&bQzFsnd}$ImWFEM~8)!6rgk(gz_ zc&StAi+L;Sr92s178mkgYuiLlE~Ihj5VEbHSKb9PI)4`9j^3L*`O3;x^h3a+@(OjL zlxVN=?++0LOtA4K+7}Glh^^_t3m}gbZS`}M5Gqe8$mpde5}kn znqitxcsSR8TqT+8RYyDT*=V<5Eqlj|?fePd$0vZ8&L6`H$LeAQo5RilTz1?|#7%@t zgY@pI^&fjb8@x>O@ObwEibGw-Qx)<=#hSZc*LM6KgF)26`0R~7gr^%`_p7&!JEc@E zg0UBEg#-lN#nJ^1!-LMsUJ+|d-fO{!^WU(@z6t{YqbDx!UlG4uvQYb~|3UNN$45KV z3tw57ka0!x{^*9fL#cQUKZdKDy2CqyR68TbxjhR5ofzIuR1@OLU4yqv>ZA(_R>wH) zM*xtE^;I_yy{(^=31Bt#G0T<0Kdwiw5p9u1OtT%2PQ#B9e~vN9E2iqLoB8Op1L{{D z=tJt#ZC|kCQJu=SNDe65j_LFI4oKW|_)Qof9(8!=`?Pt8x`rKAQk*@1xh#)E2QOEj zY;p>DEo@%{0X zt{h5IbfME>GSO|uRj~FO>*u_|q}rr9o=+%$?~Ot){+eg}1l`j9?DXnsz`}}Ur|7!t zL{i>1saqeb@)agSwpkE`=V_7^n1`py*h3vNbJ3{@xa`+}65Q=(8!ofYh8v@%tCQG0 z$pRg_8TclO`{>|5ev3SrS}zI%C6dAqCnH8w2Do#2XvVjkcfwF1Uf#&5?_m-TV*sCaQf;-?YbQM4UFsxlT(F;|%Op*S_VNuY zFFvONr`VErjOyp~t_C)}G}~%W)0aB*Ce!^To23#N=`mRrCmj&?vsILSJn^q%vo^)i z=R;60KmL!eB-T&1DB>EUR4K71=OOyQse-Zx@PO>Pi(JK&%q(B|H1Rg*_;UjHnL^^N z)bzl+F032#))!+gWNY{-Hl3a!DZg$_3~t4Icvj}S#V7XH>T7`h1?>7sG;$D+@ASXF-#8^58o>9=?jCLSfcPU!a{(s}jv2#lS}BTdE&Y9x<@ zSQ2;SY%EUIrDmwWU@0Y&)CaP#XC4pO4m{XF*PlQ>N*RuyV_X8>uwC|ud~ZEn zYEmOYPwN_c+fsb*{0_eB;WcmZ(V&@9oaRf?z6#<7QDrYCJ{zmdzHv}_n?ODAXQ0fU zjRqV1wkz*J|9}T;NYwyQA2b}co0S#Q_o=U98_$dkjWw>`yO6*+MV8kzYMt^nA`3mR zpH9V7GN`^!8vREq7d8IVRHoa@<+@Q5GS2vbvq`Jjz6(B;^1>OYFIBiv5Bp!Dp@x-n z0NJA5MIE9$6vjH}-@vGS+{a%tD1%Awx=l}E^Iwz5VyCkWW?@vNoeYtx2qf8I&*~1l zqAi~*)IljvN;Be+@K3%Om?YeM-P?tV-g>e0@<_jw;PNYDA&-k9U7Mg7V#C*;0N$tF zyM5_*k;AMJSrn7N$Nc3fb&|%3+l4h;5K#^8@Kf zxPz9@4%4l{!U+u@Cxy*e8F5}{G==xc9~uD|DFK6KqyG*IjPLFi7ZK&&X)1N?ob6~c zE*3B0$r12iwLxXsBx7&`DaFt!UD+6QH=F-`ucC;x!1)F4fT8X5?XY>nfR4Jm`=+3W zXhVM!@R;d~fBdZ58ES7%=F7{Q_-H+vxd(i2I#0$-e*$}bpJ_B7w;r2s;>NFsG%GTA ze#)nd`Tm0K``Z_VuPZJu8=T|R@=I9SpeirW6L%W(UBmJN>|I09gzy+W%b6zNM(7$z z_9ZIyE9fobuK2H)QTmTmc`OeMXe2!5UI8nxcaKi{N^WIxJfWD;-YIH#~zxIT3a({1fl zPJCIM7(N;1YR2WW_B>ca4DCVl!G*-h*sk`kW>Zps1p|@q3x@ZbB9B-CnuJ@;>gBGp?i$jyMbEdj>xquFSC14bI@dwj?zM|Z2v{!d(pFs9&4-3jZBKgoNG%$OX3SSf;XVw>fWk%QP87V2VG3gM+<Tz{?DN&cO8jcTPIa#0Kw1k) z7m5%{MN2PUanJ{XUvwY?cun1R{x#n%wVI-<2U)+ktY<$Dmz&CQ+Z`T>#%X z9Rx!v;wp3e`!tvkIYx%r?o)yhr`@> zanUbX{6W&ljn|AKtDEd&+@=^PTaCfWtDRR@lxaUBPyP`G2cC!oPV-o(Z4|s>9GJ-b zi%(GH{H?tK2(o4dbggZO3^}6;mxz1G8!jrAy6<=5?@C6>sYB)=>cj6B-s9L94XTqt zJQi+isOJRishZA`W_m#?PjSmV4YyZdM<&LN%h#G#@$5_{kh|<`Ct2olWsKh&3emzU zdE{*FYDL0(mIA;`THPf>S5jox2j81y7Aw|&cW0*ep4twJ4|m%?t*btjxWQDoZ&jdQd_ef~6PLjL=6?hALVHZ!3tnDnyo~*SMOtfJ(a^z-7Rv8kLM*M} z=Z#GyHdOn`!s6tFqibJvJ-X=ZOUv7@tKiL)G+~5Bj0R`p<3&ZEtqqsv})tcCyA0$A~rjOvUU;I4R$j6BoqciPPgWTkKS%w=%u- z4sKU`fRey^yw=$LUwnYLhZhGG=1OD1;fpQ8af~(LSoY6E;f?+hX%nDB%7HE`2zn)vC8<@ed^BR@FJO&bMkAw@?XAYmmu&b zq5Z|&MV=YAq`J+`tt2A$8^!ur&CWHluG$C3IvO3_(w=ihDW}nY-68BGXDT@Ry%e74 zOr(FYnlNDkWdIwE<`4^)-S}zbDe)cV5dj%nCefqRZ10}brT(U+qfY#}f|cvcI@hKZ zFK*pJPE)m{vAf-I*euz7Xq~~zY->AQG|=?cVk2h3YE~+qNAK>kp~#_eg#%L<*L)*G z9kEr*7?`otkZZQgs6D>@yK+v!xBC^=vh}Y_ahnfY;e$f>jQg2zN5839=4}|d#$~#c zNjGjDIbb>4A=$pvYS!))=bNf!8Iv#aklrmwT>R3Su5ZvYOB5i zr6PM@)9ZDuP}aB=+709&X3#pb)#=Qu2}8J*GSPn>_au&#*6K^qPZ)Ma`A&sa+dOz1 zROvmpz)e-{C+=_|{EEt*|6HX@>STje`h>2Sh-jk8@ndCV>HWm|+BzLQ5G1ZTPrSEHif_1>n=-8@im%{3EMTO=A(-Fc9+8-aN+HeR=4RS1cFYlZW|1& zS6X6-B<)|lVwVL{6jEKGayN5>RzwZCZzvfcFeCWXC0jp6Vihyi7%k6OoDK0cl4pCs z6mU9uXrU?ii&Y>ufJ0mrr9KZ13u7_oNg`I{@W<+HZC52FBt5Ib2FmU(re{rWJ5^YA zU6SpZU-|6SXQ?S%N3(ul1r0*ka~3NULqE~plkBLhle z@TmdP6VY-f#)Hxxx9WRUS?&GCca8#AT zKHQ}jtAz<3$%+E^|IY$wMba5w>{IYz4+?X?$MNRa`zL9Ju#=~ePq zSvL_8^D(*ypGQG?7}2AyWxZ7i1PnLhnpb1&b^V2AaC5* zRaQ7*%_<6(v&PJ&>-jWLS4fz2{RW85E%%QD8; zxXIlX^%@`Gy$_a~B=l;a32~S)_YVJaZ`CMBEY~m=9$=#r#be(*=%%tbr6yU$d#8`q zkmxbM+^A#F>Dl6=`D9NYaTXf%e$ns@t0iZi9$VC<1??1NH&Vm>AnuqRmS(&NKvqwL zX|ElVxg_YR@Zn(D*BO1@kB&Cm_izCR5tEC|-y2b^(e^&<$pj<^OBLnowBZWW+y9tE zcCp!?z-50MqO(w@o0sr?#G>JG7FRcu*StAm1qr#>YYy{Fd}I`F>_#EF9!9Z zj(|A=8SJ_3Aou3DhW>9gub|^&rF)P8pJ2%+d+pCH!zW5KGd=@rY>F01Hb13cY5}>d z@JZ?b!Bt9DbB<#NckLz~$3MfbS7eoUG<%C_e=z;p1vLY6J0O14U+D-L(RdhMGDvT6 zWk>!X=&8ETF=_lNi~&=})!{Us3#W6g5}g0J(D^tff(hbXe;j6s_X&Bdj49Xz|Cd0h zTDm==KOxbPt%@a;V4uJQB^6MX<3P1jMEL?SFuL1(e>J$=qo zSV*QndiX*GQzq@{qv-O_Yrl9u`{McvU~GS}C~`{XPH^*rWI*necy`%o7f$&g6`0Q4 zl`tfS;RkeFtTDW_WbN`pL?8X+2g`R2$$NjV@rv@YVz!H##y{oXb?+yU zCe_$STe^l*@Li(zH-5aqN)njh2|Uz+mt4dGdu6=pC*g7lll`h5UlrGd`XhkH4Y~1f z7$>pFCA2A3`t**fh!}G;;$1iVHE5-^;(Mki%!bhPMnz;j>^<23{%0}b zeYtmgHYLtCM~8k_@ltDVMTn&bB6dHH2#7Sm=e|xa-t?u7&?3c@XQ?b5u51omryp6? zzwU}ZqqSSV=#tFwe=2GA!1}41P_y$R;|rtStFeulMhlzXkMsHr?p5@;!8t)W3KQhj zB7D2de=xIFT<~zDRF8(YH)|Gn{Y}DJ`f5ir3ss{a+XuU<*7k9hV8$Gqeo!S+< z)Rcdm;GIQIL~hoThwjxSx)I5L=RCg1uGn7q7k@WxFId*$^Gh?PX4dCfV2sCTd3dXe zwnokWMM_c_Hklr;d_?Rox&_hsGjy-wD^kq=NNoJk)JXUf7}Ah-p(@_Gp7+KxZu1r= z^_pam#(bBZjXFyepbF%yROL6;ZQx9*cCq{T{OT~^c}JV{nXus|Z94QP|1}BPuVF)y z-aCAj?J%ZgB5FdwbCYhaStXM->+Lz#J6JB+D@Ksg-=aWg`Qu~HRKA3DKZ)dFE4*+W z1%Mb|zIJelh_2}t-P?!*Tn`*NnnH!37LOIEWK*%Bwxzs5PrX!D&-Cl0O0X@=%)Yoc z`Mt(``!XsW#_eok?ED2@&dG(c1rm~M^{$4(JzebytGAh;qZ~4G0(*v`ZkLDGg+ z+o{o6E9AX(QxAW*oAUP7HmbJf&S5eIZy$9=fPKiM&m}dei(?RLiM3)*ggct8tv!09Vm2lHmn$#U4%!^aAZqf*1D~k zyBv43H&=hKS%Lp&tBb9|3tRmakUzOQc6 z(go|1T@@$>^^r6-uyo&-M3>J<6S$l`KhCCweZ1u`i@HiB8t ze^+~6h5xeaDCpqf1K&A0wP2>QlW9CH*6S(>zoSEpuH@m`;LdoqoqiZz7-BND^R&=w ztDLBy79O!)8SHgrf2Fn$?n_@)D$H_Ccg2a1Nlxgs6GYG%5)`*b#9?B``b}R%JtxK@8?JN^r zfYFo_?N^rES4Jf>ipQjk$F#8%(bNiiR;i%bE!_%N89j_RE$q##>(o7PH`=lTXx}Go zb&xXspI>pJI#@_3h%ExJy)r-ScWG#GY*X;(MSn{k<#*@6>!tJ&JN1C`1)HW#D-N@j zH-Ju2*TT@IEwC|==dO3Hz9$zJEl?GX10iTv85h+suHz54|IF-0Zb=D2FIz1B<)XN4 z=gIKs;z9eka1<($>O*Z<8j zq4~m<$VYd%Lbhk21<)8;!_{)o9LOCXaOkJ9@Wo4~QFnjHch*9!EEb*od@xyurpbS} z&TaG}o_o&3-$a7!zX`FheA%Ij_Y2eofFw=*Ogk~tr;PL3*&v;etm4P>dfevR)0M~1CVOf$e zLdURz_m#hYH}ziF_yY>?Yw2DTwrRw=6+lj`MPE{l~fG`j01G#0Lp0_@x> z2>TiEE9%WArzhqM7^HU ziajY66C%B@9I<*3#Zk;bm^!1X1B>w0d*!(sW)0dDxL+zY+l?k#!5 z{Y+OIIx~bNyBe( z`h6_=tdonj_PxjQ#x|Nj^&hqiCBSFIl5v_)x)z%mgPS^UjeTc zT`plwwMK?`uBvG0{jF14mkw<`l+)#I8e*#2 zzf3i~-N5I_J!=WN)NcubI}-e+g<5#?aF`;?RTnq)x~_EUzX&Bv%qAU?`A&q=y3aZ7 zVS}Dyb<@IZui2OYwgSTw?D`pdOpUZfraKx}_zLe9yP`GboJ%0NOj=6v>s&}|HF>e4 zkY3;<`zeXC2h@zJJc;}&@xzekTtM!KD2l z0}l~nqbjj|BhNfnw*{~C1;m6LspC>qt9M)xSl4W^@7#x&Y_I1spF{7+#H#zhkld_k z=DKpJQ~S%mdo%LkXS8p!aD9qxiT+7C-CR3$JredLStB14eT2lyQv+ex$FfIABBa}e zrExUXerwWS%lB5G;}qgNh;fA)m{00ZZ|BzT01VBH>GC&no&)eupN;%dw|Yyj{WxeH zm?PEopTsJ}*3V2>9LjjVUR#G<;)ttazSIsf(_{eJHa6TABD1-#f!oimU;&J}Y&yjA zL=;>MRjo4!&$kW>l}$RnQP6Wf`l>UJYk)rGBc?x17BU6uG30JP zHXk2bDDFCxsD|imWh3LHH9QnCSt4M~O`EvB$a@<$Gh}r2;mol38cP zR?57C))aF?P?gl#`ghKy5|77kAsB?~y{5J5QakYan~g}jqI=sHy6&17@-$&{kGfa< zs?5f0t_M#erm@Iu)7N>GzbB(-0%1 z1n^|^8kcYwUyDd-JKyJ?fR#@My?TE9`X$mPGfW!l=-3Bd0@+N7VxB7v-L! zjDH!zXp3$VPXzu%jFlkJpv|ah3pYA??`C%icG=fvZE=o_CMwxcj)#cnN zh8`p1cpjkka^6PLG3 zo!aM^;_mC>Uc8S$z1vI)Aw>%K==b;uq5`44=98f0`6YQITE z%wj}Rt&&sJt2aHGT9tNG-$Kd#wPk)7NYV>WB$>&|CRvKwbtU~;C^>brt8uFPGf$R z!G4?l%v$YvGh4k|^IPrS>}k)#2sFQFHEVTm#8cK(ScbEJMa$L}%d71!qSs7~bPUo) zdTpcKKHUKBN#gC({z@lA7Nl+ltWl*Anp!)&W9zj?ne!Wa-GAU~7v(2kh?B)ouijat z1$rhYlX+XGwJe1~w>)b-Fh8Id&do*3H;MJone9xSBIiy6pD+DqYsF0dWWZ4Oy;qlU z1@@hPo~7Z};rmFa!BKJGiY8V0bIi@Q^eR=&=x`lw1NIpAdH?)`kyq%=MoRT7@o$^9 z4=+7jq0UrXlzd~f`AHV2Juex=%OIjueoiv{y_~|$u1~Uo<`@?%F1$`tkne}I7R;Zz z+kc}lr1gwj4tB_)QkDopIL zi7KA^%L#MlFoh^{GGcAJVQFQ(uD4XKW0DukjeAWqy7gsd0_C70&vgX#Y;uQz74$ES z4yLtW+2{JE&cMpdXs#hSedAk>5{f4iSQ^@#U+OqVoqCFE^+Edkd-z)qqZ1-e-73xr zqR=FtMK1FtSeZ1Q!^n;t(^aY=QkwJLm@IejW_uP=dv?lc)walIt)B6TY7Td0d%U9~ z0}X?HSu9SW?p(tltjJk<^nR{*l!Ki4y!Tr}=7n}n7@qeye5 zmG4f8dm&IRUR1L4`9bBrfuc6_Mm)E<_Ay4g@uvZsT9#o7Z)ZNO0WZ(D;VRE(g2h~{ zYEgt^yJS6Nak~`D?I%n`Z-&jxyJ<&1lV-6+v6=4)dEzt$mkBm;RRL7TIMfa5>vijY z;>6%{pY1I?u4~w2S36Vxz*jY#;VP;7(ou+Z(XryM0n>P_9O`YRz3^IJL>V=~tAI`SRPCZfqL z@x>e!_nGjO54P2UlHGYU#I_S{eYrkyXzBdi(f2;j&nEZ(8_}<&k0ut5+uGT=fmBSe zqmOlUE%hz@a`7^$)F&(v-_zmuclXuMI%#gE3U~K*WK=XCmyuqJWgBODYGs*fLxEuu z+<(*=TB>KFsb0xHS)Uys?=xz8qp)QD4bWT6O=0lo^ExdHB@|Rwm-*};E3h|CV!PHa%wcOmGB<1CAlj68`d0A8^>>WEHvm^uyF;I-xB2W8CEg8Et7V9*m~2# zN8z|Pd$=uP_{HoGYoI*gXS29Z{X^MiAua^yqx27Ll_A)5fQwlvsjPyWGp$#NL%W?F zK)p$Rj$FODK(uGy6*y>D;*%|Vm7{Mmn>Y!^vP_;kCc;**UR2h*Bs!|3%|Nh<*pdBU z3SF7SOj)Ai{sW4fC-rV3)lP|V;pHbAg!VrP`)LcAf~(41!nL!KG<-1Rgq$RAKU^}o zbSExWPT+N7yVM$QQUE3&i~c$<>GrCkA|kc;pP40@0(ZkpJ<^nrerP7*sz$w{2KJ1T zsw4m6f3Kfrc}#s!Yq7?s#+|d7>hik|7dNjBExqkCTGfU4pU1=Erfrhs!0^uV=t|g~ zks7bj9K>x9{oD!;0;eVhg6%Q0Buy!Ndi}klXZjDhhblf5IzKdbdlziW4YW6mZw$?c znDb+Evl{a2Wjx2!Y_fJQr1b)Ac>h_uL z;4AwkQw|AW%Pujt>Vk!YvRE^#S0MoBR_vkgNlBpogwMmh2I^9`UJ*) zceYj3U>W-%l9&TU=*{zm!;r{_%AU|ZDS}7HU@E7p!joq#hx|AwpFr{vjlC@ow{glT zoawwKAbo0L9nxry!cjU7MY^PxH7~C1Ran zr?BS+l@{doEW_`Wpa{0hJBniH;)XG{>ipp@HAa%w^>FiVbufe^W|g)(|CrRD`#X5G zeFfH z(+}F4w@~SADM`p>7bDEp*$GUq(_1w$ITI8gvF4nR%&`7ms$SFvC#m=O{b2pDMnZDF zOSsUI< zcTLVE+Go{j4&|>wKX_wYslqIdV;p@qXg`{QQ&K(6Q?&}FpN1; zzP}i7r4GKkE-A%1#OvjocCu4k-4o2?0e#wj?pf>AKebqmG(pVXtwcab{&!mL)5OI# z>72*$C2%sp6m=UfKctCcDQzi|5XDD?YMUE@K94Y{oQn3*q*4SQV}qs7ekho8dLQxQ z&@J*8bmXcvUhVGOadgyMLfdQWJ``q^r_djr6`<#9D z+Iz2ct@i~h;QKrL-kVh?^Bm-8Q5R{yM+~IO^hfs9a;!f;-DMC3mmhlzxD?vahcaj( z)PBBexiC{Q_6Mj7KJTX_PC2U|zT)pp@_6=0o|A(7UXJP*WM~})tm~()y2$9YTvy(& zXu>agq+6_$Q;HHNhUU7FIloy`85ALp&qDyBVFnTc8C<5ufb(=)m;F2wR~!LUVkXu0 z4Ngu1oNXHLu#d&U<8A9@k5LonneV=%<-P1!cuH){%^%#_n9DE9qe)HKniAFP^gEwp z%lfy1vOZCK4wbkQx*BQ!v(Gog`yD9@{2~_?^Wu5-wv)inWi8hy{izpKlN+PZq4|0fR1L zLG}i2$)nT%&0LKa#5ah@`9qN;#tOq?CucaknEvvWn$UV=9~NTxZMj@1HkvzkPR$RA zmE7@qz{b6U;b2vO@ZDL)?DK3@Nc?AXBheHu8?OgCQTr9r+G7P5!OV()e;Qz_-w2ZM zs64McBgC~i*P4n{UB6>mt2PYs_E_9J%69lGpsG9V2uWl!@maFG-`#P#UFlpgP7H(3 zSM8-v>O9%--s*UujxF-DaBAyC7dTty#WX<^^IGzH;e?aHmvFV8b{9wgHI z&fWerQF|Va{bAMV&;98h@`VBkO(G~a7`f%V&ROeA@V!f&c|X3Ctd6diN&NBOGYQr! zahnZGIHyrp*L4wh%28EYCt_)>E%{EOoPJM^_D`%*n2Ps{w2uJgqwJ?f%6stREeChi zRj?EN-MFiPQFVJcSKyEFDWws0{OmZ5m+X9o2fJ@TFS>3Eo^S(hR=JY6{WtL)Tv9wv zzsu7vame(>ws@mbsnGend?b`uzFntlc+qrSEc}!&A4&+lX^o(Nd$^F4Xv@ z>is%u9daA_v&<>s0&ll{yp*(D_MOPM(y=#i*QUj!VNy3nh55FQy$ZzbjUdI$2&>8> zSJpt%Q$S1Ma%Em4p)QK7HZg#pG3HP@Oq$|Z@=IUb@6F~Uf3x0|adRm=e)JAR(lQ*H zpvX!afBz8}o1wk6GhBz5(WRqIy{lzH_8vYNXzck@q@w*mIxNk4j-c;F{reISzns*Y zNywL2SEYuD`^y!(5u498odHP9TWCc&3!uE3&CuZZmwMCVZ0g09XGjCZ>+p)UavN1$ z`_;L<>ph&a12-r;evsu(PUp9$+@O7A2WGKT^g=}%7yZ$dqZqM9+Qdi86tQfuYBTDV zXLD{-z@?ajLydff{_g`BwQV9J0+(~lKovB%C4ch`-q%sIHxJ9FR5=7Q3Ur@ zZy6o=6jyOOd|<@MDov)$(oUuq%P_mGxBRls0FZihq(+wdTkKQF z#LM62(QJnC#QZNP#lT>v<`Na7XMgij8I&Rwma!IETQwsxfoGhjfxJr;oAAnfF!IJV zwc`(9FowFEaUZwrWjm;$?qQ=LHn1hF`zxW; zY;UjqFsoRz#jwu`S4N|Cv^MI+c}BZ3y5v{Go^=;Yi=C`5YMjs2>M1ejn*)Bm1%H;k z-5z_pE@#3Te`-Ei51x#4JzXyBbd_Nqd!_*R-Va|$d{0SoI=>O~8BAP`ptX{nchlsB zkN9lO48(3S=m0Nda#H5Jt}J4cR)F7Q{>HQV>#jWDkI*}3G8X)~AlMD2D3jg>Z?aW2 z9yF>4mSmlX&YaPCOd|)rqC;3dy|St^WO`3;M*Fgvw@!~nIz>Kh=oZfwe#-IpAFXF+ z{rPb{gYiZu=tjq(e-uv}uzngWtI!swC|vdhpJGKjThQ-|4=fdy8Htd}jZ*0)djAv!};W3aK5tMw%Ha!lNlv&}&h(*|$e?*za;QZH)u z>>O1w60@+PakbXaUZlT_M5Pkj;RV+x`b%9pNSc zJGNP-ndW@|y&(9!Cj5Kk)6piYcmRM7UzqQ+WR{UQJ)hf$pPJm@@{p+9-ZC6F^J(@G z=0DEOXB4v|E3N$xrEb~U#`iO6LcZ<3`Oe6ZWugTVCZVT0# zS9yQAe8uv#;dvrR+}I5;x==D&%{aozOe)(r+NW`V)|Ox3C4?~kMQeK>Ni!2QJ8^&x zf&A>HpuISrGS!x{=2>u|t(l>71U?+x{mUJ0)nWB$^~zr)RD$A?hvYaojbz6aa2fCz z2w_xhzQ zYApAaGNq&SL%?PO*ZLdAEV$Ymc!F1ZfgZg9BYmO?y}~JaTmtPV|H`+(^3~SF+sAHf ziK%OG?;7m0n~lM1EQ0X+sVY=t?rcOywr6~wsBbdyloA)Of?jNyx{21`qVD_6PrN|ekEd;_VPN4u=D7KgSp^mUhdgvw6hR-H- zkn5uxMubxjF7aRH{nS;Vr=C)$N3)MNe$L{`&Nl>GYxn)Gk7nGmiO&N#|8GXnGCwI&LGihfjEZwR`__>$~p z14|7g5kgiQM&HF#JtvM%s*;l>i4@&H`a|O0#TXk;zCF;-(Q_}e3 zxMN=C$!KqsQop=#&Xvbcq$?HK$(p#0OT3N-e$&-55HF``S!?+N#p9C^qIM`Bp(tb6 zh1w>()9Z>Q0Ph z2%)%ohTIx<2%WrpS){og`c-Lc<=qz%xQ}ybHw8dgWOV@Lqk!|2G`2wN7tK0iWOs43 zU+8UK(R?J*~{(~K`sFk}t%L}!#Ik{o-zw_=h%4)QDt7debkr$@i zaBd!(?{-K#9Qx9@ifih#jSo@~$jGnX(i1f&$1O>bkmJ0{da%NfR=$D5G9sa5?)5#F zmtb%J-CD*bGrn4xOGTSRzY#iP`T>vE^<)TRTPOHIswSML7u6H{0*n^HNcu7( zWUmd3z0E}Tt^G}%E@QoSna8*2$3q5Ag!nT|TL(MCYL#AZ_BErvCIOfT{aYds9@f0p z275u9fOHWl77pj`iBlAnA##Lb?p3m9B`Z>%;i%%Mt-zS15SKVP^7=;3j1SJCuEEX2 z>;d4j>yYr_Gykx*T9lF3%hJ3SO87S=^ajg~FYKOvM+Hmdp8lt1j8Lhi9(6QAtvs6N zJ}m9K87fHeFsmuA_&v=2C{6F2irs|o{DSCJoXqyB!XGDL_8|(LsCmD%RN?lXtFo1y?t5*TM{MGlC_B8;N zvp${%5k-xi*D-VF7#fIlL~C$stUKy{osm4^{F$rS^T4^OeCFIDZe%ve#dgdW<`S{P z@?5zL6xvAS|DOn*od%78{(!2=WDp;JQI+YqROY7aadu#B+USw%8bCOoCKGk@{rctt zLbO$6xVjZSp#1ZGNB5aJSA|+_{}SvdVJ+9WDF7emZs&Ni)(lqJp1@Bd+7hY?yeZ50 z{@?xI9nBsupS=1Xl}>B$o#Tm=c)6|qYymi~pE~9CVAy}uEw(;egRqS2DM_x(o#}of zQTVAi5+Jo<#))THKX-~Xz-FY^G8$WSgsY&XZ|L23b4ntFSQ0vmneAsjaEN|5wvudn z!!rCKfp>;$i>jWOIa3uA!tPKaKHGR7>kqMz!{gZ_5uoc zKO{iXkK&8l5z3s?k~Ga8>;cQgfqB?C_w2i^$wYNr%}#Pqw2XxnbsCXY`Hb5^ z|1;k1XPaUU9Ti` zsPp88*PYzQLYaSB8?Vk;nv#zhdFwnH2vmQbXwfm|a4!1pcBX51`rGD2+mesbNceTt zuHG}oGh){h&f5~FB8$DzQ~eJbtolSOsqTAd?(hxRcOY>n_j5DFYn{~Cx3r+4s6*n3 zLt5`+*HzY;y+3@-**M2tUQ2fX-BT3e!qRqeg=jpYay_4Zk!*#&a%@mA9)=vpec%7@*IQNoWfu)5?fhxST zyE(a$tb$CEtL}XtFR9H;p@D$dT`J~I92|etl(<-*h&=9M?sQi?Ki{qaM@A1v?$$8-fs=fQ7bH}lr3ym7;=CHrOR^R{ zNcd7~)N-I!W!#~C%CP$A`LOmN-*sSoX5>}m1Sup{RngxH){6W3GAZ;oLH^j4p_Jwr znFgj;klGv|LH!R(#Gm`eKNFE!4?wHH+oj&RNw@4PJF5W1l^lchO88)zeeJx(-if$H z>>XXB<0BUAJ39J%z157j1ET6%TJ%VdgB*s=!VP&Mz2@zr{d$|(9st+2R*9gv$}-w& z-OM*}=!PzXR<<3ZSoSw>lBaXC{|tsnFs6Q5Rl?}G-UA3{Fs{P^@b5C$pa8UhUSgAR zcJnE;@@?kWewiDQ`M)~(Qu$I(S}3cz!H>xaED7p;%F_NXM@(38Go=%rz@8=#ANnA8t+Ry!M4Eiy*gib;mDAU%b#o9DBeKQRNBlF zFoG3knKYax!D_72J!**NnI7S=+sKP@o7j+XFHsyJ64u0Db&laS;(Z>lO&WqaGQz49 zTNNL2#kD3*Rv$%Rjvs9<+>H|s<>0?yejzN(=IhzIvC?tSMK<}8&}yQ#fB?N`;=_c` z4(%{@07+DDh7Yo(=;ODKzpg__z>3K-QLeqwx%-QRVKw;ywCoc788`FGx3h$=1?wH` zXpZ+h?@h}5MOS6dj}1xMKsPI%k~tn;KRqs8jJAka(72{3H82Xp1SC>6D_+G8`F|ZZ zgPt_4(^D8zC8FqSs)x8YRNN+)*xkRs_?-zgBwf(`WUL!7k4*ux@Oj@J@~&?fRBaN% zY1n*zjo=kr^LE`Av*H2+fH)iHSCMI1T$!BSy@Wy-*}3hf>Xe-(Or{0g?XleT8%m<` z^4);Uoxze=j=$XJAhoCUBQHo-{Fq*%_9^Z2FdE0eQrL5lt>4e_b+2yUn?t)Tf7JJk ztyVPyaw4ZWQE$C7yL-u_ph3tCI6+SH9|Dhb<=|$B#&WAR{Xy_G=OoGB}7?!Tf0lNhA>_pI^27USDD3^VBsi%ZiE{pgW z0v$?2A*UDK_u&eiWKoMkxu$>7Hn5b^D6SIv`sMwvt4*JnXrdxI*rx!+$!^f1xUs+K z-KI3HgmG(65aKOo;|p==ptBav^tPtIPt|r^k3L zDb2G&;6(B6PJBw`c^2)@WtC@uMf!=70YN^&)2J+(cCxvawv|LkzQo<@De)ibDBXqZ z>gLh2uacI0ZSTy??h|taxV|i_mU51>`%fUiZ5Y=aL<}z8atFK=#M0c^*suNm6aa`- zW3ms7Nil!D#I#@l@1q_NSc3&bQR%?3s%R1~I&ppMH^5DDIpziDF|e7q2%XNt?V}{5 zgDM7%tXPfeOEv0Snjo5-XwRBa*oZ&V$iOTA{OJOD1FkguNmQo}-rgsVRicMGO_8JF zDuOFNKn~h*HnvdgbHjMz$x_3h@7;mJ!b74Ko&|oL?v|DaML-K+;_6EsS=T0*$avxu zqfoxa2IXch5;EU#ses-67rm?;+L~|_spN4Ubwl_*6u%&N7Z4}@L8AU4W&G0*PSaz4 z7Yh_WLTB-S0}e^GnaNiY2ZO}EKS@)b8m_0^S&k#d)lBCk>hg7Tx1Ro5TPlS6{0?8z z!^n~dXnouy4yvBBn}@jAIaY{zG}8AD=$48Z*6npU;rnY+>?LL;AcH=x)0(htaD+}c zrnK~wMluS%^|7G^FS$&pOw%*~vbZ^v&-o!-PRvsQy69rUJ(5O8N+aY@?_;;H{CwUY zEAJ|9Z@(v~{kI$v9ZXAM@2 zw^pt+Y65#%%Ir6D2;5H{1KMtD=;RY114~yui8<|86Jn-KOglV=nVds;JN1~(ll8>a zN#qj2+}_Se&OI9bH86MEp9uFHGq&>BtWd(~RM39K#L0e@0tS5u^IeQ4s_2!G^KSkf zxXZ>_RS!@aIhGVWt^ca+;Bd()WWcZ|Fiz<3hDW|HWe~PuPVWuVGX-21#nU=%hZ6+yV=R04%)Y$8B`DVecqV=ZEvodr(^|)b<#<BTBZ#4af_1>yQGtXLQH4gmZc9ktlN+QOS$*h~kN84FP zL(#SZ9jj;Qio?MI+koB4!Nwu4wkNYI$BaGD+iQCNnE1Qf#S_H`j00(&%Y1j9;4GTvFiweq{tTaSt-4^=-~9rGcxo1e~sn!!HrjZwzgb-BBe zOI%wiax~Q4#?fR#WkX~`id9kFWofJX7ghZ~*zNF zIF{jb^@#bJM*BQsvAu60j}|nzK)p|!jSa3(Tmx5{CHh

    #){T#i@r&j(ni-e~Dm( z$TgYfEL4E3_=Fs5gD$xd^qqwz@nF`TaP4JH&o+8tZ5!^hw{=?>EMjVXGBTgB~ zI$^e-RQW0BuI;(DBhb_KZd1pgNpnH>VePS^`LkG=u;&nw(~?D-gq{flhml3fjy!_! z1O6yi`wTV_DSjnHE=*?707yZ>BV`{nXmXBDMfo4C*ji}yxuVG(C%&QAJqgyNq~vP$ zdhN8J-+l6FmVG@<3RlI~lk}>+6xxbv_yh4$#_I?Bj4pDpW~Qj6?F)Mh2Bv{Lnf=$W z8Z)bM;RgTP!F;|pC%Aj)B$bZ=75WVd=U+wXzJZW;hJ(4jCUYdkhM5ctWmVo|04VN^ zt~)shF*q4WvF1*Hu~6JGs9IaeT-A6h{+L7!3Kw_j3rcam@@0qb(`D3|WP_SD|=IUnK50LMH zrJrlBwn0!KvM+FYImAxocPnPhM0B#PaBtuyTQ_ViI(du7HkT^QU;XzG~Tf#!z! zxRqOT`4U=!$zm+mz)a(r--IzdZE;ccJb^lXXz;mNx8z&@4*#A;ESC!rF zlw0ABLDFXQk+!u-nRrn?@Pk*$TEOwHVv_y-R9;SP-cay0vkT_E3^HT|15Azs`)&HC zwP=z_e$yRzR&F(4fHd{x zF{AqAGqcf)5rY*i>NVw;|AvzyOO zTf!=82Mbq`%J_en%B3^Gmi3dsnR@yxd)bHi`K5q-*XNg(gVUt9m0c)VHFh~C*Mmbb zq?P!cwFrtOrR&%CJVbG%CPG}Dmt|F*)W$tf(hKvBI6nOPvc@7W%e@2Ti{|W+@Bh#A zc8{^Q<0}kAo9o(XLpBs7}S=E*-LHDgUS?)~@`(i_qtk>#aNNA70vm>%W z&HaKMxAhL`UREp||Ca>-&m_+R(yt9L@zA17gtFsT9Os*A5;BK|m|T42qvU18pcSJq z)C-Ru>>gK8(45dQ_DR{(iKJ^(?pRev;DoMqoOc z=CF^1(qV;5(csyp)u0uk<8bGEsV$3Ar$S2O*Q&kQsWl^8-zojw`*q)kYa3f@oAWMQ zIKFuX>Ye$)*P)w;G45SX>LF<&WF#=v{UU;gUmBdZ;@vh{Sui3<=kRxRy>^ZCta?}6 zx6SQrB7S|{rn7mM;^=&&%Hd{twxVG+T+YLKs22kW4rO<8+r*!c$#<}x=;$Us#{~8) z8&Mb@z);&-1CFyJ@m0ngCcn6*tv&QDsB% zFNGh0jJYrYP*1&lW}|@D_4p``#1S3P+j#@uBHl%sS$d50?`JWn$Vk8Wo_>JERx)~c z-Nm<`Put>{W)h(~eaz>{-l+87%D*RM?k73YJ1;$|c2Dm`kaN>{%gU> zrpr1@$;Lg;6v&u3cDWt?y3IZ!pnEcDc`L8h!+?h0Kinv2)x)&uZH|hP;*S??;LGVRrPWY^T0H_2?sd%@&U-sV(pEWOFT`o&|+-LCm&j@ z7=jcqVi{l&cjE)_r|?~(AO8Z3y}^M;otM+Q7505u#QJA6r`6Ck^zxMDyjh89m@NoC zx4uMummOtUMO$Pb88h+*Cwjme5K6l{7ve{B`jrm|p)i$U= zSX{z6=ZJZI71E@Nf3z<#&D)hWbGg;H3@{Rdk3xoZjBI8RhFbQB7XO#~Z)jrZxBJQ3 z3@=jP0f2y|*R7)oTJkqjO?b)XV|a#ThJuxTq%4vNZdr`aZ@M#W628A#GQSoQGch;b zbgb?_7oGG)7G{IUktYjcM3f}P2W{Bi{LjH0D_0G z2Xc~FJ^BMc=FDiH8B;UGc)mUlvj*7}1q$v~i*;GHv~`;I(>+Fsy)C1#dL*O9XPs>M z&q2sLov`bfL`?HkGF{*(!nu((wn8|XlH7Cdyn8g1{X_U_9LFD1+4_yI)x0eCA9%=L zg|j&iM1M=55#l-)G^{a8;fA_ktr8hPop9j@-ZUC}9ry)yZ%3kLfqa&+Ndf%`y~^@e zWz?xB>`rrR)WL)8s2*LM>|-&3(%!}2%u~(QvBEw*$(n60B6G<84O^wr)Y1cLx%z3q zJGj8WKM6}^>5!FDEIhR3g}MU#BKUk1^z@c?OyV{q%Kn?Ab$XC7DXhom2LGEru)w{{rYf2tUWu3cNdREUrAP+vomiuTM86BmxN8?{MeAKhHe7A6zi5%l9;LSX^CiG#2t-3Yr*k z_}vwJc}A*sw7GjCrr&fSq^}&5TOsT^bOZh4toD|O&b)uv92K`yb}xTn@nA5P`xHIk z?ZcohB|76%0ex${R7>WO!trjdw8+Ah9YVw`w&Hd#xHpP8A}zNC72V-mmjxZKtE@9n zaruM_2R8EWyf#EmZed?D0sK3XGEs*9RSv9%mOKZYy)1?>Dg-BKdsvocn!{m$hsYaLS~C zj~$*74iwuO++bZP>^^8hO3Cat#wyYtvAQLX!J$|C7j9IGtdU{MEq2?SIHh+j@1yzgD+L=S zv?Uy{L3O|bQ-JsU5JBS2&*Ncn@9I&bt9s<_XW8C#xAzje)=q)@ma;}1Q)|(q?5D=H z$_H;|Gv!z4ABdi9$9JxO*G`tGA?r+g8ER=prs%AzmJH$-&r2%Ld4Ezl4I(D#r(2jm z#0ev*e7)qWaZjP^I8e#xu?F{_-rs65E(GxNb>M2hM!D#qR`pJe3+vwE?ZI?Mv8@Y6 zl3)*;y>oP?#V^Q=Pz@yJAy@tMXVZRr{X<*VgV?%}AK!3^gIn|OCmAD@F&O`c>VXPi zN-VES2xP{iCqgmg7}$f2G5k3*v6nS4Eo;F@E2=%cYSAN!QIkeq|6RjrHzfM!ncu zX$hgxAfD8gP6tP^V`rFkd|m{@1~FRfC3&Fu#=vrgUetu=OpxIb!QU|e9XEsBDmRV! z(Qcq_A5UyUj{23M`eA@xmKJA*IL{Pl^kjpk-1 ztKd*4n!+YQQa$oj4W1ty$CoC!g3) zA+tivSw6mGJDpYLN;hJ+*tN1dt0%)3S&~?n6=_z4o_J5TNJ|;68Sgdkf7E<5 z4zYve)1JJ^vO~+Qch(H3@b39{y%j-8HF{0NfUjV+%C^frL&nBRmB;0untPy;3$`-A zBZvetL{sSyrGs(yj7WH(Z<@makqQeNbgYB0`|mm1|D8Jw;`ER)|{`KIjt*w#Xx zu!{}e%kVTs@3Z}#aXUGL{%y%cDKQnHv%QL$jGU1)t&YBg*+m2y9YTt>KsmEtK|;%6 z=})`|)w`Lm2?eiwri9(9f?PooQF8s9vO_@NW?)c5{tk4QCl{;>TiW6M0ctZskQ?97j`D!C7oe)*#9t58=`O)?}*kN@^MD zNhf)^pl?SPrwV*c6(yuP-I?G}kNY3=ELnj?i7}*8;N-n%)Ifn>=8iCnm{oyArcS9T zZ+?mR47;1xMwK^$A`9Xq6D9q5TXAfjooE-rb0q_j&Sl*3dJESK zX^C5ZKy14FD6!n>`1{7&R_#3ixu-C$lszz@TD69I5sh2|bY2Pz<-d^wV(rt+esghF zeNkbq;!_ER@XF~`Zes3o?P5dm{+g@!L+)W)iOU&Iom3M2j}u;n&n+Yz3i8BnOo@huZ zw5KE(CM8PxNMp}`b;KDGVb^?qR=-BYMwlO-@~D zQjxX52^I2fEvb^c%8gSzZhU|%i6tu0H+g5J?O?Vfih*qFNhB(Dl4p4DR}I=l`YU{s z(1F|wH%<+4(<0Ut*qaxVA}-#F2x2F2q6zF%w*XMi2U!;#J;{E2zK_Hdq{6qo`W0Uy#BeDc?`RuZksABhfC}N#%*E!8IKFK8gG(rKR-m(1TFVkGr0Zy0G+KX9?o? zi>KsXAY&H*R5WC(wJ7jipP&?2Y3fKxw)eHknUyZ>YOqM8GaKhfpz;f;SYryFU&o%|7}8{OkY99B5>}v;JFm~4)_1p3UsKAoBMb16DGE}GHi5zXr2Wzs7CK+_U-$!su1b6Q z9_da{UeM5V6zg@cH1!V7Ik1a2{}uv2<_3*7E{6npQ1sv?^CbrKb0z#j6+A7TL9QzN?=(m9QR?a2xeaIuAUJxL zx`?YoI{vms$m%YoW-#imx$0QZ_2$={px^EBc|ouLMB-Y^32&?7t%`57T`78sdpJ_@a$GY=kKc6zo(>@vzt*tM5u^+oP^4S3IJ9l= zNw%Ujv@`5O-Te27Z1AEJH*mel;r}H{Q+net=D3fLqVRAt{JR@Qz2KV3`Jf!ar*28; z9{bUlH)&ol`PWD;zokldI5I0jgwAbrn|IT3Iy&F(Auj|leO7@D8IC&qR-cLe1?L{3 zXRGUAPmAmwXmVI`oRoMd=QL{NM^w)=yKIG$7Rh176Ni{0dRhUOip)IOzsQ7?vac6& z`zA6ej4e&9V?j2Ee(pfitKD&YNwixejZnxx@J(J3C_bdIw19u+HKOo)Nlh6+^9yNz z)Pyi^^nwXsHF^vP!UZ?UzG?OSqb`w10<>#`Lv2`thUfgyxHQ-ntkQ@p6)OmkccJh_I5@*jzhVz2i<$~1) zuMqwG<<`qA??u#4MhU((%nDR6p|Yqqtev|$1vy9yL}#)WT~lMDRsSUao0xBIFb~yj zsF^9FYqle5-*T7V{^tlGi}sO8W>jrgH1sdD8ixI1=7s&h3uLXnB50A)Uh0o`RQtyE zjZH(BEZU(4x@duEK_{}8Ie)hvwC>66GwBcfT(Qy8|w%L-zHZ6h@azQ@u}_Kt~sP-?QT2DfC!pYwrqs*Ek`zKqxIjHWFk;8zYk;kc#}8~FPrBBTKk zas~q|C7|?o=W1hG7h#d~T-(5DY?rlf;5`l;CrJGp#E;v%&O7lf3Do<1&`pJ~Fujh$ z0Tgv^!2@}O;>Fx}?M>K;v7vfFPuYp}O{NkxaICuNIT~%x{IYBr&tJc8JxG^{vlHtb z`8C7tuNOzAf!(%$e`_}O0W`cgyl3wXeF_@roMW2yk!=Avou0-M*0i@fkwO6!zVfJ=p-A~QWGNqg42RNfeQ zBZ06n2L7Ngri~>#Ck5i<^oyp~(>S&B?i;&vO_KT{bo#;V+Z$!!^)P*!1})!wsZ9+O zJ=VTP%C^B#W&h|Ce5<8ynzQPOf~uRVxrtMDr%MLN+tv1YtlKlIG-8HLZYM!ZkoC?; z%I2t80>sn}FT==Ua!PuiVTq@bd||OUmGi?{?x6RZw=^*f7|xsz%2ZGjG=~o4%H8~p zDyY8d+)KE5r{Q?XFjY(szb@z`$F}o#(HAu0Q}aa*zD`pLffa;j5wyv0A#QUnE-&vI z_)dbPD?XTRfcuB|?y(t#DJee@-gZBfT}4TlA=ng^*U?bkrhS~|OmgI6*MQGq?n<(0 z)X2t*fp$GbE=WC(9)s@IUrMe*^=WDX@mi_-YSUiC__4R2-@N?;Z1{Qo z0}$3UzeM~?(q|@XpBiTN&@KrX0WdUa2b!ECrDgW{G?yMn^O&6`7b#7Djse~p&vA()JHDz`O(uYY2YD(Fy|&Z!c0v1#7v`fi%t z6t-a4O>8xaXh-!gfYGrbt#b@ZkhCDnw3RD|qba}Kw7bFYT0p~AqU+};(UF7`_L>SO zNvGwt3{xXfF9OMMpHeZ8Owu@_0SUJOfdmRMq_1ufH|AK61dydOd2Ln~di2X$%ILcC zFsx=WSx(3;p3{op7y-s|Kbjv8JP%`_OrVb)18Qh-VL>qeN-nrOzg5>KSUM8s?kC0U z(=mamF;e|?(H?r;jxqlp+Xb!rs(9p8HYxcLz08s9S!;<~=N+S;Gw;Bgp9wo+j%n`3 zp1o6{)MQ0v6maQmrJOa1@9H;OSu7nu8NSpU`!bqG_I z>yOVyN%GV_F=x3S9;E`NT9GGQB#m8@*-%AFaVy+my|Xv93NiII@S)3oO10^KT7xpR z)e0&l1PoV9x3yPEGH_C`{V|_zu@+KEEk4q@F@VNun=2NTc`_Wzif^x+%7IIX&WE(N zuAYq)@)t&$LYEHzw;M>7M#=)7Np;6fInBDH#Nr9T+(6o6@*e-$bii?vqxU}tS4S%Z zs@{CoL94X8AEM4`2K6729z}T`)~Z{-3g*t;OrR5UsNina$u6&!7FxH-iZZgS5Oc3d z>VGN&KNa6>WIJfrZqqFbU-^g`*7mc&fQffPePevYq#XWr%ZmPBB3+pIM5c(K3$(8q zXB67LulhU|7Tkhb7=1IKy#PfW=iU*>N!~QYj>-{2rV){Ny4B8$cqfeP&vZ9MaDeu5 zE0!HlsetPOS>gEf1)9acjxGLvaB>3kH>?wmj_GX{ratwM}?z3l;75R(=% z3)CAVsW6UWEp%yb8_gSvoz-zB<6KWn}9G4aoZY* z2UED#({RgNq%7rzat=84=zN*}6r4>mBD)%n#T zb6Zof($X(_Z_wu1aS!Jf6Lnej5v{kYJ2B`Q`T95r95Ci(;aI#qUgBd5fS5L+e2D1Y z)1xGx7&L7nRHKh|prjD-r4#e~>nssLWk5T$Qf*qm3A%{kyf=yA7W@RMH1a#fTpt$h zlLT4ay_}$n+)mLk#!aBM8mj4@ADkRTmqmxRS~=S*K;1`d2)>5V;5$3zzkr7w;3->cRpOO!iiI>=p{DvTT<|-V#4pO^FF&bz1L8E(n?;n=hI^ zBY!xxE=xJ#8>x`1L{cJ9uwJ<3W=d8y2GrHNC-=`YkKjN7uUmX~mC=!yU0RSm1Bc=E z0);W7%3U1>_We?AyY%i#8qx*xEv@S$@t5;&F6#kzRC@#;L3e6G^>3*A+@oh>%Iu>* zEocEZ|LveCPa#pL-f_a|*j5*`HV~g8*Ymg12zfoUZlXwx2u0Kjbls~`ValvQ)zLSKJPh?b;_+uNwahTb%asPL>$db?Sa*<c+Y6l@i&(rSHsSa0uZU!t zP?c`>({tTRoC>eeMi;Qe+R`y_ zT{E6o&bHjty!%aR)PUuhudp$~?(KGhp!Zid28bMSFK+>l8Cg$rS>pK!5Ha%1>_~1O zQ5uN7{k54N0Z!8KT9JIQ{}4bUn1Oln*kn*CDV3$$Z58uewZwAKNJlRDf5Rku>m$m6 zI#)b~dO9Ac=lfw)xX(=(r&+~Qe)`6{%06ij(>u7&>4}TRoIyb=i(He7&QZi0Z0xi{ z37B~pw$A-6^W+F2c}8qBPc*pCkS?SJ^ zl(ZE1xNlfHc-Dxp_>NOO*I+}&Jdv_&cAsps1XG0B ztMfuf@CJ7FXcI8d506IpDeFBWQmMllUz5u-J5B9L-I4%rb2%)v)&Vt6k(4}>*y=Uz zjr~-~Kr?*ZmSC;De|vQZB6DoAA$XX*l6x8VmVK};f=nlR_bXm*(A&0*$Hrdkv;mMb zrC2IHljDG7)L!IBUuQxP{CXl?P4uBLB?i4kho#S<`AS6 zufGEa*^UQ|B2-A>eK{!OzE>uHsWm0#vU$P0_QL(B4Z1to(tEbIz%&fmS?T??|1$%VXKjxh!JA$Mjq+e5u**`PfLFUXMbJk>b!sPC|N)WLXx26mFoWGIduqvSO6b~u<@ zi^)5U-WEB=2@Z}#D`S%?tLf^UP*y6xg^9RtL)sHMuWpE2W9NOy{u~p|zD{AzAhS~K zshkgWH}tg82kno?dcLX5cPK-Dr0x$_c|76~eMjWZgWc_e5Wq`NliztGHdej!{tLVI z^0PH34~d?CQD(Oy*IN_iK?ULh z)BR3OO$Bc7rHRK{85R3B$u*V}L_X^UZth!ZUItw^XSJ=Y?xrFwd-xD5y#K;}Dvnk6 zccpn+;g4~2qe*BOPkMszTyog1EskkhmKP=?BdIM8-1Z2NxR9+Cygx6QUws{ocV$6> zr9GF;^=X>gjLhh^yq3)^{AK&gj@pbn=#hbCh)yD(J^I$NTD#$@#Puu3E^fGE@0Q+F zxrwuko^OB6{@+SMiQ*c1Q-k$&{*IuLPLMo{DZ!{m3WS0BoV=^*r7^O-nmv|Z5V+k> zaui-P<%JRUQS67`#Ii({vQT#| zWMSC4*M8{%=9xg$#yb7%XTFo!~QTd5Q_;)mmd0VAl|L32Sg4{sn#0-BzFSZu+wRptw6=^wI1)OvGE6uSzUo zJPlu2+WsBP5siQ}zi=)q^F2X>i;t=W6Ko&yktM5`kr{-~#w6k+HH{GAuxAtDdupL< z78VA*L4g9$N$x3tx8RW(H)#W-T!;aUjD1o8IL)7`y**J|852f5Q1&A$MKKrJUzRH7jXy zqIOx$$G5_*`ebBXr$cp&yeNR_{pkPvtwZPRWr(j}%P1yH@D9|K372vl_t9}(K&{WAtxoWiIU%;sFOJLpOUYwc_gD-AWyWI($6EO}<0 zjDyYHAQoYA8hxp<|EpfI9EFsTWo>KX=vdnKFL0Jk(B`MoWG({X+1*(M}B$taq&G#3+E?N){voj zJ0ldnxD}QBuHk>)k}ZB?{#Tq@jK7rfrA>~3dVNAf*yVs$KII?bJNiM#msGeUIEv=C z@gf;IQUAYk#rYs;^fSTA!x|D{rN7P}G)B_Y)}_A@V)+6}bs6-5;G#c&j4lm$=2S^| zxjl(rTIreXB@Z>~Y2>0!=L4^%@O3F1vkN=|kAZTLd1Z#wyZaj0}H99KXibH6e{}N+|ZJQGoL}}&os?pTV&`@TzCszDgX6WT= z3yMsHNllT>JWR<(yv)v~GgJN`%rh9>5vJZQt7}PcPpBv@?Spik(V!M{zE!l)Lvxl# z_etT8Vruv_q|?~5QSPc6AL^F}?WySSPnxnO`i7ZVp1aQ_rLt9S-e^ugfoshy`Bv91 zdMOd_uI*CNI*B!X!eO1Goo|B*!Y9qm;^BslBO{A8tpIh>!L0sA+ESN6^t{cJk<8#6 zS1-JWDbCQ!IXdksdK({}P3uDJ?7w1gh}iRJ;YG))Z3axI)8HC)^Ovy(KERYNUh_rn zgT4gwr<2vNrtgWr&+1X>8&i1><>bsW{gm<9 z`Bpuc2fP@quFGwT>9-wZQh- z#ysvgLh?3@zF@|BE%_;L-|jCahj)cek4BfHO8L~~x{hCB4HcqEVq=sBgV(6cZ8Ll? zxw0H~;=|Z`a}om-`S_IB$u(yM*x7NiR33On6^;xej;= z=#AA`|CZ?NKOWC%`Q!;*I9*jAQndE7#}GYAk-xtUL~6`$f^5I0-sA{E;@O-3eLX(| z%=ZS_)0_e_7KXv7<6rch!v?Wbl3wU+Iqdqadr#My-pZerySi^Mq3D&U`rB#e_LU@Z zwwdwUdDJ3&zoyvKBS6r!O7i+1`D%>f>H|m@lEarWcDKMYx9!(gf9(BCf_Ucfc!}}+ zsP>QAXyk2Zi`6+lh`H(523KLb*L|P{~xEJRH8TuzvsDO%^8Y4~mS9B(Ey+DN!iTZ7Y z#r#LmA{&8PPc`$EVxiX6*0xs#?bi~winJXxe-CngDQ&n;J*HP}w$`c*F9d`S`DSES z#}eLiyv!6~^hMZyox&I7tKp4PsDP&}lHkwomk-rwzB!P#`Ss#r@b2-GUliVVPpkSV z*G*x!k1&a=|6Xmzc2);KbAFh_tQjCF!D@G3p06v8;tN?S|nMcNH z%tU(GPxj-%%oID+YK2d{%>AqC4&x1O!l1a_ZBs{p4KY~M>~E*>o_E(+W0fEA)!f>X zsX3eBeU;xG4njPWC9kf-xi(uV_RSp67iNAzO>4(&M?N`7z*?jj{<|DcYl~Lrs6tP7 z#kd{}-x-ZzbZV zp|Mnc!HedGyc)DPLR$al z9n)aUm1vHqAPr)TVYF=c-Gj1BH;5;Y+!X{AoY>8qPHe$d8VtSlvcO)eFHe9$h71x< zziTmTTXtIJD*88^IvhXu%O$5^>SZh+ZaTktzwmDFTZ5QnOgO{oF8J(3J4|y z^n>6~L$Ce@_qQ8aiZzu|Y^)VA%o6o*GivvRQMNY8ccicDx1DnpZnPCd6ythHF>RHU z7!{Tq=7BizXD?w5sdo7W#1knyl+KQn6F*xHV(9-H?E9y>4t2@jEn1p{`D!np%{$jJ zirUTqP7DWN2wET{r+}v1aB0O0rpN~yC&tQB_ofP}9N=QcSl6h+b+Xv6_{|LA0l0s& zVkRJv8L7N4qGnPQ_x!?3@uq(o!c#Y;XC(^V5!8FDt0wSnx&er^dv32p0nI#0!mavl zQ(j^w-FfJKi^wVggX#+o>#y<=cz@M!`+^&i1{0q6{1EA-6~QA?+x$!^&zl$)t)11> zeJ@uXwaA=~lKl8UfjBQ}7Z32hybg{@KD}7r`}>W{3ODpZ2roYoq}0*GKrN>bso2Nc z>)9JW8oM(_XJZ-O^Ig(lKafOL^SF=8ZgCVWzbEFcX2iU(7}D=ChGjR&eS0ynum5-5 z%7yo^TlltKIX%VHLuuwq@Tg5pWH*d0c+YdSTp~pB9K?#V5h5vO`2-j9TE+R(Q@eh8 z!S6~Yc&_yFT`rVN#kX9(%OV@Z$HZx(F<3=$Y~aO01)|tQs;cPG_{A!2(Rt>c18xh|Fd~@hiKLLNdHEOjUXyWj7BqT|d`6yawy`E^)D=8|V z1ua(13&pEk?uv!_Zbh9~T)bZISL#`aI1xRZqnSkv<4H%sHv|GTE~_fqn~mm}5{6sGs@A*EeTEjjl9(K*wI539e;x9l zvQ8`BRmX@W&qt5yo?kj~%SVW}*Ocv*N>M<*^W0B^*rm9sE~^ezjr;XJ9N#!9PM63R?Um!TYq`6#+55X; zgUL<71^$NCu#n=QOYZ~SM#9n09A91CRcbwNfKMg|fY!WBKeQc_HTA55xRqU8MLsne zL+-`qHs+y69Hr+rrvFY!tSs3<;~K zJk?ey}p$mNE76{4FF zSnBO`B;_Znm}AsB>)J+N-6q8hv}1qdx$6@BYR=1JK?b7rUJn;tTFLq!WJy|0R|{>wpmM!4$CJh3V|^3BD;6gVSJpU=#XggC zijq7jx$Hog0Y0&`Xq&6IO&L+ePBuvA?1;l4T&WpES`-PtVzWL2i|4Qmt(ewOP;5JqbFd*ux>}5le-c?}RHSE`v z6r}ndY#ZyC!Ufao$7@O-Yaa)870eIM&}`-SL_71OtV_xP{3U4s-bej1tkH=1+|P$b z#)+i@pNO*t^h3d0)Tc!21u(Ht$rzC}GW}$Lshr~_l`$GoHm~hjKcF7dN=Lg~_2=Ra zz0=&>;auk>HP$5ULaMTOCCGA;%~*~f$2^qZ$wK{=%8GfJdI_{?z; zJ3+}>Ug%S@jiwR67cjFkr0Te&eK!^>k9t(Np=IUXTe9FeNRgUV){gHTd^L`JYD!8y z5;_qh)_$+bU(0-`ASr1h)d7odO7)^0Z%7V~lf>owq$QhtNM=1EQAG}oGkpr$$gtvn z=fp3;R-@F@%yN;V3-`NT(1(=0bE%iPwO#aa1zG8A?=jT#;B-7rf1j(Mq^*Eq|1u=` zkgcmv6>6DD3pt7SY~CJ77)d9q((``FTw(e+=GSTX#EI~*!PBsEuG2334xsI=--uIzd9|2o5>Y^cOT5@NvAdBYZDEgC@i(0g072aJi90wV8Qx+?peKA$)6=E7 z8G`S|>KID9OT{eyF(KJGeEtIL)JNYHDrg2M*Kx+fqYA%=UGEuJ&O64u;d306qhc!h zUllx3U^uMAbNiJxX_Kp#CkW>0Q~KVS-$i0bi68VG@=|*Ghboi8l?c&o@QNLL+`}7aSMUb6uPmUFX zhzwIp7O&e^;+q?r7MvJ?H7&m@-%Up}%F~eeTaR)h%ZsWyOnnahX)l;fv7BUxE5L2` zg)bfgz#oALATYMle^0HXmkauSx$Ql427s-DqP2ORsD47?9|%*7&&%pk+O@9^%N*Vp zCwueOBIqi7cYPl9#X&aOGgYO}>P)mfzk zm5-k-gs~a5{G6dHk$$Js?NMe7T1wtOtLK^5^y4+=fV^?((sPp9fV5rC`{Ulg;sTHg zRz(fZ{iq2siFilftx21eMpN%aGca8q7$+2RtuiX2F+U^%Wpw98Jjw$MYm07AA+5CiAKQsL`U5hEB%-N5L zhHw%8?pJQfw>+e4Ko3v+5vCHMX}yndwCdwM>U0}Og*b~rsl?Nkx^X4+d=`MA{8hBj z=q4^ha)YI@oLIyobvsP{b;$41oq$SvCXm+-zNQo&d}T7=H?hsC0`uM!dOEof=GW(Y zi+;kqBY!HKn{_Dmp18C_-w0O!CD&wW@7<8p)dYO#(2lhXfDPYB+$s0&mNbhqXk6y( z3%p&JJ^hOOQlK+PzosZUo-Ol|+Uy(onErGw|Fp0fUkTD;?kjnCH9qESBXepY=S9yr1Uv38CA-0Blyr`LI&i0dx`AuF&Xx?df*Ro7S)s z_phTw<|z#3^-@UFC=`B(Ek9*VLxx-(wl0mFNpIP~!PQ)(2V zz`wJ4;z*;{jvvsqWIx$9?@u5SBEBB&$yJVm-}cP!B@-6!k2r5zn#^6!?%jfaVyfWo zD5lA`34(7-oFL7WBWGYI+r*Ok$q)6$Z;QSHvS6=S#I!E7fSYtug76`8ukz0zDG(zg z;plgN*W=0WI{Rd1*k_}yTLo-!9XPYa@q0Cw`ITHL=$42SbssHQLY^&FOIxR(1;>Hn>F9{W24R{ z>@@bq|J=aUAZ&?{w#@fe1Az{r`s&KQbdF@sfFrloc@G%Kk}peUO~u=Hr-YJy z0KpmR75j$vS|N{*X2V{Y_34z6GdTnVqzfp|pr8HoHI~D=rr*Xa0*V|UcY~z=cH=@b8duYgr)yV=g}P{ztI1FL+tViUp`adH?NZBKO(N zeB&&Z1HjCgCD;#^X&YZ#v=>`1sb6(=1fBVPJv>`1E#50JDhy@)I9vW}I$A<;pJp`U zml#lGDH~}H{XHja(@_(zc?al|w3$H6XcBZ)f&cj`P9+Y`gU9OsDe7_wd10)L>OWKu znKt~n$en39S$7h*eXrrLIPRf+g-?V;OnSZBEulW^xW+#NXvjB}bgUiePuv~tWjmN! z5BLEx2MHiMbgDQ~9PIS@R`Km0xetOIP6Z6KD!^i+ok_`q>d@WGTgE7M=FK;lVb6Mr zXH7T*c>d#c68+k73QgrW=baf{mq{0qn~!3nh!*^e!yGUKlR2ykZfc2gJSMUUzuT*P zW^Gbuf9spUorn2cRGmuWdiI7LWZv8zK+S$x{Qb@La`22z&YrCVYT4?WH%M0^Yl}~8 zVmo~qrS1Fhzg4O!gEm0if%$^hiz^b`lCURadrD)!);_3QXs!o(c%YRvy zYf))aPo=I0S)bb-Ew`WD9QsKd{Spe9G-ni7 zqI}!52_36>Z05BZ%Qf}mh5kRuB$If>C={g;M-rwr0bGf)U;V8^IXH=X@71Vn=aa;9 z2dv$Tmg^3+pLS@u`mds=q7k55x}D;HahaC*tKZ#iKwp<8eWZmRBQ4tU&xw6)O+^*` zrn)309wY7FzNet954r6El;*kz_2*8z6ScL?!1`EmKjnz#)*b%~XT*i)P%I`-V`zS! zJ2kdE-Kn9kqz)Z=c?OSc-eSHFj`-c49L>&!Q*rIs8002yyqeye@OokM zbP17(<#bBCAnxKd&ZE(m?E+$VQ0qnKhOM`#eSl4_njJX9GL~JTrjt)X*j^weTWcRV z&5p6&YOk;)Y3HdB+TF{{mcJ|&>{?a(Uj_LC44tgHZ(TPIetFZ!Qsw9F0>|+jJ#3pm zDES(eD-Y+KOD_9(cZKs0N}0(!fN~dEO~uq0?l;N}X*zoLR`?ocS=q6%=BwD|anqEZ zczQ_#LV6#kKS&h^>tFXT^WrGNweq7PQMtJB@v|t-Byr$>C^ z$lQuEAi+8)r$&+Bb3khny_XxcepKx+ec+mb#QR%kHl}LeT5o`zwaX%}aagEB&=>Vh zYT>?G9#dtjW&^e$_{L)Wy&0RaQYnWg`v!p8$k0eL^OOW}p^&6>R>#9;m?)^CQtyY% zZMnH5F!MU7Fgg1ZiT@f()q)W|Prd;*mOp9j=Xvp&25FO-_f`Ybk9|iP&GoPDUnzVb zRMC1U|J-f?_(>sIcxLxxLgiF4F=M=z*M>enk2izlRN{Jh(E~LNc=E3*=3KrPfid}u zb55eRTQ1~JoslU7I1?_#tcDrwiq}UgorNQO3ymJ7fO29;96o2w&CSgL!bL}u z`36UrR0vFBq`0EnKIzJ`KiCYQiY^b-{$5D@y!+1ZVAJ?~o(A(Y?zq_(%+$Rf&f9iG zKuW2s<}bfY{ACEwEL?s$s@SNOsAZPo;WavtFFiR9bq=cryLf>n4Ams*$H z_vdVkY8*)#(j0bjoB?MLW5zQ_m-B-~C8q5eD^$aM+XJ!b7ZwN5f$RK=f2>#Rnu;{T zt!ito1Ic)+xi!(JTmb{O6GOmUg18;6?|tM=FQuZFuxJ3%h;85YdhF*V5wnNZR`6zb zI(`N(cN8*^#=8l~jE5$~v0VLUV`I~<|EJ6%XoWfv54SicoRFcn=A6jUNa!>}1W!s0d48PS=>F2F#DQ7aTcM29lG>c-d1l9k=rSKJ!C zv+l_~4R$DO`;}{0!@8kEN|QONMC=mDX$@{4gdjVR9=lV5vi}_kaTJDP&}3y`o>K+I z`6_PT(N4l?1$X0W9nSTKq+<7hLlfeMyT(1`mak3WOMe(NmfA}NV%*Z-;ecHb0dSbo z6UJI0gsZ>MxoY;${%uoB=|{!o%TaF52$|6|2dpBJFe_X&Uk1QIqn@}5L>2wZzXDZG z+QqI-=7fOTjmFf zBbp!0^l)}-Jspy$Sggo1ZWx6@b!;Pdy3!yL$WE z3BdkBp5bnbZNt`Z>JwPPr&s*eR6$HhITYDp$d%9rk3KFdqPh)eUCLfy!DjnMyM7%l z`1>!**N@oc?vdMAMRfA!N^Kue88mTvCk{uw8h;9!oao?hNE0EbF06meyeU7YE27Qr z$M-T37#G!vqed!+!4=91whv()kvu6lih4PWS zY|M4rUA~K;=9?+retn;4BwjDiDz6x~PX?VC^ly?bOhUs-Ibj(+T~ ztR;p~wj+f&$Vw-B$|&E$yw)bP0{`VCD4SCh$bQn;$Ot1E;$;9Uc)g-1`0oAcx#Vrv z3Q5`cZ&f;aqL6VuYC`A>Tr7ivZN5YT*?0mq7R(r76N2#`WiRpWVkSikzuhxGAR9!f z9VDReyz+q*TJE~{q~Qlg?9LMKJ7tN-0`p7fj3ktI5aD97&=nEq%|7I=!!Kq&|W zY$O80yZ4GT3YquD1`JueGCa|Rqm|1oZse2aT(R>HLL*<0Um zNkYeMbSWS1pqHt7`hp(y(SMD-q@(zfFYl#@J49k|(yc@oaG{$an>-eX1V-a_d^E|< zgCwF>s#HvWR5~mO(1R$DwvEWX&pRaOCADAU;3AMqq8`B$nTls;rHl(G$2UJ9{#kb# zT;lM9tu`6G`PJW*pL>PawE#WdZmgr_N*vY7mnWVd)IMV?DE+y9A_b3$d)U7b8jvm)`tNb3`7b8DND++P7p#bY1q`Q8)ISTTnlzVDhQV^zWRiF;i5w96p(|`>?^Zcyj1Qs=z-|3_N zIi&e${5=SdPwLr))y`hR_Gnb?dlW~&HjSMJxxMkgC?)SG9s5srQo<#d5jnkAuPm3Y z{+#dSgn=I?S!A6{I-tqq)V@&-vASr1K@8r#@g4d1wOTy!@52IC*;{&Rig*7HmTLBV zQ_=PmL~A~NuQwDdR@W+-2|FK1Ba9^5HSL>LB1UWzqup>dQlC_-g1sJ!$)PS?cYhb@ zo48J7!=EW(3Eo6`jHV06t4eKKI2OBux|%$IKyC2;sLFjexiEy;qfEZ{!jEtcJ0^k+ zC?2ZjKa!hjL|j|OZa$kwLkS{X;z*wt!B_`9d9*SZTqpi5bDA%%{quu^XQ7MlNW$+# zB9sj|AXqz-7YCi&Isi*4nW>dhs&1@N2le5%&FBW2o~lj9hmrHfN&dZO%++X)u{RPa z7FSyR_b-C`3dfWlra@Pyb%sU1;OPm}0`e}mlKn1H{YYvq(ABS_35LOh2=uM-H@Nut9Evf(bN;2S*P$!-yv=J1jJ03@E4u5f*P<&U-&Ue(un?>NW zQ0u%D$Z)vHnC~c@#_kglH~978toE0a$6s_3I_?RJ!chC~sv~U$_C&b^_i} zl{FVQ>zd61gZ4D8$yqHoe*9!`=3yt|jY%A%%+IHNi;W)Bh8}i$0XsjVDkWH`Y1c{1 z$!P27g`?6;VwuiWpeliV>z^qhRUB&Xf+TUil0VB(Wuc^?--emKQgET+e|}8ZgfkLP z-?~^F0pZK5sQOMECmG}VH#N0OE$BH7_CDa_{Uqs4>q|VE^;8kQ{WQ6FL@Y*T-Yx^Y z}>tVDy$;r8O6ZTk=bE%f?j%slqsm_+0?ysOj0Vz!YRf@EgyB>+Js*5L$@S zB*32OTJ|jgU>?hp@0NWL2f_^2yu{a5nLJn;Vkzo119 From 46f8399567c55bbd790709e110c410c26a31ad0f Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 17:53:11 -0600 Subject: [PATCH 544/590] Update cpt.md --- ALCF/notes/cpt.md | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 3d3fb51a823..18111f77928 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -11,6 +11,7 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. ## AuroraGPT V1 (Stages 1 to 4) +![different stages](./assets/cpt_images/stages_training_initial-1.png) For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that we are using. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. | Stage | Dataset Symbol | Size | Source / Path | Notes | @@ -23,7 +24,7 @@ For these runs, we have 4 stages of training with the first stage producing the ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) -![data mixing](./assets/CPT_data_mixing.png) +![data mixing](./assets/cpt_images/CPT_data_mixing.png) Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. #### Stage 1 to stage 2 (weak distribution shift) @@ -50,6 +51,7 @@ No buffer data is used at this stage, $\alpha_B=0$. > In practice, $\alpha_0 = 0.05$ is often a safe starting point. > Increase up to 25–30% only if forgetting is observed. +![stage 1 to 2](./assets/cpt_images/strategy_cpt_stage1tostage2-1.png) **Dataset construction** Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do ```bash @@ -139,14 +141,17 @@ if __name__ == '__main__': Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. At the end of this stage, we have ***$D^{CPT}_1$***. + + #### Stage 2 to stage 3 (shift to math/code datasets) ##### Naive strategy You can try the naive approach but it might not work here, stop early if loss does not recover. ##### Strategy 2 Mix in the final dataset $D^{CPT}_1$ used in Stage 1. -1. Construct a mixed dataset containing the final dataset used in stage 2 and $D_3$. +1. Construct a mixed dataset containing the final dataset $D^{CPT}_1$ used in stage 2 and $D_3$. 2. Follow the same procedure as in the previous mixing strategy. At this point, the model has seen 6T tokens and $D_3$ contains 1.5T. Here, give $D_3$ less weight. ##### Strategy 3 +![stage 2 to 3](./assets/cpt_images/strategy_cpt_stage2tostage3-1.png) If the loss is not recovering, sample from $D_0$, $D_2$ (not the final mix after stage 1), and the buffer $B$. Start with the following candidate weights (some exploration may be required): - **Mix A:** @@ -168,8 +173,8 @@ Notes: - The buffer should contain representative or difficult samples from earlier stages. - **Important:** Add samples from `D2` to the buffer at the end of this stage for use in the next training stage. -##### Strategy 4 — Reset + controlled reintroduction (if all else fails) - +##### Strategy 4 (if all else fails) +![stage 2 to 3 decay](./assets/cpt_images/strategy3_cpt_stage2tostage3_decay-1.png) If all previous strategies fail, apply the following procedure: - Take a checkpoint **before convergence** (i.e., **before cooldown**). @@ -183,7 +188,7 @@ If all previous strategies fail, apply the following procedure: This follows the general recipe described in [https://arxiv.org/pdf/2407.07263v1](https://arxiv.org/pdf/2407.07263v1) -##### Strategy 5 — Rewarm from a converged checkpoint (last resort) +##### Strategy 5 (last resort) If Strategy 4 does not work: @@ -194,6 +199,7 @@ If Strategy 4 does not work: At the end of this stage, we have ***$D^{CPT}_2$***. #### Stage 3 to stage 4 (shift to reasoning tracex) +![stage 3 to 4](./assets/cpt_images/strategy_cpt_stage3tostage4-1.png) At this point, we only have ~6% of training left and one should start the final decay. ***If we didn't use Strategy 4 above:*** 1. Try @@ -209,7 +215,7 @@ At this point, we only have ~6% of training left and one should start the final ***If we did use Strategy 4 above:*** We should keep decaying with $D^{CPT}_2$ until $LR_3/100$ then introduce the new mix at $LR_3/5$ -### Summary +![stage 3 to 4 previous devay](./assets/cpt_images/strategy3_cpt_stage3tostage4ifprevdecay-1.png) ## Legacy agpt-7b checkpoints From a73ccef9ca16b0c943d9e2d1cc44aa7cc0f183da Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:03:13 -0600 Subject: [PATCH 545/590] Update cpt.md --- ALCF/notes/cpt.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 18111f77928..9237c6806fc 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -48,10 +48,11 @@ No buffer data is used at this stage, $\alpha_B=0$. - Start conservatively: - $\alpha_0 = 0.05$–$0.10$ - $\alpha_D = 1 - \alpha_0$ - > In practice, $\alpha_0 = 0.05$ is often a safe starting point. > Increase up to 25–30% only if forgetting is observed. ![stage 1 to 2](./assets/cpt_images/strategy_cpt_stage1tostage2-1.png) + + **Dataset construction** Use [mix_datasets.py](https://github.com/zhenghh04/blendcorpus/blob/main/utils/mix_datasets.py) function to build your cpt dataset. For example, to mix the lucid papers with weight 0.9 and the dolma dataset with weight 0.1, you do ```bash From 0e259a2428331366053324dbaaef4b59ed797185 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:04:49 -0600 Subject: [PATCH 546/590] Update cpt.md --- ALCF/notes/cpt.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 9237c6806fc..fb3e6044d08 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -16,10 +16,10 @@ For these runs, we have 4 stages of training with the first stage producing the | Stage | Dataset Symbol | Size | Source / Path | Notes | |------:|----------------|----------------------|---------------|-------| -| 0 | D₀ | 4T | Olmo-mix | Pretraining dataset | -| 1 | D₁ | 2T | Dolmino and fineweb Edu | CPT stage 1 | -| 2 | D₂ | 1.5T |Open Alex, and proof pile II | CPT stage 2 | -| 3 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | CPT stage 3 | +| 0 | D₀ | 4T | Olmo-mix | Stage 1 (Pretraining) | +| 1 | D₁ | 2T | Dolmino and fineweb Edu | Stage 2 | +| 2 | D₂ | 1.5T |Open Alex, and proof pile II | Stage 3 | +| 3 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | Stage 4 | ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. From 13e0bda8c4938de704d421602f634de83a6a6930 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:05:11 -0600 Subject: [PATCH 547/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index fb3e6044d08..42d3d762076 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,7 +1,7 @@ # CPT This document serves as a strategy cookbook for our current runs. CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed while avoiding catastrophic forgetting. -In this document, we focus on three CPT approaches: a **data centric strategy**, an **optimization (LR) strategy**, and a third one mixing both. This means that the following are **fixed** across all runs: +In this document, we focus on three CPT approaches: a **data centric strategy**, an **optimization (LR) strategy**. This means that the following are **fixed** across all runs: - Model architecture - Sequence length - Optimizer (although it might be interesting testing how changing optimizers across stages affect training) From dfa754ece7ebc2e6eab93da5da8fdb25945f6a34 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:06:26 -0600 Subject: [PATCH 548/590] Update cpt.md --- ALCF/notes/cpt.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 42d3d762076..53c43e4551d 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -29,22 +29,22 @@ Note that you add data to the buffer B after the current step to be used for the #### Stage 1 to stage 2 (weak distribution shift) ##### Strategy 1: No replay -**Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN** +****Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**** Just naively continue training with $D_1$, no replay data. - Continue training using only the current dataset $D_1$ - No replay from $D_0$ or buffer data This may be sufficient under weak distribution shift but there is potential risks of forgetting ##### Strategy 2: Replay from pretraining dataset - **Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**. Then, - 1. ***Replay the pretraining dataset*** + ****Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN****. Then, + 1. *****Replay the pretraining dataset***** We mix data from: - the pretraining dataset $D_0$, - the current CPT dataset $D_1$. No buffer data is used at this stage, $\alpha_B=0$. -**Initial mixing weights** +*****Initial mixing weights**** - Start conservatively: - $\alpha_0 = 0.05$–$0.10$ - $\alpha_D = 1 - \alpha_0$ From 8f19fe2304e89f8f152a0b740c72a86b854ecbeb Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:07:19 -0600 Subject: [PATCH 549/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 53c43e4551d..5f05e744892 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -29,7 +29,7 @@ Note that you add data to the buffer B after the current step to be used for the #### Stage 1 to stage 2 (weak distribution shift) ##### Strategy 1: No replay -****Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN**** +`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN` Just naively continue training with $D_1$, no replay data. - Continue training using only the current dataset $D_1$ - No replay from $D_0$ or buffer data From f02c699227878ff8af3a933c057e0e728514e0a9 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:08:31 -0600 Subject: [PATCH 550/590] Update cpt.md --- ALCF/notes/cpt.md | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 5f05e744892..6c8771d4e9d 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -29,22 +29,23 @@ Note that you add data to the buffer B after the current step to be used for the #### Stage 1 to stage 2 (weak distribution shift) ##### Strategy 1: No replay -`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN` -Just naively continue training with $D_1$, no replay data. +`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN`. + +Naively continue training with $D_1$, no replay data. - Continue training using only the current dataset $D_1$ - No replay from $D_0$ or buffer data This may be sufficient under weak distribution shift but there is potential risks of forgetting ##### Strategy 2: Replay from pretraining dataset - ****Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN****. Then, - 1. *****Replay the pretraining dataset***** +`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN`. Then, +###### Replay the pretraining dataset We mix data from: - the pretraining dataset $D_0$, - the current CPT dataset $D_1$. No buffer data is used at this stage, $\alpha_B=0$. -*****Initial mixing weights**** +Initial mixing weight - Start conservatively: - $\alpha_0 = 0.05$–$0.10$ - $\alpha_D = 1 - \alpha_0$ From 89ca4d4b016dc1c77f95c937a36b19514abf3e96 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:09:47 -0600 Subject: [PATCH 551/590] Update cpt.md --- ALCF/notes/cpt.md | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 6c8771d4e9d..f703955e929 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -37,15 +37,14 @@ Naively continue training with $D_1$, no replay data. This may be sufficient under weak distribution shift but there is potential risks of forgetting ##### Strategy 2: Replay from pretraining dataset -`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN`. Then, -###### Replay the pretraining dataset +`Important: USE A CHECKPOINT AT LR=LR_max i.e. BEFORE COOLING DOWN`. Then,replay the pretraining dataset We mix data from: - the pretraining dataset $D_0$, - the current CPT dataset $D_1$. No buffer data is used at this stage, $\alpha_B=0$. -Initial mixing weight +###### Mixing weights - Start conservatively: - $\alpha_0 = 0.05$–$0.10$ - $\alpha_D = 1 - \alpha_0$ From 76bf567114ba7c431b0134dfe128d8f2b2c24f12 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 18:46:48 -0600 Subject: [PATCH 552/590] Update cpt.md --- ALCF/notes/cpt.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index f703955e929..7d3fb4258b2 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -141,7 +141,7 @@ if __name__ == '__main__': 3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. -At the end of this stage, we have ***$D^{CPT}_1$***. +At the end of this stage, we have $D^{CPT}_1$. #### Stage 2 to stage 3 (shift to math/code datasets) @@ -198,7 +198,7 @@ If Strategy 4 does not work: - rewarming the learning rate to a different max value, and - adjusting the data-mixing strategy by **increasing the weight of pretraining data**. -At the end of this stage, we have ***$D^{CPT}_2$***. +At the end of this stage, we have $D^{CPT}_2$. #### Stage 3 to stage 4 (shift to reasoning tracex) ![stage 3 to 4](./assets/cpt_images/strategy_cpt_stage3tostage4-1.png) At this point, we only have ~6% of training left and one should start the final decay. From 72a6437cd7a025781d9ceaaf6284b8f4fa4d2fd9 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:00:21 -0600 Subject: [PATCH 553/590] Update cpt.md --- ALCF/notes/cpt.md | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 7d3fb4258b2..94306f37ef9 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -9,6 +9,16 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, - Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. +## Recommended approach +## Recommended CPT strategies by stage + +| CPT Stages | Distribution shift | Primary strategy | Fallbacks | Notes | +|------|--------------------|------------------|-------------|-------| +| Stage 1 → 2 | Weak | Naive CPT | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | +| Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: 0.33$D_0$ + 0.33$D_2$ + 0.34$B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | +| Stage 3 → 4 | Strong | Cooldown with mix 0.33$D_0$ + 0.33$D_3$ + 0.34$B$ | Continue decay if started previously | Play with decay function,stages, and final LR value | + + ## AuroraGPT V1 (Stages 1 to 4) ![different stages](./assets/cpt_images/stages_training_initial-1.png) From a037b7f1945c2554c82eea2821de26c4a35be5ae Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:00:53 -0600 Subject: [PATCH 554/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 94306f37ef9..b07f3e88f77 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -16,7 +16,7 @@ In all that follows, we suppose the base model was trained on dataset $D_0$ and |------|--------------------|------------------|-------------|-------| | Stage 1 → 2 | Weak | Naive CPT | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | | Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: 0.33$D_0$ + 0.33$D_2$ + 0.34$B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | -| Stage 3 → 4 | Strong | Cooldown with mix 0.33$D_0$ + 0.33$D_3$ + 0.34$B$ | Continue decay if started previously | Play with decay function,stages, and final LR value | +| Stage 3 → 4 | Strong | Cooldown with mix $0.33D_0 + 0.33D_3 + 0.34B$ | Continue decay if started previously | Play with decay function,stages, and final LR value | From 41439c49c87050e928e79bf488959c866ba61428 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:02:38 -0600 Subject: [PATCH 555/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index b07f3e88f77..188a2f00153 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -16,7 +16,7 @@ In all that follows, we suppose the base model was trained on dataset $D_0$ and |------|--------------------|------------------|-------------|-------| | Stage 1 → 2 | Weak | Naive CPT | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | | Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: 0.33$D_0$ + 0.33$D_2$ + 0.34$B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | -| Stage 3 → 4 | Strong | Cooldown with mix $0.33D_0 + 0.33D_3 + 0.34B$ | Continue decay if started previously | Play with decay function,stages, and final LR value | +| Stage 3 → 4 | Strong | Cooldown with mix $0.33D_0 + 0.33D_3 + 0.34B$ | Cooldown with mix $0.05D_0 + 0.47D_3 + 0.48B$ | You will need to continue decay if used in previous stage. Play with decay function,stages, and final LR value | From 0a595fb3e77addb4003dd3a869c3e1f505ce9253 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:05:33 -0600 Subject: [PATCH 556/590] Update cpt.md --- ALCF/notes/cpt.md | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 188a2f00153..7527d4bab77 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -8,8 +8,7 @@ In this document, we focus on three CPT approaches: a **data centric strategy**, - All hyperparameters besides the LR - Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) -In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1\cdotsN$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. -## Recommended approach +In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1 \cdots N$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i, $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. ## Recommended CPT strategies by stage | CPT Stages | Distribution shift | Primary strategy | Fallbacks | Notes | @@ -26,10 +25,10 @@ For these runs, we have 4 stages of training with the first stage producing the | Stage | Dataset Symbol | Size | Source / Path | Notes | |------:|----------------|----------------------|---------------|-------| -| 0 | D₀ | 4T | Olmo-mix | Stage 1 (Pretraining) | -| 1 | D₁ | 2T | Dolmino and fineweb Edu | Stage 2 | -| 2 | D₂ | 1.5T |Open Alex, and proof pile II | Stage 3 | -| 3 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | Stage 4 | +| 1 | D₀ | 4T | Olmo-mix | Pretraining | +| 2 | D₁ | 2T | Dolmino and fineweb Edu | High quality focused data | +| 3 | D₂ | 1.5T |Open Alex, and proof pile II | Math, code, science focused | +| 4 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | reasoning traces | ## Data centric strategy ## The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. From 1ebddf53f470b7953f8bfb26540afc8415efe5a6 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:05:57 -0600 Subject: [PATCH 557/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 7527d4bab77..16c15b8c455 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -14,7 +14,7 @@ In all that follows, we suppose the base model was trained on dataset $D_0$ and | CPT Stages | Distribution shift | Primary strategy | Fallbacks | Notes | |------|--------------------|------------------|-------------|-------| | Stage 1 → 2 | Weak | Naive CPT | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | -| Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: 0.33$D_0$ + 0.33$D_2$ + 0.34$B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | +| Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: $0.33D_0 + 0.33D_2 + 0.34B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | | Stage 3 → 4 | Strong | Cooldown with mix $0.33D_0 + 0.33D_3 + 0.34B$ | Cooldown with mix $0.05D_0 + 0.47D_3 + 0.48B$ | You will need to continue decay if used in previous stage. Play with decay function,stages, and final LR value | From 48892e3ce36823407ca4b0c76f575414b4f8d8e9 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:10:45 -0600 Subject: [PATCH 558/590] Update cpt.md --- ALCF/notes/cpt.md | 37 +++++++++++++++++++++++++++---------- 1 file changed, 27 insertions(+), 10 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 16c15b8c455..9689ad6b28e 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,14 +1,31 @@ # CPT -This document serves as a strategy cookbook for our current runs. -CPT is the process of continually pre-training a model on new data. The goal of CPT is to continue training on new data while retaining previously learned knowledge and avoiding forgetting. When compared to finetuning, the goal of CPT is not to improve the model knowledge and performance on a given downstream task but to retain and improve current knowledge as more data are being streamed while avoiding catastrophic forgetting. -In this document, we focus on three CPT approaches: a **data centric strategy**, an **optimization (LR) strategy**. This means that the following are **fixed** across all runs: -- Model architecture -- Sequence length -- Optimizer (although it might be interesting testing how changing optimizers across stages affect training) -- All hyperparameters besides the LR -- Evaluation/validation tasks (needs to be fixed from the start and be consistent across stages) - -In all that follows, we suppose the base model was trained on dataset $D_0$ and label the subsequent datasets $D_i$, $i = 1 \cdots N$. What it means for us is that stage 1 is training with $D_0$, stage 2 is training with $D_1$, stage 3 with $D_3$ and stage 4 with $D_3$. We call the training data used to do CPT at stage i, $D^{CPT}_{i}$ A CPT strategy for the legacy model (agpt-7B) can be found at the end of the document. +This document serves as a **strategy cookbook** for our current runs. + +Continual pre-training (CPT) is the process of training a model on new data over time while retaining previously learned knowledge and avoiding forgetting. Unlike fine-tuning, the goal of CPT is **not** to optimize performance on a specific downstream task. Instead, CPT aims to **retain and incrementally improve general model knowledge** as new data are streamed, while mitigating catastrophic forgetting. + +In this document, we focus on two CPT approaches: a **data-centric strategy** and an **optimization (learning-rate) strategy**. As a result, the following components are **held fixed** across all runs: + +- Model architecture +- Sequence length +- Optimizer + *(although it may be interesting to explore how changing optimizers across stages affects training)* +- All hyperparameters except the learning rate +- Evaluation and validation tasks + *(these must be fixed from the start and remain consistent across stages)* + + +In what follows, we assume that the base model was trained on dataset $D_0$, and we denote subsequent datasets by $D_i$, with $i = 1, \ldots, N$. + +Under this convention: +- **Stage 1** corresponds to training on $D_0$, +- **Stage 2** corresponds to training on $D_1$, +- **Stage 3** corresponds to training on $D_2$, +- **Stage 4** corresponds to training on $D_3$, + +We denote by $D^{CPT}_i$ the **training data actually used for CPT at stage $i$** (which may be a mixture of multiple datasets, depending on the strategy). + +A CPT strategy for the legacy model (**agpt-7B**) is provided at the end of this document. + ## Recommended CPT strategies by stage | CPT Stages | Distribution shift | Primary strategy | Fallbacks | Notes | From fc44d13930dc6760f1f8fcfa0018b50e9a67b09f Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:12:17 -0600 Subject: [PATCH 559/590] Update cpt.md --- ALCF/notes/cpt.md | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 9689ad6b28e..d8364b8db2f 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -38,7 +38,13 @@ A CPT strategy for the legacy model (**agpt-7B**) is provided at the end of this ## AuroraGPT V1 (Stages 1 to 4) ![different stages](./assets/cpt_images/stages_training_initial-1.png) -For these runs, we have 4 stages of training with the first stage producing the pretrained or base model. A key element here is the learning rate scheduler that we are using. In fact, as opposed to the legacy model, we used an infinite scheduler where the LR was warmed up to $LR_{max}$ then kept constant before cooling it down to convergence. The main advantage of this is to avoid rewarming the LR when doing CPT which can lead to instabilities. Therefore, we mostly adopt a data centric strategy here but might resort to playing with the LRs if needed. The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. +For these runs, we consider **four stages of training**, with the first stage producing the pretrained (base) model. + +A key component of our setup is the **learning-rate scheduler**. Unlike the legacy model, we use an **infinite scheduler**, in which the learning rate is warmed up to $LR_{\max}$, held constant, and then cooled down to convergence. The main advantage of this approach is that it **avoids rewarming the learning rate during CPT**, which can otherwise introduce instabilities. + +As a result, we primarily adopt a **data-centric strategy** throughout these stages, resorting to learning-rate adjustments only when necessary. + +The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D_1$ has 2 Trillion tokens from Dolmino and fineweb Edu meaning the data distribution between these two stages is weak. We then have $D_2$ for stage 3 that has 1.5 trillion tokens from math, code, ans science papers. Finally, we have $D_3$ stage 4 made of 0.5 trillion tokens from reasoning traces. | Stage | Dataset Symbol | Size | Source / Path | Notes | |------:|----------------|----------------------|---------------|-------| From 8a28d36f36b975e56c60d8112f05307f5e5aa5ea Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:14:27 -0600 Subject: [PATCH 560/590] Update cpt.md --- ALCF/notes/cpt.md | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index d8364b8db2f..6a17741fe0c 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -54,10 +54,20 @@ The dataset $D_0$ for pretraining is Olmo-mix and has 4 Trillion tokens, then $D | 4 | D₃ | 0.5T |OpenMathInstruct, CoT Collection, AQUA-RAT, Llama-Nemotron Dataset, GSM8K, OpenHermes | reasoning traces | ## Data centric strategy ## -The main thing to figure out here is the data mixing strategy. To avoid catastrophic forgetting, we need to sample from the pretraining dataset $D_0$, the current one $D_i$, and we also might need to sample from a buffer $B$ that contains data from the previous stages $D_1,\cdots,D_{i-1}$. Which means we need sampling weights $\alpha_0$ for the pretraining data, $\alpha_D$ for the current dataset, and $\alpha_B$ for the buffer dataset with $\alpha_0 + \alpha_D + \alpha_D = 1$. +The main thing to determine is the **data-mixing strategy**. To avoid catastrophic forgetting, we sample from the pretraining dataset $D_0$, the current dataset $D_i$, and, when necessary, from a buffer $B$ containing data from previous stages $D_1, \ldots, D_{i-1}$. + +This requires defining sampling weights: +- $\alpha_0$ for the pretraining data $D_0$, +- $\alpha_D$ for the current dataset $D_i$, +- $\alpha_B$ for the buffer $B$, + +with the constraint +\[ +\alpha_0 + \alpha_D + \alpha_B = 1. +\] See the figure below from this [paper](https://arxiv.org/pdf/2408.14471) ![data mixing](./assets/cpt_images/CPT_data_mixing.png) -Note that you add data to the buffer B after the current step to be used for the next one i.e at sampling time, B only contains data from previous stages. +Note that data are added to the buffer $B$ **after** the current stage completes and are used only in subsequent stages. That is, at sampling time during stage $i$, the buffer $B$ contains data exclusively from **previous stages**. #### Stage 1 to stage 2 (weak distribution shift) ##### Strategy 1: No replay From 07170acb5740501d98032227dcc2ad4dad495c21 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:15:12 -0600 Subject: [PATCH 561/590] Update cpt.md --- ALCF/notes/cpt.md | 81 +---------------------------------------------- 1 file changed, 1 insertion(+), 80 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 6a17741fe0c..1e51d1f41ba 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -100,85 +100,6 @@ No buffer data is used at this stage, $\alpha_B=0$. ```bash python3 mix_datasets.py --input 0.9 /flare/Aurora_deployment/AuroraGPT/datasets/papers/papers.txt 0.1 /flare/Aurora_deployment/AuroraGPT/datasets/dolma/dolma_v1_7_file_list_v2.txt > ${debug_dir}/Megatron-DeepSpeed/ALCF/data-lists/aurora/mix_lucid_papers09_dolma01.txt ``` -For convenience, here is a copy of the ***mix_datasets.py*** script -```bash -#!/usr/bin/env python3 -import argparse -import sys - -def parse_args(): - p = argparse.ArgumentParser( - description="Mix multiple file-lists, normalize internal weights, and apply global file weights." - ) - p.add_argument( - '--inputs', - nargs='+', - required=True, - help="Pairs of file_list and global_weight, e.g.: --inputs 0.3 f1.txt 0.7 f2.tx" - ) - return p.parse_args() - - -def main(): - args = parse_args() - inp = args.inputs - if len(inp) % 2 != 0: - sys.exit("Error: --inputs must be an even number of arguments (file weight pairs).") - - # Group into (file_path, global_weight) - pairs = [] - for i in range(0, len(inp), 2): - file_path = inp[i+1] - try: - gw = float(inp[i]) - except ValueError: - sys.exit(f"Error: global weight must be a number, got '{inp[i]}'") - if gw <= 0: - sys.exit(f"Error: global weight must be positive, got {gw}") - pairs.append((gw, file_path)) - - # Compute sum of all global weights (if normalization across files is desired) - sum_global = sum(gw for gw, _ in pairs) - - for gw, file_path in pairs: - # Normalized file-level fraction (optional across all files) - file_fraction = gw / sum_global - - # Read entries and sum file-local weights - entries = [] - file_sum = 0.0 - try: - with open(file_path, 'r') as f: - for line in f: - line = line.strip() - if not line or line.startswith('#'): - continue - parts = line.split() - if len(parts) < 3: - sys.exit(f"Error: each line must have prefix weight corpus, got: '{line}'") - prefix = parts[1] - try: - w = float(parts[0]) - except ValueError: - sys.exit(f"Error: weight must be numeric, got '{parts[1]}' in file {file_path}") - corpus = parts[2] - entries.append((prefix, w, corpus)) - file_sum += w - except FileNotFoundError: - sys.exit(f"Error: cannot open file '{file_path}'") - - if file_sum <= 0: - sys.exit(f"Error: sum of weights in file '{file_path}' is non-positive: {file_sum}") - - # Print header only once - # Compute and print normalized weights - for prefix, w, corpus in entries: - new_w = (w / file_sum) * file_fraction - print(f"{new_w:.6f} {prefix} {corpus}") - -if __name__ == '__main__': - main() -``` 2. **Start building the buffer $B$** in prevision of the next stages. 3. **Run CPT**: Load your checkpoint and run CPT with the --finetube flag. Note that you might need to convert your checkpoints following [these instructions](https://github.com/argonne-lcf/Megatron-DeepSpeed/blob/main/ALCF/notes/universal_checkpoint_bug.md) to a universal checkpoint. @@ -199,7 +120,7 @@ If the loss is not recovering, sample from $D_0$, $D_2$ (not the final mix after Start with the following candidate weights (some exploration may be required): - **Mix A:** - `D0`: 0.33 - - `D2`: 0.0.33 + - `D2`: 0.33 - `B`: 0.34 - **Mix B:** From 9cb3e8eb84f203ae2a99490582f4ed624a1b9006 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:15:56 -0600 Subject: [PATCH 562/590] Update cpt.md --- ALCF/notes/cpt.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 1e51d1f41ba..dbf0cce3b30 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -165,12 +165,14 @@ At the end of this stage, we have $D^{CPT}_2$. #### Stage 3 to stage 4 (shift to reasoning tracex) ![stage 3 to 4](./assets/cpt_images/strategy_cpt_stage3tostage4-1.png) At this point, we only have ~6% of training left and one should start the final decay. + ***If we didn't use Strategy 4 above:*** 1. Try **Mix A:** - `D0`: 0.33 - `D_3`: 0.33 - `B`: 0.34 + - **Mix B:** - `D0`: 0.5 - `D_3`: 0.25 From 4092a6f589c4c5cbc9e9f5c5b994dff407feea8a Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:16:19 -0600 Subject: [PATCH 563/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index dbf0cce3b30..e3772ee9ae0 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -172,7 +172,7 @@ At this point, we only have ~6% of training left and one should start the final - `D0`: 0.33 - `D_3`: 0.33 - `B`: 0.34 - - + **Mix B:** - `D0`: 0.5 - `D_3`: 0.25 From 2a0517d57b1dac1be033402bef7e0636d171509b Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Tue, 27 Jan 2026 19:18:10 -0600 Subject: [PATCH 564/590] Update cpt.md --- ALCF/notes/cpt.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index e3772ee9ae0..9b341345c45 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -1,9 +1,9 @@ # CPT -This document serves as a **strategy cookbook** for our current runs. +This document serves as a **strategy cookbook** for doing CPT for auroraGPT V1. Continual pre-training (CPT) is the process of training a model on new data over time while retaining previously learned knowledge and avoiding forgetting. Unlike fine-tuning, the goal of CPT is **not** to optimize performance on a specific downstream task. Instead, CPT aims to **retain and incrementally improve general model knowledge** as new data are streamed, while mitigating catastrophic forgetting. -In this document, we focus on two CPT approaches: a **data-centric strategy** and an **optimization (learning-rate) strategy**. As a result, the following components are **held fixed** across all runs: +In this document, we focus on two CPT approaches: a **data-centric strategy** and an **optimization (learning-rate) strategy**. As a result, the following components are **fixed** across all runs: - Model architecture - Sequence length From d3ca3a4d551ac977a642e529f74824c9d1a07b5e Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:41:34 -0600 Subject: [PATCH 565/590] Update lb_optimizers_settings.md --- ALCF/notes/lb_optimizers_settings.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/ALCF/notes/lb_optimizers_settings.md b/ALCF/notes/lb_optimizers_settings.md index 706b2aeafb3..110580c24ae 100644 --- a/ALCF/notes/lb_optimizers_settings.md +++ b/ALCF/notes/lb_optimizers_settings.md @@ -55,5 +55,7 @@ For the learning rate, we implemented the learning rate finder routine [here](ht ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 --lr-finder ``` +This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR blows up and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. + ### Maximal Update Parametrization/Complete Parametrization We have MuP and CompleteP incorporated in AuroraGPT. Will add details about activating them soon From 915cada62749c212e479bb606d8174909cb85a5f Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:42:14 -0600 Subject: [PATCH 566/590] Create readme.md --- ALCF/notes/assets/lb_optimizers/readme.md | 1 + 1 file changed, 1 insertion(+) create mode 100644 ALCF/notes/assets/lb_optimizers/readme.md diff --git a/ALCF/notes/assets/lb_optimizers/readme.md b/ALCF/notes/assets/lb_optimizers/readme.md new file mode 100644 index 00000000000..8b137891791 --- /dev/null +++ b/ALCF/notes/assets/lb_optimizers/readme.md @@ -0,0 +1 @@ + From b6631886250a39dc6947de575b9e7c868c680970 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:42:52 -0600 Subject: [PATCH 567/590] Update readme.md --- ALCF/notes/assets/lb_optimizers/readme.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/assets/lb_optimizers/readme.md b/ALCF/notes/assets/lb_optimizers/readme.md index 8b137891791..cd73cc337a6 100644 --- a/ALCF/notes/assets/lb_optimizers/readme.md +++ b/ALCF/notes/assets/lb_optimizers/readme.md @@ -1 +1 @@ - +Images for the large_batch_optimizers_settings document From bf3c5ced065b303cb2888c2d06652f05b1885e2e Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:43:11 -0600 Subject: [PATCH 568/590] Rename lb_optimizers_settings.md to large_batch_optimizers_settings.md --- ..._optimizers_settings.md => large_batch_optimizers_settings.md} | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename ALCF/notes/{lb_optimizers_settings.md => large_batch_optimizers_settings.md} (100%) diff --git a/ALCF/notes/lb_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md similarity index 100% rename from ALCF/notes/lb_optimizers_settings.md rename to ALCF/notes/large_batch_optimizers_settings.md From ad9269595623cc60d218cdb4d983c132e2fa9768 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:43:54 -0600 Subject: [PATCH 569/590] Update cpt.md --- ALCF/notes/cpt.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/cpt.md b/ALCF/notes/cpt.md index 9b341345c45..26f3811ba5f 100644 --- a/ALCF/notes/cpt.md +++ b/ALCF/notes/cpt.md @@ -30,7 +30,7 @@ A CPT strategy for the legacy model (**agpt-7B**) is provided at the end of this | CPT Stages | Distribution shift | Primary strategy | Fallbacks | Notes | |------|--------------------|------------------|-------------|-------| -| Stage 1 → 2 | Weak | Naive CPT | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | +| Stage 1 → 2 | Weak | Naive CPT wih $D_1$ (no replay, no mixing) | 5% Replay: $D^{CPT}_{2} = 0.05D_0 + 0.95D_1$| Add $D_1$ to buffer $B$ | | Stage 2 → 3 | Strong | 5-30% replay of $D^{CPT}_{2}$ and monitor loss | Use buffer: $0.33D_0 + 0.33D_2 + 0.34B$| Add $D_2$ to buffer $B$, you might need to switch to LR centric strategy, see below | | Stage 3 → 4 | Strong | Cooldown with mix $0.33D_0 + 0.33D_3 + 0.34B$ | Cooldown with mix $0.05D_0 + 0.47D_3 + 0.48B$ | You will need to continue decay if used in previous stage. Play with decay function,stages, and final LR value | From 2292e2c5fd20fa8463d25838dd78e6f638fd7572 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:47:28 -0600 Subject: [PATCH 570/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index 110580c24ae..b94a68b76b5 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -56,6 +56,6 @@ For the learning rate, we implemented the learning rate finder routine [here](ht DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 --lr-finder ``` This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR blows up and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. - +![lr_finder](./assets/lb_optimizers/lr_finder_example.png) ### Maximal Update Parametrization/Complete Parametrization We have MuP and CompleteP incorporated in AuroraGPT. Will add details about activating them soon From 5551ac89a0e9fc14dea1ff980a545abb6c247cdd Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:48:23 -0600 Subject: [PATCH 571/590] Add files via upload --- .../lb_optimizers/lr_finder_example.png | Bin 0 -> 1086596 bytes 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 ALCF/notes/assets/lb_optimizers/lr_finder_example.png diff --git a/ALCF/notes/assets/lb_optimizers/lr_finder_example.png b/ALCF/notes/assets/lb_optimizers/lr_finder_example.png new file mode 100644 index 0000000000000000000000000000000000000000..5e5b195ecacd73f096ff7d75863106d0f4d495cd GIT binary patch literal 1086596 zcmeFZby$__x;;GAW%n`&1yoclK&6pV6jYQD=~9s{m5_#I2cf8-k`hiD=?+CeL_oSj zkQV8#Z#;;5pS{mI=luTv_Po|5E}2Z;_j&FZW8C9;A4*G#uU$pIibA2RJ#+fx1qx-) zNeYGT#)@V5m%eo8NBHuG>4`HJSKwv4;_3qmWe4TV$>SF-y?!(}m{p9}~Cw%CO&%xXtCe{I9rne%7s!QE)i2XpzIBQ(--oh5Uheg{|$bN5dVL z$(Bk7?BG@A6?_)D!*9nf-5nc5H!VK1qK2uCx7YCCjm%_U&+h}g9X{8ZzmHhmGv&Qw zliXyIn1B~9>i>SF+)?fRzu$#-_+7gZvgH5eosJ*!gVz1Oypghs#e@I<>5com{&SK4 z>q7l!CI8puq5LN%|M&IzPj>$w>qGfZ3Hm=)=szXse_x>gl%W58eg6Mjf~dRA|EHyS zl3u0ZAb8;5!J%4OPF-m)tM*rPTNY6$yby!iCQo;`bx*iVn|+kd@49v{yiH8Ip7Zf-uZ(ZQb04E!UEbcs?SukID(xBkn^ky%DVgDEO)@NBq7 zrck5pP+@h9mIS#!8ChA`uP%&&)*q?DGhMu0fzMjLKa&y~?rh3*IDTyYLz?H#`3ecz zPq#94#9uEv&yzD!rQs{HkwTGPAtf+y^2^I3%u&vlv+w12@tGdQM(_%#Bp5!FU{B|y z-?q)wGC;~#Xdu>i=8%ivXf?SfiXX4q^_(pOr4sCI@kexuylA-%6~#{B8zSQ6Skwqs z^FRLO>dF6PE$6fBzhqebZm0c7m2XTM@h$f%kAasU8Ec) z%R-*b|9IzO`_Wk69A}=!msPori(g zqtcEB9B?cO2npG*udjclDZ{R$ynOT2SRdK)>P!cRA5EFU*d>3RVxQ~7CF0`6K7s*p zal>C-g-4@Zb4tXSyeB0b{mV;B|J2NNdKhm|)gI4*>l>e}%Jh-Jwox`GdT7}5v1LvV zrZwu+QKft>+h6JUYh>Efxnut*3}+G(u4Fu-nZJ4prT;Ja2<0td3Q@*Y9Lv_wZU_)} z#os8D&h0+d-62UMl`3^c!cC8jNECb74(5k!I^Fq&qu42r$sw(|BT_m3K(6z)D&whQ z!7)|a!DqgPiFSTNuepts1&}elF?L-+9=T034NdU{+!Qq_BO{}5#pqq`>>7;h8RJ*V zD=LB{KTBlV&s^zmet*QWznPZbOxN~1Gt(Jw?2fc_U4qf$DBHG!y$N&Gnyxv1$8Xmf zF;V=OgCiqdWCFx;d<1O6(?H($nd!oa{_4c*b*7QMAE_=AFd z3-tOrqWN#Txw-AzzyBffQ{0tN&_6J6H#RunW?$od#=PyO=WK@qn4B~0rgHOyX9qg* zsU%AW@}=yLT95t2W_6a^&kp(K@nlXErg7sZHNSnhd%fi$i}t3aUypAWusqyo-4kKi z8((#0sJ$>+GtY(G9K$h-kN9J7*1?GcA4zr=p@D~@Ov~48r=|K>>)UqA`hM;Vb9Cvj zX_``>uFITJQQyYmZFb@Lix;QHem0N*O!2RY)yZ43Uf?TR-I8T%9(4FRWT+(!jeKHu zJ&DSb>T?94&fE4}wsvIABT*X--_g_FW~FCj)YX0XZ?EqX7oL=)LhkPFVtRUd!`&4JU4$k(_p!1%Yr635C0X2AMlCiG z6q+4y6`z}(2{38OU~?e>O!<1(ur_Z0fddCx?$W3~P|&dBLcUyJ(bCRyVUUrR->Rjh zb-i3hJm<+y(JS9SJw0O9`Lz=Z(JerJn8Kfg49;autpC$y^hdry$77A@D^9dWb_Im4 zH<=pk?X_pU`{2PgleFlDV#m>U4s)|Fk7$aeHa(x5BNs^mC*5X@?UR++p%*V+TyM1M z2r6n8RgBk{`%xQz(0rM9)?Z7Ox8M&n2&Sp(yp}tgBp=|Pa&?5KYRFo*l*>vvyV?&2 z(8*hUz@2<}xJn{hWt0ZfHXdWA z+}vD?V^NWjY~9`6SI`Xlrs6eS1-T^!N6JNAIf{6hCr5g!j9H}*%?=6Aam>y%%?b8- z*BPg>rrhlFIlhr1QZ+=93H_pO#A(rzWgquby-Fi>vY2A7x8_!-S()0lbNHU|{&4@D zgM%2Ot#foBGO#1^-YcvXhVg|)o7&-ln7DxEv~D3;SNVBZ;r`GsBhL!_g;ww9i> ztjOFp`oGvp`iy{_}X963fWq!+79Fl&^Qdxgp|$yRHq-KZU1}4bZc|m|ECwuDDsqQ}>Jb2KMdNU7s z-5TVfJ{jTLhjwK9ZMDV&;khAU*VrQdgOWlM-#aTKRQx}?7z`hx6Hy)+85uwz?UD~y+(-(WVN-^ndFf6>*uxhG6>>knILN%qUnS9a zfW={u<@(R$j@;x9pKVlci`SVnU1s_Q2AcTvlFCZ=sv?Dq!56dwme<7OkTW`RN=kH9 zYB$xo*Q}ETw3t#CZ^n-I(6r{cx=!_F3P*X96mNciyH#p#Sc>BS0U7*{+{)(@Wno7s z{m0jB=iShkHYAybO4paV51Vqv%d+L+^;$h@jvm!}tH$^TnwBBoIa{;7MnM_Y`5(L= zhHU^aY7L?KJdz8O&v}1)xh%_`SQNf}2M<1uiHVtScAw-2{`Cq8bH`dA&Ux!RsBCn`Gn2x$h{_t#5#f4Hox zr}t3D*N)qK+4S*>iV7JuH3m63xs4l3(8UNkTf-zIkTKoHF5cIezKOV`XJMqE{MdIoU-`XCw=vc8={| z@#pte6o_d1nMs1g2<{PcWn{R$Z0)0-s7xUOFikQhw7XC45H+X_KjTeyLu476(w6lC z{UU1^_&JS}+uX+|iX9l2uV*~xEyu}iTu+aXW93(`3!E*)33YI|`+zdZOQkBK@bCud;M zE#}bxrkQ>2Y-+E^iVz{1%~^lZZAK`2H{4shP1Q4UNl0+;Y45L>hkw*?>^=Yb`B)Kh zvlSci*^j`{&eQ)Upzbh0V#>(LZHd(`e6*S^X$M3a*s+pYjQqc>?U-O4kR-!lphI;S+_kB{s{@(H4~eo(2;W=Z8>;Br+?9FPtR7=RtfJ#uR_AY z9_KOI93q=OzsaP#JbP4$hTei2^u?xwym2ypMb>gXBO_-NG}(TaW)diQ z-~l=VEB?Q- zrDbHE^1sRMLVVT(-n}v`1c7O^=}SLHh6au;9uoi3G%0t7!!%|N?3MD_t}#34(fWQl zlM`<0t>j{RS5kU&Xf(DH{zi+o7Gg>^f_HIGRg^2K7o>I+xU))lmwoGPOh2Yw1Z2(TprD{vhKqVN+djV-Ks(rdEE7Aol}+{K58Fp| zgxZv^v~{FWx7Wy4-&}vM%2>Ur z)Q|dPj%TcnhhBJi(^XPM-mq#oYD+V5o|!(O;P}n#|x>x;@?yKE~-irm_Q5e z^Kr6ihnR$fMECAyhl=v@kKgm3Y1SuQkH)5z^O4ot{#S%bV(y)ln_Ao?*d79Fl=FEP zVKEs=X#=*YT)TI^gQg_e{7hyE4GSA;eT~;582cLz<)h1K0c_(Zw0Gb z3MVcA5xD%~#fyQGjuvCU3Lp1)sQyW6PfAKkIo~O*PfvDUAFa{pNC)ZgSxD-d>J)PW zRr3nDBUXdoEqfvoFLZ3ai@+QIw%&mjN&VP&`a4UAX|o--N#l0m@%QVG*1X2rK7aoF zYRpm!#f%j&G-T8G5^kf~&D*#QPkWbKKJO#IXFdAk)tKa|Q+I_A;(ouL`u7|@n`+Vf zNGVo3=hIVW&5`bk5K})117rbsoj}H9MU}iJq2pU!- zd$Xz}9AZo9+{F`tYe41ocV|`EDkV5_aj#0*+w+f&8&b4lmd3dX+Kiq=2k&;Woa#+% zMHHX&KEAOw^`;47f#%CH#WgiG4`J!dmu)@?^7Y007*CP#-1IoulN&!Z_?$OC7Xhtf zMJS#eQ;93Gp6UZ4I81mrJbcP5PH>fM$L1guV0IGNMExLtH(z)1Z+@&d)j-6TG4ARgBAmgDdl>!5ojO2{G?7gGE1Z5FpD`oV4Cf}Us7N1#ZN5= zMlJF*$>1~2m}tk}CA{aGhsa+)YGPS_WhS|Afq2ol;e`sL^_5*nkQ{C?u+kP&h%Uoh1RS zzO!Av{-pCrmV||eJ3VGg+2`XhJ5j!W!+PQAYvD?4PuPdmPv4qhk{FED6v<6TBXpP4*XGfx1N9#IP?dL`}-c2k8O z)%w~XGXJVU&Z8~%Qh(jz!y?oJ3LUxL^07M6M5Ehlok3%o71=rvEh+C4G-aTBBL`^(GmR^i#!a|zmoueYE8eZIOHH?dFX-&e>9 zrL-lX&iFB@zJ%O&h)@S>4We>hD5%o`X^Io!=ohTGJFq$Zw-;c=G(<3$hnrY(pYNCT zfAeOi`MS(+;6?*8eGbB;w*tfc>YkKOih#vmy!betMG)>xF#Wj%$IAfgFJ}&MfNZ&@3850kO;%11%3`onut35=9DK9TaQSP5A zudK|KQ&8LDDm1Cv{ao@gIxVzFg^vkFbr+~#?+Q$I1`y13HO73TLO~qz7m^+Um4FS) z0b-EV>7VP=D?NT*#`t`!mL>HJW^wmha`m-PO?B(1K_hGQ)~Wh~EF{bo3Mc*IFla2l(L))fC$wNG_UbodhGvYc*u8ISY5rcr21AA1v<8iSfR zs-Y^9SlFuq6n-5w0N381N`OKJGBB^ZB21QQcEBWUKupzsqAeORQNG`75uvGHd<2g@ ziy)dB0g^i8mN#ttg{+$Tp=#&VnPhtN2guR^QrLocf4Lfl5Ci#sn!gpRBdz*x;J~jM zO_FG|w7?=EvVPeI*x6kPA0U*G%`O%{17Z8;xBE;kCbq5Hh#db`^4sCG>A`}5_$WJ8 z?1=~ee4}$*lQ#JIAb~UI5=igB+7%+*G|=Q^6cjx8TW`Jwo9X`Jku>5fZM@}i2*DdpK7V%X@&_w5U!1v+W}I4|pMOcn81n9$ z=CK)L(9p7*o11%;?kKXu=_6(FqNl7IKQ7W*ar6(K2VtkSx7Eq_=te|LChOf_z2#K+ zzKUC?H#0wfdS}n7#W7lc9RBj;{E4kQceC*EWeQL4d2t}7vC4c8pX*ewyUM{kc|GQ| z`FtHIMadV{)s4e{afI$=8r0v~GFjH&ukPExv0ac7-11tlTW|4Ghtlr%Gkr{55Zwao z!PI&I5UcL+CNki4aS1tJGnko zhwo3tH`CJU1}@D~Ub%9mWYje=R6qq~<=15fe(PF3l$hs%AY4>Z>KdjO@qX^*6|UoB zmD|O&y|tr*3XpsChs|aeVyFq)Dj3eyX=N!!ETfta+C= z@AVVU_MR^VeDhMD&tZJ>HJ=+dZg3jaa%!ft_*kA05)#5yh7X~bg{7p_c5&HP-dei) z`;BBT^(E$C&`{!x8~CYJ{nmM|OjTg$_~+mu5Umg?;hAy-(GdDJy+Pq?uEz+XB;B^Q zw%13s9v$W2@K(OZ65?t8osN&=H1pPh9VlURDH(O+{f7@95MnGCB1t}d`ZS)2 zRXI)?Lx?S#HwXIp?O3{O*@1ogPQQQu9{ta)pzVT{RWej}Bzt69ngDLH3u%eel|`0a zkd=M$>XlMd|q9A}|Vg+f06y)&S!;_PYtw*Z=);!ryM2+caZ+A&^$(RJ$~ z#byMhoT6edrYp|dc#68@r1N&Ouo!^tBMK|ro;_=`NcG0DgO`3=YTjEbCulu<&3a4S z$j=59sE1X9Z3XNpa;A%hLH$1G<>h5rP|`}uq=X}$Ki^EP8b8xpm#BbpYM#HxV;?`i za!E;venseu6DLk={_{`&=H^RFmoG;TIf1Mrw~32%z)DplOU5g-I86*9XpL0QsjK?Mqfd|wpil%Yn!MO$No0{n=Z=9bzuNl zf!N$kN7ps{+uC=j9>G1_e0zO2h%WZEXxFY?<-lqqU1j3Y(b34{XAsznQa91k24`i7 zAZ>z2^Y|@(j6d^Lf!H@QHFj{%o>Letlq0q;N=sjRpZIm>nV1aja7G;F{dgV7nE01i-5vh`N?ke8F~TBclb!BII>;M=MZBfSxeUQ zrr)cog~H)&Wgocnv6E9dxE=LoqgCGNK7hT7uWv-RY}|M`@K?9pH#@}z*oo;x zO^hg{TF&b&Co#@Hk5*RIlJj^3f`bf37KygAcMq_#R*0?_{J8DTubMy+VNj5fd6PNQ zYm#dX@+pl`rPByS@!-y#?=6R*JaIKX4>YPv*w4jP;l|w^1rVZ(Ndl59{)VY(ym^m2 zwTjaH;>(wFf{=6Y&kxoy@aux_=2}N(Ool#s{MdPfVtF4;`2xVcsrA}Z;^I9BMmg*$ z2BgE7DCt42)vt{UDd9sgq31SyNm4aJDOL<6r3_ljW_o&kv!c7F&z&n@&SZ;(56)|8 zZH>1XGiGLGVqy{mDi-^dI;PkQZ}slW2Wdq?-74o-*QM|l zG!#6UE>3;P) zzG~e%3+te=@$vCs(~hF<`s7&15P=vO=;yobQBJz{rnI!wOTCX1HQ2iI&3(?BKUMGE zxx@R7+fLWaZ;iB67NydNxvb4?qWR`ZDRhLn;a99zJ|X&!zvoEUfxxV&+T)FvTv90`rmx z0OWK`^CuzmU%A9VY4n4X?tsBGL;BpGG&EEQ2MSP|p85LvM#N$`Y!1HC`RHKe*H%~I zNXX0vMP7WLi;5D=U74&wS}M?G)X;8RdL1W20s~K)WK6t`ZiL!kUZdqf*Fy?f)~g?l z=_+(|bSdB6B)rsfdQiUTAdqty)F(xQjfIEEQt7)`72k>6Y%uqL%wzmJslN67T~g^L zEnc!@3oWf>t!vK5k6r^qYHS|psc2qoi0kv1u2o<{)idnbmkHdR@-wBU9LgG10oP*w z@gEUEHr8=y6cnYKEYBVq;S5EjR8_;xrU*C+_V3&Gsmr`b3F-uw*5}QF)BV|#VBo_L zWOX4oNJhz#Qt2u@oAg%7mrX-MgIe{gXj3TDBF4s%NHjssbO`rnz^FJYmt2Yfra}Rh z74w8EBl$I_omcq#sg1bC` z;y*eOoQZ{r#C0%7I66}PnbX(UkkeeaR=_BJU_e=9W0`^F3ov|fsG;)e>W!xo0R?7f z$1-~mbKLLf=5z4cqs+2-j6MH<|NMgs!B7DnYW|~IhNr>o+3t{}Bt@8~@Mi@^RMPx3_`YwMEml#UdO#-hX~5$S5&KJ*dbq)<8+_)ntE~@g@WvyOKh@FwY9S|@^#=oe*8EoO=i$% zd!Wyjqa-lZ-TYKkR2ipNHauJ=RVTD$*~5og4f0h&u0?QedPvL~Nl6+61P85!IA5+?N0fyt$Df6=ea(99!uykUAV^;eSjDVbz1my-l-u{7o-j}x zEaw?4i5!N*-s<_mnCnCOQ?j0$8LLP()uC2B-A$H6Inme~1Y+Se>u~Jm&vNK&P9sn^ z*P-*Gc)hE{pqN1A)&})->oacm*3ex^Q)3nqYK-CTW@o!(C~RLWqY$rKa^m{+>!e|f zVA$6c;3`xKq2R@nCw^p&A@JR`wYBw9mvtKkuP}g*OET(mD)3Yf)TX1ADybF`m}wo< z6r`+`B4KJbs#Mo8+SjD! zLg@%1m|tl8lDE~fpQynG7>`(kifrDrsdbp&os{CZYu`@HfWgZEwFRVF3NX$LxH?=s ze*CWho8P}TV+?#yjnOsl^g*zW^@77sz%^}l@7}Fg3LcZVI|N9myE4La569?_ngEnn z#nRt36Xdo304yHn)=QhcCMG9CVKJI-xF(GPgtJ51mH}#EXh0EFo*T~--rn$j|5%el zCJ2zzh`Q|n3^;PwcwI$Oyg<(C+7V%0(Rqwz_nGX20eEY;tL|Xppbw>{yCmR#glU6u zA||nP$qmLT9saXQWHYA+B{b8K+Eu9zRB}`F_r8C>N>%7TfG{jUfhNrw8`TY&iMz7k z%a?LU?|^9J<(NMG-!*}?d2lKDs~9Ke5ESX?0rlDg0R1iIZryM`eMPV_H0U7OmET?Z z%Q$^RUVIE6CDKp^=mXGt^VbGV2f}xCVgLU9l2LD9#yCkh7>23(C>k$NVjv$gs-XeH z>_?5Wo3hBW$S)`;^PL@LNuTe5TDF6)@;jr={^k0EWx=^z|(PudOu&y|PDk?5@i zJ1iv`bQ6wD4CsSv=NS3Tg0=EiwzjrHaA%jO4fZ{I_H5kc>}6}|IX{62d_X;k15XF! zWK%qM&WX|L`$oa>=G%{4U8^9*U%YUkBm!0kd zc6)v7W+a3lFMF|xRYj_epmg@Md1%OE6u7;i4N&6+K&RO+3ISfMfPXK= zEQ_JxeN~kLeuXb-67V6^mZ*7^XV}>kdLL786&QSAYv*la5KJ&QTBe1e(EP8aP%Kl? z?NBf)6xn%s6_JR3NbVlHi8hbV#l<|0$euC1IYG3uRu28$d-sMWB~`}=g>Txt`Kr<} zKE9e5Zppk;0VUFR?%e6b==KH3z|wEZ}pw4X`l6P zY`$>wf?8ybs$lKrtUT)52SdDv&B8m>=+^h+^sh;VTN0Sapg~fHGvMqZu71*`Y{cHj}6&A0XyD+~+W-f^kE< zmJb8^_Xyk--58U^8foMDgyyDA>)PMU@?LMbgED3uP6HVS9pAAkx4ojnG>@gB zMhsdCxD+nLvSTa&^jBP^4GJn>LN>DSIr<^GC|`o88kbq8bWUF0HoGNVL*H&B4_9T* zRy?g)Sm)+KMAitQI*ROv4@+a{!43GF4jvEVNTr(vrX=>mngBwHSKW8p4d%1B$6xkl zQ%eaJy}!;q?(yTt1T{$1jz?jAx`35X{S=9D1A4p<)DJ^nL4(Rebvl(xNq#{%a2f&m ze1W(NU+!n`33--ml<%eu8~nP;s9`{o?!_^luCA_x#5&V+eNMp8v&d|?Ubg&vU*=o} zK(x~cTw>g(#Kh?5`HpT9{4;tB88v(G$dMAro%u~oO`UvVjIQDYM*i-(_oMz6)9xkR z0a&35mfhx0)8*wCiSql86}rZau4@fH%5zW8o=>*I!omcHTBEI3K(7$D2wtCJ*5yAw zZW(XV)cAA)qCrRD#|5aBrH!k<2$&JwSjf{9j*dofP9uE(*eI>Ey1+A=FnsD27Z$qm zaYbva{&s$|(@nNR&RTYg_g9NonZPh8W`iLoPwf&mdBi)M=}i0i`D-haf#u6EnlwF50$4X9ZZXGp6#MplQe6l^UQa&Lc3ljuUA`EX z2S3F?D(CK_A{p>LE}K!qU#rTVKI|sxBhdB9aN&pgxs#KVC&yb|okympr!9pTgYT0} zlt_WP@+_*)m0hYdKzuzUvRt=UZYVt_O4};w_-+Jh^+Fp1lPir83jFx-WI#GJ1=`+4 zdrv}FX_}i&gobD6*)?F9Hab;{3cPv4hD+gzAdCzGmR}Xui_DO`Lh~xTqUt?xrF+OA< zs+@cy2;#2O2tX#+r7Vi3pb3!+QTpkOGDd3+#D3wlEM;N)oWFtZUqu99>3Mzl@~vJ| z4|a`6s10=;3(8nl@fj(pD#+g7KK=Og>9ieUC+*F!713X@*AW1#HA2T&Sj0)m>oxry zu-m}dL?JTgS|jHb01`+~BSasuT^<|I`QmQrT4y&)MLY#iZS(4DUb`hDJZJ4u= z`9(p8-_L^2pG$$au7r2Awg&v8_&VNvH>fBEW~ya$dOnk&w8V}i9Tzo&gBV5$v{sxO zsKP|_7yO_&#)k{{5QD|Py{>)bC>98Sf>6g_)^yt zn&N3F|-_tbKR+cCRT4zy%aSsUt^@BorH;u1zosBPoYLS3(g62IVRy=~_MK z7WS>5n+Rj*EY2g7qrG7W6!eOB>c~cDrib*z#l%7}LDrd`z$_>j{Kox@dmv^?7g6Zl zM=m?#4wU9W1R(aI9-D)%d79HbpAgK>qm+X%!zh6YBp9h9u#956-Kwh<36jsXCfz1| z^ff8JZ5V;X+va|JxQPgJN%)Gz(_&+L(BM(kYWkTY-5@!XMJY6r=7K?%UtaoNAG%Rwm#&>LB zpaj$MOZkRpK7(?BP#o&Xrf1+clSUUiq(`%5iyo5EJubZl<_yeyjKVuQI&`|KmSt^5 zgC&&pDH<2+kbI=7UV&FiCm08J?G(m#@md925mDwE*RMw-!L+)Zp#ACsNV*pV7og4r z68V3SucoF3h2&YAsq{^vQSSoKWP1sRk5yKcM0|{=_$_5G2fSve9CB__ppFbzHy= z8Q>UC!k1739R3LI?9>e?tWvir)Lk#qkZ94X3>ib2zZKM29#x9fTc*l+JKqJ6{5V_U zvODDaSBf6INx4i_=U*OvuS*cE<4ro=YkpwxhxGa|_`yPB7c0QOGr|oNJw(Y-ucBi{6FA=Q~?aMR;^qaZPb`z-ec(bn`sDZfy608k&WPht78OK z$S^dOgrbqH#R5Z41FO&ZoWF3v4Xi=;ABnHqiHT_ZjkdD^g|G>n|Ld>6#5^hCFWlYV z!a1o}$#nQqmXa_P>YzKL1Ca(vIYV5^t8-eyEQFno3GXw4`vF5?!#FK(Zo6&1Dj+{;UGB$c9OX8uU4Km%dz<F(Ainr(vB{BOa}t%X%zEui?@)Z}D0 z&aEB>H*m%Lu%H%Q8Caa3T8s%~oR-Z+1%)06($V4T4dO7?CZhYEUZ4@Xp+X+SM+pQJ z%;t0m!1Vi;eScfeA`9K7jP%INS!)$-(z)FEFf2)I40Uv*1QfyxQbi&Z!{-qMzY2dV z7EK-`H;AMAF;K7XQ0x2EFCZy^t|;>_F)t><#+95^0fqQx7{C*D9j~-v1R$#c3k!=% z<*3*^0>A+3ISNb3-oTZc_R{v^;xD3p{(3tl>(EqV;cuJE!U~Wt!Ar>81jtQA8}x1e z!GoZ43*YX~M1=}LW2)mHfo3xIdLm_8bDdfB96)g37>N$-HCmWd2AmIb zHMa?apH^w-&YfbOL}!3I;&*NQd;7>GD)77>{5T1P95cBG7nbcB8**(JUc734Q;aXk z2+q*EtI&QdwJ8LI2%s=BQr^}!&1h)h2mE&8O2@uiq&jo6;C2XL(r}m|GgQ`|^>jzS zdkK%l6tKzSC?ad!9&jIsuW^V)r|o>DF{8wS1iGcvq=GMc6N7DHo~oSmBNaefo40Jy zn(%^X>n&haI>={U0o1OSMpCU6O{N&99JTc3aGe*S7oa>gIWZio9Hd)FYnLBC3|ZCF zs&&b{foP<^3SA!@QNf5vWzxaVN=hbR2~nOes;L=;S3y=X?`v#OTt8O<@<$l1E(aHv z96Wp6|HncmV&Q8Nl$z|uy#JBcez_1t-Xf^0x!H-t0hVNIXQjfNN<^jg4&%2YIN)zK z1*}uiS)c_XqNZ?CK()!z<=DGd4E@_1m)(OQhGL~LVhP7c&DArtae#EI(GZzaDb?P$ z0_t*C6G_7C5_W~Sz~1xlGeu$Aq?rzaxR&ndN5%E;o*~qi0daP9E2^EdLcYe**Q)64MPvNYaGtYd_bfil&rCo@X|MDeG63wpy^GFqK2HtB~ z?tBa1lg31^Cw;6j*1R3(3Fr3j+V!FG)%dJ`IZ=6x)G;K(h8cSP#-VFV(6?cb(~9G} z{Nay@@o{w6ck+pOH%aSnrJOaPyklb?$G&4B&VQYy2#A$FYHxqqKR)12adez>t$wiH)Uo9%|Kd{ zmAX$CBG$8*Ijy{>&8YnQ}mL0E}<~yg?VIH6ryycRjTS zXaJs86bWd}q$dBacuqq$BMYG8zbqT&D+R-KT}&YpzVKgbc4ReO_@Z2V_>f@5+W>N8+DH?dL5rb9Lg7N3Io^AV0yI7pzG%^+6imRKxfdy#;QSFLc5> zc(%jPEEmd^Ec_OSs@pSpD=*H~U&~(W9v|?3ph`;f+T2g1~FfElduJeNRtfc zZn^WVh0@$oGqz6tzpmM@zl$*3fOWLyC<`-lb}7$q9mH{N1j=;{cNpY!Yw$%w&-+Y0 zpoye#*QG7^L;lfqzdo@ASqFXLsND{%1t6v!w=&Dx*I_nUs}~M2H5`1T88$$%GqCh5 z1A&kRq(`tR!Cuvkna!SFq$bg@+!A#yad_1t$kc=*7!3){49nJOFGR73C?rne_-Jh) z_Dhq+?}*TSNbv1VY|DO5&f#13y|`46P7}+3xdUu$#2p2>jhr!Zk59ffe3xU!L=R7n zShZ#iXYC(eUopch(1DpyStuYf6(`i3c}PXwLPN8yZ}GK8OgmY7$2v>0OYQ9{fxWo& zNg?O%n0S-)|K!kfdd2&h50bW3wX#?h6 zyUYr?s%Y!m^Ds-*(PL_FK(-DJ3`7T}19@Q}CsDgr2_8OD66bN)w%$dOutWs3aX{cY z@aU9hIm8nt{0x7?lyu8h+6#qyUOT9x;IShMEVR?P`S z+NXkoSgqeKyFPb-N(d3P+=mvnR+VRE>${4ao2dYU+ z`+=S+l^TB{<~jJuHPexQUEN_(pa@>@1#T0H~nUPge3IY5<1(zTb!`gTQLc8D` z)9)$`QjWfIm&|328q<{U}Z?4WH$w zuVTXZ-B#|SmmB9PfM*6JhUf5%4Y%$1wTY>zZAObFq0}WKG|BWqhN<4N)j8J_rYYrg z>-YgSO2<_lov_f*&oQ$&O1MiygDr2|6i-({Lx$f3!8rmQp3WpUCkM~N`1~u<^oQXg z*tl_{dt9PX-P`(lRpf)5zJ4gI1t#GgJv|0p(Gt@yGcz;yWOn%44w9#M;1`is(qXSN zYv%v<>kmik5XkJ%TQ9)j3-{}bXU_sK97o9sgXIfS0zc+;xNtnA0H%`zxF2GGL1V-u zM;4xIfPqLv5dqTrWpAAclZ2@-7|;ea&1_Qp!!l3{SZyx=e*i$6Lpwy>*e@)sfzO7) zF=A#PJG%m+px!tUFdd+~t*7T8WEuRXf~F=7cvr}8f+z2Us3;7Akq;iM;4*1cg}X`? zu>eRN1aecLzImli(A&5Am{j2CVbnS%p>T2+TepU$rKy7OV)6^pFJozG+1qpjb&PfbcV3sf?2LCc-K_^+YVudvH&0Du` z8)S-O9#U@a@ZjM?85ElGNVQ{)U$(4OhW7%g;ujYeSL+~*12yq;_L5$_J3R{a;NZjO z&u8IZsGOZzi5s)DOd^Ic)JUIE^GGP6kDQ#$CfmOWU}iIKN{0AFK8hH4Sy|7c4HH+D z0AsP$$l}F|H=`EAdz_P-8;WTbVFh6CJ7BbYY`l~*ordySca#cSqdVW)Ev`+AbikYs zV%SKwTU%SZcjz6=?ooDZ3GGKt6q-i6=`%6^v20l=IBh%=m+L0`ao(vb>OEI@`FP8U_6rrbAS zV8evC6qIuWQ}O-G%%Zr7uo;{_z<%`Lf%Mg@fkb(Q zW=8~6g6(m~UKx79qA>QG0pV+DYghU(J%tSyiHV4v>l8}PYYh<;-nCqhJ82g) zvq7Bn>?L80u27_Qb~g%MFj1JJ#EGrZD;k zv38p*E;^2GLqy2BwEK+KpAwXKZ;~lD+Py(R_?Cn6on2Jl1TA?{X*3dCY=yl>@#oK> ziE2W>rXWAR z1gZ!Xpb8p-8dNu&5g@1C;Se@()l5-kgL9@rNPGi^&#-;%KYH{6TI5BDN)@n+AQ%sD zbIT(^O();J-vtIOjoil~!^U?6HrZx05fl^BBQ1W`hoI(E_-I@o{C)=8 zNoogvnk46YVapdZ*zvjl^It9`KD8+;&5=B^5Asn47I=;LLm{+R-2*kwh1$D_TCvnb~|Hk(Rq!;}7@uPMqPv4&D;DG~ZF$;%Z@Dt@3 z3TF$gLWBiey{h(9O0jwL#8quA_wsn^?X5P&$UkQznJblU{|?MUzclb6IrnEHBvyE2 z5LLnF=F;XD0BxLT^ThFBVgsb8k)aA}z3YI5XuMh1g@E*=Bwkp4KkIJNxryg#=)+VY z=84z_Nh$?YbF%mad+vU#X#X!gnG!sNOpAv(5fn!Dn8goiR4LTFqEy+T+)6iCr~Gk4 z9QKj$c42MEX`qqr3JDw*ynyL=QkJOTX!hfc4j9{=^^08l|~uJY+#mSs9H~0J282xNsAW z6RE*)ngT7L96mHGqhh?p?Hes0R}u_H5E>vhM(IL0m(`L+&qZUrpOsb4-hP%&m>o%t zOq7SCtpc+}JjBYC=Doi znpi5-t04G^!{E4~G1~?~b3Z2B#8H8D^M_)p3q6S1;4465oT!32#ShaC;L6hHKmlze9 z+)hsn`rFwDcadrj@&Z84Mqn}l!g!QVajLB80Obrg(PT~WTEGoM!ykHM`RvKx-B#B! znM=WXOb#UOk#N_=!6^(1b*Af)#ALl;^`6dOO%CE*R8TX2G5@>p%m7t^P#4IILVcQT zG@+*YMT;W-w<3l%k!zgGSjAv)4nthf49Wj7FJPAcU&a^`kfZVGpXPE(G z?CGW^SKhLv2DfkBG6P!=!~2mkWvE~cevEl34aW25gdo~x%#5lai-yuNGAq59ExCa$ z$kO8|LL}sdE~u^FPh1>%P}fVC`!IYnzk$K!%ik8_PhQ?<^d&nN*9eO+4_cjCikXb6 zsuvl5Vb(i%GstKEp+j)+P3gP-g{O;YkOv6%XpY8CeCuHjj1^wqFG`vtK|{zJg4FPa zN6yYKqN4gv_5NW){A5sxm1C|xRTq@>GG#yVY zHv%97gOc4K2}bO>K(|AaDRM)#%etnj0%z+@_knQ3IV;|4uniAMC^QhioQ#+y8hCCh zuta-?fOPRyZS8Jo%hBOX3&R;b=$DchHi{_=6N~X2jue_fCsyJHbv3NHK$_c4YlK1tl2jIA=H;y(q#GfJ!h;hV ztqXKbF$N|QBRI{~xYb#7!shCsXVQdoYLsSP-&xJq!7!2yLrw7%ZEFqEOdrlNG0(emU-^*#%ER1hm9M&oszs!q z>?+-OQY7%fvZY%N?eH(%7;-5Z4W)Z&>FC}U7rUOZIC=7q?FU2R_wg9;ROD8Qu&6u{MTRS z|NkFJ_wlf7T->zx3SF@A1E;iA|J|8Y@5Vl_(t9Xu0> zffCw+A?s6j_j?d#Rnsgt0?$gRSIcj~k$ow3F^6-aqDye%cqbUU|g8ksp zqZ1b|ZiYwUE_haUbu}%X*KlrK7I3==gDzJ zFLYkVZmxqv&Y?@xL!}j1;k1N=sIqeHGp4D|0N29+k)55Lmyikm+fN?H=IqWARX*xt^)ZV08mvL0145X@%_t$SJ9klqYj5k_Uc%Eh%{qkIXA3N6|YsTXWa`Xh@|&SBm$|3op;t%_tIg6+$f zFS9;1Nc!Bs>)q?^;BGPLPG8?@qi71QhefABt>VcEfNi?H<4w4k8 z4dk(3QkANJawGXo$pEe2Fwmw6T5ffn1gMzUGB)+J2gm}MKCX;laAurC7nm?zI-OzP zGN7}Jfq`KSqrgicj^JQO6x#(XH{ihHExUF@+y%tZYd`Dx@KDy3*`TD0GFq)Uq6C{xfQcF|A?r$06d42rSB>_yx_zDeDv$#@5_S3jMcl-Yh~^AXzXJ+6A<5 z(o;=Y`O*JJ*LlbF+_&#PyFtU=Dai^EA(4hrLb5_pq9|NOsYKC`J&Kgkl1(C^p)Dec zkU~oop(QD*-}B|Zuj{^lkH`J}?;6tQ^M1cx=XspRd7Q_|UyyeHeyg8NQP~J;Latod z514&kwl2jBXl$9aKFDUnhLO&yN?w9p;6~dtXlZGc1KkEJ!NLyrS-G5X8hp#hJ9ljT z=X%+$TIKcqyMt;TC;^|iwtnjzyY%lra!t_DqYCfjRP3f5s5dE}a^F25FfchMXMUAA zvFoaNjA@m*-kzUk*Hc>MZd^`JC_p@g4Rwc=0UWF!tyA?_CtDWW!Q*Vms(Z6 zqE97LdkIcq}`a2L@K z7dB6Z2X?x2s7Zw5Q0URa2QjgEti#^NWxO-h(vKgN$B(bisQ)3fem$pbpMd#UC}vjf zQzVt_fax-b-Bf|sZL^+cW*ThTH2Q03W9ta>7;44V$u+P~JKbQ@_?lNcm>YoH^Chv@DZ$Uc%&)F%MUOsu)h3#+`d~+>Fxw% zP1ycZUFeZNZh5#LCC+dPmnYkwqWd0=9RA~$@SPEb3k#Vo4_s$5yc^%t`}py0mAx~y zovS2ZCqx2bv@@XpMHMfC%T)Z`v8EFd2TCd{uJgP$-(wC}C6z3Ho--$S%0Y;s6HDA? zy}jq!gHb+(vGjKfe_!`J?0qaaSkB$uorgpuI+k(Zgcx%VIC(M!k6_t?Zj2Gi{%&eg z=-2PbD$kr(fbc$QYHI2j;+@#`s&1@*aEq|YL{_4jXS!&~lD#|uWYPmCPwt0BhtoM* zdRoJ=LF_be;6QW{i;j7ggnQhNy#~7`j3%!&*P)idQ%3{Yzd>BtbH+w{_UxHdP#}C~ zd#qM%*mCEu!ri?0AyGnQ#9y7`7%Po(-3QYptf<;%eZ!`WKDaanQXTtfc7!%JJk{L1 zvFvHTwDff4kt1ahu>0OEYr;2DJfrE;r;8BLZ{6>%-GdKPJ%N~h>SY^qDc3>ZkR5%D zRl8^>Mq|PtIeGGAA?57c!JmPRq&IK+@IjbZFSRJ!i;Z4#W@blOSy>N@mP^CB2VbFe z??t#uuK8AQZ@)M&0|rP?K6-qR4h0SW#E#R-&Q7{Zmo5&Uo_jl;)4UWBA&azdrm?XB z$GG%8Ac7&aNv*C@IeOqS@PA_q^+K#{_m6kRCViiRs0-Pzoy(cGisRpWIXswSAz7Y>6?G1AMSp+NvEHHB!Q`v!#Ff8)wYePK-K^i+c$Y5R1Y=v= zd;5L+_RZLK=3aIR;5##3S0m1`VmAX?VN^&SAAMgvCMJB3`WU=v>(+E})1I2lfM7SS zFP^x$RFY7(w5j6VyLW*FsTh5CLnHN2jVzIHJ1RQ*J_T0bhb~i~sSeC;|P zw^Y!Tr?3T-7JCuY^I80y&UF7oG3Math|5|}V3}m>x2e)*Z{OS0^@UGyt*Wb=Hgl#V zxUOB6doRr3cj5|n6_f63%f=Tt23<->=uaDj@rwxr8Ry)FQJ9kS>f5)kI^9CeJKc@f z(WcikH}6Xcbp(gcxTi8M%+-$G`o)Y}yzgcY&jpch>+i9IVbA{k*1kT|rt1)jLo$(a^^OUBc>%KG1MPgsejbTKHKaQKuHH?(E`{{P?j9 zcL(g?LxO2KGsqnKG@^T7$d{>Gw`x{uWrZ&owy8L!o%!93t(b1!i4*qe)2G*1cily& z0V3%=59O|NwK9K_#*oNvr>E@bjlw5|Qr8kN0s&ARFpQOsCw*dW`dPan36v=mt!)Skgl zL4$9iUOVq&mT*8I(qKi=&&Cq4OZ)ci*ti9QbmRlMR_vN9ILo4~`Oc;dw-@y)%V+j4e&QpPcH#`f*n0xHcOGERCeTsp8d$XSO=a-!w3AYDE_{ zGAc@+`)E+9+dL2e1x6Klnp(p-cWt+GrhA#Xcn{{O`u;r)3}8t@kdcuQMPhu#rD($9 z&&Echl6MDmF7t8Lty7~i#h%WnXYbx6e1;Sh1TKpQjtaSPLxDq-1fhQKYq-;dy<<<>f=xj9hzp9r+cp~>rZ$Hr{^ z+B?Tlg_C}t3gd@qFK$P6cDCj1>1Q!)G2Z;rkODI2H+`ScPX=%2xkv})p|xV!Ko5_y z%V+oPlWyC#?ErOkdEW9ziwiH|=7X*-5rzsFU^)k!I+dd_rJL{ZMB-aYegG5d} zn-)L`ddm1Kw;sK3vlF`9100u!~{$Qn%m;5-^5)FKG8X&T7%#7pN zb(1U-yEH=wZK4<1cjD%@xTwWf#NVP87C%UQe84*?E-BILaGmF~R)L!a5M}Mh{!hCU zuL3cYc)X%I1i14-S^?||qE%RNpaAB>J5`ZBnoF*W6@2u4wl+{X?FU<5uB%lZYQJX9 zR1*_<(FLpVaZTKVn!?%DH5D*OO}cUG*4=PPl}pzz=is9?IfMedVBrzMw0LsBpnyqE zBm1R{xJlo0-bZMDS=OqAWh#$)K%GS>9?>`sEAO9hHX-mz0>dg!=_^WCIPyZmLZS&SN*U> z)WL=oH*qU)rWN0!z`bQhLG+8K_={~s4&vqQz4vy()Nar=#e@q{%mA2*zI^G(!{v(} za*EazsTLjdZu604`S8wNi8}W=AH1m69XfPa+G9jmW8=2dT8`4Z1$<3DTFltUWq3e) z=>|F;0AqWkDq;3rTHzF4gqaCY>3FHeO`Z9R7yAIUA$wOM2->)Lvj-`9hlQyw`E~-x zjT+Jmp_Dj-1aaC@9(V4H@7PJW@EddEb~=qv*5#>M`c49osm6H8Z#CO7+4&1YuW7=x zMOQ=6*(+f?$?{<#F?09t+ZX0Jdr&%r7kAhACKvnD$*PIWW43#uK}H{YSbibS()YJd zCo2VJ9qG&?V1>n^s<(VfQ33|dT8gkhdU1rgmH!G;QxCMyjoR6*A*+mQ2O9@eJ# zyn3e?^FyE+=$(+ecY7kVncH0dy@ho$8#cL-RF+>}v8<~6*}(G+4(`jFDE_6OhdOIJ zfug=U{g^n?{H_G9M+VcT>8kowM-0UOXP7uX5PbU3m{zv79mqBLQGR|8vdV#jPvRcQ zpkMO1aA7b$HXaBO@!sFd#Kq*?IdY0qyiT0xA;4{HtX71@@&!-^_%KUCin2UxJRIsd zN6S%4q&Qxkq@3IgxQa?GkB1o|G$v2dz~cgZ^O@$mxGK${de-o=?l*oQ}A zbw+x(f@rywFmL?5kz#v2_Sr%_uu=XN%>vQi=dD*xPlHC`0e zxLKq?a#FsSQx$D2H1NK7ovA+rBlAYcF6!Xc();w0BZ3hSkMAer_4ng^O9fMnot0g! zZ3*QB3g-4y2{%x*egFFAgwr74AM%*}V=i!S!ZPQzb7*Q-bZDC6e6%Dza0Kq<2cVOA z-By%tEbVB;LYyuI(WHorcj2h8>+jFF??;Pi@>IY9|%Q04|+@W(hHpmL_>Ed z{vMTH4$1XAjDzC5x9WUoMIjK4aL7m$WNn?CoV0o#X?4z2igUg0&FzB?x*=*qyGn!c2M!x1-MxGFja#;)V*Kzp z&OaQCWiWku3L%Y7M0~}@6SOAK+_TfAL%?KL&3=0xm~7di!CBf34OJCUvjVy;5nuZutm9yV+UkvTbI&oDNFt)B9psnAtZ^o=MSeX#FQ6rpMC z_`!oa(5yT=x&4AYxZK0obqF+0Uzbk4VsV6|O~dDK@Cbk~v*t~-eFu0--J61oBu9uP zZdt!wI(6!V7>l{A48NA9>Xs9P#)O3CjO(G}x@I}S@tkj{H-u{Z248~L+v&%TA8d4y{+c?#7haIvbG^s{JM}_~EG{f0;Z^Y+ ziikmGiEfimRJ8NfYXvdJri^y<6~eN5q#(^L`u6P*;(}sKO7sY-E{lDSf!akq2H6{b zMV-vlfgp@=i4}-joE~7lbFQ-2k?uL!*-0>pn_x$iqq<4Aqxvc?DjMK?J63+%ichdm ztuednG1l+aX$W7g2= z^&Ae(w7}FZ7Ktt6V%-L>T)SpXk|^X@@L&mL$Cifo7JFl`aC-gvb<)e1-O$*@R{;Fx z=|Z@f-QB1JJ+iT(W;MUvi6#ztXeC4993&9N7+N$|5VykO;eM-3@r9hed2u?Kw5LeR zp+qcx|6X+bh}{Y_>EiJe=kL0E_pYEN|J?msLjj-x+Ad>Y&{N_bt-z~NkTz)JGUE2- zTkqIdlXLazRfAL0mWyv9%=KFuZ-_D*+JB>ylMgjIAxjX(A|B%&NcG&|!tKjwdb={^ zmCi9y=-KnNP0+063EO)yLzl)6pL^Be*2P$2E&Eq`!aLq|YtpLJ8byyQDJx43So!eL zqt1@E`MW_n3SGOlJ!d-f^SV&NmO#v}UcIubF4H@XA`60BO8k~TRjGmjL5nMIwQ5z{ zxpU{9@6h?m@Id4CbV%6VC0pqBYlKC8b28$B;&0z%$?36!@9#7p`2e5WjR=Pd7B=JE zsX#8+$s&Z>h%kRx4d@3eOUs?)-LUxh(@uk)S=`&!jm>#G_wJRVpml-HOHZFJkfn9H zzr@KIe`^5_B-59>su}!OP#N7&!ML&gK!z*0E%>kyH=8Y6W(}sbt5*BAmhK8+L!Gh{ ztg&Xq02d`VGUEhQYBs4LLrJUQW5;%7Ax~xvm6;S-S2Yhxu>AQs8NV;!p-HL@u6NgG zuvJDD7E?8kQ!{h!TwC93vRsmOJ2o^h?98I1#uK8i8m7tIa{9wO!Z72vUv!=xu+?IWPh0b<%%czeYy(=n8 z0|Ti8NJ|8r%q-4bhfkep#3^*|zMD<3mK4ijTaPKfay=A}yxq8V?_J7+fGCJ>Z#(AH&G7EFx?!BwT|!D682!WOTxYbE zM2sZfw+W?TTxTV(TLFOWC}9_3lV^ItWJueVCenz)*u%G;Vz1#LY#OTG^1Aq4v+1z? zU9!ec=v2_|?|=Ebh!xmv&o*2vQfb1Fv)|Y(f3*W^_g+lQ?#0|zy2nWKA`3cl~7zf>f67(vFScUpA}MSX^z|L zgdMJ@CO)iL%Zj|+1Q<~-gQg1{1w`9}XqZp0k>|DIrGOO=A4)kKsS9Wa&ST>5WV6-4aG}LYYrZpfKJGcP*?)x%yScB;;lZ3NJFjhWD`7G?Gl3H z#!Z`&X)3cO@X7X|fSQ^bOKa;kfUJUrFH+zPApIh4+rL{qCeJV0;B!KP@enRFH<+$p z8c2BFf(2=DzD99GI8m)HUc4BHeZr|CkAAUH%lpfIXz*M$JzwqS(L2~&rYxpfva>D znV9T^yBd8)*9&cMn%XJZL(TTE4jvzGXiJ=y58^Qdd?djs_U`TQC0Fsr(#7~L=iB6N zxQIf0`1uAHFu~Dl>07{%_^zE&(ei54d-kd%V-&ClU115-T&cy3b4>P}4GZ)5)X~xH zaKZ>YKdowlYFMdSURg2!b47*d#=XwZe?D13g7gl%H(q>k_~J|N@(m06YdcHy=+Of| z;*P+Y_zKC(5D!f3dR@6PoFpIc)O1K!mK4fCAsuBihlBrXa9%aYNF-Y7LO#GC1~gV; zrWSeYmVusL>q;&2f2nSgsE?76u55NXB(LWsM$<%_*YJn8Opp9h*AJtVQKe@Wo&xpj za(5%OpA(#-Q)y0RsGO?Hm;NVIX!p5xK;>`9^WPyO^XIQ$sccW7=e6E#{w`iF5#c>g zXDVD_rPi=Y$aI$Iw4*~Quc(+6cPJd~cQQ0MdF)z#zoJ361#=7(fUAnX5`To7R27kJ zN0<}c0|xjrw_6)73^N&0^z~~eFiW&f`L4+MrlzeyRid7PUWl*Qb3Q)4FFPEP30WqK z7e9zQWW@AA5*_AZk`MSc?$8T*YIjGsK8o)0O61=G6St_= zrVa}y{79;*s-hz*vG9(8Y!F>=frhT!lH%?DWExT(uVmx_H4&-DVTwR`mhj3yi}#o6 zkXh8N(>rEpFY7mEmE<3KXWWaWhK5qec2pi+<-mo3?;xvEvRRtL8G=IaVV8|0Z|%K> z;RnQM-J$2jM$1#%WLwn8X7B=n;29L%5z*zfb1(!gT-hWV+$WW z+P8L!1eL+)U|lb301_+}+%UKIM@QX9wEqCrN^8!;&W8L!9rg{o7d8OOtzU4S1Ki%uPyesvI#Ibr)mPX7V8x|)cy-2p)YA&acHFo4|ZLv*E2 z5m-veE5IXvOz7nNdR;EU5|}3RRM?@)VZ-#1TVd~~uAw0c&}(`3D=Yp2uFHV%X^@?C z>Zd2^qQv2`-Wtpw9iXLk((v{R>IJEj=sON`>QECdB{Cc~+`wGIQumLle+PjDXWCb` z{`+74J0ch~y^V8{qpTsCC;b7m^LlZeY43Um2KIvHbk3c(gkz{aW{d|NsP)gQJJV5U zNBUc}dUbV#vLT;Oh(`x)nK>OTmr7jT8ki17=XV3qZsi?p`RiTmpB9nL_VD}oJFCO+ zpVn~?a_Y80fMzZPn(vy3u+B6+1p7ZvbKrt$L|svYa13*75--T>EgebVJ2nwfZ!#8C zgC42f$gYc1N6i<16`epTOfWB<)PIbTwe7UK&0;nW>Q_EnQfYlQT1jzYX}}dMybGn= zQOs+a-`jTFsoQaemJU)wS>hFS@!GWy7Bj|DY_LIjPHyR{p8*%S1Up9&~U-G16 z2!`YhN`~o3E-Q!GRnWr(8`?#=kD?kE#5%y_`4)v59Yj$9rJu{oLkJ?qH>H%;{Ym%D zA`9bB>-FgXet;qH6UMfv*h25#4fxbi`LrnSa4Wa;wOF)BHGh(hX=i#p<&(5UTNq!G z!DjuHE{8$6Fu-AKE6M@m+H;qo^0QyQT}a+B+_Zt|ZcfnlBWrv8sg-SSpEwF#;AeiR`o|BkTTKXG3KlL{3Jrz>(N0t} zwGqlocoKqp7Y!Z|S=^x^KoJ3tP=D!e8TBVwJ=WZkIBfj*9x(2rK(8G3mcWqw^r!?=0Y{W>C&#axJhj3+5<;k z;P@hi#-0^ed|l2u$Av3a$bk}>I{+7^gl{{NiM)V%VpgF>)vwT_s6gnrmRfkvL|~}M z?N+^Xec@Ge+JThCzFI$*t>FtG%1`Ck2W&7K+KsV{bUI6)rcTy-C}TK#cDL1NsCF0x z4s0Lm)`Q&b_saQal@b^;60(-na#ay$eJQ&{U1*Z*(fOArY?if8!%x)0QdfcpI;kN zoR+1LCDL3@dkO)?WwF9Lj(QdTKdf_aI2Q#FDEj!ZJ&_k2*DBv|2E5b9tl&`(E3KEA zELhNX(xgexrp&k$21-m~qaclY-j5e@4O&SlZ{D1GRWGAL$s^Qvu;P`&e0`xtQ6B7d zI@mo>-Kuq^w4IPR!V>91yz>7y{ra(9+;qBw@mDsYdmRUnW62QEnsw_0z(pY-N+7dX zhoRT*bU$KxDGfTWn$P>Za=2UKVtZ?A{oq{>y8y>BgBIpAZPN=!oUAlQB|Fe|y5PE~v-RIldsvUGBaoc044%fOGzmMozcU1B(Emi5UN`m~|MBx76oYI?$>8`NHkGXH0> zk4Kr&UWx}Gg9vCy-wPa7=*U`45- z8b-Hc>K`!1EM#_?adSN(Ttp@MAL!CUL7^2njNN8c;7g?}F&$;&noQMkD<(@O%g@)B zWll!PHH`8|LDDDJJba9m$Z)Jgl2s*>_$AL#vGg?U(iZgUm;9#+D% zqqu=P`v5oYylvY`ib;=zfI$l(!>9pPB?Q_pQ}N4)bgo_rtFzOoTL$%Ybg8Vr^*zJB zGIY#PMV!_D)MClE_z89Ik~?6S@udY~2dO~o>(|vCQxB|J1kR-QVrQ+cDBN4E2;EW+WR(vAU&7GN4-GX zq?Mf9k+g!_@R0Y=k$A2fH}}%@>)fhSMLr>h)0~|rppn?r(!7Jj`X*+}VBCtb;&$iX z-SJBkYJDS(=7NPP#RwQSIyi`47{^vH{t)f85zmz2ThJp9uIz0$>Q8P*r<)9&H1d&} zUfOfIR$Mb{>I{c9HZs~%EbV{-x0WxEpx+lTWyiKt&G!!M*^`8E;&;{U`|%^}L&v+CKqz#vT4pmPh^VN0ru_-)z9N~^++GQvwZ6u6iKXR$b?et3_47-O z+ZT?cA_<(Dpd&REp*jOU$h|yb06Z~op_WNZQn!e=ebm;t%t^4^u$&)*3Yw}rz2Y2bLVDh{whou z$460j{JcQ4j=oiN)?08AMVX(ux85!oK?JdbzW48;7Cg)MRZdQhweKkg18{>^&HD$) zx@i7ioY;4)3zG8kdV#EpEv9V-c?x#yyxdEW;;wu5CihcLh3~6is;O1MxCP72%z9!} zkQ_IC|IF$Eftc?J?c<+fecGUYu)jp^QVY+E1NdY-_~*H`i6+aI?MIkCrf|>PKb-K> zr-wdGKW@6U?(@P46DAaK?&Gh>Gn_)hjr3p;C{)4oK1LekR+fZRIMrxF)^Iv!K>l(u zOzC>iOQQ`EJ-fe$5x{j9j2okQvrd&Z@1@bYz5bmO)W&;@X$#2bNpOS%^Bse`ien6u z_1&d>(4fJCQxU$={-=U71;=q84>N?J|EHI2NHZ7W{KVUOtnj+@1&;J*=EKgu zi=;~sLJ)x!3JMCi1en5Ui!0202`^kMXA!caPmz5k`2Q1NC9b%w=k?P7R4G>~GIgXO zgG{F?BJ?Kedns2}*QK*+{%A^oOK;xp=o)Cen-3yPa=$wUus6z-enJTei6l4{p$V4b z{Tbk()ay-dz-%sZRYvErnwm8c0}X}3 z8G&lQsr8*j&HVZEu%DrhpTp<{Jbn6fh?*L>Wn*^*g=7`3U8smUe(Urk`&=zt)*q2D zqQUzP>EgqA@|<&eK`#j%TTU;73F0a-=YL-%j=u_`(D>%LndMo94Abc&Q)u@@#h@FC zT~yLz&WzshcEb4v(_Az|Yu2tEEfh>`8Yd@WZmDOUFD` zEumz!PH=Ng+Sv-8+BstRy?r!%RP$E2H4O4IA4?u^iAepJoDu zQ||eC%^g>%qLJfnc)7)qCrMjbfp)lcw((NRO*%-gYzjKWL6Fl-IpXi19=A`01SIHg zeE;MYby&4ifRm?PBb_{@c+JXHS2oE%luc0 z`%l_i8@K6wdl3hNf=D{6-o&( zibmC%+Dm2dI6K9yR@Ikmm=7undmchL_UIv!++X+Vx;r>HP-nT4#KWHS#v6T2l%&;H zxDJ6OA*T3&c-}hjt%wlGSwF;C=i)z(+ijE=X{Epax&{9KBK*RwXxI(H>Hn%OZIa5~ z!!BIBC@Bv3h!GzwrVV1|sF(ygefI25mBYfE*{v*Egcz9>EX?-;9ngLY5j0Y_f`f;lC)w4^JPzX7E%|E_`mtX)v6SaM) zjFgm>|FzK2U679Ij|Z3FCt*89h$Ay?TC1)yV%f{G%6Mfr{r(og+P9rzmDlPmG$q{^4DQ+OICVb4XUX`%wsZ zAD<4ecrJ_QO@yFfD)4xz?i^FCnIE}XLcJpTGhjFNEe;ql;-FKfFlFR)bNY-+2;uG1 zbLc^JOoN@>TXK9ZhG4*%hR)kVR+F}Qhw|Mv*M6M+Ot#gAe-97Nv$-*qWCc8|nL!}H z`YDCEZ|BAHWQYl(xuR0pMnO-qVTDLY1UQ%6BVH_C&pb6=j>4DUbauh88(eNJ`Ra)b zKHR|LpkSPeEmfju9&xEbn(^FvA#(ayH3BXA0mL(0o0SleLPZ5OoO{*w5yl=#aPtO% z(F_Rcp(ybmhY(}iYzknaf&eZ zn8~yKeIQ=|AE7Qp+(i}95!!s;l{iqu+pGH>Ki-Z~T?~{tb&E!*3YEQAAcdxu9+bCU zFi%4BArK}(i(JM4uk&K1zOa$);ZBWjQoC*uLOM-&ut2w0B1=IhJyvVULc1IW&NH^< zhCS=wz4I6fRfo?n_8M+=a2UHID5bLUn2oKijK6R_<;L&xjlh734)X96aG(M`c-h$iB(xE zc;mQaK8id#mI?LS<@NK7H~>PZg{r~#$HT5BKx@1`%6|<|G8jav9i6doNY~7WE8(hY zYR1{xv|_idSmJ$S>EmOJWFXD>FAUgV|9|ErBV*$8G%U63}Z~Ub7sgCV#q^)AzH>NfLI14yd;|0s@u|ez(2e4p7C<&m6U)v!J|1 z?QK8vke_cBNr)fX z$ET;zN4K67XSAQ;)S8F^!pL0|`jG7eyIF#_6sf4U$1Fp`WEi^gs-@F$#Da0zjSK=n zM0YZcU&oMB8YLLl#EOa1czbzScH;z;ZvFJU9o40k?`%n{ft2oR6A+S#CapudBHj@H zx(E7~7L?F~gp4R=j=K%YWaFFkpTrOgWk}~g!?3q+Yl>+~T@9&u?PxV1)wI$Aa{I4M zi0w&yX#R00add`_Q{KI5GmmH6}chCQb;gSY>aofy4@i*-toi zGOVeupRsM57B3vnt|ZuQ6G2_h+6goCyI8>e#^;f*T_KIQd{k(CY2TZEK-I$iCxV#&~O3ObQ=uXjWCEmp~+&jh|xn@|V zlGiy%ek(eX@%9(*;RRQsz?xY`|aymeJc2hendH92) zJ83wIg7V+j=J&+^i;}p--aR3@WzL<*x}LX;Q{Mm;6)&f5$&5{}q%}*>lHfp1V|pKe zUyRj)foRP49`TB~URdd>h-!048BUEpL>1^9Kh!-`Yz5;aJD{KZi4TQy?h!RgLA;qR zJvm>-Ti{jSM;i_{yhSORT{cfo5+V8?T20SB0*B|^TA~+P;q+2Bw{ranN^1~_kh+Q| zf)L%@+*~N;#S`(ml|DH24R%4IG@(ZlVMDg~wdL{m@k*W2x zId>5yKiW5wQ;r>lviR!qdyS)7V!*=JE^KQBq5&Fxu<#U1D=@z>eGw=Ic#vCue0pX- z-&&deK0Sn)vu95u9$+&ghSNaOseELe-&JiOMnlh?^A>Y5`E^TRi6ZgYEg09f4!y>J zi4y~z4rroxyuE&5ZP2jnFdyIs_-BU16*p1b!* zSzf{p;bb20R0!80L?1Zp)_V5*`AXj_CPT_KJoc>x<BUuY183b{wg0M|bw?YON!A$; zeJ6K1A2$6_zVh0$>kV$NlFCar)3bW0nZMw2p`vWUZ<~%UcLe3`D67>vROWoBY;V>t z>$te567u{HtBA|^-DNh-o!fTcj=3*=$fu*TFz|u zT=SUC)5Zt$n_dkvTi(OtWCDF0H`SZ9xCQ89b|oEyIn?XTme)+2Q(?teln zD+bZu?F*r)U;1O(JUS-jm~*^Kx5XVx5{juuY^Rtq$`-Q3OqI@|%al%QZ~rN;VUw9e zKhA5zw20NVwqhoq{MIG*c<)PBucq!4j+O;QNiZ(1- z(=>9~nxppiAE&^x^ zN|@Vk?bI@tQI2$Kfg9`JXfAc?qPM3n09LdJxg_Gd-E*#P%`v>sRJ7NIbLl*zRaWgI_RRr6sMeWyNRQkI@$%#WWxtE~g$r~9>IIfem4 z#o6nB52q0!JQ+|ZG6{BZ@%4fkxuq*_{A}8h$B_`v1hSavEukk~wPC;l!Y-5Ydl1V% zzxPH^N!AFK^zO_X!TSY7caJ@IJq2CN-BJcq&qB?Z-;}gc;=K^|^zKv#Ga@1e4`FAg zGUUc95Xu%v1(aMmjRnoj7KnEsejkMdqQ{{QzgmXzpUp zhoR!VDA!5o5)z=I9WBw&&BD)W+_-M=Vjk``LcodkjlM$nTXJiKxerWH_WJF(_s+&q^88S&CAoprxK82`uygMF4V2ppe~KI{xO=V^VT_zv~4`)0sv_di0$6 zJ|@f-=~-EYX0KoH7&6;?gYHNiy9t&3cwPz^F4l`4l>-BW4Q`50<0enwa4CLGb!loG z2jj6D^{F!3PHgI1!?bh|3Nw<+aEIY9w=%enltSO8ep9QzT4a0m>64n4)|T244k!Ot z-Bqpxe^WuFBJwuIcSefA{ZprgNBt_Logqf)&@$<&b(eW}$XL|h=$h3W`fMJ>m;s_; z*foXNZYixE-+P4XlI5aP^2!#RYz9wPom47qG*;D}8XEBJ$W#<5$! zJ}UiXN$EkZ(}@jk9Eaaj$bt$-EHi$TA6Yth^5lM^wWGmwOZJ_=R-Eq@-RW`NET-x4 z74n4L5p8o#bg{O{q$aD9JSlF%G zyhV$CH0M>t?HEeIrq8*@3}*2B#F0f`1C0gW#AgZRPdZF``;{qG1QL38l*TjlQ(RH; z>+lY5bS`3Wi2u;;vvC3rS}ExGIT+nzkK4K4xekY@U%qIHA+d}lt!ti#OVKJL+t<12 z?B-FCN(&OL5oPb`F>3h$z6EsQ+roKGrj>4m*ehn^X&w3oQ1>0L&b+|o^o@k4h+qCq zob2QQTc+3Y65ZV|y&AaAfo>iFz<=92j$=4Tj+si+>&TD(l`xLDUl`)h=6@v*gDGfb zQ;}C*Uy;~7wwuIXa(Xwgb2xLR9|d+BR0A&v*-N7lJ5&S0P%U*EJU7UIi$-i4{%-e` zV(^ks?;v*>Sg+rXFki(^yke9?dUsk&i1-L2|Lo@KS+~~Zbw+8ZuCBgwri!mo6E(Vs zTD-S38m$zC$)(;`PvROYDa6d^cqQNM8-No3!`7W@5*>SuF(~goMN9-R30iWJD=UR2 z0;Nq7vKlgz9z_(x1l4Ymw+xhZ|CN=hem88I})LWcy)v@lGxIG%DymIOkx4f zgaAtKj@|^E&>sLh#g9b907-WrJlKizSehn=31u+=ZFuUJvO=<*2W)nIP*?zEB+k|pT=YSh0W;6x0Jo3j+JlZVj@ zmseD@kwB`e)3}ZEhe$vSpNUafew)>+Up}0i`-6Psm6s9h=>%kq%X#)pVc))*+D6%H zg=?GUtV%_t9VlyX57TdV zKZ}t&l-SfZGVi1@|H~2ro94ci?;O{nAo@3sAZ>`0!%<~A*4s=iovBWp-mROzCFF@I zVtS24zs+BNZDGc3r*M>m5aGMeekI#yJYO2r^yf=6s)M5P^5%o|x+y)CMvUmhrA{Sy z_jE9tG8l|39IIjv){F5%Vs?FQDe65h1LaGLt`Vv^qypuLDv?i(W5TEphpv1>+)hXF z0W{j>Ci^SvDf#!&QIvc)J(~U}sb_JyW`eY6NBYk9%j&1+MF*}GMSkd;pL@*IrgOcS z1CXOxIF^n^e()(GiogZ@0`Z`*%Y2mCf35`F#&j^*S`%e4nv0*vu37#ZUi8V?2uJzR zj&u`9sr`jZ+~v!S%@f@@A~e>jKffG~YPWCc}1wQ0&piW!phKph=LCjL{ zh(`k54>Q19j{Rl!`i*s<0a!hsPleWe9T>f0)};CXKNr|>htHBI`Gb9Ei+iAw`SLcV z3&Snk#?P3?=z{qWlnpSG8|Mh`Ok-E-UHhGPf}f!VSd({i5J%K%_3CygL4hGNN4)1{ z^SEC9%K8o&{it+j6!#{Dqq*{iS^1*2qL`+|BhAYI<2|ag^)TT!3pQTY72Ap#LTs>l zwKvT$p%-JQ%JdRzGIsI-m0%pGEK-Rft5*iwFz9A(VS$U0n#1sgxU!6+OYHq7A%E0r zAO^Q3g^4arP0ck#yyzdBxuinyF~PlAn^CU4&>r#5)gHPJUT&d9ET*oQnj+363-KdG zk$LWTZk4c;Jb3WQuR1d%xdqql3!5&W59gs#!Kcv?s4mUuB2FL;V+OQ*1EatCh}JeP zwz?>^%NI$CO}ZE^xAn^h?1*ROm+vM2X_7;^JlQuuFh%m)i^@p+#bsqa^g?X@m}m-p z+lCEK)jVe&#_u6&M->5n&70dJ2=H_AEB&=5O_~+mUrEmu9Wz(?{`OIyc(aN5uK6!t zvVw+}5Q5eA=jbd3ra`t$)Xlo5d8t3xUt|LygN?s1nd1?SxiU&_d7|6F@NgBIx>IA< zwkEFI)Jf?unUe%I4J<=~&*ZM3$fY!HmVCYbH42H5Nw$U>XopbJiMo(2lzlk~!qm;z z*Yu$Mi4!OOpvFg!QkXjzyabPER%pe^j3-a-b0rzG9O*FJyort5W>Ges?r87Wv4fEY zT5heOsJtnVL1V%(f%<|-hX%7JIcPd+g#J^u@8KScmbDnILR6gg-R;u}Y)_II43|O~ zD)CqS%D9-AO%u165bHsEQ|ajx>U4m35jKzb8HD2g*RGxS;UDLwxc%JDg*y%CcZ zb1TQxeu<~KgL^|VaN*qTdVhJn%a0L^li==dDLeGSo;|?bvx0S6h80FJe|h}#at*GH z*z0OnCpn4x`+)Bj_+rY|F9Air2L6+!ukZbByx@P0$r|)cBg$kxJx%b`nC_h}9uhC3^uQMuRSBaSlVe0P>Vq90eCWIX1)HV{@I~u!kmeMh*5rQgk`RK33vs_$V2eON% zT(hkLO*4i#Lf>5Mj!G?1QHAN5ej|O(QNPh5l47*ifO!(?m&R#r#uap$${YfNRG((} z6mghJc$k71VK-AqVTUPQ`MX+Jpv%Go0X*5FqC?aSEPHWfu^fMqm#1eN&_v=-Zx=WX zQ43)P$IxVXessmJ>x|jICuxc{Dqnr|$^dpU4&@vQ;fKsSqi7gy!>U&Ico5|U_a+HH zJ6G3VcAZ8xc7>J}J{5TCB~w5tbnE7gR$EujX`t7OZqe38|r%hu2&f555m-q41@UGav*I5iP|*4N&sy@gs; zOnXrS9VV)aY%DyhOig`5Qj6c0lwj{!^Rk~a<}o!wkI6My6?FF6wS)YPB6y?1^W3#7 zxkr;W0(EZuSq5>$G%&jN0{3fZM1=bZQy)i;9NEcRyJ?qbBpRQrFy#gL^9$+gVq+VA znPlw2j2Ytzvh(NX^v=y8B!hsTF$f|d%yuL3kY_}Tl#`fVmr=%o155NNKp7fz6STC% z2mxI0vHSt!xV}O|%IW6;va4eva7eyO7f(2F-IgHu+N7KuzxJ8$zfL{G!IvQ)ut*Rz zm(W|>suO*eEn7BnqSwqfR8As0guaoe+In~;=|LEB;`ck-PQQr$r5HW~@(G!ik~Z34Q4fYh67Ky<~@G=cr^QesljDlc~V37?bnZy z^PSMhv%dD*GnZ{jt$L4Lvr`NqcTqK-iW&k=Y8REI626H%n87)}M>%pOx%tU2Uy`iu z4j(n@M8z*fbQ`9mGzzSU%#HJum7GWt=3&%`6&M^ zt`zGL+TkJJ+*~2v9qCK>jP;$e-7I^DVWq(P!lMd4-)n=nI??L>mn~Pp#3l>EZ?yJ$ zxSv*l#x>R`To_f5o3;1_2U5MBZ=_zgsbkT$m^0gJkfNsls)^PtIAhhK!u$8Vm#k086>`l%4>B_L&XTpN zgp10wLo~K)$@(SynAo=QEe2js80Nrmh=UCTQJuK*M$g$N$LfOH)oBh%B!Ml2Sv8G? zw~6k-hTl%SzP-K=II3OZiu73w=K`I52nRO3#L3P%-wD_FgPbY%XF-qT3a^hHGp4cJ z9#HcTr&*+EbV~!Szn^mM=FQFsjo39fLc$U~gK_-^e+0ZL(T zJ6lFnXUup|zvA(uN5b2Xex2|(yjq#X0T-j3v;c;f$-QLljBq0s_x{J^M3R#g5;DQlI ze_xxx zoc)*otpzxISniH)f!C4zzd-hT4j%k)5`*po9Jnn{u1YOUOq^EXq#OQArYW^Vp9if z6|q`Lf43;NO%(?WJ33QLv%nbjqLm|NC{g=lZtq!C)_3x@t|AlVD3xC&814FW??u+= zF=J9#$cXPp`>AZoS~7>yNXXJMgC>iCe@Q8+;xP`5_H}f}EE%FDuywuK0r8E{$Z|7F zYP3z%x~DRDBj(f*MsfCWjowFxcXU?scrlL;1Q2$M87I2o1!Juh;4ujFv~>O^(Kle< z-pD02&ex_s4pQbcYxrl9qF9e{%@LmGCXkq3+zPj{nIMxHxXd^v?e~RDU&{=>U-ezFOT`ir7 zzYb?q+ljBv@JD^cSzZ-ju`8S{we-g+55G_^2z3QIqYoQ)dL41iW$Dl^%1MD~O$*DG zc2hno39e%W2~il?OW40bV|&!Kli+IY#3wUkvZRulT1W0^*Vu!0eXuml9`W!D zkjau~`nCokP&OSa;{bKt$Q`OM*Oj@dy!)w)4Nyi7dZ_*sN?B-YfMjo@u83TP%H}63 zmbPE@DJ!sXO~Mroo$CYbxhkEL8JpoxiU#MAW5>L}qeZAEIfMI=hQQhH2GYyq`eAfFIBx}=E6_~tayy-Lj79HIU+|u<*t2rQNmP(u(fck&x_Ln5%ld#s2@_7KQ~A_ znMeDK+jA+`e>F^*uiZT{sZHRzJWfjC5$8jFQ;ESx*J0hdc8$cc5AJmdNQ(^(^qWYf z0{*E+8`dGg>H`$E{n3%(*~J1<<1COGo}g{n5Oj;+*}hZnw5wB!1(izz$lOBQM<4o7 z+b6P~?$2q_wf~B~B?K9pvEAb^^o$|l;aqXMCDNlwefO>(PhAuN z8#k(RgdW^GjGBwt14~S4-bQ%~Cvux#6JedibQpR&qq?_?DS>!}pV{Xx?q?6~M+v{w^vO?C6dBGG zS-4{lpXugE2$b0x@zlq7_Uvl3sB*&7m&&qM8~n@r5ZW{N*6FKOso=CD92d2J%@5Lc z5y9|F!M#5GPLDd4HlY8js;#v)oy3Qt%v@!fOWYCKZ2ze-kom7|gaAboM6^1AVx?A<%CgZ`&Z$O@qLeC#~QC9YPiT z)>Qyi2KuppOr3m6Uhkp%@R1{JHs8C9r(G`%n<4_Y*_{pjMB|p)V!lStWZGuSXGD^M2^93X!DO_Bsz_(1>3z0tympkl)77#Xv| zkE|=iZO%=N(E}P-f0xYX6&Al>bBOJ|u5x@45wZDAID#GqJ8!ORB_ZbSNiR~qBV+`B z4%0)yr%9~VZEkiuD0rYXA%uY<3bQqaD8Ey=l!(?b>5qOc81)S!bC-QTVbQ6baZC;$!aBP z>8nelrRmf#t+d*|mZLBfRgC8gA7#PTW0-?qx>!;Inh0OoqMx6iFB&LR?4#{o4ABiz zqF>w>=A1Mthrl7;kMv*XnfBx2r{O~Z%F59}_L?s84E#-ebBxEN_9c79m zK`yRJJjT@kI>_S(B>LU49qxgfg=~tyu5E?xOl*Z#o8BQO2!ABnyy~`3`;K%^=W_I% zxLF0PAUfqteT>ffeGD*39-kYzJ*7Y}buYVbc5kH=Zf92c>lK<_AY0CXQ7_4I7cX5J z{e*t%pxAHO8-d30JJEZakE>lL`;coQv>BjM@#l!Z$%pRq8wK zHTGFk$(ernq<#7Q_lf!^VTF}p6KDC;!Uh6}Z5Lw6$(y*@Ez@XS3+DwqN>IGQd8FL) zVdxcZZc|n_k6ZQUT2Dn2CTx=_?0O0V`0Fwt9zN~F3GI#w7dIYhbO1DkO(CsX! zSO2`C*|YCczXwJPBEyO7FAOTM`$Hma?O!*PLG+_nDr&(>!c6QBZ1dqkTk*fK6nsD& zXdB}X<&;XVp&H`c8Ur}v`VjXyixzbkqhw}Nl)NN%G9nb^bSdlf#M+nlQbiV_$h-71 z>eekUTpC||R`_8yrHf_E;Qjx^YDK@`OVa-{h{y4DfC5{xXi%pC(+lK!`5spqK)+J7 z9guYqWMaxH8J8ixeBi7D4c>2^`d|`SBn3)NOs+wMZF~}O*~v2lZZ#6)nDcF-n;#SP z+V%M|yz8RWEnGEqX6g|1Nm9X!;z$s-bZ!Lr5rh%F>qEjmh{quI2M`(NpW79Qb)u-X z{hC)F|5wWQBC6myV;ouhm=1Fi(BKBC!EDJAspqLH+?%|C7t+47OZ_karfe4Fr!dq1 zaqkAI24SI1$PZB5*gTX*-R5!Ta4Fsxt^n=Rj;T@c4oA z$_l=7s?cCAU;cF5VieR!o6R-sJziC|9DG&jh3nxMk;bSHOM})!Vt}X6%rCTQKQ=y2 z*4W4>gC*>z9n#PB*GHpz6g9W& z2FfdOlL7HzkfLvEbftc%_)Wt%;$~WHlIvL!IUY8N;qAi zsNMWT6DtsuP6$~ajk|O8YDY02^WrnzXZG-P0n_(Xx0J};<;vMf6)RHu_U#DKzab{+ z4JxSK+Jk0+D?BiA7L~K|(V692LztgIZ_2IQ|4Ut;mfHX0>rLQ#&e#8Mvsh;A`w~%- zLH3AgF(@I~CS;9DBC>?`vP&Upgp3-=Qc@BsQnrvJMN(3vMI~vIlDeN)Gjo3b^S^)h z?eREg=A0S&em|f0`&wSF*Y&!xFsQJyFr@eT4q`H^0*poMq!DCDQq9_Q!GL!T);y)NnK$Uaa|)21%AG=oHahjJ)t*@EZu;i&J(hXecK zoQSl_u|3esdkzzS+{ksA>bF07ug!^F`gG!rT_=$Pyn)o6vBp596X5EoAj4K(p$I@} z&qgExGx_(+F=#wM?T=_IQEMPm^Y$&g4pQomblj^xethkP>j$s1+<>X{epek()l7tj zJp*Q?1p6qP90cly*E(K^V8jgS{;|y(b_ETwNFNGWE~~b@GXv|4%amr|!bwg$X7MnX z=uJtqXQ6ZEiDjCI2*vGo$z`#(8L;<~xPVw(B~6k;#z00c*~AC)kbvJyed?u~ivI7I z9s}gTkMLlL%LVcL&WiS4ylfp>h~^Bd;an>VgJv5U<@sD>M2z@k23?+&Qw-xT$6 zX~Q43t2(O=7;wt9v0+EG6jVTgt$k*-Qo8kORhSyK6a5Da*p22-X48URbgpph5~S!; z^PO>a+Xy=srXNHm%!d}33?aQGmo{?F<6mK7iTj1g_`=|;{AeA9Ngov zul;)RWx*)_w6fSD)4>=oESAAne9wO3CctxeE)KoILjz&>X-Au%fVsrywk9ic64Nm3 zZjfD6D4y}QI)GVcH((Kz1!RjBfXP%~LZ?xIiAVWAemdL`(VL2ip43FJ>y4w!tQ7iL zVS6KHgAy_?m>>RcsuQAGftD*x@Ly5>pjz^;SX#ImXi#3v@QN}h-@Is ztazV5#O{tRtf$WE{j2oaPb9KLV=b^9V^! z@G<4zhsF7ltOmCUOSt1%*#1v-E|&cHos)b0R6AE{b3`XUCX5A`LMi;IGI>Jc<1O#@ zGW{U2L+Wtov>jC)IM=+FiI?+FqA!{sV!U0btiAX4#hXfwWc$#lk^9dD(-v-n|EK#{ z^mN4r`Zrx3Y|YooicxmceK4->PJH#h*{$0SvY@m})JPi>)dmfck?_bE-zXgFd5ux| zWF;k4)fT*&qZuo_{R42?;QJ}Dhy_hVRwY{NUtazdop}pQ8W~Uz*Loat!_l0`5og*& zWD{^Z@Yv0SNG~oHc9^a{UYv*^Y?aB*WE$gGr|}d*jg?<-pvB_D-zs0QKvP0U(S}XU zP!kB@HWGFI+3~52?(N|_@9#CneX#{>2s1&^`STuNAW+UjYkm0Q*|U52j4~I2)FP;# zc71nHC0UI^ANF`pvhx%0&LY1pZIL#b(Kkua&(5>xDy=2RaB({zST8BacTt9HElJ0d%dqKR_C%fJJ(6|yX?eS=tuqMxrPF5Ooj_t>5 z`)2X%#26;7EN-*u_8%*GMyC%@a2`&ogtHtSxnohtGs>h-j)nFMa9Okqhip*TIGsII2oXh zmimxb)4W6|4za1GPH$dv>Uu8?>;3#ZAqjW*d7L?YTBN>GWG`t{gVsMNLfgg5w$y(4 z@`l8*!~7LCpb%7G=|>VAZI~F9>&QHb+Dpw2oKgf^-Bf73n~iww{dvsEuAdQR1%-K- zm(DP%uvg2mr*^s{i79R1zh(8qMz64Eo&KfL|0_6a2pq&ykIvtxt1JYf*v@Kb6#6V) z%Jt?W=z%@3aXVpX4S@Nk(Nk4*bQDq6yg7UZ=9Pgd@vSgW9D+bt=0>TC?mHjLORqW{ zcm*=y4IuD68Y*@zHP_U9@D* zi=f|I=D0R$*JKq3R6C|YYEKo@b-`akk8^c}G+^4lx%Qt`1^|3blj=Z3>8>{Hmk=dI zR6YpRiezY1B){@dih<2GYc|RrNPv&!z=c^6(AiO|=08BQv5$usdYqwJb~(j84mA^O z9(xVa$MpB}^InjN!|~Y(J&q;IOXx_9i-gq*%}AqRZ%w@mq~Km+6VAVfVD|`h!!y>% zi1R*00@&V$tBoIYl{*5$w$RYfAUrDph-2g>BG?HoAJE-cWsVKn!I;~y@klh5Y-;7T zhjLhfiC2gja5>$zYeSZ_vFD2QB;2*+4{?h({a7m#S*Wz43LixXhRv1;!bsTX2KCi8 zOtere?y`R85rit74l_GDdq>-n+Qzhti@U?w2=O2Jn8xQVMvyB}V*@XSr4% z#LcYcW(_{abVUU|3zq8W=+KMi9hNoKG8yvnCm1d3wAPF zY1%54m->w3i0Yi`sk1bKwmj?v>auU@_!aY1TW z-XQcV{MYeU15?4N)vdFR+0UfCsBEL$N3Es)Koyk{7ls;8IR`d0Oa)dtv&gf9ST-I& z4JO}G_Mk*&AvqV-N>;_1(LNyF$I@qF- zSak^iDoqNLnUf-aXl9ueH4;zhgwis1bxD~b;{cG_ZQcCVt7LaE_=O={Qsb6bXH!M; z+vqy5tKAJ-dr)nl%P9!Wq6!zx1I60#>CHr0!Q{A=7X^gtL7o4RQMd17UVFa^SvXqt z>gZ9UM3}02b)_LHX;jVquR1`??;wnUHt3*J{fp_zz^!O61gIhR{2LT?&i4ENko|s! zQIRD`Srx?mOJJ=so=+so+H^*CBJxM0^Dpo@{^y@+m6Iok1svx50}P)d2Q_g7t!rOkETcUmkv*|A0Dv1>Mw=mid&$PDff+c@1akqX)edEBSC{%%78G#o7@&{o}7TJZtpF zLh5mwxedJ+#ih5om@~~#;{OWGGG(9VjKl@x1}_y-!Y=B$tG32ySZopxQL$O{Yo zU8jJ69o3@rTzgYW7Mf^)d^Zn?LLFWOPw?!r^V5|H~wN}MypwT6p(40r-bx>|A!zF-nQ5Gyr z5zijuYioN_eMMQMqv-1-bBKg)#J(r3wt~;G4mR52Q3T=3jL$U6-Yt$|F!t+5B55+}OjmV4wI-@Y9@ltoPUR|0tl?G!4x~Kb z?v;NH6CF|TLX(}~7^q`MUkC%ArhjGwd51h!t>ZI&!LK$tOablQoJETsjq$!@ zEBcMAE5B(gO?tRkCI%>JpS4-I)W7_F$v}iTf5T1 zX*pE6Ep%RZuWtWw4#lX8vtBaPbQjz?%Dkwg@4P+%#U&T)(ZyE z)K>_7DPUH&FrY>pf)*@9V|UXdj62KMrDnE4vWv1c-LIdfJ^l6%$8 zMGq;^O4P{#_bVL#zAeJel?`NQP#X0fqb=HyfYU$tVuiUIP99d`{>?JFtmV|%OGf_` zaZl^SzghsLJmxHB&M5ZzPs?pX2n>2*R=sM{%TD`r;l?fIPw3x&=fZXWvS7RF`)2qX zW5@V!VEdoHi*kT8BRfAB8V1@(e)CRT;a>(@fUI5Y%(F(_gfv>R!4QY+&Nsq*O^y{cwuXc9kB_$3JL|UP&MX7b> zm3<}JS-!n~-TVtZ;_d3E%U?{My9P_Z-x1ao`V=B70J4bg8Qafms-(T>8|eeD02hcs zP*#PQ8OsBk+}oHkTAybEbns+=ICnQ)81u9ijKEH$I>oy%LCXMPyX?OZl$lo~OVw%F zR{0q;%pw;b?TzCf0>J0b8BplZwQKyC-?5NGg|(qTT#4-}lynPD{pIMG7b@tHR!1}2{!Znk2@b_l;2 ze~$C2t=`U`4hZFi&=P^EOkLp;v)p%~!ZgRIXM3}kPVRw)NaCB7Txtv@`$n0E*N+Yg z1=w4TZf?nFAS|TsyYOe31bkw3?FxuweGL< z-cadq&$|~F(ESKCZf(EoCg^~;Ti~XcnK??zEI(vlBGkPAHPtVotS5c=vD%sjcy$u$ z=%jwBr$Ry=e^_?se|k@^&%Lf8588C2g^GzMEsB`J5N8n?-4X=`KS<^lN>6jBq}oq^ zoOv+c;VhFSv`Ne7`x|*8=f*#h<``F&~tD&O(1KDNo5 zLn;N{{SH%PUc8FPK~Ib|d_PvF=(VPkR?Yl*dvPnHc}N~0ZbqXVhgsu{*aX@>V{;pr zM%L%%5z@hvPk#e~wF)IvrB^Os)_(H;NtlV=pZ%gU*Hi<}A8Rj-(pZ+}d zZ)fS%yrO(czyBp5wR(FWmJ>tXH-}fQTX%rPrA*ClNHAjvYz5j7NxDaG-{Y0sB5EBk zG)kvnjm{(5{>8&xa(UNNH0ZD_`xj@UQjA)ea2g>+Kjvd-JqrC^2hSUVahYr;$%?Sw z#PACz^c46Nvk{;j%{jVsF2n3a_zAf7aDd0dSOT{x*MF0y`4+G;IU=xR141_wWW zJOf~FLjtO!J7IAR?gFPKn8~#GvGF!%pRS{H`i%@H7oL-v3yHtua|QEV32929S6q&^ zTNLMA%<4Bqe6*v@!I;Dh5GY|ug)yp6F%U=Cb9=ygFoMmWhAQoPIEb!J{QoJn6Du`W zD-eIE6rK#7Q5_JlKSJkBs5bm{P%9 zHNCVP9z-_3$wqF2Y~UCfvLS-6<>E1 zw~IOj+ndD})Z{`Q2vU-09*>}zy$@)R@_eb&7B$f{e+%egJmsrZ$`#LKi_>Q zz9476YKpzaqO09k(akN&xU>(Li3Vo(5M1m;9=Aq1YS*4$F8 z5>cbY23|KAFiA!|YvwmG@Gd%#g{7G~<@~;W&I{Q^PEJN`+)X9!8cSSsfv1nBt%_FcpH|2|vnwv-lxKC9 z^B`ts)U!~Dq67hcnJilxay z=WIEY`7`z)1K6M$H_h+Lx>wz379g^_TtAZROPd&J^JB08?!2l^Hr?{zjS}*x{SHP*TyZ055o!;~s+|%~GT`$%8xKef z&YaR!S(`DLql=RG(@Fv4J|nyHViRx(6!SkZCVOs&>;J2*ud=4#i1gAn32V}VyUGu( zx;0oyIx--Q(MLmoTTnxxW0)s1mQ2vD@O$^+!+zj_QOO-MML*H!i9a@p1N0SdDipus zfKB{5d-nO8uNok3K@ilVdrxniN%<%QrM12Kp9h<*pWx5MXpUO{)Z(CyP@%$#8N49( z30ObiLU5BJ0HwC9Rn2j#YlamKi7rKV_w?l+n74| zWIPof5AbLa(%JqJOn5z`PFWQMf=ldSBcIzOnj12qLM_vTHr9T~7=69$^H1=*2w-C5 zYm>q7@%fDmX1ZpL`?hV(aPIu+^`q`YPjNvxRwzaavR4#>|bhUMhCA zi4FBUD5P>Dl7oW5yo1{PUaJA3iFzum>*T}+4?0pXF)+Er0iJ$4T^E5ibm-D$7YU`+ ziytn2z3`E8PGIi2;4@J#N>3Q!Xarbt>fE^p>gO0HLGWSfJEyBmj{;yu<)h_U!QI`| zSpQ)%dP>oO*VU9IkFkaP6#&b3Q2WP8N#U!K6~V`KnF1|+t@9PF;D&_6hY#Nzp8Q3| z@**-RfXrl=T$={@4#)nu1!Ck~tH1f?rsRMcZ=l;L7#~<5CLkYNrl3r{7^}*GJ6as8 zik}?Ubv{#60=p5r(9kDCD=D=YLwM(--%6=3 zUfaXHucLsuO(Fgy@LC8qx`&KP%Cp-~4fM$W#*p*R%%O}T#Qzh34n?jhGgtpl3-)i} z!#|z7k|S2Q>61S|CaK>lE%AEaUbKKmSDXdwzfVa*T{DmLUnwT8xQ)|s$@aK66=nIJ z6B^E^?RA}h6d)0@_6NFH0?JhZ3`8yRdXJF^&o~AJhxK;0p9=2Ak(D75{;iSwbawzg z<2tg>em-g=alIiNQ&rt9%`Ya8kI>vav6bxg0omGZy|creX`~X$td{(G>^r746FJ?k z4`}uq$_#P+M?PQp>D*WT=riZ&m*^m-Ti(G@xL`;NyT8-rL<0j413xuIt+vg;ewS=QH7YY#aG#qJ%o5N^Aa=qyAcC`u zD(wD%65$L$grm&M&$0!IZyNyG@i>AHAiM1g##dM(D?c`FV*oYS6B#fWpwIV3?~l_) zP$-3bsN;KNJxzh8XU~m^d3qpTfK=?Kd+S3(G*c8Pi~Y33@eXnEF6+~eXd?P`^I$CP z)~mQgDMDVPrmBOn4f7w4l4_No(t4(9mRb2(F_+OpDlj7>bHA>i8(-e2j2c{0zwS-? z_zmqm{&KtQ_E)Eehu!S0V#xCT{)9s76NM&1tUf5_>HBlAgn>NVb$ zmR?NQa~6oJYuWjW)KYzDKzuik*VIJo&=V5xdEhJqi-?{qq#KMj;ljc~-fTFM{-nAO zGWaZs4hQM;if6Kg%Ia?7G{-xX{;2Y}$`43TJ&dpAP1&w%u1Pu6J*lRu>Uq_%bHk7U zp^{h1uH+)R+s^*B@Zx-||Mr>t|GH)R(K5G8_lp__W*QpqPB6c`S~Wr3K3P>r4&A~y1+;1KgDGa?xt6Rdv3+`N1$J+LW+SU+}=lW_~S3L5h4EaJTX)vh*?BvdBpEvhtY|A zug}osDb#S`PT8INfr;|#@W8ma&8gA<>sQ9^zAj(BBR+L6Tb&&bT0!k2Sm?lSLC0u zR#dg@hCJ)gTp0cj{~$0f8W1DTstJ!RMiV%#7Ze6*xgNb;X&Wpjf2p@0Glo)X z`ECx%1t|W9I$?Kew-QO42&E3lGOCcB`&VJ96RbZX211A8cxrWzz{#Abm{Z-HYPfey zScz1lJQVV$_+w@YEr6i;<%+0SWM?vaWjmFn-U+LT5~EnXK4E4QSH5e{k7aUGtM2{j zTfE7v5x62H6qF~|V|Z+b(n}fuO7{rV^E@a1OhC0vefRF&UBfxkYIHE7 zg=1cwnvWF}8y31yLHSk%iI&c_q0+wxq7itGLD@AWTO#p5GmJ@8lyM54y^;2mlVUVN zp)xk-udf!zlPtsvxz*1<&;3<@DZSR)Cz@sV&=Tr8qNE${yM8@0C~Tx{!%GcPjJJTv zD(6*5$OzCzh#gN6wz;y+dnjvB(WD&YT)YW?pxC+e>4{g?H`@9;Rs4jCOV<(BL&bN! zCt`#0M?=DHtsQ$oad)xo%21AG=>yQ+UsJCsipYTSsq<^6g~OANe|#DK0|4pskTRmj z*ecZ(2z13sh0#SF%D&a9wJ#R07oUtLPnKlA!ic`?W{o0hStC!=(UiRU^yQ?Jdp8Kz zmXU?wRgFty?B-NCX(}sEQaj*HBc+GF8U3h?|M%Pg=e8x$s(-&Xa zTO5WS%CsyLqly3ubmfe}vbY84b8HgguZY!(UMWau%|HY&2QkuiKZQZ0`$`dC0}*S* zy%5~^Ey9ZKiOC5Gx6M4YE#9xn`bUhuFMrj_mHSqDhcIXeXx5RrUEeLdx#Q{g+FH=( znPyP@*}9EsaczChFduEUNr$gZL*hRIG7*ypy^zz?prEcJpKI>fvSQTnPsJC)l>7Dj zt@@WOqabtYu~!d%a0_?H5zv%rX3sPF*k)~~zL3GSt|c!c*ZUi9T^dtcZ~}33N?-|( zGrnbCC!rAx>x@`sAX~vbAnGhDW)^^{%Ro$iqCzO_TgIp3HHe{=1X_3wOsNSq(3Sv!@byxKCl*%#7{ zdXYWenr$OH9%hc%;>n*u31b#GZbFBtw(zoU%s|p5qsU$Pvn5D#Fmqc+Qq>;!_@iM8 zS@jdqC0KE1m+`I*F{;|6M&GIqY4gZ}I^U{(H@rSWwevFj{H{!FFY?woLN8g)vADhF z>Y7ycEbZh5tWG_%s+&i1+Vv|XNgHrK3~>074Z+=bEx`}YA4OK(baBKZGNAVJJ=1Uz zdu{!e@wLC`#9+$e9i_vB#4Y#B6!CDGKpl_>D%qz8@&^-JI+nfF{?u4NaI?*}Z;6z8 zw)W3yb6i|vr&?14sB6w&kbk>n-!4fLaJ!M{qPgZBE}+jstk@4Ap~S*D*!bdnd8`p% zwZXbsRCh{sse=ad8E7KT^VSvZ^Ew*(HGlq2d(m}!Q8~On`v~~cG(*>1L9Di&YA;(p zK-HvMxlYexyqGr8347Gc{)S1H-Uo+Fqz$@puJaaxwNYa)kB9-Q?l*YwBMW=oDWfqu zmpMqxp{Y8wimX~XU7PKnd2O8sP-l2O-bF$inbtapw@~=n;lqbBO}1Fg-`o%$6*lNu z$ncRPQw~{Wjr%pBY*znljVQ&SuXpBO=n5LrDdB|&0p%3Ma6q3a6K0-2Ku?fS-OF8O zvg6|7QhSO>@tmK-^rAM^`G_ES9UfcB75lpFMqrW%{!IhdzS-ixxE#W4x(cB282X1pe`@>9Wy^}QVWhme}zmAN+FP*q6okWww% zWdY!*z3zh9it{6PT10|k`oRdxHbzYAOH3N? z6N9!?f30(XO!@Twv}ds5yR+r>OWP&q6itzBqvZ#Q^)jxMtL>1=d~JDUAYHCgRf>K< zA6;Fe2`F3axXklBmF{F@EcABl2Ba`)OFiX$;7}!ok~-@!L(+a{B>STA+NG?*`=187;0iw)V)TYliV$@&e;_f z7N)ALyQP}5{Ej&1yZHdK*)UNlfd=jI?ES=!ur;Y5&@#R(*PPookUg36Jm(BYZVMqe z`$Z1;`s6JCP4L>Ubl)vlWbGDYOGX)IeY1Lgnwi^(6G?W}X$z2g4al^iAaJ4)3Fvbb za@d{EyzGjoJAod%ls>(H612SVNmYf;u4P({^_7-s!AFWmoB67+p}^@cqMPrZ&s_KdEOzNu zcB(GiSk1KNB_z{;x;`P(XU;rmSQ)iiJ*9e~S~nhb*WodrzR#KVV(Nwsm)?(2=D)2))XAgNF^9vTT{3u?rR#r_%jKTqgYYCAj!jN8h;dF1pa~S%~Kq?Dh5_ z3v|nW^)+I{)h$SBy|M@supMsZCA+(oJqn5)9uC4Tc2Eb;RxcW?ruIBOB1kvSPWp!z9bZ%|SvG{Z zfp`>nYkfeCb7z;8^*o-pWjkqa-?C@_^bX3mLL8d?Cp-W1H*;rs9Nh}|Gm;Qju6T1b zetpYaYFDPPPwMCTO8>?#rRNV2*jX4b_ey`py=Z*``Z%18H~pTtFJ}_mz^<<799P;h z^7YmeuhniX=}vN+1pN~4h^nL?H`=T!exCmR9^WyWmvT@nV6^P#UO9b@slBj-@0%y< zd{raH4a+{g2vH&#F3_P(z=SalRh9_D6e{9SSBbhCjjBFaIHyNuk(qCxaOpz3`heXa z2ahgu_=z&(4zmt24g^=w3X_)L#x>T9I3Gn=3hvge&~ z#(+Lxdv~R~Olg-!cf9e{x=0+Ux|79ZSyD7VA_ya&u2@u5 zrWwFPrWp>(&W2x+2X#w|^J2;J)cS^&&D7UrcA=Qq990&^yORjkgexd)I(TAr)0GQf;rRL!-Xr2 zCNOT>VX1;FV#fEdd*XuCh&fQQc1TRmjXJq@J^#PZe`imU%Qw4So!&bM^j9~`aPQu| z9xF={NC|qI`sBxcqS4j$uYDD6))Mi;y^89OT>8=XMjrv8d33!c5_8JNUrHj+z#Hs~ zD>=Kjx&x!ad$5O?=^e^C)$Rm_=Ya;ZmN^s+V4!48YBhTWtty<_c1&3>h6RXDlrseK zp!C+eJTIIy&F|nE>@^%c?EnumrJg>>qu`(9JpR|OnH#sGT>Y)cdadH%tcJIP2 zQ<~E{XjMI(9Q>ez1_O0X@1eY1>QMwpSguX1?h!zLCRX+3egO032Q!VYSbBX?K0)p1 zTR$_^gC|$%_@~=FzTF$L*iKYv#dEBCCFkUq*Pm>7IMCloIeR0JgU9ezfeQeusB?E? z`eLk4`5N0$XBFp;V1&6!Lw*|igE0$lEgqhpgtwo7S>=BEiuU)Ey4`mwIKGPM z;4p_a`3==!uKU*ig5iD9A&=vY#d3+kKXo-JQ~Udd*t;B?!U;T7)`;2_;b0tM7r}rMt zX7m8q=$9fNv`{Tm91RHxDOuO&;y@03fI`H_%&P`!U#G~d zGt4w+=T7V4ZtE?B`D_a@6G#gziAdIY7O3wjmP@E0NZzO$Cy@ztHBsM)04=56;eqxu zRh(z)>sD((4y3e=PO=y+fv92tNcf$#J~=H`G69g=$H>Z9wh8O-n1OXOyE4ZZdfZjZI0@ zF%hG_cuD}Sof>L&YUjDydkn?InEJDF*}#DV1@Q8%nO^)d{Iq4PqMyOHip1e##%rv_ z>_?I<(5lD>6xPQ`CMpkwBvttAV;;etCsDQ^e)=p~$>+S1t6DzTkpP8G7V%OZp8QsIL|`HqfLS zD%&ZrKf0P#PtZg$FBQ-SHo8MnWJ~l>f_oQ_Y_krdlIh)gq%5M84Jxx1YTC%L)Uyc-QO3A)(@YL`4pB_*fq>d zMYt_U4p5yMdC_lLQ^;+KbINOhZ zTIyrm0Vk@RCV>YN|soY%6gOn=_I zBgPH>3t54VmBl?r5~b30`%BtF-eNb_FFOK?v!0y*#DI4QYCuvJSNHIh(L}qw%RZ*e zeT{ym9V!XaTDj=PV>_mqkhleBO*sgQ{_6G3pV6uqW)G6-SwUCJ6@&Zo1@PVQF!$H8 z7e|@y3G;8iI57n!M9x`z-Op;sqYy^M+aznB%X3lvrmUi(t0sF9(LZ%dJXe?S=uxL6 zlNh^j%h;b`)H-TE@oLaidNt&iUv3fKJP=H+c~Qh>2ozs;*A|RDI^_zok?^(rP%&my zGZ`*8FZsZ0tw+4_;S2hJhymO`E?`rb3sSso0eS#Q5^BF5U2_VpLXzWv`1|+W3;f61 zOi^!~f(>+hPM;rLYBgkCf@amoxo}_vprC2#6Ig$0=-}q&asiJU(~|k%ZSHx&vilQl zbh3p#D8O$t|7(5p*HC>!Lhj4gERI(E`)+y7?_D#j)qmYH{>NK%4_|0>^)#898bQgL z($0~XKfJ-_3qYA}<$7I*{>RDx zb@2L*m0Gf5c+s7 zh7B6@6TTL?+60^t&+jx`-IBi4mVImsxn@kP>>wN%YMPmvK4pZ$Q1fcPZy|g8Pq2N# zOneq6FZjVzBU4jl*g~~T*=D_F>z{Rw$OY~|y4|a|IDZIQslh($hf59b5ZnHf&bVfb zqD`-EW&Q*3r)dV7Me4-9P*Or*YL4yPCu`S4F2(BVExU+`C^ z>^x5#`{>&lnhbH%`VXS3z2^D(j^e{ZVTEgwtmzGauj+Lz0Zz552Hc%q=1eQ_lgJ@@Br zIfEF-JuH99-^fkg!^gob|84jF)9P(OAE4Jgv*k7@*d?1!edjpJI%`~1RFqr(dZw>4 zcGN>$#Z&Lzs~Fm?b@#b-ae*3x#%6wRvgNj+hdG}89>a{zX%Hp@YWWwtCJl7n7HT7y&Wh zL{NzreKTlo5cefeqkm0h+ZOlzx@ydhArhTri-^HyBaX|V(>ptT6jCNaQ1=2#~rWe#&A97~hUZiELdzqSY zuy#@5PHp@t7;VW^I)jo@2B7VdrCXxzArJY3h@65#Q_hfrFCk5~Zi&(PiTtbUUilR6 z$$LuSE#7}>Cb7L4AKVQfQtZ|0@(KSRDhgK_*+S$5B< z-ZLW`JWhUVQu-0uF!eF&T{}a~X`1Xye|f_;%s95nHg?&k*h}kV;|h~pK@0XoMrsHw z#|EgDigwKq!-`S8^z~YwUz&~eeeROQ*j}*yrSTX24N58Yl**E)=4dYXwxk#bgm}$f zDD>p0$2?FJMZd5P+7t%9^{+d1{I-ldrnd>5i@4@_t-UyO;pKa0?HR8bo7BHr#WN^u z>5V-Uh93F<>9Hb;zx%KLivKaSPr~l?&H6E?NM{AFH;^#RP~IiWy;V6yeT6)?bODKM z0?k$o5A5FgQvRglpRv4b1ByW~I3;r&)vbIFH`HuqTU$!J*VGB0c_G%r2WTGGHGV7yjSVjF(O)b}RIV?un`Q`hA zzOC-u->1Y`w`fL@z--TM%LnJ@oYt+Z z4i>+3o@wz&bz9g=@b}4A%$Omen<3x|?7$>SxN!p#lR2|z_wxx`0LC^?`}?9@^K0Ce zE?J^pq;!2Dv}0}`BF6jFGah!HH(Rx7Ljvo6&k@*VKc{D3&HH$iKy0-l+nso}>D?gm zBxR7*i*siHK->y0WZO+Yf&sYFkRb-uJ*JB#uG8IV^G>ue$Z2^r^XYxXX*n%ztG=#3 z_<8?Tuj4aR!`Bw&*gJH>h(kjoa>BQ`jkdP>tbGR1-3f*w3ivMsR2i4TZ=%nT`61$S zQBMCP1OB3pVy59ppS`afez;V!8h2Ja9s6aQ4gd4xNtyC1_amLCExm3Oqd#20%Pw6& zs3bPgd$@0fSM(_G4iU(2<%t=p(6XGJIp4%@ifPyY@fwqvq0&R9R~O+rr<(n{k&#M6 z&2QC(d3n2ty6)!w<2HWZRQtU+e6ei%n)2HIsS*5x0LmuI=jXkO=e*ZIzo}FgDoD8T z*fjqt=Sb=|6ln5xP_ z4odbWcJ!LCgB=soPbj?JyODt)60d0bDRbTeM?6@({^o9xUdY@l(ztTu7;nLsAHdn| zM?HOVHIvZ7)V*Jz!O zeqBFjXPJBrsQo+9=^mU>@TkZCT^<%KW0Z-Z;=!3B7*4?qD20Fua69zLy#wY~`rk}Q zBgyMkRzI1eaXoc9p$x|c&s(pIc;mTEk8V_SOyxn2ha6mXjr+0@n z%VK_4&ruXJk3tn)qxB+Jq35!8(@U^x2<_t+_iX*pWiiI;anLv{QtAlR&*(84oGwjW zD`R}|75m)Ie@1m~9o6$%3V`NJb(3o7zucS%^iu&3es-{Z(8u+I%^3ymwYTn|M)8#Ai>xi_MR%b~wB(9`4W!gF=)13=B#mp?`*V*6=A%~7dIt0% zli_C_ZTX-KlQ_57&MALQ{DDbL1GSuNL6X9sWo^o5 zYJUrtTVDWhiR_5FqiZ6S_H6_keJKf9#ITF;LG%$`*g^rhz5y$|3mdO8Z&c z{?2Jr4=${3_RpRA&tC$kK#W+%l7=y!=$E;|lkljgF-W}oRTW~tWg|#@O35r;;|>h6 zyw*CJUtN@Peop^zC;jZHEt=6fpK|zdvcb%m^R;&rKm_k1mhap8@@wK_BTahDVEv~{ zkcC+`aOdP9e=Ek0xJ4r0EwwIbsSyR+7NYNFb2trnLGk9BjV=ItG9MIog zu9Ti8Gnvf+Lg)h3anMSsNBis!I69~F)Ai$}L9y0brG&BmxK&%IiOGE@R5P?ps-~m) zqpGqa=3RpAQ!{#1jjR=Bw6c0n&E=xVzpqzbnL+CC#lw%eXp1i_#SEFOyt49H-tpN* zB|#1{j|yJ^%0_y79)c=ra>THU;GAv+z^)t!oL3Zh17{uDoT0VT)A$b8*in4Dfj+L< z=+n0|h{H;1y$XD!n&b1yQ&g#rxgAFIR`m^AMJ4q1?41_5=%0OVqO@lD@iaA)^&%9K ziELu_U4X3ZsG_6|#A55}jgl1v|a=GhG3)zQCyyk0V_+z;W2?32mW4jx}@ zvtvq}zQGnjH9>8|oi2{8Q4P3$yoos@D#nk^NSh89!`Y2@l(ioq^=?q*$YMNBCNeQQ zLuJ7bhEyVpwtPi4hZ6Udn&(5mKKK}-pqPRHRhl@aYkbSh2dUco*)-JK?~T$iOJ>ya z-em?urHypYw>kr13kqafOZm0O-X!y(Oau~8#Ffq0c0$F&{zz7y#T>mN21O}O`q80- zszV-y{f*)dr0$TV8(3EEtjTt4pR85cMvzYiZj}zFK{m?$1gDSL>Gr;c7($IWc<_gYZ=lPrN)?cuzL5Vyf zYouCo&eixz+wWa6mg`r%dipnwX?MLR893T%qCq7%w7>D=kwSM)s%zv^-jzjxov$v5&SdW)h8Ip z#%fdYL(}knIy%c@^t<9jgtXc{A20rWD@)R-Atmahnf;PRo(ey(&}G24gLlem>*|!* zEXIS@BlCzfMG_J8*aY0=QkeX}f+6aO1f0mXS3b|GP#pK>SlkJh_(~VE)$|{h4Fu_a zzF`BWTpue-#>)@liG3?QeF6Q1bi>vker|@+c6xTvvqhLBP1YG`nP;nT-i1+Gy|4^M zY?dHKD_%vlf;naPQX$8a(jP(b-qWNp`b0UkkGpg%$s*i#Gx#pEL|fg=t^N<#Z%R0F z46UypB4}t^ib|Z0H%Ku-Y1;a2E>-X6q&Nlp!x{+iP)|w?UwcsuXa!47>`M<~JT_wr zxTAm_B$}ifLheHGBsorV6x1-u_Ky-@Y_$Mzw;Q{#o>g>(K4Amt;PyEC%{4%Ef&aXB z@yL51a;sCHUA0-8Z2S?L|$4BB5B44qzuz#ESL_q zafG)*LVZ}>+Q&Bi2Y2iAw6%Vg`cea#uBX#49{C4072T4+*Z?`Idsp9hnv>HP_y9y` zf~I_qrN2ZX)oJ&&tiXYR5|!AVeLMDqW9u$-o>Oi9(I$cH#(SUYmnvwv4s6zR8(pxO z^4!uO_FM=`?p1`bzY#LXjs7y_06Ru{tm*$D6IEj;ZIkG&rM;Snll3ZWK2(o`3(Gx` z@@l8ZE_ppDUUAQo4nrFu2GyA5R z06gy%%5W?A%8&R;@v_uZRAg|sLz$P(o@YA=Jrtcr{Hgae(N7M%J|b3esmER{6#{-@f;T$2o>^xi%l}D?7J$Wh%`0viUyCmKoXn{CHVYGnAk} z@3oPF*sW&Vo!XGoIs+fp?0D~q;jxkW1mqCv{%*zA@J<1>aE4fN!4&tT_jozwcAS_| z(h12Ot18PeQFElaFwHnDv>7vMuhbB2{zTIY&vtpH%RmF#6%cf>4;Kp_fz36bCS@)lXo7w%@VEC^VK!0-Fkh7#G-jY@7Y4<9} zBZK`KVc2Bdsjj9hsE}H{og=Hh`iZ1*E4s+ug3qHgic6;i+x+3>92G|fd0C{gW9qdl zf1`u!PAm)Wbu3SNyFkzHe)|Ur)L)r(@x#&!?)yg;H4vQo(QfmfSd>Rl72)OsYYM*#phP1_D%X&8=-<=Te>K z-;7vv(KI7#>YhMmRpVEfj5@MxYCg*1fZD)iG#v@-xNUoeZ61ov7fL8gv&bbP9#eI@<*6l4UGvXuA)B%tL2D%lz zx`D@=SfZ#+V56^`awI_k6AI6#Cv7tc)g(U7R=L5REcy#QVT3Ezn^OS>)PkaRON#7H z5~Skj+-P5N8^}br5+)jDny%)Dg#h5o981T9*T=+Y3HpTCF9m)9f(qs3W?hfAJGYyR zO;=9!Mv6bNDq8=|(UNuDJoK*I%@*M&NPhQH8RHf6;mlKqGuDo+AAK$LL}s)5Ih%hz z^liLPQ$oX3y$3ZH#8zv-6DYJ>KtFmFvslALtt#XXF93K_c6OoYE^uv#M4Lm(tWS2{ z0E2UV;~DM4F)iDT`{=v@MCEUS-hA?+wcEx~d@u+P1^8QxM{yY#CT29-EcFSC91gyq zERvP$oMN>M>8~y$?WZ()ZGG-|1+IKMosH(QIknb5q$R(-yH~GrfJSqbfa^n8AW3mP zBvVAOcYBw8{qp56tP9wDcEEBLqdTmJb_Xv=0bP_L5ufSv=N}dbz`nS5epy$_rG}nafmUv;@ z06r?{Ot{lD7>0o4Je;Zbt^EBPLO?Tk1yP6=yw1+fwkteo%M{^RdySxt9M15y!0AbT zHZOL{)q3YiFC*goTL`I@UQE2dqH0W+t9HZ2z?G8Yx|GeC=HQ%(JAlZyw$}e0p%7uw zW*<*mqlY&Uac^LzH`{!oGVPb3KUOuG}c0EE~gq+^efq<5nFQ3eGFN zdV<6R**<`Ow-*wJreFF6$22|uZ%ku+6eljFo&t9K^BrBb&`bW}gH`zfMxFQN8vy!Z zp&<^~QgQ3|b?zbpL0Mn{RFL+r=wP13sB23`Jj>?(=~cJd)2(~=a}GbQB%MpgC^maa|A{ zamRxLhA)ZQ#=-~pt>f+_hwDgZs__F8?B4GSsvT}SW>h(Hg}nWcqQN7sF{slSa?A1J zRpgo~4oPjB37E+7Pic1qGVq>8m$`->2hQcbdl#>1ug*11hm14L|b#RO!TAlh&lNpvXsJ6TgAExYU!Z3?(*a9rw1CqNR`s_P) z-C!53&2%xc`bJKhwz&c@TIVw2Q`fRfwty13WmlKrEiE`+>Ffk_uZ*SzKbSqGLh<4w zsSYt>8SIr_M{#zFsJW5Ci zU#(bX+v#--$51BvWU7Jgcb=!29W!kKwW)=xja(rQx-OlZ%bbfIRhn^ZARvLecEqZ1 z<3nCvN{0=r7jR})`PsLCbV-kS^|va&Q{%T8bt07&k~Cj4b4(KLsfTSn@)fQ{B6db8#iuz8GUrhAwR!i*(=s3pY7x|5&-uuoT7@R zhQF+b{b$ZkV#eZOkzrhIk?<@2EIow+* zG~5dmCPrGn)~$s6>c?4~X07_}yg~xfDfjh?AK&+`9q@ZX{Fi1AUj6wbMfB~PZ6@FJ zuCG5|_kpE9&6|{=HEhcu)x=kn&fC2C%H5%V>g&2bF&`!ShjX3CSi8j>u2z&bNirJYzn1JgO?8r_6G^@u*HvQpKXutG~R^x zi!7r5F9Ym7=cj#J7(AiKu=-28AH8&v{3n>G$5HU+rmw&90@TbfRAK$=>uE}UN|$V@ zpiDD9U)xv(+_3igXODQ(i4@}?wRx7()6hY6K%LT_+Tol-@4JUx+85Kz1k@V2{#~z` z*0K@9>RCobtl<{g_j|Q<^eQpHg27kRv<9 z8eK*4FS^tij}J!~o1kpE=+I`4)3CI$O$YU=OG}W>W;jxGD-~-(BqTiAzn~v-%d`>v z?XG#TZ>kLmLfW0FpOHMf$u%%HSV(p;tW+seIIQ!@uh~Cm)%kxtt90j{P3p)Ecas&E zaqQw8UJInxDN(ihu|>3gT<YX+xUVk>HWkqWtsnI7yec~=x(L0&^?y1 zr`L}%UW|)Qhs02(e$3y^RaUMu;m9zAi9%{rYHPa-*FUaJOnM!2=4nprmL9nLOxcKciib#@bf2*pwDi7>jEFw;~q@IyAeT?7wm-mN8W|@9T%C?rR61s7fg30?Bu9=qoL(ASZ zI^+D!G|o~I7&x7pG|EiRGp19WNyRK1r(GSH9gJ!$^!=I)o1#8C9kTr-kEiy zt<&&_E<$Fgy)qw#Z}I{)x_|ne*UAyY~!i(g&t(kp0y^T};BKmHRt%^!yT^4tH^^?(1}_5L?I1|lDw z9P@aoiIB7&+xm{%&`fk0wdP9>UVk}_+f>}K%}{mq=bt>z)b~s!8RMulV=q&V?8t%qJE#u$2) z2X;KjtM1&Z*U#X(X=aC>7^`(Nbf8ES0oeh89c(4Y#@zh{vo>LR2 z^>_1b2;KBY;U8CD-b(BU{iR>s>iWC<{N0PKqxICF>USv^%sx3T*Lj0?Lf)tJ8waK* z7IyV;ci%v#DdQg^akRK;qzOEGTi&&%oz7xM-~qRN`#M52qx!vYK~K|YV^x`2f~~@r zWrv)DnW@2<=U&Cln>YK^Icq83Ty$yq_Aj>1^XL0m#tvTeEm|znZTg=aeSC&U#inju zu|f&=l0k=uu0Vo3HVOAY#&Ep)^Rj!(GHg#x@V+h=0WtU&dIGp$IzRd`(?L zBmHA&BO_gE$K#jA7s#oscPxky! zs2{I*g_b`QhOO0{b3w?b5zBc2x|8?z>5KWv{b`R z;!=`gsI%W?#>e+L_M4p-t=q<#2%Xk`&bp~NtF9WW3D{XWq7!AMWc#hM(?H$i40C{T zAhYcrbsT&0`({A=Y5jZb@Je949fMcx0z%sERrig@9jkio6$9x5Ucc^3DIkSo(> zq}VStag9kf4t-K`v-UjD)Hnws)P()s#yhd;9G0XVuCbh#<$tD5)A4i2`;j44d$u8; z0cVua&Xj{P#KuLtGWiHKNk28U88L-?U$L5I7N$upb?xYZ(`-@O$7|12OJ*q&LNQ9U z028Or8LOfg&Dw}SA?Pps%MR=AH%$+d&@54u`+)Y*CVIH;Kh_a;j5Q1kz-gn)gNy` z$gK(gl3Oio{eWfu>({0c5>}Jiw`tQQN#ihF_dIozx-#GWL1)iSsctoWTwOqYYVQ@{ ze@sB*In38&;le@Go#X2YA0#Cis+*J~&!W73zhf}4V3UeD_NEmhGffEnMBO|5qz&Yd zcg&b+xO{orlG%qNaS92CdRN!K_Osl$ZWtn{n7gvvT`&0}GXn9+DjmJPkFi#Dd>s-H zFiE}9f&X*B^@VHM0MFBqq7@qEv4z<%O{H(^e||HvHVOSiJW8TJP+>8Y_`bNy7&X!s}_f<=&s^8MnCEbS^q@u6Mq1RzkwG z>Q-Bm*N#06lLELnpvXTFFj}`V;aKl-KZWG42axE=J`&#=KY&4z?P2KT6l1ryV@?rH(R#oAhCjd`hhy{;Yj%8i2BWH1c}C+$m4FkG zjP#e6yb#Q;Hn^>#ma(@VIzrUuiGZ#e{=EijC#KY-&ge40PGK+fn9X_gx1&&O(VPC$1zKZ*Ydd5|2U2KA1v}V0j{V2 zT;){F7GbDZ7Ug1bl3=RD>o06B@)_sfO1pp{FCb#Dyr`nE?MG%9R}r}Q7%1-!e#6Y8LL)Z z^@{boD_c91n9QuKFFP9hhLlNZ8fc{@0-W&P1IgQgCd(r_zaU0_yV1sd78QE-LbVu} z&<)VerE8?*v4)`k6i21n6B+{kQ#OO4Vw$3bh3RAU<@|*$yzbb-pM8C-yu9W@YAiss zKU^n_un)_a;axS!g+TMG{^huLdx7AbhxL^C*DHWUB%~0rtXaYG;yr%R}rB z8Lx5PY%HACiHT)CxBvhhG_mlBmsrRW3Wx~8M{va0Vm~&Oj^bQ2js!z~2eh)3jCXPp z?&DhWJf3j!v)jNutB6OBW@449W0T?szD+`<{dWJJb%j!frO%T|H{=i+1$T2^CGIsa5MEcqf{=MmMO zw0G}Rd#+I;Ee=GuJ@6rQM0`=I8Gd|D>>IHp5hY|5i^Mq)9_&+O6u8vHvX1>^F9I6A zyEyp9;-yPh1*PIc2YA95J`3jfMQ|*sz@(Tzs6Dvf{#X-+Se(fdMPb)^*S>})CP7^C z-j_4$tusf;tQ7q?IVJ*m7eN!5tp0X#yG>5Y;3DT)+j7J+1waX6sa378{IwF<_qt14PGpZY&-g=%1Mw zoj&1;qpS}XIM&9Z9GU<9gvq9f^f>l6z^Uf>*($^5D#N!d7TGDScJK{I_5bm+KJtGb zp219phGHGrRTTJu3`phhxylP@A}P8t$SOBP{(!&0<3yCa*fwnUVox6iOZW{jPHuk( zBm7i<8RNV$$fZl*tAbh%+&q!EAVE@2DcfFxWGous1eCAqnB8Fvns1O!mfkJu^4`i4 z|E(K%^M58fBYj9c3DJPf*8#=9Yq}C^0C>k@u%CNlye?9}w`n&pBWF=32^rw( z8>&pHZU(tP;6RfH=|{(v4L42!jR!Q`);zcnEy(2}QgDI^?hbM$!CK$4u=Lm%7@aR> zCT)=L$-b6VTFL|Onf;2>oA+P=wV}CSWNUf-rij$iN$2AIj;()PfG+E3>ejep;5S z%ep|GkT9#m=IC@N=zm_Mg=l#F64me6=Ok2+cc)&IZ}2aoD#4cg>@@?N4wJF1m^3kr zA(QzEpg3Ya0g&NsQJ(tcSUY z%gL7rUj>Pu@5n-gyQ1KcZAU%m4XxvwMzFkr6Pc$5GPt(b+p{!eDuRI>90O$wxg}5i z#z6u~f~x79rRBcD-NSng4JG^KXAoGLQDs~B1d!notQhevYb(y?sD_;grDFr3=Kzla zzZGsoQ}FrLkLXS>!a}c0n#Fg?*Vng&d&7|kX@`i898B-`0K(xmqY#?-GH*f_eSp0) z(AeX0Zt{(I@ZhS#b^|DN@pRyf0{fPiFcfve|0>@Q?=mrtJIe{=w#RVWH;esQ@wv;E zbSRPA>wkX9&XkeB@DZppHeKkB zXkpHIg8cw4f(>37@1}hbN|X($59f*qK<#efzTscF!>~2qw|84I0C7xV2nBy^Kb7(L z3V=RhX9NAwzj)OOs%$GofJFVFXYHabeu(CSGck`{`GuOlXRyyKon_lw1mpkZo zXS%!Uy9Jk34;|$=GWo9i)S1%!+8T#jM{8o|iGFuTv5h)gk=iO(8fw{JInWcl$geN? zyMs~Q?d&|+DW@dADbk`$sdPvF8N4Kq4}a2!S2yMyHM7oJNyi3vRLbh-Jg_WiEi_bT zo%8eFIqZ-O_S(0BFPhjdk}Y9+_jWX7ak|Zz(_0qLd8;QG8Xmn2M2(ikZHKUbBQPL< z?lv~hGOV5)5$GxbFX$Q4Q#Q!RJT}k14jPw~k;UYV6;=x?9Xb??N=u)Ze|OP2a6owV zYEG=g5I+1Cwm_8W3q>l4`y9Dq)^Bngd^Q-wIgnxF)7waa_vPh>DjK?a-$)-SW8)Xc zru~tN>7g`35fKs4lIBBQ6K;(ppiXpL)le02N2wV2gHDR& z-{GEbc>O>UPct7-lQ2rr|C|J6@Xj8O@*F^PxjzC+4XsZY833(?rj86l$<-IFUe+QlqKHd0^l9w zKK#^;yK3!nM&_Ce-NwKN@4bGl*FzLY+{U?=ccBo^v~Y526m{ zJmZ28k^RrFds9**SN`S#5Xci#)7-q$Fij&u>e(i{&gD`A*7LzM`U0soWm5vTr(6n! zkNRQ?H;ncpJe;(}GB-y0j)bcIc$>5AxYXQ9%bl}7#b_F3<*8Une#BjZx6aWbBx@4< zwZMku!b-~yHfBHi7fe2{#Fx6yK|^_9(nU0cc5{W9N*{H*H*X{KIC!Ifm}QAWWgBEq z;el_lu_F`bqeDK?C?{AtvL?QX0&&);-&__}xBfyVwdADM8Zp?|vR<|{u30}qYT77L z*sn=idq~!Ukp|S;ldKo<}x*y-D}o-DgNmZxw0o67L*ybTjO+v#(wn)s{GS+o6Y+ zD+)VAG>oQesQy6Gxe{FhB_I#6^*nPOj>z1$DAql21UuW>7nD<)Baz@ydN&U=*Z0`F zN%+ATwXNGA0^=N2&iEvhP)x>V0qPN3fw|-QT6;9zLvW(%w9GPCvH5~+IFDX*V4u>- z=aQe$uel4@&m^=sz3R$FMzid~6qWXaAR~6c^5*JCU+ghSe}m5XNkiEoxb6Hwc4*z7 zv5$_6bb`u@Mpv&6Gvz?Y?(1Njf-E&(zhPsSh*aTg&m{|aZ7ZyK@ShcwlssVF?4dW0 z_@s`otARuz16**5AH)Ywb2cU@ycF|lQ*_RpGkRRdtn`0tJ4fy;s6XM#g2E{P0)5`X zq7_1a*yF!wD5D<*bAC;0>yryM2acPVm<0s}2R}V2P^KI`>lXqjD^MA&_P;RddQ}=& z%{mM7eV9tB)gb_ZL#=3K1-76x`4CFFI&Z+MHP<$6dr zUEjaohBU#uFJCT(7zuZo2(`h`dcXp0A&MSrVA%3kB_>Zm1@KV`!Yy>pdVOwmSIGU$ z)P?Zq?OM7Gj)PoSUm>r|qux>=z(WJ?&lJ3y5PtQXC~WL* zEX3x*%nzpmE$C5cl=C&PJOt0(;@bZ4is|PS(3fEa88~Ud-Vsde9;I;MT^Ms@H^>s? ziq$0tq#x336)a)xBwB*EOTL~8nkd7;34|irosYvf09!+f@ngsCwJh#J)vcCt8R8Uh8MP>m+Qiq~<eJ=J%eq%3|0#d zB9zL+@=ksLsWa`YbI&r|POA8zN=iy7;QT1YYP-5?8~>xgfMkoFY<~sg#gx*n=cEv1 zO>4m-7NxG)se2nSxX@8=6E_+**{y1&1ZD#JwRtTeAwlsHc*&*n+pi+l6afAyGVvj1 zeQtNw+J9XlSzWIp&Hrm(-n-YxC?+R|cUz*qc2MsTF3gFGavyR&iH@fDOd0{m|Bl8W zbkm>8%OA6(hbM@jkJ@Y5?|RjP^>?cBd;ImT_`73A@q4ZwUzQ7>6AD2oJ~O|$-QDXS z0G616E2F3EI3u$R4T~pcPy95~;jMMuyeHTROBz@VEZtb;xE_*90f9GKL3M*6&#fRJ z>sp$m|GW;x-V@%J9>#D*r^rnAkOkb$mwQgU7akk^gNFOFQ;F69R7xk|F0YdR*M}gy z9D!8t=nxqj0f`6JS&dg?JDiy$t)V)%j#RM{VvkgU8%(d8rZBI-s-+YYfXZKg-Ld0O z?yQZBwzRc9P%&9{=DllNA{zO(&Iu+SLJh!g8pd@3e(smomSTA2AJ9j#sohop+k}4< z9c0LpqBzRx5DgGsUR@|5?@r4Qk(7e0aUpW^^3L+lj$o-JrguFZ>+bDU1!}0~vTa-9 zMSrvBsuQgk|H6kxlZE5B@N|j)8&v?mDp>06f9{sd3vh?hC5ZP@dML?mEu@Yat$P|p zK}SIzS5PCY_LpFU$Cy{7khmJ;f z7v(Htse0dOQxNSdMiGWu;8+M?d?|V)wK#0K5DDZBS`c$I8i^dY4)*IFFd*SUtfSn< z+qrH8jFJmkoB%E)cR&hZ$YbAMGNnwcS$1dpwJ8F2)sx_mF&WLcc7@i%qLPw#01S`D zatw|;Qu+#ln&Fps)x`13%H~Ehlefu{KH&H3_lj@iu#t5;_ebG~p7%>ZR}a1%v~JZP zm?JAVHy)Z{_R;@w71Txax~BpVud=D4@|x&~kfPCY#j4*7VefA`8~a-EIzr z`Gt8+eU@T(ut^7K17_PB-#|s&!KHh6W-NnFxZ#MvJ2Ej~O=Gxc}Kw<=zOKas6StXw;O)mS7 z*-qWrV&veM-nwVx%W1-%F9$wZ`15RA(x5yFa@j#Z3M6}j)}r`~={8>zgcj*z9zK{f z??o7QShfTEb&T5w`XN5^#i?)Z;wQn;f%jTC+9y@}@qoaaR)(Zh2a6V$fEV3X5eHw+ z=KE~JFBKBrzpi|6Gwv1-l>J~Xr~FedYH^P$N4sD_dJJz78G{b~bBX0v=iGR{!n(AE z$aQGnpRjI9njzL4|LR5I%8&_cMsI}^NlDW0l0hj69H{TEX&Mi9|^kR5CE~Gkw2&umbBE&tcWglC1`X6~)Ns^C9^~x-O zjRyWQeB4Yr<@o1A$M@%Tu%LL02j_)fVvb_HgdN&pkwh4~e0X*|XL0N(3_tga-_DDA z{hF8j(m|oR*+=~j{Ab5Ak3q-y(N;Tg^YEBAx{F}!P2k9ahv<>_09+mo4UIS0 zze(p;K07i6!3*&By?4RJHmhyQCVyr6@Tn zGO`aFNR;qM)B&h5C{dA2&#>6XGIKtB*Y31kRA>nT%}wAj<0E`;tvOZadCmC9n?qh! zJyNVD1|9!N{i-1w5$#2k8=4EcvcN`Fovm@e= zp5CpN#Tu8uyfc*1QjJ)ccbnI@cTy2nOqlLJ0lq7iFugh=9PH0Mpnk5>EAJtO2g*P- zor@xa9SNI>H?Lp_-O`=Acegy>hNTkL{*s$EE$Qp)J35587YGWmE>E)4ngTiCbpfMyK7MF4hid!w`M2~xT8Kg}i8mU;Xa!W3 z3`wg4_whCXRmawIdgquxJo;O-H??rvvu@e0a=%f0q_`O{z^-fp zFdfRZly7y9jo%OQzum6MveGy|x6N6^Kp)I88vh$Y4xHbL5$Zm%2zE~$`*D_4YNLY= zh^(f+auuTE7#_CwdBW!br2%aPK#m1Ag{J~o32zbxMA80@W)NmC!9}Pc6LR>>83|$% zAo)>GkBi<4xPArOpK;6U0Eq_|Q~Ma*3M7_5%`J{HeT)i+TnP#>pp}CVMt+cHl%wbN zd=s{YU#xDzUo*emZaElHTBovgYjwl>Cn@sV?UnIihVo$z*CnT_YC?Db@TOCtU%!3x z28H(+*c3q~w%d^Y15V*Xy->I{fHkrqM>rF8{T%x}`!x)kr0XvDpK~#TAtvPHDPHyv zoA6?ZO*a}EPN1+(g~JWJ-OGF@hTJC-3+CMub29wL)+VX<;-j9ykz?Y-D9hy{1;@q( z0^!?=tTB+LL1^YfpErp6@04LkgDa+`G7}LYofwbkT!Z^xn5-0;)+Jw-2q62h`K3Rq z^~JfF&YRAjPyiuoVJ_Fqne%fuPQGUllsCqmm+g^Xub=;*rx}zoGSoIC7viSkCxLeS z;g$u*QZ~|x*NusN4Xm7X#H%zc$5C7W&+M&mOT%VVml~4x`>)r)WQ?pn3^?|p_kvyC z8yL8VRGDIRpZ9v6Gi)P8N7iX9&K%35H55!OmB=N#ILf0OQLT9Lu0XYU@Dr*)UqYgE zLBS5IiT3-M>%Vb*l`ia=gvtvSpB`M{6_c3;CrwQ^o0)-822CO+XNZv>IiZP?POk1d zEt%N14r#6j(3{MjIg>(=)g(E`UqEv?hwMTdn|>Tb&~~L`&(2&+E+Fe@T#ka+Lr)i0 zG!HI7&>7UgExU?}H^d;=$J^C2)o-*3>LU-xx#xBj(HaKIeem~jBIx7j5K;0{jYFPW z=Sr7WBU)=wwj_XCLg}HB^{BUf52jXn9DAy5ZiC>?8`CvEVE#QNq-dv5llCWfd;H1; zyQidd^!a1D;)W->1PZ-Q0^-8LQZMg0^8F{37T30MzSi+f$?CnH4FX93jlkorDgny z3wS|5ye<{@<#X?KywvMB<0p>FTz`IF#V^j_zme7EG!10ljqrB8%a305bCGO;2TTPU zlC1Hr+J8_@Q3i@7{Cw;#hA0>ss})^PXv#TAbfcaNk$XqF-Uu;%Kn+(Ho!Gt=uC+9< zpy{q~pd|?!d@{;c=quWnG@7}2le3um@fU{xjo)X@)xM}~%V3P1tNdt3LNBtMTaIQM z6C|L_KL*q{i3uzDTPohpdk9Gct7s)21Ud)nNHr^&!oZt|hk!+v%5%2@&6w1sC%Hs2 z<5h=i4TJh;fj*JU0V3U@1d{^}rRC+rUVR;)5g=u64vaT9ov)vX>rE#?L$2;smF&)2 z>!#6BS}dvmh_%Q74a_lEBM_z!Sd)5Ta^D4&jx_Hr{i;w$bal-cNVB=t+i3D!)+}Y2 zYVYia>FiEtR>F|?ZmcEUN!I^Dm12r0J9 z0%j?F#E16(7_jTBfCt#!1tmU+B;Lb(h%hLI-w!x~LCEcC{J9KU=(#w|35ANgHdfjR zkM}-QervpLW$PurDUH&7b&w6?m!}|cb+d|!Yl}@feq10;1Y4a=>Y9#qBR6&sAa+so zmmLUK)xYIj( za8F3*3TKk#LXgydG|{g({eIB}fBz#(v-8T%MxDcly^u~rau##712iOY3BIpPb|7*% zKCLx>-;gTglC`6Jtu4*_y`9t37v|*}7H=(m` z2HjYccs7U#Pakw76|MUk0ibLn=??zV;xpGL2fK^l2ppzR_oQQ zxU$oKu3CklO72SgRtZ|tz!@|hlgoU5{xh$tJkeybJAmq*U8Q9C(Xi~K0_VAF0Zpak zEgY1eljGTz1~;h;Gbkixpx@+L7+gx zaPb66y4_e`sRgXTJY$xX-$qG--Rp9=-l7Qu!;Osb*WkMkoyNFmHoM+?(C5+^F!cpa zrM@N$xS6L=_ZgO-a&vQuViKJ)d@46y9tC14cKkuKo36$8@(K0Qt9Qhzum?;9_E(<9 z*u2&M`}Qxq!Y$<-!5Lq68e^_2?UDAUqdY=`#{nP^!ByOZL10j$2GZKBp2IFXSOKF> zvBZ2?apKuYTzA9lrq(kQQsTVLS!RXy8B`Yx;^|Z>u~eyOg&P7teo}r~rVzt-VA#r; zFV#Q3TH$F=?Qu9F4%Z~z>e3Kgx8y$y<_?5`FI~0_!1L@SUAMlWG+J{SXQI`)w><<3 ztrl(fCIH`$Up0k$srH0y0ajt^?BtYeni6W){0B3I&Zb8GERBv^+EKBw+(1t@A30Lv z+gc5R{^jCRx)n=hIogLT4XJ%q9N?pbXsCrF&DC$=@GVzVEtzw-hg*a4&WeiH&Lq46 ztvB!nw9tePi3aM2U$AjNIm(sF*rlmiu)#M=1I&oTv)dK4eP;Mb~gH@wWb@sRh&!m0`B7khi!d{C;{hIP&$=pB?9E5ucPEfxN( z&&p0~wVB0WO9C zh6G-u+krwdilwNs=H0N|UUkim>}qH4XvEnQ@|E5#+b+f^Mr(w0V2SfKK^aHi;=_!( zV6Oi3&PQ*MTvzj_{W_85mjCAs-W2Hh>2+~wDVZ=?FAD+%Bd*Q$Lsc@?quX z;)rVuEsC(H=rn-1*d#t|>E{c$7ah6K`bw|7v18_));Xc8S!nXl z`xy%p^`5r=!r=)ccGt{2R%8ELVNzFfGUnG2{MSEry))%e)*j*gZepaZ{WnmWWvN5C zXX)c7?^nW&rJmoH>_SXU#bj=7Sx+zIUOM*=`AD9}oNyc_A-g*1U`jo6X(X`TFAp-XA9U z+gCPih1#yVrDc^N;u3X|rWXup80p;YK4*9D`gqgl2;B++*)=|6uxDaB@Dqj%fT#G!mumn~ z2y}$OhtzvtL0=#+e7I0hJsux!o!{K`lr1e>e|=p3G>Gb7zWjN})uV_hz^UO|?9nue z0!lngG#D45?l5#am5JuwLrBHV!x*epL?eI5%PZL+B`d4Zmv6$6f!*X_T3C&K zkEc<(_!h!diEv45@Tf00LHuDSW+8B*GqGxH-u3|Twdsb!$z3C1!PqW8b18IKr{nv> z!-L~>Ul0(|`uchg;{M^2_zLgCR{~J_Aq2fYYt)1`B5yE*=XhbHP}9i%6X$FG{~J#fH%Vzg2D z)%SqI6rz+g+(I;IK zdHnTG4?utg*#KNK5Fw-%ClwJDHP^_g@8*MBKWAIsTX>fZUB;|2u(#~*0sRe8w^z*j z_-=Z6wcR`!9B83`K;L2-pc1t>YCzYO1=i!qCU4ep~ z4tmI6@nt#*4f9_;*qrC#u&zyz$O$II070g;=dTl9Vs2*kS-bUxy@9 zSU%{#5}Oi_Jogfe9OLU^+`{52skd-DG?46k1?;>Yn|sc=Dy4>2&oF|^txiaeKh1=L zF-FCN9KCK7Az3_AaAHuMECV&2Ox|?K7fjTmojPT1j%fi2GAe%K^S&qPjqo1N^Ukl& z%Q4Nb-!>E?qK9P}t7xzFT5;~>sRP&Yy$r<0LPV2GqQ#o0eDiD)8y%^1PO@roi9$dV zO_9;BYG;eW+Oof`1Rr25BoO)+zTf;u+H5h3 z;E&vR$}{|%weQ`R&&Uxe?)QYYbO1J+V<#iVO98^rU8Ue9#_0x=RfXTKmLl7>@mm3I$S78sz#wx z7QMOK5y{l<0_rAz&|B8>T3n62|Yu9pP8Ki{j zrjwvLU2II7kDI%ID?(~uJ$4d9+tO6rJ$j{)BBy{@9<5Upjs^%T9k0*LqK&J3syy?e zs5R$|n8~G%6W79*ce{-1p5uPU6}E=k99A?Ql6j#H<(Z^48u`fW{NeG;%8~chY0>^E zPt%$G%7!zgLNE_bq$Gs(I$jjMRt%|$7u@aU!KP&Eu3hs$L~aSlsvbPOIh9OMD13*` zU^Tc0EdE4*AF+rp{|5h%S11R}TenUIy2Q(dzHlk5(|Z#edd1_>v53;wLr*m#cAY%A z64aV-*4zb10=|d=i-t3OfDELw$VRLT<26qldnkj_7DL(j#VAeEP?=jl8349e3sD%W z*iuzeGVZ_w3&Z!Zz@j~mH-7(q7jh2q2}iQlj}uj1Z!&;Un3S6okrXF@$U|=+zlgO5 zEa1Jx-q3%f>MIl`bIGKLHXmROMw(|v^P@FKO42%Lxh|525wPCJhvT*v=9j4U(<>iF zv#;$OkE~b}7|(hcoEI@Mb&(NA(%_PtRlEY+4j@`Lj8{0n!`@pjqEX@3qYs{YVEKO? zhN)bq&Kvr4Zp}jsXBSeQ{vsm<^r|YoJJ6F9aTn%qRJe2hnB8>&=D2y{4;j+TYyLnl zMq3*|2K%+EGyp%b*5swhNsYJR&zDoT*MFP#@N52`VT;0Yf_Yr}>%HM7XB|o0a)Xxn zA=i4VZ0{*di4+;*?UoPqkaKmZ`nr2K{`_U>?+|=^JC`<>USGg0x(O<(yYc|k5eE*# zY;FE20l z3H&fq#mt|FQ@&Ru0WxdwT^DC7!RBi*%?9xb;2od}-DU z=j_!?ZTvCQOvlO`@=+DYA2n=12@H@jlJs*@0%EL!qM`8;H%1uu)aNf>9tFs32QAgj zb|cVH-=$8BodukLX2F=Ru*z3bGHJmJVK**8AHrM7!wdkBZo;urisc64TZ>639T)Hw zG(yD^+{~knnQz9s_@^+=F9wyXv#)Q_&qLZ6kBbd6a&9-cd*NS`6w0I&^PIj1{q8Pb zz8e1JTK2EzvR-8H-GiRE#@-5OsZTTucsnfzEUvzmwI9eXAtzGOhBMMXuX0UO~NN>3(Ufno9)`+AykS;dE8 zsb>CykJ` zM*o)gLdou~4r6lc(~BOA9Czr$WTUWgW>|X>@L?)klOQc^jCFpIgYwmoGQ7*?N_O<6;IQ6Rjfd_4*rF zK)JJ5JFZqlff_oV5)lFiqTf^!;CtJWEp}b-q@SAxLkO{!D!b$m6(Cgz$u` z@45@nC0co*kstZ2d#tzr`>_`WE?hcSS&BirGD^*=hBwZB?NsQNkl(w!QM3F~-GC2f zMvfs)iEQVtzo4!R&todRdG_Lk?uIH2@AE5vKXtup$N)%k!>EssKckfjea}(AIi@M= zuNyjiKQ5|TL^?cc&Uq@siJvMPmJ#<3j+Wend`YREo`FRvp#0hwrG{zXV-fWN^-nem z53UAxX*zzb6E_srI@+iPpgYYYGet_r!g^tKt81O!h)H}|y0W^@Nl)+X_D|EC@>EYz zewFzps~e%AKx){X9`l=@-vXe5xHL;~-n%eXm<2Q!Yk;|JB$9|d3h9BfZMIQ{>RFeO!#4r~R<)vR zEkdyG2Et(5otEcn#vDFp7~iU(pwJ2EY%$*HOju#8flm_~kj^o8f}tESpegXYIBGsK zk5r>Quf^`M;RRWL5SIyy68Va})cVk!?3RzTk|Td+uWjl&ogqExSF3-Q(ZBjO=g7PASq zh#-#`M`a=}?LQ-b{JF{bEHJI+QoF1TxI*@3Y6wbZUAWV~{)5mb#}#4>ll7;H{xoQz z&2*rg$P*kZ?B(eisnAN)G}_N_DsWq0jiX`gi+GH&2<6odtb-)oBpWag7C*xtJN!D5kriDew?JPTZt-(X|9&2vlUZ^K=N9!bPX{7M8moqeCyzXIfL6nl* zZpqZq^Vlq$f?>o*=Yhnk=KW@909eK6Hx$+hI2^hcp0^#8$DDD54rRo-ND%(+08B~Y zz%F4jCxiv=6;dm~j07)==!{oe9H zcrG=zsUsogbxw>puEP8`-~0=#9WyU_1bbquj`g!z9DS4E?=jnq;Bs8D;VhJ%K@H5h z!ec2H%_igV(N}q`xWd18lI(wD`h)ii<6_;4>9Vny@oHhXzsKg*#pKq?WO-r%VVK0_ zK(`{hN<;{O@LJ#~5dPVjn6zeQ`zoe0Ka(TGcB$mcP>t=@9HTU1FgvM2$3&zk)(wu0 zZgPIeS@ZS+eG~k`POp|X=9cIG8MwRiR7qb~l#5gA;TaUCXgtL4gnU zbBKm`cRbL*4Rp4QijWWh%ce?UJc+x8*VkXFNBJn^`E!})dVRU$@nxNew5`DduO0cL zk+yXBrh$$Nty(o3w;86ya2wK8k`-vCfWsVk`~b{+2gT4{qQpuk5*tY$ddLACe8u@y>$A1K5F(DN_zbktdTtMscUB=^EWSYduq>nA|t?fq_gc zB-dad;{$sT>q`I>({OP-#1B_cRCJm0(7|{nltc%8wA^(=uf|aC=sNl7s`aOF=fEWD z?uB+-T)20Z3eQ}9+4Dupa$pc#byX{k;d+RFuEowOCxca}hiLYM2?pnS} z8-(=~%2>+@OmrdU(~|Wk<`A!t)~)!)Yk36FV>0Cm1Jk8tWMK0&D@@a~EfWk80@G?> z9)3XE1wSz;E%R{>Zx$t}^QBHm`BX*T8|;9vjwU!Rp8CNDW8U3*U389exq8rnoe8%l z$~b_tE@>%0-939c>s%<03?GF%9ZYOi$n_bbdid|0D?1)|7ut{6@?*24R>!&&hTZK= zoy1tTV%+Jha`ONu?m#f5(m|&46bpR=n!(-YTn>O(!vw=d@Y~575?9h*9tcSP8XfFf zg0zK0hYvI11qUWfEm4%o$D{v-Ji1!*l%W%dm6=IcNAdSz_p#4)>+3}itG*W<844VYo9MRHys0h1cJ=Ns6wf1%~YkN!zvWX82?*Fe217SD(L z8-w9}q!dqupII$F@r5z_fE8n0Aj8N~c5x~;(}|c1LqL7&VqIi&ajAC4gUpCh%xn`R zGih)|7-h(v-C}C`NaJO^sj2B36)wfgXsqr#VMeXNpt}te*PG~M#m2_6)=PeelO`AW zI}uV#YdN50i&BPLR<|Le&f4q)c1Rz?lm+oTC{+<%yv9Gh-R(Y~o@!MQfAhe{){)J-+ttO}r;)Iotpq-+A;%F zu7EkRbrPZL02R>NPUtCGre^JZ_Tp)cA<<^1PA$lVC)&kq;{{P$R9iG+G_{PTYN)2_ z;8OuA_1N6?4|@Kf;7~Owj{2l0A}U$~^y_jiJg^<$@ykAf12Z5S#eeuoTneJzX8qMn zZYcJv+2`*V@E6!nPzY2~z`dGjO+|)ili3FkZ07j2Xypn9CIw0yUJ(!i5=X#~Ikb3B zY-eqE?D&2~-+AaeaqB<9cILj*b<6AH&Zo`H7Un)QLb}9q@{0b3_$rcm0*-$Me`a(O zWs&(_c6f6jpUNH*V7Q{Lp<3;pPN&VG4&5BUL1G#eD~zBnZNY{L$%NAR<6Hg{D7)Re zw=@#d@#A6 zd87nm4$r1uQwl~|*r+37?WT^f2* zVn8D;5(iZ>xmUYDrO0L1@C7Dl#~}fMvSD+jY*&Qmt@*Iz0GD3}562#xlV%vJsh1_+ zW384a@TqshE^_?UJSC{Y@q^3tInQL6G+?^KuyYicFjY08lVHTfG(-xDT?0;f5Umos zix%6+H;c&xq2BED=|6IbW&9#4r{Y@)qwhSrJ!*`mZ%YWDA1EI?pa{cIydxPeK#I zz+b`mnEVcry|x(d=D}3wA;gJv3$x7=KZ2&jwC^BRJR_^MFpz zHvp&0H*Vwse#o;SSn@#JfM-H)|dvU`{3HWXleHA82;be>3g_g*}1m1*kNHa zmC>Euh1Y@?Zu~O6qO;HgEy243+>Jsy@W&Dv4isl*O9@`)tUN?ac{D(oLHDroYR1Ut zd+J<-{*y}Rs}pZx0M5x8q)wuy8x55AU#zoboY52voymdD^gGCs0E^su@7F~|MVX#H zeHolNQ-0w!AGTE3EBL#+xpm`uZqq_^GYy49hiqdQ^ws0>8iwe3vE7h}#E8K4grffq_6vB!G?d z=x{94aA0*39(v)%r2+jt_LLAJi5&96;2@S9_D`K2Q}`|75%`9H4iL=38wGAo{sDFCdFB5_}j6yU{IpmQR8EQuBu8D^Up zh8K77gW~4Kc851Pae_P}MI5lN!bki^b_&KHvqXmd z7VdEZS+bfg%1eiyPFAmNZS{gK7cj}26Y)o&r&NyS#0i7A{;~O95;PN2ndT-m?vK9; zF(zRP$tKW*T{_D(wVsfY|1?3zHAaRkc3rm9qBjJdiAaFtV?N1hkqL%3S8 za9}tcoX2gw3r4quv4LK9KywThr_lDzrrO!^Uc!d7^|BLlr?u1Xw|~B!@DyHie3YM_(TwacKC+Tj0n*DCo`A z6fcUk_fgcXITz>O*#PuU2U^A1Y-nyb&zVJE|8_-hWybjv>boQ1n=_*;t#c|BKv!L z(`g_ChJx)@k6&nMy~E;u3K>ES82PnPOdteA=ZrRzB)uq|t zTB~&E7x?i58|MGk%n6OXTfr56M-W=TI^2Kih&*c9FNcY78m6?NuDE{$$6Ik0=JWAA zWhEP3-}Kxcb<()6t!y96!Z?ID_@5|KAoQY~#yy0%<`q$}Kv{Y3|VCRFXM+^Vb z54pJ+`1{jaufGJ?NV@shQ7k8r0W%9yu&qeBQW|a&&W(d{$c7xb+fY?N5wYBd;c`S_k4p4i7$$uS6Wfi*09d6<($2Kq5G9Q{^v^SJw z7jzi?G_J>!0kC1;LojQ+!n*0rLzA0_ep5C~>L0LvV|2exD}zz!-n*qmLf_ax@;4Vi zLLzCC8wUsJSpb1OU|s9!?(54ykQK3-_0MOE(u|C2&iDY&`(WP!Nq@Mzo}UhK!|x3?(yKaQ<-Bf1g~cIU2i8OBBfa2!j_yakG9>#!D|F zl$x*z!Rtq1AN;k7bD}RTlqGN&?+OIIlQ^)8cy^)Z1Uw>yaY*4L47+;WZ^uiqdaDKT znAYqW*~~U*()6ABii1qFaw&t8AMm%xR}XOIe%V*9{2`1C30yX+*r#J{ zl(%zuWbg1u$a5YbTJRkr2os1t&f>6zfgM)Ps5R9-` zz2Fk(7jar+4)zkBlHG+}@<68!LWyQ2KooqW09=8{UmEShLBW0-K$V;GShu#@trmUYw3I4o%P1Vmgjx!pV|b#;`R#2jQ{d zhwpzu;}m%~3Si*Qo`?ZO8MfkV5*o>HYwsPy7v#ycDa^wg1Ax|O;cE2*+ac1lgP9^*o@84GwNsWrS z))SyAcaUYZ^0dtD7MUz^M45`$(5h*G$9olavnT;cStGo7faqa^zDE&_wb2!;dIZ54 zU~4%n+B}frvix;m0kdneRm-Da?m{SwpA;N$)dFZQvI^6NeaZXw$MEfUMism# zH01MAM!UD)cOr5gXtLk+mPZ8l4&3VCG)_G(Zg2VaOuOv?Ce8WT=2{5Up zau%Qe1Dcm@SjwsNs3wR6!PPnpXU}B|X%7ahB4TYOFi&`x%7*xS~ zQb)Fv-*Dk3=Ta{0IJ2>cgeEUsc>)$Xv8SnCF9P%{5tPsQR%*v^;n`&5fs^R==nySB z9Eu=>#PKJ{F5DWgIHarOLwU7C=m*84;4b}ie?J7hYazbex&FS?>PaashP@E~#vIf6 zDcv&Ix?T1CvFEYTQ$2J3c8Ny82p>`o{ONztyuZT{MkyDuAY&*g2?QG2uLM}Bn-IbK zq!?o|BW?;33I36@UWLAIHquA@X}Ur77&cN?k2rG$8W`OjfT!@ z7P_=~rfAWp?WWq7ISNItEcpAJ_?_@*)>D?Um4SYu*EAlBx59tNJ!NTaytrKG3S)W5 z=BD}JSNlPEJTuZ?vthv9ebLRl5vR%vChhswA+6$~17m!?w?FnIeQ#amXB1ibkH2sq>r;u@4 zknvav>B0e=Ts4*tLa&tsx-jT?UeT=-f@J8_rwf>`i8YVGK^O-4ts{xcps>b1DHFPegGg8e(RS>kl7miXw!k4JsHsU&=4-{*x%6e$X0!LXflKhqWQPx zb5}&-wjnpQpvHi}z&cQ1RY9FvaRJ65>gMLrEA%)hTj_p2*L4g16h}RDsK~Hme`WFB z>NET=HCm+5*eA*5%70vG`-(h3;68T`yP%ekode&p?4pkk+i>?7$!A zy%NgC4MzF-N4Nb_EH)+2VX&rQX!nU8c?|_#?w+(E^BDsjE$1fMgE4!xltisAe`|RV z?+RdOsf3T&+1$CdQ+@jV+Tqpi)zCvF`w1KgL=`a->lUh%vni3%=c+g7u7{gEd&X!l z&I?ss9n_IUH5D1gy2^A(ic`e*ls=vX$0iMU#I7T_>Gb zC6HwT$Z%=+kU4>Z8xxYiaT;vezL0-fEe@S3+UOt7uWaER)g-G9 zAFuSdN$n;b0lg+d?IxalVZCdTu{FjwVUmEPHe(YLuK|*SC8aTOW!=S_`eB&xs;6G& zK&~%qm_x8fCI{K8;MIs>S~`2Ovyx@o*Mu0r_Pq{ZcxaKs`OV%Xv)sczhb#qy@!-n< zlZ+~tiR2S{sf+meB?5kENE)Sh6Coe5yai|G!iU8U`MI>C|1#FM;Jko!OA&W&;5hH@ zVWjVFhDkrIl1YM(g;E-&TwvrQ`f760?cuig{9eP!+$ZtV@7X{>Ng!K1Ph)CU`BL61 z>G%<3lUM*{er%X7UaO$vQAVkT@!}lhNdZcsd+&uyy%+u^d0f|;{F_$#uRc~=C1W%9 zyfG_7Fs@*DcS*7W-|>=-=6om^w9B8vFIQCKmN~lEJsa%??>=A>YCY@MtyAsc2Pi9J zB(TGHKGZ51_*|_VCh6i%t%Ly8fF)`%5xsbzBX7I?9ijXqAedzpAkFU}4*j$bKGoO; z+e>ZPH@ah&YThciU2E>fM-M*n;|QD^KEPlBmD`^=6b>xa3>(r%8(`FAQ>6n}1hSqTg~1X5%^4kBpO?j4T3C>aPs^&M zXcjTIXpaDX_c5#)0W7%!2dm6Ue84t<-QHnxf56!1y2=q zv_Liyw^i3X&{^=Q=p!p_PH>o0;D<4g|j< z6vB&rf^acu*$?jM>981-t%!tvf?S&!Wl$BVo;cB9(1;bw$Bdd~vxz`6{t&C+^ZL}q3(hZ5* zro$stuPi>wSOJv9@S&s?zf3(M6yqy7F}nQEFX_Mk{ySZM@Ne}-X>3IBdVfzwhsh>? zU~c!~n*B@+571uf5{`HK#ojma$l82$$#P*Tp zEP0Ai!Hs*v>NSg13*p$J>;R<$0RpdTvMA&-h%UI}DyE)UU7bonxBZpYpM+fS$q-!d+Uq;c~fr`}XZ_>84AiZ#`gq}MMpN9qf7%ECWMQASk1t8M#v5#=s z*3HY=sQYBU1o9e|q5P^GJ<3R#7PkTd4Dvly?hUdOcm8eEnZ zO6{F-5jj6zwg_1RY6flD4I5TsBFP6bVT;h#euZb1M(SY3iFi-^r5I>2-FY<#f>5}j ziW`;>Ey+_EtewQc#kT3e;R%526oywZ{33lU(f(Viyi{{MvRoKCK+kRH?~fEoHUcrF z@!6N#h`}z!p{fPaSpN~Y%8g#C$1rBrE>yGK^}RY2OGmG{iI=IEW~-xJ%y;AmL~!W!JCIS0Kjs{R;a!>2nO$GPvh0QLw`< z+gSWstK&^)pnWa6R z!DXQE-2CP)Pu_*|OsauIA=Bu%6G%zwh~|z{Vp(gu-oskCefa4HTo)+TpH0;^Q$8Z@ z2pW6KBl44C*bb=2L<&b>JKy5b;Gcqp0~s3nMP=x9f!Q4MRx4UUPVSKSS8mz>L6XU; ztl^Xtzu5Xt4wBv-sp0Vrs_II3o|k^;x8^k+#+Q=@AF^aQvK4xc<3=_i1eOjL0jBe{xM83Ws3M@K^Em~FUOYS4Zojf@;& z8gpS2uLfUr)j0kwP$80K9N2tKFWt~TV*R%9(SevybF<~FTU1RNdUp6n-$x#u-H#Wg z&tSq${B&$miq1#?;BoAj&z! z)D-BKAwvp^%EX2gCph)Qg?c-?->#lncD=Wz|6*Zc*8BCvGA(|5({vmc1(FoT^^@$l zNt98-0}ps7LY16?ZSCd%$rS+x&8CD17^X#cQ2!GdGrqtq*#O^*|KCtp~t!#+v~ zow)q&-PYv5BMInRVI(0MAjroz2TQO3E{-P$Vz$8!*)+45KoPc2SZC*+l9Oj8^8<7=NL#QFTGyw9^Uzd8A!!Dl(6z`1goDUz;Q@H%Pu--iamI>z6 zMeuQ6KF(Lxhb@tApdq6$wAcDVrLTnv8iKh9t; zc|r10G2DTcx`oOUIha=x?u8ASAa;m3+X7`4$EIsg1{%q!5f5O{Oz+@dv<&GQB;0^SP z?lT~0jMpUA*uTGo-{}XkAS~e9o?B9gj18;yZ+d{!SSbJ=RyB7G3>+vtCW5*VK7_^` zjn-Hn?A=crrQn;AC-Q6QsaJkRs2y~ejF%Z;d?8{4uDF&$BlJHz0QE{G_bz}US`r1y zpio)tzDu6}F}kGp#~5CNfz;v?@#TA23reqeY6un%+8H{vni#Oh$4$L=RTH#(nN`7j~`x<-T{Ns>gA9wWjQf7F#M6V<|#HnxrfVl@iqyY1jUK&&=G<@;-e&_n3L+nd<-hUDvrB z$8jDfto`cL&_QqR{a|=~W?Z>xyMZeHQ9&QuI~ew@OX{bl#v5k#A#HPO#7?pFhu#WH zuNtPL-m`Ce%B{&6TD$pItrXE3p1*4P_qd&93r1E(UN3^>&yKxu(@%XyTe}u*~^D8MrDIA%nb}$m9?~nH8z%J(p9>{5bH(e-=8!N+m$> z1socGDQ7{3b#X<-L(tQMV90qXGi}b>g-)#?ME*)N~pyv^R=E-yBW=Z z0W>Iud$61Nuop$*C%AnveB{Cog5Pt`SqDo934Oh83uFO`@Dm}dmud4IDMb7T*A#eN|JMhk2>7xGo7cpn;+j-HkimpK^*N8eSi>-9Quby4GQ zxb)ADobMYe&}=~dn`^;V^{i3Dhd}MkvN>FA+W6gNt4hcjzHuEq(@QEUrU0dQg)+x2 zBU)3tq<06a!*^bzAef@<+sIi>zAu|2Wr87imX({r5s_@D$~`~4#ea^bexdA&NCc9N zvT|Ou=#*8x7|jay&3Ow6Ov;$Lq(60|CO|sTHlx02ZJXN#wC}9dT*l_0W`6N_@wk_r>{NKD6#Y zSlEz>lO|1$0IB|xS8#QFFy3a3S<6YRHB+!IZc07mB|T~;T~+e72n8T2Ot)3j82P$h zHxhLK`!9X&uHKFm!E;ex9hZ7ti8!b99OoUar+*Rw*>b<>?rKvTPUYO}mmZ2tQb9xv z>=N)AkDC=wQQh1?Z6Ka|0&4e~vhC5~j6lD+AM3grSZJFh*|8^Btj0M*A^|QtQRP&b z-W-6Uaqir&qfuteKQ_)em;Z7Y^8XR+FY;ep?6_#rA_JW!#s6Q?-fgN7injoka$s_<4gi7q(;K`d62`i3nK7 z?QwB0RIZ~=PA{^2eHz9(x!^cvv5gzR#y1fKns$SS2ydt{=cHKv^iMQ$E`UlAs@B^E z$@so)>kelmabLz1j*aP*U#HbXGCEjond$i%y+p11ahk$UYMZ7s*3|4trTXbqv~iw- z?$_-Rp;T!={fig$PQN7{uJ5?dDx*hN{&}Y#>vo^lxKab|W6}nJrVYxIvED{AvH- zIvO0=H7x5K&8ywN%!5UmT#Ao1rR}iv%}ILSSOdw1bnh~XIn8_()dlpT*?#dAU($hW zt#Bu2vZC*sdjbbpjobQVKSL1i7Mbza{sLv6p4z@!HxT-<%K0^ZC$xW8ZG>c`p#8tk z6gC%~dJm3pshG=KG|OAx?!z0oExK9 zuN=Fn-m?9ynGudKsIg50wi#Vq^C0wbeEdDecf4(q4-k4qt8c7BbrBqBSREZg6CtHa zuY=@+<}8O`qwmIt#s&w&n16u-=eyb(IvAN^?O+g=k1R`uikDZJl5u4IHfxQ((!W}O zY`rO-sGbFUZa3a_qCpJ1z{H=pal9@VtRy$dFID$C$jXUT^{@E#<8*q)H%3IJm)u~$ z$Py8~y(X0lm<*E<4urkAzSdscAfnIXiBzlAXMnbLC(4OAudZ9fcd#_KDIlOUqen~J zjQ2C>?YaVM#-B&(Jl`_eoB(P|-zHQ;LZ+e>#8dN*C*)Ap0uszs6=d62oLlG46dTk|JU{MvFSr7xR zi8z9Gb8Nr-%|GUE{*U@1D5Drq_2eAA}7?#J!8Eo|U)Bq&Rc9#yHhn@hP6;`1;#77UAL`$!3V4J53)ynnyx8Jkg z`B3vHkNJZ7 zT8)AL>l&sJz1!vIzYOj&&L;0n;IugVx0F1*?mZp_VL&?_i~6YMis#wwqf57qn4r}icgQwfhyJ3bqoR`aCwh$Xf{~vOHPQ)gp))_)8)MMvy``HhD$H{V?&7OBcPNnS$ zX*^1usoLOYIZ9jc8Y3tsR%3tM3dlrE*{5dKylBkL{~VoFmaiQDm5xg~YWxZm-mq^5 zUEiqUc80ihd?i}0z_<2W!IB1#v70|F1(6lH4UlC;>X7polo>2rcGfP5r}&=Dm%JVt z(i7u#?eu!~W6)fUN(WgXD6bHnR<52<0s9m-si~INXExNhNEtsko$w>-dG9v!Z!dfo zQ?RRM>yO(h4;wd$t$^3Eb=qq+lieq>r$nr%?9kg3e>}QGcw`J#?=ikyzF3)(u+9$( z&HWR-i0a6ZbFao)F<28?+2H|Pr<+2rnNz3UqiLF~({nPyDN1^AjrhQz$U{(7aaBf+ zB6#hZ_MJQ<@g8So$9At!Xt!GC!c0>pr(*mtRsyiG4kKS1EzKDhoj7aOVJm3&AF+3! z!g}U#qx9N_`7ITHpZxI`>%QEv`_yv=I#N89&TV_+dXE6$r> ziRchieCY6V}_uG;i8OfC{gHPIVZpOp34k-SBRmmg4UrjYXV9 zXJD;SvYWw&* zr$<~yy)-?F=K%-sNX=?XkiRzy5fS57(Z;&ee&`+NxPKi- z3m2VRU;N|jv-`E3^@8|^=1N-&Tza>$#k7Gg($ z>p2|3;?kt_^z`Hlv->j1XI>g<4`aM^)JdhV6OX@9bBHa`HZ)uYzEb0f7bMa*Da zotG-!gKRF-D6mfU|9NklZvK}QzE!QN;Y0R6vV_x2aa=O4U^@Cbt8wqD%ortqCM%5F zscuq6Gho05Ucl0_#UAFeDMeoJks{1R*`??9vFYm2U6q(t3Q$}$tb1>?7@qp?b5>%;^VWDhQZKvBjGinSW|z~cZ-uhg#l)*K}Hoy z#Noy^%eC(_hYN47u^etX(X`NzrU&|V#0Vi1;-EE96{uv4K%j_c?Y_z}QGBaLp+@AY zLtb|4CYM zHLqHyw!5}^qxS2Uft6!ZYP2KtCHBjtWfBC(0|Z*n_}|I;+c0UXGZenwOb<>3_~n;= z;+VK{gg7_}zxgD}J@7@asRB6G^2(F4%{`}kJV3eg7QJsz8t9@b_7*iWnmSe1-71B? z7(e~k&>u*}7xQ!s7=(Bb;cgVdxekB@^><7J(ksq5y-&@%<%mL;L)5-?GgQ{hL7k{d2}S z7QnIndx{9HA0b-WfYiI)4>#R$DSG!ciMkAN#9`(RsaF=27QF~t(r_a7m#|v1zvC`y zF8=cB(%gYpt&%53KQfwg`&9XqA=70O)mqH*x}yzipD+_wdUSZt5LSe@3y|Rn;W1ib zV%Df_6W5cJhT=!PS8#4^>uw*Z+3K{RVTP%AM%Ow5+66zuEdxL9gvXC>3AM><)3aJ#EdA`Fvc@-;frFncQZ~qqvnqk z%^%4CqilF>JGeRo*Qn-1%I&N<^BP~B2B~}a@ZmF8OJTp~IxD9z$GMFx=--QjB7w)s zHCgly(S6m_egg8!)-%ZI2~s4r!v3CPO^SkE?|~snu0@hryyG`?kp5LohN9ytZr3-w zaW0~BQP#)}n|@*h4+^m1<AAC=|*{*U6>ouA@7S{R2uzU5~>~6crJs zcY37w1d{%Kz`S}^isAK#^mQDHPB*8V_|L5x0+A3ogN9H9C%{V>wN!O_3)40|NQNL{^-x0(Z>_g z{o~s;qu6Bf9@$5yf}ywQq<$Yh>k}XI#)ho>Tb2wiVW4YYn=)d=(QS1RUcji};JHKU$rKLKyWuY~jw$UjtUDuR` zYHK>6OirPUJaGgjXaSS{On4`7@ZYr>>-*gCWTFGt9*O%P z_{JhlSullxda6_L#eMr)bJ?i)3lYBf>m?u(5XGs-Ac(L9?{@e zp*34sdEzRA^W%ybiT-XA0T9EyBmSQQ3kuSsu15?GIGmB&$l~}FJ{Qe2#xRMS86e#b z^qKLs$SD0#IJfkq z66J|;cW2%fe_6)4I~IYLD;5paBv_HNr$nE}a7qe$I|a|#dRW~7Ys`Ihopmw{Qv9;O zFS15#kEJL7ERK#_#SV@blrtQOPp@UTQr%j4|4yvia*X6j1p*#}p?x8bSp*v-uV!k$NWk*z|L zHyvOas42|GT<5mQNU9-+UB2m^>S@+UYV;~!^xR^27IU^^uvaD;jZD&KM`6+WYJJKq z5er%LLDrR>lmKAWi#ZmM>pz*Y7e%kFRCMOIY(${o)md42u-D1Mhd0FPhp`Ba^r2l~ zO@@Ki3IYiV#n}J}Rrt32(zgF5g_ZwT1HTx-!7RYp3cQT ztVMWnsAViuqmeLurCxQVD~G>M90>fYh=%j;Y`baYst0X65qeqx?Vnk8pji=dbYU^I zf7<5+^_woHhiWQ?JOrod!tdEH)d36QH(H zhkpBrV0Po6e*>1#IY}p)vLM z^CPOO3&Pfq+iz-3^?OYFcdwME!56>zHULzg6M%+D(T_ zTto_dq6iD2BYG4(;e`W&72)2#crR(n_?B_H7I|?Jd!Cq@SkqOXjdz3iV z3vSA}KFb|9c66Wf#~uA7&oI+1umxIC=gu_)R8veC-J$Oi>9^^|9{3qvL&1(<%gS{& zGK!^ip4oatocB6w+#Y}BRT^w!#BC#UBn9(namkIqq~_mhRldhi z&z`>ZA}RatUt`Y-X?Ocw(ds4JiuVA0gdi!S{>j#(g7tgdbU8E~6(j>~9gBvx0w)88 z&?xwDWidZ`q(ExIGuZxhTBi#K)W287YHn-oqDeUhiyc3YEWURs^$#Wtq^AcMZ_}t* zs<-wU^2dmj?O}LQx^L~N=g>)@ce_K$pidw0OEyUvY4Nzdx*j3rskj~ZSMe(_#d6Wg z_4aS&W8;+t_S-i99vVnl)IpzMQXWx>$ufMw@Cb|<;2st(@S{$GPjF@4Wk7+=C<1mI zl&w!CEqgaeSJMo~Js!(kkZDd>zTBT5POY|#LYQYubt_mW4UQ^0SDJ=&VD=B-HvYYY zKuK)=Vo}pnR<=1_qeW9s8sDVCrR}La2gZz(S#=*?i#WAMyWmazQYzmW`>MO{-p%Ae zv*lIW-_)q>7w;DgBNU5T(tOe5#yPq@t~Y9z3e;|ZAY%S_&j4f{v2y ztXzGK9x$lsjMT6thbVt7oouWT!cE?^6@fWLuXVj>Db8gslp$EOn2n6hXvc~*i*9U~ zu%LI*?=`n&p3}M=An3csm^pgw8yn8(zI`3@+bAe>iK$G>phRy=X=kjnYv<0dRFD3G zHaWVN^{Jx?AHkIBOmHuy#N5bi{ShTwmN8c7v8oMJcS3<}>Xp?mH5s_~sFb}2bnixH zY!Vg;k(X0I?)$6bV@-(*hUgi$@QJOkBHG7AQ+M-RlY}{-g@W{;g@$Zb_tM&mGxmSV_cR`-h zrATG?IO2w^)G^i>vN+o9ysty6E<+a2c~j6vDm{+v0R6s~N%p?mN~Q*#j(Lgr1|9g+ z=vC~w%kDL$6o2)F@o8GF_d9rc7Zw(B#J1REZYw=tbq6^SW7b-4ez{jxkEvfL%Q9goZSG1TC6PU zsiS2ZZUaUoZ&Nkw+_7Uo;3{&Eu|r;^$RI8UM@RPr&@_2vczq>9_NcK^bgwfHk7>m1 zPkyLMtQ-=5ja%&Lww-EEurgyEAR6Z&rzy-u^>fJl!PQUaIv3XCRPh& zb`^=aw~30Z6$H)bhGWss9{^_Gm4eTA`(ENUEPfKG!mZ$r-MWS@9?w( zK3J4i_^};igl)QZ3^f7_{P(Y?e_j4RV9VZk$xQ-@wqsIv^8=mzSSB@NKp|TXGPgzB zwn_=6#rH+#ma*(7v=hKLzrBfX6vX`fD{*LLWZ~L$4BHi5`n>a{eu#!v7iiK(zzG?C zKDv&&5xdGs@n2uQy;m1C?p;iSZ_cWo!7nwB44M#kCIA*}85Als-m>udU$HS9kl&qg zLN8^>@w8gI&{eqEG9}dx1G|o&ub9fgc;40s_pJSP4EqeGRoBc|p@t2zNy-QYb^zlC zTCXAq~n9;XzqJ>P#N>sZ>#MRd%r1nEjw7j^iG-s9=P%^Y3*xsp-|g z;_lyui29#4ctAnqb;rSO>s-sWpNG_?R^;X0*j5!`<=m*~n_IYw0!T(aij@Pz>aSwW z3b-LYa)dsejA9~80^?Gf=J3E#0Vx)-_`U1M)!udr=5K{fa(%ZXaY50@h*g-FEU!Gy z){@1gl^-A=3y>Sf66V?B6nTr?r%{D4pKc@B_YQ38(xI~sN zLNEf?P`ouJ1_5AVRFg9xTX~2R2MnBFwQ(+z<_6$4OiFY9FyH#M+n(#zhM2}`1Mo#jHd$A{g1Dx zk2C?m3m%n{hk?&HNwa3pK3f=60vJXy-#)=~*p{}MB3=;oV*(wW**XcvYe#v~9RbN+$nBzgj z=emZ9OuTbu4PQeh9&6lO`*oz4aKIWyKSHr27DJ77KpsAV?TchDssQT6%8p*x%Hioi zcK#@fP^rV+D`k@xW>rc>!*)XVW>Z%h-6rq8BM95k8a^QlzE%uXh2FhTI%s&fLh zV{P#vmlu}v=Kwx^)o&=(f8rPh5MokRTkHX6az@aCM=dg(D$UhXPf|pLztzX~j%V(Ej(9MCLN4(jx)IV6TekQZE8USQ^ZfniN zA?eiNwLM8nAcf?1P3E(-7so1Xw92nL;1M}GVm-?fgG?lRw2mF`#Ru&QD4-KGj> zpedd>RuV9Y(;wIh_9Uv;WRjg=ywjT(W*h|n5zW~xn}&J9lV-hw_hsed9lJxf8D^b3 z>fOS)Rijk>@n_WRb>`A_u8Ztb^Y2AynM8>+Sts=U$PHT%db$E2c2FSc|7$_*rbx1(^wcPT(p+oe(9-1LG|e-4aM`Y&|yo^%EPLrMn7WB zwWg+V>7oETW~L3on${li^Xtv_KsV}{ft=j_DXF7+RV~`g{qV0M3_i)2`+FEp4kCf3 z21)|8z=?fq zuC0CBS=P$-lGIt%1eZ;fjAcbKUCZ7mjiM0JVFsr;=P#h|XBZC@`!Z?NfnBVfi;}02 z?(0jA##mLGY*Zb7PGmBS(!C0b0-t43;NJ9YL$d45+kVLmW9Qm*t*GeeckAjMZG+i4 zbRjSoZS6hYua8YyFsH`jPW$}TqG9=7X;?sIZ`;dpyEsX~?1mWxJ~#Dz<#3x)sF+yP z`O&NBvR~hJyl!_$uLgg*pfyYU9j<72_GORhJ#a{$Z8K{_q?kd>05HC@WL*GFyKET> z&Y$*py&-2(jBut^mJ8oz>>#fWDW|%5#Rc5<#ev1!1`S^yh8U)HQ5<_6J>th*VdcwN z)yVBC_J#g4ivG|FXuDXD7sUD=m9Ha_J$(m62X)?8G6QB@;Z8dC3XCXB)+sz*cGDVx zhCyi<8j*P=r65F3->59(91GLLB7;)8v6z%9x{^aq^LTQ;E#skZ$@Fo3Wj2?j{lF2% zw(v`}`-tvCmp19HZFT1I5C}X_Ql?ox37bo^>siQ6sQ)`v;+>_Vz5OyAO2j^?2)UdB zLmq#YWc<*y*%^}U6UPI6kpF1uKqX01@!LNs`wmT@3D9<3Wh$sw2m3$6i87etywFW; zfiR;OHf%sa>DRA!NkCg1JD+~6EGV~jgDAJF956vVW93ZnkU&m2n#U4FXp&A9|{}iVpsjz3DL1AI(LoQ7$Ey=A0G!F&Ld1oDU2XvsX5px(%*F!s{iE^heDGuVgSUedJxt>bhYGqsVpiqRJB>>64W|Ef!SGidF55VB2D%5gst z&B@fGw-*4dq5Sf(nQUk%(pYXwK=>Bt0S{FKF+kLr8L+wc|5eI5SRFb^%) z2PA)@bKpHm!ar4cOzo*qvG4j?Ff|gnZyv7UeQ!{K8Sgz{1d> zqR?Gt0$w%XHT_%K_OCs%>dhI{8U5+gr+z1fR1N=41Izo}UcS&ywM7l}VP1nBO%PPd z?kqr(shR8;Ggy3>%We?EEY7@$>muD_N>n8y;5_;81MR!Ox|hcP;fHcOf2p$LY^nn8 ziN@P7_|qTD|9<&f%H$lk#dcXgqio4fIgOB$Ir6T62`NFDM(k87e%$v^oKn_4&z(&F<)s*H1kzeYe;&yvEqv;ut_Ai%D-It z=q|l_gL*TvDC@QX&6m>;^qUgd=M_t zJeY$#(ejcn&rmgf7U;6A-4O*W>Up@9Wr zYzxi6At(XjM{Zy~H5JD~QFX{OzcjTj`i9x3g`MZXO`ccO_?JTC49T$!M}@oF>6m3F=oAe zluVF^GI0te@!)jE{G}MQ$-l!WZNv?im|^0bT+390+ox%Kw_gZ_J~mD$3g<`JhVT@I zIaS**Avz$R1z`_X9MU@%h)@0FG(8}_Y`;N2ECr~_j(t%cfs|XyUxfXQb}g8-8sCz! zEXetR`4L_GoGQBE3+HWvn}KEJt@dLz`MKsvLnZIXmQv!^1R9EU1y@#y(*`v(UQd_a zcE4qlzq7xu-NftlC*3_fOh&z5P8ww0<^g`S4J`kbZfZd$Wm-!X*IGIm7&N)EwP>** zqTETqW#z^8AC&}6QI?-Ke&B^0?a|>(kjv%%#!Vg9(6Cq*#`?vHLV!vp7OBQWV8uHGR zg#%;#ul4NyIXBl8eT?W!R)Oq*7Cr>tZ;n(vp?hPhaQI#rA-4zsyM;H@1wA;i%6HtB zRsc4rV$VO`d>ZD%e^u*>)uic0YnaiSa*T9o7MO!7G>3kVlP&QW9mbWEj{;OKmYrAa z8YRu){m~6(Cq1kL(2}aC63>(SqEnEV4Xk-CFhNosJ-DK4`6-eB< z2WcfrevoIzEQwEBpWl=H_b`&!U3IHmX0}&cDCj;$)-W3NMoMsBx-c6{3yb$+ zIxh;5vZ>Tn59r{-vAJu!6_-fWBJ#B<$i#1!nsG}(^BXQlAJo+(+M~Dol#7Str)kvc zv#akHN;+Wt;7N7aRm`$wcxpi=3@iHYu*+P-$H(Cg6=eOiS7_=Im-MW~zc|s#)FNEp zgX6cOhOI2yeR2!8%aB6#Ooz9Ts}vyMnZ$j3SVh{mi1Dh1N*L5hs=Boo`S1$T_>9ea zHx4Ak4+AN{Nz#Rh=d4H~8JwITSp$F0;AwwNoY&2ORer-fcyXs1_q$8BaAi zGN7h=gk|RSzaGD?oKj)AX4u2|D~H~nX|bexg4<9PPt`%^Ki1YCEBf}=@|L+ynYlR* zKFAFG%gGl;ox@!A4SL z!Em`N;0-sn7Wp35MV}&a4x`SB`Qw?xnG?EbL3mZ~*ez34*QSN9a6Q};smz5rnkoyt z_380Dp197dKWkrRYuu}+Q{FQb&Sf_5(ClOyX#25a$E?;WEB?TJ0NZXtWzui0egII& ze%JdkJG=@`@cCmYt;4GVUAm86x1ZI!tDV;2>vFK?SpV{y+mF?>zV(Ush64G-m1xbm zmt?os{Svpb2^0WIClQWrtgGG<_bFlC%}thbF3+ExEC#va?*2OR)Gev*S*Bs49YDL| z@wC5c)%cpi5h6D@l8D_er;tFhVIV5C{1xkz=Mo)nOgVv7fAInSGt=yqz9MZjrf$&65S4oL` z5qY(azqRp0vp3h*A0j5TjDLO(Z0HPs`;9bYVXC6qiw5K%pJQr*Lff`J4)X`37Cm`i zt>@yMVKg39U2>H%o+>*60s;_sp3yhn7$Jyj8Yc11_|VPqEUL1sY;Aqc-yh;$J%Fu7t_^ z!@*02XzjajrU%O>H0NCCb8Qk989z2{s>qF`tG;M%;rQ*Vib(kYyC(!B8+{%QQZ2%J z6P+AGFwJGw)&tcBB2I* zYWPOl6)#_O%VsK(b30XKqzZ>20k1r-)qQR(4cUgS~kjwKeSUW!B@ih(B9vjpxplttCywKfvzm`yE$w z9im8EcWJo`HbWc8YVwf9*U`(SywUPMlulw2{`7uA0+qy~!*48U<%eqL3|ktuxj1`r z+{)6go2y5!Sg`^fv$fL~ei5;|2}vSw=%PwJSAO%>I;3HC-@cwbtc95hzCC_P0~G6~ zadnpA=JL9Br?dWZ);bjLr$iLnYmeN#ofv!d0_2&e9l%+-?>iPVDud8u1U}D#O^4{Z z)GspfZ3FZaizMqXL)mI_#;K{hcuh7?`-2%y)y)dM#s$8t8(TH@M8-?vw^H4^yN>_v z`FZl;kHi14VE_5!r}s$*MhDNisH(M|Tq?G8Jb~So9@@lb(TC0VH4d6c!?NdN+M74` zSsqdh3H9YYt?e9L!!F`dM~!&Iyy;y(*GjIVXU|u!UQJZkgtd|ZvsyKRBgzJc(f_EGI@l3%D^i)rZB+M!qK4 zYbN8i2u<}Q)c=nufc|;+dbu1$9S4{_a8nbn!J_2`4w40}T9JgQ-XHJHG%(hfbD=%X zs|p+F3#92a6K53_Sctpi%G z-S74}`ylCP8NUG0FIDtzi}=8K!D0)^YLSG{AAhMV+ST!+tJ|*ivO0qb3*<<{SN%B+ z953Rm)n4PEol#Ly{@tXF{n5u*U2R%G^3r!VHnNYz@35|Cef1k@m$jb%^;a99NFtAs z{+zJI=HY3qpPz!qSZRvy4KEfyp&t$T(O>5=;^YS@SoyUw3KeckqIN^bsn5h3Fnw(y|k)Vo;^Fq4Q_XpWNvrO{S zH!WN6ld|)SF#iYFs&Vd}{3;jh*v#)@!IzBN@{n64fQ5B`a%9jztwTmjIt?=GB#Qgx zr9;eb1-@Qm{tPNkrLy`Jk_zD*ef$1l79#6Vn1&3nnn6F>>GVv4TLBaIw3_IfSv8Gi z12j~-AnaQtbW$HRsxt_gQuU%GoYnU^Icgjyqa)J6mxR`!5ZL(n&B&0%54lzjrpAjE zW?h_f!R*>49?SRAE^p@Eh}QpGhU2+z=+L2ei7|Utf~d@W-0Ua+L5riMG4608EofSo z*J)`j6LH)OJ$#DH+802^=!g~+oq!><&6cJ=Z2z%pI2*J*$LF(Zt7orXdpczINqZB?`QLOzG@!wQrxSqhWqaThvK|TnO4G3RiNg%^hzzyNrJL@ZnG>$2Iko z^c-;i*W1*u@fw~!eR{vdcf6Aa47oI5_lHlWg8qfG&8(1?*7JAodNyWFK4YFsb;aj> zCPBt6<2|rFB?^>Rl&)lUj**m|h{+N{{UbF^L$%7jo}sP*2*_gMNWLFbrU1RYdk@ii z{_I(M@F&uGSvcwaK3Jt?YLh1KVqAK2#|^M7Zr`zE->8G_O+yz=fZGx$+0!>ZnO>Ts zLG#sw4-pZZbMpH2(U&Af6;ZW8%<8PmpfCCDcd*I4w8;J#UkuhpL=#SY9%W^qhSa@- zTs?-q(&lEo#al}z$W<1A6+9qv?zXsX479eVCl#jGPSJrD_hhYg#8CEoi4C!>&rzGK z5yO&QQBaO-rKA~XH-phL!Sc)QPVl7Z83tMrmpRt@%~cI8Xv##pUXG5A08T9u#Jz#T z7mlCjvg5`K&D;ZOxxx;yvp2aB6G;erR(A0G5^Nszr+=xBuGZlTXQ=9351#^YXdQoc zD;664wb%0)eqQI-#wj;H|Kj6!B zp-r1Mvi^N?yU-y6nD}XPU`oL2Z9%Ck0{uw-+sR0)x18(XSGRxv{%3HNAxST{0J}`2 z>ppy$p5ZJip@|W0Wu9J{hOMhdCQ3l1m-EQA((X>}bnX{@Q$1eHqT3F0liQHio%Q=v zKW6#R{N%bB^`C*GWj}+&C`--H6qWK)sEh||MY3%S7xKebbX8T4uUWemUc4c=}3T zb^I8$m)5U&4__f$7vCQ6rB0b@g-!@S{Heza9>uvMa3^?Zh8?^0pd9yl7h=;ls-@)ci3sd|4fR^CITEGhGTAPt?UAc0l{gKM` zD_5>G(~thPg?GDwxT><5sKm8$UGXqJUvJ|tpClK=FO1(a+>C~LNm|8h|95;xstec{8*!@W}3!-`g zWy4`FulzpC5B{qKP}SZ_H+}_f?Zi|AwcijAS(@4fa3=2|nXqk}G6JAy-Wpr5LZy?_ z|0xO^Rt$*x|Ga_okI3(=SJtr!vdH(jx#v?|_gr|n`d234fL)b*2Nlff241yn0WM$< ziPC#My2}^>K&2MMefI-XY)`wqpCA@OC3C|T_u?IO^u>C-tsE!zHr=K0*`3?4aLSZs z^DfWd!P!xAi{0&Rq|6eEUy(Fw`WPEKZ`&5_ZQZO%YEf^Gf=Y$oA}pp(&J9E5id1)& z_9P7@%{j7`boa`0=g*5k;+US}C{pU4mDbNeOOh}C)*oF{i^MG(Hb}N35VN>mA6O)S zpLV~5Z2dE6?CqO}>ZnU`{rH)q?_)lB$oz#1nfEfVlEf$5QYiHcUhG&?^mkZH4)=w?`ZzZd%(EeINqJ)qoxe2j@L~YsLsd z(u>OrezQGLX+mris?O>Y#9J~`M=H&SX{a;G3ugs#t-t$IXYPg+h`dVg-h6~zjIH^* z4!8b|r-n+}4Qs|_$9**C?)3BXyAqu`yX2RrF>L*8EQfr32dmBWrei(Ju^neW>)iNv z&HFz&G~TJYh2snjehT8oN96Q9IU0lL`6&D5l>HJIGw0HrmS`5{>gG=O=%k@qJJgo& zCj;taTQL+o5i*E&kYm`TeoZ)0W8$Pq%T})Jr+w6|%+<^FOMZfxkI@mlPg>kOG>6YO zX3UrYzLSb(mR%d@bZ6$A_T75)h=rrJJ&+o4?V3_`Z4b}!csbnVl8GW%w%c(-egV6X zpixt?Hm$rqHYKG4m4mnWq+Ii8yV>Dsy0xb=9?+$cIwGv2J;TDn3R~SD1aLe570}r1 z8yl9xo(-*i$`-0}YXQjlcMl&$R)`Z76F@t?A?egMTu8@X8A z^lX1l8NoaP<3pxv#uxo-O6q?0uit+9PY?Fb-vr`mdO2{-pIa)wn9D)P+oAQQM9yMS zMlZ0`j=l)l=l<^cxUAZu1S!~qu&f`d?erpY?K7IzU( z&E`MLx=C08(u78|+JWHOc&x>Yka;UeAOgKn2HHRCrTDwJB^)HbPEDY#x{JDcPwu9= z`uZEr4I9*ry=U1t^J>9u*hd~{bC4I%^k}lrl$K8|Dn%#RN*jv*A<{qK+8x{m z2BaT{mMbux=HOxn#NBr(0H5{<%vQucP7rxytlBk`#B`Kz4D6!71MH{kf_h=Vf{4Jc zi&Xt>#-%bY+7=yO?;=PoP%bPHw@Z1u>=X+5VhR3#D`)?k(NmZ z_U|_UEXXhq9&h)x50A(dguvX+mQyG`5W4?dv$%&3RVnqus|>2OI2IJu9lSn%Pi#x* zo|e+}Xcci88rx4@2OTiPG^NGC06*iLE{yaXMU_WgG$CXjHCa>C<9y_YzzQ3=akcmd z${4tSXgpEjeUFXZ+)qXc37cmrRpe;^SBp!Yu2c1#ilb#yqAO49Hzt-*3i{_qYs|ac zm1Cz8_0W2OnVCN{YlL+Ryvq+jV6}{woi!q|Qrl2Y(7=E2*}x@e-VvcMp4;r;^$gw@GP(ycG=`8Vr(*|plA7t65lQlM2# zw>#EH2OdA(jre3==j}Fr+^A8bF3b%+QrJB#(B$HKc+;q6`Q6mRns(AK`H4)mtyuUaL%E{)6m$hn$I=Vtm1E)Oz?;TMg+(LBKns27R2_gK_obnefO zIXO!yO)yITf!42ZT$m3~MbwY{Zdplfw_qKNhx0y~udFH{%lQ!&EqN(2O-cMv6pU1R z1U9+28$ROLZgCD+u9`Aa_OgH|2KCIw*Y5zX5Ns8W>2ha5_tc`Iy#-r8Q%26#?ThIL z=B+ZXl41U6Fo1}xe`?=;I|EW@)999W;#wivBrfsk%&uSQF)5WVU5$N&DRWiw+M$rk zfx!a!qMnik!}o_n$P9(k1iV-Q2r(49$|H57x*uppPKu+Av2m#L50)+lx}Q`0^qn8eZ5G>=g*)2IJCl;fj-Ej{<0Mqf8hAgQnQiPuO6@?(<1fT(#am3 z?m>DSko?%-tkkWWJBfEW@HgCv;)%LYs)aoSx%W))nv|F`F~ZcNMw)>^l@DuGQN(9A z)WtNoqP8h`aji)cv0;$Xfs=gpCMRNUS^K}X7UL$$=CnJr%9?zR(cyYiaqlKr=GQu1 zOSS*{{s$pHK|}~T1E3dde^w3QDJw8;=_$}3+f|j}lkbA|C&Nsr9Dif;@rjbCu_<|# zyi9|Pm2N~uwT^#s2(}G_XtQtS$3QagQmN8hh^i>RXfyZ=^R@%97tv79eEPHHw6t!NT%JV@z^FHz>&DUMUoKB8>+j^)?AjUr57o_+7&03J-gaTnsP*T#LiDP3 z?VP&8eu^y;0jh)7`DCjrOhnQ^PB8){ii{j#d_>C~-h6g~E3 zxm(HzSP7vMip~-!JoH>tR=YGkzWUW=AyGMUt>dK-f>}{S{>%g6nj08}{we*Hqyp-) zJvrUFc3p~x1%0s5=$98}HHvrIcXJMs39u&x#Hw3PbQ{vfDUED{CGEH;`3y`;i`K^a7Mo zbt=s!0JPLQRu00+JANS5f#vE1Fu`Oj!&yYTa8<*ihL0SiMt&o?VdJJvUe0zis3195 zGWT?K^~+Wa7Yt&+$kWHC{9Ys|s2H2e&LL8kSl@sqe6!!})@jA+)gCaI#tAWY#j1!g z`sf}lsI2`^ZH{uZaCZ*A4FM?EWfK3Y^_HSwhHtQ#^YW@IOdBCVP+CUnfIu6&9u$?7 zU}uPX0#16jBRUJ|(Zf+uIy3gqS=aKJL zWXtGRx#AP>!IPs)!(j9|DA{TQR|`v-u5_k`j=H*yRGG_eBGueOoWBx%U1RR0 zPGrX*3*`MvK?d@$-h4V zGpJqz0y%M$?2D4V*{HMAF|ehUiErWT=Ve6p{XBH|@STj;(;Q{x<+(xpJPw^;V745h z>9g6vJF_?SD#cylj8(F~N8=+UeHtpH3p_!Qn3$|JczMWta{E`vpvCK8> zBo3|Mzn4oW@E1P=)PQ0W{H!(W8YV+Z;xOdKaJl63V>(+03*!oWBR=u>9h&|!YgYSH zr|GQf)1`?gh|y~QMBnZ6>O&@Vg#cK6XwTn#a)%xkh@AVx8u#(PO3PT4rDM z@*f9nGv{A_Z6j{}5XIz(X$8ewaU0qWBQGheJgJ1nOn~Cws|*5{U{nCjqpr|aEyWa< zvoW~d9M+;C=kS=2MYj zfYt$%ryk6gFC)t12kM#fp$KVuv9QbNnSu3F8~{eZq1$kTQ^vJdjGqVnLBD2bd%%T4 zc@j{)MQR_t00b9;lz%Gq{Rf}(9CIt?gLn)4kFTlEk@iU?xcSpxp0Bb+~ew1?W`&n)2mjqu^ei>ckf;k$J!t$r*DiuiK~ZJDF;q4 zeui*myJ(hcTFxVyo^U3HvtmZSIg`v;@cv0hM<;DAH!rUN5nI#I>oZZ(r*13n z7;MGY!C}v)k55#vGo`2e!>i%>O#LjZMiBpYL?AKy+^^l@Mj(9MOKFKQ^0Wm>ooxV0 zVi|osNM$tL%$3sb19kTWd;ui45`|;&eX=`*>oQf`wr97D7s?bl0XuD^+juij3TAf3 zOJHN9CT&Z}XMSgfC`Q{BWwd>tWIcpzJpkQ6&tAP|^WwcV=eB}Hzw+mwQ_fD@vvl?9 z>JpjoggaeQkDz8EO`G}i z?4nqL=1WY*2ciuyCtfZyzLlqdJqU$QF6iX2{YAmAEPS4OX^yA)F1NhffC5yl&1F{< zIfVy!n9AyCG<5bI4~xmmgv)#@&aW%gZ(h>P?Uc^ z$oiY81tk^>{2`u&x|vXvxLCFKJ-LLG5laavh@3_hy(DV;j*!|-0Luk=;a_!r1FZ3Z zw=0WG2m!lS#ux*M(8~6SYNnX)$xG3!nDK8RF`}q*$NZgC3zM*k_5Cf-NKyIQB}32! z8SuNj=%&g1Pv~|ecxWBUxXh7~mt7#beBEZISEiF^~_@_cEZ?Ls}NW*aHaVIVp zw?-&7s!2%Doq%0320^61?^yemuqE;YF;IUJnFn(_!2H%#aI2K*?b@{y1RMj!z}rnr z-rs8~Sp7iLPynSqY;$N!&2hFD(*nCFYCix0Jfe!&y|UpXD2ZsZ$v#G(Cs6k2>gw8L zFSr>a7mRZ1i21Ivzxea8k5AC!HO-qh2aX&#{Nz8SPJP1s|6hugvX-8AG9Xk%QRa|q zkGMU<*xB{h$xJ9f+nF4M2vSA<^83o5OZYwkVJv>n555!UUU%6Ju0=Lxs_WPZKR7be zhlSJh0ztNRfAADnz#ZBZXYYCT4x%5;O((&*xaqQ)uSKGZ51{liK>XM2kq4(35cSm2 zkxI3&m4O+=x;rDTS-2gWl%SrG$tu}{w>==Bue>a3b9nz2j$el9==9(qipuY6!jKrF z+s)utXPtKFUy%6^VM^t8F=-}GV5$5laGvpXWxo+!+Y$3T>N0-U?!G9-xtD(J*QNrp z)QziKviik^9l$*Cv_1~!$U^|lYQDPm{^gAYheNq0mz;kvsMchB*=Po!_cDgj)b?Fv zR7ZjP+v&(w+VMqEwrzRbosLWmTV-X{0y6|LkhE25au*E09jjS5>_)9=MSmpPr;RLo z$er4OrP}0`UH(BhEk-HnVyHTa4WAM})8;l2wFUDaN2rFNB>zs=IcL!Oq|6=0)K>?q zM$Yurb?31U9KP_Vprl|K_o2E1x40cR5C^tC3#R1;%Tu2}?})}k@V!^Eqe~*bz@7pm z-`~^_!U&JFmRA|ZoG$O1F-eH$0oM%a1Y}VeRt}U(_b{5~80`V1Q?&=BdIa7sA`WbS zXyONWl@2B<6f-*Y9fjTFX4D_FYXwE?E_j4u!#36($g6`wi`Y!kK~r22N&I9Zo85m zBUg8lMP8JacvCcitv9wFvy{VW4ZDDM}3ymN&a68;}`2 zoO7mJpTyRV0%xACVJg(sGAck3{(${jq)wm$qU{mWvom~0aQx1+h%#Fu8|#alUa#gy zN@k-R%0_S`?JPO{+PnH#qM>X%oEUbpeR2VKZyQ`Y?&0fmW?rym>RAPg_3IDugZ&w| ziU=MM3eE|tttq~~QLVJQx21kSt<;nfbk0SSpM{#FSeYHZd@^uH?d?w}9N5JrTj}KV zNOKC{JLr1DP`Z(E#SPotj84KwZ^U2&Wu?59Nz|XoH#V*qqoWi1>Q%+*Ty+eJm}Yas z78_B-^5&Cbf_WJQwJ}*ojvV3I45Wzfr`|yADd;OhsN&!)%c&<84G}ek_;kZyZ79i^ zx~#IRPjvI&uc}r2s|Bd12B0UK6l9_?+|THUAe3~wJWP)l%BYjYya0xIKiZ$Wz`qj2 z`4tgGlx0}|5hL0MR7$CWe#NNj3~p z{ff3)Z^hE3+sS~kE63xnL+p@3|@q$Ziui897SqrS@t~4(WL0 zCQL5Uf$0bokGr}y7ko0RSrg^%)91_*F(({Egms>hy5Hxu?`{m+f55)5`$s>><7Y{G z*Ow6Q?>7o0mdOe$Hi1wp-kvOjYOqWc#vw`L9V@huhK~2pTOi;fO}g~SCXRwA!@u39 zbUkBoan@~zExg#)(I9{~(~v;TkW zy?I>D>HGhE4P$)9I*VO2i#22^TZ_!tmu#gd$qrnsT^F+(XHry#}={0PaeCEQ~&->WUZ+p zF*AA0`~Y7jc+5t@1H1PvpnR>Tk46~~Z~}yPm2ip#x&_%VPT)_%N!CyxCj2}=b-NB7 ze!yzQx1|GefZ$Lpg-$FT2y)wp z^D9@b9Hy2?BsX>A)t6PfuO@9D-0m9XpSujgU9+aaJ)yy#Cn}M~a|!(BRjQw9ZoX&< z011~mMmj%8`rsA)Cuu%4bYh8~wCDk)c++gmQJIO1oI-)EkhqDpp#(ULW)KJo_k1&^ zQ`segUP|e1&;I=nz>9{{g~F@m6sC6rAxkt*E^LaE1+5vVOz2R2*}bSopFZaiQ@l

    ~*N2)!V16nFO+v|fzBKa<%!&tWL)-AyXo2q-W2m-M`Tmo6c`fftY$ z*fk+oz6L6ie2rh>PT{b@1HS^J+%u(qV)xA>my#=-{`+r2lV2+Q$&_h}Q9%a%ccqmm z+d(a18CNY%H6r9cH%OUvuww$9nC0%JtKK(B7doSgVw$b(O8oFH5c}?Wef|9XgZO^n z$-j$W@QoYNWg}E-YG3E@PIIcO77r;|;?arf4~WL$v?JwvG9QskXS1RUI_kdh3p-)Iq81#~w{R0R zc4Y>H_UstlnecSi6XQcOYnc$3^jJ$v>vVGR_koK_6E8;{;)4i<%SxvXmxX3=(CL!J zVIkrFYv@Wf~kxG8d^IkPKb(Xy0so#Bfj!Q>w`w2 z2g@^KY~X+swWe)bP$-&vB0$NgZXbTdlbd3lwlkQb7f9E zdQbu#BtMHmI+2o~lrcKG;Hi(KW@oIjtJd#?fUPWQfd z6$&gdxE4gHlUdFWFvP+wmGFyv>;ar9-l#=hA-!oWC1}}0*hH#1{#L0Ad+(u$etlCl{@FKO-;2w>I z6}Do<2D98K#$PyB&3h1wEyNT}y`ukg<>R1Dw0n`y7o1ysGDB)}a>QUfW=<^KJ8swS zG8BZmj(Ti8kmMt@|BCJ{=<#f4=fN!EIJr;#`Xyl9AVhZ>zx%9Sqm4xFuC87WSrGF( zwWUFaj0Qd)*n$(QAKJcd#F`W=4~q?w&nc~^i71r2Z=^vGj`W`u7VihkI>N2BDKMLF3@Q| z4x|}#J-RRwy}dta8sv1_E$*{%!1$1~(>un#C@sCOJ;?16(+R;PkjiiF?F+9$U?Gae zKX?ywvYIwcLC#Af#TdKwghFTT+?i@u7e|lMhKxMLF9CvTGepbCA`BZQHOV1X27g;( z`FWjmk=d;$bO?gVJwbcL3ji5G?_R?M+jsN1Iw8EYdI0^em7SXuxE>Ssd#KUnu;(MW zgTjcAri0j3>2<#T8pY$WAQ>rhNS_7gg{K@6S&~~>nf-VCi`>rtk%n)u=u0I0dp3cu z^rVeVAd_Mlf!#Z>OxUSSO$9Dj;T=|8=}>cccN;$#x&0li_Ti7xV$tux!3c!H=n&P2 z!bO@SKjNJNa#XIul7me`odzZ6q-Dhb-^pkHiQ9 z*kG-E&2rmAMG(7%DQ0qjSCN(~>c8{1>bUm8-zwg^s_HO$WaA+?&qAc-W_iz&-d$u3 zueIGbY1~!n9(b`WW)I?|=ZUi>h?Ed^V#h48wzpr+u{e{QJf3JMWgcfps)LNN-C3R~ zAO{y!)ArgSrDOct()yh3r4~`l0vz7R*8JY-r9x{IGVSHJ0OV2-f%#%Zaw$rxWsL^! zi5DUz3Z#otlT1&B6Q2A6M87eUSbr$F6e&|3Q%X=rfbWCbfO8EvsQWFk;cOCb(4tO( z6u6}=dx#YoTM)>^sw?S3xTNz;Ns``@%!vckD7&A=p8;I|qATTp;_98K+ALZ(Yy&%I zI!mEUHx5Yv%0Gw@_0dK+kBr-Kr-z3Tl&JN&ot;gO8CF*wWp;~Zcz@X4?|!(mGBmUs zGzY-tx_oFo;3S9>8UBC%Ef5SMyoq7sf08hR0ukHU?e)Boj)wzDFQ{X5>LaD!0ox%P zF;u6fpTT*I;3br*KtrCY>V+rUG`XzFE~bIFGowlVLkV1(Ykb}k1+irya?`zguc9&@ zGWVa9%P)mJyCxk+v9*L{tvt%l#)CD>Vdp$slB0rS%D@A(&}-q6BSIc8CEJ*ZOVT+| z*bvZvH9@prM%CBbG;MzX#}VLvQKzXV`_VAc5C}1I;^S^GfqFnIG<&qZzV!ND zzkc2Gt=WXq#ol-mf~MhpXdvin7Sn_z+S@EQNOVRxa-=lB2{*V-{bo-!)}EGt7ggjT zf#q-tR(mghhPG>GUThAIl83j~Sbcf`K6xZ-6A+#u@)y!{ z=EQnOp=16`#zLqAhj$Z=+ouuDn7_8x8Jbkte-g`6x_GwUeBSKYG+=2ucOG{I&X1@TY(l+=RSeMk5OZq1_V8O_UsL2dlTxz$suR9%(%Xa z8z6f2olK4%;GRy9*o579^VY3P#A|hi3ylu6d5BdPmlZ#4JaxjedH{5_r4}@E)KC05 zR|dQ%7q$qGsL@7n<4a6vyef6NV{{2Nhq$ z^M#DH#T;ah28ano6d8A5sUH`h7`tRRa|(vB;us4mKOowV7ds`^<*|JgS! zhYcM3+rtw(|H|`@J@Kfdd4;_d|X zOvm7$Eu-vK21id+|0!NCuVapF2aWPmNsGN6J%~A3aByLq`Thpi9u0U|`C5@X6!|uc z+4RIBQlEo_;#;l^eSM6!iOpNK_<7IDbE&70$86X6A@!9%z5CzJx24aY-+&;~8#5IW zV%P}|&UNZ@U9or}AyYIcdFclZFoOVpIA&z@X8Z?~sxEB@XEBNIhL-fx0XziRzE&s_ z`T-pH^_#OBwf4lhl> zMZ7JvR)Bh+wBA|L6OYcS4jl}W%+B?o`j-NHvCzqFt<)W92Pj~-7!DkmU|Aha#P8>J z_Z9h-{^f77sNf$Wi4UQQC~HpK=|*DQ8|9^tkgZ_4;|cZ+Nzu9dC$4U81!N}~NIP}J z$b}Rt4(ucknlaMR);5@CeW#Iq?ml6>$L39&9(Yg2MLc}IrQU9m05rLrGf#3d5knfH z?!P=|%l7RKiRBdE$2NLg=h8+NFkiL$)pB^Xd+34;DfDrN+M3SU9i5ox-X+YKkV3|J z_$ImZQ`!8)Ba=?H$FLiY$9&@zVi;o74V`d5?>HzE26M9m&@qw$u-P7Oij9neXA^Jt zCICi&bdp(}fpr+nS|&1f1a`jYqo=uCilfpjtz}5L2XRApW{ns-_PB%k%bW)$TU{D& zf(C}+nK`nLa-Y7MHKl#6kG5@s%aPl~6*mH__Hc?P12-6Dbk2YWhw*Ed?O9zJk?CNR z5$%XBrCVX>@g}WX+q-O@+a47mUWFOx8qGbd&nQ-Ks~kZ1deHjz&4 z+n*~?Q~$9=ivyy#*@~vfrj6UFsjK%1UCv&y%~VX;PtCg^IcT}%b7;zzi{BIfPw5N#c1ZE2LqZN zpm)N^cus3#p)~4f>|{q&C1sX6MAxI7J}3{mEZNQpW|zNF(^`IJ7n-Z%pqd!bvKAU< z(3wLz93&wHU(@AmoOPJU(xLS{4xb;%%hO9AjAAJ_KCj?F+wKP)=YL>LEKcY*cB^RCR4jgV?-}| z_&=nPz>9ul?Dj5A4dm2-@2L1KY6E_)3Ddb@Tb-K)Y6`Uzq=Q|4`YFMuh~&=CyZ(t- z5cEPWS$$htr1eNY68D#|bAjh#D8Ai7z8&5xRXzq-2ixXHD0~?zImmitr)AV^Prk0nzMY={m8Y)r{ z?6MqrEH?6XvD^KX9m}o{YCLCs_-}|#9un0z7;#3%Uyb{_>`AsM#6JSqG)|&=X7T)N z*GB=DP-0>nLz;_X=M<{Bn^YPt=Ody^+FlCh*xFEt+-I?%AT@uL`{kNFJMZuyfaiZ6 z4L0uBAyc2ZUl$A1awjlt1!f;(G5WInYFb(*a!;`kX%jCk1rD->16Gg~{k){gueH<3 z}Z*shOoAizpo0_ku1a%JTMC2ShW+TLy&w#N_MtF?C+OMa!%9f4 z-B@Q_a8CEB-S9}2jIv7CtA0HeoMn<(T}9X>nlgLartBHHxX14>j>MEfKA8l}Yn0Jb zAF)c!>8JmySJ`Uq4;*Ie9j-$O|Hq9UeWj!-1_0GRJs;0^E5dTM?Q1XXZ|l`df%~hk zJ?TUQkp!ZNVW8fYE=_&m5~cn)R%aI-8e2l#=Vbjxtg+eOKZF`wNQmOV*I1s>?p6z- z(20>~-+Hz+8^JkI1#jn4keb*&@@^{-4ww9s{GYH?Up-*uhqg$WPoD9#}@h^SrU&DQ=LlkAsb|Wr}gvBzzkvYFAx^ zqc>NyHB+3JO6P(dVdyApaI?93ZNd_fB|1d6%vSGg3VjB1zmn~nuzB0O3;*n6ep1^@ zfO{3fG~jeR2LZ!g$N%a+5_>y9mV@BwX)dF`N9j{MwSD>(JarcwIWR$nXd>RddpD%w z7H>h5(3w3}DdHa84>wRKwq79Th2}7kS>>=b^=~^|QzO2V@IrPa{ineDa# zviO(3xobl1=!U*&H~)A%5+?SFa)-Apd|B8W`)WMZb#K-b?YGQyKNR*f1VL&T(z{8~ zJ9X}ySfG_V6eC%FO}oT~tPRI`G^P>3(vjGbQ>Hl3iTpA&?Ec(F%LjEo^=kLBMg#J9 z`o8A^$cC@C7u(G&HBkj1XmA9Md0d`ho`GEHuWSN*c8`q|VK^NJ(crYCY-$xb*(~_? zh}*^7(o{YsgnYQP_+_>mzKo92SuSmS3$O3`6A=9n7TeV9TrsV&cKGC0dLAcX zj)jKR-uXk~l`ERjm`tvp_)S(CASub|NZ%_3DSPXev{c(F=3Yl9Yp+IRM8p)W z#=aSVi3Wj>UO8cx#Cl%N`|4GtxNfQ?O`9}%%+`)O_-NC!H?V0R zl=IRiPU(QA3r6!qfxYdk#Ws=jFDF6&zF=)U{_vggXQpe zb+7J?(ytbz#(R1a*Bzh@%E-8LbMo8}?9p1(cqW8DkU}Q4c{8$dL-rQmX&n*j&~+-o zFg}8TTZ}Tk8uzWlR)ihTwpr~EW^t>xDXkK{XWdHmaWEW1 zRYcu?Gh+-mk-k^wHLu%9l1Y%GNv=nOF-USdO2v`eKyPJo1k>c+c@%kP;uJep>(?uWwTcGZr5$C?Z_v-KG zW~5u!f;VRshao1@T6*cFZA}=iJO$WAQ~L15PL!Xn-`y}~kU`j4_9tID6{J=&hLFM& zYtG!e4(qfmYhKdCiiRq>Iy$LHvgAH`R7kWP_hfm4s~!Ufbm?+Ab&*0b?Gy*vp4M)e zmLPrmob`|{W=;7cd(MS6CNz|xIM+>2FFxd2KhLU@-(Ia+&#n)Ct(fA3OuQU3ix+e^ zlyNzedij^Xsh(cxxC6{DpZv@D&6!T|bj^fxnX(aCzb7Z3kHI;NCP<(5YhX1g;%+Ly zB3PVJ*7gU5!cYhQ>xObT4M$y>Aj)<#lm;%1D4IG@v0)F48-uy^ z>(v{bSSKqD^wzA`esW4Vwxg{6jv^B%OmSri{TQ+`%KCmO+!C`sAte6l{&nt_;zIa_ zyQuMl>`jb8be}=bGh9OJb=T9&@yX)7cWnD(7r9qagnQ!E81=i?knstpNNkEiqKl}@ z1O7Km>ZY!qPSn5;>nX!1b4-FppBwo{W7e6H21=3!#bMuVr}%)Ee=53)y&)DDLKJ48 zv^QCz=?pC~9E$*F(3GfBZ98z8uWp7Bv!c?14KB2gWSVAHv;jW8w@!xRapOz&2bD+R z#E6v?x%d9@N3FGOb#M7x?RII?sK(-r;cAZG{9@W4bivyIAd2V1wmnI) zs=#?C%TNzmWCqOzQDcC)iVSf48@P1q%(ny(5_oR8YF&A)BAe;BE;ry~ zHKA1OviY(WonTNbXS%G-*-OMLK^8o-@Hcnwulk90R8iI!n?Xr%GOMFFv5Te#8}LQh zyPX6lDr|A--t5N|(lS`)SvTFjF>dz554$@CE}9_Ih^e5eRsL3=^0QlMV#g?p#2=!Z za_(SyvIdhBlq2wMLxw{fu+>x%o0;RXfC0JejK(Mcp5a^=~pF+l;UP=-3CqP`mW$lg{_ui7nn%`ur_r z%v&)oA@`!av~My0<`^UR@Ov(X zCpcm6$(OX|4~l`!lh#`?r7j(Syi2xE5s|6E+{+={>Ml&f5CM7;aYC9taEy0zIz)<% zqZ&zQJiOv%oU}-!2UfqeUb=j_so8}e}UDBwVcR4roN5C_I_GDk_BZrD>T9$|6c{Vx&%dD!76Oj7*0q#ucLSnlhYtC|s zRv>Ua#t*spU)8L9`_>0L1PnC@HFKpMQc+oy9S!kSV`p~m97T_DJ?SnUJ!b!nh}WL4 zFPo39t|~X36d&o)*OZVCcw|4vVUk&H1Q_E?m!%=z+qWM&e#dQ;00;2XL&F}-Wy48w zjKL<6X^E3j5z9{yv=c+YV#WWivfblM!@f3u|51log@cza)Xg1_L>ICZhwu%=y~Z*6 zap-pug>(v~cZvhhD|ec;BG;T7PKx_=sCoR<^|~*%6580iv_Jmv!Gj=ZWpW?M;lS!Z zqYOJs`J1?k`4MyXzK$y>EDS*v&W37&oXd$+=|Z}E4Y|f(v$tj<4GeFp{sC zf_YDaUX5oDAnLM({UC$Kl^NgY=Gau3CM4@el$u*uXm8^oDyCR6n>^23Xa`f!u{0TQ z_X(xU-sBtIsHNeI!kjZ&cV|c6U_XMfsOd8Og{=}1<9&M+r=GkbiUyjR^Q1mKGSdgu zf4sGvVewMNwY++P@=v4`F*>nN;q2cMvgv0`u#7~pGCyMS>^Ry1Vw~@gPVEP4iXkHF zuB)3}){+vO4>+EjA%C(mc!meupF|HM1Wjl2m#2*C-dR`I3^dy)BgFZNNg&#aDXB@n zjB_n7%DhqbW(s~_mffjXtvUDqN^7cUv6IMq-$Tc&6XPi?^l2Y^lB|xoV1bg*1DED( z(4BuNggx?D(2VW<=lKA`;@DqFXKhUJvD2FW+BNoI&jIV#6wA6Ux{g3b9g-BFTHN3@$L*M|LFStci-*g92paCBpZy? z-OF%O%$6a6O)H&+hvl-w1A^{Gve4Yg)vS40s=?UK9Xb$$EUVwJVH!BXQG}x_i=8|T zfTwIDA+lN&>^jd*N+F}uzl`Up{%b946*B2)q{AQ z{68BHxDIXm02p0sw}lzRur`y@!-KE1zVuboZN!N%hGu0>|AQ8f3`oOWw^asCif_G~ zFzz5_DywVTg8^>Nck|rgQCXg0WPAdUM>82DY!^ujg9MMCcIvcH4vz$oH6~e2aPltr z&AIIuaYiBE*u&!#e%Z3XeSRdvsS1`hpL$*3MM)702Tp(qCCNh7wdFlK)s=J(%>eMLℑzs2mF{c__@X|Na}nfF@I9?9mR-Dp!wV zv0nj&P#5jQ^FTa|vB;K@nLPCb?H*(wLn#%d$qgcEnr!y1V#+mC&II(3t9?;q>k<%> z=d1<1)lSX}K09QCUiccSWBUnq?Wp#)>Nk7Z(zg!5I8#PqL(8TPeTNsNi;h$V`*m`S~Qc&*bqhDlgI4elw_rL zf+7SjjCQs%GTJig(Y;a%AyyS@Xnyh^Q2W3Yls4p0(-#2~81>EJa6(_b1nk|kBWI38~c z?js&S;whK*iUXR6A*BK08(1R{9Zb-@@7lm~W%#0e1Y~>Z^u@!lm3rwkSDwQu0j0-( z?D+A?yrfG?(~wEK2uEEXrn-0!oH4;X2ycvxH3oyP6T$By^o7uHP7*qXGxF4NC~Not zH;PxGH6bDsg7;HK&+R7U(RMd*Fi7m=rH00#t_voAw$bv1;W?aXy5oqHful(Q3H%q47S|(rdZR zPZem1;Ia@wFdKR+{3w;Re{M~J!uy$?n%=%}!OX)R)h4zInmqs_ssV5rUhUz*_@p_cD!Jzf!N;SHbkXJ zqN)3@G=Jo*CHD2kj_=Vinn0LDRH*hNrvrp1m?cLf8x7+qsWSEt7^AErHue>nhXNc) z51qdK`fWXvH2ok)g~;cQPZ2yH!&@b$Rwads1Q)>3Aef9Tc9@PECoy-+9j4={^7|B5 zKEJnIRgKy&24IfAb;;WM&RSB^mQbl*zng^Uv-MHjxO~dmvn+X-$>C{B@Qjzr-S1Nw zb3u&}Py}5$>J^He6AJ0Pl(6SF&sq~r1C4qWxDQnToVg=!>?u}BtipRfF_8NsHJW8OM5C^CKAjx0#ylK$en!C_p z5uoEbfiAJ@1ja55p<#Y?$cYn^IIbjCgr!oDL71{>vu0NL-&j~VF#%k8-tEOG3lxJc zUr}i$dA(H5r2I`*bYfAJ=gdlv%1QC&Q}LeSQSO~tF5#!tBdO8a+U}=%JV`636B^7C zg-EFl2CU@5q^_xKqswW3AHZN206NPu5G>FrcbVKyv9s2yb?ed@RIN#YTp$R^JV@#d zCrp^(64i@vFpV-$8u&zil|~_jcNai8nuyIzrE)>~yTIhsqG*wmW?ajaP!A`wl?Y7A z8OC6R;BJvkjEecetKRv0vkw6I3By2)$YO9k@%2|o@IY)4Gk&=#MphhsdjP#Y!RqPTo2U1WYSp)x0#$gg!^ce z5lffnR6FG3aL)U{5JKo`p$2tWP$8PM8*;eBVbD%+6l}jZGx9=#UHs}B`IKkzv6|=S@{Nv|~>Kcv*Tn_frAm+p>QB*vxS1ga^2~djI*)I20|3 z#n`C(P?SePJQyIlKner!jmBC!6=-g&I+WZj=mG=NASe(XFstWF{cTAr87M3Tq zgJ~NuA(fRSo_*hXY$MHkOaYMh?xA`77it~BRJ0G0h1&zynh@Yna=Co&PNVpJq(A_w z253)g-mqbspmeF8PeMphbucpqH7P)!L{_IO1r^n+uH)-Cy5XyQo49KJK%@%`e z#1L(zMS1KJ*p1r<1{mh=7_(A8Revsx?Fa@<=31`H3m__oblJRe^=bp^0W@}Yp>E{2 z79x!zi~hGlXO4;gmTKt!Aw=u|gkIV+D;sG@Y#_&z!&U|dr4ZqaeOD;;N?;Dkg8Z~@ zTDZA6;g$i@0r?W0HL`yjKYpW}87b={24Kiv(>8nj%^`s7B~e!@2r*p{lPCZgWl$mC zjPm6LpTNLy7|@U93lgE3-T6D;`5^JJ_+E$-W&>37=mh>ttQ?)TBUC+J?Gh=VkS>*f zWIGz;!!JEqPX_R#sgjr1gUiKf73RI@JZHd*6l6!xr(f75?c{P8$`gTCE~W3 zbcc$F9QgtYcWES$6ERDZUTkKr9IlJx^*1*5L5%xlD`O*s6w*NKqM%^P*T9b+kMVAo zNSxO!Zb6ooGc{UIxvNRyO-UU&!4hLuu&;aV+c_Hz*j{td={pLDR|~C!cwmb!q3G zJ%eTC2H<@bgC$5cP+U!jedR$=910tTFnEo)(1lKQcc%1re{o0vD?&UwE=(}6w9V}* z-`okH3smZvc(5C5``+m!C8BN$#kfYI$wDQjx6X0m9KW73m)(n2y4AY{nViE>q3R!> z6a|y4QOAIovzNu(2U8~>W|A*1=265TD;Uc@jDzEoxI5WQcvI0cN>|vvmRJ)onbO_39fv)Z zZ)3#yQAEixlp=DSA^-$cy8?C&@8#nG1LGgBi{%yLM$kZ$S`&;J&1FQ0ol&eH!xHY! z#7Has*RNhpsx0lgfOvftm>Qwm4KRH<)opa$Qrkzyojln@Rn>fcn?9X8k6>My^`TD@97 zz!cS0wyY`D>XDE!&?zWs|J8{o&VxNE?LLbOXU`rueE7Og5p_EopzT{H<^)Nw&jCrg z9_3Gzgj)nh-HkyU(Kp+96TYJ1=`#1K;!lawKE0 zHFD?L)B>L0`5~cXJ|VW4TCpM^d_ly3c7%Ldv}0VBYW>755v8T?xd6=k;fNXG`CLFG z;E^pYT<3U*)4}3&GEH(MnxpHnXQjz%LJ){iso*RTh#{Ksqz!u+A>W740QDKQj;3Zuu?yaDv^*ivdb1 zCOs1=6d_O(2=R`-ZnQy_YhdJx_@Iu?%xyw%PBo?DGq;2jxvhydtXxA zQPrfqZ{|Cmxc3b+xwj11QN_&27^FOm@BWuCjgdg9FhG!{ z8Gx+ra3T(D&c=0Zbq~_DQb4k+g%aSqQ{Y<(S|$g81tIhnscm3j0{%|F{PMeidsJtD zp_@P?VIXhKkUcrVBpGqcZwqCe@)6Y0f*-$;Laa*6E)MdR99>^^$bkjv5HX!0&qtIg5i z`BUDH+quQemtkrr0S}(8*6ZG_TjjalBT3_4zj;GtW+1F0VkQGrPBRxk(sk9VEUAix;}?kG+dedELCEZ&HBv z%hfNpzRQ#d9WLK7pIV`!!v(N~u#exwH&$#Ft(04l*6S%AX=9-cNvCgc)Gk*o1XhVC zF|UrcIz{Y8I-$~NSmw)K(5yLXodJ-ztClaHllV%RiZ8kaZpg6G1 zKdU_lG#?;0ZF|36*!ZgZZ41m}*WV(NVd$8gF!&*?PE&d+KszC_`evVL)wZo;`KWKa z=H}y&Qn~D-P;bo19|Dsa$zTo%XW2>Pj)j-*b?>RJTQWDlrEg7e@2lJWqK6LWL=d(V z$Wjfiq9vDxPS%O#O_Y|W?{rR3x2Tw}KQ>0npRFq1{0emG%b>GD*2@6#hmu(arec==Q@If%gGd0FtF?&{# zQ4MDJ(M2f|L<3l(8$vxPM;$)?h4SZdJ+tICTDR4nKUyH?nooIZDw-kq-<+pYG}84I zio3TY#v@eJ(4luF=_`W!_hWt?qW+O%wGYw`&T zMET4EE$Dg_gkYf~42gSUI2BJ#L7W;UH2~Owr6AXJ_m=Ra01KBr5*J7}WuJE#X^TQA z#IPy)>h%5f*Fl&IDwXmcUiQ|JqtLiw-Lj!|q*KUxI+z<*1_noIcT|7z=y(!*4@@MCtBE5j_a-bL5j0ows#Qk3V4 z=d|$q7Vi1?cq3(Q8cQKlQk8dmW7>A*t-_*(5VO9&Rpm!McdIl=aeoKf>CsJ(iu30- z%^a#9YUJThX}+3rMnl5pyBS;OymIAEfi);s=KP7mu-J$9?oGF`!ncAJTD)6^9@m6q zgR^VBLz`}qnwKjQ%0IhpG$as9l|E2|=buF9vP1*V&B-uZkjR{Kzt=<`1x%;!Bd159 zs50d<$-ta!_Hhpgj~ZO%K%ZZ=cCGLT8YDq_J5{sI6bUT=mnO`ab7a>x0wqb;cPaND zzxeF!l1yNmpE;1?XRl~Hc%%>$p0Xn!wc5cwQ7Ecy@#D)>J$6u|z&8wH!MvU^A<%JQ z9;O-MRL>7Lgt&4dr2c!+@V_;5l;1_s;*6uqe|`4@4iMvmkcrKvqr;8-6g~RAl5gT- zjf_J!pwnk(G9N6MVuP~ICKvM2qv`sV%imCa$|oeKIW1o8aOYAx#;B_f$dWTY?Ii36 z!IiIHp&qiew5o=QJ|$2v4)EEYni@1;QwY_pS$0K>3pWaCp7YU1jms3O1JgJK3`aci z&FXl@>h0&AN)L^?`jh6KPbiJ(-Hb+ZKAl#uc?suy#uIPCR zk|(T$uFAzypA6O}?sAZZ)5&K;0SIo&d6~9@2*{rJewme(wY9x;Mc*!7miNv`_-@{8 zxiHUcLH4-{RpoW}&9nX7igW0@zj-MkCwoAjv-Bn~-k5?n5mV;@15OnggI+*ZRv$ao zv2k^8=}kgLNdIy;Sv%aye?NBK2Ls$XbWICR2vsfM=9x7oWO$y6#{imu;#2)WPIraP?14JZK~Y>lJ=YMRZR8suP6*7;^u9?3_`6SSQ#`6uH5f zUa*D_vI`CnHZHT2MNJ?CEtezimk&1X=VvG1tVa&fMsI?sK^m}l1$LB^AkU84#(5-P zYk2kKgLT@7oLIiM6DDDCSU_x($tLA>C(Bg(CvdYAkOJaXT!f9 zO*df-RI8qncYv7MFO&uJy3|cu+mr}fQ*#~^T)1A#toR1c1rdc2y38sJMGMA-y)!I| zDRJaLw9h15zk&u%HatA#^LeKSnW04B|9WiJCtkPB6kpXX_M|Im*+fG_3lk3xCWi8} z2h)@x1KY=N7lqPA5}-6)igB)H=*w^lH8pSKJ5!6jD*O6wtNF2J^xE~%j3jKMvF+W( zc!eCN30Y9~rWa#D~ z4QT{XE{8HC*qi~n>r9brkjAjJt=dx@lN&iZfU?u2eCJkv+b~ifp{>PN@d&$y5wW$I z3iMASzs>ubzWlQ{{V%dP$fh-^I>FRF3*U}{7H}5y1CkF-*7)p&`Z7>z{|#H4JX5E* zz1+H*EoUfcSuf3Suv2AXBwGy>XoqD(|1A=Ei25WrsJ>)L~bFYIE2zl}P74$e;`pKEvS=!Uh zOoZPE36qt^wISaEqukOj8fzexw0G+>+fvE|&_5eqRT;481Sv?(UREgbJ{NeX1>q8I zMCdW2h=l**(M1d9qj)1F#eGS1*lf!G+y&ASs0p2EkcPZWP{W#ien3Z17Q9GHzy6ea zz&T}Qyj)1;4*^uS^6L*V_x%!Xj;nleTe61fu9(nPJ*Ih0cU7)M!lKWF00S+F%YVHQ z0b*DV((S*c+QsX^OeGvB`rl+2E?!<#DbEp^YLRTb2uH z{s0|l%W~NwgTvA{yfeV%vl%FQf!||)JS;aJ(S*k;JKGT2%Wlj zPr7~up4P?dyW6P1+$Nbcrw)y~<4TXa-EGe{2E8tMl+V0vsAE zT1SKUj+L>(?JxZsz=5C(zU~S(Hw?i@tpV1Ag$v=~S>Akt? z0Z5a_4J^HaL_}FFV3vwQ$D{-(+gxy2nqRj5&53 zVFlz59a-OL7KAwZM~@xraNC_=cT-iMBWB|{=JRnoQlH)z!8JhtGzZ~R-!&M!NR-uhk6+q}M`qCYn2 zYs?|^tfL8=&*er6AXBrF!BfL8P|xx%hR@oQvas@AN2|laeU;*S%EtlW((cp^wy*`{ zI;VQP&V5yAUhU`LL>u+9zZ-k>P_FX0uwNE@bhQZ!Ck7kdr}3xl3+?z<=epmgV%zG4 zwk5~c`OwrjEw+%st*haSi|u=oo2ESZ*rhzUQK$Ko@6e|lek|km?bhs=vB$VmYxF6) zu35KkorOEUd-^6*(PP~qf*!@B&Y4E2#{oZgxGhHN(!aK2 z(twxINkjHdX!$~0wq)#+m@Wn76^ClmH8%kQ{jz1JtNdLD*5VkYhbx zol(6Xl`(bHvExh+-l}0eMoyn%=b|A`K3YewXGN$euUo(9;jeqMK+wjbucL|Bz4@iM zdxs%jzIyvePu26y?VOQuu~@IIy(~ut%kg{v@OmbVBu5AjEZnhUD<17|SC{+M==O34pbT59Q?eg^Q-zOl>-KSi+RzO#`H-l6w#&u=3 zS*rj0E0sSXGOVKqvE6O=g5dr4N$6^}sLVzZ6$J}-V>6eD7>Ryz-rYDv=Se6w6}v20 z05qiSEyp@x;Y%3wXS;--7`3i0SwXjbh9M4`pZ{bOQcM#fc_ zX!>8h8N0wKiyJkEUiNc9{xccOFzPZ^Z6qJbn#)6OGR2q5PJi}S`iHVJ!x1Vo3C)3jR z9LfeW03V+1czLp;dh|nY)Kg0q#^O~-5nAG6^zP$P*UEQ&hQV&n9oe)~XYGRcH`&PE zJwXp{?vok>7eCiCKY5tyT3Xc|CLO~gx1qP8^soc>Z=kvuLpEo}Og{$6pDarmg+_2{ zJIr6gW5e)_nHc9e%!_U3o>N`*awZKWBaSJ&lEX3f`}HdGsq^;rE?Qh@nzIUoPq|mU zp5?K8vlvCvD={73@qG%AYk-FJ62*ybFz@}bc?Su?z@fUc>;w_#y!~g10g7%P?@o=S zIh_)t2M;z3Njvd)_v-t*SF5k0sp%|^F(@uhmt^cD3xXKq?oO-l|Kr0mf>;`iBsX50 z9lSdvqzmk|CKQV07+x`F2oOh@Q42GvZuDPJquWw*+mZZ37^s~D2H`LowC_+(HhlZA zquxp}`<+;(p&?@MsE2VdF$-(xQtZINJip8&v{_+B_M6a8KpZy^do*5KbAC1sR6f!grqCVdWl#m{;0e+1_8{w>BK9Ticne zv}57@mF-NCx0I#swCWe`MUVMY=3VXmY+KTz4%m(7>yG20+Q`{0S@cNz+e^Irw70Ly z!#&#&A&-hEU($HdPp5Fe^0iryMtYSWoyH!gvtZlFuFbvHVgbFY_o?!)gS@Nt$vP** z%QsN&3?qtJ5~n0K;}R=$=o3s%OWSNw*gEIcP%$O4%Z@w(FjLlB}Dfr(fe!2qx-&&sFJY|O3BG@jU)|1N1JX3=TM{6uNf4k4j-`k&ZJ9pE4Q=fm{ z{oPOJcR*XdK!}}`pUGsCj$w)67x5PVFCISo|M!VsSSFwSe%hb;_0U!3RpCRoKX&Sn zz5R~SO8>m$9kS)tRoV8Nnp7R<-YM@`^AUX)umAL=Ek=AAC%pUZv_1WjbJm1uT+M39 zza-r5p%VUReW2~oH?F6KefkD#%szEHpZ@lvzwOld-zR>w9RL3Df0yUqEAq+T|L<6y zcV_gHD_8WGFxn|(%_@EV^={y@2T?CyTX*R>^Y>e3Ilq4T#G&4H9|J1yf7zn5qhF4n zwixJz{wH^Q6co$9yjfLay5B$hfBJKv>;FFSe_9*<{r`^T`K+ye|E;`^*n>W`T01ng zu>AOd#tfl~%v@CXAy@HQux~=}$@1}`MvTxMVsBq)8n}A*#ej?dF}V6Ax4fnesZcWD zrmn6oLqfK-Xwl+gz=x2a$CHU6Xf&V#@{28SW6S?+zk2Wsxbb0?y|NZCP z?Lf>!t*xz{7c3YSnRzAP!}m|9UJh#|E-p^K%;u5pZ{5BUcg{#kgG@T_0}eL`0`s0BR^Hqb<>mt zVgHAZ+i(%uq)Pv}?Eo0=lclTy2hSTbV{6k-viT2Buld0%`5quSGu8eE#!L!O)~@6? z2|f7xdN_wN8EW4C9#_6-0unA}Fe8(?`?Y^U#)03$OHs+r&W_e-G4zZ6dzwWds{qiK zckRQ~?Ho+k#99WnWVLLt=~Ha*!*gnW=rPxY@vkIrbFDx8X2ak7eK#8R3%)tC^NSJT zn$e3FFJ`LOM(q!G+iMLO;%*883%*(T<+a<$=*UijyL|NG@9(0}F5A+2!?;h;qz_m3 z?z&5zU0jC5O;z1eEK_KVLdG{4LY_Q5`ojk4u>+KoRCW~-o~TSb zy!~+emk(dpayMNqZS74C%oveVup69)6o4FDAeMiE?rJ-}L_xV5M_+C^e1HD%ybw#) zDqtHp*0*Yjpl_4-ozuz#o<1Av*zDL}^-la0JLKHPf)$T@ zq^~R7n7pcX)S%#?A%&-R-u+`#vYq?TP+fy=cA;IKn6@ccad+prhO7VRG|sr)SKohc zsqx2>uYY$uef9YTo!p~ILA_VLyklM~JfnaAVHX#;KKd)?YFtM5xAlVP7;;?n;S6r@ zVzekcwDo{Z!v>#qj4f^ks|7||6T!kpcet_Jr8QH{LCu+F7U+=Yy9w@|54NX8R zWvck$IiLPuvK+&5ASA^8#YrkHLu_qrKYqK-=(R9m(LVA~RN0}Hp`*V1eO(0VjC|zd z<1-jG(%f+$Zot5>V(P2oCrn5wa-})v(r*o4`;653^g=%Qn@t-=a%UCwr+a&~;iXy0 zAAWgjH~WBp<#PS}{8Ummj2%<;;b?t$)`uUG8Ay=t95ZY4o>2l@CZho6EKYF2Vd`yQQ zbw?UZHoM_2K5(EZ^H`5vwCHu!kAC@9OPAhx`Qd$RF2hC^w&1~#5PzMHn-b4bo7_#a z|8mQR-X;Id0Z_a(f1cs6x?$AX$!3xE%m26;?(gq`DIXa2@P7H*h$M|83w#ZY&^nXv zf4$kB8~Ij3(J*K<^20}7JPr|mh>MF$WFEinq^5Rh{P)pYvg1e8P=7u=@8chij^O!M zciez3q~i187aN?q^>3W+J|e*Y|-%cUkY{Fg-7?@$3eR|{-3Wh!sbcz;n5v71^rO^ z^JeG&{_qd~vf(6q%4+S{y;=1>e2;0vQnKbvzpBeZ+;8&Jho^qwhoZp~KDQ-5^NQD- zzPPmiyy9Y@VagYd%s)>YSoLMo@y|0$JAQE~{(0gCy)W*bf1c>|cElHN%|Fk4IQEN1 z>7OV5RrgpWUBYh5f9qB zJ>G?U#qf3Z%8S}LOYQ$)kfvGF==h^hY-T{b-k4|+5s~NhRqcp~_wkDMsV@d`e9Zg~ zw6yf}pf`b@qsU`cEJ|5Pj-I`8-}C#!GrRmBF9OBXw$|I{3^klM(b6TGl6@Cg(2+4h zOqPdsm|TQSEd6Rt2^H5u%{~ibX$GODa8npMvsh|%aQ$1!+PM^_9|#X$k5UhE)GKL@ z6@@Off7 z9gn!m*S93n<9(Q#v5r4po$q_U#K&x63T77+%mMgLjUEzqOFR%*%+p(>4& z>LZR)9sj5fQ8Ha|$Mgo!zWy`U5K%cZN%!R%#4D9oLFdd7YV_?Yh4%RQ$3P<)^*ASe zvh@%)W_pIGutlEw*YY4vaPvE2b=GQ=@9g#_92J=Iy)$3l zoVmUQVpmKQ>Cs5H?0O4Ln{fn;M^C8U@*CaZNBTfk(Zs8Z4090OGB72+-!^*RR&;YfQL1 zk?F6VKmUWC=?nBHw}nbpm2yOG1pE`dQ{|yf=b%?SU~Irzs&|XByMKrl{%bW#pTIw` zjVyV6!w@N_i;3^Mnu@lrBE;V5V&~wn2JyjPk&oRhq^Of9=&`Fyk5$WlA!wDt>xvb! zA(-3QZyi1MEwr3%iXiE^+Amp@Nm2bRN+N4uah|1fCFtNmH7ED|H}WSjpQ|gh5xUx# z%mtHkH5^bK^eS6XqP3yXJTJ7a&0<+}I2LnZD|Ea=ZQjPl=Hl%;+?N&deiUyQlSD*B z*IJWdDsu*_P~1<}H~^Nv7GI1dM?IYPv1Xb|iW@Xf;!q+n9Bm+W=YF^9=1 z{copbWL%&(cqY`^XVChzps%-}WoLEC;u3dM_I>x%u(RKPosa*#Tc5V;{`(6{D6aQ6 zt7$kr+pDZ28#%~w!d^7a3xP^3`2rqHn zk7N4%EFawu=GLZbAJCnTzo7i=RebQ^!TZ>Kr}XV~sLNhcnRE90Bh^)JE`-HCX~l}N zJ!f!gsW!J3DjiLv{b&wq%{32OSMdIBf8SK@_FTTdFg#E)ye15Xfq(D9=%t%OC; z)A{lBekCWL&q(n4YJ~OQzCsnvqWR%jPIYC@cw=K<=P|2!rS>yuG-ri@_-OmQF>6%% z_ae>qzGnBF1HWmEmmC{me;r|t^&SJ0Jz-($u<)A+h9-MnC^-EdZW!eKg)cn`swu-T zLlYm=Wvg=@XEF!iL4Ca(5}N;qv2%gz`ELJz+iu%#h8adq&E~8!-FODF?k%Y^W`rqEdPKkzbSDgL+n~3lmoqK)i(vZlqgzyCRmLYjk9d_W=i+{ z$2+(HFQ3X=LTYSVn$AlDdB9q&1zOLp6&%`+44g(;S(&R|V~%i_vy8XQAHb8WE@oY|JSqWwqR60%yj5l{yc&|Xw#|H?T{^RMso&)>f& z#2F-~@QQ>2BH$ToA&)jO@vE5`&| zCr$aq4>Q`DYp>PZ9bIs2iQZ2V_e;;55vuTkzGFWX>YJbe|JzS-ep!%lcHtGRhC2(q zpc`vm`s2Xu+AA6NH&C6PEYFxeX1|96dU~r~GJ+WsfPTV-A6FohVF@p#L+rj%+4=HI z<82vBs}uU$3~_hu!$H(*ez`~M06)WY^{D4w7OcMU0<_uFqV*f*{)^r^*0-ynEsf$G zy6WADgf6`U(c5Rj#v_~iMce?IfS{LR^={U~+X56TA*2Ro4<=l9DZ=1j(i zV~cS|4(ZYUDq=}nlWXsdFZi*)_1>JEoby|1)+5!rImgoXmP>1ii}nRX`u(gAk-P*)(YBpu?>j{NE~-&!(n1o!^-uoJzdqa_fA`IBsicM-d_=La#keEH zwT+fOGg8<%lglrKK0|CG)pl%wG2*Oh)gT{R>I`>qoO+WNMm! zelrzB(&i3WpgCAt>NTc3DLFba_c%3gSm~K%KIHI(Lj(H@1^1}ls`r*qYqB}HkJ5Xn zds{{H#RFt+dh_2U_^f?&Mk>F#Po#)rA5t)HNwaWlpFb?7yeoLu>E)#xlFT=ezm zM=IlC7*@qKS@SSPp~Z5859dxk{3NY(D_8$4xxm=5B5W*vJr)}_%$wqyA+Bn294%^A z+5Gj9{rUUfj#I&&@A=2?AL zpWpsJ{Y{~gwwFSIRt~&4i%akbHmMIyXQx zQMi?70er5N`3J1VWKO<-FbIuzR^Lj!fB)yL9i7^L`V2~6==tz8o*=kvu@}ihsn|HX zN2RN3e?iG-53V4!9Z=DlAunO%Mm_TP85Ut>4~5sWyQZe*e9wn((a~+IMg?<-d+M<1 zO)o+>Hx!+iE$L_EfBe=R|3Y(#AOWWN3tSr0IVXKwlP%IE-Ai#T2cfiA&Mx8bpx@^e z9zL=3*u(&dI+`VC-axMIlTUBrnye!NBP6+RtdRuSfsM~mz6w#Ub@*!>B(4>@>i#d+ z-p!93MFTTXT4CA;ZOZuv85x3dZ$<0G4!Y+E~n;WbQcd0DrQl5KFG|} zX7D4++z7?>jeZfL%QSH~>9NPGS=MVi&tGa_&_}K~4(aM_rmHYSp)~2k2Tr9rkiz`f z@#C}5X7*xEC8`7GkA|*enE(%1H|a2?u)`o!;YY6qT;JZ>@y+b=o$8+>Vk4FOBDCzb z+Ga}sF63G|CfO#CR5)|yF@-{#rB`p7uG9#dd{IKVw)OTEFHEn%rWasB%B!qAit zB|TH!GI|}h(_}tzi%PNwyB+~mZ)qOY^EE?m@e1Ctw26l((`PfikouZbYyvgKx{Gm;3g- zpZG)?mf{158q$o-zYX4V^XI`|Kl5xY_4J|()>O~aY;LWOKTH7@uX%yLbgW{jO{&?~ zn`B?bW~7WZj^3^6U;la(=++j% z?dg1Jf)XNVR;9dn)Leg#x|rcrAL*7UIucm#6h81GiZd{XRK}#3l^kRg^rm=HyfQ&x zfi%l;q>kC5I-IwJdLKaPAi`9)K!a$#dM{eJav)^`$xQ`hW53LOK$%{ZWsH7}7L#=8 zAN120!2}~T@2C}^OEs7I7S-VZ7GGNL-v0L2Uw@TOXeawThon@OrhH=UN!h#tfxLJY zwb-$x?<*M=LuuYeA25a#k~XqO5AMr{f+o=_Ome8N9LQ;Zq{TzjH!!$O>F*U4HAOlS zS+8mDFk1{*A02=<&P=)5rM(Q#fiGaVO|ORj@}Nu7!=(fpBiG-=#|jh`!4sZPy0aoAm6l$k=!QuBV&iiIw*HNMq-l zmwtYJ7O9Up31j2}7IhTs?N0|7OTW9?+}l6qgx9Bhx2L95L3IIJ?PMZOL1fQhoKjG|5kl0LdbQ@{% z-sD2HvNxe$cb~2iM>%pEo7KKUhX&Tqd(0zwMk{}q$2f)S5gM`%?W^{dMy2`5dt-OJ zd~)eh!z`(4AT?NLp25H76M4S_dbfn}bAGOeQq3@}?&!UKO$XI8HQuGIB7Xb5vA2iV z*~H0pCqvck_H3d=dj|Z|l2cyz>%U=%MVGwEDH2&wOg_w_$AT5HhZhqr3r?CH`st|N zl#y|@?p13x=5D}I&AJ^tNOC(K*7O+5i|Nw2O)fgNL{POAE5`SsyjzX7Ts1N0Het*t zMsL*deMt38NuJU4Vlwi#@tfwwwZ@bbljL zl_a_f=hG4s((W~2qxjdc zbGfPgkQLY~u?s2KxJ<+L1EKRMHRVWGzz2RtLyhe~2>V$bmgv-es?#BD|9%l6+VLsk4=YtP`ddMAp0Rj>m&dvPeX8hYPPXd@e%34nWBA$mRrY~KJl&F3A zn=c>RJ=)m584iL8z-$9SIv@Tg5(;3z90uB@S}U|P74n_cWy8{@yfEf z5ra6JeReGm_%NGXnkbAM!vA}wEC)$xQM9&7v4a-208fDZV9+}Qx3LNO5%OyXKE2}yH^II-0SuVmuc}B%~0FO`&De{8D)~W03>)#yXT)khO z{j$tQvM|IMsebZ(_XNfXW<1(dF36=zzt2V^g$}T*I3OTke;4&@pHK#UlkN1A^#}K2UEuM#K3=Gpja_|n?={RkfnEp0Ck36eus0@{zH?YPXE>_pNQIIsyub6{pYcjj8gnZ z)agkk0U#WkU;gsmZ7!e~vA5?>KKp9CUVi)1K3z&u_1ITwmsAS>hJx_>Np~|x{cEgk z;cvXZsb0L_@^Vnz+wEtQ&O38dZ zR;b1}#sMO3=B_6I{SKgA@T>_b>9Yd-R?hWU98}h|WZV2C?JRN-X6%c}-p{Nl?#ZRj z^GhQ+js-vT)Z5E;U18zoZS+gkvYjbc@ltTE;-34|NvMGIIALt7(|&aG;{^yLFmc}w zf_MdIjj=en?njgLMi#2+j;tThAAo&gm9iH)*F$R6SR_5iyRmDR7= z|5dGOs$>UxGxz3pd~;&_OnhwgUOz-gY_-H>{B&}wtZ1<@v%+lI^r0Be63 z#iz^@$Zb8ijoW1heYj(bn$&`VUP?_AQ}riZ;rCNlnE&JanWfk@h+!KKIczR7PI(pqK^zmHL$ znl&}5xCnCiIO(!CCcDq(j*{Jg=~a*d+sqN3J*uv#L`BpJ1mp9j0ydvcX)d_)8%iia zbTuFOj_lX3-!sAgs8C0hIFk5IYFI9g;ZIbb&Tk$@vFcGfy&X|KJG6a1WsgmLf5`?W z^&2=02l`J^cScB22FWI`Rp}Sx$kOT;csN%61kgU7CZ{%}-m~R}xz>L9CQ(F-EGNsA zRPGYm2+}oCRd~}plU@@4sltsZlJ4c&2z1oUe*aw8pnc;%rRi3Vr=v5JnG48A8#1^2BuAxpSWhx-4LxaDI71n$v%%pXJZRL;P0Sp+vXe%}q?}o=RCRAn z+OBge{>92(vtN%q@a)&`&WBnIl`+fYf&=4g9L8LvlWZSo;AOoT?qA8q{Jne8pBOH0 zQ4~i^>RWXCPqIEB4Fmj5Gyc0p3x9Sm-^{$B`nFc?Z_R1vPBf@wCLy7>3{I$n2~Ht4VzA<=W@#O~ua)Fl_NUlG9t(ag_5 zjm-xHc*NkrzgVOkAAsA3m`DQniu8DVUFE|oUoX&Vxr)YDOXf%&-gXcTqi6udzs@P5 z4RBzRu@!x@XoCiKQR@^b$2@#PpfhBL%gB(ttYJTGQ2Z;}`gs^hhy_a9OUDLEEHOcg zk1gRAhZw7$peL+25!ZIy-93=@#E|?|Z&d{<5lkdmAm13p+?Jeiy_)0LAAh^_xVd?6`e&<-vf-+1Wy^&oT zjVXeSJE&9!Rvi=YUBFGWF|ghg6;;MEW8`P%f3K{zqdA8LVRHG$|8{);?H}wr+4G|H z!KR5uYFhX7ZdYN!*gJu4Th>yAzU(7#o;z7-gdGenQUz}?f z{O)N{nG=0jJS$rADynzL!=Ha@+$|d?aeC#g9{2RDI*~lOVmwRe2M1`TnU{o59?*>u zjuqmw$BL{FnDyfZKN{ZBwvp34#lGfzQ3E$#JgR2oztdVv$7u;*V;5ZAAAytmv_~7R zel}skQj+UstS}C|UXF9mg+FX0>vZ$^=V!${LQ^w)j8j`V**5dM&e08}uYJkz7ct(Q zf;wn!K=VbBa7QUHyl%m%akEIwbxkWQM3W61xH;OTRqQ#qcg`*;kU(}&>)i=(q9O$Y zqsgb%7U}pvN+C-Y;`x`(<@51D19WN0+~ZZmeRrF<(Jj9bBVh3}6g#axa0QOjO(-RlL?Q z(bqqQpv=(VAbT*Coem}o5Xh$pS!W*y1xT!8c4l^O)8*-Kd0&_mba$dAO-R-jt#{phS1fbk65^k)@x7)XIKTREO zA5ivA(P}d$OgWc+om!0=!-%Ozx&mBjAnXZpGP)E>of9$d^fU?!TPP6Ro?p7777ZG#BaDcu`W)yC_}y<;?9hY^ zEe`A$w#LLw#9dZCDj>|>`Nhty$J>mb1&1oN zBvsg1VN;6&)zz#ll>9Fem%z@05IPN_$r+8nK6W`E2}g%uW`r>1eB5cKY5%Beh>6GE~-OkgL zwKRR{6~G>Y>k(s=)X{9|dq+E!c6T!UjB?dKJ9kb4U%AMRO8=@+el!?XXxT`Fgd8)t zz^Np~GYj0au0FmkcLyK)|N84my`K5C3@06hp3iBml=r?!y;AABbDNQ?&#ijJ!iLX% zbIY%HM?C5RkIQ&-5F-YO7?sQqxWLVoS@@$6n$96!TMFGr6uiZSdcNn3mAoKDef5rM zz=S6rFq%3(+8{aLaM4`(_l!DkmzNucZ@36Y(2s*=JEiT-w(8)+ZO|9Q=k65)m@+tX zV(}<3bQEfYw)q4yxi{kSZc^1a8|2gn6tSEiS5ZtW<_+&AfWcI!%B|74P$%}n9a1wL z&y!>LwsNyISJB`M`oyu~lXo$cxsc-ZjpGOn-w8-*kkCQ{Qnt^0N^5x?Z?mN;8&uSi zKZJsizjJf->_qoI>c;oc(+j|^!Z5CQF8G1?zc_50&p@t{cBHN{*NEHSoUVBy(TDtu z=&_!d2vhw+w7n!rh&-TZdTbp*1Eh(b>KD{%_3N`ngHjxz8S4`q^iM(yty3LbF2Aj% zY4J^UfJ!`UKjTPS@q(jR4HUx~FxYS9kYT^qxV~kz%$vk?#31V?RwqDABHTMyp=zHE z+??poi*r2>u;s_l(e*qH(0VkO#Pnk;>bk*=3p}P>WafC0xM{M|VBwgUm)~JLHD>A( zfBliU^ebZ2ODu8B&L+VmozYxEt!q*N=f@ULpRXxU6^|<7UTolp+a`&HVcCh60i0K< ziX3o-Qj4`|vVjA1!*#!n?B6WlJpQTl!__l&1t9%ULABd$hV;&#n{#xG&` z$_7Wt<3FgPWx^?Zb*rr7O!w-y;F;1FV=q%2iOblI1r`>5(p`<7F?mTZtkDc$5&bl3 z*`?{uO>M5GhwU zmxkbS=V8!Y*})r+3xl1Xce(Y5I~V3qx4vLH+jS1~(iw41js+`sFJ&|%Wr>(KnKlCT zoaHe$`Ur~gfBf}Q#Zq(7^zdd^Js-yxCY^S5)w^K->ft`!xrt8)ZBqL!B_;mZJSe=c zicXl8i8Uk+vZ_Tm43vrqC3T3n`vu!k3#FoFwLLf}(pAjN(Y!KQxdhqjSa39+s)GAA z(lgt5j{B3kBtdSbu_>X(_I!{P?F8>;AtQSZARtzCnYC!DLJjkVM z?GKBSaAj^&NE7L7B?UqT*&eR+_LbFJ^u1>OGnYA|m!!eopleKFww^ z{9G|EvH)`PUwH1jqH;W4E;7Hd;PS$tuJOltpaJ0aXhc8<9qF4Nwl83Rs9BvN2hP@y z)gD-6iQhey!??Vn(QKf)O?M2xt5CCwDW?;Z+jg-2ti>B5r_<}pg+=)4m@raXhCX$* zx~cNyO=sKN-5~1fksaB;YqIhy?!aEwg5X%a73Lt;jl_D6YPu~ zCxVZSS~^#*FlEbu?PB`LGf6;D?f?9%?a0xp)T06FBdgCms$8n{(1Iq?!*H8UWW%6n zgKrN~4vU?#6+M=C$}H4gv*tXr`}tj!$6-)%qlOkYFr66I`9VWG6HuIcKj*s}#$8~0 zQz964YXM)?WAOAGzRrE7U8<%>ffhuM}p5iTi%sZigQ1JJ#!q7A%!gNf~M> zw-XZ9Q}LMav{qo=$=42=Y{2^?JIzLz_@z#8Vo@68zChZg<#D&q`}i~zw%XPd>L3N0 zS`(9A`hX1OWcRKLG2&BLCNgkhf zJ?|L_ck^$@?>j9o0*{5XOMEx+spkAib7nsEhdq0&e*NK}ZsXSOqp4RjjX)?GovKW8 z$6hz()YITDl+%52u}@4~94qT+^Il*^u4=j^vVMS=&%c6bDE4b6RUcvU&4Ay*X@+lb z-jd!4rmZ!7Bv?fuHQ4z~D$FOfl~MX#`)Rx^ixSsas>M!~9RjJ>4bIq&LDUznUa=Iq z369Xud*jpgV8`L$TQ_f75GnT^IG~X?!KHO)3%tJs$8UnEi%AseI~LG-0;@ku-(%Ul zmV+M!pmLyo&AN4EfTiQ|%=WamESTXWgsh4ePd^0W5*$fY=f(1bZ*Y9%5^eUD7ty;` z2azs$?JNJQa~JOxHr1eQ+7X-jeoFgx9UNXW!D@R@uGRHnf(0>|ak4wKm@Nr8e2&|a zIORhl9<6*{$@>r!14Rk#|K?cqaH)AmZ>^8V_O|PQ$qOM&@55=@#|k3OR3T-~c=VLV zERBI`iVYr_l5`9P!X|PMw}#VJ6?uy|RPu#vBaKb!&OTp(SG!LnQOF;kxEWRPJdjsl zu7)0Mgr&~8qGL^)Y3Q~s(pWyGbg_x`N~5V4;lJ5Pj#dQQm$+{S{5ta0cJZE_f;m!A zgJ4&r;@5X>3lc|w4}b);J9$(YJ*wzQC)T3Yn2*Z$aI+?HgXxD>w?WoU0>+Qwks+Wk z7Qid(c}3@J!LfpwP>0`dg4D54V)Uj>o6gya#}CioeoZqi)G<$+e9u}HF;E^jHFJ1= zRpPOc-k4#CY3=XV0B(c+bL)=y&^ghv{%xgQS|>a_f&ka@ccgG%{b;$rGYQJsk^M2Xf82x2&DGCnVd&77&Y80}b z$O(_Um8SwZyz@nmn&jjGKI_QP?>HB4nyg{>V!0@z`$h_E31Q6KzLf=j0>AyaDP!;| zrKDT#;R4M>n#U{Xi_8S)<#A~?w;X08c$=om+)zPsrqtq=_vgY6?-z|1Mk`0Jxe|w z%GhWz7>-Xi+L5K4^YJl9UP(O)6sIVG{;+Y=X)S{b^OVJ>o4&kWClAX+1GqDR*R!;k zMoW*d!0yO->bMnoXr-^DH>_QO_6ARYbKj^DI`gNGBvbK6)Dp8~i3L}m`qJNrU_ARB zf1H&`+bfL`P48Pl$~*^gp_BK(*89wPkxak0BZUL+4{X!7>f7l+!!Zs@7eC=t~kEVLyRL#DFqxRGnKc75lp*nTOjJ)L&UtnW64qvv17PINX1r^pD z(9L`L1#WZ37?s-))}&eX{%=l-WpU7emG{COyC}`Nb8ACs%SuhV;J36KiLu*Y$MnBb z2{ep|x{ZA+gEj??@}_*W5|4R!SX+iudOK_SR#07~K|`~&e2ub(AtyT}3Q}bqInK}n zZIUdiQ`IE+M$5W_PlH5V0QF9lMx?=HPh`WxdGY<2-hUe%w1z0d zvf6lsO#}s(p<5SR92=XOn7t~cwoucGR%+N}o zNC_qgZBYZ5Wz*ICeMI)<&MHp>G3o*zo<-HKC|-ao7)L=lyCdT$1>9@7+DmES2yywv z@i_I6l$p4f5xTVs1@5&>L2Z?Oent$9<*HTt^`b0Sqz;Q zdYwgG-H#*Ruj1vTfbNqW0$?>tAaL#K7@2EL{hDjBhkVXs4|WSaC#N3f(&u?*jkO(h zSVutkn_sh{9&9+guHtT(V^^hF_nV?^kGoY|aqRj=2X9C*Z1)}rMbBSKN6XhdX1X-1 zZykuqbqtpb-MV(~o(0rfv}tnZ2StGlM5NJEtOL-hNUx+`)&~#?n+7?@f!g;)$aq1z zU^l3lYtNotPDc7#tI#r5`!fmpw!(!i@Gu|2m?q_8_b zC0t)Un7ZhpzfL&z0{uRF^CudTU^JmimdvWz#Jw&ei z9Q7CCmIA9~!SAO8;5Of4PdrtR_=IcZJUyJ!G34p^Tz~Vcabn!mH(TO-F@GY}7$Be^XYIski89T={ zR1Kv8je_UU_A1L=&zAwV14&Q+$dMzR$`~G>^U(x9k}|mTRsI#4)=A5#oYazn@SHw% z>Xga$Aw!3javl}y;1k$USf!MAv%#ZzCCNFydXUSb82eCw3!iKX(!P1c5!9edBdk~I zhKh+!Oj6Z!Xn5MT-oUUdDg_5w;;7YclJ$l)o+Mg7=$?voE&v*zO>I8KAjKH?XJTG0 zE08NHv#*%#0baM>@ma^yZ9UWGZ2P{_<$GX|Fk-Z)vP$-f2Bv?-<*$KRB}c*pQ!fAQ z-n|EKmjJd0$q0x7`k)<~KwrTp(#To0hso(|K;yAF)i6c`KIi2Q88)n6qBJ&&u8vM{ zW3+3beFvL+*e6)*8AyX;;#e^O$H>0ro1$|Qw6r=t(SQDRliH)5qw*_zEA?9S__Ls{ z#JUJRZF5=%>hr>jSFD3Sl%$y(1?SDeNP zSU!kdw;O1q0uz(02hvL!zt%FG?g<3fBWvL9wRO$+D9Wi*@0K@XPnH;pomTaf__K}m zRW`47&YucVAUn;Nj39*t7do`jW3ZUN4U1pOimP|NIo)b$Ml;>5xy4aa;&oD*v8`=L z>hxA44s*Uq8v#S$oT3KuneyWfZt1SXEiwmtQZG)~1SjQ+9TdhzTTR^a-(|3Rux-~4 z-uc~>lEIk)ci^IZqY)4A#%DO1+HEQ=+pje@d^>H%6idsR(19}S(`ye>KDaJ2r##fsvP(9o_nHAWXw zLMcfu6s-WglyC(OOyl;JjQjZroLkCc4o6d4IlVt&*#b zZz)}`N^+l?WzmW}{3W`CeifIEcu9iXn>7qy;PHSxW$tf0+2aC&`2E@s!@=T5fdVZ9 zPNW_ZyYQs?6O=S&*`W5cg(~S)UBq4(7A|=@SE@+9O25Pu!U6C2K+*sGg0F{-nwsM6 zWX`Pmt+Vex+MXQvV%lMP!L!T^0vL%pSH!RoteRrYcsJk$<>hjZSonD{cVGLEJPqA+ z0oN#Fp47IIVMZfzlXmHLK`I5=vD2qV>)1x;IE1n}PE|LcUco9vThoM%x1jil1-fpi-D9KR@GZZBTKb_p zfSKqeDebfxsXsEQ-O72aHC--bakz2W9MCVG@Nl9(l%bjI%v1eI} z8rO7=HBI~66Y3j>^nb0RtGfgoF7*&o63l+D^CBtMAt~8=4}^I?(d~B{f5*r8+h4Q1 zrX4Q4;`mRcS#FPN%VXbWG_^9iF?Kv|x-^jC8qwRzmj#DRvVLq~sZPI0NUR_A=mag4 zFUQ2tE%)-O?^Bv|kB{;J;hYb)l-A6Ycmh3!P8tVaOxP1wf?+R#So-bjnW-Q5>0@qB zYz-4%Rw|!0n?4&4bRXC9+DJE)e5yz+8W)r5Ny>`!i6pc-(M~D0uytc?9}WbrFn96+ zOaMNpbqpSOm$X9?E(f4qO0TgY`cHI~UhbkJ3QjOduvJ~^x@L+6)9j0Pf8n^^!ayKvfi zT+Sl>t>u~G4V+LjiE{((y?^2!m-D{9R)DVBQyX7bSZT!jz0Xhy;Ou}M5&*wh#92~x z%b@Yd2Ae@{Gpktvn!DEAyno{1Lx(cO=TtR*gk`GNZZPF_s%^kECexe$&($+|pHFH? zxR9ru`a{>-uNKP_3~i3AFOKV&6z&@H`%6%PC?Vi@>LK_i z{gTGa6^Qe<&kpxZOi4+}qDeVebVG~9WJwds?Y5CX^Omadex*5K0W8ZNC>moLXAFXr zP%~e5-n?VC3&y$*&4V&%vfep`)NdwEtGJpApPYf5c z1_q>-pr~Yiw!)2%lm$^iEoqRJvoRB0#X7Uqt5uUut*J9UFClp9G&ZpDss7O?n>lM; z#UlgI8AUOD=>%x&zh2t2`YeUNNsZ77e+yq0UNyQOvK#w`D(i92S~EEHw)Qu{wHG?-O! zAcobDhiN!eTQ=mKrh&H;jp+8jT7O6vq_VA=zEiZx5^6^4`<<2fXyTZC7-$Y_d6HhmtSxJ^MYTdnuNEX zk=y7+UO({1zzKueUj!3Pm~}53qnK&vaMca!D|0jGHe(@dN@>B^d;K?wzg8FTyh`cW z@9|AHxioH}jhktL6YHNM-J!c>*kf!E!nc`r5&zR9Pc2hnbqO~oI;+7Q{3+Vt%|W_n z<*Lh#v5!=^&|rl&M)4V;!;ZST7FO^0;7{Y9>s5bThu>oor!28FCfvG(mLFHT(^r~h z1kXS)p=MPkrPbAqNV_+{F(~5}Txs_5ry$Z+8MGpN!>KD9(qrtOq!=wzS^Q zI{TbOf+ZMA@?T<`>_=OiP?MgPc9llitl@87!(aZ>Ryq37$uc_%qI|-2*s_szd*Q-)&+k%6>{OWV|25W! zl1x!74?Hrra&&SF@9J(+QrNN5B&VdTYb9^=92}ygU}E9wxnRvqM7EU&My!XW+X%v; z>N5RJIs+*4`p)q^xFPkOotHwBPpm4RxA&)?u4oA2%83e1?05z*-)pBRy$k>pgK(I$ zJnqKRg&dZHMSa#Xqe>?k9w-_`C%8@ZIefjYt z&E&!|!kQ5Ij_l+ZxE?4(1>qtkop#77PkIgS20&7(2z{B|_tSEYp;|dm$ai=S95uo9 z@4N2nTK?LTx2@c3)26r2RlQrgsg~09Aw69@x^JD6@u~Zaq{93^3_Nln0@@RY`#2S-@^{8DrCP7t*v*b6-%*7Sz^AQdFd z(E8k{Fqt)fe=xQnDYgir+vS!It`QQKyCo(jrqioJdum8bUJ?7A?qYT<)KVG^PmNpK zi+L-0jgRa`Sc7w)r9TqvS6EY2;OD{2%Hd9Gr<~I`=jT4*xozYNlV5@vAg!}^VIgAK-V9p@O zoQ1wg@wLfQ4;FI(qPG))0}iYyVa7+{ajSCGecIW9({|jSV~2O}5#xs?oOfYQUTG$d zv=Rq{`ub(}6vaeASK)01;B~L?9EQo)kW6|B?3CYXHK80JUT?+srz%T2^ zy2{GRosxf|RslGAaU*HrfSI#qm4cLIa^0Lf+~NMn6p5eA^*p6v02Ol-?rU2qWJi8} zzd2=GY-9u%>&BB|aodaUDIXtW+8Q!`S=BGDbxlztd2X2`^$oe_M%>}861HR$d?1;E z`FR!yja2dO_^Oji?Sh|=r`xv+e%W9BL}F#{X13|>W;)wcPUpBSVF3JVOFlR zT7Yt;0)N)9=U^or>74M95NQ_c=^Z-qLb|0pyPf96k9iInOxc}!$moz{Hm>$X-ao!r)U)4+34wjTsqgo>&{Ab0zikhV zRC1azA~mfE6!cnyX7+WWD$9yD~s#;#*(6ZbUch*t%sh@TE)s zJZB37X$G)!WZFD)<`0r{;YD2p2!(f7XKy-p?mjwJCl4e2<&hC*=w56+a`O#dR7Na; zeU;k>iqsSJC&{7I(J^ohSPx})*W>~ z)u`C!%$++`1Z}^~HCP@{0CZqP4x=GCPliT)Pr2&b5CDp!c&wU#-kY*Yu+UEra z7oLGfGaK???IYzaoXfcs$*fd>q;(PK)6}IUOr}wpnm2jY0}kSbefJ>eH>)OCrp*+jM~aU?0PsE3tmO1*~qeCIQxD?i9v-@~6Gnc7BPayr;Pk)?@Nzc*!; zA(tvY3hd3B4{pD0K3Ei!B6Ew+dJPBC$Yeq;a7X-Wlc=paLS2DYHZsgIT^lj?2XMf4 zx;jn>waveG_io5B#A5XyZtJYZ&IAGC8KGay;^m}3I(QD(YzZx5u2uHkGRMv*dkol6 z6~FLhaji>OepW|RoprSRxuyxOQVdHd$HIAY>V^6J&9ZUyy-$DwZHXY=vJf%Mtby8# zBIF_u9US#zZI#EuhrP0)Cl>Dq3xR8vu@g(qKFqu*;rZ9wVVx>&JpEi@YJ761=_VE) zHs;6WE}!u?Pw+4Q$k^)B?*=+}olC5opww%yi`DbRGe0-qWSsQf_up5Te5rmStV*W5 zisUnqTT2&n$U>SMt{hiiX$U%5Q|87LTNb^4ocPt|bc$s;VxSNUp?BpsrFd!u^AKTK z=}93$mg{nc&5ep1N&5zku2;~c=&o2+RXVKq?t$~f>sd&8pakn-eug_aeSB+64&B%b zlc#R$YK-Tb)|2G=J^n+di{W&!Kv=u33h>bY)t@kIe04*1nnf!yk_fP!=(4uq7w_+c z@^ECwhnd#YtBlOgU6C57U*g17U|rg_1qB5M3xAyvGv;OYy#IoO`9*cj@|I7+{Cg2c zVJa*;{`M<=&t+fr<(IEr4owGy*Yr&v>05jE_V>kwg}v;2J{>)2QkKn1>4uT%D$AIF zCA@$XruCo}{rdSc=vz;F>cApiuYm(A>IV{n&Vxmi0xo7!Np?Q=$7TOGAh>*&Qd67i zqF!+kr$V5$N!Ag@7UwR%X{GuJ++9D0@Ulp_11)*ojSo$#S0R#a>QLBGYJGUza|h03 z$JtAP0YjIxGU_Uc)m$}xvL%lO5b9u2TdilChN5eJPRQE*sm;XE{z+r>C?40UPSw<0 z!D>rFjgvHQkl6n0b+akbu59p4zbG(b;ut-~LA`tSG*k7g|8q?Ems9sYe*45INNG=A z*#NI=Ap_#yHBNkXcglp=EqlLvdjNd(8=yG3qerkolV(hxIpEM}m-+Ah;~(;1;S9a; zDIt?`50`)2$i0vmV!wx-Z>^wOdl2v90u-PP>1!zxF1?hMOS1;*F}-9L(h7wO6Vvjo zCb?Ay17k9EX&*-Nnj$Wj_sYq5<8&R!1~!jioD_ObB1{<0@Bm~H#chUDl<-4g3_OQL z8`3=CTdUd74g*`3ec>9Z8(}k=(M_eqmNwuwON5iS<5Fj zSCJ{EI5*Q&J$$%a#v+MhtfhB>$=L6X-n@Aa03Mf;IzRhwIxMx$+cjUcRtzqS8=~u* zHr+1xo$D!HHwL9~vdLk^64mUU~-cQbTwUy59zS zLT6$!765A2v}v?)SLzn|EPVcQ25o5^qA>CWZTm8CWsN+~)29R2g3tqe;TVSPTq(~j zaNiy#$GTexHa|`pIcVw|S72MBY}lr!Ulpum@U0I$=~+R6OgJW3-2Z|aWnA9=fbVmw z;$!S9wwQKQL=Okw2V{6+$nNe$?}llrbw-pi4=2JEk`3O@RhqQUNHT$BoGxoFWxMS^11aC5)7 z%o@|`A&-7X`Anww{h;IZ|L0$?h4_8$JEPp>t1*{SGNU|eDoozn<#{!@-`UL({|GSl zORrq|WH?X3F8D0%hVGW3bB|rt=+nNxaA0&Y1nE+MwkyZ%)ui8sYiZ*5LByj#p-NJG z8W234t}1sPMbxS6-Ma7(UOoIC3nmp)cDy``4pV>GDWG+lWT$@mk4+!i+7&ZARfi0@ zY1=BW94D~%*4R(khk1}@Y-bk^<*rv>i~P;Wvz|Jh9@B!1qbGy<=j|*4Sred5YN~IO zr56gYQRGPT>DOKZUiLZd9GBUV0PQYU4YyEI(n%vs94MY-nRb&5nC`deGw%ifkqmh) zSKryq@P80FykDZt1HU7jBr>3&KXd`<+>QfC{S>cJ3mg} zVfw^$QtsBHsm)|We*iD-zjH(T%gcFUsIUIu;;EHiUHxS2g5c)yMyKrUZ2ocm*yT?w zZzlZ`6|DD(`8zAkSlM`C@+BJtp&ACTt#!B1GbzS}hK8yYt<-gHZ7@HmJU(Ig3Wydm znx6oh-LvrA6B0U|pliPg>-~W791#D6D?XI$yppYsG=`GCR3f9awDQeY5+u?{reP#n z-I7q&Csrh@tH9 zdf7%Ia2R|oW`kGn50qMsij0Y4%IIVh#>lo)+?gsFoJqQOPxY1ctVJxy6=kEc5P;6+ zyAhGLZ5f}yt3v0mrqP43fj<%(^JUe55xlIgeAYx0d1sFLzF<;Sc8&{MHi@BjdEOmQ zBr8akT6S4wGtN5BTU3qbVTaH7!JrpwNM&qc+z0lv7~;aY)V6&(6)0|kn<>jX-VL`F z@hVcAnJ>1#pQ)Wc<8w#=?trL&yL&f{xN)G!fE_L-!4s$+LEWtYd$mKl514!wto69+ zinKZZMH%^pf3K$xx~P3uXT0P@oTf+c80UNGPkTk^U5GsHy0P9)>C(Or6ka#nqXws4 zj9&k)|NJ7QPMs#dFzGU}X2-NCTZ#_|C*qOa8$deMq;?d@cz5UCs!2>UMzcBEJ~s=t z69zX)#mQqT-Dxi9=b7?aQz``0Cf;&$j`>%hjItaDATy`2F7AqT?i43z|2fe9zjawq za_^Kk(4NU=b;1d=Uw0OkgC|?9y)c=5BalMifm;a)ZbrqOlRc(e8AXf{wmjQTSEf~M zn*0i_64sbK7PBJO)%vyIaRmcMn8Y+VO!%eDajuRuNdKPuL!iuhH!#`fofyOCoQ>p< z_mLNzj&jH(Cet6h-3HUl^U4NF40ex6KubwLzrQoe4Q<}p*s+TS9lkwhqN$C+2|%X) zPA5;)cXGMJArSV#kqz8klBn#2{3raT<0Q5J+Sa4jlZAUuez9#Yw%r?)BL}(b)sINI zpWQ>L^O*Zj7wYTm`Sm*rukSv6&IgQQlD)B1Pfu@|UPC;Z>1h5`$Jduko0`uN zrOI-nBOtj4f<9RBU>hOVif&!7EsjfG@?#N(49*z`1Vr2XTPd{;0sEh7TEBKIBx4%PX^d|8J6q)ZSau;M6wyamu!QcIU_MNw!SUj9529 zosT#4bke#~`^`|EO-??~VgH;0@@Z(R2e_#hbM@sx-5JLLYGs`J`n~U$w4}kXRLd$Z zDvAtS;I+ecd-vB@K?zs*^T-yncl7jM9d_WZW?dzv3j;vvESv#P4ixt>V2+k5=tWUg ze!M9|(us=0-|6b=j(6y}fCeG5p|+o-S1f`OClyDPgOF+4P9HjKSOQkd7v>*ad!y-^ z`pbjNUiv1XycSP6W_7irCi?$Q!2kc7CplkLOw0)UrnauH(yq6Oh|jzg{P-ykzv5?oF!6DXE^6AWB&%;$8woMvD7(%C3{`ZL{BGZrt8^(ij(iy zhy+<}Ac_xZh(;{ExOiyd_+Ib9DfDOR!u33DQ^D9f_Vd&*cg-!@4B%#zSB->ev``MJ z;pwCc2ml3uasKGr=tI=74^!C}$06f0-+j92&3PBHoKIAmb4AA!dO5H>e$+ck)6{vO z7S6Jg%NuFWb~bf;R_`$7zc-q+IWhpoZ$ila@tliSF;f}53dyagkK$r%90^B;Z$~z~ zc{i(5+$x8v;toal|Nnvi^>-sq3f=M&n0~ZcMyL1QGje`+-_(@3V_T-3e)fbxS=i12 z5LjM8DM?%xS}I$2P~O$*Pr*!vnq_i*NIx2r@7Sz_{TI-(^Okjl6^3uZ z$ZT$DQ45~RqNC50*B_Ush=J3Ir#eAZoS+%y6*^QQA;wJzlzhoNW!&T>d7ufl8?CK_ ze$iQ|UszmRMj)6`>Qq;@6y<=Mi#O7=bY0NFX&>H86b=9%=F^I}m} zPiDig%FP-sy#DJ$`rF^1J^u)VJjJ1k(eUEj`IB;o?76Y(;>0Oa3?WJ`u3)aHmT_EP z@P7`OT2?#?5WJxTm$q6y7YiP1EjMyn1a$y*j3`m}{F1_vdj5rG`=02%2v zuuZ$Uy!c$&ka6P{uf2P0tU`rOP__SV!Ck!@fbJ)`nV!j@BeZZ{G`$L_wT>+q#{qYe6XC1u|b3I%3e>lrPK*V59G+3xaI0Pyx|;=h30_(T_3#JtF^uVbtYq!EEu zE(%{wu>~fiOvEp~<}}TnH}C41jlYs59z<`qk*zkjv1UI~>tCQ0A2nl!iJouz@_eQ^ z!m;6DEN+oiPQb;MXDL7rb?j6aRz^Z|5n589=KlR0hQu(y%y)-3_>diae0?2#vWT%^ ztk=jA4C~9OX!rO4z8Oi)pSA8yHo__wso`|PSQ1v0u~R%)4ux><6=n0>P!VC*>ntKI z>yyuD%1DeF@f^(b5w!nboB~I>VjK;v>Fv4^vd7(cK6tns`kK7Q{!2#_g>Jo^84t4I z4N)kuZkcKKeeujB8qQbI_^M@OSJ>sEPX42BM~(X==k1--m6VPxgB+z@%U-dKM*KA_ zCp+G#uGhj+u7?OEh~iZ>K8wQIYKVl5TGNIQgEx_(HXxXmO- z?@1P+S}k#q&6FR_=9rQyS2wuK`P(b<*MIn=ZR6u>ix`vDtWN=?XFdD&;*EUm$9Jva z1j7bCW#i6r69^?*d|DAT(=Ay(mHClfRbI>K_|WVP{>!hxd+-t#hJ(UL(N zX7NU{oFL8F8^ZIObES`@-CmAv!_{NL-65&zA^EV$D$h%u&G(6?=Ob0skU4DFFo#uM zZ4b#>S6)Cc60W2qBAR`qF}CLB05Wo@mL;rf2qXQ4S?3ypgl~%r6B5KY{6}TNTE-zr zjClRWxQj{dHzOaV%bX5nnpaUN+q=B~{z5ttuDuVj$%(H~87q-wwzn6DDhwes zR5}$Tcm2FFj`IV-DzRi8A)3J(NnIL7qoDtMz=l=5lg6giMbk;%!d^GMc zInwYb%A|atdX7GxOFf(z#+BGI`q9x=jwQ%I z4y;&8Cg%uFQEWo-H4i}>{8>-9dL|nMp~SGb$40`yrPw#Eui9d3d+g_*&&s|yL4~&z zekS*d*=*fyP4!h-?UWRU@1`~xE>gf2j|>Obe|LD~M0k*lMz4hp<3!)r7_pv(@`KLx z#)-UZHP1dEqkYG8G~l*T#o#t-RS%!JVx`#*OK*T&YXG9(S+9{ zYi&i1>>n;d_x5I0<}o;=?UYqz6_sGL7}9N>Hm|B{`H3;jv^y^6!Mr%C;}rrC@K%ld zg|l!lY4ddAilPi!iMeJd_=&>PryjyWDXCHsIvr`MS=_{M<9wccL=6-n%Sbyyf%V!x z+ARDcn>s$YZv062?Y4nX>n7g`Q-cS^s=0L_ad)gn*L_~`t#!dB zvTg_W;`hIT;4{6W^o8ZN<~N(Y8+dyfZl;f6wdY#ImDcyE9N*%>zGxx{5eyn%ZbZMt zc!mTW|)ia1cqTJJ7 zF%qh-ec^sKh~>-RRkkhm^-_gE=YIzw^#Jb@2awOBvI4 zJA}x&g54p{B25vc06bV8G^U%-0yAU`Guksk1bOdLq2+iJb`@nJp+t46>|T^b9x}Ta z#LkIgJc05OxpjpUFn+r2??}kmD87ehh{2|$C*Xq8VUXn|^hktOB3jAQ< zL}0u*>O1$8znw4jZ|ctf@-2Kf6)=+GU|mcKXjUE4M~41{A zi2GkSb^erHI69qBV&~B_!M=peRt1%nEYxZ{&RNX1n_8Co^2^ts42g$;?oTmi=pRc__V}(d{C>48S;1&) z%!+?}cgz7uSF#qOC^rgQ?*qB>TJCLkv(MQ?Lp~ug_r^e~K&dnjzJ7Uu$+ck+^o-zN zbezSMfQe{2g-fo9Rpa#eMOku5T_Nxf}Xc6{{z_19j>2qbM~Q+D)}DLHeTxyGzQ zu8VYLr>#?EZ}>GR3eR&v%}*_342m0a-oZpWKTh{&IvQ0Mb6mEpKwWp2{VUcXp)_IB zlP@A1WC@$ibO&}1)G|y7lmx~5XiStW^YrqkZ&O&YpA!u2B=Mmtw+J-X^-8m+Nc(^; zMIn$8M|~2Mfu`#v*=K%ZjI4y{WBKGDhk7UQ1)!22Uq-MZeJ$`OsX+M?b^(W@WkrnD z(a|ij)VMJa2>Buw?};9K{snx`ln5aYO6|!COW50WcMr3`23BwLB71q3pItR$l2ztBlH{aue@4A! zo!&_Ma8O<`0~^(NDf&Zn_Yqv~SVu%}Te5$u$gdT8JrP|M4bNJ27qELFuq00nm{i+) zCf94aVp~Ja+7?!Ngp<(?5Rf<}TkkcE6uq2ux|{jDlqhR=B09St-qQMmK9SRplWj|W zIr*8|efUHU0ejS~iU= z82l;9=~*xb<4*wjvO4)gQOPh$cUf_4ZEdx#mldF^uzLhKRFnx9o7z)$KCJ6UtrzwS zuZy0NlPjCLhf~H{WvV9Lj=xOzh?2=9*$YyX!7&OJ`HbgVev~4VF}R^WY2gwE8hQ~m z8z-I?I=$?^EOz1?3hjWxcn%|d>$(smbQXZ>r=pUj8Z2{J1ZyhauImEUgwllua|PWx zink?9MATHq2Z+*w!z&A6eR)U5WuE==oy0Je{ znn?A_39kmNZFKA?tvf#Q-~KwH6tL8XYnkSq@-~d&EY{4=q4lxczrMCMosOYPE)<=w z=!!y&J<+2;N7{lqYHYsSO=6q}MgmMLihjXF5(;cdswN%3bgAFx9^1NxrsN=M3bj=HlX_8%F$>4Q%0I z-gAZfNk%ntPFMvwq*vd*&+R;gB!|>ubWtNDii>DdXnRB}ND3T{G9cKkR=FqzuL;fW z{5S%+pnZo9g&VhFi)O)^7N5$BC)#hS-6aI6%dgz?^5VT&PoG_dDs{SK!ZiwNcUdka z1DTw&rDUG2(PQ6f_>WeUdV><-&PdtZp~E zdR?A$z3BGQNmIRE&Wl$$v%O)wQ}QV}4bTCGM6fMZQ@iE=arPZ>QJ&e?B-zA`CB}%@ zAu0+gMFj+uViZuCD56xsP^77Vh;+u7Xiz{DlqN-xB0V&bIwlH;2qPV&McN1>5E$CP zF!P;f&}_0PoA1vryBea*`#$BCbI(1OyNd(<0Fl5TSUq9LlhcUE{tlp_QyTW@YtAIB zN9Y^a)h7mU5P>mJ@;nMiigtSJJx?Aop~nseLnOZ6N(*U{7to zFhZ1OP)w*p6Wx(FVy?)MG;s^^OekP0PmFY#K~G@v%{u)IlUf+MTy(o=PFyTH37}BY zkxG(nQm386M!_jd!VfiA&}53dSlQ4QkHfc$@-lE-3Ly)=u(7SH%Zy|U#xze)C{YkC z1Qb^VQe46A!dR0uc%7Bol0+F`8}=C?i-MNK351rDZx*A#cZ_HmA@5tcXZz;OWgjh% zeMDp~cwz6r1%YjZd@<7_#C-raXk>~8+Bmprfs|i4bJ6B=p)!1zDk9+vY<%#N0sW!; z;906MwW+ipBw*$s(*@E#I%d&IYr)#AYP**X4Nd=EG*mU5PsifJsx)AuTo7NA;t%N~ zeR13XQOhfI2$AOp%4ZL}pb$PXvtn#veM|>Y6sVTms?jF+k8(18eN@hl5gjjF98npW z#-Rxi5+ineeHpuIOEh{-o55@%aMro0qG&WnXh-cij-HNl>9ez4+zdpbW~5~S@zbJq zE|Q2NWYN$tUYJ*NI>#0sa^a_+Zgm_%Ev*Ctyq;H9X6ZTPx5-#N=Z7L9|06n$gVA&k zNJP#+LsA9;c!9#qOrqAc`)>7lZ**dj3JC}h{140!r;&i492*?wv)}Aecn!5AXU4j{ zz-V|ON3>$mzS*iwfLDjfxdXiQwL9uq-rzS3eK_TR{dyhX+zzAQ@bF}GvJnwBM92Ey zcoGURDT~%6_`%!k&2PH;{r6<_o4Df$*9DYQTM!Y?rIS%{jgX!o@xjM~1*oq9zNk)$ zadmKXl~jlD0mAqUT#FThdo_ayv__OpNLvB|4#Nj&F{oE~2rnpWCIwV-ExV{)LwPXi z1gXGms|{F%>hwG~zoUrD+J`@VXU{Kv`p(qg46LpWgn&(inQoZUVFXvQ5lb21#q?d+ z6VT^eV`woWN6d{A&}!WBRRjsOEtP>-r4ib!0V+L!<+MZ8k9ikTy8=H&0?~lQBn3kv zfq6YJaBAX^S~%w|8AfR1qyD4I`yA;{J2+wTAM&t$iC)j4hbP?6DW*W5L03r%9 zMiv1l)(7#k-6%LRG7V*926_T<8&hwC20IdgDvS-LqYxFkPMj&6__=f5h_TuuDo>)& zPY9^^n{YaWm4-XgusyRl4n<|?Up}dO1;0*6)I`Im0WufFdWK4xJ#60##2S&Ki+!iP zJ0haSBpsprX23_9Xl1Uo2vkhOi|Aqbb92$zWkHgl1{Bc}96vD7&R7P=lj)GcxpVc0A=c!lwxuD=V>6m}!4j(gdQ3wtTd|e-DqaIwI9(0LGkc@_Ktot5F zb3(wc;zQ-_ThQ6(P%;sMRbspz;NC|xM*b@f_rT;Uco$pHDp9QrA?>d~zexukDS?C8 zhEtUcXGf4kQtd_uz{v-OdtCSP%t6w3QDo5WNc=v2gm50s4!7@)b$44 z2y*vU1W^K{aX-h6#Ap*}XWWMTD%}OQ7s1FZiGvBdnV@E%3h;wj0TT-e8Zy*|5S>Vp zM4udN)CBrRq?$Xtx1L)Ktx00uhzMc23vv+@FTzQM0%l$H;6bFNMkMVe-5&UPfOSCu z*$Qut)LswSDjFY+?#T~bEFyOtIlh3+Tqa`E!z3d?n8uiVg|lM={y%XaL@^rYi)i-} zEmrjH067EYzy+-eFnnh4?pqtu&{?I2H(GRO)E@djBq)M?xR3ljzv*1tLv7-O0jd*i zRCIJ*`TqOuAn(Kd5gjSgg+!wMu^bjb z@r1@mWsW1-BFPw$XNBA`YNs07+Q*TFpx$+1{LNA5Up!JYLzRy-4BNsdu$lz-UVJg+ z#jlJ&mXP_Fa|$Fmnn6u+1)M&BI6@cVZ9k}UpahZd1}X2ks99n--#yi*aTQ{7-ER6jubw*PqTV zBf+FBNF1mWQ(4a7o$5eo8B{)$fzAu8#LY57vpya)2|*=;v|$+-gdsXum9gmhRcKzU zs;2+5+N`4v4kpgg1-KK<39^m|z7+QyfI=SzcV!^0aF?EH=8PvPZ-quN)^(d>IavSW zIIA9xDP|;k#h;T;}&Dhdyl&C z{?D#BFbm%TP*Y<*dGdsWjw`n5?TDZN4ZY9$fl%Pom=Lw~?++DD=qr|T$$>pt(l;`~ z^;2#`(So$OlSmC_O>IVC1<-C10-)bdkhG6B_qh2FC|+lLXs=Plc~2^T|KWf7EB{Jl z=Olf7d$CAorfukA2{hhFxTHU`Wc?mJ)I>sJ%xXOw-auL;0mD)Efvj-l`Dzf~0hU-6 za0qt@_{yS2rU%e|QYFhbttTp*xU{RoD~x$*4t&-4z+xf~1%oB>{gc;u(ZD=vFMqtq zUpe)`m-vgYwMnpk9twML{>hQYdJm4C(ZNXdFboekmwONfaUA4{RP-6C&9b8bSrOMc zZb(^994Ernu4W6iK8T?hMVksDvnWb&=q~voR{~Tccu^UJ_17;8Dtf#Lq_FSPX7bSv zL1fCo(WRY!zcN32T<<)&knhMsDATO2df0jf-9fS;WjC5|g6T&2n zibKQPPCbmEaU!`5cB!5K$m82nB+N?FFcFRvBb{wiAPdzYS&u^}fk)m5dgXI5^K0WodpKd*qnR1)GTZcOU{KKKv88GPCd32M zfQuTXlo4{)5p}o%C1fVZ-|CP`Ejerz0<&K3T~%c z$<)NZ7A&wwxKE_1vjmZ@y}QXkRRboR4#J8Pn*qL^M}isP1tOQ*alT62bYO}~B!;0= z+Q#rM`5Hd~lTnJi6tBX)Cc^U%Vl%lTF~sSQxenkxwF9nhJPkH4kr8j6Rf2Ri5nnFx zQG5>sUOGM>F5gA1ohJt%iAwpc0`>R@spyuiWuwjj@Ysb=XWv~d;Y3;ykE%xccL+5N z6=!MNjmQUy_CHZOgAE86-ztHk4kZ~^n{dxE4{qh3d=M8a{Z{Awx%j(Z_c=9SP73NE z6sR`b)}JA2#l%j-1rOZupZ}xTH;v=(r-G-l|OfNs%*-aFA&DiZ{>B z&e|hjB|-F{eHj6B06`jLC z|GgGw0hIJD;1Lz;!#NS5gp!%TR6=U*L>$Tcx^#J4`+69`d_;vPQ65Y5ho^ixo>h3h z?>hXX^-S@0Ad2{&F#4TA0FfJk2S>@^tngl3Y0}?>P9j8Ow($p@NaRl__XG+6p(V4z z>26h@F52_5yimFAp`TO-HFX;>DGL^&qOvimeh>AF-+9ngB)d0$iD}6@$#jNMq;g*RM%)5-C4tj@T=MCP2Jz9>mS=duqWI zN<%0W`4i0`Fz%<lIN5awolr@W3G92%-#8UCLizEdOS)UO zZmqN-#S-keF#6OWiW-cvdQc2bQqYdl^dLY2I^Q;#zliF_VH$rW~!I-4AEG{u(a1sUbZfddfT%ZeThDsmzZ# zD1f8@7l5SXSiU?QzF6Bu5LD5@kE1^IkE}{7Hs;9*bSja)O8_26fmC2x zM|W&8cN5ft5L9QIjDP`vnDU7Tjj|R&F(oi{1)SF!DDe|r{i>K40tX=WtVx$d4R}n8-YBI*R}cMsU(Ld2)_nq*L+iDM0H!+43GKeJo}iE#kBOV?KjPM1KcJfS<*N zidaSZ-G>@QpZu8cA&4}Ofl?Ys+kW6_1@vX2*A9gOX#{~@xOYSfJB!_kUONbE zm5-$p6Y)8ma zsfEk}V05=I4}R$NwLdjjr(T&+E^OQRB!@d?^7_Z!ik)w#go@4MGGz_i&+`64*y--B z#)FxU_)q?rv$U_TtCsiOdvbC>Hc-PAZ~h+B8&0GfeIoK9-)&x10!a-sy$F&11SlNh zU4;meK;Ka%7z6;Ucu?1ePS@Cu-ddAcMhsX#p#ac9T~M2ETx) zM1a&pkQ7v6F91TqW*r9%he@|V)u#zafY#ZK7}bH&I;ywHC@(?FFC4uhV#>-i&X&tT z8z3(afF3oSBFvml0{MlYgmBuVj*lP{OiD7f2P{uxm~8TwNslsYvRR;+4Q>s$;}|xK zfDq*GAzD8QP#G5keM+b~JwWlF4E#Yoj37j~DCkN}$HuE7^eOe(dX6aelhHZA_IO>O z=uk(9h*|cDKub{f1z(Cl6e!XYMLPU14onFG4}jU7Kx7FSp+~<2=0Nz}dEk1YGNVsK z7GR;^@$R4ocJJOjZfQTz4$|d^V8J(H@X`U))NjHbcF6Ek%e~T@MHlfB6c9NisA;F#1{kJk4`AFQA6D zwuLk`9#*#=`9nY-nhpsbh`XYG`0(xsvxjF4@3S_mBj~Fo7WqEw#5$-NH*dP_fnz{g z%C{VjB#w>c5!q$sLFq1hteEh1;Cqkpr|16tH#H(qMuGI+Nm!Gq2~pD(0s@Hueohj1 zmU-~2Px;Zpo}z4q+pmuUOom~Y1$Mh&W64aa3pB;4X%GHP(DE9HZ$!ndwfX+-RfHS? z-SZ1w-cyJwzBEHY3zm^6>3b{w8YG(b%~R?i9VZn5evUa=fzLkM9T5W|BWaOF5~(h7iXp&M-W!YAzSK{kb*%Y$_?3mhwIiMcND&*($%UzzvoZGm_B4(gfr zN#i__qX-=2`z#y|HRdhl(5lTz^|rYS=us}n%GH=qAM)>~iy$*inj1AH!x~64BV%Vq z)ws)7xPXYMul_Sk^>@#aa0R6qzDspmfQ;Mtz#!n~pmp(?jStRPtyixk_;NNr0FMw% z)Cz9CHlX;tnQj6*r!(903P&vw@d(Sq-@Y4O`}IFVsDAtT`)pso+XWpR*$D9c&R`dI zdKd4Kk_zbGM1ikSW6IOpkJZr%06E7@R!oqwS7>HM?+5iq!Q;8N$N#gVzx}~~`+J}M z`46oGyY2nf<3n+!@W!-p`IWyu2YR%AU3Dk#`Ov=sZ4cgx&m}D}{0*qwW5vY4(99{( zzx~8N`;XIxm;ST#&`5%UQC-UrIu^k>w^15{t`cB!=NiS$tjHpdi4+)LuO{l3K&y$w z7m(jnV0kG}j)UnXKA1q_pM8G-mK*9$$Dvh%YA^)AGjL&}l)R1L==Wgy7NG5P#QvfH z>YF78Tw(nbG0dAl%Anf5W;h9lQzdqI*W8g#4AcjB_s?|Se)o1iY@5eQgUR!Ey>1hSoKR){CkAoXA?&7FE$UOtmHmG>`bxf7&>01}7*+)F69 z=@_p58Ba+|_FN<(wxAYjfCk}TX2hDuP*`U`8m%6~72ed#YmE?nZzOw9{#niOx98+i z-B0se+F;PWq`S`LL$kqFODGdfiZ~@PNF^d@@^r!a3L-`^?dhWk2D8$@to=Nvpc$AF zVG6-r(V91H1+GYHbSPaQ0uaO9nDr1=$0*of;&EeYPR$fPoejTLv+h_UP-yboW z=5@H7Mc+a#8$}nm!hnu*vJnn5b7~w?8|!zjDX18i7YV1(GRp=ofVf!0@TLsi*gO1^ zu=5e;qFG;~RFE#qX0wUUvVji<0CyXhkJla@AR1r;^wh6@Bq3^qB`s}<`P&Ejv;X+B z5u4Yxu^WX1HRc?<RLjQX8#OA``yzziof%E z=w7*9|| zeU%IeLzbY2J5K9&L!kfzmb!yG6O}XH9v%Ka-S>BHsIs~m0=M*vrs2=(E2;oRR3GGM zQ5Qry4`;QWDfbNPw9X{`iws2+)QGq@NeTeRn2wC76YvV{F~=dw5Hn6ilgw#|><4$6 ztmGFr`3EvBzgs6TBE{crD4LFJN@IkKi}oTXt@WxOksS-=-I#If2l1ojf`ounT#<<` zV2?aAP!!3uU!uiqI%Ka43U3}LhyZ`vMdk?c7)4O`63OO|RTup;^8fE1N@I~SuU@PT zQd6xLU_iL)QKr=FaR|rx!&xVGeN-P48&tlwqH0{pQyej`DWehNG`!J;_F7%}~pkwTlTRkj}yw{{%KrQ5sc#(&c4eRW^yK6oN#WKYv9>R!lS(#wD@l1a8Bz0m zy6CO;_Tu7IpLh>DJ3)f`h360o9LheEIWg7t+W#Pqu!}vy_=GPc!aOTXTA9X-XZdO8 z8EHqv%rN_ZT6iTlaBnj8s=!^AoLPeIR`-BP5b32PT{0eZU!P8*%p}>IJsTD3k=M}D zvMPr4!n!+Z#EV|slzt+0G>|(T$TX6gnEG_;L4mq=Y5(m1t};1FTrra6`=Nn)%!BFfb?k;1*_Hhfsgg`5{Fp-YJi3mp0Z8VHwvN4f_Us92PFZv*3q@n6IMg>dcj7C79kFXj$27%R#Qx6T@Z2+|S98vW@b{MJvv9J6&m|5bN2& zI4YH5YWLaFr4^y2;|bFxU9%b5Cng@|QnD`F z-N!~cu)Mo$?m$V*JL$;xuP_m>_v@ZfjncPNo1M{($t0_ELXEjzSp#W1oW;^Gdb4CR%1hyCajO7p9t=k+|6_fj2K}IFXjHSR3~%k;>Ax>!bL zL{zxiwYpk+&_3gcs7i>Hx_zWhg6_5L8ILxSE3kTMYJB`DO6)y&Y$puHu4f><4g@lw z&GLYJl&06Tx(1j8F>#x5ErIvx?Hv!9{rH1$i&pOhUWjT^J#nLj8=rQ zwDfH(>uN#4!x3g;0>S;xUngf4G?*RmtJbCVBhXrgnvJhLHa*Q7b1$6G3XS$c3P#s| z**Q1tD#D#Pc?eghdy2xH>LGWPp3sTfG_b1WwM|VSxi>uZybTBcxBWOZAlgv)!KH-6 z{XF@9-MGK_%@2C~^QTNAG`N51(xoEv*FYm7NIGVby`7y# z#7M;5yAAa16N(|Pw(&0R|FayUX+ zK_PPfByYQ;fs_0~M&In})vH1Mva+%w)k%}&jJAN*UEDA;>`J}MUgH{=BAHAJ{oikL zL~hcLkYhTm&_%7i{Wp$06|Zm>GEjh;AvShF+Ta^lc0Y6ZFk~XCI*A{5);iY&4XvW?o{=~vl_k!vbqT{pp0Z* zv|IPJc-HXIi8ANkU-Ca$x%WF#-`*Z6%|*GA+$N-u@hq5UTcwpDGZC6|GPx~_$rI?z zxgQhKNYM3qYaw(u)CjG7Ds^5{5B_Lx{`f%#+up60y}f3HvYcEPxe2RQ9f~kZ)(gT- z5EmD}Wo4Mn24cQxOhG4v+7h|PWt*)J*+reX={shuJMPI$kk_|VPnjGs(^X#i=vPUg zKYrS~?@YwaJIaR-AI7tQ;cG-tTNW=~EL@$bw~d_j{js0pb1TP~3OXE@+%Vodx>k#? z2X8C@6sawl_07#IC8_N5*lrIS6g3v_39Pu{qSjrQ>VVPEww*qTY%80KcPm)XDryVmE0rx{{+rXCC^{s(eFR$H>nBZYTM8tBhu@^cAUg{ezN@}?6q^k)w3k4yOe{P7+) zM$Ty?UP3qhL=RL`D#zBYU3(u(VW^Z7I#bACFf=2mVezF0^YZg?t4d9rt7VvD;UjJN zl7a74JnMz& zAR_~xi1!VeF`0iQjtMY^N_%384JW~&i;IojU|34O+*-v{082PrjEdD*vvzIp|DF}{ zLN&H&37?n@Z2ck9s)(EFr#|SMhZ+e!yA$zAaR1cA#AZB^iL1#O)JN4#$qSRaZ)t5P zFf=rzp1o-!K6N$1EkoWzCa@zvxHPVo>b^Lzf2VHs{RsLkx_i3u|2Rf&n)AUEoRdF5 zJXJF@vmiF~5=1Ib`g?eK2KVQa?X;c*tvFn|%SF~RW3R+Rz^E@Yqku#obaizNdaEf> zAPk2dA-tEylta-v%^+)|I>)>lfumMSmwo`(!$){hFtt2&u2HTfrsCEycRglKptxmE zrmwF$!*KUxgn`gnT8%%oOult`C*Ap*jvVFF0t>X)?P9f5Djx2P32?9q{w%VPVt3+V zv_;ZL%&y((p*d65Qi6LE@_oZzTLda7dzvXJh0z?_+s~g~wb!h3u?2nb>q2J9NRHAZ zbG>J@x3kiuf|c_yzfB-K}wt9)15qm*GW!g%eH7gl5__w6xYluU?4wwwdU7Q`r{znU#@ny` zcZBx)3+Qk>nH&;T5)hP@mI{;hvFktn_@iiOmHODq;Lgs@cosPb%?Eyf23;5NLBV$=N!2lF-s}h`n6Rx>-KJT7;L}hy>s)(I%kodzs@FDlGUOIE(Jf5V==dwWoKS2s$+g)@AJc_Xm59bqV^f z45a6s`I(gS1!nb$@BB9T`HK08TJ|t{+6)DbhUG?=#am?i?)N{-f4W81 z>YUbHd3zj&ivib;#| zzQ`Gv`01&~#pt)AGdo$+1N1Nb%)frCuwqd?O&bv{lOywYOo;fSDJ1@>1;Zt%AHlAQ zzP^6l8o#d`931M{K*!*NCh?-MGUMa0dtxb#cH^)&Rn*=f%{aX=w<3NQ7nd^d<%4q7 zo@aKj+$4VfNns`5<2otnh>ET`M^R2!($)JB&4p|E_zd=p`95R1EvLP9==-tb0Jlu< zMY(C~boLQnQ5BV}&yV_wExeLE>(|qk@b*RNwXFRDN4Kf)XP@zs_jx-o@pe4x7~PWo zEZn`O#p7aA*7CfhrbuarxSHEpSyQ@V{?%Kxi>do}pWVB$D%*&Hjrhwi$HVhGj|boU z?k*Ut+j~*WmV-KHaKDr%x?-{YZEbfEb-WoG3b4zI5bKi3zmQN`&NHRqVr_`Z4$Yao zBu7MuFYKjz{+bHJzU%nR~U;jN`+8D>%H)TxOf;-$e;SLI#OP$y0%!Gu^AO{I}dwZ*o zopjC#o>53B&DW(O47^iZj`C^t8N@9=!Xbv-u#2wX&JD?S*Xpa;caR#GU%QRq&OK{( zPB^R{TYV$a_RQTVDDc0V<^*gvfs5+I&boNW`C+i_&w-R2#f zh-c?S54T0q&fsVMyal?%y~b;bZfs6KsRvejz0(X&>MR{k8Z=?-#|;&LP^|9PFO3bY=s`vp-BHReV` z;R?Kt!3A8(6g3r>nKxPoB^ECs+hc!5(eG+^?$A%YJ>~g+^$#m5*7#YpX_xt4H5UlD z;#+=gqPxW@Ah?`Hbw~>>AAQ9s{guo6$2|#4)g=M>^}@o!pni~^g~^kTDagpk0B$~* z?bUw{og!J#6BBNQ*{!XwUk=KmRuI)7zEpPp^(InvlZ7*PA}5(YZr;)8M+hdzEy{r^ zy@ty^Q)o?uKTkQW&bU!h^DJroPVb`K*}gWSYSIE5*Z3_9C=**#-n{mDYfJWy-Aa97 zqw-6F`=w8w@~Iz~@J%!q&`ub&+L@RpE3heBEOjFJZ;WNl{rPhRipZ8~Wv^+H1D?8Y zK^pmFh!_Rzjmm6y=lM6q9+3zjw|sx-OlNPQaUsb_nneZtDzky)#OWqKMLT^E0^}TL zkOwVWmn>OQdXR75`c>)uqs@U!8ViYr$t6FDK;PKfEd9kg{#u>JYqiTp^|$V|FEM!$ zmnH3_C7>PLToJf^rG=Nm-Cg-k>__KqJviZ!x*JkkXWtKi`{RyWx)L}47GN@|Ky@xs z!F`uH2XwWdUWqN}M!W$ofh7y5(;*B(%K{!L!&EJ7&&WSqU#s^#PN6hXp-FGuU403wqgz#P`?WR< z)jhqaV&zqmEcY-o_RYw*e|)?5-!U+l57P#Rhi{-keM_Yhlmmj=%Y-N7@vXxVwDts! z7DmV*#NK(xyW^zCKmsBFGq^MN0Fum8i;DJwt6L!4ioPz7NJR1^r0;F?N&B4HLQV{joT7@6Lkgd=&BU(AH{1( zsOhaK53!yM5Wl-+DD`%>&9(sXssyRFV0vi9lnnpIY%$FlpZD+Z$6pQD|MgbG42EB` z8ATD|z|q2#odo%)V-F7xBPc#ONb3b!gUs~X+w=z|C5o@w+JgIzb$zm6L2xq(iI~T_ z?qdanpg(ii$YO)$oQWeqvdz^C_$1Cu**CiNshG~cm%&ptrZ49;{;5>{UV|x>Y%XwA zLiM($;cd-jl}`t z09G;pouzKgA~F*>s6UD06<$P6P_iCbabZ?6HBq1tn3WtFqq5mm5=CN$yC)owDb8;h zMJaj_6uTpF=lT(%2~bCAct=NvIuj!W`dR>l zk*5cT6{rAO?X_{tWTd5^wslTEbQ?$r4eEv9qq6$2`RMcQnJPe^Lq;NuUx5)E&?`fr z7B&eR+I730blm5=H!0|8mD?mvKJD@4T@eSzm#U60?d#dZau1A2)z}(Wu)#0nYH}5x zU-Q83he|t}b|_u+=xVSqT9f@iS_P&rTde4T_g^~|vC};FuU7(|woU{W^i+T)Z*-$$ z%FD~$+S)op#A0MeK;iZ*Mn9~2wi#~7s`cvw0G|s}Vm&ecFsL8N49Q%V0OiQ6t*Z-? z?sER>Y(nY5lNVc&V1ICFk>}tqHtEK^1>3TE)wNZtDtlV?si^oI-KBZOm-plX)k~ID zFKNx4@YTS*&+4eVsjAu2;$eS4AU(h^SNcYJzO)MeqE^)bVh{iP>z?2|2N25XC=2LX|X6_8Hs_@g;`2yVY`cTh`WnURxIi6-1TX8=3i9ZC!hmlB~Sa@^<93OyNun{a0)8r1Ma80&nXwVSIY`4?uPWF0TFaU+AVkQHTw zu-}v8rB)j?vSc+-p}>U$4vCxF8Ux%nx%ipWHLj(> z!en4-AgJjo*wY~PmA8kqRP_u$8foYc0Q1p6)2J_%vNlWkp;eEmEjAs?Q z!x9}$)-wl0rtUMCnSR9%u?cZvaObk;v_h(Xa)kU&JPT>AI#VQ`V+7Ks>xGLKx1?u% zjwo!qZ7DD`=>A<80Lu&4tEe-&f^-yirHSvk2?Xlv6kvOx4jUdl(J z1uxYOMvS_>n9A0eMH_ztoZE2C{=|yz(Qg}#=7)10kyT4sR?+1v$59hp#=86b!R?J9 zGFDk>JN>M*v8%1Ohpd*jX&dIs99Ft`rYI}yWV`Hp-s~UW>V-Pb4_GyuIBU-5XSZddH>yoK31 zVI^yu1$zGTzCS!z?60!ezt?Hnl9Sl4ud%EuBHN%TtSmbbyLsmMWcIb8M6nQA)m&Mt zTGJYH_hSO$OO_diob3G%1N;x~t*JTx+Q3eLldK2LP&0@d9K#k?AKrfv+FT?S$z}uj z60Yj+Zb9>6J9;PcoOUiblRkb|^< zVGoJH^imt^nUvQneKsg@S+wR~){BWjjz-JW0xfHm_A35!aZRWK?S?c0LWSr>H6bXd zX2zyw6mHYlOnLjsB@f2~`a3sXweOe@je~z(5hZ#7j{I1;SV-umS9h}wcnh%mVdrgC zPkD3q`uMD|mh*c(jH-sg#0_#n zg7N07HmWCm{bMKen55MjcpI?)FJkBFf zyMiraP@RA>>X$2J>S%PvY%v5ln z133Anlv>N!BfiS4VC8G*`mmINAnF%5Em@Gev6K-ACV_sn!`ZU|nvx~bwgHt7&visd zlb_^2Mdi-HAPB>zpBk{Seb(KadDfRZ@#M8pKF{au3z*=AS57~xC?&ZEl=)63itTxB zm?mmnd$cm{{yGarPv(j{>Wn_IZA+L_BSwFC*Z1Ezl`$XEqE&z|@aumjPp4#5Mql8I3XFHI&z^hP44kBmAE?&R!XJoAxnm zDynrsK-rzA7$ax=CE0Frrj+|lQM`x#;qxT{YU||h0A^;T(Gj++?-M(5JLkrHFolHYWZ9<@LJ+M)dJy>Z}BhTTM<@Lpv!ERK?@w)*6LXLLKdfz2J< zM`fHGxxPR4rdW(0y<^13Rbl$)Z-HIl;=`HUAbDn`nN2BoZbYOSbMqL%{WMFADae9* zh?gPqJMUK@&}aPqs&(tMjC$tHM+Zs_=CGkz=!z0%1u9M&1hrq;)~7hQe#GtIg$oyi zs}}I}pEEorKrP7I$(dPjfS{82y0qEcQMmgQMR36^(~xF+*7wc3U5_lgKGu?bAHU?3 zm){T<$o_8R_W{apmf})b3NKD86szi{`wY{&zgZx7D2V&55c^AnR&jdgt0?G|uyAve zGvpxPh!MN%XAhI9&J=>ZYAJSiM!-_GDNrFutO!B`Tj=iSovpN2;7s?78dE2XD;Z)J zNcn$k4Y@jFYwAIdS0q5q%fq&yg)OTL!kxET!-p&_JEf)_YnK% zM!R%l-u7KuJ+Tzo1U-wZqLkXR$WY~=yr1Sa!?ZL@Cw1oZ`ByVfOB2mQrq}=G?SB8; zAC8cF&9@A5Wdaunx_zlWi6-A35cO{`GO0@Hoc&`ZBi@p>95j-mTt8}ppgIe!cyf>~ z3u2=WTH1#xOOe8vI8WRJ`U$hy=$qOJ+C(u{Ao16MBo6EK$Ws<1wQ}t?1xE zN^T%l4(adPx#e2uFA>DZ9(u4P-CYo(qc-iBv3);tgKS;~0$eDgSyM6RlX4LhS zH)z8vD|spCKE>4jWfh-{$?e`z6yupMqsDvV&1;Q!TRW*Id>$okn3iIhm9p8drax{0exv@DhpOWfx?O7^J~t+XPQ zoYzY!tB_dKqP_{w=QjQN`}?4@R8dAeyY7n3dPx9N>6JmjiWIzEMw9~4K~k|uviN3h zIi||5tqPSyl(Pj^&<~D&)tWULL0t3#SD**|jEll9`;0U+G}FwL$eE~j9}!!woU$F02Nl{!sth)%m+{LLx&{cl}(;NU7}C{&lPEdAw*a4YV& zAKEP>GK$BZ;|+iO%^T9nraGreWpU*FZ`{~YNkPYQ&jF#CBSCMkoI?0cf*4B{Sp+b3 zMXHoznwpwG!`W^Q?LKt|8(bDwud(15EhCHA=S&4ok1}VIdlcF@^ zLEf0fRR?TLEHAwAmJ}TZpK}f(19V__w(v?KLXWgz3Wy(Ee`FohO#T68) z+!qJ#=MB$yy$B?|wtMN%J54E@0#TQiD;;GYDt>7k(BsfhI4=-r;4$XwcJ|A_=1<02 zSg#5Kk$0upRhnz_tV-IstxpF--j3UAYi=ks$^oIyo$;5R< zyVG+-JBR{Cmm74YdK5?8@z5TVd70kgwq$e}%90n(QRgP7p~?~^#!WSwv*RsI2!xfVuqW;!xZI`FEfq>3!qjLYWB~jjxX)99@6wgC$djlST`btI8UvOstDRC4%#3`1 zqP@F+|FXtWeH9fMTGH9TlCjYVuezBQH$iuw(`GW=FDCt+$F{aDcRtM&4W@OM)C`6h zu9J3p)@#93ZwsEH_ZZhtMYRf955?@{uA8-vQ0|Ew4G5o!l-pfZM>xIRA5I?}czb|0!=}AbqFF0@cbVjzo-wJBah1qpLTIn9!esD;9oJ+q zdPK4Mrc-HF<=SJ;HD{tmX2PP8ENW{8dv$*5gRnNg#J*i=cutR{0b#AX5J5@t;d`kArKktDBy(#8+Ofao}puLj*=0ooqY1fC|hv&S;Y)58)@{Vq; zp)uWKEQSZNyfo6oYYeq!O)C2P>Cf&?uCCbu8vUhQk54vve3F{Q%=7iF&6%`b=A`Z{ zU=NdSS8aaQo?Ok*<`xfYXMu$MySvi=a=V0&lR{*q6LI5&6pVYLK8AS~%ZL#M#)=I< zv(8+X1ul*uhgcm{8Y;)sN1h0A92L+qRXGM2Yj~PLCNq;elgZnv%$(2Pb`;5yyV&cZGt{1;KI={_|Z*__8Wniea5DoElNc7pWgjzjaV0`k)?D zt|Yo5e8^5|w}5p3s3+z=f=3mt&-r+GCmQV(cZl;*>KwUkku(&u>tb|4Fhf6SVn==H zwW&R~^aB=6$Y)8$%m^l>>u)m8-ShBd^(M6nPlX&s+mZ;Hf{%w{VC|p*$HwK1f>0Ob zv|`LI)AYl`8!=1t4H8F8lS*LiXqV||Od)h|j=^Mie{%qfxelhIG8#{4-apP=* zdL!Swbc>$Uh#XN%pPJuC9l!ZEv32uT3H)`XU%pxT6zU7X!eC98Lj!+3@ibVLPO$aL zL3*abDPk_#=4z3Wr9QuWI~q!1g_4|8n2Wr{myqf(_52lfvwyo8OJkntfLyHStb?a< zHoM@_++Ld4nC+O3d(Q%v|bF3EcWT3TEOp{X~dYf~z6|`$9+> zbVAH@#uNexsoi!7B6fTguQm)(67Ng2TTnjxRRQ5Q3(foOg9W_R)+1JiAX}AVRCuN`2X?u`Q|P?}H9lC$9h2mK z0Uoq=2o;OzHdWLWB^JXXV1*mavvWDg6@u7+nx~7Xvqd~71{V{FaxCYy5NCsJ05F8@ zwzv(!z$Brn;(ku??gby;vuv^JhAN}%gVo1m z{SHLakK!1Q`P#{tOAAqsZ&G7=dC9wrnde4Sfd62!e(l=T$`>uoOZJ4w>Q=+aRj*%7 z3zze;R2Ug&b7M+oZv>{)G|O=!)ki+~aM}J(RaN%wvufPF?Q+KTfb0(?L_;&aRE;+} z5|WaANpG2N+P_6bA6;i$zw+RX2J^#(eh)4jSaS2`u5gcGp|Ae6c6peT$al2ai3mk0 z>r(oZLQl}HVD*|x3vW*FYN}k5Jtxw0CQF$X6XI#dq5E{!3ld)Zm zI%+e+Q$er)$y5A^3RXHzyas;;Ugo5ZGZri1XsrIs@FTOX^`XLDnZtn#gHxWaO^JAO#Hctn z_g~-FeQ`bi`0~ZaugmcFi?(&8{s%uvQ0=+?^?8r#dl{>j*U;b|@%2N{3L3q>dwojC zs&@K#u^R82|NVa&!Hz}PVxM$$Wsz<`D+U#>O0=XhN+yk}cDz}S{sN8n899r)vqZvj!Lnn6t7=E5YQan@-;-QF8)r6tX_fd$?z-2cK`8pph z>gb+ntMi8{oI&BQ{O2DnS6$usQ@pQPm&Jj*KTByJGo|)UXht71X6olFDn;#{Vh(kO zcf03vlU6h77T(lHL!k<#Qc^lJ=Cp8rrd#KnA44#vcTf1xr23DF9azvuFCTf;Wtbc* zI`2(&r#w24@`Zz_cxUXkPp_tIk-Hso>Ve&5vK127r1+Pe*j<|RALstINlxa)yCmn{ zcWG`)l#@m;L;uv|i5~xH2>m?ML~GF=Euv_M%(3*YmzKe~OvmI5DCUJ&w%7uw^XLhX z6{~&(Q4}%@R#WQpJuq=(dlC;v*%_9D!bDW19{@TDv1;MzaUVT;%Zh;kLdgK`LeeQ2 z=N`TO1R}QwFpDNjfssTzYc^h4wU8yjZC)Ak?GTif|@0V6u-6`xz@$$ ze!C;SLGFO3LD|Ni-!*B23@v5g=PKm-@7<-`eoXGVNvEce_|I3#@gox%o%}|Nskc+jFkInu1 z$tRF9D;tZ1UW-w*QwDNz>YFPoO`(ty+h>U8dk?_hGM;DdhQEB@`-Ex>^gJQ|uac8L3 z?@Ri>KEkiR(O85jwSZ=D7PVg$I42=CPeyM#yj#k>d;6dj1L19D7j=lSpqU#QFg}Bj z^vRHP6i5`k-KFWg|7!@L5n01N zc|f)K3Lj0r!R&!fclvtP995~-_y;e{<>2i(8ZvOOHt*eQ46OPAo+CjXq^*(&O!~7e zXZZ#O>I2qAgKyl;UJ@|yq=6?GD}xmFx|NtNStoY{mhe+`Q{=skkm$Z?8I=lM6(W>R zWW*EXF*k?^JYPT^%^+g=LK0?^{^crXA`>sa?z1Z5sOP^VPG9|6bfXqdOSj+DcVcZ0 z5!=QzR81=fWcv0Lh>y*6PgaJ<#SAm~9sCY=PF1Z=(%&`~t}N7Z$E5V8q8PkxaGS`` zvz>XHDntgF&BAoe7a!M=)>EvEwfpp{ov^&#rpnlH;qRXx^fTM|r)~sKA-V~IMyc?L z|B^1kq;bDCzLlXuKGQr&bWm^;(Cpt8#q=n_?qEj9n`Cc7ty(RGx<)Z{dd?c^^KH(D zrj{a5LhC5-7l|lgjr{kSeUDoH<#D(Z4f$_g*=OcG0%7QJSuC=82rlhmw14;M@&;VY z&!w|#Pz(c}B=oVgC(vm|ld(hIqh*ex_+xcsMhW`$Y*2i$E>@a32%RE0qD{cqUF>S( zt{^6`a;nc3hXODZd}g4KPlkn)VJ9D~U%7T|D$0&}WMhfZMeF2c>e9*<1f19c%b!iFz`meNpbcZ6hkm-#pek!&Y|{oClNm3tdWima@6;^CHPMXf1DR3E zzl}d=T9jV|D%%v%CBmF4&Sp4+D#txhWUNSt#K{!M23DOV~o#YLLplgcL zzELC!)ygQF6ZQQB`H^Dg4KkA)#hb@<>$VqkoP#RtAuaC4jQueRvB-0v&Ygh1Q?^19;-f;`Cvn=Z$nhobbys6s{=Ieg{-Cru)qhiMpZDE~ zG);bHo$~Bhv5QB-#?0in;I{5IktvpNQj*)OvD6jQV~1%vZQXN+GZU76#CPPX)n0A< zGuIVs#_iVjioYG*d&DE?fXo*G-DiUi45(k-_H-3{|AzPX;19pDzhSez$b=L;=A=-x zNr5mfdG@`u`C!R$CobsS(=Wo7y*z{8ktvi$?au^>Iuf)^I%H~2t-1CArAQDrj-%|9551Y|Ff~0gJ_&1vRC4m_*NGQ-;j!&mk7F&8LYM;E5JfX7gY3s zZ!wCVc(y6<*=t9G=wBn_OMc%z*}=bjLz&8EwZ6VS5)xe0|G;q>30^bx z@edT1(8qXF&+7)G>kAB5Ly*mgu+KP(Ft{PjP)ZfzK|k!&_=(Q6g^C|hO(neFhP?c< zZHZ)Dp$EuyWT2JHK%FW)=U3IDnq;B}riL&uylm-U;eyZBy#&{YsKk(IC6IXSw)jGB zfvT!1k;x_#%i!QYIeQYmABxtRWSRihIu_!(y6~xsUd<#lLi_968L7QL@Bf0AOAD=q z)q-4A-;}Xu;$u{4$jBoogAac^Av^^9z45?OMg9hd>l7W(d!Y2QyvjK(_nq0_P)S7h z@s2Bl(oPkE2l_N`lsi7V86BS5_X-%1qp^f?nZu6!sX0vs9YoGc8(_b zrQk|(m>#Ib5$_5=J8#`$uV-Cd{3Ie%{}0ye{Wp%Rfw}sBgnb7*)_eaxqJ#z!4T_La zDrp!YBcfs4W=2REC1sD0LNqk2gpe(yQZ|v45;8+ZWn_eq&Hws%&iS45oaguae_rQ# zPMzew@9+2XeUIz9-tX%uv?E%_jvZ5Lgs%r#d*z+WzxWMB-!JBf5tE4fyOj``HPE~b zrqfatz3(3<#wpE(mli)`S$TOXZcty%BR^DMrcMVUx&bycV7uybEVxm%!!Dok62=`C zSQbKTIbdOCqdF=gh75q?_X#4Et`X;L@3CyC8cNhLJj-M`z7>3{&GlsUu*!WYxntN9 z7}9;}^l9RCMS+P8@r;3OfnNBTP^8HpSX1{rZ|TVgi&fb?O5<@-t;08>_19yZN(*Sm zQB5_|jVz7BzME7q&)DXbB_%8+Wq@4%OlnztDYm&H{tBrYbKKz#EPn)(6$|X=&Bd7C z_DY5j1l+x+jz^O>4<{4>UcFapI3#4l9(H@v>z#&&D2kC7G(ai91GwL>cmd5t3eFM| zY=&HjC5JEBa4Wv?)@}2p^6;J`);ZX6^X1oLlJW6=$x2v)$GEgj5$iKzkUGV9T4voY z)7NwpLyPSE;!UCVxwez^P^Ll5bq-(1S)p|7>KbD`&v+n#K}sX0oYHxG#>1@p#!PEj zIjeI3+#{P(i?__}YspwRt$KIp#*Arc{QAQy>y`C!u(>kio#HT-!h{^Q@W z0@xi+04i7(EuU^i`MzR0eDS*?M6iRBSX!bSA4sR6>fx1ZZ??V*ge&>b;S_`wGw@qE^Rr#oTW$!PX>)|e=jR5gGrwW_dA9JEVJuJ zeS~Fij&DVeN9>P5!9@CXvR_dyWWzSdmptpCJmTix+IpH1mT$Qq1i+su}QIns|of?*1h{jn+LC7V~zepkM`k zbGUR-GEHF6&@Qcx5C*C4`;A``ejU)boag-WFVH245r(w3keB*B0f9t%4%-t#6H?u)Zp+e z6$6^u_=3w(5(bpy@d6HURjLJP3RiL(k%G7#l4r_^N<|V_6CBO3pSn6^mMo=4iqt_~ zLv?vtb(t^`Scc5CuFO`u+G0pWW$TE2kbR$@C1Mdj*WG-Ch$K`y$b83=XS;E3+iKya zhwvwNR|k(dYTF&WdIATge-EY->lF0nA!lG8bO8D_vb2`VW76dPs7fDFU|(a#L-a2j zo_@e_62chUDVJ9)Sk8(=Y=LKiyV4=Y09H5G3K!4 z#R)A|CB<1z1+WkRXjC~ z+HM?bFZ~zuBE$KG%-jb_d#j6CA=rl(RtD`9L+G8q>nd$+&ePe>NjOt64;90>%f zl|?x?d5gYUSbQxXqL-?@eQs!B{`}it-h&X^*T5Gp3FGCyQZtw&_(b4N&_nRgKJb0& zGBH|(%|S9$J$_P>@BzEG^|-e4q|Y2B2^m(u&^-r*KKtE#j3e}dh^xktA&c*>^6p*@ zw5kEo3%OJ-WD2?|n6qv_@w6=i~b)v%V z0r|&?z5Sz%!-=A1PE|K-$lSbX42mUb}4yb=i8cw?Bp0 z>EPjm;AxG>;HV>xy?<8dI5ArkY7dE5i{aK&aP6prJO$f=8-EJjjpI`d5fBtS4A}lf z9)lVw9ps5!d7dxt4|MAk;=qS?pKOKvDbX{4|5S&oG?-2u#IOkU(elHcj99Efd?ipT zh8vHKC#uez&##DUb+)C=in!WxMmaKlroYco%nH-~eu5qEnlT5>N>9 zx1z*~?;K23iEkZ=%po(XI7h&(>p$Gzy?Adnpa%pSDu_Dk)O~FZ5EO;eeZt7Eeo}7e z-Kvw;memm!47ujoFQX)r!i_Gls z6#_q@zE{R1CxbI-Sx6+`vPB+H%((bO-fbor+kvItF>-;_|7d7Zi0u{nus__40ojfr z*7OxE%w-V3ir{T16ZrJqP)}O5W@o}VxPfJYrXn25P&Jf~0wWjTXH>oYWos-7XCs_G zEmVl`Y!b$^fQ;Hl`W+G|urgjiQZi4$dZ;N=9!6K+#m9wK;L{Q49akqWl`38?`wNPJ z%QLkP=i`W#A+T%iN<$nQEK5&26XkMfU%R`V@-rA{VUMjCCiq!oa-@t&K$G!sSKh5Y zO->bY%tL7<7jhpSxOf=Jif^(&pOS8na~m^^2ttI{IfsR6aXEKMDQ@Ucuf<)p;su}r zQqH+Tyn61PlSCwG#K=NNIHf9h&W}Dk_hV$u!>UA#7Fz&=BDwS20fJvh4&~Rrm_4Zq z-$;m@W1$!J6Ptjlw+?IU#(|%u;-ZM&sE!=2=^yXIYzXN#<>yO|0fHb)vfsXad%OED zGNm|&!HGj90Cs>M=X!D%t04Ol^S-MsQaQmF4M8v*>B{lEv~c|8=$C`;vQ#IzLytFm ze8?CqpFK0Mj~4`Ris8qY=65UW1$)#6FP61jtch`J|F4JF>!}%fh+k9g)K$iiE)u>M zebBE56H1DThwlPku6Qk8ym;Q@-{TCKJ3)h7hE~TOI3DT3GxJpL?{WEc3LDl_UYD=) zjlrQwhJ^u6|NQxnyu7>uNCX%uZ-AFrI%%1KEuxtTCn9IHTeUwZr1bW>cPqT_{yFE%{53H)AFQd*h zi1lE)1fKM**bwc9P4Ig)dnMfL-vF5usKRv>yCdqgnGfzHJZon^y z-E}?uxJh#cWb38ac>v=CGbcMdRqZ^E!(*xr7|MyKY1Ofcj6?`T+=R~cKf%^xlzPX} zZ&E@<42k1xyL)D_(A3lvait{wFN3W)yHDj5b)g-ogE`E(!IM5fi3HFDK~KqKY4ZmU zhIzYW1Ruf?W0|qJ3vvK(!xvaxnudx;KV(#ifC%XEn(zK?Zq;$Q9f!(|yDX1x*h#eY z>cb1eJA*Xu2H%R`UZT;GA9>-aEvjg4TVxloms`!=ekcR^!jQOTP|-KXt`?t98U zaeq$9i;#Dy5T^*|IG;$2kPrU0Li~o^O#Xl6W1k+Eb08(y#;mZqi zHEnDf=aWu{BP4OwA`6*WIN6Z^VWg}DaCFf>6vY$FX(Gq70kD!zAKaCOuCA{Cj1fI5 zvP5inTPj6}dK}wih-3LG4j(;QPkIeRTuVfK>LTr^d-om@ds&>a?dol1#XFgZvkWmj z_@Z?(b8D2d9O!=PDv%TUCigo-oR% z4zq#Zc;>7?bCMGq{I3(sFD-3^vaA8l7dryCj-7F9d2V?RhcF(9z8ukZ5cF3F1&2ld z5SV7bLq&@fh@_;XMSlo|z{7x}T0#Op#Y3S$&OR1Ftot>93`E>G@$o<5eI|mmCjM;< z&C>wy9$xx=zs-I0C}DpNo_HCD)z^DK8WCEo;S~dhA%aK&ix~O2*$l13pWXpLJbD2b zHnQm$m>dR|S&L|jPW2)&zyiM&@;yia_99|--A8}`HXHKwVvj6|MDSlpP}spk_a1WW zjB+lwBS=zeC~(OePlY_`;E4rn4sczVo5-^5lq38Gc~j{*-RtodDqpxS2D`=o%U1Id zXW~;ld|3UBIOovF{R#*Cb67n6UgPy6UD-#vSlNcE9N60DEfR*r69lZTdVY@zvBI0uebOs0Z9lpHVhAt z@(tGWCL8|s=N%<%u+qR>6alPMf`!mSbtcaosjUMgLrj&4HH?sefWXK!;5saWzb6Jj z|71g9FRG549J}N^iWUY+6D{#fgY%9K3k$RR=(v9)@s1_7kK}omz@}Bbz|W9*kS85> z=HDaiRG&eCRq?}qsn`P;Z$V1BTe?9d1tFg_dWj{RaZ*6YEN(XKdzI(SS!1 zp`%9iOH~LH40!~Zi>sbRTWSwsp56(D7jC%z<5)hoObYyuCj!?{#3;NT2|PvoGJx3K zZR|hLzY8d=D-;lDr=Qz>T)~He2KI>WzKt{)(*2>@5XgacfK-emU7zE``o;~H@C>+i z-#ZF@BeqFqNHoQkB>*N)ZeXx2}g^cSNLI9??+dgIF;wTc^{ zN-&?G4*FyJ6o5KH{H)!k8x5@iS%RL6=1r@};}ZPLf-Q)#5qUlIyn)XVZ|q#D_cUR5 z_aoFiErwfqQ7VgJ4>?Q0uvt8UjRneK;t`3=N>m+&$hlThc)EIemgP+4Tn`Lv#B^{HOh20p%S4HnBCMr8_WQ()68p)U>vkU^I0Z^@2WWqL zf!b+f=9vwi7sC=ZFHf|4E)?uN+Z2r_!Wf}G$ga%= z#0x%!oGp5B>fzM&=QW)gk*ECnf}DCkU0hkeeMl|x;z})5vF{xPe@=}52*79?Kp>v`O&kQ7E@ZD}=pttD1A(V2~#XEp;leIBJLcF#KKQB`vBq3rxkAfs9JW z`Q?rNj1+>hA_-@#CvpRjj~}&(slY+uNyF3|Mr0`P*n{iY{-N1VZwDR8A2nklTXvLq zl%s36vfSEKHR#xngT2bR5Ce@sRE*W0qTogyAV5r|@f8@T$j+APmk3e=R1sj_c$q0n5NRe;S+?>C1}-v zTeq|i9=v=EI2~N9(TrImOyl&0x=TMy3+%MVr9>_Dj1_LPPHaJ<#1$R|7;*ECjpaF! zcgoyc5Dfy7-Q74GaPstp8eH@TxEOr7eWcS zfRTa5pX)9ydIk;y1KiY4SRy{RBSiO*CjF(O6xON0I8^|YQjDQq%#}xA)p~2 zKflx`PloHdPz@8A12SSflJl}L8gEYc0FPm32Z=O}I+?AQhC4Gp@R=NlXIXyjLyfq2 z&(Xpj)<6kOE(Y1jfzq!vxk?=?#uMka6EANfv_KLjE>G_fFs3<2FUzC2*8=kOD$Zx`G*WQpao`O?gZf1s9w$xy9*OWt?1 z;1{Q6viN2`Oq@{kQ+2q@nWEViqxxQ)@-lz1e%W4gdy8_B?tcXqqQ}TyL|Ho>g9|<- zrQ(B-9*{5FLfCnXAs;QeDIiBVV7A#SaNM;ZT*4s*2DPRprPs@0{2p}Yj(V&ap+OPZ zd3p^|>LpB#_K-#g#K2ssUPfaoaQq+GJ+RVsl_cC__jG(lAWy-uKKT9?N(YE2VJ^NE z7EoP}N5N_~K(3sMK)5tb-cl4H7?`w3^k>q~A!I$zW>&cZow+<{4C+cOp)nxir)s09 zDvfa3`Z1%)XyF0JK3DVvSDLw-h?xc6PmAFaf#At^VpImjB3-o zywr$$CyVw&9Q5axz+7zU3PrySDmNKtIWpSG8Th97X@qCAuB+*IBf0-MGDY@OQYD70-Glk+@)!#eDsw$ytFpKxL{ zj|(f^KLB7P1{EmL$%TwBk0&6qG0b^op8?-V1k;@!?Q{*cfoR>3nkDfFM>t61u}5AN zwDU}KLn8%Yho0;+OG`^)dWQS5-G(l2%3T^W%R8b!!=LNsCYiRofj0@hg+v2qRDCoPR#I<~~GH(Aq2rk3- zUx)<-Yl1(Aduq6tYl_3>^dWv|sMmYS&6rc_`3*1ng5;Z{bkBBg``YilvQ5D8~|9Qnh%-$%whYFhs5pN?IA|@_ z!t2Z0p$XDJd$WB|BSwzy00tsRw^}8U2RazhSs1zQ(=onrNWv`9BAkgN(8#12i-7f_ z>v0nvpizJq!(Uga4ES%2QuFZekUhTNHUaZc*7LJdM4|Xa>ns^I zv_vJ__a$#a&p1fo9lV3AQ1t$*Ydx!lcS5{U$=QH1hMWz{mKSLt2P~cb1{qL!@bQ0ga`e)So~`w}i1XgJ0ge?F$E6q)dp@wiQ7!Y$Ud?a~Ii?k(r_+c$Y-Ee<_QJ+}%R zsw<@TBUt`>;r$euvY7Xp8&izAk93(aZFulsJ1cqxR107JSl;6$O46l%zH=r)qN-L-ibwA*x~nMPdr#7@p{8^g1et3inKG zOO*b4Zi;=je=tdsyAktV=xVZFz3mq_%d5HjZq-%y$Br3XwLzMpN8a;=>3Qc^}aIrk@h=ZDxT7 z5q_j({*vk%tlnkeHJ@1{y}HME`3kQkt*ZbTZvLwL{{2&s|# zh+=#~0%|V`Tn=l=3!sCUL(+6?6dWhx9;h&vj+OrXF{-$`(~szJgx;u(YkGKo6g%3L z6Tig=kLR>Qc6D8+oTd`P5zY(Y5(X!U#=6wUqbpagag*#jPJx{c(Zz{cZti!b?#4|j z|MsM`aLbcmtQ zd?oRHiTD&N=w;ww=o?q?5GDn@2y69O_*)W;9_0^7nNw);X~L}` z@6phh@8Df?umO}BNy%p-TOcSREeslW-{|^HClJ0-aaKv}Lq5Ru_Kjv%bM+zY{b~b< zB5L3YH4dWF5^NBzdoU@3Apn4YvEiqXz!umwEXRO}9)b2ik$iDbiO1Wq0M1O}&=)P0 zV2Cb>n>BEfPuL7IfSXI)Ey*AixT6-_T-1&AU(2rVISDIhy=QN5dKyqy64eHp?7zO$ zQh?dN77*N30kYthOoYSK5p(v%8)+Mp{O2tm{ys>fsT5JRTY9&`Dyrg56;7%^xU%Mp ztA?&HIVls&w#ZvPG;x33W)h1_f>_O)n*J~653f=7O`9?RC0n4Hq^<-jRQmOhc_Z*F z&fx)wlKD!X^bBn#3ogig)awCbw1%KB+!n0vxY-OABG5k(YCXI1KH+SzUOrn-~iMM61Km|0pp7 z`acCd2DimH?oLZ0+WrWKp9G48v|@wzA=w{miONrQnA{=c;BX9!JjRSvUVo4f+Hq1n z+kj)uOO3wq1Na&Yc?nVmt$&kYB850Rh{2=|K4wk0O93}-kf$EX_a(%{64Nd}RoJfr zg+M{Ag0TJ~PbE1tC@2nr#}e!?p-^!cf|0`!WAq1`v!3Psk9jF>s<{d!4C7Gpmdib* zdPW~eXrJn{y__uXldzI4OoGxOle4{WqsUH&=Wm&gUZ6TNJRN^N5VAP+qOmVkGqo&C z^%))2U4kq3C@%$Bt=3eXA9(!sg>=C!x7n4L<|(05#ay zYd_Yj4)7966(9w|LP9EtbSlcqYN&T8_+KskjDDX8`bhI<11 z{nZg`g|t%QfXNZ7J@dYyYV;I+Aa&G0bU<_hy_RtDWJDMM5fzM`;EI!;!>;z}Hy{#3 z%s>V&fa@cGyu<9ppyR`7=j%6aP~6?~|ByY$$HfgVP;dL@m`APxu>i)%22fZvL=95K z0zZM0HwhUN_xX8VAKiPgqvq!3KcF2WXlrsHX3Pj00&EB$M{v&`Ez}?CC>1K|8rM^? zw@*Ca3hb#H{f#CZTHX{;No2+Za1RV5|Cs8~O4d-s2XNF8(L|=oavCxM^kodvCo!sp zNVQ=-P6kVmvE(fyo+tWLYqrGoR-1pL~GUjEuTa zSI&=3Q>oYhoTsh=l3&@tCVri)ql?ZV^zhn!nI94LXu|U_XCya2-ic;Y^T?4({bwit z)B@ayIM%c2BXknjdQAx0dMlD@k zfxvRRzL{w#@gWyDivM+G1wGRR@o;4KlhNJq126fy_lzi$JQ?_cILoo&Gx6GnDm1IG zZfJOBj?fACNei|je0+)smkdV*ql3P2(o(4%!ZrT^*Y~;zaXT@zR_Jx}gg>XlhafqEU~5&)}~`# zrtK?G!5~l1TO>1G`*P?+@S3-WWkFE5#0+1G8sfZrPpXGEN*-^7e2sm)dzdfDgy-Oj zHPP=j^f9PXoWnEZ^rIL8M=8z`iL(nCLQ?NEJ1-QnSmp?QXSd!iQ6srsF)j6ncrk%1l#izJNSa7}GVfA(mhD+;YWR(B)0{`{@zqw9w1rbv(r%|ac6r6pri$TP=@CFdm4)8zmM%5E^P#f(R7k`Dd zj0lj25TZ>^o#MUt)u|NGh&%LsyDr>Ag)Pz{2A>J`GX?STR#S62xUqx@^Y z!#TkLmI52D&*eIW5}+Dnb9Q!iG$v|not*q&uyM_Q&sG_kO>jt5z@kxT2fP7=V$1G9 zGwfih!MCFTX5ixDvcQf8>}q=Jzb7#%i5?c4k z;2MojuCH=p>VW&?YEab&&z{`|*5ks=?{lv*%ogGj6PMu&`GFRSDMf6F6rP(yFu4ZQ zx)U{=va)h9%DFp{kvCzMiD*^w`Zew8)2Dat+vj)X%94nPh-!$*0r4z_qt*HEiW%0cK`mXhK9A^K?r%!+1J z0|QL(mmpUYPa!P8bFcXy-Ud%`=doU9Nh%Y^`6u)Bp(8)zkBj(0&GVf|-l!7V!5@9JaR=6+#Y209}z;cl*I+Hl*8c;l6Ex z5Wuf7cyRCD7Sp>gTL7~0C4IzpGJNA2@ydR2NlAJBd%?LflanK6yj!3(-k(fiz@jB`28W6kq|sf_+u7OOsnr&d<659H;tH&N0U<^wil6uf zP-rpH(NtKhrmsAH9=sOwnEM3ey{TDiH@iOg@k4@&nb}{7=gvv=A>=e;F4I3Kh}BT( z(|x32Qxu(Sj%Tj7aUoHYM=aBM#1!jFbQ&+%C*U=7v#i)i`PDBF9B}X68Wc=^xb2-N zrj$JvS*hpW%9xkL4XiEQikO<>#zgV9c}dN@E)|jfBRDB zGZDB~{Lz3WZ{MsjZ?K9_1kdI@I{r#ILUCOw!@Ua&^8z|Pby9qvE@}y|dPwi$(U3pJ zJnUw*+~fFM)uGdy=^|1N*Su8;;Mdj=OA(PT?bFcZ_%&+$Ut`n1{($W>8oOOs@qaK2 zqpKoz52K?uKQ_$!eRB`F0w~F6n;(L-+=MIn3aGl8fx%<#1)@gd;i18sCl>^X(+3@Q zG(rH|wCfrCs|GfKyGZtOD>aGwA$djXK<+QSsg} zHYbPO?|9A>RL0tIE<0gSf?DL@IIbpE#IN4S$!Y9y9?$LiywfMKco7W*U89?;YvJ#3 zDd0uD%~`3@F2vZop3wrH8XR`r9|W_o>w6^>tjybSYp4N2+M>k5;>JyCLc!)y!dtd1 z0ghA()DQdR*RJjT6we6F@O7MDOQENS4D+3tp?iM=NP_nKl7&>9D4btF$|WFUPFh=E zhpg2MENl|iK=`h)K86S)CnrZf@+rvg_q=%WWHt7DT)TDa4K^P6L0xKAm!yuVP#*=>ZNUY^rK*1$frkaLPF7BiMowNn6Er;nehmJI1Jh8w>`&RJXHipWg@uO~ zW_DxrV83k#70;F}?b)1A5JHb3N_0n~3O_$TR2EBonbvu)lAe?;*zk)1pHW^} znV#f#bp zuYMUl#rLkRV2O{dts6<+1lU*&WP!}X*X;tcqb9(5{xR0)`hkk+m<=BmN!!Q9K<6M6&kchozQl?BRwSX4v%uyg|QYbF7q|r}q;`X?jh`V=L$hfL^qaO-z1=NlT$Ks>Vcn`D}Z9*ZgtfC@RyHO>cWn!qk z9BG-Eg@tCxk|nSgUxyGUi|oB8BKsmsZiqFewcZ{!sw^v0REw00DF(1p1Ki-wnzyC-4Yi)F{xhrKA6u#QwpkPET6M`2^A*_OdLL(fykgic{Dk5&!bNaM# zBg==$cuGqhIKaGm_2yo$+Lk!ah50bB3=QpJ?kAMvH72L0W05nkGJy*RKB-@sO%6AF zCqy$CozrYo7do(Y>sD-3&>32tv+&>@wpibS4&Y%&dik7)TqCv#I1n*swu~AtM`>F zs9UzE1ft9eZ3#oJbH8+{v#X2l{tIw?9U^7-j-CHq>lmABs068#%P=oD~GJdFz1?H3wK<0@@TG7J7 z;(nASRPW@918v)8n9B-r`fHGcH7EcYZ|OZbNsYDW^KV=R2L^1rKdj3d4JreXf@S{P z8^tfuqHeSb41M3V-=U9{L}4J%CDU^6!Ek0F~lUC4j*1FVvy~H_TnV!0S@dg&(F`_9IUdZIq@+? zhz$$A1JI>%Z{L1&i=DK`EC-aUN=W5U3|~Qtv_54N8z3YstZJJEOVw+K+l+7-u^xIY z5@BFK-}+y5n;hl_wSEn}+rbz$3;UmtQG{Bz;?tSoy!8T$I^;&-uX-p`4$7Fe zJ~t=FKcc&SX_?%=KiCL^;!kj1&$bsDo>?{^X&50#9F4ikOoOZRa6jnUCd-*9A!*-U~iot^~ z7|Ms%err)xq>FB?`_J97y)h#YAuDHB> z83KYV!quA|?7}KR+!+S6$OF~Rx6xEpkx(;`qr<;{zX35y6RgrYDypEUXo8eBJUUwX z>`axI#LL1$Y3y8DiDvDU8FFtXIK7+Q&cSHeNK@zCkE{C5x-*~d<3pK1Qh%Wbc?+^o zr3$mg#aMpY*473+iOlm)qgtfZfRJW>5pC2M_+n6k*Zlo!To)kvyn6L&H0Uivi@e^9m_Fcn zZ?Jl57seV}3oo5`WM^FHz7DGrULhFLw*N%#ys=^5N~q2b%E@`Zd5Yw56|aE^Nxy1U zV^e-O5ivvGx*6+GzS#ly>_D8D1T~R~@qY}0SmY}y{{H@M4f!EBdGbK5v|YQqyLn)g zK`J()VDXWQ2`(HvkERE<48kamZpEvHT74H+Os3#Q(QG+%qZq@zwi7=MSMahOx3Lic zuSSgqgqDtuTTn2$)eauhcvB`Q#>1p9NuY{|yBzn`rLo}wq+Cj1$}=r$cu>V#$DdJA zVF=l#=Zy>hZR|&_{trE7!C;kvwt@{&QBm|L57GZ5&0!#G2dN^4M@C*DAS(>=DHs@R zT*}1Z1zwFriEpEciHT-lg~Fg;Z20}a-~M1^qZTp}^S%56gdEx-cCD+{pLNpowbUZ{ zCmv?;VFiquA#ODuCc%s0I@V23Vmg`whF!P4##ABXYU*eVk+LhJo*9YK(y~3?-ywJ;e z#?+KXMn(q3eJP?g%exjQzI=ePCy|#(dH{}Z&)Jg(7Uu*hs>tJlGRvM*$(qn802?|1 zm7->DbclEv^6c4e^ddg!#+q7M$jhPG_Xhob60LHQc3OD)x7B~^9q~(bx~;*=YNZMc zI@ar??W;@a*_XK2zgcaayP?x7bh*gN-DPRJ60-Mypj}pQgSqa@YU8Y3#q{i#+sy>gL0y2AF@-RtgP=%(s47Y{`;!!s1xydtoN86i5u}tC) z$G^Y$fBep}9g?S=P#nH4FTdU<4A?|kS{l{)jot#I9)!kT7#?`cjwlHPtEBy2$;c>= zN&E}H2lQ@!JW=FvX>w{RdCzI+A+j9@!w=cb)}n}bf~VruIFJn8Y}*(M7iD*$U zWf-ix4GQcY6W9n233~x#>}Up zp;GX|qpj`+)cZ)d*m6*xtFw+P20eIi9g&I!m4!<(dsL$@+EJ3uk!=)hY|_)eVSx_1 z1$Rs<@AsU~w>ZGcKnJYDm*kP$Tlh~>*kG{*E%vn#CB#1H(ya=*1L9HWAWIa-Dqdd< zLK&pc)6JwIb9Nt%?4AR2u||wsJ3cbTk1h>qdL8M@@BsGnmgWtBzU-|}WhX%rM1g)F z@QE$^_v`le3veS5GvK*C+m^2vH14YsUxISFCh_f}DiS_~qX8=KU71R>!vBv*!yhbOAFjj-}*4DzN;V(la`eSkGuZ1aaY}vBqphNJ8i+XP+ z5wQ73d>=4fH=BWQ&Aszx?C5)P8LiCuHu^T^x zd=fO8vZ<8{H&}(@C`~M^|fot zGjSS%?(XikeRYDgw6u$`+8wc2fKXUDz5``05WL6z`xO@AQE-tXf*P6zS5lcbc){BV z;8!USMpCL_V-&;m^mOj6G3|bU1*BES`-9cdN{t^aFM~prd3&2Ow~#s*N|;#83*b9H zB0Vg_`^mKbp%KZ&+rSIV5w~a^61{I;UfzE^!!;)qjtd<;x?51csiHcsQ2J2L`ja*% z6YPd$7=sj8eDH*v#yh3bV zlJ`|M9us)u{g!9)>e}efFQmqp-j(0jU6t8?nt9$`!#nnFfD_I7|4ilo_Fq_(fY+kx zzY`Ueo)ra%at;3xSv9rW4X%Z7rP02zQCuVS<%B0>H)fde)rW3XheVeO1reSqi>x6K3g}0CfR*Dns1J5w<|k67?Rgh6 z)FkSF5yNII=Vn3coYhhui+hm(mA{aZ*LrS0ttS%I%QI!=-lizQXUE=Z4; z@%ywf*I)z%k~q-V)>^eA(2yL7{a%e$#^>^!M|j8}U+ta&MuMzkRP_d)D!iIxXIU;p zXS>iXvlKi;_X|>b7v!(3tgP4Cj$gy${^7yCRk#qZK?o=gg+6$|h=f7Abg7m3cEQ(q z*=>M74|IQwWQ2RAYsM#%CJ=$dL3TNCDhi{irH6|#zQ^#)Iy|mVko`!^LeHk(f5O1v zMn>T@12r|Z0M-cUw_}ME(}ukRX~tF6c|E{#-hTYJ*1#A$6q#%D3+h}7F;&R2%YUav zGDXc$UtixxcUjUmfEC_-sF7c<__Y*eE*haV6ydSoK*v-=ZGU|~d@u`GSr`Q6Z~axU zD|#2eAmzlOc0fy8HgBc^8(28~2N@o8Kxd=ha;CDfvJA>; z)dOqsF;?M|r=8+SIeRtD>)N$m&?^?#pxNz&T$cnsG(7g4htI)9gX9XZlh;sCAaJhD zy`?aA2;$Gz$eP2r$tIXf+lk?E2i~mDEiGo4+$1Ov#+Dg(@7`^wiBFP;`R|dCL?v=!wBwS% zCPp5{3%my}zuwj*Y4>U0zdfP<@tatu!=|P@7`Y>iN!o;VGKD;rWU@8O0|k;UnMxN~ zZJ-~8vVt7Cf&7(V2UJ709R$xz8<^uZAwpDpcjgzKL*aw6o<9Adm?aCkC9$Y0&(EOr zfE8F#y4h?fRL9$($f)B2rOE-18sMu6gIj{ID;nF`Y`?!{P8-Vu5{Q{*A2g(CV>8Bh zK*|7yCf;jnwjXLcz{DxR3a&Qy)-Qc96~hw~jd}~DlN1n8?(b0nVZ0M8`b@?UlF&N9 zgALQ+W>Ab~+yA($?MCATyAbsHwkR0VCYE|Z`)-E*_1g98wTC@3=P-|e>XHULr~_wL zJZO1vQR^}DQ|}P;QXe%i>HC~ej3MM)-|MMI8W&WsY+s$d!wBL6L1@2wizhj_!{g(M5CHTRAgE!m=q6ME z{%!qQQ3u?7u~uR;IPa}<{;zF^I}T>`JrJ|x(YX8a!4dvi2ih!p!xOH=%j^x4l;^xbqPP)U zGB-g#a}~kHP^pF?CeRYx$%QLHa=m*~)C8Wta3p|{>LA~}cTbR(fUR2dtx-3Bl-N(D zpb+Oqp0~z~O*r{}aFhLt1U8ABYX94Wk(*SY1m`E?l9xDcCdDS{WN-mf8iw@#D~9^y zbagq8uld1+FlsvWo3#wtIiWZg%ZQL7NA<``oqOE9tS{RL-am>TKg?A+-H0Y7# z@NCr}fvPy|oI1xJtWtx04}^&#h$F5s^V+q(5c3);xorn+=zo{k9$$#SlR4DT+oX?$0S`#!&*lf&?CVIgaHQ2Hg+oz03W z9nEX_(hp2k_3TLHph<5K&DK}#{Iy`5b|g;X$l?EkLFu~l@L?ciGabOUB`UE${py^% z#GhJ#L%RCC$PF{Cu|OB0dAzBfSf({428;XFrqj zgXk$+=cqSB1dr3rZqN=0l?-ZaDLA|j&)VvZ8z1*`WITrg0@P!1OUsFGc4y4ZJAp|d zYd?O|BeE)9!(ZI4TQ)`wt+~>3X~N2&6}%l9x>bJ_H|*2`NVpTb5{A98SX@L;7B)$! zB*To#m)Oy4dbuo?-nZ=(|AWocqFPJIT6Z4C|DuX*F2y?C*f z5*i*Zbn))5&!{7z4&IW~0OSYlKeKqv13Ef76BP3jmBT{d%F6KB;9cqptEWDRTS-@K z5}`w86@T|&`IkeNivlaJ&4C2|et07m1lQ%_q(6cHD)uyY%7-y~PNYVm5P*lV(58eP zaNCP{Qqm_y9B^BU@F`Jge)1$0%zK|h*NcyM6^3ocMk(5;&12n_{VP6Vlnad)X~z-D zNEu81V$s5U-p5s8^;mv`Hz|+XN=RRhb?bhJtttoyF#*=724mSkEFg9Mid8WCoqlcv zucB_!W^hj>-l961x*6_OsF+_>R<3Y$ZRJevMb)qqC-_LD1e@pl)K)x_wm)0NUqf{w z78(`BMFbC6;s>0O8l7&vCD%WmS)QZ!4)Zi9rm;#(h%JgO={AoJH=mBr z_l;fc#$6RfMO$@vj)PiJV3T#^wv|q%?YXByHCw=4w>Px7UVo|120?=~a|zK!eLi6+Y13M0fd;D<;xk4+I?Sy_89{&^eEHL?df*l*~(@RPg@M06zI z6JQSvAYyvT0qx#d)bb*=y4Kd#_I-8BF)4i&+5|K?KpB_bKSJ6@sY}x&(`9?iNf0O% z{Zp13At5mRS&ZJIB51@D&?+OS%2#-f=<0=88Gy%|0erdrZvK+dT8FrV1XF-#-u-x& zgd$On|8#Mj?~BXtf&2OsQ;axE3aC=i$IM=A5#6=xa%Sc(^6o=Jm*p7TB2?|(y=x&q zBiK6*ecey2coZeP4brWa+dw~_0237O=FQ#I#h*W~BPRk9(P0dg5WxgEMYUuKqzR=2 zr(T%9q@bikE{8G(g{?ghWQvPr0f>t-;B2HkW_1TM(Q*wB z56hw>=Jt2v9P_&dp7Y1(k%4U}k4u7ObL`@Bwg(H#$uw zCZ-B8BSV1qWCWyR8XC1|U^YpHd3z!9rtw$s#snp~M*#oI@Q~DfB=;TJ$`thmY2-1e zw9O&{M;L#T7H^=bxw)jSjunUkG!rGzv|kwQJ_^7+HF7K1Jp%m#lBnZ4p^ZT1-8VKi z*3sE{8IMW~3eP7mU;5p@zb;6H584nYIVOKKd*F4oB$Xi6A$17~hF8_qYJ($eHPCEc z!STy_g9kGPe5~z-3zx~+8XPpXSs=GXfOW^qIbc?sMMPF&)#(PCw)|wrLdfgDVV1mq zuR54H=1R&13>2+KWbGF~`$A};lYph`3(+*-2(tccD@d&`%wPdy;pIh@HgO4Y6*Mjh zHQ4l4zaB{39&)nDV}V!4sVE4oase-o@oPK}N{G)OM163o5rRp9Gk61a!y)YbQ_CCV0189072wH}Ctfiz+-UP|L*|2;U{NJNG)2fa@-b>d z-_xPUK~xvT;KEA4Su~( zkjQ!XYqmMu`ht^vJuJ>1b(VJ@q^&U&A>r?tZg{M4J{Z!8*)0JxOk?-8U;39@}^>B){I zAEFLyz@{|vV=;yi(BGKx4KfSdrPXOgh-WywSII?3f+N?<1iih!Ol&}LbMt!O6>pF* z)j@;cZxNXqPA-b+HTIi}9^>)p`2KzIuVTs)kD_ybo2@H5dMZq*vm%t2f!re%5P(($ zdlGAK$FiMQDzi9Kb=-#JF-@7oMqY^LY z$bEa84v$`XN$FWWu5N)lygHpGw^#%(vIW%7cp51L>pwN{IpwnJSnJG8yXf-j=v>zH z1`V;7cVc{QGL3KJyRE{TyF%q}B_G?H9sXr+1-G>*svDJ;sCjdZJ2&t6BY96xup zrM2Htr+~4Vu~T0jnd|H?LwidZU!{YTQsIjIwn| z0WBTvsqzEr^k#>cI~ou)v#AHt5m!jO29|B)_G*azW&5jF)KD~)p;9L`7s2L!#NN<|)XsPh=NZcS`#SjWK;4J9tpsQ~_h43!yr!nt6*dBEVL?Sd$VP~X5Lwfpw`gmC2~ zrg=Mt`je+7C#k5wNO|49y&9pJ2RhBPd0YX^Mehe79o>-qH2i>@V{+r_#~SerXqkv8 z1Zy4&Gk=5eC&Uga{Rb*i2OQW2 zNyhfr-5u#~4L1p(inOTr@JjYxnLeNXZ!H^C#Rf zR@jk_hNK_}bBPAZ+ifi>Bp|^=huyJ@%@ATQV0_fUDg{*m_&laK@_^+Vg#*`7FHC)D z`TRKsZ;?Dzr0qvSHv#fw#VLt3T_q@-M#y0cekKOB-5dP$dafSaM{L%K#>-{QPF=sG zj~c_-)wKkSCO!s%q-{VPw>2rSHRFl9fTB9<%~)T(YMj`48DvB>m<+DR*e(uG&LYpT z7(BRhhuf5;k|h`We{qwU?*$#}b=%3{LxT=%My&)rc{E4LfrGTo#ovALMZ{GhxhaFq*UYi^g$(1 zFgqLuGzT$_#2Y|Ps3F+!LF5h{TKe?aGot=Q4Gk$WINV2ZakNkY?GzRDzI>S)HSS4M zQ*R`;V$`X7_wId?mE{u|$w~4ge)x~cwQGKa;QOF@#iKI+%MB1TbaImbQf`Nb2Y|C9 z(pwa)07l`8vJ&^l$i@k%=Z=^Az&W5936P z8l34n5fL{K5kPDXb~U0_gZS?ob^$m;$OZJtcz&+z+hJqlg)XxS1ZNCtJ$e3I0V&O> z@EFF@H-N;5_t<$uXrHjKFy!(}uus-`ssLit)&e(n0H!+{=>WCNHRhs;kB*D`-evZq zOb3q(^t*G7qnbpk3pM@0P2c(d(M!cHKtb?CjAq--Qx?+!3(s_7DcswS`}m2gNPST6 zGxf>A3=ZcQgTYOTpnAZBT zeS7`kZz(bElBJ8vDb^$>je$=T;cmB8fBcy#i<##7!VrRa@IB~c zgf**xWD)5B34TQX#L2nc`K+ugm4Trlu#8C2@%gQ&EFedu1riL5#`_Jp@fd)WNk|v> zU2Q%;RZMG_()ggny0`_|();~>I&v2=`xQXqPh zk7xPa>;&l^5g}}8q@ICh=mI3DS#OAB9cV1jv#91d2+WXtNs9oKlLrZnWYafLcDlpS zVn1L|IQG^5oD(AYD6;gVAq)taDHa%58ShJlkOW}-T0+YqOMV4=XmvB=119mnrtoGP z@!xw@RkaF_){cow=gyxGMB8Hm^n?(o&4Cjr`%Dq-x+*rq`JlntGa(}*5a^T{gwaVF z@d43Oq+|q7a}p~h!uDU_!P7{0Q&WHy7$;;0&gX6-fFn(eVAUagcrDTa@E3X1{Unz_ zy=F9Q?TY4r$e4g*E>B@63wucW3UPbVo{zvvkWx!9=}q)`q}Lm6f~tWUWxF4~axVz8 zIy97F4!#7-BIvekc_cc4flg{7K|sfzIxgu+CMzQLU&il3bN0=Z15P4B$*gj$#9E5s z?%9m~$Mg3Pz#w>3M=L@o0O(BKKtX;Bp?0}`DdT6qx*(Fiu~&5P0=V_pP@Ps|6_w6> z>mS9`#fwBixSE3ll^s11O-M@@)=gziNjjrVLQRi4;AONvW;B4Nn_`WJ`OmGuEdi!r zbIe-QMqAK4&c10J@x-VpD?l&W!90Mmr7bNpSu#V2z*w275_tx9^L#|&W3W8CFf4~` z$jD`8WyOG;%ETqbA$os$(^8BQNf0;MNYIA6*(!lvLl7^Uq(>I68%ELf9Ab{U>; zmGS5VF{`gU>I#|a?e|^Y{kug)6^=a8LnLCOY`vqbhVg_uwfR^#8Qjr~L6fY$EN*mt z!?}P0ORB1le^Gbg}nPf2fv-*vMQ0 zwp*wtw6W=7GQVSj#j4J^jE?^KTlCmy7`+>MQS`-j5mXzsiH_A5q;8|UO~(`#)L%&( zA@T(KP=tI63J#z#cHs2sjWn^bxdkdd7K-8ATXMl+ckVD_hX6+Z`uimq3d{*MbLLE9 zKgZt!+Tt&Hy^aW7j?AHT?ACD3|GH(9xQ`I;`(Wp=M{*jsSXMp79!6B+8eKxeL9t_d zD@15N@&$Jod}?3#p+QreoWg4J;?FMPmKM4QXJ*lVMCE_{Ha0+(|DFl-!yIrkxZi`+ zKnSfT*f_KebIikvz7UApV#6(zaW6j<7z>Dqyg@G;ttLPl_~T37zO^C-^qQKr7bl>l zEhl&^$Y`*e+SOrLFG}zult{NNIt-YA;reD|>_Go4k~ELxM5-Hnf89?{G@>}t%hS&T zoFVp9+N)BWp<;)`-@qyoAQB=v~nGY5x2ulOP0H$?CzdP+IR6)QYW zV%LIqKg>#D__Gs3xi`-?qOTxPawljBTwMC~_io?zLRCc6m0dp{LRQHSCGBXF)Ym5eZU47vjQ3Xm07^p1kJvE_b!jchd`P6Ylv6Y+e-7W z(fkY`>U4cG6u|+2!BXbdZ@#09L_oc2)hg8DBxwOM(6C;)t2NJQ6?krn^$nCa@CORO zqnfB}opj9dRv3xRXn1xmq@8#%@*B^>L3RH(X4e*F2ZGE0uU|qd8ReI+Kk;rj<_g!Lefx^{ao&8Am$6I4-6|k^SHh(E zljSeQ#JoK>EMYnq>*ttdo^(#+GDx7H%_%!6R4k&w^f_8L#LqBxuDiSH<)F*tw#QWM zO6*@}OHEI)Gf4Ui{u2QC$L}peAfiNbdE074sZ7s1FUR z6+4VM4!DY8ndS1u5aMxADA*MH)Df1Md09-hGUvEYi9+DAC&UERKM;kr2AkN}e1PL7 z*FVB23|JUHxtP_I^JmZAfHJGlfxXUyJyOa=L+uN$ZwF$=$o!IJD_?^OrfvLxe0>F6 z)NA^^U0g9i1r-w!1%px~6chtNIz~c88l+K43o$?pK#>+{8FJ__K}pGxR*@V5VHlA9 zpO?G4|J}WNe}49JSGnF{_|E%2ah~&>GmS^x-iMByuS@2~EL^5*xS#|kxDh({7IyiH zP_6C7-u@c{QaHVT|K4_F3Z)V>ayDnW75nZ4n0j08QUUN+h!S4IoYU8tdQVVHYz>gs z6*|%|J;69xH%$V@95)II3MdBfoECdj<=cn-b+W;QK2d{u6)_Fdi5lC^0APtkH4){# zzj+8Mlk@n&Z!jWWyih_iriQc+V0_#Lv#ZLF3|HWTN!WEJ>L0(Jfp`QCQdm=-!?7-N z|G+?QTn6gE6khcW`m*`bmlE|09$;8!E>VwN;j0twB8U|W$u6nM1B7>CqmGoaS}pX^qw5F_emPdy4?yY2 zo7sZBDAcze5fKJi&T|Q}o{TXf7EDYoN^Q|qj#WG0JiMFCe*v__OSWKwgx=d@(We#x z8M@7$zCF5ku@)Y$*Y!MR3*ysCwdhYHJjKOe)md9&)%?!Mnx}|$zb97_QuA}+lPCx- z1GyksfZ7i}9tNB@5U|aKi%|QKTJcShabIhJ53+-)uK`f^ zjo6Aez?-#FE-t>fKcG-(d7k&nnYEGsPT(_t_PZH=`o@@gWS3c@=OgL@tuJpD<&1H) zO+-iC*1)Tp7oFmopNc zjl&W{U!Hnuuc)xuqzCrJb~S!Ks=>;+5Y>Wt7CACa6EiPaC`}btJp<4l>&(&>-8-kTQw)`(5vq-@Pv8z=& z1pv{u@$xRAj*N@|IEnb`6!&bSwA=13w+uvcJjkOZgl2@(bwbHfO>OlL+p7W25LH@? zYjnd5nVArq00%$KY7NA4>uyTgB6KS^da(2a4fV6s5DYA#DB$^Z42aTswEFEmqx`_g zP#aeM&gJ+7m2dgHuM=#jmJ1=QaZGWZ3-oQR?d(?RxKEBbfKjAJ?h4dKfS8+K%-<|~ zI}CTraN{({0~e6=xjxOA1*~aB&<@;O0K{C31a?rlNTQ6ey2VkaTwDQ(yGaXztgLFC2do||67$gTc@JH-wS9~5;Ep*hWNBXQT^1#d9jgm(XvkYzc$Hg8k6dthKeRajJ)Fg*)Bs-$4`?xUWN7F; zR8_)IB%TT(NDm}5Z}{+_4V+6V?~9cj4lQe9PB+G`ZS2C&x_s3tRkLQWZ_Zg+B@Rp- zwb~U+o%_#L|G(~oa=S5BTEpX3>-MsKWyR0V?>l!V*wxF%F=pp?xxnu=%U4^1cf@FfD=o@?IKn_el(X(dy@evf2j52qK_157H};5QW%~i*uwRxw zwrwz&omVT0kre2AhJEVKF;#_hl~&rQLodY3XU2#}Cncv7@hWUo1w}>9eV-VJ&&x1} zs{7&s&}ycOC?v0dCbv$DbJoL87eUO*{5l3bR`axi-|tmI-729U^2Kq+!|fT_-7E{$@tx+_;EFP>3`9Dy}{e+m$jLzs)j zN*0J?ez7f66nZD`w|&nBtrBbBTVm2tergrC<$a~MrkJ(l9>hw{{8}spzB-#hEzJY6 zhR%Zh4U5b(ki-JU^X}rkp&{ZcZaQiRYbj8`=MX3kQ41r_etO$(LE1^IvxnG=tq@iVCVYFNd*D&g{5>BrG&$FU*PH${GWWRkYiyJK+~ z|1d>v5k>`XXaE3kwXp571jW*P0tlVMf)*)0$qFZ|PLvW-Q<>Armpi{H*Iphk>eo)0 zyIB6jqFdK)Z8|!~0)}Khsk;sz+mmt*E`OG*QXtjPSsSXMX;PCD*)wJf28>(m)Y!2sjg3^brHWId=kXx&c+SRDOUDd)(IgJ5^dRm=S0sQ`N1OSc&|Tv{(R*X z71J~C3fp%niCDEX@#WwYn3a(gFBX_>T!uLgj41HC)XWvbFsq%hnCyNv7UAnErSJVf zfH)*7mw&VFf@Y`kl`B_h#_HRl)pSN;;1(PR4TXN%qKR*#za|RQ=tfgAD@VxC2V?pO z3viM@Sh^AY*KADbZ|3(MTcfL%(A!rq?S>ir{48%_#Ot=7NZdE2rcdLWqjcUapWi=C z^sb1SQ;?YCo>vO+yY(6-_>H#PMsCQp=udoKBUj&(Unm0;oOnQH7ks_3!vr1GSUfUH zRV0!;aX`WWG5$0bO(t}%5D%BDM~OSxzQ;ZJr_NRwSrD;$fEK&8r|ru-m<3tE)6gs^k#q6L8 zV;?D|ZoMkhTa$@AV)q|4JRnRdV>^5Icr}zMh(9S%a3#$0?1`c8ZSG5-k@AP|QJ5oW zj$v$WKH>T@GO`uHozq?8+J~pZZRjc$8sai|@;>+EGbkpr?a5vK%TX_pegh{* zKtQAz;y`DcUYbRXIBtv0&24(id;KcjJ2WP}QXtVD3@t!2n~2cPk#QEj6}KTg(TOP8 zF*!L(Q;%;e5;|n&go^XA-h^#NDED+d>x-z0Wn9V$*tqM)sFj+Y9u>hxGb@y6Yr&Dm z#Ca+1 zJR4&dj@@Qtm!Izm{i$Lr>-n2wOU?Z4epNjdi6J_jl{98?4Xgcv?y%zBsKnj$h0gqb zs^3#{w*1?7`S;&C+^FuZA+AXf5bh2_+%`vrS2&73Lesn)BPn$q9Sh$)Xm!CtNk8FW zEhse6jhAxD_;O&|4g5&ZBy|J3Cp@%l@Z){H5-1JTROl!j=xsd%;=n#F3SE47KVU7d9AO=YioRauDFF1>q@O4@rAVNe3gTu#7+JGcN zFuaLx9B6V3rgUG|A#KqyK=Vp@s^=7o7`^_N*I{HkMIvmM92kf4Fv<&|j}k0=vWJ=HUnL8w_M^xr3p5g!^eQaxD;=a>X8jFSa3oLqcQw)=O1a zCYoDr<9_(w5;42CqG%x>qbrge}_io!e50 z8x3q76~PfwHzFHHlLJ2PYM_)I79wu%5r7GTNJ9MF@P+`Hk^}Eo;h@LtCmaQSfK4W; zPaZ$E2W(C3=j29#PL{#N3>UuZ%v2u+yL7nrcpKvgDg)0HDvy{##V>!NgB(7HJ*WVO zs|Ar4t5;B}JqCoGNWzhvL6^EZ_)-d{HxU+yeKkajxFU{<6Qn_QD)wtabp_w#1T@r- z9cEysx%GfEEaK;-IVw!_T`s7GzUFqHqE#sY8fP)1R;HqtCO#C%Epb>x8YANJr3m(( zVPzvIG)$b7yxB{mXQI3jjb0Iz0dK#Qx%v5~9Vs z&ak?wiWorN=MJ{5Jhp}y1QW-z_SF2z;j=KGspC5Y1PJ~|pBrSRJ7EG>3w#7*Ou-Uk zn|xDVH99A*u5>o__w;l|ZINy~aRV&WKBOm&`aK?aUpJB1-hT)NcFndW0<-ffaC3Kf z*ZC`Mumio>n@+up@I#Cu>f#ggxgKA=rVB@kAyAJ|l zg;3T^$L$vbOV)`QG)tyzO6cF?_@BSOHtSgChMvzgk?acY6Ao}&Ycs5BnbUgNeK_<0 zt99;FL zn)&_SwK_3LBQQzfk0BXGGv9&$OJZQ3=9(sm`4K};;*o_f5n;mC$$p%Keq>m7TBJUh z^U73|2j6+6u?DkOo{n>x2Y)Wv2n&eXLRRO3X(+0!b?r!TyL$&@(a~uDMgQ1=nf@vF z(Gq>H{35KXG5pV7T6`q_t4U4FBW3D~7c!6|ssy?s(7Ouqhsa3U;%ye`7eB28WLm{{ z2qbiDbKd86?T0g1Ez~6hLYWA!pE`w{h2AOkmB7*@qsjLErGiPv664_B&cG--Cum)C zWFbF)f1_MShl@~I#IPdW{&p}h(sRD>W+Q&8xc)83X+wZuGy)E}{7%9^g^C8BD=`lh z9~^Sm8oIOM_17hG{{d*Gl3b0dfWv&`hgl}^>l&yyy;PMwB4|U=&d_lqy#|3_A)7rW zYUViL+Euam>0mcg1>$l%Ew8nIneA4_vyh70gDz9{sD5qg7%a+{_oso)<0q1gumDPd z^269Bq8gLIYLHI6QZR3(LCXK(tluXh7gKZH>?|E}T(xi8IJzgsFV6tuLe$Bs+w zq<k=L$2Y?tmj3}9T7@W%h%-}41cl1opie$K zArH7w+fQ7M^~fxw(TLE&zC4!B)(?#?lN@?Du+VaR(oOQ$>$woqf2`=5IS5;<2+SfP zdU4ZUVYpx1{lh?y5r`H#LI?a^05P1t%R16auy!ZEN3tQL|?2jtL^zj(D}9s16>A{6B;eK&yC322cQZh`sR;CR1pzuQot5pUg<;dd>>4yLd+Yq0Wwx z;K!$_tbE?)2@mmjk7NXHWNwSaE+kYr)04cBBaD4_?b(x&Rmm--JC|NlOq*CLVYHVy_LwvxWf zD+S>GvBTzxI$2eGbkK=l2!yg?FEJqi<$G^2RV#T*7Za=_))xbxPv7=BFM9}z1|xKO zB6?dl(1l8LlwQV@C>n8#WoS+>92qsSPi!19Gb8qj1DCOPYQ- z<1Y}+Z+i$LrP-w8Fmj3>gh)jBKTfJDT7_2#(jFJl#euZZ~@nD;>`qu+uIg# zD;s?rF<`pS{bmVx3{YH7P#%UW!K0FPwC87ZZ9Z`LQ!%8D2cdPX;c>6E=S`o6w%KM8 z!D)G8>-?Rh4*{|+SC5m-K$}~K_~%V88adq`ctQmfnB5k(MhM$@P-4Xk1|?5+(8a*~ ze%jc}Zqj>>jLkn9&(XMxbw#l8(;)b!>oV}ck$!L&I$Ki|2eU0xv7({tKFNaAq1I<7 zknK#s+o9xMvGJ};Zkx!+KHXZBDe7|Ox_#34WB}R%IO2&!yh4|U@im^Y(e4+5BO%?6gX`yxYJd;5B;F37OZwFh8tO3^pv!p9bq4WHgUx%Q) z{^9tcTA@#o$=&C(3#@KUUB2k6;eEeSQ|!G-pStpMkpnUrK|E1xjoCGV<(tI{9?cH? z;XlTD6LLy2>4nak0`0hs(|nZzJbpc)fxFTqD|UsJ6HL@E>48ZPS+XR9P@k0d#^n_Gs#oIp>(xFzp)(}O@1&Cvl+=({ND+TE9 zUcC`eYK4o53mw{~LdF7l)ByvGB{D?})@=(7gn0pXU7x#m_mv~Hw!GzyMy%74+PWOW z&oskJfbzYkd*>ePt6e}^Jey%4x!I6UXGu>pr3(|aV>X~vj)OX4#8244m>_spIOH_@ zCs;Wx-EopcKSRm#i^TARS~1!YYyJacJ+4mrEA0rZk`nLEM@4P>TG)v%wYM$0zaD7L zeXuNlg-t)XEF4U3eS?jM98cqt?=ODdlCd})fxmJ8nnh8t3f08GqMEGEfmD2r8m(kU zI@21=k*=j{19%1g8VLo#utFx*KO%P}=Bi-MA2&b|d>UV9e(PTMK|8+wr$LH_hLJXP z-Dx9bZ&8KCVQ3%m6V~t+Sz@EIQJoZY1VjNCZ>e-hrz8#~F=Z2-cKzp8jeWgJPuhL} z#2tbtyn28<%m7fs#YrxFz0+l4!C#KxYg6Gxm-!;H0|YT`*CCsC8)4|A5Z2LZT|d1Tmr`2#H3Cj6=%*R0*(h*&TuvKl>{gN=-E33k#k?MHUb6W%DN&(&!5Q7#-x~t z#utp1S@{MByEX*D+X@HPtL_iXzGH3%gCp-{i)fCNE>!2Q-^9+kqK)7H69NuYI`@1W8=98?<;DN}f*tZmhNUA2 zUEZk(Ij|=Pu5`ZE*5Zs1rxKsEo5K+858Mvj07|wqSFI6!LgD4M9`@yBX1fY`Af4A zg(>@6UL&Y-u~5XS1qe;%LyI0vL1W)uqSDOuqgw!DV|eQ^p|3y}$|+3p#Y*v8rMf9R?$**XrW> z3F{UNjJ1Z0D%VxrHOvHDSoIEqo}`5|_L3yv-MR1y%$@N8(HwgzXFKMq+lU97)u%)P zd~g8H9EZVf4>W2ae%(vG&4aZ&Un1)xpp~@gt0gXa7^Vw+=u^RVRl!@ve+B;YODG{C z8KTAO0aDoj!Y6T>!_>+go1t=yRwqWnpi>?SWNp8CIGkP^Z~QPMF8O&pMk5A1xN`B} z;k?#C^oZX2%Sa#tELiqk&Nv<1{fN-i6AlcW5(di*Z#lhXeDUqWeUbB1e6&8K-7IO3 zs~aOe59kN5$)i~?=7-J#nBy$B()>d+7ViIfOj)O&m4w$6vPX(e4eUAMlowpWk~09J z19fzJR=Y!S@h?8qPiR{nyz6%=SY9rU*T3qa@jsE)Id-f+g5h5bdl) zcSal=Q9hRTQH(FvnqPE4E9+!;-SeMZfO_4{FY%Jty)|r{zfjT&iuch=pXnf=?s@KP&MEe-n4lj2KEPQ-NV{iXBXa+=hG| zY!u0fQLgJN3PYlzfbqTEJQUc$bS-LpBGnHE(2KMjR6Zu4pgxy&&xZvYs?FccHvJfZ zXwHC=)MY&lb^>_l*uR4xL(G0aSb7m6%QlE#^u;_T5*@yB-4YO2s!@Yj1}+OaiN#@I zxE*`1KUg*FEP5c7_y`bHLp+c;nk(o*ElvypVu5~yAq;2oNjQPmBJ`O1W-uv2N)Y0c z%PKBDbw6-k2DJNcc_@_ce%RRp1{kpEC~C0z{4t?8z>RE;S{3#@eu}#B4Q2My3FT7R zR;^NHK?hTdbxuM5`b^0BUt_b}LR>YU{p$q%>o>-C_db=4_1EGE!40`@5 zsov5N)8ip$4o&qJcq)!rGVYC8cG`qGZXD_636!)w_|mIG(U4}p6UVi4NA~J(wqJBt z{Oc9}<2Tp(gY3!=rkBBP7{*eSh>-E6U^IR)T6euOBTv+EY| z#j&DqCkcQr}?cu}P9a7_TU#4-VXB#oh zM4;hVNenzt^Z|d|_1S9nW-Ls{7EJb9aUF7eMe^X~4J6Hgd*P2vaP?|dr6?*M&Wt4? zNlAMng1Sc&{pNII3)}Gglj`_auWD}0XPX(aWdC~0CU8lOcDS5Iri&CCYPuu)ZBM-NvH=6DVY<@mfQ%Ij@zVo$%9Pe@+qSI>N~m>cgKy+J0wn}Y{}LO@ zCm$oAM9-ePzy$R~_SYM1KOE0a54^d){%i>J_}BvHwc@qmD9scyI_>l3@u*ERKZYrg zqzFvdVD~RI@GZ;{*{Bkld@m4WPN%@?-m7N^P8yfATP~@72+l}aGmqBUaq$4piIuGg z=z}C!q;JHi#R)~r>UHaOo`fHtSOc1ax?;)^c=uTP#($a|Ik)(wEX*}M6cnwniUO2R ztXL10#I)}^f)DKyz*jQ|b(}MOv0%Kr9LExXw`v)~)zcd_6@Y2lq5n;cDs5Finr=&xGk5pK&$Lrja+V zwd0z0B}n|NEq)i>awW#}rL5~?U9N{eRa!4T{T>l4DmL$5ulOIo2~t4z>g7>sMZFNA zgVjZHdIPEEV9)Iit&CIOg&CRxIC@sJ^vgNT*v#<9Ug6pwMlo!HxxZwZQ9e8eF*(u> zbzlUb$Sy!($j06^P{;@DbQ)FavaY#h3P`XRDoJ!Pprldc59)jy>!+J;vU7egCgCV` zc`dQiN=j@j?8E{JN45S%S6~~UvR(2+Y85A}&BE)dd2bMGv8CwLu&9i63|sIU@tKfY|JzwOS6Z+FY2? zh{c37q!1#ZP1q6|(V$H+$Q%<(l2kMd7+sly7{?`XB?)IV%LoBdI(f!$e$ zaiJA@YQEIFSk86Ee%!G=c|(tJz9Wbq9d7W`fppZ}#Qa;>dDueN`ptaK^rH^wC{JUp zFOCc}K#1Z0!A#8V;U~DibhhIjur(Ue1ve~H&_9HPHUqzEG}(y^pi^9D>^^W`U+c^J zrur$NIC8R~kZixGVImiWR?tBI+=C%Ck(0u_yoFcF2ry^mK7Q6O3q5QzYUM)4oV{)7 zIsb76{`^B+hxLKgJiSUTX{wWMzd{b_h3YR<{7n<6h_}p=at&rwr%Qy|o|+i;JC^rQ zvx7dK!_S(co;s-#ZKTMOcZ_vUkzcbddv%}fm%nS^ltRf2h21l+ps@uug~$e;Qz0<* z<#jNl!8q$-DQU)tC<{<2zOR-$6rR5%)|a1`IDA+t)T#rjSdux z1Nn!L{1KkGfX$YMVI_@=6L~0$_ToX(ZR;AF{oSn;55{;W*6WI`jxx zaLwaMYWk98%QAA_YL^WKmGdJY7HU@;P%Ment@{*nhCty5cGRc@93#QKHAkd#g2w`3 zS(rl6ifiA6TL?G;q3-F>8x$@VHLZG&77;fk3R~l0RMWAwz$F@eL_j4#(0x+Wwzwxd z$Z9GVJBDcWx(JjH`{jpcC)F^|07>Zu5;9)tG=M+y+9-6QZ~{v@Jx^g&(#)~jU{HrS z>OvwCWsj(6dX%5pOVRZsHjW^ZKPAI^64gkrIN%6Px@?_Uk%!3rjFareS;; zVS!eO^m3G@kEdOwG7^_oG(KPQBTI)nU%#3?sud+x01D4ni0VQe_%O(=>dTnJX-!NJ z|MIK0(^lMG(ZI9Z^k>bTn*M?i)$i5mO{(ts!{Si%Wc_G8e)thqLOu024jc%oO-tHcwa*Lh;N-E^u(Y=R{D`%q`6q=cj^H6Z~`#rox!{!K_11A zk#UFa$fuMq%WK>1#T`%>m}3uitc_fmU!%t$~U zo!*}l;!7&H-WS>|Fjn?ZaBIOMHF0Sfystk26St1?VY`Hea&$30b#7+aK2WVxMd=;C ze`LIMS`Xhrn6m2dxhBUofBmEn{{kK_5=K*HzwG1 z9|`sa&rGMzZQn)hjA&v%mUYuvp+fQ>f74m`?vg3jfQI25-=%NY`@4jeQ-&&; z-sCUSQuW@l-b_WipkLv3&a2Ukqm!#wtT=ypKDa{M?SS4;V8fWQ{28ieHn{}KPT$tE zKP zIJ&vm`6t3U~xVEDEemA#f23g$BYXc)6kH z$>H)xX79{{IOq@YQOjkKlUXORAjm32n?iX`kSbunvZPEz({`h#^fWZ_rzZQ|8G!67 zK`5odWLXyMH8xYvbK;HtfxiC!T#3dbHbs4-*R~Nx7hoM?a_f~5i$;NXgPzQW4|77? zZLXfG$opV`MW7$*!I}~sKMo3>+5$v}P;+MNG8T465)8l%P;=NM|K>y1rPu%AFE#BR-pShxp^?6}u;nMvraJ5lN!SZwM4`xBhke{Xbs$T6Wvh z4ZV{O|C|*qKS1sLb3SFwOApbUmiEm#{#ah}v5q4>NloWv+pFIhxk;9j(^uYzCM(kJ zCQMKJFM5!)cier!*O75=+nGWc00CsX6n528Hu~(in5vPemq1#}GzBzlBzhCbSwkY@ zTmh^j5&2G{*r`PW8%tVTP#+MOG&G_#^{%2t5vT&=f?8TydW{}aQN;1-=a&mk6E;~t zrH!c-MWOf$-H;cdV=ibJzyl@?Ce+|J>Y&-KA+zA_*%LvBA4q@u#%gmSq#A;`k`zb= z@F@EBNoBD6W60R^+?-n3?>2I15-cwvas4760E276vvc;&YJO9Ys6wHb`Nl78Ex1@_ z@Osp!7Kz8cmqrI!hDy%;5Jr=P{dmO0n3|HJLQpPZ?ubC71+Zvo_r$qajOfqlHJ(VB zbV!iWZw$K>fDysq;NTD`x2)P&P7c^WiihwMNwOZjN~n6`{Ka@hEfW^&kDiTCpEmDppUVZ{8i>gJ1P4#Y6^%-AJaq)&ON|QZL@+mhr0z50HgSTZo zF^(|;sUzYC^OZ3U+35T;nPL;l6O-68J#Db~ujezm-BH}P;S118Mculo8v2A=!2>^3 zdNXIU!N0Tm`l@i(t($8UHio7R-#KIMqtZ+QeM@ybO-TZPFmcC)59!6 zBd1AhJfpy<(Q=(gas(X=WRV_(8TM`+P+tfDg}G2qb+iwGihwDfBqu|rVLnYCfS_{6Z(0tfZIiQqF|Anxm!&?@S?viw@yr(Ma|6_A(ZXCt-k zu+Sw_Ake#oLCAPmw$y*@O8$gCZv_o%Z0M4t`f5fD=JxoLUPAoJuP`vO&gFW;;}|Uu z3$)u^0s>`ku}FmS7-5hkXOxC?OtWK2v5J^TCQ#_+fI7Sa;)Z)4iv|r{mnsaF|H0yA z0P0-UC1OMn^`c(vo{Bq2Sk%V(*+Bb76Dfx?{a`8ys5%&z%DPPvLlds)pw1Wf1yeE= z&XE+JO=xj)BMMM0ll{zlfVpbr%JSbMgwj(fQVkRCaG-%jR|K!U&gvEX>w&!1@B#C% z&4uB&)+63tBQvYRRQUtgPiWjW$pQQwln4e*WhvIUAu`#v65@=sbyLt_E*3h5r8 z7;^K=64v#1-yqtTR|EG9(vA_{Xc}m00=n61h<^SR$d=j6@8`1kOiN+cZbL2W!ladiCl*yOp9QO!-`ZmyEl z3~bjzTK+S(h3z5W;ffxEBwg&z@+H#8#P*M41w>6*Ih>5A6nFHy8uCBvjA$MYxLC$` z)oq}}o$vmsrD&G4|D#9Y_!bIaq=Pb5gI#>mQuWrS77f+|Js`BET)o6gxdU<~o%2;O z$8YAaFm!T1nKIq*wr}+HGE^#z*yxZd4o#hKeCBPviZC7uyp(+s>^)^fw%rzY&OuOA z34Xn`P#=`Szu}151gpMR-KnEU_gR;ez2Tl zmmX?fJ0C4?6+uHw;Uj%JOu@QClNKtdmGs&qn1JPA`Nj&w<~zGu4BUlXx}PtykbzH$x&cy<;p*peVX5OOIoqW!~>~9D3bP zWc6HaKgKUq4^Jy3`v<1h_s(|i5Jy=2`FDJwnC<-hKi!!(&?s@Wqp|a{1(3imC`jE8 z0#_Is@EVh{)K7royrkET+9Nhcoep>b+~f>Us6{G5YH%e_&Usq_#%xb>P6m-G0pMhc z>p3=5eTD{FTmqzGfMxiXvPNBqt9uQeD9GBMOaf=(!0`v`V|RpHd6CUBOF&j8Jxvxc z*mlRpkw@jUZ#02tQ)Bdk9xTHwX>dYz!sAr|MqDsX`dAYEK*>E4cr*1321mgG?xGTO zRTUNcHgB<>LT)QHbe7->MMC8UIwb0te>^j^dx{h}JsLV)iF(bn{*N3j))>})&j$(c z^Q#~Z2y~y11E$br{#!GBe4B*F+Q*1F6GMr%TodCy4-6%CRRrYh& z5ZJDm)--3+J}9`AwhfS3l3YDF)AQ{wty*$@s2%`g!iaK(1^`sKtn3sY@nFTa{>9fO z5A^(nsWvOD0fc=AlO7e%C_(R2hgc$^t%)EUHw3lScB~Ig|8a%3y(s#7!|2q_q{{L( zsa}b`YBp}kd)^PncI1fgeTbZ$&}_E8J1OU{i)J(N>$M#6ouj4_#@NaPbmUvJYL!8n z5kb?PjGvI79m)gCiYX*JZS|3OLH1CNXfLZJ9L3A?w_<2O_M)uZnB3q2m;$dD1HUm1 zBU#JlY<0avF_ToZ=Y&~Y1|b=6S&ChtRXxz)&_2NK+iid{(GNuk8NfAl4PokgzVYX| zvxr2%oJY;9AjTF>$cnkl^6IOd!O=Jaa72d-P;Oy!55sOlElB!+WUK8?olM^t@pCzL zXn=H>gTM3mB|aTfV4@U|^9i#RpK@5aGrCO?FhLO$x5qFV%G#1U6k15do{EJU%#M;@ z9^bU#hQPxj;w_fm_m0eIeq^D_v&Q-WA#jq(1JW`Hzv83&XWa02@Ig4Wg(J>!g|R?02Zw>v!r5n zJKG2inp_si(#VF}FL-cu~q07`dQQ~&(sTy!|IbyDV zT|dw+cdX~xnX|UJXm3?bBlC2XorCzEcJxwfa-02Z#fxhN-o(Dyd1>h>n^@UZ(HLMh zbNuziXHM2;dq@xfn+MUlQ5*tG-0d@*2^h_YjXnHIc($x^PGkwBDj!_uo9P`Wl7ugQ zTn;Xpu!(uyLbn3xmrV4rGU9vIg6m!jLmF0h_3_8{(CsD*JA`!xJ&jo4zVs`Fi0>@2 zs|Gkl-3W^x{L$4rLn|~5K@fUb*lZSoJ-apwl(N`Myo4cLCw(Ex6Q{4&7tWC~LN;X} z@sh^O7dx`hosu8Jan?=|>;X4j;%sMUM+cX2uY_Ff0z?u`fCdg|8b5V1O|dyoP8cEb zqK5m}Hc(^Yih7Ww72F-8hg{PWUP`GwLHzsnRUR-OaqAeka>FU_m;QCB1*`#yn@j}; z+=6z3b%_bUbKsMpS<(rG~nwk`qy7H;+yhoEla%ot6^zydKt4Mh>jy6&t8hZxp$Ym&0ffheLRWKq-(->WR5&&@Gy{-_ zh{&4XlP9W%b=y?eV7^dzfDk2uv8h?-7b^q)mSyVAZqQ+CF+lTTB3GfdNMsSr33YmU zQvFWgZT=O9n^%5h)-H+a^nHven;lzVjR5**tbG=L{P-Cdz)Nr)kZ4a4V-6HKxCEcf z65#hCM*v1-KQ`Mo)!dfP`H!Br_G==1Z#NIrD>zYwY25Dab3K^E`M=`sGWeWqa9rN$ zS$W2s&cN_hP^cz@Y^+?6tU$}WeH~B5@d<-?d_Oe zRG2XkD(AN&qY)5(7IDp|lgN(VQau3chx%l+0RI9mV_DZwE4&#w%mc<^0I}Bj$KF0% zLF;L(P{aZ|rkA|+x@}&BUYv?%OC3xn4@Ild zPA?n2|2aof&YbDJk7DiMJ0-DKCgMu4+gpEd>KYoC>@sTf(uvg?yoY)%zg$m_13s(@Q1(7jI4oG7K z6BsI&A#6FRm~U%jP=dc%KP$L;9PJ^HB7Rs_y=2O4#=aV)kn;3gS?UBPBlAT-pL6t# z@!+5%`UG6RFDm*zu#4VVSYcPa%q-(Z1sdwP3oQb9cB39~LdoUD#FHLtY@Lci4HyeK zbWYK6JtTXZy8cc{i}D4i^^zT?z&`M{#fr8XTKW%`Jyd`JUFf)~AS6cv#~#V*EJg&i z6mv-2&9*Waqpa#-#$vLD@Gf7ci0w|yCY9^T=~}xwGjIWh*=gse>fQ29pCan&+Leho z0oF}J;W@MVhhA`wUMJSY2+3mp(wlooY&@$UY85r|jD7uqO`#PCQq`%}3SSplwg$ntEP;h%lQDkjs*7VXl^_Cr}kBPq`WArl+MWIZn~ zFBi?vz&0xxoS)F-Q3co+ZMZyd z$ZGRRGLo2!Fmwv4or0N7Voq}N}&3+1gI9NHyUa!4

    RB^PM-g6OzKmT|ZJPq$RPUK=7TFggWGJZ|gJ%?pef(@iiOWyLzSQl-Cc-%lAgCGT*nw| zUEOrk^m4bUJP}-^I(_l!g}Khs>@}-6q`IqR^_mM_=x9&#SGM>2c6GBj8bUfnrXo-L z^>xWt4)+?;Jx5PfA+(zM-oJNGg;0J|q7I%M11f`xpcXSWZyrxHYcpr#zODs zH53k+nA>}}Rwv(2@Nf8#zbYXwy^uOAu-I%@WSwyM1-6yjxZbbaGAsj=OXl0E4#i)m4=Y zG1)tA%YFpCshZ4cEtQ6eB<}tl4BFncwCG(4?xJ}_X658W9KJqBrXIEm#S-3zYE&KV z10@XZ7h)(Zxq^a&wsGQsqLvCS{O-$iIYisMnP)_hjdAdj6 z|LSwiRj#^S0IvP7K7NXzytp0W$jYy@_60=m`(uP<a)I#jfBaha6=)78!xf?|;X0^0!ZhMN zPt_0MSfD4mEA!{W|k1=1=~AEyeOEMoNw@|F5fT)__E z2_7HDfUUc$=)Oqx9u;NKk%xT6qpL66t4)OO+`G53oMA7%E;@YT2UaUMyPffVzjbP3 zpYmczq05kyq%fCv{n)`nhYYaC3e%?*dRrZMUZ~y6OV)OB%6^$$_}I2o=qMzoZrO5q z7hWhZrX+CUsif4dncPTBXaIAjlnu;STBGtejIWq9YAJQKREAf=X%ILLSaqB4*?Bb7 zidgqz5>;CLOdtFl<)m{*!e2$YwITqv#C5LD%UQBU&gau*H1C|LV#u z`$hiG!?n{x=DNy!4nFnIOL~i^2QJrBBri$SZvLLkxx&XG^Y@ASzlHM*7j93zD_?0a z$+5IfzcMUr>DSTlB?6~nKWE-rna=MKSp+H7|LfOmy!lLc-<|T`ZCaNP$nrNn=36N$ zQcaWujR@f8nb^?^sfW20?5tNntCN#zWd@Dy~pwR@RXOjgl%c0fD6? z?I}jDD#q-E8|AQbcRz7Q!T_GwoW~R*1|{3JpMoX((ZSr&;Dm$(Yf;o!g(vkmxw)an z{zvK%Mvm(M+D=VC-+BXp>961+z>&~1FE3BKW7aj)NCtRe86U}m2eo62puL7W0XFj! z<6$^Ati{ful5Q5n)z~m}ck8|~_}wQyeisBee>?{)x4ayCL4QD#_fgY3EzH|RoX@CY z6v!zhN{>}th|!fFPBd}v13A-!`NGGu#SLTL& zlN&EVoGtZOND(sj+i_l5;`@R5+REZ7p!1-tcODtyB|xoRZBwG`y0Qm0;n7|rB-|`4vf>{@y zux$A6k@1CK&@CcX63{mGFiX(MT8R&W*lEAK%7i@iA@~e*vBo4)QHB~l>skgEi$@8L zFr@)XdkeEOxS?%<&IPR=&@}GZ@pSE}PnDG`0PVelb;)39RAl7h+BD#CJ@uR6CUi1SSd_4G~)7fb}7}|7zLX_fbjTlhp_JByS zHs4vO>q2k+qr@ZAS9*C|xM47}KLVkAd-3L7cKGp*kn1WA)sV6zaQnk8u*UYXVN2)M zSy$jLMpBa9nfbF2@#JulZScS)Ql%${RzZ;MyrJQC#_4H$pbOdzywJMx@^TsJMkob9 z6=7paIg=lYhmUDfIk~N+pFX|8F3obpHY}=$^@(A|bZ@_5@!H9?N72c>f#4cy8_weB zsHnFndpcshIw6M-On#9etRH;<$d*V{GC&lXkM{A7mzPUBJvb<6D%YwCS}IpyB2$YQ zg4E1y6@;Q4umE!BwwGDH!n+Bi@jOb#)4c(bCChunbLwOw-2iw69HbJp6E{yWa9@TS z7@h?rwU_QKD}h?%c9+&RI4Pp^xmt^e2xFIzA3t6(P>il7Vv~iW@QwPk$z|!}ryZfs zSeKdViH~v>vDSPs==~L(r?>PnbOXmDPx)O8>0%Lsad7+FpbV2Z;S?_80b%L|XsQWc za**LL8sO=<5Fo3caUTx-FD6sy-BI@gHs9x^HRQH_){Rwr?&@>uXrWO52FbroVQjVz zNiro>{9e=V7v$t86&Xq|oR^XD&iA%inqa)lz?^9}JNYrIPo?wOTRwfDA5T?wb`@W+ zzWAuM?}DB6R|`yc{2#R!{#Ag!we{X1C+1Y(F95!O){HbW^^>$nms!6MMYq?AM&jvm zVA7R1nXQ+292W*i`7J#CHcSG=PLiy#7tS31VMQd-N6>t2 znzkf93F&Ydz+yQBCa(iL6*JnSo4u`ZeIm)Y;X&Ts-qs3; z4&PpF_YnYM<0%($wU-Wg&jCnwLPNU^^*r$~rb3rkD7*=+x^;Uo$G7aLP3@Wfp@8of z8SCom8V3ESwHM!jVbKm#1wdq{tCifI`1|uRvO3kFti5Z(xvjfHr%xyOp8ERU0^Q>zZK(qk*3k&)4@;qFT}jr6 ziER4*Gt3k~3GjB-uiFS>KM|VqNGbT!z-=c(V%i=lKM09tSgOCrO*KJ-{71x%7$Kqc zSy=_0^V(2bNTgb#-Uf)MuP)vw?M8e}_MaIt1E2D09}@#EYrNA5ylzrT%F3A^9!lWh zKSBPu#dt0QHCiw<)}#(^+wTwIMBXzYyRYh=KK%!PC0 zoH=uJ)bqFcuD?5)mF`}Y*uxHMwmYe-w&irBAE+&xT`-(fR9sVa>tH+!vTKW6L{*}xiD)YT!~CBnr)0LK%Q$*qIN1TYPr$E zQW)oX0$*D~_d?|YLeZ`u_p^7VT!?chbDKLf5#ds@YV^8t{7-Gwc9Q24vlUuf6 zOomvO;p3BuUF{sL|jF>`F%GwdR5xtfb{^fPOJ=4z7c_rftm2EAk=SuT4mP_`L1b~VzY~za&$c9u#{8Y##R@mw-VH8zha?jZ_F74S&vk3b`N7Y?}+Ok zEaW{l_gNC>vX_a5z(gBO6j$PH6aH0*bg}L_*zd{e*OM^ zi!a-WoP6O}_)laDn}P4gPO_E~;P+)z)bq%vO*LjT>tURJa;h-# zT_BkVe8uqD?ZJ3V1z@3DvpWyR;x;DX;TC8zJSu8OuXi2foP`C8oo(Bu|9OlDCRfh_ zswk&oDGSw&^FR($MrKc4`8gAeC=-3IWHAePHU-N&|(a6t;9DlTuioA`-kquZRH zJih^tL!^^iDbT+sXY5Bu#{~ot8EqP6slL8`z>fXP;2(Y9&>=R%&Wg}IOkr$k1pQ(~ zWMqgzRUQ?|9!&UuPr)0r-7ut3_~lO;OnkQYCM$fyw$tMWvytLOt|SKP$JUx*P^U2? zyqfyc^O(>4`Ab$GMq_p{Y zn11}Q0e4zr*jrD<=>+$mC$4EM3vv$o;9FfEQ?PGtu+09XsZ;FVtA8(;_g}B}Uw`yp zj>V*j^Ir;ydN!~$USABIHlq#yqo*3W!ZEdG;TA|{MDu2 zpP!LX9OlklR1{=@P$7bR>1eaNu$&iQ$@1n<8+Zye3$KM!b4)}X8MOf~hf|ob>4vgr zQRCY{eBDe!9PV|^>tX+tJ%;B01aRFu>)c2$Rkt|2z9iU1FGPtj2KX`Zf{^7Fpo`?m z#yw+-i`eGkTQ_g+n9MYIuI)VcD|;^l8n>`YpNi({&a&(({24YHD7O7xR@S1{URRXZ zsP%hGAKic+N3r0Bryu$cfdByU@8ZVZ(G$6A?6X@rlaSPt8Ss_;r8CP{NDNY(l8=1< zY2Bdf#QVy@DJTgiE1Vlh$!~+uV~7U8t2d&{Q4mXeu92Roz`t+b8f1RsyN?o`#t-C(Jp{z1VT`?vCAdE%2=`0gcnK z*$dd&+f#R=;!X^~qsZ@w`U#Q+YF#Hs$D5vIb3u?cMc>~6pA86*P;I&#gmK3W@*Q12m$20hel*-@rRh+%mXy$(Y|?it@r}q)voat1bemY+clx& zw>!h_Njx4YUF%=>ar?du z;ZJo;)8}jRwr5AHg-ZW?;QW40QPxLym|oAh36~xXxN%xQ_qG1%)wLSMzc@agYh85v zZ$|vD-&}`GF92xu)OjL!jIntvfy@|c_a>XbdT?&*dJzD)BZOvXK79u;U;9j(3vIaz z#+qKWv^)~658+obb+O$MX0{461}RbULXIHd`NO(($jwg|6lY*JBx^(-#gI#b# zz^dv%bqwGn1VA}`mVo3V!KK3rzQvZh@9WFz6RRBK7uE<_0WarT0ATr`@Uu=G0vNp- z6Q?%bJIRt20S>_GDU@18p}yvth-vH8TRCPj43V+eYlgFC$b?juK(%r~8Z+lLy^Q{yS^a4;6237x9I z4=HKPDJK5Vcmld`?~E7HXvHAv({9m@S-3@@E+1x@B5KJi(6CQ}gO&){?9kaNWYJTY8vOlr9JRG=>wuN#1HIJ^$lv1VlVW+K9S0-oqT>a+?QN-yy}u&;A(gbwgNVmNvclinmoOR!2C z13o;^ingvd8yY^(PW(i-FEJ2tYY1T9;|;J`M9?9z^crQ^N%qmgp`Jc2P-(E1JiH-u3pwhqHn%u)N5APK|go*C3q{bU=9m!HJTe5G)wSV*B)a#VDmK zC>~xV8aJn8Wqq#gAsFxZEZYQyT8QJJYT%`$9il+aMcy2aAE0Ae61jhrq(<Z&KviUtQPh3_(n(!(ch(c%RQ%Ddv?Mlmt5yao)} z{F&f7*B*F^y z%Y(UF2v#0D68XJ$G@uCBwPZb4;(+9?v|ui9Tqszp`>EnFxMEWlD=x^jb}vQvbhd(s zPP9R#rou5d^d0IAu8F9;;E^18^)M(1@jIx3pP1r1pqdM11L|G@lSZOj=x=$;B?Flz zCKhoRl!0-%euNAVdzG2FPWQuy4^@zcIO@Du|H(PtVroG7F0*)Rm(^{~H^s%&I50PH zT*9p2xX4)fqY~LV#A1!wX+m%Z8n^d#J%9i@qF=4jFy;6h_Ix`W@qOLxGS_TuXj?mh zR1nofNJaAXuZR1X1sTkBLUCl_Q87%&%_N}IzSp*V7>AM-POa+%ffNfNuWTv(qi7W`lCXsJ#x zdfSz*xV%3y`0K0RZ+ti}c%g#UddJz*_uh|43!YlA8TC1>X&d%7`Ge4v&l0OMMk`kG zV_j6`Nb%|0UsW!(cubX`LLarCEWFHz5rI`n`lq3(dV<|dSJxFo%67EWH8jwH5=UDO znOGh4xj@_KX;rp1K0g;R{b78##I2Z^m@`gDsegmU+1g&Dr*~mU=$g$i<>acl`IhpE zK%uSJ`G;Oma8 z=6J}VSMg}*EB50$5U36G8By|vk(tw|$)RoG9hx()BtD+|^}(KB6{3~)w=W_Ry$Wew zRXvQsc!w5+m7}G+4^0XuwpKOip$1Fp`b7f$uk3z{3Z;J`HtzlNb z);vG6ioi_e%%b)*;J!)#%9WZ&K@wAM1=uJ#((RB_MNa?I%K1P^lLeEW_bwrM;xYWl zZrqyHuy<=tHq)M~`cx#`{ep+M51F}c^JmAffnwjomBPcu^b(;V9 zNi6;_{sGSfG4loFAvh3h=psC*m9-Z7l?cDr#ba~`n;3>0-*8YiwqUTI-z_k`q~=?@Bq2h*vO^&m`4mt8l|`>dRd(=E zyWbI9bP$J-;im)**xgy#*{&0)obT>oBK3rHnkQP`xfXoK_7^BLA^#^ihj;nr?b~kwgSvosJy{20HqVk}?}`3u6U;B6IzV=J@vR{Q=$<(luBR9@JV$3BP?2or;y{jQ~xC zu0{ZGDiYbZ$ja_U>rf+*^PW6fpX(xy6y`eeBl?;d_~a`XUmd1%Pop(|N$R~ zQfXgAW>AG^9VqUiWgag;tQEF7X)$Xj#+W}^A0>*L<~WX@XJ!N%mZ+&nR2()cLIq7< zCh{BoW8-L;K7r+Z;N!D5t0&*;b^`2*E2@&~dWjc+kcaot8KNwaA zeWlKg_>Kxi#n)??9WN+p?@8nwQ>ppyk^1_Z!o5o8sFiG#pSj5K)utrA%@^Ca*1*i| zrwt#k%!h3GalJ$E-r_Yt^4^WA-CY!7u;LHXseB;2Gt1ju_MSnjD!<70|69xR69Rbc>c_}5-xd_CL7=sI>L}C`ar|_udNYMiiTYuaJcY|x zyMFzVfxsWBfQU(Acwl9ylBoLXAGiZA@a0R}?A`>OdNB6G9HQhJYO<`T0L0RrF2n!9l@8Tmx6ZIcGBr zt{lRo?XZXU8bkdR5$hj9gQk=bjN+%x&Ku}1?nK7#Roe=)j-cgNn80g;wOmq0-Q#fN zON)ynSr2rX{k$96xgPMEk1kjIOizDjVKif zU8(2hPXl7-S~4l7{b4mG)uZ9rbWOQhXx$$9_P!!Y3GSCQ&W6iVCe%H_^$plc4*P3zMe-Zug-(7P9id z(e}m$9WmH~=e}WNiGL0~i+fuJQ4=RY1Oy&Ev3xQE;dO&QZ`txYyvZl*x4gmPhuoVG zB}rxf!A{{>R<`|S{(AUMvGi|-2c?xP&O9^Qa4hCS=X+kpypEK$o^^W`L$((+_&xCo zE3WiBo`a)U>mZ zP;7hEgXE(=Nq-Y0)$XGt8WjwSi7ic)qGnQBPf4-K+wxBo75Jp%7V0$OVI@rpYn;(Af$?ORl+}pZ3TBL$EV`GOI>NJQR zDpc1|%5?*2>ZpS)+&^NrJYJ|yk9goX>0qQop3xEd4FpI;edif+QQ|ES$gS`%>ISCl zVHw(y3d67+2R%E?P7~U(@ZeHLTHWB&yhEeJORK?iQiDWjmY6VPtXJyd7(q)8y^EL-l*uH}GKe5@ z?0v5zCi!_ViQXZo!{Ni#G&;Qwh^vkRd%p@;p`8!83W;E_iO10B{R|!TSFNndEP~`C zF_()5aMvwxF!pe86#G|c_aoA0itJ=;7+e(z$K}))U>pw$V(D8yFK6!SI;6QHPx$`a% z9|@A2LVf&R8!aU7Su24#SiNCZV*g|}x{iECz<;4bGz!R0Nw;2-NRfeCUf z@`F^X&5q)xCI-N_Z*b7}#v;w$K|Ts#icp}FvuFf415yG@^|ae{bo8?R$w zV_OZZBM-Dhl8(?Z-)=|96aiK`f`fG~h&%ZZNPupehmA#wgXDZXM^*#2m}{-L{p4CK znS_GE{jQ}55cI-SwIglEI=sdS`A(+|Z+QX#o;L+tiO~K?$c>6;-XLt=OKfncwr_`I z9iNyUJ#8c`z`ZCwaF6jT&dhuG^kYo<`4;N5PW~8k3Hv( zP9PRR!P$xhJ!}+v6ZDF9A{?tvF`@-`0+PB0cdBlWr=Sw?2|L)wQuh1n#J?GQSWP&b zvfkMYS{j>%8dGG@aXm;B!-rb432{*ZR=9m+WcfWGD!Q}SYcDZqGf?o_@$sBQY=Ip@ z=3!xcW-I`}2#u5|k&^(TmcSpVAJo6P#sWq7Pu<;9@0ZF*w)*txE@RMmP`2~xS-__u z7>tyWEGyYxnO*jup zY}3}P{RaFgYWL9c^JsDE?4HB90@_?#zXf_cXxJZ1^L#vKi*l<yVtj18b@49__pH&UEEvv}8uFm&SPy_JK^sQ^J{Jcv2 z00^e=s_$U!&Hz`zF*VHKpD z*u4lai^Km5=Dw}~Z^T)kX8RL@SpZ_suy1TZgHU*S>^l(Jb{#md4k-#)6A~I4I|*qS z8V5Iel2cRPmoEp(XdtcvLke~pZU~2jXaFLNEZv@}w9fG;T|u2 z|L)zXxYK6Lh~WZ79r)311zUR+jjl*RL($g2uAOs~dOPP@@ETFz>j1h8Dx%+HCWZ9o zzefJMK`nq8X#+LEH!zikHD1(Z`6#sFHc`Q!{0he3nXpqo)37U0$;->5>_=7`2;>tn zr#qOia7ye6MH5!@6A<9f^Ryhw(Kk^-X-A2{s}>?67qEE50N^8?0SS@jz`6Cf3aa6! zX5IkmkzIfI%O7s zO@0xeoj+zohgqR!2VB5SEtU{M2}=V+gSYSB_f&|=fg$x~Mw=LLI5k*ky5!dLo=l=! z$}?WQDipk(g-JyBP)>Q<*2Zj~tySHKc06REs|hp6+xtVX0~elgJ%%iye2l}0lE^!F zwVxRn4x|^fDTC|=f44f`MO_sbA|iegcq4jotqYHg)Illpe2U>V@b_8vRt-S@aKh=SY1+ z5c@yYzpAMm_lQ%23evNP$ zv+O>+`>>=7XZ<(^C+Vg40bZcNP{`Zz*pdIQ3R6ay=ZT3N8mq*DE18G=9ydXL-FBK-;Rbof*iPx1HeF+#`gy`UR|gb0+Q|yOJ1B9pRgC4 z7?Kv?xU8;Q@tOq&XJA6TlGl`#VC2lz=^<4dPrBq~WYe?A?C(B#+t^5t`1%A=g?n+V zz5s&ZUB%lkD!t6RbQXd^#E5#S;g)aRySD+}K~9kyOlAzs zCIyDQu|}0ozXLn-H-cNCG4N?CKCPh3a@J=jswlD9Ppb&52w2E3d>M)uD8{=2i@i}B zFp#5)XG*}1(DJ1dbFK(zK}x(!hEG*zpi#LUTHOvYqg#8{5DT1c&Pv)?7wKQOfvL)Z zszR3q?r0;b<-Y}UmvF& z)D`Y}7slg(p?kvrsNG>p;kT`Q!r&_+l~}uB1K*gX%X1Vxawy9V_4W0IisHEkG&7En zvND>ickSCJIc_f$mq}Xoz>a>D%=WXYM>}N=)i>iO)tew4dAf7(B&X6hH8!$5^IXNB zocJq+zV^A|J8Ei~O{emhA_o)Z6IaLD=ypsnK00s5vu2axh6Tyozt8=@eMjK~N1yno z7qPKi)os`uq+EvzR#~-K;r-n^#GQ|V8P`^D4--&-Xmg>6BC)s^&5JQ3D3B-#wgCo0 z&tAjxC-I>Q?1wN2SIU#jCc)py;uzEFOlYm%# zWE=??Vgd}M&{*`>{WR2sA%x6;&&~vqH(B=9)>gK76KX&=`4}5Z%r4?a8^BnIf2MeFsj?KRFJ=3lE@)jShXWH?`tv z)pQ1MYF@QjT3WtCw97SZQ6WQ)$>^wAqCMz{$YPJEic-NR@?8l6fgbG;tKh@E%XO$p z0fA6P7bY%YHf|q=4!sa<4MUVRiwe&Hhov3R-c#VMa|lF_PqVkEi_9SZgv-it|SYs(wx@2P++w|F)_bY z<4aKI>%aBoMk8kf%5#%6ku78(qQ@rwJUz{C)CNO+5#1exyKK}ZfM|XbhMD|?-4!|j zIBg9O)NSP;L2So_YE%fc#1=nGtjCx*L8ObME(pZFSkO1iN17F5Dej_SPBqL0!3Wlg z1fNMW2h~$K5wfnfv9U>-ac0EN2x2dTf8PPWx`m) z7GsMKqIGa^aKr$_&*S5V9Y$SHhulo62e@7Tlz3m?ZN!*@$oxIkO4bW5oOeYBP9ALD zBd6etIyyIyX*@-2NJWL-;beJ_KezO?Yj3a?3Ly8W4ltnPa1lvAskPV*1N?o92}PF) z@qE3OdZaw#4?q6c9W&|#bCF=tE1#rk7lhCswE??Q0(|4L#)pMFNe2-K z^-@NN88DP$Z`P)ex){w&eyGNHVv?%5r3b(Z9hfd)ANPOleYrPW@itKSP%@V#AOsal z;-i7@=p@i-Ax7U`uGZ10uo0F8x$+tk1^1s>#wc$ioTToc<(5Seq#Xph04wxm{d#N+ zlJwzx{n2$XBgXlgNM(4HxX33|mfVWqwatC~+fB4~tW3lHLYE({ugGpLz+Z~Y) zpyIi^^TfEtpXP`Z3X-PKCAj0rmCY*-tf?s=c2u0?QkV?z`@@G1^Yv^)$i4)Ih@X^J z$HSLH3wj+8Uo$`F__OGr!Ltnk`lvTN*wOJjK*eYH6}|pQ#VA5CLxsAKbBPljr3A73 zJT{haih`p2YXr|o&5xnIa(QYBV-VKXG~~MEMoES`8Ut!2ffx<-T$?-TN+U57LPjvX zE|2OP{8`i?+_{Sg6RT9wkcyD;I(F!5)QM8XfHl5Djr-HU0EfD#3!-Qm@Fxjf@^P%b z7Jm69hG3)c!f0)6oj8Np4MjLi`$ULjP};zZ0<9cJyP@Xv zttfELPK5WIpEJ!*4cNzQ{X z0zD6-q-oYY6Ng49t9O0tKssD`8BZ07!|%Qz8~XX8^k6xqLlW4 z>4v?OXGAb|1ROeS#x5=o@K%w5q2eH6-}o*8`lAXL8%Ok@0b?_{_v-l zfS32x#s(ORj;{lJ2NES|b_T$p*Kc5ALU{0t7%$idKc>V{bHIYA*ehg%d*2C57_sxg zI1ehHr1KT`L^^L{JrP4pqt_50s-%`B7x@DNKr0f%B+Z-yQ4f?=OoO2U2{>u;g>uYx z{N*}wEKu7DYbC=KL*O9mp^2Vf9%|Q*#B?LG*n0e~>a-(n9R8<1&5Rizai+SCcnU*= zTaO&$XyGm~)aiHV4ZbNUk?~`~(IN%~iV`78o8HIz+Hg zI4cmS_+%<_QG?8TdGYOL)Rb`fl14HGwGVo1Qif0(*`X-`>jBV?ygNYX%fCm3jZIFz z1KFZ7*TuzMv-cc!G{NDaZ}=mge#89i>@4A|1O4%hU312{5>C@`(PgN`z+Y4Mx6xrr z9AdHJANKCOSSb78*sfppTs*S=yY*Z*>3;lfYBpLqUh!?bXQpGce5@j&C^SuTnB%)C1nVO1@iLrQ3r{4uXUaq}Vqdhfm z?^K3%%s>=5yPw|JI53rBXk=u8ab>*-uqz;^RXaFnMXJ1+qSi)^Wu|iQsU>we(<#lC zALO!^g(glbD=T-0dCM1r6Pc~4rP3dVJ>7+gQ>7TblBS%qyis(rlE^V>abILOH#hp} zuL0^SQ&QCBuoZ3wMNcFwNzWDiG-cPpaxZ=0RYqXuX}Ob|hO*b}Pv~yFB4=uw%A((# z%HyV87wBS_%*UNmtK84Ia)tl?=X$K(>+9dmK^iNFs6Zi(8ceXWX);N3gjjbOX39Q5 zd^S^`CEEV-?Zy%eseFV9(U~}M((!RM_4WQ3xF!JGUxc~pa{yIeO7<+ZZ*uVmXCter z6B{+?w9~*@(L`$-_?&~7<<`n(G67vcA#^hhgg=YbN6iPFXz~X$`koD8*rBNKOu5Kt zV1?6&#UnnWa~p;Go-zex?%V0IIQB@?!TMm%YzLyO)Vewbi1CPFBBMS8rG*AxBq?%k zcW^2bkIcvr#*8=lkzC5MBlJ5;k)haIWn->6U+_Y6FGo%rr3Rz2gE4A#f7@b@059fu zFa@M3adb;C&nplG&7FWpMfuINuQnP)c9RFLsvqR581?ah5xgWjGiA2GqRnZh8$%el#0BW7eB2@hkV3gbY3f|sLTx5ldS?|;0e-Y5z4G5#UG7H__0g03uSq( zu`b82$uyG-WA-emU7t8SQNc#?5bXG)8LA_5d9*qA5p_@1H)j(<#2jXoKEa46UIZxC zaNo)K7PU1s-5?c|V?lh;k|hZIj$)>xrlyv`ry?QocKb|L5Ms=p%(f*{Q!xJm&5QSo zCiZ|=W)20!%YYQ-K1Pa7>9-l-1ZVQ8I5{nmmkf_h^jvgf^uvb~C99f*&XMVPdvcJ< z#VL!;06m2;sVcR&%?KM8iep~n0po17JWEG>`%yyK#kOLu%G23L-7K4L)S+#`erhxq zkFtQt!Fc5;+I9N^=8%n#tHq&Pazd%AFf2C6FfRguU95FwsjYC^U z!NB$SZtb~Z75=%v)bMat?fL45_ECllXF0b|+&=BN2X)XgI{m<^(0C(C2xcur23f0N zICi-P^g9|1B7^K5@0sM|**_tVI~@bg?Ynv9@h0*@-9F8jnLpQ*3ugG-LhF{{;o%mh zCiS|v946QmWtdjyV-%OphGcHAYNAD4nZp+=7}S%J>;{n_Iyqvp~(?i?_$QC69nPiIt6^`(q3{`Z{X8wH*Xe6Blr68%|DZukDE zmmi>Mcfn4d>3r)0w{wJj!1&JQ+!zm#?m1GW@1nNDhXoN;XilPbJx;Dc zjFjC_LILhAjU1H*NJ6t!V0iWCK--SnbD!3T`y=O?Id0O10{?x-ei`jz82g9h(Ac!v z=|dBufondEh)Bx@@NnCPqD{{ zM63pnnj-KJg8mLKSd2R{cs43Np5orwcDvw$!$Ov-re@GqdBNTL0^Vk&E(YKJ~GFr4AV8oOhy5gB?Zv5D5K ziG&Y3?fXLPz-O7jG`S<>w!s)^rYG%tqobrf=#_)cPQNH}u#T6POL2qF zDZVCIDf5=tV@vL z@)}f!$p#AYpoc>=L1bGJ956r~@zbJHN~oqh2B_U*le?gairqs$KR*OBBbf-E2Zj{K z7?_$um0MYQBE!g5hD#t3|InMWa31*h>E=me*?sdz1o(i)9heD zk^tO-9&tE6Yg`&RIapeL$tfw}aMk$u{g?%Y-S34V=(@-R3=9p6fzpu6N{_zR!(l1| z`zHufpAQo)%OtESqHG2`%<8OeZVm*8J8jwyj2`m3-EKR*P$E$gUSc#t&PdE^A~s4E z1T5#k`V2-Bnmz&@10y4ubPaj9EP8tSR{Qu?leCsZ7_}bFQF|24PdN3Ad*=akkzWxrMKKiQNXkTAnAyVih0AOjY z+Uja*iX6UhsG$4bx{QNPC+0ZIR^~7e71;wLVBd5N_h<^Oz`U}u(VyY@R6$-sb>j&j zpYZf_As8w|TT7vJ{&>A{b=De2O9cqBFQj&o*92l+mj|lgebmA9>NIWSV6~?kjpE@Sbpu}M&B>oR_;3PC)eF^XWi8uwSF~^L#f>`- zYD9v!h#Q`onlf|pgVoR=uOknaPT2BFxXo`&mDSZ=m_H?qs;SGM(FIhu&754~06O85 z;vhxE#FPM@TKS|%Nyvdf@h}k*=fJ)rsxuSiAlqWX1c!4J(hVdQ!?cd++8F@^@`?!3 z8XFr)vV#3x0%~E~{G!OUj(1|$%3NF1VkzQ@NVxO>d<3KnqnOEyC15%j{o;7K#)!jd zF_LH2Q)`_1E8n~>o~oxKl&jJDK$~;4wBn(So*eUH$WW7e(7-yOgeZUS8BV=C`5NEy z6}$bEVR@=Q^~Lawy4F8UaJaPoe=IPS=w)@vFV<9978;Q6Oow-Evu^%w&CdHq@!8e~ zXz=n%6-J9v9_-xyCXY-g`ivt~zjT1c#0fx+tb@ZslgC1@%W^LXv(!V@;<9}usZ$h5 ze)uOq_!p8=84;#{9$EADS6}59mVdqP%Ks|`hLp4mOCuha8b?d1qz6Wxv+2%ZWlVQb z)qkSZqTGLU*`1G}Bz>T0T4U<6!!(nI@7Zb2o|~eEi8@Xp)dpW&>(2}75S=R1dngi| z=YFX49@VMp*4Eb8gr)%=@(ULxwjKB)dZ(HcRdzkjHmngOC8E;%`zkLLe#`D<-JouX}i4QlDzw(e7{3&*XN&WqM{Be32*r4UGS^J543^= zZhDVV%+Z#+PU$_Yb@lN7aTiV^0_VE#{-i9Vb6ieMq{MEhySzC!nRkY3xG+(m#J8M9 z%0cC@%{Bd=gCnK#R<1j9C3zw}xb@=S)cDG;Xf)VLs*ayjX%u#tZnTcdxPN95i zZtau!)cd)`fgbZ)M`-YqI`V}6{IVxwyE869I{TU! zrNc;*E=@w&v%H`Fub6|XvhFCXgDUXylO8)CVSu_W1_-~ftY#*@FOx40J{w7GiC!$a zaRTAW^@!sYwYMapxa!2D-w4wU7_`XqoYUqjPs{1vfs-qr=s*8C7LS4|U8clt9UJ!U zP;YMm^^&>nJLS!3VJzb6wGUFbum{rlH0K&6OwuPGq{=cY2Yo20>w4J|b5>LNxh4Jo zUhB{Aw@rd;*He5T-f?gS?732$cV74&@p(Q3)*tKY%%*aXeW$5+!rt;j$P3Xh02cf* zl7vo|CJOnCFy{Mxr313ddm8Br$Yrmas3cVXcTvbsUQg>BGud5mkE`cZLcXQ7??iyI zX1czVWDnO!VTVAmZ+TOkl_hHut(D&Nv-0(3oLwKs@&oyE*fKXfkN(%&MsE07GS>yZ z|FFyQA_69g=kj7G^U{^&cIM@aMiIL3?sL#@#EAl{pmE3wE}GIe19||Nyg2c0>*>X+ zV9eIb7lP#$a0n_(KyXB?cnf_d>o-3{R7Uju{cmG@h&p&+>9>hsUC| zvP!Y<#%7NtQwgDmqODI#n9VBQWGeSgB7^45@Nc~IuV%oQ75Iup+Iymqo~<$7)`GK0 zvWD7vo9sbNEUC!c4h_tLYhn84T~SFLwswoY+Y;h)dPt>speL?5GmIB^)%9jukip;! z%a-Y=K=Qra)nD~j*QEBmi+CK!AL9R$h_cGRn1t(E{D8Df?>WGmPM`*@2Xrr6V{#41 z6cvy*Y=}+Z{YhtefL@cg=d~BJP{#l=rZUR}#p^BRJp4sZ-8Jx6ijUVw2$F!A{zqBvc zbHfrM?$3Da?+NydF418&EZx96IPt*A=4=W(y zd*@z}HYs?s?Xh)?Z40e4=HLO?I4t@xtr*u-O=aJWe_tNo^~GIM-hI|4s{B{Ck3Gx0 z4C9nwt>AU7n#DJRnmLqzVi56C-)`Ntl*F$$zoBey*{m@!BR%F{SNA?tY!cNw(K!{w z?L6k*j{!X?e&8Z^%V^IlrWJV7@kTT)xfJ~@xw9(v8i|w1%5!Oraua-#BuThU5|?LB zH^$Jqlbh@-Ioi$qzjxc`4J>53IZC+q^#1O+4-DpQk2cJe{uVxznbZ`)adC{HTVwu| z`o`>3q2RzeFUfP(VHyWjg#VOW{NL+b=ek-gwJ$t6+M#0RF~NObefr%PvJ{dG$P&yU zAl~a8ym%Q6zM0uAztU>@AQj@$q5#+&nXV@=w)WD1%7@n0oczIM6-2}7P#&`=Vtp8f z=5O^T8uxq+glnRW#KOH!&G%mVUsHG8B=7le%5(3W!<((8C7j3J+;VLHq&Urd&i8ou z+f?3R&7vS~>#ou!b@RzwzTW74P2=`5|GJxvZy^ur22ujx3s~i4q$BAFmmMxIL1AhS zfgFjvo+Uz8hfEz0AvOXMuD+<@7<* zAcfx)KWMJZbnemqYFbx0nH+KegyMs>>6J7%B3viRsdM986HEQNwJ&w zV*3!lD50@;TP=_$kt}MoHWh}6q`imXYny?TAky=ND##CE1xuuqO)%J>*4^A*PSQJDDrrXHL8*@as z^ba#j6Laz9*D>9DJ9zf2e9*CbX2#Xls92TgKGgIpIhe_bJexpcs1*Am_ltySf*(b8 zUXI!}X<6XJtW_XBj>wc>HpC*$CIIhYix6t>$=*vC}-F%YJ5wIcr&0bFQlSLG!GJq+V+l53iYN`Mrvi$nfrQJB3{;b#dRp2>X9?T{**PP`9-H|C+%+@9(n<#&|n+GPD({xsScFV&$DYBR_w!ev8`c4W1VZ zE%=Oo@TMOxNn$W>|0R5g_nnEV`rCNtI^W!&CNG0JzTWYIrWMB~_2t5mzb!H&m9~eE zio=u`US(x<%A5u%Qd)eF6X;F~UNQzelQut3?t3tIc@S~M0m+@$-ksCVRbcPJPAnzh ztf$kE&i?K0>e*xwuIE&E&wQ~O85bWqE6??HDe46^-d`v;E^K2|kpAFdS~Xyj+on*S z#$;4-v5vo&lAXqyt68JkzxSWF=>>uUXWQ(10dwROUCAHhWdK{n$Eg6XCJ;SQ9-V+! z+ly>>sAeid2N6U1c<})Xgvfd@tfY=uGX*;!k64owp;l}OVm&R*P2`qM)V~VZsW28F zjx1nFhWt@|6~Swh{qcc_)=*(J0(SW(y%a0fG|{#PO;1J(b3|0S18?)q#azq(c#f7M zXvL@e{Sl0xVFsf(rV%@b^Wvv{y9q=H*rUdAIw`meNLc_0BdR2s%a8`A+K$}1b&DYD zLW`f@S^)glN7V~;k_j-_0!JNtMeN|fc~SX7VMx>t0-OXT;(FqMi7i51_azPk)h7jz z6)2mAY?fn+UuE;hv{QTKYZ@#f2Q)s9DE7pU6u*E$FK}m6O-j({Pd2GV05?7}lhQM| zDW&^v_8lRk(Z}mYRsK=@ChIn=2K!^taeH46atc`$6%|r6AlM8y^xn}X;ROX|Rss`& z9rQo2DKbIWa|2`JQsl9Rk!ztW%}>bXEj9^!)2u{tjS>_s5xsm>(NQ?jF*^+s!Y zj1cw1Ng7_06Fdx!C6Cfxv5<|W9oHwTk`9miq}=?l!x?=&?nZx6n-@6jA20VXR7zT&l@H7nJ+5i{QIWH_E9dM7IO-=QM%9#1btb9@O(y%de% zpcx-&ubFF1&NymrsuL%eda&SboP1S_w&+NMO6K^cnY5mlMgbGzE;W8wSYP=sS7znb z$$y^z@%8P`-e?>+?!kDrG(qPM4rZW;)|Cpbbt2=XF}O-oL{>&ZAZ2baRb%Xgx<^~6 z|E5uO@9fno{}ynpyvc=_{pW=L`*!4;>lD+-`Qu{AX8V~N(|mesEqiKR8PkX7JYF=M zkF#v#4~;#Lu6t3^RJH!WDkM;DGqFVv5HWlMNg-D}7SHkZd*m7e-ki&OQ!`ikQ4z)V zN9x;Cds|t93`V0Z8&W$TJQSHIED56Hd3xVsCaNf^)2KRE>AB^0aoZgu~-JJN8<#g;wG(Lh(LqNA<_nTD5g@jaO{cOD>Xa8s8{__ng zLyx+C@r0jlY7B|HZ^7nwc<|$`Qn64m<^T=v-5Xq`^-Xwd)Ao`5>G3K~DA+%nCwE>` zAC;2uq<>y^Z+CZ$o7 z93BLkVM)k2zTzJJ^*-0RzBcewqM_%N3)@-Tg&Z&Q#Hac24!Ta02ygp~ezGd<0x=op zt$}`b>CYEZ{FRFLJ8VCz!m0ExiR8*#T<}&_{%Yqu5qOSOfhCOlFksQNe`@ZvR{rDx z~jjJKO`{D}aT@5UEB ztIII|@cVRg#!HrXQ%)7(*waS+r{bJ4MC(L9JR!IS<&lOraOQ6wrcV@<7{7dPV<#P0 zo8FDlWjFK%x)w%bOUPjX>ZsgSt1M4~#CabX@2kq+ zDVpKm7C%}HGhjaP`!uIh65O@e@P7`GL1PF9bC8Dj)@k+#CCzUiBCmg}dKT~cK^5QJ zhFl+q3gbeneIL7Tir(3;tdbFJ=CGU+=h(AcE1MZ71Q)1Yo8o&Q=!d^o`hUN>|FsFT z)sNz!F|Ulv-UL6LRv%Q_QhsTQ1K2&_=7+HX{}%8#7>^aUT_&msSY zzL2+hZT*i$&$1X+P$`t$=W9N?hJM_;bI`oG{?eTh*(T$`MlS+EUKDS*FXPdFo&De_gMjN@dKCKJG!MS;`aYt~#x!vOL@{ zy*Qcf@xov~4rhhkkd3C@P@JFaP7U8&XO;a09#QhjAqwBJhZh(WzRKo+@35>^>gHWj z+w-vZNSC(9i?Xnwmg-cYC(bDcTzJ!13l`NaOtmVXOqZMgTYgEFHh z$-)$KB~52LYl-EN>x#GdwsQo1?9Ix`3FH>5e9RJWR=%MkH8K{5|6lSXa<5N4qrO6! zshWb{Z1^*qnN!`6zUZGJ+mb(Q8Zgj3)$Q2c8!I$$t7ow(Y1m*C=I|Tmq>tt3?k~i^ zuiqhd2^EdzLrJ@Bg0zA+4JpfNCR8gsl@wcbYs^m6?zEDyFm+57CHe7Rw)!Yh2rqq7n{PMUS=+;}a@ z>B@V_YW&Yv<@;{dFZ??ddK4XOiadkYAGrPSVBh9CrZY54!r@->jd59N;X-kb&d}}? zQ)KLop8UsUDRuJsQM2>Qzw%5qC(aJ{%3M0%wixD}df-6nk%F3v zn*847^NgMcr6ncKT71ht9%Jx+_+_zxYLR!&E=tfj?f2OhT4|l)blq!joL$eHyKh-D ztuxot^UBa(({bd%fUWTvVfx39y}uQ0HvW+n_GPeqL4CDlxjTdBK@VDPvJ6@@*&-2G z22;+a=AA0?@$MB%3Bqig^8e@k=8b!OVd$u?%UW^B=htmwztgXM&^=S0Xs^x7G)@7dBPwq3TFxJv+%6{bG z)+SqUt1&F5F~>%LufQhaW4^!2hih{ozh2YGT|5=8D7Aw59P4<`EnDE0Ql{Es(d(sj zw>k4ri~|YtU3}dOCd^#wX@VPsxMLjbJ-*yJS7fNaUM}Ol*4^AWHm9y>EH*r|?Z{>5 zu$lO9COu?-{|kPz_W6j}`L4ztjV#=8g2{p#`+bA|+>m>tQxk*?=^rO$=o{#qOCt|< zS^gkXMconGuimmg;U62dK}qrc9d!NOrO>O=Z*or9>(Ry~-XAucm$m6}N8mogOCQ4? zpMO~^!u{`w%WIDpyZXQUT!pR}bs{;v{|}Z0ns;9JwZ*jx1_i zTR+khZpW9=&%4L7wc3W4`XeH~c}36m=)o`EFza zS|+^Z4ZW-7exAvF;Gx6ZNNFpQY-{IK8XTiu2sc6y+hV91$Az?I(0)=AU=kqNO?BIlWcp(!hWAQdf-w z<$|)Zg`=6FrelA?W81OpTi;sErN-c zCLSdbwfTkOHSZXV7`*?arhV~P$X~8Ap8MtNQ{PI{72a`Ibe+z-*T367OWb|3BF0lU zAi%rakhjOFl1AJ91>3So{O_~NA-`6|a&rH9^cx>AF^)fbu`9Ep4#9%1l{hZ`$*j9$ z%+X2bM6gH0lFfl9EPuZJuC3q%67G=ga7x`J(vpgaq-cfK0qdavd{P$7XDlu$5ub+= zU}PWYv_qi(F)t0EnUH4tcD5zS6=za-8|?5y7}4=F2oQHL`J0S zvXsh}EtI7cO4(Y7WX%>?BKz`x9(QJbbMO7m{ChlRrrX^3c)wq-bDrmUp65I{D#2Xq zL1Nek8GI2I14W}X12s)Fx-v@I+MFap##Ag8%e7y(F)D=`wGb-@ny_Nt6DnIdeCbc0 zKJACi9X1{9hv%COc7Bp!EVf^IW8F>6jdKPN-_m9l>~i~YJoW7KzmMJq^U9U3HYeZ# zW`K>yYrNrUTsmjD%C)}$!Zh_c0EE^_`J$$Zd&eLB@yuI7Q^yK@zs9wC+S@4Y8?TaB zY}v6_rtkAhccq}7#cSy)yveyS`do4)erIAXd3QR{UhJY7F&f)rEx}4~(IVmU-0X?L zua;;h7h%s+)Z@oL^xE)YqwoYse_PRm4*VfkL2FEGyoLfgI92Ul3pV^NtHP|-&bA9f z-5i+Y8T-^WtgYaf3MT%n(9nz7OJ8Fg;ZF1F85ai|(ew;}!rRQcJ63AqT$p7cYl32s z1jG}*G-4q{e(XgUuDI2_&M*(^R95V`@$>ffe)#a=aA=#H)Isc|)Wcd2%P*&|7C_s& zl4dz9JN%6?D%ywkNCBkEt1$QLNpj`1MEvdTn9vaIdh)92i(Y^q#!2M((w((+gu)f% zWMpJ$0)S+R7zjQMZTH!9hk<@TQ>XNCNJ(_Poe5qpR` zqEk|KP@^nsD@N@{2Y<5Rc=%Cc%3t!=8y*1xbN8mFG;C*yDMvYF<&Qa3@K5k+Pd>x1V}0h1EQRN0FVxl5-&ZS~v|G*OoPv1Vk%#hf;{*3c?Y}43Xr140 z$)0y&($2B*Pk$Y^GY^w(ESQD}XvXs!el4PI=0%eW{9|;J(hTSkLuzLr0xn<;G`@v4wetSDJu(TYHH%)<@LfEuP;y$L}U01>vb})?%NR!O#$qDSr#$! z-XID-2OTDwJmCsO%13vxz>>z0wDuLXwZA%Tus>`&R)w-+WvwrbagZJg>ov}V+0~q@ zfIgUArRO*v9(`0#jOGbxL-KVroi3Xm9@%#Ex-y{#io$jVoDhT8_!zX0qRh(D4EhL( zJ~ScGe3E3uF6~Jb$C!^D_E_uNj7GmNJpd&voulE!CWws#W-pc8_O2jJ1iDjC5^vFI z<#OuX>r*8nEL;GcC0)er5D)P1@p)r%un4;-b)ot%-fQ_a4Jz$SOtYm7_KSbH%e>() zh2(fZ{`fTbSKv}v#>B_gl<4rguP!)P$KteH8X=Scopj5?_2Ze7Z<>bRmpc3{R*xUT zFduV>meoP5iCJgtxMIbMLQco7>**7CWeqpz51*P;FNYlD=m*|C5)l2fqlZ>eUcL+> zm}ImNJw4$F(jgTtz{BH->7(^a&<_;FTuw6{kS}75+U5t`h+KLY7TIps^#xnFRMm#G zaoDM8?@JA7lCwiAv@i;%S7Vo+9At*e(3SI{VM-D{ea<15Y>c~|rz#9=!G%>Tu`wI!*FU*z5g<1e>tPhW4 z>@xrgr1+mFbac4TzV*dsyh6Io)U$`(fQa!}hAO<1tkqo4&il%wKK7^f>cyQ8wn?!V z^qgF!T)M?4VhO*09fvaV7^|SVP+?n8AG+2j?F9bxx9~*EHPNbEMN+5K$--hx z-vwY%V0<@`AFfIG+}KgVfmOo+Xg6|)&+hf2C=FSEA8fn|Me;^UMA4pYeh_NrjV&_V z;f3|LrCeQTj>{4q=hlJ#RMClDJytkCu&Pq*v|?~Q8i&U&YLFnDBOpI-%yCv zgQ|Wz#@y)Nj>b5pfk8SHAj;GcrDnA~ZfJCtSVol;mS>vLgh}G_?q?4lKQ8Xj=o01Q z<7?`Vm~z6{Xr33IVpurMlMQ2(bs1)wyf7bu?}6a`pa8)r*kUQ zmW{mgBdf2(seWprUJu7a zqu#On-d>G*COpE)B#b(eu}m}cvcBlR0`zIdKYq+NiL>_l;HGe9(q}=~v$kWyx~7c0 z2~!%Q=!4|)Y0Yi)i@h*%Be@lt7VHr)jtl>)HbKY4x>Rga9Y%H4BvLVW1Bp!iYfOvK zYe}%{E~s+haZ0mUaTl7E5UFUXxJM(9rNe%xM@RTejXdkcIQxy0%RTXr(7& z`ZRN-HFQ0U9zI|tYlks%ijY(7Y(WyBN zmO^yM>rUqP+cn3t?(=rA|LI}OxESyo60HJ^3Ryj!y6f0|JC!6Oh7mZrx)4&+J-CID zw+~`7Fr)eH8T1~xjOS}DSO94vqw{-K_?8iShJaS;U3`GX)7Kf9x2}NqMd$j>PokMZ1i`MD+e9T53D~4 zn>zh8*71S#d}H{p>(DF>W7`0SK<=>a2i}BM4>mK;>F42_*GJ=*h_;`M+7hr!2(mgB z=)FzT1vBnPMy{nNVh^|xs1WJ53YvRKgs+1J`}fmI&6PAMYEqNN+Y#jYmj0Qd?FZKK zjekpWPU85Yx^TGg(5hLska5 zz^YrijlCf78SXxx$A&e_EyzL=8Mkc3&j|_=P(=+rdRB88q7}k9^db(%?g?$+tNYW7 zd=lQNc+1;-@fPp?!7aXY0P4X))Og?!q?;&P`(AjqQ-R|eR~45A`>uC?C--WyfZ!2o z-rG;281U7Z{1sAxckMdb2c+rgi6wC7r*%>CRP>D}0oQPc&$#AEiF!MR3`L=F3RI{* zJ@G3f&lh1k-Tox7Dln%oEy4JLyD-^t7%RMKe|j8^p~3x+9l$pHO-Efvm4kwVpFT|GMgV6*%-y1SBO-zeiU6jAFEA8s5_) zFw99$$8yZn_5FxTDnuxd!oCGun*tW~XrV>$cVjP{?nhl6B9#RQzH>mtn=6{ z_V`q`;0mF5ScSbQPSbdjmV6+kJk3f3G$3#rH`qEi8%W4U>|9)L=hqfaX*Eq;*Sdmt0f8+zz2-69X zwfw>Nzc#bZs1Gf1KsM&y-h!^vhkKU~aB~Db`QhO?<&?EMci8aAd+nPe1%Fl7q?|*& z%vz?6Eg?!+0rhG%y|6DXIJ+|VVAw**PXqd*9WghWq$ZoCnh4QB&A1z`4qUV*gsO!I zqba!mtTD^^UAb8u=#>fM(zZgv!s&+JFk!ozU_B@!5O?Oci z!DgVr=8;!Rr6fsJ{RL-4>i2EnBg@W&jKZP3sfh#kcjW1IVSf%QUg+qjwiM@4pk{V~ z>|W`@{t)d8=V$>8=$y$%rC4t1L#kVs@n$YLY@(Gl#=eba1F5m`AfPN;ZmP?| zmsxyHP0(M-z>Br*RlEW&4$=pSRRuwRRD6@=O|G2-{-OnDRNTKT^`vymQk74@iI$Yj)Z}NXho+P5w@z;`n`WwdqUjSR$n? zYgrHG2dF5dM66cY1Nw`?5v3yE){h-FaF`V5H<;4F8X5)48mwjT4-q8*z8@S+s?{XM zNSJBLUGQ$wy!!qKM|+ z?{Z<>8C>L<^h-}Wp$PT^J9buxhChbuhH^tb_sjXNVpP|%lqu||r6+1-W#k@JNP6uU z!FxykMt2)>%(E#AxGEb|ctvNk*Ti|#dV%e=$G9w;7yh;!Z5HmU`SCaM{G#1G6rs^s zfDJ8~(@LM35O6ewxLxeKuo1W78Fp2b5nf_fO&3_YYd{g~tAV5|22YBBAVN~_QuKG3 z?2j-Il0YQ~vhIz>A>)IwANv5B^d-UVRAIB#E~^hT%0V=cY$_%}`mh7e;Aa=G$BLil zvBXCPOhc;O0zE8=i@qV$6d!|qh`_ge^EeQ?@DZ*)H2RQ5G&>5?# zYN7>AhITko*w#NDFLLG(>IN;_&#D!em$zX^JNR?VsD*CTzkii4;i!&XLnlj5f!FJj zL`F}@>aEACm<<#jdKpK=`7JdDh5hF{JHzf&hmCl27Isl_7je^9DxV}e3?3wGh&!xF z7&`Z+w|-Ui=RV{cIF=$nVNGEn&5pf!_wJ{#;MK)!D-6PI)nrpCo-3Tte1$2BW<2Ki zhIWvAThivEw6+uVqiTg2MJ9DiRwDA#0H@*L#o0?dsUFmj80*2(THag??4OmCqD3<4 z<#yv(5atC^LSM)ccwZLJ(Yo>%)du2rUAHQ_Km*=L^>_Vz)aBO=s)T{_&!+dp1^uaP z-6L^R{(e); zrxW4Z{}4Txd(Nwc3>E1GY|-q;yJ>y+V1i0KSk&Q>u$g796+hn{_9`mx8HU;gFDTHP zgP&jpgG|=Pz`S=MIxY@N4o%9zIG_pfzsztjA*0S4D?Z}G;M-j5YXEo7+cyQJ-eT<$ zQvrO!N=mVGrARn!PJraE1G3iON{qvBSHw)A89(IId7{49&<<A4{Rb$@GVcxXizJtuefth`&jsAV`14QpGqu(wGPBe-Hc=e$62 z7AzL2eNvJY?Jv%^1i(d&=kinDzt#T^_uU^2SC&WJ8f-}!E_QLh9wGmXM=UAc%$qUb z^XsG9@jeM1t^a4I@8agS@80=fays3xgtkAPm!>zEim0#Nm@S{AC`s9D!#1*3uz8Yt+X~6 zGg3gJhsUaSAGssMdnYu*c~>n`rYQK+6OlYr4do8p7tj}?xaDf&cetjG3 z@e2@AYUO3hcbblA!|(e)ewCZqYNrvizC)I&c7{K zc*9>Qpy|JU<}I;a?r?-G6FB~4%~fHEr&t4D2$Eay&QI8$&;lf}t)p_ldH&!^?L)0W z{q*H86iAa+8dOzLIU^|na)dp2JON<@{5*~^N;wTNYhY;B61=u8{B8PC*bvNGmZ+U0 zewBNJDe79*5)Pa4y%lJHuvd+LrY22AEefD!#OM;sBWDT1;d|G&*5aL0nQiXA&hYtP z`?@n~?#`5GE4ZaBuU8u)-+b}p*PeHb>SpKsla3_%nM<8uLtj7euMu+khn|#aX=zFB z$sIc@IR)7G{mI#)oruc523MDIaUx`273S)}1b(SuH}JzIo ztO^L$g*u0|#0&%lzkwQ(jS$th@nIoecoH}O_}STmVFiqnOXsPgYr180CodhVsHpPY zaVH!#n!oW=Q|wjg!#XFMP;2ZtvBH#C6pQyg)4wT!44*IQzReX7p|tu&{W;Hwvc!NE zbTJj=U&myyHI1+=w*1{#u-z)@^6RbF{+t$G6d9YD`6Kog!0|*E%`co>7O?Xbzy;X# zsSG5FzV|QE8m3(1pG&%5KI<2EvTD(V2VYYK@`@!Dk(PnJiJz!Xnv4; z$1(Uq!4t(ib>bKp7@{&WMS&xzuwTud znjGtUIge{Uk}P^=vS?hQwsW;{Dn0VjMh zxvNV!$Z#wfWt$BB%P02y*0ePQ4b7R0&g z$lR(KKYJ$6iic0OTaE$6Mx(+9a;KiKyu3V-DQFFok@55+!pr#ZRw4*_+Vz9u2le$6_PxQF=7^+kvGekBN&yXIB@iz=(pOj~;aMCRIpXn4P1>;ul=6 zz*4u@)UTyJqVL(Mq1Pkd9si{T7@4VQh9Z}>3|*b%tcBGcsK|w)vuF~vh&#)&LS|(F z)` zGnTPZE9GHgqQ8y!M8(7?-kMB(t}_+hcMny2AeoBY2^D-k*n|E`Nb{XM%JNu7=0$~b zXe#yR5SLqa|IkC}R&Y~Yn-%9FLbM9uBNMx(PHbYR5uLjHAJGuM9o~<7?BV)gn?je) zvZY%N6tz42PEOw1@whSr5~4e8Kbnr->==IgByd%^*gsqT|Na6O?aaZ$ZO6jnTNo&% zh9PT?m@QI0F=du7X3l#}#UmX#{-Co*QC6y`sJsSa7abL~4lAkhN<;=Gbmq*NGu_rj z0Y6R4QKuH~%u1_vH^DCS%zk-cu2Gfs=WpcJBCN5x)rZvQY!?&~Vjx*oMElBUs!`iz zq-9!IfWx`vg7pYtJ9vAtW4XN-0bnd=r0^}ax%4icF{NcJ#uq`Fj{@g`c7v=jw#n#W zxt=9fD;8lV94GaFv$KD;36GF&0Z7drIG@%v24G8`V`#(>&haYjhhPT)w8X?dy{Z$B z?8Y*X%mAA`LBk5tdYa>KzbdRRnoL6eT7^YWUEm&)5!?zuL_HlV+lNb+jZS70DcST^R6$*T-w3ROfNppP~W9>+t^meOK#5nxPWagv6QHqu_ozV!H;H zZV!pe1(H<`W{h}ma@6_9ZjK6=*pLu2wsk*F2Yqz>Buw=0!f26fo}@DoUVkaL=I!UGBIix62C18_L1QCk(qiIzEa#CyN&gXSt<6xy?MY}hs5CQkyyt8Vc zt6I=s{Kvy-a`s`gzg2{X9jIQrxAu03D`K0gld;=gy)ka&-f7i?{^x#;xdLfj3++&3BzdWlqwW>4S`&70YQA zR!y4?#@b86d|M}nMlvhy=#ZmXm0Oa9^$hz#*AeY0{aNZ6gdR-@9o4P10WJg4E55C~ z$f5C9snCnTNKq*kov(R3rL}R}T>9mmYo(+(ejhogo}ODBcc-bpIj%V{Ql`&uHbXRz zYf-kvG$-}%xKTbyn-!^-=X-XgfcE&KvMlsZi2IHZ{g~%EH4JiRJef-eA(Uv0K9n7@ zz_>o8V0xX22!3!hf|EKd?5N(UnO2BSWi+mwjAQx(+)W8u4n~bvJ@5e$MggezNffdf z-f>j#Q|@uup#Qz)9h_wEoTRk4Z_iJ`U)mk;#B0Bhx|>tSyYNm#BsL2jCBP z9RvgjbK~u~$22rFs2Z{adMrfhtW1%5BLybWcNJI^3hBuIQg(@W_W$ho&_xW$J#Pyb zM$3pr!6xWoZhFe+GBT?3<_bdUBhMv!{kG>D=DowR-O-c_#Te3CH60 zMkmY7H@aD~Ac%S$)?s@5(lv}Vl=PEO7Xh66_tLJII|EjvDNll>zw#GJ(Bw6xby-H;<- zB%F%is>5S(L{_$tEh2Ts3%Hi`Rb6%cXp}yTIe!bZ=(L{MEP$)mUijo*>V$n#McO%K z^8K~2w%!M)Mh2}L!NGIWH75#vZ{KFck{q|Ocaa-7%pZR?Vi&Xe2MjIz1=gN)NR15>6=_H*QR^$#v+*ArbAX$t!#Rp1GzCcLWA(Teh`*|L%rU zsFE4M3{SQ%Fmiw^5b)UI@WFDU^aJwp(m-%{4ySZ=FOymiRFh$SZjMRB`K%oKzFD+F zAEsxPlj9C=JGpi{d7=O&$JZwr8AXV+WNBZZqvPd0i#NcyqHRCpvBZMe9umPf-a!iD zZUr3V0e97WQ|+?+{8`}UWMT7psTz9-JAZ3FUViw<5wf%FwH|`ob)?>z8D!S5Sx{z( zaiZ?#E5J$jR4ZrB7!0=0gjs?=Q0|6}8-Ju-h1DTi92T$_hR@n7=z1{2zZV7=jWGDI zZqB&To~(rk%wqP-2%cTiVlYuabdLfwXAl`p0wuIoR5bHN*ww37d&|l^mE1YbqjvkZO<8k}2EaEXuWID3-x_Y%qXrY9M=RQc|hEB-%+ ziogF(r4|-=ZXO=gFn{yzM%$29S(w-rK*&>&Q1$-)?f(4v^UY($Er8;%5Gz2ay!Y8i z|C&V)Fni-hk13Ad zaB$nKlgkd9(_rTw90f8?LMC31Zr)`Ou+gX{mOn^VMr#l?7J>7 z2FqhgyE>z&+hsAnvC8}^b(R$Ey_)w}mloHYSYz5w>F$ZOj2hi^{ zo$;(x6Cc9hQV(H%p4;^|1pP^-(oc_222+a!~ zYGqkP4yxean1w~*et4{JQgq+~3rKSnXrHuja5{qT#g)yDd4BL9B4cM{6P}bj+33A+ z!D(u4MoU4LJjBSH5N=p3x8*)y0VeJSxmk(tiv9lm`x-p-5 z(~Z$%xb)`D8Wj{!M(f$7TqvZA#+f?PUVrx{H05a;JexP{VkKsSvaRQBY%q8 zwBzHmZ?kCFLvGC=;&?vgq6)}|UW(yc4_ZDjI@xn63n;V!J8-uQWal~l`n~{Z>e!=(&dw7JLq>2fYG`YVjaA!kel^N6_`jQD?%!`7{`kIM-<8lgg9B;%Kk{++ ztFJJBVAQ~R{YuZ#>^&C}v&}h={QY+SzdxA@3JY5VGk3a2D0WkNB5vLC#C<&t_?_uE zl-MXTE)X|9%k!5&Itd8Tc-9*shal^{jAl+kEh~*YDm<47m-s?uH(^q@-jo z8j&6Pt#P$*fg&3#_+4>3_jIlk5Lk*c-3~h!G!N&Zs%@9NzFTU%cQ!;E7ZFTk?y93^ zEQbvdlFfg$h&4u6dSW4O|*gpHVii>_iz?VCy&IGFSCF@TLA=h`!0)t3NGUyhC4 z1V(%PSl!6T$d?>Pk(vzi05C*rrKEHQS!JrZT*oh=<|bpYOsf`e?X&o$i)fFewF@mH zIOZgTWDV%Rn+Cml`b z?c0;f9g%nU%g9867reFY^eoQZPCWaRTY}P>ODKiF<_){i?@GP|i1q;zYyX*j`^~bL>_n{A<@*vdT>SPe<8Fb+T3IRirVMINO*A6I!rIEX zzo7(K7iL<1@6u+$lUiDCshf`84bpE7HB8rp_vUtKh!KzC z*tpghymVGc*lqS$!ol~>;nUZ4yp8;_5F~1~%#t`*Z@#Rq*5EPr)N2Lv-V8us{(coJ z>uDg@C&|eb%KT9o``$X8xBPF8Px$jdze6%iCLKmzisu)qqzmt`;kYvYdEnKu4E^Q? zj*)uN(7$^3{pUYlJQT7u66DK)0|ykHDzTmFpC;P8yE|=B<01CV1Eie;`}8CtHRzwo zAGp%F-mzyF0?PqS=twteWm@d5X^`T9uE5Y{rp#$@uj`_{0Z<%U)Q;6CVeU zI#nArS=qUu-IuLe^{_oGibqh8Nyl|UI2XjsEOW`ErPbBdlq#^@)e9Ea`>JNnAWR3) zSOCo~b7-Vm99V=|Fjrk+&kkCB0U5b(-@e`Z_uDsbUVzP5-yy8diy#7cPyiYSc4y4t z_u+>NpGNF^W9<@_tA{D1%~$6xrBQ zpfwkq+RDqj0L8*fp!}ujnx@y5cm#%p`GVy*1zRV$u7Bw4LTDn&kUUVoSXbaX6F@|2 zmg&;~ns5i?dYN116%_N?s1H5wg>C$*k7G`SHXSOqRKkKfW|MV=`T5d3MQt(;hC-`h zrp};Gd@l^1-t5RgnU-IToWVC~edo)U<0uTdq^qIYII@88ts9wMK;1npDk=xbtYcBXJWC#yew-gth z&Kho5kbwf*@~gj}B_ia;joAp~@B`%eWxw1O)qwql3+py+lufcO3zxV`xB z4V3@i8tTnQt>B$X=tyc};6`NJco71>FJ8%7^6cFO+y$H8gbdE>*RSKFIsP1Kz?0o( z^oi%pnKSM8kHV1<;Nf0~Mp9sCD0TgHLET*lcWlg`144*x!GZ-B!sE;-$6$WsOu62d z5S;6aaPDlx5%fqrwk}&Tk#)Q7KLDgyB4X9z7-8>+`@0@ix_u7oK+gj+dEibpf$H%t zd-wjmTR^~4Y_oeDqIZyyyU$)wNQjy2Yd;yKH>EP9IFH#mL1PFP?VUC(Im5Wo4ULVL z5aCzH?*lNx9bB z9;v0P8&y}TAt)$F1j4MOf~to~vv=>_eW7Fjz*9zvS^pziuUmfVw2@mf*X{MA3;AuV zcL*a~i&_bAb$?CiYvyM8CCuG?>bI{Xp`T;78Kae7@)&|$lA1_BZSbrW{jUx-*d1Nc=Io7^ma|4_a1 zUcCm&RDY!JH6kKa<*Jjvpjf0wf z;g?)v0y1)PM4{VHj%PKck~=DFTajOoFIADYJl5{>z-Iq>b8jDwal_G*FRqGHKN`8Q zAk@ptE0X`fx(g;ZBO(G2nt5v<*dIE2v6l5X{EKj8*Uq+&{IZCbTySx1Q0DGaukaZg>RH$V316hi^;s%O z+9!E%ZGYQ!iGn$sWpMk}ty9p0K1oaabZAplhF@~z!vEnauhgepUAStm*@`a}ua^Jj z56adtc&{@1qn>K7@0*u?K}{2_|FiOzr+N}+q!(oJl|ks>=3u%YseO^3>=jVgOBx#+ zy-O721`|;mX(GSxq+TbAR58 znYWa4`aB#tg3~e!77IrM?@qO(oGn?6GDy0|T1OML7b4%(Xb;dYB*Ec3t-q7Q+0bP`1WltV4Lw~V*f+vCM4rc z>EXY6>((le8LOXp=-PGHdZ8*@Bb8^ElSG|nq7-88-{%!KKwg_G7CiAhD2N%CdoG&r z_jY-Hw?G^6ZNMXFcoide7KEaTwAhJ2{&J&kjF#Mm-iAw;k9cQfXG;Rl5lE9Bf#vl* z7^SD!p;Nok#Kgoc1!}sB85tQ}NtbZdUuQe?yY*Ame)#a=Ge3Ke;i94Q!(9D*8`k&w zQ=XWJ%P_jKLlJ&5k-=dTMD}J^o zAgl}3W?9Se7#kSO6A=+fLo*3@To%oDG;c296-!hNm=YM5)+udmH|2a6=7 zN5{Cq{O7JU}uM@BZ3LCId7!R8T!Wp>p1BM)n&U{f_%}S2F2tA(AI!7M6FaIPZ<+56w4EpaV z$ULEAYQQWh9!Jb~woqfC)r512Aw1WE$8?2`|>w0$Hnn$f-Oa% zb`M|gNqV{ra{P`RJLnwJN5a$C4^|b(Xko43&Yfx@OQYX>G!z1=TE{8nve_X4#4I#O z9uVx#fQ=C-wf)dWoz~Z%1N+_?dytOs#jf4G3;$}f1{C3$P9r-YDPc%Z%NTkB66G@7 z;NO>&*lG%>f@lJ9wmB?GF>=-q%qgHj9tDS&Z98^IgRHu{+u0kywGj=7Yqe{E=GU%W zOZ5YI-AflQ&WN;y+Rq-c*2eaBPZ%)2tf^UzIv1ogXfyGxI=Q|$D{q|x0|TEp^q+-V zL;9EINn|>09s{WBU{ykv%n#u#g{VA}jg1W%4F;-tja%h)~b8? z`MKxi<<&9kS@r%)3(yvf@Kal>1^XRV6hwiQIcQ)RetRZtRF&w5w;PH7xbAAx(9~SF zY12GlVvyejxJ7jA5a%cf21+Et#1~mm78sj5X-hGfxW=BIq>tP0!S9?-2%h^YWb)m_ zQsF&Dt2{b}BK{@lYRjofXRYUUi(?hS^8qnYxN(*fbkRMu33$+>+`MJW^_w@{an>4;2t=Jng-{)Nfc{yLS&z?Y^JA~d zC-leFKj`Hk)RC3*Wj}9$g$oxpn!ut7fbxJ?kIP`!ZN$c;urPLtjTps{b^sg&EGe#Q zhFJR&LSnCNAZ}X|c%K5~@>BZys_i56sjGBcx7e66&LGMcZEfi?=bMyFaKgkB(o<8f z;F7EsFZxa!bv$NZY9vl(v)cv3@1HgN)!f->0r=MQ|I;dA)yx0>i zp1Wm#Qj*F;#l4Lif0cw>C^_&LKjta;tq|~b%cU4-Po>srhHSUqni>Qx$?75*=-Ay+)3@oW{zx5Xs`tN@{pyQGq zALv~6`W{kgX_$){MYky>P2qYmgQ#-XxttDZyxjWKo8Ih zO#A`uswkO-FvVlWfxMF37w3BQcT15D0=E+;lZG+dX#tfQ+J_4e@%C$L*YQ{76oU(H z?CSD`+UZdDC?6z8uMoRE03qhL2p|HWwAF*Vmw4HyKLXf?k00HuQ0KjXJ;94wya|T) zX{KLSfV+|w+$AaLHmdXz_jyByaUJwI=*O-=$He@7j+2v^sOb33@SUcs>6ii`*Je8? z+n+Tu;=_4%$9(RBBwo+FJgL$MDS>l-2#}bu@xUc}_4aK?x)`{@pe-j>Al%4+){aEA z4TO0{2p2a z!;rEPFp+c`UPh=Gh~OXTnv;5!QD$Hpfhm1sgoTYhy#QcoY;B#(#uh4amY4&)quDc{ zrzNU%#tTRVsb%2lIUl^f{GAA}7bw1e)0$+%gBiNZUcVh2aCXkUqbPA58VZ< zyT!#PGEqe(D25ZT_9`e!Cuir#itLf>XfV4}so}Pd6(%VuC@g_Lwl56d7jk4wfU@o+ zFmx+)rJwaI{4`j6!3BBDUSuWh+sRL!>_LvFejsA?EReSQ4GnqVzP;vLC6pUfdLXf> z`u~{A!o?-f;h2n&j~PwL-BaV%$2Bx^b|R`XKtiJetr}zwoS<&W#9J%Rg##mSa{KqY z#oPrBmBkx4eOAq;G@#xN5J2wmVP6Gf4i1irr=}-_%xYI5qx~yBg@^lr6KMhx#uHrx zs9}My7rCpssp*t%RGzVi8srNkBqVNyhOS4_B5EI> z-*R-|97SIA4x}Bst5fBYB4*z^!)cBLu^QD!{s{0N4a@JtuzTGwBv8qh{QMTfW!@K0 zjR?}-p`jwk%81l~bXmdvf&6`GMa5-c)DBTjc9_i-@&v?ah|J# zLT7tB(Za(cBWpKqoQum{fSa}OMCii$g{x~x!D2i5*Iy?5-9Taq;tj!KO?%?E#7 zNbtms+V^tf#fS)j43i62EO(8ocaXdhSm@`;@Z1 z?tE7uCDfj>o{hgW?po~i!3Xkxi}U~bgP71D_Jl6i9jY>&Lsxcu{^jb+E2hbOU*}5k zN1p!et>0N9mo7)YWt>azhfmb+?T>#=RR^>m*>5)MC%76kGtgMby$G_UJh!s4au`FT zme6p5dq4C8${G*kaqSOxE^pYd;S^@h0xMC1a-j8FfWuXesJn{o(DPj#m;kzpMhq2V z6k)g%eh&@FflP@JPSP_r_Q5RW!5>^`6ynl%N?N0FOOeLH1ri9uuTK+7```3(O4?t9 z=3{|=LVWzP)vLXMs`Cp9WMUog7%k49F9NutHv$fQLDIKmy?N+3*xng4hvMobHiB>` zoU6+g1Mdo<5WA?D*g;KWu!}%y6TfR!?#r_@rUS_g!{mJ$a5&w-Ppym38d#ZTSTPGR z|1~&HB>?C~TYO*YygqTAauG85qGC1)66B<@Qoaptn z`x4ZMcrfp31{g($b=~I8i;(!X^ctTzcg_o6^fW#wTCRZuckbS~3E2%Ip)Q zZf0g?w|3|=pj>uC2A&_0>KV4LSRY{sndxSBBI*XB#qnL{0IkwfR(-5;35f|(1tm-e z;2lMxy$dp7F2W;)rpT&!d-5W#1lnicM)*WPT%Zpg7TUhUCSac^v3q$94At&{00mwy z5b>^{Y6Yhw-#9ut+P&->vakh&=~a`K78c8DHmzCH;5U{%E(n6GF<|y^e}8vNi~Dpp zAEz9uPd=O$Fcxc_O4{$E=sJys4x(1y=VOHq|2b28`^@+xF{AX=h^c`C6UO@b?x-8w z+}$G~e#q}cdEC(2>Vc`GQ+S7gxow_x8v!j|BNDGU=ZEETlDl@X7^&>=%(a8?>ty3b z^WdN$H(%dHjZPHMOua);F9UhrL$x^_)1Cp~HXyaD#fLn4#4EjTpAtI%jly8u_aN(l z5L<-O^}vU0ynTCn`%~yb1p3z_wEI!*|6wyDr3$cG{ErTyp#tvO6I@?2%pXFW-ZwNf zP&Q-V?R5NCSqD^w^C$^?xRZhEks(bpIU+)x0OwBgv!|Ht3@_aB*a1Ajv~wQ%OlnJ6qjpIeGN2w5dF1~}cWUmxztE`0rZ&5j+%V!_RX zgj@wrvL``pVN%DZM@(R4%OQo5E=y91p6fE+CkJiEf+QfbOVEk-9=W~n1CYTYPRZk< zQ)qUBf9@2T0-l+_WXX!ssz=K6@GXKQuX+blOTVRsk&ERr8awESu+Ica%nZ^^9v^ zlMo&cXbPHq*+yvz)Sq}+0;VNU+f!#Je8~PQtcsVQ%$L>x=z}NeKX8TI!3CReKk2*>TU^#qf7?Mu-AK6Crz2i%E8Xb|MY2t#Hs zbKt;yP=o>KjV??QL~V|zL?t!-D5a$eMki&^5o1G+4o)k4`En*yErmGH7eO%0PZHd| z{ZeQs8(!`<+aD^`ocr8Wj|GUJR|)fCwNs^EzN`Y^*c-9inL)URd;R+L1PM`uq1swW zgYN-&n;NZbn)qSC___B}IZ6j5&(-x{5qHut#31MtXGcaxf)iP;ErJ0<@O;chgNK0m zicmp;gcA`=Vl;WMw_1k9c1lA9osu8l3PadWMWFEZiM6E_h>%%*-n7 zRRfn*9;+;XoF>So6ivz}h;qD@d@GmcCBU1F9owtoH3TVVpm5(h&;^D|7vu{-(Su&wU4F8& zK~X(ac`RKfV6?wfSeuWbttd(=;^r^QgH*D^=rQkw7KVSf*|aOlh%aLiDfHO-zrq3c z4P#+t_I6YEo)foTsdeoN_V%lvD!Bkg|(d*x9WpErK`%uQ2c|v^%Guc2bP{EbRd3567=nt*zW$D3rfpwfaJRu+Te?T}lRh*TkWA)};COOEBBmUF6q>+B zjJcdE0c2-3Qhoxa?KBiTC=Zk(2ccPiiFUSf`|i5Bx?6$)ZmdSH6gj@SEe`|mSj=^^ zcjzYs3w^_Vdex3WLb$Ji1LA1rob3AcJfl2mCK&kFxz1y~kg2aU;u0fRla-Z~KZ4r# z6qHVZcey4@a5W*;D59&=&~UGRw-xWBqR^zeH-lF0+Q^rf zl1h32fO!h46p!$b@wZqYM<{=+%1*T|WKF z3FdGy8;QZfMfXY}x;a4jbknKI1UXLySu4?(FFBLLU3a+_?-bI5IvsE77P6 z#mk(~dy8S)>58p}kPKZe?PqAhzY3W>$s(jj(7ne>r8z)+?srWw?S>)7e29VgZps^$ zpy7)mjt#-{AP*N>+K9dk=W!FR02*4{^?pzkz_rl%8x^y7ss$H_mBl@Kg)qEW-MibiJB&>pIjE%-rf7yWeBRDZ z#B?AQT*YHtaqF{Dy6A-g9&8+qJrEMu0O_o#eeXSsTnNODqZWxzNYKBH026?v0$2pM z9Mc%r^^ZOR@B4@PCs#QGi9GGSR*i*P#1&M`OeeRZbcaPot)LeqB7uf>^F2C<; zVfkN^QJ6*ag0Q(6P2f3ad%iutioFkcpWB||nU@pv(2(;Eh2TSuLOzW6}*zq~Fz~&?GeyB)XM3UO{%C{~&K- zN2K{Q;;yqx300L##Kwe?#m!|>er7QLEpgxNP`YHE=YdST6yXjZ3AG(gWyeZh3Ekt zG`Ifl?rsFInZABE9qKU|txgGtD17wJFO3k_2(b~^qxf~si`K4RzfaUyPcQlu)Ap#g zQ9!_yGVLs@`{lx?baZr(2KEh(>@Q%cXZ>FSutkYs_ZeF=HP=MNa%=Gg$LqLe_gjX{6D_&!fo4Hmotg+KRU`R zd-2-TyJXw3S7;A1se3=zL4i*tPpT45Y;t01pge$*NW5&hFqE^nuQEKLU{mK^MI+<0Px*zHgJ&ZQyi*QI;G|a$>OX$zW`v8=iuMd`Dv8$_V zxB<%q?Q(a@bhx)MzZz=GIicBgSXQVANOAy=|24ys0WJ zEnPyOJ8Bms3G-3QM0SCkW&U}rcFw$c3nONWNl55{1_4eFPQQ|#E({jw5;Wm&pC!#l z?n0<*bRw*5Sg{UI;B8Gb2J|uYlh^3ufA;!&^t3wa0gmCnqZ5!k=S86nYIU^Z1Y1R=M8lj1nK{MltMWb|Wj-z^; z2F(v3o9UbASP5f;7#_29HOi|7*Qv3He80&~(8~E$H?^V0RC-wQb*-3~M&(3&eEc>r ziEeIg_i$cPy0;Xg)rv9@gRXIVOC^ZN{G;}M`S$I>WHE$xSujhAl_28K`cxcPEj}a7 zxJqHTMjO2X#DiEvF6Vcpr7KE{of8ulBIZPf1qB?UHZ6E1%~Ql~U~gzde5#lC^T6N! z4gan||Na}5m@E~cmx-Lgw5g9Vux6(7OC^E?8x9a`K^K-G z+;L|99iemU!lYcrML}59-*Olo;KhPmw;>w~ix+rNZv850+?zHnY{A6WN%znhNU;x( zCr2p`qK6`0LTVSl2e%Y781u1Y(B{s1$g1X`4LMt{`Lg-__zFV?_=G^l6&ZH@IwSo# zs!!aQ*C=}dTrMM#=?X652J9FXeu7K-P&558USt5eTxi6(r+^ez&Ma}IO(K^)@UXoG9GsXgf#=$Cz)rU-qt{c2d>z=Yf8MLq9W7o=N-k$ zx;jqO0%~2yA?e~oQOct#pJ)C-B``olrfkcAWvw!1@G>(q=j+$Ew&kGBJ>Q@W?O32= zxtBYqijYCkEN3pBglxoqxHnS4+MjcmH)wnY{jX`~t^mz;3dyM;wP8qfrHp}MAIjEIEgEGy^esDerOufIO5XVP0g z5NF`gE)ZW4r_kr9S0*Ahb^MBgL`w4LWrmT}36Y)ukw^Z8C;$DwEDJ!UP;2k-k((3W zK7!Ag4rg2}EcZW#MYxPEAZnxfsK0+7O0aTm#3EGdK|&GUPthfh)N*mj%|9|ZK50~!Jz_jn2f&AiCJPXC z%D^D@qc2YL!WIbZFzql;PjBK|4aR5-u}a~F;vmLCAOi`OT#ta5VP3zY80)JTk;6kI z_wmWe$sv}?^@)uRg3gnUV{ts{hKiD}wlxCGCoL?PuOBf@3rwvRvx-?aCJt#yMTOFb zJ9K#PExZRFsXSW#Smn{WAZKiLU&oR2KKky0DQKnU(b0L=IMj`4@JK@eCl zh>R#>rF`A+5JW?=xIBtD=?^~}<=F63Yltq=88O1{#@oW(@B!#L|TdIF=QJ|H@)kBxw}e;w9ILfsMgee|Wcf-9CKc z)#!oFfq-J4IMzd429k(7hb7lh(84lsQ6Zu!DA`jV*4<~PS6vhw}r5FNV z)khVOWDPNw2@Lk2J#1jWhU7h%$MPf3jh( z*rD~C4fQVl{I*3YR_pD^%>$~tmo1BxjzEUE@i;;x_TYlQtU5| zQ%pohgrqerX8k%*MT>B2aXL4xT)9$4#{iOv4J{zB5)^MpeU8^q<{6@4EXQ$J=!VyE zPGQFe2Jt;DASpWba;JwiKs(?*&+a0KWWdGk1r3V;cL{ya_1m^BL#yHi6eO0d?^$uw z*nq+cy2UiO5Wn1YLh%JFK&#?NzWw%%19AP+yLi-q&A6M=8aS)yX=qkUPt{}ca)Z=V zr*~1P?traX1vxszsVyea6qaLLq5zIrx-S{#T%@9gR-R@iz`R`OSTDj`wHm6t=uZ4A8>a@TVLS#JBB<87i9USX?45}6g_Y<9bD?2;cr1o4Os|zwN`p z7A)kde={M-w+|}?@-@I5%n9|}E-cJa-0{o$le(9W&nvKp4DUhmMB>e*yx+igPQLWH zE7QLN_uqa~apIAsK2ukx{sO$wC%k)mHl6QIo-RE~=AiWWQ1|}dB9s5)HxETwq1vxI zdpk5V8dyoX8WXKVJqHdRm6r$UzS1oPs#*ZTO9Tnlf|F1xN(Pz1;KLu%!X zmel+uATOMos|v z8`*h@(ErVV>@cU~<5k=r8c(G*lgA>~%VoRA7U3d8r8_QHs>pR7BN}d2c%--*u|N>K z&+=~ExM5KV%oX80c>A*HMT9hl7PL57e)gS5S9GFm=usdIjUiS>n%LN+<(qz21%n+K zCVxL4olzFE+34y~gB^Xp`Sa(Kkh>fcC_poSK0cUHHpM}d&O%v*Uk7R=&{srcNRJ`1P;?= z8FX}{sf8qA!aMWT7r(dsj&T5&vfe!V1c9uP7qy}+qunJhzX?s`GSW2cfBU7R z)QkPWFHrF$3F1)mh54bIn>O_J#wPbv0@`UAVU#efNV90u5!lsl08Ob=v?;J3I+Rcm zqT|_Gj47)Xxk~nkVaX`bRA(hkZa+?mu=cP_c4ZSpph6pT-Vzs{%9tj`7kA!2ieBTfjxRuHD0Sw-|)jf*_zE zArjj{5s?NZEkIgIVHB_vWQ!o(og?7TFk*lpAvw}0k^>?!((tX@efIgE^BwoQfA8-- zZnq%wJokOYy4JN8ewyvr2f?|@xbkHVS&;%hgx%zo5Cq(Jf{qSkX}da~7s{-n8pt3{ z9Sa_s6#~o?6BpH&%TZ_Xcz(q4*Q>%7M+&9KV*RW$Iy&LVLx<}S)VA&1dF+ea;BJjL z&mW9Yvg{0?aGmARlaF0p*^03l%bc=LLtY|y3tKP*k9T^SeF~Y4Iy~lzzl+VaYY(ua z{MGj4ILa3$39W=>hu6QTUvb6`D>&eGEHzy2*c5?&@J*qbYSR4KF3v8GdG)ehHE*^8 zu1s+W+FrR3h%y<*fx4(QsT&)?i3AnFe2G-Y56a@Y0}r%f)qfGppzIJhvR72#K)>6V zRT$%|ojuwX%xcl792H_%q^72RL9ixk1W-YHT+V;O(ozs}Ubv7fJUc!Nw)%BmN-_N1-x-Qc@ZsmME;!x+Ny3aPpmM3bO{oi9*N>!{HPi}bv2CLwQ z3V2Erf>z&w2@R zCs}uO(EQ_%S=b4;?~y;I16tR;RN%08AinXVwDrlF-QrGIft23c*j51+lh| zHRM&zBbaZmEyK0dl;HP9Jb#0}rQddWDCp;x@il~Kh6a*7;!anmBp%(DT{DHC^&;v% zE+Z4ZGOSH0=kIa>+P}6K!cq79Cu2d7BmJzt(G)7GzPmP{WdZwNlKxB!_3vF9l#XTC*FzL&B|((dM?A|kdC9v(L|<4!lKT)LzB=;K4(0&IsUt5(lf$=AqB zy1+O#xj0+1_p^$%UW4GtvmdPv`EA>ydNk|i&2wElPaIzEb9Ytc31Q!FBM<0J{aJIy z8ZUis^z3ZoDjV=&IMG`~d^1OE`vXdrzjO?1$$|!!MPl%?XQ8+QoQBlw?~*#k2fP7` z=t0#q?dQ4N-J({_x)^54k;T5ATMM~A;b(PYwz>G~i$9CNYT;*zqI>ffEQmw6mpjYv zE-x%gm&b1CxpN_r^?76go<1@Uit?KhL2Wcexa5#>cv;m}y$D5MB{XUTEK;$;)&vz2 z=)DcqaiOl`kfywPA!MK*DtS7=h5rd4{f6uCn(6{t`w4$88sz1=+D;ILIrd$+n+JJ4 z8w*f?63|GSfmlgty|r){3Qb)Qo2@}awugI391yz#v!a=+h)7M&6Mb0MIG>p|W2ITU z6E9>?zI^a9HgyrE&U@0@MEH2lj3w0-zwF-^7#h|RiYe=b@R1JCxm>AlXG@5uSN)f; z=)Wa=&eTN9hpeAc%HZx_ouEA9vcY|O*?NWYM1g9jCrdqp489Ia{QJ-T=P%N1vzxM@ zPS4eivdStPgQ5vfe}6OGYnb23Wg^MIgPonhqT!|3_cu53W?M9f=@@lkZ1z2)ZAIH8 zRAa?q^Kl5cd(eOwgts0b@hArkD9;DEF9t!oV2nzi)&{*e7ae+(hezCycoW(uEltfA zDrtC_VoJ8>5&E@|gWXa<{421B{Dx(P%QvQiHN*fY!|eS~9G| zVG<+h1Tlp{2dRRFLl8WDt|EvSi1oBIN57@ans{#&OSG^Boj%B2l9b^XmD=GfH!9XD zB`P=khy-<_3;I7*!23mwN;WEVs>Ng=_@6890tR`y!1QZ?{CfKM#9-)Sicrb{TrMg1 zr}jtshzkr{?hp;=KpCg{LiexUZNtz+1`Z^9wq6i&0pg8U?qK~W41K|6jH3LoB?_G; z6|_FuE2Vzi{@{5&Qr6FTfWTLrNkrd z%@p4oyF^2VvHG-8drn$L#^(b08wZaZ`}rIXg#z{5;*$3T1wD8W{MRUTZc^_TkwQVn;P&CLG)nS=iI4Q>W%r z#)XqRKxM>ZjJeSe+Q@MrNQIlDvv9S!ihv)e0z};@t@H;teev)klB>fYrwKAh5T6q+ zPB*C{T0rq19Af4G?=s!(0-~GSrh}Sb*VTd%7>%eJi^a64?i&|@6Be9}9uWYpYYNm# z51edpK0nh#HKBlmXeS6)W4$32H+l_u_X6^?i)R5TS38dX7GU&qcKi`awJaDMC` zBehn6u)u=$X&07c8R#oN+6V3c;Eu_fs=j_uIrm%O40{-MnYvkU@S$PYMuf9^TIh6e zXp0-kxySQmU_wWuuuSnoTV>pV73ViGe{ze zDP5#vZMDtZP=dNQaO;+k$(9Y1uor-|DDyVE_s2+x9><_@X#YO9b%SN}e_O5oeZ*rP z?xHpMN$ohL_TX-uynuIA+vqW)Tl*;{w+a&ZluxU)PY>$;{b&F4e{T-4`rpF|YyyVt+BIAHeD=m~txWyQ-6AxnR`4 zp-&TMPxRdDS<(~zH_7sSTGblpQG zK~YW3QN*PHS1)S*()C?V_jdT~f^-S@V z?ep7|>%q`<(*? z?l_-WxZj+2q1~w8>wL?a4I7L-p#uF9kK)_aY$2b6 zSQ(V!c2`m4pAa>;7~FOcU%d#e_PP^xFn z_*|Uw7d12b)%sB=S!mcU_x~X?3?oi?e5ySoKf$sG1QUo+9*?|fV@5Lhfl$H5fbI~d zmz_qHJ9gk#++kqT#-AtiC7S#ORR)vAw{z!v1S8u|NqPqb&ZfM3Q7`J0J0^tvx=LuE zhl6-cZ%hSmrW?q7Em`A2FhV~ueVq=SSuM00SBdIwPdCC_Vvx~Y7mrX+_46{R$k(LsU)j~?Pt{{|MO;A=C8o6^FxS&3S)KoC#0$%buHnxqOeSZc9S_DkN z1j8cWXck1CN%n#M0PPJ8NLmnN%Gd#ty&d=ixvm!5WY(=U@QDi z>vNgCZ37N=lcUz*=4dS}f(m@6fyyBsJuN#IhaxPkCaIfQI%9`SY=*h*8we=$GvS3KlRA5p9t-p9kj+W?Ru~#J*f(jh% zf^c;U56}ycNjr$Qu(e@3=aw6k8jVOYK$1Xe0^i_op$cpMcTs`KeK_SGoCBS z#Nt@?pj^_xd@@mH-l+WL%P}&OKaRL@9Z&^r@3eb3pgtzy$RSspKOb5l1`uCNNh24> zpc_>XBsRiPmS4_k0=yOufFuTIyuD`vs*gQbz7#FHh8`M-Q!K{cA8Yy#qYPvrXz^oQ z?qKm8KNN5sE%@o&j?~wz)h5#gHZ$C8L%U6%zL&>LVk_cqTP@NP!&a4GH$R+u$JO!XFH)842ZBY&SMTZbSj^yCW7cos?_vHeCYA zXu~DTUd{Z>*Ze5})Gy6?O$(S{YG8Z_`Rr?2sL(NCpA88>btoWO+ubjY$jCSY;H3&U zwH+X(O^{b1uM-&Mh^3DrCX2(6R82$s zH>^iyQD{O+;xvwD+HOg1o>T%PO2ON=hoW;F8l&h`2(_DGda0di_1H()0v@~pyqKs%LFuCz@IO|dOm?BzYKH<~%FfOdFn2vGATgLwpB?HvRVD-kJ z$mZC4?j~vCEL!sPz#hFtq-omxh$as-UAdRD3m+wC3cR64NXiB9gNsW_s$u(M=ou_(_F}rzW%@Xtr0DnahKV%jOaVE_m+0T0 zEgL@g0Q{yl^J6kRL+h&8r)dVFEI=qGcq4`#80*Qv!k|XgV%7$u86Lm?iNfSJR$R%} z&B{IUXOijq5bxS`J^pN|LR#^~KQDQ9-4`k|)pbADv}yB>^pEYsCw}|&?Q1q}>>5?< zK`U=elDKH`*~~gSjpso@me<{p+J3d~=$-YF`sU)U!omxaUTvA6Wg7V~+)Bg;H-WuY z+Rp=E7?OC3W$`oU$u$ojzKwSiarf>#1@ZmjP?L^C`y9+j+;i%u07g~?tN@uLiKZ5L zY^@`?W@9Tj7D0}0_ilNxBc2tuSyT4YvD0yta+f0!asTO0Di~kZwZ$@tDKRJ;h=>s) z3_7mBl*DpCs8~wc2F&{eTGF1z%+#c;9zL+TP5^(VC0Z$7#kFWl=+1nu2b31QmB|fD zqc}e4g;)ESXRM|*v3ifCix{3Dhr{|*rxr7N#A_#%wBv-hhE4)IX+Wb7W#Q(+&dgB9 zfCuUv2c&OXoZ9T)t-VE3GHY9#`WZ!;^jjj>c8CD{V}gc&0xc!`*4MM!EWDvs1Gu3J zakC0K-9~Mmyj0*F#P*E>d$w0#ZP7>Q-BJF-*xsA&X`Dhnk!whi9d?77PcG7C7XF)O zaIn>@Qq&(PEmoe*Yo39*2$7JMt3yvY)hFsf*!A#Esld{BwmUyqJN8J^hPuS|65^O zf*F$fvn^{@Nk$;|Z%@(d)5*^)og^eqZC<`Y!~E{YZfkR`Z(jOM$`J?Gujds_f7fmO z+pjB*M2BOO88FrG33GkU_ET2l*Fjr)?5Y+E$B(hda&&G7^#ED5*MpXk+*29(#GU7` zhE0ppNH-SX06*U}7)O9(k|Hp8dYA`eD}`Ux1jLgZrMQDxF=Z3$ip*!Um6gw_^P>w! zx>~m>_HaUDb)F;26F|49gU}@WFTtB84yMlq-p>iU6RdqWh zCF>>3vXtV@$!rBvz*FGl{#1|!;sQLZ2d|@ALrX8x-!e;{fZfLn7>Ye~$qEJJ8?7KP*KPdWp}6@rDQ|~z%cP9%wX2If zQVpE?Hn(gZ>8Y<98}Gd3H#9%wcV84T{?OIcjxw*!zZ*uVob3qW{KBUrHPzL>rRzf$ zGKF;JCjMwH%s@aDQ@k7Z4dJzxG~8$<=|2T{c|T#o#lqFvy>;s$0wkcNzM+~fDN<*J zr&$}LsjRGgtT6T^s!7~@18`Ei$O%N8({~;-1dl~tpUUV?<}%ERx=DoJxcaQxM-041 zn1ru@R1k>;%<xH6anD0zS(P36(l@Yk1({HO^B3-nb^))#%WYZ9kjjC}QWNay4zyG*A2VZ1~kkoz~d(~s<|^Mb)MPnKOhF!j=T zr|^dn_hd^Sh3_Ya_s(tlDOK?D^~vEaJ8M=9OkGS^tDVVx@ycsz=Ujl>JKlA#X5->A z_k{P12ZJdfqJ=kK@r+NvMJ5f=JqA367inX7_jV9s_TP0_qY7)dC;*Q_1QIjCjFt&w zBO?&eO>pG?B)%I9KZ;j9sEif)J%E;$!Xx)FkoBZw0eu24T%P^(?X zc)`4R64Hm=)?^1i5wY??-tC%1;xlzOt}rKts`lV-)Da zeKE4w+BLO}OIS##oL7sTn|`U>_BCbb4UTOjG_9DVw#O@Rj}rFw_Krqt(+wZd8j*6T za4fXa%Ucyhr?QIBAh% z#hz5Q4Auydbrlh@B&`IcrJG{(v4cUCyAy$+*904df-&3(bSxhv=nx<9^7Pzjhy!Yb zMwgxOGZOG-%jkG5RzO^twZk+`L*sT92C7pt)A46LbxOnHX8PCU zVdwAL@FFylqQ#j_mYGf8g>5x0OY~|Sq`2vqcr(hc@pX#$Va2mU3^r>3-ant~H*PtKZj6QM3HrKnPq$amVSeB%wehveanB_#Qp2LyymOo?5-~w?)Zei#wJtOup@3r zk@7K1C7FtPP>$GXWRCYWWou#aFmJh9C|^7t5BH59hZ3QbM@UFy6bSAUy^L5dkhw2` zFWN1dzV|p?zwms7f}%r3z`and!kRBZl6xW0K2hthN6GXDK{tz@FRqfpAQ{u4cwZmfij9>Pa3z;CcEh|{(cM@B7i6Mz?G|t03&1$ zh{H(%cyC{4=*`ZOvg`V@!ug<>*o&#q5F!D;6Y56^NnV zA<+Zu>K;U$w2=`&mA%P@>cFfpf#R^*ux8DLHg%07NQ5d$2=X2BaRtlr$bSaY(+m>k zX(P~CdxrEFj$({oMn)eR?W#y`eNHOU61meUgU!^W2^NywW%w_U*%{b<2}Fnwd(_o< zA>UXOi+f5??*OM2Kg@uE12bi$mdnJW z{U-j@NP}KzQ!iidudGxhV_raF1edhv?^M8&0dQ^JDyD+RQ0My7BMYb1FpnVO98GX! zIjN*1-Yf_~MvUAlA+3mjtjEc7mx@gnf@K4%_fB+zn5~zFzcFv zfst6TW3E29CLc2BRlf%Gy}$~qfz}Dzl0dU=ze(j$d3kxm4rmfg?sc;oFDxj~|IvwZ zPdECHKhE$Diswey{6V;@oyLV84fXTeSeN25=t&zf3L@Y_M=$z<~tRDgl0U z8G{KqQ8N#)I5}E2jzJ6KKd-N#4F@p*fzYVlMOQ9`ni-?JS7U?7&3`B z!PkzX0DooA)I@__Uzv{ULB?@{ncs-YW7%bYXoR(m6aK_B+8*y~IVVXV=G8iLCg$tc z3-cF9UQ_5SzG6NUNdQAs#sg<_QC|S*Jc`F(M4zEut~iW31FCJb8?_5Xm9RBVM^@u!?;UeU&v*G`P(2RUuxK%sfc4*jG6aF19!`cuzS`n z*|^YhKt<#xUI9wvsCkuATc3M9J)EIp7Kb*FqaAc{%b^_`2t*Pd(9kF5d}>xeq5kII zg`L)!_zfME1`F=@6llJuKw??d7R7gw*x38Zwed?>P5Y?&@{3dw&v{pE|0`~(FA;yW zNY(n06?N(StjwQt$Tz(i*eelSK6GNWOrH2A#-hLfo|n5InccnG4zfDox`+oVg4?#e zi7`UrR>}4IhOq*PyiTeFtf%zR7{ieNnTX&nrXt34uXY9ea~SZ?Cm<~Gw{OqEh(Uws zR*>-}=H8r5p3Ulq4&4g6F(^;QqIevJ9awdCHPiCyY~aIO_0=nNpi-uvUxt>h4Qv6P z#Kr70BM}FNK_7FSPMmm1fE|#LW5AL@u=j%OKAHsGFLiNDvrX-&r8_UXREb9JI6k6s z`r*^3wWOv;OD|uEw(k{z{2DVNMx_pg;DnN7yhlX2l4ehxXL&rY%>KbltOS|TCg)-$!3^s?)HULNK0U2~*bE~}a zD!BwGys4n(e2ywot}$;&E9GTCDa@ZIHWMpKmjX(QiuM%~LQRGdVy8C>m8OzL$80qVxlg7Bx#98v`-3s)rMaOB_!*~3VV5*1Jb{2*k`?b z{d8z-Tqg0af5V?BZo~2}V)~LrSHAae#>m*s25Bo?}z&+Os8E_|8=z7syv(Ml5j|4~|NE?i9h%Y~6uWS6W z`IQpF7LrpG#xq1Z1#{>GNG^=`+ma1_t&HQSO~8{*AW?AP+6F2fUqM|Dr4LS1a$pne z7U>*zW^_-#9>emzHaR|Ex$UtS0d^gW@hVA%iqAvie&CFz8oR5c;nIjHso5SQ^ficg zy4h&ft^fzD;Os#tbu&>&7) zf5)8v{NI}zz;yBMt5EoT`@HN|&IV+~-hScYMK>EjY{g||nuM)4SkpvcBV4-iobTvo zqrrhBqe&U6o(8(vQChH<|0gX?GN{IG8Cjf`r}v7Z{*8n}AaPx7cRr2Dgxb{f2wg~I z0~=cejtxF)Zb^cC6dyzbeXV~2ZZWNja&vm?iT&}n$<2AF4|d?^WAD@x^d{Ukf}6yt zML%_WPHb*J{N#?&PRP;kJC_yZ8G#9Cy@Y(t*Q?o3NTiOAIuP&>BYka18xSfLh)tM> zM(f(ATQo$ydv}zS%7trZ-Y{A9sv|e>2sF-|xvXGOHMBvVF!}^Ue4t3;BVeB>vk+Aj z&2UB4^wo2@-LjxB{LL`H7@!-Iq*}aq77~^zIyuCr_PB!({BmO#k%N?vOwi|%gq+Se zCNmMJ1BwMck!xU>$)2)xu4!*vP~+avSF2-{8;Oo?_rZe=4%sNsq@nr^+Gr>Ou3_8V zm|m2BQ5gIP+oBCJtxJjAyerNZ|L&Sf^4kIaPE$2?kYv2;}b6rCaG7 z$n$(H_QJGWz5Oy-=(_OX#CkAGXEm$f56JY}6mKF)h&hoQU*6q^b%W>96jrPkCT zEevCr!r6)T>@13vH@w%;!ArwdynRz*F{Dum=H*~AswU3yh2?R4Xh+F zZI(-Kw0T0%XK2u!h9+4Kc2O)dPh)izUB&sgck?!Ia+(DsCQ-$!%ueABYRW&dZCTgx zP4r;M9(>$NgbD$xYhV1FPDF;Qa-7*s5VH*#bruTH8o z6O#o_gHB7}lciO&e(d}w60-VL+HF0*E5~AFvIW&Q9P@MeVx{qF+DD5~R>X*VJ3VFd z+gbF@nHK>8y=VF{fo$vD>8*Rg(~J~srDV=%93Kxp9>IxzaWhA{Dvl7wAptWon~c)U z^?-vw4^iXUK7Q${&chx9w$QmVF>&tew1QC%VPiusU052eZS&3TCz1CosG5ySvl5?7%?EDIIA5LJh=P4zap+%#; zZYIGxI@nno8wc6gb0Gnd41`G=hmJh~1kPT{&}V696rW@$?Z z{1r-~+)InKpOnW0RAI>@ybVdh0ac}t^IfC_>W(NEMPMB zD=SR%=%&9~ZbxC$FOuwmn999>XxmI!;GUi+Sa`jBJ95gJ+cm=4;UK0B7#)R9ibfa9gL zM+>6+keO^WB-^^lS^=T4&!0DMqg6U|;=;Km*2m)H%VlDmRWDUE*4LO>h1@fY7sFvp zFj_tP8Vv+ZOE3fq&Y~e{+Yz%TN1y79y}9EQSZcyly@79DQUKUM*|2bE)d?GVFda-R1M&;x2ocywbAGQ;IT}2v*%uv)0z>8@=5Put8aT|)! zJh$hQ+Vb*`0Dg3(J9h$c`NnY8(szs`FkxFx_t3?5xt}%mEtLRY!yA#L?D5?EqYtFx zEJ#)|ywUZ$x|br)5|(A@+Nl?XZ54V=n=Mx?%4OeSH0;c!$8UPX}f1L41a;q3^6T1)&}Jy_>3$iu*D{3P^OD=u$j=(}8{;Yd zf7kv$el?nY_sLw^w->V%KP$%^n0Oq>T_|>4_@v#`hA~RA!9z>SO}yU!OYRu_c#irv z7vQgN7IGhYs1MP=Mj&sD(X6nCf_mT7vk#4+brVh)BCtFf)ZHT45WR(x9U&(cu$(M_ zy8B6APs53tjCVl_Rwac25~e~6E*Og7)8LiKjo}Y>qnw$knRx>2lw3bD$cZTc#>c`% zGewK>qBF4<>2z6OivmK5}{3 zp;qZa02HJ2qVd=l$Bvp7KQ4F}UcLxciQF!e*5O%K9YMBW0a=T|! z9B~@pU>)WNbR>{b@!1TGv=YvI`M`9*XDnbw4MOPJz1gvIQ#es^BwU1gv5#X={c7yG zrmmkLKEpspb+)y927)KxBvYuxLgXTI<>sP$%QwsDx=b+*a7_k;Q-O8k<~vJ|?52{Y z4&U9d19RfMnd$3Dw;xuj1vcO(WW8If6ax+jostV>_!BTzRY7^z?uVZZlKr2^hJ?{o z@Uu|9gPXutKH8oJ^g?Ikz)c~-!yt=fpv;?^WW(mW3j+XUG_K};2}Wct0I2kiaQSuQWZ*bTu?TG_0Jd$+G3fYYzI&qOK#a1D7 zd@MoUQ6rRfpg`6Z1`4N6?Jp}pCuW>&&op}Fd7p0~w6ybgAK=ztoaeaeU0D((X#I|nD=S|1-{juveTT)Y1jC~+`9Z;i| zJvDDsDdZm(&D1q$M|3F{|A@K2U#M0jeyRO7D#iIwdEJHXB8_|6*Q(>S)*Mzp`0gKQ zRw4I^L=&cz?TjESVkd&D>Xw4An4*3VY=d9dPD@oa45^8UTa>Qe;ysP-kRNO)h>*8g zIRnO3r#nx%v^b0-ZNIc@8(2yXyqo0*0eLcc0wDIyX6k3Z!2|~QWLwI>iRGn|NJ@P; z(}qXAg8i2WoEr3SoAB`PF+CRl2(Hrg+U-sa`z%4$ZAJU9b$4kI5BCZ zgDqgS)r22ki98-xVvZ=QQi{{9=RDR9Ubmia9qiP zY(uElIA;~8-GTLL8Z`kch(p5+k1}47Y_KPruwL}SnEDWvFvO20XwjR%4RB2bh?;Ee zF{A^vw8)(90d9>J02W}cMBw_TgdZA52uz(N zWPyP2q795Wf$^;0bLN~Uu&iYDhHnl#fS501(3?9MqMu#6?(hFFYVA;ie zR~sa&Ty!ujOJt$RqM}o0pR$CfVjP}VRjp&0A=tBNwbUYJDeCoH7+n5EEgqcD@W&3F zIxZpq|==&$e~$i1}FDc2%)Om zNoD9532xc)y5CHF*1FCX{@?fHHOe!;vf>rD%WJ!adDFB<)smBf{i*Ms8~)RtoDi`# z#B`ozM(BbZ?F33i5w9I*(Z1j{1_(G3-ULiVTQM}9gJ*|*3g}(=lLaB*@S_p^c}8(l zTwI)W?kezF>|Dwqp$h-ijh;GjVsDvbrj_qsx5v>%A7gYVO_|{g)q{+O^54~LC6~k} z-t>!wD{!peW+VxnsiSHbA)gl1;hu5w97y-@STi*gJoD<6qP2ChVpT{=QE4fS_1RVs z${~B5Tb5zcEToUKs0qp5p2ifx(Lc-XqGHkHy8RK+PlT z+i6xUGT16C`6+|6Ct4gC6=e!7EPOgMw3Gy)NtDLS7nJ@Bp14Qx$oj@ishK{k3lqSg z&&jlhB7z=~0O}Z3fTnMqZ*C}){$~P6?UkUB00rbU)JoEgh3*fMLORqCgc)P_V5B?) zGO#th!9w>sgeZ{J_ZFA=>?0o@MwF%~_nrF|tE-a|6J$-SRbeR(mn!9am|r}o6LOh^ z{OG7y&LF#^;kuam3nV__dyY@eVPJLY?dKk*Wa5gGvg8`aC-n$ccK{KX3cuUgLS`jf1QJPI7m2(PSHl7JG|< zcj*5}l`;Sb@IiCysoc8nzMuyJyK9zc0S}@H0P}_t2y%4a+n~QU$rnWjW=&%27T5 z)Zaj$MD+F+theY$Qw$ZtdqD~__iwvFc8IW*S)9><7mw|SbDmgVPXdq5)okJ<1|P(t zrEL`|4#G+(&Wc|8l;9LMiA5RPpB*;T50=L*{8%wCSM3Ui@i9e7q|ENQT z!!9%ZGb|k{{?%!)h)1ptjYQLxx68$pz8%D%jmX*^QAAOtD)HuK9ki@h{hJ znhsK`?C9xuaCEA`sgdiJ={mR@f?6x9nA~~VGbFHcOMT{L4mj*LEM^NHruBd-`(R*P z0^5iHI7E^*tzCODzyNp6z&)MFb77vG4D3d*%fCFf<-l?klH5E4w<&>?X(6QBcPk{NkSIGCI!}Y zR!<}uBKG2zk-C$-ZFDgOOjkn0rUB~91RW%k>cX(yH_eK zvCDcrEp+6!D;1&*WVg6dbz&we89|p&6;*26ZMwmc+@lqY1)+_`D3~lrxgW3O2tLc3 z(FH#|`Yp}>wT1eTm-8W4nTiP{9D(Mr%cTsub*4QM38-O%{-^8AtTMm&+JLCOTS z4ntAI>s%l1srks8PL>VoQVVO(PV0~bNkK8O+6*I5H!$;cB@`zL>T#UjVgi(q@t1@Z zT4V@rgST103tv@LLL7S0~!Sypm z|K{AS@tolwHIa+xny8yunL!uw`96YGJ+zJ*N!$cH&%`{_rbc!HKSL4yq2Y}ZmKQdj z&Y~mUSGE#7daQB0pa7ZCOfu9^H!_NnndlXv56|iS@n)$IDQA9BZAi5NVBg#?&S>$n zW$FltZYksvmsU-NVMT?2u_|In1L;jy{Y%%@5!7y98MZam7p~xXY3%iV^^vX?;3$UUe9VR%|20ig_r4q2xVbsW5;U}kOre-dh>6Y>8k z&aF*M0BswRrF+ABu^ZX#&3f5A5sK;}3k4cM={6NlS8w`{Ci~yNoRt0f{a`}ugwy8H zO(!;@9nJFFLb;``&LfwEq4rnTrgnK_M7yJIV?B^h#g7=ph<9yPHs>(L8`x zeIF2fNId|FJvwQp;<+r~IJ*a{>Q{1AV~^rM*q2S=%YJ}UxF~7!0h=AT_}ep_!6e|- z&3M^5PWA#O74jn0p+t4a9>vcu`><4EnuH3(!)Vqj*888FV&;Ed@o*uQxFD0xuhERx zR8;TBEMRHU@|7zKK|d3WqYL8Si9PYsB&uGv7*hGs&WDaGU>v)`>*P+!Bd!P}7f=g& z7;W@OOiXO8hg`V4L=hb-gG7Lys-L?iSF?vPYn>G0BO^bP)u7_qf=tN8{z=!Ru;r14%Q(aojD+PZONzM zle`7kyE9+`5rDyLAAGXiCS|gVomN0`?2r*Eo;T!oy}fzxyH2qaiu*d(6B{16@YTB1 zrLe{Yhhjv)K zhl!xxd<=D)#c|U1J#56zO!7#=PK6gI^4zz1s6JY=b1(EUDwwzpcVXe9a{E!?T(blUg11eF1_y!X^9>xGF*tg39#W=Y zMjlw8w?}r*R3oS7(~#BPfbM_HxN92*+CK);AkB+Kz589Rh>A;{1u#Q8kF|?1kynuw zm-k-#uAAR^K-hN~a_4ggzP43a4@zK^igJBt;7&OHB0JnRb%N8II^WK~2M;68U+A6v z_aXh=|H^0yocw70&ft-zWmEz8F=3@j(XM3KwM?PdWJr%9t;FOFRU2Tk8>NUCg5Sz-Ouui@)nWkejYvL3z z8Q?6lqXg==#4Dv}Jif-%6&8d5-Ea3uOG{UPJ6P25>?FXH`8B9)&I8O|qcj8t)ebJK zqbge%}jKC%^vo?}gzw62yx)%U;1_dh=l9cX2&EJz}zZh=#9!hk!J1r1m`y z?Jc6q|JU7s_~Y#!eY$g^eSr(Rn7L1DJw6RNd(^SbDd;7R%E>KBqUvS}V_?i931I9Z zW)-)YAe%l#{Hi!`n6SZnX^J}{apPeC^hiG)I6;1Y=CC!05O+`QTC08#jCuKfD+ll- z@)?H(_Uu{3#*OIChuea+yG{|EP5a`Kax`fmpa^C(!>A9JtSj6JL z|AF)#Lb-SN78*exc$Ag(d-7tB4C!014}_ zcX#Br@OXcOG}>Q@<;2O89#^)tB2%2f_E3m6jA3r{fiFETsM|=BuK&2Uw*lQKtF6a7b zg6n=OlgXd)6a9(=1QG65OZ+Iv>FC?0cEy70f zjusHrd;IQ^&{fyeUL!H#<{$A3&D*%-L(uBoZN)+6opk>gvtx_C330cJywHMrLCaDasM9DdD&RW!Mh@0(-p65laT8C~OY4VEc9vz!@ z?4+)=sjyaBASk`|>cy4+$A#Rs57qxCa5SZy#?`H)+uGVH;PD~lGOeqe`{>bPLYo5| zzwOxGOCMI{vtTui4=MpNcv;;HQ7HBiF@72&3fO^27M{)opY#omP)ivGjZenU0&4(VyOcjT`hrz5 zD8i&z08q6kZe(EK7P^D30CWIfF~nM<8~o>=H?gm}9r_}6n$BRjt>)r##bVYIZE=vp zK3oz`v_|q3rtcOZ*xfABunepT%&S{S_-_ zfNK+b5gZ$A?#%Ch-q+g;pErd>8`6hl{KV!NWRGuY9@spHBo3n}VwhTMW)OPQ6Ahndx z)T;qxEgbG;IHQMXc%F-vEvue5P@{uK=Yw%z>olT#ZFi$yR~w`v3vjw=Wn)5ECt+iy zVmqH5ppSc_?o`WMCNyg|N?h3j9n4lyT&?3|qEV*K%SOUN>o`i1P>2Iia-UkmV0#(I z{zX>JITboe+b;4OPmY27{kINJQSomb z&&iPjtNk2`Wo={U5(K?{rq#T#Mb9@QX1Y*8k?+{ei${J7Pk3n@;PiY9>90&rI_6$N zowfiFA+a)>SEHk?eF1Xkl%ieR+D32m`zZdeq~*VVNBwoiXTNjDc@%EwFqS3GdWPE!OnnwYU|7<4?rVHx zQnci=V77*xkbFd2vP|O5UTrG69w8nFcB_z((ho_91rvINrxRWwNjO)Wps`9GbD;1h zF0NHvTzfKPq@_3gIvtH{M^PuE*V&*Oj8utN_62uO(pc7InhQ|G8>CIU?>jZ&7f)nt z@>zfv4}(ie46Jj@2F6)t%fLkWvEdm%+Yn&CU_pcXDTlf4={jr@Y?QR)B|}ZM z;M-({?6w)2guORLuaI7XDgu$h*AQ>V*t{ZB3&02Wt}vw<(BL0Hg#vQ3%<4{zqR2tV zNS4>ZO*Qr%*^@37O(PpQS1e$Fj*X&QK!m%Tp5j$sz}DJr=G-}R&Y^bDKvS3)qRcyw zVbzef97tB>#{to8+vdQ+4yxD5Ww~~UNXuZNlF!1B?+{AsRcqIVt9?g|S`D)&e7m0N zsb2Ry@agkCKoT>!NXqq>seuw{tjT?gZ(CXngs8bV*yMVLaf??(lY@Q5)^=^KCCcx~z)yEjROiF8 zMY*rR_7UML1jJWdeu?!45(O|6yIqUWTX8*DDIW#T(Ww=P-o!8TGo~c^gLpT@NRb?Y z>V8jOUv}i!rJ1@(YI&!25wFT)eay`=Hy)B9dpo6C2Pc>`vd}$&ypDi_9-eV)Utgka zp1b?XUxk`}c*Pb-`JzQCU@}SvQ-c5I0#NPC7EaHYodbn}35^{6D@(%pPhjW=EYue6<*Zt#wXe?~+{KhY0{&nl`_;;0*z zXP&JZd6M|Abmgpk_OTsq%rcOcJ(f5DwoC;kcSQc4K+~PLz;AK5QO0bbo!<5QJP

    rTtJ+nAVR+=PK%cfHaaz9Ps(N!Ni`xw+B%MvdxA?yc5?W`Li>`iZ|gX+E_~3} zB1UVsW2xdCI%3E{ZHG8=tj0>GjiRQWF)hg$42A>$ho|E`&3>W&gLYl3z)5n)aa^ol z1=5kHpq@7K6vSQ-R3BpVzbJ^fn0YxuwjXx08(A#NMHgBWP!15&Q4*@6FRq^jKiC!MV&s6I z4Zw6ekdg~ha)h_ca5pGJOA%8@YN{|3R^9H$JyBRK#3EJ_Q|1BK03-^d zxw{RH>o&lBLf6;otrODc`c{@{SoS#_e0R&*oa(xpiRe8Cj?K8z%>=?&A_?LtB;0d}1@apE;b5JzVw z3*9(*#wWoaAQm`VRj-m!-f#^!w>y% z2m%B-nwWYVHO9KZGN9K2C>igd;I3fOFQGsrmZM9DWihuF{#m8Ig_vUD^@f#M&2B|W zfiKQGs^)Z-oYj09el!w>35lbkztE~F!+I__&ae^?M3vKO53VQ~Gti3ne1LlX5AZx6 zWi=<^wZVdkm{~^KWg!R}XSpm<`gcw6n_q-!2pT>9VEN+A#_)my{;!tre#A;}1V7;v5?NJSJV*b% zqMP&zw5b2|kCPlq9^!dVOeSE?)KwX|6pa!6!H5oGTX1X;yDe;b_cx+7q-E!#vO?@6xSTC=K{qq1pNE6yoC(Dg6a5~X0`!d%b_(T z1Wy=KXoWeTJJf%dCT&I}g|k42T8Aw5G&sQ}`U9I5A>$}r!Z3iM1m=GN((nKn{1>8- zowIA_&iNqE<9jY7)g5BQ>o;${hMOhn=RU^w-X<)(2FZTR2puA^-@9$R<|F(y00A6A z$J;832`cgaM!5ho`a%Sa*DywU@#4i;5?sUJi?kJL46C@gpMn#sG6WwmE_B7KXdQIB z{^m{MR?Y9gr`loKWZ*J&52M0R`~+Q%nIcd>xdkM%JsbVz3SzeoR>F_oI+2PYKOdiS zh_<}iuVT7rn}yM|q0k@_>9-k;aIe70g>Rc6)^I-|ce&?vsHv%u0hW17-X=gajKa_J zgJ2Sj9nE{-%;lS*sYpgDK&7^jV0y@II9G#uShjkAp#24(R{}i&LPg}qu9*2vM9IK?@CJ<`ycV9Oo$xY(4_|SVO)I2Jt6GPrMr0`CZ(?!eSp9 zL+FAnMcHuqC5W6@m6!`U&tdRfl;PeA9si};GsE3#fET%t3d2VHPe!U%Cu~+txg@W` zLL!^GYzZ*>C5qMjDbJHTv7Z1nWy%(8Cj7`k+#}B4m>m28vJrW@#FWJ*b26NX%IOZa z(#c(f`1L|Wk9o#<8ZfpYp=&s6C9K~`RKbSFJSun$@6qGF1bTOkQHr60%o!>|Ra?Pw z^5wYiAE3BZ*mZQiHlA(;6ggD-&~X)}^At;K z5*;luEF$Ry;zf8}&UU3J)SLTEfZ|8g!04~&?nN>@|1IP7F(vo_oq>R3(=BMR_s01v9o(GQ*hwWC>9cIURY&rOzm)sXmC z7Zx@r1fePw<4*kQK4#>bLbmlSmy-s@aX`FoX=;6Z; zpRim9D-UrncW zn**SsEkJc3tu#3U^L7r5p=x4}Bj~H3d&W0-gTiM0`t=ZrI&5mnr=R0+3q{@+;Fgkx zh_>62Z(D?D(>_=;-Gq3%TGt@&h9i8HH!Qh#m{~GIT=+`-L=~>&K6tQ@G-6fWl8C5W zXiLXxv7q#Z!O&s`tb{VK7#{yIr{po&-E2#UU-&vc^*_kPH8i}3yAu%=wGCkw;qf^7 zv{q*%Ey(q-<4(wEclP$WqP$Ry_no~3*e&L`889cWmHuXP0G7poc%H&tEg$}inwpyU z7;iAF`ZFl#DdulqG50xvT{pfXfV#0(7;`o?s@8T|4K^0U>kbt`PjxhgM7)5wPrlf* z*v!lfWy%*IvM1!_^AWB{5s8?h9qkHtO9C}ix6%JUzODnF%k_QhbkaGMrbvV4F)E{o zWE7PmJ2NW^5rt4@HW$jAkjONXG>02#y#hmvwn zhF7fP)S?3R5$tJ~%qI8j5BO-03+jLl`|!LSJD_~J7Te_}+G-_))gL{2lnH%PR71xq zfxywjgH}lNcqTV&6tv5^>>E*}cP+!M54fizqrsG-~m zrS)yNCEUSH7nf`lnBs&!EW8Q{CbXC=f-*~lwaq&?u$L%AMNl%*jK6erEWyV6{N>Bm z-MeXs6GSx&=Wt7+C-O5yY6B4uI>sHh{8>k_V~d+gSNEsHQDu40v`OH+;-E8k46T@) zO0ffmpscED@O1Z=jmt*`?0oA%h_)Qbg}~*n2g8SUh`V3)^rR<1==AK68$0AJ#hZ~`8ZV$_m4&5&qk-b(5UHW$ifsL!!okYmoYez7g8E`r1mi zfj&~SvnYfr6MF$?QFPWMkOEe;6LU!z68lhGH%n|uTk!^TpL;80KI77!Ht)YY{Ni;ErsDC>Q~h9Ih& zl@u%bUCLLnr3~X=ZSFvGir;_-dT_LW7f?TXmnbwcgnWrnM zahEXFi-7HVbi3I@Es0NOY{-44O4JKDF}6XgQ@IyBFJC;5X)P2 z>H({RIu=<-Gz*WL&f+6bz6>sY) zdl>1oqJxXvQc@0(pdF_A&=4lv_)xd|;0xZu{YVCHV6RKWz?zYzK)Q=al9HSYQypjJ zOBaS` z(#RWlkQ@Bf#-k74m7s?D9{~y4D#Gb~pqiPGUj2y2kPf{NS7DrN42Q-x+JsIN0sHu`j|9YCHcnqwwiD#EZYYou zfQzs@vK`pAZ6W4#*4S)1xQm&YnWz^bq`E)IwD0^nC?JwlFru=vGg$>y+Y*ElPhF=D zFLHqC1HtHRG@~6YPe$h)rZFZTq)REUL?}#zo&h$~Pcmc*6SGLl6ccJF?2s%=3qs!` z^^3rPPzG#1&Q6o;85)`$17$NHpl$s8szqf}5wJ_jvAQ%s!FIH37JIPDQzNb? z@Ersvk3@Mkr5d6%h>R^(!+Q0-zGG(3cKip5zEyU~9ukZKI(& zhXxpQO(TC52^PX~lSZ6H6f~Fp0XI>}Iv@%t1-R$V4la->{)q{m`|ovF8f*q=2mwVo z1KFzB;cwo*zi6-A-PU#!5S8D>Pj9FTXvG2Yp_>VzY(l;DDP?6XOlzSk^t9f(GYbQK zTqu?u-wR%Jg0a%>R6J4Y?FP-hC1FZY#&I^Rh+YDKNaZj#Z^D*>LO276GKobdh2;nd z=oDf*@9=S5@*8{xfQn*$BJrLJ2Xh*@ZXCliZj|vrB47JaXEGSIE-eK6R&|!NyH$zx z-dlZT$=8vje2`0$#*M&28OR__!UhnnsuEoS?i_Vv+?`mDR~t2kG-MP_T8$kYVU%={ z$4IB~*k1<*ZlM1t?kGBNJrf(^UcJh}R=YE7R#b#{J&S4EA;Lmf__%(2!9g3FRK-q= zYu&+JW>r=yMh&!yq5we>yF5*LVmQ#8vZHWx9mJi{AC{>m<$Oo4wIPUaDUu z%m3>K3lyR>1c8{+i3gSXCBeZE|Bk!RE3QSs8zs#JM{d({>G$NKVV(*(og?@kq-;rY zhV}b15-XPruYoQKX{;M*?2-}1)`;JV=xj*#x-`we*S;gVjW%LSP@ywkCV+wh@_uwZ|aCfMy!FtQ}cM_#_{ z{)jZDmqieJ(fAnt0f0{o$H_&I8g4xr8Sh|s%}8B_89L!?cELg=o(z;-KscmG$7^E} zpt?*HTWQjLdA291xx%SsQ7AazlMPl67SFykBY2B8Ix47zsgXwL0JndM_5ywOds|LC*+)}ZSs74?z)>4! z;rhsy!P9^z-XJm1B53e_%+%gSVR#juV_(O{oXrpNNUIatAp#F}wVpRJipE<&y(>rS zG4ema`iqEG0mash9Z;bIez3%k(tyq4gb?;kMMZ_y<1vuIEW`}O^pR=fX?lDH*+KjE z?K}G9G@j%Un88c>Zyj5|sJRUSUi1IzH2>t(@aLM;Ya} zNJelC^UeRGhfijr!;uOV@Ed3cJ=Bk8au-y#vI+_cHeczT=m?L30l$Xbv~I^t1i-t~ zh~Y_`GqyBy&Fa;&Eq_BqQ=5jy0SS*(k_gc$ha+99HZ$6%LlW6Hx1Q-C3xU-FnHF?U9gnV!bU(#MSD>pcTH$imTg`QCd!flcqXda%g; zhHUQW!GoD8_MT)nN1TjuLf)s1P5_kDpB=Jm5EKx2iDo4TQFYX1>>W0vvbi!*Nd5`1 ze`>*~@w@N^&FJ5MgGTl3?Y#&y5z%%=Xo4d|s#6_q3iCb77zNe=NRXDl6)eNTiAEw) zszi1{T{t{2kcFnJMhrpRO$Aom4A_2O8<*R=y8uTy?ItgmbOL(51}yP7XNhNSLBUdt z@i_zs2fvO*%f(CJsnByQ+Dr*t8iUx33;-UjE5c-_5>PL~P`J#;2z^4p>rcabQBMeT zh)zyUlFft9+8{t`a6~~Fp{VZNPG+Y5;pBWg+IB&=F;TYX!7|7J|~yuoh0_aZZE(SHUXvI|<< z+mS!~aXjO9OzXxdCz-XtjU?x*MS@$+lLfgqs1yDQKSr3g{%S`-EsB96Xs1^^)p`ceeq z*@yx*dlt;ZMxu%J`Z#o}=}~W>wB9^OGO)prhk5Y)}Pd-rZqzUco`*Cae(mqS&us{+ zN(65uqTnB%T9XPIkP)(h;TH!vJ@^o|h(dLq45Y;k)WU3UL7{}CBfzD$7xWFGmA$*> zQ)A<16>8Z-#Tafaf1x5jn2Zdq5odAriNO|D*?`|c?r{rbqnYr5{Q`K1(TRy>fVTP# z>1m&@_%4BoBX}QN&)&Ay?Wo}7;D|(Hb{q|J47)}$5VbFcuKF8vT5Cbsrnn2)6~c#2 zq(SVnHRz~esL&Z04O@{n^0;>|7uO*m?hs}7=jczqEra&d*RXe77-P7Rst49F5`x=g zY#+bFFtbMOzasf@KhxpAKDHX(vQzlG;ozA3YR?zfE?lNwUbn_yF{SN$FMx&o{{Ohh zxM@{&HBnmrm}s&MXyQ;MI&kmf)|X@dP{GBe4OKWHJN$&c8gk}9(X`_n2p85Ows1oA zhGf1n{vuZGDbR|YgaYyP&br!Ka#GUHENuZaiE7s^)xA17%SfslaBEpGD$wZEeB;h4 zFOz!NByE@-XBVmrH{r$tfq7YVN36gxc?)_IP86lxcb`)1Pu+7ie>3{+v8bXLY24yj zu%l>_1AeTnCV>OQ9oI;E48)2{1yFrI;mudl<~HFAhOJwG!%qWaWq}-V{kYfT$6HX= zCa;R{v?rwd{)Q^fM*d5Gpuv$0RcvQ~@QU0s)Y&YeEJ0LcnbS|y>`_SLan;s3omN|IZU zGcb@6kR==Y=T?|YP&x&M8tqOjSEYG5j0R|A# zq2RO(-_M1^3LG(Po!R~%4<$++V(Cfg6`3Nt%0p$uL)oyLFTs3I_*$b{Q+jc2QLLZ| zFvj=3!c1tTy?@r0op#A091RB?c7HhE`>A?XR#rqo9)Zn`&#e}R|h47&bDf^t(R zUyVI$(D94Daw+W^sWYGAUrt2X?x{bsaGR^{>0>;Lawt4EY7Zwr+4x_1c`%{v7eOLqt)8q5f#G$;} z+Kos^M9?>UvAuK$xieBQ1b+w0@quPD2&e+_s9*c~4&mJnaz;<1-~IRBe$;%iT!Z<)*(p;H(&@EZVWm+-so zRE=YS0dxYDvmXK_0M(QFwufw%r^ih_4-I9tv9W1HcPc65;{;=R;CE25+vx79p>@Iy31>45=RvTQD7Lx)b#p}S-kAi9(<2(%KV)=_xB-|DIL*Z# z;z9D7fG2SSf%tIJp#l8$dGM{Ff|8O|NSSjnxU-Uh;TBB%Y)8NZ3+)o}p+JI;z@sl6 z9NM#$6;6f@O@C|61=moavJnlEEJQ<+0eR^H=o2R3C9oTkXOXY~CHk;R%DVb=Uv%ly zVe1}&u^bZUYbCpabesY)Am%bX#lj2y)AE`~hse7?&=w-&f8FP~ha;jv2qL?JNV^Ha z18Hs`P7E!*?d234G7Sw4{gQx12q+fi10JWaKGVse>G_!0_;LenSgD_L4-Hq_;A?dNK4-?`J zu9B$#{X52=zUy1Hw~p?8y_{pZ&zlSxv}677Yso9rYBz7&COSL22#967H@+{Vgw2hk zcO=~$WJRw*2H8Ig&mkkf1}w1|?;Xw2tI&CM7Y7e)sCe){;JI1QLX87NjoACLmDL+W zb4)98^&!tgikNB`dmu$ebo3A?9nIxu&YnFKBzNeq2d@EWqjJiE%Gzyu4p(G8q^$!5 z$-@xlIgQ$Lmy^gM8AV03mnL2Xz`GR1eo-460aAZJspAG-^MtI1&eo-VE!9b{c97Entx5Cixp!W}x5KRU;g6&;SpgM&b^``x;vx zJlGgAClxO>b#<31&bR=%P%=uYmTtp<=<^NS@ZJ0Oue-a`o;h=d^cvOa0J3s|KZcyi z)g^1$)>#k!hPDVLJ`BF##G0U>OA5c}KR$?RC+?hkJq@C4fC3Gh!hxjT!o&cg^7Sn8 z&zGIqQ*VEDnP^~1BO+z6Btj98DB ztFOledNG1=_{522060m<6B+8pz-id0x3&i>D(@AC2viHQ8h zC}cC*1Ga9DMez3IZR}KBhT!}f;K7fnN>A&-5YhY4AxwHnstkPO^`2u7{#@PN3Hs z;e@+*ONwnt&4>U{7UU{x02JrgvCN3!$%(l}m52`QJ>~$Oj^xc4@&KU+ir2_CZvcMm>h6|( zq0HsJOJz0QfhfwtG7+jg%F4>e_=8WbUbBYO!_-nQuR`}6-h-8;-NH%RWij#bwwB8~ zQG$end{bRr{qDhi38dWYAM4T`f0mkuX`kZEdVci=9S6mSGl@C2h7KWGC2@8Y3pG@<*^zyF`FY$2ut+{$KL8zhQC?@cgF2ZcO{ zUly)iZG~#!rZ|$0|YZON{|2l!w0pf5!Cd(+hQE@|iHKXnnuG zad+FE8739d>kpSxQ4Kqyg(`%V&%4AEcLb?^yn!G8bW6uzYt~mUU9@AY{;+S4j$5E5 z28h2lc;4E?+n@1z{g40pSAVii!X)bMkPjs#vQ5YDY{QYPs(RB?F_0v*_`4H(Cb&jt z#li1?`qkI!&`>-c9|3+rNUF~Eh9KsgRQprTX zaJ(&Bmy0N55ydXDXuwT z#m;Qy+Wp@!*|*EcF6c*3`0IF>kJVDz0wTT(9YhvjFWw6NzCi+{3-vCaGW`DHe(^Pc z!x_P{{gFim{9fAS(YtDGJoLKm4R&-m3tXyxaP92w-Nk;}X9OPp;L(5ehj%N1-`_w; z>yNX`g0=)$8yCUS4ATD~^AuiSn#p>E@vNQ-lhx!%k6^4`?_;$;em6a2c470%MXQ!A zGDyrTbgW%=lRZ>BTW2W?n_NWGey>>Wc2?6Rrn1L(uh(BGvi0|*MK^;@qiGzjK9c$4 z@XBLX|2W;_nO4LkxVV7LBk#SBKrFLwk?KJ1vWyYi^nvND>fZc;_lU43;&L#~#<*$I zCL}S>9tm7AUB)DoDm}u$l4eVd<^tDb{?-vf8}U`Qe%+63KFd~>9c|HWxh}Wsir-$l zQGKNj@3886#y-7-xZFjtOH77e58XSf(y{$t(ur@sxw@qrddGL5nAGpmi|G{xWYTv4 zilnt0B1@KK^RT!=st(pXykMjkiZaRFWchY?LT!J}h&5C-?8m2=CB10Lt%>*-x0;U| zC<)9Gt~f>}m;ZQ!R41iH)#>OrRjd?C-^Qew94u%@@ktJ`%J;arW5uPMZ8cWf?OfUG z6yM1)|MY`YRLfLcT%b-e9Z%74Mk(MeEPVaofbh(Ku-j7hL@lM~M*yk6fclV9?BY$A zWPB;-3=FKwM$lQ|M%w8`R|pP1_a&T#Zf-W8x7lh0bNGNgcC;eOt|Gi@dj30ZFnQ3? zR9UvZdBD>A?Nrgovk#$RHTy>1W*A*?*(uGIT`rqcz8?#jxw5QY{g!D^^2qzMtFn_{ zB=au9-_~=k^ld$0%_jR!X=EVP^7rpce=q`aXtq{7BW;0h%P;2HTV&DjqP{K`rMqJP z=et5tdms6}#*@}`{>QB?yTc`!l&CAk6={{YGku`oLtAJ+)AWh6E-7iJ}mzi%9&ZW@?;kg+p2PBtm5zU;uPP6oM2 zze=E1(Mo1+0{zF(m5c7|bzGU4n?UP8WNLrDMqwp|lDkT5V@Yeg)x@AbjHNCerD6;BFe9&$k($d`G9T(vx)qvzSbi$MN{N=qiMHD=UVbV;AvfH+qq ztHe`zg+ldJ_9oWW*v8we@++nSpU?iw|NUP|3VKb%gnaF5k^qS*BVb`?8Mg$}s>x)J zWkzR37pOlOF)q+Mpt^YD#e-2_fa$D9lG-`Ie;^P;Wq`*6!^06hVK1wlsVe5T{{xr5 z+VL)E`Tu?lZm%)T^1P7o{s=BtpRdDC1r=W5)A2int{7UUR>~fK`tN#u|0&KhKy-F3 z7xoQP=2w>sC>e2ZU6y;WTjY$#vH$7Za!p~R-Xu5 zs2OY5K4fF>J9_{`mDtU{U-{eDl$;Q4I$^$`b(;LCE5~-3OPWtjR{1=#)G?f_88+>p z{pp`mUER|BQOB*jq88Msw3-vulz`x1y-d}f)PWC)kjJnR6h=g;F*36ycca;fs4wXA zrzOjnL75OmHG|Pp-9ja*6}#XaU*pw<)Gue3N%f(`Pb4;(kq-)DMlbZTzKu7N;CeOt z7Eft}zkTBoVL~aZ3zh%*wx}HcMhpi21-vVxggQFZlI);jeKOu(jXf9j|0dj@Ga@7n z%@G_?p8heL42tGUtwT|Io=A;R;&DYiX3!T?DPxW@qoHn{~+)U|hRYH7j&`c>qi}()npY{M1 z)j`P(2Q1Eeo945ofr@vyrWq8CuFwpKXX_M(9s*#10A;kKlh4<*eDf#s+k0Y+Djtra z%lB`uAMhC&w3-$Z9+&lorm6`9WEIX%fqIG;1(h|xXlp_{g@&WuveCH{R!VL(v)KLT z%-ndzIKZZi#?%TaDn3K*pcR9kK+fAo}_XN(nIXeyQ9aKSd`29<&K%+!b* zd>9$`(m~NimW4JTKAuNaRW$%6jF|q*xZ41*m=s08v6y79ReB4i-nv%nR1f$HZlUoK zH!^Z=*^vzFF&jIKlqA3g@D7} zs~HL@q=r7&J`P6D-HewTui?MNVQRzHu#;{eq!++ok+7B_`6$jypXGiM%;Ds_WBfk5 z!T#mCRwl)pY7dj2TWNIP939==-Iqo46b+_wt#fX&+uG3c`Y`k$3_2VkEyKZFUsJZV&_OS}Mu=K2jKS+}|$1i*xYqsswE<6`B z+H5-zDj(L7BUxvX)A8zKwVc#kw;*GNiY~?~xYMyubHEjL{}p^CU+0V1CLEY^=o~Tf zj5whE_U35)NIJdmL~Uz8db*uw#DwDQY#jQfnm17)e2cAj{}p*O8EezfpGfW2c@N|L z@Yt)~8CV|Sw#)p)TEWzajb<10ot+YfpGnP$r$UYgKebQ=XUJb@ecO_{i?MN)huJU@8*s{{o4aC*T$7WN)$7SU z9R%U_g<E>C;kN!)#|m&BAm~8SW5D4cl!O6TS1Fm;C?#oETb2 zlAb3EWt=|tMH!uyw9eTJ)H!^Ma&IyFtP>u~lWqc$o(W!&*+ELD3(9Oc9!eUd2k<72 z-Tscw&c&C~MEqAWjTbwS9a2 zu)iyzav=opV%)YDCC41N5(ZQTTtT3ci-K;L4Ea2&Rpvi$fDjrTFVpDBaygsM=?|k$ zVkyLqtYGTCEm>1wQuSx{I)(SYI-<`266B(5MH|hZvS@8aLrWh5hHQCh!L`uEJAux- z-5Q|L)<-iH+SS@nKy`F2!ryGXfrfyJeh7ZxnNZUC1 zuTmGe4Ea^el%&RvY5!Gwb>AO*djAB-NjBPieQg3A+Thvqay4PDBkq38na^PEE;qP}fIOyan^w)aH+IBrKi^6;+osJ-dq1Fmfyq0c_d zIrX!b3Q1P`*%2F@VlQZ#9N~F8BF&!jRiwkKBdMpaXLQtS_;P&OeSVV4%>CsT)-lIR zaBt)dWmmKh;Sw@CWqrO{L+Yvg+L4)8%yci#yD)jn-Db7^EJuA{<(luk0JBabC3D*P zOJGvCT!r=W@S-p)VSH9&AxTX_9Cy|4k4c_U%>4hLe3Jd7j8yvQn{JDCFaRcyz5Q3sB>ZT;+}iZ6(M{^Q)gSDpP0^#i!W zHNEU_-CEtXFOe8|ZC%UF3=PVvhRGvhgYG+j5$DYQ;P|e&_ta_IJYX^qAh&UkZrgdo zsRfwJtaH!dZfUAc3L5?Rp6nm9qK7_YI}0QPde8IH zwSrsby!64f{VJ1g!sl+zn4cHr{wz;9nY*T@07>X8aZBjoik4#`@>YyLXA% z*cpjUcHh_}8~Lj+W%fF$5&kCP+_S&^_kTQRFX^z`aR<{f?e}9h7q74x@sH#iaUIt0 z4Eu#go&B-V8*>inM?oylE+6$f*}WIY_URnbLwGxg$BGY{ZaL_9s30gV;2B?~K0n8de@UtTf+cc<01WS7)M5 z`es|Y8s5x6F00bun_Eo99O5e-oA>Og^hY$V!2as>gNz<>+uoT4bO$a)WBxxDP zv&c|aXZtD!&GdwA$yRK?NIz#UHYd8Is?YEG)7+%&P0piav4?oB$cB%be8*L+Q|ncJ zk?hZ2r0dI^H{O`B=d7FB3$>5yawXpfV;}np*Q{aEsG#H6=P4ij{W*U8ni4(7D36Ld z(#IrreAf0Jn6>VGol5!97tfwLZxh5k&mn*LWRA1Bnm|(X$%*w%F>7wihDoTG*>l95 zY>HDb>T_;Qp7(bA=iE+Pg#7c$Y?=XyKG#nw9qx99J+to|-753Fb1mKnKOb?b1JKa6 zd7+AM$R&I#A_fL6#_RFn2j58JJw&;Cy|@zYP>kB%?ru5r|wzbM(#Up%$d zRobCnRsg-2VPpF0Hc>HAhUh)ifp?2yu|pu;WAcgVH= z2%9SN{?R!!&Qt}n(o+g|XD<$}43cV3_!4doRzm)#>a< zt~v7c7Q@9%RVuj7Cc4vE^CXS47ZN=@XNgaq!V)KCt#7htYFO!D)8REeQ91QAdNln5 zi_&~2{OXp?KYLCp>U?kEFgJVeQ@2ZkcA1N3OiejPj&?oWR8pco-w`iQoby=gjxBb} z%avRz5j~?3pAm0SC|}pt$72*77&u?F6mmM=-eB*KyCBuY)~U5={7Mj?v9=KIMR3$+ zKY3ICVJCRl<~)&w!)b?IkpiLX^qACq$)+*l$bIvqx~bvSyw-+?W$slxY}lPM#+w~S zhf8Q=HJc9@Y3fJ54S1s3?5o)}@5(${FZ`V0$A_}0KexA|*=v8$$G|P-$7j0Q?u>pc zG=iK~&pbZ-AkCauj*9jDdSABAhSg2H2Vfb-U&JyuaeziQYUcQs58f)rX2WEj z3w0Ng1CwHtdzUz_jd>*aI$9HV!zn|8$Fiu0I{l14;@Y#fs>K=Ix$1|TzFJr5E@*1f zd_8BhJi@5&>nqa^Z=vwRZ|!G`$S-y+n_)$qbO7|>#X7~Da>cOZk(KRu9i_FM<+EIO~C--gzd)1Vg(73aXB=8 zzVKw--*bGP7U%v>KM|Xex7r8*JqL?Oa8e>^)c4-7WuDibxp}U4xKVgqzC(K~H0WdO zqvU7ycCmr0ofTgp(ixi1{88VVvo7oIEN6H!yLsN;BaZ4yP}*uHl{&f=ldtAW(m0RJ z0lYG`ct|?o-LTH-PH4AL5{eQ`nq`cZLUm8)360nNZZPrr><{~jYMqw#!+5Z z;^c&18a_Ok-QA5ThxsDfgW5?ym(4gouI-FT?D|{Vrr(d)&{FwZiH$z4iR1S^%(J!! zMdn;YW~r$PoxSa0QJbj{ku#w>AIpxEz z>PW1nS6@pP^-5{JQ1Sg0noIc#PFU@bI#od@JXxmw{^L$x{>4B1j8;M^I$$y;TR56gIdes$Q!Z(Hgy@h9 zn!WYUox5ct#Dvia>t;jHpo6yO??-4v-k{lQ(|RBLT#4SeSC>CI&^JNF z@@OgCL@-G1e0@&$GF#U+T zbVg}{8FSR@fzOAUUHf|uOpz&rLO7yok*=~w*_u(EH?i#k(yxTZuO?BtlCRut(cBjK zS#09pWqRo>0F^ltls1M9EP#?NR9l$h3EkN}kZ%u5oajjjOiCKd@a_aahF_OkJe2B3 zt55=&dH$G8XaIR${9^6LE9eWl!(l&GaN$4##HVtw5OuBQlNsK1h3J*0geMtBCjEuM zeV$AnDykGG9XG*!b`v!X>D{#n=|d49s(sDQvYszvdMK(nGNq@Zq6! zvBo#~HLDqwXxB89Yp)uHN1+C{%}ro6{&YbA^{qa9z|8BBu^9k_&IEEzsUM2uYr45j z8?EtNK&=srQPt^)H0>a}3=1RZsbzLAk_jbY3#XY%q=8m4*(fZPPU-yQ=s2K432 zh0uC@3st45*RA_deHtB3$VginuGLuaGhBAn|4SUhP_(5E9VvJj)rA_4e;N94bMJ+7 z&b>Y|cXa-$ez=TkQggURDE-U$O!MeKm(||)@$VHLIMd$WzQuI6dx1?#zryyzO6l_K za_M7#ZLe??kxMpRm4D2^mLabNtpp5vm2@lwYd(A!o9aueh`OLz)q1|wyGYk{LSCe+$yGp962MWwc7?_PP-3MlB|-}a>Ni@P9*z`ERR|JgbDF&NAsLM6rL z(RWl4Z>U7iyGIOd-5^IiW8pHR9j*Miy`2@s^Re-9$Wls!Oa~oZ_c7Xh5*=?9A+n-q zL?p7r#z`%7&{=$}zQOq6y?g#xW=hxkAPwkLq)(3MY{~qNGFA_io<2 zo2@O2`ri-`-B6GS_4l(?Q4R=zb0U4bV<+LRfI1O-DdWB;k|5UT;9KN%d= z!pPtJmPOZe$EZf&zx={N1uLuM?W|k2eDWGVYqJ4JIB`CVG2|}==>CD5o7;vBS<5gH zBag+c;t`Jkj~V95kNMj4-$Hw=cXC3VF$Khw@euLB?4cM(>euUpvD7)csoc8>gCzIC z5fg9PBCZml7~EqNuez3jAr$@kJeb{>8O;?g2me;EZexyKyy_78K+b_rT3k}nygD7a zKSY9e2q1YFn#aA+CGeQ9VFtq>YoRP_L6b?Vp-T|VU;HvYx!d5f!Pyxq!6YHE;IeLA z3hU}{Nh=fiNIA3a^;QSyduErVcJ%VpQu5;m+p{{a{Cgeuqv-o7L0!swd{M>yOVDsG zsHv$5wVt)=U-d9xff1v7=r<0bXxgizBFdApnrd=Kcm-j2$|_7?px=Mfc+|q|u9M>M2&G#=qF2ni`@V)-e%a*I6C70i8VeCr=D5Hc?$v1{*OEDl+e_SBEB? z9N9)5jfX*lvxJ7mqfmEmig>6FUVf4Nv^-jrW{ApzeB zFINU~bPo@YmdY+vOhQm8t&Y}lI%Ot1j$S7z@Su`~IMolu)HrBD5#4nxnD?nYb*#L+ zN}%3^Md3|ZqUBX;U2*37v;Xkb&pm{$(Vv5{pJHn+P#7t<>q2jSS(mTb*@PM#n;-t^ zw?D`jVAA9`2q!J_>Ot9Nk(WOE4Zbfx$tN~9d7wo7(1J7D0$)njgK`$5Y8w!2=m%sYe#HzQWeXbHS`tKtbJj&1N`sE`~ z9aLG_1{NO;yY`V>a_^&eYyxeo|*_vG!LFeN|2mHoF=Ug9XDEAa-3aJPXdc zOX6riFIWk4JJ{jp$jeU#|H3)Yz*yly-bY^7CK{=6AZ3d+bmzxpixnaW@%Ki&_R}_n z-uufU9%EO4iT)8<+UvZmcPoY~pzc+H78v~OYF+T}c|ioZ>o1KJgJ?&vg6!j3D1oak z)`K)7bME@hdGkNwr|xqlT_4id*Ok`b5&QG}ZPGSA4*%`R75YQJdPFJ~7aZ`m(kV2D zHy>VSVO;?*z|EQbmZA!yomZYCaMfA zJZ54*Wq>fUXtxS^=8n+dOHFV8O!&+&6lFgp0awJ6LM?t17S{kFgf8Xob|^dIANiDG zlIJ~kA?RnV5JH_vFf&rfESUE#!h6zMxkVYc$Ld!P3wTtzrP*)NQ*c$V3U-dN5!kIu z`SoG$gTyyJo874rb1+10mI#5C$DNZ`h)`9)7V|9BmB-W~Z2ct+J3&pc1CJaA-6H4e zSx=Pjh50}tAPO5pl`!6TW5wsz)?hI7JyBn@oCxm1{uf5dvZWy`ynXvNTWR$vUU93_x;fJ5#Dn&6EZ($&$dsHCkOkTrou?@)wLknu6RH_kwm#y0IY zef!Cr&iM#zuL9B3f;e>uh8x1vl43B@5DXF*DCoAqIx9fD804z%-rils(-tY}@Y!KR zybq!*P7|MU`q1gohPdARvN{Op;Juh3zjugeO(N)2h*aCJo@S+h_H$XVxnNW@7$=-4 z8~UfAcX$YadWSPd`rU+tgseL8>tK^^$HHI{#DC|Aopmd7H`dvbVqKvy*cf_VO$Ci> zRYtdUv;GV>ePL19*~2W%;av4Imio^n#N#j^73P;Yx>y2nVyI>P6*0sW$zO_4rrmKI zlk>>~9}YY%orQ6(Hgwe_6m>13(z=ZXa7{3klp?M+vRJ$BB?h$8(6FUTVeug9wootF zpV|XyCL)_U-M=mrBc1%~jw}tkC4uVZ`{YhhKo=H57ydP*vTze}B?D>zecXKsia2gJm%RyXI6@6$}O~ zBcri}iySLRY02j7vPk*F1G+$~W*1%}G3hh_Hh?g=TYYp~d0}K7&TUx)NWO4D%eY}S zo%Sw$0C`XhF!IGc$8eMvv_3IZ*;&kwD0-;3K@)AtC1?;CL@2dE;|T&(#L7x&P_|%M z!|MnG!v=z%U$K)RUs_~mRN zx>;$8q~QDz{j;C{7VWxg2nh+1s0thU5{wNoRK$DG($PJ~z#XzZ5wza@jpMKMY^n&f zY9KJezyNw%m~f~Jj^q!3AP*zVJa}PQ_F@PpcrZL))g9Do^$gcFZTq-5|F;#f+g&!~ zeD4LA&w!{0=KH9YG-z%)L>C)QTXoHOpWwKpEBo1+_rDBKSOSO)J3IS_VqvO{mYCWV zHN&*191c2f~4}M3-LN;)tY@ELip<8 zE|mZ*$Nn8N{!Wag&{?10ULC2!kg?#p_Xa7{7C)BJ@oxgS83JiMGGbtg{Z`rPONI& z=4muu!@Hl~^W5H)*_OKR|L*(OgSX5uj3{ab>;C}!*3LEGzQLQ0aKjJ)!8kZU;1+qD zDpeyC2|&PZ;21zOw-_NtJ%tWTT_x5Ez%b9Tdl@Q_2i;Lm)EBbc6Ys6K`%9mHb;F6Z z_Rd1~#hONjSG?4N-q}gU8@{XF_*6XhvHC9|cfZ3)7_9YAf|pH7{PUk>k$DAj9dWWH zjFa2{a;gNhA`3Rsi?TcUv@xfKNyWREq&jQiA$I}ud5(G*6?}pHmGUeXED050rpp?@ z%QRPgwJqT_nz_pT4lN8mJ5vH@gw)AaG#2^2YKC2!gq7cC73={2{TLlZlqe5g6sg|g zSUqXMp>lagU8DgKBrtB@G<(?J{QYNsULIG<5}B=6=GzyUGfM}f^^ac=HqWRk>tZ`% z@mE&l7DKa3yRvoXiFj9U6BceLCIo}(wcmR7S*$U&9}e~^HaLW@>z@=p2R4!Pa_thb zhg>!tP{etXh2CO!oF zoWJ1D8qU~&X#Bcxxn3n35ATk)mV$?CRL)rmK;K}Bo=%OL)|q>@DhPl3dZZ;q{Ss9_Hx%moZ>|& zRTX3l(v0UASYz-KgjO!5b!EEVBdq!7$Wx+xF^Ycs=@miH{3#K7wL%OIYGMnpcP|{U z#V0v_feGLNiJ+$im}7V*QyDPB2KVYY%z7ID4Z#f@K3uW3PVuLa+4CQ79wIOLy3Cp` z`m$ESD}7BBtOjc~b_*(L&GX{rvMMwqzX4tDXIBVMzC>Jq{<07|yA-mrOQ*L@svt_m z3kgER3qN1g#n0;w93U$e*3kU)gBc2<%B)z>^z63KKp6Q+#j=hXGxgluo==}fz{Msd z(!6(n@E{Pe<@#CE2Ug9lrlWEo(}r%KplB>PBPJViF(R#i@CTR;<*8wL$UUpUwolI$ z59ZCA)l*S)qm{X%Biux4-0Phc^}ZapcY9Ydyq7ZbmGqR&`zz54bH<3Kz7%_5U6Y?!>;1XoOA?WJCId`NF;ie3Taa*&+XL@Z8I^w;+7v5^QE|{e4@vijWs=e5Mv615l32 z>->OJ#1LRe|M0Md{7%2meVp?|bXPOTLyk)^O@#+W?b7vR4cHjYtv#h+#D9+G*W3Bq zb6?#Pi3kGM-zNZ>ujV)VAVph-VGn;qJ|yU+AcWmiABgEv)tEE42;4}|ooa&l?5bjK zI531jvvghu^# zJ$w^&+Y0sp=}!TT%aR17!nW{a4j%+3x?+=uDo@oF|G|n=7InS3>?1zdOZ`ZtXBI)m`(8o3ltheV~koV!9@Rh zAM+fetM^%jg*9bmWgT^)$K8abi2!K^NPZssS7E$7IZ^Tx7y(cw*-1&io8-Ri@ zk$2laU$jo8!>HdF)GuVWKE5x^^o4GMiP|-FtJCt4{Y#!*lT7_J5SsY7W?+OLb5v$bypQY_#(G$S!U|83C$DB;+0sA3g^Q z!f`p|KSZen!wumWMeiRRdo6f%dg@hVliOgU!f(wuWFA;{%<4;13Z9_*A^}(tG4eyJJW~R)of!h&`V5mdkij@@s z@1fm+fW2Ijx9%BP6R_axDy91qmOHl<@#f2f|MPp{$!X9|SvGm_}MMK%x*(iFm26;gm!<0Bl%V`ojq$iyPHog9uFM z91oK#IEa{%jgzzZc*N$ZR>3`%c)jufx&6)Xc8&E1m@sB( z!?ggFGlcb0eP(G6!sw4XT~k-sImS93nEt zc=tQNcRV-y(ODAWBWUC4b zt#JPQ3!s5(>FM=GuXMSAqn{Q+sx`(bk_ioVDPRgx?&x*GIqa@YlE<674XIY3tYJ7i zXLWVSsX!7*(njDrYw?5dYs3%|=xHZ*bnplsxdcPTrJMPLJi-#&_9 z50FAZ>E{#yn=qmVG*SW^oIlMi93xF9g(tt>AOSz-hUmxn8gZ(qY1X%4dP)H#jOF-_ z+`yv?j(G6ys#r0IUzUS!Zy-9|qVHqr6L3$^1{mlpw`M>?e!cMW_XkuCmuHR3Yr7EKUUA=hGGn z!8;9Gp8!6rhDsE9tJq4NI2?{MsrOHUo2bsB#>k)~z@aCxc{vi>hm_vspu2fw(e=?E zks0DmRgjVFiTj`(%>o56*zrnr%Q`~M5sHn#ZrrH``iOEY=_L?vZ9;AaO;3l&r1s}e z%F}JF30{L~<5q9#v*L)GX^Nf)2dBLJ^Y&JpVgxc)$c_Pfhe3a{zE}+aBlJN&L*=#U z=**Qu{lRs6s~Z@s`TB)yO&A0ql;9ikHdZ% zJVRLCBjeDhIUBFKDL=zzEN1JrZKV7InX@n)ko9~Q{ZK9Ie#kM*9)@y`Ay$ecJ*Wvd z$2nCPRAB1m5@^6s?>K>P$s;Np1SS&>phn7gat33Qfv`mUak^l(A&1R*{Zs2x1z8HSfM{|H-j5?s}>u*Z} zHIi3+*c}NYL(*P{4{(@<*!=9B_u%%x^J$B!807VQ&+j{RZv@Vb00;5VSv)QWc9QZW zGd+?D;O$!~o((1@1r+NMc!qUC)YD*q7lWW}CC)m5mMXych|N|>@*40P@g+qMY{a<& zl@J9^4}S1VaWy&LI6g!uv*toZ1b96)iEuP#gxUjM1=XMu6@NO58oKZuD5ObMl^Mjd z2?=Q+15u6Db*aDHOhb$z`4Diuo=h@58~k~4ZJ=8eXhD>akVPJd1*=AcT9OimCdPh}kFeA`{3;}x6he{e( zjv!=ay%cFbGi}uwk)T%HQiS@~`8I(|P`+EwsdaS&HV5ofCx)+Z87E@%%FD||C}7ki z8wwfd4IH57jg5^W6z~qIjc`fa2b+lc^yF7PLljdlhS_51>0NfV-DWQWM3NjLtPjN6(plN7Nu%@(de2b& zQGC2;z!v0><&a%4@2*xr308$iIn=& z5DBH=3JEw!S`4tv6%EtFyq>}tzwNlTEjPX-_wcrjS2#5G^r}T}H-spF?g#HsOaUT& zE)P|X4)|0BxF@M)Upk82pH%d`i`J*nM5-kPVB4aPf$l&oXHPoKMumq(@wItv6Y{7p z-1GJIMG6!}{1}*P9H$x2!ikg8?*`SW1n}3k`OZx0`I}*etOO&)pB#@KY;Qa}Px+S0 z)$nAM{Lb+)L;b_9QbARPk@DN4#9CN04z8~c_^h#!@u!G{{81(biZtvnrJ7;p)`yL-ucaf~yj=eXvj`!1}|nO0z2+Y_7YNUmk>imui?mkp-1qx4}GS z-w+TQu7~C`fXdcE?*N##QjAy!;tU;VF{waggtmoh0Q4X^3?#)1ym3pBveH6e!l+k_ zV`P(d)28l+@4-vZjv`6L^+ij;3nq+ITnbftGYJb>8w7z~ycc{x&~!Uq8s~eTacu~` zABJcEsqkXS@TE$W#FjiCA&8BU3J1y!aB- z-AxyL1Ncbi3z}sjAX$nGWqDN5yLZJT51MU#B)0lH4s#u@yh2_$ne)1 zHV~u+1|Yj90S4RX<4gbIBoHih;wg58WXnT`uAzKKyLN35 zuvTe|VBQ%Y>E(d=3;sgufxZ*M6besDDw0xF14!dYefS}}dKhZnVW3x(s-kBJg*l=r z8Xvuw7I)tdNeF3<7)r>Pig+u)=(!IG7?(f_H1iU?_10Q2PKc3GD3WBh0d1s1S>MzbB`cGJug0DL%4h~Qv6vTUT9cPB*edxN1_NIJ;=zsVj zKU$D{l>$PI*LwXMX{zbalkm<%$bB5b*}_; z=+JE9KZ!7d0=xY*88}9QZjsM!1j1Q{U;!da$seB^56bwRfBlv7)drMb+|ZGt-*^20 zMqed~2ZisAg7{a$8O8$^ClF~=g7z9Rr9ug3G$as&!7x7unMAeL zg`=ppXzeM$=GZ?tu$~|_QtM9ybCmyL#>)qZXDAzCmn5+#c!tk`N&2>{tA({4uqSYD zL?T)+Lt{JZh8jBlXCB$R5K$v6$mbEy#>u#{&mz&5BrsgzC?! zqpF5aBlR?Fx!30=__NSbK%NXyMQ*z_DMG-96L(6wdXTjT>$#sH85piYka)Pe;Ug**qt=0iL6@{_$zUoxG(+|u58^{);)XVq=m+wfhfiwt9`&fv=eEs z0hX8?D$*PC=#g@l-~hk|JU-m9GU?NP5pm zN?$f2EhSZnS>tD*)K2U+->XF}=A$$$?hhPfZ@q`{uV(n?uK2@+U zx|jE{vCTq!ApZPR1u&5QK+!!x@oFO?c@6T9ifki3%APE2EAzglBdAk&ahsSt$v|aU zN7PKW*3H*;_%U9sY+LQK;}2+I>WD{s{rH#dB~&uE6_$Q2_FOU68R-=?5||Zh+0y=h zgk1?dmTC9a)Ra(KC?RbKQAn0-Wy_KhlBK9D#VaX0p#?3H$Wqp9A#0>imIxt4(I$BIMo|9=Gep|Q^)PfpOg~~tx9Rc2rhE?=Hb1Rv}+?y1l%BpvsE|{ zq>70jpNzpVx4W2J28vy&N**T6`KC=br+14tM3UhC1GBhs*s61}U%%7HSD z@sA&IzEFm~PszcVa7N-F7d~9dsV=3lKm}L5ZDHPlq%pW;c2>cS_l_un*p(dI6`+9m z)cKvN;TVgsB-a2xbZ2D`czT4_Z{s68+c1-WZLFtO9%#k3 ze98;a7|#IN9E0jh+|~UI4@yt?Bj3gN$~?vpt!#l2UlL<*?ug~$sGD0p@5Z;VOvY}J z_d1IpQBhHGkz#-F;O(U=*%NS)L&tTAn`Z9&d3P<40F{}fE?KmA>HTq+mD&FIbx{6K zJLZFF*lBsK52m{~%&CY1VU&d86pp_!=x*AWQ!1Z{i}N0#?f55bA1fK^dxL+Ui2Wrc z)!31fBPDhFM1j!Otwo*PY}*fo3C=nadASn%=Hpx|rgMVyFc@@24Xk)oJ1sA|B24TS zaR4lR-g+J1{SM{GiJoRcU0}FxZq#ax+gTdXM9UW&8ibJqxPwme37+i}z2%D(1a`!mE!uzeQpa9fQ zUwt#G&~F%*AvQNrbpqpZqX<@cF_^#mMa4n`jAA4bVw|Iv3vvquqvfpnp1}X19w9Zg z%Y7M02Kb4Tc(-jU7H}n}6CUwgu0DP2_#FdA`-Y2bM;X8cCZ#X&LRwYXJ%@&J7DM_U zMHJ7I4dg%7#xU_y_JP9f&qqQK{fM!l(+RMHym?FrPJurzxO)Rx;f0|u>eyrc4!`Gk zL@*ds*s>|$`K%;@MWXp5&SNc`5piRHE<6!&^ghOo^bp*IN3!u~GZ}3|swr@SXZ5cW zUoWr;CUp=xvn7aWN5qiyxnsZ8*e;uBxVhe$jIe&Tq-dW|LI5p2)V|PlPV)^t4pXPnyH;V%=?a8%Me-b@&B_PWEBLSUnEl+tiEl#Lt6c z&5Swj@XXiuqXv+k>bOUI&@jIO758D_%RBg0I>-BwV#x#aXp@WSk4q!(56gbU{H*Sb zIq%aLe2l?U!K$xkmZv_o|E@1RH~Rk3rkR4qr{?zvE(0eTe5Ucceco`zj_K*Xv=b3_ z0_5nxl$?m6Dk1)i*+q~$*h31wKWjSvVTn6P8|ge~Qt-}e01CcMwUH-$ReK{avIU<0 zoxK;p1;A|I{L}6d7WDnixgi3LwrqK2-snyW2}BCk5HzEF#Q2vUe-6z?87kj$;tDvs zlZS`wAi4;YpaZ$TI5Tyx)jR^S)2*8~d2I&}OOBxiBXXGQ-^yWy5-+6%h`rIP`5Sx& zTMb6l@>?7Hy2rh!!UC?5RsnQ`L=KhwV zp<3ffa5);f&)HqC=ZE7RNXt$&pOcV_#Lv*(1ZExSxzv2hx&WoFz=_-s{jAhcoDmD8 ziJgs-7&8ju=RF3rp2%=N0uPmXnD`N}>9*$82snFAzosSbZjpFFSgGnkcBy~=tV*ye zxgq$r`zPWwz^jDI`z};xU{eBN@2qO(svwdE-oHONQ{WB*Eb)Z#p1`{$1vft4^R9e^ zBT|61giM_Qa$uFX*x}x?UW^QrmY!&~!NdNz@#y>kO-+BpOFtAdDb49PlzW^<&*EN% z)NcXli(LS1(3Q+p=b`wBz>WYhpb)-ef;>Sb#lZceHC*{+sta899xQl@AM9g*?>hu5 z6&v>)H^XfU5x)=(r`p@pVIve!vFL^&CaO3I)Sxq)KpSxqd2U5|Necys#vn*K(;(6L=h1ZVi5!&^3^A9M5s;!@MDt+ z{tohz&FuIZ-Y(R!qE-85gudKyO!kaqvS|R;F2@d zVL~Df@72X}vdhPy%C!t7XKoTp#f@~UCE&d0c%+Q$2N+I*Cl;wD)wysP{{n~NrG?`6 z#nEDt(|m^qqN(oPM{T8<8<`0&V|G3-0|6Ot*Xwme%5|C>{#<28gNk zBcL-;2%(~Z$*l(d@mRU+B`1_dWe^%v zu)K*tmX87S=&*#T#p)Qu0!*3zDQJQcJ_3-6-W&E$Z9qIo;zS-Kx=`#LD&b@P(WcOz z1>4ZKh9@z5ZF?TVu9g4*NEBO~(g4)TIDv%I+!x%gyI{>X1>6rw7@d86 zeM8g!Jw^-D!FcWl-?3st4|2gFB4I&u!rIq`nID^=pJQ2Bo9`T<)i81pQ7RH#^H=Kv zczY=D))CZLO!+|t#S-B@?*xhqp8UOkv@Zb>Y<}K_rS}%-tt4W&flZnz;k2;6ugL~M z#|(r3W^gfs`M6#WW04*NV@2TErTSlzuBJ8lpG5t<8<+}>tv z?C>Nzm5EGqof>2U_PP!uyKdR8WWgIpv`T<1L!1~8@XirY8s~_a-ikv7*n_d9 zbr637A_p6?Ldnu@*wf8HP7t?Bw(%+K_lqU2WMzsxtme={T3jbu=$0=(0?nLA|G>c{ zr32d#)fQ}AH0WdI+%{m=TOStln3J}j>vzV84I3VfotE160I`J0VHGyHL8Q9OKD=R= zHNgRyW;W7==RsB?02>aIeR%T%_Tzvhq$a?xVGEf+_QlXQ@9>!e@CJ~`j4mAvk7}~Gx)Wc+POFQq&=>tmP794< zO_7Vz9bB1O3=pJq_00!}M32^EwHGvT;lYi+bP>|fhzG;<6X4>?)Fa^e-KPIEBUj1# z>nRn_RnFUw2F716>Hn`0hlr{V1Exi|YkFPklRih?u~Vz27Jip8co@S=UAR~l?U;S@ z`Rq&8((b*QX`jZvU+8u}^hj~f;jD!A-ceWYZ`beXgxB!R(qG@$-fdo6{5&XZm#4aN zrny1%*o6Bg=NG;=&uP>Vr5QpdzildlCIRw9_5{ig@2fhhd9{C6% z4yn^?23C{2`RP7PPka$JuUL%RRRtahn;r?%|OD`iK_y1^-*Og0|LSaMn;IZEKW$mfdVN8BBhXO=Q5>Znyxw)eKkPM9U+kH+n zl>6ZQhj>zpDZrpnXfa~i`y51&+!@jx78}bc{{T#+;0H!e7u2yEb895+zQux))4@Yp zFwlIy?3lIn&|oGyi9^WI!&*gDQg5xh5I>&621JM=FDjE6hFuf>?W_+tLTF4|NFj>< z%~ndlCW15YGjbp;f>|6xyE36|iOAe8g)l;@Dd_DaK+2z>A34LfUGVK9Qu&dnXobG9RFGkzHHg_50n zN&VkleuI{l5<`W3kJA{Q1f!L6s@V0S42Q*2dK$?Xlv*RCF04(`a z@MQpg$eUVT<492PQqnCG9KN7ea97i?5qxV38S-lfIx7jMEO)9a6v8Ga4gu-O1LKZV zdmp(pZozo4x;UC|X5mV~8LtiYEGmlok8?0FGl?8oeQb%=ej6@ zt7Fh(%7=?oh4SW%vWmE7H4#6CtbK~wCTwPtbG2#bUo@Y&zk7G#jNJ)Cs*Uu^!I=?> z-Nh0U5^KBJW!=~Tf8=by2vd{O7HJ4kLm@~hLQ1s8Ws4-!B;R30)R}x)?3l8~+9-hW z_P#!ExEINa+(|Tir^5&n26;Ez?9{0(aHU^G&r!Qj{Z1~f-^upi3m2ZmUVz1)ESp2w zz3?Ms_0SAu5F0MhtwffXVO32b5E4fuv*sujEn(g{HOWLkr30-epvn$hmb1^)lw{hplhk)jijVjH&i95prFz6L-2B*(*M#kVoHsG9n*y1jH{KIFnz2=cl zQv0`We53Pl&{TY?gtqbn&4)qvL$P)c>uOBDdnP1jMl7tk^Q(0N|9tJBQk2HY@7$ zwy$Bf_rhkg%`!3^sQcG&gocOvKoU8H@}$UE6^QXFnr4()wP4|VT~qUvGmGlwI;CBO?rz4eM|DWdM{$WY`{z9bQ z&mA2mwsURp7BJ&h-4o#!7biaI1g-cx9nN%~zrjl-homut%7ly<&32hceb+!%xKxR? zy~gu^l{Z|}`R%Zp^TH9g+58|-ypY4d@$45b)L4%pfS?;V8ukFpt5^dA_Ga@-dgItc zfeF?J#nUts$=+Ip0|$bq^YB?j0h%!4S@|uv-{EdokQ`WScsS8{RbY7dT$(IC23Z}% zcK+oetbK`^f+6n)K=kNf+a@&ek_AQ{rMK?dnFVaLG&4h?X{x=gS|?>*gH%Bwu z`B$=_6=em{V;3fE$d04z?1!kr$YyXnUUxRh{#>=q#TEB@s%nZ=`}G(=>rR>3A^E6e zAD z_b4h=pKMN=jUl;HTC;MaRCgi)@``Ixtf4@BVU&!$Sgw^=-K1dJ;RU zqCQ&qVQ&*=tM8A}p!Jn>n-w1)Y&)-ARe(~GEKz;s;FK~|!x3wY$6M)R!NpEGIBvMZ&cA!O2t_Mr&Pu&t zfXb+ZgoM{&*i#3J!q*LaK}xR@m)AiuaKZUN3drSY3(&G z`VmAcO)A=P_Q*_iti^U%THhROf&$TZ;@c;hyZ7#0L*<~#x*N7Ogu;~=Jfj3jI%yq{ z;#E*FY|;nxGSTvhRb$#bNwsH5o+J6tBI}chv^W;PqX`4E|CeLcsYe-1Ju4>$+r;!X zl_nVG?Ip*bY$(J49RD&b5MxKrjtu+8jhT6l4o3 z*3NM!Ze(Kez{xLs|6a@SEuJ%V%@@%wL!TGnappEZh?*?@($qz=DjwG2qOaKUg^=P2 zuxyNq?wv)JW>cV|g#{nF-1yCJ$H#An3UwPTlTAVm&kgDqGF$8y}`xllmy&p?7jPVhpIM--Z+ zXzn8{^MDlc1CFy8c)OKq`uM0ab-?V;Zh?LW?LeC2Z8l4Mn{L^#&;808xBay)oYS-An!hOTX zCopbV^sv(Z`1qpFD}%;$KQuc`m&c^&yLzX5@?GYbd?|plyuO$J*AG-|l-G*3!Lx{rWQ)t1EI_ z#;@Oifw~kHz=QS0xSrUvxh+iAri(!aeY7fjV0RTfRkWnXc5f7l25Fox1|Owui7WLY zA>l3WME1q5)04x~yYfcYAUyJh$zqmBBBivfY#Ut7^tD5fDiEkOi~oEszcEvw*~c@TA`&~lW))4L zme~g&_SnpvfkTP~vy_C&f3#LrtpLR$2g1*@?0A72@=HID-8r?@(q>>1nV`Z6I=KS3 zLU>l1v~fX)Dx}K$FMV&BkxPCj6YL8%-5VwPE`oM6iB%(5=~}E=WlRSG1kvJ(bv_sm zd`D#LT(J=nDg%ok4JbSh`2K1~?yD-EIDad`fW#=MI7-whF_;AkF(%4mv3Zkr$+BfJKs{$0Gq?!8AP5vlH5oA1tyu@?(dG zh6+*E)q*ut`O2_#F~Hnq(Ef}_9|}k?s}b_R!A9XEkiE0GV)%P0uKtO-X1+i=$y=4ic=~Em-}>L7khvUvIqYdj_#eeZ{MCV3sB{P-Qx^%+WI4U51-mD z{?%Hxf4;_9U^wOd=(F8R))#5-89^^0u50@x zO!~o2^#sk2v)c&}gGZ*kJBu4thaAXC4N*?cC3uSc-)>xltzvW`qUZR+3Sbqz-7+hQ zMUo`-;708!zl7W+jJ$6=Iti(d87D|oWfu*>R_$)E_YA>h5Ssjq2_Ou?Z%;3Q(T*6a zLgI>)8DP)5zA^6a#m9f5vSye_00vEZNk1Eq!EKFa+1V^OlYw!*x4s~AOhT0P96`}j zg1d{Is8;^dIKkL}Ll3*7mH*UjvV0m#C69==fx$#@X9eKlPQ(^SY0I|!mgsIiX1lyZ zYTpGsrfY1lahpZvQudKO3TW+FhG%S5#XMlAvbuU;%6c1RW#xSro;1vP;iKQeM}1XZ zt`cNO9vp}d#jUM|JO#+_ng#RZv4Ta^GXuX%_l3Ox)JUDjA?~CD<0V7_>kAizdS5DZ zS${2ELH^wHDQgFDcOyN$PjG6*8g?m=mfQm?$wPz^hI=**ektz3D?PqgWlF*&YV0C( zP(o4+ zq=&_18zhC!k-f~^X?bwIMfI&Ej~yDd%KO#^9y1-aFZt*5@T)I4j|ML^oIH0-vY}Bo zckNg97jb3jhe!O3IPJeO{9@4`UgB6%!}+HdfP4p^5fLRpg~GR#rxpQYqJy}^^iO(*NC`y648M>sRgap7(IAMJQ4{2agE;8ccY{i zfaUU<77S&>ig~*3a%ap?q3twuZ2z4na9FzQSc(qW9VDj*NC5eo?Zgi3@fK&pqG@bM zdDWx86@g+MzwTvl+Vq_uNOT`BYdnvl1As-Y5zzA|qNVbwpg=++uW4vo9gHJ80y3XG z@f)+7Pa>nV#9d~_&ux&LazLSZ05Z@)(MaSgL+4(um{PLP2jaY_-Yy#K99^|>qd-`J z{r3=+K5trD+Es=Jlso(TZ-a^}tgb%9(?8jiuofuppOHaak;+} zWQAC~7{++@y=e)za$66osmUW1-OPbmzYs@z7$f^Rr1iyJP{eGI^rJ#Xym>xZ#oZC*f8QRR~hAp}%6!%4sY`WXJO_ub{95{>2;6`C}-a9v&wl8F2SUyvgNdqGN!K z75(=LPypW=w!M0_=={qcU`53*qk71sE~7#8L=kferscB`4jcQ9J=R{1Dz72)3?uA+ z+y>#d^6N8{pf45a<z=p$LN&?5LSP+@C8OZus6DilUm5n7C_SXJ!~lCS|r z=H0tdoUC`dEJ}k|KqsyIpu!8!2W=l=z{;4AsK!w^fK~O8{J5fNu!nD{&$=Ei<|EvfrA%3v@H{)Y^0NKdC zX)*f3ZXx!)X|5dviO1;T;u6&l!=}kV*C0qCc#_uNt7$^ajVUcDE0a$VovT^~j+G}Y z@4{EF@;+Jb@bHZ&7)%-m+Q_DKtTbO;?VU=Wq5?zRyFoIW$Hb@%8*>OzD5q!a`(()_ zIE=)hu4&YRC;;n-MY}<9>jBwGDS|l`K~oD=(4MT(xfzxG;)NtBb(-{xTUs^(8ePLc zQT~yI&|5ECnWUc!mgU9+01iz|tgi^_oNL9i;*p8eeI+zB6zOcIu<#nREUd2J-oE`S z=9q~-u7>D-VPW1i%2HXG7FB_n_QoS8a&7m+t74k6zw^M^vpi4@cQwhvgN*LrCAdZV zVAsQc#YjLKj-YOy^B%&epbJLx{B>VnM#yU>SC4rl^h5aG3p3&^XdWoB4o~WPKY<=C z2=^6X1RT)UCOt54FILFVrD_}tTK&*lEe3!U$FNi1!4v;5><(80o=r}vL+(HX+tJrcaz z+tKIWsE9YWP+YvdFC;&{Y#(_zfezMH!_TsJ-#FKR9^rh z-;RJel!8A%4&G0GCKpu*Cbzx6pI^!hR1yl*d1_=?Ny!#8be!GE#>NJ;qx|@Ay#U&? zc;cyjrUM<+8ln{y%pXtGj)uBNw%#Lt9@+EkeA~_dEgq&DKvkON`3KmREnSLhA;*KV zdcO*z`+~ZvDji(+o*BUGXE7kIgH;og8DfkI+%l*_&jG2TvUU5<=b`g-tBshunX&OT zpw;E3o1~pN@#hJ!if@F|P;28xarv0z*Mdqn?Ttraj-B9u7NC$aMu*OmEa-^-Ys2l( z?X#<4@g!xm&4EW_9p!7I!d{(>STKBqtX4Cd&6D$l4duc}fsGvfxmyPdMp z7(aCxXyB`u%b9EkR0uNk$J;^4gPs(DK5Y$+&0$BAJ__4!#p2gC%V77MKn~IrIbI2T)}c_Gh^Sj z{`Q2OQ(nyfIIqz%Gf!ysxMhynXCB*}>sK0Uv*r(cUpv?#9$E0?-R0cCgENAnhTAs( z>xmIIM0@tCZsFqSNcerQTTB(Qjzcu`>-YOe zLgoi@pwOmw1BopfgxS#WupF-F=g!XGQHP!tDRqyW#Q9;Dxuna3P-Gxt)!fjdUkWv} z6cm`+TWQy6CIao@h~un^3<%&1+X!7gHb>1Ej)jJNS4rEw;f)p*>K#!dG#c;_z(A$Z zWi?+-rLG76EFd&7+N-tOc0>|A?|QzwAx_C#SUk3##~e@pl=U=(ejY&2&uRRhOUC$* z1NtfTmIVtSUm@QW;Z%AIqkFIn9I@#@#jI6x_7zxW8k(9a^~>7YHlddcPsn6}asTmN)1uu8-@Dze6@W_AxpajWjs-XJxlC8K>HKqf=Bi{MhW6du3i*U=tv=M`nPT265vRvVsPbTtr0^ zArMO?c-z(x$kCrxZn#TQk_`idU&9Ll{6&w#eJS{g;JkrMyj`-xwiuO?X;7at6hYl5 zr+-Ho?+LcT7+q;+p+56x5iUtzWhyLDn{R{UYGY&LZ2`MK4%VQ9D;*2!*4Q9j7Qv#T zu6ZvZTIjox=08sEf7GY{{KJt4wVx%JpN%~|U>Bd^wD@|aESn$?FQbI#Bi96he}30r zzR(K>gLFS7NGWDND0o(YW5B^%*| z9Tt;Hj3_~nC9grk`T~;~n5B(q*jVACr1dz*mtEp2oP1IJ*mbS>Siclw5Dq|e)PN#X z?B9(xk#z$I7l(&uOCi_OK@xeZQ^QWmt>hCg*5Ssnh+wt1ynOrDuZvw=W@J10vswr) z%Yh^Mf#g&sp_BHyV;QkWCJry8FJb$|=%p-$qf6PJ)!Zi@GV zZ1AY_%a^SIJNWsDo@F|))Y{s*5OvoR1ZBdy!+xUxTI=)RAS2irm$~5qs36GBH#O{0 zoMR&xUDQY#%}{}wXT&o&hvu~`U)7s0GFOXK-TO0EI=T$G~VsV z<4-Bnm9W4R!r=4-4J7?-1)DIlrEp<>LY9zkp!wuM*zngydV70oq^759bRC*c##r8r z>Rq`ti!Y-xD}428KQIB|{}4LwX8FqAx_w*3K!67Ip#qF<8slh#I{+Kv+l>$I03Z*! z%wmk@7l;eO4hk31CRw0S-O|$1U)G2*FBlSdaC|Q`0l6IVCyc#DHb_SJ9U2`C$UCHH zbp95{8N8JrIF^_)X5?r!or&58)YSnVN_ihQ_==)FUS8_e%SN)1TaO<<{&s$*>-+a} z<8$nO5>R%H(fW_FC%?a2RiXYHNTLIHK<`dmAwi;@R9i9M7-l=GY-L z0mtF1rluF#SvUBE-LwwS`fP0T{x?fC?rmnVm7z-p;NB zs{BI@9*hvtjOjciNt4j**m}NTPDi}yiSdE<hjoZFR;J{0y3Q=NZ%c2 ze0$V!uv&z(d+APiAU?H1xz-~ASJu*%%>tvW>iDR zcRl{XUY&Nu$mx~c#!gyX!KV|=~;8QU~Qh&Z#2-GgPzyZy7kGXcKx1HB3$nWYW^06$(^Zhv zr{JOF;p6j{GK|t7!go(y96bzDR}pwmqFache?zw^QzCA|;xGL8IWWq>@V(49^nqi0 z4=E`vMJP?b@Kq_M6e}+#r>CQ_?ZOcg_{j{&)8r=zon8V5YAt51v=180!@JCefe(eC zk$hMejvc22GOEUP4zlre&K6`hOwM>j$hvcY#1Rmp0wufzO{_%SvbAvbOz@v)@t6NJ zv3tCi@5d2F7lwls#horo)t+x_d|tpJ*t3tL{y;;q-Zpz1oBuwY|N3)=6?9c(i*<>p z8E@ma2i`|p_kIkK8Y)h%36t0sfIbZ&Cc4VGyfD=`B7j`LCtE|Nh24$24kr+s4vi<(d-?FfP{}0Q5Tl{#HVK{6mJ}Y%I2y} zn5OBVTk;4wAjJ?jjt%6BgY7oDwKZvp@L;Wk^bajb1)Q+5np!a$+@osGa!yQ> zS=?lu$=wGJc+JB!6Oj#QNgKse;NRD-H(#ABgrnu>kARvD7IamrsUV9W#&ZG(oFamDf;v`z)a7XO@I%W%#lI)TkdEJ&@HO(L zPbs>Hq#24XFibTWF|Y&#T;J}=)kYieXS)9DKa-6B%TV2#`F#R^_fq3f8ucrDt8$HS zl}fqI>hc(%Y>5gB15fLD5h_*27*Q(0HXY*u&*P0s0=7{No;I+^1a^pgKe7QhV+-%X z`@^zZxVe`C1H8h|Bh?nudV`6!RJVn>i6T9!df8nQxxNB+MEV+Y z3moEgNhfN_jSQ1;f9`tel1bXBGiSDkjSUa`Ke=br*wFCCQ2rLg3WL~Goh0Vg37poW zt!4)geiM3@p1umUC^_5bBh3HpG_$mT|;gG zy>;T5Oav&lWw>KO=iE}o+_~%u(A4Pqx}=0yxCoU98D2jKA3mZGrk zUSy;%>hQ0S7w?LsLIH2Un&wEZ3Rn$YbAMaNv*b(ZmmHS_@?iXL_iBp&w zR-N+Ii70ImDXgnDWf2tQW%Q*_o++*SWwibIqIL$kgUR-It~@iL@X6(%Eva3Cw625& zs=+bqc@bq!G2ugdU)9&ILvt~c+p4EM@(;g%|4z8>vmzmNPY;;x15@qe+%S%NkPeJT zY8_I6qIJmUu&HgT|qiXi|To z`~B8u;5H(B5jqJBTXEg@Ss%#9aR!t3&=ZMMH`;J+C1O3QfOScOm{u1+;{1gRjXB5U zK)aK|wEQc2sL1WcIv+YxlVQ+MI@;f$Olf)TOneHhc6$IlE$w9=A0L8YC2BU}7@{$! z(RtODZi+@N0$?vtT3C2E*Tg-z)ga;+gTAL-xv~^uDEZ0F1@1<}$2(g<{1Ibel%-y- z)o(Z4TM5s05d<><sv7j#7*M!p5ZbkieyZuHdfJDB?ZHER9aE6M)X^sm^p+q2A@%j7Gs zfAQn_9*M|Pg@v(Fmi=ao>kT~Pn*ImC+@?3&K8pJ68f+KPR8-&2LkpF&F$9*%&L@db zXQw8IsOw0VI*}7IMu~YC$YwLT8cl3UOGeB{x>{~3FvkjPwpBtxFDPfnwDX7sF$ITYYfN%jC8XRj1OJa1JJVpBcr1pcnmaj zR>Bi1?mGRng_n2lIQrNFE|kKywl z)Ei*9_ixAVuU{qyh5P)L^rGR^hci>+_Lak8Qg^?}JzA8;)p<%w%q6=4<@L(H@IwE5 zVMDv@!jx}`#SSLA508)1Bja+Jp|^8;lmx`8GwsvBCyG!btPNKpuC1#!GckG7uv`!x zJ)&R`HbYJN7;$}I=gQU8gk-o)`0sE?-$DVPSrZ2q0grbW05VZ?z=Oj)hcQ2FKrHJ( z7jpOLvhvOZ9hR9|0`^AOUMmAvjlL73)vb{VF>QbBfkdsVAPwG_nmEP6%E}7zQ7M#I z{(5=65txT(@$e9c1YXWBB;*!wa61?{jGTXF>*BAC!OwL9vp1Dup4ifxku471^<)yf zwdl5OF|X$$_AF;$P=Mu#^wRrJL6*3NU~mfj378xR2SRAx@PTqu1ZtBnu!7T0`Szjs zVwf;IGH`Z;9gH7LL*Vy36rXvPF`A4pZE9wA&6_#^J)UTzTq*k()Jb)9r*2K9zRZp1a`xf~M@jfl7dxxBo=hXpkoMl1x~M@KeO$S#7skHLHp6?UB8 z$x5+)5J??2ZLe4%Ji5+#92jnfvBB~>I_tpJdP5u9hG(VDZ35vQ7L5|XjMeMc6R4jU zx=%Y54QVIgDJ*hAU4^Sv`K%S-VAmf~jW-ISH6w%`*+W?eoM=2M@dbGu9lx7syD{$h z3KCs~eZGuEgKK*5pI(3uJZH1?iE6=Pq^o-aV&q9AQ2%e%DW)wjF<9%2N*z74b&#G0 zVL~m!1tT^PAtE$V`A;0!mSHa%Wc6?I(qE7CjpTbNKELgdY&zB7eu-W6y+Q8A@||X^ ztHL2;<~A0W=$z`c|HXgxkC&9A@!<`gy1`@NNhq^*75MbDRCuA3Fwe%iwXVah;VJBj zuvQI$;Q0cerd0D4t{jy6q3OH%=s_0=!D$OS-Tg}gKr>&_#8`|rrlmAP365q>br6+G^VxaC7nA`pX~HP50icZgJC~T)dSE`19cgBr=ck);6g4zf zAwH?=*RZNgC8M9vVG@j!Nq^~6g|yHc5p8Jfi4*xA5Q1JI#XhYmEQgH3-)Q9M=G%~U^?0J${4_R{O+rZDF4c{XC5WwIwjdXX>jHLIljh%|Grq#Psfei)8agRD zUoqf99{I}jJ1-34n0QtMm?aPZ`voiU1tZ(Xe8XHD8k}A^J>UVdjH@oBmW6O|0CL3M z5H*XBLT2W8czMxLKkHZKO6=WBcSdbQ`a$B2@sq&$=B+T%&bCbm)U4s)xHB|ltzxsb zHnV)ehx?yl@>gF3>ahm?b~C(Ra9($jzt#P5Zp$uf-uIV#Vx=ySV=8W;)skgTYia3Hq!z?J(F&6aj^C0oY-RM)an$ zPxnp-jS`aSrQ>Qs(x46}oHU81V1I+Gs`y1%ec}uf-eDZPW$iTLZjm1Ej0mOu%a^OD zZny5;%Zl>yZb~)5gBV=IE^bW&+yO`n8yyP~byqau2 zsIBL`qT>_wLf(`Ea0>Y1McjwNX!8tr$sa4owdr1oZ4SZh(+Jp+e2%Y@bHn~miceK2CQJBDx%#H}#kG=pYOr9dhXD>A7ao&nSsxu%tCst!b zhV}HQdHXC}J4aBLs_ePV31iG>6|%ngux{<14{X4K?4JcvK;5!Ff#!iLu-wpnHyP^E3CV znws??YAHI!uUa%PGaoE6%VL zD^@7N1jR`e41yyfOv1(^?bM^#*w0{iibjqRdF9^4xi`n#${$28tNkv_!hl4sh_rRH z5&2cwIq!z@?f+L+8OuDvm%jP+I=>Z`L{{0^wtKUMI|m0dlR@tP@Qkb>-Cw=y9@d^E z2Spz=G>^KZPD*%B*N7y$&7_tJ(=}51!eJ(ZnwbB zMxh`f%~P01q%M5ki93VLXk1H%=0^Y5{0q=nE#&d${-o z7!Tv%JqnMB*@jYUdsrLB4aZK=)6@Te7P{ONG6td>kGY!`xD-~f)LX-Xhgp9LX}A<# zsOu}bIy&AotCvT}xbg+q*x4~xV0`mxbRzKzt>3WWCIA9`m5M^${fUIbzk0_1`I>2v zYGPsY%4aN}&lvaW+`iZuvFJ!-#}>(ghPsr$(c%0aWH;_OIA5a@lRj!Tck|w-(WV}= zw99q3MBgj%)`)&;(ydkc&HBlnIZ>UA%+zxq66TvX)n+z)`lO%xJG^C4$I-t!>g;+m z2j-T#XC_&1bIyTqmqU@q>63#N4v+(;-)>ZdfOG*g-GE%-g-(*jzB8t#?L;X;u*UVB z)szN7FZ-HfRnv1YXLMCuxrc-A?HS!157~e;8UoZ6f;tcwpVH<@J|CW_naJNF=ZQYi z%7d=qhqfDUAI4=X!LWrV*pLODklH^s<}3|csu#ck(KN{|A7((f3Dv3KA7 z=~&V$9$!5Ca_RB&GNCYb!4K+@b$$S6w~Cswhu^F!K623`Hr zLF5KkuOphrhMLK0x}mIwyLVSXj5*+FrR9u&<(XyxaVZ8SrPf=dIrcj!a1y{XkiCT+ zw1jKh@{l|N?~~ps^u%bMu0~G)YNBZb$&VGm#Lr3jrp4abG}MiGS$JD!g3jc zRyztRLcPH(!1eN(XdoacafuQ62Lmm(JT;7$Mfzs0=3@#fxPINor#CH$-2KkvQ!|U4 z3>T-0so@Z=-nh|+G;#fo4aAccg7r{{yL{KSCnno!=F zb2PfM**Xz5=MpS$qD~_q6-MrMA(V8`>#l}|vXQkX{deBv64gPx2_nP>NheWg?k^#D zV>R#5n|crqK|Dl3H%VVtYRdzstun^DjYLBL?#S7)eBH7@Ady4PruBp>F6K@1JMom;Y3W#`mXWEiDcS zs0!6|b!fYsFz589T5PaN`0wQcW%%LJZRv{`uRqJoye1dLBZTHP)h7Rsd^<&J+WIl? znHB`F(BoDhQS>zO$RE=V!UVfhFf|c4ML24VCBws+Ic^Q#3Tf;mkyuJ6k-yD%`?{$U z=&t|{iGLr6B(uY}?8=ufr}j(g%aeT8X=Y5t%_ z38wg%)J{iJRB4cV0-vlM0g*|NC+Tp6>Ek3uo(=u@Aw*%^8@>`AR!7XA=gMOo-ND<( zKoPI5sJ#aCv7(7lcYF^|(m|)4eb7;$q;Ya_kzrZSLzjYD2LmQt7Y7P}_gP80G|UOxfD=l9`Gef888XmanXA3Lyeb-0 z%F)TsAC`wU4%FFu;0!-YO${Ggn#j9^%bjidW$IJ0lLPsS z(gtdZueW?ZQl8oo`u6VwkDn9DiiM^>x+o2+ zrFsK%;4<{&0C3UL(G|gUWj)@nhpGHbxR!)|F-c2%-%Gm1;gH`3rSOqvXh(*Kf zXB3SVQySa~Gtw5#J;0C0ybQv;klGwhIdQb*l1!L(;`s5~0K@y?dBEt%4C-z!*tC1s z@vDwbuQ@qoL$5~NKD`ZAcnrFT!o4PR$xt?o_9tQZZ^&~IYbJ2_?wh?hjPg8K9bigW zu`a8K6nro;Y2qhNL{wN_vtybSzQ0vy(J30Kh|uYc7d&Z*2%wQP24E-@ zb~v3&T5oD*r!e8tqYKme*5YB~JbugysOuHl2R$+X$H`ufw;NXnM)<_B{hN*QUx(w* zuc9+lqBK6Itb3-hgIRZYq2a|~_M_@&jGqhkH7Fm|PQJDFz{f8MnZ-w*=xb11pH}Cd z$x3nYeP*#S<(kFO<0-zGD?eDYuqk@2T_641t%nSL(rf%*x=((Jm+zD*mstL0aAs<< zY-ZfSh+48G%qFr!WsUx%%)@VH{v45mm8FmFtmLsu%NP~^CnoyK7o+Q0eLCBo2N5QC z?c8Zg*x%96&=qQ9V^d5Vv=-~QfGyX~pY8~*?HeOCHedlgj2{oUw;7|8L(O+c4nAz~ zX9vLf{=|-XF6#l-_V3$ALtLF0y01752AlW|;D+8$*ztNWVDwbdMDI5H&+X-rfq~56sL)Ke zvm?}XcJhh!iCW{&?d@BRYw@DN<~0V5+(Zr68OkiZRM9oH)xBmvm{PD6Oy^&~!r_Y{r!Gs}K<&RaI($fPq2Q5+(4&tQ%!vm0PudBof?_cC4ZEh<-OvEzA<#(P^D7`(?ikkn?51UiXhNpGV06R^ zP|eU*$&W=a&&Euoc zQ!r`oFmHh=VF>d6DRiCDsViv<-Y}G`2;P-8#CaMdDVeEu$tJqyxk}@%^|8k2Qi{PJ z%Pza5T&;o5&Jzq_0PW&?^!9z+%@vC*ESvEW&=aC(JDg>gpN86$DC(ACz_@W z5)6-=f%521fCiQ;piwRN7%f|R?N#!eNJ!B{zBb|%(LYJUB@GQYH9U;_ z;s2_RkS22HJr@MJlgN(vSC0Zu=7+|OPO|zYJGI)<4R#jb7tv}3ubQM-Y zZUk&P00ZMgC_4ZqXl>iFc{2?-L@%ABJ&F0Spn!(+MO~x*=%!I~adW^Ytv~?-9dSx; z!EQJj`{s&lG<{$7W!K=SNI*C~`ypX9>BYpa-U3YzWAA52TEC#`<(?qo9@-yBSzY|P z0&(Q2^|+nuz-#n$WB%Dx7r$$3maThq@CKPn;?O)uzE#$LjtNzU@|@xI zT}OB6B(o5i7de~8(Y}?4t)%b;KgsHP(0y*~7@nAIZDD9+gggLXegC`;rxpttTqr*C z`PE{AAaYI(`vnA;JZZhC4;2B)!(#+gE*5vcLSX-SWwyI0#nke&UZY#j>wApr#sKf4 zhX)@B(;d>x#yNY1;u8pgOsm4QW74O_E@{sI*aClS?h(7R<2hH+B~7N)2p^sWq}7TY zsD#g^ql<4uVpH_QE}B(X1oVa>nnUJNg7S&Pvd`G#u4%Hec3$y_5_9%ZbjXVqq11u5It>ULnIotm*{)zB<tUgErlsmONoC-oA1ZO z!LH8o*&+iOOs{(TsfHa{zylC6#(iO)6yCs!FjI)azohbbEi)-c{WxFrrmrPf32?(r!8I!phCI{jyCa&wN zT5~O6H#{HE?yqGZ{}I}?Q<)AQ$jlarG#*^osG_qpAP zSvOuMDN3h%!EA-3q-1+*blM7D^-&h2a=OlJD&Hi0Y!=j1pN9`0UdqHUkOQQxR`8>@ zi#O>yWQy>((ee7Z%>V2q@k`F7KD+XX?!GPG`srC&vFv?0L%P2E-A#=1d>6Byn|zh^ z$Kg^uB5;hN=DY8YKPjx?y&!8;<5RQx-v~wKB^U{C7OM;toewA|(4I*mU9Tf(luLC2 zsvtcV;tPB0&WaEu2~kM{>y#dm)~I&KtvE7rNK-Gmzf&Pg9{M+0nwq4^9@anl8Vwda zy)p3HJ3{ctZv?6i#W{n`T^=J5aL!7QX~Cw+!GPw8SyNE}Omv6G93XdyQe^oi5sP5Q zUJNU$_b7}m>noVY@*M^ijSiH4h!b$FT=dhiG@H zhIt|9bt>lWt%0u~M!5xVD+V$ta$1GhOfJXb%QID19|DnJXbK~sk`Lf>xeg zGSfZVVBF;u+2GeH4xLjv9_w39dzlYI`q~lzmTgr9>^jRQc3PGd*47^3Da+Go*MRbZ zRNfg1;(B+n$ig}h1JD(t`k}R^42dXkZvSyC;J?_N@#EPge|iD_3u-T!iw|1iHv4X+ z&gsZ*-G?4hb;fUAv71m<1aWx&*SG%bj|s||AA+ZTD}?0cL0psMFvH;}L4#oeLCE}= z|GNYEVB(%#t__gQ2Y z8gRFV0m0E7lKUM-k75J_#8PSLjR>-GWHk++R8?OhBoQJFLG#=4vJ3Zt44(LKV3aii z^U)Gh69ZZg!;7TG;{dz_+FbFB6Lam1nDGsc#aw67ZQTdN=vU$xqCNN&0x+zaOorQ? z&z~{}6d^^QhA1|_w43wCfU_#nYi(PV_X&Ha76as!_vO!x8(?4!kKk%Q(?Br6JZBPT zy4}~fy2ayU$w?r1{OoNq`U!X1u^yB+fS!?2Y{q64&Ch;of(zh?P5YZW{I7fS^DEAt z`ZIx zL0Lu@7;kqm9C({a{Nu+Y9@JtfSzxhf8j(l+uV26aX@xWDgXUc|(((QFtbkF}*RmFj`)fw# zfCK2#YnQJ7{P@FVn9Jz4V+2^+4B4VTA5fFwy2W?jpC~LVe*c~g$KQoOJI?UYc_t<% z_jHdQeGr=Rz~@ZS26^I^oVP|82g$Rr_qcU1VU@5lKELn@Rmk8y|L7~iPv&g(9 zYf-C1Vjz&z&azdkwO9CU#0LA5ACc6;JGo<&2vgWBImqzRVAnlb+>nePy$Ue)|b0C{h2dU zQ&7t$C=9uS9kgLjpEYdGwx_6%q3Et(cl^>W<@#8GAA@-mVJP$|%;tKi{;Z$@TkrKc+wmo`JU4IkIHnXGMWDs!EA4Q{BL&owNfGH6%j} zIHMADE|gd~koPZ)5q<)S#wQeGwnG$yF@U)8jDSIg6lf`It^O$Lqm%$&GV?PRk_q7P zV_z7}V4olgScDL8QG_}41)_AH&WGTlkCfTgm7iyZy)7+W7|_@S?$ed)*NtxIBctDX z13umrXfs;Umq9#-T5fJlVWGMNZPM`$Bp=(vq@|?~nV8J`T+;UG z-6vHlplstGh8u@v!Th@?Y|jWo`D>;zNt^&-+ueDHdS7p)hgBmyy;xib28=x72evyc zwXgs#z>n(ET8g*37@$=Gu#eF|2E>6`nfWRtGG^@M%|vbnYsYBkUrOTY`JOtqtBTO*Z@zkmV4DcE*?7n~g`4xuiA5eg;Uhdt1x>KE_uL!2} zk;0{%06=2%K8#L{)7Ec--o>^|UPJTDVOiWTBjpKT>}{yeAT@KvzAPB>O*?=Nk2uKJ zLD#d%|NV(XsH?4SPxp?aeD2DXD-IT5F@j~k6O6h-#3wMGbD_Ak1<=;SfDP!SOE6oU z1>jnNn1&!5sq+9bY4%5lvnXh3h_*dAYL7t{&+U%IrAEwFP%Aiesvf?9 zXtfJJ@A^z;{$we+6!393x+nytN?w6mKl6(@OLtJat!5*)Akr=Bog|L8$*V7@Cv6)w z+fpSW&}JW<^~qDb<{TRSF(DwpYyhzcv<6%_9@xo4rxM!w0z#7I*j;-#6wV?JB^GmU zp(7Z^)b_@~zkEGU;{ECWHz-=J6V|qI-%pDzq}WHc4Q$d3GqfD4KRrh?Dc&_Y@nvbl zKfSk;T92Y}&%m&B1){Q^vlI4i?V&}Jbi73}$45yD}JWN`9+ zMo2hP?yMHBE}#+dF33Pe8ahst?A^$KUIy#LV++L1)X^&VKC^bp5O{u{0I#r()E{Oc zfE!~ue}V{rLhm9=IAX9N*DdaG_B94~j)UKhfA#7;aaE{-kk%sFq*xK_U%PhA@2DBt z!I7*BZG0sQ*As#`59#QUj7qe3-@A?XIaV6_D-a8RvjOZOc$6(vpE&UfV~Bpj4!NKi z=~(Bi75VAy206YYBt<$4B7e^S~4h#NNey_3{`(%2aKRj!9PnINg7 z=lL>fzb*YE>`f?EzClxPd{G(5bR8d4(RBi}I6Nu8{}&cCpe-kdUaj`i7H}z{XSo!z zX417C5HKOs4(aeJ$+xmJy}V=uT~EJ9Mn=Br8o=_kU}jQqZ(2T@quHVjt_}JlOCixQ zhyvje8a%B{1D_3)?xQ`$%P_BE1_)5O$i6Q20PI0Xpl&1h7Spl$n*5DFPX%t#T9)1!X9%R=5wpe$P{yo+Z z|FPxIlcW?*NJ`xspuxHs4x4JLaZuKQQW5{r6?2NV4%wY3u=s_QBQLGE<(r-T?;lmJ zzKctleZ#Ku`ABWhO8rlPX|XT-cAmNW{xf51?RUb8zy4`OER4#FgX_9RQ$ZU8f5O^2 zkLZEnB+%;;PiVTjjxxYUY&V-$XK?_id-i1i)g*Oot=@qfWL}@+NW1ico*)k@k-qf+obyZBs8dR2`f4e6+ zW=Uv3`HaV44%9_~!Fk>{?Ix5u!699HTnU88ZEyE&^~J!Dmc-^AWa-DOF<8~aCG2W3 zHVnzuN;wS8vrC`2{}kExZSx3Yx!5?CgJx#J;YrkPOcLV=j(tvWVVvNSxjZuJX;AX; z-UlxJ9IWX+o7k8P{Xa9A$h zh>DJOc?^2~q8K4pFpwa6{4z@V@}AFaZEhrxtxp^n=;bp*Tnv@iC1DS`A;r8Bg^>!C z4d}StFCp}KF&ZD*uMei}E^W1hY~Q(kWkp4EfJ$X0J}nnq0@uiQjp`0Gq#cx=IKUM5 zbWvR)>1>y<-2;Yf0tsB}yq6d_%^`R_LA($t!A!Itvc0SULl=Rsue!lnG*M=eTG@;D z@837xfr&*+%pa#SfSqCVVXe>xlFb5{flqgXpx_L^oePWD7X@bqfmgChs$c_PDUu97 z+>egXeURI}*Fo0Czrl!P$2#HT=EYQIINDPZWP~Atze6y)uIh{e76mH57kjY}uc~Jv zR>-P^6H$mTbwfuZ7=JHLj$gY5@ndu<^dqf0O!6lPWkbcVmuDz#pIYL~7&VCR$Ed#- z=;~g6Eioxc9=>_uon10mfo)H5b?`{r>W)t+_i@hLm0I*O*-n=;VA%mVz+QZ%J9-o7 z^DkW+%3|~)zlu6PLGRKLV`HJ3X3U~uEsB9S=W~S#Q`FMdExyK(x)z*PyHL6doV(W@ z_A^U*@8E0#FZF2qZ6wWAcT`qLNC+foeSygLCx{WGJ+eCy&AxTxae;wLWrjXYsjDJB z>oCmU&3?O7jEx8D=yo*ERb8Tmfaa*6{{+hT@mi}9%FotF<0 ziSUjir7K`&bq)yZj9M|eLmdwH7b8@bNdED~OF>5LQFxv;q#;|kiRjmOh`%(?ynwpG zFQ4H^U-#I2fZP>|SPQ?Dq}f{Bt44f#8of+AjEBSEVejb@R%Qad!+U^TwpAjQrNA>e zNte=d5_Xml?yv4aQ%MLGVp}C7L+bEkecHHass?n)#TA9`S0aTpUVtKJKw~Ya^qzLQ zP=ga(H=9kJsLXhxg{3B7_90%K|B}X$WtF#`--4~j0(YQw+Q)f8dhXAqG4bbyoMx~f z9uW#p5<$L2*k#TrOR>`cKKWN8VS@}0P#gs|zhMw{0z1>>hCW%8Q4h49O+kGg{?`~r zoaC2p-q%v8fjHA3dXvm)15BKmC*5^@mD)7H>*hQ|hNh-#NL6A(#Dx1U+nOWr23h60 zPoev7h)QbKF!h*#8?8FG#BkZ?r>oU4IEP6QY26We5kCAmP?}OsBx^8;P`DKmWxHw% z)JH|aB)mBRD55~X^`2gq;sG%*ezI&KWu`qSv0RjMHq^-k1N`h}v0(GL<4R|I2JZn} zw5YG5#S`3MPiJgG@)}|w%eVa1ONNH^KPq0|`If_0OYOyzXz3`=Q4dk~kp_M4+@O3H z<=C<5%|Ff*OE!=s0>VAEu6WRa5iIEmj3V8*humJizZJ3%$lqKZ!w!7y?-z^k7NJH) zz&Ukks7j%$bx&t1ehoZXc?kx_4g@3d$Qp)0rCMIrFx3%3)XN$^9|Kv$ZBG$sBFqt# zRD=MS6KWqj>{+%j;{faiwbMSCuk!Psp|E`MG4eP+^-=V6EGQZo9c@7WrcaxHJ?PLC zm@#)c0~&J=A@trTqw4h>y^B`><2-d;h`J$30I+>Iqzr$VRfz;fK%$@)EuDLUWR@a% zyqc!jn#aN>h2qy56ldAKuyS#^Ljy-aq80-&r-ofMz#b-zff+`F=&^ZM`yh^(XGHqk zf%Kj4{f#qvWH&;qa))rU=)nuN32H`te|5k(Y9J`mgnkk|hj0MO23G+DKuG{}5N&h}?L9L$9HhhGAsM$_ z3J8~8z(uWyO@&lEN>HKCnhWqIyE>mEqWp>(Hoq{^i;z$Q5RcPFSPt18;cKR-2$aMI z&A)1-t^NNi+uEf?hIgy!*gu?ExA~59rmySel@|Bj`ZvF5{?B;<(@uD7l!cazhf;6k#mtjjEHa}V8&J^ofA?R`a!;*9Ck$VJ9t4`t`?4UXBuSrOX_mX#c&CX0+zHVbWc6fW z(7bL9ipDQSTxk3bKZ~8rHy(pn?Go@41!<4J7-fn`D?zYW{Jve5F+3MUp^J|ZU*E)Z zMioanV2)4(&t$*DU42=85c>DJ4-KgzO4I2(ZaJASm&>_%Cb_KtP2rCB0?~BU7 zhCu!gmNx&jSPUgOkaRMzt+^_ zh)7`6cO!Q&An+Ne9_6%U?gDyWf?WOH$-{@21M7l5Wk`h-L?@1Z$dn#LWo4$p%r3Ow z!N5H`xUVT*0d+YN3WvcKQe_yqwK>J3zV;BUd;11Jm8U+9)23=L3)3*ZV4HUM;38+x zSyo;{msxm;w0@(zaK^kM{R{^VkaT<#8LvoD0SvR$(;3)-P|my6C<7J`E^0whF@*Xl z3>m#dN3W9mgWPV+XPb+{!b;nZK75K7i=WC(y@j3vnejfgx-sNrr_td*b>&D9pfzOB zS`uB*Nrm}Q=%iULBn+H2oa4^kfMSx*=weeVvBG24C-;V?S`q$VC-v8a-sdub*HXeG}s@uXg z@@eaL%=cgaG?I#Zh7?<|86TH>UA|1%0?qNb#(*G22{aXa?V!-1JLDcQd)gGre+aF5 zw?P&mI&-*24cg-27e&L3|Cb-#O9PG!lxsTNNkKYAQK-qzY({?J0q3}@Q*9o`T**n; zgGvV)pZe8Jh&TVhwii?#4SIJ%4DCWI0!XCE$JduF;1}pMKZDPAc?>H%^y$Lf74Nme z)fY?KC7hN)`T{Udo}y0~RTPnlWrcm7Wr9JJmr>Y#^Uw*LHBZi01usS=T|Wn6 zPZ%f$ip?J#iEF|k_Lv13;puRcsQEWRC^Q#j?IV375Ub}1yO%}O6lk5_>aYTfoR~^5 z^uNS%04VOTxVas`%EfpT`SK{X4rscCj=`B-t1y_+vyFCmT?Z}$(oH7nL0XsH-HE@+ z79JS%>jM5QkVi1WNWOJwjs}akK!!-oabPP{wm0pBx-rU0S;GLh4xwJ;_If==WE1#s zwJKH>Xn|^xJ)#cY;E)i$_kN^D16*U#%L6((A@~jX!7~K*p?8ntW5W(8$bLoqG_x6= z_Yw|wzT^v8Jt*I!bIvE>D5!+;gY6i91Ky0wV^kbTa~dp3&(6X1D0{qs8#ofHc;r`9 zv8YA~5N<{EE2uu8$?ELPbQYK6bj@2XE4%%T(IIbK8cxt-(hbc4#0cO z8Qq0^($?GLrYin6cI@Dfw=(|6PA?%gBKL99z&`dYYniTD#V`8a0*;>bf0vm6g+D*{ z)0D5MfLETNiAmsmW7ExhsiUoZ2oenZgI4G#z`1bgh0R+`N}(8b5vOAJ0ddqxP!|gsJ= zpkEX0LbkdwkVWP#Ui>hBHDm#AZZzd|LgTk@8}jRXQC7B%Aa3NLDsUt+?EMK5>TZB4 zrHv9oDr%pqJi0Cn6|i8y(pkgU;Gf{++zNYZ853GIL0}=9qsP1Kp$xE#jMD?a1u6AV zanx@DWq|l$(9b^-EffK&1en#VMNnT|xc!e=TR|FGJ&uD%y^W*h*Rwu+-~tI|VG+PZ zR+P_O#fN#XGDrE1_7tP)Ku+fIk|G@By!!_Vx#$UWZOEm-Z*e zO0q)jtxJMNAwdhQVtq=n`<^>e+NE7Dkr|Tk9#BfC2Zrpr5c#4AhIQ;OKjF zR>Gnv3~q$bK}%(PSZMkBP_SOM(MCZ&p4QYJn<;;JKHr`;op*x=T($yNe*^vWx-n3u zkJv92LAM*h?WC%xe_yqrXJi$tw(@Ud`~N>@Dlj23BDcr86shzn~1%Kw@d zRD8!1?nwLSn`5F7yXX0ewqMZJX2uHXB(45qF3%~|E)0kq8Xo?&nNn=hv3&OyTmW=Q z%r7!VKRYD+e7uylhGOoN<8yn^reIQCfj8W$-V9I~hawk{>Yf-r4PZS`o`h&S9y1Be zp%YuDY!h0p+yI8w14iFwi}nf{+k$Dlb>nO%AT^=%{e2LQ!_fjgDN#D>aaMXGgd zs1;_w*r2ChsSS1yGrf;Kd%1h8@D-|yxfGoH`WmWs0nMTgLu{h;;mTXfpygT@h2R~e zDs#*k(Kd#<3gIM=ESi;lP40vLw~m!f6QE8u2y}B%(o_w^W|#%xaCjSb{RXrYfG?IC znsH+zTHoq0y_V>al(kki(;$zgqCla5E`%<~s65;HQ7(B7POn5Jpbyfmz0uwdwgu^X zgP?Ou_tio?#BG=zfrWa$ZS=OgWZ1*a+(Lw}f_pIw)HLL&41?Z~D8YUKOZ?P-{wjVa?$NHcL8KG?* zOlp=1>Sza5i|6z0{^c{Oup&UsEAPazoP$?ZknCjniq2bV(s+DF7>x_wh-sJc@$IlF z>>Zu2Has+Rm$>Z1Xsy6r)4P#ZgNoW43^P-Z6<0Y#$lXJ&;LEiyVfcQ)%l2%L-qa?Cm_K>J(qf<2wZ$W%z-hFNB}l3Yqw33@6yp?5$pqwNl9 z34yc+*u79c5e*k+)zykFw%| zB8-|C3kl7H!1PwYoHSNQefA~?$nE(|fz)qIF^01gF?zXB!Gy@snD|bjq(I_{uUT4!fW02aRV-!D0HQ6k03TgXI0`U}Bmyxr+DyF|u`5kgS>%}Jf8m{l z0@`yCmCQt%;v1vQGLMtf4OZqZ#OrOoe)?S@@s6Uh@{Rj**z}QoZGkOdU9^bbHX0;X zR1CV2p`yube`ysQdoX}NANxfsFsyEEjWffsE)6LZu!A?4@rrKY^2P~tB9Xl(zm}E@ zBL?e)Tf+$i)r9YOArcO1X_d0;%YS$Wnl;oiO*7YC)nuz1y}U;6wIUkrVmlxvM?1a1 z*#z-*HgX^sHq%n>9m+H-33Dvg94`cufs$-+G^gU4eG>-2j=hOiRk(7RGO+7F4|ty2BF!XuX5hGb)Mq6`G4+z6>$*ezEweC0M@|(Mu6f(VQ8} zu4WL~Xk-K#gII+f4E(D0U4|p(THcV^ zGxa?B4UR(9x^-d^)#lDaa+A=pVUWLQe%hnBI9C0Tx2=zWgg{sNC&&$;3L`(535ErE z3as1ilrjkXt=3Ol40R#-L8lnY)>0q36{Y?zB#3E*3cpZH>dXlf_K=1w~y(O{0V zl4hI$t&MWsuYS?O=y0KUFlxh19qJ)L6p0ejHX9|(WFsCfux09eA|6A>moElSn_!DD zEkl+-5D0-be|Q01B^Z$fcN7HCT~A|05R5L+4*}ADvT{8Q4aD@HE%2&DXoFf)V`|yK zb}2E4Fk2|ZJ>+3K|o0Yq9897 zS?gik69y}KgQK?`ttvfflpVo3q9V;_`{k9x@Mk}gN`p+TOF0Az=7Lea6Q`yGs9(ad zlKSi3!>H#`0~?ZPcnI;v#i`FQNpc1;^_U^GC!9GfcN{_iYGb9&Ltr>Pdc<);w<8g% ziQ%HtjIX1_s-lLtHG{MKN?t zbL@I3K2djb>1|~0%W=X1gKfwAqSDl_)f8l4K#TMra0qoWQE!7`@EKg1i<$JfFdA`? zCR1o9Vmc1NJ-B*|d10%;)M#)S*^mxh#5M%-OA1c#?U0q74MGy;pk22NI5N;lcbm1I_U>yWhg8zdpa}|*WT0NFSc?mv>>J#F&W7pVr@YAW7GG8qr`?ej3?8AZZF{1`!byzAHHwXih-h2HmnmG z3g-}PWtZMJcbi>vp!HXDpzhKiT#t$r0gniJ*QSAXQU@^}s!B5)YNBv=!i8M3rcqBX zP3lD$2lRJMUW?q+1Gok1i&l_a`4hRM5-7i5ODCN6WxS^SlJXmVz=lI zTSt@}qe}G{h8NNji+@8J8W= z8Nn{;oQpqJ6qSbjA4D1sgha$`(Tvf(?8YV(4EE<~YwMkHhjY-%LV6VNQ#2VCRCe(Q z*8|bH)F^=T;ic3#jSmJvU4{xOwx)1)gkMpYCc;f(sqRBt4=UbEfIWeJw-tE^)q%kF zL^b|RsCIf<+B!_)_C0}iZcgCB{--lg#09R9tRT$@MI3I|pFp@#Du=a0lS50Y#cZoefst z@OvRNY5=UF>by_^8_NrTL_N?`pE4sOqeC#>fKfLCeuFR+`s;odud$heD#Qi69o#e& z|1KlcmBLgDtlBne<*=x#s5l5tKH3~aPK?Vcq3D4i0A}J!8iBIOm=%cIa=|VLK_>ff z-@Tu4n!u5BN17b^;K6cm3=r4Y0HRfhSH>t-bRmm|=gNk|{lYDf-6dQv5zTNgrhh;k z5i_pftbRvKmsM0W2l+B}8Nde905%wcd$(b5cOvPf%;mMW+Fi^Pn*-BNQ}05ps9FR1 z!X>I^AaCM_KHE$X^5C>z6cqdl=T318>O?p-u6S0<(T&yPfDlR!sBj!CszW@X!tH8x z4bboc2Ksr_M)EHuL(E`%hcUV+9|Xbe^PtS_t8mnCVt#9vg7_YU5)?8vL9#)S3h1PYjYL{rPca4!JU~G6 z*T3qu1+Imvl#iKaj3LXy!f}eM&D-9sV!8!M6ey`O>O9sx9dR9oU!8{^s3U3}WdE=g zAYnWq)diWZIv5N7@oGF&DtmeAGXh*)T?@n}Ftm7g&utvJEWk3rg04Eqk~SXp$w9bh zu%P5%y6CK-9>s}fXB-Mh|82)2X1#b}4gew~wwL8oXQvnT#dahpdJQ{B!H;GtqSWA1 z23z+G;}k%$aF#LC#QR_~YzM_pT6!roa+9H^fn)2^PUsmTMOEin!9N>AGh2e#WO=eI zL`K7@fr80}ZN99xc4DRiL5P4S6!f{!?hP^r&CC?gPm>_tF5O2?&gWv2lMoyyVcO3WA$)v|1CO7XiK6y2jW#)S~?E} zMc~@J{QSA0+SuxD82yN#DebG4t7)hRf(7vW)hora6ox3WADn7< zsMjr=I0a-_%p)>QYSK1Qgo-^GVuBcu2Ki;#;y00n0^rDPxg-2GQcD zCkS-I$8nJbCb>_etQ~8^k*FW}o@P%b+x15j#8ZzTqVha+mhOqU&Vx6E=cX-hgPvm* znvXm0PbIipVWUBiDx$HvSEKJFCW_qhwhg0Lfe8UWk(|4g7ebnYsKvTBrwRZyu9n>d|24?wa@V@c@!|%EC0^)K2uH>3@vEilm7D?!#7a=Ny2hHC;*^X103va=!GfjfNsE&{WLps+;Mb;Wcr6rq+u4%MfO z>?eqh&uDTX2qGbN&CL9qZ?~uO9-aZx#T9$X%vrFU~?MJsaT~q*V+ZTU$qnUK6ez z7!KqC=u>Ev*ya2)9t;DP9|tA`<*W)-G7{5Fdg5yOo5Rk&xz9PHG9%*p7XK)@laq7N z2PQ()>k443p)^;SbZ5YNs#!gb@Y)3%_SwsbPw7u%u=Al9!-m*cDPjY#6FlW#n}PCq zAP>0tuGa{rnvla*;}K8}h#ZnH)zpLOFf<$BQOhhS`uz?{05HYEvUu!k%phL6`cdUvlxo5~*ljk96&en&qZbU2`)? zlhfS`?}j3;Xm~MKHlc5^@Viij1(*w@S?0JCl?|Hf6*s!@Q)T)~d@czUxEJn2hwop~ za*O;Ay}7%USx~IOhAx{xqYZWT32H_OnCPicZU-T1U?pHyEK88@G~9>Wfkwh&{Bn*{ z*Dwc~_bG?ak;a85oceem$&O_n8=KL)6xgYQfEE!TUA5xgq>edx;N_~1nT!0XhrFeP zJGLHY7%1faDA}v?5m`@1hkZdid2_syIF7>ztMp1Ussd;g0VXMza09V}`=SCsu%idG zl|*3ZQNh*i)<>l6fKz)RD9GUVp{1bX5nTanZ7P2qQT;l>KI6wbi2jG=fR?JEk-dGz zqAcqj#lOBg_|_`=ejdUYe2vHsbc&`$jOdRISg)+Sb>Y(Md=FN(sn=&LVw=9OTR`AM z`=zqCC)^VgobBb-1X~54h}Kxb+I-$>+ppU=LX#Z5tR&vgP(H@?pKMK)g&gazHrkrU zlvdSU*>YC%=Rrrk3>K&M!SRkYM`9BX#j^bOE5QaChgWK6iV%j!`CFbsvGp)$E#|L9 z@Dq*YwD%Vp1AG$Gqw6Apm zqW{y7B~2_aW^9HSV7G73*oq>l-q#+5yTaa@zf}?^XoNRV+4!XL_QvMsJj6MRYsWjS z{g3D)FrbDWOi_8pe`|0M89#g29lI00#ilc4hYR-Bzu7)lNsM{bH%!Jr7Ec9bKE!5-jm0;q7gQ&Q)On^8r|z_ViXhxQ5{TN%a{;&^aVquL*A& zX$c{J2=$_OE%xxK5>6v;gXB@02SkWIYpFd$oH+Z@;cx8fb*B^ja6 zeALJ&2hVX4yPl=!!t5x885VKn`u0Xg7|~a8WthZ!$b&~Vqht^zmR??=$nH}=Cjd{2mKhK- z7^d-&A==V|w8GFrOelyE6UG3i
    W|1q(LXsI%Xx`8~NgTLikQ5W#Yc;L zXH-S zaPt#k2C`qDO=(Xo229Pl+4Ja0Oo}moK9yY#y4gcfaEZy7iqnuMJ7Eqk(-8;#Qt-4f zykphc)Q-o{ciWqAAGi;U>O$}gEWnO1I$4AQ)>2f$oB`baX1HNO1wOU`Od)TW{Imib zY=rD6;%;&pYv9=lWHz_i3h`2BCSyVZ+jS#DS4l}J-K+3jmn$|wE}Yq7<~A9#6<)K( z6Z=opOS0`@p{!q@#k{zNKr_5(Be*?fq2(G3WfcAnL$BH##^&e|2I6iB>}M|1OhkyM z-gX1Y@HCj{@#~H@a}>gT@M?(ZN(Oq(6i!ByY(Aq~_+&ewx0bDh?@Q4%*2K z8}qL!$&5zD7rEte+pq4nsN(gKSdK)S4H?q$p30)c+nFEFepo|mF41!L!fne_xERH? zPgle|ZMwa*K7DOPNy4)a6F>iJ&$4Xx;k9ebZVc(+4V&sDIpR~Rp8MK5!nf#OuR7Mv zG*Yy8EnbUo_+yAoSMryn5A`>6PF+SG=usH%jp9)rWLmjlqC61_h_`m24GXC1+1GcZ zH_wsodt+J?3xpgz8XnY89uA7o|8>BWe~?O*LGQg`X{f^Nk=z>Gcz2}1xFAfkpK=&& zcMk+Ha>y_@NJb(QFt2qSq=|LQg@lIZdUU4s7MXuAeUh`+dE~8!?11xq>|GI1Z*3Z{ z?hHl;vpTXY)e*0xum0V%x;aB{OlmRVtLLey~0w`)HpP6fk z$y}X4VQBlT-x%@~HpCOuh8)B~O96YG@VNyAQ02{E5`g?V;T&y($?Fv^pUoqarazhv z(5>55cI=+Wg5$ZW;pz^9skNRTPNl9nTWRFyQMjq`%J##kT$tBiUU~GSyoFURhjBKz z?HmXbO?xVFA@o1FsH@QNQd(~Tl*k?NO`gKaz=h%Jf`-T-&QL_v4Nc0G1Cl9xVF71$ zgMB8BVJ(4IEuhyDK<*+KSpnj!zWW9r5{|2-4?x8ASYJ%$OPIzkr@>UjHj9u7yvp}L zv>&PI!-IGTYzUd% z;miaPz6F(NqiW>jFy3xpb}b#)SW2Ni8423i0zyLWICIIY5u8nrUd(X^KqY}`77J2~ zK3Wt@hu|_ZdIP)Ukc#Ivku$^5TLk1jzhc(0Im(+pU#oCtjB`0uLkffwVP*an))h1a ztQ%~wTynHzXiF;r1$lu2A%m!r8;8=gw9+{imAh^fvvZ*p5(0XNlBE5Uk7wPnld`c% zYZ~JjcPyftVrjFWh6_o8ZN6iHO|XJXkx3s5a2%-Q2%5U$g_h30&H>Ba%FPkR%3}`u zosJlH)_8NMu(QgJSs9JBrH_1+VEyN7&pF2;;%3GND2#F;5;>mWX#LTuocx?2wt&dV zibwJvo`@_#;@u6o?L5;bABj-+x+MxNY!dBJ@rP^DY|RV!{Iapx-hnL*^Uw`OTe-a> zOd!QFp6gRHV4fDhhnc<+(b6^F;ehT)4#5+W*iMue6NVU!d-P1P24K+ug?leJq}<=9 zb5}gH7yqNdYx7#f{9MSdY*9+x+|Y|$j?GW^i~ z(UZE^$*2=N@hECXP_2s@*Oz5GGoOI&m>1$L!zo^UdOPY&=_Lb&Mzby!{>=rz8KS5S zq^Si4<84}8rnNUdVR@+3nB5u{#0?yd;}*(5?Giwp>&r7ItaPCV2pM~ShQ;_jS8h&J zgrdu9@Bp<8SCpP77mDuxYKi`LYKXnWt8iXZ&7|OZ%)uJl31dN@@Brr@_2O&bGaUc} zV>PVI2FX}-A7l(9#^wNiuZMIqagmKIYh(rC$xLFN^53y*3zz80=jFYABF*0r*ziav09lU(-)*WnB3H5;e6D#Nv*>wPMMr2G%#Pjk5Ha0f0v$Hg|u7{ml0G2fi5k;uPh9_wPtqTCP4AlPf$$h1Ys2*~21fqsAnxiECaUq&W zJix;_YN@%sqow&kp&mYNL$TC`JGM=?hKdCASj zW<10!_&L~(5{SziVD?P<%x;WfWF-lXx2MYh!iq^Bs6r2~$vdwRvWMcbvHjh{%|Uq> zzsv~_8V#$870bqfNJ>8CEj{qC+{qK;7{n>_kobH}tzJBHSc$B5ceO?F4h;>B3xm7E z-gaM)b>LH$sJjT|S~)QNW-S!IgV3EkLVdzpWp_yKilH3MlR5gas~+p)nUZ44*g~VJ z52G*-HbMU4bqBZb;@0+sC&ZFF}Na8oFW+r4{t zQp)ukVPtP}*tD{rWR8Ea`SkmAa8S!^i&k3%Dn@!f>aoD!!GLOpgJ zs(vBNJA-&!H=>P`@5`&4@vt_Ng1aGjUeE#kiqO>TV!i2GxX5Izu^})vI#6KNXhKWi zOymF*pcKi(kkM?xA?Niwst+q@n04`2^O7CxIdFFD7;@)86C1kWW^!QJl589>F77Zk zl{v82yi2US!MFuR^9UrQ?W|0X2jwQDrqo7Q9YFbCuh&b|(kmAwPchIVa*ckkYv z?45x7ybMGohZYR!m~w}Ra|2iHB}CRzsKU=jD5hiekqCM6_%=j;vop}>NIp_PG_ zs$gqK&~_#xNFjM|4Y*vFC@iL!{^(#vYb{2Lc~JsI-ix**IW{m&7DFVYZe!O+tS6me zpeRx>bzJv0TAb}eTo8lJ$6PiP%?vT;i5W>i87pZ-G8FFPk2y;@^LU&cmRP=lJ+PaZ zpQf&7I94bCoAT@|bnH~(&>|8o390!E5#XQlYyS~*FJFGgE=+Q*?tYFnUoMnB)n>Kk z@#UFrU8o>Am4)Tp(U#}sk%yIR6;rD`kEva86B%z~%r^SVR-J1yC9n?!I=fim)vG;# z${IiV<=mI|uqq9(E+6Ytrp%F8#%pqLvcn4QosttF~NAFIAke9^0{M*a1?Z|vO(xug2Y}Q!2W8=iYk)L z4e6Z_`yqj+YawqyF(!Ik4VK^p1J1Z+Ik9mRB2nVbO4>;$wH>e-iEfpR- zpqs)%Epa%}HGD|X@8DLlFl+HDbRM*tZMZn38 zhDkTFsl>S0kH8bq<3=S;;zX~B51@Wx^viJ3~DZ+ z0-git!Vtti|Jn;ef`V??6Cge`G?m|=9f_m0U~RDn33<}XQ0wzIFu==}!=VA+3xS=W z7zAt=wF?8%t)ad&B>qEv3LRWRaiR;&7UJOm=I%k^-iT~IJ!Oc4VYd$6j}<{`ZSp2# zI=9V1zz#><+v%4mf==3)snT(*+21++tTQ+kG;gRI;Y>CP03-LyZ??6y0mD<-zdw06 zttHl=msT1RwVmmfTU~qhVzt?Y)4py?L!=T1{ECiHJq(gJibAyOAa)@K3myz_4{UmV z>X#&)Yd3Dl15?5H^$iWpw$6@)+6l*0P}UYjFDV;d>xLd-c>uJu@%8ldV9n&Q-XIw9 z;}+Nv)|hl8eKT)kqm#plnTX>%aZ_OBKH0tPOo-)^FX#j3f(D}?RH zFdG?UeiXZ~7f1lSAV_dHPxJ;gAg*`AEf+R#bC%A-NHs5n%sUVQ5Z}c==mJdnrdHQD zC8p^F56P;#R8g2{OeEImjT$UX1mUF9iZ|JHUlohSj(FJ%h`~;ru@IO^I$X>fZcH}j z|G>!PMNwTH<+~%1Uf?>Jq2d~U)!pZML}a9it+fo9(_=r^?E0TsJ=fHJneV-MN&NMa zQkm6(eoM`3og4E$tv&UXY_gy9kex=_MV4x#9NRU+BsT zk_Mf3ZbbP4jz=3K+A70 zhQ*26kN)-%xfUT9GprfLKEpHon6hh3xf~X>sw*eJKVAZP1 z2*>sJGm5Ev#|sVT7h-fR38FP3|Jz~At2>YuKUm@+2EJQF84%JTY;MiT2Ff>(`V) zSP!wZS$kp)U?5E<8Sl5+8S1;Mpq8;z~}iGu&Z0La-@6 zZ=%mRIQsci@;^V7BBX{iM~jB6DQA3(jy=je`Oy9CtCtNg`V$qx{{2E07F;4KnHl=+ z`u1NEXZelA!&ggT<=PX^Qde!;ro!P(1lpySpL%6Mzx-_AJ0~pHa$@4C$f<)k^Rtsq z=;;M$YjQ`5m+=RFv$!lQ6u#I+L#ssvIW#**%&2;O%YopiQ08jU4}X1eei)qgpdocMjBee?r2X*g zD7T@s;I*v_|9-{af1f>#kOL&f^c*Wnv=K$}3yc`Zj#d`e14oH0uL6~Ir-DL(W^ol# zi}u7$veC5@15dC@zrnB7Bh2x_-*(Zm^_O^;V}@7q47~Tb8v>I}Q0|T0_v@z{6nL6- z`72A_e_o@n#${Z**4s+MYVw*qnnEEpVu?t7ldhEZ7+4J9^|}k004v;BTlz#hlIV?v z?U0C}Y2^AQ@M)0!0-ICb<2VGSM!rF$?9>=H*o(%MJyPA5R%W5y~@{`$m zb7@A6abk&mbCmqYxAy<}`>}3Ov_u3+Tm4D$DG;X2fy(AU=^;KJb=_UBFb4843KsqP zwFpNfP}O>3+ezpT67Y&e<>?e_#RrYA28NwkxPcLf0!j{$$pc;efZ#8q8neMrnm#Gv z9{h4~cWIaC~S(AQ=+dD-s6 zMBq7>Ehz5|x>24}G+e{`_f4=M!C%0H-l37VCI;x|Da4IAp$fji4<9}(;XGkt!mRIc zol3r^ztmamP}W7NEJ5EOQE1mszL66x% zqs7C}Ytw_8_U1x=b`VKuPN?Y1F(mSQ%E$(iJ5$<&p{rVxx76yLipY`tEAq!HkhC5K z+8;eQ;>v9&Kxu3yW4w#YXu`~?*Pk42J#u5yH|EB7Y>%Rr7pB5VS6d-t#J7~b`N!9l zoZV%%9-o*^#UP+lw)}! zBNuLKv+eSL@7#~+C1ymmD%B8cQ`5Bx?l%x?qUqoIr&CJr{@LF#G|e4C|`5d z4&{m;Mh%CGo4E*s+Z|7$GKKQRW)a(-MkYF8U1Qsg!mRi)Cj{9I_O+w>7rf|E9?W~& z(^)q2ANPZAAF(}F4Ie7FXRD&Upt~>C->2yKm9U!1g)W;|wU+v@Q?9^t-2eQov&$$m z;u-ri-5hN<;w{POTFs_f9{T4fWgw^+{38ncDYgX{`6Q;_f|W3j(m@LIF|u$0vZac) zG9X0B0FWuk*GbRIT#q8R#CYFZLzL-fOLhWe_iB$xT}wz^#qqGf!#F~h5F745DVMT6 zdh3cuztLM8ICSqIO?ygX(1e?#`Gjh4SY0oH)vK%9m?x@2q;bW!Yx+k0alEQ|L!GUP zh8laX-}@DTZa3`6GDR^9G zc>TteP%^`@`E3Bp>d^uI_mB-Lc89_4)Ug^vgiccuFQL4B@9L`wjO$x*3Vx`i`l&_k zQTmieBwa06oJdORU7R3SYID;5!&hAqDnX}d2on!D@roevNHd^D_nAf(4I zKc*5fR7IilL$~l)O}@ZyPw2Bb{9vG9&L}5zF?UOTR;0{;s1N8n6ljPCeR3&+DfbI= zB=Hj`<@eqByxww1bxK#!#V0J&o)nU-_)T~vo0?aV_Y}#*G;ckFQ3s7Z&LNVmv%bCa zv+IzR(2mU0(9&Z=4s$sPs!w^SK2sT#S#|RN02(Eg@EsVVwceJJACFituA{9b8F*U5 z)t%gzQA%A=MIL?2D;$>B-yD`VyJnPz-?ZDawd2FR&yIZ~k83q@)w7&6 zpqeF!H*DJ4O3qzgu@0Ywx)E3Ag6R=*_wL<-W@uIBS)+1Aas86B6^E2$WUueeK^VY| z3#M|hJb-Rrak|krJ&9?lJ>SF{qQUjm>eYihQ%hp?UX2V{>B)_i#|jzNlq!AC&EmIa2^5zRdtfXcpX2n{o zvyW9xzA^93DH)l6h2IIYsHU!a`uSAmwtlmVs!{+(tiS@1A-HYkqScoOwM0h!q|lxn zFbzP<2Jl}zL6;DPP{f;Gay-q@?%CZN#E>F%2uyt?Xr=9*nWhex7Cn}Lc(tB?Qa4^Vf`1fs(B)gVNZh}Za1jBHdr4c3lDB%Nt!!Te1 ztsrm%{#Zp#Z2^!Z80LiGOcZ=*WyNKu17Wp4Vw@UFxh9LJ@XARkWmB-}Ax$Yu-K_!Z z@3w#4H-7K|n8Jn{&Q7{sg|HR;fy&gKei%#g z=Z<`qd2MnGU{QKcKPvcxs+6LQv|Wlf_RG@=DqAr*x~7lZvSD(Py2aw{Qwip~ zcg*;+vlzmg1Ev~P(YH}9LyzXPUlXKByUMWd0-3&ZdUrL%v@6!0TCuQ-5NGTaz z1Aj{jeJ6hvs%uUPJqtBr%e1@QuMU^+%eMObF?`xl&*PQ`chJcwB?4$Cf1Nw}ortpB$Gro$^kqG-t$3RS8f|2|ZLZ6uObt z@MUEC|lP*0$)shEy%?JCr|9#xv1boM(`z)yzTglIJorf+FS18sm-kh$a65w^y@|b?8 zYDU3YBzc(|&OEU!{S_auWy;e#NCf`zbu4*u)^*@EuYFVY-nqr@>ZQNU?rZT;5@{Wd zco-S^qs92oWWL3GpUDSJ@vJhVddPWtBY$AYL9H=sjn_7*-*KdW{{EHk{J!;s zQ>%pa7RgqRXSg zpZUMB7B7@9xc|)sz*RmQSO$N|HoZCba%%I6z|v(JiLA}z5z8lrz6X6VZ|m6d$&dA< zFk)cZ8Xhgz11)!7I9a{WyeOb&f6sBHA8qd6C-0A>m7BD~laKFrT2i%UGod;<9Oh+j z*}5mHeq@01AARRt#Gh7%Cp+3W&FT!GYl*7a&u&~K=q(!?`y*b{5deHt_kgwY+ z-X6m(qc{5FIyKtc021^JC8T`^Bw#+5^`S-|H?cFn{qG-lr>-he%5j9B<4#S8@;;k6 zXshR7c-A&5GU7*je%3K<>%wm=+_6JvUn=HI$!pr#RmQZb{qGaIzZhu9eSVPgBh!53 zJFC=HwdiQh_B=MZavyc1ialelVj`Wl+$z*jng70N_WX`fUAi@|K58f@|9s2-YlLt$ z2}fT3H8YwLu^(N`ipdw#AwYh@)l2Bo#$mFDA&MJwW($8OB4$3!ExeO0>cO3*fEAJ% z73T$tlwg-T#|*8i$4_Yd7gGN3GBCdpJwJJ`oE06fGRk|$I`yhwApFA~y9_~Y1Jyvw zY5)G!zkd4gy@yeP2&Q46ZusikESY}C0y9Cw=Z}>pCl$=Mcr^P!MYQ?9%`Zz5kE>Pj zWkyi?=DBZEVg~aVDaJ-Oe`jGI$eG1A@7lYC8#Y?Zd?oFr{q))m$2Enj8(vvteBP_L z_e3XWlV)CYNb=pAzXUJ;eVOF){N)1vKmR=CMaL_Q0P5a(`=XQkFlJkn}h@nxcXdf#liKX3Y+=1awtZ=S{$`?If` zXEZRlGy*I=7loU-NwkOGwlQ>6E7kbsm4E(t%q1=Kn>>Ie!J4r?d7YEfX_Zp314m38 zvs-;n7~1Zz_|aS3oBS5@%m=;(57n{H@Qo>1&UK>HylR11TvEiQR>ALHMZy;S(r<5Z z$Ti3*P9Q@pbX_awn#U3Amxf8JaIqz}hDLt(Ev|g$D!$G|hClJ;)eiHHXXisB6Y}dl z3%k=-ngpl(7HG-#-M3kP#n1AaO~5jtqYv7U=PIukBh4%I?(fjo{;-76+*6^ ztc+>?q?KV|s2y?=JS7#p9fml_^7^7%8IM=34LHDF*lPKMCp+{XTZM0|7R#p>ZZlQQ zKa?yzT+z9?>_ZlhLDgaul4ZcH{n5KD-duLe?3B$V?Ewm@S#jTj7s^-Pzl5(}m;q-K^gv;}!d)s|uzWJ;_-|pGY?>v*B zDEs63)ua7B73qe8t%7TMoDCj&=lO}Q(EAn^X1>We2A^>2H)$fvNuA6hTk}DYrNN5k zrAq7lyem9>cgpvz zFJtfIEt_AR@v*mIAY`+giW9Cx>*M0?j|yUOQNcgj_i}PuC++)Vw^-9>*8lNhbM~ga zwstYK4a(Q|8ZgJ2*xp)qWAX+^i}-*;IdeFg*D=y~AQU z{vlf0XPuw4iv9ax9kf@yGHH7s==>@@qyA+{j(K-NN@(PflU7EO%Fe+Xxg}y&NV13d z#9%4DUOk%mVIh}EsrJWs9Rbgw_bQ5u?H+P`6JRboj6a)9ry)|OP{7~k$nnLzhBf^KHB1b3y#yN{XLs_70m=z$RS z0|S$=dGcj^dP{KB@WXnEGu~r#1jc7_HB|Zi=m@;&Y~<7X`!W3W_qn~X%Y@m=9q<2b z7M_})+Fspj=lFx8Sap4}wY{=5XKHwIf6JV>;!~T;bP5<FFW>5tYXUU+@sOQy?Vpm<$CV4yfl`k&NK@DThnFDZSC!b z;DIu^JW%Z-UI(hXtPsm5iUVOLnwkU8g9tN$k`RFyJY%h>!=vEIKYpLy=~1A`sZ(CW z{~_WG(W3P9mL?`9f()`4l{X?I63)_rE;AKO!L|GtSO!Ft*pG22^MUT>f!YhW4ERq@ zA_oE2svtr;BxDq{3U3ja1|7zRAg0h2;f;CH79GHA(dMG)&}k7wLOLt;#8xhGXW8O2 z!v|^YzrMD7kg(cy=wt*^&UlqFQ^wK*?Engsq#muXmdz4^%_lIYPrwr0}Qb-I*;Q}xAu5!F#J-h9+;hj z7uVKNO;F1Kd3l4axC;Hbg{UyqG~EU-8JZ4mn2nZ$79%QwKsu^?d$wifJ0RD22>U&pKCXex#7#zP7Q&NHLEutuYcL~~>KP?KSI z_l^1YG0{#0qmU@!4JasiHZ?Uxbvz{=su2&CJ^v4COF##ek4h36T)Kygon`Vot7?FF z*MmH)n3C&C{5J4}zz1SpE-DY04HK;Bx0W7d1sVg0%pd~2yrB+bS*|J5W?cd{9%}M- z24F;U$P?yK-xGnE?&V=`ThX;`6$OS6QQg68^`_1-4&wk;;yw+`?7Dc{-=8W%-Dq3l z17rVfgfzO%7ZL>?EsEJ}TDq*%U$CJ_g4au`Cm+2f`DYqBB4cB{sjCJT!w$Ij*n4g; z`rL8bIVcOG0G$Kk)e_?Xb9thqf-!!B%Q(nTWG~oD!q~*L1n8PtC`Jbv=Mj{l zu{VQjfOZdntc*roOAzGO0$cVCX<~WQTQ$L>@bAaWazzYLvD5H<6hamAvU5BY$_pxH zTX0ci;b5`r&^J-!G9lWf#J?s&ul9#{h)h!DC+Z81*Om{*U?Fe}al!+N=;DI!*UQ*9og&`tC;AF^sK}MPDhkUY_h7F}?iY!2{wM z*P~tFG891bsOb`<+oGz8&Hkw56dH!#%M@*-&baeE5#}K_#ykBn};%>Ny!HfdP2-V&Y=l>>%l}Uu2el7}q zi@>XV-3H4Uw-eU(D6<}aCNeEeDx=%ejP}k|OSU07F)=Y_w?OLy|HDw$KwoqPvzQnW zA=rAt;UdVqxxcOqm`5~kJm<#$$Jcj&Re5DylQ_miqs9bLq}fFUQ4o->Q4tFuA|TyH zZ_+y$jln`u#6nk4s&whi5<#g_FC7KpN)zcF{8bDHN(I3Lu zeqGPUQs@6)>YX&T`=Qr4miPZKHCAWYV%HMny{F&5;r@{Mmo4gp8h8WFjDw%c0eXje zt8+r`XwhT{g24xjKo63}X8u!PCU!f10e>1~ei_{{v1h@Pw}j7-EM92kopE<=j4Aej zFL|~i?4X%oLUi<2yw~{gPSJdJ=v&PLOPX48L8rHDvMeO&L8ECW(W$otUHBOTuIy5M zH(!zi0r78)aOHL#u9*K`7AARd$d%TBghq5BBri8>?+r$z0Wmt?GyrDb29z)=pxAA72 z`-6JXukRhjMVk*;m|&eJO;1HruhzESI@rR@3wGQm?zvxsW<;V(ysrgJ^tKmE1U#Wh z773l9#Bgm($ZBL0GihgMLGWfsB1GFY@b8#lTH?JHVCGcanKD-oH3GLmu&pYN*wcw# zDh4p%qnB+aJr$bDJZFjYJrLaQvc)`iq}I^TQ0@`dFT>8V$)-wvpwGfx0)v|SrG0^zqTjKM($N`q0QA0J+zSLO|7*FY-nqaQ~2<=?ETDJ4?%Vp6k0DSiO3h!I1hqM})R$@_J*aksEI|8#OWMrvbO^Sk``qS+39Egx8GB5U>_-A`3r&^voU)zosY^YIE zx@Qwul9MewUsW#h-hQo|^^b_jOD|zZs zi^-paodXA&+@|`(n%$w15Nk%RyjU9@6?K(lTEucUG%p);clotX?ypaef!Zz-LfQn| z>5=dlF>s6ahV`?Sf%f#?K}Y4mV?Lav5D04NVbCCCuA+>46no%LrK~tqE)j)oP}{ac ztI)9HQx8$8NGt99DRT>G{|)-qjYCNpSPMAMbwKhG3#~apyj~0^`Jn~QCw?@pEKB2^ zOaB0$ZswHRmfn!@+I~%$C-z!E!K23W`U3@>d#0iTGMM*$d3VUo7j|{%Fp7_6QEGep zbEboIYh(&ze%OanArCZlQk~qWMRYA#afoxphC|+-AcGZxX|aY#0VpnLFp2vx&q&wG z6vV~KoiB_U(Cxm}K&a7}pQ2-SiO!|r0D?&ne!c;yvUg!%d<#$SJqXx0NT|oJ!$1n8 z)ax=)tm_T2$U&%uh}MoMd=3{X2@4EJ;YAW`RrsEVeVjCla{$c@^StgOGU^WglE_(;TKBk2K@pcHArARHi}6tU~wry5(^Mi{yl6)mV; zE#F^RHLvDGiNvqbxQ6)qh-NZtI5h3nr;w5whRP(8h6DEln!I9F)j|Adrtl5Io{}FE z@+OTXK~NzgH(MXL*IZzP>qtWV1GBtHwm37dQ5U{=-F)dX?C>CMLhKmKVvl|gGXsh=|QZ4TnZWf%GKC#>N_{bHolIwG>g6&??k+hp;53Z2f21*yu5n#VnLZ zXOpspA$!ic(oAq@FuT-p)~_LB;Oeqlq4$nt{cX6;;_?JssRPQ&$^{ctHwZJdenLS# z`MMaJ+%~KaqmuE8fXlk=T)0WE^RQfMJIXabo27=zxbyRm?VOJk8+x3hS!D`wNU`(g z5U#GS1t#!*VU?MFzx*7MFj|?(hsE&tC-E?>Z8$u*>`rMpyuU&c$_xy_$>fy1e5-(f z4_&|=APajkdZT7XbmxRY75BH427`p;-Oy%?7Us=j{fh^crsuE=z!C-q^W$!khf>Mk zE~c4gdW|Fy15BAYvBirQ3+gaVfDkoLutTI5)KLZ@iI`nMb`3^6juO2dLR-Y zSTBT!(<&A_*-U4@KW`x}=KMd;mKoU0 zUCJ;%W~uerR2#s;=4YuD0#nV(^#ZW9F>0HMl6R&nVGGSMcscXi*NL7 zGy?CCbjTw9cR-6q)};oQE*By?e7_EoP6KPz+95&rLPS#SM^iMJ@xk|K8^i*|XH90) z$#EnJ%SSwih+792YjHcM;yoE)K@pFURH^eac~V(oi#mH;E{JFCQQ_X#@rFd(9VG%t z;*J_Ys0>Ca5=7e~VbqPz@7V=6Qu=cM>7kumAP|5&Yhm&#K+=0kHyXPWHW6EkDo&v? zVte`!^9%OgZ^Ojs)1dd*Lxu?BNSPeAw*J8aX4GAfI$JG&vWBhHlYOfzfA?HD zIS>p5wtC|9i`qea?#a+bwSXq8F~+~KGQ{T6h-*diwmVq=nBz(_A3ib7O< zoMgT++~1mi(8MCMHu@#hO?QjwgViV>GZ}mFyRyG^V>Jf zZ;Gyvy$c?gY&|HYn|I_-52A~mtF|tL$d6oc(A69GAsA6V>eijv7*4cfJf~2lQR~nt zt?4dB_v_#yx^`z$r%b7xdP=1{ghukpP?Q=ApVwI;hXUc4Zd=v<8;AK-_*^JT;kk}^ zfdnZ8!H=(!8!dbQy_Wt2>g53nJk<<_LC_rOs8PVVx^xPKzR~ai<#`J?6m3NFEHI&> zA|^JzNMjL8-L3xu%=J~hTD0obHRn|uaaP?1w6u@8M(*mGC%67_Q1RcN*TJ;Ch*uOu zC|=&Wq?T5yDSMl6E`ET~f#!5${){F{gfzoGI7_xAxX+k)2c+r}I)j2v&yP#e;er;_ zvbT;rP8ONN`SgaYdt3W7eEvBRh1CP|r6>my8<<@WIiU@ewI`e%&YQU*KD0yeqD=as z@M+TgQ_I&def27-2+$n(E2Uv)6}e!nyME(FiikTZj_Z(|3AZ;sMqJtcOn;z@RMJ#q3|J(8N z19AMi(yLt&?rsRjwSL}v#w=0?x*LLf9;yBN(a~>u$5Fos6_Fi_nE8CAB~DHnO%Cl| zNHld}ws_I@-fL2I={(71LW@7%-0!vu+(h;M3)M=8Ug1OfE7?KHMiu}Dx%+*|G6aBE z8SfNP#G$nJd14QrcXut@8WncEz<>a{`>(sF$qWQVgisOPYHu7vJzKW}zf({_B6O6k z2e6ZY<}Fz~AL)ytDS1oz%~amYW2JsowbLz#tN_r~g#*uG?||r*46DTY{>373>7V#{ z;>@-Z#aCV)_E+CG+XtmbGDKuu#@+;+O52+9)5gCwZhJM+47}!L0LtVPrT@n;7wNK) zxYlfO>YzIN-Mhz(Qkqf*$_I;pVqWK>t&zWz;2~uP`F{Yx*%IE1e$1*iJg%m;mdfo= zB|A8bQY#x|3+(6^;Wy#Egfs$%d|lK8^r`?0D1(&%03?zA{7K0SFVW*jksIC%G=*zq zWJKM+5uplFPf`vG7U64m#petb0cZ(z`jkdmJL0N)4v~BuvWyRjD_BhD?1F^KzC?$v zqA@)-ZhY7!-sF?w`rG`0#v2quPyW;s6Yyncx?)8|UEO`9coe#Y@^oONgGsv%3oj%n zaG6heH$0VrHd9wN6H?Lf_siYySM9**u!L~Lz|R%3Nb*k?>#I^FkPY42YhGvgToYGN zVM@!U;$Q`;BePD51V98!i%O_$T(f44V37`hA9eo;ON8!e^b;k#*+@F%V4>QXhL)6E zzO=Q2l}5>m3mRAV7)Snr)C5<4IDF=NCIdMefFh$a1COB9YD(^lN$|zspBc-dGkP$7 zP!Bi)q+L}dqTTawYSQB!Urvo*pF$L_od5G_U8LN`&*?cta>!+kPk9Ky0d3tgxscQP z7~Dqr3%%AalbQZZCtkK5akn!=zL7k51dB%cTV=qdz0)qTf;6NQvcpXyn$tH`^$b$r z0d}RY*SQ4_SFUGKDzfXP3`iOF7q$UTzo!%rJE*)77%au8#|`B6y9%`TBHdcm{fi~d z{l5f&!nd(}0-wYqHfEi7n@#$izv(R}Q&OA5R{p=uG@dCCYC2Gp;3nKYO~iI^AjM3p zr6nP>`{NnIHK^1w&>aWO5@-FqpDWHEp+bdTRwlp($mhc|nc@b@EKbo$54tPt9RyVa zB^p1Of_i2Z1xMaSRT+;J{_e|lP)@@U^krl~4Nf!+lHh!HR<@pSZaDR72s5izBYV$B z)?PqezU?9`iLU$dnm&vJX6MjE2wB@IYwbzh5kGar@$|?+EnkB-F-;B%jsF+>l-KiSP%HIJq~-i6aQJ6 zZF}rsLh{KEOh=yilK0st1QC|R_&)C^A6Z474&*T5t=ZQEp8#Jav`ucL+2Bl{upkSK zcKiG(j2lCzvolHbO9c|RzGmV2^Ys5Tw zu$KI`9J-?;)BJ;>k0UQW73+qnq;Zr`20_h$9yN2^)9`qt|9D1f=!UV5!L1kQq1oHT zENLr$QfmFq%h1&KFRPL_ygNU`)~2Q%w`_MZ@Y|5nJDFOL^flh!-mBO-bg^Q=+A9n zC=Eb{QyV4~X!>Nqe!d+aU`wA9lIHbT#LIywL?mgnxfPgUKyiK!&HX0wDx-ws6e*?f z$7tLKemFIuNK8`HBBQHuesn^je8vV46ph?)4$mSvCZWi>dX>ztsE7uUIt&8JTi8N) zp(?l#Kj*yp3a~x_)IV!z?1;e}5*-hx1<>F5;Ujt=5{sf}S7Ur7rmU;WL*h#4b;x<) zq>2iSgGg*nGHiVB3fTMcZX0YFF*PCr&s zvr&6hgO;W*P;z?r~Ic=yo;Zw`nIECxJj1cd7Ej%8??2u5^& z9LqrTY^apeyEC<&W!PwFsT2?28xp>=^@sBFPq@B?L&{m0Q5%uWD-I1jKv|rUPUr6^1bfj+>B%CVYfp<>g7{2g}4iJ{U(L=OkM*OHlU6PYFey zV#)mD&7DsPFcT<9DFk+cm&?waa3HxA&DvVgNX$=iP-s?ESNjc2qF|(rgy-nfkmuG{ zi#UI{az&!Rc};9=EPPbh=fzFDHdMFYI4{@KJfSlyHh{Ynb=jR~a^@Q2X#=POj0EzYf3 zxthb6s`ZVJvUYJ*sqU7Yn9LcUe0FL4nB}&vBNC74Dv~nGGa0ATcqE9_eC9a=1&Sr4 zY@wRfC)b61KnU%oSqSE|rWMnq1H_zq-eed}Az9?DhL7-BYFL1HX*^Za)PUEZB+@K4 z)jh&mJlEJ!8u8ND6X&I!&?qtp1=t1)rE%Hb{JB|m^XXPM=+(#CG{+7Ug1KnU4#T|yOfgd`nQ*!Bq*K#A4S5?o;sTFj{hy40V#81r z%OV6?3^XbBKz2qxpXUQJ$z2ZF%(|8!>JW5GWRi9cdt@t`nkEZ_$by4v3v_D0jlpf8 zx$AfXBaaQ;hhf?_gBqYNBxa|vx4z5w(-w3EQRHF+tQ`%#h$DgZU0Oj_MO!IU^NE&) zP_QRxgY0U7wPzkQ!O)})t^nS!6%?DDX`PcI^QG3Qma(=ay5m24H+-stJM#}@c}z$! z{2~MHNMY&Xe2l;}fx3oydoyz9BF|*<$E7}5udMLR@aPEU}nZ{>ji6d&C zBNC8~&g^qnHOO__2pY9;?a`#Lq}aTLK^3QcljR}wo!j{}7}eFo&&%JDdoL+!$w-lq zZ@@{f8$JVZ;J%0nnz1gF*u4x)XTeCXVi(1phUzQ4&TDD&L zZFRcAzTrL?s&M^y95o*9z;D`U_!78(b}x*uBRmuBn`S%J7xpdah5-vFF;z7-c#_+GS1axT0+j!C~>E! zIOg0(KS`E#X3fow50#XNrLYU#WtX_zTVHgS^~jCK@?LCeCu);z?p(O$w{hLZ1HN*s z^?CKm8Bs@n+<)@&fwbaNg}sepLsQ|p3|p1^WBy{H*7}Z37U|~c{bPeJ`MTx!AEcSq z%dE?wD1C@{--Z5>yLAdvTPIMRQcpksn!IJ5FtAXG(-8J4e-n$!(g>6_jcmKPLuIi$ z?j6Ny&c}JvgGX(~u_5{1k#d`OL8g~VOokINE(Js^mX5Ad)RQAJ{P z&>hKSd!=_a;;Cm|enSq!6M5yjZxG2rGvv#~=fW)9W4FWS(|5M16t!F56a|flfwJB2 z)~z!+oB8?Wp&PCM&yltxV4blX43GyL(&@<|G2Jch5Y)MS|9%u|KVxilGOmj7LQSYn zJ7HS2uMdsXr*g`;Y2gvpAf|=%kA0=JMvdrVMEH)`JWmV#mMWmx{s$te>m0^o!TXS=ltw1et%&q|NgbV(Y`kNjz{vH zkl^;7Z7hp#4)=1#N&3aP8U*LXe=Q(k+wF&?Ygnttr)OZ`>@tU2X^3id5|KCp`P9k9 z-G0-iO-5iab~mKP(o*7%52vPF#_?`qfKROvJb!hER>r%E+&;T(IHNy$_H1vvFm@1E zBhyPYv>yR`@*$4NY9@RqqVC^cy@caA2b9JUbj;;t7VZ(l`@2Egpibe~pKFir1k1c^Q*d)A0(jjpf zx8%o)T!h|e7_XnC1&{;fKCR3-J|Q7xlmNA8)8jzvQ5RpL zk02R^c%j73&hFT94u__@D$)m=VEdozR|2yh#?SetqiD)Nk=DeZ0=tJZGnjM8GEX}D zF7}_vhe)c!XoX0?lC#o#aod(X7lh_A$&ouK8*mZc@iSJ=IO zu`pQ5xt13#xb}@kEo~ISo>s62z5Po5$(e@6dxg(zMv=+y1O#-~w(Pi9Jf4*o)w0}? z??^>Og>iFMDvFe^kUkY*Tj>j|CmyUdLv*JN`R@6%BMI(CC~VC8`wk;bik98wReJGa zGYbo=saxKcT;jhDD(^JstVpi4+Rb9p6je{Nb;!`RmukVK}`F zgC=QOE-R!iZokm+!LT@?zQMk>t7qG1Uv$4zbgH6D>qr3|`Y}$OSxZj+pt@QL_8Qvu z)+ynsAg~N}_tJ^M`Dw_qjXgA$3wqk_^RItMuf@)y0UY7 z+F|!J+?eA5gz`#BHOHn5=)w`@3JPE5tz3<#h7Un-cRDX{coJI!un}k%mWI_mbSG4<R>sWuFyin&#xZ*FwV4$RM?p1qP>7^3O{Exc1hzV zmYE3M_Vpbu5h^!CAEXmFk&QPV1LFmq8OOO^thwuM&Cf?@d|x5(bALs+2OLbyg6uQ3 z)?gVE){RTmzPE?V4P|Mq9Vymwd-q=4tQhhH1YjTgU+l4)!Voj(0BP5(L;VcLK0zEs zYJIK5+RqJ%t@kTZC_ zyx*b)v#k{XW%p}fp?X5>^ea$<6-PcbYvMRa_W|bk9*Oz2i4-_2aj;O<8Ii{NWfzKz z_>?+bLXLyyOG9N^x2ME!M2JP7@fICmXH6#VkP50-OuU&jnR#~jOK_W9HF6x16w5DD z6#Yq2@*ytcc{`iFxiXMsxE?=7Qc{vE^|ag8 zu6@=9CJd|A|0jBhuirmCypMVM&UeWZS=W;+=K3mcmamp)m^de?Hl z>1=geu(Pe`O-0s*`mCVmV&PHZPQy_+FX31|l2Bgu=1m1U-rd-SmQcze<@A)cIe{O4 z$3Jt#AMYNJXB_8pDS<{)kz9aYi%A){RI^`_N zVZspgEqG(D+U?6h5{m>bts%Z)*DfW@zTTn+;|o3qr}@L!@8gNursIwywvDPG9nsQ~ zgp#QgcpaaDCrE^%zH=20LUpdQJ(rYIHA5wSn}~=iQuuKFG`e$R{1DvE3c{2L$=23= zqLo?VAC7$DJQF~??}nsv9%=?he2Y6OQGr5!pqfyA_Yt=SB=nACbYI zI5rz5#LD@Oerk?|Wh~C*8-~jDrT2;-qv3PxwGC=EXCS3_-O9yMVPWk5rAp>}K6bgK zn$}mw3IF@-U{twv*RF~!L(6e1;1MNUUSXm28qo{i5~d3Xks2-(c5 zS6xTBg8$+Id`snPn~2&fE-temDV#AqsVAqMzTX%PT$)WJGYzA_g^Lz(=rvFHH$Q#y&lx_zE#DJM5i=BI) zjeMXP=>f$l(dIWiYd{fENYtmqw+H~(h@pxDseQu~wgW3rD9|*ax&eb?$Mbm`q0yeBebuJFAI3v4p^>rB;VRRM%`z4<5$9_HR^HD|(oV#g?c?W+Hz_ww~ z(KW=GsLEAwfbRJEp6v>1!Y&Xsl+PnZi?(C*qXCrV5rD=IJ=-!V@@O*M75ydbvHu%a2$5qbHM?k3QTjX-kM37#;(CW!9F^nM=i zgG#I%9I?P%KcZL3iyvG>rzym3(g=rQD(gs6;)VA=Q}<+=4|>u;ENoK)>t>;@sVY*-x8x8Tzi-&+O1|${x5gJPVf;} zU$55Nuzr2@!~Kg=jjCk9V}YxWA7)T~I7(#!df}@Bh)d*G{{h|4V6Vy;*J#o7WwuMt9l#$7I zDZSEha?@u}l=7`4LRmKYH-Ey+vF1 z0B&k0Rtf0XIM>45zb4PsnQnGps@iRT|7chhGrM*P90v766%-~L;9*g4G2W(|m-F+_ zKVz|3xSt>E%v9@KExrr6%Nb#H33#f0np4?YJ773j^c^R5EtdQv&TW;J zZs|+4%hrYs;%(A5)I3pFM)`UYVAjeXe++y3_W0owXr@qF`Jr7#k!J+rviOB}7u}`@ z4q;{5hv*^HOW@+zzj^<;te@2#hW*Gocl}uJdzd+E6Pw$9? zeb&mco)nWzu;_GLM^q34*b|~rc%pV43uQdRTR^$`WSG5Jt$>TD>w%O49&nHgX`wwm?s)WcGJW#ArOEh)S)>HC!2_&PhmFK)jR;qu?M%D74@wHNpQKVSF9 z?_McH2py78lyKEvuyDBn*qwvpPG$p02?>vPt77+-8hQdN2@D#+uOCK_<={9Ph7tdY;5bsX>P^AAk_f zqVRO!I9#kevcwC3DeVL>kGtmWT?^J)97_MF1J{0Jlpt)Y;-%f$C~V$z3cJvKAEC7~ zLfnAD^)aR2+G*5@Mo=M9i`uMlPi=0zQ09Y;&Qz+ogF^;&`ceSWV>Hu038xT5fKg#A zLQh5wDTuFMyHZQBl}`Pg|(pBc3P6dmOwLA~B+CZceY3ntp!T-j6(Z)o9 z&-kR|?}ddN9l-ke%&pAuUQvL^_UC6g^9%}CoqN|cc5~>()63a~4-%V}KW=SN+u5ag z(&BR1^r@0kjK%KUzJ1OQ@MttX+q!8m60dhcfr4YkO1koUh zFs~>%@GGQcQ9w?M;}pL8?mH*U=bJQc2Lc4i#r(-UE?PA%pap;^c=(Ed{!^x>o2Hy9 zLAzJLz`O!+zXpT$^jhyV!bi3!1YIyGt@-uJl`ga<7h>n|V~#*x)9KiW5pyIiB|DQ` z4e9MThAe88aGiV#1&mO*shr$ZP!_oeaK=DiDWtQ$k2vJouyVhGf%DqHQvIA!+yfPW zH6fj|^(Y5N+~>yJ&Et#!Ur{Fu?hHPuhIw3Sy(z_@S!?r{Q%PY14FF8&c&VwX-qjRJ z(8G|Kuz5`rbGRXz)g|sjgL@n|)03fs^9F>1K|d1EYx`I>W8{)b{0v@0*v(G;JDRa@ zD3^5!V;<`L_|XtOt`;hbN&tk+(Mg>44#%`?#L4(c1Dppm36<#8Xv=s_oTeya1}P^J zNvj7;<`cM#me|F~hf6dR;7Xa7Ex2o@>xbh7XyexFaalX~fuzOF|W8 zoVhLiVL$=R8YzxL`e25CzCP#0BhzB&g9xo{>mxI#2BCiXgO86NXRP> z%`*y8R8(X{DTNR2oiQ7@>;m-^C0oLM*D&uxE_JE-m52$V1?%mZ6 zrK#~^5Mr*2)5<&nG|?*i1|BsQIz7UDDsF9UbqJ1(j~;ACK(EA5`w22XJ_Z7U5k8o4 zQ$~_zkB<9{tyu8E_S~CjMGW!z%;Qj(I6pP6+17mF<+owl-KjH#M#>O> zb|T=$c1XL>2WNGM0-b5LD&}7!Y92&@3yq~vms9b_`cjSnmlh6Li zzFb9JA=lPXlsIgn15ePlglp;OFdYNBc4Z)92q(OseZt%F&8H@e0oY7&S_ml~XkGWK zhlg{oFccnF{6iDIv^4wc8eA7aW1%D}QJonv;qmR1hE>FHb%RZ?g2>!WOa}h`&)ycU z?A{X*RlP?KfAvm7Fj!37&cF4$twkuZzy>ly*;~3{3gyaC)G@XurKqc$Cohf1`eQOl z%}%=uC|8t-ul)8~r}Jn(7>2ts8iiEcB)Q&H)`Xe@ARbOnUSbd%iP+7-gi*GpE0{y( zrE4U`M#EuRT2c1-9;3>y*NZi$ER_7x}dYzRb~N^eyb3M zVR3}>JbC&4E!YWhc(@#)3yvYP89_izm6aC*x^_)`UFlda#3((;NCgwwyby(oYTWNh&Cprn6U`lX!sD17!Kl3bbyFi zK!Pg{@FSeg9j&f($dTQ)?%cUkcZ&WLa4o?y>iE!bP6YV#V9~SRJciGl!rL}M3NQva zW*`7E*#d5`P{JtBcJ!Pa0K*`$ej9rvJ5`NylUl-wn_0_~&i7)bG4_fGx z=oP#eH2y|6IuWvzbzxUm7v>bV(4#*>CzzR=hu}rixP_(HeQ?CvpKNKBDTze^7_NmO z%qw}x;^954zn(kY@Vvun>!wYwczbNby?oU7Wh{M?!83S@ua3~E+SY9JLi_Q9CPL1^ za^>>>=ip0YT5Hk59i_T*#uc*#e?LyC)Xm!0*j*m}T}X>NaH{X#zh?VHSjyiAov^7l z6~RGM0CVs2$2^qnG{&@?#iB*r&v94=3K2FXu9kfdJHE%9vN}D)Zry(ZM+sF7x3OJAq<&cUk!9>pniB;-3J8 zRN^t3s?(+y4csxIB4mPefD}bTPf#3ugm_c1XmVo02veghFpCBy`!%L#W=6mf=mPgb zu*g79&j`>nPTAeN5zD2=pY5T=;3m+OP6Rk^Leryk?3ica7{467`*~m#s0Y9xQEAkIgvbIsERY#^+`DJXhh21kpr+P9KH=Jv^9}37#Om8Nhv+$R z9`WNnV|g21|1vwJmSHGU^k&zT#QhgG9FHD7THXKVT++J+R5d8#^n!e@UOkS%2cPM` zP}scx-&l0py0@(n_4bRR&HY~t#eVYGky&lAKR_^f;>Z7^I`|`SCte2_eeRw%O~k$x zv#Z|}d82jF*6TFNKzUy7n~PT*i;A#x52h1*?E-H3zL>wWyFVXCXnBH)^n>HwXyjlm zj46GTM)4pu@?0_t^`Z7s10goZQd&$*9e`W1k{#iL$>*;PwBRtF#7VN`UZ@7-M(Gnw z<^4iKZS$uL(;I8@-QB}YQNE*Y9FioSfk+uWL^PBJAqM`RhlZ%2SP3NGPdD8>*qc+# zm+45%0EA_)WZ<6_R4GYr7X<@C70@?-V=DuNfu7%zZ@*0(wwq0ncAZ>ppa|6W%blAsz`sF$8jc;08bunGpS;iyj@HTu%9@{(W#>HP~1O1~T z2b9%wkoXb^$NXJw{D50`mq8(4U1|L@Xw||l;}_beQKRzpT$*zNGffyIbxCatq&AJW zAM+u-2*XsP!zj)u9gXT;HM(D*mh#nS6BLW4#&ZZ{vZs$0y)(g9Sit5g03{VdrQ=Wq zm;i~H1ZqY6o@j&vB$GQ?_WkjAHCs`6m!}Y6fl9l}eK#T@HETD$_Twf#)3XJ`nS#Q? zHGdqu#V>Y!``vfD$^y7O!klDmfM+TyiWafeG(Jd_Vw>v)`IM*?*aqOmAV%ai_2fra z;($s^E-sJ!{T~3-AIE8NuJ11*u5ZQU?z}JsW@XAWx~EeYv`$SrFX-=u0{-oSZRBlh zzxN+nO&5F)`rg2MHn_y`ag-6tH=+q0H6VX{8jTfki&-7Ww(BdZ3PP+v&G?p=gN-c$ zNcL;}p|bw-z@3iVx^+voo!Bx2dpMU7IeamWn(1!|T3I%4#>Eox_r+*sCcpBUjYSp% zwk=Ym1LV6+#GL?aVw-Tcjv@XQRb_z&TWp^0oZ)2y?k2kduyq5U^8CAiS3nb~_YIx0 zCoU6&Or}AixjXo-62gk{28SNkGLOy2?w{^56F3Y$qq2?7QzbiGT)fcqr;3*LaQn%& zmHH$PN_jSJ_PMU_vD(8I2R#31|Spx zyTk>n#_h}nT~GtO8LGt)qFm-#xobQeN0d-Mw{d z-MwNxKmKEddj(~d9UO7vwd7!h8A_<$`SXUzEtI`A)QZj^K~$pl&z=H8xE3V!0lBlR;+M>wS-qPn7&mQd^3~k z@urE!fd%#h0o3uWTlYT3UE1n{96;*|!-R_$S?DFjP4L#CY4Wl{#%kM~-<%JAfNPD+ z*wBzNXnECP2bUZ?aNvVc!C!|5MImC}SYAU0nu+eF%oHU%yg}S{TW49k5yZE%e%PCL z8{OW=(eLf4M#CT^vBNs24^5krf_FUuBDs9#Lx=FQW#P(rC4-HYmQkt;j!(I->volx zN2ukpQX}A2M_BtN*P?p@36NZ@(JK!UJ9pS47V{aOn3K_GCxQ^jn;bAWaJFgt{kt0+ z>f~=!P#%sTcpTU$pZRd(P!ojMMt;OqKuvPaGQ7rCyOz=n6x4u6e*c9yaI%G7m7;MdaE ze?OrWIXrCB{-mZ~&pGRImnCn^*#2vc4|XCZsVOKBd>da^cdA_&5XBnc!yODn7S_!z z&T9YF#gyz;6J)o_l=iVcS=;Kae>=&s)xhXe-iwJ}{*MUYkJEMh=Hv4n<5^!{Tz}R2 zXoVF`1Q4Nh{72B{Mji7Pl&wcqN1$yVJL0cU5#HoSK)u$^bJJO<-dR_yIOC^{?03jO z4NH`ffJDl4GcN1{k0z%rlwT+36dF6>785)cV<3T++y@JReCG&8;>E;(IfEuKmNF?w z=V2|bW7?>r6>y2G(0rz1^ojX;b6+mFDVTh04)5hpZwf;NlqIzxA9>}B-_gT|E0Bnq zTioFtP*9K3Ju}x`ho&ufIXZ#NrL8VQUP?PKvcWq|!SV3d5yJI7)%3v!-N*T^p<`&Y zn~Lml3Bxw$N8|G-f!}X^cXV7AmqEvQT7q+#6o}0Q6UZx?bp_%OV+a;aVF^; zW+3c(nA7N42DZiGCt5fPd4C&%PoH*2Tc6ZCa`~LfrfO$QW!L>@ySbBUn zk!3+><+V!iaCs9vwr*x-hTrt;=C#jyUp^ijoQ+AysHz!eru#~aEvO2O1D9=+`Q;bG zr>1ogT5AC8Hfsz)njC`?i4ESp4=xN2{e7l#_x@gqzqkPZ$wGSvZ1y|4%ja*9n-$R& z-8D9uFPPfNd@=FxpXcKEU8XsPaL3Dja+$ z&O+~6;K;fR6^3jkW++-<0wiX55trQawg+WpVK@0IbUJ1*=QOdTV_GA>Pc zFT_K46Vx4RqLemZWKhSOjkA(IhVPnd>4VlX14n|oI5dlNBv2n!5mt-5(*s#ZHL9(6 zG}YVp>=|D!raRL_5O*d>iZ=qME3n+>OxG-;fM6L=9s-%z(ry8LH1#{Gppf>oaE>0x ztd&e^CFJ=DKW;j*S}%XRZOM&BhZfL%V!*;&qnC|&3Id;I)5m-iPc|5#UZY=-ssO%oGwZ3>HUTzFf*i5T0@ePdyv-&WNZ~?!ml^*IE2IAYs!)VZy@xvfwELBO+ znUR^ZMiOom^_K%soi+1~U`r$C|KP4Rr0!YUl(SA{|%y$-nBI(p&u;}_FARY5_ zOaPNE+K=S`r4*>UySKWF_)2hLWdG1Xv7p0>-2gZapK05~vIrSXrM#ubc+N?(E%}W= z4shRW=P5?Rxa^wM-?ChM5qClB9|M^`kHqyS9zFFVE=xQTOjZ}$9G|!46k%w7Q^&CF3eC%sCCUhyO%07y2t>`_ zF36-l4NTyQI>8iz;O0S@x#<+}`_~>OqCfg4F9V?MI(UA?%8IO}Cm7&bs;g%ys6+aoG|tdFfX3k_v^H~Vj^o6l-rlq=$R5y39%pE zo|9|5zyS>VP_8(=`(@F|RAZG;o1M9F!3zDkt09pRFV#J#O!LN8&JOdOPjEo<@GMnv&smWV3PlTZ2{L0fX z5CF;ukb)kZTGjb85WNRe5N`-DfX1s)Z8b49r73)SaynFAK6S?o+AAe zcD=ZF{rw3rjG&$|Qy(TfCWk#xQuNRkW(=YY{X!9%x)})xIOy!<4!}&%mrUiT9{5Tv zUQ0|R`ouxNDm#l3z$qgN5X=f=Fe`dpo^WA$7VbHV?%ce2+RqGK+bgs9IQaNcp9nC^ zy0oC6AO;Po?9`|FltSKB?CjBa2^@M-Z!loT=O$E4Ip&jxFAUTuMN+1`vo9gRFT#>){;!SzEB0-hF%(S9(_OH^P zo$Yjg*jdVu2>JM~ti1Cm=K}Ku_Ew+D(vK)PlZ)6s02tOL@`i>{z!Bphg;TEvIr)W; z+Lzn0|2hYuDfOmntdBkrKd@?g`OAssH;*_+H15S$^Od_L3I6p(fBniT`fyob!gdzk zOF#DpXEyGjRYH)Uw;2oDnFX#{N~%|#YeEz2}& zq(vs=L3i#DYO27HIJCs;>eZdnW3#X^s=Yh#02jCi=#Z>>RgW<%e?-D2m**VKR2QN- z(a4$$z4Gg?s|`|8Qus~3X60Vyu^+?oilYL#k1(V8h>C(}qFd*0>~=N`6D(`HY&6hg zN-!1N=UxnCDu3d<)NKqC_HI0j3i!&P zd6H{(>WRyOl&l6aMlg9Q!_pjwFa}^Z071|2%uhln4=O;swkq|Xg}mx0LB~bp?8a+w z1xaz-)XwrVeMex!mDFcIWueQMxPiAhk@bS-00ytpFUePyN|BI;0GQFeDyvmvD z^2pnwlU&(B?LFeMB6Ho<_Y(w%yZL{)()LpC%e?;J!FQ=c7D-<@aHFhyh|UUk0Jc@& z3}d@GKcg3DfY&Id+Ybh&DS$muiBXL+4w?bSiPQ{7oj^vNS0SNnl}clR1EvDp!KcsT zQYO&)208*nF5N|e z2A8o^$|v{N0QG8rQrwZ8_}%fHD7)kc_Qj>{?AbOx_n)hp!Uag3-@Hxi;K*vZ&ccUO zmovUOI3mIu>B@635B?Q@D@r=h!)4FhTb;#%6+l%cW<4EC5Vnxuwi81awr@A^$Upv4@X3v`b;vgY`oNwaE{xjC_qTxaq$7QY| zE-NskkI2O0Kv%4+78mxfor{MhMqd2qMavEiOa^~M+MTj-!>zLvk(QQb4C(AU?eu9W zx^NUxeCC%gMS$+M`1|jNZN0V=okXpZ6%@*xZORdzo2GQUO7&-3l+I+2?tQNoePR>I zvHENkB9-wU%;x}9)l3XANde*tlyMS$c$1U%G?Gg+6eQH^zuhI^1c^yW9Ii$MH zIJM^_kPxqOhWHegZ3mjs#DNC!W=NW5aWD;%s{8)RzD%o*-FCgFNQTqyyk+y|ZU{XX zm_F?|9HX8RL7)$KkzHu0PVT)pqRZ1;mf3;FmzsEuiEpk7PBNq9e#qrbLql1J<!sn z3UY31Z>K70Nuxsn{-~!FF$_x!Oj6nk7(SUM1FC<_BXExI_AX)F0yfm(*w{V9Z5(42 z#64}qdm6*SJ76V_gF5_Vt1!9r9BYCWkCdCFF2ayvtf)szOC9BsrHD*_sN;1b4*=OW z0v)_zS40f;kdP3Qw4GcF-c0Yw7;@3RZ@vJ;FBgu_r-r@JqC^9<5C>&Uvj^;L61n1v z7KP~p>yyu^{2#@IXw1X==Hm&XH2gz8>@PL zLqS*VtK$bwZtvRSE)$qwvNrVVufINt#yZrF7q>uZ2ZnXLtI&vtiZ2AGiOCnLgsBmA z*t8d+YVr^4mssK4=D)pohj~^tYA!kwN12=$UL5k~L`gJs5^US~e2vb>1~jd08#n&m zJ_G_8^wjSMT$e1qZCy-btg)eyq3a3EsS`(zT&Kv5XW4_9ZY5N#Hn?c8qnOAFP(#?& zRpYG%>A0tA0@GHJvMg(bp~uJ`939t(9252kxpH+ZY57<%qTT|6UGeBQpBvBhBJmOj z@+ss!$i0|PNP(@77{wrVxa*`1hW2G`X&Js(uYTKZ3W47hdCA6gpyXg=v~tCYhd>lQ z1Z)SC?>jK1ir%xp8QaQN8^oP$Shr_Sor5CxjoK-+$inDJ+Vh%>Ub7%&=v{kB!|qFo5cHH42N^VHh`!F>Q&YO2}DWn zc=7Tjzd4-3u;PTVGQMcO5|#iAQ{Tce3xbm_NCT)1?yT8u_^!4Acu5!d(bc$Qe8wIo zmdvb5&Hft4u*(uMYWq$%q=O2F1R5TMdo_^ij1nAhh-dP`_aZy{-jz9vNV3JH+H{=& zJB2YmJC%(=7dTIHc>76FsAUfZ{%-=r1^(qVsXY^HSD5?V1PtU9p?%h{#A2pq|9-hEudE70FbpTE>lAiFuF zReXYtjSUAv#}X1_Lv?ZzY{X^_s@3Tg6S#7)^nx%hIl`f6 z(dI2%G@CZt)8Huc^XChQYpP&CD#h>Qx{ORcG2uU#d%er-w0P?gRbpLfyzH{+21A)a zJq==adM$9C9<*ElTvGXr7olbyo2$$2{~T469cR$`d8+F35o} z4g{=U-1c?s+DkV3emJ%7hnwp!g>8@GeHW9s{@Y5{t(E(Ze0=rmuN!|IKHnmf)1;zS zV!hc~|Hr`@tyS*4V?)#C`kf^~VmH}m?=1cDLjCdYYg;)kXkN^!v75~LEy-F<+sW8A z;DjJ}0solvF8ujs)x){%2BWW%SUwx9xih48v+8}%n$H&z-gVWEL+YWl10u|MP9pqt z1~s;}4t9pd;V@*O3*(OybDPe9IP8q_2G-pesMMlX+zs;MJhtTK;*!(i)We@@(W-;# z0T{r^jm{{-oDWgmUrh6nglb`Hn9k%o9N^a_v|URyw= zCIf;VFCT*1xJIwNz1__Gb~bD_S9cIc!}_Ok6mIvG%7)7=wq$>TBJ7>4E2kBolMX)Q z@twkiF6;Uab_PrP_ zT3;&d3ngjymiFztPR3fk@9Q6x-uHc;=f3Z=T<1F1>5QQEJhiG@+ouw(5M1_$w{g7A zp8x*wy3B^pVp>kSZ*cGL?qlgg8m902B8oh&HFQnltY$Bf=wDsZ{LcJWmT89$o!w5o@0|@(xujLv}nc&^TJ2hmxA!J-fAhe8qavFzbP&RB+^~pI3H*ReWY7#*`7Eoww+19@2g#JP{RtS#@ibh)+f$FXcC)v zQD^?$p=E?a==6xa6yQD-zuiH60rb#9N$mlQ?5LDDWJyFW*?WDR?oK|?>lcQIejUy${!kE87tvZyQD)^uq>Y*lgy=-t`zK?&R+4-DE4#1}|O4{1mYOSv;T>H1a z1W*f7%1#1q;Gt6l%7CGPPt}43bZY6K+z!gH`-!E^IhR`mc_2L0d!GaD2(f3 zvu~E>v|WVEV)1t89(k3Mmk=RvAKiPl|2}@DUUqsqyvtzH-^S-*g#%QM9zB}YZ#rKS zZCPcAk*d|kNOuIgiolHN%C9648<#*N5wU>+0Y<3C-3CzJ3QVLFd5wX;KROGEDhBKX z$=+tQh4bcVYJ~nxArZGHWN=Kc{{8&%Rpj7A1jnFyYvPueVfsNCgO9W z#LC9D!+niH@KFzF&PEkTiwKG@U;b~_NyUgZro{M4DhRAD{0^e4L(tZ^u%b zrnK3jG0PQ=@2^Lm%{x#v^BSvVu!h3vNn*6LH=RxUn@-B`2z%U*dGI&WC~Hc;+J)|> zrdf{S=OaQNo#Ex>MNZP6SG04-4i^(@e?aR)I2yNq2Jn}2g7NQbP)CaqiCqa2CCv_y zsq7)V1oWYAKVL##i)W3kb8#~|unJ00yP)n(%7ln=jaCDcAf#~Larp&k#3#Jv-ro5y z!d0tINP`y497Up3gd%!%m?x33mNXU>7en0akDy{p^=2PVwJH&!{St#$B+eh7fh?J% zlE#LCiX1=|LWjq#VEZdH38@@?v(O{k1_-F!dJsm75qzm3TW5Jg+f(dVAdR5%Ax9xC)K0W~l)jF{6FD|H1$bi}z6IsxMR?HUk zXi6Lo?c0v)a4|t^B93Zrg?Txe2o~#v)&8{9*&0##(gHuxs;(rEVPOfS@c^Q4qo4S! zcjsvD1kT@or%A9zr9{?w>yM#BLHGCObeXg9F(PD@ZtfT7U)|XW!J%)yE=SD&#U!|a zM)uk&yfSl^*QO3g=J=eZZLw3G#(}1)m#s6P%+HsL@^(J=LX2sP(Ltr@xx?8U z#94B#G)SdvUd=X z$qRsRpa-tlu*vu%7eEdhK>_F%^*e*QV%=d8AXfS+4K8lLEUkIVJuZK--G}GkTzMp$ zG&&VXFOvi_IW7FO`YW=beIyYAWR?_Y?wJ#n6xO)}$hWuGc6USyCp@I!3UCw^?)>>b zZ?=(0tkR-JN-WX&;9X^P{aPAYX%1q080eeYgWP*1?_K9%gd6(0M#b5;xhS2-1xLoD z2l{G_*EAFDmu)`-@`2Tf*~!f$x^2I)ar-)lEl#TQ~MY zm+Jh*TZ)RuWE9_DTRN$2JuW96n^I9n!8{9(QLob3J9hjHuH4X?z=w;q&xRP0M>2SB zt|kDmH|1iFE@_7>qdppDcQiLUC36@1h~4D-pRC3S7hAL8QCWslJTo>Gz<53tfDI*`$7!R)b{RWBt*@_YD8<8kh!=->i#e5wR=rsTw z)A|q<#-hX(nJ78h(B8fJ{!u^qnx{hIQ};i^pjVQ0TO)VppL^!0P?GK%gOO@zM(CY! zFHk4}0cL(14smX?S!H!rG}1ZOh+Xsk8QSJf*spFUOfXxW5mGRqT5_0d#?xL9c|sLu z0sU=9?P|KqsD%r6N|j#njx6pO=}F(Zx&IE}yhCAYw)E@`U8Y{P@%oMRqRd?($K_3M z;=VbcEM2LCS|cx#+~pl+!N#&YJwLp3R&4ldg^?&y4DiN6WM*Xa$DC>QJgXFtqu7&^o?Q=czJHd@c9WG=J7Ys45v4&D%Gv5#bBa4Lyyyj7{x}IXNT1aPh8sy#bq0&9})U zaYl`SPo1&DPcJ3*$LSv?G^2|e%&F7#oJ4;n@sq~``G9}`1oC|qFx^p!7ey(R3{csg zz={RieNG@N)bDCew;0O}h{t&ug*D;;05 z^iyL-{NGu#r_+;4Wv9>lv&%tm>`MrU8Ws2Xf*|NG&^-^_K1Aw0vn)$Jd-(${=mgLMw-M=9B*1g6@@IQ zTrZwBU~P}&7c*v_e=;mWe6_5VW3-%jYQF-RJOELT9){znJi1~^AULD|OfPT92Vf!~ zTmD=xvH0p>##;O#uY-@jh;`8$fih3D3IZe=(oYNG0tFuDCu@LWcr=hv0tRxizJ9Md zPFU7qPiOx=o|z;pS4^#K<8AY(lvGt`QCyq!GLrM=V;=L{V zUWY&Q_n(rmDF$qM;`cY2)7}sp?bl#ivVCyC=Pqzki3;=%lvl6pjMjS*qGdYQ7A3`O z?xwt=w}|65!!-+fQpA+Y2tb>kS565z z-bK2#*c>ZjS`}~!GL_s=5 z@BYs1vzIP?fXeh@Eof%$_O4Q?Lz_8(){f`_zbK^0XqjF_WMPthL%FZ2x+R{v$X{4L z9YZG4)KiwAE)48SE*!aqfFRQc{3Voxt!2_3$7m#x6pD(^Fbu(rS-%Ul_=H%X!W(mq z#im_Ez$17Gt=?;Sd5gM|h-nL6Pt5H(T=$8}k_dRG0z^Y{Y4FVw6;Qi_+2jc9dra6h z1~YqC4R0zQs;e1#^j@GG6e#7PCPgHqOn!@Gqci{7ty{^(^%jJM_+g<_TYaf(nc8gg z$`q!soVK;foL~t*tmZr`)RE^N`SHH=ACBruo8meY;#TSxK(1nv{Q7AhqL`A@IQcLm zi#3#Hv$KfJZZQk)DUdObXjh0_<~M!DFY5Mk30!FOtKwdWPJaM&i?yUp-AP8Gq0N=h z9lAaW!Swnp&ukG<|KHC)KSRMb zj8Y{q5X7fjzkT+5z)}X!0YpKrCg8nCCFbOw4kZ9;%YEjj`ekk zhD8(**zRz}*EbA_&X8s=_7v&W1?a<==^Hj|8E+JV-a=W{N=HLLYD5{LCrCa02&fK# z7iETy&v9dix&_fqLAv27OWmjbq`0`a`$(1hzb0ae&rXrGMMM-%eVDHvT8Q?7E=bGK zwt5=18*p1fC@jL?d(WFDN6?X};$C3TCWr2?fQIFu*{mv?N=>|We&rw(#B1LWo7Dgr z4#dCl+OIP~hNbpckJupr?g!a`1MWa+`hR&%$<%KFMiVy88>S4-?_mCz#DpsRhb2F4 zm*3p<>_PA25T&@6!`vgb(|Kc07zIY?eS6-w{q+(V%E-@-J>d&)XPyWCGlC*$~k#fi3*+RZD4f;U#U8Nn?g4PFe&AZAJ zad(seURV?90j+Zyp!eG&LePTh*I+er75RW_WUpp#@)zTvM1m4Cpc%dj{}lAcRsb!A zO^iu=K9-xGUjk{T6F!WboXE&uyRKriN55tV4y+5O4(d-26J47;xaD>Lik3L6UeoJt z1}gH{fFq(>(!qh`;2m(w_ftIwp^Zrmb=sbI1CqzOzWiU-GOXWm6f#(<_mB(&HcoLD zB{ePRQGa-Q4z+%n8M-4`25-;y_H!eu6LKpB$*_O0{Tu+tHan!&MQJ$~7>s&0+s~do z`(^LYWo(~NNOMb64BN4JKK4@0_yKy?zHi!N9gZ=kTv+W)E z(`uP|+W6Gm)4pw1;+<6T+Fe1L$I&O|wEy>=tu?T?-^f4t!qO+Gp- zY{HWGOP6KvUSn~^)o=DU9(R~wV&WS6kh~j=-aU9bTXRG7?7(eld60~xYabO|Q6Gp! z?xqq=0&Rdh39?E_52CyV(n_}>IOW(K3PJ#hh5$w5F};c&Y-}GG+(nISAO!#;AAK_h zzO^n?axnxd2;$aY9ut^}BXp$0yYBD+?+{wP*}R8KsTk<)UVn|408W=KjWQxY-{b}h z92JH^wCbU6v*j9x`?#bfvp!S8u{R^*&j8U@XcvH&f)UKh-;{& z_y~@4-+EbvMr!mdb@U`6;o{Fle>kaj2*fiVkN^SN|8n@_K|%xwbss_rYA1S>iI0>^ z1++wM)C666$8n89dFYL_y|CW+WMJ`WwgSX6F_66=A;+R5ydSB>u2D-#kS>@JbUf1u zymp&Gh&`GVVDOL}>LBlSg!y-awBPDB1jw zFiU*Pyd}%d{vT2Wq0=rof| zvYddB&>^(MyFX#3GZtf9fhQ@`pt-;3L!2eJQuV!3CnL72^ zTOG7OL>qLX<%4vr5K$7%fkZdFMZvg1Dy*;US@#PoYZ;N(foHw;l9rJXsO00~`V&#r zbCBvy3P!3 zEn4(8_9pZk2xXUx15XcwIinM==s-(HWqfygLiiO5MiK}Ox^{$)QPJ>u>)%`f-Qriv zR!ENHDArECO$w(tfted_rbbsG<2yEe9R|9pXblRK@eRT0u9AN+o0K9PAYG$kN)a>E zC3%4DZHRG4i`GC;*)fMOU_t`%kxJOxfD49IOet#sbyfjY>>(*hKyO~ey$;aFA@1HB z3Uf>1Kc)rt)ITgEbI{7l%J0q1t;GMPex#m^aU=rqEdN&4(O5dJVD|xx%oM4_?!3{k zYV-FtxlhO+!<)zaW{dV74&gPit66EexT2_7??4Z8>g|vt{|a*+nn02k{0)%dntA5I zJLJ?cKG3QXhLu<$)Ou2&5~qJtE*uA&`h4%DV)pO-GBw)*?ja1*ArWe9RSyEkL6u4E z=!DMOf2QKsy-GKvl7QPP~fQ%(BN3ZnynC#eyqm_)hGFe zhCo~)Olt~QgT+EXvj|zr$r2dLdq1#DL_~#J#rTzCf-80fp|Ktn>eZ;`bNeqO;sP=! zegCO_ScMq+(r~IV3dlG_fuq%gY_t4>Lz@sV6eYS9p z^`AFRyR|XZkxHSaSN*}fBZhWdJ10^L#D0sl${rlnP?KVtmUJ2N=jyu_)N+7~lXAg)a`q&$;B=68yoclD+9r(riUsNP(T_mFnr&g7~HI9 z3-O2OQwmz?{6vyAN z;3nTHMjvwcr^3;p3%(OM&fC_vYuek)z;VoGpSzTO+1=Ir#iOj_V_d&^%?u}W$JVS> zzup1iFr;_m+96X}MP8QrAi-Ir17tTGjZ^V&67&%7k^GDCC^{`dcfWxJJ%BzRi3(uS zq`B4Aoi`GC7)WRY*kff0*;q{Q1XniLZ6{PKpc?r_=e6uA{3kVXE9-y?mCtSY?&WqG z`&GI=-4b%}0V*Jwx%9a}9sq0a5``!2wMpX~v(~Q*yt22jCt22^3?A%MMQmNyD4HTn zYW-3}$$GE;Jxg(vl6D2p#%IGi78)qS>UX7wtEIM-Hf%-t@0`HX2soGNnD4Blb`!e1 zyx^S`aaRH9ZYXOR&U$ZG4rN!UH3-BHL_h^$@3>|Y3Te-|bsELU+!>dW=;9!Z0T~k7 zB7D|kYb0S=1q7#5E1{f%?NQp51oL6lqjKw^6fBf;*>6WTx+eD3cggC9D1y}B7gh|O; zTpCe-_VEwC_;=oquoNYzN^1`JAYp3(xn81zkWIiMVOua+M}7XHMdRH4QdAf_N|##R zg?$5=E^7eHVd&d$VwmCxlgbLjpCJW908A6Sbttv`G{GF-Cq6<`cQV$Lt&t2pj;3-l z7vK@t%t-TdTqK+HSDjED z+ON^+k=eC$ZTw?E>wt_5tGf%_ z3^2f&1?BKWs|(y033p@-AW5k#&l)GlqUI*o3r+lZYGCkYlOzBPd&xBl{4p8|B3GK* zT7;QSx<2M*B)@~Tt!i3kBqaVM0yNqHn&qo+_`_>y_-yK!&ugVwCtwIxJPUqD>}l7C zO}G(Kjfpd_h&S?wn9?y1LTjta6uYUtfGFt_6<`HPAtHzklfqL$p-|g{2Qt1LseNj0 z`3BHW035900cm4G!$o-e+#pZ8odhQ3;*h{mxp?;*lXZKk1st&UxB?eRaWMywDFV0r zu6okZRq$$&mXMGg2cddj+}BS{;opfKzR4ZWF{ra_1w+PQh!dkVz=7c}WQ8JJ`(b3z zB?uc%`lv$_XA~lL$)XG=(r<&U5hxVzdb$F|QATn8XYYdSgeuDS7Sl^{L|nVcE1`IF zZ}GrktXu%pIHgM`_M8jfnK zf%hu|5M(wMc9ci%L}DrsHxPj^BWYx-^@$j#2b@1gvle(lRR(36R z22Gh5&%}*l$~=%vE^ef{BU_PO}V{bjG?3m9X3c1MVPBY@rT;FcfT1` z!#zR4;Q{IA*a4yz6-a)_=}fd5>{OR-07hQoj%mnmMR++`Sgg+dUv`!th%6EDtOsj& zr!{4_)rnrdziTL8T;#Ve4AOt!CR{-QC&Y1V)XCMgr@t1*K#|5#;FNn9;RI5E8*~#i zQzfu< zj7vAC{rf9a_bq)OPP(4JIiWrHtig8Z077V2LQ5%LG|=ykCeRh}tBI{W0J-T^w=hsM zZv*N6xOI9X-v#0JBxpuidQgja%jBpnMcOg06u&_&B^X%C5`Rl&Qou)M`h>)$UZ6QZ zY6wGxo(c`2rDhFZaku}%Q*h8U#3QclR|d#8>Do5GZgYo4(o&!EXe!PKDe>jClKdjY zU*{ywmYtq5b>8gRM?LJ(F8&_%Co3}t94$Hs(K?JJjh#>wlWJ0!wjH{p)P6ZGTY^c( zhZ0rRyIDX@Rq#Ff5SC~GjttM{u+0zXrUD2yGn-afgt~uFAG3@PI?xS$$Ei68J;+VZ z2hR2!Nf9cO_y#E zIB3)c)Gd6QDjZ+aHiRMJ3A4tmZ1fdh34+Baq*g_jZc_C`{!#5Gi2It)bEHIQ4C~km zsbWJ$-UgCvtx-agL)N?DoR4! zk4DEao$XMr-41o!#ZBtA?fg849zJTfPWJ>-kG`j7v;y>%`wi>@ctRYNhGD=b;~qWf zZeSp3qrxT}^y@{zyLUg>k{}u2XgG_+sbfE2zUwr5oU|;xBglGD&9QxOCIU@k5=crC zj2#d(j8B4sWwDOQ26gy-U;3Ep;ltYE{&k?AUE82uTjVmH4%#A^l=j{FqG&X63kdg6 zlei7PU}dIW?%d5cyd3Q*LPMZMSCs$w&IPSxA-M4CWkfd#C|gRMkT`&g#z;~-Tf1pG z%=$A@A@kux_LCjH0k)zUGs&JI2y*R$eDI1g#j!mQEOFtS3b32`Rv)q-)EXg_j--^z zU^LaB8`2x9Qk|~;5dW8okX(uKI>zb9E^ex`YWNjG8s!?@G?5E8S}r6`RCB#^fiqeo z6A<+vzhBbi4^ktPQi@kZqduhbf1yRx9#{zqjPSYO8$Aq`zMBBtVpvV73=C$dX}9by$VJLWE_cxeZ81C~T; zKqC=icN`Gb^%XK+&WJ|5{vOYUq>Qa=^Fok)-bZYdH9)=fBuR^h^!7OiO8s&;RWYVb zG68nrjH15R*qegl8_8`7IU;38TCYg>GpKA;rj9%yWx^kj+T=-H6dKyintUza=*A&C z!|_=mF5Vqs=l}(m%^2o}(4bM*0!TK^Y9J{_>&jK)dUa{3#w1&Sv-u2_C!9OPZzOPo zf^LII=QT_t_)CI27()Xc$6NCyBqaQLFS@$EARaxWFRHDm(^d_|C6R&xSLkzGg(00biTeym8-2FC$s@N8VVVL$y)>g>2uRtAt@oI{U# zSy#%!ln8`uuXG>+xUnz|q-5$vb~QMiVhYKn3RHEb4mpCdLRb(mS11lL;QH=r}4 zDsmCkc^wg?lpyRAAuzaiWJd#K1dKa+Cy7y<~8wNWw9Q$K;&nnq{XTENYc-9ih?;zm-<#P}h^`0+ZU zG#!?xL9X4@WDX)STtQkdQpVg@rxu|lPCNq9tS)Fph6SW_197q!_vKenl_^YAxgbH9~7<#S>GM0s{HpW(h! z_fsh253~c&K(3ghkN*}))s`&6N7DcW;+7*q-RIXkxC;kmcklMcv6?x*(U~K7JB30n zWUS-8p!P!+QhJo>!51j|-Bh18ZJI7{^sa8cvEN6j)ej?tpTV__&v&Z^4TPqukisNX z8(hj<0WXlYrAP#DHYyU|?h7BZGe0`lb- zbHk{v1K?5cqj3pUZslF#51x zJj5bM?|u+jzDpUA!|r*iMTz7o91~suR6@Vmz8u>Urhvyk{oGi^1~p7EEbDIgh(!ie+> zI69E#f1%t$B|RR?4j3K;2in#(5z;q7nwIh9EdgD)@$`(WAVv?$ zeTLRGh#_cENtvE7A!4q?kS-Mh@dGG8kPtYIQBIKxiWqcn08-+r2W)^H-$gO*$PrWY zI?{wQk{m4!SM_ey5X0ar!ZkS%YW>O~Qr_RY<|*|`rVlr67{BNiShQlA`v~(-Gst+4 zP&o_w?|Qf!a!NEc4@pF#{7@1RVxEue-w1qd439Qy%W^DagRF73N*O&kyl9EA#RHVQgD7$- z%7=GQB20)UY=A$sX}&uZE@|>3ZEYBi>w2S}WuBx6hRdKr*63!(Jf8SqxcKX~kCTPO zccMH@sD%$KS1Gv~x)hXKNp+~DG%QrOBVi}to%B)6eAY#(DIoAuK4!6#=68IGAyB95 zjDS=2Bqs;|r=eQ_P=r_@G<+3x_h6@TB&om?KdCasIvVW<;ojznDyR;|W64md+S&}Mk%yZ@VF3tAxyOGHl$3qoC&a{P zL%l)$h}s!HlY zG&~6`SBmDSBMb60D?ub#N(`h~yu*{eg zY_4dMU8X1*xGvNQj*P!!ij^_aNQehW!Z!So@Q-BpQ5 zbG!KPyE9O%3qo8+LI@!k!i7K|0BJC`0j9mlLHz6)@;ftlMiFy%@OqKx+zb+RU^JVJ zcS~SE@OUMo9WZe$+$NrBD57w>Ic+J)J8-?AvUrjHfVW|iB&uX`C=SzQL2(rPQ&edn{J5QjhN1cZ7kn~N z9pnJg>m*iWqBh8;Je_WT)A$PlP{Q4UP?J6Z$OG|@BFw2G_~vN+Q9*Y#NxDKF4xHzc z&E;dMP!}RkZWzY%GtuDT&?wxwUJ9617a%=JZjuZD%tz>fdwZ!BjWpDHS`;(0vs89<;!=V;->^&GA+))(9l{uZD4~m zuoT5@8defx^k3^PqR^!7?7uwc_jdE`K_yNemz>?}jLoBJS<@If!eaU3k3YTz;BXKL zw2OK=7d!_^9+m)A)7Mu@*av79_ABx|hY$cj7)oODP#Zafqm^jH2yu!r;^hw@JfNPD z{uz$^n}ZQJ6AQeWLa>cnp_ErlBD-N#3B~~=K21yx7`w5=QA{5S0z62q@WiK~i9V?y zi-lueMrl~=L2{;$YM_noTBeEU{lDh>a5|27hY-*+;oW-u^@gc`1p6LxMlTOWpf0-~ zONbMpcox-0>>(c2>A*q;{ENTa!2Ir-!>>P>b)39ZD}JM~b&W)C#_JRB1~;cNJ&z6y zp+jPvX7qL3@K6!KC|MLNW56(G2oc50%>o=p3J|c=P0#HeJ0Vps2S}3C%X>LMNT4$= zTHX>5{~rkL&O8Q#(Jl(M1V6+`8&Zm!Y2My^;#cLjHsg#SKMBA#Sky>@G>iGPPoU_^ z(OJ&}H!$9`3`{LKR*a8N9x%}hJnru5Mu4WIwgqgf6sZ~;Vtrg2P0t1<b-ztysqa!s@$EYhgYpxvlo@n7q}#sE<_RN)(=7jr1?1qkwMUdufHWCM^?>Y zKnp<+%}p{4OSc;^@l%Tr4?^qPIP|)6XLm{&ZjX>X?8=1LQEI-&VsuxpzyP#})Lt)= zNCXWg6>N}7X?>j3av+wO48UL%({83;S2frJ1|qW`svm4&S@|~Gwmr?1LEa)YE1p> zjrjSe{f|Jba82gh6Smg6>4}%fvGyE&>B{S9@csNZAHO@@4p5)O{1KM-YdQc|AWv9> z*$iqo1=0FMy#bIV>}-`#jy-t~`wYsGLb5|HKzy-{QM3T6L*hl#r}scXZcyI=!77rxAMjWG56;Ac(~0O&npNZ-n&F0b z!%R}gJ@jF%G}B>eXf#z{#n8OwroJ@@3>1)D+Luo0cs2asR5FBP0qlsx)gguwjaztG ziJZ8!V2!4A4Yi8n%%RS8ii8g7^9bB5Q!EKr6(+4pO?D0)e_{4sW$8fk|NzQ0jx=ap~8GhyJ769& z9BTit#bdRQ_}^oNqx*Pjy}z>=;5kE(0C59S3N?kB?E$|q6kwE^Zxw)X5*+}zi}AnW zs3LGu?*Q8VtSg%U_&lfTfRCG#QxUN5uFZUcs6iE>0Z@-7E0d(KISeG5as>(v>Ue-8 zRN8TM04Jj4?`gk4TDZQa0$~Y-2k?t;dn(Fk5*uZl0OsNfenucd<9Gq6VWXA{WIJJ# zNc)Sa(j0=^6okts;b|!@ReCd0iH{}EU~L~R2+iMJsFQV>{|c~SH!80s&{ysCAEWVi zZ0Xm-5iNzEe5yksE-)!u`wrCSLQrSSYi){71DH&yEm>RWO^US_fjfhWqXKA?95wMs zu_(}h3l~X>xQI**@tG9jgYH3ntDl;5&MM)c6A*z3XP7+Y);{gEA{;WDppJD7dg-Wm zP4!PwhN5|}z!C=@JOS|ffOrW_O@F5yKnD(e9PA$R@OhXVn#D(&KC*A12hwFI15i3O|13B5nPMU^O4mUXB>V!^Pm5HwDQ=LSv)PXA3R+S zC6Y}7y2|HDeV*cBeg2b=-#xOxa+gtx0>y`j{AA!+G=wSytT?rseYOOq6LSibr$Nop z4ajK%Xgn649M}<6O1Drx~!UmRkdxyDQ-S-R%jV_owT_FGg9ugN{{4e^aD zZZ(h#Dh+e7d=R@};f&=+suuFK$G9ZKt})*}$3X7Kl5M@OtHnFQ8qC_Pek;t`@Ox*U z|Ddg3lj%QPtHk}ghts;Q%~+sV3%QqiP%T_!?ds=_#A&a*dJ)zx+cK@&H#jV;J>Gn* z{Gb2Ltv@e5d@SFKe{JkJ@f%|HHkjPRB_lTQ+TpMfat$V9Y?!eL9d!rm{j)}472{>BA}=LX)JxsKs>DB~G7UQ^2QM5_E$ zJ9@sl(K19&YV=uIZk%&2FxM5Id->0bN+7R(QqQ z{hQ5zBdal=aE@-ACEWJVK~>N*FB8YQ-f3<4qNvWl#%`shi5aFavu+4m8m14AGA)99 zO!ih|GSem;SqIfVH?T(p(%>ojhHPE-t(@yWSQ1o^^C8_MrkPPV5Ze#-r>D6X%h99d zku@K8UR@;bol(HKKI2`kznl7k%nSX)vX#dRvN~eqKmHBHG5A_&wS;ed9JRm#<-WPUj?gYW4pGL#iH2>bCU%TqAb{r2|BYF zWyQsF7pk1vt`9uu^(GXsT=69Bg;Ok6hlGUGhpLa9d~K5Go?Q(WY7F%H;KPtEC`a2h zu2sL->Cl%P39hB0Sr~jFwYMz=wCunoIp|i*zQsgS#k7E*$vXCqb-rVJ=iSK2O$*km zEE4a3%z7N?;Yd<1EMPsl{$>3I zWuw&CJ9Qc`HVa`*?0t;07SC7Kxb?KDoB1Nbfq#AaaMvO+JFD~uH}e|nV~S;*fvaXA zwp#98kk(+H&2Fh8*-ewKn9p{yUgNtmW!=@B3?1>Iii9^_ddEFgE$)|Z1sY(59l2So z>x!F#zU$+9j3jjd(31rx6_Avaq+W`iSDR`mnKBa@_UrL*)+CiqnZ9&3T&J^}!~1|M zIht$YWCXljqYJmL#f+s(B@UMkEZ#XXQn4S%)QtoPVdnOlo`9&L*3Id;_cqiqY;0`m z-KExF1j;)@@mFbDEGTPcWG+;%-q5!dxJO+AIQkk$!{@8(J;cpM0>n)K0H6UpWWIU? zo@jVdOGy7v`q&Wnyczh-JfwBbMh@NH#xE3=GEpm7lU7T|9Qw4EHaIYy9lIR1J62Vs zEkM~-{fhOJ(`NU^98Z1YX^)=1A;NpHvvB>N4wvngq^Ank``9XLjIfF727S5zwq*jZ z6`>0tHT~l-f?D2|a7GsZ^ohU&EI~lBrYfcJ-D6gVjSdip&gL&unG4A9=&rJm&`{^} z;kUu+IM3%?WMz|>t7&u(12k{U1D0H<(auQBIUsC zJENuRmPZsmJNf?QK27Kua|Ze>nYUwAi6$Cd_-5pB@Uw2TR^`+Y%2;wxU2NUi4zx0 z3%9HS7Wg>feUX#~a_X`69vnYS#C%D26C*JF(u9lm!cx=xO#m^R`d_aSU!15CuHW5} zJ~kQ&*gOGK6g1*%EY?<~oKZD5H}}CH4knPwEL?7LFf8oTt>}?)HnMM3hTXHwK-ok7 z3V*Gny}iAPaprtuP!}>W7l7^NzCeA`Rg%Av)gxQKKI!e9MFRSVD&A~zZxu`Ij#V3DVdXa^oyZ(7O^yf^pxSiiZ+#39K*se?T>Hdaer z19Ha9*a&+bdP?PZt30u>!lOWqBhfsP&=!?H%Rl}0?cdRro05Q{lm_q_{AMZs zXf8g|(GFRP+qrY+q8Q@f>m|Is<6I9tn8la-ezy#a$V5yQd)#y?OS7@%w>pZoZ(i1f z)VJ&r|?4?*tupT9iXTzeL_Y+xZ-g zWuaS6>ghFP+VcAHQ_Y_`bWGj3E&|8S#O|y|@%RGb*$kjW4Pbz@^t6s@ zz}{{`nkj@zVncAir8Prm%Kgy@a|mEYP8l9J`VX}fl$1PU1LpaUz7MJ!>^wqzP0`Vg z&e8Q4EqTs;Y}CiR^zifQ>f!J*MDAgCZ+XPle&hlSi|FR3vwAh7<{nORQqGF$#YTF#E`^FwB%ys9dDkgvrNAB##|CgdZsaLpRX z1(+yDNJJj0el-A3$`Hc6H(QUplfXcD; z`KDShVH(~&VI}HJ_V4GH{wC?}aXTS}wuDvvMdN?}w<-fMPy=1x1C>PHBg)FQqGho2 zOIG>YnULw;go6Hzl3BB$Cw~#}H(A&Qtnx)Lkh7UcW;&jg9oV^Z=ThU0ImW=MvyYu; zUoc>o_Bs;BZGBpQJ+U$CdMe>q3knM4)Ya>n4*ukiPM&NnPt^s;UJNN%APWS#bW4v2 zB*2Xk4rFGF^+uK>idnlqse6&AjhWr6U(_!!?B4K=pCpvo>YK#Jto_saMjyj+*5h$q zqNmx5@*i($Z!P+I8$hACgv2ESed))+kcdR8EmjZB5; zwNnX?z^++J-HGBhRNfkJW?aCYS(lWYJhZuv;=2ngE?q#bvISAfMscORuc{75g52qyU=GbE!k50_s>(4bS_`_eWe!};ixm^^j|#k>$1ldQxsnQKVP#V z$w&|q#azLMJ6R2zthN3Np`%;kBMnd3nXc+RqYryymZs?a<}Azor}>LV_V4gl?}A2( zjkX?_YJDlZKwshaxjNFE{pX*2eEGe@=??Z097nN**kEuYj?Q5brb!-|=m;7KFJs5q zHAej*rPUm578h~HnaG9~FhsidrNGYOEG-QS3zOrXcO0dq=g*$awrz{t-G*8q>L4x@ z?$U{DXuPY9V33KBEE|02)qRumY12z2KZ0Cr6O%nf+BS(OP zuE6KtNlMx}d67ag+;^N}ozI}TEJ})>t{1P@Xg$iVxHy!);pBubdd!eSV0-LwUVX=Kt zZ7oOoK!H&`{D(6@_sv9rAWBy~W-MR0e*Jpz{rlFf{Fq<11>-G$BfGlKcPoGU`zh4H z774)O5YS%{@D(Ca2r5VCs6P65w{G2f=g>L{dK{&tX+6<46gc!Dh&O&TWqJmD^opy@ zio=7uBM-Tt+$1vEUN_3)FjQ5Sgcxa@sj-3m48A zXJ?SbHYFt`5lo0-!dhkH)YT4Fbf52wF)%Q=QlK5MpZiZ*2cS7{_VeJbEa|qflmfTX zWm8bwVIulngO1;eR%oxIhZqSUFTUgP3#_23s;Uc{U`Log4}jTINX!3dH2KN>_VB1a z_hH=caQ(vc3m}#)hq5&%n(o15<%glg#@&r;N8FQTTifk-0|$?;9XM4wawgsM zf3`iJm;_XG4FG}wd6toIKSU`CA`Xy{^8i+OBfDB|nx1oW& z2)4w%k>4y1DJv_(u&qR*aD|>h16-d~D~G0@UaRuhuPv>u8g1o~Mm2n7RT~frEPdbC zzrv2^hmEJ&+lHRXY;C?vYww>0yp_KLyj9+X_twg*)Sw)anJWa)YTx(2JhWUtZ%0(gEVE?E1TYYfhX)t=b?vkXZQK=i~3st{vQ9@ zzfySUOdHE0A&~3mVdwWGt62gV%f>{aBGz61{n56J!gRN`yH{nLP<-muZ6kY*P;tGf zT#|dtw`I4Z%W^e&nT@Omv#&@i>IQkVg$E_qJC51zuAH>Z`w@%hU6A&_{+g;NkdtwT z^+m?}=VzK42AoUHvfS$|0)4*;w?DjS=Fu}v8Q*Gg?Aflmey*TH!b*z>#$7y-Y2tcr z>S@*2KHd0NAGmSbAG+Tn`j55=1`1sCelzd5l~}ubky-DbKFN1@{HiLiT1?uLw2@_! z_c;2nESM|YTlGTqvVps!-t2R_ug7NT@qCjHe7csY`OCIwf6LA+vV3Z&$m|ISDK6x$ zIQQnbPxKw0$|R?+Q;N@TvB%xw(+~fBGjruC>72}ktT{oxa#6e0nlEk`zRP3dl=*d5 z{P9(%r!1Gmh^4Q;e)3MQ5MStDE$nR}k<+1C#;!i5_1zV{GI2%ENCQUM;P`OsCH;`S z!uK@m>|^-S#x!-lw<#`9@)mMmY>I%PUzB3ALl!0%6@Go8|NV48z)D8(8?5W}l)&|# z7unvFWEei-#m>B&yogzG&45|*<{NzA<2#<7H`$Vw9`ekP<_y0$Xjl2iy^7Ms%(A@7 z`fKgqThrB&Mc)pXJ^k1)H#nrfkja0Ajmy7&Y^3+N-AQX@{mylPz8@X<$7>k>_wLcl zOYeQV>!w!a*F-a8KkN%5wbqO2sp&jPw%^+rBkQ~;Y>c*bGJfvPp5D&Pf*0&~BCl7q zCO0^#n`Qj_B&1jdrelgmPiK;bvQFd!(^+|1^+*(Qx30n^z2jWN92!0+Xj3n`!ab_6 znd9q|x$&}Q@r0Z9%^99*!Zb0Jt?pG&s$BiYKy8F@fUecbH)$LE=}67|Sufr!`+Di| z!%a&SpJqHxR=j2og$S0H;tKVB13K}&8Ko9|Unei8eU&Ee_O^X6^LCBzT7>FF&oph% z8tZOiXFlpy)z$gxnST6Xglpn%w-Bh{n(Hj4FxS?NV<~6r28oOT)|ku)LGz%=*XRA2 z$HpdKrAc`AnY>TBZYCVFkah{pg9&mk4Zo1CR=Z$>VBg)z-$mijBscRQ>sP7ztS$U< zo>IwoH+vs(GVbk)JJG@Yot>hvf0Fme4|EN8+Lc}Ur{YRjGLh7KtHhF{c(x`xf5(Q* zcbA@E04ywaUt(lwaXz>tbucyvx8qpxjNNlo%2vhrPjgM)n=z&zA1M7>p5ucdug103 zP4_+-?TvRQIZgXd4T>f9m$;}@F8(9A{eF0VM}Y5l7Ha+^HeqMxG`?=*>VHCF`1bb3 zs#|1qK~H&di;%l+;yRym;<|cq%zU48{)wd`oM?WRt*n3LTJxO?(AsPzv*%lz;=Ap4 zPar#Nl%R6-s;Q-C_K~>px{fBMmIe*Vnuurxm6}yAp)JJzR?!nmz`*%1QiI?1*-0eP` zI>~(m4^?j;5HgO?$Bq~(-N|lTR{XskVec`~OYBsIqlsWwWS@SCO(dH|kC~;L&$J|4euYku4$nR0NY}+g^G~2e;7gf~itWR#MpaOnmAj*D)-B^g3g%n8*>s;%%PYXz1G!W7?dZ6?F!#}Xp zR*-+F_ZDS7iB%hTm}n*Xi%k@7CiEtFmFvBJDYX1n=&IwR z_6E%AQorv6P~yi<`Ai5Bcb30v*;g$Tv{GUr1s|^qE?tQBjL(9{o&`b9( z?yLO$51SJj|Bpi+KQVw|LCZ{r{<{8cV(l*OH9clZDye~d?UQPgzBHJ4#!uI78M03l zviGMmp4?I=e7u?O>oeZD{A$komciqZhi-rFfGc`NC}XQ_a;M8*e@tAax4yK}90%EV z3n)?&G{!9MwX)K%{BorqIR?bD2DY}J>HPG_7nM6@9*+p`Y)|9+={i^Yr{Dkg&8@yD z$(3pL`qKNVv`#H$fA*Z+xG>cq@)h_2pLgHszt%|F#EU0}hKBC=h(Zm_SIj_0F2bKRMS_Xj>L!-sEfX`WSC8Sbw- zH@QrwNM?0TR(ocDNvgr~qQhU=l5;78gM(^r72rR{#K$Xr@r?HB0hIyCwAV&Oz1Mrw zL#_!aG)p6;(6kHfDTNkc9+9={_$%&h$Lg;f*v=z>gt#O-IgR6hb8#J0?CJ+G}*d$Z=##pd2lO%+AI zS!Ox@-#-hb6qLWT!k7PQU$y${Gzz)O8h>#iPEJ1$Y2g5_n#>D~>aqn4!}hN>*}1!1 zlE;CS*>ltt75xTszcUn2B=q$35FhfEe)2NUlrxWS0_>~-9&RMKHz#y;o4Zx0V8JDs zduaEOpRy(IeY(aE-#nTn$$PUu>-h`H4n+n}%2saa3`ty3-=-B7`Fcfcq}16@|J!w((f#4YHFZvGKZa= z{dsBW0u(@=S5zcE={~5WlnsVp-u|cCCW(FCa`IPDFt}1iwP?u(!G*1 zrTyZo=H1!qEh+p3bac)GP6&7I3{Gz6I~7y}BGET7_MU-@K}lT#!6}!dh1DlJe)@Je z@I-8KGjsD@mX?;gN*4(ls%Y{SUZ_4iF^6UuWZbCaGQ6QWmoqtQH$*%vM6$nmwVt8-MOpb(4m%42Lgudmd*cBJva+$;$kM`86 zhSayPzW9(dR(J1HvKzWx7=w8XAHR<7K?{UF*Tvm_pB8_nmZbZi;I~vR_PAN|*T+Bn z@869-ve$aJo;|NfFqDyK$s9}3ny)g0nOiUy6yW)CNs-S&=SR8sx^vMly$b?;cigO4 z{O#V|SpqA+$+B)dg?ty(&V1IH{AIX*`u8B<)$;Q4xKjVmSNaF86u$%v9wQ&jlKjJ` z|M_sWM}JO!J9R+lq@5z%QS;l2LYBMqQu?)Jzw~h%U*pv$PM&n)V)lP}E~jrVlXRzF z?znx~V#fUAyO(|Q7-k+*L`1N-rV?OTyejEBK0P<4OzoOm3N|(C%!uG+>S zhe4U7a`gm2;+)@q^&d$CEf;p~S7q0OCH0S(`MQd=HCbTR*0;A#2&~e)<-YrF{uc)2!(V1S zxyo65fz|h!>vG|{fwFa*U4AiKq^a}iZCD<;Lyslu5GwnG?B^CmpZN5DSA*B(eXJJ{ zhuQauSji-xkS%#$upx}0pFI@OtP@mXlJ#Ols7+4GC!z2k*r{%f-4RZ&g1yT4ytU>G znLAsH&cz4xC!S4UbaMQ;&ZwaEO4_)TrSn0#c~l-MN5kB43;~V=v9uiB%q|iEZJR$` zlf#*)bs9D{TQPY_1Kj~~*R?Eu``-iWPX_V7KTLmJ5)@N!D=fjh>3zOpwXMRg4Da== znP*qYW1|M->I>eve(Ei`W)oc*E zB)G)&%u!U#Ji76MLptZk z{lk}^eB=T+i2rnEJA||sxk|_>Dk>;mQ~8v#Y#ea9bKws6oppI91iYW0jJ;E1xI4M# z(!tRAyW})i@oIZjd}AM29`)-AI`#KU)>u{SV$C!V&Kfhm&Dj}#Jg-+x#i~nTozdd* zPfNja+W5e{Oa{k`&)1u^^X7!IhfC0IbEono>oo_>qorCBnYD5$%p$dX*O0bJVW35N zS`!f6jcFWx!MufAbk|hQjoN=|ps%UkC3L7_($dSu(v_Nq-@fzHE|)V)m}xz}Z7!8! z1FqZcTB^>peR@ov?}&wEk@6%TKPn>@H9x}f$XelmI7@YYHl>lHH$EYNU%r|}!CK)?Ry9pFOy8m16_X~&8!d9%zIsReaFlo6t~okkof-0F`?6FX zpBcFzA6JlJ&!_mMY5n{nmuoqP7JUi<|NDK`j-bOCiEPQnPtC6Q9>~MqW41W!9JdJ1 zofI)S9Vw^%X>Wgg>#L2=^K#3R64};PR{YUp*WMwQGLl;n9J1!Kt^N3?^l#RViJM`c z-wGwcdW}^U0dkt2CEXRu=a`XW9ADa_&+qW)5O?(Ft@Yver*q7PZx>kgF~?W3|AoVF zQ`mP_a*xs^chK4>dooj9B=MTBT$OV84cAsj5zon2^8O_25S2P|xLxR&!3yDk>lKL? zY~DYyLvt zuDS5oe|U~LNt9#h4P*7ePqVXQ7+I?m5e?Wq3^m&W~{fBx_k z@8CprZb_SK+4i6wu2-w6?RmIE#_B5V`a7QMonOfBqR{-}CDB`R_@10pWB8mLWFxn= zplC;R6RVnB*6p=TeKMVFUz^X3GLsn8`jV6Ln9Qq5%qV9Lr_PQWb<=|v{>v^LYSoj; z$CTaR=Opt)qlAS;>P4j9l8XP~GajiOxiB=$$LiO&;_J81+B?ZgsI3*=rSk5)d}NZh z>H&tirCa^`wK@SI-!ZO-Hcs+16WHUTTucRJcG>PYIBR{9)}kgm`^%GWhF`RI1b*9y zaaz)%`Y%_ju|XLN2?NfAIL@jx;rvN~>C@9DF)C6|q%P@avR%^C(kbb^zShS0iR_yw zzHip+!=lwrvXhF8D~;Q&p3f<0b2(!3Wc)~;XT9{Dt)Mp98_3Pu<&&5;SJt*ob00@_ z_x}0HaXrWTZhgltTJtS#z4gu2J-YK;cY$%CWQ~w!(o>oHje$cF%q%I}O@d#=a-Uyx zRLx`JktphZtEHu1Q|&sxz0KWb{n)5V%9xkKyXzam6{7~hHoVf!?H&A zoZD_Un6-WK*NYsVWE+@?c0SZj=}sI!lh!zs1K&H7BFiUvc}!~kTZ@#g8wYVFpU91$ zNrUg5$(P60PO$7OGL9C)4r31`@Ot5)IT8Pdu%CAHx)PQI zmeXeDo_ppwck8jO^y6zBV|(@0PR;$gMoA0nYm2V9%8--O)arwmt;q7@y=#0zvq?be zFDr`*JdU4-_hgxklu^O{q#QgIZsVoM=q|vK}^|n1il$`0THC$u;C5+#8<8wWENI^U)90FV@ z-)O1u=;JEvKdy7UOIidIPXuoAW$8rJkh_~7cp(e=z~w|=4Nke^X` zBwNROfQ%wu_QU%Ma!kt)VB;)T+0`GAB8>jF#ONnvPWW&ik!x-8Lw%~)nDZM(j4dYI z2s(r2uc$7>lKyM-}gj{UE7nHKvDSvtB zNc!AHe#`j;K0xBAx5zaoz7dBTk1+j7@gJ(I4O&9^j8e(tL2>eo z&xI4PL!A0w`wdE0E`lEY>yLY+Qe#;+B_2pqG0vrOMcEJbu8k6cLRQa2uW)Bd0?-w9 z1`dZ{D;IKXm<@G>vHhB!r_t((e29sfG%t}kW=*OeE?`z|fjbNz4$hLNRnO#q^SFdE z83uCh=x+wSL-W+#xoIjTo`WA;;XDIw1ujmN*-&gZ5qxR?zbZgV2P+q6bn$FmOObe%llwTCbhrXVr( zKpNuBRQ1KwYg&A^^lQW{#$1eTd6u!`jPrPj+0N9sAJF4vmE)SwL%jsU>)*+Mh-@GJRYOx`boRvj;KwzRGRD|nANQ|x`1V&L zT=h7s?OnMbzb=}2dsBPvT`qQd&iD*t4RmLsxUz^gilzV|k1(If?;$?Up~S3_moYxZ z`UJ<%f~~1DWP~DZKS=3x$xJITxTlGi}EU()#ke! zh3zff&w#TOUFj@)YA?on03k>!-cExbudR3qV=vasuiei)T@61fZWX5;$ub)#OS^6> zD;YUE$xkI~t+f|_uAa!QDkVQNWTjCRTVneQbVtgWMwqWJp$mVS8UI&|qa@EIk;G!| zc1av;LTNjB_MFq&REIoIpfI^Ywnqf=t?P@rW zBk32JF)-bH;uT1P^C+uwT_dY1{+oL>fGMl|8vlN}PqyR6HsD&qY)tXIxu)2pQ}KVW ze*dEty-``7AA*sSJylNRM5i^j~6o>Sl+pwOI|2TG1EnDfs=$~*diusVIzL@S2+cQt-cj5ebRM5M!N{I!D zRgsA-LTMSECMe2~?}+!LsZ92j=f%^g(88++Uuv>{-iWi$p-#WLWQz#il;HN>|y+VgXdqhro4%V)yDNNt^t0$Q!CZmR(hnfM>=?k^B+5!i?o1b zNuy-tPJeToz;=nu1{khAN8I`vOYhAgX{VDxJEUeaPdHBot&xlw7@xd;Z27csILjrl zrA-RYs+;G;iwPgy1um`*P4Zp=em6@^B3l7IzQB9gPfT8h>vGpX9U&N}XMbgAWUa5y zJ>fJ#k4}amxgeS;dzcw}5g)u- z^yTgg*G36ZhZ6_IeBB}qM_sBT7??M{*?WA`N&k9p-i;owJd0DT4}UbP;%$_>`>pPL z$$NP%f^yZiZ~xPXPw1}8VB{7IIC6IXlfUl%ahdtQDSEtL_(>naT-;_2G+T(f#y)(I z>dl)HO>XKmlq{NZE20c(jI}JWQ~qIIsl@Gg7deD!bP7C}bOo8{t=GUK{uplWbJ77X zdy?RFQaQUMo5~iukp$ukR`FVnOW2vm-VopI=a>9_^w#}X;kEvVN%S>gm0*Kj9-Y*ST!pp^{Fo`6`xz{Jd4|tX z`OxL*QwWqyj&t(nzuek3wFb4N{vWI%T-rH=JU9(e1!y+wgtmps z8g{$^!=3Tim|O}ZkBqg}hWWp^hhNa=lS<9SrmKs8dF=mKe&vbkx1F{4XisTL2TRmR zi)9ryXT`=eDL4!}?k4ssnwH*CGy`~9$0nVi)#>SE>xZ3xF{fk9Uo>}o{IQ7Pc~mwS zc8XWeF5K!gyT)NQz?wgoPnriFAH(M7)@A)cs@<{kz-0?lF6lLRsp1g%@+13KtU>V~ z-#sods3n<}*B!#&-KLj+3Qz``2svqw}^OjsmCR&>1{Op3)sP z#QuPdSRY{18pTS^GOWH1m}Fouk9^1JE_O1$^EuWsI*pJt$)R|{VpPkqCuz55t`*gG zV{i#Ud7ZQ5$M)5gG?;I_hgTcVl=pg#jo5@y$x|PJdTO?m)v<{ASN$P>HRnfp1-qqf zy{)tcxI6*wZPj7im(;;Sq?oi>3|&ECEpn!jpYW~aGB#7Y>Dg-ErWa{&DafLs)i&Yg zb)XoVvd2lcv&XT)iO6x4BW|dwu=JH85$0&kNw825=WYGuc~Jc2T|BkPEI~BQBi&cn zU>%O$5jo*(TS`US~ck)zjQhMH2m>|4qdoQF0XB1Lmj+=X%c6l%BD?deHvp+-aY%;z3L z@z=|i{IxM(rLcU*{DsM{;>%0?dj9GUSE9C%<4n}Ehg^FW)3}K}mz_Uq4-NTF-78xO zuh2NvnO7fC+}QRr4CW$E$096;tUsKBYgfI3&4y09sbfd&6pqiS_N!Kx?&DD}`wGs&avNULXG(alKJ4zJrZ#yHztkD-)gT_(E_(~n+^iqV*nAFOhMN4M zMgTm+#J}2u;n)fE3A`jPtG3z$$Z{!$PUOetzhMWnX?Bbcm>f0YI!ekk{0NqNBw^3x z$(rySmo@lE_j?ysBhm{se3Gf#Q9uJ9I6A(#HGLPcQGvwS)q&`^OR|*J{%Wab55twJ z;EI3y{xXH}yKC6)f+E%jd`$DGis{AJHA{^6hhW*_46H7yB^5D0~Jc;#h3!zFG3k3M5BWEDnwRQzVHpW|k!c~~Rj zmhJe6ZzrtC!)wsGdKwso8j+*H^Ih^`PWmXG<=Lr{{e7%zS=T^#RhL$8Me#H4`+Eq~ z1Y3!(ef=T+({Br+>^~M+uT`~lzh)$MiD#t=+p8baI-{MN#pb7}0Xf?DhspFrzQVd1 z%wmUn5XPY{Bh}suBvUhoZwLb{aTxoppxL?!doT1_Fp?zZNDkMFoN#*w0{9mh$Ze5B zCBh>d*DE+=96MKVumRB++y3^3@O}HvnPGdX zOfQijNun@(yYsCUvNDVKDs{V7d4TsR-iaEj*SXMn-*jU)p;uAVc*t+5pVxaY@hO%_ z%{{o|+Ua=j8`~IUl6jtMFt!-s_!(F87&yl2D0Zc?xuz^qFV%v#XJd=d+*DA_6fv5! zN~HI6I`dD%`tgn@BD~|^F@k^xpD9aOaMl)sF<)as=gwnWEoP^LiLvFhe@I;o*G_Wd zyzh0({P?|pzPj%l`L+Bgak$2~yp&bEg?mo*Ov$Krn(GegeCNo>8RvoOE9jroqFPq@dsj@V@84lmk`N65@j0K z6!FBjCSs{Bqgcfv_kG3}RmlXtDdG_Qh-a1gK=}yaPdsVX{5O&)N5p+Kvw@PJj;wTC z2rtHAy-y;3)+Fbvpi)MAdY)JC&7sqF7qGl%N{RJG)%ixQJaQu}wRpiPr%T%zI4EVi zBBp9{r)MeeQpeX7>tg^k^_@`c2#u6tjdAI;mRD%cxVFvhmWwcUy-3e9$9(jFD|sMy zx@7ovs+3(D;h^hRj$UG!(>((hUL4P(e)X88M+Ysp9?$jAGx<~VwmY5FQ9grrw)&k` zBv{#K1mI<#UXpGppH7D-pE=v%iZQ#1J+9p(D{}4vJ5R;fI=wKgt!NTBRFz|FDt-wu z4i0Ta6Z#}J>F^&L{qw8HQS(`C;lf;ljHiay8yS*wvRLz8H~bZ`9(OMe1M!i={{d{M z-^O*SRRMAUta73|_eAW%&XP`Qxt>yWYg&w#FyWKVqjcOCJ&Y1%mq>5Pb#2F4dDKO7 z+hcQ+o&Fp_2`k4|xvp5`)j4Nvs8lmw(X`2e(sBC?v~g}Zz1ql+N7&%&yqMvQNn@P%fcOR0Wq(al*N2$uy zuU7UA!iIhBZk)SoF_xv&I-}Nk?5sE|j$Hv`+J#Y2rtEIDm3STqHIzt~XH0|U!Xqn> zACRQ)@+PaV48UhKDoS$T3WG7QU*)%HnPWt@-V7QEzdcmUqH41JPwtv7~{c+d7yDI$bR=q zW!PJgsIq^b`dQ#SInXpJXZ*F<#lUB@0Ptg$i(d=lsJq7xx9Kx z2zpOB93uWr+gGVVhu?!c>#;C_-d}XmEgR z=CII#`>z5CX`3gF7+W^r_;ptwjp8n`?pM;IwviO;nafAr~S~zF;d#afWSJuem2NM_vqmFL>RsT4p>rA*}=cHAO zIXT$mnzOy=mNBx?M5j1JNLW#@iY)G@Z`WQOQp$XfR;gxU<$4#pLb{af)Os_FO7Q_Z zDtzs-E%(~8R2lj5r*&Lr z93|I;G2!*uBRkPj9b99nG84_1EV?%wBBneB_K)CTuoh!<{4eT(JiSQi$2xFab0hv8 z>OXHpc5x*`;$)Z&V+Q8xs)}%78fF!ycn<*J4Rc4U)>x=@;&TEY?LVp*0-@f&g%jo7 zhx0H|JI{0KtM>{X9v;`BMH^Qvff0#6_p3!TdofWj{JVe8VE-zaUo20N(H!)?{wDK( zjq3!UzspXIH^VHnQMpVj%a0h?WMhC`EaCa<4O*=0Yc1A|L}`Xmv1~SzOSG~XmWFiMwVp8{0mRBM9g{vhvucKujDsy#qw!$I|-8USAR)Sv{7;4ut zntMah4cD|?&BS+{fr4mqd%aI+_HS5RiTd%|2hrq2_i^grFwy$Ccb^q*lyQU$LF#qR zju}O^n=O)9tt@(f(%nU??zNV4>7VX~igj>gGs?jLw=_a?w{{WcN3Emdx_UiMm1fEq z05>S8Ny29Th`rvnNJ{b*+xM;=mYoN9Tk1WlYD{G`wzY`jivKhHt!XwjN|3Oxs$qDK z>qvd0dL&wM=cbeSHRc2XuK+c_DYgC1%b;w^^la47UoJaHl;nfh&^jmxI_|djW|1~< zw6WSr@Dn&VbpJaz%3LO1)O^3w;r&fO68!7L-W}}ONTSfC>x(p&qa2Pt?f20!tM$th z2A&P?YOmhBBD9wIG$%eeOnjetNWbAG>>{{yKeyM3SPXD6^8}oChHDH;jAwvR`+(7e zopLnX{wfPPrysdBa!|l&&$q#BdpG&lr7M5jlZV5xYhpx*y2-a(8N@S+#=tPAxtMQfg5yAN#Ud4p zckhrdj=q@wffwg~ug7bp!Rx&#fDFyP^;aK=^=7%|+k$Tnl)QT4zTwc)kuSdA(yI5U zh^qk^&X;Go&sY9{`YptlM0WTF`0_v%rsY;SKVha6lcv0<$N!$UIs4rPT);nNtfe2O z=OCh=MWeFfuOs?@{-vw7ix#Fe-#*HGjp_7J6m9HpV66DZC#u)-=kjfj#$25s)(yTU zk7{9dbM5>a9M!JpLi_Txu(z&(rCjL8*0D-!XFS_FH5B14gtbZ)dcW+C2IsL3Cy{{> zvbuZkDBdVdmWqWZ3aTaxRJAjJ+(!-l3-1U8n`AXQXV@1eKX<&tc`V-mMt$l$Ezs@O~xwI_OcUr&my(-ZrI6 z&ja+mVyvdQtRHv;xju>RRlogopV_(sVK#xwhbbtl*RHgH2;OeP89e4)YtYe3S&l z$gScW2LAHU4| z8@qXroW{T}xV})%l^wb6SEYYDRPvFF-fxQToK9WX0;S5heM(sf^OIh6!zaC8>yx4< z+YaAnm8xZVaZKZF&UkIabi;t-0Nu}#{CJo3JSc|O<3Ltr86)nL-|8WC`4Zx{sra!3 ziA8Z2On6jU9ZdcHkx@No@fQ`lI2x0#fYeZHakfPc@t{S~1Rm)Iv_tBxPs==ZAI zr~X%^_sUGTc7_uY1y?zNv4jz;@RS)D`4|K+&A6= z^Q(-Ua_hyfiW;k5Ii9=;#xXA=PMmW#mn16MvNcB@{mQbGNi42bGyjtF*4Z*HORr;fkwm-inNA!*6^t zbB;uR@yF{azCps~#Oao!+i{2AW9-=Md}I)*ii0Pn`o{ag4xbQ-m_N)_Sz}dx{Ql3c z9v+UHKb+j2UcUkNNmyo8=5~1aJH1^=>fAOsMK4TD*_kA@J@adS8N@h6r?e9zUhjGy z_5uP(q39~Ff5WIW;dlJJy!i{K*2lJSf;~ZuwG0K)`b} z(_hKqWNJs7ffVUY@F2N>%SpxXpGhtGLdQKJjhy&yK-QVZ@Mm$K9hhKsr6EhvOMY6&5 zVxELgaVSWw8SYLz3)xQG#4+l<-p<@l*5_VO6Ddz|@+H-@x$VBnq4$tS!89 z365|5v3q^h_4_{Nl3R@1)1$E*uil0K&!Dc2DrvNT^Zcr07I4ky5c{yFZs8?Gj0eQ{ zUv>FNclp$BXLxa}rQ{dAK8saW!{6t8J|-9%cw-{x*G8#K^pV3o<+p)+H0+4%ltQ7NPWK3 z_gxV5NxQ1U&Lo8#=vj<5{ZEq!`eX;L<*9C85;#;cfp$J-~%;8_Ao6%}@}cOTPen;x8zB!7s*!awd=s^E(A6V=Zmzh@HQDZTQu?CpH#N&52Q zKlV`9>FFQIW;Y-1JG}BIaqh=dOdhhtSgwII^eRM#pIXXGWrP8iGCAZEaug!q_W);(=;-JKGStTe-7T_<>l#rE3+G=ot937XbaAR^(E@ zWhpOu`SU%@E4{ql7q*%5(h%FB6o_N7Jq)m$;+H#`_Y7(-{ouv-hZ9=&1e{+Bf)+oR zY`OhjVCIu0*mgqvY3FMS`v^m(n0sI}kuNm)#A_|EzI)U^N{mRF+m2xI3mb>|hIY9s z@0%;fj>BaggP*!>V zcHF54tAIalkAAT4$#zRiLvNPLXT$3rg}SQxR|Nb-(;saOr__4>xhw6>Tvq{nO?A2HZw@3_P3R&q+}8E}Cmj$+5qz zIPfSjZgI=6yV#x{l00}Wd_C^=TPu@+)FdapCLS4jzLMQejGM-vz@IK&$mT4p;nRsM zUzbNelx|nE>>(N3Bu%N5(XgAz{+2 zwfdYTKBYxm#yxOh3#4e64o$OmpNvfjdD#)EAH($iK^K04OMLmNP&2;!&^F>6xa`|y z8Re(TIldnJM>7}hFPa(_II`8i54 zZ|EMnl;9D3*^wL6-L|7k_M0v{9Y(SjaQlpp+>1W~<@SLr({tt6)0vt+%CVrL$%d}m zf*XeXYNKSz;lgcLrBsc&fJ6n-z0O$blf;?uxZ1SeVF?o67plSO@(Ok!9mNS7$ zX#za~+M9H>1KW^-!~1SaYisKgv%KM~ZmVq2^SC+lVHo6OQ%4Kuq;s_uC~C{AE;lxF zR^7?zkFFM(7#UG-3wP}^HgTQR7TVG*dB28Z$Yte?c#wYdEuPv%8X_LNk^5u{++hYn z23-2J&(DxSC%g<4Opt&>^GlQQ$w_H*^UbwSO)UeNMp(ZpcCA1Fe{u9X=na<{E(fS1RZOuFHk1mBX1p=ZWgBy*weyDhGclKq;vfAr+4F znf{p?pO`xk>8Rgsrhxz@9jF0X}z4i62{ZMMTK1e>1 z2Xz<3zB~82ywe29_*gX1)^8y8a!sAH6+1KsbjF35S*Z$`$rI5_Z~lzbvuEwR5&H zTOYKdmj`W}-f=c+p}w|LdrHo?uMzd!TB;v83n#}$LMB=mQfm%Wsjr+HuK0_Zc{Gqz zk6nIwdM6Olz)zZ&9l8T)3Q8a~f<&Y0MYXK$?8-r=4$W1-v^7Hn^kIl9k+T=WEZ=hr zSawy^JEU~PH5h-0#N$UmeIx|$=2~x@Sud-_qgyBu#HY(Hc>RNifRZ3hDTf~P2Fby0 zGr<<`xnoYl<)8ycHQPa0Qy*03V+HSq8AdqluLfC?dX`glAS%udH+>r}_aS;*d3~>Z z)w6c{ZPHDOQD}+lOsF9Z(yJ1K0>q+1G%8j`0~~H+vL*p2l10E&Bd~%0?`{Y7bi*27 zGu37f=o*pxtxOooui0|hO79Ou|F_^kGDJd_a`v!cG<3snZd?NWrKCJnBx+UOXS3{!aB^;56CK zZ6_#}d+Soa2gh15KSy$nhM>rW#N1LrqO@LAy#4u^`^3$<(fu) z64R|SX-ohCjF~En4x1)))8jx`xG`MIX3?l&SPGh%tl%19yeC09eml|~4K&942Sn8De>fdM%H zTT5QYPis#<^A)+GEnp+95|m_Kg3wG*49Dz}eMzn`2< z74IC=jRhYtUKYp!HH{58|GDvX-LJaW&yD0?k%M1tPPHwWn#qo1;H+BjD7ymmrx|aS zKfXwY6rs6nybK3;qCu@3v^Kze|k^F6{O&_h!Uf48fypZNd@s%-eXE!Ln`IYXD>4raDu`17F*-6z|vl?}rj)$H;Kk}3GVU1V}> zl!ORHPK;1XSx4<3!F{?;_^bi92h-fQ9_x_D=$I2-U&GR6IEx4&j|10YH{+hV`Et9T zfcpfwcyqq|1;SU-1CN4K;~R+Y4<&mKCKpNB{O@)NNXoK9gt(6GNVJiwazx@sW04Ea zNEs|vrTGMic_tz4Z56}c8SXRjDJ(x8#iJ~F=NN^<>si=J5f?X*8Zq6wJrj!r7}cbC z1X>XUd4$_%UnqMfgS;vdrF*Ghv{{&7Zy+Cf?6A#7Q!`KkMUr9mgNfO&*!dtTxUUdk zNf~I@DaY^^10>kK5Rv79gwQ|%dPT7Qw&mW4hWElu>J=cpl&d(aXB_qg71jQ-`SQ|* zT6{lx@-T2I9*ev)1@X@O{EkjeBv7mKiHUGLXd|bUi=;_gGmK&dZ>Vf$xouN<|9^!)p5`M zk0ZW5lKnFIOMhnZeU=(+wSHz9_EW~0nkWtdM1#`f;SSJewV+Hi#Z*82h&0ELc*hxY zbMt{mSs+LHWV$Z6a2u#*%ZIsc~9L8??>zhcBr zxWYG9rL_?djDB6TEDQtI`l}Aoifq62?zdRMX-SMD#qnVK|HuWuh6}j*O@_=*$EXdE z?fj;&sN!Z`TK<$O37zT8&*4m&n$3>)TY!`)gxhU}6N%=f5Z6@{fMZlPG*?P)B0(PG3z?HRsJ%r(lzvJe5JX1v$Miwq5-N8{U~T8E z;&n7dxxU({Z08jzkkO+DYnuH8arr(@qUa;&1uL8DC| zkZyh%NfX*^98H19a^q+no^ILke=3Sx%IHpE9;&fftDR)1(xO|TkxSWHL1LR$nebC? z+DWLi-ikY%fKpRaBtWgZFxx}|{b+g11)2yTf)hyEmRa0(DDQRSE2StRsH%EJ`HCGx zdU;?Tz6j8P@HLXw zF6?uS$E}dPxQz+B1&H02f~&4yC@h0;~6CT?3i zYVqmBrc%)1W2VQ?-_iz6a{41oK0!3m}KRB^>D#k%g z+u{tmc-Us63mn1~NT}F0=SzjJ2)PK=qOY~wl@43Dun(0d{r^{vZ6XvT0CMyqAf*|l%{%(d zsqk~)f?t7?Us`&?D0QL3X;NYTDTT0;zx;h_m;Z;qZ~XPitB>=*$^XQ?H8#Do!Fs#P z>DzTX6b}DSz)y-K5ae(BmeP!wVK~=4SZ=oXoB_KDp25!hOuFUF1vek(@_W-N@TKFDP(1UK0+v z%{<69d!FMfIJ!mTi~Bm!F`ZFIHJSuEiDN+oR2DaNeOMR)!C)^?F#T;hcf1n`rX$hg zM#1L0Jv}`M%QLrcj9pf_)2&Jq0tVTwIUBp0_$ErP(c2hGu5P8!yXROHn80=V*|DUM zU-Ux0Kl$a=ZVfMeKAMy+5{5=bb%62ydfO4Vxa4HNT&vzP5J^NLqU#+e z0W%N*6|;R3WN}U^m~6ENAiZfQKisyJyQe0S1CptQKQ^&he{v6>>qOeCC}kA^RrO8t zxeL{HgETy$^)!yRYScD@b8-yYRC0xf7)x?QaV|sgMyQV`1x?vp~k6lvsK{J>t!lqr$>mCc;8O`%IP@)Hi%Me7opqGtS2SjrIkHD5Gze{TOZc@--y86PQldcc%e{ZVN&uQ#=b!lYP%)L;uLnpg4>o z6=>1##}5uaflOom+daIuB+wJbtLH}Sidex&*IP2SK!WK7spw6ZOHYUaO7fJ{)P#5S z7N`O;LK^L`=OgZ6*&xe0O_(Ya-es6-Iy^i~1Gl=mx{WH$e=5g5O6aJsf0XdAC<7W{ zF~`)MQ!*uC+_&eZ6c;|nHOhS`h!;f8B8E1BW&fJj?UviO@!jy*(%0IJ(6Jx-U4tHg z8c6jLSv#ktzgr=}S6Ej8r^%r%L)9t-DL`0WX9#Gx${G-1xQ`5SMwD&~e)8Mf_6{2Ww|%dXWkFc^m0Ttd3ya5hn-Zw47__pnlqY`r&s+0S$br$?N?qmAiD za(dV}_1xqxCPpt0ib6<#$37hy!c$+avj2m~C zI~kI`5rEp>oeR{d${I*W&Jk1IZ8_(vkcpfCxnR_V4+!hX`2>`mL!=L#-e z-$!ckP<9Cz53Bz^Q}ZdS$iCAyhfa> zzI3~c4>Iv_)jd6g29|iuO_jgtE&DO_O>s(t_}QdWaf|->qsjmD9vFjQ(5TzOoNUdF z8#g>qwgAE>bygeSBMy%%?AFOTLGaQ=0ZN48QGvB=_R#@XMDkuP!%&+0p6s_sKJM zO=Y8ZvcW?QR43Lps!m#-ditXmpwe+Q{NneX2-;MxbGFo(7oNaKIX{!I3HPnM!?Nly zd3G?FOOk}h#33N@V1PGJMsJr-Vw0^=PrE9M#HagN=<*@dN*gTgKSrJ`BvmRM7m9Rh>WmqM332a*4Q5xi0 z!W;);Wos(lqHMwIP?Cp!YRHhcudgqfL7G$xtug@9JLg!2;4@V56da`)r6yDTO2b60 z{3FQkIH>%vb6_LwHc&~)cNlXHGm!^F12Lt&M#u#Em+=|JemrHJNq%@($6Vm-7P(a^5OmFBX@%^l}U55<}`0xQ_s6IYED-1Q?hUR2aCzNz69Ay`r zyZ=nG|I$rZ0k>$Y{7&9H*MH(~503*~=_nZY2yNi6KtL2}mq#0Z2p1PNa^ruMoUBN# z$%POd?Xs~N@OM~-+w|9a<-70g^dI+r^JQz~^l5oevezie%@u`;SVRI|yMQ`7xQL{c zDIC!tdle@54G{b(Db3r_^tsK6*h0{bNz21-)30Vmu0aQo7-+K}o z8em-xMOjp}_CL0C!^ClYHCa<>r8V6e9{g+rgD}J5iZh(hK6ZWBr^~3);x`>XVq>| zUJ63>4-*#BBPl#tW*tuw@a^qzT*&Z!!axTqtyTkMinE(JB`GP13a@*O6d{k5Wm-A` z_kr2@y!#gOf8I$j9IqP4l5-wm3a|aFhB*Pc@xT=mr~qPMM{@pmxio5t=T;Js4RVbf z-e~oi@vhEkRDcK(2_K7dhO^rICl>~isRxbRQn(rKTMWd;>>td|&Mq&QB7`R3!@yUR zKn*sIaCqDQpje|;0^TT&&_kc7oqObUUS2>(=#wY@*30S`uadNRnKt|FWZ5)&nN`h0 zqP~u4GrRm)p2gDnExupSQG{H`{6Kw~kP?ng1ucDrX?90YhFhI1*BL2%e&KjiQxmFa zg(NKGby9dQC8$44P}Jlrnz!W9%Hq)|diHELs&ikZycviNQ1AjDnXq8KFlk=d`tD4w zwl7pq`5^t*o|rV08hacfYIVIKZ(b2p#<1e(MKzb*OidqRAvbw{P4b1VnRe|~Kr&cp zOwVI5{$g6h?+gq8W6u15G5OB`q97{$Dg~B+E8JQcPJ}JADeNEav?ynjor|{W4Y!lt z_m!T$=fmTz$YnQjB-jF{D>CpZ%720bxG+#4JlDvvn+zf z0?8zlz7A=)y>bHHk`&rG4@(SS40S3T*k;`Ia(wqw788o~+VX3nEY773Z@h$!)z5|vt;8kTPiE|fj2h*3G13ntO#E>#|EE#y62#21> zFxY9zs2c|=x2I8|IJe^-o2@$Rh8=Q`ITj@U2>@CnSBIY^fC?mHq`($c%_AWf7P$8Q zQJh^%7R;4b)()8FhgCD4%dR6ket5QG-JR*SQ#9vt%uXgO)!HP(_EK8JvUo~6;&dk# zBW1|XAa4O29hXV(*o+7yNB$EgZ)#;Y!ZX~E0~}UyfeNnfqpE&a%Uk2MGBS|$!l2c7 z7)ceJe)5gTxko|M4Xn0i&aEp6+PP^19jMnsL29N84H}hDK;{q2ESZ_D=()?tn;6p~ zYWd>Z4w7^8#G+!7{U^5Jrq_oT1orjCjx2~d6(AATT`#tE)@0iTQM{Mh?9;mwLCDX% zc5Z|xyxWXPlPBGG!;vfn(CD9lhg}Hazj89o^>B!UgXPg<$I8-cq(Y<7#83?_Jo@|& zATiDBsQ0ACpG8wNzd;3B#H?VVaqlp2@o@Av5dpH?@~2)HkJqIfG<)6uGS05thm|Q< z$5x}uGB+-}B?~((JnopQ&CAFxdZaP`vTuwo+g@P{)ADzRT?7{XOj7?wedwyo0PTpn zp-%dwMmg2cJn#3Vv4YYUGf`=AJ{YE2loGph(};G*LhIj=%Wi_93nN(CxOH|V##zA= zHn-Jn+yee^&+dYwbFS5`bB++4p?in`2^g0N9_T#s_bNWFK@}a~zTCnT+1HX!HW0DjAofrgkJP;Y|KZ%k>uHhyx zl&i|l6;UUW114i5PzemuV*iPSE=B5IY2TfXVZb|f?9hkOSLH0wD#z0%FO0oFWZ})` z3lk%#CxT<9LMD!It(+EwS>b2lR4djkdqPf6?|d^i2Xd&hCmYYr8os&RK^_ct!~~QK zH0EL3cVyecV{yM!s6#a#wa-+}0xx8>6EcX(+03gjpxRHL9waP4HF3X*ijJS;qA#QnWBvq*fpf06(Yi`TaP;3pb#E{6`>Ub4E^? z?pb=ii!TTmOuf$e{%RIt?m69k`Xf5wsg|F2{W^#w9A3lsNTB=((5t=#u65giv{eECe@G*dUes1Q}mKYjf#Ti`y;7FbyWE7 zQHCy%a2i|DX>Z7b16?@VkX+WBY7kGjUJja?QGiBe5k^}96ox*CKKD`eNzu=msk`KCjABG@mw9ql?YFFo)_!xj%+9V`w>@X9> z#uzbFA{8#0IiE%ak4b!bdy8CSvJk*p)O>Z7vWdf)Un8#(e0?eE1%jSBaG=ndWdTk~ zt>D}OWP7hJ)@gSUfW}#l+K*>R3Fk3$?NqOTgf+kUl4lL&xuD~0-8?Vb(*2;l=a&F* zQ&jbeH?qp{7u-!XL>Ij40R9j_0@Qo<4>vu5(h1ivw0MzZ=ki@@B*x@{!q_`l6Co*X zDT=3?Rs`u&REfGu;fi!Vf|8I4la>ZGUACf`kLGcN@FUmcK7X0FRux5JKr}p-+XmQilqYH#(8_Zww!$T43~*{dyW7MK0KkOZE7gJ zX1__VWiE(*!l_k-Q~U1_bBEHZh^0)r364`}7o2+pBV!$dg=3(F;xhDu?lsjdTK4Df z-h0_GNPJJ0!KzoFXzmy)YzUp0wH+Pd(c!a{Wv%SRUo3_mopdIrFd$I%u>;hptxRqXe4r`| zESnel=Be#C#Pr{l{>j?{X#>7=FwlRJm6bJM(xd^pbo?hli#uS_VG!JEl(xdu=f7yH z-IS)y8$6z?m45v^c(i>QMVDwbZ4@SC=TK(_N_AvIg}J)LuS195id>jZDl%Nh8y~7k z^8wS2`=qAsSNDy?UzJ=|lBEs!6~m%Q{frKu5VT0p;bAl=qD-D^Oun{h=dI&)RI3`v z;EWnAH8lPhh43dCx>d?jcF+Tj%EPd>7H>8%l|aXCJi!HmHxo)G;|bk~vJ*BpOFS4X z0D|l{TDGVBB2bRWnP1OI(%EWEntdp9x&QtQz4e(F%2ncQR`z>V7*%&ajW=G-kYEs+$BolZlkz!)@UW@Mb!($eyug!ZnLS?g~;c@3pa zc=gB$Z&z-$J=L>Q)%I6ZLLY2&&(9kaD&Xs3m7>#h{8s7(NNv0oQ*Adx#beAuNv4T)lmgd z`7NC9=2I)EYEywzoQbS3b9s9GTe&=U2@^f*Ll~&hBD0YiYxR5k(PiL^HflP#MZ=+S z)ihXDG0+enFgXQy90E-?KvAyY80!4@o)^@xp~bQDM%I)^%Le#Q=tAoOl&#MKae(?Y zq4bmFC>&f>s#Y}Hz$|!OzM(B&5>Bw}YlyD|vaj9V<)(8%OG-SCnIoKL;0X=|e;87o zO{}L>F|(wTv~kvV)3~l))eR3GuDzVD&{nu_7381%O) z@v1Gv=pwQvzgKb;9D6nj>-`E)VuXH~htbRI=D2BPwh(Yu?X{z1-pJM(?70A2SL%+u&(*ht#zdgsIa`x1luF8ObNrda8T?qO0 zkpV7~EnG%!2hLL36QNHZc+nuB=ZJ@NXXJojxme%?o~Nvn($T3}aLxvB)7}nX$@g|F z)PtrwbUm0O?5C-zNeooG?1@r5)X@N^8$HQ^zCc4FKm#DnymyKoDf4AxH7NJ>fgYPu z;C>8%q)tW8K!9T*7O$uGhQ~g*I+iU0zqSug7cmdAi*mD;9VTsuG-_I&?fY0DVAJay zMowbTP+WEZX(wyrLxSbx)Pe2abD_*%Y`Z;;9ml^DnHR#iROmB)oFI8|2=XK)9>ypL ztg?tql?~2b+^mQ^woqueuA|;^7zyIl*?~GX8uH*ACFKykP@@x^Ef;9lQJ6H$PfblV zPE6FwE(6TRi1PU#-Tlz>G&I%O3Cj<;J6ej+W2UVmy5eA3i0DBbAlHFpMo%=He3PBC z%#0gSoK8tG*2MbZ7y28UTSF(Fu+0s;>quRgc6PpnFEU(hO8L(gx1L54x|osIR}_P$ zb-*N-LRd%!ejDVc&A?pwyYNCHg$g7PWrCw?LbKsPm1v4=!2W&1^c$cD{%orvCDVpq zy~x!CHM_{RLLJ%%Fu`6^Lg7Ej?RXZIik|3kovG3-qfB*jpd<&{b5xC5CKG|*Ky^6u z^aeK@`yo;XPt zFLm(csAllQPTXC0qrCbvPKqz$KGv3Mt2)6_6?4duDf88wRss&%HVgN+x=JryKmJ>e z@pFUw&1K&t2jKWC7Iz;t?;5$Gn#kNKM1eBEBS_~PbwQaMJyE0)DQwvB+D3K`_P!0a z(tlFL&Ge%tRJRaE2QD&>5CQpz^MFcZoMSL5b;-VlLk*@f`EsE!D=P~%))0fV%sinN zXs`NYPbQF*;QE;mVS>8EEh={3E0KYI_z*!u1#hUl4>~LzGES)f-5e7!49~)#Cea!` z@rbRpZYwz!+H_*Le7rhXCyz4l7v)3vcN@9E6j4XhKFQlus5GJ`yZwWHV3p7T7<`rTOn+BI?z5q} z&IMPA`ET!3M~@vnTJrYJ=BT9V?!&h=Ahi0AfSu}hRXn6n2RU4)zoKGKJ4_R7Jn*WD z5Of!yE*27G;C`qz2hoVg)uFzsL8ImK?G;S-BENlUIX^x_NQfEwbZ9tGWUM#b;&Y^r z*TXy7jaJKk9DUvPB`6Bz;$YL3O3CJW+~!FZ(4AOb z11(;t11pW|4zh1W*Y5$-eKPnRXhIU>pH68gF5}Z)aYItfJEzNOFv*^2alUfvV3h@x zq63v7tdMf`RZJK~=k!0-L7hyC&%5Ah3V7BanxbRZA!OwCsu>>E3Ycs{6I5Jq=-GIW zOe$X_+)R=S7+qXkDrYtNM@D=d#(L1B0S#>&)5V*JsHD=6cTo*GXP`Z3vn(p}v`N}k z`tK7ktPVp(L|PMW+Lc|jj?gA3hFU=GR}O^y6EL+UFZU+`X%Ftl`z88G$~2X|(9EV} z2^Wr?#w=Z7Y)rClWJXb&Sspl+?a(rRu8sS(E)z`r~M zUHT6| zKoOwfr%^BX&ypZ%dKv99%Qw;cR;ayKRHaL`419r@eI(kdT<=5Tt?K<^a<|1zmK>9-m8^OIAJs zCl@^jVH=!i9_%D(vBSHFrK^zzA*c%@_P=8i#8KEB0NV>pD1ys?tMA^uyJEimG?K*L zM$C&S<~K;wdIIgC@>B+g_yIut-@QD3vog^aOp$pkGra|j^{&(K?Pd_+FOJ=oc)wx}lS z36{f}hV?{rbB;MH%#L)&D%u)}sUd^f~t(*LRPd^v<95dj6Wkq$*Li%A46bKUpIcWqN^`p^*E%AP7NB>X-1o zHv_`^Hbi+Wu;L2NbCam0My)$Mx>}jJz-j8O>6$li(5y6)WiCKQg8b$W0|W7wXKo!_ zeRE&6MAV@EGB9;7FaJO<(zPoP#Wo+_ONu$wUos+SzC4c#I zWz+9(gZNqQ*S`|R?_jnJaD35o093;HHUp?kaRzt=iZjC}p$-g8pJ;RE(dkscOD)I% zyD84Z6Lc%f`h)~0rxuIY9O#6Cia_dI=2_scys51X_`GsH40|e&S_*xD3$h^Ai?EgS z#*5_3wyd-2o@9$?J?U3ekiT#GB~vouPk>@WWR>J=b&teahK9EZEK1;nf-hSTJnXv0 zM(}vJz7d?zZhoXiX#EOd*m%+AD|zCJgEB z=9!l?ymQc5at0^oG;N}+Y!srY=|9Cv-*C;s=9LSvRiQ40nzz{)&WpAThf zE4}j-Jep&Bp8%4X@u1id>0Q|RY5KPz@%Tc;&rcbec{RFmh#Ja)9W*LS|0vwMO^tGUd7#4`%nICsk3{dzj%q|mVenj-}?L7 zU;ndbZSMA--_LAb$9vjruS-ow4fp=c%$R#WRXfoqI;fYEgEBAv*cl;S)Hy$2m|>pW ziIR5@x9koz@Eoe)Kb*nw+@ZAHrE7lopE`|;4gsm`NKB~1)dDYC5<985a6^V5X-~(> zl`FSbL5G-XXs0D)RRMi#R;}WxGB=rYd4ucnSpX}yA(spLae#mZrqdr>-O{B?i)e28NNP3gfNosiD)E4i-1=a zP4lrfGTLlWD4I=HOU-@ss7&@3sjVn>KzMS+2_Px-5p+Kjb=ltY`-zy!%I7S&Pr?B~ zPXilL(|`nKuR+fT5`H=0Ib>=li;%%A_xwGZ)~r$}*sVvojlX&m%aSHf#%m?oSMY?C zAomj9O$UF%{UN~T#7 z?CHVF_HVN0e()i=bf!?2m3+Z~-|G^eCqvkTcPf{WAo3zjq13fu70Ia6T9f3UB^VOe z2RtOcDmziYg)HVDNV^2|?b~aWk``@n0on)@spbD=Dg1cN83`Q+!~wJ<7(OBvwRg?c z$$$vyXix2igb@@sd~yaJhar290vvTIXk^rFT#ZC_P@m@co*@Dkoxx~ zra6H$m~={c#wG>`+3;k5@2^Nz@3fN_J8NM~rPY<#2Mp=p=-^955~-q;7oH;@hI z%9%a`8!^`e5g;9H&%b!LNOkoo^=2{G@8{XApC>}}{;m>l`vvH?HF+xNpHUXVfdQ=a zxc=10QbvaE5(70h+r5$zIp(UfZ_AE0h%UN#kB~=dRe8@1Alv!6t%&QCjhH;!$(^g$ z^7edU{W6-Pq+d+oa(n~@eOtFe|LyoDq4HRn&+*vyP<9|2La>zFzEH4w+*IrI3;^_A znX!16IydHbY(R2)HxQv-CY8@a7vrTi=+~tHw?YUqDKkQDAqV(id&6)<`?uHY4=Q2W zKK^rtoH~uHLGMqMTyt7g7Ff`+i6N52w|kiFYujADJhL~$mPDE{OZ*`&h(KOiC z7Ve1^{cmlr#gB<^OVrp2kAZysl(KY+_O_IK?%8+CDam$j0*Vt6bXr6?AWVTKfxf^v zISlYDZ-&3*U|bF)K17F_pI@KE@p;3i0H-03m^|r3W`W*fxFa-VYotFNH={BfShn_7 ze@kz=yM}IqsJv<8@sb|}mPI-}33ga~5{zW;B>LJr9^Vq;u$0u4?b`4osVt1R1KKl% znfW7b1)$$j>na~S&2{p^3}1*6Jq&Wp=DAF_nF$QN15AW~pfHcD6h*OS9e@&fFJQ5x ztg&z+6B~?qspT(GqXq$FY{R(sDz)Cy9@_5@CpWMbJ!Ck;XAf!LZ{doq5-`y>))0Gk z+J~ zF5;1vy1%(ba>M4L4b!qHU+`gMgp=9^pl>PA3`TN0|A4bQD47aekPX^;Yq5RgX=Y}< zgX2!<2li&S&Dqj19-M8mmpsyvN%U<5Hse)%dsRKWTT@EE?7+YAqG)-3-lWYYKRQL# zg>6(?>)U(2{_rj)C{aheDe+cK-V)MqSn`U8)o51w`p;L{zq5*Zfm^2D@~W#8XSrJE z)i2jR6g0g*Uv_?ej4lHZT*UT@LWqTptMBhuOs~*_#ID`U!NH+4q-eI+d8W0~Ap72D zwH@ob5)xCqFcJ-8ftWu*A2YyF+LEns?=Xh zBRyr&Mj-h-BrHfN?|$Q-JSde~KOQ#+P};&hSqqJLLjBWDE;N6^HL-K{e4owZ>g&2M z1EsoWpS&7S`H{}Ha`kGzOA}X?bFqSU4|#uIuE_Pm!a@`e7~hYQC_-)nN?6qI$zXdP zG1nYX%02gnyWVXyr0`P=D;HPFqD!xRer>AqtyFr7@Xq>EB@txu9(UR^r}FP_kofuG z?QdQvqB8FX(y+SB&<~_;WtoVHH@eAM;9Wyc?H&m!ubT|Jt#iRQhO04gZ9{v%%w#SK zOa+)2I@D_tnPsy{Dim;N;Op0mx1Bq<3oJD)N&Y-^Ho8U=d)|$^QHH(>@Y??8@`TexB?{6#TDl-z5eahyR-9hE8 zpFi(hze16^NoUaIcw&73T=!+Y)TG9C=>7YJ=q}hk1Gdfz%R-7`zK`HqE!@E)V&e(a zErTzrYv2d~1_6=nv|WZMndC|v7p}3{2nn=3l-?m;=d?Ia@EA`YHa6|q1=R>O1;~^amBT%SLQP7 zV8?P$t=TGR{Sn*4zP$@I3?)_lWe1pC)O+dsdhco1T> zK)K;S8XV}wht{^ zw#=PcgXMaendTaK2*?Wt#h+#ujMz}HJsS1M@?PntojE*PYB+l>f}v@JB!9ZP$F*&$ zMQygTlhgCj`=nN%;k=m}T~AV6pH1v3&mlbx$d8CG8GY1p*G$E-rgRf^1e*hH9rYTc zs*WXOJNr_<`K~kp>BbhZ8`b-bfWQ=y6eg5}xa|A%BH?CjDLvud&zDUvW9or!n zgdlCO=6Iq-y=XR=(5b0%*$3NafMDzcIPM-ru__y^8AI3x9`=pW{M^Jyt1&1B;38mC z8HhNB_R6hAtBoy;1r$+1{%-n6@Fowo{2TXp5E>`kO)SN zKjL0p_N*d2&E)!I<&smS*9$F&Sy@{T?v8bk3`eC2wWAcb`I{Y<8f8=T$^VWq5S#{$ z(O-CGxgHYicl(vZru)URfhbz-EODFl(S-(>HOOoefI1?ky>Gp&QnV6F6+qPvuxEDa zFyyh=2^p1~Z1XC^!VM+9U^57IOqGeF{!Dtx+jCGT=tj%Z-Ur9!dRoA1v_|}nWgvL7ugKbq1+{Y|JvDu)@@sc{*pQ!!2ClEi3BIKu8}}g zWVRvdL!@wZjRK$dO~eq;1ZbS_R1^YlwF~IPssXDG=>b0t+eq_XoyxS-1SONaR}3|> zvm3be8o2tov|i+u#aX+hZ5-1RW?gDJ7O=>*$u(SOe<$(qjLgt7(w@Do8|1=~PDg`u z7$QqZMi5~6l5c@)=c$SSW_Rr@*$(TQphv@30fh6wZh!ief1t@y7v%_<0^1?$nTZh! zY@<5Ujo&YoT;hxV^Y-mJF%p*Y7dUQ@znHMp8}F7rkt|xwfA;$~5u|SCvHFl z03?K=y+a@tV4>f~@QqIcs^GQA;jHB*Tu%-14vO# zN5!@>=(B)zE3$$5tvx^`jiwFY&5nmK7Sxi>8lnj^amPzXd{a36v(f&-nMBCl8Eyf8 zF=|WD$gl+=HgjAwvwX-Cep<)7y=rW13~?YZRsnrZj*h&4pZ!f79LGWj3%i;0x%N5J$6&H z^?C-A=&4%`lSj|*_Lwp>qcK}AWP<>O!h{A4*qf)4gLf6xf0~@+s}c&T&f%9ad>l@9*77SDpN zqj>R6CcCa$gYDf^#KJ3eCzHtwo@`}56}v@EXdp*N>yHb9_>Q+KwLEkNtSDD&mW@)t z9Q5h$h3=Wq?d(n2ke4@_Rs&OHXfH!{_fWH7GAGNqTfhr*!2V^H6hEnD-KFPuF96AK z(f6ciBmj#|{kx}&gpw+5e1-&+0D6q@Y%d}xEYg|BUU22g74UkEZNNy`i1>qn8tVuL=R4ezDPZ9a@j0Y;<{*0lmN@^7l3TJ` zN-lOkfE-QG_I#ib2o`**-QHaM;r_Y*{!}sP?db_IzIKKAw+)9sJk9evBWUYv!@v4L z9mjGvH}c#$Js0|MriXzw#k75MWD`gE_9c6&zoc*21!*y#qMr}0Fu*dCO{j3{O!_Rq zWkk!0tGv+XyrH2X2!rAeGW9$p8@x8{6HTP9+YGNdcSbZGWk>>_+ZOT z&xHmk!@Dw={9@-l#=pxHUNqq6U7rB}g?Nf!d3Hk()AR9r?t=5d2?9^Bd?PJEB1J*N z9R*SVN2JfzO@_Y77-ds{40m#~2I`=>ud1=cu2 z^nsUM(a*4gKFhshS-LKy3LX|5+8T9{2DoxW_DXa8#kk|Cj&l$$fresl=k`MBy4=Pb9A( z2~RO#)o>Ar9VG}}0E(8zf=$=`yt%zJ=hK~uBrBhzxonitprw>!hD+}UFf1!^E4!2Q zl1=ru{FRw|juX)yvgt&~5M?V;brlD#9FQzpW&R2D_#TK9rKSp5xC>ZhIf(Ds^9%R< z#~UEI<8NVm=uo_d7bcVdk(V7{cfN$FLCAhMzloKMBn=8?Fr~U1H=q$IxVm;M$9T^) zRHctILTbboO&~nP3IdLQVXaMaXz1wRQh$6*F}Mg!m`QW;V~~&edLZnewTw6rR#^^C zC^?7%osaL)yAEFnr4J`?+7yd ztku1=liTNLCNm`FtYQ%1y{hbY{YvV-KeA%;QUD*6g4@A#UcGv?Vq8}diU!2)f=r+T z+G7Iy#+uT$!Qv^pC4)7SG}B4cO7=>18?iDDp0q0%)J*k4MJC1cE1gx4BLvhm6}q`B zyhLS}Ei8u4%I2>k0xI)f#e0ccnVs$S>g`98irP74+8$o5uvr?$Y^;r6R!P35kP2IX{6V>;Edw39i8ci3qloe7!Ve!AH}usWY_Z1F&YnI1>+D3t(3E zdP`BA`IE`HNa4a9#u--+IV%QG=;o_fRwh3f@pkK0M=Jt}jfjG*&J2#bi^ zS$MsZ6HzT)+6DDFrZzl7JGTw8d*!_IyD~LFTNjylr!tR$mqOfu$k*5p3vDlmB-0;v zQZx-cK0I_a!&sEXsWPX(g0>^a)urt*p!r*n zoeS(++I)H7Rz#oWWHY} zQ5pN1qn!Jz^W{QuEk$a!O>3}7(T0_4)|7+JU&Jmlb{D2!x&$QMaEw}X2CYomml!_+ zwsr(|N-%aW8>KFlJWq+aAl25oq}WdCl8SK3f#hvN zY*A{-ho8}Bu2QD_-X|q8TqFJ>r@kODcyR=%jOoJwL89_~KyUURP`&4^QoTgJbV;o< z5LO91{Ez&pAlRghgr3UfW&RLN_UwZ-SYX^|73d=`vZlzvUje=j`*NcLvd5Y3if{y= z)_y7QVMO5r;_iG^oq(p)&YFcxPe&IP4HqHP5?W+~p^~W!qzxd2@>a?M5j4Gs;YJKXx^hBu{{>MycBq|l==p>xgB3$KEV@>YG~ z`E`x`Kvz)=ts&@oh|5)W=GYsj4{JSvD~r=gtQb$;ZlioayBHwD=gi+% z;}8Z!PM=0=u++1kd=OWkE_;%pf<-)}(8DoH{6_fyj-6deyFZ|QQz#vzp5Yld@~5f- zy#~%EpN@=>;Y$Zp%K|`HCAf3=f$MlhG!1r=Lr=c^AP#S6lwa zA)(z5pXicxZn8JtVHSrfBZl_dA>jz9(y-ox*kD_#bE!ZifJK#L5QGRY70*AFybyYlZr^;%t6LgMaUk8qWbtZufga;Plu*|715ambO!TwqOaSB81WnTjh__UkME zEJ`Wkk?mLqc{r1t`Sq=xtIi!#eYD?C^arE+fV!D{06MM8x9b9eKt+0gbvSzY>1q{Qv#P$#rLs4|kxQtEyeaOO6q;9dw zcO3Boox~FGw!j@6o=TQQ-Z#ugneQ8eC^brNzVXjWzGHWPStB!L)I)Vw-_M3R!Yl^P zTL!Q$&sW6&S-k5Flji3Zf1v9nJ5>!PwmPrsh0G-N(aAlcEUB{3m_CI)5CH*7kPY@lpdr&50SXTbiBhjjh!cUU2da$Bi00kwO}%IIPg`PS#y9~q9Rorx{=|iD@jH* zV^csydVXFWjs&R~-;xRW!Dh(7iMv}v#_$&URw*iNfD`+2ZlkqWKRCDF9mQJ|i%BhWn3 zn$1^%-Rl88Vu|2gr+S6R_jb{t%HywBEo5}oK)n5;`Yg0U9F;T%z!jT(FUlvP0Ma<2 zh_v9ic2*6ow)5t5YUr9?emy;9Z=^I}UC%D;DWvvQ-<9T%U}WeJH}RqkGkli~oIn3d zZru|&9m8;5l|$*T53YezE5@H%j%_GlP%FkaN%1$f10^LAO0t!9Vkpy#YA`eu6;8K~ zy8%+7d=Y8Dis;#Qsua5%M2Y|%F`#&9Z)cqKSKQ>>NwJ|!|LKj8*T`I^bHCT~4?-?A zjA<=i-%En_^bPNN(S*96z&=wBCL<(XX)6YFwlpy7CN#(c^m?dS4QA&+GKiil}uwEcd2{xUVY07Zj}_5EG+`I#|)C`5r( zCe#gKXTrjPre3;rSp!HF2Z&&!g<@#uF*O5k9I671T=#Zy51#ILcB#S~NBRuRH||6Y zGFHI)B$24D$VkRu)r|pI>_GPV48C4?^gErTr8($ha|fA)qQ`=6p!e|n3y@7e19S{e ziQYTvgtA#Wwkw{5W^P2`6vx9VRLRqCqiHDfSL<379e3&aQM-2R%#pS{Tn~nZxHU#ppyon)AnYtDW{)}xs53W=&Pc*o{TXN z^gu1yumn1pB3!lh=|VYUd=qy5EQm6Fk~ggzr}SP>q6D`ti!2Io2)jM^dGP$MyP^J6 z7Kl6-M|pdN^(v9|qbE4Dg0v?RNhicz2K2$+fsiSz6o$A2nAQNP5of^cibwqlFI0apA_O`$0_m?945S>J_nKWL zbSD-5(B6}2m0BY=mp|aFW)j|*L=rVin{PnTDkI23kTh*V0|hzl^kFkRP=xGiuMpI= z&*z%$FAVndL}hOx=mOdy9LA0D5z;8t5xVB{>xPDnKeGrPlj-M^r74bYsIt`vk z^gdMmLX@+X@tj3d2M>DNS5%DPX?$c^l0U}bsqD7u{8k1Q2VpJhPC$LWlmA-8gUqUu#NjkuVaCMgixjl$?rb?%TZ>W00&}MrjAVAThP0K zXb49JKn7jWFOrA=K!rJScoqcsw;H7UExg5rc9|iKV5Pj?xpJdnwghHRelTI-x^{uO zv0^A0j0BKkFA~y%!@y%myM&fj~w7_D#ppcdnrLEvsXM1Q9>drA;Tn?iu6Yv zs3YF94?G#@E$yY*dh~#pYFfYVPYQ5iOqn3VgOC5JwqLlf`LE(B#F$7fz#c&E zTly@hx{Hm1aGVLvmk?d9ofUufew89s5r#70V}Lmyz#7y*zVOUwzL3s5*kCpyrdjh@ z47=w(JiHoEa|z~6_rA+S$8GF-~33^nm~kpXtXiX2}h zUoxtR!w8AM-c&HeK;a7L6;D7#2AD?_*jqH9Y9pUnd`W=_J=QSNZ^SV-&-BZGE{#{o5?$#&1UwM#*t+ToTU!;AM$P%T>IOj{DIMg{%hv&x6O_3)3on-nP<)uLCo#EM|7vWwH zDJdycRC7ZghzLMz!>&iiAbI8=pdm*{u_jdnM=L?AN2W=?`QCd0^n1UpK#dL^M?%*h z3SyP%ngf8opMer6PiHfHcoa~#c{&-8j)$`o&5|sgYz85cQ9C$kp6X*ZzAW$T$Y#?yMBgL)O3>AY&s4@3Q zfxd$QkVQCoI^&mt1ih*;oI{?@{V<{KXd%NL$VnQ~F4}MuTo{l^eYPdp()(^s9FJD! z1}1|dT^GQEEsGpn#ft(D3a1+)O#d z4?5{oet(04%cU@0d}rqS<5SH~kb3qHMvR%-w@yCurC`qsYGs6;14sg1Z!Uo%$Gn3UGhFZmIAF-uzjiuKeTU&^rFZI(Z4d#J0Dx zjf4l;Gm4ZEe`0+$e8C>Ni*ZC6a4Uj`vEDWX9BkcmssK9SE)eP!8Ylz(JI;H@9!Im^ zDTnY~>GC|p8eP31hJpj~Jg5%;3{YqusUnxjQQ^yu#(ru+X75)H% zT8iR3BA~|u1}4G-f~fw~cu@#H&^)B!{J@{XR)K!^M<)FTc|JfN6ZER(?b5~i5R_e> z?a~eE`P^tI)Y>QwS`468Bv8pf2q6WXZslseODnzpLbJkGFyz29K`z(iZ4Qg;eb&bb zixig}Fyl#TdfE$vn?B0gWHp-+st4FG9$#F8Fu`vV@@jv7bi6hk zga)#r`_Tby<-0wrfZfL`{*Pizp`K4ayqLY-A%1zE<>+H zBei4O_fLLD!9?}%ga3JP=#W^TA|zHrK}wu3(s&f|HeVb+_@=p8C*2o&>l`7>MK`aM zbwULd6hOxl;xeIuYS5;S%+1Y(bJmGkBc(_&oq{%Lt#n_QWmbZOM4UhgY7_w@ci!U2 zutvQL$`wvHJ~O~hj_0hct>2@AH?XDDAh-ZDpl~{8q=G+}X|qzy@g0Qb4syVBCIDzh zl`_+2hV=R2ILE#mF$05p{_GJJs^RGdW2xSEpXTuN`M2-7ESe{=>YeOtSCF4`tm&Ty z*3F5+>&i#evVW_=;e_+*{022e&-QY+PQJ2zy{|yw`=vR4Ih=jMn3r2=l6^(^X*7sLPiqq>~mFn%5m7*Aa;!dL$OIPA#R!QE`oac*?VVOr0u7iJg632Fj zAKc21VL)htSuO##D8`FCmZUQt4r7%{Zf!YuwD2Gs=u+$Bm?)mk zT!^uZA(7p_!Lu%uz;;JFlTWpNj0m!C zd|T=`BCj49|2kP>PodEtThTx=GpR>+o4k)FqH;W)SB_%u5lBey80y0|5+wTSh*zEl zCOb+O-?6cSu2Af%OdnzJ(Z`$f>dm{Rud+vK9~=zmYNPzF@X9?wnu$Gf-O5SBLx>TQ zoc(ziz&zE81DsI%yoX8)7)8ZO)KEr(R}@6@(YoxlI93->M!!s}#$Z>g2)Buw?T)R< zJyJ0u;MQ@rXJxu56p&#ylr`L3G>)f{_hs9F`$^FW#sZ+WN8ZJOUqM)UDc#_$cD8oFk=O_V z#0Co==<20;rrmi<*}E&9vJbgfq^@e(JU?*TAFrU=91$6qvJ{Z{0R@r6`H ztB^b*P1M0GXN!u9mqR>%DmWPi25-m52puy*l^uMaA!sZ?B_iCmozQ%Stnz@dqISp~raDC;C;ub79+o`}Et0 z;O+0Ge*ek^aBw<^gPOtVN~GisCZvx(aBQ)7lWJI$Zp&bWj&n{lrSl4Q#>zOzpIEWh zzAW`rH!WLAJ$BJu7MIuF#=TpKL%_T6E}yug9ymr(rbD1@L_jtms8yLhX)jGFS!Z8& zLZK8)ux4-pm)Rb-t=8+l*aUE%gdP)A5YmU;^icO4xH@HRxv&Qsq1X5> z&}?+F9;%H1eHbS}1UqDaT}7=0gbk~!2lL}*!<{;Au2YfNf}FkdQJgD+9b$JJ)_s5d z9)stdFW4NLx3(GdiA);tyql9F+Hp#Ei8$mx3tq(vbtjJCT|UOUbT(O{aRp-!cwm)T zlTcbI0cHMf&}=bo7L&zMPo=2-3UVR4DhhR==Wj8Y4VW9V3HFDFUd0xP$BKn1hLcE#oX}3FfS|fFFu(#u zhC$>CEdp;dLoEC-@AFuwFE@A7FP}x@mAt40<&CA%bghmWPKG20_eybjJ>hTvKF#-l zwqFV+NPVImn?#|L*?4w}LII>d$TbzTJaE}ja?N$jrPCX3!q5*}d==nqR~9OnHpB3r zZbWa=-5|0kh4w}koWDh+WSxZKJ5DCXy#e)2-SGmt?ZZGl&b394K0Do5)bX&NB@tvZ5E%fQ65BMgs{VLYun5$AzRl=WP) zblLV1hFgC)J@@D8OkvZr4z7ic`sO7SvUa|vW17Xg?HJyMzkC$GxKWMeP|p{8m|M4m z)T z%ZooHdcSRJ<6zM{_uJsgov&bO79xB%0g2{mHiQZm(boXlriazl)j2@)Jq_$^Ddbrk z0KU_Yn4=&y3i#t(?F}gH29fCFL-nQkNX3Pba7y7zHjF`t=ZVT3IHN)m_3n)kK8I)u zbJ1b=e}zq6+(J3NGXVVh=#54dwY}=6rU9vk5Klwp&;yXMJk|ApS)DWw*h)EJ?=LfHOU?gl1+kEbUjTbNeC{@Rk~G{yzhK5rl?%R_@Z z+&C{FP@D)gVLC#Axl-i-M@B)1!9e53kjoRpSiARvPpIS9xC=Igm{<-PgPtHr6}=k` z7D5X!3~1}tNYD*mwuXxX%F5pOPC&r8!t4q!U37GGpOVhL{CN0Obt{i!|!vdIJH+V^Z#LEy!EP?TAY(U0RnysI8=!Cu-qRC}J%}{=OSVc+pWTL-p zVTgZh=d0Liom7_2it>RXOTr*R*V`f+`0xh@jbaCxTm(LLs0cccNZg%Zh`S(5^9D)y zXcgLVz%CC!;z5BB;IRtr1xmSq;O!ewDwn*rA``?CO0ZKG>pvieQMaG_RF<_4*4Q1& z%J)J70T)|%TXJc_QFu-`^9KVz&EZLRkokHBjS3X#@rzMuHbROx8X01cH|~!wFR|ZO zF{D=vd zu65es!8-|J4uf?#c4ly8LJ5k;BDA7ru8|2q4lly8cxngN$DIOD$(~vzUS51*GPw}$A)gw3a=dy%twn73Mrdg~7XDk%ew?%ns#fv)5^6zZaubSX$hb{Mt6WKF z%(m;qu}{_j55=pC9mQ>dk>P^HX!?&&ix`XgWReo(_7L!?J+**Ci5{ny>KVvLv zwKskp>EVuMFuvBTFeFx;C=!1H=jU)#b3gh-=%$F@7x(+UW4xIEVypq{Sql^u#6?9! zMPpR1_{w7#j#a;5sKLSYu|c;sWyFpwg_HE^80Cwt|4^Y>+xWB#BcPtxc=$fRFIrhz z#UqOy#4l?{0LfkZ-SUauhumX74oh*H{-hYLwxL>bC`kfpJ8I^~OVgXJQCx?DvpeDH zSw+qs$H9TZXJ0?8R`g}_Z{u5l7@CgGwMYV5Bb9@I2A86%QyT^`URbs^+`bP-H?hwOqPd8M!jDn8imyOU_PF%r(UQ;lUgjL#Q;#l|(?8z7_LlkHX@eIgU z`am+P8m{4hZ_DWmwTdLu=NBB)rjq(bldc=f#mY`QrTEGgXdemBF!0|s=0oMS^cI*% z_CLlOD7pOIx|z0M(Kiigr)!d&y0NYEjs{>`I!OKZ*SjjJVxvo7)>qs7(UJKs8+bM4saa8^U$y{@@;8U=?5Z+`D3Y58^BF z7k0jUaeG@gD3baSF5`(J7KKr7)eHNAaA0)Ze$DRxtW)uXP)-WwUoP0Wa)%CCzqxy} zdAVrGZKuqVnUQ#_hOr42$ulz#n(sc<)gN5%p_+8bJG2goYsCs0UX9fJiO*c$J4rdw z!B;G5e1`2|KYibk@t>ii&f%W7jNI%qLpWL1Xp_gB< zjR_`c#;U~s8v;0l5K!b7vd72OSCn7AesmB8i*Oob9sh<@csQuWNr<7qyW=n5UitQV zS2R_(QIt}!lx=I6qQ2UF0rzY{_splv0O+tMeh9|tW$JPKtC*K4E7ew*?8Uu>qB7`K z9aPlN=VEvNO(?@kI;>zP%nbHkp^DJpooGnqxKRWNv*5j(>;mLgNMHU$j9Ns2;yIXQ z8m?}?_ueto$;HlLR^c;;7EITkN-TV$ELhF1g$MVVr$o1QGNr9Ge8CpPTWmVyu>SAB z4ieok4+L$Y+_?@f2Y|%7{igMRi{Sd<0PVrSgzi7I6AVG|6x&OQ0kvAXHbIiH!>gq~ zz0Mg*$v*@78cN|m94dh+wVXgR@DU=@Do?1Z2TI0ng<%SUSoq zQ9&rN!$2@rpc#5WI3eq>IxG&Tv88x}h~>h-ikNEtT#TsDRM9`F($~Xs9|>PNdreS3 zuLN7y@fNPGk&)PYPnp_Y38Jv2P;TPFdGW3=b$sC+9Iga|^v2*bg*Go5ukJNI9s2wI zS!%=CksVAdlF_=i2pXLTQ-<8Vl=9+-rjnJMk|LvU@z}!{1*6{JCeDB)nLQ-U#N-tH zyo-wAfAZR9REc@x(L)74jJ45TSX>?IC~!g8fH)D5E~@pD$bSM(&yJ@Z`-E^_k7TSh z=31D_WA=d~SnTe|dVhcRNKI4q4JKzA;;uJ6c5!f3866O(QhaAxwi+_O4Zyz0W6H{6 z4ogVu`9h$cI$mj(VGjgD9BA`p#yYe*bU1hW*6Bnv;qY99fKc)~Ui}@Jb&ru8hEzD; zt^`8aE|#LH!|M}fyE?E?i4YPV8VGeCS1*zIL)C3iT^I_+n1CZ>od)#XvlbSL;hE17 zRz&F$RGcRsEyF#5#ZfrO+s9GIYTTddTR)`MnA90b9`hWzY|{0rrapGD$1qy=1{35H zVPCMz9w;f@Qh}|{0V4tsxH$<$V02UufQeG4!>tJTH>#?uNDf3#eTtP0mt5YE2*VU{ zJ{yM2s62w12NI_Sll)EI6<@Ij=%*Y&vnUwjls`nQwh=J24hUy~P ziSNa!GsyAkp>Ft?QbvyiWp7luE=38kgV$FohA*`?!gK;Puz7({iZbo9$f&{@K?b1X zRA+HGJ~`PJ&geeS+2k~w?P7_7TsuDSn^DFe9!Af*C+g`vb8?&QzG{O_Ne18#-|pe7XTAlqe8QUq@~>a9;p2yYJ*dojMgTVVzm8g{6WfDEY%bYjNt-)8#vgh~x!{ zz^FcgO9I0^9;-|rhirBwyV$M?s1>rxxH_ZVa`=@!9wLVl!2ds{x!Zw!D4bTKr7R2^ zcs5pv%ho$rH9NL?$IFeY?&S=9ENe;#(`_NmhlT|wc{`m-7(4aE!1tbnZAji?q-*Wk zwDLAlN`J#sCkeehk%EQe7L&pxjYwwS8-o=?In#vr_?xI2N}<(|WeQ;gQW?yd=?07H ziP~_u?Qw85OdNz;x6Q8+)X9OvAF5A;Sd zkpWN=1Lj@c$$lEF7jT4<0Oral-+g`MreA@HxSJ-38{ykaBaeE$zEmJ_+@78YzOiAb zCixl=E<|m)TUo;zO~G_7d_e#Lbt}F@%4O10B#j$-kR>D>q|eVqVF&@Wa8FNwB$0Yi z3YBdR%&8GX!x>qSFQ}ll8VnuT0O!FM@_Hbx14v&DQ(65XIfKHtH~iCNolzAD4bfPN z_|Ted_i3n6C^R$_DzW<^i4K|)xc+nymI&|%;fo($uOW{>0OyVR{|2C8fV?Ucw!uy6 z3~&h6P|`5>hPeVaI-PR$M*kek#_xi!1e9E%$bD|Ua@;ioEWy4q5wuU$`p^>l~DEz2buKvY|R>Z6(J(M)fXvVKt$sw`F zVTJWxu})Vi>t2&1D>v=k?yv%4)4IC45|1TI&lGzsSr8Q&1=eQ@u509O`{hPQ^RHxE z{%8OZmKzTFRFd-rAN$9thBFk1@al`@ zGEW8N?IJLrooqOKOt0?Ov8T+6yieACNV+v-eg5xEd*Nq^`5fb2r|teo#&JlUQ@iq6 zR^Bg$1=&0;J9{@WYF^+UCTIb8`+WKNdix)LA^k^zkvjtay;weaOUO-b&bnW89l?w9 zzZ^5HDbMDz2Fk}RCSGD^133J@craK{UwCcB@{cFYnser%uAml{7cJpb*tAymGSm96 zW|A~Q3TxM9HY2FO;lFF)hk5wER?)31ahkOAJT`c1t3K^TUe4qi&IM2b{#BeB!IX-t z(5kGGC0P%qLp6EfYlrXcr5Ef+KZ4~yEH%`aoYs?Ez1RtmG>9#>rp}BgOGA-;;Y;ic zpW9=Ed$4%K@sIXb<>12LV2ucN6hZoDb7K#*#T738ljP?GAF@=pfXc;Or6f)G)glKYm6 zE_}n2o2+%)*c*K<<3cNI&7|LOQ4OtCnS9O$#2}{RGTSIiNsw+We1+Yr5%28^C>T^X zUq8~eyrnCxH=NmaIEQe{7xE?Zg%a)tSOKKnA6M@<>LNohND>ow_B>wv40ozQTm%Q` zA2+G;>!&r1rSnUz)XdkLnp%~V9NwLlm;AeJ*rF}?unS3vexYATiCzIvN?G`lqt%=2 zv#*+8mG*XHz5S}Hq9R?-x9)EUhTwY!F$Ll;gJSkY89$pziidvF5c?;y{p6>i zPEmGQnI~HYt<*x+_Xqb&o*8}14AzX>^M!i$4RBp}+tyuQC2aK;Zs)Tzn){RcEvEygZ}y{Z3D+{C-(<;8~w zU)bhe>ChKxZxuZD;umwliLK({UPf|1JihwNAFsO)+gdw5SGnuhsKm0djEAM&H`4jh zuj`+kJ?XoB_lk7oP?fw5BbO5%Q11tNSHJ2hKQ?{9Sd&~6f4N}ZDm(;4?8X+T~sq&zHn7Yuh$e! z3Uue3%e*N|Pn4AEy`ECPw!B)UJH{@#q^KbCRrM0bTkv0W^xDs6f4Vml>^;>dz6HuOQv(cRyc_aS4TlN^y#P1H;&p}!i{c_QsAwA^wiu9DFsa8@d%$1+b zh)bVz%)jCtk)7sf%O>Q0fk&(Cu;c0{2jOeG#~NCPQ6tZ2Nh?K$o;_sKCPwl7{ziIgCqpT3|IGUXDuo62n3$Q!OK;8i{>SJ1SIFIA zJ$bYt7^;sc4=n@=L;UC3JWpQmoR&xDjwItx#jnqN$BK6Jp*R$6BM>GH&j zi{9F%k->GW#;0=l2hxXXM$RC<_Z{C(udVY_jltcGg<>yN26|qKbT!@kJY^kKwK%_j z#~V6nUBlqWH2RGLF>$cTfLzyor&Gos2RjYW&B#(Kk_P|60|HGR3O%_WwNEh z=koIvD6hzR)zu${?`4a1*ZU7L`J61Bb`zFE8vk@TPi}SQkvWQHD!Ye$R9ZZpyMip9 zkuN=95-W=(gbn}UneYnuL!I}yEHhyB$RtGc&LI1eoPxj4LKes<{nH&`UrowC;r78FG)YMz9T}_L&@5h*WoBw6# zbR&0oN!Xv=uQkSH8hi1)DVzQ9tIMugZ?0}we5?xZ5L^5u8iW)k#@?v^B-;l2!57;L zKhIfF-o)BCR0 zjU`K)I0^2*|4{6q!NlGV@0#H~JX@on)ur26v@ycgq_M3jEaFx7fVfy!{*B zZ6~V?SynHR;eHiN#yT+PmH*VQTbI@32qxgG*NX4Cdmh7jM9KvB%MF zYyQ&)v<|bA-Q$;e_5VEfe+=^GZ$6{(PmZm1aB5JEqeJ#(^@pBQoc6=RB7$yju9_}> z*pgwBMq*}q|M)oxyS$ty9w~CN-Rd-LEFySM~ng4n6 z#$u!J+IABQr64LWhuU!3Co{aLt|Ty3Im&{%eDvEYfBhjR?b4dQGYS9X-g5XyThz-2 zq5@CGgW>-letUTaEK}q^p3|_tV$LWD+Q;o)!?*k%xHxSTa@$kMVi1 zO%6kSSnc8XM)9t2Zz>a$Kfl4@e+Cia?JbKBJFnDY}n#4|F4e-)XMT}{mv zdef17O3q*Z{@cI#bxC~t_C#&XhX|0_6}?;jGSbtWJGA6V>kVQ53lyqYL$$K8TZ0*s zYdS@JaI9D(3N5}ZtM_6$!Z2I5kM~r^8^yLuR=2}zD`-l}2WTxKp;>(1OfKwL`zK?1 z3|pBsn#w~=HYkbSrtk7n!tLU(0-BDB;^wmV%PX?QnRrOY7;r^7DLcJ~|3&(c#3Wqy z;O_tH1Yh4fHt~-RyEsK1pP%MGgC1;%<9U z8|SI6z-vF&aoVTgYs4)!R(^y*(i|_XSVfE2<=2I8|5KnP#{$my0R6?$Vf|{sfxar$ zKFTL{_BUBf973_5m|gnMDeNK5KayUDXDl*f#YSz}sXlOJCsvtA_m@r%I=`RYCAH{e z87YK`;?Y}{?)wiB_to)t(weT5ug^Qmo7X(c+?49BVZPXZ51*tyu2KR0g?|i{m-YXD zp5P`DjzPzDp5Npfc^*AD6WIDf)Xp09Bb^ju@@2fQC7^(_k9hBw2(%m2qL zV9$}iB5wBVHO5E#yZUXC!o$j*+A^`@ju^DTNg?d`>7<-RUsddHUzy|iR)NpHKl#s- zSER`CH~9=(Js;aJ`!Jt8zCgi$3{qCMFq~e&I{WdFs>Ke^gEPvuW%xgyo^eDd$t1Tl z`nbAH*eh_2YUGR610Qmo%qg|YnIv5voR0`^aQM)Pm8L&J;+!zN@JK;0z2dJUJl`WE zmZ2_rY4Ee}REze@srm>`(^wMyh2i|j0kOekx4!cm?CIKb$*|Lk9h?0oX>|Ifbxvda z0D}~+{r+>UBZxRSYodP#>B$Jw@7!s%yqxmBo_D^+u%Ve>{OOlOyO5r&+co5}yZJZ6 zXKwafA7IQPA$i}Lul_vlF7e8?$!VOdTWgb`TfsV1@ z)g%mH%eouE#-B$|I3DHGehd&#Qn@)hvomN!s570!Dl1umF{ z{ZH?HRm6@usheZ~B zBSHjbY^Koce&!{!$GLiPM4-Fw|1tI*U{PIL+sBw_@>>Xk0@4%>Sm+?V*pLteEHtGm z0wP@mhCXUkIvS)&QxQ?=BE5}~qI5;+Ei+1QBW-BIzt2GvoU`$}^E~$k9(B(-d$o7H zYwfk~sA;b~)85f{bv(tT0N-E8@*@m)xO*;G zG3m~H`(@xacZ;U$R^b@73xHW}*#B&~2h(djM61Qo%mw~tP5*{P$4ff^v{^X#@GA)Z{Fd4V69#6+dt7-S~xP6*jXTQs$5w9#0C;(w(uqWYY z7=-@Pm503&WK|#z+a6V*R!Mf1D%9g?%il3}8;f@^vGwrEmK3nwnh62GE?-xbOI|)m z-DJ`c_j z!mY8?hh1c4FS9VPYDRGm#fx6`!mpj*VE)t?=gO@jO-CKDbvdk^xX6awOoUK1#1fQd z19>v7b}YTJ+w+d{71v!~44kU$?Im0F_Z}Wk4nUq|v4HO){$vYk0L>-jZ@=X>P(>CA z@x>oVB-(-Bus2=#e+%s2kZ<`XRsr*N1t|Bux*jh4h>~xBVEjMOK=dVy4qVL#Axr+d zW_12fFmHDc^EZ? z(R~EvU(W|C_jNc~j)Km!k(TVtofS8KWr^vzTOs&cCZ`nDeS~+*UX_#k4m^7|^@HEm z-;YTB#;o(p?nI%^b$ncZ=5Dbq`TgYkqZ)TyGmYAJZ+ld6Ao15hBKLU*8&XR^>1Zry ztZL*vwljq^B|8-5H5%mxs(}eJOuH22R14y-dP8iFzAmDl29tXF$61JOo zxvvSI$7!L|iDD#c9|a}M<{*)$=f^M==Jl>x>?~s zXKXceuiLYEG_LG~)IgU{y&#;j)o!AQyH$Cqo@7p|l7A5p- zT~jIqc;X-_WlRQeA^Lh@Xo4Gp2SW(w#U*nJVevN|5PqyR$~Z%~0@m}G+*VO$+$Cb% z;7?InPgj2&ZB~_{3I9(m&h$MX*e5jl4sGTQ5YTNO#@&=d?CKpkZawazo#wv@T(q?6JCV#7GIZ}^b%-8YQx2~=agm3GE2wi?u3I%azm zx*YowY(xR_UG4?A^uiY6ni^V*&}Fxx8>an9V_+_qJM-HW{LJFLp${B z@}2;3BqI$L7Dg`Cm+Nq~WTS=ktL=J}P{7YxGMXa7l?1qkCX6YBJK@UQm(y+@C2m4KyLt8X`Tk zzY_I(SQ-!0gx_BPRB%%SPs8~OqQrZJd`?dx(Yzg#pD`~mP3z{y-P%**oSb-yFKD<^pAb5u5A zSikT=y@GY3Z+yuWt0VE~F{p*uT0(s3*hLl*#jK77%eik!7Fjd}EhCuu`2TyVJ7wzwC*7Ry&y#w=IOn8ZR88?ZvK-_VHhwcNnA?N}{f?d6vmUfp$XU9kWcidfs*Dq^55^ zIT2W{ep77gTBOpkGXrZaP8u$_+N$-VZCPcdT_zj&R%B^r< zdLHxfsTGw=k0I6coJdT=G2B!?i!)2&)j5H5_R4zCZnP*@;74l>IZjD0DM5y=GAb}B z{G+iub^bLO$d zJ8GUYXI@3J&J9(zqmN}~+XXQ?gM*zP?W4Z9gH2||Xy4qoAA+uYxWCMm+TC@=92I-I z;*}*sa+d}G>Z0G?@$KIPH`&=>Jka)v$qgXNK_Vm)SGqxSKjpfIK&C< zTZFq(WPx-F<8O*SHCZi`@0!4aUeo*k)#rV5AZ=$EnQ1=YQ*xNvZA>_eR%9)CtjKKk(Z?Ed@QE&o zh|z5a0;AXv8BG}8wh7MC+DmCT5ZWS!mi1nn)KYqktRe23_#Oy;xepR%CN}3pfZ;sco+=V%nTQTr1F__6zlJ}M zl3#AA%ic5s2HZgk<|WMyPr*#D*8ba)2NfOvv**4ui~YMxhHB2$Fm*0XNu`ypLgENM z!N`tHX77GNf+-MaA$57~GRVN9%Fcv>rixKd2cEG!rvq&{ki>plgXfwYwbfCI)WpNa zY8vUo{!Gc+TOW2W*>LmDO+sX+e-N$RoHxY!PudSYe$BX(;B4R|ld-73>(V&t07~hZ z*?A*ZAr3%a6!b!XE+({H)CWc<-=96+fYt6g25)smxLS0HVWg10eOg0guH zFjvyvV&HAMH_b~8?WAkW0>aav+_Btddi7Vvhu8mk>W7*;fvd(8+Y~{6+Dy7ZupYhm z-0>}nliHm23gSUloe=;1rdNYAAmiw%WqctOI3>pTZHQ^(@n?KWd9GVTTEqfR7HfGY z`mN1PKh`e%H|tKuzkEB<`1UV`K=$3#%WN+r9<)AQ|F-#I#vST;gffg+ zK8f)kP=tAYFKB3)#Uo!9GkLjhE1igf)L|U9ya)6Q5IELFG{^b6h@Ejqk=vg> z^o5X(^n}p{DQnlXig>T55GW>v_A`IwJ)lTBU~DQPkg};;8QAiLu49Hi6OyCK1v{*7 zFn%TJ@e{TS=PzKrjj$lo@J6KA{cAcEx|~j_CPDJ%47KlTdV|^-Ry5NPR`g(>P+?Md zTM-@16_5=h({0eUZ+*Q@W`Du`FcORAQUEytQWWJ+TopUY+Lx}t9kG;tr(i|e?>d{! zLViL@U>$tO#%eeDsyxTB-MY_!Tc?kA$!+$!1j)i}GTfrSXFB~O@BYoETzx!+>=qBh zvZ?$3hXL;N5KbeL(m*gqo6)ON8)>B`(U9gg&I>5}mAstE!w}}pp-q53;War0+R;BB z=gu4h>6DwrA>EC*bgIuhXUA#2RTECqFZ=!Zx}3`|8`7_xJHB0U0k~lq1-KioZ!9_! z$bIKImbl!^c?bn=JNU0(xy1?o!%RW{1sc)VW6g`Qx&pE~>0Iy8$e`fy;D-n=}pJ-scGK0ipH!3$0Lku-7j>*`tSjGa5x_$%J|2K=_}Cq={*5@=-#+)G#l*4((R>6 z(EbQPS#tLd0C9d0<<~`e3}nvX79%}w$<)TmC)0nRz`fQaveJXzcNX7t+VB?5ld<#& z{#cEM4re)}M`O!A?*S^tC4ti=z>&rwq1DYE&m5A1-D3bCh_ge9YY3DH#Nvzs!ECWpSAfXYc$s(qSa zuwJZm;A}!YuPw^oqrGXk>Pb)g6BN2%Vt3kyj0@pZp=T-6sZ`m8+a<1KYCmt?U~5!Z{Z8)buCnW&)gLzS-Iu!IZ}NtJ-%r04o;A-_9zS*MG)ZOr z$DMH}jwS>jd1|En{@PE+e%bqTMs@n(+DN;1@BGGU6lY?*M6Vqtr`T&+E)-k#PR*?i z?t14n{z3NLB7zu71?Vu36*{qscpSPDD<0atSZJj{qfX(w?|S(70S5UMbD*8D4K|;8 zxW7uq#er3CbC$Zr#C{)fKa<6eyYWdD;ieU49@sGzSV;qsudfy1D)V3pd?;ez&E#n# z7pLD$MVn*p_{fMgvYSDK}%E#^^;D~yz_=u4vDB7<}vpxT=DLgH*fouQE9=6 zuqW+b4?fx9Bn9eEwKLvRq9qg^9hZvZN~JthuWB~(8!A`I4jfN}ovXZ#oi%X$8)vXY zNsz|-w3f1`7kl3ZTV6VA$f-QpFVh5r9K*n;maY{S#2S`i#c9MP=f)vsrfymP2$Rl` zbEVx&x2!srGP5tXFNhsw{O*=FXx07v3G{S^v?c_b+08uhHId6>E!_4Th;Yb9w^Bq= z=?#wOBy!uf0YY&jcE+#b7vQ5AKD~cyG~cBa{I!^+Le82Y5+qQ%Usl;lL=+DVIF8T;UqsIzkH&g#vUA8@-d=4Lf15g)40`Qa`IRl6 z|Mna;Xu!Ws=L~8K}(a* z_6~FWPS*8l;$1VniTeF-&cC~JdWiZU3nulS(O_G49|cf$?&4$kTFJ1s$;`V$3baiD zz;Fxu%4B#;ic~F)j>S8UlCZ*d`@2Udqd94|`6nc<9Aaqh6#sL~r3437k&=h6HKX1n z_#6n@mDGfz{*Ck=gcly9z4p37|{jI#{k8IA_)M!+Jx%0T;z zVFrDFCav&cpcydes&lA8&r;PE+Tu0woXh=OeozU%c?df{2oWVIXZ#u4X0m0MBychp zj?MVk2r1DAAk>??RhtEy4SP6Bm<6%mGa?&K`-RW@x)uv=0P}gTQO2Ys7p$fD0q=ef zjvarDQ7#*?tnV=_>uWjQ?K&@7ve>&%-i$xqBm{`i$Tz8CLM#o(w|I%ghQ0}2K3fqA zg9xS)TUD5srZxpo?+wk1Wlc86TT^a%dvJxiozed$*#J(P$b-VUO*+aJnn=$X2uLt( zKf}&S9gkI~eoh7r@KK)Ru_P*KELEi)LZhS2uSOXUFq(=}0AGt~2~XqLN~K>{r(U}; z{8uXwU=y`EL<=HBHk6k*f-!Kg(3fFbOui=`+$HM|+?#mxdgV+W0a*Qun6LMh?Lw^w z=HT1gN;se8`E;KLK@{|@s46K0D=)#qaqn3WDQunqtpNwzU^O`;r_HrfE-<%Ss-`aT z4I4l6m_ARXvZA6uK)}lIi2f{kYfuWj-vKOKKs{IWKZ+h~%`1PmU=mb#MOPH|Pcx-|py~)}J`ntqL0Cn|!=$4|CBz9|8;2rRF&H z(_|{9Vg$t)9pu4>fHZ3n6+~G|sKI1{W`|pR5#zI&Q=oRzH9sGsUxRzquroH@^jzLg!8Adjup?hx<&QrntW+oc7wYZaOZ{p za3}l=MM?(3UdhZ~;2c|gH8p%!eS0%Aig}{Dp98e7Fw#p;dK6Fv+5;7P1Mxe*z^S~zdy}e>p7LD6 z;<8eYv9yUw4#vOQbNN{kibGaEtVmKox%obyvw`Vuj>;DIcucIwU7^;{^M!^| zjk1yLPq3(y#;b0?pby~Jh__Gnk=Zr}lN0r(%pAw>NH}wWi|6sUrrP2i$d0fUv@3m( z7c?w$*1HQFdqUpBDuA;9OwRPo!E@lm>hlh?cOjCeqs)ZI?nWX%aWGWZgd%?sx;57S;p$py#7L11k- zM%G~UFqkP5IgARm%jU<<`m+gs&y-1zlhO9iXW03lI%tqla{F|^e4&Fofqn8q!1yQ& zg+N}A#QPR0o&VA;m?9}?&Ch64PtR`yh9aJ}YqL;@nYr`kZCs37U>bV7&s5V4%llpH z@Ol5i_4b9%J4@!9P*;w6V2~wMI72E<(dSXkUe5br8rk2UWL)~|{LDu1K%ZOwV3~vS#<)|xco!#AT)uLVA?V`xb>7?~6Twu{ZK)b3Zd!P>>n z@O`f^Y9tI$BXo1UuuRM5-jc@oV=EspBOhcZU&n31`fx}tS=O1pX6qx!LZ#Fzo_0co zz3+l!&W7j>3sZXbA3xnm|LYKFoCcK(1zSLiSexVwgsj|X(@+hRpZ$&=w!aN3td?=F z-196hMX}TAJeO8XJcNiijy`BvhCn+X;vfdGFw(YGk0W2DenXDj@L5Jzx6gFon*+k= zs|%lNLni_4K{1ZJCEhWbZt&_8dPSJGeFel7V_-`er#c1>Eo7>$+&A=R+R{_>)y{n< zIJ(Tl6>rG&sow!KC?r+VvHK064W;nZTHI;H^%LyNgr*ZjgL3qBwox0dp6qy0h;k6? z>sNtFai_oU&J5T6ZF6IWvfY z?l0%=<7&<^YCMA2FnQQNA=AB3q?~_Z^{}{9cV#ob*y=VWoR!-p{x#Rh{~)!Qm$xzx za_qjxL<603wR#Bpajdm)rGJfu`_~Gq9ntd3-ydQx*YQfzZK zXMs=_wYtb!u3|~x>LQb+hNheNE_$p3I^<@P=r^wk&nAaOLoJk~L{?X#-lL3WMjane z!`J!bYhHjf=|tFK%P(t)0npr=_V632C26Hres0CHIz;YApuj4%Os46<$h7gX&b-3F z1lPcagKti~R}`qAzJcYEF1`h|;E^_A(z;@5ptd2l!oSY?iIoZCGJb+$p0h4dg3~Ft$AuoSt^tCxjSYo=m(+<{I4vZRw z7%Mb>oSP{ghd>vUr$qB+%hbj`9@DQe~f+>d&a8L*F(V0`!gov=9df;h}C z?YnjGC+Evo|CD~`AN}|*`@u(Ng>UbRif-9^+iu&jJ`H8q0F5#OS!qX_yPR8|4T~YOYZE|x}(jj%f_s#r(IQXItNbwa~ zyqd0uz>fZWvjCN%>6*(af))|S`^edyd7rz7@a)0nr;dQ!vP2;#i4MyYoO;Tr;==L9 z7$+5LvM_S8qvIYPlso*a(e~X3ie%AB)umQziC9=3|76VVI9jzwDmNNmirpOLh)&Bo zT6TXAn6qNq4GSE|pp3zXqhE>d1%sk>nI3`!`Ml^7G+qtuLF=OJ;vBSv=V5}(}mtb=OTBn z7+j7=+KCn&Juv41m0RSI{fu&RFU04ZJ-x(-o&r$$uYn?3(;hoJ z4Lm6s>4fnM+nCM81YYqX<42%HmErEw$vQoo=HXnHzdxa+WAx?2-HaCIlsx*< zbJ^^rVbKHWj4|S@n6L(@=xIl*9py%H0{aI@3zyR~rpB#bYS`wCUxFaFND#3p>W8ny zjOf2I9mRSC?qmG>vtj%~NI`NHpRsH)tI6zCA+@U#%rsoP2!yv#W@hd=k}h%~X@MvW zhJuF1w$hil4Ft)Pzht4OV8$~P0%vC1yRFF|sjYOr8>Qa>#G1NCZ@Qb=>!v4I5E+LO zcgDv;5SiV&+X4uSt)FOB(qk2s@~sD$T&Qh!3$z6Dh2b~!`*8ZNk%JL|JjGug4e)9INmfwmB+EH<{pOCI$>~Tr{^XaLo6u&4aR*?+5t<_9eqpohTAmbL?sQ{bJOS9S zY%dsMgFcXNs^B(iTUNEB&CbFLa%|}%6(8CmwN*uQ<`(mp?gQ9H=Z98S=XqM}U8mym z=8CbWHLG+kKfb~=sf}gB%N72;c%)g|b*>W=3VK{ivxnL9|Bnv=niGen*!$^VL=5MQENQ+1v>U9pg>5;Rjhb^=ytSZoB5 zI}7DsCX6!)pYZkkEaz4fsBRgKwo)ALE6+XV=wyvA%^zd}^%y`pEq9$jJ3#W)7cey5{SN--ArU{nS z9j$7|jj0>@CfpamDisziecfGtIU_Ik;e@ppr3t~CqYW!izn>4zKBy-}kjtt(Qv^#K zlY{lrL-48I@7O6p4Anrrk8fBxpRtFFY_Jfz$0(?Q1;-tSmPfy?-ni7jBoGqeI8bwH zWA6oP>T0qgt%?3mb+jR;GQmvpic4EXYy~T-OLIeGKObE7h*Zuc2icOz*EZYuBG#!0 z))O9BTb-zxz6vy)#6(~po2b@duKr-tJ;30@T~Hr>^gwj}2IO}MO1%VL=ObO~Q#VmM zL+;U}#{#oX^YjDBf@W`jpP88mv&?;R%42=>L5FoZ=r_K(@>CF>0r6EA6y%@ z_%}O={le3qpM5Sy@xYfjSmJKZDi5J_Nk>G#ADAPn@RkZWrx=;(7+SjbdyYet_vUpz z;aFpTmi-fA6|{;~L}psSZ;NUxFAEpqU;?%I3A6u6Vv!F(3IfJqV=W1g5Krp6{F%TJ zQHO=6sJd053kT-tSC6Q)4k-X#tX4u7=iK$`;fyl5@-D{~*owdJx#eyPi13H2ovdHU zcw9o^uzb6%=Ty}mmqDJCGmNMs!yauNusnesnNovwK9d#-GD?2cQdD{k{C!X@e_d%Y z3#@vQp@prqy4B4~aj`ni*8FRAtBEZht>uvv^0%8Ue`b0_FSNHi7 zV;@247<*x@rv@l+1uU!GDoQ0S`7X@cUWPH_HX7mtqH4OlURXM|B`{i z*SRFKhx|Dzk#PC@FlI06z7<#3?#O&JHu_ln7aIu^#Sp4!J^OO+$j0i`TRvoBG?mU) zBuc1@MTU-4n9x#BE(|Q;co2{UabTO|*7_bH&O-?-nId3wc^R2#P$|UnM%uVBzFb!en;<<%kPh?N~Y(^+_h@wvl*j`Vw7p;pKWr*02=f{ueO1nfx?F7jlZxA2-P1s3K zpIia;NJf{b+K~sS%+ss8@T(r)wmZG)+Mt&&Y-#!&ODU~A;bI%UhSBpnT5_?4xLi)N zP}056_)Q>nzG8gVBygsrxgu97@IEzHO5CKCT5Hn=Mk%qr_FkfX=>L2jS(&MmEZ0HEVK<>?9@<#NC;tuO0N%_rZDofG zZrTchoQm~)5vT_eT#)5O6-)7X&n&9N6w3LRE(R4k&W5;{4cRg29$*BWz1l!~%N}&R zL3;XOXN8cS0fA@0!hbDY@F;BE%T()H(p^z*UFbLv;zHqF`Y1{t&=M&-W!||3JY&J! zbfw0yrp5gJZ3eFGw0FC*yzEUt(f{KgvW?h$+^-ViQgN1Ow^y4s({BqzOzAo>@ck<@ zU;qzy%w_o5cccI?{47?jES*l3p*MOsFH}WDkdx@-9rEZvnBKOr5|heXGZyo^wvBne zFEjCe5i=mcYamR16Lup?iRJ6Jg8%mH*^cGqt={fGAbVpV+vrY7oGK>_QUT*|HE*$4 z%6n9Lv|M?qNi$Hu)@AdGrO+QqhLQiJ5VE)qT?Cn|a5#Zxv0w}C#9rP3diAoG#DjgN zDk-#!?=AWmmP00uSCr;r7-fYZ4f`u2{0WQ8j^wffeb~6Hbg5CUq_FdPxdy{xL>`bs z)}L6eJvrg2m4jn3|0#^DOoALaddYbUuV>r!54fe`o?)#Nw!S`Z}%8XDQrO$ ze>ae*zjzpk@xIb_72v-W^?MkZzy3QYyIqEbr%|YOjb(lUr<5hsslSIAUCv1J$@P=( zw#U=9?q=_<@4fk*S>(uLs$PkVuMc%59BfmEy_d%p(9Dn1z{$UX=l(jWTNckBncBrD zoh%nw2m<}UYp34sA-B1@XOy8vwer)jHWZr*g;m4*@Z<Y?$smXx&AspS7t<)Jro36CQRr?M;LkAV?%5q&p`llCSxT3koN?Z6BB5tbn>(<@H6emDe)wDzIJ zs=54q72HR6U2FNiwv{_Vl%(+cc$_=Iy4npS2x7GGrgyjeKL>5QYI#2~%H_5Az{|7xgMjH6SGaXp%(t7ehzD4L57>><8uJMa4CP;7IoC~ zIcMhcNF!84i{l@whJ6&t0zA>PyVf8y<>5|)6!Y~)mDHtzVJ8bs`H_vq?>5ASud&*h zD)sZ-I)2dq`1sGu z`}c6e0W4ufw3;%f7Oqxp1QY8R zA$Jd|sf#CD)XNPLLIor|qXrHxBJoH<+J4z#+ z&d$%EcE<48druhjV@fR|Brg4$pBQ%n{N>)iXXpK8%=Z?acNNmhbckm(vmPHwFQDi~ ztM9WeI5rz20cLI(Irsj-ta1{c9Ss+xjU^0$`*j9U8XLan@8g^*&TCwXbvuO~`ug^B z`{4j+S9qVexE`sn&uqfQZQzOIBjV1O3k*M#(A#*Pad!irV*h;g4Sb@w6a!8x`x;#t zEm4_!@$Hv`TgjS$iEw`J*)`h@%GB$?g)T0&2}2OI0YX^rxL-w8yQ1S2-)OcRXy&pb z(_Yl$Ty2Z9tuklyhES~a)E3(5Yg(9Abl=}e(>UtU=DY^%<~v^fYpMpF%E%bas1 z-`2mr$Y`A;&5MA;`Q9dv7R|eVF&cB{gC@iC5V;#Bf&M-#zn6`e zYC^HL=nF+>W?uIXT{0O0g%@1CC|$xOK0Xn+c7%*w_kaTgm*W(m74i`%CQxVN-b1lB zDU{Z{cOhapt-hnFLNTPSf`EIw+n7}tVe`;!p?t#d-Rim&vb}NE@>+3eA$h$HEFoz# z`1;oMO9f!7oKXfAu1d{uu9{nU-4>T$afZvDFHN?irp6cdSiiP$w4VB&@LUI?bUYU} z?Z@{0x%H?#54F`G+E=NNF*t(xji}wP?m7kk4A^{^T3=eQrPQev+$|Sr^!fVwd0Is= z9uJasw3T2xf9Yh!egM4S`xOqvIOg%sWzU5m56a)C+04#-zA0arxA>1R-uT+@aYUYY zQtkUtKd|%vA-(Y6T{!YNE1C8&T9?cnp>;mLDy`M6xvQh%?FS|Nf(BTq!R~xuhd$ng zmeXw1xIU4}(h|4dx+WG9W%@qO8buq9p`s{rkH*l^O648@c)$pxQIAk%J11kc@{g8( z7JSjDS4F7&q)xrjZRp2T>aJBZxd1%*6YlYzOJzxRRu9LraON2ogP{)e_yzm1SZ@5P zNEYa#%o{Z2*5gj0TxId{oJx}_@p=?A#x+20pUHN#Q7m+rw{s&2axikOcGAW+?Oiuyamd5i-H4%s2uIUzX#7Ylp~ z0-sd*CW@1*U9+hPs#_?z{noCoj&px7-`ac^vj9=wr6nchKVLC9UoVyCZ8=>QEgZLT z?)Zu~BC0fQoR|1ZAW^CnA;qlnEGao&F@$TMsz1q*p9-)spHvtvHCdb#S8{2Td{h#Ak;&KKkj&^h$y8684P*_CztTfhi3@KUeT@X0gr!B%Ee z=TgFkZ>P=<+uC+aT#Uy#;qlYi5~3j(41=sR-|w&@bz=v*!DptIU`acpq-W&@PGsukV2xPgn6jQK5t( zk>k%Lay~v@KbJ>^fQudhN#Hw{ajR&IB;y`|i@<&wbL%=~=yoV9HuE==qJ#kXkXq6Y zdl>#AY3XeZiR&=t!&g}bw3V1Fe2$a3a{skvt2Qq)>L4?-mGFMS2gvOPhErDA?KdTf zeDt{w9+das=F8gD9OihLjot#h&fU^DO5eu#dIw)fWL!NFWGe9kk^{SD93iSqkS7KS-CNaLISek-F7ld`7UGZm z`&vbSfj@g!wfWtI>cmh2MPdlp>g;DDqlNMQje^%0)p0{=?A6)pSn2z#x33UwdGNIt z;2X)@zQ4zx)>9o$c_&dR_KQoWkgWpYAs*Fj)5o-l66uc+HDs;6Spxgz_604YvRlU4 z0-A9bB5ds10H||dvDU!(k&Gvpf)}8DN~XU;!<*pqEG=P5vv5@4dG_IdvVDg@_^Iz+ z$CyNJJAplt8M_}s6mocnf219)vq`LupMC-ZX~H{PQbI&JTk+J)h9?7H#=;;Gyh0jcko<WkUD%&p%PO#>D>>e8+ua!_z@SglV0ym7S=IOx3 zn(gmZR(H)rfgD5CqR|uF>bvwedzfa2 z$G^AEkx~}D~ z(zy!BwBmgHejj@_GKXFq8hMw};tu8sLV*V<6zun~m9xUsBNSCF)~8m?@Zv!pkr}1v zzV}GJI_728OdlLS?=HOHR2l7?Ym~bkP{`8v6$zR8<)(eV5#)^?#)54PP_ zK$PI(8jQn&Z+Hx$$;#Z&yGe$>LdlaNNgq&7!aI}S0j7F+^Ggt<=rC1%W7j9aZ+>p77kzQeRgyJ>D_2721{Mw#^zF?vc8J&QkD|QHD742MWJ*<&Yns<8_C)r1*SfO zNh|Bbf}Q#B0iZ{1wmeCfI=M(;f4A^TB&A$>%og%Qk|2;sX;UH{fpde>Kcu?7{=Ry~ z`ju850Osd*sT3{FeUQ2Ad2HM7j2apBw(d38%hfY-(LR2ivZ1(>m6uVc+#7oCe&PcV0Oa$xg=irj1vJtoYrHOPv?|1O`{sTXdcpc4fodGl|DaXG{9 zO*N#2gGQidj^qHhe^=m@KHM7Ae+q)P>yn>Rm-phOlmmGEzWoG*fG-XJ`ba;>Spi4F z^dW(L_k5F=(ZWl%=7(c=@3up&@zdDm<+~=XKZWLJx*Y_2PJw&pDn(|RR195o5A;ji zx;*_xrD(x)YkYoot!C?cW%_+T%Xs4od!vH8Xh5wcAx{4+89@Kruv);$4!Ba*@w^{+j-YC_s}9=ZC%wQE1} z{k%{8+@9?b-Nys@g)b(VpY>-C5QfDMV&4K8)XeI6wyKI~=(M`9#i&2-Jo zbj%YL)(F&pq>tChzUgybp+Q93(E&?s$GJ2;yjyVSN|dlh5Y=`z*~8wh?SEs3S= z@*b>_1KHn&5|Qo?71rrdN83L!MKFeA&IBuT5a0P`nlL(`{?J=yDHxeLQ2~ z{_X`xK@-r>F$eS0$OexVNAnohw#zkL`%&KI!)tP2N*m7|ON8o5##kQ04zm5XuRdFY$hbIv*DMFKcf zHALqplf6RoJpNL?#3(ba;>YOrRR3cygpeM>w|X{38(g2J$z7QS=J2@AUllga_t~P+ zZl1LAE76}-pQ4J5p;`DQseYnys}wC=bG!>q!+FNj+62Alc_C;mLDKsFSZIWU=_$oo z%lJ*5{_8O>yX`(xVR&!kzzU`pER&4~eW3A)-QO}^*8lN{`(jOjp!;IcPWJ(dDz)DA z1@%t!z2Uq0T@E8>fi%SHT8*&RZ3qpbDfx6uk|R&1^X3bAf0O>V{GOH8ExlVc;q%cS zi9(O+HZ$%IknaN-JHw-YetNS@2ed#nW$99qU300xGY`kSx+lrZHWpoGqGWFwlkut> zI8W34ExuzD0@HKfFl3YG`|gzNR>V0j34xfAahOH09zLA)jbBA(XxYZYqVzA*Y) z=qX<1trGY66Xqp!DUJf0d`4~F9W}iCY4@l?q_R7sfhc5HAdSe>coHjOdJYzf8VWxY z_%e-fRQk#L1C1%6#tNh1NOA-T*JALs#3j8(&=iUFj7(O-T&+~u)55BjLQAfM-}Dvd zZ`s$7JL3)U-iU{UIkf#Gyd2l|rUPd?E^rsOPljw4LlT!Tf{G&&K;JC!k zUaS2P(Z36Gxi;Bi&M#4=T;K!ViaQH%8Y1_r#+Y9YLEJT8!g-ruP^x zw46kDdnPlR_6`yPGExuNMxX~9oWHl<^KO*Wx?~W%hXeCBOEHVIOZ#~=-4qLC)baL{ z*i)^JNn5TYJoBv<=Bl(RM0MOwzM(u>NzU&KJ_e9--h)7S>pLobDkdYrm&rsj*fAF1 z+saa@fSHaA{EoN311c`_Xm#XvyEmp2TrwDy3?NeLe3X@;qZ1z7wAZ~DdCtQ3J@vjl zXj!+tT)m##92m+wYkuk{d3?ytNUZ^l3`YjR)Mtd)1x;p=h`8J>eeoy3C_ZN&7DsA{ z8kauj>r?v}f`H9hU~WCP-D!TLU^4CPNW#-b~ctLPis9KDZo)$ zE!!->be^3b#x0qFbAkfNgGb98DSX1E7rEpIK`cls_S!a9cuo-qO>%Ak*DLeo)*cv^ zG!?B=df@N6*m4QzrSSZuWLiNYj`|}C*|)i1hrq(H8$j6*o2a7oVBq4e)00Xq~4hGn-z zwpZrW2}g|$wp%#1COyU>xQyQppn`K=J?9XOVPPn0suCI1Cor7=?Yg-pi>p95#YUOB+D@A>K8 zg4t1F;bIH;b~m7?8!P-QUT(uLMND*925La!%okLvii&=T6MT(bpKRCab=}GtW24li#l5LPf-{x zLcWdmQQER~?=9ys_p<6~4RU>a)-IP+)>c`tl?LyQdB&Fxq)F9zI+QLw_!~@6T=7x0 z2S#&Ldr?Vx4e<(09Wt&Q16P@pnjX;X)B!Wirosx_@<76J#Ry-ZQnzaQ7xh$IK@!aH zcYYuUFFbTi(Qf^1!pN;!aN64fA3W7pq&;vcrdE!ok)`AF?*T>cF7{STD^A7xsKiE8 zsg?t0IyB=Uj#WDG$ojJ>yWEWx(_5CA!59ariqV-1-myTmriRdB;G7O@U>D+9+C{q? z29mm0@p|$ZJdSTl)H*-~kul>nwUS&*cx=Mt(}Wiq%sf;q%9bIf>rFjWo=hT^~;Z>@C*-24iDg^z+Hpl-Q8WF{-!LAtG^Kw z8&Dt}2z(2;pA-Imlb;kWZv=0%d`Cb>=rcmQgl?9^w$GzlX_eibUAV4eQV(!_Upn;R z?m^v+-kLxwcRIFAH&iW#<}H#CUA>i-0bazEWU-cT&jbT4B1sS;t>v=!h}C`!Zsxkw zpDp`sp5Ji^U(h4IyqAOO1Lljj(+@L2_KW%KQ-2hd3-C%T9C$0@RTyMa>>^Zf%>-g~9rx@XPp`bq_v_{_63BaXiasrTKyk~|STi|m zNvXQ->z{u4FYyWhcwp8tI(!d4P=RFO{!hXX;bp=gz~zr^%c|4wmF14|lJ@OqSZX&= zrU9?-D3bx0w+)8BCR1slHJRU3mTSupWIRgwyL@sH8D@yHyZ0w z?A@Do^6gOH+mH{>e2YCQsKd(g?5>z*qi@sRCK7kLa-t43kh9S$MZ;BU9S1l_e=thR<{`#jU`e?i^vuoer+{^6 zZ`8wBhq#fU?g4TbZ){l#Bpx(kAWCro%%YItgT~*H)FzeTPKjU^7c?F$E;d_I<(t!S z6OTxbF^Kr=GA;T?#9=o)-tCmJ)#KnTa^RZUlJdfc-U?5qqR*poJTTle1O%x+VL@uV z+k5iGzOKplVW@?SnuXG;V%O^1FeLcIO;X&XiMiF|6Vm-jC4{5gsqPur@&fzqAZo&) z{iXw)q{g|EXnQ#nA0j2{;1tZXWm<)q3zJXUNVrDF;lQ5sR`Od15>L%-YYA4eW zn?aFmTbq;|jTfJc5Ap5Uw^hi$zGv-}E^GelBQg?}@x}A6B)2DAw&)$rw~hz1JDFN$ z>7t=w#i7}o8~o5KWyQeR#2=A3e1fa^LLZcntQAzRGBpW9-j-=qa$|x zmxpGELy{ascVN=&8R0Dp8yNF`FNayAMKM^>RSL&-WoYp$o{3FghjL4)P-4L}PGO1q ziZCa(SbD0!mCJE7%~9|Jer#h(JD6yze-b09R$J(<3bVb}+#c|POA&vR64Y7J3*7fY z?ky)eM}bKYQR!9mZVrT)4ebL&dTU<>s2I9W@5#|?{VmI@C)s&nA_70uvim$HoK(N8+=$kM-2BXqX1ao$Q4xqHR!7)hAOj05YsBNHrT)zp3ZtkB>+n8Ky zuIk1(Lw81(sDoaHuyy%NuQ@CV4POT&1-E4|`N?%a!9^)~II*x9XW>VaCS!-DkAlhi z^zN;z?|A3A1U}(TG5_*I;KHgUhgZ9mCbQLW?kD9BFsC48nLZ7F3(1=su2yo1`EJlM zs;Sct+n`j!D%~ND&m+*|*QP4ifnm7Zf)ys1UlHo4U@lBl_+w^y*C4YoMZ98mB%?OB zJYq^H0Rs8&A7N9l+8O6?Vd?)9b{$|*T}^w9fAmW%1PdTFY9dw;5$T{&A&P*26zPJ9 zbm^TXF$z*`h)7dWkPaffM_^SDM0&?vrNgq)mbUxPSu~N|tJx>d0~omX+;gV9^UmBE z)#B0otYA(9xe%CGz#MJ$YTuhazprG7zH0?eXCwB7peqa+x?|7V2tU=hF!z2Yp3-a; z&cd3&l}>ex{IxUC|LFi7$`o`0u@x^%(!_HMCz-4A(t4qi;{J(S)H*v>Q!uWi+<^!N zRIWevy4;stW~PLoTu)g3!zG`u`PD2m%;J8 z4;3(|%n~>YbfuFkpM9+vDbF;KWxSqmtf&ptr!2K^taHPN-CKW#`I=U$XM+g*_4Ee2 zo|59=a6^%hc-E$6y$LV{h|Q;!p0xs(aiI!JT~_I)1mw99cCHsN3$Jmhgm-aToNuwD z8mNirC44T(A%2H*!kis}YaI!ZdJ^OsBNkB3+bLc=P#R7$6>nm(y@){nr)gR7ZfT5- znj?~W5=m(vavIF}^G;VJDoW2EuXw4bN%?q-llZsQoh*$pO&)gPd@o))Wxj*ddQ|^B z==Pl5_U``DDbfyphl{!7VRX{4R(tyrR2m<2)r6! zhQf`#LH#}rfAPfJb|3Xk$BQV>D%~pHEy{mCGgYiyM11uaV>p5DKI``PhSq8@obS_=)fuQYWBK_0Ujm;-WngR`bg$BAP$(LYBego3vk6r&A;63e zr#v#oSKMA*M!o#I z99((&NBIdKpRpJz<6Fs*3tiLSuy(Q{I|K6uYSd%~?r)PeiyyYNB~G>MWr>wCUZUPh z$26zuK+L9SVuHG|I^ILOi8uL{28>OY@a&+Jr?ZZFY^{I>ecw{L!*w!!{zr=5J%drZ zn)$s+kEz6cti_LqmQ39ygz1VgU5l^LWRM>lDL<4Cz{)e3;8yoGiRtc!3(UnC89$(B z`!qQX(QzRBHV&tXE^bDN!WOryF|>_h}v1FEZU!I{$|!vDuzh z-t6fJZZz+`<1Gm?g_aHgBKX6yH-3yAlhG-BLwOwE|FQZGr8bMh6it_&&TSDQX!azs zK%u*)|GH94yI@?q(5I$ZM0}5ClD7N1Q9h8U-F&inLd1UW7>r>-nlHo zI+82D7J!#Y(qn%|z>64T&uSi^$>$@9;aW|I~ETYGJfd=|-ufMI(%j5=8{zQs-T4hb8G! zv(-%Bsc*^gy>3#nUu0uVBG%kM4s8XL;wel9w_%Hs$4q~SPDQ_xc<*dMR#fBUDj>J3Jh`nSF70BB1k6;|wtU=b;Eb zf@bE#UudqItMWFNu>!#{-LaS7>n?YEIFHBB6J2X9qs}RYf)Q3_{1AOm-WHC>SIN%`J_?#5C%WZ|?6Uj+zG%%TaDybg@M|fzmxJ$by6?PoYSU z>-J(#*J|JF-B<RTXdDr^{dHlw9_X*Lad~vO+S9pEqk9#8C5Kq@GC#L5IF@fHN}vRUjAS{hP0uCsj=7c- z!im-L_gJW}>rap0j#h5$(28gMV~*Fo{57wF1AOKEMyZj>gJ3 zCImwhiD%iMpU#E3v1ps3_W@XA)DQm&GFjrzi1x9~+rD^U2)i^mNoq5L^b-Z4cyn$t zb;N;%efRlx}J+ALCy>q=Gbi*9wi;b=z*aB6;PcJ> zPTLDv+K|9B@#cQ-x7OBiAk7odB)SaqoKauG)(2vlpMzP2BC(f7xXuhq{bxa07BVn6gxH{a5>6Bad zJF_*vN=7O6LkgIP{4gG|N7M$AAA|{uy@EDH#xoK-8dxQKo%)RN<02&C#-?2M7>inn zS`+i^C&BJgdjC+`qR&P^BE$g%2p(Qt}dr@0>T zah~b16@GS6EUEB}6-$Bne5of~3i~bKH}Z=7I#q>Gp=P{(>}d&v<_; zi6`zAL@&%A-;%u9wD^gmcQv=zzyCu>??9{%>5t29voF`!zZ0u&+ch&ZCS}rZMxy5S zbnveh=0GJPWj;CV^1VptLAiR+b zn$g9s#Z(LZxmi}ClP@9%T1~F4rja{VmU&>;_f2-Jcn+Lt=Dzj3Fnezr=s?oOo@5qV zGf?(Vli&=IAc-%x;nVjQq{cE1(E28!!t<0e4}}QR)9a#T^DLQhuw2BnvGcd&HRV+Z zDM_NAfj2w}?EPch1n9cT?ENy>keD&>=2w(aCwoFw*I$rN44In1+=G;Rs(V@9pZ|0j z#)vU&oz}?)iu$3bBE@$ln!DCb50j#O*u%>FG$%4N=oO9J^3~0*37@y#!!R^=F9*7$ zmw4%%JdauOb1J%r8llDE5Lrb4in9aU!@i72TXv+Cl}dn(C$9N0;V%gE>UER zo?ctAA#h)Z+UhX-`iQu@ejC&JjYLAo-Sw9PbD6}q=HOtt2bF<0gU?~vzKl8*_e-1m zW1j)p)-bBX-am>dllz}PTb=ywGK#YgU1bMFXZ{=(bcvDp#I-r0xNNEkTAeKatulJ1 zV|wYGF0&Pc*zD={{*j2BH*5GV?)fdy=7v(Be!ymnKXOE=e_s80$8X~|je{cI9^S0< zXi7KmNPg~GZnp&D6@UK{B_=OyYmm67KUJ{icyyO+#dr08XKV8n0x&%G2;W(Dx4;A#%$jljtSAx3;DPveKR zon-}5^N-XPy4T#v5|v8byG^J=I@mLTHG-pu{^x}XCu{5%78(d`TcS&*4)bVt9t_4H z`>AsOsM8y^?N|=UaG`xxi4vu86+b=JC(UW!W!}&7Twbxq5l6fZZ^MskuE5&u`+y?Z zTgXurA*07#PA0f|yZmZf!Wski;_otg34a)O?OS2IFA3{iRjz;ul`WMihCOydh15z8 z*N@W&(t0GXY5P!(L8D(kx7tn&;7Opo3%3gua61zb4QJ1LRjzx&JhrtWk&s62D8nx+ zq4cw>N97U$DD3cSQd};(=qIYMs+`5`Xk6smEY&TQn({fDN{jKoN`JUM7JYoATRfPp z$VqRnw$NK(eiiBop0rNF98_NX`!W7sAENZ~P?pCB(>zR$na3UqQL2N+C{l~nf9&w_ zr0-?1GX6RT!GpqZwepoBr057HMuKOlh8(m15zctnjTl$$)^Qw~ zHL7GCuNW(C$g^u^@qN~3|4R(xh8_xPQSO-sgJO73{BbBP&z@VkgY`lurf|*03>Z3x zjnp!xil*C??8`}tpcyn;?eUS>1sUeeHdb01ru_9I<8_XE>Yr;QbFCp{&r=Pl;>CtN z6pCFli)!edY{;p9*0%Xmbu=c({@woXNB*v$9B<2nN!jd_0ubNP)ht6KAj!c0BEWSi z&6bRkySs!% z*AK>)wp5TuVp*_;^$!z}G1g?nkn)(4q`vt$dy9Ssuc~{^xa#}GHXG-KrqK!JRH3gj z(!mt&b@tn|j zdgETX4KEYQLqVkce|47eU1WU`@=9dPPjv?-#+TO71u z0UH^+_CcaolY|`U9H#Q_-~^IW zHze@tolkJ`m^arlq2~^l4J3 zHhojZWj1iiW(IWC@tT`E%~dVhEDR|o)IVc21XnfxOQN3V5hax)S(nW02!A0sGu7A*$xG%ovn}-imBLz3ge@@=LW;56&RAa&Y~M66HGxdAFlZk zX2GcMbNjbB+q#RbHjw@>YP4$iZ|T|}NS2Yfz6aE6*@x%FLmRZ|54owfaDHK!=XZ4? z2t;$Je;;gBfoFS_)wHsynMlx4mWv%n(?Eg({2xLeE8P9B7?+gF4ux zS%02w4~tjeK?OU<&M@?ta2E@m9)*-gTCuTYt(-kIVTwjI>Z3f6?qj`>0^>r@jm%g3 zii!3ud-$avS~GHj16#AKHV}jWntGC6XXP!99cR&)tEW+nzVr%@2bEDS7_`{&m+DtMHVBONAJKNq&(W#Jd;~M z|KFE~xxTLGY)|x!Y_7AbR~v7{9G+ynb1XXBzOH=1^lsS=_db$>eN*VZO1n{`mnxEv zR)_JO0CORVN{^tB2}R3H-1exDP^M{-i2J8TT9w* zl*(dTR+%Aye&)H+*{{m@?VQkmWxViXTmyBen59a$%(!u-PI(#w=wpt(!{NLoNz?Lvej;PAxBpoonT=ojemj)DS82ega68twWobfXCgpOUI9?FAO9iTp&;o7c$eM#p0wMF=$15MzJrp@=iiqdRN1z zyrHu3FH{DdUenHY&ULX^fu*r2T#RgofGJiqxbNM2$n%#vZfGhKrlcNZ4P-xAxB3@^PU~|)SrEn<`B3O3dj$=)p|IEm}icIM2aP(6|I=Xt*r9{r< z7wg@ryVm&iw?s+r?H0k?ls-r{y*3W0f| ntLMhjN(o*R_(5 zdg$*@MrE09s*`nQbO1}7bMuTnxQ|$2GdyoVUslG*8^egw!%=$nYd3;6z->R2;dpb; zQ~;25Yg^4$RW4Q7_ikUifAx$qiR@#kHzHBB8Nhmu5ZV2U&GUVY;M%b>TKOwlD=-h5 zDrX*hFu+P!A9;K7>W;jHs}?n!tiYilDRMz}$nrMaV-d3%2+r8w(4$$xaH`SXMIoWi zE8$wMm*uZ4k-Gt-ergFDe?`yNHfh%B1vt8{r91U_+C>u-OLSS3JS8#=c`v~Il3!7B z9Gj{8nTjcCLHoad0o)I72vup5`O1zPeX%|8t*`TGHe+v(Z-}tOU%(-d_&8M`zF0pv z-Ya7=(`_Mh(2k`XI^=ht64n|!f9qNU;|m0BN_#)#1)K^U@jS-c-Ck!^*}r28K}dEWL<+P$n)&trj0+&e8A;!8g)-lnsa13c-K(%5 zKxIR8TAO^JUTNWrTBZf*+yW`p5nJjbfr%DmeiJ+u;5JixV(+?zdp##NB5I@4F@vSd zP{Mkma=**t_aogYrH~LyGUkyHyvO!l)x#}S517wRF$tyZHNWMc zjpN)aHIm%g&T4{GhTje4vB9U$a8RW^ol#2iw~BMB(L|e%S9*Ncs)o)QM{d6(;Qy zzIPr3%9SJ2lYH<(;5pt!rjS(P_R{e&YTuMG^S?Nrw%Uf|PoT(~a(H742BYM}AUGH` zPIIO;7z9@tyT_tqI59tAJ-wo5-Y%JR86#-tRYz4aNmol1*5j-$U|%Nb6%;?d{6(eP z5O|V;xjSdQW--E|=ZFXV#M@x6M@quob4!cCfolL!94u1;V<%5w02Y70*T47g2-MKrFqD^eQ~* z7A(1qar}RYr7<6frJc{aUsUi;H=^GTvdLv{h%K=~9X^+R#!FKmHcI~1!AVKmCi=@3 z?#N7iB_#RmI1h?PB*xzyPFukg{~kN`X+rhaaVXB?()5YAyy?F`04iS1Z%J|%uUX~z z6V%s#`MHnKb(PsMHRy63 zyvwKWCkK-H_f_UgU&Yhx>iYXB^%yUXs4bzQmWma<{#?!&7kLx+VJS`W!>y5`0v{ zM4wBw0N@k0B{>Z65OJ5zV@KSrKgQ6hyA9`~9>c~P$-Jq^j@4llO;J^+#}?ai3(`nb z&s4ImC!UR8Re$s8GjER^)VTKD#CN-*j^;+BB*x!Myih<;yZP)CpV+Gdw<_+($+tSi z3TY@ywgveq{b_RT$VtOX{yWv#kIOt(vj@2# zt~-fLyNT=ILHu@)qhCat$5Eq~5 zK9impi+ZM`dPDz3~6_|M;4I9kqhM$=0(1|OlGMn{4dSf+=dU*8*>AaA)L-MbsL z{h`mz${E^^Vl{lX-fn~1nj+2GzgGhzbiw^lLf)n6)nbAAhm!MBgJG*!n*Myc3WAhh zyGp=Aens^Ey-W|`YUzg^7NtrWq=GEDS*E77E38w-_fNr1~U8Tz6 z`^$w?y7RN?nSb{zBUJbc7$<$wE_sEHE#?Hupvrl9vCul%H+F;KpdrwA|IY>?o z91!v@KjxywBxG~7@$u!q)car?^iVjM1r@Kb5O9rWIUh6cHlku-%b&vX1u zHX5u_oL#4cGj=+KWl4VNaJ1YkPpOdtLA9cr^4l#Yu=GLDaZvt`bKE}uG)tQ}a*Kf9 zG5!>O+oXNHW|kT+ z`>N&Fx}(Y%C~ym@p%ylnSjf#%iK(cL1>Pz1O|Yx8{ouM`W_Gk}gJt38n)!1PQp@kb z_=zh*z*~XOQziPm`QFK#$jDZ_CyTJ}jrcE{a09h>Se$KeTY$F0W)EULTlS9Sx&B|v zU(2_^_wkanf1vhwt}J_iU320~w`;LvpBK7YHG~@FuWs2sZ73OdHWGiD$-wBW|1VA6 zKGjGPw0}8)`n=2ym>1@~pUYuO{>zGe-eljqnF{pJ$?r{iNuJ~20&On{51xZ-0 znI-S0bRRWmV0izlW(37`9R^64+5M_5)`8US_Z;5A%sV{uwVPe?R6QdH!FHK+W<`v z#`1DQjq+SLyn9$$tt&qHuk+ttkt>rKAmHv14YDY07fT@1SbN;@K447tVw5%N#A9%4 zxR9TP@y#cr+=A876K&7}B37Q(-b{RtY{@MH%vy(tu<{mVscY3EBG#cZ;l9kbw- zel66d-*LQ%`s!@w#9`V4CY9}d@fJBvMvJAg``8Y@caP)S+5HyEFq7w1@&c+!ySz| zD8xPHlmy{U_B}v%ufVlR9bJf#-#<*eJJ?KnG9$i>gZI0zmmF_JK7!M6_=tF$PLSSeu-pkalcf3_2r9W+d0XvB`H119vrNryHCuSMWmR z>;Pw)L0|`)6N{z(U-owQ0R1o$z7SL&yOac%Q&k(bZF)4KDr7fU#u+ho&av3aMP(LI z3_1l~ZE1k$S3umaAK#O9PzoH8Wh|IF%$e>pg9Vm?eg`*X{Ylrbb6KsrJ+tqnO6Z9Q zF5#W6cS5KyNEID)QMU=QYDZs&ZY+8D?zvl2L1< z3g5yov^b+oXb0rYdCIf}I;HnD_p^J>)DgS0?RkG?GLa5qXc`IQg)(gCsH2`RvYYgl z^Az$QRL*&~HnA*>p*Fw&JKy&!w*Oi*9srHI0y2+ct~DeZ0c&P@-uFCsvy7!dgV^eN zr&Fm;_ga<;H~ojdmF#q2)L;|~X?N|@d{qX^Z+x0ms_>;!ecmnJn?OB(>}rtkA>^@* z)z13|rroRTnqg^?c=)DE46NxJ(w*~%qU^(8j;)#jRl!a$j?Xq zOC)WGXW^|4GxKzWL`tOl?c3&A8gBdkAAr7ZFX%?B+XcRUSs#Ro^yJ@Eh|#-Qs{Z}X zc1$fb|NJ_B@A&=fCi_7Mb4HfJ9FNy~*|>D~AVMQzA9GuA8Jlacz_ImFBhSBoF`Vc`&AsFn z+Jt;&i>+A9$vH;Iv=n@ax_&qY@$=t*@zuAS zSCcc7{HKM?!&;%XPDVUvZhxt`j(M`xs>XjKUF>IjElat3z>uz-3@);Z2C*OnsTjk* z)-vXD4t@RUe5uodO_b+GdeFvuGNhxO`aJX?%T^!$BF@EJ&f#M)u!~2V7!k@%PyE}+ zo$S%MFzC*ldfM;HKv|5;(<*V=p6y?3&F62+&Sk>rrvTXo2@f&w081es@7*cp7CU7e zRtWbqn(4iZkpBGVUwrLV;-7fIKsCo*UVP`uKSDB3k2&7LG8wsy0}V!Adh7NpSHerq z*u(f)Mzl68Cl{K#puyrPtslgL>0SaOZhctTkBS$Xn+tQBHf6w?Y1&jCZ6nJVx_aGAX=~(&iZ>tEKWh@NxHnn283?GONw_j zRq|lWZdk9z2u4v&%-Xt6EL2Vzjtod3xH}h52FwW|172cElhkMRLpZe(wo&5Z*@tJr{_W9{kru_^6)5p#U0kCiitCk;S7(!=M_3#|i z@|0Uiqy4pp*F;r$>F&?Pi|k}=7aCb`o(^OFX~`UUxT?F(4e3s+6LYr$^UCK~MVo|8?!l!Nkck9npX znMKX%NX<A$D#SZ@2uZdxgRe3pwcXvWx5C@}&T2?85ys*Yki z(U}>0N|TicnI*Q<@%S%+{@uR$B^1@P?EyT#xG30fKN~#EjH5Fqm&LA^zH;i`i&W8J z2|A<)0Rj!j@93VOuo%ZXy8i~7*f#$ydiBk#{Gf{a`5SP)a@&MTJUC*dByO&G^~HX} zw;6wdo&796Oh}q`tJpL2?bEawI=zIPPd-;BdN2Q6GFPAEn$#4hVXfRp*Yw0=)OYFW z#{b~{zyH%vI{j+=Zi0-!|IYmOzs;(T{PkoWSZw=|Yq#ZL|DVk-^u^?pdjx17t!%z$ zbL@)cag%9wNG)zjni+nX^zL3n_Sl}Kh5c@M*(E=3sAHX)QFy`m4;%*-7A94L)htsV zF@;IKV(N)e1--JBUPIsEUc?ZvV#({BDzV9djx!g-rA)(U3o{OD-3uh0mm5A@$_?FD zVU0hqv#Y&GfETz%+%ci4*t-{_Yw8^%e;W$YETCLbm!aw^4KXF{bx6-{Yj7Uh^pQPZ zUkGkD>&Uk(WW{F|pmr8YU*JI`FzNw^F9b?2aB)Pk|;!R#~YKn{NSDdUcN998i1!|SR5 zr<&f~z;c9*w}q{5hpwSz8NgF|3`RSw^(flKvR@*M{ZgwzyEQcaV!7Te-O{r54?RxO zYNOrpwK|VGjBKxoq~BII|8hBydE$O>Mg;)Ji+**ORHXSHGg*oAy1Sw$B{bbycRpQ3 zX1G3{I{YWMTNd+1_32vnnigJZr~8ZIp8XIiv;0nspYHEyYWYA67+aUAc68`wm|p#c zdQW}Ry~3I-8)ccwSl2h7Ppg0Gqy1pXx*JM-Pk+RF=kl7Ik{pBa4eRp;| zy~tK1wg87qdcOTjbNBi8o>+L=s`<+q?L(U-lK!lMPeYGovn-RBPOt7UB#Dz=Bw)pu zIGr%L!e#kr)yv$PE!kWCZe5XacKgrx)hhG1x|;mNJ5{4g(e%s|Nb<*VZSybb0uZ@I!w{_4!s z#3)5e_f_JWZ9mR7evdoN$Nyzb;^&Pkvm_(u$`F#%&t_ey+>>6k!LhIATE1g{RS%sub648cwkB3yo|}g!E;Cc)-o1Md z;mgU#>YIMN@+3VybZ&0$Vo!;;(@3kxq4tY$1Fe~+PD72ma#~FSTfHV;n^=LD*3uzgGVPGB-`LWkn5h#4S8O%Dysn#G z&~g5jayR{6heP!~)9QV>K_;1Z=R*INQq}D zk5~Sl$IEBi^~j;c%+P0Z@$saQE#j(xh5*K7}piz_-hPR!HkDQC0fJ{rx>*5Tz`I_(Lq_5?nXmC?Me z({S^?q2Xb9XZaM{S05ZsMI87w#%FHuu9UOosc0#&3$;;OK716MPd%ZeR6S3hDRQ5H z%V--J33OOzDUallLMR`E@scI8uk92PRHEga}LFIwNW=++LE>~8#_=XUm~oNV|})0eyniMB3XeD!e{Eo=i(pj z(!R^3%ejTx-crIb+L237!85rFcy`C(-0I4(E*@e={ho*Us|+m{)cW=&SJYLBR|WT2 z_Hr?$>9_TLiv@RHo*R=Y>@@$6jkyCf~Z{491#T6<~bK9$* zmtooQ`t@rOvsbt89xyu@#3izA+qMFy!3~Av5>F>`w`|FDcKUGx1A`_Hor$%)68kUI z#Y9V8uIv2}ZWKaa?4bvvWlYq`3+&ssPqV`i-5^Rr1l@q0z0yRazG5xxw?lKKWQe|1K zXIH%0v?}PDJdF}6V%40|C`s1KG&%lxA(vOPg`nM7FEwy~pgvA9B?|_{PxNZ+2|Gy{ zhzqTPef)QP}SAYD$f0U4>|!%*LZv*Iv8KCMOXh*>`uKJ=<#N zAhoscbtGqQr-M=uzpNC;UY+Yw=3(9L{n2@g(;cMg4!h83*KWxWsi7xm!@%w&sKiyh zjt~`GwT}C?)MeFy2=kb0Ja$DQ{GKD37cCM;!(|kX6-~ZXqP$fSt~3ib;8@Lj=`gQj zjY#H1t#m>ybL!^)Nib(?yORxy7F4Wi+;Uxd;+k}?Nql%vmDis!oq1xl-QnfX&6lk~ z9U+Kudt+Sl25=EpnGX59GhMDmyk6twlv}Y2zaV4qqOTXXm$qhulBDmzby#jq{utLw|PB)>b9@ zSTJm%ebpB83SwoTM#8v=GRQS!2b+?$Wj?;xUjO#(O{u|+7M7L)aP}4N`~>(3e6IR3 z^CMY_EZMv#H=P6W^O8L0FB<)B@YKOET$4EI7x^%|$fJ0I=|!TX$B)IFj~x34Mi}p# zLQ?w2YJRSSXct{%GM#J9wS>Y+@K65eTven;pWT)9bQ$`ED!7JID9Eu+RvgZ-luORb z?Vy<0W0kp@zM6=QmshO|@Sf=sr%?&=3Xh2&EElrS5$;gF2S*zEZni%*^f{ZnOo%*< z>NeeGuF?PO8bJUrgPvJpERRG?$1hqj8+kxFy@7laB5vs};7u)vJ-Vc!{*#HK%N_Im z9bpJ}e~JH28-URbxuq&6sUq?2FSpU`7WMvT`6nN5V4UftvK98TAHo$)O-+Y>X?6e! zR|w|bJWnpgcY9P9hj^RS z(@W+odA%p=DHE}DnQl0ZM#+ds8FvS-nZlmnH9~nm1B}qkTfu!!)UBt$ffmBj!~{o+OepQ3GsJx zi&k}4F?lAjM~_#@{i&BS8oFoY;YiAhl{#x`BE^Nl*yZJo|GmbvGGKF}W0EDvPkpj> zdS36JYaRlOa9(e916f=9)^mQ9Aa3DOkCVVBcLJcXA1_}`Frs}7ZgMC@cZ3YM6JmWlqTKr;GB3URyHQe849us*;G#-re}3LR`Xf@C zqbYi^5b|`NRlBA^Z`_pn;p7lgwfvJUOL$EVCd8ldQ{*q>9=j$O<9~F$3x?AQiEoaw1zR-4KV}M80e5fLv%MY)cnOJ- zpm}c9hq^-v%y9Ld+HP%i!DCxL2MosR5Q?az5CjCsMH4DRx9^P;1BDgpVq{6+@Rg$f z=C*83Im>WHcSTIq6B_?gPuzsO8z2=itrVRbQhvLU)FZ7>OMFECz)x?1UnIWR9vURx z?Y7B>CIp^V#_F1(4D7afRT@hkaoVrH%`b8K>0`QokwD5_1b{wFrM zKN?p3f$QMK)B0YNN;RRDGF2V1^Y8w?Vo$s#SS3Ik_iNmhaxPvanp=_I{d~<`5m?3R zd=O@mXV1JL0?y7qZkewX_F|cIvL%^7OrJd$~0QW2cA%=FXPwG__ z%mb2pA^X|03d#Pcjae672Q;Wi?*r-f6^wD)<}owISAA!nQnY%OAv~M}7=6;*aC&HJ zPWx`>^;`15waI!~p3c_~DZjj$fW#!B@fpMi2LL4T^+&rJZu-&Z6q6sz?t0Xgn)&RE zMMLB;ha+W5#o5MVqi<+}PIzE-p?JlS?cs#N;6y9A_JyjGhz=21DD6<(8mk1YZ!+3n z68mR!Il@dQ*0oRL^ZzCT$!5N&23QVq8xpc$747(*JQl6 zJW-U&`N2VQZfcm5fqtIqa>;r)?bXZKE0q8K=FUwEPnax=RhGpXUDNV_labvct#Sy= zw8V?Api-q$BSV#@ZTv}NHbFUyJS8@=q6xg=rz}H6OX`InG%T|e42%JP4-~0&?10A7r6v76=OTk#)TRW6bMnBcvBIR^(Ex(L}^Qz=Tq`DRf(HxJ} zcR9bZH0>^Ow;#zQR3)dE+>^R;=xuvD_u|}C;D_q)1eJFy0W?KCw z+OHzNIC;^+;smSI@pM=P{uWhaCE{GQ+hD?L7r*=Sd8XHKbUeeBLfU^LQMp0@m=dGU z8cc{4*Ag|fl#cqX4+a2#5EdI@9_xJwi5p;X_+QeX3xTGgvWFOG`^uq9ufSSjseht^Z$sA$B3bn%P zD$VFkLUJuz!vXkn96G+>q;Gm#cTb|m@@3v1VlV^&|1bo}`Bv%0hrm>%HEyx`>yMSI ztKvfC?nzzd7RS>Ts4rE0HZTHPDtaH<@d(no@LfKzaStcLkliNzNUcj|`!ZquC3ID}*idkO2xgkTy!k6-rj#y=e9c`k zE3KiCne3b#1>g@%8=t6UR&wY)s(r9Z3}R(+cd==0lw_htcW$lJa7DA89Up{2&OkFz z7Sr?B^GQ_zmjD!jUXhzUvbyoAvHaEH!m81WCFU{iJA}ZKOY}!Mswa&9usGLJ8UlVV zzP5K|{Hcfy4C4@lf7*5-vXZluS~>9cakad+(doS$XnkZlA3Ayzb{By=MHb)AkB;w; z#rs$nj^@;->g6&sW3Yn{x}=&a>q-m-IOV#c;HcB^4>lRi=N3@YQ;VcFEZcD3?4yYa zwEnR*ODxWhcG)-`kPRF`mZFKC4r%Xa+r~MHuMxko7 zv}=g|7>T|iPG5*HoLN1-fSbgIATdE;dRaKAp3<{jCfl(2pO~}(lT>vVf`^nXp;%7g85V2{0 zhJr+y@v@Z&wFaG(j{$F<5anPL71hWZ{JfS4e^XtCFz;NVMIzuZjz00c7& zGkCfya8iS!UAp+iDbw68m%uKUj(v=<0!Dasy7qXBm{j=dqkx6 zo|;jRN+2w&DmQzcM+<)dgiP23f|~&3SmKGRtptY0N&%WE21cL;^~X_9X7d@o%xO(PpG#d zfEz(z+O~*N_2PrW3GQX$20vW%oYfZ+>dxKj5v0%Wct=aU-|JE3Yj$s1BiHuz76Mbw z-B9BD<)`c(^0mFPOWfA_P#NE*H&-|fxkNA2JeeEQ(zDr&P#J|pAfPoxgQ`pfkLGj; zA(sN6#F1v$pifE?{ECl(HMgWMu@_-r@DT7@Dn_v4g#C@5FAzE z7YO)cnL5@D!OqQR1wAGQ*cs6~@HNmmG5t_6-F}}@iOk##Ss*nv^~Cfn)rU3_OMra@ zd%OZs12VJkZ)x^)S!cZ_0xuBs!^R!QN_`ehfPRFvpdv%5K%ol}R3o@HS~_>O zia?=~LJ8yVp5L-&qDOJpql38}>??iNiO~-bIlF}j->b3$+#2KryETSmkF4^|A>LS> zanF5$LNJBEMYGHL;zj@1#SwFRgC#sz-DmCxLtA@ZO`5vX*Byvn+R?pP?}uC z(77aQpEnR4{?uaVY18C7OSJ_fb$9={--g!8^@eF1UJDd1$xDj7^4nHM$b*3CYI*7r z#1TcZo&cC?cxW((Rqod>GaEtK0$jIcJe1O4~ zaIc4WI;|dzYtks~=!4V#rXi9$fPRc=1$&uXP;62eATgYhl{Dwggt#H=WWNv|m_}@&6N;0CD0l|i1*X%U{;O$Qi$AYOV~>D zE`A2Zv0;y2vkQh=1y$W)iCSMAVNgm()7_|#Y3 zglu?b5SdPPpz0Kx4hbsX+XSbd*v_Bq@*zyS>h>-Th;COPHU?~W&Tc(Wg<84!5o%Mq z&t}j`komk2;JflZ+|fe>W>S0KaZZCxz%!+^vt%ioGLIDM>(>`=4G^QI1f!H$de%snev z9JNEyl}UrLI6M#lcSn zN;V1xF;ZwrGe~}#Yh3?xC!#Oi4CcOxms2Lz~VueQ|xOnzpufSg~sI>75nAkgXW#Ds zXDC{81o_~>gB95=hCC1?DD~;tJa+7_wV!w`PS_D?eRH-M9+U{GYtSd%copT>C;)BE zO0r;;6_uPuyw1+qbYd2MpdmqZp?9_XlUL2k;$52{UP^w&Auv`nQ5#qa;sU{{FK~&a zknB0o^Rj<}64|q`qrjCMGpnW72Dvk0e55S{Y)qsA=;Ac;vhQQ5@j7D!881kXZ_Uw> zN6`tl-ZmAV43RqrwVBh%ha)Y36>(GbHejT>P})(;7Mz?2#3D7W2qsYhV#my?=ncPp ztkmh$pq%;cD$j)yG2oi-ikhqRNk7WAZhM*s|9GIM4IC5^SoB~>=D}*EftM!vZx4Cl zWe%v(zDZNttmL_$CqlPi&rnc~2-BjEul7Wc5!DThcW8}kyDsFWNO$yXuYn^sFA;+S zcLOf&K+kdmtU^TlALwkx4F@zfdA-+4uydQ~RVBvQT;z$#dhDg}~n(VjLy zLnMMc4NmK98$vRWCk+8rmY+xks~|POGSTjB>!72v{q)!=kpb_$T>U-o*5&eCji;;!} zjlNNx+Fi=q1xelHkNpU$pu|dq?Uw))eSf`HM4ICvq4^59*n`w-ebnmCW@f`P9+8bs z&Fr_gdY@|n68@16u;#f{AKFK?a_ ze^S!d*Ov$A91Z}9RHKIKYp1w&02_r|5|@nQdBEhNq|hmGw>p&j57&1%0oV}%RNzv( zK`J2JVHoZLH5+uV5y&n@*q$ND^DsZp+q3C%oZ8NI{w~gmN?NwlZ%;0aysO%&&7y25 zH;I&OiI!pT2y{)Ma0HwqY)3MluctsXB;OFcsXf%U;RFx5d%wTw1ReKuf z?A-JydTyXdegc;2^&}1KC3ixMhh;p^uJ;qOC=Ea?Tges#33T(PYOpd;Zu?rv{__q3 z%OCnl$@ZK*3i)qy#3%A|idP?YwkbVf5LKvHC^1pr?jv9Qm$)A3UcVCMmcful0uFb^ zfJ44cJXh;sNtzA*$&B=uc|+eYD~n;aYhEWAWnfQG)Wv=9+~0q>f$tFQLNVg|39CL+ zW3WGgq)>UehF}QPkAv5X=Q+~?vMX9#ilRiwAB7JpoJ8AU4Hh?|%LsDTov=?4U4jg| zJJ^trP}_SH<=Ie5@sv0eWNZYO3qqZU66y8nMy033A*st3$SGXw#DxBy#9aMNNNQ8Z zbS!+G5?JF~H{d_I#{R`2q^3?UAN9F8XrXE{-=u0FbJy{F^ilkB$4;9uiYA=%{q<}6 z594&CQLWTC9vLGT?$DfDvKb}jVI|Q{Z%Ny{w6HWs7ZiIbS~?loBg7GC0;W@maV(Vus6t70 z7P>1eWPvX2X_)USa1KPp1r5Qv^?;#)D^t2KBLGkvIkF2gi6N+x1k$QegH>#J>hn|N zmH?AXE+{prW!RvIZX=#(ps_^EYoiX8dIDG_fU33kxwagd>A~Y6du!%W3IK@`P4*JZ zn$La@K<2^-7^jFXJfL?XxJbAkvsZtzuESQJ^D-@tQQYU@A}JKz$P#|Z-MHiMtMhN0 zCJ;g7Av2A${Qa@|H4rc9hG3MpiRe!M)!Eq@k>(d%O0%=YAO#!N=AA?Ws}TzPR?V1Q z>Ed)Ils}lels5AM;YL9ei5Kc zM7R8Ot7+Wno#zYGlu*w36giY32FrIIhikn0JZDeUxsE29r?k%uu@-6D`C8a58s^9LfVxJ5Y>o zR8utbC~e*pSV^VLl*YC#qJsk?4}X3r0;t$1@c1@9xd5{f(*qSYb`u>|ktkyl1Ls#W zCVueXG04=378(Uk1A{^(a+cn0TSSQjh^R62Xp}2Yv9LAXbhPwsCYf4A?SQJ#VWzyf z*V=tzXoUR$6G3d(TN%adeFPjn2 zMmc=Z!XAj;C4ggubtSKtoqV&CBQTK|N3 zHN9fTCutf`qk3URXyfu=oUbF zJ@c{_d9$#SF2Vy+y}m0=AXoZvQ%*n&-By{BJ}&?>HNWFV9-Bc9Kk$g|llVedgD z`?n_qvuC@qHJj_gQISK#e0GvzD8HJKQEZ!c4^tyg7(^>uV0$~X+%sD#I`ibM9w9<{ z>_ry7Kl#iT>05#++kFL@#mR2>jA^ojI}%!_Mrqa0D=p!OXSd#zb>+6 zc{#AjFIM&t~>km?zDL?-)~eXRo=0QG{9#`FWz%B z;kL<0kUC3QbK)IF)uUXv)CyeE4ybANoe2VZyI~&H^Tf|s8Qj6sBpv%IP}o|idZ@21 zR^XPP>i-k=-T_gc+4u06th%ugqlk!ri3RCF1w;fi76e3)D&0jC>7fo%O&OJzKEJ53l1 zjd0I~il$TSO4smM-*2Yh=5yPh*=m&m_5wZhLuDI=MxGaap_Hp=?50$4e)ii1+V6JW zv{ndu3yc3%H9`A5K0e+RY|eqdUC$UJojBU*5<2~KRADmPw&HE%{p~+|e{$D4Yq6h< zql|sBdAw}j@8p{^i`uC6OlvLiYRZlPyW>$SJ?Ml zcVoGV?)1{M|L8)?*`oz%Yi+NyCB5b~{lBHXZg68Hy(YYuFm@-%bs=$4KOc$-U^E^z zjINE4vQegXLI#-Vd1MsxGacbklOLsQTDAi8kw4AksAS;ra^o#lL`TLgv1RmM|m#LV~ z@AU4J6(PuI^NZM>{MXvg)LeRJ(b}P7NUsWuE@%Cg>!jy3Lhvn?UwsRp+<`3$5Dsh5 z=v|AG6g4Zm)@)I)#cuwD>4D_t=4MO-Y8Q6z-i@(r>@=-Z*m&IcS-mBED12dx+|?Os zE0?bEwad!10LEMZQc2EPh-+b{b10%JYYEoP1#dW|Q~UsBHEnh@072-L>$l7Mc(Z=a zjEf^kzC*f3&e~WB z;2mB1Q+)B4#YGCkqdb3oq~meo;T2Z!{-?;k4W-RiZj9&UvtF0>dA-cho??S~$F>a0 zjd3zX zbv8)1c%j~ciqzn2+xGAV{;ei+e{yw!Q5kk6R)tI2PuJ@b3sgz~)_&3;2Gqvqx`je1 zooWp!23$zG)9=h>J$}%z`p6`gjr+^o=OYCr1_k10s% zpHln&yeO9(%ufwy1dgToywLQ_%`@#9N%dV&7f>GZMzYqG%&>+8 zM`pozsnU2Rqs~h~X12{ylZL)mlSa1%;1DY}E(Hcc!Cfc#7t}@cGGe=_H*FzG7?0tn z?@#fbs1TsFE!}u0HwY}`dk_@HPqT+|{YHx7LgnRR+x?A;R}J5$>Dy0BU-9K*3uTuQ zPABz$_mgE$v;g7y$au@Q` zAcSgES+2#{caF#N+x;O^7X61y?p`lZ?^<4Y%a`xcId6vg?xGiDWA@tvT8z^=`Td)t z^~E^lz9BfN=Ju~5Y|9(<9Cc5mhZYWBPNaJ|_J*0q&tR-sCr|0*@EW`E=<(yTLnHuD z&I`@ozUi$A=K@(f*i48IYOJL`EYOEFNefyKg~{=b@_8?g+YiKrx)f9%e7rsn61i$+ zSUl?ufR1NtZ&w}BY7-Mdw|g>1biXjyev|ed)@?pEz9S`Cz~@=W-6t@syxfW->R5`7 z@UjJKYPx_l0}HyB)51W4Vq73mLaVn=$-E70LLwblxChCd0S8DG5RRF)X4}BeIgb|3 zmLMa5P+k5)-=_4Jz+{AWgUCU0s>7a%*6MjWrXRUv{-NK>D<28)|B`v4G zKFA~woGv+K66SwmrB#3NYWa2x|L=DH1}vA&P+d?SA^~D}K2fW7fqvE6S_Gr;tR;vd zA{*iljsRo3ZVAdO;Vq+z;@2uvG-%9)6?cM>DFc8xrsk?s0Uc7(dWvgwVq#)bk*B-9 zH$2vK%lhYVpWe2`WZSfETg$?@~rDrs+Zu^lfRG0kz z^Q;GR;niRjZu#^a@y4pCAaF_Rras=_h@UZ8s11#uq16i}JPvH=$S2wjH>K}Rp-j#K zED=jqiPI?fyjNUlXg0}@M>>9H6p12U#kt^nF7;JM@{VF=KOT}(5Z7(A)DFSdRyfF% zQNythm!GWHPtn1$tHp`{YT(r@{(W9F z)R%7+Aurx=3YsjN>9d`r7ViNS6WC2%9%mZLl@u0g0!&?8q*Cm>j}*L< zdtPr@$~+m?=T0C@kxUfgwiUZ+kozTir;6;VO6Y1*ciH^i_0_S=;7&hIYwF*G_mVx) zYY7*gfmM6ZqVVazTOhm}VQwN|U~*)N<0PPCv1BW<{!YMU!19I`_VRhneX8?Hd)_#d zer>WBJpG6(jmvwStJ~ar#$2|eC&5u{e~2TyGbXgp6iwdhj^!-$-22)3U{^FxLia%P zia7eku{r~cC#cl4Ro@;{yasQ}4C*n7bfol=J4caGU$+F68BN-|PAJV?GBy1vrP#JT z59%(kyc`#53p3~(mLuzWc5NZ~0P(%JyB!!bWe_xV_!F>lV`~9`!ieNvg^|V#o z+~=FS(+`hYKJjnLE#}qrvf0-{UAaiYA-16BICBHh3|SU;@(OuY7CYZl#E!0i$!~^kNm=k3IgqxawI>%c+3xPy8Lzyz%Fi)t~`6!uo#; zgtC16Ojt2+i}#R=0q?C_+V#emnD0VWrp^We38xHuQBri0x~bAd<9_45x&_hQMw=+P zQ5S|9(=do(zkg%aUQ7SAEEuoThvZ`5MrZgP{-ezz>_&c#Vd1jzOzhuGh_ZTnii7at z`}(12Ok8Uhu;L2nQT+5;%#<`~vc!?I)Pvs5q>5V`c{FG@cDc7x!CR3#RXS#iW=ny& z*fKtH*LW)W<>x?^BORA*XP%j+7#AM7b^G>*g_Y)gXLY`@#MXqg5$e#7<_4R&JTD*F z^e*czAFCg@zN+CWr`++AV zg#h`5tNIarS0H9h^c=Fag(fN`s`x|a**I}e!KDGxQetRgLcRx34Hk@c-zJjNchtinL9ZYf;S3U)9G7#z>FlWG^>7N!G7B};xX2YzGS64Wp> zX*VuT57xuwEhKXR{wIKyr-w60T0qBZ;pplAm~w+vMtk;lRVOXcuI_W2AEV1SmsH+qhR{{eC&GaOfwo7O@^!R!1LbMtOw0{}YVpJOr^ z-dJL{SZ3!`P`TI?vP-92+-LreoJO|>pFA4%k5_x0aKFJBm8W~ibm z=0>=F^D1sCT{}r9Ze}fX{=i?3@0jZN*#*~VlIw8U(p&wZavXNS>KB6#FDW^xU%a1x z{*5ALHN2x@KthzC`-20rtwkBGsawnsTp0pB3TWgt@MFxFSYjy6;Z2k(n!S+O&n9@Z zoxn^EVb0<+Lm{_+5&FX&YQ80ioK!nHe$vt#JCo?g)1if@ZV8G!8nm6ch#!!P9eUn^ z8%Q7N?SneW-QtC$Vr2^?36Zz6qtQ26itlA}k>_gCwGXl*<#)w-Fh z^DwX4xolr{yDHz<{~#?{$DJlzM0kH5P`=fV9z8l^6$-n;d_(um6cQi;ikp#fXeQm1 zBz{iL-<<gD1@WlXy*v4CC)4=f$phQN{AW_a7`rW%I9m7Wn`Uyz$>iiWR3`=K8y z6@8z49((f0x?I-w@trG4le0anKl|cWj5x;b{f&0z4S)RO#`(0y4-4Jh-Ao&Q!wk@v zMEV~u=W71q0(@EKmMvQn>72qku`_FxX7XAgnX)X7K&#o5X|7hD4^iFS*@8aEr6<~& zW87}g{{lt-TElhYEEjtj`H8#-XrOEYB)c!=P*xZy&t&F(`#Y9@f0_ACfmZ)t!KA3v0$Q~@0`y52Hwy1cwRMr$G~_?H{4OM;b&*^c=-9%s2qV9NX*bYt2}!TlI{S7A?EL}W8+5Wf~v`G!}kUm zvy(tdh?`lldDd3jLc3$QHYg_cAMgE@vL)r^NOm)OTp2}K@l=>=Q@~^8KZI#FmH(*g z&t%GjR*KSa$rx4Tp+8iTLMJM()&HY z;qfyZd`BOFhcNTr)U2R*V~=~lV56k8oE&ecv71HR(6oOJ<;YO-9CYK;{fPq`XqtBR zD-#3$0{Z0>;}cv{OU{A1Z3lQc)UsCI{u(fC@td@jPX!PNMjvR>7YG|30kpGnG3K@a zn1FB6WNa83dEwjV+Qvta$%pN74$bMsx|d1oLFnNj?C?r= zHFd)>cPoGK#XUOyGc%{&t{vL!J*N3N5k8(WyMa17&a%2)YmC6tRaXfZJ{a7G8A z0Mer1)UR$y%2fpdX~vEr6-Q4c777E-LaL1iSkUe6 z%3P0bIvwq>Sh4!`c_Fv^GDov=I7jJaXIC8H+CjhsT~B{?d{R)hY2oMPRi$!-5BEvR z$wju9+X9a&(bLM-D>;{x1rFa8Amqqeul6BhkodML4ylcI96n!I%I0ebFo5^kv_$>O z{$yr4GRkjR(#M~?!j;RVGD`&~zXD3$U;gmnLn0mGH;6S zypMLR3arD7ELCpwn%Z(jAk{GUfc)5R98Lu)eL((+n~7ZCD?qz8-MVsBqxo~gPhCs% zlrc#fw0l=_Lrro;lFqlBfW&PDW)fbtFIZR=kd!Tf(7vCna23j3mfjG(ktuEa1Ozl` z4Q-1T92q0f9PNQLNkJ+!L($X0p+A(04E7XzW9*14heWGRp7k98#-R>f{|G!o!uR29A>l4;^~gGJ0FG z@#-N{;JreTuz7q&qI~BTEKHT>f3A?{GE2)n-pwB65{?zS^}t=48}2B!`&`K(-OX6? zDrAqj`j7tw$YV~!+?}LGNAZu`I|k`@ecY7D`l?^HCqlB(AfvHuco?$rElb0cB@rZg z_!BF89GF~Ya3lUX6+wlH^_FC~ga2TXNc2x(WY=klU<@0aJH!0)U8Gmuq7P71jnes& z)(Alr?A0$Htb3Pw0rm%f2iHMzhaYSY>=yHZWsr0;o6j=dkph60A+Q7muA^~u_u^T1 zd*DMPWK{%tif!MqLz&94qmT7oz=c}SgFB>l6(%+D7$=KV8*NY3uZOu5Cc`sDm%|n(3=*gJk>54dNPqXKpMO`hoFy{s>qo>y84M z9`gxo6~hLUZlei^mP3Z+glLifV!hc;Wdn-Zrq63et~zGv4_@UUJgI_$*5W|?gSk$I zg;02;XQXpo_w5$@3e^weT#VK;4^=vX2(KnX$7zdsr9BslV>8V;8K=MgOUH)MX8Ak| zX4Wy1^bw&}g-xS|R(d6c1!zpHK9i!uu^6BK>azD5c#?3Pe0*H;rR^@}BY+Md0q^Lh zjT=jWW?*{t>a%W}39#;Y>Ah=|W+n&g*~7y*r38K>K?vVnjjN3f?o^scY;E*%&+b3J zX{do;vO9>_&~0O$udtTjx`gQ<(WpG=>&5OoU8udtSZlIeoMXNY$%v{tuS6|4!*BjvDTKp-Lz|DeYvrSk#9oj z+!q`tHHCT@uT^H`WlgZl4Z`P9B@2>cwaj?DbfI9JDwJENwynH@Yfb@(BAUi~=Thj@ z%+1Oamy%Yy$NmZX+N)Tjofq#V-98E@klPp&d;~Xtv>9am+~WC%mDw}%`-2_CE#58_ zgc;zWnlm#-08^=eW;j|m-BAKJJG7$E4t*{rbcIh23^i?xH z{KE+*>@Bdbb-X&lJi7kU>{`~QWn;ZR@v<%fv3_?f><uzIdcYajI@A_^a;Itx1kFX4IU%7kd!Ex#n8 z@$xi2F(}e`b}LrR)fue-W}=pB23n4cN_RrPjIk%dsKFL0vQU^ObI{AeWA+229e|BR zafOhC%>|px2S1c*zs%+);0|Wc%LMD^g1L0>VhPiMX6-SWeuX^%P``hgS?>{EuJ=4= z)_YXrG?b`0t)&dpdNn8~St`&buh{lDjMxj%%&)$8>b|C(RBDo^lh(-0?FR-1{Bv|u z^f^ZX-aXP0x4n7vJNLRpu>8g=v(yW&APQ12y7gS`+Bnjj*(bD%1D{~F(lIp)L)|-AX+^3y5Cz^S+nRLAMrru<&U!V3=UOq|@Ws>OD z%CtsrdSf>^&C^CX!u0>~ZkoTVthA?6Heh+fmVW-AyeDy`Z_9AY{JTodyr$4O-(m-z zgMK`SUXY<+*Q+De$45-gogXNh6mn>n140)H(IhO&Nn1C4uWM+3syu&dWo}f?O18w$ z{Yyi4$a4zD`oefSr`$8`Y29+aUXuLUnn75t$;>dnSC4bH&c>!$b zb7;SSSdGq|rJn+xbD)e*`*U7?Z@5(#$W&2Xb2Ji5py`1h8t(^PBTT_@{DUFc_Pel& z%+`by0EzcUPNa%k0bUsWOQps`A#YTN*SBVZofFHEPm;ECnGWtY?PuK8m|e z62V`l_&^Ily!Y;kYHzu4aVvU_}mb-AC+jO@SE z6PXJEdJ-(85R=trij|YoA=nA16flElFd4&gGYRu|i|6jnRTGPXYKx$x&~-}O?L5eE ze1t%Cjr|Hccm4HzjN`4Y9`Sx^lHixSLW{p!{ED?v`ufo?wb*5Vd5~%#Kv?ts&6_ip z#l~p~5IN6G?!ctyptMzL&b#XB_?hyuG7TCd2^0a!%H?O$AI?Cl3y+;gH;-n^)FyW= zJ4j2~J67b)7scuw1m(T{IFr1CO1&G#uTbZ&V)w*fXJxTGzOl^a`Ss5;0e0qdvBq=6i{tPOkSv|6u_KhPe?@LsNHIT-> zzAr$!N~Bj*RKT8uVG0p|h(egbOn4=ROUx{Bm_qCHfTmLa;-#VDa~3AhD>klDJf$LN z;}#(I{2y%Hi~EH=_w);{oQUg}*p$@l@7n73K|;c4@p1#pz7U*rm?Q(TGP6#Wv~2nB z1q39_BqSt28hXXjQi!Z@5%SR%`x(Ij%B8V%zg6l0f-A5Fg4_7 zQO+whw>KQ14x{STg?nj+@BQRh&jS|mB*`IhE59fGOft_XRVAEXp~=5WS}mW!id|lO zwOpzFD!BB+)d^rS%Cb6WP(}Qr!7itD05BLcH@73?PCrQ^5Yx&L(Qo00cEn47u)>}N zIUcvG=%6sKKkI%F2xA_iJc6a(oriuP^A)E|#79=m4bKjp3SbSWj(tf9spF%~-yBOo z*WHXVbTsC1p|4_sDSdC4dzq5ITUt^wxP@Zu$Qa{be2)Y`&6eW%h+=Mq$%6_(weiG% zb`m~^>HW%vUL)#Aq^u-f=^kmXT7j3HSVpK)nD~?thEv&07!Nd$V2^qS5=iH$;vfSA zfEhQUcm_m8F`C}x)LeeXT=_W05pkZt2*Hj&h$eFWc_pr8NLh*R7CIGRf5A_uq|_wR z|IR>{cs6TH^Km}u%ado)#k84pv2Z4W{6S3t3PXDUmICpUhYkEYcK8DLktMGv=HAW= z%r~;anT*+I&;uOjWRRj`h&&Z&pZCuJE(?=;rKl*|5kHGr*geT5eo^4Xk=co-gh&mc zVAIN>)28jNb|y`W{ZI@V4jH(b)5?x9yqc_AF1~)XX~2E0Tl%s}Bt;b#?5Ab+zR@EV zc0)6IBAvN31^R#*%DjcuF5OU=yKxxe4=@Pw0O<`hE?@E`@0k&~wznD8{0#FH&x^Z# zcplt{`X>6#A7@WV)E(F*`9r#hb8WiB$;hL<-zCjks-A9gdU4jecS5V7?r_|t4?b~m z7x-Q&iyS|?XV<#l_w!`G>KG=^=yxj2(R*bj_dTN-Q}iPmhw|oT>DOq<^rN$d4bE~o zU*BTE;o99?o$`5!gB&(XJrmV0$M5{oz!v1V-u8`4<~eytYd)V$cwaB#N#M)|Tf46c zi0h3(y+nb;G;I*I3F^(HYjhH^EF0b#8C#?S@<&EbaseC0w%E|V(;<;-SYSKFlS9R_N}efjRYGIutwAkgVQ-b4 zjg2a-wXVgpv7upGsF8~v$P0*7`T1N09mq7V>7<(X@89<{*ZCWYalzMAOh^N_J5Gah zaZvwT4U?pYB(@e00wLC^chyFTYgdgLBs;VcciR8qX<#~ix-aGHJ%=usXw%?=eflKI zWusLDAt93Us_VR}VA`h!J_zV70nH%jGEo{-jg%%MA|h~K;eQG})ajW+x1<(f?1;iE zM&a{}d}t=U1Zp@+8Hu9MDrb-}?2tRmO^&pr&a)v8upfaS!;lmdZu;x!pnSnRIorYT zR}*(9r;yB4)p8lMAE$0&rUD1~~&VwE(~CWXZ@y!}CDn(i73F&b&a_EUp>L7uF3684CkoGZxH z$lJbkTHDLFXH+-TbS8#ZuaR~4K8gxG%&>i-O)gejPcW*gU3>pGRJgJ5Qc+QV2Tl6Xp!?T)PpT0+>=n6Zc-0D8#zdxjiIG{XWA zNHPkX4h6+YA4A0@T24WGCA%;d>%lE}N+HYK*R5gWJv4E=nK>{Ya`TZQLhW zI#T$YwyQ7Ov?RCqYCxmx^X`vUte?|i`1GOkjg-MjJ!!)Ax6dhggjfEslH*E~Cf9m` zW9Q+%3?2IQ<12i+#)WRDNb;W1(8Z=ykr#lah z0=(dF;|n*$nU`(NC)2-LbH>yvvy0$_lqMUK^@0u+=b!P61QiU})0`{NoTe612t<jE>!N z|8L_v&n#QuE_43M@~Dpava>;hBgE|FP3RVha(WL6@;|NV`#N(P=>ila42K zqzVACAA!d_o% zpI&2Ad}o?isbqkQnbEg@{r}8=UYAzik3MGMZy&=gWql!I7~~br(2ss;0 znwadND$&DGHUV$b!^UUO(hgmiK|Lod6G`V28w>SA7Zx_~%OAB^QARSOKEAP|>zUk8 zMryihr)S3ZjCsQ!+2e-Ao2H46op1X1JfoPmgUzZ0hw=cvbZ?b6bQ%QqHs;ke3QYG^ z8stEu*wg};^>(bq(2hHYR-{w#Y*%*A$><`1KZaP=_3FCx(aLX*>Ap2LB3x9Igf8#cd10O8GvOvf8YU!Y zH`rIzmtwWtL%65X8#<^VQdkT8qk9^H?P&@lGM15^$FnsQUdyepEKQrIhQjXJc}59N zXnU188c1|$D%gQ72KT_~%04nQ*^dULI)Zw!y~M&RMdg|&=o+J1O~u4DnnnjvykP7X zCl{A}m1)it1%^lMB6IU@*A1`F2wvbuU6j~)mcxhtqAps-k$U$oht0dZGAYqkb^oy0 z9K~G2>O8jGZ}a(OxBB-rX$BU3Olqo_-V$_Y2VHK=0|_zILJU!W-Fnj}nqGeux=U0- zjP`E@9VpI<2`jKKwICQ6*n7efcp84IzFo7YGDqCiQ7}yn8rOrovR5q5M&9|?69q(y zQ(v-ku!+dfy@)0CH+iht3jvlEoS6zuFB|8XK3{KvaG(|hTejmpUIfbMzRFbRB%Ksr z=(`c9vICKa_=WxhYLFEfa-^j^XeP(g4p9IVG!VC| zm|#5_PZ5X&cVF=0@F*iGssmMjLBCbN0d;W3DVc}~{#WETmqC&&hRG^{4L zn8lp-Qw}MF%w>P@^N$faoC_aFOAilugY1>kBp#&W<^e{wrQTaBkmDgWLB9<=vA9@| z(>6^nel!LzY&K6GhR#QyU!L9<=`3R3wClVzwz56lc`*^usc$LnBO;21zjlUPD6Gk~ zVf|4(A+0}#o=h`0tw(Q(aNrR1^P4qOsnm4H-Qbs414%<^viu%*5cp0zum;$RZ#JI@ zi_SJ>uBFlo=Jze6`Pr`em8)G{=@IS-C_E+!}CK=n9wuKQCP^9!cfZ`N%g~o}F7H+o zS3&wi+V9Y~+XPw}fo!c%P(~+1p?5s^j;choOsI3P&B92~&6j=irtEggpaxEeCn%4L7#AGYTLNtlE|4tyhexn&%Jy9;QNHP*%k$r2 z4yQh9p)_xGi$yn6S0zn{Xny;8Bhc<1I zh*Z>H-&bV%eq6V&zW z$9yr#bD_6ec#FH4(WM9}my?gZ?cA*g&JvBeS$~pYfZ>MEvI|^ZA@ueyr&%y6CQzh! zo2cQ-pDmVv0vmpaM6^Y5Q|K@ds08wKIhela?Cfk;fa(K#0(hjwn8C!eEI8(WQ=a|9 z?IQ#`w#jdlpkKd*5nVJn?qzR{{|d{Y&dl}e`j{EC`TVzUnov6E5Ce3~)xu4zG>iiv zb7$(Vh>%i)CW_oa89drhOrSDN5)?7`qc4c;t5leJeK0D<-vfxsSq=pK4NG0?aA1H*n1ces@lCD0fO?ET;j;<%epvJ)Cu_EmU)exTTe+OB~{D%j_SROgIE&`MPVb;Hm;FKCM! zop{GvAFC8(+VeJ8DFXt8MHCzgG&8pW%?W!@0o+@GJ;2b_Q}Ug_BPa*-TQ%P;AgmF_*3YqZ^batx7>_ zGw|~4Xs1G%KW_wFVn7D4V9m8`JB6|9Y=3Yqy=xMbLlpF+%q-eo+SfESL8x1FF$rE5 zl&=nV`Rx7&BTD=g!Jo^c%#>;NiAaeYg^N8Sf1kpsf+^FNgy|(-5g!!VH3jL^vmxGK zuOV*WR8yckq3ntt%3$mcos7!}*$J~TDl*pJ`~^ok=y|1_p$~^0wl}Tyc%uwN?`~+M zJQtz>5KVu7|Ctbz^?)~YN~<>9H3k#VNF|i zhbMVQ7ML2l7rX!iwUO{q6>=t07imSZw9dFEc{yfIlOOtqU@Zy4^mEV!6S~J8uyf3L z%FtDQZu}kMM`;WEl%U=Z(@3ku`++fIMKSc2UmC1SRLZbAZeU<=6if#gWPe_sZ`W=D z98fO;WR-ESfYMC7YD5>zW$Ad@n2Ov@rVZzvZ()tOj}d>J=)aXLOn9{^hImP=X!4II z@{#671|GHKV#@O&zJNL!8k|#AW%!K~q#yR<~!AA8~Hv9MN2mZNNYXf^JFptT? zi<}P%IYE~Jpf81`JP{!sf=uKRT#$+hZ~Bl)IS76P_2}lC&a=-iONCV^zZ6A~zl@7Zno%XCHkJ%)g2WC7>o%RAjWav@qMmLrkdH=pH+d zGW+2idacmNW{M%`#3=NWH$mfg#-ioSM5TE7?tr6#BHk#j25nkCP+GOV`F0v$O^cgD~xe| z#%ADzfDva8nF$4Qx&;Xmo4B>;*EUx<(8&R7?K0`urt^~egK3J>`~`OgaL{Ncc2GkZ)KVk z4zGy;G*EBp$J;x9fB{+HHtVmnWLYRp+JVL!oL($MD4lDYS2&heveI6^8W~DFMQqop zHTsu#o<)zq$XoX2$>uG_ywEHTwJ9U#$3%3s-9!Mb|gl>4HRCmoziR`#ll~2 zk`>XD4@TXyGPvJiBE_II1;&^P-(&`rj&X6ugt(GyZ$TIL4qnY_EcCMq|DI8f4 z(jAv$K+sG=)vkRjpC4~U*xExS4+}p;7S?LZA54=oZc=~CmZcu664v=cIAY_xVFM>2 zs%t`zi|~H3wpKtD#f?18?UnSItbSmQ5U4?Dw%sAamcWmt77D1$riP<;S9(V z4=qh4DC~_qzY$xQFXY%7Wh@%GH19!TSt@Q*#)A`R1xnT9;x7Qf9n|ngE^711@D;gr^MywVxDxOnWPx>!>hBC2M z{U*v^$Ij!0T)lqVp2!ElC!$)V1Mpxr4>k`4qV`rMWZ+~%7?hP<**z2fwToch4cYJxzudX% zweDr-rvJLKkl>iAa0u%?Ul;l7pWmT?lt868w7;NeGcYd$tKQV2aMIj(fcImk zvbsb~l>AdNSC@ed$o{Qh!k)-iG`tXmj{OiG ziX0D4^Vn+)TFJ2oGcq8bM(PS20)&^~j^&lizrNY{?*Ox-@0p-6vz4OaXWu)L#ILZS zbx~LT3R?ckkcB`%NVfF0!s4gcgT}xx!1)%iZ!luz4h~()eLSOr#pEZ4NN;5!*NcH+ z4FmUJtIIPkR|ZzU`4e@1u}`&@8@!M&C}Qo2Tvaaw3ZbY`;Zz3oF6NM+A}<|%MMXtJ zw}H5HD5=Iet}M^&inO3fm6Vs)nM1ZEKY7euq;usfPzB7vDt!9uSKt>3su7^K9O}f> zS4S6G%N?u*li0xZVRzZy$b%3O<8+os;)WMm50LgF;kPZpX}a;2vDlQjCSF*$?A7Z+o!f zl+StkbK1YF{BAB={!$0_2-or)79M;S2E8IV34t%bASJ@tqxy6pW&TgDqVFk3C(lmE z1pf@bc@Ca;kVo}uF2nmCYnl%OcCwqKmM_QDmPC<>|nx|&$z`bne|}F?3D=-8m&MNUgbzkN{T2d$fH@Rl}TidH36-% zJtRj+Im%wzcL!2_u@|Cj;P@O#MGsOXEWtEt(HD$vPB`BDH-z){U_c(&U7 zvHj_(;pT$__dp~XOTA1)xi5AGl?742?6V`LHkt|NROMHEkxx^;o%|Kv`rTLpyis~{ zuG#S>d+J>G@E3ELp>;L)|MqEqpn-@J`8=w@6L-S{^jD3L+Di)l^Us^>yHUj)=y#9+ zET7FnU4p#!0~wrK1?4LpGaX%EGaz5F90Ifuj|Olg+-rHA)F))r*P~J#%HRi5Ea1g~ z$Z0u{je;o3{{fqYAwF)T+;uMLMj|JD@#rXIpcAl$Vc}$Irpip_=Fbf%}=&VU1bi|^w!TI^J@b@MTJDtpk*Jq_jST0Nryc=Y~2 zj24ix>v-(ZfKp&ULA^L3BiIhwIpfU~uk4>PaiSKCV#W@ZnuFD~c;=!pQfz{AV zCINFDagYhNvn!c9lab<#<1cj08ldZ2>w%2AFjLXKLd`4Cqum`A#2m7V8oF2&s2)_t z(gbb^4u|q+fDGi)+lPV`ee92o;63tow_C{;E9LahWTC_L3u#S;_YRM8QOEiDsob)X zhv!%c1{V!{wZ{gpdj1b36p{wua&GLlg5d)_h))BMsR%mkr+N!DQ6HWA;FS`nt^`t^ z!{A^riu*_>oL@o}18w2_WAKlt^@~F*=aW+%P!(RyYa+<^T{sLF*$E>PIm1?yabg1= zr;b?z7;he&VGs%zO_?_U*Wd4+i0zxo-?sQgm zwvJgx>okrM>0GSpEP;CmzaHLD0BCC^{FM2o1JrrDLBT2IP4djl433mA9cjzu44ZY~ zTz7MB*j#}JIpU{!p@Cf(#{B9%&{ManK05hs`_C7Tow`B?2Y3%aLP;E&Sw_uUYb^Kmc(XV|If!=Mfplbf6S zQ)aexGgC%-5#|Cp!r5yBDu`!R2GOBjdKdsIwlHee`#++eSIlZgkc=jQ`3V_xAi#WB6k_;Kzy7L&0knWcx%Q7 zlhSmoM|kwemf9#sy#y(%1elA`UVkVyTzU8CeEK!^9e5RC^_)_g6JX(MtnFo84IoWl%1wP2eDna8YmMiE98CbKRbYyb zBc)@slxgejU?>T6LGQ+?Y$%9m3W=anU*Q8(>W2w0dgHk(jz3%VD?c4P&pX~4D$5Qi zs%x(UY~U*n211J6yHYRx#RVX^TH!Fl&bcyaoS7Ch&ve@`cxpU7z*7XFAdIm>dy)2u z;~gHs1Y4LjFAER1HEzCjXgB}-8Pr+|dl$ce<0ZrBa!9+z0bI-<4VM(gA;c~gOOVKP z0V;5px$huGf>USe;jc(aOccZ(t}8RvH}th(h&!&u6k@Onvn7PHQH6geB=+Ms_!pxp z8>zs(m%%*qFlb75BOeNo&}fy(oJN%C$EUn6ll1`r8H(!Ivsup6sd!NSN@nIx`%j+K z@xfjV_Q?79>8HxZg~@3ny4tID-9gxUtbq!y%^p~RrQqle4vj$wyNUNEG-3>4=0QC+sqjb?-s zo?u{Gvehxw9YE!I2m6gz?%fd_{D#5~n0w(IWai%r2qYJV#eh;%s|Wh~eZT>U!BE7` zQkYsiiBsRe_5Zgj#|Fq%n6{h-@I7lqygSy*jNm0yC8z=smo$GIlJag3SM7N@r)$ua zUoLsds3VsRK|?Gm`}tRBaDf~~+=~E)`9L^(r3J*nOHhB7Lw@e!$&=^`nNy%g^6hZu zTMd+)aoP=Nq?ZEFi`*{MMRo;7HqI%;%5eYi@VlwvPu_#j@r8oWSQSoAQooFf{*6XN zW-g4Nt3_zAux?qEbmva#P7lYECpT~(*aJ{M4^GU++7%=%JF#*K@bl8FTqBGQ1-FOF z4<(Rw9;7ltak?VB&C0wrby=8QfNkE4^LOIqpjqoE+{tl((N3K_2@&rwAg1Y^UbIud ze`V0P$(R(Qnj$P;-@pLu=H4fuZ0QPCBU=60lc0aO!>Bi$TaI+UAvP%p$i-c@cRUUd z=dT9|y`GTdq=LlTCWH5eyc#!i1&ls$=47!C*WYn6?cXw&hgrfIo2i0#3yofzfqwrs z@2xjb3-8`s&s7HLMrP_nNIgJoV3D~1A*;w+r)|lg&j#QSSuqemfI_HVeJ&F??Wh_pP1FUFf)Kl#2la?Y>VS^U& za}j58P9uBT21~SijahdMWZiG798MZ8Ce2(RLx#<^-)A|KK08+~Eb}Ps>pfutS}Rf# zah4X29n;fIw7{+*9NeSV7!N3R0rUZeKmlPdvcu+eJpEJ!_|bVei64}{ayADFk{Xg>n0N#65 zEV<%%E_^wesT!L2=rADy<9Qsw@her~oET~}Xj>{HxCnEj{2)p+Gg_TT#`(;s9E(-6 z4E&?uP%&#piYDus!R=#Y5J%7JG*16nu23xJPTAn|^x^NOd=ir4vD>vKKKGp02<=r; zl8KKzV!VHGe{`etI&C<~SO8lDt$ot+(ykYn+JJ(B( znlpHQ9!S} z@k4zZH_Hnq=xJFA5`Q^)QcF2^J>lZ(O!_M8Uyzym2+Dfa;Qw+prPjfCWPrv*rt+K#3$#v}U9TKSn9r}VHvQcWLQWF5FfbS?Hf2oPSa(PeDmCR`Ag zrDkd3ba z*oB%lsZG$Cin$|>?e+3!jqBm+?|YVLSaSLHA;n_lk-*f#nr-9t`35oMaEAoejkH#$ zlG|2S2Gbp92~)phD_mE7FV$+j1%E4!vdGmOJO_k-2<%dQ0B`6!_n$3}IF^#)HIs>B zf*ca+BG1R1E=Ai2kM#1$*8X4|zS(H#sEUyf{&|?Dv0m)uE(|Bz1X+JOWZ4`$GENDK z9ep7~EYq*@O5E1ho+CnJ<%_*qqn*+tb|Ka9tr3C1V=O}?5acdpR!QB(Kye?`>d?eG zIXQhk_m)ZYAT53tNQasXuGSftDP=gu8(k{KX2f$Y&^3u^^I_0@i`nnnCi6Z?TPS!A z8Ta@ux@esYFANZf8ZNDN#hy4EPQt1GJ^R{^MP74J5^Xu6kz<^h{?vzeA;RrDf(-A{ zZw?-xP6WRk5vS&A<~dCcTi$Dmm-BBX2g^aLl&*v{YI0DPD=moBAUU#XS3pE=xEiX$&uP+=2~y zQPY2?anjK^vGkmQ%V@FbCn5{~*sjrdt)ha;sgsHO9n9fTG!_a7u6m||K^N?_91d&R z@7BDY1hnc>6b3|<_vrMR|9DH?gVNh!Fu}}`rVt8>5v$h_VN^*ejETf?P`R4d>ze>3 zyHp#iG>$64FtlUj@kUFSK<0xpddmX%`JuR18Oi=o2-X+zFu>9l|NKTw4Ur-2)dYM?4n*(xCR30%LuP zG*=?v(Dz;ip-cpQQ78$rHVOE#R$M9+xb16Bag}Z}uzn!Bw*MD+ClCaQ0_wSypVn0F z&^PrmdcP@f4DBICD}`{^VI>g9y|JTUs{&vB&Ql}z$#bnqf<=;^da)1LU-{HCOHtzS zlW`C8351CktG}5#fP0Q^3*+G?n5yL1H|2wK7;)?-;tEw#ts=xtg@)i`xE~+p1cWg7YE^4!cipGpHf}5D^2e5KW3G~HCg50rWm=Dz&;C#c7?R4DeeAk1 zXvE^4biw>$vO zZ5$PjQCt>h@R-8Az_M{~wZ1q!Zyec#Lv*(DFTB6)(DzE83R)EsUW;=`Ut{V@DFy8- zg}Dr2vQ~vS6AfjjOA!~w$R+YyryykA3_gqLLM?b43%>8%HK5INL_O7LeP%~eE_3#j-xB&6t|TY}HMi!eBhW_Ii&$l;xUR1zr}<4`je zWhRa=*R5;5k3@lA1NP+-Mo&0R3y0c`>|{U1i@VdcJ2lW8$kQU?Y92U1M@prYV413w zUnPfyfAoI#H|Uf!qd)SykAom_72z}UKs~vSI9=TU%zw1cs0)oC7$j!f5IJi|1tRHo z3Dc9zdkXxAlC>Z;l`O}Tq~^WwVI1Dg5#k!*0qDKLeh7#+w_rQ4pk{nYmc@rIoQ?qO z=qolLN=h|_aK{fh(HOp8wY9Bod0l1>aDDC;*H9O@1BY>9CXTqqVcJj@y#xU?&apP9 zLeP;8xosKJM1c?Fjf=HfQKx+|zR;h&aojIJvg{^DcF(ziS_a)fadATZ0rW40+Ab%C0R?kicX9GF{`p0i)#Z;Y=nmIzKRyF@ zXfI=O53pzn#N501m%vQRj!Q$VFBvXQVP)~Aq>|n1WxG1eWgQZn27Xt_D((1|$29G? zjk%hvH!Dw-9O58wGEVFBIl^_3x4GXEnhxWH&~)%+Wyn0)yqT;o3dA$h>c=t?0D3$E zQYKb75X{H9RLGEld7&v+?d$?9$vb#?d11)$R_Lq*e%*Hb2ebe+Stg~J0yrT7snzyF zSJg3W#Bvwb2YO|16=3fH*e--bs&fiq^yD}c8%hyI!Wzu{WM4T-)dBkJhbcXTLfe2R z6zM$*yyCu2P!8zD-cDfAmPZ}Gp)>EE123x#ukZd)Imv%DR55X2;5~K$F>}D96(SYp z#}G%_s{2->Qef*7F5r7ZQChK&lEBHJ0)4aqMqC*5^nJjX6QFb}!25shZf7N?!A?fx zMepxYTq{|g;5DuD&Qp%@YFN@D?bKs8HLEtSvz<-R&FZXF`HLUk8&?{Af8O>DgpAi7 zI<6(Wu!oQ!9q1(`n`DRY%dU3wgY53;Ee?J@XhWeyelsS8^b<&s`q%<_`1P)e3(}nS z%}=W&uGH3EO}g>4+3LXP^w!QtwL7+Kvib+BP+(PmQKp^d$UX8xyCbiCRq_mbn$sSwk6$79fBZe)Z!V`A`|l=-^;KU;;+JpQ79i5k z?smpf$&DH#|NWo${HOK!_`nt~ej9F@AbPh=&}i_?=7jbSJH$n+6h5*J@ch{R3KPE(jEuFefOfC$`-xbXa z?yog{v*MY--+ghftc{+pw?dR4BnkA!HvgBVr`<(ykm!_BY>PK;Gnk zCPd}b$iE10qKgDpfL6F`$M&giJZ0#0#u12iM5V@U#lcFOS&@LmF>rmSLGgT8046gHC^0olgrj<-Js%*q2RKihEVX!pZ)qD z$GGGrVs{E~Bz=Y}Rk>8hH(Qm`%l0ZRnAhvSjr_hxW8@pc8`#gU|NNJKA1qxO6<0J+ zoZa-++IoN9|Hs#LfJJ#N?KLKP6Qc&j1_B0RK?Fo80#XtYbg|Mq1VlPW@5PuXQe&hT ziXgZmy-BY|ih@+7HwC3F#g*Qce`dj47x(@=&yDPL-96tqbIzH0=bd+^T)=bFT~;%A zsn*IAcqHll1@%~FCWCS9ojHNr<4RvmTuoz2BVs_POL` zzCFd8rN`^Hrw%yosUHpfd@A*X+Re~+sW?xlmv3WijBMWztNMUPMDw|wuKdbN?R${D zcITQk(axX$*NOtBT|GrFu<5*rOjl8@F%OKdQ?@mtpGdoS4f(w7F$x&@{;$m;5g{BS zAZyw2Cm40HBuQvz;pOTzMx7=*YJZJ6L+dbL5z;$Jj3$pjNE0wQw1aZ@MtHBg`><>7 z6i+gfQD+AWt~0%rq-oGhamDBPAEr3<1pFv1mureaKArouU7ddB^_V}SxpvPsC9E7J zt1%V*v#-1=T+g*Jw9hY4u>X@WO6h*FXN8*rI=^9BF~(;*|HJ5W#_PE`Qp%{~R?j0F zNGo>d+mrCcE<5cCJI~uD{yCVDdXUXVMjusLk)Xs(;fM`0sJ}WTF*sjN;;t=pRJVKm z^wKrD_j~O~0C0sVn6^p&*SP+1U0}8zdRo(P=^fut`}zs7-m1Sx*X~*|Wtv%T#)%44 z7^C}^PcK?q-LHh-oSm;^v4sd^|JI!HQxoD4XVd-25V_k;^LRL`Xx{lP-XG}xcJ)R6 zQ}k?pUatRBJXSu7*}9@AnewqQ2mp-Fe+J)+4Zq7Z$XoGgh~U#O_|a3mbl4|%HoCtx zTi1PSk~{YnYJ#7cp?A+nm0-@ns|EQlKN z);-zG9<_YQV9kvly(_-?#a#8u<6=k@o&j(o7Z6@qz~iHldw%<7Q}?t0H;FHm5?Zoo z7%T-{W6Q9wIx^r~QN?B?v^#eGn_hgjA>J^3vsG6)sLr(vylnejoiTtBc%4qZ1%# zhFLe-}xTcDg_I-L5J~7@Ld+V7`&8IqV_*KGbEQ`74 ze(@27guu7-d$06SGWbquT6s!6M$+q|%-p0yJ~0za zOp#WeB7xp&@AFSzwksaTuvGDa=>7n);m1f+sO6{U8>~Ei+Fu_2r?ni6Z3Tt!pkuR5 zpj01BSaZ;M=|s~;@(HO@5YYhw>pe+#7v!K=H>(@HuT?*oT0CzW0cm(^XdGCH^sZ0i zx}z}cW_K$%HK#X>r&*JdjkH%b)JZuI`Q>21YxA@CJ2U&bYUv zAct2KdA_pgUoOCZ|DE3U7P#DDx;_QF(K9Xp;$s{6K*mG9#&RK& z1guYgWW=OHHPXPj&wZ)Y3G>oPKIAhyvp%f1p`4A<`PcmJ#v-wcC1JewnywcRc^e6BbA@g;s=*PD$OMZ#vPMtAs1c^9j7ouYr8+ZmeV*yB5v}1{ceR5GnsT(nj9ImC@*e+{SzP_EG&`m+ zSH1sA;pgHXfn-t+Z2))SuxRG)?G2+Mc8LVV03WX-C94=$73=-P$1N0}e4MdtPEw~; z^Vrf`?#eezUO0WZEO3iyfQ4}4xTtqsQ3(2GU1FZw=*BS^nUJ+Ou%jA0vyN{G67fOe zUjB73E4{yOzu>K^>n_9@vi0Q|?y2Yd(6TK{`tuRk(}EI^%gohLhb*g32`dmId!Avc z1u0^kXX!qEjVqidTL)yn0<{Zxkf(|OBd5zWb@=RMX0R*d{$n9?rzpcgV5D)PkINr{282Cd z{=v5PCihqS3~0c!vM(K^@R54252m|nATD+_Fc<7a)M2E^px~5aw}I84t9Fg;l_G!~ zE#Um|-J#GW?=OZD?U*-rd)kSV1d-RFJ;QZbXj2K-=bx^o3@7Qnn9d@(GX%X@g<7`K ziU%r*sV`#tT~EQg0`8l&BFr`*Nzr)&t){17b723|gIU6vYB(VH{uDBSy-h(g3v zL33{ZhyFuLGv`hbo&mx6R7KtV)Eri(srgqxATfb^f_B0HZn*pn)bmm#q3U- z>Or6lA|V24g;=JB=@NFA!3R#Dr6Hc0-Ydp!KYn*ONC9qdZjuxDVuKXU-8=dt5G$LwTIeJ0l}BhE^Q4{bDQv!=yP1gOQ2$A%N? zV{yzvEn3@~Oocc(6^N`?YGMwFaqu*s=SB$#7q?M#KqOq2g2zoDUq%7u?n&&EY;0`o zRM~O_Bq{#jcl*%;gp?~FBpQ5l{^F-JKVEPJ0W0AK2wca+|HRV$ zX5;bJqsMvI>~O`bkmV!#D{)u)=8wWC90vxgL>*ApV};1B*mNE$_ZJXG@ME; zu=qO)brL3S;0*Nvn?Eq-kjs_G;BgBay07jc=QL0VEDNY!>(=Xou7&_~f6Ym5V^eX{ zMs^&aC}Y{6(6Hv`!*W}}jBx>Z-tt4#xa5M?FVg#Vef98{ee1S;t$kDc!MfjN3g0}} zb9Xb2^Cj^nw#Pg^5&bG5aPWAm#^c{c_r=;>ZF0yM$xI5|Q7N;}WE-h6dc?S7Ph>=kS#r{dbnU`Kcgbcq&x2-5paN5BCSsc^yErv; zD$8u7K9HSU*3+3C??7LF9epcHJd0IkQbb`C1T4>II;?Z4Luu=Z4 ztUK4arw{(XbBTVXdzx6{L1d-ma=|*n>%QGrSVafWSpttzmE?VhTTaey?16&qP)_UW>dMHZ&t$!y zm(=FUq4+>4iYC&5bWwBq+C`>99sEpABghwc&A0Zx&Tc&bSUQ-4Cqak__aFO#Zk?h zrY)@xT<;H;4On%n70}13^#<*c?VCC07glUD7HqQ;%o6G8cNE&fV|krL2N3sJ?31Dz z&>AXV%~ZDo2}md%QngVe-mwh`QlOPSREQL(^Vnt&^l>aZA&Df`y~4EGRRgNa?qxC< zo!7Ao;<a-p=*A~~(RilR;8u<7D9Ek_ z8vJ<~`I-%dK??68xHvKEPLoMMRj!?p0voU^A)RsH#sRlv*w#X*0UX`9R-rC+hjp)XN)iB>dfFs9aswewh>X53=w+)~ip)U>E- zfW$8tq0!4Iv$Fm=;Q=%`x$9TymSg3G^x1>8<`r^*tdF$~Zc*!8HLRCTjjlDUsuine z%>u1siGZIuIeWn|;0_B#YsOH}8gBDlHGIE<{$hirm5$pTTfUHR1r`uOw9>&!mGX?r z-au#s3a22G!2E<+)|7oa)F7P%g{IRW2=e96d$(&Io)Po&#qB~4gN^T>r8{VEHRq;Z z>jaaj(%73z{_P-IegEp{SCcmp(K~Gemnn{`@a<+RZ-6+W&sdt^QL8!ukQ$ttnTf}B ze34xg4*mcXLy2>elzy71l%iU{{QG#^nVnw@X3}Trpoo6pRlcyf{HMnXJP8CHe_0|W z$LRa=wjP%s+1-mr+|Qq76+W$%Yks2GzV_kn5!;flOIz*w7Us@#F_q+~wk?PMAw?Id zwW_#21iQQ1hLKumW0JPlysWHjFO;hI1yxD@X8|yr3}YSc$O{KS3`co!iJA?CqxM3} zSsH~-NR?wN|2PKQ)~U|5Tk*)3w}8vj-cw1?pUS~q@c`|REgTo;mn~2%sNpFXcyU3? zqx#Y8OYR$4iG(!YiKUaDJ#uMEeAA089+}KLrm6%lQ?S8HJxKC$0_jo6I}AjYh{|)H zt`$`vTF(!~RYK=F#RUD!XkFOd|%^itJLI`hg**donW! z6h7>Pq)5!zQ)k#O5>Jo{fcN(@g^Q0R;qlT3ih75Y>D`Q#haUWGrX*czCr6`dLB$qr zn^)Hsd-K>^)#&|9zeIHDFS;Qv;rA*4VvRk&K(9f3N4Lq%D!f0Ebz4Fc1IiTwA z1RgS4W})I9xZqNu%cCWh)0fW&XTv$3Qg;Ga;{}oiMxZ~PXxyC&vRyHZ{1Iu##R0Bx ziqzibBsd89@=Q*6%xMLhAEiHY@*g#E1#X`fS)Kmvolp+%i}MfQ9WTs3G}J-Q{*J=o zumXD>a7`e;nP=>A76N)BECS|?d2a6n>OK__^iQ_30Cu|2WXW?_?`3z~h)zZ0{SId9 zw;69+1^+TrrGI<=xF-j~jRdh>Tg5XGDSA;L6{9`fvWjm1pK^N+A%M1Wk8Er(nU*IcNeaFQNbyFG#LT zR~ZfiEeS{W3P{%GOBy6W{#N6sg9Ce^qL=3{Mrky^DBFO+z-7TybvOXrdn?{Ly~|^7 zyGlPX%uQo=g;djCEBcI?H$sJ*tbwW6GykaQL5PqI`8i6n$!R#v_n)mp7wUWNPg zJ3a#zYh^_Gdv_^-jty!VJ$^-x(S<#J&Wf(CLPNIUoRWp+CxrdqwbNs7UjMXnWwDw! zkBnCGu8U6Gdt-Yi^dr&!=^%+2#9EI0)wCJw%k0D=y}h63jcoI!g%OykC@wCJY3oV@ ze1B&(3p+bs2;*sZaMdSq*Vgj;OIhLOJ%Le!3`4Utm4huPf&9sbyY^cL+~es<@rrRf zs}GL?=X7k|4_N8(`e?S{zE4*3 z*`H}Vt^*O&6XVaX;t_K@{AQmj+4V{$Jyd-3>L|Ojd0Lw)hH$SQYkTE`%%$!Tpp{M^ z4pDBz3L>m7XG4!#KMxL^7L0ISMvCXCIfzWn%#G+tLaIfe0#gaBzhHAd4Hy8Fb3~n^d>IFu%Z&1GLGe@XMaS$qb z2nsqos#$jLKE%*&Ur(hxo9~n2XmefI@7YlatcgKAHv`Ai~+G+1&Wp}KYdzq+Mh>@SDtu$wDK@*-!i|l%rR4)Wy7h}aG<$T zP@HCGAuEDP*X0@D-}^HDy3In0T1Cc(rj45|Gu(5e+HWUPd%ifA#pApFdpF$97y9uN z7obw$#M8}Nw(P8izRl$sMAW5dF+$AB<4|)vV6i zD$Bfixn#AaGheLi!6&j1z1-o9NxcI$w&FXsywS^hzYUaDkAx`y1IHB`DB0eI)y>QS z_IpZy=rs1w73CVJ5%~z!6qSra-|mUAXA6B&HD=8z>Dr>4k`B)*oId_>VU(VlT59%s z?0V(;l|x!P;Gb_Tp}VbGGsI3lhQVpjpaHM~89%DlFGLf;@%mF+{ejWXT+O~UP>SmG zFsLHpu7VG6igKjs{Pfk8jNawmAN2HuKG7^?nkou3Lbzb~e( zBk=N9(k-Ij9n7U2M|kzX(}QkLK^Wu_&?87ZPGsG)=QiANJFBs#Ih|2b6FS=AdONMI zcOv_q+smW1Drxd?GiNNy1}?IpKa{fQDO?WC=U%Vz$O?(c7gDr;4D(ki z_==!J`*xsas2m|fDGyJGkP2LR7XfuGxed5db(h_v$;KLk%r-yB@E5C1J2`!Ly3x1) zJjB+lL%-6eiyE}OnI$CBr_>Z!@VA{pY8qo87i>?R#k~G$5Hr#Jh_6il$f(@Gb&;usnasj~}MKw-OK&O}D7-w+<(2{8w zvO~6l2lZkx31LuMfSRBY0z09BGkC=6?~2UpKr*cxcA+rz zqy;t6Lk`*yOoKQ%^9x3osspdnjo_3%tXEON1i)u2X^Yi)Q?LK1?o#hY4;@JtH2 zZjMkU=Ql*Q^%#0&dF9GaDEN}D(rzNsdF|Bd-b8hR4I5oO-&VlbbI0<@0L=)_fF=t+ zW0+5)`*@C2twapC?hY0#w`YAx&nRw*R0TV_QNP)rd|`A0xU95jb$HZ&ou4lxb9dG( z011WzM6Iqiyi_%V25WZ|9IZaop6K&~iwSVp47CCW3!FcjY7yYX=pOkpJ>XKV9T-^V zJ8HAw!pV6+NU_jFR=43CB%X10j*hwL|HQs{zRmUVUmn?fmAXrn6|)^V-#fTl?I+kH zI;1NZ+BcQT<7zp3{PU&v3Z*TkAT8X`u4(_~@i-;&8T~d{W^T#6>H>l zIcOrX5Gz4!bGdw)$ld)hB!{_rEc?!#JF``)#R@-}k;!QwSyB!?8PtB-MK?A0!sUhb z#elNo{2?6%0L~dI_=waD@`Qa$sOOm;ylCAN^T27w(^Wt?;I-vM6?K&A5OFe-ALhNp zVFw{Unzg4+YBv1B%eI`6<%WNv=y)iF7_dP$WTcOakXCKYOl!Ob!o$F^A)3S;v+)@E zxyPSXqyo~eU7TGyzOYeJ%-Tc;&T!M9Im{cj^ao8}T%Np00g9uE0}=iXX6?IfN44B;0!cru*oX7o{oWqKPo; z=?I83I;Ql+d;c+fM$u2C{zh+>_XbxEj{jbr-JleE5P+qD&=66q7DSyP0i$9K5Hj(U z67wXp8w4w~_fM%CMJgXD2p5aIfE2uRfIz$txt{0b#8O;_&_-zrFNrxy%Kqq3g99h$U7+gG8i-o5aj%LB zilLoz)I#S#en?4HPSL1E-!Y>k8XXw2!^a=hQQn)+8KKJGp|CG^G$i~a1H!s@MTt2_ zZ&|eES&~Y#W zGG-e&0WJ(J%Io~d&dJH*FIxzra!0~?Kha8<;m%qSjq>)rponv1X!{%}YzV&Z^=MlH`p z(b5O$BfOXhk`FnuC-T*1&F)H+IRhCaa+RL~UXQeDR>+SsF_ovBCGvM2MYVXuo_D-p z1RywEBzq^Bw zeMRw6efbNuqNs6&s%+mq>BQDRSZAJcrcM?0RM|sC#mQ#63mP=jhr=XA>n?B9`@fCu zi(0mPQ5Ov}*J?j1aj%!hXv@?h=LgTcRJI=XXyrBT%()XdXAKi|KYt~iBl=xi1vGHe z6y!&L0bED`%~(-rwhvg*T|Y*81W|yy2QLNyPdahdZ(DP+ z<}S<%p?(f{L9(unePJ>(5m1}2&3NXZi9-p$r1d}2h)cxpVC^%in=Wl$b6QBZOMEca z{p9367sg&Il?XF20eyw@Z#*^umJi|AWk+4 zsY;WLQ%w+&)#Y$f^@pyRn>$+7?WcQpTCz$U)z*FW<~0SJ*y?W(o5SzBW821>h5HxW z6ZBMR;z?lqwCOf$NpUYv)db8O2|OtF-~{w&R&1{;mswHGvG|n3uOmsd4=Maob9bd~ zqdE}iqFgk$l^}XrYA(b|&Bvw~GKy*n;vjWH%K5(B??qA34{kNDB6H2BD%Rw?^!}k` z8&-FB806txQTZ5i`IJr^w5+Zah-FzrSL8BijLd{m3|MmjNOY93a5$|I2KX{ec!2;Q z*N}P}n4JE_V>fx{fNIpo+Or!m(w3`5JEcI;wU@PLR`v*p@c}B~G2jxCyrqQ3h+kR4zw4r^1RMRkoSkh0Dq9|?G4bd80why%oS-ajHd3{3fXgNZjFm5MwGA> zq`Xn!(-*}JIdxo?`>Z9h-E=(X<2+gXZ;Bf+7_r{KUQaz;w|A}Ab3aa>&q_3Ji3_pv z6FEorJvE(vi_U8x4uPc3!yu1=R4WTP z&ypv}18i`@Vwr`#d%sn|GIfC_9|brtO$!S=Z_ex@OU>nc*6iXf8#&yrVf{9_ol(rK zr$1Z2NX?yU;j)JQ2d1jdr&%?acJZ+WEIUA6&-@XaC;dg1rq~{-Rq%@->;3&->*<>{ zgN27d&hrTLoI>4IZu^0o+ykMO8x)DDCydBFBJ4JpJC<6=lp?pg_MpN;Ajq7Tqx>K8 zz&Zg;MC-A9W$2q+;8!WsX2XVJZG&Vod7e>>syEZza>n}oo*R?^tD@Z(OVm8i&hg}p z$6s)x_WMH=_h1c(te4X?1{48-dfEYO3gaE0Z>^Nv32}%w0Bztt+KhF_cw1gA(ZEv; z<)`uN(2kk|I(->%XT-!7a#Ei)M^=zoEas9Fi6~@?nhlkPd<@n-ZK1=;hAZWS!O_~~ znYWKi{67Y}Pjov(IO*|p3&f!d`T0*PZE;-!+t?`|rn4bh+ z8j17cp34h~_VlK^IuxrKSs#9W79aNwh&Q-K4VUe*>{eQHLyoAR6DDj%OFKe%IZ_YZ zaM+=~Wo|H4k+`G74tPxpM9S@pZk$`h2BWPb-65X!BNv8dP8rSTw-kX5m-DNbGdhsZ zeZ6t4&AY_8`Vub4#hL70*R9!WFlw(Z*+j*Hr&m|94ywQibJ%RigJ8H}ws^8~k z{ARfY_E?&FN#%xh_ujzmkw(rPjPa!vT7_HYL(K?Wez#mJH34fLKnq71wZjsqQ*_vt zODBJW=zy^x3rz?fh=01tU*^nTXV+cw@y10e{r%(Z?HhWh<>Y$D4!T@%ovUuG_sP9B zbZwC=dRISf$b%`m1tU$9)2}`sVyi~tT?o#-{{>8*Cx$5aU=1HKy#OdB7L0xTgSX4J z77L^*r52w_l|iYHI;k4;ZkcPY+J@`&nN%Cip`dIxbvh?-!R_;pU;SNApLI|kBgj^> zXotI_1vBv%7F#rhn^rNz(&ft9u8nQ#P2aLFM|R{Coca&{I_HkaAO4g2vrfgfbw8g| zD&6+`IwdtDrYAdtnZMfOcHRCruMvV6$@}g3Jx150xt(wOZMweh7isC+U#ordjqasa z*U$X)wcJ-jW3CJLm$;fNetUT8d6B#2!KjN(ym!YYMSBLEzpJ~e^hr&t^s%Z{5TDuu zjkBcfgy_e_A%bjZN=i!Z;>T)iP(>q=JQU2=dAz~ZwpIWtL1sfe=R#WnS@D@dHa5sZ z$+>k0M|FUQ^#WzG3 zBvxm#H2je_33#U1WLA*E(-O-mlLSNAM4ah#Uzk>*=duOF+B+`#TSnJ`*JVj^hN zDm-Te6Gfi|nxrSRLgN}PbkNUm2VYp=xp7`}fTn`?DBRKqgoTQ(H9;$3!T9&KNp4$( zd)y{a)w&P32bOb%3=EZ!s_H1`^Y<-wh#COUe3)k`29NqJ;z6g8!Au@q-M?zfQD5*i+*0pbIb^E*5_H=LEz2DR#Lnu{) z8w4N>)xR@IKp8(&p2qG8dqD>1(_95PE>cQ9`KA626dy> ztKk+<1a1?h2td z$W8a&qo%JruITC*xmp=~?G=*A+-MhWo4QArB!K>q&PNLDOghYpRQh2NMNg3zGVtdXv}l$e-UFP6_E zmkhV6Ap9R;!SuQ4l?gkl|Cu3%N_KKgEFZlDTDYJ8$o}CHYxnfKCgRXnghPoMFvuRz z2V?>?!53omaxCb?sv_tNvJdMbWn$BZHGd_VWc9s^SROWujEtmRGq{;Rh3Pg(!wD66 zK1{M$NE}zL70oXw=tBpL)}{*fQp(@XpMUqK-%sx?RusD6WNV}~GodjQ)Wq_f>DBm{ ze_N%$|M>z{&*B50-bnlNhs(6HKT#o{K&VOx*`oI>16;{+6x+i2!qQ!KV0Hal7lNu* zG`J*`&mU=Co}QjiqnDV-gHn8`A!?r0q=FqaTY&(r4*-xa6cvi+=jRi3RTSYUGV%93 zKvLfN1e1_uK%U%Fx!3kwgk9s=^d=R}tcPFT@lfO$85Dqh1a(HMDDBX!UqEKDZy$gJldHqnNcZ~Pd#5T zRgO2Pi_v32?J5@gLN9nTGqD9LfsOdU%o0>ksS$f7QlQb295J!@VSi!o`@DhzaT+`b zFfYbqpn2IEBSW9kT^T{CcgL@pL@mKA@e2issdiOqwVrAG-TB&bBPWX$)_%W2(dGPo zF|UNB3EnUXZ}?G{5*tz6^nlbW6boZ#;@udpV-IDptW7&VN8Ud@e~@iiZ6u{W=K5@H z9hs!4l9-x_vsC^ zKczb!3lTxWi)P{{tAC{X%heZ04@cTway-|aQ52}=ZrmWPHKmrk#@Bk{4`g|c_+VEOK`PWaZqbeQmG5{v7IK@~F1GaD zj2m9NZY!Sxvvl_VnWZ9ul?%)u5#6Tpjq|-C{Mk>>j?xY2(iz&BU`j$0Fk(F4&71BD zrbG!yNEYH9aXoee|jr6duEXv-ZPDyPFAHF7`8@2653##4S ze9Cu2FL`2N%{-NQ$K#f>`+u#nND08mk7Hx4Hs=mU9!sibk1t=^L|$W0dNQxUdFD)@ z9b5O`GZb)Av$80%-cOzD!krS#)%y}24yWoj`l))NR&+=H;Odko!UglikC#J(W?jR` z@#8$)`saj%94`?F%~h{&H_hrU*j40sXz)a4Xz)+=-}7*^(sF$#dpX(3^J_KMyLmvDCT~drf9YjIjQVRs3wjogVV*3;YEcR0PK;({XG_Rw2WmCL*n*;<`l81(FTeh&{)0}?bO2=$J^cSM*i(T!7o(A8{W&*U&=L&-AMvDQNR z5qbVvwlzWxJw3X^4s@Tt`r=74RO@`G9lMuy)|EqvdRI!YxA9gaz3frxd|We3&Z;wW z)=rFD38sW{q^G7<5f_Z=Fer;%^N)mT|NGjhXAjVU=Dc4dDr*q4WL4D-X$A<9`B@rpIa!`R$hy!OQ~ z-J>OOnEE$0(ohsjV`U2EitZ5s;W^Xlqkv95Zye^;(JdJvG|U z3;!d62_};9`N<1btQgovwSfq4=2Q$*P-piXdxoMDdnuTUXKVY35*4zz} zMN{V$?!h8!T)cP@&Ie~1aos=^vccMJDE-<62m^PrQ@q<$R?x9lu!j7oPldNo-rYxI z1YrM&w?z7HroHm|co_PJba%1c>iIxCGRh0V#n$BqOuDYQQj(dTaK_R78#N=kTl z1k{hw=FPQx;{Gg@ud<-L(z}RTlAseL$pibATdBweOH&|&#Bm$W>F=ku>c|&3Pf65@ zR^nvS61$MwI}%BEM67;o{m$&}+-=K?V*;e(&4WI2X4bl2SZGovodeKF(_qIY5;U1F ziJ8vrQ!WxVwQ!=WZj@V+Jes-J-g-N3HG=MgG_@Yc!b!ef9`1LT99}4PoBQ}lIqZ*W zy60bg(f+L9fJ&c%t7Ay5TI|KTA%CeOW;YbO>6}kJHgLNWwLZBejY@dtV`D4C3oxh! z@NC7jHbF=z)Bx%~=6CZ=cEIpM+M_C#)AEMQMwZ1)ztnUULN zygf=hnrGTc(Uqem|A|PiUx7<(_il*+{m<_tLTMF?sGPX@U9$N&?;nJtoJVd6;svxD zZ_hsLIdc+VRKSK-a*;qsn#>Y9i+ODq7Z;7lrF={WOTu|n3Bn0asuhFNIR;EkZxr>dv*yH)V*2fz8SDYrGAacLd5dRua-s(*9<$S$AY6FzS))9OX zgSO~R7681`2VF)z@uGrr4*3%ggCLHy)gfE&0;$3&m!faV%?2`~mdr8Vr=UUkI4he% ziu`_ht7H^N+lxYvS$^gex(Qfa!b4V#v&0C;YmR`#=6_7r=q!j{%!5elv!AqV)Acmq zXEh*bez8APYg=5Tg9aRGn^s2ydrx#Ai{0vdl!+h<1!U6&k2@y2W4As47LVczS`CJ^ zH@unH!Zt9|tKcMuFR1c3yEP=+~lFR*bwaTP1lKU8S4KG-^WDi{ji*Va1i?BEWcQ zDJax)!twBIF?yai$M3vp7oYk3z0?-00|!woP7Pa&(Vs9oc48pHwim|8w)@voJjhmUPU{pY^xks6Hf% zV0P-e=6_A$vtORvTAu4S-J7nMWmm!C?sCEEh(_~)gQau_KZ+GOdMaHiGjzSx{vh9MpD8C?;y~Tgpn&}R17=Vk@ zNJFeH5sGQbupi`uIT0Be#Rt1V4~~xV{tq)iFGbw$oSl6TKmpG@jmzE#HWQ6LnqWjl>OOt+@3^*AZk3M68;a2KK&Y z8(2(3fUH=uL~6_8I9vL-0~??NvPwKuNvUG-^51I?pJ#hjLdUm|q2bGePj+l2#Nem9ja>Q+gz=TeKOKlYu<2AzMV zzrUoH5yjrGbT$Lb#xt=>EG6gPr(+WP@5x~kCZl5}rB=0r#M@il1rsc=} zSuP3ypJzy|z&=!W6!O-84_5*toKVw+0@4JaP-Iifb<7C4rGwP|2-OAkxw`C+&Pi^g zMMXu{aCkm3XT(tOY5+4`79@!~v?g_mMy6KA7>Y`IT8(?(Po+PdQdNr@*cW&pK!p&j=P!mJx2!Kh2@b9IPB=7w z9E30?(9_0G|?^xj1HM43gQ4r3q+|zV=77a?{@1z9vuQ!l}eX zy5sz_*4%w;{cH^)NtNQz$!NETWoMeyOtmic7<#ciAP|9CV&Ke}TJ|5W65Z#1u>pxSqh1y2II1_Pal)Lt781y`sQgKII6;PyA z__smTs!Kim(^Jv@Ov_%~_p7(G&MVaDLYfrBL?>--?WA$4Cdm=Dw5gt;R(dYVRS)bg zOo_0K0J^UL)I&K{Yj%IUS&4NhpL%_Q1klJ~Tft{lbwxUlh?3rUt|Os;HE!e|4p2%1YV z0y#RYh$t^Dm9FmhgwD`?ATwY#H1!CHdQ40RHG`|g1&YIp6%s%@HvkPn6uNt`(0uO3 zC|00?H9ZPoI9LsTGSRxtwNl1WZs@PK?h6R06&>GCca9rxox)I=z= zDnej0TkL_GGz6)a69Tl}x1d8@uT_)waS212UTBPMT9r#TcdIYRH(qEul;w>N@4E%p{7HAG_R>YqWH_Sr}-ONcp8Z?Ts`}d+eygrro?hA9a13j?g0q}IQ5;& z-)=;HBKo~3q2sjB(bJvbtZD=$5(orjU+E66+zXAVkb`?Zs;>!Jl;E!d_QS8Ai;9fq zZ-{EcHY(7-{rsNo42G)b4tfWkVU28eg*n=6g<6^equ9Kys3jU}%vSgsp8rg=tfKMO z5E&KgYTg_fD7w}k^bFrm5)>SR$Zan&z})7lbprrkdk{a7o!zldOct&IwGU z2{d$FhWJ5u>8$6;X)V?WW;(2(bVDKr(yOTe&BsAoqKi-t?T*5PZqSCxm!g%oKw|11 zC<|gb_)fsW> zTi6>|Rsy^!P!-eHy7u&&hNh;bAWBO)_@jN0Ed$?0zJjB$AB%Hf+spUwu7y!yPbA~v zOn5Mu%AL?q*R1Lu!(ra-JD}P(s4Yj`Ta}gmfwUK06cZyAI#<55{j%h)2GBR>|v!27az155iB#!{# z7az2#O|(l+4j!+AlWY+hGnvXVDlsf|*-v8LRhTJ$z=B#)cYW`-%QGTV_ok{{)|#U9 zO6CD$pC=pTIZx3vW8+0{?f*XDu5AF3v<(^Gg`iAiQPo;w0~f#(bDH`EEujs%e}#q< zg&;e>I1nz`4{Xn#Nx*l%{Au3E>i8&kBuzlz5eBOR-95WXmK0edSk@RDX~ozu#nXOc z{~d2{R0O2;iGyn~d?F(RgH=@Hwg!DCVZ)bKAw0AOqRwezJZWOD?MnfO1g2U+b6+1i z{IR;QL!lLfLr#MaZiOi^0z%TLZSK_pC&t+jnV%v`=tD@>AaaTN#)3VSj0%9|yLA0<$c}@*5Ivn=;Hk3LuKwrqtghOa z94DWwsG1o9c0abe6%V|BaOT?sn}3%x+xK%`e@OF3 zC6})|3N$RZeVPtWKhOK7HvM@+Vs=auA5WsMa*@kP{m`n*?40XNV>;I#s;j!OoAc{G zfB5R}zaFqGMud|Ab4jkZp%Tpyd9J6+Hc7?a#>nb0Ly@W0J>U#LX^!ubW8&4FS8^6#jXt9+ zMdir1&@GalkFg^Wq;xETEI}pg&6%G#Ik6P)Onlhe|Ii4d1wBh+2)^(CTUiro#0^^( zY@o$-L(^glSP2q^BWw--#v&g>b-E8;eIXvJ?W*WD{E%wg?DpNrLMn&f=DH@m-%irK zSG?%K9j*LdMxz$id%fsA^Am_Ha0E)u8^eAtToN(mCZ)Fk{9?7(z_HZudx8BkexQxa zLi^C2Oxp{(cXdq9qi2VFe8V=7dV-{I3YFe}Q~;RROr~Jk?c*Y}5jf2h9H9_bNlk>1oN3OI@z9 z@~{Szg!xxng>~hvQ(b_7Tp?{Hzq$e1@C4kWqZEDN1K9 zd{e$#&81Gt?3p1Jy3vo6-4rLW$_TQOPy_75w%Z`XpZHe<-W>W^?h9Fmvsas-Y;qe^ zZ}~Uiu5P+f^=`G>YtTc^i1LrlPxKXG=sPnc{U6y8V z+XJoq@8||09`6vhc-*{3b_f*Ywg$-1IJH3vhR#;b)tZN^6WXERSO40zqu?o~Og|Dt zf*=|R^fF9t97#YIys;2OXp|Y2lcpVCh+GP%RNl%Vw(M?s#%jAbdj1B!eX@kOT+nOIZ7hUXN;OQC z1fJ@Z5f|Iq;P;P$Ds<|AH4pk3hpWeKf*O{p!xG9Z>=&h*)YTV?FG|P!XG;Rdd0?N` zZILS+aMjdzdYCVIpYGlDV?m&CXu5iOMf74>B5`6TuZEI|PB1)n*NW^Eiu4b)vnT`n zc)?H|6eyAy)bB!*w!_W~PiveytOVs@(MJOvdza_lURrDWGB?HZ?EWwoQWye7(oKN$ z93lfXJF0#`k;v~ParEfX2BNktxaP1M0nAAuo1J?NQ7g}j7owFw*X3bY*nQ6wR%_^6 z_w+jQu!P_e$MVdz&PiW~wO{=c{kDtCg&KaPkSw5O%jxRs8agYpfRF?TlF1O8wm7v4 z^OKk%I!+cWVZ921if{cI$lcJG`*((z@1ul0?%b003Z9cn;>R)-L^-L$`A?lb&V8}# zuG3DVtE;C`FLMsmAAf5^D|SS#<~RQNeW=<3a({AV&Be+rHR6z%F@Q5GRj`0HVPqE1 zA%^NLM|Tuir`~rO|En9eeb(E&y`JvRLaGK_=)A+DQ%~e@&j}Gx^X%GB7p%jDEfS8H zfkaENq4Fq%@k6E_L~cmQw>JqTG`6&;sy9P*1moLDv8kZD)hBjlj2nzh0hmP+4hbO0}RYIC9 zb9fjkTir`=Vd0ka(AOPl(9M4_OC9UE)thoqmbm6l-q=jDYV&`s`abxuXSkLB5(3^N z(Iz1qzJHby=&99aAJ8tG3c)at5fw~4a%kT!bn>K@IwRedQl)G-jAvWJ7tp#d9qg6p z>DX4cd7x35ZcDX7H7lI$PgqvZ_PAVFO4X329^Ja@-%3F~P#@~`@l174`-K+AYL-lC zVz(b+W&x?;LkHdQu=>f-0T)h2?-@Ulq?})o>oIK?AaBWjM4JvC@?v{^4X?Qs!@6}9 z=9%^+Dgcuisw-<0x!AX~pmJ(U%i}8AiYBWRZM;G^bbL{2HU(}@A6yqE{V7G<-A|&@ zbqRz*v0dz4tY++Nw_T$Wp32OGQi2OQtQU(%PZ(Zb(28cyrwq((j2bIj>`Gh; z!gux>o%(P*(rI|kbfZyi#MS=*2{lzm_jq|+=uG7}HE=**-ahsekK@|}-t=p=e3bbX zA)@NK!K$bM&-ug)Pk5*~LTd(AMw_t^xjv{X;M{v#QL8|^OX)@gkH@i`uW0q^sKS^q z8rqo3!7@fB`-&7({DTc0J65MJF@q8RkdQ5Q?JAd~jk)d+OJBPnGo{-s2WFt;?}v1A z{kr9|<=QCM%iJ=gGq+UJ*>ha>?)@l`3~??S;GEgm&Qy|GW7C&J8x)yt&D0wiLv_3W_TrnbI{Sl<))& zrK*gC#VI-)TYZ4iRPc>sOI~qIIGiUX`jI&-*#d+D1A^$~Y)ZF8-?g=~8QRH|yy{_; zaYMmXYDSVnfwsEJ4v^w9fD0aUB5nzzX0LWlYWvO2tOp|QOf-wncXxMxYs4t3`_Sk> zmZ!Y9J0IK7MG4#f7gw))i{))yi}9TVA zBwQo^A7S4e*W}r~AKSP3zID|)5J73BN)ZJV*-NW}qGClEvK1APC9(;twJORgRYYZ} zh{y~x?4c|L0ofab2w_8H@83BOUYcn1`u(D66Q7{_r#S z8%K*F8Hf0of>@qxa0E{uQv?oq@QoE%QIL^r434)C=jiZ9ncTV}8qw@Wd#$Xo%YPY@ zla*B*(#YGW8~(4luLj@lTosbd!5aK)VKKp@r4gjD&2g;dUsc8QK=U7LY@K3~HXd)U zB;W6`=Hk2)-e=^)yz&R!Q3DxOhK+KQNrDC_*nqH}x7*?0hpSThCbr-_EKmhyn28t> zZ}a2v6NU8Y|ClQ=^OqFN0199p#U9Vb2?bI1W~cN^f_)CxpD)q$SFhUq;xor*H|J62 z@R}fW-Y-1)AQ=D|yWWPr#{34DGM(r5cco_dTnCmat39b4m?T#iM1x?!O8b!ndWR{7 zIr2^+nHKJ5t33{bGjk&*Sj+Q|A*KbYIG&0-9vSd&)bSXqR(}Z*DNA%!-@n{~Pp@F2 zkeP8u>VxyQIMltbK%xRQ(Q_T{*A@(MzBI3L$X+ zO1lX-4c?a`buu$zBkrUy@;Fi-_+u^s8BTGpfjLUUpb1;SSU)yxvLt^f_YZOfIig3| zno4t$e$T%|=|5%%Qx~lq*W}r`EbF-$m~(i9NAD-ii}fhCp^p(-FMO{A|IaY7o=Eor zpK3Q|%Hgc3hm_wzieK-(LS&XLU@hX8%GG&g64C_h$hh&hRgCY*2S&cVxa6cgZG0Qd z2gB^*PJ?QyxWR39*B83o7sAMTqJbD^5~9`XYZrpZD+K^W9>6h=C+(Eur+YQ=24@bR zO%woR1?8776hCZ&vugj}=cE}edUm)2d9PeXwNiJ;UESx$BrwjOFz%#YUF=%|ul>vY z{XnJ^wcnH+p`H*npAtANzjb4J`tY`ifeY0D^#UoG7(O`!d2U3;J-P8BwVLnn&1ANeF-N8^?_gx1vB;x`&ep@HN~&Ic#~b`VI<(zrtuKD~62eX8 z>w9S8Eq|>~0qP2<4q5N2^&uiaK(cI%XY$(S??jgjnZPnOTE0+X7Lf!xt4G2piZ8vVG-C%=oNHe*eFt<0D%<^@x!#V{o@7Moa%%#VgC1>%5W~EsD|sLWUy$1|0oQZmbF0-+zn41VkfYtj{{L zzH;im(ryiEM7ZJEUQEpZ&3l$SyiItrH!y)Vo1Ut* zB=_)L0(jdsVU=g_U>r&hRMJ<1M&@p97AzGFE2eC*?~=@>)P&YiT>vG!MEofL^B z=KuFM*G_KRYq;S0{|F4vcbl3b%fDArsQ30V-$tH>d~Y6*@bYzzs84l6Jyp%qc=ALA zmAnte(ThPdG#^kQ)tjX5!Lo~Ss%$kI!C8ZIHaa>xLw{japK!iYuP?_vQ|~*0iGaag zmr;Au1|;P7e@NI5!4Gw-trsWy@H`CvQrQP>GZf z)iIGEfh>%w(}zx(BuILt#`3^Ihp`Z$uGX!$haW`27OwxKY8>oZbG~)U|1PYhgvlyN-kQs3aRa3n2po}M>pBm2>D}JuN5g3siXoHMPB_z% zn4O?%GC(czgDfN_jy?&ewA_}RI&a@D6c}oBidxhOo-Izz_($_gg-0TVB^>8N4_n@} zh^NGH;MN{SP5YJ;QA|2ow`^7QhZu6U$V}kW zM;^v_iBrZuj3AUK7! zWOb#Vw>mX7T!!)PaK!_c%oGzr39W}bOoDYlnh3|{Qf&@mn}9r#5F4rMN4_Nyj)BwI zZ6_CKCkp-Nolh)KTaLjr2|yY}Mee-{xi~Eyy=MW?$TtN#QCGJ_)F9z1vJHMaLy0oZWdi`=V(#(z4(zliBl)VZf}Z-4BWa`^pst!;$YZFL3H@U`{r`GXLybpJ1#4 zwJ6kLu|}-+G6gDlKAFh^BYrB4eg!6DIdxBydSkqn7_f)`$(~Ez+R0jj=Nt;2^Dwe! zVZrUN(N;A0l@_^3&a`>+Fz0$>Ok2%RU0^c4rx~f)F@Yp?00?B2sP@XY8xt4 zaCxP%&=XS<(jgZ0|rRE6TXnZ8V zytxkXM67Lx0nNmf`@o$JfOw)|Ok8zvVD<_)@FFG+o5A_8LOA#O;&do~w*t1$4Ki!L zC@HKjCiB*}zaV%HdJ80!u?f2sD4yQN)TW>-nYK=nGjxaEC41lT%!&tm_((fh=z_Ud5c6gw zLaxzMLxdTe5WmDG5&>TTpRrmGMpQ6Hfl&=P^i=IS@(``jS-;!AKfDb63b+MPZy+NBMlAbm)?%NON> z#M5|&u+WDUvKUxec}TA#RS!0*Vij2r{uV)?SRzO6$w!TXkGa-?KTXcLyOA#?ws<57 zJbL2ipvNP09`Un6%^k>g#^ujMm`h7k?_yWZkISxl5os*(!%jY5Y-V zGTorGUhfH0DnF|j@RG=X`)ebRdj*|?GZ+V8W}?j4m|{tknRg@yi9>ORi|$8c48bt> zZfqWH9kn_`B(H(O=SX;ESC=~FB8&?lM*1MQW2_TKQ2gl_L`l75;qT3-ng=MR6zUpQ z0@G;SbucoC0VG{{ASue)Ri{KOT4{#2T3_o;Kx$?2pFFwHi~01q;D@WI?7NqNm*d!- zX6R1aaEcvZj%*$AuJmff>1LTu^=l|Y>FX1E)e%nyOi$L9={~S>B}5-Tfjfetz1yo- zuh=vv!y8k36MAKahKF;ZL8#f=a`AN+-apiIBHQas_3L=#WS3Z27t;^hw~VxiSp^1& zotLGVGu^Az&TRirCDb_&-NELWua5=x3m|Y2tF~CaRVA+%eGJO@ep&d;%v*3TU<3;2 z0Z&MPTREgdl>pgx5zXU@y(BRZx+XyqIjFqTZJ1p+MfF|`SJ33OVsUy7%z#eKV2V)| z?znL@rD;Q;W)oM)bO}8OHoLuTx^nmiqqm%6FPZIL8}lu1vb2ectUVgcM>=4<|2W%| z005CA)!ikc-S>Wq_27N;KJ^j9gkH-LSg6#d;ixFeIF1>@2K!zsat~cQji-L(vD2TwEI z+(W6wdk-qEiWs;JGpr%2ebVdpj67u+5wzmw?G_H{h@&kY|4s+dyfKtg?VwrHeQtf( zu(e#!EFKuol3C{==WO-5NHLSZ0I>CzG#E4%EO~25E%wQF!fvU%qvs~W>FU-EOynLo zS{MiHUGwd$9jUtv3et(e=Cqf$Tg11h#x!$!d+^Yo6sVdH)>L04(D6`6K)fEizTaL0 zLssc8LEA07xlTG&Y;hCa(&j^YvYYRw25#nE;z)kK%JXB{#`*J`9rGp$F|=G`Su3{F zKZ%+(5}*QNrWK4>yi``T6@abv&BR?C@2t+|!Okco4V38Ix_D<6-anKwe<9kcS4cS3 zkcv*PUXwSNEjep{cHtZN>g=xt8IgmB!+GZP;;*0=CE=LSkkto5&G^B_qrnh!86#>M zwDO6r38<^xOKGx^8zp9k0-=sKtFh0d13u}tvzOBb2UH-*nKG;DKV~{BqqZdYS--#G zUlj&R5Pluuut=s$P_P(>#q}l-xG134IMxq^AHhz;47GW51Gqn;5B6xNq%iY1%|G6~ zj})Jgyr<+U^4VTYjnu8EUlQyHE+d2z3Ve!lmMpxizMdm+{GI*70=&d@|7Mlm|HinNDdauj%a zD^!i|jcO3td0>P_dCw4pLHcy zSr^E1RT&YrRuQPRM(c?~L`N&&qqqT%QyPq1-ColO!XpM{2UFoZfFCAO%}2|+Q5bSc zLRE&bZc#vd38RMuqkvHoFy8+0!%T``{-Iy`K#Y}F?+8{4QqeQHN9`FSOS2?3Yv#7+ zvs)B*XpQ(U5bg`)F7PnEcGvbbp3DG(MHk4^X6pJz!q#l3OBr;#_Z@`Me^8(&NQ3|_ zkoTMgu@i}OU>fozW$Py{1Xfp)n>yOSa`A91Ef;+do~uF=Ve!PZ-LYD0WX3)a>{f`) zMb-cOZ2!UU*6pp@OXozV3x{;w{}!_#YWhknbvlR0mZwwzg?86`r|Hy<4*MB#3yN3NX>Ia_iJvPwiu zieu)kzuiG}6-5bO*45Qjs$!qE2RtuK?ip*K%dkN+lJETlRy$U+cXwSlPgCmiM96DN z4?um%^o#mZwJwp~Et1;>bBeA8K~x}#g|u8A8OEI?=%NX#HG&x~=!`m;B*~=$q_io` z)~r%lHP;}-UnQw>C3Ff{2s;f4y%k4a%auRnSN_bYPXGkb()tqV4$5n5IaU{g$)kwv z5NK}nm>oWlx+LeS9lWTTx;B5iH_?TH2x<&3<~qdyo*knY)D6`XI%no6Ig<_3E{JOu zsOoG3yrKn=f(rpRDGVy*(ityMZCyJ%ur`&-(r8O&1jyYzLzfu7nd83RA6Z$aX!Cl*YB|l3_8E+dhjeo$ zDI9KIiu_tki8*IgK_=aNv<>*w*&z>l5~)>DExobr7IeN}R-;-;ErKB_#o{<4HEFEX zm4>Se#Q!fo35Sb-&ZerakLle8J}8cXK-M#)4$d3uEa&lK0@}?hhpdVidq<&}^7D~+ zfBl{wK#|9_G_9@rzK<}TMA}Rq;6ql5fRHi7dm{*y5dTRSZ#t*AX*&A}-s#&E)$XRJ_qwTfw%MsTP_Z#!|X5;8JPiWG#8E3W=>9fWpV6e1lB*~P`dzM$Amx)!2eSlGA1|e zRvQR|hhtlF7NW$%==0KGfT%vvTnptc+a-OnbBp{U>RWx0R_^?fT&E01Wzi3NTAsR# z^Jf?c@{{9g{pL3UUZ!Sy8p(eLDEoh4G}})09{=T+k9n8w3vTArKk(y3k z^TVpYhI$>EE17zi)}L;wiOOqsQN=T^U2*UkpP~NVjd5kyupOuUz$-?Sbvhq{ih#^ZuZ|UgM%GBT1G$>^;`SD^VTYnJ zUSEq&Q;xmXm$}=Et`9NgCJuPVU-!7oueGl99G@1<_ByXB;tdFKn3S;xOPOW7xlnZ5 zcA8R+`dXn=ydSu(e(_hjTbuD)*d156_3v4A#y@hfcJ60fjAb8tL9_c}FCU`eIBj5yebaed6qIbDI zEx#(U&e0hthGV)9hn)Y06y0l6Uf9iGWcIqvGQwZMehd}LYnZiAmecN0x_9}1<8~uv z%y#B&hRTUfgZ`31NBMu{nM5)wsY{>LGq&=;3cw7iz(j;v)Xg z$j8yniMwi~o$4WO`P<#KX)nSf^}Zn;FyKjl%4cGm#BMF4?X|umF#omAnj8)7(sh)^ z;2?|d1@rP+>Eq-8qg?GXb(0j0F(zC_v^0ki)p|C+JuK0>@V_LZ%H6KN|BPLJ#7!_` zDCckAEi8^Z_jR^i>!snc9lllX*70c7vz=OdmiD*(;s7^ji~sp#xjR%e5hb$Psm5|v-CUwK^L zt?M@57DiArFf1&-BT`M&YN`U;s!(g>mFjxw%`)FX((S&lbT_WuD&J&1^vj-hn8}+_ z8$NqPv5+aDQ?o{RS8Q&$_T2e8$4X8~x^XA_6d7%Y$j=@q_!!kq{NVzE;VQ@HqsSzn zHT=q#oK#i4MYES>iG_dluBfv*7CAg4-Mjtwh6Q^jv>F~?v4mX4$Y;5F@XA{ihR9wh zvD=mD;TJv_#hgly;+wJZxyofrepU8I60P(yGdAD3S8;CV>gqp`bO8s&bhit?Q**%| z+Q9AAPMzr4trcq!$!Ye<(t&Z-X$iYlw3b0Uta)dIH?Dc~ZdUq)^nQ~NPi3p?3QM2j z&6U}w=wo!>r36Dps>?N!EubhM66@pT!jaf(50seTxpxKCNxot&3s)-*K^CrnEZ`%~$wEw&>{3C$yc%Qy zb`a%LeYZQy)pC2JdLk_MoM|t#;e^Ro)5oU%nRz3=^p|!DgQRcoB5d-W2XZWbDX&6O zyOOtvNVz<%GXP!cZb} z^XvQ#dt0<0Y&6T)E3z2~{dBv^wBQj;0qC~sb&$bEI3S6-YeD)I(k)2wV5ABTio(R@g~~AH7${%?!8ylZx3!g39=3D=@$K;*m_1MQ&w`~RnhXKz^H9XN z?x85F)jP29WRF73Ky-lE#HXe!?PbCxriZv&KXy;FmW@MKR`#nAroZYfyW$O;M8t_o zr~q!Hb{H_|0=$m-fGD_t{T7h8wgG@=;Gm_rXLILr$T8t<+Q3xfz|mLd+FF4mB`WRaS8 z&;{4j)VrM&T^Tp*RiFXzO)O&(oU>4}-S=tSFL3(WBc&~tezLO@F@g<(>V<*USVu(! zVbXZJFEO12Ap3ieN}L$44EK-Hx!3qh3qSLb5@LB@5n^v+ZTUU9wzo%e>SrAeNeTa+ z(mvkiGB#v6ZI-)tYs{IF4tl^hZSJ`xgXUHYrn`2wx;l5xM_iX><>8W#BT)}QyVT~R zv&~cL(^15m1P%-HN{_zkE6Kc^mi~Sipx-s=kVRL6;yp~b7DJj+j3+VU0mZ|yG$&Rm z0uoh3pgCrfK3Btfo~d|f*54^Ta>Ape={=i{lZL1HK_-Po{T(dZPtW3zJn`&tI7CK0 zSYyCEgyYO0u;K&%5_i9J^NU&c=H@L@I;P+pH`fnT4?{W~n?_YN&5mHVE|Pvh703vJ zOnzD!%Z!}frKSkGv`I=g<3_WHLWogW<)|>Nc{XTWe{E{q64kF_45}Y)i%{6PWk=0$ zgxkOkt{mw4Y=r?yo*JgxJ1joR&XF~`;GwwKMdBK{276tA z3u{7t8Xzn3IanW2OTK;y?t~zVHt6{2&z6!l&6)GprKZM5C8gMiBHM=#x>G<*?1u8G zD8^$7)n5Bl2(4r2(9fUJ-CdlLr~=C$*XNJFTzpd`J#^xKI86ygrfA-c zP3=WgV&^c!BQrZ*l8-nh%$~ZZ0l=)-Yj7m}t&`1~TlDibj@OGzVG8PF!>Bv>?P8h9 z9vdk2d|$ozpe$s>s-D50&Iu#6TSD$dC*eoa*p@sh@Ua9KtDnObzXY@B6=caXG zecsb?D2ON;KzW-3?_{1Rgj*oMm}bi!5a$QIP3}bU1c}Z7s`C8W4yNCo|Bz<;GFy+&`#=68q$ zgd@`q^qa%vOjn$-umUjXYS4ef*r^2xI=(0QzGXm&g9 zSiXd#G+LS6=wDN(wuOB?m@)=*(uv57CYR)bg8=Ky8;l)9Y*F85~A9A2^T{nlOgDX!9?7&tX`63AQTKGGgCXhnKZCuf>d+@h6bWFIj}Tl^2?ome8NcH? zB~JGmKgD|lx(=}C6uCm~E3!KL&TR1|ZXRN7?pi47zHr+5NB&siglgP;g&x|lI`RQ| zEAeg3qQb-WxgZ{f0T`yAi3|Igf7G1$`5IYRYLqt^{IVFx%Oj}bd*%wr{4-~*oE-2g zwkhP3z*FutVEPpX)&;}UjKh>TH6{%$SB@T32N5U1fr`yVxCheH zLCt$kl~9*uD%|Ra-bvuE13?l6WSm;kmw5tsM!CeoH3?%rH^>+z4Bz4mnT{xV6tQ^2 zNkN;Lq_}oF`a{IbC`j;u@;Y2&qjOrEKF=vc5H(^)m;_5(Cudm`Fbs8ebhNcb4+v#g zTL&@6@CPmlt3&wq+4VP~LN1q2QLN>!IWiw`2Q+8?@GK<=27pHjC?jbIQ0x4K>8|HL zY_6oQ>F`K@@f^;{glXbPet$J}34hR8rJ!%>DH4X2O(Dxkv`7^R=g9IKYh^a)Lq!@| zDda#WsEN~GD+@r&@(FDX2=-u<8)AZBdXRT9)6UFa)cNG&_m5inZVw+;y?Lsl*Hq5W z!dPcWl%FVhcl|E9rM>mj+ON;HBY-jP&3bG{*(l^ zW>xK=bB6d`SC9e@QlOa`*EGo9{xwy?B3qN-XQ zCu1Vw3u(-O2vz7CfECMwCV_Q`3x?LX9GC$uBEFV4m_DR@z8G94V{p~)AcmV^Iy+)| znlhQmRB2ea{jknMagwyW+mFp?hcJOT)GN7y7MW_D)9|QxoGFE}dEW1Z3KJqHck5o80euQz~e49Zs0D$ybUf;wb_yz-}EnScRHl^ddGsA!P zJsiy>)9z$^y4`GApdZ!m$>$N8JLHov%E?@Lk5r!XHi-;aFhoVs!`N(2j3;}(f9Qbs z;F#$cSH?(m%dNh?KBRE%0bR`i=;2)yO9>R5iU1Obf zgPA>g9D68K6oUYsFU*3Sv?3z+l^fyJU5fGY%5ptGxV7w zJB!TfR!_%t$PP=ny=!&I0Gq!9 z;Evp%Z*%^LaU2YDFf+D`QBuHQ4#9&!zuFJ*OQab!0jh+giLP)O@&BsW&VoMUN_giExY|=k5$XQcXRFE<_6i( z$MhFW*2^tqE}KF4wdnh6+$PFx;9nCum1fm=zwAxEIQ=0^)Jy9%Hbp?}zNUd&qEg+{ zk#7N5wFx;NF&e?Z8%hA~;ZZ5OOp@CkKy}0o{&rMk+n~Ssw&W%l&M{X50(85cBV^#e z{8ytHk{-(EICD)%B_cA5cg*)%apbp*yhe%E%5n*X5T-xa0E8Hr%|A_<_qIO$iCB3Y z!2;XR=WTS+nxM!g-E0NJPx2Qg2-xl51(e*2&c&3ZcQ%vk*z61{rP%fa#hdV7%rU;h35i4n5ytv$wlyV}Ha2dBfr*o?h?GF6B20vr+Bd5w z4n(+KjQ+I2;v`@p%kNQV+Bf{*v0wGm0u}&1zF)&-b5fsE3L? zO=hE+odQC@SKqPaWMySN)MjbCslfD>F=XAi*Wq`s*Qk-LTbsk>n?Pdtuw%KK=i5&rfbPd=p2*on4R zq8AJp4&>S?{vsLG=(^%4HRH@L6A`dl@#RB4m908!J24l?V`uI+5jV3}vU?*g?83h% zVat6hYgCozL&Jo!WoOF)oY{)pWcTy>S`2J}Hip3M0mbQl?ddL6&)Q+?cQ6;wdLgQH z2QBmJ@6W+u z2{Bw1B9iTLGe$o1VHlM72!cvrEMkNCLRY#ismc($9|ArFK}Ff=p_kR&>-O#s>5XSd z>%hotdk{Ut?y<#pvhZGT(@f_{Ek<>Dm_tvWc;cFz{%VzvLyimojjzsTf@)rSl!pl- z{5_RTk4~=RQ;l*%hFCA9tCKa^dUHdJVt_OmQ7mSnC_x zD?(EPfxUl&2$|pBYUOgobyM!L6`b#>j70eroYxN6!|)FrH6oW2*q*diBx4lC26S(j zu^y9%&OddQ+IV>ARz|X@HI@@htKwS=GdY(;af19?A~hv`-VS*|qMV=r<%G&lL*=Ki z6wVzyBC~(BC$May>>|aE&2+a;bCsbH1&Eq5$BL~p*5i0umn)O$uA5U#*|aQ9*XOhH zc${MfW*oZAI=0Ql@;a;YjMs`T@l@TC=QSn~(b;v9!rEZ0izkH8sF#xW=mFFkYNVXf z($cs^SgX{?yZi>y80)clXDtvll~z|*QzwA~n;k}e_uYkXtUUdTmC^Mbfs$Ar*JQo< z^73f^lrw2GWq;$XP~qOJr2u>-&BHlqPR&1ri^h+#XKGzPY(FtEll1AK$HNML1$vf> z@ovX}%+a?h?;e*mIkmj*UtEQA?nO^W_ozNO(Up`14XqN+SI(pfqzo^egj$-ZqMpo}+KnZq@Z9@tT|H#nc^{z!F;Yrq!_E+Txp z6g;AA!gE@Wi9bl+P+9Wuk@VW0Gxf~WMs(f^rsUCna#-2ZHB=S53Gq8-e z2GU1c);fs+TvT-5sKze3tn?cSBy=GpjLn6b@0b6wY{uXDrnH*AP@ep?NuZ&EYprY} z^dT?)qTtTuzE~X$Pp@1>%JUEQT8Bh$&+pgmtIDH6jY1elN>VRrYWtyqgRYEq>q@C) ziJpAf8Q91##0h@dIi{k~>p}};4bBJ(K!$TQLnvV0NqKjm}OM(ObK)!&LYJE1Tu^i)@(?EQ8R^2!(vkcl9)^YRpZhY?z<#qSQl1W5WIf- z_Gb&453UamC6(%S4+63zn0f`Ywk7eGG(}3@@&w{hU&rVn#|6*bxJ=3lggX51+HtjRP^AhoUaNrl zqFiunnrfgWwSM1DUwts^9++G4GM`){2rGY5w6TK9Sh(!sSKYJ|{(mo6XK$x>X2m5d zmgu(lqD7%Y{->i25QPy973eg?T&_?II+7FDU)sKT zW)exLq#&iO(ZJF{p!MRxD;4+0{847#HsB$+vV=tPiYwvv&!tN77T`gRIruO~CkE-t z3kCOijX8XZL=DQG+mw)-){E2kY|*D>LT4d%w`e?OdEF!S6Yg3uB|}%aN+wUqc6$nQ z?7CLKwCq`%uExW8?}!+Au#fcX>rY3qOR-mP8;~3rbKNVlhEf=)9MCN#dlG<;jOxbf zfI;FSoWNQ<3(-!wSRT+p1ZiI8<2PoQjy5%iSmwEeXaOWdpQJ(Q&6Kej&kkG$tyT@B z9qWIWOMT}jE`(#ksN1xo@xtz`>=ub;Tya=m9A%AyGh#C<`Ytr@>IR;&y$D9 zpVE#w$tZ_hc|6oVJz|?`qZra<`x&osLLpTH>LsrW~qGi)2V{Mug;UhT@HGrp8O*cW(if2!#d56vo}_y>kH(j z$a3QZ1xPx?{v)#Z@GQLl^=BZ|l(iSsPRK2kq?qe%?mO(~P<|~wU%P2MX5mZ3YjEXx zr?b#q53NA?pTu%uxa}eDcW-TrT(O%1F^#((9ZR4#TrULX)|oo8^35fD$DZ7e zTyaJ-TK9=X-d{7lvZ)N&zHNJ$O~pX0MGDkRfx1lsV7?)zC7feH|Ebr)l<+i21x6Tc ztCZ|`a$}eOD6p5RNH8U+7cQe9W|4Y1&Cn4BEB%#m-BfautcVJ;N&K**+u}GIlH>p! zz;U+h0YFIip(U)7sqWs!!uD44D zcWM7)l?YXmxQzOI(v!`Fv<;&W>#IsYAG57@agaJx%*nxn0l;CaKxZu*-xh6P=n>gMQnJ=NXXP8=x_ffORbr4zsWB*}E5B57-NN?p`(To%dY>|v( z_tD*2rAr0jS-fHN*04(Apq205df*(=72a3c5oZPPP=)sQ8W^@5V>;tk4o`5x&%Ya1vhFVE-8Qi)n#pDHdIGffpfu79GVELEOHFU(tF zAWyt3FM|4hV^!Wwe&Z0r z!Fa{BD!-uglPQzDkfML?HGz61(9Zz>fM=W8-EIpB@1n;d#_E^iI{bI_pTP(Ea>3x) zD~g6@Ht0xQ7u*tR@%ENf3dl{pRUed@ zf${TeJM`WWW)hZWM3%9yPnnoe)cjJ=2fN&$Bmy*L3yK$d0ShimLvt~%jY_1ta&5%r zxO25E8JJ#%k7j1bj@T)udk#v|tGv`wo@p#OyyA6@5Cbk~xhBb%oyo)%B7qK5n$KIdJ-m7v$lzHjZR722ERjQ)aCiEzD!}9rJ zC!C6iYg6banFRI7g3Di=7g(bM{=Mz~L2{Wn!4uUhayX`^>CierQ*jIkNF_x73&v}?JZmbCCu>Js$o8MQg5?q5}C2w3&Vi_pFgIZVh*%ZHSQUffxE6?lvQ zLJ#ze>AyNc2Zsjh^Ix-Jx~(bFCYw4e(>3}}ZV)5xk{)wZ28|QtXiK^-t^y3zCdJEe(8sRrUUzF(QV;`zOGF3P@Ru3=Ud z1>jLakK6YzJqWtr8^B43D228|0Pje;mo2+53dM+HCB*sy5h9TI_n)-eSvd|IJRCtm zLF-~>(jU74?MD!6$0a&>%=gfW3zO`_G8<-Ti5@v+8-;CGFs-d_Ty^MMs}~EZ(F>;; z1P_87K`$C4)yk67r1bRkh))WY+y07%j-9Xgu_Bh`lV>#TKk5!ClQ&!mR&kTF?|9L& zxm!Y(Nr`o&!^*ze0u&FBwdX42-U~^D&0DqrBI3UGjN-W8U>(H?+r^=56#3lqZF2i# z!`7DDTpljMqc3-XMLYEsJ!8&DF`j6Rm9 zi82Fb0jlit@u73gGPr0^&~bGc?k~Cs84PH8i@3&R0%!}_pA(}Eh`c`dJA5p|jz%rwF71Ekt{yXs;T~Y=xL)l2j^Nkcz8xy3T)DE-<+95AeAwu~?`DUf zN+kvF=5lPaotT{Dd&Cly?3elWEu~h0Ra&5-0K^|rZ8aO-KK73yZ zRW%kEPXYe$IX(6~vY}&H%pGcHE@Ro!58W)CHlc$pHqjBpBMo`*FaN#xr`oujO~1EV zM?&wnL|KVVI z$W#4zR{)gjZK_k7HsUl?;EP#j7(fpLn$wcmsS>_GW+9w^@-s#MTEB@&;LIJ7mt5jy z6diyVK-YaC{PLosEkd9N5w8cwqa%Ea3@c?3c;T!mRMm|KtQ(9rp zXa-&e6=DEgf$oqI-i?Edr-+_Bs)ouWKiiB4{?fv~hj;7r zh+G+*H3$G+`r~7CV;>)qva3r$e_;~C5R1b=eqeX~OY*;$zNyjdAJ1y~@$V;}&64`-Nlb%@1 z1liFq9!t1JrZQgods2YUj(=9;q&OVH>SdnbYf~*OBZ<{=C>S}e>$l05Y`qsab#V&Z zx9(a?(H$_n#uaB0Hcj_g=l`?g_&gvspuv|x1{s5BL>y<`7Er#}9$eJJL&8d$_7%D( z`k|)AgdI!-8zBxMyL_@ilt9i%7i@@m0WXuM-{FTr<~JtzHfAFojT_YH%(^ zGwXZvi1KZ@Z_@@W;b9V%qy^o&>^`E$+8wS(4%$ zprqN^aTiM(Ve-^51%mR7kR?HR&3`OY0O>fF)gk3>h;ti*xN6(VO}kxinm6P+hW+R-RP*Q|_gk^)7~t;l`qfN_ zX<>1y#LP_i6zdYsfywe|6xuQ>BeNv&T9Xr>`fqwS6wW!&!aY;#%LxF0eEWNSI5#+vByxistCUOx{0 z_GRXMC3_^tyr(F^E8(w*W!xrX39dkBVupNtiKT=;!OVrt9Z#P=Rg2a=h=jXlZ+^ag z7~qU0c6OJbPYUU9b26t2Gjq@;%>^E#vQa)IqbUTa*t}Sr2Xslo=_r=71D|=7U-V&%{%Q&OgBF{{o;?>v_V+U%p%W2jZbT{py$N8=6DoFZ&3X|86r{wgETCxs zSXuyR`Hz00b;H%^FoH=*p#%n4`YRGP6G08>6jMOW0?NUrjZcR??sCyxfTTuF+e{2~ z!aSE*KY6IU0=FZGT8S;_A}J)5L^yfX%plH?#3|r;d3ho=8GWMzzd60A&f@?U2r;qE zIum6+7eHDYI<_u6)Z*7Hh&%5JY!lziV{fP&0aW0&j+tC{#@-ZhGuHB~lH<1JUi@v6 zZZ=`EfSUvoZl{WX(LfzdCic0d4f^-HDTb^^PtA15!;F$ucZZL_WWj zYvz=sVtp#4+!ufMcfCc~>U%r>@i7W?7`cN-@p^Ia8AQUQH34oQIQ~^W*X(%MHUW#q z)d>&w&1gb28DFmGifu4Q$cuk*;k~o0m_g!Y&P?+-EJ$0>Ylk{Q-0EYrIKv|)Ti;FYsl8f6o!2PWycn-wD>hk z0?5Q5EfZ4|Z|6Uo5(WPA+8NznG5hex3WCEY_@2!&+UT!zP6C1j@_Yw!?X3F^YW}{V zQ|^}8kg5z9bebKsY8dBQnrz#$#9S+Y9>!FJS=O=@z^@34d#}csL5397w`;_HLUJkS z4ER-(Wnm@l?@YTxPVan`n`wWUAGRNBZ&BO(_;|?TWhV(^Smtu2(we-9wBgd7i)9$< zjUg+g81y2}b+7Sy7hl*)8-)*OfeAVouokSL_gSuc-}|(-1}(qvgaXBoJz8Egp)ZOd zS6?*#Th4g^rMVg@RNj|-v^MehUHk20T6?aJg0^$zt1$%;HtPOkVN($e(zM#(I z4(G;=+C!(KXGTWG>r1RInN`SW9yQ86Utc+CU+UQ`!8w~8w!Hs$rmt4~mb$#&{qrf= zl?b>r%&h2Lx$-L}K!8_GQ#ip*wi&?GG7fh3C_Dh_+sJZsJ836`S~_AEe!9Ho0VH}t zZgbB}Jn;IW<@@RrAr&m85T(eAcb@ZL*qj7uv*%Vtm3qMZiviqDDqAcbf!d;8Gpl&yQ~kyswK$MCNo)E zk_QPFt!O5#N^GCV$^4%85mP^%ynO-8Ui!eD8!K9QX7_H&ba)YD-EL`)t2nt--=K)nXj~h?Q+NI4+Q)gqYxxd2UR0Zv3>Or+*GBO z*LLco$?Q%U9x@XhNx&&Y15e0RHN8uD2g~+A=WjM;L~W!rzsQhZotV0gE5!h;NLzFK z!VMCyHDeN}*ry#3)>@(DV9gs>TMrXA2%tDpv3kG<5jdiQk56{;1}Aui_IL)~CVEo= z{#J;Q7o%$3Y~_}lRDnKs)y8m6kDE@m7NF2UG+N z`1W0J{xTvXAW=9+&!#^4@cWsu$*oTOFl37zvqLzQ+zciN=YdzKl*C-^nt$MzH+nDc zZ*4dPptBPl{Z36DlCyeCQ)6BUbCJ%xa=fYspgUKhRTrtbzGKt;VFj1{H9&R(PHK+i z*O1Hip6pCUUM3w>BnA$e*cgG40yfPe`nE-#X~pM1JTYlXF%+8VE`~w~$9?YBY|*v( zcS|)TnAA^=3zRy^cjrB?;YR;=_bY+aAKUyeeqta18K-W~PEcmIU=~Wx$Y}SRMzt%h zZr7Z-NndH3(dQx?p9nO=(9cEmbwWyYC$;WsxI62nR--@B0|{J}bBc@0H9sJdE*^PY z_2byPUT(8je*$mO;-|@yD}!Q!%N3ZD0LlDd$Ynfy17$seEcDDNX<=Y#CmN54?ry-C za05XCwJGoXJ`77zpkeod(w3Kape;m5=2{0 zDcP<(8@AWDV9WNlXWon>4NdFpOLyK?uTIr2YklaxH{wU{9eH^x^KQoOUQ_tmPrTo@ zS^LDL9m4_tbJZUiFM3Dk-s&D7ZaZArAZ6z{G{p)S<@@eP(5^LoT)C8KpJ1_Bp4bdw z9fuTmPp#5Cie{V5#l_21^L-tE>QB!?<3LHD5OJ&(go)GZL>n#6G+30MT(=Gf1qebm zuRjM6P22K4wz?vOw2nF*+8iFXD=bW+IT^sUqq6BEvYc!5&Y%A|3hF_GEi*IQi_ykk z5{s|4@Lo4=d?5H_!}UhNjG>ReNM3s3H2y|u>UC;ju4H2Jk%ey&uOg|`LVlOc>X*K8 zj=LKB?T6mFRBO|o##B)tYJ@?>IT!cc>38TSn@x=|aPkEQkD{LoIGg-PMs71$BwEs0 ztE}ul+&r1#R9I4z;Xv-GjpIS5jK~GRU+SdIPWO9`Lg{WV=bDmGmW9v{ktvpD{4+!2 z2Aod8&c5Fiw^k_S&^_E}uXOwaZ8tH&Hli8B4qV2|3>8oMLpAmi+_P@vs@vIWYHBVw zZ{7@?$OOQB9&~VQ;5ilG=A+{%t{5Q2PRW?K>L=qMq3tw1EN0$*0;;z=^Jr%A#vJ9H zLp-u8ZXB$z+CasG@A^wx=N0=)>QzT+5wvSC_-;L$j=kAy z9hMoXnl8B;yZV}C?>bEl_{L~vU0SAZK9;k8n_ZZ@)2Fq!YMD=vgu%*M)2m&(B$tuy zR$N?>H6BeK?W;*ATU#+>L+1A@%Qhz3@Q1`b#krZ&RoATwPK?@{b8Ur z4-bZ}-Opg~Ew5!4I+@~9zs6m!;elB)(>g?pz{%7-az-C8Lyz2!F22s^_cb|TVRr=@ z1n+9Govq~+nyw7x?~_gOW3KTT~AG{WSE+4zN8<~RhYk% zQRDey>9ZN6!cu2Tq>@Ml7Y{RM8|Hs1YtujspYPbvZ}WGl}|8gN=USylS%ihCubZV%IKm zIy5gi7^daV`SfI8hKBe2hxDsba(&+CTUh)iD`oTQu(8&e)9*IgD=?m}FSqFk6_I3; zH5*@)p4{jU(z{4+zPeo&!%+ttsA4`K(c}zq54twr$10|u`QR6SGh(^0rPVUh{kLB^ zUwgW$DO>Ya2=I=+Y!$laQonXjZhUv#nZ-BkY`(t}4*dSMsNa_z*!8vLd3jJP_m#t7 zjKf2%ef2Z}E5?VICFk^-M_p727HiKFx^>@nmfNMfGPUQmL|=*{p41f{{P}0TIH=-R zK$u+)Zal554Ewkt^z9l_RlYv7E?C-{C@V9(DfXAN0=_Iyz%zXW@m{YpdyqqC_SJsL zwuqftIgz*X7=g032i>;#Fae0Ok7wF0FsNiWWo<_ERwdaL? zkLzp4omwoo0-Y?D=&-4+|GFufd8~EVzu&{#S zpG)o1N@NGGvh`1r4axO4nIXfGFZSpPhlciUTb8xkF!dPATt3O@`=EmxN1co|n12^| z(Y^z$yF}BHWza9Rs{L+V_Mf9cq@R5Evi=#r%F5K`X5FotR(WB4wicWv=_*-8i(BPL zS^^i{m$JP6c**a`7Je5ox&KDbH>6)x z{aI$INu>Mz?KgMISRF|YnLRaczekw8rvrzzo>Zem7MpOmJQeb zue3th>KoFA@M4z9ZAepicA`Gvb&ngO0{*{GqIoP>?;DftK?Q5ho{! zbx#SIgbijN-iS@e%(gQ5oHbVoHN|T^{XbF}`@O|G@cDgK{k|<{&1Pp#uUaY7JNEvL zfs{qs(#_IG)%nai@axAj%~-X6TMpq*NiEVIbkG@U zD>Lmctq)Yc(%M2ZwkkL`yv}U0C3x(fZ(^>8INKH@izcSVp@IrD5JUj`RXZtcT|FFB zlNwY|{IK(8xr6;oso9dlsg8&9f*J-e$zRccx<$PY!&^UOjK7RMq{0|rh|ElWamFPa zn&s;+sUF@x@0c*1xX(*lmD?I{(0-)uJlHF=sBgh+pJP&7B3{(Lz5QSO-}WKewvj5v zCYu=7K>J(gTo)~io7OQYBgvHe5^)FdRolDUI^q8~la>jK;&z}tNO($|2dH<+mN-qdQ)sz`6`U0HNR%`qQ@Jr_X;|8HnweZwZHtM zS^m@DhgScOvF{FO>U#f9>(AQODy533EQ=LX5KxdIORK=w0f@+wsVtFYL_*k9s~}TF z0fm5qfb6}eLO>ZohU}Th5W+haG`sP54PqR~b0gNG3B#->FlYIvt?+I~lO; zb+7hIy#*EiA7 zYHOk&?bkc4K-A*fMsz!fSZoiSc$q0~;WC%Dm_M1Z7PYRg#y*e?#c^LA2QBVyDLyUC zn}f^!q5>34(QJ5pApLo^VHECn_V1vE1Yf7PPlJ&xCONSY^As3fbOX<>>ZG&oa3I{CHeS6xc{}ift0_on^S>` zVA+x~-^!?`Kb$f>!!qFXywXLj$H3C#4>~2Sdx|Wo{lZU<9LC)sOz7RQFN;nQw_T!$ z2g6O({>~R)NnE(b`X0b`l00WCdRRpnq8N+n)f7*{_Un6LR&9IYArra<1K~UT&#bEt z&-P0h_f#&|ZUUh$I+h#V{{TJX9)*MPSf!^o=ZmD6@lx@FFlqSGrdRYuEe>k0v@dQN z42nnLaDbxR8n@>71E3CoaWnKY8%aO=SEvMAN4J5y^JIUmp6L;3Bk3M<7~?*U?)3B*a(7 z;vK*ZAPyLKlB65C#$CIgIR44W=5yRO2_bHa>P_@q2JZWGQLVhue1c9PjczXrch`<5 zGAQYzqPzcxQW8UuxDX}-;#8@#dVvZ{ZuxNn~Q#;=!0HPmgYo&2!kOXx*AZu7vHBIYSMzbES-EfvioppMj|G6~e(ysN_V1*D4+j(6pZeuf=g*%*e0)1qy3e zk3){ctYsICaCT$2|MeYxD`-Bdd@eU5UY_|erFwBmJPkt`ZFhnxlzp37VQ$ebdzJ5L zt-+SjgA-j+e4G*xroA#gM+{fAfdgcJ@&Au3+#JL^F8Q%FK*S@j>~LUP)f%*p&{?Ie zNcv_{v5*tRSEVAt1DMB5nepU*J>34{r{5RwElH_~v>s0hU#B{AkehfP-$k?UvtK(b zE66)#MiDJ8!hrMZ(s6ah%8uqQE@(i}8aFd;!M>49aatL>lCkFWag!Fnslh;f`b@%v!Y4*@?l_T&By!-l_nsd4#? z{oj5+KfZAwJLK=pcP{LY9q(8q=TJX+n@+tZTUj;Cfc;*^X^JMpxcR{+jt}AA^}b$q zb#qwL%ornoqzwXq>t5YDic{?Zk<)LCFjar&1r|xK5KrGU422sA@1C2HUmQJ2m3`|r zls*GUoUv!CSRk)?sPJ55Lmg2x&1%hDrdtnxMI8r(x$JtKAv=nti-Qo6_sl}A0Gc$u zv3LmfPHzp_$;@rh#*k~%!Hobrxs;M1JK`d5vzevv&1#7&lS-s8r+GG5LbQC~2+or8 z)Ql@5E|Y>3muWwf=2|7jTO89}{c9?1Np6{@xF$Vwy*+yx1XP$nD1(4Iwu3J_3JIm3(xYl1V;V})UaETPZI0px}>H1YaW|7aaZ zqy_nKRfAHw@EN)uRS#02s`SGC6YQw%&yXZYx*NmW<`*Sf5(TDcm?x}qJos-TSaGSB zYXT5Oo6$DNQ$kd-KM3-szpG=<7wCw44iH3)oI2LTb|!*9$`zStYEC@*8RgrLAVJ!% zpXuQBqvZWhh!&Y;(4(%A_|mM`sH#yokVMI&8z+6h;lrxyg>xSf=i_Hgdu_+@X zL%R>D^>3?Zh=afY-lf``#S7njwT~n{{Vxxi9C%FD`~Y|0GcW^8IsI3onHL_#!{W9E^LLyvH7DF1xzWpu0+hhv z!01G`q+VPFMZPR%vX51<`g@=5IMQXs2%3h+tRLgnJy25U>+Y(IKBWjIh1x3yb~2cS z!{w$g6Nqto0wd;*s8)k=OK*;eU^~TRN%xIB=l@c0a{V(6ex{Rr%Z?-WK7M5)%*hVhK!K#td=}QR4Y<@Qg&c zd@CVqv*he|Y=xN&Mun$rt`1naKgWx%>`g68xv#T_6`A|(vhSR`yugSQ)g$;d$=SmS zh4+SW-2j&*?Di6hu5xENuQ51I*2fw|+L`Ix+#78M$Bp4V@HyLRbxg=Krnv=Dvs?jL6TKc#X#-UmDr;$V z97sRk_LvfN<7>k#H(pIVGRegHB5%rTr=Tp&9WG-WTp7dcsu4G{Lt*Q?u;@KvQgY=%av8_~XlDg8cHym5nXk^CvH_zn8 zi~TRF^)3B`+WHnj7^cd|a;co6G5{Ad-c<=hQhu22Bx@_>E0U0jZqYhr2<#M?P$bvs znS?07#(0I7-PRFl3r%t!D|?&aUN=o(Tn-dl8Ba(6Y#hjkG2ySTq6p}DAM{s+5bB)#*8P;0|1=q^N(`G;tj$7;U7N0 z3VkkXa|=BCwGXT}UPY9X@e&mBHdjDD+{v}mv>rBfxDndZ+9e|U=TC!b|A1q+kX-oV z9|7@{iV*iVRR#-I0yqqXBH92Cit3_ClEAM@Nyr4rNi0%Jedh|?e_fozP#VUA@AE5DRq)ATsr(hpmgQK4dauv%9Lpuj?LB#7UW2OU1f_oQTg(F z9dfR3_NMgn3@Z3c&3Q&uz0a##GcXO(k8y{rRiOfrG$it&E5OFBO}Xm;QgEB=3w=s9 z2^@an2v3(LKbw{GY&&v|krFaO9GA^pM0+KhT<%+f?#R=$I+E>5E}U7F|GGO5@Uxg3 z@rtPP{z$57va4jlszQ3%=OB=vzQrktWzyL!xUOo~<&S4>yV4@$MBRXjMP!~KjSngq z)(teI$C*Jb|4SfiY13b#-AlF#lu28U%RjU9va?Ff|wC)0u8uMeUVQn-_GhH&#T_v++UvpH`3&)ETAf_ zPa(OCB7IUAEnHfeVSQrtz5p^g9v>*GOC8l+eOS^L7gp8}^zc3QcamvlAt$yucMaz-3$Z|Vak=$TC#wnj{3nJjF#w9x8dYWG;N z@Y?P^`bzv%B7H<6b!iE1jJ#~k6rzJf~dLUS5?Y< zH#>ne^^Q?~VzFNpx-GjFmV`11-`P;pp6KraaX#eGYf>psE5D#GY8H27)cl)}G;Wh( zj4$np0lC?&1I2y<=;7>ISZLbht!n2|z!?tq+eVm;5dNFzk#2se1z0C$2}w@-1Q07u z(cTV_9~-GkbE3pi)!_=Kdt;1Yn(nJ0xW8xskOG5`DT%VMJX$#8BG4O`sDZ;>hw{Ms0|H124 z3qqMFDv3Bla-{^zGFEUoP^5Qec})f5*=yaHj$aN#GhcoK5yZ9MibfsrswJK^G5d|7k>fdFHMvhkb{*Z(+X^hY_XpbhpCCHIsd0%ywD8 zyrmXk;NW@0nSP3^o+ZYYWkD>Nb7F+JW#t$=OJVZXzYD~l01gVG$}b=ftWh-Ls84W^ z2sLp;am9>_(=*qLb3Y7Gc9VV_+8-lOghGX!ED+LR{oU7lFI$`TB;srNiV-b{a(!|S z&UbHuXK~X1sOF@LTR^y@KC9I z1-92XMMbtDRVH7@@W?%JwGn({%njZ<(%gO1>K-E@E37oD{7hIjMEsb7gQO7Uan*`1 zIfo+;zX8h~QdEfT`OmE6FlL2ehjf2MV(O?B(y-cC^)}v%6C6#RiytMIKPq;?7yMV? z82ejxaIiDTv9R}zw&eeg8TuEYmwilG=5+I$*%*3`C^WXFxO`(rm`dH|um~m*3ix1Yo}M9DzIYn>G_a$OUi^SX7e6GRKYH%n5BPns9dqZxxn`{)zaT{Q2p zU!7xFy-cs0W457Cwm-%QFn=&O9i+h}CnUA!x$eAu%RYGd@mT@%n9D9UE!N)Ox9+y! zwApxh06uP||otU$39BwDo-eK`A5=kHMxj(7JJF9>ee_9|dMMog2wsqjoGN zOp>b!n?hYhh< z3T-Pni)hIDC+L*~j|hxGfzH{L>Deq=k@Xs(dn)}pw|Lr(Q@$M}C@kT2rnvs3chb3O4SI53rD&WJ%2+Wvulg%2JnJi?o{ z`Ysfze_7lv3lspC_!K4r<79FG(Y$s$h`L-+BuJNb3&pJrif7Tn)j6kJV6|D)Bo>Pw z)17mW^-ZhyeJ^x$C_`X`981!S5pqsOz!SOG6))L?UVWbO7$ok{ukX;L;YgNug~W24 zi#@Fn`-fp0+|{_QxGQg7|Dgs%l54-w67{v<94UEQP06)|@-11kG3zxZWI_=5+JD8w zK_yzT8WU&MHz8hkMnO1Yk{1tHPiA|Xk;1+3$*{F!+t~oJJWMg*MZ`ENVJ{!enA~HZ zZ06+LP8p7np%MzJdIzJx`FyX?XU#SUu(#a3RewzMeE74A4{q(-{@0`E<;Mq)o+*!h zcgx`AVfJh9Lkj=UT8Q3%?Z?Lv+Nw<pd_-OK}I~Pp>~lNKUrX}Jp8|g$pfUqya@|S;n_NmN^g6da@*R-fnM`a!+o-r z62On&l%MddE;8#~d53Es7lxPuRkNJ!S-9=&HXysE*vDG(@35kIck{!Qf2Z8iFFoVFZ;A`uy-yIgpDp3yIVtY&($P364 z4{^clK=&MKhGNjHs_MmI-MRGSfy*PH6%`s52<;8di9!R z6Kb6?F*CD|e*?7Q72>+-;-dExHC3w76Ka^ZH>8#uC3rn7)`pH$fo=5pUWYEF9ZT~M zS{g)y$riL^U+HhPTAQyf_dVCOr1zlkAaK)6#A?W_E5UNFJ=u9SDSzShmt`|v3Wo3E z^5S?hV{-9DDySebkw$JNZNG2+z=$-Vt6oHJVw-VP)m}kPxVKPzJjiWDiikbU=+h~AhOWF^mlumJk8QA(lQ$gS817xA zR6KU2%>!7{`;(zYI(WYBwkl80>yfw{%AVL2kOe4}^jLm4Oyw;Jq|Pr!8C0s5eJDt}gZroY8A9$x;9{YXy73K+O zHdf^!Wyk?iO=3IE*UQpE-QP5vM#mrQb)Fb1FH$y#&Ny{dl&xUE zi6zavHkGW>X_5fxOJzO->&_#B)D8EgM09aBAys#x!uUy5R@PE1EN2~r6dnvTlBgl*j`sD+ z`k4mR!Ka7jDFL}{pxWpM)Nf}Ig~xNd5_^=}ur#5kUPLBKgg>(R#M|<jvzc@V86p-Y z>Gs|Cq{^bpboBdf+U!^v%YggS<*inoV2SzpN0~h;(_ANt1j*=3)k|SO_Qu+I%tIqO ziv4K)q461tlhr@6?&$WiJq%LK5!&BD(sD6_PY;iHjEjy#K1h$?4oXMS8qmqD0jeOK z@*U8*OWI)Hd9PM;SeEu9|M`ekjOw^*2VqO5crI*jv_r%W2q#M3=I!g z&AHC-{%!_wLU#7U-y7htJ-trNVD8hhk)wNaPZq7!bQ%U@UI9|JtHklqvf4d_KIzNG zfDbU%OD9J5VtPdQ6>io_%nidMBIU~A+4Qi!mkcWphUbIOlv8Ml+jh@kOQ%fnkcByt zU;f8-c$UW`<|tHQh&vK}&=2%>Co}i^da)NP0~DnKGv5Gs#jsOe>Bk)o9~Nz8eW?9ZGs8ZQe!RXJHKU7! z)#d3WA8%i@+u#IU{!`C(F#GN@tF9QRO?# z<0BdsSE@YfI1LSMzVt_Pps(inw3X)jZ*CBZ-tSA6W9g~at81N0&JPQZ@coEt(0PpH ze@_J@fi+d-nJjSr!*7XahvMZ%r6s>23q=aGg5=Xy=C<8ly3D^;bSM6?iZUfrezO?M zJ5Z5HaSUq%2{W%^6zi{6?~^MVO0R6^%Xbr0aB2{i9h7mswi~@zF`Z8za6VNR?c2NEN>cPZ=rJ8=1wjLAci+hd>n+zQ8A3*C3Xw0_gh-WL+3rbKmF?Ii zB#u+*zDX7CrQgv+lOH%^SPjzgGzs~gc{`beCjyc1)K8D;kgM?CQq~VD>FOBtcztmG zz#OAA*TOO{x&= zk6vqz(?yxk5m}vNv%ZmxuW^Z+pseYkQm1G?YPZ|L33tE~SQqntB5-{^Tab7p+G65B zU*IUBJgrc|R9%X=%HK=n6GRVFoMA!P<4;*62Dw*ly=v@IyeTnnw(123XZf%#E&ui3 zPDOuwrHePpNoW>Ar7Iv3#qe?8y)oDoE&yh&LmhgRY0JZ>yFITwfDQ0^zW`K;!L*Uc zBb;`hs_DVjKj)VT$ci*rd-2 zH>j=00T=#aw$%`EagG)sX!gU^d~@t;T5Z`YlPg?x(|d&>a?okyL9=or}DirNwaNz9B7xp4?F?irR+B#+>YI zidl!$S@%)fghQ72@Iih|4f)q^7&7jcu6#I4LGG6->Y%Rj^W-spST}-2VRz)v*scUp zO}`!f_qG1y^k!cDs7IG`s$&MU7JQ5H4SeXF5iGj zhZaaq++u3;<%3AkgLYmv`DT3cVh6`2>tjLEjb^ZHezFn7E1538*kh#NwoTBaV!<1= z|4~i_mvyrm8I$({piJbHJR|V%je%4*FM@EqZ(I*yn3W$(>hoC$R52W3O=J)W1HP#Y za1U9HdHLw&3`>k~#nEFPefP=L>ul;yEC|`;0ZF1&7VW0aeuo}eGKfP|Ub$mz1F9%AO46p4h&3n#?+FURg(g!@@tB2;~+| zAd9((y>uT($4c8bDA82$ITtZDW+doboII;MtqxQAe5kQljoQd#B~D6#s|Cf+SdN1l zrE504`UnaoXIT$Qo3DIV{vc~}xxP|;q4%ChufudgKDCk8=10~S(;RLZa)M|g`{!=` z#|Ex=m$(MxB&SYFFX>ySxzBZq|0eeHt@1s89Dm@)A$sw=?@_DfYumYEGp;5-ecDR)r+dz#OGv zU0DYU59@KdoV|efFG5kX*DN2nt$2POuGiqFXIAX$o#Hiwk=PQ|N{7Cp-5?79=W|o; z^)5#|!~7Ycx)+Gdb#S`dp}OOjk{o+^q7t8OK={PO+mt=5^SbFBfY$Ih1|kGU!H)@N zJV(ntOTCyOIhi~AL@4Hpe$*!LuYSb0Y&`h4DjewH(sB9@zr~22YoKK#Biw<2WX6mrlDROs{9)nLOae#313cm5L|9_oy@nAUv^l5mtP(~bxOh(^D2(PVUev~ zzj=`HJ#io`^G)8{tu{bT7TyCs5Et`-Ew;D`@~8S-uC+hW^fuu9`T~OKGtpb6bYa zwBZ&D^XG)t1ZQ+@L9dwclvE@&LM?a2@XGn-t$dKc1ZJIG7&DYgvxLO;>vakWRG(u8 zVUv*H*K1{*A{az$h*SvZam1WfbpOydjGnv(Kfp@3phM2HkGcB!h~h}I*9a{&ubC)5x?ZnRXA+*c@Y>g!J>PZLDyMky50v!9{}_fZ<5=?KE{pLGHOT*? z&9Gqa@`&K5^dxOi?#5zmUpugwELTYTHAdPBa<&X=3ka;SjMc|Xrj*Uu)n!V2ITc+Z za1V*4Iq-dJx>vHCmvf&p0+GlvA-z2dgwCpMB7ys}dI@K}=e0S;wB@f5ZF&FsWP3Bt z54z!CYd6bPn+)mfOBF`<5vPOOm+~+&zl8XjHMFv)kvVdPYts6VAkMmStFO@A9()jV z+6*d(SiOX^^q_2X$%~>tQ0-g4@!}XAU-G}(i$Z2fSep0pxC*g!D|2uf^d@9&Jc(Y& zsCE7@msu*c`exMqw6Q{6Td+l$$)#$~z!KJ9AM`rWm6hZZqmwBVm~1K#KFP{oU=0j@Agh!e{XQpaZ6Of^3W~W~{EGMf;8l@=XmiGIybe5KgncC~OG+=K z1UeTl`oT*O(kq|xp5q!l2nD!7A|UzTdW9_uM0sO_pc2)*l!k6YPQa;1w(el#4r)qq z>x{f2j|#Vrid)HP-(x@ax{-uMdc=J-F^^$kr}958+2&E*?%rkwH2` z4J~!Jjdc2?{fg$+Y%@F?$B4Ev9Vj8BI6?pQmo@y67z+*{nm1_Oc{?bhW0q@cVAD@M zbp|^YZ~h~-7j^wF#Gzf!C#U*htMp=aicTeT^Od^udb(S~w~Glscn?MFTrNUob z&8(9%g~rIQ)~#37+^`#u{YQPsgQQ}Krqa3Ar~(6ET_ONzSLBvwt-sL})qb-k^@zvl zKn$Q)4=YIFSw3w>ihikMnVPFp3CP8}?0K%c%ij&&fiO(wrTXfH=PnxhtjY zjoF}_#({T?s~|<+89)lHpCcwL&~S=%T5O{Je+IDuAW`6xQ{3VQc&8?dwD}dTcgZD; zE-u#O71GAlQ}9}x>Tq%2n4vIvr5Bv*$)wY{h(5<%aviBg+QI1oE+aOtFPm4YM;`8Q z?x+OomY$HhP&AD4`V+AP)MP|N)^_{s{vu+zWTB6ak*weBnqwb?z4X*2e`VT7WDR^b zzLOK(VFk(>5B1!bR9+ewZiQ9Zcx_DYge*0aG9ViyrR#HO^9Zz!r)U!mr=Etlb(Mm- z&Q@J@aLd}+&i#M)HqxvqA*GB<;q-*Lp3EtJsE$k zDCKn56F^GE(>k`PzlI7yOrD< z=a-~-48AY(r?@9h9ltOM1`QGabuS9?X;meOM4DOJd`z>mB`u`E&-5(!l8_JMKUdB0 zT#OL$(2pG7kU%7A0+y)nGCP}p(#O;LULd{b=uN1vF-7{+-gN+Eq@v zTgb!e4M!j%Tgx zZirrjaB!%C#Xf>`?i+&iv)KTjq*NRp1edO`5}{{Ni~`YP6T^}lPglGCXWNF?^V9xRxx5-edy~FG#tx_MJa~Qn%z2g~ zNF79ibi@VErcV4*cF}phYFUzaw5=YTwFNtz?3pDrGG?jAH`GJ9NZS`niVX&HZAg9Z zyi{K{*MAo_Lpsb&6^u;K6tbRaLlKR5oSr$*vrxrjey%#xq6b)4cp>HqIECfhqHOiI zwlJ7HDVccqSQ_)t5N+Zb?jz-41}7UcZ>~X=xGxy~6tZ?a;Q=9=HlrhN!8|XPwp1y0 zuQ^9m{_xnVI+II`l5Q-r2o=>fq!OFtAv0JqSE$#W*J^ILDyk^d+s*)yR;f%7KUgM+ zeXD$07X!*@-zy%u2R;|@_8<50Wf2b&$r`4x$1|A`fbSsDsff%Z@JX%ghX7<1>91Yf zMy=sNP7sTM!YQY|r~2h_!#Jrxm8kZ>+p>NV*&a5@INJkLP6q4-wfoV^{qP(lTQf-ej!4GEf1TgjZFm+om|W~Xlj#sxqw z^V-;f^xsc&`DAY~t5T4Aw4qBzggAOu5W={2o@rXGC;VVZuf1kDo5XLScWFFk5RcMUZ9KG?S9! z)7B^U)mHYGwHHxM3VuEIid5E2?Vc!AKJoWorXOVXNdI(V6Y;6>3q+#8H2Ipa1-OdMVM$2(I1nJ)!&ETQ;M=YbBE_z!vEdc`8&J7@3-v!Pw$h& z07pR&6+`V%h;{%4hki>6C8kCXf-ajyexPdRyArOHMRvLVGB(``0w3s>*7%h-kr<) zu*^$~*|2Tq?A0PS1yw=I?S`1@`6W|z9rQZ|d@`?tWa(b%MV}+fb1Jv#Xe+M8iPM&1)_EqC)edZj_AIj%Qm=(9xi)A)P7?7 z9n&nAbKr0L`NjjM=lK*5`|h2yt{**YNs_$<_Ag~>9z(b3L7;Tm;!bg5kWXUvlbD?FjEspHJ%iVl-Pbn3x0poa$DvRkY9?@-YL> z_nlT^eZ}g1zw;#}b{*kyWmn-MvteHMMBfz*^&)ILbUN3}Pgzi=p9X$T_>XRp7sO7Ob51Qo9)KM_Qzlp zE=;N91PD)6VpY8lpiBa-5PMZL6tuqI& z8uBufI&4%qcor0iiWgeb_m_U^-vjT^#^9qn*~9kxfen9G|3zN8PsQ}|BLw%8Pr%58 z%G!dqbOE0Ldc)_dytai49P8hhGWqjHRNaA;^+KP}qy2+Lv`V$&Ar26NRABJv5~n}jj?yt}GYt*z5%S=YIJUxfyT z=Fy#7Mgl=7>09&Efo>fD9=oo4o{JlC&O@#v3GT2PzT*RC2}~vkEFRyBJl`GD!-7sk zt#Pt1uIq>%ez0#GzCXLjr&bVDa$uyf?;^L71xXeR0+?@V7pQ+|OUz_85X_dG@I z;4Q3&sg(P*Lh}v2(0b@QOD9%DSWG#hrYjKwX6bAQmDKUkMfe0Ozc7mE-BV;zWcM93 zvCz@jj-F4zjrf~~lDk)2<9y4Ct)XLlRxj)?F1HUt+c*l%zb)Wl?%oq!zJ4LuS>hHm zv7TXpi;`R6-VSGiPVix`d*1_sqG~=KSzeX0xX8S}__Wwpob?prz)|R|dG{&)b6ZCb zZ+LeIImtt=Psq$wx?iyi%8e1ok|5<33p&j&-Jrf2N$)~X?X?@&O)=%1Qpd$FWj8TtM7B#I7IE&gcn{T1hi zixH04T`J_Z8rbPlM&%z&FqYDo_e`Yk-@DGa03HMHd-(aGyav#&eeFEM)0-`3ug*?A zCv~BxJ5MZqNN^G6oj2Ev@d*+h!|U|PvZuUu5?@_2)Oj5Syy>zZRo-B)ls|P)LDFF zOCtH-kw=hIAfb%hA92V^>Pn|Y`YC!S@K74N7BsZxYgxC}cnveKM4$$;Oo52_mTpOI zK0wS2hH&!Ia`WY-o}i!L_@;JF4Sao~ks(Afmyqh?$|D^*FPj&SO)c21n^bu>YSqA}G>1$lz9jHZ3H z-DWD)Fjf8#JJOvSf#W}IU&(HR(LN@K_#@BHFugXWMtc5P1_Veq^c8bep%YVAuX53u zpXS@<)xWp&4MM*BB8st^T}pe#^0X+%+G_BXFB6ktaHQ|7{2O$?LJF-pFBjm)&HBWT z`IZ%8FeLKvS>6M2RLDF!AWyTL=Tsa8JN?6Kv8^))JD*;w-*P%k1-(joKkQjqSY`rI z)yrYCKUW9MdY7)^pW~ZPY*F)Xn;{G;dPMs+^oT=Wb6LGyOPFgP0j4mDKw_THQCfq2A{4M$V~ zIG+tBO98#lYM`wN_bA_+r-r3_{0eh(5CoKFy8P~x!ewE}(xY;t%Pf@S8$D`vH8QSB z*wp5v`37I>+Q5$712N=r zftv3Pt8L8az3<9QE(VNC!E^GaT(M&qY>EQDG9<|GBV+dQ68adPgzmO)nQk3wj`0B; z$@JoT(+pNMk=^;1GZZnuCdnYvCP8D0Bw!R$MKjL!{(y!zO2jJjXg~6!+Z5U z>8<%9jwz3EdvCM+G0W|79CxX)&@!De4*7s{uJrbF+eh$r7<^mf)TyG!hCiU{2Ta$Y z22g`o4H4`#mis91l%)G4dC(tyJz#n((ewcT<&|iEsN7%!P`JqETPz)6xjjCfUG1Cy zi3fM}{Y|GgeJf@=S6pqkNG7vBQo-T}h9FuKF(+=(n1M5h_AN|bb19#q21HeP_>Pt0U_PRQk!KkAw#t)1XEs!O*Shif^PI-6SC8eYX=l0A&2Lj$ zT5cTIlV?=D1VN-VC~7C3eK7|_UsUOVCMy6)T&@T%&$N1AYSi2(A`@u1idwxDDvQtM zRZ`Css5+1}vpDRG;K7qKsl0ksr#A(fTasH^_7jQbB|Ej?yuHq<I(h;oui@OFs0oYkc@@M-)q&8IKa?4Gw;z$YKa&wLXbNmANsQdAEV)$dyHa=q?f?!& zFwD|D2E4^0+HQ&JRdd^x{4|iyW+&VxMKnQ<#Agohjdp6AZr5TP#%ck2hQn=|R3uH;hEsrleiU$sIwhV3xrl}gTyL!kife#pFkSKTQ zW6sty@$vj3MW%3Jx|RgurR}LqUn!==VU%^78jGS_^`~U;r|NdzD@^5?Zw=an9*a5J zL16W|jy8V@mVBp6(>z7jg_rwP8HTCdmfkL@6$Ja{!FfSedV;)kWS->;{VqzUPOHQG zn+02~r9neD&p9s(r1Fah;8lcBMPy1fUywJShjeS2F&~$do&Odt$;zq#F4NkOLycpz z(QRW&Wqm>c>k=Jmh~o3xame>Jn}YP#V|kWR_N4*Wy&**fS{mh{nMkA=BVYH)r;c=)#43uStR;gw<|bf3~ES0fp;7uJJ;jX z$1_>9aq|u`wt?ryn?KQS{`9V-(XewtZ+O$qYeRI4Ft*9YWggY06?eB}ST)D-q(vEf z$Wf2EOi;SVt}ZXlXPCUYB|TNXf%NVqTMxzM(=pbS__Az9Z-lylTYE-unI>C8euOJr z;EN+gw1uL{uQOm7-Irg8ZeT6_2NP|`h}CV;&$A1Voh%lgC|g;^PQBX17+H&^ddUmZ zbE>NNmX?;lKE-7kNnF8c;PsshEQP3Bj71CUW&YLsDJ1VyQBUunN0_y8j7IXMoP*k_ z2D_4U(gNQn=~V9xlKUg3abzElrPxj~ZFZlBt~L?ZU;mPE zwK3XoBAz$WWi){{z~;F#Njt@N6EAiTFSh@m_VmjYXB7x0Vm_8p?vniclD3@^r0Gak z9v=6ZM)XXhN;Oz%K5rgOy9oFI!GEv%nFh84Bt!2|%J_;bZC1UvBGHF7mFcn6GpSZd z(|v$r{iB?BKy3k1`b1w(QAE|nfm$ms7w}mgkEz>ymE(DRyh7T^K2Xn;V;XrEr={$= zcS5qIZ(nkhpI<;gfS6AT^m8gE@t~`ASyukAi95ael8Jgpp`*31ey*@~vM#$*H(Et* z{zI@UxgqlDQ(vtlZG56u5;4h`e(lsLZkLhgl)1bzd4e>_UpLe2w0}Urcm^+~r-Dt9 z8wjr7U~Oo13ALB2Vjyx+dot8mr9m(Iss=QN=%&+x_@>yiroF|^uE}dr8H48<3c}}p zyrvyAUxSuv1~$^G$I2<9VY2N2@tM0mNXYcDFC_*=wFGwYaaA^JcI<1&vl!c7itW3( zRMcl@8P77Y#}Cc?lqFk~A;fe~dHK>%OTJ(&!AULk_EiB=y*t4deXtn9*5Kgaj4OCw z-Ha^&E|GVFQ^$q~lTkNo9@ zG19_VT992Hmtd6JjR@zVC+JSAc%{6qdIIZ0gW-4$JA=R(h}1)oz8eaPQYAwiDYQ5D zaI`m>5F<03r6)*_O&!_VlLt>0d=%IOWmPg5J{5Fuz#d~QE%aDQUe$XYV;-j-M;++P zvop7VMs&MVd(hM#r)o>Z#SZD=2nQm4($kaMgEr7%H5F~!TP#yL6Dx|}BGvsn(87Ac ze2&LlRutZ`&5*Or(8-Y7q3$rHlv*N6P*@}bSn05GO-V>d$R7;RCt4dXUP?A7$edeo zQaLr_Qo{Jh5PiFQ?sFZ21exIo>B-`0BACcJHR}in{D5#|iO-$c%qimEWmj|v7U223 zx+ZGUlUvN)rUyRi7T7()eUhVJuH09uibObwKm*>$F1;ffkJp@9hH5el8Q)n5t@e+{}0`U%+ zBJ#)hs6gw~0)f?Pvu~cJmpFZ~rEG*dlf)PA9$Bdk+0msRUyEarb$f;z^5Z$Q6iGO<3I@7@@xAE$B zsXjM{2nT`;|=<|rwx^q89Lcs8xhQem6lxUbfiS^)iAo40CDk2>|aJ3-Sz>7|xR*%f{iNQ>1= zcbOe!2=UsHVl_Ob-+#>y&NwYnvMk}m>3$+3g&G2-P~{++I+yG?){5VxNPCsve{1uV z@y?tMnXJ^rL>j@Vbekp4QQN@ry%w;ZnZfO~T3Q@!#-8pB4eN1i!_Umz#55&$Wm{>w z(JJ@?rwOv66GfAyd#r#|?VFaF_$R$LuAVP*7B!kpq8>A-zsOzF)nJ!*Vw?BxXHTB1 z&pd}lc|VFcPyOMCSy6PGZm?7Z!+*7PcXNZ;WFw|mZ1q!I8el6JPJeJI|N3P& zuxHsHnxkH~>H2Pmhps@-iD9OJ_k>dO`JaY|J2ZUc*u8r<@kOwjFW?s9MxON9Zih1# z6Ws+27(Eal@kd(8+@%Q_`3o7kb@D9k$5%P%%~E)LC*SSwzWr_054&eCxXRJaxK6N! zneMAP?26Df)_m;x18PQrqhJ&2elGxF5@d}GoChB_wzNEQRXcsis8~8sey%;A=qnVR zlXH~HEy-U7j-VkzsxCL%qFE6C=%S^iW%u=pE$H^k-gwaCU{|%E_Z=O(q|IcH`j@@YZc(^r1pYv>p@LTyc0L>VP4 zNt-QlDi?B7yIB)1d;EsxmWEr}E)E%?P$lY06F;C{%#OqaRSkIa@dnAy=}sKzxCvGU z!dM79VxB+$lgjN+aXgAuo`Yxa>ilD(Uxvamk7u<>y&~U8`iT z>2`7Sxn~}wLaC)M*PG}BdFm}653NDxdawj$j3pohc`lJ?oo%H%0NZ#YA1hpkla&gT zCi;XOwbq_EFmqwcmMsv7;(469AGn^;yJ~IicUH7S0y2_UuU)H=k5|Die2gc784s%F z^OU1L%A8)BZm4Eg;U1v5u=(PzOu){{_}rPP(%(ikef$)ArKtW+juj5Jlx{Ab0?RL5 z1wW6!J(}k<82P(<(hX6RS8qH&6@@A^1`BWCM!{2XlWu~?WMv&14hTWYo#sg&qW}Ss zqi%P3awTJ@3ZMH;$drTLI5fypE78?FoI*OOb@H*vj1T)kFy<~(u0+!|6nh;BwJ`Q1=IAZuu7h<>{-wDQB-;Fy$@L(~Y~0nRd-61V3PCF9!r zM9|^|cX#*oXZjyEhqBVcFV!bFCdqgX9J<0`;!l<(DX)r6C=sIm`!b=X1ch>m@&)I4 zj40IC*Q1H|1gil53wJ#ebba={PIYXk(Nou-bmimG#VzFSF~mN1yGbkAp%hBoebd zGAWc%ecT$6kZ}57^t*K`JTV$R`?9Th2=Y3ddq zd2Z1V$zvITR?g1J8H{W)6*exFQ(9 z`RU=|$P1D?L=1L<#$jFf(%JG&n3<8pcGsd$#gAPaTr20!=uEKYFwXx`Z(9Hk z2)#JS01Ia${HF4stnBTB#0k=SLCz)PF z6CHitr&gflRMe4@@CEWd9(48691PQ*{q?U*b5Q>>i7Y?8?8M!&d};CWiytB>sRD|fe+Y$tQhdi6SiUeiI_euY z*Db`qpv9a~2!7qf+6)|2yPb!DsLcGzX?RmZZsy&|0?YG5kaM^)qrS3ztSzkp2%>(v z$%&$cQ@yj~Zf)>K0-fXRF$Jg2e$-Me?YvsPZj#Ox&; z8%qv5Y^wYlm!vS2yXL&LwKXIxA&2xsTx+6!+bOmO1DH`e7?A1zSZ5NfWTP}sza!YRQcF~Y@AtAIVcNQ@;n09;GbU}?!gmOs|*8OhU&7~m;xnHK- zg^`(DhGBl^d}Md**YoXdpZEKH-sil|>%7kSaH>`L^}CwrR6cm<>3X-xMl;v4 zU%yyk1!+)ibMsxwL<|8J%hns0ej$A37DboT?q0Rity;MWcHek%cwWhilQ;AtfaxMEl*L#unFda&p=^d>hD{5Ez&0 zt9GpOoBqhql)f=NJ-wt$ePA}8B3F8DtwYxevQlp}JIH?*U2_%x7{%B3aa)N%jDD@I zoZD~lAt&OKTg}_O>cLe6H}^doO=??O?s--W)W&tMjhQ+y4AjCd_nsF6Ogewr>@4mn zgdvUTfw-|qCQk~;fcxhMmJO=`mgt6^35Q|eUwokOV(x-Z) zwKsK_2i}04IPma8r}p=Q!^1Zr7MHc>dvufOGv+TNlx0m#O}~+!n|Y)kxd9+31=DI1 za=eP|tNVFVT{k>xpwjo*=gUg}^Y67*qs5a!z|v?Q3fi(MG&EG?etlitHwqBC$zkkxz zCtTU{ZyA0Sc%)KI9pgiMht;P9EcC&oy~}8&CJ22To)CexBF*@c*yDr*$+=$3IZ1#= z;R)2_l$`6s%Row_KG7f;a6y*H zDY`Z!$Th@iAmP^b+m`%xwW82g_Zw%Ha;(pvph-BBle*Ja$8FvzclKkWe0fc$)jp^4h z8`YW{3)b=U^vv}_7qCjL=T_cWxd>s`|6VI~!Euh#(A>Nc*etIL7w#|;k97;D>mQ<(w^V?@r!?xNYs_j4TCwK*bfCr6?jA*I! z{$E$f2Y_Fa>&RdCUx3oL_w=VXbVbg<7ImAOPdb?u{a!@wG!wv7mjcaBni;cD1(zl> z`7faXKN%QhcJ{dj`^C0h&aGKAFTlU0_iUp#(RsZJM4JC`uwL4y|Jg5}clU%C1VOMY zPfCp5+|ba#Q0M2akk9)ToVO9%qEZhq;(g#HkfNbY1c|81yU=GR@;=t-Sx8{Y8b8Gi%b3SfnW=4CZJvp3NQ?A@}<|MoGjZ1!SSl(nOMX%G$#AFj9Ngj(#zk!KS zPKQ?;Z6bK%uB}oA(OZywL0=;|5)^Qn?(c`jy4}b1==k_}zVu%)0f+gZDd~l#(4|Qi;D6l!p#sJj zl(*}2rO$Rjz8e%`pr!UCIa#l9hUwtoU{x1Sg1d6_adNV({UoMVb$SXr?WKEdEAQ+W zgG~#2>iX=`X9R0d1Jut(HFq1vMSgp8y0!K6$3Zs$YT>**K60WfHpp$ax7CG233J5Q zo?`Lo*H^p169nH-DA|+KY;}V+$Fo-bx%%>FmVV?SL6Fg z!f({*+_`fI1rFWZxCSq}lCpAW{AAY+g&-u_(LU;QxOX*t+_SQ>?!-I7Ww?=?S|EcS zo%XTVxR6!gNQ8el|BB(3OuafTAn8)b_trp3!<#gH!l&o9P8Wtd#x$5&R~vKA>&X2* z={8X!4w23|*LIwIgo#OaqV}vkX|Hjbg{U?-`7XFp@+3=nxDB-pp?&+?w*~FlDo^LJ zC0Z6XL*Vm+zi)b~49hoZ=iG;iSM4X~bWc)BRmEJc4XsYS_TwLyPYL|&05_t|L&4ij z--rcV(hMe8&zMb+{v-Ok;OQwjx_QsfXL%63uhsbWDKZK>)2^+RF3+;Bw*uZ&p2Ym~ z%8KqskuEYM;a=;$m-oirUS1s*KXhyVs%zr>sfu)lROFp_*ZnI21t8Zq)g??no$-o4 z70eHS7#03BtuQ+~yG%0t%!!PWow@D1fg5{tcaNJDV=ptUva&Mo!=EdaY)$nxC9dtyRV7r435^Iju8I*4XL8Ocru!c}r~`a${4O>1-8-SwQ{z$FI@r{5=I ztwnNhD?KbMEN?iYO1Dngt*dkjN&un&MkEI*^d}Hq{#>DOlN2Hq8t;n8es;VmJYM!n zq)Wk$4LeN@X{ivMwmv!<)w<21kl#*{j#|f`aOIz(r8YJ;Vm0M7E%gip(~RMQf&!bR z4?3{9Xlsd9>Ry1*%G}$H-kJFtVpEG&dO(lD`eKcr%r@8+a1EcFAU!2Leb*cSctsxr zeSQ5i9@Wa}PzaD1eJ{WNx38`t$CCqg++&U)L40;Vhe6rCTD9ZC$)br?M+Rlpp0}3> zY7y4TC@iSL;9)t(4%e&mghf_>yVR>|hypgu8oZ_CVC#i?eS+KvCl?-{R_Hh4Xn6x^ z2FbzcUIpqVAcbW~#;d@(N$#?;ves`NA17r9;H-tCeS3L2QFFRf{&E%PCn?#;!mbP* zGxwT>--XXhtDh1Ve!!R@#*QK@^C4G1Qe`J3F!zdV**%Q0+rhpPJ_h7ANLIVB;PA)g z<K0;A$#Tm`Jr0!;fjL9s6Ege;=H%Z z8I*#y8tNQ}R^$TtYDKXB$GDBk4uou$CP4>L*>=aWd{&<2mZ0^+} zIXylUE+HcD>QG)Z;cY`&%>8}%Wl`pyfSJ2D89gPivGP7OaF8NE z$l9@^4G|kdvgm1OY%D!u@9Z4tm{zP}znrTNX+83n@uHWPt-Q{kPUByXh+r3+{K@n{ zc3fJU%jnTW4-G`ICTx~H5LH0<`H4&u#|J)0R|lq7dxfDly7Sw^I=!B}Q#_^a(Ra zBeHUGq!_6wDf%?Q*{}->?C9OrW?a524)Szw<9_Wo-&Bu4-nj6)@cC@!)T;inraPkv zp9{T?GeVbCI0L-_*5&WBK| zE(oA9yc0Ut8Y+<_%86va$gX}vX6X>1r2gIw9dZFiEaLEqa`C(WtffufsYGz}Bsuju zc%gXBz`#JP21$>+0CGU8Diz{DnBW|!g{SZ61uPqwGA#^i(@868SkH{C2cBo)x8bwM zwR(ggGQ#>jVR`k$^3dvY^idmn;$o6bK`FtWwh9E48N@S^|@$_@&f-^q%Dd$CuT92gj` zT$F{VoIXjJoPEV|&zX^2>CveH z9qr9f5I@b1KC^7qmW|TU-rzwfqs6K|Iy<2JqCjE=sk1hIzU-^Rv?Z-C<=FPpgENzX zpAEX`&-|h>dC2yNU&+oz#C7vo;;cTV5DaA^I*%#G)ml?$iy)|x)R7r^{CFUWz6_Ho z%bC_Dninn?l0=%ebQ&^-9T@|Oe}9m#zOSO0D3$nsOA zkT8Z&$_uAojhe;%y@ZHkZ(bp%`&sg8%0RFrVUi{?kU zSMEq3$!U|Sa<#J~6aXe#3Cchyc)z@2Z`g+29!-VC@1t=F=2}5ikG_jb0Y1W zoQ|}Pt>%ANP2hl$?aOCpCK2#T*;hab>}z1UbX$+h{GV{-aIcs5uYC1*Zk-} zdFnXwzVYQReqZkT>24do6>gu*j?Nkd(~eE*t5XLWkHoCGbLiZ`@8oB$mW<`MMZ^{> zPDaotp2>=iC0`kFdy(3ZFQ`d)MyYG>EN*fx5?2l-4zlNo&ZXTR{ZVA8YH^?)KCP$> zY9MP)K00!(?1r#3BL?azMW7f+uU2W3U?eJGA!?FIrPc+lHPF~)@1OrTH8lu;R~c~U zQBN_Q!L{3PM=M&h9j(EMkkmm@WbDOf-T7^`xY#zGpqoCNKFJ|(HbdIFtMnR!t(Q47U2-y;gum8SEW_5Qe1zV}z5hk8m%Az?sugsWZ~1v&9j`x{B5&`X1k8g(K9ZC_TwW&05DZqI z+~uVcce89`n(gb^zRirvN2z&T>0yh~%(t5M_QVb zzX|I=FxTei}k$iTnjGFYtLx#xMCx zx0e5uU^OzPDYv3~<+%?B1YgiLr5@2*Nca9m1KiHo)BzBrhclynV~ zFOxZ_CUrqPJsf6WimAPtN1sI1D5MQrd_VlbU{^KV))Cv((2y^*4H&F%g-8c}Fl=D3 zsX(GhRdpVVii*Ozs`LmU81S138?9D6u?fuG@YE^rRDdrb>y@ZdxB~Szb8~yGJm4{x zMpFqNYrivcd-M1SCc!>XInmpsu}<&|=fe6X+E;9KsGMXA2@TW|XX6Ziof{CSQdJ7; zzI20(6u#K%Y;ikK#{=3`-Ck>uH{C=JN=>~W%oD%x`UebMU|bHF)hag-ZkWtLZi^WZ zgRNQ$AwZn)gkysS)V*-|HUk$1Y=V+~MU?{mzLIvfRr*~y%FDaEmWCI1Ud^6nHpg0G z9U#kF-`JPGdU2iT8S_nhZ7Ai(H-@48*`F<7rZ#`zAu2$zLF*j_$h*tt#VvS5r_z8m zgtM-qgLTIy%ko8RRoX$B5~IY2>a~7^<*Ojz#cWz2de83w^J@c*F<3Yd>`LXq;xSoq z_gCV>&mo-r^zel_IVvVg;fOnRZzFy@lV=%&yRTf6RmL+CU5sT59bEo-Gn^G|Umb~{ zA_k~dAn)MMqCgD1x|WHSJqUw#L+H%6;bNu0w`ZF;Da?{J0e-9*lBkegCUCQi4Aldb zF%JU*fRsF7)KET?gi<@W58x3=oRi6EX$CZP*bcmAw53wy7w_J_Ga6g{oG87HM6Kb^ z9n-tu<=Go8m^o4tQsQJ=)jiSz9`C5B`j^myie$VIxtddNFEcV^Wn`=fM@jh=X=D32 zVvA!~lg~b1eZbbXz(8Y_#)oH*pSF8zSEDmF@amrw`sTbta?b5Q{^DpDb1hmb0Ako$Du0%dK|&UT#L*s4GLWnu z{kIn2Qt2ues5rM|+DcbZ6;RLN4M}c@*D%y`!8MyhDJP^)RaI5G+Q1%UmbI;|5fX)| zsq!Rw0sph`wMWV3*uncCE`aRr{q+h{Ds!HfPfMKV6z6RUUfXw7mzrV z>#*6{CKEg^$C^QtWUvP<;67u?L$V6}Q!tnfgH5>}>^E$UkW`KeOJZa&Z`elrNZHTb z`fk4!fXu_K$61(!D$(=v)$ILsA^T=zKp`yToIs^$;j#A}tq4vE>d29=Y0FG=4$SMh z4fV@~--OS%3Rwit=x;RdJY{NRSwbqikv6R3;nUjt8>Ekx%t%#Z=&tiEr~#RP2xL_Z zc15tbxZli#ei-xQ$5Ey>#N}`~DpWl2fuWkz4E!u{R5ZNMQj8IjjI3-(Uv5^GM706% zqEc02isz3k$ppPf%830s++aW?^xX7UcM65GD)y&^F+!{ap8MOa5tThklMzpl9wnp8 zKWmNYNR8HvQw9&vgEgP6R5(^q{1kW?9(CKO1= z09qchwe4qCZvlS1lA-x0SX@Zo&$2M1bT zV#bBRy3@rMJ!!)TE0+W+M;)Ggxy$EA<{nmfb$((|+)%Xhjpv69%iz#G3RsAa@IBJXz-*^a*>90L**|xo@=1RwHYp%VUg6Q&PGEbN_b#Q z;to7Kb{N(R>`(>PEf5#8WBjKF2L=WLDl4NojN`y1+!R$;flDwWY-tH{;FmGl+8_r< z+1E9sFYEtgr@Qx7#&aWGOML2hhW{sDdF<#!$3vN*Ehg&~`?C9^*~WVdw!We~mUNln zPaiyYg(BFsb<5!7z58Cwv{vpopkk!$MKWK#dWHI^klsnSiIAU$!o{41yR-#p$lBW4 z+Y;KnlOe1zxLn`a&xuinZ^O-Og+i$nt}FwRbaCtvsEyN+W+)^e+l(A%Gtm4Xwh&RE z^W5ur2%YWy|L>G;yK4VPy{edR!nRW?psHdUNBhMoy7;H>3=*%Go)m0QR2-VM zr+*IFb#)9vzwa^wwHnHOOy;1-9+y#9n52#Jg64DK=Coc=g4HuJ2E`PLKKxk6OSi6& zsvf-tVhviKHk8vz9ENRe>_HMIx8;0fJU0n6Y|@Mlhvbxb54jX4&})|&@#B5ED}hj` z8l9V|XPWLUVsAv!34r)ScB_o+mjGV&rC~1pto$iB*J^Qau64XmnNcMVh=erGU!GSUJ;C$Jey+%eL+j(r0Vf#G-FE2aVQ~Km8Z1bxte~rf!De(r zX0za04J3PqPXD(`%yzP+n39|T-C zp-2s+A_uF`Sb~Zd?XCzmh=|c8A}g|>kd6K?>it2Yn*Wg)LRNEWWzVce$HOP{*x_o8 z-Syqxr>ZQujP{&8y!}SIypk*1PYUd9a{yu=P|`)A7}o-84o*)&@AlWT6+qcTy2=8) z=hM^aTTQ@BiM)vEHpx06=CNiC);%~oQow@(4#SWPWiFEXi5*}*?r1(e0-4guV*gXS6A28xRkEl{LM&|`66oN_citoHB>Rmy!W9HMwxC3-lvX-PgFfR2#T2S zcrIuSk3`hWWCF2j(*ha*f*`LN9XuiKK{!QUU0WpgplB*FGSyt6wE>C-x~;c=gz_%x zLP--KrJ^nUkRGj_epW9Y{0L}wfV)0wEF|A~QF;WAIG-Lp(Wv1pzts5vIMz#n4P9k7 zLi$Fbk_;zX01ZOm&O!V^Q;J!c3xEm(Xa)~;M8Jw<@SwZ>uVtU z3-Q_lnJai76Rb#NC*%(v23JAD;lBU;HDEt&EZo%OPfAoQE-XBy|NBt|5&7{+n+j3#kqBnJk)dHJTALgnnri|;tPE`3A>b*_ zKoS8||M-wFv?QIa3;Msq=IeT2tBNOkyuE$2b|$22c132IVq*LGIe~Qy%$l}(C_tTi zUNLhDX(W+kB+IIqpeF?vbODq-`|Ehd7Cw^pS|Tn)7`eJqBut5}df2bM7};MAY1E84 zeM$*(nUOs`xAt4Te(=MBAub1!mp8RUi|(c9zF+-jN*`$D8y98HSPlj-i$L5g$Uunv z@@OKpeQm5IQDdfXl~>&Q0c{7gnDe*`PRu@o*Lil}^+OWJ5{UG8u2bF{X2?&^e*^Tk z@`N}UqD~ML1SChkjbifpRL7ufu0mA0B8#?#0=*!S3mH0zMWu8Hs-iM49dMQ%@Vy|d zIbCqU9=lqADMOhhq!h@x1IUd7stK6EwVy9v#jEGrLop$;8BECvsQ6$M7$Gx=D7UUd zV4oMhI<@DB9b3bns4a2(i;A3dJq`h|#>Ed{94q z4Ox&2I~p4u5djrRhaz*yn57PPQ10A67o{+5X=F?`k{&#>_q(1yCHMR!^urei3=h2% zQoQ_Ym)N>8U2hYBK@onY6Ky9Xm%QU z!oDx=U)W!_Z*TZf=97XU-XChLl>dk6my&hen{2m3dNKQd#WXYOLL*tTukuv>Cqbd=hdW)3ZIMFR&F zu7?-(NV*L?yh=#_8yq~zzW-pJiEaOS%^LprmC&iq{o_-)<0tczHJ-mcte9G0yH(yV zyLf8hXyEg(?bF+ZKOD&3M*duAd#WzHHcYX$I~X!GD&)%~&W#6#GG)Nu(V}kLzJ0s7 z$j9R)7FVQh6pvayliK1DGv_om^pWy`Dp z^OG{5c=EP?d~*5n<=rUi!4@rlkepWJIJbG7%;dL&uxJw@;?U8ahCvmE$qn96)Z1LK z#p4LCC1%zHgw#cnngUP1@O68td(ygaar3|?=AO-gN$t+jgQJaay6MA~*Pe$>*$5_0 zbsHU(Jd6Wax;w~U2yK5lzhcN$XnhKB%k2G+W12TM)g3Hd1x>4dxLgyh!t=y(Qwl}f zptc>-mspaMF$2a|hEg$X)MJ}KlHdeq1A8MKT2fMC)3RLL_k%n?3eYxsVc&jr<~6KO zMO5d1RQ~qu*$kdT_U7W^Vg~z0?0Uzb!ETT%T*qF|F<=BGV<6(K z7oYoZHKc}Q4~tn%>=;#4o zGy$acHut6_;`G_g7ny6!-8inB$J%yQw;M(OG;;FfSi{WFSol@X-c*4xtlnb108HS%Yd?%5FsB#eeg-+n?o5w4O%wQaw1d)_YSS*7)wd8kW|yXpElpPMv+ zM=I*w>3R)H|Bod(uQWI>v3*nTd_o<_EhQ!&TwZnkL{irCv_i1oZfxm0X`rcGGCSgm zhD!TLdT5bW0ov2j^0!up({qO%!STMwf_DS1fVM`gr%w#u0w0JL^axCov&B4a+4UQ z(4wNE40a5tDDU3Ce=Kk68%sgTF6`?4T%-$H<9=Z6{oCyR`gUYnT4&)muFZ5>VYyeJ z@>cm0YmJEN^DLRaenLoS-HRSWMX5*bxd8(%HMyM6_kDNALs;nUH5I4kTS`ccs!_f=oAf^k6t(MXNMe5 zYckuNnVAW$?ERC+4Guf+g;zpF%lhnDz*exvq7FVE^;f!Zk;~;W!P2;loLrM*PQn; zQ9BULjcCdF@Zskr4=F4EU}ojmJL6LxP2Q)%JJiI?q$3tD;cM%__Lb46hF(ChzMg!) zQ0Una5R5&tX-PP-<06F;(ko1Q-wpEmJ4!K8;nQbMd#KkGM5Iiu#nt$`t(Z zR+;YQWA7feitdcMrwch7tsO2`87MJ@*BHd4VSpQh9xs(sX%!`+9 zY2J9FnMad=$6DCs2h~v*&q@ZJX^l23S{b4g8rT`M)ZbB}qviSEj<+B~ zM>_!dpunJ@^6@51acp=w`15@KnE zxh{n+)P&SAkGFO}2~(xJvIiXM%{%F1RTtZ}_uO0T1io*4 z(ec(kTbH#==-3(Oi^*H|xj}h-2av%A26eoxB)_Ks&~PCq(hBoE`oS;Iu#Q%p%VPUX zy}L?uAZ)>D0obnvL{3;j#R`iLzfRSKm#k}WP_E*wgI)+)Pbw5mpeT7Al^|VB^f^!3 zruy34lJS&1y?0J_8PlhY7G39sZO=;5Ha!-Uxq9@axpiU08kn0{Ztec$UZLQ`tNM1k zIL*$;PV-NWi~)}9Q41i3Iv(OKc;M*hXkfoooOL5=Ug`PC_q~dWRH7e#T4==`aixlY zWGTqZC0`P+Y`dow({C~mmFAcp7FN2POl#^Uo_BZ@LTRm?63xbdyjq+0a}hM;2Vond zqXj3T-E-lI1wQTe_it%j&*vBidfRwVeDkXKF4Z9?-9{oV7Il}I=O+vN==Xtv%I>P( zrYS^Z3BPJ`X#;lq*6}`VIaD@u^3Mc8f4T&B{Gw)gMMa!xiWI}PKf!{G9BqW9gr=q@ z7~R#YA$>I$E?>Tk<1>_4Kq;cH?Sg*ISLTD{2`Hct(p~`TuguToPD8K7TKM}&UU9Wj zX>&PAfVKAUd7N%gYz3klBZ~?=Z{NNhoihQM7Pdqoa438I`WS1v z%@eAp<>20~{HV2TVtl+e_4{r=Y!QaPNDdXc1H;2F&6mJEmsHr-dqchJe+kWVbKXV{ zwcVJi!)pkx=w2axpe7adre3#+jg8R%=;D-duMpo3 zxYu1(=u($Vzr4`*^*CO@eh#uSTA#2!nM?*k;@vxlF8AN`e>Yxp{C0fFArlXEmG8g5 zRrsgN@sv~D>?(y|xDE%gEpRX6BkuDvem^!IpnxY;uoSgmAiU@=vQI$dQzi%6j*qFA zTW58OiSHArZ9V(AyA?6YiJ*bu?Y*g=eS#^EHi?SrM`%NQisv4z@92n9qI!GpvDZ>p zQC42RcI{e{?P;=l+Nos8YG#hJ8F;ac8#lf%f9Io)vtAG@BCaBb_K1NFpnd%mvmNlU z8|a`48vVz}<_lVYg{Ktw{6Lwj5;yL8+$);z=w!@qJnze5*0E{?ywSd0c##7hXZN>_ zfA-+}%KQ~VFO{HVL*f(`7D7DMPdTL*L3eJ;fvs$Q@#1hDPdDKkVCP7->qze%?Jvby3(xwjUN_+c zuc5X9JpXDFE}eQjczI03OLKv}y<~sB2;j{)N!_2yg)L+>VY8ut}-$F;I!= z%||P#>fAoSUg6AV&-SsttYIA25nT_{M#$ZMqZCVM0q;w#{|1=7eAy3j0GDD)x&1NX zAt5SDb#A}4pQ%=;n6YRL1vsbt2nMyycpp*bP`?&LvGo;p+sKy#OeHZr%yGx`uo|v; z9S`C>xb%oO6^>R`Tj9vVE>?Sz$ zS#N;z_Aq{{O|)^rY#8^Zg&h<4d=Plw)6RC7{y;OA%7-j=Y@|eGN@BCXPdo}tKcq;$ z^4JJ3j1}kDiq}^2XIxN#*9G*4b-ZiWt}UFgucJ%TX%Fs7;<--Zb79{C_99_Pr10nl zV&{fw;0Ru_=@SzWmBk<7n&^JXcafLJJgO!M2Kk(AA^_SG36KDL7D0i}pvu3{5HjbHQc=HDRLr*uG zSDFiWluPz!voKdr>C*flbnAp9rak8&&$jo6yAI$Jx&)6zXlrRjE5(S4i0V@~bmd&f z6KVOx4%Bx5VsrvTK@Jpcg?rYGf_vkluJ!XvSD9b>Ymi`eLFqEcxcI#hK{Eiw?e%`N z`ktaRR5>|?d@Z;GF=NU&5eS8|ZV;x+Kq9aKQ+t=8Vs*dwCu6ST&Y^l~Q50G@H@nXj z%vuS0B5#h__V%RZ+_S#103&i>MPhK;qCD@J()U^|cr`@xMTGXH8@WR!VGq79iCu@1%@I z6`EE#wPYkX*ED6EtLpMhqnH=HTj?4QDr+niS>M(s0)qv7QMc;$#}~rHs2=QVAbNQ< zP}$MPdU7akjri2FgKWX&qc;6 zdvlTm0^VC~RX~B?#ujE0>4hngUOci$1OEB_d#H&hh3d9OrdCu`K;mTGc!okM?t@@V z>oXt|&T5zjtPuvo54B7z_Ek9YZ3DVobKNYvQm981fukn-jhao2A z(24ss%qVag8o4S?ui zRVPnEP@LQD`rv@Hhe;Pz13;LuQ&`8)C|ZQh4G zkZ^W`g!{(YqkCZR2Agh0DP5yeiUSU&a65?paYLEj)OFZ;@5_l~F7m_+26{=wjXf7S z6;jL|eV41b+SQ?)KhC+x5_CLS`RRvP%hojL&edcrzqD1B2-Pr1OtL!7PzHSVBS=y0 z{oiw{gK!n-v-RIA3*Vl*Z(uurR2#axyT?*o{bVs0LXEx^NW}7GN4%gNyW0FatVU0@ z_iG8`_?a!x2$etN)FeJzD_#aAyG_XJc7uBHmAQ0k5xbIbHA_0sm3MR#@`f842HwBu zu$d|Oka^A9CS+DH7gw97UF|BVOrNz}>?UiPcEIDu_;_l6V*HN0N-R4HcNI3We4!6th(jC|JsP{QCyxqT#Y+|O1vC2l^U?$xD6H}oeYyQ5e~>2(%K1Fc)>$)w z`#VC3`=RmXUs=sPJvMbb_k2Lyb|=I^tCR#$?DTsN(4C}IH;^& zo6p6Zx*OXEQAR{g=oah9OMV1X%c%g$%)|Oh^I$PON9Pgf~);~CO@3|7b-8PjV zR_E-Bi-me)6Z;Fi9Hn5%AL#MkTKu7hNvHGAPO@tIlz+W96j~)aCQLnZZB*;c3-jJh z(OaC$L&?DvmZIiWi=wxRV&^i~%V|cs`LaU$9*%UX@A$0wA%nq)=O)mkj~B+CIXZJ} z5A;w%^w#rRb%OuZdVSd-zUGBM_J( z#a$k^E0llPVg9#?|Fsh*{FXy#qyJG&fcBvTY7Pntisx2868Vh`6CDlfcyIb6MPR#20yiH&z_B_ww{}tJB@uvRQ@dXdvEYJR`17DAQyOLH$;2@DBPP`9%ef>NyQvb zGQ`G#AUq6)ZT4kVKpA78|A09dXKL&632zRKo7W*Z&APCZ5^`kl&1D(4?#o!q8Ocqw zX96!iZrRvvt}wV5AgX-d;8Z?k!`iTBo=9>MbNCDf=*dfi6L+AC+-oz*Y&jAUGwn;j z2m7+I3|5Zo*|TrS{1T``m1E-o_S9_o`PpY~&pumzAltMf={^YT?`l*14;(0Z`-MUP zLPq*zKMJLI(5Zg|MbJR+iGnh?lrjvBMkEmc#~`g!c794_>$HFMDm(>)xt2`o>VCv#h-w zJSt%;=uuxgsUQTA4GQY$(KxZ z<$;FZ)9+EG?>n-qThrh`P|5u6UcD74d!|bu-w0RIZl0m*|2GRXXHQP2rhEp427vzXys!7smRFmA8{dLvpSm>QPj6u2*(3V zh|CCgjv}=EI+1wl=ii^GYjet}2+1GZIzIM(VBjz$#R*!==U4k^ebe@#RL${^A4>}% zC7ueWa>J)yU)5R%3?0)CEMb;aDj$h55D%Bz-@ffqFO*`ykYx0=^Pe zX1D_I#$3fYUneIyxRQABfD4ZRsdx3T0Wi-^`+jMErf z@Q4ZI>GSvtyL9k0@m%DM06-I2&!-c%@JOQmcD+2Jo1crMBe=~q*ze~%uwVQ-)LO#G z#L!F~QBT7nnA+jW9qxE|D!HsOR2J2xZL^#}5!FLfD)r;!qz(X1oRqWErMYu%hRPqs z(j8W{*(N~0;s_2T#%YUQh5c!HXQ%U~j2iMXsS-Ka?=~bdPoE5??|JZzLhEz?7=ay7 ztUWF0dFE3cpi}Mri#vZK%jc%QsYpJN1f|Iehiucjc}9Zw5F6#NpMMW@&zNZyLNQiH zi+y{en3xN9_|yOx9E@EnBPOGPJw_HDZM&!HYEd-PG3wG%1XGA}oi+$O!ot`3=XGkm zQ?5E*y{YNuC#wT)%*Bk^9Z~i$766{#UambvDT*nYFQuBzXPn5`r8Pl=1bUG7EY?xD zW&kqcx!^-cenlrNtx47(ak@#2zYEJPO6bj8zHC_+4neBwuM_Kp&czA{fTp{LAj=K} zzB26P67b{f^q%fpul9H!r?_mQ_c4<4yaHzhAMJcB=2!;Fp8!*Ke}}Ws7pIaNXnjpX zE~`4%Cj}cLAL{G%GQtt@I*DUj9RcX+0x&aOpr*VrS=6t&GiG*@%Adskl^kDrE{dW8r%@Aeb*-i;SBZv+0SQN1LuZ`ET zAh^4hbkmeDf|qFAqVA6MJQ%Lk&d$fVU*|Jh*!ewV5LluP+H5yS;8_lgEenSy@cAe_ zT#gl>xM~oKFF6g{P8wv=ksdGp&=F z&cK|yD6?w_hJ#ZxnXnUOCU5F*BqD0?x&1=_F((`g+PUQg!b;;G{|nC zY8KK5Cn*hg0@$iI+WQYGs;Gp4o`n0z6UQTZx3zUP*4@MId`xCu#cLLgr?>{XtIgJT z!<-zYWy_a`)Ck!0!tBMYerB!@x zaHMb1v#q$P90A28GC>iS%u~eo)jVi@!GVFFC}KcGButv%r*-&a^ zY`$JpG^OQQ90(QW#x42Tl9_f2pb9*r*9U!3pZ^!r%c^O2FPcm+N7FM@?4q`-h=rWw zYQGU6r!Hk363G{V2t$Ct5A&Ai(fF2|)KLWv4eUBQP7pFVBkQFZ}|BM|qr zWcXkswwcO}iFD|6DTSaCQp0UX^AGhE_?Q0Lu=+NWt6t@mUyN`c?xg!#$)r*qL@%oI zUJE2PYrYO=DPV9|mFg2_K-z*89}}c>u@x1na0=rEZGeo4B5+^?An`h1win8I~_uV|@3SA80N>s>uAR^d6<_4mV49Kk_qHH^|gLj(cNL0tnKh{0CBfc<3{^n{J zCeN_N>iTNx)N|zBLTX}*QNX6F6$jA`=gpy>lWp?X))8Nxd3f~H1#M(jN4lRHnL_TH zl;EvJg`Rb}$djEZ_bj%ZHQti{bX(_zvyv`n+Lpz9%US;0(35CRz!OcsDmm?8kGd#JF{rcdK! z1E9NI3B63NQw#Xd!%x+t)9}>{k{(#`IZ$lj%8+ZVz>)*pH%33G`@iD~&f{I!@kjJd z)(Kwlyi!D*Zmn_XJy-j!fK^}Wa23!7&fjc(v`ZM(3u;s-tDY0ChSF{w#IJtv-u*j) z0k<#l%;n%CAqNY5Y+8>8IQVWH8_QP93Ww4PYAM)iP#S00EEazBb z@UxFo-X^NY=z&F{SK|9lyS~(~=M!9CBIk+i?Ci8@!7Lr(h8+Dcp_HhB_M5-93SMGx z7q;qzcrqcTlBPqJ1yo)U zA?^ZV*rL+YSUx-?#-!7u-x{FmaF7G zz;_{l?_5NJ8A*%vSi^^`19_us(})V$eB@X=0vM8n?{$DJ*-q@00=(w}_X<1;vhPuf zKzf+vIFZ8WAas9hSQMeOb{PH;)D3hbV(Oua;Ja)jaeXjzN7|Kz{So+FSFd)}v&^4E zzyiO^C&#L(u8YFiGZ&@CWv=j0tsK9PLdPWL4-wlJnK`F}{SYuk>D2{00wD_| z)FO(8ij!XMq#!SuqO9_Fy=jbK!>@(Onb4G|gvRT5yw*`1=mVIu6T7;B4Q8-myF#Y- zB$VXTSp;cK4}`t+0d+1K6fu285eqbebC(wO>4CRJe$h9HJU26jlVNevDRTR?S78~CFu9SKw9=_|7fT7VbF8`_nl8_LKx zD-IVgR#4SEFP(?v1i$+3)8WCGS|uuUA;R;Wb$8#5Z(p(DqCJIGdt?$1-`D{Z!lA~F zmmY9#QZc~r-39ATfE*|WOQTDM`LgjxpUb%Q)J?RX_dn~Xs5M!CtzG%=%@Xn!L*${6 zGM`mhQaY2C9s3zd$Ga#K!!WL)Da|0ve4U7jQvlGvHpDcN45#QwfQrz|M2N5H{{yT8 z;5L}L>{{pBT7#2uYBp}!azyL$#EImFFLw!8%EG@3iV-Nca6x6}R3Ftxzev?=)CzUs z3pSJ0Z6>tO0iWvmB`U@GD<4cjA^f&nAw)D2FR<`KJ8Bea^xu1A0IvyVcI#rt2ylEP zj=eUrr0h+E<9=;{Q*LTNfw{C3cajyp(>?rruSZjJYPn(Q!Q8+K!%R&rA%Spr$y^cK z{G;RK#QwMC+;6X1>$yf?(ef6Y^~T$sC2$*J-PnZA!UQ?2mpoeF$u&@KRO zHr%X4c=d)X_a@K)J&JY;K+%FZIb{DYbDFc9U@7T7nYIw-=U$AB{9)h0gEo5)9N71- zqyN6X`F?=g&8>U3*v6@f^>074_WNt^&VI8#Eq|5Tp|Qucdm>BU{#)G@BJ z`TI8M@tSbYrb6=Q_G%o7#((GM1H=SFgil0Ls-OTi;l5?Y51hPhF}h0?}V+=yAH*=oi#e<`yp~ z{}OWgP{|e(dTjxkaRvK17+Fs1%k@lhjVbYiPIcQ3qf@62rCAiydYc*>VQR>w!I_lVvJf`HQ_N{mkx*ocL%cH)Pejn-WK>gfwgxv5{4R2S`@ zBU^Az<+Hs{T;=46&jAxEa~ywJEKa2&@n8D}U4=cOnv{tnolBvU znRNS#jH!um^&ZXcwfgsKH;(ID*!#OU0H59@KnWJIV{V>L)O`>`-Z^kCcH4=0lD)w3lFbr76N5|w|xc&$Oh||y_ zYn(~3^kcbC<*ZNd-jLw4s;o8m`WV%pE;T9`ntgQxtILX{4}7=vNhQYu8nH^CGkWtr z^<8$aAK0|uyUAd}qYc-&1uA#o$%*!!8ryL8E-0p#G=wD-IjgkV#SL^6k~4gD%|PbB z4uUCkXlFQo+ouB5?)R2H&V;AO0vHrSopaZs{pO~HBfCN;O&6P5;_GAox=JNK4%oi# zOP?B{ODip|`Ae*P;I~o9?pVe!*fBBOqBr*(jID(xTkj8V@wP0JPr{RpK~!Xm~scUWI~HF_<(0 zGmAyAI0ml);-vA5VFF1r>WHAS^@`#Uc>)`D>BqJos0b1bf$?XZAOiot?DfL*Dxe+^%kA4QeRR8gnUB$ef#w~S8RP!wZaER%XvNAI{KN=NDH|{6r3O~Z(bl#)z5+OdYV{T1`{qnQXdys?P?@{Gi3-I@m$=5 zI-V!=WDCRE#$bfRA?SGoc0>XuoWwzkT$sH1X1IU+56l|Kcj6L6f`DiRB@Rel?}Ud- zkDO-0;AaR|PXkaP0K7oAtR+{sGoBeA* zcI7SK%#O}D6|PmkkuokDEGAmT791_}Oxn&4-Q5#eytN~jE%Sxc-gg#LK7tpy6WONf|fFtZ)*67qE{KT;~RS4~7zzXdl#efl;e4 zwcGW8v-1ekFXeQ4m&Z9eK#p`925q1|vF4&_;ZM->d;_#yFgCF%)jXt+x)~@`oR&!U z2e?9^PBKS9(Pzo$!+RL!D(EpVW!wpPY_$|uGZ?$L7LtE)8ZoPo#96{^oz3lB@Ag($ zrb{(VxnHw))HhGcie2y6Rau4*QS4n5L!K{@A%i7ADiLmO~9U>IEmX-*Vx5u=)L znmdeYk^rM%s9*dza9att7^icTgS898c~1>B;Wt>fKQ<}D8;rD!E1>808_~tRhgxK=$`h{-&o9_OZyNd7B2W%Ro*+976i(QW36U}M796Lg~Oa{dlcp-l?=WHVw zM6eA`JxYQL+U}#Ywo?sFP57WRIVx0n1AcsXOOEP{ZJpTF9HtG}6gydIy{?CG$PfAB z`vzHJSdlo&-s*xNVJ!!aIwH$@{}_mjrAIgJwmHmRiq9OVO_!73*z#Qo=YVzR$wr5| zo8{g=`}!Cf+fZDN3tWS_{NMVD;8lb_7pV!^#LD5t#zO_Kg++nW-*WH78L_f%z+Z=*faU-dlK)eRSaOfp%UP z;BN1KpcLiaWT)*&W))e#xb+=lshj!jRx2JvtmhE1_D-v6$(1+IEkhPfg=~yXmmK|K zO~ZO0ct6m@LnZBBmc0)8YRjJQ3&^9bebqnwWcrmKdK4*5HI() zFfby+@cv>Uq1LO2Yre+Ua1c{kcY(bCdSHO;3CgKFwG982@VWxL)^dDktvVXk5r$*L zkoA+J!gM{rYQxMQ-nZqtvV&kMtHh|?5Ke%F9%oYP@$KuZV2(XW{o*7`^xU~Y%P=4Z zWqowaCb3y`v>DJ5jV$nB3tyjAM5XUH`p7zgn6~(Xmt%YO?`n%c5HIjI;+7X|9fjE^ zT8L;b1>D}IBUMd><^llP#9X9|VKL2%r7)6;)|)w(;M-md1HbVtOt8K(nB}+z23s-M zbarT(MAe)Qj%yi!Kv5S2S@$cyff(;OZa>^F64J-t z_WbH~Nbnanc6_ZFhjjo_5o zk2pC=Am3a#4k2Hf=ConyXO@qN9OL!IUA^~|0p9)n>n*?O^_iEH9}UC%zWX5b`yV^1 zC@?QI2kDC{3jpu)fUfOV9WfsXdz?%Ia@;vTeA^I#^pi$C5>f%)&2tSnBOK(98YXkv zidLloJ)$_Cj2)Eo9=iSlaA@NwqpN?*nsgA~N{`}exR71-F))J{>W~itdaNVf++wfw z_v8Kbji<=$NRTNo>@Bt$VNtTx&X=^@U*xK`#+LlQRimj%+k4=>RjpU>YJ9MT*ab}_ zFcTG0X-_&dwT7GnPPan7aF`|+qXaqR2AqD7!NZ}%pf;oRh8OUmQtmpB3N+4 z6wW~{a}0G-8ogBRUMYod!ZyY_4J`gV^j#r<6?^$s$mpUE# zg{4RSI-kG4Tr%i>@yGgIR-8JGgR@<$um69>z67r3^!xwL*9^b;8q!!oDwB~VNutuC zW)u@Dq*a?zx3o#S)*0U+tuvCeqlorR(QZ_iR_%LnqqMs%+UtMLC#KJxar5srub11% zr|0uL=Q+#!obx_bWqr&dXFS#u;^hd#u!!jKwl>Hnh@7g94maKtoV_JK3!&CpRQ}!@ z_wST<-H`?EmwzV@G|K@Q1r?6L7X~8yNN_n~jjhs6-54C*Gf&S3C;^So9Bv!ICl#`J z_wp7H{8WWwUxpPxyN5M2Ou~QUWh1r@9ElfB;9zIG;c`2vqOul5Sbo+8@;tl3p?elb zJ^6n~+m96|L%ul3214!YV+$`!eCnko&b?g2&sjfs?Le08_JW_zs3p`NhrfE2ch5m; zjy872KxS;z>7dLY`-%S+gV}k_%|*YR)f2gOuD5zRthaMtA^wvw8o8V{LJ?RALEEc+ z()PLk|ce_2e5V1DdbGKO^5bX?{+x}!fK2O;im&9L_d z{t6=UH|L^FM}XI;EftDBnK@d;P;teL$W;Xrl{|uLp+zz1Y72S`B*;`VTP zwTGI(J~0jCi;8X4K$6zrHaFw{&)bU2&!-AyTFN<_%b4i=&^dF(V@Jh4CYSMc>qE7CMr`HUHLoE8UMjP;B-u z1%%hj<{)*x7DwY$`JUPX4bIqGl`@^6omGTH6uah9*8__hpOli+=xIr4fDN(pWxRF88YE+94>(~0{s(C*#bS290b5a;gP=LI|&f)bGH9r?6oXa zxvrROwYXNP#C>2d*B0kotryal z<-a**5oUoB!mtT#Pg`w8#oissze~;Tov5dM5Ha<+pQS_aMYU$IX4J9 zHu>`@c3$SOWK;(+c#pGR(G7~?!n`06G1I8f1T|G;6k>i;z8;7Q*MV1B z5f!^88=5nS`LU_F*k!SE=a7rqkjEC^UuLyNI4Y#)oUvdQ-C&kw#U4d6M82pX1Z!aTz)3Jm}P362)%`h!IYzpaDiE+}*72C7?e& z&Iq!Z*%D%wBJT_XZc5??v1=K4K1(NuMml)r#DK8yPeP;JJ>th5Wb>I5&!61mP(37e zR^6~oxGpC50DVenKeN8lo`=qWCxU>y`wyYQQqO{fMG+Xx!0aw;+Jp8f{ufRu#GqMV z9}3Fu0=}1D+I~8S^SQS3ADil`1P!wh1QKZ60zm7C5@YIt9JVR7WOvp)En>&;Mz{pP z4N9#Vg%YqrkfdAkYC3=d({ZvMupb!KIRY>v2dBBWcj|~Q)E}2-0u37-=Y7B(pw
      +O#AZV6- z)ftESA(QqSDQJ;N#0D>gHvcNX_z12G=$_l=SP#z#%I+9$0_sTY$$Jquh((v62n_22 zQIv2r3mtxIahx@de}kFXhIL~VW-w7(F)JFnbua~l8wX(;xj>XV1a03O%MJQLzv>plD!rhtEX=CkWwG1)MB!+-aL)==!y5*9O_m3IoWX zbKeDX@;GEO$j<02{0gLck+Zh8=5s;9T$mau>HoPtkYCP%@f5o6<_IExzWYewxb#23 zDfbmkJ!ZRC%&|eVP8jLKa%iM!$n`VcI_Sy)?;-C5xD+SQSZZRg1^D*?X{IikIMEak z9V~}_qm6X~-vPm;qta0bGAy;=ng`h^z(AdqQX?;qXVyhyA_(VvA^%2II4G=Jd(QrB z(^Ip9VA0j?ZHDO>9e1vB7f1C%V@q)q)dejIHe7l`kaDgn0{~>~0ik2(AfOb~svttR zz#%#|+c5PQcMUO|;}X(7>{|{OU@;R$m_}xPyVicH1qef{3ObAiC8GWs)T0xIucZyM*CF>F4o3tNdijku zoVd{0zy9||>c$lurgTz6BPcr>kTg6E9CJRDF!~loCHmma_fNoK0nkPcMq~0K4nBvF zB|xJc{5USCc{c;1Wuv;Fzp(wwZ66mdUgvjWZl^KdG^$B3=E6K zx$7`YF1Y)eJcjo;MR9L0>CitZQ}>!#+g}XL>Ld3l8A_XKn6`|a|5ocP$*cXx7-i2Ba4=`vXJ{FF84|WaVa8Ti1NgUw>la?#uEM1zw9vQ$C)rz!X z@cO+#tL`3vFzL9HbRwWUqDbYPm^XW=vlW8aYZ&AXx7pejZJ^4+6qLP#VK7z-8G{&b z2@Tt3G%zvl4H3{@q%Sz`1UHE2B`pt=pf-KZ672{GT&q9C&2=~KkFo2ILWV|y)f@kq zlOrS2BKt-Wvw84vd*$WEuk@CjCQMNnHf+N1t_Y2GoD6~u!5FEfmb=}D;ss>(5`!*w z91fM*hU7Qs&1>-7y8{RdcH*C_Q`@!90BQ>Ga$#Q2C}*w|^Dnr~g!!2eoEs8E!a1-+ ze?`gr6HbI(0x(;m6cNNapkqtZoU{d?=?~{PJq9c_^3I(fhG*l`n|8vK)lkPlY9-)4 zkKbHBcCkAgx-B*V(t;yvyQ=0)_KS+rklc!3{)8&hps@(Bi3qEd0-XfS(3l>Ax0g};xPCY>70Oj{CV6k^z`~+ zO(o7Xg8qTxID2y);MK!H4N>P6MZi;TV8pV^EmChC}PZPXyr_NZ8wy*F28U z<*ncH*vbl!gaU9T_71iK;7kV>v?QBjVK@H%lyb4WT%|JA+OA7Zha)eveUo)y#v5`P zL+|pbb_ihrHmj2K<)VnlBI^9A8V45Q?+F@f126+K#6?(D8bD%~K;d6(W^h1;&f9G&!_5m~NTb=Yr@A=M4V&f<-+5K` zxw$1GT46hvGjdLHp_^FfF+Sj3WNabONgmp!cll&{=$uf~{`pK+B+6V|2o6dCMt|W! z?Q}B@bP0r9r^j&s{7^1bJYb`*K8Cq-GkJ6N1mHXU80R|zQ@mfP(+&Zb$&qU~6ghAJ zvaP}&fWHbk8^GMJeKCbQNbrsUD(nMR>VRWAZ3WDvR|F>aQYYnL7y^kVDq|LhEEB!i zU|Ml-(DUp(sGPt?OGF3IDCj^fTB+!CJjNY3hg(b3Sv5?;-WTU@qMdDqwO$!-tZ-e2 zMQb=w0>KayH^2`;LlCT;*y)LZXwZkb_?2AZ1+{9rT#%l!vf%t%{uVxL=ARl#^l+I9 z>;enfnb*8AqqJmpx^UqKD4a2q-;2p+85sUB_$}q7#pySY!?CRf+Attn;|I!1Uzq9J zr~9-F(0!uO5Q04p43)4!q!QXgM{mwS-j8PKHr4zXaseGMqe|QINpBBy_+9eeAk3HH zdb9+|eh|-b!TTH)(Fe-!FTW5Pi22-i@LLl zvp%j}YmoirO@)o~Zd}jNSAU7{h>kwl&v^9J+GNM*l%Y)_c!Mcd|CI2#Ax{;eajxy%Ln` zz%TB=--?Y)Sn~-}w2!4tR%e0T!EB>r#V=^>{SfBDuo0M=pWBu|!My>NKspY+K|B^S z0=XatMc5_yt}ecrT5JZ4&Ny8JXYV1nfV>Ys4k){sg5&QAs?RyhRXQLBYIBj?SHCIE z#8K&5XJ@AaQVD_w=L@kb0{uvL?+K|K+N-CRA!T?zOc=^ZkX=*=#{z=$*8x%BHH0+P z=NiKxJCHm#JJm&2(Rc>Gyt^IGKMbHDR6K4;?D`=`Kq9cPQ9`BgO?3SmDEo*z?|Ex_ zB>3_9h@O8O+Nc7ieAG3FoFeC1)b6PVpr^V6_S+kBSY4S&R2>e@%t*rqb(AppZ|RDS zYamx*Yt!GfpC}H4iXDi}j{r3gV#N#q*9yVq=7c0XF|-xpg^M%Xv#J1Tbp)FxT!UB7 zjy|D}|CFFrKsqGQCR!PIrWVlh3;%J#uY&4AoTLRq$p!QyeMGSCNEQqSP-eeP^m~F{ zn+^n6m~6nSP(QYD2Y&u`keK3=Y904xrL!Ab54R-o-+18x20@sqsJmY~AUl(EMeSp;9A!wroPXX?WbAfgt?M3Jg4Rg!7)z+kw{ycNZxWQ13CNEI+-zjK$7kqTEZr&t=q6Nh4NXNu=md zM{(pKxgVIy_{>#ZtfAKJ{7x>8l@tvk?;N2G`%es=h7qWpQihlJ0LIz?RXxMuKV1Vi zY=({zwo}p4vsh{hLUP#A(XmZ=uQg=#ijV^(o0_&9)0;S&G}Ep>c6cj9bm>NBH27)28k6bKvsc&(hMdAi}b_Da6d)|txM{%2tA~I+GrD~JjV*- zYK+4$TVpOX0=L1GVk0%}rO#d4tC6l-l|bvjxzlx(4zA&z*{w z_TwB!d~vzo?c+bQ{LQ@dX+e7rpV(oYWV`xujn!V##!QvYG9hW1Sm|7sJfCcvk8Qhf z(hOd8yxhf3M?*#|Vdgltf#Kpi?|SVfh7AvHFh2!lMh2il@zT7N+A2y6entrU=cgfh zz0Clw@<>lLR$!eByc+5VI+7oT*){DL+`$=zGZJAOHjf-}aGzv`^ylq!Hi^3nzS#n` zy$11hX|h_==L}#FC9|qLFscAqCp!fq>oh}ToqB1IoZ+0Ch?v^|^|2N33KvlWA#OMJ zUDIu-1ohvhKe?Khj!73BIJ%bA4t%)V&4 zd25!&TtuCAAokds6Men`2|{de0m_uFef2)*J8nrai=5HA2D8+%q3|lcP{7*feU8f_ zadC0Gc0YM@9nmEGI~+U)AAR4!c*!>9qlol!0OELNq_0tUaEG}jRHt?n^Cx;&D@jZJ z_xBGIMf#xS@Om+Tf;F|ZZE5c@{2xG|W>vf+`VHA%{`*9Zs%qR>71-DI!_K24W@CY+ zTB~{DcQ9km3J)!J`*b9k&Lhu9uDQy7=U&l+ETJuMb#WH!8>xHePw zh4Y~!!5?(2F!4v<1v9^WVDKE3K65$TuV<9;y|tnz6=zf9Y9mxR40^1YEAC(3uMvGq z6pqiAip|FfZcvHtjY-D&ohh1 z)E^o+$f5=NDxs$g_M!}GtgV1Dkp`t!8(?&Bg_EK@yo?JSTu4u-P)89x@XCD&9n$XC zaPkzB-u!elFO5Qz4OB+@w1PQ;pE&9$XHP3aF89;l!kN6J;cOi^jXNnRd-;yvT19lF*VMd` zihcv(dLcH_qtZZ=?O3tH1PqfWP*oKo3IW5F;D5>Mj{n7h85fM^pSFKBCyOyC4r4|W zgJwRvi?rfTA7w%i`sVQ4SvGtqeccv5q6pkX>kkqdb-Ska-zOENynPKF$cB87m!G)Y zp;~yk86~>2djWUzA6p}G0cSO1yoU2Wv!mafhuoYuQlcR2w0%u>W$JD;od~cZB?=th z0l+q`{`z(sG5dKvl>ESs>A=NthMr1a-Pxf)9D|zObt4L^2wdTR&ig&&fnT3q%sBN! zw;H$82Z0JGk1Lx|Hf>414drpu(jNw-o)2Um93Sr*YqoAo{r1K8-+ue#rNV!H{4@5) zHFv)~X=++4x%APx?i#x*dq*oPcbM~Jmwx?!M|;)hE-dYmidE^ zuBZ9GqCf6vEsYP}Eb3Dr)hI=+aGbq;D&U@0OIg>sir&$#^1i`rdtjCvo9-N*ab378 z?puCNZ7oxTF_VY4W>F}UM}oL0`6Fkt?982EM6EAIyPkKSmmk1EM(d#q#ccx^EDdNv z>f1C4-MD@Lrxfk4*IEs=^DoK|^p(R#Hvk=OD7Z_FJC4Kwlu3&KB~8?){#aOPv}qW46h@q!S3PpW`n+n^Jmys6qm@-m6TWS zG0rh-&D@VsoWp3WZdjObKeU)(I=~8I5?V;x$bERn8mJkggEU>7e$6v`7T+6i6*M#m zMn_E6MC(J#xCHeuf_fd0O2qI(5-Tpl{PjnBY8en89wgu>08CP?ebs?1<>2^JwjyK| z#Ze-6HRFx8-2*NZU!2y-2uO6^+V=j6=(*3`l6>#vbBY^yw5mg#lDOAD}L%x`ZM&U7X^Lo}m>K zPSK!7OdrOyvf@)5=BCLnY)cCkXYiXFMfU=O(Z;1E9Q|j$ZBI z$UnoX?XO(D+JUYilBL=Wv}{0SuisT3R@rTnQmvDW5d#iE8SihO?Z;H1ouaM=0Bu@O zr9?3IqZrJMD9x=wQ)5>rcuzAP88tibHy|W-a*dqb_oVav^l`+;){Eb&O+Mi9>!)^h z#lw;#&lIQo2UMpA>uXY!Km8Q^dFfQ@vcpd~Rl2H|ZoH}&@CwkNFx~(> zS;Ae%t-)E(c1_F0aV(AY;3YLA55|)AuqA@PM^T~=QJ@v0kb{Olt2|P9O&scUa@)REw&?AfcNS4VxyxQz-gn9b|7#k$xwj;CEV%Cj9 z83_f-#qoN#nfk|2wbcPy+#4aAN72AE1!UF+aCz|ku%78W+wrSdH-~;G@xudIgT6qT z+LQ{_h3YWJ1t}R|?gy2!^I)2 zp1Jl|P6OeXq7^$yB;ccP?IWk6j_%GsGXxs453}dl9~B@jk~_C`Yq+Xby7evzgHx6p zH+WR!-9roU_t#77-MKR-riTyp*FbsYGWwus(8N-`N$lRmISy(gImmT{xT>eu#kH*# zW6zr>{?GlH1whQ`aSP|d^oF8=)+_>aGDlGOz9@0FC-HFP>u0<( z20->~H`vC7NonMYPX;MyfG903M+amWYT$(sod6EXi>Sasvx%4|(q4m1(jgb(iNcW# z-e*H9RSDs{%JYxzkJ|DF9EEz9z9&=N=p-Q_{rUYNf!r^|!xV`!CCK>DAoaon|Ii2W zK}An5FN>aegzZF_AV6dX;cJ|71n`helP(9Z`6CZ$j#~)5jE( zqlx{Kw!B;IV{dL+G(eMR+~GetZMm{2Z;S$D^cC#oE z%AQp7&u*}W;WzomQh>8T9oqQtzVhvYY7PA zW{SM*Pp4-0djD(?R?Jn@l0K!3sE&etp-A4(!^0wj6ZXw*fV(Nr%?%-NB%LUP zbpJ#PFh4qV!|wt9kq<31r-2|ICPX1x4@|Y#f`W+fCWyn35(2jme+_`;fz#C_YxVb? z5~Rs!6a;kDb!7L3RCXKHOSab~`V^W5eg|g!xgzQEz1tAk-}l>|t`(H48e1hO?pu^K zZVG`*2uvuAJ6?+7{A?6VU^3;C_H)J)_307Jf997C)2OxjvnT!x6M8r|H5l*k zu3OQL)Nn3SvECfK(0w~|4qRSI5mo{3?DJpwrNYrV1vl&KQiSW*Xji}JTJdhBl^Kszxi@Mvt#@c9pimFNb0;)rts{xyo21*;x%REq@Vsp6c!Z!aY5achGbR-(vML}zb?62 z;?Qn-Y4~(-^h(=~>h|>|OwX}@5T?CKNPS1jYRc`z8;vAsAU}puYBngHx{g-SU3S?d z_i+Z-*o+=8*)`f+2EI~LSf8iI9l|>bT;$!%lTH<^)2=9LWjb#{wyESlpOe-2Em&gp7rQC$v!{$m(Kge9UP`;p)trxWRk5@PaI4Zv-_L27}Ol9<(2q~GI`W8VTjpA{QWS!!!bAHoarr>?249WnsuprKgwbX zV6G~p)^ZBh%)7LU9NO%~&cA+jsOV{veAp?mER*rX=?wF8%8VtlUTG{W=>5MJfog`MYx`5hUBTLl-TS4sxOYu~P8fIdOYi7D)VQ@` zOl{1U*6F`fXs`WwH*#KomCzxbyQuk!kuq}i23qwW059esFJbI$7> zO+-Ge>5T4i$+`Dz5>|gzzP+JSE&Lyc+}BTe4`wEC^UU_BgU{=&B&(Hcy(_`7Sr20~ z(xu-M9}J4pV^T)e#Fdk>GWMaiJU;SsUxJkqYO4DEf3>xUxOLfYqjw&;^y)MzMd5BD z(llVSt>DCd!~&Eo6F^iy`=wXJ@5Ou2%0g)on4~Lhqq%+RYW)YIwBoPy4)e3r zcke8DG|zZ-7;(T|&a_kFA8^xW(h7P_o6b|`Rj@LsU=5ZkSjH_xMSwmuI@;$gZ=1;; zII!MJF(ZULlzxxE;WHD%Rr4nu!BV?WTwTs>ZX4%cf-4$iC^}nNN|3Lhp5{I$J*~`> z#7!P#t?y2(ku%fQ^?Q2M?!}vfGd{@<7t{R7e(IjLhtIV3{cn9wrN%|dtrx77l!u4< z-e_;OgolQ(*ZJxN(F5ESWCeWdA|4uHi7D>6$}BZPzLK`hrTvOxnWAhdPiCc?l1InxoQG zS#vK!FR`D|Ia5-8>vYy?t481CcXm!4t2Qreh{$E3B)(sr{Nc zTY4%vbkRxwxGbNeUphh#YL=?`VzXqkDg|q>;?pA}F641`=@RJmOP4u{%h%f&9C~ z^uJT}Y{tWjuPzoMwbHw!=VYU>8`w-wSH>?e7iCxD-_h&CxnW{e-rHgvQM zw@dwu3zmPZ%8W1o>~bgJR?D<6q6gk)j@XXgQYlNg)k3xNp|QDEy&x+}hm&RK&25}_ zc}}sGZ^3%lL@j4k0e~2moncpN+aFHW-0$xmZbr==L%=ZzV1EAVRlxX+a%}RXO3su1 zSp~l|gD!h!Junw(#GkeK?z|}x9PKRL_4x-;9M}U=Jd{GuRFpgX6xhGXE7^_-Rebf$ z{&|gyqNv2$NzFb*-Mz_UEN-Bz+A`PrHfKauMgY@`{TAGBVfXXL;NspQCUOJM!D#N` zrK$XUglm<-X2W#<`SydB+9n$r4F1E+L)Ld+x9GhdJCC+n1t_F(*@Cwcw7bn^(m(#+ z!Zy6(@JUZy&U+<>h<6rSlE$E>#JR#Kf zMJN!V=OMAz=ZapbKOjan-i~vTUr>? zlY)%=yoK0Qyod$mJZU1yEYqF4|MH{9)K_VJG2PXC7rNP`)Z%Twk>c;HBfoTL&tBh{ zSi)#ij)5?s$v%lwtzDQ|j{ZcaQudzrgB5E)E?x$*_PfNiW+9+2<>@JVc5$||{S4z( zHM1y1h}B>Kr$;1n)mR_e)sla^O2cexNlMNrcRsl`@9L_sjs;$)b^q^Zw#5q>I5)fg z!GiQ>`a&Dm9CYhd}qv zyvG%LB#=B%fqZcJBN_*_il%gw(K}VXn#t(4^k#Cug|&#J_no#)y-VG5Pq2J9nJa9e z8eY0V%BU>8oSLoL+4#uT$$di{+&7Q)aU-3_MIn-9t>X69@S_vR(#yGZ2P4YjLxf4u zWqnD`9T)=_$5zfV{PyQuoIQsK^ZijJ#j-Y+$H7hZ(x~n8eBlD|_VlZ5OgVGCr5jNr zebh%K5ogJ}JO6oHuzp~JUv!9amPXfnQpKDVl=04dIW^FVq48)uHPMe6Em`qYwIBc{ zV%Fvf-6BWTrF&3Quzy^<>3OwNByn6=!CXYn)Wnmx1U9kR16>Ek+PZXTJA8kcP}%YG z;h+Ba?dPTE?#uglX=ZJ{@3Z^pY8c=e_wDld`_mh*9=;1jvHKMEJ+x3yHvSKPX3&P; zjz`@tl$YBkU`zF-Hy7!jSfKEj6KlDj1U@eDl)KTfW?}_e>C9!_P&- zd=ft#&LNh=S~#2Nrddmq4FGX`GI&tZYnzS^lgT_&5QRdnf_{^YDr=o$Zq;<2T{rL2 zA7RNgW&-q^OR9~7`N*z(!<$Z1g;I{jwd~JyAS$p~L_e7i1+l4Gz18T?E}zR+Hnkf3 zQ=E*PG`!&x*l>(nb$&?_N+(D*a6wX_!2bH(?fUKEKb;oHsdQ_fNp-NTV|ulahA0c# z@u8$Zu)D3bRUKsHm;9VRs9nARlu%Qmv>h% zl(bAi@8q;0fG75gJxcCmL45s6{YUqX+2zi3m`F-93sOia2$Gml1OJz{Db+e7knz^< zLnt}O6!7|*L16NxXbUV3d;2v z)=vp}nqVFC*n%n{Ixsy{C^!9zI{c5v)PM?%j6r-!aS_)vb-bu{rlGAOL4BY+qLf~r zxSPo?Dd^#u={J-BvTw22W5ire7pd8=fc*)sVRvFvB!e983VNp?UiOYbkU%`1wJ|As zW6hyv)wrQQdtfeiTOB=@b|xu^6nPmxVq}9@&kJ>Y*tVzz>0^D?7|PoI4Sgi`9E7*b;peI&SU52FpDO7cE4e^{UeQfszq;>^J@6lB2|a8 zE7ar)CZYp+Cy%%PpI+~iH!l1P`@j);`>bhr*w8wqyt;xgwlaEz-3oHH&>gm<*NMC1 z+`Kv$ar;N}4zBEdrnFmL?-Tf>m&!QEb}sXEv1^M{ly6tjgqM3;-;{J1%~m3|;fcXk zvK}j+`FTHc-p9w->Jgp|~&(Jy-*5;jLY{!5q5Q}K)dy;qG5b!xm^%cTn&6{>yZ4llfT{zKt#D#s?_o-Zy* zC1OHvty*|(c`apGxTcqrUk!PK@^^?Zt*ObehgIS)-Tjjv)VeZha8@H*+?2L=%fgag z)vQMia79H{g9t%Q1fz;)G#c~v%P7I5iM2~ZZf zD7Ly2>eA(Hs&MH>pB?F940l1v!@@6eS3ai&#^*by{?Npw?z>q-wR*9f|8j*C$S~A} z|9W>#q&}OsspYJV;qB!I9bMN(h*_eRJQA*5pnbKMY@1)Lxs|486{Rp+(59m3aLwqpjilH((q}=%1eolD@E-VC}BcR;}fB7VjtCdU=gMtM=U9E!l3M4bMfBUf=CDSk?=rUl5+n z^lA=O2E$qn?dnK~T+58tO|yff@LI9yT-&*A02Dt^!-%JgwY8U#J-H^?er)rkKv+|7_BvU0rR%V(h2^9qRN&#dQn$2q zMO<=ZcN;qLHuGS$=zm^?P3D^zIx~iJS|(mfO1u@&5d%)lWzh^lx7=Vd_z@Z#S-U`N5%uzkFpcLU)-}6WmQNd}#T#Ot_$S zrJviJA(v8aN^YK9@9TtTw$o)->*y7;4VD`ho@;j4KaS|Gwi}8Pwd&+L_L+db|59$VRI-$|o>FtemJ5cc zWa~TGGg_sYl&s=Q6sdv-p4&aq6%c2wt|>ROW~+3xoG(eJEpH)kp;{)B_$^qDuN{cE z*WJUo>*3ag?PL11IVQ!!isU}=U~uVkc41`y9a9(>7Ft6Nc(a_axj>(u~}+I;Vwi@|lriy8kRv*%N z;W?j!hBndWP01r`?v0ep`-=$mlaq-a;Ddn8e`L&S(Io97$E+Ml2mnM9o>T*LF zdudhLWu-(W!4AU5($+zT;meK)Jml=5w_u#$GBMMqn@V@29d44o8Mtex2WGJ^8>pabRPC!c=tQ z4ojhjTRw1y7Uj*p-9yGb#LczftfAtRaERF1S$i_?uTHfTs%Y6AmZ)FVGFG73GUmV( zVTOqW5$?yl@HIJpPf`L*awGPG;mHDXxajFNqA&mLTD{Z&8#tUz1iY&j(>%Q$&j!z2 z>lHNVvb!CvYf3L?4tn4@7=b@S+-_0vtE*U{HrnBRQi9`am$~R+hrZt*ksZ%>QK&BP ztiPl%^k%ZbvGU;=g!`Z($ZOFi?Q`FR#!Ip+@2t^2N z=`t4tw=uK#0vFp{1yclMQ}h^98>3AJ+P|Y{pJwUkQ#cJytVO30?kFFxh4>zyOO<3V zR%YAr&WY44BIs~9#QLtAg-Q)3(v>;VtW96O?&4rX9_eCwt#3OptRsLy_j=YYY<^KD zpn}=4fnb^;{AQF|LecBO?j^nDiU+V_y(+-t`JAkcj-El-lX!2=gg(@%O_KEVDJt`0 zP!+Jx3$+gGD2T#2IC;#;iqYM#rF@~{Pw|!0&`>nR-}uwI3)2Uv<)tk_Hy;TI-1v{k z+FdE9&kNiWusL@3SB{PPX8V3V9`h^5>@TuDys~F{(mnnwq`!ymw~BMMfiC-&qja|K z8_iAbq}|)RUoW}ll(FWnd&j5)rJ)iujx4YG5)HFJr?CTLpAf~x zWQtdQM~P)!tVvPyIac0Z$oWC7aYH=!^rh8<-6D=Qy#mNhX()JTG@ClN^v*_1MeI!QsyZp1;sHN9< zp8+S+E!cLP10Wk&**MHL_kOf9>uzh- z;H^TnZ6!<>q!uE9rzbmH(M8$By0DV+L?oZ?)Zwr8+qk$iPLxn!$fy=}% z_>krB&lh(QX~Rg;tgfv8hlP!0`m|X=V-2G`_A2J%`dz2$G`6kWyKq_aErjARYc%7# za{H2e&zO|Y6tCXq4_CLmt0U4pSsJs8?!DjbIW9xOa2voRnzEsF?R7^ceJBk9I(CYJ z2{{jVMBR-33>y;2o%Z2tv#*zIvxBW5RYcnT{zBhrOi;0iP*@_vwWMOKPb;TQx6XJu zIA`X#y>hI3;9)Ab(X3Y(R_#9bX!YRmn49?q!GozSbyWPl@vfr*m&f08^f=gDcbhhj zWoyau;K@t!1kGYd_KCX`FdcO=54}5)#x@03(V$+1L)~qUUL62Rja3Awcy@788X%Lxa&`*>;EpUn~+QCFYS({`dX{^ z`FkWf{+A#v>E&ZM`zP`ze*AiO7Zm?0p#+Q#^!!;ns|{QUvjaO_-^i(enVrhb^v;{@ zV@mmmn?OS9+vGabu_B9@>(-7;wH5R5n9cR zHC9T=qCJTYWW)wSI$o12=*aYJKT~nQyBm#%W&OF>+2FbZPh`E0}H1W)pET6NEBfQp2G$%U(%(ZlG zx_5|lDjB(R-&c*BB!?c&T1vhVYmX*rD(CIm%VL8gykL-cS`YXoIvWI${k0tgl!yqa zMLFk^{212do$jrce|txeaLGEz1ai2qA(j+Lcp_$tAb?KKNUqUU!5XXHKqjAG?CpK7 z^^sg^=&GsahA&vFPn-DamRCi0VVXtyXi#KKLmdH?%+)Jq1CQv)UY*C5PVL<@hbeK_ zITJP(nX@VB=PWtmpPu>p<*tL%$~o&lIPYGta!KOpPGj_Q4e=EBfrp>Be!E>Q1fh6_ zLUFia2N=C)m*v!!uT+bde+Xb43tUN;;)$R}J9{n5Z7Wzdpj{w1`)V`U+$jdV;Kgtd z6;S&8*8UQas`{xG;L|IS|8&oJTgEZgQAk|}Ug+$2f}^DKIi?efE+w+FTDI1?>nW;> z_#%A*mvz{ev3-{+tQlX=F4`)>r16!FQosXcS{iv7pFX?ibQO20F(x@MB%1C1)n~kZ zSufe+6Wa>s6~;A&wb*)Iiztu~*I|{*VmT4uV7-b@y;aNX?>#nI)@f8919gsiv5P$< z+(~moo^#`75a-KEA`Eas$mM|fz9BYxW_DVbm(O3f$hEN7b~-7&TQWAJi+%RZfTQHd z8zz-+S3~Ua{JHgX7JY7-817Kgwhc^ofnIXG0vR4i#%dWpcG301a;ZgpyJPTW+`eYwxmZkc76E&Vx` zPksFA`yYfZ5CaoNb+L1rD&G*uFiqTa4itUWpL}g@V@y$5dTWY_eolFU$NKpzy2vJP zbe%MuX>hX>RoXh;DCIRyuMcF3%i;->@lj<|mI;vhm_MzW_*1XqVKMJY?)660v|zz4 zE4jnvs`rsQ)soElyevv-zF4%onDnS~1%@|!MU3ZK=m<}~JH*NGV>4rto^*{wD0|sB z`XyOb7qHCIUq7|9ac#dvw(Yw1TTCulk^nBJ$oEUatS(T&y_V4%$OacwYUuQ8u!c-v zL$f|%E#}>zMNUjz%CbebP_D~n*zB*e2#o?NPcxIT7;N}tzzF%NEi2<@+TMk4^ptu( z;V8d&8LnoWfe1LcuZRJdz^5-UjcTY$ePEU6<^PP5)U=3D@7JDg(aApR763M^+HZG< zyB*=y#gv)4x%5QU zPS;llS|#tizcbYnyt_~-FD@j|m)kkz+E}|3Q*2K)Tc`mwUc@lV z%1$@Ez3lKoQ`C=Z2Cr*Hq5wHv^lC%Uu3*3<@!HFTu-4H)Xq$CKF1x?@C-BNA_R=R0 zr5XoMUp}$$=I?Dt(U>M{V247}9=^VFw^T+**BXntp~KVtT@n}ANxeSd_lYpr3d7*E z?#`tw)%0~JOXXzy4jZSNa(WGh9rUH`lL#+fPL1DxzJ# zeWh2+`f5x%M(_pA4zU=b+Efc~s)wVN1 zGBFU=W9=u{At;^0q(dxCK^J{2BwkF2(sp?AT4Y{Yf7w?@Z9Gz`{GT|A*c}tJWRFo3 zrFFk6Ub!F@OrK;5#zHlUDzvN0mj6;MPS0SgxN9=xC%fPc%7k>IXGIhR$}FNBUo}q@ z@@y)Yn96kgx_b+oMS--N)fRGemn)ZJrk`61_6*TM8BX0wriBpSTqu>6IzjVKH98aV z#8SPWclY}YWXMTUMwd`JLcM4#^rS_vFu0xX!#1dbLom9>wR5mORauiuQqoH;?>0v? z*c6hyjo$~q@YRFl17|4jemvv%x!<@&NuYyak{Vwj8T%nR6n~pP{c&K!N`K^wO)1YPEh2zTCS%vo~%oO5+rTU(fy|Nbv(l4uBo+2|VRK#00s zVjpv{W!@{31&efBj&+jcy-ld_?hb&!!Lpg}W3P9yS;q-+CkY1vs4??hky$U1R&QEF z1t`^Xj;UEOtTP=gUMqtYfy+WxN}G9k`M5Tia}g%{ww$P_&A<0@{{Lu>6!W@v?SHmi z;?nBwR{K)=KcaknbcbPCT3G6}n6{5Hqd)RC3FD)0l)5}(-pEFSO z)sBU2P_0hG4A~8K6%BND7F&})BXjvB#-Xk;fG=k1{S~qY{h2JJD&a{K5MWd2FRai1 z!d@uOxV-F&Pz>nXnZg{e@fJ5rvZs}F7hJzjp7Cnd6)Y6Ry5R_wqBA-|^LSi>#6B-W zrecVs-}ti421mk|-SjS(fnU24`PB?PYs>H>X3^?-}H~lJ9XGIjS zYo2J$DbKchA?T5cVxyoB;Rpe%mudhA*PtJ+) zs;^J}di(3u1^*|zTF!OPZ%enZ=eDQcxOvR<%zuyWSX+FL<91ruM0t3s_4MoJs2k-v z+q0$)W=%Ovf92n08Pb?7Jq$Sxcd2ttX;2YD2yX2ncRS?uyL-HCHl2GU3{<*!LJOnH zc>)%4Bzl&&L=clC5QVr!3PKuo<=I_a8!ep{7?%7}!9I-C!x9XZ}BBJsmb z!XMwy+6d+wf${n<$ay#~((5~($1%Qwt(n(JSV=zj9+SV%L|yJkL_3pYkiKb=_LI8r zt#&XT%Jg3z<+K`zrY7&TBr|w4yuEKbTh>FM;Hr!H3F&HK;vPr1;Og$OsVy5GpE6~N zzU`jK`$JUMV?TVKgW~bv^BBg>OLE>K6C9a0FS*{rh%miM)j5JkMb0~>Y({a=?Q2T2 z<-?uy_^B(56hE$FMU8hA(pj7G#^3E!bdU)ml{n__-f!Ln9-!q=78^xxcwBe|>774* zw0B~3OhNPw#U)D5#~dkj7wfjfILykcIQPXW*cH39XdKCYOV`LAbc`e$M7ulb<_4Nn zHQ8N6&q0#sA`%}PYTvn;d(ZO(^}f}jaRqK|FPT)!a)F91s!c#0XN5XJeUXZ!Bo%IF z*$s(qbtr(Axen{>Yq`{eNfM|QRM+^;RUZPRS6{^5B?Y+leybum`m{CFc>F_{d3Nv6 zTe2T-7bIT|6d1Deeox%}kVNx;@7(aMZep%GE(ha;I0q$YV}s50{rA?r;itM z_k+@ZdcZ!fYS0G38X_PIIzfkW~=XU1Z*K3f|n!Bf|Z1m?p*%j(XWLs@LX!YzU*(^x3+__(d{haqO;j^ePh57OXT9t2j*5|E~ z0}|en+~V`q&V^&PyN@?D`Lx^(Ab*2VW{W8_3qeQcwb5ZO%Nnl?w$&EtI@=u}xvcJ7 zW||qGUx@_!Bb#Bv&s;;|_vWwL>BYNS+-d!c?`~1&@nG?UNvIDZaau~$M7;plU>@&S z!R4a50F5pz96u#JQaqT>k?3$93T^L*@V1R~YwQ~qAvtvJT*jxm+VA+rqPq8|beCg7@7@JJ z({)Ss<#gRc1{U;|-2&~%PK3JZ$Bsn2T==KVpUm=T+{#7#N?7=rmi@ie>%vsyW+vhB z4cRv#8cp=~P!`)En-{T)Tm_MSJ)PrKPc7pL^0eE@s{f@GU{jmtt8?lW$1Dd6Ts9A`8K-W6x_UyL zExIO~(o^hR;Ri~~)s5Q@hq~#l8%|%Ytg`TSHZS1HiSr`%UmZ zbReb>2TWR|>{4%=L+2qU#$PlQc(s{|u`Wi&oHp5qefibL6-(p* z)wWhCwTjBJE*vN-2na~D4(b40ESV0L%*dWe>rYuiR8*9$sK}nON0g-^OJs%+Oq2~F zh7kxPzx!S=d02UV^G69K-n@J78K3jH=bZZ;p%8@;Qmxn%FG72I4tO-iWXFG;?8U~7 zAg5Gz?ZRSET4A;g1rBR5V{xK)5B}b480oOx35? z-Q1@=v3lhJ8h0Y5?t=v+?aiTarniMEhDbXMcU|L&Vg|RDPX6B(?qs!#kp~Nu*{E~r z*Y2=Cv_!D4F4H_ET=PL4SjW%U4U+*y%kuCM!ZpE%+*C(`HbDUHb!C70u9C$DJE zz-2a4hD;xB5;dmkNY!odAY2Vhdg1LuqMV5(bUmy}@MJf4Rr?vIMoPD-wG| z#U$<>B6kxFggjb$F&v4Xf&~mpM zGKOL{Rdr7OfqQRIw$j=8JKDozVRXN&>HS-U%aBtC^A&thA}Q&!bY4&*uh!kW%a><8 zGkE=~eMAn^GA0UH36LJlcXcpwL+N0d-dNoKZogXpO96Bl!T#_Zf;?g==$AT{bH`(lL4>fo@oQCL$VxA~n5WKA>_JNA%(Z_~8 zU!tU(=P+*LpSen+sIWbbo8^Z)hy}>R%{W7XS~BN^iL@f3*$%2@h0e|Pq7L(7!E1J$ z<*z`B3X($H2P)>AbLO;WLoo7xc$`m}G86RM^nyJ5n~g;nMeweEiQ3CefrweCAbOnX<)Zu+}0*I#qi1hDavfEEKAcD&r?c zCPOzEfme6;K^3%IfY#s;gDtekMnB*ccC-9hOW&OD%t4-f+CH>y4Cd5-`{4|Wdv9;i zQhgA?`Ync6tEPV07@4(8+87Q6y4(1%cZhw7yU;+)UH#S{!AZ&t=UT%+Z*(JBE~3J& zQ4txMXzsEuz=^AT(+Cu{ZLS)lx%s9-!Rb(ze=gnvS#?F)$dA1Fjxz*>49N8wl3&QY zv94fG|7b&+0g>?_(55Jgh7AdAfo^4dA7By1gNrV#-dikV#x8QYhG zYQwLcZV*zHnNN4Y871tW?J6cFLy448;tvFWfA(XI30gjJup*ud3fMoYmyBOgyH28rL_<3$VmETINmx6zgPv1+k;V@$oQ==>@#T@ zNvNLPc|?3GlQ}Twz;ihM$Re%%RWtUTufNAkpdjYK4IsCLGqKIwGWim3Lp8I^AOgwM z%{V@NRk}bDHnfLP47OKImQ;I?n0ji&G40&r?e0Av=Ro)l5m0MAd&opuh&E1xcx3Mh7(E-J&nJD#( zK{y2BYbLv7pSaACAeD`hd6N=wiVK#aDTtQh$Z;LlclC@(WH%HF@zp2AeoP z@6**8{Yk!majCL@Euk#!(7^7%QsuabiqfPM3)4`cGu*7+TRNNF&Thd7S=kHMR{{6?W> zgnfr=2#mAVp$NeWFF=$>j`^Da{nzVMBjO8Ekr<-O9+;!gs0MB8()9s#a~GKYUdZ&v zht<$>tdY{a@8EIpKrY{YAYC`wN%NwqcaY)Fty|lPYeN>I6a3*5nlh7vX;3aOVF`-m z*C%g{ZYEYqP-@OSC3@!v)cJUAKmZ4`x`=lIC<-n-6N<(c@0t|{KNEjF()SL@*Gbn4 zBf!?`Cop>mR}6`x6a%K0$V=%~q{SFP0Q(N3hJHn_1q+`aC;DRs-_54;Z7X2qKcevt)4b7Yi9vA*FTq_sAEj zdP%BkogWDAD*yRNM^|15t0E+7;FM@=$`6lq-mY!<L@@3N`@oLZI_sPBv9nr&IF)-rhAlEE1q zrZX2^k1kEI7~{T~#Wu>BuF#<-xPSpfza}W$!ytvcol)5&mg{txR?1^43{MX#Aa=Ry z?{@3wt9mSQp)!o|r%`FAzt~;Duli4A1M`LsVpXYxaym^EDw#pG0zhnM5%E>OU_G0R`N&_+Tt6q_hCn~rA&wLnEq zP6139X`qahG$|hDLMsnag$Sc@-u(&<>HEmwS7KFRBiJ%U{+3NhAc(E0VeYYWQhRznckl?|C+ZK%9gF4iv_px)}ygoalv;5-L~jo6@H^Y6kZ~FuJleXDJsN+(sGUZU0;jA7&=qxn1J+!#M^uTr zGvtZ&oXMxWl6|zsJ;}nM0|n+-kOE8Yowk z9sMY;S;TXzLYhOrx5EmBUDgrbp4OCi-Ey5G+P|(^r>0v#lN^%RoPuP(+TmGD_0AYBRf&trz*j`s|$><}7zWm#8H(-m(AgM2ijl-wd7;znbX zWJUlHxI(O>mrn0Lwjx*fbxQ-H_?f7|-!=qO#jJK;wKNE)EgFs}7rTZ`kRcqnU>R;a zlAU7zEkiKi_TEC2HysSk?yyU_of~n^<_A-t2U@!ps4uARK%3kkvDD2&Db@R=YR_hF zkEI1Oqp=Zj^h`K#F84>v`w=qxP*XvH@f7HoAHh@!WG z8uVWUA0aGoa0|zW<{La4wOWqRM=4r1;~6nuXn9X0fqeF>>YOvnq^HW)Bi#UroB*tx zVj*O{ccD1@@fMNN2V!!q09~w@BE}@s@pyA23+fmgIhwDRochXT`0?Dk%c(vFvxmbgVCf?>xsL+}2#2kNO6F`@*48)zjWrYf{G>2zW(GAVUw zB*j|b<*wU*jqlH{sJOB#r(j^_pE}mmki_jpARDOcJhr4x+d|aTE!k+matcr>c9}ZM zan)lp?i`26SAa`-Lat)kI8d%9(iy`1aFWT;I}lkK8c)Pv9B(*(tq)x6wvvyJeWRfz zTe=$bOflO{oxY_tJ$6i$X&7$_?#~$+S;k_IB&c-y>c}AutGh_b@-^H2TcNJ21ap+I zKGG1JTy1Nah@-yRZZIgrh$-(WQ9%McmNxx+J;GYGCMVSH)WfaVl*`P7Uqb+)o4)Xn4 zliNMQe!WaPV`vi_tF`fB7WsmME38*yh@aG)-lo~j zPx&S4Cm@%Pxy9PXQ-i1*Y&Xj%{Q5c@=KtlCc9Y`KPbe{u9t*cP%*iZ&g=YcUujtcrh zaMvw|Rt^c>wAGc9QFz0A#2#2F1-oB5Z=puwHw|(9N~CwSSVXV%gmn@ikojA znhsvuIrao}$u^+4S1H#XXZaNg)o9nR1JkOn`HF!yD~?ak)&g7&Qwu0Ye2gvaK&cXv zJ^6blP}a3M{0t?O3%~YpzzsyXQN0>XZ=VI>d9PHYxUZ3{M^~h{Ca8mH9fs}DrM_if zTvjf8@b+$VvD?h?q(rdHPe6(5mE(OacgSk))kyFZEbq4_MBj7oHA?sXPpR&&MLcaUwxao&SB6T%uUlRAtjH+?aA51u(0AI|QY2}NRIpWU{v@;|u3 z=ZnVxrbb*L^7*UyKhLI}UcroBHe)s5-Gujy{nD!qM$Fp%ly505>G@c=mapz*p&QI@b#9iBHp+ag^d;|G5KxVXs?6X_i&ejJ@9+( z>m9=z)QfOudJIU?MnIp~kU+!E4x*QOX=C8Il+xZf;v1Th^nFlea4A)oYoM%lEp)6T zNxvdC3@AbJi^al)7VI!vUc}pag=w2HL&QfNpXI+*u0MGUZdHQKm8SM!XUf$~eD~)! z*Nwv8N++97zZCbxeHuRE`ueZj&OISEa(%F!6&}Cn2c0bLbX%UZ@{GOwcU^1FGL9V; zeBBpsLbCBA!8#z+a@9JorgDk4U_uj#yP7Rp|80CZ1c(pt__0xG*f0LM;8*1vD(0e$ z;s?@EcZbjVd=TkZ9&1G!JhN?wA$IEuETJe_(4_aD`e)fT^CS6IPf zYR;<7Ry6hV_U#=j&ikORT9qQ@=`98DAd=;GuRXj?IiO?s8O!1u;^|kXt5g0;X(CI$ zwqHhF9nj7$$N+LA@>hu1TnH0)C^_eZvc3pF*HFC;lmN$1u`b`{793~&NrNOnqMrjz_ zWyOw98>o2P^UP_a|9Mpp^CmS7$-Io)D!YwZgk#6;nA zu}tQX=~s?O$x4?&`cSlN@2=iF?mz)E*R%onP^&W!`_z?uO=zEVj>XJd0_fOk?7Dvy@58fV0LZ#E z%L^GRt8AmOe<0-OB>?n2%Du3k2}E1zRJWp>!0K9up5K}is=Ap@B6GSsEe<~G*zKO& zQmYE2NUo&@aztOe2s!NWFuk2oN;WfZ?ECznj1V&{#JiuA-1)S5>Ly4|#MC_m5qgzF zt3P9^i86TmY9~)Ml-(TWN`JT?S(h`i`!H}-3HlWUR5J{w2i(5<#dkr!w$lK}UJDhr zgD?7-0}>&Bw3$)Bn8|gz7cFz}(NXg4JKZ>V+j3RANk^vbOoj+M7&|ASz_>0bBJeTx63ehJagtEo zwsMKkGqdrBx8>P&@cSYhLct~n4|-QAIny=6z=gCcrlwD8dXI4D#LD1$Q_W>%Du4(r-fFgU%d zgx3!p^EP_S5)@{8g*vAE_FAZGhy4HWR9w76X(B9#%FQw*%ETDF6%>Mzc;?{Xa6{n$ z2R=BoblrPFTwMJ79}3GCvM7X|roRrR{*j=wO!Um}4<4ob{?Y6I{{1T=c53m`d$JoJ zTR+S`^2dh2Te|B{S_Pe0{BX<9W9^mq(r{~jpeYlACx4e9w?;FE-`rP?$J(}Ue6Q6? z*;B}HV)&?W2@R|&RXs-C=h)4R2<1anX(sx1dVi0h8qH5eVDMzC6Xow0#0_45*6GVp z8}@QWUPe54F6$5|S^#~sAPl0cgRdbOkFF+xmCuXgJV!NA^e>VpTbLrq7z&Y(Do^eEPZ z7;fyCM%yZznZX*Ca?W7Yy_EtDP0u|pHj{TQ@*O0Bj%YSVzsGC4cx6y@eO*9sp|m|n zSuR1==#R%5K&61P>keN5ncZDy##qmUhE4Ew{5J=|4GPTWE1ZY@eXlDWIvT+3S+>a2 zL$5cs4N!Wq3aD$t3?AR>lPCmnUkp`ceA#3bmOBlaUmMnJp|wYJXIx@wR0uvrD5l>$ z#2upXyWlKAA=^f3ipGC*G4i&@YGkQ1^^)AH;*sYMW+L~;x}0&RCRjPkFv@6&9p3Vs zYlo&g{yFJddA8J%v$kZH@tnz(CxbeH^QWNrJpu)Ub{#C5HuPgA#xKp1&)IDUDpjl1 z+7;AF!5KqcQ^iq-GWODu?0MI=oXjbcv#sE$&$(=S@H;pi>j>nSZBYut z2+LgF8<-o-yoF}wP+~46Qo~wQDO_rbUCzr=J&6|Bn4zRCV6-Z2qIX{3TJkt1OC0Z$ z9UQ|0S6c`zET~R}@+cm`yylWomifGblzei~f=MNP574sczMy+SJz07;n4fyh&}ZOcmkY4F2-jOh$J)e2X6+k9B)6huhhie8dg>%9SyC-^n?p3hL|dq8#GIij zHp0YZ)vSR0KtF#dtm!%7Le70hI+;LIs8}tpaq2QB) z#T_D$%!FzOl>5GEEiL(+>yMc902BmKa2dp+d>_Xp$H>u}y^!s&NJ|wBp_;|ib%75d zv#y9D<@>^xnkZb?+v@_d-#wzpl}R56e>Zmws` zi1s71KN7bgAPqtCQjQg5D512~h}9P0Fz?IPK2r(9oQ!_961IBnd1`v^_yHi*1YVD` zcZ{j+vy&nPGwM)%xnuWUSG;J8@$SHwNs^kXxnb#;jAqfNx98LbUhGFyB&U}hj|h=I z9r-WrW7H3rLFcl;FTW7{-W%lRn-j1o#FY~l)5DOXabk?!_(erLHR1CC0FQHT1wu9Bq%{*Q5Tvr@dbJL?}pow!qfAtSSbEMVb47CVP?ZAFd(rOu+ zpLOXC0BO}D&rc4DTZGv%v>ApK4DU=N&+K~704mMl+Xa-&6jI;ihCSxfNn1IEd1J?< zjmyQPO^|L%^(Eg;eRCI;qkKp5b<{Zj`qny27#}WTc6(&oad!yY$*0R77lRVLVbG)a zt}YCbc{z81>u;VjC?_usK40HRNii(RTh1tPLN*^9{0-r#P$Eb|;$o9T*ICj399Fae!1Zn%11(Ya@1j3O_$@?U7J)d zq_|P@uSYNpEGz|2gV~#7^bD81sOQ?Oe?bXqN%uc}a*(WaGgSl?mIJ{`b@YZO$kTt% z{svHZl6QoU2v8rXU_9}pnyV>RBX8K}QLyP$FA}HrCP4LFP{#n+y`YKFedo@SwW)Xy zyk4F!Csfn>lZ_$bFlV2Lrr4~-ge~0(6cBFNzEw`L@_Eq0wpHY_&qP}~6o?i{(pSs{ zE42^*24ToihdK+2yx62owW}d5k)KGT<*K(KvupRtL#-r0o-UkX{RG8Gy&QRzRn2aZ zOU#mQp881KPFpOJg0xan_qS&*4iVRbLC*se`ziv_N#B-_SFd@X~djeH`heF{~FUp}XrfqylP#zOjl!=aK*F`#tk@LC* z`SDphqJrk1q>Dz6l)2YHg-bTZc$V6eAY_%$-wJl09DbI8$vV_ImE?6No-|4?jag@G zdOwF5Bn@{k`n-Yx7GTsrL6Y`&nEI>dzLt(MTG3c1U7p=H?uV7p!7ge9>tU9wEJ^LD z6JO5!;8d-JZRGugG3!xwJ5wHEx-&3l+O#V3bk10edQ-XEUYC`1Uen0Zo46N6DkLpBM-AhsGTrDPUY={J=8rA#+N%0dJX;hurpaP6s9D(l ze|8E1ZmW$2T$caM4f28VFdfa_kuy?NrS&TuIvbE$^)s~4U`f6yEhQebN7t2#@@jz< z!}@xQ#&D*#AOZkoC`<+cz;-A>VkrtDt=Hhu40MXRm@|u4#*vrBxn?d+J#*!hA26MvID>W27x#CfZ$Shwj~lPZN9h zX~!JX?)LQ%it6JuuRbAX2>mO(9}}7lO8MR;D0sV{kfji+C&~tnE^WfNdBYxYeTt zZd1drWqze`-~#^v&D=L2wnirQM;>l`BAxs1xAk20d8x;%O;T>UAT2f}t+t)*mMe!e zJB?Nl^aJhCK)4SZ1q=BY`yC(ag&{apyO>uaiNqg zy#+f*P2!(tFOvB;ZvVmEa8e|ifTxG5`5IL=$*lsFW@9!n`sJo->BvA(MbQ{S?Fmp@ z-Pm<#tE52Xyg6qy`G3Azcm)szg1PBLU_FJX-PyAF19hQRKj4#Mx6XJk7aDJY;PT(> z)diQ0x;vV_+iT2Yv-R|tUNI&}I3pRpTWF=huLY;2?N_4CJ0V^=6>tcVM&|fSdrq)*D7ePwe~@g@ ztzM^cw_Y$<>ZviTJuPd(0U5~1^tJ9f-(;b8@Hk(#(UfFkn#a@*lTYQK;OH|~qAR_k zL7)MIWr}uFIF1rb{-f*LMsR9mC=&8e>LNg(A!#Wz$ryX2i3{Cf47!f|FC!^xdmLYt zR`?KdmSXBf{&L!SNN@6UIG>YxNoL5q&5 z+`8`wp=dFLy~`wh5pzLZ0@%WW_sxgAXY{cP-wM8wFb0t&98kdy*(S98v%t>KRHBtP zXUgZe@ER05#~*kgo@qq$O*r676R)paGUc)CTMCnf6VN5qmSH$@@f?>ay~73s5d3b$ zVQ7|MXsNNKJnZ2?k_P{*6#kvXE3k6oHGjEO1h9t5-sCUSc`oip82S`rwq_GFf{ z@c+GVIjD?UgE?|={~7*g_r(tU{hIxx9w)SCHd3|{a++?GBWW-#TSd)!iU2f z(xm=tW4jzWW=6J1DYp-a5TZ*=)mVhy(me3 z7_S-;W0Q)k?Y%wDFB{Cu$8z7q*|Gc@KTi$ZT*jyN`S!TKCrS9Qg3s{Zv{qa*7gds4 zeW}*g&cFy z9lvBOFIr==_WS0=KV`}f*M0=&fXPkzmVdbR4060nsa13lWe;v1tA>-6hVXA|)*gmR z?e?cdDG@QG_h=5TP~{0YN1}jy;yYZX6Dw~FkL-ed zStS#9U68f&cp)MWpx)!57g&@AZ|Ab7exTYL9}+`nbx+N?FPei zQ>y=j&pBi=3%S<)tR;M@DVz(3QX+?uxZIC3FY0a^GG0vfs$MUpTpx0Bw?JcBw;#@O z?HFFdPY6Yu9s5C8lwO60OsI}JA?TyA>G^>_er%|$cfB8)*V#nY;$AOQRN zi=^1JImo@*e5Uu0whh(Ayi|`4X9g0&H5HEHMh)c}aSYQhQF2&W$rHcPX4|TeY>A-k6og`UiO%rNHqC^ zX^$e9=^akmLzO|b!xomZyG;)cwkZW1io#|JrOB|3YYixYS!F>fE5Rfg<8WB5_O@El z3~e7LQQk-!5(C&;!&BpD6XwoSMRcrg$Q&H0Vlc#8R2GKhI&x3)kdey~Iqqm&=-}XYB@-@n6 z_8EN(8bh|yo5XV5Gza^Mn&;K3rHuq2g^*&QPz@#Dd|RwuPIDWRlsS)FwL-jo&Qy)X zL?WUd(p`Db=jUh86#)3~f#%wldCOHu!~~T?MILV0z`+~KJAJBThuPzY5kSs%73+{> z=!(d0NyBGjo}w6KBNWjZc^SoHYfTU5gNl_?K`OLB@r^r8-yxEV(l=bU1nnPw|7GU_@@K}vnYtucgo`EZzTKEEyRBfS#jt+f7Qpq6*5G4GXiM=lSZFwUj`XbW823TO5&h% zVe?%biy^IhyE&+}uuwW`ZQ;fIiHmt6O%>NayK2-20Ak1|hiR4$S$u>FM&d4^b6Q_!d8$=pM6Z{9o< zxpA^njg}}bbF#$+asHd#X>}Lxp!5`~%iNgFY7#hBfhNtqacl)`i3VI5{)vwV)+>42 ztplepNJRky0rj1AR!gk$z`lo5gg4ychEWoxJC-ane63<8!lJQxv0{faT@S424hMCqG45n@XnC)#~>ZyVcU~ zdi5Z1PyNxLf)J)G=#SS5mz<%sq}t~c&3a^C)H`VJT6t6tl&U}XKw{q%`Dk-7iCs@& z-ts#X>s_B*EjL^hpqjb$ZXxHrN~bFsStfYa5^`{WCx%^II0^-hsZlwnA?mfg&(bn^ zG?%wGsE$&Kw74m|!l4IB{`LbEiK#uCb>7cTY8d&Jtyf$zM~W)l+d~Qq;DRwLw8Jh1 zP=F1)Z&%>H3+dz__X$a)hFI~Dp}+tBZ*PAbJ)8cTon<+eDX%+~q^(8o^b@L}@?y&br5DO8u{H+uYWHKYpt9Sf2Oe;!1N=#R;XZx!t_sL1% z!H>3dss%3+=_ck0v~%B+1_h9Bq7P*%>^p!Gk9GBWja@BPp2>3!xxLphJ)}7IID(N| z(*#8s<#4@?!sCdq<%8QXgB#X%2aWk=86H+Ur=L0z2PRDtQDAy;pI-@5K3C<5uj7b= zvw_2^No!Nvnut_R{c`{qJ14Gs6m=zg7VUu!~j{bG#AJ@o(Ybh%ylY*Dmv0fiER= z;`s3ol%|XNNRwK7Q?IwE6@t4z-j3yJ(WB?6o#Q(=P~?~7D4<-kjA?&w-lK!-d^HrQ zuxAIc)rk%uT&*xij}UiWp}O?GIIMN2+b8EOd%-C8tAONy{A51JAyTqLrP(xOyt!Jh z#RcHqI|G3nFV^}Jr&vTG6ydPsJq_8jj@y-N8Atxa zIQ}1s?!_9L_&^b7DkM#+i`JzUnj_0Gf>+n|!}%CzhKghW9zaXh&aMuz7i}2^<)g`? z+`?{VhF3R2oa*n*0H)^vpTV#~xg`jbUAH0Lm43+54K!kJ&vN*k8i6dLsQ1Smv;P3v zR0H6(8O=dy=^1qpLtR15<1tEBX-Fi^a6}?NkkP zFXJnU2S-+&HG~YN^$UTW>$<^@FHxm(Xlz=(5ZBfhHF*8;3YYSaO2^NaJVo69VwM8& zAeBpCREat{6L5;LyKoA3TnRvCIOOs1M0QHG*=_&a+9>UMuM1S#5&I|FLfLDpk&>%Pw-G^}`I_0m+|#w8z2GwGW;Ddy&UoCTi{sMU_qbhll)`;1Ir7eh%N7^- zpX4$C_~9`SMNn(RkVO48YCgI5R(VdnwUo+-eZzSfKKdPa48z;N4sJ zuHP@mfPdddj&wRDV}gTD{NK9d9;f}Y0C~NWLz^yi+i@k7_ zOmNPN_BkrEkp-L-13wMwvk8ZH1OL>cWmLY* z@8j9hK8GK`&BmF4wdthHu%kqz-e?*0FEEX)T3i=imD)J7PH=EasIgIl;(k8TU!yni z->HxISjN121&kyHBlUQrZ(R(}9ls8PVOwwaJ*8yg?W)uX-Hc(LuvbZowyv}(;KHvN z0=J!h1PB0k>kA@dU^8KtK= z##X&*A}DX_Ehv0Big)A4f~W+^DoEVr$4F2byaLLhEJiWjl+$KE{bJRuel6#0+(~{Z z{eJ=nPSCJXt0~2c^WoNhw7Km8)&36cO+--Zp4rheYzkc?jySy}#hn7?`m7@t^c{rY z%@tJP@ua0+Z!w-#FfDo#)fkbTLt3Xu@xPbbHEPHA{b-DA&CmwzB-+1K&zycS&lJtD zq<-O!G%f;0Ncke%TVdao_Q`p3)iA9Jz#nR4 zvzGO-chv4N0Rl!YAme5D)pSy+sjCYtgMc)n^(bnm8U>D>hBmEiJPu-s7i)p*vkd%inoF}BMsqgn&>)Z>b) zt$3r8UmiPvm>H5vetVRCW%JIg3*S~)N_{oVNu@KKy1i;!qj5Bs$gAtu-wL&VyM9CA zk>WqGj&^AWgEl2@$s-(HbLRhY>NlM&C%S}&H$SkJbRZNH7H`SV|4B^ef!H?eFFzf! zX|WEq_^0V#onM){JOAvv*5x@6l~F3+!eOg==WzIUNyThxYalezBgyU%R@*9dwxT6O>y( zkXuqGn0<09($_5d`~lJfCB_t}6;gm|g`PBht;YlZqonsET%{12U*|K(*;~f+(~g-^ z5s?J4#=6c)sOmGAjuy|LQ2HfOhUt9tK0ga=T*7%$EgUj(e+!Q2}zQi zy?LTp1hiRH`C3see;8#o6Zuryd#@@0()V%-eIiTq5LvTbH~L*U`omEE3mITpvqrQf za-5_)Py!jnULZ~AttIz>SlFJK@zT$^0I4g|HzL+co{_oO{+@ykbh*qqm#-zj@}JL< zJw4=A#SB+0q;-!^`7e^DhE&+c2ru>{{VC6Qoo5`e!jrxTldO%lVCot&F4HL@pj)A? zXZNH!ms@I#IACymX=FMMx7)W)1i4;jWvEk~I?ZSeMlrc|so*bPzF8Wdm`ICfh*yD1ocw%>I&JXkAz3#LqFd|JYkHu>VruB$`rtI|cg54+rn64)ZrcM6876ZqfLDj&J5TLgN)=zdKrZqwQy z5~&`Wd1bNSD@oP4;5&Ov`spvX`L4rYUVds;?PuFC9OyUYWV24&_fzo~s1d^*FPu_@5u72MUQv?xX+UvahBl*)k4#4>Au8j!niP1VNUy9S~Y?&j6%K!u+zg zZC(jonw3l46G-fAgm6wp8l$p~*R6$+^hIxM+W;5$KZt^o)6~IbjDn<&fIHeg=R{2l zOXf(zt^t8f|Lt{g^ZXC=%;HQ%{-VkZo!-~4i-OopnH3EalXS$~ew{nFrg?|*cW*I{ zV^=fmAS$*gZMag7+Y^Q0J(v1VX4Yx>lVNN86N=`MazEW<*wcqJUPqms{E)VdCeA8G zj5D53DONY>c$MnUx4}CIm(xZCO7P|O0e7EBYgklyQmi1&)n^s6epyg2M_qV&P4Kbp zG#BgGVZC^7>jF%3B8tId^Y(BT9cG4wC;^A#Gu%dDTh)7!1y-BjW7#>zcMV62-+>7xpiIy`f7qjJ6ZeHQNZ8b!y0cZ4DVH{9f+95q z@B}x0dxCe^;`m3wN>or=mpmG(y`45WwCix6V0T)v7S)6Qia*iJSMtP=rqGODTDhI( za$*M-q=EL~i{4df&$J^-wg^5JC4+F3@%AE!G7dcbTF27auQf>!nR$@He&A*YJ=8It zxs3Y69i)hk-}Gn!>X%|D8dm^in=d?p;%tgmW5=K4iq4%H?rlF%THr6XVv;WGfifu4 z5(e1a#+;5qHH=v?ibp2s(Z_QFWAgEl0DCBP`>1`oiah4ylot&EyvmFk^MXHij?Pn#BYR>k8ltcuA4Ep!E~_4YmQHU<_tbR%S;U(e%G7KKc1kpcTV zsMHB)DmIndN-<>0lwR1%os?H?+pIK3jP9p#s5#3Sc={%U+W%E^7i>>x4g_-+oi+!ex z=d`u>Ww#-~Z#(UhAoR8dZJdJXfLw5WQRiL=y~&3a<~h44A@=6`nnG}FdgL^Olop@~ zOY`HzM0NsWRh3Ou+Xkf;jq%YVwul@kEI5HCzAm0Kb}rN9eh>^d z&Tp=ooIAC*6FzSiD{qT{@$!xvD5lJe;yW^Ah~pU^h5)^5e$X*JBr@Gp9;7Pj<9ZO_ zdD!>=jpF;$?R9i2w5duV0DE3`^77$NZ}Yhlp<%RTnXj6FZe;|*brfx=pB?TsGnpbG z0Gk$EP9d_y!VoCp4>R4Dd)gks(Cr^ID;Dv|vNM0Vv}YqIhJ9SM-3`~*Z;vqeuPh)@ z7Zv3TJL>an22)!%L8RxT4Bm<$53H*-X$`5+T6I7Oxjx);Gw-;m$%ja&%XGxSS5wjw zTMxw2`QusX9=Yv^P~dNctsk!2w(CNuW*?daXmXsWO`zS|ZPk`ftsl2M!M!ZoiT z`vt;oz6IMn@j^=A1-^J=rrDnus7=8iyRHF*GPgw|vN&BbC~DWjA;AY%HUr1_TSLg& z@RLyA1XMG&@&(219kYc>70hl$x|IOFkh5g%Qa8xYtA!86akX|_G1$Lw8@d(g1=arYGVJY>z{WhMS5$Li}xtN#UQ`7Wl9;3Lu5Xa2=K%jFKZ8gyJ67$;WbwXLzX0U zOy2#kL@v0$jRVmF{N=eUhw?O!R_Wk||H0WVFX+nrylK<_2t9+M1Athc)%L`Lp3yIt zM(}&YSZ(29rd}r*P>TN*$dx2-7#(+ns$k3(ph={x^yIK&r>|wc{C1-8Fzqd9<0YW? zF^aZ1Z@@Zdf1X+2{Og=7sROo*{&X?LQJ`ta2!3Hr!L(`yxM3W4R5WDu(s{aflUB?o zw~C?H&(v#8jSxX;`hHd_z{S>Zc)K7G#lvSt?P;sbSog6uU+q8Ev6WE5xRCP(fMUhZ z^Xi>=)%Qk94NcItma>V{nMGT>h z23&6GNHrRuPgfv|VQds3zXq-WPW<+}`=GU?oUNN`OuJ5o-N~A>BYzj^8D*DuInM zuZ`U}DKJS}Q-n(zKV}Ar{6QDW)UnE0e1p-}P9B!c5r0oI79RGas|xIxs33Meowa z?z?rKt$@C@*%JEJQp;rwR-Ia@7!~)(WI%ofB&wwIApPwr$rLyB^_5fo_yl4%s1wa6 z0b#Y#S}>9E2#?%O-r=Ivxw+6*);Dvs=j}xc*pOMV0OtXhL`?L{ZSP|P18=5Z|ze3m0z)c+t1wjZN!0~ma`iRhW+G9N?><|oPCN(DEjT+>{Ca`f z6CCXL&gn<#aarY2T0AwpM*@@nCyKEn z`F_^q1*~e*&H((3p(@h+LV=_GtYb;PEGN{((XZd0w=x_#L8aljE{1 zfVqwNlp7i!oF2nBQP7@6lZWXtF-8bs1j9fM)8C)Z2ro!Nahc;a?(mK~3=81phWbo2Sl3acHw!3Q>ubs>8BwE4r!P=1RVk)yRFD zmV=b!Vd`3Y%Ij1N=9=#CMIa&#hweS;ck2>@7bEaOks8;(j1~&i+o?UrVh`w(;=m>HX{N;-`1fRrJI?TS)bgN%X`~j&bN3&^}Kp3W_ z&kP{uCmWW8AAXD}6mZ+K1Ce+a8Qka%r2{I&QY^KxXS7et|9)S{+VqRTT}nTD z!>%sRNvmz#nhA%Cmf7v$8j&jUK~G*`p9oBjqkdEjxKzm~=OiT8KNNrMY_K;_+_XKY z;*uLSOK)fx&hRmHkn8NJGuE%TSwrzjn2-ehOy7^4raqJ64X}dPadM1LR|d9w^k$e( zkH^T;Qh1Jgg?viaiSTAk*T0M`{#uCfJ0^1v*%Cd>4Xu|^>==6 zB0cS}s&P%tTIC7ySytip677QJ$~&VuU_Y20vh>?S*>+9^niLIhuxVQb+S=Q$V&IYK zGmb|hpo_Y^Ph-ztNP*nU&xv1C82%gxWlwJZ?gpCnhdEsQ>J%?(LY?hr#UTMjwwA_; zmBt3fc+HvM^cAnqi-ie$_{x>5Zlksd%741GWUA? zlq8Gs*pH#@Qy=Z!HSBW{f#4ipDPxKE(lWQM{a+Tn@YK;FPJypP+nepy;Ic&nwTz~J z1`umbI09O=gNV#sA2OZmbzo90jFo`@oD1;fRQbP)>McofZwc^qJ)D%{QJw-RBSraz zaK1NytlDP$4-d$;aZo9dQj#2ibaAA?yo2)+0qVL)5w>6}K|n3_&CMjWM5;Fs8!X&J z-atF4ZJUC0ciFS^w5q2Rug0Y3qHF-B#6n>cZt5beh-$QEz9P~Eim2@WWNZt#^3+Q- zEX5Lpj0UJI5X{2l;|7`1_nU zb%UB!E%e})R7fVf{v(t7?m^Ea@LFh3H8P?4dCw@!_XcXdAaD3Hs=?HIgz|5u_#bwS z>q|ekJYIVg+O?sBMKfH3WZ)e71v1&2XiW)@YT|1#`s8Y%ZHt2^r@~?H@*QYIbOs=3 z2^0>B?DnnI7MxDaSt96{MF%%;=yU-hsh9gzse;A?%0^-AEQnrY?$PI^vFQhMpOrmmiyU6bu99Lq=Se8y{RhrnFcZwOVv>+R+{EQ5K#B-Zzml=kqam}^p!x7kca zM{83ZR3KDh0YEI)%YJzk@l4^={gQQ{_h4F4pS~P-z*3BfRWE&`<#9HK{b`FYYnduD zt+a}@rrz9{v&S)y30?OrBhxrQFGq%KAt47LTkvOV&VEOW?G-u z1Wj~UtZsU79y9zI7%V0_nbv|WL7W2Lus@6RAO@%-ojw{s)*h$7_H{i{Xv$%$sgK=3 z(7XE*KRXfFSzbi7Ayl2RE%vGnb+uN{^bd7-Q)Mlx?Il5+CejqJSL& zg?U#vJ&()FOQ$jrC8+ztXP>Fz+p9hpa_VS@Z>F@~I8fM`N|1mo!t+KPq}(_UEfucf z@IgF`3tyr65!X851TU?%&*%MRr2j#BOfbP#2O!g`yQS_z&Da*8j?`ipGC`}fX$FqO zXIKuTdrcsVQ91UR0HVy=?WJhZORMqY+j%pyBI?BeeH-GmWwV5-e&bEq_TZvG{+cg3 zgCd`UB}vzGI(#q719P-Y)S-~~m)f&Xe1||j%}S5=PuIM&Isu(KOLhhPM_WGskFMWc zfbTDwbe99RC}zfdnH&BhBJDU`F8oP=HRYn^EdD~Do+@wv7u0%(nFSL*o-!ajLm(xy zx~Le9H({XMD{iU7YjBVHX%xej?iONuH;+{SE<%*|jE#(T_?qCjZl|Jv>VH|+I z;L;b3eoya(SgWM*t zQe2kh3MLj?`cfDvwz|OhjhTLMGlg>FTNC6Nm7IjaPH2LXB5!D~mP&Aiil-SmG;IuI z6T&$mywV*DrQW-rE-58g2>N#KPFL*B{M?vH4p#g}^R1;%O8<1^*(gWy)!JpIrDaZWrRpGuxYdn|@!iNiV*_5R@_ z5m1=Jkv2gP*lIGcbo=`=(+gF1V$AR!TRt1e|NLB|*nG6Z=>XPKvgDEz6tIqmAu(5w zVjI}Rhg(De>fDMUr1p+D&r6IEW<~>6C@5SlfOlWc(Ttjs zb||3wbm;fB*rLID(i#9|At7>r{69b&Yz+nfUVypiXb=jSV z?&x%U3J->`qB+2d#{Lyscqi6;UW$DYV5r^~Z>;%#-4D&LJGVgA%!KuDIJ#i+Lzo^@ z9G{24iuISEDFU_uM&i&Uf1iJn#yUqkd}b_u^=jKN7MAWQ*6EUp4d!oU-k(s#>~6RXE`GqAMC=%Mr{`E!FywvDE`7_{?0Y~5 z2O&0AW=2OgnRZ?7oq3`DhL(igv8DCon!JVqK3_^UHTL9N?(`l`*F}oka;boV_=O9- zb;+IEcB;Lzl1Kh@gM6}+`-b43zfhF%63v>x3eF=$JOYULbHoUOqJYnhyxQ+u<|A~o zf|PUG(%BWIum|@S_BtFZ7$HghXKQ|m7;=co`D98|CVuLh-Bjbv>vPE!HtvaPf34gQ zh&@fS^1XNMw{65`u#s=6f;h(eaRss8Qg?>)9|5r?B(N)y59x?6xH&#WM2}16W^8&D z&_nal4tc(%S-V;sZ&pts;wY*K$Z5x%ZKvFNTr$n+^IdEH2s*wgr=otxbUG{Heo#>U zlyG<=F!$vaJ6!UXJ^xlDhqp8jt)Ubc{*UE@elCg3qxOq8``y=j&ynlaHMd2bhi^A^ zeQdgBH`S|e#joG~)-F~bY?z>6MmljmAGU&RC%MDMB@}hGknFdG?jkYQtg|{x7z6r9 zM8e;($2IpRf1eQ7mN_%*YY}nJZPoQDauF;2o@xxbYBpE@f36;~FTrM*2b|Pv8;mgv z#D7Jd=f}>as8P6X_JqS)AX08Lk^#+5a={#gQw&u(15+A`8clEWaN3zEi!F{#A3EN}o1n zz1ToNB;al2fM#UlXoSc^iIUVx*BMpC`KYA+Ef0KaNl33_d1;6|pTfJdwS|s2Y$DdA zCS7T;HG{)sk&>3i)A_VEDS&V1MdQ9WbN${U9zvq0cY3isAv7!lcZ}|!Q|8w`l z-i@=|m8xFa10l?V{QVKf|%Fz__WJfGZz$Ar*WK3T2@`mn|D0BB;(q4Q|+4e z{wCG4Cw`z>RUmz7!XH1rMWyK>}Ww?dJjjmT8C{@c>gam==rVx*S*$n{kpWnTe_5e zvl`eX^TXVJJhu6L*H=Wm$Bq8#{)sZtD~1j^-?hz4=Al z+s3wsGPc+Z)=-h}s9q(1YbgsDoh!tc)fn1?MDznsHlD7I?oiz(GycrDQdohH^1m{y zyyYb}2flUv+d|k?%ax(nOdJ2`y9|PF(cFuuUn`t($owO z$uqI_Ncdd2@bHqd+P1~KeotpSP7P1r~d7hC?t5~V$E@unU@BV4iLwih4j5XAhDVo`GVW8%&_$^oVTlKDRy~U~A zQb)r5j#Of#)Ms;SQb%+A%Xh#0d{<;0c(QK=6p#ITi;G6aMqU{?F7{{acG4fIbKNem zw-kZ3Uwp7DhstB#ruqQ_Inx5HEuaIx3 z7gg5EV2p3Ju9>~sM0)RaU#mA92@RWTNOdkV_a->%kIL{8|0Nr`pc+wPb?L@Vl2h#6 zj-BvyDUW0P5M=7s+Kkibd#UFe;GrMS!#i$<`k47!dIu6Xt`rB9jQGm0DeE%i8tF#- z9ZGXy^9!HTPmD1Qai2H;y?|EN7=w?7;Jv4aXn>3G7LRssA zMTQS|Wfo83d5vq4=bmLrv>tpo*?NF*_%DS|8>kHA@-=)L^!PDX@@U=&en+)q_S4B) zTVL?;7mqij$=~-{zq;0?ojog~P&M2ltC*3ftFeXB4C{)H1f^tM@d3ZfyQfH4>UM66WUro}fkD|YFaQ5z;j9b_r)G)y)cgsc z0#*`Q-l6T=-QSVsCT&WL-CVqH>C>%WqC2z1PfqHDis)bHbEUY2wP16Ovs-SVd4m0p zepGVCV~Y9R{LK6wu2r1>N-w}3me9KDAZo$tef24_I~SH-%k<9Ou)=Ru0Dn2LY1kjZ z3`g`9(D}FgV3XW%mc|S*<3zZ!k$E%5KUn{^k^8oL?VtdG*laF61AY8YRKL`HYs&Aq zg>Ppq3*9w$1)m>$2Bv)7TgiP{^13|>ZTk%Ud{d1*T^-Uvld<2?Tdrmrdnd9L7za_)%ZB%C@k(zHr)J<2aZk@q%+|j?Yskq;sqLJ*>1e#3-8CV=-Z8u~qoeKASbamxhu`1fVye|umiX0_NhW=k2!L)Q7S1V=I^w^{j~ z-zmyrY*vd>Jktx`vdle+`U~m){!_});JkwY?zA}RU9wuPez>I_@5Zxk#HF#VXsz}KhO76A+`2UZA9TGpAC@yY$;S&Cle97bsB|Dn5ZTRWew@?34 zf3^{c|Br=C)>9g@st#MlrfGs_C98 zX{yz)gU{%MOzm6R8slHj}Xk9kfEK474l{L_wp;<^I zEiBt|ht^5z>+AFJ@m+>X*_4G<&r+G0nOVga3mE4_e)iYyN6vWW=Fe z!l#;kGn-goGyM8yxzl_NZd77GN5W61vfPcq_bC7A?(S~AP;GlJv4u{39C1jf+a80# zR1a4B#Fdw?RhC`J&?y)3t{!y1>v}p^Mo~?zkT{=GTr6zgdHbPnA)`YtkA06tVW&l5 zpC0;Z9{rljT%!hp{}S@X%Ma~kD_Qu#*YuCj$vJ_Qi^#yZf<_wVwJ#e*1l)S*4NjU$9}T&u%M+DgX6p3tzo`swF--(%^j!m zI|ggnq96URo3@I#@%+KY^Rsg4TE#ZG%t={U+$E)@CywjQOGKRh>JriV z)vcb0({HC`488Qh6%FD~?{WIkI+=pKMmAy`0!4_HL z2FnSqR%PpJ1pleR0W#;>r-!p`7x43f@zq6vsyw=t!IZXehpEL3Ls#L!W|M@XDXXv( z(YcSd=rdDA3a!!o(}+KnhhP$NOI^wts`~w9cX@0_-aO$qQ$0WXhF3Agyqp`Cu91IQ zHA91@B~jkT)AM4?&%2miev^B0_Lv{&iX&FEKf6uimf1KS?>*jDeYQr0BXlVrpTx?t0+NzNZJx};~;`~IpYWx^W+cdTdZtHiC zZq?|<>U*Q{q-%B>fE_f1jW{Go-7ND6|GMkaBi*#zzPSHICxm-J|vkye4w$+IeEtVIIB9c$Q8*SZT16M@#iq zZU@n|3w*;jTFB&}%KVN?xq=AFc~cYB+M;*4CE6 zx@y05uQj$>cYR)03f{o8kpIXx>!=~PudfuXOqb$PlA>fHd7qUpbf38<^#0GKC9DLK z=v1CEp5Z}^*FmjXPj8hsvKaHl?OsjJbAwvtqD6$=gl({M-loyU&B!01(j2**LM zrp~d8K`B9X!$pLOmBEfmG6qy4DNzSZvS7e%<%y8BUW5}q1(|lM@Twqk0pXS08@ z73{$>DYpH2B_$lZ)JRf(@m28!(3nWtCnT|}&*$2Eu&9x0^+iHve~-5rg@Mc6iY zNe7;k)Q{V(WXTC(ipae_!)@InFDMDI0w3|K^;pcq8n8%uDWMs|_yol%Y&dcV1^C)PHtc+dF0(FqCKNXmsNkp7@B{w?nUovex@&kfYn-oOO&GtK(&(Pl zV%yQ0-eOz*%C!Zb*#=Q=KhbF+Y_n>CL2W?MGCobeJzX6FUww;$1mkF=gHG8>AHnZT zX3_C|;!E#Z|D|OY_LBN$%bQfv_u=(X72plgcYFDvu~Nya+~E~9UY{bouAEtW(i0a0 zWi1!?Q<&EtZp#K%a-HSk`?`v`Fh6aq4i(jhljdW-?DI*Pp2nPerX7w`dAzSH-RWWR zYG$|&_s?$bM}TPrL%6iVM2z05q_r(|xUY3kq{O^VWR6QpV#6HhDKgDJ>d=Q#V*uYo ztkWf*b8>c8X4_lRmZpXUH^`I}z0P)uTVA2q#dAaM%g7z_MgK5uD?}LGlVgA(D3p zz}g^If^3a^4wL&Tu~pbSE`&q{j~**0klGo@0eS zg2C>dnu7Dyg-5GcL$wa1sb1kiHg(>ncYj)DHh-i%TP4x|BDvo!0rNu}*JF-FPGAbAX#Y`6^R zdjMpj&);g~-jR11;f?kpxDuQdIcSwf%~k3`IDHgc@ya}guY=e_>r!kxAM^rhBoh7I z&jT0<6){Xpy>Z&QjpMwhLd6=BWVp@f3<-!FxuZfJu z7Y&DK##UBJU>g1g42Hv}W%(hk58>905;02CsC096nKIM$$!!!D1t<|Lvg@;XZN7Nr zStvLG+KTO$ zn$%!6^0T7UsT8;k@2DT&ob87ik#py8Evl6ApXp@;!Y<(+&|7*ox-H~3H8uSu!g_F| zF`79k!bavzgfK7Wr}(I4qDID zvSH)=#lu8cZD~1f{tP$p9_`Tl@Zlvjc7QxJP_s&!RZcRvC?zFDQ4D&^qGMkjf2iv- zo_^B8B-eM79I80)aWTb=JNmajO+o@g+tRqmH*(^NKw>83xRPid_tKP7s)lL&MlR0&RGj-hlHpLMlNr`}2_#v{D-2}jg50V=v5c~li2vook zeOGjKb@5+r7yBjq;w7$3u@Ygt4KO

      pA2i*O-aZn-#(fs+6qbvY9NGP4OP;i@;k^ z?C9O8U>RaHw;KFV!mj)WI`$FtlXO@F;N6jZi2nxgWiNmxVyQ) zk#(MQI*!LMG9!R&mf8y|Xb?7}nEW+)3PgLu+xRFmGqd#+|9iCdbzG%yi1|OFxZ^Ty z8||}~ULgJ})b;_<4~!Th>C#ceB^5GpRqxK)&Kz4k%YEWjkhDPxwmwEBxSdjva74H+ zbP*wJ!=U7g$jcMB)U-^Q$X|U+A=kh`>gh}Q%G$C}T{mI095-_eq+vc|6B9l#I1V_Y zqW}gw>ctl2365Jhiu{LJX7R~Pz?jLZ;BL=Z0q(Y5=rVE^`?8TEj<4!>!mDYGXg%Gk zQ06hOez%Tin{>NJ`Z&})CQLTSG=7%Sz!s`9lVKZSF0ew!$!x4rQfrW=RXKlV2|l&; zNNcis?w#t;in)ZbsVVtLC0>ZD2aO^zsvPP$EEpLZ_`?d&BLtw7lUmay)=7=lHAjr9 zAi9N$um-x##W7d=3idw{iA#JFF38;gz^U`^+m}8ip_|4^x&9Kf>_SnUunFbRJ(DSI z|7MQdTPSS3#Rjvu7uCbxWsHO5xLwC(dCc0=L2~2mnpVCOnWbEqo%|#I?yLG*c1@Cs zTq}Qcj38GXfoWIIb|KBQo~i}lQ9~MyW?md{cLyKU0hEGK)ZIIP3)wL;iM=kf^~9IV zADD30tZ_zI{6prFvzoa-;|IzuHvf#CCl*@Yjb|&rta;qQO zQK++CUt3ISNH*!vb$9J`TbWAXP2S%qzG5B$;T(k$5GiEIRl=hY2eZl-H1I(gnHm{2 z!D>23a_`j3M4mG*n|XbPIIiF>qT@84(0c0WZ&*eLIKX5!vh@hRMMH$Jbv2FJ<@_+R zfy8Hm-hA=6VH}6YC2LP88SP0{@?DzQA(mY+8TNYX|9?mzo2GLj6tk%?KuCIUF1AFI z`4X>1hKcCkTpAIqJOb&})yQ9KF!uH}KhSCYOfLY914^?|AhIHG?i71`t%~WxbW{`E z7E4=?yu7BSrUo!sxT&ozScEkN!8t+5HQ;~#5nJp%IIvD$+*S+T91rFCw%Wz^l%qfeR{CPIX|Vr6B8U-b>))TD*p6}Qzvx=PQQnwnN& zO>GEkF7J2izG|(M1`g{Hqo@EP!8$?X^s7Ag^Vkf%GB!2!o}cKja?yPYW`M>CFrMP& zCw(o8-aj}j;pXB3)-3-ixB6^RSaNWlE-ov{BSmA>s42&8G`8Z&-0`G?w+5xJ=&0ULN1pBE1Xofq6P$H9K}oth8A@$L`FzkY}U)d#ozO!otG?3m%IFvMKo8ogHDdI4o>qK%q0{ znxwG5?qWOyM1Yspbba)H0eBGsfLB3~#zAWC^OV=+8x^x|mS$s} zI&A8`ieuKF8LCa;g@!c{6ong!dB}8>aJSJ}FWPm%3zFD+w+N+_L&$0o7AL~Et6-FI zT3pn>62liEiK%H*(`6i3su2vZH@E2P>q8mjWl;ZSzetCE;gfPtem;0o&J)@Ku)tWt zT*0T(a_cb6j@0`&6UqgnAlpt>ZHcn09P~ajs7-KEE1QiZ$aR0w6H%@EmpfDE=439` ztgG|pKH2fhj(dODJifQihfpaFwvHZI zFKyi@wpaxSx#xCqcv@yA7fSsPpLhK24!x2_zNrcWrd2WfxA*_N5x4=zBqBm9InK{} zZJ~#|#BRL#JtIUHFNCdl+*X$yN1hsqoOidS*w_g_q2n!ktL>(XM{8h&q0$cW^76>s zC~VjRYUci!E@V)U>$6N3;ryOJFnM^#%1_(Ai;^L;VD}EcN8bgSXxT(y6fnPe{=@Sf zjAdKJ7mEFOv|2{?A2`#FnZ(-8r`A_Xms+ikhA7X`zD!*0f=~RCe~6GapwY!7I-frF zR*n$&9n~|J3`?%6_YQ*9|KumExI3*ZW<45URzFB1OwGGlt@GgM<;~?M>nqu|^A3^% zC-izs{MMfSmugJTPXwR*MA&wOcCG*R%LmmvsL#KB;gd6*`rWfPT5hCaBE(sma)%G^ zURV>W>98n|2_IO!Gx_(+eT@A)=AC~pO*>T5E1V}@Mu@2C?c2BS8N>B}q2)1Yj+ZMr zB?iQcrIzFS`;vr=)KqV7H=+-5RAOB?vvj&d2&ksxuy)2$9hurV^*TCD*p%IBAN=HboEIjGAR_aQZ^Erj&$n@vc`49VT(tTapUxz+ zn7!ExU}WAJ;a9%k_}_Q>_VGg)N^ zAXsWQIV+a}yc%qhV&-(RGBh-#&=W{xP%B_`U0L8XU5BiJ5*iVUy$_T*)?Vszx^}66 z8#q+D^1@CaK8_+Yw%2x57l@P?bf05YlGF`b&CIXvWgYsLF3(-9}5y=^bhX_FwE+waQO6*Lu zT!__@Q}J`XzrEZ`8P+^zs+*sy=-nShAZi3FI5i0Ff)H9!=tQVfQexU2^2!?Ub1JP3 zLH(d{MNoJ0=`RJ`Uj_O`hZw-vn3$Ncwe`NYF_$RpV_H^)2-LNFy2XG?KxC_Tx4Joy zg4FuMuY)Bg==qGV_d;pNdORWYN|sK!cl!!8G}ei)bT!sQudj{@LDKHi-PhmGg#IS3 zxVSjE{4=p_zPMmQB6o6S+cq}44k^fLRyjl&`;8P#GN*5OACr){1N1tiN>zgIl%oRCcwMh0d8HL-=Ra_SLy(_PNXn8 z2m#L#1}cpsQPu+NTcxqD%wXB!?M!`t|Iu4>p- za~2dsw!=Ql`V$QFNT6k(bAv~ zj@o_clA?$^z(-X7iv@fsIX+%H7BtuY;5DIRJNF#E9#o-6VO^nECOIyGSV*85*=0IHkm90%N6fHtJ);BuV#t+~ zlheAEHM9zZ8eX=0QxY53R}E1;sx>8h8CULd=c|aHMc_$EhUiwITj+iGZ+UD*(xMo+ zzOnP<#$*?{u*zdPTD?*8LDV-qJqZfteh>Z99x=7`Z`GJC@NnpMWJ~n?&3e?dkNZDV z=25BXyQu!Q#$o;ap^(T%NJ&fcnaHXXN8xTro4F0}5EEKlRNf&1$Yv#!E_9pWp6L5| zF^M^`5RN+oQX@jJ$z#_>)j}=%E<;6ib@iYMP8E3I3CcO<@NH)@E0*y?t`LfRQNagA zpdsy3$VKruO!vmx@$>OjWfb-wr9>KCDN>DK#eNzZ8tX2@P*#GeiHR4bxCb&~A)I~y zZYl>C9xJ2*cl@6BlkzVbPYl-^dmDpv715S1v*81v=!jXpS@hd&#S*2` z)tNf|mk;J7+LocF&3#RrG&Cf}2)P`W^?aKeR>yftC8z@NOh4V9T;H1B<}^31uNs~) z(Ga6oRE)&*7E@`{f{z{*PS5qd|0$ z;2((z+^(nYDkv)}BbO}Q-_k^e+A8gE8SHAI6fgjhvJzy*MF&XN7@$X@)>%VS6IVlT zCD$L=jd)}QvH2S_QHT~I`dYc1fN`W{WcUuOO*`;FCO4$rlBD4Okl9V>?w8&B^=`j` zjMMwNvd>K%p`XXq))O1>3!hxqg`@%(ZIj-(mQC+|I){ns`7TZ1mwFv>zO~M%47~(w zY-kxgtX`t-=j365w_Qw{q%qa zxq+yxTKG2_g^11>rW_)afw1+00}**Apy|~xRGW6ku6=Udff|yMl9ciXYIZSE0PaM~ zeQjP^RHPZ@cThl<>$hEc{}2nl9prHns2W-|ox6tM+)qxK0DA#l6K74NWKb6cuy`j; zq+DHHSHYgVxwK1Irld2B8l#91BSccu(gGI{nal*$-1sl=D7Y{_p{=$*)a~5V=yCTO z%5()-f{taq)20Pb*=f{u(Bwoxvr@U4t-1mg)oVG5I~YLpXeIQK1lpsn2pTo5{UjH0 z6NPM-;WN5kjjq$S(xSjrpen@G?qBUG8UU$hvmYw?>=Pb(9M4$YT<*NM)RyH&!_Utz z?TpkOHO>Us+Xc~EmGRt_Le39dokG8|`OAmY{#kHqkS{P9LyRe_%N>>1`cVkGQt7r9 zuz2VtbaD)q%{IvP+d?#BA$)G=o3CUBV(z65o;i`46P2x~cFn@8BcuT393Ou~5;k%X zPtOP4TJ5kO?3^-Mb8;b4Pt`1)$UI!n0NEb7JQJyGZAj^|!NagjA>dmEthcl%dqqe{ z$P$pLkMrWJ1}d!E)s0xE*b7iv3Ca#{LZ=f-jpG>^Z+GXJphz(!hLmV-vWxew0&T_G zlob}3f$%(_agIpdjTGrd#UW5C*=|ZL9hkFmkexjc*fl&@hQK@=l#%@b-3gd{nI*pt zIfobiu##>G^}*h^OvFfR2ezcW+3~|x<;Ct~U06;fcao-fY+|Y%;r91bGRl-iHbO;6 z$iUd94vV`uM%Xn=ON-`Q*VLp3M*r^6EXjZof&yX*3`hnzV2iwgtYV#7YVvnG1fW&6 z`>*@lzz$f0F0)M`;1&y$H^E9_AO~Y5P+CZ&7%||y(vjI-6Ps@x67v^dHlFeGvxkBq zOMMnNGDL2`x7t7wA-PUo3$%}S`!Wdgn7JO_Sk1jJ!z=ZG+wr#%H5avJ%(vmxXD#X# znUClHG)}HF6#I^f{SvFo3h(pw@sDj+EK^JUixqS9%#8vB}D|};Sl^OC@6?$ zJk$ZPJdDqMZmfyNvbQLnQGFTB6KXfLGivEK_XD$aHDp~8R2e195oM#3xsBKQ|H?;> zS-LXh_;J)|7EfZdsP?8oS=Y%aA<|GF(h-=Q637&#L;?2a;Y}W^{<5!`>)lTM($`B1 zPq@?@JCl2i@@nEqKt^rhfXLT5nEUB$mkA%$m*TR>UEs0;_}oG4zKG>9k!nZfrLmqu ztDIHno)J7L9~Q(BJs1Z4GH;&noG^jIJQvI3N$L|2(j)GUJe3x;tXx}&CoxwengIw@ znQmgOg4;QSr{WnCo&m+=CqZ6{XGFb6G-p&H_W9!p(TR+hh&3y!{aM)n2J zuyhDe<*Y*U7wSXGaoIGz&#uqPTx@TM*UWKFVZUF)#*f@=Z;hJYYT+K^3&}cTAa?Ai z>4rzTZZbgH^LH?iO4rUq9xpA5a@crM!ZS!`Nzt}|OQmN|MWL)vdKQS40o3{vU?21c zYT*J{`G-(q6S_c3PJczwO-oXGs7ulC#81}Kf%S^)2J1gx z+5F(xo{)?>vQES%)JR59V|5)3D4rB|#n&4(Uztdy2rZAr>A!vUYwZl1LKFk=ZP!8+ ztf9r>B;rrw8Bv|Nc~$8zfDcgfyxNWIGoGY52(iWAq$Ls9UWiqX7=+K6NS!(&VrM*A zGLb1Qib@=Om2s#JAibV}05byp>TaiwoHMwJ2*cVLdU|@4W66em8At?$l6!7c{jlo> z!ctLB25241vyV+|1=H&b5Q#Z-)m}MW;cFeqiK3zkitz@R{=uE*r~9Zqs@V@(gZNF{ zPNk%$KMGV37gl!z#;X@RZ|%$}KZIb~vvi|5+);BY(qeVD6k%XbVu7Nnoq_yc0@XAE zhON9%Yt4hbChk`WkYg)^>3UPb4JoT)8RX(;1IZ&>y#%N(piCxdQJGpW^aI6OC04Qi z%o5`20-MY6+4Z_x#ow|j3_N+K0VQFYTIzX{vck^)9LIy$h>5E+#hcgXh$?@1GKxK+iG`xlRaOU<{R>ULZuVgrGA zA5D(5?>>ddC<9WX#JUTIPMMIp_-E7Ed~sdJrdUhM@fK!oDy+`w^+S!qV}qp%>Hh5n z&E!eT%}=o#*0l%MtO@STPjJpUKy|%@f&tg(pYNb%sh1deq8}TpHz6|B(b`%CR9Pg4 zJ2X(Dc0>bG{5;^YsyvgHfZ+~63HzlO{^F>-!D#!22h=lgcpd6_G9l%~P};!(Dl}{M zYlJoeMQ!a!G(~77hI}+|SP)z)pu0eh$`j{7^_yB-gE-s?<%@nV?(Jf$<5odFDbTPG zx~DHuiv@9P?NHtvKIO_TG0%?($!2mHQ7tqN+vaXo`w8U)Wq?Hj2kGhgnekl)2UdmJ zA2=icz@cc%Q5nSOI@csU%k1*uMLa2f7J5b{V;)b3i-3YUqA&((iSNoaVz7C2tac^| z>RzaMC!|ssHKao0*VS%H5fj!K4`ltG!)Jdn!A1z%28_7hNLhC!K_%-6A4mwA;<6#t zKUwFX{n<3OUaTst%x8HkFTO$6<~FNmCa2jM^2tTA%hrW`X+q~AY0n1M!+-?Z?*|zd zYaO>C2-gR9u+*=!BXEdRFK0C@EDSX~=d3EJLgtLZeaL9E5k6OP;(I z@q}+3I#a;p@k(S89hcT&$bA$-6pq$~a{EIUF~7RorAwC}E`!66?E@T1FXl7{K^xoz zTU|fHsmce(kM9#1dzPGE9;7)NDm zLIv09VB4+A&u5dkE|Pl2)_OQ8HBQ#BT~@rehd?*TrX047_ck=z;Dlr)FE4VbX92QS zM4-#l5qfR{4b^VksJi`-F@&R%0rl>hkO4RU#TV@-2qgt$D=UQ%2>>C<=^XE)S!6k& zo(tWS5ghBNhaE^<15!91^@oL30yTKDo*M(ivh-{yr)CJ>dK_v8Ze9h9avqeh2guks z-AcFG8Ev}H#=rG~+2H0S!&Pi&v)kai@?3$cOYY$2wn8!cnT_;dZ+`KsVH@>En8kH8 zivoMAGUAmoNY!d*76B!60=E$`9D47&9XnBFA#hj%x{GRar1b}@ygtr1_9j?Y#He?O@F9Btp3HX|VO{6&w=~t425r$$G%NM83FlpT8415v$YqiP#CL(Y5 zcYZiUY&-S|{bJ%)ECf}nK*L5ztq`|uK(b4R*zX9fDuKhu3!`(+*_%9PTj7 z+8n`ZNY^9yDnZ;$mG0{RA2JIH1Wd@l@&%4qrG)A@or3)wfy4exVj>XAVCl8u%eY{% z{h=rA+}AvK#c3O$c8;|v>PF57P#6ETp!;<`ygXtPD+f3m zl{Z*+r~CSvK9E+j0hN%Yv(|F51qTO9&xS*Vdc^8f#}Cy!&?$>j{SZRQmO_2@%hFlb z64ffnT?s1@Lw+CTL@kIbYMR@Pjqtjs{@#LvTj`s89Hj3L{)D@(^n&^ zZXG(ZjzG7#`6MJ<%p{lG9?Me9+vXV*UgJq9!6k=;0Um628BJbKD=86y94&Ce>=qUl zz6Sp8u1vm76}r=rnhpgp!^Xm@H^UalR5pG*0t(YpD0xXeB`;AK_# zus_6l8)2)%Cyqm7_9XDBJa=b?SSX=AFF#)rxby%Qkp>iGM<6PKnbg%n(sV!6uGDNJ5Spb}X!{yJf(hAz-i=p|!>vbv;29AGHx@-+J)}sM~tzQ{axxUe_Al zh7J+w*)*ULMy$Nr+uI$7gVi9C0}SCRoebcp>_xXkv?Pl&lxT>+_F#cpjbiyq2zHm~ zuN2X7%;M~KlUR9p%B!Qu=XV-R|;M#Vf9lHDU{JF7yb#!(3fFB|esP!B=G zAmH(MV@u0x0qb)u3Ru)}$I`{D0=1mqckHTp{`(I7-XiODLI+Ss%vKeoJBT@fP)U<( z)aVD|!$Lpv`uHL>y+C?rplky~@4yi-c9!~q1Bhb7OtiDs4M1jbJapD~$o;UGlU?&F zqG#+O?fP0aN5c2B(y1|M9K=fe&DFhIJ&SZJS2Yhy!<|<}y_nJ9wAExC6hAy6gBywN z&yUilrZRtu2yXLw9Uvg$Nq{dMnpbUjo#w{kNyOW03IBK0CI$r zgAoF$LfAl>S{5V}did59=~->M)m-?qD{x~_5Ejb!w_FqDl>Rb~ek=T_tKwV*zX)?B zSu?CrcKW}9Wz9!-?M;8Qw(Q&r2Lo*UtSW4>P7qiqylW~-hBVW_${eBow*e=COof;d z2VGRW6rh;oA~t>)6&?ddkPkI+6P6)4fm8h8) zci75y$VrCs3D&w>7)7Pq;oJ^8J=Ey}Rr`~51F*d@KD1{GEV&6-t$>kAsN+b_);7R4fWVQ(@fJ34Ls#$my&j!K-3JDw;nTyyrm$5I zH7c;yeVGOldFW!uPa@KJ=1PL(`LaEnoPj`^}DyCpKpm7EJpEUX&;A$}vDpi7OnO$Zqd7rVABst}GpPQSoSAtoL^o4Ohwk!a zv`tZZcFQ4bjt^Zu8g5^8m0UHhY$kJzKO|Qin~pBuo@XkJ2oxBJ9kbcrlzTE1kjgVT z6$3GMZ7!LQC=gQ}-+VJ0c?lA_NBy(0F zv~j8nbUj6WAK}?fYb+z=S(#826f>!q_=9@nE5VJO21`(-i9UMc>HH&K5qBHgvRsjs zT|%2h{R=5idY-JU@rjuoPd&YXoo-ODlPbba&8rkKa;v?3uva!r0`}6ylgx`z3zxHV ziPaG5;Hu53Q|Ahmb?W7W?T77|+Tk40=4c0j2^n=I(9k5ztBwIVs1QbAn15jpNCg4g zuZH5z8&FN>A{offWpeiII+$tVyEJ`pA}*u+@u%h{_8)E4@p*lAM zexI$b0hAHwc~C}kDBDQS8r6pd^F?zmLZt#+B%q|gkAtai+GPjIGeb!_w~PB*VY zqKI5*N~2=wWR5K8R~u-|*g&Ow&MK;j!DiC0!MI4-OBP&i_P=4;c?C?nV;71R656*F zE=Iz7ej+xk+)p>Edd{@rMA9ivBo3-FS757a&MG93Qc%!?)h_6>?~z=b?h71lg#LJS z0p@z7)O_yHx^XQ`x8cE_BLRE^)*Z#D3pLxgDK?%_AL}O{R>`Uq9f!)&@f*{w%|XEf zto&;viJkZ^!c7`BzgH-2%M5Ejki}(E_X&C$;HTQ|8*>|!AJ_ZM-<#t2-UR>W!cH=5 z*8|AAA;r>|OAk-h=@gi&b7&TDsEcb_Li?-qETaQ6*Dw#=ezff}o^+@KiVJ`7NkD4w zghwCx@iIq7MjV0t3LLfq#MT7P9hQn~Uf2`@|5q(-v{%8->)1T>khmo8o!UQe*Tt1D zKje7}!=`2aa>Dj#PcAAC+?zN0S8@j1iZ^+*W{?KxNB=dAtbt9kACtCj;Zy`?~FJ2J)Zuvk((|@Vh@2WtDQRmTm za9}26DAYmD-8)YhkT3x+Fsnb=nJq2ac^(pXZ|EvDAgwLW`%p-}VM`#x{awVFp?VG> zGg|pn6gWdpsB{I`qZUl**+SH888}kP)ma3o4okg*-=?*1xG;~I*|FZe*(JqqW1jZh z7#(c6=Q@cJulRt$wbh84g~QD+sQ*Jm#qV3Tk9rHvHFTHCuC9{uq6gL#dr8X=gdE>J zSPEJc*|;0*Z@a(Y?T#OR&^o#ye~0gP7~F#Gx6-rV&9)}yhVKZi4_)T(Le6U>wA9<% zo6-~u%^%V;p|B?jfCB8r4rPfRL|TGN47Y?F7-(sfQFUjtjxctcvi;2@YQCqT-$QRiP5bTH^HI~2?+lRFcmF2 z12ZrIA5<;-D!*c$^4eQV%lfLko$R9GYvwNR`TgrW*$^%-wr_3SeU>~TaL2vz&9i0i z&E)qc#eoZeeo*=f>JUK!6@wG%ZboMzpv1!#v_^%n#u3@IussYyXHDFU9xoS#BWrPK zUWSct2mH^^T%CzRMY6ltkgKkKGOnw^l*IxvAYd3OVj|Gm_SaWPuCd-_E9LzX!p}rd zEw>eVr6%UKyTqn+Xp%$)0;8&TOdB+1MB0xY$DLut<6yV2I(eVyfR0kO zI(OtT408=(^Uq^&fQwo4Q$2zv6}U_nDCufiLkfA4(1o%<*y~X1TyPn0djpl@MjUoR ze&Gsf68om6es+BqCuBOBDrA#HNr_k+3#uk*#?fOlKmH^bOs6hD8RK%*yX6g%M-A$5 z9vStp(GL(18UskKZa5a2S88X%!@~3c{nXBMx%kj>%7j6!qKB=Xum?h9X*?;VF(W?P zX&z76GlSr8DS2&RT@1vF_GeIBs2lWN#yKxe*UpIgw2mgXuJ79#$>xYC*GNrB7Isso z`%RUGOuyW%*?qjhsZ1E+jJm?+A9xIXGDtN_jI_->N#$)$p^*xqS@SK$hT6M-Wl?(>0teC1*D!OKm>8XrYMsjzQWUQLjLfWaq69S~NU z3aKzWN3AvPIHgm{mjAo!1{tHYk2^bQj>P%}y?!^q4$m(YZCR`fM9l zlk*?8k`kpSJ)KeWe|8pd(e;%Mp`u$*adQRlk!J@r;9!mBRX8LAEiSM^ft!olz@v5) z9o2!5J{cVffs9O?5rjI!}b`j{K=z<_0A0$N_P#mDtDPX&WKP3{n@S!6ZP@TqY zxh47v5Tc4A5X-e-^Nk{Out0y*FoVibt*=mt!v-A>UN61CmKp%~A)ia)oPL_W4Y|L4 zAHqiyoAMt^-y^*ntIzh%`}=3FV8?6#UnS@t8f%o8^T~Q$)Jp_1wHREN|Nf}GUiqJtw)WI_g>#*>{jHio5*@A!wO za2y8Y53C^Sk;U;wljK$x3M+l#jo1ZdblL(`bZGN!WfRCH7RJU`!v@qr(o&L=C52do zJ1)jN{)&(9T)r`6W8L*)lkm%fij9Akf4e|*%O)#(VU))~#iW;-w@9jF6rTI4Nv(E> zN-__)40ISw$N;ompjPP_(ZKD})P@N)uTpe@Ag9FafQl7ijgx52Bjr|g92J%SG!>m)* z^8*^&s@IHnuIW3sqcf9#bCpEx7uTMhMrnT_C|^7Yh3_yV=+Z4g<1E813Z&spRC$50 z#3c&g*9$UxUerztr_*5Whyr#$04bcs4K+2Rg4?#m0_=?FPRmYkzREdhhkg1rL5sNF zxvJt~eN&~E*Y;AovUsbNj=ArSU8>bOu4i&TVHM`C-Wn9E47v243sAbPBQK~R3j8Fh z{Dciap#|K6G)gW;1O)_6)*H1XxN#PA-3f=cy>OYfPXSJ$9mmAb%2igB{Le?A6qDj4 zz^9}qnTNFM7G~}hi2PcuA=9l)7gP2@N0Z~G_A95?yG*$+yNnX&iKE%yX$x}6!Hn(~ERXKkhplH}fAtv7GpT!sTBWW;4d#Rzbo zi}I{yZK_I|(kvE>Ri0HVImmXcYG8eB01NSF5Xz~ZC#3uiX^%@~77cnYwMeYBpjh9e z!u1O7B_}5rKdg2-813_{v*L&5sT5NCmvPE3vbs{I#24_HOZU02N#?8Cg^$`#nsbWV z5#&}0S<|OWwg|fOR<^tI4-0Fq%8OiD8;(6gMa3DLG4?&xncB|Io2F!ZTN(^m8jg49 z@9!6KUNG-IjKksJXs4$kjm-v)rD$jJNMS%lMMdi>kw|P^RXr;zN{_%`I>1-xJYt0H zR6G(A%(=O_JYr&u;B4W13NPk0GXn#beazwv52=~`fcvA-UKgI~D-{%nW)gwnf#*)? z3UFu?=8Cw5+hoiV)(UC#j@@M5^r#CJ6L)oK< zjB1zsG1~o?cKq)QDXXC=54)u!#GC~<-CbPH1A7t*(~A&y72$UWk~W2u5f7&s@uUCa zucX;SO^prjpG%B2;|Le5Ao>zxs9o7bL`1wZXUYt+6c&Dr-BJ+@-oDwET}Zfj{B7Z)d~c9S+TGLk4-IV)-ny{wY3 z%e@H`B{-q1pUu4~mCvtX+Rh1wm`k(IICXnl+E^OhIpKz@^9P3ww6p|<4Va>%1vXS$9ZqWT?ds}k<#|(@FoDO* z)JG$USp7g5V39HEfTV076D%K>msBe@?ppZ+R16eIU-$E|ZNAF;;|1BKcbQw%C(~MI zky+PM!qqr)Y|CN)_Tm9HRZ*-%8a0E`433|ySJJe|xe(rzan%&C%4(vRBO!Ob%}Q5x77U@bP#r|LajFkGI99r_O3fPwnEjx?)l)@w9Q*S4-gYqJyUYgu29l^a-RMNs#|wCW*d3G6N;vc6zxe} zRSk44OH4{qo>lK{ZM|VyA?$GbP4S?wpC7IE`@46$>U3zCm!9u}O^BK7C9iXGuJDIM zMMbr`nEk8N%jd6KR8l^Me8BS`Y{aOc>)ITwO?;+B+xlIV!Fhhg)!m1q>$La>I|xyJ zmCK~4G0R1pt$v@dmw&ai#h$6UNsqENyJg@ef3lo*mI5Qda(N7BnbmkK=|_*&F1I#`D;w=?5y+0Lq&B`T4<` zw)97>#Dzb7`V^!EEhE8URt?%uCWQq>MZ?1eCTp-*eRK2lv7SXV&&4+2|(5f`~K`V?#thqzi&> z1Vx$@0jWxr9s~xIGDJ6#jtYo0F(6Hv^bW>QM4I#_V1_Qu5$R?Aa|Q=>Kk=Vyugsz& z^S;mfJg3~}ocq=@DvcZNT_H8uxfTiN?}VoM)z{E5*8+hFJJvV0V~f8LJO&NTzBzxU zsi`S(p0+?K2N#`dYijVI&>o5X#4`s6hew0q_Cun_k83;pet?yg)eFbiLCCvQ4EyNk z0;T5zSy)-Q3p9STRA1NBXsejT?r?h5&dL!S9Ujp&66!hYe(m_q$Jw781@X-R!WGNB zbaP!+_w7S}fB{vfB#s|fozhW+{wj-=ni>TQDIdWhHu;%|@W@ETT3h0(ZP^1nU=--_ zq0z00_Bb0Yk@;5)m!L#?E#kQ2<6(>H3bq!lvHzLv_^>V>8jb^<=avTL*%v!n7X3B* zsLn@c=ez04_N#!uFU9~2YrcebaXWYJRGsST>H=Sh`#dU2xz^0C#bP2veDKxpRZ{w! zHf7m%XyFVubi)f{oi5Fx0f0rij*tSJV$(BFJfmrvuL-2T2UXzq$l(UFb1qhOX(j2nk+o)nVx9G$J1Gweg(o#Lum3Ay(j(exx$PJUpO>YE-w2qH5es*Q_AI@h@pBR`MR(4|2>noD2 zZzX?c5J|BzqyIJAi>+}McEp@df`y6+*nPAIEFY_C;hw}DVo&|aG|G0TsI+ucSkyAB zh48eP{$O`L?dkq3lcM2+fxP}~L0dWJasy{Mgl-MN_l_)1S@;zko%ymMd3x^GzVkXu z0c}}$T@%->q@<{4Cl=`smhk#NKSiZ7>46Nx9}C5Q?Y2`mX zLjgW8uxYES6{mhUHsDTioyh7VeH8ZKfxj}ebx>ebLwq+hf`P5|ayzx|A=U8ST9PdI z$HG(;CrLnR$w`pX&Ry?BznX^vyI3uKT`vByp=j+=)Kdoab?;T*o8F{ycwi;(;G%10 z-tWKv-n=wce?BR(k_{NN8Y{2JLhvXcBBJ5oS3sLR#{ODblG$)!XYjJqGd9P|5XLO7 z;i>e;LVJIx@;)ns?BC`fqw`(qprM5nxozTsf9Uu159x8%L+8JZ{3xmKsy|%VW>1@_r!u5-!!? zSs`bK!Ue$7nW`nf7h=OhOI_eE z!5P+yIK}X2h(bRwx;}2PP*75;fOtefQL!9Y%M2m1_vet=NFYL+@6NI6P<4njONS6;%q?ho%`JWQz^@1eU*)zTIcmOoKVKy#q2my4UpVj} zbS|y*y#MCn{Q6n?hV&T9IG<9^81F-JT6*cRXh!fDXrl#xk96ybpCoDclsa1<1gj$~ z4ck)8J~h3$&pFho-rB+8{TZrzr9VeHjGnX?yEu)#$`>R~ykdp$S^mYy@UZ7^zy0Rv ziKCX}1N_&jGFKL{Gh$}4mveDuoJMIwftC#ZUg?kvudTtlp1L% z++n%yEuIcZJzQBzx@J~1G!Yt|6w$Ob&JrMaQ?fMGN;614HhPr4Bu}3C`nMU1%w8Cl zKh`|yHbQv%&XNZMa8q$jE5k$5)7zEhGM2-yt>8jk7UWF8wbyp0}Tu@{G@ft0{LUS3}Raap?kHCvalby?jE zj!L@0edopRGAS>;en-!_`g-#0Cx}H{`fhD_LA%b|E!)I+=)C|O8%}ZrSLavTq-daH z?WhFOFASN+?9Zo&RD{UZM(?%^y4H3dxnphJMs@n2QgW3>VhaiLvjW#imtG$ooBY_H zS4>SOHzlk1l`qgqAL!q|`Yx2w5*eGfd_hlzkEy%=)nF3C^vQX`% znPS0@G4Q4D#gAbO?4Z6^WH#6vj8xXc)gYJ_|6-c3q#-CO8WLuZsIEG7QFkVydgZ5V zHhX_gWX)UZnFR9gH)>Jd$c51Ok6P4%`dC+P+U<1Ga|q)yW>wa~KpwIdxhS70`t=v1 z($?Wk01f!uI^jd7x%XIL*`2gg@8=DR)hgT}s~1&KV$brh+<{1p8?3LLPcQ%dxs*Qi zYJyWu1$x!hL}8co1XG%`%qys9rvnOI5TKXTLKY=Wi5j`O7EKsYX&Y^w+~stZ<~h4c zKfQ)7eTTrQ#gw$#GBjBY`{_?G#rk^5&dN$lnKsWqS5#>^>;+ni-3Sy+VbZuny%uo7 zot&H?@*M?|xLVO6K7|>|3w@)ibi2Naxf(MmMqBr(@KLE5@Fzd7_v6~}0K3&H`)4}i;=bLH)=KBUCyu3V`uU8+Vds*AIB`(U#} zig9rrXvg$j@1#`#b;3I*+A^XoS6RK!XluXJMtXmqN3gYXvWCO&P;X&7#H{xX;fe~8 zmEhP6u+;xuCd3|KnS8}74SQg?v=aIywP)?HnINI%4OQj|nuc|rm}J*2gC5o{P1qn1 zZeP8`H^C9gUqJUd1l~8#wZO51hGw$?UjfLMMvomMyh+H+6ci@Wa^V(miLrz5xB#Zf z8xs$~SP_f_?xv`PxmbgRSm%0|9^G!pQGFoCCcdB zEFBmYoe1?ZJILNw9!TFoS@$ZHMJwF6LpBS|DM}JfHf2;35@HYpSfXW|8iJT$_O3px zj3M0_vxF7}W#xFZ3vY%vCY{eLz2m1|)vGLmV{d3Dvy1}+0~6=<^h!=y9^eE{l)m6JSCj6FMd&Qy#98?63ZEzZ zSA;qizx`jBz&0yO?i*yQBv-Z?!oUg(mxCHgXj;xbFNP)xF zF?>ccT2X9%>-g}e`PRd{m6N59lVHV!HT~wYOC$X=Ge61DUw-?)V4GkX?cvp%Nk2BE-2~MeGOz75d3cIGrE@+7ekN`f>p2&~ z(ClLlYf7nZo4$kJ)OE4~uluN_ph{AUcPc$HJ{ibPv*mmD^0BX4gUduVNRDWa5Bi1T z0H6k*UW{kvIXE~#kRTAyU6cAkL9@C?@ONXv7MCvV5so4k!3=8vG~qUzBtw%q*Fi3> z(PJ98%q7z!&wLSVot>SHU*|bC|Gj9?$e@{KIo7s&xg}8m`5PDFXmp&}vm{B%V;+kFW@wxQ);bOqb83`Xf zdOJ)xT^~5|Dxt$$CnOz_vjShn%YGL7J6yh7?@S)%Gf?x_A!TIZ4io02Occf&nU-ic zFwxIOSo7Nv^5lt*S32GgW>!z72F({@U+gE>)Y97<1^XVx8u^5&FmVjaSXR_$SUkXM zn^-M|G%FDIw3r;RRPSSfjakhXd~UsXA$I9mi?vd{ZF$3gX! z7BU4)g=N|6llg$2p7gOCleqDXwgTYmNc0EYY=L*BMMUo1?ebQNJPsZk7w5yZ!I?Zh zcXU3U%QC$rJ}Wu2WFLH18w)o9MbUy= zB(t|wR))-mS&1(XP1ckv>uYS)G2k(bWnPN_>w7(DT2KDzr|@W@=+sEeU} z?>qe=x(9k$@Q1SIDrUDlaqSzz)edVya|iHK-wh){FF0Ja9qr=b1gv~FFgZ%3}J!2%0IYl7_Mb&3>km&Q7Dy__aT4i!{f^Q%3*w9wu3 zSsBISpa7n;Fu;Rp__SEzW{=}K5a(K~z^E4BhvlFMpgJXL-5CU1$GZK=&k%U(gn|9S zr%+wme^(xxF?gdY(qRe?H_n3Ctt}kqIQQ)(8Qfl;O}${u8Slp~8n=#_+A*TnmRs3V z>$iOP(Z&)3Kk0>5UGRAyURSpNZ3NIh^e??UY1A9?9YIJMudl1FkBkrJ2JAhi1KOYb#3nHhs(>9{#WzLW~c1rX-!iKiAf!rHi!yMW{|GM~K}C z>ph$i&;%+cP#YuK$wc6 z6ScNX9hK(Lg_t<6dmr-VcoP_1Z&FE`Pj57=5x>_2C*2SlG)MY&Nx6!-`PNKxKP$PE znEd4q9|H5F8K0#ObtH@y5C<}5J5x2|iq~LE5Bso-3FYaT<-G>SgwA6u`sCYoJJ{vW zd8kHTpko1QP>RNYfH!JWE>Iq1jRD!1*tK`Y+U`lu^}|JMUfmB|Dw^)O-J@TpZUAPxPe9D8 z{pBI+lluzhxjOrVZw6x$_I5;jsnEsj(`U;1A?yMicpnGhxhBm&3yF!{zvl*#im^Gir8|cDf&2-tY>H*`Pd}ACeR@(`%Fi6^?T8T+W)8dg{@YT$0hV<4 z)YO#uJV1};9rXME{M`>2V$07MbTEaxurJ2?QU_e7DCK|Ke_}Cy1GpS=)`iY}d?Pkv zF17@ZRyY;_mf_`vk)Tt$dC4@y%{8N*pVppH9Z$$ia zj`<=zh(4bzD46YR`gm^f>>mr&2z6=vyKsxgw@4wDI1lwHsN51mE8_tt;wygNnC1GQ zwC37u8KZL9vuFR&cFzQ+o=VlJjT<+rO_}RL&&P-nw7xtFKvaU(uae6zCcREDcrul= z96i%sn4fMC{5~>1iL!Cg^dpvi@-MpghNO08S%h$M^vNjU&{JgY0d$L(3WHzf9kr<% zN#)R)_~4$jm>A1IhIE?Wm7%&w^A4y5+Cd|87Ci0ZgL*)Unm8|%3Yq460aH`c#!3Kp zVKSmAgA=YlYfR(PIw;BW3)~^(K2c)%Ni!&GO-i2-VkX%emwfC#*m1=<8LFZ*K!OPX z39X-B$S@PMypBsq7{RlpU*!ZaC<=&+hu$Lr*UuvpdpiJ|Lm@MY)R{@o2T>GUJV;SS zjDo7C^NoSuJ*hDx3gmZap{p%1;KYoU_6*QR8$!auOGiWKyKD9R;YF(n9!TCYAbImC z76Pr%R_fcWiSsC5 zWg%5>Fc{(UoN6VUS!0w5-N@}wg@#jW4fuweFHE()7ufr*Edli(gi9|fLLmTn51ZbN zZ~FYA=s&Gg`Tb%|kv7nb8X`#>^w~2^E;TwR)P0)Fee(w*82bQ~!pb1efAuz=JcUg6 zSz{1=vh3ciRBdaD0Z(n|V3?)q)U_3`fickD!3)4}OET2y$S|jEJ|AjnenL?}p$vj5 zpVS5i7%qqMlWki4*1eh1)vXCbGeB8;I_2Ih>Bn|v`qh2@F4rFcNUy#Sa>I0!|N}!V7M(9E{zr6W4{HZiJ;IXv{FOv6(2ME&>TzbE5g^ z{fG#K+DYJBMOBT8t<&${mu=F2dqd21?$>kfGrCq99hCzhpd+_V_th2IkQDC9(#6EO zya66TC{U#J&MzqF=9cU#6lom_9iVT59iwH&yazEO`he~|xdF9V!lhW0DEMj1mI|~s zcagUtzF#CWJv|*AT{YD!YR)Um>Rx54ULiR%YT4=>5$eG=H*T%yDRxnN7v6#`6U|E~ z(jiQ(1m-9;;1f_Td;a`6X(|n<(e>45P(}}*_IOt=zgkzTh=2e?ymFKhM+^-qs2uMp zmVm~&h`VlcZ~uW3B~+dLgO*H#^JP=?3PnV0dm`51xp`T$ax2oBGHryFnayuERN3E8 zk}VpQZ=z`0zVUMG{-o4&+v=k=Lo`6^RHrvYw|DWoQ4c6+3~6n^9mb^8s#1SJ-u`Bb z?5ri`;$a!{jw1njg~ZrFY3Oo5jJ#POtafF2u_RIj9%0;9nmz(XE$0anWb`M74|Jqy#uS50WaJ!`{*&T28DSQRNa+R=%)u znzzS(*1>hqz)>c4tYFKBP4F;ybCiwmq88mF-li2%jUf>AMuITz4!|buU}}&ifbuT9 z0mdF+NB_N9^wFR_1+)CowzL2sLOp(f1QcN0`L!01@B>BFWb8ocunfPqYGR4>TxW@! z3%1dk&o9nS28b5@v)LS=AlUhCzViIVLT40E0bI*s@D(8_4bPlAve* ze`BPRsi(Hm#LMFX?M{#eI!yw=_P?%t?-)ehYS6@FyYNiCX5fU#LbTg*`+r_hYCD`K zeW0a|9tWI^-e=@z$5?Ih-2)c_p*bLObQwjy}OYwGLc2YP{U*}P*Y!@4V{c?qQT_=QV@@kG1%VjdJgsfVfTa*X~; z-2Tcm;SAWj4kF`WI`wRGuye0GRMyiIlqaHm_@y7rF0Ik1c|(;O3QOwe7@#E=1JmR^ z16buYM%F=Eo&(f&lgiaLBKVwMKm9Fh3E5O%n2r0M z09!UM4P{_RzjWu$9dtMxF&8A?zx-vy2u^(*Tz&Al*k<83$fT~GR8IP#=Md7=XN^Hg z4@qX5)`{`8tI|rIoX}4%R(0@`DW0uvvU~44Ie7ld%}i!G$0y?Y?_Jlp+k2D9KlOg0%JteyGhzYnjRAo*|~A+ZiVXJ zLML1G!@*fg_8sX)XBJot7CYO(jT5o80P61X{vdZrGbtQ>`Nq@}RVyy-M`p&mm}&XL zI@L-7^WfbFS6#W*JIFn?_pxpaYFcvIPGYT5;Zax503wf@bFF*RdeE7p>H9&Q5>6rkti~B=?nxP2&VS}D2%!P zg8E`y*GbQB|LoNhzrq1jTPj*w_uH2rNe{l$1HKcKAMxs*gbArE%Qn9BMV+Qx+ukx! zJe_5tY>#ma2ndrS=al$XBBbT-)^Ez_uv(h0Cn@3YMy?k?Cm0JS0mJ;#>*V2=I zTOp^b$yxGJy@Q>elPA=fMjvOG$J0N5e(Sx|BzXxa<{o~rsZohQU}LS9BD{MFF*9Q^ z%tOv{U6<}fWm2zhK5lL=qzy`6+#gd^BnbyJqZcmq_=hSCEEHunni?7v zz>amhec9CsCh+!OCxJ`%>gf%GknN0$ib^zlc6K&Olk6{OIDcT7!o<*kqy9i@L6(S$ zee`HQZ5vK-E%#QqB3T%Wo*Ed=R=;==4U9=%-Sj>%`Xg96d2-_y1R z@Td7vr~c#Exkkj{Dn~LffHUR5%j0=N;nUlD_Usu891p`;P3S5i%~u<2FN~@Oyaf&l zl4CV2?l*(T(5t5uO1RsLqtY}|by;Eh?d+I4(Dx$!clx5iX>)V)G1zrpli0GPq8YLx z$~#qs#PrDG>uLrpjb64^ij4s(a@1J?1JEe`g#v)sej=5HZYvONjpmIPqJy3C2~E)Y))8k>P8gm}liBy#DH$>g$$UYJu#2NfdN z#z{a%DBiF*|BmI$H}quRA?_DtzQ2UY%Bgf7h~j5w7?M)?q2i#R+nTL5BNfGtwTa8@u>flM+rXof^FEHJVcU`O!|^MB*?CC=HBC9FHMyuhb( zm1GL@>g4$1*lqeERAK!=t?M9BnzIbsDvOBSDR z)b3`Xao-f}TyC<_gSKk5R%hcfZ zWni&FB6Qe~wqO^qYfQ<(2*4Mw9U<6Q%*1~442ot1ZtbedbFR2(Paa) znw3r1`atSb0aJQ+=IW3k(HRm%#$NM(gCb;id;skf+*za3c7X%ckZKO{oMQ`KZny;#xyT*Oie1 ztIi)77*tgBZxf-RCV?s;$`1QRDxX_Bf~@eM@etqubyZ4IF8z1TkfB)c zqN)`n4z6ZQ$@~N(wJ)lx84IbXya9$d60FiYc!D%F%l2i3gMm%~UxvI15_c273zv<` znWqxvbC*&M4jxHWlAu3r59_zHjvqYQwUtt-4TN@zGv3_w*$EEVoo80B8Uczi_Yy5^ zyn?w;MTb7fwfx~*hd1OS!O!W*`bY^;oNOJIl|qqp471Z0!M6!mG$+8*3t-qM+iyGs zH{}g-u6$fN^H-p1AmFy}KEeuT@C%$K9|WfGN9W~P4+so*?p7oxa$l z00#BsKK0u~0)c7)b7yeIASLJntrIzieb)GuIHs1qc;LVRngm4pVt*^JGev_Z^Xf(x z=cuKAF0bzTjn%-!*b7AWUN%S_c1~WNvAbOZm+d{=6PFGq9LAX)5mp1?7^pxwdveg& z5+o1AJ34mCN%L#LkTPE2*pU zMVP1H?o@B(B=({D`*eSDofVJd z0=E}ZO;A`{yyGHB7@$yuc*hGa9P&ds_WCLfC}tZ0?~`~(=T3;=U638L!4Ffd)s3Q! zMPzBvyV>473=zyOhnJ_kZ0I-o`TN__4UT-?_A|Un5=Kcq+B(eA?pG^aZMS{+lm|!E znc$8&XZueA<{-153D- z>_F=&WZt;H!4LU7VXrZEpZ(hw=(kXLCTzF@PIL^Ue_l3LX~4@W1Y5x`|R5*s%>OoA8A_)W`)d#;5 z`u{ZY7eDyprod$%Msygj9NYA~Z2DEUhj4ozK0Kq#pj!GDrP0Ir@g?d~81CNwmkF0K|x>jacTyS29P| z0y+b-MEy&G2Y*3Jiv>k1 z7<-Y&8$FtWD;G104x{ea_Fi!iR*!*D!)x;Md#3;NHtmJ{F}rckyNCD`3j3!nw49Eh z{CdGExh`%^$MGrYiG!Zrs$Y5_hxyf`9~(heRa#;oNNsPR+>B2-(ENyvCTC(r*=u6| zI4Y~5hy*1PC<0MvV0-V+9C$R5-WP&CXdl8R>H{T#k02kc1iQcv6uOtmI> zBCg6^_6l_~IGKlHdRKEo$H6VlE>g^hs@Q*hUl{) zWi?qs4mJBwlNIQA!iSY0Bv+E2rEh5Z?XSMiX}aWc`^s++R>WhkAL5s&Y~^hw5>8wj zZljuZEEHrh?qy`q)86?A&c*5hNIWsqz7KX^*jydvWKh-ghpm0EB>2BD?O%a%XO50l zq~Azz0|!e^6(~uJXGM%*O$&8hhm4a-;JKEgOf!5`(bW<#q6SRKoZkG6?AikZ1L7Tz z!41GR;exWuy})8!q!@=ncsYds_5~84c6R9S;TI!=Gl%&2{E>Q`5{c<51k1aCn1o2y zkwZ1)U69EmmrGf$63C8Rfqg}m_NsdIrv>^*H~XBVwAZH`Ct z1$TTI0Fq&Itx%pYigKHgbLEml?g6w@TslYy3B+do#z|7spz9#(%ff)DUbF-huu&=f z3YansQ5sSfMBvs@q0c_U5%^<=jN_%Ut~Bh$1->ukc3o)(KFQcSU*|*yMkepNNQoh1 zK{+4{@<(kHFVe(vAt_Di@l%E#I)w31W6r*9BBw&>z|kr^`lg)AP=2NKL5&(+D( z&8;SVk<%lkgO8$%_F$UNFs1OXJpt2TFOk>;d^n1X!Tp_q1yGtPNrtY8ayhQZ9RTbq zyP}aq7SXva^Zx)G6994q`%BdH^f>SofmV>9mu*W35PC`iJ+SnEu_eg(H2{ zb274UYgcoCX?U_6IIP)0E<*ZhPUwG?rRMjeq=Yj>idPGC^!0a3gJ8qef_anRpdi#4 z4h&3zwl--LE^vlTY@ab};s!EeNX$`zixM%0@;vE4W;tN|s`NEM{~_JP^V0i@tf!NsQD3e?;b1-T1V7agvN%e8M(zM zpM`w*7#NFpNC3S$na5!MoKntN1JE-wgy;*-5{Sm3yzuJjvvs)u{_=NbeHO^_UR3F0 zuJO#pi-~>H5RA|r&)pI2K z5%170alMAJhR78YSUf%r6w_t7rOV^aEAvD48M(0HC_-hQ@#TE)DRZVy^^Lr%T5_GN zI@u0xVDaYSo||TV#88B;%eS?}M#2mzlgVnnmyG$IlmXNQ$iOf3qNC@=_nT<4Vjyn4 z&n&!)4CqtA>*q~l^B;t6{KXzR-nzs-X(+|`0bliE;0E=V|-!7bL# z&j78ZA-I0=-juDO3I|b(v){@V){~~?ioD#Jwbtp*6fzQ>7eUvLct^GIF-jjNRqV$f zUAfmE3}h}pQs9lC_xF6(OnY9?Z0Vl#i1_k*H z%?e>O6JunVz=Mitbl$ytHwT&qVJ~LDsdJztfFy)1iDRS(z+;Ilvg5yU zk>d|s5!;Z97*>DKL1Hf%MT($@r5sctIUrUL?_NR8Tr5OfEeH)lzy+wO!(Yq*nK3iK z@mEo+5s5a{*;XzBXif8Q)>z^+a zEY6f-mv(hkgRC@5js6eVBHwg^ApI%pPS%F( z(ges58OVG=SOmpc%4WfE8$yAICK_DM@6G^A;bps0h-=pXO_&jrcMV|8b3t2mlWIo^ z@YopvhKkbmWnsdW+k0K0tkee;8w@VWkfC(0r*srxa1_%N?^SHX92Gdgm8_{?5B)Ki zMA=-L!7eF<<Q2AtW>EGCbZqZ4I6k(}0Lc4Zw2Wk2K+D?M72_ zD`?&uqM^i9Hyk1zCf>+7{@Sw$FKZXhw;7`ba0lt`@ z0l&uvL$tQEnM?r8DnM9eH|&hW+(jr5c_TiL(MCD!{L%4=oWlU(BO=iN*=IA`-Seud zG)^f5F;M4KoeikP$>1P;QJW3UNE#~IW0KGSC6?1O!13G%p{y^CL%>opV_rLEmGsLN z@%^ZynrCZ7lWi4xwylpM-{re6ypL;-*E==V9d~q` z&T8TzYt)mmhv;<;O=dD255qsu>v4f>w~Sy1M%5 zQ-3j2|C9Up@nOJ9cHqD@ih9$4Wv#_>%$Z?!O&X!{PVQ+WdqveYhLi*>yp#Wlf<2#fqEk<|{aQK*WHxGmqat2)fK(7-d4>+R2T(js;*-G?D3m%h@zhC!nZ7L68xe40d3U ze+0;wEi4+BW!<4anx-Fy30}*VMqwam*8+1-TsZuzAf zTQ86*sin7s{A{PM+=cW;=rCegGCn)4+tS7X;Rkz~`@pC1LJ%N&T{xWT1CnmyNSdZD zftwX8p_QB@$-R34ql16Cy8%e!4)O5#K=vjJ_Z0)>8idzj1llt|dPNL-8IZO5h+ct^ zg{7t1@`c+;#n#bQZ)6TI1n6kzbcC!pD{$pMsnDc~O9~XPy-e?F$8vt>? z(p%z|Wz}&^JL?JquXdIv_)Xp@d;A(C5(iyC2AnI%ruIgHgFxr}DDqAMR5QxJTJ=F` zPf_B)-n{_m6=5+G@ZK}0Pah2Q+QM=CG=OU{;E_-dKo%n8Xt)EsH_$Ev@_1xu#QQgkSvKhvbsLG3y2_F;m&rfbwVwT*& zR1THxF_UvbQXLV*S~C~3UqtvK449}xFYi7#K4v~d{}*E~@7X5!9I(Wq&}k?Otrizj z4?#?wQ;)_R$~mZ$5F42x11IW;Zr;(Ie7g6EFGP3_A4W1yT%snWpOwdfXsJPBDO*1zKC2?Qc|?CxM^j&ya*GQ9Wf^)Mtd*I^TLECih&WQw{9riz7y>1B34|a%qE+-eTYWcB1;eB( zkhNdQ?a5&{0Mk!xx$`+Q8e+W$N;qommRgAN_wyD9@w%@enC=gyJg=KWA> z9RR(iI`g;SaCc&)+R56V4{e($K1u5=l!*On6l#_zBaL-u;<-9RZSPbWlR4WfhZ%GFz(@S@=9Vx*Y(~EI?OU}R_-hS{4k2kq2O~fmdUo#$zHs50PAQMC>Uzm|D8f4gAraf4D5}$B%1eO%#_`)GmZ8F6Ls6D3P_l zhId~1A9^?rN+B6 zOAzuQ+rk^T>oMNabuTU0>5=jns3(yZA7&t$l+15qUl?KwP>1LAo(iI`KJ z0P3R3NdrubWT~l6A*TlHfI;?~KVTM!TD1W%m=_y)H>|O-ac(+|!~ug=3UHIh!JYL% zMaLg*{$mjM;YL5aC73DxC@9c(`|csjFk;ovVr6B={dZayJ_{R(V#=jD?BS!0sdEO4 z6WaG7(mA;Uq(PpBw1I=2>z;IHPSMUj2j3|E&wm-ope=Y#;^u}!fmLOw9g9TV@K-oI zQWGZ3bDTH^JM#@-B|9*U04O0YGUFFm8aqhjD!^;8Lhf%q^e7de7TN=Z0#zjO+|3O1 zZ=lfJXY{%6q2JD>jjic2v||PZ%9Z2^>+W2mKD9T4$j54?eN&kFA`$VdT^op3-_x<81qrR2jQ> zQi_dG1gK`V@@MWzQe3OyUAmIZ&f$Z(!T1j1QE>9Ex^C2qOI03QlZ0R%sCjf@il6 z_98^UWK=~32Kg-eciONW`%uvu+HD1)6RA`D+5z|w;A@P?4!dK=Tq6s8Ykl_qP{16z zYz%g(ZcjPa$T#mgX<3E!1dvyCx4k~m&jh-ZkKVpfs3VWxHJGa?YbS? z%7H+F`S;&T2!5W!kOv{3gx$+i8)GA%ZLxs%A;@iBb+JmUbBT>- zH8rjQeV^SNVb1}MFy^x0U*vHE#`mW9vDVvuEFwSrD|{Ax6azy;vRs*a;m$bl6o7+a z{(ak5;UFU>jVP~;rTi>=RbeQKp{DRQI5s++++(i~A3jWb2B1xNd^||q5!%#j3P$3@ zpa~it(WA(whsZ9E{ZL|W5@O))C7@@;4k}QUAeO`U-G-meO2nUHlXU(S^J;jQrC<=V zftVL_T-y|S-*s&BJ6J?&tN-YT8q0}TQ4 z68rYCdg|4fH8Ply(B4voi;q@=92&A>hLvO9AM?0PJ5$;+AYQEIyXX zDq)-(LpFF+uaNNwX^3f9J2+?yKuu{H+E`6)0F(jt;xm+VLI{t%QF`u_kY($U@rVs6 z;Hm8T{6uN>Tc~cHli-TSSO9@CG%7H()Ko{m{?G4rk8(t=f!%xCs9xKFAV1Hk-6`Ho zoPeNQT|-cWrb61{BA=As=99suL#sQzY0=+5)W@&JLK;oP9!w*AryV|kXFqV2p~N@z z&nUxvT0Hgm2K%APu*yNR*85X|g`P&sL!!=-&OoWBQ(WS|H@6(8^~TOYQEV41lbi$> z6SyW{1DF=aTayKD;ikHd%f`n-=b+hu1CUu;Aidr@9ua{ZPOqR14GR{hFJHdgDgOB_ z(tql|6MaLWM(ojk?T|-$@8(-IWFqFFdX`%I5#@jRx<*GMkoaup#>v6u)^9wPXeTkh z30h8&+=dn#Wm$KLj7OZt!u13!Fv1woa8k{OUOtDzqC~^6Z~^wo`qFH>YB3MkcC4`i zX$LE`INQSgMu5q^FJ=c!gY_hT3<|?UZ66R-m9WtYa1USFHhBB`FY}?XCwN>jslb*V zOdqoBx8r8WgP9gG(CA-qc+<>1zKedBpTECTUmX0NBK`L{Jfk7d#4!B}VKdu~ z+@9-}h(Mt}a+dvT1BZ!M;*jYwAkPLq7_^ge2HKFx^x-yy6oFO@0ShH3VSYD#e? zJ-lN9InXX>0-%A$&_PWrOu_dZ7*DoV)z#evc)JhkSJ2{(Emjx6{2D^h8Y@nD5a}BM z0D#kMS(!O<(2q~+i-B3EGYAie`(bqxb;fB&1!~W%paEo$gM`?dM(iYxLRY~qNc#|< zM3?~nNhkafqGYIEM>}v}8L(1HY!w*+YlERX3a85+43v}DNWb^b-aR%9M!5O({XT2a z7aBajEclu~qLlmkgf2^AEu4AX0^DQXdn^&lZR9 z(u0Y~nEd9S0|IYmUd$yN4ky-`b!u@?if=tOtW$|iS7Vf(p+I5&Mdr{hX13d!)2=`z)6)sdcd*fXw zigK0#0%!>R%=mB;5?=u7J%NNuqmm(eF>FPh-7%;l3D=~Uf-ybYT_7gdF^vRj8~euYZ*hZa_N%btTaA(h*I6R@o1*{0+J|zWku9) zL`4q(BtqNW!SltymfrKEft0|;$Y6k`G-(p5V-v8zlgh9^%RNYGHF525@H5W z8Z9ly0&^E1lIV(ow%1&>@d7~k1G{G0 ziJ%D*jkpwC{V>2WyWq%S<6zm$=ew8w)1~u6(PA+MfpY_!s#aWQ`7I}OgwrQee-}vC z4E%J>i<~SPmD&x@hD1YaCSis1!JmXD7dVbPcmjk}mY@;?87Lq|P?rH6^dnwa=>Ran z2#y!3Dr37JHo{2(LjkDS`2$}vZg|KtW##1Lb}%!afwwVmgoyRikS=In@65I3gzN!q z$aO3LsjYDQQUC?480_Q(-^xp!fC)K&)Ix)#DpCtHCXzw808wYHTpK<7IuWb(ZO{_X z^X?j`24@XSr2w=A2?ZON8-$-p^F{D`b;h6(SVg_x{2%`~@Fb`z(++_cid``?BJc^Q{zQ zh#*Q77%w6-ne%-^q2lF_O}N*F(oNoUOsg&RqB!X zE~1M`a%xRnB5#6dx@H3kgM6;mOF(Jon3=^weMLOaJ_o?5kIWUo}Tj*P87a}uzr~E( zJEojUFzwCCsBxBDe}yUa)lT;oEwC-;X|EeeKK_HCVu z7z&80uW=dtpRMq;epmJ*-||`uji_sHwFHvtv3WNx+-GurVh_1Jb|6`*sUE?P!8G}L zn{sRRjg;wI8#}lQr_*%`?FR3dL^%1H(`64XU1ry~;A)+*J9jr1ACF=AUzv-d&`gzC z-m=!v{c#nSYRc=UCiCGBE|pM_5B{mHWsr<+y6F-9saOt5)A$8qT_&3b!(`Fr!zD}A zCcE;>JQ4rH*OOn}X45ZQy=m^NGiv#fGz&w6AwFvr$JNh#@?Vc`F+)36O?{!ACU^8Q zT*8U!lsfj~4Vf|ivrdQ;yg zq;z}c>UDW>6dzs-_b!!QlN-|xzU=mONOig~GJCgy)`8M9R8OMc+%Rb&?KM_2-4}(b0y05|Twy0OW zZ7Ei!X1)%!f4IA@Ot`OQDNiF7Tld{;R|ypK+N9Y`O@-F0uimw-ceE$<>R(#Px*AmO zgsmjE!zGz^x8dGaUq|Cx6&b%6r_LBv>P|~+BvB?8~PPbrCTlzi1`%d{kj3vq64o)W_vZ zYE^FS+ijT4VAo^M-T}CWpenUmlm12j=iMXBQq+5{xx6pkO0v*6%Rr~si_> zSB*E9#D!NS>AQ2cNjq8X+nG(fS61$?ZI>8+&Lyjpm^=cBbr zdqo*O`XKB@>YPLp7h4CQh%S%5c1Xlt?kY*}RMwN+T^q(PGkuF;GVf2u0Bd&BwSzpC zj++`RjiOK{Selu$we`B|FX3Q(roTnm#(GOLSV8asy4T&fuHXE`(gY+&?J|ZL$+AC< z82F_!@QAeU>(O85e&LS?pNA`1mM1A+F$R;+RdHDt?)jrh@Le_@Jt`t8P*VrSYZF&k z!yzBEGAC3`H?IFHcO`&jsMhqWoZ6B|eLcYGn#*iZk{0rDYwsI9xUu60i4<+xQ*Jjh z0o;GR*VEsu%nLp1lQ{w?ro^f%oLbUM><<#^wdf*CPctuyuG&^v2D`t!ul>sME#YlI z&{wx%0rT!2&#eYW(JaikmvyC%m$ljiubG;+$vLx+=LVHGbg(io*v+Vz>^i1_K4t8y z+ws!Rb3?B!b^0G4ji5H(ce{DZX+dZ|zm(QYjb>VJ5ecse0e((0hLG-~Or*v4M?X5gco)&lMtwPlF$~qYUQzQPSFwxm7Z&ld|}nbYS4W1jP_SAaa-|tvPN=f0`<@!N5Gc8+|q_xI1{Tc zPnd|~pPZzb2(oFJe3TBgYb!hzwXf~&Bd1MN$|_rZE6e_pA-*_EdcUl%oB8L*-F@WO zHtD0DaNn6wVk8d*SeI#Kt~kNObA8;`EQMCWmfbw~gamqO_*ln(5mhWN4u>81;ZmP| zE6k!{SDw;6<_ked&CQ!|o!y7Nx{4WrQB9r!MTSQp3R99MkGByGtJOd5@*^inlokSk^`FMWrwN$M=()mlo=?`^wL>lC<9>1>VeCkSGQVv--O9&3xn51k1>84YX=|;qjCR=sX%UI8dm`;bn2cGs~p0FsgwXe)%Em63i z7_;?oUm$ZpIXl&s>yOOEDv6}Wp3ZH4k#jzzbTR*n#|wMczJ&Uc)2);_z&fq|=kkAi zyv3~ZU49^3TA7kvKgK)1Z-}GG*TBsjgYVi%&s>8xuS9pZ4QD>wmB)A0tu{=Wr1=o8Bw zV6{AB$M<4rpz&3&A(lDp3wNu?$Xa@mfSFH|k~-KAHqrv47@oscEn*AwW5pu<+GI2bQ8 z>8qD1t5fpGSGoAAB%2JlBfRmgHi;d_=y~vY7xo7%Wy0(WaG}nlK3(ZqC${>FfIB+* z)%C22SbQbx`nHHjtT?MvKzl@y?jJ2)%|6c3Y<9;&5v`IuPtgaRFTfGpcl=uLK@glB zY|^2GF^0oK@1B}F>fSe)nDDruHi&-njt*@`H}fWbJ4LME<2zBv;iFAEPRU;a@3=mLJkS>HGH7LCWktT%RBHbiF2rd8GC_4W=W6pjaO@L%) zt@X9{eZQ5xM^8DA$hQ$c*834phFvv*uvZ(F6%`1qqE{!BOjMF(3uUgZFIv~Gz1P;e zW$>Np-2{OZ8Z|V~4o;e>@wi|WwL)kF(3p#*_5JcqI4L1Uk@dHX zch?y$E;Xt|vExC=S9_3d9bF;R|JSo_Ob6E^{MW}RciZnHvY4xYV=P;JH~zI=enSey z#T%T{ZLt%Pj4>C4k`uB0J|3!QKRB)`%)Pq08I;Q3N<4ngSiO?=P-oKzl=fYT;o34r zc6p*v*AR-RC9TG93GCX`o~hM(volmVL<6=pe4lZps={ozrX4Z^sd0Y-4(=pjjd5mfCXN- z80<4wjPr3MeRrQHR|wvAawVR4k#vW+BMDV6a3IIO9%0G$+ads&77{-ZMds7D$p|lU zGdl(qO})R+)hBz+(zDiTmsX&fnQz%0YRxVVm8(@HmrGw}^LPKVy{_dvM!^(? zVRR@f*xP(0gn+XDo|fDG>=+qp)%jR|#;#=IJq1eir$B|jz6)!2NB8~l$+3xc0ad)0 z@?q{)z&$cEgSDy)1xviXXJspXh1$*)ms0MKX%GJRm_`STb+5sz_2G#-d=k5b5{=ck7H@$-KVQNPgx z-t=3P&V__s(ckP7UZdF6#I1LZ+I7 zgS53Tb@R!zg~UCLFFJVl|FS|%mr|hQ6r0Z*<<93ApjufR>OK5EWc&4Z40?6#-B8v1 zzwhZN=7eKb|Fnfx(}`umn^a`JvGxB1fv|A@t+C7g%kkd@rPn+l;NRF@QB$%`q^FP~ z3wHphecIn;>|OsnHIk(3QZIq)ZX-U4?ilJ1V|rUke8WmUoPCa;>LbxWRZ+4=aXUo1 zKRcytIh0yFp8V$~uJUF6FjG zh5g_%FVIk@2aVH$v-=_&Z)we{!OGM4DZLI#6q}6iYq|Z086S6DK++s-V>2!&4k(v9 z)v}mnz~AI=O?d3KlHHJ|!tQm6fADsF1qdQ5AxYQ~EtoF8M>~^DhgU z!D>OD!imPi!L1U!AWEW>_)pXwEA{%Gm4badNs@V(r>VGtt{B5Q2P0YNZ}yoi&xnSU zfrRA#7I4PwrBC&i!`M za>`J*k+`+7^vl2sUk*{@s>{bLD>D*6<|L@R#Da@&+%#iZGgrFLg}tFZr^t`J7f?dH zSl9|Kg_A-u(!Ucp<3~_7(J1X9(^tv777dy~a!{n(vuyo5t%g(^k9^akS|otGwVgA^ zE#$oJNal3B+3os?9!+JUCBB%aMF-5gnDxAsK+6L3;`l%>PI^F_{qT4Tsr(y@ef@Nh zfI2?1)_e*mBDGw{3ReHU6Oy?s^HKN61CI8!;)G&6sBB%O^zm~3?`6p9?bi?ochi#j z!V8}EvLtWQK^V+ckuR`NaLKGMgkJ>+#9vJlxOaFJehjl+3kNC?;>ZpYhL$m(mH$|w z&xH!8QbP;m15jH+uS-@0hTRt;KT7X5yqKv~6;UxcIqb^Y6{T6_S$^eqRu2=L;k@B{ zY^{7Qo-n!d^NEv^Qc`y~jXLMAUK*-&Y)i_K;zLTEO`O;CG`KY3`>SmEiO#j3gwO3c zw!dDo@YK?~Q!@Aejk10?<6Gs;d~)Qj)sw5f)0tV@{=`fVm!T|=4*D{Gz+zQL*A|%A zm4z^GFKY)SWPKXEvglVf-DT$Bepq0)xzvw~s-#1NJttR=W8EHWli9~Z<#t-!n6(w# zqX?|xE_{A?+a%tNqK)h3mPsyt9`e(~qY@u%$Zv~JIg7M-lw+ zBK&WvITA*Ov_VPwa$F&gCn;uf*!nDV7B}HVkw7cD)TGDum-qjlr;<;I#)e@dswMG? z)`r`k=Gxyl=tV2;PU>Mp!;Eybr;!{O;%r>7(Mg5lG_a|A{qjZR|2Yrz<8qj7Y54PG zlE<{kZ)u6`3z}e5fg1B68w{`f@aUGo;SRs6;AJ(EN=q-I-yp9X>C0Jose`&sQas|a z+2ZNH7+&z?xLfMsucz+lRvQ@z@#5F!@?nhie9ZVAgkz1P!Qaj15@BsQ7xz)#f;gkV z6=h3VKPKqm3tkQzz2!HK-X?0RZ_ZaUl2f{ND9b~Wiyh;8CGOSbW4!!wh7?QbGad^; zNv%Gu$J*a~F1o7ke&Fd>99g5AeXey)P0staU7KrfXK^>^A?-ytGh=q|`t)_Ie|ZQ* z*-`F7t_0Sg+=gCVkj}Ma)}!;yM0*!z{ro6txro_t4PHEkuaWPjCitV|O=|b975{R! z-KPhu)9Cqie!7_%#3yl4XaWb59~!<{BBaoLIEWef1Jd`?x%6cK3~c^Z2kf zpZ^xHxtAp8@qosj+&9h^62Mcw90!v(_=Lw>E^f(8SHI$;WOrny6aV`e8&N_Lbq5Y- zy&!(f)j2|$&cD%^D5gHU3z3eWOWnLI)Y5$B@=UqH>{ooB8{maelop}kdnSLEx*EytSi&DYZxC2rd(-?3K?(@QYC4G`499Ob`@$lWp4&Jj{C z{MwOvBK!<~cq`Y?pQbU?z1ysKpCDdc7OVLC!#-YvF#D93khzWTOI2{}zWYBO1OL#^ zwxPUDkiBDnnJIB;KtvEtcleSyb`#fn%avT|?Wt8@8;x#iF!N1iB3KZsr&BBwrt?PB z6lJXIgw$(>lP7h>GPqiG7siIG^yti*q>h z&7_VWh{-eU3-d;e=(0MXeEMgyxFxfeq+;P0Zj}yl6e+b5j zsumR2LC7`ZaSocldPqsQ;kbr#6T`|mt`q`LKvnTSn*i|*2!F{ z+6eO@5lWtmwth<+uQhT<4C@Gk!`e8hS`&@F+d?2FnKqPX39{pX z1)|879RI~tXLHvpLfQ$(~aGLRKnwY zLN^S@B`}$M074ngJ~m;bxm=id8qz7h2$a3B~;L!q9k;<;ivR=XW%Nb)(gD?$lb7N!_Me%r@kq`DOx||lE|%l%dr!9&nEuy z<6TFA)s%R0RNXx_Jxu_p*mV*p54DmgH?wbi^J$N9nNs8(YZ+wyjotcson?Jp3sG(z z2l`m>v5i9w(q^SDqT_T4m#d}c^fI7T0!NL^2eFp5kF!w9`3v-W_`#~Jd-p~&Utamw z|E-3PXCvG?XCmZXFUKn)4D+^cNenaE4Ilpnq@wLDtMNQ^r-l52rM~=E>M4gBHut~+ zFUPWq>%hh&-j0cPT_-wUrHf;6nDxtY+uIG_;l*0!V+1!>-eSJnQo!sHJMsLTxHe_E z2tLvP*n|+!w*-+5IaTyTFaL!_=Wl?;X8=l(rOxM|pkBAOmojYKDPy@T%h-HNYl>ck zn*a8-*>IEdg5zRF7bcineiGjtKv|3{&l%*&XSX7JwnE=NkpSxPE0!+xSszWVGu<2* zgK~xRP=q}BgKZAU@Wr=WZc;Y34L2{%BQ%s*R#kF_Y2cy4c}0v>4TtHTX)RpKw#;h$ zh3#;|2Qzj@iHUiwq=HPoG^}deu zlz4R}2fO!|pS6`Le;mYXIQQWZxu~y@uuX*Jny%dw<7*#J%ZO@NyEYrsUj}&S_9)eMIaK;vG|{^4 zoEI7%+F~Ni2$X-DPMX9+g%aM(a&2YY*2a}p*J7?t?kyCb?dw|}m&kMaMtZIIW;qf5 zo=8lDiEe})g_@cRv5P0(tYY-+H6@VXzw8BoeZGOC=GbpCnguguY!@&iA44#enX$XD zCJ;6Kp`zRklzdRQ?D&nsOwa6;l_i$N-AcR?tu6-l^SlV9syVaVH|xXSQ`K}xq3Far z9!2nZA2ayerm3b7P(qz&nE*Wg)mrG&*u3I1_AIXZJpT6MBLqK`%;HDmZ;8c<@6EbK z)f>bWOvEjMv$|M={SK`AZ|8k!*bMFyd9Uus#^DBOuxTE+nCwZOjPFR~VA%nePcL>- z;99OV2pHgFX7{DKeoR=TNCgc z(>^VLZ?LTn1wNM=z$J$s-eQHrVABJj0{bs7U%il(sbx#9m<#jSFvvaE`rXB@eEO4T z0O%F3Y>ohP?bs@xCG@t^m0zs6zS$mj0A=i7-XCComs4Q8egVDJ-y#8l2C;_Y01`BHgL*Loa!RRo<(zF{VDe`1KOD!N> zN-q7E0V>khOu;tcL?EcN1Vfs_+S~h8-;mYW0~gO6;7E|CX_i-(r7KBGp|Ec zT;PtRt^SWWSB*^3>y-LNcyI8Th#hdt`~zH7tL3jyC2)^NM?APBw1pC~funNf-_ek^ zGXIz*PN|xD{{Us~<12Vo7;-2rZ?B1^yd}au`>)phWT|e_fHIOz>({oB?i$;ecANdM zhx;`f>e0igB4D)wU(;B4TWk{46zTiNrWOgJiK{!J+Q>eGLD zkd)Z!<8W{T%ckmQCWmZsF7`V@hWCbd;G|Bl=}u;PfY%0QWS(pNbo9uO!1J!KPe+x8 zy?vD;i~SHqz6BFKVrfHRcU2ZYPSEZA?Z^di?#@@gh4&U7&%GLMh`*~Z*kG?zlGY>I zkaZ+tme|2htLn6GTXRl9WMUk_sg<>NR=4DiAOGR|gt3pN;rfFPKf#t`?Pvco$la%( zxQvEK5l##4IvonCG4+d3kHhCD#&ozzN|Yuqzl?AvrBWr; zC~ndME_lY78~1hU^sr!OiRV?c0B^Fx`*1@rJCpD}0nTt6DZ`-xi<_-?E{pN?NA?sN zOa3YaZ}pm2o5t@e7F|8ruG|Vf?4#E9%(QdPY%OI6^z!Wxl$wdPY>nL+et-OUul?7ojejifaq!K*zjtrHlytIQr{rkqN!Y2M}MOTyEnpfAxrWMfVUb77r zJ_{~2?+gc|`nNOmHt^2CF@whlM+Xf@=eUA0lZNvxVtm2xiUQ-EbNq@h*>Bo zz?R$6^WJ1|EfMlgl5zUFJGlt|VKEM>z_J%xKMsmoYdJJJG@|RK4Y|z{ry>^w^Mjw? zuppZE4@7<}F9;^dEv_qZ2p-V`;}dJT(*Irt?LRn7yj<|1ou0`6JWL%e3T(!2i4hLVm2)=06@+mH}^^-Typv>er-8LFBwug-` z-T{R&BwCcS*P1LF_CVK%KZ-Z?m7Nqk>|9GQ2&VG_oJ%wN&C^dQa;3$cOgWkA z9Tu$PLiTPsxr7gXWh5}$6YP5?tbyoM^{W*1$=1n{{sPNZ!9r1zh#GJ#EA)D|*w7nW zCkbRjtfB`>j5DfS=qK1G4VGqk&yIG^`gO_gmdf^MC`7nXpv?O8zEehTKfS(#YNrbT zhyfBebY{ccg$CrDFJ!*Z-%F}^3aZQ+w%8K@x)ciw_PzN0R>&(OJQyhmmw8gt_*07L zVQkaH`v*!TL85KrH3g!=sCY$icZuWwbSUMd3_GlQBv-@$b=F)Ak}fK4YO_!`VO!&U zhx6-I%v#n+OC@?4^M)D0J~X5Bi5QuN@z?68S?G!lJBVvIsKvOrgsU%at{&LM$I1_@ zcwpU>`$+p@q)|~zBX59vvGVxS5YfZ9rin-3!Q@HfGY%8pa>=OsLWLh)E`*7;u3iOh z{+`@Y7+z}86Y7NPKBK!h1+k)ECi-Ept*k^A@-u9`c}b;|)1l-jr3}<+Jue3jk?}*8 zDUlfTlrdPA+hj0M$rrs&B6oTQ;N}&yVrMg~;x7+PGlGrrFT*dI`RcAS?p_>!6rv1< zw(wEMMhay4$v>6T8}Ra0qG6`>cmYuL*;NA^m_<7hJ!-ahPTu~vMl%0;&!QYRNSUarQOQ6-6;gef8t4RU0V$p+Fiy(F~scY zQK|HbBi@d}NI$Tbta7w38}3^(FR{%hkY0_Z%pEWu&b26Oax{&}K)L z%F^Q$trOryMZdPj?@+eSN3&Q-xlc#BMNFXH7aY~>IapeJ1BW;38?mh8$D|#@&|`cd zGQkFt-Jb5*cnU+QoKmk;USMw_5xPw{x!>zqcraT@0;OhE+8KO-Y&aI}(tMXw??B-b z-i|x&m-*OqZ3;kCYjGh8Jb&l2zSw~UUt_e4LvXuPWE~?7WNM|Dp9uqWaTXHHK1Td) zdD>mtPZ}E$Q6cV~$@uyY>L#KVj~hQcK=GD%Fij57jEy^1eSrk=`Bleo zZ$$QDz&O&M;Ine8N8aBJlC1-W{)?fJlQGvfhqO+(5fSa@TENJ>J7;xxrK1~7MC^AE z+HsiN;7C}hif6z{s2tDInsC^_mVQHv*yV0^qK=b~PQ zQ2~zuD)OIqfCKU`0^O3i<|Tq`5Z4xO8rMZv?+nf4zYNW5_Je)p!i#8mW}+9AaLA>m zkb{S@sr?H@@JRNdsZ)jn6x~u-PLT;zVl(I?lL6 zJ_@lhBhR_MSCy|C-En@!*w}lE9VYp6oAi&RvB04S$0_+){jCSNPP)!P0tr%C*ooPDssw^V@I-KG-Zv*8ws6HWHX zr{VvtN<%rN#mL=0F4LG?;NQ|VD6OjxUIfye=`}E=GOF183_e@ zz-VNI`X-~NeT}#d1WifyV~9F-xdcjW`_LNIB8tyaKgzh|bkvlVu-jyea=b+IL~uRo zLY7InF%0%ek6qcku7WUh9ao^m7LK5-#?C>W?{k7>h+6%#8{m<5A}1XCHGYh1iO1?m5B5 z0rLinqzcMQC2wAyX2Hl|wJ46mIdvhiaq^TJ<+F{nqGw=?FgjiA34GnVdpk(b4Ev5GK%A@h+B7zvi-3R z0E~~TGOMeCy#unLJiE%%K7LZD57OX1|CefT+2h089v}LFIZ8?Px_4)Pf~$PsL%VM7npVKSE-tY$71-#E#uut@UCkb70?35~5N{ z()$kD;&C5MDqqgHCk>KB<*I&RZ=6|3E6t<5id~!b(i}gF98ABrT2kU5_c4U>>wO37 zy(Mm!_TvRPpDkRw_H!;dd4$p4wsqy5e8`cXv=TjDkZF`P6j}$ z+;r;0lmQ%dC;Iuh@)&CCVYdC2 zR5Eun(7LAF!l`%R#AIyL!I;xRQD^fl455LEgNOMCh^)d-Z-C%aU9pjxN82W4loNd} zA7~7%TxyBV1BqdcZ%y-me>5_XH?UcSO)gcxU%=2gI8c+=^tog%gVN?hsi|!mBeL-x zTm}{p`r-9hT_%MuIpI;dv6L#O26_?q2+4+)vRboownGPL@r&!GudOtrRcLl|iUS#A zGwfREdoNBa!KlpBB+i2?jyo?jQe+S8RR55RM{RrTOrJ$UK>Ra9^MCx z^`k>F2`6bjNgq*8EM@lpR-#*CL9ToiaDOsGRR+s5A{Bp3=<&`d+pcSqrkm|ZgZN^i zidoViA(v>al&GxKgXD%u)Tf}SFEeKSXEGHvr3bXcO>a+qK!8PrPd`6naMe$}F%C@A zWuTlRr&3@0lw1wejCQqQAfXw)yHgAw#?2U)~~d8WO#AW+hSkMn5@FJ z>hc1jkzB84%rj{Pj2l)f!Jy0&6{i(%h?R3xIRpz7oC7}S-ZgN!k6Yr?Wanf;{xW_O z25bD2g8eBPr4@<926z7{N^qo!co$?ot}!AWgZPr7H{`Jl<{q}fWcmL4Nw z?&|HM>U~f-*BH_tntryRz1VL`9v0gujZz_b;r7)W1g<({mN;02V4fJg7H3{q)N)KD z+q~d;OIcXU^7T2OWK=3wY&bc;K}m0w&&6=9ZrSOhT1EUUK(=*#@?;*I`Ul6Twt-7&x5Pf~Yd#z6Jevb09CJDIW2S_)sIU;v=z~8|kmL%I<=2Ux%?s)HmOL^W; zRwap9u}ZSNW!4kh#WkqfW<+bYdG&KJvaDp3JNEfGMQ{NJE7`jw*`+F^@SInPMMmXQ z4viSgfURI|F4*Yt4|A-@O5_>Dtx35&p7!-2jl9J=m8Ve$H%s9H$NFOy6Dx8p?<36xiI03SR zZ5F$lWRtc7n}r>3w}j=z7$f09wrhZWePK8)06iyJ$ zbZ-bz4}A{hw!pk0rYDX*qS}LOxLie97FOF7rpN4oMP59s`~&NOU(5d#-zG+5QkQm@sTOnB2W`0V+_br>u%eP~NoZcrJE9zr%BFq4d_ zWUdIF{|8*8oZ*w*)9B4H!+8LdBWl>9} z(9!du&nIq!0sgzhi8gGOQRqx=tHy5T!{x@erC>EuS*!=wnx|NtU|HdNnmMkeXOY1h zqwP1axojsy>nx2#5N~!uW4jdd!inTRH*+@C-{w(=ifYr^aLL?hnEZUAY#GKZcBAwn z!-c-2R^pn|bVlpz=(3YbU$x!sCMv61_fSm5MH^%CMfr%$%Q{G0|MQ18G>M%7VD1-J9?*wv)YqwMC zT^gv}{@Wqzq{fFWFN9;^C@aD_Y0uFBgR@rE1z79nc%?Cw%&>*8BNaW zWt#_>N*LwjxW9Y7atYLqd*E{p$?2x6)h4K*XOMy>(xvkdei3}d+8fwIpEnVzZBI}5 zk^GSp2Qh^UdoaEuo;pdElR^^57a?U$zTr@!tfh%1nIr0;JB{Q1jrPdavypqPt89oj zqsy*f0+ccm+D-~f+r4Tsmf+5dd6so8|Hi{sn^)I;5z)`wnSH(pX?;%ate9uinmHxa z5s|R)<(Oz;P@49Vwi?+>OPpjBuASZATsK;_CA*Ox+k82xs;jwrM%ItbX?2x-URTI3 zosO|}A+!(gZLIcCQjnj|thrT)(oKp|m2cAE=Y3&xHao!(wRO)RPZivfNGRcaXf3nq zXV|C@h_GS5IsM)R#|j;VSJ}GRR3hGYr$YQ2yY^#G96bjHx0$iT-yRikM*!5*a-za% zf|rJx2sr`O#myH0e3=*SM4GO4(pIH>K9y>!)U+A`BE&X?QQJH?CXm*%?XVCHmo`5u zSS3(vAWa>l2(t}yVXP~dbKpnsb66RDo5h=uW0qQ1y$}w|W4;NpxOJ49=X%4#n90vu zjM?<|;TfAH&m>T{OBNmQsE+ME2t=~2$}Kq+R$xA&%|<18E;RSa_Amz06`6|;%^%9t z;TZe4Y~8sjq)2ufzrt#@1aM((ezesikci8P@~=S2VtTJ!3w<}w26N4!>HYEJHE=!J z&VQZ?C0Xxb`RLq@fjkeme$wPn>5XYYHR<>R@%lSpr!g?1B59M6cHf^GJ(DveMavJ9 zm-a;zTD@oji8Umx{*NsI8i+aha^~8fE=yb{u2z2UTyMvz_c8pr3`}gco3@Li5VROP zXZ2@0`Zpxn?zg%W#s6`{-wI>I{8YSG!qML)DN5x+QzPFAU~Vpn^Bybrs?0r1!bkWc}{HLbM zF@gLO;?%5IxDW1kT|)b6h%A6AUfRX|Jsq?SJtr@^$LN|zs;1XUDsy1$Hm4?t7m&4L z(zVlPZORq=32beg_s@bGE5<6x(P`kW%Anp+*T2MkLjEiF53iQyU@!@ruQ6}wZQ<@} zgfpFLBrYOB9qR7h@BcJ#;NeQ!TX7?(uDj@Id~7Bu&jCIN2B@}z4e{Sx@Bo(k zqzFLWlQX?b_S7J{4jH3*0aw*!(dT-hedbIyK&}4jajk~>sP^*+-Scnh+|Q(}$eWla{7Nj(Xa) z4$Jam?mmiAyKpTgx%z2Sy;hd_n-?wWS}n=N@9Ah*;OTq2D2zDIQwwO*=6bQny-kW1>HR2v@Z z5aQ#udGBda+bMbb4A`#b&IsMgrvYB%zHm(gFtE*No9%1S$AoyxMyz$PZQR#m5}4jq z&I6A9eYUlEvOHdLCd>t=*h3W%3F|5f661+{CFEU#awngXQ&8N@FCee!^J2hBrlme4+k9669u%)Ott!uN5x&91tJT56|u&hk85VPe`49tc+8*z` z3{i~+jr|HpG}h8*!4(3|H3umW;P*6A!)$*#r43D~b|pIt;?gL=i*0EsV2ZL}Ubb1PEn}a50$>F1Mn5H&JH$l4 zxdF$&YI^|uu;@h_AO^|Ojb-~_^0M=brDbE&&C7Oc%IJQo@RI4mumrHuf2qq$Eubz} zd7TDSlChq?Lj_2fo{mgfSmi{{vphWjybr6Cq%|)M_H<1Hc_21@BI8<_%H_%1aos32j&wh!J3Ng=nLs46lnOrTAukZm ze<9osVxm#d+wX@#1+&_JJTr82kv)8B34O=1ckxL-dwl@Wulm9cnW%ne4X(g4>O=b^ zqpj*=q?QU-YMOuZ=a^rQNNlPy=G>C0O0vQgKpAVaZXQ+FIiam{BG;XGeq>9gB7Go- z)MGP5JPQaH2%A5)ESyq?$x133x=A+e@n;=MmT5`eXIrO_l{p*4%^9wM2o+9KA)kU- zdA*^GT~;mjdJovgF-=w1t!tbeViiDKdMb#fvOub~!TSW&w?B(nqZ%VP`ei%4aJ1*? z;1+uRfOM}$Dxi8a`C>q)yYYTdX&(bukxs)l?#D8W#CXNC6XH?KISu#TcUf~sM}Nku z-JYyeHz1lVmok+`l!m4=Ewlkx!8_Y|jP^`7TWHHJxyg<;4OC{dV+fc>Of>UUTGb>{ zB+J~}SE4UBtI#Izq9)AB)_1)B?9xVyK2Q#3A8I{x@5!u!`1z+t_jd0X{x^29eH2G) z-P{KEgJR$I7vQ)L6@o=HH8wS+DYeR4G;I2d^rF#pg*4~?Ie(ooKY{E!&#NvJah5qw zjjp8Sz+0Y!TJfoA@RFzj<8i+)#b$3KRo$%L7Q&?dO-3G58HGscP>qSlmi;b z>pUV???S2uQRQV@arVwU$YR2{Y*--oGs{xO1GBSumxm*BRgS1Q?_9n`2*O6Llmklg z!(+Rg#%07{cj0aCKs!r=$ZiiKy{>5Q@b{ADXG9l-A3C-JRDwG1lV&%A2(x5~)K?Wc zW9K#s@O})8IDfNdS{^NsOV;}pBn_Gu6|Jupq&wl=ypz~Uu|Xw$sJni#$O_xH z|I3AM2_Pj7mgY9IEhQ44E2!tG$#i>W@k-4CxCx@eKQUG3wQtsL@#GoO0%q~&6sH%B* zX3vip&by^W1}Kr5 zEX(F-#RkPAsL;(@W`aL)Be_HE7TL&ojtul7ov&p@q85l3+goz$jPA7iMHY>&`T$WZ zp%+}Y>9mK^`eo^R4--(e>QFrxe`yf&?6hWA2}n-uiDJk_zI+dQ{8@Gvm*o3vDa%Qd zt#fn!r~sU0RaL4C=g~4-1Bt%m9r5&PG8xbwE1*7LRtf1qGrc8eF8w%LXJH2#9sHeD z|A~7()lxPY)MEeXf}JWt7*fNX@d(@8JX33t0bEexZEw|1eqv6%y}Vz$?;@|0i(*;} zq1^p+D>)9lq!C+q4$?P>@&#>5#1q$Jp!9F&fU9Us%Hht(+iKw~wj!@!ds063Plxzr z&@|{eb~+MecMx$7I&)LAU0FHuZ_C(WYs1A(Nto^WW?-rX#i^&_X7DMC?4ibeuu)W1 z-jBf@ERf5PQyo1~9v!}P&z&B>Q#g{mSMfH^(flfYujasaxklCJlkI604(p6q7i~ND z_n;J#W?v$Pw+6<{3qDNb{ehkM_n3fD-aNQP?#76mQ&OW>QJG&QamZ|TPz6pdup>DO zVvAP~#q26hlWnoB<`r2j#ThH^oiWl^3_9KZ1(W32YA#Ku)gzI8+-$`_C;(FQX@h&i zz>!gl|NL;iSdfKZ)FLqQD@;mXnJ`rG%pr*I7>9Na(z-DU$v}lWTX1o>1KSaQ?Fwt zWCfsi4ptNshi8B>;`mVM&4L{l7#$Aj9`9VC>hvS|V3G%v-{cYrjM+B0 z`S}?C(e-H26o+aslYClNzz-7>c$+A%2t=WFrjml21v*ua6O{%O_@EKR_ye3qJDG`X zc}|u;z*UjI+(A)y$)SQc=692`ITU&4ONsib(JKCnjF>*bZgcCObZ!)0$AObx1HWsVZuj zMIboa&nt=l56m_(WJY$W9H^)1b20ZT#$ukCBkl9&Knu|P3_vGZrNPoxr5Rhp$4|(`D7oybT zi6VGU=K>7I84Cg{S<>9b@f@~G1nBnD2ajR9v>s-^0LJUuWBYHo$>e6e0nyTYDnsI0 zVo7^i#|4NIxj5$0RR8^i0SY5%q99g22xmx~OAgyJ zDoSaIzGz3sM}ti$Mv_Y75-xmX@uxe=e4r+2t{wON+gsLjDx zBQ@@L1t>|p-k_(zG&Lgk5r#W}{{V3`fiwKH7CGKTF~13PfuOnCEP8B!Ct|>{Ajvxu z)))>lMDIIcJFy(<1o`kwClY`!(?o4#bwL>smo%F*GUSD87Nt!RWpx*MCjt8c3)V(#7agQi=YBbw^NVR%bAV64o7_Dy$LL7dHb*6sFh&EzJ zY~{%kP<8}Jx9ZM=2w=1SWUe(`_2k7A}nFIQZ!yi0x?t{j_1sP%4=C9wh z?bFIiu6pfx@_!`v9#xM9BAP( z)vr<~rfud7msl>;L+TS3@feE{_5j=ETyG2%Wi$gU*#Cc#nAQF{mVLrXPc?7KO^B)% z1nu7y4QUt!ER#CjdC*{|$KcvFP*~b5rrXO{9=KdxlDVizep2FE)U;2l16;b+1H56jVN07ZndM8z)P^z~#yUYEM}X8IzlFPK^)RWHS_1;mQaKfe7^b$sV<0ANNirr!?#&nTBSjv-I?c+Kc|(0F zSw=ovWINAwqM80wWj%su>6i|L{ZDGcD&TG?9rG&*t?@>vzS%JdR6g1Fw?L-mhE#P% z^Qa_~VCRP@^n{~)FwliNMoE!%E1QwcXN4m8hUZWY(B_YtXleap44hoVbfc<5v_=ZH zpM5BR7!h?4NaN8MbpJJout_?A3<9b5)X7t+m^+UECk7Msh;J8xu$H)q=@e*G!I8JI zaG(tDtGQJKdc#Y~6>Ul#eAfZf8Sj@Zb;+UiMBMjxvc}l~Q`pE3$l&~udboV&tk@eu znt-}MZ8#d&LaGv!WG1CN*kEnHc}Mgy*fBjziW5(F$xUG-&{Y2Q_x);_dU7i7BsQdW z@5 z#tD=8?bcCLy{0e~F?|@T4ST8um=wr%&D}CSn(5c5e*k3ZvpBYe@&UuPV4W?O+Cw>7 zcBlkhO(8k}P>I9s`GdV3&lM2Eo7bcCLCxl(&f6qObPrPFR#tMbbi4b7*Tl`u zr-52&jHQ_eWzOXpS=$LqU2j(s!+edV3@l1#%QIti+= zmdc7vAa8f$t*TSS48_j5PqpmW2uOlZOIT=+YxwcJ&`K{8knysqc|En*EE2#I%@z~L zUPl1Zd~lR1o!Lr%?WDnr!;^UwGI3 zLF>4*8)an{d%5fP)A)gD6(?0{%_)AA)uSokTBTG$);XI{2@?W3V>=v?_rF@qBjxQ#65}rwsJi zw#O@i>}7&~**tJkdmV2b62(2)c~YAXY@(r8E8swDVh;2M2D-AK-VxZI5eplAxz?Uf z2OuhaF8uzwMm~27PsgVno+OHhVX;ro`i99-M$JkZBM2iE?N>saXIoY!N1xP2ui`lv zvHT}mujTKE$dSP-6^WZO0bPub_q+Osmdb5cmF6a;m*le(_)$m+t?77xfNSUg@#UD_ zwCVX%+C>SJ-48`hCAvlZMHTH`8w1NpFPsMsCfby1&dOrJvsvlhyAC^j0n!O z8i4;Coj!PNk|#e6^yq@a8s(?RzJMA7JLP?#-A z`sAo391&m*?8z=hTk&0;=+f%)1juv>8BoWh=7<5OI)LDXcZp++l@#nq)V*c=OhlHg<^7GH1!DZ`h z^Fa$v4deMpjnV4~R2)y7Om@Ud4A>IYO6r)0c82)Abl3;;QPANNq4vrILH~fws#iQ? z%~G>yQyKdi)4gYb67w$&J%Ubkb!Ky)+zO3^i3?wi2}{ub+CNZmxrH5T8QfPx$#5Rn z92e8;mzBZ?rD0I@*LFJipyrD9vD_~q?+!=Z#xggj3pVRA&Poq?T6o;uo(t$QxscHW zlw)t`iwqL|(<<8A8(UqBroCPLQvJyGd4tms-QI~N;wQwgYbWwo7_S}>OZV-d+m;GL zcDqKKBvLQC`iGvCXko(16gD$tk|5IIfI7CA@?9)0BRp@mbc^ac>y|Iy{Hok=p@ z8m+k3p1Gnxo6G`~Ox`*9g6U+CrOIP|^vX>@r?!BOwXTmdOFaQO+m5wJb*G>t3VOXl zn$w5R07YX)&S@R_0nDaco6t{Sx^G7wG-}=)6F6`+h9OwL-RBw&^44^e4b$sQ(=UP@ z7~8jh^lgrTc&^>d*gxvy4GK1TgJ&yG=CFP9pa63{@fP_q40bq-<*t*OXnKiEkTlpf zN7xu>LkBK-ufLSgoqEwJFv9#tlm+Ji$YYgaOtY?3P+Q;O?v?XXZzmeCQdAnt*||I88v^06>b zk*VyKGLKL9EA&jKh(Y)WTiCXx2b{Wl*mrxdO6LRvZ8M z^05E0YS@w7dm>~uwr3uMwN`Ev*Y33Z!)I7BNnRBL6R3@5yHWSs;^3es&f^*oQN93; z)q-S2fPvUxqfEOcV(onS+(smZs6ZdaM%##$oHi?IjF+oyJx~`If%y5!>*GKDVs@?N z*SkNRcRKa*-u0Mtt;C4;A08F`ar1w9t|!G@`q}mn2W0s-9B_}&(95n?Hhf)F`l-y`F0`J$vtB$uy6vlpaK`TtRL0v+!k9zi3|p9x@m zdtBkGodu2qa$TisDx=Xu0PVzDGd@Q}*;F4Q?}5qxmYCk`1U5J>n$o?x)rHy3 zMqUAQhd<8LHd79?mP)W z)EFTo2~rs%XNyk$a4s-dku|XL&TJ?J1@ltRp#@%DugJbK7qb}u4WYR|o!z$!g|j2YJl;};xkM#A1~)}cUk9Cq zFV_VSQ2$+p$10t1pCsf#Yrv{_CRe@&T~Pm53qY2feyv0LaW;Zo@=NjbjH81u>gJdE zipI+L5k_3fDV>+M>o0OJbUV1O-{S!DLSLKMI?EsMG?hN0g4OVnb0W9-8bnua=I9Jl zAHzWilmIr-_G_cWIhqVPj)A3>_p9|xP9M6zxgfX&tolf7qy26hpI=5ylH6f;TEpd- z%+t+O6zCdDKvT+&PcSL2YqukL_aZ70IYe|b&tiJ z++cV6&NLdK=pGMgx~fVk=?hQpt=H9V@ygtu?30*#dCu{tgRmNoQoHBPL5oQ(sd?Cz z)q;NMUpgEmlyhTCEk3knhP9UHfiJW;!b_X*MqbqUINSy8*o~5>Vhy;L9s^?eXOc z%0BN(=3S@^js1*G!TT;LN{Cz02O@Lb-%_QyMDyKOr4SDld&~D64iej#tz4i1p||>z z1UUJ=*6b1)AOcOW4w(hN;Y{3S!&KE~lEjS?oU{{vrJXQn!J=sORbX(iC+f0G*|ZFY zbn4hfdZBmLLSoriXs0l9GqWnWG{+uXLj8#~*qMfE%>qBRsQF!&Fk9^@LvNkx%Ud8f ztn8)n;(PnNqDRo-l-rWUBIUs0 z#j|&MA>|;jfBYhOr4Z#3%k(Adlg@Xo63vA0p77{|RKyi4uxGzCe<6z|^JW7)D~6LxDW6%Z6;6HpP58TJfAkqwCg zvO)kQ5FrdnSRn!a=S6Mr{r@z5KgB?rJnu8k_@3`M&+}?u;42-_!yp-oI#Vi(z}(BS zBCGF;F!TR0PD3Z3wzOYf&hCjr^|w|l0ff2OYTJp2bCB?9zey{!=gbtv&F?_ne8;fl z|G)Y&PQ}Z&Y!}x`5P8C>j9pRXmk?vYGc>x{XK`-9iS~+v>5{EUR47bod)1_xR@buv zwbrZ)P8GXpqY6UwBrFsg@JgvEPQR0MYj~1ikIe^O8?n{Lo{T?_^9sy-U`fQtLv@o=efEXcZd%=vUJiJ=<44dr^&* za&MkWssy`d8w19<)RLPVy^xwJT{`Pr!jir>?|rB;8WuqJ{@l*vqk48m-AwRbhTWEB--vq@AYWS zI$w{g({}mICZ3xrB$LlizP$`3-7#;-j$lGt9E9@JM4YA?_syw{ol7#6=|`Z<1$`Qa z;a>Zv)*C>dHL-^>5>ilYJEem?+TN)mA3ryt1q(tOtQW-A`HOB*PS8D;{@8QtE7Vj? zt58bdT76ESRkvQnd^)cb)mdUh-69}ZMkVCBctlwDYgFTiWP>SrNJIg>(xDx*Lu%4I zg`?V~_q$3rBwKP`aOY{Y=bVldKiV{QFZ}SW*(e98-+!HfF>}8br(vMSHIqpHN-Q*GWmwW1arN-^R`FJD_bHLsP8ZA?XlEAhpO%9*7G;G^$BY(D%c|>OVVrd-SnfK zVHS!pX!DdB@O!EOg2@PcVGY&kq6167C{H}KpvBtz(CLT0J!e@Zb38C*Z=YJ+?^IK$ z4lIj3p2Tyu+RX_=gbL>hty?4Yj`z&>BLkU!gVEv}m5bxdPwKcSgq{gMm8X;I(d;0U zEh={HK!49K)s*5o@8D^-*4{E+S`cecSl=Q@`-6&H5oB}X_jo{!6*^AEfBxp?Vs@)k z;CYo_c`<8A$U;bvZDaS>5@v}kcn0Az-~(jOM(C^88U^iGylzj-v@w78Ggmq z)h9C2&>lh!8(oiBwDSFk*#*MHY?j#rSe8PfH4 z$%bv5|8yo=Qb?6rQz2VHD&D<5-Lqb-Uj+=)dfMYe@(=?`%5_lq24$svbovmQ=B^L3 zyxi?sE0jD^>V2A0F zx8lV;CxvYb6*FVWjk|F^IZO^XnA;6DL3<7WrBMpp$ zuEk~NKZT6t>BIlQCPWYE3mcP zdYeAez}E@5uEH}bQ=O&}qHzEB3#wxL<#F7w$IxSS>U=Og>(0pH;B4Y)`(MEqis0v` zq$%D~HAUMVlmCaS2aKzqwt+%we2qtq|C4`0Chty!7ejoZ?MbhRAGNN7e*YR5`w^ z>#QFRErkb;gI{M@-nE^Z&w}VaDK}Oo#7=RG-h8Zj+$d$w4pf=T#HjoN&K5RgJstDu z$an8}EPyqC-K~>0)Xj25P!@{gs_}v9_k*Vn^O;V%BtS0Yb+s$RCxukN_qS$c&Pi!R zF2^E6nu85w?MNhuFdIrHla4}qg|?wWHdkaAogC}vfiW#R>gJHU4ow05%n`%N@Z84t zjg`}%#rK~^)@7jq0Up2MU9v4dcZxLke-;*mRG0v?7#dTPOD7Qa0j2f?um)mCHpzuK zu*kk2{>qX6vlEYj`9FKDM!-Woy;P&08&(FD|ccV*V_B! zxBeKDPwFKTz>uVq6yjWy6m3=GL@wp!E9Z11sU--TO`yh_=TdeuWMs6-oy3_6-Qbf# z;-WND!`-OKS zm0sA8($&8KFh*(=r%l@ro_&~JE9Y{j`XFljr*nD-5vLBmow3~M&loNg9LT3qhRbh_ zR5;!XCi#4?Nr0FHcl}<9%x;=O<#A9b4~7FW7&D)rg(AWtD)#@vWd(2!F_1tU*j8Y_ zTkaq$5qR;>zsI@({8s!M^SdycslQ1E=4kG3ymV4l#rq3{KIb4 z&8JGZQhY;=t0F+xN{r+L#!v75K1)P=m`~n4jvPA_w-q0iyqJ2D?)uG_XAnTO*{%7~?I_u@# z*82SBiRWbc?*%WHJlYOT!GF|eu%#jzN8a~M?gfgbaP8(M{Pcv`2@d zZGs1~;Ha&zAA8R-N&~WKL>V2spFs<|R?6U8uKF?ri5N%+r43X15 z*%uY_P^T866px)v<{qoH@G}uo@pRqBT5E-xd7l*G|GqY!r{i=L7`_RdlLM}a;I)Sl zEB_}^q+bSHi9|Ny5e*?&ly>59$?h_$JH=R(l1GymBnTpIAP8U13J520rP3}9!mOY= zA4!)IfA`i%2_i;&OB~&LJwUDs?~rH0n21%4A?Jf1f>@aj6>H`BUDFyJUIo;q)M6)G zJAt+AS6|}f8kjDRzIuPmf9gv3bEKkC%8)%eaYZjzlbpGoX!ArtqIOaRH$N|VZX09; zoYZ&J{TV3DH@Ceva2v^Y;8VleN%!M-3xT6eNSDK=SYPUT2mo7NNKBMo^7}Rosnu&k zN)xPfK0==QFcM&eX1n&Tx?w%HR^%BTD(aGYCn%z)J%Q|k$N70~(j>D_ixtqDMEkHQ zxJl0mw1Yj#r{nyqyciXnhpgH%jt@>vF7;*`P)((Utdm_U=G#`^78w zLjus^s@FYX5d3`?8`>90@<}S@QUP|)nx?_gj@~>kF0N1-d%v)HKN9Yo3{S;9ZEJ^w z1})VU2umPQs%@LTt90f;ST1LNr(yhWj!8$3`??lFShAi%xg_6SjqHoXbuczW`XE8Xk8b&t~0gG43( zG>Zd35fRt8Mcra;=qdXVM$pe1W|p)+g)fPRVxhp~OV`bx^la~Mt);%1T_^jwj!Gba zW^;Tc2LWc%l+|HvP!a+=b+yyeQ?ynB zH$G2AIUe{$;9Q6eb?%bj32y;D!S10} zJ%p)f%zs@F&Z88=qjZfNnbC=R7rBE1k~K`ynkUJ(0ZQ5t?Q%;|dB|0~v=!Kw{!y7Y zPLd4xQ8Nf&z*QlLAT`srPP78_zILJpvT~O{v=7 z7zu@3L-G!B9=wS1x%l#cO1< zcx%gGEnjy7QIEg@EBvVl0SIzEfUgjjpJ=SEaV&Awff74_X8rQ3&IM3$a-J{UxiT)v zEmzwWstWt>#xgccW34ZRD^9#Pm8X`cXkFG(f-YPbx8h!qLHF9Xl)5<2=L>I8A>7CJ zY-2RPA6Yg~W&5S;W*b6V-1>lRI#-NKirpKuM_EIQfomTS+Tk52R)gS9tx7VWo6wFLz6Xxt*UD-xxJoh0f-l3{$Aa=_0^)v@wS}WE&U^|il2}9fXVzdHM zoJWXV$>lfpHb9`Eym>_h@Lu|HIeSM~8YBduktcGOP7MUTZ*oW$+@cNN0ZP+VVe%ou zwz2x;FbI=0{*6@gm&+niVa;Hz5QKw}V%v$YUa5Ya(f2vUf#gg&%=0!Y!S&`5ulMLW zW0AG70zpPUrsQcG!F;e&BdxFfN;bkFq1Lt_CW<|LP!4@HDNlQA#WLlc%upG&J*&?9 zobu*YU0r2+OWkt5ap%;_by|hI<){Ir4nT~B5TYocF9jtzX|PmGo?EX+iBDguQtUCxPS?$(LIJqPfc1-Lv zF0F<$S2%_gUb_l~4IHaYQ7oxV-&&pooYR%jskrgmHsu;KktfRI5%SBiUhqaLtJB4^ z80T8@I75x%P46gGK=KuhoXr=0wE1xg4K2DHiMP!4N6 zSjK}&WNtE*Vj11y+IBC?%stlj=(l$gW~S~Q_VMg_j;IinCyRs&h#Wd#T>P+$u03~L z{UaY1KnDgU2X^0_2M=qX>na!DH1!LwH{ak@Bpsb88bgnd`?Mgj(Rq?ss>KdZ%LZ3k zM2J+Hnhyc2p0yF*aJp;>E2Y(^y&Z& zOvdb1h_@@p^&)ly0Eo;AynR=djK)2ZH~_ zA7piEYi=HLl6%A7Xw&i%_D;U0cmX3D1zZd_doD(;qU0nkfN-><@Qxtt6)wX_xip}A zZIFuW$5J&2J;rkuA&?oR6V>RZ*`&^3qioSrG|$7!NNO4glV{fFNHTYmXUG zj{(5x(Rn-Aby%Y9Fh4J#>a58h_htjpUT7-Q=1twHo)vQ(L=s`d^gHtaeoo z4z``Db)tW_Pz5?0dFw5h$KVCEFwgav5=gC~WS_Dh$##}yRyd1W`>GF{@7N+#D}i%91ybK9Um=0nOOHE|H{&V?FR^Y%kRm zx+KXVcgi9}+Izg;81of0Eq>T0u2P-=3Zs5?4J@eD{Gu*I&j0`Q-aBODrJj!#&q&C> z0wD;1&nrFU-wz*Ki@)Hmv=xW(hO!C1zG%n0$XBiWr@SL@TVEeTzqY(P`3$q2Skf1A zTkrDQWSuxB7vT6BNqlk6zQK4S&gk5{Qq}61-D4qM^C?~(C_a4k33$F&!c-M?i5!kc zPY#hFtqg@F)s9WDRb9RfVpAlu@Dq1^4HPi=F4GM;`TQ&nwL8N8iqf*K>Q1QBDeIEc z$P+#4LF#<->Z}%$fqsdX=UjM7zt3b2m1y=|hh*f3T7Zz1LaA=*akrYP1eyF9yC+mC zY}~K@e7dIVKrK9U@JLGCR*@Ktmgmmi=4BQDPmS9J`;g(6*)HC^&uJl@ny~fS=$xXm zLcimR_6h73*>Cf0<}o?<>xkEF-0#E8I*()(N&+6w#ay5!-WO=_QWwX{=Edf2haD)t z=a7?5YSojYC!WGu#sS-dWIVVbr?#5vIVZ8TLbR9I#{cDmuMZ)cWs$|y+HOl4~>z%(|NY*Z8UDX2<%05@uf3OR2tS~ zCX$YT{eQQVOSU)krcXr#$mQA_Ku9qa^(eOu$zcxN!Lagyidp0A$=y1wl$l{q9e zD#1rriNoIoP7 zg-UZ@2eXpLU0CyT4-I#V${y=}$~DAQn#ar@aRO3%W2DFxB$E@;;f@OTxoj3}JjN>IS6!I%3%Uo9ij{bR zxW_C@0BSrrqUaGk3fQ(DwCW{!_^7n|$Ij0iGF&Wpv-SrC#!VxxAmtk19YukSk;+j7&wkX>p zEIFbMB(^BusS<~i#%IlJUcHn_J8b+ig1X^2T!IC>Mn&-F4j|EsGVnDdExY!)iVL}- zF@|ZRc4F|8giQK&toKl#E!aoZCr~iM?QVs99SVi%2z|E6T3{m1pD4D?bd0C0z~pw? zw|av4APRGkfRY7$L^pj++U>hdE~D(qpu`H_1BHDuQVJG&snWNJxn-Yr#GqZk?=-xbjQ_>Dls?d>tXr?#phrbrL_lRl5%mK&?yO)*?R zxB*=#`<_LmguAiU2husc!$rBoO8ia&D_uKUzhgA9bOsf6!`BCO#aILRL7xPc_MvNx`Tll5BkG=m zng?c$r=ew(wZ$Iu<&xez6c|H}&!R8Wz!=*;g119Sj?Mq4oC zoRkyEgB@u|Vy7|n+=wtF284i6kJ)Y8G~r!ege1?G5rl*UHvh59s!^|qmKjn`JML+3 zZgQbv^nvYV;JFX=CtZ1XVS27-P14b`)lM21?*3&&5$8!J zl^BlqE;oL3Q_!J#I zq7s+RT=Ij2bOH}g6NWs7yES^j1l?r3zj|}aD27mfLp^vEcL!&qrI05`s{P8gJeRyxYbB{B#-A>w3dgbc z4rj7Sq@xby4&=}ZBLWz&IJGF7cQbnA=q1-{+9zc+DdHhd6YNh*`*EQLgEm>C)vS)w zQkhadN&5|7Gj7p^JV*ZR1P^HqZJupB;cej`h)LiZYkvv_`?s&U23aMV@*N?ssHt(C z%iQ%#P&+orQih;{GlyeBP=}<2iasyc$h(b3!y%75%AJHf+e&i1I&ex*NNh-YNtmk7 zSwJ<;yKf{+(XveKF5AyaV&yVpw-`-zJeAK+LJ3Ex`!z#-Z4*)lAy`KyWbuLqt=dq$ zn<$>z6^$9>%L@v!QmmUDnlszPX-`9im(ddYneplWzHmn}Bd|9EWtXpFjJU@1} z+n-m7SeJxn^F@ z3Ws@#+K@w{WeHZozRG^@rogE;ki>?WscT;4F{8*>O5D%&*b57+)9NVt<++jQ@6XSt zcYa_nmAZXkzq!&mH3R+Np2 zV@+nuuY&E=GBwhY?ikSW2_*7L$aWu(Zq-=hx+h}DrXlMLb9U#nS{YW2C;9}_Q}3|W3#2l2 zr{<=xd+B+VUF3#32LXNkIG}`DUXrL`)o+jWCh*XW^=fiSPNM)#K|HOEZcusI9uZ>4 z1Te5)yax4y-&SKZU3e#5(o;k*l5QdBvG15VdW^Wz+ML{6ue~UVOY>v0T{@$(nA4@m zl!<81lN(sAu7W20z|5cm?d!DTdO2Mumu{JkRaJNwwkuK9sslD8e6jqK^Eub+*Zp*h zzCAhVo$c+>?vCk!e2;|B-^hFBK!}@6w5to`pZ8qO?jqLCr*?i73mbcvT&EA9tTY4X z7O#pmJWAKv=sBUV&}b1?nWk|6Fppyw63XqWEYqFrx?n0tf7vfnsC1?mvbbT@bC0gO zn#*(@?vMWWp??HCuU0Fk-7ZZn1~ZMnd5sN8HmjeT+mT`^dw)-_*5-;+6Ii8{enQt& z=$%t4n;EjdE&c70$xL>KpM6pIf!i=C`4D%(ql( zNW2T*Aamb4U$tuS8e8IVnA5yM4O!$ON;&mZh~K7-h5LrreB#kKd{U)lgSTL(P>e0T zAWCUEu!ml)b~}kT{n$UpO+^<32h939gUtgp;g+SJT~4dwD(Ac3$3T-3t|J7*^1mmE zPV1oun(J6?9F>!K(K4ByQ&7zkve~VCiymoA_s1{5Q_Dj?$j{fW8|3X!()9aDZ>R-Hvc>eApbzjL$1>Wy>Z`6B!Ktc7 z|IDH!v-;GsJt=)pVd0GBl2}8wb*a8ywO$aoq7*Zm5(XHH6d#Xe)w;lWp@b=H@Y4jD z)S>R7u<{<4xWXOv^IF!+tGyO#&L2(hK@ zd9Je*g5_}M8IHb8PgLlScDJ>Hr>26H>F*Oof39!d$JDQ0iLr<(hfND8t4J>seQlMh zZS!E?AOzNw!OCa~oOv`Ct&n|*E0R(mA6c%;nM|@QNdw-ymCIKl+#H_9Tu0-s&<>vcY(;i>U<bVbfJmCdAuF-Znec z^P)(I*E5@V?*dEc(I8V4#Q)XRIo!LaJ~1|)4ej=Pd{erdE+K)Ay-#m`KWpBf|J0L| z87t@DXpS7OyK;)XCk*Iu(w?@v;S}pTDrYEfJ6d1n_`q=NEJmfeEWwt981_21j1Y^n_gy<{;0B>VO<-@u6e}>hqFsd4``x%FbvgIYbHZ*TPMXy@ zq4nm+BbEfYB0N)U+E$-n0ir#7<8L}OERG>{_f-h;Z}z|~aU^N!4kwjiLO{Q^U5U%M z6l%5ra#*IyN+~LJ_JRGvTccEMD*3Lxx2}Fak{+L}H4xA797>i?D40I3&X*3yAb?9w zJ~~;LbA3MT=;ZwZtv8pAMrKYi?ofPax20;0m9JyLO8SKPuvIywg2G(fWb^(Z2@*L( zMOG9e<%U?50n!LT`mk0OQF4t7wQ?zBiBXN4h^p@4+GTKpad%;1iWXb5?U$`y4FODC zIZ$Ijawn`ibd7Q4I7$P}Wa|$y0vpBp9m*$O4&LGr z^n6v^G6^?m3@os?M!{uJQ+=0KnA2H~wDDLin5?P7oU8LDDu^E}@h-f1a`t<8So37F zTDj-w)q+%S2NC^A9zdi%GD5gNRqp@%(x~iYT-gEZHf777iXTXGxJ>CHw$@8_iEoiK zgDE3-`fi`!(P;4|N-d00jsRrQL7FQK56~V0h7WU=50K|ryrJCBXiRylOT#%HX-3DC zAV7xg4h~6l9hQkKak)})FfFHPCC_6dT;9o`PtH{pQ@Yn>^63V(#~4fnfjzW2>9`yt zztB*Yz|R=>%BqkCZ!;G%s`3*yment}{{y&rzk7*8D|milqburif@l{4HUVdUf?QVE zzTCWgV`JOy4*9L=-{uBxX9!vb;MxzL+wHk<`|A9VL`NO>?V&z#;=jjoFME(T^$|rK z3HN8xJLF+v?HyZ(4uOx@$zC%ud5dv)rZ}c6I>E3Th17%z5YP0B5ELb$3%|)3AF{w$*P0i0N_DFpTb#6-hxCt7VO*^r_X_oKbPNV zca-H2Ec*&AoSI$ga;3&@bJV@nEZ!%VHNr&#3cBfd1V?9 zLVfWP6&wR6))2C3vq`sHl{%W?IJUK1F@&HahNnDeNGRIhqoUp0GEoF`#!ZM{kAgY8 z+C$ab)}2sH*%RH#7duSl78|pJyGM%JZ%+&Anx-h$6triUrT~<;y+M@0YQd{@2*JBF zV0J)D+sK%*1|782rwoahLdb;=S7)k#3W>A#wnolk@)ek|;5Qe886n*dw2kjs*4hfz z`OtbM6t=s$)>!9>_Nee`X7MzX;LDu?hOnmFs2Al^s-4#X`ovQ*Tif7jH%ML#j-5W= z!+&;~GiLhPie<>g@KS?$3d476mV1Cw&wwLAPDB&V4k`B9bSBOwM5gQr(AQReidWAw zMUT$NBfjRDMxI{W2JDs}G%RO7zn8=nierWBl36R+E%q}$u@JD#X~uiyn%#_a_eK>Z zcJEHXQ3}DB!*-la{lZ!=NfE(31b(0z1teePdc5p0Z?KdA6!EhlIxvgTVhuidqEw4- zA%wXsWEaydrFD*QHi>thmRel4b+bWC+Is5B zPt?j=381%5#LD?^)odHJ#>e{l`YTjHrgt7sb>IMwaGtsUQyL1y zEk(dn=FN^0Uz$h|x?YFZ`lJ5y7cKlQ=6QAe3e#Z`G~qQL;Edq@H7HgjCll>8>;nxu z<=X?qrzd~YpD2f?aKi|>%Vyvv`UcC0am5^qA29AOxTdc^eHQGwxGr0JeuRT$5Lam_ zo3tS3hihqcNmEB+YOFIKqQEh%rehf+oA2Ikn58)jEkH^&mh;93&4$}r$sEPHJx4n5 zh3h~cRS&(|T$fsP19$0}OQI{SgFB1fs=;W9R4VaxxlHzpvuZFp? zZqL{?Tmiv!fOt~ZRVIM*L*)CXwU9_Rt0L-hHmXsX8WFvR-oSSx9={f4qE8DutT~l> zYS>G^@)O`11qq2E~(@ zMcz8b@B7j1i`zuWc28%|vkw{F%k6A2P6I0?WxCMveCpB3Y!zTAiiE-2F(ZA)q!XIg zUNN&MzH}NTa_Hxe8Xan%&o-!$z-yxoL<}v@Ce0^mh1CLmVzIRtyfb6Fzsu*eWf*jE zd>nr5T*J*vGO(HH^rxp+|L~RdN}SAg&p-Q5XfuVm*Nwu7PuI*lWnLlyAyFdYRJ3}5 zoCvKOW;RHqY(DgAjzSVS=_(nz2vLh%p+B-;yMS9qZC_A7)dE1=LfL{SQGfd0*Xt3? zo#@VNP#{#Jg%i;%7IyL*%3E^~l#1}!RFandU1A3WJ=-@i&ty9-MU{Tjix}3Jc;ZP+ zvDHGAwV2_UvMF?N<#NbkbG37tuz7Jto;DSUW`XJ2z7FWIRl~s))|#gR8qT_c*kVN? zJ*}H#9wP$N84y=}J=T|0U8QUp4dq3Tt%;dZADcYMe({Pbt7(0%!EYj60f;eQo7&Ot zhk$iMh1RubZ7H@jnI`veDOABnhQ0OwoZjbuGrbwG7_sSqPUNl6Cm(NEyAuoy))lZ% zEf&3JtJhb+n<^N+&Z9L6mBB=RxQhNItMq?eNN9-Xm~~Uv>|430{d44@Jd@fLEzrtP zw?W=j;K{fqU)nB<1?dEA13hL|ZZ56U-nSH(#dK0_giMJ~M)g%%FF%!11G&WwE2arB zP@D6=TnC!+6sGuAx?)ajyU$Fy#yPpBpT-p3(uUo_9@&32`V;#`EhS#%(pGi}dGE`bwD(Zm|vqc zprDmE#hXAB{Mr}icyP|R{_j3^Xx2jeJ{#M+a?c{WLm{LnL3i&_>{XAgGWQ1}Dv}r% zPc7hBjg}zNaGw5+R}Xeb->Cenw_Scg5qmz3`2W6Ib)BXfRQtGy6D*`HCdkUG*bLTe z_eR&=6vPiXD3#bg;2cu{RlY@!c7&j4qI~{iBUY=PF>C-_Q5dg;@nlef=2YSa0LNV( zoi*1-*D9sBqJRQM6`eE>7%-ct11!Y$CHfN;#3&V%_{J5g@dp3v%NKhKK8cgr`K!~b}P1ObC!q_8p|1j=Na8(RI^@D z><}ya_$B{Y^n&?zPoH1a4J4-9$! z*aQ6FGw3J+b3QH6p`048kKO+4OS`9GC8T#P;>$E3%ujJk`DA39cVE&P^`=0?%8Nv& zDTL|(IY+$Sc zHCnHQ#h3+O`^(x`KV&NljlMpUIV9YCnKt;bBwj)6hJ(u1T3Zog}p@R%mK9skO-WWXj{d3k7l(j&w;+}rX@?81fl9piqRXe z6FQv%S%|3z+nL(!MQmYjcSExc5NDBk$IQd5b{ouV1=CulM^*=s#ucqiw@G|v^F(tp<)bTlzum`QKl2i*TBr{Nz192C&$exP30kn4bvwAe#8D3%vCnW; zaIB$)3I;MF@!E_+FqPR}pIndi6ZTWxORa!J>+|;1&HTlq7b<~O7jaF**9cugknc0T zA<2&DUqVAwRMCP~v7jc!9N7Jm{7IU0j1@{KEUl3arEYI)fC>|zSdVOc3-&#i& zWc}RbIo-h7C>Fru8RE4+cEGCZvw>dY&+bL%Bj7UIuB!6rnl*V4&L`Uv-QgN#>?V=0uV#bLf!?GQDuBmtknL*>$&hc+h&KFkY zq!)N*YD`h8;2~oQHGf&(Gwi{r6-6;XUu40UsDU*{q6x@&BQWQK7hb!~Iiiq|&h%3$ z1v5ck)y?+NHT=s4w_ITh9YQg_>(2^Ui@(QCM-F+_5H_eM5q*}rg@hF38KONMiYX+I zW<+P+f+zE}C;R+q+*|q*gC;*&`u~~c4}XNZE&qCq$RN=Luf^>(en{jyI+Yu}EG6Xt zn52j^V1T!jfzcs=aViqgo_<0+TB#HAv6Vv+?2QCrWOzazP2VMiaM%UZ+l=YI+Fj3M zPSJ?z$)_O@=&Wsr%8tKG@sYg$FB*KmRqC5P<`ZxGYuLgn$zJE1!}F)b$-J>lkrD=}LXNx9k7RQ0U@fzgDL^%O|dn>zEJS z&EI`3#;(j?b>;vn>t1czOtZR7aWdlA|$Ch3yIC)ZFcBhnNn# zWLo1|dx@y_$fIvL+$0~!+rDsJhqldnA`P%VVgAv0y#L(+_ciHmqgO||m{Tw2M2hWB zX#D8_LpybW^!ef?QC}h&p)+^wZjhzicGH3Rht*CyUdl6kqWzWCx2M7dHYZyQSU;mx zygQ!G7M&M9)Xj}!e(Y@M9ihMJx5%V%;F~?K2R$>c%{Y-1eNPPjUPD zzU>IB&5^ofJDRVEAIv*s9G|((=0Sh0zHMXM^P(FMG>ikiqfKUOt_yp2{9(RjBpy{? z@u~5jr!z=vZ+YkXxH3vMmqhDJA0Cp*HHH=ba0%W~WcTPkC z_CxEa{VD1|WPRysJWzD$M8v|ZZghy5{rmnwbEboMl<}X}_!g4V-D6 ztD0-Kn2Fp+400958()d$dM%1P3VQpB?HjKXVAHr&Uw@&!`}}j0E=*FUHdnC|-*j{F zfOJN$$hB;g1WIhpRtR5WY?M&I;v#=JDE?qT#mQERZ{DTqvwpsx=K9tN_Od=x z`79Zo+Tfwgt$~sJ*-&}p?nI%RYNaKjsQ*Ae(tUB=P-cR&yYEC( zg8jbqq4J|@xU6rD&-wXEew=|@!fZ$IgFH8(V`Aj@^2Dp^=8_(UihcimgLj8jQ*U07 z;xl@sgOPNOQ+!*dbi=(ZYqAeX5FnT48TQlPGi-yomRk3aueTFS;Cy&kjXOEUE`Rnw zjxOq_*|J)B_wHkLeOyiF>UiAymCF;i!XN$B_W8eQn@Cg5Tw6(*xypSRx;}84+WdXO zxwpRJOS;*pdu`IuUK4q~<+}E}iNoc8&{qpuqr|xyBMyJ)c&PgKZ0Du%MbTX9H5PHS z{ORuTtXgLL>O^P+xBFJO1y~biN8tSOYp}9C;sVMm83utf z{RT_xsxv$&gI^w>d^@3XqnTG{SvPxkC!Rb0EiQ<9ChspBANt21?2Wv{8QwL{$+e|I zr)EvAk;RbK zE4+-w_g3fcYY2Fh4j$Vbins&kD|}Le+=RS4if}jP6zfaWPM^yB%bc;2*!z(EMB-*K zPxh>t7Us;2RV7mN4l&0U8zTN78%J)z+(h`6O{A=PU4zc&o4iP(JDsUZ{nP!0ZXN#{ zsa##@?pVBo$piN}mqUq=^>aT4phE_%ZHCvdj>6*IT7?{`+qNg8WrGMN&azGGMjxj| zP;ft+@^#ehN#2yW2{J|mX_%q?BB1a_i|+yzF*Cp6t7>^It4lHR)`Q9`H8-3ao?V&w z1VY!uCv(^schnJ{Fjs>+vA(%a@JGsn_bnFlm%r^*aaF}yKEZ!wb)?=~N1^^^fP!~R z(ME69g+&-$&L?Vj+^vHx7LYK4r{{jlKb8|CZ>?D$q$7l&qc{xWt3#TqS1Tr@R9v(1 zM=Iu~$cx38+`rmgIC}YWz-xbV^=kP9=iNnf*D4>Shzl~@v&{(UDRI9be1XoX)Ej-* zesW?xDfk#NrB;&ul&QFPhxeu9$}#e`eeq(5`H_BWN_Q$AU0U2Q|4(Q!b5va3op$<` zC&9VM?vUkJDPP>RcnD8FT$^IzSs#D>+{=+T--eoRw`#0Uc(){uta_=eG(6e2ihT#d zGOkkf?iXzqmD3c~+GY64wiwl>{Ut`%+u6@t)h0UPa%$Oj(~8T1`#8j1U-0UzfDZj< zU&3bYexcF9%a}6)+4B|=T6PK7K1H{UMSru{&I?12zvU+tBP)`&a1oB}y)?4!2h!R2 zz-RITMt1YI!GW9$m&ONOP8F|qhW*v>A7u7Y6u3ZwhKC*@4Id@_v1;}@@`JHVjEPt5 zxoeGeQf>IjTr!YuX%U`LDVF_-WD4E)H%M3xf#ljF^+?0Rb$_f{_C{t#`+x({6ovzX zTR7s+Wf(pYh;@`}Mn{%H!1#Qx2E^xntjya=l- z08hN^ebUD~-`P#d+Vpz-)pqG4_VZz(=3;(F zK%m}BO~Jb(F*Q&_!KChwriy$pU`HHu6bGXByE-)Mq|G8Ajd$w7+b6uFPi9MV?*+EP zmpBqNTJtUXko(NsT%WKtZHPGhYG$LMfg|+3aJYb2a1nR*NKZILA;;;GYm;>8+PX#g z$%K+WVT^>>?Q3jv55U{&Z@9zc1y@8Oh^Jz#ENerHo|{(q>U7uO&owPeTk=T3McP5+DF- z|7eU_&6%y|mqg9`0c4&sz8HvJHNoBR5&jdyN9O%a^Q{%0mlqe9IuQxoUDvWB+bid% zU$j@2WmqCWPwy3e@8Q zq&om&?ET@}RaX7sR?D%-uq~q}+FYd}!(JyhnbcGFZSh{8RvP;^nqY1m-^@{&^I{~) zIMvTuWO$=bHN*3D6Wl0V)`eiAr(rz96Z$%I>E`v5+C!N~ zji*FEN~@Izd)wLgQKb5Fv&6XSsPW5hHw zaL3op8QpxC4qgpXejq)+?WjRTEG81m%Pf5Lj@l zWfhIwG%?lgbqS)b${xBRVjKE0-uS1Atk&%yVL|k1Gp<}C#}n1kywMrpPA>2En;jDH z?x-4q=)0cQVgeM&pOlADLEXBX$){QX_Ds0`?XHq ztqL|qP{Px*^4kHYHZJF2fyGSnBE2Tz!Q03>r>m!J|MZRD-T5jg<`an8UZVf~r%|ui z!9osiZWLXf3w|nl=&md)*^b5BfdQd8Mr)JWkR6Y?^ubhA4F_ur<40M-^-;Tt1j`BDm{s;F!Db+k}a zZT-P*s&1U12<*jPSbxe0Q5_>Wf6i1Wy7X|LbvKZspQc~%w`W1WeEQ=vcW8WYwl=Mq z!R()SP2Dqpxt^bu?uuKOvBSL-1{`$p;)^}e-qz1lMJCe5fJBe11(*d?kbl=xyFZuP z>iLn80mIL<{pl@DW2rpj)XE}0(+0nr1GJ6zGIf4L&ZEVTuYzyLDspZ3L!9;G;6hz4 z7oh?yT?4w5hNxD5?q4Eh^&{6v0^55yIvJG62Ao&&Nz}cJk-VV-Cm8QF!AYk16++yY z_-mTQP_(bv=pRzS{py&K%ebElMDfJRL#5QeG%@ied+n1j>FqQQ3`jSbEkUTwr%LP2 z`#8j0U)(}s$53Iikd7D3)ZMC$_TZTZ`ajQiiPqNw?}V`n68Sh2ENQ{d*_TiAw7YEh z%Au`u4Tow8rx9y9R(IvaGayXx*~ksl3oQ>o^hjJ>=}cK%>Jae8pIQtrFOdxy^T#>V zye|0I%6rQ*`fSjQ)wsTfacaHc>8S=BFw1?LeNnPYv`c2I4xru=>IJjgq_ynwiqBF+ zpLAY}wzzO}Glyn`oRbmp6JZz<(aw$&i~TY4L9^lN)8=3-53&>m7BBuEd+z}i)z!9v zPGS5~XC?^_1*E;*Y1^lGUJyHTlCC3H&u zv5kqkPx?%3nU~=Poh^9jq(eq$Q7*q9ukCN_io1G3RCVec;(GxH`*t>Z9PnoFX^o>r z`08(eP<X2+$S||gWo3NKXJiKhx_B@w$_uDvmqOcCK$&!254^!Z^@G?CO4e!e* z`j<2;We2_5W3#ih=HD=^Oj`xMP6tH5@xW+Og)*6mTqyuv|O>=L#0|A8J8paZg z+)+ma6zhod>Te=Gxmnc93|19$kJWwVMuy`NVo1>@Br}01C=N=dapLJ#2T6NOLDpVo zQw=1gnicNyy~AcwYPQlSgZ@_XCR&{W!)j7KA@-18r_(synNH#G*;Bi_pCh_ZLrm0M zZ<8mUT$}owROI@x0&>7MSET8A^4>IO9Mk9qKqPr1y1Uy)AYusk3M!ykp9x&=cL3J+ z6}5aSX$)*n-z8^wZ)|~%KsdA4*Y110Y+&)^ol6tzT+xQe&GF4CyXQ1L1<^pFs~YD? z(*++^^qna)okm+xn$rP0lV`4?Wg3!NV+*+O?xyYo+?i6^5h+A z*H!&g38dwXZ}lxbkgtM0(rRb80S%UkdJdU)W;MkGfN~Mbhr zamxW$hp8x-yE;7@#tg-;?vJ%_y#S`>yBu?1Pr`?2X=;BPV|gEz?lUl9W4@jUN!pW_ z(nXzgD+%+um1$d;A%aZAalXbHNk;8KF@Jhc0<0yre>*bwWrDV+90VlCY@Jk|F>xaT zL+ZqF@=@YJeWmUIQbDWs%#GwZf$<_~S_d^2Wn_vf}P>h1574p-RhZe zrp?HqC|;2TSCQfoW^Ti5>X-nou2j8Hoi3Q3_mp+0K&<^2Es{weD9V#dez(C z27LS3*U=|Q0;2nhfMU9fl=!4O(kTPK>0X(2WbW@$IE#&gDOp+9(?c*-sB6F}k7QV{ zxXnfd*&HS_6jRLk*0_LA^gBQK!zL4d`%#?kp7MyCtNBAY;VPJ`{it*4=@xD*L^yza zTAy+wGO*FbqX;w0Q*3yglKzuRTmJ0viZ>H)UyXOmb;&jD4E{#!=eHCT-`zca;~ef= zFJ7-s?OXeOnLJMK&1|wg>qo1#mrMPk=Kg)(3wE=|{O9-+4(>xCcT_d6{7LcM%U?~7 zymsYyrSl?gIBj-*SYP#NLxWnEhQjRR5>G=@&x)MYsIt-cLbE*$t~5$4hU8rxu^oJ!dS`@}3M-CRp0V1U64>0??OnT+0xtHlVW3-s zPfVEzy^?47+g@i=QXRj^OKrtFPnKoc%=SN&cDcWFiuz6egRYZKYVa#k)Ldp;PI&R; zO9r}Zr|uT6ENs#mAna9(N5>Hf&dIqUf>OYlA4+ zPYt_7m!{t*Hd4!`IWDr+s4Dp`ThDaiml;9XReDWZ@o5|yOLMo?1>cU0Kltk~jS_8I z=xx}<1I3}Di?w&|y2rpRnJ12yl3?EZ4nXSAiUmLUr4PZPLtzGs_0?TD#Y*M(edELv zL%{Pp`&OgRh_RcLy^tRAJEbyRdU**;G2NOpU_nA(17ArOclzlt+Btp{^|oyugRswe z>tn4wE;=})KeFPC)lZ)U0(7Wd7SR9}^ zdXq9jIExN5A}57D!`M@pPsN@eLp^JH5(@F{Pv^1j|EnV}mNQGgX&(!2KyRmb(0bzF zdV>5MMahvkxwN^KXNEYB<@?v7t6ra)rNa5`fW&q?7+8smMcLsl+a4}hu%MDLIHR=G z@NoKd+Il9Ma)5j^;D|YCM>%d^bF*OjK_4QY!B;v3?~JN*a78drd4!?a+4Hoht^Tq=}c4mUMyxlIgr@#kSPj&U2#J;EE!Vcv_;J@<_hy`U)40UF1EH;(t~KW$*QS1x_DKdd~K^NDA#HlK=cb za01Lb-4=~)V8QMR&eH6UJY)5SvT)=?lj8Eb$VSq##*(vNX?rB^f_u~4lDT)QHLq30 zRHmm}@!X8I2-=uljsBuxla0i3x>{blI^oZoIt@=Sdx=lU`upYfXBJXY? zZCsl65CVNhAO$Xor2`jf#Qco4vXL+4xZ_y4wB!7M>5$i{_999|EgR$REK-xA;|1~) z9ycP}%%U%c73WdNv&nIxDMz}BBHehL z@(+}u!ZcyvkNc>qOOcipp{8Em7&jz(vkgVAox0i?M~hKv2u4!WwmtLX*AXk4PFD|e zH0PuXVU>+8`6De-uH2~kEVj=Y{DH#^U7+nG=RH|H((_J!QIm88J;O}umKXDQf|`CHz5Hd?Y(ZUZ^gf& z?>iwCD|VT=Supw*Z>}w`biyoO4k$X90ZiYSP}|e!&fe1rGsnkVCKLL2^;<5l&44_q zCGQGS6t$HopKafTF%WNLsJ+%A*|ROtO{6G>TA{lQpu@JoHv;*%@c52c&KK`m3**cu zzB#`(BkHR36(?CX5Vg3g4C9j|p6HjSy})h5sGk1ffX3=2m=@aU)xonLS8kU1^tf!W zahtk;ws|{6=0-2ZvEwUEKX^D(k4xA_l`)G30D}{nq2FbBQQL9gKFqgpjGGt-;zG>0 zv3=k0s)@x>=+Sk#9oy)`g1%x(sr=g6m%}VH8Liux5K|h@%{xp#H>Mq+1Ew?(C7tZW z;~0Nn-=5v&d=o2-hN-hY<>6}~TX#)6Nm{z`cr;^NI*`}JiZMLrJTQS`HF-t*Ru@;L z#j=lsz872z(kWPstC;g$9g32E?58jyvY&ZGD1fxI5cNpsEBZ2G-NWr`Aw%mPEu@=* z*%>~IH7#&XI8ZH(wA`D$6V#;(uP!nUl5{Q>tvA=y<;L#oTTKC%IGWcNoxeI|_k%|G zw2ct-rI)9=eA+Qzq5Ni4G1c@KepxU|HdTw!GhIf%l!ncu%w}vPs^)Q4C6hhwjb|vX zLhDefQ}wQEl-VrfBjqBhU5FEs7^a|CQ}R`gC}t+57X#6TTd`4(*oWk{AzLckcw1FI zz4ZD4;m1wq_&av$;BZvU*ktwyFHh%sDcB*orAxL{g59`W&o>d1Br(y*_n>r?#={ny zt$Z~gRw9el)2$&w!?!Ejs)#gRn!g8c9G@@Gs~BGevI@^FDHkm4xrc;flRf)!{X0U; z7w8Wz%@oc7LE<&m=KpjNdqYouJEv!$1-S8nF(Oi3ibd(BaO8Vvo z8I7)1AU1wazO=Hq+lyuF!wc!#O4+%xc>kKUOLQpU1fOk-V7JYMQ)$MoBHBbrxEhdO z8GtJ8@SHvD;hPvIUKj$F9@6C|Th!*U!wiYYN(~JBh5<|Mjv{En9D=MQf41W zU>U_i*M?W8<;x-6_3xP zcq8pJaI@e>t>KzM9{I{@0S>GDQx#YiX#4H_oii8zO73p=Zc%IZ(2bTvf7@U^BZcN& zb^x3xJ0re$9h)gbe|aq|S8cmfxJW|wq^@i{QbtBivm!#@?c#V0^ioy0&Qrnh88wZH z&d~l=L%>@==7dS2PwQ+S^}{UgP8Qt=aqYe`rpAR2BUy!u;^^XMV?nPU(kHO@Ki|QF z>GFli-y3g|3esuJ3XRXaY5YYt15p`eIbUV2^;y8GedL|s4f}g1l^kz~o|3j}hq+m5 zrR+$rV8=x2VQ!>&Pq3}BVT^Je^@Hr;i(LKt*b^{m6k3u*FXXC`?!QqIU!MP<7_Nhx zd1Z*3(c1;@nSQ8Yd=mS72RCNV*hjF~9s|gE)C*rwhvvshykH_JjKaFYOiGfU#Cen} z8!k*suX@#U2qkl}lb0ID+@oJUcLCV&rJ#KJ6aDYZ&T5@!v8re+1V~`}G8`7A5!+ZI zc($a&>&pgj`0l!DI_-bI9~(bZ7xMyaNuH-av&Xos#dH*GRm1LyQh5c`#m*fF%5FR@ z++^Ezdqc>#J4xPBr)nuG0wPR&<-({@oV-}Midp!l9m~v5S=8f%g zRN`D;A>zn7OfFICfk#T@ZykiKI`AKkeDav$+^Ud)0o`D-X=%5_b_dHo<`(PzMjXLV zS6EqFb7bySuGY zFjVR{Rd`k$6Apw5fJITdr?Fzu3ex2v9l)t%y}9p$xEY1D)lR5wH&Sf&lj0^1MXsng!J_1B(VB6+Znsl!{9NV-NDJ1CrV2+6qwnq`BlsvMFH zFa|A>VpoG=We;A_n7hh6)^_&bctoUNe}dfddxc8UT7YTWb~fNth~<=^$S?LleH1ha zz8vPze{rP91RwfRv-Y{zwP|NKfX4q&^=kA?&BCv+jx#`;!hGTCwCxx+Zfy$J^&9dt zBdB`&{(8Y0jV?Flao$#~-RujOcvg(1Sm+Fbp%X_UlskNOB{m+Ao!nC!Uv}C-ZXJ>B zD7i7dpG-!ITF?wOnV5Uy?rmG1bVq3D(@?%WZo)Ifyk`KkeOvE}E;W=6Gd{=6=3n*A zVWT%_JEhyp?qj6cdXl&|9;d3UGr>w78GY=3L&9q3R3(G@Ql{>bvDQQWyLMw*E!kSK z%vNPVXvdN(1-RY~+{L%#Q(e@6uRWJ1r*9JG(xSX902%b2V4#h-G40cZjVHDxss!NB z0(s`;R-?8ZI*Gj4JZr~6A5%-uGgj?hUCS3KG;(4Agm_A~<{xsI9qHScjHS5loEbZ^ zu^V5UKN*j+ql`X$)l-L2DGYZ<8kjwYCkJp7$vjvfYIrqsDST9nATCNDHY;m)@a$P= zYoPbXY?p7;N{+JoBW-2w^=t>&VA9+&f;Lv>Un{>;TAzGXyzzrg3mJL>#bE@$O)3`5 zpd5!+BWcHK={|w=8}CCCc~>zo!Mjbz@hJ@koz2(Wts*#7YnGGkrd8WPi{017$2RO6 zVanRSL|;2TbH};@Y3#wIofFg~U{s~>6`ZlcKkwQ%_vvcp!9%@2x)q-3PiBko;!H9U zR|<-?7%0TKn|k?VKJ8ju_BEPi5W^^+7Oq2wrU>&dw_8sQ;8cE(h_Z8>cqHbva|1p2 zeO$;^bJ+u)-ibivxI~A_O|n|Y(1;-mGShXqlLmRL7_z8`hW&_%*tYSx!uy&AjGisN znVO(W)l*KN+N~3Mu9pi$aMA{LI(3M13fFnprOIb`#tIdwU4SO`SIJE)raE$zYp;AL zQdhm@0C8eGV=+s~aV95kocMDkl&ccgK>;KI{Ni(Cx1*C9Go5~rlbQuaj54{?B`4Vk zv9MxpnwCibUAX5_b{Z{~$jWHgu=zjX>Ec{Pw!bm1jQGiS_6&^|w&hyR`0^tJ=UniC zOxr5%*btW&LuTi6iknGeL-CnyT9itDUYW{x)gSll2@2zl0O{e`4$ z0;<^Z5hncaKCUO`?_jR?cQ$#qYIOh$=cQ_;!0y9mW1%`TX*v{Orv=LLOe;~Js$p^2 zngXdG=RUt2q7DR3QqHqre(cp#)E;wH=F_bf+p%IKhufxipPWbG&~X@hfSHe{NgbEI zBeWNHTfX!Jz^?q{5U&c;$NbE(L`6}nGDedsHFf!GA%V4WuEa(9MUCzCSAlN6#VK@) z?D4Qdh1fYt_m|zvj&!^dt|8kxUqMX|KDtJQs9al2Os9b`+FG)Y}uR!YWe z0bJ;a_Oquid9&zCTKBhxVAj(1_kvJnuMvg5_3sdM^aFkp07!1 zor652RIQ1SF+ZIQHHU-*nfa&*qmKAc{y~wYUDTz;go*;heK6v~j7=Z)@i%9L=OQ3u z@f~uy-=g_`BeZ4sN+kY4HZi13T%j4OBZH(AsTJy_JsNz}$%l=~wpBx=2{gn?%SIA< z@alvZBYK;0lti-?HfT}rV?|9#?ET#siCLQUJc;$UuXuo)auclGiv_Uu0etldC)~Km zc-iSeQ=jn5-2<)JXL62e@{GpHBC_6ruQ#i zGomD2Uw*BhDnUB9p!0^fvB72riNg)sXJHa2ut&d4#L4DNd*K-?X$@~!_G6iFU{ur8 zhhR8(x3f*yW72H2y!6iG`1oAb`Fo6&lz&l^v}fm2^58p#qd;>ta_;pT`-Y}6<1Otd4QHv$ZcF|H8TQFV=hh8^*tIE0T@$A8+9F&o7 z6^1NS*4$PaM7lPKL~d*Yrdg}v%8=g*Oi?iSh?rS8PGbm)ltJn#Lp82rty=rebhDw- z1u5vK^-v_9RwON6PjUN7wtn}0icp++jk~uZw_lX>N6%^?Gk{*8(s=7Nt13iI)!poF zMdR_|fX-0Hi8U>5_aPbWSQzbipEWhSVj$dK%fuUAJXQK@zWh3h_f+K~`)ksaPl7sw zmmRIF7}c%d$Cr7jlIRAimSpz+h3T^76`^b1t!9IQGaB;l%QiJyt2%ZgnP+_s!;d8_ zNs$z-bwX?Kw|qBHy0XXdL#dkcBR)+_?{&~5)=o$_;i)P<%YNqdW9 z$?p@q8n^N^C=LP&1x22;o1~Wu{aulk+n%GU8$Vly`e-Wqa1>{ zax8SRi0IRVYSkAA;^9ep@cBE|#zRw1ZPc*?6bPbQ(@zuy#py)>(vP9GkPgaK)I|HZCoo?G z%yqPGL=53cY}A^32Ez;ay3IvdnCMt_M{S9t<1n+kj%t)a6O0qC6eTTc?4iPyh4MS8 zrPE8dXCJD}=DEC|SkWW;5zVdYBwf>0w&v2uVjxdzzO3Z%vNPx5QJJ*_Ut7292*UeV*ruzNdSz37If>-gJsV zc=DAr?R=ROE|;dG?qgts@?cT5C`iE@U=RkV%kqe11JV2s79*@iBv)T!n;2nwynjhW zuSmMM@YZT%s}7CU1yU@Xf3_w5A@v)jsH(ow7};uf)RV_0|A2pCBbR|mn3wIW+$itj z3+aO86lL1^xy#y0GcmrW-kdR2)F^GTneH!Nv^Q^X675Wn4a4@g)osKR%jTi<3+e(M zb~Apg)IPSOw_mMmqw}jN8~wGaw>Xh(cq0^#yNjhUf5Q`s>(w3#nKJEB1WGldpU>es zp4o+8vMmki(J1SL*+?)$4$eFV(@lze`2k4a$<)lFD(-;}^4NZ> z!koDo(+r2$$^o&BPNLbnYpK^IOE1=N!P-z87wy8aKI71vO@6Y(wajSz@Z{XkkeHyI zI3wk1@l!9}!65Yl`)X~+(JZ7g&jEQtWpfv7-V|Gmb`|FI*UZoUR@nC>vQ_sueh-3A ztF8I@2Rp{{ys)+E4%H!C(NXO6Jjx+WLCUfrMeZ3=r;hLVQ?x|!!Tufw+;1FOJAe4* zeE*JLu6i9+J($ksMKu(7A>iO%+KaT_?nh4JyyNswzW@1LAL^8wC4bEb?gjVH?cv!% zlRXZ#gBE;)VO%}hGu7SmXgj>8T2=H~|WqAV0q`RUpgM$Fi^FUG^|oGq+xKImBr z#rVdIRH-@j9Tl;!R9@Sm!!>%Uii$)EY#7^{UY*YGGw^%R+zi@{U0xDusak%Bzai|l zXEXDBp@PY~Q!YjRB8!dAA89E|##q}r+bei8KITV9^g6CG&raBAJUaY)+hf>-BQ!yV zd2LT3RQNTyt#XXD%8y)X9=_x}yuH1kuQ{{+1gl$@Qd(}N|ERx%%>1B+-N=z*Tq?#l zXJ{(irs8OFZYzj(4N8XQN~le?xbX=p?2MEtV(W;@(gxCjX@>v{d(E{q)%&s!M4C65QF=X->Ds;jJ}GlZ*3M zmkx(}E^_6G!3xx-kAl{$$yTi9)>_9mpSa$&T$M7Kf3?Un5w?RFT1*^SXmMJ;%L9#- zcy#?GnU9G!JdY&{bV*LEOg@mb_-u2LaK)D2MwPGnJ{|8vg>H^hz!kNkkA3R>T~+)Pma5cia9ULfEs;ze#uWGCw^_G7u3!2uPBS2mJ@%4z?fR?cyV5} z4vl|3-vj!**cuxl2j#UH(Oxe(vOBGzYRWv55+@_Kb5Ho0D)ZmfN=WkKagkZu8WYu< zaQ!O+^T$jVC>4&&HnOJtN*CXyg}>XOg(K zEnx|tg)}KG*{UB235I7XLdT3JR$7eTWwBK8o}nKvFcV^K-ry~29!nhB>RIBo8}GZU z%g(w?w0jrKMh)i<{c-6dZBT1va>PJXOVhf~sqq{pM?3W@#xVJJx-=-On>(SwHO{|& zwK(rtJMw-wKdMrZk}Ba?v3-yIx|^SrC%eTQiuW>&kw0t*nTh-x8KZwUUd&O+EIm5- zQZ`-C)=*@Ri2Z{*^$%H}N4Wg%Kd7p>wk;lt*NpsLd#=|)J|4Bv5UTD{kcn7oZ?V}k z!LAYVKB71yV!KTBL~pFp3bsy3cl-C}PmIk6%h|n3G+LL!z*Dk&pP^DNvKcew4i6cO z<=HR1YbnANZ=Y{ruZ2>Vq18c11#cPCN9gzOeE(X~Z>in0MOMKd_IV_JEUuE5ilQZq ztldkwhJxO>kNz4RZ33?`!r2kCj*dJLOA`q<4&NWem?>3`DoD$$QjzpO=mPm@af3Q@ z`Y8LyA;(WvH1pHrDz)i6Oi}*4-vviJdIz3-L+9g8k~3^nO4W;|qaqVanZ~QF?y8Fk zphyEF!DA-3>hdB;zp+VC{)(52?=@Ss5Q$hEzf(_LRp(TT9%$le-q41mue4z#OJ+<5 zz(#ugFI7^UQr*-2(~?XZMXW$R7%A!^!3^Sim(s_*10VH@^DC^VLRUyUw!D8li4Y|X z+Qy@+MTXNYb_iZ==s>pL!;|uk*kREd^k=u!*T%vmqO)?KS=w4JO5#yAyHL>5wBGD` zkf9CvO596p6V45;PDGE?Tn@f@a7MQR$^wb3ZHtZK-7oiXV~f8$chv!yKyOb!0pObNLkmv1 z(Zdet_)iM|mOI;}92cJWq+T>L2QbyHZs90*@TrB94Z|9=f?bl2&KJqK3@PW?RQl9_z_n$r{|G#-EskrpJqoDK- z>ovVRuvq}>q9`%ww9ozVqF+bSr!^oge}{#voiyak<~DRX-=%(y9r0m_>vRApK_aF~ zRCE3$Ij{9v*3Wpo^FrPUIfd#f=l!`i#9$)4s5&DAc_A6$^T_inHgYBhSAP=A(tfu2 zh;QJF%jF6gcHL3VpNNl_4Bh;ZRMdK%>t{FbEhN?OW*D&kMz=#M(Cd6?QM|<30+_16 zNvic*H=VhrzqPR{kCf6na4U*S?#ynJg(SjD43IlHY@{_dJ+a|J&9$B;>igW@I6!&m(x-04Zf>-8SPtpFt&UZf2$C*)9rbpTq#xeuCMRhQaVG^m z+V->Zj8w|9$B>kFXg$7**BFpn6;TpK?68mO#Z4XF-MfnENEcPFixAceG1JdLmUIwuh?zea6Uf> zH1(*`g-w#(@sgUXWGOxno@KIo9&!?_$Vp-CjSDEngVN-PO_a|?f;xy+FxIWB!=Ar( zLc@=sE?xNbI-vt4znPO3QWAUg`QhfDL?gUR7q`g>DSaA9yG4Kg6ycjIV>7`Bl_sui z(&v;1VWi&|Z}lu{KhYM{ZY6E*yrMMe=eAou`_o5);5hiGHmMcs*+J5ernY?cd!jl& z7G&T2tl-)YuC{#kF7eqALFUcRey>Ni-W^ZIhrltHx4%;|!0VVA{$+RgdTto-UM1(= z6@Mn!FC}yB#Z6i!QD#SCAhv3B$+tvd@y~m07Bs)K>{btvu2D)7wCsIF^k(;d?N-nh z226SsmEua&EE#9dO_~)Hy|UwZXY6aTi!7+ZBtYWOz)3o-q6lXr}5GFhUf+;UTzwrL@e=HT!rU6M#3}{{5%-BOx@h zR%mWkQp_8_FtTgjZ!!@CPU)DrS>RfFFJ&bDzGf7PSB2nCcI2sUR`Jw~Cwj-oJssvt zFp%$6l8wJ~WWBA*DgMZdM*j9QKrj&9UnyBP=^1Y`<-0I)gYEi?FkrH_Wu$JmArJxo zxi%L-{#in>lGU#uJ^PsSD5^*_l9Oz+eF!+Wnj%?=sH4K|W;=<`mo(g}*kY_@=B5<% z+PXv@4ReNmJldk>SXyZgPu%pPEnl3M?csg>o|(zQ=>>6)LUI2(5V4-?NhX*u(^*o% zl4JQfs_`BWvZs#BO+<*R#CY4h2^j%4#$US}As?RI3D0Z}V zv_Aj2RP5yqt>3B&Tg3je@~h+-y+r?ai$kyYi>=LO_`U!qa}6ybf?gV9Y9suU?NRF_Oi<;g*T2+E+Kh!`6614+C2a^oruEmM(^QF`8;M`U^)B7);nXe7l1{B^)+3n= zZK#vWCDNK7qM$xIl+>vV{E}tmfjbv^3ggww`vXZf4aVV%9!kUYGqZOHy+-P zt?3{=-fFY<7P9=d`3iEm9;(&@-~NL(?Kk-dO@Af4jU;PcH*~n?XN~>h%O_T4LfcPd ztNKu!u$CCepi+>QY3=bYp@Dq_kW9M&i(|M#r63H}>6KdVwa2;F9#>ng z9sK2sCqDaH@E=0$g3IHz-q94$jWc_=eka{B!jw;r8%FW6LFH5tsy?&HtBXUGv&njn z^A+zE2O@LUS|DUQO&SPNsT+?aBAt05xOq%)e<*!ze@8?Oc2T1iI}qqKo22LXX}}BZ$rWbuqegnxmA`F`fFWJm^4M7M4@Fq>cB=w0{jX@w z3-ac-S`FCRrwtp}*$M2o-F zZUt?W@B6t^;XS!xzBcfr(zZIPANil|`oG$uAK~k4-~Bu|FJM!PZEotcdLcl9l7v!wAMM;K zI^JX*i%=^qyib*1Q(&1jj4XHPHG@!5Jp1YvCk7JSFJN! zmoxXlgs|gn?V&-Wl=Y7*CU`5;UB7{J_J0W2{*RB|e_w9!uLwzY@#XQq$ql=tn9ZKF z5Bz=qy;=2iX5A}|)PK4@F#T!AFX;Pyzu3BlOYwVJ#$LF1E`VZiFx}!Jb!$($wYl2% zl$?&iyF@x^;Q3Zb=9fZ+9n>!UvpoHgKe;Vn*URN#9$o)J=In-a`{z*QZ|*G#%h)2x z-F&kvpTQQMe)D}Oi>$3WLR;3+Lm!-fzac49p(C%r&MI{n&jXIo&MqyA$W#s|x2W+H ze!($`r)6%@w%+62E4UoT9D&h1JBv&aIiR;HYFy^>7OmZM zsrPODKjT6ERe(*hBV4ffSAbhe=AT>8|CRvLw>CHDqC!==aOr$g({I>PVM`{$S^y;* z{F_mPt$Z9-#us(wCvLS|GW(neD~UAT%31Mxm}w+crX`Fb^lUF@7&d&>w1w+4nYm*B z>)hu5)}a5Ni97!a@G}!4efj^t+xb@mNn`c-@gGCSzwz5EcD$!#qMdls*(xhdx@?fe zF>^h9WAiTbGL@d5PUCi8y6cC37z(bioBxyr#B&W;Tvv$S{`(wNP4gD1hU>Grz5&ho;gyWx0fper zNpy{Qt%ge~TXJzlQfT)V#DxE`Xt=% zpSg@i*|aT4n=eX(udeo&w<3x*-*`o)VhcC9`99RpiMD@&I8gtmgxH?Gg|u%Zwa(76 zn471y)j!VWZr2>uD40Q3=IK@%j||GJ2p^v9x~k0I!1?)X=nXbpy!d9h&P5liJ@V;U zsWbjqH706u3XLe%l16%J#Om+|OI3r>i^S!Iq(@!(qW)Drk}$qJ{@+r*kZN5{{G9p? zHv750`&!1v#$rYS{QUhzi(^LSU(mVfquDrJRkZ17)%VAq=BFnvSQ*W6&LY2tOE=-3 zY1*d{T9^7H(?NYv|1+VzwwicX@d!S3?PZ=+n7&u{Ra67wh46mxMp8&vSQxI|-QAtn zy#K-uqtBj1l9twG&QSf=$9yij^>6>3j(P?k z7F~1W;^SGw#KrY8vg-6N$jHcel3m1$NbQk3r(R9zyLa!rnysy@GIN+TRMOjwb+_=W z>oQ(Dv-&q2ugT)elYcSIml9e3{(ma0{0j`%2lk7zf6?_90@?8O-xOtoFupwfH_!W0 zBJ1D(KL{)Y_&}<{RlXz3M5SPLWwDL=;~{UcojZ4G>*4>%QS0sFQ|>P#Og&f|QJiEI zuaN9>SirjK+NK{#pEGire;-`e(Xkj^Gdeozov4ysbBhACq&G^+dOKCwRC9hlfByVw ze*QzjY-%2r!R#6W3=I4N0?(d3W28K#yuYifE7kI2y<(mv>}4^WSn(d;h)zi#dXaE6 zp|wj5M#+Vg(3;vB`~_DR1ABXZF)^|Jk+yX6nUS`GG&G(#9L`AG!NK9|A@2_(BVI{K zNk-$6xr<}@S}*_jzHOdR68`@x3Wa2K@9b`S7J+T3Bnbyn&d-w+;k>YiSK<*?yfz1vZtuI@!vQ&qovUvJl(d8AYJ&l+Ku&>dE%NBG|Gyc7W+fZ zirjl#Rtt|=M89~UiS+2+y?#&jA~!cT71P{f%?V2ThMN;Tt%jS+4%>9r`5EC`Tq3EO zd+XT3;v%-gtk`>HWrf#y$#!W0mgop!>!T!Egjbg)Ygg}V&tF>$Fi{X59sOD91& z^JH_98qF-!>e57@ZFoZZ`ubXnoE)rwK&fVD>87_;`QRImx-n+R9g6F+=_zB5aNG-` zIDPu`!+KoFwmqfg-ZWav3)88#xc&enLqwrWcP`G1dY)3r($=f`(OZo1haY}O&1tm3rglA49wFg;j}Qu~<+jDrk;*Z^>=Yc8QN4K7Mq^(7OnR+xELFrt3;& zWM)QL;%`59ioJO07Ffn$Uh~wj)Ni}z`!|a%Mp~I{gtG9b*Q~ z(tV!TX3U+=i2n=fk-&@SXl#0|uxECUvw(DjFd98+*5$l7nyOP7WZQR_>M4j<0AY!s zEAIjo)ZIbwdYeW~?}zQ%v5)n_-9bL9hBZS?@mgT5r{eIWRIjeh+^4=)@S3zj_Gbm& z_Y94V`A}&%o(^}KY12P@sErTbmd@`G4>De~XiCd(d0gQAYExYu1r3j>SFpPMiM;&$ zbPN1#PmW2W`47-=8@r{ZrVgujbac?DWSgSzo9VnGxm&o3v-yFAHS&*df6|nTx$s2(YO-=B zozZ(`bA|8q4j&jkcfs73|Ef`@X=me)9vx3(Zpsavqu9Uq9QW(pZ~pVkFY_xaInjmL z#T}9J&VDlGSaYA~|JW4^%d8IZeIu;#WcP`aCoc;S{Mw_2%h5xV$*)YtwDVJW7j~z)-4;$nseqfw3Ik z_W`o~5O6LD3<|V?{(BwbI=Nf$Q%TNnQa;jq1Hzlx&dGC|wC^8`aB)sh$!3EZW3oC) zD+MoV*xA{2-l{xhUVn+XEI}!~zGhxsx9SlEhSox|)1p@BVTE=*;dSRnbfTi7Qg8PC zl|iu8Cv|!A1EE^o1$G960QmNzWEk)1h1qd>#DEL!XJ0=R_3R%>ul1R%5nz{3R1UNz zxRL4Cplvh?>UKIiX?ZzbNIv$nmd*?a3MviUYC=LLd%{C3=A}HQL`xnuZy@GgF(E&9z1=_sn^^w zZiDx$=bU)V`%4ieXs97*6h%y2oCk=Lv4(#*0dalpd`;Y@ojC_NZk$5`y+JhozAJmY z=tYD{-P1nvIKv_>%W-~!7wg5@9~YnGv5SWLIhsJ>_Ea+Z3hu2HS#@luMm@+}7k8}$ zd*Oa@dMC@O5H6O{fC~j68uO_-aSV9l2Eu+&M~;b#iW)HE>pb`D`n79&VWQp+%0WJ7 zzeT*f#_B?g`Rej4kJ}yxjZ8ziFz1Dj=CfTk(S#k(&K;IhQlf@HZ4_ov!^Gut-clAStKc11~vZD{ms;XP=DVX(p+V0PD+7Veirb=r)PC#7O$zX6cRzK?!Vsoq3o}_JBR{(5>)-@G$X04 zmSVnihs}zYtw|ESi3I(9(!9TTczD!H$DU3uoKaqy$u9Pun3#b2kWb7u`2w*Q0)mGprw*9pLNLHZgN0WNlA$)xVQR9gHwKg z^)w^nEXj|hfN$en97)$AG%*As61;e$--GBgE>^q=WYK#2jOGJYt}6YilO!D-mX;Z(RRg@8^2aXh=>S7vmOcB-2z|}R~{m~^yyX3XjVsm zy##w=D#3wW;n{!SKok)?d%tm-?V^fsSy?l^n^2%9_~yn|XK88a5is4lY_pzLonST1 zP?IcoHnqG9fbPKOQBqK(nq$iqi=CYum#140 z&WpUm3F(H~^hnoT^9&kZNSW904VD7*GkR~w$H$l2B(EY0$nYs>X`vd`X^MQ+Y%`R#MmK9iI``?!%>s0HYOkbp%~tglvhZv=A$)&BO>TwLvPQS(GHAt zhY8i<_E`{QPW@NtVW#tEn*N7r0mBV0>{CmF(i6 zLVLUL?)k|W?`TE?f;D4imUr=V{*1(pKST+_;|PR!iMVq}a+2!?r-obD#k>#6J*&n4 z`NyT0N%}f35UTT^`QB!E*ocqwAG;?+o9NK%?iI`?xQPa`6Fo(g42z%br}W?huZM! z=IWyT*PlO{gFMB-Uw8{)vx}Ea!W_nOxw7`4dPGf0i(DGRKoAGdXkSF2ZfZ3Ii_!#lhQls1bO+b zh5|k~Lt2BKYLM|oa%9uo_d%_YnD6WIb(@1sO|#UE!hIayx)PDeIjk=Raoq_CtOy`8 zUw6S4_*e=VqoxBWxtbR>CljJ+7eMp(bu6z!f_DnZcMkGf>c4yc{#e#xw8-)aAOr$h z2@8X)Fi1{3J=D}xWCme-1j6Z$l!Wy4)fA%LFzTTY?vc-~PEZj_i-D?3#>pDR&bT)$ z$)&E}Y}?zhq_a8tgx>$Ids<~$9GjU%Z_M$n4R3ZF>LqU>E`w&=S+?jJawYsyj z6G1_mojK+Rm=X79lG@v`A`au%|LN5pU$Cz-YP~Ru+}sLh?JHNfd^B$+! zV_7!u9*UPExJ|_*Q%*B@ehawzN=%`-sk&$2kCa_qE~2@W~iNdr3#I4);;3 zp~e%4>q0>NUI)yqO;ds0B=V)-ojKeMRD~%^1p*xcc=3uZg>T?o&*jUP`x|9tyiaSD zKL_NNd2$Va`6&SGYte9V(sU?ZXiDzyb{O}De3UtBUTqY{4+m0Zo*5BYov%IZva&?# zu>c^Lv^;iJR>>FwFguh~$zCi7Ni=$Gz!63UaJ;)Q`sQChnm0)Ch$4g4wkRCJWkhMy zX)aFatQ0u7ZkN^NwW5&}!luhgLDK;Vx6tBJ(53KJU8@n8A3kkA_2C1g=9-~khdAic zu`EK^agkG1rA6Ep;xG|Kw^~#52L&QaeS2_VrbpAP+6|)1tXkBla&tc}H`fCQIfSIL z4D6Iu+GoIx${f9d_lW1zt|5`c>G$7%|LquFh;KQBCaX9Ruf2|v09x3NX145YVCgv1llQl>O6>ycK)d}|X16?_94cLCm4>Y5#DVu>W|GNo@jP#Zyycpef) z{dSMwr*i=>B%2S{j}RU#ZGE`fP4-abj_+8_1x z^|qg0Zr|4iTCdRQFkV=frc-%g#;-n3E*|;AGYGTpBZ-Qs48V5f5(itqvk3H@7hIbr zB<@$0l|vwJd(O13oLu0`AZL1GKj^w`w;OPb+_#^;3%m8^1kOKV>F6*TsTFX~zT*T- zVQWqK{XDB`PD=ai_uqeydquy(@sa`GKz&|rvu;{|Rg?9?_zQ-ekOyJk2W!;WM?ZOT zWDGEtetU+&w{jJ8FZVeAj93JN3IbOc2x;JRvys+R1ky~`$g7wf*dsi5Bg3%iR=c|n*H|=TPqTA=S8)(kaTnq!8B0Nkzw>$S3U-@ zm)u;d;cWj~w{G<#j^p|8bc=?}6a30e*E}Q*g6t-9++k*f8gl71UuzC*PdS9ddQz5o z5d6jU>({jlEg&vpKo>*LT}_*TaKiiH$(^)rJCN8jaACIzqN(Qwn@Zq8Dh*OGbw@Pv z_S2AtSlMV8ZGpXZm^m29y>|2jIe5z;X}o2FlsRHa5C9;pfQgic-+M^zJr;XmmfAs^pUx-7p}+9{T(HHSuSEm-vkUTb}E{ zXd-BEx&{IG4x^b0-->dXK^XQzZ0TVl#1uZFVIJ}YtX+!F@P%d+@%c9DNzLmu0zg|b zLRbjNFQ;qXMxNJ#;k@`HW+2R}SxLzZ$xH;s3;fR!m_UQdnKEa$xtD}4D0LF5^blj~R&SF+_FC4n7$ z=mc<*&};&{8pJ@4d9Nl(R4t&j`LMGA_gl@VsteUzsz$sA&xy9i6Ih zp?thor~}UPJHmR5Co41{IKVwG+Gn8{E#uTON}^T0eE`*${rTfQ5H9b{8-=zuZ@$@1 z`*U?DPpu&~KDo#P@_Oz_f&m$JfbnSp4pK6K6DA+Nfk%IO7`B*W4pYHsRoY`5%JG^# zr-zW>yhAb1`tyd9Jb{O6e$Akxk#|GwjZ@^vYy49Cs2*caX z{rg{kCh+(OPP|U&X+Q@-4v2}oPZu&>13@={sQtSt!uTwZJjT4!98(5@H5}oCl*$X3 z`ZK^kC}`E9T-*Ei5j+X%kSmqC?Jx0*05=eXx-TyFFh~N)nzRci%cwMtULnwE%A3H( zJy6Mnc6$EeF6)jmxHIe~Vb$2KF39mhr^}I{0@7dd$qVbHIgCh+y%tPC{TP53$%A`9 zX6P20mr{ryz2X=CYPSIEuy=ode`}7p*0%&7)@ZYXlGP0iw+woC2;tnpCmeEU-Yqcl zGgzsvgSo6<#<#>>aVXH27mcV>mZ{ zrVo=dfcKCT>nU(`Tnn_Q7^!vGgjUdS;vx*siRdstNy$?kpgxQ|Jo!BnU~9>Hxw9quqr zmujW~W0rh^N0YAY+*;*cKweZCY615s!^9*C{6rlj4hc4ttcnka+BRMH`XM8hs(4OR z5bP=yIHN;O^1zpb0o#vE(}@B@V>G3Ny3&!hXh~cicr6gP1fU_`7-Ao3`M8619J_u! zzdhTGUdVnbrr8A21+DnREDu;9_-8y`Ut*-d44aBfn&RXv$Ge4I5?&V*6M@ho05=f& z#Sjmve%M17c9SrB$YP*{q)mb=;oz@aUJlnL_NIHQ3T0a0?)$C zXz0013(ix|=yeJavWhggzCc$atSHw?803iM+x7_lOr?7B*LMR0&2y8eCt2^K39u4% zZd&T`nPY$}uZ}K)5+Q8BKGrxVFa-yM`)$uM!AA|?21FrkT%EzfSsA5@G0tYSJ4r47 zg8EN*6F-+1bV`NNs5v2aW*LYLS`ax2Vk}z%<%@xcsg&3O%wQP^@Q_r<*-JTi-ov%L z7b12oKAwP@CjIfp9|=65etVDO_2WsF+-$Assu>a$1D+?t9Ey?|t45LFZ2|Ku0sxxjiu+Uew1-Jgch)oj$t|sY?isIb(@ZG3O);cLIDHu zW~eB#pmh&^ywY&MmX?EN#5;*tet0|Ax5wqt4T5{{u7Fu#YKYbrh+Lo}2Vl{i7o=?% z01rizVxXvoDJi;m=meb8TKlUXb27&-g2xk|9qZZ`cS!+Qvb1i5g(B!HYDFD9CEo8% zR4|L1nwqxrp{_ieAO)%?3t(l131i!L9(-`+l6}Zo13%(6#wTF51{$A|!a)S&F%iwl z8uUmX@qob@!7KjK{3l2r!5=iX4B-#*fl=&PUILtY80iEA`Q=d;n1H7Tjak)$B$_TW zF!T{j6lfRLL5@gLDid%6HABwBiP=7giW}C;UAgjLTFk(};9EkP^GU>hdguV7n3o$8 zbdW^q!2B8uE`$VJnBtIMECNQ&9KsQkA|e0Fj0bcYzxdZ*fBi}DO;4QjAa{sc{+%Dz z9QeSTa_&pV2ZX?FB>@cAGjKSFt6v%&1mZ{wX3I|lJMf6hNWLnJk247P5VbtZOQ{+5 z5FLW2#qdyppibOSAREjVhc<~pIWQa=zyA7bGYzDgM9V61ap|%D)cevL{F<1vvokUn zMoYy6(&C)Z^vmt^+E&Fth>9Xi2f>NCwkyD}k9I)vc>xv&>`TDc<+|`5vPLfmQ<3en zNK$PH`S(jg)Sqh!;WjxW)Gz3KM4rBjwKjQ5pGZxYvX+d16 zVydQS0_0;s>>6~)xI&a2F96Qh^AKx-8cch>wtogwvqYJ`RkQ*!6d{awKb35Qi3Hci zckBMg5$3p>imI!ty}|V|lU$@4r08HqaZVbFijQs(@NpwOz&N(TsrtErq|1Z#qp0Kg65<)aqfA=&1zH{#<-GMD{-NP82o9`p8n_&d+cGxN-t8DkktiwrGxV?;>H zj3uFxb+Qx{$x1NvZ5hR76yUR$G)vNmL|B+O*euo}c?Jp8s*Y?|U5Y zJIC*Ne)CY@?|pyn&*!?%>paivx@XS(&hDVol!GpL_TRjAuT;!b?cCk(S}B9#teo6j z2-b(?oJ@hvydnsmkLZrq`wUvk%&MTJ&u?zqs_d{GAyq>@3cj}hb!F&^;@P~W(G z8*r$U2Z$V5DnNGZgDuH4OrpPM&z_TTj9Q+3VXz%yO$v%e<%0MPJNxRnO^S0Uh;DKe zci>&Q>h8gDPM{0tHn+`BYMT27`f5j(D6T=6n~V>grwokk2c}Scj))G1n|FiRC#|bH zz#K~pJcbNC1GF_c^N=2Q)_4tr3h`|UMPd@S;{xlQ z!>n-`t#_@GaJ?H*0$YO!uDWlemI5@}R$+qc_-pUp6p&5*i;}Do>z{8^Yb$t!jPN5V z{ajSrMc^~-oh$O{np?Mt@4Hq%6Dn5k#m?sV&AH`?s~(*l>Cx~`1_EJKo1mjbKNYA& zN^RcGWUtD70nD`IEjIBY%HOnq_g?fJBJ@HYYpu>ieXU*xxaRa*WZ~eM(rJ8{`= zKx=MAvt@!?Ykt%TASA#nozz`F^Z;RoQ@LZML1m}~)Z(jEsml02$4QeWAypLt;`p|; zHm9L#KLA9)e^xwsi_&6cyG5;xCh=V5VpHLJ>-K?Bw6xrHE8RWuTVtunQ)I;ebNmJt z_@OQsg3U4n)}FMvsn&;TI@FF(Xm??zw?1sTD8RV9xM2%ow*x665KXi2sKxQ1mFv1A(fo@HkEZ#4`NM0=lL&KBd$MeX%Ogv z3AOho<|GI zW~5@%dG)$)&P|j_%H4jf2Htk*(giRN)iiNXF?PNmu6}iPcYYh_=C9fbRw-v)^Lsk1^Tyx^FBVO4K{ z37_rTftK{BmT+X7Wl1Z?Psr1k?-9{E-jd$I;0cbW|vC6!4YbUT) zF4oo>q0bN7c`{(9^&N2q54@0ha}q1S#GbZH7=Jdg>Geh+zV4?_ZYuPsFyDz+lB#^pS}Dt2nQGvgsu7J}f4NTXuI!>xG$cOh7LI0sHV9Tq)xN zodAMDscdzGGBXlF$NupfR!(Smw&gr3$uww1*4Mm05tQzIJL#;o4_zFuO#*pkVH7-% zZ8Zk%qp(PUOXgN6f!855PNyX@p={5Y`301fKKdRTFC-m%3SN>T3!ljIpT%9Sn1D^+ zYvR_geb6#ym#{ZDSeYU+HJzC<2loS*c?wbGA*8sfYc>xZA%0b1cyV0^+JDw;5p8Qc zuO3_O#QO?V8$D>}_~g^ilSA8Tk1-I=8&=39RrKF4`fg_ckMe3Q-~aQnX}$(RsPc|K zb^Q3iQ_TpaOkdqk8fQYJ7hE0{aO@xc^E!;?La>t9phM1H>VL|>U4qxXk5Q@E!29QP z{Yx#%`#l~{fN~LG7OL;FTCAMm0-8+B^c!vaFWjc)xlyLL6$;>xK)C zHPzMOu*4KLJ21U&Gidp-yz7C?hoEQGDwEnhVfE1$>TBiGgxMA1)5G2$X-#+?9|?we z_~nv-hL^#PMJ0h|rn0$jyfwI(s^-G>M4(kw=b0*CVmc9o$lsNo^;h=8Ibt1b(3oLX z1aFwhYN@u%T2~t4It4=SLtUVW9WZf?`mcn>6^aih>dKHILv~{C&u{kJIqc$g0I15A zIusW^%8pk@QfgXh5Ti-E)kyoS$Ub2A7Onl~-{h%d1I5Z7*TAN<*LpArP}KoBcOa}o zKTg~#JFX0GqFyNu-#L5-j?Ub2&s)Rdb}?+q721jhD)N^A0*WE0Pjz;sVFRVOE;h?) z@$rhfe8~PqjisSxwyUtGiR9CJNq=Q%Q|dTetfJz<064KTU&CJTzObyE_9NNSJOtk# zL#l=CZ&iasR0zt}uR+4|n`*K4T{!A@c$I%by>WJ*v+R^rIo53V_ zG%i$n)n6&xTx#Gd@0-O~*p|jjFCkMQ4hJ=Of?X~G2UeZnEKGh6hy!n_NEdexa*wvlRi-vKCR+OS$qwa3Ux=ymR zw9GRWpZ{QYUHPLlY}xF)$7AT!qS!UvV7zLJT1&qL zu`9aq(P5XRZ2d;r8s78sz$5#094*OS95oB=ixT=w)lsSil%`om`Wgr`uwrE%D1>ex zZLIn)1X1Gywq2{^2CsR2gybARYgrV;t;~5H+Lni%Wb^9YVe@(O{>~Ydh2kYrkk!D{ zzDRTw5%A9g^Bdc``lHM}8ghThH*i&+jemVo>G7eFHdB1xwCNHQq9BI#O)ngh-+4$# zQ^3PT;Kq5xSZffL)2a?-&{{m+EF9oCIHa}g~1XWq9G)G{*4q^ft54Hng23snKM zfg(mGxDw^#uJF11Ci<2kYT7a&gDMa-17Nep&U`{=lC2Ga&(OV6Gyp`q@yW5BLP zBjo9eWG=XsSR;9Vh#PyU$v*Xi;(ID((vFr0y=uB|RnLB@3LJV>1 zTSS+(IJMs2f1ko~6{zd+@*E2SL>6UL$m~;4sLBS~eqESAVQolu1tPG9uhQwCfA=P3 zP}JOhlQs(ac2LQsv9hD>!@veMBvOQeyDJoT)o(H4_X-clf;ovv&wlR`4w_QV{SUr* z8Zj#-z|DW!Vv<}7yk7Fo?Ls%Zv{dg)+J$m91SAzQ`+h#L1zXg@NAvkaEoZk(?&xc_SS_q5AmzV2N7#gPxeV{Z$Y0- z)Smii@i%zsRpCd#=GOZW?_RuoIYl$j2L{Bk6pm-kodhA)2BF@cQbW8;g%T?wCA_wHGs{PXe&QHcUKDS*v7oT)rIzwhHJ(ZBqlRC7cYQxf-K_UC%4!{4`>V(EE>=_X zM8@uCuyW-}5%q$Cf<6y%L@ZpiA!jc6`w?OG(wg?8Q^Z1_6L)B0Tf#VweT*Aq@`Z*6CB*PkAqc@<}2ul^Jf`r8st!b>|iEKP1 zL81KBV4Q__9AS3f>BoqqEb*a*+M39#(Kf~MSRvdRrbU^ zc=V{$J=Kb2tLOWe&=h(-<}Y%Z%wD-d3H3j-LeSq4b9ej@S{zRw02FJgEhC@*;cgRW z$Xi5lt^!KIUpb9f868Y}P6Ve!KdV8&E954sP%nLu9(e93>Ejsq3XO%^Sw7`~PC z4&mC^JErW+B4H%Fw#kxYxuh91)q%<{y|6zQKS9Ccl)H3W3a%okQtE(u*A<{+5ZvDf z2`d6z(!e!RImptKpR1RZW}zHNb6t=p?!NX|vu4efyQ%=v^r@ww7CDep7dIJq$A*SO z!02pZscwT3s0RC{b;oVhKy8@>m=m*-#$l?QlR;s0KXda8AXu7-PC-xVdJmZs*hrHh z5f0jCjRQ4P;4FCXtOlK!esF|qEfN4wksk*})WfXd7==nS6WgaY5jov+Q}9ISiXzh3!5!HV4tNSe9ERxjRvhlF9W-?M1W~@ zMRLyLeEOEyzB5DfPK6p|iYk=Fn)V`}-@D}5Ruh_rMwa`}YUut4gwyD_<0OP|53^iN zh=Ge#_yk}=eCaNS8U)?7SOCTDZ)Xt_OH2YVfL0LtkNUHXz@4{YDjk=L8ir4U|wo?IA=_*iL!`#ENTJp@N#$A8bYV zKc!ck%eoh*`c9YZLkgF#`BAN#u%Q;a7K{euSEcO z1GUq@PUVnZLZ#H9ANR_cxx7f?wTQ{tfOll`t^ zfxFgMXT4`Z3}_b8GPy=v=DR4Wm}DzULd{Xh4yKor)qa)~Srv->qd$F!X5b_5*e5}) z7Pfg54(r@Pwl<*)OBF<_C2V54#*Sd9oDh(-q_dgt-)+DYURRBt&9)qz5}d^!efx~; zzx?jOAK38v@Caz%dt?}=$%#a@buSvza2-bBYdsiiNujivtME3T~}+U zJhX&QMs$t#L`V?944mB9@h)%npO(tf_NkD(s92gKs>d#zH?J2XQ!QM)_VRlrfJF9J ziltX5R$qbtA~$;3^t(+sUOVx~N7%LYO!ZS$7Pf88YTF!LrUC;m+y)TxdaUfoo&hLrVx(9_`JA{$KRn*zc<@3G2i|^L}D+b)(ZRm2p3Vwb>iHaKQp4 zr~vM7z?60FK94bPO$8QJz&Gu?9=>B z!daK={v6@)BM__qZ(y0F2lOqbA$%rMzjwtiPzJ;CI2#ZBGY0WwAKBJ#^yzj;L0#Nq zxU`_rbFtt!$ve^FoB~-eaDyH8GS?R8Mr)P$A;U?@Y@lG%90nn%SNsjyhKdlL;}Yp; zT{9?pJOt(DqNN?M$mGWMZUe&Ma&0_x{>E!;`v6lo=!p!C%N;|IOIePW6f9B#k1n`< zguBz@)jVqgYaKhh01ZorqE6x+PLdy_E2`T*=(fy=uR)_k3iPZ)a%DA$UJ&kWoa#0A z8liIShp7>pC0Y8_i)x1=0v4V)6j8ltG*pHaV7sV?8%H5*Q)z-2ZJH&3&P6kQ<;*jg z#dgKqF^h0OzG-{<-YXJ4c1~bM(jfJR#&!`KOJR89`_f4v<%;IML26e+TYI?{zr zDgf!q9-bgF?&Bu^*eQPgqQRkvUuV0f{$Q63wxI|%tr=MhF`xpEP>cSvNRjAPdwGm8 z9b8ve=X7WEqhttvAy-ioma{l*Ns?fL;$E_C^@lnJ&ikvI=v@M)u6D4Yg)9OJoP{el z^=GT3^Hfhn4=j@&A43nK^rutH;3eM4KFNPdzrbTcvAF2^Wbq4zXDO>3GC9F@ikN; zfZb#>tM`+^!-n~ScU)d@;`HemvSoY@dowWy@EiR?)WcL*I<9Q^XrOcY>yYhCvKAr~ zwY=0xGR_hk|5=S4#AK(NgN`M$gXwAkGTS8*#_%|1x!a^Ast0c$NJfn z<;OXyfPpt&K(Nm;%2*r@@62t?g~DHix~tlSMg^Q5e*_v0N4ccf{qf^vJ@jz!@Zo+? z!nDd?Ob6@AN}>ENl`GkX^<;Y8?j+=WjWTg%_tt{w+!5>$f8w7~UDvlNZ13raY)bim z_O|Q;Ja6Ip`t|D~aHXc_BnNZm7GSs$tOOc>J_6&KVT46+HH$_6#wKNH#k^yG{t%c~ zAzDN~v~;i7Tk^|BlYgMr?ccz9{mZ|_4;#K^H>V})?H>ywZQ~81IE~ZUn;}|}JbQ})_3Nj8KUbf6@ zeh+E&e2=0I%~*h%*fn}yB5BETLi8R(#j*B&G6asvEDY{2&Jus{EXEytXfbw&j-c)c#yS~oJa3zy$id^{#E+6I zQFHV>0R6xV)n{1_wTlooVZFr#BW`145*t;#EiUqzF_ww$s9%A5ms1~1xdZDy0 zx@Fb2?)qetY`|9=Ji6QQ*bmqFjmINR%FDYX?I(!!LTLk8VDj`WmV(OB|9q{IQIO6O=@RY@B<76Fh64^8 zbYj4~+2_H2+7=u?d)D-vR8FoNl2iH^$VT4p((l=PP zm_-hFDQQJ)PR#`O(r|NrVnAnrD$Xd86R9HqEaWWX@z9T^X4y|epby<(gQL}P*c`~+ z7#{;T{hzzmrhWS3Q+ZA=A3b2g0#0#Is}%?%#SBKuS0Lt0o1{xtXdn8B0+=>^dIV}> z3s-bWc0X;uU{*^AD4ot-3n$;~hrR&S>Q)lLAbm!^5Rq=k6+Xzx(U43~*WR;cV^d8k zdI8pF8L%r4Ie@N|7m;#ZX;^%-^R7cpP~Ap5pbbO)x2#x@dv*oK&zv#oSnpLYZtk_v zLClDNX;aP^($~)-(fE)9y;9L}CkM(U?lmw>jr79+nvK}B0dc`-wS?u&2j_4$xNgN&us_D-FL)>-)9#e^KJZhKP=X(gH%R)DQf)Y7a& ztg&K zG}8snd$|49deYz)I%Ig{50I{O>ItAaj0`$ElyiCB3~+81y0F}L97_n=%SovN-xosjX`14y5B|9vI#UD0(w zr^#4`eH-jBK8hgrvz>$l28 zRCieoG)vDCXBpsF(He+O_ovVGx5+W^S^MBlqDwy*r}Y&aAav`ocjHkng4&B9WX&%$g z_#bq%lXrR+lMl{=vk7jU6@bG7)4X-Hajag_NenKIjwoS88v|M zc-&D3q8WDK!iABT(!`+`fpH7btj0z;73X`ubRuEx#UJvm-6J@4FG2E-UU?O0%jaQ< z6_^cD`=?SN5(cf1km(6_(r~>`E71=*Xa0O6z`pRgjWMLT4ZJ)bhQKeSE*ne5E!?>Q?+Ed(2IY@twQ+>W!!O{b~XF>{v>Zz58&@ zYd(rZ`C!On9Eg3G?>}2rU5R`#`@zG9%X8Rj9wZGBQICQ7V>zR<@1I?IYyKAK#TgCw zg=Galfcfrde-#(|%_DI5Ji;ewxn^fs_+*obU~xo5o6I;jTKu5)_dtwX-~BKFR=|+v}{_bDq_pBG?Hc zQzozsXPf`_TOYcw@XzNufSA^p-sl9EtmdM>twfPd)n8MJ+W2XdH*{1_0PUKQPq*13 zOmJZ%bq5Hz_WsAtQcC-FQfq$FvVx5OhCPO6Byqe&c`BM@Uv$L56OM(6#^V7Me1VzO z&v;m%oSFa_Z{g|^*}GLNs`Ov^acq|>$TdF{+P#|?GDnZ?TiO!7d?-aGJYy4@QNkPdont>bETsoH7s;o{`#Wn4krAZs-dBjrvC+^rZF7M`U_f& zU4aE~B`@g9UN15!SKk&Vi&4!wPn{yKfI{eQ18 z_oS&5;?C~5#h8&eW0b^dMe>TQT3^o(j79oMR1fjsh(h-5TeyA;(u(}4jBKUle4Njq z;=sl&R74kWN*vL%V?3Td7rBgSL{WX39CE;zojZ0EA%^<@9t z8^kXdtEqs1s41b*Fr!4YWJ&QR4dx1BVYc&1%9>ql1yz`^DHl zm>#0v!mZ3{Hhbh6KwP9Do7ns?rek>l*_H^Po`1qyHen7C&4S=nz%m$V7Jn}*T`?RG z%GqdGqKqIJ+6YE*v|Q`tj=ZES;j$&VSgCE3q504AoqK9t3L0^{4_mp2g)&xdSxzp@ z)!``5e0qsr-Ms*Hk4MEStLa&8Xd}0V|H^oQ@8P3IHHV|%Al3Tt3}>Ezgzs_sqf9cj zG;GVaH{;KPiQr(n9;Y)xDJ2<~KHJh&hGhK=sLTO=TB5wtScO9}TG-Q&kE1*EMP0v6 z{C#gNTiOwsYzyGbUKkD>^iO!|X!?wm87|~dR+0r~G2nux>A^nGBTXOPI)X0j@i5dH zEGw`_qCSY{E?0=seux~s@!d1^ke-=8X6!pMeP1dTB>FA=5PoQ+lZghK&(R9 zI{GCW>p1@c?yO-Ja#8E8I=yq-rr(7}DaL{PBnA7=Rw-rT68ms+%RFD$00=w7CGxA} zCWKEF94+?%VV2c-vKA%PaFl!#Gc42BgTb4GGBBW*bg}$z!)fhpmF<0zd@Wp`f-L+E zyx#AVeJcsjnwpwU>vCSOL7>UU{wh1h*(32bidRwNW6^b_-KWt`~#J0L9 zBTN@~7(%h1_+*KBF~xZhc^aptob=Ka*RigWM%{<;JQxtzGu%!L`|Vsmc73Uzfnb$F zPfO&#KV0%|ws_Q578)3rD5+qS{LRqkB+>#6`#QNzrfgw4h2g4 z7OgYBiX3n3HmH^J^guLgtz5M#uom+Zz9WCQ;KMa5F8>63=exG}}O0B(0jU%8>>vFuwjMRO%#Cl%TDr|Zf zS48v?{gm@JbPF3z=`XT~In3@_@h3dOnNnyq%?lDt)uE$l+R zM-_!=3V3u0rkn^wwSP{=@``9bdl*6sV`AMIN4aCq(1t?~IF{n)1I}?1IlA;NyfBen zKW`bz`L+VNzcMn|fmC2yK!%nZufq_;uU%&k#)Ko@)X$J(r0T7HW)U&n40L8&RNfwy zxeFsXY^6=3WCYMhyC=48P^1oqpb>t{5;zbqtVtQI$n;j>r2~RH=>gjoU zQ0CcJXdEooPWrFQ573%078bdHI;H}uOwpEuIS_$uFk~kq;~-dYY~o+i^E51O$}scn z4*lME0nn%wp|`{aqkhcX*BR+|--QAdfhKR`1vE~;DMIU3Ky7rv&nPU9pE`9bUWi*C z38{_z8+N$_xR%}t?#jd-Odoz(9jaxZ8AeWB(<G1YT zqMUaLy@_bKx^t!*t&<#K`Y5TbPK*}D^Y8^snY?&$!5AGjr*PfadfQQzBVvL;Ig`ptoSa6VptA# zg5%RDYa4(KE}U6t3uv~z0}xI#SAY~Q(v#5jmx`ta3fJSVq~dQ0y*TT5`*!pz__E)Y zeit;)kd_3=%ehrZ5HyBxcm|mGeryguSkbhQ(^BBynv&$%s`SE*>mdJgS$o;dzs5Gb zieS5^X%@2BjxNJ5^+TTPMNa^}RkN&Z2&Y?dQt0DROWML|h(NE)L4!}F0^!O2aqE|( z2G+C(tu=xmTK!DLT(;jr=v_#9RgRe1I($IT;( z{+ew{uZ?;|5)C}`&T^m_{(V{=rxFMU_^V zfl0&LSWM=u(fbII^z->vbv$!<(eGV`PtDtmL3>5Oce{A|ewfIXZu_%wvzHjwO`|f# zHf9=o!JU2{s&8qeD5?w{&Y+u(b%aNzyq?}kH}5Q-C;g9-0*s+khW&Et(^jCRA_^;L zuMk!?*#9kb07jMJ3$UY=lO-kZ?_6!WD-dl6LbIg7A|tVhpi;xX+V6zB1; zUn^-3ES&w}78UErD{?!ueHzYbjy@H>2}xT&U))ycM90bjO8I2cctcE2#y>Ua^UdAyBlF)h@4-tx9D}4f>z|pK# zr3UTe7_^y#NmR0J0s@dZbbfX)fg>VohWZGgbO&-C104z(&gY@;TG$G~H#nESZ_m9_ z({D)mkZ-OX3r;zjvSQNxe~o;$^>?LDzS(=p_O^Qe9~OPSCGbm!FaNmzAM;+S#^Ff zk0#6SBM`p(68IkBjTc>E;5e)5Q5l4Y`M_9ala@2_O<)Lmkul{|r;35%X=`5x#@+2A z8#EOS%6GCQ8CYRh4F$>?sU8}vG9Y*R9+!j3d%+_>D(OXeG0&WzpdX9HBfY&k|Kgcn zl9J$nbbJ5iK=GWCJhoMF&;uLSsjWg7Wdm8xs#G{=ZD0;nYiosEtL<3rY_rnU6U$;1TH$j@~F$oU{ReD0X4$S<#L zk?*T0ukl?a>>~sBo5HU_(B^b;CTV;wQyY^s!b|au9?wP72FoR*XMD>5)@z(-bD2eAh zGW{?%>i{x>mk`=G+t)s=@gAFff$mX4{=hCDV3cGEeVh`DU@qzx$mN5zcjdbTYoJw9-DxM%FVM~l+cC3H74ktU1QHdX z+>?)XRpiH|VfPxeG-u`|RQ}&?TZxCYDKr0X%VU7uVhAU@*l^9@8C$G=6v_hn-sevO zSdx!%=uHQe*N)ip(lqUUna)u?zbF`@Q1tx$rM)C8?Fb&Qv*=m+zk&{VHEi?Rj>|}c zG;V`p0Xtmy`p+h{HPL@7J(MGj8*iIP&pL`AL-6jyc{Ae?*v}!!d9<7eDW%E2IM*M$t%6V zu!7a^&Wtsgk1w6jTANMX0Rz-_LGPLb$p7>yW<74ERD}6y-*saX!VW%0!UU}l=YWN? z_%Z!PyUp3LrSJDstw0X49{zUxtFTKrZ8c2^Cp|<|pdAf%sMrYpyvOh>RYgm5zP(%p z2gxzvCH0<=oiJ;^bCg0cb9=`=9O&RT^sB%DUDCHk_>CXTT|###tq2ukznKRkq5(#( zTC;+_dfCy7(ZM7WHBSTtMl9p6xO@%-VMi}pce4Gm`$4!84)(X$$UM(QMg~oLx~7H= ztbnn$DyqFHia>!|n!}{%Fm*ZIRTUM&{UopU1~5ctc9Gpnzsgzx#96lAFW)BMK54~G z!5TLoP}2`|WPdxz^C^fcENLlz9y_>h8V!FgDAZz$MIf6z$$#3?tcZV^JYp9eClJ%s zN&+voSeyJ9y$B;Q$Che7oMP$;NcTsyWAm07W!L9+$>pY4rK z?BnEc2w24s$W3)=`@42550ISw)Z-ps$@MwLzR@ZeTYlzH{n&_-Ib`$T+7XI3U`w7; zzqwuwH`J_wLD-ZxCowGtJts^(v?H^*?`7*|e6V*UXHFrwE04b6Fkv%@2M)5M;f@nL z_Z$fdV5*dNdYt=~o@iBD#<8_y8tT1X46*XTNSK#Db;Bna{#<*!XJ5U+yM-H@Bnb!F z%-C#D9zong`icENQYziBT*q?bo75l25d(TO=NFplPWfjgWB?vJX-U;0+>iW(u=u)2 zxv*hML}JnyfLYSC-}d;#uywRN(q`sW=w128)#gOnr|0dNj2$-=ZKsIbz0hs!JD^L0 zfmgk&S8u9LB-tqSVwX(U&9WOjVuZgMrX~qs+WCOb&1RRfcn)b{pAZf)ZUQhLQ`P83 zS_WfJ*o((i`}`ZFp4egbX}`~5isCNo){TR|L?<|VZP4o12l*zm(VfmJN~fY*3vOpt zOyj(`rTVSG#jSDvQLXGG^l6ZW1Id(g&SRdFk}amEk;q6zHdV!OpVdM(3;v>Gs$KBA zih(oT=6r>7pMI2002r@T1h3#u&Qb?Gcfcib{3-;vuT{ruw*ixk1V)j!@d#|=e;ib* zF}vgGRPjb<#imGijb?HDS9czNQ@^3nnS1`-^Uw&i?UBVGyJU@`3kA?FzT+$Jy#7(( z3nvH)X_(4yjd@VUy2M1+?9c^oPIjWA&x6@`ES#7x1T&6tKvX{qe%2wr>?kKp+i+Yv z);V^-FDJj%Z+JEqvqNmq@`_hv0IhS{nHbsw9-9UPS{CWIkh6Bz_BEY+W z;-!&G;(~KPPRH#YSJ1)vi|w2qc0(6x$yl`bbdgxXu-{ zE6ygmxHhcu5xj76M)1t%f3M$R($7GvM z*u;TLi`Hdmu){DmB9h&u>?>#29jEeChzY)iQxi~}u9l9;cG z`FH)69o;(4CuMYcwpq8is87uyW}=YI*uqW#6oz-ppPPfpECya(*^-~sx{y*Uf#x=Pf6ZNg_p8SA zP&6|UQ8Bg+Z}=6j;6ri7u|48|5+BMqoD1i|g{#kr@*GWBBLy!w;!t&`2j)hT5!Wq8 z!DA+7T$9e`TWZKRFc|#xchh#UC|BO3#@upH$*O_Ps3V)-O9(KDpN|y_kGu%_phED{ zhE-%0S;*VfbnLI-v}30bdBD-=(FxW&r|XOjZ*oU+iWw%WPSyQN`v+Kt{b3Xz-%+Mw z#=(y`r&|u*DWo_FJwuEQCMlQ;__%(Iz>8K6n2p~DU95LY^L(^Q(E#AfHcO!VEWa#U z3}EQX>TE2cVYXDA+44Pr#RT*?NV;~TDpCXM*)~8SmF`I0X@ik5i&?55E=)$#O18+# z3foo znx_iGjnj>9bS7~WfV{qc$jkI^eM~4a;VWyBc?TqQ`EEHl3yRpk%UK!Z62MjYg92dd z9w;D|Vi*9jIX5H=hm@+cD1N|56y% z$Zr!$A*pl1@uA(W0Xw(joUps02m<9g+fPEnf$O+3oJ=7B^I5QX#d8({=#pcqe~#x% z0(YT>3xx7Mm%lVup+FLE20?L5d$xS6g3c{;Oc-7Ir-yp^|L!|=4gughnbc_*!jF%kiYX7JSX41bGa5h&QWXREp!Vpl(QB753x4f z3JQ}4xO*E+ZL&`b^COi*AsKIB8&cBz2lz=Xib-NNohYjnVLsUr0eWt8^iBOmi^xnX zaj|VqE+Wht&T@{j-T8KJ2p_`CC(~rkqNl!5ug68{)}<;I;!8u1LTrP8!-=9Zbidkg z4weIk*J2ETpP@VLD4lZ10x}7pJrbE}=kpuesEFm6AK!u}$8PpyBRg-%t6!FCTT+Sm+$(A@X9%DeVgW zp~7(KdfPq2{D$g>4iJdvSs()wY$Y>vKi_Z;35GAZLG(u64n#zIBiIe^t_i>aoNkqd zuomLd^K}@uG|Mjp{pq^XG-)4#t2@L+OqU7sxgQZQ3=0_A7bl(q5jHWdtTaYh6qD;F z*WmAy{eBWQ;2y0K@KZbB3#s;%g>^bxpmKsEv0sO7h(>I&v= zcQp1qH#BEBR1AFZ4{B4vTXEKU#lQoz%*~%K7k!rvyK!FcN6a#Co5RcUHnY(=G|Z*A z&PLjX+FRe^uM@I<;x^nd-Cm+(yswv}_g|=Dp}k95uz?Pkwc!tM&u}2I`*=r0ceYOH zPWgp~F(Os_G=8{ZmNw83oeOy*k8k1AL%A#pxq%1ffYK1pQ6y#Wzsk%rdg%@RCc;SE zcf-IdRjbb{JNA)+XV_tEHyn`E``m65IKswS7*~1t5*&(=tgD|34lX_sEJ!Y&ZOP@Q5y`TiN%*E`vj~x4veU?f93id{azh&PhXT-nu zvScF+>i8N!p#*k_B<=g&{0tb!Vj<35k6ildy~LvQfoa3c-Z+T*-6!z$$cURSD~oR?0zzP~t90SkGg zzOw4bSbt*UUK|^pKFmnS12Y?_6@xt7DJZ;w5AJiY8mPjX%~!}uQz=9~Q$qHiQQ|uh zZz*Hv24A`O1W%3X7vFZ%-*NNb@WiDSwzn3h0j=z>7fp2DEeBwuWk>}QzU z>4TQn$hOiAoCYD7A52nV*B6`?<#o+$=jR+~juqtWHNIFU@XH1M6R5gFt(^qOeo{nv z?-J?kUOpn7-5r4i8!zVimi{f}>vtAXs<0&;IdHWG3f>F$+qh7QXh-N%4C0dpS?89aKP`zpR zVl+f^T$a$AFuS9qqCnteK}8}$R~AM+#@=F|kv}H}MoF(mYZM+ybUFvs7uft!uyG>T z^kL{G5mt#8tz_ERD@JvfBE`6nRt&^Sp%#zSQ3ZCf#Mx!D{4%_~dCj~#!|K}OwmbXo zEqP_ma>l!8sfe^09oxcKBmiS^=&#S26`X0-Aj+Pv0;H2NLEad@k=GKE7m{g?ic8Uk zI3TT8h`{WF+T>mXtg|x|FNL6w>S{uxJ-amfprw}dcU0wj5L9&AJkk+YQ9=P1!r0$8 z&urv<(lNp5e)X=p7@x{XTXbo%N$u4_5^HP*;EPZ(0(@%inUDh*z`{UtLh|dXPZ#Oe zSdIh(J%AFdbq-CC!uGQ$JbLW@mU|4_8#_>fEoKtb%xeB+veHYmkea;s;(LBIIT5Q#}ur4n0-$y#U5dw0zn!l$J}L5HJgTg`;ETT{O$01__F*=f`=M=wWv`q>OsW zMpmrzi@H0&sZ6bS!SSb6A(b9lRV=+oVZob0yEBy#as@?>o>_z^kYaB)U*V@2Mq>;3 zV+)i?Tj2U~okMy6)iMU?kL3Z$*CMjohvnHfVCNe|N&{)X_sfU)AsXo!w<2bmak>K4 z@)3>{gWx6*iYfHJt5+?0)3`$YnL~lV6W+|ff@#thDwX-{4gV`VuF2cGi6S~;rV%Lh z9rO_68*1LaMDf|1bG|agOd!XO%@s#eC2Y-BTEr7TD-=Huj zxW7rQNt4pRj$naj!^vN!4wCWgZWAQzv^{@HC?d>(1pz12!D<5-?nbhTEs`UyyyNn6 zE+!p_EoZg@h|g!QA=(^*_ti7;aUQJYQ}N9K>+!VH=nAA8O82X}wVYmsi7W^9#etrp z_P-rO-DGsqR#(5YE5oTJ5lk%|aA1==rw7r`b~QnSb9JK*)Xv+|_5-1i>zglS>c+$F z8o|y{3}sw9^K@}F4IVfu+vz#%>)G%;HsYRw1B_YSR=Lu|h9P%5@v^_4 ztKyW(`I17m^CG`hS3hNk97?E7n*=CLKLZYjBG`=mKm7tb)5D~3C~#y;*^{a(12cDsXOGWOpgfd-!SqajE8q)EgRPms0bv&qNHKd#&ZuMrHYM%KDZz|n%Gb9%dLLtYoi&W$q7G7o< z$J)ypZ!@0Oat22>1r7QhQtqvkIkDhY8>=D#Tls zylu-YGlmxoQ9H(cxOH&H?R_I|FC}f9G+0gl4xCHH^yOinVJCi$mpLo|HHk+?Hu@&vJbtSq#1mYHjqO__1(iuLID)oli7NPcQvo&A~DRsCAYgl7Y~e6^2MB_ zhK0U-6IL3817?y)s%kX9*+|?jPL}FGj zmfbhop|aw=y=MiT}+)NF&?{9({rDW0cz$%H`x)aF=5WhUkjN)YS2q?!}Ia zP9I~`OSe=yTDh>d2d?Whg9mmeZPvP%dB`}?UaXXt6C zuS4F&#jngW#^?>!JtpCu+#kff6poW?mGg+|p>OwQ9_5y6>L}&!IpbbX9*5qM(R>s< zG-Ei~L%iJFd_^S9CM6kxJvQ@gd>rw{=D*qyk1|Y*J&2ba`Xo}YC5{Pq}qnf?Fz zjFVk&8ox%oV{KZc+RmyF@n?Fiur`G`Av=Zk#g)oMm%1R8+gHmEW#TK#L0`wue?6rh z6u^eLK`!E3!^)g&&HJcWJb)6yKNvV1fR0w7DVT|~x9)q6Zx91jH;b{z_)4v&DT%E` zy_2@HrSQHE^Ealj8Os;pNlP0@HBTDY*=~-Vx)fo6ph~@I3P&%3gP&gO>+uRnfofWn z>svflrg-`29aqyE{O2$J(+cDe(Em+pmp=x9c2p!h@T}+AS%YFVQgxdO5{eIwrbg2gdfhPo%edB8n{Rp~$e47ih0NT3SmUf~ z1h{J21esP)x)3mqsQx*H3LD$_@3U){ek~ubfin{6`>yI*_86=ahyUN#;W~OZp(-C+ zd;CF&6r~3`Ku@{haP{pQDU!wCOPDp*IqE?u8AN@v&bvSv!MpI1@8TTs9VU?UfJAR^ z^y4-TplhfJZ9n)*nSS)Ftm_l5=o49C*0fI|O~dTymY&74m>d{PNjFh~7NKyP%6`n5 zV>xtRXL$UpaE|Q{cPz=LcY#H8_suqNp0aA5RLHwfyq@Ae?fJckSzc233LE>_xA(VK zs+#2R*BlpN>P!94xh9NM?U-Z1HZJ*+*v> zJGZgPhBF#_Y#y+Nb3-^J$mfm~sM4YEnxAjv$6d2y+c&4M@*L^9yqT)|?YNmZEr?3k zNoaFpphnEFCjq$#<|`l1C7F1Zy^*VDtQNxz-Ze_vV+SR%0|oo@50u2SK==dt77Y1D zgsjY2^6BWe80t3gz}EG)A~*fH48V}Xl%R8ek>&U6zbLMdPvg=Sz}p;<{r4-Vz7j36=tep(Noi?mv6BYWTPwDdIm0B zSdSSxb-(@c0L1g~nkyU1dy&4c^*tJn#yd_!Onx~6Y$RHILUTR%Y0rvbpX$k-_T$8@ zq~5drCi+o}x!-8jocE{*=1-Ga-`K@KW{cVV%*N++2EK&0do|oq`+IP(zI-?I7%wzp(?~2YeLHHorC@!L zs=2CR=dgXofW|l{XbkTnXy;>*rd1h5J!W;ppBBU(u+$}f^4f2ubqDOi5BwVHI~8L~ zjI)~Gjk-H_z_jT*%OAvJ@i?X-w*>=mhQTPfN7^|9=OK?&xp)7P^ze-zNSJ%E6}jgJu-xUeK_YW9(R;$tEnsgn@GDAI@Re$e zNkZkwrjX>9I?HgMCeCeaunXzRbxV0k4ytl! za$Alb$0-7qbXOl^P}c0p-yS#xrkibj_dbz<(&J@1mJ{Bju7?uGzy1oqq6p+#J;}9q z<{djJpg2AJ5^h6Jq73e7>%j@lFv9vF@Ud4tH~@Am#M<=yQ|r*i{-oG)TL9dQ0QY(X z7yhM++2phmYmRB>Fht_EDvCxie`xr7nN#TOcC?B%eH%ny_Z%_O#VCB(Dp5hZMrGi9NEx z*s~Q+%$GHnln-XU!mF>ladarB{1SQ7lr@!($JhZq|D>l3ltlMnD>QDc;G7$*am)%2NQ?~lPw|G$4hlA58}bdtmjt{LyAD-juS`xR8hQmeX;kyO0mDj zMNgGa$J9a;am>6zVXWi;yX2R4|HPra zVp1t5EJ_i_``0>}2{7@;8YA}jvP){>;yqCC4NTXaVvhZ$WevdZa|ezCcY^ctjV6!T zg@`%9w$?R#+XLqeN83dh)RtgUK6jeZuW zVxcaJBKKfLZg`yODcP!F$*&kY*`7W+2>Kdh+IL#DSaV8*&z;NwNWPOH5@C}eamt%) z9_aL;C^W8$wm6KxHr2gb$oHZbfoT!5j4K9Cwu5MZa0gF8kC4#sjoYrY()SB*Pm;P6 zB%^2ee-#>Oq_IiLoc$KN6ng{|Y^;{!Tv>L)^}EQN$Uh3jUq?kDsyMD})awnzYUED= z6jRFCILSd!e1TMKM9&K}8h{|{9_$4xADzC|H8C7ONLN$xO9Kp~f~7u0N}lQAF&I_N z3Ggnv%mCcQsBLFBE^ztwFah}7tskXmFDLqP_MF`^P8c_)2KN2l1#sO{Ca7Ctl*Ptir6}FEF z%D3*p35ZV{A)NLI78$aHR6`NHiFDLZf_wR4 z)B|q=nN9xtenGzt^}Nz8?B~&FWtqPEJzz>Pn0GU& z_PGc69rU^0;6^zeoKtp$sknpMrNU=t7ST*X*1 zgY>|w7MecqH(F%Fbv3D!?*zFo%09yz&{)R<1n!edXiJx36O?V9vKt^2ntOJHyGD|b zCgVC+`gwXdNR2&KGT>Mnf*xC%GHLqCGjHfg_kwfc@2O_4(8@H%xO0>vI^4CS(|+V?vT6@yB5(V+?jCV!l7;He#QYglZAY%|wkH1b1s=hj8X! zurV8QJggS{>uZP^_HNk7toXf)=52By(Eeow*vjJG z#=kjt0|5pM0R>7qz2zkkkXYDS!O zHd-`w*Odb~Ek=R2t$UqyUK&39kAMGS7rwp!%>VqlXWfFan=kclyWR5E{g7K9{nGfv z!picxFRor*7k{GC(<2l%w(&!ixtSY`1`HTbcB8Ao49GO0*x1t>!0CwJ1$&24-uX0Xhlb2l^nTvY?l3H7sI{eyn-tn1f*{jJT>vnE*krzPu3Lh;HD zcO^A7GZ@vr&8UAdhR@}mn#-J%X@1lCx=2O8zH@O-gmUIYiMY69?S3R^bySHB56b11 zt)f2r1~p9(7F*_vZa#8|%SKdZ7}~qs&JIi6o`soMzt-S2#Jz*8HlQf`_72<I zwu=_o1yQ2YKOL-S=8f>aKleIDY-h8B^cmC2-mN z&+XeYrV0eC9iqB-WpqS|R5p~jfaXCH`t=^Z=s$N*KRT}P%&B;qpsQ`rRN95VYBFlV zgoAA4bH?c4Sn!-rGN0>Dc0Gpa#{a=SLRqt|VCS*gTkTz+S&9<9?@;XDz{VtJR%YA$HnvAnb_OE zt&w7oVUJw%6TNi@Z+a|=Zk|%D)wGo9R-~@+wiSOf(j>SBY2)yB_+-){hTp`^ju40<0da*1MqGS`~SN3Gv~X^d40!knx`Cl zCd5~{;iMdsPRi@^30W#QDKYYQ_WObNz;l*Mafc3 zn-W4=$kIYdC|igy)4r)>DUyAykjk26ri2t(O0qR1L}ban{2u4^e!WdKY3|Q`U)T4# ze%E!+ACvTczn16oJkRnt=R6L(#(tCcj)Z&3uJOo{Igg<1xoiB=+n?miuM$oLtaUFd z4c91hPR(VzG2T4EciB$nQ2{KDdK7i4Hn)yhZ2F?wN=jQ0h zZse#7)Mt9zekVuA@?<7)&^u6(fWfCO`L0#~SbmA4>32(%6 zy#jRp4cCkKv4&X3`u;WSr?M0B$Pi(_=Pz8y=Or&M&z|QEdB(T%Ap`lS1aViry!iY# z-zYDWRJYpoTF+re?B0qg^jTkYv5w^qh*xlbroMZwtcW3N9Q#5i@10v_JJE{qZ`G!h z%n35E4;~u}RNO1W=~r4}g(Hh&rp{Cx7tDsmKZ+kn+q-k;EXmZl0mmfvmd5|YRV^FN ze813{R}1hKX}Dl&Ree)@b9Tr2Own#Vx$cpPa|7fc>W4`IS{vj=ZY4 zV<#cW9MlMuT5y-{pkgoB>yErCT4^OUrEh@A=quDUguR{f*D^F)RrM-K>3zdhaBk`C z!%|@tzG0501n+3x2hOp{Mf3sMP3Z-umXk5x4{0wbDA?9zJ-SaCWDo4^{geWl(s%9L zxvaH*TH@qX^Qu)wjG=VPalxenHNN4Rq94~usieoI4vxt_C4w(c| zCkF?)5hF()tu8vj>i$}IpghTaET7oCpqA14oCiDXgQ1upv!Om*hmK=%dHN{}-n#Kg zmwRd`jTtkBZnkKV$MZMw1+8Xr-i_tSleylS8ed*nCsx(S(PxFzMs94y!JFM<-^>Yb z#`SkOLTBH&TX2tU{N7+XA)FV#S{kUco};e4RsJlUco>geO0ukE!noM;upFCCnC*733`nAx$T(Q9e%C`b8dO2Cwz0;2Fhf#*HdC#mX_~-elD(9MGKz#QeJeE_~11quNL- zD=SIKRI`8q57vh?l7Z;m;6@ZUwR+>&`)K{cePn!8jU25qwn*jq7PYikaJ7H}pQxDF z5x5CWXSm3#^JW~=0LE+J91oo=eu~`@QS3;7;n{4MQoT*pL z?ufE-90NH%#W;)~$gwCZ$CWE=DiPk{d`c`{Ji*He!upsM*5fIHQv?J?-~noL>ybEN zCGNj?9cWJO)sSxE$#3+nTVuWCB_+o~w)4v3x!d9gGvSubssOn%TP?0};uzjY)_{RR z0GXe|^?o^va?Y>R;@PFT{kX17bNLTH{BZbHMY6pVyEel}8dA)#UmcaaGxDF^K}%P6 z#G|u~w)L~RTfr1MD7iql9y^rT?kzpVoZoveHRF=G?VcY${}P!Y4+luimbHHS%{Mq4 zvl7Rg?YCntruj=;$%sdYQj<^gZ~5)NTOsE8_s02XbObyxKJqH+`t?DnvXW}dHa#$= z&+huQYlpd_2CJ#0#VW^p1!p{0a*dlB_5fRfuuU{2o$>Vjmho%3rE|k0h0C?7cj)3p zDoFzN2d4D3GJ#jvZ%ArC?yVKy6YgDW-(2e}MW=%tg3s^t=$`A6%sG-n~ z&Rj8Yhp{+HPFbYrB=v$>c!k{oO^CWXW_9oVo@+gwL;gwM zeDb2j80KKq-`#z{$gymN&C@*jz4RK+BFFnJylZ=E5}Uao3FpQ??N{Y%hQQ|oiQpG7 z(Rx}`w7hkM2;;)tv$?^VdZ7~n?lJLyGVk({0*0q2STTuNGPR#sb!9Z{kh$TB{yS7z z!}WgEzRuI#-L~XF8EfRWb3fCyp1a6X}}YG9#fgknIiPzo}a;i=52aZnTyXC zd1tVqDNDO+7>GuVY>>GG*BXA(26evScDj9d3|tDG_4fwIi+{y+AR5!xR%M$q53s#1 zJ^NG2x;8I+J2?VcI?~MNG>=a|6~ptU#+03pq@!7+Li(WS1Xt781UIJJS8)NCsUBF( zS_Yuzd@P-{d0%k6X0}H#ez^0uaOVfr=0vhHsWwe|*n4;^j~#rtnyc5d`S>4A z*@A zICX2)`-%+(g~whg+T5{*mOOL%G^-^we}DUSAJ@G27Nw;ru)fAWYPL_?u)GtVGrLYC%#JMW0Hp4Z2l%&4GG`z+9y738)!mN7n1{XJ#{OHX;;XYG_| zvaMDdcX{4-Q&yhU8}}KLwTk>jS#q{MxUR;eDN{z0lk#7#cyn{}<6bhZy@^`d+QSB= zzT(t-3t|YJ{C{7N9471_Un*)cEAV&n^70rn&gzc;P)v|XKm2O)+j^42MT^GqR=GkK3ZG&%!^bY+6ai?N=Pc-&cK`jen&;EWT)5uA)Km;J-e!wB85s$k zadA0X&GkppnZV~@f(n#d6ha;LEa%hhxUlNqUWf@2gWnGl2i2N{wDd$TClo3ma-$Hm zzP^52SDySHI;hSO7b&39r#KE8B6TK>LHX|-UgdS`xMz5DpBB=_JohJ^q#?!3p$a}& zK5|HZv>ZEnG{3giXfuN!5Na~i%Y7~aUbtk5fLGL*iul@GV}P0Lql*OeXq0u{7n?gb z1O(dA4rPqKbrx`UeZ~O4oZ3-kw$iK+R>SE9o(8A4;C-i8n&}_4chY#OY zY;B3C>TxfM9C_G8I^i@a_RVKb^#9Eb{AzqGkqX`e2wYZ zKlN#KSulw%PO|NqPes-+a@=V51Mxe?nhIM92?_r$^ju=cPoaP8@GI2QE#WQPQEbY+ zdn-2gGZ;|Wih&W^fN1c7iLFD&q<1MMZT%o79WwYB|{@RASy%@Zh()}PtWh=maHuodCz8k~rLxj=({fi_kJWEWJ zU)D$>--Z+WM+^6#?zt#UNsagD0}tU* z!=jaKyT*34GA>})G<~xR6}UafrF{QrA))u+dbE(obI(!SA1DCMrSbh!E?*}q)_$2_7yyq zx3skM??O@94rDA>IxWqU*Jo^p$jJuPca;N|*sD1Gm96kgwW{jkX0A97kP8xL)N3J6 zy&kB^oF>i!R()g2E)iRmf7gcbeO(PDapLSbIyo(=+1IIG8aa_UqNyY2>fM>()x-** z)6Ky=nREYO#*-uD&C8Ij=RrQTK7;8oPFpTg+RIyWeUW6)w?6Ni`}nci0aSO~W%pmC z`^vNNrM>nkrA*-Jm-<(V%Wq&X759%XWQnXYTD?M|dqMJi%y3f~g1*Fke6l z=fiEI>($TexQFq{%eiX`^E{kI=_U>4Cpoq9gK0JAm}0%GY*MlkUF{Q?^&f`HWi~76 z+&>Zi;G1-MHxFHF6yT`Biii%NN_12I*HjnWdK;<9L=s!VA zAo^gu|7a;YHLwScM;)k*wsk?C!pjzQFrn((fuZ-(fr0NlEAkl~A(DB1cg?ubQio$@vQisKo1L$#su<~^!Ae-7oFUxvux&~5? zAQ`w78#~TJ->|}=RR#VIH@_fQte%y@!fjI$n;lrsQFJ_o8O}upzT3g}=N^ zwpF5L&bgAQ5!2*7kvies_3QuWu5wo}HzK+L^;sdP<&N$ZkvZ6Hr41M++X*$0tm=OK z0#rUfAm-`MUFn$c2pDGpd`i}RbjJ$cWmkAn`M}0JG+``n zh05t-{q2Zs>IeeS2rEm4Dw9=g2AT^xTRTnbFwQ-9#_V98_9l7UMIN7a3!y{Xn*K$t2F7uZ?l23}E4yE_JDZg>$HeA-ZeYPS2s_WAidqkq84FDaraU26lW8YZshR2$_#Mg z?Rp&a#x0!a=w6y2A(|wT(%)U*h8nrU)u|YO5Dqw*tHlZ*p^f}UkAoogc`R-f?uuC! zl9KW*H4GoD+EDj~XWmCj#hGU(-d30!)>}xJg6wVK2jP?~c_8-2X#S;UB@;mVSqh1g$a|=c6_4!9h)rg~B%;YTrCQ zYjeQ0si#-JS&?vScdSao^PM~Fj4fVu`GMa0iB9gKS? zaspghY|pWf0H{yc3Ur}$C@y8_=~k@xxrzH6R!FV4@t%WeuS6TCS3@5WbUiOAxsa^? zKo2uho&k}L=}^RwZJP4NK9eR<%E!b>0snOZ_oHjF{z#3AhZy)IQ`f^>-5vZ=2e{Sn z9Hmobwgs(~C!Y5p`zJKmn@BYbnml8ue5$U8 zI&1UUw|s&^vI{8{rKOkK`ZPMgk}$(Q2_vDX0nVU_3yw@%BC?yyal#Qom)CB#;Mz^9 zY|d`Bfn#t(kh{`mg-|IQTIWIvwTa9)m^$_#dIZ*V{7J(T!4x{fo~fV>Yq0`+ z-&2r-+4ckQ1Ws(yvzuAm6HINItGoL_`h2bpvIr1I2@N3v+?rlaFt||!xW&iLEur=^ zxjmFu-(M0L-jb|CQ`Tj4w`G;@-MjZJT6AXErxrG#PL$|Gf+QwQo_ryNQ6_42MC;Sm zheoOy5a(JsW>szXj;;XNd87Lzk5^0?iQuJdn;- zgZWtj$D4JKX=MQBG22I7gt$7Xb-%qI{YVo-efDd!Cn4f@qNk++eK8&S7)!(zMJGvBb&29UNCHeo#E!OR7jtFJL^J{~Mlr2xaz%Oh6r%fVcMLB^ z^ezWKd(AFMdOudkKKDS>=a4pvuZ4lmISS_JRiDcfny7hu2l+uZh5!Dn~<RH8=Eu(UoTPLc54)Z`rZ~{^goo#IgdiNY@qLWP*;z(a-IX*vJX={sTIUErxRf%_+c7X%S<1I>Xb8> z!MOtE7x~iEn%~Vm!nYGpWY1_Ee(Jr>pJzE{$vvI+*XA{s%zAvUjRz`*y$yJtQGu!)oD77(90qJLu1aW9m z<9I&fVsV}@sY7gP5fRpHtC}JHGA#sK`9*p~4mx)@3 zW^p;Fjw7yqGR`g1L?_hEFqEBJK;=l%BU*HHF3?5|xvSi%XCtw4?{C zl#BxugZFTo)H!Z~NzIXLUwxQK;--<8>5&|jcou6cidetm=>j>NDq=#U0#Rs<%`ap*5e^Ka~gfZQh!HAVr zV}e3*>SRPc<+`PFj?hRBV$pl^=1tZvU5q`e(Xg>5xO@mBekfboOwXYKP8y z=`Q{rZi1q+ATXcp>Ek@`F0-xS9Yl2U`7j&=2ZaZ<6#QvB_8>keq1iZUtDUAKqrE;J zY<5)`PpY)^>P6xUF)}tjiB`)QzScR?H8eWI)qzng3Jdq(Ux4uiN$dwMMXXU!BSDrL z_GpD6p0TVNB19z0^p}>*8b_R&;-16yxTuvk+?QpWk^K^D^@^QH`kx8b4KyuMN~L7jl)h!5%Vm$L#IxTeSM0F~G)DSiT-`n$u${7)DR7aAG}@Ct&XFC%&_^ z{y`lk$*8R zT)%g3%nU&QMd52$TWBg zw?dHEUeP&ZxS&blI+*flh6&t&EQ4}DYDDSr$Y30^(+dj%V{BNE>x&C}77Ux~>t4^0}+7K6`)=Ae+S zsjaQeG-;%!z#BK_-?U>^)`6J^_J7`x$AK+7cuR|>s{wXt)8fDx%nYwV39sx&~^@Y{Wo7h9Ni2%;ex)c5dJ|3#uc zBc^6jp|Cj*U%8;I7OtuUSDwS5_pAPj zD_KQ$`^MxrZaKn>&^kTRVhTlr@{AypwclZjh)AC%U?%Le1u<47#_^n{@!4dWs5d@! zpsxvGOnJBsraG= zJc*yoBd6y$dVvIx4rrL%ZPtOCDoYLo+W7C$6ydtky=O&f?gPW{To|UlriOagh+T>( z5s0O}?iw;MOcKS8Y=huL2YvbX>o7xT=YkF9{} zp;6z|LjW3Je6DlaQp_!~1%$c-vDXctBs!^o%D)4T$+)>X6G$^LGi*4Vfl$mLq~S65 zK^JKu9I*wHjdU<-f>;DIz?)6XnR#eI3rI#?{C!*S+%rR^kp3C>rYt9sW<6kvQ34mA zenB*KY|pcwBnZ2Cp;9l}_m4u=IzQ0#W`Tv6s?2FDa}P%;8<_Jq3iG=``q&T^4H8xe^;b0NMdU?u^f{;7-7L~yHLL?q<1_G7^1-n z!;bvnn0gbLI$OMG<&#bNXNWwGrfMAy?^p+0PTXmB8H~J61*scPVC$kM=N@G};;#H; zR}bbDHVnQ(b>=jdewY(XQ=VyU$;6^w`A(G^we1rJR||JWTbdpb2%~jn?+Qdak<#!O zm|>d4k34M}O{y6&Vg$`>SV=QZXfz5`7tMk-K`NW_ulf$nD+%0dZcJ*rkJsBlEijIE zonFJd5O)_Yao%Pu^Ai0aO})kQQ(^ON@vt87-$;@#s2Tg5b?eW8^kz5DbJiu|GQ`aH zXuJa^m!!73tdHH9dqn_^VL4o?bWYDcjEf|Ox+)M|<;Xr5C_x_<4NOFnfTxX{>Qy*C z#I%j+udUn}2Wg|gx3Qd60%#;&$1)`NSs;Q2yCjF>9<`{6d2r54G$)y}*rw3IL*I-z z`Rx?jsg8hc?qdyFT-NU=R&pABSw5fn#jC!^9q( zb_5dUyg=e0NOTNY4*=B4V?CKY-NeH*x;oo4D}kpxlvP^Ewm+y%A*xdvX-mVZnuH9i zTe8K^!iRq)!1#8~sS(Cu-S?|`Mvwi77>x$$WXURNVP&oo$pS0NTjgujh1tGn zPBpmj7V}^N&qnu1(!Xj;c=6^`fZ4$M7=xu?Cj)}GDq`@ZNpwv@Gs&eQ8fEz257+~! zI}fzIwovu~2#4`-bBJ_{cww8EL>|ZLw<)*>w+3I$vEhEweHPGesne5Y=%i_x8n*klsm@wR%6co+NqQM8OKq>r0yfHVRHEY4mP_F*q4{+jaM@or( z@fVz0b?JGmBlW*w0OYHB16NM+JZjGD=c(l&tM{^SJd#(P5AFV z?1%uhgYY{+ox^c{J&2?Q-#wq6!b*J@#frGjSQUPDCyZ z`bPJrIT{KT!QO*MxrZ6-J3xWRuwCFu%bCEJlleP7S_K>Fg0A9K@cPQl7c9i=ZJMG) zZXcLUTDVhk6qX&(OyOBfnP+F5Sc~bLoX4dU-aDunc3bi&lKr&#h8-ayQ#nv5F~sZL z9ZkMAK>>>~DrpDJd_=r*A49`i*_-4L<$qkM#abG~rix{}Q#E4~rlQi3MLIbH!5#0T zM?d9h_DYhH{!yK$B!gM(fb--UT1tpXwlRT@(+-2NNsV^v?63= z^4>f137BI~qmr_&t}Xz}RlxY}p!dI-X2D_nPU*7r(l=!{QNR ze-!LqqZWwhy(x~SbSiCgIb0e%q7q)4m`xnPSC$U&_p%L>z2_*M8)9Wu23K|+3&%_i z0MN*clY>N0kt*G{9A@?&%EWfSg>8nvb)QC>0@qSrlnO&dbnY}0kk*9ee!$sW%?-eu zUu)nkP|TDEERb5HY1TI8u;tS9kQSHNO~{>WC#qcr2n0OH9w(u%SXEBJ53{7vdzmF2 zw@?88*&SnkCsIV5VKfXgrUN7)leqAD9Y3qm-W~*uMcA<|M#a)3cNchRok(6*Vo6T{ zDFSwI2Ds7+APN*olFwX=bp`6Kh&99?8AMt}w2WVT_00;9^%4;s(JB%sL6k2|>$pt} z4?r0?n;>HfX^AHplkqO@xssTeH~}>HG{DlFkgNF_{1^62p-1<#s8bK0%f4g7{i`3! zx`?_80A9uNpP9D|@%r?~p^-8|gz6BL8!DGFoWJ;LJH`s; zJ%0Qn4PhzwWs z1q^<#>EVU2+U3sj_dhx0&&9lC-;keR{~=SAXSDv|lM4HJX7(RG$cb6cQ?mc?$#XtD zu;L#+=-j|FA^-3}jzK67PXC8bPL<^$$^Y;{PBG8T`lk>6Zz=z2h5p|xrNi54^}Jeu zPe0NBB=pnkGWgs79rnSc7~+Tj2ZudvAZA^h|AwLg!Hfz=8y(B(H4lyVT=Z01wf~De zrQ@|Om)1N@9Q|e8sD@G-&Mk22}&0N};cg{Bg_d+1E;Ir|%OnKDp?O zgyn0cvSOJ5*#Y0))4sZiO&O(CUkHx8Y}uJ|MQ|}=o<8-Wnd>}cKK1**{qBDs{M(Ye z|BL@E&-<(Kx1ayFJbzn~_kZ!f<#~TK{`T|#mgjFv^8PRWw>vz^Td6|%H!bvEO0_L&UX7iISe?0buf61|{K8#AS*pvNO0$0fU!y{yp>)i^cr&1i9|+?(N{7BT_u0 zJ$p_aKMvMLNi4*6OfQ;m_`r>(L02lhaG*IqrB($h;oL$e-o>3Ogo3ygg}mPk{`pND z$PDjcN+XS?`spXa^(bRR>G1iOWN^zot-=D8_qJH!9Yl5xTq`8!!`FA5zLs0?J9tsv z4{s|_@*#jeH5YX0$W8&{{^q3prfy2Xg<5I8Y9j8HH2;O_Zcv)+D)XUtxcHdp-(``x za9V-`U}aR!ji!9&9aIoBqjDrv^~;=Xz+2`cmrQejQ7#}J&MjWzJ&tq5`QRazp>ulr z_W9+=Bh#=yV$f9z)_6VR#Q@)_bfew7X+muVzGMar$ zDQoMIKN$D*^);g`D;)&YV~}B0@nJB@!Atgjt@V2FFCYBioNA0=Nb~L68;ILHFf?>F zE+<3j>tj<62+=H;#ULbn&4nv@-+UqdK;y$D<^2n%SD@NLn({QnrcOCiTyB`+*(ppi z68WKCHp9q%;f22FkUP8%<*lP=I0|8f>qFC4Ue%ea2wadbR*bTx8(EJYyfOXocGTD* zpKCo@U0f*POl4`<%u_(*f>eJy1r`LwJ4kBXK%HHo@CTPmv(s2Tg!t`CxkGcbk(Rfu zcHa~b5HQVpWbf5?ZEg23v2h}mNA&mi7jlaTdDl#{(;V`0OAn%35C`N0%~2!F7MGht zRZ3JZ3Hh<+Rz~qIHIEZBgreZolu%L{Svd8%o>;Wh2M{i{9CS2Y=tC);r!Xx1Ls8+~ zg7eKEpsnBjGV?r5AqZCeb$B_{%ntC8He=Mlif~P(k7VqJ=w(C6*sn2QiZ(JaVdn5! z57UXJ5>?Er%quRwB*P>-2k)A>W+=JeIyD&uU(L{Sp*haOH)7kY_U{kL8T=;mgN55r z$8f@GB=ISkrg&qsvJX?m^w(F%A0AR8D9jmmiV{nO+)~xS@9g_!D8=VIoGQix!^4HG zS4AGMx6i$-5k59KIXSPaOaPPe*7HkSynI+3w^zTREJR(m+N{FTS9>>#fF=rnQ8DG#-^@wS-NgW@NN?OCofn?XW;9}PWK$Z4qhSbNYYCR87~%X}oK zu31AaNhs_WX~%0A_(_@qk9i7)zfkL^pQ!Y%6Ed`Z*WgYvKPZg;wCPot1j=mWI4|S& z6K0s*X}(V<6kK=rs169pj5YQo%qWDUbl`>=4?FAKpZ^j^*JD_k?-?Yi| zmLZ&%yow6nTOZm$jz;ECFM22!CAvGP2ou?=8GD|kIyq7MQ&#(jSD5NPlyuKqyiii| z-bubWl%G|38Vz$ITjsWts;DZ&kgQZm}w z+N229#FjjN?%MF-sZ_e~8S#v0H;V^?*;UHU@*^?ea%CkE5BNHraIMgclds0gc&C4z+r^de$ zZhgVvf&~q!H9Pgm52nfFX7&*F1k-=}Wi9P6yD1a4brmqZZ0+%ZJ}Or!wg9Z^+$|`v z%Q5_DS%-ei-Yb;VgQIyR`zb0gEsH-~9A~#7chl)52AF0W)}T5ducX8~`{Vs^9mwTcFRwH@zYw9 zla?mpu9WxW2_H7n9Gvru(RF=!=|LHo!$5d(hK5YC9!)s`k&%rLvtmT8Wzw{${1yDf zlivC?Ha31y@!>N%yM4-2aXzKLZrwUEr0q==f|!kc66yJjKkpu^LN$dsnqC(>_S#;S zvs(bC8^J+sef_Cr{y(A}`V9U+h0lD4&bJiuv3XHcbW$D}K^j$K1Y$7MyMVIU`o1aJX&6`v9U4Y~lUw^%F-`DOgzl|04pOk2IZtIHQ{5|jec;Nf* zzYmv4b#iPu*sjGlx}~dB;?F)Aq~Mn)MA@8s_;5+hypYtxI=^G+0jLVl+kHC6M>h1? zv&}y8PI}<`N;I!JT3J=)?&TE$<|0gf%ldO)W=4i0NN68;OEvk_97j=Cxt*hse##w@ z4G}@Z_(U>Gyw8>|Z=|oR4^?%hi3)_I18*7!8} z!jBAB^Is;iIoeOjL$piqtfjli28|P=XDU0{Wd@u+{e#aE`;8cUfei`8tb;Y__wLog z89RCQtY<;p!+bkDiO81NYY#ldqfkboyiTt9aKy&z5olG?oHAv~xPyG0w;jXA3T@3+ zH8zeAlTG03uu1#j*j@kl_{BcbmaELo&27q$C8OZkr(^p-{Pf;ty=y_THviJs*XI-6 z{5EHxKYn&^ll^|QosDpE|4@LLzd2w2Ba}O>`xE~NoPyuF^_7Z%jMeJ%=g(u-*g5c% z`y#x~G&FRr>*&#=tx)Qn+Iw41Pfr93i7_-GlNK6>K~YC&-;TJG+!Z#{#@)js z1O=(}DE^Ug?2V_(pbYH`N7r%j)3{UYrPfyMh2(3XD6^}+?q zwq9_AtyEORY16fFytDfJw1`bnSFYH-nuvOF<>0lRIX6@|@!k-*;+P9wUdGujNmczL zFIHK}SXbWre&woFA`2F1V1kJ^9GE~%z7zF*mm#plHBvvdKQ1*EQW~KVd=i(8q}_q9 zA_810HihXd`e^f=wH%bjfAqWe{u(ci_r=f!XLR)iqLAY?hM-z3=8s)uuydz~jEpvF zF|^~%d}avir#R3nOj=|@_Y2oau!vps^Bzdd*eO#Y^&PubVr~rZAvcZn>s5C39eerG z0K6lckBt&cY25)m7NHsLip@lY<)k`v!m5LEWW>hGsKq|2*Y#8N0Dk0Q4?T9&T#>O> z3RU&V$GG`wbScoSFyso8$Im-BI9Qm>mRbzWfL8NsXmSz~5mAD$#WsY5hR(sy>gwv+ z@%#K|gk!Vm8+x{XevQ(-wK#pTw>ZV;uUv@)M-*z=)}Z~Z2ow8EOI*&LErNNyZQo(G zM&pW{T^oY9dQdv7#{f3X2ummJZ^xX05^qi>zj9)$y*{^<}hPibI!4vH5M;@T)Jd)7EwQ zGfgW}bWvUR1Z)LC-~_={%kPBa36C5*R?$UMPnAbZl2pTz75e@1RabZS^YRblS?)Gw z%8UvvsFM`<&_Q@_i#elCLSq@DAI!t7@am>@XhDg#`YQ9`!@?^GyF`n4wE#=(Z*9B# zjcHjTs0^AzE;L5qtQhA79QpFgFWb&K_1?7_IfHf-#Xpa+Vtf-PhGX7L!J3XoB_&(H zbH5QQRv55yIGRTuTDiW#gtd_P_;~&$mM9^MAZc+Lz<^ZAZGZd=C0L^0UhUU|W-UZ} zjrE)78&Ht`Fz@E|PKlAjhp)$YRLeHyHKgznU|q)rYcV=e+R;%}NH_5o&R{SGzi74W zdhvSM^5qLnp3Te7k%t~K+_mco>W%ouj&%#%^)yUVFGImS47MRAk$PBMsp=gw>Pb@o$2&E>`h-viAbRlOgMT&AL=quRf0YJe6#l z(l0t}iolQ1v>Sz`Arhy%VCJ8o)vjQoM}lp0z#o77fflh@eW<0fcz4!AT+Pp!mJTRa zIJz-fT3QzPeCWq>H@0(ldB{gG>-#(I+|$|DQE?meLDzK>G{CF0$mhaL91r*n`RD>E zM6a0j>%ui?%Xrv3$l)A0IUP)U-vpHP$@MK#3r(ME z!Zd`mPFW0dK1ts34j$C7ryZ(V()79Hql@#GqojHxrmnUDUn?o!_5hK*f_L`Frflzr|dc_k!jPG z{q)mM-KgR#jMy;e>#x5KhcO5#uyCj6J91<<#*E3>zV)lSob(_w^D!C}Zd=!H+;Clc z19aI&G>=#R`R6v75ejoHdfH95?CHYO(e||CWVodxee$W0YO}DSNHP=hl zCK7sM*0HOvWOp zHIbQuUZQd-fg_8-wK`feylp7h53&MOGeLvY|X6rfq;zCOOT`j>LW)Iegd%J(` z6?E#o;U%><5(!mzx!&qFJhE!wG$ z$)x9?HgbcnWXIIQC=VMsa^t>z`yx9Ja4y=wc<_Yz>(Yt=!N#U})i^g^P(SqKYc9+x zZr{Ai6AIe2G+rG;Wz|uq36;HD=KBv#NPoYr~^UJV>o`*`xYzDPPh$nBrVJZ*Fd0k3&#jPrMIfZM2x9LV5C{JVFp+iJp?Yqrw%M^Y>Hm1q;wK}(Q_8=E!TlWY_hP3 zshg3FsX2-`5*v}t;2Se$$44n2?mA5|y7F0#TpxenT?Al~5KM3sl9Ma7a(NS}mmC2a zms*^njVo3huH%ohg!7l&^~!tL4?h&b$g0%D-$G^fZa7Xb#_q8u#oJ|O7hz79H|DJg z7#bK{R86>wKu3u5heTsGrbHLQw5a&VY{2xm5KB?aSPaG~q_KjUc(XeGcxeP25KNJ& zQxzbl5T#?fPKle!&mqk8b zrHw%Rg?Q?|$8`)dFOJ#qWk-j7MxVe&vL7V4PVnNHbFm`{_8|D}EP z4+1y>ikv*Mv6Ci=gcj_U$p}OJztFsSYVi3tXlS_Np*=5N48xJSs++X`Io1bjIePr~ zP>>7k2Ct#g>ecgns-8}pHcis#@l@`*%?6vQl+QU+)zI~5J5>V_xHV$B0_j-u#r)mC zw+sE3e~X!+mvVmfPITysA=y{CyjUHh0T=n~@P=O>js|yl!xx#iig&wsaRPZV2sAfn zYkSr4m+;Hksv+7^gGs9!%Q1Teg&UjvN_p?iY%69JNP%>?sT4<}kTmFr?j7W#ZCbrXZsZ||+pLpI_<-Pj#)(a3RN49dy1 zn{OMtKablRMfP2I)r{G*=ggR~0{(^5R^Rm;gEI6&nG|PvOGPQU32n>v(2Y;6`Ju5X zULT$u@h;rjQ#7|gcZSXf+bD%DtMC!`I z7?)U`%*NsufPo?cat~Yq%Q2C{$8$d~9}At1l-tFeEI!&R@VLXH%C|bXCE;LCEXv%k z;c+;Oz*^JNnZ4U==#H1Z z3RyXM`n06acI8#8jv_D$23gN0Xv+}G)86tZX7nb%=SwS@+qTWy$)+PCiGw=$zJL8N>R+Fi9Y<)6Nm6E)(Qp)n zP^Un8ZtkD^jG%!?ks|EbCX*45?X#|Z{0p<=uJwqIh(n?PWCY5g@nJ4@z+{Ct49KWW=G!M`ump!!e+Z*05B`~ z6yuzwU&Nlzff7m`=ufJ1(=kQ*3NFT#x*t3{pl^*3#)L@31g$tRnqtwq)1t*Ub{s>l zegwvn6#54A3~1>`Y7VzL`6`^`q-ym)W8x=|Pp21F@jeW?;` z7}{F{F#R^O&dTePr$s0GvmQWnE&@oBX+RulRwqlk5nPaC47-?Nb@B~)?C6-O1>H%= z`YGR`GNTYeuvO-+VS0Lc9X>1NbKdZaFTP-+8S`9hOaSCoc;Uh)SO)JU_OU5_?OU)6 zKf~LdBx4l@I-G4X>A~gQD0bh2?U&F4t_y2*2^tOty$IbXw)!6;+uZ#Yt^B+wQ~^&% z7giGdJeAP+WSC`UpFlmdRVICrzEj_Ac#%`dI#vJrrj$F)=A_dHYrkV47)h%$FTDJ?2gs1McU~ z`;%$`&xNFq>l~BvJ6gE?{_Kul{}T1VpVFRu-p76lN(@hQ8-^np=cSsp5OEawy>Qho zsop3>|MWXxY2G=uQz>|W@&mpqg|q7+suh6Fs!KVUF+kEHfY88U$pV=E1pZrpqv0<& z|I3yw8!ITdSLW-ZDqfOn5Ew}L{JGIZP(Bivxpz+V3T5Tf=z#Z!Hn^l}J_r6EW{da5`GBRgz*f1WvWCHmcq@Hv&xmm;uy!-fsZ2X2GY^Y{N0 zB0nzzn`!tbTJg`4#SeWcgF_2!CLbA2<%JZM}1T1Pt^O02?a1()@6sjS*6WNR;NIV9g3(V!=e|*HAvM%aY>e z>Zh1s!+jQ;FT)Bl-o;qXQJ~L82$o{s0OQP!y?*&#UzuEE(`pr!4T!#PnU*Xn5Khnl z9yhpEpKutQb?$!t+~=UcqcGLUiSgvKZeLanUFE_bjyrm*W|>2+7lC6&d6z;yif^po zy5fzufh$5uNofI2Jn(Sivc%hUr;`%NR=%qbMSP0`vBlrt|9PrYYTXYLegt-iZOgPe zT4EssVmf2D{QP|7(3i*IaUwd{#HiM>apQja>Gyhd4y~m7rAxa@%8?&8D%q&|XMdwi ze5!mrzC=&?rYxz*N-1(b03Qk?!M+-@{_oq&nY!uiq9ICem``6rgzS%_gapAOs|UU! znqlWarxFU04Mz_9fvA2%C&#J-fOiKZ zWoK`n%}sC6frcA~mvloW!_951Zi2;15+3XW!HBwdZ1dVYEjK|PQLy(EM^Ghn!tEUy zBY|{gS*?nd+?~GyzTZM0^G!%#6~RkZ*%eO(9YnFz^6nkHHt)-D_dvyR}i_uXPG&3Ca`#Q ztOEuIyAg&>=sW_3_D0yIaG)2lO<(RCfjBWA#>2E^4ghu>(p@;ZFZvs`v8NZYa^;Ra zrkGezfK6y=Z{LXWAhPs8WAtEKlUkQv1GF%s%x=^oVc|7^qY#^Hw6eN|_{t!@Kb5db zAX)`z)l9JYgd>!8u;sH-U{Q|hxhJTkffIL8_52L%M}2tc6^-lKw>3jm1qfHhQ)EBY zf_g_FsfMFQ5IRn%Fp4&I9lH7R^y$m7r1>Z}c?bB)DDiC*gV%>m*y@XSb za~2#DLYgMgu`dy}%^F(|0^1dEq6jU$US?xS!r&{f50kOT=gY~+3m7A$qK{us0%gWI zk_8apbmnxd10L!owO6mJrzbl%cO4c64f}NVBXyH#680SA6Tr@+ZeS2ZS{(@9mZYYg z0DqA2yNT5>m6=&Wz655Od*_^+u^~+D7etO5fU&>B$)B_CikwqVJxY)$F)(_j(rQ;4 z{yosy1up<%vL2Y)Q|K!k5@W5N0grR%wq}py6j^}6@XVAY&!0ZR3fcF-wyLDDFDvF;_KHOiW9oi0$+G~-iSMX%0s)FCX9xhtuya(sO zD{=pu2)t0B0B}}HxKNOWx(aouh8rgJNQ(f+zu+ORABnTB9cw9$P{|(?0o+Bk@{Kb3 zF2VtYAF}n9KOn;csJeU-Z@yzgq)u^r9|FSHh$uA>_>m)xoL|Z0er2HZBrydOovv!# z+=(zKJ5K`1j;oNe-NZsX}!NH{P5Af-wg;B^50Ky>VcfRYZ~K4Qm{NVtSzK-W`X4 zWOiIOs4nw^HRwi#nwWISEkG*9zyd=gC~1@oa!zB=Y#A3tnilU#yLC%)<%$(Xa1aO{ zscxEH`Up-P8LYsMYMB_}c#$IqgO?BuECLtFo&%7)d(expU_#=3a*M zQQR056vV%zXfAI47-a{R4RUk(PoLtO6M}%Ty7tmQ38=mTp%Ws2QzBdLYbfhV7&UCv zZUj5`uYB7-Z~MLPfb~x_jrWFZL%JTuw4N1|H!CjIx((Qcf1pVnwbC<@#K(Xf#oWF_ zsi~>J1^poO;Yja7{-O;OU6G2sjob>tv~|#-9&budAw)HX{E6t@iO+Mn23SZ9Aed$E z{BAuvJG*e$sMi&#sicWYA`9vYf#oi=YcU{DHn!~Yi%*J3DD>$b)PlsCZG^gloSzE? zEq79JdS73vxl^kA#_|ylJjAZz5ahYuR3R+?w?l^jm!4i&czw$z0!tANizSQqUP7#k zqFhR&x2| zKZP$w@Gn-wD3w6)ZCnGL^1Y{hNy4a;kg;jJ28pVzk4i-%==SM$)S&-SoAN!eevp%p3!iQw z@{}buhkpC*bjtpRB8^Q0XL@8C#sX19RTl8F3B{n-45#x1&mD2+&Q;}K;TZZ+-T_0< z3DZD(!;NIBW&u;4x{1lPHzn_H4g2%C$?>_Vk)KL+upe;R1GP%1!keV>@^Xq^P!8Mn zd`}duTdY*?C3vN%^;`oV!X25Qtv&rPCASXt>@Trt%7FZm8@NRZgwtQcCkO|32f44; zOSF&{^T)B>4Z}ydHO)J==F(8VGWi_vki_b2HX!#8J?I05j48Ab@^VVaAky~-CCDGE?*f8;2h?Oj@lksb?OS1>74gTh7ZMW6d}fI&E(wL-NKk?& zPWt(y`1BlM;jGO4==Hd<9O88IU>u4fC;;1;b9r||bNeXVmrx)Q?tw1ff%(tBOgl(k z0)d)H;9O7odiXd^1azilcLb3_nbKjiS!~G?T@cV|Iyl^UQ-ZW};bmz{NfehsHp|-k zQ(MwgRDk~Qv+h9`hNDR@+6BI3qX~nD{7YT&hZ5YZDfs}THn*VPtHM!#8 zQSa)`eZ19^R}0V`0T-PLfH2`?w{^oabpkL@Uqz8A9WZRP2I7=~{Za5|7acT=^v6-w zPOu1G+Vye_yv+MnE;H(YX%qsOd52o-2S~ZM_dzP2r+nX0Rb3qn;=A3DYzU3(*|TTs zX4oA#AOge+24VwY7ZsSu{*)qvibpO8C&1eBw7m>YDn%@m33*fEguKl<^g89CN_j)T zith`gIi;ES1i=Y_dx}%afKR~@ElKefD6P18v`6;ZJY=ZB3$!=Zth=+b7RNFVDNKN; z)3?Mmxkp7sfzV|k=>~Y#H>`V#^qrm{lrsIOjF#Cz=hC###?A3gI3SxKMyB@g<`018 z@fLh6Xr-*!vIYw0Fqr_k463jwj~ZEkB@Dv{sz+V1Ro1Wv;gnK9LS1?oSa;YwFf+Wu&hg;fs0E) z7IF^|;^cvDYtka$eHUM06?bm-x+#0OHmg@a8v!P zQVMp+QKWCcQhjRLllNhi`^Tgb@P7vqlV6o)Pd`OIJQU~I&|pWBdLlo6zj(} z?p#Sa779oHpv%9U2OdE!g0@x@MG}Z=O)9AB0cP8<$85j3Q?5*7Q|1~fq{8P7>t8`D zTtlun0?0X{z*uWpS*oxD54$Q@?qlECH^TTlMe&A7$tR+`f7~f~0wWSELrv}BVCNwr zyU-x>N5pz1vEE}uHeL&()ELfKtf?8whR6b%+7}bNhuJe*VsUn&V57p2VvB7k_6H_N zpe}y9w`8ZSZ9MWLCMA|O5ctexlYEYn3C@Ac+rFk0x&Ayk6jY)N!bYP+e%Enyd}=z z;K6suHyeSpD7LY&ByJt*K4PsUu2T37Gxa>S1_y7{(oO}e+y1T$oi(Ir5q$;1_!Ei> zfY-ASB>@MU)5imiSv3ji4NSkf+w*&0U;jR*j6S8cYYU>c-E|`fW+!~bjgdOxXo1U* zeRTOnU0pcx*JLILFA?*mECS#hMDb$wtkQe`4;a4w*`Fy(bH*sP`y&p5qBQmm23suG zjxjiaWSddRPSw_T@18)1jpgSzDS3rF0&4V3OWI?M;mOI^wFFhmqMU)Uyf_RyyFAKp z#)6TNc~DVN;gJkoLX|eLCJ7R7MW`>zYw5v6IB0vi^q_;^0Z^?6{<;aPT9|UsK3ht~ z9*gk~cIPUfj*#_@TB!1pM6#MsNG&b)5uBRmv0rz*Kp+hiYc(nxQm_dsHKjL^uel0G zHTDfk5U|%dvB?b=xfwG6w5_cLXgUdo2mCz7bA7*UUmsMYm+BWUmSMTZiR)cMxri|; z8ScwqE&YIcr$2Mt2xro%w8C5-_AMMSOG!H=pO{2{$CF`js-7UWGPToR& zlDIb%29dx~98l*bYLCKxsw#Pa8EGEL$RyMv;*B-={Tc<8A3et|!vf*r1zXcN z(9>D5$p{2XV%;~8zCc-^Nwx><357>a{jI*S4O%}#pY2V4=Yb&wbGO1W9zaLvLKFz3 zXJkw^orq)uHtftIgRc?N{oW@~{1eg)ke25$7%m))WYli3gX1cc=GIQM3I|SOoLvqe zArgkRB<53&TVu{=O+)4tB+pQE^oPfFSRXpn&@var?DyB9Yz0xTVfItZqYDJaT^!@i z2T*?_6iKx#pgeV?&_pZdUq{GmToQX-YzqYbreTHvB7Y598N$zQ-7S#la4aoBp_(LC z2rm6a)hvQu>(NONYubePKgM);!q^XdY+9p$-0vb+jCWEvkwZHo+q9)Vg2N5Vr zEc6f%7Z)ci9&QiG0NBIT_w9?uT7H9SqdB6Y=%cXKz;9gmrMJth!KukdOY2^lsw+_A zTFAw1fJ+};Z5@u72vJNB9MUZ*hM|N4Kq*CRz7bHp1t=}e>JqK^6h2)#wZn1Qr~E5c zoD)j%MpP2er3!^fN*N{}w^)X%vTsK68+EBtJx6EdaEPff{C$tTW5xA!kVcDj$k=cP zo>k(0k>BOs@mG6&7I*(HFa8CeaPbUZ+ky}X&Vo%2{`(GrYirbI$ zK36Ar!*@Wpw2+{nV0Ii<&-(3)UvNJW;6@~sVp{-?`J={dTegP)#ZLe!F6MByUdYT` zO$yF9Kj;K(gH`9-Z3SywgNWN23~SerF#MBA^j7In6&V2yrp1#acAkc-k7P?3p)x=n zB9t#J@YqYf9r7U_A2z^p7s)xPRq}IicZ?Yox5n-Dt(L`gp8zif@b+_dobv0UDBs_H z`%Sb$3yxhtV4zqeM?(gA<7ap(Xmz_EFFTFhLWzP#TMyJ zz6h?0V+sNlM!Cob zuG))q*zAG*N=9jvtC%e6jAH?A7-xG3!{9xI$=`skyR|k*WKq^;t<)#f=2WS-pI^g9 zWF(>;rURlRVI~kLJ5J|!7)vVMT`Pr(BPy#~D;2XHdL11DaaMtF8sLdVTixXY(gAQS zqs0e~Q?^uTR>|Jf>*ywaYPBlz9f*oamxf4K=maadGe*f~E(#GUoCeBT1mZ(rbQhYH zKC@coG8_eB@X5t0iUAf^Pu4;M@q7e41udi*_MjQwqv_%cgnc;Jng~wk`+Pao8}HpIb+W^ZrskU`#||bEB37fp0oim z#%)wJucbBHTG!S15=Cnpumvi+UTVXl0joUgCa%Ab}ug2fgwHHVWmsLhw>iXc2MKM|jmg*FsN2uHn8+S!91Bw3v!v~vV z&7;Uu>1s`yaR$x@l-F&C|F69-kEc3s-#^VfGyR%prlwVe%rusiGHoQGY1&X!DygK2 z2q6@*oYOqbJdseOMWG4FBqB@J(@d6-M3x9=vhOGB!Ew&{UH57FemkY-_4@w*_+t`A zpU-=_m+QXn>&{QYfgopBNtM-T21^+T!tqa>HY08uz}^>tY!xgbfpN-j*}%YuWtSN^$V6REiLm?uXlS?)^BQx3Kp$G zdmchf6)lb_u9FGUXtmG_w(xYe2=#9*SE)kryGR`u*=pqT{YB8&h=jL7!5mF4Q9IEu zQ{;-m&SK63DU<+qqhJ&zO@UhT(t=(8`0TUZHIpAY!5)+%*+o&)YPCWUZg>D+Q&HG6-Iv{1 zh$86{7K>H%6wzB1kw>aWc!vNvokv+pdi##TRa0!9!7mmDPx{>!DXDOC-xUiCg)qn&)Tx~%Uj;}M|=0=cMGkCP{f zIfY+47w|_GL{)DsXEYr`#9NNTRyVYAG0(LV?1}KSW3T>&fZVcuBMP+x5cl7c{;DI) ztzi;KZ*^sUL}#9S=26*rQ2sIF|Huh_n57Gp1|~(N?)Lt)M0{e|{%=ica%g5vhJnwD zYUR+t*u`3G6rc#yKK@9yJkIoGS^xBD({LPQlkDsEic7mV4miP>d$2O}Ctqr010I8G z2c}e6|Es--G;g6?8}jrX2_oB$yB-Je@FuAx5bL_ z_*Y8t@%z&TotN=kJ>Ww{$?yq>EiLz^FFoG2t-1JGRoPdr&!P?Gs3*%sh=4?er7~Vv z?ArVv={2?2aKMR~UFJs!E24k=-clr^%>_aEiTzV#S4Sf0DS9f>q{254Y86cms^f@- zle;`!cSPyU)GsmIvL$K2)Abp!uIH1|_M$9Au;%kQDG1_!j%qc_a<8DpAeiU*q!iGh z&LO=z*e+fSEztpR81l;cy*xb!kd5t4e^p2vTNFnStDX1ViS9K3B|E@|-x_Tp4Pp%~ z#-*!QW008aN#70~U^-uT=M2^tx8*qsuWt$^0*o_93MastLM8~&K3x5bkxaH>~rU`Vp=!ak8YN;n7(N=jQthhEJOoMV_^lCyo2Y_P{hUEIz^ z)PjyFV0~(N~fpRBfT<;hNQHnr6kAN&WlOei|&O+Qsu|f4~_r z0)TN4!>Sd##bXOwE?_Vd$eS`F`)vuZRI zr)Xb&Al^(C<0lI=k)Sp=$u{xAc$rGPUjnZ|Dxz!*o#v4v@vRtW3bI~`{M%YuAelDj70Q!gX};-YzxP(amAooT%Fwu)bn9 zdWsE03^~Uzy`&XiN8FEE9hK>ejgPQIBPqq17ks~O}T9t{1 zD!+a6Q%*aB#@*#vi}zk!Ap(7XU3bJW(u=C_pUOv@%WN3vZS#-VOm8jb4b9MJnxHHs&;xlA5)6 zNk&&98qsz85RA%*UQZF7m*)@Cpq-Fw*D%2CWNLV4Pme!{8Z@X-JQ^m>l%LOHz4}#( z+_uM)HE8iQDp{flAl&d$T}4H(N=CdUW{x-;-adkV(HFsMQ`m)E0C^3#S6eP|p!uzG zE&DgN@W3h*JxNF$QImW@$w-X}m@D8cYM2N?!Q^+xU(?b++#Z1)z!Xkr!wH2CVUrr3 zhYOFB@;Y5)4})brys~5L=~Epsu}ZO6r@{NXoA`O_O=-WQQ68Xm&R6QS=jB;|P{flI z05SzNNf_&k5h1k_Ul1?Xsj3v{b6{Z#`q0@rk0v1}jyb7gzAlX<6AP8}kup8xEZ`rX z$&c`D9-f*&1DtZcJtLF&vIN3hC?Ao${`&u2IB96CJ7U;Tu&1pVhBjPN$+&Z8k%^0$ zD@KsCgl#LuFwh5R_fNC3AZ@eA5%jri4l!iXV3M9mCOdcS+JKP!9>7t0ZHC|=Jv?*het0gzKv@hDD%A4%qKTq6(T50S&8A6&7%a?Q{oV*l<|Vlo(!Xq4Sa1Y05n-5^_cUT+F#Z)6}`z zpxXy~T<2k!Z{!Ht$8|%@lv|#L;_mEG$wRB;`=0UIaaOvdpc@@lx99U{X7mV6QGjSq zq-^kwTelSH0aaK?{AvGH!V7n)%yKq--0k;jCXU~e{Le>ITUuHk(4-ogeC*)aC{3bI9@8?F1%pggC9FaM z0NF@cs3;ix{TeGosGe>Knex9kiow`sFv@kk`azy~YT0W#8O5Oey3hFHSG)twN$SPZ zDirjzxBDP3ra6&;O-5-@NWqKH$T@Q1nC#RNcD@2cDO@Ows>6jkDTwqE4B&(obArmb z`B@^56WVIeL@OZV^s;^^WmRYF{X=NGNo#$`tT5y9oDxyBBaym5m*JYByeF6q^*KU^ zy1F{uzHo5!%+8w(& zkWK1UvwWE)63g%HJ+70`LXW3@tQlL|I-jPu2nZNRh_i zvx7ZzW_CkC;HlQPcOV;snT>QK z{rxv&NnLgZlC42<7--?}GG#GoW`W1HaPGW&Z@B=veG>0nsJnH%7&8oUt;)%K+C^s27rLR}*xqp~rC*GquVUrCbi=r?4_o5a&uB+q15HL~6y*K@L(Y9}{E zk+2S`Y`GvZ1T~?AE>;=Zwm`lhItj?wEcm8PDtnU_Al*jPu|Wf2H6k20ri1tG7+v1o z(X+XS5|8%v)#&nTd=>~gTR{TX577bZn_delsIYGiYoXKwtW675tGP5=5F4@)k9|&B zeF=#wm7K0>LNvTX$5$3|sbyn&%Tf4>93dS@nz3pv5!9{K*vt=|FZn|F+bORUfa*=M zuyeZg+jYSZ{Th5+!@#=g>N#M~_+q~l3zRT~F@=VlHiFaR3s_I}fM;)RZP3C>oMMXG zXrc>x8(PCInEtkH^QAhdOQ~GT#4&`^xc+N_HT4q zxiWi~oJTxDMBXo56tA>M`Q}EBu&4~L+aJBJ6K!mzGu&ucwWYA;`ct}IdTgVj!7Sc6d5$*^rY z0=~qqLkU-~Kbg$3Y?t|ZqEYBC=w9-Ih}lNW!%Q60E5igjbaOt@I?e1tc;ttLN^gQQ z%G$0ni^N(XA{T)`lr&9g;t$gK4bZeZT9>{OWUkqmKzfK`PB4Vpp|Y%BkSdIU`mQ=! zXR|4IY@~o(=J>N8OnA@O4>;AvQNt&eVU|*pY_J7F;H4OBh%v>R1GyS1@2^Eo>b(g8UR_^xpUHZ~v|^!$->c6b zl3jRVc%r2VaFz1a051KZ_eEiOprmtOV&f$7g*S`hl8Ulf;DC%rSAs(OD?jCm)=+Qo z3G!g+4}q|d0CodWK*dBAu|mhh32$0#Zb3$pEeW;ar-+^0hJ6_Bhg==CJLhs4`rnSHl9=PJa==M4?5g zSWd&q0rl$|ma=HLbJ^&uK1FE{ddQ^RSwv4dl6 zn{>{Kh66<2$cQg`=A(Qrp4?fqBtxZMIFx!hGu*S5J8(jLmw*0pjjsg$>V13s`-H%p zUKC}e*05)-K-BV3G$6rBT66LV*qeL|iT)a!1B2jf82JmjNCQ3A--+Um+%NEZrcL3qp*p_l!jQ3o(RMaejyaSt zo1Pt@Yz|!!m~Xmwd+u9qJ&Ndtw`TM_U?AAZdV-n^PbKa$YTUPGcOV>J;u1Td#p8~a5o}`o=IJsNFugh5H^&ztIxyyarZ?d10NKaXo~lYfHi%a z3;2XZaJ{%5jeL%X2_?$l%@o+P zn+8*ou=ze5AE*`xb>J*g=LMXHk4+;0hKjq;CHrT?8k{m)g3!i?u#m zxEp^X7?1|N4+EXt)hC3}$$<}%=J4xzq;jpm<#O>{uEBjO{#mWg64v_b?c*9aP1YO_ z(XOZAeAxU4KtaWW&Vk#k;mIuA20PI+i7_x8f9m!_=N5@V2oBN<)GvDGZ0i;F%88cJ zTw>$>jvL^A0$so+DRM+rB6nrg{^`hhm91S_P0TYQqo*YeZqA;PVfp;={Yk5;IhPG& z)!FccE##lp(GYgWqQ$9{PxNT)yeGaMj5a0k>w{OYv7^?^sF*ut~GX^~JjcFvZExZ5Nis z5jr<{W`YW{T)>u;UZ*KdV93KN2epU+qxco|uIuK73kRYb&nz{`} ze=r#>3gyezKgb?a@lSgHjEoCoW%OT8;F-Z5rzv%nqJg-F3kqD0K=(Zi`aXZa;f3)-y+vN# z&&V*>PM0_oiFd|zqDX_hL8z`>-qs3c(zNi=^5_P{T#vP)O0EHoxsED{c&aMs*Umr= z>LaNJF+%W2|02Htsiu{P(}thAW>07JtFxwK(5V7fgkp#^-FO-$ufSa*)x%SNDtKy! zOc0J$@}b5>USF8vod<@7TEc$N8G_bazLL&J0b$NP%3fu7xY^JZJbImG;iAw7-TWuC zP61>gSFjQ7BM1S<%Lj$I+T}MmsG;ul9K2)Zhd|(aebI?3Q>imDPMcw^HB3hN%XoZS zXLi54jg99W7l}t3Rb>@^NbVWE0o1dtr{^g$jt+J9$W~8q-2&p`;%bdX&mjZMC(Pn@ zg#TF@<_wN zLSI8MM#(1(y-47i)^z>6Yy@sLzL-ELAfRB3hb&#q@^~aG0TQ3DmC6LEk z*~1exGh2)MBoF*4;gj%030#7n`jqw&SXJH8zrs1aVlCwxDYK;QPsnBd0sOg%cA_!w5 zgDzsXb3p5OD;PHd<{;KD8$);st0J$x`x%cQeq;7%#qi@z+L$(x&?G<$qAvx(*?qzB z7c(11z! zKZ4$H;2G9@a=bb^B9~Rs+EZ9NzN-=LE&y$c$!y zVq9D%`c>8QTe@lho(js$Y@&nrf6AP>WmKm@^*}Tzg zj07*Daz#^EZ^-0Tf1SdtB=NkQpU%k6!aXeo@BnxtEjsI4)Og6Oz0e1uuN%U_d*LyG zwJPxb@z<^|0a6MF%|t7ou4+7oKzPAP7V2A`eL>;p$k)+i=0M)%5Wbcc&qK5V^|mHi zno<4t-0-){%l}CJ012@z^_UVehmaPJwv2hea~|~ONjsOkD5kO`86qmppc*if1|h)$ z*WMp8f!{~e_V9qs(W6^J*l@Js7#`r3wV}cu^pN*VdT4uPmE$3@-@Vl(Of`KG+SjC@ z=~LBMrmtFDX0}jVo8%cHSJI;gxSA<>{f9P;f2kn7aWyjnMHqsWc+b>f=W`+Sjq=L+ z^@Og~k5;DUp_17Ms~-Th@!FuFJ7mYFBVt(rVgZSR!`NELVi<@gVf7}=s68OA;E6j| zfjTG#H_QYb#wMfj89nJ&AVQHjc-Nu0EIfptbsWcnnmr!Y}_yTal>Gv|z3 zU(S$5O55j;UcMZHvhdiZCn|O&U{Q|>P)VV(XD;-B=g?qq#B`P~18dNAOnJ%BA2bsL zAV)!8RjkQ$bmDKhWHt#19t|Z&&CB7TRQBs%-C*-?CMU1d7qx;yKogYEh%jHDrH4g! zWb+I$$qL-Ed2nQUl?k}s7S$3@{n~Q;08AZ))+Sjhsn`F&p~fYO?Cu{&Wj)`-Ds(2n zzMiPR`=FG(9!c$`#ZhJA5R-9l3M%A5Eog^QgexTs0Fz`8IcH8lB|HH6$1LL$H1;%CF4c$_fFNJ!Svgc`KUJqx;|6)>fFaU zi;WHBK_yX#+3RbTGt+qgRCqAeldL6wXHYLDV%CUmKUOjbQ05vWNeIq8Ui~uNz?&_2 z$u6#BqpyBA9IcW^1>MRt8|UNJ&1;z4-AXG-(~Cr+A~~xi`2cd(v7P&}*k^ zZH0)C(r3rP3gdj-E4`XJDaysInV6LWKvb{`gD1+yD66IgL$Lto@xbU4EA8n-eMk|i z;*An;&yEWCLHsN-PsSZq30QQxI+BP!V`#&-<5ZXqy5fp2MdleP3vfUmG?n9s8`gvN zS1^ife*afyVz>@ndEh+EseJYW~SQ@ zy{hWSkpZoJDMx+G#u}s)I*AO`8JDpHhf2S)x{AP8sEGTdZ!3}hM z=PH25R6qXHnTTsVh{FlJ{QK!HMMnSr`@*=)zQR?Q_orW0bs4(O+9zo~6WoKKf(oK7 zI#?(WyQ%?vhj@JQS`lGpt}%iM)O?f@!ler|+oXwehNxwcZq>=wP^swaPI|iOqhK3VDsX{AY!yEmipSwA zMBql8B?3(M0%H^}3bZD;1zDDa$)fd$nIc5Xrv~v1a3u-_VQ2WE7ax0lI z(*PA3*CHE$8Rq|?q}}UtW!Yi9OFcl}2-QP$bGV&j+{0T_%O#$LwGhYRW>l1dWUY$P zb$48{L0CmqRX#17OD$B(=pMQ}Q>V#(3himYQ_XSl6xtU~qzYj@O*n!39HrVR1}z*f z^9Nm3p;&d3m?oH<+>DapM8NnoCviR9)NctN>aF84)%AJVPV81d^{Ai}Xu@8F;n2SO zaZkN(O!?ruUlD))*H>4;^$~}r`N%&5@o;OInhkACRxN_1)7W- zI+`%Efi39Ks-xrrHe{^Bn2oUKMF0w$(~0|+A|#Y90Yto`0*@l`mO_0cr%W_Yac+^kF0V(h(%mtV)uDKxQG3=YnXe%KmQsfB4q@C8_S~;RwPqABB3J3)u zZneu#3A!l2M}dTBVFd^ZGbrZqlM@Z+V_1n^JzKq&d-4It$D1ahpy)`LbI^%3sG6ec zU11ge?T`X{tXY-dpV7wtDxEdHf=2%W7=tK<$`{r^SR$MQZ9O@b44~{uQlfC+#TO%6ft@VSHB57By0A$P z9UIS$0_QQ>;;C2)-`NY|pkn8P=1cn;5ZNLWF<|yz5Yqk&#DVnRjRHgwpMA zxd6uT27&x&GBE*^wvA_Q$~s%qB24x)9LY)Tu( zDgX~Y{}@bSS*{0%-*D@iePNlX*nzp+Kx?pETaaA&zY7$|@*73Z^Spdn*Zp8Er6K^M zT{+OtZf|Sz275gZHG1A~KSE%l#DU!Or+taJ5_XDex(jCPcp_01U$AKWwd%&SI$sD& zC={vA^6kAJE>O`Pj4@Q`T;K}Cn~t=4=zWUtldO%1pyV+&LO6)Fus49+9ww$?bn%|B zS3n`CNw&EK<8)MYOMq+Xm3_((RBLiZo(|`!XHSJ0(Cwq~WwgU=mZ1=A)O_Fx*F8Qz z|I7*$ya%YYo1=!Lz^zBqF;%1-D|Q(NXY&2T2C8+6zedXy5C9A6`!}q|X_4*4+GtZ^ zWg*+#1GYNz`JB+uPztJ1=_sDZVp42Elk~j21lM{b(zpdy<3^NKDIg0r$s!^xp=!!W z!?bjyLNpNeWC=MWc?38sb$i0WvD@9}@6=arWSEPVw>^PN;!PQw;}M1UBQT1TRj)g4 z)CCHj^(HgkBRS#synlR?@ag1L-w2z{gbMparml4}Q|nyzr3kAUnW|bmz9@!pt#30h z7MB<+K`!(^(|BD_bz>tUzGIF)4_D+~1<8XbK3oX^tgUn0Ks4U2KJz4btVmfpp>*w7g{YzhL6{yRzE|EsUYZcl3pk>i?Gd?f$ct?4M1@ ziig4K_Q(Y@$~xJQf6oV_;9NzZ?(%gai~s*&TZ`=9PU92#(8 z9rI~XVWL{iS5Z6upV4+@qvczdvvrs{_AaukpYZshrHziJadrBG%rF1MO_Fs z8!_lWH`9m{yFYDN4=(A>y$Dt|EtK%jr_z^P<+@7*Jz^hG{=2rEB10BypoF~^3`#VIf5(`X$x2~%}L`*E@RyT0zt=dYg= z;~o}i!`t6vNC6;(PhA>9k;CMIupRh<6jl^n__i7v8b0bBr}5ekASQ2Q7YBMk#XLrA z5vt}h($Wf;nk7W<0q%44NW>Cc!L#~H(&RhmtoKwu+nO`rcNJepAi8@!?yQ zpUuq7Yy()=5K1gGwCqjimQMsTDF7{oOM`c#QIq&yK6R>){O%g6f-1mthw_P3+l1_k zQbh}RYdg9pJWP)Tzcc}{l41=_Fw+@(HHC;hN225C{fXsgQo>kvJ}789g}5`&Bw7}Z zm9yarcQp&b=5cwVip z-F)g&aEplB2VxI}3^A^Y7A-Pz_~c=F7qA>GGYDe@KuqUabmM0DT(eJC#%}$4wM*A* z0!}Oi+aosSqJeLDr%02_JMM-KV03Qh*c)SDw~OW^oKNOJSM~DVr^K8CB8W?ojzBqT zNaGa-?#e(gCniba`8!V6#!8&aH{ZQ;3mRz&&kZyc>XjEQ@`Z-WnqC7_HQ>z;Fwfe^ z8p`7uDb<$O)=Q&v)u%qdTYNzM(tAQ|<9xtSjH0s4@xHaa+~tzC@k=&FjUc6W|y8 z6k%~ArgrE>?UFJ*Ub+F|aToONKs?Jwoit6{8bMz$%-nt%s$&{NciK{8EjaKQLs&(g z(k%iEVj{ZHocU=EQP5#N+mOz2wA zE$C;8e!|cJpL(-I3mgpFmw-^@PzOESK*Qx%A&aDy>D_HgJ!*Mi#0Tn;Y$MHziY$}x z^B;XfG?-}72o1eXG*10^F*rD??00>_K8eS>!#eol#heKfWzM@rgR$a+Hq=JH9O)Ak zDA&)Swhff}wv1w`#MR>y4w?>inu%7Ul|IDH3WkXoUBgX5hIr>F7!3nRLAjx!{C?Oc z<=>*C-Ae{k?W=u#csD+NlBNG>UX~u7+U8ZYoU2By|MF zXOAv?s?mD=c1(8LVy^t~JcG7lD8sGh1mRZ2ht%R|1pqaE9qyEI=W zNbe5BYGaPiLpQ3=a}f3tA?BsM66A^-+2i`ed+t!uX+$Kt^)3$z8TksA z_XqA({U;L1UeH~%sqgts8k{jv?KXAAkXuz4kE`XBZ`ov7nf(@ z#*PnxH&$?G!RPqY8y4a!eT}NQVbm@YlR4C7fWD{5vZ6v%gc~7(XKs@PJG}}xl#%2C zE>}u3kywK_25L74g>+W_?tON_d%4t1fYO3G3Q|V4x7rcVoue7_$p>&jWPnh|464DI zKtOS8CcyLS$n88ELeu|(kqSWHp-$%dGe9c`6$|5-y0trt3n)pO5JRVlLwL1 zY{G!VMrhR+x_`0FkgBewr~_=NL4%<8qPPe1gIs>$p_hSUY+A+=D^)4F>nNigm@41V zmHU9m%Q)st?S-F-tlaV(Ak7S%1t*LOjI5+)wg}{3F_O-+BWQf*qpGB&^b{foJFgMC za~;G}^71z7)|6O5#AbCx3)D*I(15?>G&HD0mZiP=xAe`MH|Gaq!Z`T$0GTZ*r0i6l zI<*%U^Ds(JmDYqCB|&T=nM6RpoVR8b6$}u~prbY?i&rmAFf=)TI3%qyu*!y!8C*6I z;h*v$?Ck1b7RMIBQAa^ZNdQDIb9&fFohWPa~Tu1J59oLNpo8PM0S_t8v z1~45BlhgkATsj9WL_6#@jl@oh&(`5a`C~DxiN2b*S$wKNsvFu*2bnG zv;V0t{M44wdT#O~zE?3aW0i!L6c~b4h1KY;buRJ4fiHI_LbiE;mKxES%R_WYhogbI z70{5os-j0%DRc(dS6h-Lrd4eVAAdERTdzACtgx?LRnY=2^=?pKn!cw3-SkY;%AxmL^Ksu{gNv!m-pu4GvEIb^ML=k>@!fvI>sK409-M0l9xm zvgtJfT8MOyw)$H&cF0U+yaq~R)Fhli$F*@Ui>>|XuhhJs8|r8c_MomloA4THX&4-; zf%4N2^oyuPD5{=jV@1G9#<+erACr@$P1BF-M(qOmnLSeMKwTCw1A#;6C)~iv_AXtaLn{t4sjsbtY$8CDPWAQEqg%jh|)#j9OT%ZLru_4*Ns76G?tVBxDE7j;0;t-Skbc`#6dxmDO zII=BktaXUm2dSaFVCeqS)r&AB%!n%y{*)_?jAK7-TIQk8fp>sSspQL)6&vnQnEP4X#43_CO1G$`sgeE?hwM{!Jf_+XfG0*@X2W6 zL*k^DEJ$Sw7K1Q5lVZIeCu_(*k6#*hmN=ydVP+d@4enGFphl$xhZV4;=8aGK`@e&_ z>7rQ@VVU?F=e>!cq4wh7=`A&21(@A54A*u9(y{@opeW7MGchrtn@12J)>3s@^=pci zQ^1?qQe)N$J2wL(5-a&+ABcv76^N{iUF?uOVo8As72qnXIaOxlW-3wW7R4h$0q5At zBK;=>-UJS=Lc{e4k!Kwu^h4x?=Rt2W~D??NxXuml2Nq1PlmaH;7UZ`<$Ow! zF|KNL5B$?A03;2_CeI?dkIZi5P&XY_HNmK$-oL{s5df~)B4A)P z;rt<{&84t}U>VBoiD6nZMB52+Jv*1coxS}p$re+;URQF}0cr|)amaV^vg%&05Zw{^ zXrVSoLG`NLZJJHl*{9`3c_M3lNz){ijWDua&KReYJ9IB zhMbbRgNlAhHxf8bq7vN_lBd#GV&`xs60~Ef9vR0Srae zdqVVn@8&MYck%d}Dwj*xpANtD-p~YzFVSC4?RETfl2Q(2BZ3Cc?Sbdpj}$EdR$ttr zKz5YGAZqpld7-*F-IF|2MyEm}&>(Zf$jRBc@$8~FZ}4uve=f_xjEyrdP?R*{?gkTt z&}`sn4r|Z?O|`pn^|6e?!a_L@XZ zjV|8@KV&cvZf0O$a47ZXnkry4l<6<;WWyr%BBQ=lGr$y?)pOh10fjWRRI(jRwE=RJ zUxUB&#qrd}t=r)s${@K3LipIi!T3g&u$9$-d!U;+0*oS$PAHIfUn2tz1svt;ZvD|m z;ZF%CkF}Cmwem&dxlV>n6@EoTY;3H-SokyI!NC5^#XYteLqvOSacb={AOUj;J0y(! z9IkbIZ#UWm=HqD1B72<)083n$N;tG(ah#b4@ejMuX=qY1JmswqgPYvpP=dx^;p;Ox zIcO`Aqh@t7oTW%z%SRDP_7d_zo|yzTP-%>=T(V%n<}O|vMLCEQwl=9T?pY$GpH74a zNI!pA=NX%pZb$GM!bcGB!YCbGUtfQmm{T zVDgxT`>b74$QMJE4qYV{=kaV~vS!1OVEOEtLpoV(;{g zhgX%yVY_p%_S4SDy6hoJSEmMt5Kzz+b0DZl3IkK8mSRZL5d@D%Gkfz&0l6IkE`X8p z>+w=hv=L*_XEtTb0LHAu%;S4XpJZko<(_5o8XAS>+Zin9nn6${TAVY@J#ZRF~ubs zKM(9b_Qz+X5V4}BnT~EM%ix?gCKfHxzuPY85rq9pL>gkm##+T*LR7z=EI#D->U-eD zscX*>Enm0lg-dDgs3Z*(`n&m*13~C=CJME_fbjE>Q^^6y7;EXgha?ez;8b$Kcn=I* zIY15sCwLlaZ)dTQv}7rWEpibV&&3rja{uDp3$`{ZC|ok%6@T}}5Qb}u9XF{bctUCk zX~+*S^L_-4L2MIVJ9M}HH5tMlQ1pK=dxr%(ZWdQY5n?G=(r{T`Q8Ni z@S6yTw(s3LSNiC^A0R(HALvsA;v1aA@8;jczC(h4DrGT%IL5BAiI9Np@Iiw8O+8B_ z*n`VzDQ5=H(LfL`EZ$1{wB}1Wt0^)%lse&5vN(#Ss(56O!C8tow4x1cQEj3&!qQ~h zD(_)d_##CNl2A%`xzB4DoFS{(ZnbvC&h=si+gu%+x9iZV#gtQ*{S^QVlxJJj2lx~ zP*Bp7-Wf;j$c;EM4_;nxaB^{)WwYpej%Lu(s#mI({eZxapw~<0m^xDoKG1tZ8-`qPt;;< zAynHyExcm(CStUAK5nOu3w6s|Au~CL*kd^;7?%exo>&KY;RZNgUqBeK(XoLbZ+%NT z39(jVD%}V(NwH=9i0O6yTP^_JnLXk`;8tuZ=q^M)Kg7djcBsiK*tj-k@CRRSs>-jrAAC#sjCB4fEW>KnE#H*uE~upiuO~hS&&la@)8#} z1+bZg*0*%z339m~mSBMF+M7-T#Cfj{&UojA8K>MAj|T^@O(dtlt8+Cwiuy9rSd&gb zD77vx?Y8gVf1S*_#pCmBFbb8+nQDYAl2R0E#S}JW_Y>*Vx3;z^TTlh;ZhFR7jBbJCn-afj7`+-538+|CX|yaAy=-|zB1c;o{)c$vjmnEjtCuoNqCM}u?U>5A z_|N7|D58+I!S?4$E@)MxSBFT6rPeD=%nFi$ zvgffbED^SWM+I#+L>3XK-BMop&d0Dx!FWaR^Vvwn)rr6M(muluH!fiPqz{+?>8gBMp*ZO0L*G^M5y$oTFI z+9I};MqS-;UD14$4$??Pf(*6=)ntg3MZpB+zfR7q1Ds`$C=;io#CW@zlyp3pWr~#U z&^1gMI*)5^3My~F@~5l5yDucrmB?-#NC1(NXAfacggGHhEw*QnN@A-2^WpEiVLLJt4<9c8mCJ5H6`;{4h(UWC$2KJb&->_^^AtUNq*3g0T{v@=bSD; zU^0Q`K+@;Yj}Z^R-d-TA{0;t^GJZ-=0oqz6Gwl!!?A*OuF-e6$4WGKt%c>yJ(FS*7 zDZpy1pmM%@6<@$V$*zh6`VaEy{;6I!1z z5@DoLa&jrNvV-TkQeuU~f3cy^L)^GQC;%jm62pw5DXLeFzCMjuo{*2V(o$LpiE+n2 zK9kBJDuUnKm4)sGx3G?4*ob0Y2^fG(jR;E5-m#-1@b69tzasN?XrMtU_$q zxbXLhACXOyU`sEN?EpeWfbMEyShy2F37v6LEm;Lh9}QUV)V=h#46YF zn%$+rNVy8!#o1`y%S0YeU*7=mziN0lmSph~zM;4-s(n5iUG@10=PsJCx?Utu4F3e> z6Vy&SVbh50R|@nmHhSy^!9SOln1;Z@B%-X92WKTquH!hjUaM2%sx6?s{m#WJ$+P4j zj=MG}^X@{OI&Mouu&X(Km%DrC&hkIg0g-6>=Qbd(xOwm1y-H~_ zOl7#s8A|~KgA4h1BO=NDs5g>YF6pUe=@4-TXNi~sF{$k0H!)f?I$QIAU^L*V{3)>~ zb2b>sCDsJROeQ6J>k z7UZNhARLXr+#ZyNk(cBUTMFCTiz0YFA;jUHeY&2A6$tFB$mRgdK7xaU#=~#b@Nk3( zQ!SJ(Y>jZ6@JC#sn>-q*)8_S!CY1;#eEN9~3S>dLEC4g58Lr$H*u>@SV8Ud%5NGT9 z_3Kn%?1B*{z>^~Q2Ec5E?z`8o2TSU^&Ple1eHW*AlaHaENHL`!GGV~`SckRl?54?Z zlqRe~@6jQ|cSY_JOV5~kVO-=vNPhHqcg;#1x=gYI4d}@8p+g3w{Q6*0MIx@y1|s7T zTLrNdl30D1ZemE#^HS1En?(t%%Kiu6cUtSj+_>=#;elEnL+7YqL?a9f{3!$gXYBe5 z0UIPQYFVb#>Cjh<8}`=l*I1PAppVlV%7%D^EGU-LNsx~`l4$Z&fj~lmivLWQmhH&+ zQIALf%(8)M8sL_dVz>y^B}(~gQxJV~a;g6ktyO@BaKDDp>^ytQn%8R) z_C%qAD~{C#XD~!#6$J<+s!GahL>Hho>t%HyrpO(xiC;|cAvLBG59F=dtGH;0MVl_J z8t*%dyn(n_>#8>%|%?2irM&C1OAE_dAGLi6(5}RD{)Ljj*Hv zG{cC_TC9~9&0D1lUJ7)Cu<;0lg(ilhCVK(Uh;x@Jws55o51fZ9azAXq4{l`>Qu;%L zWRle*=Gve^tre#docKIYbDAG`SM4sfw+@R{zhzi6x;A!$@ z1&H3rCIFC|yYFbihND1$DC}Gh`46goUa&OX25tbA_r&lX6$}sxB!&&h;*=zS;+Buv z=WHS^Aa7g(v^p_Ng17r{%41B5jwx{ekks_LpTvoW@pdyY8A6PO;WA613gnM}V&^D| zt=^OcsMQ^RPKSv=?NXrhJ8L}A=+JMD*_gd0Ea7rOpr z7(RIiI)-pRM0_`gJSesFF93s+@KWeWk%=Xk{O=cV5Ei0HN6-GhuRusMp`f(jSgMV< zo`moapiGD?JoSpsDVOCyZk>gZ2u!MD(dT$;Uuue{y`WnRT^~@%5p2YoryCJWWsigrY^Wx)fS~?USLLb|` z{qEm?euSin@4mLfY=)DEQqe|50F30v;D+%+Ir^FGyP$6+yab;YNr~kD^bwLI{SXI( z(=v1H9g<2n{-6H0AzTl67mo-(HN7s$W^e|hTxwhb8~(N78nWoPfj)Im9JK?8Ecu`x zVf*2S9##>@p-nJZcC|dhNy(3oFqZK{o_5fe@Tuom1Z|y`^**U?@Q3m6Vf*?_gEmM2TnEz=k%a@)-uk<$d4CC;Yqe#25nUIS3od| zv$&b0t|96*2cPPHUSx#O0V0ZSU}ih~bJUCEaY^sLOGuf^iw};eq33v$EU=YRks8|$ zbW#nUE;?qW3IOrK_Y;$ditk7+Hoyznn75St2{(MXO+a*^Qh>$B65hzuKH2v38T$v}8I_|?=f?kYR%cf~^DgV)+Y`3d zG^s^0BKrF|PW?f|d{JAL<^4HX5T!;FINH{U{mU9)ntt`=moySUuZ}WC{SS`NiS11v z`=tolqU!1)@@AbX9%Oos24g4jkp`WGtxr#SyAFRp24kDjhMu>dOZ++e$?Das^)g7< zX|r;$3)>(JQPmD^$<^i2<^P9YLft1+UQ$kFAt!qQZFJkrJbb8?(-1kD0 zwV^Z8-uu#}}0v#U#`dk7)Z{+jM|B<)IioK--(^f(gFPUn7slZ2~!`~8Opt12pX z)pR3bh)8;Q;x(|BJaBSNI;h}9zB|hP?T`Nb=YM#bYq*uxsp#uOF|lhwxA33~(pjoR z=yB43)qVI>tCc_YRO%?Nm)S>OYygjY34T)wI8iS$Pdc}2Ql8%~{oX@;eA~JRwL4;G zx42sf?=u@0d=Y-#0EpUofc^K3iW@R;L8B!5!5EDH@Sf68#=*ggNaEwy`C|1mllZ5L z0X~_7@l<+LO`o$DeB=)wQvD-m>-K2`2AAAfC{-rFF6>bM{4}lsRH}MGBD9V&U zbR`z)y+5JxRQgBOZ>=oB{!#S~YS##Y+=Q`BF`61MatjC^ryhZY)fs(-i%3}MwuOX1 z0v5I-^07pp*Z-{biSKbdi7TyNM}o7&%et^*;R&TPKmYtQ5$7lj4(=kkZ$c>aQvT<8N_xEl^mg{tmim;goF;BKKT*n&SI%hc+HirUU>Er19lu(@Kc zU*QM7e92tpsEe8sansxN>1aYY8Gwqwy|Q9L{h(z<<6a;?x}$xF)qz%^vq+^>$OW!) zdwsB(PXa2Ko4XR;YRAY2@rLeqQ5T&yz5V6y{<${vXibtbk-jvT-5O}6+JojA3fTek zZyOPcf_I!n(PII%NmEr@W{unT`-%uipaM&Z*So(8pWpV~SlByG_XT|`4d}~aqDmd* zlhP@1ND;n^Y0&`3j8p24s)_L%lx37q90?r;7k|-9p(~LMslVCD>b-w#?cxxu5|z%b&Q36y(FGP-ppuefwT=vhfc`#K)_Wcn+xy&d}rZZ(q0AAOx*k0 zBT?L?W%k>sr`+uS-gAoodMcgr;aJlQHP^hKzNk#%H1aS*v`;54D2a zPp7P*(Mpd)-@g;dYMPPFVQ^!psXm|j zmp)jpe_#Ljq5SPA@Sp?;@Qd}qF8<&Epo=tDb5PibKwrJ<`{Co4x_%dT6w1E$-fy1q z%xPjtQQILY*M=@zU`paj)KEZt_z#~e>EqA!C>I0dsc1n08F|>!>zY>EXvRlF zpg&ytN9#UnW*@$4!IK35+|I!aPIDy9Y0inHD;5DYJxbg@c(%!l{=;kD5ehLG9WWGO zkx21ZyBf69eIivw;&2jrSgY~gz8%|R^U)3crU6t*e}sceprAx!c+;GPRCyAk2yp*1 zNaa|v4!~gib+7WH*YcsrM(tWJOn5#zKxUE%3QCH#x*`XJ6C|mJmd0}=lmv^|1K>^K z<1g074g2e=&p-eC5b6okWRj2for;XmTfpbFJ{{Gkv;w2)XzY!dt9w)4FoWvpHKk$~!hA3X8jAAEmJ-ue65 z|1QrvtMS37|6QIBmgL Date: Wed, 28 Jan 2026 09:49:51 -0600 Subject: [PATCH 572/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index b94a68b76b5..ae87d159395 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -1,4 +1,4 @@ -# Megatron-DeepSpeed, optimizers, hyperparameters and CPT +# Megatron-DeepSpeed, optimizers, hyperparameters Single command to test and run Megatron-DeepSpeed: @@ -55,7 +55,7 @@ For the learning rate, we implemented the learning rate finder routine [here](ht ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 --lr-finder ``` -This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR blows up and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. +This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR start increasing and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. ![lr_finder](./assets/lb_optimizers/lr_finder_example.png) ### Maximal Update Parametrization/Complete Parametrization -We have MuP and CompleteP incorporated in AuroraGPT. Will add details about activating them soon +We have MuP and CompleteP incorporated in AuroraGPT in the `lb-optimizers` branch. From 06f86a9a6cb5aa630e86f6caca2849b73c92e223 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 09:53:03 -0600 Subject: [PATCH 573/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index ae87d159395..367d735e754 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -44,12 +44,14 @@ To add a custom scheduler, you have to modify the following files: ## Hyperparameter tuning #### Init variance -Weight initialization is key to training LLMs,and to avoid spikes in losses. Here, we initialize the wights following this [paper](https://arxiv.org/pdf/2312.16903). The default variance value at initialization is 0.02. To add custom variances, one can use `--init-method-std, `--adjust-word-embedding-init`, and `--word-embedding-init-std`. For our runs, we do +Weight initialization is key to training LLMs,and to avoid spikes in losses. Here, we initialize the weights following this [paper](https://arxiv.org/pdf/2312.16903). The default variance value at initialization is 0.02. To add custom variances, one can use `--init-method-std, `--adjust-word-embedding-init`, and `--word-embedding-init-std`. For our runs, we do ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_TOKENS=$((22*10**9)) GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 ``` -where `d=hidden size`. - +where `d=hidden size`. In general, the initialization should be +```bash +--init-method-std sqrt{ 2 / (5 * d) } --adjust-word-embedding-init --word-embedding-init-std sqrt{ 2 / 5 } +``` ### Learning rate For the learning rate, we implemented the learning rate finder routine [here](https://sgugger.github.io/how-do-you-find-a-good-learning-rate.html) and [here](https://arxiv.org/pdf/1506.01186). This is activated with the `--lr-finder` and run for `TRAIN_ITERS` steps. For example, for a 1000 steps: ```bash From c2d9f8b5dd01b7ff5beec1f955bc1057aecf3c49 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Wed, 28 Jan 2026 12:52:40 -0600 Subject: [PATCH 574/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 138 +++++++++++++++++- 1 file changed, 137 insertions(+), 1 deletion(-) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index 367d735e754..a3d9b066c39 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -57,7 +57,143 @@ For the learning rate, we implemented the learning rate finder routine [here](ht ```bash DATA_FILE_LIST=./ALCF/data-lists/aurora/books.txt TRAIN_ITERS=1000 GRAD_ACC_STEPS=16 LR_DECAY_STYLE=constant LR=0.0002 LR_WARMUP_FRACTION=0.01 OPT=muon bash train_alcf.sh --lr_constant_plus_cooldown --init-method-std ${sqrt(2/5d)} --adjust-word-embedding-init --word-embedding-init-std 0.632 --lr-finder ``` -This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR start increasing and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. +This approach allows to find the largest LR one can train with without the model divergence. Training with large LR is crucial for infinite schedulers as well for large batch training. We increase the LR following a power law at each step and monitor the LR curve in particular the decaying phase and the blow up phase. To tune the LR, you identify the point where the LR start increasing and divide that LR value by 10. You can also pick the LR corresponding to the steepest descent phase. The learning rates are stored in the output folder (the one set in **SAVE=**). ![lr_finder](./assets/lb_optimizers/lr_finder_example.png) +After running the LR-finder routine, you can modify the code below to plot it and find suitable LR candidates: +```bash + +# ---------- Helper: find ALL local minima vs log10(LR) ---------- +def find_all_minima_lrs(learning_rates, losses, smooth_frac=0.03, min_log_sep=0.0): + """ + Return ALL learning rates where the loss switches from decreasing->increasing + (local minima of loss with respect to log10(LR)). + + Parameters + ---------- + learning_rates : array-like + losses : array-like + smooth_frac : fraction of length for moving-average smoothing (reduce noise). + Try 0.02–0.05 if needed. + min_log_sep : minimum spacing in log10(LR) between reported minima. + Set 0.0 to return all raw minima (no de-dupe). + + Returns + ------- + list of LRs (floats), sorted ascending (left→right). + """ + lr = np.asarray(learning_rates, float) + y = np.asarray(losses, float) + + # ensure LR is increasing + order = np.argsort(lr) + lr, y = lr[order], y[order] + + # light smoothing to avoid jitter creating fake minima + w = max(5, int(len(y) * smooth_frac) | 1) # odd window + if w >= len(y): + w = (len(y) - 1) | 1 + y_s = np.convolve(y, np.ones(w) / w, mode="same") + + # first derivative w.r.t. log10(LR) to match your log-x axis + x = np.log10(lr) + dy = np.gradient(y_s, x) + + # adaptive epsilon so tiny flat noise doesn't cause false flips + eps = 1e-12 + 0.02 * np.median(np.abs(dy)) + + # minima where slope crosses 0 from negative -> positive + neg = dy[:-1] < -eps + pos = dy[1:] > eps + idx = np.where(neg & pos)[0] + + # sub-sample minimum position by interpolating where dy == 0 + xmins = [] + for i in idx: + x0, x1 = x[i], x[i+1] + y0, y1 = dy[i], dy[i+1] + denom = (y1 - y0) + xz = x0 if denom == 0 else x0 - y0 * (x1 - x0) / denom + if np.isfinite(xz): + xmins.append(xz) + + # optional de-dup: enforce separation in log space (default 0.0 → keep all) + xmins = sorted(xmins) + if min_log_sep > 0.0 and len(xmins) > 1: + kept = [xmins[0]] + for xz in xmins[1:]: + if all(abs(xz - p) >= min_log_sep for p in kept): + kept.append(xz) + xmins = kept + + return [10**px for px in xmins] + +# ---------- Safe savefig wrapper (uses your savefig if present) ---------- +def savefig_safe(fig, name): + try: + if 'savefig' in globals() and callable(globals()['savefig']): + globals()['savefig'](fig, name, None) + else: + fig.savefig(f"{name}.png", dpi=150, bbox_inches="tight") + except Exception: + fig.savefig(f"{name}.png", dpi=150, bbox_inches="tight") + +# Small utility to optionally place a dot at a given x if it lies inside data range +def scatter_if_in_range(ax, x_target, x_all, y_all, color, size=45, zorder=6): + x_min, x_max = np.min(x_all), np.max(x_all) + if (x_target >= x_min) and (x_target <= x_max): + y_val = np.interp(np.log10(x_target), np.log10(x_all), y_all) + ax.scatter(x_target, y_val, color=color, s=size, zorder=zorder) + +# ===================================================== +# =============== LAMB (optional) =================== +# ===================================================== +try: + fig, ax = plt.subplots(figsize=(18, 12)) + i = 0 + for opt in optims_lamb: + fname = opt + '/lr_finder_agpt_olmo/lr_finder/lr_finder_data.npz' + data = np.load(fname) + learning_rates = data['learning_rates'][:430] + losses = data['losses'][:430] + + ax.plot(learning_rates, losses, linewidth=4, label=opt, color=f'C{i+1}') + + # absolute minimum (diamond) + min_idx = np.argmin(losses) + min_lr = learning_rates[min_idx] + ax.scatter(min_lr, losses[min_idx], s=100, color=f'C{i+1}', marker='D', zorder=5) + + # ALL minima where loss turns up + minima_lrs = find_all_minima_lrs(learning_rates, losses, + smooth_frac=0.03, min_log_sep=0.0) # set >0 to de-dupe + suggested_lrs = [lr / 10.0 for lr in minima_lrs] + + print(f"{opt} minima LRs (all): ", [f"{lr:.3e}" for lr in minima_lrs]) + print(f"{opt} suggested LRs (min/10):", [f"{lr:.3e}" for lr in suggested_lrs]) + + # draw minima (solid dashed) + suggested (dotted) + for lrj in minima_lrs: + #ax.axvline(x=lrj, color=f'C{i+1}', linestyle='--', alpha=0.6, linewidth=1.8) + scatter_if_in_range(ax, lrj, learning_rates, losses, color=f'C{i+1}') + for lrj in suggested_lrs: + ax.axvline(x=lrj, color=f'C{i+1}', linestyle=':', alpha=0.5, linewidth=1.6,label=f"{opt} suggested LR: {lrj:.2e}") + scatter_if_in_range(ax, lrj, learning_rates, losses, color=f'C{i+1}', size=35, zorder=6) + ax.axvline(x=0.0002, color=f'C{i+1}', linestyle='-', alpha=0.5, linewidth=1.6,label=f"{opt} current LR: 0.0002") + ax.set_xscale('log') + ax.set_xlabel('Learning Rate') + ax.set_ylabel('Loss') + ax.set_title('Learning Rate Finder Comparison (LAMB)', fontsize=32) + ax.legend(fontsize=14) + i += 1 + + savefig_safe(fig, 'figs_agpt/lr_finder_agpt_olmo_lamb') + plt.show() +except NameError: + # optims_lamb not defined; skip + pass + + + +``` ### Maximal Update Parametrization/Complete Parametrization We have MuP and CompleteP incorporated in AuroraGPT in the `lb-optimizers` branch. From 2808ff4c116521974a56492f782970028871c7fd Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Thu, 29 Jan 2026 13:58:27 -0600 Subject: [PATCH 575/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 1 + 1 file changed, 1 insertion(+) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index a3d9b066c39..3f0c300bfa0 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -1,4 +1,5 @@ # Megatron-DeepSpeed, optimizers, hyperparameters +**Important** For large batch training with infinite schedulers, It is crucial to tune the learning rate as these schedulers benefit from larger learning rate values. See below for the `lr_finder` routine implemented in MDS to do so. Single command to test and run Megatron-DeepSpeed: From 3d0fd17c83142efd186d81313092ea2cf67f0092 Mon Sep 17 00:00:00 2001 From: Marieme Ngom <50598247+mngom2@users.noreply.github.com> Date: Fri, 30 Jan 2026 15:57:53 -0600 Subject: [PATCH 576/590] Update large_batch_optimizers_settings.md --- ALCF/notes/large_batch_optimizers_settings.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ALCF/notes/large_batch_optimizers_settings.md b/ALCF/notes/large_batch_optimizers_settings.md index 3f0c300bfa0..c6fc7c409dc 100644 --- a/ALCF/notes/large_batch_optimizers_settings.md +++ b/ALCF/notes/large_batch_optimizers_settings.md @@ -1,5 +1,5 @@ # Megatron-DeepSpeed, optimizers, hyperparameters -**Important** For large batch training with infinite schedulers, It is crucial to tune the learning rate as these schedulers benefit from larger learning rate values. See below for the `lr_finder` routine implemented in MDS to do so. +`**Important** For large batch training with infinite schedulers, It is crucial to tune the learning rate as these schedulers benefit from larger learning rate values. See below for the `lr_finder` routine implemented in MDS to do so.` Single command to test and run Megatron-DeepSpeed: From 22b5b24f81f9319000da33fb81517e5a3faff2f2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:01:32 -0600 Subject: [PATCH 577/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index 687ed2e53e6..e211ccbcf66 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -96,7 +96,7 @@ setup() { # # 4. Use these (^) to build our launch command # ezpz_setup || exit ########################################################################## - install_dependencies + # install_dependencies # Set command line arguments to pass to `"${EXEC}"` setParams || exit # Create `deepspeed_config.json` from runtime params from ^ @@ -361,7 +361,7 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - LAUNCHER="ezpz-launch $(which python3) ${EXEC}" + LAUNCHER="ezpz launch $(which python3) ${EXEC}" # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job # echo "ezpz_launch: $(which ezpz_launch)" # export -f ezpz_launch From 88dbdc4d48dccf7eaa5ff2f482f1c9f5761806ec Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:01:55 -0600 Subject: [PATCH 578/590] chore: Replace `print` with `logger.info` in `megatron/timers.py` --- megatron/timers.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/megatron/timers.py b/megatron/timers.py index 89b2da3795e..c432b4c2af0 100644 --- a/megatron/timers.py +++ b/megatron/timers.py @@ -5,12 +5,15 @@ from abc import ABC from abc import abstractmethod import time +import ezpz import torch from deepspeed.accelerator import get_accelerator from tensorboard.summary import Writer from packaging import version +logger = ezpz.get_logger(__name__) + try: import wandb except Exception: @@ -309,7 +312,7 @@ def log(self, names, rank=None, normalizer=1.0, reset=True, barrier=False): if rank is None: rank = torch.distributed.get_world_size() - 1 if rank == torch.distributed.get_rank() and output_string is not None: - print(output_string, flush=True) + logger.info(output_string) #, flush=True) def write( self, From fc0dad5f48b26efdf80255577a34cc3b937a3caf Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:02:26 -0600 Subject: [PATCH 579/590] chore: Update `megatron/training.py` --- megatron/training.py | 104 +++++++++++++++++++++---------------------- 1 file changed, 52 insertions(+), 52 deletions(-) diff --git a/megatron/training.py b/megatron/training.py index 385ed9064aa..89c17e0d80d 100644 --- a/megatron/training.py +++ b/megatron/training.py @@ -9,6 +9,7 @@ _TRAIN_START_TIME = time.time() import ezpz +import ezpz.dist from collections import OrderedDict from datetime import datetime @@ -25,7 +26,6 @@ from deepspeed.runtime.data_pipeline.data_routing.helper import ( convert_to_random_ltd, ) -import ezpz as ez import torch import torch.distributed as tdist from torch.nn.parallel.distributed import DistributedDataParallel as torchDDP @@ -82,10 +82,10 @@ # from deepspeed import comm as dist -RANK: int = ez.get_rank() -WORLD_SIZE: int = ez.get_world_size() -# DEVICE_TYPE: str = ez.get_torch_device() -DEVICE_TYPE: str = ez.dist.get_torch_device_type() +RANK: int = ezpz.get_rank() +WORLD_SIZE: int = ezpz.get_world_size() +# DEVICE_TYPE: str = ezpz.get_torch_device() +DEVICE_TYPE: str = ezpz.get_torch_device_type() DEVICE: torch.device = torch.device(DEVICE_TYPE) log: logging.Logger = logging.getLogger(__name__) @@ -208,9 +208,9 @@ def pretrain( extra_trace_path = "" os.makedirs(args.trace_dir, exist_ok=True) PerfTrace.initialize_log( - f"{args.trace_dir}/trace-{ez.get_rank()}-of-{ez.get_world_size()}.pfw", + f"{args.trace_dir}/trace-{ezpz.get_rank()}-of-{ezpz.get_world_size()}.pfw", f"{args.data_cache_path}:{extra_trace_path}:{args.data_path}:{args.save}:{args.load}", - process_id=ez.get_rank(), + process_id=ezpz.get_rank(), ) timers = get_timers() assert args is not None @@ -409,7 +409,7 @@ def setup_teacher_model(args, model_provider): @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def get_model( model_provider_func, model_type=ModelType.encoder_or_decoder, wrap_with_ddp=True ): @@ -425,9 +425,9 @@ def get_model( mpu.get_pipeline_model_parallel_world_size() > 1 and args.virtual_pipeline_model_parallel_size is not None ): - assert ( - model_type != ModelType.encoder_and_decoder - ), "Interleaved schedule not supported for model with both encoder and decoder" + assert model_type != ModelType.encoder_and_decoder, ( + "Interleaved schedule not supported for model with both encoder and decoder" + ) model = [] for i in range(args.virtual_pipeline_model_parallel_size): mpu.set_virtual_pipeline_model_parallel_rank(i) @@ -446,9 +446,9 @@ def get_model( add_decoder = True if model_type == ModelType.encoder_and_decoder: if mpu.get_pipeline_model_parallel_world_size() > 1: - assert ( - args.pipeline_model_parallel_split_rank is not None - ), "Split rank needs to be specified for model with both encoder and decoder" + assert args.pipeline_model_parallel_split_rank is not None, ( + "Split rank needs to be specified for model with both encoder and decoder" + ) rank = mpu.get_pipeline_model_parallel_rank() split_rank = args.pipeline_model_parallel_split_rank world_size = mpu.get_pipeline_model_parallel_world_size() @@ -474,9 +474,9 @@ def get_model( # Disallow training and inference with Transformer Engine # for non-GPT models args.allow_transformer_engine = all([type(m) == GPTModel for m in model]) - assert ( - args.allow_transformer_engine or args.transformer_impl == "local" - ), "Transformer Engine is only approved for GPT models" + assert args.allow_transformer_engine or args.transformer_impl == "local", ( + "Transformer Engine is only approved for GPT models" + ) # Set tensor model parallel attributes if not set. # Only parameters that are already tensor model parallel have these @@ -558,7 +558,7 @@ def get_model( @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def get_optimizer_param_scheduler(optimizer): """Build the learning rate scheduler.""" args = get_args() @@ -580,7 +580,9 @@ def get_optimizer_param_scheduler(optimizer): else: lr_constant_steps = args.lr_constant_iters * args.global_batch_size if args.lr_constant_plus_cooldown: - lr_constant_plus_cooldown_steps = args.lr_constant_plus_cooldown_frac * lr_decay_steps + lr_constant_plus_cooldown_steps = ( + args.lr_constant_plus_cooldown_frac * lr_decay_steps + ) else: lr_constant_plus_cooldown_steps = 0 lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps @@ -603,7 +605,9 @@ def get_optimizer_param_scheduler(optimizer): else: lr_constant_steps = args.lr_constant_samples if args.lr_constant_plus_cooldown: - lr_constant_plus_cooldown_steps = args.lr_constant_plus_cooldown_frac * lr_decay_steps + lr_constant_plus_cooldown_steps = ( + args.lr_constant_plus_cooldown_frac * lr_decay_steps + ) else: lr_constant_plus_cooldown_steps = 0 lr_cooldown_steps = args.lr_cooldown_fraction * lr_decay_steps @@ -650,9 +654,9 @@ def load_model_weights_only(model_provider_func): model=model[0], config=args.deepspeed_config_dict ) - assert not isinstance( - model, deepspeed.PipelineEngine - ), "Weight loading only mode is not supported in pipeline parallelism yet." + assert not isinstance(model, deepspeed.PipelineEngine), ( + "Weight loading only mode is not supported in pipeline parallelism yet." + ) model = [model] print_datetime("before load checkpoint") if args.load is not None: @@ -664,7 +668,7 @@ def load_model_weights_only(model_provider_func): @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def setup_model_and_optimizer( model_provider_func, model_type, @@ -712,7 +716,6 @@ def setup_model_and_optimizer( if teacher: optimizer = None else: - optimizer = get_megatron_optimizer( model, no_wd_decay_cond, scale_lr_cond, lr_mult ) @@ -780,7 +783,7 @@ def setup_model_and_optimizer( ) tds0 = time.time() if os.environ.get("PYINSTRUMENT_PROFILER", None): - profiler = ez.profile.get_context_manager(rank=RANK, outdir=args.save) + profiler = ezpz.profile.get_context_manager(rank=RANK, outdir=args.save) else: profiler = Profile("deepspeed.initialize") log.info("Calling 'deepspeed.initialize'...") @@ -1002,7 +1005,7 @@ def train_step( @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def save_checkpoint_and_time(iteration, model, optimizer, opt_param_scheduler): timers = get_timers() assert timers is not None @@ -1121,7 +1124,7 @@ def train( # convenience handles for DP sync dp_group = mpu.get_data_parallel_group() - dev = torch.device(DEVICE_TYPE) + dev = torch.device(DEVICE_TYPE) # Main LR finder loop for i in range(finder_iters): @@ -1182,37 +1185,34 @@ def train( if smoothed_loss < best_loss or batch_num == 1: best_loss = smoothed_loss - # explode = (batch_num > 1 and smoothed_loss > 4 * best_loss) - + # explode = (batch_num > 1 and smoothed_loss > 4 * best_loss) - # if explode: - # if mpu.get_data_parallel_rank() == 0: - # log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") - # break + # if explode: + # if mpu.get_data_parallel_rank() == 0: + # log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") + # break ## Record the best loss (use the same global smoothed_loss) - #if smoothed_loss < best_loss or batch_num == 1: + # if smoothed_loss < best_loss or batch_num == 1: # best_loss = smoothed_loss # --- GLOBALIZE the "loss exploding" decision (any rank => all ranks) --- - #explode_local = (batch_num > 1 and smoothed_loss > 4 * best_loss) - #print(f"Rank {mpu.get_data_parallel_rank()}: explode_local={explode_local}") - #if mpu.get_data_parallel_rank() == 0: - + # explode_local = (batch_num > 1 and smoothed_loss > 4 * best_loss) + # print(f"Rank {mpu.get_data_parallel_rank()}: explode_local={explode_local}") + # if mpu.get_data_parallel_rank() == 0: + # print(f"Iter {i}: batch_num={batch_num}, smoothed_loss={smoothed_loss:.8f}, best_loss={best_loss:.8f}, ratio={smoothed_loss/best_loss:.2f}") - #if tdist.is_available() and tdist.is_initialized(): + # if tdist.is_available() and tdist.is_initialized(): # _exp = torch.tensor([1 if explode_local else 0], device=dev) # tdist.all_reduce(_exp, op=tdist.ReduceOp.MAX, group=dp_group) # explode = bool(_exp.item()) - #else: + # else: # explode = explode_local - - - #if explode: + # if explode: # if mpu.get_data_parallel_rank() == 0: # log.info(f"Loss exploding at lr={curr_lr:.8f}, stopping LR finder") - # Keep everyone in lockstep before breaking + # Keep everyone in lockstep before breaking # if tdist.is_available() and tdist.is_initialized(): # tdist.barrier(group=dp_group) # break @@ -1224,7 +1224,7 @@ def train( # Print progress if (i + 1) % args.log_interval == 0 and mpu.get_data_parallel_rank() == 0: log.info( - f"LR Finder: iteration {i+1}/{finder_iters}, " + f"LR Finder: iteration {i + 1}/{finder_iters}, " f"lr: {curr_lr:.8f}, loss: {smoothed_loss:.4f}" ) @@ -1702,7 +1702,7 @@ def cyclic_iter(iter): @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): """Build pretraining datasets.""" @@ -1731,7 +1731,7 @@ def build_train_valid_test_datasets(build_train_valid_test_datasets_provider): @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider): """Build pretraining data loaders.""" args = get_args() @@ -1741,9 +1741,9 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider log.info("> building train, validation, and test datasets ...") # Backward compatibility, assume fixed batch size. if args.iteration > 0 and args.consumed_train_samples == 0: - assert ( - args.train_samples is None - ), "only backward compatiblity support for iteration-based training" + assert args.train_samples is None, ( + "only backward compatiblity support for iteration-based training" + ) args.consumed_train_samples = args.iteration * args.global_batch_size if args.iteration > 0 and args.consumed_valid_samples == 0: if args.train_samples is None: @@ -1802,7 +1802,7 @@ def build_train_valid_test_data_loaders(build_train_valid_test_datasets_provider @dlp.log -@ez.dist.timeitlogit(rank=RANK) +@ezpz.dist.timeitlogit(rank=RANK) def build_train_valid_test_data_iterators(build_train_valid_test_datasets_provider): """Build pretraining data iterators.""" From 08e872815f03edb1122a5cf63c0418b58cc408e1 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:02:39 -0600 Subject: [PATCH 580/590] chore: Update `pretrain_gpt_alcf.py` --- pretrain_gpt_alcf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pretrain_gpt_alcf.py b/pretrain_gpt_alcf.py index a2d974418f7..674038d7ae4 100644 --- a/pretrain_gpt_alcf.py +++ b/pretrain_gpt_alcf.py @@ -623,7 +623,7 @@ def main(): model = main() dist.log_summary() - if wandb.run is not None: + if wandb is not None and wandb.run is not None: print(f"wandb.run.name: {wandb.run.name}") print(f"wandb.run.url: {wandb.run.url}") wandb.finish() From 6027ea68528afe62646fd0db42d77ee1ff6b78f7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:02:56 -0600 Subject: [PATCH 581/590] chore: Update `train_alcf.sh` --- train_alcf.sh | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/train_alcf.sh b/train_alcf.sh index 9bc98c4c3ac..17624893566 100644 --- a/train_alcf.sh +++ b/train_alcf.sh @@ -10,14 +10,14 @@ GIT_BRANCH=$(git branch --show-current) && export GIT_BRANCH source <(curl -L https://bit.ly/ezpz-utils) ezpz_setup_env -if command -v "ezpz-test"; then - log_message INFO "${GREEN}✓${RESET} ezpz is already installed." - # printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" -else - log_message WARNING "${RED}✗${RESET} ezpz is not installed." - log_message INFO "Installing ezpz..." - python3 -m pip install "git+https://github.com/saforem2/ezpz" || exit -fi +# if command -v "ezpz-test"; then +# log_message INFO "${GREEN}✓${RESET} ezpz is already installed." +# # printf "[!! %s] ezpz is already installed.\n" "$(printGreen "INFO")" +# else +# log_message WARNING "${RED}✗${RESET} ezpz is not installed." +# log_message INFO "Installing ezpz..." +# python3 -m pip install "git+https://github.com/saforem2/ezpz" || exit +# fi ##################################### @@ -62,6 +62,6 @@ train_aGPT() { # Kill any existing MPI processes -ezpz_kill_mpi || exit +# ezpz_kill_mpi || exit train_aGPT "$@" From 3ea917b68fc7063b09f14e369146636c53e24f84 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Mon, 9 Feb 2026 08:58:10 -0600 Subject: [PATCH 582/590] chore: Update `tools/cooldown_generator/make_cooldown_cmds.py` --- .../cooldown_generator/make_cooldown_cmds.py | 158 +++++++++++++++--- 1 file changed, 132 insertions(+), 26 deletions(-) diff --git a/tools/cooldown_generator/make_cooldown_cmds.py b/tools/cooldown_generator/make_cooldown_cmds.py index e7516a1e72e..46e876115b2 100755 --- a/tools/cooldown_generator/make_cooldown_cmds.py +++ b/tools/cooldown_generator/make_cooldown_cmds.py @@ -1,33 +1,111 @@ #!/usr/bin/env python3 +from typing import Any, Optional import argparse from pathlib import Path from textwrap import dedent + +#!/bin/bash --login +# PBS -q prod +# PBS -A AuroraGPT +# PBS -l walltime=06:00:00 +# PBS -l filesystems=flare:home +# PBS -l select=256 +# PBS -j oe + + +_FILE_PATH = Path(os.path.abspath(__file__)).parent +_MEGATRON_PATH = _FILE_PATH.parent.parent + + +def get_header_template( + queue: str = "prod", + project: str = "AuroraGPT", + walltime: str = "06:00:00", + filesystems: str = "flare:home", + nodes: int = 256, +) -> str: + return "\n".join( + [ + f"#PBS -q {queue}", + f"#PBS -A {project}", + f"#PBS -l walltime={walltime}", + f"#PBS -l filesystems={filesystems}", + f"#PBS -l select={nodes}", + "#PBS -j oe", + "", + "cd ${PBS_O_WORKDIR}", + ] + ) + + def fmt_float(x: float) -> str: return f"{x:.8f}".rstrip("0").rstrip(".") + def build_command( load_path: str, data_file_list: str, train_script: str, - train_iters: int, - lr_cooldown_frac: float, - grad_acc_steps: int, - opt: str, - min_lr: float, - override_ckpt_opt_param: bool, - extra_args: str, + train_iters: Optional[int] = None, + lr_cooldown_frac: float = 0.05, + grad_acc_steps: Optional[int] = None, + opt: Optional[str] = None, + min_lr: Optional[float] = None, + override_ckpt_opt_param: bool = True, + extra_args: Optional[str] = None, + model_arch: str = "AuroraGPT-2B", + train_tokens: Optional[int] = None, + global_batch_size: Optional[int] = None, + sequence_length: Optional[int] = None, + lr: Optional[float] = None, + micro_batch: Optional[int] = None, + use_activation_checkpointing: Optional[bool] = None, + tokenizer_type: str = "HFTokenizer", + tokenizer_model: str = "google/gemma-7b", + zero_stage: Optional[str | int] = None, ) -> str: + act_ckpt_val = "1" if use_activation_checkpointing else "0" + override_ckpt_val = "1" if override_ckpt_opt_param else "0" env_lines = [ + f"MODEL_ARCH={model_arch}", "LR_DECAY_STYLE=constant", - f"OPT={opt}", - "OVERRIDE_CKPT_OPT_PARAM=1" if override_ckpt_opt_param else "", - f"TRAIN_ITERS={train_iters}", - f"GRAD_ACC_STEPS={grad_acc_steps}", f"LOAD={load_path}", f"DATA_FILE_LIST={data_file_list}", + f"USE_ACTIVATION_CHECKPOINTING={act_ckpt_val}", + f"OVERRIDE_CKPT_OPT_PARAM={override_ckpt_val}", + f"TOKENIZER_TYPE={tokenizer_type}", + f"TOKENIZER_MODEL={tokenizer_model}", ] - env_block = " \\\n".join([l for l in env_lines if l]) + if opt is not None: + env_lines.append(f"OPT={opt}") + if grad_acc_steps is not None: + env_lines.append(f"GRAD_ACC_STEPS={grad_acc_steps}") + if lr is not None: + env_lines.append(f"LR={lr}") + if micro_batch is not None: + env_lines.append(f"MICRO_BATCH={micro_batch}") + if zero_stage is not None: + env_lines.append(f"ZERO_STAGE={zero_stage}") + + # ---- TRAIN {ITERS, TOKENS} setup --------------------------------------- + if train_iters is None and train_tokens is None: + raise ValueError("One of {train_iters, train_tokens} required!") + if train_iters is not None: + assert train_tokens is None, ( + f"Only one of {train_tokens, train_iters} should be specified." + ) + if train_tokens is not None: + assert train_iters is None, ( + f"Only one of {train_tokens, train_iters} should be specified." + ) + assert global_batch_size is not None and sequence_length is not None + train_iters = train_tokens * global_batch_size * sequence_length + + assert train_iters is not None + env_lines.append(f"TRAIN_ITERS={train_iters}") + + env_block = " \\\n".join([line for line in env_lines if line]) extra_line = "" if extra_args: @@ -43,15 +121,21 @@ def build_command( """).strip() return cmd + def parse_pairs(pairs_args): records = [] next_id = 1 for item in pairs_args: parts = item.split(":") if len(parts) == 2: - S = int(parts[0]); R = int(parts[1]); cid = next_id; next_id += 1 + S = int(parts[0]) + R = int(parts[1]) + cid = next_id + next_id += 1 elif len(parts) == 3: - cid = int(parts[0]); S = int(parts[1]); R = int(parts[2]) + cid = int(parts[0]) + S = int(parts[1]) + R = int(parts[2]) else: raise SystemExit(f"Bad --pairs entry: {item}") if S <= 0 or R <= 0: @@ -59,11 +143,12 @@ def parse_pairs(pairs_args): records.append({"id": cid, "S": S, "R": R}) return records + def main(): p = argparse.ArgumentParser( description="Emit Megatron-DeepSpeed cooldown commands so LR cooldown starts at resume.\n" - "Provide checkpoint iteration(s) S and cooldown step(s) R.\n" - "For each pair, sets TRAIN_ITERS T=S+R and lr_constant_plus_cooldown_frac f=S/T." + "Provide checkpoint iteration(s) S and cooldown step(s) R.\n" + "For each pair, sets TRAIN_ITERS T=S+R and lr_constant_plus_cooldown_frac f=S/T." ) p.add_argument("--load", required=True) p.add_argument("--data-file-list", required=True) @@ -75,32 +160,53 @@ def main(): p.add_argument("--extra-args", default="") p.add_argument("--emit-sh", action="store_true", default=None) p.add_argument("--split-by-id", action="store_true") - p.add_argument("--include-header", default=None, type=str) - + p.add_argument("--queue", default="prod", type=str) + p.add_argument("--project", default="AuroraGPT", type=str) + p.add_argument("--walltime", default="06:00:00", type=str) + p.add_argument("--filesystems", default="flare:home", type=str) + p.add_argument("--nodes", default=256, type=int) + p.add_argument("--include-header", type=str, default=None) + # p.add_argument() + # p.add_argument("--include-header", default="") + # p.add_argument("--include-header", default=None, type=) p.add_argument("--checkpoint-iters", "-S", type=int, nargs="+") p.add_argument("--cooldown-steps", "-R", type=int, nargs="+") p.add_argument("--checkpoint-ids", type=int, nargs="+") p.add_argument("--pairs", type=str, nargs="*") args = p.parse_args() + if args.include_header is None: + args.include_header = get_header_template( + queue=args.queue, + project=args.project, + walltime=args.walltime, + filesystems=args.filesystems, + nodes=args.nodes, + ) override_flag = not args.no_override_ckpt_opt if args.pairs: records = parse_pairs(args.pairs) else: if not args.checkpoint_iters or not args.cooldown_steps: - raise SystemExit("Provide either --pairs OR both --checkpoint-iters and --cooldown-steps.") + raise SystemExit( + "Provide either --pairs OR both --checkpoint-iters and --cooldown-steps." + ) ids = args.checkpoint_ids or list(range(1, len(args.checkpoint_iters) + 1)) if len(ids) != len(args.checkpoint_iters): - raise SystemExit("--checkpoint-ids must match length of --checkpoint-iters.") - records = [{"id": cid, "S": int(S), "R": int(R)} - for cid, S in zip(ids, args.checkpoint_iters) - for R in args.cooldown_steps] + raise SystemExit( + "--checkpoint-ids must match length of --checkpoint-iters." + ) + records = [ + {"id": cid, "S": int(S), "R": int(R)} + for cid, S in zip(ids, args.checkpoint_iters) + for R in args.cooldown_steps + ] lines = [] # header = "# Auto-generated cooldown commands\nset -euo pipefail\n\n" if args.include_header: - if (hfp :=Path(args.include_header)).is_file(): + if (hfp := Path(args.include_header)).is_file(): with hfp.open("r") as f: lines.extend(f.readlines()) else: @@ -139,8 +245,8 @@ def main(): # cid, S, R = rec["id"], rec["S"], rec["R"] # # args.emit_sh.write_text("\n".join(lines)) - # print(f"# Wrote script to: {args.emit_sh}") + # print(f"# Wrote script to: {args.emit_sh}") + if __name__ == "__main__": main() - From b1bbe013186603c6239f04cc4e10e330f94d95a5 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 10 Feb 2026 07:23:25 -0600 Subject: [PATCH 583/590] chore: Update `ALCF/helpers.sh` --- ALCF/helpers.sh | 28 ++++------------------------ 1 file changed, 4 insertions(+), 24 deletions(-) diff --git a/ALCF/helpers.sh b/ALCF/helpers.sh index e211ccbcf66..8fa0343e1ed 100644 --- a/ALCF/helpers.sh +++ b/ALCF/helpers.sh @@ -361,26 +361,7 @@ setupLauncher() { make_ds_hostfile || exit export LAUNCHER="deepspeed --hostfile $hfds --launcher MPICH ${EXEC}" else - LAUNCHER="ezpz launch $(which python3) ${EXEC}" - # source <(curl -L https://bit.ly/ezpz-utils) && ezpz_setup_job - # echo "ezpz_launch: $(which ezpz_launch)" - # export -f ezpz_launch - # export LAUNCHER="ezpz_launch) $(which python3) -Wignore ${EXEC}" - # if [[ -n "${DIST_LAUNCH}" ]]; then - # mn=$(ezpz_get_machine_name) - # if [[ "${mn}" == "aurora" || "${mn}" == "sunspot" ]]; then - # LAUNCHER="${DIST_LAUNCH} --pmi=pmix --genvall $(which python3) -Wignore ${EXEC}" - # elif [[ "${mn}" == "sophia" ]]; then - # LAUNCHER="${DIST_LAUNCH} $(which python3) -Wignore ${EXEC}" - # else - # LAUNCHER="${DIST_LAUNCH} --genvall $(which python3) -Wignore ${EXEC}" - # fi - # export LAUNCHER="${LAUNCHER}" - # else - # echo "[setupLauncher][INFO]: Saving environment to: .env-${PBS_JOBID}" - # printenv | tee ".env-${PBS_JOBID}" - # echo "[setupLauncher][ERROR]: DIST_LAUNCH not found in environment !!" - # fi + LAUNCHER="ezpz launch $(which python3) ${EXEC}" fi printf "Launching with: %s\n" "$(printRed "${dist_launcher}")" printf " %s" "$(printMagenta "${LAUNCHER}")" @@ -502,7 +483,6 @@ set_ccl_vars_on_aurora() { export CCL_BCAST=double_tree } - get_model_arch_AuroraGPT_2B() { # AuroraGPT-2B export HEADS=16 @@ -730,9 +710,9 @@ setParams() { "llama3-3B" | "llama-3B") get_model_arch_llama3_3B_customNlayers ;; - "2B" | "AuroraGPT-2B" | "AuroraGPT_2B" | "Aurora-GPT-2B" | "AuroraGPT2B" | "Aurora_GPT_2B" | "aurora-gpt-2b" | "aurora_gpt_2b") - get_model_arch_AuroraGPT_2B - ;; + "2B" | "AuroraGPT-2B" | "AuroraGPT_2B" | "Aurora-GPT-2B" | "AuroraGPT2B" | "Aurora_GPT_2B" | "aurora-gpt-2b" | "aurora_gpt_2b") + get_model_arch_AuroraGPT_2B + ;; "7B" | "AuroraGPT-7B" | "aurora-gpt-7b" | "llama-3.1-7B" | "llama-3.1-7b" | "llama-3.2-7B" | "llama-3.2-7b") get_model_arch_7B ;; From 5f11d7bc1bfb34af6ddf5025bc7631ea3fcfc20d Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 10 Feb 2026 12:16:43 -0600 Subject: [PATCH 584/590] chore: Update `tools/cooldown_generator/make_cooldown_cmds.py` --- .../cooldown_generator/make_cooldown_cmds.py | 129 +++++++++++++++--- 1 file changed, 109 insertions(+), 20 deletions(-) diff --git a/tools/cooldown_generator/make_cooldown_cmds.py b/tools/cooldown_generator/make_cooldown_cmds.py index 46e876115b2..75cb031883b 100755 --- a/tools/cooldown_generator/make_cooldown_cmds.py +++ b/tools/cooldown_generator/make_cooldown_cmds.py @@ -1,21 +1,16 @@ #!/usr/bin/env python3 + +import os from typing import Any, Optional import argparse from pathlib import Path from textwrap import dedent +import ezpz +# _FILE_PATH = Path(os.path.abspath(__file__)).parent +# _MEGATRON_PATH = _FILE_PATH.parent.parent -#!/bin/bash --login -# PBS -q prod -# PBS -A AuroraGPT -# PBS -l walltime=06:00:00 -# PBS -l filesystems=flare:home -# PBS -l select=256 -# PBS -j oe - - -_FILE_PATH = Path(os.path.abspath(__file__)).parent -_MEGATRON_PATH = _FILE_PATH.parent.parent +logger = ezpz.get_logger(__name__) def get_header_template( @@ -35,6 +30,7 @@ def get_header_template( "#PBS -j oe", "", "cd ${PBS_O_WORKDIR}", + "\n", ] ) @@ -43,6 +39,70 @@ def fmt_float(x: float) -> str: return f"{x:.8f}".rstrip("0").rstrip(".") +def get_total_iters_from_cooldown_percent( + checkpoint_iter: Optional[int] = None, + cooldown_percent: Optional[float] = None, + cooldown_steps: Optional[int] = None, + train_iters: Optional[int] = None, +) -> dict: + if checkpoint_iter is None and train_iters is None: + raise ValueError("Expected one of {checkpoint_iter, train_iters}") + if cooldown_percent is None and cooldown_steps is None: + raise ValueError("Expected one of {cooldown_percent, cooldown_iters}") + if checkpoint_iter is None: + assert train_iters is not None + if cooldown_percent is None: + assert cooldown_steps is not None + checkpoint_iter = train_iters - cooldown_steps + cooldown_percent = (train_iters - cooldown_steps) / train_iters + elif cooldown_steps is None: + assert cooldown_percent is not None + cooldown_steps = int(train_iters * cooldown_percent) + checkpoint_iter = train_iters - cooldown_steps + else: + raise ValueError( + "Expected one of {cooldown_percent, cooldown_iters} to be specified" + ) + assert ( + checkpoint_iter is not None + and cooldown_percent is not None + and cooldown_steps is not None + and train_iters is not None + ) + return { + "checkpoint_iter": checkpoint_iter, + "cooldown_percent": cooldown_percent, + "cooldown_iters": cooldown_steps, + "train_iters": train_iters, + } + if train_iters is None: + assert checkpoint_iter is not None + if cooldown_percent is None: + assert cooldown_steps is not None + train_iters = checkpoint_iter + cooldown_steps + cooldown_percent = (train_iters - cooldown_steps / train_iters) + elif cooldown_steps is None: + assert cooldown_percent is not None + cooldown_steps = int(cooldown_percent * checkpoint_iter) + train_iters = checkpoint_iter + cooldown_steps + else: + raise ValueError( + "Expected one of {cooldown_percent, cooldown_iters}" + ) + assert ( + checkpoint_iter is not None + and cooldown_percent is not None + and cooldown_steps is not None + and train_iters is not None + ) + return { + "checkpoint_iter": checkpoint_iter, + "cooldown_percent": cooldown_percent, + "cooldown_iters": cooldown_steps, + "train_iters": train_iters, + } + + def build_command( load_path: str, data_file_list: str, @@ -166,14 +226,21 @@ def main(): p.add_argument("--filesystems", default="flare:home", type=str) p.add_argument("--nodes", default=256, type=int) p.add_argument("--include-header", type=str, default=None) - # p.add_argument() - # p.add_argument("--include-header", default="") - # p.add_argument("--include-header", default=None, type=) + p.add_argument("--train-iters", type=int, required=False) + p.add_argument("--train-tokens", type=int, required=False) + p.add_argument("--global-batch", type=int, required=False) + p.add_argument("--sequence-length", type=int, required=False) + p.add_argument("--lr", type=float, required=False) + p.add_argument("--micro-batch", type=int, required=False) + p.add_argument("--use-activation-checkpointing", type=int, required=False) + p.add_argument("--tokenizer-type", type=str, required=False) + p.add_argument("--tokenizer-model", type=str, required=False) + p.add_argument("--zero-stage", type=int, required=False) p.add_argument("--checkpoint-iters", "-S", type=int, nargs="+") p.add_argument("--cooldown-steps", "-R", type=int, nargs="+") + p.add_argument("--cooldown-percent", type=float, required=False) p.add_argument("--checkpoint-ids", type=int, nargs="+") p.add_argument("--pairs", type=str, nargs="*") - args = p.parse_args() if args.include_header is None: args.include_header = get_header_template( @@ -188,10 +255,6 @@ def main(): if args.pairs: records = parse_pairs(args.pairs) else: - if not args.checkpoint_iters or not args.cooldown_steps: - raise SystemExit( - "Provide either --pairs OR both --checkpoint-iters and --cooldown-steps." - ) ids = args.checkpoint_ids or list(range(1, len(args.checkpoint_iters) + 1)) if len(ids) != len(args.checkpoint_iters): raise SystemExit( @@ -230,9 +293,35 @@ def main(): override_ckpt_opt_param=override_flag, extra_args=args.extra_args.strip(), ) + latest_fp = Path(args.load).parent.joinpath("latest") + latest_ckpt_iter = Path(args.load).parent.joinpath("latest_checkpointed_iteration.txt") + ckpt_parent = Path(args.load).parent + if latest_fp.is_file(): + logger.info(f"Found 'latest' in {ckpt_parent}!") + with latest_fp.open("r") as f: + _latest = f.read().rstrip("\n").lstrip("global_step") + assert int(_latest) == int(S), f"{_latest=} != {S=}" + else: + logger.info(f"No 'latest' in {ckpt_parent}!") + logger.info(f"Writing global_step{S} to {latest_fp}") + with latest_fp.open("w") as f: + f.write(f"global_step{S}") + + if latest_ckpt_iter.is_file(): + logger.info(f"Found 'latest_checkpointed_iteration.txt' in {ckpt_parent}!") + with latest_ckpt_iter.open("r") as f: + _latest = f.read().rstrip("\n") + assert int(_latest) == int(S), f"{_latest=} != {S=}" + else: + logger.info(f"No 'latest_checkpointed_iteration.txt' in {ckpt_parent}!") + logger.info(f"Writing {S} to {latest_ckpt_iter}") + with latest_ckpt_iter.open("w") as f: + f.write(f"{S}") + block = f"{tag}\n{cmd}\n" if args.emit_sh: - outfile = f"cooldown_{cid}.sh" + outfile = f"cooldown_id{cid}_s{S}_r{R}_t{T}.sh" + logger.info(f"Writing:\n{block}\nto:\n{outfile}") with open(outfile, "w") as f: f.write("".join(lines)) f.writelines(block + "\n") From b5617bac7b4e2303e738ef7db0307915da905941 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 10 Feb 2026 12:25:00 -0600 Subject: [PATCH 585/590] chore: Fix https://github.com/argonne-lcf/Megatron-DeepSpeed/pull/96#discussion_r2512411230 --- tools/cooldown_generator/build_checkpoints_from_tokens.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/cooldown_generator/build_checkpoints_from_tokens.py b/tools/cooldown_generator/build_checkpoints_from_tokens.py index bc6e8c4611b..0bcd040099d 100755 --- a/tools/cooldown_generator/build_checkpoints_from_tokens.py +++ b/tools/cooldown_generator/build_checkpoints_from_tokens.py @@ -31,7 +31,7 @@ def main(): runs_mod = {k: int(round(v / r) * r) for k, v in runs.items()} # % of the FINAL (ttokens-1) step count, then rounded rollback - cooldown_iters = int(c * runs[ttokens - 1]) + cooldown_iters = int(c * runs_mod[ttokens - 1]) runs_rollback = {k: int(round((v - cooldown_iters) / r) * r) for k, v in runs_mod.items()} with open(args.out, "w") as f: From 65d0a416136235c02ea74acfafa95646c6894a70 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 24 Feb 2026 09:40:38 -0600 Subject: [PATCH 586/590] feat: Add `train_aGPT_2B_sophiag_stage3.sh` --- train_aGPT_2B_sophiag_stage3.sh | 46 +++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) create mode 100644 train_aGPT_2B_sophiag_stage3.sh diff --git a/train_aGPT_2B_sophiag_stage3.sh b/train_aGPT_2B_sophiag_stage3.sh new file mode 100644 index 00000000000..48735f2d1d8 --- /dev/null +++ b/train_aGPT_2B_sophiag_stage3.sh @@ -0,0 +1,46 @@ +#!/bin/bash --login +#PBS -q prod +#PBS -A AuroraGPT +#PBS -j oe +#PBS -l walltime=06:00:00,filesystems=flare:home +#PBS -l select=256 + +setup_env() { + cd "${PBS_O_WORKDIR}" || { + echo "Failed to change directory to ${PBS_O_WORKDIR}" + exit 1 + } + # shellcheck disable=SC1090 + source <(curl -L https://bit.ly/ezpz-utils) + ezpz_setup_env + log_message INFO "Using: $(which python3)" +} + +# 7,064,155,541,716 [7.064 T][@ end of stage2] +# + 706,610,881,663 [0.706 T] +# ------------------------------------------------- +# 7,770,766,423,379 [7.770 T][@ end of stage4] + +train_model() { + MODEL_ARCH=AuroraGPT-2B \ + TRAIN_TOKENS=7770766423379 \ + OPT=sophiag \ + LR=2.17e-5 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/nvidia-math1-code2.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" \ + "$@" +} + +main() { + setup_env + train_model "$@" +} + +main "$@" From 81b1d1a623c56a86ef7c81f8a638ef8b5b41e2e7 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 24 Feb 2026 09:41:14 -0600 Subject: [PATCH 587/590] feat: Add `ALCF/data-lists/aurora/nvidia-math1-code2.txt` --- ALCF/data-lists/aurora/nvidia-math1-code2.txt | 210 ++++++++++++++++++ 1 file changed, 210 insertions(+) create mode 100644 ALCF/data-lists/aurora/nvidia-math1-code2.txt diff --git a/ALCF/data-lists/aurora/nvidia-math1-code2.txt b/ALCF/data-lists/aurora/nvidia-math1-code2.txt new file mode 100644 index 00000000000..9218c5e6674 --- /dev/null +++ b/ALCF/data-lists/aurora/nvidia-math1-code2.txt @@ -0,0 +1,210 @@ +0.0066041769 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0006_of_0014_text_document 4plus +0.0064896352 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0001_of_0014_text_document 4plus +0.0056249649 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0014_of_0014_text_document 4plus +0.0065620092 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0011_of_0014_text_document 4plus +0.0051356758 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0013_of_0014_text_document 4plus +0.0050103722 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0007_of_0014_text_document 4plus +0.0052755781 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0008_of_0014_text_document 4plus +0.0050491088 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0005_of_0014_text_document 4plus +0.0052196623 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0003_of_0014_text_document 4plus +0.0049355830 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0009_of_0014_text_document 4plus +0.0050814598 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0010_of_0014_text_document 4plus +0.0051993360 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0004_of_0014_text_document 4plus +0.0052450016 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0012_of_0014_text_document 4plus +0.0050743459 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus/fused_0002_of_0014_text_document 4plus +0.0059846329 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0004_of_0018_text_document 4plus_MIND +0.0059845899 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0009_of_0018_text_document 4plus_MIND +0.0059808878 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0016_of_0018_text_document 4plus_MIND +0.0059890326 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0010_of_0018_text_document 4plus_MIND +0.0059734562 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0003_of_0018_text_document 4plus_MIND +0.0059810124 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0014_of_0018_text_document 4plus_MIND +0.0059867130 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0001_of_0018_text_document 4plus_MIND +0.0059810021 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0005_of_0018_text_document 4plus_MIND +0.0059782283 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0008_of_0018_text_document 4plus_MIND +0.0059782940 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0017_of_0018_text_document 4plus_MIND +0.0059710269 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0013_of_0018_text_document 4plus_MIND +0.0059844574 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0002_of_0018_text_document 4plus_MIND +0.0059633344 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0015_of_0018_text_document 4plus_MIND +0.0059780891 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0007_of_0018_text_document 4plus_MIND +0.0059854627 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0011_of_0018_text_document 4plus_MIND +0.0053505435 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0018_of_0018_text_document 4plus_MIND +0.0059814337 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0012_of_0018_text_document 4plus_MIND +0.0059837015 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/4plus_MIND/fused_0006_of_0018_text_document 4plus_MIND +0.0041796606 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0007_of_0028_text_document 3 +0.0041769267 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0025_of_0028_text_document 3 +0.0041584894 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0022_of_0028_text_document 3 +0.0041582932 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0027_of_0028_text_document 3 +0.0041745608 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0014_of_0028_text_document 3 +0.0041765056 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0023_of_0028_text_document 3 +0.0041771952 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0026_of_0028_text_document 3 +0.0041790022 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0008_of_0028_text_document 3 +0.0041780952 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0002_of_0028_text_document 3 +0.0041744548 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0019_of_0028_text_document 3 +0.0041766825 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0015_of_0028_text_document 3 +0.0041795456 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0020_of_0028_text_document 3 +0.0041702190 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0004_of_0028_text_document 3 +0.0041877268 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0016_of_0028_text_document 3 +0.0041684592 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0003_of_0028_text_document 3 +0.0041732037 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0021_of_0028_text_document 3 +0.0047718597 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0028_of_0028_text_document 3 +0.0041760609 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0001_of_0028_text_document 3 +0.0041803800 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0018_of_0028_text_document 3 +0.0041766392 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0012_of_0028_text_document 3 +0.0041878393 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0006_of_0028_text_document 3 +0.0041838998 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0024_of_0028_text_document 3 +0.0041798020 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0011_of_0028_text_document 3 +0.0041927431 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0009_of_0028_text_document 3 +0.0041697497 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0013_of_0028_text_document 3 +0.0041929834 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0017_of_0028_text_document 3 +0.0041780669 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0010_of_0028_text_document 3 +0.0041583608 /flare/AuroraGPT/datasets/NVIDIA/NVIDIA-CC-MATH-V1-fused-tok/3/fused_0005_of_0028_text_document 3 +0.0038871792 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0021_of_0150_content_document Synthetic-Code +0.0056804901 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0041_of_0150_content_document Synthetic-Code +0.0038909922 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0017_of_0150_content_document Synthetic-Code +0.0056810745 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0034_of_0150_content_document Synthetic-Code +0.0056795634 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0074_of_0150_content_document Synthetic-Code +0.0056790712 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0046_of_0150_content_document Synthetic-Code +0.0056805038 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0086_of_0150_content_document Synthetic-Code +0.0056793563 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0057_of_0150_content_document Synthetic-Code +0.0038903737 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0014_of_0150_content_document Synthetic-Code +0.0051452715 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0148_of_0150_content_document Synthetic-Code +0.0056803707 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0096_of_0150_content_document Synthetic-Code +0.0038861114 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0022_of_0150_content_document Synthetic-Code +0.0056814237 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0066_of_0150_content_document Synthetic-Code +0.0051442917 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0144_of_0150_content_document Synthetic-Code +0.0038872163 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0009_of_0150_content_document Synthetic-Code +0.0051453408 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0146_of_0150_content_document Synthetic-Code +0.0056806584 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0045_of_0150_content_document Synthetic-Code +0.0056803566 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0089_of_0150_content_document Synthetic-Code +0.0056804375 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0094_of_0150_content_document Synthetic-Code +0.0038751647 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0106_of_0150_content_document Synthetic-Code +0.0056804373 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0064_of_0150_content_document Synthetic-Code +0.0038742748 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0121_of_0150_content_document Synthetic-Code +0.0056796959 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0052_of_0150_content_document Synthetic-Code +0.0056797172 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0055_of_0150_content_document Synthetic-Code +0.0056801534 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0060_of_0150_content_document Synthetic-Code +0.0056800629 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0071_of_0150_content_document Synthetic-Code +0.0056810417 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0059_of_0150_content_document Synthetic-Code +0.0038883756 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0024_of_0150_content_document Synthetic-Code +0.0038751090 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0114_of_0150_content_document Synthetic-Code +0.0038862753 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0013_of_0150_content_document Synthetic-Code +0.0038766522 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0112_of_0150_content_document Synthetic-Code +0.0038864154 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0019_of_0150_content_document Synthetic-Code +0.0038784007 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0122_of_0150_content_document Synthetic-Code +0.0038891191 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0029_of_0150_content_document Synthetic-Code +0.0038896526 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0012_of_0150_content_document Synthetic-Code +0.0056811154 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0081_of_0150_content_document Synthetic-Code +0.0038757149 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0127_of_0150_content_document Synthetic-Code +0.0038879644 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0003_of_0150_content_document Synthetic-Code +0.0056810409 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0039_of_0150_content_document Synthetic-Code +0.0056802797 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0073_of_0150_content_document Synthetic-Code +0.0031175146 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0133_of_0150_content_document Synthetic-Code +0.0056801069 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0091_of_0150_content_document Synthetic-Code +0.0038889065 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0026_of_0150_content_document Synthetic-Code +0.0038861094 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0031_of_0150_content_document Synthetic-Code +0.0056808751 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0095_of_0150_content_document Synthetic-Code +0.0031174551 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0140_of_0150_content_document Synthetic-Code +0.0056798206 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0079_of_0150_content_document Synthetic-Code +0.0056806840 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0042_of_0150_content_document Synthetic-Code +0.0031183970 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0138_of_0150_content_document Synthetic-Code +0.0038883052 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0011_of_0150_content_document Synthetic-Code +0.0056801098 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0088_of_0150_content_document Synthetic-Code +0.0056795778 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0067_of_0150_content_document Synthetic-Code +0.0038752524 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0126_of_0150_content_document Synthetic-Code +0.0056808165 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0085_of_0150_content_document Synthetic-Code +0.0038770843 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0125_of_0150_content_document Synthetic-Code +0.0038900937 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0025_of_0150_content_document Synthetic-Code +0.0038886996 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0015_of_0150_content_document Synthetic-Code +0.0038772575 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0120_of_0150_content_document Synthetic-Code +0.0038873478 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0018_of_0150_content_document Synthetic-Code +0.0051452386 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0147_of_0150_content_document Synthetic-Code +0.0056790260 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0087_of_0150_content_document Synthetic-Code +0.0038854542 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0007_of_0150_content_document Synthetic-Code +0.0056800477 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0082_of_0150_content_document Synthetic-Code +0.0031174740 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0130_of_0150_content_document Synthetic-Code +0.0038878863 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0010_of_0150_content_document Synthetic-Code +0.0056799969 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0093_of_0150_content_document Synthetic-Code +0.0056811633 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0077_of_0150_content_document Synthetic-Code +0.0038867735 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0002_of_0150_content_document Synthetic-Code +0.0056814006 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0049_of_0150_content_document Synthetic-Code +0.0056812120 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0072_of_0150_content_document Synthetic-Code +0.0038767087 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0115_of_0150_content_document Synthetic-Code +0.0038912735 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0005_of_0150_content_document Synthetic-Code +0.0056807723 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0048_of_0150_content_document Synthetic-Code +0.0031171573 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0139_of_0150_content_document Synthetic-Code +0.0056796890 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0043_of_0150_content_document Synthetic-Code +0.0051439656 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0142_of_0150_content_document Synthetic-Code +0.0038750602 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0100_of_0150_content_document Synthetic-Code +0.0040740067 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0141_of_0150_content_document Synthetic-Code +0.0056798520 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0056_of_0150_content_document Synthetic-Code +0.0038745755 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0103_of_0150_content_document Synthetic-Code +0.0038881197 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0023_of_0150_content_document Synthetic-Code +0.0038756101 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0098_of_0150_content_document Synthetic-Code +0.0038755106 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0110_of_0150_content_document Synthetic-Code +0.0038884248 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0027_of_0150_content_document Synthetic-Code +0.0038763646 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0107_of_0150_content_document Synthetic-Code +0.0056809407 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0058_of_0150_content_document Synthetic-Code +0.0056811519 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0078_of_0150_content_document Synthetic-Code +0.0056811760 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0069_of_0150_content_document Synthetic-Code +0.0031187430 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0136_of_0150_content_document Synthetic-Code +0.0051432204 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0143_of_0150_content_document Synthetic-Code +0.0051418979 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0149_of_0150_content_document Synthetic-Code +0.0056808752 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0038_of_0150_content_document Synthetic-Code +0.0038757175 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0124_of_0150_content_document Synthetic-Code +0.0056817464 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0090_of_0150_content_document Synthetic-Code +0.0031173594 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0132_of_0150_content_document Synthetic-Code +0.0038888720 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0004_of_0150_content_document Synthetic-Code +0.0038763645 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0116_of_0150_content_document Synthetic-Code +0.0038766972 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0113_of_0150_content_document Synthetic-Code +0.0056793052 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0036_of_0150_content_document Synthetic-Code +0.0045001179 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0128_of_0150_content_document Synthetic-Code +0.0056798868 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0092_of_0150_content_document Synthetic-Code +0.0056808054 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0054_of_0150_content_document Synthetic-Code +0.0056793768 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0044_of_0150_content_document Synthetic-Code +0.0031175993 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0134_of_0150_content_document Synthetic-Code +0.0038751691 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0109_of_0150_content_document Synthetic-Code +0.0056809684 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0068_of_0150_content_document Synthetic-Code +0.0038766209 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0099_of_0150_content_document Synthetic-Code +0.0038881890 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0020_of_0150_content_document Synthetic-Code +0.0056813348 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0065_of_0150_content_document Synthetic-Code +0.0038746094 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0118_of_0150_content_document Synthetic-Code +0.0038768599 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0104_of_0150_content_document Synthetic-Code +0.0038772660 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0108_of_0150_content_document Synthetic-Code +0.0036717869 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0150_of_0150_content_document Synthetic-Code +0.0038880646 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0006_of_0150_content_document Synthetic-Code +0.0056797327 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0047_of_0150_content_document Synthetic-Code +0.0038767864 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0123_of_0150_content_document Synthetic-Code +0.0038768383 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0097_of_0150_content_document Synthetic-Code +0.0056799871 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0084_of_0150_content_document Synthetic-Code +0.0056806478 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0051_of_0150_content_document Synthetic-Code +0.0051445019 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0145_of_0150_content_document Synthetic-Code +0.0056802167 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0075_of_0150_content_document Synthetic-Code +0.0056805148 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0061_of_0150_content_document Synthetic-Code +0.0056802122 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0053_of_0150_content_document Synthetic-Code +0.0056802732 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0083_of_0150_content_document Synthetic-Code +0.0056814370 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0080_of_0150_content_document Synthetic-Code +0.0038761274 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0119_of_0150_content_document Synthetic-Code +0.0038759382 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0111_of_0150_content_document Synthetic-Code +0.0031181519 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0129_of_0150_content_document Synthetic-Code +0.0038721431 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0102_of_0150_content_document Synthetic-Code +0.0056796768 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0070_of_0150_content_document Synthetic-Code +0.0038780303 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0105_of_0150_content_document Synthetic-Code +0.0038889541 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0030_of_0150_content_document Synthetic-Code +0.0038775153 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0101_of_0150_content_document Synthetic-Code +0.0038857488 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0016_of_0150_content_document Synthetic-Code +0.0056799540 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0076_of_0150_content_document Synthetic-Code +0.0056799475 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0035_of_0150_content_document Synthetic-Code +0.0056820564 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0037_of_0150_content_document Synthetic-Code +0.0056815333 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0050_of_0150_content_document Synthetic-Code +0.0056804766 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0063_of_0150_content_document Synthetic-Code +0.0056806236 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0040_of_0150_content_document Synthetic-Code +0.0038897764 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0028_of_0150_content_document Synthetic-Code +0.0056798964 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0062_of_0150_content_document Synthetic-Code +0.0038870109 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0032_of_0150_content_document Synthetic-Code +0.0031185882 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0135_of_0150_content_document Synthetic-Code +0.0056793760 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0033_of_0150_content_document Synthetic-Code +0.0031172349 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0131_of_0150_content_document Synthetic-Code +0.0031169274 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0137_of_0150_content_document Synthetic-Code +0.0038881875 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0001_of_0150_content_document Synthetic-Code +0.0038902657 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0008_of_0150_content_document Synthetic-Code +0.0038778521 /flare/AuroraGPT/datasets/NVIDIA/Nemotron-Pretraining-Code-v2-fused-tok/Synthetic-Code/fused_0117_of_0150_content_document Synthetic-Code From 81ac9c58374d21528879f6cae6e249b9bfa09d87 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 24 Feb 2026 09:45:08 -0600 Subject: [PATCH 588/590] feat: Add `train_aGPT_2B_sophiag_stage2.sh` --- train_aGPT_2B_sophiag_stage2.sh | 41 +++++++++++++++++++++++++++++++++ 1 file changed, 41 insertions(+) create mode 100644 train_aGPT_2B_sophiag_stage2.sh diff --git a/train_aGPT_2B_sophiag_stage2.sh b/train_aGPT_2B_sophiag_stage2.sh new file mode 100644 index 00000000000..b11047932dc --- /dev/null +++ b/train_aGPT_2B_sophiag_stage2.sh @@ -0,0 +1,41 @@ +#!/bin/bash --login +#PBS -q prod +#PBS -A AuroraGPT +#PBS -j oe +#PBS -l walltime=06:00:00,filesystems=flare:home +#PBS -l select=256 + +setup_env() { + cd "${PBS_O_WORKDIR}" || { + echo "Failed to change directory to ${PBS_O_WORKDIR}" + exit 1 + } + # shellcheck disable=SC1090 + source <(curl -L https://bit.ly/ezpz-utils) + ezpz_setup_env + log_message INFO "Using: $(which python3)" +} + +train_model() { + MODEL_ARCH=AuroraGPT-2B \ + TRAIN_TOKENS=7064155541716 \ + OPT=sophiag \ + LR=2.17e-5 \ + GRAD_ACC_STEPS=2 \ + MICRO_BATCH=1 \ + USE_ACTIVATION_CHECKPOINTING=0 \ + ZERO_STAGE=0 \ + LR_DECAY_STYLE=constant \ + TOKENIZER_TYPE=HFTokenizer \ + TOKENIZER_MODEL=google/gemma-7b \ + DATA_FILE_LIST=ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt \ + bash "${PBS_O_WORKDIR}/train_alcf.sh" \ + "$@" +} + +main() { + setup_env + train_model "$@" +} + +main "$@" From b8629a66809c2c468e735d98da92f5afb8cd9fa2 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 24 Feb 2026 09:46:14 -0600 Subject: [PATCH 589/590] chore: Update `train_aGPT_2B_large_batch.sh` --- train_aGPT_2B_large_batch.sh | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-) diff --git a/train_aGPT_2B_large_batch.sh b/train_aGPT_2B_large_batch.sh index bc78bfd9f2c..655f61e222e 100644 --- a/train_aGPT_2B_large_batch.sh +++ b/train_aGPT_2B_large_batch.sh @@ -6,22 +6,6 @@ #PBS -l select=256 -# Optimizer Plans -# -# 1. Ongoing: -# - Lamb -# - 256 Nodes (50M Tok / batch, LR=2e-4) -# - 512 Nodes (100M Tok / batch, LR=2e-4) -# 1. Up-next: -# 256 Nodes @ 50M Tok / batch -# - Sophia: -# - LR=2.17e-5 -# - Muon -# - LR=1.82e-5 -# - Muon-clip -# - LR=2.28e-5 - - setup_env() { cd "${PBS_O_WORKDIR}" || { echo "Failed to change directory to ${PBS_O_WORKDIR}" @@ -35,7 +19,7 @@ setup_env() { train_model() { MODEL_ARCH=AuroraGPT-2B \ - OPT=muonclip \ + OPT=sophiag \ LR=2.28e-5 \ GRAD_ACC_STEPS=2 \ MICRO_BATCH=1 \ From ff45fa8ed9be8d8f331007602a4253443ae54293 Mon Sep 17 00:00:00 2001 From: Sam Foreman Date: Tue, 24 Feb 2026 10:19:57 -0600 Subject: [PATCH 590/590] feat: Add stage2 data list --- .../dolmino-mix-1124-fused-file-list.txt | 324 ++++++++++++++++++ 1 file changed, 324 insertions(+) create mode 100644 ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt diff --git a/ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt b/ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt new file mode 100644 index 00000000000..86c181c9232 --- /dev/null +++ b/ALCF/data-lists/aurora/dolmino-mix-1124-fused-file-list.txt @@ -0,0 +1,324 @@ +0.0015052651 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/wiki/wiki-0001_text_document wiki +0.0031947101 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0032_of_0286_text_document dclm +0.0031550160 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0157_of_0286_text_document dclm +0.0029098600 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/wiki/wiki-0000_text_document wiki +0.0031729901 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0266_of_0286_text_document dclm +0.0031557786 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0213_of_0286_text_document dclm +0.0016219887 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/stackexchange/fused_0001_of_0001_text_document stackexchange +0.0031494994 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0002_of_0286_text_document dclm +0.0031384534 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0252_of_0286_text_document dclm +0.0045391538 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/math/fused_0001_of_0003_text_document math +0.0031296876 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0155_of_0286_text_document dclm +0.0032052303 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0209_of_0286_text_document dclm +0.0082405926 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/math/fused_0002_of_0003_text_document math +0.0031716142 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0164_of_0286_text_document dclm +0.0031444920 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0162_of_0286_text_document dclm +0.0008143753 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/math/fused_0003_of_0003_text_document math +0.0031529612 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0010_of_0286_text_document dclm +0.0031307002 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0125_of_0286_text_document dclm +0.0035483860 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0022_of_0025_text_document pes2o +0.0031687411 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0110_of_0286_text_document dclm +0.0031371333 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0085_of_0286_text_document dclm +0.0012068614 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0001_of_0025_text_document pes2o +0.0031888175 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0254_of_0286_text_document dclm +0.0031190496 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0096_of_0286_text_document dclm +0.0012027663 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0006_of_0025_text_document pes2o +0.0032149646 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0199_of_0286_text_document dclm +0.0032173151 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0103_of_0286_text_document dclm +0.0033666555 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0009_of_0025_text_document pes2o +0.0031355279 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0149_of_0286_text_document dclm +0.0032267198 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0207_of_0286_text_document dclm +0.0033626399 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0014_of_0025_text_document pes2o +0.0031670295 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0047_of_0286_text_document dclm +0.0031653788 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0234_of_0286_text_document dclm +0.0012156084 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0007_of_0025_text_document pes2o +0.0031082385 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0079_of_0286_text_document dclm +0.0027714846 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0102_of_0286_text_document dclm +0.0035431033 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0016_of_0025_text_document pes2o +0.0032059612 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0056_of_0286_text_document dclm +0.0032011452 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0242_of_0286_text_document dclm +0.0033600802 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0018_of_0025_text_document pes2o +0.0032213690 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0226_of_0286_text_document dclm +0.0031784799 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0018_of_0286_text_document dclm +0.0033597072 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0017_of_0025_text_document pes2o +0.0031490267 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0007_of_0286_text_document dclm +0.0031139241 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0036_of_0286_text_document dclm +0.0033764500 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0021_of_0025_text_document pes2o +0.0031707085 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0178_of_0286_text_document dclm +0.0031665816 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0050_of_0286_text_document dclm +0.0033649051 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0015_of_0025_text_document pes2o +0.0031835001 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0161_of_0286_text_document dclm +0.0032264217 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0026_of_0286_text_document dclm +0.0033746150 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0019_of_0025_text_document pes2o +0.0031562112 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0137_of_0286_text_document dclm +0.0032215595 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0003_of_0286_text_document dclm +0.0033586666 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0010_of_0025_text_document pes2o +0.0031278647 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0059_of_0286_text_document dclm +0.0031861383 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0117_of_0286_text_document dclm +0.0033667562 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0023_of_0025_text_document pes2o +0.0031633491 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0211_of_0286_text_document dclm +0.0031985336 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0160_of_0286_text_document dclm +0.0035498167 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0011_of_0025_text_document pes2o +0.0032010168 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0151_of_0286_text_document dclm +0.0031483384 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0163_of_0286_text_document dclm +0.0014734011 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0008_of_0025_text_document pes2o +0.0031805931 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0021_of_0286_text_document dclm +0.0032041236 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0248_of_0286_text_document dclm +0.0033757003 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0024_of_0025_text_document pes2o +0.0031895551 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0086_of_0286_text_document dclm +0.0027796348 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0133_of_0286_text_document dclm +0.0012108242 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0002_of_0025_text_document pes2o +0.0027788350 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0153_of_0286_text_document dclm +0.0031746045 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0170_of_0286_text_document dclm +0.0035471278 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0012_of_0025_text_document pes2o +0.0031470627 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0024_of_0286_text_document dclm +0.0031703677 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0131_of_0286_text_document dclm +0.0012096160 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0004_of_0025_text_document pes2o +0.0027936870 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0201_of_0286_text_document dclm +0.0032065070 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0006_of_0286_text_document dclm +0.0012110298 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0003_of_0025_text_document pes2o +0.0032189979 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0124_of_0286_text_document dclm +0.0031205252 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0088_of_0286_text_document dclm +0.0012101216 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0005_of_0025_text_document pes2o +0.0031902809 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0005_of_0286_text_document dclm +0.0030890769 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0277_of_0286_text_document dclm +0.0039357783 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0025_of_0025_text_document pes2o +0.0030715945 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0063_of_0286_text_document dclm +0.0031737420 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0058_of_0286_text_document dclm +0.0035518882 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0013_of_0025_text_document pes2o +0.0027717404 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0038_of_0286_text_document dclm +0.0031438131 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0171_of_0286_text_document dclm +0.0035414461 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/pes2o/fused_0020_of_0025_text_document pes2o +0.0031684634 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0262_of_0286_text_document dclm +0.0032265582 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0188_of_0286_text_document dclm +0.0006686033 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0007_of_0007_text_document flan +0.0030672209 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0205_of_0286_text_document dclm +0.0031284188 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0141_of_0286_text_document dclm +0.0030282017 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0004_of_0007_text_document flan +0.0031805798 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0270_of_0286_text_document dclm +0.0031827255 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0017_of_0286_text_document dclm +0.0030449227 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0002_of_0007_text_document flan +0.0031340682 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0132_of_0286_text_document dclm +0.0031388612 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0046_of_0286_text_document dclm +0.0037764005 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0005_of_0007_text_document flan +0.0031988945 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0061_of_0286_text_document dclm +0.0031538174 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0167_of_0286_text_document dclm +0.0027763045 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0003_of_0007_text_document flan +0.0031378665 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0075_of_0286_text_document dclm +0.0030807466 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0127_of_0286_text_document dclm +0.0033007839 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0006_of_0007_text_document flan +0.0031965237 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0247_of_0286_text_document dclm +0.0031418999 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0136_of_0286_text_document dclm +0.0031028829 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/flan/fused_0001_of_0007_text_document flan +0.0031666885 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0049_of_0286_text_document dclm +0.0032167048 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0093_of_0286_text_document dclm +0.0028354767 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0067_of_0286_text_document dclm +0.0030981970 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0172_of_0286_text_document dclm +0.0032208863 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0112_of_0286_text_document dclm +0.0031998742 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0081_of_0286_text_document dclm +0.0032197401 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0219_of_0286_text_document dclm +0.0031433616 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0051_of_0286_text_document dclm +0.0032045225 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0118_of_0286_text_document dclm +0.0031755993 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0076_of_0286_text_document dclm +0.0031269355 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0256_of_0286_text_document dclm +0.0031600071 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0115_of_0286_text_document dclm +0.0031659900 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0239_of_0286_text_document dclm +0.0031155813 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0004_of_0286_text_document dclm +0.0031664035 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0030_of_0286_text_document dclm +0.0032035219 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0231_of_0286_text_document dclm +0.0032066816 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0073_of_0286_text_document dclm +0.0031779161 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0235_of_0286_text_document dclm +0.0028517473 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0193_of_0286_text_document dclm +0.0031348925 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0250_of_0286_text_document dclm +0.0031874754 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0173_of_0286_text_document dclm +0.0027849144 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0121_of_0286_text_document dclm +0.0031283684 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0268_of_0286_text_document dclm +0.0027784129 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0221_of_0286_text_document dclm +0.0027866484 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0092_of_0286_text_document dclm +0.0032043660 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0123_of_0286_text_document dclm +0.0027896062 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0203_of_0286_text_document dclm +0.0031824291 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0122_of_0286_text_document dclm +0.0031773425 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0236_of_0286_text_document dclm +0.0031805916 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0246_of_0286_text_document dclm +0.0031991257 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0261_of_0286_text_document dclm +0.0031333196 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0146_of_0286_text_document dclm +0.0031920086 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0020_of_0286_text_document dclm +0.0032207971 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0183_of_0286_text_document dclm +0.0031957595 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0025_of_0286_text_document dclm +0.0028059214 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0053_of_0286_text_document dclm +0.0031544420 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0138_of_0286_text_document dclm +0.0031346562 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0064_of_0286_text_document dclm +0.0031245839 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0040_of_0286_text_document dclm +0.0031792991 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0174_of_0286_text_document dclm +0.0031546504 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0034_of_0286_text_document dclm +0.0031792773 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0177_of_0286_text_document dclm +0.0030766268 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0214_of_0286_text_document dclm +0.0031631241 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0264_of_0286_text_document dclm +0.0031832719 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0107_of_0286_text_document dclm +0.0031436963 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0009_of_0286_text_document dclm +0.0031547059 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0253_of_0286_text_document dclm +0.0030778083 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0212_of_0286_text_document dclm +0.0032180233 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0035_of_0286_text_document dclm +0.0030883104 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0194_of_0286_text_document dclm +0.0027991340 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0033_of_0286_text_document dclm +0.0030850749 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0068_of_0286_text_document dclm +0.0027763803 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0128_of_0286_text_document dclm +0.0032038067 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0225_of_0286_text_document dclm +0.0031695389 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0176_of_0286_text_document dclm +0.0031668089 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0280_of_0286_text_document dclm +0.0031728570 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0228_of_0286_text_document dclm +0.0032221463 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0180_of_0286_text_document dclm +0.0031527493 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0104_of_0286_text_document dclm +0.0031359081 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0098_of_0286_text_document dclm +0.0031887354 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0044_of_0286_text_document dclm +0.0031531199 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0148_of_0286_text_document dclm +0.0032231247 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0216_of_0286_text_document dclm +0.0031634484 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0078_of_0286_text_document dclm +0.0021278969 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0286_of_0286_text_document dclm +0.0032313950 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0054_of_0286_text_document dclm +0.0030389176 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0265_of_0286_text_document dclm +0.0032293310 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0186_of_0286_text_document dclm +0.0031684214 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0140_of_0286_text_document dclm +0.0031982496 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0197_of_0286_text_document dclm +0.0028035962 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0052_of_0286_text_document dclm +0.0027851853 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0057_of_0286_text_document dclm +0.0030869438 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0031_of_0286_text_document dclm +0.0027925782 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0245_of_0286_text_document dclm +0.0031451600 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0143_of_0286_text_document dclm +0.0031331606 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0182_of_0286_text_document dclm +0.0031184014 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0251_of_0286_text_document dclm +0.0030317013 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0269_of_0286_text_document dclm +0.0027760587 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0116_of_0286_text_document dclm +0.0030998495 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0233_of_0286_text_document dclm +0.0032297441 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0045_of_0286_text_document dclm +0.0031679263 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0062_of_0286_text_document dclm +0.0031498640 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0192_of_0286_text_document dclm +0.0030975464 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0114_of_0286_text_document dclm +0.0031156953 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0175_of_0286_text_document dclm +0.0031531682 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0224_of_0286_text_document dclm +0.0031905406 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0144_of_0286_text_document dclm +0.0031558186 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0039_of_0286_text_document dclm +0.0032048414 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0016_of_0286_text_document dclm +0.0031296540 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0267_of_0286_text_document dclm +0.0031244483 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0042_of_0286_text_document dclm +0.0031814069 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0008_of_0286_text_document dclm +0.0031832452 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0099_of_0286_text_document dclm +0.0027639504 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0218_of_0286_text_document dclm +0.0028148554 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0037_of_0286_text_document dclm +0.0031475968 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0208_of_0286_text_document dclm +0.0031457669 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0187_of_0286_text_document dclm +0.0031595927 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0012_of_0286_text_document dclm +0.0032093698 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0185_of_0286_text_document dclm +0.0031415708 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0043_of_0286_text_document dclm +0.0031920289 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0055_of_0286_text_document dclm +0.0027814197 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0065_of_0286_text_document dclm +0.0031088925 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0200_of_0286_text_document dclm +0.0030863506 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0147_of_0286_text_document dclm +0.0032243705 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0255_of_0286_text_document dclm +0.0031587061 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0048_of_0286_text_document dclm +0.0031215730 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0283_of_0286_text_document dclm +0.0031680983 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0113_of_0286_text_document dclm +0.0030982127 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0159_of_0286_text_document dclm +0.0031348350 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0243_of_0286_text_document dclm +0.0031501876 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0158_of_0286_text_document dclm +0.0031510752 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0196_of_0286_text_document dclm +0.0031915438 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0101_of_0286_text_document dclm +0.0027835534 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0139_of_0286_text_document dclm +0.0031138654 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0273_of_0286_text_document dclm +0.0031421083 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0276_of_0286_text_document dclm +0.0032250782 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0156_of_0286_text_document dclm +0.0031456717 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0041_of_0286_text_document dclm +0.0031777396 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0244_of_0286_text_document dclm +0.0032174460 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0080_of_0286_text_document dclm +0.0032111763 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0285_of_0286_text_document dclm +0.0031166419 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0238_of_0286_text_document dclm +0.0030693775 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0074_of_0286_text_document dclm +0.0031426514 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0142_of_0286_text_document dclm +0.0031817773 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0229_of_0286_text_document dclm +0.0032073815 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0109_of_0286_text_document dclm +0.0032215419 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0202_of_0286_text_document dclm +0.0031582602 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0189_of_0286_text_document dclm +0.0027776870 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0204_of_0286_text_document dclm +0.0032060557 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0095_of_0286_text_document dclm +0.0031725788 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0168_of_0286_text_document dclm +0.0032214427 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0154_of_0286_text_document dclm +0.0031215996 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0011_of_0286_text_document dclm +0.0031405866 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0100_of_0286_text_document dclm +0.0027680490 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0257_of_0286_text_document dclm +0.0032077676 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0232_of_0286_text_document dclm +0.0031674161 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0108_of_0286_text_document dclm +0.0032152229 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0130_of_0286_text_document dclm +0.0031782480 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0119_of_0286_text_document dclm +0.0031424302 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0230_of_0286_text_document dclm +0.0031558210 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0022_of_0286_text_document dclm +0.0032239387 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0066_of_0286_text_document dclm +0.0031894807 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0166_of_0286_text_document dclm +0.0031265981 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0217_of_0286_text_document dclm +0.0027970662 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0191_of_0286_text_document dclm +0.0031562327 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0249_of_0286_text_document dclm +0.0031096728 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0077_of_0286_text_document dclm +0.0031055863 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0082_of_0286_text_document dclm +0.0028031178 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0120_of_0286_text_document dclm +0.0031768558 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0090_of_0286_text_document dclm +0.0027699660 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0184_of_0286_text_document dclm +0.0031072803 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0091_of_0286_text_document dclm +0.0031908168 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0220_of_0286_text_document dclm +0.0032086453 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0134_of_0286_text_document dclm +0.0031220883 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0070_of_0286_text_document dclm +0.0031385199 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0259_of_0286_text_document dclm +0.0032271458 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0135_of_0286_text_document dclm +0.0030638732 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0150_of_0286_text_document dclm +0.0030769679 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0260_of_0286_text_document dclm +0.0032303114 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0014_of_0286_text_document dclm +0.0031762445 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0072_of_0286_text_document dclm +0.0031839177 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0278_of_0286_text_document dclm +0.0031614587 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0027_of_0286_text_document dclm +0.0032287822 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0179_of_0286_text_document dclm +0.0031706969 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0169_of_0286_text_document dclm +0.0031927919 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0241_of_0286_text_document dclm +0.0027680498 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0105_of_0286_text_document dclm +0.0032086731 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0106_of_0286_text_document dclm +0.0031408533 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0129_of_0286_text_document dclm +0.0031964728 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0023_of_0286_text_document dclm +0.0031728416 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0019_of_0286_text_document dclm +0.0031440213 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0206_of_0286_text_document dclm +0.0030863232 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0271_of_0286_text_document dclm +0.0031409528 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0029_of_0286_text_document dclm +0.0031407346 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0028_of_0286_text_document dclm +0.0031271659 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0190_of_0286_text_document dclm +0.0027747169 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0111_of_0286_text_document dclm +0.0032174905 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0279_of_0286_text_document dclm +0.0031577664 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0084_of_0286_text_document dclm +0.0031890982 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0165_of_0286_text_document dclm +0.0032084469 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0275_of_0286_text_document dclm +0.0031989553 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0215_of_0286_text_document dclm +0.0031470140 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0069_of_0286_text_document dclm +0.0031501623 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0284_of_0286_text_document dclm +0.0031378629 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0258_of_0286_text_document dclm +0.0031844465 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0126_of_0286_text_document dclm +0.0030853355 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0181_of_0286_text_document dclm +0.0031349147 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0152_of_0286_text_document dclm +0.0031207958 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0083_of_0286_text_document dclm +0.0031629135 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0274_of_0286_text_document dclm +0.0031174416 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0001_of_0286_text_document dclm +0.0028260097 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0237_of_0286_text_document dclm +0.0031286683 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0272_of_0286_text_document dclm +0.0031256223 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0282_of_0286_text_document dclm +0.0031479668 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0087_of_0286_text_document dclm +0.0030992436 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0223_of_0286_text_document dclm +0.0032101406 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0071_of_0286_text_document dclm +0.0032023398 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0263_of_0286_text_document dclm +0.0031790937 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0198_of_0286_text_document dclm +0.0031543190 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0094_of_0286_text_document dclm +0.0031423711 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0097_of_0286_text_document dclm +0.0030922585 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0195_of_0286_text_document dclm +0.0031964924 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0227_of_0286_text_document dclm +0.0031985342 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0015_of_0286_text_document dclm +0.0031781367 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0013_of_0286_text_document dclm +0.0031596116 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0240_of_0286_text_document dclm +0.0031463073 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0281_of_0286_text_document dclm +0.0030829725 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0060_of_0286_text_document dclm +0.0031550240 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0089_of_0286_text_document dclm +0.0030955622 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0222_of_0286_text_document dclm +0.0030790804 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0210_of_0286_text_document dclm +0.0031393625 /flare/AuroraGPT/datasets/dolmino-mix-1124/data-fused-tok/dclm/fused_0145_of_0286_text_document dclm